model_4kai894c / checkpoint-2622 /trainer_state.json
ugaoo's picture
Upload folder using huggingface_hub
63e9359 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.987052551408987,
"eval_steps": 500,
"global_step": 2622,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002284843869002285,
"grad_norm": 3.668196201324463,
"learning_rate": 5.0000000000000004e-08,
"loss": 1.5687,
"step": 1
},
{
"epoch": 0.00456968773800457,
"grad_norm": 3.6277146339416504,
"learning_rate": 1.0000000000000001e-07,
"loss": 1.5714,
"step": 2
},
{
"epoch": 0.006854531607006854,
"grad_norm": 3.813422918319702,
"learning_rate": 1.5000000000000002e-07,
"loss": 1.58,
"step": 3
},
{
"epoch": 0.00913937547600914,
"grad_norm": 3.4566409587860107,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.5604,
"step": 4
},
{
"epoch": 0.011424219345011425,
"grad_norm": 3.287661552429199,
"learning_rate": 2.5000000000000004e-07,
"loss": 1.5425,
"step": 5
},
{
"epoch": 0.013709063214013708,
"grad_norm": 3.318340301513672,
"learning_rate": 3.0000000000000004e-07,
"loss": 1.5477,
"step": 6
},
{
"epoch": 0.015993907083015995,
"grad_norm": 3.407221555709839,
"learning_rate": 3.5000000000000004e-07,
"loss": 1.5848,
"step": 7
},
{
"epoch": 0.01827875095201828,
"grad_norm": 3.732999563217163,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.5884,
"step": 8
},
{
"epoch": 0.020563594821020565,
"grad_norm": 3.532766580581665,
"learning_rate": 4.5000000000000003e-07,
"loss": 1.5892,
"step": 9
},
{
"epoch": 0.02284843869002285,
"grad_norm": 3.5676348209381104,
"learning_rate": 5.000000000000001e-07,
"loss": 1.5619,
"step": 10
},
{
"epoch": 0.02513328255902513,
"grad_norm": 3.1015849113464355,
"learning_rate": 5.5e-07,
"loss": 1.5649,
"step": 11
},
{
"epoch": 0.027418126428027417,
"grad_norm": 3.163240909576416,
"learning_rate": 6.000000000000001e-07,
"loss": 1.5807,
"step": 12
},
{
"epoch": 0.0297029702970297,
"grad_norm": 2.894922971725464,
"learning_rate": 6.5e-07,
"loss": 1.5454,
"step": 13
},
{
"epoch": 0.03198781416603199,
"grad_norm": 2.8211843967437744,
"learning_rate": 7.000000000000001e-07,
"loss": 1.5801,
"step": 14
},
{
"epoch": 0.03427265803503427,
"grad_norm": 2.676609516143799,
"learning_rate": 7.5e-07,
"loss": 1.5446,
"step": 15
},
{
"epoch": 0.03655750190403656,
"grad_norm": 2.6186320781707764,
"learning_rate": 8.000000000000001e-07,
"loss": 1.5443,
"step": 16
},
{
"epoch": 0.03884234577303884,
"grad_norm": 2.460139513015747,
"learning_rate": 8.500000000000001e-07,
"loss": 1.5489,
"step": 17
},
{
"epoch": 0.04112718964204113,
"grad_norm": 2.368126630783081,
"learning_rate": 9.000000000000001e-07,
"loss": 1.5317,
"step": 18
},
{
"epoch": 0.04341203351104341,
"grad_norm": 2.244192123413086,
"learning_rate": 9.500000000000001e-07,
"loss": 1.4805,
"step": 19
},
{
"epoch": 0.0456968773800457,
"grad_norm": 2.242701292037964,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.5478,
"step": 20
},
{
"epoch": 0.04798172124904798,
"grad_norm": 2.13895583152771,
"learning_rate": 1.0500000000000001e-06,
"loss": 1.5194,
"step": 21
},
{
"epoch": 0.05026656511805026,
"grad_norm": 2.0152103900909424,
"learning_rate": 1.1e-06,
"loss": 1.5067,
"step": 22
},
{
"epoch": 0.05255140898705255,
"grad_norm": 1.9156895875930786,
"learning_rate": 1.1500000000000002e-06,
"loss": 1.5145,
"step": 23
},
{
"epoch": 0.05483625285605483,
"grad_norm": 1.7710504531860352,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.5147,
"step": 24
},
{
"epoch": 0.05712109672505712,
"grad_norm": 1.807431936264038,
"learning_rate": 1.25e-06,
"loss": 1.5357,
"step": 25
},
{
"epoch": 0.0594059405940594,
"grad_norm": 1.6638832092285156,
"learning_rate": 1.3e-06,
"loss": 1.489,
"step": 26
},
{
"epoch": 0.06169078446306169,
"grad_norm": 1.5708481073379517,
"learning_rate": 1.3500000000000002e-06,
"loss": 1.4768,
"step": 27
},
{
"epoch": 0.06397562833206398,
"grad_norm": 1.615577220916748,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.5159,
"step": 28
},
{
"epoch": 0.06626047220106626,
"grad_norm": 1.5125129222869873,
"learning_rate": 1.45e-06,
"loss": 1.4972,
"step": 29
},
{
"epoch": 0.06854531607006854,
"grad_norm": 1.479811668395996,
"learning_rate": 1.5e-06,
"loss": 1.4674,
"step": 30
},
{
"epoch": 0.07083015993907082,
"grad_norm": 1.4502017498016357,
"learning_rate": 1.5500000000000002e-06,
"loss": 1.4811,
"step": 31
},
{
"epoch": 0.07311500380807312,
"grad_norm": 1.3617135286331177,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.4872,
"step": 32
},
{
"epoch": 0.0753998476770754,
"grad_norm": 1.367607831954956,
"learning_rate": 1.6500000000000003e-06,
"loss": 1.4699,
"step": 33
},
{
"epoch": 0.07768469154607768,
"grad_norm": 1.3374927043914795,
"learning_rate": 1.7000000000000002e-06,
"loss": 1.4659,
"step": 34
},
{
"epoch": 0.07996953541507996,
"grad_norm": 1.354506254196167,
"learning_rate": 1.75e-06,
"loss": 1.4351,
"step": 35
},
{
"epoch": 0.08225437928408226,
"grad_norm": 1.2532024383544922,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.4358,
"step": 36
},
{
"epoch": 0.08453922315308454,
"grad_norm": 1.2684043645858765,
"learning_rate": 1.85e-06,
"loss": 1.4534,
"step": 37
},
{
"epoch": 0.08682406702208682,
"grad_norm": 1.2418140172958374,
"learning_rate": 1.9000000000000002e-06,
"loss": 1.4624,
"step": 38
},
{
"epoch": 0.0891089108910891,
"grad_norm": 1.2266045808792114,
"learning_rate": 1.9500000000000004e-06,
"loss": 1.4282,
"step": 39
},
{
"epoch": 0.0913937547600914,
"grad_norm": 1.180330753326416,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.4107,
"step": 40
},
{
"epoch": 0.09367859862909368,
"grad_norm": 1.1651424169540405,
"learning_rate": 2.05e-06,
"loss": 1.4041,
"step": 41
},
{
"epoch": 0.09596344249809596,
"grad_norm": 1.181652307510376,
"learning_rate": 2.1000000000000002e-06,
"loss": 1.4558,
"step": 42
},
{
"epoch": 0.09824828636709824,
"grad_norm": 1.2221183776855469,
"learning_rate": 2.15e-06,
"loss": 1.4449,
"step": 43
},
{
"epoch": 0.10053313023610053,
"grad_norm": 1.085172414779663,
"learning_rate": 2.2e-06,
"loss": 1.4235,
"step": 44
},
{
"epoch": 0.10281797410510282,
"grad_norm": 1.0497649908065796,
"learning_rate": 2.25e-06,
"loss": 1.3891,
"step": 45
},
{
"epoch": 0.1051028179741051,
"grad_norm": 1.0502350330352783,
"learning_rate": 2.3000000000000004e-06,
"loss": 1.4048,
"step": 46
},
{
"epoch": 0.10738766184310738,
"grad_norm": 1.0798920392990112,
"learning_rate": 2.35e-06,
"loss": 1.4383,
"step": 47
},
{
"epoch": 0.10967250571210967,
"grad_norm": 1.067581057548523,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.4128,
"step": 48
},
{
"epoch": 0.11195734958111196,
"grad_norm": 1.062606930732727,
"learning_rate": 2.4500000000000003e-06,
"loss": 1.4438,
"step": 49
},
{
"epoch": 0.11424219345011424,
"grad_norm": 1.0157577991485596,
"learning_rate": 2.5e-06,
"loss": 1.4257,
"step": 50
},
{
"epoch": 0.11652703731911652,
"grad_norm": 1.0165379047393799,
"learning_rate": 2.55e-06,
"loss": 1.407,
"step": 51
},
{
"epoch": 0.1188118811881188,
"grad_norm": 1.0268282890319824,
"learning_rate": 2.6e-06,
"loss": 1.3942,
"step": 52
},
{
"epoch": 0.1210967250571211,
"grad_norm": 1.0133647918701172,
"learning_rate": 2.6500000000000005e-06,
"loss": 1.3737,
"step": 53
},
{
"epoch": 0.12338156892612338,
"grad_norm": 1.0097134113311768,
"learning_rate": 2.7000000000000004e-06,
"loss": 1.3994,
"step": 54
},
{
"epoch": 0.12566641279512566,
"grad_norm": 1.1268850564956665,
"learning_rate": 2.7500000000000004e-06,
"loss": 1.3676,
"step": 55
},
{
"epoch": 0.12795125666412796,
"grad_norm": 0.981015682220459,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.3819,
"step": 56
},
{
"epoch": 0.13023610053313023,
"grad_norm": 1.0456632375717163,
"learning_rate": 2.85e-06,
"loss": 1.4031,
"step": 57
},
{
"epoch": 0.13252094440213252,
"grad_norm": 1.0366231203079224,
"learning_rate": 2.9e-06,
"loss": 1.4017,
"step": 58
},
{
"epoch": 0.13480578827113482,
"grad_norm": 0.9980257749557495,
"learning_rate": 2.95e-06,
"loss": 1.4261,
"step": 59
},
{
"epoch": 0.1370906321401371,
"grad_norm": 0.990281879901886,
"learning_rate": 3e-06,
"loss": 1.3699,
"step": 60
},
{
"epoch": 0.13937547600913938,
"grad_norm": 1.0530250072479248,
"learning_rate": 3.05e-06,
"loss": 1.3656,
"step": 61
},
{
"epoch": 0.14166031987814165,
"grad_norm": 0.9878147840499878,
"learning_rate": 3.1000000000000004e-06,
"loss": 1.3712,
"step": 62
},
{
"epoch": 0.14394516374714394,
"grad_norm": 0.9554497599601746,
"learning_rate": 3.1500000000000003e-06,
"loss": 1.3507,
"step": 63
},
{
"epoch": 0.14623000761614624,
"grad_norm": 1.0152994394302368,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.3531,
"step": 64
},
{
"epoch": 0.1485148514851485,
"grad_norm": 0.9816209077835083,
"learning_rate": 3.2500000000000002e-06,
"loss": 1.3733,
"step": 65
},
{
"epoch": 0.1507996953541508,
"grad_norm": 1.014113187789917,
"learning_rate": 3.3000000000000006e-06,
"loss": 1.3798,
"step": 66
},
{
"epoch": 0.15308453922315307,
"grad_norm": 1.005303978919983,
"learning_rate": 3.3500000000000005e-06,
"loss": 1.3877,
"step": 67
},
{
"epoch": 0.15536938309215537,
"grad_norm": 1.109976887702942,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.4184,
"step": 68
},
{
"epoch": 0.15765422696115766,
"grad_norm": 1.033060908317566,
"learning_rate": 3.45e-06,
"loss": 1.4043,
"step": 69
},
{
"epoch": 0.15993907083015993,
"grad_norm": 0.9719234108924866,
"learning_rate": 3.5e-06,
"loss": 1.3481,
"step": 70
},
{
"epoch": 0.16222391469916222,
"grad_norm": 1.0430618524551392,
"learning_rate": 3.5500000000000003e-06,
"loss": 1.3227,
"step": 71
},
{
"epoch": 0.16450875856816452,
"grad_norm": 1.0481953620910645,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.3174,
"step": 72
},
{
"epoch": 0.1667936024371668,
"grad_norm": 0.9868738055229187,
"learning_rate": 3.65e-06,
"loss": 1.356,
"step": 73
},
{
"epoch": 0.16907844630616908,
"grad_norm": 1.0015943050384521,
"learning_rate": 3.7e-06,
"loss": 1.3462,
"step": 74
},
{
"epoch": 0.17136329017517135,
"grad_norm": 1.0458308458328247,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.3962,
"step": 75
},
{
"epoch": 0.17364813404417365,
"grad_norm": 1.0376830101013184,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.3523,
"step": 76
},
{
"epoch": 0.17593297791317594,
"grad_norm": 0.9821555018424988,
"learning_rate": 3.85e-06,
"loss": 1.3559,
"step": 77
},
{
"epoch": 0.1782178217821782,
"grad_norm": 0.9579638838768005,
"learning_rate": 3.900000000000001e-06,
"loss": 1.3073,
"step": 78
},
{
"epoch": 0.1805026656511805,
"grad_norm": 0.9736194014549255,
"learning_rate": 3.95e-06,
"loss": 1.3494,
"step": 79
},
{
"epoch": 0.1827875095201828,
"grad_norm": 1.0055922269821167,
"learning_rate": 4.000000000000001e-06,
"loss": 1.3697,
"step": 80
},
{
"epoch": 0.18507235338918507,
"grad_norm": 0.9767876267433167,
"learning_rate": 4.05e-06,
"loss": 1.3225,
"step": 81
},
{
"epoch": 0.18735719725818736,
"grad_norm": 1.003092885017395,
"learning_rate": 4.1e-06,
"loss": 1.335,
"step": 82
},
{
"epoch": 0.18964204112718963,
"grad_norm": 0.9898741245269775,
"learning_rate": 4.15e-06,
"loss": 1.3103,
"step": 83
},
{
"epoch": 0.19192688499619193,
"grad_norm": 0.9903189539909363,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.3741,
"step": 84
},
{
"epoch": 0.19421172886519422,
"grad_norm": 0.9661535620689392,
"learning_rate": 4.25e-06,
"loss": 1.3381,
"step": 85
},
{
"epoch": 0.1964965727341965,
"grad_norm": 0.9668599367141724,
"learning_rate": 4.3e-06,
"loss": 1.3511,
"step": 86
},
{
"epoch": 0.19878141660319879,
"grad_norm": 0.9633579254150391,
"learning_rate": 4.350000000000001e-06,
"loss": 1.3841,
"step": 87
},
{
"epoch": 0.20106626047220105,
"grad_norm": 0.9665766358375549,
"learning_rate": 4.4e-06,
"loss": 1.3211,
"step": 88
},
{
"epoch": 0.20335110434120335,
"grad_norm": 1.0263577699661255,
"learning_rate": 4.450000000000001e-06,
"loss": 1.3398,
"step": 89
},
{
"epoch": 0.20563594821020564,
"grad_norm": 1.0054337978363037,
"learning_rate": 4.5e-06,
"loss": 1.3598,
"step": 90
},
{
"epoch": 0.2079207920792079,
"grad_norm": 0.9768564701080322,
"learning_rate": 4.5500000000000005e-06,
"loss": 1.3386,
"step": 91
},
{
"epoch": 0.2102056359482102,
"grad_norm": 0.9710814356803894,
"learning_rate": 4.600000000000001e-06,
"loss": 1.306,
"step": 92
},
{
"epoch": 0.2124904798172125,
"grad_norm": 0.9943618774414062,
"learning_rate": 4.65e-06,
"loss": 1.3368,
"step": 93
},
{
"epoch": 0.21477532368621477,
"grad_norm": 1.0000272989273071,
"learning_rate": 4.7e-06,
"loss": 1.3561,
"step": 94
},
{
"epoch": 0.21706016755521707,
"grad_norm": 0.9748716950416565,
"learning_rate": 4.75e-06,
"loss": 1.3216,
"step": 95
},
{
"epoch": 0.21934501142421933,
"grad_norm": 0.977959930896759,
"learning_rate": 4.800000000000001e-06,
"loss": 1.3275,
"step": 96
},
{
"epoch": 0.22162985529322163,
"grad_norm": 0.9991240501403809,
"learning_rate": 4.85e-06,
"loss": 1.3143,
"step": 97
},
{
"epoch": 0.22391469916222392,
"grad_norm": 1.0590916872024536,
"learning_rate": 4.9000000000000005e-06,
"loss": 1.3467,
"step": 98
},
{
"epoch": 0.2261995430312262,
"grad_norm": 0.9592604041099548,
"learning_rate": 4.95e-06,
"loss": 1.3568,
"step": 99
},
{
"epoch": 0.2284843869002285,
"grad_norm": 0.9900586605072021,
"learning_rate": 5e-06,
"loss": 1.3162,
"step": 100
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.9882398843765259,
"learning_rate": 4.999998060367119e-06,
"loss": 1.3348,
"step": 101
},
{
"epoch": 0.23305407463823305,
"grad_norm": 0.9522809982299805,
"learning_rate": 4.999992241471486e-06,
"loss": 1.3004,
"step": 102
},
{
"epoch": 0.23533891850723535,
"grad_norm": 0.9822378754615784,
"learning_rate": 4.9999825433221295e-06,
"loss": 1.3326,
"step": 103
},
{
"epoch": 0.2376237623762376,
"grad_norm": 0.9944847822189331,
"learning_rate": 4.999968965934098e-06,
"loss": 1.3429,
"step": 104
},
{
"epoch": 0.2399086062452399,
"grad_norm": 1.052456021308899,
"learning_rate": 4.9999515093284605e-06,
"loss": 1.3476,
"step": 105
},
{
"epoch": 0.2421934501142422,
"grad_norm": 0.9862610697746277,
"learning_rate": 4.999930173532304e-06,
"loss": 1.3638,
"step": 106
},
{
"epoch": 0.24447829398324447,
"grad_norm": 0.9718945622444153,
"learning_rate": 4.999904958578735e-06,
"loss": 1.3013,
"step": 107
},
{
"epoch": 0.24676313785224677,
"grad_norm": 0.9535952210426331,
"learning_rate": 4.9998758645068805e-06,
"loss": 1.3317,
"step": 108
},
{
"epoch": 0.24904798172124903,
"grad_norm": 1.1905543804168701,
"learning_rate": 4.999842891361885e-06,
"loss": 1.3325,
"step": 109
},
{
"epoch": 0.25133282559025133,
"grad_norm": 1.0306485891342163,
"learning_rate": 4.9998060391949145e-06,
"loss": 1.3198,
"step": 110
},
{
"epoch": 0.2536176694592536,
"grad_norm": 1.0334984064102173,
"learning_rate": 4.999765308063152e-06,
"loss": 1.3075,
"step": 111
},
{
"epoch": 0.2559025133282559,
"grad_norm": 1.0020740032196045,
"learning_rate": 4.9997206980298e-06,
"loss": 1.3324,
"step": 112
},
{
"epoch": 0.25818735719725816,
"grad_norm": 0.9771923422813416,
"learning_rate": 4.9996722091640805e-06,
"loss": 1.3072,
"step": 113
},
{
"epoch": 0.26047220106626046,
"grad_norm": 0.9955299496650696,
"learning_rate": 4.999619841541234e-06,
"loss": 1.3501,
"step": 114
},
{
"epoch": 0.26275704493526275,
"grad_norm": 1.0125700235366821,
"learning_rate": 4.9995635952425205e-06,
"loss": 1.3387,
"step": 115
},
{
"epoch": 0.26504188880426505,
"grad_norm": 1.005936622619629,
"learning_rate": 4.999503470355215e-06,
"loss": 1.342,
"step": 116
},
{
"epoch": 0.26732673267326734,
"grad_norm": 0.9978262782096863,
"learning_rate": 4.999439466972616e-06,
"loss": 1.2954,
"step": 117
},
{
"epoch": 0.26961157654226964,
"grad_norm": 0.9668537974357605,
"learning_rate": 4.999371585194039e-06,
"loss": 1.3318,
"step": 118
},
{
"epoch": 0.2718964204112719,
"grad_norm": 1.0156077146530151,
"learning_rate": 4.999299825124814e-06,
"loss": 1.2681,
"step": 119
},
{
"epoch": 0.2741812642802742,
"grad_norm": 0.99967360496521,
"learning_rate": 4.999224186876293e-06,
"loss": 1.2666,
"step": 120
},
{
"epoch": 0.27646610814927647,
"grad_norm": 1.0085562467575073,
"learning_rate": 4.999144670565842e-06,
"loss": 1.3261,
"step": 121
},
{
"epoch": 0.27875095201827876,
"grad_norm": 1.0338691473007202,
"learning_rate": 4.999061276316851e-06,
"loss": 1.2943,
"step": 122
},
{
"epoch": 0.28103579588728106,
"grad_norm": 0.9880859851837158,
"learning_rate": 4.99897400425872e-06,
"loss": 1.3035,
"step": 123
},
{
"epoch": 0.2833206397562833,
"grad_norm": 0.9832742810249329,
"learning_rate": 4.998882854526872e-06,
"loss": 1.3015,
"step": 124
},
{
"epoch": 0.2856054836252856,
"grad_norm": 0.976040780544281,
"learning_rate": 4.998787827262743e-06,
"loss": 1.3325,
"step": 125
},
{
"epoch": 0.2878903274942879,
"grad_norm": 1.0309007167816162,
"learning_rate": 4.998688922613788e-06,
"loss": 1.2998,
"step": 126
},
{
"epoch": 0.2901751713632902,
"grad_norm": 1.0828396081924438,
"learning_rate": 4.998586140733477e-06,
"loss": 1.3093,
"step": 127
},
{
"epoch": 0.2924600152322925,
"grad_norm": 0.9725452661514282,
"learning_rate": 4.998479481781299e-06,
"loss": 1.2811,
"step": 128
},
{
"epoch": 0.2947448591012947,
"grad_norm": 0.9891279339790344,
"learning_rate": 4.998368945922757e-06,
"loss": 1.3104,
"step": 129
},
{
"epoch": 0.297029702970297,
"grad_norm": 1.022490382194519,
"learning_rate": 4.998254533329369e-06,
"loss": 1.3425,
"step": 130
},
{
"epoch": 0.2993145468392993,
"grad_norm": 1.00505530834198,
"learning_rate": 4.99813624417867e-06,
"loss": 1.3494,
"step": 131
},
{
"epoch": 0.3015993907083016,
"grad_norm": 1.033308982849121,
"learning_rate": 4.998014078654211e-06,
"loss": 1.278,
"step": 132
},
{
"epoch": 0.3038842345773039,
"grad_norm": 1.0194460153579712,
"learning_rate": 4.997888036945556e-06,
"loss": 1.2963,
"step": 133
},
{
"epoch": 0.30616907844630614,
"grad_norm": 1.005299687385559,
"learning_rate": 4.997758119248286e-06,
"loss": 1.3187,
"step": 134
},
{
"epoch": 0.30845392231530844,
"grad_norm": 1.0271679162979126,
"learning_rate": 4.997624325763994e-06,
"loss": 1.3106,
"step": 135
},
{
"epoch": 0.31073876618431073,
"grad_norm": 1.0343165397644043,
"learning_rate": 4.997486656700289e-06,
"loss": 1.3355,
"step": 136
},
{
"epoch": 0.31302361005331303,
"grad_norm": 1.0498188734054565,
"learning_rate": 4.997345112270792e-06,
"loss": 1.3126,
"step": 137
},
{
"epoch": 0.3153084539223153,
"grad_norm": 0.9742498993873596,
"learning_rate": 4.997199692695138e-06,
"loss": 1.3006,
"step": 138
},
{
"epoch": 0.3175932977913176,
"grad_norm": 1.0044124126434326,
"learning_rate": 4.997050398198977e-06,
"loss": 1.3298,
"step": 139
},
{
"epoch": 0.31987814166031986,
"grad_norm": 1.0173184871673584,
"learning_rate": 4.99689722901397e-06,
"loss": 1.3286,
"step": 140
},
{
"epoch": 0.32216298552932215,
"grad_norm": 0.9835124611854553,
"learning_rate": 4.99674018537779e-06,
"loss": 1.2937,
"step": 141
},
{
"epoch": 0.32444782939832445,
"grad_norm": 1.0389831066131592,
"learning_rate": 4.996579267534122e-06,
"loss": 1.3077,
"step": 142
},
{
"epoch": 0.32673267326732675,
"grad_norm": 1.0412015914916992,
"learning_rate": 4.996414475732664e-06,
"loss": 1.3131,
"step": 143
},
{
"epoch": 0.32901751713632904,
"grad_norm": 1.0527534484863281,
"learning_rate": 4.9962458102291254e-06,
"loss": 1.3075,
"step": 144
},
{
"epoch": 0.3313023610053313,
"grad_norm": 1.036034345626831,
"learning_rate": 4.9960732712852236e-06,
"loss": 1.3198,
"step": 145
},
{
"epoch": 0.3335872048743336,
"grad_norm": 1.0121785402297974,
"learning_rate": 4.99589685916869e-06,
"loss": 1.3346,
"step": 146
},
{
"epoch": 0.33587204874333587,
"grad_norm": 1.0597130060195923,
"learning_rate": 4.9957165741532635e-06,
"loss": 1.3025,
"step": 147
},
{
"epoch": 0.33815689261233817,
"grad_norm": 1.0982815027236938,
"learning_rate": 4.995532416518693e-06,
"loss": 1.3177,
"step": 148
},
{
"epoch": 0.34044173648134046,
"grad_norm": 1.012061357498169,
"learning_rate": 4.995344386550738e-06,
"loss": 1.2905,
"step": 149
},
{
"epoch": 0.3427265803503427,
"grad_norm": 1.0748074054718018,
"learning_rate": 4.995152484541166e-06,
"loss": 1.3191,
"step": 150
},
{
"epoch": 0.345011424219345,
"grad_norm": 1.0346341133117676,
"learning_rate": 4.994956710787752e-06,
"loss": 1.2923,
"step": 151
},
{
"epoch": 0.3472962680883473,
"grad_norm": 1.0333645343780518,
"learning_rate": 4.99475706559428e-06,
"loss": 1.3272,
"step": 152
},
{
"epoch": 0.3495811119573496,
"grad_norm": 1.0411094427108765,
"learning_rate": 4.9945535492705385e-06,
"loss": 1.3102,
"step": 153
},
{
"epoch": 0.3518659558263519,
"grad_norm": 1.0394591093063354,
"learning_rate": 4.994346162132329e-06,
"loss": 1.2912,
"step": 154
},
{
"epoch": 0.3541507996953541,
"grad_norm": 1.1258337497711182,
"learning_rate": 4.994134904501452e-06,
"loss": 1.295,
"step": 155
},
{
"epoch": 0.3564356435643564,
"grad_norm": 1.0196075439453125,
"learning_rate": 4.993919776705718e-06,
"loss": 1.2935,
"step": 156
},
{
"epoch": 0.3587204874333587,
"grad_norm": 1.020180583000183,
"learning_rate": 4.993700779078943e-06,
"loss": 1.3118,
"step": 157
},
{
"epoch": 0.361005331302361,
"grad_norm": 1.1170531511306763,
"learning_rate": 4.993477911960948e-06,
"loss": 1.2924,
"step": 158
},
{
"epoch": 0.3632901751713633,
"grad_norm": 1.0637717247009277,
"learning_rate": 4.993251175697554e-06,
"loss": 1.2797,
"step": 159
},
{
"epoch": 0.3655750190403656,
"grad_norm": 1.046305775642395,
"learning_rate": 4.993020570640592e-06,
"loss": 1.3142,
"step": 160
},
{
"epoch": 0.36785986290936784,
"grad_norm": 1.039476752281189,
"learning_rate": 4.992786097147892e-06,
"loss": 1.2773,
"step": 161
},
{
"epoch": 0.37014470677837014,
"grad_norm": 1.0379183292388916,
"learning_rate": 4.992547755583288e-06,
"loss": 1.3057,
"step": 162
},
{
"epoch": 0.37242955064737243,
"grad_norm": 1.0063403844833374,
"learning_rate": 4.992305546316617e-06,
"loss": 1.3108,
"step": 163
},
{
"epoch": 0.3747143945163747,
"grad_norm": 1.0467029809951782,
"learning_rate": 4.992059469723716e-06,
"loss": 1.2675,
"step": 164
},
{
"epoch": 0.376999238385377,
"grad_norm": 0.9822115898132324,
"learning_rate": 4.991809526186424e-06,
"loss": 1.2987,
"step": 165
},
{
"epoch": 0.37928408225437926,
"grad_norm": 0.9957991242408752,
"learning_rate": 4.9915557160925795e-06,
"loss": 1.2927,
"step": 166
},
{
"epoch": 0.38156892612338156,
"grad_norm": 1.020486831665039,
"learning_rate": 4.991298039836021e-06,
"loss": 1.2891,
"step": 167
},
{
"epoch": 0.38385376999238385,
"grad_norm": 0.9941042065620422,
"learning_rate": 4.991036497816587e-06,
"loss": 1.3279,
"step": 168
},
{
"epoch": 0.38613861386138615,
"grad_norm": 1.030573844909668,
"learning_rate": 4.990771090440114e-06,
"loss": 1.2715,
"step": 169
},
{
"epoch": 0.38842345773038844,
"grad_norm": 0.9810742735862732,
"learning_rate": 4.990501818118436e-06,
"loss": 1.2808,
"step": 170
},
{
"epoch": 0.3907083015993907,
"grad_norm": 1.0300201177597046,
"learning_rate": 4.990228681269383e-06,
"loss": 1.3079,
"step": 171
},
{
"epoch": 0.392993145468393,
"grad_norm": 1.0107353925704956,
"learning_rate": 4.989951680316787e-06,
"loss": 1.2872,
"step": 172
},
{
"epoch": 0.3952779893373953,
"grad_norm": 1.0361515283584595,
"learning_rate": 4.989670815690469e-06,
"loss": 1.2784,
"step": 173
},
{
"epoch": 0.39756283320639757,
"grad_norm": 1.0452970266342163,
"learning_rate": 4.989386087826248e-06,
"loss": 1.2976,
"step": 174
},
{
"epoch": 0.39984767707539987,
"grad_norm": 1.0585196018218994,
"learning_rate": 4.9890974971659405e-06,
"loss": 1.2921,
"step": 175
},
{
"epoch": 0.4021325209444021,
"grad_norm": 1.018211007118225,
"learning_rate": 4.988805044157353e-06,
"loss": 1.3046,
"step": 176
},
{
"epoch": 0.4044173648134044,
"grad_norm": 1.0587507486343384,
"learning_rate": 4.9885087292542865e-06,
"loss": 1.2901,
"step": 177
},
{
"epoch": 0.4067022086824067,
"grad_norm": 1.0261503458023071,
"learning_rate": 4.988208552916535e-06,
"loss": 1.3081,
"step": 178
},
{
"epoch": 0.408987052551409,
"grad_norm": 1.0412943363189697,
"learning_rate": 4.9879045156098846e-06,
"loss": 1.3052,
"step": 179
},
{
"epoch": 0.4112718964204113,
"grad_norm": 1.0323666334152222,
"learning_rate": 4.987596617806111e-06,
"loss": 1.3048,
"step": 180
},
{
"epoch": 0.4135567402894136,
"grad_norm": 1.0095067024230957,
"learning_rate": 4.9872848599829825e-06,
"loss": 1.3292,
"step": 181
},
{
"epoch": 0.4158415841584158,
"grad_norm": 0.9761002659797668,
"learning_rate": 4.986969242624254e-06,
"loss": 1.2884,
"step": 182
},
{
"epoch": 0.4181264280274181,
"grad_norm": 1.0436338186264038,
"learning_rate": 4.986649766219671e-06,
"loss": 1.3211,
"step": 183
},
{
"epoch": 0.4204112718964204,
"grad_norm": 1.0505225658416748,
"learning_rate": 4.986326431264969e-06,
"loss": 1.2863,
"step": 184
},
{
"epoch": 0.4226961157654227,
"grad_norm": 1.006611943244934,
"learning_rate": 4.985999238261867e-06,
"loss": 1.2812,
"step": 185
},
{
"epoch": 0.424980959634425,
"grad_norm": 1.0494719743728638,
"learning_rate": 4.985668187718073e-06,
"loss": 1.3105,
"step": 186
},
{
"epoch": 0.42726580350342724,
"grad_norm": 0.9847164750099182,
"learning_rate": 4.985333280147281e-06,
"loss": 1.2811,
"step": 187
},
{
"epoch": 0.42955064737242954,
"grad_norm": 1.0337165594100952,
"learning_rate": 4.984994516069168e-06,
"loss": 1.2876,
"step": 188
},
{
"epoch": 0.43183549124143183,
"grad_norm": 1.0178074836730957,
"learning_rate": 4.984651896009396e-06,
"loss": 1.2597,
"step": 189
},
{
"epoch": 0.43412033511043413,
"grad_norm": 1.0170668363571167,
"learning_rate": 4.984305420499612e-06,
"loss": 1.2916,
"step": 190
},
{
"epoch": 0.4364051789794364,
"grad_norm": 1.0148853063583374,
"learning_rate": 4.983955090077445e-06,
"loss": 1.2785,
"step": 191
},
{
"epoch": 0.43869002284843867,
"grad_norm": 1.0563602447509766,
"learning_rate": 4.983600905286502e-06,
"loss": 1.295,
"step": 192
},
{
"epoch": 0.44097486671744096,
"grad_norm": 0.9817858338356018,
"learning_rate": 4.983242866676376e-06,
"loss": 1.2832,
"step": 193
},
{
"epoch": 0.44325971058644326,
"grad_norm": 1.0299488306045532,
"learning_rate": 4.982880974802638e-06,
"loss": 1.2952,
"step": 194
},
{
"epoch": 0.44554455445544555,
"grad_norm": 0.9951279163360596,
"learning_rate": 4.982515230226837e-06,
"loss": 1.2901,
"step": 195
},
{
"epoch": 0.44782939832444785,
"grad_norm": 1.0001885890960693,
"learning_rate": 4.982145633516501e-06,
"loss": 1.2554,
"step": 196
},
{
"epoch": 0.4501142421934501,
"grad_norm": 1.0821017026901245,
"learning_rate": 4.981772185245135e-06,
"loss": 1.2903,
"step": 197
},
{
"epoch": 0.4523990860624524,
"grad_norm": 1.0269831418991089,
"learning_rate": 4.981394885992223e-06,
"loss": 1.3077,
"step": 198
},
{
"epoch": 0.4546839299314547,
"grad_norm": 1.025965929031372,
"learning_rate": 4.981013736343221e-06,
"loss": 1.2771,
"step": 199
},
{
"epoch": 0.456968773800457,
"grad_norm": 0.9828860759735107,
"learning_rate": 4.980628736889562e-06,
"loss": 1.2788,
"step": 200
},
{
"epoch": 0.45925361766945927,
"grad_norm": 1.077913761138916,
"learning_rate": 4.9802398882286515e-06,
"loss": 1.2815,
"step": 201
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.1024688482284546,
"learning_rate": 4.97984719096387e-06,
"loss": 1.3135,
"step": 202
},
{
"epoch": 0.4638233054074638,
"grad_norm": 1.0494202375411987,
"learning_rate": 4.979450645704567e-06,
"loss": 1.3027,
"step": 203
},
{
"epoch": 0.4661081492764661,
"grad_norm": 1.0050199031829834,
"learning_rate": 4.979050253066064e-06,
"loss": 1.3016,
"step": 204
},
{
"epoch": 0.4683929931454684,
"grad_norm": 1.0264744758605957,
"learning_rate": 4.978646013669652e-06,
"loss": 1.343,
"step": 205
},
{
"epoch": 0.4706778370144707,
"grad_norm": 1.001989722251892,
"learning_rate": 4.978237928142594e-06,
"loss": 1.3088,
"step": 206
},
{
"epoch": 0.472962680883473,
"grad_norm": 1.0501984357833862,
"learning_rate": 4.977825997118119e-06,
"loss": 1.2875,
"step": 207
},
{
"epoch": 0.4752475247524752,
"grad_norm": 1.0487364530563354,
"learning_rate": 4.977410221235421e-06,
"loss": 1.2917,
"step": 208
},
{
"epoch": 0.4775323686214775,
"grad_norm": 1.0768541097640991,
"learning_rate": 4.976990601139662e-06,
"loss": 1.3,
"step": 209
},
{
"epoch": 0.4798172124904798,
"grad_norm": 0.9696170687675476,
"learning_rate": 4.9765671374819715e-06,
"loss": 1.2822,
"step": 210
},
{
"epoch": 0.4821020563594821,
"grad_norm": 0.9987464547157288,
"learning_rate": 4.9761398309194385e-06,
"loss": 1.3076,
"step": 211
},
{
"epoch": 0.4843869002284844,
"grad_norm": 1.0254422426223755,
"learning_rate": 4.975708682115118e-06,
"loss": 1.281,
"step": 212
},
{
"epoch": 0.48667174409748665,
"grad_norm": 1.0040076971054077,
"learning_rate": 4.9752736917380274e-06,
"loss": 1.2821,
"step": 213
},
{
"epoch": 0.48895658796648894,
"grad_norm": 1.004184365272522,
"learning_rate": 4.9748348604631416e-06,
"loss": 1.2641,
"step": 214
},
{
"epoch": 0.49124143183549124,
"grad_norm": 1.0694694519042969,
"learning_rate": 4.9743921889714005e-06,
"loss": 1.2853,
"step": 215
},
{
"epoch": 0.49352627570449353,
"grad_norm": 1.0564874410629272,
"learning_rate": 4.973945677949699e-06,
"loss": 1.2882,
"step": 216
},
{
"epoch": 0.49581111957349583,
"grad_norm": 1.0076894760131836,
"learning_rate": 4.973495328090891e-06,
"loss": 1.2868,
"step": 217
},
{
"epoch": 0.49809596344249807,
"grad_norm": 1.0476043224334717,
"learning_rate": 4.973041140093786e-06,
"loss": 1.2642,
"step": 218
},
{
"epoch": 0.5003808073115004,
"grad_norm": 1.050991415977478,
"learning_rate": 4.972583114663153e-06,
"loss": 1.2751,
"step": 219
},
{
"epoch": 0.5026656511805027,
"grad_norm": 0.9902971386909485,
"learning_rate": 4.972121252509712e-06,
"loss": 1.2685,
"step": 220
},
{
"epoch": 0.504950495049505,
"grad_norm": 1.0011303424835205,
"learning_rate": 4.971655554350137e-06,
"loss": 1.2829,
"step": 221
},
{
"epoch": 0.5072353389185073,
"grad_norm": 1.010233998298645,
"learning_rate": 4.971186020907054e-06,
"loss": 1.277,
"step": 222
},
{
"epoch": 0.5095201827875095,
"grad_norm": 1.0275652408599854,
"learning_rate": 4.970712652909042e-06,
"loss": 1.2971,
"step": 223
},
{
"epoch": 0.5118050266565118,
"grad_norm": 1.0285537242889404,
"learning_rate": 4.970235451090629e-06,
"loss": 1.231,
"step": 224
},
{
"epoch": 0.5140898705255141,
"grad_norm": 1.0604579448699951,
"learning_rate": 4.969754416192292e-06,
"loss": 1.269,
"step": 225
},
{
"epoch": 0.5163747143945163,
"grad_norm": 1.0375958681106567,
"learning_rate": 4.969269548960456e-06,
"loss": 1.2712,
"step": 226
},
{
"epoch": 0.5186595582635186,
"grad_norm": 1.037304401397705,
"learning_rate": 4.9687808501474925e-06,
"loss": 1.2826,
"step": 227
},
{
"epoch": 0.5209444021325209,
"grad_norm": 1.0280749797821045,
"learning_rate": 4.968288320511718e-06,
"loss": 1.2726,
"step": 228
},
{
"epoch": 0.5232292460015232,
"grad_norm": 1.0595530271530151,
"learning_rate": 4.967791960817395e-06,
"loss": 1.281,
"step": 229
},
{
"epoch": 0.5255140898705255,
"grad_norm": 0.9964226484298706,
"learning_rate": 4.967291771834727e-06,
"loss": 1.3188,
"step": 230
},
{
"epoch": 0.5277989337395278,
"grad_norm": 1.0433804988861084,
"learning_rate": 4.966787754339861e-06,
"loss": 1.274,
"step": 231
},
{
"epoch": 0.5300837776085301,
"grad_norm": 1.079641580581665,
"learning_rate": 4.966279909114883e-06,
"loss": 1.2991,
"step": 232
},
{
"epoch": 0.5323686214775324,
"grad_norm": 1.0351816415786743,
"learning_rate": 4.965768236947821e-06,
"loss": 1.2659,
"step": 233
},
{
"epoch": 0.5346534653465347,
"grad_norm": 1.0495244264602661,
"learning_rate": 4.96525273863264e-06,
"loss": 1.2898,
"step": 234
},
{
"epoch": 0.536938309215537,
"grad_norm": 1.0479910373687744,
"learning_rate": 4.964733414969241e-06,
"loss": 1.2536,
"step": 235
},
{
"epoch": 0.5392231530845393,
"grad_norm": 1.0365879535675049,
"learning_rate": 4.964210266763461e-06,
"loss": 1.2369,
"step": 236
},
{
"epoch": 0.5415079969535415,
"grad_norm": 1.0398730039596558,
"learning_rate": 4.9636832948270745e-06,
"loss": 1.2669,
"step": 237
},
{
"epoch": 0.5437928408225438,
"grad_norm": 1.0146657228469849,
"learning_rate": 4.963152499977786e-06,
"loss": 1.2893,
"step": 238
},
{
"epoch": 0.546077684691546,
"grad_norm": 1.0974043607711792,
"learning_rate": 4.962617883039233e-06,
"loss": 1.2452,
"step": 239
},
{
"epoch": 0.5483625285605483,
"grad_norm": 0.9900649189949036,
"learning_rate": 4.962079444840985e-06,
"loss": 1.2215,
"step": 240
},
{
"epoch": 0.5506473724295506,
"grad_norm": 1.003464937210083,
"learning_rate": 4.9615371862185394e-06,
"loss": 1.2744,
"step": 241
},
{
"epoch": 0.5529322162985529,
"grad_norm": 1.004382848739624,
"learning_rate": 4.960991108013322e-06,
"loss": 1.271,
"step": 242
},
{
"epoch": 0.5552170601675552,
"grad_norm": 1.0129280090332031,
"learning_rate": 4.960441211072686e-06,
"loss": 1.2874,
"step": 243
},
{
"epoch": 0.5575019040365575,
"grad_norm": 1.040189266204834,
"learning_rate": 4.9598874962499096e-06,
"loss": 1.2918,
"step": 244
},
{
"epoch": 0.5597867479055598,
"grad_norm": 1.0145982503890991,
"learning_rate": 4.959329964404197e-06,
"loss": 1.2713,
"step": 245
},
{
"epoch": 0.5620715917745621,
"grad_norm": 1.0469987392425537,
"learning_rate": 4.958768616400672e-06,
"loss": 1.2689,
"step": 246
},
{
"epoch": 0.5643564356435643,
"grad_norm": 1.0191642045974731,
"learning_rate": 4.958203453110384e-06,
"loss": 1.2718,
"step": 247
},
{
"epoch": 0.5666412795125666,
"grad_norm": 1.0718231201171875,
"learning_rate": 4.957634475410298e-06,
"loss": 1.3128,
"step": 248
},
{
"epoch": 0.5689261233815689,
"grad_norm": 1.0109634399414062,
"learning_rate": 4.957061684183301e-06,
"loss": 1.2586,
"step": 249
},
{
"epoch": 0.5712109672505712,
"grad_norm": 0.9942657947540283,
"learning_rate": 4.956485080318198e-06,
"loss": 1.328,
"step": 250
},
{
"epoch": 0.5734958111195735,
"grad_norm": 1.0184757709503174,
"learning_rate": 4.955904664709707e-06,
"loss": 1.2815,
"step": 251
},
{
"epoch": 0.5757806549885758,
"grad_norm": 1.015625,
"learning_rate": 4.955320438258465e-06,
"loss": 1.2585,
"step": 252
},
{
"epoch": 0.5780654988575781,
"grad_norm": 0.9848981499671936,
"learning_rate": 4.954732401871018e-06,
"loss": 1.2866,
"step": 253
},
{
"epoch": 0.5803503427265804,
"grad_norm": 1.0482749938964844,
"learning_rate": 4.954140556459826e-06,
"loss": 1.2732,
"step": 254
},
{
"epoch": 0.5826351865955827,
"grad_norm": 1.0250680446624756,
"learning_rate": 4.95354490294326e-06,
"loss": 1.3053,
"step": 255
},
{
"epoch": 0.584920030464585,
"grad_norm": 1.0545597076416016,
"learning_rate": 4.952945442245598e-06,
"loss": 1.2638,
"step": 256
},
{
"epoch": 0.5872048743335873,
"grad_norm": 1.044873833656311,
"learning_rate": 4.952342175297028e-06,
"loss": 1.2683,
"step": 257
},
{
"epoch": 0.5894897182025894,
"grad_norm": 1.0361744165420532,
"learning_rate": 4.951735103033644e-06,
"loss": 1.2887,
"step": 258
},
{
"epoch": 0.5917745620715917,
"grad_norm": 1.0238685607910156,
"learning_rate": 4.951124226397441e-06,
"loss": 1.2736,
"step": 259
},
{
"epoch": 0.594059405940594,
"grad_norm": 1.0217833518981934,
"learning_rate": 4.950509546336323e-06,
"loss": 1.2681,
"step": 260
},
{
"epoch": 0.5963442498095963,
"grad_norm": 1.0546188354492188,
"learning_rate": 4.949891063804091e-06,
"loss": 1.2582,
"step": 261
},
{
"epoch": 0.5986290936785986,
"grad_norm": 1.0834907293319702,
"learning_rate": 4.94926877976045e-06,
"loss": 1.2487,
"step": 262
},
{
"epoch": 0.6009139375476009,
"grad_norm": 1.062184453010559,
"learning_rate": 4.948642695171e-06,
"loss": 1.3188,
"step": 263
},
{
"epoch": 0.6031987814166032,
"grad_norm": 1.0373252630233765,
"learning_rate": 4.948012811007242e-06,
"loss": 1.277,
"step": 264
},
{
"epoch": 0.6054836252856055,
"grad_norm": 1.0140316486358643,
"learning_rate": 4.947379128246571e-06,
"loss": 1.2617,
"step": 265
},
{
"epoch": 0.6077684691546078,
"grad_norm": 1.054410696029663,
"learning_rate": 4.946741647872277e-06,
"loss": 1.238,
"step": 266
},
{
"epoch": 0.6100533130236101,
"grad_norm": 1.0967663526535034,
"learning_rate": 4.94610037087354e-06,
"loss": 1.2682,
"step": 267
},
{
"epoch": 0.6123381568926123,
"grad_norm": 1.043338656425476,
"learning_rate": 4.945455298245436e-06,
"loss": 1.2572,
"step": 268
},
{
"epoch": 0.6146230007616146,
"grad_norm": 1.0187970399856567,
"learning_rate": 4.944806430988927e-06,
"loss": 1.2613,
"step": 269
},
{
"epoch": 0.6169078446306169,
"grad_norm": 1.0666472911834717,
"learning_rate": 4.9441537701108654e-06,
"loss": 1.2611,
"step": 270
},
{
"epoch": 0.6191926884996192,
"grad_norm": 1.0025635957717896,
"learning_rate": 4.943497316623988e-06,
"loss": 1.2519,
"step": 271
},
{
"epoch": 0.6214775323686215,
"grad_norm": 1.0135135650634766,
"learning_rate": 4.942837071546919e-06,
"loss": 1.2759,
"step": 272
},
{
"epoch": 0.6237623762376238,
"grad_norm": 0.9985151886940002,
"learning_rate": 4.942173035904164e-06,
"loss": 1.2844,
"step": 273
},
{
"epoch": 0.6260472201066261,
"grad_norm": 0.9952817559242249,
"learning_rate": 4.941505210726112e-06,
"loss": 1.2356,
"step": 274
},
{
"epoch": 0.6283320639756284,
"grad_norm": 1.0448962450027466,
"learning_rate": 4.9408335970490305e-06,
"loss": 1.2587,
"step": 275
},
{
"epoch": 0.6306169078446306,
"grad_norm": 1.011099100112915,
"learning_rate": 4.940158195915067e-06,
"loss": 1.2729,
"step": 276
},
{
"epoch": 0.6329017517136329,
"grad_norm": 1.052904725074768,
"learning_rate": 4.939479008372247e-06,
"loss": 1.2536,
"step": 277
},
{
"epoch": 0.6351865955826352,
"grad_norm": 1.058173418045044,
"learning_rate": 4.938796035474469e-06,
"loss": 1.2807,
"step": 278
},
{
"epoch": 0.6374714394516374,
"grad_norm": 1.022147536277771,
"learning_rate": 4.938109278281506e-06,
"loss": 1.2887,
"step": 279
},
{
"epoch": 0.6397562833206397,
"grad_norm": 1.0064011812210083,
"learning_rate": 4.937418737859004e-06,
"loss": 1.2192,
"step": 280
},
{
"epoch": 0.642041127189642,
"grad_norm": 1.0092360973358154,
"learning_rate": 4.936724415278479e-06,
"loss": 1.3159,
"step": 281
},
{
"epoch": 0.6443259710586443,
"grad_norm": 1.076401710510254,
"learning_rate": 4.936026311617316e-06,
"loss": 1.2872,
"step": 282
},
{
"epoch": 0.6466108149276466,
"grad_norm": 1.057209849357605,
"learning_rate": 4.935324427958766e-06,
"loss": 1.257,
"step": 283
},
{
"epoch": 0.6488956587966489,
"grad_norm": 1.1738762855529785,
"learning_rate": 4.934618765391946e-06,
"loss": 1.2547,
"step": 284
},
{
"epoch": 0.6511805026656512,
"grad_norm": 1.0405137538909912,
"learning_rate": 4.933909325011838e-06,
"loss": 1.2766,
"step": 285
},
{
"epoch": 0.6534653465346535,
"grad_norm": 1.0377894639968872,
"learning_rate": 4.933196107919286e-06,
"loss": 1.2624,
"step": 286
},
{
"epoch": 0.6557501904036558,
"grad_norm": 1.032714605331421,
"learning_rate": 4.932479115220991e-06,
"loss": 1.2527,
"step": 287
},
{
"epoch": 0.6580350342726581,
"grad_norm": 1.0755581855773926,
"learning_rate": 4.9317583480295175e-06,
"loss": 1.2966,
"step": 288
},
{
"epoch": 0.6603198781416603,
"grad_norm": 1.0262556076049805,
"learning_rate": 4.931033807463283e-06,
"loss": 1.2585,
"step": 289
},
{
"epoch": 0.6626047220106626,
"grad_norm": 1.0510430335998535,
"learning_rate": 4.930305494646562e-06,
"loss": 1.2662,
"step": 290
},
{
"epoch": 0.6648895658796649,
"grad_norm": 1.035854458808899,
"learning_rate": 4.9295734107094825e-06,
"loss": 1.2346,
"step": 291
},
{
"epoch": 0.6671744097486672,
"grad_norm": 1.0485846996307373,
"learning_rate": 4.928837556788023e-06,
"loss": 1.2978,
"step": 292
},
{
"epoch": 0.6694592536176694,
"grad_norm": 1.02550208568573,
"learning_rate": 4.928097934024013e-06,
"loss": 1.2478,
"step": 293
},
{
"epoch": 0.6717440974866717,
"grad_norm": 1.0328837633132935,
"learning_rate": 4.927354543565131e-06,
"loss": 1.2788,
"step": 294
},
{
"epoch": 0.674028941355674,
"grad_norm": 0.9913997054100037,
"learning_rate": 4.926607386564898e-06,
"loss": 1.2423,
"step": 295
},
{
"epoch": 0.6763137852246763,
"grad_norm": 1.0034306049346924,
"learning_rate": 4.925856464182685e-06,
"loss": 1.2562,
"step": 296
},
{
"epoch": 0.6785986290936786,
"grad_norm": 1.0546495914459229,
"learning_rate": 4.925101777583701e-06,
"loss": 1.2598,
"step": 297
},
{
"epoch": 0.6808834729626809,
"grad_norm": 1.0412935018539429,
"learning_rate": 4.924343327938999e-06,
"loss": 1.2744,
"step": 298
},
{
"epoch": 0.6831683168316832,
"grad_norm": 1.0731669664382935,
"learning_rate": 4.923581116425471e-06,
"loss": 1.2912,
"step": 299
},
{
"epoch": 0.6854531607006854,
"grad_norm": 1.0394880771636963,
"learning_rate": 4.922815144225843e-06,
"loss": 1.276,
"step": 300
},
{
"epoch": 0.6877380045696877,
"grad_norm": 1.0383579730987549,
"learning_rate": 4.92204541252868e-06,
"loss": 1.255,
"step": 301
},
{
"epoch": 0.69002284843869,
"grad_norm": 1.0251744985580444,
"learning_rate": 4.92127192252838e-06,
"loss": 1.2688,
"step": 302
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.017650842666626,
"learning_rate": 4.9204946754251724e-06,
"loss": 1.2818,
"step": 303
},
{
"epoch": 0.6945925361766946,
"grad_norm": 1.0219080448150635,
"learning_rate": 4.919713672425116e-06,
"loss": 1.2828,
"step": 304
},
{
"epoch": 0.6968773800456969,
"grad_norm": 1.0862151384353638,
"learning_rate": 4.918928914740098e-06,
"loss": 1.2514,
"step": 305
},
{
"epoch": 0.6991622239146992,
"grad_norm": 1.0639281272888184,
"learning_rate": 4.918140403587831e-06,
"loss": 1.2739,
"step": 306
},
{
"epoch": 0.7014470677837015,
"grad_norm": 1.0512444972991943,
"learning_rate": 4.9173481401918556e-06,
"loss": 1.2576,
"step": 307
},
{
"epoch": 0.7037319116527038,
"grad_norm": 1.0291866064071655,
"learning_rate": 4.916552125781529e-06,
"loss": 1.2934,
"step": 308
},
{
"epoch": 0.7060167555217061,
"grad_norm": 1.0338629484176636,
"learning_rate": 4.915752361592032e-06,
"loss": 1.263,
"step": 309
},
{
"epoch": 0.7083015993907082,
"grad_norm": 1.0358542203903198,
"learning_rate": 4.914948848864365e-06,
"loss": 1.2453,
"step": 310
},
{
"epoch": 0.7105864432597105,
"grad_norm": 1.1184923648834229,
"learning_rate": 4.914141588845344e-06,
"loss": 1.2653,
"step": 311
},
{
"epoch": 0.7128712871287128,
"grad_norm": 1.0791000127792358,
"learning_rate": 4.913330582787598e-06,
"loss": 1.2659,
"step": 312
},
{
"epoch": 0.7151561309977151,
"grad_norm": 1.0901819467544556,
"learning_rate": 4.912515831949571e-06,
"loss": 1.2208,
"step": 313
},
{
"epoch": 0.7174409748667174,
"grad_norm": 1.0219902992248535,
"learning_rate": 4.9116973375955166e-06,
"loss": 1.2711,
"step": 314
},
{
"epoch": 0.7197258187357197,
"grad_norm": 1.014364242553711,
"learning_rate": 4.910875100995499e-06,
"loss": 1.2877,
"step": 315
},
{
"epoch": 0.722010662604722,
"grad_norm": 1.0699234008789062,
"learning_rate": 4.910049123425386e-06,
"loss": 1.2425,
"step": 316
},
{
"epoch": 0.7242955064737243,
"grad_norm": 1.0614267587661743,
"learning_rate": 4.9092194061668535e-06,
"loss": 1.2475,
"step": 317
},
{
"epoch": 0.7265803503427266,
"grad_norm": 1.0620336532592773,
"learning_rate": 4.908385950507378e-06,
"loss": 1.2618,
"step": 318
},
{
"epoch": 0.7288651942117289,
"grad_norm": 1.0389032363891602,
"learning_rate": 4.90754875774024e-06,
"loss": 1.2742,
"step": 319
},
{
"epoch": 0.7311500380807312,
"grad_norm": 0.9754124879837036,
"learning_rate": 4.9067078291645144e-06,
"loss": 1.25,
"step": 320
},
{
"epoch": 0.7334348819497334,
"grad_norm": 1.056058406829834,
"learning_rate": 4.905863166085076e-06,
"loss": 1.2451,
"step": 321
},
{
"epoch": 0.7357197258187357,
"grad_norm": 1.0641580820083618,
"learning_rate": 4.9050147698125944e-06,
"loss": 1.2532,
"step": 322
},
{
"epoch": 0.738004569687738,
"grad_norm": 1.0407251119613647,
"learning_rate": 4.904162641663532e-06,
"loss": 1.3103,
"step": 323
},
{
"epoch": 0.7402894135567403,
"grad_norm": 1.0477187633514404,
"learning_rate": 4.9033067829601385e-06,
"loss": 1.2658,
"step": 324
},
{
"epoch": 0.7425742574257426,
"grad_norm": 1.0202401876449585,
"learning_rate": 4.902447195030459e-06,
"loss": 1.2569,
"step": 325
},
{
"epoch": 0.7448591012947449,
"grad_norm": 1.0629253387451172,
"learning_rate": 4.9015838792083196e-06,
"loss": 1.247,
"step": 326
},
{
"epoch": 0.7471439451637472,
"grad_norm": 1.0284748077392578,
"learning_rate": 4.900716836833333e-06,
"loss": 1.2659,
"step": 327
},
{
"epoch": 0.7494287890327495,
"grad_norm": 1.0653586387634277,
"learning_rate": 4.899846069250894e-06,
"loss": 1.2673,
"step": 328
},
{
"epoch": 0.7517136329017517,
"grad_norm": 1.0795682668685913,
"learning_rate": 4.898971577812179e-06,
"loss": 1.2778,
"step": 329
},
{
"epoch": 0.753998476770754,
"grad_norm": 1.0359232425689697,
"learning_rate": 4.8980933638741426e-06,
"loss": 1.2732,
"step": 330
},
{
