9b-117 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
0d11299 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 3276,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002442002442002442,
"grad_norm": 2.5787229537963867,
"learning_rate": 1.829268292682927e-08,
"loss": 1.9266729354858398,
"step": 2
},
{
"epoch": 0.004884004884004884,
"grad_norm": 2.9470772743225098,
"learning_rate": 5.48780487804878e-08,
"loss": 2.098057270050049,
"step": 4
},
{
"epoch": 0.007326007326007326,
"grad_norm": 24.35196304321289,
"learning_rate": 9.146341463414634e-08,
"loss": 2.419158458709717,
"step": 6
},
{
"epoch": 0.009768009768009768,
"grad_norm": 13.176837921142578,
"learning_rate": 1.2804878048780488e-07,
"loss": 1.9603787660598755,
"step": 8
},
{
"epoch": 0.01221001221001221,
"grad_norm": 3.178710460662842,
"learning_rate": 1.6463414634146343e-07,
"loss": 2.1654319763183594,
"step": 10
},
{
"epoch": 0.014652014652014652,
"grad_norm": 6.474632263183594,
"learning_rate": 2.0121951219512198e-07,
"loss": 1.944001317024231,
"step": 12
},
{
"epoch": 0.017094017094017096,
"grad_norm": 2.9987783432006836,
"learning_rate": 2.378048780487805e-07,
"loss": 1.6851489543914795,
"step": 14
},
{
"epoch": 0.019536019536019536,
"grad_norm": 16.205787658691406,
"learning_rate": 2.74390243902439e-07,
"loss": 2.074493169784546,
"step": 16
},
{
"epoch": 0.02197802197802198,
"grad_norm": 28.914915084838867,
"learning_rate": 3.1097560975609756e-07,
"loss": 2.760730504989624,
"step": 18
},
{
"epoch": 0.02442002442002442,
"grad_norm": 5.000414848327637,
"learning_rate": 3.475609756097561e-07,
"loss": 1.890244960784912,
"step": 20
},
{
"epoch": 0.026862026862026864,
"grad_norm": 12.402643203735352,
"learning_rate": 3.8414634146341466e-07,
"loss": 2.127265691757202,
"step": 22
},
{
"epoch": 0.029304029304029304,
"grad_norm": 2.161700487136841,
"learning_rate": 4.207317073170732e-07,
"loss": 2.077669143676758,
"step": 24
},
{
"epoch": 0.031746031746031744,
"grad_norm": 2.237544059753418,
"learning_rate": 4.573170731707317e-07,
"loss": 1.9492441415786743,
"step": 26
},
{
"epoch": 0.03418803418803419,
"grad_norm": 3.4599592685699463,
"learning_rate": 4.939024390243903e-07,
"loss": 2.0288307666778564,
"step": 28
},
{
"epoch": 0.03663003663003663,
"grad_norm": 13.558135032653809,
"learning_rate": 5.304878048780488e-07,
"loss": 2.0438828468322754,
"step": 30
},
{
"epoch": 0.03907203907203907,
"grad_norm": 1.968347430229187,
"learning_rate": 5.670731707317073e-07,
"loss": 1.4716084003448486,
"step": 32
},
{
"epoch": 0.04151404151404151,
"grad_norm": 2.296098470687866,
"learning_rate": 6.036585365853659e-07,
"loss": 1.916358470916748,
"step": 34
},
{
"epoch": 0.04395604395604396,
"grad_norm": 4.799354553222656,
"learning_rate": 6.402439024390244e-07,
"loss": 1.500044345855713,
"step": 36
},
{
"epoch": 0.0463980463980464,
"grad_norm": 3.422872543334961,
"learning_rate": 6.768292682926829e-07,
"loss": 1.9229261875152588,
"step": 38
},
{
"epoch": 0.04884004884004884,
"grad_norm": 5.7842278480529785,
"learning_rate": 7.134146341463414e-07,
"loss": 1.8417783975601196,
"step": 40
},
{
"epoch": 0.05128205128205128,
"grad_norm": 1.7502555847167969,
"learning_rate": 7.5e-07,
"loss": 1.8713977336883545,
"step": 42
},
{
"epoch": 0.05372405372405373,
"grad_norm": 3.2419745922088623,
"learning_rate": 7.865853658536586e-07,
"loss": 1.8525761365890503,
"step": 44
},
{
"epoch": 0.05616605616605617,
"grad_norm": 4.0050272941589355,
"learning_rate": 8.231707317073171e-07,
"loss": 1.7802633047103882,
"step": 46
},
{
"epoch": 0.05860805860805861,
"grad_norm": 9.282805442810059,
"learning_rate": 8.597560975609755e-07,
"loss": 1.744933009147644,
"step": 48
},
{
"epoch": 0.06105006105006105,
"grad_norm": 11.279670715332031,
"learning_rate": 8.963414634146341e-07,
"loss": 1.6575506925582886,
"step": 50
},
{
"epoch": 0.06349206349206349,
"grad_norm": 4.024263381958008,
"learning_rate": 9.329268292682927e-07,
"loss": 1.7523654699325562,
"step": 52
},
{
"epoch": 0.06593406593406594,
"grad_norm": 9.84774398803711,
"learning_rate": 9.695121951219512e-07,
"loss": 1.5431915521621704,
"step": 54
},
{
"epoch": 0.06837606837606838,
"grad_norm": 2.255718469619751,
"learning_rate": 1.0060975609756098e-06,
"loss": 1.6816049814224243,
"step": 56
},
{
"epoch": 0.07081807081807082,
"grad_norm": 6.438508033752441,
"learning_rate": 1.0426829268292682e-06,
"loss": 1.6770493984222412,
"step": 58
},
{
"epoch": 0.07326007326007326,
"grad_norm": 3.231410026550293,
"learning_rate": 1.0792682926829268e-06,
"loss": 1.7219147682189941,
"step": 60
},
{
"epoch": 0.0757020757020757,
"grad_norm": 4.533494472503662,
"learning_rate": 1.1158536585365854e-06,
"loss": 1.3120664358139038,
"step": 62
},
{
"epoch": 0.07814407814407814,
"grad_norm": 1.4741380214691162,
"learning_rate": 1.152439024390244e-06,
"loss": 1.839809536933899,
"step": 64
},
{
"epoch": 0.08058608058608059,
"grad_norm": 5.4159135818481445,
"learning_rate": 1.1890243902439024e-06,
"loss": 1.4975742101669312,
"step": 66
},
{
"epoch": 0.08302808302808302,
"grad_norm": 1.5194345712661743,
"learning_rate": 1.225609756097561e-06,
"loss": 1.2997660636901855,
"step": 68
},
{
"epoch": 0.08547008547008547,
"grad_norm": 1.0652521848678589,
"learning_rate": 1.2621951219512194e-06,
"loss": 1.325761318206787,
"step": 70
},
{
"epoch": 0.08791208791208792,
"grad_norm": 1.5348917245864868,
"learning_rate": 1.298780487804878e-06,
"loss": 1.6478363275527954,
"step": 72
},
{
"epoch": 0.09035409035409035,
"grad_norm": 3.0826334953308105,
"learning_rate": 1.3353658536585366e-06,
"loss": 1.5671182870864868,
"step": 74
},
{
"epoch": 0.0927960927960928,
"grad_norm": 3.705549478530884,
"learning_rate": 1.3719512195121952e-06,
"loss": 1.528605341911316,
"step": 76
},
{
"epoch": 0.09523809523809523,
"grad_norm": 1.423170804977417,
"learning_rate": 1.4085365853658536e-06,
"loss": 1.433777928352356,
"step": 78
},
{
"epoch": 0.09768009768009768,
"grad_norm": 5.42734432220459,
"learning_rate": 1.4451219512195122e-06,
"loss": 1.3220820426940918,
"step": 80
},
{
"epoch": 0.10012210012210013,
"grad_norm": 13.278860092163086,
"learning_rate": 1.4817073170731708e-06,
"loss": 1.230900526046753,
"step": 82
},
{
"epoch": 0.10256410256410256,
"grad_norm": 9.673623085021973,
"learning_rate": 1.5182926829268292e-06,
"loss": 1.2397403717041016,
"step": 84
},
{
"epoch": 0.10500610500610501,
"grad_norm": 2.096724510192871,
"learning_rate": 1.5548780487804878e-06,
"loss": 1.429734230041504,
"step": 86
},
{
"epoch": 0.10744810744810745,
"grad_norm": 2.136627435684204,
"learning_rate": 1.5914634146341464e-06,
"loss": 1.6148204803466797,
"step": 88
},
{
"epoch": 0.10989010989010989,
"grad_norm": 0.8591102361679077,
"learning_rate": 1.628048780487805e-06,
"loss": 1.2777928113937378,
"step": 90
},
{
"epoch": 0.11233211233211234,
"grad_norm": 2.6411468982696533,
"learning_rate": 1.6646341463414637e-06,
"loss": 1.6813223361968994,
"step": 92
},
{
"epoch": 0.11477411477411477,
"grad_norm": 9.360551834106445,
"learning_rate": 1.7012195121951218e-06,
"loss": 1.2510548830032349,
"step": 94
},
{
"epoch": 0.11721611721611722,
"grad_norm": 1.5671329498291016,
"learning_rate": 1.7378048780487804e-06,
"loss": 1.549704909324646,
"step": 96
},
{
"epoch": 0.11965811965811966,
"grad_norm": 2.817790985107422,
"learning_rate": 1.774390243902439e-06,
"loss": 1.4941961765289307,
"step": 98
},
{
"epoch": 0.1221001221001221,
"grad_norm": 1.9618436098098755,
"learning_rate": 1.8109756097560976e-06,
"loss": 1.208415150642395,
"step": 100
},
{
"epoch": 0.12454212454212454,
"grad_norm": 2.0582926273345947,
"learning_rate": 1.847560975609756e-06,
"loss": 1.4008541107177734,
"step": 102
},
{
"epoch": 0.12698412698412698,
"grad_norm": 1.3583874702453613,
"learning_rate": 1.8841463414634146e-06,
"loss": 1.0804157257080078,
"step": 104
},
{
"epoch": 0.12942612942612944,
"grad_norm": 0.7717381119728088,
"learning_rate": 1.9207317073170733e-06,
"loss": 0.9862698316574097,
"step": 106
},
{
"epoch": 0.13186813186813187,
"grad_norm": 4.132621765136719,
"learning_rate": 1.9573170731707316e-06,
"loss": 1.4770206212997437,
"step": 108
},
{
"epoch": 0.1343101343101343,
"grad_norm": 3.617838144302368,
"learning_rate": 1.9939024390243905e-06,
"loss": 1.7098721265792847,
"step": 110
},
{
"epoch": 0.13675213675213677,
"grad_norm": 1.887754201889038,
"learning_rate": 2.030487804878049e-06,
"loss": 1.4937952756881714,
"step": 112
},
{
"epoch": 0.1391941391941392,
"grad_norm": 1.5456370115280151,
"learning_rate": 2.0670731707317072e-06,
"loss": 1.2423112392425537,
"step": 114
},
{
"epoch": 0.14163614163614163,
"grad_norm": 1.8167121410369873,
"learning_rate": 2.1036585365853656e-06,
"loss": 1.4288495779037476,
"step": 116
},
{
"epoch": 0.14407814407814407,
"grad_norm": 2.20339035987854,
"learning_rate": 2.1402439024390245e-06,
"loss": 1.3045105934143066,
"step": 118
},
{
"epoch": 0.14652014652014653,
"grad_norm": 1.4721053838729858,
"learning_rate": 2.176829268292683e-06,
"loss": 1.4094347953796387,
"step": 120
},
{
"epoch": 0.14896214896214896,
"grad_norm": 2.694047451019287,
"learning_rate": 2.2134146341463417e-06,
"loss": 1.4338178634643555,
"step": 122
},
{
"epoch": 0.1514041514041514,
"grad_norm": 2.1025125980377197,
"learning_rate": 2.25e-06,
"loss": 1.388542652130127,
"step": 124
},
{
"epoch": 0.15384615384615385,
"grad_norm": 2.5838522911071777,
"learning_rate": 2.2865853658536584e-06,
"loss": 1.2388739585876465,
"step": 126
},
{
"epoch": 0.1562881562881563,
"grad_norm": 1.6586570739746094,
"learning_rate": 2.3231707317073173e-06,
"loss": 1.4730716943740845,
"step": 128
},
{
"epoch": 0.15873015873015872,
"grad_norm": 1.9007089138031006,
"learning_rate": 2.3597560975609757e-06,
"loss": 1.3335144519805908,
"step": 130
},
{
"epoch": 0.16117216117216118,
"grad_norm": 2.044327974319458,
"learning_rate": 2.3963414634146345e-06,
"loss": 1.1245752573013306,
"step": 132
},
{
"epoch": 0.16361416361416362,
"grad_norm": 4.774345874786377,
"learning_rate": 2.432926829268293e-06,
"loss": 1.1957733631134033,
"step": 134
},
{
"epoch": 0.16605616605616605,
"grad_norm": 3.8634209632873535,
"learning_rate": 2.4695121951219513e-06,
"loss": 1.208402395248413,
"step": 136
},
{
"epoch": 0.1684981684981685,
"grad_norm": 2.314042329788208,
"learning_rate": 2.5060975609756097e-06,
"loss": 1.1570374965667725,
"step": 138
},
{
"epoch": 0.17094017094017094,
"grad_norm": 2.155163526535034,
"learning_rate": 2.5426829268292685e-06,
"loss": 1.1594734191894531,
"step": 140
},
{
"epoch": 0.17338217338217338,
"grad_norm": 1.5824326276779175,
"learning_rate": 2.579268292682927e-06,
"loss": 1.420199990272522,
"step": 142
},
{
"epoch": 0.17582417582417584,
"grad_norm": 1.5222316980361938,
"learning_rate": 2.6158536585365853e-06,
"loss": 1.1563900709152222,
"step": 144
},
{
"epoch": 0.17826617826617827,
"grad_norm": 2.792137384414673,
"learning_rate": 2.652439024390244e-06,
"loss": 1.0181738138198853,
"step": 146
},
{
"epoch": 0.1807081807081807,
"grad_norm": 3.344496965408325,
"learning_rate": 2.6890243902439025e-06,
"loss": 1.031973958015442,
"step": 148
},
{
"epoch": 0.18315018315018314,
"grad_norm": 1.3484665155410767,
"learning_rate": 2.7256097560975613e-06,
"loss": 1.370530366897583,
"step": 150
},
{
"epoch": 0.1855921855921856,
"grad_norm": 3.3972597122192383,
"learning_rate": 2.7621951219512197e-06,
"loss": 1.4999196529388428,
"step": 152
},
{
"epoch": 0.18803418803418803,
"grad_norm": 1.831746220588684,
"learning_rate": 2.798780487804878e-06,
"loss": 1.3834564685821533,
"step": 154
},
{
"epoch": 0.19047619047619047,
"grad_norm": 5.442370414733887,
"learning_rate": 2.8353658536585365e-06,
"loss": 1.667514443397522,
"step": 156
},
{
"epoch": 0.19291819291819293,
"grad_norm": 2.3951525688171387,
"learning_rate": 2.8719512195121953e-06,
"loss": 1.1068520545959473,
"step": 158
},
{
"epoch": 0.19536019536019536,
"grad_norm": 1.1164201498031616,
"learning_rate": 2.9085365853658537e-06,
"loss": 0.9049057960510254,
"step": 160
},
{
"epoch": 0.1978021978021978,
"grad_norm": 1.9392452239990234,
"learning_rate": 2.945121951219512e-06,
"loss": 1.3560075759887695,
"step": 162
},
{
"epoch": 0.20024420024420025,
"grad_norm": 1.669164776802063,
"learning_rate": 2.981707317073171e-06,
"loss": 1.2463322877883911,
"step": 164
},
{
"epoch": 0.2026862026862027,
"grad_norm": 1.7429471015930176,
"learning_rate": 3.0182926829268293e-06,
"loss": 1.3172338008880615,
"step": 166
},
{
"epoch": 0.20512820512820512,
"grad_norm": 2.3932223320007324,
"learning_rate": 3.054878048780488e-06,
"loss": 1.1728577613830566,
"step": 168
},
{
"epoch": 0.20757020757020758,
"grad_norm": 2.543950080871582,
"learning_rate": 3.0914634146341465e-06,
"loss": 0.945122480392456,
"step": 170
},
{
"epoch": 0.21001221001221002,
"grad_norm": 1.3421686887741089,
"learning_rate": 3.128048780487805e-06,
"loss": 1.0504443645477295,
"step": 172
},
{
"epoch": 0.21245421245421245,
"grad_norm": 1.8919559717178345,
"learning_rate": 3.1646341463414637e-06,
"loss": 1.332210659980774,
"step": 174
},
{
"epoch": 0.2148962148962149,
"grad_norm": 2.054534673690796,
"learning_rate": 3.201219512195122e-06,
"loss": 1.2866244316101074,
"step": 176
},
{
"epoch": 0.21733821733821734,
"grad_norm": 1.249266505241394,
"learning_rate": 3.237804878048781e-06,
"loss": 1.387539029121399,
"step": 178
},
{
"epoch": 0.21978021978021978,
"grad_norm": 1.3586231470108032,
"learning_rate": 3.2743902439024393e-06,
"loss": 0.9814241528511047,
"step": 180
},
{
"epoch": 0.2222222222222222,
"grad_norm": 5.887375354766846,
"learning_rate": 3.310975609756098e-06,
"loss": 0.9087549448013306,
"step": 182
},
{
"epoch": 0.22466422466422467,
"grad_norm": 1.151898741722107,
"learning_rate": 3.3475609756097565e-06,
"loss": 1.3200916051864624,
"step": 184
},
{
"epoch": 0.2271062271062271,
"grad_norm": 5.016763210296631,
"learning_rate": 3.3841463414634145e-06,
"loss": 1.3280826807022095,
"step": 186
},
{
"epoch": 0.22954822954822954,
"grad_norm": 1.9523217678070068,
"learning_rate": 3.420731707317073e-06,
"loss": 1.3782356977462769,
"step": 188
},
{
"epoch": 0.231990231990232,
"grad_norm": 2.2894701957702637,
"learning_rate": 3.4573170731707317e-06,
"loss": 1.3895543813705444,
"step": 190
},
{
"epoch": 0.23443223443223443,
"grad_norm": 2.341043710708618,
"learning_rate": 3.49390243902439e-06,
"loss": 1.3415981531143188,
"step": 192
},
{
"epoch": 0.23687423687423687,
"grad_norm": 2.133942127227783,
"learning_rate": 3.5304878048780485e-06,
"loss": 1.3378491401672363,
"step": 194
},
{
"epoch": 0.23931623931623933,
"grad_norm": 8.46426773071289,
"learning_rate": 3.5670731707317073e-06,
"loss": 1.0105526447296143,
"step": 196
},
{
"epoch": 0.24175824175824176,
"grad_norm": 2.380206823348999,
"learning_rate": 3.6036585365853657e-06,
"loss": 1.3137314319610596,
"step": 198
},
{
"epoch": 0.2442002442002442,
"grad_norm": 2.650566816329956,
"learning_rate": 3.6402439024390245e-06,
"loss": 1.4053981304168701,
"step": 200
},
{
"epoch": 0.24664224664224665,
"grad_norm": 1.5482836961746216,
"learning_rate": 3.676829268292683e-06,
"loss": 1.3324224948883057,
"step": 202
},
{
"epoch": 0.2490842490842491,
"grad_norm": 2.6213765144348145,
"learning_rate": 3.7134146341463417e-06,
"loss": 1.0133942365646362,
"step": 204
},
{
"epoch": 0.2515262515262515,
"grad_norm": 2.176896095275879,
"learning_rate": 3.75e-06,
"loss": 1.2807339429855347,
"step": 206
},
{
"epoch": 0.25396825396825395,
"grad_norm": 1.500716209411621,
"learning_rate": 3.7865853658536585e-06,
"loss": 1.1330087184906006,
"step": 208
},
{
"epoch": 0.2564102564102564,
"grad_norm": 6.454537868499756,
"learning_rate": 3.823170731707317e-06,
"loss": 0.9686428904533386,
"step": 210
},
{
"epoch": 0.2588522588522589,
"grad_norm": 3.82650089263916,
"learning_rate": 3.859756097560976e-06,
"loss": 0.6518009901046753,
"step": 212
},
{
"epoch": 0.2612942612942613,
"grad_norm": 1.4536477327346802,
"learning_rate": 3.896341463414634e-06,
"loss": 1.1154720783233643,
"step": 214
},
{
"epoch": 0.26373626373626374,
"grad_norm": 2.921241283416748,
"learning_rate": 3.932926829268293e-06,
"loss": 1.2995545864105225,
"step": 216
},
{
"epoch": 0.2661782661782662,
"grad_norm": 2.156456232070923,
"learning_rate": 3.969512195121952e-06,
"loss": 1.1956074237823486,
"step": 218
},
{
"epoch": 0.2686202686202686,
"grad_norm": 1.802549123764038,
"learning_rate": 4.00609756097561e-06,
"loss": 1.3290941715240479,
"step": 220
},
{
"epoch": 0.27106227106227104,
"grad_norm": 2.224459648132324,
"learning_rate": 4.0426829268292685e-06,
"loss": 0.9570510387420654,
"step": 222
},
{
"epoch": 0.27350427350427353,
"grad_norm": 1.6716811656951904,
"learning_rate": 4.079268292682927e-06,
"loss": 0.9313968420028687,
"step": 224
},
{
"epoch": 0.27594627594627597,
"grad_norm": 1.8278779983520508,
"learning_rate": 4.115853658536586e-06,
"loss": 1.3462107181549072,
"step": 226
},
{
"epoch": 0.2783882783882784,
"grad_norm": 2.975511312484741,
"learning_rate": 4.152439024390244e-06,
"loss": 1.3946359157562256,
"step": 228
},
{
"epoch": 0.28083028083028083,
"grad_norm": 1.9801675081253052,
"learning_rate": 4.189024390243902e-06,
"loss": 1.3629244565963745,
"step": 230
},
{
"epoch": 0.28327228327228327,
"grad_norm": 12.177047729492188,
"learning_rate": 4.2256097560975605e-06,
"loss": 0.9982041120529175,
"step": 232
},
{
"epoch": 0.2857142857142857,
"grad_norm": 3.7087788581848145,
"learning_rate": 4.26219512195122e-06,
"loss": 1.0232059955596924,
"step": 234
},
{
"epoch": 0.28815628815628813,
"grad_norm": 3.9933722019195557,
"learning_rate": 4.298780487804878e-06,
"loss": 1.2940337657928467,
"step": 236
},
{
"epoch": 0.2905982905982906,
"grad_norm": 1.5372240543365479,
"learning_rate": 4.3353658536585365e-06,
"loss": 1.266202688217163,
"step": 238
},
{
"epoch": 0.29304029304029305,
"grad_norm": 1.9267133474349976,
"learning_rate": 4.371951219512195e-06,
"loss": 1.1469693183898926,
"step": 240
},
{
"epoch": 0.2954822954822955,
"grad_norm": 2.542509078979492,
"learning_rate": 4.408536585365853e-06,
"loss": 0.9890923500061035,
"step": 242
},
{
"epoch": 0.2979242979242979,
"grad_norm": 1.5382633209228516,
"learning_rate": 4.4451219512195125e-06,
"loss": 1.3382046222686768,
"step": 244
},
{
"epoch": 0.30036630036630035,
"grad_norm": 1.2644575834274292,
"learning_rate": 4.481707317073171e-06,
"loss": 0.774017333984375,
"step": 246
},
{
"epoch": 0.3028083028083028,
"grad_norm": 5.0008769035339355,
"learning_rate": 4.518292682926829e-06,
"loss": 1.216355323791504,
"step": 248
},
{
"epoch": 0.3052503052503053,
"grad_norm": 2.470137596130371,
"learning_rate": 4.554878048780488e-06,
"loss": 0.9469658136367798,
"step": 250
},
{
"epoch": 0.3076923076923077,
"grad_norm": 2.0575084686279297,
"learning_rate": 4.591463414634147e-06,
"loss": 1.045383334159851,
"step": 252
},
{
"epoch": 0.31013431013431014,
"grad_norm": 2.9232146739959717,
"learning_rate": 4.628048780487805e-06,
"loss": 1.3018403053283691,
"step": 254
},
{
"epoch": 0.3125763125763126,
"grad_norm": 1.5343422889709473,
"learning_rate": 4.664634146341464e-06,
"loss": 1.282840609550476,
"step": 256
},
{
"epoch": 0.315018315018315,
"grad_norm": 2.349189043045044,
"learning_rate": 4.701219512195122e-06,
"loss": 1.33484947681427,
"step": 258
},
{
"epoch": 0.31746031746031744,
"grad_norm": 2.369356393814087,
"learning_rate": 4.7378048780487805e-06,
"loss": 1.2917380332946777,
"step": 260
},
{
"epoch": 0.3199023199023199,
"grad_norm": 5.888512134552002,
"learning_rate": 4.77439024390244e-06,
"loss": 1.5345417261123657,
"step": 262
},
{
"epoch": 0.32234432234432236,
"grad_norm": 5.536413669586182,
"learning_rate": 4.810975609756098e-06,
"loss": 0.6992707252502441,
"step": 264
},
{
"epoch": 0.3247863247863248,
"grad_norm": 3.4682443141937256,
"learning_rate": 4.8475609756097565e-06,
"loss": 1.0846771001815796,
"step": 266
},
{
"epoch": 0.32722832722832723,
"grad_norm": 2.0366370677948,
"learning_rate": 4.884146341463414e-06,
"loss": 1.337081789970398,
"step": 268
},
{
"epoch": 0.32967032967032966,
"grad_norm": 5.923638343811035,
"learning_rate": 4.920731707317073e-06,
"loss": 0.9941681027412415,
"step": 270
},
{
"epoch": 0.3321123321123321,
"grad_norm": 1.7837181091308594,
"learning_rate": 4.957317073170732e-06,
"loss": 1.2944457530975342,
"step": 272
},
{
"epoch": 0.33455433455433453,
"grad_norm": 4.933067321777344,
"learning_rate": 4.99390243902439e-06,
"loss": 0.9657446146011353,
"step": 274
},
{
"epoch": 0.336996336996337,
"grad_norm": 4.120548725128174,
"learning_rate": 5.0304878048780485e-06,
"loss": 1.0117385387420654,
"step": 276
},
{
"epoch": 0.33943833943833945,
"grad_norm": 8.162941932678223,
"learning_rate": 5.067073170731707e-06,
"loss": 1.4234728813171387,
"step": 278
},
{
"epoch": 0.3418803418803419,
"grad_norm": 1.219032883644104,
"learning_rate": 5.103658536585366e-06,
"loss": 1.3093342781066895,
"step": 280
},
{
"epoch": 0.3443223443223443,
"grad_norm": 2.771096706390381,
"learning_rate": 5.1402439024390245e-06,
"loss": 1.2924383878707886,
"step": 282
},
{
"epoch": 0.34676434676434675,
"grad_norm": 1.962974190711975,
"learning_rate": 5.176829268292683e-06,
"loss": 1.3206619024276733,
"step": 284
},
{
"epoch": 0.3492063492063492,
"grad_norm": 3.3968982696533203,
"learning_rate": 5.213414634146341e-06,
"loss": 1.1162430047988892,
"step": 286
},
{
"epoch": 0.3516483516483517,
"grad_norm": 3.5822386741638184,
"learning_rate": 5.2500000000000006e-06,
"loss": 1.673839807510376,
"step": 288
},
{
"epoch": 0.3540903540903541,
"grad_norm": 7.095156192779541,
"learning_rate": 5.286585365853659e-06,
"loss": 0.9515596628189087,
"step": 290
},
{
"epoch": 0.35653235653235654,
"grad_norm": 2.219644784927368,
"learning_rate": 5.323170731707317e-06,
"loss": 1.309720516204834,
"step": 292
},
{
"epoch": 0.358974358974359,
"grad_norm": 1.6270936727523804,
"learning_rate": 5.359756097560976e-06,
"loss": 1.0194257497787476,
"step": 294
},
{
"epoch": 0.3614163614163614,
"grad_norm": 1.4493879079818726,
"learning_rate": 5.396341463414634e-06,
"loss": 1.3390767574310303,
"step": 296
},
{
"epoch": 0.36385836385836384,
"grad_norm": 17.245960235595703,
"learning_rate": 5.432926829268293e-06,
"loss": 1.389981746673584,
"step": 298
},
{
"epoch": 0.3663003663003663,
"grad_norm": 2.71707820892334,
"learning_rate": 5.469512195121952e-06,
"loss": 1.2446229457855225,
"step": 300
},
{
"epoch": 0.36874236874236876,
"grad_norm": 3.7474751472473145,
"learning_rate": 5.50609756097561e-06,
"loss": 0.9005425572395325,
"step": 302
},
{
"epoch": 0.3711843711843712,
"grad_norm": 2.9197535514831543,
"learning_rate": 5.5426829268292685e-06,
"loss": 1.202176570892334,
"step": 304
},
{
"epoch": 0.37362637362637363,
"grad_norm": 2.3822195529937744,
"learning_rate": 5.579268292682927e-06,
"loss": 1.329774260520935,
"step": 306
},
{
"epoch": 0.37606837606837606,
"grad_norm": 1.2802468538284302,
"learning_rate": 5.615853658536586e-06,
"loss": 1.3126323223114014,
"step": 308
},
{
"epoch": 0.3785103785103785,
"grad_norm": 1.8697741031646729,
"learning_rate": 5.652439024390244e-06,
"loss": 1.345770239830017,
"step": 310
},
{
"epoch": 0.38095238095238093,
"grad_norm": 4.083343505859375,
"learning_rate": 5.689024390243902e-06,
"loss": 0.9912192821502686,
"step": 312
},
{
"epoch": 0.3833943833943834,
"grad_norm": 6.874935150146484,
"learning_rate": 5.7256097560975605e-06,
"loss": 1.3861531019210815,
"step": 314
},
{
"epoch": 0.38583638583638585,
"grad_norm": 2.4264907836914062,
"learning_rate": 5.76219512195122e-06,
"loss": 0.6633478999137878,
"step": 316
},
{
"epoch": 0.3882783882783883,
"grad_norm": 4.533822536468506,
"learning_rate": 5.798780487804878e-06,
"loss": 0.9467814564704895,
"step": 318
},
{
"epoch": 0.3907203907203907,
"grad_norm": 2.7442786693573,
"learning_rate": 5.8353658536585365e-06,
"loss": 1.3469103574752808,
"step": 320
},
{
"epoch": 0.39316239316239315,
"grad_norm": 4.531688213348389,
"learning_rate": 5.871951219512195e-06,
"loss": 1.3630614280700684,
"step": 322
},
{
"epoch": 0.3956043956043956,
"grad_norm": 2.2029385566711426,
"learning_rate": 5.908536585365853e-06,
"loss": 1.3367185592651367,
"step": 324
},
{
"epoch": 0.398046398046398,
"grad_norm": 2.591531276702881,
"learning_rate": 5.9451219512195126e-06,
"loss": 1.2546979188919067,
"step": 326
},
{
"epoch": 0.4004884004884005,
"grad_norm": 3.0475008487701416,
"learning_rate": 5.981707317073171e-06,
"loss": 1.2662358283996582,
"step": 328
},
{
"epoch": 0.40293040293040294,
"grad_norm": 3.1861870288848877,
"learning_rate": 5.9999984668716506e-06,
"loss": 0.9014570116996765,
"step": 330
},
{
"epoch": 0.4053724053724054,
"grad_norm": 5.928226470947266,
"learning_rate": 5.9999862018553e-06,
"loss": 0.6253970861434937,
"step": 332
},
{
"epoch": 0.4078144078144078,
"grad_norm": 1.3099563121795654,
"learning_rate": 5.9999616718783144e-06,
"loss": 1.0681229829788208,
"step": 334
},
{
"epoch": 0.41025641025641024,
"grad_norm": 2.3551321029663086,
"learning_rate": 5.999924877052124e-06,
"loss": 1.3677372932434082,
"step": 336
},
{
"epoch": 0.4126984126984127,
"grad_norm": 1.5692805051803589,
"learning_rate": 5.999875817543871e-06,
"loss": 0.9468737840652466,
"step": 338
},
{
"epoch": 0.41514041514041516,
"grad_norm": 25.2011661529541,
"learning_rate": 5.999814493576417e-06,
"loss": 1.0896673202514648,
"step": 340
},
{
"epoch": 0.4175824175824176,
"grad_norm": 2.671222686767578,
"learning_rate": 5.999740905428328e-06,
"loss": 1.6776376962661743,
"step": 342
},
{
"epoch": 0.42002442002442003,
"grad_norm": 1.7033958435058594,
"learning_rate": 5.999655053433888e-06,
"loss": 0.9407004117965698,
"step": 344
},
{
"epoch": 0.42246642246642246,
"grad_norm": 2.3176705837249756,
"learning_rate": 5.999556937983088e-06,
"loss": 1.1251733303070068,
"step": 346
},
{
"epoch": 0.4249084249084249,
"grad_norm": 2.094940662384033,
"learning_rate": 5.999446559521627e-06,
"loss": 1.3193105459213257,
"step": 348
},
{
"epoch": 0.42735042735042733,
"grad_norm": 3.736506462097168,
"learning_rate": 5.999323918550911e-06,
"loss": 0.863933801651001,
"step": 350
},
{
"epoch": 0.4297924297924298,
"grad_norm": 1.4415168762207031,
"learning_rate": 5.999189015628049e-06,
"loss": 1.0673351287841797,
"step": 352
},
{
"epoch": 0.43223443223443225,
"grad_norm": 2.9786107540130615,
"learning_rate": 5.999041851365851e-06,
"loss": 1.2724120616912842,
"step": 354
},
{
"epoch": 0.4346764346764347,
"grad_norm": 1.8949273824691772,
"learning_rate": 5.998882426432824e-06,
"loss": 0.8991950154304504,
"step": 356
},
{
"epoch": 0.4371184371184371,
"grad_norm": 3.24233078956604,
"learning_rate": 5.998710741553175e-06,
"loss": 1.1615791320800781,
"step": 358
},
{
"epoch": 0.43956043956043955,
"grad_norm": 1.7800151109695435,
"learning_rate": 5.9985267975067975e-06,
"loss": 1.2164958715438843,
"step": 360
},
{
"epoch": 0.442002442002442,
"grad_norm": 6.551607608795166,
"learning_rate": 5.998330595129277e-06,
"loss": 1.063248634338379,
"step": 362
},
{
"epoch": 0.4444444444444444,
"grad_norm": 1.5826555490493774,
"learning_rate": 5.9981221353118815e-06,
"loss": 1.1169027090072632,
"step": 364
},
{
"epoch": 0.4468864468864469,
"grad_norm": 2.175288200378418,
"learning_rate": 5.9979014190015636e-06,
"loss": 1.3833471536636353,
"step": 366
},
{
"epoch": 0.44932844932844934,
"grad_norm": 1.5917333364486694,
"learning_rate": 5.9976684472009465e-06,
"loss": 1.027613878250122,
"step": 368
},
{
"epoch": 0.4517704517704518,
"grad_norm": 2.8905162811279297,
"learning_rate": 5.997423220968328e-06,
"loss": 1.0340604782104492,
"step": 370
},
{
"epoch": 0.4542124542124542,
"grad_norm": 1.940250277519226,
"learning_rate": 5.997165741417676e-06,
"loss": 1.3127802610397339,
"step": 372
},
{
"epoch": 0.45665445665445664,
"grad_norm": 2.4297854900360107,
"learning_rate": 5.9968960097186155e-06,
"loss": 0.8877848386764526,
"step": 374
},
{
"epoch": 0.4590964590964591,
"grad_norm": 2.4118640422821045,
"learning_rate": 5.996614027096429e-06,
"loss": 1.2320910692214966,
"step": 376
},
{
"epoch": 0.46153846153846156,
"grad_norm": 4.578921318054199,
"learning_rate": 5.996319794832055e-06,
"loss": 0.9276782870292664,
"step": 378
},
{
"epoch": 0.463980463980464,
"grad_norm": 2.2359204292297363,
"learning_rate": 5.996013314262069e-06,
"loss": 0.8818802833557129,
"step": 380
},
{
"epoch": 0.46642246642246643,
"grad_norm": 1.6342476606369019,
"learning_rate": 5.995694586778692e-06,
"loss": 1.2811360359191895,
"step": 382
},
{
"epoch": 0.46886446886446886,
"grad_norm": 1.650488257408142,
"learning_rate": 5.995363613829776e-06,
"loss": 1.0443174839019775,
"step": 384
},
{
"epoch": 0.4713064713064713,
"grad_norm": 4.0077738761901855,
"learning_rate": 5.995020396918799e-06,
"loss": 1.2094539403915405,
"step": 386
},
{
"epoch": 0.47374847374847373,
"grad_norm": 3.2971303462982178,
"learning_rate": 5.994664937604858e-06,
"loss": 1.602729320526123,
"step": 388
},
{
"epoch": 0.47619047619047616,
"grad_norm": 1.8406124114990234,
"learning_rate": 5.994297237502663e-06,
"loss": 0.7992144227027893,
"step": 390
},
{
"epoch": 0.47863247863247865,
"grad_norm": 3.1878228187561035,
"learning_rate": 5.993917298282529e-06,
"loss": 0.85492342710495,
"step": 392
},
{
"epoch": 0.4810744810744811,
"grad_norm": 1.531112790107727,
"learning_rate": 5.993525121670369e-06,
"loss": 0.9317103624343872,
"step": 394
},
{
"epoch": 0.4835164835164835,
"grad_norm": 1.492789626121521,
"learning_rate": 5.993120709447684e-06,
"loss": 1.2386692762374878,
"step": 396
},
{
"epoch": 0.48595848595848595,
"grad_norm": 1.2817524671554565,
"learning_rate": 5.9927040634515595e-06,
"loss": 1.2882177829742432,
"step": 398
},
{
"epoch": 0.4884004884004884,
"grad_norm": 2.322162389755249,
"learning_rate": 5.9922751855746495e-06,
"loss": 0.5464396476745605,
"step": 400
},
{
"epoch": 0.4908424908424908,
"grad_norm": 2.907667636871338,
"learning_rate": 5.991834077765176e-06,
"loss": 1.2339115142822266,
"step": 402
},
{
"epoch": 0.4932844932844933,
"grad_norm": 1.4344274997711182,
"learning_rate": 5.991380742026916e-06,
"loss": 1.337497591972351,
"step": 404
},
{
"epoch": 0.49572649572649574,
"grad_norm": 3.8610916137695312,
"learning_rate": 5.990915180419194e-06,
"loss": 0.6470708250999451,
"step": 406
},
{
"epoch": 0.4981684981684982,
"grad_norm": 8.258731842041016,
"learning_rate": 5.9904373950568695e-06,
"loss": 0.8996188640594482,
"step": 408
},
{
"epoch": 0.5006105006105006,
"grad_norm": 1.479630470275879,
