diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,6665 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1892, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.004228329809725159, + "grad_norm": 3.432370901107788, + "learning_rate": 1.0526315789473685e-07, + "loss": 1.8266634941101074, + "step": 2 + }, + { + "epoch": 0.008456659619450317, + "grad_norm": 1.2269554138183594, + "learning_rate": 3.1578947368421055e-07, + "loss": 2.214822292327881, + "step": 4 + }, + { + "epoch": 0.012684989429175475, + "grad_norm": 16.31293487548828, + "learning_rate": 5.263157894736843e-07, + "loss": 2.2101941108703613, + "step": 6 + }, + { + "epoch": 0.016913319238900635, + "grad_norm": 2.3812522888183594, + "learning_rate": 7.368421052631579e-07, + "loss": 1.9961841106414795, + "step": 8 + }, + { + "epoch": 0.021141649048625793, + "grad_norm": 12.308541297912598, + "learning_rate": 9.473684210526317e-07, + "loss": 2.0121607780456543, + "step": 10 + }, + { + "epoch": 0.02536997885835095, + "grad_norm": 2.0651066303253174, + "learning_rate": 1.1578947368421053e-06, + "loss": 1.6350789070129395, + "step": 12 + }, + { + "epoch": 0.02959830866807611, + "grad_norm": 1.8003875017166138, + "learning_rate": 1.3684210526315791e-06, + "loss": 1.6400768756866455, + "step": 14 + }, + { + "epoch": 0.03382663847780127, + "grad_norm": 2.4902076721191406, + "learning_rate": 1.5789473684210526e-06, + "loss": 1.7653813362121582, + "step": 16 + }, + { + "epoch": 0.03805496828752643, + "grad_norm": 1.1994597911834717, + "learning_rate": 1.7894736842105265e-06, + "loss": 1.7891666889190674, + "step": 18 + }, + { + "epoch": 0.042283298097251586, + "grad_norm": 5.830203533172607, + "learning_rate": 2.0000000000000003e-06, + "loss": 1.6629786491394043, + "step": 20 + }, + { + "epoch": 0.046511627906976744, + "grad_norm": 4.855837821960449, + "learning_rate": 2.2105263157894738e-06, + "loss": 1.5011135339736938, + "step": 22 + }, + { + "epoch": 0.0507399577167019, + "grad_norm": 2.163144111633301, + "learning_rate": 2.4210526315789477e-06, + "loss": 1.759260654449463, + "step": 24 + }, + { + "epoch": 0.05496828752642706, + "grad_norm": 3.066650152206421, + "learning_rate": 2.631578947368421e-06, + "loss": 1.3511682748794556, + "step": 26 + }, + { + "epoch": 0.05919661733615222, + "grad_norm": 0.960111677646637, + "learning_rate": 2.842105263157895e-06, + "loss": 1.6441280841827393, + "step": 28 + }, + { + "epoch": 0.06342494714587738, + "grad_norm": 0.9521470069885254, + "learning_rate": 3.052631578947369e-06, + "loss": 1.2195796966552734, + "step": 30 + }, + { + "epoch": 0.06765327695560254, + "grad_norm": 9.232256889343262, + "learning_rate": 3.2631578947368423e-06, + "loss": 0.9830036759376526, + "step": 32 + }, + { + "epoch": 0.07188160676532769, + "grad_norm": 1.2219663858413696, + "learning_rate": 3.473684210526316e-06, + "loss": 1.166562557220459, + "step": 34 + }, + { + "epoch": 0.07610993657505286, + "grad_norm": 7.223867893218994, + "learning_rate": 3.6842105263157896e-06, + "loss": 1.348006248474121, + "step": 36 + }, + { + "epoch": 0.080338266384778, + "grad_norm": 1.8092782497406006, + "learning_rate": 3.894736842105263e-06, + "loss": 1.2814323902130127, + "step": 38 + }, + { + "epoch": 0.08456659619450317, + "grad_norm": 3.420692205429077, + "learning_rate": 4.105263157894737e-06, + "loss": 1.2029392719268799, + "step": 40 + }, + { + "epoch": 0.08879492600422834, + "grad_norm": 4.487104415893555, + "learning_rate": 4.315789473684211e-06, + "loss": 1.688148021697998, + "step": 42 + }, + { + "epoch": 0.09302325581395349, + "grad_norm": 2.5123021602630615, + "learning_rate": 4.526315789473685e-06, + "loss": 1.5875585079193115, + "step": 44 + }, + { + "epoch": 0.09725158562367865, + "grad_norm": 1.3677341938018799, + "learning_rate": 4.736842105263158e-06, + "loss": 1.2614831924438477, + "step": 46 + }, + { + "epoch": 0.1014799154334038, + "grad_norm": 0.9786781668663025, + "learning_rate": 4.947368421052632e-06, + "loss": 1.4942141771316528, + "step": 48 + }, + { + "epoch": 0.10570824524312897, + "grad_norm": 1.5975029468536377, + "learning_rate": 5.157894736842106e-06, + "loss": 1.4608821868896484, + "step": 50 + }, + { + "epoch": 0.10993657505285412, + "grad_norm": 2.5109217166900635, + "learning_rate": 5.36842105263158e-06, + "loss": 0.7047278881072998, + "step": 52 + }, + { + "epoch": 0.11416490486257928, + "grad_norm": 1.130075454711914, + "learning_rate": 5.578947368421052e-06, + "loss": 1.0526067018508911, + "step": 54 + }, + { + "epoch": 0.11839323467230443, + "grad_norm": 1.5286709070205688, + "learning_rate": 5.789473684210527e-06, + "loss": 1.3589348793029785, + "step": 56 + }, + { + "epoch": 0.1226215644820296, + "grad_norm": 1.0047403573989868, + "learning_rate": 6e-06, + "loss": 1.3783363103866577, + "step": 58 + }, + { + "epoch": 0.12684989429175475, + "grad_norm": 0.9190431237220764, + "learning_rate": 6.2105263157894745e-06, + "loss": 1.393243670463562, + "step": 60 + }, + { + "epoch": 0.13107822410147993, + "grad_norm": 1.551571011543274, + "learning_rate": 6.421052631578948e-06, + "loss": 0.7647256851196289, + "step": 62 + }, + { + "epoch": 0.13530655391120508, + "grad_norm": 1.6368632316589355, + "learning_rate": 6.631578947368421e-06, + "loss": 1.0364748239517212, + "step": 64 + }, + { + "epoch": 0.13953488372093023, + "grad_norm": 1.258395791053772, + "learning_rate": 6.842105263157896e-06, + "loss": 1.3412930965423584, + "step": 66 + }, + { + "epoch": 0.14376321353065538, + "grad_norm": 1.8389757871627808, + "learning_rate": 7.052631578947369e-06, + "loss": 1.3744611740112305, + "step": 68 + }, + { + "epoch": 0.14799154334038056, + "grad_norm": 1.820074439048767, + "learning_rate": 7.263157894736843e-06, + "loss": 1.3497071266174316, + "step": 70 + }, + { + "epoch": 0.1522198731501057, + "grad_norm": 1.3421845436096191, + "learning_rate": 7.473684210526316e-06, + "loss": 1.3445059061050415, + "step": 72 + }, + { + "epoch": 0.15644820295983086, + "grad_norm": 1.226967692375183, + "learning_rate": 7.68421052631579e-06, + "loss": 0.9300884008407593, + "step": 74 + }, + { + "epoch": 0.160676532769556, + "grad_norm": 1.4437878131866455, + "learning_rate": 7.894736842105265e-06, + "loss": 0.8529413938522339, + "step": 76 + }, + { + "epoch": 0.1649048625792812, + "grad_norm": 1.5913375616073608, + "learning_rate": 8.105263157894736e-06, + "loss": 1.2803168296813965, + "step": 78 + }, + { + "epoch": 0.16913319238900634, + "grad_norm": 4.461759567260742, + "learning_rate": 8.315789473684212e-06, + "loss": 1.0896272659301758, + "step": 80 + }, + { + "epoch": 0.1733615221987315, + "grad_norm": 0.8871933221817017, + "learning_rate": 8.526315789473685e-06, + "loss": 1.082736611366272, + "step": 82 + }, + { + "epoch": 0.17758985200845667, + "grad_norm": 0.7336280941963196, + "learning_rate": 8.736842105263158e-06, + "loss": 0.8834536075592041, + "step": 84 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 1.5130850076675415, + "learning_rate": 8.947368421052632e-06, + "loss": 1.7085888385772705, + "step": 86 + }, + { + "epoch": 0.18604651162790697, + "grad_norm": 1.08708655834198, + "learning_rate": 9.157894736842105e-06, + "loss": 1.3114171028137207, + "step": 88 + }, + { + "epoch": 0.19027484143763213, + "grad_norm": 4.0093159675598145, + "learning_rate": 9.36842105263158e-06, + "loss": 0.9898357391357422, + "step": 90 + }, + { + "epoch": 0.1945031712473573, + "grad_norm": 1.3196293115615845, + "learning_rate": 9.578947368421054e-06, + "loss": 1.3104819059371948, + "step": 92 + }, + { + "epoch": 0.19873150105708245, + "grad_norm": 0.9962700605392456, + "learning_rate": 9.789473684210527e-06, + "loss": 0.9705811142921448, + "step": 94 + }, + { + "epoch": 0.2029598308668076, + "grad_norm": 1.550417423248291, + "learning_rate": 1e-05, + "loss": 1.2923219203948975, + "step": 96 + }, + { + "epoch": 0.20718816067653276, + "grad_norm": 1.149327039718628, + "learning_rate": 9.999972492846047e-06, + "loss": 1.385302186012268, + "step": 98 + }, + { + "epoch": 0.21141649048625794, + "grad_norm": 0.963651716709137, + "learning_rate": 9.999889971720468e-06, + "loss": 1.326028823852539, + "step": 100 + }, + { + "epoch": 0.2156448202959831, + "grad_norm": 0.5960947871208191, + "learning_rate": 9.999752437632121e-06, + "loss": 1.2462652921676636, + "step": 102 + }, + { + "epoch": 0.21987315010570824, + "grad_norm": 1.4711657762527466, + "learning_rate": 9.999559892262412e-06, + "loss": 1.3037302494049072, + "step": 104 + }, + { + "epoch": 0.22410147991543342, + "grad_norm": 0.8557540774345398, + "learning_rate": 9.999312337965291e-06, + "loss": 1.2940036058425903, + "step": 106 + }, + { + "epoch": 0.22832980972515857, + "grad_norm": 0.9147750735282898, + "learning_rate": 9.999009777767202e-06, + "loss": 1.0358223915100098, + "step": 108 + }, + { + "epoch": 0.23255813953488372, + "grad_norm": 0.9510343670845032, + "learning_rate": 9.99865221536707e-06, + "loss": 1.2810924053192139, + "step": 110 + }, + { + "epoch": 0.23678646934460887, + "grad_norm": 2.302175998687744, + "learning_rate": 9.998239655136237e-06, + "loss": 1.3574622869491577, + "step": 112 + }, + { + "epoch": 0.24101479915433405, + "grad_norm": 1.0104137659072876, + "learning_rate": 9.997772102118417e-06, + "loss": 1.5118541717529297, + "step": 114 + }, + { + "epoch": 0.2452431289640592, + "grad_norm": 1.0515953302383423, + "learning_rate": 9.997249562029637e-06, + "loss": 1.3338898420333862, + "step": 116 + }, + { + "epoch": 0.24947145877378435, + "grad_norm": 1.4034860134124756, + "learning_rate": 9.996672041258154e-06, + "loss": 1.3421039581298828, + "step": 118 + }, + { + "epoch": 0.2536997885835095, + "grad_norm": 1.8375295400619507, + "learning_rate": 9.996039546864395e-06, + "loss": 0.8561623692512512, + "step": 120 + }, + { + "epoch": 0.25792811839323465, + "grad_norm": 0.7544057965278625, + "learning_rate": 9.995352086580858e-06, + "loss": 1.3005398511886597, + "step": 122 + }, + { + "epoch": 0.26215644820295986, + "grad_norm": 1.8997758626937866, + "learning_rate": 9.994609668812022e-06, + "loss": 1.1488518714904785, + "step": 124 + }, + { + "epoch": 0.266384778012685, + "grad_norm": 1.157852053642273, + "learning_rate": 9.993812302634237e-06, + "loss": 1.3442846536636353, + "step": 126 + }, + { + "epoch": 0.27061310782241016, + "grad_norm": 1.033600926399231, + "learning_rate": 9.99295999779563e-06, + "loss": 0.9459542036056519, + "step": 128 + }, + { + "epoch": 0.2748414376321353, + "grad_norm": 0.8347329497337341, + "learning_rate": 9.99205276471597e-06, + "loss": 0.976822018623352, + "step": 130 + }, + { + "epoch": 0.27906976744186046, + "grad_norm": 2.2171669006347656, + "learning_rate": 9.991090614486546e-06, + "loss": 1.2778072357177734, + "step": 132 + }, + { + "epoch": 0.2832980972515856, + "grad_norm": 1.6110771894454956, + "learning_rate": 9.990073558870028e-06, + "loss": 0.8382304906845093, + "step": 134 + }, + { + "epoch": 0.28752642706131076, + "grad_norm": 2.4742887020111084, + "learning_rate": 9.989001610300333e-06, + "loss": 1.1078466176986694, + "step": 136 + }, + { + "epoch": 0.2917547568710359, + "grad_norm": 1.2871977090835571, + "learning_rate": 9.987874781882461e-06, + "loss": 1.3787493705749512, + "step": 138 + }, + { + "epoch": 0.2959830866807611, + "grad_norm": 0.9194169044494629, + "learning_rate": 9.986693087392341e-06, + "loss": 0.9308235049247742, + "step": 140 + }, + { + "epoch": 0.30021141649048627, + "grad_norm": 1.7333077192306519, + "learning_rate": 9.985456541276667e-06, + "loss": 1.5560063123703003, + "step": 142 + }, + { + "epoch": 0.3044397463002114, + "grad_norm": 2.4464614391326904, + "learning_rate": 9.984165158652708e-06, + "loss": 1.0934826135635376, + "step": 144 + }, + { + "epoch": 0.3086680761099366, + "grad_norm": 7.818751335144043, + "learning_rate": 9.982818955308135e-06, + "loss": 0.6715948581695557, + "step": 146 + }, + { + "epoch": 0.3128964059196617, + "grad_norm": 1.1906154155731201, + "learning_rate": 9.98141794770083e-06, + "loss": 1.2456437349319458, + "step": 148 + }, + { + "epoch": 0.3171247357293869, + "grad_norm": 0.9795631766319275, + "learning_rate": 9.979962152958669e-06, + "loss": 1.2671079635620117, + "step": 150 + }, + { + "epoch": 0.321353065539112, + "grad_norm": 1.0989006757736206, + "learning_rate": 9.97845158887933e-06, + "loss": 1.175504446029663, + "step": 152 + }, + { + "epoch": 0.32558139534883723, + "grad_norm": 4.20810079574585, + "learning_rate": 9.976886273930063e-06, + "loss": 0.7261967658996582, + "step": 154 + }, + { + "epoch": 0.3298097251585624, + "grad_norm": 1.0932376384735107, + "learning_rate": 9.975266227247474e-06, + "loss": 1.249506950378418, + "step": 156 + }, + { + "epoch": 0.33403805496828753, + "grad_norm": 0.9015107154846191, + "learning_rate": 9.973591468637286e-06, + "loss": 1.2704322338104248, + "step": 158 + }, + { + "epoch": 0.3382663847780127, + "grad_norm": 1.1831700801849365, + "learning_rate": 9.971862018574094e-06, + "loss": 1.2559826374053955, + "step": 160 + }, + { + "epoch": 0.34249471458773784, + "grad_norm": 2.6710400581359863, + "learning_rate": 9.97007789820112e-06, + "loss": 1.0979704856872559, + "step": 162 + }, + { + "epoch": 0.346723044397463, + "grad_norm": 1.7830309867858887, + "learning_rate": 9.968239129329953e-06, + "loss": 0.9583532214164734, + "step": 164 + }, + { + "epoch": 0.35095137420718814, + "grad_norm": 1.5103063583374023, + "learning_rate": 9.96634573444028e-06, + "loss": 0.8929807543754578, + "step": 166 + }, + { + "epoch": 0.35517970401691334, + "grad_norm": 1.8013784885406494, + "learning_rate": 9.964397736679615e-06, + "loss": 1.2005603313446045, + "step": 168 + }, + { + "epoch": 0.3594080338266385, + "grad_norm": 0.7990474700927734, + "learning_rate": 9.962395159863014e-06, + "loss": 1.2118266820907593, + "step": 170 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 0.6552523374557495, + "learning_rate": 9.960338028472781e-06, + "loss": 1.0488790273666382, + "step": 172 + }, + { + "epoch": 0.3678646934460888, + "grad_norm": 1.0434958934783936, + "learning_rate": 9.958226367658176e-06, + "loss": 1.3508553504943848, + "step": 174 + }, + { + "epoch": 0.37209302325581395, + "grad_norm": 0.9309784770011902, + "learning_rate": 9.9560602032351e-06, + "loss": 1.035404086112976, + "step": 176 + }, + { + "epoch": 0.3763213530655391, + "grad_norm": 2.7481086254119873, + "learning_rate": 9.953839561685783e-06, + "loss": 0.8625020980834961, + "step": 178 + }, + { + "epoch": 0.38054968287526425, + "grad_norm": 1.0653841495513916, + "learning_rate": 9.95156447015846e-06, + "loss": 1.2866113185882568, + "step": 180 + }, + { + "epoch": 0.38477801268498946, + "grad_norm": 1.2197877168655396, + "learning_rate": 9.94923495646704e-06, + "loss": 1.5259649753570557, + "step": 182 + }, + { + "epoch": 0.3890063424947146, + "grad_norm": 1.296986699104309, + "learning_rate": 9.946851049090764e-06, + "loss": 1.3405705690383911, + "step": 184 + }, + { + "epoch": 0.39323467230443976, + "grad_norm": 1.3075778484344482, + "learning_rate": 9.944412777173854e-06, + "loss": 1.113606572151184, + "step": 186 + }, + { + "epoch": 0.3974630021141649, + "grad_norm": 1.096283197402954, + "learning_rate": 9.941920170525168e-06, + "loss": 0.9938672184944153, + "step": 188 + }, + { + "epoch": 0.40169133192389006, + "grad_norm": 1.641861915588379, + "learning_rate": 9.939373259617822e-06, + "loss": 1.0212286710739136, + "step": 190 + }, + { + "epoch": 0.4059196617336152, + "grad_norm": 1.1199278831481934, + "learning_rate": 9.936772075588825e-06, + "loss": 1.296032428741455, + "step": 192 + }, + { + "epoch": 0.41014799154334036, + "grad_norm": 0.3750116527080536, + "learning_rate": 9.934116650238697e-06, + "loss": 1.004199743270874, + "step": 194 + }, + { + "epoch": 0.4143763213530655, + "grad_norm": 1.306297779083252, + "learning_rate": 9.931407016031078e-06, + "loss": 1.2771074771881104, + "step": 196 + }, + { + "epoch": 0.4186046511627907, + "grad_norm": 2.30645489692688, + "learning_rate": 9.928643206092338e-06, + "loss": 1.0745476484298706, + "step": 198 + }, + { + "epoch": 0.42283298097251587, + "grad_norm": 0.7231057286262512, + "learning_rate": 9.925825254211161e-06, + "loss": 1.1826379299163818, + "step": 200 + }, + { + "epoch": 0.427061310782241, + "grad_norm": 1.4271743297576904, + "learning_rate": 9.922953194838143e-06, + "loss": 0.589704692363739, + "step": 202 + }, + { + "epoch": 0.4312896405919662, + "grad_norm": 1.7569785118103027, + "learning_rate": 9.920027063085363e-06, + "loss": 1.2287952899932861, + "step": 204 + }, + { + "epoch": 0.4355179704016913, + "grad_norm": 11.933446884155273, + "learning_rate": 9.917046894725957e-06, + "loss": 1.3055689334869385, + "step": 206 + }, + { + "epoch": 0.4397463002114165, + "grad_norm": 4.293759346008301, + "learning_rate": 9.91401272619368e-06, + "loss": 0.7848652601242065, + "step": 208 + }, + { + "epoch": 0.4439746300211416, + "grad_norm": 1.3277677297592163, + "learning_rate": 9.910924594582463e-06, + "loss": 1.2941670417785645, + "step": 210 + }, + { + "epoch": 0.44820295983086683, + "grad_norm": 1.0579293966293335, + "learning_rate": 9.907782537645955e-06, + "loss": 1.2195998430252075, + "step": 212 + }, + { + "epoch": 0.452431289640592, + "grad_norm": 1.1869871616363525, + "learning_rate": 9.904586593797062e-06, + "loss": 1.4210292100906372, + "step": 214 + }, + { + "epoch": 0.45665961945031713, + "grad_norm": 0.9697230458259583, + "learning_rate": 9.901336802107485e-06, + "loss": 0.6687322854995728, + "step": 216 + }, + { + "epoch": 0.4608879492600423, + "grad_norm": 4.1858978271484375, + "learning_rate": 9.89803320230723e-06, + "loss": 1.0143953561782837, + "step": 218 + }, + { + "epoch": 0.46511627906976744, + "grad_norm": 0.8843514919281006, + "learning_rate": 9.894675834784136e-06, + "loss": 1.274765133857727, + "step": 220 + }, + { + "epoch": 0.4693446088794926, + "grad_norm": 0.8239166736602783, + "learning_rate": 9.891264740583367e-06, + "loss": 1.2279239892959595, + "step": 222 + }, + { + "epoch": 0.47357293868921774, + "grad_norm": 1.0150994062423706, + "learning_rate": 9.88779996140692e-06, + "loss": 1.3455586433410645, + "step": 224 + }, + { + "epoch": 0.47780126849894294, + "grad_norm": 0.9176246523857117, + "learning_rate": 9.884281539613109e-06, + "loss": 1.2493900060653687, + "step": 226 + }, + { + "epoch": 0.4820295983086681, + "grad_norm": 0.6774841547012329, + "learning_rate": 9.880709518216061e-06, + "loss": 1.2465935945510864, + "step": 228 + }, + { + "epoch": 0.48625792811839325, + "grad_norm": 1.1564357280731201, + "learning_rate": 9.87708394088517e-06, + "loss": 0.9222732782363892, + "step": 230 + }, + { + "epoch": 0.4904862579281184, + "grad_norm": 0.8378649950027466, + "learning_rate": 9.873404851944573e-06, + "loss": 1.2515840530395508, + "step": 232 + }, + { + "epoch": 0.49471458773784355, + "grad_norm": 3.0550146102905273, + "learning_rate": 9.869672296372611e-06, + "loss": 0.9386910796165466, + "step": 234 + }, + { + "epoch": 0.4989429175475687, + "grad_norm": 6.6074090003967285, + "learning_rate": 9.86588631980128e-06, + "loss": 0.9068935513496399, + "step": 236 + }, + { + "epoch": 