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": 10.609610557556152, + "learning_rate": 1.0526315789473685e-07, + "loss": 1.8229680061340332, + "step": 2 + }, + { + "epoch": 0.008456659619450317, + "grad_norm": 1.2035231590270996, + "learning_rate": 3.1578947368421055e-07, + "loss": 2.2103376388549805, + "step": 4 + }, + { + "epoch": 0.012684989429175475, + "grad_norm": 13.37847900390625, + "learning_rate": 5.263157894736843e-07, + "loss": 2.2032673358917236, + "step": 6 + }, + { + "epoch": 0.016913319238900635, + "grad_norm": 1.7112228870391846, + "learning_rate": 7.368421052631579e-07, + "loss": 1.9956222772598267, + "step": 8 + }, + { + "epoch": 0.021141649048625793, + "grad_norm": 18.974014282226562, + "learning_rate": 9.473684210526317e-07, + "loss": 2.0128746032714844, + "step": 10 + }, + { + "epoch": 0.02536997885835095, + "grad_norm": 2.057965040206909, + "learning_rate": 1.1578947368421053e-06, + "loss": 1.6338518857955933, + "step": 12 + }, + { + "epoch": 0.02959830866807611, + "grad_norm": 1.3819606304168701, + "learning_rate": 1.3684210526315791e-06, + "loss": 1.6392173767089844, + "step": 14 + }, + { + "epoch": 0.03382663847780127, + "grad_norm": 2.3822357654571533, + "learning_rate": 1.5789473684210526e-06, + "loss": 1.7625806331634521, + "step": 16 + }, + { + "epoch": 0.03805496828752643, + "grad_norm": 1.3974515199661255, + "learning_rate": 1.7894736842105265e-06, + "loss": 1.7834045886993408, + "step": 18 + }, + { + "epoch": 0.042283298097251586, + "grad_norm": 6.584742069244385, + "learning_rate": 2.0000000000000003e-06, + "loss": 1.6609768867492676, + "step": 20 + }, + { + "epoch": 0.046511627906976744, + "grad_norm": 12.285764694213867, + "learning_rate": 2.2105263157894738e-06, + "loss": 1.493046522140503, + "step": 22 + }, + { + "epoch": 0.0507399577167019, + "grad_norm": 1.8775111436843872, + "learning_rate": 2.4210526315789477e-06, + "loss": 1.7588095664978027, + "step": 24 + }, + { + "epoch": 0.05496828752642706, + "grad_norm": 3.3788158893585205, + "learning_rate": 2.631578947368421e-06, + "loss": 1.365150809288025, + "step": 26 + }, + { + "epoch": 0.05919661733615222, + "grad_norm": 1.0759273767471313, + "learning_rate": 2.842105263157895e-06, + "loss": 1.637587070465088, + "step": 28 + }, + { + "epoch": 0.06342494714587738, + "grad_norm": 2.3137385845184326, + "learning_rate": 3.052631578947369e-06, + "loss": 1.2199206352233887, + "step": 30 + }, + { + "epoch": 0.06765327695560254, + "grad_norm": 2.1346375942230225, + "learning_rate": 3.2631578947368423e-06, + "loss": 0.9973738789558411, + "step": 32 + }, + { + "epoch": 0.07188160676532769, + "grad_norm": 1.4109710454940796, + "learning_rate": 3.473684210526316e-06, + "loss": 1.1656228303909302, + "step": 34 + }, + { + "epoch": 0.07610993657505286, + "grad_norm": 1.980239748954773, + "learning_rate": 3.6842105263157896e-06, + "loss": 1.3496501445770264, + "step": 36 + }, + { + "epoch": 0.080338266384778, + "grad_norm": 2.8445024490356445, + "learning_rate": 3.894736842105263e-06, + "loss": 1.2816426753997803, + "step": 38 + }, + { + "epoch": 0.08456659619450317, + "grad_norm": 17.366552352905273, + "learning_rate": 4.105263157894737e-06, + "loss": 1.2112103700637817, + "step": 40 + }, + { + "epoch": 0.08879492600422834, + "grad_norm": 2.2819364070892334, + "learning_rate": 4.315789473684211e-06, + "loss": 1.6811118125915527, + "step": 42 + }, + { + "epoch": 0.09302325581395349, + "grad_norm": 1.8612358570098877, + "learning_rate": 4.526315789473685e-06, + "loss": 1.5811426639556885, + "step": 44 + }, + { + "epoch": 0.09725158562367865, + "grad_norm": 1.0594091415405273, + "learning_rate": 4.736842105263158e-06, + "loss": 1.2700392007827759, + "step": 46 + }, + { + "epoch": 0.1014799154334038, + "grad_norm": 1.014484167098999, + "learning_rate": 4.947368421052632e-06, + "loss": 1.4919196367263794, + "step": 48 + }, + { + "epoch": 0.10570824524312897, + "grad_norm": 1.3634133338928223, + "learning_rate": 5.157894736842106e-06, + "loss": 1.4592434167861938, + "step": 50 + }, + { + "epoch": 0.10993657505285412, + "grad_norm": 6.238326072692871, + "learning_rate": 5.36842105263158e-06, + "loss": 0.7135005593299866, + "step": 52 + }, + { + "epoch": 0.11416490486257928, + "grad_norm": 1.0045223236083984, + "learning_rate": 5.578947368421052e-06, + "loss": 1.0439128875732422, + "step": 54 + }, + { + "epoch": 0.11839323467230443, + "grad_norm": 1.1906667947769165, + "learning_rate": 5.789473684210527e-06, + "loss": 1.3552573919296265, + "step": 56 + }, + { + "epoch": 0.1226215644820296, + "grad_norm": 0.9425953030586243, + "learning_rate": 6e-06, + "loss": 1.3783701658248901, + "step": 58 + }, + { + "epoch": 0.12684989429175475, + "grad_norm": 0.8465375304222107, + "learning_rate": 6.2105263157894745e-06, + "loss": 1.395394206047058, + "step": 60 + }, + { + "epoch": 0.13107822410147993, + "grad_norm": 4.073490619659424, + "learning_rate": 6.421052631578948e-06, + "loss": 0.7701259851455688, + "step": 62 + }, + { + "epoch": 0.13530655391120508, + "grad_norm": 1.6913920640945435, + "learning_rate": 6.631578947368421e-06, + "loss": 1.0387282371520996, + "step": 64 + }, + { + "epoch": 0.13953488372093023, + "grad_norm": 2.9340786933898926, + "learning_rate": 6.842105263157896e-06, + "loss": 1.3393646478652954, + "step": 66 + }, + { + "epoch": 0.14376321353065538, + "grad_norm": 1.5377577543258667, + "learning_rate": 7.052631578947369e-06, + "loss": 1.3740901947021484, + "step": 68 + }, + { + "epoch": 0.14799154334038056, + "grad_norm": 1.4497699737548828, + "learning_rate": 7.263157894736843e-06, + "loss": 1.3476486206054688, + "step": 70 + }, + { + "epoch": 0.1522198731501057, + "grad_norm": 0.9461106657981873, + "learning_rate": 7.473684210526316e-06, + "loss": 1.3423476219177246, + "step": 72 + }, + { + "epoch": 0.15644820295983086, + "grad_norm": 0.9822519421577454, + "learning_rate": 7.68421052631579e-06, + "loss": 0.9378120303153992, + "step": 74 + }, + { + "epoch": 0.160676532769556, + "grad_norm": 1.8004499673843384, + "learning_rate": 7.894736842105265e-06, + "loss": 0.8492949604988098, + "step": 76 + }, + { + "epoch": 0.1649048625792812, + "grad_norm": 1.6665072441101074, + "learning_rate": 8.105263157894736e-06, + "loss": 1.2846704721450806, + "step": 78 + }, + { + "epoch": 0.16913319238900634, + "grad_norm": 4.724425792694092, + "learning_rate": 8.315789473684212e-06, + "loss": 1.0786850452423096, + "step": 80 + }, + { + "epoch": 0.1733615221987315, + "grad_norm": 2.139698028564453, + "learning_rate": 8.526315789473685e-06, + "loss": 1.0848716497421265, + "step": 82 + }, + { + "epoch": 0.17758985200845667, + "grad_norm": 1.4293001890182495, + "learning_rate": 8.736842105263158e-06, + "loss": 0.8646807670593262, + "step": 84 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.022289991378784, + "learning_rate": 8.947368421052632e-06, + "loss": 1.7042136192321777, + "step": 86 + }, + { + "epoch": 0.18604651162790697, + "grad_norm": 2.3917741775512695, + "learning_rate": 9.157894736842105e-06, + "loss": 1.3036689758300781, + "step": 88 + }, + { + "epoch": 0.19027484143763213, + "grad_norm": 1.6193654537200928, + "learning_rate": 9.36842105263158e-06, + "loss": 0.991129994392395, + "step": 90 + }, + { + "epoch": 0.1945031712473573, + "grad_norm": 2.860992670059204, + "learning_rate": 9.578947368421054e-06, + "loss": 1.3122997283935547, + "step": 92 + }, + { + "epoch": 0.19873150105708245, + "grad_norm": 0.9149028658866882, + "learning_rate": 9.789473684210527e-06, + "loss": 0.9763924479484558, + "step": 94 + }, + { + "epoch": 0.2029598308668076, + "grad_norm": 1.0236929655075073, + "learning_rate": 1e-05, + "loss": 1.294936180114746, + "step": 96 + }, + { + "epoch": 0.20718816067653276, + "grad_norm": 1.3747087717056274, + "learning_rate": 9.999972492846047e-06, + "loss": 1.3845425844192505, + "step": 98 + }, + { + "epoch": 0.21141649048625794, + "grad_norm": 0.9011280536651611, + "learning_rate": 9.999889971720468e-06, + "loss": 1.3260761499404907, + "step": 100 + }, + { + "epoch": 0.2156448202959831, + "grad_norm": 1.9983714818954468, + "learning_rate": 9.999752437632121e-06, + "loss": 1.2449053525924683, + "step": 102 + }, + { + "epoch": 0.21987315010570824, + "grad_norm": 1.0277795791625977, + "learning_rate": 9.999559892262412e-06, + "loss": 1.3070706129074097, + "step": 104 + }, + { + "epoch": 0.22410147991543342, + "grad_norm": 1.7305063009262085, + "learning_rate": 9.999312337965291e-06, + "loss": 1.296847939491272, + "step": 106 + }, + { + "epoch": 0.22832980972515857, + "grad_norm": 1.2090764045715332, + "learning_rate": 9.999009777767202e-06, + "loss": 1.0146769285202026, + "step": 108 + }, + { + "epoch": 0.23255813953488372, + "grad_norm": 1.313328742980957, + "learning_rate": 9.99865221536707e-06, + "loss": 1.284561276435852, + "step": 110 + }, + { + "epoch": 0.23678646934460887, + "grad_norm": 2.268934965133667, + "learning_rate": 9.998239655136237e-06, + "loss": 1.355086088180542, + "step": 112 + }, + { + "epoch": 0.24101479915433405, + "grad_norm": 0.8743160367012024, + "learning_rate": 9.997772102118417e-06, + "loss": 1.5229133367538452, + "step": 114 + }, + { + "epoch": 0.2452431289640592, + "grad_norm": 2.378490686416626, + "learning_rate": 9.997249562029637e-06, + "loss": 1.328493595123291, + "step": 116 + }, + { + "epoch": 0.24947145877378435, + "grad_norm": 1.3889822959899902, + "learning_rate": 9.996672041258154e-06, + "loss": 1.3420453071594238, + "step": 118 + }, + { + "epoch": 0.2536997885835095, + "grad_norm": 1.834333896636963, + "learning_rate": 9.996039546864395e-06, + "loss": 0.8575987815856934, + "step": 120 + }, + { + "epoch": 0.25792811839323465, + "grad_norm": 6.752488613128662, + "learning_rate": 9.995352086580858e-06, + "loss": 1.2992044687271118, + "step": 122 + }, + { + "epoch": 0.26215644820295986, + "grad_norm": 0.7973200678825378, + "learning_rate": 9.994609668812022e-06, + "loss": 1.148786187171936, + "step": 124 + }, + { + "epoch": 0.266384778012685, + "grad_norm": 0.9675792455673218, + "learning_rate": 9.993812302634237e-06, + "loss": 1.3500239849090576, + "step": 126 + }, + { + "epoch": 0.27061310782241016, + "grad_norm": 0.8263481259346008, + "learning_rate": 9.99295999779563e-06, + "loss": 0.9418169856071472, + "step": 128 + }, + { + "epoch": 0.2748414376321353, + "grad_norm": 0.9413770437240601, + "learning_rate": 9.99205276471597e-06, + "loss": 0.9810768961906433, + "step": 130 + }, + { + "epoch": 0.27906976744186046, + "grad_norm": 1.322837233543396, + "learning_rate": 9.991090614486546e-06, + "loss": 1.2799700498580933, + "step": 132 + }, + { + "epoch": 0.2832980972515856, + "grad_norm": 2.4948737621307373, + "learning_rate": 9.990073558870028e-06, + "loss": 0.8641204833984375, + "step": 134 + }, + { + "epoch": 0.28752642706131076, + "grad_norm": 2.1680777072906494, + "learning_rate": 9.989001610300333e-06, + "loss": 1.1166131496429443, + "step": 136 + }, + { + "epoch": 0.2917547568710359, + "grad_norm": 1.2369376420974731, + "learning_rate": 9.987874781882461e-06, + "loss": 1.380534291267395, + "step": 138 + }, + { + "epoch": 0.2959830866807611, + "grad_norm": 0.8044794201850891, + "learning_rate": 9.986693087392341e-06, + "loss": 0.9397739171981812, + "step": 140 + }, + { + "epoch": 0.30021141649048627, + "grad_norm": 0.7351009845733643, + "learning_rate": 9.985456541276667e-06, + "loss": 1.5624629259109497, + "step": 142 + }, + { + "epoch": 0.3044397463002114, + "grad_norm": 5.393040180206299, + "learning_rate": 9.984165158652708e-06, + "loss": 1.0967549085617065, + "step": 144 + }, + { + "epoch": 0.3086680761099366, + "grad_norm": 2.55499267578125, + "learning_rate": 9.982818955308135e-06, + "loss": 0.6742914319038391, + "step": 146 + }, + { + "epoch": 0.3128964059196617, + "grad_norm": 1.0078721046447754, + "learning_rate": 9.98141794770083e-06, + "loss": 1.2606271505355835, + "step": 148 + }, + { + "epoch": 0.3171247357293869, + "grad_norm": 0.7713409066200256, + "learning_rate": 9.979962152958669e-06, + "loss": 1.267325520515442, + "step": 150 + }, + { + "epoch": 0.321353065539112, + "grad_norm": 3.7035458087921143, + "learning_rate": 9.97845158887933e-06, + "loss": 1.1791411638259888, + "step": 152 + }, + { + "epoch": 0.32558139534883723, + "grad_norm": 3.4524617195129395, + "learning_rate": 9.976886273930063e-06, + "loss": 0.7506053447723389, + "step": 154 + }, + { + "epoch": 0.3298097251585624, + "grad_norm": 1.0968884229660034, + "learning_rate": 9.975266227247474e-06, + "loss": 1.2544300556182861, + "step": 156 + }, + { + "epoch": 0.33403805496828753, + "grad_norm": 0.9143646955490112, + "learning_rate": 9.973591468637286e-06, + "loss": 1.2751351594924927, + "step": 158 + }, + { + "epoch": 0.3382663847780127, + "grad_norm": 0.9158719778060913, + "learning_rate": 9.971862018574094e-06, + "loss": 1.255610704421997, + "step": 160 + }, + { + "epoch": 0.34249471458773784, + "grad_norm": 1.8719379901885986, + "learning_rate": 9.97007789820112e-06, + "loss": 1.094406008720398, + "step": 162 + }, + { + "epoch": 0.346723044397463, + "grad_norm": 1.4100563526153564, + "learning_rate": 9.968239129329953e-06, + "loss": 0.9665659070014954, + "step": 164 + }, + { + "epoch": 0.35095137420718814, + "grad_norm": 0.9090960621833801, + "learning_rate": 9.96634573444028e-06, + "loss": 0.8870795965194702, + "step": 166 + }, + { + "epoch": 0.35517970401691334, + "grad_norm": 1.367483377456665, + "learning_rate": 9.964397736679615e-06, + "loss": 1.206871747970581, + "step": 168 + }, + { + "epoch": 0.3594080338266385, + "grad_norm": 0.7214027643203735, + "learning_rate": 9.962395159863014e-06, + "loss": 1.2130553722381592, + "step": 170 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 0.9371772408485413, + "learning_rate": 9.960338028472781e-06, + "loss": 1.0388234853744507, + "step": 172 + }, + { + "epoch": 0.3678646934460888, + "grad_norm": 0.9869715571403503, + "learning_rate": 9.958226367658176e-06, + "loss": 1.3533731698989868, + "step": 174 + }, + { + "epoch": 0.37209302325581395, + "grad_norm": 1.5355944633483887, + "learning_rate": 9.9560602032351e-06, + "loss": 1.0356580018997192, + "step": 176 + }, + { + "epoch": 0.3763213530655391, + "grad_norm": 1.7834546566009521, + "learning_rate": 9.953839561685783e-06, + "loss": 0.8860217928886414, + "step": 178 + }, + { + "epoch": 0.38054968287526425, + "grad_norm": 0.7529613375663757, + "learning_rate": 9.95156447015846e-06, + "loss": 1.2859153747558594, + "step": 180 + }, + { + "epoch": 0.38477801268498946, + "grad_norm": 1.165433645248413, + "learning_rate": 9.94923495646704e-06, + "loss": 1.528343915939331, + "step": 182 + }, + { + "epoch": 0.3890063424947146, + "grad_norm": 1.2064543962478638, + "learning_rate": 9.946851049090764e-06, + "loss": 1.3444085121154785, + "step": 184 + }, + { + "epoch": 0.39323467230443976, + "grad_norm": 1.1523786783218384, + "learning_rate": 9.944412777173854e-06, + "loss": 1.1340974569320679, + "step": 186 + }, + { + "epoch": 0.3974630021141649, + "grad_norm": 0.9289480447769165, + "learning_rate": 9.941920170525168e-06, + "loss": 0.9834291338920593, + "step": 188 + }, + { + "epoch": 0.40169133192389006, + "grad_norm": 1.0578315258026123, + "learning_rate": 9.939373259617822e-06, + "loss": 1.0226718187332153, + "step": 190 + }, + { + "epoch": 0.4059196617336152, + "grad_norm": 1.1814934015274048, + "learning_rate": 9.936772075588825e-06, + "loss": 1.299898624420166, + "step": 192 + }, + { + "epoch": 0.41014799154334036, + "grad_norm": 1.1240030527114868, + "learning_rate": 9.934116650238697e-06, + "loss": 1.0035775899887085, + "step": 194 + }, + { + "epoch": 0.4143763213530655, + "grad_norm": 3.881202220916748, + "learning_rate": 9.931407016031078e-06, + "loss": 1.2822128534317017, + "step": 196 + }, + { + "epoch": 0.4186046511627907, + "grad_norm": 1.5710660219192505, + "learning_rate": 9.928643206092338e-06, + "loss": 1.0776773691177368, + "step": 198 + }, + { + "epoch": 0.42283298097251587, + "grad_norm": 1.0270471572875977, + "learning_rate": 9.925825254211161e-06, + "loss": 1.182267427444458, + "step": 200 + }, + { + "epoch": 0.427061310782241, + "grad_norm": 2.6018481254577637, + "learning_rate": 9.922953194838143e-06, + "loss": 0.5917832255363464, + "step": 202 + }, + { + "epoch": 0.4312896405919662, + "grad_norm": 0.7647141218185425, + "learning_rate": 9.920027063085363e-06, + "loss": 1.2310165166854858, + "step": 204 + }, + { + "epoch": 0.4355179704016913, + "grad_norm": 1.6856355667114258, + "learning_rate": 9.917046894725957e-06, + "loss": 1.308571457862854, + "step": 206 + }, + { + "epoch": 0.4397463002114165, + "grad_norm": 1.7268143892288208, + "learning_rate": 9.91401272619368e-06, + "loss": 0.7988643646240234, + "step": 208 + }, + { + "epoch": 0.4439746300211416, + "grad_norm": 0.8977281451225281, + "learning_rate": 9.910924594582463e-06, + "loss": 1.2860561609268188, + "step": 210 + }, + { + "epoch": 0.44820295983086683, + "grad_norm": 25.122228622436523, + "learning_rate": 9.907782537645955e-06, + "loss": 1.2171250581741333, + "step": 212 + }, + { + "epoch": 0.452431289640592, + "grad_norm": 1.5233873128890991, + "learning_rate": 9.904586593797062e-06, + "loss": 1.4229017496109009, + "step": 214 + }, + { + "epoch": 0.45665961945031713, + "grad_norm": 1.27586829662323, + "learning_rate": 9.901336802107485e-06, + "loss": 0.6469933390617371, + "step": 216 + }, + { + "epoch": 0.4608879492600423, + "grad_norm": 5.587992191314697, + "learning_rate": 9.89803320230723e-06, + "loss": 1.011372447013855, + "step": 218 + }, + { + "epoch": 0.46511627906976744, + "grad_norm": 1.1823580265045166, + "learning_rate": 9.894675834784136e-06, + "loss": 1.261022686958313, + "step": 220 + }, + { + "epoch": 0.4693446088794926, + "grad_norm": 1.058074712753296, + "learning_rate": 9.891264740583367e-06, + "loss": 1.2349940538406372, + "step": 222 + }, + { + "epoch": 0.47357293868921774, + "grad_norm": 0.8530188202857971, + "learning_rate": 9.88779996140692e-06, + "loss": 1.3473305702209473, + "step": 224 + }, + { + "epoch": 0.47780126849894294, + "grad_norm": 0.8320378661155701, + "learning_rate": 9.884281539613109e-06, + "loss": 1.242853045463562, + "step": 226 + }, + { + "epoch": 0.4820295983086681, + "grad_norm": 0.9211630821228027, + "learning_rate": 9.880709518216061e-06, + "loss": 1.2367050647735596, + "step": 228 + }, + { + "epoch": 0.48625792811839325, + "grad_norm": 1.6035828590393066, + "learning_rate": 9.87708394088517e-06, + "loss": 0.9145620465278625, + "step": 230 + }, + { + "epoch": 0.4904862579281184, + "grad_norm": 1.8635786771774292, + "learning_rate": 9.873404851944573e-06, + "loss": 1.2485287189483643, + "step": 232 + }, + { + "epoch": 0.49471458773784355, + "grad_norm": 0.9878160953521729, + "learning_rate": 9.869672296372611e-06, + "loss": 0.9443119168281555, + "step": 234 + }, + { + "epoch": 0.4989429175475687, + "grad_norm": 1.9962083101272583, + "learning_rate": 9.86588631980128e-06, + "loss": 0.8812016248703003, + "step": 236 + }, + { + "epoch": 0.5031712473572939, + "grad_norm": 4.044755935668945, + "learning_rate": 9.86204696851566e-06, + "loss": 0.8383482694625854, + "step": 