yt-gui-4b / trainer_state.json
ying18's picture
Upload folder using huggingface_hub
4ee537a 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": 408.9283752441406,
"learning_rate": 4.310344827586207e-08,
"loss": 1.0777758359909058,
"step": 1
},
{
"epoch": 0.0017391304347826088,
"grad_norm": 479.2920837402344,
"learning_rate": 8.620689655172414e-08,
"loss": 1.880948781967163,
"step": 2
},
{
"epoch": 0.0026086956521739132,
"grad_norm": 209.09791564941406,
"learning_rate": 1.2931034482758622e-07,
"loss": 1.7428491115570068,
"step": 3
},
{
"epoch": 0.0034782608695652175,
"grad_norm": 423.4102478027344,
"learning_rate": 1.7241379310344828e-07,
"loss": 3.529658317565918,
"step": 4
},
{
"epoch": 0.004347826086956522,
"grad_norm": 348.4798583984375,
"learning_rate": 2.1551724137931036e-07,
"loss": 1.4600706100463867,
"step": 5
},
{
"epoch": 0.0052173913043478265,
"grad_norm": 440.9241638183594,
"learning_rate": 2.5862068965517245e-07,
"loss": 1.9179795980453491,
"step": 6
},
{
"epoch": 0.00608695652173913,
"grad_norm": 538.0116577148438,
"learning_rate": 3.0172413793103453e-07,
"loss": 1.489146113395691,
"step": 7
},
{
"epoch": 0.006956521739130435,
"grad_norm": 313.45281982421875,
"learning_rate": 3.4482758620689656e-07,
"loss": 3.4452528953552246,
"step": 8
},
{
"epoch": 0.00782608695652174,
"grad_norm": 478.4830322265625,
"learning_rate": 3.8793103448275865e-07,
"loss": 2.1786704063415527,
"step": 9
},
{
"epoch": 0.008695652173913044,
"grad_norm": 434.273681640625,
"learning_rate": 4.3103448275862073e-07,
"loss": 2.80996036529541,
"step": 10
},
{
"epoch": 0.009565217391304347,
"grad_norm": 278.2303771972656,
"learning_rate": 4.7413793103448276e-07,
"loss": 3.2558488845825195,
"step": 11
},
{
"epoch": 0.010434782608695653,
"grad_norm": 547.8084106445312,
"learning_rate": 5.172413793103449e-07,
"loss": 2.853588581085205,
"step": 12
},
{
"epoch": 0.011304347826086957,
"grad_norm": 730.852294921875,
"learning_rate": 5.60344827586207e-07,
"loss": 1.6315467357635498,
"step": 13
},
{
"epoch": 0.01217391304347826,
"grad_norm": 308.340087890625,
"learning_rate": 6.034482758620691e-07,
"loss": 1.8621805906295776,
"step": 14
},
{
"epoch": 0.013043478260869565,
"grad_norm": 137.95724487304688,
"learning_rate": 6.465517241379311e-07,
"loss": 0.9975405931472778,
"step": 15
},
{
"epoch": 0.01391304347826087,
"grad_norm": 303.2461242675781,
"learning_rate": 6.896551724137931e-07,
"loss": 1.908968448638916,
"step": 16
},
{
"epoch": 0.014782608695652174,
"grad_norm": 184.87095642089844,
"learning_rate": 7.327586206896552e-07,
"loss": 4.061191082000732,
"step": 17
},
{
"epoch": 0.01565217391304348,
"grad_norm": 560.7840576171875,
"learning_rate": 7.758620689655173e-07,
"loss": 1.1566259860992432,
"step": 18
},
{
"epoch": 0.01652173913043478,
"grad_norm": 400.3383483886719,
"learning_rate": 8.189655172413794e-07,
"loss": 2.058223247528076,
"step": 19
},
{
"epoch": 0.017391304347826087,
"grad_norm": 135.8885955810547,
"learning_rate": 8.620689655172415e-07,
"loss": 1.6466004848480225,
"step": 20
},
{
"epoch": 0.018260869565217393,
"grad_norm": 472.2660217285156,
"learning_rate": 9.051724137931035e-07,
"loss": 3.454291820526123,
"step": 21
},
{
"epoch": 0.019130434782608695,
"grad_norm": 570.48779296875,
"learning_rate": 9.482758620689655e-07,
"loss": 0.7950913906097412,
"step": 22
},
{
"epoch": 0.02,
"grad_norm": 158.8548583984375,
"learning_rate": 9.913793103448276e-07,
"loss": 2.105966091156006,
"step": 23
},
{
"epoch": 0.020869565217391306,
"grad_norm": 119.38016510009766,
"learning_rate": 1.0344827586206898e-06,
"loss": 3.6958322525024414,
"step": 24
},
{
"epoch": 0.021739130434782608,
"grad_norm": 2187.12890625,
"learning_rate": 1.0775862068965518e-06,
"loss": 1.509929895401001,
"step": 25
},
{
"epoch": 0.022608695652173914,
"grad_norm": 179.49208068847656,
"learning_rate": 1.120689655172414e-06,
"loss": 3.1071619987487793,
"step": 26
},
{
"epoch": 0.023478260869565216,
"grad_norm": 235.8142547607422,
"learning_rate": 1.163793103448276e-06,
"loss": 1.5640977621078491,
"step": 27
},
{
"epoch": 0.02434782608695652,
"grad_norm": 730.990966796875,
"learning_rate": 1.2068965517241381e-06,
"loss": 6.2143449783325195,
"step": 28
},
{
"epoch": 0.025217391304347827,
"grad_norm": 312.9621887207031,
"learning_rate": 1.25e-06,
"loss": 2.45662784576416,
"step": 29
},
{
"epoch": 0.02608695652173913,
"grad_norm": 702.923583984375,
"learning_rate": 1.2931034482758623e-06,
"loss": 2.8522918224334717,
"step": 30
},
{
"epoch": 0.026956521739130435,
"grad_norm": 100.41119384765625,
"learning_rate": 1.336206896551724e-06,
"loss": 1.4748220443725586,
"step": 31
},
{
"epoch": 0.02782608695652174,
"grad_norm": 912.0726318359375,
"learning_rate": 1.3793103448275862e-06,
"loss": 3.5353848934173584,
"step": 32
},
{
"epoch": 0.028695652173913042,
"grad_norm": 218.8749542236328,
"learning_rate": 1.4224137931034484e-06,
"loss": 1.64268958568573,
"step": 33
},
{
"epoch": 0.029565217391304348,
"grad_norm": 125.46963500976562,
"learning_rate": 1.4655172413793104e-06,
"loss": 0.9044767618179321,
"step": 34
},
{
"epoch": 0.030434782608695653,
"grad_norm": 198.825439453125,
"learning_rate": 1.5086206896551726e-06,
"loss": 2.7218778133392334,
"step": 35
},
{
"epoch": 0.03130434782608696,
"grad_norm": 173.73460388183594,
"learning_rate": 1.5517241379310346e-06,
"loss": 1.0479382276535034,
"step": 36
},
{
"epoch": 0.03217391304347826,
"grad_norm": 77.9815902709961,
"learning_rate": 1.5948275862068968e-06,
"loss": 0.9507594108581543,
"step": 37
},
{
"epoch": 0.03304347826086956,
"grad_norm": 57.68743896484375,
"learning_rate": 1.6379310344827587e-06,
"loss": 0.5203785300254822,
"step": 38
},
{
"epoch": 0.03391304347826087,
"grad_norm": 346.4000244140625,
"learning_rate": 1.681034482758621e-06,
"loss": 1.4837923049926758,
"step": 39
},
{
"epoch": 0.034782608695652174,
"grad_norm": 115.84912109375,
"learning_rate": 1.724137931034483e-06,
"loss": 0.48359355330467224,
"step": 40
},
{
"epoch": 0.035652173913043476,
"grad_norm": 254.86424255371094,
"learning_rate": 1.7672413793103449e-06,
"loss": 1.7111670970916748,
"step": 41
},
{
"epoch": 0.036521739130434785,
"grad_norm": 122.12332153320312,
"learning_rate": 1.810344827586207e-06,
"loss": 1.3477897644042969,
"step": 42
},
{
"epoch": 0.03739130434782609,
"grad_norm": 100.55317687988281,
"learning_rate": 1.8534482758620693e-06,
"loss": 0.9901793003082275,
"step": 43
},
{
"epoch": 0.03826086956521739,
"grad_norm": 93.47430419921875,
"learning_rate": 1.896551724137931e-06,
"loss": 0.9722920656204224,
"step": 44
},
{
"epoch": 0.0391304347826087,
"grad_norm": 710.9335327148438,
"learning_rate": 1.9396551724137932e-06,
"loss": 2.0677783489227295,
"step": 45
},
{
"epoch": 0.04,
"grad_norm": 197.51223754882812,
"learning_rate": 1.982758620689655e-06,
"loss": 3.213724136352539,
"step": 46
},
{
"epoch": 0.0408695652173913,
"grad_norm": 369.5987243652344,
"learning_rate": 2.025862068965517e-06,
"loss": 1.9037742614746094,
"step": 47
},
{
"epoch": 0.04173913043478261,
"grad_norm": 462.7992858886719,
"learning_rate": 2.0689655172413796e-06,
"loss": 1.7205874919891357,
"step": 48
},
{
"epoch": 0.042608695652173914,
"grad_norm": 185.6064453125,
"learning_rate": 2.1120689655172416e-06,
"loss": 2.038105010986328,
"step": 49
},
{
"epoch": 0.043478260869565216,
"grad_norm": 495.98736572265625,
"learning_rate": 2.1551724137931035e-06,
"loss": 1.542877197265625,
"step": 50
},
{
"epoch": 0.04434782608695652,
"grad_norm": 174.29568481445312,
"learning_rate": 2.1982758620689655e-06,
"loss": 3.509795904159546,
"step": 51
},
{
"epoch": 0.04521739130434783,
"grad_norm": 88.45124053955078,
"learning_rate": 2.241379310344828e-06,
"loss": 1.086487054824829,
"step": 52
},
{
"epoch": 0.04608695652173913,
"grad_norm": 288.6754455566406,
"learning_rate": 2.28448275862069e-06,
"loss": 1.2725701332092285,
"step": 53
},
{
"epoch": 0.04695652173913043,
"grad_norm": 76.34807586669922,
"learning_rate": 2.327586206896552e-06,
"loss": 0.4325469434261322,
"step": 54
},
{
"epoch": 0.04782608695652174,
"grad_norm": 86.93162536621094,
"learning_rate": 2.370689655172414e-06,
"loss": 0.6717068552970886,
"step": 55
},
{
"epoch": 0.04869565217391304,
"grad_norm": 50.85002136230469,
"learning_rate": 2.4137931034482762e-06,
"loss": 1.7667793035507202,
"step": 56
},
{
"epoch": 0.049565217391304345,
"grad_norm": 707.333740234375,
"learning_rate": 2.4568965517241382e-06,
"loss": 1.9895107746124268,
"step": 57
},
{
"epoch": 0.050434782608695654,
"grad_norm": 211.9180450439453,
"learning_rate": 2.5e-06,
"loss": 1.9432262182235718,
"step": 58
},
{
"epoch": 0.051304347826086956,
"grad_norm": 164.1087646484375,
"learning_rate": 2.499994827098767e-06,
"loss": 1.2299290895462036,
"step": 59
},
{
"epoch": 0.05217391304347826,
"grad_norm": 144.11956787109375,
"learning_rate": 2.4999793084378814e-06,
"loss": 2.0348281860351562,
"step": 60
},
{
"epoch": 0.05304347826086957,
"grad_norm": 305.6357727050781,
"learning_rate": 2.499953444145787e-06,
"loss": 0.9155855178833008,
"step": 61
},
{
"epoch": 0.05391304347826087,
"grad_norm": 214.6992950439453,
"learning_rate": 2.4999172344365515e-06,
"loss": 4.967019557952881,
"step": 62
},
{
"epoch": 0.05478260869565217,
"grad_norm": 125.02455139160156,
"learning_rate": 2.4998706796098705e-06,
"loss": 0.7700686454772949,
"step": 63
},
{
"epoch": 0.05565217391304348,
"grad_norm": 91.49518585205078,
"learning_rate": 2.499813780051062e-06,
"loss": 0.6872555017471313,
"step": 64
},
{
"epoch": 0.05652173913043478,
"grad_norm": 66.5513687133789,
"learning_rate": 2.499746536231063e-06,
"loss": 2.322596311569214,
"step": 65
},
{
"epoch": 0.057391304347826085,
"grad_norm": 184.43043518066406,
"learning_rate": 2.4996689487064263e-06,
"loss": 2.804111957550049,
"step": 66
},
{
"epoch": 0.058260869565217394,
"grad_norm": 168.07064819335938,
"learning_rate": 2.4995810181193166e-06,
"loss": 0.940452516078949,
"step": 67
},
{
"epoch": 0.059130434782608696,
"grad_norm": 199.48654174804688,
"learning_rate": 2.499482745197503e-06,
"loss": 1.4553430080413818,
"step": 68
},
{
"epoch": 0.06,
"grad_norm": 690.7571411132812,
"learning_rate": 2.499374130754356e-06,
"loss": 1.210028886795044,
"step": 69
},
{
"epoch": 0.06086956521739131,
"grad_norm": 782.9773559570312,
"learning_rate": 2.499255175688838e-06,
"loss": 2.8736820220947266,
"step": 70
},
{
"epoch": 0.06173913043478261,
"grad_norm": 187.41598510742188,
"learning_rate": 2.4991258809854985e-06,
"loss": 1.842674732208252,
"step": 71
},
{
"epoch": 0.06260869565217392,
"grad_norm": 178.23484802246094,
"learning_rate": 2.498986247714462e-06,
"loss": 2.2488696575164795,
"step": 72
},
{
"epoch": 0.06347826086956522,
"grad_norm": 146.09564208984375,
"learning_rate": 2.4988362770314244e-06,
"loss": 2.0136961936950684,
"step": 73
},
{
"epoch": 0.06434782608695652,
"grad_norm": 327.2138671875,
"learning_rate": 2.4986759701776387e-06,
"loss": 1.7963201999664307,
"step": 74
},
{
"epoch": 0.06521739130434782,
"grad_norm": 179.66259765625,
"learning_rate": 2.498505328479907e-06,
"loss": 1.2915806770324707,
"step": 75
},
{
"epoch": 0.06608695652173913,
"grad_norm": 96.14694213867188,
"learning_rate": 2.4983243533505704e-06,
"loss": 0.6629434823989868,
"step": 76
},
{
"epoch": 0.06695652173913043,
"grad_norm": 401.1381530761719,
"learning_rate": 2.498133046287495e-06,
"loss": 0.855819582939148,
"step": 77
},
{
"epoch": 0.06782608695652174,
"grad_norm": 46.46095657348633,
"learning_rate": 2.49793140887406e-06,
"loss": 1.2157412767410278,
"step": 78
},
{
"epoch": 0.06869565217391305,
"grad_norm": 225.5874481201172,
"learning_rate": 2.497719442779147e-06,
"loss": 0.7673141360282898,
"step": 79
},
{
"epoch": 0.06956521739130435,
"grad_norm": 221.45071411132812,
"learning_rate": 2.4974971497571233e-06,
"loss": 1.9723515510559082,
"step": 80
},
{
"epoch": 0.07043478260869565,
"grad_norm": 48.89097595214844,
"learning_rate": 2.4972645316478286e-06,
"loss": 0.7122063636779785,
"step": 81
},
{
"epoch": 0.07130434782608695,
"grad_norm": 449.5947570800781,
"learning_rate": 2.49702159037656e-06,
"loss": 1.8776814937591553,
"step": 82
},
{
"epoch": 0.07217391304347825,
"grad_norm": 104.74349212646484,
"learning_rate": 2.4967683279540542e-06,
"loss": 0.721406102180481,
"step": 83
},
{
"epoch": 0.07304347826086957,
"grad_norm": 631.10986328125,
"learning_rate": 2.4965047464764753e-06,
"loss": 1.6470348834991455,
"step": 84
},
{
"epoch": 0.07391304347826087,
"grad_norm": 165.18170166015625,
"learning_rate": 2.496230848125392e-06,
"loss": 2.532701015472412,
"step": 85
},
{
"epoch": 0.07478260869565218,
"grad_norm": 68.72663879394531,
"learning_rate": 2.495946635167763e-06,
"loss": 0.7041295766830444,
"step": 86
},
{
"epoch": 0.07565217391304348,
"grad_norm": 681.0164794921875,
"learning_rate": 2.495652109955917e-06,
"loss": 0.6914689540863037,
"step": 87
},
{
"epoch": 0.07652173913043478,
"grad_norm": 112.05097961425781,
"learning_rate": 2.495347274927533e-06,
"loss": 0.5509334802627563,
"step": 88
},
{
"epoch": 0.07739130434782608,
"grad_norm": 90.69129180908203,
"learning_rate": 2.495032132605623e-06,
"loss": 0.8257150650024414,
"step": 89
},
{
"epoch": 0.0782608695652174,
"grad_norm": 364.2044982910156,
"learning_rate": 2.4947066855985067e-06,
"loss": 2.681523323059082,
"step": 90
},
{
"epoch": 0.0791304347826087,
"grad_norm": 144.56988525390625,
"learning_rate": 2.494370936599792e-06,
"loss": 0.6214126944541931,
"step": 91
},
{
"epoch": 0.08,
"grad_norm": 43.279747009277344,
"learning_rate": 2.4940248883883526e-06,
"loss": 0.7092267274856567,
"step": 92
},
{
"epoch": 0.0808695652173913,
"grad_norm": 199.32327270507812,
"learning_rate": 2.493668543828307e-06,
"loss": 1.6910313367843628,
"step": 93
},
{
"epoch": 0.0817391304347826,
"grad_norm": 143.92161560058594,
"learning_rate": 2.493301905868991e-06,
"loss": 1.143976092338562,
"step": 94
},
{
"epoch": 0.08260869565217391,
"grad_norm": 71.49710083007812,
"learning_rate": 2.4929249775449354e-06,
"loss": 0.6278894543647766,
"step": 95
},
{
"epoch": 0.08347826086956522,
"grad_norm": 53.63858413696289,
"learning_rate": 2.4925377619758416e-06,
"loss": 0.7950785160064697,
"step": 96
},
{
"epoch": 0.08434782608695653,
"grad_norm": 68.49473571777344,
"learning_rate": 2.4921402623665535e-06,
"loss": 0.7202810645103455,
"step": 97
},
{
"epoch": 0.08521739130434783,
"grad_norm": 616.7372436523438,
"learning_rate": 2.491732482007033e-06,
"loss": 1.79442298412323,
"step": 98
},
{
"epoch": 0.08608695652173913,
"grad_norm": 143.07440185546875,
"learning_rate": 2.491314424272334e-06,
"loss": 2.269249439239502,
"step": 99
},
{
"epoch": 0.08695652173913043,
"grad_norm": 217.4616241455078,
"learning_rate": 2.4908860926225676e-06,
"loss": 0.6722813844680786,
"step": 100
},
{
"epoch": 0.08782608695652173,
"grad_norm": 111.15724182128906,
"learning_rate": 2.4904474906028838e-06,
"loss": 0.948624312877655,
"step": 101
},
{
"epoch": 0.08869565217391304,
"grad_norm": 45.984622955322266,
"learning_rate": 2.489998621843434e-06,
"loss": 0.35971760749816895,
"step": 102
},
{
"epoch": 0.08956521739130435,
"grad_norm": 56.90412902832031,
"learning_rate": 2.489539490059344e-06,
"loss": 2.1533353328704834,
"step": 103
},
{
"epoch": 0.09043478260869565,
"grad_norm": 457.8325500488281,
"learning_rate": 2.489070099050683e-06,
"loss": 0.7864513397216797,
"step": 104
},
{
"epoch": 0.09130434782608696,
"grad_norm": 36.61105728149414,
"learning_rate": 2.488590452702433e-06,
"loss": 0.4267352521419525,
"step": 105
},
{
"epoch": 0.09217391304347826,
"grad_norm": 42.19702911376953,
"learning_rate": 2.4881005549844544e-06,
"loss": 0.44679027795791626,
"step": 106
},
{
"epoch": 0.09304347826086956,
"grad_norm": 126.22087860107422,
"learning_rate": 2.4876004099514554e-06,
"loss": 3.5700860023498535,
"step": 107
},
{
"epoch": 0.09391304347826086,
"grad_norm": 29.303131103515625,
"learning_rate": 2.487090021742957e-06,
"loss": 4.308433532714844,
"step": 108
},
{
"epoch": 0.09478260869565218,
"grad_norm": 48.11052322387695,
"learning_rate": 2.486569394583261e-06,
"loss": 0.2773371934890747,
"step": 109
},
{
"epoch": 0.09565217391304348,
"grad_norm": 123.01402282714844,
"learning_rate": 2.4860385327814108e-06,
"loss": 1.3421229124069214,
"step": 110
},
{
"epoch": 0.09652173913043478,
"grad_norm": 108.20648193359375,
"learning_rate": 2.48549744073116e-06,
"loss": 0.6545270681381226,
"step": 111
},
{
"epoch": 0.09739130434782609,
"grad_norm": 49.05990982055664,
"learning_rate": 2.4849461229109335e-06,
"loss": 0.4957168996334076,
"step": 112
},
{
"epoch": 0.09826086956521739,
"grad_norm": 437.2885437011719,
"learning_rate": 2.484384583883791e-06,
"loss": 2.001176595687866,
"step": 113
},
{
"epoch": 0.09913043478260869,
"grad_norm": 203.33876037597656,
"learning_rate": 2.483812828297391e-06,
"loss": 1.96820068359375,
"step": 114
},
{
"epoch": 0.1,
"grad_norm": 296.4373779296875,
"learning_rate": 2.4832308608839495e-06,
"loss": 3.3025102615356445,
"step": 115
},
{
"epoch": 0.10086956521739131,
"grad_norm": 1142.8726806640625,
"learning_rate": 2.4826386864602024e-06,
"loss": 1.7267082929611206,
"step": 116
},
{
"epoch": 0.10173913043478261,
"grad_norm": 76.05111694335938,
"learning_rate": 2.4820363099273647e-06,
"loss": 1.9300884008407593,
"step": 117
},
{
"epoch": 0.10260869565217391,
"grad_norm": 174.35528564453125,
"learning_rate": 2.4814237362710923e-06,
"loss": 2.4974822998046875,
"step": 118
},
{
"epoch": 0.10347826086956521,
"grad_norm": 70.0567626953125,
"learning_rate": 2.480800970561438e-06,
"loss": 1.2597458362579346,
"step": 119
},
{
"epoch": 0.10434782608695652,
"grad_norm": 315.5959167480469,
"learning_rate": 2.480168017952809e-06,
"loss": 1.3150229454040527,
"step": 120
},
{
"epoch": 0.10521739130434783,
"grad_norm": 72.14741516113281,
"learning_rate": 2.47952488368393e-06,
"loss": 0.9447697401046753,
"step": 121
},
{
"epoch": 0.10608695652173913,
"grad_norm": 230.29095458984375,
"learning_rate": 2.478871573077791e-06,
"loss": 1.2176882028579712,
"step": 122
},
{
"epoch": 0.10695652173913044,
"grad_norm": 275.8350524902344,
"learning_rate": 2.4782080915416107e-06,
"loss": 1.6070802211761475,
"step": 123
},
{
"epoch": 0.10782608695652174,
"grad_norm": 74.05403137207031,
"learning_rate": 2.4775344445667883e-06,
"loss": 2.00801944732666,
"step": 124
},
{
"epoch": 0.10869565217391304,
"grad_norm": 516.320068359375,
"learning_rate": 2.4768506377288582e-06,
"loss": 0.9637946486473083,
"step": 125
},
{
"epoch": 0.10956521739130434,
"grad_norm": 55.961585998535156,
"learning_rate": 2.4761566766874457e-06,
"loss": 1.8744444847106934,
"step": 126
},
{
"epoch": 0.11043478260869566,
"grad_norm": 119.298828125,
"learning_rate": 2.475452567186217e-06,
"loss": 0.9833845496177673,
"step": 127
},
{
"epoch": 0.11130434782608696,
"grad_norm": 204.20599365234375,
"learning_rate": 2.474738315052835e-06,
"loss": 1.1487748622894287,
"step": 128
},
{
"epoch": 0.11217391304347826,
"grad_norm": 325.07928466796875,
"learning_rate": 2.474013926198908e-06,
"loss": 0.6357107162475586,
"step": 129
},
{
"epoch": 0.11304347826086956,
"grad_norm": 58.65763854980469,
"learning_rate": 2.4732794066199444e-06,
"loss": 0.43879246711730957,
"step": 130
},
{
"epoch": 0.11391304347826087,
"grad_norm": 61.1912841796875,
"learning_rate": 2.472534762395299e-06,
"loss": 0.40946540236473083,
"step": 131
},
{
"epoch": 0.11478260869565217,
"grad_norm": 145.53366088867188,
"learning_rate": 2.471779999688125e-06,
"loss": 1.1725125312805176,
"step": 132
},
{
"epoch": 0.11565217391304349,
"grad_norm": 1644.3331298828125,
"learning_rate": 2.471015124745324e-06,
"loss": 1.2450734376907349,
"step": 133
},
{
"epoch": 0.11652173913043479,
"grad_norm": 118.49866485595703,
"learning_rate": 2.470240143897492e-06,
"loss": 1.2511043548583984,
"step": 134
},
{
"epoch": 0.11739130434782609,
"grad_norm": 230.3250274658203,
"learning_rate": 2.4694550635588676e-06,
"loss": 1.7083921432495117,
"step": 135
},
{
"epoch": 0.11826086956521739,
"grad_norm": 73.51437377929688,
"learning_rate": 2.4686598902272794e-06,
"loss": 0.6450086832046509,
"step": 136
},
{
"epoch": 0.1191304347826087,
"grad_norm": 64.80231475830078,
"learning_rate": 2.4678546304840934e-06,
"loss": 0.5642832517623901,
"step": 137
},
{
"epoch": 0.12,
"grad_norm": 90.22604370117188,
"learning_rate": 2.4670392909941552e-06,
"loss": 1.7450593709945679,
"step": 138
},
{
"epoch": 0.1208695652173913,
"grad_norm": 57.59726333618164,
"learning_rate": 2.4662138785057387e-06,
"loss": 2.830716848373413,
"step": 139
},
{
"epoch": 0.12173913043478261,
"grad_norm": 64.79215240478516,
"learning_rate": 2.465378399850487e-06,
"loss": 0.5736865997314453,
"step": 140
},
{
"epoch": 0.12260869565217392,
"grad_norm": 403.09857177734375,
"learning_rate": 2.4645328619433575e-06,
"loss": 0.7428206205368042,
"step": 141
},
{
"epoch": 0.12347826086956522,
"grad_norm": 91.38631439208984,
"learning_rate": 2.4636772717825653e-06,
"loss": 0.6577591896057129,
"step": 142
},
{
"epoch": 0.12434782608695652,
"grad_norm": 56.21049118041992,
"learning_rate": 2.4628116364495236e-06,
"loss": 0.4712821841239929,
"step": 143
},
{
"epoch": 0.12521739130434784,
"grad_norm": 59.32276153564453,
"learning_rate": 2.4619359631087863e-06,
