yt-gui-2b / trainer_state.json
ying18's picture
Upload folder using huggingface_hub
09e48ab verified
Raw
History Blame Contribute Delete
213 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 575.0,
"global_step": 1150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008695652173913044,
"grad_norm": 89.0245590209961,
"learning_rate": 4.310344827586207e-08,
"loss": 0.34853503108024597,
"step": 1
},
{
"epoch": 0.0017391304347826088,
"grad_norm": 1189.884521484375,
"learning_rate": 8.620689655172414e-08,
"loss": 3.0463950634002686,
"step": 2
},
{
"epoch": 0.0026086956521739132,
"grad_norm": 93.21183776855469,
"learning_rate": 1.2931034482758622e-07,
"loss": 1.3009825944900513,
"step": 3
},
{
"epoch": 0.0034782608695652175,
"grad_norm": 391.48583984375,
"learning_rate": 1.7241379310344828e-07,
"loss": 1.7774399518966675,
"step": 4
},
{
"epoch": 0.004347826086956522,
"grad_norm": 310.1289978027344,
"learning_rate": 2.1551724137931036e-07,
"loss": 1.0683984756469727,
"step": 5
},
{
"epoch": 0.0052173913043478265,
"grad_norm": 367.3047180175781,
"learning_rate": 2.5862068965517245e-07,
"loss": 1.4958542585372925,
"step": 6
},
{
"epoch": 0.00608695652173913,
"grad_norm": 160.18492126464844,
"learning_rate": 3.0172413793103453e-07,
"loss": 1.6158473491668701,
"step": 7
},
{
"epoch": 0.006956521739130435,
"grad_norm": 173.58932495117188,
"learning_rate": 3.4482758620689656e-07,
"loss": 2.8637938499450684,
"step": 8
},
{
"epoch": 0.00782608695652174,
"grad_norm": 334.5928955078125,
"learning_rate": 3.8793103448275865e-07,
"loss": 0.8292187452316284,
"step": 9
},
{
"epoch": 0.008695652173913044,
"grad_norm": 368.74224853515625,
"learning_rate": 4.3103448275862073e-07,
"loss": 2.565538167953491,
"step": 10
},
{
"epoch": 0.009565217391304347,
"grad_norm": 362.66180419921875,
"learning_rate": 4.7413793103448276e-07,
"loss": 1.1161541938781738,
"step": 11
},
{
"epoch": 0.010434782608695653,
"grad_norm": 761.982421875,
"learning_rate": 5.172413793103449e-07,
"loss": 2.4958276748657227,
"step": 12
},
{
"epoch": 0.011304347826086957,
"grad_norm": 87.07658386230469,
"learning_rate": 5.60344827586207e-07,
"loss": 1.041823387145996,
"step": 13
},
{
"epoch": 0.01217391304347826,
"grad_norm": 494.1841735839844,
"learning_rate": 6.034482758620691e-07,
"loss": 1.5525039434432983,
"step": 14
},
{
"epoch": 0.013043478260869565,
"grad_norm": 107.69103240966797,
"learning_rate": 6.465517241379311e-07,
"loss": 0.7673182487487793,
"step": 15
},
{
"epoch": 0.01391304347826087,
"grad_norm": 531.2876586914062,
"learning_rate": 6.896551724137931e-07,
"loss": 1.6066687107086182,
"step": 16
},
{
"epoch": 0.014782608695652174,
"grad_norm": 597.859619140625,
"learning_rate": 7.327586206896552e-07,
"loss": 1.50773286819458,
"step": 17
},
{
"epoch": 0.01565217391304348,
"grad_norm": 7288.8623046875,
"learning_rate": 7.758620689655173e-07,
"loss": 5.417834758758545,
"step": 18
},
{
"epoch": 0.01652173913043478,
"grad_norm": 515.7960815429688,
"learning_rate": 8.189655172413794e-07,
"loss": 1.674267292022705,
"step": 19
},
{
"epoch": 0.017391304347826087,
"grad_norm": 81.13020324707031,
"learning_rate": 8.620689655172415e-07,
"loss": 1.736510157585144,
"step": 20
},
{
"epoch": 0.018260869565217393,
"grad_norm": 1084.963623046875,
"learning_rate": 9.051724137931035e-07,
"loss": 1.3609899282455444,
"step": 21
},
{
"epoch": 0.019130434782608695,
"grad_norm": 268.39093017578125,
"learning_rate": 9.482758620689655e-07,
"loss": 2.3733294010162354,
"step": 22
},
{
"epoch": 0.02,
"grad_norm": 339.6881408691406,
"learning_rate": 9.913793103448276e-07,
"loss": 2.0576863288879395,
"step": 23
},
{
"epoch": 0.020869565217391306,
"grad_norm": 2633.229736328125,
"learning_rate": 1.0344827586206898e-06,
"loss": 1.618069052696228,
"step": 24
},
{
"epoch": 0.021739130434782608,
"grad_norm": 213.87811279296875,
"learning_rate": 1.0775862068965518e-06,
"loss": 1.755639910697937,
"step": 25
},
{
"epoch": 0.022608695652173914,
"grad_norm": 179.05685424804688,
"learning_rate": 1.120689655172414e-06,
"loss": 3.2130239009857178,
"step": 26
},
{
"epoch": 0.023478260869565216,
"grad_norm": 275.2119140625,
"learning_rate": 1.163793103448276e-06,
"loss": 0.9658058881759644,
"step": 27
},
{
"epoch": 0.02434782608695652,
"grad_norm": 135.9149169921875,
"learning_rate": 1.2068965517241381e-06,
"loss": 1.6293413639068604,
"step": 28
},
{
"epoch": 0.025217391304347827,
"grad_norm": 577.9674682617188,
"learning_rate": 1.25e-06,
"loss": 2.5478925704956055,
"step": 29
},
{
"epoch": 0.02608695652173913,
"grad_norm": 85.50240325927734,
"learning_rate": 1.2931034482758623e-06,
"loss": 0.6629786491394043,
"step": 30
},
{
"epoch": 0.026956521739130435,
"grad_norm": 110.3420639038086,
"learning_rate": 1.336206896551724e-06,
"loss": 0.34647268056869507,
"step": 31
},
{
"epoch": 0.02782608695652174,
"grad_norm": 250.96575927734375,
"learning_rate": 1.3793103448275862e-06,
"loss": 3.2381415367126465,
"step": 32
},
{
"epoch": 0.028695652173913042,
"grad_norm": 153.7061309814453,
"learning_rate": 1.4224137931034484e-06,
"loss": 1.756481409072876,
"step": 33
},
{
"epoch": 0.029565217391304348,
"grad_norm": 536.8954467773438,
"learning_rate": 1.4655172413793104e-06,
"loss": 1.451122760772705,
"step": 34
},
{
"epoch": 0.030434782608695653,
"grad_norm": 352.06256103515625,
"learning_rate": 1.5086206896551726e-06,
"loss": 1.693876028060913,
"step": 35
},
{
"epoch": 0.03130434782608696,
"grad_norm": 139.7254638671875,
"learning_rate": 1.5517241379310346e-06,
"loss": 0.6885156631469727,
"step": 36
},
{
"epoch": 0.03217391304347826,
"grad_norm": 79.29289245605469,
"learning_rate": 1.5948275862068968e-06,
"loss": 0.9699131846427917,
"step": 37
},
{
"epoch": 0.03304347826086956,
"grad_norm": 132.2869110107422,
"learning_rate": 1.6379310344827587e-06,
"loss": 0.789880633354187,
"step": 38
},
{
"epoch": 0.03391304347826087,
"grad_norm": 111.19515991210938,
"learning_rate": 1.681034482758621e-06,
"loss": 1.4578804969787598,
"step": 39
},
{
"epoch": 0.034782608695652174,
"grad_norm": 317.183837890625,
"learning_rate": 1.724137931034483e-06,
"loss": 1.220886468887329,
"step": 40
},
{
"epoch": 0.035652173913043476,
"grad_norm": 232.15017700195312,
"learning_rate": 1.7672413793103449e-06,
"loss": 0.9733860492706299,
"step": 41
},
{
"epoch": 0.036521739130434785,
"grad_norm": 111.05082702636719,
"learning_rate": 1.810344827586207e-06,
"loss": 0.4968069791793823,
"step": 42
},
{
"epoch": 0.03739130434782609,
"grad_norm": 66.22428131103516,
"learning_rate": 1.8534482758620693e-06,
"loss": 0.4948171079158783,
"step": 43
},
{
"epoch": 0.03826086956521739,
"grad_norm": 513.5997314453125,
"learning_rate": 1.896551724137931e-06,
"loss": 2.0543527603149414,
"step": 44
},
{
"epoch": 0.0391304347826087,
"grad_norm": 210.15696716308594,
"learning_rate": 1.9396551724137932e-06,
"loss": 1.344010353088379,
"step": 45
},
{
"epoch": 0.04,
"grad_norm": 353.21820068359375,
"learning_rate": 1.982758620689655e-06,
"loss": 2.415740966796875,
"step": 46
},
{
"epoch": 0.0408695652173913,
"grad_norm": 394.0849914550781,
"learning_rate": 2.025862068965517e-06,
"loss": 1.0793585777282715,
"step": 47
},
{
"epoch": 0.04173913043478261,
"grad_norm": 149.27621459960938,
"learning_rate": 2.0689655172413796e-06,
"loss": 0.8092156648635864,
"step": 48
},
{
"epoch": 0.042608695652173914,
"grad_norm": 80.21371459960938,
"learning_rate": 2.1120689655172416e-06,
"loss": 0.7981650233268738,
"step": 49
},
{
"epoch": 0.043478260869565216,
"grad_norm": 172.74063110351562,
"learning_rate": 2.1551724137931035e-06,
"loss": 0.7529181241989136,
"step": 50
},
{
"epoch": 0.04434782608695652,
"grad_norm": 70.95804595947266,
"learning_rate": 2.1982758620689655e-06,
"loss": 2.2753522396087646,
"step": 51
},
{
"epoch": 0.04521739130434783,
"grad_norm": 202.04214477539062,
"learning_rate": 2.241379310344828e-06,
"loss": 2.475764751434326,
"step": 52
},
{
"epoch": 0.04608695652173913,
"grad_norm": 194.72572326660156,
"learning_rate": 2.28448275862069e-06,
"loss": 1.2179863452911377,
"step": 53
},
{
"epoch": 0.04695652173913043,
"grad_norm": 156.66160583496094,
"learning_rate": 2.327586206896552e-06,
"loss": 1.6020311117172241,
"step": 54
},
{
"epoch": 0.04782608695652174,
"grad_norm": 75.63845825195312,
"learning_rate": 2.370689655172414e-06,
"loss": 0.642714262008667,
"step": 55
},
{
"epoch": 0.04869565217391304,
"grad_norm": 317.29595947265625,
"learning_rate": 2.4137931034482762e-06,
"loss": 0.5819599628448486,
"step": 56
},
{
"epoch": 0.049565217391304345,
"grad_norm": 91.986572265625,
"learning_rate": 2.4568965517241382e-06,
"loss": 1.2792414426803589,
"step": 57
},
{
"epoch": 0.050434782608695654,
"grad_norm": 106.760498046875,
"learning_rate": 2.5e-06,
"loss": 2.098811149597168,
"step": 58
},
{
"epoch": 0.051304347826086956,
"grad_norm": 179.09872436523438,
"learning_rate": 2.499994827098767e-06,
"loss": 1.3645347356796265,
"step": 59
},
{
"epoch": 0.05217391304347826,
"grad_norm": 149.22055053710938,
"learning_rate": 2.4999793084378814e-06,
"loss": 1.8922319412231445,
"step": 60
},
{
"epoch": 0.05304347826086957,
"grad_norm": 88.80119323730469,
"learning_rate": 2.499953444145787e-06,
"loss": 1.0340819358825684,
"step": 61
},
{
"epoch": 0.05391304347826087,
"grad_norm": 137.12167358398438,
"learning_rate": 2.4999172344365515e-06,
"loss": 0.8423926830291748,
"step": 62
},
{
"epoch": 0.05478260869565217,
"grad_norm": 94.94590759277344,
"learning_rate": 2.4998706796098705e-06,
"loss": 0.39722388982772827,
"step": 63
},
{
"epoch": 0.05565217391304348,
"grad_norm": 75.57413482666016,
"learning_rate": 2.499813780051062e-06,
"loss": 0.5243494510650635,
"step": 64
},
{
"epoch": 0.05652173913043478,
"grad_norm": 88.9786148071289,
"learning_rate": 2.499746536231063e-06,
"loss": 2.822748899459839,
"step": 65
},
{
"epoch": 0.057391304347826085,
"grad_norm": 79.60736846923828,
"learning_rate": 2.4996689487064263e-06,
"loss": 2.185532331466675,
"step": 66
},
{
"epoch": 0.058260869565217394,
"grad_norm": 508.6560974121094,
"learning_rate": 2.4995810181193166e-06,
"loss": 2.56195068359375,
"step": 67
},
{
"epoch": 0.059130434782608696,
"grad_norm": 160.19577026367188,
"learning_rate": 2.499482745197503e-06,
"loss": 1.345895528793335,
"step": 68
},
{
"epoch": 0.06,
"grad_norm": 157.58526611328125,
"learning_rate": 2.499374130754356e-06,
"loss": 0.9514150023460388,
"step": 69
},
{
"epoch": 0.06086956521739131,
"grad_norm": 233.49813842773438,
"learning_rate": 2.499255175688838e-06,
"loss": 4.562856674194336,
"step": 70
},
{
"epoch": 0.06173913043478261,
"grad_norm": 419.5863952636719,
"learning_rate": 2.4991258809854985e-06,
"loss": 1.8423646688461304,
"step": 71
},
{
"epoch": 0.06260869565217392,
"grad_norm": 114.12431335449219,
"learning_rate": 2.498986247714462e-06,
"loss": 0.7551250457763672,
"step": 72
},
{
"epoch": 0.06347826086956522,
"grad_norm": 62.691314697265625,
"learning_rate": 2.4988362770314244e-06,
"loss": 1.181515097618103,
"step": 73
},
{
"epoch": 0.06434782608695652,
"grad_norm": 397.2389221191406,
"learning_rate": 2.4986759701776387e-06,
"loss": 2.3253417015075684,
"step": 74
},
{
"epoch": 0.06521739130434782,
"grad_norm": 108.50537109375,
"learning_rate": 2.498505328479907e-06,
"loss": 1.093691110610962,
"step": 75
},
{
"epoch": 0.06608695652173913,
"grad_norm": 41.83134078979492,
"learning_rate": 2.4983243533505704e-06,
"loss": 0.34358829259872437,
"step": 76
},
{
"epoch": 0.06695652173913043,
"grad_norm": 61.207183837890625,
"learning_rate": 2.498133046287495e-06,
"loss": 0.499338835477829,
"step": 77
},
{
"epoch": 0.06782608695652174,
"grad_norm": 572.0919799804688,
"learning_rate": 2.49793140887406e-06,
"loss": 1.6153039932250977,
"step": 78
},
{
"epoch": 0.06869565217391305,
"grad_norm": 174.57875061035156,
"learning_rate": 2.497719442779147e-06,
"loss": 0.8222454786300659,
"step": 79
},
{
"epoch": 0.06956521739130435,
"grad_norm": 138.34426879882812,
"learning_rate": 2.4974971497571233e-06,
"loss": 1.7004735469818115,
"step": 80
},
{
"epoch": 0.07043478260869565,
"grad_norm": 105.51431274414062,
"learning_rate": 2.4972645316478286e-06,
"loss": 0.3894553780555725,
"step": 81
},
{
"epoch": 0.07130434782608695,
"grad_norm": 2002.7835693359375,
"learning_rate": 2.49702159037656e-06,
"loss": 3.763047695159912,
"step": 82
},
{
"epoch": 0.07217391304347825,
"grad_norm": 514.9981689453125,
"learning_rate": 2.4967683279540542e-06,
"loss": 0.8840932846069336,
"step": 83
},
{
"epoch": 0.07304347826086957,
"grad_norm": 97.48663330078125,
"learning_rate": 2.4965047464764753e-06,
"loss": 2.1965551376342773,
"step": 84
},
{
"epoch": 0.07391304347826087,
"grad_norm": 120.31857299804688,
"learning_rate": 2.496230848125392e-06,
"loss": 1.0099523067474365,
"step": 85
},
{
"epoch": 0.07478260869565218,
"grad_norm": 58.51309585571289,
"learning_rate": 2.495946635167763e-06,
"loss": 0.46549540758132935,
"step": 86
},
{
"epoch": 0.07565217391304348,
"grad_norm": 525.2761840820312,
"learning_rate": 2.495652109955917e-06,
"loss": 1.2519830465316772,
"step": 87
},
{
"epoch": 0.07652173913043478,
"grad_norm": 268.23516845703125,
"learning_rate": 2.495347274927533e-06,
"loss": 1.0990965366363525,
"step": 88
},
{
"epoch": 0.07739130434782608,
"grad_norm": 215.14605712890625,
"learning_rate": 2.495032132605623e-06,
"loss": 0.9456059336662292,
"step": 89
},
{
"epoch": 0.0782608695652174,
"grad_norm": 136.0934600830078,
"learning_rate": 2.4947066855985067e-06,
"loss": 1.0525486469268799,
"step": 90
},
{
"epoch": 0.0791304347826087,
"grad_norm": 56.83368682861328,
"learning_rate": 2.494370936599792e-06,
"loss": 0.2791644036769867,
"step": 91
},
{
"epoch": 0.08,
"grad_norm": 210.65115356445312,
"learning_rate": 2.4940248883883526e-06,
"loss": 1.1349866390228271,
"step": 92
},
{
"epoch": 0.0808695652173913,
"grad_norm": 484.526123046875,
"learning_rate": 2.493668543828307e-06,
"loss": 1.4696400165557861,
"step": 93
},
{
"epoch": 0.0817391304347826,
"grad_norm": 166.41043090820312,
"learning_rate": 2.493301905868991e-06,
"loss": 0.9334723949432373,
"step": 94
},
{
"epoch": 0.08260869565217391,
"grad_norm": 228.30569458007812,
"learning_rate": 2.4929249775449354e-06,
"loss": 0.9395383596420288,
"step": 95
},
{
"epoch": 0.08347826086956522,
"grad_norm": 92.01274871826172,
"learning_rate": 2.4925377619758416e-06,
"loss": 0.6681497097015381,
"step": 96
},
{
"epoch": 0.08434782608695653,
"grad_norm": 155.4160614013672,
"learning_rate": 2.4921402623665535e-06,
"loss": 1.4780633449554443,
"step": 97
},
{
"epoch": 0.08521739130434783,
"grad_norm": 116.00241088867188,
"learning_rate": 2.491732482007033e-06,
"loss": 1.5823075771331787,
"step": 98
},
{
"epoch": 0.08608695652173913,
"grad_norm": 264.5154113769531,
"learning_rate": 2.491314424272334e-06,
"loss": 1.3235678672790527,
"step": 99
},
{
"epoch": 0.08695652173913043,
"grad_norm": 96.83821868896484,
"learning_rate": 2.4908860926225676e-06,
"loss": 2.860171318054199,
"step": 100
},
{
"epoch": 0.08782608695652173,
"grad_norm": 647.5700073242188,
"learning_rate": 2.4904474906028838e-06,
"loss": 0.9455481767654419,
"step": 101
},
{
"epoch": 0.08869565217391304,
"grad_norm": 46.34527587890625,
"learning_rate": 2.489998621843434e-06,
"loss": 1.0399949550628662,
"step": 102
},
{
"epoch": 0.08956521739130435,
"grad_norm": 43.286598205566406,
"learning_rate": 2.489539490059344e-06,
"loss": 0.32844942808151245,
"step": 103
},
{
"epoch": 0.09043478260869565,
"grad_norm": 71.72222900390625,
"learning_rate": 2.489070099050683e-06,
"loss": 0.4192119240760803,
"step": 104
},
{
"epoch": 0.09130434782608696,
"grad_norm": 77.59490203857422,
"learning_rate": 2.488590452702433e-06,
"loss": 0.7706747055053711,
"step": 105
},
{
"epoch": 0.09217391304347826,
"grad_norm": 62.34656524658203,
"learning_rate": 2.4881005549844544e-06,
"loss": 0.2663648724555969,
"step": 106
},
{
"epoch": 0.09304347826086956,
"grad_norm": 76.3553237915039,
"learning_rate": 2.4876004099514554e-06,
"loss": 0.5939281582832336,
"step": 107
},
{
"epoch": 0.09391304347826086,
"grad_norm": 96.17376708984375,
"learning_rate": 2.487090021742957e-06,
"loss": 0.5613781213760376,
"step": 108
},
{
"epoch": 0.09478260869565218,
"grad_norm": 39.40800476074219,
"learning_rate": 2.486569394583261e-06,
"loss": 0.59663325548172,
"step": 109
},
{
"epoch": 0.09565217391304348,
"grad_norm": 373.3201599121094,
"learning_rate": 2.4860385327814108e-06,
"loss": 0.39191126823425293,
"step": 110
},
{
"epoch": 0.09652173913043478,
"grad_norm": 172.11978149414062,
"learning_rate": 2.48549744073116e-06,
"loss": 0.8129326701164246,
"step": 111
},
{
"epoch": 0.09739130434782609,
"grad_norm": 742.5328369140625,
"learning_rate": 2.4849461229109335e-06,
"loss": 2.371474504470825,
"step": 112
},
{
"epoch": 0.09826086956521739,
"grad_norm": 175.04432678222656,
"learning_rate": 2.484384583883791e-06,
"loss": 1.9925346374511719,
"step": 113
},
{
"epoch": 0.09913043478260869,
"grad_norm": 53.41303253173828,
"learning_rate": 2.483812828297391e-06,
"loss": 0.2166566103696823,
"step": 114
},
{
"epoch": 0.1,
"grad_norm": 102.07194519042969,
"learning_rate": 2.4832308608839495e-06,
"loss": 0.5669763684272766,
"step": 115
},
{
"epoch": 0.10086956521739131,
"grad_norm": 85.9627685546875,
"learning_rate": 2.4826386864602024e-06,
"loss": 0.3785554766654968,
"step": 116
},
{
"epoch": 0.10173913043478261,
"grad_norm": 119.04127502441406,
"learning_rate": 2.4820363099273647e-06,
"loss": 0.9650425314903259,
"step": 117
},
{
"epoch": 0.10260869565217391,
"grad_norm": 248.23486328125,
"learning_rate": 2.4814237362710923e-06,
"loss": 1.5368950366973877,
"step": 118
},
{
"epoch": 0.10347826086956521,
"grad_norm": 94.58474731445312,
"learning_rate": 2.480800970561438e-06,
"loss": 1.9859470129013062,
"step": 119
},
{
"epoch": 0.10434782608695652,
"grad_norm": 87.78974151611328,
"learning_rate": 2.480168017952809e-06,
"loss": 0.7612618207931519,
"step": 120
},
{
"epoch": 0.10521739130434783,
"grad_norm": 275.7257385253906,
"learning_rate": 2.47952488368393e-06,
"loss": 1.101609468460083,
"step": 121
},
{
"epoch": 0.10608695652173913,
"grad_norm": 113.06259155273438,
"learning_rate": 2.478871573077791e-06,
"loss": 0.6124528646469116,
"step": 122
},
{
"epoch": 0.10695652173913044,
"grad_norm": 89.37154388427734,
"learning_rate": 2.4782080915416107e-06,
"loss": 0.9808050990104675,
"step": 123
},
{
"epoch": 0.10782608695652174,
"grad_norm": 107.48558807373047,
"learning_rate": 2.4775344445667883e-06,
"loss": 1.5289537906646729,
"step": 124
},
{
"epoch": 0.10869565217391304,
"grad_norm": 64.69747161865234,
"learning_rate": 2.4768506377288582e-06,
"loss": 1.0372562408447266,
"step": 125
},
{
"epoch": 0.10956521739130434,
"grad_norm": 212.69064331054688,
"learning_rate": 2.4761566766874457e-06,
"loss": 1.7842557430267334,
"step": 126
},
{
"epoch": 0.11043478260869566,
"grad_norm": 144.72122192382812,
"learning_rate": 2.475452567186217e-06,
"loss": 0.3821652829647064,
"step": 127
},
{
"epoch": 0.11130434782608696,
"grad_norm": 215.56402587890625,
"learning_rate": 2.474738315052835e-06,
"loss": 0.6543242931365967,
"step": 128
},
{
"epoch": 0.11217391304347826,
"grad_norm": 147.76312255859375,
"learning_rate": 2.474013926198908e-06,
"loss": 0.552076518535614,
"step": 129
},
{
"epoch": 0.11304347826086956,
"grad_norm": 132.79061889648438,
"learning_rate": 2.4732794066199444e-06,
"loss": 0.5234610438346863,
"step": 130
},
{
"epoch": 0.11391304347826087,
"grad_norm": 510.35894775390625,
"learning_rate": 2.472534762395299e-06,
"loss": 1.2488858699798584,
"step": 131
},
{
"epoch": 0.11478260869565217,
"grad_norm": 325.57958984375,
"learning_rate": 2.471779999688125e-06,
"loss": 1.0867407321929932,
"step": 132
},
{
"epoch": 0.11565217391304349,
"grad_norm": 546.8318481445312,
"learning_rate": 2.471015124745324e-06,
"loss": 2.032482624053955,
"step": 133
},
{
"epoch": 0.11652173913043479,
"grad_norm": 367.03155517578125,
"learning_rate": 2.470240143897492e-06,
"loss": 2.6561994552612305,
"step": 134
},
{
"epoch": 0.11739130434782609,
"grad_norm": 39.405216217041016,
"learning_rate": 2.4694550635588676e-06,
"loss": 0.20321732759475708,
"step": 135
},
{
"epoch": 0.11826086956521739,
"grad_norm": 231.1078643798828,
"learning_rate": 2.4686598902272794e-06,
"loss": 0.6954633593559265,
"step": 136
},
{
"epoch": 0.1191304347826087,
"grad_norm": 90.90170288085938,
"learning_rate": 2.4678546304840934e-06,
"loss": 2.372281074523926,
"step": 137
},
{
"epoch": 0.12,
"grad_norm": 166.088134765625,
"learning_rate": 2.4670392909941552e-06,
"loss": 1.6563303470611572,
"step": 138
},
{
"epoch": 0.1208695652173913,
"grad_norm": 96.55294036865234,
"learning_rate": 2.4662138785057387e-06,
"loss": 2.6319773197174072,
"step": 139
},
{
"epoch": 0.12173913043478261,
"grad_norm": 96.869384765625,
"learning_rate": 2.465378399850487e-06,
"loss": 0.6885479688644409,
"step": 140
},
{
"epoch": 0.12260869565217392,
"grad_norm": 169.0710906982422,
"learning_rate": 2.4645328619433575e-06,
"loss": 1.1338417530059814,
"step": 141
},
{
"epoch": 0.12347826086956522,
"grad_norm": 248.34930419921875,
"learning_rate": 2.4636772717825653e-06,
"loss": 1.3146528005599976,
"step": 142
},
{
"epoch": 0.12434782608695652,
"grad_norm": 51.504817962646484,
"learning_rate": 2.4628116364495236e-06,