238 + }, + { + "epoch": 0.507399577167019, + "grad_norm": 1.297173023223877, + "learning_rate": 9.858154289453363e-06, + "loss": 0.9133360981941223, + "step": 240 + }, + { + "epoch": 0.5116279069767442, + "grad_norm": 0.8767004013061523, + "learning_rate": 9.854208330203958e-06, + "loss": 1.3159452676773071, + "step": 242 + }, + { + "epoch": 0.5158562367864693, + "grad_norm": 3.041069507598877, + "learning_rate": 9.850209139008377e-06, + "loss": 1.3311914205551147, + "step": 244 + }, + { + "epoch": 0.5200845665961945, + "grad_norm": 1.778280258178711, + "learning_rate": 9.846156764758344e-06, + "loss": 1.2870732545852661, + "step": 246 + }, + { + "epoch": 0.5243128964059197, + "grad_norm": 1.5028165578842163, + "learning_rate": 9.842051256995759e-06, + "loss": 1.237648367881775, + "step": 248 + }, + { + "epoch": 0.5285412262156448, + "grad_norm": 1.4685592651367188, + "learning_rate": 9.837892665912105e-06, + "loss": 1.1841355562210083, + "step": 250 + }, + { + "epoch": 0.53276955602537, + "grad_norm": 3.6482200622558594, + "learning_rate": 9.833681042347827e-06, + "loss": 1.1595423221588135, + "step": 252 + }, + { + "epoch": 0.5369978858350951, + "grad_norm": 1.2831525802612305, + "learning_rate": 9.829416437791721e-06, + "loss": 1.7401180267333984, + "step": 254 + }, + { + "epoch": 0.5412262156448203, + "grad_norm": 1.2406295537948608, + "learning_rate": 9.825098904380284e-06, + "loss": 1.2138323783874512, + "step": 256 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 1.241223931312561, + "learning_rate": 9.820728494897101e-06, + "loss": 1.388733983039856, + "step": 258 + }, + { + "epoch": 0.5496828752642706, + "grad_norm": 0.9619345664978027, + "learning_rate": 9.816305262772182e-06, + "loss": 0.9950351715087891, + "step": 260 + }, + { + "epoch": 0.5539112050739958, + "grad_norm": 1.1318546533584595, + "learning_rate": 9.811829262081315e-06, + "loss": 1.1954119205474854, + "step": 262 + }, + { + "epoch": 0.5581395348837209, + "grad_norm": 1.0538593530654907, + "learning_rate": 9.807300547545411e-06, + "loss": 1.0519647598266602, + "step": 264 + }, + { + "epoch": 0.5623678646934461, + "grad_norm": 0.9785088300704956, + "learning_rate": 9.802719174529821e-06, + "loss": 1.4954670667648315, + "step": 266 + }, + { + "epoch": 0.5665961945031712, + "grad_norm": 1.3046112060546875, + "learning_rate": 9.798085199043672e-06, + "loss": 0.5226970911026001, + "step": 268 + }, + { + "epoch": 0.5708245243128964, + "grad_norm": 3.686903715133667, + "learning_rate": 9.793398677739177e-06, + "loss": 1.0078847408294678, + "step": 270 + }, + { + "epoch": 0.5750528541226215, + "grad_norm": 1.5437880754470825, + "learning_rate": 9.788659667910942e-06, + "loss": 1.1723384857177734, + "step": 272 + }, + { + "epoch": 0.5792811839323467, + "grad_norm": 0.9986572265625, + "learning_rate": 9.783868227495262e-06, + "loss": 0.9541176557540894, + "step": 274 + }, + { + "epoch": 0.5835095137420718, + "grad_norm": 1.0669474601745605, + "learning_rate": 9.779024415069427e-06, + "loss": 0.9416250586509705, + "step": 276 + }, + { + "epoch": 0.587737843551797, + "grad_norm": 0.4588918685913086, + "learning_rate": 9.774128289850987e-06, + "loss": 0.8676360845565796, + "step": 278 + }, + { + "epoch": 0.5919661733615222, + "grad_norm": 1.3972774744033813, + "learning_rate": 9.769179911697037e-06, + "loss": 1.2311569452285767, + "step": 280 + }, + { + "epoch": 0.5961945031712473, + "grad_norm": 1.0819487571716309, + "learning_rate": 9.764179341103493e-06, + "loss": 1.4673097133636475, + "step": 282 + }, + { + "epoch": 0.6004228329809725, + "grad_norm": 0.45675915479660034, + "learning_rate": 9.759126639204337e-06, + "loss": 0.9332614541053772, + "step": 284 + }, + { + "epoch": 0.6046511627906976, + "grad_norm": 3.103182315826416, + "learning_rate": 9.75402186777088e-06, + "loss": 1.3119271993637085, + "step": 286 + }, + { + "epoch": 0.6088794926004228, + "grad_norm": 3.2873263359069824, + "learning_rate": 9.748865089211004e-06, + "loss": 0.4315389394760132, + "step": 288 + }, + { + "epoch": 0.6131078224101479, + "grad_norm": 0.7815708518028259, + "learning_rate": 9.743656366568398e-06, + "loss": 0.9495774507522583, + "step": 290 + }, + { + "epoch": 0.6173361522198731, + "grad_norm": 4.356925010681152, + "learning_rate": 9.738395763521792e-06, + "loss": 0.8789107203483582, + "step": 292 + }, + { + "epoch": 0.6215644820295984, + "grad_norm": 7.836264610290527, + "learning_rate": 9.733083344384164e-06, + "loss": 0.7236508727073669, + "step": 294 + }, + { + "epoch": 0.6257928118393234, + "grad_norm": 3.2982900142669678, + "learning_rate": 9.727719174101981e-06, + "loss": 0.7367111444473267, + "step": 296 + }, + { + "epoch": 0.6300211416490487, + "grad_norm": 4.145655155181885, + "learning_rate": 9.722303318254376e-06, + "loss": 1.2766035795211792, + "step": 298 + }, + { + "epoch": 0.6342494714587738, + "grad_norm": 1.1087963581085205, + "learning_rate": 9.71683584305236e-06, + "loss": 1.2390812635421753, + "step": 300 + }, + { + "epoch": 0.638477801268499, + "grad_norm": 1.0106208324432373, + "learning_rate": 9.71131681533802e-06, + "loss": 0.7159666419029236, + "step": 302 + }, + { + "epoch": 0.642706131078224, + "grad_norm": 1.1209036111831665, + "learning_rate": 9.705746302583682e-06, + "loss": 0.8513851165771484, + "step": 304 + }, + { + "epoch": 0.6469344608879493, + "grad_norm": 1.2404656410217285, + "learning_rate": 9.700124372891104e-06, + "loss": 0.9119742512702942, + "step": 306 + }, + { + "epoch": 0.6511627906976745, + "grad_norm": 0.9948480725288391, + "learning_rate": 9.694451094990636e-06, + "loss": 1.2315819263458252, + "step": 308 + }, + { + "epoch": 0.6553911205073996, + "grad_norm": 3.201961040496826, + "learning_rate": 9.68872653824038e-06, + "loss": 1.204048991203308, + "step": 310 + }, + { + "epoch": 0.6596194503171248, + "grad_norm": 0.8589417338371277, + "learning_rate": 9.68295077262534e-06, + "loss": 1.2040612697601318, + "step": 312 + }, + { + "epoch": 0.6638477801268499, + "grad_norm": 1.525591254234314, + "learning_rate": 9.677123868756573e-06, + "loss": 1.229467511177063, + "step": 314 + }, + { + "epoch": 0.6680761099365751, + "grad_norm": 1.6267969608306885, + "learning_rate": 9.671245897870318e-06, + "loss": 1.0793099403381348, + "step": 316 + }, + { + "epoch": 0.6723044397463002, + "grad_norm": 1.1831475496292114, + "learning_rate": 9.665316931827131e-06, + "loss": 0.7765912413597107, + "step": 318 + }, + { + "epoch": 0.6765327695560254, + "grad_norm": 4.7630534172058105, + "learning_rate": 9.659337043111006e-06, + "loss": 1.242596983909607, + "step": 320 + }, + { + "epoch": 0.6807610993657506, + "grad_norm": 0.9527212381362915, + "learning_rate": 9.653306304828483e-06, + "loss": 1.307033658027649, + "step": 322 + }, + { + "epoch": 0.6849894291754757, + "grad_norm": 0.9060173630714417, + "learning_rate": 9.647224790707762e-06, + "loss": 1.0850989818572998, + "step": 324 + }, + { + "epoch": 0.6892177589852009, + "grad_norm": 1.742342233657837, + "learning_rate": 9.641092575097795e-06, + "loss": 1.2613813877105713, + "step": 326 + }, + { + "epoch": 0.693446088794926, + "grad_norm": 2.3110523223876953, + "learning_rate": 9.63490973296738e-06, + "loss": 1.169764518737793, + "step": 328 + }, + { + "epoch": 0.6976744186046512, + "grad_norm": 1.8334541320800781, + "learning_rate": 9.62867633990425e-06, + "loss": 0.8696043491363525, + "step": 330 + }, + { + "epoch": 0.7019027484143763, + "grad_norm": 0.8396955132484436, + "learning_rate": 9.622392472114136e-06, + "loss": 1.2531880140304565, + "step": 332 + }, + { + "epoch": 0.7061310782241015, + "grad_norm": 1.9480777978897095, + "learning_rate": 9.61605820641985e-06, + "loss": 1.2122329473495483, + "step": 334 + }, + { + "epoch": 0.7103594080338267, + "grad_norm": 1.1044273376464844, + "learning_rate": 9.60967362026033e-06, + "loss": 1.2058602571487427, + "step": 336 + }, + { + "epoch": 0.7145877378435518, + "grad_norm": 5.243762016296387, + "learning_rate": 9.60323879168971e-06, + "loss": 1.3080583810806274, + "step": 338 + }, + { + "epoch": 0.718816067653277, + "grad_norm": 2.4650166034698486, + "learning_rate": 9.596753799376355e-06, + "loss": 1.2157717943191528, + "step": 340 + }, + { + "epoch": 0.7230443974630021, + "grad_norm": 1.497857928276062, + "learning_rate": 9.5902187226019e-06, + "loss": 1.1279668807983398, + "step": 342 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.8875362873077393, + "learning_rate": 9.583633641260285e-06, + "loss": 0.9240359663963318, + "step": 344 + }, + { + "epoch": 0.7315010570824524, + "grad_norm": 2.311155080795288, + "learning_rate": 9.576998635856774e-06, + "loss": 0.8495903015136719, + "step": 346 + }, + { + "epoch": 0.7357293868921776, + "grad_norm": 2.6599528789520264, + "learning_rate": 9.570313787506974e-06, + "loss": 0.9777283072471619, + "step": 348 + }, + { + "epoch": 0.7399577167019028, + "grad_norm": 0.6879770159721375, + "learning_rate": 9.563579177935844e-06, + "loss": 1.2996944189071655, + "step": 350 + }, + { + "epoch": 0.7441860465116279, + "grad_norm": 2.7574212551116943, + "learning_rate": 9.556794889476687e-06, + "loss": 1.1595312356948853, + "step": 352 + }, + { + "epoch": 0.7484143763213531, + "grad_norm": 1.8974783420562744, + "learning_rate": 9.549961005070161e-06, + "loss": 0.7094411849975586, + "step": 354 + }, + { + "epoch": 0.7526427061310782, + "grad_norm": 2.260322332382202, + "learning_rate": 9.543077608263245e-06, + "loss": 1.025563359260559, + "step": 356 + }, + { + "epoch": 0.7568710359408034, + "grad_norm": 1.0396032333374023, + "learning_rate": 9.536144783208229e-06, + "loss": 1.2061057090759277, + "step": 358 + }, + { + "epoch": 0.7610993657505285, + "grad_norm": 1.6093682050704956, + "learning_rate": 9.529162614661687e-06, + "loss": 0.7457548379898071, + "step": 360 + }, + { + "epoch": 0.7653276955602537, + "grad_norm": 1.2853434085845947, + "learning_rate": 9.522131187983433e-06, + "loss": 1.0880964994430542, + "step": 362 + }, + { + "epoch": 0.7695560253699789, + "grad_norm": 2.579901933670044, + "learning_rate": 9.515050589135485e-06, + "loss": 1.052270531654358, + "step": 364 + }, + { + "epoch": 0.773784355179704, + "grad_norm": 1.3049057722091675, + "learning_rate": 9.507920904681006e-06, + "loss": 1.2584377527236938, + "step": 366 + }, + { + "epoch": 0.7780126849894292, + "grad_norm": 1.1259350776672363, + "learning_rate": 9.500742221783257e-06, + "loss": 1.2016220092773438, + "step": 368 + }, + { + "epoch": 0.7822410147991543, + "grad_norm": 5.495427131652832, + "learning_rate": 9.493514628204516e-06, + "loss": 0.8634753227233887, + "step": 370 + }, + { + "epoch": 0.7864693446088795, + "grad_norm": 0.8340354561805725, + "learning_rate": 9.48623821230502e-06, + "loss": 1.1856554746627808, + "step": 372 + }, + { + "epoch": 0.7906976744186046, + "grad_norm": 0.9506821036338806, + "learning_rate": 9.478913063041878e-06, + "loss": 1.2069045305252075, + "step": 374 + }, + { + "epoch": 0.7949260042283298, + "grad_norm": 1.031280517578125, + "learning_rate": 9.471539269967982e-06, + "loss": 1.2852314710617065, + "step": 376 + }, + { + "epoch": 0.7991543340380549, + "grad_norm": 1.9207712411880493, + "learning_rate": 9.464116923230913e-06, + "loss": 0.9563255906105042, + "step": 378 + }, + { + "epoch": 0.8033826638477801, + "grad_norm": 3.2214620113372803, + "learning_rate": 9.456646113571844e-06, + "loss": 1.2986749410629272, + "step": 380 + }, + { + "epoch": 0.8076109936575053, + "grad_norm": 1.0131698846817017, + "learning_rate": 9.449126932324424e-06, + "loss": 1.1665711402893066, + "step": 382 + }, + { + "epoch": 0.8118393234672304, + "grad_norm": 0.9806280732154846, + "learning_rate": 9.441559471413665e-06, + "loss": 0.9619807004928589, + "step": 384 + }, + { + "epoch": 0.8160676532769556, + "grad_norm": 3.777022123336792, + "learning_rate": 9.433943823354816e-06, + "loss": 0.9550887942314148, + "step": 386 + }, + { + "epoch": 0.8202959830866807, + "grad_norm": 1.4736453294754028, + "learning_rate": 9.426280081252232e-06, + "loss": 1.4764561653137207, + "step": 388 + }, + { + "epoch": 0.8245243128964059, + "grad_norm": 2.512502431869507, + "learning_rate": 9.418568338798245e-06, + "loss": 1.1447840929031372, + "step": 390 + }, + { + "epoch": 0.828752642706131, + "grad_norm": 2.4704666137695312, + "learning_rate": 9.410808690271999e-06, + "loss": 0.9280427098274231, + "step": 392 + }, + { + "epoch": 0.8329809725158562, + "grad_norm": 0.955447256565094, + "learning_rate": 9.40300123053832e-06, + "loss": 1.2348592281341553, + "step": 394 + }, + { + "epoch": 0.8372093023255814, + "grad_norm": 0.47231826186180115, + "learning_rate": 9.395146055046535e-06, + "loss": 0.7616856098175049, + "step": 396 + }, + { + "epoch": 0.8414376321353065, + "grad_norm": 0.8924692273139954, + "learning_rate": 9.387243259829323e-06, + "loss": 0.6794538497924805, + "step": 398 + }, + { + "epoch": 0.8456659619450317, + "grad_norm": 1.3153858184814453, + "learning_rate": 9.37929294150153e-06, + "loss": 1.122640609741211, + "step": 400 + }, + { + "epoch": 0.8498942917547568, + "grad_norm": 1.957002878189087, + "learning_rate": 9.371295197258992e-06, + "loss": 0.9678458571434021, + "step": 402 + }, + { + "epoch": 0.854122621564482, + "grad_norm": 1.661242127418518, + "learning_rate": 9.363250124877345e-06, + "loss": 0.8274236917495728, + "step": 404 + }, + { + "epoch": 0.8583509513742071, + "grad_norm": 1.2411015033721924, + "learning_rate": 9.35515782271083e-06, + "loss": 0.9946659803390503, + "step": 406 + }, + { + "epoch": 0.8625792811839323, + "grad_norm": 1.8923640251159668, + "learning_rate": 9.347018389691093e-06, + "loss": 0.9779168367385864, + "step": 408 + }, + { + "epoch": 0.8668076109936576, + "grad_norm": 9.132567405700684, + "learning_rate": 9.338831925325974e-06, + "loss": 0.8799471855163574, + "step": 410 + }, + { + "epoch": 0.8710359408033826, + "grad_norm": 1.8964372873306274, + "learning_rate": 9.330598529698287e-06, + "loss": 1.1820063591003418, + "step": 412 + }, + { + "epoch": 0.8752642706131079, + "grad_norm": 1.4616568088531494, + "learning_rate": 9.322318303464603e-06, + "loss": 1.0221836566925049, + "step": 414 + }, + { + "epoch": 0.879492600422833, + "grad_norm": 7.450461387634277, + "learning_rate": 9.313991347854015e-06, + "loss": 1.1125426292419434, + "step": 416 + }, + { + "epoch": 0.8837209302325582, + "grad_norm": 0.9838056564331055, + "learning_rate": 9.305617764666897e-06, + "loss": 1.2110792398452759, + "step": 418 + }, + { + "epoch": 0.8879492600422833, + "grad_norm": 1.8934524059295654, + "learning_rate": 9.297197656273672e-06, + "loss": 1.1955455541610718, + "step": 420 + }, + { + "epoch": 0.8921775898520085, + "grad_norm": 3.026000738143921, + "learning_rate": 9.288731125613544e-06, + "loss": 0.939410924911499, + "step": 422 + }, + { + "epoch": 0.8964059196617337, + "grad_norm": 1.1672602891921997, + "learning_rate": 9.280218276193255e-06, + "loss": 1.2234692573547363, + "step": 424 + }, + { + "epoch": 0.9006342494714588, + "grad_norm": 0.8628170490264893, + "learning_rate": 9.271659212085807e-06, + "loss": 1.0855259895324707, + "step": 426 + }, + { + "epoch": 0.904862579281184, + "grad_norm": 1.3510693311691284, + "learning_rate": 9.263054037929199e-06, + "loss": 0.878591775894165, + "step": 428 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 1.7558636665344238, + "learning_rate": 9.254402858925142e-06, + "loss": 1.1033222675323486, + "step": 430 + }, + { + "epoch": 0.9133192389006343, + "grad_norm": 1.3664168119430542, + "learning_rate": 9.245705780837774e-06, + "loss": 1.2542273998260498, + "step": 432 + }, + { + "epoch": 0.9175475687103594, + "grad_norm": 2.6412429809570312, + "learning_rate": 9.236962909992368e-06, + "loss": 0.5288431644439697, + "step": 434 + }, + { + "epoch": 0.9217758985200846, + "grad_norm": 5.031874656677246, + "learning_rate": 9.228174353274037e-06, + "loss": 0.3886476755142212, + "step": 436 + }, + { + "epoch": 0.9260042283298098, + "grad_norm": 0.47040000557899475, + "learning_rate": 9.219340218126413e-06, + "loss": 0.9618459939956665, + "step": 438 + }, + { + "epoch": 0.9302325581395349, + "grad_norm": 3.2738940715789795, + "learning_rate": 9.210460612550348e-06, + "loss": 1.1283197402954102, + "step": 440 + }, + { + "epoch": 0.9344608879492601, + "grad_norm": 7.959907531738281, + "learning_rate": 9.20153564510259e-06, + "loss": 0.755107045173645, + "step": 442 + }, + { + "epoch": 0.9386892177589852, + "grad_norm": 1.6950106620788574, + "learning_rate": 9.19256542489445e-06, + "loss": 1.24224853515625, + "step": 444 + }, + { + "epoch": 0.9429175475687104, + "grad_norm": 1.1565735340118408, + "learning_rate": 9.183550061590476e-06, + "loss": 1.0946691036224365, + "step": 446 + }, + { + "epoch": 0.9471458773784355, + "grad_norm": 2.8532538414001465, + "learning_rate": 9.174489665407103e-06, + "loss": 1.1125609874725342, + "step": 448 + }, + { + "epoch": 0.9513742071881607, + "grad_norm": 14.895803451538086, + "learning_rate": 9.165384347111317e-06, + "loss": 0.7967692017555237, + "step": 450 + }, + { + "epoch": 0.9556025369978859, + "grad_norm": 1.1173516511917114, + "learning_rate": 9.156234218019291e-06, + "loss": 0.9280839562416077, + "step": 452 + }, + { + "epoch": 0.959830866807611, + "grad_norm": 1.9179232120513916, + "learning_rate": 9.147039389995029e-06, + "loss": 1.3109300136566162, + "step": 454 + }, + { + "epoch": 0.9640591966173362, + "grad_norm": 18.68795394897461, + "learning_rate": 9.137799975448998e-06, + "loss": 0.6728971600532532, + "step": 456 + }, + { + "epoch": 0.9682875264270613, + "grad_norm": 7.34496545791626, + "learning_rate": 9.128516087336755e-06, + "loss": 0.8823999166488647, + "step": 458 + }, + { + "epoch": 0.9725158562367865, + "grad_norm": 1.7849724292755127, + "learning_rate": 9.119187839157559e-06, + "loss": 1.2286086082458496, + "step": 460 + }, + { + "epoch": 0.9767441860465116, + "grad_norm": 1.213707685470581, + "learning_rate": 9.109815344952995e-06, + "loss": 1.2127104997634888, + "step": 462 + }, + { + "epoch": 0.9809725158562368, + "grad_norm": 1.2118189334869385, + "learning_rate": 9.100398719305566e-06, + "loss": 0.7322313785552979, + "step": 464 + }, + { + "epoch": 0.985200845665962, + "grad_norm": 1.270865559577942, + "learning_rate": 9.090938077337305e-06, + "loss": 1.1820967197418213, + "step": 466 + }, + { + "epoch": 0.9894291754756871, + "grad_norm": 12.68324089050293, + "learning_rate": 9.081433534708363e-06, + "loss": 0.978464663028717, + "step": 468 + }, + { + "epoch": 0.9936575052854123, + "grad_norm": 4.5913310050964355, + "learning_rate": 9.071885207615591e-06, + "loss": 1.239919900894165, + "step": 470 + }, + { + "epoch": 0.9978858350951374, + "grad_norm": 1.2919870615005493, + "learning_rate": 9.062293212791123e-06, + "loss": 1.2460461854934692, + "step": 472 + }, + { + "epoch": 1.0021141649048626, + "grad_norm": 1.8249655961990356, + "learning_rate": 9.052657667500951e-06, + "loss": 1.0005391836166382, + "step": 474 + }, + { + "epoch": 1.0063424947145878, + "grad_norm": 0.7612274885177612, + "learning_rate": 9.042978689543487e-06, + "loss": 