csm_1500checkpoint_lora_16bit / trainer_state.json
crazyup37's picture
Upload folder using huggingface_hub
78071e5 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"eval_steps": 500,
"global_step": 1500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.013377926421404682,
"grad_norm": 0.5197079181671143,
"learning_rate": 0.0,
"loss": 4.2636,
"step": 1
},
{
"epoch": 0.026755852842809364,
"grad_norm": 0.5626901984214783,
"learning_rate": 4e-05,
"loss": 4.3971,
"step": 2
},
{
"epoch": 0.04013377926421405,
"grad_norm": 0.5167903304100037,
"learning_rate": 8e-05,
"loss": 4.3249,
"step": 3
},
{
"epoch": 0.05351170568561873,
"grad_norm": 0.4764951169490814,
"learning_rate": 0.00012,
"loss": 4.2031,
"step": 4
},
{
"epoch": 0.06688963210702341,
"grad_norm": 0.45488491654396057,
"learning_rate": 0.00016,
"loss": 4.3914,
"step": 5
},
{
"epoch": 0.0802675585284281,
"grad_norm": 0.568274736404419,
"learning_rate": 0.0002,
"loss": 4.2346,
"step": 6
},
{
"epoch": 0.09364548494983277,
"grad_norm": 0.5974003076553345,
"learning_rate": 0.0001999555061179088,
"loss": 4.131,
"step": 7
},
{
"epoch": 0.10702341137123746,
"grad_norm": 0.6204471588134766,
"learning_rate": 0.00019991101223581757,
"loss": 4.2256,
"step": 8
},
{
"epoch": 0.12040133779264214,
"grad_norm": 0.7143808603286743,
"learning_rate": 0.00019986651835372636,
"loss": 3.8449,
"step": 9
},
{
"epoch": 0.13377926421404682,
"grad_norm": 0.7799420356750488,
"learning_rate": 0.00019982202447163517,
"loss": 4.4301,
"step": 10
},
{
"epoch": 0.14715719063545152,
"grad_norm": 0.8880407214164734,
"learning_rate": 0.00019977753058954395,
"loss": 4.2266,
"step": 11
},
{
"epoch": 0.1605351170568562,
"grad_norm": 0.7776209712028503,
"learning_rate": 0.00019973303670745273,
"loss": 4.3208,
"step": 12
},
{
"epoch": 0.17391304347826086,
"grad_norm": 0.9125858545303345,
"learning_rate": 0.0001996885428253615,
"loss": 4.4363,
"step": 13
},
{
"epoch": 0.18729096989966554,
"grad_norm": 0.9000256657600403,
"learning_rate": 0.00019964404894327032,
"loss": 4.2917,
"step": 14
},
{
"epoch": 0.20066889632107024,
"grad_norm": 0.9995108246803284,
"learning_rate": 0.00019959955506117908,
"loss": 4.1784,
"step": 15
},
{
"epoch": 0.2140468227424749,
"grad_norm": 0.9209024310112,
"learning_rate": 0.0001995550611790879,
"loss": 4.7852,
"step": 16
},
{
"epoch": 0.22742474916387959,
"grad_norm": 0.9421981573104858,
"learning_rate": 0.00019951056729699667,
"loss": 4.8501,
"step": 17
},
{
"epoch": 0.2408026755852843,
"grad_norm": 0.9213201403617859,
"learning_rate": 0.00019946607341490545,
"loss": 4.7923,
"step": 18
},
{
"epoch": 0.25418060200668896,
"grad_norm": 0.9378194212913513,
"learning_rate": 0.00019942157953281423,
"loss": 4.9593,
"step": 19
},
{
"epoch": 0.26755852842809363,
"grad_norm": 1.0096492767333984,
"learning_rate": 0.00019937708565072304,
"loss": 4.7099,
"step": 20
},
{
"epoch": 0.2809364548494983,
"grad_norm": 0.8903587460517883,
"learning_rate": 0.00019933259176863183,
"loss": 4.3746,
"step": 21
},
{
"epoch": 0.29431438127090304,
"grad_norm": 0.7808490991592407,
"learning_rate": 0.0001992880978865406,
"loss": 4.5873,
"step": 22
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.8145670294761658,
"learning_rate": 0.0001992436040044494,
"loss": 4.7924,
"step": 23
},
{
"epoch": 0.3210702341137124,
"grad_norm": 0.7945849299430847,
"learning_rate": 0.0001991991101223582,
"loss": 4.8881,
"step": 24
},
{
"epoch": 0.33444816053511706,
"grad_norm": 0.7871395349502563,
"learning_rate": 0.00019915461624026696,
"loss": 4.6922,
"step": 25
},
{
"epoch": 0.34782608695652173,
"grad_norm": 0.9111238718032837,
"learning_rate": 0.00019911012235817577,
"loss": 4.9982,
"step": 26
},
{
"epoch": 0.3612040133779264,
"grad_norm": 0.7121369242668152,
"learning_rate": 0.00019906562847608455,
"loss": 4.4756,
"step": 27
},
{
"epoch": 0.3745819397993311,
"grad_norm": 0.7118422389030457,
"learning_rate": 0.00019902113459399333,
"loss": 5.1389,
"step": 28
},
{
"epoch": 0.3879598662207358,
"grad_norm": 0.7100292444229126,
"learning_rate": 0.0001989766407119021,
"loss": 4.7691,
"step": 29
},
{
"epoch": 0.4013377926421405,
"grad_norm": 0.708591639995575,
"learning_rate": 0.00019893214682981092,
"loss": 4.8721,
"step": 30
},
{
"epoch": 0.41471571906354515,
"grad_norm": 0.6711616516113281,
"learning_rate": 0.0001988876529477197,
"loss": 4.9152,
"step": 31
},
{
"epoch": 0.4280936454849498,
"grad_norm": 0.7158232927322388,
"learning_rate": 0.0001988431590656285,
"loss": 4.828,
"step": 32
},
{
"epoch": 0.4414715719063545,
"grad_norm": 0.6246087551116943,
"learning_rate": 0.00019879866518353727,
"loss": 4.8452,
"step": 33
},
{
"epoch": 0.45484949832775917,
"grad_norm": 0.6088873744010925,
"learning_rate": 0.00019875417130144608,
"loss": 4.9702,
"step": 34
},
{
"epoch": 0.4682274247491639,
"grad_norm": 0.5798126459121704,
"learning_rate": 0.00019870967741935483,
"loss": 4.9838,
"step": 35
},
{
"epoch": 0.4816053511705686,
"grad_norm": 0.6268919706344604,
"learning_rate": 0.00019866518353726364,
"loss": 4.7636,
"step": 36
},
{
"epoch": 0.49498327759197325,
"grad_norm": 0.5649904012680054,
"learning_rate": 0.00019862068965517243,
"loss": 4.506,
"step": 37
},
{
"epoch": 0.5083612040133779,
"grad_norm": 0.5947792530059814,
"learning_rate": 0.0001985761957730812,
"loss": 4.8057,
"step": 38
},
{
"epoch": 0.5217391304347826,
"grad_norm": 0.6204257011413574,
"learning_rate": 0.00019853170189099,
"loss": 5.0511,
"step": 39
},
{
"epoch": 0.5351170568561873,
"grad_norm": 0.5972265601158142,
"learning_rate": 0.0001984872080088988,
"loss": 4.924,
"step": 40
},
{
"epoch": 0.5484949832775919,
"grad_norm": 0.6117077469825745,
"learning_rate": 0.00019844271412680758,
"loss": 4.8729,
"step": 41
},
{
"epoch": 0.5618729096989966,
"grad_norm": 0.5085508823394775,
"learning_rate": 0.00019839822024471637,
"loss": 4.3616,
"step": 42
},
{
"epoch": 0.5752508361204013,
"grad_norm": 0.550647497177124,
"learning_rate": 0.00019835372636262515,
"loss": 5.2512,
"step": 43
},
{
"epoch": 0.5886287625418061,
"grad_norm": 0.48329588770866394,
"learning_rate": 0.00019830923248053396,
"loss": 4.9501,
"step": 44
},
{
"epoch": 0.6020066889632107,
"grad_norm": 0.6313246488571167,
"learning_rate": 0.0001982647385984427,
"loss": 4.5767,
"step": 45
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.5111928582191467,
"learning_rate": 0.00019822024471635152,
"loss": 4.5586,
"step": 46
},
{
"epoch": 0.6287625418060201,
"grad_norm": 0.5264492630958557,
"learning_rate": 0.0001981757508342603,
"loss": 4.7033,
"step": 47
},
{
"epoch": 0.6421404682274248,
"grad_norm": 0.5058289170265198,
"learning_rate": 0.0001981312569521691,
"loss": 4.8396,
"step": 48
},
{
"epoch": 0.6555183946488294,
"grad_norm": 0.5688439607620239,
"learning_rate": 0.00019808676307007787,
"loss": 5.1887,
"step": 49
},
{
"epoch": 0.6688963210702341,
"grad_norm": 0.5488842129707336,
"learning_rate": 0.00019804226918798665,
"loss": 4.6075,
"step": 50
},
{
"epoch": 0.6822742474916388,
"grad_norm": 0.5358632206916809,
"learning_rate": 0.00019799777530589546,
"loss": 5.0205,
"step": 51
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.47869494557380676,
"learning_rate": 0.00019795328142380422,
"loss": 4.8495,
"step": 52
},
{
"epoch": 0.7090301003344481,
"grad_norm": 0.49378660321235657,
"learning_rate": 0.00019790878754171303,
"loss": 4.6563,
"step": 53
},
{
"epoch": 0.7224080267558528,
"grad_norm": 0.5167868733406067,
"learning_rate": 0.0001978642936596218,
"loss": 5.2558,
"step": 54
},
{
"epoch": 0.7357859531772575,
"grad_norm": 0.5230040550231934,
"learning_rate": 0.0001978197997775306,
"loss": 4.7769,
"step": 55
},
{
"epoch": 0.7491638795986622,
"grad_norm": 0.4822310507297516,
"learning_rate": 0.00019777530589543937,
"loss": 4.9282,
"step": 56
},
{
"epoch": 0.7625418060200669,
"grad_norm": 0.500045895576477,
"learning_rate": 0.00019773081201334818,
"loss": 5.0399,
"step": 57
},
{
"epoch": 0.7759197324414716,
"grad_norm": 0.4740642309188843,
"learning_rate": 0.00019768631813125696,
"loss": 4.8041,
"step": 58
},
{
"epoch": 0.7892976588628763,
"grad_norm": 0.45918184518814087,
"learning_rate": 0.00019764182424916575,
"loss": 4.6304,
"step": 59
},
{
"epoch": 0.802675585284281,
"grad_norm": 0.53122878074646,
"learning_rate": 0.00019759733036707453,
"loss": 4.8377,
"step": 60
},
{
"epoch": 0.8160535117056856,
"grad_norm": 0.4925791919231415,
"learning_rate": 0.00019755283648498334,
"loss": 5.0919,
"step": 61
},
{
"epoch": 0.8294314381270903,
"grad_norm": 0.4777262806892395,
"learning_rate": 0.0001975083426028921,
"loss": 4.8379,
"step": 62
},
{
"epoch": 0.842809364548495,
"grad_norm": 0.49119675159454346,
"learning_rate": 0.0001974638487208009,
"loss": 5.0819,
"step": 63
},
{
"epoch": 0.8561872909698997,
"grad_norm": 0.4732685089111328,
"learning_rate": 0.00019741935483870969,
"loss": 4.8948,
"step": 64
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.46269145607948303,
"learning_rate": 0.00019737486095661847,
"loss": 4.824,
"step": 65
},
{
"epoch": 0.882943143812709,
"grad_norm": 0.49532708525657654,
"learning_rate": 0.00019733036707452725,
"loss": 4.8986,
"step": 66
},
{
"epoch": 0.8963210702341137,
"grad_norm": 0.5253002643585205,
"learning_rate": 0.00019728587319243606,
"loss": 4.9073,
"step": 67
},
{
"epoch": 0.9096989966555183,
"grad_norm": 0.5069419145584106,
"learning_rate": 0.00019724137931034484,
"loss": 4.8962,
"step": 68
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.5038817524909973,
"learning_rate": 0.00019719688542825363,
"loss": 4.8711,
"step": 69
},
{
"epoch": 0.9364548494983278,
"grad_norm": 0.4987100064754486,
"learning_rate": 0.0001971523915461624,
"loss": 4.8816,
"step": 70
},
{
"epoch": 0.9498327759197325,
"grad_norm": 0.47370976209640503,
"learning_rate": 0.00019710789766407122,
"loss": 4.9675,
"step": 71
},
{
"epoch": 0.9632107023411371,
"grad_norm": 0.5081727504730225,
"learning_rate": 0.00019706340378197997,
"loss": 4.2768,
"step": 72
},
{
"epoch": 0.9765886287625418,
"grad_norm": 0.45571258664131165,
"learning_rate": 0.00019701890989988878,
"loss": 4.6182,
"step": 73
},
{
"epoch": 0.9899665551839465,
"grad_norm": 0.5216127634048462,
"learning_rate": 0.00019697441601779756,
"loss": 4.7126,
"step": 74
},
{
"epoch": 1.0,
"grad_norm": 0.5393329858779907,
"learning_rate": 0.00019692992213570635,
"loss": 4.4919,
"step": 75
},
{
"epoch": 1.0133779264214047,
"grad_norm": 0.4506986737251282,
"learning_rate": 0.00019688542825361513,
"loss": 4.5089,
"step": 76
},
{
"epoch": 1.0267558528428093,
"grad_norm": 0.4328899085521698,
"learning_rate": 0.00019684093437152394,
"loss": 4.7518,
"step": 77
},
{
"epoch": 1.040133779264214,
"grad_norm": 0.4397362470626831,
"learning_rate": 0.00019679644048943272,
"loss": 4.5069,
"step": 78
},
{
"epoch": 1.0535117056856187,
"grad_norm": 0.4604664146900177,
"learning_rate": 0.0001967519466073415,
"loss": 4.7054,
"step": 79
},
{
"epoch": 1.0668896321070234,
"grad_norm": 0.4398234784603119,
"learning_rate": 0.00019670745272525029,
"loss": 4.2743,
"step": 80
},
{
"epoch": 1.080267558528428,
"grad_norm": 0.4570735692977905,
"learning_rate": 0.0001966629588431591,
"loss": 4.8012,
"step": 81
},
{
"epoch": 1.0936454849498327,
"grad_norm": 0.4814144968986511,
"learning_rate": 0.00019661846496106785,
"loss": 4.6449,
"step": 82
},
{
"epoch": 1.1070234113712374,
"grad_norm": 0.4526231288909912,
"learning_rate": 0.00019657397107897666,
"loss": 4.5546,
"step": 83
},
{
"epoch": 1.120401337792642,
"grad_norm": 0.4847906827926636,
"learning_rate": 0.00019652947719688544,
"loss": 4.4421,
"step": 84
},
{
"epoch": 1.1337792642140467,
"grad_norm": 0.5136271715164185,
"learning_rate": 0.00019648498331479422,
"loss": 4.7136,
"step": 85
},
{
"epoch": 1.1471571906354514,
"grad_norm": 0.49209895730018616,
"learning_rate": 0.000196440489432703,
"loss": 4.3145,
"step": 86
},
{
"epoch": 1.160535117056856,
"grad_norm": 0.4972032904624939,
"learning_rate": 0.00019639599555061182,
"loss": 4.0408,
"step": 87
},
{
"epoch": 1.1739130434782608,
"grad_norm": 0.5077862739562988,
"learning_rate": 0.0001963515016685206,
"loss": 4.4074,
"step": 88
},
{
"epoch": 1.1872909698996654,
"grad_norm": 0.5293861031532288,
"learning_rate": 0.00019630700778642935,
"loss": 4.5385,
"step": 89
},
{
"epoch": 1.2006688963210703,
"grad_norm": 0.5062645673751831,
"learning_rate": 0.00019626251390433816,
"loss": 4.5141,
"step": 90
},
{
"epoch": 1.214046822742475,
"grad_norm": 0.49655866622924805,
"learning_rate": 0.00019621802002224695,
"loss": 4.4765,
"step": 91
},
{
"epoch": 1.2274247491638797,
"grad_norm": 0.6059755086898804,
"learning_rate": 0.00019617352614015573,
"loss": 4.568,
"step": 92
},
{
"epoch": 1.2408026755852843,
"grad_norm": 0.5442761778831482,
"learning_rate": 0.0001961290322580645,
"loss": 4.7724,
"step": 93
},
{
"epoch": 1.254180602006689,
"grad_norm": 0.5426056385040283,
"learning_rate": 0.00019608453837597332,
"loss": 4.5308,
"step": 94
},
{
"epoch": 1.2675585284280937,
"grad_norm": 0.525372326374054,
"learning_rate": 0.0001960400444938821,
"loss": 4.394,
"step": 95
},
{
"epoch": 1.2809364548494984,
"grad_norm": 0.5407588481903076,
"learning_rate": 0.00019599555061179089,
"loss": 4.7347,
"step": 96
},
{
"epoch": 1.294314381270903,
"grad_norm": 0.5726659893989563,
"learning_rate": 0.00019595105672969967,
"loss": 4.9446,
"step": 97
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.6211283206939697,
"learning_rate": 0.00019590656284760848,
"loss": 4.697,
"step": 98
},
{
"epoch": 1.3210702341137124,
"grad_norm": 0.5627567172050476,
"learning_rate": 0.00019586206896551723,
"loss": 4.4892,
"step": 99
},
{
"epoch": 1.334448160535117,
"grad_norm": 0.6174790859222412,
"learning_rate": 0.00019581757508342604,
"loss": 4.5686,
"step": 100
},
{
"epoch": 1.3478260869565217,
"grad_norm": 0.5586990118026733,
"learning_rate": 0.00019577308120133482,
"loss": 4.3916,
"step": 101
},
{
"epoch": 1.3612040133779264,
"grad_norm": 0.5655365586280823,
"learning_rate": 0.0001957285873192436,
"loss": 4.1789,
"step": 102
},
{
"epoch": 1.374581939799331,
"grad_norm": 0.5834594964981079,
"learning_rate": 0.0001956840934371524,
"loss": 4.3316,
"step": 103
},
{
"epoch": 1.3879598662207357,
"grad_norm": 0.6065447926521301,
"learning_rate": 0.0001956395995550612,
"loss": 4.5167,
"step": 104
},
{
"epoch": 1.4013377926421404,
"grad_norm": 0.5250216722488403,
"learning_rate": 0.00019559510567296998,
"loss": 4.1718,
"step": 105
},
{
"epoch": 1.414715719063545,
"grad_norm": 0.5861116051673889,
"learning_rate": 0.00019555061179087876,
"loss": 4.3077,
"step": 106
},
{
"epoch": 1.4280936454849498,
"grad_norm": 0.6138104796409607,
"learning_rate": 0.00019550611790878755,
"loss": 4.4748,
"step": 107
},
{
"epoch": 1.4414715719063544,
"grad_norm": 0.6742071509361267,
"learning_rate": 0.00019546162402669636,
"loss": 4.8769,
"step": 108
},
{
"epoch": 1.4548494983277591,
"grad_norm": 0.6634951233863831,
"learning_rate": 0.0001954171301446051,
"loss": 4.6423,
"step": 109
},
{
"epoch": 1.468227424749164,
"grad_norm": 0.626646876335144,
"learning_rate": 0.00019537263626251392,
"loss": 4.4654,
"step": 110
},
{
"epoch": 1.4816053511705687,
"grad_norm": 0.6306963562965393,
"learning_rate": 0.0001953281423804227,
"loss": 4.7021,
"step": 111
},
{
"epoch": 1.4949832775919734,
"grad_norm": 0.620370626449585,
"learning_rate": 0.00019528364849833149,
"loss": 4.587,
"step": 112
},
{
"epoch": 1.508361204013378,
"grad_norm": 0.6410287618637085,
"learning_rate": 0.00019523915461624027,
"loss": 4.8089,
"step": 113
},
{
"epoch": 1.5217391304347827,
"grad_norm": 0.676434338092804,
"learning_rate": 0.00019519466073414908,
"loss": 4.668,
"step": 114
},
{
"epoch": 1.5351170568561874,
"grad_norm": 0.5756319761276245,
"learning_rate": 0.00019515016685205786,
"loss": 4.3223,
"step": 115
},
{
"epoch": 1.548494983277592,
"grad_norm": 0.5850693583488464,
"learning_rate": 0.00019510567296996664,
"loss": 4.2343,
"step": 116
},
{
"epoch": 1.5618729096989967,
"grad_norm": 0.6172360777854919,
"learning_rate": 0.00019506117908787542,
"loss": 4.6102,
"step": 117
},
{
"epoch": 1.5752508361204014,
"grad_norm": 0.5887568593025208,
"learning_rate": 0.00019501668520578423,
"loss": 4.8097,
"step": 118
},
{
"epoch": 1.588628762541806,
"grad_norm": 0.5763369798660278,
"learning_rate": 0.000194972191323693,
"loss": 4.2001,
"step": 119
},
{
"epoch": 1.6020066889632107,
"grad_norm": 0.6158986687660217,
"learning_rate": 0.0001949276974416018,
"loss": 4.7075,
"step": 120
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.5540957450866699,
"learning_rate": 0.00019488320355951058,
"loss": 4.452,
"step": 121
},
{
"epoch": 1.62876254180602,
"grad_norm": 0.6193795204162598,
"learning_rate": 0.00019483870967741936,
"loss": 4.4583,
"step": 122
},
{
"epoch": 1.6421404682274248,
"grad_norm": 0.6699966788291931,
"learning_rate": 0.00019479421579532815,
"loss": 4.3728,
"step": 123
},
{
"epoch": 1.6555183946488294,
"grad_norm": 0.5904677510261536,
"learning_rate": 0.00019474972191323696,
"loss": 4.5452,
"step": 124
},
{
"epoch": 1.6688963210702341,
"grad_norm": 0.6137760281562805,
"learning_rate": 0.00019470522803114574,
"loss": 4.2853,
"step": 125
},
{
"epoch": 1.6822742474916388,
"grad_norm": 0.6396192908287048,
"learning_rate": 0.00019466073414905452,
"loss": 4.4258,
"step": 126
},
{
"epoch": 1.6956521739130435,
"grad_norm": 0.6190487742424011,
"learning_rate": 0.0001946162402669633,
"loss": 4.9866,
"step": 127
},
{
"epoch": 1.7090301003344481,
"grad_norm": 0.6971675157546997,
"learning_rate": 0.0001945717463848721,
"loss": 4.2126,
"step": 128
},
{
"epoch": 1.7224080267558528,
"grad_norm": 0.6245931386947632,
"learning_rate": 0.00019452725250278087,
"loss": 4.8477,
"step": 129
},
{
"epoch": 1.7357859531772575,
"grad_norm": 0.5675052404403687,
"learning_rate": 0.00019448275862068965,
"loss": 4.4097,
"step": 130
},
{
"epoch": 1.7491638795986622,
"grad_norm": 0.6594040393829346,
"learning_rate": 0.00019443826473859846,
"loss": 4.3747,
"step": 131
},
{
"epoch": 1.7625418060200668,
"grad_norm": 0.6377655267715454,
"learning_rate": 0.00019439377085650724,
"loss": 4.2733,
"step": 132
},
{
"epoch": 1.7759197324414715,
"grad_norm": 0.6167862415313721,
"learning_rate": 0.00019434927697441602,
"loss": 4.5694,
"step": 133
},
{
"epoch": 1.7892976588628762,
"grad_norm": 0.577671468257904,
"learning_rate": 0.0001943047830923248,
"loss": 4.5006,
"step": 134
},
{
"epoch": 1.8026755852842808,
"grad_norm": 0.6361016035079956,
"learning_rate": 0.00019426028921023362,
"loss": 4.9907,
"step": 135
},
{
"epoch": 1.8160535117056855,
"grad_norm": 0.6445321440696716,
"learning_rate": 0.00019421579532814237,
"loss": 4.779,
"step": 136
},
{
"epoch": 1.8294314381270902,
"grad_norm": 0.5955402851104736,
"learning_rate": 0.00019417130144605118,
"loss": 4.6026,
"step": 137
},
{
"epoch": 1.8428093645484949,
"grad_norm": 0.6807080507278442,
"learning_rate": 0.00019412680756395996,
"loss": 4.7124,
"step": 138
},
{
"epoch": 1.8561872909698995,
"grad_norm": 0.5799288153648376,
"learning_rate": 0.00019408231368186875,
"loss": 4.0701,
"step": 139
},
{
"epoch": 1.8695652173913042,
"grad_norm": 0.6187757253646851,
"learning_rate": 0.00019403781979977753,
"loss": 4.705,
"step": 140
},
{
"epoch": 1.8829431438127089,
"grad_norm": 0.6614826917648315,
"learning_rate": 0.00019399332591768634,
"loss": 4.8146,
"step": 141
},
{
"epoch": 1.8963210702341136,
"grad_norm": 0.6204859614372253,
"learning_rate": 0.00019394883203559512,
"loss": 4.3041,
"step": 142
},
{
"epoch": 1.9096989966555182,
"grad_norm": 0.6527450680732727,
"learning_rate": 0.0001939043381535039,
"loss": 4.4493,
"step": 143
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.6470615267753601,
"learning_rate": 0.00019385984427141268,
"loss": 4.7771,
"step": 144
},
{
"epoch": 1.9364548494983278,
"grad_norm": 0.5642555952072144,
"learning_rate": 0.0001938153503893215,
"loss": 4.3344,
"step": 145
},
{
"epoch": 1.9498327759197325,
"grad_norm": 0.6206467151641846,
"learning_rate": 0.00019377085650723025,
"loss": 4.2191,
"step": 146
},
{
"epoch": 1.9632107023411371,
"grad_norm": 0.6079016923904419,
"learning_rate": 0.00019372636262513906,
"loss": 4.7397,
"step": 147
},
{
"epoch": 1.9765886287625418,
"grad_norm": 0.6197662353515625,
"learning_rate": 0.00019368186874304784,
"loss": 4.5342,
"step": 148
},
{
"epoch": 1.9899665551839465,
"grad_norm": 0.6556297540664673,
"learning_rate": 0.00019363737486095662,
"loss": 4.6709,
"step": 149
},
{
"epoch": 2.0,
"grad_norm": 0.7837930917739868,
"learning_rate": 0.0001935928809788654,
"loss": 4.6215,
"step": 150
},
{
"epoch": 2.0133779264214047,
"grad_norm": 0.5267267227172852,
"learning_rate": 0.00019354838709677422,
"loss": 4.2695,
"step": 151
},
{
"epoch": 2.0267558528428093,
"grad_norm": 0.5862157344818115,
"learning_rate": 0.000193503893214683,
"loss": 4.3702,
"step": 152
},
{
"epoch": 2.040133779264214,
"grad_norm": 0.538254976272583,
"learning_rate": 0.00019345939933259178,
"loss": 4.3953,
"step": 153
},
{
"epoch": 2.0535117056856187,
"grad_norm": 0.5977053642272949,
"learning_rate": 0.00019341490545050056,
"loss": 4.2156,
"step": 154
},
{
"epoch": 2.0668896321070234,
"grad_norm": 0.606006383895874,
"learning_rate": 0.00019337041156840937,
"loss": 4.2802,
"step": 155
},
{
"epoch": 2.080267558528428,
"grad_norm": 0.6071277856826782,
"learning_rate": 0.00019332591768631813,
"loss": 4.5545,
"step": 156
},
{
"epoch": 2.0936454849498327,
"grad_norm": 0.6281546354293823,
"learning_rate": 0.00019328142380422694,
"loss": 4.6105,
"step": 157
},
{
"epoch": 2.1070234113712374,
"grad_norm": 0.5703116655349731,
"learning_rate": 0.00019323692992213572,
"loss": 4.2751,
"step": 158
},
{
"epoch": 2.120401337792642,
"grad_norm": 0.6587452292442322,
"learning_rate": 0.0001931924360400445,
"loss": 4.6342,
"step": 159
},
{
"epoch": 2.1337792642140467,
"grad_norm": 0.6141905784606934,
"learning_rate": 0.00019314794215795328,
"loss": 4.4345,
"step": 160
},
{
"epoch": 2.1471571906354514,
"grad_norm": 0.6741939187049866,
"learning_rate": 0.0001931034482758621,
"loss": 4.0257,
"step": 161
},
{
"epoch": 2.160535117056856,
"grad_norm": 0.6468759179115295,
"learning_rate": 0.00019305895439377088,
"loss": 4.2313,
"step": 162
},
{
"epoch": 2.1739130434782608,
"grad_norm": 0.6703383326530457,
"learning_rate": 0.00019301446051167966,
"loss": 4.2164,
"step": 163
},
{
"epoch": 2.1872909698996654,
"grad_norm": 0.710967481136322,
"learning_rate": 0.00019296996662958844,
"loss": 4.3398,
"step": 164
},
{
"epoch": 2.20066889632107,
"grad_norm": 0.6862124800682068,
"learning_rate": 0.00019292547274749725,
"loss": 4.3379,
"step": 165
},
{
"epoch": 2.2140468227424748,
"grad_norm": 0.6288430690765381,
"learning_rate": 0.000192880978865406,
"loss": 4.3487,
"step": 166
},
{
"epoch": 2.2274247491638794,
"grad_norm": 0.6358796954154968,
"learning_rate": 0.00019283648498331481,
"loss": 4.1656,
"step": 167
},
{
"epoch": 2.240802675585284,
"grad_norm": 0.6818917393684387,
"learning_rate": 0.0001927919911012236,
"loss": 4.5363,
"step": 168
},
{
"epoch": 2.254180602006689,
"grad_norm": 0.6996105313301086,
"learning_rate": 0.00019274749721913238,
"loss": 4.3208,
"step": 169
},
{
"epoch": 2.2675585284280935,
"grad_norm": 0.6730326414108276,
"learning_rate": 0.00019270300333704116,
"loss": 4.1401,
"step": 170
},
{
"epoch": 2.280936454849498,
"grad_norm": 0.7022603750228882,
"learning_rate": 0.00019265850945494994,
"loss": 4.5761,
"step": 171
},
{
"epoch": 2.294314381270903,
"grad_norm": 0.6525995135307312,
"learning_rate": 0.00019261401557285875,
"loss": 4.4017,
"step": 172
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.7066033482551575,
"learning_rate": 0.0001925695216907675,
"loss": 4.0037,
"step": 173
},
{
"epoch": 2.321070234113712,
"grad_norm": 0.6708059310913086,
"learning_rate": 0.00019252502780867632,
"loss": 4.1947,
"step": 174
},
{
"epoch": 2.334448160535117,
"grad_norm": 0.8711172342300415,
"learning_rate": 0.0001924805339265851,
"loss": 3.9958,
"step": 175
},
{
"epoch": 2.3478260869565215,
"grad_norm": 0.7258634567260742,
"learning_rate": 0.00019243604004449388,
"loss": 4.4682,
"step": 176
},
{
"epoch": 2.361204013377926,
"grad_norm": 0.7693021893501282,
"learning_rate": 0.00019239154616240267,
"loss": 4.54,
"step": 177
},
{
"epoch": 2.374581939799331,
"grad_norm": 0.7271276116371155,
"learning_rate": 0.00019234705228031148,
"loss": 4.2942,
"step": 178
},
{
"epoch": 2.387959866220736,
"grad_norm": 0.6836609244346619,
"learning_rate": 0.00019230255839822026,
"loss": 4.3099,
"step": 179
},
{
"epoch": 2.4013377926421406,
"grad_norm": 0.731164813041687,
"learning_rate": 0.00019225806451612904,
"loss": 4.4077,
"step": 180
},
{
"epoch": 2.4147157190635453,
"grad_norm": 0.7575274109840393,
"learning_rate": 0.00019221357063403782,
"loss": 4.6572,
"step": 181
},
{
"epoch": 2.42809364548495,
"grad_norm": 0.8461325168609619,
"learning_rate": 0.00019216907675194663,
"loss": 4.4922,
"step": 182
},
{
"epoch": 2.4414715719063547,
"grad_norm": 0.7225251197814941,
"learning_rate": 0.0001921245828698554,
"loss": 4.0372,
"step": 183
},
{
"epoch": 2.4548494983277593,
"grad_norm": 3.563720703125,
"learning_rate": 0.0001920800889877642,
"loss": 4.5412,
"step": 184
},
{
"epoch": 2.468227424749164,
"grad_norm": 0.8452121019363403,
"learning_rate": 0.00019203559510567298,
"loss": 4.4961,
"step": 185
},
{
"epoch": 2.4816053511705687,
"grad_norm": 0.8734024167060852,
"learning_rate": 0.00019199110122358176,
"loss": 4.0884,
"step": 186
},
{
"epoch": 2.4949832775919734,
"grad_norm": 1.1765823364257812,
"learning_rate": 0.00019194660734149054,
"loss": 4.2228,