"epoch": 0.7562833206397562,
"grad_norm": 1.0286237001419067,
"learning_rate": 4.897211428799512e-06,
"loss": 1.2455,
"step": 331
},
{
"epoch": 0.7585681645087585,
"grad_norm": 1.0179105997085571,
"learning_rate": 4.896325773956793e-06,
"loss": 1.2413,
"step": 332
},
{
"epoch": 0.7608530083777608,
"grad_norm": 1.0381865501403809,
"learning_rate": 4.895436400720264e-06,
"loss": 1.2409,
"step": 333
},
{
"epoch": 0.7631378522467631,
"grad_norm": 0.9918906688690186,
"learning_rate": 4.894543310469968e-06,
"loss": 1.2556,
"step": 334
},
{
"epoch": 0.7654226961157654,
"grad_norm": 1.0300416946411133,
"learning_rate": 4.8936465045917204e-06,
"loss": 1.2325,
"step": 335
},
{
"epoch": 0.7677075399847677,
"grad_norm": 1.052534580230713,
"learning_rate": 4.8927459844770995e-06,
"loss": 1.2561,
"step": 336
},
{
"epoch": 0.76999238385377,
"grad_norm": 1.0454604625701904,
"learning_rate": 4.891841751523448e-06,
"loss": 1.2845,
"step": 337
},
{
"epoch": 0.7722772277227723,
"grad_norm": 1.0518709421157837,
"learning_rate": 4.8909338071338706e-06,
"loss": 1.2485,
"step": 338
},
{
"epoch": 0.7745620715917746,
"grad_norm": 1.0326422452926636,
"learning_rate": 4.890022152717231e-06,
"loss": 1.2757,
"step": 339
},
{
"epoch": 0.7768469154607769,
"grad_norm": 1.2617943286895752,
"learning_rate": 4.889106789688148e-06,
"loss": 1.2656,
"step": 340
},
{
"epoch": 0.7791317593297792,
"grad_norm": 1.0038459300994873,
"learning_rate": 4.888187719466996e-06,
"loss": 1.2636,
"step": 341
},
{
"epoch": 0.7814166031987814,
"grad_norm": 1.1393420696258545,
"learning_rate": 4.887264943479903e-06,
"loss": 1.2621,
"step": 342
},
{
"epoch": 0.7837014470677837,
"grad_norm": 1.0969446897506714,
"learning_rate": 4.8863384631587446e-06,
"loss": 1.2208,
"step": 343
},
{
"epoch": 0.785986290936786,
"grad_norm": 1.034393310546875,
"learning_rate": 4.885408279941148e-06,
"loss": 1.2101,
"step": 344
},
{
"epoch": 0.7882711348057883,
"grad_norm": 1.1397764682769775,
"learning_rate": 4.884474395270484e-06,
"loss": 1.2823,
"step": 345
},
{
"epoch": 0.7905559786747905,
"grad_norm": 1.1488789319992065,
"learning_rate": 4.883536810595867e-06,
"loss": 1.2615,
"step": 346
},
{
"epoch": 0.7928408225437928,
"grad_norm": 1.0274580717086792,
"learning_rate": 4.8825955273721524e-06,
"loss": 1.2334,
"step": 347
},
{
"epoch": 0.7951256664127951,
"grad_norm": 1.0355713367462158,
"learning_rate": 4.8816505470599365e-06,
"loss": 1.2224,
"step": 348
},
{
"epoch": 0.7974105102817974,
"grad_norm": 1.0540703535079956,
"learning_rate": 4.880701871125551e-06,
"loss": 1.262,
"step": 349
},
{
"epoch": 0.7996953541507997,
"grad_norm": 1.0765819549560547,
"learning_rate": 4.879749501041062e-06,
"loss": 1.2731,
"step": 350
},
{
"epoch": 0.801980198019802,
"grad_norm": 1.0639638900756836,
"learning_rate": 4.878793438284268e-06,
"loss": 1.2673,
"step": 351
},
{
"epoch": 0.8042650418888042,
"grad_norm": 1.0149368047714233,
"learning_rate": 4.877833684338698e-06,
"loss": 1.2479,
"step": 352
},
{
"epoch": 0.8065498857578065,
"grad_norm": 1.1710435152053833,
"learning_rate": 4.876870240693608e-06,
"loss": 1.2775,
"step": 353
},
{
"epoch": 0.8088347296268088,
"grad_norm": 1.1317570209503174,
"learning_rate": 4.875903108843979e-06,
"loss": 1.2732,
"step": 354
},
{
"epoch": 0.8111195734958111,
"grad_norm": 1.0417158603668213,
"learning_rate": 4.874932290290517e-06,
"loss": 1.2647,
"step": 355
},
{
"epoch": 0.8134044173648134,
"grad_norm": 1.073765516281128,
"learning_rate": 4.873957786539646e-06,
"loss": 1.2738,
"step": 356
},
{
"epoch": 0.8156892612338157,
"grad_norm": 1.018481731414795,
"learning_rate": 4.872979599103511e-06,
"loss": 1.2509,
"step": 357
},
{
"epoch": 0.817974105102818,
"grad_norm": 1.0737470388412476,
"learning_rate": 4.8719977294999695e-06,
"loss": 1.232,
"step": 358
},
{
"epoch": 0.8202589489718203,
"grad_norm": 1.0921229124069214,
"learning_rate": 4.871012179252597e-06,
"loss": 1.2342,
"step": 359
},
{
"epoch": 0.8225437928408226,
"grad_norm": 1.0502641201019287,
"learning_rate": 4.870022949890676e-06,
"loss": 1.2463,
"step": 360
},
{
"epoch": 0.8248286367098249,
"grad_norm": 1.1755155324935913,
"learning_rate": 4.869030042949202e-06,
"loss": 1.2625,
"step": 361
},
{
"epoch": 0.8271134805788272,
"grad_norm": 1.0167341232299805,
"learning_rate": 4.868033459968874e-06,
"loss": 1.2563,
"step": 362
},
{
"epoch": 0.8293983244478293,
"grad_norm": 1.0481575727462769,
"learning_rate": 4.8670332024960954e-06,
"loss": 1.2541,
"step": 363
},
{
"epoch": 0.8316831683168316,
"grad_norm": 1.0657804012298584,
"learning_rate": 4.866029272082973e-06,
"loss": 1.2444,
"step": 364
},
{
"epoch": 0.8339680121858339,
"grad_norm": 1.0473397970199585,
"learning_rate": 4.865021670287311e-06,
"loss": 1.2356,
"step": 365
},
{
"epoch": 0.8362528560548362,
"grad_norm": 1.011077880859375,
"learning_rate": 4.864010398672612e-06,
"loss": 1.2417,
"step": 366
},
{
"epoch": 0.8385376999238385,
"grad_norm": 1.0485464334487915,
"learning_rate": 4.862995458808073e-06,
"loss": 1.2728,
"step": 367
},
{
"epoch": 0.8408225437928408,
"grad_norm": 1.0683908462524414,
"learning_rate": 4.861976852268582e-06,
"loss": 1.2354,
"step": 368
},
{
"epoch": 0.8431073876618431,
"grad_norm": 1.0323604345321655,
"learning_rate": 4.860954580634718e-06,
"loss": 1.2665,
"step": 369
},
{
"epoch": 0.8453922315308454,
"grad_norm": 1.024782419204712,
"learning_rate": 4.859928645492746e-06,
"loss": 1.2515,
"step": 370
},
{
"epoch": 0.8476770753998477,
"grad_norm": 1.02902090549469,
"learning_rate": 4.858899048434614e-06,
"loss": 1.2274,
"step": 371
},
{
"epoch": 0.84996191926885,
"grad_norm": 1.0355148315429688,
"learning_rate": 4.857865791057957e-06,
"loss": 1.2289,
"step": 372
},
{
"epoch": 0.8522467631378522,
"grad_norm": 1.0638132095336914,
"learning_rate": 4.856828874966086e-06,
"loss": 1.2245,
"step": 373
},
{
"epoch": 0.8545316070068545,
"grad_norm": 1.0459909439086914,
"learning_rate": 4.8557883017679895e-06,
"loss": 1.2347,
"step": 374
},
{
"epoch": 0.8568164508758568,
"grad_norm": 1.0818232297897339,
"learning_rate": 4.854744073078333e-06,
"loss": 1.2564,
"step": 375
},
{
"epoch": 0.8591012947448591,
"grad_norm": 1.0551162958145142,
"learning_rate": 4.853696190517452e-06,
"loss": 1.2809,
"step": 376
},
{
"epoch": 0.8613861386138614,
"grad_norm": 1.0419256687164307,
"learning_rate": 4.8526446557113525e-06,
"loss": 1.2532,
"step": 377
},
{
"epoch": 0.8636709824828637,
"grad_norm": 1.058478832244873,
"learning_rate": 4.851589470291707e-06,
"loss": 1.229,
"step": 378
},
{
"epoch": 0.865955826351866,
"grad_norm": 1.0275694131851196,
"learning_rate": 4.850530635895854e-06,
"loss": 1.2555,
"step": 379
},
{
"epoch": 0.8682406702208683,
"grad_norm": 1.0653144121170044,
"learning_rate": 4.849468154166794e-06,
"loss": 1.2397,
"step": 380
},
{
"epoch": 0.8705255140898706,
"grad_norm": 1.0227371454238892,
"learning_rate": 4.8484020267531855e-06,
"loss": 1.2568,
"step": 381
},
{
"epoch": 0.8728103579588729,
"grad_norm": 1.0583505630493164,
"learning_rate": 4.847332255309346e-06,
"loss": 1.2489,
"step": 382
},
{
"epoch": 0.8750952018278751,
"grad_norm": 1.0397239923477173,
"learning_rate": 4.846258841495246e-06,
"loss": 1.273,
"step": 383
},
{
"epoch": 0.8773800456968773,
"grad_norm": 1.020776391029358,
"learning_rate": 4.845181786976509e-06,
"loss": 1.2257,
"step": 384
},
{
"epoch": 0.8796648895658796,
"grad_norm": 1.0420705080032349,
"learning_rate": 4.844101093424407e-06,
"loss": 1.296,
"step": 385
},
{
"epoch": 0.8819497334348819,
"grad_norm": 1.0465624332427979,
"learning_rate": 4.84301676251586e-06,
"loss": 1.2514,
"step": 386
},
{
"epoch": 0.8842345773038842,
"grad_norm": 1.0915330648422241,
"learning_rate": 4.841928795933429e-06,
"loss": 1.2664,
"step": 387
},
{
"epoch": 0.8865194211728865,
"grad_norm": 1.0246195793151855,
"learning_rate": 4.84083719536532e-06,
"loss": 1.2499,
"step": 388
},
{
"epoch": 0.8888042650418888,
"grad_norm": 1.0145692825317383,
"learning_rate": 4.839741962505376e-06,
"loss": 1.2638,
"step": 389
},
{
"epoch": 0.8910891089108911,
"grad_norm": 1.05404531955719,
"learning_rate": 4.838643099053077e-06,
"loss": 1.1875,
"step": 390
},
{
"epoch": 0.8933739527798934,
"grad_norm": 1.1422752141952515,
"learning_rate": 4.837540606713538e-06,
"loss": 1.2496,
"step": 391
},
{
"epoch": 0.8956587966488957,
"grad_norm": 1.0648959875106812,
"learning_rate": 4.8364344871975e-06,
"loss": 1.2375,
"step": 392
},
{
"epoch": 0.897943640517898,
"grad_norm": 1.0459322929382324,
"learning_rate": 4.835324742221338e-06,
"loss": 1.2419,
"step": 393
},
{
"epoch": 0.9002284843869002,
"grad_norm": 1.0693044662475586,
"learning_rate": 4.834211373507048e-06,
"loss": 1.2485,
"step": 394
},
{
"epoch": 0.9025133282559025,
"grad_norm": 1.0930724143981934,
"learning_rate": 4.833094382782255e-06,
"loss": 1.2389,
"step": 395
},
{
"epoch": 0.9047981721249048,
"grad_norm": 1.1270296573638916,
"learning_rate": 4.831973771780197e-06,
"loss": 1.2033,
"step": 396
},
{
"epoch": 0.9070830159939071,
"grad_norm": 1.044074535369873,
"learning_rate": 4.830849542239735e-06,
"loss": 1.2464,
"step": 397
},
{
"epoch": 0.9093678598629094,
"grad_norm": 1.0138458013534546,
"learning_rate": 4.829721695905343e-06,
"loss": 1.2473,
"step": 398
},
{
"epoch": 0.9116527037319117,
"grad_norm": 1.1201279163360596,
"learning_rate": 4.828590234527107e-06,
"loss": 1.2729,
"step": 399
},
{
"epoch": 0.913937547600914,
"grad_norm": 1.0771571397781372,
"learning_rate": 4.8274551598607214e-06,
"loss": 1.2665,
"step": 400
},
{
"epoch": 0.9162223914699162,
"grad_norm": 1.0691912174224854,
"learning_rate": 4.8263164736674905e-06,
"loss": 1.2094,
"step": 401
},
{
"epoch": 0.9185072353389185,
"grad_norm": 1.0740418434143066,
"learning_rate": 4.8251741777143205e-06,
"loss": 1.2879,
"step": 402
},
{
"epoch": 0.9207920792079208,
"grad_norm": 1.0185081958770752,
"learning_rate": 4.824028273773719e-06,
"loss": 1.2459,
"step": 403
},
{
"epoch": 0.9230769230769231,
"grad_norm": 1.0672869682312012,
"learning_rate": 4.822878763623792e-06,
"loss": 1.2394,
"step": 404
},
{
"epoch": 0.9253617669459253,
"grad_norm": 1.08120858669281,
"learning_rate": 4.821725649048242e-06,
"loss": 1.2918,
"step": 405
},
{
"epoch": 0.9276466108149276,
"grad_norm": 1.0407681465148926,
"learning_rate": 4.820568931836364e-06,
"loss": 1.2443,
"step": 406
},
{
"epoch": 0.9299314546839299,
"grad_norm": 1.0847117900848389,
"learning_rate": 4.8194086137830445e-06,
"loss": 1.2505,
"step": 407
},
{
"epoch": 0.9322162985529322,
"grad_norm": 1.0484883785247803,
"learning_rate": 4.818244696688754e-06,
"loss": 1.2469,
"step": 408
},
{
"epoch": 0.9345011424219345,
"grad_norm": 1.0654011964797974,
"learning_rate": 4.817077182359553e-06,
"loss": 1.2544,
"step": 409
},
{
"epoch": 0.9367859862909368,
"grad_norm": 1.108176589012146,
"learning_rate": 4.815906072607079e-06,
"loss": 1.2387,
"step": 410
},
{
"epoch": 0.9390708301599391,
"grad_norm": 1.0624432563781738,
"learning_rate": 4.8147313692485495e-06,
"loss": 1.2488,
"step": 411
},
{
"epoch": 0.9413556740289414,
"grad_norm": 1.0391454696655273,
"learning_rate": 4.813553074106761e-06,
"loss": 1.2514,
"step": 412
},
{
"epoch": 0.9436405178979437,
"grad_norm": 1.1086232662200928,
"learning_rate": 4.812371189010081e-06,
"loss": 1.2694,
"step": 413
},
{
"epoch": 0.945925361766946,
"grad_norm": 1.0448237657546997,
"learning_rate": 4.8111857157924465e-06,
"loss": 1.2366,
"step": 414
},
{
"epoch": 0.9482102056359482,
"grad_norm": 1.0393203496932983,
"learning_rate": 4.809996656293367e-06,
"loss": 1.2747,
"step": 415
},
{
"epoch": 0.9504950495049505,
"grad_norm": 1.083590030670166,
"learning_rate": 4.8088040123579106e-06,
"loss": 1.2167,
"step": 416
},
{
"epoch": 0.9527798933739527,
"grad_norm": 1.071567177772522,
"learning_rate": 4.807607785836711e-06,
"loss": 1.2108,
"step": 417
},
{
"epoch": 0.955064737242955,
"grad_norm": 1.0953818559646606,
"learning_rate": 4.8064079785859615e-06,
"loss": 1.2381,
"step": 418
},
{
"epoch": 0.9573495811119573,
"grad_norm": 1.0628875494003296,
"learning_rate": 4.8052045924674105e-06,
"loss": 1.232,
"step": 419
},
{
"epoch": 0.9596344249809596,
"grad_norm": 1.0838161706924438,
"learning_rate": 4.803997629348359e-06,
"loss": 1.2699,
"step": 420
},
{
"epoch": 0.9619192688499619,
"grad_norm": 0.9980992078781128,
"learning_rate": 4.802787091101659e-06,
"loss": 1.2473,
"step": 421
},
{
"epoch": 0.9642041127189642,
"grad_norm": 1.094283103942871,
"learning_rate": 4.801572979605712e-06,
"loss": 1.2656,
"step": 422
},
{
"epoch": 0.9664889565879665,
"grad_norm": 1.0554611682891846,
"learning_rate": 4.800355296744461e-06,
"loss": 1.2584,
"step": 423
},
{
"epoch": 0.9687738004569688,
"grad_norm": 1.1019188165664673,
"learning_rate": 4.799134044407392e-06,
"loss": 1.2877,
"step": 424
},
{
"epoch": 0.9710586443259711,
"grad_norm": 1.087965726852417,
"learning_rate": 4.797909224489531e-06,
"loss": 1.2662,
"step": 425
},
{
"epoch": 0.9733434881949733,
"grad_norm": 1.08269202709198,
"learning_rate": 4.796680838891438e-06,
"loss": 1.2419,
"step": 426
},
{
"epoch": 0.9756283320639756,
"grad_norm": 1.071199893951416,
"learning_rate": 4.795448889519207e-06,
"loss": 1.2489,
"step": 427
},
{
"epoch": 0.9779131759329779,
"grad_norm": 1.0306544303894043,
"learning_rate": 4.794213378284462e-06,
"loss": 1.2467,
"step": 428
},
{
"epoch": 0.9801980198019802,
"grad_norm": 1.0567327737808228,
"learning_rate": 4.792974307104353e-06,
"loss": 1.2637,
"step": 429
},
{
"epoch": 0.9824828636709825,
"grad_norm": 1.0448797941207886,
"learning_rate": 4.7917316779015554e-06,
"loss": 1.2244,
"step": 430
},
{
"epoch": 0.9847677075399848,
"grad_norm": 1.0123138427734375,
"learning_rate": 4.790485492604264e-06,
"loss": 1.2326,
"step": 431
},
{
"epoch": 0.9870525514089871,
"grad_norm": 1.0484559535980225,
"learning_rate": 4.789235753146192e-06,
"loss": 1.2436,
"step": 432
},
{
"epoch": 0.9893373952779894,
"grad_norm": 1.0161617994308472,
"learning_rate": 4.787982461466568e-06,
"loss": 1.2185,
"step": 433
},
{
"epoch": 0.9916222391469917,
"grad_norm": 1.0779787302017212,
"learning_rate": 4.786725619510134e-06,
"loss": 1.2256,
"step": 434
},
{
"epoch": 0.993907083015994,
"grad_norm": 1.061590552330017,
"learning_rate": 4.785465229227139e-06,
"loss": 1.2747,
"step": 435
},
{
"epoch": 0.9961919268849961,
"grad_norm": 1.102403163909912,
"learning_rate": 4.784201292573337e-06,
"loss": 1.2561,
"step": 436
},
{
"epoch": 0.9984767707539984,
"grad_norm": 0.9936567544937134,
"learning_rate": 4.782933811509988e-06,
"loss": 1.2409,
"step": 437
},
{
"epoch": 1.0,
"grad_norm": 0.9936567544937134,
"learning_rate": 4.781662788003851e-06,
"loss": 1.2271,
"step": 438
},
{
"epoch": 1.0022848438690022,
"grad_norm": 1.4983975887298584,
"learning_rate": 4.780388224027179e-06,
"loss": 1.2312,
"step": 439
},
{
"epoch": 1.0045696877380046,
"grad_norm": 1.0163486003875732,
"learning_rate": 4.779110121557723e-06,
"loss": 1.1992,
"step": 440
},
{
"epoch": 1.0068545316070068,
"grad_norm": 1.0127511024475098,
"learning_rate": 4.777828482578722e-06,
"loss": 1.2135,
"step": 441
},
{
"epoch": 1.0091393754760092,
"grad_norm": 1.0328449010849,
"learning_rate": 4.776543309078903e-06,
"loss": 1.2143,
"step": 442
},
{
"epoch": 1.0114242193450114,
"grad_norm": 1.1579132080078125,
"learning_rate": 4.7752546030524775e-06,
"loss": 1.2051,
"step": 443
},
{
"epoch": 1.0137090632140138,
"grad_norm": 1.0556671619415283,
"learning_rate": 4.77396236649914e-06,
"loss": 1.2136,
"step": 444
},
{
"epoch": 1.015993907083016,
"grad_norm": 1.0315356254577637,
"learning_rate": 4.772666601424061e-06,
"loss": 1.2444,
"step": 445
},
{
"epoch": 1.0182787509520184,
"grad_norm": 1.0836431980133057,
"learning_rate": 4.771367309837888e-06,
"loss": 1.1967,
"step": 446
},
{
"epoch": 1.0205635948210205,
"grad_norm": 1.0516763925552368,
"learning_rate": 4.7700644937567385e-06,
"loss": 1.2012,
"step": 447
},
{
"epoch": 1.022848438690023,
"grad_norm": 1.0447187423706055,
"learning_rate": 4.768758155202202e-06,
"loss": 1.2281,
"step": 448
},
{
"epoch": 1.0251332825590251,
"grad_norm": 1.0647971630096436,
"learning_rate": 4.767448296201332e-06,
"loss": 1.1907,
"step": 449
},
{
"epoch": 1.0274181264280273,
"grad_norm": 1.0792133808135986,
"learning_rate": 4.766134918786646e-06,
"loss": 1.2346,
"step": 450
},
{
"epoch": 1.0297029702970297,
"grad_norm": 1.1311832666397095,
"learning_rate": 4.764818024996117e-06,
"loss": 1.2652,
"step": 451
},
{
"epoch": 1.031987814166032,
"grad_norm": 1.090455174446106,
"learning_rate": 4.763497616873181e-06,
"loss": 1.2258,
"step": 452
},
{
"epoch": 1.0342726580350343,
"grad_norm": 1.0815131664276123,
"learning_rate": 4.7621736964667204e-06,
"loss": 1.2233,
"step": 453
},
{
"epoch": 1.0365575019040365,
"grad_norm": 1.112673044204712,
"learning_rate": 4.760846265831073e-06,
"loss": 1.2136,
"step": 454
},
{
"epoch": 1.038842345773039,
"grad_norm": 1.068178653717041,
"learning_rate": 4.759515327026019e-06,
"loss": 1.214,
"step": 455
},
{
"epoch": 1.041127189642041,
"grad_norm": 1.079059362411499,
"learning_rate": 4.758180882116788e-06,
"loss": 1.2024,
"step": 456
},
{
"epoch": 1.0434120335110435,
"grad_norm": 1.1010431051254272,
"learning_rate": 4.756842933174044e-06,
"loss": 1.2239,
"step": 457
},
{
"epoch": 1.0456968773800457,
"grad_norm": 1.1039162874221802,
"learning_rate": 4.755501482273892e-06,
"loss": 1.2212,
"step": 458
},
{
"epoch": 1.047981721249048,
"grad_norm": 1.0536257028579712,
"learning_rate": 4.754156531497869e-06,
"loss": 1.1672,
"step": 459
},
{
"epoch": 1.0502665651180503,
"grad_norm": 1.085410475730896,
"learning_rate": 4.752808082932943e-06,
"loss": 1.2471,
"step": 460
},
{
"epoch": 1.0525514089870525,
"grad_norm": 1.0432335138320923,
"learning_rate": 4.751456138671512e-06,
"loss": 1.2345,
"step": 461
},
{
"epoch": 1.0548362528560549,
"grad_norm": 1.072708249092102,
"learning_rate": 4.750100700811395e-06,
"loss": 1.2328,
"step": 462
},
{
"epoch": 1.057121096725057,
"grad_norm": 1.0852991342544556,
"learning_rate": 4.748741771455835e-06,
"loss": 1.19,
"step": 463
},
{
"epoch": 1.0594059405940595,
"grad_norm": 1.0906398296356201,
"learning_rate": 4.747379352713489e-06,
"loss": 1.229,
"step": 464
},
{
"epoch": 1.0616907844630616,
"grad_norm": 1.055309772491455,
"learning_rate": 4.746013446698432e-06,
"loss": 1.2419,
"step": 465
},
{
"epoch": 1.063975628332064,
"grad_norm": 1.0667710304260254,
"learning_rate": 4.744644055530149e-06,
"loss": 1.1943,
"step": 466
},
{
"epoch": 1.0662604722010662,
"grad_norm": 1.091098427772522,
"learning_rate": 4.743271181333533e-06,
"loss": 1.171,
"step": 467
},
{
"epoch": 1.0685453160700686,
"grad_norm": 1.0701195001602173,
"learning_rate": 4.741894826238882e-06,
"loss": 1.2163,
"step": 468
},
{
"epoch": 1.0708301599390708,
"grad_norm": 1.0526652336120605,
"learning_rate": 4.740514992381893e-06,
"loss": 1.2329,
"step": 469
},
{
"epoch": 1.073115003808073,
"grad_norm": 1.0725401639938354,
"learning_rate": 4.739131681903666e-06,
"loss": 1.1793,
"step": 470
},
{
"epoch": 1.0753998476770754,
"grad_norm": 1.126091480255127,
"learning_rate": 4.737744896950689e-06,
"loss": 1.1769,
"step": 471
},
{
"epoch": 1.0776846915460776,
"grad_norm": 1.039844036102295,
"learning_rate": 4.736354639674847e-06,
"loss": 1.205,
"step": 472
},
{
"epoch": 1.07996953541508,
"grad_norm": 1.0432695150375366,
"learning_rate": 4.734960912233411e-06,
"loss": 1.1909,
"step": 473
},
{
"epoch": 1.0822543792840822,
"grad_norm": 1.0890403985977173,
"learning_rate": 4.7335637167890366e-06,
"loss": 1.1928,
"step": 474
},
{
"epoch": 1.0845392231530846,
"grad_norm": 1.0710458755493164,
"learning_rate": 4.732163055509759e-06,
"loss": 1.2402,
"step": 475
},
{
"epoch": 1.0868240670220868,
"grad_norm": 1.0940525531768799,
"learning_rate": 4.730758930568997e-06,
"loss": 1.181,
"step": 476
},
{
"epoch": 1.0891089108910892,
"grad_norm": 1.177641749382019,
"learning_rate": 4.729351344145536e-06,
"loss": 1.2232,
"step": 477
},
{
"epoch": 1.0913937547600914,
"grad_norm": 1.1401522159576416,
"learning_rate": 4.72794029842354e-06,
"loss": 1.1878,
"step": 478
},
{
"epoch": 1.0936785986290938,
"grad_norm": 1.0874228477478027,
"learning_rate": 4.726525795592535e-06,
"loss": 1.1658,
"step": 479
},
{
"epoch": 1.095963442498096,
"grad_norm": 1.0325064659118652,
"learning_rate": 4.725107837847414e-06,
"loss": 1.2084,
"step": 480
},
{
"epoch": 1.0982482863670981,
"grad_norm": 1.0492074489593506,
"learning_rate": 4.723686427388434e-06,
"loss": 1.2208,
"step": 481
},
{
"epoch": 1.1005331302361006,
"grad_norm": 1.1399495601654053,
"learning_rate": 4.722261566421204e-06,
"loss": 1.2158,
"step": 482
},
{
"epoch": 1.1028179741051027,
"grad_norm": 1.1156668663024902,
"learning_rate": 4.72083325715669e-06,
"loss": 1.2252,
"step": 483
},
{
"epoch": 1.1051028179741051,
"grad_norm": 1.072943091392517,
"learning_rate": 4.719401501811209e-06,
"loss": 1.2381,
"step": 484
},
{
"epoch": 1.1073876618431073,
"grad_norm": 1.0337257385253906,
"learning_rate": 4.717966302606424e-06,
"loss": 1.1782,
"step": 485
},
{
"epoch": 1.1096725057121097,
"grad_norm": 1.0744901895523071,
"learning_rate": 4.716527661769344e-06,
"loss": 1.2412,
"step": 486
},
{
"epoch": 1.111957349581112,
"grad_norm": 1.0331535339355469,
"learning_rate": 4.715085581532316e-06,
"loss": 1.1869,
"step": 487
},
{
"epoch": 1.1142421934501143,
"grad_norm": 1.0795518159866333,
"learning_rate": 4.7136400641330245e-06,
"loss": 1.214,
"step": 488
},
{
"epoch": 1.1165270373191165,
"grad_norm": 1.084125280380249,
"learning_rate": 4.71219111181449e-06,
"loss": 1.2049,
"step": 489
},
{
"epoch": 1.118811881188119,
"grad_norm": 1.1166882514953613,
"learning_rate": 4.710738726825059e-06,
"loss": 1.2143,
"step": 490
},
{
"epoch": 1.121096725057121,
"grad_norm": 1.0764187574386597,
"learning_rate": 4.709282911418408e-06,
"loss": 1.2301,
"step": 491
},
{
"epoch": 1.1233815689261233,
"grad_norm": 1.069149374961853,
"learning_rate": 4.7078236678535335e-06,
"loss": 1.2094,
"step": 492
},
{
"epoch": 1.1256664127951257,
"grad_norm": 1.0744988918304443,
"learning_rate": 4.7063609983947535e-06,
"loss": 1.1893,
"step": 493
},
{
"epoch": 1.1279512566641279,
"grad_norm": 1.090267300605774,
"learning_rate": 4.704894905311701e-06,
"loss": 1.1575,
"step": 494
},
{
"epoch": 1.1302361005331303,
"grad_norm": 1.067543625831604,
"learning_rate": 4.703425390879323e-06,
"loss": 1.1801,
"step": 495
},
{
"epoch": 1.1325209444021325,
"grad_norm": 1.0365897417068481,
"learning_rate": 4.701952457377874e-06,
"loss": 1.2197,
"step": 496
},
{
"epoch": 1.1348057882711349,
"grad_norm": 1.066163420677185,
"learning_rate": 4.700476107092913e-06,
"loss": 1.2156,
"step": 497
},
{
"epoch": 1.137090632140137,
"grad_norm": 1.1297317743301392,
"learning_rate": 4.698996342315303e-06,
"loss": 1.2064,
"step": 498
},
{
"epoch": 1.1393754760091395,
"grad_norm": 1.069610834121704,
"learning_rate": 4.697513165341204e-06,
"loss": 1.1986,
"step": 499
},
{
"epoch": 1.1416603198781416,
"grad_norm": 1.0844234228134155,
"learning_rate": 4.696026578472073e-06,
"loss": 1.1892,
"step": 500
},
{
"epoch": 1.1439451637471438,
"grad_norm": 1.079389214515686,
"learning_rate": 4.694536584014653e-06,
"loss": 1.201,
"step": 501
},
{
"epoch": 1.1462300076161462,
"grad_norm": 1.075682520866394,
"learning_rate": 4.693043184280978e-06,
"loss": 1.2261,
"step": 502
},
{
"epoch": 1.1485148514851484,
"grad_norm": 1.0244231224060059,
"learning_rate": 4.69154638158837e-06,
"loss": 1.2048,
"step": 503
},
{
"epoch": 1.1507996953541508,
"grad_norm": 1.0907280445098877,
"learning_rate": 4.690046178259423e-06,
"loss": 1.2202,
"step": 504
},
{
"epoch": 1.153084539223153,
"grad_norm": 1.097701907157898,
"learning_rate": 4.688542576622013e-06,
"loss": 1.1781,
"step": 505
},
{
"epoch": 1.1553693830921554,
"grad_norm": 1.0993037223815918,
"learning_rate": 4.687035579009288e-06,
"loss": 1.2113,
"step": 506
},
{
"epoch": 1.1576542269611576,
"grad_norm": 1.085300087928772,
"learning_rate": 4.685525187759666e-06,
"loss": 1.1996,
"step": 507
},
{
"epoch": 1.15993907083016,
"grad_norm": 1.0483977794647217,
"learning_rate": 4.684011405216832e-06,
"loss": 1.2343,
"step": 508
},
{
"epoch": 1.1622239146991622,
"grad_norm": 1.064441442489624,
"learning_rate": 4.682494233729729e-06,
"loss": 1.2405,
"step": 509
},
{
"epoch": 1.1645087585681646,
"grad_norm": 1.05643630027771,
"learning_rate": 4.680973675652564e-06,
"loss": 1.2112,
"step": 510
},
{
"epoch": 1.1667936024371668,
"grad_norm": 1.0908793210983276,
"learning_rate": 4.679449733344796e-06,
"loss": 1.2077,
"step": 511
},
{
"epoch": 1.1690784463061692,
"grad_norm": 1.1248105764389038,
"learning_rate": 4.677922409171136e-06,
"loss": 1.1987,
"step": 512
},
{
"epoch": 1.1713632901751714,
"grad_norm": 1.0605143308639526,
"learning_rate": 4.6763917055015414e-06,
"loss": 1.2027,
"step": 513
},
{
"epoch": 1.1736481340441736,
"grad_norm": 1.0496442317962646,
"learning_rate": 4.674857624711216e-06,
"loss": 1.2259,
"step": 514
},
{
"epoch": 1.175932977913176,
"grad_norm": 1.1550832986831665,
"learning_rate": 4.673320169180601e-06,
"loss": 1.2418,
"step": 515
},
{
"epoch": 1.1782178217821782,
"grad_norm": 1.1532083749771118,
"learning_rate": 4.671779341295378e-06,
"loss": 1.2265,
"step": 516
},
{
"epoch": 1.1805026656511806,
"grad_norm": 1.081101417541504,
"learning_rate": 4.670235143446457e-06,
"loss": 1.2078,
"step": 517
},
{
"epoch": 1.1827875095201827,
"grad_norm": 1.0701441764831543,
"learning_rate": 4.668687578029983e-06,
"loss": 1.2252,
"step": 518
},
{
"epoch": 1.1850723533891852,
"grad_norm": 1.0859651565551758,
"learning_rate": 4.667136647447319e-06,
"loss": 1.2131,
"step": 519
},
{
"epoch": 1.1873571972581873,
"grad_norm": 1.122533917427063,
"learning_rate": 4.6655823541050575e-06,
"loss": 1.1608,
"step": 520
},
{
"epoch": 1.1896420411271897,
"grad_norm": 1.0594309568405151,
"learning_rate": 4.664024700415002e-06,
"loss": 1.2151,
"step": 521
},
{
"epoch": 1.191926884996192,
"grad_norm": 1.1281721591949463,
"learning_rate": 4.662463688794175e-06,
"loss": 1.2197,
"step": 522
},
{
"epoch": 1.194211728865194,
"grad_norm": 1.1049555540084839,
"learning_rate": 4.660899321664808e-06,
"loss": 1.2416,
"step": 523
},
{
"epoch": 1.1964965727341965,
"grad_norm": 1.1022320985794067,
"learning_rate": 4.65933160145434e-06,
"loss": 1.1842,
"step": 524
},
{
"epoch": 1.1987814166031987,
"grad_norm": 1.1182475090026855,
"learning_rate": 4.657760530595411e-06,
"loss": 1.1417,
"step": 525
},
{
"epoch": 1.201066260472201,
"grad_norm": 1.1111788749694824,
"learning_rate": 4.656186111525863e-06,
"loss": 1.2092,
"step": 526
},
{
"epoch": 1.2033511043412033,
"grad_norm": 1.0595399141311646,
"learning_rate": 4.654608346688731e-06,
"loss": 1.1549,
"step": 527
},
{
"epoch": 1.2056359482102057,
"grad_norm": 1.1151765584945679,
"learning_rate": 4.6530272385322426e-06,
"loss": 1.2469,
"step": 528
},
{
"epoch": 1.2079207920792079,
"grad_norm": 1.0334808826446533,
"learning_rate": 4.651442789509813e-06,
"loss": 1.229,
"step": 529
},
{
"epoch": 1.2102056359482103,
"grad_norm": 1.0765459537506104,
"learning_rate": 4.649855002080044e-06,
"loss": 1.2163,
"step": 530
},
{
"epoch": 1.2124904798172125,
"grad_norm": 1.1509029865264893,
"learning_rate": 4.648263878706712e-06,
"loss": 1.2101,
"step": 531
},
{
"epoch": 1.2147753236862147,
"grad_norm": 1.1111629009246826,
"learning_rate": 4.646669421858776e-06,
"loss": 1.2379,
"step": 532
},
{
"epoch": 1.217060167555217,
"grad_norm": 1.1683619022369385,
"learning_rate": 4.645071634010363e-06,
"loss": 1.2011,
"step": 533
},
{
"epoch": 1.2193450114242192,
"grad_norm": 1.2373298406600952,
"learning_rate": 4.643470517640772e-06,
"loss": 1.1502,
"step": 534
},
{
"epoch": 1.2216298552932217,
"grad_norm": 1.080675482749939,
"learning_rate": 4.641866075234463e-06,
"loss": 1.2173,
"step": 535
},
{
"epoch": 1.2239146991622238,
"grad_norm": 1.0971184968948364,
"learning_rate": 4.640258309281062e-06,
"loss": 1.2117,
"step": 536
},
{
"epoch": 1.2261995430312262,
"grad_norm": 1.183856725692749,
"learning_rate": 4.638647222275349e-06,
"loss": 1.2137,
"step": 537
},
{
"epoch": 1.2284843869002284,
"grad_norm": 1.2277085781097412,
"learning_rate": 4.637032816717256e-06,
"loss": 1.1977,
"step": 538
},
{
"epoch": 1.2307692307692308,
"grad_norm": 1.1087970733642578,
"learning_rate": 4.6354150951118676e-06,
"loss": 1.2256,
"step": 539
},
{
"epoch": 1.233054074638233,
"grad_norm": 1.0706229209899902,
"learning_rate": 4.633794059969413e-06,
"loss": 1.2429,
"step": 540
},
{
"epoch": 1.2353389185072354,
"grad_norm": 1.1261042356491089,
"learning_rate": 4.632169713805262e-06,
"loss": 1.219,
"step": 541
},
{
"epoch": 1.2376237623762376,
"grad_norm": 1.1817506551742554,
"learning_rate": 4.630542059139923e-06,
"loss": 1.2367,
"step": 542
},
{
"epoch": 1.23990860624524,
"grad_norm": 1.1145075559616089,
"learning_rate": 4.628911098499039e-06,
"loss": 1.2029,
"step": 543
},
{
"epoch": 1.2421934501142422,
"grad_norm": 1.1309086084365845,
"learning_rate": 4.62727683441338e-06,
"loss": 1.2374,
"step": 544
},
{
"epoch": 1.2444782939832444,
"grad_norm": 1.0949103832244873,
"learning_rate": 4.6256392694188445e-06,
"loss": 1.2204,
"step": 545
},
{
"epoch": 1.2467631378522468,
"grad_norm": 1.2004865407943726,
"learning_rate": 4.6239984060564535e-06,
"loss": 1.2327,
"step": 546
},
{
"epoch": 1.249047981721249,
"grad_norm": 1.286232829093933,
"learning_rate": 4.622354246872344e-06,
"loss": 1.1838,
"step": 547
},
{
"epoch": 1.2513328255902514,
"grad_norm": 1.0974533557891846,
"learning_rate": 4.620706794417769e-06,
"loss": 1.1678,
"step": 548
},
{
"epoch": 1.2536176694592536,
"grad_norm": 1.0960924625396729,
"learning_rate": 4.61905605124909e-06,
"loss": 1.2314,
"step": 549
},
{
"epoch": 1.255902513328256,
"grad_norm": 1.1535454988479614,
"learning_rate": 4.617402019927776e-06,
"loss": 1.1928,
"step": 550
},
{
"epoch": 1.2581873571972582,
"grad_norm": 1.2693071365356445,
"learning_rate": 4.615744703020396e-06,
"loss": 1.1966,
"step": 551
},
{
"epoch": 1.2604722010662606,
"grad_norm": 1.1645997762680054,
"learning_rate": 4.614084103098623e-06,
"loss": 1.2251,
"step": 552
},
{
"epoch": 1.2627570449352628,
"grad_norm": 1.1186461448669434,
"learning_rate": 4.6124202227392175e-06,
"loss": 1.2037,
"step": 553
},
{
"epoch": 1.265041888804265,
"grad_norm": 1.1100102663040161,
"learning_rate": 4.610753064524034e-06,
"loss": 1.2011,
"step": 554
},
{
"epoch": 1.2673267326732673,
"grad_norm": 1.1173806190490723,
"learning_rate": 4.609082631040012e-06,
"loss": 1.1871,
"step": 555
},
{
"epoch": 1.2696115765422697,
"grad_norm": 1.1128157377243042,
"learning_rate": 4.6074089248791735e-06,
"loss": 1.1965,
"step": 556
},
{
"epoch": 1.271896420411272,
"grad_norm": 1.0940717458724976,
"learning_rate": 4.60573194863862e-06,
"loss": 1.193,
"step": 557
},
{
"epoch": 1.2741812642802741,
"grad_norm": 1.0955843925476074,
"learning_rate": 4.604051704920526e-06,
"loss": 1.187,
"step": 558
},
{
"epoch": 1.2764661081492765,
"grad_norm": 1.0802319049835205,
"learning_rate": 4.602368196332134e-06,
"loss": 1.1753,
"step": 559
},
{
"epoch": 1.2787509520182787,
"grad_norm": 1.1677578687667847,
"learning_rate": 4.600681425485757e-06,
"loss": 1.1964,
"step": 560
},
{
"epoch": 1.2810357958872811,
"grad_norm": 1.0803853273391724,
"learning_rate": 4.598991394998768e-06,
"loss": 1.2052,
"step": 561
},
{
"epoch": 1.2833206397562833,
"grad_norm": 1.128667950630188,
"learning_rate": 4.5972981074935975e-06,
"loss": 1.1774,
"step": 562
},
{
"epoch": 1.2856054836252855,
"grad_norm": 1.0685629844665527,
"learning_rate": 4.59560156559773e-06,
"loss": 1.1897,
"step": 563
},
{
"epoch": 1.287890327494288,
"grad_norm": 1.1464303731918335,
"learning_rate": 4.593901771943702e-06,
"loss": 1.1809,
"step": 564
},
{
"epoch": 1.2901751713632903,
"grad_norm": 1.1095281839370728,
"learning_rate": 4.592198729169091e-06,
"loss": 1.2118,
"step": 565
},
{
"epoch": 1.2924600152322925,
"grad_norm": 1.0897274017333984,
"learning_rate": 4.5904924399165215e-06,
"loss": 1.177,
"step": 566
},
{
"epoch": 1.2947448591012947,
"grad_norm": 1.0702495574951172,
"learning_rate": 4.588782906833653e-06,
"loss": 1.1872,
"step": 567
},
{
"epoch": 1.297029702970297,
"grad_norm": 1.0990184545516968,
"learning_rate": 4.587070132573178e-06,
"loss": 1.1903,
"step": 568
},
{
"epoch": 1.2993145468392993,
"grad_norm": 1.121097207069397,
"learning_rate": 4.58535411979282e-06,
"loss": 1.2469,
"step": 569
},
{
"epoch": 1.3015993907083017,
"grad_norm": 1.0787534713745117,
"learning_rate": 4.583634871155326e-06,
"loss": 1.1995,
"step": 570
},
{
"epoch": 1.3038842345773038,
"grad_norm": 1.0721417665481567,
"learning_rate": 4.581912389328466e-06,
"loss": 1.1889,
"step": 571
},
{
"epoch": 1.306169078446306,
"grad_norm": 1.1017696857452393,
"learning_rate": 4.580186676985024e-06,
"loss": 1.2133,
"step": 572
},
{
"epoch": 1.3084539223153084,
"grad_norm": 1.1040468215942383,
"learning_rate": 4.578457736802801e-06,
"loss": 1.1894,
"step": 573
},
{
"epoch": 1.3107387661843108,
"grad_norm": 1.0856465101242065,
"learning_rate": 4.576725571464604e-06,
"loss": 1.2234,
"step": 574
},
{
"epoch": 1.313023610053313,
"grad_norm": 1.0786073207855225,
"learning_rate": 4.574990183658244e-06,
"loss": 1.1989,
"step": 575
},
{
"epoch": 1.3153084539223152,
"grad_norm": 1.0701881647109985,
"learning_rate": 4.573251576076532e-06,
"loss": 1.2095,
"step": 576
},
{
"epoch": 1.3175932977913176,
"grad_norm": 1.0697689056396484,
"learning_rate": 4.5715097514172794e-06,
"loss": 1.2198,
"step": 577
},
{
"epoch": 1.3198781416603198,
"grad_norm": 1.1303515434265137,
"learning_rate": 4.569764712383284e-06,
"loss": 1.2456,
"step": 578
},
{
"epoch": 1.3221629855293222,
"grad_norm": 1.1471296548843384,
"learning_rate": 4.5680164616823355e-06,
"loss": 1.2155,
"step": 579
},
{
"epoch": 1.3244478293983244,
"grad_norm": 1.0679783821105957,
"learning_rate": 4.566265002027204e-06,
"loss": 1.2346,
"step": 580
},
{
"epoch": 1.3267326732673268,
"grad_norm": 1.087063193321228,
"learning_rate": 4.564510336135642e-06,
"loss": 1.1735,
"step": 581
},
{
"epoch": 1.329017517136329,
"grad_norm": 1.190617322921753,
"learning_rate": 4.562752466730374e-06,
"loss": 1.2472,
"step": 582
},
{
"epoch": 1.3313023610053314,
"grad_norm": 1.0759129524230957,
"learning_rate": 4.560991396539099e-06,
"loss": 1.2263,
"step": 583
},
{
"epoch": 1.3335872048743336,
"grad_norm": 1.080640196800232,
"learning_rate": 4.559227128294479e-06,
"loss": 1.1773,
"step": 584
},
{
"epoch": 1.3358720487433358,
"grad_norm": 1.0868074893951416,
"learning_rate": 4.5574596647341414e-06,
"loss": 1.254,
"step": 585
},
{
"epoch": 1.3381568926123382,
"grad_norm": 1.0621445178985596,
"learning_rate": 4.55568900860067e-06,
"loss": 1.2091,
"step": 586
},
{
"epoch": 1.3404417364813406,
"grad_norm": 1.1124675273895264,
"learning_rate": 4.553915162641602e-06,
"loss": 1.2093,
"step": 587
},
{
"epoch": 1.3427265803503428,
"grad_norm": 1.0877987146377563,
"learning_rate": 4.552138129609428e-06,
"loss": 1.2399,
"step": 588
},
{
"epoch": 1.345011424219345,
"grad_norm": 1.1441737413406372,
"learning_rate": 4.550357912261579e-06,
"loss": 1.2274,
"step": 589
},
{
"epoch": 1.3472962680883473,
"grad_norm": 1.131813645362854,
"learning_rate": 4.548574513360431e-06,
"loss": 1.2296,
"step": 590
},
{
"epoch": 1.3495811119573495,
"grad_norm": 1.0901340246200562,
"learning_rate": 4.546787935673294e-06,
"loss": 1.2151,
"step": 591
},
{
"epoch": 1.351865955826352,
"grad_norm": 1.1450812816619873,
"learning_rate": 4.544998181972412e-06,
"loss": 1.2054,
"step": 592
},
{
"epoch": 1.3541507996953541,
"grad_norm": 1.0988374948501587,
"learning_rate": 4.543205255034958e-06,
"loss": 1.2133,
"step": 593
},
{
"epoch": 1.3564356435643563,
"grad_norm": 1.0787577629089355,
"learning_rate": 4.541409157643027e-06,
"loss": 1.1972,
"step": 594
},
{
"epoch": 1.3587204874333587,
"grad_norm": 1.0694243907928467,
"learning_rate": 4.539609892583637e-06,
"loss": 1.2182,
"step": 595
},
{
"epoch": 1.3610053313023611,
"grad_norm": 1.065865397453308,
"learning_rate": 4.537807462648716e-06,
"loss": 1.2057,
"step": 596
},
{
"epoch": 1.3632901751713633,
"grad_norm": 1.0816274881362915,
"learning_rate": 4.5360018706351075e-06,
"loss": 1.1846,
"step": 597
},
{
"epoch": 1.3655750190403655,
"grad_norm": 1.105301022529602,
"learning_rate": 4.5341931193445585e-06,
"loss": 1.2219,
"step": 598
},
{
"epoch": 1.367859862909368,
"grad_norm": 1.1194454431533813,
"learning_rate": 4.5323812115837215e-06,
"loss": 1.2021,
"step": 599
},
{
"epoch": 1.37014470677837,
"grad_norm": 1.0899710655212402,
"learning_rate": 4.530566150164145e-06,
"loss": 1.173,
"step": 600
},
{
"epoch": 1.3724295506473725,
"grad_norm": 1.0824511051177979,
"learning_rate": 4.528747937902271e-06,
"loss": 1.2131,
"step": 601
},
{
"epoch": 1.3747143945163747,
"grad_norm": 1.0796427726745605,
"learning_rate": 4.52692657761943e-06,
"loss": 1.1911,
"step": 602
},
{
"epoch": 1.376999238385377,
"grad_norm": 1.1047371625900269,
"learning_rate": 4.525102072141839e-06,
"loss": 1.1734,
"step": 603
},
{
"epoch": 1.3792840822543793,
"grad_norm": 1.1101821660995483,
"learning_rate": 4.523274424300596e-06,
"loss": 1.2274,
"step": 604
},
{
"epoch": 1.3815689261233817,
"grad_norm": 1.1410820484161377,
"learning_rate": 4.521443636931671e-06,
"loss": 1.2,
"step": 605
},
{
"epoch": 1.3838537699923839,
"grad_norm": 1.0687155723571777,
"learning_rate": 4.5196097128759095e-06,
"loss": 1.2028,
"step": 606
},
{
"epoch": 1.386138613861386,
"grad_norm": 1.0923937559127808,
"learning_rate": 4.517772654979024e-06,
"loss": 1.2522,
"step": 607
},
{
"epoch": 1.3884234577303884,
"grad_norm": 1.1132218837738037,
"learning_rate": 4.515932466091587e-06,
"loss": 1.1797,
"step": 608
},
{
"epoch": 1.3907083015993906,
"grad_norm": 1.182809591293335,
"learning_rate": 4.514089149069033e-06,
"loss": 1.1885,
"step": 609
},
{
"epoch": 1.392993145468393,
"grad_norm": 1.064723253250122,
"learning_rate": 4.512242706771647e-06,
"loss": 1.174,
"step": 610
},
{
"epoch": 1.3952779893373952,
"grad_norm": 1.1065499782562256,
"learning_rate": 4.510393142064567e-06,
"loss": 1.1919,
"step": 611
},
{
"epoch": 1.3975628332063976,
"grad_norm": 1.0989713668823242,
"learning_rate": 4.508540457817772e-06,
"loss": 1.1835,
"step": 612
},
{
"epoch": 1.3998476770753998,
"grad_norm": 1.0850595235824585,
"learning_rate": 4.506684656906085e-06,
"loss": 1.1945,
"step": 613
},
{
"epoch": 1.4021325209444022,
"grad_norm": 1.1323665380477905,
"learning_rate": 4.5048257422091655e-06,
"loss": 1.209,
"step": 614
},
{
"epoch": 1.4044173648134044,
"grad_norm": 1.1112160682678223,
"learning_rate": 4.5029637166115e-06,
"loss": 1.1742,
"step": 615
},
{
"epoch": 1.4067022086824066,
"grad_norm": 1.1052254438400269,
"learning_rate": 4.5010985830024086e-06,
"loss": 1.1916,
"step": 616
},
{
"epoch": 1.408987052551409,
"grad_norm": 1.0695083141326904,
"learning_rate": 4.4992303442760286e-06,
"loss": 1.1829,
"step": 617
},
{
"epoch": 1.4112718964204114,
"grad_norm": 1.0871409177780151,
"learning_rate": 4.497359003331318e-06,
"loss": 1.2053,
"step": 618
},
{
"epoch": 1.4135567402894136,
"grad_norm": 1.093496322631836,
"learning_rate": 4.495484563072049e-06,
"loss": 1.1825,
"step": 619
},
{
"epoch": 1.4158415841584158,
"grad_norm": 1.0716845989227295,
"learning_rate": 4.493607026406802e-06,
"loss": 1.1911,
"step": 620
},
{
"epoch": 1.4181264280274182,
"grad_norm": 1.1274534463882446,
"learning_rate": 4.4917263962489635e-06,
"loss": 1.1737,
"step": 621
},
{
"epoch": 1.4204112718964204,
"grad_norm": 1.082309603691101,
"learning_rate": 4.489842675516718e-06,
"loss": 1.1986,
"step": 622
},
{
"epoch": 1.4226961157654228,
"grad_norm": 1.0890616178512573,
"learning_rate": 4.487955867133047e-06,
"loss": 1.2273,
"step": 623
},
{
"epoch": 1.424980959634425,
"grad_norm": 1.0633172988891602,
"learning_rate": 4.486065974025723e-06,
"loss": 1.1834,
"step": 624
},
{
"epoch": 1.4272658035034271,
"grad_norm": 1.0931994915008545,
"learning_rate": 4.484172999127305e-06,
"loss": 1.1976,
"step": 625
},
{
"epoch": 1.4295506473724295,
"grad_norm": 1.1375906467437744,
"learning_rate": 4.482276945375135e-06,
"loss": 1.2093,
"step": 626
},
{
"epoch": 1.431835491241432,
"grad_norm": 1.3243980407714844,
"learning_rate": 4.480377815711331e-06,
"loss": 1.2102,
"step": 627
},
{
"epoch": 1.4341203351104341,
"grad_norm": 1.0940284729003906,
"learning_rate": 4.478475613082783e-06,
"loss": 1.1888,
"step": 628
},
{
"epoch": 1.4364051789794363,
"grad_norm": 1.1363506317138672,
"learning_rate": 4.4765703404411534e-06,
"loss": 1.1833,
"step": 629
},
{
"epoch": 1.4386900228484387,
"grad_norm": 1.1287343502044678,
"learning_rate": 4.474662000742864e-06,
"loss": 1.2344,
"step": 630
},
{
"epoch": 1.440974866717441,
"grad_norm": 1.3280748128890991,
"learning_rate": 4.472750596949098e-06,
"loss": 1.2025,
"step": 631
},
{
"epoch": 1.4432597105864433,
"grad_norm": 1.1119890213012695,
"learning_rate": 4.470836132025793e-06,
"loss": 1.1586,
"step": 632
},
{
"epoch": 1.4455445544554455,
"grad_norm": 1.066416621208191,
"learning_rate": 4.4689186089436365e-06,
"loss": 1.1717,
"step": 633
},
{
"epoch": 1.447829398324448,
"grad_norm": 1.0481845140457153,
"learning_rate": 4.4669980306780605e-06,
"loss": 1.1949,
"step": 634
},
{
"epoch": 1.45011424219345,
"grad_norm": 1.094254732131958,
"learning_rate": 4.4650744002092384e-06,
"loss": 1.2005,
"step": 635
},
{
"epoch": 1.4523990860624525,
"grad_norm": 1.1029901504516602,
"learning_rate": 4.46314772052208e-06,
"loss": 1.1956,
"step": 636
},
{
"epoch": 1.4546839299314547,
"grad_norm": 1.129492163658142,
"learning_rate": 4.461217994606225e-06,
"loss": 1.2053,
"step": 637
},
{
"epoch": 1.4569687738004569,
"grad_norm": 1.1537097692489624,
"learning_rate": 4.459285225456044e-06,
"loss": 1.1668,
"step": 638
},
{
"epoch": 1.4592536176694593,
"grad_norm": 1.0732276439666748,
"learning_rate": 4.457349416070626e-06,
"loss": 1.2107,
"step": 639
},
{
"epoch": 1.4615384615384617,
"grad_norm": 1.186018943786621,
"learning_rate": 4.455410569453777e-06,
"loss": 1.1789,
"step": 640
},
{
"epoch": 1.4638233054074639,
"grad_norm": 1.1642903089523315,
"learning_rate": 4.453468688614019e-06,
"loss": 1.2169,
"step": 641
},
{
"epoch": 1.466108149276466,
"grad_norm": 1.1415472030639648,
"learning_rate": 4.451523776564581e-06,
"loss": 1.1716,
"step": 642
},
{
"epoch": 1.4683929931454685,
"grad_norm": 1.1644552946090698,
"learning_rate": 4.449575836323394e-06,
"loss": 1.1497,
"step": 643
},
{
"epoch": 1.4706778370144706,
"grad_norm": 1.2200912237167358,
"learning_rate": 4.447624870913091e-06,
"loss": 1.2289,
"step": 644
},
{
"epoch": 1.472962680883473,
"grad_norm": 1.1080158948898315,
"learning_rate": 4.445670883360996e-06,
"loss": 1.1378,
"step": 645
},
{
"epoch": 1.4752475247524752,
"grad_norm": 1.1372804641723633,
"learning_rate": 4.443713876699124e-06,
"loss": 1.1639,
"step": 646
},
{
"epoch": 1.4775323686214774,
"grad_norm": 1.1383754014968872,
"learning_rate": 4.441753853964174e-06,
"loss": 1.1558,
"step": 647
},
{
"epoch": 1.4798172124904798,
"grad_norm": 1.1565297842025757,
"learning_rate": 4.439790818197527e-06,
"loss": 1.242,
"step": 648
},
{
"epoch": 1.4821020563594822,
"grad_norm": 1.156384825706482,
"learning_rate": 4.4378247724452375e-06,
"loss": 1.2241,
"step": 649
},
{
"epoch": 1.4843869002284844,
"grad_norm": 1.2158401012420654,
"learning_rate": 4.43585571975803e-06,
"loss": 1.1977,
"step": 650
},
{
"epoch": 1.4866717440974866,
"grad_norm": 1.2885035276412964,
"learning_rate": 4.433883663191297e-06,
"loss": 1.1916,
"step": 651
},
{
"epoch": 1.488956587966489,
"grad_norm": 1.1530733108520508,
"learning_rate": 4.431908605805092e-06,
"loss": 1.2362,
"step": 652
},
{
"epoch": 1.4912414318354912,
"grad_norm": 1.1524220705032349,
"learning_rate": 4.429930550664121e-06,
"loss": 1.2263,
"step": 653
},
{
"epoch": 1.4935262757044936,
"grad_norm": 1.1547584533691406,
"learning_rate": 4.427949500837749e-06,
"loss": 1.1478,
"step": 654
},
{
"epoch": 1.4958111195734958,
"grad_norm": 1.1083858013153076,
"learning_rate": 4.425965459399979e-06,
"loss": 1.2183,
"step": 655
},
{
"epoch": 1.498095963442498,
"grad_norm": 1.130506992340088,
"learning_rate": 4.423978429429463e-06,
"loss": 1.1923,
"step": 656
},
{
"epoch": 1.5003808073115004,
"grad_norm": 1.106553077697754,
"learning_rate": 4.421988414009488e-06,
"loss": 1.192,
"step": 657
},
{
"epoch": 1.5026656511805028,
"grad_norm": 1.186131477355957,
"learning_rate": 4.419995416227973e-06,
"loss": 1.19,
"step": 658
},
{
"epoch": 1.504950495049505,
"grad_norm": 1.0756375789642334,
"learning_rate": 4.417999439177465e-06,
"loss": 1.1992,
"step": 659
},
{