1.0985958576202393, + "step": 476 + }, + { + "epoch": 1.0105708245243128, + "grad_norm": 2.120131492614746, + "learning_rate": 9.033256397248124e-06, + "loss": 0.4463256001472473, + "step": 478 + }, + { + "epoch": 1.014799154334038, + "grad_norm": 2.451970338821411, + "learning_rate": 9.02349090947379e-06, + "loss": 0.6860926151275635, + "step": 480 + }, + { + "epoch": 1.0190274841437632, + "grad_norm": 2.22886061668396, + "learning_rate": 9.013682345607502e-06, + "loss": 0.7149063944816589, + "step": 482 + }, + { + "epoch": 1.0232558139534884, + "grad_norm": 1.6330233812332153, + "learning_rate": 9.003830825562887e-06, + "loss": 0.9997832179069519, + "step": 484 + }, + { + "epoch": 1.0274841437632136, + "grad_norm": 3.6653683185577393, + "learning_rate": 8.99393646977874e-06, + "loss": 0.9798961877822876, + "step": 486 + }, + { + "epoch": 1.0317124735729386, + "grad_norm": 1.2772337198257446, + "learning_rate": 8.983999399217536e-06, + "loss": 0.8699902892112732, + "step": 488 + }, + { + "epoch": 1.0359408033826638, + "grad_norm": 1.8453668355941772, + "learning_rate": 8.974019735363952e-06, + "loss": 0.8166494369506836, + "step": 490 + }, + { + "epoch": 1.040169133192389, + "grad_norm": 5.692193984985352, + "learning_rate": 8.963997600223387e-06, + "loss": 0.8104552626609802, + "step": 492 + }, + { + "epoch": 1.0443974630021142, + "grad_norm": 1.1926912069320679, + "learning_rate": 8.953933116320474e-06, + "loss": 1.1175689697265625, + "step": 494 + }, + { + "epoch": 1.0486257928118394, + "grad_norm": 2.828871011734009, + "learning_rate": 8.943826406697569e-06, + "loss": 0.9687854051589966, + "step": 496 + }, + { + "epoch": 1.0528541226215644, + "grad_norm": 1.3092215061187744, + "learning_rate": 8.93367759491326e-06, + "loss": 0.7396000623703003, + "step": 498 + }, + { + "epoch": 1.0570824524312896, + "grad_norm": 3.2904458045959473, + "learning_rate": 8.923486805040845e-06, + "loss": 0.8625678420066833, + "step": 500 + }, + { + "epoch": 1.0613107822410148, + "grad_norm": 1.6755874156951904, + "learning_rate": 8.913254161666828e-06, + "loss": 1.1327816247940063, + "step": 502 + }, + { + "epoch": 1.06553911205074, + "grad_norm": 1.0441874265670776, + "learning_rate": 8.902979789889382e-06, + "loss": 0.998689591884613, + "step": 504 + }, + { + "epoch": 1.069767441860465, + "grad_norm": 2.697725772857666, + "learning_rate": 8.892663815316833e-06, + "loss": 0.8176366686820984, + "step": 506 + }, + { + "epoch": 1.0739957716701902, + "grad_norm": 0.9283720254898071, + "learning_rate": 8.882306364066114e-06, + "loss": 0.8020952939987183, + "step": 508 + }, + { + "epoch": 1.0782241014799154, + "grad_norm": 1.7760288715362549, + "learning_rate": 8.871907562761226e-06, + "loss": 1.022519826889038, + "step": 510 + }, + { + "epoch": 1.0824524312896406, + "grad_norm": 1.2100770473480225, + "learning_rate": 8.861467538531692e-06, + "loss": 0.9932355880737305, + "step": 512 + }, + { + "epoch": 1.0866807610993658, + "grad_norm": 2.784210205078125, + "learning_rate": 8.850986419011006e-06, + "loss": 1.080140233039856, + "step": 514 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 1.1554030179977417, + "learning_rate": 8.840464332335062e-06, + "loss": 0.8839500546455383, + "step": 516 + }, + { + "epoch": 1.095137420718816, + "grad_norm": 1.2442573308944702, + "learning_rate": 8.829901407140599e-06, + "loss": 0.8389308452606201, + "step": 518 + }, + { + "epoch": 1.0993657505285412, + "grad_norm": 1.8293520212173462, + "learning_rate": 8.81929777256362e-06, + "loss": 0.646580696105957, + "step": 520 + }, + { + "epoch": 1.1035940803382664, + "grad_norm": 1.5187311172485352, + "learning_rate": 8.808653558237818e-06, + "loss": 1.3373504877090454, + "step": 522 + }, + { + "epoch": 1.1078224101479917, + "grad_norm": 10.04066276550293, + "learning_rate": 8.797968894292992e-06, + "loss": 0.7550784945487976, + "step": 524 + }, + { + "epoch": 1.1120507399577166, + "grad_norm": 1.687009572982788, + "learning_rate": 8.787243911353447e-06, + "loss": 1.267970085144043, + "step": 526 + }, + { + "epoch": 1.1162790697674418, + "grad_norm": 1.0987988710403442, + "learning_rate": 8.776478740536409e-06, + "loss": 0.8014870882034302, + "step": 528 + }, + { + "epoch": 1.120507399577167, + "grad_norm": 2.7739126682281494, + "learning_rate": 8.76567351345042e-06, + "loss": 0.46113139390945435, + "step": 530 + }, + { + "epoch": 1.1247357293868923, + "grad_norm": 0.998328685760498, + "learning_rate": 8.754828362193719e-06, + "loss": 1.03612220287323, + "step": 532 + }, + { + "epoch": 1.1289640591966172, + "grad_norm": 0.9924682974815369, + "learning_rate": 8.743943419352638e-06, + "loss": 0.8442054986953735, + "step": 534 + }, + { + "epoch": 1.1331923890063424, + "grad_norm": 0.8140501379966736, + "learning_rate": 8.733018817999977e-06, + "loss": 1.0020159482955933, + "step": 536 + }, + { + "epoch": 1.1374207188160677, + "grad_norm": 28.491947174072266, + "learning_rate": 8.722054691693374e-06, + "loss": 0.8022699952125549, + "step": 538 + }, + { + "epoch": 1.1416490486257929, + "grad_norm": 3.403494358062744, + "learning_rate": 8.711051174473682e-06, + "loss": 0.7669907808303833, + "step": 540 + }, + { + "epoch": 1.145877378435518, + "grad_norm": 0.92375648021698, + "learning_rate": 8.700008400863316e-06, + "loss": 1.0623034238815308, + "step": 542 + }, + { + "epoch": 1.150105708245243, + "grad_norm": 1.1923998594284058, + "learning_rate": 8.688926505864624e-06, + "loss": 0.7859268188476562, + "step": 544 + }, + { + "epoch": 1.1543340380549683, + "grad_norm": 3.2635719776153564, + "learning_rate": 8.67780562495822e-06, + "loss": 1.1548057794570923, + "step": 546 + }, + { + "epoch": 1.1585623678646935, + "grad_norm": 2.591111898422241, + "learning_rate": 8.666645894101346e-06, + "loss": 1.2081559896469116, + "step": 548 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 1.3697686195373535, + "learning_rate": 8.655447449726189e-06, + "loss": 0.8476346731185913, + "step": 550 + }, + { + "epoch": 1.1670190274841437, + "grad_norm": 0.9091811776161194, + "learning_rate": 8.644210428738237e-06, + "loss": 1.058759093284607, + "step": 552 + }, + { + "epoch": 1.1712473572938689, + "grad_norm": 1.3113665580749512, + "learning_rate": 8.632934968514582e-06, + "loss": 1.1236515045166016, + "step": 554 + }, + { + "epoch": 1.175475687103594, + "grad_norm": 1.228116512298584, + "learning_rate": 8.621621206902256e-06, + "loss": 0.9948160648345947, + "step": 556 + }, + { + "epoch": 1.1797040169133193, + "grad_norm": 1.1882266998291016, + "learning_rate": 8.61026928221654e-06, + "loss": 1.0244168043136597, + "step": 558 + }, + { + "epoch": 1.1839323467230445, + "grad_norm": 0.6281694173812866, + "learning_rate": 8.598879333239274e-06, + "loss": 0.9882432222366333, + "step": 560 + }, + { + "epoch": 1.1881606765327695, + "grad_norm": 0.907728910446167, + "learning_rate": 8.587451499217161e-06, + "loss": 0.6845173239707947, + "step": 562 + }, + { + "epoch": 1.1923890063424947, + "grad_norm": 0.9854462742805481, + "learning_rate": 8.575985919860065e-06, + "loss": 1.0217852592468262, + "step": 564 + }, + { + "epoch": 1.1966173361522199, + "grad_norm": 2.510847806930542, + "learning_rate": 8.564482735339296e-06, + "loss": 0.8860641717910767, + "step": 566 + }, + { + "epoch": 1.200845665961945, + "grad_norm": 1.835896611213684, + "learning_rate": 8.552942086285909e-06, + "loss": 1.2221957445144653, + "step": 568 + }, + { + "epoch": 1.20507399577167, + "grad_norm": 1.8435322046279907, + "learning_rate": 8.541364113788974e-06, + "loss": 1.15140700340271, + "step": 570 + }, + { + "epoch": 1.2093023255813953, + "grad_norm": 1.7908110618591309, + "learning_rate": 8.529748959393855e-06, + "loss": 1.0449750423431396, + "step": 572 + }, + { + "epoch": 1.2135306553911205, + "grad_norm": 1.0777274370193481, + "learning_rate": 8.518096765100485e-06, + "loss": 0.7578415274620056, + "step": 574 + }, + { + "epoch": 1.2177589852008457, + "grad_norm": 6.83182430267334, + "learning_rate": 8.50640767336162e-06, + "loss": 0.7553078532218933, + "step": 576 + }, + { + "epoch": 1.221987315010571, + "grad_norm": 1.0021545886993408, + "learning_rate": 8.494681827081096e-06, + "loss": 0.9235932230949402, + "step": 578 + }, + { + "epoch": 1.226215644820296, + "grad_norm": 3.1951746940612793, + "learning_rate": 8.482919369612103e-06, + "loss": 0.8074833154678345, + "step": 580 + }, + { + "epoch": 1.230443974630021, + "grad_norm": 2.2242960929870605, + "learning_rate": 8.471120444755405e-06, + "loss": 1.0809502601623535, + "step": 582 + }, + { + "epoch": 1.2346723044397463, + "grad_norm": 2.378070116043091, + "learning_rate": 8.4592851967576e-06, + "loss": 0.9454922080039978, + "step": 584 + }, + { + "epoch": 1.2389006342494715, + "grad_norm": 5.33587121963501, + "learning_rate": 8.447413770309349e-06, + "loss": 0.846709132194519, + "step": 586 + }, + { + "epoch": 1.2431289640591967, + "grad_norm": 1.2231918573379517, + "learning_rate": 8.435506310543607e-06, + "loss": 0.8623060584068298, + "step": 588 + }, + { + "epoch": 1.2473572938689217, + "grad_norm": 2.068950653076172, + "learning_rate": 8.42356296303386e-06, + "loss": 0.6422075033187866, + "step": 590 + }, + { + "epoch": 1.251585623678647, + "grad_norm": 0.6553975939750671, + "learning_rate": 8.411583873792325e-06, + "loss": 0.8084491491317749, + "step": 592 + }, + { + "epoch": 1.255813953488372, + "grad_norm": 2.4686522483825684, + "learning_rate": 8.399569189268182e-06, + "loss": 1.0660851001739502, + "step": 594 + }, + { + "epoch": 1.2600422832980973, + "grad_norm": 0.973196804523468, + "learning_rate": 8.387519056345778e-06, + "loss": 0.710346519947052, + "step": 596 + }, + { + "epoch": 1.2642706131078225, + "grad_norm": 1.0025427341461182, + "learning_rate": 8.375433622342824e-06, + "loss": 1.09117591381073, + "step": 598 + }, + { + "epoch": 1.2684989429175475, + "grad_norm": 2.191934108734131, + "learning_rate": 8.363313035008613e-06, + "loss": 1.3207026720046997, + "step": 600 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 1.1381713151931763, + "learning_rate": 8.351157442522187e-06, + "loss": 0.9467790126800537, + "step": 602 + }, + { + "epoch": 1.276955602536998, + "grad_norm": 1.3978862762451172, + "learning_rate": 8.338966993490554e-06, + "loss": 1.0857586860656738, + "step": 604 + }, + { + "epoch": 1.2811839323467231, + "grad_norm": 1.6825940608978271, + "learning_rate": 8.326741836946852e-06, + "loss": 0.734017550945282, + "step": 606 + }, + { + "epoch": 1.285412262156448, + "grad_norm": 1.5627394914627075, + "learning_rate": 8.314482122348523e-06, + "loss": 1.0762019157409668, + "step": 608 + }, + { + "epoch": 1.2896405919661733, + "grad_norm": 0.9331506490707397, + "learning_rate": 8.30218799957551e-06, + "loss": 0.7343036532402039, + "step": 610 + }, + { + "epoch": 1.2938689217758985, + "grad_norm": 1.0899161100387573, + "learning_rate": 8.289859618928404e-06, + "loss": 0.6366298198699951, + "step": 612 + }, + { + "epoch": 1.2980972515856237, + "grad_norm": 1.8112653493881226, + "learning_rate": 8.277497131126608e-06, + "loss": 0.9008564352989197, + "step": 614 + }, + { + "epoch": 1.302325581395349, + "grad_norm": 1.308631420135498, + "learning_rate": 8.265100687306503e-06, + "loss": 0.5234721899032593, + "step": 616 + }, + { + "epoch": 1.306553911205074, + "grad_norm": 2.8963139057159424, + "learning_rate": 8.252670439019595e-06, + "loss": 0.8586516380310059, + "step": 618 + }, + { + "epoch": 1.3107822410147991, + "grad_norm": 10.228710174560547, + "learning_rate": 8.240206538230663e-06, + "loss": 1.1341429948806763, + "step": 620 + }, + { + "epoch": 1.3150105708245243, + "grad_norm": 0.6819997429847717, + "learning_rate": 8.227709137315905e-06, + "loss": 1.0599015951156616, + "step": 622 + }, + { + "epoch": 1.3192389006342495, + "grad_norm": 1.8252311944961548, + "learning_rate": 8.215178389061064e-06, + "loss": 0.6584379076957703, + "step": 624 + }, + { + "epoch": 1.3234672304439745, + "grad_norm": 1.418512225151062, + "learning_rate": 8.202614446659572e-06, + "loss": 1.120978593826294, + "step": 626 + }, + { + "epoch": 1.3276955602536997, + "grad_norm": 2.267376184463501, + "learning_rate": 8.190017463710675e-06, + "loss": 1.1850979328155518, + "step": 628 + }, + { + "epoch": 1.331923890063425, + "grad_norm": 1.201109528541565, + "learning_rate": 8.177387594217548e-06, + "loss": 0.9416132569313049, + "step": 630 + }, + { + "epoch": 1.3361522198731501, + "grad_norm": 1.1283001899719238, + "learning_rate": 8.164724992585421e-06, + "loss": 0.6001866459846497, + "step": 632 + }, + { + "epoch": 1.3403805496828753, + "grad_norm": 1.0230671167373657, + "learning_rate": 8.152029813619685e-06, + "loss": 0.7112642526626587, + "step": 634 + }, + { + "epoch": 1.3446088794926006, + "grad_norm": 0.9196541905403137, + "learning_rate": 8.139302212524006e-06, + "loss": 0.7454161047935486, + "step": 636 + }, + { + "epoch": 1.3488372093023255, + "grad_norm": 1.743666172027588, + "learning_rate": 8.126542344898419e-06, + "loss": 1.0068058967590332, + "step": 638 + }, + { + "epoch": 1.3530655391120507, + "grad_norm": 0.8174201250076294, + "learning_rate": 8.113750366737431e-06, + "loss": 0.6959707736968994, + "step": 640 + }, + { + "epoch": 1.357293868921776, + "grad_norm": 1.4081785678863525, + "learning_rate": 8.100926434428116e-06, + "loss": 1.1708544492721558, + "step": 642 + }, + { + "epoch": 1.361522198731501, + "grad_norm": 2.978811025619507, + "learning_rate": 8.0880707047482e-06, + "loss": 1.1028854846954346, + "step": 644 + }, + { + "epoch": 1.3657505285412261, + "grad_norm": 6.359747886657715, + "learning_rate": 8.075183334864139e-06, + "loss": 0.7414219975471497, + "step": 646 + }, + { + "epoch": 1.3699788583509513, + "grad_norm": 3.093681573867798, + "learning_rate": 8.062264482329209e-06, + "loss": 0.6780709624290466, + "step": 648 + }, + { + "epoch": 1.3742071881606766, + "grad_norm": 1.2487008571624756, + "learning_rate": 8.049314305081575e-06, + "loss": 0.8003671765327454, + "step": 650 + }, + { + "epoch": 1.3784355179704018, + "grad_norm": 10.850486755371094, + "learning_rate": 8.036332961442355e-06, + "loss": 0.7324129939079285, + "step": 652 + }, + { + "epoch": 1.382663847780127, + "grad_norm": 1.330004096031189, + "learning_rate": 8.023320610113691e-06, + "loss": 1.13347327709198, + "step": 654 + }, + { + "epoch": 1.386892177589852, + "grad_norm": 1.5921283960342407, + "learning_rate": 8.010277410176801e-06, + "loss": 1.1417347192764282, + "step": 656 + }, + { + "epoch": 1.3911205073995772, + "grad_norm": 1.6330335140228271, + "learning_rate": 7.997203521090053e-06, + "loss": 0.8730330467224121, + "step": 658 + }, + { + "epoch": 1.3953488372093024, + "grad_norm": 3.4455583095550537, + "learning_rate": 7.984099102686986e-06, + "loss": 0.9122240543365479, + "step": 660 + }, + { + "epoch": 1.3995771670190273, + "grad_norm": 0.7939943671226501, + "learning_rate": 7.970964315174384e-06, + "loss": 1.0506963729858398, + "step": 662 + }, + { + "epoch": 1.4038054968287526, + "grad_norm": 0.2450345903635025, + "learning_rate": 7.9577993191303e-06, + "loss": 0.8770999312400818, + "step": 664 + }, + { + "epoch": 1.4080338266384778, + "grad_norm": 2.8335092067718506, + "learning_rate": 7.944604275502102e-06, + "loss": 1.263124942779541, + "step": 666 + }, + { + "epoch": 1.412262156448203, + "grad_norm": 0.9448780417442322, + "learning_rate": 7.931379345604495e-06, + "loss": 1.0469601154327393, + "step": 668 + }, + { + "epoch": 1.4164904862579282, + "grad_norm": 0.852315366268158, + "learning_rate": 7.918124691117563e-06, + "loss": 0.6510782241821289, + "step": 670 + }, + { + "epoch": 1.4207188160676534, + "grad_norm": 11.286778450012207, + "learning_rate": 7.904840474084783e-06, + "loss": 1.0516870021820068, + "step": 672 + }, + { + "epoch": 1.4249471458773784, + "grad_norm": 0.9891210198402405, + "learning_rate": 7.891526856911046e-06, + "loss": 0.5895065069198608, + "step": 674 + }, + { + "epoch": 1.4291754756871036, + "grad_norm": 2.2937374114990234, + "learning_rate": 7.878184002360664e-06, + "loss": 0.7422171831130981, + "step": 676 + }, + { + "epoch": 1.4334038054968288, + "grad_norm": 1.4257853031158447, + "learning_rate": 7.864812073555402e-06, + "loss": 1.2657513618469238, + "step": 678 + }, + { + "epoch": 1.437632135306554, + "grad_norm": 1.4582903385162354, + "learning_rate": 7.851411233972461e-06, + "loss": 1.1611875295639038, + "step": 680 + }, + { + "epoch": 1.441860465116279, + "grad_norm": 2.666832685470581, + "learning_rate": 7.837981647442486e-06, + "loss": 1.106019139289856, + "step": 682 + }, + { + "epoch": 1.4460887949260042, + "grad_norm": 0.8176199793815613, + "learning_rate": 7.82452347814757e-06, + "loss": 1.0742647647857666, + "step": 684 + }, + { + "epoch": 1.4503171247357294, + "grad_norm": 1.7066468000411987, + "learning_rate": 7.811036890619239e-06, + "loss": 1.4130909442901611, + "step": 686 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 0.9444884061813354, + "learning_rate": 7.797522049736441e-06, + "loss": 1.1267884969711304, + "step": 688 + }, + { + "epoch": 1.4587737843551798, + "grad_norm": 1.9853451251983643, + "learning_rate": 7.78397912072354e-06, + "loss": 0.48357418179512024, + "step": 690 + }, + { + "epoch": 1.463002114164905, + "grad_norm": 1.1945173740386963, + "learning_rate": 7.770408269148281e-06, + "loss": 1.0998144149780273, + "step": 692 + }, + { + "epoch": 1.46723044397463, + "grad_norm": 1.174439787864685, + "learning_rate": 7.756809660919778e-06, + "loss": 0.5874847173690796, + "step": 694 + }, + { + "epoch": 1.4714587737843552, + "grad_norm": 2.8170855045318604, + "learning_rate": 7.74318346228648e-06, + "loss": 1.0440444946289062, + "step": 696 + }, + { + "epoch": 1.4756871035940804, + "grad_norm": 3.926302194595337, + "learning_rate": 7.72952983983414e-06, + "loss": 1.1133286952972412, + "step": 698 + }, + { + "epoch": 1.4799154334038054, + "grad_norm": 2.762296199798584, + "learning_rate": 7.715848960483776e-06, + "loss": 0.9230332374572754, + "step": 700 + }, + { + "epoch": 1.4841437632135306, + "grad_norm": 0.7554068565368652, + "learning_rate": 7.702140991489637e-06, + "loss": 1.0790983438491821, + "step": 702 + }, + { + "epoch": 1.4883720930232558, + "grad_norm": 2.199388265609741, + "learning_rate": 7.688406100437149e-06, + "loss": 1.008788824081421, + "step": 704 + }, + { + "epoch": 1.492600422832981, + "grad_norm": 0.8400372266769409, + "learning_rate": 7.674644455240878e-06, + "loss": 0.7245333194732666, + "step": 706 + }, + { + "epoch": 1.4968287526427062, + "grad_norm": 0.5919277667999268, + "learning_rate": 7.660856224142456e-06, + "loss": 0.2714691162109375, + "step": 708 + }, + { + "epoch": 1.5010570824524314, + "grad_norm": 2.098564386367798, + "learning_rate": 7.647041575708555e-06, + "loss": 0.6564630270004272, + "step": 710 + }, + { + "epoch": 1.5052854122621564, + "grad_norm": 2.421992301940918, + "learning_rate": 7.633200678828804e-06, + "loss": 0.5243597626686096, + "step": 712 + }, + { + "epoch": 1.5095137420718816, + "grad_norm": 1.5564970970153809, + "learning_rate": 7.6193337027137215e-06, + "loss": 0.8967912793159485, + "step": 714 + }, + { + "epoch": 1.5137420718816068, + "grad_norm": 0.9343472123146057, + "learning_rate": 7.605440816892667e-06, + "loss": 1.0377099514007568, + "step": 716 + }, + { + "epoch": 1.5179704016913318, + "grad_norm": 1.1493507623672485, + "learning_rate": 7.591522191211746e-06, + "loss": 0.9455903172492981, + "step": 718 + }, + { + "epoch": 1.522198731501057, + "grad_norm": 0.3293185234069824, + "learning_rate": 7.577577995831755e-06, + "loss": 0.8772664070129395, + "step": 720 + }, + { + "epoch": 1.5264270613107822, + "grad_norm": 1.9101412296295166, + "learning_rate": 7.563608401226081e-06, + "loss": 0.8627620339393616, + "step": 722 + }, + { + "epoch": 1.5306553911205074, + "grad_norm": 1.1184234619140625, + "learning_rate": 7.549613578178632e-06, + "loss": 0.901220977306366, + "step": 724 + }, + { + "epoch": 1.5348837209302326, + "grad_norm": 1.1049782037734985, + "learning_rate": 7.535593697781741e-06, + "loss": 0.8831064701080322, + "step": 726 + }, + { + "epoch": 1.5391120507399578, + "grad_norm": 1.2415213584899902, + "learning_rate": 7.52154893143408e-06, + "loss": 0.6195402145385742, + "step": 728 + }, + { + "epoch": 1.543340380549683, + "grad_norm": 6.555059432983398, + "learning_rate": 7.50747945083856e-06, + "loss": 0.8089724779129028, + "step": 730 + }, + { + "epoch": 1.547568710359408, + "grad_norm": 0.7224189043045044, + "learning_rate": 7.493385428000234e-06, + "loss": 1.0039284229278564, + "step": 732 + }, + { + "epoch": 1.5517970401691332, + "grad_norm": 0.48552051186561584, + "learning_rate": 7.479267035224191e-06, + "loss": 0.2228410691022873, + "step": 734 + }, + { + "epoch": 1.5560253699788582, + "grad_norm": 1.266108751296997, + "learning_rate": 7.465124445113459e-06, + "loss": 0.9717158675193787, + "step": 736 + }, + { + "epoch": 1.5602536997885834, + "grad_norm": 0.41860461235046387, + "learning_rate": 7.450957830566878e-06, + "loss": 0.7111193537712097, + "step": 738 + }, + { + "epoch": 1.5644820295983086, + "grad_norm": 0.7583433985710144, + "learning_rate": 7.436767364777009e-06, + "loss": 1.0799121856689453, + "step": 740 + }, + { + "epoch": 1.5687103594080338, + "grad_norm": 2.399691581726074, + "learning_rate": 7.422553221227993e-06, + "loss": 0.7972431778907776, + "step": 742 + }, + { + "epoch": 1.572938689217759, + "grad_norm": 1.444593071937561, + "learning_rate": 7.4083155736934484e-06, + "loss": 1.103126883506775, + "step": 744 + }, + { + "epoch": 1.5771670190274842, + "grad_norm": 1.4233990907669067, + "learning_rate": 7.394054596234336e-06, + "loss": 1.1827806234359741, + "step": 746 + }, + { + "epoch": 1.5813953488372094, + "grad_norm": 0.9901655316352844, + "learning_rate": 7.379770463196831e-06, + "loss": 0.8610133528709412, + "step": 748 + }, + { + "epoch": 1.5856236786469344, + "grad_norm": 1.1883097887039185, + "learning_rate": 7.365463349210203e-06, + "loss": 0.051627811044454575, + "step": 750 + }, + { + "epoch": 1.5898520084566596, + "grad_norm": 35.91194534301758, + "learning_rate": 7.351133429184666e-06, + "loss": 0.8116692900657654, + "step": 752 + }, + { + "epoch": 1.5940803382663846, + "grad_norm": 3.005581855773926, + "learning_rate": 7.336780878309251e-06, + "loss": 1.147510051727295, + "step": 754 + }, + { + "epoch": 1.5983086680761098, + "grad_norm": 15.526881217956543, + "learning_rate": 7.322405872049657e-06, + "loss": 1.0913742780685425, + "step": 756 + }, + { + "epoch": 1.602536997885835, + "grad_norm": 0.9753276705741882, + "learning_rate": 7.308008586146111e-06, + "loss": 1.1604857444763184, + "step": 758 + }, + { + "epoch": 1.6067653276955602, + "grad_norm": 0.9733684659004211, + "learning_rate": 7.293589196611221e-06, + "loss": 1.0768887996673584, + "step": 760 + }, + { + "epoch": 1.6109936575052854, + "grad_norm": 0.829832136631012, + "learning_rate": 7.279147879727811e-06, + "loss": 1.0790703296661377, + "step": 762 + }, + { + "epoch": 1.6152219873150107, + "grad_norm": 1.547074556350708, + "learning_rate": 7.264684812046787e-06, + "loss": 0.9501691460609436, + "step": 764 + }, + { + "epoch": 1.6194503171247359, + "grad_norm": 2.463318347930908, + "learning_rate": 7.250200170384961e-06, + "loss": 0.7793781161308289, + "step": 766 + }, + { + "epoch": 1.6236786469344608, + "grad_norm": 0.9056198000907898, + "learning_rate": 7.235694131822894e-06, + "loss": 1.0757365226745605, + "step": 768 + }, + { + "epoch": 1.627906976744186, + "grad_norm": 1.8374583721160889, + "learning_rate": 7.221166873702738e-06, + "loss": 1.0771503448486328, + "step": 770 + }, + { + "epoch": 1.6321353065539113, + "grad_norm": 1.0796993970870972, + "learning_rate": 7.206618573626058e-06, + "loss": 1.0819135904312134, + "step": 772 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 0.8046290278434753, + "learning_rate": 7.19204940945167e-06, + "loss": 0.8486723303794861, + "step": 774 + }, + { + "epoch": 1.6405919661733614, + "grad_norm": 1.069205641746521, + "learning_rate": 7.177459559293456e-06, + "loss": 1.0997098684310913, + "step": 776 + }, + { + "epoch": 1.6448202959830867, + "grad_norm": 2.0412747859954834, + "learning_rate": 7.162849201518198e-06, + "loss": 0.6558449864387512, + "step": 778 + }, + { + "epoch": 1.6490486257928119, + "grad_norm": 0.8818774819374084, + "learning_rate": 7.1482185147433905e-06, + "loss": 1.0440751314163208, + "step": 780 + }, + { + "epoch": 1.653276955602537, + "grad_norm": 1.0151294469833374, + "learning_rate": 7.133567677835055e-06, + "loss": 0.7639485001564026, + "step": 782 + }, + { + "epoch": 1.6575052854122623, + "grad_norm": 1.3014827966690063, + "learning_rate": 7.11889686990556e-06, + "loss": 1.1806261539459229, + "step": 784 + }, + { + "epoch": 1.6617336152219875, + "grad_norm": 4.210557460784912, + "learning_rate": 7.104206270311423e-06, + "loss": 0.47801291942596436, + "step": 786 + }, + { + "epoch": 1.6659619450317125, + "grad_norm": 0.9024162888526917, + "learning_rate": 7.089496058651133e-06, + "loss": 1.0849629640579224, + "step": 788 + }, + { + "epoch": 1.6701902748414377, + "grad_norm": 0.8603456616401672, + "learning_rate": 7.0747664147629335e-06, + "loss": 0.7134051322937012, + "step": 790 + }, + { + "epoch": 1.6744186046511627, + "grad_norm": 0.6777817606925964, + "learning_rate": 7.060017518722637e-06, + "loss": 0.4346533715724945, + "step": 792 + }, + { + "epoch": 1.6786469344608879, + "grad_norm": 1.0490171909332275, + "learning_rate": 7.0452495508414276e-06, + "loss": 1.048262357711792, + "step": 794 + }, + { + "epoch": 1.682875264270613, + "grad_norm": 0.5446838736534119, + "learning_rate": 7.030462691663641e-06, + "loss": 0.5887590646743774, + "step": 796 + }, + { + "epoch": 1.6871035940803383, + "grad_norm": 1.2877373695373535, + "learning_rate": 7.015657121964575e-06, + "loss": 1.178261637687683, + "step": 798 + }, + { + "epoch": 1.6913319238900635, + "grad_norm": 1.3940520286560059, + "learning_rate": 7.000833022748267e-06, + "loss": 1.04930579662323, + "step": 800 + }, + { + "epoch": 1.6955602536997887, + "grad_norm": 6.209669589996338, + "learning_rate": 6.985990575245283e-06, + "loss": 0.5518321990966797, + "step": 802 + }, + { + "epoch": 1.699788583509514, + "grad_norm": 0.7903432846069336, + "learning_rate": 6.97112996091051e-06, + "loss": 1.0536116361618042, + "step": 804 + }, + { + "epoch": 1.7040169133192389, + "grad_norm": 38.261756896972656, + "learning_rate": 6.956251361420926e-06, + "loss": 0.8339720964431763, + "step": 806 + }, + { + "epoch": 1.708245243128964, + "grad_norm": 26.985219955444336, + "learning_rate": 6.94135495867339e-06, + "loss": 0.7612593173980713, + "step": 808 + }, + { + "epoch": 1.712473572938689, + "grad_norm": 1.8201414346694946, + "learning_rate": 6.926440934782407e-06, + "loss": 1.081141471862793, + "step": 810 + }, + { + "epoch": 1.7167019027484143, + "grad_norm": 1.9191442728042603, + "learning_rate": 6.9115094720779155e-06, + "loss": 1.2877556085586548, + "step": 812 + }, + { + "epoch": 1.7209302325581395, + "grad_norm": 0.9197869896888733, + "learning_rate": 6.896560753103041e-06, + "loss": 1.0282819271087646, + "step": 814 + }, + { + "epoch": 1.7251585623678647, + "grad_norm": 3.420444965362549, + "learning_rate": 6.881594960611881e-06, + "loss": 1.1128859519958496, + "step": 816 + }, + { + "epoch": 1.72938689217759, + "grad_norm": 0.6601259112358093, + "learning_rate": 6.866612277567261e-06, + "loss": 0.8981441259384155, + "step": 818 + }, + { + "epoch": 1.733615221987315, + "grad_norm": 9.56154727935791, + "learning_rate": 6.851612887138501e-06, + "loss": 0.9813768267631531, + "step": 820 + }, + { + "epoch": 1.7378435517970403, + "grad_norm": 1.4616731405258179, + "learning_rate": 6.836596972699175e-06, + "loss": 0.6598048210144043, + "step": 822 + }, + { + "epoch": 1.7420718816067653, + "grad_norm": 0.5075655579566956, + "learning_rate": 6.82156471782487e-06, + "loss": 0.6283214092254639, + "step": 824 + }, + { + "epoch": 1.7463002114164905, + "grad_norm": 2.205639123916626, + "learning_rate": 6.806516306290942e-06, + "loss": 0.8650239109992981, + "step": 826 + }, + { + "epoch": 1.7505285412262155, + "grad_norm": 0.7252737283706665, + "learning_rate": 6.791451922070265e-06, + "loss": 0.8854203224182129, + "step": 828 + }, + { + "epoch": 1.7547568710359407, + "grad_norm": 1.720704197883606, + "learning_rate": 6.77637174933099e-06, + "loss": 0.7364405393600464, + "step": 830 + }, + { + "epoch": 1.758985200845666, + "grad_norm": 1.0492388010025024, + "learning_rate": 6.7612759724342895e-06, + "loss": 1.0686368942260742, + "step": 832 + }, + { + "epoch": 1.763213530655391, + "grad_norm": 5.734156608581543, + "learning_rate": 6.746164775932096e-06, + "loss": 0.7557425498962402, + "step": 834 + }, + { + "epoch": 1.7674418604651163, + "grad_norm": 0.9327018857002258, + "learning_rate": 6.731038344564862e-06, + "loss": 0.7879675626754761, + "step": 836 + }, + { + "epoch": 1.7716701902748415, + "grad_norm": 1.0288727283477783, + "learning_rate": 6.715896863259286e-06, + "loss": 1.0945039987564087, + "step": 838 + }, + { + "epoch": 1.7758985200845667, + "grad_norm": 1.405065655708313, + "learning_rate": 6.700740517126061e-06, + "loss": 0.8011113405227661, + "step": 840 + }, + { + "epoch": 1.7801268498942917, + "grad_norm": 2.722510814666748, + "learning_rate": 6.685569491457608e-06, + "loss": 0.34896984696388245, + "step": 842 + }, + { + "epoch": 1.784355179704017, + "grad_norm": 2.036210060119629, + "learning_rate": 6.670383971725809e-06, + "loss": 0.7101260423660278, + "step": 844 + }, + { + "epoch": 1.7885835095137421, + "grad_norm": 1.5442942380905151, + "learning_rate": 6.655184143579746e-06, + "loss": 0.9554334282875061, + "step": 846 + }, + { + "epoch": 1.792811839323467, + "grad_norm": 1.249420166015625, + "learning_rate": 6.639970192843423e-06, + "loss": 0.7113332152366638, + "step": 848 + }, + { + "epoch": 1.7970401691331923, + "grad_norm": 5.022181510925293, + "learning_rate": 6.624742305513502e-06, + "loss": 0.6313583254814148, + "step": 850 + }, + { + "epoch": 1.8012684989429175, + "grad_norm": 1.2391078472137451, + "learning_rate": 6.609500667757022e-06, + "loss": 0.9306125640869141, + "step": 852 + }, + { + "epoch": 1.8054968287526427, + "grad_norm": 0.7820528149604797, + "learning_rate": 6.594245465909127e-06, + "loss": 0.7952921390533447, + "step": 854 + }, + { + "epoch": 1.809725158562368, + "grad_norm": 0.9432372450828552, + "learning_rate": 6.578976886470791e-06, + "loss": 1.0426123142242432, + "step": 856 + }, + { + "epoch": 1.8139534883720931, + "grad_norm": 0.8079444169998169, + "learning_rate": 6.56369511610653e-06, + "loss": 0.592127799987793, + "step": 858 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 0.7049525380134583, + "learning_rate": 6.548400341642128e-06, + "loss": 1.0183756351470947, + "step": 860 + }, + { + "epoch": 1.8224101479915433, + "grad_norm": 1.2878975868225098, + "learning_rate": 6.533092750062345e-06, + "loss": 1.027207374572754, + "step": 862 + }, + { + "epoch": 1.8266384778012685, + "grad_norm": 0.6170883178710938, + "learning_rate": 6.5177725285086416e-06, + "loss": 0.54091876745224, + "step": 864 + }, + { + "epoch": 1.8308668076109935, + "grad_norm": 0.8307227492332458, + "learning_rate": 6.5024398642768785e-06, + "loss": 0.7404611706733704, + "step": 866 + }, + { + "epoch": 1.8350951374207187, + "grad_norm": 0.6512665152549744, + "learning_rate": 6.487094944815036e-06, + "loss": 1.0693024396896362, + "step": 868 + }, + { + "epoch": 1.839323467230444, + "grad_norm": 0.9685786962509155, + "learning_rate": 6.47173795772092e-06, + "loss": 1.0882283449172974, + "step": 870 + }, + { + "epoch": 1.8435517970401691, + "grad_norm": 0.46740958094596863, + "learning_rate": 6.456369090739869e-06, + "loss": 0.7198358178138733, + "step": 872 + }, + { + "epoch": 1.8477801268498943, + "grad_norm": 0.6628546714782715, + "learning_rate": 6.440988531762454e-06, + "loss": 1.0333024263381958, + "step": 874 + }, + { + "epoch": 1.8520084566596196, + "grad_norm": 2.340125322341919, + "learning_rate": 6.42559646882219e-06, + "loss": 0.8775447607040405, + "step": 876 + }, + { + "epoch": 1.8562367864693448, + "grad_norm": 3.4817638397216797, + "learning_rate": 6.410193090093227e-06, + "loss": 0.30013445019721985, + "step": 878 + }, + { + "epoch": 1.8604651162790697, + "grad_norm": 2.009491205215454, + "learning_rate": 6.394778583888064e-06, + "loss": 0.7175704836845398, + "step": 880 + }, + { + "epoch": 1.864693446088795, + "grad_norm": 1.6280615329742432, + "learning_rate": 6.379353138655227e-06, + "loss": 1.1051079034805298, + "step": 882 + }, + { + "epoch": 1.86892177589852, + "grad_norm": 0.6645678281784058, + "learning_rate": 6.3639169429769805e-06, + "loss": 0.7477265000343323, + "step": 884 + }, + { + "epoch": 1.8731501057082451, + "grad_norm": 1.2300724983215332, + "learning_rate": 6.348470185567025e-06, + "loss": 1.0531162023544312, + "step": 886 + }, + { + "epoch": 1.8773784355179703, + "grad_norm": 0.8753910064697266, + "learning_rate": 6.333013055268169e-06, + "loss": 1.0897793769836426, + "step": 888 + }, + { + "epoch": 1.8816067653276956, + "grad_norm": 1.3083759546279907, + "learning_rate": 6.317545741050044e-06, + "loss": 0.5681002140045166, + "step": 890 + }, + { + "epoch": 1.8858350951374208, + "grad_norm": 2.6303460597991943, + "learning_rate": 6.3020684320067805e-06, + "loss": 0.799943208694458, + "step": 892 + }, + { + "epoch": 1.890063424947146, + "grad_norm": 2.919250011444092, + "learning_rate": 6.2865813173546976e-06, + "loss": 1.0557307004928589, + "step": 894 + }, + { + "epoch": 1.8942917547568712, + "grad_norm": 3.601170778274536, + "learning_rate": 6.271084586429994e-06, + "loss": 1.3714882135391235, + "step": 896 + }, + { + "epoch": 1.8985200845665962, + "grad_norm": 0.9073897004127502, + "learning_rate": 6.255578428686433e-06, + "loss": 1.0622445344924927, + "step": 898 + }, + { + "epoch": 1.9027484143763214, + "grad_norm": 1.07796049118042, + "learning_rate": 6.240063033693023e-06, + "loss": 0.4852241277694702, + "step": 900 + }, + { + "epoch": 1.9069767441860463, + "grad_norm": 0.7135300040245056, + "learning_rate": 6.2245385911316994e-06, + "loss": 1.0725690126419067, + "step": 902 + }, + { + "epoch": 1.9112050739957716, + "grad_norm": 2.251589059829712, + "learning_rate": 6.2090052907950096e-06, + "loss": 0.5318590998649597, + "step": 904 + }, + { + "epoch": 1.9154334038054968, + "grad_norm": 1.0892387628555298, + "learning_rate": 6.193463322583792e-06, + "loss": 0.8591321706771851, + "step": 906 + }, + { + "epoch": 1.919661733615222, + "grad_norm": 1.2523810863494873, + "learning_rate": 6.177912876504851e-06, + "loss": 1.1854162216186523, + "step": 908 + }, + { + "epoch": 1.9238900634249472, + "grad_norm": 1.0911489725112915, + "learning_rate": 6.16235414266864e-06, + "loss": 0.7868256568908691, + "step": 910 + }, + { + "epoch": 1.9281183932346724, + "grad_norm": 1.2987313270568848, + "learning_rate": 6.146787311286927e-06, + "loss": 1.1219470500946045, + "step": 912 + }, + { + "epoch": 1.9323467230443976, + "grad_norm": 1.3436657190322876, + "learning_rate": 6.131212572670483e-06, + "loss": 0.7778401374816895, + "step": 914 + }, + { + "epoch": 1.9365750528541226, + "grad_norm": 0.8392052054405212, + "learning_rate": 6.115630117226741e-06, + "loss": 1.0792145729064941, + "step": 916 + }, + { + "epoch": 1.9408033826638478, + "grad_norm": 0.6756947040557861, + "learning_rate": 6.1000401354574855e-06, + "loss": 0.6716108322143555, + "step": 918 + }, + { + "epoch": 1.945031712473573, + "grad_norm": 1.118888258934021, + "learning_rate": 6.084442817956502e-06, + "loss": 0.9379401206970215, + "step": 920 + }, + { + "epoch": 1.949260042283298, + "grad_norm": 1.1046290397644043, + "learning_rate": 6.0688383554072625e-06, + "loss": 0.94891756772995, + "step": 922 + }, + { + "epoch": 1.9534883720930232, + "grad_norm": 1.4434776306152344, + "learning_rate": 6.053226938580596e-06, + "loss": 0.963381290435791, + "step": 924 + }, + { + "epoch": 1.9577167019027484, + "grad_norm": 0.7047730684280396, + "learning_rate": 6.03760875833234e-06, + "loss": 0.871060848236084, + "step": 926 + }, + { + "epoch": 1.9619450317124736, + "grad_norm": 0.9204599261283875, + "learning_rate": 6.021984005601027e-06, + "loss": 1.1106597185134888, + "step": 928 + }, + { + "epoch": 1.9661733615221988, + "grad_norm": 1.0530664920806885, + "learning_rate": 6.006352871405534e-06, + "loss": 0.7726105451583862, + "step": 930 + }, + { + "epoch": 1.970401691331924, + "grad_norm": 1.1999796628952026, + "learning_rate": 5.990715546842755e-06, + "loss": 1.2282732725143433, + "step": 932 + }, + { + "epoch": 1.9746300211416492, + "grad_norm": 1.2779529094696045, + "learning_rate": 5.9750722230852694e-06, + "loss": 0.8220756649971008, + "step": 934 + }, + { + "epoch": 1.9788583509513742, + "grad_norm": 0.7741147875785828, + "learning_rate": 5.959423091378992e-06, + "loss": 0.6703850030899048, + "step": 936 + }, + { + "epoch": 1.9830866807610994, + "grad_norm": 1.5844155550003052, + "learning_rate": 5.943768343040845e-06, + "loss": 0.7882522940635681, + "step": 938 + }, + { + "epoch": 1.9873150105708244, + "grad_norm": 0.7715420722961426, + "learning_rate": 5.928108169456419e-06, + "loss": 1.0299077033996582, + "step": 940 + }, + { + "epoch": 1.9915433403805496, + "grad_norm": 0.7235667109489441, + "learning_rate": 5.912442762077624e-06, + "loss": 1.1129530668258667, + "step": 942 + }, + { + "epoch": 1.9957716701902748, + "grad_norm": 1.0659117698669434, + "learning_rate": 5.89677231242036e-06, + "loss": 1.0971673727035522, + "step": 944 + }, + { + "epoch": 2.0, + "grad_norm": 1.6687339544296265, + "learning_rate": 5.8810970120621714e-06, + "loss": 0.4226009249687195, + "step": 946 + }, + { + "epoch": 2.004228329809725, + "grad_norm": 3.520862579345703, + "learning_rate": 5.8654170526399014e-06, + "loss": 0.6749889254570007, + "step": 948 + }, + { + "epoch": 2.0084566596194504, + "grad_norm": 0.9890263080596924, + "learning_rate": 5.849732625847354e-06, + "loss": 0.9129278063774109, + "step": 950 + }, + { + "epoch": 2.0126849894291756, + "grad_norm": 1.1954376697540283, + "learning_rate": 5.834043923432947e-06, + "loss": 0.8943532705307007, + "step": 