0.5031712473572939, + "grad_norm": 1.8940776586532593, + "learning_rate": 9.86204696851566e-06, + "loss": 0.8527886271476746, + "step": 238 + }, + { + "epoch": 0.507399577167019, + "grad_norm": 0.9940058588981628, + "learning_rate": 9.858154289453363e-06, + "loss": 0.9246642589569092, + "step": 240 + }, + { + "epoch": 0.5116279069767442, + "grad_norm": 1.563374400138855, + "learning_rate": 9.854208330203958e-06, + "loss": 1.3151363134384155, + "step": 242 + }, + { + "epoch": 0.5158562367864693, + "grad_norm": 1.2777539491653442, + "learning_rate": 9.850209139008377e-06, + "loss": 1.3305803537368774, + "step": 244 + }, + { + "epoch": 0.5200845665961945, + "grad_norm": 1.3406856060028076, + "learning_rate": 9.846156764758344e-06, + "loss": 1.2857502698898315, + "step": 246 + }, + { + "epoch": 0.5243128964059197, + "grad_norm": 0.9291679263114929, + "learning_rate": 9.842051256995759e-06, + "loss": 1.2356584072113037, + "step": 248 + }, + { + "epoch": 0.5285412262156448, + "grad_norm": 1.5110595226287842, + "learning_rate": 9.837892665912105e-06, + "loss": 1.1822853088378906, + "step": 250 + }, + { + "epoch": 0.53276955602537, + "grad_norm": 2.680776596069336, + "learning_rate": 9.833681042347827e-06, + "loss": 1.1697964668273926, + "step": 252 + }, + { + "epoch": 0.5369978858350951, + "grad_norm": 1.3377050161361694, + "learning_rate": 9.829416437791721e-06, + "loss": 1.739481806755066, + "step": 254 + }, + { + "epoch": 0.5412262156448203, + "grad_norm": 0.7585676312446594, + "learning_rate": 9.825098904380284e-06, + "loss": 1.2119654417037964, + "step": 256 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 1.3904414176940918, + "learning_rate": 9.820728494897101e-06, + "loss": 1.393256425857544, + "step": 258 + }, + { + "epoch": 0.5496828752642706, + "grad_norm": 0.6582812070846558, + "learning_rate": 9.816305262772182e-06, + "loss": 1.0005115270614624, + "step": 260 + }, + { + "epoch": 0.5539112050739958, + "grad_norm": 1.2003450393676758, + "learning_rate": 9.811829262081315e-06, + "loss": 1.1924750804901123, + "step": 262 + }, + { + "epoch": 0.5581395348837209, + "grad_norm": 1.018638014793396, + "learning_rate": 9.807300547545411e-06, + "loss": 1.0629571676254272, + "step": 264 + }, + { + "epoch": 0.5623678646934461, + "grad_norm": 0.9702784419059753, + "learning_rate": 9.802719174529821e-06, + "loss": 1.493300437927246, + "step": 266 + }, + { + "epoch": 0.5665961945031712, + "grad_norm": 5.1235432624816895, + "learning_rate": 9.798085199043672e-06, + "loss": 0.5702834129333496, + "step": 268 + }, + { + "epoch": 0.5708245243128964, + "grad_norm": 1.9747250080108643, + "learning_rate": 9.793398677739177e-06, + "loss": 1.0085912942886353, + "step": 270 + }, + { + "epoch": 0.5750528541226215, + "grad_norm": 3.398075580596924, + "learning_rate": 9.788659667910942e-06, + "loss": 1.183624029159546, + "step": 272 + }, + { + "epoch": 0.5792811839323467, + "grad_norm": 0.7080333232879639, + "learning_rate": 9.783868227495262e-06, + "loss": 0.9508528113365173, + "step": 274 + }, + { + "epoch": 0.5835095137420718, + "grad_norm": 1.1537398099899292, + "learning_rate": 9.779024415069427e-06, + "loss": 0.9382362365722656, + "step": 276 + }, + { + "epoch": 0.587737843551797, + "grad_norm": 0.5110741853713989, + "learning_rate": 9.774128289850987e-06, + "loss": 0.8599207401275635, + "step": 278 + }, + { + "epoch": 0.5919661733615222, + "grad_norm": 0.7918158769607544, + "learning_rate": 9.769179911697037e-06, + "loss": 1.2367205619812012, + "step": 280 + }, + { + "epoch": 0.5961945031712473, + "grad_norm": 2.872875213623047, + "learning_rate": 9.764179341103493e-06, + "loss": 1.4585726261138916, + "step": 282 + }, + { + "epoch": 0.6004228329809725, + "grad_norm": 0.48448535799980164, + "learning_rate": 9.759126639204337e-06, + "loss": 0.9426642060279846, + "step": 284 + }, + { + "epoch": 0.6046511627906976, + "grad_norm": 0.855935275554657, + "learning_rate": 9.75402186777088e-06, + "loss": 1.3067773580551147, + "step": 286 + }, + { + "epoch": 0.6088794926004228, + "grad_norm": 1.1592859029769897, + "learning_rate": 9.748865089211004e-06, + "loss": 0.436039537191391, + "step": 288 + }, + { + "epoch": 0.6131078224101479, + "grad_norm": 0.9701468348503113, + "learning_rate": 9.743656366568398e-06, + "loss": 0.9499627947807312, + "step": 290 + }, + { + "epoch": 0.6173361522198731, + "grad_norm": 1.0178031921386719, + "learning_rate": 9.738395763521792e-06, + "loss": 0.8715479969978333, + "step": 292 + }, + { + "epoch": 0.6215644820295984, + "grad_norm": 2.44124174118042, + "learning_rate": 9.733083344384164e-06, + "loss": 0.7320042252540588, + "step": 294 + }, + { + "epoch": 0.6257928118393234, + "grad_norm": 3.050933837890625, + "learning_rate": 9.727719174101981e-06, + "loss": 0.7356696724891663, + "step": 296 + }, + { + "epoch": 0.6300211416490487, + "grad_norm": 0.9884108901023865, + "learning_rate": 9.722303318254376e-06, + "loss": 1.2744605541229248, + "step": 298 + }, + { + "epoch": 0.6342494714587738, + "grad_norm": 0.8869503736495972, + "learning_rate": 9.71683584305236e-06, + "loss": 1.2401574850082397, + "step": 300 + }, + { + "epoch": 0.638477801268499, + "grad_norm": 1.286820888519287, + "learning_rate": 9.71131681533802e-06, + "loss": 0.7463349103927612, + "step": 302 + }, + { + "epoch": 0.642706131078224, + "grad_norm": 1.124804973602295, + "learning_rate": 9.705746302583682e-06, + "loss": 0.8460999131202698, + "step": 304 + }, + { + "epoch": 0.6469344608879493, + "grad_norm": 0.7509139180183411, + "learning_rate": 9.700124372891104e-06, + "loss": 0.9095317125320435, + "step": 306 + }, + { + "epoch": 0.6511627906976745, + "grad_norm": 0.9610084295272827, + "learning_rate": 9.694451094990636e-06, + "loss": 1.2296950817108154, + "step": 308 + }, + { + "epoch": 0.6553911205073996, + "grad_norm": 1.255344033241272, + "learning_rate": 9.68872653824038e-06, + "loss": 1.2034916877746582, + "step": 310 + }, + { + "epoch": 0.6596194503171248, + "grad_norm": 0.7059822082519531, + "learning_rate": 9.68295077262534e-06, + "loss": 1.202559232711792, + "step": 312 + }, + { + "epoch": 0.6638477801268499, + "grad_norm": 1.5127094984054565, + "learning_rate": 9.677123868756573e-06, + "loss": 1.2318086624145508, + "step": 314 + }, + { + "epoch": 0.6680761099365751, + "grad_norm": 1.7505037784576416, + "learning_rate": 9.671245897870318e-06, + "loss": 1.0772197246551514, + "step": 316 + }, + { + "epoch": 0.6723044397463002, + "grad_norm": 1.8178893327713013, + "learning_rate": 9.665316931827131e-06, + "loss": 0.7960694432258606, + "step": 318 + }, + { + "epoch": 0.6765327695560254, + "grad_norm": 1.0382561683654785, + "learning_rate": 9.659337043111006e-06, + "loss": 1.2425750494003296, + "step": 320 + }, + { + "epoch": 0.6807610993657506, + "grad_norm": 0.8479052186012268, + "learning_rate": 9.653306304828483e-06, + "loss": 1.3043532371520996, + "step": 322 + }, + { + "epoch": 0.6849894291754757, + "grad_norm": 0.9018455743789673, + "learning_rate": 9.647224790707762e-06, + "loss": 1.0811829566955566, + "step": 324 + }, + { + "epoch": 0.6892177589852009, + "grad_norm": 1.5794569253921509, + "learning_rate": 9.641092575097795e-06, + "loss": 1.2593729496002197, + "step": 326 + }, + { + "epoch": 0.693446088794926, + "grad_norm": 1.3980871438980103, + "learning_rate": 9.63490973296738e-06, + "loss": 1.1707936525344849, + "step": 328 + }, + { + "epoch": 0.6976744186046512, + "grad_norm": 3.612877368927002, + "learning_rate": 9.62867633990425e-06, + "loss": 0.8749436140060425, + "step": 330 + }, + { + "epoch": 0.7019027484143763, + "grad_norm": 0.9410237669944763, + "learning_rate": 9.622392472114136e-06, + "loss": 1.248960018157959, + "step": 332 + }, + { + "epoch": 0.7061310782241015, + "grad_norm": 2.758704423904419, + "learning_rate": 9.61605820641985e-06, + "loss": 1.2263383865356445, + "step": 334 + }, + { + "epoch": 0.7103594080338267, + "grad_norm": 1.2853227853775024, + "learning_rate": 9.60967362026033e-06, + "loss": 1.2079390287399292, + "step": 336 + }, + { + "epoch": 0.7145877378435518, + "grad_norm": 0.9742172956466675, + "learning_rate": 9.60323879168971e-06, + "loss": 1.3035683631896973, + "step": 338 + }, + { + "epoch": 0.718816067653277, + "grad_norm": 17.1197509765625, + "learning_rate": 9.596753799376355e-06, + "loss": 1.21787428855896, + "step": 340 + }, + { + "epoch": 0.7230443974630021, + "grad_norm": 2.2225992679595947, + "learning_rate": 9.5902187226019e-06, + "loss": 1.1275969743728638, + "step": 342 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 1.881629228591919, + "learning_rate": 9.583633641260285e-06, + "loss": 0.9096772074699402, + "step": 344 + }, + { + "epoch": 0.7315010570824524, + "grad_norm": 5.478016376495361, + "learning_rate": 9.576998635856774e-06, + "loss": 0.8505641222000122, + "step": 346 + }, + { + "epoch": 0.7357293868921776, + "grad_norm": 2.770082712173462, + "learning_rate": 9.570313787506974e-06, + "loss": 0.9856930375099182, + "step": 348 + }, + { + "epoch": 0.7399577167019028, + "grad_norm": 0.8123994469642639, + "learning_rate": 9.563579177935844e-06, + "loss": 1.2990679740905762, + "step": 350 + }, + { + "epoch": 0.7441860465116279, + "grad_norm": 1.2549108266830444, + "learning_rate": 9.556794889476687e-06, + "loss": 1.1582320928573608, + "step": 352 + }, + { + "epoch": 0.7484143763213531, + "grad_norm": 1.7971625328063965, + "learning_rate": 9.549961005070161e-06, + "loss": 0.7151243090629578, + "step": 354 + }, + { + "epoch": 0.7526427061310782, + "grad_norm": 1.4521124362945557, + "learning_rate": 9.543077608263245e-06, + "loss": 1.0172699689865112, + "step": 356 + }, + { + "epoch": 0.7568710359408034, + "grad_norm": 0.8418075442314148, + "learning_rate": 9.536144783208229e-06, + "loss": 1.2074182033538818, + "step": 358 + }, + { + "epoch": 0.7610993657505285, + "grad_norm": 1.7443615198135376, + "learning_rate": 9.529162614661687e-06, + "loss": 0.7586445212364197, + "step": 360 + }, + { + "epoch": 0.7653276955602537, + "grad_norm": 1.3051332235336304, + "learning_rate": 9.522131187983433e-06, + "loss": 1.0901813507080078, + "step": 362 + }, + { + "epoch": 0.7695560253699789, + "grad_norm": 3.2068612575531006, + "learning_rate": 9.515050589135485e-06, + "loss": 1.0601897239685059, + "step": 364 + }, + { + "epoch": 0.773784355179704, + "grad_norm": 1.4952013492584229, + "learning_rate": 9.507920904681006e-06, + "loss": 1.2570055723190308, + "step": 366 + }, + { + "epoch": 0.7780126849894292, + "grad_norm": 1.013555884361267, + "learning_rate": 9.500742221783257e-06, + "loss": 1.2038854360580444, + "step": 368 + }, + { + "epoch": 0.7822410147991543, + "grad_norm": 6.852783203125, + "learning_rate": 9.493514628204516e-06, + "loss": 0.8750179409980774, + "step": 370 + }, + { + "epoch": 0.7864693446088795, + "grad_norm": 1.2793877124786377, + "learning_rate": 9.48623821230502e-06, + "loss": 1.1862928867340088, + "step": 372 + }, + { + "epoch": 0.7906976744186046, + "grad_norm": 1.5481107234954834, + "learning_rate": 9.478913063041878e-06, + "loss": 1.2054961919784546, + "step": 374 + }, + { + "epoch": 0.7949260042283298, + "grad_norm": 1.898176908493042, + "learning_rate": 9.471539269967982e-06, + "loss": 1.2891910076141357, + "step": 376 + }, + { + "epoch": 0.7991543340380549, + "grad_norm": 1.576489806175232, + "learning_rate": 9.464116923230913e-06, + "loss": 0.9752786755561829, + "step": 378 + }, + { + "epoch": 0.8033826638477801, + "grad_norm": 3.02227520942688, + "learning_rate": 9.456646113571844e-06, + "loss": 1.2916446924209595, + "step": 380 + }, + { + "epoch": 0.8076109936575053, + "grad_norm": 1.7446725368499756, + "learning_rate": 9.449126932324424e-06, + "loss": 1.1669645309448242, + "step": 382 + }, + { + "epoch": 0.8118393234672304, + "grad_norm": 1.2364778518676758, + "learning_rate": 9.441559471413665e-06, + "loss": 0.9683005809783936, + "step": 384 + }, + { + "epoch": 0.8160676532769556, + "grad_norm": 11.812475204467773, + "learning_rate": 9.433943823354816e-06, + "loss": 0.9517707824707031, + "step": 386 + }, + { + "epoch": 0.8202959830866807, + "grad_norm": 1.0091938972473145, + "learning_rate": 9.426280081252232e-06, + "loss": 1.4731582403182983, + "step": 388 + }, + { + "epoch": 0.8245243128964059, + "grad_norm": 6.173403739929199, + "learning_rate": 9.418568338798245e-06, + "loss": 1.1501412391662598, + "step": 390 + }, + { + "epoch": 0.828752642706131, + "grad_norm": 5.444221019744873, + "learning_rate": 9.410808690271999e-06, + "loss": 0.9310405850410461, + "step": 392 + }, + { + "epoch": 0.8329809725158562, + "grad_norm": 1.5666706562042236, + "learning_rate": 9.40300123053832e-06, + "loss": 1.2317243814468384, + "step": 394 + }, + { + "epoch": 0.8372093023255814, + "grad_norm": 0.24349795281887054, + "learning_rate": 9.395146055046535e-06, + "loss": 0.7611055374145508, + "step": 396 + }, + { + "epoch": 0.8414376321353065, + "grad_norm": 0.864414632320404, + "learning_rate": 9.387243259829323e-06, + "loss": 0.6852919459342957, + "step": 398 + }, + { + "epoch": 0.8456659619450317, + "grad_norm": 1.5123212337493896, + "learning_rate": 9.37929294150153e-06, + "loss": 1.1211647987365723, + "step": 400 + }, + { + "epoch": 0.8498942917547568, + "grad_norm": 1.4040285348892212, + "learning_rate": 9.371295197258992e-06, + "loss": 0.9623957872390747, + "step": 402 + }, + { + "epoch": 0.854122621564482, + "grad_norm": 0.5657250285148621, + "learning_rate": 9.363250124877345e-06, + "loss": 0.8230786919593811, + "step": 404 + }, + { + "epoch": 0.8583509513742071, + "grad_norm": 1.237465500831604, + "learning_rate": 9.35515782271083e-06, + "loss": 0.9845194220542908, + "step": 406 + }, + { + "epoch": 0.8625792811839323, + "grad_norm": 1.0880403518676758, + "learning_rate": 9.347018389691093e-06, + "loss": 0.9817683696746826, + "step": 408 + }, + { + "epoch": 0.8668076109936576, + "grad_norm": 1.9139715433120728, + "learning_rate": 9.338831925325974e-06, + "loss": 0.8787358999252319, + "step": 410 + }, + { + "epoch": 0.8710359408033826, + "grad_norm": 1.7283661365509033, + "learning_rate": 9.330598529698287e-06, + "loss": 1.183264970779419, + "step": 412 + }, + { + "epoch": 0.8752642706131079, + "grad_norm": 1.1281251907348633, + "learning_rate": 9.322318303464603e-06, + "loss": 1.0252187252044678, + "step": 414 + }, + { + "epoch": 0.879492600422833, + "grad_norm": 2.6932523250579834, + "learning_rate": 9.313991347854015e-06, + "loss": 1.1178538799285889, + "step": 416 + }, + { + "epoch": 0.8837209302325582, + "grad_norm": 0.7762623429298401, + "learning_rate": 9.305617764666897e-06, + "loss": 1.2115470170974731, + "step": 418 + }, + { + "epoch": 0.8879492600422833, + "grad_norm": 1.0163838863372803, + "learning_rate": 9.297197656273672e-06, + "loss": 1.1681904792785645, + "step": 420 + }, + { + "epoch": 0.8921775898520085, + "grad_norm": 2.3892319202423096, + "learning_rate": 9.288731125613544e-06, + "loss": 0.9400791525840759, + "step": 422 + }, + { + "epoch": 0.8964059196617337, + "grad_norm": 3.914783000946045, + "learning_rate": 9.280218276193255e-06, + "loss": 1.229254961013794, + "step": 424 + }, + { + "epoch": 0.9006342494714588, + "grad_norm": 0.9981458187103271, + "learning_rate": 9.271659212085807e-06, + "loss": 1.0940930843353271, + "step": 426 + }, + { + "epoch": 0.904862579281184, + "grad_norm": 2.4068081378936768, + "learning_rate": 9.263054037929199e-06, + "loss": 0.8796516060829163, + "step": 428 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 1.0804468393325806, + "learning_rate": 9.254402858925142e-06, + "loss": 1.1219546794891357, + "step": 430 + }, + { + "epoch": 0.9133192389006343, + "grad_norm": 2.030319929122925, + "learning_rate": 9.245705780837774e-06, + "loss": 1.2639557123184204, + "step": 432 + }, + { + "epoch": 0.9175475687103594, + "grad_norm": 2.617784261703491, + "learning_rate": 9.236962909992368e-06, + "loss": 0.5257906913757324, + "step": 434 + }, + { + "epoch": 0.9217758985200846, + "grad_norm": 1.9403786659240723, + "learning_rate": 9.228174353274037e-06, + "loss": 0.39105135202407837, + "step": 436 + }, + { + "epoch": 0.9260042283298098, + "grad_norm": 1.7514019012451172, + "learning_rate": 9.219340218126413e-06, + "loss": 0.9517390727996826, + "step": 438 + }, + { + "epoch": 0.9302325581395349, + "grad_norm": 0.9263060688972473, + "learning_rate": 9.210460612550348e-06, + "loss": 1.1432080268859863, + "step": 440 + }, + { + "epoch": 0.9344608879492601, + "grad_norm": 5.8498077392578125, + "learning_rate": 9.20153564510259e-06, + "loss": 0.7478289008140564, + "step": 442 + }, + { + "epoch": 0.9386892177589852, + "grad_norm": 2.669020652770996, + "learning_rate": 9.19256542489445e-06, + "loss": 1.2597752809524536, + "step": 444 + }, + { + "epoch": 0.9429175475687104, + "grad_norm": 0.9071623682975769, + "learning_rate": 9.183550061590476e-06, + "loss": 1.0969961881637573, + "step": 446 + }, + { + "epoch": 0.9471458773784355, + "grad_norm": 1.810524344444275, + "learning_rate": 9.174489665407103e-06, + "loss": 1.091011643409729, + "step": 448 + }, + { + "epoch": 0.9513742071881607, + "grad_norm": 2.8751273155212402, + "learning_rate": 9.165384347111317e-06, + "loss": 0.800588071346283, + "step": 450 + }, + { + "epoch": 0.9556025369978859, + "grad_norm": 0.9461812973022461, + "learning_rate": 9.156234218019291e-06, + "loss": 0.8863937854766846, + "step": 452 + }, + { + "epoch": 0.959830866807611, + "grad_norm": 1.2746022939682007, + "learning_rate": 9.147039389995029e-06, + "loss": 1.314546823501587, + "step": 454 + }, + { + "epoch": 0.9640591966173362, + "grad_norm": 1.4575163125991821, + "learning_rate": 9.137799975448998e-06, + "loss": 0.6644535064697266, + "step": 456 + }, + { + "epoch": 0.9682875264270613, + "grad_norm": 2.342315912246704, + "learning_rate": 9.128516087336755e-06, + "loss": 0.8800123929977417, + "step": 458 + }, + { + "epoch": 0.9725158562367865, + "grad_norm": 0.9644564986228943, + "learning_rate": 9.119187839157559e-06, + "loss": 1.223119854927063, + "step": 460 + }, + { + "epoch": 0.9767441860465116, + "grad_norm": 2.072160243988037, + "learning_rate": 9.109815344952995e-06, + "loss": 1.2090517282485962, + "step": 462 + }, + { + "epoch": 0.9809725158562368, + "grad_norm": 1.4577010869979858, + "learning_rate": 9.100398719305566e-06, + "loss": 0.7334827184677124, + "step": 464 + }, + { + "epoch": 0.985200845665962, + "grad_norm": 0.8280290961265564, + "learning_rate": 9.090938077337305e-06, + "loss": 1.1758804321289062, + "step": 466 + }, + { + "epoch": 0.9894291754756871, + "grad_norm": 3.2668943405151367, + "learning_rate": 9.081433534708363e-06, + "loss": 0.9873254895210266, + "step": 468 + }, + { + "epoch": 0.9936575052854123, + "grad_norm": 1.8736488819122314, + "learning_rate": 9.071885207615591e-06, + "loss": 1.2415308952331543, + "step": 470 + }, + { + "epoch": 0.9978858350951374, + "grad_norm": 0.8245468735694885, + "learning_rate": 9.062293212791123e-06, + "loss": 1.246222972869873, + "step": 472 + }, + { + "epoch": 1.0021141649048626, + "grad_norm": 1.9057750701904297, + "learning_rate": 9.052657667500951e-06, + "loss": 1.0147171020507812, + "step": 474 + }, + { + "epoch": 1.0063424947145878, + "grad_norm": 