"epoch": 1.5072353389185071,
"grad_norm": 1.1160579919815063,
"learning_rate": 4.416000485955135e-06,
"loss": 1.1747,
"step": 660
},
{
"epoch": 1.5095201827875095,
"grad_norm": 1.1178874969482422,
"learning_rate": 4.413998559662771e-06,
"loss": 1.1654,
"step": 661
},
{
"epoch": 1.511805026656512,
"grad_norm": 1.0919098854064941,
"learning_rate": 4.411993663406774e-06,
"loss": 1.2013,
"step": 662
},
{
"epoch": 1.5140898705255141,
"grad_norm": 1.058582067489624,
"learning_rate": 4.409985800298155e-06,
"loss": 1.1823,
"step": 663
},
{
"epoch": 1.5163747143945163,
"grad_norm": 1.0792784690856934,
"learning_rate": 4.407974973452527e-06,
"loss": 1.2013,
"step": 664
},
{
"epoch": 1.5186595582635185,
"grad_norm": 1.112774133682251,
"learning_rate": 4.405961185990103e-06,
"loss": 1.2005,
"step": 665
},
{
"epoch": 1.520944402132521,
"grad_norm": 1.1190800666809082,
"learning_rate": 4.403944441035691e-06,
"loss": 1.2146,
"step": 666
},
{
"epoch": 1.5232292460015233,
"grad_norm": 1.1045669317245483,
"learning_rate": 4.401924741718685e-06,
"loss": 1.2217,
"step": 667
},
{
"epoch": 1.5255140898705255,
"grad_norm": 1.1048752069473267,
"learning_rate": 4.399902091173065e-06,
"loss": 1.1944,
"step": 668
},
{
"epoch": 1.5277989337395277,
"grad_norm": 1.0909706354141235,
"learning_rate": 4.397876492537392e-06,
"loss": 1.2058,
"step": 669
},
{
"epoch": 1.53008377760853,
"grad_norm": 1.1354328393936157,
"learning_rate": 4.3958479489548e-06,
"loss": 1.2164,
"step": 670
},
{
"epoch": 1.5323686214775325,
"grad_norm": 1.1490201950073242,
"learning_rate": 4.393816463572993e-06,
"loss": 1.182,
"step": 671
},
{
"epoch": 1.5346534653465347,
"grad_norm": 1.1395319700241089,
"learning_rate": 4.391782039544239e-06,
"loss": 1.2201,
"step": 672
},
{
"epoch": 1.5369383092155369,
"grad_norm": 1.0788644552230835,
"learning_rate": 4.389744680025366e-06,
"loss": 1.2212,
"step": 673
},
{
"epoch": 1.5392231530845393,
"grad_norm": 1.0663102865219116,
"learning_rate": 4.387704388177759e-06,
"loss": 1.1872,
"step": 674
},
{
"epoch": 1.5415079969535415,
"grad_norm": 1.1177600622177124,
"learning_rate": 4.3856611671673505e-06,
"loss": 1.2032,
"step": 675
},
{
"epoch": 1.5437928408225439,
"grad_norm": 1.1109418869018555,
"learning_rate": 4.383615020164621e-06,
"loss": 1.2041,
"step": 676
},
{
"epoch": 1.546077684691546,
"grad_norm": 1.096182107925415,
"learning_rate": 4.3815659503445875e-06,
"loss": 1.1988,
"step": 677
},
{
"epoch": 1.5483625285605482,
"grad_norm": 1.2027829885482788,
"learning_rate": 4.379513960886807e-06,
"loss": 1.1812,
"step": 678
},
{
"epoch": 1.5506473724295506,
"grad_norm": 1.0674421787261963,
"learning_rate": 4.377459054975363e-06,
"loss": 1.1948,
"step": 679
},
{
"epoch": 1.552932216298553,
"grad_norm": 1.0463448762893677,
"learning_rate": 4.375401235798866e-06,
"loss": 1.2174,
"step": 680
},
{
"epoch": 1.5552170601675552,
"grad_norm": 1.1295356750488281,
"learning_rate": 4.373340506550447e-06,
"loss": 1.2013,
"step": 681
},
{
"epoch": 1.5575019040365574,
"grad_norm": 1.116245150566101,
"learning_rate": 4.3712768704277535e-06,
"loss": 1.1983,
"step": 682
},
{
"epoch": 1.5597867479055598,
"grad_norm": 1.1043322086334229,
"learning_rate": 4.369210330632942e-06,
"loss": 1.2042,
"step": 683
},
{
"epoch": 1.5620715917745622,
"grad_norm": 1.1363909244537354,
"learning_rate": 4.367140890372674e-06,
"loss": 1.1793,
"step": 684
},
{
"epoch": 1.5643564356435644,
"grad_norm": 1.099576473236084,
"learning_rate": 4.365068552858116e-06,
"loss": 1.1849,
"step": 685
},
{
"epoch": 1.5666412795125666,
"grad_norm": 1.0956041812896729,
"learning_rate": 4.3629933213049245e-06,
"loss": 1.169,
"step": 686
},
{
"epoch": 1.5689261233815688,
"grad_norm": 1.1022474765777588,
"learning_rate": 4.36091519893325e-06,
"loss": 1.2376,
"step": 687
},
{
"epoch": 1.5712109672505712,
"grad_norm": 1.1013610363006592,
"learning_rate": 4.35883418896773e-06,
"loss": 1.1665,
"step": 688
},
{
"epoch": 1.5734958111195736,
"grad_norm": 1.1273926496505737,
"learning_rate": 4.356750294637478e-06,
"loss": 1.1723,
"step": 689
},
{
"epoch": 1.5757806549885758,
"grad_norm": 1.1341313123703003,
"learning_rate": 4.3546635191760875e-06,
"loss": 1.1813,
"step": 690
},
{
"epoch": 1.578065498857578,
"grad_norm": 1.0935291051864624,
"learning_rate": 4.352573865821621e-06,
"loss": 1.1932,
"step": 691
},
{
"epoch": 1.5803503427265804,
"grad_norm": 1.189144492149353,
"learning_rate": 4.350481337816606e-06,
"loss": 1.1798,
"step": 692
},
{
"epoch": 1.5826351865955828,
"grad_norm": 1.2104555368423462,
"learning_rate": 4.348385938408033e-06,
"loss": 1.1895,
"step": 693
},
{
"epoch": 1.584920030464585,
"grad_norm": 1.1346546411514282,
"learning_rate": 4.346287670847345e-06,
"loss": 1.1896,
"step": 694
},
{
"epoch": 1.5872048743335871,
"grad_norm": 1.131858468055725,
"learning_rate": 4.344186538390438e-06,
"loss": 1.1895,
"step": 695
},
{
"epoch": 1.5894897182025893,
"grad_norm": 1.1175106763839722,
"learning_rate": 4.342082544297652e-06,
"loss": 1.2215,
"step": 696
},
{
"epoch": 1.5917745620715917,
"grad_norm": 1.1056832075119019,
"learning_rate": 4.3399756918337675e-06,
"loss": 1.1617,
"step": 697
},
{
"epoch": 1.5940594059405941,
"grad_norm": 1.0777372121810913,
"learning_rate": 4.337865984268002e-06,
"loss": 1.2716,
"step": 698
},
{
"epoch": 1.5963442498095963,
"grad_norm": 1.1472179889678955,
"learning_rate": 4.335753424874e-06,
"loss": 1.1842,
"step": 699
},
{
"epoch": 1.5986290936785985,
"grad_norm": 1.0825904607772827,
"learning_rate": 4.333638016929835e-06,
"loss": 1.2377,
"step": 700
},
{
"epoch": 1.600913937547601,
"grad_norm": 1.0884586572647095,
"learning_rate": 4.331519763717998e-06,
"loss": 1.2003,
"step": 701
},
{
"epoch": 1.6031987814166033,
"grad_norm": 1.132827639579773,
"learning_rate": 4.329398668525396e-06,
"loss": 1.186,
"step": 702
},
{
"epoch": 1.6054836252856055,
"grad_norm": 1.1931272745132446,
"learning_rate": 4.327274734643346e-06,
"loss": 1.21,
"step": 703
},
{
"epoch": 1.6077684691546077,
"grad_norm": 1.060774803161621,
"learning_rate": 4.3251479653675705e-06,
"loss": 1.1893,
"step": 704
},
{
"epoch": 1.61005331302361,
"grad_norm": 1.0615568161010742,
"learning_rate": 4.323018363998189e-06,
"loss": 1.1814,
"step": 705
},
{
"epoch": 1.6123381568926123,
"grad_norm": 1.0800261497497559,
"learning_rate": 4.320885933839718e-06,
"loss": 1.1995,
"step": 706
},
{
"epoch": 1.6146230007616147,
"grad_norm": 1.1502355337142944,
"learning_rate": 4.318750678201064e-06,
"loss": 1.219,
"step": 707
},
{
"epoch": 1.6169078446306169,
"grad_norm": 1.07515287399292,
"learning_rate": 4.316612600395515e-06,
"loss": 1.1787,
"step": 708
},
{
"epoch": 1.619192688499619,
"grad_norm": 1.075150728225708,
"learning_rate": 4.31447170374074e-06,
"loss": 1.2062,
"step": 709
},
{
"epoch": 1.6214775323686215,
"grad_norm": 1.1458789110183716,
"learning_rate": 4.312327991558782e-06,
"loss": 1.2009,
"step": 710
},
{
"epoch": 1.6237623762376239,
"grad_norm": 1.1655830144882202,
"learning_rate": 4.3101814671760546e-06,
"loss": 1.2004,
"step": 711
},
{
"epoch": 1.626047220106626,
"grad_norm": 1.1394225358963013,
"learning_rate": 4.30803213392333e-06,
"loss": 1.2059,
"step": 712
},
{
"epoch": 1.6283320639756282,
"grad_norm": 1.109095573425293,
"learning_rate": 4.305879995135745e-06,
"loss": 1.1727,
"step": 713
},
{
"epoch": 1.6306169078446306,
"grad_norm": 1.072530746459961,
"learning_rate": 4.303725054152785e-06,
"loss": 1.2059,
"step": 714
},
{
"epoch": 1.632901751713633,
"grad_norm": 1.079399824142456,
"learning_rate": 4.3015673143182864e-06,
"loss": 1.1929,
"step": 715
},
{
"epoch": 1.6351865955826352,
"grad_norm": 1.1387250423431396,
"learning_rate": 4.299406778980428e-06,
"loss": 1.1924,
"step": 716
},
{
"epoch": 1.6374714394516374,
"grad_norm": 1.1101268529891968,
"learning_rate": 4.297243451491724e-06,
"loss": 1.1678,
"step": 717
},
{
"epoch": 1.6397562833206396,
"grad_norm": 1.0745279788970947,
"learning_rate": 4.295077335209027e-06,
"loss": 1.1632,
"step": 718
},
{
"epoch": 1.642041127189642,
"grad_norm": 1.1062053442001343,
"learning_rate": 4.29290843349351e-06,
"loss": 1.183,
"step": 719
},
{
"epoch": 1.6443259710586444,
"grad_norm": 1.1918151378631592,
"learning_rate": 4.290736749710672e-06,
"loss": 1.2,
"step": 720
},
{
"epoch": 1.6466108149276466,
"grad_norm": 1.1596369743347168,
"learning_rate": 4.28856228723033e-06,
"loss": 1.2232,
"step": 721
},
{
"epoch": 1.6488956587966488,
"grad_norm": 1.1261903047561646,
"learning_rate": 4.28638504942661e-06,
"loss": 1.1614,
"step": 722
},
{
"epoch": 1.6511805026656512,
"grad_norm": 1.1628916263580322,
"learning_rate": 4.284205039677946e-06,
"loss": 1.1866,
"step": 723
},
{
"epoch": 1.6534653465346536,
"grad_norm": 1.114009976387024,
"learning_rate": 4.282022261367074e-06,
"loss": 1.2027,
"step": 724
},
{
"epoch": 1.6557501904036558,
"grad_norm": 1.2100892066955566,
"learning_rate": 4.279836717881022e-06,
"loss": 1.1922,
"step": 725
},
{
"epoch": 1.658035034272658,
"grad_norm": 1.122144103050232,
"learning_rate": 4.277648412611114e-06,
"loss": 1.178,
"step": 726
},
{
"epoch": 1.6603198781416602,
"grad_norm": 1.0997899770736694,
"learning_rate": 4.275457348952955e-06,
"loss": 1.2276,
"step": 727
},
{
"epoch": 1.6626047220106626,
"grad_norm": 1.1784963607788086,
"learning_rate": 4.273263530306435e-06,
"loss": 1.1889,
"step": 728
},
{
"epoch": 1.664889565879665,
"grad_norm": 1.1602816581726074,
"learning_rate": 4.271066960075715e-06,
"loss": 1.1671,
"step": 729
},
{
"epoch": 1.6671744097486672,
"grad_norm": 1.132406234741211,
"learning_rate": 4.268867641669225e-06,
"loss": 1.2017,
"step": 730
},
{
"epoch": 1.6694592536176693,
"grad_norm": 1.1267459392547607,
"learning_rate": 4.266665578499664e-06,
"loss": 1.2135,
"step": 731
},
{
"epoch": 1.6717440974866717,
"grad_norm": 1.0951600074768066,
"learning_rate": 4.2644607739839875e-06,
"loss": 1.2463,
"step": 732
},
{
"epoch": 1.6740289413556741,
"grad_norm": 1.073875069618225,
"learning_rate": 4.262253231543401e-06,
"loss": 1.1879,
"step": 733
},
{
"epoch": 1.6763137852246763,
"grad_norm": 1.1010491847991943,
"learning_rate": 4.260042954603366e-06,
"loss": 1.1812,
"step": 734
},
{
"epoch": 1.6785986290936785,
"grad_norm": 1.1129403114318848,
"learning_rate": 4.2578299465935805e-06,
"loss": 1.2281,
"step": 735
},
{
"epoch": 1.680883472962681,
"grad_norm": 1.3570629358291626,
"learning_rate": 4.255614210947985e-06,
"loss": 1.2013,
"step": 736
},
{
"epoch": 1.6831683168316833,
"grad_norm": 1.104535460472107,
"learning_rate": 4.2533957511047485e-06,
"loss": 1.1708,
"step": 737
},
{
"epoch": 1.6854531607006855,
"grad_norm": 1.1040364503860474,
"learning_rate": 4.25117457050627e-06,
"loss": 1.2154,
"step": 738
},
{
"epoch": 1.6877380045696877,
"grad_norm": 1.0932183265686035,
"learning_rate": 4.24895067259917e-06,
"loss": 1.2182,
"step": 739
},
{
"epoch": 1.6900228484386899,
"grad_norm": 1.0946629047393799,
"learning_rate": 4.246724060834284e-06,
"loss": 1.2058,
"step": 740
},
{
"epoch": 1.6923076923076923,
"grad_norm": 1.060051679611206,
"learning_rate": 4.24449473866666e-06,
"loss": 1.1919,
"step": 741
},
{
"epoch": 1.6945925361766947,
"grad_norm": 1.0483872890472412,
"learning_rate": 4.242262709555552e-06,
"loss": 1.1638,
"step": 742
},
{
"epoch": 1.6968773800456969,
"grad_norm": 1.0956077575683594,
"learning_rate": 4.240027976964412e-06,
"loss": 1.1805,
"step": 743
},
{
"epoch": 1.699162223914699,
"grad_norm": 1.1185762882232666,
"learning_rate": 4.237790544360889e-06,
"loss": 1.1923,
"step": 744
},
{
"epoch": 1.7014470677837015,
"grad_norm": 1.0720428228378296,
"learning_rate": 4.2355504152168235e-06,
"loss": 1.1895,
"step": 745
},
{
"epoch": 1.7037319116527039,
"grad_norm": 1.1020833253860474,
"learning_rate": 4.2333075930082345e-06,
"loss": 1.1845,
"step": 746
},
{
"epoch": 1.706016755521706,
"grad_norm": 1.107071876525879,
"learning_rate": 4.231062081215326e-06,
"loss": 1.1751,
"step": 747
},
{
"epoch": 1.7083015993907082,
"grad_norm": 1.1301578283309937,
"learning_rate": 4.228813883322472e-06,
"loss": 1.151,
"step": 748
},
{
"epoch": 1.7105864432597104,
"grad_norm": 1.096433401107788,
"learning_rate": 4.226563002818215e-06,
"loss": 1.1728,
"step": 749
},
{
"epoch": 1.7128712871287128,
"grad_norm": 1.1047178506851196,
"learning_rate": 4.224309443195261e-06,
"loss": 1.1947,
"step": 750
},
{
"epoch": 1.7151561309977152,
"grad_norm": 1.110417127609253,
"learning_rate": 4.222053207950472e-06,
"loss": 1.2186,
"step": 751
},
{
"epoch": 1.7174409748667174,
"grad_norm": 1.1027036905288696,
"learning_rate": 4.219794300584863e-06,
"loss": 1.2164,
"step": 752
},
{
"epoch": 1.7197258187357196,
"grad_norm": 1.109299898147583,
"learning_rate": 4.217532724603595e-06,
"loss": 1.2202,
"step": 753
},
{
"epoch": 1.722010662604722,
"grad_norm": 1.143134593963623,
"learning_rate": 4.2152684835159695e-06,
"loss": 1.1837,
"step": 754
},
{
"epoch": 1.7242955064737244,
"grad_norm": 1.0689709186553955,
"learning_rate": 4.213001580835423e-06,
"loss": 1.1874,
"step": 755
},
{
"epoch": 1.7265803503427266,
"grad_norm": 1.128243327140808,
"learning_rate": 4.2107320200795236e-06,
"loss": 1.1756,
"step": 756
},
{
"epoch": 1.7288651942117288,
"grad_norm": 1.067436933517456,
"learning_rate": 4.208459804769963e-06,
"loss": 1.2212,
"step": 757
},
{
"epoch": 1.7311500380807312,
"grad_norm": 1.1413803100585938,
"learning_rate": 4.206184938432552e-06,
"loss": 1.1491,
"step": 758
},
{
"epoch": 1.7334348819497334,
"grad_norm": 1.141803503036499,
"learning_rate": 4.203907424597214e-06,
"loss": 1.2636,
"step": 759
},
{
"epoch": 1.7357197258187358,
"grad_norm": 1.1099580526351929,
"learning_rate": 4.2016272667979814e-06,
"loss": 1.2192,
"step": 760
},
{
"epoch": 1.738004569687738,
"grad_norm": 1.100486397743225,
"learning_rate": 4.199344468572992e-06,
"loss": 1.2044,
"step": 761
},
{
"epoch": 1.7402894135567402,
"grad_norm": 1.0598186254501343,
"learning_rate": 4.197059033464476e-06,
"loss": 1.1983,
"step": 762
},
{
"epoch": 1.7425742574257426,
"grad_norm": 1.0922818183898926,
"learning_rate": 4.194770965018758e-06,
"loss": 1.2194,
"step": 763
},
{
"epoch": 1.744859101294745,
"grad_norm": 1.0975127220153809,
"learning_rate": 4.1924802667862485e-06,
"loss": 1.1465,
"step": 764
},
{
"epoch": 1.7471439451637472,
"grad_norm": 1.0934858322143555,
"learning_rate": 4.190186942321438e-06,
"loss": 1.1544,
"step": 765
},
{
"epoch": 1.7494287890327493,
"grad_norm": 1.0712271928787231,
"learning_rate": 4.187890995182893e-06,
"loss": 1.1893,
"step": 766
},
{
"epoch": 1.7517136329017517,
"grad_norm": 1.1157736778259277,
"learning_rate": 4.1855924289332485e-06,
"loss": 1.2362,
"step": 767
},
{
"epoch": 1.7539984767707542,
"grad_norm": 1.2225691080093384,
"learning_rate": 4.183291247139204e-06,
"loss": 1.22,
"step": 768
},
{
"epoch": 1.7562833206397563,
"grad_norm": 1.1402082443237305,
"learning_rate": 4.180987453371519e-06,
"loss": 1.2024,
"step": 769
},
{
"epoch": 1.7585681645087585,
"grad_norm": 1.1124638319015503,
"learning_rate": 4.178681051205004e-06,
"loss": 1.17,
"step": 770
},
{
"epoch": 1.7608530083777607,
"grad_norm": 1.1337512731552124,
"learning_rate": 4.176372044218519e-06,
"loss": 1.1862,
"step": 771
},
{
"epoch": 1.7631378522467631,
"grad_norm": 1.0702142715454102,
"learning_rate": 4.174060435994962e-06,
"loss": 1.2038,
"step": 772
},
{
"epoch": 1.7654226961157655,
"grad_norm": 1.112242579460144,
"learning_rate": 4.171746230121273e-06,
"loss": 1.2146,
"step": 773
},
{
"epoch": 1.7677075399847677,
"grad_norm": 1.1164225339889526,
"learning_rate": 4.169429430188418e-06,
"loss": 1.1768,
"step": 774
},
{
"epoch": 1.76999238385377,
"grad_norm": 1.091208577156067,
"learning_rate": 4.16711003979139e-06,
"loss": 1.193,
"step": 775
},
{
"epoch": 1.7722772277227723,
"grad_norm": 1.138411283493042,
"learning_rate": 4.164788062529203e-06,
"loss": 1.203,
"step": 776
},
{
"epoch": 1.7745620715917747,
"grad_norm": 1.168305516242981,
"learning_rate": 4.1624635020048835e-06,
"loss": 1.2154,
"step": 777
},
{
"epoch": 1.7768469154607769,
"grad_norm": 1.0742619037628174,
"learning_rate": 4.160136361825465e-06,
"loss": 1.214,
"step": 778
},
{
"epoch": 1.779131759329779,
"grad_norm": 1.076762080192566,
"learning_rate": 4.1578066456019885e-06,
"loss": 1.1834,
"step": 779
},
{
"epoch": 1.7814166031987813,
"grad_norm": 1.1189744472503662,
"learning_rate": 4.155474356949487e-06,
"loss": 1.191,
"step": 780
},
{
"epoch": 1.7837014470677837,
"grad_norm": 1.0916801691055298,
"learning_rate": 4.153139499486988e-06,
"loss": 1.2104,
"step": 781
},
{
"epoch": 1.785986290936786,
"grad_norm": 1.1265934705734253,
"learning_rate": 4.150802076837506e-06,
"loss": 1.2366,
"step": 782
},
{
"epoch": 1.7882711348057883,
"grad_norm": 1.1008100509643555,
"learning_rate": 4.148462092628032e-06,
"loss": 1.1919,
"step": 783
},
{
"epoch": 1.7905559786747904,
"grad_norm": 1.5858978033065796,
"learning_rate": 4.146119550489536e-06,
"loss": 1.1927,
"step": 784
},
{
"epoch": 1.7928408225437928,
"grad_norm": 1.1155521869659424,
"learning_rate": 4.143774454056954e-06,
"loss": 1.1948,
"step": 785
},
{
"epoch": 1.7951256664127953,
"grad_norm": 1.1289353370666504,
"learning_rate": 4.141426806969189e-06,
"loss": 1.1719,
"step": 786
},
{
"epoch": 1.7974105102817974,
"grad_norm": 1.1492801904678345,
"learning_rate": 4.139076612869098e-06,
"loss": 1.169,
"step": 787
},
{
"epoch": 1.7996953541507996,
"grad_norm": 1.0931838750839233,
"learning_rate": 4.1367238754034935e-06,
"loss": 1.1581,
"step": 788
},
{
"epoch": 1.801980198019802,
"grad_norm": 1.0901176929473877,
"learning_rate": 4.134368598223132e-06,
"loss": 1.2223,
"step": 789
},
{
"epoch": 1.8042650418888042,
"grad_norm": 1.0907678604125977,
"learning_rate": 4.132010784982711e-06,
"loss": 1.1839,
"step": 790
},
{
"epoch": 1.8065498857578066,
"grad_norm": 1.1389234066009521,
"learning_rate": 4.129650439340866e-06,
"loss": 1.1765,
"step": 791
},
{
"epoch": 1.8088347296268088,
"grad_norm": 1.0889054536819458,
"learning_rate": 4.12728756496016e-06,
"loss": 1.1913,
"step": 792
},
{
"epoch": 1.811119573495811,
"grad_norm": 1.090705156326294,
"learning_rate": 4.12492216550708e-06,
"loss": 1.1692,
"step": 793
},
{
"epoch": 1.8134044173648134,
"grad_norm": 1.1290946006774902,
"learning_rate": 4.12255424465203e-06,
"loss": 1.2114,
"step": 794
},
{
"epoch": 1.8156892612338158,
"grad_norm": 1.108325719833374,
"learning_rate": 4.120183806069328e-06,
"loss": 1.1941,
"step": 795
},
{
"epoch": 1.817974105102818,
"grad_norm": 1.0901302099227905,
"learning_rate": 4.1178108534371995e-06,
"loss": 1.1709,
"step": 796
},
{
"epoch": 1.8202589489718202,
"grad_norm": 1.1386867761611938,
"learning_rate": 4.11543539043777e-06,
"loss": 1.2008,
"step": 797
},
{
"epoch": 1.8225437928408226,
"grad_norm": 1.1768696308135986,
"learning_rate": 4.11305742075706e-06,
"loss": 1.1735,
"step": 798
},
{
"epoch": 1.824828636709825,
"grad_norm": 1.093137264251709,
"learning_rate": 4.1106769480849795e-06,
"loss": 1.1952,
"step": 799
},
{
"epoch": 1.8271134805788272,
"grad_norm": 1.12264084815979,
"learning_rate": 4.108293976115325e-06,
"loss": 1.2118,
"step": 800
},
{
"epoch": 1.8293983244478293,
"grad_norm": 1.089824914932251,
"learning_rate": 4.105908508545766e-06,
"loss": 1.1856,
"step": 801
},
{
"epoch": 1.8316831683168315,
"grad_norm": 1.3425248861312866,
"learning_rate": 4.1035205490778505e-06,
"loss": 1.1942,
"step": 802
},
{
"epoch": 1.833968012185834,
"grad_norm": 1.097115159034729,
"learning_rate": 4.101130101416988e-06,
"loss": 1.2083,
"step": 803
},
{
"epoch": 1.8362528560548363,
"grad_norm": 1.0973830223083496,
"learning_rate": 4.098737169272452e-06,
"loss": 1.2033,
"step": 804
},
{
"epoch": 1.8385376999238385,
"grad_norm": 1.127233624458313,
"learning_rate": 4.096341756357371e-06,
"loss": 1.1941,
"step": 805
},
{
"epoch": 1.8408225437928407,
"grad_norm": 1.097070574760437,
"learning_rate": 4.093943866388723e-06,
"loss": 1.1971,
"step": 806
},
{
"epoch": 1.8431073876618431,
"grad_norm": 1.0978144407272339,
"learning_rate": 4.091543503087327e-06,
"loss": 1.2029,
"step": 807
},
{
"epoch": 1.8453922315308455,
"grad_norm": 1.0643872022628784,
"learning_rate": 4.089140670177843e-06,
"loss": 1.1532,
"step": 808
},
{
"epoch": 1.8476770753998477,
"grad_norm": 1.1128400564193726,
"learning_rate": 4.086735371388762e-06,
"loss": 1.1851,
"step": 809
},
{
"epoch": 1.84996191926885,
"grad_norm": 1.1439098119735718,
"learning_rate": 4.0843276104524e-06,
"loss": 1.1816,
"step": 810
},
{
"epoch": 1.852246763137852,
"grad_norm": 1.1020617485046387,
"learning_rate": 4.0819173911048965e-06,
"loss": 1.2081,
"step": 811
},
{
"epoch": 1.8545316070068545,
"grad_norm": 1.0913503170013428,
"learning_rate": 4.079504717086203e-06,
"loss": 1.1892,
"step": 812
},
{
"epoch": 1.856816450875857,
"grad_norm": 1.1332125663757324,
"learning_rate": 4.077089592140082e-06,
"loss": 1.182,
"step": 813
},
{
"epoch": 1.859101294744859,
"grad_norm": 1.0584102869033813,
"learning_rate": 4.074672020014098e-06,
"loss": 1.2169,
"step": 814
},
{
"epoch": 1.8613861386138613,
"grad_norm": 1.1291085481643677,
"learning_rate": 4.072252004459612e-06,
"loss": 1.1796,
"step": 815
},
{
"epoch": 1.8636709824828637,
"grad_norm": 1.0868487358093262,
"learning_rate": 4.069829549231778e-06,
"loss": 1.1832,
"step": 816
},
{
"epoch": 1.865955826351866,
"grad_norm": 1.1105777025222778,
"learning_rate": 4.067404658089535e-06,
"loss": 1.2242,
"step": 817
},
{
"epoch": 1.8682406702208683,
"grad_norm": 1.1203657388687134,
"learning_rate": 4.0649773347956005e-06,
"loss": 1.1755,
"step": 818
},
{
"epoch": 1.8705255140898704,
"grad_norm": 1.1000312566757202,
"learning_rate": 4.062547583116469e-06,
"loss": 1.1829,
"step": 819
},
{
"epoch": 1.8728103579588729,
"grad_norm": 1.1125813722610474,
"learning_rate": 4.060115406822402e-06,
"loss": 1.2013,
"step": 820
},
{
"epoch": 1.8750952018278753,
"grad_norm": 1.0868099927902222,
"learning_rate": 4.057680809687421e-06,
"loss": 1.1749,
"step": 821
},
{
"epoch": 1.8773800456968774,
"grad_norm": 1.1242718696594238,
"learning_rate": 4.055243795489307e-06,
"loss": 1.1601,
"step": 822
},
{
"epoch": 1.8796648895658796,
"grad_norm": 1.1220780611038208,
"learning_rate": 4.052804368009589e-06,
"loss": 1.197,
"step": 823
},
{
"epoch": 1.8819497334348818,
"grad_norm": 1.0715032815933228,
"learning_rate": 4.050362531033545e-06,
"loss": 1.1834,
"step": 824
},
{
"epoch": 1.8842345773038842,
"grad_norm": 1.1281424760818481,
"learning_rate": 4.0479182883501855e-06,
"loss": 1.1653,
"step": 825
},
{
"epoch": 1.8865194211728866,
"grad_norm": 1.0727750062942505,
"learning_rate": 4.045471643752258e-06,
"loss": 1.1907,
"step": 826
},
{
"epoch": 1.8888042650418888,
"grad_norm": 1.122889757156372,
"learning_rate": 4.043022601036238e-06,
"loss": 1.1935,
"step": 827
},
{
"epoch": 1.891089108910891,
"grad_norm": 1.1605268716812134,
"learning_rate": 4.040571164002319e-06,
"loss": 1.211,
"step": 828
},
{
"epoch": 1.8933739527798934,
"grad_norm": 1.0915296077728271,
"learning_rate": 4.038117336454411e-06,
"loss": 1.1614,
"step": 829
},
{
"epoch": 1.8956587966488958,
"grad_norm": 1.1206241846084595,
"learning_rate": 4.035661122200135e-06,
"loss": 1.1592,
"step": 830
},
{
"epoch": 1.897943640517898,
"grad_norm": 1.1132665872573853,
"learning_rate": 4.033202525050813e-06,
"loss": 1.1865,
"step": 831
},
{
"epoch": 1.9002284843869002,
"grad_norm": 1.0694245100021362,
"learning_rate": 4.0307415488214675e-06,
"loss": 1.1767,
"step": 832
},
{
"epoch": 1.9025133282559024,
"grad_norm": 1.1312873363494873,
"learning_rate": 4.028278197330808e-06,
"loss": 1.2344,
"step": 833
},
{
"epoch": 1.9047981721249048,
"grad_norm": 1.0864746570587158,
"learning_rate": 4.025812474401236e-06,
"loss": 1.2146,
"step": 834
},
{
"epoch": 1.9070830159939072,
"grad_norm": 1.0774086713790894,
"learning_rate": 4.023344383858826e-06,
"loss": 1.1496,
"step": 835
},
{
"epoch": 1.9093678598629094,
"grad_norm": 1.0805225372314453,
"learning_rate": 4.0208739295333314e-06,
"loss": 1.2098,
"step": 836
},
{
"epoch": 1.9116527037319115,
"grad_norm": 1.0994813442230225,
"learning_rate": 4.018401115258172e-06,
"loss": 1.1881,
"step": 837
},
{
"epoch": 1.913937547600914,
"grad_norm": 1.0419007539749146,
"learning_rate": 4.015925944870428e-06,
"loss": 1.1935,
"step": 838
},
{
"epoch": 1.9162223914699164,
"grad_norm": 1.1358449459075928,
"learning_rate": 4.013448422210838e-06,
"loss": 1.1989,
"step": 839
},
{
"epoch": 1.9185072353389185,
"grad_norm": 1.121999740600586,
"learning_rate": 4.010968551123788e-06,
"loss": 1.2108,
"step": 840
},
{
"epoch": 1.9207920792079207,
"grad_norm": 1.1504106521606445,
"learning_rate": 4.008486335457312e-06,
"loss": 1.1768,
"step": 841
},
{
"epoch": 1.9230769230769231,
"grad_norm": 1.128138780593872,
"learning_rate": 4.006001779063078e-06,
"loss": 1.1992,
"step": 842
},
{
"epoch": 1.9253617669459253,
"grad_norm": 1.1590732336044312,
"learning_rate": 4.003514885796388e-06,
"loss": 1.181,
"step": 843
},
{
"epoch": 1.9276466108149277,
"grad_norm": 1.0851722955703735,
"learning_rate": 4.001025659516171e-06,
"loss": 1.1711,
"step": 844
},
{
"epoch": 1.92993145468393,
"grad_norm": 1.066331148147583,
"learning_rate": 3.998534104084974e-06,
"loss": 1.1728,
"step": 845
},
{
"epoch": 1.932216298552932,
"grad_norm": 1.110464096069336,
"learning_rate": 3.99604022336896e-06,
"loss": 1.178,
"step": 846
},
{
"epoch": 1.9345011424219345,
"grad_norm": 1.1028679609298706,
"learning_rate": 3.993544021237899e-06,
"loss": 1.2122,
"step": 847
},
{
"epoch": 1.936785986290937,
"grad_norm": 1.1760601997375488,
"learning_rate": 3.991045501565163e-06,
"loss": 1.2103,
"step": 848
},
{
"epoch": 1.939070830159939,
"grad_norm": 1.1260336637496948,
"learning_rate": 3.988544668227721e-06,
"loss": 1.1443,
"step": 849
},
{
"epoch": 1.9413556740289413,
"grad_norm": 1.1055935621261597,
"learning_rate": 3.9860415251061334e-06,
"loss": 1.1795,
"step": 850
},
{
"epoch": 1.9436405178979437,
"grad_norm": 1.1292855739593506,
"learning_rate": 3.983536076084541e-06,
"loss": 1.182,
"step": 851
},
{
"epoch": 1.945925361766946,
"grad_norm": 1.1108464002609253,
"learning_rate": 3.981028325050667e-06,
"loss": 1.1876,
"step": 852
},
{
"epoch": 1.9482102056359483,
"grad_norm": 1.1306401491165161,
"learning_rate": 3.978518275895802e-06,
"loss": 1.1645,
"step": 853
},
{
"epoch": 1.9504950495049505,
"grad_norm": 1.1031887531280518,
"learning_rate": 3.976005932514807e-06,
"loss": 1.2047,
"step": 854
},
{
"epoch": 1.9527798933739526,
"grad_norm": 1.0953725576400757,
"learning_rate": 3.973491298806101e-06,
"loss": 1.1756,
"step": 855
},
{
"epoch": 1.955064737242955,
"grad_norm": 1.109553575515747,
"learning_rate": 3.970974378671656e-06,
"loss": 1.2228,
"step": 856
},
{
"epoch": 1.9573495811119574,
"grad_norm": 1.1159707307815552,
"learning_rate": 3.968455176016993e-06,
"loss": 1.2037,
"step": 857
},
{
"epoch": 1.9596344249809596,
"grad_norm": 1.1045714616775513,
"learning_rate": 3.965933694751175e-06,
"loss": 1.196,
"step": 858
},
{
"epoch": 1.9619192688499618,
"grad_norm": 1.110876202583313,
"learning_rate": 3.963409938786801e-06,
"loss": 1.1772,
"step": 859
},
{
"epoch": 1.9642041127189642,
"grad_norm": 1.1226321458816528,
"learning_rate": 3.9608839120399975e-06,
"loss": 1.1875,
"step": 860
},
{
"epoch": 1.9664889565879666,
"grad_norm": 1.1401004791259766,
"learning_rate": 3.958355618430417e-06,
"loss": 1.2137,
"step": 861
},
{
"epoch": 1.9687738004569688,
"grad_norm": 1.0866281986236572,
"learning_rate": 3.95582506188123e-06,
"loss": 1.2001,
"step": 862
},
{
"epoch": 1.971058644325971,
"grad_norm": 1.1426069736480713,
"learning_rate": 3.9532922463191145e-06,
"loss": 1.1794,
"step": 863
},
{
"epoch": 1.9733434881949732,
"grad_norm": 1.1191396713256836,
"learning_rate": 3.950757175674257e-06,
"loss": 1.2118,
"step": 864
},
{
"epoch": 1.9756283320639756,
"grad_norm": 1.0993397235870361,
"learning_rate": 3.948219853880344e-06,
"loss": 1.2209,
"step": 865
},
{
"epoch": 1.977913175932978,
"grad_norm": 1.0973010063171387,
"learning_rate": 3.945680284874553e-06,
"loss": 1.1738,
"step": 866
},
{
"epoch": 1.9801980198019802,
"grad_norm": 1.2131692171096802,
"learning_rate": 3.943138472597549e-06,
"loss": 1.1833,
"step": 867
},
{
"epoch": 1.9824828636709824,
"grad_norm": 1.1128953695297241,
"learning_rate": 3.940594420993479e-06,
"loss": 1.1925,
"step": 868
},
{
"epoch": 1.9847677075399848,
"grad_norm": 1.0862925052642822,
"learning_rate": 3.938048134009962e-06,
"loss": 1.1965,
"step": 869
},
{
"epoch": 1.9870525514089872,
"grad_norm": 1.1464707851409912,
"learning_rate": 3.935499615598088e-06,
"loss": 1.1579,
"step": 870
},
{
"epoch": 1.9893373952779894,
"grad_norm": 1.1059821844100952,
"learning_rate": 3.932948869712412e-06,
"loss": 1.169,
"step": 871
},
{
"epoch": 1.9916222391469915,
"grad_norm": 1.1403911113739014,
"learning_rate": 3.930395900310939e-06,
"loss": 1.1586,
"step": 872
},
{
"epoch": 1.993907083015994,
"grad_norm": 1.0881669521331787,
"learning_rate": 3.9278407113551295e-06,
"loss": 1.2262,
"step": 873
},
{
"epoch": 1.9961919268849961,
"grad_norm": 1.1039564609527588,
"learning_rate": 3.925283306809885e-06,
"loss": 1.1951,
"step": 874
},
{
"epoch": 1.9984767707539985,
"grad_norm": 1.1270387172698975,
"learning_rate": 3.9227236906435484e-06,
"loss": 1.1808,
"step": 875
},
{
"epoch": 2.0,
"grad_norm": 1.307372808456421,
"learning_rate": 3.92016186682789e-06,
"loss": 1.1959,
"step": 876
},
{
"epoch": 2.002284843869002,
"grad_norm": 1.3203482627868652,
"learning_rate": 3.917597839338108e-06,
"loss": 1.1606,
"step": 877
},
{
"epoch": 2.0045696877380044,
"grad_norm": 1.1091164350509644,
"learning_rate": 3.915031612152823e-06,
"loss": 1.1532,
"step": 878
},
{
"epoch": 2.006854531607007,
"grad_norm": 1.1063039302825928,
"learning_rate": 3.912463189254063e-06,
"loss": 1.1635,
"step": 879
},
{
"epoch": 2.009139375476009,
"grad_norm": 1.1640090942382812,
"learning_rate": 3.909892574627267e-06,
"loss": 1.1174,
"step": 880
},
{
"epoch": 2.0114242193450114,
"grad_norm": 1.147412657737732,
"learning_rate": 3.907319772261273e-06,
"loss": 1.1285,
"step": 881
},
{
"epoch": 2.0137090632140136,
"grad_norm": 1.1952067613601685,
"learning_rate": 3.904744786148316e-06,
"loss": 1.1657,
"step": 882
},
{
"epoch": 2.015993907083016,
"grad_norm": 1.1174241304397583,
"learning_rate": 3.902167620284017e-06,
"loss": 1.1424,
"step": 883
},
{
"epoch": 2.0182787509520184,
"grad_norm": 1.0976516008377075,
"learning_rate": 3.899588278667382e-06,
"loss": 1.1328,
"step": 884
},
{
"epoch": 2.0205635948210205,
"grad_norm": 1.1376157999038696,
"learning_rate": 3.897006765300791e-06,
"loss": 1.155,
"step": 885
},
{
"epoch": 2.0228484386900227,
"grad_norm": 1.175310730934143,
"learning_rate": 3.8944230841899935e-06,
"loss": 1.1799,
"step": 886
},
{
"epoch": 2.025133282559025,
"grad_norm": 1.1553736925125122,
"learning_rate": 3.8918372393441036e-06,
"loss": 1.1656,
"step": 887
},
{
"epoch": 2.0274181264280275,
"grad_norm": 1.0809762477874756,
"learning_rate": 3.889249234775596e-06,
"loss": 1.158,
"step": 888
},
{
"epoch": 2.0297029702970297,
"grad_norm": 1.2475727796554565,
"learning_rate": 3.886659074500291e-06,
"loss": 1.1958,
"step": 889
},
{
"epoch": 2.031987814166032,
"grad_norm": 1.2172832489013672,
"learning_rate": 3.884066762537357e-06,
"loss": 1.1703,
"step": 890
},
{
"epoch": 2.034272658035034,
"grad_norm": 1.1374081373214722,
"learning_rate": 3.8814723029093014e-06,
"loss": 1.1384,
"step": 891
},
{
"epoch": 2.0365575019040367,
"grad_norm": 1.1220778226852417,
"learning_rate": 3.878875699641964e-06,
"loss": 1.1368,
"step": 892
},
{
"epoch": 2.038842345773039,
"grad_norm": 1.0943710803985596,
"learning_rate": 3.876276956764509e-06,
"loss": 1.1345,
"step": 893
},
{
"epoch": 2.041127189642041,
"grad_norm": 1.079895257949829,
"learning_rate": 3.873676078309423e-06,
"loss": 1.1469,
"step": 894
},
{
"epoch": 2.0434120335110433,
"grad_norm": 1.164766550064087,
"learning_rate": 3.871073068312506e-06,
"loss": 1.1458,
"step": 895
},
{
"epoch": 2.045696877380046,
"grad_norm": 1.1422792673110962,
"learning_rate": 3.868467930812864e-06,
"loss": 1.1286,
"step": 896
},
{
"epoch": 2.047981721249048,
"grad_norm": 1.104988694190979,
"learning_rate": 3.865860669852906e-06,
"loss": 1.1316,
"step": 897
},
{
"epoch": 2.0502665651180503,
"grad_norm": 1.157842755317688,
"learning_rate": 3.8632512894783345e-06,
"loss": 1.1515,
"step": 898
},
{
"epoch": 2.0525514089870525,
"grad_norm": 1.255118727684021,
"learning_rate": 3.860639793738143e-06,
"loss": 1.1806,
"step": 899
},
{
"epoch": 2.0548362528560546,
"grad_norm": 1.2516144514083862,
"learning_rate": 3.858026186684604e-06,
"loss": 1.1973,
"step": 900
},
{
"epoch": 2.0571210967250573,
"grad_norm": 1.143211007118225,
"learning_rate": 3.85541047237327e-06,
"loss": 1.1654,
"step": 901
},
{
"epoch": 2.0594059405940595,
"grad_norm": 1.1019172668457031,
"learning_rate": 3.852792654862959e-06,
"loss": 1.1534,
"step": 902
},
{
"epoch": 2.0616907844630616,
"grad_norm": 1.1939572095870972,
"learning_rate": 3.850172738215757e-06,
"loss": 1.1145,
"step": 903
},
{
"epoch": 2.063975628332064,
"grad_norm": 1.2124632596969604,
"learning_rate": 3.847550726497004e-06,
"loss": 1.1037,
"step": 904
},
{
"epoch": 2.0662604722010665,
"grad_norm": 1.144073486328125,
"learning_rate": 3.844926623775293e-06,
"loss": 1.1605,
"step": 905
},
{
"epoch": 2.0685453160700686,
"grad_norm": 1.1347826719284058,
"learning_rate": 3.84230043412246e-06,
"loss": 1.1421,
"step": 906
},
{
"epoch": 2.070830159939071,
"grad_norm": 1.1252721548080444,
"learning_rate": 3.8396721616135805e-06,
"loss": 1.1767,
"step": 907
},
{
"epoch": 2.073115003808073,
"grad_norm": 1.224308729171753,
"learning_rate": 3.837041810326961e-06,
"loss": 1.1325,
"step": 908
},
{
"epoch": 2.075399847677075,
"grad_norm": 1.25150728225708,
"learning_rate": 3.8344093843441345e-06,
"loss": 1.1235,
"step": 909
},
{
"epoch": 2.077684691546078,
"grad_norm": 1.244081974029541,
"learning_rate": 3.831774887749854e-06,
"loss": 1.2037,
"step": 910
},
{
"epoch": 2.07996953541508,
"grad_norm": 1.14827561378479,
"learning_rate": 3.829138324632082e-06,
"loss": 1.1378,
"step": 911
},
{
"epoch": 2.082254379284082,
"grad_norm": 1.1389387845993042,
"learning_rate": 3.826499699081992e-06,
"loss": 1.1785,
"step": 912
},
{
"epoch": 2.0845392231530844,
"grad_norm": 1.2572706937789917,
"learning_rate": 3.823859015193957e-06,
"loss": 1.1474,
"step": 913
},
{
"epoch": 2.086824067022087,
"grad_norm": 1.1964038610458374,
"learning_rate": 3.8212162770655405e-06,
"loss": 1.1508,
"step": 914
},
{
"epoch": 2.089108910891089,
"grad_norm": 1.2023462057113647,
"learning_rate": 3.818571488797496e-06,
"loss": 1.1401,
"step": 915
},
{
"epoch": 2.0913937547600914,
"grad_norm": 1.1288152933120728,
"learning_rate": 3.815924654493759e-06,
"loss": 1.1395,
"step": 916
},
{
"epoch": 2.0936785986290936,
"grad_norm": 1.15272855758667,
"learning_rate": 3.8132757782614405e-06,
"loss": 1.1357,
"step": 917
},
{
"epoch": 2.095963442498096,
"grad_norm": 1.176313877105713,
"learning_rate": 3.810624864210816e-06,
"loss": 1.1529,
"step": 918
},
{
"epoch": 2.0982482863670984,
"grad_norm": 1.2011158466339111,
"learning_rate": 3.807971916455325e-06,
"loss": 1.1301,
"step": 919
},
{
"epoch": 2.1005331302361006,
"grad_norm": 1.135551929473877,
"learning_rate": 3.8053169391115665e-06,
"loss": 1.1545,
"step": 920
},
{
"epoch": 2.1028179741051027,
"grad_norm": 1.1023021936416626,
"learning_rate": 3.802659936299283e-06,
"loss": 1.1619,
"step": 921
},
{
"epoch": 2.105102817974105,
"grad_norm": 1.1410151720046997,
"learning_rate": 3.800000912141363e-06,
"loss": 1.168,
"step": 922
},
{
"epoch": 2.1073876618431076,
"grad_norm": 1.1780641078948975,
"learning_rate": 3.797339870763831e-06,
"loss": 1.1895,
"step": 923
},
{
"epoch": 2.1096725057121097,
"grad_norm": 1.13074791431427,
"learning_rate": 3.7946768162958424e-06,
"loss": 1.1786,
"step": 924
},
{
"epoch": 2.111957349581112,
"grad_norm": 1.1389769315719604,
"learning_rate": 3.792011752869676e-06,
"loss": 1.1401,
"step": 925
},
{
"epoch": 2.114242193450114,
"grad_norm": 1.1424660682678223,
"learning_rate": 3.7893446846207254e-06,
"loss": 1.1269,
"step": 926
},
{
"epoch": 2.1165270373191167,
"grad_norm": 1.2282761335372925,
"learning_rate": 3.7866756156874996e-06,
"loss": 1.1739,
"step": 927
},
{
"epoch": 2.118811881188119,
"grad_norm": 1.1365275382995605,
"learning_rate": 3.7840045502116073e-06,
"loss": 1.1506,
"step": 928
},
{
"epoch": 2.121096725057121,
"grad_norm": 1.1741012334823608,
"learning_rate": 3.7813314923377603e-06,
"loss": 1.1436,
"step": 929
},
{
"epoch": 2.1233815689261233,
"grad_norm": 1.1063398122787476,
"learning_rate": 3.778656446213757e-06,
"loss": 1.1567,
"step": 930
},
{
"epoch": 2.1256664127951255,
"grad_norm": 1.1437269449234009,
"learning_rate": 3.775979415990485e-06,
"loss": 1.1524,
"step": 931
},
{
"epoch": 2.127951256664128,
"grad_norm": 1.180048942565918,
"learning_rate": 3.773300405821908e-06,
"loss": 1.1283,
"step": 932
},
{
"epoch": 2.1302361005331303,
"grad_norm": 1.179042100906372,
"learning_rate": 3.7706194198650635e-06,
"loss": 1.1285,
"step": 933
},
{
"epoch": 2.1325209444021325,
"grad_norm": 1.1634424924850464,
"learning_rate": 3.767936462280054e-06,
"loss": 1.1491,
"step": 934
},
{
"epoch": 2.1348057882711347,
"grad_norm": 1.1401522159576416,
"learning_rate": 3.7652515372300415e-06,
"loss": 1.1256,
"step": 935
},
{
"epoch": 2.1370906321401373,
"grad_norm": 1.240664005279541,
"learning_rate": 3.762564648881242e-06,
"loss": 1.1735,
"step": 936
},
{
"epoch": 2.1393754760091395,
"grad_norm": 1.23284113407135,
"learning_rate": 3.7598758014029158e-06,
"loss": 1.1421,
"step": 937
},
{
"epoch": 2.1416603198781416,
"grad_norm": 1.1896634101867676,
"learning_rate": 3.757184998967366e-06,
"loss": 1.1041,
"step": 938
},
{
"epoch": 2.143945163747144,
"grad_norm": 1.1241672039031982,
"learning_rate": 3.7544922457499256e-06,
"loss": 1.1566,
"step": 939
},
{
"epoch": 2.146230007616146,
"grad_norm": 1.1452815532684326,
"learning_rate": 3.751797545928959e-06,
"loss": 1.1952,
"step": 940
},
{
"epoch": 2.1485148514851486,
"grad_norm": 1.1131870746612549,
"learning_rate": 3.7491009036858483e-06,
"loss": 1.14,
"step": 941
},
{
"epoch": 2.150799695354151,
"grad_norm": 1.16527259349823,
"learning_rate": 3.7464023232049895e-06,
"loss": 1.1648,
"step": 942
},
{
"epoch": 2.153084539223153,
"grad_norm": 1.1289446353912354,
"learning_rate": 3.7437018086737876e-06,
"loss": 1.179,
"step": 943
},
{
"epoch": 2.155369383092155,
"grad_norm": 1.1474318504333496,
"learning_rate": 3.740999364282647e-06,
"loss": 1.2018,
"step": 944
},
{
"epoch": 2.157654226961158,
"grad_norm": 1.177672266960144,
"learning_rate": 3.7382949942249695e-06,
"loss": 1.1756,
"step": 945
},
{
"epoch": 2.15993907083016,
"grad_norm": 1.152292251586914,
"learning_rate": 3.7355887026971417e-06,
"loss": 1.1387,
"step": 946
},
{
"epoch": 2.162223914699162,
"grad_norm": 1.1584752798080444,
"learning_rate": 3.7328804938985335e-06,
"loss": 1.1648,
"step": 947
},
{
"epoch": 2.1645087585681644,
"grad_norm": 1.1106222867965698,
"learning_rate": 3.7301703720314897e-06,
"loss": 1.1875,
"step": 948
},
{
"epoch": 2.166793602437167,
"grad_norm": 1.177903413772583,
"learning_rate": 3.727458341301324e-06,
"loss": 1.1488,
"step": 949
},
{
"epoch": 2.169078446306169,
"grad_norm": 1.149046778678894,
"learning_rate": 3.7247444059163106e-06,
"loss": 1.1601,
"step": 950
},
{
"epoch": 2.1713632901751714,
"grad_norm": 1.1731359958648682,
"learning_rate": 3.7220285700876812e-06,
"loss": 1.1743,
"step": 951
},
{
"epoch": 2.1736481340441736,
"grad_norm": 1.1620594263076782,
"learning_rate": 3.719310838029615e-06,
"loss": 1.2199,
"step": 952
},
{
"epoch": 2.1759329779131757,
"grad_norm": 1.1532787084579468,
"learning_rate": 3.716591213959234e-06,
"loss": 1.1403,
"step": 953
},
{
"epoch": 2.1782178217821784,
"grad_norm": 1.1255860328674316,
"learning_rate": 3.7138697020965945e-06,
"loss": 1.1262,
"step": 954
},
{
"epoch": 2.1805026656511806,
"grad_norm": 1.1715703010559082,
"learning_rate": 3.7111463066646858e-06,
"loss": 1.1371,
"step": 955
},
{
"epoch": 2.1827875095201827,
"grad_norm": 1.121799111366272,
"learning_rate": 3.7084210318894177e-06,
"loss": 1.1942,
"step": 956
},
{
"epoch": 2.185072353389185,
"grad_norm": 1.1630092859268188,
"learning_rate": 3.7056938819996146e-06,
"loss": 1.1928,
"step": 957
},
{
"epoch": 2.1873571972581876,
"grad_norm": 1.132199764251709,
"learning_rate": 3.702964861227013e-06,
"loss": 1.1706,
"step": 958
},
{
"epoch": 2.1896420411271897,
"grad_norm": 1.1396610736846924,
"learning_rate": 3.7002339738062513e-06,
"loss": 1.1628,
"step": 959
},
{
"epoch": 2.191926884996192,
"grad_norm": 1.0978549718856812,
"learning_rate": 3.6975012239748664e-06,
"loss": 1.1966,
"step": 960
},
{
"epoch": 2.194211728865194,
"grad_norm": 1.1913878917694092,
"learning_rate": 3.694766615973281e-06,
"loss": 1.159,
"step": 961
},
{
"epoch": 2.1964965727341963,
"grad_norm": 1.1240078210830688,
"learning_rate": 3.6920301540448054e-06,
"loss": 1.1767,
"step": 962
},
{
"epoch": 2.198781416603199,
"grad_norm": 1.2754088640213013,
"learning_rate": 3.6892918424356238e-06,
"loss": 1.1382,
"step": 963
},
{
"epoch": 2.201066260472201,
"grad_norm": 1.1081241369247437,
"learning_rate": 3.6865516853947923e-06,
"loss": 1.1634,
"step": 964
},
{
"epoch": 2.2033511043412033,
"grad_norm": 1.1287072896957397,
"learning_rate": 3.683809687174229e-06,
"loss": 1.1354,
"step": 965
},
{
"epoch": 2.2056359482102055,
"grad_norm": 1.1273839473724365,
"learning_rate": 3.6810658520287106e-06,
"loss": 1.147,
"step": 966
},
{
"epoch": 2.207920792079208,
"grad_norm": 1.2143324613571167,
"learning_rate": 3.6783201842158633e-06,
"loss": 1.1793,
"step": 967
},
{
"epoch": 2.2102056359482103,
"grad_norm": 1.2204785346984863,
"learning_rate": 3.6755726879961575e-06,
"loss": 1.1915,
"step": 968
},
{
"epoch": 2.2124904798172125,
"grad_norm": 1.168456792831421,
"learning_rate": 3.6728233676328988e-06,
"loss": 1.1502,
"step": 969
},
{
"epoch": 2.2147753236862147,
"grad_norm": 1.1369385719299316,
"learning_rate": 3.670072227392226e-06,
"loss": 1.1787,
"step": 970
},
{
"epoch": 2.217060167555217,
"grad_norm": 1.1360613107681274,
"learning_rate": 3.6673192715431016e-06,
"loss": 1.1279,
"step": 971
},
{
"epoch": 2.2193450114242195,
"grad_norm": 1.2034990787506104,
"learning_rate": 3.6645645043573044e-06,
"loss": 1.1586,
"step": 972
},
{
"epoch": 2.2216298552932217,
"grad_norm": 1.133101224899292,
"learning_rate": 3.661807930109422e-06,
"loss": 1.1521,
"step": 973
},
{
"epoch": 2.223914699162224,
"grad_norm": 1.185314655303955,
"learning_rate": 3.6590495530768493e-06,
"loss": 1.1347,
"step": 974
},
{
"epoch": 2.226199543031226,
"grad_norm": 1.1746361255645752,
"learning_rate": 3.656289377539778e-06,
"loss": 1.1749,
"step": 975
},
{
"epoch": 2.2284843869002287,
"grad_norm": 1.1611051559448242,
"learning_rate": 3.65352740778119e-06,
"loss": 1.1399,
"step": 976
},
{
"epoch": 2.230769230769231,
"grad_norm": 1.1828947067260742,
"learning_rate": 3.650763648086849e-06,
"loss": 1.1815,
"step": 977
},
{
"epoch": 2.233054074638233,
"grad_norm": 1.1289913654327393,
"learning_rate": 3.6479981027453002e-06,
"loss": 1.1524,
"step": 978
},
{
"epoch": 2.235338918507235,
"grad_norm": 1.1266071796417236,
"learning_rate": 3.6452307760478583e-06,
"loss": 1.1502,
"step": 979
},
{
"epoch": 2.237623762376238,
"grad_norm": 1.128389596939087,
"learning_rate": 3.6424616722886004e-06,
"loss": 1.1611,
"step": 980
},
{
"epoch": 2.23990860624524,
"grad_norm": 1.216874122619629,
"learning_rate": 3.6396907957643623e-06,
"loss": 1.1693,
"step": 981
},
{
"epoch": 2.242193450114242,
"grad_norm": 1.1498830318450928,
"learning_rate": 3.6369181507747305e-06,
"loss": 1.1304,
"step": 982
},
{
"epoch": 2.2444782939832444,
"grad_norm": 1.1365691423416138,
"learning_rate": 3.634143741622036e-06,
"loss": 1.1673,
"step": 983
},
{
"epoch": 2.2467631378522466,
"grad_norm": 1.1243813037872314,
"learning_rate": 3.631367572611348e-06,
"loss": 1.1665,
"step": 984
},
{
"epoch": 2.249047981721249,
"grad_norm": 1.1227095127105713,
"learning_rate": 3.6285896480504633e-06,
"loss": 1.1566,
"step": 985
},
{
"epoch": 2.2513328255902514,
"grad_norm": 1.1476339101791382,
"learning_rate": 3.6258099722499063e-06,
"loss": 1.1759,
"step": 986
},
{
"epoch": 2.2536176694592536,
"grad_norm": 1.130340814590454,
"learning_rate": 3.623028549522918e-06,
"loss": 1.1835,
"step": 987
},
{
"epoch": 2.2559025133282558,
"grad_norm": 1.1270296573638916,
"learning_rate": 3.620245384185448e-06,
"loss": 1.177,
"step": 988
},
{
"epoch": 2.258187357197258,