"loss": 0.7832269668579102,
"step": 143
},
{
"epoch": 0.12521739130434784,
"grad_norm": 48.48396301269531,
"learning_rate": 2.4619359631087863e-06,
"loss": 0.41868847608566284,
"step": 144
},
{
"epoch": 0.12608695652173912,
"grad_norm": 130.2314453125,
"learning_rate": 2.461050259007988e-06,
"loss": 1.262451171875,
"step": 145
},
{
"epoch": 0.12695652173913044,
"grad_norm": 116.0643081665039,
"learning_rate": 2.4601545314777843e-06,
"loss": 2.908426284790039,
"step": 146
},
{
"epoch": 0.12782608695652173,
"grad_norm": 45.22342300415039,
"learning_rate": 2.459248787931792e-06,
"loss": 0.38275718688964844,
"step": 147
},
{
"epoch": 0.12869565217391304,
"grad_norm": 107.33024597167969,
"learning_rate": 2.4583330358665248e-06,
"loss": 1.5183453559875488,
"step": 148
},
{
"epoch": 0.12956521739130436,
"grad_norm": 272.31988525390625,
"learning_rate": 2.4574072828613354e-06,
"loss": 2.774092197418213,
"step": 149
},
{
"epoch": 0.13043478260869565,
"grad_norm": 120.921142578125,
"learning_rate": 2.4564715365783505e-06,
"loss": 0.8305889964103699,
"step": 150
},
{
"epoch": 0.13130434782608696,
"grad_norm": 550.00634765625,
"learning_rate": 2.4555258047624064e-06,
"loss": 2.5018184185028076,
"step": 151
},
{
"epoch": 0.13217391304347825,
"grad_norm": 181.6162109375,
"learning_rate": 2.4545700952409875e-06,
"loss": 1.7048686742782593,
"step": 152
},
{
"epoch": 0.13304347826086957,
"grad_norm": 189.5089111328125,
"learning_rate": 2.453604415924158e-06,
"loss": 0.5605365037918091,
"step": 153
},
{
"epoch": 0.13391304347826086,
"grad_norm": 144.29432678222656,
"learning_rate": 2.452628774804501e-06,
"loss": 0.6180440187454224,
"step": 154
},
{
"epoch": 0.13478260869565217,
"grad_norm": 199.5081329345703,
"learning_rate": 2.4516431799570485e-06,
"loss": 0.5576478242874146,
"step": 155
},
{
"epoch": 0.1356521739130435,
"grad_norm": 130.9627227783203,
"learning_rate": 2.4506476395392157e-06,
"loss": 0.912418007850647,
"step": 156
},
{
"epoch": 0.13652173913043478,
"grad_norm": 231.45712280273438,
"learning_rate": 2.4496421617907347e-06,
"loss": 1.0736489295959473,
"step": 157
},
{
"epoch": 0.1373913043478261,
"grad_norm": 337.33953857421875,
"learning_rate": 2.4486267550335848e-06,
"loss": 3.842414379119873,
"step": 158
},
{
"epoch": 0.13826086956521738,
"grad_norm": 51.868228912353516,
"learning_rate": 2.447601427671924e-06,
"loss": 0.4069295823574066,
"step": 159
},
{
"epoch": 0.1391304347826087,
"grad_norm": 78.15707397460938,
"learning_rate": 2.44656618819202e-06,
"loss": 1.728126883506775,
"step": 160
},
{
"epoch": 0.14,
"grad_norm": 71.64280700683594,
"learning_rate": 2.445521045162179e-06,
"loss": 1.297673225402832,
"step": 161
},
{
"epoch": 0.1408695652173913,
"grad_norm": 812.8759155273438,
"learning_rate": 2.4444660072326762e-06,
"loss": 0.9168662428855896,
"step": 162
},
{
"epoch": 0.14173913043478262,
"grad_norm": 57.765438079833984,
"learning_rate": 2.4434010831356824e-06,
"loss": 0.16194552183151245,
"step": 163
},
{
"epoch": 0.1426086956521739,
"grad_norm": 179.18353271484375,
"learning_rate": 2.4423262816851927e-06,
"loss": 0.6879428625106812,
"step": 164
},
{
"epoch": 0.14347826086956522,
"grad_norm": 68.3979263305664,
"learning_rate": 2.441241611776955e-06,
"loss": 0.5606850385665894,
"step": 165
},
{
"epoch": 0.1443478260869565,
"grad_norm": 56.23814010620117,
"learning_rate": 2.4401470823883926e-06,
"loss": 0.40830135345458984,
"step": 166
},
{
"epoch": 0.14521739130434783,
"grad_norm": 1063.44482421875,
"learning_rate": 2.4390427025785345e-06,
"loss": 2.612975835800171,
"step": 167
},
{
"epoch": 0.14608695652173914,
"grad_norm": 390.76300048828125,
"learning_rate": 2.437928481487936e-06,
"loss": 0.7162667512893677,
"step": 168
},
{
"epoch": 0.14695652173913043,
"grad_norm": 120.74208068847656,
"learning_rate": 2.436804428338607e-06,
"loss": 0.9277188181877136,
"step": 169
},
{
"epoch": 0.14782608695652175,
"grad_norm": 242.52012634277344,
"learning_rate": 2.4356705524339317e-06,
"loss": 0.4081736207008362,
"step": 170
},
{
"epoch": 0.14869565217391303,
"grad_norm": 181.28346252441406,
"learning_rate": 2.4345268631585966e-06,
"loss": 0.7378160953521729,
"step": 171
},
{
"epoch": 0.14956521739130435,
"grad_norm": 83.74777221679688,
"learning_rate": 2.433373369978508e-06,
"loss": 0.963422954082489,
"step": 172
},
{
"epoch": 0.15043478260869564,
"grad_norm": 166.9161834716797,
"learning_rate": 2.4322100824407148e-06,
"loss": 0.4587669372558594,
"step": 173
},
{
"epoch": 0.15130434782608695,
"grad_norm": 41.084144592285156,
"learning_rate": 2.431037010173332e-06,
"loss": 0.19999371469020844,
"step": 174
},
{
"epoch": 0.15217391304347827,
"grad_norm": 97.73751068115234,
"learning_rate": 2.4298541628854596e-06,
"loss": 0.47642168402671814,
"step": 175
},
{
"epoch": 0.15304347826086956,
"grad_norm": 65.47894287109375,
"learning_rate": 2.4286615503671005e-06,
"loss": 0.832300066947937,
"step": 176
},
{
"epoch": 0.15391304347826087,
"grad_norm": 455.4596252441406,
"learning_rate": 2.4274591824890815e-06,
"loss": 1.4678276777267456,
"step": 177
},
{
"epoch": 0.15478260869565216,
"grad_norm": 214.0467071533203,
"learning_rate": 2.426247069202971e-06,
"loss": 2.0047309398651123,
"step": 178
},
{
"epoch": 0.15565217391304348,
"grad_norm": 68.03226470947266,
"learning_rate": 2.4250252205409962e-06,
"loss": 0.2981765866279602,
"step": 179
},
{
"epoch": 0.1565217391304348,
"grad_norm": 139.72267150878906,
"learning_rate": 2.4237936466159624e-06,
"loss": 0.3308563828468323,
"step": 180
},
{
"epoch": 0.15739130434782608,
"grad_norm": 288.9244384765625,
"learning_rate": 2.422552357621165e-06,
"loss": 1.022907018661499,
"step": 181
},
{
"epoch": 0.1582608695652174,
"grad_norm": 201.04246520996094,
"learning_rate": 2.421301363830309e-06,
"loss": 1.0035773515701294,
"step": 182
},
{
"epoch": 0.1591304347826087,
"grad_norm": 76.2557144165039,
"learning_rate": 2.4200406755974216e-06,
"loss": 0.6603399515151978,
"step": 183
},
{
"epoch": 0.16,
"grad_norm": 65.17644500732422,
"learning_rate": 2.418770303356769e-06,
"loss": 0.4546869695186615,
"step": 184
},
{
"epoch": 0.1608695652173913,
"grad_norm": 132.39784240722656,
"learning_rate": 2.417490257622766e-06,
"loss": 0.7756654024124146,
"step": 185
},
{
"epoch": 0.1617391304347826,
"grad_norm": 92.98999786376953,
"learning_rate": 2.4162005489898937e-06,
"loss": 0.43417030572891235,
"step": 186
},
{
"epoch": 0.16260869565217392,
"grad_norm": 103.86365509033203,
"learning_rate": 2.4149011881326085e-06,
"loss": 0.6076518297195435,
"step": 187
},
{
"epoch": 0.1634782608695652,
"grad_norm": 42.93599319458008,
"learning_rate": 2.4135921858052556e-06,
"loss": 0.18312108516693115,
"step": 188
},
{
"epoch": 0.16434782608695653,
"grad_norm": 743.072021484375,
"learning_rate": 2.412273552841978e-06,
"loss": 0.9056973457336426,
"step": 189
},
{
"epoch": 0.16521739130434782,
"grad_norm": 60.07756423950195,
"learning_rate": 2.410945300156628e-06,
"loss": 0.49955257773399353,
"step": 190
},
{
"epoch": 0.16608695652173913,
"grad_norm": 592.865478515625,
"learning_rate": 2.4096074387426788e-06,
"loss": 1.0758520364761353,
"step": 191
},
{
"epoch": 0.16695652173913045,
"grad_norm": 94.58853912353516,
"learning_rate": 2.40825997967313e-06,
"loss": 1.2590835094451904,
"step": 192
},
{
"epoch": 0.16782608695652174,
"grad_norm": 469.5373840332031,
"learning_rate": 2.406902934100417e-06,
"loss": 0.955599308013916,
"step": 193
},
{
"epoch": 0.16869565217391305,
"grad_norm": 257.1993103027344,
"learning_rate": 2.405536313256321e-06,
"loss": 0.8033830523490906,
"step": 194
},
{
"epoch": 0.16956521739130434,
"grad_norm": 249.76739501953125,
"learning_rate": 2.4041601284518734e-06,
"loss": 2.418315887451172,
"step": 195
},
{
"epoch": 0.17043478260869566,
"grad_norm": 37.73283767700195,
"learning_rate": 2.4027743910772634e-06,
"loss": 0.2811780571937561,
"step": 196
},
{
"epoch": 0.17130434782608694,
"grad_norm": 147.74484252929688,
"learning_rate": 2.401379112601742e-06,
"loss": 0.5272519588470459,
"step": 197
},
{
"epoch": 0.17217391304347826,
"grad_norm": 136.44337463378906,
"learning_rate": 2.3999743045735305e-06,
"loss": 1.1866902112960815,
"step": 198
},
{
"epoch": 0.17304347826086958,
"grad_norm": 145.77249145507812,
"learning_rate": 2.3985599786197216e-06,
"loss": 1.8868203163146973,
"step": 199
},
{
"epoch": 0.17391304347826086,
"grad_norm": 224.68118286132812,
"learning_rate": 2.3971361464461855e-06,
"loss": 0.8317220211029053,
"step": 200
},
{
"epoch": 0.17478260869565218,
"grad_norm": 349.6592102050781,
"learning_rate": 2.3957028198374703e-06,
"loss": 3.0272631645202637,
"step": 201
},
{
"epoch": 0.17565217391304347,
"grad_norm": 197.87518310546875,
"learning_rate": 2.394260010656708e-06,
"loss": 0.8683797121047974,
"step": 202
},
{
"epoch": 0.17652173913043478,
"grad_norm": 61.49953842163086,
"learning_rate": 2.3928077308455134e-06,
"loss": 0.3128216862678528,
"step": 203
},
{
"epoch": 0.17739130434782607,
"grad_norm": 345.0967102050781,
"learning_rate": 2.3913459924238865e-06,
"loss": 4.4456787109375,
"step": 204
},
{
"epoch": 0.1782608695652174,
"grad_norm": 143.3613739013672,
"learning_rate": 2.3898748074901127e-06,
"loss": 2.1473073959350586,
"step": 205
},
{
"epoch": 0.1791304347826087,
"grad_norm": 658.1432495117188,
"learning_rate": 2.3883941882206635e-06,
"loss": 1.0932512283325195,
"step": 206
},
{
"epoch": 0.18,
"grad_norm": 115.35150146484375,
"learning_rate": 2.386904146870094e-06,
"loss": 1.336449384689331,
"step": 207
},
{
"epoch": 0.1808695652173913,
"grad_norm": 64.45808410644531,
"learning_rate": 2.3854046957709425e-06,
"loss": 0.5701490640640259,
"step": 208
},
{
"epoch": 0.1817391304347826,
"grad_norm": 328.601318359375,
"learning_rate": 2.3838958473336302e-06,
"loss": 0.5811010599136353,
"step": 209
},
{
"epoch": 0.1826086956521739,
"grad_norm": 103.48978424072266,
"learning_rate": 2.3823776140463545e-06,
"loss": 0.3989741802215576,
"step": 210
},
{
"epoch": 0.18347826086956523,
"grad_norm": 168.3060760498047,
"learning_rate": 2.3808500084749893e-06,
"loss": 1.1245806217193604,
"step": 211
},
{
"epoch": 0.18434782608695652,
"grad_norm": 94.9990463256836,
"learning_rate": 2.379313043262978e-06,
"loss": 1.641979694366455,
"step": 212
},
{
"epoch": 0.18521739130434783,
"grad_norm": 586.6797485351562,
"learning_rate": 2.3777667311312326e-06,
"loss": 0.616899311542511,
"step": 213
},
{
"epoch": 0.18608695652173912,
"grad_norm": 66.73456573486328,
"learning_rate": 2.376211084878024e-06,
"loss": 1.5442198514938354,
"step": 214
},
{
"epoch": 0.18695652173913044,
"grad_norm": 227.69723510742188,
"learning_rate": 2.37464611737888e-06,
"loss": 1.409386396408081,
"step": 215
},
{
"epoch": 0.18782608695652173,
"grad_norm": 119.45149993896484,
"learning_rate": 2.3730718415864763e-06,
"loss": 0.7249864339828491,
"step": 216
},
{
"epoch": 0.18869565217391304,
"grad_norm": 41.28050994873047,
"learning_rate": 2.371488270530529e-06,
"loss": 0.8691262602806091,
"step": 217
},
{
"epoch": 0.18956521739130436,
"grad_norm": 111.85698699951172,
"learning_rate": 2.36989541731769e-06,
"loss": 2.2550487518310547,
"step": 218
},
{
"epoch": 0.19043478260869565,
"grad_norm": 140.46804809570312,
"learning_rate": 2.3682932951314343e-06,
"loss": 0.7733378410339355,
"step": 219
},
{
"epoch": 0.19130434782608696,
"grad_norm": 63.75563430786133,
"learning_rate": 2.366681917231954e-06,
"loss": 0.27650198340415955,
"step": 220
},
{
"epoch": 0.19217391304347825,
"grad_norm": 423.31854248046875,
"learning_rate": 2.365061296956047e-06,
"loss": 0.7498263716697693,
"step": 221
},
{
"epoch": 0.19304347826086957,
"grad_norm": 118.12458038330078,
"learning_rate": 2.3634314477170066e-06,
"loss": 0.8427287340164185,
"step": 222
},
{
"epoch": 0.19391304347826088,
"grad_norm": 244.00949096679688,
"learning_rate": 2.3617923830045126e-06,
"loss": 1.4409527778625488,
"step": 223
},
{
"epoch": 0.19478260869565217,
"grad_norm": 33.63735580444336,
"learning_rate": 2.3601441163845157e-06,
"loss": 0.21779245138168335,
"step": 224
},
{
"epoch": 0.1956521739130435,
"grad_norm": 129.28109741210938,
"learning_rate": 2.3584866614991287e-06,
"loss": 0.7803496718406677,
"step": 225
},
{
"epoch": 0.19652173913043477,
"grad_norm": 46.4891242980957,
"learning_rate": 2.3568200320665123e-06,
"loss": 1.8099173307418823,
"step": 226
},
{
"epoch": 0.1973913043478261,
"grad_norm": 162.5518798828125,
"learning_rate": 2.355144241880762e-06,
"loss": 1.1460251808166504,
"step": 227
},
{
"epoch": 0.19826086956521738,
"grad_norm": 184.81568908691406,
"learning_rate": 2.3534593048117936e-06,
"loss": 0.7761359214782715,
"step": 228
},
{
"epoch": 0.1991304347826087,
"grad_norm": 90.52983856201172,
"learning_rate": 2.3517652348052268e-06,
"loss": 0.8244172930717468,
"step": 229
},
{
"epoch": 0.2,
"grad_norm": 89.99091339111328,
"learning_rate": 2.3500620458822734e-06,
"loss": 1.079744577407837,
"step": 230
},
{
"epoch": 0.2008695652173913,
"grad_norm": 53.900211334228516,
"learning_rate": 2.348349752139618e-06,
"loss": 0.2871057391166687,
"step": 231
},
{
"epoch": 0.20173913043478262,
"grad_norm": 207.09825134277344,
"learning_rate": 2.3466283677493025e-06,
"loss": 1.8728336095809937,
"step": 232
},
{
"epoch": 0.2026086956521739,
"grad_norm": 205.58786010742188,
"learning_rate": 2.34489790695861e-06,
"loss": 2.6323258876800537,
"step": 233
},
{
"epoch": 0.20347826086956522,
"grad_norm": 47.92216110229492,
"learning_rate": 2.3431583840899444e-06,
"loss": 0.3071705102920532,
"step": 234
},
{
"epoch": 0.20434782608695654,
"grad_norm": 178.76068115234375,
"learning_rate": 2.3414098135407137e-06,
"loss": 1.2083189487457275,
"step": 235
},
{
"epoch": 0.20521739130434782,
"grad_norm": 265.9340515136719,
"learning_rate": 2.3396522097832103e-06,
"loss": 2.1280312538146973,
"step": 236
},
{
"epoch": 0.20608695652173914,
"grad_norm": 43.541725158691406,
"learning_rate": 2.337885587364492e-06,
"loss": 0.4030447006225586,
"step": 237
},
{
"epoch": 0.20695652173913043,
"grad_norm": 213.9405517578125,
"learning_rate": 2.3361099609062584e-06,
"loss": 1.5078468322753906,
"step": 238
},
{
"epoch": 0.20782608695652174,
"grad_norm": 315.79046630859375,
"learning_rate": 2.3343253451047354e-06,
"loss": 0.5508573055267334,
"step": 239
},
{
"epoch": 0.20869565217391303,
"grad_norm": 482.69744873046875,
"learning_rate": 2.332531754730549e-06,
"loss": 2.2617902755737305,
"step": 240
},
{
"epoch": 0.20956521739130435,
"grad_norm": 46.16420364379883,
"learning_rate": 2.3307292046286035e-06,
"loss": 0.6573113203048706,
"step": 241
},
{
"epoch": 0.21043478260869566,
"grad_norm": 134.3013458251953,
"learning_rate": 2.3289177097179612e-06,
"loss": 0.4778707027435303,
"step": 242
},
{
"epoch": 0.21130434782608695,
"grad_norm": 123.29259490966797,
"learning_rate": 2.3270972849917176e-06,
"loss": 0.6680418848991394,
"step": 243
},
{
"epoch": 0.21217391304347827,
"grad_norm": 372.6416320800781,
"learning_rate": 2.325267945516876e-06,
"loss": 0.6925360560417175,
"step": 244
},
{
"epoch": 0.21304347826086956,
"grad_norm": 56.82334899902344,
"learning_rate": 2.3234297064342243e-06,
"loss": 0.23991772532463074,
"step": 245
},
{
"epoch": 0.21391304347826087,
"grad_norm": 106.43640899658203,
"learning_rate": 2.321582582958209e-06,
"loss": 1.3718976974487305,
"step": 246
},
{
"epoch": 0.21478260869565216,
"grad_norm": 58.7371940612793,
"learning_rate": 2.3197265903768095e-06,
"loss": 0.2983436584472656,
"step": 247
},
{
"epoch": 0.21565217391304348,
"grad_norm": 69.20380401611328,
"learning_rate": 2.3178617440514128e-06,
"loss": 0.7256472110748291,
"step": 248
},
{
"epoch": 0.2165217391304348,
"grad_norm": 207.98634338378906,
"learning_rate": 2.315988059416684e-06,
"loss": 1.7729918956756592,
"step": 249
},
{
"epoch": 0.21739130434782608,
"grad_norm": 115.02137756347656,
"learning_rate": 2.3141055519804393e-06,
"loss": 1.7988307476043701,
"step": 250
},
{
"epoch": 0.2182608695652174,
"grad_norm": 190.162109375,
"learning_rate": 2.3122142373235195e-06,
"loss": 0.6468178629875183,
"step": 251
},
{
"epoch": 0.21913043478260869,
"grad_norm": 86.558349609375,
"learning_rate": 2.3103141310996583e-06,
"loss": 0.5797602534294128,
"step": 252
},
{
"epoch": 0.22,
"grad_norm": 67.32994842529297,
"learning_rate": 2.3084052490353552e-06,
"loss": 0.23084008693695068,
"step": 253
},
{
"epoch": 0.22086956521739132,
"grad_norm": 559.7723999023438,
"learning_rate": 2.3064876069297436e-06,
"loss": 0.9130882024765015,
"step": 254
},
{
"epoch": 0.2217391304347826,
"grad_norm": 235.3927764892578,
"learning_rate": 2.3045612206544603e-06,
"loss": 1.2320632934570312,
"step": 255
},
{
"epoch": 0.22260869565217392,
"grad_norm": 35.825828552246094,
"learning_rate": 2.302626106153515e-06,
"loss": 0.17873142659664154,
"step": 256
},
{
"epoch": 0.2234782608695652,
"grad_norm": 136.6010284423828,
"learning_rate": 2.300682279443157e-06,
"loss": 1.607312560081482,
"step": 257
},
{
"epoch": 0.22434782608695653,
"grad_norm": 145.7604522705078,
"learning_rate": 2.298729756611745e-06,
"loss": 1.3880008459091187,
"step": 258
},
{
"epoch": 0.2252173913043478,
"grad_norm": 64.47338104248047,
"learning_rate": 2.296768553819611e-06,
"loss": 1.442543864250183,
"step": 259
},
{
"epoch": 0.22608695652173913,
"grad_norm": 169.54971313476562,
"learning_rate": 2.2947986872989286e-06,
"loss": 0.8224241137504578,
"step": 260
},
{
"epoch": 0.22695652173913045,
"grad_norm": 45.167747497558594,
"learning_rate": 2.2928201733535767e-06,
"loss": 0.2948722541332245,
"step": 261
},
{
"epoch": 0.22782608695652173,
"grad_norm": 49.49987030029297,
"learning_rate": 2.290833028359008e-06,
"loss": 0.3207906186580658,
"step": 262
},
{
"epoch": 0.22869565217391305,
"grad_norm": 51.862159729003906,
"learning_rate": 2.28883726876211e-06,
"loss": 0.5076830387115479,
"step": 263
},
{
"epoch": 0.22956521739130434,
"grad_norm": 64.63931274414062,
"learning_rate": 2.28683291108107e-06,
"loss": 0.2963521182537079,
"step": 264
},
{
"epoch": 0.23043478260869565,
"grad_norm": 53.02265930175781,
"learning_rate": 2.2848199719052388e-06,
"loss": 1.478598952293396,
"step": 265
},
{
"epoch": 0.23130434782608697,
"grad_norm": 71.54142761230469,
"learning_rate": 2.2827984678949937e-06,
"loss": 0.5322201251983643,
"step": 266
},
{
"epoch": 0.23217391304347826,
"grad_norm": 254.59388732910156,
"learning_rate": 2.2807684157815998e-06,
"loss": 1.5792579650878906,
"step": 267
},
{
"epoch": 0.23304347826086957,
"grad_norm": 207.41798400878906,
"learning_rate": 2.2787298323670707e-06,
"loss": 0.4023154377937317,
"step": 268
},
{
"epoch": 0.23391304347826086,
"grad_norm": 50.321533203125,
"learning_rate": 2.276682734524032e-06,
"loss": 0.29348844289779663,
"step": 269
},
{
"epoch": 0.23478260869565218,
"grad_norm": 50.237552642822266,
"learning_rate": 2.27462713919558e-06,
"loss": 0.7310285568237305,
"step": 270
},
{
"epoch": 0.23565217391304347,
"grad_norm": 83.97930908203125,
"learning_rate": 2.2725630633951413e-06,
"loss": 0.2441037893295288,
"step": 271
},
{
"epoch": 0.23652173913043478,
"grad_norm": 132.39205932617188,
"learning_rate": 2.2704905242063313e-06,
"loss": 1.0067497491836548,
"step": 272
},
{
"epoch": 0.2373913043478261,
"grad_norm": 121.62727355957031,
"learning_rate": 2.268409538782816e-06,
"loss": 1.845292568206787,
"step": 273
},
{
"epoch": 0.2382608695652174,
"grad_norm": 68.42473602294922,
"learning_rate": 2.2663201243481656e-06,
"loss": 0.4180206060409546,
"step": 274
},
{
"epoch": 0.2391304347826087,
"grad_norm": 63.45404815673828,
"learning_rate": 2.2642222981957163e-06,
"loss": 3.666378974914551,
"step": 275
},
{
"epoch": 0.24,
"grad_norm": 49.06626510620117,
"learning_rate": 2.2621160776884236e-06,
"loss": 0.9683353304862976,
"step": 276
},
{
"epoch": 0.2408695652173913,
"grad_norm": 827.3475341796875,
"learning_rate": 2.260001480258721e-06,
"loss": 0.9524872303009033,
"step": 277
},
{
"epoch": 0.2417391304347826,
"grad_norm": 62.60003662109375,
"learning_rate": 2.2578785234083734e-06,
"loss": 0.34575188159942627,
"step": 278
},
{
"epoch": 0.2426086956521739,
"grad_norm": 64.26739501953125,
"learning_rate": 2.2557472247083356e-06,
"loss": 1.157808780670166,
"step": 279
},
{
"epoch": 0.24347826086956523,
"grad_norm": 134.6650848388672,
"learning_rate": 2.2536076017986033e-06,
"loss": 1.6884756088256836,
"step": 280
},
{
"epoch": 0.24434782608695652,
"grad_norm": 95.88436889648438,
"learning_rate": 2.251459672388069e-06,
"loss": 0.7435152530670166,
"step": 281
},
{
"epoch": 0.24521739130434783,
"grad_norm": 44.57541275024414,
"learning_rate": 2.249303454254377e-06,
"loss": 1.0371330976486206,
"step": 282
},
{
"epoch": 0.24608695652173912,
"grad_norm": 36.62701416015625,
"learning_rate": 2.2471389652437704e-06,
"loss": 0.3761155307292938,
"step": 283
},
{
"epoch": 0.24695652173913044,
"grad_norm": 118.8804931640625,
"learning_rate": 2.244966223270952e-06,
"loss": 0.48897528648376465,
"step": 284
},
{
"epoch": 0.24782608695652175,
"grad_norm": 68.78189849853516,
"learning_rate": 2.2427852463189274e-06,
"loss": 0.5494109988212585,
"step": 285
},
{
"epoch": 0.24869565217391304,
"grad_norm": 169.46231079101562,
"learning_rate": 2.2405960524388633e-06,
"loss": 1.9049848318099976,
"step": 286
},
{
"epoch": 0.24956521739130436,