1.0847645998001099, + "learning_rate": 9.042978689543487e-06, + "loss": 1.1072622537612915, + "step": 476 + }, + { + "epoch": 1.0105708245243128, + "grad_norm": 1.8684371709823608, + "learning_rate": 9.033256397248124e-06, + "loss": 0.4752121865749359, + "step": 478 + }, + { + "epoch": 1.014799154334038, + "grad_norm": 1.4569878578186035, + "learning_rate": 9.02349090947379e-06, + "loss": 0.7129391431808472, + "step": 480 + }, + { + "epoch": 1.0190274841437632, + "grad_norm": 1.3255629539489746, + "learning_rate": 9.013682345607502e-06, + "loss": 0.7047132849693298, + "step": 482 + }, + { + "epoch": 1.0232558139534884, + "grad_norm": 1.213117003440857, + "learning_rate": 9.003830825562887e-06, + "loss": 0.9994310736656189, + "step": 484 + }, + { + "epoch": 1.0274841437632136, + "grad_norm": 2.4814367294311523, + "learning_rate": 8.99393646977874e-06, + "loss": 1.052198052406311, + "step": 486 + }, + { + "epoch": 1.0317124735729386, + "grad_norm": 1.1184537410736084, + "learning_rate": 8.983999399217536e-06, + "loss": 0.8507804870605469, + "step": 488 + }, + { + "epoch": 1.0359408033826638, + "grad_norm": 1.351128339767456, + "learning_rate": 8.974019735363952e-06, + "loss": 0.8119800090789795, + "step": 490 + }, + { + "epoch": 1.040169133192389, + "grad_norm": 2.5962376594543457, + "learning_rate": 8.963997600223387e-06, + "loss": 0.7182390689849854, + "step": 492 + }, + { + "epoch": 1.0443974630021142, + "grad_norm": 1.0359458923339844, + "learning_rate": 8.953933116320474e-06, + "loss": 1.1185437440872192, + "step": 494 + }, + { + "epoch": 1.0486257928118394, + "grad_norm": 1.4469959735870361, + "learning_rate": 8.943826406697569e-06, + "loss": 0.9576400518417358, + "step": 496 + }, + { + "epoch": 1.0528541226215644, + "grad_norm": 0.6539305448532104, + "learning_rate": 8.93367759491326e-06, + "loss": 0.7852102518081665, + "step": 498 + }, + { + "epoch": 1.0570824524312896, + "grad_norm": 2.3659756183624268, + "learning_rate": 8.923486805040845e-06, + "loss": 0.8582470417022705, + "step": 500 + }, + { + "epoch": 1.0613107822410148, + "grad_norm": 0.8050792217254639, + "learning_rate": 8.913254161666828e-06, + "loss": 1.1146094799041748, + "step": 502 + }, + { + "epoch": 1.06553911205074, + "grad_norm": 0.8499988317489624, + "learning_rate": 8.902979789889382e-06, + "loss": 0.9811311960220337, + "step": 504 + }, + { + "epoch": 1.069767441860465, + "grad_norm": 3.1883771419525146, + "learning_rate": 8.892663815316833e-06, + "loss": 0.783465564250946, + "step": 506 + }, + { + "epoch": 1.0739957716701902, + "grad_norm": 0.7172738313674927, + "learning_rate": 8.882306364066114e-06, + "loss": 0.786201536655426, + "step": 508 + }, + { + "epoch": 1.0782241014799154, + "grad_norm": 3.9731550216674805, + "learning_rate": 8.871907562761226e-06, + "loss": 0.9575165510177612, + "step": 510 + }, + { + "epoch": 1.0824524312896406, + "grad_norm": 1.3529167175292969, + "learning_rate": 8.861467538531692e-06, + "loss": 1.0198792219161987, + "step": 512 + }, + { + "epoch": 1.0866807610993658, + "grad_norm": 2.1442151069641113, + "learning_rate": 8.850986419011006e-06, + "loss": 1.0819895267486572, + "step": 514 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 0.8071472644805908, + "learning_rate": 8.840464332335062e-06, + "loss": 0.8569697737693787, + "step": 516 + }, + { + "epoch": 1.095137420718816, + "grad_norm": 0.8328543901443481, + "learning_rate": 8.829901407140599e-06, + "loss": 0.8482993841171265, + "step": 518 + }, + { + "epoch": 1.0993657505285412, + "grad_norm": 3.250950813293457, + "learning_rate": 8.81929777256362e-06, + "loss": 0.6445708274841309, + "step": 520 + }, + { + "epoch": 1.1035940803382664, + "grad_norm": 1.7141265869140625, + "learning_rate": 8.808653558237818e-06, + "loss": 1.2630366086959839, + "step": 522 + }, + { + "epoch": 1.1078224101479917, + "grad_norm": 1.430616855621338, + "learning_rate": 8.797968894292992e-06, + "loss": 0.748252809047699, + "step": 524 + }, + { + "epoch": 1.1120507399577166, + "grad_norm": 0.9419235587120056, + "learning_rate": 8.787243911353447e-06, + "loss": 1.3854765892028809, + "step": 526 + }, + { + "epoch": 1.1162790697674418, + "grad_norm": 0.7226510643959045, + "learning_rate": 8.776478740536409e-06, + "loss": 0.8130208253860474, + "step": 528 + }, + { + "epoch": 1.120507399577167, + "grad_norm": 8.920578002929688, + "learning_rate": 8.76567351345042e-06, + "loss": 0.4950494170188904, + "step": 530 + }, + { + "epoch": 1.1247357293868923, + "grad_norm": 1.1824902296066284, + "learning_rate": 8.754828362193719e-06, + "loss": 1.010925054550171, + "step": 532 + }, + { + "epoch": 1.1289640591966172, + "grad_norm": 4.8336591720581055, + "learning_rate": 8.743943419352638e-06, + "loss": 0.7442973852157593, + "step": 534 + }, + { + "epoch": 1.1331923890063424, + "grad_norm": 0.5645107626914978, + "learning_rate": 8.733018817999977e-06, + "loss": 0.9534879922866821, + "step": 536 + }, + { + "epoch": 1.1374207188160677, + "grad_norm": 1.3391566276550293, + "learning_rate": 8.722054691693374e-06, + "loss": 0.7376333475112915, + "step": 538 + }, + { + "epoch": 1.1416490486257929, + "grad_norm": 0.9142791628837585, + "learning_rate": 8.711051174473682e-06, + "loss": 0.6905903220176697, + "step": 540 + }, + { + "epoch": 1.145877378435518, + "grad_norm": 1.6403955221176147, + "learning_rate": 8.700008400863316e-06, + "loss": 1.0252349376678467, + "step": 542 + }, + { + "epoch": 1.150105708245243, + "grad_norm": 2.4788246154785156, + "learning_rate": 8.688926505864624e-06, + "loss": 0.7351314425468445, + "step": 544 + }, + { + "epoch": 1.1543340380549683, + "grad_norm": 0.69508296251297, + "learning_rate": 8.67780562495822e-06, + "loss": 1.1282206773757935, + "step": 546 + }, + { + "epoch": 1.1585623678646935, + "grad_norm": 2.1606674194335938, + "learning_rate": 8.666645894101346e-06, + "loss": 1.2914485931396484, + "step": 548 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 1.1642086505889893, + "learning_rate": 8.655447449726189e-06, + "loss": 0.7788194417953491, + "step": 550 + }, + { + "epoch": 1.1670190274841437, + "grad_norm": 3.890846014022827, + "learning_rate": 8.644210428738237e-06, + "loss": 1.0123552083969116, + "step": 552 + }, + { + "epoch": 1.1712473572938689, + "grad_norm": 1.6349087953567505, + "learning_rate": 8.632934968514582e-06, + "loss": 1.1295274496078491, + "step": 554 + }, + { + "epoch": 1.175475687103594, + "grad_norm": 2.3135898113250732, + "learning_rate": 8.621621206902256e-06, + "loss": 1.0003678798675537, + "step": 556 + }, + { + "epoch": 1.1797040169133193, + "grad_norm": 0.964402437210083, + "learning_rate": 8.61026928221654e-06, + "loss": 1.014490008354187, + "step": 558 + }, + { + "epoch": 1.1839323467230445, + "grad_norm": 0.7214915156364441, + "learning_rate": 8.598879333239274e-06, + "loss": 0.9661449193954468, + "step": 560 + }, + { + "epoch": 1.1881606765327695, + "grad_norm": 2.6707189083099365, + "learning_rate": 8.587451499217161e-06, + "loss": 0.7001445889472961, + "step": 562 + }, + { + "epoch": 1.1923890063424947, + "grad_norm": 1.21587336063385, + "learning_rate": 8.575985919860065e-06, + "loss": 1.0277830362319946, + "step": 564 + }, + { + "epoch": 1.1966173361522199, + "grad_norm": 1.8789758682250977, + "learning_rate": 8.564482735339296e-06, + "loss": 0.9529078006744385, + "step": 566 + }, + { + "epoch": 1.200845665961945, + "grad_norm": 1.6645126342773438, + "learning_rate": 8.552942086285909e-06, + "loss": 1.2264162302017212, + "step": 568 + }, + { + "epoch": 1.20507399577167, + "grad_norm": 5.157077312469482, + "learning_rate": 8.541364113788974e-06, + "loss": 1.144323706626892, + "step": 570 + }, + { + "epoch": 1.2093023255813953, + "grad_norm": 1.8206926584243774, + "learning_rate": 8.529748959393855e-06, + "loss": 1.0146433115005493, + "step": 572 + }, + { + "epoch": 1.2135306553911205, + "grad_norm": 0.943773627281189, + "learning_rate": 8.518096765100485e-06, + "loss": 0.7863532900810242, + "step": 574 + }, + { + "epoch": 1.2177589852008457, + "grad_norm": 2.501394271850586, + "learning_rate": 8.50640767336162e-06, + "loss": 0.7742471694946289, + "step": 576 + }, + { + "epoch": 1.221987315010571, + "grad_norm": 4.286005020141602, + "learning_rate": 8.494681827081096e-06, + "loss": 0.8926150798797607, + "step": 578 + }, + { + "epoch": 1.226215644820296, + "grad_norm": 1.3842778205871582, + "learning_rate": 8.482919369612103e-06, + "loss": 0.8287017345428467, + "step": 580 + }, + { + "epoch": 1.230443974630021, + "grad_norm": 2.123948335647583, + "learning_rate": 8.471120444755405e-06, + "loss": 1.1018996238708496, + "step": 582 + }, + { + "epoch": 1.2346723044397463, + "grad_norm": 2.2407541275024414, + "learning_rate": 8.4592851967576e-06, + "loss": 0.9453326463699341, + "step": 584 + }, + { + "epoch": 1.2389006342494715, + "grad_norm": 5.7265753746032715, + "learning_rate": 8.447413770309349e-06, + "loss": 0.8476324081420898, + "step": 586 + }, + { + "epoch": 1.2431289640591967, + "grad_norm": 1.9978158473968506, + "learning_rate": 8.435506310543607e-06, + "loss": 0.8746148347854614, + "step": 588 + }, + { + "epoch": 1.2473572938689217, + "grad_norm": 1.0323681831359863, + "learning_rate": 8.42356296303386e-06, + "loss": 0.6095293164253235, + "step": 590 + }, + { + "epoch": 1.251585623678647, + "grad_norm": 0.5390802621841431, + "learning_rate": 8.411583873792325e-06, + "loss": 0.80368572473526, + "step": 592 + }, + { + "epoch": 1.255813953488372, + "grad_norm": 8.916613578796387, + "learning_rate": 8.399569189268182e-06, + "loss": 1.0682650804519653, + "step": 594 + }, + { + "epoch": 1.2600422832980973, + "grad_norm": 1.0128202438354492, + "learning_rate": 8.387519056345778e-06, + "loss": 0.6816946268081665, + "step": 596 + }, + { + "epoch": 1.2642706131078225, + "grad_norm": 0.8855728507041931, + "learning_rate": 8.375433622342824e-06, + "loss": 1.101008653640747, + "step": 598 + }, + { + "epoch": 1.2684989429175475, + "grad_norm": 2.1038267612457275, + "learning_rate": 8.363313035008613e-06, + "loss": 1.2884970903396606, + "step": 600 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 0.8188600540161133, + "learning_rate": 8.351157442522187e-06, + "loss": 0.8854770660400391, + "step": 602 + }, + { + "epoch": 1.276955602536998, + "grad_norm": 0.8462247848510742, + "learning_rate": 8.338966993490554e-06, + "loss": 1.0687255859375, + "step": 604 + }, + { + "epoch": 1.2811839323467231, + "grad_norm": 2.9716718196868896, + "learning_rate": 8.326741836946852e-06, + "loss": 0.7634568810462952, + "step": 606 + }, + { + "epoch": 1.285412262156448, + "grad_norm": 0.8664849996566772, + "learning_rate": 8.314482122348523e-06, + "loss": 1.0502450466156006, + "step": 608 + }, + { + "epoch": 1.2896405919661733, + "grad_norm": 0.9873576760292053, + "learning_rate": 8.30218799957551e-06, + "loss": 0.7701109647750854, + "step": 610 + }, + { + "epoch": 1.2938689217758985, + "grad_norm": 1.0947819948196411, + "learning_rate": 8.289859618928404e-06, + "loss": 0.6398302912712097, + "step": 612 + }, + { + "epoch": 1.2980972515856237, + "grad_norm": 0.7757117748260498, + "learning_rate": 8.277497131126608e-06, + "loss": 0.8694761395454407, + "step": 614 + }, + { + "epoch": 1.302325581395349, + "grad_norm": 1.2524510622024536, + "learning_rate": 8.265100687306503e-06, + "loss": 0.5030971169471741, + "step": 616 + }, + { + "epoch": 1.306553911205074, + "grad_norm": 16.640033721923828, + "learning_rate": 8.252670439019595e-06, + "loss": 0.8920400142669678, + "step": 618 + }, + { + "epoch": 1.3107822410147991, + "grad_norm": 0.6471233367919922, + "learning_rate": 8.240206538230663e-06, + "loss": 1.1126651763916016, + "step": 620 + }, + { + "epoch": 1.3150105708245243, + "grad_norm": 0.9322137832641602, + "learning_rate": 8.227709137315905e-06, + "loss": 0.9956121444702148, + "step": 622 + }, + { + "epoch": 1.3192389006342495, + "grad_norm": 1.3649861812591553, + "learning_rate": 8.215178389061064e-06, + "loss": 0.7928463816642761, + "step": 624 + }, + { + "epoch": 1.3234672304439745, + "grad_norm": 0.7223655581474304, + "learning_rate": 8.202614446659572e-06, + "loss": 1.0491870641708374, + "step": 626 + }, + { + "epoch": 1.3276955602536997, + "grad_norm": 1.4594831466674805, + "learning_rate": 8.190017463710675e-06, + "loss": 1.17411470413208, + "step": 628 + }, + { + "epoch": 1.331923890063425, + "grad_norm": 1.3069385290145874, + "learning_rate": 8.177387594217548e-06, + "loss": 1.0363198518753052, + "step": 630 + }, + { + "epoch": 1.3361522198731501, + "grad_norm": 1.1741656064987183, + "learning_rate": 8.164724992585421e-06, + "loss": 0.6287016272544861, + "step": 632 + }, + { + "epoch": 1.3403805496828753, + "grad_norm": 0.9149936437606812, + "learning_rate": 8.152029813619685e-06, + "loss": 0.7200716137886047, + "step": 634 + }, + { + "epoch": 1.3446088794926006, + "grad_norm": 1.0878347158432007, + "learning_rate": 8.139302212524006e-06, + "loss": 0.7308053374290466, + "step": 636 + }, + { + "epoch": 1.3488372093023255, + "grad_norm": 2.570563316345215, + "learning_rate": 8.126542344898419e-06, + "loss": 0.9812169075012207, + "step": 638 + }, + { + "epoch": 1.3530655391120507, + "grad_norm": 0.668743371963501, + "learning_rate": 8.113750366737431e-06, + "loss": 0.7291630506515503, + "step": 640 + }, + { + "epoch": 1.357293868921776, + "grad_norm": 0.8278720378875732, + "learning_rate": 8.100926434428116e-06, + "loss": 1.1781340837478638, + "step": 642 + }, + { + "epoch": 1.361522198731501, + "grad_norm": 0.877626359462738, + "learning_rate": 8.0880707047482e-06, + "loss": 1.0660470724105835, + "step": 644 + }, + { + "epoch": 1.3657505285412261, + "grad_norm": 2.164764404296875, + "learning_rate": 8.075183334864139e-06, + "loss": 0.7516101002693176, + "step": 646 + }, + { + "epoch": 1.3699788583509513, + "grad_norm": 1.7909237146377563, + "learning_rate": 8.062264482329209e-06, + "loss": 0.7041962146759033, + "step": 648 + }, + { + "epoch": 1.3742071881606766, + "grad_norm": 4.3711113929748535, + "learning_rate": 8.049314305081575e-06, + "loss": 0.7470508217811584, + "step": 650 + }, + { + "epoch": 1.3784355179704018, + "grad_norm": 2.861943244934082, + "learning_rate": 8.036332961442355e-06, + "loss": 0.7375006079673767, + "step": 652 + }, + { + "epoch": 1.382663847780127, + "grad_norm": 0.8263391852378845, + "learning_rate": 8.023320610113691e-06, + "loss": 1.152714729309082, + "step": 654 + }, + { + "epoch": 1.386892177589852, + "grad_norm": 1.3392091989517212, + "learning_rate": 8.010277410176801e-06, + "loss": 0.9912066459655762, + "step": 656 + }, + { + "epoch": 1.3911205073995772, + "grad_norm": 1.6691114902496338, + "learning_rate": 7.997203521090053e-06, + "loss": 0.865416944026947, + "step": 658 + }, + { + "epoch": 1.3953488372093024, + "grad_norm": 3.069655418395996, + "learning_rate": 7.984099102686986e-06, + "loss": 0.9387435913085938, + "step": 660 + }, + { + "epoch": 1.3995771670190273, + "grad_norm": 0.9359487891197205, + "learning_rate": 7.970964315174384e-06, + "loss": 1.0533168315887451, + "step": 662 + }, + { + "epoch": 1.4038054968287526, + "grad_norm": 0.21687138080596924, + "learning_rate": 7.9577993191303e-06, + "loss": 0.8734056949615479, + "step": 664 + }, + { + "epoch": 1.4080338266384778, + "grad_norm": 1.5798527002334595, + "learning_rate": 7.944604275502102e-06, + "loss": 1.199033498764038, + "step": 666 + }, + { + "epoch": 1.412262156448203, + "grad_norm": 0.8036489486694336, + "learning_rate": 7.931379345604495e-06, + "loss": 1.0926624536514282, + "step": 668 + }, + { + "epoch": 1.4164904862579282, + "grad_norm": 1.0274549722671509, + "learning_rate": 7.918124691117563e-06, + "loss": 0.6228892207145691, + "step": 670 + }, + { + "epoch": 1.4207188160676534, + "grad_norm": 0.9182811975479126, + "learning_rate": 7.904840474084783e-06, + "loss": 1.0273898839950562, + "step": 672 + }, + { + "epoch": 1.4249471458773784, + "grad_norm": 1.2766584157943726, + "learning_rate": 7.891526856911046e-06, + "loss": 0.5703936815261841, + "step": 674 + }, + { + "epoch": 1.4291754756871036, + "grad_norm": 1.939756989479065, + "learning_rate": 7.878184002360664e-06, + "loss": 0.765852689743042, + "step": 676 + }, + { + "epoch": 1.4334038054968288, + "grad_norm": 1.1129051446914673, + "learning_rate": 7.864812073555402e-06, + "loss": 1.2686443328857422, + "step": 678 + }, + { + "epoch": 1.437632135306554, + "grad_norm": 0.8963062763214111, + "learning_rate": 7.851411233972461e-06, + "loss": 1.1558243036270142, + "step": 680 + }, + { + "epoch": 1.441860465116279, + "grad_norm": 2.708163022994995, + "learning_rate": 7.837981647442486e-06, + "loss": 1.1563526391983032, + "step": 682 + }, + { + "epoch": 1.4460887949260042, + "grad_norm": 1.1730352640151978, + "learning_rate": 7.82452347814757e-06, + "loss": 1.0920851230621338, + "step": 684 + }, + { + "epoch": 1.4503171247357294, + "grad_norm": 1.5551517009735107, + "learning_rate": 7.811036890619239e-06, + "loss": 1.3649554252624512, + "step": 686 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 0.9818048477172852, + "learning_rate": 7.797522049736441e-06, + "loss": 1.030195951461792, + "step": 688 + }, + { + "epoch": 1.4587737843551798, + "grad_norm": 1.4609994888305664, + "learning_rate": 7.78397912072354e-06, + "loss": 0.4242745041847229, + "step": 690 + }, + { + "epoch": 1.463002114164905, + "grad_norm": 1.0644174814224243, + "learning_rate": 7.770408269148281e-06, + "loss": 1.1275460720062256, + "step": 692 + }, + { + "epoch": 1.46723044397463, + "grad_norm": 0.8054934144020081, + "learning_rate": 7.756809660919778e-06, + "loss": 0.5492128133773804, + "step": 694 + }, + { + "epoch": 1.4714587737843552, + "grad_norm": 2.121662139892578, + "learning_rate": 7.74318346228648e-06, + "loss": 1.03450608253479, + "step": 696 + }, + { + "epoch": 1.4756871035940804, + "grad_norm": 0.9237878918647766, + "learning_rate": 7.72952983983414e-06, + "loss": 1.1372120380401611, + "step": 698 + }, + { + "epoch": 1.4799154334038054, + "grad_norm": 4.046713829040527, + "learning_rate": 7.715848960483776e-06, + "loss": 0.9173552393913269, + "step": 700 + }, + { + "epoch": 1.4841437632135306, + "grad_norm": 0.6067910194396973, + "learning_rate": 7.702140991489637e-06, + "loss": 1.0187067985534668, + "step": 702 + }, + { + "epoch": 1.4883720930232558, + "grad_norm": 1.7863223552703857, + "learning_rate": 7.688406100437149e-06, + "loss": 0.9936438202857971, + "step": 704 + }, + { + "epoch": 1.492600422832981, + "grad_norm": 0.6458945274353027, + "learning_rate": 7.674644455240878e-06, + "loss": 0.6985223293304443, + "step": 706 + }, + { + "epoch": 1.4968287526427062, + "grad_norm": 0.42708370089530945, + "learning_rate": 7.660856224142456e-06, + "loss": 0.2622559368610382, + "step": 708 + }, + { + "epoch": 1.5010570824524314, + "grad_norm": 1.8386787176132202, + "learning_rate": 7.647041575708555e-06, + "loss": 0.5994659066200256, + "step": 710 + }, + { + "epoch": 1.5052854122621564, + "grad_norm": 2.634805202484131, + "learning_rate": 7.633200678828804e-06, + "loss": 0.5527669787406921, + "step": 712 + }, + { + "epoch": 1.5095137420718816, + "grad_norm": 1.2726631164550781, + "learning_rate": 