"loss": 0.8708121180534363,
"step": 144
},
{
"epoch": 0.12608695652173912,
"grad_norm": 146.76007080078125,
"learning_rate": 2.461050259007988e-06,
"loss": 1.2219767570495605,
"step": 145
},
{
"epoch": 0.12695652173913044,
"grad_norm": 235.25128173828125,
"learning_rate": 2.4601545314777843e-06,
"loss": 1.182314157485962,
"step": 146
},
{
"epoch": 0.12782608695652173,
"grad_norm": 195.0425262451172,
"learning_rate": 2.459248787931792e-06,
"loss": 1.1465587615966797,
"step": 147
},
{
"epoch": 0.12869565217391304,
"grad_norm": 265.4162902832031,
"learning_rate": 2.4583330358665248e-06,
"loss": 1.257580041885376,
"step": 148
},
{
"epoch": 0.12956521739130436,
"grad_norm": 127.14356231689453,
"learning_rate": 2.4574072828613354e-06,
"loss": 0.7291069030761719,
"step": 149
},
{
"epoch": 0.13043478260869565,
"grad_norm": 63.75065231323242,
"learning_rate": 2.4564715365783505e-06,
"loss": 0.3326197564601898,
"step": 150
},
{
"epoch": 0.13130434782608696,
"grad_norm": 271.7769470214844,
"learning_rate": 2.4555258047624064e-06,
"loss": 1.3682246208190918,
"step": 151
},
{
"epoch": 0.13217391304347825,
"grad_norm": 55.215667724609375,
"learning_rate": 2.4545700952409875e-06,
"loss": 0.5347506403923035,
"step": 152
},
{
"epoch": 0.13304347826086957,
"grad_norm": 87.39543151855469,
"learning_rate": 2.453604415924158e-06,
"loss": 0.3665112853050232,
"step": 153
},
{
"epoch": 0.13391304347826086,
"grad_norm": 88.65547180175781,
"learning_rate": 2.452628774804501e-06,
"loss": 0.6229044795036316,
"step": 154
},
{
"epoch": 0.13478260869565217,
"grad_norm": 60.733619689941406,
"learning_rate": 2.4516431799570485e-06,
"loss": 0.48716792464256287,
"step": 155
},
{
"epoch": 0.1356521739130435,
"grad_norm": 50.772193908691406,
"learning_rate": 2.4506476395392157e-06,
"loss": 2.28497314453125,
"step": 156
},
{
"epoch": 0.13652173913043478,
"grad_norm": 151.53134155273438,
"learning_rate": 2.4496421617907347e-06,
"loss": 0.5507988929748535,
"step": 157
},
{
"epoch": 0.1373913043478261,
"grad_norm": 423.7794494628906,
"learning_rate": 2.4486267550335848e-06,
"loss": 1.9924039840698242,
"step": 158
},
{
"epoch": 0.13826086956521738,
"grad_norm": 27.168025970458984,
"learning_rate": 2.447601427671924e-06,
"loss": 0.21292328834533691,
"step": 159
},
{
"epoch": 0.1391304347826087,
"grad_norm": 40.74406814575195,
"learning_rate": 2.44656618819202e-06,
"loss": 0.9094164371490479,
"step": 160
},
{
"epoch": 0.14,
"grad_norm": 487.17852783203125,
"learning_rate": 2.445521045162179e-06,
"loss": 1.524223804473877,
"step": 161
},
{
"epoch": 0.1408695652173913,
"grad_norm": 94.6084213256836,
"learning_rate": 2.4444660072326762e-06,
"loss": 0.6373294591903687,
"step": 162
},
{
"epoch": 0.14173913043478262,
"grad_norm": 48.60924530029297,
"learning_rate": 2.4434010831356824e-06,
"loss": 0.3453695476055145,
"step": 163
},
{
"epoch": 0.1426086956521739,
"grad_norm": 30.039562225341797,
"learning_rate": 2.4423262816851927e-06,
"loss": 0.6582419872283936,
"step": 164
},
{
"epoch": 0.14347826086956522,
"grad_norm": 94.57225036621094,
"learning_rate": 2.441241611776955e-06,
"loss": 1.2551779747009277,
"step": 165
},
{
"epoch": 0.1443478260869565,
"grad_norm": 49.300682067871094,
"learning_rate": 2.4401470823883926e-06,
"loss": 0.36331892013549805,
"step": 166
},
{
"epoch": 0.14521739130434783,
"grad_norm": 72.4591064453125,
"learning_rate": 2.4390427025785345e-06,
"loss": 0.8550442457199097,
"step": 167
},
{
"epoch": 0.14608695652173914,
"grad_norm": 59.477577209472656,
"learning_rate": 2.437928481487936e-06,
"loss": 0.5492538213729858,
"step": 168
},
{
"epoch": 0.14695652173913043,
"grad_norm": 39.80845260620117,
"learning_rate": 2.436804428338607e-06,
"loss": 0.5414822101593018,
"step": 169
},
{
"epoch": 0.14782608695652175,
"grad_norm": 115.1328125,
"learning_rate": 2.4356705524339317e-06,
"loss": 2.7509379386901855,
"step": 170
},
{
"epoch": 0.14869565217391303,
"grad_norm": 1049.0916748046875,
"learning_rate": 2.4345268631585966e-06,
"loss": 2.3620762825012207,
"step": 171
},
{
"epoch": 0.14956521739130435,
"grad_norm": 1196.13671875,
"learning_rate": 2.433373369978508e-06,
"loss": 3.821645498275757,
"step": 172
},
{
"epoch": 0.15043478260869564,
"grad_norm": 57.267784118652344,
"learning_rate": 2.4322100824407148e-06,
"loss": 0.8352653384208679,
"step": 173
},
{
"epoch": 0.15130434782608695,
"grad_norm": 104.67648315429688,
"learning_rate": 2.431037010173332e-06,
"loss": 1.2190207242965698,
"step": 174
},
{
"epoch": 0.15217391304347827,
"grad_norm": 49.657554626464844,
"learning_rate": 2.4298541628854596e-06,
"loss": 0.38449758291244507,
"step": 175
},
{
"epoch": 0.15304347826086956,
"grad_norm": 91.279296875,
"learning_rate": 2.4286615503671005e-06,
"loss": 3.13693904876709,
"step": 176
},
{
"epoch": 0.15391304347826087,
"grad_norm": 988.6467895507812,
"learning_rate": 2.4274591824890815e-06,
"loss": 3.2508978843688965,
"step": 177
},
{
"epoch": 0.15478260869565216,
"grad_norm": 65.45477294921875,
"learning_rate": 2.426247069202971e-06,
"loss": 0.881549060344696,
"step": 178
},
{
"epoch": 0.15565217391304348,
"grad_norm": 51.99641036987305,
"learning_rate": 2.4250252205409962e-06,
"loss": 0.8144845366477966,
"step": 179
},
{
"epoch": 0.1565217391304348,
"grad_norm": 106.11524200439453,
"learning_rate": 2.4237936466159624e-06,
"loss": 1.2375469207763672,
"step": 180
},
{
"epoch": 0.15739130434782608,
"grad_norm": 75.31381225585938,
"learning_rate": 2.422552357621165e-06,
"loss": 0.9956674575805664,
"step": 181
},
{
"epoch": 0.1582608695652174,
"grad_norm": 65.4555892944336,
"learning_rate": 2.421301363830309e-06,
"loss": 0.7255182266235352,
"step": 182
},
{
"epoch": 0.1591304347826087,
"grad_norm": 57.11001205444336,
"learning_rate": 2.4200406755974216e-06,
"loss": 0.6393588781356812,
"step": 183
},
{
"epoch": 0.16,
"grad_norm": 137.87217712402344,
"learning_rate": 2.418770303356769e-06,
"loss": 0.836662769317627,
"step": 184
},
{
"epoch": 0.1608695652173913,
"grad_norm": 74.18389129638672,
"learning_rate": 2.417490257622766e-06,
"loss": 0.896056056022644,
"step": 185
},
{
"epoch": 0.1617391304347826,
"grad_norm": 93.32887268066406,
"learning_rate": 2.4162005489898937e-06,
"loss": 0.886600136756897,
"step": 186
},
{
"epoch": 0.16260869565217392,
"grad_norm": 32.393306732177734,
"learning_rate": 2.4149011881326085e-06,
"loss": 0.4953998327255249,
"step": 187
},
{
"epoch": 0.1634782608695652,
"grad_norm": 660.8528442382812,
"learning_rate": 2.4135921858052556e-06,
"loss": 1.6165666580200195,
"step": 188
},
{
"epoch": 0.16434782608695653,
"grad_norm": 89.49231719970703,
"learning_rate": 2.412273552841978e-06,
"loss": 1.4204996824264526,
"step": 189
},
{
"epoch": 0.16521739130434782,
"grad_norm": 58.8868522644043,
"learning_rate": 2.410945300156628e-06,
"loss": 0.8278652429580688,
"step": 190
},
{
"epoch": 0.16608695652173913,
"grad_norm": 177.73793029785156,
"learning_rate": 2.4096074387426788e-06,
"loss": 1.5063395500183105,
"step": 191
},
{
"epoch": 0.16695652173913045,
"grad_norm": 74.32484436035156,
"learning_rate": 2.40825997967313e-06,
"loss": 0.8496602773666382,
"step": 192
},
{
"epoch": 0.16782608695652174,
"grad_norm": 205.34027099609375,
"learning_rate": 2.406902934100417e-06,
"loss": 1.9253334999084473,
"step": 193
},
{
"epoch": 0.16869565217391305,
"grad_norm": 58.65646743774414,
"learning_rate": 2.405536313256321e-06,
"loss": 0.8478743433952332,
"step": 194
},
{
"epoch": 0.16956521739130434,
"grad_norm": 35.247108459472656,
"learning_rate": 2.4041601284518734e-06,
"loss": 0.5699338912963867,
"step": 195
},
{
"epoch": 0.17043478260869566,
"grad_norm": 39.290225982666016,
"learning_rate": 2.4027743910772634e-06,
"loss": 0.3849886357784271,
"step": 196
},
{
"epoch": 0.17130434782608694,
"grad_norm": 530.9067993164062,
"learning_rate": 2.401379112601742e-06,
"loss": 1.6984246969223022,
"step": 197
},
{
"epoch": 0.17217391304347826,
"grad_norm": 130.42115783691406,
"learning_rate": 2.3999743045735305e-06,
"loss": 1.3066246509552002,
"step": 198
},
{
"epoch": 0.17304347826086958,
"grad_norm": 194.60684204101562,
"learning_rate": 2.3985599786197216e-06,
"loss": 1.220478892326355,
"step": 199
},
{
"epoch": 0.17391304347826086,
"grad_norm": 59.08544921875,
"learning_rate": 2.3971361464461855e-06,
"loss": 3.240159749984741,
"step": 200
},
{
"epoch": 0.17478260869565218,
"grad_norm": 470.42596435546875,
"learning_rate": 2.3957028198374703e-06,
"loss": 3.4826736450195312,
"step": 201
},
{
"epoch": 0.17565217391304347,
"grad_norm": 41.92544937133789,
"learning_rate": 2.394260010656708e-06,
"loss": 0.7275924682617188,
"step": 202
},
{
"epoch": 0.17652173913043478,
"grad_norm": 74.81078338623047,
"learning_rate": 2.3928077308455134e-06,
"loss": 1.0109493732452393,
"step": 203
},
{
"epoch": 0.17739130434782607,
"grad_norm": 104.2004165649414,
"learning_rate": 2.3913459924238865e-06,
"loss": 0.7825707197189331,
"step": 204
},
{
"epoch": 0.1782608695652174,
"grad_norm": 151.63543701171875,
"learning_rate": 2.3898748074901127e-06,
"loss": 1.7156990766525269,
"step": 205
},
{
"epoch": 0.1791304347826087,
"grad_norm": 67.58226776123047,
"learning_rate": 2.3883941882206635e-06,
"loss": 0.598099410533905,
"step": 206
},
{
"epoch": 0.18,
"grad_norm": 58.410621643066406,
"learning_rate": 2.386904146870094e-06,
"loss": 2.046065330505371,
"step": 207
},
{
"epoch": 0.1808695652173913,
"grad_norm": 687.547607421875,
"learning_rate": 2.3854046957709425e-06,
"loss": 5.820431709289551,
"step": 208
},
{
"epoch": 0.1817391304347826,
"grad_norm": 74.79448699951172,
"learning_rate": 2.3838958473336302e-06,
"loss": 1.7740070819854736,
"step": 209
},
{
"epoch": 0.1826086956521739,
"grad_norm": 75.154296875,
"learning_rate": 2.3823776140463545e-06,
"loss": 0.4735368490219116,
"step": 210
},
{
"epoch": 0.18347826086956523,
"grad_norm": 171.7770233154297,
"learning_rate": 2.3808500084749893e-06,
"loss": 2.0269219875335693,
"step": 211
},
{
"epoch": 0.18434782608695652,
"grad_norm": 199.6459503173828,
"learning_rate": 2.379313043262978e-06,
"loss": 1.8065171241760254,
"step": 212
},
{
"epoch": 0.18521739130434783,
"grad_norm": 134.1776885986328,
"learning_rate": 2.3777667311312326e-06,
"loss": 0.736809253692627,
"step": 213
},
{
"epoch": 0.18608695652173912,
"grad_norm": 37.64540100097656,
"learning_rate": 2.376211084878024e-06,
"loss": 0.9797202348709106,
"step": 214
},
{
"epoch": 0.18695652173913044,
"grad_norm": 203.0355224609375,
"learning_rate": 2.37464611737888e-06,
"loss": 0.7420639395713806,
"step": 215
},
{
"epoch": 0.18782608695652173,
"grad_norm": 473.5257568359375,
"learning_rate": 2.3730718415864763e-06,
"loss": 2.3235082626342773,
"step": 216
},
{
"epoch": 0.18869565217391304,
"grad_norm": 99.54174041748047,
"learning_rate": 2.371488270530529e-06,
"loss": 1.4670581817626953,
"step": 217
},
{
"epoch": 0.18956521739130436,
"grad_norm": 22.580745697021484,
"learning_rate": 2.36989541731769e-06,
"loss": 0.1934395730495453,
"step": 218
},
{
"epoch": 0.19043478260869565,
"grad_norm": 234.8290252685547,
"learning_rate": 2.3682932951314343e-06,
"loss": 2.1711068153381348,
"step": 219
},
{
"epoch": 0.19130434782608696,
"grad_norm": 102.78935241699219,
"learning_rate": 2.366681917231954e-06,
"loss": 4.88736629486084,
"step": 220
},
{
"epoch": 0.19217391304347825,
"grad_norm": 291.0909118652344,
"learning_rate": 2.365061296956047e-06,
"loss": 3.1556975841522217,
"step": 221
},
{
"epoch": 0.19304347826086957,
"grad_norm": 152.71365356445312,
"learning_rate": 2.3634314477170066e-06,
"loss": 2.949265480041504,
"step": 222
},
{
"epoch": 0.19391304347826088,
"grad_norm": 360.668701171875,
"learning_rate": 2.3617923830045126e-06,
"loss": 1.2918152809143066,
"step": 223
},
{
"epoch": 0.19478260869565217,
"grad_norm": 50.754844665527344,
"learning_rate": 2.3601441163845157e-06,
"loss": 0.9370805621147156,
"step": 224
},
{
"epoch": 0.1956521739130435,
"grad_norm": 160.21510314941406,
"learning_rate": 2.3584866614991287e-06,
"loss": 0.8767625689506531,
"step": 225
},
{
"epoch": 0.19652173913043477,
"grad_norm": 106.48426818847656,
"learning_rate": 2.3568200320665123e-06,
"loss": 1.4009802341461182,
"step": 226
},
{
"epoch": 0.1973913043478261,
"grad_norm": 20.296323776245117,
"learning_rate": 2.355144241880762e-06,
"loss": 0.35235852003097534,
"step": 227
},
{
"epoch": 0.19826086956521738,
"grad_norm": 41.92637634277344,
"learning_rate": 2.3534593048117936e-06,
"loss": 0.2995027005672455,
"step": 228
},
{
"epoch": 0.1991304347826087,
"grad_norm": 21.888343811035156,
"learning_rate": 2.3517652348052268e-06,
"loss": 0.3009972870349884,
"step": 229
},
{
"epoch": 0.2,
"grad_norm": 75.54927062988281,
"learning_rate": 2.3500620458822734e-06,
"loss": 0.5800594687461853,
"step": 230
},
{
"epoch": 0.2008695652173913,
"grad_norm": 24.868005752563477,
"learning_rate": 2.348349752139618e-06,
"loss": 0.34900957345962524,
"step": 231
},
{
"epoch": 0.20173913043478262,
"grad_norm": 82.75250244140625,
"learning_rate": 2.3466283677493025e-06,
"loss": 1.5277791023254395,
"step": 232
},
{
"epoch": 0.2026086956521739,
"grad_norm": 34.98390579223633,
"learning_rate": 2.34489790695861e-06,
"loss": 0.31576430797576904,
"step": 233
},
{
"epoch": 0.20347826086956522,
"grad_norm": 44.74645233154297,
"learning_rate": 2.3431583840899444e-06,
"loss": 0.7691624164581299,
"step": 234
},
{
"epoch": 0.20434782608695654,
"grad_norm": 255.3861083984375,
"learning_rate": 2.3414098135407137e-06,
"loss": 2.0969600677490234,
"step": 235
},
{
"epoch": 0.20521739130434782,
"grad_norm": 300.41107177734375,
"learning_rate": 2.3396522097832103e-06,
"loss": 1.7521913051605225,
"step": 236
},
{
"epoch": 0.20608695652173914,
"grad_norm": 46.60157775878906,
"learning_rate": 2.337885587364492e-06,
"loss": 0.3463154733181,
"step": 237
},
{
"epoch": 0.20695652173913043,
"grad_norm": 17.11665916442871,
"learning_rate": 2.3361099609062584e-06,
"loss": 0.23372775316238403,
"step": 238
},
{
"epoch": 0.20782608695652174,
"grad_norm": 533.9444580078125,
"learning_rate": 2.3343253451047354e-06,
"loss": 0.9582382440567017,
"step": 239
},
{
"epoch": 0.20869565217391303,
"grad_norm": 319.76348876953125,
"learning_rate": 2.332531754730549e-06,
"loss": 0.8091366291046143,
"step": 240
},
{
"epoch": 0.20956521739130435,
"grad_norm": 55.66841125488281,
"learning_rate": 2.3307292046286035e-06,
"loss": 0.5547176599502563,
"step": 241
},
{
"epoch": 0.21043478260869566,
"grad_norm": 23.923696517944336,
"learning_rate": 2.3289177097179612e-06,
"loss": 0.21241050958633423,
"step": 242
},
{
"epoch": 0.21130434782608695,
"grad_norm": 144.29461669921875,
"learning_rate": 2.3270972849917176e-06,
"loss": 1.2355529069900513,
"step": 243
},
{
"epoch": 0.21217391304347827,
"grad_norm": 169.2731475830078,
"learning_rate": 2.325267945516876e-06,
"loss": 0.8365756273269653,
"step": 244
},
{
"epoch": 0.21304347826086956,
"grad_norm": 379.0030822753906,
"learning_rate": 2.3234297064342243e-06,
"loss": 2.4593677520751953,
"step": 245
},
{
"epoch": 0.21391304347826087,
"grad_norm": 586.454833984375,
"learning_rate": 2.321582582958209e-06,
"loss": 1.582472801208496,
"step": 246
},
{
"epoch": 0.21478260869565216,
"grad_norm": 525.1233520507812,
"learning_rate": 2.3197265903768095e-06,
"loss": 2.1884984970092773,
"step": 247
},
{
"epoch": 0.21565217391304348,
"grad_norm": 135.0796356201172,
"learning_rate": 2.3178617440514128e-06,
"loss": 0.952675461769104,
"step": 248
},
{
"epoch": 0.2165217391304348,
"grad_norm": 56.53429412841797,
"learning_rate": 2.315988059416684e-06,
"loss": 3.0985960960388184,
"step": 249
},
{
"epoch": 0.21739130434782608,
"grad_norm": 273.1423645019531,
"learning_rate": 2.3141055519804393e-06,
"loss": 2.3993539810180664,
"step": 250
},
{
"epoch": 0.2182608695652174,
"grad_norm": 643.1798706054688,
"learning_rate": 2.3122142373235195e-06,
"loss": 1.1015483140945435,
"step": 251
},
{
"epoch": 0.21913043478260869,
"grad_norm": 173.88906860351562,
"learning_rate": 2.3103141310996583e-06,
"loss": 3.3863284587860107,
"step": 252
},
{
"epoch": 0.22,
"grad_norm": 109.98249816894531,
"learning_rate": 2.3084052490353552e-06,
"loss": 2.4739880561828613,
"step": 253
},
{
"epoch": 0.22086956521739132,
"grad_norm": 57.38041305541992,
"learning_rate": 2.3064876069297436e-06,
"loss": 0.7926726341247559,
"step": 254
},
{
"epoch": 0.2217391304347826,
"grad_norm": 31.205821990966797,
"learning_rate": 2.3045612206544603e-06,
"loss": 0.2903473973274231,
"step": 255
},
{
"epoch": 0.22260869565217392,
"grad_norm": 86.18379974365234,
"learning_rate": 2.302626106153515e-06,
"loss": 0.5054085850715637,
"step": 256
},
{
"epoch": 0.2234782608695652,
"grad_norm": 266.5611877441406,
"learning_rate": 2.300682279443157e-06,
"loss": 2.424168109893799,
"step": 257
},
{
"epoch": 0.22434782608695653,
"grad_norm": 285.8017272949219,
"learning_rate": 2.298729756611745e-06,
"loss": 1.507334589958191,
"step": 258
},
{
"epoch": 0.2252173913043478,
"grad_norm": 89.92759704589844,
"learning_rate": 2.296768553819611e-06,
"loss": 2.383544921875,
"step": 259
},
{
"epoch": 0.22608695652173913,
"grad_norm": 191.19302368164062,
"learning_rate": 2.2947986872989286e-06,
"loss": 2.563422679901123,
"step": 260
},
{
"epoch": 0.22695652173913045,
"grad_norm": 39.09922790527344,
"learning_rate": 2.2928201733535767e-06,
"loss": 2.133596420288086,
"step": 261
},
{
"epoch": 0.22782608695652173,
"grad_norm": 21.8812198638916,
"learning_rate": 2.290833028359008e-06,
"loss": 0.2714746296405792,
"step": 262
},
{
"epoch": 0.22869565217391305,
"grad_norm": 1493.7266845703125,
"learning_rate": 2.28883726876211e-06,
"loss": 3.6048572063446045,
"step": 263
},
{
"epoch": 0.22956521739130434,
"grad_norm": 46.37892150878906,
"learning_rate": 2.28683291108107e-06,
"loss": 0.38325613737106323,
"step": 264
},
{
"epoch": 0.23043478260869565,
"grad_norm": 87.6407241821289,
"learning_rate": 2.2848199719052388e-06,
"loss": 2.0260558128356934,
"step": 265
},
{
"epoch": 0.23130434782608697,
"grad_norm": 191.0821075439453,
"learning_rate": 2.2827984678949937e-06,
"loss": 1.9020317792892456,
"step": 266
},
{
"epoch": 0.23217391304347826,
"grad_norm": 41.589229583740234,
"learning_rate": 2.2807684157815998e-06,
"loss": 0.6306068897247314,
"step": 267
},
{
"epoch": 0.23304347826086957,
"grad_norm": 152.72744750976562,
"learning_rate": 2.2787298323670707e-06,
"loss": 0.5510439872741699,
"step": 268
},
{
"epoch": 0.23391304347826086,
"grad_norm": 34.26531982421875,
"learning_rate": 2.276682734524032e-06,
"loss": 0.18592610955238342,
"step": 269
},
{
"epoch": 0.23478260869565218,
"grad_norm": 36.68274688720703,
"learning_rate": 2.27462713919558e-06,
"loss": 0.3657134473323822,
"step": 270
},
{
"epoch": 0.23565217391304347,
"grad_norm": 23.140466690063477,
"learning_rate": 2.2725630633951413e-06,
"loss": 2.244689702987671,
"step": 271
},
{
"epoch": 0.23652173913043478,
"grad_norm": 80.58472442626953,
"learning_rate": 2.2704905242063313e-06,
"loss": 1.7534723281860352,
"step": 272
},
{
"epoch": 0.2373913043478261,
"grad_norm": 20.709491729736328,
"learning_rate": 2.268409538782816e-06,
"loss": 0.21615323424339294,
"step": 273
},
{
"epoch": 0.2382608695652174,
"grad_norm": 45.99274826049805,
"learning_rate": 2.2663201243481656e-06,
"loss": 0.43217071890830994,
"step": 274
},
{
"epoch": 0.2391304347826087,
"grad_norm": 132.41433715820312,
"learning_rate": 2.2642222981957163e-06,
"loss": 1.454113245010376,
"step": 275
},
{
"epoch": 0.24,
"grad_norm": 339.4913330078125,
"learning_rate": 2.2621160776884236e-06,
"loss": 0.7456318736076355,
"step": 276
},
{
"epoch": 0.2408695652173913,
"grad_norm": 331.543212890625,
"learning_rate": 2.260001480258721e-06,
"loss": 0.9685505628585815,
"step": 277
},
{
"epoch": 0.2417391304347826,
"grad_norm": 100.98051452636719,
"learning_rate": 2.2578785234083734e-06,
"loss": 1.7122559547424316,
"step": 278
},
{
"epoch": 0.2426086956521739,
"grad_norm": 64.42782592773438,
"learning_rate": 2.2557472247083356e-06,
"loss": 1.0805375576019287,
"step": 279
},
{
"epoch": 0.24347826086956523,
"grad_norm": 63.56621170043945,
"learning_rate": 2.2536076017986033e-06,
"loss": 0.4447496831417084,
"step": 280
},
{
"epoch": 0.24434782608695652,
"grad_norm": 73.55606079101562,
"learning_rate": 2.251459672388069e-06,
"loss": 0.5761306285858154,
"step": 281
},
{
"epoch": 0.24521739130434783,
"grad_norm": 59.72388458251953,
"learning_rate": 2.249303454254377e-06,
"loss": 0.7934373617172241,
"step": 282
},
{
"epoch": 0.24608695652173912,
"grad_norm": 81.59115600585938,
"learning_rate": 2.2471389652437704e-06,
"loss": 0.4326229691505432,
"step": 283
},
{
"epoch": 0.24695652173913044,
"grad_norm": 135.9369354248047,
"learning_rate": 2.244966223270952e-06,
"loss": 0.8459509611129761,
"step": 284
},
{
"epoch": 0.24782608695652175,
"grad_norm": 62.44116973876953,
"learning_rate": 2.2427852463189274e-06,
"loss": 0.3225611448287964,
"step": 285
},
{
"epoch": 0.24869565217391304,
"grad_norm": 24.50501823425293,
"learning_rate": 2.2405960524388633e-06,
"loss": 0.27419164776802063,
"step": 286
},
{
"epoch": 0.24956521739130436,
"grad_norm": 110.9744644165039,
"learning_rate": 2.238398659749933e-06,
"loss": 0.5947219133377075,
"step": 287
},
{
"epoch": 0.25043478260869567,
"grad_norm": 125.97201538085938,