"grad_norm": 1.170242190361023,
"learning_rate": 3.6174604805561524e-06,
"loss": 1.1589,
"step": 989
},
{
"epoch": 2.2604722010662606,
"grad_norm": 1.1653361320495605,
"learning_rate": 3.6146738429563837e-06,
"loss": 1.1349,
"step": 990
},
{
"epoch": 2.2627570449352628,
"grad_norm": 1.195779800415039,
"learning_rate": 3.6118854757101855e-06,
"loss": 1.1427,
"step": 991
},
{
"epoch": 2.265041888804265,
"grad_norm": 1.1496593952178955,
"learning_rate": 3.609095383144284e-06,
"loss": 1.1922,
"step": 992
},
{
"epoch": 2.2673267326732676,
"grad_norm": 1.1693689823150635,
"learning_rate": 3.6063035695880838e-06,
"loss": 1.1701,
"step": 993
},
{
"epoch": 2.2696115765422697,
"grad_norm": 1.1333836317062378,
"learning_rate": 3.60351003937366e-06,
"loss": 1.163,
"step": 994
},
{
"epoch": 2.271896420411272,
"grad_norm": 1.1533620357513428,
"learning_rate": 3.6007147968357505e-06,
"loss": 1.1117,
"step": 995
},
{
"epoch": 2.274181264280274,
"grad_norm": 1.1433489322662354,
"learning_rate": 3.5979178463117505e-06,
"loss": 1.1871,
"step": 996
},
{
"epoch": 2.2764661081492763,
"grad_norm": 1.1923670768737793,
"learning_rate": 3.5951191921417063e-06,
"loss": 1.1502,
"step": 997
},
{
"epoch": 2.278750952018279,
"grad_norm": 1.13682222366333,
"learning_rate": 3.5923188386683067e-06,
"loss": 1.1314,
"step": 998
},
{
"epoch": 2.281035795887281,
"grad_norm": 1.2247021198272705,
"learning_rate": 3.589516790236879e-06,
"loss": 1.1392,
"step": 999
},
{
"epoch": 2.2833206397562833,
"grad_norm": 1.1454575061798096,
"learning_rate": 3.586713051195378e-06,
"loss": 1.1473,
"step": 1000
},
{
"epoch": 2.2856054836252855,
"grad_norm": 1.1719251871109009,
"learning_rate": 3.583907625894384e-06,
"loss": 1.1642,
"step": 1001
},
{
"epoch": 2.2878903274942877,
"grad_norm": 1.171013355255127,
"learning_rate": 3.5811005186870927e-06,
"loss": 1.1811,
"step": 1002
},
{
"epoch": 2.2901751713632903,
"grad_norm": 1.1246569156646729,
"learning_rate": 3.578291733929311e-06,
"loss": 1.1172,
"step": 1003
},
{
"epoch": 2.2924600152322925,
"grad_norm": 1.1219955682754517,
"learning_rate": 3.5754812759794465e-06,
"loss": 1.1369,
"step": 1004
},
{
"epoch": 2.2947448591012947,
"grad_norm": 1.1373074054718018,
"learning_rate": 3.572669149198506e-06,
"loss": 1.1703,
"step": 1005
},
{
"epoch": 2.297029702970297,
"grad_norm": 1.204938530921936,
"learning_rate": 3.569855357950084e-06,
"loss": 1.1089,
"step": 1006
},
{
"epoch": 2.2993145468392995,
"grad_norm": 1.1538746356964111,
"learning_rate": 3.567039906600357e-06,
"loss": 1.1508,
"step": 1007
},
{
"epoch": 2.3015993907083017,
"grad_norm": 1.135554313659668,
"learning_rate": 3.5642227995180787e-06,
"loss": 1.1507,
"step": 1008
},
{
"epoch": 2.303884234577304,
"grad_norm": 1.1385624408721924,
"learning_rate": 3.5614040410745737e-06,
"loss": 1.1578,
"step": 1009
},
{
"epoch": 2.306169078446306,
"grad_norm": 1.144099473953247,
"learning_rate": 3.5585836356437266e-06,
"loss": 1.1549,
"step": 1010
},
{
"epoch": 2.3084539223153087,
"grad_norm": 1.1651406288146973,
"learning_rate": 3.555761587601976e-06,
"loss": 1.1613,
"step": 1011
},
{
"epoch": 2.310738766184311,
"grad_norm": 1.1579418182373047,
"learning_rate": 3.552937901328315e-06,
"loss": 1.1261,
"step": 1012
},
{
"epoch": 2.313023610053313,
"grad_norm": 1.2125619649887085,
"learning_rate": 3.550112581204273e-06,
"loss": 1.1778,
"step": 1013
},
{
"epoch": 2.315308453922315,
"grad_norm": 1.2006595134735107,
"learning_rate": 3.5472856316139193e-06,
"loss": 1.1755,
"step": 1014
},
{
"epoch": 2.3175932977913174,
"grad_norm": 1.1417663097381592,
"learning_rate": 3.5444570569438465e-06,
"loss": 1.1604,
"step": 1015
},
{
"epoch": 2.31987814166032,
"grad_norm": 1.1562373638153076,
"learning_rate": 3.5416268615831737e-06,
"loss": 1.17,
"step": 1016
},
{
"epoch": 2.322162985529322,
"grad_norm": 1.4638527631759644,
"learning_rate": 3.5387950499235323e-06,
"loss": 1.159,
"step": 1017
},
{
"epoch": 2.3244478293983244,
"grad_norm": 1.1422604322433472,
"learning_rate": 3.5359616263590637e-06,
"loss": 1.1391,
"step": 1018
},
{
"epoch": 2.3267326732673266,
"grad_norm": 1.1229259967803955,
"learning_rate": 3.5331265952864065e-06,
"loss": 1.2002,
"step": 1019
},
{
"epoch": 2.329017517136329,
"grad_norm": 1.1740734577178955,
"learning_rate": 3.530289961104698e-06,
"loss": 1.1347,
"step": 1020
},
{
"epoch": 2.3313023610053314,
"grad_norm": 1.1417142152786255,
"learning_rate": 3.527451728215561e-06,
"loss": 1.1979,
"step": 1021
},
{
"epoch": 2.3335872048743336,
"grad_norm": 1.1507670879364014,
"learning_rate": 3.5246119010230994e-06,
"loss": 1.1522,
"step": 1022
},
{
"epoch": 2.3358720487433358,
"grad_norm": 1.1694669723510742,
"learning_rate": 3.521770483933891e-06,
"loss": 1.1215,
"step": 1023
},
{
"epoch": 2.3381568926123384,
"grad_norm": 1.134131669998169,
"learning_rate": 3.5189274813569807e-06,
"loss": 1.1648,
"step": 1024
},
{
"epoch": 2.3404417364813406,
"grad_norm": 1.144411325454712,
"learning_rate": 3.516082897703873e-06,
"loss": 1.1526,
"step": 1025
},
{
"epoch": 2.3427265803503428,
"grad_norm": 1.1161473989486694,
"learning_rate": 3.5132367373885267e-06,
"loss": 1.1495,
"step": 1026
},
{
"epoch": 2.345011424219345,
"grad_norm": 1.143911600112915,
"learning_rate": 3.5103890048273464e-06,
"loss": 1.1724,
"step": 1027
},
{
"epoch": 2.347296268088347,
"grad_norm": 1.2014847993850708,
"learning_rate": 3.507539704439177e-06,
"loss": 1.1622,
"step": 1028
},
{
"epoch": 2.3495811119573498,
"grad_norm": 1.2054967880249023,
"learning_rate": 3.5046888406452966e-06,
"loss": 1.1539,
"step": 1029
},
{
"epoch": 2.351865955826352,
"grad_norm": 1.1581631898880005,
"learning_rate": 3.5018364178694077e-06,
"loss": 1.16,
"step": 1030
},
{
"epoch": 2.354150799695354,
"grad_norm": 1.1760412454605103,
"learning_rate": 3.4989824405376314e-06,
"loss": 1.1736,
"step": 1031
},
{
"epoch": 2.3564356435643563,
"grad_norm": 1.1309703588485718,
"learning_rate": 3.4961269130785047e-06,
"loss": 1.166,
"step": 1032
},
{
"epoch": 2.3587204874333585,
"grad_norm": 1.181168556213379,
"learning_rate": 3.493269839922967e-06,
"loss": 1.1701,
"step": 1033
},
{
"epoch": 2.361005331302361,
"grad_norm": 1.1288000345230103,
"learning_rate": 3.490411225504355e-06,
"loss": 1.1546,
"step": 1034
},
{
"epoch": 2.3632901751713633,
"grad_norm": 1.1293566226959229,
"learning_rate": 3.4875510742584006e-06,
"loss": 1.1389,
"step": 1035
},
{
"epoch": 2.3655750190403655,
"grad_norm": 1.161147952079773,
"learning_rate": 3.484689390623218e-06,
"loss": 1.2149,
"step": 1036
},
{
"epoch": 2.3678598629093677,
"grad_norm": 1.1328603029251099,
"learning_rate": 3.4818261790393e-06,
"loss": 1.1373,
"step": 1037
},
{
"epoch": 2.3701447067783703,
"grad_norm": 1.14400315284729,
"learning_rate": 3.478961443949509e-06,
"loss": 1.1547,
"step": 1038
},
{
"epoch": 2.3724295506473725,
"grad_norm": 1.1662416458129883,
"learning_rate": 3.4760951897990734e-06,
"loss": 1.1395,
"step": 1039
},
{
"epoch": 2.3747143945163747,
"grad_norm": 1.3636192083358765,
"learning_rate": 3.473227421035578e-06,
"loss": 1.1528,
"step": 1040
},
{
"epoch": 2.376999238385377,
"grad_norm": 1.1440818309783936,
"learning_rate": 3.4703581421089566e-06,
"loss": 1.1655,
"step": 1041
},
{
"epoch": 2.3792840822543795,
"grad_norm": 1.2090609073638916,
"learning_rate": 3.4674873574714886e-06,
"loss": 1.0997,
"step": 1042
},
{
"epoch": 2.3815689261233817,
"grad_norm": 1.1551830768585205,
"learning_rate": 3.464615071577788e-06,
"loss": 1.1652,
"step": 1043
},
{
"epoch": 2.383853769992384,
"grad_norm": 1.1676479578018188,
"learning_rate": 3.4617412888847984e-06,
"loss": 1.1596,
"step": 1044
},
{
"epoch": 2.386138613861386,
"grad_norm": 1.1927589178085327,
"learning_rate": 3.458866013851788e-06,
"loss": 1.1462,
"step": 1045
},
{
"epoch": 2.388423457730388,
"grad_norm": 1.2201225757598877,
"learning_rate": 3.455989250940338e-06,
"loss": 1.1703,
"step": 1046
},
{
"epoch": 2.390708301599391,
"grad_norm": 1.1762723922729492,
"learning_rate": 3.45311100461434e-06,
"loss": 1.1706,
"step": 1047
},
{
"epoch": 2.392993145468393,
"grad_norm": 1.163713812828064,
"learning_rate": 3.4502312793399873e-06,
"loss": 1.1208,
"step": 1048
},
{
"epoch": 2.395277989337395,
"grad_norm": 1.3621329069137573,
"learning_rate": 3.4473500795857674e-06,
"loss": 1.1493,
"step": 1049
},
{
"epoch": 2.3975628332063974,
"grad_norm": 1.1427867412567139,
"learning_rate": 3.4444674098224555e-06,
"loss": 1.1454,
"step": 1050
},
{
"epoch": 2.3998476770754,
"grad_norm": 1.1881464719772339,
"learning_rate": 3.4415832745231092e-06,
"loss": 1.1094,
"step": 1051
},
{
"epoch": 2.402132520944402,
"grad_norm": 1.134605050086975,
"learning_rate": 3.4386976781630594e-06,
"loss": 1.1676,
"step": 1052
},
{
"epoch": 2.4044173648134044,
"grad_norm": 1.1633696556091309,
"learning_rate": 3.4358106252199043e-06,
"loss": 1.1258,
"step": 1053
},
{
"epoch": 2.4067022086824066,
"grad_norm": 1.1574831008911133,
"learning_rate": 3.4329221201735015e-06,
"loss": 1.1499,
"step": 1054
},
{
"epoch": 2.408987052551409,
"grad_norm": 1.169659972190857,
"learning_rate": 3.430032167505962e-06,
"loss": 1.1197,
"step": 1055
},
{
"epoch": 2.4112718964204114,
"grad_norm": 1.191874623298645,
"learning_rate": 3.4271407717016456e-06,
"loss": 1.1673,
"step": 1056
},
{
"epoch": 2.4135567402894136,
"grad_norm": 1.1627485752105713,
"learning_rate": 3.424247937247148e-06,
"loss": 1.1634,
"step": 1057
},
{
"epoch": 2.4158415841584158,
"grad_norm": 1.1143652200698853,
"learning_rate": 3.421353668631299e-06,
"loss": 1.1509,
"step": 1058
},
{
"epoch": 2.418126428027418,
"grad_norm": 1.1409255266189575,
"learning_rate": 3.418457970345153e-06,
"loss": 1.1155,
"step": 1059
},
{
"epoch": 2.4204112718964206,
"grad_norm": 1.1744626760482788,
"learning_rate": 3.415560846881984e-06,
"loss": 1.176,
"step": 1060
},
{
"epoch": 2.4226961157654228,
"grad_norm": 1.1737017631530762,
"learning_rate": 3.4126623027372763e-06,
"loss": 1.1526,
"step": 1061
},
{
"epoch": 2.424980959634425,
"grad_norm": 1.1165390014648438,
"learning_rate": 3.4097623424087196e-06,
"loss": 1.1041,
"step": 1062
},
{
"epoch": 2.427265803503427,
"grad_norm": 1.1141180992126465,
"learning_rate": 3.4068609703961997e-06,
"loss": 1.153,
"step": 1063
},
{
"epoch": 2.4295506473724293,
"grad_norm": 1.193962574005127,
"learning_rate": 3.4039581912017946e-06,
"loss": 1.1109,
"step": 1064
},
{
"epoch": 2.431835491241432,
"grad_norm": 1.154571533203125,
"learning_rate": 3.401054009329765e-06,
"loss": 1.1416,
"step": 1065
},
{
"epoch": 2.434120335110434,
"grad_norm": 1.2508164644241333,
"learning_rate": 3.398148429286547e-06,
"loss": 1.122,
"step": 1066
},
{
"epoch": 2.4364051789794363,
"grad_norm": 1.1385433673858643,
"learning_rate": 3.3952414555807493e-06,
"loss": 1.167,
"step": 1067
},
{
"epoch": 2.4386900228484385,
"grad_norm": 1.1891573667526245,
"learning_rate": 3.392333092723141e-06,
"loss": 1.1732,
"step": 1068
},
{
"epoch": 2.440974866717441,
"grad_norm": 1.230272889137268,
"learning_rate": 3.389423345226647e-06,
"loss": 1.1714,
"step": 1069
},
{
"epoch": 2.4432597105864433,
"grad_norm": 1.1934008598327637,
"learning_rate": 3.386512217606339e-06,
"loss": 1.1613,
"step": 1070
},
{
"epoch": 2.4455445544554455,
"grad_norm": 1.1615201234817505,
"learning_rate": 3.383599714379435e-06,
"loss": 1.1551,
"step": 1071
},
{
"epoch": 2.4478293983244477,
"grad_norm": 1.1495729684829712,
"learning_rate": 3.3806858400652825e-06,
"loss": 1.1311,
"step": 1072
},
{
"epoch": 2.4501142421934503,
"grad_norm": 1.1871975660324097,
"learning_rate": 3.37777059918536e-06,
"loss": 1.1294,
"step": 1073
},
{
"epoch": 2.4523990860624525,
"grad_norm": 1.1563098430633545,
"learning_rate": 3.374853996263264e-06,
"loss": 1.1412,
"step": 1074
},
{
"epoch": 2.4546839299314547,
"grad_norm": 1.27406644821167,
"learning_rate": 3.3719360358247054e-06,
"loss": 1.1515,
"step": 1075
},
{
"epoch": 2.456968773800457,
"grad_norm": 1.227537751197815,
"learning_rate": 3.369016722397504e-06,
"loss": 1.1525,
"step": 1076
},
{
"epoch": 2.459253617669459,
"grad_norm": 1.1642208099365234,
"learning_rate": 3.366096060511575e-06,
"loss": 1.1476,
"step": 1077
},
{
"epoch": 2.4615384615384617,
"grad_norm": 1.1597189903259277,
"learning_rate": 3.363174054698928e-06,
"loss": 1.1567,
"step": 1078
},
{
"epoch": 2.463823305407464,
"grad_norm": 1.2041665315628052,
"learning_rate": 3.3602507094936576e-06,
"loss": 1.1265,
"step": 1079
},
{
"epoch": 2.466108149276466,
"grad_norm": 1.2899173498153687,
"learning_rate": 3.357326029431939e-06,
"loss": 1.157,
"step": 1080
},
{
"epoch": 2.4683929931454682,
"grad_norm": 1.2047252655029297,
"learning_rate": 3.3544000190520144e-06,
"loss": 1.1466,
"step": 1081
},
{
"epoch": 2.470677837014471,
"grad_norm": 1.2135494947433472,
"learning_rate": 3.351472682894193e-06,
"loss": 1.1474,
"step": 1082
},
{
"epoch": 2.472962680883473,
"grad_norm": 1.1497875452041626,
"learning_rate": 3.348544025500841e-06,
"loss": 1.0876,
"step": 1083
},
{
"epoch": 2.4752475247524752,
"grad_norm": 1.260158658027649,
"learning_rate": 3.3456140514163756e-06,
"loss": 1.1698,
"step": 1084
},
{
"epoch": 2.4775323686214774,
"grad_norm": 1.2946076393127441,
"learning_rate": 3.342682765187254e-06,
"loss": 1.1504,
"step": 1085
},
{
"epoch": 2.47981721249048,
"grad_norm": 1.2317827939987183,
"learning_rate": 3.3397501713619736e-06,
"loss": 1.121,
"step": 1086
},
{
"epoch": 2.4821020563594822,
"grad_norm": 1.1697484254837036,
"learning_rate": 3.336816274491057e-06,
"loss": 1.185,
"step": 1087
},
{
"epoch": 2.4843869002284844,
"grad_norm": 1.1670335531234741,
"learning_rate": 3.333881079127052e-06,
"loss": 1.1294,
"step": 1088
},
{
"epoch": 2.4866717440974866,
"grad_norm": 1.2509000301361084,
"learning_rate": 3.3309445898245184e-06,
"loss": 1.178,
"step": 1089
},
{
"epoch": 2.4889565879664888,
"grad_norm": 1.178087592124939,
"learning_rate": 3.328006811140026e-06,
"loss": 1.1516,
"step": 1090
},
{
"epoch": 2.4912414318354914,
"grad_norm": 1.2256996631622314,
"learning_rate": 3.3250677476321442e-06,
"loss": 1.1598,
"step": 1091
},
{
"epoch": 2.4935262757044936,
"grad_norm": 1.1501520872116089,
"learning_rate": 3.322127403861437e-06,
"loss": 1.1678,
"step": 1092
},
{
"epoch": 2.4958111195734958,
"grad_norm": 1.2015535831451416,
"learning_rate": 3.319185784390453e-06,
"loss": 1.159,
"step": 1093
},
{
"epoch": 2.498095963442498,
"grad_norm": 1.203966736793518,
"learning_rate": 3.3162428937837233e-06,
"loss": 1.1335,
"step": 1094
},
{
"epoch": 2.5003808073115,
"grad_norm": 1.2212433815002441,
"learning_rate": 3.313298736607748e-06,
"loss": 1.1801,
"step": 1095
},
{
"epoch": 2.5026656511805028,
"grad_norm": 1.222063422203064,
"learning_rate": 3.3103533174309967e-06,
"loss": 1.1536,
"step": 1096
},
{
"epoch": 2.504950495049505,
"grad_norm": 1.1464064121246338,
"learning_rate": 3.3074066408238927e-06,
"loss": 1.0955,
"step": 1097
},
{
"epoch": 2.507235338918507,
"grad_norm": 1.2135506868362427,
"learning_rate": 3.3044587113588134e-06,
"loss": 1.1545,
"step": 1098
},
{
"epoch": 2.5095201827875098,
"grad_norm": 1.208592414855957,
"learning_rate": 3.3015095336100795e-06,
"loss": 1.1504,
"step": 1099
},
{
"epoch": 2.511805026656512,
"grad_norm": 1.119957685470581,
"learning_rate": 3.2985591121539495e-06,
"loss": 1.1504,
"step": 1100
},
{
"epoch": 2.514089870525514,
"grad_norm": 1.1869513988494873,
"learning_rate": 3.2956074515686105e-06,
"loss": 1.1694,
"step": 1101
},
{
"epoch": 2.5163747143945163,
"grad_norm": 1.1720670461654663,
"learning_rate": 3.2926545564341715e-06,
"loss": 1.1518,
"step": 1102
},
{
"epoch": 2.5186595582635185,
"grad_norm": 1.1721378564834595,
"learning_rate": 3.2897004313326608e-06,
"loss": 1.1388,
"step": 1103
},
{
"epoch": 2.520944402132521,
"grad_norm": 1.1609101295471191,
"learning_rate": 3.2867450808480115e-06,
"loss": 1.1309,
"step": 1104
},
{
"epoch": 2.5232292460015233,
"grad_norm": 1.1172316074371338,
"learning_rate": 3.2837885095660598e-06,
"loss": 1.1626,
"step": 1105
},
{
"epoch": 2.5255140898705255,
"grad_norm": 1.2131626605987549,
"learning_rate": 3.280830722074536e-06,
"loss": 1.1458,
"step": 1106
},
{
"epoch": 2.5277989337395277,
"grad_norm": 1.2087211608886719,
"learning_rate": 3.2778717229630584e-06,
"loss": 1.1665,
"step": 1107
},
{
"epoch": 2.53008377760853,
"grad_norm": 1.171183705329895,
"learning_rate": 3.2749115168231238e-06,
"loss": 1.1922,
"step": 1108
},
{
"epoch": 2.5323686214775325,
"grad_norm": 1.1399853229522705,
"learning_rate": 3.271950108248102e-06,
"loss": 1.1292,
"step": 1109
},
{
"epoch": 2.5346534653465347,
"grad_norm": 1.1595377922058105,
"learning_rate": 3.268987501833231e-06,
"loss": 1.1481,
"step": 1110
},
{
"epoch": 2.536938309215537,
"grad_norm": 1.2080212831497192,
"learning_rate": 3.2660237021756047e-06,
"loss": 1.17,
"step": 1111
},
{
"epoch": 2.5392231530845395,
"grad_norm": 1.150943398475647,
"learning_rate": 3.26305871387417e-06,
"loss": 1.1193,
"step": 1112
},
{
"epoch": 2.5415079969535412,
"grad_norm": 1.1977055072784424,
"learning_rate": 3.260092541529718e-06,
"loss": 1.1656,
"step": 1113
},
{
"epoch": 2.543792840822544,
"grad_norm": 1.1943557262420654,
"learning_rate": 3.257125189744877e-06,
"loss": 1.1436,
"step": 1114
},
{
"epoch": 2.546077684691546,
"grad_norm": 1.1852405071258545,
"learning_rate": 3.254156663124106e-06,
"loss": 1.1743,
"step": 1115
},
{
"epoch": 2.5483625285605482,
"grad_norm": 1.1327954530715942,
"learning_rate": 3.2511869662736855e-06,
"loss": 1.1275,
"step": 1116
},
{
"epoch": 2.550647372429551,
"grad_norm": 1.1748192310333252,
"learning_rate": 3.248216103801713e-06,
"loss": 1.1277,
"step": 1117
},
{
"epoch": 2.552932216298553,
"grad_norm": 1.2016124725341797,
"learning_rate": 3.2452440803180953e-06,
"loss": 1.1692,
"step": 1118
},
{
"epoch": 2.5552170601675552,
"grad_norm": 1.1545820236206055,
"learning_rate": 3.24227090043454e-06,
"loss": 1.1335,
"step": 1119
},
{
"epoch": 2.5575019040365574,
"grad_norm": 1.168172836303711,
"learning_rate": 3.239296568764547e-06,
"loss": 1.1515,
"step": 1120
},
{
"epoch": 2.5597867479055596,
"grad_norm": 1.1570290327072144,
"learning_rate": 3.236321089923408e-06,
"loss": 1.1921,
"step": 1121
},
{
"epoch": 2.5620715917745622,
"grad_norm": 1.1722872257232666,
"learning_rate": 3.233344468528192e-06,
"loss": 1.1842,
"step": 1122
},
{
"epoch": 2.5643564356435644,
"grad_norm": 1.2346643209457397,
"learning_rate": 3.2303667091977397e-06,
"loss": 1.1987,
"step": 1123
},
{
"epoch": 2.5666412795125666,
"grad_norm": 1.1846752166748047,
"learning_rate": 3.2273878165526603e-06,
"loss": 1.1672,
"step": 1124
},
{
"epoch": 2.568926123381569,
"grad_norm": 1.1800742149353027,
"learning_rate": 3.224407795215319e-06,
"loss": 1.1405,
"step": 1125
},
{
"epoch": 2.571210967250571,
"grad_norm": 1.2667362689971924,
"learning_rate": 3.2214266498098357e-06,
"loss": 1.097,
"step": 1126
},
{
"epoch": 2.5734958111195736,
"grad_norm": 1.1848291158676147,
"learning_rate": 3.218444384962071e-06,
"loss": 1.1309,
"step": 1127
},
{
"epoch": 2.575780654988576,
"grad_norm": 1.2490592002868652,
"learning_rate": 3.215461005299624e-06,
"loss": 1.1677,
"step": 1128
},
{
"epoch": 2.578065498857578,
"grad_norm": 1.1780728101730347,
"learning_rate": 3.2124765154518245e-06,
"loss": 1.1438,
"step": 1129
},
{
"epoch": 2.5803503427265806,
"grad_norm": 1.1854690313339233,
"learning_rate": 3.209490920049724e-06,
"loss": 1.1854,
"step": 1130
},
{
"epoch": 2.5826351865955828,
"grad_norm": 1.1640013456344604,
"learning_rate": 3.2065042237260897e-06,
"loss": 1.1421,
"step": 1131
},
{
"epoch": 2.584920030464585,
"grad_norm": 1.1875327825546265,
"learning_rate": 3.2035164311153967e-06,
"loss": 1.1617,
"step": 1132
},
{
"epoch": 2.587204874333587,
"grad_norm": 1.188707709312439,
"learning_rate": 3.200527546853822e-06,
"loss": 1.1618,
"step": 1133
},
{
"epoch": 2.5894897182025893,
"grad_norm": 1.1984432935714722,
"learning_rate": 3.1975375755792358e-06,
"loss": 1.1647,
"step": 1134
},
{
"epoch": 2.591774562071592,
"grad_norm": 1.2000738382339478,
"learning_rate": 3.1945465219311964e-06,
"loss": 1.1555,
"step": 1135
},
{
"epoch": 2.594059405940594,
"grad_norm": 1.1691455841064453,
"learning_rate": 3.19155439055094e-06,
"loss": 1.1337,
"step": 1136
},
{
"epoch": 2.5963442498095963,
"grad_norm": 1.3328726291656494,
"learning_rate": 3.1885611860813747e-06,
"loss": 1.1662,
"step": 1137
},
{
"epoch": 2.5986290936785985,
"grad_norm": 1.159832239151001,
"learning_rate": 3.185566913167076e-06,
"loss": 1.1231,
"step": 1138
},
{
"epoch": 2.6009139375476007,
"grad_norm": 1.139963984489441,
"learning_rate": 3.1825715764542765e-06,
"loss": 1.1348,
"step": 1139
},
{
"epoch": 2.6031987814166033,
"grad_norm": 1.1485956907272339,
"learning_rate": 3.1795751805908578e-06,
"loss": 1.1157,
"step": 1140
},
{
"epoch": 2.6054836252856055,
"grad_norm": 1.1575555801391602,
"learning_rate": 3.1765777302263464e-06,
"loss": 1.1738,
"step": 1141
},
{
"epoch": 2.6077684691546077,
"grad_norm": 1.1840925216674805,
"learning_rate": 3.173579230011905e-06,
"loss": 1.1345,
"step": 1142
},
{
"epoch": 2.6100533130236103,
"grad_norm": 1.144516944885254,
"learning_rate": 3.1705796846003267e-06,
"loss": 1.1219,
"step": 1143
},
{
"epoch": 2.612338156892612,
"grad_norm": 1.1362788677215576,
"learning_rate": 3.1675790986460233e-06,
"loss": 1.1382,
"step": 1144
},
{
"epoch": 2.6146230007616147,
"grad_norm": 1.2047488689422607,
"learning_rate": 3.1645774768050224e-06,
"loss": 1.1427,
"step": 1145
},
{
"epoch": 2.616907844630617,
"grad_norm": 1.1238912343978882,
"learning_rate": 3.1615748237349626e-06,
"loss": 1.1298,
"step": 1146
},
{
"epoch": 2.619192688499619,
"grad_norm": 1.1865835189819336,
"learning_rate": 3.158571144095076e-06,
"loss": 1.1537,
"step": 1147
},
{
"epoch": 2.6214775323686217,
"grad_norm": 1.1614654064178467,
"learning_rate": 3.155566442546194e-06,
"loss": 1.1608,
"step": 1148
},
{
"epoch": 2.623762376237624,
"grad_norm": 1.1808875799179077,
"learning_rate": 3.1525607237507296e-06,
"loss": 1.1168,
"step": 1149
},
{
"epoch": 2.626047220106626,
"grad_norm": 1.2047436237335205,
"learning_rate": 3.1495539923726757e-06,
"loss": 1.1402,
"step": 1150
},
{
"epoch": 2.6283320639756282,
"grad_norm": 1.190082311630249,
"learning_rate": 3.146546253077597e-06,
"loss": 1.1257,
"step": 1151
},
{
"epoch": 2.6306169078446304,
"grad_norm": 1.1529537439346313,
"learning_rate": 3.1435375105326198e-06,
"loss": 1.1332,
"step": 1152
},
{
"epoch": 2.632901751713633,
"grad_norm": 1.1505589485168457,
"learning_rate": 3.1405277694064306e-06,
"loss": 1.1654,
"step": 1153
},
{
"epoch": 2.6351865955826352,
"grad_norm": 1.1244786977767944,
"learning_rate": 3.1375170343692642e-06,
"loss": 1.1625,
"step": 1154
},
{
"epoch": 2.6374714394516374,
"grad_norm": 1.2105226516723633,
"learning_rate": 3.134505310092895e-06,
"loss": 1.1743,
"step": 1155
},
{
"epoch": 2.6397562833206396,
"grad_norm": 1.1880476474761963,
"learning_rate": 3.131492601250636e-06,
"loss": 1.1119,
"step": 1156
},
{
"epoch": 2.642041127189642,
"grad_norm": 1.1480690240859985,
"learning_rate": 3.1284789125173257e-06,
"loss": 1.128,
"step": 1157
},
{
"epoch": 2.6443259710586444,
"grad_norm": 1.174415111541748,
"learning_rate": 3.1254642485693255e-06,
"loss": 1.1545,
"step": 1158
},
{
"epoch": 2.6466108149276466,
"grad_norm": 1.1595776081085205,
"learning_rate": 3.1224486140845063e-06,
"loss": 1.1502,
"step": 1159
},
{
"epoch": 2.648895658796649,
"grad_norm": 1.1877551078796387,
"learning_rate": 3.1194320137422483e-06,
"loss": 1.1487,
"step": 1160
},
{
"epoch": 2.6511805026656514,
"grad_norm": 1.1842609643936157,
"learning_rate": 3.116414452223429e-06,
"loss": 1.1329,
"step": 1161
},
{
"epoch": 2.6534653465346536,
"grad_norm": 1.1747876405715942,
"learning_rate": 3.1133959342104186e-06,
"loss": 1.138,
"step": 1162
},
{
"epoch": 2.655750190403656,
"grad_norm": 1.1728745698928833,
"learning_rate": 3.110376464387069e-06,
"loss": 1.1593,
"step": 1163
},
{
"epoch": 2.658035034272658,
"grad_norm": 1.1547205448150635,
"learning_rate": 3.1073560474387114e-06,
"loss": 1.1244,
"step": 1164
},
{
"epoch": 2.66031987814166,
"grad_norm": 1.1600390672683716,
"learning_rate": 3.1043346880521456e-06,
"loss": 1.1664,
"step": 1165
},
{
"epoch": 2.662604722010663,
"grad_norm": 1.1489025354385376,
"learning_rate": 3.1013123909156347e-06,
"loss": 1.1869,
"step": 1166
},
{
"epoch": 2.664889565879665,
"grad_norm": 1.1556684970855713,
"learning_rate": 3.0982891607188948e-06,
"loss": 1.1408,
"step": 1167
},
{
"epoch": 2.667174409748667,
"grad_norm": 1.1727960109710693,
"learning_rate": 3.095265002153092e-06,
"loss": 1.1676,
"step": 1168
},
{
"epoch": 2.6694592536176693,
"grad_norm": 1.1960148811340332,
"learning_rate": 3.0922399199108326e-06,
"loss": 1.138,
"step": 1169
},
{
"epoch": 2.6717440974866715,
"grad_norm": 1.1238528490066528,
"learning_rate": 3.0892139186861563e-06,
"loss": 1.1308,
"step": 1170
},
{
"epoch": 2.674028941355674,
"grad_norm": 1.1751792430877686,
"learning_rate": 3.0861870031745266e-06,
"loss": 1.1518,
"step": 1171
},
{
"epoch": 2.6763137852246763,
"grad_norm": 1.1722700595855713,
"learning_rate": 3.0831591780728282e-06,
"loss": 1.0687,
"step": 1172
},
{
"epoch": 2.6785986290936785,
"grad_norm": 1.1473584175109863,
"learning_rate": 3.0801304480793563e-06,
"loss": 1.1313,
"step": 1173
},
{
"epoch": 2.680883472962681,
"grad_norm": 1.1742531061172485,
"learning_rate": 3.0771008178938112e-06,
"loss": 1.1293,
"step": 1174
},
{
"epoch": 2.6831683168316833,
"grad_norm": 1.1747832298278809,
"learning_rate": 3.074070292217288e-06,
"loss": 1.1334,
"step": 1175
},
{
"epoch": 2.6854531607006855,
"grad_norm": 1.1738357543945312,
"learning_rate": 3.0710388757522724e-06,
"loss": 1.1663,
"step": 1176
},
{
"epoch": 2.6877380045696877,
"grad_norm": 1.1576931476593018,
"learning_rate": 3.068006573202634e-06,
"loss": 1.1508,
"step": 1177
},
{
"epoch": 2.69002284843869,
"grad_norm": 1.1925418376922607,
"learning_rate": 3.0649733892736143e-06,
"loss": 1.135,
"step": 1178
},
{
"epoch": 2.6923076923076925,
"grad_norm": 1.1554484367370605,
"learning_rate": 3.061939328671824e-06,
"loss": 1.1328,
"step": 1179
},
{
"epoch": 2.6945925361766947,
"grad_norm": 1.153615117073059,
"learning_rate": 3.0589043961052344e-06,
"loss": 1.1356,
"step": 1180
},
{
"epoch": 2.696877380045697,
"grad_norm": 1.1935803890228271,
"learning_rate": 3.05586859628317e-06,
"loss": 1.1266,
"step": 1181
},
{
"epoch": 2.699162223914699,
"grad_norm": 1.1908665895462036,
"learning_rate": 3.0528319339163003e-06,
"loss": 1.1393,
"step": 1182
},
{
"epoch": 2.7014470677837013,
"grad_norm": 1.194982647895813,
"learning_rate": 3.0497944137166326e-06,
"loss": 1.1349,
"step": 1183
},
{
"epoch": 2.703731911652704,
"grad_norm": 1.1403220891952515,
"learning_rate": 3.0467560403975066e-06,
"loss": 1.1269,
"step": 1184
},
{
"epoch": 2.706016755521706,
"grad_norm": 1.1594319343566895,
"learning_rate": 3.043716818673586e-06,
"loss": 1.1316,
"step": 1185
},
{
"epoch": 2.7083015993907082,
"grad_norm": 1.1441702842712402,
"learning_rate": 3.0406767532608495e-06,
"loss": 1.1605,
"step": 1186
},
{
"epoch": 2.7105864432597104,
"grad_norm": 1.2065651416778564,
"learning_rate": 3.0376358488765863e-06,
"loss": 1.1708,
"step": 1187
},
{
"epoch": 2.7128712871287126,
"grad_norm": 1.1729774475097656,
"learning_rate": 3.034594110239386e-06,
"loss": 1.1563,
"step": 1188
},
{
"epoch": 2.7151561309977152,
"grad_norm": 1.1797168254852295,
"learning_rate": 3.0315515420691354e-06,
"loss": 1.1732,
"step": 1189
},
{
"epoch": 2.7174409748667174,
"grad_norm": 1.1579010486602783,
"learning_rate": 3.0285081490870057e-06,
"loss": 1.1375,
"step": 1190
},
{
"epoch": 2.7197258187357196,
"grad_norm": 1.1517558097839355,
"learning_rate": 3.0254639360154475e-06,
"loss": 1.1594,
"step": 1191
},
{
"epoch": 2.7220106626047222,
"grad_norm": 1.2115283012390137,
"learning_rate": 3.0224189075781886e-06,
"loss": 1.1334,
"step": 1192
},
{
"epoch": 2.7242955064737244,
"grad_norm": 1.1372671127319336,
"learning_rate": 3.0193730685002153e-06,
"loss": 1.2064,
"step": 1193
},
{
"epoch": 2.7265803503427266,
"grad_norm": 1.2300207614898682,
"learning_rate": 3.0163264235077777e-06,
"loss": 1.1767,
"step": 1194
},
{
"epoch": 2.728865194211729,
"grad_norm": 1.1949130296707153,
"learning_rate": 3.0132789773283734e-06,
"loss": 1.1474,
"step": 1195
},
{
"epoch": 2.731150038080731,
"grad_norm": 1.1411316394805908,
"learning_rate": 3.0102307346907442e-06,
"loss": 1.1565,
"step": 1196
},
{
"epoch": 2.7334348819497336,
"grad_norm": 1.153688669204712,
"learning_rate": 3.0071817003248667e-06,
"loss": 1.1368,
"step": 1197
},
{
"epoch": 2.735719725818736,
"grad_norm": 1.2030675411224365,
"learning_rate": 3.0041318789619465e-06,
"loss": 1.1567,
"step": 1198
},
{
"epoch": 2.738004569687738,
"grad_norm": 1.1593271493911743,
"learning_rate": 3.001081275334412e-06,
"loss": 1.161,
"step": 1199
},
{
"epoch": 2.74028941355674,
"grad_norm": 1.181596279144287,
"learning_rate": 2.9980298941759035e-06,
"loss": 1.122,
"step": 1200
},
{
"epoch": 2.7425742574257423,
"grad_norm": 1.1973023414611816,
"learning_rate": 2.9949777402212677e-06,
"loss": 1.157,
"step": 1201
},
{
"epoch": 2.744859101294745,
"grad_norm": 1.1745750904083252,
"learning_rate": 2.9919248182065512e-06,
"loss": 1.1843,
"step": 1202
},
{
"epoch": 2.747143945163747,
"grad_norm": 1.1595338582992554,
"learning_rate": 2.9888711328689933e-06,
"loss": 1.1741,
"step": 1203
},
{
"epoch": 2.7494287890327493,
"grad_norm": 1.165958046913147,
"learning_rate": 2.985816688947017e-06,
"loss": 1.1217,
"step": 1204
},
{
"epoch": 2.751713632901752,
"grad_norm": 1.1976659297943115,
"learning_rate": 2.9827614911802205e-06,
"loss": 1.1443,
"step": 1205
},
{
"epoch": 2.753998476770754,
"grad_norm": 1.1795979738235474,
"learning_rate": 2.9797055443093744e-06,
"loss": 1.1624,
"step": 1206
},
{
"epoch": 2.7562833206397563,
"grad_norm": 1.1748056411743164,
"learning_rate": 2.9766488530764105e-06,
"loss": 1.1715,
"step": 1207
},
{
"epoch": 2.7585681645087585,
"grad_norm": 1.191728115081787,
"learning_rate": 2.9735914222244165e-06,
"loss": 1.1472,
"step": 1208
},
{
"epoch": 2.7608530083777607,
"grad_norm": 1.168042540550232,
"learning_rate": 2.970533256497627e-06,
"loss": 1.1389,
"step": 1209
},
{
"epoch": 2.7631378522467633,
"grad_norm": 1.2145792245864868,
"learning_rate": 2.9674743606414163e-06,
"loss": 1.1975,
"step": 1210
},
{
"epoch": 2.7654226961157655,
"grad_norm": 1.1970194578170776,
"learning_rate": 2.9644147394022925e-06,
"loss": 1.1911,
"step": 1211
},
{
"epoch": 2.7677075399847677,
"grad_norm": 1.181572675704956,
"learning_rate": 2.96135439752789e-06,
"loss": 1.1384,
"step": 1212
},
{
"epoch": 2.76999238385377,
"grad_norm": 1.1687877178192139,
"learning_rate": 2.95829333976696e-06,
"loss": 1.118,
"step": 1213
},
{
"epoch": 2.772277227722772,
"grad_norm": 1.171147108078003,
"learning_rate": 2.955231570869365e-06,
"loss": 1.1387,
"step": 1214
},
{
"epoch": 2.7745620715917747,
"grad_norm": 1.184873104095459,
"learning_rate": 2.9521690955860715e-06,
"loss": 1.1535,
"step": 1215
},
{
"epoch": 2.776846915460777,
"grad_norm": 1.1838406324386597,
"learning_rate": 2.9491059186691416e-06,
"loss": 1.157,
"step": 1216
},
{
"epoch": 2.779131759329779,
"grad_norm": 1.1575403213500977,
"learning_rate": 2.9460420448717264e-06,
"loss": 1.1179,
"step": 1217
},
{
"epoch": 2.7814166031987813,
"grad_norm": 1.1447091102600098,
"learning_rate": 2.9429774789480576e-06,
"loss": 1.1745,
"step": 1218
},
{
"epoch": 2.7837014470677834,
"grad_norm": 1.1797469854354858,
"learning_rate": 2.9399122256534412e-06,
"loss": 1.1278,
"step": 1219
},
{
"epoch": 2.785986290936786,
"grad_norm": 1.1838300228118896,
"learning_rate": 2.936846289744252e-06,
"loss": 1.1605,
"step": 1220
},
{
"epoch": 2.7882711348057883,
"grad_norm": 1.1908303499221802,
"learning_rate": 2.9337796759779197e-06,
"loss": 1.1133,
"step": 1221
},
{
"epoch": 2.7905559786747904,
"grad_norm": 1.192145824432373,
"learning_rate": 2.930712389112929e-06,
"loss": 1.1247,
"step": 1222
},
{
"epoch": 2.792840822543793,
"grad_norm": 1.1842783689498901,
"learning_rate": 2.9276444339088095e-06,
"loss": 1.0794,
"step": 1223
},
{
"epoch": 2.7951256664127953,
"grad_norm": 1.152113914489746,
"learning_rate": 2.924575815126125e-06,
"loss": 1.1397,
"step": 1224
},
{
"epoch": 2.7974105102817974,
"grad_norm": 1.220105767250061,
"learning_rate": 2.921506537526471e-06,
"loss": 1.1426,
"step": 1225
},
{
"epoch": 2.7996953541507996,
"grad_norm": 1.2473626136779785,
"learning_rate": 2.9184366058724655e-06,
"loss": 1.1645,
"step": 1226
},
{
"epoch": 2.801980198019802,
"grad_norm": 1.2051730155944824,
"learning_rate": 2.915366024927741e-06,
"loss": 1.1498,
"step": 1227
},
{
"epoch": 2.8042650418888044,
"grad_norm": 1.2148674726486206,
"learning_rate": 2.912294799456936e-06,
"loss": 1.1334,
"step": 1228
},
{
"epoch": 2.8065498857578066,
"grad_norm": 1.176007866859436,
"learning_rate": 2.9092229342256915e-06,
"loss": 1.1177,
"step": 1229
},
{
"epoch": 2.808834729626809,
"grad_norm": 1.1746013164520264,
"learning_rate": 2.90615043400064e-06,
"loss": 1.1708,
"step": 1230
},
{
"epoch": 2.811119573495811,
"grad_norm": 1.1968902349472046,
"learning_rate": 2.9030773035493997e-06,
"loss": 1.1497,
"step": 1231
},
{
"epoch": 2.813404417364813,
"grad_norm": 1.1996639966964722,
"learning_rate": 2.9000035476405657e-06,
"loss": 1.1179,
"step": 1232
},
{
"epoch": 2.815689261233816,
"grad_norm": 1.168249487876892,
"learning_rate": 2.8969291710437054e-06,
"loss": 1.1528,
"step": 1233
},
{
"epoch": 2.817974105102818,
"grad_norm": 1.1561291217803955,
"learning_rate": 2.893854178529347e-06,
"loss": 1.1591,
"step": 1234
},
{
"epoch": 2.82025894897182,
"grad_norm": 1.1894656419754028,
"learning_rate": 2.890778574868977e-06,
"loss": 1.1606,
"step": 1235
},
{
"epoch": 2.822543792840823,
"grad_norm": 1.1667678356170654,
"learning_rate": 2.8877023648350284e-06,
"loss": 1.1799,
"step": 1236
},
{
"epoch": 2.824828636709825,
"grad_norm": 1.176679253578186,
"learning_rate": 2.884625553200876e-06,
"loss": 1.1994,
"step": 1237
},
{
"epoch": 2.827113480578827,
"grad_norm": 1.1774452924728394,
"learning_rate": 2.8815481447408273e-06,
"loss": 1.1716,
"step": 1238
},
{
"epoch": 2.8293983244478293,
"grad_norm": 1.2223371267318726,
"learning_rate": 2.878470144230118e-06,
"loss": 1.131,
"step": 1239
},
{
"epoch": 2.8316831683168315,
"grad_norm": 1.1785866022109985,
"learning_rate": 2.875391556444898e-06,
"loss": 1.1411,
"step": 1240
},
{
"epoch": 2.833968012185834,
"grad_norm": 1.1710691452026367,
"learning_rate": 2.8723123861622338e-06,
"loss": 1.1718,
"step": 1241
},
{
"epoch": 2.8362528560548363,
"grad_norm": 1.1571224927902222,
"learning_rate": 2.8692326381600926e-06,
"loss": 1.1529,
"step": 1242
},
{
"epoch": 2.8385376999238385,
"grad_norm": 1.1815292835235596,
"learning_rate": 2.8661523172173392e-06,
"loss": 1.1522,
"step": 1243
},
{
"epoch": 2.8408225437928407,
"grad_norm": 1.1292341947555542,
"learning_rate": 2.8630714281137263e-06,
"loss": 1.1394,
"step": 1244
},
{
"epoch": 2.843107387661843,
"grad_norm": 1.1632921695709229,
"learning_rate": 2.8599899756298887e-06,
"loss": 1.1778,
"step": 1245
},
{
"epoch": 2.8453922315308455,
"grad_norm": 1.2025562524795532,
"learning_rate": 2.856907964547337e-06,
"loss": 1.1442,
"step": 1246
},
{
"epoch": 2.8476770753998477,
"grad_norm": 1.2091012001037598,
"learning_rate": 2.8538253996484465e-06,
"loss": 1.1388,
"step": 1247
},
{
"epoch": 2.84996191926885,
"grad_norm": 1.1544710397720337,
"learning_rate": 2.8507422857164523e-06,
"loss": 1.1402,
"step": 1248
},
{
"epoch": 2.852246763137852,
"grad_norm": 1.1689878702163696,
"learning_rate": 2.847658627535442e-06,
"loss": 1.1477,
"step": 1249
},
{
"epoch": 2.8545316070068543,
"grad_norm": 1.173471212387085,
"learning_rate": 2.844574429890347e-06,
"loss": 1.1661,
"step": 1250
},
{
"epoch": 2.856816450875857,
"grad_norm": 1.1636780500411987,
"learning_rate": 2.8414896975669374e-06,
"loss": 1.1465,
"step": 1251
},
{
"epoch": 2.859101294744859,
"grad_norm": 1.1479815244674683,
"learning_rate": 2.8384044353518104e-06,
"loss": 1.1568,
"step": 1252
},
{
"epoch": 2.8613861386138613,
"grad_norm": 1.2065167427062988,
"learning_rate": 2.835318648032388e-06,
"loss": 1.1585,
"step": 1253
},
{
"epoch": 2.863670982482864,
"grad_norm": 1.1619449853897095,
"learning_rate": 2.832232340396904e-06,
"loss": 1.1238,
"step": 1254
},
{
"epoch": 2.865955826351866,
"grad_norm": 1.17509126663208,
"learning_rate": 2.8291455172344045e-06,
"loss": 1.1614,
"step": 1255
},
{
"epoch": 2.8682406702208683,
"grad_norm": 1.1738674640655518,
"learning_rate": 2.82605818333473e-06,
"loss": 1.1523,
"step": 1256
},
{
"epoch": 2.8705255140898704,
"grad_norm": 1.1643424034118652,
"learning_rate": 2.8229703434885165e-06,
"loss": 1.1565,
"step": 1257
},
{
"epoch": 2.8728103579588726,
"grad_norm": 1.1405576467514038,
"learning_rate": 2.819882002487185e-06,
"loss": 1.1638,
"step": 1258
},
{
"epoch": 2.8750952018278753,
"grad_norm": 1.1445550918579102,
"learning_rate": 2.816793165122933e-06,
"loss": 1.129,
"step": 1259
},
{
"epoch": 2.8773800456968774,
"grad_norm": 1.1620489358901978,
"learning_rate": 2.8137038361887297e-06,
"loss": 1.128,
"step": 1260
},
{
"epoch": 2.8796648895658796,
"grad_norm": 1.2187519073486328,
"learning_rate": 2.8106140204783054e-06,
"loss": 1.1312,
"step": 1261
},
{
"epoch": 2.881949733434882,
"grad_norm": 1.178515076637268,
"learning_rate": 2.8075237227861475e-06,
"loss": 1.1075,
"step": 1262
},
{
"epoch": 2.884234577303884,
"grad_norm": 1.1716185808181763,
"learning_rate": 2.80443294790749e-06,
"loss": 1.1553,
"step": 1263
},
{
"epoch": 2.8865194211728866,
"grad_norm": 1.2317765951156616,
"learning_rate": 2.8013417006383078e-06,
"loss": 1.1377,
"step": 1264
},
{
"epoch": 2.888804265041889,
"grad_norm": 1.1073942184448242,
"learning_rate": 2.798249985775309e-06,
"loss": 1.1507,
"step": 1265
},
{
"epoch": 2.891089108910891,
"grad_norm": 1.149707317352295,
"learning_rate": 2.795157808115927e-06,
"loss": 1.1161,
"step": 1266
},
{
"epoch": 2.8933739527798936,
"grad_norm": 1.2311087846755981,
"learning_rate": 2.7920651724583124e-06,
"loss": 1.1481,
"step": 1267
},
{
"epoch": 2.895658796648896,
"grad_norm": 1.1760973930358887,
"learning_rate": 2.788972083601329e-06,
"loss": 1.16,
"step": 1268
},
{
"epoch": 2.897943640517898,
"grad_norm": 1.149901270866394,
"learning_rate": 2.785878546344541e-06,
"loss": 1.1187,
"step": 1269
},
{
"epoch": 2.9002284843869,
"grad_norm": 1.2037584781646729,
"learning_rate": 2.7827845654882112e-06,
"loss": 1.1508,
"step": 1270
},
{
"epoch": 2.9025133282559024,
"grad_norm": 1.2114837169647217,
"learning_rate": 2.7796901458332877e-06,
"loss": 1.1243,
"step": 1271
},
{
"epoch": 2.904798172124905,
"grad_norm": 1.179283857345581,
"learning_rate": 2.776595292181401e-06,
"loss": 1.1351,
"step": 1272
},
{
"epoch": 2.907083015993907,
"grad_norm": 1.2145215272903442,
"learning_rate": 2.7735000093348556e-06,
"loss": 1.1295,
"step": 1273
},
{
"epoch": 2.9093678598629094,
"grad_norm": 1.1994370222091675,
"learning_rate": 2.7704043020966222e-06,
"loss": 1.1405,
"step": 1274
},
{
"epoch": 2.9116527037319115,
"grad_norm": 1.218436360359192,
"learning_rate": 2.7673081752703275e-06,
"loss": 1.163,
"step": 1275
},
{
"epoch": 2.9139375476009137,
"grad_norm": 1.1954008340835571,
"learning_rate": 2.764211633660252e-06,
"loss": 1.1209,
"step": 1276
},
{
"epoch": 2.9162223914699164,
"grad_norm": 1.1329401731491089,
"learning_rate": 2.7611146820713187e-06,
"loss": 1.1394,
"step": 1277
},
{
"epoch": 2.9185072353389185,
"grad_norm": 1.153887391090393,
"learning_rate": 2.7580173253090876e-06,
"loss": 1.1236,
"step": 1278
},
{
"epoch": 2.9207920792079207,
"grad_norm": 1.2343658208847046,
"learning_rate": 2.754919568179746e-06,
"loss": 1.1582,
"step": 1279
},
{
"epoch": 2.9230769230769234,
"grad_norm": 1.1790735721588135,
"learning_rate": 2.7518214154901025e-06,
"loss": 1.148,
"step": 1280
},
{
"epoch": 2.925361766945925,
"grad_norm": 1.1787333488464355,
"learning_rate": 2.7487228720475812e-06,
"loss": 1.0793,
"step": 1281
},
{
"epoch": 2.9276466108149277,
"grad_norm": 1.1795237064361572,
"learning_rate": 2.745623942660211e-06,
"loss": 1.1811,
"step": 1282
},
{
"epoch": 2.92993145468393,
"grad_norm": 1.1928547620773315,
"learning_rate": 2.7425246321366205e-06,
"loss": 1.1207,
"step": 1283
},
{
"epoch": 2.932216298552932,
"grad_norm": 1.2683132886886597,
"learning_rate": 2.7394249452860296e-06,
"loss": 1.1868,
"step": 1284
},
{
"epoch": 2.9345011424219347,
"grad_norm": 1.2315943241119385,
"learning_rate": 2.7363248869182407e-06,
"loss": 1.1732,
"step": 1285
},
{
"epoch": 2.936785986290937,
"grad_norm": 1.178771734237671,
"learning_rate": 2.7332244618436355e-06,
"loss": 1.1377,
"step": 1286
},
{
"epoch": 2.939070830159939,
"grad_norm": 1.1433113813400269,
"learning_rate": 2.7301236748731623e-06,
"loss": 1.1556,
"step": 1287
},
{
"epoch": 2.9413556740289413,
"grad_norm": 1.1987029314041138,
"learning_rate": 2.7270225308183318e-06,
"loss": 1.16,
"step": 1288
},
{
"epoch": 2.9436405178979435,
"grad_norm": 1.1818667650222778,
"learning_rate": 2.7239210344912085e-06,
"loss": 1.1443,
"step": 1289
},
{
"epoch": 2.945925361766946,
"grad_norm": 1.2523951530456543,
"learning_rate": 2.720819190704405e-06,
"loss": 1.1512,
"step": 1290
},
{
"epoch": 2.9482102056359483,
"grad_norm": 1.186334490776062,
"learning_rate": 2.7177170042710706e-06,
"loss": 1.1347,
"step": 1291
},
{
"epoch": 2.9504950495049505,
"grad_norm": 1.1355229616165161,
"learning_rate": 2.714614480004888e-06,
"loss": 1.1717,
"step": 1292
},
{
"epoch": 2.9527798933739526,
"grad_norm": 1.1879637241363525,
"learning_rate": 2.7115116227200634e-06,
"loss": 1.1737,
"step": 1293
},
{
"epoch": 2.955064737242955,
"grad_norm": 1.160522699356079,
"learning_rate": 2.7084084372313207e-06,
"loss": 1.1603,
"step": 1294
},
{
"epoch": 2.9573495811119574,
"grad_norm": 1.15951406955719,
"learning_rate": 2.705304928353892e-06,
"loss": 1.1809,
"step": 1295
},
{
"epoch": 2.9596344249809596,
"grad_norm": 1.132550835609436,
"learning_rate": 2.702201100903511e-06,
"loss": 1.1537,
"step": 1296
},
{
"epoch": 2.961919268849962,
"grad_norm": 1.1578466892242432,
"learning_rate": 2.6990969596964066e-06,
"loss": 1.1578,
"step": 1297
},
{
"epoch": 2.9642041127189644,
"grad_norm": 1.1534026861190796,
"learning_rate": 2.6959925095492957e-06,
"loss": 1.1392,
"step": 1298
},
{
"epoch": 2.9664889565879666,
"grad_norm": 1.1865522861480713,
"learning_rate": 2.6928877552793716e-06,
"loss": 1.1292,
"step": 1299
},
{
"epoch": 2.968773800456969,
"grad_norm": 1.1986700296401978,
"learning_rate": 2.689782701704301e-06,
"loss": 1.1591,
"step": 1300
},
{
"epoch": 2.971058644325971,
"grad_norm": 1.1916639804840088,
"learning_rate": 2.6866773536422157e-06,
"loss": 1.1757,
"step": 1301
},
{
"epoch": 2.973343488194973,
"grad_norm": 1.1540440320968628,
"learning_rate": 2.6835717159117044e-06,
"loss": 1.132,
"step": 1302
},
{
"epoch": 2.975628332063976,
"grad_norm": 1.175986647605896,
"learning_rate": 2.6804657933318035e-06,
"loss": 1.158,
"step": 1303
},
{
"epoch": 2.977913175932978,
"grad_norm": 1.1727259159088135,
"learning_rate": 2.6773595907219937e-06,
"loss": 1.1454,
"step": 1304
},
{
"epoch": 2.98019801980198,
"grad_norm": 1.1897666454315186,
"learning_rate": 2.674253112902189e-06,
"loss": 1.1385,
"step": 1305
},
{
"epoch": 2.9824828636709824,
"grad_norm": 1.2285549640655518,
"learning_rate": 2.6711463646927296e-06,
"loss": 1.1057,
"step": 1306
},
{
"epoch": 2.9847677075399845,
"grad_norm": 1.1508749723434448,
"learning_rate": 2.668039350914377e-06,
"loss": 1.1537,
"step": 1307
},
{
"epoch": 2.987052551408987,
"grad_norm": 1.1747997999191284,
"learning_rate": 2.6649320763883045e-06,
"loss": 1.1447,
"step": 1308
},
{
"epoch": 2.9893373952779894,
"grad_norm": 1.1757659912109375,
"learning_rate": 2.6618245459360896e-06,
"loss": 1.1312,
"step": 1309
},
{
"epoch": 2.9916222391469915,
"grad_norm": 1.1855767965316772,
"learning_rate": 2.658716764379706e-06,
"loss": 1.1559,
"step": 1310
},
{
"epoch": 2.993907083015994,
"grad_norm": 1.1734507083892822,
"learning_rate": 2.6556087365415183e-06,
"loss": 1.1291,
"step": 1311
},
{
"epoch": 2.996191926884996,
"grad_norm": 1.1714792251586914,
"learning_rate": 2.6525004672442734e-06,
"loss": 1.1225,
"step": 1312
},
{
"epoch": 2.9984767707539985,
"grad_norm": 1.1722139120101929,
"learning_rate": 2.6493919613110923e-06,
"loss": 1.103,
"step": 1313
},
{
"epoch": 3.0,
"grad_norm": 1.6373590230941772,
"learning_rate": 2.6462832235654625e-06,
"loss": 1.1512,
"step": 1314
},
{
"epoch": 3.002284843869002,
"grad_norm": 1.1781227588653564,