7.6193337027137215e-06, + "loss": 0.863373875617981, + "step": 714 + }, + { + "epoch": 1.5137420718816068, + "grad_norm": 0.8802553415298462, + "learning_rate": 7.605440816892667e-06, + "loss": 1.0387914180755615, + "step": 716 + }, + { + "epoch": 1.5179704016913318, + "grad_norm": 1.4086440801620483, + "learning_rate": 7.591522191211746e-06, + "loss": 0.9854521155357361, + "step": 718 + }, + { + "epoch": 1.522198731501057, + "grad_norm": 0.3206987977027893, + "learning_rate": 7.577577995831755e-06, + "loss": 0.9069070816040039, + "step": 720 + }, + { + "epoch": 1.5264270613107822, + "grad_norm": 1.200541615486145, + "learning_rate": 7.563608401226081e-06, + "loss": 0.868287205696106, + "step": 722 + }, + { + "epoch": 1.5306553911205074, + "grad_norm": 1.112833023071289, + "learning_rate": 7.549613578178632e-06, + "loss": 0.9251552224159241, + "step": 724 + }, + { + "epoch": 1.5348837209302326, + "grad_norm": 2.230614185333252, + "learning_rate": 7.535593697781741e-06, + "loss": 0.9085221886634827, + "step": 726 + }, + { + "epoch": 1.5391120507399578, + "grad_norm": 2.21995210647583, + "learning_rate": 7.52154893143408e-06, + "loss": 0.6506403684616089, + "step": 728 + }, + { + "epoch": 1.543340380549683, + "grad_norm": 8.375800132751465, + "learning_rate": 7.50747945083856e-06, + "loss": 0.8118706941604614, + "step": 730 + }, + { + "epoch": 1.547568710359408, + "grad_norm": 1.259455680847168, + "learning_rate": 7.493385428000234e-06, + "loss": 1.0121232271194458, + "step": 732 + }, + { + "epoch": 1.5517970401691332, + "grad_norm": 0.633515477180481, + "learning_rate": 7.479267035224191e-06, + "loss": 0.22526824474334717, + "step": 734 + }, + { + "epoch": 1.5560253699788582, + "grad_norm": 0.9651268720626831, + "learning_rate": 7.465124445113459e-06, + "loss": 0.9825803637504578, + "step": 736 + }, + { + "epoch": 1.5602536997885834, + "grad_norm": 0.45610707998275757, + "learning_rate": 7.450957830566878e-06, + "loss": 0.7875441908836365, + "step": 738 + }, + { + "epoch": 1.5644820295983086, + "grad_norm": 1.135189175605774, + "learning_rate": 7.436767364777009e-06, + "loss": 1.0916999578475952, + "step": 740 + }, + { + "epoch": 1.5687103594080338, + "grad_norm": 3.6160378456115723, + "learning_rate": 7.422553221227993e-06, + "loss": 0.7896254658699036, + "step": 742 + }, + { + "epoch": 1.572938689217759, + "grad_norm": 1.3152095079421997, + "learning_rate": 7.4083155736934484e-06, + "loss": 1.0520005226135254, + "step": 744 + }, + { + "epoch": 1.5771670190274842, + "grad_norm": 1.3807041645050049, + "learning_rate": 7.394054596234336e-06, + "loss": 1.1856319904327393, + "step": 746 + }, + { + "epoch": 1.5813953488372094, + "grad_norm": 0.9735748767852783, + "learning_rate": 7.379770463196831e-06, + "loss": 0.9112440347671509, + "step": 748 + }, + { + "epoch": 1.5856236786469344, + "grad_norm": 0.5606073141098022, + "learning_rate": 7.365463349210203e-06, + "loss": 0.08642717450857162, + "step": 750 + }, + { + "epoch": 1.5898520084566596, + "grad_norm": 1.1244453191757202, + "learning_rate": 7.351133429184666e-06, + "loss": 0.8228130340576172, + "step": 752 + }, + { + "epoch": 1.5940803382663846, + "grad_norm": 0.866052508354187, + "learning_rate": 7.336780878309251e-06, + "loss": 1.1204414367675781, + "step": 754 + }, + { + "epoch": 1.5983086680761098, + "grad_norm": 0.6875949501991272, + "learning_rate": 7.322405872049657e-06, + "loss": 1.0807037353515625, + "step": 756 + }, + { + "epoch": 1.602536997885835, + "grad_norm": 1.0158826112747192, + "learning_rate": 7.308008586146111e-06, + "loss": 1.131313681602478, + "step": 758 + }, + { + "epoch": 1.6067653276955602, + "grad_norm": 1.2606338262557983, + "learning_rate": 7.293589196611221e-06, + "loss": 1.0428601503372192, + "step": 760 + }, + { + "epoch": 1.6109936575052854, + "grad_norm": 0.8860005140304565, + "learning_rate": 7.279147879727811e-06, + "loss": 1.0722112655639648, + "step": 762 + }, + { + "epoch": 1.6152219873150107, + "grad_norm": 2.396479606628418, + "learning_rate": 7.264684812046787e-06, + "loss": 1.0421149730682373, + "step": 764 + }, + { + "epoch": 1.6194503171247359, + "grad_norm": 4.678726673126221, + "learning_rate": 7.250200170384961e-06, + "loss": 0.8483679890632629, + "step": 766 + }, + { + "epoch": 1.6236786469344608, + "grad_norm": 0.9464835524559021, + "learning_rate": 7.235694131822894e-06, + "loss": 1.0866531133651733, + "step": 768 + }, + { + "epoch": 1.627906976744186, + "grad_norm": 1.586958885192871, + "learning_rate": 7.221166873702738e-06, + "loss": 1.0823144912719727, + "step": 770 + }, + { + "epoch": 1.6321353065539113, + "grad_norm": 0.7832309603691101, + "learning_rate": 7.206618573626058e-06, + "loss": 1.072551965713501, + "step": 772 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 0.7828298807144165, + "learning_rate": 7.19204940945167e-06, + "loss": 0.7548497319221497, + "step": 774 + }, + { + "epoch": 1.6405919661733614, + "grad_norm": 0.7542831301689148, + "learning_rate": 7.177459559293456e-06, + "loss": 1.105739712715149, + "step": 776 + }, + { + "epoch": 1.6448202959830867, + "grad_norm": 0.9643332362174988, + "learning_rate": 7.162849201518198e-06, + "loss": 0.6944323778152466, + "step": 778 + }, + { + "epoch": 1.6490486257928119, + "grad_norm": 0.860163688659668, + "learning_rate": 7.1482185147433905e-06, + "loss": 1.0175727605819702, + "step": 780 + }, + { + "epoch": 1.653276955602537, + "grad_norm": 1.0091168880462646, + "learning_rate": 7.133567677835055e-06, + "loss": 0.785447359085083, + "step": 782 + }, + { + "epoch": 1.6575052854122623, + "grad_norm": 1.735445261001587, + "learning_rate": 7.11889686990556e-06, + "loss": 1.1903661489486694, + "step": 784 + }, + { + "epoch": 1.6617336152219875, + "grad_norm": 1.3719654083251953, + "learning_rate": 7.104206270311423e-06, + "loss": 0.4472939670085907, + "step": 786 + }, + { + "epoch": 1.6659619450317125, + "grad_norm": 0.8533653616905212, + "learning_rate": 7.089496058651133e-06, + "loss": 1.0958935022354126, + "step": 788 + }, + { + "epoch": 1.6701902748414377, + "grad_norm": 1.089091420173645, + "learning_rate": 7.0747664147629335e-06, + "loss": 0.7530285716056824, + "step": 790 + }, + { + "epoch": 1.6744186046511627, + "grad_norm": 1.5687955617904663, + "learning_rate": 7.060017518722637e-06, + "loss": 0.4044314920902252, + "step": 792 + }, + { + "epoch": 1.6786469344608879, + "grad_norm": 1.1035963296890259, + "learning_rate": 7.0452495508414276e-06, + "loss": 1.0479317903518677, + "step": 794 + }, + { + "epoch": 1.682875264270613, + "grad_norm": 1.9705493450164795, + "learning_rate": 7.030462691663641e-06, + "loss": 0.6204204559326172, + "step": 796 + }, + { + "epoch": 1.6871035940803383, + "grad_norm": 1.6793073415756226, + "learning_rate": 7.015657121964575e-06, + "loss": 1.1809488534927368, + "step": 798 + }, + { + "epoch": 1.6913319238900635, + "grad_norm": 1.0700589418411255, + "learning_rate": 7.000833022748267e-06, + "loss": 1.0959999561309814, + "step": 800 + }, + { + "epoch": 1.6955602536997887, + "grad_norm": 3.814356803894043, + "learning_rate": 6.985990575245283e-06, + "loss": 0.5697633028030396, + "step": 802 + }, + { + "epoch": 1.699788583509514, + "grad_norm": 0.6834250092506409, + "learning_rate": 6.97112996091051e-06, + "loss": 1.0314223766326904, + "step": 804 + }, + { + "epoch": 1.7040169133192389, + "grad_norm": 2.3212087154388428, + "learning_rate": 6.956251361420926e-06, + "loss": 0.8077873587608337, + "step": 806 + }, + { + "epoch": 1.708245243128964, + "grad_norm": 1.250013828277588, + "learning_rate": 6.94135495867339e-06, + "loss": 0.7262765169143677, + "step": 808 + }, + { + "epoch": 1.712473572938689, + "grad_norm": 0.9661218523979187, + "learning_rate": 6.926440934782407e-06, + "loss": 1.0879294872283936, + "step": 810 + }, + { + "epoch": 1.7167019027484143, + "grad_norm": 3.035346031188965, + "learning_rate": 6.9115094720779155e-06, + "loss": 1.2979084253311157, + "step": 812 + }, + { + "epoch": 1.7209302325581395, + "grad_norm": 1.3294485807418823, + "learning_rate": 6.896560753103041e-06, + "loss": 1.0285578966140747, + "step": 814 + }, + { + "epoch": 1.7251585623678647, + "grad_norm": 0.8368066549301147, + "learning_rate": 6.881594960611881e-06, + "loss": 1.12623929977417, + "step": 816 + }, + { + "epoch": 1.72938689217759, + "grad_norm": 0.7728613615036011, + "learning_rate": 6.866612277567261e-06, + "loss": 0.887596070766449, + "step": 818 + }, + { + "epoch": 1.733615221987315, + "grad_norm": 4.624102592468262, + "learning_rate": 6.851612887138501e-06, + "loss": 0.965933084487915, + "step": 820 + }, + { + "epoch": 1.7378435517970403, + "grad_norm": 0.7787982225418091, + "learning_rate": 6.836596972699175e-06, + "loss": 0.6407999396324158, + "step": 822 + }, + { + "epoch": 1.7420718816067653, + "grad_norm": 0.3913739025592804, + "learning_rate": 6.82156471782487e-06, + "loss": 0.633185625076294, + "step": 824 + }, + { + "epoch": 1.7463002114164905, + "grad_norm": 14.031166076660156, + "learning_rate": 6.806516306290942e-06, + "loss": 0.8862119317054749, + "step": 826 + }, + { + "epoch": 1.7505285412262155, + "grad_norm": 0.7377650737762451, + "learning_rate": 6.791451922070265e-06, + "loss": 0.9257691502571106, + "step": 828 + }, + { + "epoch": 1.7547568710359407, + "grad_norm": 2.8704004287719727, + "learning_rate": 6.77637174933099e-06, + "loss": 0.7364992499351501, + "step": 830 + }, + { + "epoch": 1.758985200845666, + "grad_norm": 1.2103379964828491, + "learning_rate": 6.7612759724342895e-06, + "loss": 1.0885521173477173, + "step": 832 + }, + { + "epoch": 1.763213530655391, + "grad_norm": 11.273382186889648, + "learning_rate": 6.746164775932096e-06, + "loss": 0.8072609901428223, + "step": 834 + }, + { + "epoch": 1.7674418604651163, + "grad_norm": 0.7183442711830139, + "learning_rate": 6.731038344564862e-06, + "loss": 0.785464346408844, + "step": 836 + }, + { + "epoch": 1.7716701902748415, + "grad_norm": 0.7529750466346741, + "learning_rate": 6.715896863259286e-06, + "loss": 1.0955817699432373, + "step": 838 + }, + { + "epoch": 1.7758985200845667, + "grad_norm": 1.1023916006088257, + "learning_rate": 6.700740517126061e-06, + "loss": 0.8635135889053345, + "step": 840 + }, + { + "epoch": 1.7801268498942917, + "grad_norm": 1.5844699144363403, + "learning_rate": 6.685569491457608e-06, + "loss": 0.35527199506759644, + "step": 842 + }, + { + "epoch": 1.784355179704017, + "grad_norm": 0.9075146317481995, + "learning_rate": 6.670383971725809e-06, + "loss": 0.6582541465759277, + "step": 844 + }, + { + "epoch": 1.7885835095137421, + "grad_norm": 6.298500061035156, + "learning_rate": 6.655184143579746e-06, + "loss": 0.9488147497177124, + "step": 846 + }, + { + "epoch": 1.792811839323467, + "grad_norm": 0.954853892326355, + "learning_rate": 6.639970192843423e-06, + "loss": 0.7089717984199524, + "step": 848 + }, + { + "epoch": 1.7970401691331923, + "grad_norm": 1.1899363994598389, + "learning_rate": 6.624742305513502e-06, + "loss": 0.6988234519958496, + "step": 850 + }, + { + "epoch": 1.8012684989429175, + "grad_norm": 1.186838150024414, + "learning_rate": 6.609500667757022e-06, + "loss": 0.9463410377502441, + "step": 852 + }, + { + "epoch": 1.8054968287526427, + "grad_norm": 0.8933315873146057, + "learning_rate": 6.594245465909127e-06, + "loss": 0.7844456434249878, + "step": 854 + }, + { + "epoch": 1.809725158562368, + "grad_norm": 0.6125417351722717, + "learning_rate": 6.578976886470791e-06, + "loss": 1.0353469848632812, + "step": 856 + }, + { + "epoch": 1.8139534883720931, + "grad_norm": 0.9905397295951843, + "learning_rate": 6.56369511610653e-06, + "loss": 0.5985690951347351, + "step": 858 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 2.4766664505004883, + "learning_rate": 6.548400341642128e-06, + "loss": 1.0091896057128906, + "step": 860 + }, + { + "epoch": 1.8224101479915433, + "grad_norm": 1.0009154081344604, + "learning_rate": 6.533092750062345e-06, + "loss": 1.0143516063690186, + "step": 862 + }, + { + "epoch": 1.8266384778012685, + "grad_norm": 1.4460644721984863, + "learning_rate": 6.5177725285086416e-06, + "loss": 0.5153546929359436, + "step": 864 + }, + { + "epoch": 1.8308668076109935, + "grad_norm": 1.3811520338058472, + "learning_rate": 6.5024398642768785e-06, + "loss": 0.7575209736824036, + "step": 866 + }, + { + "epoch": 1.8350951374207187, + "grad_norm": 0.7286960482597351, + "learning_rate": 6.487094944815036e-06, + "loss": 1.0820739269256592, + "step": 868 + }, + { + "epoch": 1.839323467230444, + "grad_norm": 1.5409274101257324, + "learning_rate": 6.47173795772092e-06, + "loss": 1.0705763101577759, + "step": 870 + }, + { + "epoch": 1.8435517970401691, + "grad_norm": 0.4065259099006653, + "learning_rate": 6.456369090739869e-06, + "loss": 0.7662294507026672, + "step": 872 + }, + { + "epoch": 1.8477801268498943, + "grad_norm": 0.9650294780731201, + "learning_rate": 6.440988531762454e-06, + "loss": 1.0480531454086304, + "step": 874 + }, + { + "epoch": 1.8520084566596196, + "grad_norm": 1.9147449731826782, + "learning_rate": 6.42559646882219e-06, + "loss": 0.8815977573394775, + "step": 876 + }, + { + "epoch": 1.8562367864693448, + "grad_norm": 2.392439126968384, + "learning_rate": 6.410193090093227e-06, + "loss": 0.30479690432548523, + "step": 878 + }, + { + "epoch": 1.8604651162790697, + "grad_norm": 4.508836269378662, + "learning_rate": 6.394778583888064e-06, + "loss": 0.7044318318367004, + "step": 880 + }, + { + "epoch": 1.864693446088795, + "grad_norm": 0.8778241872787476, + "learning_rate": 6.379353138655227e-06, + "loss": 1.0541292428970337, + "step": 882 + }, + { + "epoch": 1.86892177589852, + "grad_norm": 0.8489974141120911, + "learning_rate": 6.3639169429769805e-06, + "loss": 0.7641475796699524, + "step": 884 + }, + { + "epoch": 1.8731501057082451, + "grad_norm": 1.3360224962234497, + "learning_rate": 6.348470185567025e-06, + "loss": 1.0800727605819702, + "step": 886 + }, + { + "epoch": 1.8773784355179703, + "grad_norm": 0.8119631409645081, + "learning_rate": 6.333013055268169e-06, + "loss": 1.0908565521240234, + "step": 888 + }, + { + "epoch": 1.8816067653276956, + "grad_norm": 1.3364355564117432, + "learning_rate": 6.317545741050044e-06, + "loss": 0.6000439524650574, + "step": 890 + }, + { + "epoch": 1.8858350951374208, + "grad_norm": 3.690155029296875, + "learning_rate": 6.3020684320067805e-06, + "loss": 0.8028129935264587, + "step": 892 + }, + { + "epoch": 1.890063424947146, + "grad_norm": 1.5241397619247437, + "learning_rate": 6.2865813173546976e-06, + "loss": 1.0165501832962036, + "step": 894 + }, + { + "epoch": 1.8942917547568712, + "grad_norm": 1.612496256828308, + "learning_rate": 6.271084586429994e-06, + "loss": 1.4110640287399292, + "step": 896 + }, + { + "epoch": 1.8985200845665962, + "grad_norm": 0.8049323558807373, + "learning_rate": 6.255578428686433e-06, + "loss": 1.0340838432312012, + "step": 898 + }, + { + "epoch": 1.9027484143763214, + "grad_norm": 1.7223362922668457, + "learning_rate": 6.240063033693023e-06, + "loss": 0.5132641792297363, + "step": 900 + }, + { + "epoch": 1.9069767441860463, + "grad_norm": 0.9025138020515442, + "learning_rate": 6.2245385911316994e-06, + "loss": 1.0327919721603394, + "step": 902 + }, + { + "epoch": 1.9112050739957716, + "grad_norm": 2.6092336177825928, + "learning_rate": 6.2090052907950096e-06, + "loss": 0.5075667500495911, + "step": 904 + }, + { + "epoch": 1.9154334038054968, + "grad_norm": 1.4433766603469849, + "learning_rate": 6.193463322583792e-06, + "loss": 0.8623234629631042, + "step": 906 + }, + { + "epoch": 1.919661733615222, + "grad_norm": 1.913678526878357, + "learning_rate": 6.177912876504851e-06, + "loss": 1.154498815536499, + "step": 908 + }, + { + "epoch": 1.9238900634249472, + "grad_norm": 0.7821553945541382, + "learning_rate": 6.16235414266864e-06, + "loss": 0.791877031326294, + "step": 910 + }, + { + "epoch": 1.9281183932346724, + "grad_norm": 0.6865972876548767, + "learning_rate": 6.146787311286927e-06, + "loss": 0.9883949160575867, + "step": 912 + }, + { + "epoch": 1.9323467230443976, + "grad_norm": 0.8712502717971802, + "learning_rate": 6.131212572670483e-06, + "loss": 0.6685277223587036, + "step": 914 + }, + { + "epoch": 1.9365750528541226, + "grad_norm": 0.7266220450401306, + "learning_rate": 6.115630117226741e-06, + "loss": 1.1048094034194946, + "step": 916 + }, + { + "epoch": 1.9408033826638478, + "grad_norm": 0.751155436038971, + "learning_rate": 6.1000401354574855e-06, + "loss": 0.6578397154808044, + "step": 918 + }, + { + "epoch": 1.945031712473573, + "grad_norm": 1.9933252334594727, + "learning_rate": 6.084442817956502e-06, + "loss": 0.8734629154205322, + "step": 920 + }, + { + "epoch": 1.949260042283298, + "grad_norm": 2.034205436706543, + "learning_rate": 6.0688383554072625e-06, + "loss": 0.9261573553085327, + "step": 922 + }, + { + "epoch": 1.9534883720930232, + "grad_norm": 1.4356502294540405, + "learning_rate": 6.053226938580596e-06, + "loss": 0.9625775814056396, + "step": 924 + }, + { + "epoch": 1.9577167019027484, + "grad_norm": 0.3833523988723755, + "learning_rate": 6.03760875833234e-06, + "loss": 0.8684751987457275, + "step": 926 + }, + { + "epoch": 1.9619450317124736, + "grad_norm": 1.4802016019821167, + "learning_rate": 6.021984005601027e-06, + "loss": 1.1143702268600464, + "step": 928 + }, + { + "epoch": 1.9661733615221988, + "grad_norm": 0.6003625392913818, + "learning_rate": 6.006352871405534e-06, + "loss": 0.7562581300735474, + "step": 930 + }, + { + "epoch": 1.970401691331924, + "grad_norm": 0.7981434464454651, + "learning_rate": 5.990715546842755e-06, + "loss": 1.2102357149124146, + "step": 932 + }, + { + "epoch": 1.9746300211416492, + "grad_norm": 1.7700377702713013, + "learning_rate": 5.9750722230852694e-06, + "loss": 0.8212454319000244, + "step": 934 + }, + { + "epoch": 1.9788583509513742, + "grad_norm": 1.064813494682312, + "learning_rate": 5.959423091378992e-06, + "loss": 0.7084140777587891, + "step": 936 + }, + { + "epoch": 1.9830866807610994, + "grad_norm": 1.5844800472259521, + "learning_rate": 5.943768343040845e-06, + "loss": 0.8270998597145081, + "step": 938 + }, + { + "epoch": 1.9873150105708244, + "grad_norm": 0.8206368088722229, + "learning_rate": 5.928108169456419e-06, + "loss": 1.0288496017456055, + "step": 940 + }, + { + "epoch": 1.9915433403805496, + "grad_norm": 0.76253741979599, + "learning_rate": 5.912442762077624e-06, + "loss": 1.049124836921692, + "step": 942 + }, + { + "epoch": 1.9957716701902748, + "grad_norm": 2.269414186477661, + "learning_rate": 5.89677231242036e-06, + "loss": 1.1043143272399902, + "step": 944 + }, + { + "epoch": 2.0, + "grad_norm": 1.246523141860962, + "learning_rate": 5.8810970120621714e-06, + "loss": 0.522446870803833, + "step": 946 + }, + { + "epoch": 2.004228329809725, + "grad_norm": 3.056257486343384, + "learning_rate": 5.8654170526399014e-06, + "loss": 0.7663121223449707, + "step": 948 + }, + { + "epoch": 2.0084566596194504, + "grad_norm": 0.66264808177948, + "learning_rate": 5.849732625847354e-06, + "loss": 0.926199197769165, + "step": 950 + }, + { + "epoch": 2.0126849894291756, + "grad_norm": 1.3620792627334595, + "learning_rate": 5.834043923432947e-06, + "loss": 