"step": 187
},
{
"epoch": 2.508361204013378,
"grad_norm": 0.750206708908081,
"learning_rate": 0.00019190211345939935,
"loss": 4.4305,
"step": 188
},
{
"epoch": 2.5217391304347827,
"grad_norm": 0.7574430704116821,
"learning_rate": 0.00019185761957730814,
"loss": 4.1511,
"step": 189
},
{
"epoch": 2.5351170568561874,
"grad_norm": 0.7105517387390137,
"learning_rate": 0.00019181312569521692,
"loss": 4.4793,
"step": 190
},
{
"epoch": 2.548494983277592,
"grad_norm": 0.7495557069778442,
"learning_rate": 0.0001917686318131257,
"loss": 4.1335,
"step": 191
},
{
"epoch": 2.5618729096989967,
"grad_norm": 0.8001168966293335,
"learning_rate": 0.0001917241379310345,
"loss": 4.7898,
"step": 192
},
{
"epoch": 2.5752508361204014,
"grad_norm": 0.7402104735374451,
"learning_rate": 0.00019167964404894327,
"loss": 4.4482,
"step": 193
},
{
"epoch": 2.588628762541806,
"grad_norm": 0.748267650604248,
"learning_rate": 0.00019163515016685207,
"loss": 4.3167,
"step": 194
},
{
"epoch": 2.6020066889632107,
"grad_norm": 0.8291250467300415,
"learning_rate": 0.00019159065628476086,
"loss": 4.058,
"step": 195
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.6945542693138123,
"learning_rate": 0.00019154616240266964,
"loss": 3.9751,
"step": 196
},
{
"epoch": 2.62876254180602,
"grad_norm": 0.7307319045066833,
"learning_rate": 0.00019150166852057842,
"loss": 4.2736,
"step": 197
},
{
"epoch": 2.6421404682274248,
"grad_norm": 0.7489168047904968,
"learning_rate": 0.00019145717463848723,
"loss": 4.3075,
"step": 198
},
{
"epoch": 2.6555183946488294,
"grad_norm": 0.9727582931518555,
"learning_rate": 0.00019141268075639601,
"loss": 4.6474,
"step": 199
},
{
"epoch": 2.668896321070234,
"grad_norm": 0.6776256561279297,
"learning_rate": 0.0001913681868743048,
"loss": 4.4217,
"step": 200
},
{
"epoch": 2.682274247491639,
"grad_norm": 0.7305111885070801,
"learning_rate": 0.00019132369299221358,
"loss": 4.2804,
"step": 201
},
{
"epoch": 2.6956521739130435,
"grad_norm": 0.7196978330612183,
"learning_rate": 0.0001912791991101224,
"loss": 4.3941,
"step": 202
},
{
"epoch": 2.709030100334448,
"grad_norm": 0.7988458871841431,
"learning_rate": 0.00019123470522803114,
"loss": 4.437,
"step": 203
},
{
"epoch": 2.722408026755853,
"grad_norm": 0.7004797458648682,
"learning_rate": 0.00019119021134593995,
"loss": 4.4986,
"step": 204
},
{
"epoch": 2.7357859531772575,
"grad_norm": 0.677796483039856,
"learning_rate": 0.00019114571746384874,
"loss": 4.0851,
"step": 205
},
{
"epoch": 2.749163879598662,
"grad_norm": 0.7527475357055664,
"learning_rate": 0.00019110122358175752,
"loss": 4.4469,
"step": 206
},
{
"epoch": 2.762541806020067,
"grad_norm": 1.1659115552902222,
"learning_rate": 0.0001910567296996663,
"loss": 4.3284,
"step": 207
},
{
"epoch": 2.7759197324414715,
"grad_norm": 0.7238364815711975,
"learning_rate": 0.0001910122358175751,
"loss": 4.2605,
"step": 208
},
{
"epoch": 2.789297658862876,
"grad_norm": 0.7537760734558105,
"learning_rate": 0.0001909677419354839,
"loss": 4.3775,
"step": 209
},
{
"epoch": 2.802675585284281,
"grad_norm": 0.6874127388000488,
"learning_rate": 0.00019092324805339267,
"loss": 4.3404,
"step": 210
},
{
"epoch": 2.8160535117056855,
"grad_norm": 0.7045959830284119,
"learning_rate": 0.00019087875417130146,
"loss": 4.1568,
"step": 211
},
{
"epoch": 2.82943143812709,
"grad_norm": 0.7249194383621216,
"learning_rate": 0.00019083426028921027,
"loss": 4.1969,
"step": 212
},
{
"epoch": 2.842809364548495,
"grad_norm": 0.8331268429756165,
"learning_rate": 0.00019078976640711902,
"loss": 4.3169,
"step": 213
},
{
"epoch": 2.8561872909698995,
"grad_norm": 0.7171936631202698,
"learning_rate": 0.0001907452725250278,
"loss": 4.5123,
"step": 214
},
{
"epoch": 2.869565217391304,
"grad_norm": 0.759919285774231,
"learning_rate": 0.0001907007786429366,
"loss": 4.5412,
"step": 215
},
{
"epoch": 2.882943143812709,
"grad_norm": 0.7451274991035461,
"learning_rate": 0.0001906562847608454,
"loss": 4.5253,
"step": 216
},
{
"epoch": 2.8963210702341136,
"grad_norm": 0.6564481258392334,
"learning_rate": 0.00019061179087875418,
"loss": 4.1092,
"step": 217
},
{
"epoch": 2.9096989966555182,
"grad_norm": 0.7339865565299988,
"learning_rate": 0.00019056729699666296,
"loss": 4.5092,
"step": 218
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.7113937735557556,
"learning_rate": 0.00019052280311457177,
"loss": 4.3355,
"step": 219
},
{
"epoch": 2.936454849498328,
"grad_norm": 0.7306456565856934,
"learning_rate": 0.00019047830923248053,
"loss": 4.5745,
"step": 220
},
{
"epoch": 2.9498327759197327,
"grad_norm": 0.7971818447113037,
"learning_rate": 0.00019043381535038933,
"loss": 4.2903,
"step": 221
},
{
"epoch": 2.9632107023411374,
"grad_norm": 0.7757331728935242,
"learning_rate": 0.00019038932146829812,
"loss": 4.2832,
"step": 222
},
{
"epoch": 2.976588628762542,
"grad_norm": 0.7326288223266602,
"learning_rate": 0.0001903448275862069,
"loss": 4.2444,
"step": 223
},
{
"epoch": 2.9899665551839467,
"grad_norm": 0.7363834381103516,
"learning_rate": 0.00019030033370411568,
"loss": 4.6744,
"step": 224
},
{
"epoch": 3.0,
"grad_norm": 0.8835271596908569,
"learning_rate": 0.0001902558398220245,
"loss": 4.432,
"step": 225
},
{
"epoch": 3.0133779264214047,
"grad_norm": 0.6591921448707581,
"learning_rate": 0.00019021134593993327,
"loss": 4.1353,
"step": 226
},
{
"epoch": 3.0267558528428093,
"grad_norm": 0.6895263195037842,
"learning_rate": 0.00019016685205784206,
"loss": 4.1253,
"step": 227
},
{
"epoch": 3.040133779264214,
"grad_norm": 0.6476898789405823,
"learning_rate": 0.00019012235817575084,
"loss": 4.0354,
"step": 228
},
{
"epoch": 3.0535117056856187,
"grad_norm": 0.6398957967758179,
"learning_rate": 0.00019007786429365965,
"loss": 4.024,
"step": 229
},
{
"epoch": 3.0668896321070234,
"grad_norm": 0.7483389973640442,
"learning_rate": 0.0001900333704115684,
"loss": 4.1405,
"step": 230
},
{
"epoch": 3.080267558528428,
"grad_norm": 0.7003724575042725,
"learning_rate": 0.0001899888765294772,
"loss": 4.3593,
"step": 231
},
{
"epoch": 3.0936454849498327,
"grad_norm": 0.7426732182502747,
"learning_rate": 0.000189944382647386,
"loss": 4.485,
"step": 232
},
{
"epoch": 3.1070234113712374,
"grad_norm": 0.6957541108131409,
"learning_rate": 0.00018989988876529478,
"loss": 4.1017,
"step": 233
},
{
"epoch": 3.120401337792642,
"grad_norm": 0.8613067865371704,
"learning_rate": 0.00018985539488320356,
"loss": 4.3038,
"step": 234
},
{
"epoch": 3.1337792642140467,
"grad_norm": 0.8375754952430725,
"learning_rate": 0.00018981090100111237,
"loss": 4.4356,
"step": 235
},
{
"epoch": 3.1471571906354514,
"grad_norm": 0.7878522872924805,
"learning_rate": 0.00018976640711902115,
"loss": 3.9916,
"step": 236
},
{
"epoch": 3.160535117056856,
"grad_norm": 0.7463901042938232,
"learning_rate": 0.00018972191323692993,
"loss": 3.6761,
"step": 237
},
{
"epoch": 3.1739130434782608,
"grad_norm": 0.7360939979553223,
"learning_rate": 0.00018967741935483872,
"loss": 3.9573,
"step": 238
},
{
"epoch": 3.1872909698996654,
"grad_norm": 0.891861081123352,
"learning_rate": 0.00018963292547274753,
"loss": 4.1853,
"step": 239
},
{
"epoch": 3.20066889632107,
"grad_norm": 0.8589549660682678,
"learning_rate": 0.00018958843159065628,
"loss": 4.0679,
"step": 240
},
{
"epoch": 3.2140468227424748,
"grad_norm": 0.9534163475036621,
"learning_rate": 0.0001895439377085651,
"loss": 3.732,
"step": 241
},
{
"epoch": 3.2274247491638794,
"grad_norm": 0.8968185186386108,
"learning_rate": 0.00018949944382647387,
"loss": 4.2217,
"step": 242
},
{
"epoch": 3.240802675585284,
"grad_norm": 0.81589275598526,
"learning_rate": 0.00018945494994438266,
"loss": 4.428,
"step": 243
},
{
"epoch": 3.254180602006689,
"grad_norm": 0.929050862789154,
"learning_rate": 0.00018941045606229144,
"loss": 4.3468,
"step": 244
},
{
"epoch": 3.2675585284280935,
"grad_norm": 0.8535035252571106,
"learning_rate": 0.00018936596218020025,
"loss": 3.8489,
"step": 245
},
{
"epoch": 3.280936454849498,
"grad_norm": 0.9484681487083435,
"learning_rate": 0.00018932146829810903,
"loss": 4.0132,
"step": 246
},
{
"epoch": 3.294314381270903,
"grad_norm": 0.8190047144889832,
"learning_rate": 0.0001892769744160178,
"loss": 4.3574,
"step": 247
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.8764749765396118,
"learning_rate": 0.0001892324805339266,
"loss": 4.3103,
"step": 248
},
{
"epoch": 3.321070234113712,
"grad_norm": 0.8929185271263123,
"learning_rate": 0.0001891879866518354,
"loss": 4.3606,
"step": 249
},
{
"epoch": 3.334448160535117,
"grad_norm": 0.9096692204475403,
"learning_rate": 0.00018914349276974416,
"loss": 4.0047,
"step": 250
},
{
"epoch": 3.3478260869565215,
"grad_norm": 0.885143518447876,
"learning_rate": 0.00018909899888765297,
"loss": 4.182,
"step": 251
},
{
"epoch": 3.361204013377926,
"grad_norm": 0.7724215984344482,
"learning_rate": 0.00018905450500556175,
"loss": 3.9529,
"step": 252
},
{
"epoch": 3.374581939799331,
"grad_norm": 0.8351865410804749,
"learning_rate": 0.00018901001112347053,
"loss": 3.9533,
"step": 253
},
{
"epoch": 3.387959866220736,
"grad_norm": 0.8684999942779541,
"learning_rate": 0.00018896551724137932,
"loss": 3.8594,
"step": 254
},
{
"epoch": 3.4013377926421406,
"grad_norm": 0.8903334736824036,
"learning_rate": 0.0001889210233592881,
"loss": 3.9248,
"step": 255
},
{
"epoch": 3.4147157190635453,
"grad_norm": 0.826690137386322,
"learning_rate": 0.0001888765294771969,
"loss": 4.0389,
"step": 256
},
{
"epoch": 3.42809364548495,
"grad_norm": 0.8306142687797546,
"learning_rate": 0.00018883203559510566,
"loss": 3.8168,
"step": 257
},
{
"epoch": 3.4414715719063547,
"grad_norm": 0.9032199382781982,
"learning_rate": 0.00018878754171301447,
"loss": 4.178,
"step": 258
},
{
"epoch": 3.4548494983277593,
"grad_norm": 0.9081966280937195,
"learning_rate": 0.00018874304783092326,
"loss": 4.2583,
"step": 259
},
{
"epoch": 3.468227424749164,
"grad_norm": 0.8424077033996582,
"learning_rate": 0.00018869855394883204,
"loss": 4.3285,
"step": 260
},
{
"epoch": 3.4816053511705687,
"grad_norm": 0.8302170038223267,
"learning_rate": 0.00018865406006674082,
"loss": 4.1346,
"step": 261
},
{
"epoch": 3.4949832775919734,
"grad_norm": 0.8747193217277527,
"learning_rate": 0.00018860956618464963,
"loss": 4.0747,
"step": 262
},
{
"epoch": 3.508361204013378,
"grad_norm": 0.8613927364349365,
"learning_rate": 0.0001885650723025584,
"loss": 4.2346,
"step": 263
},
{
"epoch": 3.5217391304347827,
"grad_norm": 0.8321558833122253,
"learning_rate": 0.0001885205784204672,
"loss": 3.9781,
"step": 264
},
{
"epoch": 3.5351170568561874,
"grad_norm": 0.8961741328239441,
"learning_rate": 0.00018847608453837598,
"loss": 4.311,
"step": 265
},
{
"epoch": 3.548494983277592,
"grad_norm": 0.7703898549079895,
"learning_rate": 0.00018843159065628479,
"loss": 4.1163,
"step": 266
},
{
"epoch": 3.5618729096989967,
"grad_norm": 0.880051851272583,
"learning_rate": 0.00018838709677419354,
"loss": 3.8032,
"step": 267
},
{
"epoch": 3.5752508361204014,
"grad_norm": 0.8287038207054138,
"learning_rate": 0.00018834260289210235,
"loss": 4.1627,
"step": 268
},
{
"epoch": 3.588628762541806,
"grad_norm": 0.9726569652557373,
"learning_rate": 0.00018829810901001113,
"loss": 4.4055,
"step": 269
},
{
"epoch": 3.6020066889632107,
"grad_norm": 0.8071132898330688,
"learning_rate": 0.00018825361512791992,
"loss": 4.1709,
"step": 270
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.8310988545417786,
"learning_rate": 0.0001882091212458287,
"loss": 4.2359,
"step": 271
},
{
"epoch": 3.62876254180602,
"grad_norm": 0.8713561296463013,
"learning_rate": 0.0001881646273637375,
"loss": 4.1247,
"step": 272
},
{
"epoch": 3.6421404682274248,
"grad_norm": 0.8964342474937439,
"learning_rate": 0.0001881201334816463,
"loss": 4.0794,
"step": 273
},
{
"epoch": 3.6555183946488294,
"grad_norm": 0.9901681542396545,
"learning_rate": 0.00018807563959955507,
"loss": 4.0217,
"step": 274
},
{
"epoch": 3.668896321070234,
"grad_norm": 0.9279042482376099,
"learning_rate": 0.00018803114571746385,
"loss": 4.3244,
"step": 275
},
{
"epoch": 3.682274247491639,
"grad_norm": 0.8105964660644531,
"learning_rate": 0.00018798665183537266,
"loss": 3.9041,
"step": 276
},
{
"epoch": 3.6956521739130435,
"grad_norm": 0.8511622548103333,
"learning_rate": 0.00018794215795328142,
"loss": 3.8969,
"step": 277
},
{
"epoch": 3.709030100334448,
"grad_norm": 0.9072037935256958,
"learning_rate": 0.00018789766407119023,
"loss": 4.2185,
"step": 278
},
{
"epoch": 3.722408026755853,
"grad_norm": 0.9792962670326233,
"learning_rate": 0.000187853170189099,
"loss": 4.1915,
"step": 279
},
{
"epoch": 3.7357859531772575,
"grad_norm": 0.8579828143119812,
"learning_rate": 0.0001878086763070078,
"loss": 3.8903,
"step": 280
},
{
"epoch": 3.749163879598662,
"grad_norm": 0.9866719841957092,
"learning_rate": 0.00018776418242491658,
"loss": 4.2022,
"step": 281
},
{
"epoch": 3.762541806020067,
"grad_norm": 0.9251964688301086,
"learning_rate": 0.00018771968854282539,
"loss": 3.9536,
"step": 282
},
{
"epoch": 3.7759197324414715,
"grad_norm": 1.0300836563110352,
"learning_rate": 0.00018767519466073417,
"loss": 4.2908,
"step": 283
},
{
"epoch": 3.789297658862876,
"grad_norm": 1.0194575786590576,
"learning_rate": 0.00018763070077864295,
"loss": 4.1851,
"step": 284
},
{
"epoch": 3.802675585284281,
"grad_norm": 0.8165330290794373,
"learning_rate": 0.00018758620689655173,
"loss": 4.138,
"step": 285
},
{
"epoch": 3.8160535117056855,
"grad_norm": 1.0104280710220337,
"learning_rate": 0.00018754171301446054,
"loss": 3.9481,
"step": 286
},
{
"epoch": 3.82943143812709,
"grad_norm": 0.9972538352012634,
"learning_rate": 0.0001874972191323693,
"loss": 4.3932,
"step": 287
},
{
"epoch": 3.842809364548495,
"grad_norm": 0.96323162317276,
"learning_rate": 0.0001874527252502781,
"loss": 4.1133,
"step": 288
},
{
"epoch": 3.8561872909698995,
"grad_norm": 0.8500615954399109,
"learning_rate": 0.0001874082313681869,
"loss": 4.2205,
"step": 289
},
{
"epoch": 3.869565217391304,
"grad_norm": 0.8451250195503235,
"learning_rate": 0.00018736373748609567,
"loss": 4.1371,
"step": 290
},
{
"epoch": 3.882943143812709,
"grad_norm": 0.9399815201759338,
"learning_rate": 0.00018731924360400445,
"loss": 4.5237,
"step": 291
},
{
"epoch": 3.8963210702341136,
"grad_norm": 0.8061622977256775,
"learning_rate": 0.00018727474972191326,
"loss": 4.1033,
"step": 292
},
{
"epoch": 3.9096989966555182,
"grad_norm": 0.7987121343612671,
"learning_rate": 0.00018723025583982205,
"loss": 3.9311,
"step": 293
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.9041138291358948,
"learning_rate": 0.00018718576195773083,
"loss": 4.0252,
"step": 294
},
{
"epoch": 3.936454849498328,
"grad_norm": 1.0002484321594238,
"learning_rate": 0.0001871412680756396,
"loss": 4.4605,
"step": 295
},
{
"epoch": 3.9498327759197327,
"grad_norm": 0.9991098046302795,
"learning_rate": 0.0001870967741935484,
"loss": 4.1528,
"step": 296
},
{
"epoch": 3.9632107023411374,
"grad_norm": 1.2179397344589233,
"learning_rate": 0.00018705228031145718,
"loss": 4.5224,
"step": 297
},
{
"epoch": 3.976588628762542,
"grad_norm": 0.8279774785041809,
"learning_rate": 0.00018700778642936596,
"loss": 3.9464,
"step": 298
},
{
"epoch": 3.9899665551839467,
"grad_norm": 0.8012803792953491,
"learning_rate": 0.00018696329254727477,
"loss": 4.0139,
"step": 299
},
{
"epoch": 4.0,
"grad_norm": 0.9700272083282471,
"learning_rate": 0.00018691879866518355,
"loss": 3.8306,
"step": 300
},
{
"epoch": 4.013377926421405,
"grad_norm": 0.7136749625205994,
"learning_rate": 0.00018687430478309233,
"loss": 3.9253,
"step": 301
},
{
"epoch": 4.026755852842809,
"grad_norm": 0.7885096669197083,
"learning_rate": 0.00018682981090100111,
"loss": 3.927,
"step": 302
},
{
"epoch": 4.040133779264214,
"grad_norm": 0.7801666855812073,
"learning_rate": 0.00018678531701890992,
"loss": 3.6482,
"step": 303
},
{
"epoch": 4.053511705685619,
"grad_norm": 0.7857955098152161,
"learning_rate": 0.00018674082313681868,
"loss": 4.0665,
"step": 304
},
{
"epoch": 4.066889632107023,
"grad_norm": 0.707421064376831,
"learning_rate": 0.0001866963292547275,
"loss": 3.9142,
"step": 305
},
{
"epoch": 4.080267558528428,
"grad_norm": 0.7936912775039673,
"learning_rate": 0.00018665183537263627,
"loss": 4.1227,
"step": 306
},
{
"epoch": 4.093645484949833,
"grad_norm": 0.8899754881858826,
"learning_rate": 0.00018660734149054505,
"loss": 3.7661,
"step": 307
},
{
"epoch": 4.107023411371237,
"grad_norm": 0.7760347723960876,
"learning_rate": 0.00018656284760845384,
"loss": 3.8921,
"step": 308
},
{
"epoch": 4.120401337792642,
"grad_norm": 0.8672968745231628,
"learning_rate": 0.00018651835372636265,
"loss": 3.6037,
"step": 309
},
{
"epoch": 4.133779264214047,
"grad_norm": 0.8046863675117493,
"learning_rate": 0.0001864738598442714,
"loss": 3.9117,
"step": 310
},
{
"epoch": 4.147157190635451,
"grad_norm": 0.9172897934913635,
"learning_rate": 0.0001864293659621802,
"loss": 3.7229,
"step": 311
},
{
"epoch": 4.160535117056856,
"grad_norm": 0.9616653919219971,
"learning_rate": 0.000186384872080089,
"loss": 3.8851,
"step": 312
},
{
"epoch": 4.173913043478261,
"grad_norm": 0.9659278988838196,
"learning_rate": 0.0001863403781979978,
"loss": 4.005,
"step": 313
},
{
"epoch": 4.187290969899665,
"grad_norm": 0.9171205163002014,
"learning_rate": 0.00018629588431590656,
"loss": 3.8634,
"step": 314
},
{
"epoch": 4.20066889632107,
"grad_norm": 0.9968683123588562,
"learning_rate": 0.00018625139043381537,
"loss": 3.7321,
"step": 315
},
{
"epoch": 4.214046822742475,
"grad_norm": 0.8762083053588867,
"learning_rate": 0.00018620689655172415,
"loss": 3.931,
"step": 316
},
{
"epoch": 4.2274247491638794,
"grad_norm": 0.9815887212753296,
"learning_rate": 0.00018616240266963293,
"loss": 3.9975,
"step": 317
},
{
"epoch": 4.240802675585284,
"grad_norm": 1.0065505504608154,
"learning_rate": 0.00018611790878754171,
"loss": 3.8364,
"step": 318
},
{
"epoch": 4.254180602006689,
"grad_norm": 0.9785431623458862,
"learning_rate": 0.00018607341490545052,
"loss": 3.8822,
"step": 319
},
{
"epoch": 4.2675585284280935,
"grad_norm": 1.077799677848816,
"learning_rate": 0.00018602892102335928,
"loss": 3.8299,
"step": 320
},
{
"epoch": 4.280936454849498,
"grad_norm": 0.8109619617462158,
"learning_rate": 0.0001859844271412681,
"loss": 3.8096,
"step": 321
},
{
"epoch": 4.294314381270903,
"grad_norm": 0.967856764793396,
"learning_rate": 0.00018593993325917687,
"loss": 3.8639,
"step": 322
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.8657905459403992,
"learning_rate": 0.00018589543937708568,
"loss": 3.7556,
"step": 323
},
{
"epoch": 4.321070234113712,
"grad_norm": 0.9641517400741577,
"learning_rate": 0.00018585094549499444,
"loss": 3.9702,
"step": 324
},
{
"epoch": 4.334448160535117,
"grad_norm": 0.9664435982704163,
"learning_rate": 0.00018580645161290325,
"loss": 3.8754,
"step": 325
},
{
"epoch": 4.3478260869565215,
"grad_norm": 0.8322617411613464,
"learning_rate": 0.00018576195773081203,
"loss": 3.83,
"step": 326
},
{
"epoch": 4.361204013377926,
"grad_norm": 1.0363450050354004,
"learning_rate": 0.0001857174638487208,
"loss": 3.9825,
"step": 327
},
{
"epoch": 4.374581939799331,
"grad_norm": 1.0125840902328491,
"learning_rate": 0.0001856729699666296,
"loss": 3.6525,
"step": 328
},
{
"epoch": 4.3879598662207355,
"grad_norm": 0.9922601580619812,
"learning_rate": 0.0001856284760845384,
"loss": 4.2373,
"step": 329
},
{
"epoch": 4.40133779264214,
"grad_norm": 0.9070426225662231,
"learning_rate": 0.00018558398220244716,
"loss": 3.9623,
"step": 330
},
{
"epoch": 4.414715719063545,
"grad_norm": 0.9369637370109558,
"learning_rate": 0.00018553948832035597,
"loss": 3.9297,
"step": 331
},
{
"epoch": 4.4280936454849495,
"grad_norm": 1.108876347541809,
"learning_rate": 0.00018549499443826475,
"loss": 3.7325,
"step": 332
},
{
"epoch": 4.441471571906354,
"grad_norm": 0.9405660629272461,
"learning_rate": 0.00018545050055617356,
"loss": 3.8615,
"step": 333
},
{
"epoch": 4.454849498327759,
"grad_norm": 0.9730128645896912,
"learning_rate": 0.00018540600667408231,
"loss": 4.1794,
"step": 334
},
{
"epoch": 4.468227424749164,
"grad_norm": 0.9341335892677307,
"learning_rate": 0.00018536151279199112,
"loss": 3.9422,
"step": 335
},
{
"epoch": 4.481605351170568,
"grad_norm": 0.9262625575065613,
"learning_rate": 0.0001853170189098999,
"loss": 3.9819,
"step": 336
},
{
"epoch": 4.494983277591973,
"grad_norm": 1.0419141054153442,
"learning_rate": 0.00018527252502780866,
"loss": 3.7481,
"step": 337
},
{
"epoch": 4.508361204013378,
"grad_norm": 0.8986826539039612,
"learning_rate": 0.00018522803114571747,
"loss": 4.1195,
"step": 338
},
{
"epoch": 4.521739130434782,
"grad_norm": 0.9502431154251099,
"learning_rate": 0.00018518353726362625,
"loss": 3.8521,
"step": 339
},
{
"epoch": 4.535117056856187,
"grad_norm": 0.8936267495155334,
"learning_rate": 0.00018513904338153504,
"loss": 3.647,
"step": 340
},
{
"epoch": 4.548494983277592,
"grad_norm": 0.8870158195495605,
"learning_rate": 0.00018509454949944382,
"loss": 3.7624,
"step": 341
},
{
"epoch": 4.561872909698996,
"grad_norm": 0.9030978679656982,
"learning_rate": 0.00018505005561735263,
"loss": 3.8018,
"step": 342
},
{
"epoch": 4.575250836120401,
"grad_norm": 0.8690946698188782,
"learning_rate": 0.0001850055617352614,
"loss": 3.6183,
"step": 343
},
{
"epoch": 4.588628762541806,
"grad_norm": 0.9812071323394775,
"learning_rate": 0.0001849610678531702,
"loss": 4.1231,
"step": 344
},
{
"epoch": 4.602006688963211,
"grad_norm": 0.9404383301734924,
"learning_rate": 0.00018491657397107897,
"loss": 3.9645,
"step": 345
},
{
"epoch": 4.615384615384615,
"grad_norm": 1.0422123670578003,
"learning_rate": 0.00018487208008898778,
"loss": 3.9031,
"step": 346
},
{
"epoch": 4.6287625418060205,
"grad_norm": 0.9838129281997681,
"learning_rate": 0.00018482758620689654,
"loss": 3.9985,
"step": 347
},
{
"epoch": 4.642140468227424,
"grad_norm": 0.9232532978057861,
"learning_rate": 0.00018478309232480535,
"loss": 4.0343,
"step": 348
},
{
"epoch": 4.65551839464883,
"grad_norm": 0.9242956042289734,
"learning_rate": 0.00018473859844271413,
"loss": 4.0669,
"step": 349
},
{
"epoch": 4.668896321070234,
"grad_norm": 0.919269859790802,
"learning_rate": 0.0001846941045606229,
"loss": 4.0549,
"step": 350
},
{
"epoch": 4.682274247491639,
"grad_norm": 0.93565833568573,
"learning_rate": 0.0001846496106785317,
"loss": 4.1306,
"step": 351
},
{
"epoch": 4.695652173913043,
"grad_norm": 0.9001899361610413,
"learning_rate": 0.0001846051167964405,
"loss": 3.8916,
"step": 352
},
{
"epoch": 4.709030100334449,
"grad_norm": 0.8896821737289429,
"learning_rate": 0.0001845606229143493,
"loss": 3.8377,
"step": 353
},
{
"epoch": 4.722408026755852,
"grad_norm": 1.0137807130813599,
"learning_rate": 0.00018451612903225807,
"loss": 3.9923,
"step": 354
},
{
"epoch": 4.735785953177258,
"grad_norm": 1.075823426246643,
"learning_rate": 0.00018447163515016685,
"loss": 4.0706,
"step": 355
},
{
"epoch": 4.749163879598662,
"grad_norm": 1.0076895952224731,
"learning_rate": 0.00018442714126807566,
"loss": 4.0759,
"step": 356
},
{
"epoch": 4.762541806020067,
"grad_norm": 0.9387428164482117,
"learning_rate": 0.00018438264738598442,
"loss": 3.6959,
"step": 357
},
{
"epoch": 4.775919732441472,
"grad_norm": 0.8920648097991943,
"learning_rate": 0.00018433815350389323,
"loss": 3.9213,
"step": 358
},
{
"epoch": 4.789297658862877,
"grad_norm": 1.0252491235733032,
"learning_rate": 0.000184293659621802,
"loss": 3.9118,
"step": 359
},
{
"epoch": 4.802675585284281,
"grad_norm": 1.0382707118988037,
"learning_rate": 0.0001842491657397108,
"loss": 4.0172,
"step": 360
},
{
"epoch": 4.816053511705686,
"grad_norm": 1.07838773727417,
"learning_rate": 0.00018420467185761957,
"loss": 3.8531,
"step": 361
},
{
"epoch": 4.829431438127091,
"grad_norm": 0.9974546432495117,
"learning_rate": 0.00018416017797552838,
"loss": 4.0387,
"step": 362
},
{
"epoch": 4.842809364548495,
"grad_norm": 1.024491548538208,
"learning_rate": 0.00018411568409343717,
"loss": 3.9504,
"step": 363
},
{
"epoch": 4.8561872909699,
"grad_norm": 0.9236369132995605,
"learning_rate": 0.00018407119021134595,
"loss": 3.7119,
"step": 364
},
{
"epoch": 4.869565217391305,
"grad_norm": 0.935644268989563,
"learning_rate": 0.00018402669632925473,
"loss": 4.0077,
"step": 365
},
{
"epoch": 4.882943143812709,
"grad_norm": 0.9328681230545044,
"learning_rate": 0.00018398220244716354,
"loss": 3.9133,
"step": 366
},
{
"epoch": 4.896321070234114,
"grad_norm": 0.9596607089042664,
"learning_rate": 0.0001839377085650723,
"loss": 3.8003,
"step": 367
},
{
"epoch": 4.909698996655519,
"grad_norm": 0.9878052473068237,
"learning_rate": 0.0001838932146829811,
"loss": 3.8805,
"step": 368
},
{
"epoch": 4.923076923076923,
"grad_norm": 1.00381600856781,
"learning_rate": 0.0001838487208008899,
"loss": 4.0264,
"step": 369
},
{
"epoch": 4.936454849498328,
"grad_norm": 1.024754524230957,
"learning_rate": 0.00018380422691879867,
"loss": 3.7291,
"step": 370
},
{
"epoch": 4.949832775919733,
"grad_norm": 0.9670823812484741,
"learning_rate": 0.00018375973303670745,
"loss": 3.9418,
"step": 371
},
{
"epoch": 4.963210702341137,
"grad_norm": 0.9736581444740295,
"learning_rate": 0.00018371523915461626,
"loss": 3.8813,
"step": 372
},
{
"epoch": 4.976588628762542,
"grad_norm": 0.9752672910690308,
"learning_rate": 0.00018367074527252504,
"loss": 3.6717,
"step": 373
},
{
"epoch": 4.989966555183947,
"grad_norm": 1.1268304586410522,
"learning_rate": 0.00018362625139043383,
"loss": 3.9782,
"step": 374
},
{
"epoch": 5.0,
"grad_norm": 1.7933701276779175,
"learning_rate": 0.0001835817575083426,
"loss": 3.001,
"step": 375
},
{
"epoch": 5.013377926421405,