"learning_rate": 2.643174258831232e-06,
"loss": 1.1117,
"step": 1315
},
{
"epoch": 3.0045696877380044,
"grad_norm": 1.1707483530044556,
"learning_rate": 2.6400650719326014e-06,
"loss": 1.0957,
"step": 1316
},
{
"epoch": 3.006854531607007,
"grad_norm": 1.192963719367981,
"learning_rate": 2.6369556676941165e-06,
"loss": 1.1014,
"step": 1317
},
{
"epoch": 3.009139375476009,
"grad_norm": 1.1995853185653687,
"learning_rate": 2.6338460509406577e-06,
"loss": 1.1066,
"step": 1318
},
{
"epoch": 3.0114242193450114,
"grad_norm": 1.1763046979904175,
"learning_rate": 2.630736226497438e-06,
"loss": 1.1006,
"step": 1319
},
{
"epoch": 3.0137090632140136,
"grad_norm": 1.2747207880020142,
"learning_rate": 2.6276261991899915e-06,
"loss": 1.1031,
"step": 1320
},
{
"epoch": 3.015993907083016,
"grad_norm": 1.2675756216049194,
"learning_rate": 2.6245159738441673e-06,
"loss": 1.1416,
"step": 1321
},
{
"epoch": 3.0182787509520184,
"grad_norm": 1.1917039155960083,
"learning_rate": 2.6214055552861213e-06,
"loss": 1.1025,
"step": 1322
},
{
"epoch": 3.0205635948210205,
"grad_norm": 1.153493881225586,
"learning_rate": 2.618294948342309e-06,
"loss": 1.0846,
"step": 1323
},
{
"epoch": 3.0228484386900227,
"grad_norm": 1.2151880264282227,
"learning_rate": 2.61518415783948e-06,
"loss": 1.0835,
"step": 1324
},
{
"epoch": 3.025133282559025,
"grad_norm": 1.2547942399978638,
"learning_rate": 2.6120731886046673e-06,
"loss": 1.1196,
"step": 1325
},
{
"epoch": 3.0274181264280275,
"grad_norm": 1.2668670415878296,
"learning_rate": 2.608962045465181e-06,
"loss": 1.0744,
"step": 1326
},
{
"epoch": 3.0297029702970297,
"grad_norm": 1.2131301164627075,
"learning_rate": 2.6058507332486014e-06,
"loss": 1.1415,
"step": 1327
},
{
"epoch": 3.031987814166032,
"grad_norm": 1.230054259300232,
"learning_rate": 2.602739256782772e-06,
"loss": 1.0854,
"step": 1328
},
{
"epoch": 3.034272658035034,
"grad_norm": 1.1794737577438354,
"learning_rate": 2.599627620895791e-06,
"loss": 1.1035,
"step": 1329
},
{
"epoch": 3.0365575019040367,
"grad_norm": 1.1808216571807861,
"learning_rate": 2.5965158304160017e-06,
"loss": 1.1099,
"step": 1330
},
{
"epoch": 3.038842345773039,
"grad_norm": 1.1999027729034424,
"learning_rate": 2.593403890171989e-06,
"loss": 1.0721,
"step": 1331
},
{
"epoch": 3.041127189642041,
"grad_norm": 1.2132296562194824,
"learning_rate": 2.5902918049925718e-06,
"loss": 1.0796,
"step": 1332
},
{
"epoch": 3.0434120335110433,
"grad_norm": 1.2564518451690674,
"learning_rate": 2.5871795797067924e-06,
"loss": 1.1248,
"step": 1333
},
{
"epoch": 3.045696877380046,
"grad_norm": 1.2996151447296143,
"learning_rate": 2.584067219143908e-06,
"loss": 1.1724,
"step": 1334
},
{
"epoch": 3.047981721249048,
"grad_norm": 1.2047386169433594,
"learning_rate": 2.5809547281333904e-06,
"loss": 1.1231,
"step": 1335
},
{
"epoch": 3.0502665651180503,
"grad_norm": 1.1555238962173462,
"learning_rate": 2.57784211150491e-06,
"loss": 1.1404,
"step": 1336
},
{
"epoch": 3.0525514089870525,
"grad_norm": 1.1934109926223755,
"learning_rate": 2.5747293740883355e-06,
"loss": 1.1408,
"step": 1337
},
{
"epoch": 3.0548362528560546,
"grad_norm": 1.2881134748458862,
"learning_rate": 2.5716165207137196e-06,
"loss": 1.0963,
"step": 1338
},
{
"epoch": 3.0571210967250573,
"grad_norm": 1.2293163537979126,
"learning_rate": 2.5685035562112964e-06,
"loss": 1.0969,
"step": 1339
},
{
"epoch": 3.0594059405940595,
"grad_norm": 1.1550770998001099,
"learning_rate": 2.5653904854114735e-06,
"loss": 1.0955,
"step": 1340
},
{
"epoch": 3.0616907844630616,
"grad_norm": 1.1883620023727417,
"learning_rate": 2.562277313144823e-06,
"loss": 1.1117,
"step": 1341
},
{
"epoch": 3.063975628332064,
"grad_norm": 1.179228663444519,
"learning_rate": 2.559164044242072e-06,
"loss": 1.1012,
"step": 1342
},
{
"epoch": 3.0662604722010665,
"grad_norm": 1.3026498556137085,
"learning_rate": 2.5560506835341003e-06,
"loss": 1.1422,
"step": 1343
},
{
"epoch": 3.0685453160700686,
"grad_norm": 1.2500083446502686,
"learning_rate": 2.5529372358519307e-06,
"loss": 1.1184,
"step": 1344
},
{
"epoch": 3.070830159939071,
"grad_norm": 1.2166169881820679,
"learning_rate": 2.549823706026717e-06,
"loss": 1.0977,
"step": 1345
},
{
"epoch": 3.073115003808073,
"grad_norm": 1.183247685432434,
"learning_rate": 2.546710098889746e-06,
"loss": 1.1148,
"step": 1346
},
{
"epoch": 3.075399847677075,
"grad_norm": 1.185694694519043,
"learning_rate": 2.543596419272419e-06,
"loss": 1.1361,
"step": 1347
},
{
"epoch": 3.077684691546078,
"grad_norm": 1.1907778978347778,
"learning_rate": 2.5404826720062544e-06,
"loss": 1.1144,
"step": 1348
},
{
"epoch": 3.07996953541508,
"grad_norm": 1.2550450563430786,
"learning_rate": 2.5373688619228713e-06,
"loss": 1.1247,
"step": 1349
},
{
"epoch": 3.082254379284082,
"grad_norm": 1.2171252965927124,
"learning_rate": 2.53425499385399e-06,
"loss": 1.116,
"step": 1350
},
{
"epoch": 3.0845392231530844,
"grad_norm": 1.1736383438110352,
"learning_rate": 2.5311410726314183e-06,
"loss": 1.1364,
"step": 1351
},
{
"epoch": 3.086824067022087,
"grad_norm": 1.1960982084274292,
"learning_rate": 2.528027103087049e-06,
"loss": 1.0907,
"step": 1352
},
{
"epoch": 3.089108910891089,
"grad_norm": 1.1903541088104248,
"learning_rate": 2.5249130900528464e-06,
"loss": 1.08,
"step": 1353
},
{
"epoch": 3.0913937547600914,
"grad_norm": 1.2436158657073975,
"learning_rate": 2.5217990383608448e-06,
"loss": 1.0973,
"step": 1354
},
{
"epoch": 3.0936785986290936,
"grad_norm": 1.2662099599838257,
"learning_rate": 2.518684952843138e-06,
"loss": 1.1526,
"step": 1355
},
{
"epoch": 3.095963442498096,
"grad_norm": 1.1999155282974243,
"learning_rate": 2.5155708383318726e-06,
"loss": 1.1421,
"step": 1356
},
{
"epoch": 3.0982482863670984,
"grad_norm": 1.2298964262008667,
"learning_rate": 2.5124566996592388e-06,
"loss": 1.1227,
"step": 1357
},
{
"epoch": 3.1005331302361006,
"grad_norm": 1.2199616432189941,
"learning_rate": 2.509342541657465e-06,
"loss": 1.111,
"step": 1358
},
{
"epoch": 3.1028179741051027,
"grad_norm": 1.1966962814331055,
"learning_rate": 2.506228369158811e-06,
"loss": 1.0991,
"step": 1359
},
{
"epoch": 3.105102817974105,
"grad_norm": 1.2209211587905884,
"learning_rate": 2.503114186995558e-06,
"loss": 1.0993,
"step": 1360
},
{
"epoch": 3.1073876618431076,
"grad_norm": 1.1826435327529907,
"learning_rate": 2.5e-06,
"loss": 1.1074,
"step": 1361
},
{
"epoch": 3.1096725057121097,
"grad_norm": 1.1526274681091309,
"learning_rate": 2.496885813004443e-06,
"loss": 1.106,
"step": 1362
},
{
"epoch": 3.111957349581112,
"grad_norm": 1.248950481414795,
"learning_rate": 2.4937716308411896e-06,
"loss": 1.1285,
"step": 1363
},
{
"epoch": 3.114242193450114,
"grad_norm": 1.2252432107925415,
"learning_rate": 2.4906574583425357e-06,
"loss": 1.1145,
"step": 1364
},
{
"epoch": 3.1165270373191167,
"grad_norm": 1.2389261722564697,
"learning_rate": 2.487543300340762e-06,
"loss": 1.1149,
"step": 1365
},
{
"epoch": 3.118811881188119,
"grad_norm": 1.1963584423065186,
"learning_rate": 2.4844291616681283e-06,
"loss": 1.1128,
"step": 1366
},
{
"epoch": 3.121096725057121,
"grad_norm": 1.1800618171691895,
"learning_rate": 2.4813150471568624e-06,
"loss": 1.1178,
"step": 1367
},
{
"epoch": 3.1233815689261233,
"grad_norm": 1.2248040437698364,
"learning_rate": 2.4782009616391556e-06,
"loss": 1.1204,
"step": 1368
},
{
"epoch": 3.1256664127951255,
"grad_norm": 1.2025740146636963,
"learning_rate": 2.475086909947154e-06,
"loss": 1.1018,
"step": 1369
},
{
"epoch": 3.127951256664128,
"grad_norm": 1.1733136177062988,
"learning_rate": 2.4719728969129517e-06,
"loss": 1.0994,
"step": 1370
},
{
"epoch": 3.1302361005331303,
"grad_norm": 1.2059153318405151,
"learning_rate": 2.468858927368582e-06,
"loss": 1.1036,
"step": 1371
},
{
"epoch": 3.1325209444021325,
"grad_norm": 1.1862396001815796,
"learning_rate": 2.465745006146011e-06,
"loss": 1.1116,
"step": 1372
},
{
"epoch": 3.1348057882711347,
"grad_norm": 1.163080096244812,
"learning_rate": 2.462631138077129e-06,
"loss": 1.1233,
"step": 1373
},
{
"epoch": 3.1370906321401373,
"grad_norm": 1.2319633960723877,
"learning_rate": 2.4595173279937464e-06,
"loss": 1.1132,
"step": 1374
},
{
"epoch": 3.1393754760091395,
"grad_norm": 1.2363816499710083,
"learning_rate": 2.456403580727582e-06,
"loss": 1.1246,
"step": 1375
},
{
"epoch": 3.1416603198781416,
"grad_norm": 1.272760272026062,
"learning_rate": 2.4532899011102553e-06,
"loss": 1.1538,
"step": 1376
},
{
"epoch": 3.143945163747144,
"grad_norm": 1.2700350284576416,
"learning_rate": 2.4501762939732834e-06,
"loss": 1.1289,
"step": 1377
},
{
"epoch": 3.146230007616146,
"grad_norm": 1.213341474533081,
"learning_rate": 2.4470627641480705e-06,
"loss": 1.0985,
"step": 1378
},
{
"epoch": 3.1485148514851486,
"grad_norm": 1.1804118156433105,
"learning_rate": 2.4439493164659005e-06,
"loss": 1.1479,
"step": 1379
},
{
"epoch": 3.150799695354151,
"grad_norm": 1.2232515811920166,
"learning_rate": 2.4408359557579294e-06,
"loss": 1.1518,
"step": 1380
},
{
"epoch": 3.153084539223153,
"grad_norm": 1.1940504312515259,
"learning_rate": 2.4377226868551783e-06,
"loss": 1.1409,
"step": 1381
},
{
"epoch": 3.155369383092155,
"grad_norm": 1.1823115348815918,
"learning_rate": 2.434609514588527e-06,
"loss": 1.0937,
"step": 1382
},
{
"epoch": 3.157654226961158,
"grad_norm": 1.2245404720306396,
"learning_rate": 2.4314964437887044e-06,
"loss": 1.1219,
"step": 1383
},
{
"epoch": 3.15993907083016,
"grad_norm": 1.1697916984558105,
"learning_rate": 2.4283834792862817e-06,
"loss": 1.1052,
"step": 1384
},
{
"epoch": 3.162223914699162,
"grad_norm": 1.2614245414733887,
"learning_rate": 2.4252706259116657e-06,
"loss": 1.1095,
"step": 1385
},
{
"epoch": 3.1645087585681644,
"grad_norm": 1.2271556854248047,
"learning_rate": 2.4221578884950903e-06,
"loss": 1.1381,
"step": 1386
},
{
"epoch": 3.166793602437167,
"grad_norm": 1.2251800298690796,
"learning_rate": 2.419045271866611e-06,
"loss": 1.1286,
"step": 1387
},
{
"epoch": 3.169078446306169,
"grad_norm": 1.2880722284317017,
"learning_rate": 2.415932780856093e-06,
"loss": 1.1236,
"step": 1388
},
{
"epoch": 3.1713632901751714,
"grad_norm": 1.240476131439209,
"learning_rate": 2.4128204202932093e-06,
"loss": 1.1369,
"step": 1389
},
{
"epoch": 3.1736481340441736,
"grad_norm": 1.246504545211792,
"learning_rate": 2.409708195007429e-06,
"loss": 1.1193,
"step": 1390
},
{
"epoch": 3.1759329779131757,
"grad_norm": 1.251848578453064,
"learning_rate": 2.406596109828012e-06,
"loss": 1.1451,
"step": 1391
},
{
"epoch": 3.1782178217821784,
"grad_norm": 1.2193578481674194,
"learning_rate": 2.403484169584e-06,
"loss": 1.1143,
"step": 1392
},
{
"epoch": 3.1805026656511806,
"grad_norm": 1.217989206314087,
"learning_rate": 2.4003723791042107e-06,
"loss": 1.0959,
"step": 1393
},
{
"epoch": 3.1827875095201827,
"grad_norm": 1.2069993019104004,
"learning_rate": 2.3972607432172283e-06,
"loss": 1.1417,
"step": 1394
},
{
"epoch": 3.185072353389185,
"grad_norm": 1.21833074092865,
"learning_rate": 2.3941492667514e-06,
"loss": 1.1575,
"step": 1395
},
{
"epoch": 3.1873571972581876,
"grad_norm": 1.2144832611083984,
"learning_rate": 2.3910379545348203e-06,
"loss": 1.1602,
"step": 1396
},
{
"epoch": 3.1896420411271897,
"grad_norm": 1.229985237121582,
"learning_rate": 2.387926811395334e-06,
"loss": 1.1053,
"step": 1397
},
{
"epoch": 3.191926884996192,
"grad_norm": 1.237524151802063,
"learning_rate": 2.384815842160521e-06,
"loss": 1.1359,
"step": 1398
},
{
"epoch": 3.194211728865194,
"grad_norm": 1.2233461141586304,
"learning_rate": 2.381705051657692e-06,
"loss": 1.1003,
"step": 1399
},
{
"epoch": 3.1964965727341963,
"grad_norm": 1.2777819633483887,
"learning_rate": 2.3785944447138804e-06,
"loss": 1.1138,
"step": 1400
},
{
"epoch": 3.198781416603199,
"grad_norm": 1.2171168327331543,
"learning_rate": 2.3754840261558336e-06,
"loss": 1.1149,
"step": 1401
},
{
"epoch": 3.201066260472201,
"grad_norm": 1.2730437517166138,
"learning_rate": 2.372373800810009e-06,
"loss": 1.1022,
"step": 1402
},
{
"epoch": 3.2033511043412033,
"grad_norm": 1.2138127088546753,
"learning_rate": 2.369263773502562e-06,
"loss": 1.1267,
"step": 1403
},
{
"epoch": 3.2056359482102055,
"grad_norm": 1.2073695659637451,
"learning_rate": 2.3661539490593423e-06,
"loss": 1.1151,
"step": 1404
},
{
"epoch": 3.207920792079208,
"grad_norm": 1.2361245155334473,
"learning_rate": 2.3630443323058843e-06,
"loss": 1.1224,
"step": 1405
},
{
"epoch": 3.2102056359482103,
"grad_norm": 1.2753217220306396,
"learning_rate": 2.3599349280673985e-06,
"loss": 1.1328,
"step": 1406
},
{
"epoch": 3.2124904798172125,
"grad_norm": 1.182378888130188,
"learning_rate": 2.356825741168768e-06,
"loss": 1.1256,
"step": 1407
},
{
"epoch": 3.2147753236862147,
"grad_norm": 1.20160710811615,
"learning_rate": 2.353716776434538e-06,
"loss": 1.1306,
"step": 1408
},
{
"epoch": 3.217060167555217,
"grad_norm": 1.1793500185012817,
"learning_rate": 2.3506080386889086e-06,
"loss": 1.0768,
"step": 1409
},
{
"epoch": 3.2193450114242195,
"grad_norm": 1.2438910007476807,
"learning_rate": 2.3474995327557266e-06,
"loss": 1.1582,
"step": 1410
},
{
"epoch": 3.2216298552932217,
"grad_norm": 1.2653754949569702,
"learning_rate": 2.3443912634584817e-06,
"loss": 1.0914,
"step": 1411
},
{
"epoch": 3.223914699162224,
"grad_norm": 1.237338662147522,
"learning_rate": 2.3412832356202943e-06,
"loss": 1.0945,
"step": 1412
},
{
"epoch": 3.226199543031226,
"grad_norm": 1.262833595275879,
"learning_rate": 2.3381754540639108e-06,
"loss": 1.1204,
"step": 1413
},
{
"epoch": 3.2284843869002287,
"grad_norm": 1.2164596319198608,
"learning_rate": 2.3350679236116955e-06,
"loss": 1.1442,
"step": 1414
},
{
"epoch": 3.230769230769231,
"grad_norm": 1.1981042623519897,
"learning_rate": 2.331960649085623e-06,
"loss": 1.0826,
"step": 1415
},
{
"epoch": 3.233054074638233,
"grad_norm": 1.289568305015564,
"learning_rate": 2.3288536353072704e-06,
"loss": 1.1027,
"step": 1416
},
{
"epoch": 3.235338918507235,
"grad_norm": 1.3058327436447144,
"learning_rate": 2.325746887097812e-06,
"loss": 1.1548,
"step": 1417
},
{
"epoch": 3.237623762376238,
"grad_norm": 1.2601237297058105,
"learning_rate": 2.3226404092780067e-06,
"loss": 1.1143,
"step": 1418
},
{
"epoch": 3.23990860624524,
"grad_norm": 1.2124943733215332,
"learning_rate": 2.3195342066681965e-06,
"loss": 1.1185,
"step": 1419
},
{
"epoch": 3.242193450114242,
"grad_norm": 1.2472761869430542,
"learning_rate": 2.316428284088296e-06,
"loss": 1.1504,
"step": 1420
},
{
"epoch": 3.2444782939832444,
"grad_norm": 1.2425949573516846,
"learning_rate": 2.3133226463577847e-06,
"loss": 1.1339,
"step": 1421
},
{
"epoch": 3.2467631378522466,
"grad_norm": 1.3503245115280151,
"learning_rate": 2.3102172982957e-06,
"loss": 1.1633,
"step": 1422
},
{
"epoch": 3.249047981721249,
"grad_norm": 1.2171080112457275,
"learning_rate": 2.3071122447206292e-06,
"loss": 1.0951,
"step": 1423
},
{
"epoch": 3.2513328255902514,
"grad_norm": 1.2225638628005981,
"learning_rate": 2.3040074904507047e-06,
"loss": 1.0744,
"step": 1424
},
{
"epoch": 3.2536176694592536,
"grad_norm": 1.209750771522522,
"learning_rate": 2.300903040303594e-06,
"loss": 1.1328,
"step": 1425
},
{
"epoch": 3.2559025133282558,
"grad_norm": 1.207655668258667,
"learning_rate": 2.29779889909649e-06,
"loss": 1.1473,
"step": 1426
},
{
"epoch": 3.258187357197258,
"grad_norm": 1.1904206275939941,
"learning_rate": 2.2946950716461094e-06,
"loss": 1.1165,
"step": 1427
},
{
"epoch": 3.2604722010662606,
"grad_norm": 1.2216424942016602,
"learning_rate": 2.2915915627686797e-06,
"loss": 1.1054,
"step": 1428
},
{
"epoch": 3.2627570449352628,
"grad_norm": 1.2471741437911987,
"learning_rate": 2.2884883772799375e-06,
"loss": 1.1189,
"step": 1429
},
{
"epoch": 3.265041888804265,
"grad_norm": 1.2008448839187622,
"learning_rate": 2.285385519995113e-06,
"loss": 1.1131,
"step": 1430
},
{
"epoch": 3.2673267326732676,
"grad_norm": 1.2054619789123535,
"learning_rate": 2.2822829957289303e-06,
"loss": 1.0846,
"step": 1431
},
{
"epoch": 3.2696115765422697,
"grad_norm": 1.2204397916793823,
"learning_rate": 2.2791808092955957e-06,
"loss": 1.1471,
"step": 1432
},
{
"epoch": 3.271896420411272,
"grad_norm": 1.2435766458511353,
"learning_rate": 2.2760789655087923e-06,
"loss": 1.1167,
"step": 1433
},
{
"epoch": 3.274181264280274,
"grad_norm": 1.2288284301757812,
"learning_rate": 2.272977469181669e-06,
"loss": 1.1297,
"step": 1434
},
{
"epoch": 3.2764661081492763,
"grad_norm": 1.2264825105667114,
"learning_rate": 2.2698763251268386e-06,
"loss": 1.1011,
"step": 1435
},
{
"epoch": 3.278750952018279,
"grad_norm": 1.2367379665374756,
"learning_rate": 2.266775538156365e-06,
"loss": 1.1069,
"step": 1436
},
{
"epoch": 3.281035795887281,
"grad_norm": 1.201027274131775,
"learning_rate": 2.26367511308176e-06,
"loss": 1.1203,
"step": 1437
},
{
"epoch": 3.2833206397562833,
"grad_norm": 1.2541130781173706,
"learning_rate": 2.2605750547139716e-06,
"loss": 1.1234,
"step": 1438
},
{
"epoch": 3.2856054836252855,
"grad_norm": 1.2395600080490112,
"learning_rate": 2.25747536786338e-06,
"loss": 1.1031,
"step": 1439
},
{
"epoch": 3.2878903274942877,
"grad_norm": 1.2372959852218628,
"learning_rate": 2.254376057339789e-06,
"loss": 1.155,
"step": 1440
},
{
"epoch": 3.2901751713632903,
"grad_norm": 1.2344226837158203,
"learning_rate": 2.25127712795242e-06,
"loss": 1.1239,
"step": 1441
},
{
"epoch": 3.2924600152322925,
"grad_norm": 1.1589802503585815,
"learning_rate": 2.2481785845098988e-06,
"loss": 1.088,
"step": 1442
},
{
"epoch": 3.2947448591012947,
"grad_norm": 1.2531473636627197,
"learning_rate": 2.245080431820255e-06,
"loss": 1.1412,
"step": 1443
},
{
"epoch": 3.297029702970297,
"grad_norm": 1.2219151258468628,
"learning_rate": 2.241982674690913e-06,
"loss": 1.1138,
"step": 1444
},
{
"epoch": 3.2993145468392995,
"grad_norm": 1.2261440753936768,
"learning_rate": 2.2388853179286817e-06,
"loss": 1.102,
"step": 1445
},
{
"epoch": 3.3015993907083017,
"grad_norm": 1.2558192014694214,
"learning_rate": 2.2357883663397485e-06,
"loss": 1.159,
"step": 1446
},
{
"epoch": 3.303884234577304,
"grad_norm": 1.2133665084838867,
"learning_rate": 2.232691824729673e-06,
"loss": 1.1163,
"step": 1447
},
{
"epoch": 3.306169078446306,
"grad_norm": 1.212631106376648,
"learning_rate": 2.2295956979033786e-06,
"loss": 1.102,
"step": 1448
},
{
"epoch": 3.3084539223153087,
"grad_norm": 1.2700281143188477,
"learning_rate": 2.226499990665145e-06,
"loss": 1.0938,
"step": 1449
},
{
"epoch": 3.310738766184311,
"grad_norm": 1.2251685857772827,
"learning_rate": 2.2234047078185997e-06,
"loss": 1.1142,
"step": 1450
},
{
"epoch": 3.313023610053313,
"grad_norm": 1.223984956741333,
"learning_rate": 2.2203098541667136e-06,
"loss": 1.1267,
"step": 1451
},
{
"epoch": 3.315308453922315,
"grad_norm": 1.2254488468170166,
"learning_rate": 2.2172154345117896e-06,
"loss": 1.1202,
"step": 1452
},
{
"epoch": 3.3175932977913174,
"grad_norm": 1.2008295059204102,
"learning_rate": 2.2141214536554594e-06,
"loss": 1.1317,
"step": 1453
},
{
"epoch": 3.31987814166032,
"grad_norm": 1.2000782489776611,
"learning_rate": 2.211027916398672e-06,
"loss": 1.1056,
"step": 1454
},
{
"epoch": 3.322162985529322,
"grad_norm": 1.2368084192276,
"learning_rate": 2.207934827541688e-06,
"loss": 1.1003,
"step": 1455
},
{
"epoch": 3.3244478293983244,
"grad_norm": 1.246311068534851,
"learning_rate": 2.204842191884074e-06,
"loss": 1.1092,
"step": 1456
},
{
"epoch": 3.3267326732673266,
"grad_norm": 1.2252134084701538,
"learning_rate": 2.201750014224692e-06,
"loss": 1.116,
"step": 1457
},
{
"epoch": 3.329017517136329,
"grad_norm": 1.2413280010223389,
"learning_rate": 2.1986582993616926e-06,
"loss": 1.1387,
"step": 1458
},
{
"epoch": 3.3313023610053314,
"grad_norm": 1.2302745580673218,
"learning_rate": 2.1955670520925105e-06,
"loss": 1.0876,
"step": 1459
},
{
"epoch": 3.3335872048743336,
"grad_norm": 1.2019201517105103,
"learning_rate": 2.192476277213853e-06,
"loss": 1.0869,
"step": 1460
},
{
"epoch": 3.3358720487433358,
"grad_norm": 1.2414172887802124,
"learning_rate": 2.189385979521696e-06,
"loss": 1.113,
"step": 1461
},
{
"epoch": 3.3381568926123384,
"grad_norm": 1.2234618663787842,
"learning_rate": 2.186296163811272e-06,
"loss": 1.0967,
"step": 1462
},
{
"epoch": 3.3404417364813406,
"grad_norm": 1.2113620042800903,
"learning_rate": 2.183206834877068e-06,
"loss": 1.1307,
"step": 1463
},
{
"epoch": 3.3427265803503428,
"grad_norm": 1.2364552021026611,
"learning_rate": 2.1801179975128162e-06,
"loss": 1.0811,
"step": 1464
},
{
"epoch": 3.345011424219345,
"grad_norm": 1.2303386926651,
"learning_rate": 2.1770296565114847e-06,
"loss": 1.1306,
"step": 1465
},
{
"epoch": 3.347296268088347,
"grad_norm": 1.2606472969055176,
"learning_rate": 2.1739418166652712e-06,
"loss": 1.1429,
"step": 1466
},
{
"epoch": 3.3495811119573498,
"grad_norm": 1.211198091506958,
"learning_rate": 2.1708544827655963e-06,
"loss": 1.1043,
"step": 1467
},
{
"epoch": 3.351865955826352,
"grad_norm": 1.2446391582489014,
"learning_rate": 2.1677676596030963e-06,
"loss": 1.1265,
"step": 1468
},
{
"epoch": 3.354150799695354,
"grad_norm": 1.1910406351089478,
"learning_rate": 2.1646813519676136e-06,
"loss": 1.0555,
"step": 1469
},
{
"epoch": 3.3564356435643563,
"grad_norm": 1.2461347579956055,
"learning_rate": 2.1615955646481905e-06,
"loss": 1.1319,
"step": 1470
},
{
"epoch": 3.3587204874333585,
"grad_norm": 1.2559690475463867,
"learning_rate": 2.158510302433064e-06,
"loss": 1.1291,
"step": 1471
},
{
"epoch": 3.361005331302361,
"grad_norm": 1.209735631942749,
"learning_rate": 2.155425570109654e-06,
"loss": 1.1482,
"step": 1472
},
{
"epoch": 3.3632901751713633,
"grad_norm": 1.2300963401794434,
"learning_rate": 2.1523413724645593e-06,
"loss": 1.1061,
"step": 1473
},
{
"epoch": 3.3655750190403655,
"grad_norm": 1.2133387327194214,
"learning_rate": 2.149257714283549e-06,
"loss": 1.1067,
"step": 1474
},
{
"epoch": 3.3678598629093677,
"grad_norm": 1.228406310081482,
"learning_rate": 2.1461746003515547e-06,
"loss": 1.1392,
"step": 1475
},
{
"epoch": 3.3701447067783703,
"grad_norm": 1.2098608016967773,
"learning_rate": 2.143092035452664e-06,
"loss": 1.1148,
"step": 1476
},
{
"epoch": 3.3724295506473725,
"grad_norm": 1.2088247537612915,
"learning_rate": 2.140010024370112e-06,
"loss": 1.0957,
"step": 1477
},
{
"epoch": 3.3747143945163747,
"grad_norm": 1.2421760559082031,
"learning_rate": 2.136928571886275e-06,
"loss": 1.1514,
"step": 1478
},
{
"epoch": 3.376999238385377,
"grad_norm": 1.2158358097076416,
"learning_rate": 2.133847682782662e-06,
"loss": 1.1306,
"step": 1479
},
{
"epoch": 3.3792840822543795,
"grad_norm": 1.242077350616455,
"learning_rate": 2.1307673618399078e-06,
"loss": 1.1236,
"step": 1480
},
{
"epoch": 3.3815689261233817,
"grad_norm": 1.2471917867660522,
"learning_rate": 2.1276876138377662e-06,
"loss": 1.1039,
"step": 1481
},
{
"epoch": 3.383853769992384,
"grad_norm": 1.2467808723449707,
"learning_rate": 2.124608443555102e-06,
"loss": 1.1047,
"step": 1482
},
{
"epoch": 3.386138613861386,
"grad_norm": 1.2468382120132446,
"learning_rate": 2.1215298557698834e-06,
"loss": 1.1494,
"step": 1483
},
{
"epoch": 3.388423457730388,
"grad_norm": 1.211882472038269,
"learning_rate": 2.118451855259173e-06,
"loss": 1.0893,
"step": 1484
},
{
"epoch": 3.390708301599391,
"grad_norm": 1.219880223274231,
"learning_rate": 2.1153744467991243e-06,
"loss": 1.1203,
"step": 1485
},
{
"epoch": 3.392993145468393,
"grad_norm": 1.2225430011749268,
"learning_rate": 2.112297635164972e-06,
"loss": 1.1183,
"step": 1486
},
{
"epoch": 3.395277989337395,
"grad_norm": 1.229330062866211,
"learning_rate": 2.1092214251310235e-06,
"loss": 1.129,
"step": 1487
},
{
"epoch": 3.3975628332063974,
"grad_norm": 1.2349112033843994,
"learning_rate": 2.1061458214706536e-06,
"loss": 1.1212,
"step": 1488
},
{
"epoch": 3.3998476770754,
"grad_norm": 1.222354531288147,
"learning_rate": 2.103070828956295e-06,
"loss": 1.1216,
"step": 1489
},
{
"epoch": 3.402132520944402,
"grad_norm": 1.21956205368042,
"learning_rate": 2.0999964523594343e-06,
"loss": 1.1289,
"step": 1490
},
{
"epoch": 3.4044173648134044,
"grad_norm": 1.2225059270858765,
"learning_rate": 2.0969226964506007e-06,
"loss": 1.0957,
"step": 1491
},
{
"epoch": 3.4067022086824066,
"grad_norm": 1.2281376123428345,
"learning_rate": 2.09384956599936e-06,
"loss": 1.1346,
"step": 1492
},
{
"epoch": 3.408987052551409,
"grad_norm": 1.262031078338623,
"learning_rate": 2.090777065774308e-06,
"loss": 1.1269,
"step": 1493
},
{
"epoch": 3.4112718964204114,
"grad_norm": 1.2495571374893188,
"learning_rate": 2.087705200543064e-06,
"loss": 1.0793,
"step": 1494
},
{
"epoch": 3.4135567402894136,
"grad_norm": 1.1917531490325928,
"learning_rate": 2.0846339750722596e-06,
"loss": 1.0997,
"step": 1495
},
{
"epoch": 3.4158415841584158,
"grad_norm": 1.2409162521362305,
"learning_rate": 2.0815633941275345e-06,
"loss": 1.1022,
"step": 1496
},
{
"epoch": 3.418126428027418,
"grad_norm": 1.2216984033584595,
"learning_rate": 2.078493462473529e-06,
"loss": 1.1114,
"step": 1497
},
{
"epoch": 3.4204112718964206,
"grad_norm": 1.3940880298614502,
"learning_rate": 2.0754241848738755e-06,
"loss": 1.1224,
"step": 1498
},
{
"epoch": 3.4226961157654228,
"grad_norm": 1.1798889636993408,
"learning_rate": 2.0723555660911914e-06,
"loss": 1.1202,
"step": 1499
},
{
"epoch": 3.424980959634425,
"grad_norm": 1.2345131635665894,
"learning_rate": 2.0692876108870713e-06,
"loss": 1.1367,
"step": 1500
},
{
"epoch": 3.427265803503427,
"grad_norm": 1.2564736604690552,
"learning_rate": 2.066220324022081e-06,
"loss": 1.1096,
"step": 1501
},
{
"epoch": 3.4295506473724293,
"grad_norm": 1.2020221948623657,
"learning_rate": 2.063153710255749e-06,
"loss": 1.1245,
"step": 1502
},
{
"epoch": 3.431835491241432,
"grad_norm": 1.217174768447876,
"learning_rate": 2.0600877743465596e-06,
"loss": 1.0887,
"step": 1503
},
{
"epoch": 3.434120335110434,
"grad_norm": 1.3399721384048462,
"learning_rate": 2.0570225210519433e-06,
"loss": 1.1233,
"step": 1504
},
{
"epoch": 3.4364051789794363,
"grad_norm": 1.2029287815093994,
"learning_rate": 2.0539579551282745e-06,
"loss": 1.1148,
"step": 1505
},
{
"epoch": 3.4386900228484385,
"grad_norm": 1.2201212644577026,
"learning_rate": 2.050894081330859e-06,
"loss": 1.1215,
"step": 1506
},
{
"epoch": 3.440974866717441,
"grad_norm": 1.2642061710357666,
"learning_rate": 2.0478309044139293e-06,
"loss": 1.1203,
"step": 1507
},
{
"epoch": 3.4432597105864433,
"grad_norm": 1.225867748260498,
"learning_rate": 2.0447684291306354e-06,
"loss": 1.1241,
"step": 1508
},
{
"epoch": 3.4455445544554455,
"grad_norm": 1.2447335720062256,
"learning_rate": 2.0417066602330402e-06,
"loss": 1.1254,
"step": 1509
},
{
"epoch": 3.4478293983244477,
"grad_norm": 1.2231181859970093,
"learning_rate": 2.0386456024721102e-06,
"loss": 1.1325,
"step": 1510
},
{
"epoch": 3.4501142421934503,
"grad_norm": 1.2640868425369263,
"learning_rate": 2.035585260597708e-06,
"loss": 1.094,
"step": 1511
},
{
"epoch": 3.4523990860624525,
"grad_norm": 1.2731873989105225,
"learning_rate": 2.0325256393585846e-06,
"loss": 1.1422,
"step": 1512
},
{
"epoch": 3.4546839299314547,
"grad_norm": 1.2461851835250854,
"learning_rate": 2.029466743502374e-06,
"loss": 1.1663,
"step": 1513
},
{
"epoch": 3.456968773800457,
"grad_norm": 1.2134253978729248,
"learning_rate": 2.026408577775584e-06,
"loss": 1.114,
"step": 1514
},
{
"epoch": 3.459253617669459,
"grad_norm": 1.2513582706451416,
"learning_rate": 2.02335114692359e-06,
"loss": 1.1553,
"step": 1515
},
{
"epoch": 3.4615384615384617,
"grad_norm": 1.226331114768982,
"learning_rate": 2.0202944556906265e-06,
"loss": 1.15,
"step": 1516
},
{
"epoch": 3.463823305407464,
"grad_norm": 1.1947654485702515,
"learning_rate": 2.0172385088197804e-06,
"loss": 1.139,
"step": 1517
},
{
"epoch": 3.466108149276466,
"grad_norm": 1.2045434713363647,
"learning_rate": 2.014183311052984e-06,
"loss": 1.0923,
"step": 1518
},
{
"epoch": 3.4683929931454682,
"grad_norm": 1.2555713653564453,
"learning_rate": 2.011128867131007e-06,
"loss": 1.1475,
"step": 1519
},
{
"epoch": 3.470677837014471,
"grad_norm": 1.2348402738571167,
"learning_rate": 2.008075181793449e-06,
"loss": 1.1061,
"step": 1520
},
{
"epoch": 3.472962680883473,
"grad_norm": 1.3485833406448364,
"learning_rate": 2.005022259778733e-06,
"loss": 1.1452,
"step": 1521
},
{
"epoch": 3.4752475247524752,
"grad_norm": 1.2618969678878784,
"learning_rate": 2.0019701058240973e-06,
"loss": 1.1021,
"step": 1522
},
{
"epoch": 3.4775323686214774,
"grad_norm": 1.262932300567627,
"learning_rate": 1.998918724665589e-06,
"loss": 1.1169,
"step": 1523
},
{
"epoch": 3.47981721249048,
"grad_norm": 1.190786600112915,
"learning_rate": 1.995868121038054e-06,
"loss": 1.1142,
"step": 1524
},
{
"epoch": 3.4821020563594822,
"grad_norm": 1.2386054992675781,
"learning_rate": 1.992818299675134e-06,
"loss": 1.1276,
"step": 1525
},
{
"epoch": 3.4843869002284844,
"grad_norm": 1.3026626110076904,
"learning_rate": 1.9897692653092566e-06,
"loss": 1.1079,
"step": 1526
},
{
"epoch": 3.4866717440974866,
"grad_norm": 1.2284817695617676,
"learning_rate": 1.986721022671627e-06,
"loss": 1.1326,
"step": 1527
},
{
"epoch": 3.4889565879664888,
"grad_norm": 1.2360379695892334,
"learning_rate": 1.983673576492223e-06,
"loss": 1.1193,
"step": 1528
},
{
"epoch": 3.4912414318354914,
"grad_norm": 1.232100248336792,
"learning_rate": 1.980626931499785e-06,
"loss": 1.1155,
"step": 1529
},
{
"epoch": 3.4935262757044936,
"grad_norm": 1.261894702911377,
"learning_rate": 1.9775810924218126e-06,
"loss": 1.1127,
"step": 1530
},
{
"epoch": 3.4958111195734958,
"grad_norm": 1.3262462615966797,
"learning_rate": 1.974536063984553e-06,
"loss": 1.1199,
"step": 1531
},
{
"epoch": 3.498095963442498,
"grad_norm": 1.255900502204895,
"learning_rate": 1.9714918509129956e-06,
"loss": 1.1239,
"step": 1532
},
{
"epoch": 3.5003808073115,
"grad_norm": 1.2368223667144775,
"learning_rate": 1.968448457930865e-06,
"loss": 1.1097,
"step": 1533
},
{
"epoch": 3.5026656511805028,
"grad_norm": 1.2056398391723633,
"learning_rate": 1.9654058897606143e-06,
"loss": 1.091,
"step": 1534
},
{
"epoch": 3.504950495049505,
"grad_norm": 1.2451571226119995,
"learning_rate": 1.9623641511234145e-06,
"loss": 1.0939,
"step": 1535
},
{
"epoch": 3.507235338918507,
"grad_norm": 1.2372757196426392,
"learning_rate": 1.959323246739151e-06,
"loss": 1.0783,
"step": 1536
},
{
"epoch": 3.5095201827875098,
"grad_norm": 1.2244229316711426,
"learning_rate": 1.956283181326415e-06,
"loss": 1.0861,
"step": 1537
},
{
"epoch": 3.511805026656512,
"grad_norm": 1.2195265293121338,
"learning_rate": 1.9532439596024942e-06,
"loss": 1.1434,
"step": 1538
},
{
"epoch": 3.514089870525514,
"grad_norm": 1.2447067499160767,
"learning_rate": 1.9502055862833686e-06,
"loss": 1.103,
"step": 1539
},
{
"epoch": 3.5163747143945163,
"grad_norm": 1.2313573360443115,
"learning_rate": 1.947168066083701e-06,
"loss": 1.1518,
"step": 1540
},
{
"epoch": 3.5186595582635185,
"grad_norm": 1.2975883483886719,
"learning_rate": 1.9441314037168307e-06,
"loss": 1.1102,
"step": 1541
},
{
"epoch": 3.520944402132521,
"grad_norm": 1.2976816892623901,
"learning_rate": 1.941095603894767e-06,
"loss": 1.1408,
"step": 1542
},
{
"epoch": 3.5232292460015233,
"grad_norm": 1.3002218008041382,
"learning_rate": 1.9380606713281773e-06,
"loss": 1.125,
"step": 1543
},
{
"epoch": 3.5255140898705255,
"grad_norm": 1.225427508354187,
"learning_rate": 1.935026610726387e-06,
"loss": 1.0761,
"step": 1544
},
{
"epoch": 3.5277989337395277,
"grad_norm": 1.235769510269165,
"learning_rate": 1.931993426797367e-06,
"loss": 1.122,
"step": 1545
},
{
"epoch": 3.53008377760853,
"grad_norm": 1.2742570638656616,
"learning_rate": 1.9289611242477284e-06,
"loss": 1.0982,
"step": 1546
},
{
"epoch": 3.5323686214775325,
"grad_norm": 1.2655776739120483,
"learning_rate": 1.9259297077827134e-06,
"loss": 1.1124,
"step": 1547
},
{
"epoch": 3.5346534653465347,
"grad_norm": 1.1953189373016357,
"learning_rate": 1.92289918210619e-06,
"loss": 1.0874,
"step": 1548
},
{
"epoch": 3.536938309215537,
"grad_norm": 1.21706223487854,
"learning_rate": 1.9198695519206446e-06,
"loss": 1.1172,
"step": 1549
},
{
"epoch": 3.5392231530845395,
"grad_norm": 1.2621902227401733,
"learning_rate": 1.916840821927173e-06,
"loss": 1.1191,
"step": 1550
},
{
"epoch": 3.5415079969535412,
"grad_norm": 1.2300255298614502,
"learning_rate": 1.913812996825475e-06,
"loss": 1.1316,
"step": 1551
},
{
"epoch": 3.543792840822544,
"grad_norm": 1.238046407699585,
"learning_rate": 1.910786081313845e-06,
"loss": 1.1013,
"step": 1552
},
{
"epoch": 3.546077684691546,
"grad_norm": 1.2778140306472778,
"learning_rate": 1.907760080089168e-06,
"loss": 1.1419,
"step": 1553
},
{
"epoch": 3.5483625285605482,
"grad_norm": 1.211096167564392,
"learning_rate": 1.904734997846909e-06,
"loss": 1.1061,
"step": 1554
},
{
"epoch": 3.550647372429551,
"grad_norm": 1.293352484703064,
"learning_rate": 1.9017108392811065e-06,
"loss": 1.1339,
"step": 1555
},
{
"epoch": 3.552932216298553,
"grad_norm": 1.2381699085235596,
"learning_rate": 1.8986876090843668e-06,
"loss": 1.1017,
"step": 1556
},
{
"epoch": 3.5552170601675552,
"grad_norm": 1.2436529397964478,
"learning_rate": 1.8956653119478552e-06,
"loss": 1.1027,
"step": 1557
},
{
"epoch": 3.5575019040365574,
"grad_norm": 1.2499804496765137,
"learning_rate": 1.89264395256129e-06,
"loss": 1.1445,
"step": 1558
},
{
"epoch": 3.5597867479055596,
"grad_norm": 1.194018840789795,
"learning_rate": 1.889623535612931e-06,
"loss": 1.11,
"step": 1559
},
{
"epoch": 3.5620715917745622,
"grad_norm": 1.1932462453842163,
"learning_rate": 1.8866040657895816e-06,
"loss": 1.1377,
"step": 1560
},
{
"epoch": 3.5643564356435644,
"grad_norm": 1.254319667816162,
"learning_rate": 1.883585547776571e-06,
"loss": 1.1212,
"step": 1561
},
{
"epoch": 3.5666412795125666,
"grad_norm": 1.2139774560928345,
"learning_rate": 1.8805679862577519e-06,
"loss": 1.0974,
"step": 1562
},
{
"epoch": 3.568926123381569,
"grad_norm": 1.2384382486343384,
"learning_rate": 1.8775513859154937e-06,
"loss": 1.1471,
"step": 1563
},
{
"epoch": 3.571210967250571,
"grad_norm": 1.2261179685592651,
"learning_rate": 1.8745357514306754e-06,
"loss": 1.1437,
"step": 1564
},
{
"epoch": 3.5734958111195736,
"grad_norm": 1.2182419300079346,
"learning_rate": 1.8715210874826745e-06,
"loss": 1.1039,
"step": 1565
},
{
"epoch": 3.575780654988576,
"grad_norm": 1.2260403633117676,
"learning_rate": 1.868507398749364e-06,
"loss": 1.1445,
"step": 1566
},
{
"epoch": 3.578065498857578,
"grad_norm": 1.255858302116394,
"learning_rate": 1.8654946899071052e-06,
"loss": 1.1592,
"step": 1567
},
{
"epoch": 3.5803503427265806,
"grad_norm": 1.248203992843628,
"learning_rate": 1.8624829656307364e-06,
"loss": 1.0864,
"step": 1568
},
{
"epoch": 3.5826351865955828,
"grad_norm": 1.2393451929092407,
"learning_rate": 1.8594722305935691e-06,
"loss": 1.1392,
"step": 1569
},
{
"epoch": 3.584920030464585,
"grad_norm": 1.317401647567749,
"learning_rate": 1.85646248946738e-06,
"loss": 1.0891,
"step": 1570
},
{
"epoch": 3.587204874333587,
"grad_norm": 1.198999285697937,
"learning_rate": 1.8534537469224035e-06,
"loss": 1.1002,
"step": 1571
},
{
"epoch": 3.5894897182025893,
"grad_norm": 1.2434310913085938,
"learning_rate": 1.8504460076273247e-06,
"loss": 1.1502,
"step": 1572
},
{
"epoch": 3.591774562071592,
"grad_norm": 1.1986984014511108,
"learning_rate": 1.847439276249271e-06,
"loss": 1.0861,
"step": 1573
},
{
"epoch": 3.594059405940594,
"grad_norm": 1.2537792921066284,
"learning_rate": 1.8444335574538063e-06,
"loss": 1.12,
"step": 1574
},
{
"epoch": 3.5963442498095963,
"grad_norm": 1.2118167877197266,
"learning_rate": 1.8414288559049236e-06,
"loss": 1.1455,
"step": 1575
},
{
"epoch": 3.5986290936785985,
"grad_norm": 1.2423104047775269,
"learning_rate": 1.8384251762650384e-06,
"loss": 1.1503,
"step": 1576
},
{
"epoch": 3.6009139375476007,
"grad_norm": 1.2497063875198364,
"learning_rate": 1.8354225231949773e-06,
"loss": 1.1163,
"step": 1577
},
{
"epoch": 3.6031987814166033,
"grad_norm": 1.2626383304595947,
"learning_rate": 1.8324209013539775e-06,
"loss": 1.0694,
"step": 1578
},
{
"epoch": 3.6054836252856055,
"grad_norm": 1.2420138120651245,
"learning_rate": 1.8294203153996739e-06,
"loss": 1.1202,
"step": 1579
},
{
"epoch": 3.6077684691546077,
"grad_norm": 1.1990013122558594,
"learning_rate": 1.8264207699880953e-06,
"loss": 1.1155,
"step": 1580
},
{
"epoch": 3.6100533130236103,
"grad_norm": 1.2583009004592896,
"learning_rate": 1.8234222697736542e-06,
"loss": 1.1337,
"step": 1581
},
{
"epoch": 3.612338156892612,
"grad_norm": 1.257130742073059,
"learning_rate": 1.8204248194091429e-06,
"loss": 1.1192,
"step": 1582
},
{
"epoch": 3.6146230007616147,
"grad_norm": 1.2139629125595093,
"learning_rate": 1.817428423545724e-06,
"loss": 1.1196,
"step": 1583
},
{
"epoch": 3.616907844630617,
"grad_norm": 1.2340987920761108,
"learning_rate": 1.8144330868329244e-06,
"loss": 1.1476,
"step": 1584
},
{
"epoch": 3.619192688499619,
"grad_norm": 1.2045873403549194,
"learning_rate": 1.8114388139186257e-06,
"loss": 1.1168,
"step": 1585
},
{
"epoch": 3.6214775323686217,
"grad_norm": 1.2221472263336182,
"learning_rate": 1.8084456094490607e-06,
"loss": 1.1362,
"step": 1586
},
{
"epoch": 3.623762376237624,
"grad_norm": 1.1829921007156372,
"learning_rate": 1.8054534780688038e-06,
"loss": 1.0699,
"step": 1587
},
{
"epoch": 3.626047220106626,
"grad_norm": 1.2308509349822998,
"learning_rate": 1.8024624244207644e-06,
"loss": 1.0782,
"step": 1588
},
{
"epoch": 3.6283320639756282,
"grad_norm": 1.2337130308151245,
"learning_rate": 1.7994724531461788e-06,
"loss": 1.1058,
"step": 1589
},
{
"epoch": 3.6306169078446304,
"grad_norm": 1.219441533088684,
"learning_rate": 1.796483568884604e-06,
"loss": 1.0977,
"step": 1590
},
{
"epoch": 3.632901751713633,
"grad_norm": 1.2322226762771606,
"learning_rate": 1.7934957762739108e-06,
"loss": 1.1019,
"step": 1591
},
{
"epoch": 3.6351865955826352,
"grad_norm": 1.271531105041504,
"learning_rate": 1.7905090799502766e-06,
"loss": 1.1092,
"step": 1592
},
{
"epoch": 3.6374714394516374,
"grad_norm": 1.269434928894043,
"learning_rate": 1.7875234845481761e-06,
"loss": 1.149,
"step": 1593
},
{
"epoch": 3.6397562833206396,
"grad_norm": 1.2015925645828247,
"learning_rate": 1.7845389947003766e-06,
"loss": 1.1274,
"step": 1594
},
{
"epoch": 3.642041127189642,
"grad_norm": 1.217633843421936,
"learning_rate": 1.7815556150379298e-06,
"loss": 1.1014,
"step": 1595
},
{
"epoch": 3.6443259710586444,
"grad_norm": 1.2365998029708862,
"learning_rate": 1.7785733501901654e-06,
"loss": 1.1053,
"step": 1596
},
{
"epoch": 3.6466108149276466,
"grad_norm": 1.193575143814087,
"learning_rate": 1.7755922047846812e-06,
"loss": 1.1157,
"step": 1597
},
{
"epoch": 3.648895658796649,
"grad_norm": 1.2818313837051392,
"learning_rate": 1.7726121834473406e-06,
"loss": 1.0953,
"step": 1598
},
{
"epoch": 3.6511805026656514,
"grad_norm": 1.2339566946029663,
"learning_rate": 1.7696332908022607e-06,
"loss": 1.1119,
"step": 1599
},
{
"epoch": 3.6534653465346536,
"grad_norm": 1.2497464418411255,
"learning_rate": 1.7666555314718093e-06,
"loss": 1.1107,
"step": 1600
},
{
"epoch": 3.655750190403656,
"grad_norm": 1.2073878049850464,
"learning_rate": 1.7636789100765922e-06,
"loss": 1.0952,
"step": 1601
},
{
"epoch": 3.658035034272658,
"grad_norm": 1.287691354751587,
"learning_rate": 1.7607034312354531e-06,
"loss": 1.1212,
"step": 1602
},
{
"epoch": 3.66031987814166,
"grad_norm": 1.2657358646392822,
"learning_rate": 1.7577290995654611e-06,
"loss": 1.0946,
"step": 1603
},
{
"epoch": 3.662604722010663,
"grad_norm": 1.2619253396987915,
"learning_rate": 1.7547559196819056e-06,
"loss": 1.1182,
"step": 1604
},
{
"epoch": 3.664889565879665,
"grad_norm": 1.2667503356933594,
"learning_rate": 1.7517838961982876e-06,
"loss": 1.1519,
"step": 1605
},
{
"epoch": 3.667174409748667,
"grad_norm": 1.3911538124084473,
"learning_rate": 1.7488130337263153e-06,
"loss": 1.1396,
"step": 1606
},
{
"epoch": 3.6694592536176693,
"grad_norm": 1.267244815826416,
"learning_rate": 1.745843336875895e-06,
"loss": 1.1137,
"step": 1607
},
{
"epoch": 3.6717440974866715,
"grad_norm": 1.242616057395935,
"learning_rate": 1.7428748102551237e-06,
"loss": 1.1537,
"step": 1608
},
{
"epoch": 3.674028941355674,
"grad_norm": 1.3466277122497559,
"learning_rate": 1.7399074584702827e-06,
"loss": 1.1017,
"step": 1609
},
{
"epoch": 3.6763137852246763,
"grad_norm": 1.269391655921936,
"learning_rate": 1.7369412861258307e-06,
"loss": 1.1189,
"step": 1610
},
{
"epoch": 3.6785986290936785,
"grad_norm": 1.2533386945724487,
"learning_rate": 1.733976297824396e-06,
"loss": 1.151,
"step": 1611
},
{
"epoch": 3.680883472962681,
"grad_norm": 1.2193639278411865,
"learning_rate": 1.73101249816677e-06,
"loss": 1.1344,
"step": 1612
},
{
"epoch": 3.6831683168316833,
"grad_norm": 1.24978768825531,
"learning_rate": 1.7280498917518987e-06,
"loss": 1.1214,
"step": 1613
},
{
"epoch": 3.6854531607006855,
"grad_norm": 1.272676706314087,
"learning_rate": 1.7250884831768773e-06,
"loss": 1.1371,
"step": 1614
},
{
"epoch": 3.6877380045696877,
"grad_norm": 1.2939724922180176,
"learning_rate": 1.722128277036942e-06,
"loss": 1.153,
"step": 1615
},
{
"epoch": 3.69002284843869,
"grad_norm": 1.2304813861846924,
"learning_rate": 1.7191692779254646e-06,
"loss": 1.0707,
"step": 1616
},
{
"epoch": 3.6923076923076925,
"grad_norm": 1.2000212669372559,
"learning_rate": 1.716211490433941e-06,
"loss": 1.1075,
"step": 1617
},
{
"epoch": 3.6945925361766947,
"grad_norm": 1.2244328260421753,
"learning_rate": 1.7132549191519893e-06,
"loss": 1.0832,
"step": 1618
},
{
"epoch": 3.696877380045697,
"grad_norm": 1.2140921354293823,
"learning_rate": 1.7102995686673399e-06,
"loss": 1.1065,
"step": 1619
},
{
"epoch": 3.699162223914699,
"grad_norm": 1.2332017421722412,
"learning_rate": 1.7073454435658294e-06,
"loss": 1.134,
"step": 1620
},
{
"epoch": 3.7014470677837013,
"grad_norm": 1.306863784790039,
"learning_rate": 1.7043925484313911e-06,
"loss": 1.1334,
"step": 1621
},
{
"epoch": 3.703731911652704,
"grad_norm": 1.248136281967163,
"learning_rate": 1.7014408878460514e-06,
"loss": 1.1586,
"step": 1622
},
{
"epoch": 3.706016755521706,
"grad_norm": 1.2233672142028809,
"learning_rate": 1.698490466389921e-06,
"loss": 1.1328,
"step": 1623
},
{
"epoch": 3.7083015993907082,
"grad_norm": 1.2379913330078125,
"learning_rate": 1.6955412886411882e-06,
"loss": 1.1071,
"step": 1624
},
{
"epoch": 3.7105864432597104,
"grad_norm": 1.2285481691360474,
"learning_rate": 1.692593359176109e-06,
"loss": 1.1009,
"step": 1625
},
{
"epoch": 3.7128712871287126,
"grad_norm": 1.2411075830459595,
"learning_rate": 1.689646682569005e-06,
"loss": 1.149,
"step": 1626
},
{
"epoch": 3.7151561309977152,
"grad_norm": 1.2382875680923462,
"learning_rate": 1.6867012633922524e-06,
"loss": 1.1003,
"step": 1627
},
{
"epoch": 3.7174409748667174,
"grad_norm": 1.2481483221054077,
"learning_rate": 1.6837571062162783e-06,
"loss": 1.0967,
"step": 1628
},
{
"epoch": 3.7197258187357196,
"grad_norm": 1.2253104448318481,
"learning_rate": 1.680814215609548e-06,
"loss": 1.0991,
"step": 1629
},
{
"epoch": 3.7220106626047222,
"grad_norm": 1.240035057067871,
"learning_rate": 1.6778725961385642e-06,
"loss": 1.1111,
"step": 1630
},
{
"epoch": 3.7242955064737244,
"grad_norm": 1.254558801651001,
"learning_rate": 1.6749322523678562e-06,
"loss": 1.1436,
"step": 1631
},
{
"epoch": 3.7265803503427266,
"grad_norm": 1.2620891332626343,
"learning_rate": 1.6719931888599749e-06,
"loss": 1.1519,
"step": 1632
},
{
"epoch": 3.728865194211729,
"grad_norm": 1.2564220428466797,
"learning_rate": 1.6690554101754824e-06,
"loss": 1.1506,
"step": 1633
},
{
"epoch": 3.731150038080731,
"grad_norm": 1.1939340829849243,
"learning_rate": 1.6661189208729492e-06,
"loss": 1.1191,
"step": 1634
},
{
"epoch": 3.7334348819497336,
"grad_norm": 1.2243472337722778,
"learning_rate": 1.6631837255089437e-06,
"loss": 1.1417,
"step": 1635
},
{
"epoch": 3.735719725818736,
"grad_norm": 1.261639952659607,
"learning_rate": 1.660249828638028e-06,
"loss": 1.0566,
"step": 1636
},
{
"epoch": 3.738004569687738,
"grad_norm": 1.2218812704086304,
"learning_rate": 1.657317234812746e-06,
"loss": 1.0979,
"step": 1637
},
{
"epoch": 3.74028941355674,
"grad_norm": 1.243190884590149,
"learning_rate": 1.6543859485836252e-06,
"loss": 1.0964,
"step": 1638
},
{
"epoch": 3.7425742574257423,
"grad_norm": 1.2348755598068237,
"learning_rate": 1.6514559744991592e-06,
"loss": 1.1623,
"step": 1639
},
{
"epoch": 3.744859101294745,
"grad_norm": 1.1991803646087646,
"learning_rate": 1.6485273171058074e-06,
"loss": 1.1372,
"step": 1640
},
{
"epoch": 3.747143945163747,
"grad_norm": 1.2339205741882324,
"learning_rate": 1.645599980947986e-06,