0.8497916460037231, + "step": 952 + }, + { + "epoch": 2.016913319238901, + "grad_norm": 1.00492525100708, + "learning_rate": 5.818351137197373e-06, + "loss": 1.0884807109832764, + "step": 954 + }, + { + "epoch": 2.0211416490486256, + "grad_norm": 0.7676149606704712, + "learning_rate": 5.802654458991245e-06, + "loss": 1.0033034086227417, + "step": 956 + }, + { + "epoch": 2.025369978858351, + "grad_norm": 0.6691883206367493, + "learning_rate": 5.786954080712763e-06, + "loss": 0.6479755640029907, + "step": 958 + }, + { + "epoch": 2.029598308668076, + "grad_norm": 1.8425374031066895, + "learning_rate": 5.7712501943053576e-06, + "loss": 0.9424212574958801, + "step": 960 + }, + { + "epoch": 2.033826638477801, + "grad_norm": 1.5509213209152222, + "learning_rate": 5.755542991755352e-06, + "loss": 0.8941831588745117, + "step": 962 + }, + { + "epoch": 2.0380549682875264, + "grad_norm": 6.5713300704956055, + "learning_rate": 5.739832665089604e-06, + "loss": 0.7240588665008545, + "step": 964 + }, + { + "epoch": 2.0422832980972516, + "grad_norm": 1.1142792701721191, + "learning_rate": 5.724119406373171e-06, + "loss": 0.6287683844566345, + "step": 966 + }, + { + "epoch": 2.046511627906977, + "grad_norm": 4.120797157287598, + "learning_rate": 5.708403407706954e-06, + "loss": 0.740992546081543, + "step": 968 + }, + { + "epoch": 2.050739957716702, + "grad_norm": 2.123906373977661, + "learning_rate": 5.692684861225349e-06, + "loss": 0.9332213401794434, + "step": 970 + }, + { + "epoch": 2.0549682875264272, + "grad_norm": 2.3211607933044434, + "learning_rate": 5.676963959093904e-06, + "loss": 0.5057845115661621, + "step": 972 + }, + { + "epoch": 2.059196617336152, + "grad_norm": 0.8356497883796692, + "learning_rate": 5.66124089350696e-06, + "loss": 1.009197473526001, + "step": 974 + }, + { + "epoch": 2.063424947145877, + "grad_norm": 7.17715311050415, + "learning_rate": 5.645515856685316e-06, + "loss": 0.9003053307533264, + "step": 976 + }, + { + "epoch": 2.0676532769556024, + "grad_norm": 0.9301227331161499, + "learning_rate": 5.629789040873862e-06, + "loss": 0.9196462631225586, + "step": 978 + }, + { + "epoch": 2.0718816067653276, + "grad_norm": 4.0430145263671875, + "learning_rate": 5.614060638339238e-06, + "loss": 0.7268422245979309, + "step": 980 + }, + { + "epoch": 2.076109936575053, + "grad_norm": 4.7748188972473145, + "learning_rate": 5.598330841367487e-06, + "loss": 0.7036841511726379, + "step": 982 + }, + { + "epoch": 2.080338266384778, + "grad_norm": 2.2791452407836914, + "learning_rate": 5.582599842261693e-06, + "loss": 0.6845812797546387, + "step": 984 + }, + { + "epoch": 2.0845665961945032, + "grad_norm": 1.4591090679168701, + "learning_rate": 5.566867833339644e-06, + "loss": 0.7946863174438477, + "step": 986 + }, + { + "epoch": 2.0887949260042284, + "grad_norm": 1.36851167678833, + "learning_rate": 5.551135006931467e-06, + "loss": 0.9647249579429626, + "step": 988 + }, + { + "epoch": 2.0930232558139537, + "grad_norm": 1.2267500162124634, + "learning_rate": 5.535401555377287e-06, + "loss": 0.8777101039886475, + "step": 990 + }, + { + "epoch": 2.097251585623679, + "grad_norm": 2.077432632446289, + "learning_rate": 5.51966767102487e-06, + "loss": 0.5560998916625977, + "step": 992 + }, + { + "epoch": 2.1014799154334036, + "grad_norm": 1.9206690788269043, + "learning_rate": 5.503933546227273e-06, + "loss": 0.6140750646591187, + "step": 994 + }, + { + "epoch": 2.105708245243129, + "grad_norm": 3.3833982944488525, + "learning_rate": 5.4881993733404916e-06, + "loss": 0.6901674270629883, + "step": 996 + }, + { + "epoch": 2.109936575052854, + "grad_norm": 0.9377997517585754, + "learning_rate": 5.472465344721113e-06, + "loss": 0.7952163815498352, + "step": 998 + }, + { + "epoch": 2.1141649048625792, + "grad_norm": 4.0864057540893555, + "learning_rate": 5.456731652723959e-06, + "loss": 0.9408197402954102, + "step": 1000 + }, + { + "epoch": 2.1183932346723044, + "grad_norm": 0.6997259259223938, + "learning_rate": 5.440998489699731e-06, + "loss": 0.2278219759464264, + "step": 1002 + }, + { + "epoch": 2.1226215644820297, + "grad_norm": 2.893739700317383, + "learning_rate": 5.425266047992672e-06, + "loss": 0.6165475845336914, + "step": 1004 + }, + { + "epoch": 2.126849894291755, + "grad_norm": 0.8575990200042725, + "learning_rate": 5.4095345199382e-06, + "loss": 0.8270474076271057, + "step": 1006 + }, + { + "epoch": 2.13107822410148, + "grad_norm": 2.945441722869873, + "learning_rate": 5.393804097860568e-06, + "loss": 0.8482134342193604, + "step": 1008 + }, + { + "epoch": 2.1353065539112053, + "grad_norm": 1.1243295669555664, + "learning_rate": 5.3780749740705065e-06, + "loss": 0.6041120290756226, + "step": 1010 + }, + { + "epoch": 2.13953488372093, + "grad_norm": 3.6248466968536377, + "learning_rate": 5.362347340862869e-06, + "loss": 0.48887473344802856, + "step": 1012 + }, + { + "epoch": 2.1437632135306552, + "grad_norm": 0.8395848274230957, + "learning_rate": 5.3466213905142935e-06, + "loss": 0.5460640788078308, + "step": 1014 + }, + { + "epoch": 2.1479915433403804, + "grad_norm": 0.8920825719833374, + "learning_rate": 5.3308973152808395e-06, + "loss": 0.7917551398277283, + "step": 1016 + }, + { + "epoch": 2.1522198731501057, + "grad_norm": 0.8760237693786621, + "learning_rate": 5.315175307395644e-06, + "loss": 0.8787223100662231, + "step": 1018 + }, + { + "epoch": 2.156448202959831, + "grad_norm": 1.4028397798538208, + "learning_rate": 5.299455559066571e-06, + "loss": 0.6539137363433838, + "step": 1020 + }, + { + "epoch": 2.160676532769556, + "grad_norm": 0.9284740686416626, + "learning_rate": 5.283738262473859e-06, + "loss": 0.6028532981872559, + "step": 1022 + }, + { + "epoch": 2.1649048625792813, + "grad_norm": 1.5914058685302734, + "learning_rate": 5.268023609767772e-06, + "loss": 0.9632537364959717, + "step": 1024 + }, + { + "epoch": 2.1691331923890065, + "grad_norm": 1.118626594543457, + "learning_rate": 5.2523117930662545e-06, + "loss": 0.8900113701820374, + "step": 1026 + }, + { + "epoch": 2.1733615221987317, + "grad_norm": 1.0411882400512695, + "learning_rate": 5.2366030044525785e-06, + "loss": 0.88353431224823, + "step": 1028 + }, + { + "epoch": 2.177589852008457, + "grad_norm": 1.1495089530944824, + "learning_rate": 5.220897435972992e-06, + "loss": 0.8420016765594482, + "step": 1030 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 0.9872754812240601, + "learning_rate": 5.205195279634385e-06, + "loss": 0.8878940343856812, + "step": 1032 + }, + { + "epoch": 2.186046511627907, + "grad_norm": 0.8607329726219177, + "learning_rate": 5.1894967274019245e-06, + "loss": 1.0447912216186523, + "step": 1034 + }, + { + "epoch": 2.190274841437632, + "grad_norm": 0.7523432970046997, + "learning_rate": 5.173801971196716e-06, + "loss": 0.6362751126289368, + "step": 1036 + }, + { + "epoch": 2.1945031712473573, + "grad_norm": 5.9548540115356445, + "learning_rate": 5.158111202893464e-06, + "loss": 0.44573909044265747, + "step": 1038 + }, + { + "epoch": 2.1987315010570825, + "grad_norm": 1.2779605388641357, + "learning_rate": 5.142424614318112e-06, + "loss": 0.9875360131263733, + "step": 1040 + }, + { + "epoch": 2.2029598308668077, + "grad_norm": 2.124209403991699, + "learning_rate": 5.12674239724551e-06, + "loss": 0.3518787622451782, + "step": 1042 + }, + { + "epoch": 2.207188160676533, + "grad_norm": 0.7920838594436646, + "learning_rate": 5.111064743397061e-06, + "loss": 0.9684445858001709, + "step": 1044 + }, + { + "epoch": 2.211416490486258, + "grad_norm": 0.8072055578231812, + "learning_rate": 5.09539184443838e-06, + "loss": 0.4609597623348236, + "step": 1046 + }, + { + "epoch": 2.2156448202959833, + "grad_norm": 1.860408067703247, + "learning_rate": 5.079723891976956e-06, + "loss": 0.7287152409553528, + "step": 1048 + }, + { + "epoch": 2.219873150105708, + "grad_norm": 2.7022407054901123, + "learning_rate": 5.064061077559803e-06, + "loss": 0.5810937285423279, + "step": 1050 + }, + { + "epoch": 2.2241014799154333, + "grad_norm": 1.2696058750152588, + "learning_rate": 5.048403592671116e-06, + "loss": 0.8156327605247498, + "step": 1052 + }, + { + "epoch": 2.2283298097251585, + "grad_norm": 0.9386839270591736, + "learning_rate": 5.032751628729943e-06, + "loss": 1.0226250886917114, + "step": 1054 + }, + { + "epoch": 2.2325581395348837, + "grad_norm": 0.5965794920921326, + "learning_rate": 5.017105377087829e-06, + "loss": 0.9035366773605347, + "step": 1056 + }, + { + "epoch": 2.236786469344609, + "grad_norm": 0.3065854012966156, + "learning_rate": 5.0014650290264854e-06, + "loss": 0.6141324639320374, + "step": 1058 + }, + { + "epoch": 2.241014799154334, + "grad_norm": 0.6253087520599365, + "learning_rate": 4.9858307757554536e-06, + "loss": 0.7122809290885925, + "step": 1060 + }, + { + "epoch": 2.2452431289640593, + "grad_norm": 0.6780251264572144, + "learning_rate": 4.97020280840976e-06, + "loss": 0.8688558340072632, + "step": 1062 + }, + { + "epoch": 2.2494714587737845, + "grad_norm": 4.11038875579834, + "learning_rate": 4.954581318047582e-06, + "loss": 0.8956253528594971, + "step": 1064 + }, + { + "epoch": 2.2536997885835097, + "grad_norm": 1.39890456199646, + "learning_rate": 4.938966495647918e-06, + "loss": 0.5270295143127441, + "step": 1066 + }, + { + "epoch": 2.2579281183932345, + "grad_norm": 2.004953384399414, + "learning_rate": 4.923358532108244e-06, + "loss": 0.7445597052574158, + "step": 1068 + }, + { + "epoch": 2.2621564482029597, + "grad_norm": 3.2700140476226807, + "learning_rate": 4.907757618242186e-06, + "loss": 0.7111672759056091, + "step": 1070 + }, + { + "epoch": 2.266384778012685, + "grad_norm": 1.6366181373596191, + "learning_rate": 4.892163944777181e-06, + "loss": 0.9749158620834351, + "step": 1072 + }, + { + "epoch": 2.27061310782241, + "grad_norm": 0.5849347710609436, + "learning_rate": 4.876577702352155e-06, + "loss": 0.575994074344635, + "step": 1074 + }, + { + "epoch": 2.2748414376321353, + "grad_norm": 4.112948417663574, + "learning_rate": 4.860999081515183e-06, + "loss": 0.7574913501739502, + "step": 1076 + }, + { + "epoch": 2.2790697674418605, + "grad_norm": 1.8271198272705078, + "learning_rate": 4.845428272721161e-06, + "loss": 0.21380315721035004, + "step": 1078 + }, + { + "epoch": 2.2832980972515857, + "grad_norm": 0.6084461212158203, + "learning_rate": 4.8298654663294865e-06, + "loss": 0.1430993676185608, + "step": 1080 + }, + { + "epoch": 2.287526427061311, + "grad_norm": 0.3571109473705292, + "learning_rate": 4.814310852601717e-06, + "loss": 0.45104092359542847, + "step": 1082 + }, + { + "epoch": 2.291754756871036, + "grad_norm": 0.889111340045929, + "learning_rate": 4.7987646216992545e-06, + "loss": 0.3398406505584717, + "step": 1084 + }, + { + "epoch": 2.295983086680761, + "grad_norm": 0.28270912170410156, + "learning_rate": 4.783226963681018e-06, + "loss": 0.14048698544502258, + "step": 1086 + }, + { + "epoch": 2.300211416490486, + "grad_norm": 18.686655044555664, + "learning_rate": 4.76769806850112e-06, + "loss": 0.579992949962616, + "step": 1088 + }, + { + "epoch": 2.3044397463002113, + "grad_norm": 0.886323869228363, + "learning_rate": 4.752178126006544e-06, + "loss": 0.8695411682128906, + "step": 1090 + }, + { + "epoch": 2.3086680761099365, + "grad_norm": 1.4230542182922363, + "learning_rate": 4.736667325934819e-06, + "loss": 0.5931753516197205, + "step": 1092 + }, + { + "epoch": 2.3128964059196617, + "grad_norm": 1.0836153030395508, + "learning_rate": 4.721165857911709e-06, + "loss": 0.884992778301239, + "step": 1094 + }, + { + "epoch": 2.317124735729387, + "grad_norm": 1.3372290134429932, + "learning_rate": 4.705673911448891e-06, + "loss": 0.8042216897010803, + "step": 1096 + }, + { + "epoch": 2.321353065539112, + "grad_norm": 1.6718645095825195, + "learning_rate": 4.690191675941631e-06, + "loss": 0.8200917840003967, + "step": 1098 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 4.638646602630615, + "learning_rate": 4.674719340666477e-06, + "loss": 0.4548729360103607, + "step": 1100 + }, + { + "epoch": 2.3298097251585626, + "grad_norm": 1.269760012626648, + "learning_rate": 4.659257094778949e-06, + "loss": 0.6183570027351379, + "step": 1102 + }, + { + "epoch": 2.3340380549682873, + "grad_norm": 1.3631540536880493, + "learning_rate": 4.643805127311212e-06, + "loss": 0.9078647494316101, + "step": 1104 + }, + { + "epoch": 2.3382663847780125, + "grad_norm": 0.8146481513977051, + "learning_rate": 4.628363627169774e-06, + "loss": 0.8710769414901733, + "step": 1106 + }, + { + "epoch": 2.3424947145877377, + "grad_norm": 1.6327295303344727, + "learning_rate": 4.612932783133182e-06, + "loss": 0.896135151386261, + "step": 1108 + }, + { + "epoch": 2.346723044397463, + "grad_norm": 0.9389740824699402, + "learning_rate": 4.5975127838497036e-06, + "loss": 1.004955530166626, + "step": 1110 + }, + { + "epoch": 2.350951374207188, + "grad_norm": 0.33022138476371765, + "learning_rate": 4.582103817835021e-06, + "loss": 0.5390142202377319, + "step": 1112 + }, + { + "epoch": 2.3551797040169133, + "grad_norm": 2.952239990234375, + "learning_rate": 4.566706073469938e-06, + "loss": 0.7841445803642273, + "step": 1114 + }, + { + "epoch": 2.3594080338266386, + "grad_norm": 1.6382222175598145, + "learning_rate": 4.551319738998067e-06, + "loss": 0.5400862097740173, + "step": 1116 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 2.5125372409820557, + "learning_rate": 4.535945002523524e-06, + "loss": 0.962687611579895, + "step": 1118 + }, + { + "epoch": 2.367864693446089, + "grad_norm": 1.6231260299682617, + "learning_rate": 4.5205820520086415e-06, + "loss": 0.9014071226119995, + "step": 1120 + }, + { + "epoch": 2.3720930232558137, + "grad_norm": 1.2299134731292725, + "learning_rate": 4.505231075271662e-06, + "loss": 0.7179723381996155, + "step": 1122 + }, + { + "epoch": 2.376321353065539, + "grad_norm": 1.7179030179977417, + "learning_rate": 4.489892259984446e-06, + "loss": 0.6040488481521606, + "step": 1124 + }, + { + "epoch": 2.380549682875264, + "grad_norm": 3.960186719894409, + "learning_rate": 4.474565793670167e-06, + "loss": 0.8589836359024048, + "step": 1126 + }, + { + "epoch": 2.3847780126849893, + "grad_norm": 21.694618225097656, + "learning_rate": 4.459251863701038e-06, + "loss": 0.6169445514678955, + "step": 1128 + }, + { + "epoch": 2.3890063424947146, + "grad_norm": 1.0652092695236206, + "learning_rate": 4.443950657296004e-06, + "loss": 0.6405174732208252, + "step": 1130 + }, + { + "epoch": 2.3932346723044398, + "grad_norm": 2.09255313873291, + "learning_rate": 4.4286623615184585e-06, + "loss": 0.6314718127250671, + "step": 1132 + }, + { + "epoch": 2.397463002114165, + "grad_norm": 1.2623828649520874, + "learning_rate": 4.413387163273964e-06, + "loss": 0.9738779664039612, + "step": 1134 + }, + { + "epoch": 2.40169133192389, + "grad_norm": 4.026132106781006, + "learning_rate": 4.3981252493079506e-06, + "loss": 0.6053364872932434, + "step": 1136 + }, + { + "epoch": 2.4059196617336154, + "grad_norm": 1.6954033374786377, + "learning_rate": 4.382876806203453e-06, + "loss": 0.7640853524208069, + "step": 1138 + }, + { + "epoch": 2.41014799154334, + "grad_norm": 0.5995383262634277, + "learning_rate": 4.3676420203788125e-06, + "loss": 0.8452503085136414, + "step": 1140 + }, + { + "epoch": 2.4143763213530653, + "grad_norm": 1.4022237062454224, + "learning_rate": 4.352421078085405e-06, + "loss": 0.9150977730751038, + "step": 1142 + }, + { + "epoch": 2.4186046511627906, + "grad_norm": 1.1480714082717896, + "learning_rate": 4.337214165405368e-06, + "loss": 0.47342145442962646, + "step": 1144 + }, + { + "epoch": 2.4228329809725158, + "grad_norm": 3.3833422660827637, + "learning_rate": 4.322021468249316e-06, + "loss": 0.6564927697181702, + "step": 1146 + }, + { + "epoch": 2.427061310782241, + "grad_norm": 0.8458242416381836, + "learning_rate": 4.306843172354076e-06, + "loss": 0.787387490272522, + "step": 1148 + }, + { + "epoch": 2.431289640591966, + "grad_norm": 1.818563461303711, + "learning_rate": 4.291679463280416e-06, + "loss": 0.5408893823623657, + "step": 1150 + }, + { + "epoch": 2.4355179704016914, + "grad_norm": 0.5807977914810181, + "learning_rate": 4.276530526410768e-06, + "loss": 0.523176908493042, + "step": 1152 + }, + { + "epoch": 2.4397463002114166, + "grad_norm": 0.718356192111969, + "learning_rate": 4.261396546946972e-06, + "loss": 0.4822205901145935, + "step": 1154 + }, + { + "epoch": 2.443974630021142, + "grad_norm": 0.7175132036209106, + "learning_rate": 4.246277709908009e-06, + "loss": 0.8414809703826904, + "step": 1156 + }, + { + "epoch": 2.448202959830867, + "grad_norm": 1.0660876035690308, + "learning_rate": 4.231174200127736e-06, + "loss": 0.48800867795944214, + "step": 1158 + }, + { + "epoch": 2.452431289640592, + "grad_norm": 0.8460950255393982, + "learning_rate": 4.216086202252624e-06, + "loss": 0.8563118577003479, + "step": 1160 + }, + { + "epoch": 2.456659619450317, + "grad_norm": 0.835564911365509, + "learning_rate": 4.201013900739513e-06, + "loss": 0.8780640363693237, + "step": 1162 + }, + { + "epoch": 2.460887949260042, + "grad_norm": 0.8362522721290588, + "learning_rate": 4.185957479853344e-06, + "loss": 0.6378390789031982, + "step": 1164 + }, + { + "epoch": 2.4651162790697674, + "grad_norm": 1.15602707862854, + "learning_rate": 4.1709171236649096e-06, + "loss": 0.9949501156806946, + "step": 1166 + }, + { + "epoch": 2.4693446088794926, + "grad_norm": 1.0430980920791626, + "learning_rate": 4.155893016048607e-06, + "loss": 1.102187991142273, + "step": 1168 + }, + { + "epoch": 2.473572938689218, + "grad_norm": 1.029948115348816, + "learning_rate": 4.140885340680189e-06, + "loss": 0.7335580587387085, + "step": 1170 + }, + { + "epoch": 2.477801268498943, + "grad_norm": 2.0752034187316895, + "learning_rate": 4.125894281034517e-06, + "loss": 0.67908775806427, + "step": 1172 + }, + { + "epoch": 2.482029598308668, + "grad_norm": 0.9801737666130066, + "learning_rate": 4.110920020383316e-06, + "loss": 0.6409579515457153, + "step": 1174 + }, + { + "epoch": 2.4862579281183934, + "grad_norm": 4.96562385559082, + "learning_rate": 4.0959627417929386e-06, + "loss": 0.66353440284729, + "step": 1176 + }, + { + "epoch": 2.4904862579281186, + "grad_norm": 1.388858437538147, + "learning_rate": 4.081022628122127e-06, + "loss": 0.7435889840126038, + "step": 1178 + }, + { + "epoch": 2.4947145877378434, + "grad_norm": 3.9195332527160645, + "learning_rate": 4.06609986201977e-06, + "loss": 0.5540118217468262, + "step": 1180 + }, + { + "epoch": 2.4989429175475686, + "grad_norm": 0.45918816328048706, + "learning_rate": 4.051194625922681e-06, + "loss": 0.34914258122444153, + "step": 1182 + }, + { + "epoch": 2.503171247357294, + "grad_norm": 0.49878963828086853, + "learning_rate": 4.036307102053359e-06, + "loss": 0.12513917684555054, + "step": 1184 + }, + { + "epoch": 2.507399577167019, + "grad_norm": 1.552511215209961, + "learning_rate": 4.021437472417763e-06, + "loss": 0.8537716269493103, + "step": 1186 + }, + { + "epoch": 2.511627906976744, + "grad_norm": 3.08944034576416, + "learning_rate": 4.006585918803091e-06, + "loss": 0.7219171524047852, + "step": 1188 + }, + { + "epoch": 2.5158562367864694, + "grad_norm": 2.770634412765503, + "learning_rate": 