"grad_norm": 28.044527053833008,
"learning_rate": 2.238398659749933e-06,
"loss": 0.15847647190093994,
"step": 287
},
{
"epoch": 0.25043478260869567,
"grad_norm": 45.294769287109375,
"learning_rate": 2.2361930864391696e-06,
"loss": 0.2587065100669861,
"step": 288
},
{
"epoch": 0.25130434782608696,
"grad_norm": 436.86737060546875,
"learning_rate": 2.2339793507613128e-06,
"loss": 0.5645865201950073,
"step": 289
},
{
"epoch": 0.25217391304347825,
"grad_norm": 60.853641510009766,
"learning_rate": 2.2317574710386615e-06,
"loss": 0.26602864265441895,
"step": 290
},
{
"epoch": 0.2530434782608696,
"grad_norm": 87.82923889160156,
"learning_rate": 2.2295274656609174e-06,
"loss": 1.062896966934204,
"step": 291
},
{
"epoch": 0.2539130434782609,
"grad_norm": 38.46051788330078,
"learning_rate": 2.2272893530850375e-06,
"loss": 0.3079209327697754,
"step": 292
},
{
"epoch": 0.25478260869565217,
"grad_norm": 60.85812759399414,
"learning_rate": 2.2250431518350777e-06,
"loss": 0.7608625888824463,
"step": 293
},
{
"epoch": 0.25565217391304346,
"grad_norm": 119.5645523071289,
"learning_rate": 2.2227888805020424e-06,
"loss": 0.6012105941772461,
"step": 294
},
{
"epoch": 0.2565217391304348,
"grad_norm": 85.98336029052734,
"learning_rate": 2.2205265577437274e-06,
"loss": 2.0251975059509277,
"step": 295
},
{
"epoch": 0.2573913043478261,
"grad_norm": 300.4337158203125,
"learning_rate": 2.218256202284568e-06,
"loss": 1.1006145477294922,
"step": 296
},
{
"epoch": 0.2582608695652174,
"grad_norm": 194.73614501953125,
"learning_rate": 2.215977832915485e-06,
"loss": 1.0468895435333252,
"step": 297
},
{
"epoch": 0.2591304347826087,
"grad_norm": 56.733272552490234,
"learning_rate": 2.2136914684937245e-06,
"loss": 0.5945777893066406,
"step": 298
},
{
"epoch": 0.26,
"grad_norm": 76.87342071533203,
"learning_rate": 2.211397127942707e-06,
"loss": 0.6825685501098633,
"step": 299
},
{
"epoch": 0.2608695652173913,
"grad_norm": 83.68196868896484,
"learning_rate": 2.2090948302518676e-06,
"loss": 0.964439868927002,
"step": 300
},
{
"epoch": 0.2617391304347826,
"grad_norm": 44.92335891723633,
"learning_rate": 2.2067845944764997e-06,
"loss": 1.8274710178375244,
"step": 301
},
{
"epoch": 0.26260869565217393,
"grad_norm": 61.318336486816406,
"learning_rate": 2.2044664397375977e-06,
"loss": 0.432664155960083,
"step": 302
},
{
"epoch": 0.2634782608695652,
"grad_norm": 398.97222900390625,
"learning_rate": 2.202140385221699e-06,
"loss": 1.1320068836212158,
"step": 303
},
{
"epoch": 0.2643478260869565,
"grad_norm": 81.70048522949219,
"learning_rate": 2.1998064501807238e-06,
"loss": 0.6652305126190186,
"step": 304
},
{
"epoch": 0.26521739130434785,
"grad_norm": 72.06222534179688,
"learning_rate": 2.1974646539318164e-06,
"loss": 0.780421257019043,
"step": 305
},
{
"epoch": 0.26608695652173914,
"grad_norm": 84.46443176269531,
"learning_rate": 2.1951150158571862e-06,
"loss": 0.9983062148094177,
"step": 306
},
{
"epoch": 0.2669565217391304,
"grad_norm": 71.51422119140625,
"learning_rate": 2.1927575554039465e-06,
"loss": 0.4108102321624756,
"step": 307
},
{
"epoch": 0.2678260869565217,
"grad_norm": 101.89991760253906,
"learning_rate": 2.1903922920839535e-06,
"loss": 0.39721137285232544,
"step": 308
},
{
"epoch": 0.26869565217391306,
"grad_norm": 43.29629898071289,
"learning_rate": 2.188019245473644e-06,
"loss": 0.4325398802757263,
"step": 309
},
{
"epoch": 0.26956521739130435,
"grad_norm": 56.989990234375,
"learning_rate": 2.1856384352138767e-06,
"loss": 0.4168890416622162,
"step": 310
},
{
"epoch": 0.27043478260869563,
"grad_norm": 88.65746307373047,
"learning_rate": 2.1832498810097643e-06,
"loss": 0.8762995004653931,
"step": 311
},
{
"epoch": 0.271304347826087,
"grad_norm": 152.4781951904297,
"learning_rate": 2.1808536026305154e-06,
"loss": 3.134538412094116,
"step": 312
},
{
"epoch": 0.27217391304347827,
"grad_norm": 296.46527099609375,
"learning_rate": 2.178449619909268e-06,
"loss": 2.159223794937134,
"step": 313
},
{
"epoch": 0.27304347826086955,
"grad_norm": 61.19507598876953,
"learning_rate": 2.176037952742927e-06,
"loss": 0.3710782527923584,
"step": 314
},
{
"epoch": 0.27391304347826084,
"grad_norm": 42.50086212158203,
"learning_rate": 2.1736186210919974e-06,
"loss": 2.1895368099212646,
"step": 315
},
{
"epoch": 0.2747826086956522,
"grad_norm": 55.04458236694336,
"learning_rate": 2.171191644980421e-06,
"loss": 0.35886889696121216,
"step": 316
},
{
"epoch": 0.2756521739130435,
"grad_norm": 176.07586669921875,
"learning_rate": 2.168757044495411e-06,
"loss": 1.5998897552490234,
"step": 317
},
{
"epoch": 0.27652173913043476,
"grad_norm": 28.953813552856445,
"learning_rate": 2.166314839787283e-06,
"loss": 1.7540143728256226,
"step": 318
},
{
"epoch": 0.2773913043478261,
"grad_norm": 156.856689453125,
"learning_rate": 2.1638650510692916e-06,
"loss": 1.4405931234359741,
"step": 319
},
{
"epoch": 0.2782608695652174,
"grad_norm": 129.12429809570312,
"learning_rate": 2.161407698617461e-06,
"loss": 0.5596458911895752,
"step": 320
},
{
"epoch": 0.2791304347826087,
"grad_norm": 23.430402755737305,
"learning_rate": 2.1589428027704174e-06,
"loss": 0.15443159639835358,
"step": 321
},
{
"epoch": 0.28,
"grad_norm": 68.46884155273438,
"learning_rate": 2.156470383929221e-06,
"loss": 1.5564182996749878,
"step": 322
},
{
"epoch": 0.2808695652173913,
"grad_norm": 57.81913375854492,
"learning_rate": 2.1539904625571973e-06,
"loss": 0.5841474533081055,
"step": 323
},
{
"epoch": 0.2817391304347826,
"grad_norm": 116.9150619506836,
"learning_rate": 2.151503059179768e-06,
"loss": 0.4702639579772949,
"step": 324
},
{
"epoch": 0.2826086956521739,
"grad_norm": 100.0834732055664,
"learning_rate": 2.149008194384281e-06,
"loss": 0.4966408610343933,
"step": 325
},
{
"epoch": 0.28347826086956524,
"grad_norm": 49.64389419555664,
"learning_rate": 2.1465058888198376e-06,
"loss": 1.6600520610809326,
"step": 326
},
{
"epoch": 0.2843478260869565,
"grad_norm": 73.4475326538086,
"learning_rate": 2.1439961631971256e-06,
"loss": 0.8089355230331421,
"step": 327
},
{
"epoch": 0.2852173913043478,
"grad_norm": 436.1669921875,
"learning_rate": 2.1414790382882455e-06,
"loss": 1.8465827703475952,
"step": 328
},
{
"epoch": 0.28608695652173916,
"grad_norm": 201.4010467529297,
"learning_rate": 2.138954534926539e-06,
"loss": 0.731433093547821,
"step": 329
},
{
"epoch": 0.28695652173913044,
"grad_norm": 47.61180114746094,
"learning_rate": 2.1364226740064166e-06,
"loss": 0.9296051263809204,
"step": 330
},
{
"epoch": 0.28782608695652173,
"grad_norm": 198.73399353027344,
"learning_rate": 2.1338834764831847e-06,
"loss": 1.222153902053833,
"step": 331
},
{
"epoch": 0.288695652173913,
"grad_norm": 151.0773468017578,
"learning_rate": 2.131336963372872e-06,
"loss": 0.44245728850364685,
"step": 332
},
{
"epoch": 0.28956521739130436,
"grad_norm": 255.4181365966797,
"learning_rate": 2.1287831557520554e-06,
"loss": 0.7557501196861267,
"step": 333
},
{
"epoch": 0.29043478260869565,
"grad_norm": 431.8282470703125,
"learning_rate": 2.126222074757687e-06,
"loss": 1.8245795965194702,
"step": 334
},
{
"epoch": 0.29130434782608694,
"grad_norm": 395.6104431152344,
"learning_rate": 2.1236537415869167e-06,
"loss": 1.3758153915405273,
"step": 335
},
{
"epoch": 0.2921739130434783,
"grad_norm": 107.71965026855469,
"learning_rate": 2.1210781774969193e-06,
"loss": 0.5021981000900269,
"step": 336
},
{
"epoch": 0.29304347826086957,
"grad_norm": 258.7772521972656,
"learning_rate": 2.1184954038047157e-06,
"loss": 1.785605788230896,
"step": 337
},
{
"epoch": 0.29391304347826086,
"grad_norm": 1326.66943359375,
"learning_rate": 2.1159054418869994e-06,
"loss": 0.5972184538841248,
"step": 338
},
{
"epoch": 0.29478260869565215,
"grad_norm": 138.0780029296875,
"learning_rate": 2.1133083131799582e-06,
"loss": 0.374930739402771,
"step": 339
},
{
"epoch": 0.2956521739130435,
"grad_norm": 61.293521881103516,
"learning_rate": 2.1107040391790965e-06,
"loss": 1.069291353225708,
"step": 340
},
{
"epoch": 0.2965217391304348,
"grad_norm": 184.9951934814453,
"learning_rate": 2.1080926414390575e-06,
"loss": 2.037717819213867,
"step": 341
},
{
"epoch": 0.29739130434782607,
"grad_norm": 258.1242980957031,
"learning_rate": 2.105474141573446e-06,
"loss": 0.7847839593887329,
"step": 342
},
{
"epoch": 0.2982608695652174,
"grad_norm": 198.63629150390625,
"learning_rate": 2.102848561254647e-06,
"loss": 0.5499667525291443,
"step": 343
},
{
"epoch": 0.2991304347826087,
"grad_norm": 39.34695053100586,
"learning_rate": 2.1002159222136494e-06,
"loss": 0.2706986665725708,
"step": 344
},
{
"epoch": 0.3,
"grad_norm": 43.83790969848633,
"learning_rate": 2.0975762462398637e-06,
"loss": 0.3612968921661377,
"step": 345
},
{
"epoch": 0.3008695652173913,
"grad_norm": 131.55844116210938,
"learning_rate": 2.094929555180943e-06,
"loss": 0.6884905695915222,
"step": 346
},
{
"epoch": 0.3017391304347826,
"grad_norm": 158.2893829345703,
"learning_rate": 2.0922758709426018e-06,
"loss": 0.8799026012420654,
"step": 347
},
{
"epoch": 0.3026086956521739,
"grad_norm": 700.5369873046875,
"learning_rate": 2.089615215488434e-06,
"loss": 3.039418935775757,
"step": 348
},
{
"epoch": 0.3034782608695652,
"grad_norm": 228.27171325683594,
"learning_rate": 2.0869476108397324e-06,
"loss": 1.057826042175293,
"step": 349
},
{
"epoch": 0.30434782608695654,
"grad_norm": 47.666015625,
"learning_rate": 2.084273079075306e-06,
"loss": 0.2861756384372711,
"step": 350
},
{
"epoch": 0.30521739130434783,
"grad_norm": 273.7340087890625,
"learning_rate": 2.081591642331296e-06,
"loss": 1.052135944366455,
"step": 351
},
{
"epoch": 0.3060869565217391,
"grad_norm": 64.19417572021484,
"learning_rate": 2.0789033228009943e-06,
"loss": 0.5136332511901855,
"step": 352
},
{
"epoch": 0.30695652173913046,
"grad_norm": 75.7625732421875,
"learning_rate": 2.07620814273466e-06,
"loss": 0.409485787153244,
"step": 353
},
{
"epoch": 0.30782608695652175,
"grad_norm": 140.1576690673828,
"learning_rate": 2.073506124439332e-06,
"loss": 0.9932225942611694,
"step": 354
},
{
"epoch": 0.30869565217391304,
"grad_norm": 69.31475067138672,
"learning_rate": 2.0707972902786504e-06,
"loss": 0.3649193048477173,
"step": 355
},
{
"epoch": 0.3095652173913043,
"grad_norm": 61.380828857421875,
"learning_rate": 2.0680816626726648e-06,
"loss": 0.2584507465362549,
"step": 356
},
{
"epoch": 0.31043478260869567,
"grad_norm": 47.857200622558594,
"learning_rate": 2.0653592640976517e-06,
"loss": 0.3207821846008301,
"step": 357
},
{
"epoch": 0.31130434782608696,
"grad_norm": 48.18440246582031,
"learning_rate": 2.0626301170859305e-06,
"loss": 0.3509800136089325,
"step": 358
},
{
"epoch": 0.31217391304347825,
"grad_norm": 129.9929962158203,
"learning_rate": 2.0598942442256736e-06,
"loss": 0.9874853491783142,
"step": 359
},
{
"epoch": 0.3130434782608696,
"grad_norm": 79.7518539428711,
"learning_rate": 2.057151668160721e-06,
"loss": 0.8659735918045044,
"step": 360
},
{
"epoch": 0.3139130434782609,
"grad_norm": 466.22161865234375,
"learning_rate": 2.0544024115903935e-06,
"loss": 0.9067327380180359,
"step": 361
},
{
"epoch": 0.31478260869565217,
"grad_norm": 178.5908203125,
"learning_rate": 2.0516464972693027e-06,
"loss": 0.774965763092041,
"step": 362
},
{
"epoch": 0.31565217391304345,
"grad_norm": 82.63011169433594,
"learning_rate": 2.0488839480071653e-06,
"loss": 0.5534073710441589,
"step": 363
},
{
"epoch": 0.3165217391304348,
"grad_norm": 112.5412368774414,
"learning_rate": 2.0461147866686116e-06,
"loss": 0.8640887141227722,
"step": 364
},
{
"epoch": 0.3173913043478261,
"grad_norm": 108.6795654296875,
"learning_rate": 2.0433390361729994e-06,
"loss": 1.061457633972168,
"step": 365
},
{
"epoch": 0.3182608695652174,
"grad_norm": 124.81582641601562,
"learning_rate": 2.0405567194942215e-06,
"loss": 1.418869972229004,
"step": 366
},
{
"epoch": 0.3191304347826087,
"grad_norm": 79.2950668334961,
"learning_rate": 2.0377678596605173e-06,
"loss": 0.9944026470184326,
"step": 367
},
{
"epoch": 0.32,
"grad_norm": 75.5735855102539,
"learning_rate": 2.03497247975428e-06,
"loss": 0.4054933488368988,
"step": 368
},
{
"epoch": 0.3208695652173913,
"grad_norm": 52.94697570800781,
"learning_rate": 2.03217060291187e-06,
"loss": 0.682574987411499,
"step": 369
},
{
"epoch": 0.3217391304347826,
"grad_norm": 95.7962417602539,
"learning_rate": 2.029362252323417e-06,
"loss": 2.769209623336792,
"step": 370
},
{
"epoch": 0.3226086956521739,
"grad_norm": 33.21056365966797,
"learning_rate": 2.0265474512326345e-06,
"loss": 0.21800105273723602,
"step": 371
},
{
"epoch": 0.3234782608695652,
"grad_norm": 59.40690231323242,
"learning_rate": 2.023726222936623e-06,
"loss": 0.4955858588218689,
"step": 372
},
{
"epoch": 0.3243478260869565,
"grad_norm": 144.7593536376953,
"learning_rate": 2.020898590785679e-06,
"loss": 0.6549877524375916,
"step": 373
},
{
"epoch": 0.32521739130434785,
"grad_norm": 96.69871520996094,
"learning_rate": 2.0180645781831017e-06,
"loss": 1.5156177282333374,
"step": 374
},
{
"epoch": 0.32608695652173914,
"grad_norm": 33.73798370361328,
"learning_rate": 2.015224208584998e-06,
"loss": 1.3318308591842651,
"step": 375
},
{
"epoch": 0.3269565217391304,
"grad_norm": 71.1515884399414,
"learning_rate": 2.0123775055000912e-06,
"loss": 1.002931833267212,
"step": 376
},
{
"epoch": 0.3278260869565217,
"grad_norm": 116.67143249511719,
"learning_rate": 2.009524492489523e-06,
"loss": 2.1182701587677,
"step": 377
},
{
"epoch": 0.32869565217391306,
"grad_norm": 96.71919250488281,
"learning_rate": 2.00666519316666e-06,
"loss": 0.42493805289268494,
"step": 378
},
{
"epoch": 0.32956521739130434,
"grad_norm": 184.78302001953125,
"learning_rate": 2.0037996311969007e-06,
"loss": 0.7065421342849731,
"step": 379
},
{
"epoch": 0.33043478260869563,
"grad_norm": 189.4550323486328,
"learning_rate": 2.000927830297474e-06,
"loss": 0.6880316734313965,
"step": 380
},
{
"epoch": 0.331304347826087,
"grad_norm": 58.27922439575195,
"learning_rate": 1.9980498142372483e-06,
"loss": 0.3300993740558624,
"step": 381
},
{
"epoch": 0.33217391304347826,
"grad_norm": 33.25176239013672,
"learning_rate": 1.995165606836532e-06,
"loss": 0.2718711495399475,
"step": 382
},
{
"epoch": 0.33304347826086955,
"grad_norm": 46.02870178222656,
"learning_rate": 1.992275231966877e-06,
"loss": 0.23181107640266418,
"step": 383
},
{
"epoch": 0.3339130434782609,
"grad_norm": 160.23724365234375,
"learning_rate": 1.989378713550882e-06,
"loss": 0.4160184860229492,
"step": 384
},
{
"epoch": 0.3347826086956522,
"grad_norm": 35.04530715942383,
"learning_rate": 1.986476075561992e-06,
"loss": 0.27640804648399353,
"step": 385
},
{
"epoch": 0.3356521739130435,
"grad_norm": 63.7939567565918,
"learning_rate": 1.9835673420243033e-06,
"loss": 0.6397489309310913,
"step": 386
},
{
"epoch": 0.33652173913043476,
"grad_norm": 32.473228454589844,
"learning_rate": 1.980652537012361e-06,
"loss": 0.9566623568534851,
"step": 387
},
{
"epoch": 0.3373913043478261,
"grad_norm": 29.80388069152832,
"learning_rate": 1.9777316846509638e-06,
"loss": 0.3189791142940521,
"step": 388
},
{
"epoch": 0.3382608695652174,
"grad_norm": 26.440004348754883,
"learning_rate": 1.97480480911496e-06,
"loss": 0.5863786935806274,
"step": 389
},
{
"epoch": 0.3391304347826087,
"grad_norm": 63.7293815612793,
"learning_rate": 1.97187193462905e-06,
"loss": 1.088039755821228,
"step": 390
},
{
"epoch": 0.34,
"grad_norm": 45.85379409790039,
"learning_rate": 1.968933085467587e-06,
"loss": 1.0376955270767212,
"step": 391
},
{
"epoch": 0.3408695652173913,
"grad_norm": 816.1434936523438,
"learning_rate": 1.9659882859543733e-06,
"loss": 1.0288197994232178,
"step": 392
},
{
"epoch": 0.3417391304347826,
"grad_norm": 76.29949188232422,
"learning_rate": 1.9630375604624596e-06,
"loss": 1.2274682521820068,
"step": 393
},
{
"epoch": 0.3426086956521739,
"grad_norm": 87.45259094238281,
"learning_rate": 1.960080933413945e-06,
"loss": 0.652985155582428,
"step": 394
},
{
"epoch": 0.34347826086956523,
"grad_norm": 89.88771057128906,
"learning_rate": 1.957118429279772e-06,
"loss": 2.4390976428985596,
"step": 395
},
{
"epoch": 0.3443478260869565,
"grad_norm": 42.63681411743164,
"learning_rate": 1.9541500725795277e-06,
"loss": 1.928074836730957,
"step": 396
},
{
"epoch": 0.3452173913043478,
"grad_norm": 301.94671630859375,
"learning_rate": 1.951175887881237e-06,
"loss": 1.0027554035186768,
"step": 397
},
{
"epoch": 0.34608695652173915,
"grad_norm": 98.58251953125,
"learning_rate": 1.9481958998011626e-06,
"loss": 1.190100908279419,
"step": 398
},
{
"epoch": 0.34695652173913044,
"grad_norm": 255.9281768798828,
"learning_rate": 1.945210133003598e-06,
"loss": 0.8519782423973083,
"step": 399
},
{
"epoch": 0.34782608695652173,
"grad_norm": 145.28924560546875,
"learning_rate": 1.9422186122006663e-06,
"loss": 3.067471981048584,
"step": 400
},
{
"epoch": 0.348695652173913,
"grad_norm": 27.750356674194336,
"learning_rate": 1.939221362152115e-06,
"loss": 0.2958664000034332,
"step": 401
},
{
"epoch": 0.34956521739130436,
"grad_norm": 46.79102325439453,
"learning_rate": 1.9362184076651082e-06,
"loss": 0.24861398339271545,
"step": 402
},
{
"epoch": 0.35043478260869565,
"grad_norm": 391.1067810058594,
"learning_rate": 1.9332097735940265e-06,
"loss": 0.544256329536438,
"step": 403
},
{
"epoch": 0.35130434782608694,
"grad_norm": 25.616535186767578,
"learning_rate": 1.9301954848402553e-06,
"loss": 0.2894223630428314,
"step": 404
},
{
"epoch": 0.3521739130434783,
"grad_norm": 25.0776424407959,
"learning_rate": 1.9271755663519854e-06,
"loss": 1.0108537673950195,
"step": 405
},
{
"epoch": 0.35304347826086957,
"grad_norm": 447.6708679199219,
"learning_rate": 1.924150043123999e-06,
"loss": 1.112715482711792,
"step": 406
},
{
"epoch": 0.35391304347826086,
"grad_norm": 95.64755249023438,
"learning_rate": 1.92111894019747e-06,
"loss": 0.4365922808647156,
"step": 407
},
{
"epoch": 0.35478260869565215,
"grad_norm": 64.63569641113281,
"learning_rate": 1.9180822826597516e-06,
"loss": 0.3381706476211548,
"step": 408
},
{
"epoch": 0.3556521739130435,
"grad_norm": 52.62809371948242,
"learning_rate": 1.915040095644171e-06,
"loss": 0.5142425298690796,
"step": 409
},
{
"epoch": 0.3565217391304348,
"grad_norm": 41.533145904541016,
"learning_rate": 1.9119924043298207e-06,
"loss": 0.2016807198524475,
"step": 410
},
{
"epoch": 0.35739130434782607,
"grad_norm": 83.78719329833984,
"learning_rate": 1.9089392339413513e-06,
"loss": 0.9166589975357056,
"step": 411
},
{
"epoch": 0.3582608695652174,
"grad_norm": 68.33119201660156,
"learning_rate": 1.90588060974876e-06,
"loss": 0.4687981605529785,
"step": 412
},
{
"epoch": 0.3591304347826087,
"grad_norm": 58.70621109008789,
"learning_rate": 1.9028165570671852e-06,
"loss": 0.3701656460762024,
"step": 413
},
{
"epoch": 0.36,
"grad_norm": 82.72734069824219,
"learning_rate": 1.8997471012566932e-06,
"loss": 0.4450571537017822,
"step": 414
},
{
"epoch": 0.36086956521739133,
"grad_norm": 76.3667221069336,
"learning_rate": 1.896672267722071e-06,
"loss": 1.8794634342193604,
"step": 415
},
{
"epoch": 0.3617391304347826,
"grad_norm": 82.80426025390625,
"learning_rate": 1.8935920819126152e-06,
"loss": 1.0356271266937256,
"step": 416
},
{
"epoch": 0.3626086956521739,
"grad_norm": 50.333797454833984,
"learning_rate": 1.8905065693219205e-06,
"loss": 0.2752545475959778,
"step": 417
},
{
"epoch": 0.3634782608695652,
"grad_norm": 58.108375549316406,
"learning_rate": 1.8874157554876701e-06,
"loss": 0.27540749311447144,
"step": 418
},
{
"epoch": 0.36434782608695654,
"grad_norm": 47.063690185546875,
"learning_rate": 1.8843196659914235e-06,
"loss": 0.2742215096950531,
"step": 419
},
{
"epoch": 0.3652173913043478,
"grad_norm": 922.4559326171875,
"learning_rate": 1.881218326458405e-06,
"loss": 1.5976967811584473,
"step": 420
},
{
"epoch": 0.3660869565217391,
"grad_norm": 33.08988571166992,
"learning_rate": 1.8781117625572915e-06,
"loss": 0.24918586015701294,
"step": 421
},
{
"epoch": 0.36695652173913046,
"grad_norm": 25.701887130737305,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.29249221086502075,
"step": 422
},
{
"epoch": 0.36782608695652175,
"grad_norm": 100.09066772460938,
"learning_rate": 1.8718830645414758e-06,
"loss": 2.1004226207733154,
"step": 423
},
{
"epoch": 0.36869565217391304,
"grad_norm": 64.07415008544922,
"learning_rate": 1.868760981979477e-06,
"loss": 0.2944200038909912,
"step": 424
},
{
"epoch": 0.3695652173913043,
"grad_norm": 56.589622497558594,
"learning_rate": 1.8656337781543632e-06,
"loss": 0.40554293990135193,
"step": 425
},
{
"epoch": 0.37043478260869567,
"grad_norm": 436.8203430175781,
"learning_rate": 1.8625014789488813e-06,
"loss": 0.8376158475875854,
"step": 426
},
{
"epoch": 0.37130434782608696,
"grad_norm": 72.22208404541016,
"learning_rate": 1.85936411028795e-06,
"loss": 0.5231876969337463,
"step": 427
},
{
"epoch": 0.37217391304347824,
"grad_norm": 109.61917114257812,
"learning_rate": 1.8562216981384467e-06,
"loss": 0.4562901258468628,
"step": 428
},
{
"epoch": 0.3730434782608696,
"grad_norm": 119.82129669189453,
"learning_rate": 1.853074268508992e-06,
"loss": 0.7159764170646667,
"step": 429
},
{
"epoch": 0.3739130434782609,
"grad_norm": 307.2554931640625,
"learning_rate": 1.8499218474497332e-06,
"loss": 1.2101820707321167,
"step": 430
},
{
"epoch": 0.37478260869565216,
"grad_norm": 150.56834411621094,