"loss": 1.1225,
"step": 1641
},
{
"epoch": 3.7494287890327493,
"grad_norm": 1.2044298648834229,
"learning_rate": 1.642673970568062e-06,
"loss": 1.1325,
"step": 1642
},
{
"epoch": 3.751713632901752,
"grad_norm": 1.242209792137146,
"learning_rate": 1.6397492905063423e-06,
"loss": 1.1234,
"step": 1643
},
{
"epoch": 3.753998476770754,
"grad_norm": 1.2190120220184326,
"learning_rate": 1.6368259453010723e-06,
"loss": 1.1428,
"step": 1644
},
{
"epoch": 3.7562833206397563,
"grad_norm": 1.2304496765136719,
"learning_rate": 1.6339039394884254e-06,
"loss": 1.1203,
"step": 1645
},
{
"epoch": 3.7585681645087585,
"grad_norm": 1.220682144165039,
"learning_rate": 1.6309832776024964e-06,
"loss": 1.1342,
"step": 1646
},
{
"epoch": 3.7608530083777607,
"grad_norm": 1.2088897228240967,
"learning_rate": 1.6280639641752944e-06,
"loss": 1.1027,
"step": 1647
},
{
"epoch": 3.7631378522467633,
"grad_norm": 1.2930717468261719,
"learning_rate": 1.6251460037367368e-06,
"loss": 1.1363,
"step": 1648
},
{
"epoch": 3.7654226961157655,
"grad_norm": 1.2154346704483032,
"learning_rate": 1.6222294008146405e-06,
"loss": 1.1303,
"step": 1649
},
{
"epoch": 3.7677075399847677,
"grad_norm": 1.2410285472869873,
"learning_rate": 1.619314159934718e-06,
"loss": 1.0836,
"step": 1650
},
{
"epoch": 3.76999238385377,
"grad_norm": 1.2346444129943848,
"learning_rate": 1.6164002856205657e-06,
"loss": 1.0972,
"step": 1651
},
{
"epoch": 3.772277227722772,
"grad_norm": 1.241414189338684,
"learning_rate": 1.613487782393661e-06,
"loss": 1.1629,
"step": 1652
},
{
"epoch": 3.7745620715917747,
"grad_norm": 1.216896891593933,
"learning_rate": 1.6105766547733537e-06,
"loss": 1.1095,
"step": 1653
},
{
"epoch": 3.776846915460777,
"grad_norm": 1.2589776515960693,
"learning_rate": 1.6076669072768597e-06,
"loss": 1.0993,
"step": 1654
},
{
"epoch": 3.779131759329779,
"grad_norm": 1.238924503326416,
"learning_rate": 1.6047585444192509e-06,
"loss": 1.1178,
"step": 1655
},
{
"epoch": 3.7814166031987813,
"grad_norm": 1.2369179725646973,
"learning_rate": 1.601851570713453e-06,
"loss": 1.1309,
"step": 1656
},
{
"epoch": 3.7837014470677834,
"grad_norm": 1.211485743522644,
"learning_rate": 1.598945990670236e-06,
"loss": 1.1047,
"step": 1657
},
{
"epoch": 3.785986290936786,
"grad_norm": 1.2450000047683716,
"learning_rate": 1.5960418087982067e-06,
"loss": 1.1116,
"step": 1658
},
{
"epoch": 3.7882711348057883,
"grad_norm": 1.2489582300186157,
"learning_rate": 1.5931390296038014e-06,
"loss": 1.1134,
"step": 1659
},
{
"epoch": 3.7905559786747904,
"grad_norm": 1.234943151473999,
"learning_rate": 1.5902376575912815e-06,
"loss": 1.0784,
"step": 1660
},
{
"epoch": 3.792840822543793,
"grad_norm": 1.2086708545684814,
"learning_rate": 1.587337697262724e-06,
"loss": 1.1164,
"step": 1661
},
{
"epoch": 3.7951256664127953,
"grad_norm": 1.2318179607391357,
"learning_rate": 1.5844391531180167e-06,
"loss": 1.1039,
"step": 1662
},
{
"epoch": 3.7974105102817974,
"grad_norm": 1.2397037744522095,
"learning_rate": 1.5815420296548476e-06,
"loss": 1.1214,
"step": 1663
},
{
"epoch": 3.7996953541507996,
"grad_norm": 1.2473440170288086,
"learning_rate": 1.5786463313687016e-06,
"loss": 1.1327,
"step": 1664
},
{
"epoch": 3.801980198019802,
"grad_norm": 1.2378629446029663,
"learning_rate": 1.5757520627528522e-06,
"loss": 1.1323,
"step": 1665
},
{
"epoch": 3.8042650418888044,
"grad_norm": 1.2788701057434082,
"learning_rate": 1.5728592282983552e-06,
"loss": 1.0891,
"step": 1666
},
{
"epoch": 3.8065498857578066,
"grad_norm": 1.208809733390808,
"learning_rate": 1.569967832494038e-06,
"loss": 1.1107,
"step": 1667
},
{
"epoch": 3.808834729626809,
"grad_norm": 1.2094557285308838,
"learning_rate": 1.5670778798264997e-06,
"loss": 1.0916,
"step": 1668
},
{
"epoch": 3.811119573495811,
"grad_norm": 1.2201573848724365,
"learning_rate": 1.5641893747800968e-06,
"loss": 1.1196,
"step": 1669
},
{
"epoch": 3.813404417364813,
"grad_norm": 1.2208703756332397,
"learning_rate": 1.5613023218369416e-06,
"loss": 1.1172,
"step": 1670
},
{
"epoch": 3.815689261233816,
"grad_norm": 1.2877635955810547,
"learning_rate": 1.5584167254768918e-06,
"loss": 1.0968,
"step": 1671
},
{
"epoch": 3.817974105102818,
"grad_norm": 1.2119520902633667,
"learning_rate": 1.5555325901775453e-06,
"loss": 1.0905,
"step": 1672
},
{
"epoch": 3.82025894897182,
"grad_norm": 1.286763072013855,
"learning_rate": 1.5526499204142332e-06,
"loss": 1.1188,
"step": 1673
},
{
"epoch": 3.822543792840823,
"grad_norm": 1.273573637008667,
"learning_rate": 1.5497687206600134e-06,
"loss": 1.1156,
"step": 1674
},
{
"epoch": 3.824828636709825,
"grad_norm": 1.231566071510315,
"learning_rate": 1.5468889953856602e-06,
"loss": 1.0922,
"step": 1675
},
{
"epoch": 3.827113480578827,
"grad_norm": 1.228342890739441,
"learning_rate": 1.5440107490596623e-06,
"loss": 1.112,
"step": 1676
},
{
"epoch": 3.8293983244478293,
"grad_norm": 1.2194006443023682,
"learning_rate": 1.5411339861482121e-06,
"loss": 1.0727,
"step": 1677
},
{
"epoch": 3.8316831683168315,
"grad_norm": 1.2403062582015991,
"learning_rate": 1.538258711115202e-06,
"loss": 1.1327,
"step": 1678
},
{
"epoch": 3.833968012185834,
"grad_norm": 1.281432867050171,
"learning_rate": 1.5353849284222128e-06,
"loss": 1.1665,
"step": 1679
},
{
"epoch": 3.8362528560548363,
"grad_norm": 1.2705928087234497,
"learning_rate": 1.5325126425285122e-06,
"loss": 1.1195,
"step": 1680
},
{
"epoch": 3.8385376999238385,
"grad_norm": 1.2461146116256714,
"learning_rate": 1.5296418578910438e-06,
"loss": 1.1043,
"step": 1681
},
{
"epoch": 3.8408225437928407,
"grad_norm": 1.3064427375793457,
"learning_rate": 1.5267725789644232e-06,
"loss": 1.0931,
"step": 1682
},
{
"epoch": 3.843107387661843,
"grad_norm": 1.23982834815979,
"learning_rate": 1.5239048102009275e-06,
"loss": 1.1033,
"step": 1683
},
{
"epoch": 3.8453922315308455,
"grad_norm": 1.236802339553833,
"learning_rate": 1.5210385560504914e-06,
"loss": 1.1129,
"step": 1684
},
{
"epoch": 3.8476770753998477,
"grad_norm": 1.2742433547973633,
"learning_rate": 1.5181738209607006e-06,
"loss": 1.1648,
"step": 1685
},
{
"epoch": 3.84996191926885,
"grad_norm": 1.2181257009506226,
"learning_rate": 1.5153106093767827e-06,
"loss": 1.1038,
"step": 1686
},
{
"epoch": 3.852246763137852,
"grad_norm": 1.2690078020095825,
"learning_rate": 1.5124489257415998e-06,
"loss": 1.0908,
"step": 1687
},
{
"epoch": 3.8545316070068543,
"grad_norm": 1.2566394805908203,
"learning_rate": 1.509588774495645e-06,
"loss": 1.1146,
"step": 1688
},
{
"epoch": 3.856816450875857,
"grad_norm": 1.2618845701217651,
"learning_rate": 1.5067301600770337e-06,
"loss": 1.0876,
"step": 1689
},
{
"epoch": 3.859101294744859,
"grad_norm": 1.197676181793213,
"learning_rate": 1.5038730869214957e-06,
"loss": 1.1181,
"step": 1690
},
{
"epoch": 3.8613861386138613,
"grad_norm": 1.2773524522781372,
"learning_rate": 1.5010175594623688e-06,
"loss": 1.1646,
"step": 1691
},
{
"epoch": 3.863670982482864,
"grad_norm": 1.2414312362670898,
"learning_rate": 1.4981635821305934e-06,
"loss": 1.0974,
"step": 1692
},
{
"epoch": 3.865955826351866,
"grad_norm": 1.2599167823791504,
"learning_rate": 1.4953111593547038e-06,
"loss": 1.1338,
"step": 1693
},
{
"epoch": 3.8682406702208683,
"grad_norm": 1.2631725072860718,
"learning_rate": 1.4924602955608232e-06,
"loss": 1.1158,
"step": 1694
},
{
"epoch": 3.8705255140898704,
"grad_norm": 1.2844053506851196,
"learning_rate": 1.4896109951726542e-06,
"loss": 1.1067,
"step": 1695
},
{
"epoch": 3.8728103579588726,
"grad_norm": 1.236351490020752,
"learning_rate": 1.4867632626114742e-06,
"loss": 1.08,
"step": 1696
},
{
"epoch": 3.8750952018278753,
"grad_norm": 1.2237566709518433,
"learning_rate": 1.4839171022961279e-06,
"loss": 1.0924,
"step": 1697
},
{
"epoch": 3.8773800456968774,
"grad_norm": 1.2487142086029053,
"learning_rate": 1.4810725186430208e-06,
"loss": 1.1441,
"step": 1698
},
{
"epoch": 3.8796648895658796,
"grad_norm": 1.2455718517303467,
"learning_rate": 1.4782295160661103e-06,
"loss": 1.1043,
"step": 1699
},
{
"epoch": 3.881949733434882,
"grad_norm": 1.2438057661056519,
"learning_rate": 1.4753880989769014e-06,
"loss": 1.1375,
"step": 1700
},
{
"epoch": 3.884234577303884,
"grad_norm": 1.1872096061706543,
"learning_rate": 1.4725482717844397e-06,
"loss": 1.1017,
"step": 1701
},
{
"epoch": 3.8865194211728866,
"grad_norm": 1.303624153137207,
"learning_rate": 1.4697100388953033e-06,
"loss": 1.1154,
"step": 1702
},
{
"epoch": 3.888804265041889,
"grad_norm": 1.2645589113235474,
"learning_rate": 1.4668734047135947e-06,
"loss": 1.1316,
"step": 1703
},
{
"epoch": 3.891089108910891,
"grad_norm": 1.2637914419174194,
"learning_rate": 1.464038373640938e-06,
"loss": 1.1156,
"step": 1704
},
{
"epoch": 3.8933739527798936,
"grad_norm": 1.2358670234680176,
"learning_rate": 1.4612049500764685e-06,
"loss": 1.1031,
"step": 1705
},
{
"epoch": 3.895658796648896,
"grad_norm": 1.2059921026229858,
"learning_rate": 1.4583731384168275e-06,
"loss": 1.1147,
"step": 1706
},
{
"epoch": 3.897943640517898,
"grad_norm": 1.2406716346740723,
"learning_rate": 1.4555429430561552e-06,
"loss": 1.1354,
"step": 1707
},
{
"epoch": 3.9002284843869,
"grad_norm": 1.2390797138214111,
"learning_rate": 1.4527143683860822e-06,
"loss": 1.1389,
"step": 1708
},
{
"epoch": 3.9025133282559024,
"grad_norm": 1.236459732055664,
"learning_rate": 1.449887418795728e-06,
"loss": 1.0942,
"step": 1709
},
{
"epoch": 3.904798172124905,
"grad_norm": 1.2225269079208374,
"learning_rate": 1.4470620986716857e-06,
"loss": 1.105,
"step": 1710
},
{
"epoch": 3.907083015993907,
"grad_norm": 1.2086801528930664,
"learning_rate": 1.4442384123980247e-06,
"loss": 1.0923,
"step": 1711
},
{
"epoch": 3.9093678598629094,
"grad_norm": 1.2351250648498535,
"learning_rate": 1.4414163643562755e-06,
"loss": 1.1468,
"step": 1712
},
{
"epoch": 3.9116527037319115,
"grad_norm": 1.251083493232727,
"learning_rate": 1.4385959589254271e-06,
"loss": 1.104,
"step": 1713
},
{
"epoch": 3.9139375476009137,
"grad_norm": 1.2320407629013062,
"learning_rate": 1.4357772004819221e-06,
"loss": 1.0572,
"step": 1714
},
{
"epoch": 3.9162223914699164,
"grad_norm": 1.2320278882980347,
"learning_rate": 1.4329600933996452e-06,
"loss": 1.1187,
"step": 1715
},
{
"epoch": 3.9185072353389185,
"grad_norm": 1.242316722869873,
"learning_rate": 1.4301446420499165e-06,
"loss": 1.1347,
"step": 1716
},
{
"epoch": 3.9207920792079207,
"grad_norm": 1.2707537412643433,
"learning_rate": 1.4273308508014943e-06,
"loss": 1.0991,
"step": 1717
},
{
"epoch": 3.9230769230769234,
"grad_norm": 1.2760437726974487,
"learning_rate": 1.4245187240205537e-06,
"loss": 1.1327,
"step": 1718
},
{
"epoch": 3.925361766945925,
"grad_norm": 1.2317287921905518,
"learning_rate": 1.4217082660706893e-06,
"loss": 1.1126,
"step": 1719
},
{
"epoch": 3.9276466108149277,
"grad_norm": 1.2433048486709595,
"learning_rate": 1.4188994813129075e-06,
"loss": 1.0856,
"step": 1720
},
{
"epoch": 3.92993145468393,
"grad_norm": 1.241335391998291,
"learning_rate": 1.4160923741056159e-06,
"loss": 1.1551,
"step": 1721
},
{
"epoch": 3.932216298552932,
"grad_norm": 1.2815061807632446,
"learning_rate": 1.4132869488046224e-06,
"loss": 1.1107,
"step": 1722
},
{
"epoch": 3.9345011424219347,
"grad_norm": 1.237248420715332,
"learning_rate": 1.410483209763122e-06,
"loss": 1.1473,
"step": 1723
},
{
"epoch": 3.936785986290937,
"grad_norm": 1.2185921669006348,
"learning_rate": 1.4076811613316933e-06,
"loss": 1.1024,
"step": 1724
},
{
"epoch": 3.939070830159939,
"grad_norm": 1.2314651012420654,
"learning_rate": 1.4048808078582943e-06,
"loss": 1.1163,
"step": 1725
},
{
"epoch": 3.9413556740289413,
"grad_norm": 1.220934271812439,
"learning_rate": 1.4020821536882503e-06,
"loss": 1.1442,
"step": 1726
},
{
"epoch": 3.9436405178979435,
"grad_norm": 1.239525556564331,
"learning_rate": 1.3992852031642497e-06,
"loss": 1.1055,
"step": 1727
},
{
"epoch": 3.945925361766946,
"grad_norm": 1.2437052726745605,
"learning_rate": 1.3964899606263405e-06,
"loss": 1.1038,
"step": 1728
},
{
"epoch": 3.9482102056359483,
"grad_norm": 1.2381411790847778,
"learning_rate": 1.3936964304119158e-06,
"loss": 1.1116,
"step": 1729
},
{
"epoch": 3.9504950495049505,
"grad_norm": 1.2382675409317017,
"learning_rate": 1.390904616855716e-06,
"loss": 1.089,
"step": 1730
},
{
"epoch": 3.9527798933739526,
"grad_norm": 1.287566900253296,
"learning_rate": 1.388114524289815e-06,
"loss": 1.1342,
"step": 1731
},
{
"epoch": 3.955064737242955,
"grad_norm": 1.223980188369751,
"learning_rate": 1.385326157043616e-06,
"loss": 1.1262,
"step": 1732
},
{
"epoch": 3.9573495811119574,
"grad_norm": 1.2570685148239136,
"learning_rate": 1.3825395194438478e-06,
"loss": 1.1184,
"step": 1733
},
{
"epoch": 3.9596344249809596,
"grad_norm": 1.213205337524414,
"learning_rate": 1.379754615814553e-06,
"loss": 1.1161,
"step": 1734
},
{
"epoch": 3.961919268849962,
"grad_norm": 1.2163954973220825,
"learning_rate": 1.3769714504770825e-06,
"loss": 1.0887,
"step": 1735
},
{
"epoch": 3.9642041127189644,
"grad_norm": 1.2263234853744507,
"learning_rate": 1.374190027750094e-06,
"loss": 1.0997,
"step": 1736
},
{
"epoch": 3.9664889565879666,
"grad_norm": 1.2027472257614136,
"learning_rate": 1.3714103519495378e-06,
"loss": 1.0897,
"step": 1737
},
{
"epoch": 3.968773800456969,
"grad_norm": 1.250954270362854,
"learning_rate": 1.3686324273886531e-06,
"loss": 1.1171,
"step": 1738
},
{
"epoch": 3.971058644325971,
"grad_norm": 1.2648512125015259,
"learning_rate": 1.3658562583779644e-06,
"loss": 1.0782,
"step": 1739
},
{
"epoch": 3.973343488194973,
"grad_norm": 1.2568615674972534,
"learning_rate": 1.3630818492252695e-06,
"loss": 1.1025,
"step": 1740
},
{
"epoch": 3.975628332063976,
"grad_norm": 1.304031491279602,
"learning_rate": 1.3603092042356387e-06,
"loss": 1.0991,
"step": 1741
},
{
"epoch": 3.977913175932978,
"grad_norm": 1.263646125793457,
"learning_rate": 1.3575383277114007e-06,
"loss": 1.149,
"step": 1742
},
{
"epoch": 3.98019801980198,
"grad_norm": 1.2489157915115356,
"learning_rate": 1.354769223952142e-06,
"loss": 1.1303,
"step": 1743
},
{
"epoch": 3.9824828636709824,
"grad_norm": 1.2458829879760742,
"learning_rate": 1.3520018972547e-06,
"loss": 1.1047,
"step": 1744
},
{
"epoch": 3.9847677075399845,
"grad_norm": 1.2482655048370361,
"learning_rate": 1.3492363519131519e-06,
"loss": 1.1002,
"step": 1745
},
{
"epoch": 3.987052551408987,
"grad_norm": 1.2878859043121338,
"learning_rate": 1.3464725922188109e-06,
"loss": 1.1344,
"step": 1746
},
{
"epoch": 3.9893373952779894,
"grad_norm": 1.227795124053955,
"learning_rate": 1.3437106224602226e-06,
"loss": 1.1131,
"step": 1747
},
{
"epoch": 3.9916222391469915,
"grad_norm": 1.2362041473388672,
"learning_rate": 1.3409504469231507e-06,
"loss": 1.1014,
"step": 1748
},
{
"epoch": 3.993907083015994,
"grad_norm": 1.2575008869171143,
"learning_rate": 1.3381920698905788e-06,
"loss": 1.0759,
"step": 1749
},
{
"epoch": 3.996191926884996,
"grad_norm": 1.225038766860962,
"learning_rate": 1.3354354956426973e-06,
"loss": 1.1333,
"step": 1750
},
{
"epoch": 3.9984767707539985,
"grad_norm": 1.2729361057281494,
"learning_rate": 1.3326807284568984e-06,
"loss": 1.0851,
"step": 1751
},
{
"epoch": 4.0,
"grad_norm": 1.2729361057281494,
"learning_rate": 1.3299277726077742e-06,
"loss": 1.1285,
"step": 1752
},
{
"epoch": 4.002284843869003,
"grad_norm": 1.7938593626022339,
"learning_rate": 1.327176632367102e-06,
"loss": 1.1102,
"step": 1753
},
{
"epoch": 4.004569687738004,
"grad_norm": 1.2422128915786743,
"learning_rate": 1.3244273120038434e-06,
"loss": 1.0829,
"step": 1754
},
{
"epoch": 4.006854531607007,
"grad_norm": 1.2498319149017334,
"learning_rate": 1.3216798157841373e-06,
"loss": 1.0981,
"step": 1755
},
{
"epoch": 4.009139375476009,
"grad_norm": 1.2300645112991333,
"learning_rate": 1.3189341479712892e-06,
"loss": 1.0846,
"step": 1756
},
{
"epoch": 4.011424219345011,
"grad_norm": 1.227906346321106,
"learning_rate": 1.3161903128257714e-06,
"loss": 1.1038,
"step": 1757
},
{
"epoch": 4.013709063214014,
"grad_norm": 1.248955488204956,
"learning_rate": 1.3134483146052088e-06,
"loss": 1.0592,
"step": 1758
},
{
"epoch": 4.015993907083016,
"grad_norm": 1.2333277463912964,
"learning_rate": 1.3107081575643766e-06,
"loss": 1.0929,
"step": 1759
},
{
"epoch": 4.018278750952018,
"grad_norm": 1.2562737464904785,
"learning_rate": 1.3079698459551955e-06,
"loss": 1.1445,
"step": 1760
},
{
"epoch": 4.020563594821021,
"grad_norm": 1.2237579822540283,
"learning_rate": 1.30523338402672e-06,
"loss": 1.1176,
"step": 1761
},
{
"epoch": 4.022848438690023,
"grad_norm": 1.2542140483856201,
"learning_rate": 1.3024987760251345e-06,
"loss": 1.0441,
"step": 1762
},
{
"epoch": 4.025133282559025,
"grad_norm": 1.2324272394180298,
"learning_rate": 1.2997660261937489e-06,
"loss": 1.0705,
"step": 1763
},
{
"epoch": 4.027418126428027,
"grad_norm": 1.271546721458435,
"learning_rate": 1.2970351387729875e-06,
"loss": 1.0677,
"step": 1764
},
{
"epoch": 4.02970297029703,
"grad_norm": 1.2349035739898682,
"learning_rate": 1.2943061180003862e-06,
"loss": 1.0884,
"step": 1765
},
{
"epoch": 4.031987814166032,
"grad_norm": 1.266993761062622,
"learning_rate": 1.291578968110584e-06,
"loss": 1.1132,
"step": 1766
},
{
"epoch": 4.034272658035034,
"grad_norm": 1.2842568159103394,
"learning_rate": 1.288853693335314e-06,
"loss": 1.0574,
"step": 1767
},
{
"epoch": 4.036557501904037,
"grad_norm": 1.2419517040252686,
"learning_rate": 1.286130297903406e-06,
"loss": 1.0794,
"step": 1768
},
{
"epoch": 4.0388423457730385,
"grad_norm": 1.3006045818328857,
"learning_rate": 1.2834087860407679e-06,
"loss": 1.1689,
"step": 1769
},
{
"epoch": 4.041127189642041,
"grad_norm": 1.2670701742172241,
"learning_rate": 1.2806891619703858e-06,
"loss": 1.091,
"step": 1770
},
{
"epoch": 4.043412033511044,
"grad_norm": 1.2653870582580566,
"learning_rate": 1.2779714299123194e-06,
"loss": 1.0717,
"step": 1771
},
{
"epoch": 4.0456968773800455,
"grad_norm": 1.2377089262008667,
"learning_rate": 1.2752555940836891e-06,
"loss": 1.1001,
"step": 1772
},
{
"epoch": 4.047981721249048,
"grad_norm": 1.2586807012557983,
"learning_rate": 1.2725416586986766e-06,
"loss": 1.1034,
"step": 1773
},
{
"epoch": 4.05026656511805,
"grad_norm": 1.2842795848846436,
"learning_rate": 1.2698296279685113e-06,
"loss": 1.0828,
"step": 1774
},
{
"epoch": 4.0525514089870525,
"grad_norm": 1.2472583055496216,
"learning_rate": 1.267119506101467e-06,
"loss": 1.0991,
"step": 1775
},
{
"epoch": 4.054836252856055,
"grad_norm": 1.280466914176941,
"learning_rate": 1.2644112973028593e-06,
"loss": 1.0746,
"step": 1776
},
{
"epoch": 4.057121096725057,
"grad_norm": 1.2883317470550537,
"learning_rate": 1.2617050057750322e-06,
"loss": 1.0711,
"step": 1777
},
{
"epoch": 4.0594059405940595,
"grad_norm": 1.2602671384811401,
"learning_rate": 1.2590006357173535e-06,
"loss": 1.0842,
"step": 1778
},
{
"epoch": 4.061690784463062,
"grad_norm": 1.245571494102478,
"learning_rate": 1.256298191326214e-06,
"loss": 1.0934,
"step": 1779
},
{
"epoch": 4.063975628332064,
"grad_norm": 1.2416950464248657,
"learning_rate": 1.2535976767950111e-06,
"loss": 1.0815,
"step": 1780
},
{
"epoch": 4.0662604722010665,
"grad_norm": 1.2669044733047485,
"learning_rate": 1.2508990963141527e-06,
"loss": 1.0961,
"step": 1781
},
{
"epoch": 4.068545316070068,
"grad_norm": 1.2506545782089233,
"learning_rate": 1.248202454071042e-06,
"loss": 1.0364,
"step": 1782
},
{
"epoch": 4.070830159939071,
"grad_norm": 1.232606291770935,
"learning_rate": 1.2455077542500748e-06,
"loss": 1.0741,
"step": 1783
},
{
"epoch": 4.0731150038080735,
"grad_norm": 1.2768986225128174,
"learning_rate": 1.2428150010326356e-06,
"loss": 1.1313,
"step": 1784
},
{
"epoch": 4.075399847677075,
"grad_norm": 1.2605336904525757,
"learning_rate": 1.2401241985970857e-06,
"loss": 1.1041,
"step": 1785
},
{
"epoch": 4.077684691546078,
"grad_norm": 1.264955759048462,
"learning_rate": 1.2374353511187593e-06,
"loss": 1.0845,
"step": 1786
},
{
"epoch": 4.07996953541508,
"grad_norm": 1.2806904315948486,
"learning_rate": 1.23474846276996e-06,
"loss": 1.0702,
"step": 1787
},
{
"epoch": 4.082254379284082,
"grad_norm": 1.280799150466919,
"learning_rate": 1.2320635377199469e-06,
"loss": 1.1091,
"step": 1788
},
{
"epoch": 4.084539223153085,
"grad_norm": 1.2656049728393555,
"learning_rate": 1.2293805801349375e-06,
"loss": 1.104,
"step": 1789
},
{
"epoch": 4.086824067022087,
"grad_norm": 1.304160714149475,
"learning_rate": 1.2266995941780934e-06,
"loss": 1.0992,
"step": 1790
},
{
"epoch": 4.089108910891089,
"grad_norm": 1.2726844549179077,
"learning_rate": 1.224020584009516e-06,
"loss": 1.0866,
"step": 1791
},
{
"epoch": 4.091393754760092,
"grad_norm": 1.2950150966644287,
"learning_rate": 1.221343553786244e-06,
"loss": 1.0824,
"step": 1792
},
{
"epoch": 4.093678598629094,
"grad_norm": 1.2644156217575073,
"learning_rate": 1.2186685076622416e-06,
"loss": 1.0926,
"step": 1793
},
{
"epoch": 4.095963442498096,
"grad_norm": 1.2528104782104492,
"learning_rate": 1.2159954497883933e-06,
"loss": 1.0754,
"step": 1794
},
{
"epoch": 4.098248286367098,
"grad_norm": 1.2485560178756714,
"learning_rate": 1.2133243843125012e-06,
"loss": 1.0922,
"step": 1795
},
{
"epoch": 4.1005331302361006,
"grad_norm": 1.2408583164215088,
"learning_rate": 1.210655315379275e-06,
"loss": 1.0541,
"step": 1796
},
{
"epoch": 4.102817974105103,
"grad_norm": 1.2372719049453735,
"learning_rate": 1.2079882471303244e-06,
"loss": 1.1071,
"step": 1797
},
{
"epoch": 4.105102817974105,
"grad_norm": 1.2654283046722412,
"learning_rate": 1.2053231837041576e-06,
"loss": 1.1031,
"step": 1798
},
{
"epoch": 4.1073876618431076,
"grad_norm": 1.2387542724609375,
"learning_rate": 1.2026601292361696e-06,
"loss": 1.0945,
"step": 1799
},
{
"epoch": 4.109672505712109,
"grad_norm": 1.2538166046142578,
"learning_rate": 1.1999990878586374e-06,
"loss": 1.1081,
"step": 1800
},
{
"epoch": 4.111957349581112,
"grad_norm": 1.2714624404907227,
"learning_rate": 1.1973400637007179e-06,
"loss": 1.119,
"step": 1801
},
{
"epoch": 4.1142421934501145,
"grad_norm": 1.2416982650756836,
"learning_rate": 1.1946830608884335e-06,
"loss": 1.1051,
"step": 1802
},
{
"epoch": 4.116527037319116,
"grad_norm": 1.2721920013427734,
"learning_rate": 1.192028083544675e-06,
"loss": 1.1145,
"step": 1803
},
{
"epoch": 4.118811881188119,
"grad_norm": 1.3008158206939697,
"learning_rate": 1.1893751357891854e-06,
"loss": 1.0883,
"step": 1804
},
{
"epoch": 4.1210967250571215,
"grad_norm": 1.2460572719573975,
"learning_rate": 1.1867242217385599e-06,
"loss": 1.1278,
"step": 1805
},
{
"epoch": 4.123381568926123,
"grad_norm": 1.2478859424591064,
"learning_rate": 1.1840753455062407e-06,
"loss": 1.1538,
"step": 1806
},
{
"epoch": 4.125666412795126,
"grad_norm": 1.2536922693252563,
"learning_rate": 1.1814285112025044e-06,
"loss": 1.0893,
"step": 1807
},
{
"epoch": 4.127951256664128,
"grad_norm": 1.2688579559326172,
"learning_rate": 1.1787837229344602e-06,
"loss": 1.0969,
"step": 1808
},
{
"epoch": 4.13023610053313,
"grad_norm": 1.2402559518814087,
"learning_rate": 1.1761409848060438e-06,
"loss": 1.0974,
"step": 1809
},
{
"epoch": 4.132520944402133,
"grad_norm": 1.2625503540039062,
"learning_rate": 1.173500300918007e-06,
"loss": 1.1134,
"step": 1810
},
{
"epoch": 4.134805788271135,
"grad_norm": 1.3044812679290771,
"learning_rate": 1.1708616753679177e-06,
"loss": 1.0311,
"step": 1811
},
{
"epoch": 4.137090632140137,
"grad_norm": 1.2588051557540894,
"learning_rate": 1.168225112250147e-06,
"loss": 1.0805,
"step": 1812
},
{
"epoch": 4.139375476009139,
"grad_norm": 1.3075844049453735,
"learning_rate": 1.1655906156558653e-06,
"loss": 1.1115,
"step": 1813
},
{
"epoch": 4.141660319878142,
"grad_norm": 1.283777117729187,
"learning_rate": 1.1629581896730394e-06,
"loss": 1.0912,
"step": 1814
},
{
"epoch": 4.143945163747144,
"grad_norm": 1.2859684228897095,
"learning_rate": 1.1603278383864206e-06,
"loss": 1.0894,
"step": 1815
},
{
"epoch": 4.146230007616146,
"grad_norm": 1.2679686546325684,
"learning_rate": 1.1576995658775405e-06,
"loss": 1.0892,
"step": 1816
},
{
"epoch": 4.148514851485149,
"grad_norm": 1.3373825550079346,
"learning_rate": 1.155073376224708e-06,
"loss": 1.0529,
"step": 1817
},
{
"epoch": 4.15079969535415,
"grad_norm": 1.2458816766738892,
"learning_rate": 1.152449273502996e-06,
"loss": 1.098,
"step": 1818
},
{
"epoch": 4.153084539223153,
"grad_norm": 1.270310878753662,
"learning_rate": 1.1498272617842438e-06,
"loss": 1.1336,
"step": 1819
},
{
"epoch": 4.155369383092156,
"grad_norm": 1.2812644243240356,
"learning_rate": 1.147207345137042e-06,
"loss": 1.0748,
"step": 1820
},
{
"epoch": 4.157654226961157,
"grad_norm": 1.2596019506454468,
"learning_rate": 1.1445895276267307e-06,
"loss": 1.08,
"step": 1821
},
{
"epoch": 4.15993907083016,
"grad_norm": 1.231399416923523,
"learning_rate": 1.1419738133153963e-06,
"loss": 1.1007,
"step": 1822
},
{
"epoch": 4.162223914699163,
"grad_norm": 1.272637963294983,
"learning_rate": 1.139360206261858e-06,
"loss": 1.0918,
"step": 1823
},
{
"epoch": 4.164508758568164,
"grad_norm": 1.2707762718200684,
"learning_rate": 1.1367487105216655e-06,
"loss": 1.1117,
"step": 1824
},
{
"epoch": 4.166793602437167,
"grad_norm": 1.2958250045776367,
"learning_rate": 1.1341393301470945e-06,
"loss": 1.0975,
"step": 1825
},
{
"epoch": 4.169078446306169,
"grad_norm": 1.2849308252334595,
"learning_rate": 1.131532069187136e-06,
"loss": 1.1389,
"step": 1826
},
{
"epoch": 4.171363290175171,
"grad_norm": 1.2934991121292114,
"learning_rate": 1.1289269316874944e-06,
"loss": 1.1119,
"step": 1827
},
{
"epoch": 4.173648134044174,
"grad_norm": 1.2864172458648682,
"learning_rate": 1.1263239216905775e-06,
"loss": 1.0708,
"step": 1828
},
{
"epoch": 4.175932977913176,
"grad_norm": 1.262462854385376,
"learning_rate": 1.1237230432354912e-06,
"loss": 1.0952,
"step": 1829
},
{
"epoch": 4.178217821782178,
"grad_norm": 1.2699599266052246,
"learning_rate": 1.1211243003580368e-06,
"loss": 1.1096,
"step": 1830
},
{
"epoch": 4.18050266565118,
"grad_norm": 1.2328681945800781,
"learning_rate": 1.1185276970906994e-06,
"loss": 1.086,
"step": 1831
},
{
"epoch": 4.182787509520183,
"grad_norm": 1.2541395425796509,
"learning_rate": 1.1159332374626434e-06,
"loss": 1.1087,
"step": 1832
},
{
"epoch": 4.185072353389185,
"grad_norm": 1.269580364227295,
"learning_rate": 1.11334092549971e-06,
"loss": 1.0761,
"step": 1833
},
{
"epoch": 4.187357197258187,
"grad_norm": 1.2778130769729614,
"learning_rate": 1.1107507652244043e-06,
"loss": 1.0812,
"step": 1834
},
{
"epoch": 4.18964204112719,
"grad_norm": 1.2541184425354004,
"learning_rate": 1.1081627606558964e-06,
"loss": 1.0944,
"step": 1835
},
{
"epoch": 4.191926884996192,
"grad_norm": 1.2608321905136108,
"learning_rate": 1.105576915810008e-06,
"loss": 1.1131,
"step": 1836
},
{
"epoch": 4.194211728865194,
"grad_norm": 1.2532767057418823,
"learning_rate": 1.1029932346992101e-06,
"loss": 1.0743,
"step": 1837
},
{
"epoch": 4.196496572734197,
"grad_norm": 1.2791781425476074,
"learning_rate": 1.1004117213326188e-06,
"loss": 1.0794,
"step": 1838
},
{
"epoch": 4.1987814166031985,
"grad_norm": 1.2686764001846313,
"learning_rate": 1.0978323797159837e-06,
"loss": 1.0632,
"step": 1839
},
{
"epoch": 4.201066260472201,
"grad_norm": 1.2395896911621094,
"learning_rate": 1.0952552138516844e-06,
"loss": 1.1195,
"step": 1840
},
{
"epoch": 4.203351104341204,
"grad_norm": 1.2782765626907349,
"learning_rate": 1.0926802277387278e-06,
"loss": 1.0806,
"step": 1841
},
{
"epoch": 4.2056359482102055,
"grad_norm": 1.289858341217041,
"learning_rate": 1.0901074253727338e-06,
"loss": 1.0537,
"step": 1842
},
{
"epoch": 4.207920792079208,
"grad_norm": 1.2895413637161255,
"learning_rate": 1.0875368107459377e-06,
"loss": 1.0683,
"step": 1843
},
{
"epoch": 4.21020563594821,
"grad_norm": 1.2379080057144165,
"learning_rate": 1.0849683878471778e-06,
"loss": 1.1203,
"step": 1844
},
{
"epoch": 4.2124904798172125,
"grad_norm": 1.2541252374649048,
"learning_rate": 1.0824021606618915e-06,
"loss": 1.1216,
"step": 1845
},
{
"epoch": 4.214775323686215,
"grad_norm": 1.2726837396621704,
"learning_rate": 1.079838133172111e-06,
"loss": 1.1148,
"step": 1846
},
{
"epoch": 4.217060167555217,
"grad_norm": 1.256637692451477,
"learning_rate": 1.077276309356453e-06,
"loss": 1.0731,
"step": 1847
},
{
"epoch": 4.2193450114242195,
"grad_norm": 1.2752294540405273,
"learning_rate": 1.0747166931901152e-06,
"loss": 1.1408,
"step": 1848
},
{
"epoch": 4.221629855293221,
"grad_norm": 1.3284248113632202,
"learning_rate": 1.0721592886448718e-06,
"loss": 1.0814,
"step": 1849
},
{
"epoch": 4.223914699162224,
"grad_norm": 1.2901482582092285,
"learning_rate": 1.0696040996890613e-06,
"loss": 1.0844,
"step": 1850
},
{
"epoch": 4.2261995430312265,
"grad_norm": 1.2776188850402832,
"learning_rate": 1.0670511302875892e-06,
"loss": 1.1033,
"step": 1851
},
{
"epoch": 4.228484386900228,
"grad_norm": 1.312633991241455,
"learning_rate": 1.0645003844019123e-06,
"loss": 1.0575,
"step": 1852
},
{
"epoch": 4.230769230769231,
"grad_norm": 1.2696623802185059,
"learning_rate": 1.0619518659900387e-06,
"loss": 1.0974,
"step": 1853
},
{
"epoch": 4.2330540746382335,
"grad_norm": 1.32041597366333,
"learning_rate": 1.0594055790065228e-06,
"loss": 1.0983,
"step": 1854
},
{
"epoch": 4.235338918507235,
"grad_norm": 1.2588789463043213,
"learning_rate": 1.0568615274024521e-06,
"loss": 1.1283,
"step": 1855
},
{
"epoch": 4.237623762376238,
"grad_norm": 1.2674574851989746,
"learning_rate": 1.0543197151254475e-06,
"loss": 1.1213,
"step": 1856
},
{
"epoch": 4.23990860624524,
"grad_norm": 1.2549033164978027,
"learning_rate": 1.0517801461196568e-06,
"loss": 1.0895,
"step": 1857
},
{
"epoch": 4.242193450114242,
"grad_norm": 1.3073495626449585,
"learning_rate": 1.049242824325743e-06,
"loss": 1.0966,
"step": 1858
},
{
"epoch": 4.244478293983245,
"grad_norm": 1.2715238332748413,
"learning_rate": 1.0467077536808867e-06,
"loss": 1.0803,
"step": 1859
},
{
"epoch": 4.246763137852247,
"grad_norm": 1.2352854013442993,
"learning_rate": 1.0441749381187718e-06,
"loss": 1.0713,
"step": 1860
},
{
"epoch": 4.249047981721249,
"grad_norm": 1.235982894897461,
"learning_rate": 1.0416443815695831e-06,
"loss": 1.0961,
"step": 1861
},
{
"epoch": 4.251332825590251,
"grad_norm": 1.288275122642517,
"learning_rate": 1.0391160879600034e-06,
"loss": 1.1036,
"step": 1862
},
{
"epoch": 4.253617669459254,
"grad_norm": 1.2842768430709839,
"learning_rate": 1.0365900612132007e-06,
"loss": 1.0729,
"step": 1863
},
{
"epoch": 4.255902513328256,
"grad_norm": 1.2533109188079834,
"learning_rate": 1.0340663052488257e-06,
"loss": 1.1035,
"step": 1864
},
{
"epoch": 4.258187357197258,
"grad_norm": 1.3068199157714844,
"learning_rate": 1.0315448239830083e-06,
"loss": 1.0985,
"step": 1865
},
{
"epoch": 4.260472201066261,
"grad_norm": 1.2762842178344727,
"learning_rate": 1.029025621328345e-06,
"loss": 1.0613,
"step": 1866
},
{
"epoch": 4.262757044935263,
"grad_norm": 1.3192155361175537,
"learning_rate": 1.0265087011939004e-06,
"loss": 1.0855,
"step": 1867
},
{
"epoch": 4.265041888804265,
"grad_norm": 1.2514667510986328,
"learning_rate": 1.0239940674851943e-06,
"loss": 1.0835,
"step": 1868
},
{
"epoch": 4.267326732673268,
"grad_norm": 1.2197521924972534,
"learning_rate": 1.0214817241041985e-06,
"loss": 1.0858,
"step": 1869
},
{
"epoch": 4.269611576542269,
"grad_norm": 1.271289348602295,
"learning_rate": 1.0189716749493345e-06,
"loss": 1.1279,
"step": 1870
},
{
"epoch": 4.271896420411272,
"grad_norm": 1.274689793586731,
"learning_rate": 1.0164639239154603e-06,
"loss": 1.0582,
"step": 1871
},
{
"epoch": 4.274181264280275,
"grad_norm": 1.2924543619155884,
"learning_rate": 1.0139584748938678e-06,
"loss": 1.0721,
"step": 1872
},
{
"epoch": 4.276466108149276,
"grad_norm": 1.2583580017089844,
"learning_rate": 1.01145533177228e-06,
"loss": 1.0925,
"step": 1873
},
{
"epoch": 4.278750952018279,
"grad_norm": 1.3069345951080322,
"learning_rate": 1.0089544984348378e-06,
"loss": 1.1106,
"step": 1874
},
{
"epoch": 4.281035795887281,
"grad_norm": 1.2459675073623657,
"learning_rate": 1.0064559787621018e-06,
"loss": 1.0756,
"step": 1875
},
{
"epoch": 4.283320639756283,
"grad_norm": 1.2553149461746216,
"learning_rate": 1.0039597766310408e-06,
"loss": 1.1231,
"step": 1876
},
{
"epoch": 4.285605483625286,
"grad_norm": 1.2836211919784546,
"learning_rate": 1.0014658959150267e-06,
"loss": 1.0919,
"step": 1877
},
{
"epoch": 4.287890327494288,
"grad_norm": 1.303640365600586,
"learning_rate": 9.98974340483829e-07,
"loss": 1.1021,
"step": 1878
},
{
"epoch": 4.29017517136329,
"grad_norm": 1.2877558469772339,
"learning_rate": 9.96485114203612e-07,
"loss": 1.103,
"step": 1879
},
{
"epoch": 4.292460015232292,
"grad_norm": 1.2467522621154785,
"learning_rate": 9.93998220936922e-07,
"loss": 1.0964,
"step": 1880
},
{
"epoch": 4.294744859101295,
"grad_norm": 1.29486083984375,
"learning_rate": 9.915136645426885e-07,
"loss": 1.0804,
"step": 1881
},
{
"epoch": 4.297029702970297,
"grad_norm": 1.274101734161377,
"learning_rate": 9.890314488762123e-07,
"loss": 1.1205,
"step": 1882
},
{
"epoch": 4.299314546839299,
"grad_norm": 1.2509732246398926,
"learning_rate": 9.865515777891626e-07,
"loss": 1.0654,
"step": 1883
},
{
"epoch": 4.301599390708302,
"grad_norm": 1.2939627170562744,
"learning_rate": 9.840740551295728e-07,
"loss": 1.1175,
"step": 1884
},
{
"epoch": 4.303884234577304,
"grad_norm": 1.2567757368087769,
"learning_rate": 9.815988847418292e-07,
"loss": 1.1102,
"step": 1885
},
{
"epoch": 4.306169078446306,
"grad_norm": 1.2803035974502563,
"learning_rate": 9.791260704666688e-07,
"loss": 1.0835,
"step": 1886
},
{
"epoch": 4.308453922315309,
"grad_norm": 1.264268398284912,
"learning_rate": 9.766556161411744e-07,
"loss": 1.0827,
"step": 1887
},
{
"epoch": 4.31073876618431,
"grad_norm": 1.2779146432876587,
"learning_rate": 9.741875255987643e-07,
"loss": 1.0971,
"step": 1888
},
{
"epoch": 4.313023610053313,
"grad_norm": 1.2511680126190186,
"learning_rate": 9.717218026691917e-07,
"loss": 1.0842,
"step": 1889
},
{
"epoch": 4.315308453922316,
"grad_norm": 1.2907341718673706,
"learning_rate": 9.692584511785333e-07,
"loss": 1.083,
"step": 1890
},
{
"epoch": 4.317593297791317,
"grad_norm": 1.30160653591156,
"learning_rate": 9.667974749491865e-07,
"loss": 1.1403,
"step": 1891
},
{
"epoch": 4.31987814166032,
"grad_norm": 1.2925405502319336,
"learning_rate": 9.643388777998652e-07,
"loss": 1.0918,
"step": 1892
},
{
"epoch": 4.322162985529322,
"grad_norm": 1.2780344486236572,
"learning_rate": 9.618826635455898e-07,
"loss": 1.0845,
"step": 1893
},
{
"epoch": 4.324447829398324,
"grad_norm": 1.2781273126602173,
"learning_rate": 9.594288359976817e-07,
"loss": 1.1027,
"step": 1894
},
{
"epoch": 4.326732673267327,
"grad_norm": 1.2914685010910034,
"learning_rate": 9.569773989637628e-07,
"loss": 1.0967,
"step": 1895
},
{
"epoch": 4.329017517136329,
"grad_norm": 1.2581839561462402,
"learning_rate": 9.54528356247742e-07,
"loss": 1.1195,
"step": 1896
},
{
"epoch": 4.331302361005331,
"grad_norm": 1.286125659942627,
"learning_rate": 9.520817116498154e-07,
"loss": 1.097,
"step": 1897
},
{
"epoch": 4.333587204874334,
"grad_norm": 1.2758541107177734,
"learning_rate": 9.496374689664564e-07,
"loss": 1.1183,
"step": 1898
},
{
"epoch": 4.335872048743336,
"grad_norm": 1.2484683990478516,
"learning_rate": 9.471956319904105e-07,
"loss": 1.0762,
"step": 1899
},
{
"epoch": 4.338156892612338,
"grad_norm": 1.2502278089523315,
"learning_rate": 9.447562045106939e-07,
"loss": 1.1314,
"step": 1900
},
{
"epoch": 4.34044173648134,
"grad_norm": 1.322770118713379,
"learning_rate": 9.423191903125803e-07,
"loss": 1.116,
"step": 1901
},
{
"epoch": 4.342726580350343,
"grad_norm": 1.2737778425216675,
"learning_rate": 9.398845931775991e-07,
"loss": 1.0773,
"step": 1902
},
{
"epoch": 4.345011424219345,
"grad_norm": 1.2634154558181763,
"learning_rate": 9.374524168835319e-07,
"loss": 1.1247,
"step": 1903
},
{
"epoch": 4.347296268088347,
"grad_norm": 1.2655125856399536,
"learning_rate": 9.350226652044001e-07,
"loss": 1.0569,
"step": 1904
},
{
"epoch": 4.34958111195735,
"grad_norm": 1.3117061853408813,
"learning_rate": 9.325953419104666e-07,
"loss": 1.1352,
"step": 1905
},
{
"epoch": 4.3518659558263515,
"grad_norm": 1.3022981882095337,
"learning_rate": 9.301704507682236e-07,
"loss": 1.0685,
"step": 1906
},
{
"epoch": 4.354150799695354,
"grad_norm": 1.2460218667984009,
"learning_rate": 9.277479955403887e-07,
"loss": 1.0802,
"step": 1907
},
{
"epoch": 4.356435643564357,
"grad_norm": 1.2989197969436646,
"learning_rate": 9.253279799859027e-07,
"loss": 1.0962,
"step": 1908
},
{
"epoch": 4.3587204874333585,
"grad_norm": 1.275721788406372,
"learning_rate": 9.229104078599186e-07,
"loss": 1.0964,
"step": 1909
},
{
"epoch": 4.361005331302361,
"grad_norm": 1.2741754055023193,
"learning_rate": 9.204952829137967e-07,
"loss": 1.0707,
"step": 1910
},
{
"epoch": 4.363290175171363,
"grad_norm": 1.2817518711090088,
"learning_rate": 9.180826088951039e-07,
"loss": 1.1041,
"step": 1911
},
{
"epoch": 4.3655750190403655,
"grad_norm": 1.2775477170944214,
"learning_rate": 9.156723895476007e-07,
"loss": 1.1187,
"step": 1912
},
{
"epoch": 4.367859862909368,
"grad_norm": 1.2695951461791992,
"learning_rate": 9.13264628611239e-07,
"loss": 1.0301,
"step": 1913
},
{
"epoch": 4.37014470677837,
"grad_norm": 1.2773480415344238,
"learning_rate": 9.10859329822158e-07,
"loss": 1.1125,
"step": 1914
},
{
"epoch": 4.3724295506473725,
"grad_norm": 1.297052025794983,
"learning_rate": 9.084564969126733e-07,
"loss": 1.0619,
"step": 1915
},
{
"epoch": 4.374714394516375,
"grad_norm": 1.2900358438491821,
"learning_rate": 9.060561336112775e-07,
"loss": 1.0316,
"step": 1916
},
{
"epoch": 4.376999238385377,
"grad_norm": 1.2947112321853638,
"learning_rate": 9.036582436426289e-07,
"loss": 1.0878,
"step": 1917
},
{
"epoch": 4.3792840822543795,
"grad_norm": 1.2751237154006958,
"learning_rate": 9.012628307275478e-07,
"loss": 1.1047,
"step": 1918
},
{
"epoch": 4.381568926123381,
"grad_norm": 1.2209362983703613,
"learning_rate": 8.988698985830127e-07,
"loss": 1.0891,
"step": 1919
},
{
"epoch": 4.383853769992384,
"grad_norm": 1.2782630920410156,
"learning_rate": 8.964794509221508e-07,
"loss": 1.1201,
"step": 1920
},
{
"epoch": 4.3861386138613865,
"grad_norm": 1.3166882991790771,
"learning_rate": 8.940914914542342e-07,
"loss": 1.1204,
"step": 1921
},
{
"epoch": 4.388423457730388,
"grad_norm": 1.2406049966812134,
"learning_rate": 8.917060238846764e-07,
"loss": 1.0871,
"step": 1922
},
{
"epoch": 4.390708301599391,
"grad_norm": 1.3386300802230835,
"learning_rate": 8.893230519150203e-07,
"loss": 1.1023,
"step": 1923
},
{
"epoch": 4.392993145468393,
"grad_norm": 1.345792531967163,
"learning_rate": 8.869425792429406e-07,
"loss": 1.081,
"step": 1924
},
{
"epoch": 4.395277989337395,
"grad_norm": 1.264641284942627,
"learning_rate": 8.84564609562231e-07,
"loss": 1.1115,
"step": 1925
},
{
"epoch": 4.397562833206398,
"grad_norm": 1.2565428018569946,
"learning_rate": 8.821891465628008e-07,
"loss": 1.0721,
"step": 1926
},
{
"epoch": 4.3998476770754,
"grad_norm": 1.252386450767517,
"learning_rate": 8.798161939306726e-07,
"loss": 1.1019,
"step": 1927
},
{
"epoch": 4.402132520944402,
"grad_norm": 1.2885632514953613,
"learning_rate": 8.774457553479717e-07,
"loss": 1.1058,
"step": 1928
},
{
"epoch": 4.404417364813405,
"grad_norm": 1.2969319820404053,
"learning_rate": 8.750778344929211e-07,
"loss": 1.1088,
"step": 1929
},
{
"epoch": 4.406702208682407,
"grad_norm": 1.285504698753357,
"learning_rate": 8.727124350398409e-07,
"loss": 1.0635,
"step": 1930
},
{
"epoch": 4.408987052551409,
"grad_norm": 1.2720234394073486,
"learning_rate": 8.703495606591339e-07,
"loss": 1.1078,
"step": 1931
},
{
"epoch": 4.411271896420411,
"grad_norm": 1.2709497213363647,
"learning_rate": 8.679892150172895e-07,
"loss": 1.1214,
"step": 1932
},
{
"epoch": 4.413556740289414,
"grad_norm": 1.2678945064544678,
"learning_rate": 8.656314017768694e-07,
"loss": 1.106,
"step": 1933
},
{
"epoch": 4.415841584158416,
"grad_norm": 1.2668631076812744,
"learning_rate": 8.632761245965071e-07,
"loss": 1.1026,
"step": 1934
},
{
"epoch": 4.418126428027418,
"grad_norm": 1.313750147819519,
"learning_rate": 8.609233871309025e-07,
"loss": 1.0727,
"step": 1935
},
{
"epoch": 4.420411271896421,
"grad_norm": 1.2804429531097412,
"learning_rate": 8.58573193030812e-07,
"loss": 1.0807,
"step": 1936
},
{
"epoch": 4.422696115765422,
"grad_norm": 1.3013975620269775,
"learning_rate": 8.562255459430463e-07,
"loss": 1.0885,
"step": 1937
},
{
"epoch": 4.424980959634425,
"grad_norm": 1.257980465888977,
"learning_rate": 8.538804495104653e-07,
"loss": 1.1049,
"step": 1938
},
{
"epoch": 4.427265803503428,
"grad_norm": 1.2810685634613037,
"learning_rate": 8.515379073719687e-07,
"loss": 1.0769,
"step": 1939
},
{
"epoch": 4.429550647372429,
"grad_norm": 1.2664494514465332,
"learning_rate": 8.491979231624956e-07,
"loss": 1.0969,
"step": 1940
},
{
"epoch": 4.431835491241432,
"grad_norm": 1.228659987449646,
"learning_rate": 8.468605005130129e-07,
"loss": 1.1036,
"step": 1941
},
{
"epoch": 4.434120335110434,
"grad_norm": 1.2892229557037354,
"learning_rate": 8.445256430505136e-07,
"loss": 1.0789,
"step": 1942
},
{
"epoch": 4.436405178979436,
"grad_norm": 1.253233790397644,
"learning_rate": 8.421933543980126e-07,
"loss": 1.0987,
"step": 1943
},
{
"epoch": 4.438690022848439,
"grad_norm": 1.2383204698562622,
"learning_rate": 8.398636381745356e-07,
"loss": 1.0822,
"step": 1944
},
{
"epoch": 4.440974866717441,
"grad_norm": 1.2798322439193726,
"learning_rate": 8.375364979951176e-07,
"loss": 1.0943,
"step": 1945
},
{
"epoch": 4.443259710586443,
"grad_norm": 1.2920564413070679,
"learning_rate": 8.352119374707979e-07,
"loss": 1.1353,
"step": 1946
},
{
"epoch": 4.445544554455446,
"grad_norm": 1.282945156097412,
"learning_rate": 8.328899602086105e-07,
"loss": 1.0711,
"step": 1947
},
{
"epoch": 4.447829398324448,
"grad_norm": 1.2641621828079224,
"learning_rate": 8.305705698115835e-07,
"loss": 1.082,
"step": 1948
},
{
"epoch": 4.45011424219345,
"grad_norm": 1.2811444997787476,
"learning_rate": 8.282537698787288e-07,
"loss": 1.0967,
"step": 1949
},
{
"epoch": 4.452399086062452,
"grad_norm": 1.280187964439392,
"learning_rate": 8.259395640050385e-07,
"loss": 1.0791,
"step": 1950
},
{
"epoch": 4.454683929931455,
"grad_norm": 1.2966340780258179,
"learning_rate": 8.236279557814828e-07,
"loss": 1.099,
"step": 1951
},
{
"epoch": 4.456968773800457,
"grad_norm": 1.2679872512817383,
"learning_rate": 8.213189487949972e-07,
"loss": 1.0891,
"step": 1952
},
{
"epoch": 4.459253617669459,
"grad_norm": 1.289146900177002,
"learning_rate": 8.19012546628481e-07,
"loss": 1.0827,
"step": 1953
},
{
"epoch": 4.461538461538462,
"grad_norm": 1.2679554224014282,
"learning_rate": 8.167087528607961e-07,
"loss": 1.0916,
"step": 1954
},
{
"epoch": 4.463823305407463,
"grad_norm": 1.2724610567092896,
"learning_rate": 8.144075710667524e-07,
"loss": 1.0855,
"step": 1955
},
{
"epoch": 4.466108149276466,
"grad_norm": 1.2689306735992432,
"learning_rate": 8.121090048171074e-07,
"loss": 1.116,
"step": 1956
},
{
"epoch": 4.468392993145469,
"grad_norm": 1.3125536441802979,
"learning_rate": 8.098130576785623e-07,
"loss": 1.1,
"step": 1957
},
{
"epoch": 4.47067783701447,
"grad_norm": 1.2809594869613647,
"learning_rate": 8.075197332137522e-07,
"loss": 1.0517,
"step": 1958
},
{
"epoch": 4.472962680883473,
"grad_norm": 1.2979421615600586,
"learning_rate": 8.052290349812419e-07,
"loss": 1.0963,
"step": 1959
},
{
"epoch": 4.475247524752476,
"grad_norm": 1.252105951309204,
"learning_rate": 8.029409665355242e-07,
"loss": 1.0863,
"step": 1960
},
{
"epoch": 4.477532368621477,
"grad_norm": 1.2643629312515259,
"learning_rate": 8.006555314270079e-07,
"loss": 1.0525,
"step": 1961
},
{
"epoch": 4.47981721249048,
"grad_norm": 1.2668061256408691,
"learning_rate": 7.98372733202018e-07,
"loss": 1.1325,
"step": 1962
},
{
"epoch": 4.482102056359482,
"grad_norm": 1.292205572128296,
"learning_rate": 7.960925754027871e-07,
"loss": 1.0877,
"step": 1963
},
{
"epoch": 4.484386900228484,
"grad_norm": 1.2493971586227417,
"learning_rate": 7.938150615674487e-07,
"loss": 1.1041,
"step": 1964
},
{
"epoch": 4.486671744097487,
"grad_norm": 1.2565317153930664,
"learning_rate": 7.915401952300372e-07,
"loss": 1.0833,
"step": 1965
},
{
"epoch": 4.488956587966489,
"grad_norm": 1.3105082511901855,
"learning_rate": 7.892679799204766e-07,
"loss": 1.1045,
"step": 1966
},
{
"epoch": 4.491241431835491,
"grad_norm": 1.2809815406799316,
"learning_rate": 7.869984191645768e-07,
"loss": 1.1091,
"step": 1967
},
{
"epoch": 4.493526275704493,
"grad_norm": 1.2692221403121948,
"learning_rate": 7.847315164840311e-07,
"loss": 1.1174,