"learning_rate": 2.2361930864391696e-06,
"loss": 0.7606515288352966,
"step": 288
},
{
"epoch": 0.25130434782608696,
"grad_norm": 44.94565200805664,
"learning_rate": 2.2339793507613128e-06,
"loss": 1.7262614965438843,
"step": 289
},
{
"epoch": 0.25217391304347825,
"grad_norm": 43.218563079833984,
"learning_rate": 2.2317574710386615e-06,
"loss": 0.9977378845214844,
"step": 290
},
{
"epoch": 0.2530434782608696,
"grad_norm": 62.10539627075195,
"learning_rate": 2.2295274656609174e-06,
"loss": 0.7950140237808228,
"step": 291
},
{
"epoch": 0.2539130434782609,
"grad_norm": 88.0399169921875,
"learning_rate": 2.2272893530850375e-06,
"loss": 0.6887070536613464,
"step": 292
},
{
"epoch": 0.25478260869565217,
"grad_norm": 180.0225372314453,
"learning_rate": 2.2250431518350777e-06,
"loss": 1.004729986190796,
"step": 293
},
{
"epoch": 0.25565217391304346,
"grad_norm": 46.612953186035156,
"learning_rate": 2.2227888805020424e-06,
"loss": 0.5626980662345886,
"step": 294
},
{
"epoch": 0.2565217391304348,
"grad_norm": 36.29937744140625,
"learning_rate": 2.2205265577437274e-06,
"loss": 1.8379230499267578,
"step": 295
},
{
"epoch": 0.2573913043478261,
"grad_norm": 90.00869750976562,
"learning_rate": 2.218256202284568e-06,
"loss": 0.6438568830490112,
"step": 296
},
{
"epoch": 0.2582608695652174,
"grad_norm": 220.3472442626953,
"learning_rate": 2.215977832915485e-06,
"loss": 2.487238645553589,
"step": 297
},
{
"epoch": 0.2591304347826087,
"grad_norm": 437.23553466796875,
"learning_rate": 2.2136914684937245e-06,
"loss": 2.046386480331421,
"step": 298
},
{
"epoch": 0.26,
"grad_norm": 32.39080810546875,
"learning_rate": 2.211397127942707e-06,
"loss": 0.8442739248275757,
"step": 299
},
{
"epoch": 0.2608695652173913,
"grad_norm": 96.92252349853516,
"learning_rate": 2.2090948302518676e-06,
"loss": 0.981604278087616,
"step": 300
},
{
"epoch": 0.2617391304347826,
"grad_norm": 25.78602409362793,
"learning_rate": 2.2067845944764997e-06,
"loss": 0.22559808194637299,
"step": 301
},
{
"epoch": 0.26260869565217393,
"grad_norm": 87.19547271728516,
"learning_rate": 2.2044664397375977e-06,
"loss": 0.5117440223693848,
"step": 302
},
{
"epoch": 0.2634782608695652,
"grad_norm": 102.25084686279297,
"learning_rate": 2.202140385221699e-06,
"loss": 0.3494492173194885,
"step": 303
},
{
"epoch": 0.2643478260869565,
"grad_norm": 361.44439697265625,
"learning_rate": 2.1998064501807238e-06,
"loss": 1.724427580833435,
"step": 304
},
{
"epoch": 0.26521739130434785,
"grad_norm": 50.31657409667969,
"learning_rate": 2.1974646539318164e-06,
"loss": 1.3169705867767334,
"step": 305
},
{
"epoch": 0.26608695652173914,
"grad_norm": 68.76226043701172,
"learning_rate": 2.1951150158571862e-06,
"loss": 1.1930396556854248,
"step": 306
},
{
"epoch": 0.2669565217391304,
"grad_norm": 28.221277236938477,
"learning_rate": 2.1927575554039465e-06,
"loss": 0.20072108507156372,
"step": 307
},
{
"epoch": 0.2678260869565217,
"grad_norm": 126.85070037841797,
"learning_rate": 2.1903922920839535e-06,
"loss": 0.5540367960929871,
"step": 308
},
{
"epoch": 0.26869565217391306,
"grad_norm": 63.662803649902344,
"learning_rate": 2.188019245473644e-06,
"loss": 0.37292587757110596,
"step": 309
},
{
"epoch": 0.26956521739130435,
"grad_norm": 132.9088134765625,
"learning_rate": 2.1856384352138767e-06,
"loss": 1.94122314453125,
"step": 310
},
{
"epoch": 0.27043478260869563,
"grad_norm": 63.81268310546875,
"learning_rate": 2.1832498810097643e-06,
"loss": 1.948023796081543,
"step": 311
},
{
"epoch": 0.271304347826087,
"grad_norm": 45.0087890625,
"learning_rate": 2.1808536026305154e-06,
"loss": 0.484003484249115,
"step": 312
},
{
"epoch": 0.27217391304347827,
"grad_norm": 83.67201232910156,
"learning_rate": 2.178449619909268e-06,
"loss": 0.7649593353271484,
"step": 313
},
{
"epoch": 0.27304347826086955,
"grad_norm": 63.45158386230469,
"learning_rate": 2.176037952742927e-06,
"loss": 0.9464139342308044,
"step": 314
},
{
"epoch": 0.27391304347826084,
"grad_norm": 368.7110595703125,
"learning_rate": 2.1736186210919974e-06,
"loss": 0.5560407638549805,
"step": 315
},
{
"epoch": 0.2747826086956522,
"grad_norm": 51.621891021728516,
"learning_rate": 2.171191644980421e-06,
"loss": 1.0907920598983765,
"step": 316
},
{
"epoch": 0.2756521739130435,
"grad_norm": 36.33308029174805,
"learning_rate": 2.168757044495411e-06,
"loss": 0.24775516986846924,
"step": 317
},
{
"epoch": 0.27652173913043476,
"grad_norm": 35.794700622558594,
"learning_rate": 2.166314839787283e-06,
"loss": 0.4381987750530243,
"step": 318
},
{
"epoch": 0.2773913043478261,
"grad_norm": 63.283451080322266,
"learning_rate": 2.1638650510692916e-06,
"loss": 0.7726845741271973,
"step": 319
},
{
"epoch": 0.2782608695652174,
"grad_norm": 43.84602737426758,
"learning_rate": 2.161407698617461e-06,
"loss": 0.2759818732738495,
"step": 320
},
{
"epoch": 0.2791304347826087,
"grad_norm": 248.9683380126953,
"learning_rate": 2.1589428027704174e-06,
"loss": 3.1019601821899414,
"step": 321
},
{
"epoch": 0.28,
"grad_norm": 228.894287109375,
"learning_rate": 2.156470383929221e-06,
"loss": 1.7762932777404785,
"step": 322
},
{
"epoch": 0.2808695652173913,
"grad_norm": 17.12674331665039,
"learning_rate": 2.1539904625571973e-06,
"loss": 1.9585199356079102,
"step": 323
},
{
"epoch": 0.2817391304347826,
"grad_norm": 286.8267822265625,
"learning_rate": 2.151503059179768e-06,
"loss": 2.046537160873413,
"step": 324
},
{
"epoch": 0.2826086956521739,
"grad_norm": 157.6624298095703,
"learning_rate": 2.149008194384281e-06,
"loss": 0.9991042017936707,
"step": 325
},
{
"epoch": 0.28347826086956524,
"grad_norm": 231.44180297851562,
"learning_rate": 2.1465058888198376e-06,
"loss": 0.49778541922569275,
"step": 326
},
{
"epoch": 0.2843478260869565,
"grad_norm": 85.47337341308594,
"learning_rate": 2.1439961631971256e-06,
"loss": 1.76524817943573,
"step": 327
},
{
"epoch": 0.2852173913043478,
"grad_norm": 40.157833099365234,
"learning_rate": 2.1414790382882455e-06,
"loss": 0.41400855779647827,
"step": 328
},
{
"epoch": 0.28608695652173916,
"grad_norm": 45.85874557495117,
"learning_rate": 2.138954534926539e-06,
"loss": 0.6939744353294373,
"step": 329
},
{
"epoch": 0.28695652173913044,
"grad_norm": 45.054229736328125,
"learning_rate": 2.1364226740064166e-06,
"loss": 0.637873113155365,
"step": 330
},
{
"epoch": 0.28782608695652173,
"grad_norm": 49.50555419921875,
"learning_rate": 2.1338834764831847e-06,
"loss": 0.7904736995697021,
"step": 331
},
{
"epoch": 0.288695652173913,
"grad_norm": 345.99200439453125,
"learning_rate": 2.131336963372872e-06,
"loss": 0.8211960792541504,
"step": 332
},
{
"epoch": 0.28956521739130436,
"grad_norm": 289.2395324707031,
"learning_rate": 2.1287831557520554e-06,
"loss": 0.7229791879653931,
"step": 333
},
{
"epoch": 0.29043478260869565,
"grad_norm": 37.53415298461914,
"learning_rate": 2.126222074757687e-06,
"loss": 0.255531907081604,
"step": 334
},
{
"epoch": 0.29130434782608694,
"grad_norm": 36.17793655395508,
"learning_rate": 2.1236537415869167e-06,
"loss": 1.4109452962875366,
"step": 335
},
{
"epoch": 0.2921739130434783,
"grad_norm": 564.267333984375,
"learning_rate": 2.1210781774969193e-06,
"loss": 1.4007601737976074,
"step": 336
},
{
"epoch": 0.29304347826086957,
"grad_norm": 216.04855346679688,
"learning_rate": 2.1184954038047157e-06,
"loss": 1.8424882888793945,
"step": 337
},
{
"epoch": 0.29391304347826086,
"grad_norm": 512.345703125,
"learning_rate": 2.1159054418869994e-06,
"loss": 2.3541903495788574,
"step": 338
},
{
"epoch": 0.29478260869565215,
"grad_norm": 20.033504486083984,
"learning_rate": 2.1133083131799582e-06,
"loss": 0.20643578469753265,
"step": 339
},
{
"epoch": 0.2956521739130435,
"grad_norm": 26.708755493164062,
"learning_rate": 2.1107040391790965e-06,
"loss": 0.33382076025009155,
"step": 340
},
{
"epoch": 0.2965217391304348,
"grad_norm": 44.59811782836914,
"learning_rate": 2.1080926414390575e-06,
"loss": 2.5688414573669434,
"step": 341
},
{
"epoch": 0.29739130434782607,
"grad_norm": 20.14144515991211,
"learning_rate": 2.105474141573446e-06,
"loss": 0.21918675303459167,
"step": 342
},
{
"epoch": 0.2982608695652174,
"grad_norm": 47.1740608215332,
"learning_rate": 2.102848561254647e-06,
"loss": 0.3276013433933258,
"step": 343
},
{
"epoch": 0.2991304347826087,
"grad_norm": 20.712509155273438,
"learning_rate": 2.1002159222136494e-06,
"loss": 0.19047638773918152,
"step": 344
},
{
"epoch": 0.3,
"grad_norm": 141.69659423828125,
"learning_rate": 2.0975762462398637e-06,
"loss": 0.823175847530365,
"step": 345
},
{
"epoch": 0.3008695652173913,
"grad_norm": 29.51601219177246,
"learning_rate": 2.094929555180943e-06,
"loss": 0.3437114357948303,
"step": 346
},
{
"epoch": 0.3017391304347826,
"grad_norm": 303.27880859375,
"learning_rate": 2.0922758709426018e-06,
"loss": 2.6932425498962402,
"step": 347
},
{
"epoch": 0.3026086956521739,
"grad_norm": 57.637489318847656,
"learning_rate": 2.089615215488434e-06,
"loss": 0.5431745052337646,
"step": 348
},
{
"epoch": 0.3034782608695652,
"grad_norm": 44.37310028076172,
"learning_rate": 2.0869476108397324e-06,
"loss": 0.3254171311855316,
"step": 349
},
{
"epoch": 0.30434782608695654,
"grad_norm": 74.85758209228516,
"learning_rate": 2.084273079075306e-06,
"loss": 3.3909125328063965,
"step": 350
},
{
"epoch": 0.30521739130434783,
"grad_norm": 163.68577575683594,
"learning_rate": 2.081591642331296e-06,
"loss": 4.116057872772217,
"step": 351
},
{
"epoch": 0.3060869565217391,
"grad_norm": 149.66868591308594,
"learning_rate": 2.0789033228009943e-06,
"loss": 2.0456604957580566,
"step": 352
},
{
"epoch": 0.30695652173913046,
"grad_norm": 41.50107192993164,
"learning_rate": 2.07620814273466e-06,
"loss": 0.3820236325263977,
"step": 353
},
{
"epoch": 0.30782608695652175,
"grad_norm": 190.1525115966797,
"learning_rate": 2.073506124439332e-06,
"loss": 0.9645884037017822,
"step": 354
},
{
"epoch": 0.30869565217391304,
"grad_norm": 176.3890838623047,
"learning_rate": 2.0707972902786504e-06,
"loss": 1.704902172088623,
"step": 355
},
{
"epoch": 0.3095652173913043,
"grad_norm": 64.47674560546875,
"learning_rate": 2.0680816626726648e-06,
"loss": 0.3840813636779785,
"step": 356
},
{
"epoch": 0.31043478260869567,
"grad_norm": 121.12983703613281,
"learning_rate": 2.0653592640976517e-06,
"loss": 0.5144632458686829,
"step": 357
},
{
"epoch": 0.31130434782608696,
"grad_norm": 121.5889892578125,
"learning_rate": 2.0626301170859305e-06,
"loss": 4.148053169250488,
"step": 358
},
{
"epoch": 0.31217391304347825,
"grad_norm": 152.54835510253906,
"learning_rate": 2.0598942442256736e-06,
"loss": 1.0699222087860107,
"step": 359
},
{
"epoch": 0.3130434782608696,
"grad_norm": 84.40726470947266,
"learning_rate": 2.057151668160721e-06,
"loss": 0.4839213490486145,
"step": 360
},
{
"epoch": 0.3139130434782609,
"grad_norm": 170.79342651367188,
"learning_rate": 2.0544024115903935e-06,
"loss": 0.7821235656738281,
"step": 361
},
{
"epoch": 0.31478260869565217,
"grad_norm": 231.14315795898438,
"learning_rate": 2.0516464972693027e-06,
"loss": 2.7148711681365967,
"step": 362
},
{
"epoch": 0.31565217391304345,
"grad_norm": 77.24949645996094,
"learning_rate": 2.0488839480071653e-06,
"loss": 0.8683456182479858,
"step": 363
},
{
"epoch": 0.3165217391304348,
"grad_norm": 22.853055953979492,
"learning_rate": 2.0461147866686116e-06,
"loss": 1.615559458732605,
"step": 364
},
{
"epoch": 0.3173913043478261,
"grad_norm": 108.22183990478516,
"learning_rate": 2.0433390361729994e-06,
"loss": 0.41780614852905273,
"step": 365
},
{
"epoch": 0.3182608695652174,
"grad_norm": 83.64887237548828,
"learning_rate": 2.0405567194942215e-06,
"loss": 0.9844322204589844,
"step": 366
},
{
"epoch": 0.3191304347826087,
"grad_norm": 169.87159729003906,
"learning_rate": 2.0377678596605173e-06,
"loss": 1.897526502609253,
"step": 367
},
{
"epoch": 0.32,
"grad_norm": 51.68992614746094,
"learning_rate": 2.03497247975428e-06,
"loss": 0.25149208307266235,
"step": 368
},
{
"epoch": 0.3208695652173913,
"grad_norm": 28.302207946777344,
"learning_rate": 2.03217060291187e-06,
"loss": 0.28001725673675537,
"step": 369
},
{
"epoch": 0.3217391304347826,
"grad_norm": 96.13204956054688,
"learning_rate": 2.029362252323417e-06,
"loss": 0.35651934146881104,
"step": 370
},
{
"epoch": 0.3226086956521739,
"grad_norm": 33.033992767333984,
"learning_rate": 2.0265474512326345e-06,
"loss": 1.3933261632919312,
"step": 371
},
{
"epoch": 0.3234782608695652,
"grad_norm": 46.49980926513672,
"learning_rate": 2.023726222936623e-06,
"loss": 0.4730550944805145,
"step": 372
},
{
"epoch": 0.3243478260869565,
"grad_norm": 42.75638198852539,
"learning_rate": 2.020898590785679e-06,
"loss": 0.47161561250686646,
"step": 373
},
{
"epoch": 0.32521739130434785,
"grad_norm": 53.752933502197266,
"learning_rate": 2.0180645781831017e-06,
"loss": 0.6370646953582764,
"step": 374
},
{
"epoch": 0.32608695652173914,
"grad_norm": 29.790605545043945,
"learning_rate": 2.015224208584998e-06,
"loss": 0.26960355043411255,
"step": 375
},
{
"epoch": 0.3269565217391304,
"grad_norm": 83.5875473022461,
"learning_rate": 2.0123775055000912e-06,
"loss": 1.126031756401062,
"step": 376
},
{
"epoch": 0.3278260869565217,
"grad_norm": 18.911537170410156,
"learning_rate": 2.009524492489523e-06,
"loss": 0.24288177490234375,
"step": 377
},
{
"epoch": 0.32869565217391306,
"grad_norm": 34.443687438964844,
"learning_rate": 2.00666519316666e-06,
"loss": 0.41903847455978394,
"step": 378
},
{
"epoch": 0.32956521739130434,
"grad_norm": 141.90786743164062,
"learning_rate": 2.0037996311969007e-06,
"loss": 0.7340172529220581,
"step": 379
},
{
"epoch": 0.33043478260869563,
"grad_norm": 26.994577407836914,
"learning_rate": 2.000927830297474e-06,
"loss": 0.30741408467292786,
"step": 380
},
{
"epoch": 0.331304347826087,
"grad_norm": 98.02938079833984,
"learning_rate": 1.9980498142372483e-06,
"loss": 0.8007850050926208,
"step": 381
},
{
"epoch": 0.33217391304347826,
"grad_norm": 149.39999389648438,
"learning_rate": 1.995165606836532e-06,
"loss": 0.7947365641593933,
"step": 382
},
{
"epoch": 0.33304347826086955,
"grad_norm": 174.021240234375,
"learning_rate": 1.992275231966877e-06,
"loss": 1.2335515022277832,
"step": 383
},
{
"epoch": 0.3339130434782609,
"grad_norm": 66.7916030883789,
"learning_rate": 1.989378713550882e-06,
"loss": 0.5394859313964844,
"step": 384
},
{
"epoch": 0.3347826086956522,
"grad_norm": 127.73109436035156,
"learning_rate": 1.986476075561992e-06,
"loss": 0.5928844213485718,
"step": 385
},
{
"epoch": 0.3356521739130435,
"grad_norm": 100.7883071899414,
"learning_rate": 1.9835673420243033e-06,
"loss": 0.709867000579834,
"step": 386
},
{
"epoch": 0.33652173913043476,
"grad_norm": 65.74164581298828,
"learning_rate": 1.980652537012361e-06,
"loss": 1.158227801322937,
"step": 387
},
{
"epoch": 0.3373913043478261,
"grad_norm": 952.95654296875,
"learning_rate": 1.9777316846509638e-06,
"loss": 1.4253524541854858,
"step": 388
},
{
"epoch": 0.3382608695652174,
"grad_norm": 54.963623046875,
"learning_rate": 1.97480480911496e-06,
"loss": 2.205634593963623,
"step": 389
},
{
"epoch": 0.3391304347826087,
"grad_norm": 35.07461166381836,
"learning_rate": 1.97187193462905e-06,
"loss": 0.24558481574058533,
"step": 390
},
{
"epoch": 0.34,
"grad_norm": 30.49419403076172,
"learning_rate": 1.968933085467587e-06,
"loss": 0.5441511869430542,
"step": 391
},
{
"epoch": 0.3408695652173913,
"grad_norm": 60.02323532104492,
"learning_rate": 1.9659882859543733e-06,
"loss": 0.732552707195282,
"step": 392
},
{
"epoch": 0.3417391304347826,
"grad_norm": 78.89702606201172,
"learning_rate": 1.9630375604624596e-06,
"loss": 0.5341556668281555,
"step": 393
},
{
"epoch": 0.3426086956521739,
"grad_norm": 54.77236557006836,
"learning_rate": 1.960080933413945e-06,
"loss": 0.5032380819320679,
"step": 394
},
{
"epoch": 0.34347826086956523,
"grad_norm": 28.26173210144043,
"learning_rate": 1.957118429279772e-06,
"loss": 0.31716644763946533,
"step": 395
},
{
"epoch": 0.3443478260869565,
"grad_norm": 393.0223388671875,
"learning_rate": 1.9541500725795277e-06,
"loss": 2.0188708305358887,
"step": 396
},
{
"epoch": 0.3452173913043478,
"grad_norm": 43.31639862060547,
"learning_rate": 1.951175887881237e-06,
"loss": 0.236699640750885,
"step": 397
},
{
"epoch": 0.34608695652173915,
"grad_norm": 104.89363861083984,
"learning_rate": 1.9481958998011626e-06,
"loss": 3.3953709602355957,
"step": 398
},
{
"epoch": 0.34695652173913044,
"grad_norm": 273.7563781738281,
"learning_rate": 1.945210133003598e-06,
"loss": 0.9807060956954956,
"step": 399
},
{
"epoch": 0.34782608695652173,
"grad_norm": 140.30453491210938,
"learning_rate": 1.9422186122006663e-06,
"loss": 1.3066749572753906,
"step": 400
},
{
"epoch": 0.348695652173913,
"grad_norm": 30.06455421447754,
"learning_rate": 1.939221362152115e-06,
"loss": 0.4493914544582367,
"step": 401
},
{
"epoch": 0.34956521739130436,
"grad_norm": 83.47148132324219,
"learning_rate": 1.9362184076651082e-06,
"loss": 0.6781956553459167,
"step": 402
},
{
"epoch": 0.35043478260869565,
"grad_norm": 26.22336196899414,
"learning_rate": 1.9332097735940265e-06,
"loss": 0.2689898610115051,
"step": 403
},
{
"epoch": 0.35130434782608694,
"grad_norm": 25.663368225097656,
"learning_rate": 1.9301954848402553e-06,
"loss": 0.8483200073242188,
"step": 404
},
{
"epoch": 0.3521739130434783,
"grad_norm": 49.023719787597656,
"learning_rate": 1.9271755663519854e-06,
"loss": 0.41670069098472595,
"step": 405
},
{
"epoch": 0.35304347826086957,
"grad_norm": 42.6588020324707,
"learning_rate": 1.924150043123999e-06,
"loss": 0.41812556982040405,
"step": 406
},
{
"epoch": 0.35391304347826086,
"grad_norm": 43.46489715576172,
"learning_rate": 1.92111894019747e-06,
"loss": 0.3116633892059326,
"step": 407
},
{
"epoch": 0.35478260869565215,
"grad_norm": 50.448482513427734,
"learning_rate": 1.9180822826597516e-06,
"loss": 0.7443313598632812,
"step": 408
},
{
"epoch": 0.3556521739130435,
"grad_norm": 32.91218185424805,
"learning_rate": 1.915040095644171e-06,
"loss": 0.27192604541778564,
"step": 409
},
{
"epoch": 0.3565217391304348,
"grad_norm": 48.53046798706055,
"learning_rate": 1.9119924043298207e-06,
"loss": 0.28298115730285645,
"step": 410
},
{
"epoch": 0.35739130434782607,
"grad_norm": 39.88666534423828,
"learning_rate": 1.9089392339413513e-06,
"loss": 0.4056350588798523,
"step": 411
},
{
"epoch": 0.3582608695652174,
"grad_norm": 78.02606964111328,
"learning_rate": 1.90588060974876e-06,
"loss": 0.8998953700065613,
"step": 412
},
{
"epoch": 0.3591304347826087,
"grad_norm": 49.831180572509766,
"learning_rate": 1.9028165570671852e-06,
"loss": 0.715080201625824,
"step": 413
},
{
"epoch": 0.36,
"grad_norm": 80.9101791381836,
"learning_rate": 1.8997471012566932e-06,
"loss": 0.7034376859664917,
"step": 414
},
{
"epoch": 0.36086956521739133,
"grad_norm": 79.74562072753906,
"learning_rate": 1.896672267722071e-06,
"loss": 2.466092824935913,
"step": 415
},
{
"epoch": 0.3617391304347826,
"grad_norm": 63.741458892822266,
"learning_rate": 1.8935920819126152e-06,
"loss": 0.2804633378982544,
"step": 416
},
{
"epoch": 0.3626086956521739,
"grad_norm": 415.62396240234375,
"learning_rate": 1.8905065693219205e-06,
"loss": 2.430732488632202,
"step": 417
},
{
"epoch": 0.3634782608695652,
"grad_norm": 240.4092559814453,
"learning_rate": 1.8874157554876701e-06,
"loss": 1.0029200315475464,
"step": 418
},
{
"epoch": 0.36434782608695654,
"grad_norm": 104.24790954589844,
"learning_rate": 1.8843196659914235e-06,
"loss": 1.635483741760254,
"step": 419
},
{
"epoch": 0.3652173913043478,
"grad_norm": 14.044332504272461,
"learning_rate": 1.881218326458405e-06,
"loss": 0.2149471938610077,
"step": 420
},
{
"epoch": 0.3660869565217391,
"grad_norm": 31.932836532592773,
"learning_rate": 1.8781117625572915e-06,
"loss": 3.244011878967285,
"step": 421
},
{
"epoch": 0.36695652173913046,
"grad_norm": 111.29270935058594,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.5952858328819275,
"step": 422
},
{
"epoch": 0.36782608695652175,
"grad_norm": 152.03514099121094,
"learning_rate": 1.8718830645414758e-06,
"loss": 1.3452274799346924,
"step": 423
},
{
"epoch": 0.36869565217391304,
"grad_norm": 224.88206481933594,
"learning_rate": 1.868760981979477e-06,
"loss": 1.570754051208496,
"step": 424
},
{
"epoch": 0.3695652173913043,
"grad_norm": 12.70956802368164,
"learning_rate": 1.8656337781543632e-06,
"loss": 0.11816589534282684,
"step": 425
},
{
"epoch": 0.37043478260869567,
"grad_norm": 27.472719192504883,
"learning_rate": 1.8625014789488813e-06,
"loss": 0.3050633370876312,
"step": 426
},
{
"epoch": 0.37130434782608696,
"grad_norm": 43.99699783325195,
"learning_rate": 1.85936411028795e-06,
"loss": 0.7790147662162781,
"step": 427
},
{
"epoch": 0.37217391304347824,
"grad_norm": 17.860076904296875,
"learning_rate": 1.8562216981384467e-06,
"loss": 0.481042742729187,
"step": 428
},
{
"epoch": 0.3730434782608696,
"grad_norm": 69.86940002441406,
"learning_rate": 1.853074268508992e-06,
"loss": 0.6527888774871826,
"step": 429
},
{
"epoch": 0.3739130434782609,
"grad_norm": 33.59516525268555,
"learning_rate": 1.8499218474497332e-06,
"loss": 0.2900844216346741,
"step": 430
},
{
"epoch": 0.37478260869565216,
"grad_norm": 174.12254333496094,
"learning_rate": 1.8467644610521313e-06,
"loss": 1.1463508605957031,
"step": 431
},
{
"epoch": 0.37565217391304345,