"grad_norm": 0.8035010099411011,
"learning_rate": 0.00018353726362625142,
"loss": 3.7943,
"step": 376
},
{
"epoch": 5.026755852842809,
"grad_norm": 0.8016420006752014,
"learning_rate": 0.00018349276974416017,
"loss": 3.7454,
"step": 377
},
{
"epoch": 5.040133779264214,
"grad_norm": 0.6844643354415894,
"learning_rate": 0.00018344827586206896,
"loss": 3.699,
"step": 378
},
{
"epoch": 5.053511705685619,
"grad_norm": 0.8649943470954895,
"learning_rate": 0.00018340378197997777,
"loss": 3.7197,
"step": 379
},
{
"epoch": 5.066889632107023,
"grad_norm": 0.9685015678405762,
"learning_rate": 0.00018335928809788655,
"loss": 3.6952,
"step": 380
},
{
"epoch": 5.080267558528428,
"grad_norm": 0.8728330135345459,
"learning_rate": 0.00018331479421579533,
"loss": 3.7164,
"step": 381
},
{
"epoch": 5.093645484949833,
"grad_norm": 0.962504506111145,
"learning_rate": 0.0001832703003337041,
"loss": 3.6123,
"step": 382
},
{
"epoch": 5.107023411371237,
"grad_norm": 0.9194462895393372,
"learning_rate": 0.00018322580645161292,
"loss": 3.275,
"step": 383
},
{
"epoch": 5.120401337792642,
"grad_norm": 0.9851329326629639,
"learning_rate": 0.00018318131256952168,
"loss": 3.6222,
"step": 384
},
{
"epoch": 5.133779264214047,
"grad_norm": 1.0702580213546753,
"learning_rate": 0.0001831368186874305,
"loss": 3.8728,
"step": 385
},
{
"epoch": 5.147157190635451,
"grad_norm": 1.3237228393554688,
"learning_rate": 0.00018309232480533927,
"loss": 3.8948,
"step": 386
},
{
"epoch": 5.160535117056856,
"grad_norm": 1.0076218843460083,
"learning_rate": 0.00018304783092324805,
"loss": 3.8894,
"step": 387
},
{
"epoch": 5.173913043478261,
"grad_norm": 1.084722876548767,
"learning_rate": 0.00018300333704115683,
"loss": 3.7398,
"step": 388
},
{
"epoch": 5.187290969899665,
"grad_norm": 0.9112711548805237,
"learning_rate": 0.00018295884315906564,
"loss": 3.5901,
"step": 389
},
{
"epoch": 5.20066889632107,
"grad_norm": 0.9451406002044678,
"learning_rate": 0.00018291434927697443,
"loss": 3.6313,
"step": 390
},
{
"epoch": 5.214046822742475,
"grad_norm": 0.8901047706604004,
"learning_rate": 0.0001828698553948832,
"loss": 3.3191,
"step": 391
},
{
"epoch": 5.2274247491638794,
"grad_norm": 0.9838565587997437,
"learning_rate": 0.000182825361512792,
"loss": 3.882,
"step": 392
},
{
"epoch": 5.240802675585284,
"grad_norm": 0.9839156866073608,
"learning_rate": 0.0001827808676307008,
"loss": 3.6068,
"step": 393
},
{
"epoch": 5.254180602006689,
"grad_norm": 0.9328583478927612,
"learning_rate": 0.00018273637374860956,
"loss": 3.6856,
"step": 394
},
{
"epoch": 5.2675585284280935,
"grad_norm": 0.8705796003341675,
"learning_rate": 0.00018269187986651837,
"loss": 3.7282,
"step": 395
},
{
"epoch": 5.280936454849498,
"grad_norm": 0.9675374031066895,
"learning_rate": 0.00018264738598442715,
"loss": 3.6588,
"step": 396
},
{
"epoch": 5.294314381270903,
"grad_norm": 1.145280361175537,
"learning_rate": 0.00018260289210233593,
"loss": 3.8843,
"step": 397
},
{
"epoch": 5.3076923076923075,
"grad_norm": 0.9769694805145264,
"learning_rate": 0.0001825583982202447,
"loss": 3.7207,
"step": 398
},
{
"epoch": 5.321070234113712,
"grad_norm": 0.9277816414833069,
"learning_rate": 0.00018251390433815352,
"loss": 3.712,
"step": 399
},
{
"epoch": 5.334448160535117,
"grad_norm": 1.1015180349349976,
"learning_rate": 0.0001824694104560623,
"loss": 3.7941,
"step": 400
},
{
"epoch": 5.3478260869565215,
"grad_norm": 1.2234200239181519,
"learning_rate": 0.0001824249165739711,
"loss": 3.6559,
"step": 401
},
{
"epoch": 5.361204013377926,
"grad_norm": 0.9358471035957336,
"learning_rate": 0.00018238042269187987,
"loss": 3.7665,
"step": 402
},
{
"epoch": 5.374581939799331,
"grad_norm": 0.8287034630775452,
"learning_rate": 0.00018233592880978868,
"loss": 3.8265,
"step": 403
},
{
"epoch": 5.3879598662207355,
"grad_norm": 1.0219204425811768,
"learning_rate": 0.00018229143492769743,
"loss": 3.6829,
"step": 404
},
{
"epoch": 5.40133779264214,
"grad_norm": 1.0601041316986084,
"learning_rate": 0.00018224694104560624,
"loss": 3.5879,
"step": 405
},
{
"epoch": 5.414715719063545,
"grad_norm": 1.2221566438674927,
"learning_rate": 0.00018220244716351503,
"loss": 3.6023,
"step": 406
},
{
"epoch": 5.4280936454849495,
"grad_norm": 0.9589087963104248,
"learning_rate": 0.0001821579532814238,
"loss": 3.9109,
"step": 407
},
{
"epoch": 5.441471571906354,
"grad_norm": 1.088295340538025,
"learning_rate": 0.0001821134593993326,
"loss": 3.5695,
"step": 408
},
{
"epoch": 5.454849498327759,
"grad_norm": 1.1284915208816528,
"learning_rate": 0.0001820689655172414,
"loss": 3.6004,
"step": 409
},
{
"epoch": 5.468227424749164,
"grad_norm": 1.0108689069747925,
"learning_rate": 0.00018202447163515018,
"loss": 3.7797,
"step": 410
},
{
"epoch": 5.481605351170568,
"grad_norm": 0.8550918102264404,
"learning_rate": 0.00018197997775305896,
"loss": 3.5758,
"step": 411
},
{
"epoch": 5.494983277591973,
"grad_norm": 0.8765145540237427,
"learning_rate": 0.00018193548387096775,
"loss": 3.7625,
"step": 412
},
{
"epoch": 5.508361204013378,
"grad_norm": 1.0253541469573975,
"learning_rate": 0.00018189098998887656,
"loss": 3.6033,
"step": 413
},
{
"epoch": 5.521739130434782,
"grad_norm": 1.0475622415542603,
"learning_rate": 0.0001818464961067853,
"loss": 3.813,
"step": 414
},
{
"epoch": 5.535117056856187,
"grad_norm": 1.053133249282837,
"learning_rate": 0.00018180200222469412,
"loss": 3.4779,
"step": 415
},
{
"epoch": 5.548494983277592,
"grad_norm": 1.0151216983795166,
"learning_rate": 0.0001817575083426029,
"loss": 3.9038,
"step": 416
},
{
"epoch": 5.561872909698996,
"grad_norm": 1.4666434526443481,
"learning_rate": 0.00018171301446051169,
"loss": 3.4735,
"step": 417
},
{
"epoch": 5.575250836120401,
"grad_norm": 1.1043344736099243,
"learning_rate": 0.00018166852057842047,
"loss": 3.7449,
"step": 418
},
{
"epoch": 5.588628762541806,
"grad_norm": 0.900745153427124,
"learning_rate": 0.00018162402669632925,
"loss": 3.7401,
"step": 419
},
{
"epoch": 5.602006688963211,
"grad_norm": 0.9771101474761963,
"learning_rate": 0.00018157953281423806,
"loss": 3.8328,
"step": 420
},
{
"epoch": 5.615384615384615,
"grad_norm": 0.9099516272544861,
"learning_rate": 0.00018153503893214682,
"loss": 3.6245,
"step": 421
},
{
"epoch": 5.6287625418060205,
"grad_norm": 0.9844585657119751,
"learning_rate": 0.00018149054505005563,
"loss": 3.5776,
"step": 422
},
{
"epoch": 5.642140468227424,
"grad_norm": 1.0481154918670654,
"learning_rate": 0.0001814460511679644,
"loss": 3.6304,
"step": 423
},
{
"epoch": 5.65551839464883,
"grad_norm": 0.9971081614494324,
"learning_rate": 0.0001814015572858732,
"loss": 3.7863,
"step": 424
},
{
"epoch": 5.668896321070234,
"grad_norm": 0.9247872829437256,
"learning_rate": 0.00018135706340378197,
"loss": 3.7319,
"step": 425
},
{
"epoch": 5.682274247491639,
"grad_norm": 0.9895725846290588,
"learning_rate": 0.00018131256952169078,
"loss": 3.78,
"step": 426
},
{
"epoch": 5.695652173913043,
"grad_norm": 1.0847641229629517,
"learning_rate": 0.00018126807563959956,
"loss": 3.8662,
"step": 427
},
{
"epoch": 5.709030100334449,
"grad_norm": 0.986259937286377,
"learning_rate": 0.00018122358175750835,
"loss": 3.5621,
"step": 428
},
{
"epoch": 5.722408026755852,
"grad_norm": 0.9166681170463562,
"learning_rate": 0.00018117908787541713,
"loss": 3.6153,
"step": 429
},
{
"epoch": 5.735785953177258,
"grad_norm": 1.1331177949905396,
"learning_rate": 0.00018113459399332594,
"loss": 3.5976,
"step": 430
},
{
"epoch": 5.749163879598662,
"grad_norm": 0.8743540644645691,
"learning_rate": 0.0001810901001112347,
"loss": 3.2511,
"step": 431
},
{
"epoch": 5.762541806020067,
"grad_norm": 1.0700207948684692,
"learning_rate": 0.0001810456062291435,
"loss": 3.7634,
"step": 432
},
{
"epoch": 5.775919732441472,
"grad_norm": 0.9412694573402405,
"learning_rate": 0.00018100111234705229,
"loss": 3.6264,
"step": 433
},
{
"epoch": 5.789297658862877,
"grad_norm": 1.0398496389389038,
"learning_rate": 0.00018095661846496107,
"loss": 3.744,
"step": 434
},
{
"epoch": 5.802675585284281,
"grad_norm": 0.9605004787445068,
"learning_rate": 0.00018091212458286985,
"loss": 3.5532,
"step": 435
},
{
"epoch": 5.816053511705686,
"grad_norm": 1.0449095964431763,
"learning_rate": 0.00018086763070077866,
"loss": 4.0611,
"step": 436
},
{
"epoch": 5.829431438127091,
"grad_norm": 0.9342606663703918,
"learning_rate": 0.00018082313681868744,
"loss": 3.9957,
"step": 437
},
{
"epoch": 5.842809364548495,
"grad_norm": 0.9687880873680115,
"learning_rate": 0.00018077864293659622,
"loss": 3.9299,
"step": 438
},
{
"epoch": 5.8561872909699,
"grad_norm": 1.1390576362609863,
"learning_rate": 0.000180734149054505,
"loss": 3.6552,
"step": 439
},
{
"epoch": 5.869565217391305,
"grad_norm": 0.9280988574028015,
"learning_rate": 0.00018068965517241382,
"loss": 3.7828,
"step": 440
},
{
"epoch": 5.882943143812709,
"grad_norm": 1.2928193807601929,
"learning_rate": 0.00018064516129032257,
"loss": 3.6292,
"step": 441
},
{
"epoch": 5.896321070234114,
"grad_norm": 1.0959875583648682,
"learning_rate": 0.00018060066740823138,
"loss": 3.4293,
"step": 442
},
{
"epoch": 5.909698996655519,
"grad_norm": 1.0713289976119995,
"learning_rate": 0.00018055617352614016,
"loss": 3.7767,
"step": 443
},
{
"epoch": 5.923076923076923,
"grad_norm": 0.9309440851211548,
"learning_rate": 0.00018051167964404895,
"loss": 3.5473,
"step": 444
},
{
"epoch": 5.936454849498328,
"grad_norm": 1.0999056100845337,
"learning_rate": 0.00018046718576195773,
"loss": 3.9694,
"step": 445
},
{
"epoch": 5.949832775919733,
"grad_norm": 1.1073781251907349,
"learning_rate": 0.00018042269187986654,
"loss": 3.6882,
"step": 446
},
{
"epoch": 5.963210702341137,
"grad_norm": 1.0430257320404053,
"learning_rate": 0.00018037819799777532,
"loss": 3.4009,
"step": 447
},
{
"epoch": 5.976588628762542,
"grad_norm": 1.1132690906524658,
"learning_rate": 0.0001803337041156841,
"loss": 3.8832,
"step": 448
},
{
"epoch": 5.989966555183947,
"grad_norm": 1.0147771835327148,
"learning_rate": 0.00018028921023359289,
"loss": 3.7117,
"step": 449
},
{
"epoch": 6.0,
"grad_norm": 1.458959698677063,
"learning_rate": 0.0001802447163515017,
"loss": 3.5745,
"step": 450
},
{
"epoch": 6.013377926421405,
"grad_norm": 0.8363592028617859,
"learning_rate": 0.00018020022246941045,
"loss": 3.5835,
"step": 451
},
{
"epoch": 6.026755852842809,
"grad_norm": 0.8071937561035156,
"learning_rate": 0.00018015572858731926,
"loss": 3.5923,
"step": 452
},
{
"epoch": 6.040133779264214,
"grad_norm": 0.7746313214302063,
"learning_rate": 0.00018011123470522804,
"loss": 3.5688,
"step": 453
},
{
"epoch": 6.053511705685619,
"grad_norm": 0.689179539680481,
"learning_rate": 0.00018006674082313682,
"loss": 3.412,
"step": 454
},
{
"epoch": 6.066889632107023,
"grad_norm": 0.8438050746917725,
"learning_rate": 0.0001800222469410456,
"loss": 3.403,
"step": 455
},
{
"epoch": 6.080267558528428,
"grad_norm": 0.7670062780380249,
"learning_rate": 0.00017997775305895442,
"loss": 3.5029,
"step": 456
},
{
"epoch": 6.093645484949833,
"grad_norm": 0.8185870051383972,
"learning_rate": 0.0001799332591768632,
"loss": 3.4584,
"step": 457
},
{
"epoch": 6.107023411371237,
"grad_norm": 0.9618543386459351,
"learning_rate": 0.00017988876529477198,
"loss": 3.6538,
"step": 458
},
{
"epoch": 6.120401337792642,
"grad_norm": 0.959724485874176,
"learning_rate": 0.00017984427141268076,
"loss": 3.5284,
"step": 459
},
{
"epoch": 6.133779264214047,
"grad_norm": 0.8044765591621399,
"learning_rate": 0.00017979977753058955,
"loss": 3.2198,
"step": 460
},
{
"epoch": 6.147157190635451,
"grad_norm": 0.8287092447280884,
"learning_rate": 0.00017975528364849833,
"loss": 3.4977,
"step": 461
},
{
"epoch": 6.160535117056856,
"grad_norm": 0.8855329155921936,
"learning_rate": 0.0001797107897664071,
"loss": 3.5008,
"step": 462
},
{
"epoch": 6.173913043478261,
"grad_norm": 0.8839483857154846,
"learning_rate": 0.00017966629588431592,
"loss": 3.6135,
"step": 463
},
{
"epoch": 6.187290969899665,
"grad_norm": 0.963446319103241,
"learning_rate": 0.0001796218020022247,
"loss": 3.6156,
"step": 464
},
{
"epoch": 6.20066889632107,
"grad_norm": 0.896743655204773,
"learning_rate": 0.00017957730812013348,
"loss": 3.6623,
"step": 465
},
{
"epoch": 6.214046822742475,
"grad_norm": 0.9268617033958435,
"learning_rate": 0.00017953281423804227,
"loss": 3.4343,
"step": 466
},
{
"epoch": 6.2274247491638794,
"grad_norm": 0.8335449695587158,
"learning_rate": 0.00017948832035595108,
"loss": 3.5716,
"step": 467
},
{
"epoch": 6.240802675585284,
"grad_norm": 0.7771849036216736,
"learning_rate": 0.00017944382647385983,
"loss": 3.5191,
"step": 468
},
{
"epoch": 6.254180602006689,
"grad_norm": 0.9157488346099854,
"learning_rate": 0.00017939933259176864,
"loss": 3.5583,
"step": 469
},
{
"epoch": 6.2675585284280935,
"grad_norm": 0.9348477721214294,
"learning_rate": 0.00017935483870967742,
"loss": 3.3137,
"step": 470
},
{
"epoch": 6.280936454849498,
"grad_norm": 0.8791135549545288,
"learning_rate": 0.0001793103448275862,
"loss": 3.5111,
"step": 471
},
{
"epoch": 6.294314381270903,
"grad_norm": 0.9963672757148743,
"learning_rate": 0.000179265850945495,
"loss": 3.7518,
"step": 472
},
{
"epoch": 6.3076923076923075,
"grad_norm": 0.9291539192199707,
"learning_rate": 0.0001792213570634038,
"loss": 3.4524,
"step": 473
},
{
"epoch": 6.321070234113712,
"grad_norm": 0.9349279403686523,
"learning_rate": 0.00017917686318131258,
"loss": 3.4753,
"step": 474
},
{
"epoch": 6.334448160535117,
"grad_norm": 0.8984476327896118,
"learning_rate": 0.00017913236929922136,
"loss": 3.7325,
"step": 475
},
{
"epoch": 6.3478260869565215,
"grad_norm": 0.8452139496803284,
"learning_rate": 0.00017908787541713015,
"loss": 3.8021,
"step": 476
},
{
"epoch": 6.361204013377926,
"grad_norm": 0.9418376088142395,
"learning_rate": 0.00017904338153503895,
"loss": 3.7426,
"step": 477
},
{
"epoch": 6.374581939799331,
"grad_norm": 1.0661097764968872,
"learning_rate": 0.0001789988876529477,
"loss": 3.7556,
"step": 478
},
{
"epoch": 6.3879598662207355,
"grad_norm": 0.9645984768867493,
"learning_rate": 0.00017895439377085652,
"loss": 3.3353,
"step": 479
},
{
"epoch": 6.40133779264214,
"grad_norm": 0.9243470430374146,
"learning_rate": 0.0001789098998887653,
"loss": 3.5729,
"step": 480
},
{
"epoch": 6.414715719063545,
"grad_norm": 0.885061502456665,
"learning_rate": 0.00017886540600667408,
"loss": 3.5699,
"step": 481
},
{
"epoch": 6.4280936454849495,
"grad_norm": 0.9025402069091797,
"learning_rate": 0.00017882091212458287,
"loss": 3.4532,
"step": 482
},
{
"epoch": 6.441471571906354,
"grad_norm": 0.9760842323303223,
"learning_rate": 0.00017877641824249168,
"loss": 3.7222,
"step": 483
},
{
"epoch": 6.454849498327759,
"grad_norm": 1.2709609270095825,
"learning_rate": 0.00017873192436040046,
"loss": 3.756,
"step": 484
},
{
"epoch": 6.468227424749164,
"grad_norm": 0.904513955116272,
"learning_rate": 0.00017868743047830924,
"loss": 3.657,
"step": 485
},
{
"epoch": 6.481605351170568,
"grad_norm": 1.158915400505066,
"learning_rate": 0.00017864293659621802,
"loss": 3.5897,
"step": 486
},
{
"epoch": 6.494983277591973,
"grad_norm": 0.9457879066467285,
"learning_rate": 0.00017859844271412683,
"loss": 3.4394,
"step": 487
},
{
"epoch": 6.508361204013378,
"grad_norm": 1.04762601852417,
"learning_rate": 0.0001785539488320356,
"loss": 3.39,
"step": 488
},
{
"epoch": 6.521739130434782,
"grad_norm": 0.9370948076248169,
"learning_rate": 0.0001785094549499444,
"loss": 3.4816,
"step": 489
},
{
"epoch": 6.535117056856187,
"grad_norm": 1.1307988166809082,
"learning_rate": 0.00017846496106785318,
"loss": 3.5079,
"step": 490
},
{
"epoch": 6.548494983277592,
"grad_norm": 1.0025054216384888,
"learning_rate": 0.00017842046718576196,
"loss": 3.3447,
"step": 491
},
{
"epoch": 6.561872909698996,
"grad_norm": 0.9893412590026855,
"learning_rate": 0.00017837597330367074,
"loss": 3.4698,
"step": 492
},
{
"epoch": 6.575250836120401,
"grad_norm": 1.0476347208023071,
"learning_rate": 0.00017833147942157955,
"loss": 3.4955,
"step": 493
},
{
"epoch": 6.588628762541806,
"grad_norm": 1.0508891344070435,
"learning_rate": 0.00017828698553948834,
"loss": 3.6568,
"step": 494
},
{
"epoch": 6.602006688963211,
"grad_norm": 1.0397465229034424,
"learning_rate": 0.00017824249165739712,
"loss": 3.5087,
"step": 495
},
{
"epoch": 6.615384615384615,
"grad_norm": 0.9884181022644043,
"learning_rate": 0.0001781979977753059,
"loss": 3.6778,
"step": 496
},
{
"epoch": 6.6287625418060205,
"grad_norm": 1.1187562942504883,
"learning_rate": 0.0001781535038932147,
"loss": 3.4345,
"step": 497
},
{
"epoch": 6.642140468227424,
"grad_norm": 1.1133880615234375,
"learning_rate": 0.00017810901001112347,
"loss": 3.4959,
"step": 498
},
{
"epoch": 6.65551839464883,
"grad_norm": 0.9368589520454407,
"learning_rate": 0.00017806451612903228,
"loss": 3.628,
"step": 499
},
{
"epoch": 6.668896321070234,
"grad_norm": 1.0427212715148926,
"learning_rate": 0.00017802002224694106,
"loss": 3.6073,
"step": 500
},
{
"epoch": 6.682274247491639,
"grad_norm": 0.9564261436462402,
"learning_rate": 0.00017797552836484984,
"loss": 3.4406,
"step": 501
},
{
"epoch": 6.695652173913043,
"grad_norm": 0.9754629135131836,
"learning_rate": 0.00017793103448275862,
"loss": 3.7025,
"step": 502
},
{
"epoch": 6.709030100334449,
"grad_norm": 0.9669683575630188,
"learning_rate": 0.0001778865406006674,
"loss": 3.5872,
"step": 503
},
{
"epoch": 6.722408026755852,
"grad_norm": 0.959619402885437,
"learning_rate": 0.00017784204671857621,
"loss": 3.5124,
"step": 504
},
{
"epoch": 6.735785953177258,
"grad_norm": 0.981737494468689,
"learning_rate": 0.00017779755283648497,
"loss": 3.4038,
"step": 505
},
{
"epoch": 6.749163879598662,
"grad_norm": 0.9372640252113342,
"learning_rate": 0.00017775305895439378,
"loss": 3.6725,
"step": 506
},
{
"epoch": 6.762541806020067,
"grad_norm": 0.947066605091095,
"learning_rate": 0.00017770856507230256,
"loss": 3.807,
"step": 507
},
{
"epoch": 6.775919732441472,
"grad_norm": 0.7659755349159241,
"learning_rate": 0.00017766407119021134,
"loss": 3.0411,
"step": 508
},
{
"epoch": 6.789297658862877,
"grad_norm": 1.0432168245315552,
"learning_rate": 0.00017761957730812013,
"loss": 3.6259,
"step": 509
},
{
"epoch": 6.802675585284281,
"grad_norm": 1.0104693174362183,
"learning_rate": 0.00017757508342602894,
"loss": 3.5971,
"step": 510
},
{
"epoch": 6.816053511705686,
"grad_norm": 0.9007440805435181,
"learning_rate": 0.00017753058954393772,
"loss": 3.5894,
"step": 511
},
{
"epoch": 6.829431438127091,
"grad_norm": 0.8829946517944336,
"learning_rate": 0.0001774860956618465,
"loss": 3.6094,
"step": 512
},
{
"epoch": 6.842809364548495,
"grad_norm": 0.9823127388954163,
"learning_rate": 0.00017744160177975528,
"loss": 3.6568,
"step": 513
},
{
"epoch": 6.8561872909699,
"grad_norm": 1.0247899293899536,
"learning_rate": 0.0001773971078976641,
"loss": 3.259,
"step": 514
},
{
"epoch": 6.869565217391305,
"grad_norm": 0.9435336589813232,
"learning_rate": 0.00017735261401557285,
"loss": 3.4978,
"step": 515
},
{
"epoch": 6.882943143812709,
"grad_norm": 1.0135424137115479,
"learning_rate": 0.00017730812013348166,
"loss": 3.4765,
"step": 516
},
{
"epoch": 6.896321070234114,
"grad_norm": 1.1327738761901855,
"learning_rate": 0.00017726362625139044,
"loss": 3.6927,
"step": 517
},
{
"epoch": 6.909698996655519,
"grad_norm": 0.9335159659385681,
"learning_rate": 0.00017721913236929922,
"loss": 3.7278,
"step": 518
},
{
"epoch": 6.923076923076923,
"grad_norm": 0.8229056000709534,
"learning_rate": 0.000177174638487208,
"loss": 3.549,
"step": 519
},
{
"epoch": 6.936454849498328,
"grad_norm": 0.8996124267578125,
"learning_rate": 0.00017713014460511681,
"loss": 3.5093,
"step": 520
},
{
"epoch": 6.949832775919733,
"grad_norm": 0.9984102845191956,
"learning_rate": 0.0001770856507230256,
"loss": 3.4772,
"step": 521
},
{
"epoch": 6.963210702341137,
"grad_norm": 0.9136049747467041,
"learning_rate": 0.00017704115684093438,
"loss": 3.6487,
"step": 522
},
{
"epoch": 6.976588628762542,
"grad_norm": 0.9707899689674377,
"learning_rate": 0.00017699666295884316,
"loss": 3.7471,
"step": 523
},
{
"epoch": 6.989966555183947,
"grad_norm": 0.919865071773529,
"learning_rate": 0.00017695216907675197,
"loss": 3.8013,
"step": 524
},
{
"epoch": 7.0,
"grad_norm": 1.125541090965271,
"learning_rate": 0.00017690767519466073,
"loss": 3.4869,
"step": 525
},
{
"epoch": 7.013377926421405,
"grad_norm": 0.8512810468673706,
"learning_rate": 0.00017686318131256954,
"loss": 3.3334,
"step": 526
},
{
"epoch": 7.026755852842809,
"grad_norm": 0.8450623750686646,
"learning_rate": 0.00017681868743047832,
"loss": 3.5769,
"step": 527
},
{
"epoch": 7.040133779264214,
"grad_norm": 0.8526298403739929,
"learning_rate": 0.0001767741935483871,
"loss": 3.4546,
"step": 528
},
{
"epoch": 7.053511705685619,
"grad_norm": 0.75905442237854,
"learning_rate": 0.00017672969966629588,
"loss": 3.4228,
"step": 529
},
{
"epoch": 7.066889632107023,
"grad_norm": 0.8442811965942383,
"learning_rate": 0.0001766852057842047,
"loss": 3.5766,
"step": 530
},
{
"epoch": 7.080267558528428,
"grad_norm": 0.9584814310073853,
"learning_rate": 0.00017664071190211347,
"loss": 3.5312,
"step": 531
},
{
"epoch": 7.093645484949833,
"grad_norm": 0.9741052985191345,
"learning_rate": 0.00017659621802002226,
"loss": 3.4877,
"step": 532
},
{
"epoch": 7.107023411371237,
"grad_norm": 0.8638135194778442,
"learning_rate": 0.00017655172413793104,
"loss": 3.5701,
"step": 533
},
{
"epoch": 7.120401337792642,
"grad_norm": 1.0128440856933594,
"learning_rate": 0.00017650723025583985,
"loss": 3.3629,
"step": 534
},
{
"epoch": 7.133779264214047,
"grad_norm": 0.9763593673706055,
"learning_rate": 0.0001764627363737486,
"loss": 3.5651,
"step": 535
},
{
"epoch": 7.147157190635451,
"grad_norm": 0.8706293702125549,
"learning_rate": 0.00017641824249165741,
"loss": 3.454,
"step": 536
},
{
"epoch": 7.160535117056856,
"grad_norm": 0.9227468967437744,
"learning_rate": 0.0001763737486095662,
"loss": 3.5528,
"step": 537
},
{
"epoch": 7.173913043478261,
"grad_norm": 0.7493206262588501,
"learning_rate": 0.00017632925472747498,
"loss": 3.2662,
"step": 538
},
{
"epoch": 7.187290969899665,
"grad_norm": 0.8414123058319092,
"learning_rate": 0.00017628476084538376,
"loss": 3.3864,
"step": 539
},
{
"epoch": 7.20066889632107,
"grad_norm": 0.8352764248847961,
"learning_rate": 0.00017624026696329257,
"loss": 3.2407,
"step": 540
},
{
"epoch": 7.214046822742475,
"grad_norm": 0.7413480281829834,
"learning_rate": 0.00017619577308120135,
"loss": 3.3989,
"step": 541
},
{
"epoch": 7.2274247491638794,
"grad_norm": 0.7661281228065491,
"learning_rate": 0.00017615127919911014,
"loss": 3.3792,
"step": 542
},
{
"epoch": 7.240802675585284,
"grad_norm": 0.86900395154953,
"learning_rate": 0.00017610678531701892,
"loss": 3.2191,
"step": 543
},
{
"epoch": 7.254180602006689,
"grad_norm": 0.8536344170570374,
"learning_rate": 0.0001760622914349277,
"loss": 3.3366,
"step": 544
},
{
"epoch": 7.2675585284280935,
"grad_norm": 0.8729544878005981,
"learning_rate": 0.00017601779755283648,
"loss": 3.4976,
"step": 545
},
{
"epoch": 7.280936454849498,
"grad_norm": 0.8263023495674133,
"learning_rate": 0.00017597330367074526,
"loss": 3.4102,
"step": 546
},
{
"epoch": 7.294314381270903,
"grad_norm": 0.748373806476593,
"learning_rate": 0.00017592880978865407,
"loss": 3.332,
"step": 547
},
{
"epoch": 7.3076923076923075,
"grad_norm": 0.7606791853904724,
"learning_rate": 0.00017588431590656286,
"loss": 3.361,
"step": 548
},
{
"epoch": 7.321070234113712,
"grad_norm": 0.9155070781707764,
"learning_rate": 0.00017583982202447164,
"loss": 3.657,
"step": 549
},
{
"epoch": 7.334448160535117,
"grad_norm": 0.7440597414970398,
"learning_rate": 0.00017579532814238042,
"loss": 3.335,
"step": 550
},
{
"epoch": 7.3478260869565215,
"grad_norm": 0.8781002759933472,
"learning_rate": 0.00017575083426028923,
"loss": 3.5579,
"step": 551
},
{
"epoch": 7.361204013377926,
"grad_norm": 0.7886701822280884,
"learning_rate": 0.00017570634037819799,
"loss": 3.4636,
"step": 552
},
{
"epoch": 7.374581939799331,
"grad_norm": 0.8931376934051514,
"learning_rate": 0.0001756618464961068,
"loss": 3.4375,
"step": 553
},
{
"epoch": 7.3879598662207355,
"grad_norm": 0.7599623799324036,
"learning_rate": 0.00017561735261401558,
"loss": 3.6551,
"step": 554
},
{
"epoch": 7.40133779264214,
"grad_norm": 0.7692762613296509,
"learning_rate": 0.00017557285873192436,
"loss": 3.6373,
"step": 555
},
{
"epoch": 7.414715719063545,
"grad_norm": 0.8861828446388245,
"learning_rate": 0.00017552836484983314,
"loss": 3.4791,
"step": 556
},
{
"epoch": 7.4280936454849495,
"grad_norm": 0.9560372829437256,
"learning_rate": 0.00017548387096774195,
"loss": 3.7291,
"step": 557
},
{
"epoch": 7.441471571906354,
"grad_norm": 0.8745344281196594,
"learning_rate": 0.00017543937708565073,
"loss": 3.3071,
"step": 558
},
{
"epoch": 7.454849498327759,
"grad_norm": 0.8178285360336304,
"learning_rate": 0.00017539488320355952,
"loss": 3.4738,
"step": 559
},
{
"epoch": 7.468227424749164,
"grad_norm": 0.8611259460449219,
"learning_rate": 0.0001753503893214683,
"loss": 3.25,
"step": 560
},
{
"epoch": 7.481605351170568,
"grad_norm": 0.8623505234718323,
"learning_rate": 0.0001753058954393771,
"loss": 3.3701,
"step": 561
},
{
"epoch": 7.494983277591973,
"grad_norm": 0.76930171251297,
"learning_rate": 0.00017526140155728586,
"loss": 3.332,
"step": 562
},
{
"epoch": 7.508361204013378,
"grad_norm": 0.8986758589744568,
"learning_rate": 0.00017521690767519467,
"loss": 3.3927,
"step": 563
},
{
"epoch": 7.521739130434782,
"grad_norm": 0.9844257831573486,
"learning_rate": 0.00017517241379310346,
"loss": 3.5664,
"step": 564
},
{
"epoch": 7.535117056856187,