"learning_rate": 1.8467644610521313e-06,
"loss": 0.5387594699859619,
"step": 431
},
{
"epoch": 0.37565217391304345,
"grad_norm": 80.94709014892578,
"learning_rate": 1.8436021354487428e-06,
"loss": 1.0055359601974487,
"step": 432
},
{
"epoch": 0.3765217391304348,
"grad_norm": 50.03790283203125,
"learning_rate": 1.8404348968130048e-06,
"loss": 0.3763952851295471,
"step": 433
},
{
"epoch": 0.3773913043478261,
"grad_norm": 82.0984115600586,
"learning_rate": 1.8372627713590178e-06,
"loss": 0.4752083420753479,
"step": 434
},
{
"epoch": 0.3782608695652174,
"grad_norm": 59.69884490966797,
"learning_rate": 1.8340857853413277e-06,
"loss": 1.8268396854400635,
"step": 435
},
{
"epoch": 0.3791304347826087,
"grad_norm": 32.66505813598633,
"learning_rate": 1.830903965054711e-06,
"loss": 0.35361823439598083,
"step": 436
},
{
"epoch": 0.38,
"grad_norm": 44.198997497558594,
"learning_rate": 1.8277173368339546e-06,
"loss": 0.5606723427772522,
"step": 437
},
{
"epoch": 0.3808695652173913,
"grad_norm": 47.64988327026367,
"learning_rate": 1.8245259270536398e-06,
"loss": 0.3799149990081787,
"step": 438
},
{
"epoch": 0.38173913043478264,
"grad_norm": 51.07253646850586,
"learning_rate": 1.8213297621279225e-06,
"loss": 3.3815064430236816,
"step": 439
},
{
"epoch": 0.3826086956521739,
"grad_norm": 93.61296081542969,
"learning_rate": 1.8181288685103148e-06,
"loss": 1.108884334564209,
"step": 440
},
{
"epoch": 0.3834782608695652,
"grad_norm": 53.49833679199219,
"learning_rate": 1.8149232726934687e-06,
"loss": 0.667290449142456,
"step": 441
},
{
"epoch": 0.3843478260869565,
"grad_norm": 119.75688934326172,
"learning_rate": 1.8117130012089518e-06,
"loss": 0.32770204544067383,
"step": 442
},
{
"epoch": 0.38521739130434784,
"grad_norm": 507.785888671875,
"learning_rate": 1.8084980806270324e-06,
"loss": 0.593798041343689,
"step": 443
},
{
"epoch": 0.38608695652173913,
"grad_norm": 76.85552978515625,
"learning_rate": 1.8052785375564569e-06,
"loss": 1.3338149785995483,
"step": 444
},
{
"epoch": 0.3869565217391304,
"grad_norm": 419.838623046875,
"learning_rate": 1.8020543986442306e-06,
"loss": 1.6986769437789917,
"step": 445
},
{
"epoch": 0.38782608695652177,
"grad_norm": 35.90026092529297,
"learning_rate": 1.7988256905753976e-06,
"loss": 0.2494622766971588,
"step": 446
},
{
"epoch": 0.38869565217391305,
"grad_norm": 124.28018951416016,
"learning_rate": 1.7955924400728174e-06,
"loss": 2.188328742980957,
"step": 447
},
{
"epoch": 0.38956521739130434,
"grad_norm": 62.259708404541016,
"learning_rate": 1.7923546738969477e-06,
"loss": 0.7146010398864746,
"step": 448
},
{
"epoch": 0.39043478260869563,
"grad_norm": 39.062965393066406,
"learning_rate": 1.78911241884562e-06,
"loss": 0.2697073817253113,
"step": 449
},
{
"epoch": 0.391304347826087,
"grad_norm": 46.29635238647461,
"learning_rate": 1.7858657017538177e-06,
"loss": 0.7018026113510132,
"step": 450
},
{
"epoch": 0.39217391304347826,
"grad_norm": 126.32167053222656,
"learning_rate": 1.782614549493457e-06,
"loss": 0.32994115352630615,
"step": 451
},
{
"epoch": 0.39304347826086955,
"grad_norm": 162.45706176757812,
"learning_rate": 1.77935898897316e-06,
"loss": 0.5059078931808472,
"step": 452
},
{
"epoch": 0.3939130434782609,
"grad_norm": 31.517831802368164,
"learning_rate": 1.7760990471380364e-06,
"loss": 0.856207549571991,
"step": 453
},
{
"epoch": 0.3947826086956522,
"grad_norm": 103.00402069091797,
"learning_rate": 1.7728347509694566e-06,
"loss": 0.6330320835113525,
"step": 454
},
{
"epoch": 0.39565217391304347,
"grad_norm": 100.9277114868164,
"learning_rate": 1.7695661274848326e-06,
"loss": 1.1151174306869507,
"step": 455
},
{
"epoch": 0.39652173913043476,
"grad_norm": 38.331363677978516,
"learning_rate": 1.7662932037373896e-06,
"loss": 0.2668941617012024,
"step": 456
},
{
"epoch": 0.3973913043478261,
"grad_norm": 211.7262420654297,
"learning_rate": 1.7630160068159463e-06,
"loss": 1.712091326713562,
"step": 457
},
{
"epoch": 0.3982608695652174,
"grad_norm": 35.68760299682617,
"learning_rate": 1.7597345638446877e-06,
"loss": 0.5435832738876343,
"step": 458
},
{
"epoch": 0.3991304347826087,
"grad_norm": 59.29582977294922,
"learning_rate": 1.7564489019829435e-06,
"loss": 0.4060443639755249,
"step": 459
},
{
"epoch": 0.4,
"grad_norm": 100.3779067993164,
"learning_rate": 1.7531590484249595e-06,
"loss": 1.9934556484222412,
"step": 460
},
{
"epoch": 0.4008695652173913,
"grad_norm": 51.10364532470703,
"learning_rate": 1.749865030399676e-06,
"loss": 0.3774723410606384,
"step": 461
},
{
"epoch": 0.4017391304347826,
"grad_norm": 64.10908508300781,
"learning_rate": 1.7465668751705017e-06,
"loss": 1.5249488353729248,
"step": 462
},
{
"epoch": 0.4026086956521739,
"grad_norm": 53.03550720214844,
"learning_rate": 1.7432646100350852e-06,
"loss": 0.35316765308380127,
"step": 463
},
{
"epoch": 0.40347826086956523,
"grad_norm": 74.80030822753906,
"learning_rate": 1.739958262325094e-06,
"loss": 1.9750590324401855,
"step": 464
},
{
"epoch": 0.4043478260869565,
"grad_norm": 51.76995086669922,
"learning_rate": 1.7366478594059834e-06,
"loss": 2.0024967193603516,
"step": 465
},
{
"epoch": 0.4052173913043478,
"grad_norm": 117.95697021484375,
"learning_rate": 1.7333334286767735e-06,
"loss": 1.369372844696045,
"step": 466
},
{
"epoch": 0.40608695652173915,
"grad_norm": 1017.1922607421875,
"learning_rate": 1.7300149975698213e-06,
"loss": 1.206782579421997,
"step": 467
},
{
"epoch": 0.40695652173913044,
"grad_norm": 79.6695785522461,
"learning_rate": 1.7266925935505918e-06,
"loss": 1.1343293190002441,
"step": 468
},
{
"epoch": 0.4078260869565217,
"grad_norm": 120.3302230834961,
"learning_rate": 1.7233662441174348e-06,
"loss": 0.6617238521575928,
"step": 469
},
{
"epoch": 0.40869565217391307,
"grad_norm": 106.52935791015625,
"learning_rate": 1.7200359768013528e-06,
"loss": 0.402765691280365,
"step": 470
},
{
"epoch": 0.40956521739130436,
"grad_norm": 149.852783203125,
"learning_rate": 1.7167018191657764e-06,
"loss": 1.327116847038269,
"step": 471
},
{
"epoch": 0.41043478260869565,
"grad_norm": 137.79925537109375,
"learning_rate": 1.7133637988063349e-06,
"loss": 1.6466482877731323,
"step": 472
},
{
"epoch": 0.41130434782608694,
"grad_norm": 144.7108154296875,
"learning_rate": 1.7100219433506268e-06,
"loss": 0.9789918661117554,
"step": 473
},
{
"epoch": 0.4121739130434783,
"grad_norm": 40.42278289794922,
"learning_rate": 1.7066762804579941e-06,
"loss": 1.3459813594818115,
"step": 474
},
{
"epoch": 0.41304347826086957,
"grad_norm": 72.24211120605469,
"learning_rate": 1.7033268378192902e-06,
"loss": 1.6319739818572998,
"step": 475
},
{
"epoch": 0.41391304347826086,
"grad_norm": 72.87616729736328,
"learning_rate": 1.6999736431566533e-06,
"loss": 0.6043583154678345,
"step": 476
},
{
"epoch": 0.4147826086956522,
"grad_norm": 541.5030517578125,
"learning_rate": 1.6966167242232742e-06,
"loss": 1.5408962965011597,
"step": 477
},
{
"epoch": 0.4156521739130435,
"grad_norm": 59.899375915527344,
"learning_rate": 1.6932561088031695e-06,
"loss": 0.5158736705780029,
"step": 478
},
{
"epoch": 0.4165217391304348,
"grad_norm": 83.0235824584961,
"learning_rate": 1.6898918247109502e-06,
"loss": 0.48279476165771484,
"step": 479
},
{
"epoch": 0.41739130434782606,
"grad_norm": 48.84733200073242,
"learning_rate": 1.6865238997915907e-06,
"loss": 0.2537239193916321,
"step": 480
},
{
"epoch": 0.4182608695652174,
"grad_norm": 122.0486068725586,
"learning_rate": 1.6831523619202e-06,
"loss": 0.5510621666908264,
"step": 481
},
{
"epoch": 0.4191304347826087,
"grad_norm": 67.17446899414062,
"learning_rate": 1.6797772390017897e-06,
"loss": 0.5196292400360107,
"step": 482
},
{
"epoch": 0.42,
"grad_norm": 58.56965637207031,
"learning_rate": 1.6763985589710443e-06,
"loss": 0.33230191469192505,
"step": 483
},
{
"epoch": 0.42086956521739133,
"grad_norm": 234.8605194091797,
"learning_rate": 1.673016349792088e-06,
"loss": 0.7219239473342896,
"step": 484
},
{
"epoch": 0.4217391304347826,
"grad_norm": 66.78538513183594,
"learning_rate": 1.6696306394582562e-06,
"loss": 0.2938106060028076,
"step": 485
},
{
"epoch": 0.4226086956521739,
"grad_norm": 108.08150482177734,
"learning_rate": 1.6662414559918605e-06,
"loss": 0.4802602529525757,
"step": 486
},
{
"epoch": 0.4234782608695652,
"grad_norm": 78.89554595947266,
"learning_rate": 1.6628488274439592e-06,
"loss": 0.4338192939758301,
"step": 487
},
{
"epoch": 0.42434782608695654,
"grad_norm": 59.538387298583984,
"learning_rate": 1.6594527818941242e-06,
"loss": 0.7179808616638184,
"step": 488
},
{
"epoch": 0.4252173913043478,
"grad_norm": 98.61905670166016,
"learning_rate": 1.656053347450208e-06,
"loss": 1.9393256902694702,
"step": 489
},
{
"epoch": 0.4260869565217391,
"grad_norm": 281.14459228515625,
"learning_rate": 1.6526505522481134e-06,
"loss": 1.6723341941833496,
"step": 490
},
{
"epoch": 0.42695652173913046,
"grad_norm": 39.20327377319336,
"learning_rate": 1.6492444244515577e-06,
"loss": 0.23699325323104858,
"step": 491
},
{
"epoch": 0.42782608695652175,
"grad_norm": 71.64807891845703,
"learning_rate": 1.645834992251841e-06,
"loss": 2.458763837814331,
"step": 492
},
{
"epoch": 0.42869565217391303,
"grad_norm": 70.16139221191406,
"learning_rate": 1.6424222838676127e-06,
"loss": 0.8271002769470215,
"step": 493
},
{
"epoch": 0.4295652173913043,
"grad_norm": 160.2700653076172,
"learning_rate": 1.6390063275446387e-06,
"loss": 1.1022875308990479,
"step": 494
},
{
"epoch": 0.43043478260869567,
"grad_norm": 133.14254760742188,
"learning_rate": 1.6355871515555665e-06,
"loss": 0.6330558657646179,
"step": 495
},
{
"epoch": 0.43130434782608695,
"grad_norm": 384.534423828125,
"learning_rate": 1.6321647841996914e-06,
"loss": 1.0990197658538818,
"step": 496
},
{
"epoch": 0.43217391304347824,
"grad_norm": 37.18069839477539,
"learning_rate": 1.6287392538027225e-06,
"loss": 0.3156867027282715,
"step": 497
},
{
"epoch": 0.4330434782608696,
"grad_norm": 112.12581634521484,
"learning_rate": 1.6253105887165488e-06,
"loss": 0.6666814684867859,
"step": 498
},
{
"epoch": 0.4339130434782609,
"grad_norm": 98.95524597167969,
"learning_rate": 1.6218788173190038e-06,
"loss": 0.9090654850006104,
"step": 499
},
{
"epoch": 0.43478260869565216,
"grad_norm": 54.75837707519531,
"learning_rate": 1.6184439680136305e-06,
"loss": 0.5066555738449097,
"step": 500
},
{
"epoch": 0.4356521739130435,
"grad_norm": 186.33668518066406,
"learning_rate": 1.6150060692294467e-06,
"loss": 0.8995145559310913,
"step": 501
},
{
"epoch": 0.4365217391304348,
"grad_norm": 120.6266098022461,
"learning_rate": 1.6115651494207099e-06,
"loss": 2.663893938064575,
"step": 502
},
{
"epoch": 0.4373913043478261,
"grad_norm": 58.315223693847656,
"learning_rate": 1.608121237066681e-06,
"loss": 2.4864797592163086,
"step": 503
},
{
"epoch": 0.43826086956521737,
"grad_norm": 1452.934326171875,
"learning_rate": 1.6046743606713905e-06,
"loss": 2.9407260417938232,
"step": 504
},
{
"epoch": 0.4391304347826087,
"grad_norm": 99.23982238769531,
"learning_rate": 1.6012245487633994e-06,
"loss": 0.7989898920059204,
"step": 505
},
{
"epoch": 0.44,
"grad_norm": 48.1356315612793,
"learning_rate": 1.597771829895566e-06,
"loss": 1.13506281375885,
"step": 506
},
{
"epoch": 0.4408695652173913,
"grad_norm": 124.52139282226562,
"learning_rate": 1.5943162326448086e-06,
"loss": 1.1770362854003906,
"step": 507
},
{
"epoch": 0.44173913043478263,
"grad_norm": 61.51244354248047,
"learning_rate": 1.5908577856118673e-06,
"loss": 0.7881090044975281,
"step": 508
},
{
"epoch": 0.4426086956521739,
"grad_norm": 73.799560546875,
"learning_rate": 1.5873965174210715e-06,
"loss": 1.9321670532226562,
"step": 509
},
{
"epoch": 0.4434782608695652,
"grad_norm": 41.300926208496094,
"learning_rate": 1.5839324567200976e-06,
"loss": 0.30828559398651123,
"step": 510
},
{
"epoch": 0.4443478260869565,
"grad_norm": 52.224727630615234,
"learning_rate": 1.5804656321797362e-06,
"loss": 0.2625771760940552,
"step": 511
},
{
"epoch": 0.44521739130434784,
"grad_norm": 361.1880798339844,
"learning_rate": 1.5769960724936532e-06,
"loss": 0.5133645534515381,
"step": 512
},
{
"epoch": 0.44608695652173913,
"grad_norm": 154.89776611328125,
"learning_rate": 1.573523806378151e-06,
"loss": 1.1001667976379395,
"step": 513
},
{
"epoch": 0.4469565217391304,
"grad_norm": 25.168838500976562,
"learning_rate": 1.570048862571934e-06,
"loss": 0.4305715560913086,
"step": 514
},
{
"epoch": 0.44782608695652176,
"grad_norm": 49.302490234375,
"learning_rate": 1.5665712698358676e-06,
"loss": 0.379476398229599,
"step": 515
},
{
"epoch": 0.44869565217391305,
"grad_norm": 27.884868621826172,
"learning_rate": 1.5630910569527419e-06,
"loss": 0.3568403720855713,
"step": 516
},
{
"epoch": 0.44956521739130434,
"grad_norm": 74.5727767944336,
"learning_rate": 1.5596082527270327e-06,
"loss": 0.3433043956756592,
"step": 517
},
{
"epoch": 0.4504347826086956,
"grad_norm": 98.89971160888672,
"learning_rate": 1.5561228859846636e-06,
"loss": 0.48701316118240356,
"step": 518
},
{
"epoch": 0.45130434782608697,
"grad_norm": 50.903289794921875,
"learning_rate": 1.5526349855727674e-06,
"loss": 0.4016726613044739,
"step": 519
},
{
"epoch": 0.45217391304347826,
"grad_norm": 40.44261169433594,
"learning_rate": 1.5491445803594474e-06,
"loss": 0.37266242504119873,
"step": 520
},
{
"epoch": 0.45304347826086955,
"grad_norm": 78.33443450927734,
"learning_rate": 1.5456516992335375e-06,
"loss": 0.8275619745254517,
"step": 521
},
{
"epoch": 0.4539130434782609,
"grad_norm": 175.40965270996094,
"learning_rate": 1.542156371104364e-06,
"loss": 1.6764061450958252,
"step": 522
},
{
"epoch": 0.4547826086956522,
"grad_norm": 200.28219604492188,
"learning_rate": 1.5386586249015074e-06,
"loss": 0.42932283878326416,
"step": 523
},
{
"epoch": 0.45565217391304347,
"grad_norm": 301.8988952636719,
"learning_rate": 1.5351584895745592e-06,
"loss": 0.5816707611083984,
"step": 524
},
{
"epoch": 0.45652173913043476,
"grad_norm": 236.58929443359375,
"learning_rate": 1.5316559940928877e-06,
"loss": 1.8768939971923828,
"step": 525
},
{
"epoch": 0.4573913043478261,
"grad_norm": 44.522884368896484,
"learning_rate": 1.528151167445393e-06,
"loss": 0.45754727721214294,
"step": 526
},
{
"epoch": 0.4582608695652174,
"grad_norm": 37.146820068359375,
"learning_rate": 1.5246440386402716e-06,
"loss": 0.3753184676170349,
"step": 527
},
{
"epoch": 0.4591304347826087,
"grad_norm": 40.38521194458008,
"learning_rate": 1.5211346367047719e-06,
"loss": 0.8287581205368042,
"step": 528
},
{
"epoch": 0.46,
"grad_norm": 204.2262725830078,
"learning_rate": 1.5176229906849577e-06,
"loss": 1.6905226707458496,
"step": 529
},
{
"epoch": 0.4608695652173913,
"grad_norm": 117.0674819946289,
"learning_rate": 1.514109129645466e-06,
"loss": 0.5957748293876648,
"step": 530
},
{
"epoch": 0.4617391304347826,
"grad_norm": 125.92433166503906,
"learning_rate": 1.5105930826692664e-06,
"loss": 0.9250260591506958,
"step": 531
},
{
"epoch": 0.46260869565217394,
"grad_norm": 69.9134750366211,
"learning_rate": 1.5070748788574208e-06,
"loss": 0.364722341299057,
"step": 532
},
{
"epoch": 0.46347826086956523,
"grad_norm": 151.97052001953125,
"learning_rate": 1.5035545473288426e-06,
"loss": 0.8764115571975708,
"step": 533
},
{
"epoch": 0.4643478260869565,
"grad_norm": 36.5955924987793,
"learning_rate": 1.5000321172200557e-06,
"loss": 0.29237812757492065,
"step": 534
},
{
"epoch": 0.4652173913043478,
"grad_norm": 44.44923782348633,
"learning_rate": 1.4965076176849528e-06,
"loss": 0.25572437047958374,
"step": 535
},
{
"epoch": 0.46608695652173915,
"grad_norm": 58.24061584472656,
"learning_rate": 1.4929810778945545e-06,
"loss": 3.842434883117676,
"step": 536
},
{
"epoch": 0.46695652173913044,
"grad_norm": 79.52149963378906,
"learning_rate": 1.4894525270367682e-06,
"loss": 0.7761034965515137,
"step": 537
},
{
"epoch": 0.4678260869565217,
"grad_norm": 85.65235900878906,
"learning_rate": 1.485921994316146e-06,
"loss": 1.0048233270645142,
"step": 538
},
{
"epoch": 0.46869565217391307,
"grad_norm": 147.44503784179688,
"learning_rate": 1.4823895089536435e-06,
"loss": 1.2945889234542847,
"step": 539
},
{
"epoch": 0.46956521739130436,
"grad_norm": 77.540771484375,
"learning_rate": 1.4788551001863768e-06,
"loss": 1.0739409923553467,
"step": 540
},
{
"epoch": 0.47043478260869565,
"grad_norm": 19.513778686523438,
"learning_rate": 1.4753187972673824e-06,
"loss": 0.13495752215385437,
"step": 541
},
{
"epoch": 0.47130434782608693,
"grad_norm": 559.4796142578125,
"learning_rate": 1.471780629465373e-06,
"loss": 2.378528118133545,
"step": 542
},
{
"epoch": 0.4721739130434783,
"grad_norm": 138.12265014648438,
"learning_rate": 1.468240626064497e-06,
"loss": 0.45228928327560425,
"step": 543
},
{
"epoch": 0.47304347826086957,
"grad_norm": 37.85926055908203,
"learning_rate": 1.4646988163640955e-06,
"loss": 0.41717642545700073,
"step": 544
},
{
"epoch": 0.47391304347826085,
"grad_norm": 812.1907958984375,
"learning_rate": 1.4611552296784581e-06,
"loss": 0.6520176529884338,
"step": 545
},
{
"epoch": 0.4747826086956522,
"grad_norm": 47.911109924316406,
"learning_rate": 1.4576098953365845e-06,
"loss": 0.28447192907333374,
"step": 546
},
{
"epoch": 0.4756521739130435,
"grad_norm": 198.42416381835938,
"learning_rate": 1.454062842681937e-06,
"loss": 1.3370239734649658,
"step": 547
},
{
"epoch": 0.4765217391304348,
"grad_norm": 866.588623046875,
"learning_rate": 1.4505141010722004e-06,
"loss": 0.7387341856956482,
"step": 548
},
{
"epoch": 0.47739130434782606,
"grad_norm": 63.434783935546875,
"learning_rate": 1.446963699879039e-06,
"loss": 0.347465455532074,
"step": 549
},
{
"epoch": 0.4782608695652174,
"grad_norm": 159.80320739746094,
"learning_rate": 1.4434116684878516e-06,
"loss": 1.3235949277877808,
"step": 550
},
{
"epoch": 0.4791304347826087,
"grad_norm": 109.88274383544922,
"learning_rate": 1.4398580362975311e-06,
"loss": 1.1254695653915405,
"step": 551
},
{
"epoch": 0.48,
"grad_norm": 37.5253791809082,
"learning_rate": 1.436302832720218e-06,
"loss": 0.21883612871170044,
"step": 552
},
{
"epoch": 0.4808695652173913,
"grad_norm": 75.48155212402344,
"learning_rate": 1.4327460871810606e-06,
"loss": 0.28479528427124023,
"step": 553
},
{
"epoch": 0.4817391304347826,
"grad_norm": 886.3295288085938,
"learning_rate": 1.429187829117967e-06,
"loss": 1.2821072340011597,
"step": 554
},
{
"epoch": 0.4826086956521739,
"grad_norm": 217.08229064941406,
"learning_rate": 1.4256280879813661e-06,
"loss": 0.937677264213562,
"step": 555
},
{
"epoch": 0.4834782608695652,
"grad_norm": 103.811279296875,
"learning_rate": 1.4220668932339606e-06,
"loss": 0.334438294172287,
"step": 556
},
{
"epoch": 0.48434782608695653,
"grad_norm": 133.6400909423828,
"learning_rate": 1.4185042743504843e-06,
"loss": 0.8252410888671875,
"step": 557
},
{
"epoch": 0.4852173913043478,
"grad_norm": 64.17788696289062,
"learning_rate": 1.4149402608174583e-06,
"loss": 0.6373386383056641,
"step": 558
},
{
"epoch": 0.4860869565217391,
"grad_norm": 115.58035278320312,
"learning_rate": 1.4113748821329462e-06,
"loss": 0.42149341106414795,
"step": 559
},
{
"epoch": 0.48695652173913045,
"grad_norm": 54.42445373535156,
"learning_rate": 1.4078081678063118e-06,
"loss": 1.273005723953247,
"step": 560
},
{
"epoch": 0.48782608695652174,
"grad_norm": 18.693782806396484,
"learning_rate": 1.4042401473579717e-06,
"loss": 0.0848182663321495,
"step": 561
},
{
"epoch": 0.48869565217391303,
"grad_norm": 139.10169982910156,
"learning_rate": 1.4006708503191539e-06,
"loss": 2.338060140609741,
"step": 562
},
{
"epoch": 0.4895652173913044,
"grad_norm": 260.83331298828125,
"learning_rate": 1.3971003062316527e-06,
"loss": 0.7347593307495117,
"step": 563
},
{
"epoch": 0.49043478260869566,
"grad_norm": 33.69926071166992,
"learning_rate": 1.3935285446475822e-06,
"loss": 0.2858821749687195,
"step": 564
},
{
"epoch": 0.49130434782608695,
"grad_norm": 38.54191589355469,
"learning_rate": 1.3899555951291348e-06,
"loss": 0.42253923416137695,
"step": 565
},
{
"epoch": 0.49217391304347824,
"grad_norm": 118.90116119384766,
"learning_rate": 1.3863814872483347e-06,
"loss": 2.127577543258667,
"step": 566
},
{
"epoch": 0.4930434782608696,
"grad_norm": 25.04210662841797,
"learning_rate": 1.3828062505867928e-06,
"loss": 0.14584410190582275,
"step": 567
},
{
"epoch": 0.49391304347826087,
"grad_norm": 89.78531646728516,
"learning_rate": 1.3792299147354626e-06,
"loss": 1.3909008502960205,
"step": 568
},
{
"epoch": 0.49478260869565216,
"grad_norm": 139.40478515625,
"learning_rate": 1.3756525092943965e-06,
"loss": 1.3039909601211548,
"step": 569
},
{
"epoch": 0.4956521739130435,
"grad_norm": 55.25202941894531,
"learning_rate": 1.3720740638724977e-06,
"loss": 0.2234216183423996,
"step": 570
},
{
"epoch": 0.4965217391304348,
"grad_norm": 107.28520965576172,
"learning_rate": 1.3684946080872779e-06,
"loss": 0.6456013321876526,
"step": 571
},
{
"epoch": 0.4973913043478261,
"grad_norm": 144.71206665039062,
"learning_rate": 1.3649141715646117e-06,
"loss": 2.418483018875122,
"step": 572
},
{
"epoch": 0.49826086956521737,
"grad_norm": 52.84182357788086,
"learning_rate": 1.361332783938489e-06,
"loss": 0.2772313356399536,
"step": 573
},
{
"epoch": 0.4991304347826087,
"grad_norm": 49.391849517822266,
"learning_rate": 1.3577504748507744e-06,
"loss": 0.43491166830062866,
"step": 574
},
{