"learning_rate": 5.989947388110331e-06,
"loss": 1.2700155973434448,
"step": 410
},
{
"epoch": 0.503052503052503,
"grad_norm": 3.2398624420166016,
"learning_rate": 5.9894451618054856e-06,
"loss": 0.6065315008163452,
"step": 412
},
{
"epoch": 0.5054945054945055,
"grad_norm": 1.5686074495315552,
"learning_rate": 5.9889307184237445e-06,
"loss": 1.3284578323364258,
"step": 414
},
{
"epoch": 0.5079365079365079,
"grad_norm": 2.6033775806427,
"learning_rate": 5.988404060302022e-06,
"loss": 1.281160593032837,
"step": 416
},
{
"epoch": 0.5103785103785103,
"grad_norm": 1.3352185487747192,
"learning_rate": 5.987865189832714e-06,
"loss": 1.219364881515503,
"step": 418
},
{
"epoch": 0.5128205128205128,
"grad_norm": 1.3756248950958252,
"learning_rate": 5.987314109463697e-06,
"loss": 1.2490348815917969,
"step": 420
},
{
"epoch": 0.5152625152625152,
"grad_norm": 1.8197388648986816,
"learning_rate": 5.986750821698305e-06,
"loss": 1.2416220903396606,
"step": 422
},
{
"epoch": 0.5177045177045178,
"grad_norm": 2.0134706497192383,
"learning_rate": 5.986175329095333e-06,
"loss": 1.2700915336608887,
"step": 424
},
{
"epoch": 0.5201465201465202,
"grad_norm": 2.1429672241210938,
"learning_rate": 5.985587634269013e-06,
"loss": 1.0763859748840332,
"step": 426
},
{
"epoch": 0.5225885225885226,
"grad_norm": 2.1284050941467285,
"learning_rate": 5.9849877398890085e-06,
"loss": 1.2657442092895508,
"step": 428
},
{
"epoch": 0.525030525030525,
"grad_norm": 3.2052974700927734,
"learning_rate": 5.984375648680401e-06,
"loss": 0.9026008248329163,
"step": 430
},
{
"epoch": 0.5274725274725275,
"grad_norm": 8.207277297973633,
"learning_rate": 5.983751363423675e-06,
"loss": 0.5902690887451172,
"step": 432
},
{
"epoch": 0.5299145299145299,
"grad_norm": 6.7014594078063965,
"learning_rate": 5.983114886954711e-06,
"loss": 1.2768715620040894,
"step": 434
},
{
"epoch": 0.5323565323565324,
"grad_norm": 1.3282718658447266,
"learning_rate": 5.982466222164768e-06,
"loss": 1.230344295501709,
"step": 436
},
{
"epoch": 0.5347985347985348,
"grad_norm": 1.8573893308639526,
"learning_rate": 5.981805372000472e-06,
"loss": 1.258583664894104,
"step": 438
},
{
"epoch": 0.5372405372405372,
"grad_norm": 15.122943878173828,
"learning_rate": 5.981132339463799e-06,
"loss": 1.2159916162490845,
"step": 440
},
{
"epoch": 0.5396825396825397,
"grad_norm": 4.852167129516602,
"learning_rate": 5.98044712761207e-06,
"loss": 1.0581938028335571,
"step": 442
},
{
"epoch": 0.5421245421245421,
"grad_norm": 2.0339410305023193,
"learning_rate": 5.9797497395579285e-06,
"loss": 0.8306668996810913,
"step": 444
},
{
"epoch": 0.5445665445665445,
"grad_norm": 6.8808488845825195,
"learning_rate": 5.97904017846933e-06,
"loss": 0.9840553998947144,
"step": 446
},
{
"epoch": 0.5470085470085471,
"grad_norm": 1.4884564876556396,
"learning_rate": 5.978318447569527e-06,
"loss": 1.1001070737838745,
"step": 448
},
{
"epoch": 0.5494505494505495,
"grad_norm": 3.1171839237213135,
"learning_rate": 5.977584550137057e-06,
"loss": 0.9998340606689453,
"step": 450
},
{
"epoch": 0.5518925518925519,
"grad_norm": 4.206235408782959,
"learning_rate": 5.976838489505721e-06,
"loss": 0.7569481730461121,
"step": 452
},
{
"epoch": 0.5543345543345544,
"grad_norm": 1.3813527822494507,
"learning_rate": 5.9760802690645775e-06,
"loss": 1.2668659687042236,
"step": 454
},
{
"epoch": 0.5567765567765568,
"grad_norm": 1.965630054473877,
"learning_rate": 5.975309892257918e-06,
"loss": 1.2358734607696533,
"step": 456
},
{
"epoch": 0.5592185592185592,
"grad_norm": 0.8634360432624817,
"learning_rate": 5.974527362585258e-06,
"loss": 0.6871505379676819,
"step": 458
},
{
"epoch": 0.5616605616605617,
"grad_norm": 5.055013656616211,
"learning_rate": 5.973732683601318e-06,
"loss": 1.1548914909362793,
"step": 460
},
{
"epoch": 0.5641025641025641,
"grad_norm": 2.3719708919525146,
"learning_rate": 5.972925858916006e-06,
"loss": 1.1725798845291138,
"step": 462
},
{
"epoch": 0.5665445665445665,
"grad_norm": 14.694830894470215,
"learning_rate": 5.972106892194404e-06,
"loss": 0.5542956590652466,
"step": 464
},
{
"epoch": 0.568986568986569,
"grad_norm": 1.5816266536712646,
"learning_rate": 5.9712757871567556e-06,
"loss": 1.4963278770446777,
"step": 466
},
{
"epoch": 0.5714285714285714,
"grad_norm": 1.5916846990585327,
"learning_rate": 5.9704325475784355e-06,
"loss": 1.3050663471221924,
"step": 468
},
{
"epoch": 0.5738705738705738,
"grad_norm": 2.6258790493011475,
"learning_rate": 5.969577177289948e-06,
"loss": 1.0056266784667969,
"step": 470
},
{
"epoch": 0.5763125763125763,
"grad_norm": 7.421305179595947,
"learning_rate": 5.968709680176896e-06,
"loss": 1.2716175317764282,
"step": 472
},
{
"epoch": 0.5787545787545788,
"grad_norm": 4.030231952667236,
"learning_rate": 5.967830060179975e-06,
"loss": 0.8685106039047241,
"step": 474
},
{
"epoch": 0.5811965811965812,
"grad_norm": 3.304006576538086,
"learning_rate": 5.966938321294947e-06,
"loss": 1.0120376348495483,
"step": 476
},
{
"epoch": 0.5836385836385837,
"grad_norm": 1.996673583984375,
"learning_rate": 5.966034467572626e-06,
"loss": 1.166401982307434,
"step": 478
},
{
"epoch": 0.5860805860805861,
"grad_norm": 2.437133550643921,
"learning_rate": 5.965118503118861e-06,
"loss": 1.2958388328552246,
"step": 480
},
{
"epoch": 0.5885225885225885,
"grad_norm": 1.4910950660705566,
"learning_rate": 5.964190432094512e-06,
"loss": 0.523586094379425,
"step": 482
},
{
"epoch": 0.590964590964591,
"grad_norm": 0.8698585629463196,
"learning_rate": 5.963250258715435e-06,
"loss": 0.954540491104126,
"step": 484
},
{
"epoch": 0.5934065934065934,
"grad_norm": 0.9387339949607849,
"learning_rate": 5.962297987252463e-06,
"loss": 1.014317274093628,
"step": 486
},
{
"epoch": 0.5958485958485958,
"grad_norm": 5.339751720428467,
"learning_rate": 5.961333622031385e-06,
"loss": 0.9420929551124573,
"step": 488
},
{
"epoch": 0.5982905982905983,
"grad_norm": 2.236091136932373,
"learning_rate": 5.9603571674329294e-06,
"loss": 0.9940929412841797,
"step": 490
},
{
"epoch": 0.6007326007326007,
"grad_norm": 2.2979278564453125,
"learning_rate": 5.959368627892738e-06,
"loss": 0.7589210271835327,
"step": 492
},
{
"epoch": 0.6031746031746031,
"grad_norm": 1.245233178138733,
"learning_rate": 5.958368007901353e-06,
"loss": 0.9640493988990784,
"step": 494
},
{
"epoch": 0.6056166056166056,
"grad_norm": 3.310840368270874,
"learning_rate": 5.9573553120041916e-06,
"loss": 1.2264776229858398,
"step": 496
},
{
"epoch": 0.608058608058608,
"grad_norm": 5.373754024505615,
"learning_rate": 5.956330544801528e-06,
"loss": 0.5740343332290649,
"step": 498
},
{
"epoch": 0.6105006105006106,
"grad_norm": 1.4315255880355835,
"learning_rate": 5.95529371094847e-06,
"loss": 0.8721321821212769,
"step": 500
},
{
"epoch": 0.612942612942613,
"grad_norm": 1.5925936698913574,
"learning_rate": 5.954244815154941e-06,
"loss": 1.222565770149231,
"step": 502
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.4787653684616089,
"learning_rate": 5.953183862185656e-06,
"loss": 1.238010048866272,
"step": 504
},
{
"epoch": 0.6178266178266179,
"grad_norm": 2.44746994972229,
"learning_rate": 5.9521108568601015e-06,
"loss": 1.2495218515396118,
"step": 506
},
{
"epoch": 0.6202686202686203,
"grad_norm": 2.0077357292175293,
"learning_rate": 5.951025804052512e-06,
"loss": 1.2221843004226685,
"step": 508
},
{
"epoch": 0.6227106227106227,
"grad_norm": 2.2311477661132812,
"learning_rate": 5.949928708691852e-06,
"loss": 1.2433747053146362,
"step": 510
},
{
"epoch": 0.6251526251526252,
"grad_norm": 1.3252453804016113,
"learning_rate": 5.948819575761785e-06,
"loss": 1.070593237876892,
"step": 512
},
{
"epoch": 0.6275946275946276,
"grad_norm": 1.3435025215148926,
"learning_rate": 5.947698410300662e-06,
"loss": 1.0740947723388672,
"step": 514
},
{
"epoch": 0.63003663003663,
"grad_norm": 1.2309428453445435,
"learning_rate": 5.94656521740149e-06,
"loss": 0.9620496034622192,
"step": 516
},
{
"epoch": 0.6324786324786325,
"grad_norm": 1.51472008228302,
"learning_rate": 5.945420002211912e-06,
"loss": 1.2541457414627075,
"step": 518
},
{
"epoch": 0.6349206349206349,
"grad_norm": 2.0422983169555664,
"learning_rate": 5.944262769934184e-06,
"loss": 0.7579051852226257,
"step": 520
},
{
"epoch": 0.6373626373626373,
"grad_norm": 10.48106575012207,
"learning_rate": 5.9430935258251516e-06,
"loss": 0.9636937975883484,
"step": 522
},
{
"epoch": 0.6398046398046398,
"grad_norm": 12.823934555053711,
"learning_rate": 5.941912275196223e-06,
"loss": 1.010987639427185,
"step": 524
},
{
"epoch": 0.6422466422466423,
"grad_norm": 2.1455140113830566,
"learning_rate": 5.94071902341335e-06,
"loss": 1.2731401920318604,
"step": 526
},
{
"epoch": 0.6446886446886447,
"grad_norm": 1.989286184310913,
"learning_rate": 5.9395137758969996e-06,
"loss": 1.321770191192627,
"step": 528
},
{
"epoch": 0.6471306471306472,
"grad_norm": 1.7523471117019653,
"learning_rate": 5.938296538122131e-06,
"loss": 1.0642083883285522,
"step": 530
},
{
"epoch": 0.6495726495726496,
"grad_norm": 2.0933523178100586,
"learning_rate": 5.9370673156181685e-06,
"loss": 1.0688331127166748,
"step": 532
},
{
"epoch": 0.652014652014652,
"grad_norm": 4.951768398284912,
"learning_rate": 5.9358261139689804e-06,
"loss": 1.530338168144226,
"step": 534
},
{
"epoch": 0.6544566544566545,
"grad_norm": 1.5651706457138062,
"learning_rate": 5.9345729388128516e-06,
"loss": 0.8295682668685913,
"step": 536
},
{
"epoch": 0.6568986568986569,
"grad_norm": 1.4690457582473755,
"learning_rate": 5.9333077958424564e-06,
"loss": 1.2547093629837036,
"step": 538
},
{
"epoch": 0.6593406593406593,
"grad_norm": 1.7123324871063232,
"learning_rate": 5.932030690804835e-06,
"loss": 1.2759506702423096,
"step": 540
},
{
"epoch": 0.6617826617826618,
"grad_norm": 2.439643621444702,
"learning_rate": 5.9307416295013655e-06,
"loss": 0.9515459537506104,
"step": 542
},
{
"epoch": 0.6642246642246642,
"grad_norm": 1.4160524606704712,
"learning_rate": 5.929440617787741e-06,
"loss": 1.2433102130889893,
"step": 544
},
{
"epoch": 0.6666666666666666,
"grad_norm": 6.129720211029053,
"learning_rate": 5.928127661573937e-06,
"loss": 1.1501506567001343,
"step": 546
},
{
"epoch": 0.6691086691086691,
"grad_norm": 2.4493350982666016,
"learning_rate": 5.9268027668241925e-06,
"loss": 1.2322951555252075,
"step": 548
},
{
"epoch": 0.6715506715506715,
"grad_norm": 5.7734575271606445,
"learning_rate": 5.925465939556974e-06,
"loss": 1.1571426391601562,
"step": 550
},
{
"epoch": 0.673992673992674,
"grad_norm": 3.0620312690734863,
"learning_rate": 5.924117185844955e-06,
"loss": 1.010740876197815,
"step": 552
},
{
"epoch": 0.6764346764346765,
"grad_norm": 1.3649016618728638,
"learning_rate": 5.922756511814986e-06,
"loss": 1.0325241088867188,
"step": 554
},
{
"epoch": 0.6788766788766789,
"grad_norm": 3.0845694541931152,
"learning_rate": 5.921383923648067e-06,
"loss": 0.7497116923332214,
"step": 556
},
{
"epoch": 0.6813186813186813,
"grad_norm": 3.1487061977386475,
"learning_rate": 5.919999427579318e-06,
"loss": 1.1814972162246704,
"step": 558
},
{
"epoch": 0.6837606837606838,
"grad_norm": 0.6284694671630859,
"learning_rate": 5.918603029897952e-06,
"loss": 1.168214201927185,
"step": 560
},
{
"epoch": 0.6862026862026862,
"grad_norm": 5.0077080726623535,
"learning_rate": 5.917194736947248e-06,
"loss": 1.011269211769104,
"step": 562
},
{
"epoch": 0.6886446886446886,
"grad_norm": 1.6639766693115234,
"learning_rate": 5.915774555124519e-06,
"loss": 1.3079068660736084,
"step": 564
},
{
"epoch": 0.6910866910866911,
"grad_norm": 1.6810539960861206,
"learning_rate": 5.914342490881085e-06,
"loss": 1.096473217010498,
"step": 566
},
{
"epoch": 0.6935286935286935,
"grad_norm": 4.131659030914307,
"learning_rate": 5.912898550722242e-06,
"loss": 1.2962534427642822,
"step": 568
},
{
"epoch": 0.6959706959706959,
"grad_norm": 1.2126718759536743,
"learning_rate": 5.911442741207234e-06,
"loss": 0.9613256454467773,
"step": 570
},
{
"epoch": 0.6984126984126984,
"grad_norm": 2.5516252517700195,
"learning_rate": 5.909975068949225e-06,
"loss": 0.9773088693618774,
"step": 572
},
{
"epoch": 0.7008547008547008,
"grad_norm": 2.4411351680755615,
"learning_rate": 5.908495540615263e-06,
"loss": 1.427431583404541,
"step": 574
},
{
"epoch": 0.7032967032967034,
"grad_norm": 1.301823377609253,
"learning_rate": 5.907004162926255e-06,
"loss": 1.2127736806869507,
"step": 576
},
{
"epoch": 0.7057387057387058,
"grad_norm": 1.3317594528198242,
"learning_rate": 5.9055009426569374e-06,
"loss": 1.190283179283142,
"step": 578
},
{
"epoch": 0.7081807081807082,
"grad_norm": 2.30523681640625,
"learning_rate": 5.903985886635838e-06,
"loss": 1.1487045288085938,
"step": 580
},
{
"epoch": 0.7106227106227107,
"grad_norm": 1.789974331855774,
"learning_rate": 5.902459001745253e-06,
"loss": 1.20576810836792,
"step": 582
},
{
"epoch": 0.7130647130647131,
"grad_norm": 1.5357120037078857,
"learning_rate": 5.9009202949212125e-06,
"loss": 1.0540030002593994,
"step": 584
},
{
"epoch": 0.7155067155067155,
"grad_norm": 1.0633633136749268,
"learning_rate": 5.899369773153447e-06,
"loss": 1.2952874898910522,
"step": 586
},
{
"epoch": 0.717948717948718,
"grad_norm": 3.676633596420288,
"learning_rate": 5.897807443485358e-06,
"loss": 1.0641299486160278,
"step": 588
},
{
"epoch": 0.7203907203907204,
"grad_norm": 1.8231834173202515,
"learning_rate": 5.8962333130139875e-06,
"loss": 0.9424347281455994,
"step": 590
},
{
"epoch": 0.7228327228327228,
"grad_norm": 1.9420549869537354,
"learning_rate": 5.894647388889982e-06,
"loss": 0.8484236001968384,
"step": 592
},
{
"epoch": 0.7252747252747253,
"grad_norm": 1.0987299680709839,
"learning_rate": 5.893049678317563e-06,
"loss": 1.2070276737213135,
"step": 594
},
{
"epoch": 0.7277167277167277,
"grad_norm": 10.05299186706543,
"learning_rate": 5.891440188554491e-06,
"loss": 0.9539278745651245,
"step": 596
},
{
"epoch": 0.7301587301587301,
"grad_norm": 1.7478141784667969,
"learning_rate": 5.889818926912037e-06,
"loss": 0.9240214824676514,
"step": 598
},
{
"epoch": 0.7326007326007326,
"grad_norm": 1.9144340753555298,
"learning_rate": 5.888185900754946e-06,
"loss": 1.1822292804718018,
"step": 600
},
{
"epoch": 0.7350427350427351,
"grad_norm": 1.3329391479492188,
"learning_rate": 5.886541117501406e-06,
"loss": 1.3204286098480225,
"step": 602
},
{
"epoch": 0.7374847374847375,
"grad_norm": 2.1655688285827637,
"learning_rate": 5.884884584623009e-06,
"loss": 1.1428865194320679,
"step": 604
},
{
"epoch": 0.73992673992674,
"grad_norm": 1.6737140417099,
"learning_rate": 5.883216309644725e-06,
"loss": 1.4259554147720337,
"step": 606
},
{
"epoch": 0.7423687423687424,
"grad_norm": 1.2776116132736206,
"learning_rate": 5.881536300144861e-06,
"loss": 1.2691248655319214,
"step": 608
},
{
"epoch": 0.7448107448107448,
"grad_norm": 1.9943733215332031,
"learning_rate": 5.879844563755031e-06,
"loss": 0.8886740207672119,
"step": 610
},
{
"epoch": 0.7472527472527473,
"grad_norm": 1.6197903156280518,
"learning_rate": 5.878141108160117e-06,
"loss": 1.2487074136734009,
"step": 612
},
{
"epoch": 0.7496947496947497,
"grad_norm": 1.7818470001220703,
"learning_rate": 5.876425941098242e-06,
"loss": 1.2276005744934082,
"step": 614
},
{
"epoch": 0.7521367521367521,
"grad_norm": 2.3711366653442383,
"learning_rate": 5.8746990703607235e-06,
"loss": 0.8518514633178711,
"step": 616
},
{
"epoch": 0.7545787545787546,
"grad_norm": 1.3806644678115845,
"learning_rate": 5.872960503792047e-06,
"loss": 1.5368404388427734,
"step": 618
},
{
"epoch": 0.757020757020757,
"grad_norm": 2.896021604537964,
"learning_rate": 5.871210249289828e-06,
"loss": 1.0128453969955444,
"step": 620
},
{
"epoch": 0.7594627594627594,
"grad_norm": 4.764344215393066,
"learning_rate": 5.869448314804772e-06,
"loss": 0.9351692199707031,
"step": 622
},
{
"epoch": 0.7619047619047619,
"grad_norm": 1.9629191160202026,
"learning_rate": 5.867674708340647e-06,
"loss": 1.2186893224716187,
"step": 624
},
{
"epoch": 0.7643467643467643,
"grad_norm": 1.4997833967208862,
"learning_rate": 5.86588943795424e-06,
"loss": 1.2730145454406738,
"step": 626
},
{
"epoch": 0.7667887667887668,
"grad_norm": 5.824042797088623,
"learning_rate": 5.864092511755323e-06,
"loss": 1.0189692974090576,
"step": 628
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.6058506965637207,
"learning_rate": 5.8622839379066135e-06,
"loss": 1.2564375400543213,
"step": 630
},
{
"epoch": 0.7716727716727717,
"grad_norm": 2.6748597621917725,
"learning_rate": 5.860463724623742e-06,
"loss": 0.8755788207054138,
"step": 632
},
{
"epoch": 0.7741147741147741,
"grad_norm": 1.938493013381958,
"learning_rate": 5.858631880175212e-06,
"loss": 0.8998976945877075,
"step": 634
},
{
"epoch": 0.7765567765567766,
"grad_norm": 1.6986690759658813,
"learning_rate": 5.856788412882361e-06,
"loss": 1.2397321462631226,
"step": 636
},
{
"epoch": 0.778998778998779,
"grad_norm": 2.0110723972320557,
"learning_rate": 5.854933331119328e-06,
"loss": 1.24760103225708,
"step": 638
},
{
"epoch": 0.7814407814407814,
"grad_norm": 2.173943042755127,
"learning_rate": 5.853066643313007e-06,
"loss": 0.8890904188156128,
"step": 640
},
{
"epoch": 0.7838827838827839,
"grad_norm": 3.4975063800811768,
"learning_rate": 5.851188357943019e-06,
"loss": 1.00565767288208,
"step": 642
},
{
"epoch": 0.7863247863247863,
"grad_norm": 1.0548205375671387,
"learning_rate": 5.849298483541663e-06,
"loss": 1.1251481771469116,
"step": 644
},
{
"epoch": 0.7887667887667887,
"grad_norm": 1.5539727210998535,
"learning_rate": 5.847397028693887e-06,
"loss": 0.9152241945266724,
"step": 646
},
{
"epoch": 0.7912087912087912,
"grad_norm": 1.818673014640808,
"learning_rate": 5.845484002037241e-06,
"loss": 1.19895601272583,
"step": 648
},
{
"epoch": 0.7936507936507936,
"grad_norm": 1.390134572982788,
"learning_rate": 5.843559412261842e-06,
"loss": 0.5183348655700684,
"step": 650
},
{
"epoch": 0.796092796092796,
"grad_norm": 1.5669151544570923,
"learning_rate": 5.841623268110333e-06,
"loss": 0.7738546133041382,
"step": 652
},
{
"epoch": 0.7985347985347986,
"grad_norm": 3.074685573577881,
"learning_rate": 5.839675578377848e-06,
"loss": 1.1633089780807495,
"step": 654
},
{
"epoch": 0.800976800976801,
"grad_norm": 1.443999171257019,
"learning_rate": 5.837716351911962e-06,
"loss": 1.2039074897766113,
"step": 656
},
{
"epoch": 0.8034188034188035,
"grad_norm": 3.9893510341644287,
"learning_rate": 5.83574559761266e-06,
"loss": 0.4816018044948578,
"step": 658
},
{
"epoch": 0.8058608058608059,
"grad_norm": 1.8784825801849365,
"learning_rate": 5.833763324432294e-06,
"loss": 1.053369402885437,
"step": 660
},
{
"epoch": 0.8083028083028083,
"grad_norm": 2.2638516426086426,
"learning_rate": 5.831769541375539e-06,
"loss": 1.1880583763122559,
"step": 662
},
{
"epoch": 0.8107448107448108,
"grad_norm": 1.5638916492462158,
"learning_rate": 5.829764257499357e-06,
"loss": 1.2887765169143677,
"step": 664
},
{
"epoch": 0.8131868131868132,
"grad_norm": 1.4829241037368774,
"learning_rate": 5.827747481912953e-06,
"loss": 0.8841955065727234,
"step": 666
},
{
"epoch": 0.8156288156288156,
"grad_norm": 2.4060006141662598,
"learning_rate": 5.825719223777735e-06,
"loss": 1.04268217086792,
"step": 668
},
{
"epoch": 0.818070818070818,
"grad_norm": 1.3124898672103882,
"learning_rate": 5.823679492307271e-06,
"loss": 1.3766071796417236,
"step": 670
},
{
"epoch": 0.8205128205128205,
"grad_norm": 1.4175595045089722,
"learning_rate": 5.8216282967672454e-06,
"loss": 1.2438055276870728,
"step": 672
},
{
"epoch": 0.8229548229548229,
"grad_norm": 1.81400465965271,
"learning_rate": 5.819565646475425e-06,
"loss": 1.001070261001587,
"step": 674
},
{
"epoch": 0.8253968253968254,
"grad_norm": 2.5499863624572754,
"learning_rate": 5.817491550801603e-06,
"loss": 0.8810223937034607,
"step": 676
},
{
"epoch": 0.8278388278388278,
"grad_norm": 3.1251838207244873,
"learning_rate": 5.815406019167574e-06,
"loss": 1.2103674411773682,
"step": 678
},
{
"epoch": 0.8302808302808303,
"grad_norm": 1.698996663093567,
"learning_rate": 5.8133090610470735e-06,
"loss": 1.2406985759735107,
"step": 680
},
{
"epoch": 0.8327228327228328,
"grad_norm": 2.1086080074310303,
"learning_rate": 5.811200685965747e-06,
"loss": 1.1365587711334229,
"step": 682
},
{
"epoch": 0.8351648351648352,
"grad_norm": 8.87724494934082,
"learning_rate": 5.809080903501101e-06,
"loss": 0.6978262662887573,
"step": 684
},
{
"epoch": 0.8376068376068376,
"grad_norm": 1.3921948671340942,
"learning_rate": 5.806949723282462e-06,
"loss": 1.3278273344039917,
"step": 686
},
{
"epoch": 0.8400488400488401,
"grad_norm": 2.7483460903167725,
"learning_rate": 5.804807154990933e-06,
"loss": 1.2896418571472168,
"step": 688
},
{
"epoch": 0.8424908424908425,
"grad_norm": 5.320318698883057,
"learning_rate": 5.802653208359347e-06,
"loss": 0.9632086753845215,
"step": 690
},
{
"epoch": 0.8449328449328449,
"grad_norm": 2.63219952583313,
"learning_rate": 5.800487893172224e-06,
"loss": 1.249232292175293,
"step": 692
},
{
"epoch": 0.8473748473748474,
"grad_norm": 1.0053170919418335,
"learning_rate": 5.798311219265727e-06,
"loss": 0.9811009168624878,
"step": 694
},
{
"epoch": 0.8498168498168498,
"grad_norm": 2.8511881828308105,
"learning_rate": 5.796123196527619e-06,
"loss": 1.1377757787704468,
"step": 696
},
{
"epoch": 0.8522588522588522,
"grad_norm": 2.364295482635498,
"learning_rate": 5.793923834897213e-06,
"loss": 1.006821632385254,
"step": 698
},
{
"epoch": 0.8547008547008547,
"grad_norm": 2.2293732166290283,
"learning_rate": 5.791713144365334e-06,
"loss": 1.0615328550338745,
"step": 700
},
{
"epoch": 0.8571428571428571,
"grad_norm": 3.746429204940796,
"learning_rate": 5.789491134974266e-06,
"loss": 1.0453742742538452,
"step": 702
},
{
"epoch": 0.8595848595848596,
"grad_norm": 2.0390639305114746,
"learning_rate": 5.787257816817712e-06,
"loss": 0.9529613852500916,
"step": 704
},
{
"epoch": 0.8620268620268621,
"grad_norm": 2.2642388343811035,
"learning_rate": 5.785013200040747e-06,
"loss": 0.39921677112579346,
"step": 706
},
{
"epoch": 0.8644688644688645,
"grad_norm": 1.2055710554122925,
"learning_rate": 5.782757294839766e-06,
"loss": 1.1861385107040405,
"step": 708
},
{
"epoch": 0.8669108669108669,
"grad_norm": 1.462980031967163,
"learning_rate": 5.780490111462451e-06,
"loss": 0.872178852558136,
"step": 710
},
{
"epoch": 0.8693528693528694,
"grad_norm": 128.80056762695312,
"learning_rate": 5.77821166020771e-06,
"loss": 1.3495750427246094,
"step": 712
},
{
"epoch": 0.8717948717948718,
"grad_norm": 2.8078572750091553,
"learning_rate": 5.77592195142564e-06,
"loss": 1.5414416790008545,
"step": 714
},
{
"epoch": 0.8742368742368742,
"grad_norm": 2.91050124168396,
"learning_rate": 5.773620995517472e-06,
"loss": 1.402683138847351,
"step": 716
},
{
"epoch": 0.8766788766788767,
"grad_norm": 1.1673686504364014,
"learning_rate": 5.771308802935534e-06,
"loss": 0.9155862927436829,
"step": 718
},
{
"epoch": 0.8791208791208791,
"grad_norm": 3.338332414627075,
"learning_rate": 5.768985384183194e-06,
"loss": 1.1884852647781372,
"step": 720
},
{
"epoch": 0.8815628815628815,
"grad_norm": 10.983979225158691,
"learning_rate": 5.766650749814816e-06,
"loss": 0.8515965938568115,
"step": 722
},
{
"epoch": 0.884004884004884,
"grad_norm": 2.596907377243042,
"learning_rate": 5.764304910435715e-06,
"loss": 1.2161564826965332,
"step": 724
},
{
"epoch": 0.8864468864468864,
"grad_norm": 1.6692111492156982,
"learning_rate": 5.761947876702101e-06,
"loss": 1.1635143756866455,
"step": 726
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.1050961017608643,
"learning_rate": 5.759579659321041e-06,
"loss": 1.1528334617614746,
"step": 728
},
{
"epoch": 0.8913308913308914,
"grad_norm": 2.1501071453094482,
"learning_rate": 5.7572002690504e-06,
"loss": 1.3882144689559937,
"step": 730
},
{
"epoch": 0.8937728937728938,
"grad_norm": 4.581213474273682,
"learning_rate": 5.7548097166988e-06,
"loss": 1.1679284572601318,
"step": 732
},
{
"epoch": 0.8962148962148963,
"grad_norm": 1.7328788042068481,
"learning_rate": 5.752408013125568e-06,
"loss": 0.9865559339523315,
"step": 734
},
{
"epoch": 0.8986568986568987,
"grad_norm": 1.568103551864624,
"learning_rate": 5.749995169240684e-06,
"loss": 0.772152841091156,
"step": 736
},
{
"epoch": 0.9010989010989011,
"grad_norm": 7.975260257720947,
"learning_rate": 5.747571196004737e-06,
"loss": 0.8893488645553589,
"step": 738
},
{
"epoch": 0.9035409035409036,
"grad_norm": 6.543982982635498,
"learning_rate": 5.7451361044288705e-06,
"loss": 1.0185538530349731,
"step": 740
},
{
"epoch": 0.905982905982906,
"grad_norm": 1.160656452178955,
"learning_rate": 5.742689905574733e-06,
"loss": 0.8777042031288147,
"step": 742
},
{
"epoch": 0.9084249084249084,
"grad_norm": 2.757848024368286,
"learning_rate": 5.740232610554433e-06,
"loss": 0.8858698010444641,
"step": 744
},
{
"epoch": 0.9108669108669109,
"grad_norm": 9.06092643737793,
"learning_rate": 5.73776423053048e-06,
"loss": 1.1307774782180786,
"step": 746
},
{
"epoch": 0.9133089133089133,
"grad_norm": 2.1863319873809814,
"learning_rate": 5.73528477671574e-06,
"loss": 0.8516602516174316,
"step": 748
},
{
"epoch": 0.9157509157509157,
"grad_norm": 1.554819941520691,
"learning_rate": 5.732794260373384e-06,
"loss": 1.333770751953125,
"step": 750
},
{
"epoch": 0.9181929181929182,
"grad_norm": 3.050586223602295,
"learning_rate": 5.730292692816835e-06,
"loss": 1.222764492034912,
"step": 752
},
{
"epoch": 0.9206349206349206,
"grad_norm": 1.5035123825073242,
"learning_rate": 5.727780085409714e-06,
"loss": 1.230542540550232,
"step": 754
},
{
"epoch": 0.9230769230769231,
"grad_norm": 2.754014730453491,
"learning_rate": 5.7252564495657985e-06,
"loss": 1.2115579843521118,
"step": 756
},
{
"epoch": 0.9255189255189256,
"grad_norm": 1.9241371154785156,
"learning_rate": 5.722721796748957e-06,
"loss": 1.2471449375152588,
"step": 758
},
{
"epoch": 0.927960927960928,
"grad_norm": 2.088291645050049,
"learning_rate": 5.720176138473106e-06,
"loss": 0.7556540369987488,
"step": 760
},
{
"epoch": 0.9304029304029304,
"grad_norm": 2.329336166381836,
"learning_rate": 5.717619486302159e-06,
"loss": 1.0959203243255615,
"step": 762
},
{
"epoch": 0.9328449328449329,
"grad_norm": 1.3153835535049438,
"learning_rate": 5.715051851849965e-06,
"loss": 0.8686593770980835,
"step": 764
},
{
"epoch": 0.9352869352869353,
"grad_norm": 6.791528224945068,
"learning_rate": 5.712473246780264e-06,
"loss": 0.5506616830825806,
"step": 766
},
{
"epoch": 0.9377289377289377,
"grad_norm": 1.5540168285369873,
"learning_rate": 5.7098836828066295e-06,
"loss": 0.8565468788146973,
"step": 768
},
{
"epoch": 0.9401709401709402,
"grad_norm": 5.5268754959106445,
"learning_rate": 5.707283171692419e-06,
"loss": 0.6587256789207458,
"step": 770
},
{
"epoch": 0.9426129426129426,
"grad_norm": 1.8966621160507202,
"learning_rate": 5.704671725250719e-06,
"loss": 1.0538215637207031,
"step": 772
},
{
"epoch": 0.945054945054945,
"grad_norm": 1.3468095064163208,
"learning_rate": 5.702049355344287e-06,
"loss": 1.2241626977920532,
"step": 774
},
{
"epoch": 0.9474969474969475,
"grad_norm": 1.4727566242218018,
"learning_rate": 5.699416073885503e-06,
"loss": 1.1616984605789185,
"step": 776
},
{
"epoch": 0.9499389499389499,
"grad_norm": 3.3165957927703857,
"learning_rate": 5.6967718928363175e-06,
"loss": 0.8437950611114502,
"step": 778
},
{
"epoch": 0.9523809523809523,
"grad_norm": 1.1607977151870728,
"learning_rate": 5.694116824208188e-06,
"loss": 0.9664669632911682,
"step": 780
},
{
"epoch": 0.9548229548229549,
"grad_norm": 2.6319656372070312,
"learning_rate": 5.691450880062035e-06,
"loss": 1.088547706604004,
"step": 782
},
{
"epoch": 0.9572649572649573,
"grad_norm": 1.914591670036316,
"learning_rate": 5.688774072508178e-06,
"loss": 1.203494906425476,
"step": 784
},
{
"epoch": 0.9597069597069597,
"grad_norm": 3.445983648300171,
"learning_rate": 5.686086413706286e-06,
"loss": 1.2426470518112183,
"step": 786
},
{
"epoch": 0.9621489621489622,
"grad_norm": 4.115734100341797,
"learning_rate": 5.683387915865322e-06,
"loss": 1.3359789848327637,
"step": 788
},
{
"epoch": 0.9645909645909646,
"grad_norm": 1.785029649734497,
"learning_rate": 5.680678591243486e-06,
"loss": 1.2645583152770996,
"step": 790
},
{
"epoch": 0.967032967032967,
"grad_norm": 3.400036573410034,
"learning_rate": 5.677958452148158e-06,
"loss": 1.0313962697982788,
"step": 792
},
{
"epoch": 0.9694749694749695,
"grad_norm": 3.4706273078918457,
"learning_rate": 5.6752275109358456e-06,
"loss": 1.2580047845840454,
"step": 794
},
{
"epoch": 0.9719169719169719,
"grad_norm": 3.956390857696533,
"learning_rate": 5.672485780012126e-06,
"loss": 1.1567012071609497,
"step": 796
},
{
"epoch": 0.9743589743589743,
"grad_norm": 1.5805697441101074,
"learning_rate": 5.66973327183159e-06,
"loss": 1.1679713726043701,
"step": 798
},
{
"epoch": 0.9768009768009768,
"grad_norm": 1.9347152709960938,
"learning_rate": 5.6669699988977845e-06,
"loss": 1.2190362215042114,
"step": 800
},
{
"epoch": 0.9792429792429792,
"grad_norm": 3.247502565383911,
"learning_rate": 5.664195973763155e-06,
"loss": 0.8972685933113098,
"step": 802
},
{
"epoch": 0.9816849816849816,
"grad_norm": 1.9501919746398926,
"learning_rate": 5.661411209028994e-06,
"loss": 1.1029160022735596,
"step": 804
},
{
"epoch": 0.9841269841269841,
"grad_norm": 3.1804869174957275,
"learning_rate": 5.658615717345374e-06,
"loss": 1.500832438468933,
"step": 806
},
{
"epoch": 0.9865689865689866,
"grad_norm": 2.697408676147461,
"learning_rate": 5.655809511411103e-06,
"loss": 0.88798987865448,
"step": 808
},
{
"epoch": 0.989010989010989,
"grad_norm": 1.2146772146224976,
"learning_rate": 5.652992603973652e-06,
"loss": 1.2545315027236938,
"step": 810
},
{
"epoch": 0.9914529914529915,
"grad_norm": 1.7389800548553467,
"learning_rate": 5.650165007829109e-06,
"loss": 1.2045774459838867,
"step": 812
},
{
"epoch": 0.9938949938949939,
"grad_norm": 1.8084465265274048,
"learning_rate": 5.647326735822117e-06,
"loss": 1.2816147804260254,
"step": 814
},
{
"epoch": 0.9963369963369964,
"grad_norm": 1.249566912651062,
"learning_rate": 5.644477800845813e-06,
"loss": 1.1596295833587646,
"step": 816
},
{
"epoch": 0.9987789987789988,
"grad_norm": 2.499196767807007,
"learning_rate": 5.641618215841772e-06,
"loss": 1.0810341835021973,
"step": 818
},
{
"epoch": 1.0012210012210012,
"grad_norm": 1.9975930452346802,
"learning_rate": 5.63874799379995e-06,
"loss": 0.9549547433853149,
"step": 820
},
{
"epoch": 1.0036630036630036,
"grad_norm": 1.4195305109024048,
"learning_rate": 5.635867147758619e-06,