"grad_norm": 99.09007263183594,
"learning_rate": 1.8436021354487428e-06,
"loss": 0.6034988164901733,
"step": 432
},
{
"epoch": 0.3765217391304348,
"grad_norm": 25.11251449584961,
"learning_rate": 1.8404348968130048e-06,
"loss": 0.3490484952926636,
"step": 433
},
{
"epoch": 0.3773913043478261,
"grad_norm": 121.11735534667969,
"learning_rate": 1.8372627713590178e-06,
"loss": 1.488389253616333,
"step": 434
},
{
"epoch": 0.3782608695652174,
"grad_norm": 74.66360473632812,
"learning_rate": 1.8340857853413277e-06,
"loss": 0.6942107081413269,
"step": 435
},
{
"epoch": 0.3791304347826087,
"grad_norm": 29.348190307617188,
"learning_rate": 1.830903965054711e-06,
"loss": 0.2921854853630066,
"step": 436
},
{
"epoch": 0.38,
"grad_norm": 49.405059814453125,
"learning_rate": 1.8277173368339546e-06,
"loss": 0.5783102512359619,
"step": 437
},
{
"epoch": 0.3808695652173913,
"grad_norm": 177.93511962890625,
"learning_rate": 1.8245259270536398e-06,
"loss": 1.0955631732940674,
"step": 438
},
{
"epoch": 0.38173913043478264,
"grad_norm": 56.21971893310547,
"learning_rate": 1.8213297621279225e-06,
"loss": 0.8876115679740906,
"step": 439
},
{
"epoch": 0.3826086956521739,
"grad_norm": 34.57631301879883,
"learning_rate": 1.8181288685103148e-06,
"loss": 0.4269735515117645,
"step": 440
},
{
"epoch": 0.3834782608695652,
"grad_norm": 23.442337036132812,
"learning_rate": 1.8149232726934687e-06,
"loss": 0.13463447988033295,
"step": 441
},
{
"epoch": 0.3843478260869565,
"grad_norm": 214.44754028320312,
"learning_rate": 1.8117130012089518e-06,
"loss": 3.299225330352783,
"step": 442
},
{
"epoch": 0.38521739130434784,
"grad_norm": 29.900230407714844,
"learning_rate": 1.8084980806270324e-06,
"loss": 0.3806462287902832,
"step": 443
},
{
"epoch": 0.38608695652173913,
"grad_norm": 33.27299499511719,
"learning_rate": 1.8052785375564569e-06,
"loss": 0.2677356004714966,
"step": 444
},
{
"epoch": 0.3869565217391304,
"grad_norm": 18.99176597595215,
"learning_rate": 1.8020543986442306e-06,
"loss": 0.5586559772491455,
"step": 445
},
{
"epoch": 0.38782608695652177,
"grad_norm": 152.81407165527344,
"learning_rate": 1.7988256905753976e-06,
"loss": 3.0094780921936035,
"step": 446
},
{
"epoch": 0.38869565217391305,
"grad_norm": 57.23750305175781,
"learning_rate": 1.7955924400728174e-06,
"loss": 0.6309852600097656,
"step": 447
},
{
"epoch": 0.38956521739130434,
"grad_norm": 21.0781307220459,
"learning_rate": 1.7923546738969477e-06,
"loss": 0.3555275797843933,
"step": 448
},
{
"epoch": 0.39043478260869563,
"grad_norm": 118.61656951904297,
"learning_rate": 1.78911241884562e-06,
"loss": 2.067031145095825,
"step": 449
},
{
"epoch": 0.391304347826087,
"grad_norm": 139.69496154785156,
"learning_rate": 1.7858657017538177e-06,
"loss": 0.717292070388794,
"step": 450
},
{
"epoch": 0.39217391304347826,
"grad_norm": 64.45001220703125,
"learning_rate": 1.782614549493457e-06,
"loss": 0.27325439453125,
"step": 451
},
{
"epoch": 0.39304347826086955,
"grad_norm": 151.939453125,
"learning_rate": 1.77935898897316e-06,
"loss": 1.0139319896697998,
"step": 452
},
{
"epoch": 0.3939130434782609,
"grad_norm": 47.88231658935547,
"learning_rate": 1.7760990471380364e-06,
"loss": 0.5249274969100952,
"step": 453
},
{
"epoch": 0.3947826086956522,
"grad_norm": 22.23899269104004,
"learning_rate": 1.7728347509694566e-06,
"loss": 2.606963634490967,
"step": 454
},
{
"epoch": 0.39565217391304347,
"grad_norm": 36.164764404296875,
"learning_rate": 1.7695661274848326e-06,
"loss": 0.4062786102294922,
"step": 455
},
{
"epoch": 0.39652173913043476,
"grad_norm": 32.981910705566406,
"learning_rate": 1.7662932037373896e-06,
"loss": 0.39916491508483887,
"step": 456
},
{
"epoch": 0.3973913043478261,
"grad_norm": 102.16033935546875,
"learning_rate": 1.7630160068159463e-06,
"loss": 0.3857983946800232,
"step": 457
},
{
"epoch": 0.3982608695652174,
"grad_norm": 34.86853790283203,
"learning_rate": 1.7597345638446877e-06,
"loss": 0.3358461856842041,
"step": 458
},
{
"epoch": 0.3991304347826087,
"grad_norm": 37.31612777709961,
"learning_rate": 1.7564489019829435e-06,
"loss": 0.4881976544857025,
"step": 459
},
{
"epoch": 0.4,
"grad_norm": 68.10031127929688,
"learning_rate": 1.7531590484249595e-06,
"loss": 0.6506434679031372,
"step": 460
},
{
"epoch": 0.4008695652173913,
"grad_norm": 50.20039749145508,
"learning_rate": 1.749865030399676e-06,
"loss": 0.5034558176994324,
"step": 461
},
{
"epoch": 0.4017391304347826,
"grad_norm": 52.464500427246094,
"learning_rate": 1.7465668751705017e-06,
"loss": 0.20062991976737976,
"step": 462
},
{
"epoch": 0.4026086956521739,
"grad_norm": 146.697265625,
"learning_rate": 1.7432646100350852e-06,
"loss": 0.9521932601928711,
"step": 463
},
{
"epoch": 0.40347826086956523,
"grad_norm": 24.662134170532227,
"learning_rate": 1.739958262325094e-06,
"loss": 0.2981290817260742,
"step": 464
},
{
"epoch": 0.4043478260869565,
"grad_norm": 313.09942626953125,
"learning_rate": 1.7366478594059834e-06,
"loss": 1.9756301641464233,
"step": 465
},
{
"epoch": 0.4052173913043478,
"grad_norm": 21.066162109375,
"learning_rate": 1.7333334286767735e-06,
"loss": 0.3434368371963501,
"step": 466
},
{
"epoch": 0.40608695652173915,
"grad_norm": 44.15741729736328,
"learning_rate": 1.7300149975698213e-06,
"loss": 0.6520743370056152,
"step": 467
},
{
"epoch": 0.40695652173913044,
"grad_norm": 124.7403564453125,
"learning_rate": 1.7266925935505918e-06,
"loss": 1.0646790266036987,
"step": 468
},
{
"epoch": 0.4078260869565217,
"grad_norm": 37.87189865112305,
"learning_rate": 1.7233662441174348e-06,
"loss": 0.6379439234733582,
"step": 469
},
{
"epoch": 0.40869565217391307,
"grad_norm": 28.195348739624023,
"learning_rate": 1.7200359768013528e-06,
"loss": 0.21586567163467407,
"step": 470
},
{
"epoch": 0.40956521739130436,
"grad_norm": 20.992639541625977,
"learning_rate": 1.7167018191657764e-06,
"loss": 0.4772540330886841,
"step": 471
},
{
"epoch": 0.41043478260869565,
"grad_norm": 37.61357116699219,
"learning_rate": 1.7133637988063349e-06,
"loss": 0.2454538345336914,
"step": 472
},
{
"epoch": 0.41130434782608694,
"grad_norm": 19.473154067993164,
"learning_rate": 1.7100219433506268e-06,
"loss": 0.15232905745506287,
"step": 473
},
{
"epoch": 0.4121739130434783,
"grad_norm": 121.49858093261719,
"learning_rate": 1.7066762804579941e-06,
"loss": 1.8549721240997314,
"step": 474
},
{
"epoch": 0.41304347826086957,
"grad_norm": 198.25717163085938,
"learning_rate": 1.7033268378192902e-06,
"loss": 1.8685684204101562,
"step": 475
},
{
"epoch": 0.41391304347826086,
"grad_norm": 69.36676788330078,
"learning_rate": 1.6999736431566533e-06,
"loss": 0.790854811668396,
"step": 476
},
{
"epoch": 0.4147826086956522,
"grad_norm": 287.62042236328125,
"learning_rate": 1.6966167242232742e-06,
"loss": 0.9392005801200867,
"step": 477
},
{
"epoch": 0.4156521739130435,
"grad_norm": 28.407100677490234,
"learning_rate": 1.6932561088031695e-06,
"loss": 0.30870962142944336,
"step": 478
},
{
"epoch": 0.4165217391304348,
"grad_norm": 115.53044891357422,
"learning_rate": 1.6898918247109502e-06,
"loss": 0.7587680816650391,
"step": 479
},
{
"epoch": 0.41739130434782606,
"grad_norm": 290.6261291503906,
"learning_rate": 1.6865238997915907e-06,
"loss": 1.2941091060638428,
"step": 480
},
{
"epoch": 0.4182608695652174,
"grad_norm": 21.97109031677246,
"learning_rate": 1.6831523619202e-06,
"loss": 0.40053749084472656,
"step": 481
},
{
"epoch": 0.4191304347826087,
"grad_norm": 132.6692352294922,
"learning_rate": 1.6797772390017897e-06,
"loss": 0.7245676517486572,
"step": 482
},
{
"epoch": 0.42,
"grad_norm": 73.60350036621094,
"learning_rate": 1.6763985589710443e-06,
"loss": 0.23950672149658203,
"step": 483
},
{
"epoch": 0.42086956521739133,
"grad_norm": 55.66895294189453,
"learning_rate": 1.673016349792088e-06,
"loss": 1.1645352840423584,
"step": 484
},
{
"epoch": 0.4217391304347826,
"grad_norm": 33.097633361816406,
"learning_rate": 1.6696306394582562e-06,
"loss": 0.14924953877925873,
"step": 485
},
{
"epoch": 0.4226086956521739,
"grad_norm": 441.25787353515625,
"learning_rate": 1.6662414559918605e-06,
"loss": 0.9890114665031433,
"step": 486
},
{
"epoch": 0.4234782608695652,
"grad_norm": 38.39339065551758,
"learning_rate": 1.6628488274439592e-06,
"loss": 3.9636688232421875,
"step": 487
},
{
"epoch": 0.42434782608695654,
"grad_norm": 67.4628677368164,
"learning_rate": 1.6594527818941242e-06,
"loss": 0.7358426451683044,
"step": 488
},
{
"epoch": 0.4252173913043478,
"grad_norm": 1039.60498046875,
"learning_rate": 1.656053347450208e-06,
"loss": 4.3725128173828125,
"step": 489
},
{
"epoch": 0.4260869565217391,
"grad_norm": 69.62320709228516,
"learning_rate": 1.6526505522481134e-06,
"loss": 1.933509111404419,
"step": 490
},
{
"epoch": 0.42695652173913046,
"grad_norm": 14.788506507873535,
"learning_rate": 1.6492444244515577e-06,
"loss": 0.12991610169410706,
"step": 491
},
{
"epoch": 0.42782608695652175,
"grad_norm": 52.727142333984375,
"learning_rate": 1.645834992251841e-06,
"loss": 0.41092461347579956,
"step": 492
},
{
"epoch": 0.42869565217391303,
"grad_norm": 207.02723693847656,
"learning_rate": 1.6424222838676127e-06,
"loss": 1.159395456314087,
"step": 493
},
{
"epoch": 0.4295652173913043,
"grad_norm": 18.497066497802734,
"learning_rate": 1.6390063275446387e-06,
"loss": 0.5801668167114258,
"step": 494
},
{
"epoch": 0.43043478260869567,
"grad_norm": 17.380565643310547,
"learning_rate": 1.6355871515555665e-06,
"loss": 0.16514955461025238,
"step": 495
},
{
"epoch": 0.43130434782608695,
"grad_norm": 55.24445724487305,
"learning_rate": 1.6321647841996914e-06,
"loss": 0.7165956497192383,
"step": 496
},
{
"epoch": 0.43217391304347824,
"grad_norm": 86.4374008178711,
"learning_rate": 1.6287392538027225e-06,
"loss": 0.26114028692245483,
"step": 497
},
{
"epoch": 0.4330434782608696,
"grad_norm": 193.58786010742188,
"learning_rate": 1.6253105887165488e-06,
"loss": 1.7449705600738525,
"step": 498
},
{
"epoch": 0.4339130434782609,
"grad_norm": 22.42270851135254,
"learning_rate": 1.6218788173190038e-06,
"loss": 0.29870879650115967,
"step": 499
},
{
"epoch": 0.43478260869565216,
"grad_norm": 163.53623962402344,
"learning_rate": 1.6184439680136305e-06,
"loss": 0.5438425540924072,
"step": 500
},
{
"epoch": 0.4356521739130435,
"grad_norm": 42.024986267089844,
"learning_rate": 1.6150060692294467e-06,
"loss": 1.6341073513031006,
"step": 501
},
{
"epoch": 0.4365217391304348,
"grad_norm": 58.69599151611328,
"learning_rate": 1.6115651494207099e-06,
"loss": 2.2464120388031006,
"step": 502
},
{
"epoch": 0.4373913043478261,
"grad_norm": 36.36306381225586,
"learning_rate": 1.608121237066681e-06,
"loss": 0.27695053815841675,
"step": 503
},
{
"epoch": 0.43826086956521737,
"grad_norm": 68.21349334716797,
"learning_rate": 1.6046743606713905e-06,
"loss": 1.1064863204956055,
"step": 504
},
{
"epoch": 0.4391304347826087,
"grad_norm": 78.13377380371094,
"learning_rate": 1.6012245487633994e-06,
"loss": 0.2547552287578583,
"step": 505
},
{
"epoch": 0.44,
"grad_norm": 14.796619415283203,
"learning_rate": 1.597771829895566e-06,
"loss": 0.3707121014595032,
"step": 506
},
{
"epoch": 0.4408695652173913,
"grad_norm": 170.20506286621094,
"learning_rate": 1.5943162326448086e-06,
"loss": 0.44317924976348877,
"step": 507
},
{
"epoch": 0.44173913043478263,
"grad_norm": 40.973167419433594,
"learning_rate": 1.5908577856118673e-06,
"loss": 1.9060090780258179,
"step": 508
},
{
"epoch": 0.4426086956521739,
"grad_norm": 262.3411865234375,
"learning_rate": 1.5873965174210715e-06,
"loss": 1.479581356048584,
"step": 509
},
{
"epoch": 0.4434782608695652,
"grad_norm": 18.57925796508789,
"learning_rate": 1.5839324567200976e-06,
"loss": 0.2211242914199829,
"step": 510
},
{
"epoch": 0.4443478260869565,
"grad_norm": 45.77786636352539,
"learning_rate": 1.5804656321797362e-06,
"loss": 1.033563256263733,
"step": 511
},
{
"epoch": 0.44521739130434784,
"grad_norm": 191.57579040527344,
"learning_rate": 1.5769960724936532e-06,
"loss": 2.080202579498291,
"step": 512
},
{
"epoch": 0.44608695652173913,
"grad_norm": 54.21561050415039,
"learning_rate": 1.573523806378151e-06,
"loss": 0.6189067363739014,
"step": 513
},
{
"epoch": 0.4469565217391304,
"grad_norm": 154.7218475341797,
"learning_rate": 1.570048862571934e-06,
"loss": 0.5640586614608765,
"step": 514
},
{
"epoch": 0.44782608695652176,
"grad_norm": 717.6767578125,
"learning_rate": 1.5665712698358676e-06,
"loss": 4.7780375480651855,
"step": 515
},
{
"epoch": 0.44869565217391305,
"grad_norm": 29.672758102416992,
"learning_rate": 1.5630910569527419e-06,
"loss": 0.19629834592342377,
"step": 516
},
{
"epoch": 0.44956521739130434,
"grad_norm": 64.20498657226562,
"learning_rate": 1.5596082527270327e-06,
"loss": 1.5060056447982788,
"step": 517
},
{
"epoch": 0.4504347826086956,
"grad_norm": 48.58476257324219,
"learning_rate": 1.5561228859846636e-06,
"loss": 0.4594525694847107,
"step": 518
},
{
"epoch": 0.45130434782608697,
"grad_norm": 175.27867126464844,
"learning_rate": 1.5526349855727674e-06,
"loss": 2.5710599422454834,
"step": 519
},
{
"epoch": 0.45217391304347826,
"grad_norm": 88.2507553100586,
"learning_rate": 1.5491445803594474e-06,
"loss": 0.6433696746826172,
"step": 520
},
{
"epoch": 0.45304347826086955,
"grad_norm": 46.259891510009766,
"learning_rate": 1.5456516992335375e-06,
"loss": 0.34182000160217285,
"step": 521
},
{
"epoch": 0.4539130434782609,
"grad_norm": 34.02726745605469,
"learning_rate": 1.542156371104364e-06,
"loss": 0.3597108721733093,
"step": 522
},
{
"epoch": 0.4547826086956522,
"grad_norm": 38.61172103881836,
"learning_rate": 1.5386586249015074e-06,
"loss": 2.7518372535705566,
"step": 523
},
{
"epoch": 0.45565217391304347,
"grad_norm": 33.561893463134766,
"learning_rate": 1.5351584895745592e-06,
"loss": 0.20846080780029297,
"step": 524
},
{
"epoch": 0.45652173913043476,
"grad_norm": 64.6435775756836,
"learning_rate": 1.5316559940928877e-06,
"loss": 0.4035212993621826,
"step": 525
},
{
"epoch": 0.4573913043478261,
"grad_norm": 37.619441986083984,
"learning_rate": 1.528151167445393e-06,
"loss": 0.207579106092453,
"step": 526
},
{
"epoch": 0.4582608695652174,
"grad_norm": 177.21896362304688,
"learning_rate": 1.5246440386402716e-06,
"loss": 0.2880960702896118,
"step": 527
},
{
"epoch": 0.4591304347826087,
"grad_norm": 48.139530181884766,
"learning_rate": 1.5211346367047719e-06,
"loss": 0.3652404546737671,
"step": 528
},
{
"epoch": 0.46,
"grad_norm": 727.7797241210938,
"learning_rate": 1.5176229906849577e-06,
"loss": 0.5256285667419434,
"step": 529
},
{
"epoch": 0.4608695652173913,
"grad_norm": 119.33316802978516,
"learning_rate": 1.514109129645466e-06,
"loss": 0.1947031468153,
"step": 530
},
{
"epoch": 0.4617391304347826,
"grad_norm": 77.66236877441406,
"learning_rate": 1.5105930826692664e-06,
"loss": 0.638593852519989,
"step": 531
},
{
"epoch": 0.46260869565217394,
"grad_norm": 79.5414047241211,
"learning_rate": 1.5070748788574208e-06,
"loss": 0.6514780521392822,
"step": 532
},
{
"epoch": 0.46347826086956523,
"grad_norm": 83.55230712890625,
"learning_rate": 1.5035545473288426e-06,
"loss": 1.4839876890182495,
"step": 533
},
{
"epoch": 0.4643478260869565,
"grad_norm": 139.66659545898438,
"learning_rate": 1.5000321172200557e-06,
"loss": 2.5634307861328125,
"step": 534
},
{
"epoch": 0.4652173913043478,
"grad_norm": 69.7106704711914,
"learning_rate": 1.4965076176849528e-06,
"loss": 1.4087194204330444,
"step": 535
},
{
"epoch": 0.46608695652173915,
"grad_norm": 19.372730255126953,
"learning_rate": 1.4929810778945545e-06,
"loss": 0.24766401946544647,
"step": 536
},
{
"epoch": 0.46695652173913044,
"grad_norm": 68.26744842529297,
"learning_rate": 1.4894525270367682e-06,
"loss": 0.539479672908783,
"step": 537
},
{
"epoch": 0.4678260869565217,
"grad_norm": 51.5244255065918,
"learning_rate": 1.485921994316146e-06,
"loss": 0.8009777069091797,
"step": 538
},
{
"epoch": 0.46869565217391307,
"grad_norm": 86.90782165527344,
"learning_rate": 1.4823895089536435e-06,
"loss": 1.678908109664917,
"step": 539
},
{
"epoch": 0.46956521739130436,
"grad_norm": 221.81761169433594,
"learning_rate": 1.4788551001863768e-06,
"loss": 1.2407617568969727,
"step": 540
},
{
"epoch": 0.47043478260869565,
"grad_norm": 105.43783569335938,
"learning_rate": 1.4753187972673824e-06,
"loss": 0.5548800230026245,
"step": 541
},
{
"epoch": 0.47130434782608693,
"grad_norm": 27.868223190307617,
"learning_rate": 1.471780629465373e-06,
"loss": 1.8065273761749268,
"step": 542
},
{
"epoch": 0.4721739130434783,
"grad_norm": 31.52602195739746,
"learning_rate": 1.468240626064497e-06,
"loss": 0.40217941999435425,
"step": 543
},
{
"epoch": 0.47304347826086957,
"grad_norm": 32.67758560180664,
"learning_rate": 1.4646988163640955e-06,
"loss": 0.41841572523117065,
"step": 544
},
{
"epoch": 0.47391304347826085,
"grad_norm": 48.6520881652832,
"learning_rate": 1.4611552296784581e-06,
"loss": 1.557771921157837,
"step": 545
},
{
"epoch": 0.4747826086956522,
"grad_norm": 105.38114166259766,
"learning_rate": 1.4576098953365845e-06,
"loss": 0.5157253742218018,
"step": 546
},
{
"epoch": 0.4756521739130435,
"grad_norm": 131.1835479736328,
"learning_rate": 1.454062842681937e-06,
"loss": 1.6056303977966309,
"step": 547
},
{
"epoch": 0.4765217391304348,
"grad_norm": 50.23271179199219,
"learning_rate": 1.4505141010722004e-06,
"loss": 0.32590997219085693,
"step": 548
},
{
"epoch": 0.47739130434782606,
"grad_norm": 36.35626983642578,
"learning_rate": 1.446963699879039e-06,
"loss": 0.20287835597991943,
"step": 549
},
{
"epoch": 0.4782608695652174,
"grad_norm": 35.27216339111328,
"learning_rate": 1.4434116684878516e-06,
"loss": 0.28870874643325806,
"step": 550
},
{
"epoch": 0.4791304347826087,
"grad_norm": 216.48423767089844,
"learning_rate": 1.4398580362975311e-06,
"loss": 1.723888874053955,
"step": 551
},
{
"epoch": 0.48,
"grad_norm": 76.89714813232422,
"learning_rate": 1.436302832720218e-06,
"loss": 0.3109407126903534,
"step": 552
},
{
"epoch": 0.4808695652173913,
"grad_norm": 334.2703857421875,
"learning_rate": 1.4327460871810606e-06,
"loss": 1.328261137008667,
"step": 553
},
{
"epoch": 0.4817391304347826,
"grad_norm": 47.806217193603516,
"learning_rate": 1.429187829117967e-06,
"loss": 0.5149089097976685,
"step": 554
},
{
"epoch": 0.4826086956521739,
"grad_norm": 44.39208221435547,
"learning_rate": 1.4256280879813661e-06,
"loss": 0.4990388751029968,
"step": 555
},
{
"epoch": 0.4834782608695652,
"grad_norm": 38.07180404663086,
"learning_rate": 1.4220668932339606e-06,
"loss": 0.6276671886444092,
"step": 556
},
{
"epoch": 0.48434782608695653,
"grad_norm": 48.151390075683594,
"learning_rate": 1.4185042743504843e-06,
"loss": 0.39081475138664246,
"step": 557
},
{
"epoch": 0.4852173913043478,
"grad_norm": 72.2470703125,
"learning_rate": 1.4149402608174583e-06,
"loss": 1.7942055463790894,
"step": 558
},
{
"epoch": 0.4860869565217391,
"grad_norm": 28.45145034790039,
"learning_rate": 1.4113748821329462e-06,
"loss": 0.34298908710479736,
"step": 559
},
{
"epoch": 0.48695652173913045,
"grad_norm": 41.73386001586914,
"learning_rate": 1.4078081678063118e-06,
"loss": 0.5306379795074463,
"step": 560
},
{
"epoch": 0.48782608695652174,
"grad_norm": 26.68669891357422,
"learning_rate": 1.4042401473579717e-06,
"loss": 0.27347320318222046,
"step": 561
},
{
"epoch": 0.48869565217391303,
"grad_norm": 19.17247200012207,
"learning_rate": 1.4006708503191539e-06,
"loss": 0.13084499537944794,
"step": 562
},
{
"epoch": 0.4895652173913044,
"grad_norm": 38.385780334472656,
"learning_rate": 1.3971003062316527e-06,
"loss": 0.25119465589523315,
"step": 563
},
{
"epoch": 0.49043478260869566,
"grad_norm": 46.42362976074219,
"learning_rate": 1.3935285446475822e-06,
"loss": 2.279939889907837,
"step": 564
},
{
"epoch": 0.49130434782608695,
"grad_norm": 102.55758666992188,
"learning_rate": 1.3899555951291348e-06,
"loss": 2.410283088684082,
"step": 565
},
{
"epoch": 0.49217391304347824,
"grad_norm": 59.88905334472656,
"learning_rate": 1.3863814872483347e-06,
"loss": 0.506309449672699,
"step": 566
},
{
"epoch": 0.4930434782608696,
"grad_norm": 22.88003158569336,
"learning_rate": 1.3828062505867928e-06,
"loss": 0.29034480452537537,
"step": 567
},
{
"epoch": 0.49391304347826087,
"grad_norm": 42.26963806152344,
"learning_rate": 1.3792299147354626e-06,
"loss": 0.2248980700969696,
"step": 568
},
{
"epoch": 0.49478260869565216,
"grad_norm": 86.31819152832031,
"learning_rate": 1.3756525092943965e-06,
"loss": 0.36558055877685547,
"step": 569
},
{
"epoch": 0.4956521739130435,
"grad_norm": 694.8186645507812,
"learning_rate": 1.3720740638724977e-06,
"loss": 5.301316261291504,
"step": 570
},
{
"epoch": 0.4965217391304348,
"grad_norm": 55.80915451049805,
"learning_rate": 1.3684946080872779e-06,
"loss": 0.4611801505088806,
"step": 571
},
{
"epoch": 0.4973913043478261,
"grad_norm": 105.0939712524414,
"learning_rate": 1.3649141715646117e-06,
"loss": 0.7904858589172363,
"step": 572
},
{
"epoch": 0.49826086956521737,
"grad_norm": 36.488460540771484,
"learning_rate": 1.361332783938489e-06,
"loss": 0.4929240345954895,
"step": 573
},
{
"epoch": 0.4991304347826087,
"grad_norm": 41.54960632324219,
"learning_rate": 1.3577504748507744e-06,
"loss": 0.5713452100753784,
"step": 574
},
{
"epoch": 0.5,
"grad_norm": 384.3897399902344,
"learning_rate": 1.3541672739509568e-06,