952 + }, + { + "epoch": 2.016913319238901, + "grad_norm": 1.6151528358459473, + "learning_rate": 5.818351137197373e-06, + "loss": 1.044058084487915, + "step": 954 + }, + { + "epoch": 2.0211416490486256, + "grad_norm": 2.398728132247925, + "learning_rate": 5.802654458991245e-06, + "loss": 1.0897283554077148, + "step": 956 + }, + { + "epoch": 2.025369978858351, + "grad_norm": 1.079189419746399, + "learning_rate": 5.786954080712763e-06, + "loss": 0.6574183106422424, + "step": 958 + }, + { + "epoch": 2.029598308668076, + "grad_norm": 0.8927987217903137, + "learning_rate": 5.7712501943053576e-06, + "loss": 0.8874375224113464, + "step": 960 + }, + { + "epoch": 2.033826638477801, + "grad_norm": 1.4453222751617432, + "learning_rate": 5.755542991755352e-06, + "loss": 0.9133772253990173, + "step": 962 + }, + { + "epoch": 2.0380549682875264, + "grad_norm": 1.265455961227417, + "learning_rate": 5.739832665089604e-06, + "loss": 0.7647688388824463, + "step": 964 + }, + { + "epoch": 2.0422832980972516, + "grad_norm": 0.928623378276825, + "learning_rate": 5.724119406373171e-06, + "loss": 0.6144131422042847, + "step": 966 + }, + { + "epoch": 2.046511627906977, + "grad_norm": 2.308833122253418, + "learning_rate": 5.708403407706954e-06, + "loss": 0.720012903213501, + "step": 968 + }, + { + "epoch": 2.050739957716702, + "grad_norm": 2.1895899772644043, + "learning_rate": 5.692684861225349e-06, + "loss": 0.9889361262321472, + "step": 970 + }, + { + "epoch": 2.0549682875264272, + "grad_norm": 1.9990642070770264, + "learning_rate": 5.676963959093904e-06, + "loss": 0.49514392018318176, + "step": 972 + }, + { + "epoch": 2.059196617336152, + "grad_norm": 2.567124366760254, + "learning_rate": 5.66124089350696e-06, + "loss": 0.9481661915779114, + "step": 974 + }, + { + "epoch": 2.063424947145877, + "grad_norm": 0.7161730527877808, + "learning_rate": 5.645515856685316e-06, + "loss": 0.8480053544044495, + "step": 976 + }, + { + "epoch": 2.0676532769556024, + "grad_norm": 1.70330810546875, + "learning_rate": 5.629789040873862e-06, + "loss": 0.9336463809013367, + "step": 978 + }, + { + "epoch": 2.0718816067653276, + "grad_norm": 1.2546781301498413, + "learning_rate": 5.614060638339238e-06, + "loss": 0.6825820803642273, + "step": 980 + }, + { + "epoch": 2.076109936575053, + "grad_norm": 1.5388237237930298, + "learning_rate": 5.598330841367487e-06, + "loss": 0.6645275950431824, + "step": 982 + }, + { + "epoch": 2.080338266384778, + "grad_norm": 1.2570374011993408, + "learning_rate": 5.582599842261693e-06, + "loss": 0.6670082807540894, + "step": 984 + }, + { + "epoch": 2.0845665961945032, + "grad_norm": 1.2153971195220947, + "learning_rate": 5.566867833339644e-06, + "loss": 0.8871865272521973, + "step": 986 + }, + { + "epoch": 2.0887949260042284, + "grad_norm": 1.158151626586914, + "learning_rate": 5.551135006931467e-06, + "loss": 1.0278443098068237, + "step": 988 + }, + { + "epoch": 2.0930232558139537, + "grad_norm": 1.0243680477142334, + "learning_rate": 5.535401555377287e-06, + "loss": 0.9156352281570435, + "step": 990 + }, + { + "epoch": 2.097251585623679, + "grad_norm": 1.636741280555725, + "learning_rate": 5.51966767102487e-06, + "loss": 0.5614994168281555, + "step": 992 + }, + { + "epoch": 2.1014799154334036, + "grad_norm": 2.685683012008667, + "learning_rate": 5.503933546227273e-06, + "loss": 0.6497789621353149, + "step": 994 + }, + { + "epoch": 2.105708245243129, + "grad_norm": 0.8283545970916748, + "learning_rate": 5.4881993733404916e-06, + "loss": 0.7263455390930176, + "step": 996 + }, + { + "epoch": 2.109936575052854, + "grad_norm": 0.9021750688552856, + "learning_rate": 5.472465344721113e-06, + "loss": 0.8096359372138977, + "step": 998 + }, + { + "epoch": 2.1141649048625792, + "grad_norm": 1.2764297723770142, + "learning_rate": 5.456731652723959e-06, + "loss": 0.8352269530296326, + "step": 1000 + }, + { + "epoch": 2.1183932346723044, + "grad_norm": 0.4778595268726349, + "learning_rate": 5.440998489699731e-06, + "loss": 0.34087488055229187, + "step": 1002 + }, + { + "epoch": 2.1226215644820297, + "grad_norm": 2.0514400005340576, + "learning_rate": 5.425266047992672e-06, + "loss": 0.6780738234519958, + "step": 1004 + }, + { + "epoch": 2.126849894291755, + "grad_norm": 1.1958345174789429, + "learning_rate": 5.4095345199382e-06, + "loss": 0.8946940898895264, + "step": 1006 + }, + { + "epoch": 2.13107822410148, + "grad_norm": 4.016479969024658, + "learning_rate": 5.393804097860568e-06, + "loss": 0.8184226155281067, + "step": 1008 + }, + { + "epoch": 2.1353065539112053, + "grad_norm": 0.7505455613136292, + "learning_rate": 5.3780749740705065e-06, + "loss": 0.5445267558097839, + "step": 1010 + }, + { + "epoch": 2.13953488372093, + "grad_norm": 2.3954951763153076, + "learning_rate": 5.362347340862869e-06, + "loss": 0.45180371403694153, + "step": 1012 + }, + { + "epoch": 2.1437632135306552, + "grad_norm": 1.7521181106567383, + "learning_rate": 5.3466213905142935e-06, + "loss": 0.56569504737854, + "step": 1014 + }, + { + "epoch": 2.1479915433403804, + "grad_norm": 1.0249477624893188, + "learning_rate": 5.3308973152808395e-06, + "loss": 0.7223609685897827, + "step": 1016 + }, + { + "epoch": 2.1522198731501057, + "grad_norm": 1.0132323503494263, + "learning_rate": 5.315175307395644e-06, + "loss": 0.8352236151695251, + "step": 1018 + }, + { + "epoch": 2.156448202959831, + "grad_norm": 4.833903789520264, + "learning_rate": 5.299455559066571e-06, + "loss": 0.6990109086036682, + "step": 1020 + }, + { + "epoch": 2.160676532769556, + "grad_norm": 1.2555663585662842, + "learning_rate": 5.283738262473859e-06, + "loss": 0.6474488377571106, + "step": 1022 + }, + { + "epoch": 2.1649048625792813, + "grad_norm": 1.4123871326446533, + "learning_rate": 5.268023609767772e-06, + "loss": 1.0053154230117798, + "step": 1024 + }, + { + "epoch": 2.1691331923890065, + "grad_norm": 1.166477918624878, + "learning_rate": 5.2523117930662545e-06, + "loss": 0.9545771479606628, + "step": 1026 + }, + { + "epoch": 2.1733615221987317, + "grad_norm": 0.7529900670051575, + "learning_rate": 5.2366030044525785e-06, + "loss": 0.8685329556465149, + "step": 1028 + }, + { + "epoch": 2.177589852008457, + "grad_norm": 0.9365928173065186, + "learning_rate": 5.220897435972992e-06, + "loss": 0.8475508093833923, + "step": 1030 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 0.8649470806121826, + "learning_rate": 5.205195279634385e-06, + "loss": 0.8967899084091187, + "step": 1032 + }, + { + "epoch": 2.186046511627907, + "grad_norm": 6.396571636199951, + "learning_rate": 5.1894967274019245e-06, + "loss": 1.0467503070831299, + "step": 1034 + }, + { + "epoch": 2.190274841437632, + "grad_norm": 0.882828414440155, + "learning_rate": 5.173801971196716e-06, + "loss": 0.6571377515792847, + "step": 1036 + }, + { + "epoch": 2.1945031712473573, + "grad_norm": 4.263335704803467, + "learning_rate": 5.158111202893464e-06, + "loss": 0.3250623047351837, + "step": 1038 + }, + { + "epoch": 2.1987315010570825, + "grad_norm": 1.2508608102798462, + "learning_rate": 5.142424614318112e-06, + "loss": 0.9743232131004333, + "step": 1040 + }, + { + "epoch": 2.2029598308668077, + "grad_norm": 2.8953521251678467, + "learning_rate": 5.12674239724551e-06, + "loss": 0.32331445813179016, + "step": 1042 + }, + { + "epoch": 2.207188160676533, + "grad_norm": 1.6485546827316284, + "learning_rate": 5.111064743397061e-06, + "loss": 0.98199063539505, + "step": 1044 + }, + { + "epoch": 2.211416490486258, + "grad_norm": 0.9450780153274536, + "learning_rate": 5.09539184443838e-06, + "loss": 0.4515163004398346, + "step": 1046 + }, + { + "epoch": 2.2156448202959833, + "grad_norm": 2.1766693592071533, + "learning_rate": 5.079723891976956e-06, + "loss": 0.7173015475273132, + "step": 1048 + }, + { + "epoch": 2.219873150105708, + "grad_norm": 1.850635290145874, + "learning_rate": 5.064061077559803e-06, + "loss": 0.612816333770752, + "step": 1050 + }, + { + "epoch": 2.2241014799154333, + "grad_norm": 1.9140737056732178, + "learning_rate": 5.048403592671116e-06, + "loss": 0.7503929734230042, + "step": 1052 + }, + { + "epoch": 2.2283298097251585, + "grad_norm": 1.1067070960998535, + "learning_rate": 5.032751628729943e-06, + "loss": 0.9953061938285828, + "step": 1054 + }, + { + "epoch": 2.2325581395348837, + "grad_norm": 1.3405230045318604, + "learning_rate": 5.017105377087829e-06, + "loss": 0.932704746723175, + "step": 1056 + }, + { + "epoch": 2.236786469344609, + "grad_norm": 0.25487297773361206, + "learning_rate": 5.0014650290264854e-06, + "loss": 0.5900425314903259, + "step": 1058 + }, + { + "epoch": 2.241014799154334, + "grad_norm": 1.0683791637420654, + "learning_rate": 4.9858307757554536e-06, + "loss": 0.7317000031471252, + "step": 1060 + }, + { + "epoch": 2.2452431289640593, + "grad_norm": 0.8899886608123779, + "learning_rate": 4.97020280840976e-06, + "loss": 0.9242777824401855, + "step": 1062 + }, + { + "epoch": 2.2494714587737845, + "grad_norm": 1.7678951025009155, + "learning_rate": 4.954581318047582e-06, + "loss": 0.8662177324295044, + "step": 1064 + }, + { + "epoch": 2.2536997885835097, + "grad_norm": 2.116676092147827, + "learning_rate": 4.938966495647918e-06, + "loss": 0.4849522113800049, + "step": 1066 + }, + { + "epoch": 2.2579281183932345, + "grad_norm": 0.7686178684234619, + "learning_rate": 4.923358532108244e-06, + "loss": 0.7458196878433228, + "step": 1068 + }, + { + "epoch": 2.2621564482029597, + "grad_norm": 3.4403514862060547, + "learning_rate": 4.907757618242186e-06, + "loss": 0.7814114093780518, + "step": 1070 + }, + { + "epoch": 2.266384778012685, + "grad_norm": 1.1714199781417847, + "learning_rate": 4.892163944777181e-06, + "loss": 0.9176709055900574, + "step": 1072 + }, + { + "epoch": 2.27061310782241, + "grad_norm": 0.35950955748558044, + "learning_rate": 4.876577702352155e-06, + "loss": 0.616875171661377, + "step": 1074 + }, + { + "epoch": 2.2748414376321353, + "grad_norm": 3.0460031032562256, + "learning_rate": 4.860999081515183e-06, + "loss": 0.737525224685669, + "step": 1076 + }, + { + "epoch": 2.2790697674418605, + "grad_norm": 1.9951536655426025, + "learning_rate": 4.845428272721161e-06, + "loss": 0.2432975023984909, + "step": 1078 + }, + { + "epoch": 2.2832980972515857, + "grad_norm": 0.32823455333709717, + "learning_rate": 4.8298654663294865e-06, + "loss": 0.12714038789272308, + "step": 1080 + }, + { + "epoch": 2.287526427061311, + "grad_norm": 1.1101831197738647, + "learning_rate": 4.814310852601717e-06, + "loss": 0.5033318400382996, + "step": 1082 + }, + { + "epoch": 2.291754756871036, + "grad_norm": 0.9737103581428528, + "learning_rate": 4.7987646216992545e-06, + "loss": 0.33340218663215637, + "step": 1084 + }, + { + "epoch": 2.295983086680761, + "grad_norm": 2.3898680210113525, + "learning_rate": 4.783226963681018e-06, + "loss": 0.1616176962852478, + "step": 1086 + }, + { + "epoch": 2.300211416490486, + "grad_norm": 2.862590789794922, + "learning_rate": 4.76769806850112e-06, + "loss": 0.5990738272666931, + "step": 1088 + }, + { + "epoch": 2.3044397463002113, + "grad_norm": 1.0708791017532349, + "learning_rate": 4.752178126006544e-06, + "loss": 0.9059069156646729, + "step": 1090 + }, + { + "epoch": 2.3086680761099365, + "grad_norm": 1.654615044593811, + "learning_rate": 4.736667325934819e-06, + "loss": 0.5798441171646118, + "step": 1092 + }, + { + "epoch": 2.3128964059196617, + "grad_norm": 0.8037099242210388, + "learning_rate": 4.721165857911709e-06, + "loss": 0.8460623025894165, + "step": 1094 + }, + { + "epoch": 2.317124735729387, + "grad_norm": 1.493412733078003, + "learning_rate": 4.705673911448891e-06, + "loss": 0.9418038129806519, + "step": 1096 + }, + { + "epoch": 2.321353065539112, + "grad_norm": 1.2563211917877197, + "learning_rate": 4.690191675941631e-06, + "loss": 0.9124325513839722, + "step": 1098 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 2.238421678543091, + "learning_rate": 4.674719340666477e-06, + "loss": 0.4924452006816864, + "step": 1100 + }, + { + "epoch": 2.3298097251585626, + "grad_norm": 0.9232540726661682, + "learning_rate": 4.659257094778949e-06, + "loss": 0.6002163290977478, + "step": 1102 + }, + { + "epoch": 2.3340380549682873, + "grad_norm": 0.8689524531364441, + "learning_rate": 4.643805127311212e-06, + "loss": 0.8683549165725708, + "step": 1104 + }, + { + "epoch": 2.3382663847780125, + "grad_norm": 1.146340012550354, + "learning_rate": 4.628363627169774e-06, + "loss": 0.8270120620727539, + "step": 1106 + }, + { + "epoch": 2.3424947145877377, + "grad_norm": 0.9631582498550415, + "learning_rate": 4.612932783133182e-06, + "loss": 0.8559405207633972, + "step": 1108 + }, + { + "epoch": 2.346723044397463, + "grad_norm": 3.16713809967041, + "learning_rate": 4.5975127838497036e-06, + "loss": 1.0118556022644043, + "step": 1110 + }, + { + "epoch": 2.350951374207188, + "grad_norm": 0.3909473121166229, + "learning_rate": 4.582103817835021e-06, + "loss": 0.560530960559845, + "step": 1112 + }, + { + "epoch": 2.3551797040169133, + "grad_norm": 1.8440072536468506, + "learning_rate": 4.566706073469938e-06, + "loss": 0.8543758392333984, + "step": 1114 + }, + { + "epoch": 2.3594080338266386, + "grad_norm": 2.5913584232330322, + "learning_rate": 4.551319738998067e-06, + "loss": 0.6503856778144836, + "step": 1116 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 3.878394842147827, + "learning_rate": 4.535945002523524e-06, + "loss": 1.0811299085617065, + "step": 1118 + }, + { + "epoch": 2.367864693446089, + "grad_norm": 0.8683692216873169, + "learning_rate": 4.5205820520086415e-06, + "loss": 0.9283587336540222, + "step": 1120 + }, + { + "epoch": 2.3720930232558137, + "grad_norm": 0.9555020928382874, + "learning_rate": 4.505231075271662e-06, + "loss": 0.8046873807907104, + "step": 1122 + }, + { + "epoch": 2.376321353065539, + "grad_norm": 0.9591534733772278, + "learning_rate": 4.489892259984446e-06, + "loss": 0.5752630233764648, + "step": 1124 + }, + { + "epoch": 2.380549682875264, + "grad_norm": 5.281083106994629, + "learning_rate": 4.474565793670167e-06, + "loss": 0.9272839426994324, + "step": 1126 + }, + { + "epoch": 2.3847780126849893, + "grad_norm": 2.165510892868042, + "learning_rate": 4.459251863701038e-06, + "loss": 0.5945338010787964, + "step": 1128 + }, + { + "epoch": 2.3890063424947146, + "grad_norm": 2.092430591583252, + "learning_rate": 4.443950657296004e-06, + "loss": 0.7835938930511475, + "step": 1130 + }, + { + "epoch": 2.3932346723044398, + "grad_norm": 2.0051982402801514, + "learning_rate": 4.4286623615184585e-06, + "loss": 0.6009646654129028, + "step": 1132 + }, + { + "epoch": 2.397463002114165, + "grad_norm": 1.992108941078186, + "learning_rate": 4.413387163273964e-06, + "loss": 0.9296039938926697, + "step": 1134 + }, + { + "epoch": 2.40169133192389, + "grad_norm": 2.8401803970336914, + "learning_rate": 4.3981252493079506e-06, + "loss": 0.6294530034065247, + "step": 1136 + }, + { + "epoch": 2.4059196617336154, + "grad_norm": 1.4414680004119873, + "learning_rate": 4.382876806203453e-06, + "loss": 0.7423921823501587, + "step": 1138 + }, + { + "epoch": 2.41014799154334, + "grad_norm": 0.9113880395889282, + "learning_rate": 4.3676420203788125e-06, + "loss": 0.9104927182197571, + "step": 1140 + }, + { + "epoch": 2.4143763213530653, + "grad_norm": 1.2235362529754639, + "learning_rate": 4.352421078085405e-06, + "loss": 0.9704469442367554, + "step": 1142 + }, + { + "epoch": 2.4186046511627906, + "grad_norm": 2.3451757431030273, + "learning_rate": 4.337214165405368e-06, + "loss": 0.49127838015556335, + "step": 1144 + }, + { + "epoch": 2.4228329809725158, + "grad_norm": 3.8002395629882812, + "learning_rate": 4.322021468249316e-06, + "loss": 0.6594963669776917, + "step": 1146 + }, + { + "epoch": 2.427061310782241, + "grad_norm": 0.9705352783203125, + "learning_rate": 4.306843172354076e-06, + "loss": 0.6815480589866638, + "step": 1148 + }, + { + "epoch": 2.431289640591966, + "grad_norm": 1.9302246570587158, + "learning_rate": 4.291679463280416e-06, + "loss": 0.5339173078536987, + "step": 1150 + }, + { + "epoch": 2.4355179704016914, + "grad_norm": 5.06795597076416, + "learning_rate": 4.276530526410768e-06, + "loss": 0.600986659526825, + "step": 1152 + }, + { + "epoch": 2.4397463002114166, + "grad_norm": 0.9209796190261841, + "learning_rate": 4.261396546946972e-06, + "loss": 0.5248280167579651, + "step": 1154 + }, + { + "epoch": 2.443974630021142, + "grad_norm": 1.0871615409851074, + "learning_rate": 4.246277709908009e-06, + "loss": 0.855151891708374, + "step": 1156 + }, + { + "epoch": 2.448202959830867, + "grad_norm": 7.535534381866455, + "learning_rate": 4.231174200127736e-06, + "loss": 0.5068182945251465, + "step": 1158 + }, + { + "epoch": 2.452431289640592, + "grad_norm": 1.43401300907135, + "learning_rate": 4.216086202252624e-06, + "loss": 0.8611742258071899, + "step": 1160 + }, + { + "epoch": 2.456659619450317, + "grad_norm": 1.1048697233200073, + "learning_rate": 4.201013900739513e-06, + "loss": 0.9013813734054565, + "step": 1162 + }, + { + "epoch": 2.460887949260042, + "grad_norm": 1.087829828262329, + "learning_rate": 4.185957479853344e-06, + "loss": 0.7009021043777466, + "step": 1164 + }, + { + "epoch": 2.4651162790697674, + "grad_norm": 1.2774741649627686, + "learning_rate": 4.1709171236649096e-06, + "loss": 1.0311435461044312, + "step": 1166 + }, + { + "epoch": 2.4693446088794926, + "grad_norm": 1.685935139656067, + "learning_rate": 4.155893016048607e-06, + "loss": 0.9932872653007507, + "step": 1168 + }, + { + "epoch": 2.473572938689218, + "grad_norm": 1.1975206136703491, + "learning_rate": 4.140885340680189e-06, + "loss": 0.6658318638801575, + "step": 1170 + }, + { + "epoch": 2.477801268498943, + "grad_norm": 3.333099603652954, + "learning_rate": 4.125894281034517e-06, + "loss": 0.7946887016296387, + "step": 1172 + }, + { + "epoch": 2.482029598308668, + "grad_norm": 1.138535499572754, + "learning_rate": 4.110920020383316e-06, + "loss": 0.7088789939880371, + "step": 1174 + }, + { + "epoch": 2.4862579281183934, + "grad_norm": 0.7843462228775024, + "learning_rate": 4.0959627417929386e-06, + "loss": 0.6703089475631714, + "step": 1176 + }, + { + "epoch": 2.4904862579281186, + "grad_norm": 2.2591617107391357, + "learning_rate": 4.081022628122127e-06, + "loss": 0.7685297727584839, + "step": 1178 + }, + { + "epoch": 2.4947145877378434, + "grad_norm": 3.7504637241363525, + "learning_rate": 4.06609986201977e-06, + "loss": 0.5655168890953064, + "step": 1180 + }, + { + "epoch": 2.4989429175475686, + "grad_norm": 1.1752467155456543, + "learning_rate": 4.051194625922681e-06, + "loss": 0.3887418210506439, + "step": 1182 + }, + { + "epoch": 2.503171247357294, + "grad_norm": 0.8317999243736267, + "learning_rate": 4.036307102053359e-06, + "loss": 0.1403968781232834, + "step": 1184 + }, + { + "epoch": 2.507399577167019, + "grad_norm": 2.430363893508911, + "learning_rate": 4.021437472417763e-06, + "loss": 0.9367678761482239, + "step": 1186 + }, + { + "epoch": 2.511627906976744, + "grad_norm": 2.165428638458252, + "learning_rate": 4.006585918803091e-06, + "loss": 0.6472379565238953, + "step": 1188 + }, + { + "epoch": 2.5158562367864694, + "grad_norm": 