"epoch": 0.5,
"grad_norm": 57.58726119995117,
"learning_rate": 1.3541672739509568e-06,
"loss": 0.5723316669464111,
"step": 575
},
{
"epoch": 0.5008695652173913,
"grad_norm": 58.86819839477539,
"learning_rate": 1.3505832108959077e-06,
"loss": 0.41305217146873474,
"step": 576
},
{
"epoch": 0.5017391304347826,
"grad_norm": 71.6548843383789,
"learning_rate": 1.346998315349633e-06,
"loss": 0.4881059527397156,
"step": 577
},
{
"epoch": 0.5026086956521739,
"grad_norm": 46.89256286621094,
"learning_rate": 1.3434126169830306e-06,
"loss": 0.2294202446937561,
"step": 578
},
{
"epoch": 0.5034782608695653,
"grad_norm": 71.92279815673828,
"learning_rate": 1.3398261454736413e-06,
"loss": 0.49881815910339355,
"step": 579
},
{
"epoch": 0.5043478260869565,
"grad_norm": 86.85327911376953,
"learning_rate": 1.3362389305054061e-06,
"loss": 0.8657968044281006,
"step": 580
},
{
"epoch": 0.5052173913043478,
"grad_norm": 103.21882629394531,
"learning_rate": 1.3326510017684194e-06,
"loss": 0.3297439515590668,
"step": 581
},
{
"epoch": 0.5060869565217392,
"grad_norm": 264.4145812988281,
"learning_rate": 1.3290623889586817e-06,
"loss": 0.5024732351303101,
"step": 582
},
{
"epoch": 0.5069565217391304,
"grad_norm": 283.6190185546875,
"learning_rate": 1.3254731217778577e-06,
"loss": 1.4738558530807495,
"step": 583
},
{
"epoch": 0.5078260869565218,
"grad_norm": 87.96891021728516,
"learning_rate": 1.3218832299330259e-06,
"loss": 0.2829592227935791,
"step": 584
},
{
"epoch": 0.508695652173913,
"grad_norm": 54.880271911621094,
"learning_rate": 1.3182927431364358e-06,
"loss": 0.5073628425598145,
"step": 585
},
{
"epoch": 0.5095652173913043,
"grad_norm": 30.446800231933594,
"learning_rate": 1.314701691105262e-06,
"loss": 2.0656630992889404,
"step": 586
},
{
"epoch": 0.5104347826086957,
"grad_norm": 122.26018524169922,
"learning_rate": 1.3111101035613551e-06,
"loss": 0.21750657260417938,
"step": 587
},
{
"epoch": 0.5113043478260869,
"grad_norm": 512.86962890625,
"learning_rate": 1.3075180102310008e-06,
"loss": 1.0234020948410034,
"step": 588
},
{
"epoch": 0.5121739130434783,
"grad_norm": 116.8537826538086,
"learning_rate": 1.3039254408446683e-06,
"loss": 0.21366870403289795,
"step": 589
},
{
"epoch": 0.5130434782608696,
"grad_norm": 49.10830307006836,
"learning_rate": 1.3003324251367692e-06,
"loss": 0.4299643635749817,
"step": 590
},
{
"epoch": 0.5139130434782608,
"grad_norm": 182.322265625,
"learning_rate": 1.2967389928454075e-06,
"loss": 1.6993330717086792,
"step": 591
},
{
"epoch": 0.5147826086956522,
"grad_norm": 609.0872802734375,
"learning_rate": 1.2931451737121353e-06,
"loss": 2.356942653656006,
"step": 592
},
{
"epoch": 0.5156521739130435,
"grad_norm": 126.05340576171875,
"learning_rate": 1.289550997481708e-06,
"loss": 0.6911107301712036,
"step": 593
},
{
"epoch": 0.5165217391304348,
"grad_norm": 236.08570861816406,
"learning_rate": 1.2859564939018344e-06,
"loss": 0.6245572566986084,
"step": 594
},
{
"epoch": 0.5173913043478261,
"grad_norm": 65.2644271850586,
"learning_rate": 1.2823616927229345e-06,
"loss": 0.532337486743927,
"step": 595
},
{
"epoch": 0.5182608695652174,
"grad_norm": 602.57080078125,
"learning_rate": 1.2787666236978902e-06,
"loss": 2.266624927520752,
"step": 596
},
{
"epoch": 0.5191304347826087,
"grad_norm": 21.916934967041016,
"learning_rate": 1.2751713165818005e-06,
"loss": 0.23428139090538025,
"step": 597
},
{
"epoch": 0.52,
"grad_norm": 92.19395446777344,
"learning_rate": 1.2715758011317359e-06,
"loss": 0.9039852023124695,
"step": 598
},
{
"epoch": 0.5208695652173913,
"grad_norm": 123.28501892089844,
"learning_rate": 1.2679801071064895e-06,
"loss": 0.8925650119781494,
"step": 599
},
{
"epoch": 0.5217391304347826,
"grad_norm": 250.67852783203125,
"learning_rate": 1.2643842642663345e-06,
"loss": 3.8089053630828857,
"step": 600
},
{
"epoch": 0.5226086956521739,
"grad_norm": 72.26994323730469,
"learning_rate": 1.2607883023727737e-06,
"loss": 0.532624363899231,
"step": 601
},
{
"epoch": 0.5234782608695652,
"grad_norm": 133.5413360595703,
"learning_rate": 1.2571922511882973e-06,
"loss": 1.2747904062271118,
"step": 602
},
{
"epoch": 0.5243478260869565,
"grad_norm": 67.44408416748047,
"learning_rate": 1.2535961404761322e-06,
"loss": 0.37231168150901794,
"step": 603
},
{
"epoch": 0.5252173913043479,
"grad_norm": 81.25239562988281,
"learning_rate": 1.25e-06,
"loss": 2.603219985961914,
"step": 604
},
{
"epoch": 0.5260869565217391,
"grad_norm": 102.40206146240234,
"learning_rate": 1.2464038595238682e-06,
"loss": 1.3927900791168213,
"step": 605
},
{
"epoch": 0.5269565217391304,
"grad_norm": 56.06349563598633,
"learning_rate": 1.2428077488117033e-06,
"loss": 0.1289176046848297,
"step": 606
},
{
"epoch": 0.5278260869565218,
"grad_norm": 38.40553665161133,
"learning_rate": 1.2392116976272265e-06,
"loss": 0.22410118579864502,
"step": 607
},
{
"epoch": 0.528695652173913,
"grad_norm": 102.11624145507812,
"learning_rate": 1.2356157357336657e-06,
"loss": 0.35848140716552734,
"step": 608
},
{
"epoch": 0.5295652173913044,
"grad_norm": 138.64425659179688,
"learning_rate": 1.2320198928935109e-06,
"loss": 0.9172069430351257,
"step": 609
},
{
"epoch": 0.5304347826086957,
"grad_norm": 41.466331481933594,
"learning_rate": 1.2284241988682645e-06,
"loss": 0.3112112879753113,
"step": 610
},
{
"epoch": 0.5313043478260869,
"grad_norm": 61.450721740722656,
"learning_rate": 1.2248286834182e-06,
"loss": 0.5715629458427429,
"step": 611
},
{
"epoch": 0.5321739130434783,
"grad_norm": 79.89807891845703,
"learning_rate": 1.2212333763021104e-06,
"loss": 1.294171929359436,
"step": 612
},
{
"epoch": 0.5330434782608696,
"grad_norm": 72.71001434326172,
"learning_rate": 1.2176383072770661e-06,
"loss": 0.518186092376709,
"step": 613
},
{
"epoch": 0.5339130434782609,
"grad_norm": 137.88656616210938,
"learning_rate": 1.2140435060981662e-06,
"loss": 0.35097479820251465,
"step": 614
},
{
"epoch": 0.5347826086956522,
"grad_norm": 67.36344909667969,
"learning_rate": 1.2104490025182927e-06,
"loss": 1.6497039794921875,
"step": 615
},
{
"epoch": 0.5356521739130434,
"grad_norm": 26.461103439331055,
"learning_rate": 1.2068548262878653e-06,
"loss": 0.3913857638835907,
"step": 616
},
{
"epoch": 0.5365217391304348,
"grad_norm": 70.71717834472656,
"learning_rate": 1.2032610071545931e-06,
"loss": 0.38299569487571716,
"step": 617
},
{
"epoch": 0.5373913043478261,
"grad_norm": 31.943819046020508,
"learning_rate": 1.1996675748632314e-06,
"loss": 0.3052636384963989,
"step": 618
},
{
"epoch": 0.5382608695652173,
"grad_norm": 70.05712127685547,
"learning_rate": 1.1960745591553319e-06,
"loss": 0.3777497708797455,
"step": 619
},
{
"epoch": 0.5391304347826087,
"grad_norm": 75.02300262451172,
"learning_rate": 1.1924819897689998e-06,
"loss": 0.2484484165906906,
"step": 620
},
{
"epoch": 0.54,
"grad_norm": 158.2064971923828,
"learning_rate": 1.1888898964386453e-06,
"loss": 2.01037859916687,
"step": 621
},
{
"epoch": 0.5408695652173913,
"grad_norm": 97.98067474365234,
"learning_rate": 1.1852983088947382e-06,
"loss": 0.857545018196106,
"step": 622
},
{
"epoch": 0.5417391304347826,
"grad_norm": 66.65267181396484,
"learning_rate": 1.181707256863564e-06,
"loss": 0.4367102086544037,
"step": 623
},
{
"epoch": 0.542608695652174,
"grad_norm": 115.09444427490234,
"learning_rate": 1.1781167700669743e-06,
"loss": 0.7836287021636963,
"step": 624
},
{
"epoch": 0.5434782608695652,
"grad_norm": 45.53085708618164,
"learning_rate": 1.1745268782221425e-06,
"loss": 0.19425027072429657,
"step": 625
},
{
"epoch": 0.5443478260869565,
"grad_norm": 220.548583984375,
"learning_rate": 1.1709376110413183e-06,
"loss": 0.7339515089988708,
"step": 626
},
{
"epoch": 0.5452173913043479,
"grad_norm": 51.19890213012695,
"learning_rate": 1.1673489982315808e-06,
"loss": 1.0075253248214722,
"step": 627
},
{
"epoch": 0.5460869565217391,
"grad_norm": 88.91786193847656,
"learning_rate": 1.1637610694945939e-06,
"loss": 0.270574688911438,
"step": 628
},
{
"epoch": 0.5469565217391305,
"grad_norm": 52.1123161315918,
"learning_rate": 1.160173854526359e-06,
"loss": 0.6707390546798706,
"step": 629
},
{
"epoch": 0.5478260869565217,
"grad_norm": 28.796892166137695,
"learning_rate": 1.1565873830169698e-06,
"loss": 0.9050424098968506,
"step": 630
},
{
"epoch": 0.548695652173913,
"grad_norm": 25.490886688232422,
"learning_rate": 1.1530016846503673e-06,
"loss": 0.17329643666744232,
"step": 631
},
{
"epoch": 0.5495652173913044,
"grad_norm": 47.696041107177734,
"learning_rate": 1.1494167891040927e-06,
"loss": 0.2002371996641159,
"step": 632
},
{
"epoch": 0.5504347826086956,
"grad_norm": 62.62557601928711,
"learning_rate": 1.1458327260490434e-06,
"loss": 0.37586089968681335,
"step": 633
},
{
"epoch": 0.551304347826087,
"grad_norm": 57.04502487182617,
"learning_rate": 1.1422495251492258e-06,
"loss": 0.3957309126853943,
"step": 634
},
{
"epoch": 0.5521739130434783,
"grad_norm": 27.15058708190918,
"learning_rate": 1.1386672160615112e-06,
"loss": 0.13363079726696014,
"step": 635
},
{
"epoch": 0.5530434782608695,
"grad_norm": 43.407196044921875,
"learning_rate": 1.135085828435389e-06,
"loss": 0.8801892995834351,
"step": 636
},
{
"epoch": 0.5539130434782609,
"grad_norm": 76.85043334960938,
"learning_rate": 1.1315053919127223e-06,
"loss": 0.47352680563926697,
"step": 637
},
{
"epoch": 0.5547826086956522,
"grad_norm": 2558.25390625,
"learning_rate": 1.1279259361275025e-06,
"loss": 2.9988725185394287,
"step": 638
},
{
"epoch": 0.5556521739130434,
"grad_norm": 28.422143936157227,
"learning_rate": 1.124347490705604e-06,
"loss": 0.1473201960325241,
"step": 639
},
{
"epoch": 0.5565217391304348,
"grad_norm": 29.4410343170166,
"learning_rate": 1.1207700852645376e-06,
"loss": 0.13121181726455688,
"step": 640
},
{
"epoch": 0.5573913043478261,
"grad_norm": 207.7372283935547,
"learning_rate": 1.1171937494132076e-06,
"loss": 2.1873674392700195,
"step": 641
},
{
"epoch": 0.5582608695652174,
"grad_norm": 50.66484069824219,
"learning_rate": 1.1136185127516658e-06,
"loss": 0.4882809519767761,
"step": 642
},
{
"epoch": 0.5591304347826087,
"grad_norm": 43.21005630493164,
"learning_rate": 1.1100444048708654e-06,
"loss": 0.30610162019729614,
"step": 643
},
{
"epoch": 0.56,
"grad_norm": 33.58243179321289,
"learning_rate": 1.1064714553524182e-06,
"loss": 0.36231523752212524,
"step": 644
},
{
"epoch": 0.5608695652173913,
"grad_norm": 27.419002532958984,
"learning_rate": 1.1028996937683477e-06,
"loss": 0.1604674905538559,
"step": 645
},
{
"epoch": 0.5617391304347826,
"grad_norm": 61.23972702026367,
"learning_rate": 1.0993291496808463e-06,
"loss": 0.6611549258232117,
"step": 646
},
{
"epoch": 0.5626086956521739,
"grad_norm": 32.585357666015625,
"learning_rate": 1.0957598526420287e-06,
"loss": 0.18001686036586761,
"step": 647
},
{
"epoch": 0.5634782608695652,
"grad_norm": 121.16118621826172,
"learning_rate": 1.0921918321936888e-06,
"loss": 0.638570249080658,
"step": 648
},
{
"epoch": 0.5643478260869565,
"grad_norm": 23.482879638671875,
"learning_rate": 1.0886251178670543e-06,
"loss": 0.552463948726654,
"step": 649
},
{
"epoch": 0.5652173913043478,
"grad_norm": 106.4288558959961,
"learning_rate": 1.0850597391825423e-06,
"loss": 0.3047104477882385,
"step": 650
},
{
"epoch": 0.5660869565217391,
"grad_norm": 25.377906799316406,
"learning_rate": 1.0814957256495165e-06,
"loss": 0.3209930658340454,
"step": 651
},
{
"epoch": 0.5669565217391305,
"grad_norm": 55.308128356933594,
"learning_rate": 1.07793310676604e-06,
"loss": 0.33284878730773926,
"step": 652
},
{
"epoch": 0.5678260869565217,
"grad_norm": 80.74183654785156,
"learning_rate": 1.0743719120186345e-06,
"loss": 0.35490426421165466,
"step": 653
},
{
"epoch": 0.568695652173913,
"grad_norm": 63.32366943359375,
"learning_rate": 1.0708121708820335e-06,
"loss": 1.6498006582260132,
"step": 654
},
{
"epoch": 0.5695652173913044,
"grad_norm": 102.98860931396484,
"learning_rate": 1.0672539128189403e-06,
"loss": 0.24597136676311493,
"step": 655
},
{
"epoch": 0.5704347826086956,
"grad_norm": 70.00321197509766,
"learning_rate": 1.0636971672797818e-06,
"loss": 0.8502458930015564,
"step": 656
},
{
"epoch": 0.571304347826087,
"grad_norm": 89.56877899169922,
"learning_rate": 1.060141963702469e-06,
"loss": 0.5706131458282471,
"step": 657
},
{
"epoch": 0.5721739130434783,
"grad_norm": 17.826444625854492,
"learning_rate": 1.0565883315121484e-06,
"loss": 0.12732738256454468,
"step": 658
},
{
"epoch": 0.5730434782608695,
"grad_norm": 346.88287353515625,
"learning_rate": 1.0530363001209615e-06,
"loss": 3.1927261352539062,
"step": 659
},
{
"epoch": 0.5739130434782609,
"grad_norm": 50.20029830932617,
"learning_rate": 1.0494858989277996e-06,
"loss": 0.23988930881023407,
"step": 660
},
{
"epoch": 0.5747826086956521,
"grad_norm": 28.96661376953125,
"learning_rate": 1.0459371573180634e-06,
"loss": 0.20075128972530365,
"step": 661
},
{
"epoch": 0.5756521739130435,
"grad_norm": 29.971525192260742,
"learning_rate": 1.0423901046634155e-06,
"loss": 0.284665048122406,
"step": 662
},
{
"epoch": 0.5765217391304348,
"grad_norm": 189.16937255859375,
"learning_rate": 1.038844770321542e-06,
"loss": 0.3455438017845154,
"step": 663
},
{
"epoch": 0.577391304347826,
"grad_norm": 118.93046569824219,
"learning_rate": 1.035301183635905e-06,
"loss": 0.6655479669570923,
"step": 664
},
{
"epoch": 0.5782608695652174,
"grad_norm": 97.87029266357422,
"learning_rate": 1.0317593739355031e-06,
"loss": 0.32840508222579956,
"step": 665
},
{
"epoch": 0.5791304347826087,
"grad_norm": 22.744279861450195,
"learning_rate": 1.0282193705346272e-06,
"loss": 0.10876381397247314,
"step": 666
},
{
"epoch": 0.58,
"grad_norm": 32.968849182128906,
"learning_rate": 1.024681202732618e-06,
"loss": 0.18748174607753754,
"step": 667
},
{
"epoch": 0.5808695652173913,
"grad_norm": 115.26560974121094,
"learning_rate": 1.0211448998136234e-06,
"loss": 1.8718297481536865,
"step": 668
},
{
"epoch": 0.5817391304347826,
"grad_norm": 40.707210540771484,
"learning_rate": 1.017610491046357e-06,
"loss": 0.14855729043483734,
"step": 669
},
{
"epoch": 0.5826086956521739,
"grad_norm": 472.1355285644531,
"learning_rate": 1.0140780056838543e-06,
"loss": 2.8436179161071777,
"step": 670
},
{
"epoch": 0.5834782608695652,
"grad_norm": 42.647972106933594,
"learning_rate": 1.010547472963232e-06,
"loss": 0.3507687449455261,
"step": 671
},
{
"epoch": 0.5843478260869566,
"grad_norm": 38.82874298095703,
"learning_rate": 1.007018922105446e-06,
"loss": 0.32309889793395996,
"step": 672
},
{
"epoch": 0.5852173913043478,
"grad_norm": 47.9738655090332,
"learning_rate": 1.0034923823150477e-06,
"loss": 2.2989041805267334,
"step": 673
},
{
"epoch": 0.5860869565217391,
"grad_norm": 41.57319641113281,
"learning_rate": 9.999678827799447e-07,
"loss": 0.24739569425582886,
"step": 674
},
{
"epoch": 0.5869565217391305,
"grad_norm": 47.57366180419922,
"learning_rate": 9.964454526711576e-07,
"loss": 0.3390384912490845,
"step": 675
},
{
"epoch": 0.5878260869565217,
"grad_norm": 66.8327407836914,
"learning_rate": 9.929251211425794e-07,
"loss": 0.39729148149490356,
"step": 676
},
{
"epoch": 0.5886956521739131,
"grad_norm": 99.55630493164062,
"learning_rate": 9.89406917330734e-07,
"loss": 0.1376858949661255,
"step": 677
},
{
"epoch": 0.5895652173913043,
"grad_norm": 175.07766723632812,
"learning_rate": 9.858908703545342e-07,
"loss": 0.899673342704773,
"step": 678
},
{
"epoch": 0.5904347826086956,
"grad_norm": 90.50199127197266,
"learning_rate": 9.823770093150425e-07,
"loss": 0.7324085235595703,
"step": 679
},
{
"epoch": 0.591304347826087,
"grad_norm": 86.7284927368164,
"learning_rate": 9.788653632952283e-07,
"loss": 0.42626532912254333,
"step": 680
},
{
"epoch": 0.5921739130434782,
"grad_norm": 29.3996639251709,
"learning_rate": 9.75355961359729e-07,
"loss": 0.13454458117485046,
"step": 681
},
{
"epoch": 0.5930434782608696,
"grad_norm": 47.4012451171875,
"learning_rate": 9.718488325546072e-07,
"loss": 1.6617497205734253,
"step": 682
},
{
"epoch": 0.5939130434782609,
"grad_norm": 437.3231506347656,
"learning_rate": 9.68344005907113e-07,
"loss": 1.1107021570205688,
"step": 683
},
{
"epoch": 0.5947826086956521,
"grad_norm": 32.076087951660156,
"learning_rate": 9.648415104254414e-07,
"loss": 0.15177524089813232,
"step": 684
},
{
"epoch": 0.5956521739130435,
"grad_norm": 220.3333282470703,
"learning_rate": 9.613413750984935e-07,
"loss": 1.4707931280136108,
"step": 685
},
{
"epoch": 0.5965217391304348,
"grad_norm": 1715.71728515625,
"learning_rate": 9.578436288956366e-07,
"loss": 2.5795419216156006,
"step": 686
},
{
"epoch": 0.5973913043478261,
"grad_norm": 44.38836669921875,
"learning_rate": 9.543483007664632e-07,
"loss": 0.749055027961731,
"step": 687
},
{
"epoch": 0.5982608695652174,
"grad_norm": 320.1933288574219,
"learning_rate": 9.508554196405532e-07,
"loss": 1.3178150653839111,
"step": 688
},
{
"epoch": 0.5991304347826087,
"grad_norm": 94.76158905029297,
"learning_rate": 9.473650144272328e-07,
"loss": 0.2434827983379364,
"step": 689
},
{
"epoch": 0.6,
"grad_norm": 99.07818603515625,
"learning_rate": 9.438771140153365e-07,
"loss": 2.946225166320801,
"step": 690
},
{
"epoch": 0.6008695652173913,
"grad_norm": 69.13819885253906,
"learning_rate": 9.403917472729676e-07,
"loss": 0.8060781359672546,
"step": 691
},
{
"epoch": 0.6017391304347826,
"grad_norm": 45.17094039916992,
"learning_rate": 9.369089430472583e-07,
"loss": 0.25677287578582764,
"step": 692
},
{
"epoch": 0.6026086956521739,
"grad_norm": 67.40509033203125,
"learning_rate": 9.334287301641326e-07,
"loss": 0.895323634147644,
"step": 693
},
{
"epoch": 0.6034782608695652,
"grad_norm": 139.18455505371094,
"learning_rate": 9.299511374280662e-07,
"loss": 3.637253522872925,
"step": 694
},
{
"epoch": 0.6043478260869565,
"grad_norm": 29.99213981628418,
"learning_rate": 9.26476193621849e-07,
"loss": 0.9718486070632935,
"step": 695
},
{
"epoch": 0.6052173913043478,
"grad_norm": 168.57510375976562,
"learning_rate": 9.230039275063473e-07,
"loss": 0.6895861029624939,
"step": 696
},
{
"epoch": 0.6060869565217392,
"grad_norm": 30.356876373291016,
"learning_rate": 9.195343678202639e-07,
"loss": 0.3929281234741211,
"step": 697
},
{
"epoch": 0.6069565217391304,
"grad_norm": 100.34367370605469,
"learning_rate": 9.160675432799028e-07,
"loss": 1.1606305837631226,
"step": 698
},
{
"epoch": 0.6078260869565217,
"grad_norm": 126.29876708984375,
"learning_rate": 9.126034825789289e-07,
"loss": 0.7983366847038269,
"step": 699
},
{
"epoch": 0.6086956521739131,
"grad_norm": 25.401891708374023,
"learning_rate": 9.091422143881329e-07,
"loss": 1.1823790073394775,
"step": 700
},
{
"epoch": 0.6095652173913043,
"grad_norm": 37.92864227294922,
"learning_rate": 9.056837673551919e-07,
"loss": 1.6165263652801514,
"step": 701
},
{
"epoch": 0.6104347826086957,
"grad_norm": 118.26660919189453,
"learning_rate": 9.022281701044343e-07,
"loss": 1.1245553493499756,
"step": 702
},
{
"epoch": 0.611304347826087,
"grad_norm": 59.895263671875,
"learning_rate": 8.98775451236601e-07,
"loss": 0.46273553371429443,
"step": 703
},
{
"epoch": 0.6121739130434782,
"grad_norm": 35.208133697509766,
"learning_rate": 8.953256393286099e-07,
"loss": 0.2643696069717407,
"step": 704
},
{
"epoch": 0.6130434782608696,
"grad_norm": 535.8164672851562,
"learning_rate": 8.918787629333192e-07,
"loss": 1.2478746175765991,
"step": 705
},
{
"epoch": 0.6139130434782609,
"grad_norm": 44.985755920410156,
"learning_rate": 8.884348505792905e-07,
"loss": 0.19614474475383759,
"step": 706
},
{
"epoch": 0.6147826086956522,
"grad_norm": 104.74789428710938,
"learning_rate": 8.849939307705538e-07,
"loss": 0.4358571469783783,
"step": 707
},
{
"epoch": 0.6156521739130435,
"grad_norm": 112.15499877929688,
"learning_rate": 8.815560319863698e-07,
"loss": 1.2568089962005615,
"step": 708
},
{
"epoch": 0.6165217391304347,
"grad_norm": 28.575714111328125,
"learning_rate": 8.781211826809966e-07,
"loss": 0.19895046949386597,
"step": 709
},
{
"epoch": 0.6173913043478261,
"grad_norm": 212.95989990234375,
"learning_rate": 8.746894112834514e-07,
"loss": 0.42774924635887146,
"step": 710
},
{
"epoch": 0.6182608695652174,
"grad_norm": 141.77655029296875,
"learning_rate": 8.712607461972777e-07,
"loss": 0.4444720447063446,
"step": 711
},
{
"epoch": 0.6191304347826087,
"grad_norm": 151.486328125,
"learning_rate": 8.678352158003093e-07,
"loss": 0.4183366894721985,
"step": 712
},
{
"epoch": 0.62,
"grad_norm": 212.81463623046875,
"learning_rate": 8.644128484444339e-07,
"loss": 0.467068612575531,
"step": 713
},
{
"epoch": 0.6208695652173913,
"grad_norm": 146.66627502441406,
"learning_rate": 8.609936724553616e-07,
"loss": 1.342170238494873,
"step": 714
},
{
"epoch": 0.6217391304347826,
"grad_norm": 59.82827377319336,
"learning_rate": 8.575777161323876e-07,
"loss": 0.2382092922925949,
"step": 715
},
{
"epoch": 0.6226086956521739,
"grad_norm": 94.89098358154297,
"learning_rate": 8.541650077481598e-07,
"loss": 1.388150691986084,
"step": 716
},
{
"epoch": 0.6234782608695653,
"grad_norm": 131.10488891601562,
"learning_rate": 8.507555755484429e-07,
"loss": 0.5265352725982666,
"step": 717
},
{
"epoch": 0.6243478260869565,
"grad_norm": 24.01419448852539,
"learning_rate": 8.47349447751887e-07,
"loss": 0.129073366522789,
"step": 718
},
{
"epoch": 0.6252173913043478,
"grad_norm": 60.04438781738281,