3.991752622775551e-06, + "loss": 0.5876720547676086, + "step": 1190 + }, + { + "epoch": 2.5200845665961946, + "grad_norm": 0.8611694574356079, + "learning_rate": 3.97693776567815e-06, + "loss": 0.8831681609153748, + "step": 1192 + }, + { + "epoch": 2.52431289640592, + "grad_norm": 1.6146912574768066, + "learning_rate": 3.962141528628465e-06, + "loss": 0.5676584839820862, + "step": 1194 + }, + { + "epoch": 2.528541226215645, + "grad_norm": 2.445615291595459, + "learning_rate": 3.947364092516437e-06, + "loss": 0.8747268319129944, + "step": 1196 + }, + { + "epoch": 2.53276955602537, + "grad_norm": 1.0791140794754028, + "learning_rate": 3.932605638002163e-06, + "loss": 0.8514158725738525, + "step": 1198 + }, + { + "epoch": 2.536997885835095, + "grad_norm": 4.930743217468262, + "learning_rate": 3.917866345513677e-06, + "loss": 0.7207362651824951, + "step": 1200 + }, + { + "epoch": 2.54122621564482, + "grad_norm": 1.0629011392593384, + "learning_rate": 3.903146395244749e-06, + "loss": 0.9177104234695435, + "step": 1202 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 3.7193496227264404, + "learning_rate": 3.888445967152691e-06, + "loss": 0.33925938606262207, + "step": 1204 + }, + { + "epoch": 2.5496828752642706, + "grad_norm": 1.2090929746627808, + "learning_rate": 3.873765240956138e-06, + "loss": 1.0040740966796875, + "step": 1206 + }, + { + "epoch": 2.553911205073996, + "grad_norm": 2.6415557861328125, + "learning_rate": 3.859104396132863e-06, + "loss": 0.990748941898346, + "step": 1208 + }, + { + "epoch": 2.558139534883721, + "grad_norm": 2.192115545272827, + "learning_rate": 3.84446361191759e-06, + "loss": 0.9163404703140259, + "step": 1210 + }, + { + "epoch": 2.5623678646934462, + "grad_norm": 3.1367952823638916, + "learning_rate": 3.829843067299788e-06, + "loss": 0.6369563341140747, + "step": 1212 + }, + { + "epoch": 2.5665961945031714, + "grad_norm": 1.139792799949646, + "learning_rate": 3.8152429410214834e-06, + "loss": 0.8746934533119202, + "step": 1214 + }, + { + "epoch": 2.570824524312896, + "grad_norm": 2.18265438079834, + "learning_rate": 3.8006634115750886e-06, + "loss": 0.5589424967765808, + "step": 1216 + }, + { + "epoch": 2.5750528541226214, + "grad_norm": 1.11671781539917, + "learning_rate": 3.7861046572012105e-06, + "loss": 0.49788519740104675, + "step": 1218 + }, + { + "epoch": 2.5792811839323466, + "grad_norm": 2.9104678630828857, + "learning_rate": 3.771566855886467e-06, + "loss": 0.4844942092895508, + "step": 1220 + }, + { + "epoch": 2.583509513742072, + "grad_norm": 0.8437510132789612, + "learning_rate": 3.7570501853613227e-06, + "loss": 0.4949146509170532, + "step": 1222 + }, + { + "epoch": 2.587737843551797, + "grad_norm": 3.4612996578216553, + "learning_rate": 3.742554823097905e-06, + "loss": 0.7449167966842651, + "step": 1224 + }, + { + "epoch": 2.5919661733615222, + "grad_norm": 0.35922661423683167, + "learning_rate": 3.7280809463078436e-06, + "loss": 0.33840128779411316, + "step": 1226 + }, + { + "epoch": 2.5961945031712474, + "grad_norm": 1.0260803699493408, + "learning_rate": 3.7136287319400944e-06, + "loss": 0.7077632546424866, + "step": 1228 + }, + { + "epoch": 2.6004228329809727, + "grad_norm": 1.4719406366348267, + "learning_rate": 3.6991983566787847e-06, + "loss": 0.7926508188247681, + "step": 1230 + }, + { + "epoch": 2.604651162790698, + "grad_norm": 2.892934799194336, + "learning_rate": 3.6847899969410515e-06, + "loss": 0.6512648463249207, + "step": 1232 + }, + { + "epoch": 2.6088794926004226, + "grad_norm": 6.283431529998779, + "learning_rate": 3.6704038288748812e-06, + "loss": 0.7001049518585205, + "step": 1234 + }, + { + "epoch": 2.613107822410148, + "grad_norm": 0.9000669121742249, + "learning_rate": 3.6560400283569543e-06, + "loss": 0.7005571722984314, + "step": 1236 + }, + { + "epoch": 2.617336152219873, + "grad_norm": 1.3975003957748413, + "learning_rate": 3.641698770990507e-06, + "loss": 0.8292574882507324, + "step": 1238 + }, + { + "epoch": 2.6215644820295982, + "grad_norm": 0.9919524192810059, + "learning_rate": 3.6273802321031738e-06, + "loss": 0.6096389889717102, + "step": 1240 + }, + { + "epoch": 2.6257928118393234, + "grad_norm": 1.2537214756011963, + "learning_rate": 3.6130845867448384e-06, + "loss": 0.874812126159668, + "step": 1242 + }, + { + "epoch": 2.6300211416490487, + "grad_norm": 1.2933173179626465, + "learning_rate": 3.598812009685514e-06, + "loss": 0.8653225898742676, + "step": 1244 + }, + { + "epoch": 2.634249471458774, + "grad_norm": 2.738640308380127, + "learning_rate": 3.584562675413189e-06, + "loss": 0.5823553800582886, + "step": 1246 + }, + { + "epoch": 2.638477801268499, + "grad_norm": 2.1639490127563477, + "learning_rate": 3.5703367581316985e-06, + "loss": 0.5077102780342102, + "step": 1248 + }, + { + "epoch": 2.6427061310782243, + "grad_norm": 0.9785355925559998, + "learning_rate": 3.5561344317585957e-06, + "loss": 0.9332320094108582, + "step": 1250 + }, + { + "epoch": 2.646934460887949, + "grad_norm": 9.316052436828613, + "learning_rate": 3.5419558699230305e-06, + "loss": 0.4240643084049225, + "step": 1252 + }, + { + "epoch": 2.6511627906976747, + "grad_norm": 1.8104251623153687, + "learning_rate": 3.52780124596361e-06, + "loss": 0.6087803840637207, + "step": 1254 + }, + { + "epoch": 2.6553911205073994, + "grad_norm": 1.1407182216644287, + "learning_rate": 3.5136707329263055e-06, + "loss": 0.8913679122924805, + "step": 1256 + }, + { + "epoch": 2.6596194503171247, + "grad_norm": 0.6544500589370728, + "learning_rate": 3.499564503562315e-06, + "loss": 0.22951415181159973, + "step": 1258 + }, + { + "epoch": 2.66384778012685, + "grad_norm": 0.4541882574558258, + "learning_rate": 3.4854827303259575e-06, + "loss": 0.44140568375587463, + "step": 1260 + }, + { + "epoch": 2.668076109936575, + "grad_norm": 4.755753040313721, + "learning_rate": 3.47142558537257e-06, + "loss": 0.6385686993598938, + "step": 1262 + }, + { + "epoch": 2.6723044397463003, + "grad_norm": 3.3811631202697754, + "learning_rate": 3.4573932405563966e-06, + "loss": 0.6937937140464783, + "step": 1264 + }, + { + "epoch": 2.6765327695560255, + "grad_norm": 5.263582229614258, + "learning_rate": 3.4433858674284894e-06, + "loss": 0.3113529682159424, + "step": 1266 + }, + { + "epoch": 2.6807610993657507, + "grad_norm": 1.3352901935577393, + "learning_rate": 3.429403637234614e-06, + "loss": 0.5825721621513367, + "step": 1268 + }, + { + "epoch": 2.6849894291754755, + "grad_norm": 1.189872145652771, + "learning_rate": 3.4154467209131515e-06, + "loss": 0.8304330706596375, + "step": 1270 + }, + { + "epoch": 2.689217758985201, + "grad_norm": 2.888284206390381, + "learning_rate": 3.401515289093011e-06, + "loss": 0.6794753074645996, + "step": 1272 + }, + { + "epoch": 2.693446088794926, + "grad_norm": 0.586315393447876, + "learning_rate": 3.3876095120915433e-06, + "loss": 0.32759660482406616, + "step": 1274 + }, + { + "epoch": 2.697674418604651, + "grad_norm": 1.7027820348739624, + "learning_rate": 3.3737295599124592e-06, + "loss": 0.3890809416770935, + "step": 1276 + }, + { + "epoch": 2.7019027484143763, + "grad_norm": 2.2275924682617188, + "learning_rate": 3.3598756022437507e-06, + "loss": 0.8529485464096069, + "step": 1278 + }, + { + "epoch": 2.7061310782241015, + "grad_norm": 2.0538721084594727, + "learning_rate": 3.3460478084556157e-06, + "loss": 0.6433102488517761, + "step": 1280 + }, + { + "epoch": 2.7103594080338267, + "grad_norm": 0.8136905431747437, + "learning_rate": 3.33224634759839e-06, + "loss": 1.027058482170105, + "step": 1282 + }, + { + "epoch": 2.714587737843552, + "grad_norm": 0.376529723405838, + "learning_rate": 3.318471388400476e-06, + "loss": 0.5974118709564209, + "step": 1284 + }, + { + "epoch": 2.718816067653277, + "grad_norm": 0.36886441707611084, + "learning_rate": 3.3047230992662834e-06, + "loss": 0.8987646698951721, + "step": 1286 + }, + { + "epoch": 2.723044397463002, + "grad_norm": 0.6800359487533569, + "learning_rate": 3.291001648274175e-06, + "loss": 0.4526841640472412, + "step": 1288 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 0.7832357883453369, + "learning_rate": 3.277307203174395e-06, + "loss": 0.8279471397399902, + "step": 1290 + }, + { + "epoch": 2.7315010570824523, + "grad_norm": 0.7864487767219543, + "learning_rate": 3.263639931387039e-06, + "loss": 0.9575145840644836, + "step": 1292 + }, + { + "epoch": 2.7357293868921775, + "grad_norm": 5.665610313415527, + "learning_rate": 3.2500000000000015e-06, + "loss": 0.9256079792976379, + "step": 1294 + }, + { + "epoch": 2.7399577167019027, + "grad_norm": 0.6406017541885376, + "learning_rate": 3.236387575766916e-06, + "loss": 0.25800952315330505, + "step": 1296 + }, + { + "epoch": 2.744186046511628, + "grad_norm": 0.8897634148597717, + "learning_rate": 3.222802825105138e-06, + "loss": 0.49922552704811096, + "step": 1298 + }, + { + "epoch": 2.748414376321353, + "grad_norm": 0.8615605235099792, + "learning_rate": 3.209245914093708e-06, + "loss": 0.6280668377876282, + "step": 1300 + }, + { + "epoch": 2.7526427061310783, + "grad_norm": 1.0332123041152954, + "learning_rate": 3.1957170084713053e-06, + "loss": 0.6883167624473572, + "step": 1302 + }, + { + "epoch": 2.7568710359408035, + "grad_norm": 0.7236963510513306, + "learning_rate": 3.182216273634235e-06, + "loss": 0.5495542287826538, + "step": 1304 + }, + { + "epoch": 2.7610993657505283, + "grad_norm": 2.5494091510772705, + "learning_rate": 3.1687438746344086e-06, + "loss": 0.7590460181236267, + "step": 1306 + }, + { + "epoch": 2.765327695560254, + "grad_norm": 1.2892471551895142, + "learning_rate": 3.1552999761773166e-06, + "loss": 0.8913841843605042, + "step": 1308 + }, + { + "epoch": 2.7695560253699787, + "grad_norm": 1.4586549997329712, + "learning_rate": 3.141884742620013e-06, + "loss": 0.9050283432006836, + "step": 1310 + }, + { + "epoch": 2.773784355179704, + "grad_norm": 0.8337520956993103, + "learning_rate": 3.1284983379691247e-06, + "loss": 0.7146180272102356, + "step": 1312 + }, + { + "epoch": 2.778012684989429, + "grad_norm": 0.9738405346870422, + "learning_rate": 3.1151409258788235e-06, + "loss": 0.4363844096660614, + "step": 1314 + }, + { + "epoch": 2.7822410147991543, + "grad_norm": 1.9106115102767944, + "learning_rate": 3.101812669648838e-06, + "loss": 0.8666545748710632, + "step": 1316 + }, + { + "epoch": 2.7864693446088795, + "grad_norm": 0.2939179539680481, + "learning_rate": 3.0885137322224546e-06, + "loss": 0.5686661005020142, + "step": 1318 + }, + { + "epoch": 2.7906976744186047, + "grad_norm": 1.287136197090149, + "learning_rate": 3.075244276184529e-06, + "loss": 0.7593466639518738, + "step": 1320 + }, + { + "epoch": 2.79492600422833, + "grad_norm": 22.401884078979492, + "learning_rate": 3.0620044637594936e-06, + "loss": 0.3257228434085846, + "step": 1322 + }, + { + "epoch": 2.7991543340380547, + "grad_norm": 1.1438804864883423, + "learning_rate": 3.0487944568093685e-06, + "loss": 0.8988646864891052, + "step": 1324 + }, + { + "epoch": 2.8033826638477803, + "grad_norm": 2.6712405681610107, + "learning_rate": 3.035614416831801e-06, + "loss": 0.21109738945960999, + "step": 1326 + }, + { + "epoch": 2.807610993657505, + "grad_norm": 2.2659363746643066, + "learning_rate": 3.0224645049580746e-06, + "loss": 0.3605520725250244, + "step": 1328 + }, + { + "epoch": 2.8118393234672303, + "grad_norm": 1.1416643857955933, + "learning_rate": 3.0093448819511416e-06, + "loss": 0.4961351454257965, + "step": 1330 + }, + { + "epoch": 2.8160676532769555, + "grad_norm": 1.1021716594696045, + "learning_rate": 2.9962557082036677e-06, + "loss": 0.859085738658905, + "step": 1332 + }, + { + "epoch": 2.8202959830866807, + "grad_norm": 1.0480974912643433, + "learning_rate": 2.9831971437360606e-06, + "loss": 0.6091691255569458, + "step": 1334 + }, + { + "epoch": 2.824524312896406, + "grad_norm": 1.2733410596847534, + "learning_rate": 2.9701693481945132e-06, + "loss": 0.9933862686157227, + "step": 1336 + }, + { + "epoch": 2.828752642706131, + "grad_norm": 0.9854733943939209, + "learning_rate": 2.957172480849064e-06, + "loss": 0.9590897560119629, + "step": 1338 + }, + { + "epoch": 2.8329809725158563, + "grad_norm": 0.9489989876747131, + "learning_rate": 2.9442067005916375e-06, + "loss": 0.6300552487373352, + "step": 1340 + }, + { + "epoch": 2.8372093023255816, + "grad_norm": 1.5359514951705933, + "learning_rate": 2.931272165934105e-06, + "loss": 0.8570085167884827, + "step": 1342 + }, + { + "epoch": 2.8414376321353068, + "grad_norm": 0.8532953262329102, + "learning_rate": 2.918369035006351e-06, + "loss": 0.6232531070709229, + "step": 1344 + }, + { + "epoch": 2.8456659619450315, + "grad_norm": 1.8602925539016724, + "learning_rate": 2.9054974655543335e-06, + "loss": 0.66148442029953, + "step": 1346 + }, + { + "epoch": 2.8498942917547567, + "grad_norm": 0.8604276180267334, + "learning_rate": 2.89265761493816e-06, + "loss": 0.7914894819259644, + "step": 1348 + }, + { + "epoch": 2.854122621564482, + "grad_norm": 8.636595726013184, + "learning_rate": 2.8798496401301644e-06, + "loss": 0.5552013516426086, + "step": 1350 + }, + { + "epoch": 2.858350951374207, + "grad_norm": 1.1168638467788696, + "learning_rate": 2.8670736977129843e-06, + "loss": 0.7921260595321655, + "step": 1352 + }, + { + "epoch": 2.8625792811839323, + "grad_norm": 0.7934768795967102, + "learning_rate": 2.854329943877647e-06, + "loss": 0.8490140438079834, + "step": 1354 + }, + { + "epoch": 2.8668076109936576, + "grad_norm": 2.7815663814544678, + "learning_rate": 2.841618534421664e-06, + "loss": 0.32287347316741943, + "step": 1356 + }, + { + "epoch": 2.8710359408033828, + "grad_norm": 0.9619285464286804, + "learning_rate": 2.8289396247471227e-06, + "loss": 1.0223133563995361, + "step": 1358 + }, + { + "epoch": 2.875264270613108, + "grad_norm": 1.4004619121551514, + "learning_rate": 2.816293369858788e-06, + "loss": 0.6217314004898071, + "step": 1360 + }, + { + "epoch": 2.879492600422833, + "grad_norm": 1.1949740648269653, + "learning_rate": 2.803679924362206e-06, + "loss": 0.5934377908706665, + "step": 1362 + }, + { + "epoch": 2.883720930232558, + "grad_norm": 1.1073276996612549, + "learning_rate": 2.7910994424618165e-06, + "loss": 1.0066415071487427, + "step": 1364 + }, + { + "epoch": 2.887949260042283, + "grad_norm": 1.755104899406433, + "learning_rate": 2.778552077959064e-06, + "loss": 0.7661958336830139, + "step": 1366 + }, + { + "epoch": 2.8921775898520083, + "grad_norm": 0.7925071716308594, + "learning_rate": 2.766037984250521e-06, + "loss": 0.8940379619598389, + "step": 1368 + }, + { + "epoch": 2.8964059196617336, + "grad_norm": 0.8228067755699158, + "learning_rate": 2.753557314326011e-06, + "loss": 0.8905459642410278, + "step": 1370 + }, + { + "epoch": 2.9006342494714588, + "grad_norm": 3.0453977584838867, + "learning_rate": 2.7411102207667373e-06, + "loss": 0.922730565071106, + "step": 1372 + }, + { + "epoch": 2.904862579281184, + "grad_norm": 0.6524626612663269, + "learning_rate": 2.728696855743418e-06, + "loss": 0.7393578886985779, + "step": 1374 + }, + { + "epoch": 2.909090909090909, + "grad_norm": 0.7990800738334656, + "learning_rate": 2.7163173710144323e-06, + "loss": 0.9326273798942566, + "step": 1376 + }, + { + "epoch": 2.9133192389006344, + "grad_norm": 2.748785972595215, + "learning_rate": 2.703971917923951e-06, + "loss": 0.638850748538971, + "step": 1378 + }, + { + "epoch": 2.9175475687103596, + "grad_norm": 1.0631226301193237, + "learning_rate": 2.6916606474000965e-06, + "loss": 0.6000032424926758, + "step": 1380 + }, + { + "epoch": 2.9217758985200843, + "grad_norm": 1.7124921083450317, + "learning_rate": 2.679383709953103e-06, + "loss": 1.021500587463379, + "step": 1382 + }, + { + "epoch": 2.92600422832981, + "grad_norm": 4.830338954925537, + "learning_rate": 2.6671412556734578e-06, + "loss": 0.7385914921760559, + "step": 1384 + }, + { + "epoch": 2.9302325581395348, + "grad_norm": 1.3860539197921753, + "learning_rate": 2.6549334342300844e-06, + "loss": 0.6159816980361938, + "step": 1386 + }, + { + "epoch": 2.93446088794926, + "grad_norm": 1.2965255975723267, + "learning_rate": 2.642760394868503e-06, + "loss": 0.8611303567886353, + "step": 1388 + }, + { + "epoch": 2.938689217758985, + "grad_norm": 0.3740333914756775, + "learning_rate": 2.6306222864090157e-06, + "loss": 0.6633927822113037, + "step": 1390 + }, + { + "epoch": 2.9429175475687104, + "grad_norm": 1.4134292602539062, + "learning_rate": 2.61851925724487e-06, + "loss": 0.8371663093566895, + "step": 1392 + }, + { + "epoch": 2.9471458773784356, + "grad_norm": 0.6531880497932434, + "learning_rate": 2.606451455340459e-06, + "loss": 0.2249889224767685, + "step": 1394 + }, + { + "epoch": 2.951374207188161, + "grad_norm": 1.194057583808899, + "learning_rate": 2.5944190282295155e-06, + "loss": 0.46292445063591003, + "step": 1396 + }, + { + "epoch": 2.955602536997886, + "grad_norm": 2.151322603225708, + "learning_rate": 2.5824221230132895e-06, + "loss": 0.8421627283096313, + "step": 1398 + }, + { + "epoch": 2.9598308668076108, + "grad_norm": 1.3173991441726685, + "learning_rate": 2.5704608863587676e-06, + "loss": 0.9961873292922974, + "step": 1400 + }, + { + "epoch": 2.9640591966173364, + "grad_norm": 3.125861644744873, + "learning_rate": 2.558535464496877e-06, + "loss": 0.7260453104972839, + "step": 1402 + }, + { + "epoch": 2.968287526427061, + "grad_norm": 2.708327054977417, + "learning_rate": 2.5466460032206872e-06, + "loss": 0.49916917085647583, + "step": 1404 + }, + { + "epoch": 2.9725158562367864, + "grad_norm": 2.326904773712158, + "learning_rate": 2.53479264788364e-06, + "loss": 0.7809391617774963, + "step": 1406 + }, + { + "epoch": 2.9767441860465116, + "grad_norm": 1.1031649112701416, + "learning_rate": 2.522975543397769e-06, + "loss": 0.6572524905204773, + "step": 1408 + }, + { + "epoch": 2.980972515856237, + "grad_norm": 4.386587142944336, + "learning_rate": 2.511194834231922e-06, + "loss": 0.6572096347808838, + "step": 1410 + }, + { + "epoch": 2.985200845665962, + "grad_norm": 0.8324053883552551, + "learning_rate": 2.49945066441e-06, + "loss": 0.8871416449546814, + "step": 1412 + }, + { + "epoch": 2.989429175475687, + "grad_norm": 0.6973029971122742, + "learning_rate": 2.4877431775092004e-06, + "loss": 0.6600456833839417, + "step": 1414 + }, + { + "epoch": 2.9936575052854124, + "grad_norm": 8.532988548278809, + "learning_rate": 2.4760725166582544e-06, + "loss": 0.734959065914154, + "step": 1416 + }, + { + "epoch": 2.997885835095137, + "grad_norm": 0.5599045157432556, + "learning_rate": 2.464438824535675e-06, + "loss": 0.4596215486526489, + "step": 1418 + }, + { + "epoch": 3.0021141649048624, + "grad_norm": 0.7426612377166748, + "learning_rate": 2.452842243368028e-06, + "loss": 0.822529673576355, + "step": 1420 + }, + { + "epoch": 3.0063424947145876, + "grad_norm": 1.1374843120574951, + "learning_rate": 2.4412829149281735e-06, + "loss": 0.47854307293891907, + "step": 1422 + }, + { + "epoch": 3.010570824524313, + "grad_norm": 1.9394512176513672, + "learning_rate": 2.4297609805335472e-06, + "loss": 0.8591348528862, + "step": 1424 + }, + { + "epoch": 3.014799154334038, + "grad_norm": 6.236659526824951, + "learning_rate": 