"loss": 1.1853629350662231,
"step": 822
},
{
"epoch": 1.006105006105006,
"grad_norm": 4.497268199920654,
"learning_rate": 5.632975690804315e-06,
"loss": 0.4928017556667328,
"step": 824
},
{
"epoch": 1.0085470085470085,
"grad_norm": 2.1255745887756348,
"learning_rate": 5.630073636071774e-06,
"loss": 1.0843961238861084,
"step": 826
},
{
"epoch": 1.010989010989011,
"grad_norm": 34.958438873291016,
"learning_rate": 5.627160996743874e-06,
"loss": 0.8150432705879211,
"step": 828
},
{
"epoch": 1.0134310134310134,
"grad_norm": 1.4282941818237305,
"learning_rate": 5.624237786051572e-06,
"loss": 0.857114851474762,
"step": 830
},
{
"epoch": 1.0158730158730158,
"grad_norm": 33.52021789550781,
"learning_rate": 5.621304017273851e-06,
"loss": 0.8700833320617676,
"step": 832
},
{
"epoch": 1.0183150183150182,
"grad_norm": 1.4326083660125732,
"learning_rate": 5.618359703737651e-06,
"loss": 0.8282259702682495,
"step": 834
},
{
"epoch": 1.0207570207570207,
"grad_norm": 3.305655002593994,
"learning_rate": 5.615404858817814e-06,
"loss": 1.1669970750808716,
"step": 836
},
{
"epoch": 1.0231990231990231,
"grad_norm": 2.0486841201782227,
"learning_rate": 5.612439495937022e-06,
"loss": 1.2340257167816162,
"step": 838
},
{
"epoch": 1.0256410256410255,
"grad_norm": 2.4268410205841064,
"learning_rate": 5.609463628565738e-06,
"loss": 0.8751161694526672,
"step": 840
},
{
"epoch": 1.028083028083028,
"grad_norm": 1.9834953546524048,
"learning_rate": 5.60647727022214e-06,
"loss": 0.8645414710044861,
"step": 842
},
{
"epoch": 1.0305250305250304,
"grad_norm": 3.6179487705230713,
"learning_rate": 5.603480434472062e-06,
"loss": 1.101366400718689,
"step": 844
},
{
"epoch": 1.032967032967033,
"grad_norm": 3.077362537384033,
"learning_rate": 5.600473134928934e-06,
"loss": 0.8427292704582214,
"step": 846
},
{
"epoch": 1.0354090354090355,
"grad_norm": 1.3050235509872437,
"learning_rate": 5.59745538525372e-06,
"loss": 1.1196715831756592,
"step": 848
},
{
"epoch": 1.037851037851038,
"grad_norm": 3.206444263458252,
"learning_rate": 5.594427199154852e-06,
"loss": 0.9106602072715759,
"step": 850
},
{
"epoch": 1.0402930402930404,
"grad_norm": 4.214219570159912,
"learning_rate": 5.591388590388172e-06,
"loss": 0.6607577204704285,
"step": 852
},
{
"epoch": 1.0427350427350428,
"grad_norm": 2.264488935470581,
"learning_rate": 5.588339572756869e-06,
"loss": 1.0989412069320679,
"step": 854
},
{
"epoch": 1.0451770451770452,
"grad_norm": 5.119571208953857,
"learning_rate": 5.5852801601114125e-06,
"loss": 0.8569447994232178,
"step": 856
},
{
"epoch": 1.0476190476190477,
"grad_norm": 1.9778999090194702,
"learning_rate": 5.582210366349495e-06,
"loss": 1.1780451536178589,
"step": 858
},
{
"epoch": 1.05006105006105,
"grad_norm": 1.977215051651001,
"learning_rate": 5.579130205415967e-06,
"loss": 0.7887912392616272,
"step": 860
},
{
"epoch": 1.0525030525030525,
"grad_norm": 1.1233766078948975,
"learning_rate": 5.576039691302768e-06,
"loss": 1.0990511178970337,
"step": 862
},
{
"epoch": 1.054945054945055,
"grad_norm": 1.574094533920288,
"learning_rate": 5.572938838048874e-06,
"loss": 1.0752408504486084,
"step": 864
},
{
"epoch": 1.0573870573870574,
"grad_norm": 4.407149791717529,
"learning_rate": 5.569827659740223e-06,
"loss": 1.4731539487838745,
"step": 866
},
{
"epoch": 1.0598290598290598,
"grad_norm": 3.7272164821624756,
"learning_rate": 5.5667061705096594e-06,
"loss": 0.41928231716156006,
"step": 868
},
{
"epoch": 1.0622710622710623,
"grad_norm": 1.6230946779251099,
"learning_rate": 5.563574384536862e-06,
"loss": 0.9580034017562866,
"step": 870
},
{
"epoch": 1.0647130647130647,
"grad_norm": 3.506577253341675,
"learning_rate": 5.560432316048287e-06,
"loss": 0.9923728704452515,
"step": 872
},
{
"epoch": 1.0671550671550671,
"grad_norm": 1.629608392715454,
"learning_rate": 5.557279979317098e-06,
"loss": 0.747920036315918,
"step": 874
},
{
"epoch": 1.0695970695970696,
"grad_norm": 1.8776957988739014,
"learning_rate": 5.554117388663104e-06,
"loss": 1.0746995210647583,
"step": 876
},
{
"epoch": 1.072039072039072,
"grad_norm": 10.093416213989258,
"learning_rate": 5.550944558452692e-06,
"loss": 1.016481876373291,
"step": 878
},
{
"epoch": 1.0744810744810744,
"grad_norm": 2.4766669273376465,
"learning_rate": 5.547761503098766e-06,
"loss": 0.7514116764068604,
"step": 880
},
{
"epoch": 1.0769230769230769,
"grad_norm": 2.3521194458007812,
"learning_rate": 5.544568237060677e-06,
"loss": 1.1210784912109375,
"step": 882
},
{
"epoch": 1.0793650793650793,
"grad_norm": 9.225933074951172,
"learning_rate": 5.541364774844158e-06,
"loss": 0.7435236573219299,
"step": 884
},
{
"epoch": 1.0818070818070817,
"grad_norm": 2.2107040882110596,
"learning_rate": 5.538151131001262e-06,
"loss": 0.681244432926178,
"step": 886
},
{
"epoch": 1.0842490842490842,
"grad_norm": 1.9917274713516235,
"learning_rate": 5.534927320130289e-06,
"loss": 1.1450310945510864,
"step": 888
},
{
"epoch": 1.0866910866910866,
"grad_norm": 1.4945895671844482,
"learning_rate": 5.531693356875729e-06,
"loss": 0.8382049202919006,
"step": 890
},
{
"epoch": 1.089133089133089,
"grad_norm": 1.4547278881072998,
"learning_rate": 5.5284492559281846e-06,
"loss": 1.0185482501983643,
"step": 892
},
{
"epoch": 1.0915750915750915,
"grad_norm": 6.223243713378906,
"learning_rate": 5.525195032024317e-06,
"loss": 1.1147940158843994,
"step": 894
},
{
"epoch": 1.0940170940170941,
"grad_norm": 1.6152573823928833,
"learning_rate": 5.521930699946763e-06,
"loss": 1.0109710693359375,
"step": 896
},
{
"epoch": 1.0964590964590966,
"grad_norm": 1.5743906497955322,
"learning_rate": 5.5186562745240845e-06,
"loss": 0.8494668006896973,
"step": 898
},
{
"epoch": 1.098901098901099,
"grad_norm": 1.0578030347824097,
"learning_rate": 5.51537177063069e-06,
"loss": 0.7963146567344666,
"step": 900
},
{
"epoch": 1.1013431013431014,
"grad_norm": 2.8918213844299316,
"learning_rate": 5.51207720318677e-06,
"loss": 0.5992125272750854,
"step": 902
},
{
"epoch": 1.1037851037851039,
"grad_norm": 0.9322580099105835,
"learning_rate": 5.50877258715823e-06,
"loss": 0.6189204454421997,
"step": 904
},
{
"epoch": 1.1062271062271063,
"grad_norm": 7.108280181884766,
"learning_rate": 5.5054579375566246e-06,
"loss": 0.9952929019927979,
"step": 906
},
{
"epoch": 1.1086691086691087,
"grad_norm": 2.261410713195801,
"learning_rate": 5.5021332694390855e-06,
"loss": 1.095886468887329,
"step": 908
},
{
"epoch": 1.1111111111111112,
"grad_norm": 1.4210469722747803,
"learning_rate": 5.498798597908252e-06,
"loss": 0.8107478022575378,
"step": 910
},
{
"epoch": 1.1135531135531136,
"grad_norm": 1.540653944015503,
"learning_rate": 5.495453938112209e-06,
"loss": 1.1111472845077515,
"step": 912
},
{
"epoch": 1.115995115995116,
"grad_norm": 4.085680961608887,
"learning_rate": 5.492099305244411e-06,
"loss": 0.7442325353622437,
"step": 914
},
{
"epoch": 1.1184371184371185,
"grad_norm": 23.62496566772461,
"learning_rate": 5.488734714543617e-06,
"loss": 0.7848599553108215,
"step": 916
},
{
"epoch": 1.120879120879121,
"grad_norm": 3.1475861072540283,
"learning_rate": 5.4853601812938225e-06,
"loss": 1.0528494119644165,
"step": 918
},
{
"epoch": 1.1233211233211233,
"grad_norm": 3.1182055473327637,
"learning_rate": 5.481975720824186e-06,
"loss": 1.0924468040466309,
"step": 920
},
{
"epoch": 1.1257631257631258,
"grad_norm": 5.0743727684021,
"learning_rate": 5.478581348508961e-06,
"loss": 0.9042542576789856,
"step": 922
},
{
"epoch": 1.1282051282051282,
"grad_norm": 2.026230812072754,
"learning_rate": 5.475177079767427e-06,
"loss": 0.7077540755271912,
"step": 924
},
{
"epoch": 1.1306471306471306,
"grad_norm": 2.2267894744873047,
"learning_rate": 5.47176293006382e-06,
"loss": 0.6727372407913208,
"step": 926
},
{
"epoch": 1.133089133089133,
"grad_norm": 6.727903366088867,
"learning_rate": 5.468338914907262e-06,
"loss": 0.556076169013977,
"step": 928
},
{
"epoch": 1.1355311355311355,
"grad_norm": 3.433774471282959,
"learning_rate": 5.4649050498516865e-06,
"loss": 1.0447014570236206,
"step": 930
},
{
"epoch": 1.137973137973138,
"grad_norm": 5.706188201904297,
"learning_rate": 5.461461350495773e-06,
"loss": 1.1313908100128174,
"step": 932
},
{
"epoch": 1.1404151404151404,
"grad_norm": 4.141596794128418,
"learning_rate": 5.458007832482875e-06,
"loss": 0.9544519782066345,
"step": 934
},
{
"epoch": 1.1428571428571428,
"grad_norm": 13.47986125946045,
"learning_rate": 5.454544511500949e-06,
"loss": 0.944713830947876,
"step": 936
},
{
"epoch": 1.1452991452991452,
"grad_norm": 1.946338415145874,
"learning_rate": 5.45107140328248e-06,
"loss": 1.0003188848495483,
"step": 938
},
{
"epoch": 1.1477411477411477,
"grad_norm": 1.6882654428482056,
"learning_rate": 5.447588523604413e-06,
"loss": 1.104058861732483,
"step": 940
},
{
"epoch": 1.15018315018315,
"grad_norm": 1.3386666774749756,
"learning_rate": 5.444095888288082e-06,
"loss": 1.0284168720245361,
"step": 942
},
{
"epoch": 1.1526251526251525,
"grad_norm": 2.4667913913726807,
"learning_rate": 5.440593513199136e-06,
"loss": 0.8081319332122803,
"step": 944
},
{
"epoch": 1.155067155067155,
"grad_norm": 1.5070215463638306,
"learning_rate": 5.4370814142474705e-06,
"loss": 0.7546210289001465,
"step": 946
},
{
"epoch": 1.1575091575091574,
"grad_norm": 4.182633876800537,
"learning_rate": 5.433559607387149e-06,
"loss": 0.559450089931488,
"step": 948
},
{
"epoch": 1.1599511599511598,
"grad_norm": 1.8414170742034912,
"learning_rate": 5.430028108616334e-06,
"loss": 1.0437990427017212,
"step": 950
},
{
"epoch": 1.1623931623931625,
"grad_norm": 2.0820398330688477,
"learning_rate": 5.42648693397722e-06,
"loss": 1.0472098588943481,
"step": 952
},
{
"epoch": 1.164835164835165,
"grad_norm": 2.6912789344787598,
"learning_rate": 5.4229360995559475e-06,
"loss": 0.8688966631889343,
"step": 954
},
{
"epoch": 1.1672771672771673,
"grad_norm": 3.620826482772827,
"learning_rate": 5.4193756214825425e-06,
"loss": 0.9058981537818909,
"step": 956
},
{
"epoch": 1.1697191697191698,
"grad_norm": 1.5990374088287354,
"learning_rate": 5.415805515930839e-06,
"loss": 0.7346622347831726,
"step": 958
},
{
"epoch": 1.1721611721611722,
"grad_norm": 3.7363603115081787,
"learning_rate": 5.4122257991184005e-06,
"loss": 1.1916764974594116,
"step": 960
},
{
"epoch": 1.1746031746031746,
"grad_norm": 2.6380817890167236,
"learning_rate": 5.408636487306453e-06,
"loss": 0.9414402842521667,
"step": 962
},
{
"epoch": 1.177045177045177,
"grad_norm": 0.9934802055358887,
"learning_rate": 5.4050375967998095e-06,
"loss": 0.748233437538147,
"step": 964
},
{
"epoch": 1.1794871794871795,
"grad_norm": 1.6639022827148438,
"learning_rate": 5.401429143946796e-06,
"loss": 1.1463696956634521,
"step": 966
},
{
"epoch": 1.181929181929182,
"grad_norm": 2.9646964073181152,
"learning_rate": 5.397811145139173e-06,
"loss": 1.188360333442688,
"step": 968
},
{
"epoch": 1.1843711843711844,
"grad_norm": 4.223168849945068,
"learning_rate": 5.3941836168120655e-06,
"loss": 0.7662742137908936,
"step": 970
},
{
"epoch": 1.1868131868131868,
"grad_norm": 1.6976840496063232,
"learning_rate": 5.3905465754438905e-06,
"loss": 1.205042839050293,
"step": 972
},
{
"epoch": 1.1892551892551892,
"grad_norm": 1.893657922744751,
"learning_rate": 5.3869000375562746e-06,
"loss": 0.8556756377220154,
"step": 974
},
{
"epoch": 1.1916971916971917,
"grad_norm": 2.201516628265381,
"learning_rate": 5.383244019713987e-06,
"loss": 0.7409634590148926,
"step": 976
},
{
"epoch": 1.1941391941391941,
"grad_norm": 2.08370041847229,
"learning_rate": 5.379578538524857e-06,
"loss": 0.8899902701377869,
"step": 978
},
{
"epoch": 1.1965811965811965,
"grad_norm": 2.196622133255005,
"learning_rate": 5.3759036106397045e-06,
"loss": 1.10507071018219,
"step": 980
},
{
"epoch": 1.199023199023199,
"grad_norm": 1.8301552534103394,
"learning_rate": 5.372219252752263e-06,
"loss": 1.1082658767700195,
"step": 982
},
{
"epoch": 1.2014652014652014,
"grad_norm": 2.458317518234253,
"learning_rate": 5.368525481599099e-06,
"loss": 1.1760859489440918,
"step": 984
},
{
"epoch": 1.2039072039072038,
"grad_norm": 7.877130508422852,
"learning_rate": 5.3648223139595424e-06,
"loss": 0.7288396954536438,
"step": 986
},
{
"epoch": 1.2063492063492063,
"grad_norm": 8.319933891296387,
"learning_rate": 5.361109766655608e-06,
"loss": 0.7217950820922852,
"step": 988
},
{
"epoch": 1.2087912087912087,
"grad_norm": 2.940162181854248,
"learning_rate": 5.357387856551917e-06,
"loss": 0.896350085735321,
"step": 990
},
{
"epoch": 1.2112332112332111,
"grad_norm": 1.7916265726089478,
"learning_rate": 5.353656600555624e-06,
"loss": 1.1697304248809814,
"step": 992
},
{
"epoch": 1.2136752136752136,
"grad_norm": 2.0972354412078857,
"learning_rate": 5.3499160156163355e-06,
"loss": 1.0300666093826294,
"step": 994
},
{
"epoch": 1.2161172161172162,
"grad_norm": 5.365437984466553,
"learning_rate": 5.346166118726038e-06,
"loss": 1.0767557621002197,
"step": 996
},
{
"epoch": 1.2185592185592187,
"grad_norm": 2.915233612060547,
"learning_rate": 5.342406926919017e-06,
"loss": 1.0875110626220703,
"step": 998
},
{
"epoch": 1.221001221001221,
"grad_norm": 3.7588627338409424,
"learning_rate": 5.338638457271783e-06,
"loss": 0.754725992679596,
"step": 1000
},
{
"epoch": 1.2234432234432235,
"grad_norm": 1.6952946186065674,
"learning_rate": 5.33486072690299e-06,
"loss": 0.8764317631721497,
"step": 1002
},
{
"epoch": 1.225885225885226,
"grad_norm": 4.733268737792969,
"learning_rate": 5.33107375297336e-06,
"loss": 1.1004637479782104,
"step": 1004
},
{
"epoch": 1.2283272283272284,
"grad_norm": 2.42063307762146,
"learning_rate": 5.327277552685608e-06,
"loss": 0.8175575733184814,
"step": 1006
},
{
"epoch": 1.2307692307692308,
"grad_norm": 1.254707932472229,
"learning_rate": 5.323472143284354e-06,
"loss": 0.7044642567634583,
"step": 1008
},
{
"epoch": 1.2332112332112333,
"grad_norm": 1.5242104530334473,
"learning_rate": 5.319657542056059e-06,
"loss": 1.1522612571716309,
"step": 1010
},
{
"epoch": 1.2356532356532357,
"grad_norm": 0.8862828612327576,
"learning_rate": 5.315833766328934e-06,
"loss": 0.6255777478218079,
"step": 1012
},
{
"epoch": 1.2380952380952381,
"grad_norm": 2.1223502159118652,
"learning_rate": 5.312000833472867e-06,
"loss": 1.1391960382461548,
"step": 1014
},
{
"epoch": 1.2405372405372406,
"grad_norm": 3.2272346019744873,
"learning_rate": 5.308158760899344e-06,
"loss": 1.0914331674575806,
"step": 1016
},
{
"epoch": 1.242979242979243,
"grad_norm": 1.3873460292816162,
"learning_rate": 5.304307566061369e-06,
"loss": 1.0861551761627197,
"step": 1018
},
{
"epoch": 1.2454212454212454,
"grad_norm": 2.2356748580932617,
"learning_rate": 5.300447266453383e-06,
"loss": 0.7876569032669067,
"step": 1020
},
{
"epoch": 1.2478632478632479,
"grad_norm": 12.627351760864258,
"learning_rate": 5.296577879611189e-06,
"loss": 0.8845806121826172,
"step": 1022
},
{
"epoch": 1.2503052503052503,
"grad_norm": 1.6576119661331177,
"learning_rate": 5.292699423111867e-06,
"loss": 1.0714237689971924,
"step": 1024
},
{
"epoch": 1.2527472527472527,
"grad_norm": 1.7833123207092285,
"learning_rate": 5.2888119145736995e-06,
"loss": 1.0774952173233032,
"step": 1026
},
{
"epoch": 1.2551892551892552,
"grad_norm": 2.1097469329833984,
"learning_rate": 5.284915371656087e-06,
"loss": 1.020277500152588,
"step": 1028
},
{
"epoch": 1.2576312576312576,
"grad_norm": 2.2238245010375977,
"learning_rate": 5.281009812059471e-06,
"loss": 0.9611441493034363,
"step": 1030
},
{
"epoch": 1.26007326007326,
"grad_norm": 3.3395566940307617,
"learning_rate": 5.277095253525251e-06,
"loss": 1.1791523694992065,
"step": 1032
},
{
"epoch": 1.2625152625152625,
"grad_norm": 8.59581470489502,
"learning_rate": 5.273171713835705e-06,
"loss": 1.2326103448867798,
"step": 1034
},
{
"epoch": 1.264957264957265,
"grad_norm": 2.7946317195892334,
"learning_rate": 5.269239210813909e-06,
"loss": 0.7500050067901611,
"step": 1036
},
{
"epoch": 1.2673992673992673,
"grad_norm": 1.8145307302474976,
"learning_rate": 5.265297762323656e-06,
"loss": 0.6441988348960876,
"step": 1038
},
{
"epoch": 1.2698412698412698,
"grad_norm": 1.4063717126846313,
"learning_rate": 5.261347386269376e-06,
"loss": 0.7451719641685486,
"step": 1040
},
{
"epoch": 1.2722832722832722,
"grad_norm": 1.3844637870788574,
"learning_rate": 5.25738810059605e-06,
"loss": 1.1439226865768433,
"step": 1042
},
{
"epoch": 1.2747252747252746,
"grad_norm": 1.3475596904754639,
"learning_rate": 5.253419923289135e-06,
"loss": 0.8882352113723755,
"step": 1044
},
{
"epoch": 1.277167277167277,
"grad_norm": 2.1434133052825928,
"learning_rate": 5.24944287237448e-06,
"loss": 1.088852882385254,
"step": 1046
},
{
"epoch": 1.2796092796092795,
"grad_norm": 4.796410083770752,
"learning_rate": 5.245456965918238e-06,
"loss": 0.7765331268310547,
"step": 1048
},
{
"epoch": 1.282051282051282,
"grad_norm": 3.4816200733184814,
"learning_rate": 5.241462222026794e-06,
"loss": 0.4130716323852539,
"step": 1050
},
{
"epoch": 1.2844932844932844,
"grad_norm": 1.4668511152267456,
"learning_rate": 5.237458658846679e-06,
"loss": 1.024600625038147,
"step": 1052
},
{
"epoch": 1.2869352869352868,
"grad_norm": 2.391469717025757,
"learning_rate": 5.2334462945644816e-06,
"loss": 0.8750287294387817,
"step": 1054
},
{
"epoch": 1.2893772893772895,
"grad_norm": 9.1031494140625,
"learning_rate": 5.229425147406772e-06,
"loss": 0.7807760834693909,
"step": 1056
},
{
"epoch": 1.291819291819292,
"grad_norm": 2.396406888961792,
"learning_rate": 5.225395235640017e-06,
"loss": 1.2667182683944702,
"step": 1058
},
{
"epoch": 1.2942612942612943,
"grad_norm": 1.8769642114639282,
"learning_rate": 5.221356577570503e-06,
"loss": 1.1226063966751099,
"step": 1060
},
{
"epoch": 1.2967032967032968,
"grad_norm": 3.907742500305176,
"learning_rate": 5.217309191544241e-06,
"loss": 0.6234646439552307,
"step": 1062
},
{
"epoch": 1.2991452991452992,
"grad_norm": 1.400547742843628,
"learning_rate": 5.213253095946892e-06,
"loss": 1.0653111934661865,
"step": 1064
},
{
"epoch": 1.3015873015873016,
"grad_norm": 3.261629104614258,
"learning_rate": 5.209188309203678e-06,
"loss": 0.7619715332984924,
"step": 1066
},
{
"epoch": 1.304029304029304,
"grad_norm": 2.847996234893799,
"learning_rate": 5.20511484977931e-06,
"loss": 1.0668418407440186,
"step": 1068
},
{
"epoch": 1.3064713064713065,
"grad_norm": 10.302043914794922,
"learning_rate": 5.201032736177884e-06,
"loss": 0.8345942497253418,
"step": 1070
},
{
"epoch": 1.308913308913309,
"grad_norm": 3.0986499786376953,
"learning_rate": 5.196941986942818e-06,
"loss": 1.0029431581497192,
"step": 1072
},
{
"epoch": 1.3113553113553114,
"grad_norm": 2.5847387313842773,
"learning_rate": 5.192842620656753e-06,
"loss": 0.7861806750297546,
"step": 1074
},
{
"epoch": 1.3137973137973138,
"grad_norm": 7.488101005554199,
"learning_rate": 5.1887346559414754e-06,
"loss": 0.7227391600608826,
"step": 1076
},
{
"epoch": 1.3162393162393162,
"grad_norm": 2.137644052505493,
"learning_rate": 5.184618111457831e-06,
"loss": 1.077176809310913,
"step": 1078
},
{
"epoch": 1.3186813186813187,
"grad_norm": 1.8668829202651978,
"learning_rate": 5.18049300590564e-06,
"loss": 0.9580538868904114,
"step": 1080
},
{
"epoch": 1.321123321123321,
"grad_norm": 2.09850811958313,
"learning_rate": 5.176359358023612e-06,
"loss": 1.007798433303833,
"step": 1082
},
{
"epoch": 1.3235653235653235,
"grad_norm": 1.8359295129776,
"learning_rate": 5.172217186589259e-06,
"loss": 1.0966116189956665,
"step": 1084
},
{
"epoch": 1.326007326007326,
"grad_norm": 3.7404189109802246,
"learning_rate": 5.1680665104188175e-06,
"loss": 1.103602647781372,
"step": 1086
},
{
"epoch": 1.3284493284493284,
"grad_norm": 2.0235581398010254,
"learning_rate": 5.163907348367152e-06,
"loss": 1.0951250791549683,
"step": 1088
},
{
"epoch": 1.3308913308913308,
"grad_norm": 1.9195114374160767,
"learning_rate": 5.159739719327678e-06,
"loss": 0.6674078702926636,
"step": 1090
},
{
"epoch": 1.3333333333333333,
"grad_norm": 32.633548736572266,
"learning_rate": 5.155563642232271e-06,
"loss": 0.9193532466888428,
"step": 1092
},
{
"epoch": 1.3357753357753357,
"grad_norm": 1.8298438787460327,
"learning_rate": 5.151379136051185e-06,
"loss": 1.1200990676879883,
"step": 1094
},
{
"epoch": 1.3382173382173383,
"grad_norm": 2.3729898929595947,
"learning_rate": 5.147186219792962e-06,
"loss": 0.5541606545448303,
"step": 1096
},
{
"epoch": 1.3406593406593408,
"grad_norm": 2.2689027786254883,
"learning_rate": 5.142984912504351e-06,
"loss": 1.203918695449829,
"step": 1098
},
{
"epoch": 1.3431013431013432,
"grad_norm": 1.5877676010131836,
"learning_rate": 5.138775233270214e-06,
"loss": 1.0981378555297852,
"step": 1100
},
{
"epoch": 1.3455433455433456,
"grad_norm": 1.9208431243896484,
"learning_rate": 5.134557201213442e-06,
"loss": 1.0556029081344604,
"step": 1102
},
{
"epoch": 1.347985347985348,
"grad_norm": 3.21431827545166,
"learning_rate": 5.130330835494878e-06,
"loss": 0.8595597147941589,
"step": 1104
},
{
"epoch": 1.3504273504273505,
"grad_norm": 1.5517840385437012,
"learning_rate": 5.126096155313212e-06,
"loss": 1.117148995399475,
"step": 1106
},
{
"epoch": 1.352869352869353,
"grad_norm": 6.060091495513916,
"learning_rate": 5.12185317990491e-06,
"loss": 0.9276140928268433,
"step": 1108
},
{
"epoch": 1.3553113553113554,
"grad_norm": 3.9731998443603516,
"learning_rate": 5.117601928544116e-06,
"loss": 1.1391878128051758,
"step": 1110
},
{
"epoch": 1.3577533577533578,
"grad_norm": 1.3449074029922485,
"learning_rate": 5.113342420542571e-06,
"loss": 1.115125298500061,
"step": 1112
},
{
"epoch": 1.3601953601953602,
"grad_norm": 2.480891466140747,
"learning_rate": 5.109074675249519e-06,
"loss": 1.0627572536468506,
"step": 1114
},
{
"epoch": 1.3626373626373627,
"grad_norm": 7.083606719970703,
"learning_rate": 5.104798712051628e-06,
"loss": 0.46798741817474365,
"step": 1116
},
{
"epoch": 1.3650793650793651,
"grad_norm": 2.1328299045562744,
"learning_rate": 5.1005145503728895e-06,
"loss": 1.190787672996521,
"step": 1118
},
{
"epoch": 1.3675213675213675,
"grad_norm": 1.7239574193954468,
"learning_rate": 5.096222209674545e-06,
"loss": 1.1722108125686646,
"step": 1120
},
{
"epoch": 1.36996336996337,
"grad_norm": 10.229917526245117,
"learning_rate": 5.091921709454986e-06,
"loss": 0.9836089015007019,
"step": 1122
},
{
"epoch": 1.3724053724053724,
"grad_norm": 1.4455071687698364,
"learning_rate": 5.087613069249669e-06,
"loss": 0.9871035814285278,
"step": 1124
},
{
"epoch": 1.3748473748473748,
"grad_norm": 3.920485734939575,
"learning_rate": 5.08329630863103e-06,
"loss": 1.271277904510498,
"step": 1126
},
{
"epoch": 1.3772893772893773,
"grad_norm": 3.3860907554626465,
"learning_rate": 5.078971447208389e-06,
"loss": 1.0518757104873657,
"step": 1128
},
{
"epoch": 1.3797313797313797,
"grad_norm": 1.6341384649276733,
"learning_rate": 5.074638504627869e-06,
"loss": 0.8103001117706299,
"step": 1130
},
{
"epoch": 1.3821733821733821,
"grad_norm": 6.043010234832764,
"learning_rate": 5.070297500572299e-06,
"loss": 0.7450224757194519,
"step": 1132
},
{
"epoch": 1.3846153846153846,
"grad_norm": 4.167097091674805,
"learning_rate": 5.065948454761128e-06,
"loss": 0.5298635363578796,
"step": 1134
},
{
"epoch": 1.387057387057387,
"grad_norm": 1.658325433731079,
"learning_rate": 5.06159138695034e-06,
"loss": 0.8351241946220398,
"step": 1136
},
{
"epoch": 1.3894993894993894,
"grad_norm": 2.1943113803863525,
"learning_rate": 5.057226316932354e-06,
"loss": 1.1314613819122314,
"step": 1138
},
{
"epoch": 1.3919413919413919,
"grad_norm": 2.5029122829437256,
"learning_rate": 5.052853264535944e-06,
"loss": 0.8110060095787048,
"step": 1140
},
{
"epoch": 1.3943833943833943,
"grad_norm": 5.180810928344727,
"learning_rate": 5.048472249626143e-06,
"loss": 0.941763162612915,
"step": 1142
},
{
"epoch": 1.3968253968253967,
"grad_norm": 3.3859798908233643,
"learning_rate": 5.044083292104154e-06,
"loss": 0.6929471492767334,
"step": 1144
},
{
"epoch": 1.3992673992673992,
"grad_norm": 1.9731738567352295,
"learning_rate": 5.039686411907261e-06,
"loss": 1.062778115272522,
"step": 1146
},
{
"epoch": 1.4017094017094016,
"grad_norm": 9.248348236083984,
"learning_rate": 5.035281629008738e-06,
"loss": 0.907078742980957,
"step": 1148
},
{
"epoch": 1.404151404151404,
"grad_norm": 1.8255606889724731,
"learning_rate": 5.030868963417756e-06,
"loss": 0.7409341931343079,
"step": 1150
},
{
"epoch": 1.4065934065934065,
"grad_norm": 1.724865198135376,
"learning_rate": 5.026448435179296e-06,
"loss": 0.9270036816596985,
"step": 1152
},
{
"epoch": 1.409035409035409,
"grad_norm": 2.166942596435547,
"learning_rate": 5.022020064374055e-06,
"loss": 0.9618323445320129,
"step": 1154
},
{
"epoch": 1.4114774114774113,
"grad_norm": 6.376034736633301,
"learning_rate": 5.017583871118354e-06,
"loss": 0.7564554810523987,
"step": 1156
},
{
"epoch": 1.4139194139194138,
"grad_norm": 2.646700859069824,
"learning_rate": 5.013139875564054e-06,
"loss": 0.9836897253990173,
"step": 1158
},
{
"epoch": 1.4163614163614164,
"grad_norm": 3.115039825439453,
"learning_rate": 5.008688097898451e-06,
"loss": 0.9553764462471008,
"step": 1160
},
{
"epoch": 1.4188034188034189,
"grad_norm": 1.7480710744857788,
"learning_rate": 5.004228558344195e-06,
"loss": 0.7161940932273865,
"step": 1162
},
{
"epoch": 1.4212454212454213,
"grad_norm": 7.046204090118408,
"learning_rate": 4.9997612771592e-06,
"loss": 0.7383595705032349,
"step": 1164
},
{
"epoch": 1.4236874236874237,
"grad_norm": 2.7713022232055664,
"learning_rate": 4.9952862746365396e-06,
"loss": 1.1871694326400757,
"step": 1166
},
{
"epoch": 1.4261294261294262,
"grad_norm": 0.5916857123374939,
"learning_rate": 4.990803571104367e-06,
"loss": 1.031806230545044,
"step": 1168
},
{
"epoch": 1.4285714285714286,
"grad_norm": 3.3403420448303223,
"learning_rate": 4.9863131869258165e-06,
"loss": 0.8283597230911255,
"step": 1170
},
{
"epoch": 1.431013431013431,
"grad_norm": 2.6560847759246826,
"learning_rate": 4.981815142498913e-06,
"loss": 1.10054612159729,
"step": 1172
},
{
"epoch": 1.4334554334554335,
"grad_norm": 1.0640076398849487,
"learning_rate": 4.977309458256478e-06,
"loss": 1.0439369678497314,
"step": 1174
},
{
"epoch": 1.435897435897436,
"grad_norm": 2.3129022121429443,
"learning_rate": 4.972796154666038e-06,
"loss": 1.143113136291504,
"step": 1176
},
{
"epoch": 1.4383394383394383,
"grad_norm": 3.9890077114105225,
"learning_rate": 4.968275252229734e-06,
"loss": 1.0084482431411743,
"step": 1178
},
{
"epoch": 1.4407814407814408,
"grad_norm": 2.268794536590576,
"learning_rate": 4.96374677148422e-06,
"loss": 0.9494715929031372,
"step": 1180
},
{
"epoch": 1.4432234432234432,
"grad_norm": 1.620401382446289,
"learning_rate": 4.959210733000581e-06,
"loss": 0.7894858717918396,
"step": 1182
},
{
"epoch": 1.4456654456654456,
"grad_norm": 2.264465808868408,
"learning_rate": 4.954667157384227e-06,
"loss": 1.153456211090088,
"step": 1184
},
{
"epoch": 1.448107448107448,
"grad_norm": 2.0204758644104004,
"learning_rate": 4.9501160652748135e-06,
"loss": 1.0387881994247437,
"step": 1186
},
{
"epoch": 1.4505494505494505,
"grad_norm": 4.494890213012695,
"learning_rate": 4.945557477346135e-06,
"loss": 0.8388749957084656,
"step": 1188
},
{
"epoch": 1.452991452991453,
"grad_norm": 5.138271808624268,
"learning_rate": 4.940991414306042e-06,
"loss": 1.0478739738464355,
"step": 1190
},
{
"epoch": 1.4554334554334554,
"grad_norm": 1.7801316976547241,
"learning_rate": 4.936417896896333e-06,
"loss": 0.8928017616271973,
"step": 1192
},
{
"epoch": 1.4578754578754578,
"grad_norm": 5.119191646575928,
"learning_rate": 4.931836945892678e-06,
"loss": 0.7815787196159363,
"step": 1194
},
{
"epoch": 1.4603174603174602,
"grad_norm": 2.283468008041382,
"learning_rate": 4.92724858210451e-06,
"loss": 0.7689943909645081,
"step": 1196
},
{
"epoch": 1.462759462759463,
"grad_norm": 1.3747724294662476,
"learning_rate": 4.922652826374934e-06,
"loss": 1.0252704620361328,
"step": 1198
},
{
"epoch": 1.4652014652014653,
"grad_norm": 3.9742517471313477,
"learning_rate": 4.918049699580636e-06,
"loss": 1.176680326461792,
"step": 1200
},
{
"epoch": 1.4676434676434678,
"grad_norm": 2.1025235652923584,
"learning_rate": 4.913439222631789e-06,
"loss": 1.0160706043243408,
"step": 1202
},
{
"epoch": 1.4700854700854702,
"grad_norm": 4.321649074554443,
"learning_rate": 4.9088214164719465e-06,
"loss": 1.0500729084014893,
"step": 1204
},
{
"epoch": 1.4725274725274726,
"grad_norm": 4.172434329986572,
"learning_rate": 4.9041963020779625e-06,
"loss": 0.9391635060310364,
"step": 1206
},
{
"epoch": 1.474969474969475,
"grad_norm": 8.891393661499023,
"learning_rate": 4.899563900459888e-06,
"loss": 0.8474172949790955,
"step": 1208
},
{
"epoch": 1.4774114774114775,
"grad_norm": 5.215544700622559,
"learning_rate": 4.894924232660875e-06,
"loss": 0.8866111040115356,
"step": 1210
},
{
"epoch": 1.47985347985348,
"grad_norm": 3.0435311794281006,
"learning_rate": 4.890277319757084e-06,
"loss": 1.029344916343689,
"step": 1212
},
{
"epoch": 1.4822954822954824,
"grad_norm": 1.7593703269958496,
"learning_rate": 4.8856231828575886e-06,
"loss": 1.1196095943450928,
"step": 1214
},
{
"epoch": 1.4847374847374848,
"grad_norm": 2.1020376682281494,
"learning_rate": 4.880961843104277e-06,
"loss": 1.0347650051116943,
"step": 1216
},
{
"epoch": 1.4871794871794872,
"grad_norm": 1.6005417108535767,
"learning_rate": 4.876293321671754e-06,
"loss": 0.2354688048362732,
"step": 1218
},
{
"epoch": 1.4896214896214897,
"grad_norm": 1.6489475965499878,
"learning_rate": 4.871617639767254e-06,
"loss": 0.6391555070877075,
"step": 1220
},
{
"epoch": 1.492063492063492,
"grad_norm": 2.582890748977661,
"learning_rate": 4.8669348186305345e-06,
"loss": 1.1244972944259644,
"step": 1222
},
{
"epoch": 1.4945054945054945,
"grad_norm": 3.2143290042877197,
"learning_rate": 4.862244879533784e-06,
"loss": 0.8394033312797546,
"step": 1224
},
{
"epoch": 1.496947496947497,
"grad_norm": 2.631812810897827,
"learning_rate": 4.857547843781526e-06,
"loss": 1.1298140287399292,
"step": 1226
},
{
"epoch": 1.4993894993894994,
"grad_norm": 3.6919734477996826,
"learning_rate": 4.8528437327105185e-06,
"loss": 0.4192117154598236,
"step": 1228
},
{
"epoch": 1.5018315018315018,
"grad_norm": 8.116545677185059,
"learning_rate": 4.848132567689667e-06,
"loss": 0.8103526830673218,
"step": 1230
},
{
"epoch": 1.5042735042735043,
"grad_norm": 1.7116774320602417,
"learning_rate": 4.843414370119911e-06,
"loss": 0.6593791246414185,
"step": 1232
},
{
"epoch": 1.5067155067155067,
"grad_norm": 3.8022818565368652,
"learning_rate": 4.838689161434145e-06,
"loss": 1.126307725906372,
"step": 1234
},
{
"epoch": 1.5091575091575091,
"grad_norm": 1.9455960988998413,
"learning_rate": 4.8339569630971035e-06,
"loss": 0.9250009655952454,
"step": 1236
},
{
"epoch": 1.5115995115995116,
"grad_norm": 1.9260501861572266,
"learning_rate": 4.829217796605279e-06,
"loss": 0.8760790228843689,
"step": 1238
},
{
"epoch": 1.514041514041514,
"grad_norm": 2.5248501300811768,
"learning_rate": 4.824471683486813e-06,
"loss": 1.0439480543136597,
"step": 1240
},
{
"epoch": 1.5164835164835164,
"grad_norm": 1.3928033113479614,
"learning_rate": 4.819718645301406e-06,
"loss": 1.0599212646484375,
"step": 1242
},
{
"epoch": 1.5189255189255189,
"grad_norm": 6.743286609649658,
"learning_rate": 4.814958703640217e-06,
"loss": 0.6897165775299072,
"step": 1244
},
{
"epoch": 1.5213675213675213,
"grad_norm": 3.9725565910339355,
"learning_rate": 4.810191880125759e-06,
"loss": 0.8072797060012817,
"step": 1246
},
{
"epoch": 1.5238095238095237,
"grad_norm": 2.459388256072998,
"learning_rate": 4.805418196411814e-06,
"loss": 0.8643714785575867,
"step": 1248
},
{
"epoch": 1.5262515262515262,
"grad_norm": 2.454529047012329,
"learning_rate": 4.80063767418332e-06,
"loss": 1.1347299814224243,
"step": 1250
},
{
"epoch": 1.5286935286935286,
"grad_norm": 6.470121383666992,
"learning_rate": 4.795850335156287e-06,
"loss": 0.845814049243927,
"step": 1252
},
{
"epoch": 1.531135531135531,
"grad_norm": 1.3302001953125,
"learning_rate": 4.791056201077684e-06,
"loss": 1.0788911581039429,
"step": 1254
},
{
"epoch": 1.5335775335775335,
"grad_norm": 2.616361379623413,
"learning_rate": 4.786255293725351e-06,
"loss": 1.188575029373169,
"step": 1256
},
{
"epoch": 1.536019536019536,
"grad_norm": 0.9566062092781067,
"learning_rate": 4.781447634907898e-06,
"loss": 0.3384564220905304,
"step": 1258
},
{
"epoch": 1.5384615384615383,
"grad_norm": 4.599119186401367,
"learning_rate": 4.776633246464597e-06,
"loss": 0.7301375269889832,
"step": 1260
},
{
"epoch": 1.5409035409035408,
"grad_norm": 3.47401762008667,
"learning_rate": 4.7718121502653e-06,
"loss": 0.5812518000602722,
"step": 1262
},
{
"epoch": 1.5433455433455432,
"grad_norm": 2.5774216651916504,
"learning_rate": 4.76698436821032e-06,
"loss": 0.7762341499328613,
"step": 1264
},
{
"epoch": 1.5457875457875456,
"grad_norm": 2.030524492263794,
"learning_rate": 4.762149922230351e-06,
"loss": 0.7970203757286072,
"step": 1266
},
{
"epoch": 1.5482295482295483,
"grad_norm": 1.4194061756134033,
"learning_rate": 4.757308834286349e-06,
"loss": 1.1093299388885498,
"step": 1268
},
{
"epoch": 1.5506715506715507,
"grad_norm": 13.08940601348877,
"learning_rate": 4.752461126369447e-06,
"loss": 0.5118930339813232,
"step": 1270
},
{
"epoch": 1.5531135531135531,
"grad_norm": 1.599934697151184,
"learning_rate": 4.747606820500848e-06,
"loss": 1.1399681568145752,
"step": 1272
},
{
"epoch": 1.5555555555555556,
"grad_norm": 3.1104736328125,
"learning_rate": 4.7427459387317305e-06,
"loss": 0.9280550479888916,
"step": 1274
},
{
"epoch": 1.557997557997558,
"grad_norm": 2.4784369468688965,
"learning_rate": 4.73787850314314e-06,
"loss": 1.3923664093017578,
"step": 1276
},
{
"epoch": 1.5604395604395604,
"grad_norm": 1.778232216835022,
"learning_rate": 4.7330045358458965e-06,
"loss": 0.9597633481025696,
"step": 1278
},
{
"epoch": 1.5628815628815629,
"grad_norm": 1.8193713426589966,
"learning_rate": 4.728124058980489e-06,
"loss": 1.052099585533142,
"step": 1280
},
{
"epoch": 1.5653235653235653,
"grad_norm": 3.915374279022217,
"learning_rate": 4.723237094716978e-06,
"loss": 0.6439841389656067,
"step": 1282
},
{
"epoch": 1.5677655677655677,
"grad_norm": 1.5543129444122314,
"learning_rate": 4.718343665254892e-06,
"loss": 1.0367350578308105,
"step": 1284
},
{
"epoch": 1.5702075702075702,
"grad_norm": 1.7215509414672852,
"learning_rate": 4.713443792823132e-06,
"loss": 1.059045672416687,
"step": 1286
},
{
"epoch": 1.5726495726495726,
"grad_norm": 2.069295644760132,
"learning_rate": 4.708537499679864e-06,
"loss": 0.5439974069595337,
"step": 1288
},
{
"epoch": 1.575091575091575,
"grad_norm": 2.160947561264038,
"learning_rate": 4.7036248081124194e-06,
"loss": 0.8789887428283691,
"step": 1290
},
{
"epoch": 1.5775335775335775,
"grad_norm": 2.9418277740478516,
"learning_rate": 4.698705740437199e-06,
"loss": 0.36467528343200684,
"step": 1292
},
{
"epoch": 1.5799755799755801,
"grad_norm": 1.783849835395813,
"learning_rate": 4.693780318999566e-06,
"loss": 0.7312822341918945,
"step": 1294
},
{
"epoch": 1.5824175824175826,
"grad_norm": 1.2510713338851929,
"learning_rate": 4.688848566173743e-06,
"loss": 1.0698741674423218,
"step": 1296
},
{
"epoch": 1.584859584859585,
"grad_norm": 1.7499157190322876,
"learning_rate": 4.683910504362718e-06,
"loss": 1.1414761543273926,
"step": 1298
},
{
"epoch": 1.5873015873015874,
"grad_norm": 2.5625784397125244,
"learning_rate": 4.678966155998138e-06,
"loss": 0.963818371295929,
"step": 1300
},
{
"epoch": 1.5897435897435899,
"grad_norm": 3.9518566131591797,
"learning_rate": 4.674015543540204e-06,
"loss": 1.1232004165649414,
"step": 1302
},
{
"epoch": 1.5921855921855923,
"grad_norm": 4.137270450592041,
"learning_rate": 4.669058689477577e-06,
"loss": 1.0733479261398315,
"step": 1304
},
{
"epoch": 1.5946275946275947,
"grad_norm": 1.5602595806121826,
"learning_rate": 4.664095616327265e-06,
"loss": 0.6966079473495483,
"step": 1306
},
{
"epoch": 1.5970695970695972,
"grad_norm": 1.9313784837722778,
"learning_rate": 4.6591263466345315e-06,
"loss": 1.07318115234375,
"step": 1308
},
{
"epoch": 1.5995115995115996,
"grad_norm": 3.182425022125244,
"learning_rate": 4.654150902972787e-06,
"loss": 0.9745839834213257,
"step": 1310
},
{
"epoch": 1.601953601953602,
"grad_norm": 1.5878440141677856,
"learning_rate": 4.649169307943489e-06,
"loss": 1.0333638191223145,
"step": 1312
},
{
"epoch": 1.6043956043956045,
"grad_norm": 2.5433406829833984,
"learning_rate": 4.644181584176036e-06,
"loss": 1.1256206035614014,
"step": 1314
},
{
"epoch": 1.606837606837607,
"grad_norm": 2.123772621154785,
"learning_rate": 4.639187754327669e-06,
"loss": 1.1688880920410156,
"step": 1316
},
{
"epoch": 1.6092796092796093,
"grad_norm": 2.0329692363739014,
"learning_rate": 4.634187841083365e-06,
"loss": 1.0112664699554443,
"step": 1318
},
{
"epoch": 1.6117216117216118,
"grad_norm": 1.9080368280410767,
"learning_rate": 4.629181867155735e-06,
"loss": 1.2758021354675293,
"step": 1320
},
{
"epoch": 1.6141636141636142,
"grad_norm": 1.2908200025558472,
"learning_rate": 4.624169855284923e-06,
"loss": 1.077506184577942,
"step": 1322
},
{
"epoch": 1.6166056166056166,
"grad_norm": 4.5929975509643555,
"learning_rate": 4.619151828238503e-06,
"loss": 1.1034421920776367,
"step": 1324
},
{
"epoch": 1.619047619047619,
"grad_norm": 3.1997547149658203,
"learning_rate": 4.6141278088113674e-06,
"loss": 0.7947360277175903,
"step": 1326
},
{
"epoch": 1.6214896214896215,
"grad_norm": 1.4481651782989502,
"learning_rate": 4.609097819825635e-06,
"loss": 1.0906190872192383,
"step": 1328
},
{
"epoch": 1.623931623931624,
"grad_norm": 1.5239222049713135,
"learning_rate": 4.604061884130537e-06,
"loss": 1.1238385438919067,
"step": 1330
},
{
"epoch": 1.6263736263736264,
"grad_norm": 17.961843490600586,
"learning_rate": 4.599020024602325e-06,
"loss": 0.872590959072113,
"step": 1332
},
{
"epoch": 1.6288156288156288,
"grad_norm": 2.320810556411743,
"learning_rate": 4.593972264144154e-06,
"loss": 1.0335891246795654,
"step": 1334
},
{
"epoch": 1.6312576312576312,
"grad_norm": 1.0677226781845093,
"learning_rate": 4.588918625685987e-06,
"loss": 0.6070804595947266,
"step": 1336
},
{
"epoch": 1.6336996336996337,
"grad_norm": 6.536888122558594,
"learning_rate": 4.5838591321844895e-06,
"loss": 1.0951392650604248,
"step": 1338
},
{
"epoch": 1.636141636141636,
"grad_norm": 15.38590145111084,
"learning_rate": 4.578793806622921e-06,
"loss": 0.5872366428375244,
"step": 1340
},
{
"epoch": 1.6385836385836385,
"grad_norm": 2.773411273956299,
"learning_rate": 4.573722672011037e-06,
"loss": 0.44960087537765503,
"step": 1342
},
{
"epoch": 1.641025641025641,
"grad_norm": 1.476400375366211,
"learning_rate": 4.568645751384979e-06,
"loss": 1.1929024457931519,
"step": 1344
},
{
"epoch": 1.6434676434676434,
"grad_norm": 1.6975760459899902,
"learning_rate": 4.563563067807173e-06,
"loss": 1.1604607105255127,
"step": 1346
},
{
"epoch": 1.6459096459096458,
"grad_norm": 27.678466796875,
"learning_rate": 4.558474644366225e-06,
"loss": 0.6984640955924988,
"step": 1348
},
{
"epoch": 1.6483516483516483,
"grad_norm": 4.300885200500488,
"learning_rate": 4.553380504176812e-06,
"loss": 0.7492251396179199,
"step": 1350
},
{
"epoch": 1.6507936507936507,
"grad_norm": 1.742562174797058,
"learning_rate": 4.548280670379582e-06,
"loss": 1.089888334274292,
"step": 1352
},
{
"epoch": 1.6532356532356531,
"grad_norm": 1.481261968612671,
"learning_rate": 4.543175166141047e-06,
"loss": 1.1312068700790405,
"step": 1354
},
{
"epoch": 1.6556776556776556,
"grad_norm": 2.4384007453918457,
"learning_rate": 4.538064014653475e-06,
"loss": 1.1034740209579468,
"step": 1356
},
{
"epoch": 1.658119658119658,
"grad_norm": 1.1855982542037964,
"learning_rate": 4.53294723913479e-06,
"loss": 0.5771256685256958,
"step": 1358
},
{
"epoch": 1.6605616605616604,
"grad_norm": 2.1966373920440674,
"learning_rate": 4.527824862828463e-06,
"loss": 0.789630651473999,
"step": 1360
},
{
"epoch": 1.6630036630036629,
"grad_norm": 2.3562395572662354,
"learning_rate": 4.522696909003407e-06,
"loss": 1.1317930221557617,
"step": 1362
},
{
"epoch": 1.6654456654456653,
"grad_norm": 1.2977938652038574,
"learning_rate": 4.517563400953871e-06,
"loss": 1.0996814966201782,
"step": 1364
},
{
"epoch": 1.6678876678876677,
"grad_norm": 2.646000385284424,
"learning_rate": 4.512424361999335e-06,
"loss": 1.086019515991211,
"step": 1366
},
{
"epoch": 1.6703296703296702,
"grad_norm": 2.2506117820739746,
"learning_rate": 4.507279815484404e-06,
"loss": 1.0686463117599487,
"step": 1368
},
{
"epoch": 1.6727716727716728,
"grad_norm": 3.5996057987213135,
"learning_rate": 4.5021297847787e-06,
"loss": 1.1250067949295044,
"step": 1370
},
{
"epoch": 1.6752136752136753,
"grad_norm": 1.2507672309875488,
"learning_rate": 4.496974293276761e-06,
"loss": 0.6428678631782532,
"step": 1372
},
{
"epoch": 1.6776556776556777,
"grad_norm": 16.312124252319336,
"learning_rate": 4.491813364397928e-06,
"loss": 0.3962022364139557,
"step": 1374
},
{
"epoch": 1.6800976800976801,
"grad_norm": 4.418454647064209,
"learning_rate": 4.486647021586242e-06,
"loss": 1.1909222602844238,
"step": 1376
},
{
"epoch": 1.6825396825396826,
"grad_norm": 1.5935609340667725,
"learning_rate": 4.48147528831034e-06,
"loss": 1.1969082355499268,
"step": 1378
},
{
"epoch": 1.684981684981685,
"grad_norm": 0.7602794766426086,
"learning_rate": 4.476298188063342e-06,
"loss": 1.0184684991836548,
"step": 1380
},
{
"epoch": 1.6874236874236874,
"grad_norm": 4.655670642852783,
"learning_rate": 4.471115744362751e-06,
"loss": 1.003674030303955,
"step": 1382
},
{
"epoch": 1.6898656898656899,
"grad_norm": 2.1681458950042725,
"learning_rate": 4.465927980750341e-06,
"loss": 1.1820536851882935,
"step": 1384
},
{
"epoch": 1.6923076923076923,
"grad_norm": 1.1264445781707764,
"learning_rate": 4.460734920792055e-06,
"loss": 1.1060798168182373,
"step": 1386
},
{
"epoch": 1.6947496947496947,
"grad_norm": 1.4772337675094604,
"learning_rate": 4.455536588077892e-06,
"loss": 1.1331332921981812,
"step": 1388
},
{
"epoch": 1.6971916971916972,
"grad_norm": 6.615632057189941,
"learning_rate": 4.450333006221807e-06,
"loss": 0.4286938011646271,
"step": 1390
},
{
"epoch": 1.6996336996336996,
"grad_norm": 1.7158353328704834,
"learning_rate": 4.445124198861595e-06,
"loss": 1.0181334018707275,
"step": 1392
},
{
"epoch": 1.702075702075702,
"grad_norm": 1.6327100992202759,
"learning_rate": 4.439910189658791e-06,
"loss": 0.9523332118988037,
"step": 1394
},
{
"epoch": 1.7045177045177047,
"grad_norm": 2.332667350769043,
"learning_rate": 4.43469100229856e-06,
"loss": 1.1586647033691406,
"step": 1396
},
{
"epoch": 1.7069597069597071,
"grad_norm": 5.861179828643799,
"learning_rate": 4.4294666604895896e-06,
"loss": 0.8811005353927612,
"step": 1398
},
{
"epoch": 1.7094017094017095,
"grad_norm": 8.61947250366211,
"learning_rate": 4.42423718796398e-06,
"loss": 0.43517377972602844,
"step": 1400
},
{
"epoch": 1.711843711843712,
"grad_norm": 3.1414754390716553,
"learning_rate": 4.41900260847714e-06,
"loss": 0.9462797045707703,
"step": 1402
},
{
"epoch": 1.7142857142857144,
"grad_norm": 3.1354236602783203,
"learning_rate": 4.413762945807678e-06,
"loss": 1.1175096035003662,
"step": 1404
},
{
"epoch": 1.7167277167277168,
"grad_norm": 3.440849542617798,
"learning_rate": 4.40851822375729e-06,
"loss": 0.5517919063568115,
"step": 1406
},
{
"epoch": 1.7191697191697193,
"grad_norm": 2.9297659397125244,
"learning_rate": 4.403268466150658e-06,
"loss": 1.0800657272338867,
"step": 1408
},
{
"epoch": 1.7216117216117217,
"grad_norm": 3.0104856491088867,
"learning_rate": 4.398013696835337e-06,
"loss": 0.7462184429168701,
"step": 1410
},
{
"epoch": 1.7240537240537241,
"grad_norm": 7.846835136413574,
"learning_rate": 4.39275393968165e-06,
"loss": 0.8748076558113098,
"step": 1412
},
{
"epoch": 1.7264957264957266,
"grad_norm": 1.4625284671783447,
"learning_rate": 4.3874892185825736e-06,
"loss": 1.1260775327682495,
"step": 1414
},
{
"epoch": 1.728937728937729,
"grad_norm": 7.651988506317139,
"learning_rate": 4.3822195574536385e-06,
"loss": 0.911249041557312,
"step": 1416
},
{
"epoch": 1.7313797313797314,
"grad_norm": 1.9895511865615845,
"learning_rate": 4.3769449802328134e-06,
"loss": 1.1256855726242065,
"step": 1418
},
{
"epoch": 1.7338217338217339,
"grad_norm": 3.724893093109131,
"learning_rate": 4.371665510880398e-06,
"loss": 0.8602355122566223,
"step": 1420
},
{
"epoch": 1.7362637362637363,
"grad_norm": 6.229957580566406,
"learning_rate": 4.366381173378918e-06,
"loss": 0.8753630518913269,
"step": 1422
},
{
"epoch": 1.7387057387057387,
"grad_norm": 3.7382354736328125,
"learning_rate": 4.36109199173301e-06,
"loss": 0.8494032025337219,
"step": 1424
},
{
"epoch": 1.7411477411477412,
"grad_norm": 2.8685145378112793,
"learning_rate": 4.355797989969319e-06,
"loss": 0.7920373678207397,
"step": 1426
},
{
"epoch": 1.7435897435897436,
"grad_norm": 2.0900766849517822,
"learning_rate": 4.350499192136381e-06,
"loss": 1.128261685371399,
"step": 1428
},
{
"epoch": 1.746031746031746,
"grad_norm": 1.9242794513702393,
"learning_rate": 4.345195622304524e-06,
"loss": 1.1854214668273926,
"step": 1430
},
{
"epoch": 1.7484737484737485,
"grad_norm": 6.52469539642334,
"learning_rate": 4.3398873045657505e-06,
"loss": 0.7444266080856323,
"step": 1432
},
{
"epoch": 1.750915750915751,
"grad_norm": 2.14506459236145,
"learning_rate": 4.334574263033629e-06,
"loss": 0.6873703002929688,
"step": 1434
},
{
"epoch": 1.7533577533577533,
"grad_norm": 3.528667688369751,
"learning_rate": 4.329256521843191e-06,
"loss": 0.7260130643844604,
"step": 1436
},
{
"epoch": 1.7557997557997558,
"grad_norm": 3.055670976638794,
"learning_rate": 4.323934105150814e-06,
"loss": 0.5219581723213196,
"step": 1438
},
{
"epoch": 1.7582417582417582,
"grad_norm": 1.5880358219146729,
"learning_rate": 4.318607037134113e-06,
"loss": 1.0635778903961182,
"step": 1440
},
{
"epoch": 1.7606837606837606,
"grad_norm": 1.6739574670791626,
"learning_rate": 4.313275341991835e-06,
"loss": 0.9275929927825928,
"step": 1442
},
{
"epoch": 1.763125763125763,
"grad_norm": 2.3931827545166016,
"learning_rate": 4.307939043943743e-06,
"loss": 0.7328923940658569,
"step": 1444
},
{
"epoch": 1.7655677655677655,
"grad_norm": 1.5773603916168213,
"learning_rate": 4.302598167230514e-06,
"loss": 1.1174166202545166,
"step": 1446
},
{
"epoch": 1.768009768009768,
"grad_norm": 1.3908194303512573,
"learning_rate": 4.297252736113619e-06,
"loss": 1.104804277420044,
"step": 1448
},
{
"epoch": 1.7704517704517704,
"grad_norm": 2.2045950889587402,
"learning_rate": 4.291902774875222e-06,
"loss": 1.0062525272369385,
"step": 1450
},
{
"epoch": 1.7728937728937728,
"grad_norm": 4.392017841339111,
"learning_rate": 4.2865483078180606e-06,
"loss": 0.49355483055114746,
"step": 1452
},
{
"epoch": 1.7753357753357752,
"grad_norm": 1.741620421409607,
"learning_rate": 4.281189359265346e-06,
"loss": 0.801204264163971,
"step": 1454
},
{
"epoch": 1.7777777777777777,
"grad_norm": 4.615424633026123,
"learning_rate": 4.275825953560643e-06,
"loss": 0.5927553772926331,
"step": 1456
},
{
"epoch": 1.7802197802197801,
"grad_norm": 1.8360828161239624,
"learning_rate": 4.270458115067766e-06,
"loss": 1.1054664850234985,
"step": 1458
},
{
"epoch": 1.7826617826617825,
"grad_norm": 28.914661407470703,
"learning_rate": 4.265085868170663e-06,
"loss": 1.1111050844192505,
"step": 1460
},
{
"epoch": 1.785103785103785,
"grad_norm": 2.3152248859405518,
"learning_rate": 4.259709237273311e-06,
"loss": 0.9055600762367249,
"step": 1462
},
{
"epoch": 1.7875457875457874,
"grad_norm": 1.7210235595703125,
"learning_rate": 4.2543282467995985e-06,
"loss": 0.7722167372703552,
"step": 1464
},
{
"epoch": 1.7899877899877898,
"grad_norm": 2.618022918701172,
"learning_rate": 4.248942921193221e-06,
"loss": 0.8464701175689697,
"step": 1466
},
{
"epoch": 1.7924297924297923,
"grad_norm": 6.523756980895996,
"learning_rate": 4.243553284917566e-06,
"loss": 0.848709225654602,
"step": 1468
},
{
"epoch": 1.7948717948717947,
"grad_norm": 2.5969009399414062,
"learning_rate": 4.2381593624556e-06,
"loss": 1.0547089576721191,
"step": 1470
},
{
"epoch": 1.7973137973137974,
"grad_norm": 1.5815178155899048,
"learning_rate": 4.232761178309761e-06,
"loss": 1.0379605293273926,
"step": 1472
},
{
"epoch": 1.7997557997557998,
"grad_norm": 2.534233570098877,
"learning_rate": 4.22735875700185e-06,
"loss": 0.7152073979377747,
"step": 1474
},
{
"epoch": 1.8021978021978022,
"grad_norm": 2.298525810241699,
"learning_rate": 4.22195212307291e-06,
"loss": 0.691750168800354,
"step": 1476
},
{
"epoch": 1.8046398046398047,
"grad_norm": 1.962921380996704,
"learning_rate": 4.216541301083124e-06,
"loss": 0.769491970539093,
"step": 1478
},
{
"epoch": 1.807081807081807,
"grad_norm": 2.270589828491211,
"learning_rate": 4.211126315611698e-06,
"loss": 0.4893834888935089,
"step": 1480
},
{
"epoch": 1.8095238095238095,
"grad_norm": 2.069971799850464,
"learning_rate": 4.205707191256752e-06,
"loss": 0.7786384224891663,
"step": 1482
},
{
"epoch": 1.811965811965812,
"grad_norm": 1.248873233795166,
"learning_rate": 4.200283952635208e-06,
"loss": 1.2088333368301392,
"step": 1484
},
{
"epoch": 1.8144078144078144,
"grad_norm": 1.4751837253570557,
"learning_rate": 4.1948566243826756e-06,
"loss": 1.0881723165512085,
"step": 1486
},
{
"epoch": 1.8168498168498168,
"grad_norm": 2.016812562942505,
"learning_rate": 4.189425231153342e-06,
"loss": 0.6552821397781372,
"step": 1488
},
{
"epoch": 1.8192918192918193,
"grad_norm": 1.4564069509506226,
"learning_rate": 4.183989797619862e-06,
"loss": 0.8445190191268921,
"step": 1490
},
{
"epoch": 1.8217338217338217,
"grad_norm": 1.54276704788208,
"learning_rate": 4.178550348473242e-06,
"loss": 1.0718238353729248,
"step": 1492
},
{
"epoch": 1.8241758241758241,
"grad_norm": 3.247328996658325,
"learning_rate": 4.17310690842273e-06,
"loss": 0.7224799990653992,
"step": 1494
},
{
"epoch": 1.8266178266178266,
"grad_norm": 1.877454161643982,
"learning_rate": 4.167659502195704e-06,
"loss": 0.8526558876037598,
"step": 1496
},
{
"epoch": 1.8290598290598292,
"grad_norm": 1.8572306632995605,
"learning_rate": 4.162208154537559e-06,
"loss": 1.5043039321899414,
"step": 1498
},
{
"epoch": 1.8315018315018317,
"grad_norm": 2.081958055496216,
"learning_rate": 4.156752890211592e-06,
"loss": 1.2200380563735962,
"step": 1500
},
{
"epoch": 1.833943833943834,
"grad_norm": 7.550018787384033,
"learning_rate": 4.151293733998894e-06,
"loss": 0.8529073596000671,
"step": 1502
},
{
"epoch": 1.8363858363858365,
"grad_norm": 1.3198851346969604,
"learning_rate": 4.145830710698235e-06,
"loss": 0.8785728812217712,
"step": 1504
},
{
"epoch": 1.838827838827839,
"grad_norm": 3.390364170074463,
"learning_rate": 4.140363845125951e-06,
"loss": 0.7647203803062439,
"step": 1506
},
{
"epoch": 1.8412698412698414,
"grad_norm": 1.699028730392456,
"learning_rate": 4.134893162115832e-06,
"loss": 0.8751652836799622,
"step": 1508
},
{
"epoch": 1.8437118437118438,
"grad_norm": 1.77727472782135,
"learning_rate": 4.129418686519011e-06,
"loss": 1.1023961305618286,
"step": 1510
},
{
"epoch": 1.8461538461538463,
"grad_norm": 5.108009338378906,
"learning_rate": 4.123940443203845e-06,
"loss": 1.0676683187484741,
"step": 1512
},
{
"epoch": 1.8485958485958487,
"grad_norm": 4.390260696411133,
"learning_rate": 4.118458457055811e-06,
"loss": 1.1111550331115723,
"step": 1514
},
{
"epoch": 1.8510378510378511,
"grad_norm": 9.013320922851562,
"learning_rate": 4.112972752977384e-06,
"loss": 0.8402547836303711,
"step": 1516
},
{
"epoch": 1.8534798534798536,
"grad_norm": 2.0569653511047363,
"learning_rate": 4.107483355887933e-06,
"loss": 1.160443902015686,
"step": 1518
},
{
"epoch": 1.855921855921856,
"grad_norm": 2.105151891708374,
"learning_rate": 4.101990290723597e-06,
"loss": 0.9111926555633545,
"step": 1520
},
{
"epoch": 1.8583638583638584,
"grad_norm": 2.9433350563049316,
"learning_rate": 4.096493582437182e-06,
"loss": 0.9556376338005066,
"step": 1522
},
{
"epoch": 1.8608058608058609,
"grad_norm": 3.2840516567230225,
"learning_rate": 4.090993255998042e-06,
"loss": 0.953590989112854,
"step": 1524
},
{
"epoch": 1.8632478632478633,
"grad_norm": 1.5776071548461914,
"learning_rate": 4.0854893363919666e-06,
"loss": 0.9239129424095154,
"step": 1526
},
{
"epoch": 1.8656898656898657,
"grad_norm": 3.6965951919555664,
"learning_rate": 4.079981848621067e-06,
"loss": 0.7926357984542847,
"step": 1528
},
{
"epoch": 1.8681318681318682,
"grad_norm": 1.4808310270309448,
"learning_rate": 4.074470817703664e-06,
"loss": 0.9207745790481567,
"step": 1530
},
{
"epoch": 1.8705738705738706,
"grad_norm": 1.6555721759796143,
"learning_rate": 4.0689562686741735e-06,
"loss": 0.7711512446403503,
"step": 1532
},
{
"epoch": 1.873015873015873,
"grad_norm": 1.7683871984481812,
"learning_rate": 4.063438226582994e-06,
"loss": 0.8535028696060181,
"step": 1534
},
{
"epoch": 1.8754578754578755,
"grad_norm": 4.361166954040527,
"learning_rate": 4.057916716496388e-06,
"loss": 1.119881272315979,
"step": 1536
},
{
"epoch": 1.877899877899878,
"grad_norm": 1.3381246328353882,
"learning_rate": 4.052391763496376e-06,
"loss": 0.9711114168167114,
"step": 1538
},
{
"epoch": 1.8803418803418803,
"grad_norm": 1.1705869436264038,
"learning_rate": 4.046863392680614e-06,
"loss": 1.0517574548721313,
"step": 1540
},
{
"epoch": 1.8827838827838828,
"grad_norm": 2.8563392162323,
"learning_rate": 4.041331629162289e-06,
"loss": 1.060086965560913,
"step": 1542
},
{
"epoch": 1.8852258852258852,
"grad_norm": 1.8419415950775146,
"learning_rate": 4.035796498069995e-06,
"loss": 0.9262551069259644,
"step": 1544
},
{
"epoch": 1.8876678876678876,
"grad_norm": 2.282602310180664,
"learning_rate": 4.030258024547625e-06,
"loss": 1.1138805150985718,
"step": 1546
},
{
"epoch": 1.89010989010989,
"grad_norm": 1.8432765007019043,
"learning_rate": 4.024716233754256e-06,
"loss": 1.1018257141113281,
"step": 1548
},
{
"epoch": 1.8925518925518925,
"grad_norm": 1.7539535760879517,
"learning_rate": 4.019171150864034e-06,
"loss": 0.6556220650672913,
"step": 1550
},
{
"epoch": 1.894993894993895,
"grad_norm": 1.8477240800857544,
"learning_rate": 4.013622801066059e-06,
"loss": 1.1449788808822632,
"step": 1552
},
{
"epoch": 1.8974358974358974,
"grad_norm": 6.399501323699951,
"learning_rate": 4.008071209564272e-06,
"loss": 0.8968675136566162,
"step": 1554
},
{
"epoch": 1.8998778998778998,
"grad_norm": 2.375927448272705,
"learning_rate": 4.00251640157734e-06,
"loss": 1.178645372390747,
"step": 1556
},
{
"epoch": 1.9023199023199022,
"grad_norm": 2.230632781982422,
"learning_rate": 3.996958402338539e-06,
"loss": 0.9704781174659729,
"step": 1558
},
{
"epoch": 1.9047619047619047,
"grad_norm": 1.7731071710586548,
"learning_rate": 3.991397237095644e-06,
"loss": 0.6532866358757019,
"step": 1560
},
{
"epoch": 1.907203907203907,
"grad_norm": 1.6434381008148193,
"learning_rate": 3.985832931110811e-06,
"loss": 1.0863354206085205,
"step": 1562
},
{
"epoch": 1.9096459096459095,
"grad_norm": 1.637665867805481,
"learning_rate": 3.980265509660464e-06,
"loss": 0.777664065361023,
"step": 1564
},
{
"epoch": 1.912087912087912,
"grad_norm": 1.7624237537384033,
"learning_rate": 3.974694998035176e-06,
"loss": 1.0450448989868164,
"step": 1566
},
{
"epoch": 1.9145299145299144,
"grad_norm": 2.4146924018859863,
"learning_rate": 3.969121421539562e-06,
"loss": 0.9237840175628662,
"step": 1568
},
{
"epoch": 1.9169719169719168,
"grad_norm": 1.4809205532073975,
"learning_rate": 3.963544805492158e-06,
"loss": 0.7602840662002563,
"step": 1570
},
{
"epoch": 1.9194139194139193,
"grad_norm": 1.7811788320541382,
"learning_rate": 3.957965175225305e-06,
"loss": 1.1881275177001953,
"step": 1572
},
{
"epoch": 1.9218559218559217,
"grad_norm": 6.949507236480713,
"learning_rate": 3.95238255608504e-06,
"loss": 0.7732429504394531,
"step": 1574
},
{
"epoch": 1.9242979242979243,
"grad_norm": 7.414094924926758,
"learning_rate": 3.946796973430975e-06,
"loss": 1.035898208618164,
"step": 1576
},
{
"epoch": 1.9267399267399268,
"grad_norm": 3.1300573348999023,
"learning_rate": 3.941208452636183e-06,
"loss": 0.7754787802696228,
"step": 1578
},
{
"epoch": 1.9291819291819292,
"grad_norm": 3.253720998764038,
"learning_rate": 3.935617019087088e-06,
"loss": 1.0728245973587036,
"step": 1580
},
{
"epoch": 1.9316239316239316,
"grad_norm": 3.1016621589660645,
"learning_rate": 3.930022698183341e-06,
"loss": 0.32128599286079407,
"step": 1582
},
{
"epoch": 1.934065934065934,
"grad_norm": 4.842744827270508,
"learning_rate": 3.924425515337713e-06,
"loss": 0.5422120094299316,
"step": 1584
},
{
"epoch": 1.9365079365079365,
"grad_norm": 3.4001028537750244,
"learning_rate": 3.918825495975971e-06,
"loss": 0.6693114638328552,
"step": 1586
},
{
"epoch": 1.938949938949939,
"grad_norm": 1.329004168510437,
"learning_rate": 3.913222665536772e-06,
"loss": 1.1855244636535645,
"step": 1588
},
{
"epoch": 1.9413919413919414,
"grad_norm": 1.492633581161499,
"learning_rate": 3.90761704947154e-06,
"loss": 1.1521289348602295,
"step": 1590
},
{
"epoch": 1.9438339438339438,
"grad_norm": 2.899778127670288,
"learning_rate": 3.902008673244353e-06,
"loss": 1.049696445465088,
"step": 1592
},
{
"epoch": 1.9462759462759462,
"grad_norm": 1.1973531246185303,
"learning_rate": 3.896397562331828e-06,
"loss": 0.781209409236908,
"step": 1594
},
{
"epoch": 1.9487179487179487,
"grad_norm": 2.6766295433044434,
"learning_rate": 3.890783742223003e-06,
"loss": 0.6280114650726318,
"step": 1596
},
{
"epoch": 1.9511599511599511,
"grad_norm": 1.4150158166885376,
"learning_rate": 3.885167238419226e-06,
"loss": 0.7452678680419922,
"step": 1598
},
{
"epoch": 1.9536019536019538,
"grad_norm": 1.5957354307174683,
"learning_rate": 3.879548076434033e-06,
"loss": 0.7975012063980103,
"step": 1600
},
{
"epoch": 1.9560439560439562,
"grad_norm": 1.440159559249878,
"learning_rate": 3.873926281793037e-06,
"loss": 1.0404465198516846,
"step": 1602
},
{
"epoch": 1.9584859584859586,
"grad_norm": 2.519695997238159,
"learning_rate": 3.868301880033808e-06,
"loss": 1.2595645189285278,
"step": 1604
},
{
"epoch": 1.960927960927961,
"grad_norm": 3.7468674182891846,
"learning_rate": 3.862674896705762e-06,
"loss": 1.4057308435440063,
"step": 1606
},
{
"epoch": 1.9633699633699635,
"grad_norm": 1.4645074605941772,
"learning_rate": 3.85704535737004e-06,
"loss": 1.1884080171585083,
"step": 1608
},
{
"epoch": 1.965811965811966,
"grad_norm": 1.430010199546814,
"learning_rate": 3.851413287599392e-06,
"loss": 1.058860182762146,
"step": 1610
},
{
"epoch": 1.9682539682539684,
"grad_norm": 4.485745906829834,
"learning_rate": 3.845778712978068e-06,
"loss": 0.9613910913467407,
"step": 1612
},
{
"epoch": 1.9706959706959708,
"grad_norm": 2.261005401611328,
"learning_rate": 3.840141659101692e-06,
"loss": 1.1407309770584106,
"step": 1614
},
{
"epoch": 1.9731379731379732,
"grad_norm": 1.5672752857208252,
"learning_rate": 3.834502151577152e-06,
"loss": 0.93012934923172,
"step": 1616
},
{
"epoch": 1.9755799755799757,
"grad_norm": 1.4915902614593506,
"learning_rate": 3.828860216022482e-06,
"loss": 1.0084210634231567,
"step": 1618
},
{
"epoch": 1.978021978021978,
"grad_norm": 2.3556222915649414,
"learning_rate": 3.8232158780667465e-06,
"loss": 1.1114431619644165,
"step": 1620
},
{
"epoch": 1.9804639804639805,
"grad_norm": 1.559486746788025,
"learning_rate": 3.817569163349921e-06,
"loss": 0.762478232383728,
"step": 1622
},
{
"epoch": 1.982905982905983,
"grad_norm": 2.0981836318969727,
"learning_rate": 3.8119200975227767e-06,
"loss": 0.570152759552002,
"step": 1624
},
{
"epoch": 1.9853479853479854,
"grad_norm": 3.3158349990844727,
"learning_rate": 3.8062687062467704e-06,
"loss": 0.8712815642356873,
"step": 1626
},
{
"epoch": 1.9877899877899878,
"grad_norm": 6.709053039550781,
"learning_rate": 3.8006150151939168e-06,
"loss": 1.0017497539520264,
"step": 1628
},
{
"epoch": 1.9902319902319903,
"grad_norm": 1.431030511856079,
"learning_rate": 3.7949590500466817e-06,
"loss": 1.154350757598877,
"step": 1630
},
{
"epoch": 1.9926739926739927,
"grad_norm": 1.6267858743667603,
"learning_rate": 3.7893008364978575e-06,
"loss": 1.0441209077835083,
"step": 1632
},
{
"epoch": 1.9951159951159951,
"grad_norm": 1.9177896976470947,
"learning_rate": 3.783640400250453e-06,
"loss": 0.8070008754730225,
"step": 1634
},
{
"epoch": 1.9975579975579976,
"grad_norm": 1.5680376291275024,
"learning_rate": 3.777977767017572e-06,
"loss": 1.167231798171997,
"step": 1636
},
{
"epoch": 2.0,
"grad_norm": 1.836150884628296,
"learning_rate": 3.7723129625222997e-06,
"loss": 1.096339464187622,
"step": 1638
},
{
"epoch": 2.0024420024420024,
"grad_norm": 6.821539878845215,
"learning_rate": 3.7666460124975845e-06,
"loss": 0.9670911431312561,
"step": 1640
},
{
"epoch": 2.004884004884005,
"grad_norm": 2.0258708000183105,
"learning_rate": 3.76097694268612e-06,
"loss": 0.6899262070655823,
"step": 1642
},
{
"epoch": 2.0073260073260073,
"grad_norm": 1.1317782402038574,
"learning_rate": 3.75530577884023e-06,
"loss": 0.9158855080604553,
"step": 1644
},
{
"epoch": 2.0097680097680097,
"grad_norm": 2.7936630249023438,
"learning_rate": 3.7496325467217498e-06,
"loss": 0.6676980257034302,
"step": 1646
},
{
"epoch": 2.012210012210012,
"grad_norm": 1.8216701745986938,
"learning_rate": 3.7439572721019104e-06,
"loss": 0.7151694297790527,
"step": 1648
},
{
"epoch": 2.0146520146520146,
"grad_norm": 26.17021369934082,
"learning_rate": 3.738279980761222e-06,
"loss": 0.6760666370391846,
"step": 1650
},
{
"epoch": 2.017094017094017,
"grad_norm": 5.065937042236328,
"learning_rate": 3.732600698489355e-06,
"loss": 0.7696459293365479,
"step": 1652
},
{
"epoch": 2.0195360195360195,
"grad_norm": 2.4799294471740723,
"learning_rate": 3.726919451085024e-06,
"loss": 1.1132228374481201,
"step": 1654
},
{
"epoch": 2.021978021978022,
"grad_norm": 1.5045807361602783,
"learning_rate": 3.721236264355871e-06,
"loss": 0.5454859733581543,
"step": 1656
},
{
"epoch": 2.0244200244200243,
"grad_norm": 2.9693870544433594,
"learning_rate": 3.715551164118346e-06,
"loss": 1.0050889253616333,
"step": 1658
},
{
"epoch": 2.0268620268620268,
"grad_norm": 4.3840484619140625,
"learning_rate": 3.709864176197592e-06,
"loss": 0.969306230545044,
"step": 1660
},
{
"epoch": 2.029304029304029,
"grad_norm": 2.8534724712371826,
"learning_rate": 3.7041753264273282e-06,
"loss": 1.00910222530365,
"step": 1662
},
{
"epoch": 2.0317460317460316,
"grad_norm": 3.8745787143707275,
"learning_rate": 3.6984846406497306e-06,
"loss": 0.4947141706943512,
"step": 1664
},
{
"epoch": 2.034188034188034,
"grad_norm": 10.617118835449219,
"learning_rate": 3.692792144715316e-06,
"loss": 0.6221736073493958,
"step": 1666
},
{
"epoch": 2.0366300366300365,
"grad_norm": 1.7398251295089722,
"learning_rate": 3.687097864482821e-06,
"loss": 0.9439553022384644,
"step": 1668
},
{
"epoch": 2.039072039072039,
"grad_norm": 2.146132469177246,
"learning_rate": 3.681401825819092e-06,
"loss": 0.6233932971954346,
"step": 1670
},
{
"epoch": 2.0415140415140414,
"grad_norm": 2.6649765968322754,
"learning_rate": 3.6757040545989615e-06,
"loss": 0.9565668106079102,
"step": 1672
},
{
"epoch": 2.043956043956044,
"grad_norm": 2.2828168869018555,
"learning_rate": 3.6700045767051313e-06,
"loss": 0.9368692636489868,
"step": 1674
},
{
"epoch": 2.0463980463980462,
"grad_norm": 1.828999638557434,
"learning_rate": 3.664303418028056e-06,
"loss": 0.6456412076950073,
"step": 1676
},
{
"epoch": 2.0488400488400487,
"grad_norm": 5.089777946472168,
"learning_rate": 3.6586006044658275e-06,
"loss": 1.019284725189209,
"step": 1678
},
{
"epoch": 2.051282051282051,
"grad_norm": 3.3917853832244873,
"learning_rate": 3.6528961619240525e-06,
"loss": 1.0033758878707886,
"step": 1680
},
{
"epoch": 2.0537240537240535,
"grad_norm": 2.755739450454712,
"learning_rate": 3.647190116315739e-06,
"loss": 0.6449636816978455,
"step": 1682
},
{
"epoch": 2.056166056166056,
"grad_norm": 1.7534053325653076,
"learning_rate": 3.6414824935611777e-06,
"loss": 0.7625139355659485,
"step": 1684
},
{
"epoch": 2.0586080586080584,
"grad_norm": 26.373376846313477,
"learning_rate": 3.6357733195878213e-06,
"loss": 0.8886415958404541,
"step": 1686
},
{
"epoch": 2.061050061050061,
"grad_norm": 6.317865371704102,
"learning_rate": 3.6300626203301706e-06,
"loss": 0.8263817429542542,
"step": 1688
},
{
"epoch": 2.0634920634920633,
"grad_norm": 4.633421421051025,
"learning_rate": 3.6243504217296562e-06,
"loss": 0.7733551859855652,
"step": 1690
},
{
"epoch": 2.065934065934066,
"grad_norm": 0.8595446348190308,
"learning_rate": 3.6186367497345175e-06,
"loss": 0.6480957865715027,
"step": 1692
},
{
"epoch": 2.0683760683760686,
"grad_norm": 1.6937971115112305,
"learning_rate": 3.6129216302996884e-06,
"loss": 0.746891975402832,
"step": 1694
},
{
"epoch": 2.070818070818071,
"grad_norm": 1.8328354358673096,
"learning_rate": 3.6072050893866774e-06,
"loss": 0.960361659526825,
"step": 1696
},
{
"epoch": 2.0732600732600734,
"grad_norm": 1.8361425399780273,
"learning_rate": 3.6014871529634497e-06,
"loss": 0.9544206261634827,
"step": 1698
},
{
"epoch": 2.075702075702076,
"grad_norm": 4.454921722412109,
"learning_rate": 3.5957678470043117e-06,
"loss": 0.9407209753990173,
"step": 1700
},
{
"epoch": 2.0781440781440783,
"grad_norm": 3.5452890396118164,
"learning_rate": 3.5900471974897883e-06,
"loss": 0.5001405477523804,
"step": 1702
},
{
"epoch": 2.0805860805860807,
"grad_norm": 1.039163589477539,
"learning_rate": 3.5843252304065106e-06,
"loss": 0.5040726661682129,
"step": 1704
},
{
"epoch": 2.083028083028083,
"grad_norm": 6.359805583953857,
"learning_rate": 3.5786019717470907e-06,
"loss": 0.4261728525161743,
"step": 1706
},
{
"epoch": 2.0854700854700856,
"grad_norm": 4.610110759735107,
"learning_rate": 3.5728774475100127e-06,
"loss": 0.9747272729873657,
"step": 1708
},
{
"epoch": 2.087912087912088,
"grad_norm": 2.2436630725860596,
"learning_rate": 3.5671516836995057e-06,
"loss": 0.6898809671401978,
"step": 1710
},
{
"epoch": 2.0903540903540905,
"grad_norm": 12.148319244384766,
"learning_rate": 3.561424706325434e-06,
"loss": 0.6288307905197144,
"step": 1712
},
{
"epoch": 2.092796092796093,
"grad_norm": 1.7018226385116577,
"learning_rate": 3.555696541403169e-06,
"loss": 0.620037317276001,
"step": 1714
},
{
"epoch": 2.0952380952380953,
"grad_norm": 2.4741909503936768,
"learning_rate": 3.549967214953482e-06,
"loss": 0.6937116980552673,
"step": 1716
},
{
"epoch": 2.0976800976800978,
"grad_norm": 2.7370715141296387,
"learning_rate": 3.5442367530024167e-06,
"loss": 0.9124626517295837,
"step": 1718
},
{
"epoch": 2.1001221001221,
"grad_norm": 2.0893752574920654,
"learning_rate": 3.5385051815811774e-06,
"loss": 0.9028443694114685,
"step": 1720
},
{
"epoch": 2.1025641025641026,
"grad_norm": 4.761720180511475,
"learning_rate": 3.5327725267260076e-06,
"loss": 1.0133492946624756,
"step": 1722
},
{
"epoch": 2.105006105006105,
"grad_norm": 1.624951720237732,
"learning_rate": 3.5270388144780724e-06,
"loss": 0.6351829767227173,
"step": 1724
},
{
"epoch": 2.1074481074481075,
"grad_norm": 16.627628326416016,
"learning_rate": 3.5213040708833395e-06,
"loss": 0.5484714508056641,
"step": 1726
},
{
"epoch": 2.10989010989011,
"grad_norm": 1.2873952388763428,
"learning_rate": 3.5155683219924626e-06,
"loss": 0.5492695569992065,
"step": 1728
},
{
"epoch": 2.1123321123321124,
"grad_norm": 2.2124252319335938,
"learning_rate": 3.509831593860662e-06,
"loss": 0.841966986656189,
"step": 1730
},
{
"epoch": 2.114774114774115,
"grad_norm": 2.9345152378082275,
"learning_rate": 3.5040939125476042e-06,
"loss": 0.5698384642601013,
"step": 1732
},
{
"epoch": 2.1172161172161172,
"grad_norm": 0.9468244910240173,
"learning_rate": 3.4983553041172902e-06,
"loss": 0.2970229685306549,
"step": 1734
},
{
"epoch": 2.1196581196581197,
"grad_norm": 4.071929931640625,
"learning_rate": 3.4926157946379273e-06,
"loss": 0.9440501928329468,
"step": 1736
},
{
"epoch": 2.122100122100122,
"grad_norm": 2.8341457843780518,
"learning_rate": 3.4868754101818194e-06,
"loss": 0.8834357261657715,
"step": 1738
},
{
"epoch": 2.1245421245421245,
"grad_norm": 1.979362964630127,
"learning_rate": 3.481134176825245e-06,
"loss": 1.0380572080612183,
"step": 1740
},
{
"epoch": 2.126984126984127,
"grad_norm": 7.127715587615967,
"learning_rate": 3.4753921206483347e-06,
"loss": 0.8644207715988159,
"step": 1742
},
{
"epoch": 2.1294261294261294,
"grad_norm": 2.524230480194092,
"learning_rate": 3.4696492677349627e-06,
"loss": 0.6818966269493103,
"step": 1744
},
{
"epoch": 2.131868131868132,
"grad_norm": 1.9305305480957031,
"learning_rate": 3.463905644172619e-06,
"loss": 0.9881414175033569,
"step": 1746
},
{
"epoch": 2.1343101343101343,
"grad_norm": 3.519512891769409,
"learning_rate": 3.4581612760522945e-06,
"loss": 0.4184856712818146,
"step": 1748
},
{
"epoch": 2.1367521367521367,
"grad_norm": 1.6307930946350098,
"learning_rate": 3.4524161894683634e-06,
"loss": 1.1337333917617798,
"step": 1750
},
{
"epoch": 2.139194139194139,
"grad_norm": 2.0223984718322754,
"learning_rate": 3.4466704105184634e-06,
"loss": 0.8873319029808044,
"step": 1752
},
{
"epoch": 2.1416361416361416,
"grad_norm": 3.159184694290161,
"learning_rate": 3.4409239653033754e-06,
"loss": 0.9011868834495544,
"step": 1754
},
{
"epoch": 2.144078144078144,
"grad_norm": 1.4487037658691406,
"learning_rate": 3.435176879926911e-06,
"loss": 1.0243741273880005,
"step": 1756
},
{
"epoch": 2.1465201465201464,
"grad_norm": 3.106032371520996,
"learning_rate": 3.4294291804957846e-06,
"loss": 0.4617737829685211,
"step": 1758
},
{
"epoch": 2.148962148962149,
"grad_norm": 1.5856150388717651,
"learning_rate": 3.4236808931195045e-06,
"loss": 0.9131142497062683,
"step": 1760
},
{
"epoch": 2.1514041514041513,
"grad_norm": 0.6227497458457947,
"learning_rate": 3.417932043910248e-06,
"loss": 0.6714634299278259,
"step": 1762
},
{
"epoch": 2.1538461538461537,
"grad_norm": 3.6048402786254883,
"learning_rate": 3.412182658982743e-06,
"loss": 0.6586474776268005,
"step": 1764
},
{
"epoch": 2.156288156288156,
"grad_norm": 1.531855583190918,
"learning_rate": 3.4064327644541536e-06,
"loss": 0.6113312244415283,
"step": 1766
},
{
"epoch": 2.1587301587301586,
"grad_norm": 25.597637176513672,
"learning_rate": 3.400682386443958e-06,
"loss": 0.7350863218307495,
"step": 1768
},
{
"epoch": 2.161172161172161,
"grad_norm": 5.0424346923828125,
"learning_rate": 3.39493155107383e-06,
"loss": 0.5671433806419373,
"step": 1770
},
{
"epoch": 2.1636141636141635,
"grad_norm": 3.7835681438446045,
"learning_rate": 3.389180284467522e-06,
"loss": 0.7587950825691223,
"step": 1772
},
{
"epoch": 2.166056166056166,
"grad_norm": 11.762646675109863,
"learning_rate": 3.3834286127507445e-06,
"loss": 0.8601915240287781,
"step": 1774
},
{
"epoch": 2.1684981684981683,
"grad_norm": 1.298113226890564,
"learning_rate": 3.3776765620510475e-06,
"loss": 0.9438109397888184,
"step": 1776
},
{
"epoch": 2.1709401709401708,
"grad_norm": 5.715275764465332,
"learning_rate": 3.3719241584977056e-06,
"loss": 0.6178626418113708,
"step": 1778
},
{
"epoch": 2.173382173382173,
"grad_norm": 2.35664963722229,
"learning_rate": 3.366171428221592e-06,
"loss": 0.9353488087654114,
"step": 1780
},
{
"epoch": 2.1758241758241756,
"grad_norm": 4.974782466888428,
"learning_rate": 3.360418397355067e-06,
"loss": 0.7427996397018433,
"step": 1782
},
{
"epoch": 2.178266178266178,
"grad_norm": 1.669165849685669,
"learning_rate": 3.3546650920318567e-06,
"loss": 0.9448103308677673,
"step": 1784
},
{
"epoch": 2.1807081807081805,
"grad_norm": 5.4477691650390625,
"learning_rate": 3.3489115383869322e-06,
"loss": 0.9031893014907837,
"step": 1786
},
{
"epoch": 2.183150183150183,
"grad_norm": 3.41313099861145,
"learning_rate": 3.3431577625563934e-06,
"loss": 0.9105832576751709,
"step": 1788
},
{
"epoch": 2.185592185592186,
"grad_norm": 2.2663333415985107,
"learning_rate": 3.337403790677349e-06,
"loss": 0.8743823766708374,
"step": 1790
},
{
"epoch": 2.1880341880341883,
"grad_norm": 3.956235408782959,
"learning_rate": 3.3316496488878e-06,
"loss": 0.9286958575248718,
"step": 1792
},
{
"epoch": 2.1904761904761907,
"grad_norm": 1.6148682832717896,
"learning_rate": 3.3258953633265155e-06,
"loss": 0.5856344699859619,
"step": 1794
},
{
"epoch": 2.192918192918193,
"grad_norm": 1.766203761100769,
"learning_rate": 3.320140960132923e-06,
"loss": 0.6644143462181091,
"step": 1796
},
{
"epoch": 2.1953601953601956,
"grad_norm": 1.4154376983642578,
"learning_rate": 3.3143864654469796e-06,
"loss": 0.7322801947593689,
"step": 1798
},
{
"epoch": 2.197802197802198,
"grad_norm": 1.9144586324691772,
"learning_rate": 3.3086319054090614e-06,
"loss": 0.9201781749725342,
"step": 1800
},
{
"epoch": 2.2002442002442004,
"grad_norm": 4.5645365715026855,
"learning_rate": 3.302877306159839e-06,
"loss": 0.9563385844230652,
"step": 1802
},
{
"epoch": 2.202686202686203,
"grad_norm": 5.516610145568848,
"learning_rate": 3.2971226938401616e-06,
"loss": 0.6716048717498779,
"step": 1804
},
{
"epoch": 2.2051282051282053,
"grad_norm": 2.037088632583618,
"learning_rate": 3.2913680945909394e-06,
"loss": 0.9080408215522766,
"step": 1806
},
{
"epoch": 2.2075702075702077,
"grad_norm": 16.09855842590332,
"learning_rate": 3.2856135345530203e-06,
"loss": 1.0019384622573853,
"step": 1808
},
{
"epoch": 2.21001221001221,
"grad_norm": 1.9463958740234375,
"learning_rate": 3.279859039867078e-06,
"loss": 0.6507654786109924,
"step": 1810
},
{
"epoch": 2.2124542124542126,
"grad_norm": 1.6095830202102661,
"learning_rate": 3.2741046366734852e-06,
"loss": 0.7600730061531067,
"step": 1812
},
{
"epoch": 2.214896214896215,
"grad_norm": 1.3766435384750366,
"learning_rate": 3.2683503511122013e-06,
"loss": 0.41193750500679016,
"step": 1814
},
{
"epoch": 2.2173382173382175,
"grad_norm": 2.845604419708252,
"learning_rate": 3.262596209322651e-06,
"loss": 1.0406476259231567,
"step": 1816
},
{
"epoch": 2.21978021978022,
"grad_norm": 1.6436043977737427,
"learning_rate": 3.2568422374436078e-06,
"loss": 0.9313133358955383,
"step": 1818
},
{
"epoch": 2.2222222222222223,
"grad_norm": 1.925812005996704,
"learning_rate": 3.2510884616130685e-06,
"loss": 0.9973028898239136,
"step": 1820
},
{
"epoch": 2.2246642246642248,
"grad_norm": 3.855332612991333,
"learning_rate": 3.2453349079681436e-06,
"loss": 0.4581510126590729,
"step": 1822
},
{
"epoch": 2.227106227106227,
"grad_norm": 4.27201509475708,
"learning_rate": 3.239581602644933e-06,
"loss": 0.761246383190155,
"step": 1824
},
{
"epoch": 2.2295482295482296,
"grad_norm": 10.0543794631958,
"learning_rate": 3.2338285717784087e-06,
"loss": 0.5892999172210693,
"step": 1826
},
{
"epoch": 2.231990231990232,
"grad_norm": 4.678505897521973,
"learning_rate": 3.228075841502295e-06,
"loss": 0.5877086520195007,
"step": 1828
},
{
"epoch": 2.2344322344322345,
"grad_norm": 3.173227310180664,
"learning_rate": 3.2223234379489524e-06,
"loss": 0.6748300790786743,
"step": 1830
},
{
"epoch": 2.236874236874237,
"grad_norm": 1.8820854425430298,
"learning_rate": 3.2165713872492563e-06,
"loss": 0.9059861898422241,
"step": 1832
},
{
"epoch": 2.2393162393162394,
"grad_norm": 1.2828683853149414,
"learning_rate": 3.2108197155324782e-06,
"loss": 0.9060832262039185,
"step": 1834
},
{
"epoch": 2.241758241758242,
"grad_norm": 2.243968963623047,
"learning_rate": 3.2050684489261707e-06,
"loss": 0.8909225463867188,
"step": 1836
},
{
"epoch": 2.244200244200244,
"grad_norm": 3.2561757564544678,
"learning_rate": 3.199317613556042e-06,
"loss": 0.9934486746788025,
"step": 1838
},
{
"epoch": 2.2466422466422467,
"grad_norm": 3.1067919731140137,
"learning_rate": 3.193567235545847e-06,
"loss": 0.44754478335380554,
"step": 1840
},
{
"epoch": 2.249084249084249,
"grad_norm": 1.4694087505340576,
"learning_rate": 3.1878173410172574e-06,
"loss": 0.9427831172943115,
"step": 1842
},
{
"epoch": 2.2515262515262515,
"grad_norm": 1.6658433675765991,
"learning_rate": 3.182067956089754e-06,
"loss": 1.1688846349716187,
"step": 1844
},
{
"epoch": 2.253968253968254,
"grad_norm": 17.416038513183594,
"learning_rate": 3.1763191068804963e-06,
"loss": 0.6192197799682617,
"step": 1846
},
{
"epoch": 2.2564102564102564,
"grad_norm": 1.2540676593780518,
"learning_rate": 3.170570819504217e-06,
"loss": 0.6497691869735718,
"step": 1848
},
{
"epoch": 2.258852258852259,
"grad_norm": 2.8082709312438965,
"learning_rate": 3.1648231200730907e-06,
"loss": 0.676390528678894,
"step": 1850
},
{
"epoch": 2.2612942612942613,
"grad_norm": 12.676959037780762,
"learning_rate": 3.1590760346966254e-06,
"loss": 0.967694878578186,
"step": 1852
},
{
"epoch": 2.2637362637362637,
"grad_norm": 22.334165573120117,
"learning_rate": 3.1533295894815374e-06,
"loss": 1.0598804950714111,
"step": 1854
},
{
"epoch": 2.266178266178266,
"grad_norm": 2.7481689453125,
"learning_rate": 3.1475838105316378e-06,
"loss": 0.897810697555542,
"step": 1856
},
{
"epoch": 2.2686202686202686,
"grad_norm": 5.33841609954834,
"learning_rate": 3.1418387239477055e-06,
"loss": 0.8070839643478394,
"step": 1858
},
{
"epoch": 2.271062271062271,
"grad_norm": 3.1426467895507812,
"learning_rate": 3.1360943558273812e-06,
"loss": 1.0592520236968994,
"step": 1860
},
{
"epoch": 2.2735042735042734,
"grad_norm": 3.4249966144561768,
"learning_rate": 3.130350732265038e-06,
"loss": 0.6386770009994507,
"step": 1862
},
{
"epoch": 2.275946275946276,
"grad_norm": 2.3887171745300293,
"learning_rate": 3.1246078793516656e-06,
"loss": 0.6159527897834778,
"step": 1864
},
{
"epoch": 2.2783882783882783,
"grad_norm": 2.1484906673431396,
"learning_rate": 3.1188658231747566e-06,
"loss": 0.8985686302185059,
"step": 1866
},
{
"epoch": 2.2808302808302807,
"grad_norm": 2.638803005218506,
"learning_rate": 3.11312458981818e-06,
"loss": 0.9426345229148865,
"step": 1868
},
{
"epoch": 2.283272283272283,
"grad_norm": 1.8906961679458618,
"learning_rate": 3.1073842053620734e-06,
"loss": 0.9898474812507629,
"step": 1870
},
{
"epoch": 2.2857142857142856,
"grad_norm": 1.6906073093414307,
"learning_rate": 3.1016446958827097e-06,
"loss": 0.9557514786720276,
"step": 1872
},
{
"epoch": 2.288156288156288,
"grad_norm": 4.265973091125488,
"learning_rate": 3.0959060874523965e-06,
"loss": 0.16929063200950623,
"step": 1874
},
{
"epoch": 2.2905982905982905,
"grad_norm": 1.2474678754806519,
"learning_rate": 3.0901684061393386e-06,
"loss": 0.96248859167099,
"step": 1876
},
{
"epoch": 2.293040293040293,
"grad_norm": 2.5638835430145264,
"learning_rate": 3.0844316780075377e-06,
"loss": 0.6548967957496643,
"step": 1878
},
{
"epoch": 2.2954822954822953,
"grad_norm": 4.610739707946777,
"learning_rate": 3.078695929116661e-06,
"loss": 0.7896809577941895,
"step": 1880
},
{
"epoch": 2.2979242979242978,
"grad_norm": 2.039637804031372,
"learning_rate": 3.072961185521929e-06,
"loss": 1.022554874420166,
"step": 1882
},
{
"epoch": 2.3003663003663,
"grad_norm": 1.1219255924224854,
"learning_rate": 3.067227473273993e-06,
"loss": 0.6635209321975708,
"step": 1884
},
{
"epoch": 2.3028083028083026,
"grad_norm": 1.9538935422897339,
"learning_rate": 3.0614948184188238e-06,
"loss": 1.2081166505813599,
"step": 1886
},
{
"epoch": 2.305250305250305,
"grad_norm": 4.5801849365234375,
"learning_rate": 3.0557632469975845e-06,
"loss": 0.9475079774856567,
"step": 1888
},
{
"epoch": 2.3076923076923075,
"grad_norm": 2.044386148452759,
"learning_rate": 3.05003278504652e-06,
"loss": 1.0499695539474487,
"step": 1890
},
{
"epoch": 2.31013431013431,
"grad_norm": 2.1628005504608154,
"learning_rate": 3.044303458596832e-06,
"loss": 0.95216965675354,
"step": 1892
},
{
"epoch": 2.3125763125763124,
"grad_norm": 5.278742790222168,
"learning_rate": 3.0385752936745665e-06,
"loss": 1.0832526683807373,
"step": 1894
},
{
"epoch": 2.315018315018315,
"grad_norm": 1.9575459957122803,
"learning_rate": 3.032848316300494e-06,
"loss": 1.029971718788147,
"step": 1896
},
{
"epoch": 2.317460317460317,
"grad_norm": 2.02197527885437,
"learning_rate": 3.027122552489988e-06,
"loss": 0.9606547951698303,
"step": 1898
},
{
"epoch": 2.3199023199023197,
"grad_norm": 1.8357181549072266,
"learning_rate": 3.02139802825291e-06,
"loss": 0.9799268245697021,
"step": 1900
},
{
"epoch": 2.3223443223443225,
"grad_norm": 3.0944628715515137,
"learning_rate": 3.01567476959349e-06,
"loss": 0.9936726093292236,
"step": 1902
},
{
"epoch": 2.324786324786325,
"grad_norm": 4.7253265380859375,
"learning_rate": 3.009952802510212e-06,
"loss": 0.510124146938324,
"step": 1904
},
{
"epoch": 2.3272283272283274,
"grad_norm": 3.323378324508667,
"learning_rate": 3.004232152995688e-06,
"loss": 0.4245060682296753,
"step": 1906
},
{
"epoch": 2.32967032967033,
"grad_norm": 1.937008023262024,
"learning_rate": 2.99851284703655e-06,
"loss": 0.6247161626815796,
"step": 1908
},
{
"epoch": 2.3321123321123323,
"grad_norm": 1.282409429550171,
"learning_rate": 2.992794910613323e-06,
"loss": 0.4279539883136749,
"step": 1910
},
{
"epoch": 2.3345543345543347,
"grad_norm": 15.080961227416992,
"learning_rate": 2.987078369700312e-06,
"loss": 0.7808930277824402,
"step": 1912
},
{
"epoch": 2.336996336996337,
"grad_norm": 1.419805884361267,
"learning_rate": 2.981363250265483e-06,
"loss": 0.8222759962081909,
"step": 1914
},
{
"epoch": 2.3394383394383396,
"grad_norm": 2.3081250190734863,
"learning_rate": 2.9756495782703445e-06,
"loss": 0.9174440503120422,
"step": 1916
},
{
"epoch": 2.341880341880342,
"grad_norm": 1.4504395723342896,
"learning_rate": 2.9699373796698293e-06,
"loss": 0.6940646171569824,
"step": 1918
},
{
"epoch": 2.3443223443223444,
"grad_norm": 9.539470672607422,
"learning_rate": 2.9642266804121804e-06,
"loss": 0.27768972516059875,
"step": 1920
},
{
"epoch": 2.346764346764347,
"grad_norm": 7.798355579376221,
"learning_rate": 2.958517506438823e-06,
"loss": 0.7792479991912842,
"step": 1922
},
{
"epoch": 2.3492063492063493,
"grad_norm": 1.2411043643951416,
"learning_rate": 2.9528098836842615e-06,
"loss": 0.5147164463996887,
"step": 1924
},
{
"epoch": 2.3516483516483517,
"grad_norm": 1.255823016166687,
"learning_rate": 2.947103838075948e-06,
"loss": 0.9378372430801392,
"step": 1926
},
{
"epoch": 2.354090354090354,
"grad_norm": 2.5060043334960938,
"learning_rate": 2.9413993955341732e-06,
"loss": 0.4387669563293457,
"step": 1928
},
{
"epoch": 2.3565323565323566,
"grad_norm": 1.6388636827468872,
"learning_rate": 2.935696581971944e-06,
"loss": 0.47966766357421875,
"step": 1930
},
{
"epoch": 2.358974358974359,
"grad_norm": 2.1432793140411377,
"learning_rate": 2.929995423294869e-06,
"loss": 0.6485490798950195,
"step": 1932
},
{
"epoch": 2.3614163614163615,
"grad_norm": 1.9438990354537964,
"learning_rate": 2.9242959454010385e-06,
"loss": 0.9063615798950195,
"step": 1934
},
{
"epoch": 2.363858363858364,
"grad_norm": 1.496005654335022,
"learning_rate": 2.918598174180907e-06,
"loss": 0.6603356599807739,
"step": 1936
},
{
"epoch": 2.3663003663003663,
"grad_norm": 1.7516268491744995,
"learning_rate": 2.9129021355171796e-06,
"loss": 0.9123400449752808,
"step": 1938
},
{
"epoch": 2.3687423687423688,
"grad_norm": 23.957849502563477,
"learning_rate": 2.907207855284685e-06,
"loss": 0.7359025478363037,
"step": 1940
},
{
"epoch": 2.371184371184371,
"grad_norm": 0.49576354026794434,
"learning_rate": 2.9015153593502697e-06,
"loss": 0.68361896276474,
"step": 1942
},
{
"epoch": 2.3736263736263736,
"grad_norm": 3.961500883102417,
"learning_rate": 2.8958246735726717e-06,
"loss": 0.8321384787559509,
"step": 1944
},
{
"epoch": 2.376068376068376,
"grad_norm": 3.9573895931243896,
"learning_rate": 2.890135823802409e-06,
"loss": 0.20095598697662354,
"step": 1946
},
{
"epoch": 2.3785103785103785,
"grad_norm": 1.3203811645507812,
"learning_rate": 2.8844488358816547e-06,
"loss": 0.46073901653289795,
"step": 1948
},
{
"epoch": 2.380952380952381,
"grad_norm": 4.666232585906982,
"learning_rate": 2.8787637356441306e-06,
"loss": 0.45975521206855774,
"step": 1950
},
{
"epoch": 2.3833943833943834,
"grad_norm": 2.1029303073883057,
"learning_rate": 2.8730805489149762e-06,
"loss": 0.4719840884208679,
"step": 1952
},
{
"epoch": 2.385836385836386,
"grad_norm": 14.054176330566406,
"learning_rate": 2.867399301510646e-06,
"loss": 0.8675229549407959,
"step": 1954
},
{
"epoch": 2.3882783882783882,
"grad_norm": 2.523514986038208,
"learning_rate": 2.861720019238778e-06,
"loss": 0.9857223629951477,
"step": 1956
},
{
"epoch": 2.3907203907203907,
"grad_norm": 5.2092132568359375,
"learning_rate": 2.8560427278980903e-06,
"loss": 0.6426673531532288,
"step": 1958
},
{
"epoch": 2.393162393162393,
"grad_norm": 1.7665150165557861,
"learning_rate": 2.850367453278251e-06,
"loss": 1.0697640180587769,
"step": 1960
},
{
"epoch": 2.3956043956043955,
"grad_norm": 2.8153493404388428,
"learning_rate": 2.8446942211597713e-06,
"loss": 0.6665030121803284,
"step": 1962
},
{
"epoch": 2.398046398046398,
"grad_norm": 1.6511778831481934,
"learning_rate": 2.83902305731388e-06,
"loss": 0.9055560231208801,
"step": 1964
},
{
"epoch": 2.4004884004884004,
"grad_norm": 2.6607825756073,
"learning_rate": 2.833353987502415e-06,
"loss": 0.5983095765113831,
"step": 1966
},
{
"epoch": 2.402930402930403,
"grad_norm": 2.3817784786224365,
"learning_rate": 2.8276870374777003e-06,
"loss": 0.602225661277771,
"step": 1968
},
{
"epoch": 2.4053724053724053,
"grad_norm": 8.560592651367188,
"learning_rate": 2.822022232982428e-06,
"loss": 0.4158218204975128,
"step": 1970
},
{
"epoch": 2.4078144078144077,
"grad_norm": 9.470584869384766,
"learning_rate": 2.816359599749548e-06,
"loss": 0.20221352577209473,
"step": 1972
},
{
"epoch": 2.41025641025641,
"grad_norm": 0.9254062175750732,
"learning_rate": 2.810699163502143e-06,
"loss": 0.18855728209018707,
"step": 1974
},
{
"epoch": 2.4126984126984126,
"grad_norm": 3.3264057636260986,
"learning_rate": 2.805040949953319e-06,
"loss": 0.40398913621902466,
"step": 1976
},
{
"epoch": 2.415140415140415,
"grad_norm": 4.203930854797363,
"learning_rate": 2.7993849848060827e-06,
"loss": 0.9113072752952576,
"step": 1978
},
{
"epoch": 2.4175824175824174,
"grad_norm": 1.2429267168045044,
"learning_rate": 2.7937312937532303e-06,
"loss": 0.46490153670310974,
"step": 1980
},
{
"epoch": 2.42002442002442,
"grad_norm": 5.597394943237305,
"learning_rate": 2.7880799024772236e-06,
"loss": 0.9845388531684875,
"step": 1982
},
{
"epoch": 2.4224664224664223,
"grad_norm": 0.8237308263778687,
"learning_rate": 2.7824308366500807e-06,
"loss": 0.49944135546684265,
"step": 1984
},
{
"epoch": 2.4249084249084247,
"grad_norm": 1.4086517095565796,
"learning_rate": 2.776784121933254e-06,
"loss": 1.1026017665863037,
"step": 1986
},
{
"epoch": 2.427350427350427,
"grad_norm": 2.9017109870910645,
"learning_rate": 2.7711397839775176e-06,
"loss": 0.977454662322998,
"step": 1988
},
{
"epoch": 2.42979242979243,
"grad_norm": 1.8910452127456665,
"learning_rate": 2.765497848422848e-06,
"loss": 1.0892168283462524,
"step": 1990
},
{
"epoch": 2.4322344322344325,
"grad_norm": 5.9468464851379395,
"learning_rate": 2.759858340898308e-06,
"loss": 0.9162377715110779,
"step": 1992
},
{
"epoch": 2.434676434676435,
"grad_norm": 2.0583372116088867,
"learning_rate": 2.7542212870219323e-06,
"loss": 0.9635336995124817,
"step": 1994
},
{
"epoch": 2.4371184371184373,
"grad_norm": 3.0031464099884033,
"learning_rate": 2.7485867124006084e-06,
"loss": 0.8859571218490601,
"step": 1996
},
{
"epoch": 2.4395604395604398,
"grad_norm": 1.2768526077270508,
"learning_rate": 2.742954642629961e-06,
"loss": 1.0290803909301758,
"step": 1998
},
{
"epoch": 2.442002442002442,
"grad_norm": 2.5159692764282227,
"learning_rate": 2.7373251032942382e-06,
"loss": 0.6687625646591187,
"step": 2000
},
{
"epoch": 2.4444444444444446,
"grad_norm": 1.5561347007751465,
"learning_rate": 2.7316981199661915e-06,
"loss": 0.9961126446723938,
"step": 2002
},
{
"epoch": 2.446886446886447,
"grad_norm": 1.4128899574279785,
"learning_rate": 2.726073718206963e-06,
"loss": 0.6645013093948364,
"step": 2004
},
{
"epoch": 2.4493284493284495,
"grad_norm": 2.9358339309692383,
"learning_rate": 2.7204519235659677e-06,
"loss": 0.717438280582428,
"step": 2006
},
{
"epoch": 2.451770451770452,
"grad_norm": 2.506085157394409,
"learning_rate": 2.7148327615807745e-06,
"loss": 1.0376380681991577,
"step": 2008
},
{
"epoch": 2.4542124542124544,
"grad_norm": 2.029444694519043,
"learning_rate": 2.7092162577769975e-06,
"loss": 0.5659610033035278,
"step": 2010
},
{
"epoch": 2.456654456654457,
"grad_norm": 1.5038423538208008,
"learning_rate": 2.7036024376681736e-06,
"loss": 0.5658340454101562,
"step": 2012
},
{
"epoch": 2.4590964590964592,
"grad_norm": 2.677021026611328,
"learning_rate": 2.697991326755648e-06,
"loss": 1.0553836822509766,
"step": 2014
},
{
"epoch": 2.4615384615384617,
"grad_norm": 1.5988397598266602,
"learning_rate": 2.692382950528461e-06,
"loss": 0.9713739156723022,
"step": 2016
},
{
"epoch": 2.463980463980464,
"grad_norm": 2.0507140159606934,
"learning_rate": 2.6867773344632283e-06,
"loss": 0.9353849291801453,
"step": 2018
},
{
"epoch": 2.4664224664224665,
"grad_norm": 4.076539039611816,
"learning_rate": 2.6811745040240296e-06,
"loss": 0.7117088437080383,
"step": 2020
},
{
"epoch": 2.468864468864469,
"grad_norm": 1.9486643075942993,
"learning_rate": 2.6755744846622883e-06,
"loss": 0.5674237012863159,
"step": 2022
},
{
"epoch": 2.4713064713064714,
"grad_norm": 3.0820834636688232,
"learning_rate": 2.6699773018166596e-06,
"loss": 0.6507576704025269,
"step": 2024
},
{
"epoch": 2.473748473748474,
"grad_norm": 8.020563125610352,
"learning_rate": 2.6643829809129126e-06,
"loss": 0.9150622487068176,
"step": 2026
},
{
"epoch": 2.4761904761904763,
"grad_norm": 3.7450058460235596,
"learning_rate": 2.6587915473638174e-06,
"loss": 0.6196163892745972,
"step": 2028
},
{
"epoch": 2.4786324786324787,
"grad_norm": 1.4104655981063843,
"learning_rate": 2.653203026569026e-06,
"loss": 0.920057475566864,
"step": 2030
},
{
"epoch": 2.481074481074481,
"grad_norm": 3.1829826831817627,
"learning_rate": 2.647617443914961e-06,
"loss": 1.105471134185791,
"step": 2032
},
{
"epoch": 2.4835164835164836,
"grad_norm": 4.825045585632324,
"learning_rate": 2.6420348247746955e-06,
"loss": 0.5420823693275452,
"step": 2034
},
{
"epoch": 2.485958485958486,
"grad_norm": 3.713477373123169,
"learning_rate": 2.6364551945078424e-06,
"loss": 1.034119963645935,
"step": 2036
},
{
"epoch": 2.4884004884004884,
"grad_norm": 1.7292916774749756,
"learning_rate": 2.630878578460438e-06,
"loss": 0.6454814076423645,
"step": 2038
},
{
"epoch": 2.490842490842491,
"grad_norm": 6.596334457397461,
"learning_rate": 2.6253050019648242e-06,
"loss": 0.7713245153427124,
"step": 2040
},
{
"epoch": 2.4932844932844933,
"grad_norm": 2.4643394947052,
"learning_rate": 2.619734490339537e-06,
"loss": 0.9619376063346863,
"step": 2042
},
{
"epoch": 2.4957264957264957,
"grad_norm": 1.9008095264434814,
"learning_rate": 2.614167068889189e-06,
"loss": 0.7063919305801392,
"step": 2044
},
{
"epoch": 2.498168498168498,
"grad_norm": 3.062089443206787,
"learning_rate": 2.6086027629043567e-06,
"loss": 0.5970872044563293,
"step": 2046
},
{
"epoch": 2.5006105006105006,
"grad_norm": 1.492427110671997,
"learning_rate": 2.603041597661461e-06,
"loss": 0.6669678092002869,
"step": 2048
},
{
"epoch": 2.503052503052503,
"grad_norm": 1.6604772806167603,
"learning_rate": 2.5974835984226602e-06,
"loss": 0.7523837089538574,
"step": 2050
},
{
"epoch": 2.5054945054945055,
"grad_norm": 7.79230260848999,
"learning_rate": 2.591928790435728e-06,
"loss": 1.1376758813858032,
"step": 2052
},
{
"epoch": 2.507936507936508,
"grad_norm": 1.4952296018600464,
"learning_rate": 2.5863771989339402e-06,
"loss": 0.6450526714324951,
"step": 2054
},
{
"epoch": 2.5103785103785103,
"grad_norm": 1.8876447677612305,
"learning_rate": 2.5808288491359664e-06,
"loss": 0.926227331161499,
"step": 2056
},
{
"epoch": 2.5128205128205128,
"grad_norm": 3.5542497634887695,
"learning_rate": 2.575283766245744e-06,
"loss": 0.4095945358276367,
"step": 2058
},
{
"epoch": 2.515262515262515,
"grad_norm": 2.479278087615967,
"learning_rate": 2.569741975452376e-06,
"loss": 0.2681538462638855,
"step": 2060
},
{
"epoch": 2.5177045177045176,
"grad_norm": 4.712648391723633,
"learning_rate": 2.564203501930006e-06,
"loss": 0.6008268594741821,
"step": 2062
},
{
"epoch": 2.52014652014652,
"grad_norm": 2.5217740535736084,
"learning_rate": 2.558668370837712e-06,
"loss": 0.9255989789962769,
"step": 2064
},
{
"epoch": 2.5225885225885225,
"grad_norm": 1.3460025787353516,
"learning_rate": 2.5531366073193857e-06,
"loss": 0.628387987613678,
"step": 2066
},
{
"epoch": 2.525030525030525,
"grad_norm": 5.685485363006592,
"learning_rate": 2.5476082365036257e-06,
"loss": 0.7797712683677673,
"step": 2068
},
{
"epoch": 2.5274725274725274,
"grad_norm": 1.2678775787353516,
"learning_rate": 2.5420832835036126e-06,
"loss": 1.1243481636047363,
"step": 2070
},
{
"epoch": 2.52991452991453,
"grad_norm": 1.9160879850387573,
"learning_rate": 2.5365617734170076e-06,
"loss": 0.7206214666366577,
"step": 2072
},
{
"epoch": 2.5323565323565322,
"grad_norm": 7.4982829093933105,
"learning_rate": 2.531043731325827e-06,
"loss": 0.9347074627876282,
"step": 2074
},
{
"epoch": 2.5347985347985347,
"grad_norm": 2.430102586746216,
"learning_rate": 2.5255291822963357e-06,
"loss": 0.8390841484069824,
"step": 2076
},
{
"epoch": 2.537240537240537,
"grad_norm": 4.842471599578857,
"learning_rate": 2.5200181513789336e-06,
"loss": 0.9063934683799744,
"step": 2078
},
{
"epoch": 2.5396825396825395,
"grad_norm": 3.8723435401916504,
"learning_rate": 2.5145106636080333e-06,
"loss": 0.6532993316650391,
"step": 2080
},
{
"epoch": 2.542124542124542,
"grad_norm": 2.0189385414123535,
"learning_rate": 2.5090067440019583e-06,
"loss": 0.9501420855522156,
"step": 2082
},
{
"epoch": 2.5445665445665444,
"grad_norm": 2.7918736934661865,
"learning_rate": 2.503506417562817e-06,
"loss": 0.9153074026107788,
"step": 2084
},
{
"epoch": 2.547008547008547,
"grad_norm": 1.4488838911056519,
"learning_rate": 2.498009709276403e-06,
"loss": 0.8253804445266724,
"step": 2086
},
{
"epoch": 2.5494505494505493,
"grad_norm": 2.9465348720550537,
"learning_rate": 2.492516644112067e-06,
"loss": 0.9089316129684448,
"step": 2088
},
{
"epoch": 2.5518925518925517,
"grad_norm": 3.4904873371124268,
"learning_rate": 2.487027247022616e-06,
"loss": 0.7865045666694641,
"step": 2090
},
{
"epoch": 2.554334554334554,
"grad_norm": 2.272559642791748,
"learning_rate": 2.4815415429441893e-06,
"loss": 0.6603869199752808,
"step": 2092
},
{
"epoch": 2.5567765567765566,
"grad_norm": 2.10170841217041,
"learning_rate": 2.476059556796156e-06,
"loss": 0.6427494287490845,
"step": 2094
},
{
"epoch": 2.559218559218559,
"grad_norm": 1.574903130531311,
"learning_rate": 2.4705813134809898e-06,
"loss": 0.8682103157043457,
"step": 2096
},
{
"epoch": 2.5616605616605614,
"grad_norm": 1.1644954681396484,
"learning_rate": 2.4651068378841686e-06,
"loss": 0.03524484112858772,
"step": 2098
},
{
"epoch": 2.564102564102564,
"grad_norm": 2.8488667011260986,
"learning_rate": 2.4596361548740497e-06,
"loss": 0.5300520658493042,
"step": 2100
},
{
"epoch": 2.5665445665445663,
"grad_norm": 1.7959370613098145,
"learning_rate": 2.4541692893017665e-06,
"loss": 0.8741487860679626,
"step": 2102
},
{
"epoch": 2.5689865689865687,
"grad_norm": 4.016081809997559,
"learning_rate": 2.4487062660011067e-06,
"loss": 0.6247915029525757,
"step": 2104
},
{
"epoch": 2.571428571428571,
"grad_norm": 2.1877832412719727,
"learning_rate": 2.4432471097884097e-06,
"loss": 0.869472861289978,
"step": 2106
},
{
"epoch": 2.5738705738705736,
"grad_norm": 5.9354143142700195,
"learning_rate": 2.437791845462442e-06,
"loss": 0.5715877413749695,
"step": 2108
},
{
"epoch": 2.576312576312576,
"grad_norm": 5.521259307861328,
"learning_rate": 2.4323404978042953e-06,
"loss": 0.5813296437263489,
"step": 2110
},
{
"epoch": 2.578754578754579,
"grad_norm": 1.3393948078155518,
"learning_rate": 2.4268930915772708e-06,
"loss": 0.9946026802062988,
"step": 2112
},
{
"epoch": 2.5811965811965814,
"grad_norm": 2.754868984222412,
"learning_rate": 2.4214496515267584e-06,
"loss": 0.27745679020881653,
"step": 2114
},
{
"epoch": 2.583638583638584,
"grad_norm": 2.2351222038269043,
"learning_rate": 2.4160102023801384e-06,
"loss": 0.6880282163619995,
"step": 2116
},
{
"epoch": 2.586080586080586,
"grad_norm": 2.2096810340881348,
"learning_rate": 2.410574768846658e-06,
"loss": 1.0098536014556885,
"step": 2118
},
{
"epoch": 2.5885225885225887,
"grad_norm": 1.1249773502349854,
"learning_rate": 2.4051433756173248e-06,
"loss": 0.7218607068061829,
"step": 2120
},
{
"epoch": 2.590964590964591,
"grad_norm": 5.084073543548584,
"learning_rate": 2.399716047364792e-06,
"loss": 0.9250280857086182,
"step": 2122
},
{
"epoch": 2.5934065934065935,
"grad_norm": 1.690556526184082,
"learning_rate": 2.394292808743248e-06,
"loss": 0.837040364742279,
"step": 2124
},
{
"epoch": 2.595848595848596,
"grad_norm": 2.8815386295318604,
"learning_rate": 2.3888736843883023e-06,
"loss": 0.4333546459674835,
"step": 2126
},
{
"epoch": 2.5982905982905984,
"grad_norm": 4.304377555847168,
"learning_rate": 2.383458698916877e-06,
"loss": 1.0332998037338257,
"step": 2128
},
{
"epoch": 2.600732600732601,
"grad_norm": 2.969372272491455,
"learning_rate": 2.3780478769270907e-06,
"loss": 0.6737803220748901,
"step": 2130
},
{
"epoch": 2.6031746031746033,
"grad_norm": 1.7612605094909668,
"learning_rate": 2.372641242998151e-06,
"loss": 0.527601957321167,
"step": 2132
},
{
"epoch": 2.6056166056166057,
"grad_norm": 1.2742631435394287,
"learning_rate": 2.367238821690239e-06,
"loss": 0.8370689749717712,
"step": 2134
},
{
"epoch": 2.608058608058608,
"grad_norm": 6.330650329589844,
"learning_rate": 2.361840637544402e-06,
"loss": 0.5774081945419312,
"step": 2136
},
{
"epoch": 2.6105006105006106,
"grad_norm": 5.526278972625732,
"learning_rate": 2.3564467150824347e-06,
"loss": 0.7380045652389526,
"step": 2138
},
{
"epoch": 2.612942612942613,
"grad_norm": 1.5810257196426392,
"learning_rate": 2.351057078806779e-06,
"loss": 0.6179096698760986,
"step": 2140
},
{
"epoch": 2.6153846153846154,
"grad_norm": 2.142446517944336,
"learning_rate": 2.3456717532004014e-06,
"loss": 0.8616040349006653,
"step": 2142
},
{
"epoch": 2.617826617826618,
"grad_norm": 3.2861132621765137,
"learning_rate": 2.340290762726691e-06,
"loss": 0.8608450293540955,
"step": 2144
},
{
"epoch": 2.6202686202686203,
"grad_norm": 3.9236459732055664,
"learning_rate": 2.334914131829338e-06,
"loss": 1.0178287029266357,
"step": 2146
},
{
"epoch": 2.6227106227106227,
"grad_norm": 2.530313491821289,
"learning_rate": 2.3295418849322353e-06,
"loss": 0.6632241606712341,
"step": 2148
},
{
"epoch": 2.625152625152625,
"grad_norm": 1.9764374494552612,
"learning_rate": 2.3241740464393575e-06,
"loss": 1.0255072116851807,
"step": 2150
},
{
"epoch": 2.6275946275946276,
"grad_norm": 2.218528985977173,
"learning_rate": 2.318810640734654e-06,
"loss": 1.0632559061050415,
"step": 2152
},
{
"epoch": 2.63003663003663,
"grad_norm": 2.4311797618865967,
"learning_rate": 2.3134516921819398e-06,
"loss": 1.0157952308654785,
"step": 2154
},
{
"epoch": 2.6324786324786325,
"grad_norm": 1.8771414756774902,
"learning_rate": 2.308097225124779e-06,
"loss": 0.9064328074455261,
"step": 2156
},
{
"epoch": 2.634920634920635,
"grad_norm": 2.203409433364868,
"learning_rate": 2.302747263886381e-06,
"loss": 0.7967960238456726,
"step": 2158
},
{
"epoch": 2.6373626373626373,
"grad_norm": 2.97739315032959,
"learning_rate": 2.2974018327694855e-06,
"loss": 0.5664879083633423,
"step": 2160
},
{
"epoch": 2.6398046398046398,
"grad_norm": 3.0944299697875977,
"learning_rate": 2.2920609560562573e-06,
"loss": 0.7291885018348694,
"step": 2162
},
{
"epoch": 2.642246642246642,
"grad_norm": 6.805998802185059,
"learning_rate": 2.286724658008166e-06,
"loss": 0.5031374096870422,
"step": 2164
},
{
"epoch": 2.6446886446886446,
"grad_norm": 2.784329414367676,
"learning_rate": 2.281392962865888e-06,
"loss": 0.88991379737854,
"step": 2166
},
{
"epoch": 2.647130647130647,
"grad_norm": 1.5047616958618164,
"learning_rate": 2.2760658948491872e-06,
"loss": 0.6943953037261963,
"step": 2168
},
{
"epoch": 2.6495726495726495,
"grad_norm": 1.8669089078903198,
"learning_rate": 2.27074347815681e-06,
"loss": 0.999853789806366,
"step": 2170
},
{
"epoch": 2.652014652014652,
"grad_norm": 1.8199567794799805,
"learning_rate": 2.265425736966371e-06,
"loss": 1.0209516286849976,
"step": 2172
},
{
"epoch": 2.6544566544566544,
"grad_norm": 2.209888219833374,
"learning_rate": 2.2601126954342515e-06,
"loss": 1.0429832935333252,
"step": 2174
},
{
"epoch": 2.656898656898657,
"grad_norm": 3.112511157989502,
"learning_rate": 2.2548043776954764e-06,
"loss": 0.7639720439910889,
"step": 2176
},
{
"epoch": 2.659340659340659,
"grad_norm": 2.3887898921966553,
"learning_rate": 2.2495008078636194e-06,
"loss": 0.8997243642807007,
"step": 2178
},
{
"epoch": 2.6617826617826617,
"grad_norm": 6.595883846282959,
"learning_rate": 2.244202010030682e-06,
"loss": 1.002269983291626,
"step": 2180
},
{
"epoch": 2.664224664224664,
"grad_norm": 2.121857166290283,
"learning_rate": 2.2389080082669896e-06,
"loss": 0.9555788636207581,
"step": 2182
},
{
"epoch": 2.6666666666666665,
"grad_norm": 2.1821184158325195,
"learning_rate": 2.2336188266210827e-06,
"loss": 0.6721835732460022,
"step": 2184
},
{
"epoch": 2.669108669108669,
"grad_norm": 2.1556482315063477,
"learning_rate": 2.2283344891196015e-06,
"loss": 0.9556472897529602,
"step": 2186
},
{
"epoch": 2.6715506715506714,
"grad_norm": 2.405789613723755,
"learning_rate": 2.2230550197671873e-06,
"loss": 1.1528511047363281,
"step": 2188
},
{
"epoch": 2.6739926739926743,
"grad_norm": 1.6833081245422363,
"learning_rate": 2.217780442546361e-06,
"loss": 0.974722683429718,
"step": 2190
},
{
"epoch": 2.6764346764346767,
"grad_norm": 4.5762505531311035,
"learning_rate": 2.2125107814174263e-06,
"loss": 0.9278739094734192,
"step": 2192
},
{
"epoch": 2.678876678876679,
"grad_norm": 9.507789611816406,
"learning_rate": 2.20724606031835e-06,
"loss": 0.32617825269699097,
"step": 2194
},
{
"epoch": 2.6813186813186816,
"grad_norm": 1.8784890174865723,
"learning_rate": 2.2019863031646634e-06,
"loss": 1.0722205638885498,
"step": 2196
},
{
"epoch": 2.683760683760684,
"grad_norm": 18.138153076171875,
"learning_rate": 2.1967315338493424e-06,
"loss": 0.8764669299125671,
"step": 2198
},
{
"epoch": 2.6862026862026864,
"grad_norm": 2.8644320964813232,
"learning_rate": 2.1914817762427106e-06,
"loss": 0.9633726477622986,
"step": 2200
},
{
"epoch": 2.688644688644689,
"grad_norm": 11.419971466064453,
"learning_rate": 2.186237054192323e-06,
"loss": 0.6935689449310303,
"step": 2202
},
{
"epoch": 2.6910866910866913,
"grad_norm": 6.860535621643066,
"learning_rate": 2.1809973915228602e-06,
"loss": 0.7737261056900024,
"step": 2204
},
{
"epoch": 2.6935286935286937,
"grad_norm": 3.685971260070801,
"learning_rate": 2.17576281203602e-06,
"loss": 0.41188710927963257,
"step": 2206
},
{
"epoch": 2.695970695970696,
"grad_norm": 2.738652229309082,
"learning_rate": 2.170533339510412e-06,
"loss": 0.9475657939910889,
"step": 2208
},
{
"epoch": 2.6984126984126986,
"grad_norm": 5.204038619995117,
"learning_rate": 2.16530899770144e-06,
"loss": 0.6033111810684204,
"step": 2210
},
{
"epoch": 2.700854700854701,
"grad_norm": 2.083590507507324,
"learning_rate": 2.1600898103412107e-06,
"loss": 0.9223611354827881,
"step": 2212
},
{
"epoch": 2.7032967032967035,
"grad_norm": 1.6157281398773193,
"learning_rate": 2.1548758011384053e-06,
"loss": 0.6177644729614258,
"step": 2214
},
{
"epoch": 2.705738705738706,
"grad_norm": 3.5457189083099365,
"learning_rate": 2.149666993778194e-06,
"loss": 0.4836501181125641,
"step": 2216
},
{
"epoch": 2.7081807081807083,
"grad_norm": 3.0739972591400146,
"learning_rate": 2.1444634119221075e-06,
"loss": 0.5934576988220215,
"step": 2218
},
{
"epoch": 2.7106227106227108,
"grad_norm": 4.0983781814575195,
"learning_rate": 2.1392650792079454e-06,
"loss": 0.3090117573738098,
"step": 2220
},
{
"epoch": 2.713064713064713,
"grad_norm": 5.047109603881836,
"learning_rate": 2.1340720192496594e-06,
"loss": 0.5931053757667542,
"step": 2222
},
{
"epoch": 2.7155067155067156,
"grad_norm": 4.083251476287842,
"learning_rate": 2.12888425563725e-06,
"loss": 0.6489290595054626,
"step": 2224
},
{
"epoch": 2.717948717948718,
"grad_norm": 1.615699291229248,
"learning_rate": 2.123701811936659e-06,
"loss": 0.7429230213165283,
"step": 2226
},
{
"epoch": 2.7203907203907205,
"grad_norm": 4.504569053649902,
"learning_rate": 2.118524711689661e-06,
"loss": 0.693551778793335,
"step": 2228
},
{
"epoch": 2.722832722832723,
"grad_norm": 1.5205539464950562,
"learning_rate": 2.1133529784137585e-06,
"loss": 1.0012576580047607,
"step": 2230
},
{
"epoch": 2.7252747252747254,
"grad_norm": 3.8578217029571533,
"learning_rate": 2.1081866356020726e-06,
"loss": 0.6832969784736633,
"step": 2232
},
{
"epoch": 2.727716727716728,
"grad_norm": 3.5652737617492676,
"learning_rate": 2.103025706723239e-06,
"loss": 1.0270230770111084,
"step": 2234
},
{
"epoch": 2.7301587301587302,
"grad_norm": 10.974838256835938,
"learning_rate": 2.0978702152212996e-06,
"loss": 1.0029627084732056,
"step": 2236
},
{
"epoch": 2.7326007326007327,
"grad_norm": 5.731579303741455,
"learning_rate": 2.092720184515597e-06,
"loss": 0.956536054611206,
"step": 2238
},
{
"epoch": 2.735042735042735,
"grad_norm": 2.9073221683502197,
"learning_rate": 2.0875756380006654e-06,
"loss": 0.6759551167488098,
"step": 2240
},
{
"epoch": 2.7374847374847375,
"grad_norm": 1.188262939453125,
"learning_rate": 2.0824365990461295e-06,
"loss": 1.0032782554626465,
"step": 2242
},
{
"epoch": 2.73992673992674,
"grad_norm": 3.1204891204833984,
"learning_rate": 2.077303090996594e-06,
"loss": 0.6075339913368225,
"step": 2244
},
{
"epoch": 2.7423687423687424,
"grad_norm": 2.2777647972106934,
"learning_rate": 2.072175137171538e-06,
"loss": 0.6075271964073181,
"step": 2246
},
{
"epoch": 2.744810744810745,
"grad_norm": 4.017678737640381,
"learning_rate": 2.067052760865211e-06,
"loss": 0.7910500168800354,
"step": 2248
},
{
"epoch": 2.7472527472527473,
"grad_norm": 2.2465343475341797,
"learning_rate": 2.0619359853465268e-06,
"loss": 0.8944975137710571,
"step": 2250
},
{
"epoch": 2.7496947496947497,
"grad_norm": 2.802507162094116,
"learning_rate": 2.0568248338589546e-06,
"loss": 1.0018844604492188,
"step": 2252
},
{
"epoch": 2.752136752136752,
"grad_norm": 1.5768181085586548,
"learning_rate": 2.051719329620418e-06,
"loss": 0.6374676823616028,
"step": 2254
},
{
"epoch": 2.7545787545787546,
"grad_norm": 5.736408233642578,
"learning_rate": 2.046619495823189e-06,
"loss": 0.7124966979026794,
"step": 2256
},
{
"epoch": 2.757020757020757,
"grad_norm": 1.7231215238571167,
"learning_rate": 2.041525355633775e-06,
"loss": 0.9689019918441772,
"step": 2258
},
{
"epoch": 2.7594627594627594,
"grad_norm": 3.9329147338867188,
"learning_rate": 2.0364369321928277e-06,
"loss": 1.0872962474822998,
"step": 2260
},
{
"epoch": 2.761904761904762,
"grad_norm": 1.9231138229370117,
"learning_rate": 2.031354248615021e-06,
"loss": 1.0194430351257324,
"step": 2262
},
{
"epoch": 2.7643467643467643,
"grad_norm": 1.4887934923171997,
"learning_rate": 2.0262773279889648e-06,
"loss": 0.9722414612770081,
"step": 2264
},
{
"epoch": 2.7667887667887667,
"grad_norm": 1.6510133743286133,
"learning_rate": 2.0212061933770793e-06,
"loss": 0.789564311504364,
"step": 2266
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.8453580737113953,
"learning_rate": 2.016140867815512e-06,
"loss": 0.655318021774292,
"step": 2268
},
{
"epoch": 2.7716727716727716,
"grad_norm": 1.5889426469802856,
"learning_rate": 2.0110813743140134e-06,
"loss": 0.5706247091293335,
"step": 2270
},
{
"epoch": 2.774114774114774,
"grad_norm": 1.9540982246398926,
"learning_rate": 2.0060277358558474e-06,
"loss": 0.16020342707633972,
"step": 2272
},
{
"epoch": 2.7765567765567765,
"grad_norm": 1.7977136373519897,
"learning_rate": 2.0009799753976755e-06,
"loss": 1.0128493309020996,
"step": 2274
},
{
"epoch": 2.778998778998779,
"grad_norm": 1.6063421964645386,
"learning_rate": 1.995938115869464e-06,
"loss": 0.9149772524833679,
"step": 2276
},
{
"epoch": 2.7814407814407813,
"grad_norm": 1.2842918634414673,
"learning_rate": 1.990902180174366e-06,
"loss": 0.5837323665618896,
"step": 2278
},
{
"epoch": 2.7838827838827838,
"grad_norm": 12.791586875915527,
"learning_rate": 1.985872191188634e-06,
"loss": 0.33608847856521606,
"step": 2280
},
{
"epoch": 2.786324786324786,
"grad_norm": 1.533747911453247,
"learning_rate": 1.9808481717614975e-06,
"loss": 0.9576466679573059,
"step": 2282
},
{
"epoch": 2.7887667887667886,
"grad_norm": 1.179326057434082,
"learning_rate": 1.9758301447150778e-06,
"loss": 0.9575373530387878,
"step": 2284
},
{
"epoch": 2.791208791208791,
"grad_norm": 1.6882575750350952,
"learning_rate": 1.9708181328442657e-06,
"loss": 0.885867178440094,
"step": 2286
},
{
"epoch": 2.7936507936507935,
"grad_norm": 1.46290922164917,
"learning_rate": 1.9658121589166368e-06,
"loss": 0.9062331318855286,
"step": 2288
},
{
"epoch": 2.796092796092796,
"grad_norm": 6.2452216148376465,
"learning_rate": 1.960812245672332e-06,
"loss": 0.2752612829208374,
"step": 2290
},
{
"epoch": 2.7985347985347984,
"grad_norm": 2.4964230060577393,
"learning_rate": 1.955818415823965e-06,
"loss": 0.7961699366569519,
"step": 2292
},
{
"epoch": 2.800976800976801,
"grad_norm": 1.7693415880203247,
"learning_rate": 1.950830692056512e-06,
"loss": 0.781195342540741,
"step": 2294
},
{
"epoch": 2.8034188034188032,
"grad_norm": 1.656353235244751,
"learning_rate": 1.9458490970272132e-06,
"loss": 0.9211829900741577,
"step": 2296
},
{
"epoch": 2.8058608058608057,
"grad_norm": 5.0975494384765625,
"learning_rate": 1.9408736533654693e-06,
"loss": 0.43949294090270996,
"step": 2298
},
{
"epoch": 2.808302808302808,
"grad_norm": 1.9063336849212646,
"learning_rate": 1.9359043836727363e-06,
"loss": 0.6156795024871826,
"step": 2300
},
{
"epoch": 2.8107448107448105,
"grad_norm": 1.8999279737472534,
"learning_rate": 1.9309413105224242e-06,
"loss": 0.9818214774131775,
"step": 2302
},
{
"epoch": 2.813186813186813,
"grad_norm": 0.6018390655517578,
"learning_rate": 1.9259844564597955e-06,
"loss": 0.6223698258399963,
"step": 2304
},
{
"epoch": 2.8156288156288154,
"grad_norm": 1.268938660621643,
"learning_rate": 1.9210338440018625e-06,
"loss": 0.7103510499000549,
"step": 2306
},
{
"epoch": 2.818070818070818,
"grad_norm": 1.6142417192459106,
"learning_rate": 1.916089495637282e-06,
"loss": 0.6377496123313904,
"step": 2308
},
{
"epoch": 2.8205128205128203,
"grad_norm": 2.1147496700286865,
"learning_rate": 1.911151433826258e-06,
"loss": 0.995940089225769,
"step": 2310
},
{
"epoch": 2.8229548229548227,
"grad_norm": 1.3524115085601807,
"learning_rate": 1.9062196810004356e-06,
"loss": 0.6259192824363708,
"step": 2312
},
{
"epoch": 2.825396825396825,
"grad_norm": 7.68358039855957,
"learning_rate": 1.9012942595628017e-06,
"loss": 0.7580180764198303,
"step": 2314
},
{
"epoch": 2.8278388278388276,
"grad_norm": 19.87650489807129,
"learning_rate": 1.8963751918875811e-06,
"loss": 0.7949362993240356,
"step": 2316
},
{
"epoch": 2.8302808302808304,
"grad_norm": 2.42893123626709,
"learning_rate": 1.8914625003201375e-06,
"loss": 1.0362216234207153,
"step": 2318
},
{
"epoch": 2.832722832722833,
"grad_norm": 2.1275134086608887,
"learning_rate": 1.8865562071768689e-06,
"loss": 0.6284617185592651,
"step": 2320
},
{
"epoch": 2.8351648351648353,
"grad_norm": 4.848043918609619,
"learning_rate": 1.8816563347451088e-06,
"loss": 0.6487046480178833,
"step": 2322
},
{
"epoch": 2.8376068376068377,
"grad_norm": 1.1281238794326782,
"learning_rate": 1.8767629052830238e-06,
"loss": 0.23647232353687286,
"step": 2324
},
{
"epoch": 2.84004884004884,
"grad_norm": 1.886053442955017,
"learning_rate": 1.8718759410195113e-06,
"loss": 0.8259285092353821,
"step": 2326
},
{
"epoch": 2.8424908424908426,
"grad_norm": 2.046250343322754,
"learning_rate": 1.8669954641541034e-06,
"loss": 0.7600533962249756,
"step": 2328
},
{
"epoch": 2.844932844932845,
"grad_norm": 2.649106025695801,
"learning_rate": 1.8621214968568598e-06,
"loss": 0.7137502431869507,
"step": 2330
},
{
"epoch": 2.8473748473748475,
"grad_norm": 2.0803020000457764,
"learning_rate": 1.8572540612682694e-06,
"loss": 0.965351402759552,
"step": 2332
},
{
"epoch": 2.84981684981685,
"grad_norm": 2.1527020931243896,
"learning_rate": 1.8523931794991518e-06,
"loss": 0.10668709874153137,
"step": 2334
},
{
"epoch": 2.8522588522588523,
"grad_norm": 1.98074209690094,
"learning_rate": 1.8475388736305535e-06,
"loss": 0.9549130797386169,
"step": 2336
},
{
"epoch": 2.8547008547008548,
"grad_norm": 1.5347501039505005,
"learning_rate": 1.8426911657136517e-06,
"loss": 0.5754918456077576,
"step": 2338
},
{
"epoch": 2.857142857142857,
"grad_norm": 2.49215030670166,
"learning_rate": 1.8378500777696496e-06,
"loss": 1.0284221172332764,
"step": 2340
},
{
"epoch": 2.8595848595848596,
"grad_norm": 1.4292778968811035,
"learning_rate": 1.8330156317896794e-06,
"loss": 0.9731957912445068,
"step": 2342
},
{
"epoch": 2.862026862026862,
"grad_norm": 5.513943195343018,
"learning_rate": 1.8281878497347007e-06,
"loss": 0.320040762424469,
"step": 2344
},
{
"epoch": 2.8644688644688645,
"grad_norm": 5.377645969390869,
"learning_rate": 1.8233667535354031e-06,
"loss": 0.896674633026123,
"step": 2346
},
{
"epoch": 2.866910866910867,
"grad_norm": 0.5098855495452881,
"learning_rate": 1.8185523650921033e-06,
"loss": 0.4520077109336853,
"step": 2348
},
{
"epoch": 2.8693528693528694,
"grad_norm": 1.4593082666397095,
"learning_rate": 1.813744706274649e-06,
"loss": 0.5917829275131226,
"step": 2350
},
{
"epoch": 2.871794871794872,
"grad_norm": 1.9374688863754272,
"learning_rate": 1.8089437989223166e-06,
"loss": 0.7799688577651978,
"step": 2352
},
{
"epoch": 2.8742368742368742,
"grad_norm": 3.2081990242004395,
"learning_rate": 1.8041496648437138e-06,
"loss": 0.5761586427688599,
"step": 2354
},
{
"epoch": 2.8766788766788767,
"grad_norm": 1.424476981163025,
"learning_rate": 1.7993623258166794e-06,
"loss": 0.8961036801338196,
"step": 2356
},
{
"epoch": 2.879120879120879,
"grad_norm": 6.229803562164307,
"learning_rate": 1.7945818035881866e-06,
"loss": 0.676258385181427,
"step": 2358
},
{
"epoch": 2.8815628815628815,
"grad_norm": 2.3079984188079834,
"learning_rate": 1.789808119874241e-06,
"loss": 0.5127954483032227,
"step": 2360
},
{
"epoch": 2.884004884004884,
"grad_norm": 3.7900733947753906,
"learning_rate": 1.7850412963597837e-06,
"loss": 0.7495479583740234,
"step": 2362
},
{
"epoch": 2.8864468864468864,
"grad_norm": 1.7501633167266846,
"learning_rate": 1.7802813546985936e-06,
"loss": 0.9721358418464661,
"step": 2364
},
{
"epoch": 2.888888888888889,
"grad_norm": 1.494238257408142,
"learning_rate": 1.7755283165131876e-06,
"loss": 1.0429596900939941,
"step": 2366
},
{
"epoch": 2.8913308913308913,
"grad_norm": 9.534012794494629,
"learning_rate": 1.7707822033947214e-06,
"loss": 0.7053688764572144,
"step": 2368
},
{
"epoch": 2.8937728937728937,
"grad_norm": 1.7080990076065063,
"learning_rate": 1.7660430369028975e-06,
"loss": 1.0583832263946533,
"step": 2370
},
{
"epoch": 2.896214896214896,
"grad_norm": 2.404010534286499,
"learning_rate": 1.7613108385658554e-06,
"loss": 0.9982654452323914,
"step": 2372
},
{
"epoch": 2.8986568986568986,
"grad_norm": 1.0838375091552734,
"learning_rate": 1.7565856298800887e-06,
"loss": 0.767935037612915,
"step": 2374
},
{
"epoch": 2.901098901098901,
"grad_norm": 2.0502405166625977,
"learning_rate": 1.7518674323103325e-06,
"loss": 0.8051881790161133,
"step": 2376
},
{
"epoch": 2.9035409035409034,
"grad_norm": 2.304351806640625,
"learning_rate": 1.7471562672894812e-06,
"loss": 0.8764021396636963,
"step": 2378
},
{
"epoch": 2.905982905982906,
"grad_norm": 3.1706900596618652,
"learning_rate": 1.7424521562184747e-06,
"loss": 0.686414897441864,
"step": 2380
},
{
"epoch": 2.9084249084249083,
"grad_norm": 2.4662387371063232,
"learning_rate": 1.7377551204662173e-06,
"loss": 0.9558066129684448,
"step": 2382
},
{
"epoch": 2.9108669108669107,
"grad_norm": 1.8197959661483765,
"learning_rate": 1.7330651813694656e-06,
"loss": 0.8915336728096008,
"step": 2384
},
{
"epoch": 2.913308913308913,
"grad_norm": 1.8730552196502686,
"learning_rate": 1.7283823602327467e-06,
"loss": 0.9963794946670532,
"step": 2386
},
{
"epoch": 2.9157509157509156,
"grad_norm": 3.505786418914795,
"learning_rate": 1.7237066783282454e-06,
"loss": 0.7860287427902222,
"step": 2388
},
{
"epoch": 2.918192918192918,
"grad_norm": 10.96102523803711,
"learning_rate": 1.719038156895725e-06,
"loss": 0.5170654058456421,
"step": 2390
},
{
"epoch": 2.9206349206349205,
"grad_norm": 2.723895311355591,
"learning_rate": 1.7143768171424114e-06,
"loss": 0.5465037822723389,
"step": 2392
},
{
"epoch": 2.9230769230769234,
"grad_norm": 2.4230728149414062,
"learning_rate": 1.709722680242917e-06,
"loss": 0.8477396368980408,
"step": 2394
},
{
"epoch": 2.925518925518926,
"grad_norm": 4.999912261962891,
"learning_rate": 1.7050757673391256e-06,
"loss": 0.5552250742912292,
"step": 2396
},
{
"epoch": 2.927960927960928,
"grad_norm": 4.302630424499512,
"learning_rate": 1.7004360995401131e-06,
"loss": 0.6446186304092407,
"step": 2398
},
{
"epoch": 2.9304029304029307,
"grad_norm": 11.290936470031738,
"learning_rate": 1.6958036979220378e-06,
"loss": 0.5395065546035767,
"step": 2400
},
{
"epoch": 2.932844932844933,
"grad_norm": 2.418199062347412,
"learning_rate": 1.6911785835280544e-06,
"loss": 0.6396113634109497,
"step": 2402
},
{
"epoch": 2.9352869352869355,
"grad_norm": 3.1459553241729736,
"learning_rate": 1.686560777368212e-06,
"loss": 0.3191334009170532,
"step": 2404
},
{
"epoch": 2.937728937728938,
"grad_norm": 8.619038581848145,
"learning_rate": 1.6819503004193633e-06,
"loss": 1.0637816190719604,
"step": 2406
},
{
"epoch": 2.9401709401709404,
"grad_norm": 2.521735191345215,
"learning_rate": 1.6773471736250668e-06,
"loss": 0.6778561472892761,
"step": 2408
},
{
"epoch": 2.942612942612943,
"grad_norm": 2.928496837615967,
"learning_rate": 1.672751417895491e-06,
"loss": 0.6122907996177673,
"step": 2410
},
{
"epoch": 2.9450549450549453,
"grad_norm": 3.0468804836273193,
"learning_rate": 1.6681630541073222e-06,
"loss": 0.9394041895866394,
"step": 2412
},
{
"epoch": 2.9474969474969477,
"grad_norm": 2.0013132095336914,
"learning_rate": 1.6635821031036671e-06,
"loss": 0.9378724694252014,
"step": 2414
},
{
"epoch": 2.94993894993895,
"grad_norm": 1.7880492210388184,
"learning_rate": 1.6590085856939593e-06,
"loss": 0.9468516111373901,
"step": 2416
},
{
"epoch": 2.9523809523809526,
"grad_norm": 2.087956666946411,
"learning_rate": 1.654442522653865e-06,
"loss": 0.6785021424293518,
"step": 2418
},
{
"epoch": 2.954822954822955,
"grad_norm": 1.5312798023223877,
"learning_rate": 1.6498839347251873e-06,
"loss": 1.0038193464279175,
"step": 2420
},
{
"epoch": 2.9572649572649574,
"grad_norm": 1.5011390447616577,
"learning_rate": 1.6453328426157738e-06,
"loss": 0.8261772990226746,
"step": 2422
},
{
"epoch": 2.95970695970696,
"grad_norm": 2.2196717262268066,
"learning_rate": 1.6407892669994208e-06,
"loss": 0.5461646914482117,
"step": 2424
},
{
"epoch": 2.9621489621489623,
"grad_norm": 1.7109367847442627,
"learning_rate": 1.63625322851578e-06,
"loss": 0.9957700967788696,
"step": 2426
},
{
"epoch": 2.9645909645909647,
"grad_norm": 15.042463302612305,
"learning_rate": 1.6317247477702666e-06,
"loss": 0.5697113871574402,
"step": 2428
},
{
"epoch": 2.967032967032967,
"grad_norm": 2.1879959106445312,
"learning_rate": 1.6272038453339615e-06,
"loss": 0.9809165000915527,
"step": 2430
},
{
"epoch": 2.9694749694749696,
"grad_norm": 2.687741756439209,
"learning_rate": 1.6226905417435226e-06,
"loss": 0.851597249507904,
"step": 2432
},
{
"epoch": 2.971916971916972,
"grad_norm": 2.330556631088257,
"learning_rate": 1.618184857501088e-06,
"loss": 0.93310546875,
"step": 2434
},
{
"epoch": 2.9743589743589745,
"grad_norm": 2.613746404647827,
"learning_rate": 1.6136868130741836e-06,
"loss": 0.4884783625602722,
"step": 2436
},
{
"epoch": 2.976800976800977,
"grad_norm": 7.81420373916626,
"learning_rate": 1.6091964288956337e-06,
"loss": 0.5040377974510193,
"step": 2438
},
{
"epoch": 2.9792429792429793,
"grad_norm": 1.7252824306488037,
"learning_rate": 1.6047137253634603e-06,
"loss": 0.934016227722168,
"step": 2440
},
{
"epoch": 2.9816849816849818,
"grad_norm": 1.4692366123199463,
"learning_rate": 1.600238722840801e-06,
"loss": 0.9195917248725891,
"step": 2442
},
{
"epoch": 2.984126984126984,
"grad_norm": 2.0103518962860107,
"learning_rate": 1.595771441655804e-06,
"loss": 0.9216701984405518,
"step": 2444
},
{
"epoch": 2.9865689865689866,
"grad_norm": 2.632267951965332,
"learning_rate": 1.5913119021015509e-06,
"loss": 0.8953844308853149,
"step": 2446
},
{
"epoch": 2.989010989010989,
"grad_norm": 1.5844495296478271,
"learning_rate": 1.5868601244359465e-06,
"loss": 0.9133291244506836,
"step": 2448
},
{
"epoch": 2.9914529914529915,
"grad_norm": 8.298254013061523,
"learning_rate": 1.5824161288816458e-06,
"loss": 0.5930345058441162,
"step": 2450
},
{
"epoch": 2.993894993894994,
"grad_norm": 1.963354229927063,
"learning_rate": 1.5779799356259456e-06,
"loss": 0.9954817891120911,
"step": 2452
},
{
"epoch": 2.9963369963369964,
"grad_norm": 1.8422874212265015,
"learning_rate": 1.5735515648207054e-06,
"loss": 0.8938966989517212,
"step": 2454
},
{
"epoch": 2.998778998778999,
"grad_norm": 1.3020809888839722,
"learning_rate": 1.5691310365822446e-06,
"loss": 0.4988337457180023,
"step": 2456
},
{
"epoch": 3.001221001221001,
"grad_norm": 10.537405967712402,
"learning_rate": 1.5647183709912631e-06,
"loss": 0.6655434370040894,
"step": 2458
},
{
"epoch": 3.0036630036630036,
"grad_norm": 3.749023199081421,
"learning_rate": 1.560313588092739e-06,
"loss": 0.5545458197593689,
"step": 2460
},
{
"epoch": 3.006105006105006,
"grad_norm": 1.9818220138549805,
"learning_rate": 1.5559167078958468e-06,
"loss": 0.9352074265480042,
"step": 2462
},
{
"epoch": 3.0085470085470085,
"grad_norm": 1.7658523321151733,
"learning_rate": 1.5515277503738572e-06,
"loss": 0.7911304235458374,
"step": 2464
},
{
"epoch": 3.010989010989011,
"grad_norm": 3.5951449871063232,
"learning_rate": 1.5471467354640563e-06,
"loss": 0.4669531285762787,
"step": 2466
},
{
"epoch": 3.0134310134310134,
"grad_norm": 1.7102817296981812,
"learning_rate": 1.5427736830676453e-06,
"loss": 0.7882851362228394,
"step": 2468
},
{
"epoch": 3.015873015873016,
"grad_norm": 0.4721863567829132,
"learning_rate": 1.53840861304966e-06,
"loss": 0.2778557240962982,
"step": 2470
},
{
"epoch": 3.0183150183150182,
"grad_norm": 9.13770580291748,
"learning_rate": 1.534051545238872e-06,
"loss": 0.35145601630210876,
"step": 2472
},
{
"epoch": 3.0207570207570207,
"grad_norm": 2.7468252182006836,
"learning_rate": 1.5297024994277019e-06,
"loss": 0.8385328054428101,
"step": 2474
},
{
"epoch": 3.023199023199023,
"grad_norm": 17.04658317565918,
"learning_rate": 1.5253614953721315e-06,
"loss": 0.506757915019989,
"step": 2476
},
{
"epoch": 3.0256410256410255,
"grad_norm": 2.0790674686431885,
"learning_rate": 1.521028552791611e-06,
"loss": 0.5367285013198853,
"step": 2478
},
{
"epoch": 3.028083028083028,
"grad_norm": 1.9070419073104858,
"learning_rate": 1.5167036913689707e-06,
"loss": 0.844150185585022,
"step": 2480
},
{
"epoch": 3.0305250305250304,
"grad_norm": 7.032018184661865,
"learning_rate": 1.5123869307503316e-06,
"loss": 0.8483667969703674,
"step": 2482
},
{
"epoch": 3.032967032967033,
"grad_norm": 2.8655481338500977,
"learning_rate": 1.5080782905450147e-06,
"loss": 0.46071624755859375,
"step": 2484
},
{
"epoch": 3.0354090354090353,
"grad_norm": 2.500929594039917,
"learning_rate": 1.5037777903254555e-06,
"loss": 0.7813271880149841,
"step": 2486
},
{
"epoch": 3.0378510378510377,
"grad_norm": 6.008790969848633,
"learning_rate": 1.499485449627111e-06,
"loss": 0.6232810616493225,
"step": 2488
},
{
"epoch": 3.04029304029304,
"grad_norm": 2.7102444171905518,
"learning_rate": 1.4952012879483739e-06,
"loss": 0.8550852537155151,
"step": 2490
},
{
"epoch": 3.0427350427350426,
"grad_norm": 3.5206353664398193,
"learning_rate": 1.4909253247504813e-06,
"loss": 0.3453124165534973,
"step": 2492
},
{
"epoch": 3.045177045177045,
"grad_norm": 0.4782972037792206,
"learning_rate": 1.4866575794574298e-06,
"loss": 0.27831053733825684,
"step": 2494
},
{
"epoch": 3.0476190476190474,
"grad_norm": 5.067359447479248,
"learning_rate": 1.4823980714558842e-06,
"loss": 0.6302961111068726,
"step": 2496
},
{
"epoch": 3.05006105006105,
"grad_norm": 15.672896385192871,
"learning_rate": 1.4781468200950903e-06,
"loss": 0.3658387064933777,
"step": 2498
},
{
"epoch": 3.0525030525030523,
"grad_norm": 1.5754889249801636,
"learning_rate": 1.4739038446867882e-06,
"loss": 0.4961889386177063,
"step": 2500
},
{
"epoch": 3.0549450549450547,
"grad_norm": 8.665702819824219,
"learning_rate": 1.4696691645051233e-06,
"loss": 0.4041464328765869,
"step": 2502
},
{
"epoch": 3.057387057387057,
"grad_norm": 2.04341459274292,
"learning_rate": 1.4654427987865584e-06,
"loss": 0.7974004745483398,
"step": 2504
},
{
"epoch": 3.0598290598290596,
"grad_norm": 2.483825445175171,
"learning_rate": 1.461224766729788e-06,
"loss": 0.5479510426521301,
"step": 2506
},
{
"epoch": 3.062271062271062,
"grad_norm": 1.7536518573760986,
"learning_rate": 1.4570150874956491e-06,
"loss": 0.6952835917472839,
"step": 2508
},
{
"epoch": 3.064713064713065,
"grad_norm": 2.361480951309204,
"learning_rate": 1.452813780207038e-06,
"loss": 0.8374277949333191,
"step": 2510
},
{
"epoch": 3.0671550671550674,
"grad_norm": 3.7606232166290283,
"learning_rate": 1.4486208639488147e-06,
"loss": 0.42150750756263733,
"step": 2512
},
{
"epoch": 3.06959706959707,
"grad_norm": 2.443471670150757,
"learning_rate": 1.44443635776773e-06,
"loss": 0.7532929182052612,
"step": 2514
},
{
"epoch": 3.0720390720390722,
"grad_norm": 3.036879301071167,
"learning_rate": 1.4402602806723224e-06,
"loss": 0.5805554389953613,
"step": 2516
},
{
"epoch": 3.0744810744810747,
"grad_norm": 2.405412197113037,
"learning_rate": 1.4360926516328486e-06,
"loss": 0.1354294717311859,
"step": 2518
},
{
"epoch": 3.076923076923077,
"grad_norm": 2.239072322845459,
"learning_rate": 1.4319334895811822e-06,
"loss": 0.4422219395637512,
"step": 2520
},
{
"epoch": 3.0793650793650795,
"grad_norm": 1.6425889730453491,
"learning_rate": 1.4277828134107404e-06,
"loss": 0.6835185289382935,
"step": 2522
},
{
"epoch": 3.081807081807082,
"grad_norm": 2.7413876056671143,
"learning_rate": 1.4236406419763888e-06,
"loss": 0.9346991777420044,
"step": 2524
},
{
"epoch": 3.0842490842490844,
"grad_norm": 13.284109115600586,
"learning_rate": 1.4195069940943602e-06,
"loss": 0.5621562004089355,
"step": 2526
},
{
"epoch": 3.086691086691087,
"grad_norm": 2.8617191314697266,
"learning_rate": 1.4153818885421693e-06,
"loss": 0.48587676882743835,
"step": 2528
},
{
"epoch": 3.0891330891330893,
"grad_norm": 4.067028045654297,
"learning_rate": 1.4112653440585253e-06,
"loss": 0.06718393415212631,
"step": 2530
},
{
"epoch": 3.0915750915750917,
"grad_norm": 2.965784788131714,
"learning_rate": 1.4071573793432477e-06,
"loss": 0.36484605073928833,
"step": 2532
},
{
"epoch": 3.094017094017094,
"grad_norm": 6.96315336227417,
"learning_rate": 1.4030580130571828e-06,
"loss": 0.9084134101867676,
"step": 2534
},
{
"epoch": 3.0964590964590966,
"grad_norm": 4.8996663093566895,
"learning_rate": 1.3989672638221164e-06,
"loss": 0.4888841509819031,
"step": 2536
},
{
"epoch": 3.098901098901099,
"grad_norm": 15.704608917236328,
"learning_rate": 1.3948851502206916e-06,
"loss": 0.7055212259292603,
"step": 2538
},
{
"epoch": 3.1013431013431014,
"grad_norm": 3.72160005569458,
"learning_rate": 1.3908116907963219e-06,
"loss": 0.9027807712554932,
"step": 2540
},
{
"epoch": 3.103785103785104,
"grad_norm": 2.0347139835357666,
"learning_rate": 1.3867469040531097e-06,
"loss": 0.5173296928405762,
"step": 2542
},
{
"epoch": 3.1062271062271063,
"grad_norm": 2.45859694480896,
"learning_rate": 1.3826908084557599e-06,
"loss": 0.515606701374054,
"step": 2544
},
{
"epoch": 3.1086691086691087,
"grad_norm": 4.653799533843994,
"learning_rate": 1.3786434224294974e-06,
"loss": 0.09697027504444122,
"step": 2546
},
{
"epoch": 3.111111111111111,
"grad_norm": 2.246595621109009,
"learning_rate": 1.374604764359983e-06,
"loss": 0.5722851753234863,
"step": 2548
},
{
"epoch": 3.1135531135531136,
"grad_norm": 3.75177264213562,
"learning_rate": 1.3705748525932296e-06,
"loss": 0.6366295218467712,
"step": 2550
},
{
"epoch": 3.115995115995116,
"grad_norm": 12.912460327148438,
"learning_rate": 1.36655370543552e-06,
"loss": 0.19046643376350403,
"step": 2552
},
{
"epoch": 3.1184371184371185,
"grad_norm": 2.1945950984954834,
"learning_rate": 1.3625413411533215e-06,
"loss": 0.8602652549743652,
"step": 2554
},
{
"epoch": 3.120879120879121,
"grad_norm": 14.562986373901367,
"learning_rate": 1.3585377779732054e-06,
"loss": 0.4864516258239746,
"step": 2556
},
{
"epoch": 3.1233211233211233,
"grad_norm": 1.2465832233428955,
"learning_rate": 1.3545430340817623e-06,
"loss": 0.414315789937973,
"step": 2558
},
{
"epoch": 3.1257631257631258,
"grad_norm": 2.1674346923828125,
"learning_rate": 1.350557127625521e-06,
"loss": 0.8689194321632385,
"step": 2560
},
{
"epoch": 3.128205128205128,
"grad_norm": 4.005629062652588,
"learning_rate": 1.3465800767108644e-06,
"loss": 0.9341198801994324,
"step": 2562
},
{
"epoch": 3.1306471306471306,
"grad_norm": 3.285874605178833,
"learning_rate": 1.342611899403951e-06,
"loss": 0.8401010632514954,
"step": 2564
},
{
"epoch": 3.133089133089133,
"grad_norm": 4.772563934326172,
"learning_rate": 1.338652613730625e-06,
"loss": 0.5630367398262024,
"step": 2566
},
{
"epoch": 3.1355311355311355,
"grad_norm": 3.344740629196167,
"learning_rate": 1.334702237676345e-06,
"loss": 0.7592120170593262,
"step": 2568
},
{
"epoch": 3.137973137973138,
"grad_norm": 9.296796798706055,
"learning_rate": 1.3307607891860916e-06,
"loss": 0.48091375827789307,
"step": 2570
},
{
"epoch": 3.1404151404151404,
"grad_norm": 3.7448983192443848,
"learning_rate": 1.3268282861642965e-06,
"loss": 0.523597776889801,
"step": 2572
},
{
"epoch": 3.142857142857143,
"grad_norm": 2.7864246368408203,
"learning_rate": 1.3229047464747492e-06,
"loss": 0.8306266665458679,
"step": 2574
},
{
"epoch": 3.1452991452991452,
"grad_norm": 4.969945430755615,
"learning_rate": 1.3189901879405287e-06,
"loss": 0.8894456028938293,
"step": 2576
},
{
"epoch": 3.1477411477411477,
"grad_norm": 2.3445777893066406,
"learning_rate": 1.3150846283439124e-06,
"loss": 0.7661553621292114,
"step": 2578
},
{
"epoch": 3.15018315018315,
"grad_norm": 5.691545486450195,
"learning_rate": 1.3111880854263006e-06,
"loss": 0.714350163936615,
"step": 2580
},
{
"epoch": 3.1526251526251525,
"grad_norm": 10.326680183410645,
"learning_rate": 1.3073005768881332e-06,
"loss": 0.8273946046829224,
"step": 2582
},
{
"epoch": 3.155067155067155,
"grad_norm": 15.851977348327637,
"learning_rate": 1.3034221203888117e-06,
"loss": 0.32498669624328613,
"step": 2584
},
{
"epoch": 3.1575091575091574,
"grad_norm": 9.003250122070312,
"learning_rate": 1.2995527335466172e-06,
"loss": 0.5692976713180542,
"step": 2586
},
{
"epoch": 3.15995115995116,
"grad_norm": 6.924123287200928,
"learning_rate": 1.2956924339386318e-06,
"loss": 0.3022007346153259,
"step": 2588
},
{
"epoch": 3.1623931623931623,
"grad_norm": 4.58242130279541,
"learning_rate": 1.291841239100656e-06,
"loss": 0.29833054542541504,
"step": 2590
},
{
"epoch": 3.1648351648351647,
"grad_norm": 2.056873083114624,
"learning_rate": 1.287999166527133e-06,
"loss": 0.7911575436592102,
"step": 2592
},
{
"epoch": 3.167277167277167,
"grad_norm": 2.0765960216522217,
"learning_rate": 1.2841662336710662e-06,
"loss": 0.6330872774124146,
"step": 2594
},
{
"epoch": 3.1697191697191696,
"grad_norm": 6.344074726104736,
"learning_rate": 1.2803424579439411e-06,
"loss": 0.5566316246986389,
"step": 2596
},
{
"epoch": 3.172161172161172,
"grad_norm": 2.674805164337158,
"learning_rate": 1.276527856715646e-06,
"loss": 0.8628856539726257,
"step": 2598
},
{
"epoch": 3.1746031746031744,
"grad_norm": 2.659057140350342,
"learning_rate": 1.2727224473143933e-06,
"loss": 0.7332071661949158,
"step": 2600
},
{
"epoch": 3.177045177045177,
"grad_norm": 2.530466318130493,
"learning_rate": 1.2689262470266403e-06,
"loss": 0.8011347651481628,
"step": 2602
},
{
"epoch": 3.1794871794871793,
"grad_norm": 3.901975154876709,
"learning_rate": 1.2651392730970107e-06,
"loss": 0.8330449461936951,
"step": 2604
},
{
"epoch": 3.1819291819291817,
"grad_norm": 43.94447326660156,
"learning_rate": 1.2613615427282177e-06,
"loss": 0.676027774810791,
"step": 2606
},
{
"epoch": 3.1843711843711846,
"grad_norm": 3.9286386966705322,
"learning_rate": 1.2575930730809836e-06,
"loss": 0.7716042399406433,
"step": 2608
},
{
"epoch": 3.186813186813187,
"grad_norm": 3.0505211353302,
"learning_rate": 1.253833881273963e-06,
"loss": 0.6885058879852295,
"step": 2610
},
{
"epoch": 3.1892551892551895,
"grad_norm": 1.979110598564148,
"learning_rate": 1.250083984383665e-06,
"loss": 0.8547409176826477,
"step": 2612
},
{
"epoch": 3.191697191697192,
"grad_norm": 1.8880133628845215,
"learning_rate": 1.2463433994443755e-06,
"loss": 0.7033931612968445,
"step": 2614
},
{
"epoch": 3.1941391941391943,
"grad_norm": 5.2559099197387695,
"learning_rate": 1.2426121434480828e-06,
"loss": 0.5308555960655212,
"step": 2616
},
{
"epoch": 3.1965811965811968,
"grad_norm": 15.855555534362793,
"learning_rate": 1.2388902333443913e-06,
"loss": 0.5126664042472839,
"step": 2618
},
{
"epoch": 3.199023199023199,
"grad_norm": 6.698633670806885,
"learning_rate": 1.2351776860404577e-06,
"loss": 0.2311715930700302,
"step": 2620
},
{
"epoch": 3.2014652014652016,
"grad_norm": 2.60094952583313,
"learning_rate": 1.2314745184009012e-06,
"loss": 0.5172269344329834,
"step": 2622
},
{
"epoch": 3.203907203907204,
"grad_norm": 6.668694496154785,
"learning_rate": 1.2277807472477384e-06,
"loss": 0.2329966425895691,
"step": 2624
},
{
"epoch": 3.2063492063492065,
"grad_norm": 3.4587221145629883,
"learning_rate": 1.2240963893602952e-06,
"loss": 0.9453647136688232,
"step": 2626
},
{
"epoch": 3.208791208791209,
"grad_norm": 10.398512840270996,
"learning_rate": 1.2204214614751442e-06,
"loss": 0.14541763067245483,
"step": 2628
},
{
"epoch": 3.2112332112332114,
"grad_norm": 2.5785880088806152,
"learning_rate": 1.2167559802860135e-06,
"loss": 0.9672622084617615,
"step": 2630
},
{
"epoch": 3.213675213675214,
"grad_norm": 6.969937801361084,
"learning_rate": 1.213099962443726e-06,
"loss": 0.556788980960846,
"step": 2632
},
{
"epoch": 3.2161172161172162,
"grad_norm": 1.9294971227645874,
"learning_rate": 1.2094534245561098e-06,
"loss": 0.7081160545349121,
"step": 2634
},
{
"epoch": 3.2185592185592187,
"grad_norm": 3.8238325119018555,
"learning_rate": 1.205816383187935e-06,
"loss": 0.610231876373291,
"step": 2636
},
{
"epoch": 3.221001221001221,
"grad_norm": 8.861764907836914,
"learning_rate": 1.2021888548608272e-06,
"loss": 0.6349549889564514,
"step": 2638
},
{
"epoch": 3.2234432234432235,
"grad_norm": 1.7522162199020386,
"learning_rate": 1.1985708560532047e-06,
"loss": 0.7171330451965332,
"step": 2640
},
{
"epoch": 3.225885225885226,
"grad_norm": 2.3566970825195312,
"learning_rate": 1.19496240320019e-06,
"loss": 0.8892612457275391,
"step": 2642
},
{
"epoch": 3.2283272283272284,
"grad_norm": 2.1190340518951416,
"learning_rate": 1.1913635126935477e-06,
"loss": 0.7615914344787598,
"step": 2644
},
{
"epoch": 3.230769230769231,
"grad_norm": 1.7972434759140015,
"learning_rate": 1.1877742008816002e-06,
"loss": 0.7856362462043762,
"step": 2646
},
{
"epoch": 3.2332112332112333,
"grad_norm": 2.146169662475586,
"learning_rate": 1.1841944840691619e-06,
"loss": 0.8134780526161194,
"step": 2648
},
{
"epoch": 3.2356532356532357,
"grad_norm": 11.170672416687012,
"learning_rate": 1.1806243785174574e-06,
"loss": 0.25675392150878906,
"step": 2650
},
{
"epoch": 3.238095238095238,
"grad_norm": 3.0035762786865234,
"learning_rate": 1.177063900444053e-06,
"loss": 0.7596380114555359,
"step": 2652
},
{
"epoch": 3.2405372405372406,
"grad_norm": 2.9616620540618896,
"learning_rate": 1.173513066022781e-06,
"loss": 0.9850119352340698,
"step": 2654
},
{
"epoch": 3.242979242979243,
"grad_norm": 2.4156839847564697,
"learning_rate": 1.1699718913836657e-06,
"loss": 0.5272483229637146,
"step": 2656
},
{
"epoch": 3.2454212454212454,
"grad_norm": 6.468486785888672,
"learning_rate": 1.1664403926128516e-06,
"loss": 0.38367509841918945,
"step": 2658
},
{
"epoch": 3.247863247863248,
"grad_norm": 7.794209957122803,
"learning_rate": 1.1629185857525296e-06,
"loss": 0.4385049343109131,
"step": 2660
},
{
"epoch": 3.2503052503052503,
"grad_norm": 3.511357545852661,
"learning_rate": 1.1594064868008634e-06,
"loss": 1.021645188331604,
"step": 2662
},
{
"epoch": 3.2527472527472527,
"grad_norm": 3.927022695541382,
"learning_rate": 1.1559041117119183e-06,
"loss": 0.6152275800704956,
"step": 2664
},
{
"epoch": 3.255189255189255,
"grad_norm": 2.2393136024475098,
"learning_rate": 1.1524114763955876e-06,
"loss": 0.48958325386047363,
"step": 2666
},
{
"epoch": 3.2576312576312576,
"grad_norm": 4.180426597595215,
"learning_rate": 1.1489285967175207e-06,
"loss": 0.8316766619682312,
"step": 2668
},
{
"epoch": 3.26007326007326,
"grad_norm": 2.2399017810821533,
"learning_rate": 1.145455488499051e-06,
"loss": 0.8960772156715393,
"step": 2670
},
{
"epoch": 3.2625152625152625,
"grad_norm": 8.40261459350586,
"learning_rate": 1.1419921675171245e-06,
"loss": 0.48031508922576904,
"step": 2672
},
{
"epoch": 3.264957264957265,
"grad_norm": 4.309727191925049,
"learning_rate": 1.1385386495042272e-06,
"loss": 0.13968098163604736,
"step": 2674
},
{
"epoch": 3.2673992673992673,
"grad_norm": 2.9554507732391357,
"learning_rate": 1.1350949501483145e-06,
"loss": 0.909633457660675,
"step": 2676
},
{
"epoch": 3.2698412698412698,
"grad_norm": 3.3006162643432617,
"learning_rate": 1.1316610850927388e-06,
"loss": 0.7681498527526855,
"step": 2678
},
{
"epoch": 3.272283272283272,
"grad_norm": 0.4983324706554413,
"learning_rate": 1.12823706993618e-06,
"loss": 0.42390134930610657,
"step": 2680
},
{
"epoch": 3.2747252747252746,
"grad_norm": 5.958628177642822,
"learning_rate": 1.1248229202325736e-06,
"loss": 0.5118038654327393,
"step": 2682
},
{
"epoch": 3.277167277167277,
"grad_norm": 3.1510939598083496,
"learning_rate": 1.1214186514910398e-06,
"loss": 0.4903255105018616,
"step": 2684
},
{
"epoch": 3.2796092796092795,
"grad_norm": 3.030272960662842,
"learning_rate": 1.118024279175814e-06,
"loss": 0.9587437510490417,
"step": 2686
},
{
"epoch": 3.282051282051282,
"grad_norm": 0.38044917583465576,
"learning_rate": 1.1146398187061776e-06,
"loss": 0.13130967319011688,
"step": 2688
},
{
"epoch": 3.2844932844932844,
"grad_norm": 3.4803404808044434,
"learning_rate": 1.1112652854563824e-06,
"loss": 0.8778082132339478,
"step": 2690
},
{
"epoch": 3.286935286935287,
"grad_norm": 2.587324619293213,
"learning_rate": 1.10790069475559e-06,
"loss": 0.7979317903518677,
"step": 2692
},
{
"epoch": 3.2893772893772892,
"grad_norm": 0.5913499593734741,
"learning_rate": 1.1045460618877912e-06,
"loss": 0.06984243541955948,
"step": 2694
},
{
"epoch": 3.2918192918192917,
"grad_norm": 0.8156366944313049,
"learning_rate": 1.1012014020917484e-06,
"loss": 0.2010909765958786,
"step": 2696
},
{
"epoch": 3.294261294261294,
"grad_norm": 1.7881795167922974,
"learning_rate": 1.0978667305609147e-06,
"loss": 0.7086883783340454,
"step": 2698
},
{
"epoch": 3.2967032967032965,
"grad_norm": 2.393721342086792,
"learning_rate": 1.0945420624433758e-06,
"loss": 0.8595046997070312,
"step": 2700
},
{
"epoch": 3.299145299145299,
"grad_norm": 2.1565139293670654,
"learning_rate": 1.0912274128417696e-06,
"loss": 0.9005102515220642,
"step": 2702
},
{
"epoch": 3.3015873015873014,
"grad_norm": 3.553394317626953,
"learning_rate": 1.0879227968132316e-06,
"loss": 0.7882212996482849,
"step": 2704
},
{
"epoch": 3.304029304029304,
"grad_norm": 3.0672333240509033,
"learning_rate": 1.084628229369311e-06,
"loss": 0.7716812491416931,
"step": 2706
},
{
"epoch": 3.3064713064713063,
"grad_norm": 2.4524006843566895,
"learning_rate": 1.0813437254759165e-06,
"loss": 0.5214710831642151,
"step": 2708
},
{
"epoch": 3.3089133089133087,
"grad_norm": 2.5456132888793945,
"learning_rate": 1.0780693000532372e-06,
"loss": 0.6923003196716309,
"step": 2710
},
{
"epoch": 3.311355311355311,
"grad_norm": 2.900570869445801,
"learning_rate": 1.0748049679756843e-06,
"loss": 0.8628150820732117,
"step": 2712
},
{
"epoch": 3.3137973137973136,
"grad_norm": 3.343024253845215,
"learning_rate": 1.0715507440718147e-06,
"loss": 0.5487645864486694,
"step": 2714
},
{
"epoch": 3.316239316239316,
"grad_norm": 3.0560784339904785,
"learning_rate": 1.0683066431242717e-06,
"loss": 0.8819491267204285,
"step": 2716
},
{
"epoch": 3.3186813186813184,
"grad_norm": 4.643116474151611,
"learning_rate": 1.0650726798697113e-06,
"loss": 0.7186728715896606,
"step": 2718
},
{
"epoch": 3.3211233211233213,
"grad_norm": 3.3299341201782227,
"learning_rate": 1.061848868998739e-06,
"loss": 0.8435643315315247,
"step": 2720
},
{
"epoch": 3.3235653235653237,
"grad_norm": 3.4163544178009033,
"learning_rate": 1.0586352251558425e-06,
"loss": 0.5442382097244263,
"step": 2722
},
{
"epoch": 3.326007326007326,
"grad_norm": 0.8344938158988953,
"learning_rate": 1.0554317629393234e-06,
"loss": 0.06635025888681412,
"step": 2724
},
{
"epoch": 3.3284493284493286,
"grad_norm": 3.7720038890838623,
"learning_rate": 1.052238496901234e-06,
"loss": 0.4655779004096985,
"step": 2726
},
{
"epoch": 3.330891330891331,
"grad_norm": 1.7745321989059448,
"learning_rate": 1.0490554415473079e-06,
"loss": 0.7798499464988708,
"step": 2728
},
{
"epoch": 3.3333333333333335,
"grad_norm": 11.497058868408203,
"learning_rate": 1.0458826113368965e-06,
"loss": 0.7589215040206909,
"step": 2730
},
{
"epoch": 3.335775335775336,
"grad_norm": 0.8510182499885559,
"learning_rate": 1.042720020682902e-06,
"loss": 0.5679793953895569,
"step": 2732
},
{
"epoch": 3.3382173382173383,
"grad_norm": 3.5748512744903564,
"learning_rate": 1.039567683951713e-06,
"loss": 0.6550238132476807,
"step": 2734
},
{
"epoch": 3.340659340659341,
"grad_norm": 1.754301905632019,
"learning_rate": 1.036425615463138e-06,
"loss": 0.7326245903968811,
"step": 2736
},
{
"epoch": 3.343101343101343,
"grad_norm": 0.8980134129524231,
"learning_rate": 1.0332938294903407e-06,
"loss": 0.4887623190879822,
"step": 2738
},
{
"epoch": 3.3455433455433456,
"grad_norm": 1.9559557437896729,
"learning_rate": 1.0301723402597769e-06,
"loss": 0.4706956148147583,
"step": 2740
},
{
"epoch": 3.347985347985348,
"grad_norm": 10.616726875305176,
"learning_rate": 1.0270611619511266e-06,
"loss": 0.8596802353858948,
"step": 2742
},
{
"epoch": 3.3504273504273505,
"grad_norm": 24.7109317779541,
"learning_rate": 1.0239603086972323e-06,
"loss": 0.584933340549469,
"step": 2744
},
{
"epoch": 3.352869352869353,
"grad_norm": 2.8988747596740723,
"learning_rate": 1.020869794584034e-06,
"loss": 0.7101993560791016,
"step": 2746
},
{
"epoch": 3.3553113553113554,
"grad_norm": 93.7816162109375,
"learning_rate": 1.017789633650505e-06,
"loss": 0.5839419960975647,
"step": 2748
},
{
"epoch": 3.357753357753358,
"grad_norm": 5.284997463226318,
"learning_rate": 1.014719839888588e-06,
"loss": 0.3563012480735779,
"step": 2750
},
{
"epoch": 3.3601953601953602,
"grad_norm": 1.8024640083312988,
"learning_rate": 1.011660427243132e-06,
"loss": 0.3267917037010193,
"step": 2752
},
{
"epoch": 3.3626373626373627,
"grad_norm": 1.5629299879074097,
"learning_rate": 1.0086114096118286e-06,
"loss": 0.4166877269744873,
"step": 2754
},
{
"epoch": 3.365079365079365,
"grad_norm": 3.879136323928833,
"learning_rate": 1.005572800845148e-06,
"loss": 0.7401219606399536,
"step": 2756
},
{
"epoch": 3.3675213675213675,
"grad_norm": 2.1835029125213623,
"learning_rate": 1.0025446147462803e-06,
"loss": 0.8265122175216675,
"step": 2758
},
{
"epoch": 3.36996336996337,
"grad_norm": 4.126181602478027,
"learning_rate": 9.995268650710657e-07,
"loss": 1.0622559785842896,
"step": 2760
},
{
"epoch": 3.3724053724053724,
"grad_norm": 2.4242002964019775,
"learning_rate": 9.965195655279379e-07,
"loss": 0.4468766152858734,
"step": 2762
},
{
"epoch": 3.374847374847375,
"grad_norm": 2.549320697784424,
"learning_rate": 9.935227297778605e-07,
"loss": 0.8873687386512756,
"step": 2764
},
{
"epoch": 3.3772893772893773,
"grad_norm": 3.020458936691284,
"learning_rate": 9.905363714342615e-07,
"loss": 0.4191344082355499,
"step": 2766
},
{
"epoch": 3.3797313797313797,
"grad_norm": 2.900737762451172,
"learning_rate": 9.875605040629773e-07,
"loss": 0.37247806787490845,
"step": 2768
},
{
"epoch": 3.382173382173382,
"grad_norm": 5.969982624053955,
"learning_rate": 9.845951411821863e-07,
"loss": 0.44880709052085876,
"step": 2770
},
{
"epoch": 3.3846153846153846,
"grad_norm": 5.617712497711182,
"learning_rate": 9.816402962623497e-07,
"loss": 0.8781357407569885,
"step": 2772
},
{
"epoch": 3.387057387057387,
"grad_norm": 4.446506023406982,
"learning_rate": 9.786959827261491e-07,
"loss": 0.4436875283718109,
"step": 2774
},
{
"epoch": 3.3894993894993894,
"grad_norm": 2.733320713043213,
"learning_rate": 9.757622139484275e-07,
"loss": 0.8267475962638855,
"step": 2776
},
{
"epoch": 3.391941391941392,
"grad_norm": 1.7333195209503174,
"learning_rate": 9.72839003256126e-07,
"loss": 0.741727352142334,
"step": 2778
},
{
"epoch": 3.3943833943833943,
"grad_norm": 0.6535694599151611,
"learning_rate": 9.699263639282255e-07,
"loss": 0.3470302224159241,
"step": 2780
},
{
"epoch": 3.3968253968253967,
"grad_norm": 2.3038716316223145,
"learning_rate": 9.67024309195685e-07,
"loss": 0.538809061050415,
"step": 2782
},
{
"epoch": 3.399267399267399,
"grad_norm": 2.897294044494629,
"learning_rate": 9.641328522413813e-07,
"loss": 0.20831730961799622,
"step": 2784
},
{
"epoch": 3.4017094017094016,
"grad_norm": 5.283458709716797,
"learning_rate": 9.61252006200051e-07,
"loss": 0.4685713052749634,
"step": 2786
},
{
"epoch": 3.404151404151404,
"grad_norm": 2.1215171813964844,
"learning_rate": 9.583817841582285e-07,
"loss": 0.9289071559906006,
"step": 2788
},
{
"epoch": 3.4065934065934065,
"grad_norm": 2.2103238105773926,
"learning_rate": 9.555221991541877e-07,
"loss": 0.814532995223999,
"step": 2790
},
{
"epoch": 3.409035409035409,
"grad_norm": 1.8449485301971436,
"learning_rate": 9.526732641778836e-07,
"loss": 0.4528883099555969,
"step": 2792
},
{
"epoch": 3.4114774114774113,
"grad_norm": 1.9408596754074097,
"learning_rate": 9.498349921708914e-07,
"loss": 0.7237257957458496,
"step": 2794
},
{
"epoch": 3.413919413919414,
"grad_norm": 2.033916711807251,
"learning_rate": 9.470073960263484e-07,
"loss": 0.8568973541259766,
"step": 2796
},
{
"epoch": 3.416361416361416,
"grad_norm": 2.8969850540161133,
"learning_rate": 9.441904885888982e-07,
"loss": 0.4384755790233612,
"step": 2798
},
{
"epoch": 3.4188034188034186,
"grad_norm": 0.7930925488471985,
"learning_rate": 9.413842826546254e-07,
"loss": 0.06476318836212158,
"step": 2800
},
{
"epoch": 3.421245421245421,
"grad_norm": 2.1533210277557373,
"learning_rate": 9.385887909710069e-07,
"loss": 0.825805127620697,
"step": 2802
},
{
"epoch": 3.4236874236874235,
"grad_norm": 0.8457545042037964,
"learning_rate": 9.358040262368445e-07,
"loss": 0.6605730652809143,
"step": 2804
},
{
"epoch": 3.426129426129426,
"grad_norm": 2.488839626312256,
"learning_rate": 9.330300011022163e-07,
"loss": 0.5794928669929504,
"step": 2806
},
{
"epoch": 3.4285714285714284,
"grad_norm": 2.972033739089966,
"learning_rate": 9.302667281684098e-07,
"loss": 0.8902238011360168,
"step": 2808
},
{
"epoch": 3.4310134310134313,
"grad_norm": 3.6812944412231445,
"learning_rate": 9.27514219987874e-07,
"loss": 0.6202410459518433,
"step": 2810
},
{
"epoch": 3.4334554334554337,
"grad_norm": 1.0500991344451904,
"learning_rate": 9.247724890641543e-07,
"loss": 0.10720151662826538,
"step": 2812
},
{
"epoch": 3.435897435897436,
"grad_norm": 7.567105293273926,
"learning_rate": 9.22041547851843e-07,
"loss": 0.6870253682136536,
"step": 2814
},
{
"epoch": 3.4383394383394386,
"grad_norm": 21.982929229736328,
"learning_rate": 9.193214087565146e-07,
"loss": 0.7681679725646973,
"step": 2816
},
{
"epoch": 3.440781440781441,
"grad_norm": 3.4593942165374756,
"learning_rate": 9.166120841346786e-07,
"loss": 0.9266019463539124,
"step": 2818
},
{
"epoch": 3.4432234432234434,
"grad_norm": 3.5219388008117676,
"learning_rate": 9.139135862937139e-07,
"loss": 0.7612605094909668,
"step": 2820
},
{
"epoch": 3.445665445665446,
"grad_norm": 4.131385803222656,
"learning_rate": 9.112259274918228e-07,
"loss": 0.9317520260810852,
"step": 2822
},
{
"epoch": 3.4481074481074483,
"grad_norm": 2.044438600540161,
"learning_rate": 9.085491199379652e-07,
"loss": 0.8517608046531677,
"step": 2824
},
{
"epoch": 3.4505494505494507,
"grad_norm": 2.4360508918762207,
"learning_rate": 9.058831757918119e-07,
"loss": 0.6897678971290588,
"step": 2826
},
{
"epoch": 3.452991452991453,
"grad_norm": 2.0640690326690674,
"learning_rate": 9.032281071636829e-07,
"loss": 0.8173401355743408,
"step": 2828
},
{
"epoch": 3.4554334554334556,
"grad_norm": 16.34990692138672,
"learning_rate": 9.00583926114497e-07,
"loss": 0.5656797289848328,
"step": 2830
},
{
"epoch": 3.457875457875458,
"grad_norm": 3.1184754371643066,
"learning_rate": 8.979506446557141e-07,
"loss": 0.49873775243759155,
"step": 2832
},
{
"epoch": 3.4603174603174605,
"grad_norm": 1.7745251655578613,
"learning_rate": 8.953282747492816e-07,
"loss": 0.9596657752990723,
"step": 2834
},
{
"epoch": 3.462759462759463,
"grad_norm": 7.131638050079346,
"learning_rate": 8.927168283075807e-07,
"loss": 0.5169793367385864,
"step": 2836
},
{
"epoch": 3.4652014652014653,
"grad_norm": 101.0040283203125,
"learning_rate": 8.901163171933706e-07,
"loss": 0.4887714684009552,
"step": 2838
},
{
"epoch": 3.4676434676434678,
"grad_norm": 2.2968623638153076,
"learning_rate": 8.875267532197368e-07,
"loss": 0.8718173503875732,
"step": 2840
},
{
"epoch": 3.47008547008547,
"grad_norm": 1.8435990810394287,
"learning_rate": 8.849481481500356e-07,
"loss": 0.7854589819908142,
"step": 2842
},
{
"epoch": 3.4725274725274726,
"grad_norm": 2.054354667663574,
"learning_rate": 8.823805136978413e-07,
"loss": 0.8386743068695068,
"step": 2844
},
{
"epoch": 3.474969474969475,
"grad_norm": 1.7228000164031982,
"learning_rate": 8.798238615268934e-07,
"loss": 0.8005945682525635,
"step": 2846
},
{
"epoch": 3.4774114774114775,
"grad_norm": 2.3333404064178467,
"learning_rate": 8.772782032510439e-07,
"loss": 0.7422125935554504,
"step": 2848
},
{
"epoch": 3.47985347985348,
"grad_norm": 6.355716228485107,
"learning_rate": 8.747435504342022e-07,
"loss": 0.5096930861473083,
"step": 2850
},
{
"epoch": 3.4822954822954824,
"grad_norm": 1.6502275466918945,
"learning_rate": 8.722199145902857e-07,
"loss": 0.7535804510116577,
"step": 2852
},
{
"epoch": 3.484737484737485,
"grad_norm": 0.4867725968360901,
"learning_rate": 8.697073071831658e-07,
"loss": 0.14731785655021667,
"step": 2854
},
{
"epoch": 3.4871794871794872,
"grad_norm": 7.8371686935424805,
"learning_rate": 8.672057396266161e-07,
"loss": 0.7259136438369751,
"step": 2856
},
{
"epoch": 3.4896214896214897,
"grad_norm": 4.284936904907227,
"learning_rate": 8.647152232842599e-07,
"loss": 0.30577749013900757,
"step": 2858
},
{
"epoch": 3.492063492063492,
"grad_norm": 1.8026747703552246,
"learning_rate": 8.622357694695207e-07,
"loss": 0.7950439453125,
"step": 2860
},
{
"epoch": 3.4945054945054945,
"grad_norm": 10.06855583190918,
"learning_rate": 8.597673894455675e-07,
"loss": 0.7480056881904602,
"step": 2862
},
{
"epoch": 3.496947496947497,
"grad_norm": 1.7933481931686401,
"learning_rate": 8.573100944252662e-07,
"loss": 0.4216347336769104,
"step": 2864
},
{
"epoch": 3.4993894993894994,
"grad_norm": 1.7255240678787231,
"learning_rate": 8.548638955711298e-07,
"loss": 0.3292645812034607,
"step": 2866
},
{
"epoch": 3.501831501831502,
"grad_norm": 3.3586931228637695,
"learning_rate": 8.524288039952625e-07,
"loss": 0.8613098859786987,
"step": 2868
},
{
"epoch": 3.5042735042735043,
"grad_norm": 7.2187700271606445,
"learning_rate": 8.500048307593161e-07,
"loss": 0.49458879232406616,
"step": 2870
},
{
"epoch": 3.5067155067155067,
"grad_norm": 13.894682884216309,
"learning_rate": 8.475919868744322e-07,
"loss": 0.39427265524864197,
"step": 2872
},
{
"epoch": 3.509157509157509,
"grad_norm": 2.581885814666748,
"learning_rate": 8.451902833012003e-07,
"loss": 0.918152391910553,
"step": 2874
},
{
"epoch": 3.5115995115995116,
"grad_norm": 3.96673321723938,
"learning_rate": 8.427997309496005e-07,
"loss": 0.30580323934555054,
"step": 2876
},
{
"epoch": 3.514041514041514,
"grad_norm": 3.975219964981079,
"learning_rate": 8.404203406789602e-07,
"loss": 0.7915468215942383,
"step": 2878
},
{
"epoch": 3.5164835164835164,
"grad_norm": 2.8507955074310303,
"learning_rate": 8.380521232978992e-07,
"loss": 0.6845412254333496,
"step": 2880
},
{
"epoch": 3.518925518925519,
"grad_norm": 5.622760772705078,
"learning_rate": 8.356950895642862e-07,
"loss": 0.6595726013183594,
"step": 2882
},
{
"epoch": 3.5213675213675213,
"grad_norm": 4.254746437072754,
"learning_rate": 8.333492501851838e-07,
"loss": 0.9753207564353943,
"step": 2884
},
{
"epoch": 3.5238095238095237,
"grad_norm": 2.7020394802093506,
"learning_rate": 8.310146158168066e-07,
"loss": 0.9157518148422241,
"step": 2886
},
{
"epoch": 3.526251526251526,
"grad_norm": 2.108231544494629,
"learning_rate": 8.28691197064466e-07,
"loss": 0.7969650030136108,
"step": 2888
},
{
"epoch": 3.5286935286935286,
"grad_norm": 3.530876398086548,
"learning_rate": 8.263790044825285e-07,
"loss": 0.9127578735351562,
"step": 2890
},
{
"epoch": 3.531135531135531,
"grad_norm": 2.557037830352783,
"learning_rate": 8.240780485743608e-07,
"loss": 0.7240651845932007,
"step": 2892
},
{
"epoch": 3.5335775335775335,
"grad_norm": 6.49147891998291,
"learning_rate": 8.217883397922905e-07,
"loss": 0.3955632746219635,
"step": 2894
},
{
"epoch": 3.536019536019536,
"grad_norm": 2.565272808074951,
"learning_rate": 8.195098885375495e-07,
"loss": 0.6618751287460327,
"step": 2896
},
{
"epoch": 3.5384615384615383,
"grad_norm": 1.6365941762924194,
"learning_rate": 8.172427051602347e-07,
"loss": 0.7428027391433716,
"step": 2898
},
{
"epoch": 3.5409035409035408,
"grad_norm": 2.283928394317627,
"learning_rate": 8.149867999592548e-07,
"loss": 0.7936429381370544,
"step": 2900
},
{
"epoch": 3.543345543345543,
"grad_norm": 3.6295547485351562,
"learning_rate": 8.127421831822881e-07,
"loss": 0.12481510639190674,
"step": 2902
},
{
"epoch": 3.5457875457875456,
"grad_norm": 1.7329978942871094,
"learning_rate": 8.105088650257343e-07,
"loss": 0.7854657769203186,
"step": 2904
},
{
"epoch": 3.548229548229548,
"grad_norm": 3.5314977169036865,
"learning_rate": 8.082868556346662e-07,
"loss": 0.8098005056381226,
"step": 2906
},
{
"epoch": 3.5506715506715505,
"grad_norm": 2.5535361766815186,
"learning_rate": 8.060761651027867e-07,
"loss": 0.9487941265106201,
"step": 2908
},
{
"epoch": 3.553113553113553,
"grad_norm": 3.6165003776550293,
"learning_rate": 8.038768034723816e-07,
"loss": 0.52418053150177,
"step": 2910
},
{
"epoch": 3.5555555555555554,
"grad_norm": 2.5802009105682373,
"learning_rate": 8.016887807342732e-07,
"loss": 0.3970736563205719,
"step": 2912
},
{
"epoch": 3.557997557997558,
"grad_norm": 2.033656358718872,
"learning_rate": 7.995121068277767e-07,
"loss": 0.591442883014679,
"step": 2914
},
{
"epoch": 3.5604395604395602,
"grad_norm": 1.6943453550338745,
"learning_rate": 7.973467916406536e-07,
"loss": 0.8024305701255798,
"step": 2916
},
{
"epoch": 3.5628815628815627,
"grad_norm": 7.8730692863464355,
"learning_rate": 7.951928450090673e-07,
"loss": 0.6496803164482117,
"step": 2918
},
{
"epoch": 3.565323565323565,
"grad_norm": 18.40151596069336,
"learning_rate": 7.93050276717538e-07,
"loss": 0.5158238410949707,
"step": 2920
},
{
"epoch": 3.5677655677655675,
"grad_norm": 1.7893723249435425,
"learning_rate": 7.909190964988997e-07,
"loss": 0.826363742351532,
"step": 2922
},
{
"epoch": 3.57020757020757,
"grad_norm": 9.468801498413086,
"learning_rate": 7.887993140342538e-07,
"loss": 0.5767655968666077,
"step": 2924
},
{
"epoch": 3.5726495726495724,
"grad_norm": 2.3637073040008545,
"learning_rate": 7.86690938952927e-07,
"loss": 0.48466330766677856,
"step": 2926
},
{
"epoch": 3.575091575091575,
"grad_norm": 2.659562587738037,
"learning_rate": 7.845939808324267e-07,
"loss": 0.8443353772163391,
"step": 2928
},
{
"epoch": 3.5775335775335773,
"grad_norm": 2.3245725631713867,
"learning_rate": 7.825084491983969e-07,
"loss": 0.764642059803009,
"step": 2930
},
{
"epoch": 3.57997557997558,
"grad_norm": 5.520932197570801,
"learning_rate": 7.804343535245766e-07,
"loss": 0.6558569669723511,
"step": 2932
},
{
"epoch": 3.5824175824175826,
"grad_norm": 3.9519784450531006,
"learning_rate": 7.78371703232755e-07,
"loss": 0.6942023634910583,
"step": 2934
},
{
"epoch": 3.584859584859585,
"grad_norm": 1.9188674688339233,
"learning_rate": 7.763205076927297e-07,
"loss": 0.46283501386642456,
"step": 2936
},
{
"epoch": 3.5873015873015874,
"grad_norm": 0.12963134050369263,
"learning_rate": 7.742807762222653e-07,
"loss": 0.4600965678691864,
"step": 2938
},
{
"epoch": 3.58974358974359,
"grad_norm": 2.5248825550079346,
"learning_rate": 7.722525180870466e-07,
"loss": 0.557766854763031,
"step": 2940
},
{
"epoch": 3.5921855921855923,
"grad_norm": 2.4224584102630615,
"learning_rate": 7.702357425006438e-07,
"loss": 0.8053780794143677,
"step": 2942
},
{
"epoch": 3.5946275946275947,
"grad_norm": 2.104917526245117,
"learning_rate": 7.682304586244618e-07,
"loss": 0.7758908271789551,
"step": 2944
},
{
"epoch": 3.597069597069597,
"grad_norm": 3.0547473430633545,
"learning_rate": 7.662366755677073e-07,
"loss": 0.8851046562194824,
"step": 2946
},
{
"epoch": 3.5995115995115996,
"grad_norm": 1.9801969528198242,
"learning_rate": 7.642544023873402e-07,
"loss": 0.5296459794044495,
"step": 2948
},
{
"epoch": 3.601953601953602,
"grad_norm": 3.0353381633758545,
"learning_rate": 7.622836480880383e-07,
"loss": 0.5360254049301147,
"step": 2950
},
{
"epoch": 3.6043956043956045,
"grad_norm": 8.00523567199707,
"learning_rate": 7.603244216221524e-07,
"loss": 0.20186254382133484,
"step": 2952
},
{
"epoch": 3.606837606837607,
"grad_norm": 4.944789409637451,
"learning_rate": 7.583767318896664e-07,
"loss": 0.466735303401947,
"step": 2954
},
{
"epoch": 3.6092796092796093,
"grad_norm": 5.835378646850586,
"learning_rate": 7.564405877381587e-07,
"loss": 0.6394435167312622,
"step": 2956
},
{
"epoch": 3.6117216117216118,
"grad_norm": 2.267725944519043,
"learning_rate": 7.545159979627594e-07,
"loss": 0.6844528913497925,
"step": 2958
},
{
"epoch": 3.614163614163614,
"grad_norm": 2.5582151412963867,
"learning_rate": 7.52602971306113e-07,
"loss": 0.7534583210945129,
"step": 2960
},
{
"epoch": 3.6166056166056166,
"grad_norm": 5.141064167022705,
"learning_rate": 7.507015164583367e-07,
"loss": 0.5010976791381836,
"step": 2962
},
{
"epoch": 3.619047619047619,
"grad_norm": 3.67368745803833,
"learning_rate": 7.488116420569816e-07,
"loss": 0.8287992477416992,
"step": 2964
},
{
"epoch": 3.6214896214896215,
"grad_norm": 2.036011219024658,
"learning_rate": 7.46933356686993e-07,
"loss": 0.9010288715362549,
"step": 2966
},
{
"epoch": 3.623931623931624,
"grad_norm": 2.759631633758545,
"learning_rate": 7.450666688806727e-07,
"loss": 0.9595599174499512,
"step": 2968
},
{
"epoch": 3.6263736263736264,
"grad_norm": 1.9032822847366333,
"learning_rate": 7.432115871176391e-07,
"loss": 0.4801540672779083,
"step": 2970
},
{
"epoch": 3.628815628815629,
"grad_norm": 2.809022903442383,
"learning_rate": 7.413681198247886e-07,
"loss": 0.4220638573169708,
"step": 2972
},
{
"epoch": 3.6312576312576312,
"grad_norm": 3.1490209102630615,
"learning_rate": 7.395362753762583e-07,
"loss": 0.48024827241897583,
"step": 2974
},
{
"epoch": 3.6336996336996337,
"grad_norm": 2.781191110610962,
"learning_rate": 7.37716062093387e-07,
"loss": 0.7418580055236816,
"step": 2976
},
{
"epoch": 3.636141636141636,
"grad_norm": 7.007016658782959,
"learning_rate": 7.359074882446775e-07,
"loss": 0.3912465572357178,
"step": 2978
},
{
"epoch": 3.6385836385836385,
"grad_norm": 1.995826005935669,
"learning_rate": 7.341105620457597e-07,
"loss": 0.9312414526939392,
"step": 2980
},
{
"epoch": 3.641025641025641,
"grad_norm": 4.350949287414551,
"learning_rate": 7.32325291659353e-07,
"loss": 0.891286313533783,
"step": 2982
},
{
"epoch": 3.6434676434676434,
"grad_norm": 2.1877682209014893,
"learning_rate": 7.305516851952288e-07,
"loss": 0.837358832359314,
"step": 2984
},
{
"epoch": 3.645909645909646,
"grad_norm": 1.86771559715271,
"learning_rate": 7.287897507101736e-07,
"loss": 0.9316425919532776,
"step": 2986
},
{
"epoch": 3.6483516483516483,
"grad_norm": 2.0179667472839355,
"learning_rate": 7.270394962079537e-07,
"loss": 0.8129684925079346,
"step": 2988
},
{
"epoch": 3.6507936507936507,
"grad_norm": 5.377006530761719,
"learning_rate": 7.253009296392767e-07,
"loss": 0.902230978012085,
"step": 2990
},
{
"epoch": 3.653235653235653,
"grad_norm": 2.3976590633392334,
"learning_rate": 7.235740589017581e-07,
"loss": 0.4519374966621399,
"step": 2992
},
{
"epoch": 3.6556776556776556,
"grad_norm": 3.464656114578247,
"learning_rate": 7.218588918398821e-07,
"loss": 0.3994402587413788,
"step": 2994
},
{
"epoch": 3.658119658119658,
"grad_norm": 1.6983356475830078,
"learning_rate": 7.201554362449699e-07,
"loss": 0.8617055416107178,
"step": 2996
},
{
"epoch": 3.6605616605616604,
"grad_norm": 1.0819239616394043,
"learning_rate": 7.184636998551395e-07,
"loss": 0.44496503472328186,
"step": 2998
},
{
"epoch": 3.663003663003663,
"grad_norm": 8.184094429016113,
"learning_rate": 7.167836903552761e-07,
"loss": 0.2283964604139328,
"step": 3000
},
{
"epoch": 3.6654456654456653,
"grad_norm": 2.11317777633667,
"learning_rate": 7.151154153769916e-07,
"loss": 0.3644329309463501,
"step": 3002
},
{
"epoch": 3.6678876678876677,
"grad_norm": 5.163084506988525,
"learning_rate": 7.134588824985951e-07,
"loss": 0.2088758945465088,
"step": 3004
},
{
"epoch": 3.67032967032967,
"grad_norm": 1.7118715047836304,
"learning_rate": 7.118140992450537e-07,
"loss": 0.9201338887214661,
"step": 3006
},
{
"epoch": 3.672771672771673,
"grad_norm": 6.890610694885254,
"learning_rate": 7.101810730879629e-07,
"loss": 0.5677740573883057,
"step": 3008
},
{
"epoch": 3.6752136752136755,
"grad_norm": 0.9095839858055115,
"learning_rate": 7.085598114455092e-07,
"loss": 0.4820370078086853,
"step": 3010
},
{
"epoch": 3.677655677655678,
"grad_norm": 1.6080018281936646,
"learning_rate": 7.069503216824375e-07,
"loss": 0.8840917348861694,
"step": 3012
},
{
"epoch": 3.6800976800976803,
"grad_norm": 2.5662214756011963,
"learning_rate": 7.053526111100178e-07,
"loss": 0.8205640316009521,
"step": 3014
},
{
"epoch": 3.682539682539683,
"grad_norm": 7.3714470863342285,
"learning_rate": 7.037666869860122e-07,
"loss": 0.6835364699363708,
"step": 3016
},
{
"epoch": 3.684981684981685,
"grad_norm": 1.6335556507110596,
"learning_rate": 7.021925565146419e-07,
"loss": 0.48589709401130676,
"step": 3018
},
{
"epoch": 3.6874236874236876,
"grad_norm": 3.9356753826141357,
"learning_rate": 7.006302268465534e-07,
"loss": 0.47200772166252136,
"step": 3020
},
{
"epoch": 3.68986568986569,
"grad_norm": 1.7010576725006104,
"learning_rate": 6.990797050787876e-07,
"loss": 0.4901275038719177,
"step": 3022
},
{
"epoch": 3.6923076923076925,
"grad_norm": 3.1634457111358643,
"learning_rate": 6.975409982547468e-07,
"loss": 0.982101559638977,
"step": 3024
},
{
"epoch": 3.694749694749695,
"grad_norm": 0.936671793460846,
"learning_rate": 6.960141133641622e-07,
"loss": 0.5470721125602722,
"step": 3026
},
{
"epoch": 3.6971916971916974,
"grad_norm": 5.059283256530762,
"learning_rate": 6.94499057343063e-07,
"loss": 0.5916606783866882,
"step": 3028
},
{
"epoch": 3.6996336996337,
"grad_norm": 1.649591088294983,
"learning_rate": 6.929958370737447e-07,
"loss": 0.9363369941711426,
"step": 3030
},
{
"epoch": 3.7020757020757022,
"grad_norm": 3.894197702407837,
"learning_rate": 6.915044593847377e-07,
"loss": 0.7085144519805908,
"step": 3032
},
{
"epoch": 3.7045177045177047,
"grad_norm": 2.26426100730896,
"learning_rate": 6.900249310507757e-07,
"loss": 0.5215721726417542,
"step": 3034
},
{
"epoch": 3.706959706959707,
"grad_norm": 1.1034021377563477,
"learning_rate": 6.885572587927659e-07,
"loss": 0.5011436939239502,
"step": 3036
},
{
"epoch": 3.7094017094017095,
"grad_norm": 1.3475278615951538,
"learning_rate": 6.871014492777585e-07,
"loss": 0.504115641117096,
"step": 3038
},
{
"epoch": 3.711843711843712,
"grad_norm": 6.192190647125244,
"learning_rate": 6.856575091189154e-07,
"loss": 0.9073065519332886,
"step": 3040
},
{
"epoch": 3.7142857142857144,
"grad_norm": 2.2396318912506104,
"learning_rate": 6.842254448754811e-07,
"loss": 0.7537841796875,
"step": 3042
},
{
"epoch": 3.716727716727717,
"grad_norm": 1.6438734531402588,
"learning_rate": 6.828052630527522e-07,
"loss": 0.5031915307044983,
"step": 3044
},
{
"epoch": 3.7191697191697193,
"grad_norm": 1.9333980083465576,
"learning_rate": 6.813969701020478e-07,
"loss": 0.9249406456947327,
"step": 3046
},
{
"epoch": 3.7216117216117217,
"grad_norm": 2.5807371139526367,
"learning_rate": 6.800005724206831e-07,
"loss": 0.5361707806587219,
"step": 3048
},
{
"epoch": 3.724053724053724,
"grad_norm": 2.0083351135253906,
"learning_rate": 6.786160763519335e-07,
"loss": 0.741145133972168,
"step": 3050
},
{
"epoch": 3.7264957264957266,
"grad_norm": 2.013129711151123,
"learning_rate": 6.772434881850144e-07,
"loss": 0.8314730525016785,
"step": 3052
},
{
"epoch": 3.728937728937729,
"grad_norm": 4.733554840087891,
"learning_rate": 6.75882814155045e-07,
"loss": 0.4873526394367218,
"step": 3054
},
{
"epoch": 3.7313797313797314,
"grad_norm": 3.0452795028686523,
"learning_rate": 6.745340604430266e-07,
"loss": 0.9440727233886719,
"step": 3056
},
{
"epoch": 3.733821733821734,
"grad_norm": 2.133902072906494,
"learning_rate": 6.731972331758076e-07,
"loss": 0.6994129419326782,
"step": 3058
},
{
"epoch": 3.7362637362637363,
"grad_norm": 7.259641647338867,
"learning_rate": 6.718723384260628e-07,
"loss": 0.43192028999328613,
"step": 3060
},
{
"epoch": 3.7387057387057387,
"grad_norm": 2.170210599899292,
"learning_rate": 6.705593822122592e-07,
"loss": 0.46401798725128174,
"step": 3062
},
{
"epoch": 3.741147741147741,
"grad_norm": 1.0304683446884155,
"learning_rate": 6.692583704986346e-07,
"loss": 0.3507814407348633,
"step": 3064
},
{
"epoch": 3.7435897435897436,
"grad_norm": 3.386765718460083,
"learning_rate": 6.679693091951654e-07,
"loss": 0.5371965169906616,
"step": 3066
},
{
"epoch": 3.746031746031746,
"grad_norm": 1.7468204498291016,
"learning_rate": 6.666922041575441e-07,
"loss": 0.8100628852844238,
"step": 3068
},
{
"epoch": 3.7484737484737485,
"grad_norm": 2.1092400550842285,
"learning_rate": 6.654270611871487e-07,
"loss": 0.8633115887641907,
"step": 3070
},
{
"epoch": 3.750915750915751,
"grad_norm": 5.546787261962891,
"learning_rate": 6.641738860310198e-07,
"loss": 0.5384469628334045,
"step": 3072
},
{
"epoch": 3.7533577533577533,
"grad_norm": 1.0326019525527954,
"learning_rate": 6.629326843818317e-07,
"loss": 0.45187580585479736,
"step": 3074
},
{
"epoch": 3.755799755799756,
"grad_norm": 4.467280864715576,
"learning_rate": 6.617034618778696e-07,
"loss": 0.855986475944519,
"step": 3076
},
{
"epoch": 3.758241758241758,
"grad_norm": 3.7619566917419434,
"learning_rate": 6.60486224103e-07,
"loss": 0.7572266459465027,
"step": 3078
},
{
"epoch": 3.7606837606837606,
"grad_norm": 6.183402061462402,
"learning_rate": 6.592809765866497e-07,
"loss": 0.527399480342865,
"step": 3080
},
{
"epoch": 3.763125763125763,
"grad_norm": 17.207387924194336,
"learning_rate": 6.580877248037769e-07,
"loss": 0.8154140114784241,
"step": 3082
},
{
"epoch": 3.7655677655677655,
"grad_norm": 1.9882452487945557,
"learning_rate": 6.569064741748489e-07,
"loss": 0.911526083946228,
"step": 3084
},
{
"epoch": 3.768009768009768,
"grad_norm": 4.858429908752441,
"learning_rate": 6.557372300658159e-07,
"loss": 0.9475221633911133,
"step": 3086
},
{
"epoch": 3.7704517704517704,
"grad_norm": 4.474874019622803,
"learning_rate": 6.545799977880883e-07,
"loss": 0.23142358660697937,
"step": 3088
},
{
"epoch": 3.772893772893773,
"grad_norm": 3.1783666610717773,
"learning_rate": 6.534347825985101e-07,
"loss": 0.8077917098999023,
"step": 3090
},
{
"epoch": 3.7753357753357752,
"grad_norm": 3.9100658893585205,
"learning_rate": 6.523015896993381e-07,
"loss": 0.8263789415359497,
"step": 3092
},
{
"epoch": 3.7777777777777777,
"grad_norm": 1.6414414644241333,
"learning_rate": 6.511804242382148e-07,
"loss": 0.927480936050415,
"step": 3094
},
{
"epoch": 3.78021978021978,
"grad_norm": 2.777785301208496,
"learning_rate": 6.500712913081486e-07,
"loss": 0.7517139315605164,
"step": 3096
},
{
"epoch": 3.7826617826617825,
"grad_norm": 3.0369672775268555,
"learning_rate": 6.489741959474876e-07,
"loss": 0.5767747163772583,
"step": 3098
},
{
"epoch": 3.785103785103785,
"grad_norm": 30.473583221435547,
"learning_rate": 6.478891431398988e-07,
"loss": 0.7141063213348389,
"step": 3100
},
{
"epoch": 3.7875457875457874,
"grad_norm": 1.8245930671691895,
"learning_rate": 6.468161378143442e-07,
"loss": 0.8820724487304688,
"step": 3102
},
{
"epoch": 3.78998778998779,
"grad_norm": 3.724031448364258,
"learning_rate": 6.457551848450591e-07,
"loss": 0.8075670599937439,
"step": 3104
},
{
"epoch": 3.7924297924297923,
"grad_norm": 2.1518399715423584,
"learning_rate": 6.447062890515303e-07,
"loss": 0.8505239486694336,
"step": 3106
},
{
"epoch": 3.7948717948717947,
"grad_norm": 10.370966911315918,
"learning_rate": 6.436694551984721e-07,
"loss": 0.4985530376434326,
"step": 3108
},
{
"epoch": 3.797313797313797,
"grad_norm": 3.9164273738861084,
"learning_rate": 6.426446879958085e-07,
"loss": 0.5315531492233276,
"step": 3110
},
{
"epoch": 3.7997557997557996,
"grad_norm": 3.175743818283081,
"learning_rate": 6.416319920986471e-07,
"loss": 0.5992456674575806,
"step": 3112
},
{
"epoch": 3.802197802197802,
"grad_norm": 3.047924518585205,
"learning_rate": 6.406313721072623e-07,
"loss": 0.536521315574646,
"step": 3114
},
{
"epoch": 3.8046398046398044,
"grad_norm": 8.903552055358887,
"learning_rate": 6.396428325670712e-07,
"loss": 0.34863823652267456,
"step": 3116
},
{
"epoch": 3.807081807081807,
"grad_norm": 2.5433661937713623,
"learning_rate": 6.386663779686147e-07,
"loss": 0.5361265540122986,
"step": 3118
},
{
"epoch": 3.8095238095238093,
"grad_norm": 7.335467338562012,
"learning_rate": 6.377020127475377e-07,
"loss": 0.9320744872093201,
"step": 3120
},
{
"epoch": 3.8119658119658117,
"grad_norm": 40.987064361572266,
"learning_rate": 6.367497412845654e-07,
"loss": 0.6013365387916565,
"step": 3122
},
{
"epoch": 3.814407814407814,
"grad_norm": 2.2245728969573975,
"learning_rate": 6.358095679054882e-07,
"loss": 0.6405077576637268,
"step": 3124
},
{
"epoch": 3.8168498168498166,
"grad_norm": 2.7145586013793945,
"learning_rate": 6.348814968811384e-07,
"loss": 0.8977913856506348,
"step": 3126
},
{
"epoch": 3.819291819291819,
"grad_norm": 5.800877094268799,
"learning_rate": 6.33965532427373e-07,
"loss": 0.4331892728805542,
"step": 3128
},
{
"epoch": 3.8217338217338215,
"grad_norm": 2.0423827171325684,
"learning_rate": 6.330616787050528e-07,
"loss": 0.8526274561882019,
"step": 3130
},
{
"epoch": 3.824175824175824,
"grad_norm": 0.7753233909606934,
"learning_rate": 6.321699398200254e-07,
"loss": 0.3737678825855255,
"step": 3132
},
{
"epoch": 3.8266178266178263,
"grad_norm": 6.823727130889893,
"learning_rate": 6.312903198231041e-07,
"loss": 0.9473689198493958,
"step": 3134
},
{
"epoch": 3.8290598290598292,
"grad_norm": 3.5573318004608154,
"learning_rate": 6.304228227100528e-07,
"loss": 0.8825540542602539,
"step": 3136
},
{
"epoch": 3.8315018315018317,
"grad_norm": 2.7311320304870605,
"learning_rate": 6.295674524215642e-07,
"loss": 0.5429124236106873,
"step": 3138
},
{
"epoch": 3.833943833943834,
"grad_norm": 2.064741611480713,
"learning_rate": 6.287242128432448e-07,
"loss": 0.4811321198940277,
"step": 3140
},
{
"epoch": 3.8363858363858365,
"grad_norm": 4.351254463195801,
"learning_rate": 6.278931078055954e-07,
"loss": 0.6814494132995605,
"step": 3142
},
{
"epoch": 3.838827838827839,
"grad_norm": 7.926380634307861,
"learning_rate": 6.270741410839952e-07,
"loss": 0.2247064709663391,
"step": 3144
},
{
"epoch": 3.8412698412698414,
"grad_norm": 4.106725215911865,
"learning_rate": 6.26267316398683e-07,
"loss": 0.4575510025024414,
"step": 3146
},
{
"epoch": 3.843711843711844,
"grad_norm": 0.16320517659187317,
"learning_rate": 6.254726374147418e-07,
"loss": 0.10908607393503189,
"step": 3148
},
{
"epoch": 3.8461538461538463,
"grad_norm": 5.839258193969727,
"learning_rate": 6.246901077420817e-07,
"loss": 0.6451608538627625,
"step": 3150
},
{
"epoch": 3.8485958485958487,
"grad_norm": 5.578673362731934,
"learning_rate": 6.239197309354223e-07,
"loss": 0.7072908282279968,
"step": 3152
},
{
"epoch": 3.851037851037851,
"grad_norm": 3.6340432167053223,
"learning_rate": 6.231615104942785e-07,
"loss": 0.471763551235199,
"step": 3154
},
{
"epoch": 3.8534798534798536,
"grad_norm": 1.6580287218093872,
"learning_rate": 6.224154498629434e-07,
"loss": 0.7280194163322449,
"step": 3156
},
{
"epoch": 3.855921855921856,
"grad_norm": 1.7522375583648682,
"learning_rate": 6.216815524304732e-07,
"loss": 0.7933505773544312,
"step": 3158
},
{
"epoch": 3.8583638583638584,
"grad_norm": 11.935617446899414,
"learning_rate": 6.209598215306708e-07,
"loss": 0.06976834684610367,
"step": 3160
},
{
"epoch": 3.860805860805861,
"grad_norm": 2.845559597015381,
"learning_rate": 6.202502604420724e-07,
"loss": 0.3656473457813263,
"step": 3162
},
{
"epoch": 3.8632478632478633,
"grad_norm": 1.7497023344039917,
"learning_rate": 6.195528723879306e-07,
"loss": 0.8561981320381165,
"step": 3164
},
{
"epoch": 3.8656898656898657,
"grad_norm": 2.9220542907714844,
"learning_rate": 6.188676605362014e-07,
"loss": 0.7740746140480042,
"step": 3166
},
{
"epoch": 3.868131868131868,
"grad_norm": 1.9189082384109497,
"learning_rate": 6.181946279995291e-07,
"loss": 0.7662597298622131,
"step": 3168
},
{
"epoch": 3.8705738705738706,
"grad_norm": 4.2788310050964355,
"learning_rate": 6.17533777835232e-07,
"loss": 0.6690661907196045,
"step": 3170
},
{
"epoch": 3.873015873015873,
"grad_norm": 6.616257190704346,
"learning_rate": 6.168851130452888e-07,
"loss": 0.5250207781791687,
"step": 3172
},
{
"epoch": 3.8754578754578755,
"grad_norm": 2.653538227081299,
"learning_rate": 6.162486365763252e-07,
"loss": 0.5411117076873779,
"step": 3174
},
{
"epoch": 3.877899877899878,
"grad_norm": 2.170850992202759,
"learning_rate": 6.156243513196e-07,
"loss": 0.8127894401550293,
"step": 3176
},
{
"epoch": 3.8803418803418803,
"grad_norm": 3.3351333141326904,
"learning_rate": 6.150122601109919e-07,
"loss": 0.9016299843788147,
"step": 3178
},
{
"epoch": 3.8827838827838828,
"grad_norm": 2.948333501815796,
"learning_rate": 6.144123657309872e-07,
"loss": 0.8240965604782104,
"step": 3180
},
{
"epoch": 3.885225885225885,
"grad_norm": 1.4079304933547974,
"learning_rate": 6.138246709046672e-07,
"loss": 0.7397529482841492,
"step": 3182
},
{
"epoch": 3.8876678876678876,
"grad_norm": 4.149062633514404,
"learning_rate": 6.132491783016945e-07,
"loss": 0.34523075819015503,
"step": 3184
},
{
"epoch": 3.89010989010989,
"grad_norm": 2.017380714416504,
"learning_rate": 6.126858905363034e-07,
"loss": 0.8450878262519836,
"step": 3186
},
{
"epoch": 3.8925518925518925,
"grad_norm": 1.749881386756897,
"learning_rate": 6.121348101672848e-07,
"loss": 0.36155635118484497,
"step": 3188
},
{
"epoch": 3.894993894993895,
"grad_norm": 1.3024982213974,
"learning_rate": 6.115959396979775e-07,
"loss": 0.4078894257545471,
"step": 3190
},
{
"epoch": 3.8974358974358974,
"grad_norm": 2.626460313796997,
"learning_rate": 6.110692815762551e-07,
"loss": 0.8648070693016052,
"step": 3192
},
{
"epoch": 3.8998778998779,
"grad_norm": 5.356231689453125,
"learning_rate": 6.105548381945153e-07,
"loss": 0.8094540238380432,
"step": 3194
},
{
"epoch": 3.9023199023199022,
"grad_norm": 25.908906936645508,
"learning_rate": 6.100526118896693e-07,
"loss": 0.5891746878623962,
"step": 3196
},
{
"epoch": 3.9047619047619047,
"grad_norm": 5.511157035827637,
"learning_rate": 6.09562604943131e-07,
"loss": 0.07191010564565659,
"step": 3198
},
{
"epoch": 3.907203907203907,
"grad_norm": 6.299953460693359,
"learning_rate": 6.090848195808064e-07,
"loss": 0.8390135765075684,
"step": 3200
},
{
"epoch": 3.9096459096459095,
"grad_norm": 1.8649171590805054,
"learning_rate": 6.086192579730838e-07,
"loss": 0.5590109825134277,
"step": 3202
},
{
"epoch": 3.912087912087912,
"grad_norm": 2.802351713180542,
"learning_rate": 6.081659222348244e-07,
"loss": 0.985985517501831,
"step": 3204
},
{
"epoch": 3.9145299145299144,
"grad_norm": 2.154928207397461,
"learning_rate": 6.07724814425351e-07,
"loss": 0.9891282916069031,
"step": 3206
},
{
"epoch": 3.916971916971917,
"grad_norm": 6.227573871612549,
"learning_rate": 6.07295936548441e-07,
"loss": 1.0907913446426392,
"step": 3208
},
{
"epoch": 3.9194139194139193,
"grad_norm": 2.676931381225586,
"learning_rate": 6.068792905523156e-07,
"loss": 0.5496243238449097,
"step": 3210
},
{
"epoch": 3.9218559218559217,
"grad_norm": 10.38682746887207,
"learning_rate": 6.064748783296311e-07,
"loss": 0.2229357361793518,
"step": 3212
},
{
"epoch": 3.9242979242979246,
"grad_norm": 2.008678674697876,
"learning_rate": 6.060827017174708e-07,
"loss": 0.8315946459770203,
"step": 3214
},
{
"epoch": 3.926739926739927,
"grad_norm": 4.7980146408081055,
"learning_rate": 6.057027624973373e-07,
"loss": 0.8503268957138062,
"step": 3216
},
{
"epoch": 3.9291819291819294,
"grad_norm": 4.067953109741211,
"learning_rate": 6.053350623951424e-07,
"loss": 0.563690721988678,
"step": 3218
},
{
"epoch": 3.931623931623932,
"grad_norm": 2.8761026859283447,
"learning_rate": 6.049796030812015e-07,
"loss": 0.6627534627914429,
"step": 3220
},
{
"epoch": 3.9340659340659343,
"grad_norm": 7.254218101501465,
"learning_rate": 6.046363861702239e-07,
"loss": 0.5418643355369568,
"step": 3222
},
{
"epoch": 3.9365079365079367,
"grad_norm": 1.3696813583374023,
"learning_rate": 6.043054132213079e-07,
"loss": 0.7259221076965332,
"step": 3224
},
{
"epoch": 3.938949938949939,
"grad_norm": 5.845968246459961,
"learning_rate": 6.039866857379309e-07,
"loss": 0.4918738901615143,
"step": 3226
},
{
"epoch": 3.9413919413919416,
"grad_norm": 1.7601078748703003,
"learning_rate": 6.036802051679455e-07,
"loss": 0.9122508764266968,
"step": 3228
},
{
"epoch": 3.943833943833944,
"grad_norm": 2.2423689365386963,
"learning_rate": 6.0338597290357e-07,
"loss": 0.4838387668132782,
"step": 3230
},
{
"epoch": 3.9462759462759465,
"grad_norm": 2.109056234359741,
"learning_rate": 6.031039902813847e-07,
"loss": 0.7914977073669434,
"step": 3232
},
{
"epoch": 3.948717948717949,
"grad_norm": 4.0477375984191895,
"learning_rate": 6.028342585823242e-07,
"loss": 0.48134973645210266,
"step": 3234
},
{
"epoch": 3.9511599511599513,
"grad_norm": 1.566765546798706,
"learning_rate": 6.025767790316716e-07,
"loss": 0.8497970104217529,
"step": 3236
},
{
"epoch": 3.9536019536019538,
"grad_norm": 8.391966819763184,
"learning_rate": 6.023315527990545e-07,
"loss": 0.7990399599075317,
"step": 3238
},
{
"epoch": 3.956043956043956,
"grad_norm": 3.6553773880004883,
"learning_rate": 6.020985809984372e-07,
"loss": 0.4292301535606384,
"step": 3240
},
{
"epoch": 3.9584859584859586,
"grad_norm": 4.039731025695801,
"learning_rate": 6.018778646881181e-07,
"loss": 0.5612732172012329,
"step": 3242
},
{
"epoch": 3.960927960927961,
"grad_norm": 1.7811068296432495,
"learning_rate": 6.016694048707235e-07,
"loss": 0.8350145816802979,
"step": 3244
},
{
"epoch": 3.9633699633699635,
"grad_norm": 2.358139991760254,
"learning_rate": 6.014732024932027e-07,
"loss": 0.4561384320259094,
"step": 3246
},
{
"epoch": 3.965811965811966,
"grad_norm": 1.5930027961730957,
"learning_rate": 6.012892584468251e-07,
"loss": 0.8480937480926514,
"step": 3248
},
{
"epoch": 3.9682539682539684,
"grad_norm": 6.148706912994385,
"learning_rate": 6.011175735671757e-07,
"loss": 0.5076125860214233,
"step": 3250
},
{
"epoch": 3.970695970695971,
"grad_norm": 16.231203079223633,
"learning_rate": 6.009581486341496e-07,
"loss": 0.18412072956562042,
"step": 3252
},
{
"epoch": 3.9731379731379732,
"grad_norm": 2.707773447036743,
"learning_rate": 6.008109843719513e-07,
"loss": 0.8144515156745911,
"step": 3254
},
{
"epoch": 3.9755799755799757,
"grad_norm": 2.6421728134155273,
"learning_rate": 6.006760814490892e-07,
"loss": 0.7697687745094299,
"step": 3256
},
{
"epoch": 3.978021978021978,
"grad_norm": 1.7280405759811401,
"learning_rate": 6.005534404783732e-07,
"loss": 0.37729355692863464,
"step": 3258
},
{
"epoch": 3.9804639804639805,
"grad_norm": 3.762451171875,
"learning_rate": 6.004430620169124e-07,
"loss": 0.4543501138687134,
"step": 3260
},
{
"epoch": 3.982905982905983,
"grad_norm": 2.905902147293091,
"learning_rate": 6.003449465661124e-07,
"loss": 0.5729525089263916,
"step": 3262
},
{
"epoch": 3.9853479853479854,
"grad_norm": 5.909022331237793,
"learning_rate": 6.002590945716726e-07,
"loss": 0.4344090521335602,
"step": 3264
},
{
"epoch": 3.987789987789988,
"grad_norm": 50.008880615234375,
"learning_rate": 6.001855064235843e-07,
"loss": 0.8903095722198486,
"step": 3266
},
{
"epoch": 3.9902319902319903,
"grad_norm": 2.487112283706665,
"learning_rate": 6.001241824561287e-07,
"loss": 0.9123827219009399,
"step": 3268
},
{
"epoch": 3.9926739926739927,
"grad_norm": 2.6051173210144043,
"learning_rate": 6.000751229478767e-07,
"loss": 0.6394712924957275,
"step": 3270
},
{
"epoch": 3.995115995115995,
"grad_norm": 3.082603693008423,
"learning_rate": 6.000383281216857e-07,
"loss": 0.8653581142425537,
"step": 3272
},
{
"epoch": 3.9975579975579976,
"grad_norm": 1.5350114107131958,
"learning_rate": 6.000137981446999e-07,
"loss": 0.7252496480941772,
"step": 3274
},
{
"epoch": 4.0,
"grad_norm": 2.818603992462158,
"learning_rate": 6.000015331283497e-07,
"loss": 0.47882503271102905,
"step": 3276
},
{
"epoch": 4.0,
"step": 3276,
"total_flos": 3.438047841308639e+18,
"train_loss": 0.884952375524301,
"train_runtime": 10014.0246,
"train_samples_per_second": 5.234,
"train_steps_per_second": 0.327
}
],
"logging_steps": 2,
"max_steps": 3276,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.438047841308639e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}