3.2490549087524414, + "learning_rate": 3.991752622775551e-06, + "loss": 0.6814102530479431, + "step": 1190 + }, + { + "epoch": 2.5200845665961946, + "grad_norm": 1.2743620872497559, + "learning_rate": 3.97693776567815e-06, + "loss": 0.8341734409332275, + "step": 1192 + }, + { + "epoch": 2.52431289640592, + "grad_norm": 0.9633957743644714, + "learning_rate": 3.962141528628465e-06, + "loss": 0.6245855093002319, + "step": 1194 + }, + { + "epoch": 2.528541226215645, + "grad_norm": 3.993730306625366, + "learning_rate": 3.947364092516437e-06, + "loss": 0.9198729395866394, + "step": 1196 + }, + { + "epoch": 2.53276955602537, + "grad_norm": 0.7451082468032837, + "learning_rate": 3.932605638002163e-06, + "loss": 0.8981125354766846, + "step": 1198 + }, + { + "epoch": 2.536997885835095, + "grad_norm": 2.992547035217285, + "learning_rate": 3.917866345513677e-06, + "loss": 0.7112588882446289, + "step": 1200 + }, + { + "epoch": 2.54122621564482, + "grad_norm": 0.8251287341117859, + "learning_rate": 3.903146395244749e-06, + "loss": 0.8541390299797058, + "step": 1202 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 7.015642166137695, + "learning_rate": 3.888445967152691e-06, + "loss": 0.28073298931121826, + "step": 1204 + }, + { + "epoch": 2.5496828752642706, + "grad_norm": 1.2791178226470947, + "learning_rate": 3.873765240956138e-06, + "loss": 1.0198954343795776, + "step": 1206 + }, + { + "epoch": 2.553911205073996, + "grad_norm": 1.1683493852615356, + "learning_rate": 3.859104396132863e-06, + "loss": 0.9235135912895203, + "step": 1208 + }, + { + "epoch": 2.558139534883721, + "grad_norm": 0.969961941242218, + "learning_rate": 3.84446361191759e-06, + "loss": 0.8913417458534241, + "step": 1210 + }, + { + "epoch": 2.5623678646934462, + "grad_norm": 2.899411201477051, + "learning_rate": 3.829843067299788e-06, + "loss": 0.644756019115448, + "step": 1212 + }, + { + "epoch": 2.5665961945031714, + "grad_norm": 0.8847817182540894, + "learning_rate": 3.8152429410214834e-06, + "loss": 0.9014704823493958, + "step": 1214 + }, + { + "epoch": 2.570824524312896, + "grad_norm": 2.237797737121582, + "learning_rate": 3.8006634115750886e-06, + "loss": 0.5822610855102539, + "step": 1216 + }, + { + "epoch": 2.5750528541226214, + "grad_norm": 1.428152084350586, + "learning_rate": 3.7861046572012105e-06, + "loss": 0.5281535983085632, + "step": 1218 + }, + { + "epoch": 2.5792811839323466, + "grad_norm": 1.8579766750335693, + "learning_rate": 3.771566855886467e-06, + "loss": 0.4593788981437683, + "step": 1220 + }, + { + "epoch": 2.583509513742072, + "grad_norm": 1.465030550956726, + "learning_rate": 3.7570501853613227e-06, + "loss": 0.5122543573379517, + "step": 1222 + }, + { + "epoch": 2.587737843551797, + "grad_norm": 4.409819602966309, + "learning_rate": 3.742554823097905e-06, + "loss": 0.6159162521362305, + "step": 1224 + }, + { + "epoch": 2.5919661733615222, + "grad_norm": 1.14607572555542, + "learning_rate": 3.7280809463078436e-06, + "loss": 0.3357817828655243, + "step": 1226 + }, + { + "epoch": 2.5961945031712474, + "grad_norm": 2.055947780609131, + "learning_rate": 3.7136287319400944e-06, + "loss": 0.6871626973152161, + "step": 1228 + }, + { + "epoch": 2.6004228329809727, + "grad_norm": 3.8593578338623047, + "learning_rate": 3.6991983566787847e-06, + "loss": 0.753556489944458, + "step": 1230 + }, + { + "epoch": 2.604651162790698, + "grad_norm": 4.141547203063965, + "learning_rate": 3.6847899969410515e-06, + "loss": 0.5551658272743225, + "step": 1232 + }, + { + "epoch": 2.6088794926004226, + "grad_norm": 1.9681657552719116, + "learning_rate": 3.6704038288748812e-06, + "loss": 0.677258312702179, + "step": 1234 + }, + { + "epoch": 2.613107822410148, + "grad_norm": 1.475700855255127, + "learning_rate": 3.6560400283569543e-06, + "loss": 0.74213707447052, + "step": 1236 + }, + { + "epoch": 2.617336152219873, + "grad_norm": 1.0191318988800049, + "learning_rate": 3.641698770990507e-06, + "loss": 0.8331818580627441, + "step": 1238 + }, + { + "epoch": 2.6215644820295982, + "grad_norm": 1.145545244216919, + "learning_rate": 3.6273802321031738e-06, + "loss": 0.674269437789917, + "step": 1240 + }, + { + "epoch": 2.6257928118393234, + "grad_norm": 1.565872311592102, + "learning_rate": 3.6130845867448384e-06, + "loss": 0.8225013017654419, + "step": 1242 + }, + { + "epoch": 2.6300211416490487, + "grad_norm": 1.5022656917572021, + "learning_rate": 3.598812009685514e-06, + "loss": 0.9349151849746704, + "step": 1244 + }, + { + "epoch": 2.634249471458774, + "grad_norm": 10.347308158874512, + "learning_rate": 3.584562675413189e-06, + "loss": 0.6068240404129028, + "step": 1246 + }, + { + "epoch": 2.638477801268499, + "grad_norm": 3.0747714042663574, + "learning_rate": 3.5703367581316985e-06, + "loss": 0.5004937648773193, + "step": 1248 + }, + { + "epoch": 2.6427061310782243, + "grad_norm": 1.4242684841156006, + "learning_rate": 3.5561344317585957e-06, + "loss": 0.8734555244445801, + "step": 1250 + }, + { + "epoch": 2.646934460887949, + "grad_norm": 2.3883349895477295, + "learning_rate": 3.5419558699230305e-06, + "loss": 0.5485278964042664, + "step": 1252 + }, + { + "epoch": 2.6511627906976747, + "grad_norm": 2.4504101276397705, + "learning_rate": 3.52780124596361e-06, + "loss": 0.5928559303283691, + "step": 1254 + }, + { + "epoch": 2.6553911205073994, + "grad_norm": 0.8869137763977051, + "learning_rate": 3.5136707329263055e-06, + "loss": 0.8405211567878723, + "step": 1256 + }, + { + "epoch": 2.6596194503171247, + "grad_norm": 1.0010604858398438, + "learning_rate": 3.499564503562315e-06, + "loss": 0.24002832174301147, + "step": 1258 + }, + { + "epoch": 2.66384778012685, + "grad_norm": 0.49673745036125183, + "learning_rate": 3.4854827303259575e-06, + "loss": 0.42692404985427856, + "step": 1260 + }, + { + "epoch": 2.668076109936575, + "grad_norm": 2.6645731925964355, + "learning_rate": 3.47142558537257e-06, + "loss": 0.5938810110092163, + "step": 1262 + }, + { + "epoch": 2.6723044397463003, + "grad_norm": 2.7925870418548584, + "learning_rate": 3.4573932405563966e-06, + "loss": 0.6461769938468933, + "step": 1264 + }, + { + "epoch": 2.6765327695560255, + "grad_norm": 1.5583502054214478, + "learning_rate": 3.4433858674284894e-06, + "loss": 0.35567256808280945, + "step": 1266 + }, + { + "epoch": 2.6807610993657507, + "grad_norm": 0.887532114982605, + "learning_rate": 3.429403637234614e-06, + "loss": 0.6280087828636169, + "step": 1268 + }, + { + "epoch": 2.6849894291754755, + "grad_norm": 1.1162208318710327, + "learning_rate": 3.4154467209131515e-06, + "loss": 0.869036853313446, + "step": 1270 + }, + { + "epoch": 2.689217758985201, + "grad_norm": 2.711535930633545, + "learning_rate": 3.401515289093011e-06, + "loss": 0.6568902134895325, + "step": 1272 + }, + { + "epoch": 2.693446088794926, + "grad_norm": 0.3214070498943329, + "learning_rate": 3.3876095120915433e-06, + "loss": 0.412013441324234, + "step": 1274 + }, + { + "epoch": 2.697674418604651, + "grad_norm": 1.451460838317871, + "learning_rate": 3.3737295599124592e-06, + "loss": 0.3878551125526428, + "step": 1276 + }, + { + "epoch": 2.7019027484143763, + "grad_norm": 1.3137333393096924, + "learning_rate": 3.3598756022437507e-06, + "loss": 0.9230884313583374, + "step": 1278 + }, + { + "epoch": 2.7061310782241015, + "grad_norm": 1.9198719263076782, + "learning_rate": 3.3460478084556157e-06, + "loss": 0.5730487704277039, + "step": 1280 + }, + { + "epoch": 2.7103594080338267, + "grad_norm": 0.9012884497642517, + "learning_rate": 3.33224634759839e-06, + "loss": 1.046750545501709, + "step": 1282 + }, + { + "epoch": 2.714587737843552, + "grad_norm": 0.44456279277801514, + "learning_rate": 3.318471388400476e-06, + "loss": 0.6401486396789551, + "step": 1284 + }, + { + "epoch": 2.718816067653277, + "grad_norm": 0.2594262361526489, + "learning_rate": 3.3047230992662834e-06, + "loss": 0.9082249999046326, + "step": 1286 + }, + { + "epoch": 2.723044397463002, + "grad_norm": 1.2604821920394897, + "learning_rate": 3.291001648274175e-06, + "loss": 0.4901333749294281, + "step": 1288 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 1.0342061519622803, + "learning_rate": 3.277307203174395e-06, + "loss": 0.9255348443984985, + "step": 1290 + }, + { + "epoch": 2.7315010570824523, + "grad_norm": 1.1304163932800293, + "learning_rate": 3.263639931387039e-06, + "loss": 1.0709456205368042, + "step": 1292 + }, + { + "epoch": 2.7357293868921775, + "grad_norm": 1.1676743030548096, + "learning_rate": 3.2500000000000015e-06, + "loss": 0.8695792555809021, + "step": 1294 + }, + { + "epoch": 2.7399577167019027, + "grad_norm": 0.8143407702445984, + "learning_rate": 3.236387575766916e-06, + "loss": 0.2564300000667572, + "step": 1296 + }, + { + "epoch": 2.744186046511628, + "grad_norm": 0.8985607624053955, + "learning_rate": 3.222802825105138e-06, + "loss": 0.5201926827430725, + "step": 1298 + }, + { + "epoch": 2.748414376321353, + "grad_norm": 3.784648895263672, + "learning_rate": 3.209245914093708e-06, + "loss": 0.6484920382499695, + "step": 1300 + }, + { + "epoch": 2.7526427061310783, + "grad_norm": 1.2287521362304688, + "learning_rate": 3.1957170084713053e-06, + "loss": 0.697693407535553, + "step": 1302 + }, + { + "epoch": 2.7568710359408035, + "grad_norm": 0.9096108078956604, + "learning_rate": 3.182216273634235e-06, + "loss": 0.6163878440856934, + "step": 1304 + }, + { + "epoch": 2.7610993657505283, + "grad_norm": 2.062570095062256, + "learning_rate": 3.1687438746344086e-06, + "loss": 0.7268993854522705, + "step": 1306 + }, + { + "epoch": 2.765327695560254, + "grad_norm": 2.1470589637756348, + "learning_rate": 3.1552999761773166e-06, + "loss": 0.7619702816009521, + "step": 1308 + }, + { + "epoch": 2.7695560253699787, + "grad_norm": 1.1526496410369873, + "learning_rate": 3.141884742620013e-06, + "loss": 0.9826087355613708, + "step": 1310 + }, + { + "epoch": 2.773784355179704, + "grad_norm": 1.6171549558639526, + "learning_rate": 3.1284983379691247e-06, + "loss": 0.710430920124054, + "step": 1312 + }, + { + "epoch": 2.778012684989429, + "grad_norm": 2.406308650970459, + "learning_rate": 3.1151409258788235e-06, + "loss": 0.4855649173259735, + "step": 1314 + }, + { + "epoch": 2.7822410147991543, + "grad_norm": 1.0452547073364258, + "learning_rate": 3.101812669648838e-06, + "loss": 0.7923295497894287, + "step": 1316 + }, + { + "epoch": 2.7864693446088795, + "grad_norm": 1.4424408674240112, + "learning_rate": 3.0885137322224546e-06, + "loss": 0.551530659198761, + "step": 1318 + }, + { + "epoch": 2.7906976744186047, + "grad_norm": 1.417302131652832, + "learning_rate": 3.075244276184529e-06, + "loss": 0.8195257186889648, + "step": 1320 + }, + { + "epoch": 2.79492600422833, + "grad_norm": 2.7174646854400635, + "learning_rate": 3.0620044637594936e-06, + "loss": 0.32833823561668396, + "step": 1322 + }, + { + "epoch": 2.7991543340380547, + "grad_norm": 1.0890730619430542, + "learning_rate": 3.0487944568093685e-06, + "loss": 0.8992734551429749, + "step": 1324 + }, + { + "epoch": 2.8033826638477803, + "grad_norm": 3.6498396396636963, + "learning_rate": 3.035614416831801e-06, + "loss": 0.22638478875160217, + "step": 1326 + }, + { + "epoch": 2.807610993657505, + "grad_norm": 2.6539711952209473, + "learning_rate": 3.0224645049580746e-06, + "loss": 0.3162754476070404, + "step": 1328 + }, + { + "epoch": 2.8118393234672303, + "grad_norm": 0.9214186072349548, + "learning_rate": 3.0093448819511416e-06, + "loss": 0.4994361698627472, + "step": 1330 + }, + { + "epoch": 2.8160676532769555, + "grad_norm": 2.294234275817871, + "learning_rate": 2.9962557082036677e-06, + "loss": 0.96239173412323, + "step": 1332 + }, + { + "epoch": 2.8202959830866807, + "grad_norm": 2.4485044479370117, + "learning_rate": 2.9831971437360606e-06, + "loss": 0.5888533592224121, + "step": 1334 + }, + { + "epoch": 2.824524312896406, + "grad_norm": 2.017869234085083, + "learning_rate": 2.9701693481945132e-06, + "loss": 0.9644511938095093, + "step": 1336 + }, + { + "epoch": 2.828752642706131, + "grad_norm": 1.0167880058288574, + "learning_rate": 2.957172480849064e-06, + "loss": 0.92995685338974, + "step": 1338 + }, + { + "epoch": 2.8329809725158563, + "grad_norm": 1.2603472471237183, + "learning_rate": 2.9442067005916375e-06, + "loss": 0.6602538228034973, + "step": 1340 + }, + { + "epoch": 2.8372093023255816, + "grad_norm": 1.268857717514038, + "learning_rate": 2.931272165934105e-06, + "loss": 0.9626938700675964, + "step": 1342 + }, + { + "epoch": 2.8414376321353068, + "grad_norm": 1.172420620918274, + "learning_rate": 2.918369035006351e-06, + "loss": 0.6409440040588379, + "step": 1344 + }, + { + "epoch": 2.8456659619450315, + "grad_norm": 1.6903941631317139, + "learning_rate": 2.9054974655543335e-06, + "loss": 0.5651144981384277, + "step": 1346 + }, + { + "epoch": 2.8498942917547567, + "grad_norm": 1.0837018489837646, + "learning_rate": 2.89265761493816e-06, + "loss": 0.973859429359436, + "step": 1348 + }, + { + "epoch": 2.854122621564482, + "grad_norm": 2.4711711406707764, + "learning_rate": 2.8798496401301644e-06, + "loss": 0.49223342537879944, + "step": 1350 + }, + { + "epoch": 2.858350951374207, + "grad_norm": 3.472890615463257, + "learning_rate": 2.8670736977129843e-06, + "loss": 0.811894416809082, + "step": 1352 + }, + { + "epoch": 2.8625792811839323, + "grad_norm": 0.8379427194595337, + "learning_rate": 2.854329943877647e-06, + "loss": 0.8549838066101074, + "step": 1354 + }, + { + "epoch": 2.8668076109936576, + "grad_norm": 7.758663177490234, + "learning_rate": 2.841618534421664e-06, + "loss": 0.3211415112018585, + "step": 1356 + }, + { + "epoch": 2.8710359408033828, + "grad_norm": 1.014855980873108, + "learning_rate": 2.8289396247471227e-06, + "loss": 0.9500829577445984, + "step": 1358 + }, + { + "epoch": 2.875264270613108, + "grad_norm": 2.9077837467193604, + "learning_rate": 2.816293369858788e-06, + "loss": 0.6253378391265869, + "step": 1360 + }, + { + "epoch": 2.879492600422833, + "grad_norm": 4.78998327255249, + "learning_rate": 2.803679924362206e-06, + "loss": 0.5822049379348755, + "step": 1362 + }, + { + "epoch": 2.883720930232558, + "grad_norm": 1.0024492740631104, + "learning_rate": 2.7910994424618165e-06, + "loss": 0.9647387862205505, + "step": 1364 + }, + { + "epoch": 2.887949260042283, + "grad_norm": 2.4475505352020264, + "learning_rate": 2.778552077959064e-06, + "loss": 0.7316675186157227, + "step": 1366 + }, + { + "epoch": 2.8921775898520083, + "grad_norm": 0.9589321613311768, + "learning_rate": 2.766037984250521e-06, + "loss": 0.8495209813117981, + "step": 1368 + }, + { + "epoch": 2.8964059196617336, + "grad_norm": 1.0323636531829834, + "learning_rate": 2.753557314326011e-06, + "loss": 0.9080972075462341, + "step": 1370 + }, + { + "epoch": 2.9006342494714588, + "grad_norm": 1.1502765417099, + "learning_rate": 2.7411102207667373e-06, + "loss": 1.0378683805465698, + "step": 1372 + }, + { + "epoch": 2.904862579281184, + "grad_norm": 1.1626389026641846, + "learning_rate": 2.728696855743418e-06, + "loss": 0.7391030192375183, + "step": 1374 + }, + { + "epoch": 2.909090909090909, + "grad_norm": 0.8892068862915039, + "learning_rate": 2.7163173710144323e-06, + "loss": 0.8448020219802856, + "step": 1376 + }, + { + "epoch": 2.9133192389006344, + "grad_norm": 2.1172099113464355, + "learning_rate": 2.703971917923951e-06, + "loss": 0.5612887144088745, + "step": 1378 + }, + { + "epoch": 2.9175475687103596, + "grad_norm": 1.078094244003296, + "learning_rate": 2.6916606474000965e-06, + "loss": 0.6208653450012207, + "step": 1380 + }, + { + "epoch": 2.9217758985200843, + "grad_norm": 1.234604001045227, + "learning_rate": 2.679383709953103e-06, + "loss": 0.9919822812080383, + "step": 1382 + }, + { + "epoch": 2.92600422832981, + "grad_norm": 1.8408221006393433, + "learning_rate": 2.6671412556734578e-06, + "loss": 0.6899129748344421, + "step": 1384 + }, + { + "epoch": 2.9302325581395348, + "grad_norm": 2.869661569595337, + "learning_rate": 2.6549334342300844e-06, + "loss": 0.6663805246353149, + "step": 1386 + }, + { + "epoch": 2.93446088794926, + "grad_norm": 1.1273760795593262, + "learning_rate": 2.642760394868503e-06, + "loss": 0.840316891670227, + "step": 1388 + }, + { + "epoch": 2.938689217758985, + "grad_norm": 0.445371150970459, + "learning_rate": 2.6306222864090157e-06, + "loss": 0.6287580132484436, + "step": 1390 + }, + { + "epoch": 2.9429175475687104, + "grad_norm": 0.9976673722267151, + "learning_rate": 2.61851925724487e-06, + "loss": 1.0280648469924927, + "step": 1392 + }, + { + "epoch": 2.9471458773784356, + "grad_norm": 0.831870973110199, + "learning_rate": 2.606451455340459e-06, + "loss": 0.24350617825984955, + "step": 1394 + }, + { + "epoch": 2.951374207188161, + "grad_norm": 0.8883412480354309, + "learning_rate": 2.5944190282295155e-06, + "loss": 0.47680070996284485, + "step": 1396 + }, + { + "epoch": 2.955602536997886, + "grad_norm": 1.984863042831421, + "learning_rate": 2.5824221230132895e-06, + "loss": 0.972530722618103, + "step": 1398 + }, + { + "epoch": 2.9598308668076108, + "grad_norm": 0.8938445448875427, + "learning_rate": 2.5704608863587676e-06, + "loss": 0.9972274899482727, + "step": 1400 + }, + { + "epoch": 2.9640591966173364, + "grad_norm": 2.3360066413879395, + "learning_rate": 2.558535464496877e-06, + "loss": 0.6626827716827393, + "step": 1402 + }, + { + "epoch": 2.968287526427061, + "grad_norm": 0.7893721461296082, + "learning_rate": 2.5466460032206872e-06, + "loss": 0.4602081775665283, + "step": 1404 + }, + { + "epoch": 2.9725158562367864, + "grad_norm": 1.3927929401397705, + "learning_rate": 2.53479264788364e-06, + "loss": 0.7620840668678284, + "step": 1406 + }, + { + "epoch": 2.9767441860465116, + "grad_norm": 1.0482085943222046, + "learning_rate": 2.522975543397769e-06, + "loss": 0.7396769523620605, + "step": 1408 + }, + { + "epoch": 2.980972515856237, + "grad_norm": 2.9075868129730225, + "learning_rate": 2.511194834231922e-06, + "loss": 0.6746079325675964, + "step": 1410 + }, + { + "epoch": 2.985200845665962, + "grad_norm": 2.8968935012817383, + "learning_rate": 2.49945066441e-06, + "loss": 0.8901047110557556, + "step": 1412 + }, + { + "epoch": 2.989429175475687, + "grad_norm": 0.8228777647018433, + "learning_rate": 2.4877431775092004e-06, + "loss": 0.6506577730178833, + "step": 1414 + }, + { + "epoch": 2.9936575052854124, + "grad_norm": 3.279634475708008, + "learning_rate": 2.4760725166582544e-06, + "loss": 0.7807758450508118, + "step": 1416 + }, + { + "epoch": 2.997885835095137, + "grad_norm": 1.5837316513061523, + "learning_rate": 2.464438824535675e-06, + "loss": 0.4736751317977905, + "step": 1418 + }, + { + "epoch": 3.0021141649048624, + "grad_norm": 1.010254144668579, + "learning_rate": 2.452842243368028e-06, + "loss": 0.8213729858398438, + "step": 1420 + }, + { + "epoch": 3.0063424947145876, + "grad_norm": 1.7557199001312256, + "learning_rate": 2.4412829149281735e-06, + "loss": 0.49449074268341064, + "step": 1422 + }, + { + "epoch": 3.010570824524313, + "grad_norm": 3.6423122882843018, + "learning_rate": 2.4297609805335472e-06, + "loss": 0.7657237648963928, + "step": 1424 + }, + { + "epoch": 3.014799154334038, + "grad_norm": 4.304173946380615, + "learning_rate": 