"loss": 1.2220113277435303,
"step": 575
},
{
"epoch": 0.5008695652173913,
"grad_norm": 10.600460052490234,
"learning_rate": 1.3505832108959077e-06,
"loss": 0.30977481603622437,
"step": 576
},
{
"epoch": 0.5017391304347826,
"grad_norm": 13.342951774597168,
"learning_rate": 1.346998315349633e-06,
"loss": 0.13482525944709778,
"step": 577
},
{
"epoch": 0.5026086956521739,
"grad_norm": 40.32766342163086,
"learning_rate": 1.3434126169830306e-06,
"loss": 0.6127514839172363,
"step": 578
},
{
"epoch": 0.5034782608695653,
"grad_norm": 16.961238861083984,
"learning_rate": 1.3398261454736413e-06,
"loss": 0.556333601474762,
"step": 579
},
{
"epoch": 0.5043478260869565,
"grad_norm": 26.407115936279297,
"learning_rate": 1.3362389305054061e-06,
"loss": 0.20388464629650116,
"step": 580
},
{
"epoch": 0.5052173913043478,
"grad_norm": 40.9092903137207,
"learning_rate": 1.3326510017684194e-06,
"loss": 0.35018491744995117,
"step": 581
},
{
"epoch": 0.5060869565217392,
"grad_norm": 427.9647216796875,
"learning_rate": 1.3290623889586817e-06,
"loss": 1.300086259841919,
"step": 582
},
{
"epoch": 0.5069565217391304,
"grad_norm": 24.641036987304688,
"learning_rate": 1.3254731217778577e-06,
"loss": 3.908027410507202,
"step": 583
},
{
"epoch": 0.5078260869565218,
"grad_norm": 104.31623840332031,
"learning_rate": 1.3218832299330259e-06,
"loss": 0.953208327293396,
"step": 584
},
{
"epoch": 0.508695652173913,
"grad_norm": 20.830305099487305,
"learning_rate": 1.3182927431364358e-06,
"loss": 0.14480295777320862,
"step": 585
},
{
"epoch": 0.5095652173913043,
"grad_norm": 24.25160789489746,
"learning_rate": 1.314701691105262e-06,
"loss": 0.5349597334861755,
"step": 586
},
{
"epoch": 0.5104347826086957,
"grad_norm": 23.42707633972168,
"learning_rate": 1.3111101035613551e-06,
"loss": 0.4934194087982178,
"step": 587
},
{
"epoch": 0.5113043478260869,
"grad_norm": 17.85067367553711,
"learning_rate": 1.3075180102310008e-06,
"loss": 0.14509335160255432,
"step": 588
},
{
"epoch": 0.5121739130434783,
"grad_norm": 51.71796417236328,
"learning_rate": 1.3039254408446683e-06,
"loss": 0.2814500033855438,
"step": 589
},
{
"epoch": 0.5130434782608696,
"grad_norm": 446.1482849121094,
"learning_rate": 1.3003324251367692e-06,
"loss": 1.8456300497055054,
"step": 590
},
{
"epoch": 0.5139130434782608,
"grad_norm": 45.47982406616211,
"learning_rate": 1.2967389928454075e-06,
"loss": 0.28155648708343506,
"step": 591
},
{
"epoch": 0.5147826086956522,
"grad_norm": 65.72172546386719,
"learning_rate": 1.2931451737121353e-06,
"loss": 0.6789445281028748,
"step": 592
},
{
"epoch": 0.5156521739130435,
"grad_norm": 90.84854888916016,
"learning_rate": 1.289550997481708e-06,
"loss": 0.7465357780456543,
"step": 593
},
{
"epoch": 0.5165217391304348,
"grad_norm": 23.170902252197266,
"learning_rate": 1.2859564939018344e-06,
"loss": 0.6693512201309204,
"step": 594
},
{
"epoch": 0.5173913043478261,
"grad_norm": 30.57984733581543,
"learning_rate": 1.2823616927229345e-06,
"loss": 2.390178918838501,
"step": 595
},
{
"epoch": 0.5182608695652174,
"grad_norm": 87.88422393798828,
"learning_rate": 1.2787666236978902e-06,
"loss": 0.5040265321731567,
"step": 596
},
{
"epoch": 0.5191304347826087,
"grad_norm": 27.1507568359375,
"learning_rate": 1.2751713165818005e-06,
"loss": 0.19294409453868866,
"step": 597
},
{
"epoch": 0.52,
"grad_norm": 74.13288116455078,
"learning_rate": 1.2715758011317359e-06,
"loss": 0.3155352473258972,
"step": 598
},
{
"epoch": 0.5208695652173913,
"grad_norm": 40.7326774597168,
"learning_rate": 1.2679801071064895e-06,
"loss": 0.26849883794784546,
"step": 599
},
{
"epoch": 0.5217391304347826,
"grad_norm": 26.95990562438965,
"learning_rate": 1.2643842642663345e-06,
"loss": 0.1746445596218109,
"step": 600
},
{
"epoch": 0.5226086956521739,
"grad_norm": 40.006038665771484,
"learning_rate": 1.2607883023727737e-06,
"loss": 0.6402615308761597,
"step": 601
},
{
"epoch": 0.5234782608695652,
"grad_norm": 195.54122924804688,
"learning_rate": 1.2571922511882973e-06,
"loss": 1.3552939891815186,
"step": 602
},
{
"epoch": 0.5243478260869565,
"grad_norm": 58.347862243652344,
"learning_rate": 1.2535961404761322e-06,
"loss": 0.6402984857559204,
"step": 603
},
{
"epoch": 0.5252173913043479,
"grad_norm": 44.041996002197266,
"learning_rate": 1.25e-06,
"loss": 0.39661285281181335,
"step": 604
},
{
"epoch": 0.5260869565217391,
"grad_norm": 22.41481590270996,
"learning_rate": 1.2464038595238682e-06,
"loss": 0.1455352008342743,
"step": 605
},
{
"epoch": 0.5269565217391304,
"grad_norm": 39.84624099731445,
"learning_rate": 1.2428077488117033e-06,
"loss": 0.7259137034416199,
"step": 606
},
{
"epoch": 0.5278260869565218,
"grad_norm": 84.27421569824219,
"learning_rate": 1.2392116976272265e-06,
"loss": 0.5845558643341064,
"step": 607
},
{
"epoch": 0.528695652173913,
"grad_norm": 72.01571655273438,
"learning_rate": 1.2356157357336657e-06,
"loss": 0.561434805393219,
"step": 608
},
{
"epoch": 0.5295652173913044,
"grad_norm": 50.12057876586914,
"learning_rate": 1.2320198928935109e-06,
"loss": 0.3289182782173157,
"step": 609
},
{
"epoch": 0.5304347826086957,
"grad_norm": 53.193702697753906,
"learning_rate": 1.2284241988682645e-06,
"loss": 0.3097253739833832,
"step": 610
},
{
"epoch": 0.5313043478260869,
"grad_norm": 59.04848098754883,
"learning_rate": 1.2248286834182e-06,
"loss": 1.2296556234359741,
"step": 611
},
{
"epoch": 0.5321739130434783,
"grad_norm": 320.2524719238281,
"learning_rate": 1.2212333763021104e-06,
"loss": 1.0541026592254639,
"step": 612
},
{
"epoch": 0.5330434782608696,
"grad_norm": 75.85768127441406,
"learning_rate": 1.2176383072770661e-06,
"loss": 0.5213727951049805,
"step": 613
},
{
"epoch": 0.5339130434782609,
"grad_norm": 34.386817932128906,
"learning_rate": 1.2140435060981662e-06,
"loss": 0.651544451713562,
"step": 614
},
{
"epoch": 0.5347826086956522,
"grad_norm": 34.9528694152832,
"learning_rate": 1.2104490025182927e-06,
"loss": 0.30887371301651,
"step": 615
},
{
"epoch": 0.5356521739130434,
"grad_norm": 17.99839210510254,
"learning_rate": 1.2068548262878653e-06,
"loss": 0.24556462466716766,
"step": 616
},
{
"epoch": 0.5365217391304348,
"grad_norm": 199.3224334716797,
"learning_rate": 1.2032610071545931e-06,
"loss": 0.7938846349716187,
"step": 617
},
{
"epoch": 0.5373913043478261,
"grad_norm": 430.3277587890625,
"learning_rate": 1.1996675748632314e-06,
"loss": 0.9159368276596069,
"step": 618
},
{
"epoch": 0.5382608695652173,
"grad_norm": 206.21693420410156,
"learning_rate": 1.1960745591553319e-06,
"loss": 0.6134743690490723,
"step": 619
},
{
"epoch": 0.5391304347826087,
"grad_norm": 23.450517654418945,
"learning_rate": 1.1924819897689998e-06,
"loss": 0.2543744444847107,
"step": 620
},
{
"epoch": 0.54,
"grad_norm": 34.240299224853516,
"learning_rate": 1.1888898964386453e-06,
"loss": 0.6171222925186157,
"step": 621
},
{
"epoch": 0.5408695652173913,
"grad_norm": 27.85487937927246,
"learning_rate": 1.1852983088947382e-06,
"loss": 0.49735331535339355,
"step": 622
},
{
"epoch": 0.5417391304347826,
"grad_norm": 47.37606430053711,
"learning_rate": 1.181707256863564e-06,
"loss": 0.7097009420394897,
"step": 623
},
{
"epoch": 0.542608695652174,
"grad_norm": 36.9967155456543,
"learning_rate": 1.1781167700669743e-06,
"loss": 0.8801654577255249,
"step": 624
},
{
"epoch": 0.5434782608695652,
"grad_norm": 78.14496612548828,
"learning_rate": 1.1745268782221425e-06,
"loss": 0.5738281011581421,
"step": 625
},
{
"epoch": 0.5443478260869565,
"grad_norm": 129.960693359375,
"learning_rate": 1.1709376110413183e-06,
"loss": 1.9200137853622437,
"step": 626
},
{
"epoch": 0.5452173913043479,
"grad_norm": 35.930938720703125,
"learning_rate": 1.1673489982315808e-06,
"loss": 0.2007402628660202,
"step": 627
},
{
"epoch": 0.5460869565217391,
"grad_norm": 115.76295471191406,
"learning_rate": 1.1637610694945939e-06,
"loss": 0.6953235864639282,
"step": 628
},
{
"epoch": 0.5469565217391305,
"grad_norm": 48.088626861572266,
"learning_rate": 1.160173854526359e-06,
"loss": 0.4470199942588806,
"step": 629
},
{
"epoch": 0.5478260869565217,
"grad_norm": 54.21675491333008,
"learning_rate": 1.1565873830169698e-06,
"loss": 0.4299599528312683,
"step": 630
},
{
"epoch": 0.548695652173913,
"grad_norm": 23.21803092956543,
"learning_rate": 1.1530016846503673e-06,
"loss": 0.1876332014799118,
"step": 631
},
{
"epoch": 0.5495652173913044,
"grad_norm": 18.29347038269043,
"learning_rate": 1.1494167891040927e-06,
"loss": 0.24449440836906433,
"step": 632
},
{
"epoch": 0.5504347826086956,
"grad_norm": 60.130741119384766,
"learning_rate": 1.1458327260490434e-06,
"loss": 0.390960693359375,
"step": 633
},
{
"epoch": 0.551304347826087,
"grad_norm": 148.6415252685547,
"learning_rate": 1.1422495251492258e-06,
"loss": 0.8560740351676941,
"step": 634
},
{
"epoch": 0.5521739130434783,
"grad_norm": 97.03267669677734,
"learning_rate": 1.1386672160615112e-06,
"loss": 0.29729336500167847,
"step": 635
},
{
"epoch": 0.5530434782608695,
"grad_norm": 40.719295501708984,
"learning_rate": 1.135085828435389e-06,
"loss": 0.31616446375846863,
"step": 636
},
{
"epoch": 0.5539130434782609,
"grad_norm": 20.09977149963379,
"learning_rate": 1.1315053919127223e-06,
"loss": 0.3076978921890259,
"step": 637
},
{
"epoch": 0.5547826086956522,
"grad_norm": 86.53784942626953,
"learning_rate": 1.1279259361275025e-06,
"loss": 0.47761139273643494,
"step": 638
},
{
"epoch": 0.5556521739130434,
"grad_norm": 18.16761589050293,
"learning_rate": 1.124347490705604e-06,
"loss": 0.16975855827331543,
"step": 639
},
{
"epoch": 0.5565217391304348,
"grad_norm": 64.80789184570312,
"learning_rate": 1.1207700852645376e-06,
"loss": 0.30884677171707153,
"step": 640
},
{
"epoch": 0.5573913043478261,
"grad_norm": 26.878433227539062,
"learning_rate": 1.1171937494132076e-06,
"loss": 0.22658881545066833,
"step": 641
},
{
"epoch": 0.5582608695652174,
"grad_norm": 59.25761413574219,
"learning_rate": 1.1136185127516658e-06,
"loss": 0.3831648528575897,
"step": 642
},
{
"epoch": 0.5591304347826087,
"grad_norm": 10.84101676940918,
"learning_rate": 1.1100444048708654e-06,
"loss": 0.1251206248998642,
"step": 643
},
{
"epoch": 0.56,
"grad_norm": 44.02790832519531,
"learning_rate": 1.1064714553524182e-06,
"loss": 0.9723340272903442,
"step": 644
},
{
"epoch": 0.5608695652173913,
"grad_norm": 25.542354583740234,
"learning_rate": 1.1028996937683477e-06,
"loss": 0.21140162646770477,
"step": 645
},
{
"epoch": 0.5617391304347826,
"grad_norm": 41.091453552246094,
"learning_rate": 1.0993291496808463e-06,
"loss": 0.5855872631072998,
"step": 646
},
{
"epoch": 0.5626086956521739,
"grad_norm": 73.25472259521484,
"learning_rate": 1.0957598526420287e-06,
"loss": 2.3975653648376465,
"step": 647
},
{
"epoch": 0.5634782608695652,
"grad_norm": 41.84149169921875,
"learning_rate": 1.0921918321936888e-06,
"loss": 1.8163971900939941,
"step": 648
},
{
"epoch": 0.5643478260869565,
"grad_norm": 56.16062927246094,
"learning_rate": 1.0886251178670543e-06,
"loss": 1.6685190200805664,
"step": 649
},
{
"epoch": 0.5652173913043478,
"grad_norm": 41.29354476928711,
"learning_rate": 1.0850597391825423e-06,
"loss": 0.5237247347831726,
"step": 650
},
{
"epoch": 0.5660869565217391,
"grad_norm": 148.20823669433594,
"learning_rate": 1.0814957256495165e-06,
"loss": 0.5414335131645203,
"step": 651
},
{
"epoch": 0.5669565217391305,
"grad_norm": 25.504301071166992,
"learning_rate": 1.07793310676604e-06,
"loss": 0.21537749469280243,
"step": 652
},
{
"epoch": 0.5678260869565217,
"grad_norm": 16.029409408569336,
"learning_rate": 1.0743719120186345e-06,
"loss": 0.228776216506958,
"step": 653
},
{
"epoch": 0.568695652173913,
"grad_norm": 41.803714752197266,
"learning_rate": 1.0708121708820335e-06,
"loss": 1.168609619140625,
"step": 654
},
{
"epoch": 0.5695652173913044,
"grad_norm": 13.706403732299805,
"learning_rate": 1.0672539128189403e-06,
"loss": 0.12180668115615845,
"step": 655
},
{
"epoch": 0.5704347826086956,
"grad_norm": 16.999237060546875,
"learning_rate": 1.0636971672797818e-06,
"loss": 0.12567901611328125,
"step": 656
},
{
"epoch": 0.571304347826087,
"grad_norm": 30.821714401245117,
"learning_rate": 1.060141963702469e-06,
"loss": 1.806490421295166,
"step": 657
},
{
"epoch": 0.5721739130434783,
"grad_norm": 41.3328857421875,
"learning_rate": 1.0565883315121484e-06,
"loss": 0.31122148036956787,
"step": 658
},
{
"epoch": 0.5730434782608695,
"grad_norm": 35.64419937133789,
"learning_rate": 1.0530363001209615e-06,
"loss": 0.3361220955848694,
"step": 659
},
{
"epoch": 0.5739130434782609,
"grad_norm": 38.72517776489258,
"learning_rate": 1.0494858989277996e-06,
"loss": 0.3497597575187683,
"step": 660
},
{
"epoch": 0.5747826086956521,
"grad_norm": 71.2376480102539,
"learning_rate": 1.0459371573180634e-06,
"loss": 1.4939545392990112,
"step": 661
},
{
"epoch": 0.5756521739130435,
"grad_norm": 120.57617950439453,
"learning_rate": 1.0423901046634155e-06,
"loss": 1.3161383867263794,
"step": 662
},
{
"epoch": 0.5765217391304348,
"grad_norm": 404.88800048828125,
"learning_rate": 1.038844770321542e-06,
"loss": 1.8383941650390625,
"step": 663
},
{
"epoch": 0.577391304347826,
"grad_norm": 46.807308197021484,
"learning_rate": 1.035301183635905e-06,
"loss": 0.5617116093635559,
"step": 664
},
{
"epoch": 0.5782608695652174,
"grad_norm": 20.975719451904297,
"learning_rate": 1.0317593739355031e-06,
"loss": 0.1383107751607895,
"step": 665
},
{
"epoch": 0.5791304347826087,
"grad_norm": 10.463642120361328,
"learning_rate": 1.0282193705346272e-06,
"loss": 0.10751871764659882,
"step": 666
},
{
"epoch": 0.58,
"grad_norm": 19.149900436401367,
"learning_rate": 1.024681202732618e-06,
"loss": 0.2069488763809204,
"step": 667
},
{
"epoch": 0.5808695652173913,
"grad_norm": 282.3392028808594,
"learning_rate": 1.0211448998136234e-06,
"loss": 1.053381323814392,
"step": 668
},
{
"epoch": 0.5817391304347826,
"grad_norm": 23.324705123901367,
"learning_rate": 1.017610491046357e-06,
"loss": 0.27052515745162964,
"step": 669
},
{
"epoch": 0.5826086956521739,
"grad_norm": 49.36856460571289,
"learning_rate": 1.0140780056838543e-06,
"loss": 0.3036606013774872,
"step": 670
},
{
"epoch": 0.5834782608695652,
"grad_norm": 36.350608825683594,
"learning_rate": 1.010547472963232e-06,
"loss": 0.42312198877334595,
"step": 671
},
{
"epoch": 0.5843478260869566,
"grad_norm": 53.28272247314453,
"learning_rate": 1.007018922105446e-06,
"loss": 0.4440731704235077,
"step": 672
},
{
"epoch": 0.5852173913043478,
"grad_norm": 56.8600959777832,
"learning_rate": 1.0034923823150477e-06,
"loss": 0.3036220967769623,
"step": 673
},
{
"epoch": 0.5860869565217391,
"grad_norm": 359.54510498046875,
"learning_rate": 9.999678827799447e-07,
"loss": 1.8597688674926758,
"step": 674
},
{
"epoch": 0.5869565217391305,
"grad_norm": 83.56299591064453,
"learning_rate": 9.964454526711576e-07,
"loss": 0.3423331677913666,
"step": 675
},
{
"epoch": 0.5878260869565217,
"grad_norm": 29.95453643798828,
"learning_rate": 9.929251211425794e-07,
"loss": 0.21496626734733582,
"step": 676
},
{
"epoch": 0.5886956521739131,
"grad_norm": 166.31478881835938,
"learning_rate": 9.89406917330734e-07,
"loss": 0.2252083122730255,
"step": 677
},
{
"epoch": 0.5895652173913043,
"grad_norm": 50.53282165527344,
"learning_rate": 9.858908703545342e-07,
"loss": 0.3894593119621277,
"step": 678
},
{
"epoch": 0.5904347826086956,
"grad_norm": 14.309340476989746,
"learning_rate": 9.823770093150425e-07,
"loss": 0.14528369903564453,
"step": 679
},
{
"epoch": 0.591304347826087,
"grad_norm": 19.853036880493164,
"learning_rate": 9.788653632952283e-07,
"loss": 0.28800415992736816,
"step": 680
},
{
"epoch": 0.5921739130434782,
"grad_norm": 285.6701354980469,
"learning_rate": 9.75355961359729e-07,
"loss": 1.3987009525299072,
"step": 681
},
{
"epoch": 0.5930434782608696,
"grad_norm": 27.742000579833984,
"learning_rate": 9.718488325546072e-07,
"loss": 0.2805967926979065,
"step": 682
},
{
"epoch": 0.5939130434782609,
"grad_norm": 36.778892517089844,
"learning_rate": 9.68344005907113e-07,
"loss": 0.21426287293434143,
"step": 683
},
{
"epoch": 0.5947826086956521,
"grad_norm": 18.567581176757812,
"learning_rate": 9.648415104254414e-07,
"loss": 0.14293357729911804,
"step": 684
},
{
"epoch": 0.5956521739130435,
"grad_norm": 27.923137664794922,
"learning_rate": 9.613413750984935e-07,
"loss": 0.6410040855407715,
"step": 685
},
{
"epoch": 0.5965217391304348,
"grad_norm": 33.96150207519531,
"learning_rate": 9.578436288956366e-07,
"loss": 0.2727791666984558,
"step": 686
},
{
"epoch": 0.5973913043478261,
"grad_norm": 57.6609001159668,
"learning_rate": 9.543483007664632e-07,
"loss": 0.8150521516799927,
"step": 687
},
{
"epoch": 0.5982608695652174,
"grad_norm": 149.72584533691406,
"learning_rate": 9.508554196405532e-07,
"loss": 0.6287965774536133,
"step": 688
},
{
"epoch": 0.5991304347826087,
"grad_norm": 47.13349914550781,
"learning_rate": 9.473650144272328e-07,
"loss": 0.7446504831314087,
"step": 689
},
{
"epoch": 0.6,
"grad_norm": 25.518701553344727,
"learning_rate": 9.438771140153365e-07,
"loss": 0.2882513999938965,
"step": 690
},
{
"epoch": 0.6008695652173913,
"grad_norm": 47.83425521850586,
"learning_rate": 9.403917472729676e-07,
"loss": 1.567733645439148,
"step": 691
},
{
"epoch": 0.6017391304347826,
"grad_norm": 17.986549377441406,
"learning_rate": 9.369089430472583e-07,
"loss": 0.18506132066249847,
"step": 692
},
{
"epoch": 0.6026086956521739,
"grad_norm": 69.72415161132812,
"learning_rate": 9.334287301641326e-07,
"loss": 2.853628158569336,
"step": 693
},
{
"epoch": 0.6034782608695652,
"grad_norm": 77.92742156982422,
"learning_rate": 9.299511374280662e-07,
"loss": 0.4185067415237427,
"step": 694
},
{
"epoch": 0.6043478260869565,
"grad_norm": 53.21318054199219,
"learning_rate": 9.26476193621849e-07,
"loss": 0.6572995185852051,
"step": 695
},
{
"epoch": 0.6052173913043478,
"grad_norm": 57.128116607666016,
"learning_rate": 9.230039275063473e-07,
"loss": 0.4568495750427246,
"step": 696
},
{
"epoch": 0.6060869565217392,
"grad_norm": 39.83743667602539,
"learning_rate": 9.195343678202639e-07,
"loss": 0.25283700227737427,
"step": 697
},
{
"epoch": 0.6069565217391304,
"grad_norm": 274.8430480957031,
"learning_rate": 9.160675432799028e-07,
"loss": 1.0550380945205688,
"step": 698
},
{
"epoch": 0.6078260869565217,
"grad_norm": 22.163984298706055,
"learning_rate": 9.126034825789289e-07,
"loss": 0.5107827186584473,
"step": 699
},
{
"epoch": 0.6086956521739131,
"grad_norm": 13.485553741455078,
"learning_rate": 9.091422143881329e-07,
"loss": 2.6540143489837646,
"step": 700
},
{
"epoch": 0.6095652173913043,
"grad_norm": 96.43163299560547,
"learning_rate": 9.056837673551919e-07,
"loss": 0.4711366891860962,
"step": 701
},
{
"epoch": 0.6104347826086957,
"grad_norm": 16.987398147583008,
"learning_rate": 9.022281701044343e-07,
"loss": 0.1364254355430603,
"step": 702
},
{
"epoch": 0.611304347826087,
"grad_norm": 31.27009391784668,
"learning_rate": 8.98775451236601e-07,
"loss": 0.1886814832687378,
"step": 703
},
{
"epoch": 0.6121739130434782,
"grad_norm": 29.242935180664062,
"learning_rate": 8.953256393286099e-07,
"loss": 0.15994791686534882,
"step": 704
},
{
"epoch": 0.6130434782608696,
"grad_norm": 112.5531005859375,
"learning_rate": 8.918787629333192e-07,
"loss": 0.7741197943687439,
"step": 705
},
{
"epoch": 0.6139130434782609,
"grad_norm": 21.623342514038086,
"learning_rate": 8.884348505792905e-07,
"loss": 0.38500142097473145,
"step": 706
},
{
"epoch": 0.6147826086956522,
"grad_norm": 27.701379776000977,
"learning_rate": 8.849939307705538e-07,
"loss": 0.4803548753261566,
"step": 707
},
{
"epoch": 0.6156521739130435,
"grad_norm": 123.16143035888672,
"learning_rate": 8.815560319863698e-07,
"loss": 0.8539928197860718,
"step": 708
},
{
"epoch": 0.6165217391304347,
"grad_norm": 17.093400955200195,
"learning_rate": 8.781211826809966e-07,
"loss": 0.2064400315284729,
"step": 709
},
{
"epoch": 0.6173913043478261,
"grad_norm": 57.94111251831055,
"learning_rate": 8.746894112834514e-07,
"loss": 0.7408286333084106,
"step": 710
},
{
"epoch": 0.6182608695652174,
"grad_norm": 259.595458984375,
"learning_rate": 8.712607461972777e-07,
"loss": 0.3563030958175659,
"step": 711
},
{
"epoch": 0.6191304347826087,
"grad_norm": 17.271316528320312,
"learning_rate": 8.678352158003093e-07,
"loss": 0.18964272737503052,
"step": 712
},
{
"epoch": 0.62,
"grad_norm": 27.675994873046875,
"learning_rate": 8.644128484444339e-07,
"loss": 0.3412845730781555,
"step": 713
},
{
"epoch": 0.6208695652173913,
"grad_norm": 79.86682891845703,
"learning_rate": 8.609936724553616e-07,
"loss": 0.26230788230895996,
"step": 714
},
{
"epoch": 0.6217391304347826,
"grad_norm": 222.57064819335938,
"learning_rate": 8.575777161323876e-07,
"loss": 0.9684433937072754,
"step": 715
},
{
"epoch": 0.6226086956521739,
"grad_norm": 74.30072021484375,
"learning_rate": 8.541650077481598e-07,
"loss": 1.392247200012207,
"step": 716
},
{
"epoch": 0.6234782608695653,
"grad_norm": 29.02860450744629,
"learning_rate": 8.507555755484429e-07,
"loss": 2.5200443267822266,
"step": 717
},
{
"epoch": 0.6243478260869565,
"grad_norm": 24.11271095275879,
"learning_rate": 8.47349447751887e-07,
"loss": 0.1644997000694275,
"step": 718
},
{
"epoch": 0.6252173913043478,
"grad_norm": 37.57582473754883,
"learning_rate": 8.439466525497925e-07,
"loss": 0.5687346458435059,