"grad_norm": 0.983921229839325,
"learning_rate": 0.00017512791991101224,
"loss": 3.3888,
"step": 565
},
{
"epoch": 7.548494983277592,
"grad_norm": 0.8052308559417725,
"learning_rate": 0.00017508342602892102,
"loss": 3.4809,
"step": 566
},
{
"epoch": 7.561872909698996,
"grad_norm": 0.7996425032615662,
"learning_rate": 0.00017503893214682983,
"loss": 3.4793,
"step": 567
},
{
"epoch": 7.575250836120401,
"grad_norm": 0.8453391194343567,
"learning_rate": 0.0001749944382647386,
"loss": 3.3199,
"step": 568
},
{
"epoch": 7.588628762541806,
"grad_norm": 0.8720147013664246,
"learning_rate": 0.0001749499443826474,
"loss": 3.5612,
"step": 569
},
{
"epoch": 7.602006688963211,
"grad_norm": 0.9093672633171082,
"learning_rate": 0.00017490545050055618,
"loss": 3.0509,
"step": 570
},
{
"epoch": 7.615384615384615,
"grad_norm": 0.8936579823493958,
"learning_rate": 0.000174860956618465,
"loss": 3.4408,
"step": 571
},
{
"epoch": 7.6287625418060205,
"grad_norm": 0.7683162689208984,
"learning_rate": 0.00017481646273637374,
"loss": 3.3536,
"step": 572
},
{
"epoch": 7.642140468227424,
"grad_norm": 0.7943581342697144,
"learning_rate": 0.00017477196885428255,
"loss": 3.4542,
"step": 573
},
{
"epoch": 7.65551839464883,
"grad_norm": 0.8183353543281555,
"learning_rate": 0.00017472747497219133,
"loss": 3.603,
"step": 574
},
{
"epoch": 7.668896321070234,
"grad_norm": 0.7816463708877563,
"learning_rate": 0.00017468298109010012,
"loss": 3.7388,
"step": 575
},
{
"epoch": 7.682274247491639,
"grad_norm": 0.8167930245399475,
"learning_rate": 0.0001746384872080089,
"loss": 3.7743,
"step": 576
},
{
"epoch": 7.695652173913043,
"grad_norm": 0.832392156124115,
"learning_rate": 0.0001745939933259177,
"loss": 3.7488,
"step": 577
},
{
"epoch": 7.709030100334449,
"grad_norm": 0.9362333416938782,
"learning_rate": 0.0001745494994438265,
"loss": 3.5722,
"step": 578
},
{
"epoch": 7.722408026755852,
"grad_norm": 1.0247248411178589,
"learning_rate": 0.00017450500556173527,
"loss": 3.2048,
"step": 579
},
{
"epoch": 7.735785953177258,
"grad_norm": 0.8833767175674438,
"learning_rate": 0.00017446051167964406,
"loss": 3.389,
"step": 580
},
{
"epoch": 7.749163879598662,
"grad_norm": 0.8344758749008179,
"learning_rate": 0.00017441601779755287,
"loss": 3.5264,
"step": 581
},
{
"epoch": 7.762541806020067,
"grad_norm": 0.9771448373794556,
"learning_rate": 0.00017437152391546162,
"loss": 3.041,
"step": 582
},
{
"epoch": 7.775919732441472,
"grad_norm": 0.8279567956924438,
"learning_rate": 0.00017432703003337043,
"loss": 3.6273,
"step": 583
},
{
"epoch": 7.789297658862877,
"grad_norm": 0.957206130027771,
"learning_rate": 0.0001742825361512792,
"loss": 3.3831,
"step": 584
},
{
"epoch": 7.802675585284281,
"grad_norm": 0.860619843006134,
"learning_rate": 0.000174238042269188,
"loss": 3.4566,
"step": 585
},
{
"epoch": 7.816053511705686,
"grad_norm": 0.8725448846817017,
"learning_rate": 0.00017419354838709678,
"loss": 3.6594,
"step": 586
},
{
"epoch": 7.829431438127091,
"grad_norm": 0.8343111276626587,
"learning_rate": 0.00017414905450500556,
"loss": 3.3423,
"step": 587
},
{
"epoch": 7.842809364548495,
"grad_norm": 0.9043267965316772,
"learning_rate": 0.00017410456062291437,
"loss": 3.221,
"step": 588
},
{
"epoch": 7.8561872909699,
"grad_norm": 0.9563114643096924,
"learning_rate": 0.00017406006674082312,
"loss": 3.5143,
"step": 589
},
{
"epoch": 7.869565217391305,
"grad_norm": 0.9726302027702332,
"learning_rate": 0.00017401557285873193,
"loss": 3.4373,
"step": 590
},
{
"epoch": 7.882943143812709,
"grad_norm": 0.9203178882598877,
"learning_rate": 0.00017397107897664072,
"loss": 3.6014,
"step": 591
},
{
"epoch": 7.896321070234114,
"grad_norm": 0.9120233654975891,
"learning_rate": 0.0001739265850945495,
"loss": 3.1429,
"step": 592
},
{
"epoch": 7.909698996655519,
"grad_norm": 0.7576518058776855,
"learning_rate": 0.00017388209121245828,
"loss": 3.2065,
"step": 593
},
{
"epoch": 7.923076923076923,
"grad_norm": 0.9629240036010742,
"learning_rate": 0.0001738375973303671,
"loss": 3.4788,
"step": 594
},
{
"epoch": 7.936454849498328,
"grad_norm": 0.8390881419181824,
"learning_rate": 0.00017379310344827587,
"loss": 3.2857,
"step": 595
},
{
"epoch": 7.949832775919733,
"grad_norm": 0.8708979487419128,
"learning_rate": 0.00017374860956618466,
"loss": 3.3321,
"step": 596
},
{
"epoch": 7.963210702341137,
"grad_norm": 0.7076835632324219,
"learning_rate": 0.00017370411568409344,
"loss": 3.5905,
"step": 597
},
{
"epoch": 7.976588628762542,
"grad_norm": 1.016526222229004,
"learning_rate": 0.00017365962180200225,
"loss": 3.4053,
"step": 598
},
{
"epoch": 7.989966555183947,
"grad_norm": 0.7592278718948364,
"learning_rate": 0.000173615127919911,
"loss": 3.3968,
"step": 599
},
{
"epoch": 8.0,
"grad_norm": 1.0106462240219116,
"learning_rate": 0.0001735706340378198,
"loss": 3.3722,
"step": 600
},
{
"epoch": 8.013377926421406,
"grad_norm": 0.740808367729187,
"learning_rate": 0.0001735261401557286,
"loss": 3.394,
"step": 601
},
{
"epoch": 8.02675585284281,
"grad_norm": 0.6732498407363892,
"learning_rate": 0.00017348164627363738,
"loss": 3.297,
"step": 602
},
{
"epoch": 8.040133779264215,
"grad_norm": 0.8319197297096252,
"learning_rate": 0.00017343715239154616,
"loss": 3.2898,
"step": 603
},
{
"epoch": 8.053511705685619,
"grad_norm": 0.7834349870681763,
"learning_rate": 0.00017339265850945497,
"loss": 3.2341,
"step": 604
},
{
"epoch": 8.066889632107024,
"grad_norm": 0.705737292766571,
"learning_rate": 0.00017334816462736375,
"loss": 3.3429,
"step": 605
},
{
"epoch": 8.080267558528428,
"grad_norm": 0.8270958065986633,
"learning_rate": 0.00017330367074527253,
"loss": 3.0458,
"step": 606
},
{
"epoch": 8.093645484949834,
"grad_norm": 0.7254801392555237,
"learning_rate": 0.00017325917686318132,
"loss": 3.5143,
"step": 607
},
{
"epoch": 8.107023411371237,
"grad_norm": 0.8450751900672913,
"learning_rate": 0.00017321468298109013,
"loss": 3.1507,
"step": 608
},
{
"epoch": 8.120401337792643,
"grad_norm": 0.7936837673187256,
"learning_rate": 0.00017317018909899888,
"loss": 3.3979,
"step": 609
},
{
"epoch": 8.133779264214047,
"grad_norm": 0.6496401429176331,
"learning_rate": 0.0001731256952169077,
"loss": 3.3613,
"step": 610
},
{
"epoch": 8.147157190635452,
"grad_norm": 0.8721235990524292,
"learning_rate": 0.00017308120133481647,
"loss": 3.4299,
"step": 611
},
{
"epoch": 8.160535117056856,
"grad_norm": 0.7671874761581421,
"learning_rate": 0.00017303670745272525,
"loss": 3.3333,
"step": 612
},
{
"epoch": 8.173913043478262,
"grad_norm": 0.6427144408226013,
"learning_rate": 0.00017299221357063404,
"loss": 3.3304,
"step": 613
},
{
"epoch": 8.187290969899665,
"grad_norm": 0.7999966144561768,
"learning_rate": 0.00017294771968854285,
"loss": 3.4016,
"step": 614
},
{
"epoch": 8.200668896321071,
"grad_norm": 0.8216206431388855,
"learning_rate": 0.00017290322580645163,
"loss": 3.1724,
"step": 615
},
{
"epoch": 8.214046822742475,
"grad_norm": 0.7364024519920349,
"learning_rate": 0.0001728587319243604,
"loss": 3.34,
"step": 616
},
{
"epoch": 8.22742474916388,
"grad_norm": 0.7688239812850952,
"learning_rate": 0.0001728142380422692,
"loss": 3.2114,
"step": 617
},
{
"epoch": 8.240802675585284,
"grad_norm": 0.8786870837211609,
"learning_rate": 0.000172769744160178,
"loss": 3.378,
"step": 618
},
{
"epoch": 8.25418060200669,
"grad_norm": 0.9048855900764465,
"learning_rate": 0.00017272525027808676,
"loss": 3.1801,
"step": 619
},
{
"epoch": 8.267558528428093,
"grad_norm": 0.657189130783081,
"learning_rate": 0.00017268075639599557,
"loss": 3.1389,
"step": 620
},
{
"epoch": 8.280936454849499,
"grad_norm": 0.8015987873077393,
"learning_rate": 0.00017263626251390435,
"loss": 3.4621,
"step": 621
},
{
"epoch": 8.294314381270903,
"grad_norm": 0.8232793807983398,
"learning_rate": 0.00017259176863181313,
"loss": 3.1763,
"step": 622
},
{
"epoch": 8.307692307692308,
"grad_norm": 0.7447130680084229,
"learning_rate": 0.00017254727474972192,
"loss": 3.1266,
"step": 623
},
{
"epoch": 8.321070234113712,
"grad_norm": 0.7649840116500854,
"learning_rate": 0.00017250278086763072,
"loss": 3.1959,
"step": 624
},
{
"epoch": 8.334448160535118,
"grad_norm": 0.7119699120521545,
"learning_rate": 0.0001724582869855395,
"loss": 3.5626,
"step": 625
},
{
"epoch": 8.347826086956522,
"grad_norm": 0.8238518834114075,
"learning_rate": 0.00017241379310344826,
"loss": 3.1873,
"step": 626
},
{
"epoch": 8.361204013377927,
"grad_norm": 0.8248497843742371,
"learning_rate": 0.00017236929922135707,
"loss": 3.5686,
"step": 627
},
{
"epoch": 8.37458193979933,
"grad_norm": 0.8704475164413452,
"learning_rate": 0.00017232480533926585,
"loss": 3.3195,
"step": 628
},
{
"epoch": 8.387959866220736,
"grad_norm": 0.8160929083824158,
"learning_rate": 0.00017228031145717464,
"loss": 3.3808,
"step": 629
},
{
"epoch": 8.40133779264214,
"grad_norm": 0.8537085652351379,
"learning_rate": 0.00017223581757508342,
"loss": 3.3638,
"step": 630
},
{
"epoch": 8.414715719063546,
"grad_norm": 0.876519501209259,
"learning_rate": 0.00017219132369299223,
"loss": 3.2019,
"step": 631
},
{
"epoch": 8.42809364548495,
"grad_norm": 0.6573703289031982,
"learning_rate": 0.000172146829810901,
"loss": 3.4998,
"step": 632
},
{
"epoch": 8.441471571906355,
"grad_norm": 0.8822937607765198,
"learning_rate": 0.0001721023359288098,
"loss": 3.4281,
"step": 633
},
{
"epoch": 8.454849498327759,
"grad_norm": 0.764872670173645,
"learning_rate": 0.00017205784204671858,
"loss": 3.3693,
"step": 634
},
{
"epoch": 8.468227424749164,
"grad_norm": 0.7492384910583496,
"learning_rate": 0.00017201334816462739,
"loss": 3.5672,
"step": 635
},
{
"epoch": 8.481605351170568,
"grad_norm": 0.8037416934967041,
"learning_rate": 0.00017196885428253614,
"loss": 3.3804,
"step": 636
},
{
"epoch": 8.494983277591974,
"grad_norm": 0.8380945324897766,
"learning_rate": 0.00017192436040044495,
"loss": 3.2272,
"step": 637
},
{
"epoch": 8.508361204013378,
"grad_norm": 0.8467932939529419,
"learning_rate": 0.00017187986651835373,
"loss": 3.2649,
"step": 638
},
{
"epoch": 8.521739130434783,
"grad_norm": 0.751542866230011,
"learning_rate": 0.00017183537263626252,
"loss": 3.4135,
"step": 639
},
{
"epoch": 8.535117056856187,
"grad_norm": 0.7618190050125122,
"learning_rate": 0.0001717908787541713,
"loss": 3.483,
"step": 640
},
{
"epoch": 8.548494983277592,
"grad_norm": 0.9661890864372253,
"learning_rate": 0.0001717463848720801,
"loss": 3.2201,
"step": 641
},
{
"epoch": 8.561872909698996,
"grad_norm": 0.8655393719673157,
"learning_rate": 0.0001717018909899889,
"loss": 3.4142,
"step": 642
},
{
"epoch": 8.575250836120402,
"grad_norm": 0.796047031879425,
"learning_rate": 0.00017165739710789767,
"loss": 3.3558,
"step": 643
},
{
"epoch": 8.588628762541806,
"grad_norm": 1.0098161697387695,
"learning_rate": 0.00017161290322580645,
"loss": 3.553,
"step": 644
},
{
"epoch": 8.602006688963211,
"grad_norm": 1.1880302429199219,
"learning_rate": 0.00017156840934371526,
"loss": 3.3581,
"step": 645
},
{
"epoch": 8.615384615384615,
"grad_norm": 0.9361609220504761,
"learning_rate": 0.00017152391546162402,
"loss": 3.4543,
"step": 646
},
{
"epoch": 8.62876254180602,
"grad_norm": 0.8794479966163635,
"learning_rate": 0.00017147942157953283,
"loss": 3.3654,
"step": 647
},
{
"epoch": 8.642140468227424,
"grad_norm": 0.9263080954551697,
"learning_rate": 0.0001714349276974416,
"loss": 3.4376,
"step": 648
},
{
"epoch": 8.65551839464883,
"grad_norm": 1.0015815496444702,
"learning_rate": 0.0001713904338153504,
"loss": 3.3533,
"step": 649
},
{
"epoch": 8.668896321070234,
"grad_norm": 0.8525484204292297,
"learning_rate": 0.00017134593993325918,
"loss": 3.3897,
"step": 650
},
{
"epoch": 8.68227424749164,
"grad_norm": 0.7196484804153442,
"learning_rate": 0.00017130144605116799,
"loss": 3.2428,
"step": 651
},
{
"epoch": 8.695652173913043,
"grad_norm": 0.8779593706130981,
"learning_rate": 0.00017125695216907677,
"loss": 3.5471,
"step": 652
},
{
"epoch": 8.709030100334449,
"grad_norm": 0.9256909489631653,
"learning_rate": 0.00017121245828698555,
"loss": 3.1776,
"step": 653
},
{
"epoch": 8.722408026755852,
"grad_norm": 0.7774620652198792,
"learning_rate": 0.00017116796440489433,
"loss": 3.4836,
"step": 654
},
{
"epoch": 8.735785953177258,
"grad_norm": 0.8112596273422241,
"learning_rate": 0.00017112347052280314,
"loss": 3.5194,
"step": 655
},
{
"epoch": 8.749163879598662,
"grad_norm": 0.7350602746009827,
"learning_rate": 0.0001710789766407119,
"loss": 3.2248,
"step": 656
},
{
"epoch": 8.762541806020067,
"grad_norm": 0.8231781125068665,
"learning_rate": 0.0001710344827586207,
"loss": 3.4659,
"step": 657
},
{
"epoch": 8.775919732441471,
"grad_norm": 0.8921564221382141,
"learning_rate": 0.0001709899888765295,
"loss": 3.2712,
"step": 658
},
{
"epoch": 8.789297658862877,
"grad_norm": 0.8921830058097839,
"learning_rate": 0.00017094549499443827,
"loss": 3.5071,
"step": 659
},
{
"epoch": 8.80267558528428,
"grad_norm": 0.7809077501296997,
"learning_rate": 0.00017090100111234705,
"loss": 3.6639,
"step": 660
},
{
"epoch": 8.816053511705686,
"grad_norm": 0.9431234002113342,
"learning_rate": 0.00017085650723025586,
"loss": 3.2795,
"step": 661
},
{
"epoch": 8.82943143812709,
"grad_norm": 0.9707314968109131,
"learning_rate": 0.00017081201334816465,
"loss": 3.3395,
"step": 662
},
{
"epoch": 8.842809364548495,
"grad_norm": 0.7547470331192017,
"learning_rate": 0.00017076751946607343,
"loss": 3.5316,
"step": 663
},
{
"epoch": 8.856187290969899,
"grad_norm": 0.8989250659942627,
"learning_rate": 0.0001707230255839822,
"loss": 3.4029,
"step": 664
},
{
"epoch": 8.869565217391305,
"grad_norm": 1.0237400531768799,
"learning_rate": 0.00017067853170189102,
"loss": 3.5014,
"step": 665
},
{
"epoch": 8.882943143812708,
"grad_norm": 0.7289263010025024,
"learning_rate": 0.00017063403781979978,
"loss": 3.4211,
"step": 666
},
{
"epoch": 8.896321070234114,
"grad_norm": 0.7978695034980774,
"learning_rate": 0.00017058954393770856,
"loss": 3.5421,
"step": 667
},
{
"epoch": 8.909698996655518,
"grad_norm": 0.7401835918426514,
"learning_rate": 0.00017054505005561737,
"loss": 3.3419,
"step": 668
},
{
"epoch": 8.923076923076923,
"grad_norm": 0.8952983617782593,
"learning_rate": 0.00017050055617352615,
"loss": 3.1322,
"step": 669
},
{
"epoch": 8.936454849498327,
"grad_norm": 0.6922047734260559,
"learning_rate": 0.00017045606229143493,
"loss": 3.5872,
"step": 670
},
{
"epoch": 8.949832775919733,
"grad_norm": 0.8618977665901184,
"learning_rate": 0.00017041156840934371,
"loss": 3.1366,
"step": 671
},
{
"epoch": 8.963210702341136,
"grad_norm": 0.7933799624443054,
"learning_rate": 0.00017036707452725252,
"loss": 3.3108,
"step": 672
},
{
"epoch": 8.976588628762542,
"grad_norm": 0.718401312828064,
"learning_rate": 0.00017032258064516128,
"loss": 3.3771,
"step": 673
},
{
"epoch": 8.989966555183946,
"grad_norm": 0.8096804618835449,
"learning_rate": 0.0001702780867630701,
"loss": 3.3225,
"step": 674
},
{
"epoch": 9.0,
"grad_norm": 1.0055694580078125,
"learning_rate": 0.00017023359288097887,
"loss": 3.5545,
"step": 675
},
{
"epoch": 9.013377926421406,
"grad_norm": 0.710986852645874,
"learning_rate": 0.00017018909899888765,
"loss": 3.3333,
"step": 676
},
{
"epoch": 9.02675585284281,
"grad_norm": 0.672132134437561,
"learning_rate": 0.00017014460511679644,
"loss": 2.9995,
"step": 677
},
{
"epoch": 9.040133779264215,
"grad_norm": 0.6752933263778687,
"learning_rate": 0.00017010011123470525,
"loss": 3.3571,
"step": 678
},
{
"epoch": 9.053511705685619,
"grad_norm": 0.6553521156311035,
"learning_rate": 0.00017005561735261403,
"loss": 3.3407,
"step": 679
},
{
"epoch": 9.066889632107024,
"grad_norm": 0.7492311596870422,
"learning_rate": 0.0001700111234705228,
"loss": 3.325,
"step": 680
},
{
"epoch": 9.080267558528428,
"grad_norm": 0.736139714717865,
"learning_rate": 0.0001699666295884316,
"loss": 3.2626,
"step": 681
},
{
"epoch": 9.093645484949834,
"grad_norm": 0.7131486535072327,
"learning_rate": 0.0001699221357063404,
"loss": 3.2612,
"step": 682
},
{
"epoch": 9.107023411371237,
"grad_norm": 0.7037603855133057,
"learning_rate": 0.00016987764182424916,
"loss": 3.2418,
"step": 683
},
{
"epoch": 9.120401337792643,
"grad_norm": 0.685518205165863,
"learning_rate": 0.00016983314794215797,
"loss": 3.4854,
"step": 684
},
{
"epoch": 9.133779264214047,
"grad_norm": 0.6050254106521606,
"learning_rate": 0.00016978865406006675,
"loss": 3.2312,
"step": 685
},
{
"epoch": 9.147157190635452,
"grad_norm": 0.6932830810546875,
"learning_rate": 0.00016974416017797553,
"loss": 3.4634,
"step": 686
},
{
"epoch": 9.160535117056856,
"grad_norm": 0.7055158615112305,
"learning_rate": 0.00016969966629588431,
"loss": 3.1408,
"step": 687
},
{
"epoch": 9.173913043478262,
"grad_norm": 0.6887643337249756,
"learning_rate": 0.00016965517241379312,
"loss": 3.0697,
"step": 688
},
{
"epoch": 9.187290969899665,
"grad_norm": 0.7201237082481384,
"learning_rate": 0.0001696106785317019,
"loss": 3.303,
"step": 689
},
{
"epoch": 9.200668896321071,
"grad_norm": 0.6617894768714905,
"learning_rate": 0.0001695661846496107,
"loss": 3.4846,
"step": 690
},
{
"epoch": 9.214046822742475,
"grad_norm": 0.8979818224906921,
"learning_rate": 0.00016952169076751947,
"loss": 3.1898,
"step": 691
},
{
"epoch": 9.22742474916388,
"grad_norm": 0.9507981538772583,
"learning_rate": 0.00016947719688542828,
"loss": 3.4748,
"step": 692
},
{
"epoch": 9.240802675585284,
"grad_norm": 0.7935391068458557,
"learning_rate": 0.00016943270300333704,
"loss": 3.3661,
"step": 693
},
{
"epoch": 9.25418060200669,
"grad_norm": 0.7437114715576172,
"learning_rate": 0.00016938820912124584,
"loss": 3.4407,
"step": 694
},
{
"epoch": 9.267558528428093,
"grad_norm": 0.680610179901123,
"learning_rate": 0.00016934371523915463,
"loss": 3.3135,
"step": 695
},
{
"epoch": 9.280936454849499,
"grad_norm": 0.846716582775116,
"learning_rate": 0.0001692992213570634,
"loss": 3.3923,
"step": 696
},
{
"epoch": 9.294314381270903,
"grad_norm": 0.9567786455154419,
"learning_rate": 0.0001692547274749722,
"loss": 3.1405,
"step": 697
},
{
"epoch": 9.307692307692308,
"grad_norm": 0.7509344816207886,
"learning_rate": 0.000169210233592881,
"loss": 3.4031,
"step": 698
},
{
"epoch": 9.321070234113712,
"grad_norm": 0.8118243217468262,
"learning_rate": 0.00016916573971078978,
"loss": 3.317,
"step": 699
},
{
"epoch": 9.334448160535118,
"grad_norm": 0.7445617318153381,
"learning_rate": 0.00016912124582869857,
"loss": 3.3989,
"step": 700
},
{
"epoch": 9.347826086956522,
"grad_norm": 0.7520869970321655,
"learning_rate": 0.00016907675194660735,
"loss": 3.08,
"step": 701
},
{
"epoch": 9.361204013377927,
"grad_norm": 0.7466426491737366,
"learning_rate": 0.00016903225806451616,
"loss": 3.3338,
"step": 702
},
{
"epoch": 9.37458193979933,
"grad_norm": 0.7595514059066772,
"learning_rate": 0.0001689877641824249,
"loss": 3.08,
"step": 703
},
{
"epoch": 9.387959866220736,
"grad_norm": 0.713771641254425,
"learning_rate": 0.00016894327030033372,
"loss": 3.236,
"step": 704
},
{
"epoch": 9.40133779264214,
"grad_norm": 0.670863687992096,
"learning_rate": 0.0001688987764182425,
"loss": 3.4491,
"step": 705
},
{
"epoch": 9.414715719063546,
"grad_norm": 0.8842789530754089,
"learning_rate": 0.0001688542825361513,
"loss": 3.3444,
"step": 706
},
{
"epoch": 9.42809364548495,
"grad_norm": 0.8298172950744629,
"learning_rate": 0.00016880978865406007,
"loss": 3.3818,
"step": 707
},
{
"epoch": 9.441471571906355,
"grad_norm": 0.7407504320144653,
"learning_rate": 0.00016876529477196885,
"loss": 3.2294,
"step": 708
},
{
"epoch": 9.454849498327759,
"grad_norm": 0.6642070412635803,
"learning_rate": 0.00016872080088987766,
"loss": 3.3009,
"step": 709
},
{
"epoch": 9.468227424749164,
"grad_norm": 0.7627503275871277,
"learning_rate": 0.00016867630700778642,
"loss": 3.3486,
"step": 710
},
{
"epoch": 9.481605351170568,
"grad_norm": 0.7307603359222412,
"learning_rate": 0.00016863181312569523,
"loss": 3.0257,
"step": 711
},
{
"epoch": 9.494983277591974,
"grad_norm": 0.7932866215705872,
"learning_rate": 0.000168587319243604,
"loss": 3.4191,
"step": 712
},
{
"epoch": 9.508361204013378,
"grad_norm": 0.7457575798034668,
"learning_rate": 0.0001685428253615128,
"loss": 3.2283,
"step": 713
},
{
"epoch": 9.521739130434783,
"grad_norm": 0.6718200445175171,
"learning_rate": 0.00016849833147942157,
"loss": 3.4569,
"step": 714
},
{
"epoch": 9.535117056856187,
"grad_norm": 0.8189072608947754,
"learning_rate": 0.00016845383759733038,
"loss": 3.2585,
"step": 715
},
{
"epoch": 9.548494983277592,
"grad_norm": 0.6895336508750916,
"learning_rate": 0.00016840934371523917,
"loss": 3.2417,
"step": 716
},
{
"epoch": 9.561872909698996,
"grad_norm": 0.723173975944519,
"learning_rate": 0.00016836484983314795,
"loss": 3.3243,
"step": 717
},
{
"epoch": 9.575250836120402,
"grad_norm": 0.8354344964027405,
"learning_rate": 0.00016832035595105673,
"loss": 3.3585,
"step": 718
},
{
"epoch": 9.588628762541806,
"grad_norm": 0.6736294031143188,
"learning_rate": 0.00016827586206896554,
"loss": 3.188,
"step": 719
},
{
"epoch": 9.602006688963211,
"grad_norm": 0.7790263295173645,
"learning_rate": 0.0001682313681868743,
"loss": 3.3171,
"step": 720
},
{
"epoch": 9.615384615384615,
"grad_norm": 0.6426937580108643,
"learning_rate": 0.0001681868743047831,
"loss": 3.0854,
"step": 721
},
{
"epoch": 9.62876254180602,
"grad_norm": 0.7029106020927429,
"learning_rate": 0.0001681423804226919,
"loss": 3.3629,
"step": 722
},
{
"epoch": 9.642140468227424,
"grad_norm": 0.8353022933006287,
"learning_rate": 0.00016809788654060067,
"loss": 3.3715,
"step": 723
},
{
"epoch": 9.65551839464883,
"grad_norm": 0.8578335642814636,
"learning_rate": 0.00016805339265850945,
"loss": 3.3554,
"step": 724
},
{
"epoch": 9.668896321070234,
"grad_norm": 0.6998556852340698,
"learning_rate": 0.00016800889877641826,
"loss": 3.3043,
"step": 725
},
{
"epoch": 9.68227424749164,
"grad_norm": 0.7134855389595032,
"learning_rate": 0.00016796440489432704,
"loss": 3.5856,
"step": 726
},
{
"epoch": 9.695652173913043,
"grad_norm": 0.6636050939559937,
"learning_rate": 0.00016791991101223583,
"loss": 3.2156,
"step": 727
},
{
"epoch": 9.709030100334449,
"grad_norm": 0.7757130861282349,
"learning_rate": 0.0001678754171301446,
"loss": 3.4974,
"step": 728
},
{
"epoch": 9.722408026755852,
"grad_norm": 0.770648181438446,
"learning_rate": 0.00016783092324805342,
"loss": 3.3251,
"step": 729
},
{
"epoch": 9.735785953177258,
"grad_norm": 0.7728201746940613,
"learning_rate": 0.00016778642936596217,
"loss": 3.2666,
"step": 730
},
{
"epoch": 9.749163879598662,
"grad_norm": 0.8277239203453064,
"learning_rate": 0.00016774193548387098,
"loss": 3.3867,
"step": 731
},
{
"epoch": 9.762541806020067,
"grad_norm": 0.6534886360168457,
"learning_rate": 0.00016769744160177977,
"loss": 3.175,
"step": 732
},
{
"epoch": 9.775919732441471,
"grad_norm": 0.8508428335189819,
"learning_rate": 0.00016765294771968855,
"loss": 3.2084,
"step": 733
},
{
"epoch": 9.789297658862877,
"grad_norm": 0.7656721472740173,
"learning_rate": 0.00016760845383759733,
"loss": 3.1426,
"step": 734
},
{
"epoch": 9.80267558528428,
"grad_norm": 0.9495553970336914,
"learning_rate": 0.00016756395995550614,
"loss": 3.1623,
"step": 735
},
{
"epoch": 9.816053511705686,
"grad_norm": 0.7998641729354858,
"learning_rate": 0.00016751946607341492,
"loss": 3.3893,
"step": 736
},
{
"epoch": 9.82943143812709,
"grad_norm": 0.8124551177024841,
"learning_rate": 0.0001674749721913237,
"loss": 3.2012,
"step": 737
},
{
"epoch": 9.842809364548495,
"grad_norm": 0.6332049369812012,
"learning_rate": 0.0001674304783092325,
"loss": 3.3384,
"step": 738
},
{
"epoch": 9.856187290969899,
"grad_norm": 0.7114555835723877,
"learning_rate": 0.0001673859844271413,
"loss": 3.0802,
"step": 739
},
{
"epoch": 9.869565217391305,
"grad_norm": 0.7175182700157166,
"learning_rate": 0.00016734149054505005,
"loss": 3.2572,
"step": 740
},
{
"epoch": 9.882943143812708,
"grad_norm": 0.7724816799163818,
"learning_rate": 0.00016729699666295886,
"loss": 3.1078,
"step": 741
},
{
"epoch": 9.896321070234114,
"grad_norm": 0.7834901213645935,
"learning_rate": 0.00016725250278086764,
"loss": 3.2513,
"step": 742
},
{
"epoch": 9.909698996655518,
"grad_norm": 0.663495659828186,
"learning_rate": 0.00016720800889877643,
"loss": 3.327,
"step": 743
},
{
"epoch": 9.923076923076923,
"grad_norm": 0.7828975319862366,
"learning_rate": 0.0001671635150166852,
"loss": 3.609,
"step": 744
},
{
"epoch": 9.936454849498327,
"grad_norm": 0.6747825145721436,
"learning_rate": 0.00016711902113459402,
"loss": 3.4479,
"step": 745
},
{
"epoch": 9.949832775919733,
"grad_norm": 0.7816379070281982,
"learning_rate": 0.0001670745272525028,
"loss": 3.2369,
"step": 746
},
{
"epoch": 9.963210702341136,
"grad_norm": 0.7011098265647888,
"learning_rate": 0.00016703003337041158,
"loss": 2.9103,
"step": 747
},
{
"epoch": 9.976588628762542,
"grad_norm": 0.7165176868438721,
"learning_rate": 0.00016698553948832036,
"loss": 3.1669,
"step": 748
},
{
"epoch": 9.989966555183946,
"grad_norm": 0.766315758228302,
"learning_rate": 0.00016694104560622915,
"loss": 3.1138,
"step": 749
},
{
"epoch": 10.0,
"grad_norm": 0.876315176486969,
"learning_rate": 0.00016689655172413793,
"loss": 3.4851,
"step": 750
},
{
"epoch": 10.013377926421406,
"grad_norm": 0.807686984539032,
"learning_rate": 0.0001668520578420467,
"loss": 3.193,
"step": 751
},
{
"epoch": 10.02675585284281,
"grad_norm": 0.7085704803466797,
"learning_rate": 0.00016680756395995552,
"loss": 3.4797,
"step": 752
},
{
"epoch": 10.040133779264215,
"grad_norm": 0.7119605541229248,
"learning_rate": 0.0001667630700778643,
"loss": 3.1713,
"step": 753
},
{
"epoch": 10.053511705685619,
"grad_norm": 0.6569423675537109,
"learning_rate": 0.00016671857619577309,
"loss": 3.1661,
"step": 754
},
{
"epoch": 10.066889632107024,
"grad_norm": 0.8173550367355347,
"learning_rate": 0.00016667408231368187,
"loss": 2.8467,
"step": 755
},
{
"epoch": 10.080267558528428,
"grad_norm": 0.7261365056037903,
"learning_rate": 0.00016662958843159068,
"loss": 3.3679,
"step": 756
},
{
"epoch": 10.093645484949834,
"grad_norm": 0.7997227311134338,
"learning_rate": 0.00016658509454949943,
"loss": 3.0985,
"step": 757
},
{
"epoch": 10.107023411371237,
"grad_norm": 0.653391420841217,
"learning_rate": 0.00016654060066740824,
"loss": 3.2156,
"step": 758
},
{
"epoch": 10.120401337792643,
"grad_norm": 0.6799002289772034,
"learning_rate": 0.00016649610678531703,
"loss": 3.3302,
"step": 759
},
{
"epoch": 10.133779264214047,
"grad_norm": 0.6444498896598816,
"learning_rate": 0.0001664516129032258,
"loss": 3.2813,
"step": 760
},
{
"epoch": 10.147157190635452,
"grad_norm": 1.064769983291626,
"learning_rate": 0.0001664071190211346,
"loss": 3.1852,
"step": 761
},
{
"epoch": 10.160535117056856,
"grad_norm": 0.6534339189529419,
"learning_rate": 0.0001663626251390434,
"loss": 3.1563,
"step": 762
},
{
"epoch": 10.173913043478262,
"grad_norm": 0.6909127235412598,
"learning_rate": 0.00016631813125695218,
"loss": 3.2728,
"step": 763
},
{
"epoch": 10.187290969899665,
"grad_norm": 0.6549767851829529,
"learning_rate": 0.00016627363737486096,
"loss": 3.0491,
"step": 764
},
{
"epoch": 10.200668896321071,
"grad_norm": 0.678054928779602,
"learning_rate": 0.00016622914349276975,
"loss": 3.4807,
"step": 765
},
{
"epoch": 10.214046822742475,
"grad_norm": 0.613358199596405,
"learning_rate": 0.00016618464961067856,
"loss": 3.1746,
"step": 766
},
{
"epoch": 10.22742474916388,
"grad_norm": 0.6624737977981567,
"learning_rate": 0.0001661401557285873,
"loss": 2.8528,
"step": 767
},
{
"epoch": 10.240802675585284,
"grad_norm": 0.65067458152771,
"learning_rate": 0.00016609566184649612,
"loss": 3.1843,
"step": 768
},
{
"epoch": 10.25418060200669,
"grad_norm": 0.6192435622215271,
"learning_rate": 0.0001660511679644049,
"loss": 3.3162,
"step": 769
},
{
"epoch": 10.267558528428093,
"grad_norm": 0.6456341743469238,
"learning_rate": 0.00016600667408231369,
"loss": 3.2302,
"step": 770
},
{
"epoch": 10.280936454849499,
"grad_norm": 2.357724189758301,
"learning_rate": 0.00016596218020022247,
"loss": 3.2741,
"step": 771
},
{
"epoch": 10.294314381270903,
"grad_norm": 0.6833475828170776,
"learning_rate": 0.00016591768631813128,
"loss": 3.1516,
"step": 772
},
{
"epoch": 10.307692307692308,
"grad_norm": 0.5557199716567993,
"learning_rate": 0.00016587319243604006,
"loss": 3.281,
"step": 773
},
{
"epoch": 10.321070234113712,
"grad_norm": 0.6617905497550964,
"learning_rate": 0.00016582869855394884,
"loss": 3.375,
"step": 774
},
{
"epoch": 10.334448160535118,
"grad_norm": 0.5671921372413635,
"learning_rate": 0.00016578420467185762,
"loss": 3.4335,
"step": 775
},
{
"epoch": 10.347826086956522,
"grad_norm": 0.8487278819084167,
"learning_rate": 0.00016573971078976643,
"loss": 3.143,
"step": 776
},
{
"epoch": 10.361204013377927,
"grad_norm": 0.6489982604980469,
"learning_rate": 0.0001656952169076752,
"loss": 3.3258,
"step": 777
},
{
"epoch": 10.37458193979933,
"grad_norm": 0.8773537278175354,
"learning_rate": 0.000165650723025584,
"loss": 3.0466,
"step": 778
},
{
"epoch": 10.387959866220736,
"grad_norm": 0.5961865782737732,
"learning_rate": 0.00016560622914349278,
"loss": 3.3417,
"step": 779
},
{
"epoch": 10.40133779264214,
"grad_norm": 0.6149600148200989,
"learning_rate": 0.00016556173526140156,
"loss": 3.0622,
"step": 780
},
{
"epoch": 10.414715719063546,
"grad_norm": 0.7591158151626587,
"learning_rate": 0.00016551724137931035,
"loss": 3.2078,
"step": 781
},
{
"epoch": 10.42809364548495,
"grad_norm": 0.7915151119232178,
"learning_rate": 0.00016547274749721916,
"loss": 3.3082,
"step": 782
},
{
"epoch": 10.441471571906355,
"grad_norm": 0.8709903359413147,
"learning_rate": 0.00016542825361512794,
"loss": 3.3073,
"step": 783
},
{
"epoch": 10.454849498327759,
"grad_norm": 0.6593959331512451,
"learning_rate": 0.00016538375973303672,
"loss": 3.1574,
"step": 784
},
{
"epoch": 10.468227424749164,
"grad_norm": 0.8101013898849487,
"learning_rate": 0.0001653392658509455,
"loss": 3.3631,
"step": 785
},
{
"epoch": 10.481605351170568,
"grad_norm": 0.8200273513793945,
"learning_rate": 0.0001652947719688543,
"loss": 3.0447,
"step": 786
},
{
"epoch": 10.494983277591974,
"grad_norm": 0.7304090857505798,
"learning_rate": 0.00016525027808676307,
"loss": 3.3382,
"step": 787
},
{
"epoch": 10.508361204013378,
"grad_norm": 0.7059088349342346,
"learning_rate": 0.00016520578420467188,
"loss": 3.085,
"step": 788
},
{
"epoch": 10.521739130434783,
"grad_norm": 0.6664522886276245,
"learning_rate": 0.00016516129032258066,
"loss": 3.3198,
"step": 789
},
{
"epoch": 10.535117056856187,
"grad_norm": 0.9230799078941345,
"learning_rate": 0.00016511679644048944,
"loss": 3.2502,
"step": 790
},
{
"epoch": 10.548494983277592,
"grad_norm": 0.6974027752876282,
"learning_rate": 0.00016507230255839822,
"loss": 3.2432,
"step": 791
},
{
"epoch": 10.561872909698996,
"grad_norm": 0.7186788320541382,
"learning_rate": 0.000165027808676307,
"loss": 3.2273,
"step": 792
},
{
"epoch": 10.575250836120402,
"grad_norm": 0.6168047189712524,
"learning_rate": 0.00016498331479421582,
"loss": 3.2319,
"step": 793
},
{
"epoch": 10.588628762541806,
"grad_norm": 0.6219142079353333,
"learning_rate": 0.00016493882091212457,
"loss": 2.9733,
"step": 794
},
{
"epoch": 10.602006688963211,
"grad_norm": 0.573359489440918,
"learning_rate": 0.00016489432703003338,
"loss": 3.214,
"step": 795
},
{
"epoch": 10.615384615384615,
"grad_norm": 0.678263783454895,
"learning_rate": 0.00016484983314794216,
"loss": 3.2331,
"step": 796
},
{
"epoch": 10.62876254180602,
"grad_norm": 0.6593761444091797,
"learning_rate": 0.00016480533926585095,
"loss": 3.3605,
"step": 797
},
{
"epoch": 10.642140468227424,
"grad_norm": 0.8732627034187317,
"learning_rate": 0.00016476084538375973,
"loss": 3.3531,
"step": 798
},
{
"epoch": 10.65551839464883,
"grad_norm": 0.7198925614356995,
"learning_rate": 0.00016471635150166854,
"loss": 3.2927,
"step": 799
},
{
"epoch": 10.668896321070234,
"grad_norm": 0.7275107502937317,
"learning_rate": 0.00016467185761957732,
"loss": 3.3695,
"step": 800
},
{
"epoch": 10.68227424749164,
"grad_norm": 0.7077828049659729,
"learning_rate": 0.0001646273637374861,
"loss": 2.846,
"step": 801
},
{
"epoch": 10.695652173913043,
"grad_norm": 0.7579251527786255,
"learning_rate": 0.00016458286985539488,
"loss": 3.1917,
"step": 802
},
{
"epoch": 10.709030100334449,
"grad_norm": 0.7607265114784241,
"learning_rate": 0.0001645383759733037,
"loss": 3.2919,
"step": 803
},
{
"epoch": 10.722408026755852,
"grad_norm": 0.7122685313224792,
"learning_rate": 0.00016449388209121245,
"loss": 3.7108,
"step": 804
},
{
"epoch": 10.735785953177258,
"grad_norm": 0.7256726622581482,
"learning_rate": 0.00016444938820912126,
"loss": 3.3209,
"step": 805
},
{
"epoch": 10.749163879598662,
"grad_norm": 0.7903631925582886,
"learning_rate": 0.00016440489432703004,
"loss": 3.398,
"step": 806
},
{
"epoch": 10.762541806020067,
"grad_norm": 1.78204345703125,
"learning_rate": 0.00016436040044493882,
"loss": 3.1118,
"step": 807
},
{
"epoch": 10.775919732441471,
"grad_norm": 0.7647016644477844,
"learning_rate": 0.0001643159065628476,
"loss": 3.1889,
"step": 808
},
{
"epoch": 10.789297658862877,
"grad_norm": 0.8039811253547668,
"learning_rate": 0.00016427141268075642,
"loss": 3.0104,
"step": 809
},
{
"epoch": 10.80267558528428,
"grad_norm": 0.6011155843734741,
"learning_rate": 0.0001642269187986652,
"loss": 3.0988,
"step": 810
},
{
"epoch": 10.816053511705686,
"grad_norm": 0.8137276768684387,
"learning_rate": 0.00016418242491657398,
"loss": 3.4308,
"step": 811
},
{
"epoch": 10.82943143812709,
"grad_norm": 0.6501771807670593,
"learning_rate": 0.00016413793103448276,
"loss": 3.1405,
"step": 812
},
{
"epoch": 10.842809364548495,
"grad_norm": 0.678032636642456,
"learning_rate": 0.00016409343715239157,
"loss": 3.2243,
"step": 813
},
{
"epoch": 10.856187290969899,
"grad_norm": 0.6830305457115173,
"learning_rate": 0.00016404894327030033,
"loss": 3.2685,
"step": 814
},
{
"epoch": 10.869565217391305,
"grad_norm": 0.7482068538665771,
"learning_rate": 0.00016400444938820914,
"loss": 3.3363,
"step": 815
},
{
"epoch": 10.882943143812708,
"grad_norm": 0.6592227816581726,
"learning_rate": 0.00016395995550611792,
"loss": 3.2914,
"step": 816
},
{
"epoch": 10.896321070234114,
"grad_norm": 0.7520759105682373,
"learning_rate": 0.0001639154616240267,
"loss": 3.1371,
"step": 817
},
{
"epoch": 10.909698996655518,
"grad_norm": 0.6802201271057129,
"learning_rate": 0.00016387096774193548,
"loss": 3.2925,
"step": 818
},
{
"epoch": 10.923076923076923,
"grad_norm": 0.7528939247131348,
"learning_rate": 0.0001638264738598443,
"loss": 3.2147,
"step": 819
},
{
"epoch": 10.936454849498327,
"grad_norm": 0.7070727348327637,
"learning_rate": 0.00016378197997775308,
"loss": 3.2649,
"step": 820
},
{
"epoch": 10.949832775919733,
"grad_norm": 0.6121620535850525,
"learning_rate": 0.00016373748609566186,
"loss": 3.3999,
"step": 821
},
{
"epoch": 10.963210702341136,
"grad_norm": 0.7355679273605347,
"learning_rate": 0.00016369299221357064,
"loss": 3.2561,
"step": 822
},
{
"epoch": 10.976588628762542,
"grad_norm": 0.7294445037841797,
"learning_rate": 0.00016364849833147945,
"loss": 2.7019,
"step": 823
},
{
"epoch": 10.989966555183946,
"grad_norm": 0.8628729581832886,
"learning_rate": 0.0001636040044493882,
"loss": 3.3655,
"step": 824
},
{
"epoch": 11.0,
"grad_norm": 0.8784325122833252,
"learning_rate": 0.00016355951056729702,
"loss": 3.504,
"step": 825
},
{
"epoch": 11.013377926421406,
"grad_norm": 0.6880869269371033,
"learning_rate": 0.0001635150166852058,
"loss": 3.1878,
"step": 826
},
{
"epoch": 11.02675585284281,
"grad_norm": 0.5625393390655518,
"learning_rate": 0.00016347052280311458,
"loss": 3.2757,
"step": 827
},
{
"epoch": 11.040133779264215,
"grad_norm": 0.5854038596153259,
"learning_rate": 0.00016342602892102336,
"loss": 3.0994,
"step": 828
},
{
"epoch": 11.053511705685619,
"grad_norm": 0.6682130098342896,
"learning_rate": 0.00016338153503893217,
"loss": 3.1935,
"step": 829
},
{
"epoch": 11.066889632107024,
"grad_norm": 0.6216278076171875,
"learning_rate": 0.00016333704115684095,
"loss": 3.1273,
"step": 830
},
{
"epoch": 11.080267558528428,
"grad_norm": 0.61285001039505,
"learning_rate": 0.0001632925472747497,
"loss": 3.2377,
"step": 831
},
{
"epoch": 11.093645484949834,
"grad_norm": 0.6559188365936279,
"learning_rate": 0.00016324805339265852,
"loss": 3.1084,
"step": 832
},
{
"epoch": 11.107023411371237,
"grad_norm": 0.6322848200798035,
"learning_rate": 0.0001632035595105673,
"loss": 3.3178,
"step": 833
},
{
"epoch": 11.120401337792643,
"grad_norm": 0.6306194067001343,
"learning_rate": 0.00016315906562847608,
"loss": 3.2012,
"step": 834
},
{
"epoch": 11.133779264214047,
"grad_norm": 0.6923161149024963,
"learning_rate": 0.00016311457174638487,
"loss": 2.9328,
"step": 835
},
{
"epoch": 11.147157190635452,
"grad_norm": 0.6900002360343933,
"learning_rate": 0.00016307007786429368,
"loss": 3.1436,
"step": 836
},
{
"epoch": 11.160535117056856,
"grad_norm": 0.817669153213501,
"learning_rate": 0.00016302558398220246,
"loss": 3.0935,
"step": 837
},
{
"epoch": 11.173913043478262,
"grad_norm": 0.7544119954109192,
"learning_rate": 0.00016298109010011124,
"loss": 3.0646,
"step": 838
},
{
"epoch": 11.187290969899665,
"grad_norm": 0.7996231913566589,
"learning_rate": 0.00016293659621802002,
"loss": 3.1975,
"step": 839
},
{
"epoch": 11.200668896321071,
"grad_norm": 0.6186792850494385,
"learning_rate": 0.00016289210233592883,
"loss": 3.31,
"step": 840
},
{
"epoch": 11.214046822742475,
"grad_norm": 0.6926666498184204,
"learning_rate": 0.0001628476084538376,
"loss": 3.0765,
"step": 841
},
{
"epoch": 11.22742474916388,
"grad_norm": 0.7475588917732239,
"learning_rate": 0.0001628031145717464,
"loss": 3.1743,
"step": 842
},
{
"epoch": 11.240802675585284,
"grad_norm": 0.5520989298820496,
"learning_rate": 0.00016275862068965518,
"loss": 3.4243,
"step": 843
},
{
"epoch": 11.25418060200669,
"grad_norm": 0.6556730270385742,
"learning_rate": 0.00016271412680756396,
"loss": 3.3293,
"step": 844
},
{
"epoch": 11.267558528428093,
"grad_norm": 0.6509199738502502,
"learning_rate": 0.00016266963292547274,
"loss": 2.999,
"step": 845
},
{
"epoch": 11.280936454849499,
"grad_norm": 0.6254273653030396,
"learning_rate": 0.00016262513904338155,
"loss": 3.1869,
"step": 846
},
{
"epoch": 11.294314381270903,
"grad_norm": 0.7454530000686646,
"learning_rate": 0.00016258064516129034,
"loss": 3.3694,
"step": 847
},
{
"epoch": 11.307692307692308,
"grad_norm": 0.7563592791557312,
"learning_rate": 0.00016253615127919912,
"loss": 3.0057,
"step": 848
},
{
"epoch": 11.321070234113712,
"grad_norm": 0.6986783742904663,
"learning_rate": 0.0001624916573971079,
"loss": 2.9893,
"step": 849
},
{
"epoch": 11.334448160535118,
"grad_norm": 0.7260631322860718,
"learning_rate": 0.0001624471635150167,
"loss": 3.1733,
"step": 850
},
{
"epoch": 11.347826086956522,
"grad_norm": 0.7522863745689392,
"learning_rate": 0.00016240266963292547,
"loss": 2.9829,
"step": 851
},
{
"epoch": 11.361204013377927,
"grad_norm": 0.7290140986442566,
"learning_rate": 0.00016235817575083428,
"loss": 3.1887,
"step": 852
},
{
"epoch": 11.37458193979933,
"grad_norm": 0.6470169425010681,
"learning_rate": 0.00016231368186874306,
"loss": 3.2537,
"step": 853
},
{
"epoch": 11.387959866220736,
"grad_norm": 0.863742470741272,
"learning_rate": 0.00016226918798665184,
"loss": 3.3443,
"step": 854
},
{
"epoch": 11.40133779264214,
"grad_norm": 0.7363939881324768,
"learning_rate": 0.00016222469410456062,
"loss": 3.3653,
"step": 855
},
{
"epoch": 11.414715719063546,
"grad_norm": 0.6548926830291748,
"learning_rate": 0.00016218020022246943,
"loss": 3.0373,
"step": 856
},
{
"epoch": 11.42809364548495,
"grad_norm": 0.8087872862815857,
"learning_rate": 0.00016213570634037821,
"loss": 3.0118,
"step": 857
},
{
"epoch": 11.441471571906355,
"grad_norm": 0.677811324596405,
"learning_rate": 0.000162091212458287,
"loss": 3.0339,
"step": 858
},
{
"epoch": 11.454849498327759,
"grad_norm": 0.6907945275306702,
"learning_rate": 0.00016204671857619578,
"loss": 2.9496,
"step": 859
},
{
"epoch": 11.468227424749164,
"grad_norm": 0.6940027475357056,
"learning_rate": 0.0001620022246941046,
"loss": 3.2825,
"step": 860
},
{
"epoch": 11.481605351170568,
"grad_norm": 0.7132136225700378,
"learning_rate": 0.00016195773081201334,
"loss": 3.1271,
"step": 861
},
{
"epoch": 11.494983277591974,
"grad_norm": 0.5997372269630432,
"learning_rate": 0.00016191323692992215,
"loss": 3.1292,
"step": 862
},
{
"epoch": 11.508361204013378,
"grad_norm": 0.6468494534492493,
"learning_rate": 0.00016186874304783094,
"loss": 3.32,
"step": 863
},
{
"epoch": 11.521739130434783,
"grad_norm": 0.5792532563209534,
"learning_rate": 0.00016182424916573972,
"loss": 3.3657,
"step": 864
},
{
"epoch": 11.535117056856187,
"grad_norm": 0.8242068290710449,
"learning_rate": 0.0001617797552836485,
"loss": 3.0094,
"step": 865
},
{
"epoch": 11.548494983277592,
"grad_norm": 0.9260333776473999,
"learning_rate": 0.0001617352614015573,
"loss": 3.1123,
"step": 866
},
{
"epoch": 11.561872909698996,
"grad_norm": 0.6337956786155701,
"learning_rate": 0.0001616907675194661,
"loss": 3.1761,
"step": 867
},
{
"epoch": 11.575250836120402,
"grad_norm": 0.6010364294052124,
"learning_rate": 0.00016164627363737487,
"loss": 3.2987,
"step": 868
},
{
"epoch": 11.588628762541806,
"grad_norm": 0.7492111921310425,
"learning_rate": 0.00016160177975528366,
"loss": 2.9085,
"step": 869
},
{
"epoch": 11.602006688963211,
"grad_norm": 0.6329553127288818,
"learning_rate": 0.00016155728587319247,
"loss": 3.1841,
"step": 870
},
{
"epoch": 11.615384615384615,
"grad_norm": 0.768527626991272,
"learning_rate": 0.00016151279199110122,
"loss": 3.0061,
"step": 871
},
{
"epoch": 11.62876254180602,
"grad_norm": 0.6333640813827515,
"learning_rate": 0.00016146829810901003,
"loss": 3.3046,
"step": 872
},
{
"epoch": 11.642140468227424,
"grad_norm": 0.7457571625709534,
"learning_rate": 0.00016142380422691881,
"loss": 3.2728,
"step": 873
},
{
"epoch": 11.65551839464883,
"grad_norm": 0.6389586925506592,
"learning_rate": 0.0001613793103448276,
"loss": 3.419,
"step": 874
},
{
"epoch": 11.668896321070234,
"grad_norm": 0.8885436058044434,
"learning_rate": 0.00016133481646273638,
"loss": 3.0938,
"step": 875
},
{
"epoch": 11.68227424749164,
"grad_norm": 0.7936431765556335,
"learning_rate": 0.00016129032258064516,
"loss": 3.305,
"step": 876
},
{
"epoch": 11.695652173913043,
"grad_norm": 0.6133994460105896,
"learning_rate": 0.00016124582869855397,
"loss": 3.3474,
"step": 877
},
{
"epoch": 11.709030100334449,
"grad_norm": 0.6638192534446716,
"learning_rate": 0.00016120133481646273,
"loss": 3.1418,
"step": 878
},
{
"epoch": 11.722408026755852,
"grad_norm": 0.6820496320724487,
"learning_rate": 0.00016115684093437154,
"loss": 3.1421,
"step": 879
},
{
"epoch": 11.735785953177258,
"grad_norm": 0.6057732105255127,
"learning_rate": 0.00016111234705228032,
"loss": 3.091,
"step": 880
},
{
"epoch": 11.749163879598662,
"grad_norm": 0.6267048716545105,
"learning_rate": 0.0001610678531701891,
"loss": 3.1289,
"step": 881
},
{
"epoch": 11.762541806020067,
"grad_norm": 0.6822847723960876,
"learning_rate": 0.00016102335928809788,
"loss": 3.157,
"step": 882
},
{
"epoch": 11.775919732441471,
"grad_norm": 0.6809714436531067,
"learning_rate": 0.0001609788654060067,
"loss": 3.0806,
"step": 883
},
{
"epoch": 11.789297658862877,
"grad_norm": 0.5546092391014099,
"learning_rate": 0.00016093437152391547,
"loss": 3.1853,
"step": 884
},
{
"epoch": 11.80267558528428,
"grad_norm": 0.7375029921531677,
"learning_rate": 0.00016088987764182426,
"loss": 3.1287,
"step": 885
},
{
"epoch": 11.816053511705686,
"grad_norm": 0.6246840953826904,
"learning_rate": 0.00016084538375973304,
"loss": 3.1331,
"step": 886
},
{
"epoch": 11.82943143812709,
"grad_norm": 0.6088026762008667,
"learning_rate": 0.00016080088987764185,
"loss": 3.3781,
"step": 887
},
{
"epoch": 11.842809364548495,
"grad_norm": 0.7996237874031067,
"learning_rate": 0.0001607563959955506,
"loss": 3.0161,
"step": 888
},
{
"epoch": 11.856187290969899,
"grad_norm": 0.6221441626548767,
"learning_rate": 0.0001607119021134594,
"loss": 3.1491,
"step": 889
},
{
"epoch": 11.869565217391305,
"grad_norm": 0.6276041269302368,
"learning_rate": 0.0001606674082313682,
"loss": 3.2575,
"step": 890
},
{
"epoch": 11.882943143812708,
"grad_norm": 0.6394500136375427,
"learning_rate": 0.00016062291434927698,
"loss": 3.3437,
"step": 891
},
{
"epoch": 11.896321070234114,
"grad_norm": 0.7674509286880493,
"learning_rate": 0.00016057842046718576,
"loss": 3.1995,
"step": 892
},
{
"epoch": 11.909698996655518,
"grad_norm": 0.7502215504646301,
"learning_rate": 0.00016053392658509457,
"loss": 3.3129,
"step": 893
},
{
"epoch": 11.923076923076923,
"grad_norm": 0.6078189611434937,
"learning_rate": 0.00016048943270300335,
"loss": 3.1623,
"step": 894
},
{
"epoch": 11.936454849498327,
"grad_norm": 0.6113708019256592,
"learning_rate": 0.00016044493882091213,
"loss": 3.6063,
"step": 895
},
{
"epoch": 11.949832775919733,
"grad_norm": 0.6606878638267517,
"learning_rate": 0.00016040044493882092,
"loss": 3.3216,
"step": 896
},
{
"epoch": 11.963210702341136,
"grad_norm": 0.7055956125259399,
"learning_rate": 0.00016035595105672973,
"loss": 3.2006,
"step": 897
},
{
"epoch": 11.976588628762542,
"grad_norm": 0.7424116730690002,
"learning_rate": 0.00016031145717463848,
"loss": 3.1298,
"step": 898
},
{
"epoch": 11.989966555183946,
"grad_norm": 0.6675695180892944,
"learning_rate": 0.0001602669632925473,
"loss": 3.1116,
"step": 899
},
{
"epoch": 12.0,
"grad_norm": 0.9356181621551514,
"learning_rate": 0.00016022246941045607,
"loss": 3.2461,
"step": 900
},
{
"epoch": 12.013377926421406,
"grad_norm": 0.8539507985115051,
"learning_rate": 0.00016017797552836486,
"loss": 3.0671,
"step": 901
},
{
"epoch": 12.02675585284281,
"grad_norm": 0.573266327381134,
"learning_rate": 0.00016013348164627364,
"loss": 3.0789,
"step": 902
},
{
"epoch": 12.040133779264215,
"grad_norm": 0.5849746465682983,
"learning_rate": 0.00016008898776418245,
"loss": 3.1623,
"step": 903
},
{
"epoch": 12.053511705685619,
"grad_norm": 0.6523334980010986,
"learning_rate": 0.00016004449388209123,
"loss": 3.0243,
"step": 904
},
{
"epoch": 12.066889632107024,
"grad_norm": 0.6428223848342896,
"learning_rate": 0.00016,
"loss": 3.1241,
"step": 905
},
{
"epoch": 12.080267558528428,
"grad_norm": 0.5881937742233276,
"learning_rate": 0.0001599555061179088,
"loss": 3.1744,
"step": 906
},
{
"epoch": 12.093645484949834,
"grad_norm": 0.7523583173751831,
"learning_rate": 0.0001599110122358176,
"loss": 3.0821,
"step": 907
},
{
"epoch": 12.107023411371237,
"grad_norm": 0.6120390295982361,
"learning_rate": 0.00015986651835372636,
"loss": 3.16,
"step": 908
},
{
"epoch": 12.120401337792643,
"grad_norm": 0.6486253142356873,
"learning_rate": 0.00015982202447163517,
"loss": 3.1402,
"step": 909
},
{
"epoch": 12.133779264214047,
"grad_norm": 0.6322839260101318,
"learning_rate": 0.00015977753058954395,
"loss": 3.1325,
"step": 910
},
{
"epoch": 12.147157190635452,
"grad_norm": 0.5858875513076782,
"learning_rate": 0.00015973303670745273,
"loss": 3.1705,
"step": 911
},
{
"epoch": 12.160535117056856,
"grad_norm": 0.6495780348777771,
"learning_rate": 0.00015968854282536152,
"loss": 3.1977,
"step": 912
},
{
"epoch": 12.173913043478262,
"grad_norm": 0.6483474969863892,
"learning_rate": 0.00015964404894327033,
"loss": 2.9101,
"step": 913
},
{
"epoch": 12.187290969899665,
"grad_norm": 0.6021110415458679,
"learning_rate": 0.0001595995550611791,
"loss": 3.2106,
"step": 914
},
{
"epoch": 12.200668896321071,
"grad_norm": 0.5101630687713623,
"learning_rate": 0.00015955506117908786,
"loss": 3.1316,
"step": 915
},
{
"epoch": 12.214046822742475,
"grad_norm": 0.6226193904876709,
"learning_rate": 0.00015951056729699667,
"loss": 3.0897,
"step": 916
},
{
"epoch": 12.22742474916388,
"grad_norm": 0.6268473267555237,
"learning_rate": 0.00015946607341490546,
"loss": 3.2036,
"step": 917
},
{
"epoch": 12.240802675585284,
"grad_norm": 0.7825391292572021,
"learning_rate": 0.00015942157953281424,
"loss": 3.0082,
"step": 918
},
{
"epoch": 12.25418060200669,
"grad_norm": 0.7881148457527161,
"learning_rate": 0.00015937708565072302,
"loss": 3.0779,
"step": 919
},
{
"epoch": 12.267558528428093,
"grad_norm": 0.6725586652755737,
"learning_rate": 0.00015933259176863183,
"loss": 3.04,
"step": 920
},
{
"epoch": 12.280936454849499,
"grad_norm": 0.5831689238548279,
"learning_rate": 0.0001592880978865406,
"loss": 3.3319,
"step": 921
},
{
"epoch": 12.294314381270903,
"grad_norm": 0.6057907342910767,
"learning_rate": 0.0001592436040044494,
"loss": 3.0869,
"step": 922
},
{
"epoch": 12.307692307692308,
"grad_norm": 0.771857500076294,
"learning_rate": 0.00015919911012235818,
"loss": 3.1472,
"step": 923
},
{
"epoch": 12.321070234113712,
"grad_norm": 0.7447528839111328,
"learning_rate": 0.000159154616240267,
"loss": 3.174,
"step": 924
},
{
"epoch": 12.334448160535118,
"grad_norm": 0.5772632956504822,
"learning_rate": 0.00015911012235817574,
"loss": 3.1767,
"step": 925
},
{
"epoch": 12.347826086956522,
"grad_norm": 0.6952618956565857,
"learning_rate": 0.00015906562847608455,
"loss": 3.1597,
"step": 926
},
{
"epoch": 12.361204013377927,
"grad_norm": 0.600922703742981,
"learning_rate": 0.00015902113459399333,
"loss": 3.3612,
"step": 927
},
{
"epoch": 12.37458193979933,
"grad_norm": 0.7571472525596619,
"learning_rate": 0.00015897664071190212,
"loss": 2.9405,
"step": 928
},
{
"epoch": 12.387959866220736,
"grad_norm": 0.9343985915184021,
"learning_rate": 0.0001589321468298109,
"loss": 3.2886,
"step": 929
},
{
"epoch": 12.40133779264214,
"grad_norm": 0.7046729922294617,
"learning_rate": 0.0001588876529477197,
"loss": 3.4421,
"step": 930
},
{
"epoch": 12.414715719063546,
"grad_norm": 0.7591777443885803,
"learning_rate": 0.0001588431590656285,
"loss": 3.005,
"step": 931
},
{
"epoch": 12.42809364548495,
"grad_norm": 0.6508903503417969,
"learning_rate": 0.00015879866518353727,
"loss": 2.8554,
"step": 932
},
{
"epoch": 12.441471571906355,
"grad_norm": 0.6557784676551819,
"learning_rate": 0.00015875417130144606,
"loss": 3.3268,
"step": 933
},
{
"epoch": 12.454849498327759,
"grad_norm": 0.6941578984260559,
"learning_rate": 0.00015870967741935487,
"loss": 3.2088,
"step": 934
},
{
"epoch": 12.468227424749164,
"grad_norm": 0.6824263334274292,
"learning_rate": 0.00015866518353726362,
"loss": 3.0897,
"step": 935
},
{
"epoch": 12.481605351170568,
"grad_norm": 0.7324599027633667,
"learning_rate": 0.00015862068965517243,
"loss": 3.1443,
"step": 936
},
{
"epoch": 12.494983277591974,
"grad_norm": 0.577022135257721,
"learning_rate": 0.0001585761957730812,
"loss": 3.0896,
"step": 937
},
{
"epoch": 12.508361204013378,
"grad_norm": 0.6165060997009277,
"learning_rate": 0.00015853170189099,
"loss": 2.7546,
"step": 938
},
{
"epoch": 12.521739130434783,
"grad_norm": 0.561906635761261,
"learning_rate": 0.00015848720800889878,
"loss": 3.4192,
"step": 939
},
{
"epoch": 12.535117056856187,
"grad_norm": 0.5894923806190491,
"learning_rate": 0.0001584427141268076,
"loss": 3.0388,
"step": 940
},
{
"epoch": 12.548494983277592,
"grad_norm": 0.6261674761772156,
"learning_rate": 0.00015839822024471637,
"loss": 3.0705,
"step": 941
},
{
"epoch": 12.561872909698996,