"learning_rate": 8.439466525497925e-07,
"loss": 0.2663729190826416,
"step": 719
},
{
"epoch": 0.6260869565217392,
"grad_norm": 75.26876831054688,
"learning_rate": 8.405472181058766e-07,
"loss": 0.7231850028038025,
"step": 720
},
{
"epoch": 0.6269565217391304,
"grad_norm": 50.57442092895508,
"learning_rate": 8.371511725560416e-07,
"loss": 0.44643306732177734,
"step": 721
},
{
"epoch": 0.6278260869565218,
"grad_norm": 45.66079330444336,
"learning_rate": 8.3375854400814e-07,
"loss": 0.21311675012111664,
"step": 722
},
{
"epoch": 0.628695652173913,
"grad_norm": 125.3226547241211,
"learning_rate": 8.303693605417444e-07,
"loss": 0.28283005952835083,
"step": 723
},
{
"epoch": 0.6295652173913043,
"grad_norm": 311.9544982910156,
"learning_rate": 8.26983650207912e-07,
"loss": 2.3147706985473633,
"step": 724
},
{
"epoch": 0.6304347826086957,
"grad_norm": 40.3516960144043,
"learning_rate": 8.236014410289558e-07,
"loss": 0.2545880079269409,
"step": 725
},
{
"epoch": 0.6313043478260869,
"grad_norm": 61.10118865966797,
"learning_rate": 8.202227609982102e-07,
"loss": 0.23725251853466034,
"step": 726
},
{
"epoch": 0.6321739130434783,
"grad_norm": 188.16836547851562,
"learning_rate": 8.168476380798e-07,
"loss": 0.7833305597305298,
"step": 727
},
{
"epoch": 0.6330434782608696,
"grad_norm": 40.73017883300781,
"learning_rate": 8.134761002084093e-07,
"loss": 0.4744747281074524,
"step": 728
},
{
"epoch": 0.6339130434782608,
"grad_norm": 108.56316375732422,
"learning_rate": 8.1010817528905e-07,
"loss": 2.7104363441467285,
"step": 729
},
{
"epoch": 0.6347826086956522,
"grad_norm": 275.402587890625,
"learning_rate": 8.067438911968305e-07,
"loss": 0.34983405470848083,
"step": 730
},
{
"epoch": 0.6356521739130435,
"grad_norm": 117.67816925048828,
"learning_rate": 8.033832757767261e-07,
"loss": 0.19223889708518982,
"step": 731
},
{
"epoch": 0.6365217391304347,
"grad_norm": 37.56853103637695,
"learning_rate": 8.00026356843347e-07,
"loss": 0.273229718208313,
"step": 732
},
{
"epoch": 0.6373913043478261,
"grad_norm": 76.53671264648438,
"learning_rate": 7.9667316218071e-07,
"loss": 0.48756998777389526,
"step": 733
},
{
"epoch": 0.6382608695652174,
"grad_norm": 116.2587890625,
"learning_rate": 7.933237195420063e-07,
"loss": 1.9217708110809326,
"step": 734
},
{
"epoch": 0.6391304347826087,
"grad_norm": 46.10844039916992,
"learning_rate": 7.899780566493737e-07,
"loss": 0.38386499881744385,
"step": 735
},
{
"epoch": 0.64,
"grad_norm": 144.63941955566406,
"learning_rate": 7.866362011936657e-07,
"loss": 0.1522190421819687,
"step": 736
},
{
"epoch": 0.6408695652173914,
"grad_norm": 37.94704818725586,
"learning_rate": 7.832981808342238e-07,
"loss": 2.6867988109588623,
"step": 737
},
{
"epoch": 0.6417391304347826,
"grad_norm": 22.422178268432617,
"learning_rate": 7.799640231986476e-07,
"loss": 0.12112091481685638,
"step": 738
},
{
"epoch": 0.6426086956521739,
"grad_norm": 66.90673065185547,
"learning_rate": 7.766337558825655e-07,
"loss": 0.5523828864097595,
"step": 739
},
{
"epoch": 0.6434782608695652,
"grad_norm": 40.05387878417969,
"learning_rate": 7.733074064494084e-07,
"loss": 0.45325636863708496,
"step": 740
},
{
"epoch": 0.6443478260869565,
"grad_norm": 43.796695709228516,
"learning_rate": 7.699850024301792e-07,
"loss": 0.5340378284454346,
"step": 741
},
{
"epoch": 0.6452173913043479,
"grad_norm": 122.44947814941406,
"learning_rate": 7.666665713232266e-07,
"loss": 0.8238842487335205,
"step": 742
},
{
"epoch": 0.6460869565217391,
"grad_norm": 1319.661376953125,
"learning_rate": 7.633521405940167e-07,
"loss": 1.1156479120254517,
"step": 743
},
{
"epoch": 0.6469565217391304,
"grad_norm": 108.50177764892578,
"learning_rate": 7.600417376749065e-07,
"loss": 0.5874735116958618,
"step": 744
},
{
"epoch": 0.6478260869565218,
"grad_norm": 1420.3948974609375,
"learning_rate": 7.567353899649151e-07,
"loss": 0.9687666296958923,
"step": 745
},
{
"epoch": 0.648695652173913,
"grad_norm": 341.1820068359375,
"learning_rate": 7.534331248294988e-07,
"loss": 2.0976004600524902,
"step": 746
},
{
"epoch": 0.6495652173913044,
"grad_norm": 45.289859771728516,
"learning_rate": 7.50134969600324e-07,
"loss": 0.34963545203208923,
"step": 747
},
{
"epoch": 0.6504347826086957,
"grad_norm": 16.49801254272461,
"learning_rate": 7.468409515750408e-07,
"loss": 0.06994298100471497,
"step": 748
},
{
"epoch": 0.6513043478260869,
"grad_norm": 73.21430206298828,
"learning_rate": 7.435510980170571e-07,
"loss": 1.5250405073165894,
"step": 749
},
{
"epoch": 0.6521739130434783,
"grad_norm": 21.338817596435547,
"learning_rate": 7.402654361553124e-07,
"loss": 0.1742645502090454,
"step": 750
},
{
"epoch": 0.6530434782608696,
"grad_norm": 29.892614364624023,
"learning_rate": 7.369839931840543e-07,
"loss": 0.19237443804740906,
"step": 751
},
{
"epoch": 0.6539130434782608,
"grad_norm": 47.94505310058594,
"learning_rate": 7.337067962626109e-07,
"loss": 0.1351660043001175,
"step": 752
},
{
"epoch": 0.6547826086956522,
"grad_norm": 18.830358505249023,
"learning_rate": 7.304338725151682e-07,
"loss": 0.10270270705223083,
"step": 753
},
{
"epoch": 0.6556521739130434,
"grad_norm": 57.59758377075195,
"learning_rate": 7.271652490305439e-07,
"loss": 0.23244795203208923,
"step": 754
},
{
"epoch": 0.6565217391304348,
"grad_norm": 30.00511932373047,
"learning_rate": 7.239009528619645e-07,
"loss": 0.2164381593465805,
"step": 755
},
{
"epoch": 0.6573913043478261,
"grad_norm": 74.43519592285156,
"learning_rate": 7.206410110268404e-07,
"loss": 0.3441828191280365,
"step": 756
},
{
"epoch": 0.6582608695652173,
"grad_norm": 457.5736083984375,
"learning_rate": 7.173854505065439e-07,
"loss": 1.7203702926635742,
"step": 757
},
{
"epoch": 0.6591304347826087,
"grad_norm": 70.8633041381836,
"learning_rate": 7.141342982461822e-07,
"loss": 0.6303427219390869,
"step": 758
},
{
"epoch": 0.66,
"grad_norm": 63.65497589111328,
"learning_rate": 7.1088758115438e-07,
"loss": 0.4415142238140106,
"step": 759
},
{
"epoch": 0.6608695652173913,
"grad_norm": 98.77738189697266,
"learning_rate": 7.076453261030525e-07,
"loss": 1.5135620832443237,
"step": 760
},
{
"epoch": 0.6617391304347826,
"grad_norm": 218.7939910888672,
"learning_rate": 7.044075599271827e-07,
"loss": 0.5372211933135986,
"step": 761
},
{
"epoch": 0.662608695652174,
"grad_norm": 394.105224609375,
"learning_rate": 7.011743094246026e-07,
"loss": 1.5432229042053223,
"step": 762
},
{
"epoch": 0.6634782608695652,
"grad_norm": 74.53784942626953,
"learning_rate": 6.979456013557697e-07,
"loss": 0.7411608099937439,
"step": 763
},
{
"epoch": 0.6643478260869565,
"grad_norm": 21.297935485839844,
"learning_rate": 6.947214624435433e-07,
"loss": 0.12155219167470932,
"step": 764
},
{
"epoch": 0.6652173913043479,
"grad_norm": 52.77621078491211,
"learning_rate": 6.915019193729677e-07,
"loss": 0.7553197741508484,
"step": 765
},
{
"epoch": 0.6660869565217391,
"grad_norm": 87.89788055419922,
"learning_rate": 6.882869987910482e-07,
"loss": 0.7316939830780029,
"step": 766
},
{
"epoch": 0.6669565217391304,
"grad_norm": 34.747337341308594,
"learning_rate": 6.850767273065319e-07,
"loss": 5.6733927726745605,
"step": 767
},
{
"epoch": 0.6678260869565218,
"grad_norm": 30.68366241455078,
"learning_rate": 6.818711314896853e-07,
"loss": 0.26603609323501587,
"step": 768
},
{
"epoch": 0.668695652173913,
"grad_norm": 59.08272933959961,
"learning_rate": 6.78670237872078e-07,
"loss": 1.0568660497665405,
"step": 769
},
{
"epoch": 0.6695652173913044,
"grad_norm": 77.0888671875,
"learning_rate": 6.754740729463605e-07,
"loss": 0.3410334587097168,
"step": 770
},
{
"epoch": 0.6704347826086956,
"grad_norm": 27.843608856201172,
"learning_rate": 6.722826631660459e-07,
"loss": 0.8075916171073914,
"step": 771
},
{
"epoch": 0.671304347826087,
"grad_norm": 34.91166305541992,
"learning_rate": 6.690960349452894e-07,
"loss": 0.5702222585678101,
"step": 772
},
{
"epoch": 0.6721739130434783,
"grad_norm": 52.66334915161133,
"learning_rate": 6.659142146586725e-07,
"loss": 0.2313537299633026,
"step": 773
},
{
"epoch": 0.6730434782608695,
"grad_norm": 121.50244903564453,
"learning_rate": 6.627372286409829e-07,
"loss": 0.47095775604248047,
"step": 774
},
{
"epoch": 0.6739130434782609,
"grad_norm": 57.6857795715332,
"learning_rate": 6.595651031869954e-07,
"loss": 0.3306769132614136,
"step": 775
},
{
"epoch": 0.6747826086956522,
"grad_norm": 1026.720947265625,
"learning_rate": 6.563978645512576e-07,
"loss": 1.298703670501709,
"step": 776
},
{
"epoch": 0.6756521739130434,
"grad_norm": 35.41200256347656,
"learning_rate": 6.532355389478689e-07,
"loss": 0.44216975569725037,
"step": 777
},
{
"epoch": 0.6765217391304348,
"grad_norm": 730.8540649414062,
"learning_rate": 6.500781525502674e-07,
"loss": 0.6264848113059998,
"step": 778
},
{
"epoch": 0.6773913043478261,
"grad_norm": 152.3806610107422,
"learning_rate": 6.469257314910085e-07,
"loss": 1.5880484580993652,
"step": 779
},
{
"epoch": 0.6782608695652174,
"grad_norm": 714.1800537109375,
"learning_rate": 6.437783018615533e-07,
"loss": 1.167626976966858,
"step": 780
},
{
"epoch": 0.6791304347826087,
"grad_norm": 216.6626739501953,
"learning_rate": 6.406358897120504e-07,
"loss": 0.7450761795043945,
"step": 781
},
{
"epoch": 0.68,
"grad_norm": 444.0139465332031,
"learning_rate": 6.37498521051119e-07,
"loss": 0.7422975301742554,
"step": 782
},
{
"epoch": 0.6808695652173913,
"grad_norm": 634.5380859375,
"learning_rate": 6.34366221845637e-07,
"loss": 0.9782680869102478,
"step": 783
},
{
"epoch": 0.6817391304347826,
"grad_norm": 34.93865203857422,
"learning_rate": 6.312390180205233e-07,
"loss": 0.24270987510681152,
"step": 784
},
{
"epoch": 0.6826086956521739,
"grad_norm": 25.33079719543457,
"learning_rate": 6.281169354585249e-07,
"loss": 0.17209675908088684,
"step": 785
},
{
"epoch": 0.6834782608695652,
"grad_norm": 69.52841186523438,
"learning_rate": 6.250000000000004e-07,
"loss": 0.3119470477104187,
"step": 786
},
{
"epoch": 0.6843478260869565,
"grad_norm": 76.1631088256836,
"learning_rate": 6.218882374427089e-07,
"loss": 0.24853132665157318,
"step": 787
},
{
"epoch": 0.6852173913043478,
"grad_norm": 68.14842224121094,
"learning_rate": 6.187816735415954e-07,
"loss": 0.57975834608078,
"step": 788
},
{
"epoch": 0.6860869565217391,
"grad_norm": 110.76408386230469,
"learning_rate": 6.156803340085772e-07,
"loss": 0.8176261782646179,
"step": 789
},
{
"epoch": 0.6869565217391305,
"grad_norm": 101.59113311767578,
"learning_rate": 6.125842445123305e-07,
"loss": 0.9113748073577881,
"step": 790
},
{
"epoch": 0.6878260869565217,
"grad_norm": 99.70103454589844,
"learning_rate": 6.094934306780799e-07,
"loss": 0.49612969160079956,
"step": 791
},
{
"epoch": 0.688695652173913,
"grad_norm": 51.14309310913086,
"learning_rate": 6.064079180873851e-07,
"loss": 0.22991034388542175,
"step": 792
},
{
"epoch": 0.6895652173913044,
"grad_norm": 414.42901611328125,
"learning_rate": 6.033277322779289e-07,
"loss": 2.277712821960449,
"step": 793
},
{
"epoch": 0.6904347826086956,
"grad_norm": 53.07019805908203,
"learning_rate": 6.002528987433068e-07,
"loss": 0.18213284015655518,
"step": 794
},
{
"epoch": 0.691304347826087,
"grad_norm": 57.624019622802734,
"learning_rate": 5.971834429328151e-07,
"loss": 0.23929773271083832,
"step": 795
},
{
"epoch": 0.6921739130434783,
"grad_norm": 122.63893127441406,
"learning_rate": 5.9411939025124e-07,
"loss": 1.7961044311523438,
"step": 796
},
{
"epoch": 0.6930434782608695,
"grad_norm": 78.53750610351562,
"learning_rate": 5.910607660586488e-07,
"loss": 0.47248125076293945,
"step": 797
},
{
"epoch": 0.6939130434782609,
"grad_norm": 39.46408462524414,
"learning_rate": 5.880075956701792e-07,
"loss": 0.24108974635601044,
"step": 798
},
{
"epoch": 0.6947826086956522,
"grad_norm": 116.72267150878906,
"learning_rate": 5.849599043558295e-07,
"loss": 1.8262581825256348,
"step": 799
},
{
"epoch": 0.6956521739130435,
"grad_norm": 114.56092071533203,
"learning_rate": 5.819177173402487e-07,
"loss": 1.46011483669281,
"step": 800
},
{
"epoch": 0.6965217391304348,
"grad_norm": 376.9473571777344,
"learning_rate": 5.7888105980253e-07,
"loss": 0.5265900492668152,
"step": 801
},
{
"epoch": 0.697391304347826,
"grad_norm": 38.614078521728516,
"learning_rate": 5.758499568760012e-07,
"loss": 0.33868408203125,
"step": 802
},
{
"epoch": 0.6982608695652174,
"grad_norm": 68.75576782226562,
"learning_rate": 5.728244336480152e-07,
"loss": 0.29384180903434753,
"step": 803
},
{
"epoch": 0.6991304347826087,
"grad_norm": 34.4858512878418,
"learning_rate": 5.698045151597446e-07,
"loss": 0.2218216061592102,
"step": 804
},
{
"epoch": 0.7,
"grad_norm": 57.50834655761719,
"learning_rate": 5.667902264059738e-07,
"loss": 0.5318577289581299,
"step": 805
},
{
"epoch": 0.7008695652173913,
"grad_norm": 19.928810119628906,
"learning_rate": 5.63781592334892e-07,
"loss": 1.9398574829101562,
"step": 806
},
{
"epoch": 0.7017391304347826,
"grad_norm": 30.564926147460938,
"learning_rate": 5.607786378478855e-07,
"loss": 0.19323492050170898,
"step": 807
},
{
"epoch": 0.7026086956521739,
"grad_norm": 115.8697738647461,
"learning_rate": 5.577813877993337e-07,
"loss": 0.483132928609848,
"step": 808
},
{
"epoch": 0.7034782608695652,
"grad_norm": 356.4931335449219,
"learning_rate": 5.547898669964026e-07,
"loss": 0.6454263925552368,
"step": 809
},
{
"epoch": 0.7043478260869566,
"grad_norm": 73.97138214111328,
"learning_rate": 5.518041001988379e-07,
"loss": 2.951838254928589,
"step": 810
},
{
"epoch": 0.7052173913043478,
"grad_norm": 78.91866302490234,
"learning_rate": 5.488241121187632e-07,
"loss": 0.5331160426139832,
"step": 811
},
{
"epoch": 0.7060869565217391,
"grad_norm": 89.43058013916016,
"learning_rate": 5.458499274204724e-07,
"loss": 1.7172069549560547,
"step": 812
},
{
"epoch": 0.7069565217391305,
"grad_norm": 107.33296203613281,
"learning_rate": 5.428815707202283e-07,
"loss": 0.5896658897399902,
"step": 813
},
{
"epoch": 0.7078260869565217,
"grad_norm": 749.8272094726562,
"learning_rate": 5.399190665860555e-07,
"loss": 1.6135833263397217,
"step": 814
},
{
"epoch": 0.7086956521739131,
"grad_norm": 56.0038948059082,
"learning_rate": 5.369624395375405e-07,
"loss": 0.3628445267677307,
"step": 815
},
{
"epoch": 0.7095652173913043,
"grad_norm": 361.686767578125,
"learning_rate": 5.34011714045627e-07,
"loss": 1.8731521368026733,
"step": 816
},
{
"epoch": 0.7104347826086956,
"grad_norm": 67.81514739990234,
"learning_rate": 5.310669145324132e-07,
"loss": 0.3275434374809265,
"step": 817
},
{
"epoch": 0.711304347826087,
"grad_norm": 78.96922302246094,
"learning_rate": 5.281280653709503e-07,
"loss": 0.4090062081813812,
"step": 818
},
{
"epoch": 0.7121739130434782,
"grad_norm": 113.99473571777344,
"learning_rate": 5.251951908850407e-07,
"loss": 1.428281307220459,
"step": 819
},
{
"epoch": 0.7130434782608696,
"grad_norm": 100.74822998046875,
"learning_rate": 5.22268315349037e-07,
"loss": 0.336014986038208,
"step": 820
},
{
"epoch": 0.7139130434782609,
"grad_norm": 20.198261260986328,
"learning_rate": 5.193474629876392e-07,
"loss": 0.11257904767990112,
"step": 821
},
{
"epoch": 0.7147826086956521,
"grad_norm": 32.469383239746094,
"learning_rate": 5.164326579756971e-07,
"loss": 0.2793066203594208,
"step": 822
},
{
"epoch": 0.7156521739130435,
"grad_norm": 107.70726013183594,
"learning_rate": 5.13523924438008e-07,
"loss": 0.7388822436332703,
"step": 823
},
{
"epoch": 0.7165217391304348,
"grad_norm": 43.007850646972656,
"learning_rate": 5.106212864491185e-07,
"loss": 0.3432861268520355,
"step": 824
},
{
"epoch": 0.717391304347826,
"grad_norm": 62.87594223022461,
"learning_rate": 5.077247680331232e-07,
"loss": 0.9271078109741211,
"step": 825
},
{
"epoch": 0.7182608695652174,
"grad_norm": 131.6029052734375,
"learning_rate": 5.048343931634684e-07,
"loss": 1.322260856628418,
"step": 826
},
{
"epoch": 0.7191304347826087,
"grad_norm": 34.05308151245117,
"learning_rate": 5.019501857627521e-07,
"loss": 0.1705697476863861,
"step": 827
},
{
"epoch": 0.72,
"grad_norm": 543.477783203125,
"learning_rate": 4.990721697025262e-07,
"loss": 0.8439666628837585,
"step": 828
},
{
"epoch": 0.7208695652173913,
"grad_norm": 47.18918991088867,
"learning_rate": 4.962003688030994e-07,
"loss": 1.6587836742401123,
"step": 829
},
{
"epoch": 0.7217391304347827,
"grad_norm": 546.7724609375,
"learning_rate": 4.933348068333399e-07,
"loss": 2.029395580291748,
"step": 830
},
{
"epoch": 0.7226086956521739,
"grad_norm": 21.91991424560547,
"learning_rate": 4.904755075104774e-07,
"loss": 0.09440520405769348,
"step": 831
},
{
"epoch": 0.7234782608695652,
"grad_norm": 41.543025970458984,
"learning_rate": 4.876224944999089e-07,
"loss": 0.15690022706985474,
"step": 832
},
{
"epoch": 0.7243478260869565,
"grad_norm": 108.59095001220703,
"learning_rate": 4.847757914150018e-07,
"loss": 0.3057916760444641,
"step": 833
},
{
"epoch": 0.7252173913043478,
"grad_norm": 63.55922317504883,
"learning_rate": 4.819354218168988e-07,
"loss": 0.4306486248970032,
"step": 834
},
{
"epoch": 0.7260869565217392,
"grad_norm": 335.4298095703125,
"learning_rate": 4.791014092143212e-07,
"loss": 1.2242369651794434,
"step": 835
},
{
"epoch": 0.7269565217391304,
"grad_norm": 77.75059509277344,
"learning_rate": 4.7627377706337713e-07,
"loss": 0.37284398078918457,
"step": 836
},
{
"epoch": 0.7278260869565217,
"grad_norm": 122.30418395996094,
"learning_rate": 4.734525487673659e-07,
"loss": 1.8792517185211182,
"step": 837
},
{
"epoch": 0.7286956521739131,
"grad_norm": 25.31227684020996,
"learning_rate": 4.706377476765832e-07,
"loss": 0.21242603659629822,
"step": 838
},
{
"epoch": 0.7295652173913043,
"grad_norm": 93.61386108398438,
"learning_rate": 4.6782939708813035e-07,
"loss": 2.2290825843811035,
"step": 839
},
{
"epoch": 0.7304347826086957,
"grad_norm": 22.625442504882812,
"learning_rate": 4.650275202457197e-07,
"loss": 0.1646450161933899,
"step": 840
},
{
"epoch": 0.731304347826087,
"grad_norm": 70.74088287353516,
"learning_rate": 4.622321403394832e-07,
"loss": 0.6447851657867432,
"step": 841
},
{
"epoch": 0.7321739130434782,
"grad_norm": 67.04022216796875,
"learning_rate": 4.594432805057786e-07,
"loss": 0.32657402753829956,
"step": 842
},
{
"epoch": 0.7330434782608696,
"grad_norm": 40.18130111694336,
"learning_rate": 4.566609638270006e-07,
"loss": 0.18354171514511108,
"step": 843
},
{
"epoch": 0.7339130434782609,
"grad_norm": 18.427295684814453,
"learning_rate": 4.538852133313888e-07,
"loss": 0.100941501557827,
"step": 844
},
{
"epoch": 0.7347826086956522,
"grad_norm": 712.8763427734375,
"learning_rate": 4.511160519928352e-07,
"loss": 1.044054388999939,
"step": 845
},
{
"epoch": 0.7356521739130435,
"grad_norm": 49.283782958984375,
"learning_rate": 4.483535027306975e-07,
"loss": 0.24343112111091614,
"step": 846
},
{
"epoch": 0.7365217391304347,
"grad_norm": 47.670867919921875,
"learning_rate": 4.4559758840960656e-07,
"loss": 0.395723819732666,
"step": 847
},
{
"epoch": 0.7373913043478261,
"grad_norm": 124.79226684570312,
"learning_rate": 4.4284833183927927e-07,
"loss": 0.5531738996505737,
"step": 848
},
{
"epoch": 0.7382608695652174,
"grad_norm": 63.03324890136719,
"learning_rate": 4.401057557743267e-07,
"loss": 0.55580073595047,
"step": 849
},
{
"epoch": 0.7391304347826086,
"grad_norm": 126.1244125366211,
"learning_rate": 4.373698829140698e-07,
"loss": 0.5353561639785767,
"step": 850
},
{
"epoch": 0.74,
"grad_norm": 15.690673828125,
"learning_rate": 4.3464073590234896e-07,
"loss": 0.08779134601354599,
"step": 851
},
{
"epoch": 0.7408695652173913,
"grad_norm": 65.6275863647461,
"learning_rate": 4.319183373273361e-07,
"loss": 0.9080826640129089,
"step": 852
},
{
"epoch": 0.7417391304347826,
"grad_norm": 74.13648223876953,
"learning_rate": 4.2920270972134997e-07,
"loss": 0.3586421608924866,
"step": 853
},
{
"epoch": 0.7426086956521739,
"grad_norm": 51.53708267211914,
"learning_rate": 4.264938755606679e-07,
"loss": 0.6314980983734131,
"step": 854
},
{
"epoch": 0.7434782608695653,
"grad_norm": 43.11322021484375,
"learning_rate": 4.23791857265341e-07,
"loss": 0.15180379152297974,
"step": 855
},
{
"epoch": 0.7443478260869565,
"grad_norm": 102.40668487548828,
"learning_rate": 4.210966771990063e-07,
"loss": 2.185192823410034,
"step": 856
},
{
"epoch": 0.7452173913043478,
"grad_norm": 248.11219787597656,
"learning_rate": 4.184083576687047e-07,
"loss": 0.5532682538032532,
"step": 857
},
{
"epoch": 0.7460869565217392,
"grad_norm": 63.720947265625,
"learning_rate": 4.157269209246946e-07,
"loss": 0.88653564453125,
"step": 858
},
{
"epoch": 0.7469565217391304,
"grad_norm": 43.15514373779297,
"learning_rate": 4.130523891602682e-07,
"loss": 0.27271735668182373,
"step": 859
},
{
"epoch": 0.7478260869565218,
"grad_norm": 138.06358337402344,
"learning_rate": 4.1038478451156607e-07,
"loss": 0.5255391001701355,
"step": 860
},
{
"epoch": 0.7486956521739131,
"grad_norm": 271.3030090332031,
"learning_rate": 4.077241290573987e-07,
"loss": 1.7646262645721436,
"step": 861
},
{
"epoch": 0.7495652173913043,
"grad_norm": 45.729671478271484,
"learning_rate": 4.0507044481905707e-07,
"loss": 0.23271086812019348,
"step": 862
},
{
"epoch": 0.7504347826086957,
"grad_norm": 46.020179748535156,
"learning_rate": 4.0242375376013635e-07,
"loss": 0.3205915689468384,
"step": 863
},
{
"epoch": 0.7513043478260869,
"grad_norm": 44.1895637512207,