2.4182765810444263e-06, + "loss": 0.32937896251678467, + "step": 1426 + }, + { + "epoch": 3.019027484143763, + "grad_norm": 0.9608882665634155, + "learning_rate": 2.4068298568622083e-06, + "loss": 0.480927973985672, + "step": 1428 + }, + { + "epoch": 3.0232558139534884, + "grad_norm": 0.861669659614563, + "learning_rate": 2.395420947927694e-06, + "loss": 0.6863812804222107, + "step": 1430 + }, + { + "epoch": 3.0274841437632136, + "grad_norm": 0.7586754560470581, + "learning_rate": 2.3840499937193804e-06, + "loss": 0.7264788746833801, + "step": 1432 + }, + { + "epoch": 3.031712473572939, + "grad_norm": 0.7941664457321167, + "learning_rate": 2.3727171332517496e-06, + "loss": 0.6077902913093567, + "step": 1434 + }, + { + "epoch": 3.035940803382664, + "grad_norm": 2.874917507171631, + "learning_rate": 2.361422505073575e-06, + "loss": 0.45975029468536377, + "step": 1436 + }, + { + "epoch": 3.040169133192389, + "grad_norm": 0.7910526394844055, + "learning_rate": 2.3501662472662233e-06, + "loss": 0.1154491975903511, + "step": 1438 + }, + { + "epoch": 3.044397463002114, + "grad_norm": 1.0831025838851929, + "learning_rate": 2.3389484974419685e-06, + "loss": 0.6795452833175659, + "step": 1440 + }, + { + "epoch": 3.048625792811839, + "grad_norm": 1.4130436182022095, + "learning_rate": 2.327769392742309e-06, + "loss": 0.48243653774261475, + "step": 1442 + }, + { + "epoch": 3.0528541226215644, + "grad_norm": 20.217180252075195, + "learning_rate": 2.316629069836291e-06, + "loss": 0.13504913449287415, + "step": 1444 + }, + { + "epoch": 3.0570824524312896, + "grad_norm": 1.7556240558624268, + "learning_rate": 2.305527664918838e-06, + "loss": 0.5326657891273499, + "step": 1446 + }, + { + "epoch": 3.061310782241015, + "grad_norm": 3.73991060256958, + "learning_rate": 2.294465313709085e-06, + "loss": 0.596695601940155, + "step": 1448 + }, + { + "epoch": 3.06553911205074, + "grad_norm": 1.3301008939743042, + "learning_rate": 2.2834421514487205e-06, + "loss": 0.5062366127967834, + "step": 1450 + }, + { + "epoch": 3.0697674418604652, + "grad_norm": 5.256744384765625, + "learning_rate": 2.27245831290033e-06, + "loss": 0.586754322052002, + "step": 1452 + }, + { + "epoch": 3.0739957716701904, + "grad_norm": 1.3477658033370972, + "learning_rate": 2.261513932345755e-06, + "loss": 0.6754856705665588, + "step": 1454 + }, + { + "epoch": 3.0782241014799157, + "grad_norm": 0.2527630925178528, + "learning_rate": 2.250609143584443e-06, + "loss": 0.2013336420059204, + "step": 1456 + }, + { + "epoch": 3.0824524312896404, + "grad_norm": 1.301119327545166, + "learning_rate": 2.239744079931818e-06, + "loss": 0.5202870965003967, + "step": 1458 + }, + { + "epoch": 3.0866807610993656, + "grad_norm": 1.696247935295105, + "learning_rate": 2.228918874217649e-06, + "loss": 0.6424930095672607, + "step": 1460 + }, + { + "epoch": 3.090909090909091, + "grad_norm": 1.6689664125442505, + "learning_rate": 2.2181336587844244e-06, + "loss": 0.6810600757598877, + "step": 1462 + }, + { + "epoch": 3.095137420718816, + "grad_norm": 1.8422038555145264, + "learning_rate": 2.207388565485736e-06, + "loss": 0.48403334617614746, + "step": 1464 + }, + { + "epoch": 3.0993657505285412, + "grad_norm": 1.6538422107696533, + "learning_rate": 2.196683725684666e-06, + "loss": 0.5302621126174927, + "step": 1466 + }, + { + "epoch": 3.1035940803382664, + "grad_norm": 1.053499460220337, + "learning_rate": 2.1860192702521826e-06, + "loss": 0.70964115858078, + "step": 1468 + }, + { + "epoch": 3.1078224101479917, + "grad_norm": 1.0107591152191162, + "learning_rate": 2.1753953295655354e-06, + "loss": 0.7231332063674927, + "step": 1470 + }, + { + "epoch": 3.112050739957717, + "grad_norm": 4.499339580535889, + "learning_rate": 2.164812033506673e-06, + "loss": 0.45468053221702576, + "step": 1472 + }, + { + "epoch": 3.116279069767442, + "grad_norm": 1.8995431661605835, + "learning_rate": 2.1542695114606367e-06, + "loss": 0.8919910192489624, + "step": 1474 + }, + { + "epoch": 3.120507399577167, + "grad_norm": 0.9064427614212036, + "learning_rate": 2.1437678923139944e-06, + "loss": 0.7613162398338318, + "step": 1476 + }, + { + "epoch": 3.124735729386892, + "grad_norm": 4.857295989990234, + "learning_rate": 2.1333073044532628e-06, + "loss": 0.24337275326251984, + "step": 1478 + }, + { + "epoch": 3.1289640591966172, + "grad_norm": 1.37673020362854, + "learning_rate": 2.122887875763327e-06, + "loss": 0.7974050641059875, + "step": 1480 + }, + { + "epoch": 3.1331923890063424, + "grad_norm": 1.0172052383422852, + "learning_rate": 2.1125097336258897e-06, + "loss": 0.792412519454956, + "step": 1482 + }, + { + "epoch": 3.1374207188160677, + "grad_norm": 0.7591030597686768, + "learning_rate": 2.1021730049179117e-06, + "loss": 0.20275385677814484, + "step": 1484 + }, + { + "epoch": 3.141649048625793, + "grad_norm": 1.3773444890975952, + "learning_rate": 2.09187781601005e-06, + "loss": 0.6799913644790649, + "step": 1486 + }, + { + "epoch": 3.145877378435518, + "grad_norm": 1.1301944255828857, + "learning_rate": 2.0816242927651254e-06, + "loss": 0.8414097428321838, + "step": 1488 + }, + { + "epoch": 3.1501057082452433, + "grad_norm": 20.24962615966797, + "learning_rate": 2.071412560536581e-06, + "loss": 0.622582733631134, + "step": 1490 + }, + { + "epoch": 3.1543340380549685, + "grad_norm": 12.852417945861816, + "learning_rate": 2.0612427441669475e-06, + "loss": 0.5549498200416565, + "step": 1492 + }, + { + "epoch": 3.1585623678646932, + "grad_norm": 0.9046342372894287, + "learning_rate": 2.051114967986311e-06, + "loss": 0.3630253076553345, + "step": 1494 + }, + { + "epoch": 3.1627906976744184, + "grad_norm": 0.17015591263771057, + "learning_rate": 2.0410293558108074e-06, + "loss": 0.48407110571861267, + "step": 1496 + }, + { + "epoch": 3.1670190274841437, + "grad_norm": 0.6826212406158447, + "learning_rate": 2.0309860309410986e-06, + "loss": 0.7625746726989746, + "step": 1498 + }, + { + "epoch": 3.171247357293869, + "grad_norm": 1.1693699359893799, + "learning_rate": 2.0209851161608613e-06, + "loss": 0.8112456798553467, + "step": 1500 + }, + { + "epoch": 3.175475687103594, + "grad_norm": 1.1140645742416382, + "learning_rate": 2.011026733735301e-06, + "loss": 0.2727448344230652, + "step": 1502 + }, + { + "epoch": 3.1797040169133193, + "grad_norm": 1.5750521421432495, + "learning_rate": 2.001111005409643e-06, + "loss": 0.4191259443759918, + "step": 1504 + }, + { + "epoch": 3.1839323467230445, + "grad_norm": 1.5685886144638062, + "learning_rate": 1.9912380524076503e-06, + "loss": 0.59795081615448, + "step": 1506 + }, + { + "epoch": 3.1881606765327697, + "grad_norm": 3.5002119541168213, + "learning_rate": 1.981407995430136e-06, + "loss": 0.31429192423820496, + "step": 1508 + }, + { + "epoch": 3.192389006342495, + "grad_norm": 8.43346881866455, + "learning_rate": 1.9716209546535e-06, + "loss": 0.24339474737644196, + "step": 1510 + }, + { + "epoch": 3.1966173361522197, + "grad_norm": 0.41249513626098633, + "learning_rate": 1.9618770497282487e-06, + "loss": 0.04826962575316429, + "step": 1512 + }, + { + "epoch": 3.200845665961945, + "grad_norm": 1.9549050331115723, + "learning_rate": 1.9521763997775294e-06, + "loss": 0.5833825469017029, + "step": 1514 + }, + { + "epoch": 3.20507399577167, + "grad_norm": 1.0116013288497925, + "learning_rate": 1.942519123395689e-06, + "loss": 0.7083150148391724, + "step": 1516 + }, + { + "epoch": 3.2093023255813953, + "grad_norm": 1.2317473888397217, + "learning_rate": 1.932905338646812e-06, + "loss": 0.533999502658844, + "step": 1518 + }, + { + "epoch": 3.2135306553911205, + "grad_norm": 0.7542660236358643, + "learning_rate": 1.9233351630632747e-06, + "loss": 0.7182834148406982, + "step": 1520 + }, + { + "epoch": 3.2177589852008457, + "grad_norm": 0.8418310880661011, + "learning_rate": 1.9138087136443227e-06, + "loss": 0.7226573824882507, + "step": 1522 + }, + { + "epoch": 3.221987315010571, + "grad_norm": 3.589010715484619, + "learning_rate": 1.904326106854626e-06, + "loss": 0.7369678020477295, + "step": 1524 + }, + { + "epoch": 3.226215644820296, + "grad_norm": 1.0073081254959106, + "learning_rate": 1.8948874586228622e-06, + "loss": 0.6824883818626404, + "step": 1526 + }, + { + "epoch": 3.2304439746300213, + "grad_norm": 1.857325792312622, + "learning_rate": 1.8854928843402987e-06, + "loss": 0.07193010300397873, + "step": 1528 + }, + { + "epoch": 3.234672304439746, + "grad_norm": 2.487086057662964, + "learning_rate": 1.8761424988593796e-06, + "loss": 0.6838529109954834, + "step": 1530 + }, + { + "epoch": 3.2389006342494713, + "grad_norm": 1.8821792602539062, + "learning_rate": 1.8668364164923248e-06, + "loss": 0.6269945502281189, + "step": 1532 + }, + { + "epoch": 3.2431289640591965, + "grad_norm": 2.7210841178894043, + "learning_rate": 1.8575747510097294e-06, + "loss": 0.7840203046798706, + "step": 1534 + }, + { + "epoch": 3.2473572938689217, + "grad_norm": 1.0130170583724976, + "learning_rate": 1.8483576156391747e-06, + "loss": 0.5027771592140198, + "step": 1536 + }, + { + "epoch": 3.251585623678647, + "grad_norm": 1.0865229368209839, + "learning_rate": 1.8391851230638443e-06, + "loss": 0.5004664659500122, + "step": 1538 + }, + { + "epoch": 3.255813953488372, + "grad_norm": 0.9608802795410156, + "learning_rate": 1.8300573854211452e-06, + "loss": 0.40708398818969727, + "step": 1540 + }, + { + "epoch": 3.2600422832980973, + "grad_norm": 2.830068826675415, + "learning_rate": 1.8209745143013363e-06, + "loss": 0.8195509910583496, + "step": 1542 + }, + { + "epoch": 3.2642706131078225, + "grad_norm": 1.6520243883132935, + "learning_rate": 1.811936620746167e-06, + "loss": 0.8325942158699036, + "step": 1544 + }, + { + "epoch": 3.2684989429175477, + "grad_norm": 1.8620424270629883, + "learning_rate": 1.8029438152475159e-06, + "loss": 0.6713284850120544, + "step": 1546 + }, + { + "epoch": 3.2727272727272725, + "grad_norm": 0.7371749877929688, + "learning_rate": 1.7939962077460448e-06, + "loss": 0.1721542775630951, + "step": 1548 + }, + { + "epoch": 3.276955602536998, + "grad_norm": 2.975339412689209, + "learning_rate": 1.785093907629849e-06, + "loss": 0.5900051593780518, + "step": 1550 + }, + { + "epoch": 3.281183932346723, + "grad_norm": 0.7683355808258057, + "learning_rate": 1.7762370237331233e-06, + "loss": 0.696302056312561, + "step": 1552 + }, + { + "epoch": 3.285412262156448, + "grad_norm": 0.09710486233234406, + "learning_rate": 1.7674256643348348e-06, + "loss": 0.07380301505327225, + "step": 1554 + }, + { + "epoch": 3.2896405919661733, + "grad_norm": 0.8191556334495544, + "learning_rate": 1.7586599371573887e-06, + "loss": 0.6632875800132751, + "step": 1556 + }, + { + "epoch": 3.2938689217758985, + "grad_norm": 1.1252812147140503, + "learning_rate": 1.7499399493653225e-06, + "loss": 0.3820130527019501, + "step": 1558 + }, + { + "epoch": 3.2980972515856237, + "grad_norm": 2.1938154697418213, + "learning_rate": 1.7412658075639921e-06, + "loss": 0.9419984817504883, + "step": 1560 + }, + { + "epoch": 3.302325581395349, + "grad_norm": 3.291370391845703, + "learning_rate": 1.7326376177982635e-06, + "loss": 0.5527592897415161, + "step": 1562 + }, + { + "epoch": 3.306553911205074, + "grad_norm": 0.8782255053520203, + "learning_rate": 1.7240554855512225e-06, + "loss": 0.39016205072402954, + "step": 1564 + }, + { + "epoch": 3.3107822410147993, + "grad_norm": 4.873465061187744, + "learning_rate": 1.7155195157428878e-06, + "loss": 0.5568993091583252, + "step": 1566 + }, + { + "epoch": 3.3150105708245245, + "grad_norm": 0.27702853083610535, + "learning_rate": 1.707029812728916e-06, + "loss": 0.10718473792076111, + "step": 1568 + }, + { + "epoch": 3.3192389006342493, + "grad_norm": 1.6705442667007446, + "learning_rate": 1.6985864802993366e-06, + "loss": 0.5119863152503967, + "step": 1570 + }, + { + "epoch": 3.3234672304439745, + "grad_norm": 3.1223764419555664, + "learning_rate": 1.690189621677284e-06, + "loss": 0.5850059986114502, + "step": 1572 + }, + { + "epoch": 3.3276955602536997, + "grad_norm": 5.79948616027832, + "learning_rate": 1.681839339517728e-06, + "loss": 0.16349083185195923, + "step": 1574 + }, + { + "epoch": 3.331923890063425, + "grad_norm": 2.860550880432129, + "learning_rate": 1.67353573590622e-06, + "loss": 0.22142480313777924, + "step": 1576 + }, + { + "epoch": 3.33615221987315, + "grad_norm": 1.07570481300354, + "learning_rate": 1.6652789123576513e-06, + "loss": 0.4672536253929138, + "step": 1578 + }, + { + "epoch": 3.3403805496828753, + "grad_norm": 1.3741143941879272, + "learning_rate": 1.6570689698150094e-06, + "loss": 0.47741004824638367, + "step": 1580 + }, + { + "epoch": 3.3446088794926006, + "grad_norm": 1.0435084104537964, + "learning_rate": 1.6489060086481373e-06, + "loss": 0.750064492225647, + "step": 1582 + }, + { + "epoch": 3.3488372093023258, + "grad_norm": 2.3415040969848633, + "learning_rate": 1.6407901286525152e-06, + "loss": 0.5267655253410339, + "step": 1584 + }, + { + "epoch": 3.353065539112051, + "grad_norm": 2.695880889892578, + "learning_rate": 1.6327214290480382e-06, + "loss": 0.5559815764427185, + "step": 1586 + }, + { + "epoch": 3.3572938689217757, + "grad_norm": 2.3424551486968994, + "learning_rate": 1.6247000084778006e-06, + "loss": 0.6173189282417297, + "step": 1588 + }, + { + "epoch": 3.361522198731501, + "grad_norm": 1.036224126815796, + "learning_rate": 1.6167259650068887e-06, + "loss": 0.23196779191493988, + "step": 1590 + }, + { + "epoch": 3.365750528541226, + "grad_norm": 0.900652289390564, + "learning_rate": 1.6087993961211904e-06, + "loss": 0.5860919952392578, + "step": 1592 + }, + { + "epoch": 3.3699788583509513, + "grad_norm": 1.160366415977478, + "learning_rate": 1.6009203987261957e-06, + "loss": 0.22235609591007233, + "step": 1594 + }, + { + "epoch": 3.3742071881606766, + "grad_norm": 1.2114521265029907, + "learning_rate": 1.593089069145809e-06, + "loss": 0.7520530223846436, + "step": 1596 + }, + { + "epoch": 3.3784355179704018, + "grad_norm": 4.511026382446289, + "learning_rate": 1.5853055031211844e-06, + "loss": 0.5549476742744446, + "step": 1598 + }, + { + "epoch": 3.382663847780127, + "grad_norm": 1.0976760387420654, + "learning_rate": 1.577569795809544e-06, + "loss": 0.19076868891716003, + "step": 1600 + }, + { + "epoch": 3.386892177589852, + "grad_norm": 3.270397186279297, + "learning_rate": 1.5698820417830137e-06, + "loss": 0.4793318808078766, + "step": 1602 + }, + { + "epoch": 3.3911205073995774, + "grad_norm": 1.316583514213562, + "learning_rate": 1.5622423350274784e-06, + "loss": 0.8752365708351135, + "step": 1604 + }, + { + "epoch": 3.395348837209302, + "grad_norm": 1.2119370698928833, + "learning_rate": 1.5546507689414217e-06, + "loss": 0.8328059911727905, + "step": 1606 + }, + { + "epoch": 3.3995771670190273, + "grad_norm": 1.942264199256897, + "learning_rate": 1.5471074363347894e-06, + "loss": 0.4663934111595154, + "step": 1608 + }, + { + "epoch": 3.4038054968287526, + "grad_norm": 1.1383877992630005, + "learning_rate": 1.539612429427852e-06, + "loss": 0.7278372645378113, + "step": 1610 + }, + { + "epoch": 3.4080338266384778, + "grad_norm": 1.6411103010177612, + "learning_rate": 1.5321658398500808e-06, + "loss": 0.7146084904670715, + "step": 1612 + }, + { + "epoch": 3.412262156448203, + "grad_norm": 1.0693954229354858, + "learning_rate": 1.5247677586390255e-06, + "loss": 0.5955060720443726, + "step": 1614 + }, + { + "epoch": 3.416490486257928, + "grad_norm": 1.0056612491607666, + "learning_rate": 1.5174182762392007e-06, + "loss": 0.7104098796844482, + "step": 1616 + }, + { + "epoch": 3.4207188160676534, + "grad_norm": 0.8999010920524597, + "learning_rate": 1.5101174825009815e-06, + "loss": 0.814241886138916, + "step": 1618 + }, + { + "epoch": 3.4249471458773786, + "grad_norm": 1.3140491247177124, + "learning_rate": 1.5028654666795042e-06, + "loss": 0.7213199138641357, + "step": 1620 + }, + { + "epoch": 3.429175475687104, + "grad_norm": 7.2188897132873535, + "learning_rate": 1.495662317433576e-06, + "loss": 0.47718608379364014, + "step": 1622 + }, + { + "epoch": 3.4334038054968286, + "grad_norm": 2.354926347732544, + "learning_rate": 1.4885081228245909e-06, + "loss": 0.7822759747505188, + "step": 1624 + }, + { + "epoch": 3.4376321353065538, + "grad_norm": 1.0315642356872559, + "learning_rate": 1.4814029703154515e-06, + "loss": 0.6101397275924683, + "step": 1626 + }, + { + "epoch": 3.441860465116279, + "grad_norm": 0.9578467607498169, + "learning_rate": 1.4743469467695017e-06, + "loss": 0.8166185617446899, + "step": 1628 + }, + { + "epoch": 3.446088794926004, + "grad_norm": 1.6009467840194702, + "learning_rate": 1.4673401384494651e-06, + "loss": 0.5006849765777588, + "step": 1630 + }, + { + "epoch": 3.4503171247357294, + "grad_norm": 1.023461937904358, + "learning_rate": 1.4603826310163874e-06, + "loss": 0.6561769247055054, + "step": 1632 + }, + { + "epoch": 3.4545454545454546, + "grad_norm": 0.9615203738212585, + "learning_rate": 1.453474509528593e-06, + "loss": 0.46822845935821533, + "step": 1634 + }, + { + "epoch": 3.45877378435518, + "grad_norm": 1.0977420806884766, + "learning_rate": 1.446615858440642e-06, + "loss": 0.7399229407310486, + "step": 1636 + }, + { + "epoch": 3.463002114164905, + "grad_norm": 0.2824209928512573, + "learning_rate": 1.4398067616022993e-06, + "loss": 0.2356806993484497, + "step": 1638 + }, + { + "epoch": 3.46723044397463, + "grad_norm": 0.8522822856903076, + "learning_rate": 1.43304730225751e-06, + "loss": 0.42114752531051636, + "step": 1640 + }, + { + "epoch": 3.471458773784355, + "grad_norm": 0.5968304872512817, + "learning_rate": 1.4263375630433802e-06, + "loss": 0.131179079413414, + "step": 1642 + }, + { + "epoch": 3.47568710359408, + "grad_norm": 1.0254826545715332, + "learning_rate": 1.4196776259891665e-06, + "loss": 0.7695028781890869, + "step": 1644 + }, + { + "epoch": 3.4799154334038054, + "grad_norm": 1.133607029914856, + "learning_rate": 1.4130675725152762e-06, + "loss": 0.6593310236930847, + "step": 1646 + }, + { + "epoch": 3.4841437632135306, + "grad_norm": 0.1946207731962204, + "learning_rate": 1.4065074834322677e-06, + "loss": 0.4490881860256195, + "step": 1648 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 2.0469253063201904, + "learning_rate": 1.3999974389398669e-06, + "loss": 0.40921321511268616, + "step": 1650 + }, + { + "epoch": 3.492600422832981, + "grad_norm": 12.335336685180664, + "learning_rate": 1.3935375186259817e-06, + "loss": 0.4644951820373535, + "step": 1652 + }, + { + "epoch": 3.496828752642706, + "grad_norm": 1.9725359678268433, + "learning_rate": 1.3871278014657332e-06, + "loss": 0.64243483543396, + "step": 1654 + }, + { + "epoch": 3.5010570824524314, + "grad_norm": 2.0758938789367676, + "learning_rate": 1.3807683658204915e-06, + "loss": 0.7746992707252502, + "step": 1656 + }, + { + "epoch": 3.5052854122621566, + "grad_norm": 0.24769353866577148, + "learning_rate": 1.3744592894369093e-06, + "loss": 0.5714927315711975, + "step": 1658 + }, + { + "epoch": 3.5095137420718814, + "grad_norm": 0.7157841920852661, + "learning_rate": 1.3682006494459806e-06, + "loss": 0.5627008676528931, + "step": 1660 + }, + { + "epoch": 3.513742071881607, + "grad_norm": 1.2036532163619995, + "learning_rate": 