"step": 1968
},
{
"epoch": 4.495811119573496,
"grad_norm": 1.2772730588912964,
"learning_rate": 7.82467275396405e-07,
"loss": 1.0935,
"step": 1969
},
{
"epoch": 4.498095963442498,
"grad_norm": 1.2484936714172363,
"learning_rate": 7.802056994151369e-07,
"loss": 1.1058,
"step": 1970
},
{
"epoch": 4.5003808073115,
"grad_norm": 1.2626773118972778,
"learning_rate": 7.779467920495285e-07,
"loss": 1.0788,
"step": 1971
},
{
"epoch": 4.502665651180503,
"grad_norm": 1.2814165353775024,
"learning_rate": 7.756905568047393e-07,
"loss": 1.1073,
"step": 1972
},
{
"epoch": 4.5049504950495045,
"grad_norm": 1.2587292194366455,
"learning_rate": 7.734369971817854e-07,
"loss": 1.095,
"step": 1973
},
{
"epoch": 4.507235338918507,
"grad_norm": 1.2431440353393555,
"learning_rate": 7.711861166775289e-07,
"loss": 1.0936,
"step": 1974
},
{
"epoch": 4.50952018278751,
"grad_norm": 1.277116060256958,
"learning_rate": 7.689379187846743e-07,
"loss": 1.0982,
"step": 1975
},
{
"epoch": 4.5118050266565115,
"grad_norm": 1.2661057710647583,
"learning_rate": 7.666924069917659e-07,
"loss": 1.1299,
"step": 1976
},
{
"epoch": 4.514089870525514,
"grad_norm": 1.2476885318756104,
"learning_rate": 7.644495847831771e-07,
"loss": 1.0804,
"step": 1977
},
{
"epoch": 4.516374714394516,
"grad_norm": 1.2756448984146118,
"learning_rate": 7.622094556391108e-07,
"loss": 1.1115,
"step": 1978
},
{
"epoch": 4.5186595582635185,
"grad_norm": 1.2629826068878174,
"learning_rate": 7.599720230355889e-07,
"loss": 1.1049,
"step": 1979
},
{
"epoch": 4.520944402132521,
"grad_norm": 1.2404588460922241,
"learning_rate": 7.577372904444486e-07,
"loss": 1.0842,
"step": 1980
},
{
"epoch": 4.523229246001523,
"grad_norm": 1.2706893682479858,
"learning_rate": 7.555052613333402e-07,
"loss": 1.0619,
"step": 1981
},
{
"epoch": 4.5255140898705255,
"grad_norm": 1.2344950437545776,
"learning_rate": 7.532759391657168e-07,
"loss": 1.1276,
"step": 1982
},
{
"epoch": 4.527798933739528,
"grad_norm": 1.2568199634552002,
"learning_rate": 7.510493274008304e-07,
"loss": 1.1366,
"step": 1983
},
{
"epoch": 4.53008377760853,
"grad_norm": 1.330805778503418,
"learning_rate": 7.488254294937306e-07,
"loss": 1.0899,
"step": 1984
},
{
"epoch": 4.5323686214775325,
"grad_norm": 1.26802659034729,
"learning_rate": 7.466042488952521e-07,
"loss": 1.1331,
"step": 1985
},
{
"epoch": 4.534653465346535,
"grad_norm": 1.2709177732467651,
"learning_rate": 7.443857890520162e-07,
"loss": 1.1079,
"step": 1986
},
{
"epoch": 4.536938309215537,
"grad_norm": 1.2737971544265747,
"learning_rate": 7.421700534064202e-07,
"loss": 1.0821,
"step": 1987
},
{
"epoch": 4.5392231530845395,
"grad_norm": 1.2980129718780518,
"learning_rate": 7.399570453966346e-07,
"loss": 1.1353,
"step": 1988
},
{
"epoch": 4.541507996953541,
"grad_norm": 1.3661080598831177,
"learning_rate": 7.377467684565992e-07,
"loss": 1.1105,
"step": 1989
},
{
"epoch": 4.543792840822544,
"grad_norm": 1.2837845087051392,
"learning_rate": 7.355392260160141e-07,
"loss": 1.0746,
"step": 1990
},
{
"epoch": 4.5460776846915465,
"grad_norm": 1.3188300132751465,
"learning_rate": 7.333344215003358e-07,
"loss": 1.1099,
"step": 1991
},
{
"epoch": 4.548362528560548,
"grad_norm": 1.2677329778671265,
"learning_rate": 7.31132358330775e-07,
"loss": 1.0969,
"step": 1992
},
{
"epoch": 4.550647372429551,
"grad_norm": 1.2950503826141357,
"learning_rate": 7.289330399242859e-07,
"loss": 1.0845,
"step": 1993
},
{
"epoch": 4.552932216298553,
"grad_norm": 1.3180036544799805,
"learning_rate": 7.267364696935653e-07,
"loss": 1.0998,
"step": 1994
},
{
"epoch": 4.555217060167555,
"grad_norm": 1.3282592296600342,
"learning_rate": 7.245426510470452e-07,
"loss": 1.0614,
"step": 1995
},
{
"epoch": 4.557501904036558,
"grad_norm": 1.3138798475265503,
"learning_rate": 7.223515873888867e-07,
"loss": 1.1126,
"step": 1996
},
{
"epoch": 4.55978674790556,
"grad_norm": 1.268276572227478,
"learning_rate": 7.201632821189786e-07,
"loss": 1.1083,
"step": 1997
},
{
"epoch": 4.562071591774562,
"grad_norm": 1.2721282243728638,
"learning_rate": 7.179777386329276e-07,
"loss": 1.1111,
"step": 1998
},
{
"epoch": 4.564356435643564,
"grad_norm": 1.292134404182434,
"learning_rate": 7.157949603220537e-07,
"loss": 1.1061,
"step": 1999
},
{
"epoch": 4.566641279512567,
"grad_norm": 1.289412260055542,
"learning_rate": 7.136149505733902e-07,
"loss": 1.0958,
"step": 2000
},
{
"epoch": 4.568926123381569,
"grad_norm": 1.267577886581421,
"learning_rate": 7.1143771276967e-07,
"loss": 1.1199,
"step": 2001
},
{
"epoch": 4.571210967250571,
"grad_norm": 1.307735562324524,
"learning_rate": 7.092632502893282e-07,
"loss": 1.0818,
"step": 2002
},
{
"epoch": 4.573495811119574,
"grad_norm": 1.2823678255081177,
"learning_rate": 7.070915665064915e-07,
"loss": 1.0677,
"step": 2003
},
{
"epoch": 4.575780654988575,
"grad_norm": 1.2783132791519165,
"learning_rate": 7.049226647909744e-07,
"loss": 1.1186,
"step": 2004
},
{
"epoch": 4.578065498857578,
"grad_norm": 1.3067246675491333,
"learning_rate": 7.027565485082765e-07,
"loss": 1.1063,
"step": 2005
},
{
"epoch": 4.580350342726581,
"grad_norm": 1.2940545082092285,
"learning_rate": 7.005932210195738e-07,
"loss": 1.1105,
"step": 2006
},
{
"epoch": 4.582635186595582,
"grad_norm": 1.2702580690383911,
"learning_rate": 6.984326856817145e-07,
"loss": 1.1062,
"step": 2007
},
{
"epoch": 4.584920030464585,
"grad_norm": 1.3609583377838135,
"learning_rate": 6.96274945847216e-07,
"loss": 1.0907,
"step": 2008
},
{
"epoch": 4.587204874333588,
"grad_norm": 1.2673293352127075,
"learning_rate": 6.941200048642558e-07,
"loss": 1.0699,
"step": 2009
},
{
"epoch": 4.589489718202589,
"grad_norm": 1.26469886302948,
"learning_rate": 6.919678660766705e-07,
"loss": 1.079,
"step": 2010
},
{
"epoch": 4.591774562071592,
"grad_norm": 1.269822597503662,
"learning_rate": 6.898185328239468e-07,
"loss": 1.1097,
"step": 2011
},
{
"epoch": 4.594059405940594,
"grad_norm": 1.3368850946426392,
"learning_rate": 6.876720084412181e-07,
"loss": 1.1076,
"step": 2012
},
{
"epoch": 4.596344249809596,
"grad_norm": 1.2837021350860596,
"learning_rate": 6.85528296259261e-07,
"loss": 1.0565,
"step": 2013
},
{
"epoch": 4.598629093678599,
"grad_norm": 1.3289958238601685,
"learning_rate": 6.833873996044866e-07,
"loss": 1.0746,
"step": 2014
},
{
"epoch": 4.600913937547601,
"grad_norm": 1.3046926259994507,
"learning_rate": 6.812493217989368e-07,
"loss": 1.1318,
"step": 2015
},
{
"epoch": 4.603198781416603,
"grad_norm": 1.321399211883545,
"learning_rate": 6.791140661602826e-07,
"loss": 1.0987,
"step": 2016
},
{
"epoch": 4.605483625285606,
"grad_norm": 1.2870993614196777,
"learning_rate": 6.769816360018114e-07,
"loss": 1.0998,
"step": 2017
},
{
"epoch": 4.607768469154608,
"grad_norm": 1.2852782011032104,
"learning_rate": 6.748520346324302e-07,
"loss": 1.1106,
"step": 2018
},
{
"epoch": 4.61005331302361,
"grad_norm": 1.3015333414077759,
"learning_rate": 6.727252653566543e-07,
"loss": 1.1247,
"step": 2019
},
{
"epoch": 4.612338156892612,
"grad_norm": 1.2432022094726562,
"learning_rate": 6.70601331474604e-07,
"loss": 1.1086,
"step": 2020
},
{
"epoch": 4.614623000761615,
"grad_norm": 1.3117856979370117,
"learning_rate": 6.684802362820028e-07,
"loss": 1.0555,
"step": 2021
},
{
"epoch": 4.616907844630617,
"grad_norm": 1.3180614709854126,
"learning_rate": 6.663619830701662e-07,
"loss": 1.0653,
"step": 2022
},
{
"epoch": 4.619192688499619,
"grad_norm": 1.2816510200500488,
"learning_rate": 6.642465751260008e-07,
"loss": 1.1072,
"step": 2023
},
{
"epoch": 4.621477532368622,
"grad_norm": 1.2523847818374634,
"learning_rate": 6.621340157319998e-07,
"loss": 1.0875,
"step": 2024
},
{
"epoch": 4.623762376237623,
"grad_norm": 1.2979053258895874,
"learning_rate": 6.60024308166233e-07,
"loss": 1.1289,
"step": 2025
},
{
"epoch": 4.626047220106626,
"grad_norm": 1.3084304332733154,
"learning_rate": 6.57917455702349e-07,
"loss": 1.0658,
"step": 2026
},
{
"epoch": 4.628332063975629,
"grad_norm": 1.2778873443603516,
"learning_rate": 6.55813461609563e-07,
"loss": 1.1038,
"step": 2027
},
{
"epoch": 4.63061690784463,
"grad_norm": 1.283216953277588,
"learning_rate": 6.537123291526554e-07,
"loss": 1.0868,
"step": 2028
},
{
"epoch": 4.632901751713633,
"grad_norm": 1.2661612033843994,
"learning_rate": 6.516140615919677e-07,
"loss": 1.0919,
"step": 2029
},
{
"epoch": 4.635186595582635,
"grad_norm": 1.2835884094238281,
"learning_rate": 6.495186621833946e-07,
"loss": 1.0823,
"step": 2030
},
{
"epoch": 4.637471439451637,
"grad_norm": 1.2521917819976807,
"learning_rate": 6.474261341783791e-07,
"loss": 1.1061,
"step": 2031
},
{
"epoch": 4.63975628332064,
"grad_norm": 1.2880902290344238,
"learning_rate": 6.453364808239129e-07,
"loss": 1.1834,
"step": 2032
},
{
"epoch": 4.642041127189642,
"grad_norm": 1.284623384475708,
"learning_rate": 6.432497053625228e-07,
"loss": 1.1079,
"step": 2033
},
{
"epoch": 4.644325971058644,
"grad_norm": 1.3427263498306274,
"learning_rate": 6.411658110322708e-07,
"loss": 1.0893,
"step": 2034
},
{
"epoch": 4.646610814927646,
"grad_norm": 1.2741354703903198,
"learning_rate": 6.3908480106675e-07,
"loss": 1.0809,
"step": 2035
},
{
"epoch": 4.648895658796649,
"grad_norm": 1.3016525506973267,
"learning_rate": 6.370066786950766e-07,
"loss": 1.1164,
"step": 2036
},
{
"epoch": 4.651180502665651,
"grad_norm": 1.3607993125915527,
"learning_rate": 6.349314471418849e-07,
"loss": 1.051,
"step": 2037
},
{
"epoch": 4.653465346534653,
"grad_norm": 1.2685065269470215,
"learning_rate": 6.328591096273262e-07,
"loss": 1.0946,
"step": 2038
},
{
"epoch": 4.655750190403656,
"grad_norm": 1.3018466234207153,
"learning_rate": 6.307896693670587e-07,
"loss": 1.0744,
"step": 2039
},
{
"epoch": 4.658035034272658,
"grad_norm": 1.2309743165969849,
"learning_rate": 6.28723129572247e-07,
"loss": 1.062,
"step": 2040
},
{
"epoch": 4.66031987814166,
"grad_norm": 1.2842991352081299,
"learning_rate": 6.266594934495534e-07,
"loss": 1.1086,
"step": 2041
},
{
"epoch": 4.662604722010663,
"grad_norm": 1.304504156112671,
"learning_rate": 6.245987642011344e-07,
"loss": 1.085,
"step": 2042
},
{
"epoch": 4.6648895658796645,
"grad_norm": 1.3170430660247803,
"learning_rate": 6.225409450246378e-07,
"loss": 1.1018,
"step": 2043
},
{
"epoch": 4.667174409748667,
"grad_norm": 1.2923517227172852,
"learning_rate": 6.20486039113194e-07,
"loss": 1.0784,
"step": 2044
},
{
"epoch": 4.66945925361767,
"grad_norm": 1.282779335975647,
"learning_rate": 6.184340496554125e-07,
"loss": 1.1002,
"step": 2045
},
{
"epoch": 4.6717440974866715,
"grad_norm": 1.2920050621032715,
"learning_rate": 6.1638497983538e-07,
"loss": 1.1296,
"step": 2046
},
{
"epoch": 4.674028941355674,
"grad_norm": 1.3137156963348389,
"learning_rate": 6.143388328326494e-07,
"loss": 1.1436,
"step": 2047
},
{
"epoch": 4.676313785224677,
"grad_norm": 1.3416036367416382,
"learning_rate": 6.122956118222417e-07,
"loss": 1.0999,
"step": 2048
},
{
"epoch": 4.6785986290936785,
"grad_norm": 1.2436286211013794,
"learning_rate": 6.102553199746347e-07,
"loss": 1.0668,
"step": 2049
},
{
"epoch": 4.680883472962681,
"grad_norm": 1.2815945148468018,
"learning_rate": 6.082179604557617e-07,
"loss": 1.0659,
"step": 2050
},
{
"epoch": 4.683168316831683,
"grad_norm": 1.2817378044128418,
"learning_rate": 6.061835364270077e-07,
"loss": 1.0815,
"step": 2051
},
{
"epoch": 4.6854531607006855,
"grad_norm": 1.2972486019134521,
"learning_rate": 6.041520510452003e-07,
"loss": 1.0921,
"step": 2052
},
{
"epoch": 4.687738004569688,
"grad_norm": 1.270463466644287,
"learning_rate": 6.021235074626077e-07,
"loss": 1.0981,
"step": 2053
},
{
"epoch": 4.69002284843869,
"grad_norm": 1.2780388593673706,
"learning_rate": 6.000979088269354e-07,
"loss": 1.0588,
"step": 2054
},
{
"epoch": 4.6923076923076925,
"grad_norm": 1.2750800848007202,
"learning_rate": 5.980752582813157e-07,
"loss": 1.0553,
"step": 2055
},
{
"epoch": 4.694592536176694,
"grad_norm": 1.2434440851211548,
"learning_rate": 5.960555589643097e-07,
"loss": 1.1097,
"step": 2056
},
{
"epoch": 4.696877380045697,
"grad_norm": 1.2729369401931763,
"learning_rate": 5.940388140098971e-07,
"loss": 1.0809,
"step": 2057
},
{
"epoch": 4.6991622239146995,
"grad_norm": 1.2687339782714844,
"learning_rate": 5.92025026547473e-07,
"loss": 1.1042,
"step": 2058
},
{
"epoch": 4.701447067783701,
"grad_norm": 1.2984797954559326,
"learning_rate": 5.900141997018455e-07,
"loss": 1.091,
"step": 2059
},
{
"epoch": 4.703731911652704,
"grad_norm": 1.3156907558441162,
"learning_rate": 5.880063365932264e-07,
"loss": 1.0744,
"step": 2060
},
{
"epoch": 4.706016755521706,
"grad_norm": 1.2851028442382812,
"learning_rate": 5.860014403372291e-07,
"loss": 1.0939,
"step": 2061
},
{
"epoch": 4.708301599390708,
"grad_norm": 1.2899366617202759,
"learning_rate": 5.839995140448651e-07,
"loss": 1.0905,
"step": 2062
},
{
"epoch": 4.710586443259711,
"grad_norm": 1.2499885559082031,
"learning_rate": 5.820005608225345e-07,
"loss": 1.0711,
"step": 2063
},
{
"epoch": 4.712871287128713,
"grad_norm": 1.2643736600875854,
"learning_rate": 5.800045837720272e-07,
"loss": 1.1536,
"step": 2064
},
{
"epoch": 4.715156130997715,
"grad_norm": 1.31736421585083,
"learning_rate": 5.780115859905125e-07,
"loss": 1.0854,
"step": 2065
},
{
"epoch": 4.717440974866717,
"grad_norm": 1.2984356880187988,
"learning_rate": 5.760215705705371e-07,
"loss": 1.0978,
"step": 2066
},
{
"epoch": 4.71972581873572,
"grad_norm": 1.3062434196472168,
"learning_rate": 5.740345406000219e-07,
"loss": 1.1127,
"step": 2067
},
{
"epoch": 4.722010662604722,
"grad_norm": 1.2795484066009521,
"learning_rate": 5.720504991622528e-07,
"loss": 1.0881,
"step": 2068
},
{
"epoch": 4.724295506473724,
"grad_norm": 1.311261773109436,
"learning_rate": 5.700694493358791e-07,
"loss": 1.0802,
"step": 2069
},
{
"epoch": 4.726580350342727,
"grad_norm": 1.2954013347625732,
"learning_rate": 5.680913941949093e-07,
"loss": 1.0841,
"step": 2070
},
{
"epoch": 4.728865194211729,
"grad_norm": 1.2342921495437622,
"learning_rate": 5.66116336808703e-07,
"loss": 1.1164,
"step": 2071
},
{
"epoch": 4.731150038080731,
"grad_norm": 1.2950637340545654,
"learning_rate": 5.641442802419706e-07,
"loss": 1.0904,
"step": 2072
},
{
"epoch": 4.733434881949734,
"grad_norm": 1.2827471494674683,
"learning_rate": 5.621752275547637e-07,
"loss": 1.0693,
"step": 2073
},
{
"epoch": 4.735719725818735,
"grad_norm": 1.266324520111084,
"learning_rate": 5.602091818024732e-07,
"loss": 1.1204,
"step": 2074
},
{
"epoch": 4.738004569687738,
"grad_norm": 1.2678186893463135,
"learning_rate": 5.582461460358265e-07,
"loss": 1.0701,
"step": 2075
},
{
"epoch": 4.740289413556741,
"grad_norm": 1.2703529596328735,
"learning_rate": 5.562861233008774e-07,
"loss": 1.0974,
"step": 2076
},
{
"epoch": 4.742574257425742,
"grad_norm": 1.2783465385437012,
"learning_rate": 5.543291166390047e-07,
"loss": 1.109,
"step": 2077
},
{
"epoch": 4.744859101294745,
"grad_norm": 1.2746630907058716,
"learning_rate": 5.523751290869098e-07,
"loss": 1.0967,
"step": 2078
},
{
"epoch": 4.747143945163748,
"grad_norm": 1.2512940168380737,
"learning_rate": 5.504241636766064e-07,
"loss": 1.0544,
"step": 2079
},
{
"epoch": 4.749428789032749,
"grad_norm": 1.29820716381073,
"learning_rate": 5.484762234354196e-07,
"loss": 1.0888,
"step": 2080
},
{
"epoch": 4.751713632901752,
"grad_norm": 1.2650960683822632,
"learning_rate": 5.465313113859816e-07,
"loss": 1.0936,
"step": 2081
},
{
"epoch": 4.753998476770754,
"grad_norm": 1.2510480880737305,
"learning_rate": 5.445894305462235e-07,
"loss": 1.1078,
"step": 2082
},
{
"epoch": 4.756283320639756,
"grad_norm": 1.258803367614746,
"learning_rate": 5.426505839293752e-07,
"loss": 1.1022,
"step": 2083
},
{
"epoch": 4.758568164508759,
"grad_norm": 1.252769112586975,
"learning_rate": 5.407147745439567e-07,
"loss": 1.103,
"step": 2084
},
{
"epoch": 4.760853008377761,
"grad_norm": 1.2640199661254883,
"learning_rate": 5.387820053937751e-07,
"loss": 1.09,
"step": 2085
},
{
"epoch": 4.763137852246763,
"grad_norm": 1.3083833456039429,
"learning_rate": 5.368522794779213e-07,
"loss": 1.1041,
"step": 2086
},
{
"epoch": 4.765422696115765,
"grad_norm": 1.2743126153945923,
"learning_rate": 5.349255997907632e-07,
"loss": 1.1266,
"step": 2087
},
{
"epoch": 4.767707539984768,
"grad_norm": 1.2865676879882812,
"learning_rate": 5.330019693219405e-07,
"loss": 1.1083,
"step": 2088
},
{
"epoch": 4.76999238385377,
"grad_norm": 1.2562209367752075,
"learning_rate": 5.310813910563645e-07,
"loss": 1.1015,
"step": 2089
},
{
"epoch": 4.772277227722772,
"grad_norm": 1.2887158393859863,
"learning_rate": 5.29163867974207e-07,
"loss": 1.103,
"step": 2090
},
{
"epoch": 4.774562071591775,
"grad_norm": 1.2844069004058838,
"learning_rate": 5.272494030509023e-07,
"loss": 1.0665,
"step": 2091
},
{
"epoch": 4.776846915460776,
"grad_norm": 1.29599928855896,
"learning_rate": 5.253379992571367e-07,
"loss": 1.0937,
"step": 2092
},
{
"epoch": 4.779131759329779,
"grad_norm": 1.267896056175232,
"learning_rate": 5.23429659558847e-07,
"loss": 1.0889,
"step": 2093
},
{
"epoch": 4.781416603198782,
"grad_norm": 1.2771025896072388,
"learning_rate": 5.215243869172173e-07,
"loss": 1.1208,
"step": 2094
},
{
"epoch": 4.783701447067783,
"grad_norm": 1.2694123983383179,
"learning_rate": 5.196221842886704e-07,
"loss": 1.0945,
"step": 2095
},
{
"epoch": 4.785986290936786,
"grad_norm": 1.3127267360687256,
"learning_rate": 5.177230546248654e-07,
"loss": 1.0761,
"step": 2096
},
{
"epoch": 4.788271134805788,
"grad_norm": 1.2706137895584106,
"learning_rate": 5.158270008726954e-07,
"loss": 1.0699,
"step": 2097
},
{
"epoch": 4.79055597867479,
"grad_norm": 1.30772066116333,
"learning_rate": 5.139340259742776e-07,
"loss": 1.1368,
"step": 2098
},
{
"epoch": 4.792840822543793,
"grad_norm": 1.2849667072296143,
"learning_rate": 5.12044132866954e-07,
"loss": 1.0956,
"step": 2099
},
{
"epoch": 4.795125666412795,
"grad_norm": 1.3047093152999878,
"learning_rate": 5.101573244832833e-07,
"loss": 1.1047,
"step": 2100
},
{
"epoch": 4.797410510281797,
"grad_norm": 1.327357292175293,
"learning_rate": 5.082736037510371e-07,
"loss": 1.1343,
"step": 2101
},
{
"epoch": 4.7996953541508,
"grad_norm": 1.2878531217575073,
"learning_rate": 5.063929735931985e-07,
"loss": 1.1009,
"step": 2102
},
{
"epoch": 4.801980198019802,
"grad_norm": 1.2837164402008057,
"learning_rate": 5.04515436927952e-07,
"loss": 1.057,
"step": 2103
},
{
"epoch": 4.804265041888804,
"grad_norm": 1.3011730909347534,
"learning_rate": 5.026409966686827e-07,
"loss": 1.0795,
"step": 2104
},
{
"epoch": 4.806549885757806,
"grad_norm": 1.266371250152588,
"learning_rate": 5.007696557239727e-07,
"loss": 1.1324,
"step": 2105
},
{
"epoch": 4.808834729626809,
"grad_norm": 1.266931176185608,
"learning_rate": 4.989014169975919e-07,
"loss": 1.1123,
"step": 2106
},
{
"epoch": 4.811119573495811,
"grad_norm": 1.2623364925384521,
"learning_rate": 4.970362833885001e-07,
"loss": 1.1137,
"step": 2107
},
{
"epoch": 4.813404417364813,
"grad_norm": 1.2980166673660278,
"learning_rate": 4.951742577908355e-07,
"loss": 1.0762,
"step": 2108
},
{
"epoch": 4.815689261233816,
"grad_norm": 1.2445948123931885,
"learning_rate": 4.933153430939147e-07,
"loss": 1.0828,
"step": 2109
},
{
"epoch": 4.817974105102818,
"grad_norm": 1.2619681358337402,
"learning_rate": 4.91459542182228e-07,
"loss": 1.0586,
"step": 2110
},
{
"epoch": 4.82025894897182,
"grad_norm": 1.2932782173156738,
"learning_rate": 4.896068579354338e-07,
"loss": 1.0878,
"step": 2111
},
{
"epoch": 4.822543792840823,
"grad_norm": 1.2484990358352661,
"learning_rate": 4.877572932283528e-07,
"loss": 1.1227,
"step": 2112
},
{
"epoch": 4.8248286367098245,
"grad_norm": 1.3282291889190674,
"learning_rate": 4.859108509309676e-07,
"loss": 1.1142,
"step": 2113
},
{
"epoch": 4.827113480578827,
"grad_norm": 1.266524314880371,
"learning_rate": 4.840675339084136e-07,
"loss": 1.0458,
"step": 2114
},
{
"epoch": 4.82939832444783,
"grad_norm": 1.2742689847946167,
"learning_rate": 4.822273450209767e-07,
"loss": 1.1246,
"step": 2115
},
{
"epoch": 4.8316831683168315,
"grad_norm": 1.246359944343567,
"learning_rate": 4.803902871240909e-07,
"loss": 1.1243,
"step": 2116
},
{
"epoch": 4.833968012185834,
"grad_norm": 1.295499563217163,
"learning_rate": 4.785563630683301e-07,
"loss": 1.0989,
"step": 2117
},
{
"epoch": 4.836252856054836,
"grad_norm": 1.2707738876342773,
"learning_rate": 4.767255756994049e-07,
"loss": 1.132,
"step": 2118
},
{
"epoch": 4.8385376999238385,
"grad_norm": 1.2814133167266846,
"learning_rate": 4.7489792785816116e-07,
"loss": 1.0663,
"step": 2119
},
{
"epoch": 4.840822543792841,
"grad_norm": 1.3483997583389282,
"learning_rate": 4.7307342238056996e-07,
"loss": 1.0886,
"step": 2120
},
{
"epoch": 4.843107387661843,
"grad_norm": 1.3065505027770996,
"learning_rate": 4.7125206209772955e-07,
"loss": 1.0979,
"step": 2121
},
{
"epoch": 4.8453922315308455,
"grad_norm": 1.2689411640167236,
"learning_rate": 4.6943384983585553e-07,
"loss": 1.1147,
"step": 2122
},
{
"epoch": 4.847677075399847,
"grad_norm": 1.2981805801391602,
"learning_rate": 4.676187884162783e-07,
"loss": 1.0844,
"step": 2123
},
{
"epoch": 4.84996191926885,
"grad_norm": 1.2894749641418457,
"learning_rate": 4.6580688065544164e-07,
"loss": 1.1104,
"step": 2124
},
{
"epoch": 4.8522467631378525,
"grad_norm": 1.2899259328842163,
"learning_rate": 4.6399812936489376e-07,
"loss": 1.1248,
"step": 2125
},
{
"epoch": 4.854531607006854,
"grad_norm": 1.302597165107727,
"learning_rate": 4.621925373512845e-07,
"loss": 1.0847,
"step": 2126
},
{
"epoch": 4.856816450875857,
"grad_norm": 1.3532848358154297,
"learning_rate": 4.603901074163639e-07,
"loss": 1.1002,
"step": 2127
},
{
"epoch": 4.859101294744859,
"grad_norm": 1.2945899963378906,
"learning_rate": 4.5859084235697236e-07,
"loss": 1.067,
"step": 2128
},
{
"epoch": 4.861386138613861,
"grad_norm": 1.3042470216751099,
"learning_rate": 4.5679474496504227e-07,
"loss": 1.0435,
"step": 2129
},
{
"epoch": 4.863670982482864,
"grad_norm": 1.3194514513015747,
"learning_rate": 4.550018180275886e-07,
"loss": 1.1081,
"step": 2130
},
{
"epoch": 4.865955826351866,
"grad_norm": 1.285345196723938,
"learning_rate": 4.5321206432670684e-07,
"loss": 1.0663,
"step": 2131
},
{
"epoch": 4.868240670220868,
"grad_norm": 1.2676403522491455,
"learning_rate": 4.5142548663957006e-07,
"loss": 1.0603,
"step": 2132
},
{
"epoch": 4.870525514089871,
"grad_norm": 1.2659584283828735,
"learning_rate": 4.4964208773842167e-07,
"loss": 1.1275,
"step": 2133
},
{
"epoch": 4.872810357958873,
"grad_norm": 1.3056273460388184,
"learning_rate": 4.478618703905724e-07,
"loss": 1.0908,
"step": 2134
},
{
"epoch": 4.875095201827875,
"grad_norm": 1.2657009363174438,
"learning_rate": 4.460848373583981e-07,
"loss": 1.0994,
"step": 2135
},
{
"epoch": 4.877380045696877,
"grad_norm": 1.2820390462875366,
"learning_rate": 4.4431099139933043e-07,
"loss": 1.1124,
"step": 2136
},
{
"epoch": 4.87966488956588,
"grad_norm": 1.2907642126083374,
"learning_rate": 4.4254033526585917e-07,
"loss": 1.1237,
"step": 2137
},
{
"epoch": 4.881949733434882,
"grad_norm": 1.3053464889526367,
"learning_rate": 4.407728717055215e-07,
"loss": 1.1263,
"step": 2138
},
{
"epoch": 4.884234577303884,
"grad_norm": 1.2887327671051025,
"learning_rate": 4.3900860346090146e-07,
"loss": 1.1195,
"step": 2139
},
{
"epoch": 4.886519421172887,
"grad_norm": 1.331275463104248,
"learning_rate": 4.3724753326962634e-07,
"loss": 1.0566,
"step": 2140
},
{
"epoch": 4.888804265041889,
"grad_norm": 1.290602445602417,
"learning_rate": 4.354896638643591e-07,
"loss": 1.0698,
"step": 2141
},
{
"epoch": 4.891089108910891,
"grad_norm": 1.3247865438461304,
"learning_rate": 4.3373499797279646e-07,
"loss": 1.0968,
"step": 2142
},
{
"epoch": 4.893373952779894,
"grad_norm": 1.2982738018035889,
"learning_rate": 4.319835383176657e-07,
"loss": 1.1062,
"step": 2143
},
{
"epoch": 4.895658796648895,
"grad_norm": 1.2912477254867554,
"learning_rate": 4.3023528761671624e-07,
"loss": 1.0881,
"step": 2144
},
{
"epoch": 4.897943640517898,
"grad_norm": 1.2807631492614746,
"learning_rate": 4.284902485827214e-07,
"loss": 1.1088,
"step": 2145
},
{
"epoch": 4.900228484386901,
"grad_norm": 1.339166283607483,
"learning_rate": 4.2674842392346844e-07,
"loss": 1.0903,
"step": 2146
},
{
"epoch": 4.902513328255902,
"grad_norm": 1.2816892862319946,
"learning_rate": 4.250098163417571e-07,
"loss": 1.1297,
"step": 2147
},
{
"epoch": 4.904798172124905,
"grad_norm": 1.2935080528259277,
"learning_rate": 4.232744285353968e-07,
"loss": 1.1105,
"step": 2148
},
{
"epoch": 4.907083015993907,
"grad_norm": 1.3085383176803589,
"learning_rate": 4.215422631971994e-07,
"loss": 1.1381,
"step": 2149
},
{
"epoch": 4.909367859862909,
"grad_norm": 1.2890045642852783,
"learning_rate": 4.1981332301497563e-07,
"loss": 1.1191,
"step": 2150
},
{
"epoch": 4.911652703731912,
"grad_norm": 1.2732130289077759,
"learning_rate": 4.180876106715348e-07,
"loss": 1.0815,
"step": 2151
},
{
"epoch": 4.913937547600914,
"grad_norm": 1.2872875928878784,
"learning_rate": 4.1636512884467404e-07,
"loss": 1.1037,
"step": 2152
},
{
"epoch": 4.916222391469916,
"grad_norm": 1.283339023590088,
"learning_rate": 4.146458802071804e-07,
"loss": 1.0855,
"step": 2153
},
{
"epoch": 4.918507235338918,
"grad_norm": 1.2937120199203491,
"learning_rate": 4.129298674268226e-07,
"loss": 1.1246,
"step": 2154
},
{
"epoch": 4.920792079207921,
"grad_norm": 1.3152278661727905,
"learning_rate": 4.112170931663473e-07,
"loss": 1.1028,
"step": 2155
},
{
"epoch": 4.923076923076923,
"grad_norm": 1.2731553316116333,
"learning_rate": 4.0950756008347886e-07,
"loss": 1.102,
"step": 2156
},
{
"epoch": 4.925361766945925,
"grad_norm": 1.2939971685409546,
"learning_rate": 4.0780127083090956e-07,
"loss": 1.0797,
"step": 2157
},
{
"epoch": 4.927646610814928,
"grad_norm": 1.2959516048431396,
"learning_rate": 4.060982280562989e-07,
"loss": 1.0698,
"step": 2158
},
{
"epoch": 4.9299314546839295,
"grad_norm": 1.289335012435913,
"learning_rate": 4.0439843440227027e-07,
"loss": 1.0682,
"step": 2159
},
{
"epoch": 4.932216298552932,
"grad_norm": 1.2806504964828491,
"learning_rate": 4.0270189250640273e-07,
"loss": 1.0625,
"step": 2160
},
{
"epoch": 4.934501142421935,
"grad_norm": 1.2870630025863647,
"learning_rate": 4.010086050012324e-07,
"loss": 1.0861,
"step": 2161
},
{
"epoch": 4.9367859862909365,
"grad_norm": 1.2742153406143188,
"learning_rate": 3.993185745142436e-07,
"loss": 1.0794,
"step": 2162
},
{
"epoch": 4.939070830159939,
"grad_norm": 1.2898108959197998,
"learning_rate": 3.976318036678667e-07,
"loss": 1.086,
"step": 2163
},
{
"epoch": 4.941355674028942,
"grad_norm": 1.2475721836090088,
"learning_rate": 3.9594829507947573e-07,
"loss": 1.0961,
"step": 2164
},
{
"epoch": 4.9436405178979435,
"grad_norm": 1.2730660438537598,
"learning_rate": 3.9426805136138086e-07,
"loss": 1.0954,
"step": 2165
},
{
"epoch": 4.945925361766946,
"grad_norm": 1.2580710649490356,
"learning_rate": 3.9259107512082674e-07,
"loss": 1.073,
"step": 2166
},
{
"epoch": 4.948210205635948,
"grad_norm": 1.2623883485794067,
"learning_rate": 3.9091736895998907e-07,
"loss": 1.0568,
"step": 2167
},
{
"epoch": 4.9504950495049505,
"grad_norm": 1.280324101448059,
"learning_rate": 3.8924693547596647e-07,
"loss": 1.1061,
"step": 2168
},
{
"epoch": 4.952779893373953,
"grad_norm": 1.276982069015503,
"learning_rate": 3.875797772607831e-07,
"loss": 1.0939,
"step": 2169
},
{
"epoch": 4.955064737242955,
"grad_norm": 1.2728464603424072,
"learning_rate": 3.859158969013777e-07,
"loss": 1.0977,
"step": 2170
},
{
"epoch": 4.9573495811119574,
"grad_norm": 1.3238383531570435,
"learning_rate": 3.842552969796037e-07,
"loss": 1.0895,
"step": 2171
},
{
"epoch": 4.95963442498096,
"grad_norm": 1.2770859003067017,
"learning_rate": 3.8259798007222515e-07,
"loss": 1.0748,
"step": 2172
},
{
"epoch": 4.961919268849962,
"grad_norm": 1.2510464191436768,
"learning_rate": 3.8094394875091117e-07,
"loss": 1.1028,
"step": 2173
},
{
"epoch": 4.9642041127189644,
"grad_norm": 1.284297227859497,
"learning_rate": 3.7929320558223153e-07,
"loss": 1.0561,
"step": 2174
},
{
"epoch": 4.966488956587966,
"grad_norm": 1.3043745756149292,
"learning_rate": 3.7764575312765644e-07,
"loss": 1.0553,
"step": 2175
},
{
"epoch": 4.968773800456969,
"grad_norm": 1.2932723760604858,
"learning_rate": 3.7600159394354657e-07,
"loss": 1.1137,
"step": 2176
},
{
"epoch": 4.971058644325971,
"grad_norm": 1.2768579721450806,
"learning_rate": 3.743607305811556e-07,
"loss": 1.1118,
"step": 2177
},
{
"epoch": 4.973343488194973,
"grad_norm": 1.2791515588760376,
"learning_rate": 3.7272316558662094e-07,
"loss": 1.0952,
"step": 2178
},
{
"epoch": 4.975628332063976,
"grad_norm": 1.2823758125305176,
"learning_rate": 3.7108890150096166e-07,
"loss": 1.0373,
"step": 2179
},
{
"epoch": 4.9779131759329776,
"grad_norm": 1.3007476329803467,
"learning_rate": 3.6945794086007706e-07,
"loss": 1.066,
"step": 2180
},
{
"epoch": 4.98019801980198,
"grad_norm": 1.3126449584960938,
"learning_rate": 3.678302861947386e-07,
"loss": 1.077,
"step": 2181
},
{
"epoch": 4.982482863670983,
"grad_norm": 1.3028063774108887,
"learning_rate": 3.6620594003058756e-07,
"loss": 1.1003,
"step": 2182
},
{
"epoch": 4.9847677075399845,
"grad_norm": 1.2599045038223267,
"learning_rate": 3.645849048881331e-07,
"loss": 1.0912,
"step": 2183
},
{
"epoch": 4.987052551408987,
"grad_norm": 1.275844931602478,
"learning_rate": 3.629671832827447e-07,
"loss": 1.1011,
"step": 2184
},
{
"epoch": 4.989337395277989,
"grad_norm": 1.259772777557373,
"learning_rate": 3.6135277772465226e-07,
"loss": 1.0799,
"step": 2185
},
{
"epoch": 4.9916222391469915,
"grad_norm": 1.2567813396453857,
"learning_rate": 3.597416907189388e-07,
"loss": 1.12,
"step": 2186
},
{
"epoch": 4.993907083015994,
"grad_norm": 1.2930412292480469,
"learning_rate": 3.5813392476553714e-07,
"loss": 1.0947,
"step": 2187
},
{
"epoch": 4.996191926884996,
"grad_norm": 1.2559951543807983,
"learning_rate": 3.565294823592291e-07,
"loss": 1.0462,
"step": 2188
},
{
"epoch": 4.9984767707539985,
"grad_norm": 1.2923510074615479,
"learning_rate": 3.5492836598963715e-07,
"loss": 1.0844,
"step": 2189
},
{
"epoch": 5.0,
"grad_norm": 1.6771284341812134,
"learning_rate": 3.5333057814122414e-07,
"loss": 1.074,
"step": 2190
},
{
"epoch": 5.002284843869003,
"grad_norm": 1.487131118774414,
"learning_rate": 3.517361212932879e-07,
"loss": 1.0851,
"step": 2191
},
{
"epoch": 5.004569687738004,
"grad_norm": 1.3028117418289185,
"learning_rate": 3.50144997919957e-07,
"loss": 1.0614,
"step": 2192
},
{
"epoch": 5.006854531607007,
"grad_norm": 1.3251090049743652,
"learning_rate": 3.485572104901869e-07,
"loss": 1.0994,
"step": 2193
},
{
"epoch": 5.009139375476009,
"grad_norm": 1.2535759210586548,
"learning_rate": 3.4697276146775817e-07,
"loss": 1.0573,
"step": 2194
},
{
"epoch": 5.011424219345011,
"grad_norm": 1.320404052734375,
"learning_rate": 3.4539165331127e-07,
"loss": 1.07,
"step": 2195
},
{
"epoch": 5.013709063214014,
"grad_norm": 1.2744762897491455,
"learning_rate": 3.438138884741377e-07,
"loss": 1.064,
"step": 2196
},
{
"epoch": 5.015993907083016,
"grad_norm": 1.2865326404571533,
"learning_rate": 3.422394694045894e-07,
"loss": 1.0918,
"step": 2197
},
{
"epoch": 5.018278750952018,
"grad_norm": 1.3152594566345215,
"learning_rate": 3.406683985456605e-07,
"loss": 1.0854,
"step": 2198
},
{
"epoch": 5.020563594821021,
"grad_norm": 1.290440320968628,
"learning_rate": 3.3910067833519215e-07,
"loss": 1.1513,
"step": 2199
},
{
"epoch": 5.022848438690023,
"grad_norm": 1.2581883668899536,
"learning_rate": 3.375363112058255e-07,
"loss": 1.0887,
"step": 2200
},
{
"epoch": 5.025133282559025,
"grad_norm": 1.3203215599060059,
"learning_rate": 3.359752995849985e-07,
"loss": 1.0834,
"step": 2201
},
{
"epoch": 5.027418126428027,
"grad_norm": 1.2735775709152222,
"learning_rate": 3.344176458949436e-07,
"loss": 1.0951,
"step": 2202
},
{
"epoch": 5.02970297029703,
"grad_norm": 1.2835285663604736,
"learning_rate": 3.328633525526812e-07,
"loss": 1.0732,
"step": 2203
},
{
"epoch": 5.031987814166032,
"grad_norm": 1.2780119180679321,
"learning_rate": 3.3131242197001754e-07,
"loss": 1.0888,
"step": 2204
},
{
"epoch": 5.034272658035034,
"grad_norm": 1.295435905456543,
"learning_rate": 3.2976485655354243e-07,
"loss": 1.0786,
"step": 2205
},
{
"epoch": 5.036557501904037,
"grad_norm": 1.2823036909103394,
"learning_rate": 3.2822065870462216e-07,
"loss": 1.0481,
"step": 2206
},
{
"epoch": 5.0388423457730385,
"grad_norm": 1.2593215703964233,
"learning_rate": 3.266798308193991e-07,
"loss": 1.0899,
"step": 2207
},
{
"epoch": 5.041127189642041,
"grad_norm": 1.272403359413147,
"learning_rate": 3.2514237528878503e-07,
"loss": 1.0549,
"step": 2208
},
{
"epoch": 5.043412033511044,
"grad_norm": 1.2577238082885742,
"learning_rate": 3.236082944984589e-07,
"loss": 1.1043,
"step": 2209
},
{
"epoch": 5.0456968773800455,
"grad_norm": 1.2967652082443237,
"learning_rate": 3.2207759082886454e-07,
"loss": 1.1109,
"step": 2210
},
{
"epoch": 5.047981721249048,
"grad_norm": 1.263169288635254,
"learning_rate": 3.205502666552041e-07,
"loss": 1.0551,
"step": 2211
},
{
"epoch": 5.05026656511805,
"grad_norm": 1.2983647584915161,
"learning_rate": 3.1902632434743563e-07,
"loss": 1.0928,
"step": 2212
},
{
"epoch": 5.0525514089870525,
"grad_norm": 1.308638095855713,
"learning_rate": 3.175057662702707e-07,
"loss": 1.0659,
"step": 2213
},
{
"epoch": 5.054836252856055,
"grad_norm": 1.3027294874191284,
"learning_rate": 3.1598859478316814e-07,
"loss": 1.0821,
"step": 2214
},
{
"epoch": 5.057121096725057,
"grad_norm": 1.280255675315857,
"learning_rate": 3.1447481224033376e-07,
"loss": 1.1082,
"step": 2215
},
{
"epoch": 5.0594059405940595,
"grad_norm": 1.2913918495178223,
"learning_rate": 3.1296442099071235e-07,
"loss": 1.0666,
"step": 2216
},
{
"epoch": 5.061690784463062,
"grad_norm": 1.2657502889633179,
"learning_rate": 3.114574233779877e-07,
"loss": 1.1096,
"step": 2217
},
{
"epoch": 5.063975628332064,
"grad_norm": 1.2814840078353882,
"learning_rate": 3.099538217405779e-07,
"loss": 1.0969,
"step": 2218
},
{
"epoch": 5.0662604722010665,
"grad_norm": 1.2882423400878906,
"learning_rate": 3.08453618411631e-07,
"loss": 1.0349,
"step": 2219
},
{
"epoch": 5.068545316070068,
"grad_norm": 1.3120388984680176,
"learning_rate": 3.069568157190214e-07,
"loss": 1.1251,
"step": 2220
},
{
"epoch": 5.070830159939071,
"grad_norm": 1.2447402477264404,
"learning_rate": 3.054634159853481e-07,
"loss": 1.092,
"step": 2221
},
{
"epoch": 5.0731150038080735,
"grad_norm": 1.2926756143569946,
"learning_rate": 3.03973421527928e-07,
"loss": 1.0963,
"step": 2222
},
{
"epoch": 5.075399847677075,
"grad_norm": 1.255759835243225,
"learning_rate": 3.0248683465879587e-07,
"loss": 1.0646,
"step": 2223
},
{
"epoch": 5.077684691546078,
"grad_norm": 1.2734875679016113,
"learning_rate": 3.0100365768469753e-07,
"loss": 1.0786,
"step": 2224
},
{
"epoch": 5.07996953541508,
"grad_norm": 1.299229621887207,
"learning_rate": 2.995238929070876e-07,
"loss": 1.0652,
"step": 2225
},
{
"epoch": 5.082254379284082,
"grad_norm": 1.2811152935028076,
"learning_rate": 2.980475426221269e-07,
"loss": 1.085,
"step": 2226
},
{
"epoch": 5.084539223153085,
"grad_norm": 1.2939437627792358,
"learning_rate": 2.9657460912067753e-07,
"loss": 1.1039,
"step": 2227
},
{
"epoch": 5.086824067022087,
"grad_norm": 1.2752275466918945,
"learning_rate": 2.951050946882991e-07,
"loss": 1.0901,
"step": 2228
},
{
"epoch": 5.089108910891089,
"grad_norm": 1.2718929052352905,
"learning_rate": 2.9363900160524734e-07,
"loss": 1.0727,
"step": 2229
},
{
"epoch": 5.091393754760092,
"grad_norm": 1.33417546749115,
"learning_rate": 2.9217633214646713e-07,
"loss": 1.0994,
"step": 2230
},
{
"epoch": 5.093678598629094,
"grad_norm": 1.3067365884780884,
"learning_rate": 2.907170885815927e-07,
"loss": 1.1239,
"step": 2231
},
{
"epoch": 5.095963442498096,
"grad_norm": 1.3246252536773682,
"learning_rate": 2.892612731749414e-07,
"loss": 1.1102,
"step": 2232
},
{
"epoch": 5.098248286367098,
"grad_norm": 1.3094791173934937,
"learning_rate": 2.8780888818551014e-07,
"loss": 1.0473,
"step": 2233
},
{
"epoch": 5.1005331302361006,
"grad_norm": 1.2624950408935547,
"learning_rate": 2.8635993586697555e-07,
"loss": 1.1001,
"step": 2234
},
{
"epoch": 5.102817974105103,
"grad_norm": 1.3205031156539917,
"learning_rate": 2.8491441846768497e-07,
"loss": 1.0823,
"step": 2235
},
{
"epoch": 5.105102817974105,
"grad_norm": 1.2623871564865112,
"learning_rate": 2.8347233823065644e-07,
"loss": 1.0909,
"step": 2236
},
{
"epoch": 5.1073876618431076,
"grad_norm": 1.262770414352417,
"learning_rate": 2.820336973935764e-07,
"loss": 1.078,
"step": 2237
},
{
"epoch": 5.109672505712109,
"grad_norm": 1.2645272016525269,
"learning_rate": 2.805984981887916e-07,
"loss": 1.0378,
"step": 2238
},
{
"epoch": 5.111957349581112,
"grad_norm": 1.2669461965560913,
"learning_rate": 2.791667428433106e-07,
"loss": 1.0587,
"step": 2239
},
{
"epoch": 5.1142421934501145,
"grad_norm": 1.270908236503601,
"learning_rate": 2.777384335787969e-07,
"loss": 1.0601,
"step": 2240
},
{
"epoch": 5.116527037319116,
"grad_norm": 1.3161948919296265,
"learning_rate": 2.763135726115662e-07,
"loss": 1.0927,
"step": 2241
},
{
"epoch": 5.118811881188119,
"grad_norm": 1.2516229152679443,
"learning_rate": 2.7489216215258574e-07,
"loss": 1.0928,
"step": 2242
},
{
"epoch": 5.1210967250571215,
"grad_norm": 1.303397297859192,
"learning_rate": 2.734742044074659e-07,
"loss": 1.1116,
"step": 2243
},
{
"epoch": 5.123381568926123,
"grad_norm": 1.3074421882629395,
"learning_rate": 2.720597015764609e-07,
"loss": 1.0671,
"step": 2244
},
{
"epoch": 5.125666412795126,
"grad_norm": 1.3157670497894287,
"learning_rate": 2.706486558544644e-07,
"loss": 1.0835,
"step": 2245
},
{
"epoch": 5.127951256664128,
"grad_norm": 1.2734421491622925,
"learning_rate": 2.692410694310035e-07,
"loss": 1.0818,
"step": 2246
},
{
"epoch": 5.13023610053313,
"grad_norm": 1.315611720085144,
"learning_rate": 2.678369444902407e-07,
"loss": 1.0806,
"step": 2247
},
{
"epoch": 5.132520944402133,
"grad_norm": 1.2910858392715454,
"learning_rate": 2.664362832109643e-07,
"loss": 1.044,
"step": 2248
},
{
"epoch": 5.134805788271135,
"grad_norm": 1.3554924726486206,
"learning_rate": 2.6503908776658956e-07,
"loss": 1.0666,
"step": 2249
},
{
"epoch": 5.137090632140137,
"grad_norm": 1.3038955926895142,
"learning_rate": 2.6364536032515346e-07,
"loss": 1.0704,
"step": 2250
},
{
"epoch": 5.139375476009139,
"grad_norm": 1.2819843292236328,
"learning_rate": 2.62255103049312e-07,
"loss": 1.107,
"step": 2251
},
{
"epoch": 5.141660319878142,
"grad_norm": 1.2703412771224976,
"learning_rate": 2.6086831809633535e-07,
"loss": 1.104,
"step": 2252
},
{
"epoch": 5.143945163747144,
"grad_norm": 1.2854193449020386,
"learning_rate": 2.594850076181074e-07,
"loss": 1.0775,
"step": 2253
},
{
"epoch": 5.146230007616146,
"grad_norm": 1.294018268585205,
"learning_rate": 2.581051737611193e-07,
"loss": 1.0748,
"step": 2254
},
{
"epoch": 5.148514851485149,
"grad_norm": 1.3650598526000977,
"learning_rate": 2.567288186664674e-07,
"loss": 1.1148,
"step": 2255
},
{
"epoch": 5.15079969535415,
"grad_norm": 1.281511664390564,
"learning_rate": 2.553559444698517e-07,
"loss": 1.0872,
"step": 2256
},
{
"epoch": 5.153084539223153,
"grad_norm": 1.2767432928085327,
"learning_rate": 2.539865533015684e-07,
"loss": 1.0875,
"step": 2257
},
{
"epoch": 5.155369383092156,
"grad_norm": 1.314460277557373,
"learning_rate": 2.52620647286512e-07,
"loss": 1.1094,
"step": 2258
},
{
"epoch": 5.157654226961157,
"grad_norm": 1.315940260887146,
"learning_rate": 2.5125822854416644e-07,
"loss": 1.0909,
"step": 2259
},
{
"epoch": 5.15993907083016,
"grad_norm": 1.3246798515319824,
"learning_rate": 2.498992991886054e-07,
"loss": 1.0659,
"step": 2260
},
{
"epoch": 5.162223914699163,
"grad_norm": 1.3555991649627686,
"learning_rate": 2.485438613284888e-07,
"loss": 1.0844,
"step": 2261
},
{
"epoch": 5.164508758568164,
"grad_norm": 1.320139765739441,
"learning_rate": 2.471919170670578e-07,
"loss": 1.0691,
"step": 2262
},
{
"epoch": 5.166793602437167,
"grad_norm": 1.2948224544525146,
"learning_rate": 2.4584346850213216e-07,
"loss": 1.0834,
"step": 2263
},
{
"epoch": 5.169078446306169,
"grad_norm": 1.2399736642837524,
"learning_rate": 2.4449851772610917e-07,
"loss": 1.0964,
"step": 2264
},
{
"epoch": 5.171363290175171,
"grad_norm": 1.3757954835891724,
"learning_rate": 2.4315706682595613e-07,
"loss": 1.099,
"step": 2265
},
{
"epoch": 5.173648134044174,
"grad_norm": 1.273930549621582,
"learning_rate": 2.4181911788321243e-07,
"loss": 1.0505,
"step": 2266
},
{
"epoch": 5.175932977913176,
"grad_norm": 1.2793792486190796,
"learning_rate": 2.40484672973981e-07,
"loss": 1.0856,
"step": 2267
},
{
"epoch": 5.178217821782178,
"grad_norm": 1.3076246976852417,
"learning_rate": 2.391537341689276e-07,
"loss": 1.105,
"step": 2268
},
{
"epoch": 5.18050266565118,
"grad_norm": 1.3009895086288452,
"learning_rate": 2.3782630353328007e-07,
"loss": 1.0957,
"step": 2269
},
{
"epoch": 5.182787509520183,
"grad_norm": 1.2714428901672363,
"learning_rate": 2.3650238312682012e-07,
"loss": 1.0758,
"step": 2270
},
{
"epoch": 5.185072353389185,
"grad_norm": 1.292037844657898,
"learning_rate": 2.3518197500388278e-07,
"loss": 1.0341,
"step": 2271
},
{
"epoch": 5.187357197258187,
"grad_norm": 1.3077787160873413,
"learning_rate": 2.3386508121335472e-07,
"loss": 1.0913,
"step": 2272
},
{
"epoch": 5.18964204112719,
"grad_norm": 1.2806161642074585,
"learning_rate": 2.3255170379866799e-07,
"loss": 1.092,
"step": 2273
},
{
"epoch": 5.191926884996192,
"grad_norm": 1.2643924951553345,
"learning_rate": 2.3124184479779767e-07,
"loss": 1.0798,
"step": 2274
},
{
"epoch": 5.194211728865194,
"grad_norm": 1.323236107826233,
"learning_rate": 2.2993550624326178e-07,
"loss": 1.1076,
"step": 2275
},
{
"epoch": 5.196496572734197,
"grad_norm": 1.2984707355499268,
"learning_rate": 2.2863269016211276e-07,
"loss": 1.1025,
"step": 2276
},
{
"epoch": 5.1987814166031985,
"grad_norm": 1.3057256937026978,
"learning_rate": 2.273333985759396e-07,
"loss": 1.0655,
"step": 2277
},
{
"epoch": 5.201066260472201,
"grad_norm": 1.284069299697876,
"learning_rate": 2.2603763350086071e-07,
"loss": 1.0813,
"step": 2278
},
{
"epoch": 5.203351104341204,
"grad_norm": 1.2818840742111206,
"learning_rate": 2.2474539694752245e-07,
"loss": 1.1045,
"step": 2279
},
{
"epoch": 5.2056359482102055,
"grad_norm": 1.304898977279663,
"learning_rate": 2.234566909210975e-07,
"loss": 1.086,
"step": 2280
},
{
"epoch": 5.207920792079208,
"grad_norm": 1.3013020753860474,
"learning_rate": 2.2217151742127851e-07,
"loss": 1.0749,
"step": 2281
},
{
"epoch": 5.21020563594821,
"grad_norm": 1.2945225238800049,
"learning_rate": 2.2088987844227695e-07,
"loss": 1.1133,
"step": 2282
},
{
"epoch": 5.2124904798172125,
"grad_norm": 1.2707927227020264,
"learning_rate": 2.1961177597282112e-07,
"loss": 1.0757,
"step": 2283
},
{
"epoch": 5.214775323686215,
"grad_norm": 1.3085978031158447,
"learning_rate": 2.1833721199614992e-07,
"loss": 1.0901,
"step": 2284
},
{
"epoch": 5.217060167555217,
"grad_norm": 1.2786146402359009,
"learning_rate": 2.1706618849001188e-07,
"loss": 1.0862,
"step": 2285
},
{
"epoch": 5.2193450114242195,
"grad_norm": 1.270601511001587,
"learning_rate": 2.1579870742666347e-07,
"loss": 1.0941,
"step": 2286
},
{
"epoch": 5.221629855293221,
"grad_norm": 1.3005554676055908,
"learning_rate": 2.145347707728618e-07,
"loss": 1.0716,
"step": 2287
},
{
"epoch": 5.223914699162224,
"grad_norm": 1.3076614141464233,
"learning_rate": 2.1327438048986627e-07,
"loss": 1.0895,
"step": 2288
},
{
"epoch": 5.2261995430312265,
"grad_norm": 1.3139954805374146,
"learning_rate": 2.120175385334322e-07,
"loss": 1.0546,
"step": 2289
},
{
"epoch": 5.228484386900228,
"grad_norm": 1.2856640815734863,
"learning_rate": 2.1076424685380848e-07,
"loss": 1.1181,
"step": 2290
},
{
"epoch": 5.230769230769231,
"grad_norm": 1.2663482427597046,
"learning_rate": 2.0951450739573664e-07,
"loss": 1.0945,
"step": 2291
},
{
"epoch": 5.2330540746382335,
"grad_norm": 1.2739810943603516,
"learning_rate": 2.082683220984452e-07,
"loss": 1.0752,
"step": 2292
},
{
"epoch": 5.235338918507235,
"grad_norm": 1.2725884914398193,
"learning_rate": 2.0702569289564683e-07,
"loss": 1.0558,
"step": 2293
},
{
"epoch": 5.237623762376238,
"grad_norm": 1.3000556230545044,
"learning_rate": 2.0578662171553798e-07,
"loss": 1.0901,
"step": 2294
},
{
"epoch": 5.23990860624524,
"grad_norm": 1.307210087776184,
"learning_rate": 2.0455111048079262e-07,
"loss": 1.0859,
"step": 2295
},
{
"epoch": 5.242193450114242,
"grad_norm": 1.265657663345337,
"learning_rate": 2.033191611085622e-07,
"loss": 1.1058,
"step": 2296
},
{
"epoch": 5.244478293983245,
"grad_norm": 1.2767589092254639,
"learning_rate": 2.020907755104698e-07,
"loss": 1.1249,
"step": 2297
},
{
"epoch": 5.246763137852247,
"grad_norm": 1.278253436088562,
"learning_rate": 2.0086595559260846e-07,
"loss": 1.0785,
"step": 2298
},
{
"epoch": 5.249047981721249,
"grad_norm": 1.2918927669525146,
"learning_rate": 1.9964470325554015e-07,
"loss": 1.0703,
"step": 2299
},
{
"epoch": 5.251332825590251,
"grad_norm": 1.2632293701171875,
"learning_rate": 1.9842702039428923e-07,
"loss": 1.0665,
"step": 2300
},
{
"epoch": 5.253617669459254,
"grad_norm": 1.3661508560180664,
"learning_rate": 1.9721290889834127e-07,
"loss": 1.0423,
"step": 2301
},
{
"epoch": 5.255902513328256,
"grad_norm": 1.307930827140808,
"learning_rate": 1.96002370651642e-07,
"loss": 1.0686,
"step": 2302
},
{
"epoch": 5.258187357197258,
"grad_norm": 1.2975925207138062,
"learning_rate": 1.9479540753259003e-07,
"loss": 1.0971,
"step": 2303
},
{
"epoch": 5.260472201066261,
"grad_norm": 1.295094609260559,
"learning_rate": 1.935920214140388e-07,
"loss": 1.0575,
"step": 2304
},
{
"epoch": 5.262757044935263,
"grad_norm": 1.3307294845581055,
"learning_rate": 1.9239221416328945e-07,
"loss": 1.099,
"step": 2305
},
{
"epoch": 5.265041888804265,
"grad_norm": 1.276889681816101,
"learning_rate": 1.911959876420902e-07,
"loss": 1.0865,
"step": 2306
},
{
"epoch": 5.267326732673268,
"grad_norm": 1.2554212808609009,
"learning_rate": 1.90003343706634e-07,
"loss": 1.0758,
"step": 2307
},
{
"epoch": 5.269611576542269,
"grad_norm": 1.3033947944641113,
"learning_rate": 1.8881428420755381e-07,
"loss": 1.0934,
"step": 2308
},
{
"epoch": 5.271896420411272,
"grad_norm": 1.3277416229248047,
"learning_rate": 1.876288109899199e-07,
"loss": 1.0907,
"step": 2309
},
{
"epoch": 5.274181264280275,
"grad_norm": 1.28794264793396,
"learning_rate": 1.864469258932397e-07,
"loss": 1.1042,
"step": 2310
},
{
"epoch": 5.276466108149276,
"grad_norm": 1.3302158117294312,
"learning_rate": 1.8526863075145091e-07,
"loss": 1.1262,
"step": 2311
},
{
"epoch": 5.278750952018279,
"grad_norm": 1.2931283712387085,
"learning_rate": 1.840939273929221e-07,
"loss": 1.0817,
"step": 2312
},
{
"epoch": 5.281035795887281,
"grad_norm": 1.305420994758606,
"learning_rate": 1.8292281764044794e-07,
"loss": 1.0748,
"step": 2313
},
{
"epoch": 5.283320639756283,
"grad_norm": 1.2627601623535156,
"learning_rate": 1.8175530331124598e-07,
"loss": 1.1136,
"step": 2314
},
{
"epoch": 5.285605483625286,
"grad_norm": 1.2889796495437622,
"learning_rate": 1.8059138621695643e-07,
"loss": 1.0994,
"step": 2315
},
{
"epoch": 5.287890327494288,
"grad_norm": 1.522375226020813,
"learning_rate": 1.794310681636366e-07,
"loss": 1.1194,
"step": 2316
},
{
"epoch": 5.29017517136329,
"grad_norm": 1.283361554145813,
"learning_rate": 1.7827435095175872e-07,
"loss": 1.109,
"step": 2317
},
{
"epoch": 5.292460015232292,
"grad_norm": 1.3082143068313599,
"learning_rate": 1.77121236376209e-07,
"loss": 1.0764,
"step": 2318
},
{
"epoch": 5.294744859101295,
"grad_norm": 1.32005774974823,
"learning_rate": 1.7597172622628193e-07,
"loss": 1.0917,
"step": 2319
},
{
"epoch": 5.297029702970297,
"grad_norm": 1.2927436828613281,
"learning_rate": 1.7482582228568052e-07,
"loss": 1.0659,
"step": 2320
},
{
"epoch": 5.299314546839299,
"grad_norm": 1.27031672000885,
"learning_rate": 1.736835263325104e-07,
"loss": 1.0974,
"step": 2321
},
{
"epoch": 5.301599390708302,
"grad_norm": 1.2902473211288452,
"learning_rate": 1.725448401392793e-07,
"loss": 1.0911,
"step": 2322
},
{
"epoch": 5.303884234577304,
"grad_norm": 1.2780988216400146,
"learning_rate": 1.7140976547289438e-07,
"loss": 1.0803,
"step": 2323
},
{
"epoch": 5.306169078446306,
"grad_norm": 1.2977415323257446,
"learning_rate": 1.702783040946579e-07,
"loss": 1.0697,
"step": 2324
},
{
"epoch": 5.308453922315309,
"grad_norm": 1.3391021490097046,
"learning_rate": 1.691504577602654e-07,
"loss": 1.0812,
"step": 2325
},
{
"epoch": 5.31073876618431,
"grad_norm": 1.2561603784561157,
"learning_rate": 1.6802622821980336e-07,
"loss": 1.0968,
"step": 2326
},
{
"epoch": 5.313023610053313,
"grad_norm": 1.3253194093704224,
"learning_rate": 1.669056172177455e-07,
"loss": 1.0868,
"step": 2327
},
{
"epoch": 5.315308453922316,
"grad_norm": 1.2892109155654907,
"learning_rate": 1.6578862649295173e-07,
"loss": 1.0838,
"step": 2328
},
{
"epoch": 5.317593297791317,
"grad_norm": 1.2944484949111938,
"learning_rate": 1.646752577786631e-07,
"loss": 1.081,
"step": 2329
},
{
"epoch": 5.31987814166032,
"grad_norm": 1.2977781295776367,
"learning_rate": 1.6356551280250072e-07,
"loss": 1.0709,
"step": 2330
},
{
"epoch": 5.322162985529322,
"grad_norm": 1.251643419265747,
"learning_rate": 1.6245939328646322e-07,
"loss": 1.0717,
"step": 2331
},
{
"epoch": 5.324447829398324,
"grad_norm": 1.2781786918640137,
"learning_rate": 1.613569009469232e-07,
"loss": 1.0875,
"step": 2332
},
{
"epoch": 5.326732673267327,
"grad_norm": 1.2685924768447876,
"learning_rate": 1.6025803749462415e-07,
"loss": 1.1008,
"step": 2333
},
{
"epoch": 5.329017517136329,
"grad_norm": 1.3481427431106567,
"learning_rate": 1.5916280463468077e-07,
"loss": 1.1083,
"step": 2334
},
{
"epoch": 5.331302361005331,
"grad_norm": 1.2659457921981812,
"learning_rate": 1.5807120406657161e-07,
"loss": 1.0551,
"step": 2335
},
{
"epoch": 5.333587204874334,
"grad_norm": 1.2692278623580933,
"learning_rate": 1.5698323748414123e-07,
"loss": 1.0878,
"step": 2336
},
{
"epoch": 5.335872048743336,
"grad_norm": 1.7504945993423462,
"learning_rate": 1.5589890657559337e-07,
"loss": 1.0552,
"step": 2337
},
{
"epoch": 5.338156892612338,
"grad_norm": 1.3026984930038452,
"learning_rate": 1.5481821302349125e-07,
"loss": 1.076,
"step": 2338
},
{
"epoch": 5.34044173648134,
"grad_norm": 1.316756248474121,
"learning_rate": 1.5374115850475436e-07,
"loss": 1.1045,
"step": 2339
},
{
"epoch": 5.342726580350343,
"grad_norm": 1.3154984712600708,
"learning_rate": 1.5266774469065494e-07,
"loss": 1.0935,
"step": 2340
},
{
"epoch": 5.345011424219345,
"grad_norm": 1.276450753211975,
"learning_rate": 1.515979732468151e-07,
"loss": 1.0826,
"step": 2341
},
{
"epoch": 5.347296268088347,
"grad_norm": 1.292073369026184,
"learning_rate": 1.5053184583320697e-07,
"loss": 1.0719,
"step": 2342
},
{
"epoch": 5.34958111195735,
"grad_norm": 1.280203938484192,
"learning_rate": 1.4946936410414643e-07,
"loss": 1.0969,
"step": 2343
},
{
"epoch": 5.3518659558263515,
"grad_norm": 1.3046621084213257,
"learning_rate": 1.4841052970829385e-07,
"loss": 1.1077,
"step": 2344
},
{
"epoch": 5.354150799695354,
"grad_norm": 1.3253684043884277,
"learning_rate": 1.4735534428864857e-07,
"loss": 1.1174,
"step": 2345
},
{
"epoch": 5.356435643564357,
"grad_norm": 1.2969188690185547,
"learning_rate": 1.4630380948254864e-07,
"loss": 1.0404,
"step": 2346
},
{
"epoch": 5.3587204874333585,
"grad_norm": 1.2858332395553589,
"learning_rate": 1.4525592692166717e-07,
"loss": 1.0949,
"step": 2347
},
{
"epoch": 5.361005331302361,
"grad_norm": 1.2641419172286987,
"learning_rate": 1.442116982320105e-07,
"loss": 1.0805,
"step": 2348
},
{
"epoch": 5.363290175171363,
"grad_norm": 1.2731258869171143,
"learning_rate": 1.4317112503391433e-07,
"loss": 1.0358,
"step": 2349
},
{
"epoch": 5.3655750190403655,
"grad_norm": 1.269422173500061,
"learning_rate": 1.4213420894204326e-07,
"loss": 1.0712,
"step": 2350
},
{
"epoch": 5.367859862909368,
"grad_norm": 1.2818864583969116,
"learning_rate": 1.4110095156538633e-07,
"loss": 1.092,
"step": 2351
},
{
"epoch": 5.37014470677837,
"grad_norm": 1.2838257551193237,
"learning_rate": 1.4007135450725518e-07,
"loss": 1.0933,
"step": 2352
},
{
"epoch": 5.3724295506473725,
"grad_norm": 1.3167815208435059,
"learning_rate": 1.3904541936528266e-07,
"loss": 1.0955,
"step": 2353
},
{
"epoch": 5.374714394516375,
"grad_norm": 1.3507665395736694,
"learning_rate": 1.380231477314181e-07,
"loss": 1.0991,
"step": 2354
},
{
"epoch": 5.376999238385377,
"grad_norm": 1.2883590459823608,
"learning_rate": 1.3700454119192714e-07,
"loss": 1.1111,
"step": 2355
},
{
"epoch": 5.3792840822543795,
"grad_norm": 1.3028126955032349,
"learning_rate": 1.3598960132738813e-07,
"loss": 1.0776,
"step": 2356
},
{
"epoch": 5.381568926123381,
"grad_norm": 1.3255176544189453,
"learning_rate": 1.3497832971268943e-07,
"loss": 1.0694,
"step": 2357
},
{
"epoch": 5.383853769992384,
"grad_norm": 1.275755763053894,
"learning_rate": 1.3397072791702798e-07,
"loss": 1.0773,
"step": 2358
},
{
"epoch": 5.3861386138613865,
"grad_norm": 1.298366665840149,
"learning_rate": 1.3296679750390533e-07,
"loss": 1.0787,
"step": 2359
},
{
"epoch": 5.388423457730388,
"grad_norm": 1.302127480506897,
"learning_rate": 1.3196654003112653e-07,
"loss": 1.1149,
"step": 2360
},
{
"epoch": 5.390708301599391,
"grad_norm": 1.2876914739608765,
"learning_rate": 1.309699570507983e-07,
"loss": 1.0791,
"step": 2361
},
{
"epoch": 5.392993145468393,
"grad_norm": 1.277055263519287,
"learning_rate": 1.2997705010932394e-07,
"loss": 1.1154,
"step": 2362
},
{
"epoch": 5.395277989337395,
"grad_norm": 1.2551474571228027,
"learning_rate": 1.289878207474035e-07,
"loss": 1.0869,
"step": 2363
},
{
"epoch": 5.397562833206398,
"grad_norm": 1.2961163520812988,
"learning_rate": 1.2800227050003056e-07,
"loss": 1.1141,
"step": 2364
},
{
"epoch": 5.3998476770754,
"grad_norm": 1.2772424221038818,
"learning_rate": 1.2702040089648954e-07,
"loss": 1.0393,
"step": 2365
},
{
"epoch": 5.402132520944402,
"grad_norm": 1.3062280416488647,
"learning_rate": 1.2604221346035422e-07,
"loss": 1.0961,
"step": 2366
},
{
"epoch": 5.404417364813405,
"grad_norm": 1.3183172941207886,
"learning_rate": 1.250677097094835e-07,
"loss": 1.0627,
"step": 2367
},
{
"epoch": 5.406702208682407,
"grad_norm": 1.2865747213363647,
"learning_rate": 1.2409689115602114e-07,
"loss": 1.0516,
"step": 2368
},
{
"epoch": 5.408987052551409,
"grad_norm": 1.300234317779541,
"learning_rate": 1.2312975930639293e-07,
"loss": 1.1068,
"step": 2369
},
{
"epoch": 5.411271896420411,
"grad_norm": 1.2960008382797241,
"learning_rate": 1.2216631566130288e-07,
"loss": 1.08,
"step": 2370
},
{
"epoch": 5.413556740289414,
"grad_norm": 1.243540644645691,
"learning_rate": 1.2120656171573236e-07,
"loss": 1.0736,
"step": 2371
},
{
"epoch": 5.415841584158416,
"grad_norm": 1.3193697929382324,
"learning_rate": 1.202504989589387e-07,
"loss": 1.0834,
"step": 2372
},
{
"epoch": 5.418126428027418,
"grad_norm": 1.3288146257400513,
"learning_rate": 1.1929812887444937e-07,
"loss": 1.0567,
"step": 2373
},
{
"epoch": 5.420411271896421,
"grad_norm": 1.3119516372680664,
"learning_rate": 1.1834945294006362e-07,
"loss": 1.0836,
"step": 2374
},
{
"epoch": 5.422696115765422,
"grad_norm": 1.295262098312378,
"learning_rate": 1.1740447262784782e-07,
"loss": 1.0935,
"step": 2375
},
{
"epoch": 5.424980959634425,
"grad_norm": 1.3412234783172607,
"learning_rate": 1.1646318940413375e-07,
"loss": 1.0805,
"step": 2376
},
{
"epoch": 5.427265803503428,
"grad_norm": 1.295854926109314,
"learning_rate": 1.1552560472951669e-07,
"loss": 1.0854,
"step": 2377
},
{
"epoch": 5.429550647372429,
"grad_norm": 1.2979471683502197,
"learning_rate": 1.1459172005885255e-07,
"loss": 1.0677,
"step": 2378
},
{
"epoch": 5.431835491241432,
"grad_norm": 1.291488528251648,
"learning_rate": 1.136615368412558e-07,
"loss": 1.0763,
"step": 2379
},
{
"epoch": 5.434120335110434,
"grad_norm": 1.2876052856445312,
"learning_rate": 1.1273505652009826e-07,
"loss": 1.0806,
"step": 2380
},
{
"epoch": 5.436405178979436,
"grad_norm": 1.2822387218475342,
"learning_rate": 1.1181228053300463e-07,
"loss": 1.0812,
"step": 2381
},
{
"epoch": 5.438690022848439,
"grad_norm": 1.2904083728790283,
"learning_rate": 1.1089321031185291e-07,
"loss": 1.0875,
"step": 2382
},
{
"epoch": 5.440974866717441,
"grad_norm": 1.2913849353790283,
"learning_rate": 1.0997784728276978e-07,
"loss": 1.109,
"step": 2383
},
{
"epoch": 5.443259710586443,
"grad_norm": 1.2997492551803589,
"learning_rate": 1.0906619286612935e-07,
"loss": 1.1058,
"step": 2384
},
{
"epoch": 5.445544554455446,
"grad_norm": 1.2887829542160034,
"learning_rate": 1.0815824847655226e-07,
"loss": 1.073,
"step": 2385
},
{
"epoch": 5.447829398324448,
"grad_norm": 1.3243333101272583,
"learning_rate": 1.0725401552290154e-07,
"loss": 1.0456,
"step": 2386
},
{
"epoch": 5.45011424219345,
"grad_norm": 1.2761950492858887,
"learning_rate": 1.0635349540828038e-07,
"loss": 1.064,
"step": 2387
},
{
"epoch": 5.452399086062452,
"grad_norm": 1.2706801891326904,
"learning_rate": 1.054566895300324e-07,
"loss": 1.0803,
"step": 2388
},
{
"epoch": 5.454683929931455,
"grad_norm": 1.2992305755615234,
"learning_rate": 1.0456359927973614e-07,
"loss": 1.0788,
"step": 2389
},
{
"epoch": 5.456968773800457,
"grad_norm": 1.3025919198989868,
"learning_rate": 1.0367422604320637e-07,
"loss": 1.0695,
"step": 2390
},
{
"epoch": 5.459253617669459,
"grad_norm": 1.3010618686676025,
"learning_rate": 1.0278857120048836e-07,
"loss": 1.0801,
"step": 2391
},
{
"epoch": 5.461538461538462,
"grad_norm": 1.2742735147476196,
"learning_rate": 1.0190663612585833e-07,
"loss": 1.0734,
"step": 2392
},
{
"epoch": 5.463823305407463,
"grad_norm": 1.289499282836914,
"learning_rate": 1.0102842218782105e-07,
"loss": 1.0861,
"step": 2393
},
{
"epoch": 5.466108149276466,
"grad_norm": 1.313836932182312,
"learning_rate": 1.0015393074910618e-07,
"loss": 1.103,
"step": 2394
},
{
"epoch": 5.468392993145469,
"grad_norm": 1.3196077346801758,
"learning_rate": 9.928316316666742e-08,
"loss": 1.1171,
"step": 2395
},
{
"epoch": 5.47067783701447,
"grad_norm": 1.2959681749343872,
"learning_rate": 9.841612079168089e-08,
"loss": 1.0958,
"step": 2396
},
{
"epoch": 5.472962680883473,
"grad_norm": 1.2794181108474731,
"learning_rate": 9.755280496954123e-08,
"loss": 1.0765,
"step": 2397
},
{
"epoch": 5.475247524752476,
"grad_norm": 1.3223141431808472,
"learning_rate": 9.669321703986157e-08,
"loss": 1.0987,
"step": 2398
},
{
"epoch": 5.477532368621477,
"grad_norm": 1.325971007347107,
"learning_rate": 9.58373583364694e-08,
"loss": 1.0707,
"step": 2399
},
{
"epoch": 5.47981721249048,
"grad_norm": 1.3212562799453735,
"learning_rate": 9.498523018740601e-08,
"loss": 1.0905,
"step": 2400
},
{
"epoch": 5.482102056359482,
"grad_norm": 1.2809903621673584,
"learning_rate": 9.413683391492456e-08,
"loss": 1.0842,
"step": 2401
},
{
"epoch": 5.484386900228484,
"grad_norm": 1.2972325086593628,
"learning_rate": 9.329217083548669e-08,
"loss": 1.0978,
"step": 2402
},
{
"epoch": 5.486671744097487,
"grad_norm": 1.2997273206710815,
"learning_rate": 9.245124225976093e-08,
"loss": 1.0782,
"step": 2403
},
{
"epoch": 5.488956587966489,
"grad_norm": 1.2999190092086792,
"learning_rate": 9.161404949262209e-08,
"loss": 1.0756,
"step": 2404
},
{
"epoch": 5.491241431835491,
"grad_norm": 1.2872076034545898,
"learning_rate": 9.078059383314685e-08,
"loss": 1.0967,
"step": 2405
},
{
"epoch": 5.493526275704493,
"grad_norm": 1.298041582107544,
"learning_rate": 8.99508765746146e-08,
"loss": 1.0547,
"step": 2406
},
{
"epoch": 5.495811119573496,
"grad_norm": 1.2661224603652954,
"learning_rate": 8.912489900450182e-08,
"loss": 1.1012,
"step": 2407
},
{
"epoch": 5.498095963442498,
"grad_norm": 1.3128606081008911,
"learning_rate": 8.83026624044836e-08,
"loss": 1.0688,
"step": 2408
},
{
"epoch": 5.5003808073115,
"grad_norm": 1.2731242179870605,
"learning_rate": 8.748416805042986e-08,
"loss": 1.0853,
"step": 2409
},
{
"epoch": 5.502665651180503,
"grad_norm": 1.2822498083114624,
"learning_rate": 8.666941721240301e-08,
"loss": 1.0503,
"step": 2410
},
{
"epoch": 5.5049504950495045,
"grad_norm": 1.3088481426239014,
"learning_rate": 8.585841115465676e-08,
"loss": 1.065,
"step": 2411
},
{
"epoch": 5.507235338918507,
"grad_norm": 1.3032389879226685,
"learning_rate": 8.505115113563528e-08,
"loss": 1.0494,
"step": 2412
},
{
"epoch": 5.50952018278751,
"grad_norm": 1.3076543807983398,
"learning_rate": 8.424763840796823e-08,
"loss": 1.1008,
"step": 2413
},
{
"epoch": 5.5118050266565115,
"grad_norm": 1.4247517585754395,
"learning_rate": 8.344787421847216e-08,
"loss": 1.066,
"step": 2414
},
{
"epoch": 5.514089870525514,
"grad_norm": 1.3126006126403809,
"learning_rate": 8.26518598081455e-08,
"loss": 1.0893,
"step": 2415
},
{
"epoch": 5.516374714394516,
"grad_norm": 1.2457414865493774,
"learning_rate": 8.185959641216878e-08,
"loss": 1.0683,
"step": 2416
},
{
"epoch": 5.5186595582635185,
"grad_norm": 1.3255144357681274,
"learning_rate": 8.107108525990281e-08,
"loss": 1.113,
"step": 2417
},
{
"epoch": 5.520944402132521,
"grad_norm": 1.3008490800857544,
"learning_rate": 8.028632757488469e-08,
"loss": 1.0738,
"step": 2418
},
{
"epoch": 5.523229246001523,
"grad_norm": 1.2679911851882935,
"learning_rate": 7.950532457482785e-08,
"loss": 1.0643,
"step": 2419
},
{
"epoch": 5.5255140898705255,
"grad_norm": 1.2714024782180786,
"learning_rate": 7.872807747162009e-08,
"loss": 1.0638,
"step": 2420
},
{
"epoch": 5.527798933739528,
"grad_norm": 1.3037550449371338,
"learning_rate": 7.795458747132028e-08,
"loss": 1.1542,
"step": 2421
},
{
"epoch": 5.53008377760853,
"grad_norm": 1.2944037914276123,
"learning_rate": 7.71848557741578e-08,
"loss": 1.0635,
"step": 2422
},
{
"epoch": 5.5323686214775325,
"grad_norm": 1.2680352926254272,
"learning_rate": 7.641888357452998e-08,
"loss": 1.085,
"step": 2423
},
{
"epoch": 5.534653465346535,
"grad_norm": 1.3408043384552002,
"learning_rate": 7.565667206100108e-08,
"loss": 1.093,
"step": 2424
},
{
"epoch": 5.536938309215537,
"grad_norm": 1.340598702430725,
"learning_rate": 7.48982224162989e-08,
"loss": 1.1008,
"step": 2425
},
{
"epoch": 5.5392231530845395,
"grad_norm": 1.3032371997833252,
"learning_rate": 7.414353581731536e-08,
"loss": 1.0971,
"step": 2426
},
{
"epoch": 5.541507996953541,
"grad_norm": 1.2875107526779175,
"learning_rate": 7.339261343510207e-08,
"loss": 1.0771,
"step": 2427
},
{
"epoch": 5.543792840822544,
"grad_norm": 1.2665351629257202,
"learning_rate": 7.264545643486997e-08,
"loss": 1.0736,
"step": 2428
},
{
"epoch": 5.5460776846915465,
"grad_norm": 1.3342937231063843,
"learning_rate": 7.190206597598726e-08,
"loss": 1.0984,
"step": 2429
},
{
"epoch": 5.548362528560548,
"grad_norm": 1.290933609008789,
"learning_rate": 7.116244321197729e-08,
"loss": 1.0919,
"step": 2430
},
{
"epoch": 5.550647372429551,
"grad_norm": 1.2763770818710327,
"learning_rate": 7.042658929051816e-08,
"loss": 1.0922,
"step": 2431
},
{
"epoch": 5.552932216298553,
"grad_norm": 1.307851791381836,
"learning_rate": 6.969450535343841e-08,
"loss": 1.1177,
"step": 2432
},
{
"epoch": 5.555217060167555,
"grad_norm": 1.2933450937271118,
"learning_rate": 6.896619253671743e-08,
"loss": 1.0963,
"step": 2433
},
{
"epoch": 5.557501904036558,
"grad_norm": 1.2725058794021606,
"learning_rate": 6.824165197048316e-08,
"loss": 1.0783,
"step": 2434
},
{
"epoch": 5.55978674790556,
"grad_norm": 1.2997236251831055,
"learning_rate": 6.752088477900903e-08,
"loss": 1.0688,
"step": 2435
},
{
"epoch": 5.562071591774562,
"grad_norm": 1.2684441804885864,
"learning_rate": 6.680389208071459e-08,
"loss": 1.1119,
"step": 2436
},
{
"epoch": 5.564356435643564,
"grad_norm": 1.2946056127548218,
"learning_rate": 6.609067498816207e-08,
"loss": 1.0673,
"step": 2437
},
{
"epoch": 5.566641279512567,
"grad_norm": 1.2829228639602661,
"learning_rate": 6.538123460805428e-08,
"loss": 1.0584,
"step": 2438
},
{
"epoch": 5.568926123381569,
"grad_norm": 1.2506768703460693,
"learning_rate": 6.467557204123508e-08,
"loss": 1.1196,
"step": 2439
},
{
"epoch": 5.571210967250571,
"grad_norm": 1.3308897018432617,
"learning_rate": 6.397368838268497e-08,
"loss": 1.105,
"step": 2440
},
{
"epoch": 5.573495811119574,
"grad_norm": 1.247656226158142,
"learning_rate": 6.327558472152134e-08,
"loss": 1.0863,
"step": 2441
},
{
"epoch": 5.575780654988575,
"grad_norm": 1.3870344161987305,
"learning_rate": 6.25812621409963e-08,
"loss": 1.127,
"step": 2442
},
{
"epoch": 5.578065498857578,
"grad_norm": 1.300175428390503,
"learning_rate": 6.189072171849414e-08,
"loss": 1.1026,
"step": 2443
},
{
"epoch": 5.580350342726581,
"grad_norm": 1.2674915790557861,
"learning_rate": 6.120396452553162e-08,
"loss": 1.1207,
"step": 2444
},
{
"epoch": 5.582635186595582,
"grad_norm": 1.3113480806350708,
"learning_rate": 6.052099162775327e-08,
"loss": 1.0931,
"step": 2445
},
{
"epoch": 5.584920030464585,
"grad_norm": 1.277754783630371,
"learning_rate": 5.984180408493273e-08,
"loss": 1.0537,
"step": 2446
},
{
"epoch": 5.587204874333588,
"grad_norm": 1.3144258260726929,
"learning_rate": 5.9166402950970035e-08,
"loss": 1.066,
"step": 2447
},
{
"epoch": 5.589489718202589,
"grad_norm": 1.2873685359954834,
"learning_rate": 5.8494789273888796e-08,
"loss": 1.0881,
"step": 2448
},
{
"epoch": 5.591774562071592,
"grad_norm": 1.2995277643203735,
"learning_rate": 5.782696409583649e-08,
"loss": 1.1082,
"step": 2449
},
{
"epoch": 5.594059405940594,
"grad_norm": 1.2935672998428345,
"learning_rate": 5.716292845308169e-08,
"loss": 1.1162,
"step": 2450
},
{
"epoch": 5.596344249809596,
"grad_norm": 1.2738548517227173,
"learning_rate": 5.650268337601239e-08,
"loss": 1.074,
"step": 2451
},
{
"epoch": 5.598629093678599,
"grad_norm": 1.2765862941741943,
"learning_rate": 5.584622988913546e-08,
"loss": 1.0489,
"step": 2452
},
{
"epoch": 5.600913937547601,
"grad_norm": 1.3241945505142212,
"learning_rate": 5.519356901107359e-08,
"loss": 1.0913,
"step": 2453
},
{
"epoch": 5.603198781416603,
"grad_norm": 1.292481780052185,
"learning_rate": 5.454470175456472e-08,
"loss": 1.0842,
"step": 2454
},
{
"epoch": 5.605483625285606,
"grad_norm": 1.2790048122406006,
"learning_rate": 5.3899629126460686e-08,
"loss": 1.0938,
"step": 2455
},
{
"epoch": 5.607768469154608,
"grad_norm": 1.2987381219863892,
"learning_rate": 5.32583521277244e-08,
"loss": 1.0684,
"step": 2456
},
{
"epoch": 5.61005331302361,
"grad_norm": 1.2727071046829224,
"learning_rate": 5.2620871753429335e-08,
"loss": 1.0597,
"step": 2457
},
{
"epoch": 5.612338156892612,
"grad_norm": 1.2885806560516357,
"learning_rate": 5.1987188992758396e-08,
"loss": 1.0773,
"step": 2458
},
{
"epoch": 5.614623000761615,
"grad_norm": 1.3235818147659302,
"learning_rate": 5.135730482900059e-08,
"loss": 1.0766,
"step": 2459
},
{
"epoch": 5.616907844630617,
"grad_norm": 1.3038285970687866,
"learning_rate": 5.073122023955101e-08,
"loss": 1.0565,
"step": 2460
},
{
"epoch": 5.619192688499619,
"grad_norm": 1.3045719861984253,
"learning_rate": 5.010893619590951e-08,
"loss": 1.1003,
"step": 2461
},
{
"epoch": 5.621477532368622,
"grad_norm": 1.3066750764846802,
"learning_rate": 4.949045366367783e-08,
"loss": 1.0772,
"step": 2462
},
{
"epoch": 5.623762376237623,
"grad_norm": 1.2741373777389526,
"learning_rate": 4.8875773602559404e-08,
"loss": 1.071,
"step": 2463
},
{
"epoch": 5.626047220106626,
"grad_norm": 1.2700215578079224,
"learning_rate": 4.8264896966357386e-08,
"loss": 1.1054,
"step": 2464
},
{
"epoch": 5.628332063975629,
"grad_norm": 1.3035249710083008,
"learning_rate": 4.7657824702972144e-08,
"loss": 1.1259,
"step": 2465
},
{
"epoch": 5.63061690784463,
"grad_norm": 1.3194633722305298,
"learning_rate": 4.705455775440237e-08,
"loss": 1.0753,
"step": 2466
},
{
"epoch": 5.632901751713633,
"grad_norm": 1.31345534324646,
"learning_rate": 4.645509705674095e-08,
"loss": 1.0838,
"step": 2467
},
{
"epoch": 5.635186595582635,
"grad_norm": 1.3296149969100952,
"learning_rate": 4.585944354017435e-08,
"loss": 1.0917,
"step": 2468
},
{
"epoch": 5.637471439451637,
"grad_norm": 1.316845417022705,
"learning_rate": 4.526759812898268e-08,
"loss": 1.0999,
"step": 2469
},
{
"epoch": 5.63975628332064,
"grad_norm": 1.2568460702896118,
"learning_rate": 4.46795617415352e-08,
"loss": 1.0767,
"step": 2470
},
{
"epoch": 5.642041127189642,
"grad_norm": 1.2605866193771362,
"learning_rate": 4.4095335290292865e-08,
"loss": 1.0988,
"step": 2471
},
{
"epoch": 5.644325971058644,
"grad_norm": 1.3232251405715942,
"learning_rate": 4.3514919681802714e-08,
"loss": 1.0881,
"step": 2472
},
{
"epoch": 5.646610814927646,
"grad_norm": 1.2982079982757568,
"learning_rate": 4.293831581669933e-08,
"loss": 1.0808,
"step": 2473
},
{
"epoch": 5.648895658796649,
"grad_norm": 1.2938297986984253,
"learning_rate": 4.236552458970311e-08,
"loss": 1.0641,
"step": 2474
},
{
"epoch": 5.651180502665651,
"grad_norm": 1.2853021621704102,
"learning_rate": 4.1796546889617265e-08,
"loss": 1.0893,
"step": 2475
},
{
"epoch": 5.653465346534653,
"grad_norm": 1.3274376392364502,
"learning_rate": 4.123138359932805e-08,
"loss": 1.1138,
"step": 2476
},
{
"epoch": 5.655750190403656,
"grad_norm": 1.296823501586914,
"learning_rate": 4.0670035595803684e-08,
"loss": 1.0647,
"step": 2477
},
{
"epoch": 5.658035034272658,
"grad_norm": 1.3049858808517456,
"learning_rate": 4.011250375009018e-08,
"loss": 1.0294,
"step": 2478
},
{
"epoch": 5.66031987814166,
"grad_norm": 1.2485992908477783,
"learning_rate": 3.955878892731441e-08,
"loss": 1.101,
"step": 2479
},
{
"epoch": 5.662604722010663,
"grad_norm": 1.3331776857376099,
"learning_rate": 3.900889198667851e-08,
"loss": 1.1112,
"step": 2480
},
{
"epoch": 5.6648895658796645,
"grad_norm": 1.3384984731674194,
"learning_rate": 3.846281378146105e-08,
"loss": 1.0834,
"step": 2481
},
{
"epoch": 5.667174409748667,
"grad_norm": 1.2879012823104858,
"learning_rate": 3.792055515901533e-08,
"loss": 1.0664,
"step": 2482
},
{
"epoch": 5.66945925361767,
"grad_norm": 1.2924339771270752,
"learning_rate": 3.738211696076716e-08,
"loss": 1.0791,
"step": 2483
},
{
"epoch": 5.6717440974866715,
"grad_norm": 1.2877463102340698,
"learning_rate": 3.6847500022214597e-08,
"loss": 1.0931,
"step": 2484
},
{
"epoch": 5.674028941355674,
"grad_norm": 1.2863349914550781,
"learning_rate": 3.631670517292629e-08,
"loss": 1.0637,
"step": 2485
},
{
"epoch": 5.676313785224677,
"grad_norm": 1.2646994590759277,
"learning_rate": 3.5789733236539505e-08,
"loss": 1.0603,
"step": 2486
},
{
"epoch": 5.6785986290936785,
"grad_norm": 1.2881466150283813,
"learning_rate": 3.5266585030760416e-08,
"loss": 1.1304,
"step": 2487
},
{
"epoch": 5.680883472962681,
"grad_norm": 1.2763415575027466,
"learning_rate": 3.474726136736106e-08,
"loss": 1.0914,
"step": 2488
},
{
"epoch": 5.683168316831683,
"grad_norm": 1.2673840522766113,
"learning_rate": 3.423176305217907e-08,
"loss": 1.0831,
"step": 2489
},
{
"epoch": 5.6854531607006855,
"grad_norm": 1.2843284606933594,
"learning_rate": 3.372009088511707e-08,
"loss": 1.0711,
"step": 2490
},
{
"epoch": 5.687738004569688,
"grad_norm": 1.3351730108261108,
"learning_rate": 3.3212245660139695e-08,
"loss": 1.1146,
"step": 2491
},
{
"epoch": 5.69002284843869,
"grad_norm": 1.3806109428405762,
"learning_rate": 3.270822816527325e-08,
"loss": 1.0438,
"step": 2492
},
{
"epoch": 5.6923076923076925,
"grad_norm": 1.3130910396575928,
"learning_rate": 3.2208039182605456e-08,
"loss": 1.1089,
"step": 2493
},
{
"epoch": 5.694592536176694,
"grad_norm": 1.2872965335845947,
"learning_rate": 3.1711679488282135e-08,
"loss": 1.1009,
"step": 2494
},
{
"epoch": 5.696877380045697,
"grad_norm": 1.2836264371871948,
"learning_rate": 3.121914985250801e-08,
"loss": 1.0712,
"step": 2495
},
{
"epoch": 5.6991622239146995,
"grad_norm": 1.3084354400634766,
"learning_rate": 3.0730451039544527e-08,
"loss": 1.0736,
"step": 2496
},
{
"epoch": 5.701447067783701,
"grad_norm": 1.3039108514785767,
"learning_rate": 3.0245583807708425e-08,
"loss": 1.1067,
"step": 2497
},
{
"epoch": 5.703731911652704,
"grad_norm": 1.291584849357605,
"learning_rate": 2.976454890937175e-08,
"loss": 1.1144,
"step": 2498
},
{
"epoch": 5.706016755521706,
"grad_norm": 1.3033367395401,
"learning_rate": 2.9287347090958816e-08,
"loss": 1.0828,
"step": 2499
},
{
"epoch": 5.708301599390708,
"grad_norm": 1.2630605697631836,
"learning_rate": 2.8813979092947032e-08,
"loss": 1.0922,
"step": 2500
},
{
"epoch": 5.710586443259711,
"grad_norm": 1.3124958276748657,
"learning_rate": 2.834444564986438e-08,
"loss": 1.0406,
"step": 2501
},
{
"epoch": 5.712871287128713,
"grad_norm": 1.2817994356155396,
"learning_rate": 2.7878747490288627e-08,
"loss": 1.0972,
"step": 2502
},
{
"epoch": 5.715156130997715,
"grad_norm": 1.2732033729553223,
"learning_rate": 2.7416885336847278e-08,
"loss": 1.0886,
"step": 2503
},
{
"epoch": 5.717440974866717,
"grad_norm": 1.3142942190170288,
"learning_rate": 2.6958859906213996e-08,
"loss": 1.0905,
"step": 2504
},
{
"epoch": 5.71972581873572,
"grad_norm": 1.2737411260604858,
"learning_rate": 2.6504671909109993e-08,
"loss": 1.0746,
"step": 2505
},
{
"epoch": 5.722010662604722,
"grad_norm": 1.3381272554397583,
"learning_rate": 2.6054322050301782e-08,
"loss": 1.0316,
"step": 2506
},
{
"epoch": 5.724295506473724,
"grad_norm": 1.3035916090011597,
"learning_rate": 2.560781102859983e-08,
"loss": 1.1255,
"step": 2507
},
{
"epoch": 5.726580350342727,
"grad_norm": 1.3344190120697021,
"learning_rate": 2.516513953685823e-08,
"loss": 1.08,
"step": 2508
},
{
"epoch": 5.728865194211729,
"grad_norm": 1.2997770309448242,
"learning_rate": 2.4726308261973366e-08,
"loss": 1.0356,
"step": 2509
},
{
"epoch": 5.731150038080731,
"grad_norm": 1.3305389881134033,
"learning_rate": 2.4291317884882205e-08,
"loss": 1.0781,
"step": 2510
},
{
"epoch": 5.733434881949734,
"grad_norm": 1.3318109512329102,
"learning_rate": 2.3860169080562046e-08,
"loss": 1.0829,
"step": 2511
},
{
"epoch": 5.735719725818735,
"grad_norm": 1.2981702089309692,
"learning_rate": 2.3432862518029397e-08,
"loss": 1.0837,
"step": 2512
},
{
"epoch": 5.738004569687738,
"grad_norm": 1.3197274208068848,
"learning_rate": 2.3009398860338315e-08,
"loss": 1.1062,
"step": 2513
},
{
"epoch": 5.740289413556741,
"grad_norm": 1.2988827228546143,
"learning_rate": 2.2589778764580128e-08,
"loss": 1.0966,
"step": 2514
},
{
"epoch": 5.742574257425742,
"grad_norm": 1.2950987815856934,
"learning_rate": 2.217400288188204e-08,
"loss": 1.0548,
"step": 2515
},
{
"epoch": 5.744859101294745,
"grad_norm": 1.3110990524291992,
"learning_rate": 2.176207185740603e-08,
"loss": 1.0633,
"step": 2516
},
{
"epoch": 5.747143945163748,
"grad_norm": 1.2705743312835693,
"learning_rate": 2.135398633034802e-08,
"loss": 1.0739,
"step": 2517
},
{
"epoch": 5.749428789032749,
"grad_norm": 1.2896722555160522,
"learning_rate": 2.094974693393731e-08,
"loss": 1.1057,
"step": 2518
},
{
"epoch": 5.751713632901752,
"grad_norm": 1.2686283588409424,
"learning_rate": 2.054935429543409e-08,
"loss": 1.0857,
"step": 2519
},
{
"epoch": 5.753998476770754,
"grad_norm": 1.2696059942245483,
"learning_rate": 2.0152809036130538e-08,
"loss": 1.0561,
"step": 2520
},
{
"epoch": 5.756283320639756,
"grad_norm": 1.2757072448730469,
"learning_rate": 1.9760111771348345e-08,
"loss": 1.0842,
"step": 2521
},
{
"epoch": 5.758568164508759,
"grad_norm": 1.275511384010315,
"learning_rate": 1.937126311043813e-08,
"loss": 1.1007,
"step": 2522
},
{
"epoch": 5.760853008377761,
"grad_norm": 1.3018161058425903,
"learning_rate": 1.8986263656779193e-08,
"loss": 1.0786,
"step": 2523
},
{
"epoch": 5.763137852246763,
"grad_norm": 1.3318055868148804,
"learning_rate": 1.8605114007777258e-08,
"loss": 1.0835,
"step": 2524
},
{
"epoch": 5.765422696115765,
"grad_norm": 1.3084681034088135,
"learning_rate": 1.822781475486507e-08,
"loss": 1.0443,
"step": 2525
},
{
"epoch": 5.767707539984768,
"grad_norm": 1.2735244035720825,
"learning_rate": 1.7854366483499863e-08,
"loss": 1.1164,
"step": 2526
},
{
"epoch": 5.76999238385377,
"grad_norm": 1.2862602472305298,
"learning_rate": 1.7484769773163657e-08,
"loss": 1.0733,
"step": 2527
},
{
"epoch": 5.772277227722772,
"grad_norm": 1.3294838666915894,
"learning_rate": 1.7119025197362416e-08,
"loss": 1.0981,
"step": 2528
},
{
"epoch": 5.774562071591775,
"grad_norm": 1.3161261081695557,
"learning_rate": 1.6757133323623832e-08,
"loss": 1.1003,
"step": 2529
},
{
"epoch": 5.776846915460776,
"grad_norm": 1.2910784482955933,
"learning_rate": 1.6399094713498154e-08,
"loss": 1.0992,
"step": 2530
},
{
"epoch": 5.779131759329779,
"grad_norm": 1.3026399612426758,
"learning_rate": 1.6044909922555973e-08,
"loss": 1.095,
"step": 2531
},
{
"epoch": 5.781416603198782,
"grad_norm": 1.3220399618148804,
"learning_rate": 1.569457950038794e-08,
"loss": 1.1457,
"step": 2532
},
{
"epoch": 5.783701447067783,
"grad_norm": 1.3009707927703857,
"learning_rate": 1.5348103990604214e-08,
"loss": 1.0756,
"step": 2533
},
{
"epoch": 5.785986290936786,
"grad_norm": 1.2853416204452515,
"learning_rate": 1.5005483930833066e-08,
"loss": 1.0826,
"step": 2534
},
{
"epoch": 5.788271134805788,
"grad_norm": 1.2637382745742798,
"learning_rate": 1.4666719852719779e-08,
"loss": 1.0563,
"step": 2535
},
{
"epoch": 5.79055597867479,
"grad_norm": 1.267993450164795,
"learning_rate": 1.4331812281927204e-08,
"loss": 1.0781,
"step": 2536
},
{
"epoch": 5.792840822543793,
"grad_norm": 1.2659962177276611,
"learning_rate": 1.4000761738133528e-08,
"loss": 1.0997,
"step": 2537
},
{
"epoch": 5.795125666412795,
"grad_norm": 1.311763882637024,
"learning_rate": 1.367356873503145e-08,
"loss": 1.0806,
"step": 2538
},
{
"epoch": 5.797410510281797,
"grad_norm": 1.2864328622817993,
"learning_rate": 1.3350233780329014e-08,
"loss": 1.0602,
"step": 2539
},
{
"epoch": 5.7996953541508,
"grad_norm": 1.2898355722427368,
"learning_rate": 1.303075737574655e-08,
"loss": 1.0884,
"step": 2540
},
{
"epoch": 5.801980198019802,
"grad_norm": 1.2957934141159058,
"learning_rate": 1.2715140017018346e-08,
"loss": 1.0537,
"step": 2541
},
{
"epoch": 5.804265041888804,
"grad_norm": 1.2888224124908447,
"learning_rate": 1.2403382193889036e-08,
"loss": 1.0742,
"step": 2542
},
{
"epoch": 5.806549885757806,
"grad_norm": 1.2837013006210327,
"learning_rate": 1.2095484390115541e-08,
"loss": 1.0882,
"step": 2543
},
{
"epoch": 5.808834729626809,
"grad_norm": 1.3324216604232788,
"learning_rate": 1.1791447083465136e-08,
"loss": 1.1201,
"step": 2544
},
{
"epoch": 5.811119573495811,
"grad_norm": 1.2848390340805054,
"learning_rate": 1.1491270745714044e-08,
"loss": 1.0827,
"step": 2545
},
{
"epoch": 5.813404417364813,
"grad_norm": 1.2867978811264038,
"learning_rate": 1.1194955842647736e-08,
"loss": 1.0474,
"step": 2546
},
{
"epoch": 5.815689261233816,
"grad_norm": 1.2755886316299438,
"learning_rate": 1.090250283406008e-08,
"loss": 1.083,
"step": 2547
},
{
"epoch": 5.817974105102818,
"grad_norm": 1.3513338565826416,
"learning_rate": 1.0613912173752239e-08,
"loss": 1.1062,
"step": 2548
},
{
"epoch": 5.82025894897182,
"grad_norm": 1.2771309614181519,
"learning_rate": 1.0329184309532114e-08,
"loss": 1.1171,
"step": 2549
},
{
"epoch": 5.822543792840823,
"grad_norm": 1.28047776222229,
"learning_rate": 1.0048319683213792e-08,
"loss": 1.0738,
"step": 2550
},
{
"epoch": 5.8248286367098245,
"grad_norm": 1.3358758687973022,
"learning_rate": 9.771318730616708e-09,
"loss": 1.1125,
"step": 2551
},
{
"epoch": 5.827113480578827,
"grad_norm": 1.2923208475112915,
"learning_rate": 9.498181881564816e-09,
"loss": 1.0802,
"step": 2552
},
{
"epoch": 5.82939832444783,
"grad_norm": 1.274404525756836,
"learning_rate": 9.22890955988659e-09,
"loss": 1.0557,
"step": 2553
},
{
"epoch": 5.8316831683168315,
"grad_norm": 1.318577527999878,
"learning_rate": 8.963502183413353e-09,
"loss": 1.0755,
"step": 2554
},
{
"epoch": 5.833968012185834,
"grad_norm": 1.2825225591659546,
"learning_rate": 8.701960163979283e-09,
"loss": 1.1189,
"step": 2555
},
{
"epoch": 5.836252856054836,
"grad_norm": 1.2995526790618896,
"learning_rate": 8.444283907421136e-09,
"loss": 1.1139,
"step": 2556
},
{
"epoch": 5.8385376999238385,
"grad_norm": 1.3120074272155762,
"learning_rate": 8.190473813576571e-09,
"loss": 1.093,
"step": 2557
},
{
"epoch": 5.840822543792841,
"grad_norm": 1.2919102907180786,
"learning_rate": 7.940530276284163e-09,
"loss": 1.0544,
"step": 2558
},
{
"epoch": 5.843107387661843,
"grad_norm": 1.2919046878814697,
"learning_rate": 7.694453683383396e-09,
"loss": 1.0666,
"step": 2559
},
{
"epoch": 5.8453922315308455,
"grad_norm": 1.2774494886398315,
"learning_rate": 7.452244416712162e-09,
"loss": 1.0835,
"step": 2560
},
{
"epoch": 5.847677075399847,
"grad_norm": 1.2975057363510132,
"learning_rate": 7.2139028521087114e-09,
"loss": 1.0652,
"step": 2561
},
{
"epoch": 5.84996191926885,
"grad_norm": 1.3106671571731567,
"learning_rate": 6.979429359408874e-09,
"loss": 1.0754,
"step": 2562
},
{
"epoch": 5.8522467631378525,
"grad_norm": 1.2943204641342163,
"learning_rate": 6.748824302446611e-09,
"loss": 1.0634,
"step": 2563
},
{
"epoch": 5.854531607006854,
"grad_norm": 1.2920920848846436,
"learning_rate": 6.522088039053187e-09,
"loss": 1.1051,
"step": 2564
},
{
"epoch": 5.856816450875857,
"grad_norm": 1.2778362035751343,
"learning_rate": 6.2992209210571695e-09,
"loss": 1.0863,
"step": 2565
},
{
"epoch": 5.859101294744859,
"grad_norm": 1.2808789014816284,
"learning_rate": 6.0802232942822056e-09,
"loss": 1.0971,
"step": 2566
},
{
"epoch": 5.861386138613861,
"grad_norm": 1.2962878942489624,
"learning_rate": 5.86509549854869e-09,
"loss": 1.1123,
"step": 2567
},
{
"epoch": 5.863670982482864,
"grad_norm": 1.2766149044036865,
"learning_rate": 5.653837867671819e-09,
"loss": 1.0895,
"step": 2568
},
{
"epoch": 5.865955826351866,
"grad_norm": 1.3137177228927612,
"learning_rate": 5.4464507294613165e-09,
"loss": 1.0533,
"step": 2569
},
{
"epoch": 5.868240670220868,
"grad_norm": 1.3188068866729736,
"learning_rate": 5.242934405720879e-09,
"loss": 1.0744,
"step": 2570
},
{
"epoch": 5.870525514089871,
"grad_norm": 1.277276635169983,
"learning_rate": 5.0432892122484476e-09,
"loss": 1.0666,
"step": 2571
},
{
"epoch": 5.872810357958873,
"grad_norm": 1.2813732624053955,
"learning_rate": 4.847515458834273e-09,
"loss": 1.0539,
"step": 2572
},
{
"epoch": 5.875095201827875,
"grad_norm": 1.2894470691680908,
"learning_rate": 4.655613449262298e-09,
"loss": 1.0949,
"step": 2573
},
{
"epoch": 5.877380045696877,
"grad_norm": 1.2572760581970215,
"learning_rate": 4.46758348130738e-09,
"loss": 1.0463,
"step": 2574
},
{
"epoch": 5.87966488956588,
"grad_norm": 1.3162351846694946,
"learning_rate": 4.283425846737521e-09,
"loss": 1.0896,
"step": 2575
},
{
"epoch": 5.881949733434882,
"grad_norm": 1.3233922719955444,
"learning_rate": 4.1031408313108035e-09,
"loss": 1.0681,
"step": 2576
},
{
"epoch": 5.884234577303884,
"grad_norm": 1.281401515007019,
"learning_rate": 3.926728714776784e-09,
"loss": 1.084,
"step": 2577
},
{
"epoch": 5.886519421172887,
"grad_norm": 1.2803399562835693,
"learning_rate": 3.754189770875383e-09,
"loss": 1.0735,
"step": 2578
},
{
"epoch": 5.888804265041889,
"grad_norm": 1.2727675437927246,
"learning_rate": 3.5855242673363267e-09,
"loss": 1.1046,
"step": 2579
},
{
"epoch": 5.891089108910891,
"grad_norm": 1.277085542678833,
"learning_rate": 3.420732465878596e-09,
"loss": 1.0548,
"step": 2580
},
{
"epoch": 5.893373952779894,
"grad_norm": 1.3056312799453735,
"learning_rate": 3.2598146222109773e-09,
"loss": 1.0506,
"step": 2581
},
{
"epoch": 5.895658796648895,
"grad_norm": 1.2820820808410645,
"learning_rate": 3.102770986030679e-09,
"loss": 1.0938,
"step": 2582
},
{
"epoch": 5.897943640517898,
"grad_norm": 1.3365559577941895,
"learning_rate": 2.9496018010233275e-09,
"loss": 1.1194,
"step": 2583
},
{
"epoch": 5.900228484386901,
"grad_norm": 1.2692946195602417,
"learning_rate": 2.8003073048621376e-09,
"loss": 1.0947,
"step": 2584
},
{
"epoch": 5.902513328255902,
"grad_norm": 1.3099660873413086,
"learning_rate": 2.6548877292090215e-09,
"loss": 1.0575,
"step": 2585
},
{
"epoch": 5.904798172124905,
"grad_norm": 1.2974201440811157,
"learning_rate": 2.5133432997118125e-09,
"loss": 1.0671,
"step": 2586
},
{
"epoch": 5.907083015993907,
"grad_norm": 1.2797954082489014,
"learning_rate": 2.3756742360062092e-09,
"loss": 1.0683,
"step": 2587
},
{
"epoch": 5.909367859862909,
"grad_norm": 1.284724473953247,
"learning_rate": 2.241880751714387e-09,
"loss": 1.0578,
"step": 2588
},
{
"epoch": 5.911652703731912,
"grad_norm": 1.2749953269958496,
"learning_rate": 2.1119630544438884e-09,
"loss": 1.0864,
"step": 2589
},
{
"epoch": 5.913937547600914,
"grad_norm": 1.2969084978103638,
"learning_rate": 1.9859213457892877e-09,
"loss": 1.1049,
"step": 2590
},
{
"epoch": 5.916222391469916,
"grad_norm": 1.273667573928833,
"learning_rate": 1.863755821330249e-09,
"loss": 1.072,
"step": 2591
},
{
"epoch": 5.918507235338918,
"grad_norm": 1.2989200353622437,
"learning_rate": 1.7454666706318014e-09,
"loss": 1.0686,
"step": 2592
},
{
"epoch": 5.920792079207921,
"grad_norm": 1.3151251077651978,
"learning_rate": 1.6310540772437877e-09,
"loss": 1.0944,
"step": 2593
},
{
"epoch": 5.923076923076923,
"grad_norm": 1.2775644063949585,
"learning_rate": 1.520518218701139e-09,
"loss": 1.0753,
"step": 2594
},
{
"epoch": 5.925361766945925,
"grad_norm": 1.2895221710205078,
"learning_rate": 1.4138592665230433e-09,
"loss": 1.0934,
"step": 2595
},
{
"epoch": 5.927646610814928,
"grad_norm": 1.3031322956085205,
"learning_rate": 1.3110773862126669e-09,
"loss": 1.0793,
"step": 2596
},
{
"epoch": 5.9299314546839295,
"grad_norm": 1.292992115020752,
"learning_rate": 1.2121727372574332e-09,
"loss": 1.0425,
"step": 2597
},
{
"epoch": 5.932216298552932,
"grad_norm": 1.2997634410858154,
"learning_rate": 1.117145473128467e-09,
"loss": 1.0975,
"step": 2598
},
{
"epoch": 5.934501142421935,
"grad_norm": 1.313364863395691,
"learning_rate": 1.0259957412800393e-09,
"loss": 1.064,
"step": 2599
},
{
"epoch": 5.9367859862909365,
"grad_norm": 1.288731336593628,
"learning_rate": 9.387236831495672e-10,
"loss": 1.0707,
"step": 2600
},
{
"epoch": 5.939070830159939,
"grad_norm": 1.3007681369781494,
"learning_rate": 8.553294341578921e-10,
"loss": 1.0797,
"step": 2601
},
{
"epoch": 5.941355674028942,
"grad_norm": 1.2979000806808472,
"learning_rate": 7.75813123708169e-10,
"loss": 1.0696,
"step": 2602
},
{
"epoch": 5.9436405178979435,
"grad_norm": 1.2918391227722168,
"learning_rate": 7.001748751866988e-10,
"loss": 1.1133,
"step": 2603
},
{
"epoch": 5.945925361766946,
"grad_norm": 1.3077112436294556,
"learning_rate": 6.284148059615413e-10,
"loss": 1.0668,
"step": 2604
},
{
"epoch": 5.948210205635948,
"grad_norm": 1.2914233207702637,
"learning_rate": 5.605330273836251e-10,
"loss": 1.0955,
"step": 2605
},
{
"epoch": 5.9504950495049505,
"grad_norm": 1.2764394283294678,
"learning_rate": 4.96529644785082e-10,
"loss": 1.0705,
"step": 2606
},
{
"epoch": 5.952779893373953,
"grad_norm": 1.3059388399124146,
"learning_rate": 4.364047574803576e-10,
"loss": 1.0532,
"step": 2607
},
{
"epoch": 5.955064737242955,
"grad_norm": 1.3395713567733765,
"learning_rate": 3.801584587659335e-10,
"loss": 1.0662,
"step": 2608
},
{
"epoch": 5.9573495811119574,
"grad_norm": 1.2883552312850952,
"learning_rate": 3.277908359194948e-10,
"loss": 1.0876,
"step": 2609
},
{
"epoch": 5.95963442498096,
"grad_norm": 1.2938013076782227,
"learning_rate": 2.7930197020020753e-10,
"loss": 1.0563,
"step": 2610
},
{
"epoch": 5.961919268849962,
"grad_norm": 1.2887256145477295,
"learning_rate": 2.3469193684844125e-10,
"loss": 1.1038,
"step": 2611
},
{
"epoch": 5.9642041127189644,
"grad_norm": 1.2994153499603271,
"learning_rate": 1.9396080508576887e-10,
"loss": 1.0705,
"step": 2612
},
{
"epoch": 5.966488956587966,
"grad_norm": 1.2864863872528076,
"learning_rate": 1.5710863811524424e-10,
"loss": 1.0907,
"step": 2613
},
{
"epoch": 5.968773800456969,
"grad_norm": 1.3223236799240112,
"learning_rate": 1.2413549312029204e-10,
"loss": 1.0387,
"step": 2614
},
{
"epoch": 5.971058644325971,
"grad_norm": 1.2479461431503296,
"learning_rate": 9.504142126581794e-11,
"loss": 1.0723,
"step": 2615
},
{
"epoch": 5.973343488194973,
"grad_norm": 1.2943627834320068,
"learning_rate": 6.982646769709833e-11,
"loss": 1.1221,
"step": 2616
},
{
"epoch": 5.975628332063976,
"grad_norm": 1.3019261360168457,
"learning_rate": 4.8490671540613045e-11,
"loss": 1.1253,
"step": 2617
},
{
"epoch": 5.9779131759329776,
"grad_norm": 1.3006083965301514,
"learning_rate": 3.103406590265756e-11,
"loss": 1.1149,
"step": 2618
},
{
"epoch": 5.98019801980198,
"grad_norm": 1.2991771697998047,
"learning_rate": 1.745667787128591e-11,
"loss": 1.0608,
"step": 2619
},
{
"epoch": 5.982482863670983,
"grad_norm": 1.355244517326355,
"learning_rate": 7.758528514645313e-12,
"loss": 1.0679,
"step": 2620
},
{
"epoch": 5.9847677075399845,
"grad_norm": 1.272252082824707,
"learning_rate": 1.9396328809762054e-12,
"loss": 1.0655,
"step": 2621
},
{
"epoch": 5.987052551408987,
"grad_norm": 1.2879427671432495,
"learning_rate": 0.0,
"loss": 1.1149,
"step": 2622
}
],
"logging_steps": 1,
"max_steps": 2622,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 437,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.618637151628165e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}