"learning_rate": 3.9978407778635057e-07,
"loss": 0.19646534323692322,
"step": 864
},
{
"epoch": 0.7521739130434782,
"grad_norm": 74.87690734863281,
"learning_rate": 3.971514387453533e-07,
"loss": 0.7989358901977539,
"step": 865
},
{
"epoch": 0.7530434782608696,
"grad_norm": 49.57170486450195,
"learning_rate": 3.9452585842655434e-07,
"loss": 0.25784504413604736,
"step": 866
},
{
"epoch": 0.7539130434782608,
"grad_norm": 45.67580795288086,
"learning_rate": 3.9190735856094245e-07,
"loss": 0.2808690369129181,
"step": 867
},
{
"epoch": 0.7547826086956522,
"grad_norm": 114.64430236816406,
"learning_rate": 3.892959608209035e-07,
"loss": 0.5124662518501282,
"step": 868
},
{
"epoch": 0.7556521739130435,
"grad_norm": 44.195045471191406,
"learning_rate": 3.866916868200421e-07,
"loss": 0.17964337766170502,
"step": 869
},
{
"epoch": 0.7565217391304347,
"grad_norm": 39.690452575683594,
"learning_rate": 3.840945581130008e-07,
"loss": 0.21143034100532532,
"step": 870
},
{
"epoch": 0.7573913043478261,
"grad_norm": 22.441852569580078,
"learning_rate": 3.815045961952846e-07,
"loss": 0.09567797183990479,
"step": 871
},
{
"epoch": 0.7582608695652174,
"grad_norm": 42.52641296386719,
"learning_rate": 3.7892182250308135e-07,
"loss": 0.1978321522474289,
"step": 872
},
{
"epoch": 0.7591304347826087,
"grad_norm": 59.721248626708984,
"learning_rate": 3.763462584130835e-07,
"loss": 0.3400568664073944,
"step": 873
},
{
"epoch": 0.76,
"grad_norm": 718.6740112304688,
"learning_rate": 3.737779252423132e-07,
"loss": 1.0350968837738037,
"step": 874
},
{
"epoch": 0.7608695652173914,
"grad_norm": 363.79840087890625,
"learning_rate": 3.712168442479447e-07,
"loss": 1.3022434711456299,
"step": 875
},
{
"epoch": 0.7617391304347826,
"grad_norm": 33.88823699951172,
"learning_rate": 3.686630366271286e-07,
"loss": 0.2878289520740509,
"step": 876
},
{
"epoch": 0.7626086956521739,
"grad_norm": 19.302053451538086,
"learning_rate": 3.661165235168157e-07,
"loss": 0.13586270809173584,
"step": 877
},
{
"epoch": 0.7634782608695653,
"grad_norm": 47.885318756103516,
"learning_rate": 3.6357732599358364e-07,
"loss": 0.17454132437705994,
"step": 878
},
{
"epoch": 0.7643478260869565,
"grad_norm": 57.5684700012207,
"learning_rate": 3.610454650734614e-07,
"loss": 0.27047663927078247,
"step": 879
},
{
"epoch": 0.7652173913043478,
"grad_norm": 41.52002716064453,
"learning_rate": 3.5852096171175496e-07,
"loss": 0.499724805355072,
"step": 880
},
{
"epoch": 0.7660869565217391,
"grad_norm": 28.597131729125977,
"learning_rate": 3.560038368028748e-07,
"loss": 1.7706043720245361,
"step": 881
},
{
"epoch": 0.7669565217391304,
"grad_norm": 30.22584342956543,
"learning_rate": 3.5349411118016275e-07,
"loss": 0.2314690351486206,
"step": 882
},
{
"epoch": 0.7678260869565218,
"grad_norm": 464.5823059082031,
"learning_rate": 3.5099180561571966e-07,
"loss": 1.1635401248931885,
"step": 883
},
{
"epoch": 0.768695652173913,
"grad_norm": 1775.8023681640625,
"learning_rate": 3.4849694082023197e-07,
"loss": 3.254634380340576,
"step": 884
},
{
"epoch": 0.7695652173913043,
"grad_norm": 35.192115783691406,
"learning_rate": 3.460095374428028e-07,
"loss": 0.19749431312084198,
"step": 885
},
{
"epoch": 0.7704347826086957,
"grad_norm": 56.71238708496094,
"learning_rate": 3.4352961607077976e-07,
"loss": 0.20573453605175018,
"step": 886
},
{
"epoch": 0.7713043478260869,
"grad_norm": 28.682952880859375,
"learning_rate": 3.4105719722958326e-07,
"loss": 0.1291721910238266,
"step": 887
},
{
"epoch": 0.7721739130434783,
"grad_norm": 76.26278686523438,
"learning_rate": 3.3859230138253945e-07,
"loss": 0.2512245178222656,
"step": 888
},
{
"epoch": 0.7730434782608696,
"grad_norm": 26.25237464904785,
"learning_rate": 3.361349489307085e-07,
"loss": 0.1554962396621704,
"step": 889
},
{
"epoch": 0.7739130434782608,
"grad_norm": 173.60386657714844,
"learning_rate": 3.336851602127174e-07,
"loss": 0.5558016300201416,
"step": 890
},
{
"epoch": 0.7747826086956522,
"grad_norm": 40.35780334472656,
"learning_rate": 3.3124295550458956e-07,
"loss": 0.2118375301361084,
"step": 891
},
{
"epoch": 0.7756521739130435,
"grad_norm": 99.60546875,
"learning_rate": 3.2880835501957917e-07,
"loss": 2.0392446517944336,
"step": 892
},
{
"epoch": 0.7765217391304348,
"grad_norm": 46.36368179321289,
"learning_rate": 3.2638137890800297e-07,
"loss": 0.7251726984977722,
"step": 893
},
{
"epoch": 0.7773913043478261,
"grad_norm": 84.32417297363281,
"learning_rate": 3.239620472570737e-07,
"loss": 1.1115362644195557,
"step": 894
},
{
"epoch": 0.7782608695652173,
"grad_norm": 223.76760864257812,
"learning_rate": 3.2155038009073186e-07,
"loss": 0.4985317587852478,
"step": 895
},
{
"epoch": 0.7791304347826087,
"grad_norm": 35.70825958251953,
"learning_rate": 3.19146397369485e-07,
"loss": 0.21705618500709534,
"step": 896
},
{
"epoch": 0.78,
"grad_norm": 28.218017578125,
"learning_rate": 3.1675011899023597e-07,
"loss": 0.23779720067977905,
"step": 897
},
{
"epoch": 0.7808695652173913,
"grad_norm": 54.67338562011719,
"learning_rate": 3.143615647861235e-07,
"loss": 0.19339659810066223,
"step": 898
},
{
"epoch": 0.7817391304347826,
"grad_norm": 61.03525161743164,
"learning_rate": 3.1198075452635575e-07,
"loss": 0.25105974078178406,
"step": 899
},
{
"epoch": 0.782608695652174,
"grad_norm": 157.67063903808594,
"learning_rate": 3.0960770791604683e-07,
"loss": 1.4121520519256592,
"step": 900
},
{
"epoch": 0.7834782608695652,
"grad_norm": 37.75104522705078,
"learning_rate": 3.0724244459605355e-07,
"loss": 0.48645269870758057,
"step": 901
},
{
"epoch": 0.7843478260869565,
"grad_norm": 61.33236312866211,
"learning_rate": 3.0488498414281377e-07,
"loss": 2.4960999488830566,
"step": 902
},
{
"epoch": 0.7852173913043479,
"grad_norm": 223.14263916015625,
"learning_rate": 3.025353460681836e-07,
"loss": 0.5283726453781128,
"step": 903
},
{
"epoch": 0.7860869565217391,
"grad_norm": 80.25823211669922,
"learning_rate": 3.0019354981927663e-07,
"loss": 0.9475429058074951,
"step": 904
},
{
"epoch": 0.7869565217391304,
"grad_norm": 180.53810119628906,
"learning_rate": 2.978596147783011e-07,
"loss": 0.6295070648193359,
"step": 905
},
{
"epoch": 0.7878260869565218,
"grad_norm": 42.07815170288086,
"learning_rate": 2.9553356026240223e-07,
"loss": 0.37310606241226196,
"step": 906
},
{
"epoch": 0.788695652173913,
"grad_norm": 170.53236389160156,
"learning_rate": 2.9321540552350075e-07,
"loss": 0.4908299446105957,
"step": 907
},
{
"epoch": 0.7895652173913044,
"grad_norm": 33.67927932739258,
"learning_rate": 2.9090516974813285e-07,
"loss": 0.15058505535125732,
"step": 908
},
{
"epoch": 0.7904347826086957,
"grad_norm": 83.08329772949219,
"learning_rate": 2.886028720572932e-07,
"loss": 0.4044651985168457,
"step": 909
},
{
"epoch": 0.7913043478260869,
"grad_norm": 28.819438934326172,
"learning_rate": 2.863085315062755e-07,
"loss": 0.24753952026367188,
"step": 910
},
{
"epoch": 0.7921739130434783,
"grad_norm": 57.011356353759766,
"learning_rate": 2.840221670845153e-07,
"loss": 0.5404356718063354,
"step": 911
},
{
"epoch": 0.7930434782608695,
"grad_norm": 121.65554809570312,
"learning_rate": 2.817437977154319e-07,
"loss": 0.7699071168899536,
"step": 912
},
{
"epoch": 0.7939130434782609,
"grad_norm": 158.49783325195312,
"learning_rate": 2.7947344225627293e-07,
"loss": 0.7568212747573853,
"step": 913
},
{
"epoch": 0.7947826086956522,
"grad_norm": 82.36697387695312,
"learning_rate": 2.772111194979582e-07,
"loss": 0.3591994047164917,
"step": 914
},
{
"epoch": 0.7956521739130434,
"grad_norm": 78.17515563964844,
"learning_rate": 2.749568481649224e-07,
"loss": 0.5096107721328735,
"step": 915
},
{
"epoch": 0.7965217391304348,
"grad_norm": 36.79532241821289,
"learning_rate": 2.727106469149628e-07,
"loss": 0.5681661367416382,
"step": 916
},
{
"epoch": 0.7973913043478261,
"grad_norm": 110.35755157470703,
"learning_rate": 2.704725343390828e-07,
"loss": 0.5396387577056885,
"step": 917
},
{
"epoch": 0.7982608695652174,
"grad_norm": 37.197662353515625,
"learning_rate": 2.6824252896133913e-07,
"loss": 0.3228336572647095,
"step": 918
},
{
"epoch": 0.7991304347826087,
"grad_norm": 120.28250885009766,
"learning_rate": 2.6602064923868745e-07,
"loss": 0.42075109481811523,
"step": 919
},
{
"epoch": 0.8,
"grad_norm": 23.663742065429688,
"learning_rate": 2.6380691356083095e-07,
"loss": 0.12314199656248093,
"step": 920
},
{
"epoch": 0.8008695652173913,
"grad_norm": 24.69415283203125,
"learning_rate": 2.6160134025006753e-07,
"loss": 0.2467924952507019,
"step": 921
},
{
"epoch": 0.8017391304347826,
"grad_norm": 35.963863372802734,
"learning_rate": 2.5940394756113723e-07,
"loss": 0.3273763060569763,
"step": 922
},
{
"epoch": 0.802608695652174,
"grad_norm": 316.12353515625,
"learning_rate": 2.5721475368107293e-07,
"loss": 2.5479865074157715,
"step": 923
},
{
"epoch": 0.8034782608695652,
"grad_norm": 129.46871948242188,
"learning_rate": 2.550337767290487e-07,
"loss": 0.36221736669540405,
"step": 924
},
{
"epoch": 0.8043478260869565,
"grad_norm": 23.237716674804688,
"learning_rate": 2.5286103475623006e-07,
"loss": 0.1654108464717865,
"step": 925
},
{
"epoch": 0.8052173913043478,
"grad_norm": 35.187705993652344,
"learning_rate": 2.5069654574562384e-07,
"loss": 0.20308451354503632,
"step": 926
},
{
"epoch": 0.8060869565217391,
"grad_norm": 14.903914451599121,
"learning_rate": 2.4854032761193095e-07,
"loss": 0.32085609436035156,
"step": 927
},
{
"epoch": 0.8069565217391305,
"grad_norm": 75.45279693603516,
"learning_rate": 2.46392398201397e-07,
"loss": 0.2958734929561615,
"step": 928
},
{
"epoch": 0.8078260869565217,
"grad_norm": 44.90689468383789,
"learning_rate": 2.4425277529166456e-07,
"loss": 0.18282578885555267,
"step": 929
},
{
"epoch": 0.808695652173913,
"grad_norm": 312.2851257324219,
"learning_rate": 2.421214765916265e-07,
"loss": 0.9054602384567261,
"step": 930
},
{
"epoch": 0.8095652173913044,
"grad_norm": 189.71900939941406,
"learning_rate": 2.399985197412795e-07,
"loss": 0.6979514360427856,
"step": 931
},
{
"epoch": 0.8104347826086956,
"grad_norm": 59.73307800292969,
"learning_rate": 2.3788392231157643e-07,
"loss": 0.31773829460144043,
"step": 932
},
{
"epoch": 0.811304347826087,
"grad_norm": 136.75038146972656,
"learning_rate": 2.3577770180428365e-07,
"loss": 0.9085092544555664,
"step": 933
},
{
"epoch": 0.8121739130434783,
"grad_norm": 28.879255294799805,
"learning_rate": 2.336798756518342e-07,
"loss": 0.2065652459859848,
"step": 934
},
{
"epoch": 0.8130434782608695,
"grad_norm": 63.14043426513672,
"learning_rate": 2.3159046121718446e-07,
"loss": 0.2690063416957855,
"step": 935
},
{
"epoch": 0.8139130434782609,
"grad_norm": 37.704376220703125,
"learning_rate": 2.295094757936689e-07,
"loss": 0.30769312381744385,
"step": 936
},
{
"epoch": 0.8147826086956522,
"grad_norm": 154.78057861328125,
"learning_rate": 2.2743693660485898e-07,
"loss": 0.49501848220825195,
"step": 937
},
{
"epoch": 0.8156521739130435,
"grad_norm": 61.00483703613281,
"learning_rate": 2.253728608044198e-07,
"loss": 0.36444005370140076,
"step": 938
},
{
"epoch": 0.8165217391304348,
"grad_norm": 80.73182678222656,
"learning_rate": 2.23317265475968e-07,
"loss": 0.3402242660522461,
"step": 939
},
{
"epoch": 0.8173913043478261,
"grad_norm": 92.48069763183594,
"learning_rate": 2.2127016763292959e-07,
"loss": 0.6537570357322693,
"step": 940
},
{
"epoch": 0.8182608695652174,
"grad_norm": 42.01014709472656,
"learning_rate": 2.1923158421840063e-07,
"loss": 0.2773856818675995,
"step": 941
},
{
"epoch": 0.8191304347826087,
"grad_norm": 110.08451080322266,
"learning_rate": 2.1720153210500666e-07,
"loss": 0.5434136390686035,
"step": 942
},
{
"epoch": 0.82,
"grad_norm": 196.44781494140625,
"learning_rate": 2.151800280947615e-07,
"loss": 1.0605320930480957,
"step": 943
},
{
"epoch": 0.8208695652173913,
"grad_norm": 261.63934326171875,
"learning_rate": 2.1316708891893037e-07,
"loss": 0.9559299945831299,
"step": 944
},
{
"epoch": 0.8217391304347826,
"grad_norm": 37.74660110473633,
"learning_rate": 2.1116273123789013e-07,
"loss": 0.24474279582500458,
"step": 945
},
{
"epoch": 0.8226086956521739,
"grad_norm": 29.111513137817383,
"learning_rate": 2.091669716409922e-07,
"loss": 0.24549072980880737,
"step": 946
},
{
"epoch": 0.8234782608695652,
"grad_norm": 514.94140625,
"learning_rate": 2.071798266464234e-07,
"loss": 0.7626498937606812,
"step": 947
},
{
"epoch": 0.8243478260869566,
"grad_norm": 29.138675689697266,
"learning_rate": 2.0520131270107186e-07,
"loss": 0.1367233395576477,
"step": 948
},
{
"epoch": 0.8252173913043478,
"grad_norm": 72.38389587402344,
"learning_rate": 2.0323144618038928e-07,
"loss": 0.33508968353271484,
"step": 949
},
{
"epoch": 0.8260869565217391,
"grad_norm": 37.920196533203125,
"learning_rate": 2.012702433882552e-07,
"loss": 0.1984211504459381,
"step": 950
},
{
"epoch": 0.8269565217391305,
"grad_norm": 195.53111267089844,
"learning_rate": 1.9931772055684306e-07,
"loss": 0.4054205119609833,
"step": 951
},
{
"epoch": 0.8278260869565217,
"grad_norm": 111.53182983398438,
"learning_rate": 1.973738938464853e-07,
"loss": 0.5514519214630127,
"step": 952
},
{
"epoch": 0.828695652173913,
"grad_norm": 47.2476921081543,
"learning_rate": 1.954387793455402e-07,
"loss": 0.2770058810710907,
"step": 953
},
{
"epoch": 0.8295652173913044,
"grad_norm": 277.3168029785156,
"learning_rate": 1.9351239307025677e-07,
"loss": 2.0201573371887207,
"step": 954
},
{
"epoch": 0.8304347826086956,
"grad_norm": 31.699995040893555,
"learning_rate": 1.9159475096464486e-07,
"loss": 0.15156015753746033,
"step": 955
},
{
"epoch": 0.831304347826087,
"grad_norm": 59.57673263549805,
"learning_rate": 1.896858689003421e-07,
"loss": 0.39648258686065674,
"step": 956
},
{
"epoch": 0.8321739130434782,
"grad_norm": 80.7887954711914,
"learning_rate": 1.8778576267648106e-07,
"loss": 0.3375861346721649,
"step": 957
},
{
"epoch": 0.8330434782608696,
"grad_norm": 88.37937927246094,
"learning_rate": 1.858944480195611e-07,
"loss": 0.6140657663345337,
"step": 958
},
{
"epoch": 0.8339130434782609,
"grad_norm": 206.142822265625,
"learning_rate": 1.8401194058331656e-07,
"loss": 0.4756943881511688,
"step": 959
},
{
"epoch": 0.8347826086956521,
"grad_norm": 66.22407531738281,
"learning_rate": 1.8213825594858763e-07,
"loss": 0.20961140096187592,
"step": 960
},
{
"epoch": 0.8356521739130435,
"grad_norm": 44.93798065185547,
"learning_rate": 1.8027340962319077e-07,
"loss": 0.301072895526886,
"step": 961
},
{
"epoch": 0.8365217391304348,
"grad_norm": 156.2875213623047,
"learning_rate": 1.7841741704179155e-07,
"loss": 0.5325546264648438,
"step": 962
},
{
"epoch": 0.837391304347826,
"grad_norm": 111.82328796386719,
"learning_rate": 1.7657029356577622e-07,
"loss": 0.8474414348602295,
"step": 963
},
{
"epoch": 0.8382608695652174,
"grad_norm": 100.40882873535156,
"learning_rate": 1.7473205448312408e-07,
"loss": 0.3510282635688782,
"step": 964
},
{
"epoch": 0.8391304347826087,
"grad_norm": 28.670320510864258,
"learning_rate": 1.7290271500828236e-07,
"loss": 0.13728836178779602,
"step": 965
},
{
"epoch": 0.84,
"grad_norm": 36.97054672241211,
"learning_rate": 1.7108229028203888e-07,
"loss": 0.7176310420036316,
"step": 966
},
{
"epoch": 0.8408695652173913,
"grad_norm": 153.516845703125,
"learning_rate": 1.6927079537139706e-07,
"loss": 1.897335410118103,
"step": 967
},
{
"epoch": 0.8417391304347827,
"grad_norm": 220.7427520751953,
"learning_rate": 1.6746824526945163e-07,
"loss": 0.2686039209365845,
"step": 968
},
{
"epoch": 0.8426086956521739,
"grad_norm": 68.17507934570312,
"learning_rate": 1.656746548952648e-07,
"loss": 0.6522201895713806,
"step": 969
},
{
"epoch": 0.8434782608695652,
"grad_norm": 47.48100280761719,
"learning_rate": 1.6389003909374188e-07,
"loss": 0.14523625373840332,
"step": 970
},
{
"epoch": 0.8443478260869566,
"grad_norm": 72.62852478027344,
"learning_rate": 1.621144126355087e-07,
"loss": 0.6294858455657959,
"step": 971
},
{
"epoch": 0.8452173913043478,
"grad_norm": 55.98668670654297,
"learning_rate": 1.6034779021678978e-07,
"loss": 0.4580530822277069,
"step": 972
},
{
"epoch": 0.8460869565217392,
"grad_norm": 138.6509552001953,
"learning_rate": 1.5859018645928647e-07,
"loss": 0.6365782022476196,
"step": 973
},
{
"epoch": 0.8469565217391304,
"grad_norm": 355.2127685546875,
"learning_rate": 1.5684161591005596e-07,
"loss": 0.7044453620910645,
"step": 974
},
{
"epoch": 0.8478260869565217,
"grad_norm": 27.501501083374023,
"learning_rate": 1.551020930413902e-07,
"loss": 0.41071757674217224,
"step": 975
},
{
"epoch": 0.8486956521739131,
"grad_norm": 66.75444793701172,
"learning_rate": 1.5337163225069758e-07,
"loss": 0.27846837043762207,
"step": 976
},
{
"epoch": 0.8495652173913043,
"grad_norm": 120.44483184814453,
"learning_rate": 1.5165024786038248e-07,
"loss": 2.411691188812256,
"step": 977
},
{
"epoch": 0.8504347826086956,
"grad_norm": 90.27301788330078,
"learning_rate": 1.499379541177269e-07,
"loss": 1.5884079933166504,
"step": 978
},
{
"epoch": 0.851304347826087,
"grad_norm": 59.322750091552734,
"learning_rate": 1.4823476519477327e-07,
"loss": 0.32314616441726685,
"step": 979
},
{
"epoch": 0.8521739130434782,
"grad_norm": 17.40428352355957,
"learning_rate": 1.465406951882066e-07,
"loss": 0.13642948865890503,
"step": 980
},
{
"epoch": 0.8530434782608696,
"grad_norm": 97.4061279296875,
"learning_rate": 1.4485575811923795e-07,
"loss": 0.21890735626220703,
"step": 981
},
{
"epoch": 0.8539130434782609,
"grad_norm": 79.38174438476562,
"learning_rate": 1.4317996793348778e-07,
"loss": 0.5916029214859009,
"step": 982
},
{
"epoch": 0.8547826086956521,
"grad_norm": 31.276325225830078,
"learning_rate": 1.4151333850087156e-07,
"loss": 0.335168719291687,
"step": 983
},
{
"epoch": 0.8556521739130435,
"grad_norm": 41.58761215209961,
"learning_rate": 1.3985588361548493e-07,
"loss": 0.20474891364574432,
"step": 984
},
{
"epoch": 0.8565217391304348,
"grad_norm": 295.994384765625,
"learning_rate": 1.3820761699548783e-07,
"loss": 1.3919013738632202,
"step": 985
},
{
"epoch": 0.8573913043478261,
"grad_norm": 128.51441955566406,
"learning_rate": 1.3656855228299334e-07,
"loss": 0.8177456259727478,
"step": 986
},
{
"epoch": 0.8582608695652174,
"grad_norm": 31.59058380126953,
"learning_rate": 1.349387030439532e-07,
"loss": 0.2200441062450409,
"step": 987
},
{
"epoch": 0.8591304347826086,
"grad_norm": 27.698577880859375,
"learning_rate": 1.3331808276804637e-07,
"loss": 0.2795862853527069,
"step": 988
},
{
"epoch": 0.86,
"grad_norm": 25.443546295166016,
"learning_rate": 1.3170670486856594e-07,
"loss": 0.1198812872171402,
"step": 989
},
{
"epoch": 0.8608695652173913,
"grad_norm": 44.49654769897461,
"learning_rate": 1.3010458268231024e-07,
"loss": 0.16760699450969696,
"step": 990
},
{
"epoch": 0.8617391304347826,
"grad_norm": 128.61355590820312,
"learning_rate": 1.2851172946947114e-07,
"loss": 0.3723005950450897,
"step": 991
},
{
"epoch": 0.8626086956521739,
"grad_norm": 50.373104095458984,
"learning_rate": 1.269281584135243e-07,
"loss": 0.23255759477615356,
"step": 992
},
{
"epoch": 0.8634782608695653,
"grad_norm": 48.41141891479492,
"learning_rate": 1.2535388262112023e-07,
"loss": 0.2823089063167572,
"step": 993
},
{
"epoch": 0.8643478260869565,
"grad_norm": 249.34901428222656,
"learning_rate": 1.2378891512197622e-07,
"loss": 2.868992567062378,
"step": 994
},
{
"epoch": 0.8652173913043478,
"grad_norm": 63.32387161254883,
"learning_rate": 1.2223326886876802e-07,
"loss": 0.5655359625816345,
"step": 995
},
{
"epoch": 0.8660869565217392,
"grad_norm": 31.75794792175293,
"learning_rate": 1.2068695673702237e-07,
"loss": 0.165120929479599,
"step": 996
},
{
"epoch": 0.8669565217391304,
"grad_norm": 32.65224838256836,
"learning_rate": 1.1914999152501106e-07,
"loss": 1.9765350818634033,
"step": 997
},
{
"epoch": 0.8678260869565217,
"grad_norm": 19.945507049560547,
"learning_rate": 1.1762238595364573e-07,
"loss": 0.15097635984420776,
"step": 998
},
{
"epoch": 0.8686956521739131,
"grad_norm": 63.010711669921875,
"learning_rate": 1.1610415266637005e-07,
"loss": 0.45428237318992615,
"step": 999
},
{
"epoch": 0.8695652173913043,
"grad_norm": 1365.6044921875,
"learning_rate": 1.1459530422905749e-07,
"loss": 0.8992246985435486,
"step": 1000
},
{
"epoch": 0.8704347826086957,
"grad_norm": 72.2736587524414,
"learning_rate": 1.1309585312990659e-07,
"loss": 0.4428975582122803,
"step": 1001
},
{
"epoch": 0.871304347826087,
"grad_norm": 199.09161376953125,
"learning_rate": 1.1160581177933686e-07,
"loss": 0.5998109579086304,
"step": 1002
},
{
"epoch": 0.8721739130434782,
"grad_norm": 59.35085678100586,
"learning_rate": 1.1012519250988743e-07,
"loss": 0.3049287796020508,
"step": 1003
},
{
"epoch": 0.8730434782608696,
"grad_norm": 67.05830383300781,
"learning_rate": 1.086540075761136e-07,
"loss": 0.5077136158943176,
"step": 1004
},
{
"epoch": 0.8739130434782608,
"grad_norm": 90.06703186035156,
"learning_rate": 1.0719226915448688e-07,
"loss": 0.5149082541465759,
"step": 1005
},
{
"epoch": 0.8747826086956522,
"grad_norm": 14.814482688903809,
"learning_rate": 1.0573998934329214e-07,
"loss": 0.17059670388698578,
"step": 1006
},
{
"epoch": 0.8756521739130435,
"grad_norm": 81.87882232666016,
"learning_rate": 1.0429718016252977e-07,
"loss": 0.22996926307678223,
"step": 1007
},
{
"epoch": 0.8765217391304347,
"grad_norm": 24.417051315307617,