"step": 719
},
{
"epoch": 0.6260869565217392,
"grad_norm": 30.849266052246094,
"learning_rate": 8.405472181058766e-07,
"loss": 0.39806708693504333,
"step": 720
},
{
"epoch": 0.6269565217391304,
"grad_norm": 43.13644790649414,
"learning_rate": 8.371511725560416e-07,
"loss": 0.4927103519439697,
"step": 721
},
{
"epoch": 0.6278260869565218,
"grad_norm": 108.84441375732422,
"learning_rate": 8.3375854400814e-07,
"loss": 0.4438703954219818,
"step": 722
},
{
"epoch": 0.628695652173913,
"grad_norm": 272.2270812988281,
"learning_rate": 8.303693605417444e-07,
"loss": 0.44671696424484253,
"step": 723
},
{
"epoch": 0.6295652173913043,
"grad_norm": 13.665102005004883,
"learning_rate": 8.26983650207912e-07,
"loss": 0.19029748439788818,
"step": 724
},
{
"epoch": 0.6304347826086957,
"grad_norm": 30.839147567749023,
"learning_rate": 8.236014410289558e-07,
"loss": 0.40884655714035034,
"step": 725
},
{
"epoch": 0.6313043478260869,
"grad_norm": 31.748336791992188,
"learning_rate": 8.202227609982102e-07,
"loss": 0.2191096842288971,
"step": 726
},
{
"epoch": 0.6321739130434783,
"grad_norm": 15.656108856201172,
"learning_rate": 8.168476380798e-07,
"loss": 0.2678936719894409,
"step": 727
},
{
"epoch": 0.6330434782608696,
"grad_norm": 87.03897094726562,
"learning_rate": 8.134761002084093e-07,
"loss": 0.6398818492889404,
"step": 728
},
{
"epoch": 0.6339130434782608,
"grad_norm": 40.4995002746582,
"learning_rate": 8.1010817528905e-07,
"loss": 0.36810439825057983,
"step": 729
},
{
"epoch": 0.6347826086956522,
"grad_norm": 24.480937957763672,
"learning_rate": 8.067438911968305e-07,
"loss": 0.24613440036773682,
"step": 730
},
{
"epoch": 0.6356521739130435,
"grad_norm": 543.9742431640625,
"learning_rate": 8.033832757767261e-07,
"loss": 0.8075257539749146,
"step": 731
},
{
"epoch": 0.6365217391304347,
"grad_norm": 638.1714477539062,
"learning_rate": 8.00026356843347e-07,
"loss": 0.6318769454956055,
"step": 732
},
{
"epoch": 0.6373913043478261,
"grad_norm": 165.76344299316406,
"learning_rate": 7.9667316218071e-07,
"loss": 1.911271572113037,
"step": 733
},
{
"epoch": 0.6382608695652174,
"grad_norm": 55.17957305908203,
"learning_rate": 7.933237195420063e-07,
"loss": 0.2594420909881592,
"step": 734
},
{
"epoch": 0.6391304347826087,
"grad_norm": 39.89908218383789,
"learning_rate": 7.899780566493737e-07,
"loss": 0.5331699252128601,
"step": 735
},
{
"epoch": 0.64,
"grad_norm": 14.860429763793945,
"learning_rate": 7.866362011936657e-07,
"loss": 0.1340891420841217,
"step": 736
},
{
"epoch": 0.6408695652173914,
"grad_norm": 28.789398193359375,
"learning_rate": 7.832981808342238e-07,
"loss": 0.2346506118774414,
"step": 737
},
{
"epoch": 0.6417391304347826,
"grad_norm": 42.433109283447266,
"learning_rate": 7.799640231986476e-07,
"loss": 1.6639212369918823,
"step": 738
},
{
"epoch": 0.6426086956521739,
"grad_norm": 28.04407501220703,
"learning_rate": 7.766337558825655e-07,
"loss": 0.15990275144577026,
"step": 739
},
{
"epoch": 0.6434782608695652,
"grad_norm": 57.953121185302734,
"learning_rate": 7.733074064494084e-07,
"loss": 1.4669609069824219,
"step": 740
},
{
"epoch": 0.6443478260869565,
"grad_norm": 22.84865951538086,
"learning_rate": 7.699850024301792e-07,
"loss": 0.1251482218503952,
"step": 741
},
{
"epoch": 0.6452173913043479,
"grad_norm": 19.29429054260254,
"learning_rate": 7.666665713232266e-07,
"loss": 0.1471455693244934,
"step": 742
},
{
"epoch": 0.6460869565217391,
"grad_norm": 276.4205322265625,
"learning_rate": 7.633521405940167e-07,
"loss": 1.033355712890625,
"step": 743
},
{
"epoch": 0.6469565217391304,
"grad_norm": 14.182607650756836,
"learning_rate": 7.600417376749065e-07,
"loss": 0.14121916890144348,
"step": 744
},
{
"epoch": 0.6478260869565218,
"grad_norm": 100.85777282714844,
"learning_rate": 7.567353899649151e-07,
"loss": 0.9635795950889587,
"step": 745
},
{
"epoch": 0.648695652173913,
"grad_norm": 51.979434967041016,
"learning_rate": 7.534331248294988e-07,
"loss": 0.7416068911552429,
"step": 746
},
{
"epoch": 0.6495652173913044,
"grad_norm": 40.61363220214844,
"learning_rate": 7.50134969600324e-07,
"loss": 0.46879589557647705,
"step": 747
},
{
"epoch": 0.6504347826086957,
"grad_norm": 77.65470123291016,
"learning_rate": 7.468409515750408e-07,
"loss": 1.1316441297531128,
"step": 748
},
{
"epoch": 0.6513043478260869,
"grad_norm": 87.22828674316406,
"learning_rate": 7.435510980170571e-07,
"loss": 0.46902915835380554,
"step": 749
},
{
"epoch": 0.6521739130434783,
"grad_norm": 32.40696716308594,
"learning_rate": 7.402654361553124e-07,
"loss": 0.22087864577770233,
"step": 750
},
{
"epoch": 0.6530434782608696,
"grad_norm": 19.364961624145508,
"learning_rate": 7.369839931840543e-07,
"loss": 0.17575204372406006,
"step": 751
},
{
"epoch": 0.6539130434782608,
"grad_norm": 180.41539001464844,
"learning_rate": 7.337067962626109e-07,
"loss": 2.696441650390625,
"step": 752
},
{
"epoch": 0.6547826086956522,
"grad_norm": 17.09427261352539,
"learning_rate": 7.304338725151682e-07,
"loss": 0.1562216579914093,
"step": 753
},
{
"epoch": 0.6556521739130434,
"grad_norm": 39.40911865234375,
"learning_rate": 7.271652490305439e-07,
"loss": 2.157348394393921,
"step": 754
},
{
"epoch": 0.6565217391304348,
"grad_norm": 65.94867706298828,
"learning_rate": 7.239009528619645e-07,
"loss": 0.365697979927063,
"step": 755
},
{
"epoch": 0.6573913043478261,
"grad_norm": 19.496864318847656,
"learning_rate": 7.206410110268404e-07,
"loss": 0.26776450872421265,
"step": 756
},
{
"epoch": 0.6582608695652173,
"grad_norm": 41.736671447753906,
"learning_rate": 7.173854505065439e-07,
"loss": 0.29309695959091187,
"step": 757
},
{
"epoch": 0.6591304347826087,
"grad_norm": 28.019617080688477,
"learning_rate": 7.141342982461822e-07,
"loss": 0.260204017162323,
"step": 758
},
{
"epoch": 0.66,
"grad_norm": 139.86886596679688,
"learning_rate": 7.1088758115438e-07,
"loss": 0.4147573411464691,
"step": 759
},
{
"epoch": 0.6608695652173913,
"grad_norm": 27.24856185913086,
"learning_rate": 7.076453261030525e-07,
"loss": 0.24827474355697632,
"step": 760
},
{
"epoch": 0.6617391304347826,
"grad_norm": 16.776628494262695,
"learning_rate": 7.044075599271827e-07,
"loss": 0.594458818435669,
"step": 761
},
{
"epoch": 0.662608695652174,
"grad_norm": 50.973960876464844,
"learning_rate": 7.011743094246026e-07,
"loss": 0.4260900020599365,
"step": 762
},
{
"epoch": 0.6634782608695652,
"grad_norm": 80.60476684570312,
"learning_rate": 6.979456013557697e-07,
"loss": 0.21795789897441864,
"step": 763
},
{
"epoch": 0.6643478260869565,
"grad_norm": 43.76759719848633,
"learning_rate": 6.947214624435433e-07,
"loss": 3.43902850151062,
"step": 764
},
{
"epoch": 0.6652173913043479,
"grad_norm": 23.111234664916992,
"learning_rate": 6.915019193729677e-07,
"loss": 0.1663995385169983,
"step": 765
},
{
"epoch": 0.6660869565217391,
"grad_norm": 52.544776916503906,
"learning_rate": 6.882869987910482e-07,
"loss": 0.20054547488689423,
"step": 766
},
{
"epoch": 0.6669565217391304,
"grad_norm": 192.2727813720703,
"learning_rate": 6.850767273065319e-07,
"loss": 0.5959427952766418,
"step": 767
},
{
"epoch": 0.6678260869565218,
"grad_norm": 304.39874267578125,
"learning_rate": 6.818711314896853e-07,
"loss": 2.1625492572784424,
"step": 768
},
{
"epoch": 0.668695652173913,
"grad_norm": 337.6100158691406,
"learning_rate": 6.78670237872078e-07,
"loss": 1.0090219974517822,
"step": 769
},
{
"epoch": 0.6695652173913044,
"grad_norm": 29.8946475982666,
"learning_rate": 6.754740729463605e-07,
"loss": 0.1833086609840393,
"step": 770
},
{
"epoch": 0.6704347826086956,
"grad_norm": 36.62465286254883,
"learning_rate": 6.722826631660459e-07,
"loss": 0.44645485281944275,
"step": 771
},
{
"epoch": 0.671304347826087,
"grad_norm": 21.076013565063477,
"learning_rate": 6.690960349452894e-07,
"loss": 0.24147431552410126,
"step": 772
},
{
"epoch": 0.6721739130434783,
"grad_norm": 15.08715534210205,
"learning_rate": 6.659142146586725e-07,
"loss": 0.7292581796646118,
"step": 773
},
{
"epoch": 0.6730434782608695,
"grad_norm": 19.222023010253906,
"learning_rate": 6.627372286409829e-07,
"loss": 0.23283301293849945,
"step": 774
},
{
"epoch": 0.6739130434782609,
"grad_norm": 13.040077209472656,
"learning_rate": 6.595651031869954e-07,
"loss": 0.10966317355632782,
"step": 775
},
{
"epoch": 0.6747826086956522,
"grad_norm": 24.842538833618164,
"learning_rate": 6.563978645512576e-07,
"loss": 0.36463961005210876,
"step": 776
},
{
"epoch": 0.6756521739130434,
"grad_norm": 11.793009757995605,
"learning_rate": 6.532355389478689e-07,
"loss": 0.13175241649150848,
"step": 777
},
{
"epoch": 0.6765217391304348,
"grad_norm": 979.1425170898438,
"learning_rate": 6.500781525502674e-07,
"loss": 0.7108925580978394,
"step": 778
},
{
"epoch": 0.6773913043478261,
"grad_norm": 467.5769348144531,
"learning_rate": 6.469257314910085e-07,
"loss": 2.7429962158203125,
"step": 779
},
{
"epoch": 0.6782608695652174,
"grad_norm": 65.76836395263672,
"learning_rate": 6.437783018615533e-07,
"loss": 0.265237033367157,
"step": 780
},
{
"epoch": 0.6791304347826087,
"grad_norm": 74.19712829589844,
"learning_rate": 6.406358897120504e-07,
"loss": 1.0486525297164917,
"step": 781
},
{
"epoch": 0.68,
"grad_norm": 83.15819549560547,
"learning_rate": 6.37498521051119e-07,
"loss": 4.692592620849609,
"step": 782
},
{
"epoch": 0.6808695652173913,
"grad_norm": 142.2218017578125,
"learning_rate": 6.34366221845637e-07,
"loss": 2.3217759132385254,
"step": 783
},
{
"epoch": 0.6817391304347826,
"grad_norm": 125.5001449584961,
"learning_rate": 6.312390180205233e-07,
"loss": 0.7036529779434204,
"step": 784
},
{
"epoch": 0.6826086956521739,
"grad_norm": 513.9854125976562,
"learning_rate": 6.281169354585249e-07,
"loss": 0.7563449144363403,
"step": 785
},
{
"epoch": 0.6834782608695652,
"grad_norm": 24.0953369140625,
"learning_rate": 6.250000000000004e-07,
"loss": 0.16569630801677704,
"step": 786
},
{
"epoch": 0.6843478260869565,
"grad_norm": 15.65578842163086,
"learning_rate": 6.218882374427089e-07,
"loss": 0.15750561654567719,
"step": 787
},
{
"epoch": 0.6852173913043478,
"grad_norm": 11.407544136047363,
"learning_rate": 6.187816735415954e-07,
"loss": 0.22518837451934814,
"step": 788
},
{
"epoch": 0.6860869565217391,
"grad_norm": 32.0155143737793,
"learning_rate": 6.156803340085772e-07,
"loss": 0.19885610044002533,
"step": 789
},
{
"epoch": 0.6869565217391305,
"grad_norm": 123.2927017211914,
"learning_rate": 6.125842445123305e-07,
"loss": 1.346293330192566,
"step": 790
},
{
"epoch": 0.6878260869565217,
"grad_norm": 29.865440368652344,
"learning_rate": 6.094934306780799e-07,
"loss": 0.21172043681144714,
"step": 791
},
{
"epoch": 0.688695652173913,
"grad_norm": 53.6719970703125,
"learning_rate": 6.064079180873851e-07,
"loss": 0.6194855570793152,
"step": 792
},
{
"epoch": 0.6895652173913044,
"grad_norm": 19.695430755615234,
"learning_rate": 6.033277322779289e-07,
"loss": 0.506497859954834,
"step": 793
},
{
"epoch": 0.6904347826086956,
"grad_norm": 37.548980712890625,
"learning_rate": 6.002528987433068e-07,
"loss": 0.2043655812740326,
"step": 794
},
{
"epoch": 0.691304347826087,
"grad_norm": 133.07432556152344,
"learning_rate": 5.971834429328151e-07,
"loss": 0.5282706618309021,
"step": 795
},
{
"epoch": 0.6921739130434783,
"grad_norm": 63.763484954833984,
"learning_rate": 5.9411939025124e-07,
"loss": 0.48551690578460693,
"step": 796
},
{
"epoch": 0.6930434782608695,
"grad_norm": 22.28136444091797,
"learning_rate": 5.910607660586488e-07,
"loss": 3.107473373413086,
"step": 797
},
{
"epoch": 0.6939130434782609,
"grad_norm": 80.345703125,
"learning_rate": 5.880075956701792e-07,
"loss": 0.7340165972709656,
"step": 798
},
{
"epoch": 0.6947826086956522,
"grad_norm": 106.55564880371094,
"learning_rate": 5.849599043558295e-07,
"loss": 0.5571392178535461,
"step": 799
},
{
"epoch": 0.6956521739130435,
"grad_norm": 28.005748748779297,
"learning_rate": 5.819177173402487e-07,
"loss": 0.7100469470024109,
"step": 800
},
{
"epoch": 0.6965217391304348,
"grad_norm": 189.35089111328125,
"learning_rate": 5.7888105980253e-07,
"loss": 0.5779204964637756,
"step": 801
},
{
"epoch": 0.697391304347826,
"grad_norm": 35.16061782836914,
"learning_rate": 5.758499568760012e-07,
"loss": 0.32489460706710815,
"step": 802
},
{
"epoch": 0.6982608695652174,
"grad_norm": 27.95366668701172,
"learning_rate": 5.728244336480152e-07,
"loss": 0.836237907409668,
"step": 803
},
{
"epoch": 0.6991304347826087,
"grad_norm": 13.76199722290039,
"learning_rate": 5.698045151597446e-07,
"loss": 0.14633211493492126,
"step": 804
},
{
"epoch": 0.7,
"grad_norm": 37.43208694458008,
"learning_rate": 5.667902264059738e-07,
"loss": 0.34416839480400085,
"step": 805
},
{
"epoch": 0.7008695652173913,
"grad_norm": 104.45726776123047,
"learning_rate": 5.63781592334892e-07,
"loss": 4.663516521453857,
"step": 806
},
{
"epoch": 0.7017391304347826,
"grad_norm": 44.0323371887207,
"learning_rate": 5.607786378478855e-07,
"loss": 1.9291681051254272,
"step": 807
},
{
"epoch": 0.7026086956521739,
"grad_norm": 11.187407493591309,
"learning_rate": 5.577813877993337e-07,
"loss": 0.11268677562475204,
"step": 808
},
{
"epoch": 0.7034782608695652,
"grad_norm": 33.402320861816406,
"learning_rate": 5.547898669964026e-07,
"loss": 0.6048992872238159,
"step": 809
},
{
"epoch": 0.7043478260869566,
"grad_norm": 96.39419555664062,
"learning_rate": 5.518041001988379e-07,
"loss": 0.5013086795806885,
"step": 810
},
{
"epoch": 0.7052173913043478,
"grad_norm": 12.453415870666504,
"learning_rate": 5.488241121187632e-07,
"loss": 0.10803544521331787,
"step": 811
},
{
"epoch": 0.7060869565217391,
"grad_norm": 162.5682373046875,
"learning_rate": 5.458499274204724e-07,
"loss": 0.2619629502296448,
"step": 812
},
{
"epoch": 0.7069565217391305,
"grad_norm": 27.30363655090332,
"learning_rate": 5.428815707202283e-07,
"loss": 0.277357816696167,
"step": 813
},
{
"epoch": 0.7078260869565217,
"grad_norm": 15.934202194213867,
"learning_rate": 5.399190665860555e-07,
"loss": 0.12446855753660202,
"step": 814
},
{
"epoch": 0.7086956521739131,
"grad_norm": 23.342369079589844,
"learning_rate": 5.369624395375405e-07,
"loss": 0.16745714843273163,
"step": 815
},
{
"epoch": 0.7095652173913043,
"grad_norm": 21.166223526000977,
"learning_rate": 5.34011714045627e-07,
"loss": 0.35413119196891785,
"step": 816
},
{
"epoch": 0.7104347826086956,
"grad_norm": 34.772212982177734,
"learning_rate": 5.310669145324132e-07,
"loss": 0.30249226093292236,
"step": 817
},
{
"epoch": 0.711304347826087,
"grad_norm": 38.343257904052734,
"learning_rate": 5.281280653709503e-07,
"loss": 1.0035111904144287,
"step": 818
},
{
"epoch": 0.7121739130434782,
"grad_norm": 18.245254516601562,
"learning_rate": 5.251951908850407e-07,
"loss": 0.11645388603210449,
"step": 819
},
{
"epoch": 0.7130434782608696,
"grad_norm": 69.61096954345703,
"learning_rate": 5.22268315349037e-07,
"loss": 2.7902565002441406,
"step": 820
},
{
"epoch": 0.7139130434782609,
"grad_norm": 10.546656608581543,
"learning_rate": 5.193474629876392e-07,
"loss": 0.15585359930992126,
"step": 821
},
{
"epoch": 0.7147826086956521,
"grad_norm": 19.00957679748535,
"learning_rate": 5.164326579756971e-07,
"loss": 0.6916980147361755,
"step": 822
},
{
"epoch": 0.7156521739130435,
"grad_norm": 18.706710815429688,
"learning_rate": 5.13523924438008e-07,
"loss": 0.1863079071044922,
"step": 823
},
{
"epoch": 0.7165217391304348,
"grad_norm": 97.32372283935547,
"learning_rate": 5.106212864491185e-07,
"loss": 0.5349218845367432,
"step": 824
},
{
"epoch": 0.717391304347826,
"grad_norm": 21.3338565826416,
"learning_rate": 5.077247680331232e-07,
"loss": 0.2499256134033203,
"step": 825
},
{
"epoch": 0.7182608695652174,
"grad_norm": 14.833037376403809,
"learning_rate": 5.048343931634684e-07,
"loss": 0.7161562442779541,
"step": 826
},
{
"epoch": 0.7191304347826087,
"grad_norm": 55.144065856933594,
"learning_rate": 5.019501857627521e-07,
"loss": 0.3763906955718994,
"step": 827
},
{
"epoch": 0.72,
"grad_norm": 46.12803268432617,
"learning_rate": 4.990721697025262e-07,
"loss": 0.2610902786254883,
"step": 828
},
{
"epoch": 0.7208695652173913,
"grad_norm": 14.994999885559082,
"learning_rate": 4.962003688030994e-07,
"loss": 0.2125650942325592,
"step": 829
},
{
"epoch": 0.7217391304347827,
"grad_norm": 41.04648208618164,
"learning_rate": 4.933348068333399e-07,
"loss": 0.5806182622909546,
"step": 830
},
{
"epoch": 0.7226086956521739,
"grad_norm": 25.080535888671875,
"learning_rate": 4.904755075104774e-07,
"loss": 0.21986179053783417,
"step": 831
},
{
"epoch": 0.7234782608695652,
"grad_norm": 49.48688888549805,
"learning_rate": 4.876224944999089e-07,
"loss": 0.44575807452201843,
"step": 832
},
{
"epoch": 0.7243478260869565,
"grad_norm": 19.422386169433594,
"learning_rate": 4.847757914150018e-07,
"loss": 0.09860061854124069,
"step": 833
},
{
"epoch": 0.7252173913043478,
"grad_norm": 13.360942840576172,
"learning_rate": 4.819354218168988e-07,
"loss": 0.2722475528717041,
"step": 834
},
{
"epoch": 0.7260869565217392,
"grad_norm": 30.36014747619629,
"learning_rate": 4.791014092143212e-07,
"loss": 0.179307758808136,
"step": 835
},
{
"epoch": 0.7269565217391304,
"grad_norm": 134.56378173828125,
"learning_rate": 4.7627377706337713e-07,
"loss": 2.5140609741210938,
"step": 836
},
{
"epoch": 0.7278260869565217,
"grad_norm": 43.07133102416992,
"learning_rate": 4.734525487673659e-07,
"loss": 0.24129435420036316,
"step": 837
},
{
"epoch": 0.7286956521739131,
"grad_norm": 83.6541748046875,
"learning_rate": 4.706377476765832e-07,
"loss": 0.7725095152854919,
"step": 838
},
{
"epoch": 0.7295652173913043,
"grad_norm": 33.3406982421875,
"learning_rate": 4.6782939708813035e-07,
"loss": 3.358060359954834,
"step": 839
},
{
"epoch": 0.7304347826086957,
"grad_norm": 28.123666763305664,
"learning_rate": 4.650275202457197e-07,
"loss": 0.4341234564781189,
"step": 840
},
{
"epoch": 0.731304347826087,
"grad_norm": 62.922752380371094,
"learning_rate": 4.622321403394832e-07,
"loss": 0.44603094458580017,
"step": 841
},
{
"epoch": 0.7321739130434782,
"grad_norm": 31.46849250793457,
"learning_rate": 4.594432805057786e-07,
"loss": 0.22092771530151367,
"step": 842
},
{
"epoch": 0.7330434782608696,
"grad_norm": 91.02661895751953,
"learning_rate": 4.566609638270006e-07,
"loss": 1.3896340131759644,
"step": 843
},
{
"epoch": 0.7339130434782609,
"grad_norm": 103.98423767089844,
"learning_rate": 4.538852133313888e-07,
"loss": 0.8688936233520508,
"step": 844
},
{
"epoch": 0.7347826086956522,
"grad_norm": 184.05401611328125,
"learning_rate": 4.511160519928352e-07,
"loss": 0.35446828603744507,
"step": 845
},
{
"epoch": 0.7356521739130435,
"grad_norm": 343.9791259765625,
"learning_rate": 4.483535027306975e-07,
"loss": 0.5675253868103027,
"step": 846
},
{
"epoch": 0.7365217391304347,
"grad_norm": 384.6269226074219,
"learning_rate": 4.4559758840960656e-07,
"loss": 0.48597678542137146,
"step": 847
},
{
"epoch": 0.7373913043478261,
"grad_norm": 9.321802139282227,
"learning_rate": 4.4284833183927927e-07,
"loss": 0.16912364959716797,
"step": 848
},
{
"epoch": 0.7382608695652174,
"grad_norm": 13.2305326461792,
"learning_rate": 4.401057557743267e-07,
"loss": 0.10761234164237976,
"step": 849
},
{
"epoch": 0.7391304347826086,
"grad_norm": 48.739925384521484,
"learning_rate": 4.373698829140698e-07,
"loss": 0.330234557390213,
"step": 850
},
{
"epoch": 0.74,
"grad_norm": 15.21980094909668,
"learning_rate": 4.3464073590234896e-07,
"loss": 0.17559093236923218,
"step": 851
},
{
"epoch": 0.7408695652173913,
"grad_norm": 46.84281539916992,
"learning_rate": 4.319183373273361e-07,
"loss": 0.2624436318874359,
"step": 852
},
{
"epoch": 0.7417391304347826,
"grad_norm": 21.39683723449707,
"learning_rate": 4.2920270972134997e-07,
"loss": 0.22666919231414795,
"step": 853
},
{
"epoch": 0.7426086956521739,
"grad_norm": 65.08245086669922,
"learning_rate": 4.264938755606679e-07,
"loss": 0.4538169801235199,
"step": 854
},
{
"epoch": 0.7434782608695653,
"grad_norm": 48.59606170654297,
"learning_rate": 4.23791857265341e-07,
"loss": 0.30713391304016113,
"step": 855
},
{
"epoch": 0.7443478260869565,
"grad_norm": 361.3922119140625,
"learning_rate": 4.210966771990063e-07,
"loss": 0.6252782344818115,
"step": 856
},
{
"epoch": 0.7452173913043478,
"grad_norm": 16.02417755126953,
"learning_rate": 4.184083576687047e-07,
"loss": 0.233723446726799,
"step": 857
},
{
"epoch": 0.7460869565217392,
"grad_norm": 51.16859817504883,
"learning_rate": 4.157269209246946e-07,
"loss": 0.6555522084236145,
"step": 858
},
{
"epoch": 0.7469565217391304,
"grad_norm": 112.98526000976562,
"learning_rate": 4.130523891602682e-07,
"loss": 0.48441505432128906,
"step": 859
},
{
"epoch": 0.7478260869565218,
"grad_norm": 74.0064697265625,
"learning_rate": 4.1038478451156607e-07,
"loss": 0.9179386496543884,
"step": 860
},
{
"epoch": 0.7486956521739131,
"grad_norm": 29.101930618286133,
"learning_rate": 4.077241290573987e-07,
"loss": 0.20464551448822021,
"step": 861
},
{
"epoch": 0.7495652173913043,
"grad_norm": 14.749531745910645,
"learning_rate": 4.0507044481905707e-07,
"loss": 0.097953662276268,
"step": 862
},
{
"epoch": 0.7504347826086957,
"grad_norm": 24.57115936279297,
"learning_rate": 4.0242375376013635e-07,
"loss": 0.43444204330444336,
"step": 863
},
{
"epoch": 0.7513043478260869,
"grad_norm": 31.40062141418457,
"learning_rate": 3.9978407778635057e-07,
"loss": 2.5366439819335938,
"step": 864
},
{
"epoch": 0.7521739130434782,
"grad_norm": 14.996513366699219,
"learning_rate": 3.971514387453533e-07,
"loss": 0.5662025213241577,
"step": 865
},
{
"epoch": 0.7530434782608696,
"grad_norm": 1001.7551879882812,
"learning_rate": 3.9452585842655434e-07,
"loss": 3.2340145111083984,
"step": 866
},
{
"epoch": 0.7539130434782608,
"grad_norm": 24.1766414642334,
"learning_rate": 3.9190735856094245e-07,
"loss": 0.2192693054676056,
"step": 867
},
{
"epoch": 0.7547826086956522,
"grad_norm": 25.643310546875,
"learning_rate": 3.892959608209035e-07,
"loss": 0.29429829120635986,
"step": 868
},
{
"epoch": 0.7556521739130435,
"grad_norm": 21.222827911376953,
"learning_rate": 3.866916868200421e-07,
"loss": 1.9577664136886597,
"step": 869
},
{
"epoch": 0.7565217391304347,
"grad_norm": 14.45306396484375,
"learning_rate": 3.840945581130008e-07,
"loss": 0.13284695148468018,
"step": 870
},
{
"epoch": 0.7573913043478261,
"grad_norm": 16.601198196411133,
"learning_rate": 3.815045961952846e-07,
"loss": 0.13604813814163208,
"step": 871
},
{
"epoch": 0.7582608695652174,
"grad_norm": 26.860652923583984,
"learning_rate": 3.7892182250308135e-07,
"loss": 0.1866011619567871,
"step": 872
},
{
"epoch": 0.7591304347826087,
"grad_norm": 21.43741226196289,
"learning_rate": 3.763462584130835e-07,
"loss": 0.19510798156261444,
"step": 873
},
{
"epoch": 0.76,
"grad_norm": 14.244462013244629,
"learning_rate": 3.737779252423132e-07,
"loss": 0.23197822272777557,
"step": 874
},
{
"epoch": 0.7608695652173914,
"grad_norm": 38.0366325378418,
"learning_rate": 3.712168442479447e-07,
"loss": 0.2798696756362915,
"step": 875
},
{
"epoch": 0.7617391304347826,
"grad_norm": 86.11932373046875,
"learning_rate": 3.686630366271286e-07,
"loss": 0.6115829944610596,
"step": 876
},
{
"epoch": 0.7626086956521739,
"grad_norm": 35.93330764770508,
"learning_rate": 3.661165235168157e-07,
"loss": 0.30920931696891785,
"step": 877
},
{
"epoch": 0.7634782608695653,
"grad_norm": 114.89765167236328,
"learning_rate": 3.6357732599358364e-07,
"loss": 0.3880614638328552,
"step": 878
},
{
"epoch": 0.7643478260869565,
"grad_norm": 78.57884216308594,
"learning_rate": 3.610454650734614e-07,
"loss": 0.920116662979126,
"step": 879
},
{
"epoch": 0.7652173913043478,
"grad_norm": 160.9378662109375,
"learning_rate": 3.5852096171175496e-07,
"loss": 0.4273170232772827,
"step": 880
},
{
"epoch": 0.7660869565217391,
"grad_norm": 31.745664596557617,
"learning_rate": 3.560038368028748e-07,
"loss": 0.603442907333374,
"step": 881
},
{
"epoch": 0.7669565217391304,
"grad_norm": 16.943307876586914,
"learning_rate": 3.5349411118016275e-07,
"loss": 0.28441041707992554,
"step": 882
},
{
"epoch": 0.7678260869565218,
"grad_norm": 41.021705627441406,
"learning_rate": 3.5099180561571966e-07,
"loss": 0.21559254825115204,
"step": 883
},
{
"epoch": 0.768695652173913,
"grad_norm": 51.68123245239258,
"learning_rate": 3.4849694082023197e-07,
"loss": 0.3459382653236389,
"step": 884
},
{
"epoch": 0.7695652173913043,
"grad_norm": 52.75640106201172,
"learning_rate": 3.460095374428028e-07,
"loss": 0.4530704915523529,
"step": 885
},
{
"epoch": 0.7704347826086957,
"grad_norm": 20.905052185058594,
"learning_rate": 3.4352961607077976e-07,
"loss": 0.15954455733299255,
"step": 886
},
{
"epoch": 0.7713043478260869,
"grad_norm": 81.68609619140625,
"learning_rate": 3.4105719722958326e-07,
"loss": 0.4723793864250183,
"step": 887
},
{
"epoch": 0.7721739130434783,
"grad_norm": 208.91615295410156,
"learning_rate": 3.3859230138253945e-07,
"loss": 0.5330352783203125,
"step": 888
},
{
"epoch": 0.7730434782608696,
"grad_norm": 55.5845947265625,
"learning_rate": 3.361349489307085e-07,
"loss": 0.3786437213420868,
"step": 889
},
{
"epoch": 0.7739130434782608,
"grad_norm": 10.599721908569336,
"learning_rate": 3.336851602127174e-07,
"loss": 0.12644389271736145,
"step": 890
},
{
"epoch": 0.7747826086956522,
"grad_norm": 25.941383361816406,
"learning_rate": 3.3124295550458956e-07,
"loss": 0.6826114654541016,
"step": 891
},
{
"epoch": 0.7756521739130435,
"grad_norm": 234.76441955566406,
"learning_rate": 3.2880835501957917e-07,
"loss": 0.7245470285415649,
"step": 892
},
{
"epoch": 0.7765217391304348,
"grad_norm": 38.8992805480957,
"learning_rate": 3.2638137890800297e-07,
"loss": 0.3965342938899994,
"step": 893
},
{
"epoch": 0.7773913043478261,
"grad_norm": 14.15921688079834,
"learning_rate": 3.239620472570737e-07,
"loss": 0.11045745015144348,
"step": 894
},
{
"epoch": 0.7782608695652173,
"grad_norm": 153.725341796875,
"learning_rate": 3.2155038009073186e-07,
"loss": 0.25505897402763367,
"step": 895
},
{
"epoch": 0.7791304347826087,
"grad_norm": 56.865970611572266,
"learning_rate": 3.19146397369485e-07,
"loss": 2.3822150230407715,
"step": 896
},
{
"epoch": 0.78,
"grad_norm": 173.7816925048828,
"learning_rate": 3.1675011899023597e-07,
"loss": 1.72023606300354,
"step": 897
},
{
"epoch": 0.7808695652173913,
"grad_norm": 108.04584503173828,
"learning_rate": 3.143615647861235e-07,
"loss": 1.7656439542770386,
"step": 898
},
{
"epoch": 0.7817391304347826,
"grad_norm": 13.3953857421875,
"learning_rate": 3.1198075452635575e-07,
"loss": 0.17273323237895966,
"step": 899
},
{
"epoch": 0.782608695652174,
"grad_norm": 13.916914939880371,
"learning_rate": 3.0960770791604683e-07,
"loss": 0.12718284130096436,
"step": 900
},
{
"epoch": 0.7834782608695652,
"grad_norm": 18.278545379638672,
"learning_rate": 3.0724244459605355e-07,
"loss": 0.1581583023071289,
"step": 901
},
{
"epoch": 0.7843478260869565,
"grad_norm": 14.384292602539062,
"learning_rate": 3.0488498414281377e-07,
"loss": 0.1329336315393448,
"step": 902
},
{
"epoch": 0.7852173913043479,
"grad_norm": 38.441036224365234,
"learning_rate": 3.025353460681836e-07,
"loss": 0.2602463662624359,
"step": 903
},
{
"epoch": 0.7860869565217391,
"grad_norm": 13.173408508300781,
"learning_rate": 3.0019354981927663e-07,
"loss": 0.0946318507194519,
"step": 904
},
{
"epoch": 0.7869565217391304,
"grad_norm": 30.28549575805664,
"learning_rate": 2.978596147783011e-07,
"loss": 0.28622427582740784,
"step": 905
},
{
"epoch": 0.7878260869565218,
"grad_norm": 34.16470718383789,
"learning_rate": 2.9553356026240223e-07,
"loss": 0.32762569189071655,
"step": 906
},
{
"epoch": 0.788695652173913,
"grad_norm": 70.8321304321289,
"learning_rate": 2.9321540552350075e-07,
"loss": 0.5961053967475891,
"step": 907
},
{
"epoch": 0.7895652173913044,
"grad_norm": 25.97980308532715,
"learning_rate": 2.9090516974813285e-07,
"loss": 0.18774400651454926,
"step": 908
},
{
"epoch": 0.7904347826086957,
"grad_norm": 98.06208801269531,
"learning_rate": 2.886028720572932e-07,
"loss": 0.5097171068191528,
"step": 909
},
{
"epoch": 0.7913043478260869,
"grad_norm": 45.080291748046875,
"learning_rate": 2.863085315062755e-07,
"loss": 0.39569947123527527,
"step": 910
},
{
"epoch": 0.7921739130434783,
"grad_norm": 20.53847885131836,
"learning_rate": 2.840221670845153e-07,
"loss": 0.15165160596370697,
"step": 911
},
{
"epoch": 0.7930434782608695,
"grad_norm": 16.014846801757812,
"learning_rate": 2.817437977154319e-07,
"loss": 0.18782924115657806,
"step": 912
},
{
"epoch": 0.7939130434782609,
"grad_norm": 188.85281372070312,
"learning_rate": 2.7947344225627293e-07,
"loss": 0.3212510645389557,
"step": 913
},
{
"epoch": 0.7947826086956522,
"grad_norm": 26.141756057739258,
"learning_rate": 2.772111194979582e-07,
"loss": 0.4376610219478607,
"step": 914
},
{
"epoch": 0.7956521739130434,
"grad_norm": 141.9932861328125,
"learning_rate": 2.749568481649224e-07,
"loss": 0.20592373609542847,
"step": 915
},
{
"epoch": 0.7965217391304348,
"grad_norm": 15.722429275512695,
"learning_rate": 2.727106469149628e-07,
"loss": 0.2650931775569916,
"step": 916
},
{
"epoch": 0.7973913043478261,
"grad_norm": 15.70014762878418,
"learning_rate": 2.704725343390828e-07,
"loss": 0.12402468174695969,
"step": 917
},
{
"epoch": 0.7982608695652174,
"grad_norm": 47.780677795410156,
"learning_rate": 2.6824252896133913e-07,
"loss": 2.14485764503479,
"step": 918
},
{
"epoch": 0.7991304347826087,
"grad_norm": 475.0081481933594,
"learning_rate": 2.6602064923868745e-07,
"loss": 0.6344518661499023,
"step": 919
},
{
"epoch": 0.8,
"grad_norm": 18.999794006347656,
"learning_rate": 2.6380691356083095e-07,
"loss": 0.10200716555118561,
"step": 920
},
{
"epoch": 0.8008695652173913,
"grad_norm": 15.091756820678711,
"learning_rate": 2.6160134025006753e-07,
"loss": 0.20342788100242615,
"step": 921
},
{
"epoch": 0.8017391304347826,
"grad_norm": 30.662235260009766,
"learning_rate": 2.5940394756113723e-07,
"loss": 0.1940402090549469,
"step": 922
},
{
"epoch": 0.802608695652174,
"grad_norm": 333.36737060546875,
"learning_rate": 2.5721475368107293e-07,
"loss": 1.2806131839752197,
"step": 923
},
{
"epoch": 0.8034782608695652,
"grad_norm": 1181.9429931640625,
"learning_rate": 2.550337767290487e-07,
"loss": 2.2810826301574707,
"step": 924
},
{
"epoch": 0.8043478260869565,
"grad_norm": 392.4900207519531,
"learning_rate": 2.5286103475623006e-07,
"loss": 2.231879711151123,
"step": 925
},
{
"epoch": 0.8052173913043478,
"grad_norm": 76.45416259765625,
"learning_rate": 2.5069654574562384e-07,
"loss": 2.3904948234558105,
"step": 926
},
{
"epoch": 0.8060869565217391,
"grad_norm": 21.617643356323242,
"learning_rate": 2.4854032761193095e-07,
"loss": 0.5770492553710938,
"step": 927
},
{
"epoch": 0.8069565217391305,
"grad_norm": 19.855632781982422,
"learning_rate": 2.46392398201397e-07,
"loss": 0.21884986758232117,
"step": 928
},
{
"epoch": 0.8078260869565217,
"grad_norm": 1042.6053466796875,
"learning_rate": 2.4425277529166456e-07,
"loss": 2.133256196975708,
"step": 929
},
{
"epoch": 0.808695652173913,
"grad_norm": 21.168500900268555,
"learning_rate": 2.421214765916265e-07,
"loss": 0.2394556999206543,
"step": 930
},
{
"epoch": 0.8095652173913044,
"grad_norm": 73.12419128417969,
"learning_rate": 2.399985197412795e-07,
"loss": 0.748104453086853,
"step": 931
},
{
"epoch": 0.8104347826086956,
"grad_norm": 294.3224792480469,
"learning_rate": 2.3788392231157643e-07,
"loss": 0.4757748246192932,
"step": 932
},
{
"epoch": 0.811304347826087,
"grad_norm": 16.401180267333984,
"learning_rate": 2.3577770180428365e-07,
"loss": 0.21083730459213257,
"step": 933
},
{
"epoch": 0.8121739130434783,
"grad_norm": 195.46060180664062,
"learning_rate": 2.336798756518342e-07,
"loss": 1.0245471000671387,
"step": 934
},
{
"epoch": 0.8130434782608695,
"grad_norm": 21.594966888427734,
"learning_rate": 2.3159046121718446e-07,
"loss": 0.18150049448013306,
"step": 935
},
{
"epoch": 0.8139130434782609,
"grad_norm": 43.387813568115234,
"learning_rate": 2.295094757936689e-07,
"loss": 0.4377374053001404,
"step": 936
},
{
"epoch": 0.8147826086956522,
"grad_norm": 18.1859073638916,
"learning_rate": 2.2743693660485898e-07,
"loss": 0.11544045805931091,
"step": 937
},
{
"epoch": 0.8156521739130435,
"grad_norm": 25.381126403808594,
"learning_rate": 2.253728608044198e-07,
"loss": 0.1694791316986084,
"step": 938
},
{
"epoch": 0.8165217391304348,
"grad_norm": 21.008285522460938,
"learning_rate": 2.23317265475968e-07,
"loss": 0.14152568578720093,
"step": 939
},
{
"epoch": 0.8173913043478261,
"grad_norm": 12.06684398651123,
"learning_rate": 2.2127016763292959e-07,
"loss": 0.1699100136756897,
"step": 940
},
{
"epoch": 0.8182608695652174,
"grad_norm": 25.395811080932617,
"learning_rate": 2.1923158421840063e-07,
"loss": 0.23998036980628967,
"step": 941
},
{
"epoch": 0.8191304347826087,
"grad_norm": 40.64665222167969,
"learning_rate": 2.1720153210500666e-07,
"loss": 0.49315786361694336,
"step": 942
},
{
"epoch": 0.82,
"grad_norm": 14.912999153137207,
"learning_rate": 2.151800280947615e-07,
"loss": 0.10635095834732056,
"step": 943
},
{
"epoch": 0.8208695652173913,
"grad_norm": 770.4041137695312,
"learning_rate": 2.1316708891893037e-07,
"loss": 0.7287003397941589,
"step": 944
},
{
"epoch": 0.8217391304347826,
"grad_norm": 29.556732177734375,
"learning_rate": 2.1116273123789013e-07,
"loss": 0.20710718631744385,
"step": 945
},
{
"epoch": 0.8226086956521739,
"grad_norm": 156.5335235595703,
"learning_rate": 2.091669716409922e-07,
"loss": 2.0706892013549805,
"step": 946
},
{
"epoch": 0.8234782608695652,
"grad_norm": 9.638541221618652,
"learning_rate": 2.071798266464234e-07,
"loss": 0.09939944744110107,
"step": 947
},
{
"epoch": 0.8243478260869566,
"grad_norm": 23.857845306396484,
"learning_rate": 2.0520131270107186e-07,
"loss": 0.29403024911880493,
"step": 948
},
{
"epoch": 0.8252173913043478,
"grad_norm": 200.6664276123047,
"learning_rate": 2.0323144618038928e-07,
"loss": 0.34521716833114624,
"step": 949
},
{
"epoch": 0.8260869565217391,
"grad_norm": 31.494474411010742,
"learning_rate": 2.012702433882552e-07,
"loss": 0.30632948875427246,
"step": 950
},
{
"epoch": 0.8269565217391305,
"grad_norm": 62.87589645385742,
"learning_rate": 1.9931772055684306e-07,
"loss": 0.23135101795196533,
"step": 951
},
{
"epoch": 0.8278260869565217,
"grad_norm": 38.245269775390625,
"learning_rate": 1.973738938464853e-07,
"loss": 0.22904567420482635,
"step": 952
},
{
"epoch": 0.828695652173913,
"grad_norm": 23.652069091796875,
"learning_rate": 1.954387793455402e-07,
"loss": 0.1981177031993866,
"step": 953
},
{
"epoch": 0.8295652173913044,
"grad_norm": 59.89875793457031,
"learning_rate": 1.9351239307025677e-07,
"loss": 0.4341650605201721,
"step": 954
},
{
"epoch": 0.8304347826086956,
"grad_norm": 30.332138061523438,
"learning_rate": 1.9159475096464486e-07,
"loss": 0.3079957365989685,
"step": 955
},
{
"epoch": 0.831304347826087,
"grad_norm": 43.875213623046875,
"learning_rate": 1.896858689003421e-07,
"loss": 0.6518553495407104,
"step": 956
},
{
"epoch": 0.8321739130434782,
"grad_norm": 177.08558654785156,
"learning_rate": 1.8778576267648106e-07,
"loss": 0.4483970105648041,
"step": 957
},
{
"epoch": 0.8330434782608696,
"grad_norm": 84.51854705810547,
"learning_rate": 1.858944480195611e-07,
"loss": 1.7023810148239136,
"step": 958
},
{
"epoch": 0.8339130434782609,
"grad_norm": 379.30810546875,
"learning_rate": 1.8401194058331656e-07,
"loss": 1.0865992307662964,
"step": 959
},
{
"epoch": 0.8347826086956521,
"grad_norm": 47.01554870605469,
"learning_rate": 1.8213825594858763e-07,
"loss": 0.33431529998779297,
"step": 960
},
{
"epoch": 0.8356521739130435,
"grad_norm": 62.90696334838867,
"learning_rate": 1.8027340962319077e-07,
"loss": 0.25681382417678833,
"step": 961
},
{
"epoch": 0.8365217391304348,
"grad_norm": 57.070430755615234,
"learning_rate": 1.7841741704179155e-07,
"loss": 0.2682826519012451,
"step": 962
},
{
"epoch": 0.837391304347826,
"grad_norm": 99.64236450195312,
"learning_rate": 1.7657029356577622e-07,
"loss": 0.9251591563224792,
"step": 963
},
{
"epoch": 0.8382608695652174,
"grad_norm": 16.333833694458008,
"learning_rate": 1.7473205448312408e-07,
"loss": 0.17130306363105774,
"step": 964
},
{
"epoch": 0.8391304347826087,
"grad_norm": 9.948195457458496,
"learning_rate": 1.7290271500828236e-07,
"loss": 0.09407632052898407,
"step": 965
},
{
"epoch": 0.84,
"grad_norm": 68.90872955322266,
"learning_rate": 1.7108229028203888e-07,
"loss": 0.3542691171169281,
"step": 966
},
{
"epoch": 0.8408695652173913,
"grad_norm": 33.98243713378906,
"learning_rate": 1.6927079537139706e-07,
"loss": 0.9918839931488037,
"step": 967
},
{
"epoch": 0.8417391304347827,
"grad_norm": 70.26067352294922,
"learning_rate": 1.6746824526945163e-07,
"loss": 0.5727920532226562,
"step": 968
},
{
"epoch": 0.8426086956521739,
"grad_norm": 57.77675247192383,
"learning_rate": 1.656746548952648e-07,
"loss": 2.840130567550659,
"step": 969
},
{
"epoch": 0.8434782608695652,
"grad_norm": 20.764402389526367,
"learning_rate": 1.6389003909374188e-07,
"loss": 0.12498998641967773,
"step": 970
},
{
"epoch": 0.8443478260869566,
"grad_norm": 54.63706588745117,
"learning_rate": 1.621144126355087e-07,
"loss": 0.3165168762207031,
"step": 971
},
{
"epoch": 0.8452173913043478,
"grad_norm": 42.974571228027344,
"learning_rate": 1.6034779021678978e-07,
"loss": 0.46324342489242554,
"step": 972
},
{
"epoch": 0.8460869565217392,
"grad_norm": 22.0351619720459,
"learning_rate": 1.5859018645928647e-07,
"loss": 0.18558716773986816,
"step": 973
},
{
"epoch": 0.8469565217391304,
"grad_norm": 166.74853515625,
"learning_rate": 1.5684161591005596e-07,
"loss": 0.4066867232322693,
"step": 974
},
{
"epoch": 0.8478260869565217,
"grad_norm": 187.2076416015625,
"learning_rate": 1.551020930413902e-07,
"loss": 0.25349587202072144,
"step": 975
},
{
"epoch": 0.8486956521739131,
"grad_norm": 22.417333602905273,
"learning_rate": 1.5337163225069758e-07,
"loss": 0.5066444873809814,
"step": 976
},
{
"epoch": 0.8495652173913043,
"grad_norm": 19.15497589111328,
"learning_rate": 1.5165024786038248e-07,
"loss": 0.19603180885314941,
"step": 977
},
{
"epoch": 0.8504347826086956,
"grad_norm": 374.83221435546875,
"learning_rate": 1.499379541177269e-07,
"loss": 0.4640807509422302,
"step": 978
},
{
"epoch": 0.851304347826087,
"grad_norm": 27.334835052490234,
"learning_rate": 1.4823476519477327e-07,
"loss": 0.1612674742937088,
"step": 979
},
{
"epoch": 0.8521739130434782,
"grad_norm": 20.078479766845703,
"learning_rate": 1.465406951882066e-07,
"loss": 0.19494037330150604,
"step": 980
},
{
"epoch": 0.8530434782608696,
"grad_norm": 48.08445358276367,
"learning_rate": 1.4485575811923795e-07,
"loss": 0.7136974334716797,
"step": 981
},
{
"epoch": 0.8539130434782609,
"grad_norm": 23.6761474609375,
"learning_rate": 1.4317996793348778e-07,
"loss": 0.2031172215938568,
"step": 982
},
{
"epoch": 0.8547826086956521,
"grad_norm": 54.8809928894043,
"learning_rate": 1.4151333850087156e-07,
"loss": 1.6473133563995361,
"step": 983
},
{
"epoch": 0.8556521739130435,
"grad_norm": 145.7986602783203,
"learning_rate": 1.3985588361548493e-07,
"loss": 0.9481346011161804,
"step": 984
},
{
"epoch": 0.8565217391304348,
"grad_norm": 23.77951431274414,
"learning_rate": 1.3820761699548783e-07,
"loss": 0.2665241062641144,
"step": 985
},
{
"epoch": 0.8573913043478261,
"grad_norm": 20.761913299560547,
"learning_rate": 1.3656855228299334e-07,
"loss": 0.19814646244049072,
"step": 986
},
{
"epoch": 0.8582608695652174,
"grad_norm": 12.282244682312012,
"learning_rate": 1.349387030439532e-07,
"loss": 0.10802216827869415,
"step": 987
},
{
"epoch": 0.8591304347826086,
"grad_norm": 38.31984329223633,
"learning_rate": 1.3331808276804637e-07,
"loss": 0.41473907232284546,
"step": 988
},
{
"epoch": 0.86,
"grad_norm": 103.03263092041016,
"learning_rate": 1.3170670486856594e-07,
"loss": 0.5369864702224731,
"step": 989
},
{
"epoch": 0.8608695652173913,
"grad_norm": 17.439985275268555,
"learning_rate": 1.3010458268231024e-07,
"loss": 0.2077348530292511,
"step": 990
},
{
"epoch": 0.8617391304347826,
"grad_norm": 126.5348129272461,
"learning_rate": 1.2851172946947114e-07,
"loss": 0.3166094422340393,
"step": 991
},
{
"epoch": 0.8626086956521739,
"grad_norm": 20.52012062072754,
"learning_rate": 1.269281584135243e-07,
"loss": 0.20435956120491028,
"step": 992
},
{
"epoch": 0.8634782608695653,
"grad_norm": 41.801353454589844,
"learning_rate": 1.2535388262112023e-07,
"loss": 2.5809578895568848,
"step": 993
},
{
"epoch": 0.8643478260869565,
"grad_norm": 41.66728210449219,
"learning_rate": 1.2378891512197622e-07,
"loss": 0.19843876361846924,
"step": 994
},
{
"epoch": 0.8652173913043478,
"grad_norm": 26.75448226928711,
"learning_rate": 1.2223326886876802e-07,
"loss": 0.13447028398513794,
"step": 995
},
{
"epoch": 0.8660869565217392,
"grad_norm": 122.18651580810547,
"learning_rate": 1.2068695673702237e-07,
"loss": 1.2578997611999512,
"step": 996
},
{
"epoch": 0.8669565217391304,
"grad_norm": 127.31121063232422,
"learning_rate": 1.1914999152501106e-07,
"loss": 1.6063017845153809,
"step": 997
},
{
"epoch": 0.8678260869565217,
"grad_norm": 13.917580604553223,
"learning_rate": 1.1762238595364573e-07,
"loss": 0.11013950407505035,
"step": 998
},
{
"epoch": 0.8686956521739131,
"grad_norm": 11.51262378692627,
"learning_rate": 1.1610415266637005e-07,
"loss": 2.1191117763519287,
"step": 999
},
{
"epoch": 0.8695652173913043,
"grad_norm": 138.78277587890625,
"learning_rate": 1.1459530422905749e-07,
"loss": 1.2765779495239258,
"step": 1000
},
{
"epoch": 0.8704347826086957,
"grad_norm": 81.51953887939453,
"learning_rate": 1.1309585312990659e-07,
"loss": 1.0723211765289307,
"step": 1001
},
{
"epoch": 0.871304347826087,
"grad_norm": 37.29446029663086,
"learning_rate": 1.1160581177933686e-07,
"loss": 0.3659045100212097,
"step": 1002
},
{
"epoch": 0.8721739130434782,
"grad_norm": 44.24388885498047,
"learning_rate": 1.1012519250988743e-07,
"loss": 0.46708160638809204,
"step": 1003
},
{
"epoch": 0.8730434782608696,
"grad_norm": 19.889142990112305,
"learning_rate": 1.086540075761136e-07,
"loss": 0.15364417433738708,
"step": 1004
},
{
"epoch": 0.8739130434782608,
"grad_norm": 64.91902160644531,
"learning_rate": 1.0719226915448688e-07,
"loss": 0.45765507221221924,
"step": 1005
},
{
"epoch": 0.8747826086956522,
"grad_norm": 85.66766357421875,
"learning_rate": 1.0573998934329214e-07,
"loss": 1.5718563795089722,
"step": 1006
},
{
"epoch": 0.8756521739130435,
"grad_norm": 62.06061553955078,
"learning_rate": 1.0429718016252977e-07,
"loss": 0.683064877986908,
"step": 1007
},
{
"epoch": 0.8765217391304347,
"grad_norm": 32.65688705444336,
"learning_rate": 1.028638535538147e-07,
"loss": 0.3193044066429138,
"step": 1008
},
{
"epoch": 0.8773913043478261,
"grad_norm": 183.94735717773438,
"learning_rate": 1.0144002138027854e-07,
"loss": 1.448814868927002,
"step": 1009
},
{
"epoch": 0.8782608695652174,
"grad_norm": 36.874168395996094,
"learning_rate": 1.0002569542646972e-07,
"loss": 0.2153337299823761,
"step": 1010
},
{
"epoch": 0.8791304347826087,
"grad_norm": 356.7084045410156,
"learning_rate": 9.862088739825832e-08,
"loss": 1.8259605169296265,
"step": 1011
},
{
"epoch": 0.88,
"grad_norm": 47.07126998901367,
"learning_rate": 9.722560892273719e-08,
"loss": 2.571669101715088,
"step": 1012
},
{
"epoch": 0.8808695652173913,
"grad_norm": 51.92734909057617,
"learning_rate": 9.58398715481268e-08,
"loss": 0.23773817718029022,
"step": 1013
},
{
"epoch": 0.8817391304347826,
"grad_norm": 26.753847122192383,
"learning_rate": 9.446368674367906e-08,
"loss": 0.17599496245384216,
"step": 1014
},
{
"epoch": 0.8826086956521739,
"grad_norm": 72.0635986328125,
"learning_rate": 9.30970658995832e-08,
"loss": 0.2927898168563843,
"step": 1015
},
{
"epoch": 0.8834782608695653,
"grad_norm": 71.94131469726562,
"learning_rate": 9.174002032687063e-08,
"loss": 2.358839273452759,
"step": 1016
},
{
"epoch": 0.8843478260869565,
"grad_norm": 18.351661682128906,
"learning_rate": 9.039256125732143e-08,
"loss": 0.20827649533748627,
"step": 1017
},
{
"epoch": 0.8852173913043478,
"grad_norm": 49.1838264465332,
"learning_rate": 8.905469984337209e-08,
"loss": 2.6017208099365234,
"step": 1018
},
{
"epoch": 0.8860869565217391,
"grad_norm": 45.20689010620117,
"learning_rate": 8.772644715802275e-08,
"loss": 0.8399229049682617,
"step": 1019
},
{
"epoch": 0.8869565217391304,
"grad_norm": 171.54916381835938,
"learning_rate": 8.640781419474484e-08,
"loss": 5.565194129943848,
"step": 1020
},
{
"epoch": 0.8878260869565218,
"grad_norm": 131.12217712402344,
"learning_rate": 8.509881186739149e-08,
"loss": 1.9047163724899292,
"step": 1021
},
{
"epoch": 0.888695652173913,
"grad_norm": 38.416561126708984,
"learning_rate": 8.379945101010644e-08,
"loss": 0.22399160265922546,
"step": 1022
},
{
"epoch": 0.8895652173913043,
"grad_norm": 26.03103256225586,
"learning_rate": 8.250974237723433e-08,
"loss": 0.5413999557495117,
"step": 1023
},
{
"epoch": 0.8904347826086957,
"grad_norm": 83.9515380859375,
"learning_rate": 8.122969664323161e-08,
"loss": 0.6338434815406799,
"step": 1024
},
{
"epoch": 0.8913043478260869,
"grad_norm": 8.502718925476074,
"learning_rate": 7.995932440257842e-08,
"loss": 0.07525776326656342,
"step": 1025
},
{
"epoch": 0.8921739130434783,
"grad_norm": 7.0947957038879395,
"learning_rate": 7.86986361696912e-08,
"loss": 0.08530958741903305,
"step": 1026
},
{
"epoch": 0.8930434782608696,
"grad_norm": 26.290821075439453,
"learning_rate": 7.74476423788352e-08,
"loss": 0.24022355675697327,
"step": 1027
},
{
"epoch": 0.8939130434782608,
"grad_norm": 160.67538452148438,
"learning_rate": 7.620635338403779e-08,
"loss": 0.9223116636276245,
"step": 1028
},
{
"epoch": 0.8947826086956522,
"grad_norm": 220.85218811035156,
"learning_rate": 7.497477945900375e-08,
"loss": 0.8716686964035034,
"step": 1029
},
{
"epoch": 0.8956521739130435,
"grad_norm": 40.53730392456055,
"learning_rate": 7.375293079702958e-08,
"loss": 0.41684240102767944,
"step": 1030
},
{
"epoch": 0.8965217391304348,
"grad_norm": 21.480506896972656,
"learning_rate": 7.254081751091893e-08,
"loss": 0.22680559754371643,
"step": 1031
},
{
"epoch": 0.8973913043478261,
"grad_norm": 938.2930908203125,
"learning_rate": 7.133844963289976e-08,
"loss": 1.2768101692199707,
"step": 1032
},
{
"epoch": 0.8982608695652174,
"grad_norm": 13.128506660461426,
"learning_rate": 7.014583711454054e-08,
"loss": 0.13842159509658813,
"step": 1033
},
{
"epoch": 0.8991304347826087,
"grad_norm": 146.43795776367188,
"learning_rate": 6.896298982666791e-08,
"loss": 0.6112872362136841,
"step": 1034
},
{
"epoch": 0.9,
"grad_norm": 19.700376510620117,
"learning_rate": 6.778991755928555e-08,
"loss": 0.16835694015026093,
"step": 1035
},
{
"epoch": 0.9008695652173913,
"grad_norm": 28.842063903808594,
"learning_rate": 6.662663002149267e-08,
"loss": 0.15688148140907288,
"step": 1036
},
{
"epoch": 0.9017391304347826,
"grad_norm": 110.23502349853516,
"learning_rate": 6.547313684140342e-08,
"loss": 1.0011162757873535,
"step": 1037
},
{
"epoch": 0.9026086956521739,
"grad_norm": 12.960235595703125,
"learning_rate": 6.432944756606815e-08,
"loss": 0.10485135018825531,
"step": 1038
},
{
"epoch": 0.9034782608695652,
"grad_norm": 16.311511993408203,
"learning_rate": 6.319557166139329e-08,
"loss": 0.11134529113769531,
"step": 1039
},
{
"epoch": 0.9043478260869565,
"grad_norm": 51.93461990356445,
"learning_rate": 6.20715185120642e-08,
"loss": 1.3551883697509766,
"step": 1040
},
{
"epoch": 0.9052173913043479,
"grad_norm": 42.39728546142578,
"learning_rate": 6.095729742146575e-08,
"loss": 0.4451691508293152,
"step": 1041
},
{
"epoch": 0.9060869565217391,
"grad_norm": 15.628239631652832,
"learning_rate": 5.985291761160736e-08,
"loss": 0.0780700296163559,
"step": 1042
},
{
"epoch": 0.9069565217391304,
"grad_norm": 30.696569442749023,
"learning_rate": 5.875838822304525e-08,
"loss": 0.2690103352069855,
"step": 1043
},
{
"epoch": 0.9078260869565218,
"grad_norm": 25.735858917236328,
"learning_rate": 5.767371831480739e-08,
"loss": 0.6541193723678589,
"step": 1044
},
{
"epoch": 0.908695652173913,
"grad_norm": 19.593347549438477,
"learning_rate": 5.659891686431801e-08,
"loss": 0.15605857968330383,
"step": 1045
},
{
"epoch": 0.9095652173913044,
"grad_norm": 142.44479370117188,
"learning_rate": 5.553399276732402e-08,
"loss": 0.5841786861419678,
"step": 1046
},
{
"epoch": 0.9104347826086957,
"grad_norm": 14.684908866882324,
"learning_rate": 5.44789548378212e-08,
"loss": 0.2255973070859909,
"step": 1047
},
{
"epoch": 0.9113043478260869,
"grad_norm": 254.1807403564453,
"learning_rate": 5.343381180798038e-08,
"loss": 2.0936872959136963,
"step": 1048
},
{
"epoch": 0.9121739130434783,
"grad_norm": 70.35183715820312,
"learning_rate": 5.2398572328076217e-08,
"loss": 0.48743361234664917,
"step": 1049
},
{
"epoch": 0.9130434782608695,
"grad_norm": 48.76604461669922,
"learning_rate": 5.137324496641546e-08,
"loss": 0.7166627645492554,
"step": 1050
},
{
"epoch": 0.9139130434782609,
"grad_norm": 75.04993438720703,
"learning_rate": 5.035783820926565e-08,
"loss": 0.2657568156719208,
"step": 1051
},
{
"epoch": 0.9147826086956522,
"grad_norm": 10.705974578857422,
"learning_rate": 4.9352360460784705e-08,
"loss": 0.17824089527130127,
"step": 1052
},
{
"epoch": 0.9156521739130434,
"grad_norm": 17.205289840698242,
"learning_rate": 4.835682004295197e-08,
"loss": 0.11702036112546921,
"step": 1053
},
{
"epoch": 0.9165217391304348,
"grad_norm": 53.99785614013672,
"learning_rate": 4.7371225195499416e-08,
"loss": 1.031096339225769,
"step": 1054
},
{
"epoch": 0.9173913043478261,
"grad_norm": 30.115753173828125,
"learning_rate": 4.6395584075842325e-08,
"loss": 0.1978806108236313,
"step": 1055
},
{
"epoch": 0.9182608695652174,
"grad_norm": 24.67990493774414,
"learning_rate": 4.5429904759013154e-08,
"loss": 1.6106297969818115,
"step": 1056
},
{
"epoch": 0.9191304347826087,
"grad_norm": 25.526687622070312,
"learning_rate": 4.447419523759364e-08,
"loss": 0.17992830276489258,
"step": 1057
},
{
"epoch": 0.92,
"grad_norm": 9.904977798461914,
"learning_rate": 4.352846342164985e-08,
"loss": 0.07753743976354599,
"step": 1058
},
{
"epoch": 0.9208695652173913,
"grad_norm": 39.5291862487793,
"learning_rate": 4.2592717138664755e-08,
"loss": 0.4158613085746765,
"step": 1059
},
{
"epoch": 0.9217391304347826,
"grad_norm": 57.925018310546875,
"learning_rate": 4.1666964133475624e-08,
"loss": 0.7720763683319092,
"step": 1060
},
{
"epoch": 0.922608695652174,
"grad_norm": 118.61414337158203,
"learning_rate": 4.075121206820867e-08,
"loss": 0.6265521049499512,
"step": 1061
},
{
"epoch": 0.9234782608695652,
"grad_norm": 10.207385063171387,
"learning_rate": 3.984546852221605e-08,
"loss": 0.12685824930667877,
"step": 1062
},
{
"epoch": 0.9243478260869565,
"grad_norm": 37.19662094116211,
"learning_rate": 3.89497409920124e-08,
"loss": 0.48499342799186707,
"step": 1063
},
{
"epoch": 0.9252173913043479,
"grad_norm": 61.965572357177734,
"learning_rate": 3.806403689121399e-08,
"loss": 0.5296880006790161,
"step": 1064
},
{
"epoch": 0.9260869565217391,
"grad_norm": 21.725605010986328,
"learning_rate": 3.7188363550476616e-08,
"loss": 0.1394590437412262,
"step": 1065
},
{
"epoch": 0.9269565217391305,
"grad_norm": 16.654356002807617,
"learning_rate": 3.6322728217434985e-08,
"loss": 0.1430962085723877,
"step": 1066
},
{
"epoch": 0.9278260869565217,
"grad_norm": 18.601428985595703,
"learning_rate": 3.546713805664265e-08,
"loss": 0.16020169854164124,
"step": 1067
},
{
"epoch": 0.928695652173913,
"grad_norm": 25.137908935546875,
"learning_rate": 3.462160014951352e-08,
"loss": 0.17368170619010925,
"step": 1068
},
{
"epoch": 0.9295652173913044,
"grad_norm": 81.3788833618164,
"learning_rate": 3.3786121494261557e-08,
"loss": 1.6028938293457031,
"step": 1069
},
{
"epoch": 0.9304347826086956,
"grad_norm": 34.46930694580078,
"learning_rate": 3.296070900584466e-08,
"loss": 0.2389589548110962,
"step": 1070
},
{
"epoch": 0.931304347826087,
"grad_norm": 45.515052795410156,
"learning_rate": 3.2145369515906714e-08,
"loss": 0.4151531159877777,
"step": 1071
},
{
"epoch": 0.9321739130434783,
"grad_norm": 62.968353271484375,
"learning_rate": 3.134010977272048e-08,
"loss": 0.42569679021835327,
"step": 1072
},
{
"epoch": 0.9330434782608695,
"grad_norm": 30.476491928100586,
"learning_rate": 3.054493644113268e-08,
"loss": 0.5280812978744507,
"step": 1073
},
{
"epoch": 0.9339130434782609,
"grad_norm": 8.987117767333984,
"learning_rate": 2.975985610250823e-08,
"loss": 0.2796494662761688,
"step": 1074
},
{
"epoch": 0.9347826086956522,
"grad_norm": 24.991226196289062,
"learning_rate": 2.8984875254676054e-08,
"loss": 0.13575339317321777,
"step": 1075
},
{
"epoch": 0.9356521739130435,
"grad_norm": 128.30857849121094,
"learning_rate": 2.8220000311875163e-08,
"loss": 0.4232993721961975,
"step": 1076
},
{
"epoch": 0.9365217391304348,
"grad_norm": 19.623804092407227,
"learning_rate": 2.746523760470146e-08,
"loss": 0.3343874216079712,
"step": 1077
},
{
"epoch": 0.9373913043478261,
"grad_norm": 13.422455787658691,
"learning_rate": 2.6720593380055866e-08,
"loss": 0.19377243518829346,
"step": 1078
},
{
"epoch": 0.9382608695652174,
"grad_norm": 25.436166763305664,
"learning_rate": 2.598607380109211e-08,
"loss": 0.15421633422374725,
"step": 1079
},
{
"epoch": 0.9391304347826087,
"grad_norm": 11.200416564941406,
"learning_rate": 2.526168494716541e-08,
"loss": 0.09467314183712006,
"step": 1080
},
{
"epoch": 0.94,
"grad_norm": 44.611289978027344,
"learning_rate": 2.454743281378319e-08,
"loss": 0.39623236656188965,
"step": 1081
},
{
"epoch": 0.9408695652173913,
"grad_norm": 84.19932556152344,
"learning_rate": 2.3843323312554705e-08,
"loss": 0.8663069605827332,
"step": 1082
},
{
"epoch": 0.9417391304347826,
"grad_norm": 41.349239349365234,
"learning_rate": 2.3149362271141924e-08,
"loss": 0.22057053446769714,
"step": 1083
},
{
"epoch": 0.9426086956521739,
"grad_norm": 43.514793395996094,
"learning_rate": 2.246555543321219e-08,
"loss": 0.3831818997859955,
"step": 1084
},
{
"epoch": 0.9434782608695652,
"grad_norm": 8.41614055633545,
"learning_rate": 2.179190845838966e-08,
"loss": 0.11816976964473724,
"step": 1085
},
{
"epoch": 0.9443478260869566,
"grad_norm": 344.6490783691406,
"learning_rate": 2.112842692220951e-08,
"loss": 2.3136343955993652,
"step": 1086
},
{
"epoch": 0.9452173913043478,
"grad_norm": 41.778865814208984,
"learning_rate": 2.0475116316070603e-08,
"loss": 0.3573746085166931,
"step": 1087
},
{
"epoch": 0.9460869565217391,
"grad_norm": 12.83102798461914,
"learning_rate": 1.9831982047190806e-08,
"loss": 0.281247615814209,
"step": 1088
},
{
"epoch": 0.9469565217391305,
"grad_norm": 142.6578826904297,
"learning_rate": 1.9199029438562583e-08,
"loss": 0.8952249884605408,
"step": 1089
},
{
"epoch": 0.9478260869565217,
"grad_norm": 21.0310115814209,
"learning_rate": 1.8576263728907752e-08,
"loss": 0.13993816077709198,
"step": 1090
},
{
"epoch": 0.948695652173913,
"grad_norm": 27.782527923583984,
"learning_rate": 1.7963690072635294e-08,
"loss": 0.32220780849456787,
"step": 1091
},
{
"epoch": 0.9495652173913044,
"grad_norm": 61.29800033569336,
"learning_rate": 1.736131353979792e-08,
"loss": 0.9174994826316833,
"step": 1092
},
{
"epoch": 0.9504347826086956,
"grad_norm": 11.923197746276855,
"learning_rate": 1.6769139116050575e-08,
"loss": 0.24690890312194824,
"step": 1093
},
{
"epoch": 0.951304347826087,
"grad_norm": 33.5105094909668,
"learning_rate": 1.6187171702608946e-08,
"loss": 1.097381591796875,
"step": 1094
},
{
"epoch": 0.9521739130434783,
"grad_norm": 20.5721492767334,
"learning_rate": 1.5615416116208925e-08,
"loss": 0.19984576106071472,
"step": 1095
},
{
"epoch": 0.9530434782608695,
"grad_norm": 15.237557411193848,
"learning_rate": 1.505387708906694e-08,
"loss": 0.131778746843338,
"step": 1096
},
{
"epoch": 0.9539130434782609,
"grad_norm": 46.54298400878906,
"learning_rate": 1.4502559268840394e-08,
"loss": 0.19282548129558563,
"step": 1097
},
{
"epoch": 0.9547826086956521,
"grad_norm": 12.827240943908691,
"learning_rate": 1.3961467218589347e-08,
"loss": 0.14774011075496674,
"step": 1098
},
{
"epoch": 0.9556521739130435,
"grad_norm": 24.567174911499023,
"learning_rate": 1.3430605416739212e-08,
"loss": 1.1719119548797607,
"step": 1099
},
{
"epoch": 0.9565217391304348,
"grad_norm": 78.66695404052734,
"learning_rate": 1.2909978257042849e-08,
"loss": 0.35071104764938354,
"step": 1100
},
{
"epoch": 0.957391304347826,
"grad_norm": 101.13919830322266,
"learning_rate": 1.2399590048544908e-08,
"loss": 0.9971999526023865,
"step": 1101
},
{
"epoch": 0.9582608695652174,
"grad_norm": 10.783035278320312,
"learning_rate": 1.1899445015546018e-08,
"loss": 0.06760130077600479,
"step": 1102
},
{
"epoch": 0.9591304347826087,
"grad_norm": 10.997241973876953,
"learning_rate": 1.1409547297567408e-08,
"loss": 0.1593540608882904,
"step": 1103
},
{
"epoch": 0.96,
"grad_norm": 43.574588775634766,
"learning_rate": 1.0929900949317174e-08,
"loss": 0.35229095816612244,
"step": 1104
},
{
"epoch": 0.9608695652173913,
"grad_norm": 75.2266845703125,
"learning_rate": 1.0460509940656282e-08,
"loss": 0.2805933654308319,
"step": 1105
},
{
"epoch": 0.9617391304347827,
"grad_norm": 22.76318359375,
"learning_rate": 1.0001378156566238e-08,
"loss": 0.20996904373168945,
"step": 1106
},
{
"epoch": 0.9626086956521739,
"grad_norm": 12.350744247436523,
"learning_rate": 9.552509397116328e-09,
"loss": 0.14843696355819702,
"step": 1107
},
{
"epoch": 0.9634782608695652,
"grad_norm": 14.665205001831055,
"learning_rate": 9.113907377432534e-09,
"loss": 0.15522554516792297,
"step": 1108
},
{
"epoch": 0.9643478260869566,
"grad_norm": 86.0486068725586,
"learning_rate": 8.685575727666734e-09,
"loss": 0.4995441138744354,
"step": 1109
},
{
"epoch": 0.9652173913043478,
"grad_norm": 35.30141830444336,
"learning_rate": 8.26751799296685e-09,
"loss": 0.23674023151397705,
"step": 1110
},
{
"epoch": 0.9660869565217391,
"grad_norm": 38.97645568847656,
"learning_rate": 7.859737633446745e-09,
"loss": 2.3994851112365723,
"step": 1111
},
{
"epoch": 0.9669565217391304,
"grad_norm": 203.57070922851562,
"learning_rate": 7.462238024158742e-09,
"loss": 0.8177639245986938,
"step": 1112
},
{
"epoch": 0.9678260869565217,
"grad_norm": 15.906163215637207,
"learning_rate": 7.075022455064612e-09,
"loss": 0.1885625571012497,
"step": 1113
},
{
"epoch": 0.9686956521739131,
"grad_norm": 29.28157615661621,
"learning_rate": 6.6980941310092195e-09,
"loss": 0.17579418420791626,
"step": 1114
},
{
"epoch": 0.9695652173913043,
"grad_norm": 28.175338745117188,
"learning_rate": 6.3314561716930345e-09,
"loss": 0.19931229948997498,
"step": 1115
},
{
"epoch": 0.9704347826086956,
"grad_norm": 28.38059425354004,
"learning_rate": 5.975111611647294e-09,
"loss": 0.17908689379692078,
"step": 1116
},
{
"epoch": 0.971304347826087,
"grad_norm": 45.8642463684082,
"learning_rate": 5.6290634002084676e-09,
"loss": 0.3295555114746094,
"step": 1117
},
{
"epoch": 0.9721739130434782,
"grad_norm": 51.021018981933594,
"learning_rate": 5.293314401493555e-09,
"loss": 0.5562708973884583,
"step": 1118
},
{
"epoch": 0.9730434782608696,
"grad_norm": 14.835803985595703,
"learning_rate": 4.967867394376907e-09,
"loss": 0.12446411699056625,
"step": 1119
},
{
"epoch": 0.9739130434782609,
"grad_norm": 18.287609100341797,
"learning_rate": 4.652725072466918e-09,
"loss": 0.1523091197013855,
"step": 1120
},
{
"epoch": 0.9747826086956521,
"grad_norm": 120.3004150390625,
"learning_rate": 4.347890044083675e-09,
"loss": 1.0910179615020752,
"step": 1121
},
{
"epoch": 0.9756521739130435,
"grad_norm": 55.653377532958984,
"learning_rate": 4.053364832237589e-09,
"loss": 1.1721607446670532,
"step": 1122
},
{
"epoch": 0.9765217391304348,
"grad_norm": 26.870397567749023,
"learning_rate": 3.769151874608301e-09,
"loss": 2.370091438293457,
"step": 1123
},
{
"epoch": 0.9773913043478261,
"grad_norm": 27.775001525878906,
"learning_rate": 3.4952535235248386e-09,
"loss": 0.622848391532898,
"step": 1124
},
{
"epoch": 0.9782608695652174,
"grad_norm": 45.21672439575195,
"learning_rate": 3.2316720459457662e-09,
"loss": 0.950424075126648,
"step": 1125
},
{
"epoch": 0.9791304347826087,
"grad_norm": 11.92493724822998,
"learning_rate": 2.9784096234405922e-09,
"loss": 0.10984307527542114,
"step": 1126
},
{
"epoch": 0.98,
"grad_norm": 14.86650276184082,
"learning_rate": 2.7354683521715887e-09,
"loss": 0.1362229436635971,
"step": 1127
},
{
"epoch": 0.9808695652173913,
"grad_norm": 25.460845947265625,
"learning_rate": 2.5028502428768586e-09,
"loss": 0.19979998469352722,
"step": 1128
},
{
"epoch": 0.9817391304347826,
"grad_norm": 65.58799743652344,
"learning_rate": 2.28055722085313e-09,
"loss": 0.3446487784385681,
"step": 1129
},
{
"epoch": 0.9826086956521739,
"grad_norm": 31.524560928344727,
"learning_rate": 2.0685911259402113e-09,
"loss": 0.3461800813674927,
"step": 1130
},
{
"epoch": 0.9834782608695652,
"grad_norm": 35.86664962768555,
"learning_rate": 1.866953712505587e-09,
"loss": 0.217980295419693,
"step": 1131
},
{
"epoch": 0.9843478260869565,
"grad_norm": 85.88794708251953,
"learning_rate": 1.6756466494298462e-09,
"loss": 1.1602758169174194,
"step": 1132
},
{
"epoch": 0.9852173913043478,
"grad_norm": 26.239646911621094,
"learning_rate": 1.494671520092944e-09,
"loss": 1.4768943786621094,
"step": 1133
},
{
"epoch": 0.9860869565217392,
"grad_norm": 85.36619567871094,
"learning_rate": 1.3240298223615722e-09,
"loss": 0.3386858105659485,
"step": 1134
},
{
"epoch": 0.9869565217391304,
"grad_norm": 14.699938774108887,
"learning_rate": 1.1637229685758367e-09,
"loss": 0.14901739358901978,
"step": 1135
},
{
"epoch": 0.9878260869565217,
"grad_norm": 52.28146743774414,
"learning_rate": 1.0137522855380166e-09,
"loss": 0.5103245973587036,
"step": 1136
},
{
"epoch": 0.9886956521739131,
"grad_norm": 53.130882263183594,
"learning_rate": 8.741190145017397e-10,
"loss": 0.29068705439567566,
"step": 1137
},
{
"epoch": 0.9895652173913043,
"grad_norm": 21.53912925720215,
"learning_rate": 7.448243111619901e-10,
"loss": 0.24485062062740326,
"step": 1138
},
{
"epoch": 0.9904347826086957,
"grad_norm": 47.211883544921875,
"learning_rate": 6.258692456442838e-10,
"loss": 0.2894359827041626,
"step": 1139
},
{
"epoch": 0.991304347826087,
"grad_norm": 39.86217498779297,
"learning_rate": 5.172548024973134e-10,
"loss": 0.19115138053894043,
"step": 1140
},
{
"epoch": 0.9921739130434782,
"grad_norm": 28.460140228271484,
"learning_rate": 4.1898188068392744e-10,
"loss": 0.7010704278945923,
"step": 1141
},
{
"epoch": 0.9930434782608696,
"grad_norm": 61.9373664855957,
"learning_rate": 3.3105129357405285e-10,
"loss": 0.3816262483596802,
"step": 1142
},
{
"epoch": 0.9939130434782608,
"grad_norm": 10.007226943969727,
"learning_rate": 2.534637689373398e-10,
"loss": 0.1578034907579422,
"step": 1143
},
{
"epoch": 0.9947826086956522,
"grad_norm": 106.2221450805664,
"learning_rate": 1.8621994893816553e-10,
"loss": 0.3024161458015442,
"step": 1144
},
{
"epoch": 0.9956521739130435,
"grad_norm": 13.402230262756348,
"learning_rate": 1.293203901296669e-10,
"loss": 0.11111289262771606,
"step": 1145
},
{
"epoch": 0.9965217391304347,
"grad_norm": 26.003263473510742,
"learning_rate": 8.276556344888332e-11,
"loss": 0.3413895070552826,
"step": 1146
},
{
"epoch": 0.9973913043478261,
"grad_norm": 14.795051574707031,
"learning_rate": 4.655585421356468e-11,
"loss": 0.11750486493110657,
"step": 1147
},
{
"epoch": 0.9982608695652174,
"grad_norm": 26.09149932861328,
"learning_rate": 2.069156211856327e-11,
"loss": 0.23041675984859467,
"step": 1148
},
{
"epoch": 0.9991304347826087,
"grad_norm": 26.49252700805664,
"learning_rate": 5.172901233196914e-12,
"loss": 0.2072083055973053,
"step": 1149
},
{
"epoch": 1.0,
"grad_norm": 216.6020050048828,
"learning_rate": 0.0,
"loss": 0.846635103225708,
"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
}