"grad_norm": 0.695101261138916,
"learning_rate": 0.00015835372636262515,
"loss": 3.1684,
"step": 942
},
{
"epoch": 12.575250836120402,
"grad_norm": 0.6176817417144775,
"learning_rate": 0.00015830923248053393,
"loss": 3.0708,
"step": 943
},
{
"epoch": 12.588628762541806,
"grad_norm": 0.6548507213592529,
"learning_rate": 0.00015826473859844274,
"loss": 3.1569,
"step": 944
},
{
"epoch": 12.602006688963211,
"grad_norm": 0.6046382188796997,
"learning_rate": 0.0001582202447163515,
"loss": 3.2479,
"step": 945
},
{
"epoch": 12.615384615384615,
"grad_norm": 0.7103912234306335,
"learning_rate": 0.0001581757508342603,
"loss": 3.127,
"step": 946
},
{
"epoch": 12.62876254180602,
"grad_norm": 0.7131765484809875,
"learning_rate": 0.0001581312569521691,
"loss": 3.0975,
"step": 947
},
{
"epoch": 12.642140468227424,
"grad_norm": 0.6442859768867493,
"learning_rate": 0.00015808676307007787,
"loss": 3.2885,
"step": 948
},
{
"epoch": 12.65551839464883,
"grad_norm": 0.6430050134658813,
"learning_rate": 0.00015804226918798666,
"loss": 3.0397,
"step": 949
},
{
"epoch": 12.668896321070234,
"grad_norm": 0.6894303560256958,
"learning_rate": 0.00015799777530589546,
"loss": 3.0756,
"step": 950
},
{
"epoch": 12.68227424749164,
"grad_norm": 0.7319600582122803,
"learning_rate": 0.00015795328142380425,
"loss": 3.0457,
"step": 951
},
{
"epoch": 12.695652173913043,
"grad_norm": 0.6445140838623047,
"learning_rate": 0.00015790878754171303,
"loss": 2.9828,
"step": 952
},
{
"epoch": 12.709030100334449,
"grad_norm": 0.7522070407867432,
"learning_rate": 0.0001578642936596218,
"loss": 2.8942,
"step": 953
},
{
"epoch": 12.722408026755852,
"grad_norm": 0.7962691783905029,
"learning_rate": 0.00015781979977753062,
"loss": 3.0985,
"step": 954
},
{
"epoch": 12.735785953177258,
"grad_norm": 0.6391687393188477,
"learning_rate": 0.00015777530589543938,
"loss": 3.1752,
"step": 955
},
{
"epoch": 12.749163879598662,
"grad_norm": 0.7632976174354553,
"learning_rate": 0.00015773081201334816,
"loss": 3.3505,
"step": 956
},
{
"epoch": 12.762541806020067,
"grad_norm": 0.7491022944450378,
"learning_rate": 0.00015768631813125697,
"loss": 3.0721,
"step": 957
},
{
"epoch": 12.775919732441471,
"grad_norm": 0.6163421273231506,
"learning_rate": 0.00015764182424916572,
"loss": 3.3242,
"step": 958
},
{
"epoch": 12.789297658862877,
"grad_norm": 0.6831198334693909,
"learning_rate": 0.00015759733036707453,
"loss": 3.1409,
"step": 959
},
{
"epoch": 12.80267558528428,
"grad_norm": 0.812300980091095,
"learning_rate": 0.00015755283648498332,
"loss": 2.9606,
"step": 960
},
{
"epoch": 12.816053511705686,
"grad_norm": 0.6904334425926208,
"learning_rate": 0.00015750834260289213,
"loss": 3.0398,
"step": 961
},
{
"epoch": 12.82943143812709,
"grad_norm": 0.6349720358848572,
"learning_rate": 0.00015746384872080088,
"loss": 3.1033,
"step": 962
},
{
"epoch": 12.842809364548495,
"grad_norm": 0.6837566494941711,
"learning_rate": 0.0001574193548387097,
"loss": 3.2353,
"step": 963
},
{
"epoch": 12.856187290969899,
"grad_norm": 0.5852749943733215,
"learning_rate": 0.00015737486095661847,
"loss": 3.0972,
"step": 964
},
{
"epoch": 12.869565217391305,
"grad_norm": 0.6641372442245483,
"learning_rate": 0.00015733036707452725,
"loss": 3.2243,
"step": 965
},
{
"epoch": 12.882943143812708,
"grad_norm": 0.6613900065422058,
"learning_rate": 0.00015728587319243604,
"loss": 3.1263,
"step": 966
},
{
"epoch": 12.896321070234114,
"grad_norm": 0.6126120090484619,
"learning_rate": 0.00015724137931034485,
"loss": 3.069,
"step": 967
},
{
"epoch": 12.909698996655518,
"grad_norm": 0.6764604449272156,
"learning_rate": 0.0001571968854282536,
"loss": 3.1397,
"step": 968
},
{
"epoch": 12.923076923076923,
"grad_norm": 0.6447578072547913,
"learning_rate": 0.0001571523915461624,
"loss": 3.1839,
"step": 969
},
{
"epoch": 12.936454849498327,
"grad_norm": 0.5872016549110413,
"learning_rate": 0.0001571078976640712,
"loss": 3.3144,
"step": 970
},
{
"epoch": 12.949832775919733,
"grad_norm": 0.626276969909668,
"learning_rate": 0.00015706340378198,
"loss": 3.1295,
"step": 971
},
{
"epoch": 12.963210702341136,
"grad_norm": 0.6829231381416321,
"learning_rate": 0.00015701890989988876,
"loss": 3.2261,
"step": 972
},
{
"epoch": 12.976588628762542,
"grad_norm": 0.6197345852851868,
"learning_rate": 0.00015697441601779757,
"loss": 3.1117,
"step": 973
},
{
"epoch": 12.989966555183946,
"grad_norm": 0.6137062907218933,
"learning_rate": 0.00015692992213570635,
"loss": 3.1548,
"step": 974
},
{
"epoch": 13.0,
"grad_norm": 0.7483121752738953,
"learning_rate": 0.00015688542825361513,
"loss": 3.4944,
"step": 975
},
{
"epoch": 13.013377926421406,
"grad_norm": 0.6102525591850281,
"learning_rate": 0.00015684093437152392,
"loss": 3.1403,
"step": 976
},
{
"epoch": 13.02675585284281,
"grad_norm": 0.7258747220039368,
"learning_rate": 0.00015679644048943272,
"loss": 2.8474,
"step": 977
},
{
"epoch": 13.040133779264215,
"grad_norm": 0.6891087293624878,
"learning_rate": 0.00015675194660734148,
"loss": 2.9722,
"step": 978
},
{
"epoch": 13.053511705685619,
"grad_norm": 0.6320910453796387,
"learning_rate": 0.0001567074527252503,
"loss": 2.9807,
"step": 979
},
{
"epoch": 13.066889632107024,
"grad_norm": 0.8684266209602356,
"learning_rate": 0.00015666295884315907,
"loss": 2.8448,
"step": 980
},
{
"epoch": 13.080267558528428,
"grad_norm": 0.7126099467277527,
"learning_rate": 0.00015661846496106788,
"loss": 3.1915,
"step": 981
},
{
"epoch": 13.093645484949834,
"grad_norm": 0.7001529335975647,
"learning_rate": 0.00015657397107897664,
"loss": 3.2017,
"step": 982
},
{
"epoch": 13.107023411371237,
"grad_norm": 0.7901191711425781,
"learning_rate": 0.00015652947719688545,
"loss": 3.1273,
"step": 983
},
{
"epoch": 13.120401337792643,
"grad_norm": 0.5769410729408264,
"learning_rate": 0.00015648498331479423,
"loss": 2.8582,
"step": 984
},
{
"epoch": 13.133779264214047,
"grad_norm": 0.5969700813293457,
"learning_rate": 0.000156440489432703,
"loss": 3.118,
"step": 985
},
{
"epoch": 13.147157190635452,
"grad_norm": 0.5789377093315125,
"learning_rate": 0.0001563959955506118,
"loss": 2.8121,
"step": 986
},
{
"epoch": 13.160535117056856,
"grad_norm": 0.7945278882980347,
"learning_rate": 0.0001563515016685206,
"loss": 2.9212,
"step": 987
},
{
"epoch": 13.173913043478262,
"grad_norm": 0.60884690284729,
"learning_rate": 0.00015630700778642936,
"loss": 3.1993,
"step": 988
},
{
"epoch": 13.187290969899665,
"grad_norm": 0.616142213344574,
"learning_rate": 0.00015626251390433817,
"loss": 3.107,
"step": 989
},
{
"epoch": 13.200668896321071,
"grad_norm": 0.6428812146186829,
"learning_rate": 0.00015621802002224695,
"loss": 3.1401,
"step": 990
},
{
"epoch": 13.214046822742475,
"grad_norm": 0.5723693370819092,
"learning_rate": 0.00015617352614015576,
"loss": 3.0474,
"step": 991
},
{
"epoch": 13.22742474916388,
"grad_norm": 0.5820907950401306,
"learning_rate": 0.00015612903225806451,
"loss": 3.2211,
"step": 992
},
{
"epoch": 13.240802675585284,
"grad_norm": 0.555957555770874,
"learning_rate": 0.00015608453837597332,
"loss": 3.1306,
"step": 993
},
{
"epoch": 13.25418060200669,
"grad_norm": 0.528698205947876,
"learning_rate": 0.0001560400444938821,
"loss": 3.0989,
"step": 994
},
{
"epoch": 13.267558528428093,
"grad_norm": 0.5959749817848206,
"learning_rate": 0.0001559955506117909,
"loss": 2.9128,
"step": 995
},
{
"epoch": 13.280936454849499,
"grad_norm": 0.6702240705490112,
"learning_rate": 0.00015595105672969967,
"loss": 2.9969,
"step": 996
},
{
"epoch": 13.294314381270903,
"grad_norm": 0.5363825559616089,
"learning_rate": 0.00015590656284760845,
"loss": 3.2027,
"step": 997
},
{
"epoch": 13.307692307692308,
"grad_norm": 0.6402661204338074,
"learning_rate": 0.00015586206896551724,
"loss": 3.2733,
"step": 998
},
{
"epoch": 13.321070234113712,
"grad_norm": 0.700517475605011,
"learning_rate": 0.00015581757508342602,
"loss": 3.2629,
"step": 999
},
{
"epoch": 13.334448160535118,
"grad_norm": 0.5500949621200562,
"learning_rate": 0.00015577308120133483,
"loss": 3.172,
"step": 1000
},
{
"epoch": 13.347826086956522,
"grad_norm": 0.6199147701263428,
"learning_rate": 0.0001557285873192436,
"loss": 3.0556,
"step": 1001
},
{
"epoch": 13.361204013377927,
"grad_norm": 0.5900529026985168,
"learning_rate": 0.0001556840934371524,
"loss": 3.1248,
"step": 1002
},
{
"epoch": 13.37458193979933,
"grad_norm": 0.7272413372993469,
"learning_rate": 0.00015563959955506118,
"loss": 3.1142,
"step": 1003
},
{
"epoch": 13.387959866220736,
"grad_norm": 0.6461951732635498,
"learning_rate": 0.00015559510567296998,
"loss": 3.1145,
"step": 1004
},
{
"epoch": 13.40133779264214,
"grad_norm": 0.5750373005867004,
"learning_rate": 0.00015555061179087874,
"loss": 3.2117,
"step": 1005
},
{
"epoch": 13.414715719063546,
"grad_norm": 0.6486302614212036,
"learning_rate": 0.00015550611790878755,
"loss": 2.9018,
"step": 1006
},
{
"epoch": 13.42809364548495,
"grad_norm": 0.6897476315498352,
"learning_rate": 0.00015546162402669633,
"loss": 3.2987,
"step": 1007
},
{
"epoch": 13.441471571906355,
"grad_norm": 0.5997576713562012,
"learning_rate": 0.00015541713014460511,
"loss": 2.974,
"step": 1008
},
{
"epoch": 13.454849498327759,
"grad_norm": 0.6484793424606323,
"learning_rate": 0.0001553726362625139,
"loss": 2.9827,
"step": 1009
},
{
"epoch": 13.468227424749164,
"grad_norm": 0.562312126159668,
"learning_rate": 0.0001553281423804227,
"loss": 3.2243,
"step": 1010
},
{
"epoch": 13.481605351170568,
"grad_norm": 0.7257137298583984,
"learning_rate": 0.0001552836484983315,
"loss": 3.1081,
"step": 1011
},
{
"epoch": 13.494983277591974,
"grad_norm": 0.7201404571533203,
"learning_rate": 0.00015523915461624027,
"loss": 3.1468,
"step": 1012
},
{
"epoch": 13.508361204013378,
"grad_norm": 0.666539192199707,
"learning_rate": 0.00015519466073414905,
"loss": 3.0081,
"step": 1013
},
{
"epoch": 13.521739130434783,
"grad_norm": 0.6867642998695374,
"learning_rate": 0.00015515016685205786,
"loss": 2.7889,
"step": 1014
},
{
"epoch": 13.535117056856187,
"grad_norm": 0.5799785256385803,
"learning_rate": 0.00015510567296996662,
"loss": 3.2995,
"step": 1015
},
{
"epoch": 13.548494983277592,
"grad_norm": 0.6155371069908142,
"learning_rate": 0.00015506117908787543,
"loss": 3.2227,
"step": 1016
},
{
"epoch": 13.561872909698996,
"grad_norm": 0.7604040503501892,
"learning_rate": 0.0001550166852057842,
"loss": 3.0384,
"step": 1017
},
{
"epoch": 13.575250836120402,
"grad_norm": 0.8445917963981628,
"learning_rate": 0.000154972191323693,
"loss": 2.9566,
"step": 1018
},
{
"epoch": 13.588628762541806,
"grad_norm": 0.7978566288948059,
"learning_rate": 0.00015492769744160177,
"loss": 3.0175,
"step": 1019
},
{
"epoch": 13.602006688963211,
"grad_norm": 0.5899437069892883,
"learning_rate": 0.00015488320355951058,
"loss": 3.2418,
"step": 1020
},
{
"epoch": 13.615384615384615,
"grad_norm": 0.7204627990722656,
"learning_rate": 0.00015483870967741937,
"loss": 3.1568,
"step": 1021
},
{
"epoch": 13.62876254180602,
"grad_norm": 0.6504855155944824,
"learning_rate": 0.00015479421579532815,
"loss": 2.8952,
"step": 1022
},
{
"epoch": 13.642140468227424,
"grad_norm": 0.8101251125335693,
"learning_rate": 0.00015474972191323693,
"loss": 3.1975,
"step": 1023
},
{
"epoch": 13.65551839464883,
"grad_norm": 0.6161416172981262,
"learning_rate": 0.00015470522803114574,
"loss": 3.1565,
"step": 1024
},
{
"epoch": 13.668896321070234,
"grad_norm": 0.6131258606910706,
"learning_rate": 0.0001546607341490545,
"loss": 3.0382,
"step": 1025
},
{
"epoch": 13.68227424749164,
"grad_norm": 0.8008583784103394,
"learning_rate": 0.0001546162402669633,
"loss": 2.995,
"step": 1026
},
{
"epoch": 13.695652173913043,
"grad_norm": 0.7101227045059204,
"learning_rate": 0.0001545717463848721,
"loss": 3.0704,
"step": 1027
},
{
"epoch": 13.709030100334449,
"grad_norm": 0.7988458275794983,
"learning_rate": 0.00015452725250278087,
"loss": 3.1424,
"step": 1028
},
{
"epoch": 13.722408026755852,
"grad_norm": 0.6013655662536621,
"learning_rate": 0.00015448275862068965,
"loss": 2.9986,
"step": 1029
},
{
"epoch": 13.735785953177258,
"grad_norm": 0.6368236541748047,
"learning_rate": 0.00015443826473859846,
"loss": 3.0193,
"step": 1030
},
{
"epoch": 13.749163879598662,
"grad_norm": 0.8222694396972656,
"learning_rate": 0.00015439377085650724,
"loss": 3.0888,
"step": 1031
},
{
"epoch": 13.762541806020067,
"grad_norm": 0.7270404696464539,
"learning_rate": 0.00015434927697441603,
"loss": 3.0929,
"step": 1032
},
{
"epoch": 13.775919732441471,
"grad_norm": 0.7292355298995972,
"learning_rate": 0.0001543047830923248,
"loss": 2.7676,
"step": 1033
},
{
"epoch": 13.789297658862877,
"grad_norm": 0.6662157773971558,
"learning_rate": 0.00015426028921023362,
"loss": 3.1984,
"step": 1034
},
{
"epoch": 13.80267558528428,
"grad_norm": 0.6350163817405701,
"learning_rate": 0.00015421579532814237,
"loss": 3.3042,
"step": 1035
},
{
"epoch": 13.816053511705686,
"grad_norm": 0.5999907851219177,
"learning_rate": 0.00015417130144605118,
"loss": 3.0983,
"step": 1036
},
{
"epoch": 13.82943143812709,
"grad_norm": 0.5942257642745972,
"learning_rate": 0.00015412680756395997,
"loss": 3.0474,
"step": 1037
},
{
"epoch": 13.842809364548495,
"grad_norm": 0.662589430809021,
"learning_rate": 0.00015408231368186875,
"loss": 2.9251,
"step": 1038
},
{
"epoch": 13.856187290969899,
"grad_norm": 0.5817089080810547,
"learning_rate": 0.00015403781979977753,
"loss": 3.0716,
"step": 1039
},
{
"epoch": 13.869565217391305,
"grad_norm": 0.6019257307052612,
"learning_rate": 0.0001539933259176863,
"loss": 3.1754,
"step": 1040
},
{
"epoch": 13.882943143812708,
"grad_norm": 0.6301860213279724,
"learning_rate": 0.00015394883203559512,
"loss": 3.2066,
"step": 1041
},
{
"epoch": 13.896321070234114,
"grad_norm": 0.6468888521194458,
"learning_rate": 0.00015390433815350388,
"loss": 3.3001,
"step": 1042
},
{
"epoch": 13.909698996655518,
"grad_norm": 0.6510801911354065,
"learning_rate": 0.0001538598442714127,
"loss": 3.2198,
"step": 1043
},
{
"epoch": 13.923076923076923,
"grad_norm": 0.5692014694213867,
"learning_rate": 0.00015381535038932147,
"loss": 3.3257,
"step": 1044
},
{
"epoch": 13.936454849498327,
"grad_norm": 0.594219982624054,
"learning_rate": 0.00015377085650723025,
"loss": 2.9918,
"step": 1045
},
{
"epoch": 13.949832775919733,
"grad_norm": 0.6501769423484802,
"learning_rate": 0.00015372636262513903,
"loss": 2.9653,
"step": 1046
},
{
"epoch": 13.963210702341136,
"grad_norm": 0.6310623288154602,
"learning_rate": 0.00015368186874304784,
"loss": 3.19,
"step": 1047
},
{
"epoch": 13.976588628762542,
"grad_norm": 0.5795436501502991,
"learning_rate": 0.00015363737486095663,
"loss": 3.272,
"step": 1048
},
{
"epoch": 13.989966555183946,
"grad_norm": 0.5421392917633057,
"learning_rate": 0.0001535928809788654,
"loss": 3.2109,
"step": 1049
},
{
"epoch": 14.0,
"grad_norm": 0.833959698677063,
"learning_rate": 0.0001535483870967742,
"loss": 3.3017,
"step": 1050
},
{
"epoch": 14.013377926421406,
"grad_norm": 0.6856208443641663,
"learning_rate": 0.000153503893214683,
"loss": 3.0784,
"step": 1051
},
{
"epoch": 14.02675585284281,
"grad_norm": 0.5841811895370483,
"learning_rate": 0.00015345939933259176,
"loss": 2.848,
"step": 1052
},
{
"epoch": 14.040133779264215,
"grad_norm": 0.557906985282898,
"learning_rate": 0.00015341490545050057,
"loss": 3.1564,
"step": 1053
},
{
"epoch": 14.053511705685619,
"grad_norm": 0.5468619465827942,
"learning_rate": 0.00015337041156840935,
"loss": 3.1237,
"step": 1054
},
{
"epoch": 14.066889632107024,
"grad_norm": 0.7213225364685059,
"learning_rate": 0.00015332591768631813,
"loss": 2.8993,
"step": 1055
},
{
"epoch": 14.080267558528428,
"grad_norm": 0.7413175106048584,
"learning_rate": 0.0001532814238042269,
"loss": 3.0028,
"step": 1056
},
{
"epoch": 14.093645484949834,
"grad_norm": 0.6072244644165039,
"learning_rate": 0.00015323692992213572,
"loss": 2.8534,
"step": 1057
},
{
"epoch": 14.107023411371237,
"grad_norm": 0.683262288570404,
"learning_rate": 0.0001531924360400445,
"loss": 2.8982,
"step": 1058
},
{
"epoch": 14.120401337792643,
"grad_norm": 0.5880157351493835,
"learning_rate": 0.0001531479421579533,
"loss": 2.9567,
"step": 1059
},
{
"epoch": 14.133779264214047,
"grad_norm": 0.7519298195838928,
"learning_rate": 0.00015310344827586207,
"loss": 2.9615,
"step": 1060
},
{
"epoch": 14.147157190635452,
"grad_norm": 0.7747945189476013,
"learning_rate": 0.00015305895439377088,
"loss": 2.7823,
"step": 1061
},
{
"epoch": 14.160535117056856,
"grad_norm": 0.6560395956039429,
"learning_rate": 0.00015301446051167963,
"loss": 3.1791,
"step": 1062
},
{
"epoch": 14.173913043478262,
"grad_norm": 0.6388076543807983,
"learning_rate": 0.00015296996662958844,
"loss": 2.9868,
"step": 1063
},
{
"epoch": 14.187290969899665,
"grad_norm": 0.7349525690078735,
"learning_rate": 0.00015292547274749723,
"loss": 2.9291,
"step": 1064
},
{
"epoch": 14.200668896321071,
"grad_norm": 0.7184433341026306,
"learning_rate": 0.000152880978865406,
"loss": 2.7889,
"step": 1065
},
{
"epoch": 14.214046822742475,
"grad_norm": 0.6776930093765259,
"learning_rate": 0.0001528364849833148,
"loss": 3.1572,
"step": 1066
},
{
"epoch": 14.22742474916388,
"grad_norm": 0.818756103515625,
"learning_rate": 0.0001527919911012236,
"loss": 2.767,
"step": 1067
},
{
"epoch": 14.240802675585284,
"grad_norm": 0.6005066633224487,
"learning_rate": 0.00015274749721913238,
"loss": 3.1796,
"step": 1068
},
{
"epoch": 14.25418060200669,
"grad_norm": 0.6367926001548767,
"learning_rate": 0.00015270300333704117,
"loss": 2.9999,
"step": 1069
},
{
"epoch": 14.267558528428093,
"grad_norm": 0.6823679208755493,
"learning_rate": 0.00015265850945494995,
"loss": 2.7663,
"step": 1070
},
{
"epoch": 14.280936454849499,
"grad_norm": 0.6238808631896973,
"learning_rate": 0.00015261401557285876,
"loss": 3.1908,
"step": 1071
},
{
"epoch": 14.294314381270903,
"grad_norm": 0.6983721256256104,
"learning_rate": 0.0001525695216907675,
"loss": 3.0264,
"step": 1072
},
{
"epoch": 14.307692307692308,
"grad_norm": 0.7568501234054565,
"learning_rate": 0.00015252502780867632,
"loss": 3.0474,
"step": 1073
},
{
"epoch": 14.321070234113712,
"grad_norm": 0.6250051259994507,
"learning_rate": 0.0001524805339265851,
"loss": 3.114,
"step": 1074
},
{
"epoch": 14.334448160535118,
"grad_norm": 0.5907386541366577,
"learning_rate": 0.0001524360400444939,
"loss": 3.1823,
"step": 1075
},
{
"epoch": 14.347826086956522,
"grad_norm": 0.6719332337379456,
"learning_rate": 0.00015239154616240267,
"loss": 2.994,
"step": 1076
},
{
"epoch": 14.361204013377927,
"grad_norm": 0.5911534428596497,
"learning_rate": 0.00015234705228031148,
"loss": 3.4178,
"step": 1077
},
{
"epoch": 14.37458193979933,
"grad_norm": 0.8071689009666443,
"learning_rate": 0.00015230255839822026,
"loss": 2.9633,
"step": 1078
},
{
"epoch": 14.387959866220736,
"grad_norm": 0.5957038998603821,
"learning_rate": 0.00015225806451612902,
"loss": 3.1126,
"step": 1079
},
{
"epoch": 14.40133779264214,
"grad_norm": 0.6604459285736084,
"learning_rate": 0.00015221357063403783,
"loss": 2.9193,
"step": 1080
},
{
"epoch": 14.414715719063546,
"grad_norm": 0.626081109046936,
"learning_rate": 0.0001521690767519466,
"loss": 2.9321,
"step": 1081
},
{
"epoch": 14.42809364548495,
"grad_norm": 0.5767174959182739,
"learning_rate": 0.0001521245828698554,
"loss": 3.1518,
"step": 1082
},
{
"epoch": 14.441471571906355,
"grad_norm": 0.6444874405860901,
"learning_rate": 0.00015208008898776417,
"loss": 3.1948,
"step": 1083
},
{
"epoch": 14.454849498327759,
"grad_norm": 0.668171763420105,
"learning_rate": 0.00015203559510567298,
"loss": 3.016,
"step": 1084
},
{
"epoch": 14.468227424749164,
"grad_norm": 0.6998944878578186,
"learning_rate": 0.00015199110122358176,
"loss": 2.9558,
"step": 1085
},
{
"epoch": 14.481605351170568,
"grad_norm": 0.5896235704421997,
"learning_rate": 0.00015194660734149055,
"loss": 3.0997,
"step": 1086
},
{
"epoch": 14.494983277591974,
"grad_norm": 0.6724826097488403,
"learning_rate": 0.00015190211345939933,
"loss": 2.9369,
"step": 1087
},
{
"epoch": 14.508361204013378,
"grad_norm": 0.5710486769676208,
"learning_rate": 0.00015185761957730814,
"loss": 3.4724,
"step": 1088
},
{
"epoch": 14.521739130434783,
"grad_norm": 0.9997962117195129,
"learning_rate": 0.0001518131256952169,
"loss": 2.9058,
"step": 1089
},
{
"epoch": 14.535117056856187,
"grad_norm": 0.668074905872345,
"learning_rate": 0.0001517686318131257,
"loss": 3.2401,
"step": 1090
},
{
"epoch": 14.548494983277592,
"grad_norm": 0.6180433630943298,
"learning_rate": 0.00015172413793103449,
"loss": 2.8975,
"step": 1091
},
{
"epoch": 14.561872909698996,
"grad_norm": 0.6412661075592041,
"learning_rate": 0.00015167964404894327,
"loss": 2.9882,
"step": 1092
},
{
"epoch": 14.575250836120402,
"grad_norm": 0.715288519859314,
"learning_rate": 0.00015163515016685205,
"loss": 3.4004,
"step": 1093
},
{
"epoch": 14.588628762541806,
"grad_norm": 0.689164400100708,
"learning_rate": 0.00015159065628476086,
"loss": 3.0705,
"step": 1094
},
{
"epoch": 14.602006688963211,
"grad_norm": 0.7713497281074524,
"learning_rate": 0.00015154616240266964,
"loss": 3.159,
"step": 1095
},
{
"epoch": 14.615384615384615,
"grad_norm": 0.6725841164588928,
"learning_rate": 0.00015150166852057843,
"loss": 2.8938,
"step": 1096
},
{
"epoch": 14.62876254180602,
"grad_norm": 0.658108651638031,
"learning_rate": 0.0001514571746384872,
"loss": 2.8747,
"step": 1097
},
{
"epoch": 14.642140468227424,
"grad_norm": 0.5711888074874878,
"learning_rate": 0.00015141268075639602,
"loss": 2.8989,
"step": 1098
},
{
"epoch": 14.65551839464883,
"grad_norm": 0.6184161305427551,
"learning_rate": 0.00015136818687430477,
"loss": 3.0904,
"step": 1099
},
{
"epoch": 14.668896321070234,
"grad_norm": 0.5937799215316772,
"learning_rate": 0.00015132369299221358,
"loss": 3.1637,
"step": 1100
},
{
"epoch": 14.68227424749164,
"grad_norm": 0.591673731803894,
"learning_rate": 0.00015127919911012236,
"loss": 2.9547,
"step": 1101
},
{
"epoch": 14.695652173913043,
"grad_norm": 0.7317401170730591,
"learning_rate": 0.00015123470522803115,
"loss": 3.2043,
"step": 1102
},
{
"epoch": 14.709030100334449,
"grad_norm": 0.5784003734588623,
"learning_rate": 0.00015119021134593993,
"loss": 3.102,
"step": 1103
},
{
"epoch": 14.722408026755852,
"grad_norm": 0.7077385187149048,
"learning_rate": 0.00015114571746384874,
"loss": 3.011,
"step": 1104
},
{
"epoch": 14.735785953177258,
"grad_norm": 0.6472675204277039,
"learning_rate": 0.00015110122358175752,
"loss": 3.2075,
"step": 1105
},
{
"epoch": 14.749163879598662,
"grad_norm": 0.6789306998252869,
"learning_rate": 0.0001510567296996663,
"loss": 2.9458,
"step": 1106
},
{
"epoch": 14.762541806020067,
"grad_norm": 0.6602732539176941,
"learning_rate": 0.00015101223581757509,
"loss": 2.9941,
"step": 1107
},
{
"epoch": 14.775919732441471,
"grad_norm": 0.7484832406044006,
"learning_rate": 0.0001509677419354839,
"loss": 3.0358,
"step": 1108
},
{
"epoch": 14.789297658862877,
"grad_norm": 0.704139769077301,
"learning_rate": 0.00015092324805339265,
"loss": 3.1006,
"step": 1109
},
{
"epoch": 14.80267558528428,
"grad_norm": 0.6545978784561157,
"learning_rate": 0.00015087875417130146,
"loss": 3.0369,
"step": 1110
},
{
"epoch": 14.816053511705686,
"grad_norm": 0.5718163847923279,
"learning_rate": 0.00015083426028921024,
"loss": 3.1683,
"step": 1111
},
{
"epoch": 14.82943143812709,
"grad_norm": 0.5773367285728455,
"learning_rate": 0.00015078976640711902,
"loss": 3.2753,
"step": 1112
},
{
"epoch": 14.842809364548495,
"grad_norm": 0.6617185473442078,
"learning_rate": 0.0001507452725250278,
"loss": 2.9713,
"step": 1113
},
{
"epoch": 14.856187290969899,
"grad_norm": 0.6748194098472595,
"learning_rate": 0.00015070077864293662,
"loss": 3.0961,
"step": 1114
},
{
"epoch": 14.869565217391305,
"grad_norm": 0.6942034959793091,
"learning_rate": 0.0001506562847608454,
"loss": 3.0778,
"step": 1115
},
{
"epoch": 14.882943143812708,
"grad_norm": 1.0203640460968018,
"learning_rate": 0.00015061179087875418,
"loss": 3.0705,
"step": 1116
},
{
"epoch": 14.896321070234114,
"grad_norm": 0.5746601223945618,
"learning_rate": 0.00015056729699666296,
"loss": 3.1204,
"step": 1117
},
{
"epoch": 14.909698996655518,
"grad_norm": 0.7374005317687988,
"learning_rate": 0.00015052280311457177,
"loss": 3.1289,
"step": 1118
},
{
"epoch": 14.923076923076923,
"grad_norm": 0.5524411201477051,
"learning_rate": 0.00015047830923248053,
"loss": 3.2795,
"step": 1119
},
{
"epoch": 14.936454849498327,
"grad_norm": 0.7024741768836975,
"learning_rate": 0.0001504338153503893,
"loss": 3.0675,
"step": 1120
},
{
"epoch": 14.949832775919733,
"grad_norm": 0.7431137561798096,
"learning_rate": 0.00015038932146829812,
"loss": 3.1222,
"step": 1121
},
{
"epoch": 14.963210702341136,
"grad_norm": 0.6568113565444946,
"learning_rate": 0.0001503448275862069,
"loss": 3.1523,
"step": 1122
},
{
"epoch": 14.976588628762542,
"grad_norm": 0.6193330883979797,
"learning_rate": 0.00015030033370411569,
"loss": 3.1632,
"step": 1123
},
{
"epoch": 14.989966555183946,
"grad_norm": 0.6371363401412964,
"learning_rate": 0.00015025583982202447,
"loss": 3.0525,
"step": 1124
},
{
"epoch": 15.0,
"grad_norm": 0.652542233467102,
"learning_rate": 0.00015021134593993328,
"loss": 3.1846,
"step": 1125
},
{
"epoch": 15.013377926421406,
"grad_norm": 0.6337831616401672,
"learning_rate": 0.00015016685205784203,
"loss": 2.9292,
"step": 1126
},
{
"epoch": 15.02675585284281,
"grad_norm": 0.85350501537323,
"learning_rate": 0.00015012235817575084,
"loss": 2.913,
"step": 1127
},
{
"epoch": 15.040133779264215,
"grad_norm": 0.6439313888549805,
"learning_rate": 0.00015007786429365962,
"loss": 3.0809,
"step": 1128
},
{
"epoch": 15.053511705685619,
"grad_norm": 0.5232247114181519,
"learning_rate": 0.0001500333704115684,
"loss": 3.1348,
"step": 1129
},
{
"epoch": 15.066889632107024,
"grad_norm": 0.6082741618156433,
"learning_rate": 0.0001499888765294772,
"loss": 2.9282,
"step": 1130
},
{
"epoch": 15.080267558528428,
"grad_norm": 0.5736444592475891,
"learning_rate": 0.000149944382647386,
"loss": 2.9891,
"step": 1131
},
{
"epoch": 15.093645484949834,
"grad_norm": 0.7732790112495422,
"learning_rate": 0.00014989988876529478,
"loss": 3.0461,
"step": 1132
},
{
"epoch": 15.107023411371237,
"grad_norm": 0.618357241153717,
"learning_rate": 0.00014985539488320356,
"loss": 3.0296,
"step": 1133
},
{
"epoch": 15.120401337792643,
"grad_norm": 0.7245836853981018,
"learning_rate": 0.00014981090100111235,
"loss": 3.0078,
"step": 1134
},
{
"epoch": 15.133779264214047,
"grad_norm": 0.6738787293434143,
"learning_rate": 0.00014976640711902116,
"loss": 3.1171,
"step": 1135
},
{
"epoch": 15.147157190635452,
"grad_norm": 0.5802761316299438,
"learning_rate": 0.0001497219132369299,
"loss": 3.1403,
"step": 1136
},
{
"epoch": 15.160535117056856,
"grad_norm": 0.5941367745399475,
"learning_rate": 0.00014967741935483872,
"loss": 2.9149,
"step": 1137
},
{
"epoch": 15.173913043478262,
"grad_norm": 0.8884940147399902,
"learning_rate": 0.0001496329254727475,
"loss": 2.8983,
"step": 1138
},
{
"epoch": 15.187290969899665,
"grad_norm": 0.7229192852973938,
"learning_rate": 0.00014958843159065628,
"loss": 2.9774,
"step": 1139
},
{
"epoch": 15.200668896321071,
"grad_norm": 0.6714467406272888,
"learning_rate": 0.00014954393770856507,
"loss": 2.9219,
"step": 1140
},
{
"epoch": 15.214046822742475,
"grad_norm": 0.6785704493522644,
"learning_rate": 0.00014949944382647388,
"loss": 3.0205,
"step": 1141
},
{
"epoch": 15.22742474916388,
"grad_norm": 0.6349677443504333,
"learning_rate": 0.00014945494994438266,
"loss": 3.1601,
"step": 1142
},
{
"epoch": 15.240802675585284,
"grad_norm": 0.557123064994812,
"learning_rate": 0.00014941045606229144,
"loss": 2.6297,
"step": 1143
},
{
"epoch": 15.25418060200669,
"grad_norm": 0.6714944243431091,
"learning_rate": 0.00014936596218020022,
"loss": 2.7951,
"step": 1144
},
{
"epoch": 15.267558528428093,
"grad_norm": 0.6747463345527649,
"learning_rate": 0.00014932146829810903,
"loss": 2.7909,
"step": 1145
},
{
"epoch": 15.280936454849499,
"grad_norm": 0.5717387199401855,
"learning_rate": 0.0001492769744160178,
"loss": 3.2896,
"step": 1146
},
{
"epoch": 15.294314381270903,
"grad_norm": 0.6589123010635376,
"learning_rate": 0.0001492324805339266,
"loss": 2.8332,
"step": 1147
},
{
"epoch": 15.307692307692308,
"grad_norm": 0.6273646950721741,
"learning_rate": 0.00014918798665183538,
"loss": 3.0084,
"step": 1148
},
{
"epoch": 15.321070234113712,
"grad_norm": 0.6551377773284912,
"learning_rate": 0.00014914349276974416,
"loss": 2.8147,
"step": 1149
},
{
"epoch": 15.334448160535118,
"grad_norm": 0.6751659512519836,
"learning_rate": 0.00014909899888765295,
"loss": 3.1345,
"step": 1150
},
{
"epoch": 15.347826086956522,
"grad_norm": 0.677094042301178,
"learning_rate": 0.00014905450500556175,
"loss": 3.1958,
"step": 1151
},
{
"epoch": 15.361204013377927,
"grad_norm": 0.6613426804542542,
"learning_rate": 0.00014901001112347054,
"loss": 3.081,
"step": 1152
},
{
"epoch": 15.37458193979933,
"grad_norm": 0.7645783424377441,
"learning_rate": 0.00014896551724137932,
"loss": 2.8799,
"step": 1153
},
{
"epoch": 15.387959866220736,
"grad_norm": 0.5698953866958618,
"learning_rate": 0.0001489210233592881,
"loss": 3.1691,
"step": 1154
},
{
"epoch": 15.40133779264214,
"grad_norm": 0.6581351161003113,
"learning_rate": 0.0001488765294771969,
"loss": 3.2365,
"step": 1155
},
{
"epoch": 15.414715719063546,
"grad_norm": 0.7809271812438965,
"learning_rate": 0.00014883203559510567,
"loss": 2.833,
"step": 1156
},
{
"epoch": 15.42809364548495,
"grad_norm": 0.6226280927658081,
"learning_rate": 0.00014878754171301448,
"loss": 3.1502,
"step": 1157
},
{
"epoch": 15.441471571906355,
"grad_norm": 0.5494824051856995,
"learning_rate": 0.00014874304783092326,
"loss": 3.2195,
"step": 1158
},
{
"epoch": 15.454849498327759,
"grad_norm": 0.5729116797447205,
"learning_rate": 0.00014869855394883204,
"loss": 2.9463,
"step": 1159
},
{
"epoch": 15.468227424749164,
"grad_norm": 0.6673750877380371,
"learning_rate": 0.00014865406006674082,
"loss": 3.1514,
"step": 1160
},
{
"epoch": 15.481605351170568,
"grad_norm": 0.6746686697006226,
"learning_rate": 0.0001486095661846496,
"loss": 3.0088,
"step": 1161
},
{
"epoch": 15.494983277591974,
"grad_norm": 0.6898564100265503,
"learning_rate": 0.00014856507230255842,
"loss": 2.8468,
"step": 1162
},
{
"epoch": 15.508361204013378,
"grad_norm": 0.7262438535690308,
"learning_rate": 0.00014852057842046717,
"loss": 2.8017,
"step": 1163
},
{
"epoch": 15.521739130434783,
"grad_norm": 0.6878666877746582,
"learning_rate": 0.00014847608453837598,
"loss": 2.9486,
"step": 1164
},
{
"epoch": 15.535117056856187,
"grad_norm": 0.6375080347061157,
"learning_rate": 0.00014843159065628476,
"loss": 3.0757,
"step": 1165
},
{
"epoch": 15.548494983277592,
"grad_norm": 0.6540268063545227,
"learning_rate": 0.00014838709677419355,
"loss": 3.1262,
"step": 1166
},
{
"epoch": 15.561872909698996,
"grad_norm": 0.6036689877510071,
"learning_rate": 0.00014834260289210233,
"loss": 3.0607,
"step": 1167
},
{
"epoch": 15.575250836120402,
"grad_norm": 0.5899893641471863,
"learning_rate": 0.00014829810901001114,
"loss": 3.1222,
"step": 1168
},
{
"epoch": 15.588628762541806,
"grad_norm": 0.7268028259277344,
"learning_rate": 0.00014825361512791992,
"loss": 3.1777,
"step": 1169
},
{
"epoch": 15.602006688963211,
"grad_norm": 0.6990141272544861,
"learning_rate": 0.0001482091212458287,
"loss": 3.2142,
"step": 1170
},
{
"epoch": 15.615384615384615,
"grad_norm": 0.6009657382965088,
"learning_rate": 0.00014816462736373748,
"loss": 3.141,
"step": 1171
},
{
"epoch": 15.62876254180602,
"grad_norm": 0.6287830471992493,
"learning_rate": 0.0001481201334816463,
"loss": 3.0621,
"step": 1172
},
{
"epoch": 15.642140468227424,
"grad_norm": 0.6720128655433655,
"learning_rate": 0.00014807563959955505,
"loss": 3.0353,
"step": 1173
},
{
"epoch": 15.65551839464883,
"grad_norm": 0.6694427132606506,
"learning_rate": 0.00014803114571746386,
"loss": 3.0171,
"step": 1174
},
{
"epoch": 15.668896321070234,
"grad_norm": 0.5630237460136414,
"learning_rate": 0.00014798665183537264,
"loss": 3.1444,
"step": 1175
},
{
"epoch": 15.68227424749164,
"grad_norm": 0.7139558792114258,
"learning_rate": 0.00014794215795328142,
"loss": 3.0012,
"step": 1176
},
{
"epoch": 15.695652173913043,
"grad_norm": 0.6374551057815552,
"learning_rate": 0.0001478976640711902,
"loss": 2.9123,
"step": 1177
},
{
"epoch": 15.709030100334449,
"grad_norm": 0.5957819223403931,
"learning_rate": 0.00014785317018909902,
"loss": 3.09,
"step": 1178
},
{
"epoch": 15.722408026755852,
"grad_norm": 0.6083621382713318,
"learning_rate": 0.0001478086763070078,
"loss": 3.0231,
"step": 1179
},
{
"epoch": 15.735785953177258,
"grad_norm": 0.6169192790985107,
"learning_rate": 0.00014776418242491658,
"loss": 2.9863,
"step": 1180
},
{
"epoch": 15.749163879598662,
"grad_norm": 0.6058081984519958,
"learning_rate": 0.00014771968854282536,
"loss": 3.0261,
"step": 1181
},
{
"epoch": 15.762541806020067,
"grad_norm": 0.5816760659217834,
"learning_rate": 0.00014767519466073417,
"loss": 3.1593,
"step": 1182
},
{
"epoch": 15.775919732441471,
"grad_norm": 0.6246895790100098,
"learning_rate": 0.00014763070077864293,
"loss": 3.1029,
"step": 1183
},
{
"epoch": 15.789297658862877,
"grad_norm": 0.56280517578125,
"learning_rate": 0.00014758620689655174,
"loss": 2.9778,
"step": 1184
},
{
"epoch": 15.80267558528428,
"grad_norm": 0.5743212699890137,
"learning_rate": 0.00014754171301446052,
"loss": 2.8799,
"step": 1185
},
{
"epoch": 15.816053511705686,
"grad_norm": 0.6163922548294067,
"learning_rate": 0.0001474972191323693,
"loss": 3.0226,
"step": 1186
},
{
"epoch": 15.82943143812709,
"grad_norm": 0.5892409682273865,
"learning_rate": 0.00014745272525027808,
"loss": 3.1167,
"step": 1187
},
{
"epoch": 15.842809364548495,
"grad_norm": 0.7977785468101501,
"learning_rate": 0.0001474082313681869,
"loss": 2.8427,
"step": 1188
},
{
"epoch": 15.856187290969899,
"grad_norm": 0.7396023273468018,
"learning_rate": 0.00014736373748609568,
"loss": 2.7809,
"step": 1189
},
{
"epoch": 15.869565217391305,
"grad_norm": 0.58844393491745,
"learning_rate": 0.00014731924360400446,
"loss": 3.0624,
"step": 1190
},
{
"epoch": 15.882943143812708,
"grad_norm": 0.6903204321861267,
"learning_rate": 0.00014727474972191324,
"loss": 3.1246,
"step": 1191
},
{
"epoch": 15.896321070234114,
"grad_norm": 0.5902391672134399,
"learning_rate": 0.00014723025583982205,
"loss": 3.0505,
"step": 1192
},
{
"epoch": 15.909698996655518,
"grad_norm": 0.575752317905426,
"learning_rate": 0.0001471857619577308,
"loss": 2.8508,
"step": 1193
},
{
"epoch": 15.923076923076923,
"grad_norm": 0.7248224020004272,
"learning_rate": 0.00014714126807563961,
"loss": 3.1438,
"step": 1194
},
{
"epoch": 15.936454849498327,
"grad_norm": 0.5669791102409363,
"learning_rate": 0.0001470967741935484,
"loss": 3.1765,
"step": 1195
},
{
"epoch": 15.949832775919733,
"grad_norm": 0.6656806468963623,
"learning_rate": 0.00014705228031145718,
"loss": 3.1456,
"step": 1196
},
{
"epoch": 15.963210702341136,
"grad_norm": 0.6073266863822937,
"learning_rate": 0.00014700778642936596,
"loss": 3.1836,
"step": 1197
},
{
"epoch": 15.976588628762542,
"grad_norm": 0.8209658861160278,
"learning_rate": 0.00014696329254727477,
"loss": 2.8457,
"step": 1198
},
{
"epoch": 15.989966555183946,
"grad_norm": 0.6495081186294556,
"learning_rate": 0.00014691879866518355,
"loss": 3.0161,
"step": 1199
},
{
"epoch": 16.0,
"grad_norm": 0.7522635459899902,
"learning_rate": 0.00014687430478309234,
"loss": 2.8293,
"step": 1200
},
{
"epoch": 16.013377926421406,
"grad_norm": 0.8024417161941528,
"learning_rate": 0.00014682981090100112,
"loss": 2.909,
"step": 1201
},
{
"epoch": 16.02675585284281,
"grad_norm": 0.7147983908653259,
"learning_rate": 0.0001467853170189099,
"loss": 2.7664,
"step": 1202
},
{
"epoch": 16.040133779264213,
"grad_norm": 0.602427065372467,
"learning_rate": 0.00014674082313681868,
"loss": 2.8189,
"step": 1203
},
{
"epoch": 16.05351170568562,
"grad_norm": 0.7274264097213745,
"learning_rate": 0.00014669632925472747,
"loss": 2.9957,
"step": 1204
},
{
"epoch": 16.066889632107024,
"grad_norm": 0.6042147278785706,
"learning_rate": 0.00014665183537263628,
"loss": 2.8925,
"step": 1205
},
{
"epoch": 16.08026755852843,
"grad_norm": 0.592339813709259,
"learning_rate": 0.00014660734149054506,
"loss": 3.1463,
"step": 1206
},
{
"epoch": 16.093645484949832,
"grad_norm": 0.6099987030029297,
"learning_rate": 0.00014656284760845384,
"loss": 3.2264,
"step": 1207
},
{
"epoch": 16.107023411371237,
"grad_norm": 0.9324999451637268,
"learning_rate": 0.00014651835372636262,
"loss": 2.916,
"step": 1208
},
{
"epoch": 16.120401337792643,
"grad_norm": 0.6000798344612122,
"learning_rate": 0.00014647385984427143,
"loss": 2.9242,
"step": 1209
},
{
"epoch": 16.13377926421405,
"grad_norm": 0.609199047088623,
"learning_rate": 0.0001464293659621802,
"loss": 2.9405,
"step": 1210
},
{
"epoch": 16.14715719063545,
"grad_norm": 0.6789796948432922,
"learning_rate": 0.000146384872080089,
"loss": 3.003,
"step": 1211
},
{
"epoch": 16.160535117056856,
"grad_norm": 0.6567651629447937,
"learning_rate": 0.00014634037819799778,
"loss": 2.8733,
"step": 1212
},
{
"epoch": 16.17391304347826,
"grad_norm": 0.5860549211502075,
"learning_rate": 0.00014629588431590656,
"loss": 3.1429,
"step": 1213
},
{
"epoch": 16.187290969899667,
"grad_norm": 0.6623414158821106,
"learning_rate": 0.00014625139043381534,
"loss": 3.1741,
"step": 1214
},
{
"epoch": 16.20066889632107,
"grad_norm": 0.8366180062294006,
"learning_rate": 0.00014620689655172415,
"loss": 2.8627,
"step": 1215
},
{
"epoch": 16.214046822742475,
"grad_norm": 0.616780698299408,
"learning_rate": 0.00014616240266963294,
"loss": 2.7619,
"step": 1216
},
{
"epoch": 16.22742474916388,
"grad_norm": 0.6345306634902954,
"learning_rate": 0.00014611790878754172,
"loss": 2.9104,
"step": 1217
},
{
"epoch": 16.240802675585286,
"grad_norm": 0.6326844096183777,
"learning_rate": 0.0001460734149054505,
"loss": 2.8453,
"step": 1218
},
{
"epoch": 16.254180602006688,
"grad_norm": 0.5441793203353882,
"learning_rate": 0.0001460289210233593,
"loss": 3.2924,
"step": 1219
},
{
"epoch": 16.267558528428093,
"grad_norm": 0.604637086391449,
"learning_rate": 0.00014598442714126807,
"loss": 3.0362,
"step": 1220
},
{
"epoch": 16.2809364548495,
"grad_norm": 0.6100621819496155,
"learning_rate": 0.00014593993325917687,
"loss": 2.917,
"step": 1221
},
{
"epoch": 16.294314381270905,
"grad_norm": 0.6323224902153015,
"learning_rate": 0.00014589543937708566,
"loss": 3.144,
"step": 1222
},
{
"epoch": 16.307692307692307,
"grad_norm": 0.595485270023346,
"learning_rate": 0.00014585094549499444,
"loss": 2.8869,
"step": 1223
},
{
"epoch": 16.321070234113712,
"grad_norm": 0.6350538730621338,
"learning_rate": 0.00014580645161290322,
"loss": 2.7253,
"step": 1224
},
{
"epoch": 16.334448160535118,
"grad_norm": 0.5804395079612732,
"learning_rate": 0.00014576195773081203,
"loss": 3.1523,
"step": 1225
},
{
"epoch": 16.347826086956523,
"grad_norm": 0.5905717015266418,
"learning_rate": 0.00014571746384872081,
"loss": 3.1419,
"step": 1226
},
{
"epoch": 16.361204013377925,
"grad_norm": 0.6824894547462463,
"learning_rate": 0.0001456729699666296,
"loss": 2.8953,
"step": 1227
},
{
"epoch": 16.37458193979933,
"grad_norm": 0.5840978622436523,
"learning_rate": 0.00014562847608453838,
"loss": 3.1229,
"step": 1228
},
{
"epoch": 16.387959866220736,
"grad_norm": 0.7102469801902771,
"learning_rate": 0.0001455839822024472,
"loss": 2.6571,
"step": 1229
},
{
"epoch": 16.401337792642142,
"grad_norm": 0.6148349046707153,
"learning_rate": 0.00014553948832035594,
"loss": 3.0084,
"step": 1230
},
{
"epoch": 16.414715719063544,
"grad_norm": 0.60859215259552,
"learning_rate": 0.00014549499443826475,
"loss": 3.1348,
"step": 1231
},
{
"epoch": 16.42809364548495,
"grad_norm": 0.6059513688087463,
"learning_rate": 0.00014545050055617354,
"loss": 2.9678,
"step": 1232
},
{
"epoch": 16.441471571906355,
"grad_norm": 0.6012231707572937,
"learning_rate": 0.00014540600667408232,
"loss": 2.9279,
"step": 1233
},
{
"epoch": 16.45484949832776,
"grad_norm": 0.6185587644577026,
"learning_rate": 0.0001453615127919911,
"loss": 2.9922,
"step": 1234
},
{
"epoch": 16.468227424749163,
"grad_norm": 0.5989127159118652,
"learning_rate": 0.0001453170189098999,
"loss": 3.0736,
"step": 1235
},
{
"epoch": 16.48160535117057,
"grad_norm": 0.673633337020874,
"learning_rate": 0.0001452725250278087,
"loss": 2.9703,
"step": 1236
},
{
"epoch": 16.494983277591974,
"grad_norm": 0.6815900206565857,
"learning_rate": 0.00014522803114571747,
"loss": 3.0171,
"step": 1237
},
{
"epoch": 16.50836120401338,
"grad_norm": 0.7687232494354248,
"learning_rate": 0.00014518353726362626,
"loss": 2.8988,
"step": 1238
},
{
"epoch": 16.52173913043478,
"grad_norm": 0.7992755174636841,
"learning_rate": 0.00014513904338153507,
"loss": 3.0347,
"step": 1239
},
{
"epoch": 16.535117056856187,
"grad_norm": 0.8673639893531799,
"learning_rate": 0.00014509454949944382,
"loss": 2.9654,
"step": 1240
},
{
"epoch": 16.548494983277592,
"grad_norm": 0.6200671792030334,
"learning_rate": 0.00014505005561735263,
"loss": 3.1075,
"step": 1241
},
{
"epoch": 16.561872909698998,
"grad_norm": 0.8055624961853027,
"learning_rate": 0.0001450055617352614,
"loss": 2.877,
"step": 1242
},
{
"epoch": 16.5752508361204,
"grad_norm": 0.6428245902061462,
"learning_rate": 0.0001449610678531702,
"loss": 3.0326,
"step": 1243
},
{
"epoch": 16.588628762541806,
"grad_norm": 0.844804584980011,
"learning_rate": 0.00014491657397107898,
"loss": 2.8093,
"step": 1244
},
{
"epoch": 16.60200668896321,
"grad_norm": 0.5699613690376282,
"learning_rate": 0.00014487208008898776,
"loss": 3.1914,
"step": 1245
},
{
"epoch": 16.615384615384617,
"grad_norm": 0.6638582348823547,
"learning_rate": 0.00014482758620689657,
"loss": 2.7484,
"step": 1246
},
{
"epoch": 16.62876254180602,
"grad_norm": 0.7390914559364319,
"learning_rate": 0.00014478309232480533,
"loss": 3.0355,
"step": 1247
},
{
"epoch": 16.642140468227424,
"grad_norm": 0.6177923083305359,
"learning_rate": 0.00014473859844271413,
"loss": 3.0067,
"step": 1248
},
{
"epoch": 16.65551839464883,
"grad_norm": 0.6234062314033508,
"learning_rate": 0.00014469410456062292,
"loss": 2.8153,
"step": 1249
},
{
"epoch": 16.668896321070235,
"grad_norm": 0.8505418300628662,
"learning_rate": 0.0001446496106785317,
"loss": 2.8292,
"step": 1250
},
{
"epoch": 16.682274247491637,
"grad_norm": 0.8339266180992126,
"learning_rate": 0.00014460511679644048,
"loss": 2.8072,
"step": 1251
},
{
"epoch": 16.695652173913043,
"grad_norm": 0.5782635807991028,
"learning_rate": 0.0001445606229143493,
"loss": 3.1079,
"step": 1252
},
{
"epoch": 16.70903010033445,
"grad_norm": 0.687126874923706,
"learning_rate": 0.00014451612903225807,
"loss": 3.1401,
"step": 1253
},
{
"epoch": 16.722408026755854,
"grad_norm": 0.7313762307167053,
"learning_rate": 0.00014447163515016686,
"loss": 2.9167,
"step": 1254
},
{
"epoch": 16.735785953177256,
"grad_norm": 0.8815247416496277,
"learning_rate": 0.00014442714126807564,
"loss": 3.0294,
"step": 1255
},
{
"epoch": 16.74916387959866,
"grad_norm": 0.7636277675628662,
"learning_rate": 0.00014438264738598445,
"loss": 3.0758,
"step": 1256
},
{
"epoch": 16.762541806020067,
"grad_norm": 0.5961578488349915,
"learning_rate": 0.0001443381535038932,
"loss": 3.1027,
"step": 1257
},
{
"epoch": 16.775919732441473,
"grad_norm": 0.6840028762817383,
"learning_rate": 0.000144293659621802,
"loss": 2.9192,
"step": 1258
},
{
"epoch": 16.789297658862875,
"grad_norm": 0.7895340323448181,
"learning_rate": 0.0001442491657397108,
"loss": 2.8772,
"step": 1259
},
{
"epoch": 16.80267558528428,
"grad_norm": 0.8516091704368591,
"learning_rate": 0.00014420467185761958,
"loss": 2.9168,
"step": 1260
},
{
"epoch": 16.816053511705686,
"grad_norm": 0.6745076179504395,
"learning_rate": 0.00014416017797552836,
"loss": 3.1352,
"step": 1261
},
{
"epoch": 16.82943143812709,
"grad_norm": 0.6744667887687683,
"learning_rate": 0.00014411568409343717,
"loss": 2.8439,
"step": 1262
},
{
"epoch": 16.842809364548494,
"grad_norm": 0.6307089924812317,
"learning_rate": 0.00014407119021134595,
"loss": 3.2054,
"step": 1263
},
{
"epoch": 16.8561872909699,
"grad_norm": 0.6480753421783447,
"learning_rate": 0.00014402669632925473,
"loss": 3.039,
"step": 1264
},
{
"epoch": 16.869565217391305,
"grad_norm": 0.6143667697906494,
"learning_rate": 0.00014398220244716352,
"loss": 2.9475,
"step": 1265
},
{
"epoch": 16.88294314381271,
"grad_norm": 0.6289299130439758,
"learning_rate": 0.00014393770856507233,
"loss": 3.1437,
"step": 1266
},
{
"epoch": 16.896321070234112,
"grad_norm": 0.6618160009384155,
"learning_rate": 0.00014389321468298108,
"loss": 2.8641,
"step": 1267
},
{
"epoch": 16.909698996655518,
"grad_norm": 0.6053375601768494,
"learning_rate": 0.0001438487208008899,
"loss": 2.9129,
"step": 1268
},
{
"epoch": 16.923076923076923,
"grad_norm": 0.5706185102462769,
"learning_rate": 0.00014380422691879867,
"loss": 3.005,
"step": 1269
},
{
"epoch": 16.93645484949833,
"grad_norm": 0.6779253482818604,
"learning_rate": 0.00014375973303670746,
"loss": 3.1071,
"step": 1270
},
{
"epoch": 16.94983277591973,
"grad_norm": 0.6679616570472717,
"learning_rate": 0.00014371523915461624,
"loss": 3.1792,
"step": 1271
},
{
"epoch": 16.963210702341136,
"grad_norm": 0.6018584966659546,
"learning_rate": 0.00014367074527252505,
"loss": 2.9947,
"step": 1272
},
{
"epoch": 16.976588628762542,
"grad_norm": 0.6106094717979431,
"learning_rate": 0.00014362625139043383,
"loss": 3.1965,
"step": 1273
},
{
"epoch": 16.989966555183948,
"grad_norm": 0.5486257672309875,
"learning_rate": 0.0001435817575083426,
"loss": 3.0975,
"step": 1274
},
{
"epoch": 17.0,
"grad_norm": 0.6516265273094177,
"learning_rate": 0.0001435372636262514,
"loss": 3.1064,
"step": 1275
},
{
"epoch": 17.013377926421406,
"grad_norm": 0.6434454917907715,
"learning_rate": 0.0001434927697441602,
"loss": 2.8903,
"step": 1276
},
{
"epoch": 17.02675585284281,
"grad_norm": 0.6237186193466187,
"learning_rate": 0.00014344827586206896,
"loss": 3.135,
"step": 1277
},
{
"epoch": 17.040133779264213,
"grad_norm": 0.5920026302337646,
"learning_rate": 0.00014340378197997777,
"loss": 2.8612,
"step": 1278
},
{
"epoch": 17.05351170568562,
"grad_norm": 0.6545232534408569,
"learning_rate": 0.00014335928809788655,
"loss": 2.8934,
"step": 1279
},
{
"epoch": 17.066889632107024,
"grad_norm": 0.7839710116386414,
"learning_rate": 0.00014331479421579533,
"loss": 2.965,
"step": 1280
},
{
"epoch": 17.08026755852843,
"grad_norm": 0.6448978781700134,
"learning_rate": 0.00014327030033370412,
"loss": 2.9343,
"step": 1281
},
{
"epoch": 17.093645484949832,
"grad_norm": 0.5713958144187927,
"learning_rate": 0.00014322580645161293,
"loss": 3.1566,
"step": 1282
},
{
"epoch": 17.107023411371237,
"grad_norm": 0.666409969329834,
"learning_rate": 0.0001431813125695217,
"loss": 2.7705,
"step": 1283
},
{
"epoch": 17.120401337792643,
"grad_norm": 0.6068354249000549,
"learning_rate": 0.0001431368186874305,
"loss": 3.088,
"step": 1284
},
{
"epoch": 17.13377926421405,
"grad_norm": 0.8292580246925354,
"learning_rate": 0.00014309232480533927,
"loss": 2.8939,
"step": 1285
},
{
"epoch": 17.14715719063545,
"grad_norm": 0.6789494156837463,
"learning_rate": 0.00014304783092324806,
"loss": 3.2317,
"step": 1286
},
{
"epoch": 17.160535117056856,
"grad_norm": 0.6030963063240051,
"learning_rate": 0.00014300333704115684,
"loss": 2.9248,
"step": 1287
},
{
"epoch": 17.17391304347826,
"grad_norm": 0.7090041041374207,
"learning_rate": 0.00014295884315906562,
"loss": 2.9097,
"step": 1288
},
{
"epoch": 17.187290969899667,
"grad_norm": 0.5750879645347595,
"learning_rate": 0.00014291434927697443,
"loss": 2.9585,
"step": 1289
},
{
"epoch": 17.20066889632107,
"grad_norm": 0.6379792094230652,
"learning_rate": 0.0001428698553948832,
"loss": 2.794,
"step": 1290
},
{
"epoch": 17.214046822742475,
"grad_norm": 0.6736400127410889,
"learning_rate": 0.000142825361512792,
"loss": 2.8812,
"step": 1291
},
{
"epoch": 17.22742474916388,
"grad_norm": 0.6580933332443237,
"learning_rate": 0.00014278086763070078,
"loss": 2.8029,
"step": 1292
},
{
"epoch": 17.240802675585286,
"grad_norm": 0.6550527215003967,
"learning_rate": 0.00014273637374860959,
"loss": 3.1135,
"step": 1293
},
{
"epoch": 17.254180602006688,
"grad_norm": 0.6616796255111694,
"learning_rate": 0.00014269187986651834,
"loss": 2.8916,
"step": 1294
},
{
"epoch": 17.267558528428093,
"grad_norm": 0.7247623801231384,
"learning_rate": 0.00014264738598442715,
"loss": 2.8482,
"step": 1295
},
{
"epoch": 17.2809364548495,
"grad_norm": 0.7138639092445374,
"learning_rate": 0.00014260289210233593,
"loss": 2.9283,
"step": 1296
},
{
"epoch": 17.294314381270905,
"grad_norm": 0.6413894891738892,
"learning_rate": 0.00014255839822024472,
"loss": 3.0273,
"step": 1297
},
{
"epoch": 17.307692307692307,
"grad_norm": 0.6106882095336914,
"learning_rate": 0.0001425139043381535,
"loss": 3.062,
"step": 1298
},
{
"epoch": 17.321070234113712,
"grad_norm": 0.6762199997901917,
"learning_rate": 0.0001424694104560623,
"loss": 2.7864,
"step": 1299
},
{
"epoch": 17.334448160535118,
"grad_norm": 0.65083909034729,
"learning_rate": 0.0001424249165739711,
"loss": 3.115,
"step": 1300
},
{
"epoch": 17.347826086956523,
"grad_norm": 0.7381249666213989,
"learning_rate": 0.00014238042269187987,
"loss": 2.6981,
"step": 1301
},
{
"epoch": 17.361204013377925,
"grad_norm": 0.5674475431442261,
"learning_rate": 0.00014233592880978865,
"loss": 2.8646,
"step": 1302
},
{
"epoch": 17.37458193979933,
"grad_norm": 0.6201330423355103,
"learning_rate": 0.00014229143492769746,
"loss": 3.207,
"step": 1303
},
{
"epoch": 17.387959866220736,
"grad_norm": 0.7004446983337402,
"learning_rate": 0.00014224694104560622,
"loss": 2.7623,
"step": 1304
},
{
"epoch": 17.401337792642142,
"grad_norm": 0.7278717756271362,
"learning_rate": 0.00014220244716351503,
"loss": 2.7967,
"step": 1305
},
{
"epoch": 17.414715719063544,
"grad_norm": 0.6384133696556091,
"learning_rate": 0.0001421579532814238,
"loss": 2.7542,
"step": 1306
},
{
"epoch": 17.42809364548495,
"grad_norm": 0.6443619132041931,
"learning_rate": 0.0001421134593993326,
"loss": 2.9376,
"step": 1307
},
{
"epoch": 17.441471571906355,
"grad_norm": 0.6889017224311829,
"learning_rate": 0.00014206896551724138,
"loss": 2.8868,
"step": 1308
},
{
"epoch": 17.45484949832776,
"grad_norm": 0.624864935874939,
"learning_rate": 0.00014202447163515019,
"loss": 3.0459,
"step": 1309
},
{
"epoch": 17.468227424749163,
"grad_norm": 0.5964120626449585,
"learning_rate": 0.00014197997775305897,
"loss": 2.8604,
"step": 1310
},
{
"epoch": 17.48160535117057,
"grad_norm": 0.5584320425987244,
"learning_rate": 0.00014193548387096775,
"loss": 3.1267,
"step": 1311
},
{
"epoch": 17.494983277591974,
"grad_norm": 0.6566155552864075,
"learning_rate": 0.00014189098998887653,
"loss": 2.7059,
"step": 1312
},
{
"epoch": 17.50836120401338,
"grad_norm": 0.7166509032249451,
"learning_rate": 0.00014184649610678534,
"loss": 3.1076,
"step": 1313
},
{
"epoch": 17.52173913043478,
"grad_norm": 0.6076642870903015,
"learning_rate": 0.0001418020022246941,
"loss": 3.1121,
"step": 1314
},
{
"epoch": 17.535117056856187,
"grad_norm": 0.6293672323226929,
"learning_rate": 0.0001417575083426029,
"loss": 3.0719,
"step": 1315
},
{
"epoch": 17.548494983277592,
"grad_norm": 0.6696231365203857,
"learning_rate": 0.0001417130144605117,
"loss": 2.4281,
"step": 1316
},
{
"epoch": 17.561872909698998,
"grad_norm": 0.7928171157836914,
"learning_rate": 0.00014166852057842047,
"loss": 2.8602,
"step": 1317
},
{
"epoch": 17.5752508361204,
"grad_norm": 0.5897494554519653,
"learning_rate": 0.00014162402669632925,
"loss": 3.1556,
"step": 1318
},
{
"epoch": 17.588628762541806,
"grad_norm": 0.6028451323509216,
"learning_rate": 0.00014157953281423806,
"loss": 3.1228,
"step": 1319
},
{
"epoch": 17.60200668896321,
"grad_norm": 0.6237207651138306,
"learning_rate": 0.00014153503893214685,
"loss": 2.9691,
"step": 1320
},
{
"epoch": 17.615384615384617,
"grad_norm": 0.6401494741439819,
"learning_rate": 0.00014149054505005563,
"loss": 3.0659,
"step": 1321
},
{
"epoch": 17.62876254180602,
"grad_norm": 0.7098166942596436,
"learning_rate": 0.0001414460511679644,
"loss": 2.9606,
"step": 1322
},
{
"epoch": 17.642140468227424,
"grad_norm": 0.6416228413581848,
"learning_rate": 0.00014140155728587322,
"loss": 3.1056,
"step": 1323
},
{
"epoch": 17.65551839464883,
"grad_norm": 0.7303211092948914,
"learning_rate": 0.00014135706340378198,
"loss": 2.8604,
"step": 1324
},
{
"epoch": 17.668896321070235,
"grad_norm": 0.64544677734375,
"learning_rate": 0.00014131256952169079,
"loss": 2.6751,
"step": 1325
},
{
"epoch": 17.682274247491637,
"grad_norm": 0.6870211362838745,
"learning_rate": 0.00014126807563959957,
"loss": 2.9802,
"step": 1326
},
{
"epoch": 17.695652173913043,
"grad_norm": 0.6570687294006348,
"learning_rate": 0.00014122358175750835,
"loss": 3.0496,
"step": 1327
},
{
"epoch": 17.70903010033445,
"grad_norm": 0.7057302594184875,
"learning_rate": 0.00014117908787541713,
"loss": 2.845,
"step": 1328
},
{
"epoch": 17.722408026755854,
"grad_norm": 0.8613574504852295,
"learning_rate": 0.00014113459399332591,
"loss": 3.0314,
"step": 1329
},
{
"epoch": 17.735785953177256,
"grad_norm": 0.7408957481384277,
"learning_rate": 0.00014109010011123472,
"loss": 2.8651,
"step": 1330
},
{
"epoch": 17.74916387959866,
"grad_norm": 0.6553664803504944,
"learning_rate": 0.00014104560622914348,
"loss": 2.9647,
"step": 1331
},
{
"epoch": 17.762541806020067,
"grad_norm": 0.5991332530975342,
"learning_rate": 0.0001410011123470523,
"loss": 2.9796,
"step": 1332
},
{
"epoch": 17.775919732441473,
"grad_norm": 0.6124044060707092,
"learning_rate": 0.00014095661846496107,
"loss": 3.0682,
"step": 1333
},
{
"epoch": 17.789297658862875,
"grad_norm": 0.5788628458976746,
"learning_rate": 0.00014091212458286985,
"loss": 3.0357,
"step": 1334
},
{
"epoch": 17.80267558528428,
"grad_norm": 0.6785842776298523,
"learning_rate": 0.00014086763070077864,
"loss": 3.1128,
"step": 1335
},
{
"epoch": 17.816053511705686,
"grad_norm": 0.5994388461112976,
"learning_rate": 0.00014082313681868745,
"loss": 2.9087,
"step": 1336
},
{
"epoch": 17.82943143812709,
"grad_norm": 0.6150069236755371,
"learning_rate": 0.00014077864293659623,
"loss": 2.9488,
"step": 1337
},
{
"epoch": 17.842809364548494,
"grad_norm": 0.6211126446723938,
"learning_rate": 0.000140734149054505,
"loss": 2.8006,
"step": 1338
},
{
"epoch": 17.8561872909699,
"grad_norm": 0.6093603372573853,
"learning_rate": 0.0001406896551724138,
"loss": 3.0013,
"step": 1339
},
{
"epoch": 17.869565217391305,
"grad_norm": 0.6861109137535095,
"learning_rate": 0.0001406451612903226,
"loss": 2.9423,
"step": 1340
},
{
"epoch": 17.88294314381271,
"grad_norm": 0.6148517727851868,
"learning_rate": 0.00014060066740823136,
"loss": 2.8471,
"step": 1341
},
{
"epoch": 17.896321070234112,
"grad_norm": 0.8285694718360901,
"learning_rate": 0.00014055617352614017,
"loss": 3.055,
"step": 1342
},
{
"epoch": 17.909698996655518,
"grad_norm": 0.5955973863601685,
"learning_rate": 0.00014051167964404895,
"loss": 3.0088,
"step": 1343
},
{
"epoch": 17.923076923076923,
"grad_norm": 0.6020825505256653,
"learning_rate": 0.00014046718576195773,
"loss": 3.1445,
"step": 1344
},
{
"epoch": 17.93645484949833,
"grad_norm": 0.6115384101867676,
"learning_rate": 0.00014042269187986651,
"loss": 3.0889,
"step": 1345
},
{
"epoch": 17.94983277591973,
"grad_norm": 0.6469634175300598,
"learning_rate": 0.00014037819799777532,
"loss": 3.1916,
"step": 1346
},
{
"epoch": 17.963210702341136,
"grad_norm": 0.6653386354446411,
"learning_rate": 0.0001403337041156841,
"loss": 2.8805,
"step": 1347
},
{
"epoch": 17.976588628762542,
"grad_norm": 0.6167243719100952,
"learning_rate": 0.0001402892102335929,
"loss": 2.9819,
"step": 1348
},
{
"epoch": 17.989966555183948,
"grad_norm": 0.6281883716583252,
"learning_rate": 0.00014024471635150167,
"loss": 3.1109,
"step": 1349
},
{
"epoch": 18.0,
"grad_norm": 0.6747295260429382,
"learning_rate": 0.00014020022246941048,
"loss": 2.8772,
"step": 1350
},
{
"epoch": 18.013377926421406,
"grad_norm": 0.5834116339683533,
"learning_rate": 0.00014015572858731924,
"loss": 2.9399,
"step": 1351
},
{
"epoch": 18.02675585284281,
"grad_norm": 0.620858371257782,
"learning_rate": 0.00014011123470522805,
"loss": 2.7351,
"step": 1352
},
{
"epoch": 18.040133779264213,
"grad_norm": 0.5407689809799194,
"learning_rate": 0.00014006674082313683,
"loss": 2.8329,
"step": 1353
},
{
"epoch": 18.05351170568562,
"grad_norm": 0.6045056581497192,
"learning_rate": 0.0001400222469410456,
"loss": 2.7382,
"step": 1354
},
{
"epoch": 18.066889632107024,
"grad_norm": 0.5433570146560669,
"learning_rate": 0.0001399777530589544,
"loss": 2.9154,
"step": 1355
},
{
"epoch": 18.08026755852843,
"grad_norm": 0.6174083352088928,
"learning_rate": 0.0001399332591768632,
"loss": 2.7438,
"step": 1356
},
{
"epoch": 18.093645484949832,
"grad_norm": 0.6720690727233887,
"learning_rate": 0.00013988876529477198,
"loss": 3.1324,
"step": 1357
},
{
"epoch": 18.107023411371237,
"grad_norm": 0.648423433303833,
"learning_rate": 0.00013984427141268077,
"loss": 2.7802,
"step": 1358
},
{
"epoch": 18.120401337792643,
"grad_norm": 0.6625978350639343,
"learning_rate": 0.00013979977753058955,
"loss": 2.9152,
"step": 1359
},
{
"epoch": 18.13377926421405,
"grad_norm": 0.5362007021903992,
"learning_rate": 0.00013975528364849836,
"loss": 2.9552,
"step": 1360
},
{
"epoch": 18.14715719063545,
"grad_norm": 0.6275555491447449,
"learning_rate": 0.00013971078976640711,
"loss": 3.0742,
"step": 1361
},
{
"epoch": 18.160535117056856,
"grad_norm": 0.5755884647369385,
"learning_rate": 0.00013966629588431592,
"loss": 2.8705,
"step": 1362
},
{
"epoch": 18.17391304347826,
"grad_norm": 0.5092719793319702,
"learning_rate": 0.0001396218020022247,
"loss": 2.9684,
"step": 1363
},
{
"epoch": 18.187290969899667,
"grad_norm": 0.7400075197219849,
"learning_rate": 0.0001395773081201335,
"loss": 2.9726,
"step": 1364
},
{
"epoch": 18.20066889632107,
"grad_norm": 0.6478124260902405,
"learning_rate": 0.00013953281423804227,
"loss": 2.7427,
"step": 1365
},
{
"epoch": 18.214046822742475,
"grad_norm": 0.6313418745994568,
"learning_rate": 0.00013948832035595108,
"loss": 2.7713,
"step": 1366
},
{
"epoch": 18.22742474916388,
"grad_norm": 0.5571421980857849,
"learning_rate": 0.00013944382647385986,
"loss": 2.9221,
"step": 1367
},
{
"epoch": 18.240802675585286,
"grad_norm": 0.5346395373344421,
"learning_rate": 0.00013939933259176862,
"loss": 2.9842,
"step": 1368
},
{
"epoch": 18.254180602006688,
"grad_norm": 0.5828048586845398,
"learning_rate": 0.00013935483870967743,
"loss": 2.8683,
"step": 1369
},
{
"epoch": 18.267558528428093,
"grad_norm": 0.6446037888526917,
"learning_rate": 0.0001393103448275862,
"loss": 2.8,
"step": 1370
},
{
"epoch": 18.2809364548495,
"grad_norm": 0.612689197063446,
"learning_rate": 0.000139265850945495,
"loss": 3.0642,
"step": 1371
},
{
"epoch": 18.294314381270905,
"grad_norm": 0.5511941909790039,
"learning_rate": 0.00013922135706340377,
"loss": 3.0633,
"step": 1372
},
{
"epoch": 18.307692307692307,
"grad_norm": 0.7538149356842041,
"learning_rate": 0.00013917686318131258,
"loss": 2.7224,
"step": 1373
},
{
"epoch": 18.321070234113712,
"grad_norm": 0.6194874048233032,
"learning_rate": 0.00013913236929922137,
"loss": 2.9357,
"step": 1374
},
{
"epoch": 18.334448160535118,
"grad_norm": 0.5770833492279053,
"learning_rate": 0.00013908787541713015,
"loss": 2.8585,
"step": 1375
},
{
"epoch": 18.347826086956523,
"grad_norm": 0.609080970287323,
"learning_rate": 0.00013904338153503893,
"loss": 2.7763,
"step": 1376
},
{
"epoch": 18.361204013377925,
"grad_norm": 0.5578462481498718,
"learning_rate": 0.00013899888765294774,
"loss": 3.0324,
"step": 1377
},
{
"epoch": 18.37458193979933,
"grad_norm": 0.5949610471725464,
"learning_rate": 0.0001389543937708565,
"loss": 2.7332,
"step": 1378
},
{
"epoch": 18.387959866220736,
"grad_norm": 0.6248785257339478,
"learning_rate": 0.0001389098998887653,
"loss": 2.7355,
"step": 1379
},
{
"epoch": 18.401337792642142,
"grad_norm": 0.606239378452301,
"learning_rate": 0.0001388654060066741,
"loss": 2.9883,
"step": 1380
},
{
"epoch": 18.414715719063544,
"grad_norm": 0.6222496628761292,
"learning_rate": 0.00013882091212458287,
"loss": 2.7384,
"step": 1381
},
{
"epoch": 18.42809364548495,
"grad_norm": 0.6253412365913391,
"learning_rate": 0.00013877641824249165,
"loss": 2.7555,
"step": 1382
},
{
"epoch": 18.441471571906355,
"grad_norm": 0.6204626560211182,
"learning_rate": 0.00013873192436040046,
"loss": 2.7279,
"step": 1383
},
{
"epoch": 18.45484949832776,
"grad_norm": 0.7254919409751892,
"learning_rate": 0.00013868743047830924,
"loss": 2.9053,
"step": 1384
},
{
"epoch": 18.468227424749163,
"grad_norm": 0.6207154393196106,
"learning_rate": 0.00013864293659621803,
"loss": 3.0648,
"step": 1385
},
{
"epoch": 18.48160535117057,
"grad_norm": 0.6959066390991211,
"learning_rate": 0.0001385984427141268,
"loss": 2.9641,
"step": 1386
},
{
"epoch": 18.494983277591974,
"grad_norm": 0.6345707774162292,
"learning_rate": 0.00013855394883203562,
"loss": 2.9452,
"step": 1387
},
{
"epoch": 18.50836120401338,
"grad_norm": 0.5806639790534973,
"learning_rate": 0.00013850945494994437,
"loss": 2.9375,
"step": 1388
},
{
"epoch": 18.52173913043478,
"grad_norm": 0.6498666405677795,
"learning_rate": 0.00013846496106785318,
"loss": 2.867,
"step": 1389
},
{
"epoch": 18.535117056856187,
"grad_norm": 0.629264771938324,
"learning_rate": 0.00013842046718576197,
"loss": 2.822,
"step": 1390
},
{
"epoch": 18.548494983277592,
"grad_norm": 0.6734644174575806,
"learning_rate": 0.00013837597330367075,
"loss": 2.9065,
"step": 1391
},
{
"epoch": 18.561872909698998,
"grad_norm": 0.5705899000167847,
"learning_rate": 0.00013833147942157953,
"loss": 2.9909,
"step": 1392
},
{
"epoch": 18.5752508361204,
"grad_norm": 0.6786744594573975,
"learning_rate": 0.00013828698553948834,
"loss": 3.0667,
"step": 1393
},
{
"epoch": 18.588628762541806,
"grad_norm": 0.6044118404388428,
"learning_rate": 0.00013824249165739712,
"loss": 2.8345,
"step": 1394
},
{
"epoch": 18.60200668896321,
"grad_norm": 0.5928333401679993,
"learning_rate": 0.0001381979977753059,
"loss": 2.974,
"step": 1395
},
{
"epoch": 18.615384615384617,
"grad_norm": 0.636883556842804,
"learning_rate": 0.0001381535038932147,
"loss": 3.1034,
"step": 1396
},
{
"epoch": 18.62876254180602,
"grad_norm": 0.6029159426689148,
"learning_rate": 0.0001381090100111235,
"loss": 3.0303,
"step": 1397
},
{
"epoch": 18.642140468227424,
"grad_norm": 0.6479122638702393,
"learning_rate": 0.00013806451612903225,
"loss": 3.1782,
"step": 1398
},
{
"epoch": 18.65551839464883,
"grad_norm": 0.6547753810882568,
"learning_rate": 0.00013802002224694106,
"loss": 3.0594,
"step": 1399
},
{
"epoch": 18.668896321070235,
"grad_norm": 0.6506614089012146,
"learning_rate": 0.00013797552836484984,
"loss": 3.0955,
"step": 1400
},
{
"epoch": 18.682274247491637,
"grad_norm": 0.6073411107063293,
"learning_rate": 0.00013793103448275863,
"loss": 2.7452,
"step": 1401
},
{
"epoch": 18.695652173913043,
"grad_norm": 0.6307429075241089,
"learning_rate": 0.0001378865406006674,
"loss": 2.8329,
"step": 1402
},
{
"epoch": 18.70903010033445,
"grad_norm": 0.7205286026000977,
"learning_rate": 0.00013784204671857622,
"loss": 2.8228,
"step": 1403
},
{
"epoch": 18.722408026755854,
"grad_norm": 0.6154866814613342,
"learning_rate": 0.000137797552836485,
"loss": 2.7081,
"step": 1404
},
{
"epoch": 18.735785953177256,
"grad_norm": 0.548799455165863,
"learning_rate": 0.00013775305895439378,
"loss": 2.9336,
"step": 1405
},
{
"epoch": 18.74916387959866,
"grad_norm": 0.5723371505737305,
"learning_rate": 0.00013770856507230257,
"loss": 2.9661,
"step": 1406
},
{
"epoch": 18.762541806020067,
"grad_norm": 0.5917731523513794,
"learning_rate": 0.00013766407119021137,
"loss": 2.8458,
"step": 1407
},
{
"epoch": 18.775919732441473,
"grad_norm": 0.6160814166069031,
"learning_rate": 0.00013761957730812013,
"loss": 2.8453,
"step": 1408
},
{
"epoch": 18.789297658862875,
"grad_norm": 0.5995500683784485,
"learning_rate": 0.0001375750834260289,
"loss": 3.1361,
"step": 1409
},
{
"epoch": 18.80267558528428,
"grad_norm": 0.5574924945831299,
"learning_rate": 0.00013753058954393772,
"loss": 2.9808,
"step": 1410
},
{
"epoch": 18.816053511705686,
"grad_norm": 0.6444510221481323,
"learning_rate": 0.0001374860956618465,
"loss": 2.9204,
"step": 1411
},
{
"epoch": 18.82943143812709,
"grad_norm": 0.6153264045715332,
"learning_rate": 0.0001374416017797553,
"loss": 3.036,
"step": 1412
},
{
"epoch": 18.842809364548494,
"grad_norm": 0.5951060652732849,
"learning_rate": 0.00013739710789766407,
"loss": 3.1012,
"step": 1413
},
{
"epoch": 18.8561872909699,
"grad_norm": 0.5688861012458801,
"learning_rate": 0.00013735261401557288,
"loss": 3.1568,
"step": 1414
},
{
"epoch": 18.869565217391305,
"grad_norm": 0.6256094574928284,
"learning_rate": 0.00013730812013348163,
"loss": 2.7721,
"step": 1415
},
{
"epoch": 18.88294314381271,
"grad_norm": 0.646250307559967,
"learning_rate": 0.00013726362625139044,
"loss": 3.0091,
"step": 1416
},
{
"epoch": 18.896321070234112,
"grad_norm": 0.6805879473686218,
"learning_rate": 0.00013721913236929923,
"loss": 2.8397,
"step": 1417
},
{
"epoch": 18.909698996655518,
"grad_norm": 0.6164728999137878,
"learning_rate": 0.000137174638487208,
"loss": 2.8318,
"step": 1418
},
{
"epoch": 18.923076923076923,
"grad_norm": 0.6298549771308899,
"learning_rate": 0.0001371301446051168,
"loss": 2.8493,
"step": 1419
},
{
"epoch": 18.93645484949833,
"grad_norm": 0.5760109424591064,
"learning_rate": 0.0001370856507230256,
"loss": 3.1736,
"step": 1420
},
{
"epoch": 18.94983277591973,
"grad_norm": 0.6126035451889038,
"learning_rate": 0.00013704115684093438,
"loss": 3.1313,
"step": 1421
},
{
"epoch": 18.963210702341136,
"grad_norm": 0.6092283129692078,
"learning_rate": 0.00013699666295884316,
"loss": 3.0601,
"step": 1422
},
{
"epoch": 18.976588628762542,
"grad_norm": 0.6506980657577515,
"learning_rate": 0.00013695216907675195,
"loss": 2.7787,
"step": 1423
},
{
"epoch": 18.989966555183948,
"grad_norm": 0.6060482263565063,
"learning_rate": 0.00013690767519466076,
"loss": 3.062,
"step": 1424
},
{
"epoch": 19.0,
"grad_norm": 0.7881284952163696,
"learning_rate": 0.0001368631813125695,
"loss": 2.9882,
"step": 1425
},
{
"epoch": 19.013377926421406,
"grad_norm": 0.5459823608398438,
"learning_rate": 0.00013681868743047832,
"loss": 2.9312,
"step": 1426
},
{
"epoch": 19.02675585284281,
"grad_norm": 0.71108078956604,
"learning_rate": 0.0001367741935483871,
"loss": 2.6003,
"step": 1427
},
{
"epoch": 19.040133779264213,
"grad_norm": 0.6824572682380676,
"learning_rate": 0.00013672969966629589,
"loss": 2.8354,
"step": 1428
},
{
"epoch": 19.05351170568562,
"grad_norm": 0.607400119304657,
"learning_rate": 0.00013668520578420467,
"loss": 2.95,
"step": 1429
},
{
"epoch": 19.066889632107024,
"grad_norm": 0.5925526022911072,
"learning_rate": 0.00013664071190211348,
"loss": 2.7101,
"step": 1430
},
{
"epoch": 19.08026755852843,
"grad_norm": 0.6208476424217224,
"learning_rate": 0.00013659621802002226,
"loss": 2.7759,
"step": 1431
},
{
"epoch": 19.093645484949832,
"grad_norm": 0.6047778725624084,
"learning_rate": 0.00013655172413793104,
"loss": 3.0377,
"step": 1432
},
{
"epoch": 19.107023411371237,
"grad_norm": 0.5979378819465637,
"learning_rate": 0.00013650723025583983,
"loss": 2.8671,
"step": 1433
},
{
"epoch": 19.120401337792643,
"grad_norm": 0.6338753700256348,
"learning_rate": 0.00013646273637374863,
"loss": 2.7869,
"step": 1434
},
{
"epoch": 19.13377926421405,
"grad_norm": 0.6263737678527832,
"learning_rate": 0.0001364182424916574,
"loss": 2.6591,
"step": 1435
},
{
"epoch": 19.14715719063545,
"grad_norm": 0.646990180015564,
"learning_rate": 0.0001363737486095662,
"loss": 2.896,
"step": 1436
},
{
"epoch": 19.160535117056856,
"grad_norm": 0.5691138505935669,
"learning_rate": 0.00013632925472747498,
"loss": 2.8774,
"step": 1437
},
{
"epoch": 19.17391304347826,
"grad_norm": 0.5838844180107117,
"learning_rate": 0.00013628476084538376,
"loss": 2.9722,
"step": 1438
},
{
"epoch": 19.187290969899667,
"grad_norm": 0.6154463887214661,
"learning_rate": 0.00013624026696329255,
"loss": 3.0143,
"step": 1439
},
{
"epoch": 19.20066889632107,
"grad_norm": 0.6480549573898315,
"learning_rate": 0.00013619577308120136,
"loss": 2.8118,
"step": 1440
},
{
"epoch": 19.214046822742475,
"grad_norm": 0.7675592303276062,
"learning_rate": 0.00013615127919911014,
"loss": 2.865,
"step": 1441
},
{
"epoch": 19.22742474916388,
"grad_norm": 0.7231382131576538,
"learning_rate": 0.00013610678531701892,
"loss": 2.6942,
"step": 1442
},
{
"epoch": 19.240802675585286,
"grad_norm": 0.6359425187110901,
"learning_rate": 0.0001360622914349277,
"loss": 2.5976,
"step": 1443
},
{
"epoch": 19.254180602006688,
"grad_norm": 0.6486908793449402,
"learning_rate": 0.0001360177975528365,
"loss": 2.9494,
"step": 1444
},
{
"epoch": 19.267558528428093,
"grad_norm": 0.5930846929550171,
"learning_rate": 0.00013597330367074527,
"loss": 3.1295,
"step": 1445
},
{
"epoch": 19.2809364548495,
"grad_norm": 0.6988996267318726,
"learning_rate": 0.00013592880978865408,
"loss": 2.9627,
"step": 1446
},
{
"epoch": 19.294314381270905,
"grad_norm": 0.5971337556838989,
"learning_rate": 0.00013588431590656286,
"loss": 2.9636,
"step": 1447
},
{
"epoch": 19.307692307692307,
"grad_norm": 0.6479155421257019,
"learning_rate": 0.00013583982202447164,
"loss": 2.6575,
"step": 1448
},
{
"epoch": 19.321070234113712,
"grad_norm": 0.6771759986877441,
"learning_rate": 0.00013579532814238043,
"loss": 3.0667,
"step": 1449
},
{
"epoch": 19.334448160535118,
"grad_norm": 0.5785907506942749,
"learning_rate": 0.0001357508342602892,
"loss": 2.8839,
"step": 1450
},
{
"epoch": 19.347826086956523,
"grad_norm": 0.6315357089042664,
"learning_rate": 0.00013570634037819802,
"loss": 2.6925,
"step": 1451
},
{
"epoch": 19.361204013377925,
"grad_norm": 0.6047807931900024,
"learning_rate": 0.00013566184649610677,
"loss": 2.7948,
"step": 1452
},
{
"epoch": 19.37458193979933,
"grad_norm": 0.6079906225204468,
"learning_rate": 0.00013561735261401558,
"loss": 3.1798,
"step": 1453
},
{
"epoch": 19.387959866220736,
"grad_norm": 0.5820274353027344,
"learning_rate": 0.00013557285873192436,
"loss": 3.1159,
"step": 1454
},
{
"epoch": 19.401337792642142,
"grad_norm": 0.562022864818573,
"learning_rate": 0.00013552836484983315,
"loss": 2.9869,
"step": 1455
},
{
"epoch": 19.414715719063544,
"grad_norm": 0.6663182973861694,
"learning_rate": 0.00013548387096774193,
"loss": 2.8038,
"step": 1456
},
{
"epoch": 19.42809364548495,
"grad_norm": 0.7092719078063965,
"learning_rate": 0.00013543937708565074,
"loss": 3.0159,
"step": 1457
},
{
"epoch": 19.441471571906355,
"grad_norm": 0.5990714430809021,
"learning_rate": 0.00013539488320355952,
"loss": 2.8858,
"step": 1458
},
{
"epoch": 19.45484949832776,
"grad_norm": 0.600847065448761,
"learning_rate": 0.0001353503893214683,
"loss": 2.9352,
"step": 1459
},
{
"epoch": 19.468227424749163,
"grad_norm": 0.6393849849700928,
"learning_rate": 0.00013530589543937709,
"loss": 2.6911,
"step": 1460
},
{
"epoch": 19.48160535117057,
"grad_norm": 0.6170421838760376,
"learning_rate": 0.0001352614015572859,
"loss": 2.749,
"step": 1461
},
{
"epoch": 19.494983277591974,
"grad_norm": 0.53690105676651,
"learning_rate": 0.00013521690767519465,
"loss": 2.9516,
"step": 1462
},
{
"epoch": 19.50836120401338,
"grad_norm": 0.5976501703262329,
"learning_rate": 0.00013517241379310346,
"loss": 2.9996,
"step": 1463
},
{
"epoch": 19.52173913043478,
"grad_norm": 0.5892135500907898,
"learning_rate": 0.00013512791991101224,
"loss": 3.2039,
"step": 1464
},
{
"epoch": 19.535117056856187,
"grad_norm": 0.65968918800354,
"learning_rate": 0.00013508342602892102,
"loss": 2.5897,
"step": 1465
},
{
"epoch": 19.548494983277592,
"grad_norm": 0.591454267501831,
"learning_rate": 0.0001350389321468298,
"loss": 3.0104,
"step": 1466
},
{
"epoch": 19.561872909698998,
"grad_norm": 0.6272184252738953,
"learning_rate": 0.00013499443826473862,
"loss": 2.7266,
"step": 1467
},
{
"epoch": 19.5752508361204,
"grad_norm": 0.6142420172691345,
"learning_rate": 0.0001349499443826474,
"loss": 2.8358,
"step": 1468
},
{
"epoch": 19.588628762541806,
"grad_norm": 0.6268441677093506,
"learning_rate": 0.00013490545050055618,
"loss": 3.0196,
"step": 1469
},
{
"epoch": 19.60200668896321,
"grad_norm": 0.6512436866760254,
"learning_rate": 0.00013486095661846496,
"loss": 2.9558,
"step": 1470
},
{
"epoch": 19.615384615384617,
"grad_norm": 0.5983771681785583,
"learning_rate": 0.00013481646273637377,
"loss": 2.9958,
"step": 1471
},
{
"epoch": 19.62876254180602,
"grad_norm": 0.6994190216064453,
"learning_rate": 0.00013477196885428253,
"loss": 2.8019,
"step": 1472
},
{
"epoch": 19.642140468227424,
"grad_norm": 0.5878567695617676,
"learning_rate": 0.00013472747497219134,
"loss": 2.7007,
"step": 1473
},
{
"epoch": 19.65551839464883,
"grad_norm": 0.6140199303627014,
"learning_rate": 0.00013468298109010012,
"loss": 2.9212,
"step": 1474
},
{
"epoch": 19.668896321070235,
"grad_norm": 0.648714542388916,
"learning_rate": 0.0001346384872080089,
"loss": 3.0053,
"step": 1475
},
{
"epoch": 19.682274247491637,
"grad_norm": 0.5991750359535217,
"learning_rate": 0.00013459399332591769,
"loss": 2.9129,
"step": 1476
},
{
"epoch": 19.695652173913043,
"grad_norm": 0.5538223385810852,
"learning_rate": 0.0001345494994438265,
"loss": 2.9097,
"step": 1477
},
{
"epoch": 19.70903010033445,
"grad_norm": 0.5864409804344177,
"learning_rate": 0.00013450500556173528,
"loss": 2.9348,
"step": 1478
},
{
"epoch": 19.722408026755854,
"grad_norm": 0.6004533767700195,
"learning_rate": 0.00013446051167964406,
"loss": 2.8845,
"step": 1479
},
{
"epoch": 19.735785953177256,
"grad_norm": 0.6316581964492798,
"learning_rate": 0.00013441601779755284,
"loss": 2.8619,
"step": 1480
},
{
"epoch": 19.74916387959866,
"grad_norm": 0.593138575553894,
"learning_rate": 0.00013437152391546165,
"loss": 3.041,
"step": 1481
},
{
"epoch": 19.762541806020067,
"grad_norm": 0.5826678276062012,
"learning_rate": 0.0001343270300333704,
"loss": 2.7432,
"step": 1482
},
{
"epoch": 19.775919732441473,
"grad_norm": 0.6341697573661804,
"learning_rate": 0.00013428253615127922,
"loss": 2.9755,
"step": 1483
},
{
"epoch": 19.789297658862875,
"grad_norm": 0.5894901156425476,
"learning_rate": 0.000134238042269188,
"loss": 2.9754,
"step": 1484
},
{
"epoch": 19.80267558528428,
"grad_norm": 0.5840655565261841,
"learning_rate": 0.00013419354838709678,
"loss": 2.9959,
"step": 1485
},
{
"epoch": 19.816053511705686,
"grad_norm": 0.6006319522857666,
"learning_rate": 0.00013414905450500556,
"loss": 3.0196,
"step": 1486
},
{
"epoch": 19.82943143812709,
"grad_norm": 0.5647453665733337,
"learning_rate": 0.00013410456062291437,
"loss": 2.9662,
"step": 1487
},
{
"epoch": 19.842809364548494,
"grad_norm": 0.6583006978034973,
"learning_rate": 0.00013406006674082316,
"loss": 2.8719,
"step": 1488
},
{
"epoch": 19.8561872909699,
"grad_norm": 0.6041131615638733,
"learning_rate": 0.00013401557285873194,
"loss": 3.0019,
"step": 1489
},
{
"epoch": 19.869565217391305,
"grad_norm": 0.5524600148200989,
"learning_rate": 0.00013397107897664072,
"loss": 2.7409,
"step": 1490
},
{
"epoch": 19.88294314381271,
"grad_norm": 0.6532869338989258,
"learning_rate": 0.0001339265850945495,
"loss": 2.9336,
"step": 1491
},
{
"epoch": 19.896321070234112,
"grad_norm": 0.6459875106811523,
"learning_rate": 0.00013388209121245828,
"loss": 2.9572,
"step": 1492
},
{
"epoch": 19.909698996655518,
"grad_norm": 0.6051417589187622,
"learning_rate": 0.00013383759733036707,
"loss": 2.864,
"step": 1493
},
{
"epoch": 19.923076923076923,
"grad_norm": 0.6565695405006409,
"learning_rate": 0.00013379310344827588,
"loss": 2.865,
"step": 1494
},
{
"epoch": 19.93645484949833,
"grad_norm": 0.6118014454841614,
"learning_rate": 0.00013374860956618466,
"loss": 2.7594,
"step": 1495
},
{
"epoch": 19.94983277591973,
"grad_norm": 0.6801209449768066,
"learning_rate": 0.00013370411568409344,
"loss": 2.7532,
"step": 1496
},
{
"epoch": 19.963210702341136,
"grad_norm": 0.5785267353057861,
"learning_rate": 0.00013365962180200222,
"loss": 3.0618,
"step": 1497
},
{
"epoch": 19.976588628762542,
"grad_norm": 0.6344903707504272,
"learning_rate": 0.00013361512791991103,
"loss": 2.7908,
"step": 1498
},
{
"epoch": 19.989966555183948,
"grad_norm": 0.6073011159896851,
"learning_rate": 0.0001335706340378198,
"loss": 3.0699,
"step": 1499
},
{
"epoch": 20.0,
"grad_norm": 0.6989748477935791,
"learning_rate": 0.0001335261401557286,
"loss": 2.7932,
"step": 1500
}
],
"logging_steps": 1,
"max_steps": 4500,
"num_input_tokens_seen": 0,
"num_train_epochs": 60,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.6403780569344e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}