"learning_rate": 1.028638535538147e-07,
"loss": 0.13168440759181976,
"step": 1008
},
{
"epoch": 0.8773913043478261,
"grad_norm": 39.086612701416016,
"learning_rate": 1.0144002138027854e-07,
"loss": 0.10339369624853134,
"step": 1009
},
{
"epoch": 0.8782608695652174,
"grad_norm": 148.6310577392578,
"learning_rate": 1.0002569542646972e-07,
"loss": 0.3061930537223816,
"step": 1010
},
{
"epoch": 0.8791304347826087,
"grad_norm": 24.111289978027344,
"learning_rate": 9.862088739825832e-08,
"loss": 0.15695056319236755,
"step": 1011
},
{
"epoch": 0.88,
"grad_norm": 326.5613098144531,
"learning_rate": 9.722560892273719e-08,
"loss": 0.9801582098007202,
"step": 1012
},
{
"epoch": 0.8808695652173913,
"grad_norm": 105.73284149169922,
"learning_rate": 9.58398715481268e-08,
"loss": 1.9168795347213745,
"step": 1013
},
{
"epoch": 0.8817391304347826,
"grad_norm": 26.056100845336914,
"learning_rate": 9.446368674367906e-08,
"loss": 0.12567181885242462,
"step": 1014
},
{
"epoch": 0.8826086956521739,
"grad_norm": 61.44002151489258,
"learning_rate": 9.30970658995832e-08,
"loss": 0.5126979351043701,
"step": 1015
},
{
"epoch": 0.8834782608695653,
"grad_norm": 223.68104553222656,
"learning_rate": 9.174002032687063e-08,
"loss": 0.5104771256446838,
"step": 1016
},
{
"epoch": 0.8843478260869565,
"grad_norm": 65.98806762695312,
"learning_rate": 9.039256125732143e-08,
"loss": 0.7175702452659607,
"step": 1017
},
{
"epoch": 0.8852173913043478,
"grad_norm": 133.05563354492188,
"learning_rate": 8.905469984337209e-08,
"loss": 1.0224708318710327,
"step": 1018
},
{
"epoch": 0.8860869565217391,
"grad_norm": 17.997692108154297,
"learning_rate": 8.772644715802275e-08,
"loss": 0.14961503446102142,
"step": 1019
},
{
"epoch": 0.8869565217391304,
"grad_norm": 153.3426513671875,
"learning_rate": 8.640781419474484e-08,
"loss": 1.0152846574783325,
"step": 1020
},
{
"epoch": 0.8878260869565218,
"grad_norm": 48.524696350097656,
"learning_rate": 8.509881186739149e-08,
"loss": 0.16481637954711914,
"step": 1021
},
{
"epoch": 0.888695652173913,
"grad_norm": 264.16156005859375,
"learning_rate": 8.379945101010644e-08,
"loss": 0.8960983753204346,
"step": 1022
},
{
"epoch": 0.8895652173913043,
"grad_norm": 181.564453125,
"learning_rate": 8.250974237723433e-08,
"loss": 0.5841391682624817,
"step": 1023
},
{
"epoch": 0.8904347826086957,
"grad_norm": 86.51627349853516,
"learning_rate": 8.122969664323161e-08,
"loss": 0.37853074073791504,
"step": 1024
},
{
"epoch": 0.8913043478260869,
"grad_norm": 28.679731369018555,
"learning_rate": 7.995932440257842e-08,
"loss": 0.22333145141601562,
"step": 1025
},
{
"epoch": 0.8921739130434783,
"grad_norm": 57.46906280517578,
"learning_rate": 7.86986361696912e-08,
"loss": 0.2742474675178528,
"step": 1026
},
{
"epoch": 0.8930434782608696,
"grad_norm": 16.50856590270996,
"learning_rate": 7.74476423788352e-08,
"loss": 0.09008025377988815,
"step": 1027
},
{
"epoch": 0.8939130434782608,
"grad_norm": 56.63728713989258,
"learning_rate": 7.620635338403779e-08,
"loss": 0.29144954681396484,
"step": 1028
},
{
"epoch": 0.8947826086956522,
"grad_norm": 97.25035858154297,
"learning_rate": 7.497477945900375e-08,
"loss": 0.47765645384788513,
"step": 1029
},
{
"epoch": 0.8956521739130435,
"grad_norm": 33.8045768737793,
"learning_rate": 7.375293079702958e-08,
"loss": 0.15769723057746887,
"step": 1030
},
{
"epoch": 0.8965217391304348,
"grad_norm": 204.56980895996094,
"learning_rate": 7.254081751091893e-08,
"loss": 1.3275394439697266,
"step": 1031
},
{
"epoch": 0.8973913043478261,
"grad_norm": 20.713909149169922,
"learning_rate": 7.133844963289976e-08,
"loss": 0.1542229950428009,
"step": 1032
},
{
"epoch": 0.8982608695652174,
"grad_norm": 34.86420822143555,
"learning_rate": 7.014583711454054e-08,
"loss": 0.3445169925689697,
"step": 1033
},
{
"epoch": 0.8991304347826087,
"grad_norm": 43.30035400390625,
"learning_rate": 6.896298982666791e-08,
"loss": 0.206044003367424,
"step": 1034
},
{
"epoch": 0.9,
"grad_norm": 36.75419616699219,
"learning_rate": 6.778991755928555e-08,
"loss": 0.2290279120206833,
"step": 1035
},
{
"epoch": 0.9008695652173913,
"grad_norm": 139.7773895263672,
"learning_rate": 6.662663002149267e-08,
"loss": 0.32020455598831177,
"step": 1036
},
{
"epoch": 0.9017391304347826,
"grad_norm": 715.498046875,
"learning_rate": 6.547313684140342e-08,
"loss": 2.748922109603882,
"step": 1037
},
{
"epoch": 0.9026086956521739,
"grad_norm": 54.58631896972656,
"learning_rate": 6.432944756606815e-08,
"loss": 1.2262675762176514,
"step": 1038
},
{
"epoch": 0.9034782608695652,
"grad_norm": 63.51945114135742,
"learning_rate": 6.319557166139329e-08,
"loss": 0.33816516399383545,
"step": 1039
},
{
"epoch": 0.9043478260869565,
"grad_norm": 143.623779296875,
"learning_rate": 6.20715185120642e-08,
"loss": 0.2825736999511719,
"step": 1040
},
{
"epoch": 0.9052173913043479,
"grad_norm": 38.613399505615234,
"learning_rate": 6.095729742146575e-08,
"loss": 1.2623558044433594,
"step": 1041
},
{
"epoch": 0.9060869565217391,
"grad_norm": 268.6465148925781,
"learning_rate": 5.985291761160736e-08,
"loss": 0.4511108994483948,
"step": 1042
},
{
"epoch": 0.9069565217391304,
"grad_norm": 27.145917892456055,
"learning_rate": 5.875838822304525e-08,
"loss": 0.3145505487918854,
"step": 1043
},
{
"epoch": 0.9078260869565218,
"grad_norm": 28.057687759399414,
"learning_rate": 5.767371831480739e-08,
"loss": 0.20480278134346008,
"step": 1044
},
{
"epoch": 0.908695652173913,
"grad_norm": 61.51593780517578,
"learning_rate": 5.659891686431801e-08,
"loss": 1.557562232017517,
"step": 1045
},
{
"epoch": 0.9095652173913044,
"grad_norm": 24.377716064453125,
"learning_rate": 5.553399276732402e-08,
"loss": 0.12674099206924438,
"step": 1046
},
{
"epoch": 0.9104347826086957,
"grad_norm": 214.2312469482422,
"learning_rate": 5.44789548378212e-08,
"loss": 2.065974712371826,
"step": 1047
},
{
"epoch": 0.9113043478260869,
"grad_norm": 66.00595092773438,
"learning_rate": 5.343381180798038e-08,
"loss": 0.231009840965271,
"step": 1048
},
{
"epoch": 0.9121739130434783,
"grad_norm": 41.069725036621094,
"learning_rate": 5.2398572328076217e-08,
"loss": 1.4642194509506226,
"step": 1049
},
{
"epoch": 0.9130434782608695,
"grad_norm": 24.79140853881836,
"learning_rate": 5.137324496641546e-08,
"loss": 0.17041954398155212,
"step": 1050
},
{
"epoch": 0.9139130434782609,
"grad_norm": 70.95574188232422,
"learning_rate": 5.035783820926565e-08,
"loss": 0.42414528131484985,
"step": 1051
},
{
"epoch": 0.9147826086956522,
"grad_norm": 57.33642578125,
"learning_rate": 4.9352360460784705e-08,
"loss": 1.2034730911254883,
"step": 1052
},
{
"epoch": 0.9156521739130434,
"grad_norm": 35.925148010253906,
"learning_rate": 4.835682004295197e-08,
"loss": 0.23591327667236328,
"step": 1053
},
{
"epoch": 0.9165217391304348,
"grad_norm": 108.28092193603516,
"learning_rate": 4.7371225195499416e-08,
"loss": 0.3866674602031708,
"step": 1054
},
{
"epoch": 0.9173913043478261,
"grad_norm": 62.49339294433594,
"learning_rate": 4.6395584075842325e-08,
"loss": 0.27010294795036316,
"step": 1055
},
{
"epoch": 0.9182608695652174,
"grad_norm": 18.06407356262207,
"learning_rate": 4.5429904759013154e-08,
"loss": 0.1015242338180542,
"step": 1056
},
{
"epoch": 0.9191304347826087,
"grad_norm": 63.693092346191406,
"learning_rate": 4.447419523759364e-08,
"loss": 0.4283607006072998,
"step": 1057
},
{
"epoch": 0.92,
"grad_norm": 208.6317901611328,
"learning_rate": 4.352846342164985e-08,
"loss": 0.2625974714756012,
"step": 1058
},
{
"epoch": 0.9208695652173913,
"grad_norm": 56.89769744873047,
"learning_rate": 4.2592717138664755e-08,
"loss": 0.3733076751232147,
"step": 1059
},
{
"epoch": 0.9217391304347826,
"grad_norm": 27.7113037109375,
"learning_rate": 4.1666964133475624e-08,
"loss": 0.2847471833229065,
"step": 1060
},
{
"epoch": 0.922608695652174,
"grad_norm": 107.4829330444336,
"learning_rate": 4.075121206820867e-08,
"loss": 1.5052435398101807,
"step": 1061
},
{
"epoch": 0.9234782608695652,
"grad_norm": 46.84351348876953,
"learning_rate": 3.984546852221605e-08,
"loss": 0.5343829989433289,
"step": 1062
},
{
"epoch": 0.9243478260869565,
"grad_norm": 47.403709411621094,
"learning_rate": 3.89497409920124e-08,
"loss": 0.23441624641418457,
"step": 1063
},
{
"epoch": 0.9252173913043479,
"grad_norm": 64.3302001953125,
"learning_rate": 3.806403689121399e-08,
"loss": 0.36695361137390137,
"step": 1064
},
{
"epoch": 0.9260869565217391,
"grad_norm": 26.435649871826172,
"learning_rate": 3.7188363550476616e-08,
"loss": 0.21069590747356415,
"step": 1065
},
{
"epoch": 0.9269565217391305,
"grad_norm": 27.971710205078125,
"learning_rate": 3.6322728217434985e-08,
"loss": 0.1252860426902771,
"step": 1066
},
{
"epoch": 0.9278260869565217,
"grad_norm": 27.180171966552734,
"learning_rate": 3.546713805664265e-08,
"loss": 0.12937042117118835,
"step": 1067
},
{
"epoch": 0.928695652173913,
"grad_norm": 75.67200469970703,
"learning_rate": 3.462160014951352e-08,
"loss": 0.5296516418457031,
"step": 1068
},
{
"epoch": 0.9295652173913044,
"grad_norm": 59.014827728271484,
"learning_rate": 3.3786121494261557e-08,
"loss": 1.271408200263977,
"step": 1069
},
{
"epoch": 0.9304347826086956,
"grad_norm": 30.555908203125,
"learning_rate": 3.296070900584466e-08,
"loss": 0.1608971804380417,
"step": 1070
},
{
"epoch": 0.931304347826087,
"grad_norm": 195.51910400390625,
"learning_rate": 3.2145369515906714e-08,
"loss": 1.3339788913726807,
"step": 1071
},
{
"epoch": 0.9321739130434783,
"grad_norm": 242.42306518554688,
"learning_rate": 3.134010977272048e-08,
"loss": 0.43515992164611816,
"step": 1072
},
{
"epoch": 0.9330434782608695,
"grad_norm": 191.44927978515625,
"learning_rate": 3.054493644113268e-08,
"loss": 0.41988927125930786,
"step": 1073
},
{
"epoch": 0.9339130434782609,
"grad_norm": 80.3126449584961,
"learning_rate": 2.975985610250823e-08,
"loss": 0.24759146571159363,
"step": 1074
},
{
"epoch": 0.9347826086956522,
"grad_norm": 427.2878723144531,
"learning_rate": 2.8984875254676054e-08,
"loss": 3.118825912475586,
"step": 1075
},
{
"epoch": 0.9356521739130435,
"grad_norm": 46.329322814941406,
"learning_rate": 2.8220000311875163e-08,
"loss": 0.7188136577606201,
"step": 1076
},
{
"epoch": 0.9365217391304348,
"grad_norm": 18.084016799926758,
"learning_rate": 2.746523760470146e-08,
"loss": 0.34340792894363403,
"step": 1077
},
{
"epoch": 0.9373913043478261,
"grad_norm": 45.552825927734375,
"learning_rate": 2.6720593380055866e-08,
"loss": 0.22152048349380493,
"step": 1078
},
{
"epoch": 0.9382608695652174,
"grad_norm": 116.45755004882812,
"learning_rate": 2.598607380109211e-08,
"loss": 0.4861157536506653,
"step": 1079
},
{
"epoch": 0.9391304347826087,
"grad_norm": 49.15678024291992,
"learning_rate": 2.526168494716541e-08,
"loss": 1.147416591644287,
"step": 1080
},
{
"epoch": 0.94,
"grad_norm": 192.302734375,
"learning_rate": 2.454743281378319e-08,
"loss": 1.1673444509506226,
"step": 1081
},
{
"epoch": 0.9408695652173913,
"grad_norm": 106.5459976196289,
"learning_rate": 2.3843323312554705e-08,
"loss": 0.3712887167930603,
"step": 1082
},
{
"epoch": 0.9417391304347826,
"grad_norm": 60.91425323486328,
"learning_rate": 2.3149362271141924e-08,
"loss": 0.35229623317718506,
"step": 1083
},
{
"epoch": 0.9426086956521739,
"grad_norm": 34.21283721923828,
"learning_rate": 2.246555543321219e-08,
"loss": 0.1597195863723755,
"step": 1084
},
{
"epoch": 0.9434782608695652,
"grad_norm": 262.74908447265625,
"learning_rate": 2.179190845838966e-08,
"loss": 1.804097056388855,
"step": 1085
},
{
"epoch": 0.9443478260869566,
"grad_norm": 289.5881042480469,
"learning_rate": 2.112842692220951e-08,
"loss": 1.9632463455200195,
"step": 1086
},
{
"epoch": 0.9452173913043478,
"grad_norm": 239.2292022705078,
"learning_rate": 2.0475116316070603e-08,
"loss": 0.6926017999649048,
"step": 1087
},
{
"epoch": 0.9460869565217391,
"grad_norm": 54.8002815246582,
"learning_rate": 1.9831982047190806e-08,
"loss": 0.24581407010555267,
"step": 1088
},
{
"epoch": 0.9469565217391305,
"grad_norm": 31.117341995239258,
"learning_rate": 1.9199029438562583e-08,
"loss": 0.12802839279174805,
"step": 1089
},
{
"epoch": 0.9478260869565217,
"grad_norm": 237.68649291992188,
"learning_rate": 1.8576263728907752e-08,
"loss": 1.619250774383545,
"step": 1090
},
{
"epoch": 0.948695652173913,
"grad_norm": 44.76021957397461,
"learning_rate": 1.7963690072635294e-08,
"loss": 0.4234163761138916,
"step": 1091
},
{
"epoch": 0.9495652173913044,
"grad_norm": 23.631999969482422,
"learning_rate": 1.736131353979792e-08,
"loss": 0.10550929605960846,
"step": 1092
},
{
"epoch": 0.9504347826086956,
"grad_norm": 37.710105895996094,
"learning_rate": 1.6769139116050575e-08,
"loss": 0.1747075319290161,
"step": 1093
},
{
"epoch": 0.951304347826087,
"grad_norm": 69.56475067138672,
"learning_rate": 1.6187171702608946e-08,
"loss": 0.10597416013479233,
"step": 1094
},
{
"epoch": 0.9521739130434783,
"grad_norm": 63.06141662597656,
"learning_rate": 1.5615416116208925e-08,
"loss": 0.6295781135559082,
"step": 1095
},
{
"epoch": 0.9530434782608695,
"grad_norm": 79.49552917480469,
"learning_rate": 1.505387708906694e-08,
"loss": 0.5494324564933777,
"step": 1096
},
{
"epoch": 0.9539130434782609,
"grad_norm": 176.24462890625,
"learning_rate": 1.4502559268840394e-08,
"loss": 2.4804582595825195,
"step": 1097
},
{
"epoch": 0.9547826086956521,
"grad_norm": 41.16799545288086,
"learning_rate": 1.3961467218589347e-08,
"loss": 0.268565833568573,
"step": 1098
},
{
"epoch": 0.9556521739130435,
"grad_norm": 49.025325775146484,
"learning_rate": 1.3430605416739212e-08,
"loss": 0.1450543850660324,
"step": 1099
},
{
"epoch": 0.9565217391304348,
"grad_norm": 25.342472076416016,
"learning_rate": 1.2909978257042849e-08,
"loss": 0.10634391009807587,
"step": 1100
},
{
"epoch": 0.957391304347826,
"grad_norm": 57.96340560913086,
"learning_rate": 1.2399590048544908e-08,
"loss": 0.15887653827667236,
"step": 1101
},
{
"epoch": 0.9582608695652174,
"grad_norm": 27.70884132385254,
"learning_rate": 1.1899445015546018e-08,
"loss": 0.20732025802135468,
"step": 1102
},
{
"epoch": 0.9591304347826087,
"grad_norm": 121.35015106201172,
"learning_rate": 1.1409547297567408e-08,
"loss": 0.29734688997268677,
"step": 1103
},
{
"epoch": 0.96,
"grad_norm": 36.666786193847656,
"learning_rate": 1.0929900949317174e-08,
"loss": 0.21187996864318848,
"step": 1104
},
{
"epoch": 0.9608695652173913,
"grad_norm": 27.686845779418945,
"learning_rate": 1.0460509940656282e-08,
"loss": 0.16417719423770905,
"step": 1105
},
{
"epoch": 0.9617391304347827,
"grad_norm": 93.84732818603516,
"learning_rate": 1.0001378156566238e-08,
"loss": 0.2765503525733948,
"step": 1106
},
{
"epoch": 0.9626086956521739,
"grad_norm": 120.61705780029297,
"learning_rate": 9.552509397116328e-09,
"loss": 0.4149501323699951,
"step": 1107
},
{
"epoch": 0.9634782608695652,
"grad_norm": 150.25572204589844,
"learning_rate": 9.113907377432534e-09,
"loss": 0.5060743093490601,
"step": 1108
},
{
"epoch": 0.9643478260869566,
"grad_norm": 22.960622787475586,
"learning_rate": 8.685575727666734e-09,
"loss": 0.13305577635765076,
"step": 1109
},
{
"epoch": 0.9652173913043478,
"grad_norm": 20.679061889648438,
"learning_rate": 8.26751799296685e-09,
"loss": 0.8489625453948975,
"step": 1110
},
{
"epoch": 0.9660869565217391,
"grad_norm": 272.46759033203125,
"learning_rate": 7.859737633446745e-09,
"loss": 1.1429026126861572,
"step": 1111
},
{
"epoch": 0.9669565217391304,
"grad_norm": 40.7579460144043,
"learning_rate": 7.462238024158742e-09,
"loss": 0.35908207297325134,
"step": 1112
},
{
"epoch": 0.9678260869565217,
"grad_norm": 32.05560302734375,
"learning_rate": 7.075022455064612e-09,
"loss": 0.5842143297195435,
"step": 1113
},
{
"epoch": 0.9686956521739131,
"grad_norm": 90.32498931884766,
"learning_rate": 6.6980941310092195e-09,
"loss": 0.8600713014602661,
"step": 1114
},
{
"epoch": 0.9695652173913043,
"grad_norm": 50.42799377441406,
"learning_rate": 6.3314561716930345e-09,
"loss": 0.33498960733413696,
"step": 1115
},
{
"epoch": 0.9704347826086956,
"grad_norm": 89.94180297851562,
"learning_rate": 5.975111611647294e-09,
"loss": 0.18054115772247314,
"step": 1116
},
{
"epoch": 0.971304347826087,
"grad_norm": 25.460975646972656,
"learning_rate": 5.6290634002084676e-09,
"loss": 0.13954642415046692,
"step": 1117
},
{
"epoch": 0.9721739130434782,
"grad_norm": 223.30673217773438,
"learning_rate": 5.293314401493555e-09,
"loss": 0.7651514410972595,
"step": 1118
},
{
"epoch": 0.9730434782608696,
"grad_norm": 30.91139030456543,
"learning_rate": 4.967867394376907e-09,
"loss": 0.17856208980083466,
"step": 1119
},
{
"epoch": 0.9739130434782609,
"grad_norm": 390.0420227050781,
"learning_rate": 4.652725072466918e-09,
"loss": 0.7921420931816101,
"step": 1120
},
{
"epoch": 0.9747826086956521,
"grad_norm": 43.3977165222168,
"learning_rate": 4.347890044083675e-09,
"loss": 0.7626063823699951,
"step": 1121
},
{
"epoch": 0.9756521739130435,
"grad_norm": 112.77864074707031,
"learning_rate": 4.053364832237589e-09,
"loss": 0.3900560140609741,
"step": 1122
},
{
"epoch": 0.9765217391304348,
"grad_norm": 87.15504455566406,
"learning_rate": 3.769151874608301e-09,
"loss": 2.3130087852478027,
"step": 1123
},
{
"epoch": 0.9773913043478261,
"grad_norm": 134.98789978027344,
"learning_rate": 3.4952535235248386e-09,
"loss": 2.125074863433838,
"step": 1124
},
{
"epoch": 0.9782608695652174,
"grad_norm": 15.371259689331055,
"learning_rate": 3.2316720459457662e-09,
"loss": 0.11161771416664124,
"step": 1125
},
{
"epoch": 0.9791304347826087,
"grad_norm": 36.147560119628906,
"learning_rate": 2.9784096234405922e-09,
"loss": 0.22557353973388672,
"step": 1126
},
{
"epoch": 0.98,
"grad_norm": 34.5765380859375,
"learning_rate": 2.7354683521715887e-09,
"loss": 1.8557195663452148,
"step": 1127
},
{
"epoch": 0.9808695652173913,
"grad_norm": 50.32831573486328,
"learning_rate": 2.5028502428768586e-09,
"loss": 1.8108928203582764,
"step": 1128
},
{
"epoch": 0.9817391304347826,
"grad_norm": 89.60624694824219,
"learning_rate": 2.28055722085313e-09,
"loss": 0.8523838520050049,
"step": 1129
},
{
"epoch": 0.9826086956521739,
"grad_norm": 59.01401138305664,
"learning_rate": 2.0685911259402113e-09,
"loss": 1.187254786491394,
"step": 1130
},
{
"epoch": 0.9834782608695652,
"grad_norm": 33.60834884643555,
"learning_rate": 1.866953712505587e-09,
"loss": 0.1633327603340149,
"step": 1131
},
{
"epoch": 0.9843478260869565,
"grad_norm": 55.18779754638672,
"learning_rate": 1.6756466494298462e-09,
"loss": 1.3394694328308105,
"step": 1132
},
{
"epoch": 0.9852173913043478,
"grad_norm": 22.594966888427734,
"learning_rate": 1.494671520092944e-09,
"loss": 0.11512511968612671,
"step": 1133
},
{
"epoch": 0.9860869565217392,
"grad_norm": 68.56832885742188,
"learning_rate": 1.3240298223615722e-09,
"loss": 0.22050261497497559,
"step": 1134
},
{
"epoch": 0.9869565217391304,
"grad_norm": 109.36688995361328,
"learning_rate": 1.1637229685758367e-09,
"loss": 0.4645639657974243,
"step": 1135
},
{
"epoch": 0.9878260869565217,
"grad_norm": 344.52276611328125,
"learning_rate": 1.0137522855380166e-09,
"loss": 1.3248834609985352,
"step": 1136
},
{
"epoch": 0.9886956521739131,
"grad_norm": 323.62261962890625,
"learning_rate": 8.741190145017397e-10,
"loss": 1.0131908655166626,
"step": 1137
},
{
"epoch": 0.9895652173913043,
"grad_norm": 465.9772644042969,
"learning_rate": 7.448243111619901e-10,
"loss": 1.5983543395996094,
"step": 1138
},
{
"epoch": 0.9904347826086957,
"grad_norm": 142.00509643554688,
"learning_rate": 6.258692456442838e-10,
"loss": 1.8868300914764404,
"step": 1139
},
{
"epoch": 0.991304347826087,
"grad_norm": 16.73135757446289,
"learning_rate": 5.172548024973134e-10,
"loss": 0.11686738580465317,
"step": 1140
},
{
"epoch": 0.9921739130434782,
"grad_norm": 82.9582748413086,
"learning_rate": 4.1898188068392744e-10,
"loss": 0.5375820994377136,
"step": 1141
},
{
"epoch": 0.9930434782608696,
"grad_norm": 134.6139373779297,
"learning_rate": 3.3105129357405285e-10,
"loss": 2.217926502227783,
"step": 1142
},
{
"epoch": 0.9939130434782608,
"grad_norm": 56.488040924072266,
"learning_rate": 2.534637689373398e-10,
"loss": 0.3822355270385742,
"step": 1143
},
{
"epoch": 0.9947826086956522,
"grad_norm": 289.47637939453125,
"learning_rate": 1.8621994893816553e-10,
"loss": 1.1271135807037354,
"step": 1144
},
{
"epoch": 0.9956521739130435,
"grad_norm": 40.066505432128906,
"learning_rate": 1.293203901296669e-10,
"loss": 0.19184362888336182,
"step": 1145
},
{
"epoch": 0.9965217391304347,
"grad_norm": 50.626251220703125,
"learning_rate": 8.276556344888332e-11,
"loss": 1.3221231698989868,
"step": 1146
},
{
"epoch": 0.9973913043478261,
"grad_norm": 36.42313003540039,
"learning_rate": 4.655585421356468e-11,
"loss": 2.6915531158447266,
"step": 1147
},
{
"epoch": 0.9982608695652174,
"grad_norm": 56.5333137512207,
"learning_rate": 2.069156211856327e-11,
"loss": 0.5547555685043335,
"step": 1148
},
{
"epoch": 0.9991304347826087,
"grad_norm": 28.36837387084961,
"learning_rate": 5.172901233196914e-12,
"loss": 0.2727220058441162,
"step": 1149
},
{
"epoch": 1.0,
"grad_norm": 155.49249267578125,
"learning_rate": 0.0,
"loss": 0.28436529636383057,
"step": 1150
}
],
"logging_steps": 1,
"max_steps": 1150,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 575,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}