2.4182765810444263e-06, + "loss": 0.4311809539794922, + "step": 1426 + }, + { + "epoch": 3.019027484143763, + "grad_norm": 0.8986531496047974, + "learning_rate": 2.4068298568622083e-06, + "loss": 0.4538921117782593, + "step": 1428 + }, + { + "epoch": 3.0232558139534884, + "grad_norm": 3.3296451568603516, + "learning_rate": 2.395420947927694e-06, + "loss": 0.6588428020477295, + "step": 1430 + }, + { + "epoch": 3.0274841437632136, + "grad_norm": 0.9242261648178101, + "learning_rate": 2.3840499937193804e-06, + "loss": 0.7460368275642395, + "step": 1432 + }, + { + "epoch": 3.031712473572939, + "grad_norm": 0.9263319969177246, + "learning_rate": 2.3727171332517496e-06, + "loss": 0.7190459966659546, + "step": 1434 + }, + { + "epoch": 3.035940803382664, + "grad_norm": 2.1857717037200928, + "learning_rate": 2.361422505073575e-06, + "loss": 0.5604714751243591, + "step": 1436 + }, + { + "epoch": 3.040169133192389, + "grad_norm": 0.5939431190490723, + "learning_rate": 2.3501662472662233e-06, + "loss": 0.13830041885375977, + "step": 1438 + }, + { + "epoch": 3.044397463002114, + "grad_norm": 1.2216732501983643, + "learning_rate": 2.3389484974419685e-06, + "loss": 0.851690411567688, + "step": 1440 + }, + { + "epoch": 3.048625792811839, + "grad_norm": 0.8498998880386353, + "learning_rate": 2.327769392742309e-06, + "loss": 0.49222394824028015, + "step": 1442 + }, + { + "epoch": 3.0528541226215644, + "grad_norm": 3.8317887783050537, + "learning_rate": 2.316629069836291e-06, + "loss": 0.11679022014141083, + "step": 1444 + }, + { + "epoch": 3.0570824524312896, + "grad_norm": 2.6142446994781494, + "learning_rate": 2.305527664918838e-06, + "loss": 0.40334203839302063, + "step": 1446 + }, + { + "epoch": 3.061310782241015, + "grad_norm": 2.526332378387451, + "learning_rate": 2.294465313709085e-06, + "loss": 0.5736421346664429, + "step": 1448 + }, + { + "epoch": 3.06553911205074, + "grad_norm": 0.9735236763954163, + "learning_rate": 2.2834421514487205e-06, + "loss": 0.5415090918540955, + "step": 1450 + }, + { + "epoch": 3.0697674418604652, + "grad_norm": 2.981172561645508, + "learning_rate": 2.27245831290033e-06, + "loss": 0.6346736550331116, + "step": 1452 + }, + { + "epoch": 3.0739957716701904, + "grad_norm": 1.0074353218078613, + "learning_rate": 2.261513932345755e-06, + "loss": 0.7936796545982361, + "step": 1454 + }, + { + "epoch": 3.0782241014799157, + "grad_norm": 1.2012394666671753, + "learning_rate": 2.250609143584443e-06, + "loss": 0.21504582464694977, + "step": 1456 + }, + { + "epoch": 3.0824524312896404, + "grad_norm": 1.1107393503189087, + "learning_rate": 2.239744079931818e-06, + "loss": 0.48933297395706177, + "step": 1458 + }, + { + "epoch": 3.0866807610993656, + "grad_norm": 2.5664312839508057, + "learning_rate": 2.228918874217649e-06, + "loss": 0.7915838956832886, + "step": 1460 + }, + { + "epoch": 3.090909090909091, + "grad_norm": 0.9385368824005127, + "learning_rate": 2.2181336587844244e-06, + "loss": 0.7082565426826477, + "step": 1462 + }, + { + "epoch": 3.095137420718816, + "grad_norm": 0.9007372856140137, + "learning_rate": 2.207388565485736e-06, + "loss": 0.4987450838088989, + "step": 1464 + }, + { + "epoch": 3.0993657505285412, + "grad_norm": 2.6377320289611816, + "learning_rate": 2.196683725684666e-06, + "loss": 0.5514395236968994, + "step": 1466 + }, + { + "epoch": 3.1035940803382664, + "grad_norm": 1.0533729791641235, + "learning_rate": 2.1860192702521826e-06, + "loss": 0.8255153298377991, + "step": 1468 + }, + { + "epoch": 3.1078224101479917, + "grad_norm": 1.022743821144104, + "learning_rate": 2.1753953295655354e-06, + "loss": 0.733326256275177, + "step": 1470 + }, + { + "epoch": 3.112050739957717, + "grad_norm": 3.6769115924835205, + "learning_rate": 2.164812033506673e-06, + "loss": 0.47708556056022644, + "step": 1472 + }, + { + "epoch": 3.116279069767442, + "grad_norm": 1.3268944025039673, + "learning_rate": 2.1542695114606367e-06, + "loss": 0.9431726336479187, + "step": 1474 + }, + { + "epoch": 3.120507399577167, + "grad_norm": 0.9920517802238464, + "learning_rate": 2.1437678923139944e-06, + "loss": 0.7208531498908997, + "step": 1476 + }, + { + "epoch": 3.124735729386892, + "grad_norm": 24.436758041381836, + "learning_rate": 2.1333073044532628e-06, + "loss": 0.2605661153793335, + "step": 1478 + }, + { + "epoch": 3.1289640591966172, + "grad_norm": 1.1512051820755005, + "learning_rate": 2.122887875763327e-06, + "loss": 0.7009040713310242, + "step": 1480 + }, + { + "epoch": 3.1331923890063424, + "grad_norm": 0.9686696529388428, + "learning_rate": 2.1125097336258897e-06, + "loss": 0.784214437007904, + "step": 1482 + }, + { + "epoch": 3.1374207188160677, + "grad_norm": 0.3542078733444214, + "learning_rate": 2.1021730049179117e-06, + "loss": 0.19213739037513733, + "step": 1484 + }, + { + "epoch": 3.141649048625793, + "grad_norm": 2.5765812397003174, + "learning_rate": 2.09187781601005e-06, + "loss": 0.7180091142654419, + "step": 1486 + }, + { + "epoch": 3.145877378435518, + "grad_norm": 1.3088688850402832, + "learning_rate": 2.0816242927651254e-06, + "loss": 0.9011077284812927, + "step": 1488 + }, + { + "epoch": 3.1501057082452433, + "grad_norm": 11.290595054626465, + "learning_rate": 2.071412560536581e-06, + "loss": 0.5567789077758789, + "step": 1490 + }, + { + "epoch": 3.1543340380549685, + "grad_norm": 2.581059694290161, + "learning_rate": 2.0612427441669475e-06, + "loss": 0.5206165909767151, + "step": 1492 + }, + { + "epoch": 3.1585623678646932, + "grad_norm": 1.5132797956466675, + "learning_rate": 2.051114967986311e-06, + "loss": 0.3697885274887085, + "step": 1494 + }, + { + "epoch": 3.1627906976744184, + "grad_norm": 0.19122721254825592, + "learning_rate": 2.0410293558108074e-06, + "loss": 0.6020835041999817, + "step": 1496 + }, + { + "epoch": 3.1670190274841437, + "grad_norm": 1.0161772966384888, + "learning_rate": 2.0309860309410986e-06, + "loss": 0.7800092101097107, + "step": 1498 + }, + { + "epoch": 3.171247357293869, + "grad_norm": 2.0083847045898438, + "learning_rate": 2.0209851161608613e-06, + "loss": 0.8102139234542847, + "step": 1500 + }, + { + "epoch": 3.175475687103594, + "grad_norm": 2.8630428314208984, + "learning_rate": 2.011026733735301e-06, + "loss": 0.23411083221435547, + "step": 1502 + }, + { + "epoch": 3.1797040169133193, + "grad_norm": 0.7848597764968872, + "learning_rate": 2.001111005409643e-06, + "loss": 0.40008798241615295, + "step": 1504 + }, + { + "epoch": 3.1839323467230445, + "grad_norm": 4.700876712799072, + "learning_rate": 1.9912380524076503e-06, + "loss": 0.5403401255607605, + "step": 1506 + }, + { + "epoch": 3.1881606765327697, + "grad_norm": 1.7473686933517456, + "learning_rate": 1.981407995430136e-06, + "loss": 0.3486909866333008, + "step": 1508 + }, + { + "epoch": 3.192389006342495, + "grad_norm": 2.9273171424865723, + "learning_rate": 1.9716209546535e-06, + "loss": 0.2166144996881485, + "step": 1510 + }, + { + "epoch": 3.1966173361522197, + "grad_norm": 0.23416613042354584, + "learning_rate": 1.9618770497282487e-06, + "loss": 0.04864403232932091, + "step": 1512 + }, + { + "epoch": 3.200845665961945, + "grad_norm": 2.1424765586853027, + "learning_rate": 1.9521763997775294e-06, + "loss": 0.7215962409973145, + "step": 1514 + }, + { + "epoch": 3.20507399577167, + "grad_norm": 1.2655258178710938, + "learning_rate": 1.942519123395689e-06, + "loss": 0.6798702478408813, + "step": 1516 + }, + { + "epoch": 3.2093023255813953, + "grad_norm": 1.9534128904342651, + "learning_rate": 1.932905338646812e-06, + "loss": 0.4002809226512909, + "step": 1518 + }, + { + "epoch": 3.2135306553911205, + "grad_norm": 1.3802556991577148, + "learning_rate": 1.9233351630632747e-06, + "loss": 0.657711386680603, + "step": 1520 + }, + { + "epoch": 3.2177589852008457, + "grad_norm": 1.261686086654663, + "learning_rate": 1.9138087136443227e-06, + "loss": 0.7206029295921326, + "step": 1522 + }, + { + "epoch": 3.221987315010571, + "grad_norm": 2.353743314743042, + "learning_rate": 1.904326106854626e-06, + "loss": 0.7720223665237427, + "step": 1524 + }, + { + "epoch": 3.226215644820296, + "grad_norm": 1.2002711296081543, + "learning_rate": 1.8948874586228622e-06, + "loss": 0.6227959394454956, + "step": 1526 + }, + { + "epoch": 3.2304439746300213, + "grad_norm": 3.1747052669525146, + "learning_rate": 1.8854928843402987e-06, + "loss": 0.09834546595811844, + "step": 1528 + }, + { + "epoch": 3.234672304439746, + "grad_norm": 10.57388973236084, + "learning_rate": 1.8761424988593796e-06, + "loss": 0.6881338357925415, + "step": 1530 + }, + { + "epoch": 3.2389006342494713, + "grad_norm": 2.4264867305755615, + "learning_rate": 1.8668364164923248e-06, + "loss": 0.7460035085678101, + "step": 1532 + }, + { + "epoch": 3.2431289640591965, + "grad_norm": 1.3846110105514526, + "learning_rate": 1.8575747510097294e-06, + "loss": 0.715628445148468, + "step": 1534 + }, + { + "epoch": 3.2473572938689217, + "grad_norm": 1.2345430850982666, + "learning_rate": 1.8483576156391747e-06, + "loss": 0.45820537209510803, + "step": 1536 + }, + { + "epoch": 3.251585623678647, + "grad_norm": 1.1302199363708496, + "learning_rate": 1.8391851230638443e-06, + "loss": 0.4934888780117035, + "step": 1538 + }, + { + "epoch": 3.255813953488372, + "grad_norm": 0.9766497611999512, + "learning_rate": 1.8300573854211452e-06, + "loss": 0.4033285975456238, + "step": 1540 + }, + { + "epoch": 3.2600422832980973, + "grad_norm": 8.77962875366211, + "learning_rate": 1.8209745143013363e-06, + "loss": 0.992950439453125, + "step": 1542 + }, + { + "epoch": 3.2642706131078225, + "grad_norm": 1.1472126245498657, + "learning_rate": 1.811936620746167e-06, + "loss": 0.8389906287193298, + "step": 1544 + }, + { + "epoch": 3.2684989429175477, + "grad_norm": 1.3688768148422241, + "learning_rate": 1.8029438152475159e-06, + "loss": 0.6483197212219238, + "step": 1546 + }, + { + "epoch": 3.2727272727272725, + "grad_norm": 0.6395139098167419, + "learning_rate": 1.7939962077460448e-06, + "loss": 0.19004815816879272, + "step": 1548 + }, + { + "epoch": 3.276955602536998, + "grad_norm": 2.324686288833618, + "learning_rate": 1.785093907629849e-06, + "loss": 0.5779064893722534, + "step": 1550 + }, + { + "epoch": 3.281183932346723, + "grad_norm": 0.9998548626899719, + "learning_rate": 1.7762370237331233e-06, + "loss": 0.7248525619506836, + "step": 1552 + }, + { + "epoch": 3.285412262156448, + "grad_norm": 0.19214117527008057, + "learning_rate": 1.7674256643348348e-06, + "loss": 0.07821393758058548, + "step": 1554 + }, + { + "epoch": 3.2896405919661733, + "grad_norm": 1.257800817489624, + "learning_rate": 1.7586599371573887e-06, + "loss": 0.6245786547660828, + "step": 1556 + }, + { + "epoch": 3.2938689217758985, + "grad_norm": 1.0998709201812744, + "learning_rate": 1.7499399493653225e-06, + "loss": 0.43618279695510864, + "step": 1558 + }, + { + "epoch": 3.2980972515856237, + "grad_norm": 2.291769504547119, + "learning_rate": 1.7412658075639921e-06, + "loss": 0.9212561845779419, + "step": 1560 + }, + { + "epoch": 3.302325581395349, + "grad_norm": 7.099034309387207, + "learning_rate": 1.7326376177982635e-06, + "loss": 0.533191978931427, + "step": 1562 + }, + { + "epoch": 3.306553911205074, + "grad_norm": 1.251503586769104, + "learning_rate": 1.7240554855512225e-06, + "loss": 0.366091251373291, + "step": 1564 + }, + { + "epoch": 3.3107822410147993, + "grad_norm": 1.1921592950820923, + "learning_rate": 1.7155195157428878e-06, + "loss": 0.46826988458633423, + "step": 1566 + }, + { + "epoch": 3.3150105708245245, + "grad_norm": 0.3862226903438568, + "learning_rate": 1.707029812728916e-06, + "loss": 0.12568050622940063, + "step": 1568 + }, + { + "epoch": 3.3192389006342493, + "grad_norm": 2.235180139541626, + "learning_rate": 1.6985864802993366e-06, + "loss": 0.42736196517944336, + "step": 1570 + }, + { + "epoch": 3.3234672304439745, + "grad_norm": 4.545248031616211, + "learning_rate": 1.690189621677284e-06, + "loss": 0.5469956994056702, + "step": 1572 + }, + { + "epoch": 3.3276955602536997, + "grad_norm": 2.818523406982422, + "learning_rate": 1.681839339517728e-06, + "loss": 0.17828865349292755, + "step": 1574 + }, + { + "epoch": 3.331923890063425, + "grad_norm": 2.0850532054901123, + "learning_rate": 1.67353573590622e-06, + "loss": 0.22279733419418335, + "step": 1576 + }, + { + "epoch": 3.33615221987315, + "grad_norm": 1.682937502861023, + "learning_rate": 1.6652789123576513e-06, + "loss": 0.4197770655155182, + "step": 1578 + }, + { + "epoch": 3.3403805496828753, + "grad_norm": 1.6053372621536255, + "learning_rate": 1.6570689698150094e-06, + "loss": 0.430209219455719, + "step": 1580 + }, + { + "epoch": 3.3446088794926006, + "grad_norm": 1.104996681213379, + "learning_rate": 1.6489060086481373e-06, + "loss": 0.8024388551712036, + "step": 1582 + }, + { + "epoch": 3.3488372093023258, + "grad_norm": 1.862497329711914, + "learning_rate": 1.6407901286525152e-06, + "loss": 0.5327326655387878, + "step": 1584 + }, + { + "epoch": 3.353065539112051, + "grad_norm": 1.4750189781188965, + "learning_rate": 1.6327214290480382e-06, + "loss": 0.5549164414405823, + "step": 1586 + }, + { + "epoch": 3.3572938689217757, + "grad_norm": 2.866227388381958, + "learning_rate": 1.6247000084778006e-06, + "loss": 0.7654522061347961, + "step": 1588 + }, + { + "epoch": 3.361522198731501, + "grad_norm": 1.1869614124298096, + "learning_rate": 1.6167259650068887e-06, + "loss": 0.21528974175453186, + "step": 1590 + }, + { + "epoch": 3.365750528541226, + "grad_norm": 1.1272586584091187, + "learning_rate": 1.6087993961211904e-06, + "loss": 0.6685192584991455, + "step": 1592 + }, + { + "epoch": 3.3699788583509513, + "grad_norm": 1.4778391122817993, + "learning_rate": 1.6009203987261957e-06, + "loss": 0.24215056002140045, + "step": 1594 + }, + { + "epoch": 3.3742071881606766, + "grad_norm": 1.2095019817352295, + "learning_rate": 1.593089069145809e-06, + "loss": 0.7534793019294739, + "step": 1596 + }, + { + "epoch": 3.3784355179704018, + "grad_norm": 1.9421671628952026, + "learning_rate": 1.5853055031211844e-06, + "loss": 0.43499889969825745, + "step": 1598 + }, + { + "epoch": 3.382663847780127, + "grad_norm": 0.5567390322685242, + "learning_rate": 1.577569795809544e-06, + "loss": 0.25693437457084656, + "step": 1600 + }, + { + "epoch": 3.386892177589852, + "grad_norm": 10.710603713989258, + "learning_rate": 1.5698820417830137e-06, + "loss": 0.4831811785697937, + "step": 1602 + }, + { + "epoch": 3.3911205073995774, + "grad_norm": 1.0061455965042114, + "learning_rate": 1.5622423350274784e-06, + "loss": 0.7833957076072693, + "step": 1604 + }, + { + "epoch": 3.395348837209302, + "grad_norm": 0.9558950066566467, + "learning_rate": 1.5546507689414217e-06, + "loss": 0.7573481202125549, + "step": 1606 + }, + { + "epoch": 3.3995771670190273, + "grad_norm": 6.826617240905762, + "learning_rate": 1.5471074363347894e-06, + "loss": 0.389364093542099, + "step": 1608 + }, + { + "epoch": 3.4038054968287526, + "grad_norm": 1.3320974111557007, + "learning_rate": 1.539612429427852e-06, + "loss": 0.8547945022583008, + "step": 1610 + }, + { + "epoch": 3.4080338266384778, + "grad_norm": 1.4288395643234253, + "learning_rate": 1.5321658398500808e-06, + "loss": 0.6717844605445862, + "step": 1612 + }, + { + "epoch": 3.412262156448203, + "grad_norm": 2.4894814491271973, + "learning_rate": 1.5247677586390255e-06, + "loss": 0.5470836758613586, + "step": 1614 + }, + { + "epoch": 3.416490486257928, + "grad_norm": 1.803152322769165, + "learning_rate": 1.5174182762392007e-06, + "loss": 0.7587092518806458, + "step": 1616 + }, + { + "epoch": 3.4207188160676534, + "grad_norm": 2.1030826568603516, + "learning_rate": 1.5101174825009815e-06, + "loss": 0.7187841534614563, + "step": 1618 + }, + { + "epoch": 3.4249471458773786, + "grad_norm": 1.0092041492462158, + "learning_rate": 1.5028654666795042e-06, + "loss": 0.7382463812828064, + "step": 1620 + }, + { + "epoch": 3.429175475687104, + "grad_norm": 2.2474474906921387, + "learning_rate": 1.495662317433576e-06, + "loss": 0.5108706951141357, + "step": 1622 + }, + { + "epoch": 3.4334038054968286, + "grad_norm": 1.1113120317459106, + "learning_rate": 1.4885081228245909e-06, + "loss": 0.6870447397232056, + "step": 1624 + }, + { + "epoch": 3.4376321353065538, + "grad_norm": 1.7375661134719849, + "learning_rate": 1.4814029703154515e-06, + "loss": 0.6048234701156616, + "step": 1626 + }, + { + "epoch": 3.441860465116279, + "grad_norm": 1.2149378061294556, + "learning_rate": 1.4743469467695017e-06, + "loss": 0.8593543767929077, + "step": 1628 + }, + { + "epoch": 3.446088794926004, + "grad_norm": 2.055464744567871, + "learning_rate": 1.4673401384494651e-06, + "loss": 0.4548225700855255, + "step": 1630 + }, + { + "epoch": 3.4503171247357294, + "grad_norm": 1.39235258102417, + "learning_rate": 1.4603826310163874e-06, + "loss": 0.5391915440559387, + "step": 1632 + }, + { + "epoch": 3.4545454545454546, + "grad_norm": 0.9247079491615295, + "learning_rate": 1.453474509528593e-06, + "loss": 0.5066539645195007, + "step": 1634 + }, + { + "epoch": 3.45877378435518, + "grad_norm": 1.8016786575317383, + "learning_rate": 1.446615858440642e-06, + "loss": 0.7724549770355225, + "step": 1636 + }, + { + "epoch": 3.463002114164905, + "grad_norm": 0.17383824288845062, + "learning_rate": 1.4398067616022993e-06, + "loss": 0.2838648855686188, + "step": 1638 + }, + { + "epoch": 3.46723044397463, + "grad_norm": 0.8914324641227722, + "learning_rate": 1.43304730225751e-06, + "loss": 0.44119375944137573, + "step": 1640 + }, + { + "epoch": 3.471458773784355, + "grad_norm": 0.41787609457969666, + "learning_rate": 1.4263375630433802e-06, + "loss": 0.12083140760660172, + "step": 1642 + }, + { + "epoch": 3.47568710359408, + "grad_norm": 1.2530478239059448, + "learning_rate": 1.4196776259891665e-06, + "loss": 0.7062510848045349, + "step": 1644 + }, + { + "epoch": 3.4799154334038054, + "grad_norm": 1.6841150522232056, + "learning_rate": 1.4130675725152762e-06, + "loss": 0.691648006439209, + "step": 1646 + }, + { + "epoch": 3.4841437632135306, + "grad_norm": 0.22874891757965088, + "learning_rate": 1.4065074834322677e-06, + "loss": 0.43919339776039124, + "step": 1648 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 1.317173719406128, + "learning_rate": 1.3999974389398669e-06, + "loss": 0.38067111372947693, + "step": 1650 + }, + { + "epoch": 3.492600422832981, + "grad_norm": 5.007179260253906, + "learning_rate": 1.3935375186259817e-06, + "loss": 0.5524361729621887, + "step": 1652 + }, + { + "epoch": 3.496828752642706, + "grad_norm": 2.7014377117156982, + "learning_rate": 1.3871278014657332e-06, + "loss": 0.6528635025024414, + "step": 1654 + }, + { + "epoch": 3.5010570824524314, + "grad_norm": 1.3773154020309448, + "learning_rate": 1.3807683658204915e-06, + "loss": 0.818991482257843, + "step": 1656 + }, + { + "epoch": 3.5052854122621566, + "grad_norm": 0.31587138772010803, + "learning_rate": 1.3744592894369093e-06, + "loss": 0.4350362718105316, + "step": 1658 + }, + { + "epoch": 3.5095137420718814, + "grad_norm": 2.6180124282836914, + "learning_rate": 1.3682006494459806e-06, + "loss": 0.5732263922691345, + "step": 1660 + }, + { + "epoch": 3.513742071881607, + "grad_norm": 2.136174201965332, + "learning_rate": 1.361992522362096e-06, + "loss": 0.46314525604248047, + "step": 1662 + }, + { + "epoch": 3.517970401691332, + "grad_norm": 0.8930491209030151, + "learning_rate": 1.355834984082101e-06, + "loss": 0.7437731027603149, + "step": 1664 + }, + { + "epoch": 3.522198731501057, + "grad_norm": 2.2886319160461426, + "learning_rate": 1.3497281098843746e-06, + "loss": 0.6952565908432007, + "step": 1666 + }, + { + "epoch": 3.526427061310782, + "grad_norm": 0.17518752813339233, + "learning_rate": 1.34367197442791e-06, + "loss": 0.08985922485589981, + "step": 1668 + }, + { + "epoch": 3.5306553911205074, + "grad_norm": 0.5296573042869568, + "learning_rate": 1.3376666517513931e-06, + "loss": 0.5701056718826294, + "step": 1670 + }, + { + "epoch": 3.5348837209302326, + "grad_norm": 1.2680672407150269, + "learning_rate": 1.3317122152723078e-06, + "loss": 0.41330066323280334, + "step": 1672 + }, + { + "epoch": 3.539112050739958, + "grad_norm": 1.1988015174865723, + "learning_rate": 1.3258087377860323e-06, + "loss": 0.7451457977294922, + "step": 1674 + }, + { + "epoch": 3.543340380549683, + "grad_norm": 1.5698060989379883, + "learning_rate": 1.3199562914649519e-06, + "loss": 0.41681385040283203, + "step": 1676 + }, + { + "epoch": 3.547568710359408, + "grad_norm": 0.8924346566200256, + "learning_rate": 1.314154947857571e-06, + "loss": 0.7350165843963623, + "step": 1678 + }, + { + "epoch": 3.5517970401691334, + "grad_norm": 0.8909021019935608, + "learning_rate": 1.3084047778876487e-06, + "loss": 0.3733225166797638, + "step": 1680 + }, + { + "epoch": 3.556025369978858, + "grad_norm": 1.1042284965515137, + "learning_rate": 1.3027058518533228e-06, + "loss": 0.4569874107837677, + "step": 1682 + }, + { + "epoch": 3.5602536997885834, + "grad_norm": 1.6617507934570312, + "learning_rate": 1.2970582394262516e-06, + "loss": 0.45771390199661255, + "step": 1684 + }, + { + "epoch": 3.5644820295983086, + "grad_norm": 2.2305591106414795, + "learning_rate": 1.291462009650768e-06, + "loss": 0.13002893328666687, + "step": 1686 + }, + { + "epoch": 3.568710359408034, + "grad_norm": 1.593942642211914, + "learning_rate": 1.2859172309430285e-06, + "loss": 0.5540592670440674, + "step": 1688 + }, + { + "epoch": 3.572938689217759, + "grad_norm": 0.1893344223499298, + "learning_rate": 1.280423971090181e-06, + "loss": 0.15982471406459808, + "step": 1690 + }, + { + "epoch": 3.5771670190274842, + "grad_norm": 3.0836663246154785, + "learning_rate": 1.2749822972495333e-06, + "loss": 0.30629515647888184, + "step": 1692 + }, + { + "epoch": 3.5813953488372094, + "grad_norm": 19.792110443115234, + "learning_rate": 1.2695922759477356e-06, + "loss": 0.5756056308746338, + "step": 1694 + }, + { + "epoch": 3.585623678646934, + "grad_norm": 2.6506264209747314, + "learning_rate": 1.2642539730799622e-06, + "loss": 0.27927032113075256, + "step": 1696 + }, + { + "epoch": 3.58985200845666, + "grad_norm": 0.9805445075035095, + "learning_rate": 1.2589674539091115e-06, + "loss": 0.700019121170044, + "step": 1698 + }, + { + "epoch": 3.5940803382663846, + "grad_norm": 1.2981892824172974, + "learning_rate": 1.2537327830650042e-06, + "loss": 0.8670184016227722, + "step": 1700 + }, + { + "epoch": 3.59830866807611, + "grad_norm": 2.947838068008423, + "learning_rate": 1.248550024543594e-06, + "loss": 0.7257493734359741, + "step": 1702 + }, + { + "epoch": 3.602536997885835, + "grad_norm": 2.9678969383239746, + "learning_rate": 1.2434192417061862e-06, + "loss": 0.2969245910644531, + "step": 1704 + }, + { + "epoch": 3.6067653276955602, + "grad_norm": 1.1396725177764893, + "learning_rate": 1.2383404972786625e-06, + "loss": 0.43581053614616394, + "step": 1706 + }, + { + "epoch": 3.6109936575052854, + "grad_norm": 33.62403106689453, + "learning_rate": 1.2333138533507139e-06, + "loss": 0.6185458302497864, + "step": 1708 + }, + { + "epoch": 3.6152219873150107, + "grad_norm": 2.351975202560425, + "learning_rate": 1.2283393713750816e-06, + "loss": 0.5312186479568481, + "step": 1710 + }, + { + "epoch": 3.619450317124736, + "grad_norm": 1.5491169691085815, + "learning_rate": 1.2234171121668069e-06, + "loss": 0.4756861925125122, + "step": 1712 + }, + { + "epoch": 3.6236786469344606, + "grad_norm": 0.2598097622394562, + "learning_rate": 1.2185471359024855e-06, + "loss": 0.24854713678359985, + "step": 1714 + }, + { + "epoch": 3.6279069767441863, + "grad_norm": 4.190101146697998, + "learning_rate": 1.2137295021195341e-06, + "loss": 0.6819196343421936, + "step": 1716 + }, + { + "epoch": 3.632135306553911, + "grad_norm": 0.1816449761390686, + "learning_rate": 1.2089642697154613e-06, + "loss": 0.39375177025794983, + "step": 1718 + }, + { + "epoch": 3.6363636363636362, + "grad_norm": 2.7164671421051025, + "learning_rate": 1.2042514969471475e-06, + "loss": 0.3496692180633545, + "step": 1720 + }, + { + "epoch": 3.6405919661733614, + "grad_norm": 2.0159873962402344, + "learning_rate": 1.1995912414301329e-06, + "loss": 0.6999395489692688, + "step": 1722 + }, + { + "epoch": 3.6448202959830867, + "grad_norm": 1.0355256795883179, + "learning_rate": 1.1949835601379144e-06, + "loss": 0.5229198336601257, + "step": 1724 + }, + { + "epoch": 3.649048625792812, + "grad_norm": 1.0544854402542114, + "learning_rate": 1.1904285094012463e-06, + "loss": 0.7571031451225281, + "step": 1726 + }, + { + "epoch": 3.653276955602537, + "grad_norm": 1.3050400018692017, + "learning_rate": 1.185926144907454e-06, + "loss": 0.8300658464431763, + "step": 1728 + }, + { + "epoch": 3.6575052854122623, + "grad_norm": 61.86153030395508, + "learning_rate": 1.1814765216997522e-06, + "loss": 0.6350105404853821, + "step": 1730 + }, + { + "epoch": 3.6617336152219875, + "grad_norm": 1.3892080783843994, + "learning_rate": 1.1770796941765727e-06, + "loss": 0.7083321809768677, + "step": 1732 + }, + { + "epoch": 3.6659619450317127, + "grad_norm": 1.7146962881088257, + "learning_rate": 1.1727357160908978e-06, + "loss": 0.5864725112915039, + "step": 1734 + }, + { + "epoch": 3.6701902748414374, + "grad_norm": 2.1355414390563965, + "learning_rate": 1.1684446405496052e-06, + "loss": 0.7335898876190186, + "step": 1736 + }, + { + "epoch": 3.6744186046511627, + "grad_norm": 8.273016929626465, + "learning_rate": 1.1642065200128179e-06, + "loss": 0.11255266517400742, + "step": 1738 + }, + { + "epoch": 3.678646934460888, + "grad_norm": 1.5386680364608765, + "learning_rate": 1.1600214062932613e-06, + "loss": 0.4220409691333771, + "step": 1740 + }, + { + "epoch": 3.682875264270613, + "grad_norm": 1.507468342781067, + "learning_rate": 1.1558893505556315e-06, + "loss": 0.2292194664478302, + "step": 1742 + }, + { + "epoch": 3.6871035940803383, + "grad_norm": 1.0740567445755005, + "learning_rate": 1.1518104033159715e-06, + "loss": 0.8184658288955688, + "step": 1744 + }, + { + "epoch": 3.6913319238900635, + "grad_norm": 0.20174740254878998, + "learning_rate": 1.1477846144410493e-06, + "loss": 0.3239232897758484, + "step": 1746 + }, + { + "epoch": 3.6955602536997887, + "grad_norm": 3.3497393131256104, + "learning_rate": 1.1438120331477506e-06, + "loss": 0.3809773325920105, + "step": 1748 + }, + { + "epoch": 3.699788583509514, + "grad_norm": 2.0099411010742188, + "learning_rate": 1.1398927080024796e-06, + "loss": 0.8483526706695557, + "step": 1750 + }, + { + "epoch": 3.704016913319239, + "grad_norm": 2.228614091873169, + "learning_rate": 1.1360266869205593e-06, + "loss": 0.27999040484428406, + "step": 1752 + }, + { + "epoch": 3.708245243128964, + "grad_norm": 2.1806023120880127, + "learning_rate": 1.1322140171656507e-06, + "loss": 0.5907219052314758, + "step": 1754 + }, + { + "epoch": 3.712473572938689, + "grad_norm": 1.0620449781417847, + "learning_rate": 1.1284547453491734e-06, + "loss": 0.5546761155128479, + "step": 1756 + }, + { + "epoch": 3.7167019027484143, + "grad_norm": 3.0138957500457764, + "learning_rate": 1.1247489174297372e-06, + "loss": 0.4685841202735901, + "step": 1758 + }, + { + "epoch": 3.7209302325581395, + "grad_norm": 0.2721787691116333, + "learning_rate": 1.1210965787125743e-06, + "loss": 0.32701820135116577, + "step": 1760 + }, + { + "epoch": 3.7251585623678647, + "grad_norm": 1.143666386604309, + "learning_rate": 1.1174977738489955e-06, + "loss": 0.645359992980957, + "step": 1762 + }, + { + "epoch": 3.72938689217759, + "grad_norm": 5.229813098907471, + "learning_rate": 1.113952546835835e-06, + "loss": 0.42560404539108276, + "step": 1764 + }, + { + "epoch": 3.733615221987315, + "grad_norm": 1.8011236190795898, + "learning_rate": 1.1104609410149174e-06, + "loss": 0.6710249185562134, + "step": 1766 + }, + { + "epoch": 3.7378435517970403, + "grad_norm": 1.6186012029647827, + "learning_rate": 1.1070229990725262e-06, + "loss": 0.7012709379196167, + "step": 1768 + }, + { + "epoch": 3.7420718816067655, + "grad_norm": 0.5522105097770691, + "learning_rate": 1.1036387630388841e-06, + "loss": 0.14224903285503387, + "step": 1770 + }, + { + "epoch": 3.7463002114164903, + "grad_norm": 1.2747918367385864, + "learning_rate": 1.100308274287636e-06, + "loss": 0.39697420597076416, + "step": 1772 + }, + { + "epoch": 3.7505285412262155, + "grad_norm": 3.56330943107605, + "learning_rate": 1.0970315735353434e-06, + "loss": 0.8253507614135742, + "step": 1774 + }, + { + "epoch": 3.7547568710359407, + "grad_norm": 1.9991868734359741, + "learning_rate": 1.0938087008409909e-06, + "loss": 0.738379716873169, + "step": 1776 + }, + { + "epoch": 3.758985200845666, + "grad_norm": 3.5614612102508545, + "learning_rate": 1.0906396956054919e-06, + "loss": 0.7684906721115112, + "step": 1778 + }, + { + "epoch": 3.763213530655391, + "grad_norm": 1.3542511463165283, + "learning_rate": 1.0875245965712077e-06, + "loss": 0.9081143736839294, + "step": 1780 + }, + { + "epoch": 3.7674418604651163, + "grad_norm": 2.4983162879943848, + "learning_rate": 1.0844634418214763e-06, + "loss": 0.7231796383857727, + "step": 1782 + }, + { + "epoch": 3.7716701902748415, + "grad_norm": 3.263745069503784, + "learning_rate": 1.0814562687801448e-06, + "loss": 0.11066766083240509, + "step": 1784 + }, + { + "epoch": 3.7758985200845667, + "grad_norm": 1.9571565389633179, + "learning_rate": 1.0785031142111103e-06, + "loss": 0.5418827533721924, + "step": 1786 + }, + { + "epoch": 3.780126849894292, + "grad_norm": 4.58798885345459, + "learning_rate": 1.075604014217875e-06, + "loss": 0.45577073097229004, + "step": 1788 + }, + { + "epoch": 3.7843551797040167, + "grad_norm": 2.979398727416992, + "learning_rate": 1.0727590042431009e-06, + "loss": 0.2022068351507187, + "step": 1790 + }, + { + "epoch": 3.7885835095137423, + "grad_norm": 0.2828064560890198, + "learning_rate": 1.0699681190681777e-06, + "loss": 0.6389042139053345, + "step": 1792 + }, + { + "epoch": 3.792811839323467, + "grad_norm": 1.1751878261566162, + "learning_rate": 1.067231392812798e-06, + "loss": 0.67657470703125, + "step": 1794 + }, + { + "epoch": 3.7970401691331923, + "grad_norm": 25.275117874145508, + "learning_rate": 1.0645488589345396e-06, + "loss": 0.41407984495162964, + "step": 1796 + }, + { + "epoch": 3.8012684989429175, + "grad_norm": 7.4821457862854, + "learning_rate": 1.0619205502284574e-06, + "loss": 0.7293649315834045, + "step": 1798 + }, + { + "epoch": 3.8054968287526427, + "grad_norm": 1.6770122051239014, + "learning_rate": 1.0593464988266811e-06, + "loss": 0.781909704208374, + "step": 1800 + }, + { + "epoch": 3.809725158562368, + "grad_norm": 1.3441821336746216, + "learning_rate": 1.0568267361980225e-06, + "loss": 0.440990149974823, + "step": 1802 + }, + { + "epoch": 3.813953488372093, + "grad_norm": 1.1365444660186768, + "learning_rate": 1.0543612931475935e-06, + "loss": 0.700577437877655, + "step": 1804 + }, + { + "epoch": 3.8181818181818183, + "grad_norm": 1.5654172897338867, + "learning_rate": 1.0519501998164247e-06, + "loss": 0.8008032441139221, + "step": 1806 + }, + { + "epoch": 3.822410147991543, + "grad_norm": 4.419487476348877, + "learning_rate": 1.049593485681101e-06, + "loss": 0.16393083333969116, + "step": 1808 + }, + { + "epoch": 3.8266384778012688, + "grad_norm": 8.012113571166992, + "learning_rate": 1.0472911795533997e-06, + "loss": 0.7106217741966248, + "step": 1810 + }, + { + "epoch": 3.8308668076109935, + "grad_norm": 0.9460861086845398, + "learning_rate": 1.045043309579938e-06, + "loss": 0.86064213514328, + "step": 1812 + }, + { + "epoch": 3.8350951374207187, + "grad_norm": 1.2017807960510254, + "learning_rate": 1.0428499032418297e-06, + "loss": 0.6377520561218262, + "step": 1814 + }, + { + "epoch": 3.839323467230444, + "grad_norm": 1.4761308431625366, + "learning_rate": 1.0407109873543484e-06, + "loss": 0.4453623592853546, + "step": 1816 + }, + { + "epoch": 3.843551797040169, + "grad_norm": 3.0813045501708984, + "learning_rate": 1.0386265880665996e-06, + "loss": 0.46757772564888, + "step": 1818 + }, + { + "epoch": 3.8477801268498943, + "grad_norm": 1.3441790342330933, + "learning_rate": 1.036596730861203e-06, + "loss": 0.3905222713947296, + "step": 1820 + }, + { + "epoch": 3.8520084566596196, + "grad_norm": 0.276379257440567, + "learning_rate": 1.0346214405539783e-06, + "loss": 0.6184788346290588, + "step": 1822 + }, + { + "epoch": 3.8562367864693448, + "grad_norm": 2.5074493885040283, + "learning_rate": 1.0327007412936419e-06, + "loss": 0.5457111597061157, + "step": 1824 + }, + { + "epoch": 3.8604651162790695, + "grad_norm": 4.493834972381592, + "learning_rate": 1.0308346565615158e-06, + "loss": 0.4523421823978424, + "step": 1826 + }, + { + "epoch": 3.864693446088795, + "grad_norm": 0.907687246799469, + "learning_rate": 1.0290232091712337e-06, + "loss": 0.8290807604789734, + "step": 1828 + }, + { + "epoch": 3.86892177589852, + "grad_norm": 1.0880918502807617, + "learning_rate": 1.0272664212684687e-06, + "loss": 0.761888325214386, + "step": 1830 + }, + { + "epoch": 3.873150105708245, + "grad_norm": 1.2926299571990967, + "learning_rate": 1.025564314330659e-06, + "loss": 0.4618053138256073, + "step": 1832 + }, + { + "epoch": 3.8773784355179703, + "grad_norm": 0.9145944118499756, + "learning_rate": 1.0239169091667447e-06, + "loss": 0.4251665771007538, + "step": 1834 + }, + { + "epoch": 3.8816067653276956, + "grad_norm": 0.30925077199935913, + "learning_rate": 1.022324225916917e-06, + "loss": 0.38876229524612427, + "step": 1836 + }, + { + "epoch": 3.8858350951374208, + "grad_norm": 1.2880512475967407, + "learning_rate": 1.0207862840523672e-06, + "loss": 0.3642764985561371, + "step": 1838 + }, + { + "epoch": 3.890063424947146, + "grad_norm": 1.0917878150939941, + "learning_rate": 1.0193031023750534e-06, + "loss": 0.7018687129020691, + "step": 1840 + }, + { + "epoch": 3.894291754756871, + "grad_norm": 1.4084701538085938, + "learning_rate": 1.017874699017468e-06, + "loss": 0.6713329553604126, + "step": 1842 + }, + { + "epoch": 3.898520084566596, + "grad_norm": 1.1642215251922607, + "learning_rate": 1.0165010914424142e-06, + "loss": 0.5418899059295654, + "step": 1844 + }, + { + "epoch": 3.9027484143763216, + "grad_norm": 8.336616516113281, + "learning_rate": 1.015182296442798e-06, + "loss": 0.5039263367652893, + "step": 1846 + }, + { + "epoch": 3.9069767441860463, + "grad_norm": 2.7992241382598877, + "learning_rate": 1.0139183301414176e-06, + "loss": 0.10513779520988464, + "step": 1848 + }, + { + "epoch": 3.9112050739957716, + "grad_norm": 1.2841367721557617, + "learning_rate": 1.0127092079907682e-06, + "loss": 0.6863033175468445, + "step": 1850 + }, + { + "epoch": 3.9154334038054968, + "grad_norm": 0.8896276950836182, + "learning_rate": 1.0115549447728551e-06, + "loss": 0.548151433467865, + "step": 1852 + }, + { + "epoch": 3.919661733615222, + "grad_norm": 0.9178200960159302, + "learning_rate": 1.0104555545990088e-06, + "loss": 0.6384239196777344, + "step": 1854 + }, + { + "epoch": 3.923890063424947, + "grad_norm": 1.911726713180542, + "learning_rate": 1.009411050909716e-06, + "loss": 0.3711985945701599, + "step": 1856 + }, + { + "epoch": 3.9281183932346724, + "grad_norm": 3.756681442260742, + "learning_rate": 1.008421446474455e-06, + "loss": 0.6350414752960205, + "step": 1858 + }, + { + "epoch": 3.9323467230443976, + "grad_norm": 9.237351417541504, + "learning_rate": 1.0074867533915367e-06, + "loss": 0.27239492535591125, + "step": 1860 + }, + { + "epoch": 3.9365750528541223, + "grad_norm": 2.622500419616699, + "learning_rate": 1.0066069830879593e-06, + "loss": 0.749444842338562, + "step": 1862 + }, + { + "epoch": 3.940803382663848, + "grad_norm": 1.1784473657608032, + "learning_rate": 1.0057821463192687e-06, + "loss": 0.7133820056915283, + "step": 1864 + }, + { + "epoch": 3.9450317124735728, + "grad_norm": 0.9698504209518433, + "learning_rate": 1.005012253169426e-06, + "loss": 0.6835721135139465, + "step": 1866 + }, + { + "epoch": 3.949260042283298, + "grad_norm": 0.07213737815618515, + "learning_rate": 1.0042973130506839e-06, + "loss": 0.0018214905867353082, + "step": 1868 + }, + { + "epoch": 3.953488372093023, + "grad_norm": 1.0467723608016968, + "learning_rate": 1.003637334703472e-06, + "loss": 0.7706689238548279, + "step": 1870 + }, + { + "epoch": 3.9577167019027484, + "grad_norm": 1.0635846853256226, + "learning_rate": 1.0030323261962922e-06, + "loss": 0.5568016767501831, + "step": 1872 + }, + { + "epoch": 3.9619450317124736, + "grad_norm": 3.191056489944458, + "learning_rate": 1.002482294925615e-06, + "loss": 0.41130486130714417, + "step": 1874 + }, + { + "epoch": 3.966173361522199, + "grad_norm": 1.0538538694381714, + "learning_rate": 1.0019872476157952e-06, + "loss": 0.7130852341651917, + "step": 1876 + }, + { + "epoch": 3.970401691331924, + "grad_norm": 5.397729396820068, + "learning_rate": 1.0015471903189834e-06, + "loss": 0.2227753847837448, + "step": 1878 + }, + { + "epoch": 3.974630021141649, + "grad_norm": 1.2139853239059448, + "learning_rate": 1.0011621284150583e-06, + "loss": 0.348341166973114, + "step": 1880 + }, + { + "epoch": 3.9788583509513744, + "grad_norm": 2.312925100326538, + "learning_rate": 1.000832066611556e-06, + "loss": 0.5929352641105652, + "step": 1882 + }, + { + "epoch": 3.983086680761099, + "grad_norm": 1.4073094129562378, + "learning_rate": 1.0005570089436137e-06, + "loss": 0.8656057119369507, + "step": 1884 + }, + { + "epoch": 3.9873150105708244, + "grad_norm": 1.118908405303955, + "learning_rate": 1.0003369587739234e-06, + "loss": 0.5605166554450989, + "step": 1886 + }, + { + "epoch": 3.9915433403805496, + "grad_norm": 1.562491774559021, + "learning_rate": 1.0001719187926854e-06, + "loss": 0.7328466176986694, + "step": 1888 + }, + { + "epoch": 3.995771670190275, + "grad_norm": 1.2813456058502197, + "learning_rate": 1.0000618910175811e-06, + "loss": 0.7010554671287537, + "step": 1890 + }, + { + "epoch": 4.0, + "grad_norm": 1.1999653577804565, + "learning_rate": 1.000006876793743e-06, + "loss": 0.36771833896636963, + "step": 1892 + }, + { + "epoch": 4.0, + "step": 1892, + "total_flos": 3.5845266747435254e+18, + "train_loss": 0.828627192888465, + "train_runtime": 8553.1396, + "train_samples_per_second": 6.636, + "train_steps_per_second": 0.221 + } + ], + "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 +}