1.361992522362096e-06, + "loss": 0.45404481887817383, + "step": 1662 + }, + { + "epoch": 3.517970401691332, + "grad_norm": 0.7401357889175415, + "learning_rate": 1.355834984082101e-06, + "loss": 0.6829448938369751, + "step": 1664 + }, + { + "epoch": 3.522198731501057, + "grad_norm": 1.6080371141433716, + "learning_rate": 1.3497281098843746e-06, + "loss": 0.6879425048828125, + "step": 1666 + }, + { + "epoch": 3.526427061310782, + "grad_norm": 0.28612497448921204, + "learning_rate": 1.34367197442791e-06, + "loss": 0.09749031066894531, + "step": 1668 + }, + { + "epoch": 3.5306553911205074, + "grad_norm": 0.3093745708465576, + "learning_rate": 1.3376666517513931e-06, + "loss": 0.6256831288337708, + "step": 1670 + }, + { + "epoch": 3.5348837209302326, + "grad_norm": 1.3815644979476929, + "learning_rate": 1.3317122152723078e-06, + "loss": 0.4126441180706024, + "step": 1672 + }, + { + "epoch": 3.539112050739958, + "grad_norm": 1.8580849170684814, + "learning_rate": 1.3258087377860323e-06, + "loss": 0.6396852135658264, + "step": 1674 + }, + { + "epoch": 3.543340380549683, + "grad_norm": 1.6190978288650513, + "learning_rate": 1.3199562914649519e-06, + "loss": 0.5272851586341858, + "step": 1676 + }, + { + "epoch": 3.547568710359408, + "grad_norm": 1.0420379638671875, + "learning_rate": 1.314154947857571e-06, + "loss": 0.6790235042572021, + "step": 1678 + }, + { + "epoch": 3.5517970401691334, + "grad_norm": 1.7040330171585083, + "learning_rate": 1.3084047778876487e-06, + "loss": 0.3993278443813324, + "step": 1680 + }, + { + "epoch": 3.556025369978858, + "grad_norm": 0.9581378698348999, + "learning_rate": 1.3027058518533228e-06, + "loss": 0.5089682936668396, + "step": 1682 + }, + { + "epoch": 3.5602536997885834, + "grad_norm": 1.3544453382492065, + "learning_rate": 1.2970582394262516e-06, + "loss": 0.3977218270301819, + "step": 1684 + }, + { + "epoch": 3.5644820295983086, + "grad_norm": 4.566845893859863, + "learning_rate": 1.291462009650768e-06, + "loss": 0.14128461480140686, + "step": 1686 + }, + { + "epoch": 3.568710359408034, + "grad_norm": 2.174217700958252, + "learning_rate": 1.2859172309430285e-06, + "loss": 0.6097127795219421, + "step": 1688 + }, + { + "epoch": 3.572938689217759, + "grad_norm": 0.07479295134544373, + "learning_rate": 1.280423971090181e-06, + "loss": 0.1652495265007019, + "step": 1690 + }, + { + "epoch": 3.5771670190274842, + "grad_norm": 1.667214274406433, + "learning_rate": 1.2749822972495333e-06, + "loss": 0.3264021873474121, + "step": 1692 + }, + { + "epoch": 3.5813953488372094, + "grad_norm": 3.953310489654541, + "learning_rate": 1.2695922759477356e-06, + "loss": 0.5752360820770264, + "step": 1694 + }, + { + "epoch": 3.585623678646934, + "grad_norm": 9.876787185668945, + "learning_rate": 1.2642539730799622e-06, + "loss": 0.2854745388031006, + "step": 1696 + }, + { + "epoch": 3.58985200845666, + "grad_norm": 1.3300018310546875, + "learning_rate": 1.2589674539091115e-06, + "loss": 0.7096505165100098, + "step": 1698 + }, + { + "epoch": 3.5940803382663846, + "grad_norm": 1.2227599620819092, + "learning_rate": 1.2537327830650042e-06, + "loss": 0.8019452691078186, + "step": 1700 + }, + { + "epoch": 3.59830866807611, + "grad_norm": 6.207125663757324, + "learning_rate": 1.248550024543594e-06, + "loss": 0.5870368480682373, + "step": 1702 + }, + { + "epoch": 3.602536997885835, + "grad_norm": 10.3837308883667, + "learning_rate": 1.2434192417061862e-06, + "loss": 0.31027811765670776, + "step": 1704 + }, + { + "epoch": 3.6067653276955602, + "grad_norm": 0.9408029913902283, + "learning_rate": 1.2383404972786625e-06, + "loss": 0.3645901083946228, + "step": 1706 + }, + { + "epoch": 3.6109936575052854, + "grad_norm": 2.304697036743164, + "learning_rate": 1.2333138533507139e-06, + "loss": 0.6051492094993591, + "step": 1708 + }, + { + "epoch": 3.6152219873150107, + "grad_norm": 1.9065330028533936, + "learning_rate": 1.2283393713750816e-06, + "loss": 0.6444299221038818, + "step": 1710 + }, + { + "epoch": 3.619450317124736, + "grad_norm": 1.3381016254425049, + "learning_rate": 1.2234171121668069e-06, + "loss": 0.33283212780952454, + "step": 1712 + }, + { + "epoch": 3.6236786469344606, + "grad_norm": 0.5783998370170593, + "learning_rate": 1.2185471359024855e-06, + "loss": 0.24882088601589203, + "step": 1714 + }, + { + "epoch": 3.6279069767441863, + "grad_norm": 2.346623659133911, + "learning_rate": 1.2137295021195341e-06, + "loss": 0.5580241084098816, + "step": 1716 + }, + { + "epoch": 3.632135306553911, + "grad_norm": 0.4007739722728729, + "learning_rate": 1.2089642697154613e-06, + "loss": 0.3992137908935547, + "step": 1718 + }, + { + "epoch": 3.6363636363636362, + "grad_norm": 3.3852832317352295, + "learning_rate": 1.2042514969471475e-06, + "loss": 0.3962453603744507, + "step": 1720 + }, + { + "epoch": 3.6405919661733614, + "grad_norm": 2.093113422393799, + "learning_rate": 1.1995912414301329e-06, + "loss": 0.6258540153503418, + "step": 1722 + }, + { + "epoch": 3.6448202959830867, + "grad_norm": 1.9229166507720947, + "learning_rate": 1.1949835601379144e-06, + "loss": 0.5780177712440491, + "step": 1724 + }, + { + "epoch": 3.649048625792812, + "grad_norm": 1.8796145915985107, + "learning_rate": 1.1904285094012463e-06, + "loss": 0.841484010219574, + "step": 1726 + }, + { + "epoch": 3.653276955602537, + "grad_norm": 1.08918035030365, + "learning_rate": 1.185926144907454e-06, + "loss": 0.9029679298400879, + "step": 1728 + }, + { + "epoch": 3.6575052854122623, + "grad_norm": 8.495342254638672, + "learning_rate": 1.1814765216997522e-06, + "loss": 0.7344114184379578, + "step": 1730 + }, + { + "epoch": 3.6617336152219875, + "grad_norm": 1.06622314453125, + "learning_rate": 1.1770796941765727e-06, + "loss": 0.663408100605011, + "step": 1732 + }, + { + "epoch": 3.6659619450317127, + "grad_norm": 1.00587797164917, + "learning_rate": 1.1727357160908978e-06, + "loss": 0.6155779361724854, + "step": 1734 + }, + { + "epoch": 3.6701902748414374, + "grad_norm": 1.2634364366531372, + "learning_rate": 1.1684446405496052e-06, + "loss": 0.6922056674957275, + "step": 1736 + }, + { + "epoch": 3.6744186046511627, + "grad_norm": 4.482985019683838, + "learning_rate": 1.1642065200128179e-06, + "loss": 0.13790041208267212, + "step": 1738 + }, + { + "epoch": 3.678646934460888, + "grad_norm": 2.3408749103546143, + "learning_rate": 1.1600214062932613e-06, + "loss": 0.4103460907936096, + "step": 1740 + }, + { + "epoch": 3.682875264270613, + "grad_norm": 2.902986764907837, + "learning_rate": 1.1558893505556315e-06, + "loss": 0.2777998447418213, + "step": 1742 + }, + { + "epoch": 3.6871035940803383, + "grad_norm": 1.138956904411316, + "learning_rate": 1.1518104033159715e-06, + "loss": 0.8116011619567871, + "step": 1744 + }, + { + "epoch": 3.6913319238900635, + "grad_norm": 0.24477116763591766, + "learning_rate": 1.1477846144410493e-06, + "loss": 0.3933503329753876, + "step": 1746 + }, + { + "epoch": 3.6955602536997887, + "grad_norm": 2.183744430541992, + "learning_rate": 1.1438120331477506e-06, + "loss": 0.33288443088531494, + "step": 1748 + }, + { + "epoch": 3.699788583509514, + "grad_norm": 0.967767596244812, + "learning_rate": 1.1398927080024796e-06, + "loss": 0.8717489838600159, + "step": 1750 + }, + { + "epoch": 3.704016913319239, + "grad_norm": 4.215768337249756, + "learning_rate": 1.1360266869205593e-06, + "loss": 0.259102463722229, + "step": 1752 + }, + { + "epoch": 3.708245243128964, + "grad_norm": 12.4045991897583, + "learning_rate": 1.1322140171656507e-06, + "loss": 0.45215776562690735, + "step": 1754 + }, + { + "epoch": 3.712473572938689, + "grad_norm": 0.9202834963798523, + "learning_rate": 1.1284547453491734e-06, + "loss": 0.6496840119361877, + "step": 1756 + }, + { + "epoch": 3.7167019027484143, + "grad_norm": 3.1042778491973877, + "learning_rate": 1.1247489174297372e-06, + "loss": 0.3921501338481903, + "step": 1758 + }, + { + "epoch": 3.7209302325581395, + "grad_norm": 0.2115270495414734, + "learning_rate": 1.1210965787125743e-06, + "loss": 0.3873896896839142, + "step": 1760 + }, + { + "epoch": 3.7251585623678647, + "grad_norm": 1.2216523885726929, + "learning_rate": 1.1174977738489955e-06, + "loss": 0.7566300630569458, + "step": 1762 + }, + { + "epoch": 3.72938689217759, + "grad_norm": 0.7948517799377441, + "learning_rate": 1.113952546835835e-06, + "loss": 0.4671517014503479, + "step": 1764 + }, + { + "epoch": 3.733615221987315, + "grad_norm": 1.520883560180664, + "learning_rate": 1.1104609410149174e-06, + "loss": 0.6633574962615967, + "step": 1766 + }, + { + "epoch": 3.7378435517970403, + "grad_norm": 1.2283849716186523, + "learning_rate": 1.1070229990725262e-06, + "loss": 0.6171925663948059, + "step": 1768 + }, + { + "epoch": 3.7420718816067655, + "grad_norm": 0.25562340021133423, + "learning_rate": 1.1036387630388841e-06, + "loss": 0.13521787524223328, + "step": 1770 + }, + { + "epoch": 3.7463002114164903, + "grad_norm": 3.072812557220459, + "learning_rate": 1.100308274287636e-06, + "loss": 0.38412439823150635, + "step": 1772 + }, + { + "epoch": 3.7505285412262155, + "grad_norm": 2.7034623622894287, + "learning_rate": 1.0970315735353434e-06, + "loss": 0.8012639880180359, + "step": 1774 + }, + { + "epoch": 3.7547568710359407, + "grad_norm": 1.0362725257873535, + "learning_rate": 1.0938087008409909e-06, + "loss": 0.7081127166748047, + "step": 1776 + }, + { + "epoch": 3.758985200845666, + "grad_norm": 0.7398330569267273, + "learning_rate": 1.0906396956054919e-06, + "loss": 0.7581815123558044, + "step": 1778 + }, + { + "epoch": 3.763213530655391, + "grad_norm": 1.3581154346466064, + "learning_rate": 1.0875245965712077e-06, + "loss": 0.8964499831199646, + "step": 1780 + }, + { + "epoch": 3.7674418604651163, + "grad_norm": 1.4774433374404907, + "learning_rate": 1.0844634418214763e-06, + "loss": 0.6864523887634277, + "step": 1782 + }, + { + "epoch": 3.7716701902748415, + "grad_norm": 1.9803892374038696, + "learning_rate": 1.0814562687801448e-06, + "loss": 0.11675049364566803, + "step": 1784 + }, + { + "epoch": 3.7758985200845667, + "grad_norm": 1.9036139249801636, + "learning_rate": 1.0785031142111103e-06, + "loss": 0.46811643242836, + "step": 1786 + }, + { + "epoch": 3.780126849894292, + "grad_norm": 2.764491081237793, + "learning_rate": 1.075604014217875e-06, + "loss": 0.4404146671295166, + "step": 1788 + }, + { + "epoch": 3.7843551797040167, + "grad_norm": 2.860501527786255, + "learning_rate": 1.0727590042431009e-06, + "loss": 0.09559009224176407, + "step": 1790 + }, + { + "epoch": 3.7885835095137423, + "grad_norm": 0.3354916572570801, + "learning_rate": 1.0699681190681777e-06, + "loss": 0.5750871300697327, + "step": 1792 + }, + { + "epoch": 3.792811839323467, + "grad_norm": 1.4355374574661255, + "learning_rate": 1.067231392812798e-06, + "loss": 0.7551760077476501, + "step": 1794 + }, + { + "epoch": 3.7970401691331923, + "grad_norm": 1.2081681489944458, + "learning_rate": 1.0645488589345396e-06, + "loss": 0.4457293748855591, + "step": 1796 + }, + { + "epoch": 3.8012684989429175, + "grad_norm": 3.3389010429382324, + "learning_rate": 1.0619205502284574e-06, + "loss": 0.8002062439918518, + "step": 1798 + }, + { + "epoch": 3.8054968287526427, + "grad_norm": 0.7985680103302002, + "learning_rate": 1.0593464988266811e-06, + "loss": 0.7525597810745239, + "step": 1800 + }, + { + "epoch": 3.809725158562368, + "grad_norm": 1.3357630968093872, + "learning_rate": 1.0568267361980225e-06, + "loss": 0.5861201286315918, + "step": 1802 + }, + { + "epoch": 3.813953488372093, + "grad_norm": 1.1272422075271606, + "learning_rate": 1.0543612931475935e-06, + "loss": 0.6722506284713745, + "step": 1804 + }, + { + "epoch": 3.8181818181818183, + "grad_norm": 1.450942039489746, + "learning_rate": 1.0519501998164247e-06, + "loss": 0.7887763381004333, + "step": 1806 + }, + { + "epoch": 3.822410147991543, + "grad_norm": 3.5325205326080322, + "learning_rate": 1.049593485681101e-06, + "loss": 0.16805249452590942, + "step": 1808 + }, + { + "epoch": 3.8266384778012688, + "grad_norm": 1.6206281185150146, + "learning_rate": 1.0472911795533997e-06, + "loss": 0.7603231072425842, + "step": 1810 + }, + { + "epoch": 3.8308668076109935, + "grad_norm": 1.4564605951309204, + "learning_rate": 1.045043309579938e-06, + "loss": 0.9016331434249878, + "step": 1812 + }, + { + "epoch": 3.8350951374207187, + "grad_norm": 0.8649386763572693, + "learning_rate": 1.0428499032418297e-06, + "loss": 0.6357676386833191, + "step": 1814 + }, + { + "epoch": 3.839323467230444, + "grad_norm": 0.9624889492988586, + "learning_rate": 1.0407109873543484e-06, + "loss": 0.4430287182331085, + "step": 1816 + }, + { + "epoch": 3.843551797040169, + "grad_norm": 2.074636220932007, + "learning_rate": 1.0386265880665996e-06, + "loss": 0.48888924717903137, + "step": 1818 + }, + { + "epoch": 3.8477801268498943, + "grad_norm": 0.9712066054344177, + "learning_rate": 1.036596730861203e-06, + "loss": 0.42531001567840576, + "step": 1820 + }, + { + "epoch": 3.8520084566596196, + "grad_norm": 0.3619837760925293, + "learning_rate": 1.0346214405539783e-06, + "loss": 0.4907451868057251, + "step": 1822 + }, + { + "epoch": 3.8562367864693448, + "grad_norm": 1.5417896509170532, + "learning_rate": 1.0327007412936419e-06, + "loss": 0.42517468333244324, + "step": 1824 + }, + { + "epoch": 3.8604651162790695, + "grad_norm": 3.981353759765625, + "learning_rate": 1.0308346565615158e-06, + "loss": 0.44803956151008606, + "step": 1826 + }, + { + "epoch": 3.864693446088795, + "grad_norm": 1.1224679946899414, + "learning_rate": 1.0290232091712337e-06, + "loss": 0.7744050025939941, + "step": 1828 + }, + { + "epoch": 3.86892177589852, + "grad_norm": 1.0671626329421997, + "learning_rate": 1.0272664212684687e-06, + "loss": 0.7128170132637024, + "step": 1830 + }, + { + "epoch": 3.873150105708245, + "grad_norm": 1.0114058256149292, + "learning_rate": 1.025564314330659e-06, + "loss": 0.49502280354499817, + "step": 1832 + }, + { + "epoch": 3.8773784355179703, + "grad_norm": 0.9319264888763428, + "learning_rate": 1.0239169091667447e-06, + "loss": 0.4061003029346466, + "step": 1834 + }, + { + "epoch": 3.8816067653276956, + "grad_norm": 0.2779659926891327, + "learning_rate": 1.022324225916917e-06, + "loss": 0.42833930253982544, + "step": 1836 + }, + { + "epoch": 3.8858350951374208, + "grad_norm": 0.24720549583435059, + "learning_rate": 1.0207862840523672e-06, + "loss": 0.36041924357414246, + "step": 1838 + }, + { + "epoch": 3.890063424947146, + "grad_norm": 1.8889857530593872, + "learning_rate": 1.0193031023750534e-06, + "loss": 0.6448405981063843, + "step": 1840 + }, + { + "epoch": 3.894291754756871, + "grad_norm": 0.8650567531585693, + "learning_rate": 1.017874699017468e-06, + "loss": 0.6602128744125366, + "step": 1842 + }, + { + "epoch": 3.898520084566596, + "grad_norm": 1.2487726211547852, + "learning_rate": 1.0165010914424142e-06, + "loss": 0.4448367655277252, + "step": 1844 + }, + { + "epoch": 3.9027484143763216, + "grad_norm": 5.131511688232422, + "learning_rate": 1.015182296442798e-06, + "loss": 0.43166065216064453, + "step": 1846 + }, + { + "epoch": 3.9069767441860463, + "grad_norm": 0.9479066729545593, + "learning_rate": 1.0139183301414176e-06, + "loss": 0.0830211341381073, + "step": 1848 + }, + { + "epoch": 3.9112050739957716, + "grad_norm": 1.2188801765441895, + "learning_rate": 1.0127092079907682e-06, + "loss": 0.649996280670166, + "step": 1850 + }, + { + "epoch": 3.9154334038054968, + "grad_norm": 0.41656494140625, + "learning_rate": 1.0115549447728551e-06, + "loss": 0.5674780607223511, + "step": 1852 + }, + { + "epoch": 3.919661733615222, + "grad_norm": 0.7337782979011536, + "learning_rate": 1.0104555545990088e-06, + "loss": 0.6366553902626038, + "step": 1854 + }, + { + "epoch": 3.923890063424947, + "grad_norm": 0.8625988960266113, + "learning_rate": 1.009411050909716e-06, + "loss": 0.35564541816711426, + "step": 1856 + }, + { + "epoch": 3.9281183932346724, + "grad_norm": 1.8072608709335327, + "learning_rate": 1.008421446474455e-06, + "loss": 0.6551876068115234, + "step": 1858 + }, + { + "epoch": 3.9323467230443976, + "grad_norm": 2.6653308868408203, + "learning_rate": 1.0074867533915367e-06, + "loss": 0.19086767733097076, + "step": 1860 + }, + { + "epoch": 3.9365750528541223, + "grad_norm": 5.576467990875244, + "learning_rate": 1.0066069830879593e-06, + "loss": 0.7493979930877686, + "step": 1862 + }, + { + "epoch": 3.940803382663848, + "grad_norm": 1.6586158275604248, + "learning_rate": 1.0057821463192687e-06, + "loss": 0.676756739616394, + "step": 1864 + }, + { + "epoch": 3.9450317124735728, + "grad_norm": 2.383901596069336, + "learning_rate": 1.005012253169426e-06, + "loss": 0.6339821815490723, + "step": 1866 + }, + { + "epoch": 3.949260042283298, + "grad_norm": 0.387660413980484, + "learning_rate": 1.0042973130506839e-06, + "loss": 0.004307010676711798, + "step": 1868 + }, + { + "epoch": 3.953488372093023, + "grad_norm": 1.16718327999115, + "learning_rate": 1.003637334703472e-06, + "loss": 0.7943294048309326, + "step": 1870 + }, + { + "epoch": 3.9577167019027484, + "grad_norm": 1.2603765726089478, + "learning_rate": 1.0030323261962922e-06, + "loss": 0.5086092948913574, + "step": 1872 + }, + { + "epoch": 3.9619450317124736, + "grad_norm": 10.705541610717773, + "learning_rate": 1.002482294925615e-06, + "loss": 0.5583027601242065, + "step": 1874 + }, + { + "epoch": 3.966173361522199, + "grad_norm": 1.3444347381591797, + "learning_rate": 1.0019872476157952e-06, + "loss": 0.6536118388175964, + "step": 1876 + }, + { + "epoch": 3.970401691331924, + "grad_norm": 8.445291519165039, + "learning_rate": 1.0015471903189834e-06, + "loss": 0.24211537837982178, + "step": 1878 + }, + { + "epoch": 3.974630021141649, + "grad_norm": 1.1503583192825317, + "learning_rate": 1.0011621284150583e-06, + "loss": 0.4144171178340912, + "step": 1880 + }, + { + "epoch": 3.9788583509513744, + "grad_norm": 5.531966686248779, + "learning_rate": 1.000832066611556e-06, + "loss": 0.6373070478439331, + "step": 1882 + }, + { + "epoch": 3.983086680761099, + "grad_norm": 1.393034815788269, + "learning_rate": 1.0005570089436137e-06, + "loss": 0.8069078922271729, + "step": 1884 + }, + { + "epoch": 3.9873150105708244, + "grad_norm": 1.1248987913131714, + "learning_rate": 1.0003369587739234e-06, + "loss": 0.6021481156349182, + "step": 1886 + }, + { + "epoch": 3.9915433403805496, + "grad_norm": 1.783697247505188, + "learning_rate": 1.0001719187926854e-06, + "loss": 0.717491626739502, + "step": 1888 + }, + { + "epoch": 3.995771670190275, + "grad_norm": 0.9147213101387024, + "learning_rate": 1.0000618910175811e-06, + "loss": 0.7117428779602051, + "step": 1890 + }, + { + "epoch": 4.0, + "grad_norm": 1.990949273109436, + "learning_rate": 1.000006876793743e-06, + "loss": 0.3688989579677582, + "step": 1892 + }, + { + "epoch": 4.0, + "step": 1892, + "total_flos": 3.5845266747435254e+18, + "train_loss": 0.8250222063222683, + "train_runtime": 8638.0168, + "train_samples_per_second": 6.571, + "train_steps_per_second": 0.219 + } + ], + "logging_steps": 2, + "max_steps": 1892, + "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": 3.5845266747435254e+18, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +}