diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,6665 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1892, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.004228329809725159, + "grad_norm": 3.061702013015747, + "learning_rate": 1.0526315789473685e-07, + "loss": 1.8276734352111816, + "step": 2 + }, + { + "epoch": 0.008456659619450317, + "grad_norm": 1.3492687940597534, + "learning_rate": 3.1578947368421055e-07, + "loss": 2.2111785411834717, + "step": 4 + }, + { + "epoch": 0.012684989429175475, + "grad_norm": 5.953800201416016, + "learning_rate": 5.263157894736843e-07, + "loss": 2.202695846557617, + "step": 6 + }, + { + "epoch": 0.016913319238900635, + "grad_norm": 1.8159960508346558, + "learning_rate": 7.368421052631579e-07, + "loss": 1.9954826831817627, + "step": 8 + }, + { + "epoch": 0.021141649048625793, + "grad_norm": 17.191267013549805, + "learning_rate": 9.473684210526317e-07, + "loss": 2.010864496231079, + "step": 10 + }, + { + "epoch": 0.02536997885835095, + "grad_norm": 1.6861594915390015, + "learning_rate": 1.1578947368421053e-06, + "loss": 1.6342254877090454, + "step": 12 + }, + { + "epoch": 0.02959830866807611, + "grad_norm": 1.4505009651184082, + "learning_rate": 1.3684210526315791e-06, + "loss": 1.6408352851867676, + "step": 14 + }, + { + "epoch": 0.03382663847780127, + "grad_norm": 1.612177848815918, + "learning_rate": 1.5789473684210526e-06, + "loss": 1.7670618295669556, + "step": 16 + }, + { + "epoch": 0.03805496828752643, + "grad_norm": 1.3589969873428345, + "learning_rate": 1.7894736842105265e-06, + "loss": 1.7907778024673462, + "step": 18 + }, + { + "epoch": 0.042283298097251586, + "grad_norm": 4.852035999298096, + "learning_rate": 2.0000000000000003e-06, + "loss": 1.697686791419983, + "step": 20 + }, + { + "epoch": 0.046511627906976744, + "grad_norm": 3.9681975841522217, + "learning_rate": 2.2105263157894738e-06, + "loss": 1.5277243852615356, + "step": 22 + }, + { + "epoch": 0.0507399577167019, + "grad_norm": 2.176203966140747, + "learning_rate": 2.4210526315789477e-06, + "loss": 1.76160728931427, + "step": 24 + }, + { + "epoch": 0.05496828752642706, + "grad_norm": 2.0155324935913086, + "learning_rate": 2.631578947368421e-06, + "loss": 1.3579041957855225, + "step": 26 + }, + { + "epoch": 0.05919661733615222, + "grad_norm": 0.8821945190429688, + "learning_rate": 2.842105263157895e-06, + "loss": 1.6458429098129272, + "step": 28 + }, + { + "epoch": 0.06342494714587738, + "grad_norm": 1.1895365715026855, + "learning_rate": 3.052631578947369e-06, + "loss": 1.2144521474838257, + "step": 30 + }, + { + "epoch": 0.06765327695560254, + "grad_norm": 4.421765327453613, + "learning_rate": 3.2631578947368423e-06, + "loss": 0.9850426316261292, + "step": 32 + }, + { + "epoch": 0.07188160676532769, + "grad_norm": 1.656004548072815, + "learning_rate": 3.473684210526316e-06, + "loss": 1.1624037027359009, + "step": 34 + }, + { + "epoch": 0.07610993657505286, + "grad_norm": 1.4637020826339722, + "learning_rate": 3.6842105263157896e-06, + "loss": 1.3481690883636475, + "step": 36 + }, + { + "epoch": 0.080338266384778, + "grad_norm": 2.7491376399993896, + "learning_rate": 3.894736842105263e-06, + "loss": 1.2821086645126343, + "step": 38 + }, + { + "epoch": 0.08456659619450317, + "grad_norm": 5.14992094039917, + "learning_rate": 4.105263157894737e-06, + "loss": 1.2038065195083618, + "step": 40 + }, + { + "epoch": 0.08879492600422834, + "grad_norm": 2.3842504024505615, + "learning_rate": 4.315789473684211e-06, + "loss": 1.6893579959869385, + "step": 42 + }, + { + "epoch": 0.09302325581395349, + "grad_norm": 2.7394402027130127, + "learning_rate": 4.526315789473685e-06, + "loss": 1.5890274047851562, + "step": 44 + }, + { + "epoch": 0.09725158562367865, + "grad_norm": 1.237416386604309, + "learning_rate": 4.736842105263158e-06, + "loss": 1.2647711038589478, + "step": 46 + }, + { + "epoch": 0.1014799154334038, + "grad_norm": 0.8126721382141113, + "learning_rate": 4.947368421052632e-06, + "loss": 1.4983834028244019, + "step": 48 + }, + { + "epoch": 0.10570824524312897, + "grad_norm": 2.277120351791382, + "learning_rate": 5.157894736842106e-06, + "loss": 1.4656639099121094, + "step": 50 + }, + { + "epoch": 0.10993657505285412, + "grad_norm": 2.99977707862854, + "learning_rate": 5.36842105263158e-06, + "loss": 0.6993181109428406, + "step": 52 + }, + { + "epoch": 0.11416490486257928, + "grad_norm": 0.7835460305213928, + "learning_rate": 5.578947368421052e-06, + "loss": 1.0457398891448975, + "step": 54 + }, + { + "epoch": 0.11839323467230443, + "grad_norm": 1.7906148433685303, + "learning_rate": 5.789473684210527e-06, + "loss": 1.3629032373428345, + "step": 56 + }, + { + "epoch": 0.1226215644820296, + "grad_norm": 0.9121655821800232, + "learning_rate": 6e-06, + "loss": 1.3880715370178223, + "step": 58 + }, + { + "epoch": 0.12684989429175475, + "grad_norm": 1.2699471712112427, + "learning_rate": 6.2105263157894745e-06, + "loss": 1.4011625051498413, + "step": 60 + }, + { + "epoch": 0.13107822410147993, + "grad_norm": 3.4582390785217285, + "learning_rate": 6.421052631578948e-06, + "loss": 0.7704225182533264, + "step": 62 + }, + { + "epoch": 0.13530655391120508, + "grad_norm": 2.431262254714966, + "learning_rate": 6.631578947368421e-06, + "loss": 1.047868013381958, + "step": 64 + }, + { + "epoch": 0.13953488372093023, + "grad_norm": 1.2116672992706299, + "learning_rate": 6.842105263157896e-06, + "loss": 1.3387612104415894, + "step": 66 + }, + { + "epoch": 0.14376321353065538, + "grad_norm": 2.2332615852355957, + "learning_rate": 7.052631578947369e-06, + "loss": 1.374855875968933, + "step": 68 + }, + { + "epoch": 0.14799154334038056, + "grad_norm": 1.4368053674697876, + "learning_rate": 7.263157894736843e-06, + "loss": 1.3531306982040405, + "step": 70 + }, + { + "epoch": 0.1522198731501057, + "grad_norm": 0.9301667809486389, + "learning_rate": 7.473684210526316e-06, + "loss": 1.344086766242981, + "step": 72 + }, + { + "epoch": 0.15644820295983086, + "grad_norm": 1.1713248491287231, + "learning_rate": 7.68421052631579e-06, + "loss": 0.931689441204071, + "step": 74 + }, + { + "epoch": 0.160676532769556, + "grad_norm": 1.1938786506652832, + "learning_rate": 7.894736842105265e-06, + "loss": 0.8424839377403259, + "step": 76 + }, + { + "epoch": 0.1649048625792812, + "grad_norm": 2.048682451248169, + "learning_rate": 8.105263157894736e-06, + "loss": 1.2822306156158447, + "step": 78 + }, + { + "epoch": 0.16913319238900634, + "grad_norm": 2.50176739692688, + "learning_rate": 8.315789473684212e-06, + "loss": 1.0914721488952637, + "step": 80 + }, + { + "epoch": 0.1733615221987315, + "grad_norm": 1.033888578414917, + "learning_rate": 8.526315789473685e-06, + "loss": 1.0837591886520386, + "step": 82 + }, + { + "epoch": 0.17758985200845667, + "grad_norm": 1.0602306127548218, + "learning_rate": 8.736842105263158e-06, + "loss": 0.8805471658706665, + "step": 84 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.8846888542175293, + "learning_rate": 8.947368421052632e-06, + "loss": 1.7070754766464233, + "step": 86 + }, + { + "epoch": 0.18604651162790697, + "grad_norm": 1.389636516571045, + "learning_rate": 9.157894736842105e-06, + "loss": 1.3066718578338623, + "step": 88 + }, + { + "epoch": 0.19027484143763213, + "grad_norm": 1.4127076864242554, + "learning_rate": 9.36842105263158e-06, + "loss": 0.9866373538970947, + "step": 90 + }, + { + "epoch": 0.1945031712473573, + "grad_norm": 1.7008750438690186, + "learning_rate": 9.578947368421054e-06, + "loss": 1.3172839879989624, + "step": 92 + }, + { + "epoch": 0.19873150105708245, + "grad_norm": 1.2962291240692139, + "learning_rate": 9.789473684210527e-06, + "loss": 0.9815294742584229, + "step": 94 + }, + { + "epoch": 0.2029598308668076, + "grad_norm": 1.9650954008102417, + "learning_rate": 1e-05, + "loss": 1.2962567806243896, + "step": 96 + }, + { + "epoch": 0.20718816067653276, + "grad_norm": 1.64167320728302, + "learning_rate": 9.999972492846047e-06, + "loss": 1.388351321220398, + "step": 98 + }, + { + "epoch": 0.21141649048625794, + "grad_norm": 0.951882541179657, + "learning_rate": 9.999889971720468e-06, + "loss": 1.328765869140625, + "step": 100 + }, + { + "epoch": 0.2156448202959831, + "grad_norm": 2.4114439487457275, + "learning_rate": 9.999752437632121e-06, + "loss": 1.2472662925720215, + "step": 102 + }, + { + "epoch": 0.21987315010570824, + "grad_norm": 1.0009742975234985, + "learning_rate": 9.999559892262412e-06, + "loss": 1.3068406581878662, + "step": 104 + }, + { + "epoch": 0.22410147991543342, + "grad_norm": 1.6812602281570435, + "learning_rate": 9.999312337965291e-06, + "loss": 1.2981398105621338, + "step": 106 + }, + { + "epoch": 0.22832980972515857, + "grad_norm": 1.4475797414779663, + "learning_rate": 9.999009777767202e-06, + "loss": 1.0255334377288818, + "step": 108 + }, + { + "epoch": 0.23255813953488372, + "grad_norm": 1.2929273843765259, + "learning_rate": 9.99865221536707e-06, + "loss": 1.2872642278671265, + "step": 110 + }, + { + "epoch": 0.23678646934460887, + "grad_norm": 1.9674358367919922, + "learning_rate": 9.998239655136237e-06, + "loss": 1.3555434942245483, + "step": 112 + }, + { + "epoch": 0.24101479915433405, + "grad_norm": 1.1659306287765503, + "learning_rate": 9.997772102118417e-06, + "loss": 1.5545711517333984, + "step": 114 + }, + { + "epoch": 0.2452431289640592, + "grad_norm": 1.0404155254364014, + "learning_rate": 9.997249562029637e-06, + "loss": 1.3228392601013184, + "step": 116 + }, + { + "epoch": 0.24947145877378435, + "grad_norm": 2.40277361869812, + "learning_rate": 9.996672041258154e-06, + "loss": 1.3454302549362183, + "step": 118 + }, + { + "epoch": 0.2536997885835095, + "grad_norm": 2.02721905708313, + "learning_rate": 9.996039546864395e-06, + "loss": 0.845862627029419, + "step": 120 + }, + { + "epoch": 0.25792811839323465, + "grad_norm": 0.8644340634346008, + "learning_rate": 9.995352086580858e-06, + "loss": 1.2982769012451172, + "step": 122 + }, + { + "epoch": 0.26215644820295986, + "grad_norm": 0.8157991766929626, + "learning_rate": 9.994609668812022e-06, + "loss": 1.1449761390686035, + "step": 124 + }, + { + "epoch": 0.266384778012685, + "grad_norm": 2.311751365661621, + "learning_rate": 9.993812302634237e-06, + "loss": 1.3440643548965454, + "step": 126 + }, + { + "epoch": 0.27061310782241016, + "grad_norm": 2.628192901611328, + "learning_rate": 9.99295999779563e-06, + "loss": 0.9439712762832642, + "step": 128 + }, + { + "epoch": 0.2748414376321353, + "grad_norm": 1.76107656955719, + "learning_rate": 9.99205276471597e-06, + "loss": 0.9737116694450378, + "step": 130 + }, + { + "epoch": 0.27906976744186046, + "grad_norm": 1.3215941190719604, + "learning_rate": 9.991090614486546e-06, + "loss": 1.2664296627044678, + "step": 132 + }, + { + "epoch": 0.2832980972515856, + "grad_norm": 1.1459059715270996, + "learning_rate": 9.990073558870028e-06, + "loss": 0.8418352007865906, + "step": 134 + }, + { + "epoch": 0.28752642706131076, + "grad_norm": 8.8709077835083, + "learning_rate": 9.989001610300333e-06, + "loss": 1.111986517906189, + "step": 136 + }, + { + "epoch": 0.2917547568710359, + "grad_norm": 1.0686612129211426, + "learning_rate": 9.987874781882461e-06, + "loss": 1.3773900270462036, + "step": 138 + }, + { + "epoch": 0.2959830866807611, + "grad_norm": 0.9512626528739929, + "learning_rate": 9.986693087392341e-06, + "loss": 0.9291329383850098, + "step": 140 + }, + { + "epoch": 0.30021141649048627, + "grad_norm": 0.9209861159324646, + "learning_rate": 9.985456541276667e-06, + "loss": 1.558704137802124, + "step": 142 + }, + { + "epoch": 0.3044397463002114, + "grad_norm": 2.955430269241333, + "learning_rate": 9.984165158652708e-06, + "loss": 1.087266445159912, + "step": 144 + }, + { + "epoch": 0.3086680761099366, + "grad_norm": 6.303870677947998, + "learning_rate": 9.982818955308135e-06, + "loss": 0.6911844611167908, + "step": 146 + }, + { + "epoch": 0.3128964059196617, + "grad_norm": 1.2819095849990845, + "learning_rate": 9.98141794770083e-06, + "loss": 1.2495205402374268, + "step": 148 + }, + { + "epoch": 0.3171247357293869, + "grad_norm": 0.7855477333068848, + "learning_rate": 9.979962152958669e-06, + "loss": 1.2678008079528809, + "step": 150 + }, + { + "epoch": 0.321353065539112, + "grad_norm": 1.1494128704071045, + "learning_rate": 9.97845158887933e-06, + "loss": 1.1808208227157593, + "step": 152 + }, + { + "epoch": 0.32558139534883723, + "grad_norm": 3.952270269393921, + "learning_rate": 9.976886273930063e-06, + "loss": 0.7254101634025574, + "step": 154 + }, + { + "epoch": 0.3298097251585624, + "grad_norm": 1.3308897018432617, + "learning_rate": 9.975266227247474e-06, + "loss": 1.2565922737121582, + "step": 156 + }, + { + "epoch": 0.33403805496828753, + "grad_norm": 1.1793373823165894, + "learning_rate": 9.973591468637286e-06, + "loss": 1.2734328508377075, + "step": 158 + }, + { + "epoch": 0.3382663847780127, + "grad_norm": 1.140128493309021, + "learning_rate": 9.971862018574094e-06, + "loss": 1.263818621635437, + "step": 160 + }, + { + "epoch": 0.34249471458773784, + "grad_norm": 3.7501380443573, + "learning_rate": 9.97007789820112e-06, + "loss": 1.1025978326797485, + "step": 162 + }, + { + "epoch": 0.346723044397463, + "grad_norm": 1.1143251657485962, + "learning_rate": 9.968239129329953e-06, + "loss": 0.9754167199134827, + "step": 164 + }, + { + "epoch": 0.35095137420718814, + "grad_norm": 0.9423706531524658, + "learning_rate": 9.96634573444028e-06, + "loss": 0.8752356171607971, + "step": 166 + }, + { + "epoch": 0.35517970401691334, + "grad_norm": 1.0236040353775024, + "learning_rate": 9.964397736679615e-06, + "loss": 1.2032804489135742, + "step": 168 + }, + { + "epoch": 0.3594080338266385, + "grad_norm": 1.2067400217056274, + "learning_rate": 9.962395159863014e-06, + "loss": 1.2084052562713623, + "step": 170 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 0.9626054167747498, + "learning_rate": 9.960338028472781e-06, + "loss": 1.0512429475784302, + "step": 172 + }, + { + "epoch": 0.3678646934460888, + "grad_norm": 2.082251787185669, + "learning_rate": 9.958226367658176e-06, + "loss": 1.3552473783493042, + "step": 174 + }, + { + "epoch": 0.37209302325581395, + "grad_norm": 1.5437448024749756, + "learning_rate": 9.9560602032351e-06, + "loss": 1.041031837463379, + "step": 176 + }, + { + "epoch": 0.3763213530655391, + "grad_norm": 1.3151897192001343, + "learning_rate": 9.953839561685783e-06, + "loss": 0.8862758874893188, + "step": 178 + }, + { + "epoch": 0.38054968287526425, + "grad_norm": 0.7156529426574707, + "learning_rate": 9.95156447015846e-06, + "loss": 1.2866286039352417, + "step": 180 + }, + { + "epoch": 0.38477801268498946, + "grad_norm": 1.0742555856704712, + "learning_rate": 9.94923495646704e-06, + "loss": 1.5316616296768188, + "step": 182 + }, + { + "epoch": 0.3890063424947146, + "grad_norm": 0.8927209973335266, + "learning_rate": 9.946851049090764e-06, + "loss": 1.3418999910354614, + "step": 184 + }, + { + "epoch": 0.39323467230443976, + "grad_norm": 1.2585376501083374, + "learning_rate": 9.944412777173854e-06, + "loss": 1.1069649457931519, + "step": 186 + }, + { + "epoch": 0.3974630021141649, + "grad_norm": 0.7968735694885254, + "learning_rate": 9.941920170525168e-06, + "loss": 0.9848789572715759, + "step": 188 + }, + { + "epoch": 0.40169133192389006, + "grad_norm": 1.3116044998168945, + "learning_rate": 9.939373259617822e-06, + "loss": 1.0180245637893677, + "step": 190 + }, + { + "epoch": 0.4059196617336152, + "grad_norm": 1.1694923639297485, + "learning_rate": 9.936772075588825e-06, + "loss": 1.291759967803955, + "step": 192 + }, + { + "epoch": 0.41014799154334036, + "grad_norm": 0.585599958896637, + "learning_rate": 9.934116650238697e-06, + "loss": 1.0020666122436523, + "step": 194 + }, + { + "epoch": 0.4143763213530655, + "grad_norm": 1.1528569459915161, + "learning_rate": 9.931407016031078e-06, + "loss": 1.2772297859191895, + "step": 196 + }, + { + "epoch": 0.4186046511627907, + "grad_norm": 1.0158859491348267, + "learning_rate": 9.928643206092338e-06, + "loss": 1.0836846828460693, + "step": 198 + }, + { + "epoch": 0.42283298097251587, + "grad_norm": 0.7987179756164551, + "learning_rate": 9.925825254211161e-06, + "loss": 1.180878758430481, + "step": 200 + }, + { + "epoch": 0.427061310782241, + "grad_norm": 0.7621904611587524, + "learning_rate": 9.922953194838143e-06, + "loss": 0.5957623720169067, + "step": 202 + }, + { + "epoch": 0.4312896405919662, + "grad_norm": 1.1555891036987305, + "learning_rate": 9.920027063085363e-06, + "loss": 1.229513168334961, + "step": 204 + }, + { + "epoch": 0.4355179704016913, + "grad_norm": 1.4978004693984985, + "learning_rate": 9.917046894725957e-06, + "loss": 1.342698574066162, + "step": 206 + }, + { + "epoch": 0.4397463002114165, + "grad_norm": 1.7540870904922485, + "learning_rate": 9.91401272619368e-06, + "loss": 0.8040276765823364, + "step": 208 + }, + { + "epoch": 0.4439746300211416, + "grad_norm": 0.8134081363677979, + "learning_rate": 9.910924594582463e-06, + "loss": 1.2937664985656738, + "step": 210 + }, + { + "epoch": 0.44820295983086683, + "grad_norm": 1.3962867259979248, + "learning_rate": 9.907782537645955e-06, + "loss": 1.2151477336883545, + "step": 212 + }, + { + "epoch": 0.452431289640592, + "grad_norm": 1.61188542842865, + "learning_rate": 9.904586593797062e-06, + "loss": 1.4202353954315186, + "step": 214 + }, + { + "epoch": 0.45665961945031713, + "grad_norm": 1.581294059753418, + "learning_rate": 9.901336802107485e-06, + "loss": 0.675844132900238, + "step": 216 + }, + { + "epoch": 0.4608879492600423, + "grad_norm": 1.2479804754257202, + "learning_rate": 9.89803320230723e-06, + "loss": 1.012795329093933, + "step": 218 + }, + { + "epoch": 0.46511627906976744, + "grad_norm": 0.9484511017799377, + "learning_rate": 9.894675834784136e-06, + "loss": 1.279051661491394, + "step": 220 + }, + { + "epoch": 0.4693446088794926, + "grad_norm": 0.7353774905204773, + "learning_rate": 9.891264740583367e-06, + "loss": 1.2331568002700806, + "step": 222 + }, + { + "epoch": 0.47357293868921774, + "grad_norm": 1.1934891939163208, + "learning_rate": 9.88779996140692e-06, + "loss": 1.350287675857544, + "step": 224 + }, + { + "epoch": 0.47780126849894294, + "grad_norm": 0.9104489088058472, + "learning_rate": 9.884281539613109e-06, + "loss": 1.2547072172164917, + "step": 226 + }, + { + "epoch": 0.4820295983086681, + "grad_norm": 0.930194079875946, + "learning_rate": 9.880709518216061e-06, + "loss": 1.2463382482528687, + "step": 228 + }, + { + "epoch": 0.48625792811839325, + "grad_norm": 1.1311328411102295, + "learning_rate": 9.87708394088517e-06, + "loss": 0.9104933738708496, + "step": 230 + }, + { + "epoch": 0.4904862579281184, + "grad_norm": 1.0686322450637817, + "learning_rate": 9.873404851944573e-06, + "loss": 1.2515439987182617, + "step": 232 + }, + { + "epoch": 0.49471458773784355, + "grad_norm": 3.3511571884155273, + "learning_rate": 9.869672296372611e-06, + "loss": 0.9356804490089417, + "step": 234 + }, + { + "epoch": 0.4989429175475687, + "grad_norm": 1.7483917474746704, + "learning_rate": 9.86588631980128e-06, + "loss": 0.8896931409835815, + "step": 236 + }, + { + "epoch": 0.5031712473572939, + "grad_norm": 2.1715588569641113, + "learning_rate": 9.86204696851566e-06, + "loss": 0.8375036120414734, + "step": 238 + }, + { + "epoch": 0.507399577167019, + "grad_norm": 1.314707636833191, + "learning_rate": 9.858154289453363e-06, + "loss": 0.9143441915512085, + "step": 240 + }, + { + "epoch": 0.5116279069767442, + "grad_norm": 1.1541309356689453, + "learning_rate": 9.854208330203958e-06, + "loss": 1.3134888410568237, + "step": 242 + }, + { + "epoch": 0.5158562367864693, + "grad_norm": 1.4157994985580444, + "learning_rate": 9.850209139008377e-06, + "loss": 1.3324049711227417, + "step": 244 + }, + { + "epoch": 0.5200845665961945, + "grad_norm": 1.5940454006195068, + "learning_rate": 9.846156764758344e-06, + "loss": 1.2825590372085571, + "step": 246 + }, + { + "epoch": 0.5243128964059197, + "grad_norm": 1.454811930656433, + "learning_rate": 9.842051256995759e-06, + "loss": 1.2391467094421387, + "step": 248 + }, + { + "epoch": 0.5285412262156448, + "grad_norm": 4.359763145446777, + "learning_rate": 9.837892665912105e-06, + "loss": 1.1844691038131714, + "step": 250 + }, + { + "epoch": 0.53276955602537, + "grad_norm": 3.7198870182037354, + "learning_rate": 9.833681042347827e-06, + "loss": 1.1712547540664673, + "step": 252 + }, + { + "epoch": 0.5369978858350951, + "grad_norm": 2.235729455947876, + "learning_rate": 9.829416437791721e-06, + "loss": 1.7381950616836548, + "step": 254 + }, + { + "epoch": 0.5412262156448203, + "grad_norm": 0.6969403028488159, + "learning_rate": 9.825098904380284e-06, + "loss": 1.212001085281372, + "step": 256 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 1.8164496421813965, + "learning_rate": 9.820728494897101e-06, + "loss": 1.3891276121139526, + "step": 258 + }, + { + "epoch": 0.5496828752642706, + "grad_norm": 2.7816109657287598, + "learning_rate": 9.816305262772182e-06, + "loss": 0.9970607161521912, + "step": 260 + }, + { + "epoch": 0.5539112050739958, + "grad_norm": 0.8812597393989563, + "learning_rate": 9.811829262081315e-06, + "loss": 1.1956220865249634, + "step": 262 + }, + { + "epoch": 0.5581395348837209, + "grad_norm": 1.4800692796707153, + "learning_rate": 9.807300547545411e-06, + "loss": 1.0499297380447388, + "step": 264 + }, + { + "epoch": 0.5623678646934461, + "grad_norm": 0.8405512571334839, + "learning_rate": 9.802719174529821e-06, + "loss": 1.4957261085510254, + "step": 266 + }, + { + "epoch": 0.5665961945031712, + "grad_norm": 1.4424947500228882, + "learning_rate": 9.798085199043672e-06, + "loss": 0.5037683844566345, + "step": 268 + }, + { + "epoch": 0.5708245243128964, + "grad_norm": 1.6350433826446533, + "learning_rate": 9.793398677739177e-06, + "loss": 1.0034993886947632, + "step": 270 + }, + { + "epoch": 0.5750528541226215, + "grad_norm": 1.80682373046875, + "learning_rate": 9.788659667910942e-06, + "loss": 1.1760437488555908, + "step": 272 + }, + { + "epoch": 0.5792811839323467, + "grad_norm": 1.05997896194458, + "learning_rate": 9.783868227495262e-06, + "loss": 0.9529615640640259, + "step": 274 + }, + { + "epoch": 0.5835095137420718, + "grad_norm": 1.0263748168945312, + "learning_rate": 9.779024415069427e-06, + "loss": 0.9358394742012024, + "step": 276 + }, + { + "epoch": 0.587737843551797, + "grad_norm": 0.36564901471138, + "learning_rate": 9.774128289850987e-06, + "loss": 0.8667243123054504, + "step": 278 + }, + { + "epoch": 0.5919661733615222, + "grad_norm": 1.1363834142684937, + "learning_rate": 9.769179911697037e-06, + "loss": 1.2310261726379395, + "step": 280 + }, + { + "epoch": 0.5961945031712473, + "grad_norm": 1.1211930513381958, + "learning_rate": 9.764179341103493e-06, + "loss": 1.462058663368225, + "step": 282 + }, + { + "epoch": 0.6004228329809725, + "grad_norm": 0.9814391136169434, + "learning_rate": 9.759126639204337e-06, + "loss": 0.9405817985534668, + "step": 284 + }, + { + "epoch": 0.6046511627906976, + "grad_norm": 1.2274624109268188, + "learning_rate": 9.75402186777088e-06, + "loss": 1.3044648170471191, + "step": 286 + }, + { + "epoch": 0.6088794926004228, + "grad_norm": 1.5367823839187622, + "learning_rate": 9.748865089211004e-06, + "loss": 0.42142757773399353, + "step": 288 + }, + { + "epoch": 0.6131078224101479, + "grad_norm": 0.7510353326797485, + "learning_rate": 9.743656366568398e-06, + "loss": 0.9516089558601379, + "step": 290 + }, + { + "epoch": 0.6173361522198731, + "grad_norm": 2.0352747440338135, + "learning_rate": 9.738395763521792e-06, + "loss": 0.8791732788085938, + "step": 292 + }, + { + "epoch": 0.6215644820295984, + "grad_norm": 2.7773170471191406, + "learning_rate": 9.733083344384164e-06, + "loss": 0.7142581939697266, + "step": 294 + }, + { + "epoch": 0.6257928118393234, + "grad_norm": 1.7828168869018555, + "learning_rate": 9.727719174101981e-06, + "loss": 0.7309005260467529, + "step": 296 + }, + { + "epoch": 0.6300211416490487, + "grad_norm": 1.7246986627578735, + "learning_rate": 9.722303318254376e-06, + "loss": 1.2760108709335327, + "step": 298 + }, + { + "epoch": 0.6342494714587738, + "grad_norm": 1.3417115211486816, + "learning_rate": 9.71683584305236e-06, + "loss": 1.2378621101379395, + "step": 300 + }, + { + "epoch": 0.638477801268499, + "grad_norm": 1.669295310974121, + "learning_rate": 9.71131681533802e-06, + "loss": 0.7345954179763794, + "step": 302 + }, + { + "epoch": 0.642706131078224, + "grad_norm": 2.6989686489105225, + "learning_rate": 9.705746302583682e-06, + "loss": 0.8533758521080017, + "step": 304 + }, + { + "epoch": 0.6469344608879493, + "grad_norm": 1.0442886352539062, + "learning_rate": 9.700124372891104e-06, + "loss": 0.9097632169723511, + "step": 306 + }, + { + "epoch": 0.6511627906976745, + "grad_norm": 1.2805988788604736, + "learning_rate": 9.694451094990636e-06, + "loss": 1.2282127141952515, + "step": 308 + }, + { + "epoch": 0.6553911205073996, + "grad_norm": 1.9768813848495483, + "learning_rate": 9.68872653824038e-06, + "loss": 1.1992337703704834, + "step": 310 + }, + { + "epoch": 0.6596194503171248, + "grad_norm": 0.8558915853500366, + "learning_rate": 9.68295077262534e-06, + "loss": 1.2059025764465332, + "step": 312 + }, + { + "epoch": 0.6638477801268499, + "grad_norm": 1.5876579284667969, + "learning_rate": 9.677123868756573e-06, + "loss": 1.225184440612793, + "step": 314 + }, + { + "epoch": 0.6680761099365751, + "grad_norm": 0.5755170583724976, + "learning_rate": 9.671245897870318e-06, + "loss": 1.0713036060333252, + "step": 316 + }, + { + "epoch": 0.6723044397463002, + "grad_norm": 1.978392243385315, + "learning_rate": 9.665316931827131e-06, + "loss": 0.7629618048667908, + "step": 318 + }, + { + "epoch": 0.6765327695560254, + "grad_norm": 1.2458664178848267, + "learning_rate": 9.659337043111006e-06, + "loss": 1.2438840866088867, + "step": 320 + }, + { + "epoch": 0.6807610993657506, + "grad_norm": 1.383314609527588, + "learning_rate": 9.653306304828483e-06, + "loss": 1.3052635192871094, + "step": 322 + }, + { + "epoch": 0.6849894291754757, + "grad_norm": 1.1016979217529297, + "learning_rate": 9.647224790707762e-06, + "loss": 1.0821185111999512, + "step": 324 + }, + { + "epoch": 0.6892177589852009, + "grad_norm": 1.3163999319076538, + "learning_rate": 9.641092575097795e-06, + "loss": 1.2659895420074463, + "step": 326 + }, + { + "epoch": 0.693446088794926, + "grad_norm": 1.9359853267669678, + "learning_rate": 9.63490973296738e-06, + "loss": 1.1682963371276855, + "step": 328 + }, + { + "epoch": 0.6976744186046512, + "grad_norm": 4.193031311035156, + "learning_rate": 9.62867633990425e-06, + "loss": 0.8624615669250488, + "step": 330 + }, + { + "epoch": 0.7019027484143763, + "grad_norm": 0.9525843858718872, + "learning_rate": 9.622392472114136e-06, + "loss": 1.2672665119171143, + "step": 332 + }, + { + "epoch": 0.7061310782241015, + "grad_norm": 2.293287754058838, + "learning_rate": 9.61605820641985e-06, + "loss": 1.2097090482711792, + "step": 334 + }, + { + "epoch": 0.7103594080338267, + "grad_norm": 1.2369450330734253, + "learning_rate": 9.60967362026033e-06, + "loss": 1.2070491313934326, + "step": 336 + }, + { + "epoch": 0.7145877378435518, + "grad_norm": 0.9647306799888611, + "learning_rate": 9.60323879168971e-06, + "loss": 1.3035504817962646, + "step": 338 + }, + { + "epoch": 0.718816067653277, + "grad_norm": 1.6533912420272827, + "learning_rate": 9.596753799376355e-06, + "loss": 1.1961348056793213, + "step": 340 + }, + { + "epoch": 0.7230443974630021, + "grad_norm": 2.9031484127044678, + "learning_rate": 9.5902187226019e-06, + "loss": 1.1233775615692139, + "step": 342 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 1.1715766191482544, + "learning_rate": 9.583633641260285e-06, + "loss": 0.9001784920692444, + "step": 344 + }, + { + "epoch": 0.7315010570824524, + "grad_norm": 3.3942534923553467, + "learning_rate": 9.576998635856774e-06, + "loss": 0.8353261947631836, + "step": 346 + }, + { + "epoch": 0.7357293868921776, + "grad_norm": 2.7805113792419434, + "learning_rate": 9.570313787506974e-06, + "loss": 0.9958086609840393, + "step": 348 + }, + { + "epoch": 0.7399577167019028, + "grad_norm": 0.9791728258132935, + "learning_rate": 9.563579177935844e-06, + "loss": 1.3026225566864014, + "step": 350 + }, + { + "epoch": 0.7441860465116279, + "grad_norm": 1.1662095785140991, + "learning_rate": 9.556794889476687e-06, + "loss": 1.1562281847000122, + "step": 352 + }, + { + "epoch": 0.7484143763213531, + "grad_norm": 2.089334011077881, + "learning_rate": 9.549961005070161e-06, + "loss": 0.7143859267234802, + "step": 354 + }, + { + "epoch": 0.7526427061310782, + "grad_norm": 8.907992362976074, + "learning_rate": 9.543077608263245e-06, + "loss": 1.0247828960418701, + "step": 356 + }, + { + "epoch": 0.7568710359408034, + "grad_norm": 1.3047367334365845, + "learning_rate": 9.536144783208229e-06, + "loss": 1.2090858221054077, + "step": 358 + }, + { + "epoch": 0.7610993657505285, + "grad_norm": 1.3646140098571777, + "learning_rate": 9.529162614661687e-06, + "loss": 0.7477072477340698, + "step": 360 + }, + { + "epoch": 0.7653276955602537, + "grad_norm": 1.8884954452514648, + "learning_rate": 9.522131187983433e-06, + "loss": 1.0810452699661255, + "step": 362 + }, + { + "epoch": 0.7695560253699789, + "grad_norm": 14.129027366638184, + "learning_rate": 9.515050589135485e-06, + "loss": 1.0486022233963013, + "step": 364 + }, + { + "epoch": 0.773784355179704, + "grad_norm": 2.512070894241333, + "learning_rate": 9.507920904681006e-06, + "loss": 1.261042833328247, + "step": 366 + }, + { + "epoch": 0.7780126849894292, + "grad_norm": 1.4172213077545166, + "learning_rate": 9.500742221783257e-06, + "loss": 1.2038286924362183, + "step": 368 + }, + { + "epoch": 0.7822410147991543, + "grad_norm": 6.898794174194336, + "learning_rate": 9.493514628204516e-06, + "loss": 0.8664312362670898, + "step": 370 + }, + { + "epoch": 0.7864693446088795, + "grad_norm": 4.238753318786621, + "learning_rate": 9.48623821230502e-06, + "loss": 1.192126750946045, + "step": 372 + }, + { + "epoch": 0.7906976744186046, + "grad_norm": 1.3229868412017822, + "learning_rate": 9.478913063041878e-06, + "loss": 1.2105952501296997, + "step": 374 + }, + { + "epoch": 0.7949260042283298, + "grad_norm": 2.623147487640381, + "learning_rate": 9.471539269967982e-06, + "loss": 1.2897117137908936, + "step": 376 + }, + { + "epoch": 0.7991543340380549, + "grad_norm": 1.1712391376495361, + "learning_rate": 9.464116923230913e-06, + "loss": 0.9584426879882812, + "step": 378 + }, + { + "epoch": 0.8033826638477801, + "grad_norm": 3.0442938804626465, + "learning_rate": 9.456646113571844e-06, + "loss": 1.3029778003692627, + "step": 380 + }, + { + "epoch": 0.8076109936575053, + "grad_norm": 1.2261182069778442, + "learning_rate": 9.449126932324424e-06, + "loss": 1.1654752492904663, + "step": 382 + }, + { + "epoch": 0.8118393234672304, + "grad_norm": 1.098416805267334, + "learning_rate": 9.441559471413665e-06, + "loss": 0.9697389006614685, + "step": 384 + }, + { + "epoch": 0.8160676532769556, + "grad_norm": 2.7409515380859375, + "learning_rate": 9.433943823354816e-06, + "loss": 0.9527001976966858, + "step": 386 + }, + { + "epoch": 0.8202959830866807, + "grad_norm": 1.223421573638916, + "learning_rate": 9.426280081252232e-06, + "loss": 1.4748138189315796, + "step": 388 + }, + { + "epoch": 0.8245243128964059, + "grad_norm": 6.851096153259277, + "learning_rate": 9.418568338798245e-06, + "loss": 1.1421523094177246, + "step": 390 + }, + { + "epoch": 0.828752642706131, + "grad_norm": 2.2886765003204346, + "learning_rate": 9.410808690271999e-06, + "loss": 0.940636932849884, + "step": 392 + }, + { + "epoch": 0.8329809725158562, + "grad_norm": 2.155710458755493, + "learning_rate": 9.40300123053832e-06, + "loss": 1.232945442199707, + "step": 394 + }, + { + "epoch": 0.8372093023255814, + "grad_norm": 0.3345288336277008, + "learning_rate": 9.395146055046535e-06, + "loss": 0.7638389468193054, + "step": 396 + }, + { + "epoch": 0.8414376321353065, + "grad_norm": 1.7543056011199951, + "learning_rate": 9.387243259829323e-06, + "loss": 0.6838975548744202, + "step": 398 + }, + { + "epoch": 0.8456659619450317, + "grad_norm": 1.306065320968628, + "learning_rate": 9.37929294150153e-06, + "loss": 1.1204407215118408, + "step": 400 + }, + { + "epoch": 0.8498942917547568, + "grad_norm": 1.5666264295578003, + "learning_rate": 9.371295197258992e-06, + "loss": 0.9662034511566162, + "step": 402 + }, + { + "epoch": 0.854122621564482, + "grad_norm": 1.3164730072021484, + "learning_rate": 9.363250124877345e-06, + "loss": 0.8297775387763977, + "step": 404 + }, + { + "epoch": 0.8583509513742071, + "grad_norm": 2.3461363315582275, + "learning_rate": 9.35515782271083e-06, + "loss": 0.9872302412986755, + "step": 406 + }, + { + "epoch": 0.8625792811839323, + "grad_norm": 1.0589134693145752, + "learning_rate": 9.347018389691093e-06, + "loss": 1.0059325695037842, + "step": 408 + }, + { + "epoch": 0.8668076109936576, + "grad_norm": 1.1941379308700562, + "learning_rate": 9.338831925325974e-06, + "loss": 0.8844459056854248, + "step": 410 + }, + { + "epoch": 0.8710359408033826, + "grad_norm": 1.0028619766235352, + "learning_rate": 9.330598529698287e-06, + "loss": 1.1726793050765991, + "step": 412 + }, + { + "epoch": 0.8752642706131079, + "grad_norm": 1.0313185453414917, + "learning_rate": 9.322318303464603e-06, + "loss": 1.0125584602355957, + "step": 414 + }, + { + "epoch": 0.879492600422833, + "grad_norm": 1.7959996461868286, + "learning_rate": 9.313991347854015e-06, + "loss": 1.1153017282485962, + "step": 416 + }, + { + "epoch": 0.8837209302325582, + "grad_norm": 1.3663700819015503, + "learning_rate": 9.305617764666897e-06, + "loss": 1.2116621732711792, + "step": 418 + }, + { + "epoch": 0.8879492600422833, + "grad_norm": 1.9488954544067383, + "learning_rate": 9.297197656273672e-06, + "loss": 1.2355561256408691, + "step": 420 + }, + { + "epoch": 0.8921775898520085, + "grad_norm": 2.1836400032043457, + "learning_rate": 9.288731125613544e-06, + "loss": 0.9339479804039001, + "step": 422 + }, + { + "epoch": 0.8964059196617337, + "grad_norm": 1.1514095067977905, + "learning_rate": 9.280218276193255e-06, + "loss": 1.2223608493804932, + "step": 424 + }, + { + "epoch": 0.9006342494714588, + "grad_norm": 3.1102023124694824, + "learning_rate": 9.271659212085807e-06, + "loss": 1.0882915258407593, + "step": 426 + }, + { + "epoch": 0.904862579281184, + "grad_norm": 2.123033285140991, + "learning_rate": 9.263054037929199e-06, + "loss": 0.8707346320152283, + "step": 428 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 1.359169602394104, + "learning_rate": 9.254402858925142e-06, + "loss": 1.107509732246399, + "step": 430 + }, + { + "epoch": 0.9133192389006343, + "grad_norm": 1.7108887434005737, + "learning_rate": 9.245705780837774e-06, + "loss": 1.254375696182251, + "step": 432 + }, + { + "epoch": 0.9175475687103594, + "grad_norm": 2.344334125518799, + "learning_rate": 9.236962909992368e-06, + "loss": 0.5221960544586182, + "step": 434 + }, + { + "epoch": 0.9217758985200846, + "grad_norm": 2.77628231048584, + "learning_rate": 9.228174353274037e-06, + "loss": 0.3872157633304596, + "step": 436 + }, + { + "epoch": 0.9260042283298098, + "grad_norm": 0.7304924130439758, + "learning_rate": 9.219340218126413e-06, + "loss": 0.9492573738098145, + "step": 438 + }, + { + "epoch": 0.9302325581395349, + "grad_norm": 0.7027140855789185, + "learning_rate": 9.210460612550348e-06, + "loss": 1.127206563949585, + "step": 440 + }, + { + "epoch": 0.9344608879492601, + "grad_norm": 1.4444884061813354, + "learning_rate": 9.20153564510259e-06, + "loss": 0.7414720058441162, + "step": 442 + }, + { + "epoch": 0.9386892177589852, + "grad_norm": 1.4010649919509888, + "learning_rate": 9.19256542489445e-06, + "loss": 1.249424934387207, + "step": 444 + }, + { + "epoch": 0.9429175475687104, + "grad_norm": 0.9628981947898865, + "learning_rate": 9.183550061590476e-06, + "loss": 1.0961284637451172, + "step": 446 + }, + { + "epoch": 0.9471458773784355, + "grad_norm": 3.5854969024658203, + "learning_rate": 9.174489665407103e-06, + "loss": 1.0860631465911865, + "step": 448 + }, + { + "epoch": 0.9513742071881607, + "grad_norm": 2.347378969192505, + "learning_rate": 9.165384347111317e-06, + "loss": 0.7947496175765991, + "step": 450 + }, + { + "epoch": 0.9556025369978859, + "grad_norm": 0.9492086172103882, + "learning_rate": 9.156234218019291e-06, + "loss": 0.9331413507461548, + "step": 452 + }, + { + "epoch": 0.959830866807611, + "grad_norm": 1.919443964958191, + "learning_rate": 9.147039389995029e-06, + "loss": 1.318381905555725, + "step": 454 + }, + { + "epoch": 0.9640591966173362, + "grad_norm": 4.199268817901611, + "learning_rate": 9.137799975448998e-06, + "loss": 0.6622285842895508, + "step": 456 + }, + { + "epoch": 0.9682875264270613, + "grad_norm": 6.41878080368042, + "learning_rate": 9.128516087336755e-06, + "loss": 0.8811239004135132, + "step": 458 + }, + { + "epoch": 0.9725158562367865, + "grad_norm": 1.428720474243164, + "learning_rate": 9.119187839157559e-06, + "loss": 1.228691816329956, + "step": 460 + }, + { + "epoch": 0.9767441860465116, + "grad_norm": 1.1554924249649048, + "learning_rate": 9.109815344952995e-06, + "loss": 1.215306282043457, + "step": 462 + }, + { + "epoch": 0.9809725158562368, + "grad_norm": 1.901812195777893, + "learning_rate": 9.100398719305566e-06, + "loss": 0.7152163982391357, + "step": 464 + }, + { + "epoch": 0.985200845665962, + "grad_norm": 3.8220489025115967, + "learning_rate": 9.090938077337305e-06, + "loss": 1.184203863143921, + "step": 466 + }, + { + "epoch": 0.9894291754756871, + "grad_norm": 2.5080857276916504, + "learning_rate": 9.081433534708363e-06, + "loss": 0.9817585945129395, + "step": 468 + }, + { + "epoch": 0.9936575052854123, + "grad_norm": 1.1134026050567627, + "learning_rate": 9.071885207615591e-06, + "loss": 1.24997079372406, + "step": 470 + }, + { + "epoch": 0.9978858350951374, + "grad_norm": 3.835284471511841, + "learning_rate": 9.062293212791123e-06, + "loss": 1.2497031688690186, + "step": 472 + }, + { + "epoch": 1.0021141649048626, + "grad_norm": 1.8404693603515625, + "learning_rate": 9.052657667500951e-06, + "loss": 1.0115447044372559, + "step": 474 + }, + { + "epoch": 1.0063424947145878, + "grad_norm": 1.1473926305770874, + "learning_rate": 9.042978689543487e-06, + "loss": 1.1021543741226196, + "step": 476 + }, + { + "epoch": 1.0105708245243128, + "grad_norm": 9.168399810791016, + "learning_rate": 9.033256397248124e-06, + "loss": 0.4507979452610016, + "step": 478 + }, + { + "epoch": 1.014799154334038, + "grad_norm": 1.804311752319336, + "learning_rate": 9.02349090947379e-06, + "loss": 0.6752689480781555, + "step": 480 + }, + { + "epoch": 1.0190274841437632, + "grad_norm": 1.3448916673660278, + "learning_rate": 9.013682345607502e-06, + "loss": 0.6984230279922485, + "step": 482 + }, + { + "epoch": 1.0232558139534884, + "grad_norm": 1.341726541519165, + "learning_rate": 9.003830825562887e-06, + "loss": 1.025214672088623, + "step": 484 + }, + { + "epoch": 1.0274841437632136, + "grad_norm": 2.1331229209899902, + "learning_rate": 8.99393646977874e-06, + "loss": 1.0407471656799316, + "step": 486 + }, + { + "epoch": 1.0317124735729386, + "grad_norm": 1.393306016921997, + "learning_rate": 8.983999399217536e-06, + "loss": 0.874075174331665, + "step": 488 + }, + { + "epoch": 1.0359408033826638, + "grad_norm": 2.908888339996338, + "learning_rate": 8.974019735363952e-06, + "loss": 0.8063684701919556, + "step": 490 + }, + { + "epoch": 1.040169133192389, + "grad_norm": 1.5344609022140503, + "learning_rate": 8.963997600223387e-06, + "loss": 0.6792523264884949, + "step": 492 + }, + { + "epoch": 1.0443974630021142, + "grad_norm": 1.2014660835266113, + "learning_rate": 8.953933116320474e-06, + "loss": 1.140464425086975, + "step": 494 + }, + { + "epoch": 1.0486257928118394, + "grad_norm": 1.4177594184875488, + "learning_rate": 8.943826406697569e-06, + "loss": 0.9537199139595032, + "step": 496 + }, + { + "epoch": 1.0528541226215644, + "grad_norm": 0.9527637362480164, + "learning_rate": 8.93367759491326e-06, + "loss": 0.7929669618606567, + "step": 498 + }, + { + "epoch": 1.0570824524312896, + "grad_norm": 1.0293605327606201, + "learning_rate": 8.923486805040845e-06, + "loss": 0.8405656218528748, + "step": 500 + }, + { + "epoch": 1.0613107822410148, + "grad_norm": 1.9201829433441162, + "learning_rate": 8.913254161666828e-06, + "loss": 1.1188585758209229, + "step": 502 + }, + { + "epoch": 1.06553911205074, + "grad_norm": 0.9364558458328247, + "learning_rate": 8.902979789889382e-06, + "loss": 1.0248576402664185, + "step": 504 + }, + { + "epoch": 1.069767441860465, + "grad_norm": 24.75840950012207, + "learning_rate": 8.892663815316833e-06, + "loss": 0.8036352396011353, + "step": 506 + }, + { + "epoch": 1.0739957716701902, + "grad_norm": 7.285903453826904, + "learning_rate": 8.882306364066114e-06, + "loss": 0.7909092903137207, + "step": 508 + }, + { + "epoch": 1.0782241014799154, + "grad_norm": 5.538660049438477, + "learning_rate": 8.871907562761226e-06, + "loss": 0.8920198082923889, + "step": 510 + }, + { + "epoch": 1.0824524312896406, + "grad_norm": 1.2150629758834839, + "learning_rate": 8.861467538531692e-06, + "loss": 0.9995055794715881, + "step": 512 + }, + { + "epoch": 1.0866807610993658, + "grad_norm": 1.3044167757034302, + "learning_rate": 8.850986419011006e-06, + "loss": 1.081452488899231, + "step": 514 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 1.423043131828308, + "learning_rate": 8.840464332335062e-06, + "loss": 0.9120286703109741, + "step": 516 + }, + { + "epoch": 1.095137420718816, + "grad_norm": 1.006951093673706, + "learning_rate": 8.829901407140599e-06, + "loss": 0.8389664888381958, + "step": 518 + }, + { + "epoch": 1.0993657505285412, + "grad_norm": 6.160975933074951, + "learning_rate": 8.81929777256362e-06, + "loss": 0.631398618221283, + "step": 520 + }, + { + "epoch": 1.1035940803382664, + "grad_norm": 1.0400751829147339, + "learning_rate": 8.808653558237818e-06, + "loss": 1.2913062572479248, + "step": 522 + }, + { + "epoch": 1.1078224101479917, + "grad_norm": 3.1153197288513184, + "learning_rate": 8.797968894292992e-06, + "loss": 0.734929621219635, + "step": 524 + }, + { + "epoch": 1.1120507399577166, + "grad_norm": 2.215916156768799, + "learning_rate": 8.787243911353447e-06, + "loss": 1.2899256944656372, + "step": 526 + }, + { + "epoch": 1.1162790697674418, + "grad_norm": 1.1769880056381226, + "learning_rate": 8.776478740536409e-06, + "loss": 0.8642110824584961, + "step": 528 + }, + { + "epoch": 1.120507399577167, + "grad_norm": 1.7178997993469238, + "learning_rate": 8.76567351345042e-06, + "loss": 0.4671909213066101, + "step": 530 + }, + { + "epoch": 1.1247357293868923, + "grad_norm": 1.4589587450027466, + "learning_rate": 8.754828362193719e-06, + "loss": 1.0046720504760742, + "step": 532 + }, + { + "epoch": 1.1289640591966172, + "grad_norm": 1.6525269746780396, + "learning_rate": 8.743943419352638e-06, + "loss": 0.8204197287559509, + "step": 534 + }, + { + "epoch": 1.1331923890063424, + "grad_norm": 0.7637603878974915, + "learning_rate": 8.733018817999977e-06, + "loss": 0.9065571427345276, + "step": 536 + }, + { + "epoch": 1.1374207188160677, + "grad_norm": 1.7804979085922241, + "learning_rate": 8.722054691693374e-06, + "loss": 0.6969019174575806, + "step": 538 + }, + { + "epoch": 1.1416490486257929, + "grad_norm": 1.2744324207305908, + "learning_rate": 8.711051174473682e-06, + "loss": 0.7288663983345032, + "step": 540 + }, + { + "epoch": 1.145877378435518, + "grad_norm": 1.005840539932251, + "learning_rate": 8.700008400863316e-06, + "loss": 1.0268869400024414, + "step": 542 + }, + { + "epoch": 1.150105708245243, + "grad_norm": 2.5568649768829346, + "learning_rate": 8.688926505864624e-06, + "loss": 0.7326473593711853, + "step": 544 + }, + { + "epoch": 1.1543340380549683, + "grad_norm": 0.8771763443946838, + "learning_rate": 8.67780562495822e-06, + "loss": 1.1725282669067383, + "step": 546 + }, + { + "epoch": 1.1585623678646935, + "grad_norm": 1.8093726634979248, + "learning_rate": 8.666645894101346e-06, + "loss": 1.2458208799362183, + "step": 548 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 1.292094111442566, + "learning_rate": 8.655447449726189e-06, + "loss": 0.8068552017211914, + "step": 550 + }, + { + "epoch": 1.1670190274841437, + "grad_norm": 1.4168565273284912, + "learning_rate": 8.644210428738237e-06, + "loss": 1.020805835723877, + "step": 552 + }, + { + "epoch": 1.1712473572938689, + "grad_norm": 1.578667402267456, + "learning_rate": 8.632934968514582e-06, + "loss": 1.1674864292144775, + "step": 554 + }, + { + "epoch": 1.175475687103594, + "grad_norm": 1.1650516986846924, + "learning_rate": 8.621621206902256e-06, + "loss": 0.9501446485519409, + "step": 556 + }, + { + "epoch": 1.1797040169133193, + "grad_norm": 1.4428889751434326, + "learning_rate": 8.61026928221654e-06, + "loss": 1.0647276639938354, + "step": 558 + }, + { + "epoch": 1.1839323467230445, + "grad_norm": 0.9134294390678406, + "learning_rate": 8.598879333239274e-06, + "loss": 1.0058863162994385, + "step": 560 + }, + { + "epoch": 1.1881606765327695, + "grad_norm": 0.7719096541404724, + "learning_rate": 8.587451499217161e-06, + "loss": 0.7630645036697388, + "step": 562 + }, + { + "epoch": 1.1923890063424947, + "grad_norm": 1.1328723430633545, + "learning_rate": 8.575985919860065e-06, + "loss": 1.0797064304351807, + "step": 564 + }, + { + "epoch": 1.1966173361522199, + "grad_norm": 2.174982786178589, + "learning_rate": 8.564482735339296e-06, + "loss": 0.9507155418395996, + "step": 566 + }, + { + "epoch": 1.200845665961945, + "grad_norm": 1.6919512748718262, + "learning_rate": 8.552942086285909e-06, + "loss": 1.2098662853240967, + "step": 568 + }, + { + "epoch": 1.20507399577167, + "grad_norm": 1.5530591011047363, + "learning_rate": 8.541364113788974e-06, + "loss": 1.1700780391693115, + "step": 570 + }, + { + "epoch": 1.2093023255813953, + "grad_norm": 1.35764479637146, + "learning_rate": 8.529748959393855e-06, + "loss": 0.9838692545890808, + "step": 572 + }, + { + "epoch": 1.2135306553911205, + "grad_norm": 1.2259602546691895, + "learning_rate": 8.518096765100485e-06, + "loss": 0.8313025236129761, + "step": 574 + }, + { + "epoch": 1.2177589852008457, + "grad_norm": 4.002320289611816, + "learning_rate": 8.50640767336162e-06, + "loss": 0.7877824902534485, + "step": 576 + }, + { + "epoch": 1.221987315010571, + "grad_norm": 2.437662363052368, + "learning_rate": 8.494681827081096e-06, + "loss": 0.9264514446258545, + "step": 578 + }, + { + "epoch": 1.226215644820296, + "grad_norm": 1.8718764781951904, + "learning_rate": 8.482919369612103e-06, + "loss": 0.7877235412597656, + "step": 580 + }, + { + "epoch": 1.230443974630021, + "grad_norm": 1.0379586219787598, + "learning_rate": 8.471120444755405e-06, + "loss": 1.0842227935791016, + "step": 582 + }, + { + "epoch": 1.2346723044397463, + "grad_norm": 2.0494489669799805, + "learning_rate": 8.4592851967576e-06, + "loss": 0.8893200159072876, + "step": 584 + }, + { + "epoch": 1.2389006342494715, + "grad_norm": 2.2364420890808105, + "learning_rate": 8.447413770309349e-06, + "loss": 0.8317912817001343, + "step": 586 + }, + { + "epoch": 1.2431289640591967, + "grad_norm": 1.1253223419189453, + "learning_rate": 8.435506310543607e-06, + "loss": 0.9167896509170532, + "step": 588 + }, + { + "epoch": 1.2473572938689217, + "grad_norm": 1.2269498109817505, + "learning_rate": 8.42356296303386e-06, + "loss": 0.6032033562660217, + "step": 590 + }, + { + "epoch": 1.251585623678647, + "grad_norm": 1.6549971103668213, + "learning_rate": 8.411583873792325e-06, + "loss": 0.8433573842048645, + "step": 592 + }, + { + "epoch": 1.255813953488372, + "grad_norm": 2.2118799686431885, + "learning_rate": 8.399569189268182e-06, + "loss": 1.0482879877090454, + "step": 594 + }, + { + "epoch": 1.2600422832980973, + "grad_norm": 0.9963634610176086, + "learning_rate": 8.387519056345778e-06, + "loss": 0.6777889728546143, + "step": 596 + }, + { + "epoch": 1.2642706131078225, + "grad_norm": 0.9078873991966248, + "learning_rate": 8.375433622342824e-06, + "loss": 1.0934808254241943, + "step": 598 + }, + { + "epoch": 1.2684989429175475, + "grad_norm": 1.9698995351791382, + "learning_rate": 8.363313035008613e-06, + "loss": 1.270100712776184, + "step": 600 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 1.191821813583374, + "learning_rate": 8.351157442522187e-06, + "loss": 0.9545477032661438, + "step": 602 + }, + { + "epoch": 1.276955602536998, + "grad_norm": 1.6409633159637451, + "learning_rate": 8.338966993490554e-06, + "loss": 1.127218246459961, + "step": 604 + }, + { + "epoch": 1.2811839323467231, + "grad_norm": 1.2714064121246338, + "learning_rate": 8.326741836946852e-06, + "loss": 0.7959360480308533, + "step": 606 + }, + { + "epoch": 1.285412262156448, + "grad_norm": 1.068260669708252, + "learning_rate": 8.314482122348523e-06, + "loss": 1.0974210500717163, + "step": 608 + }, + { + "epoch": 1.2896405919661733, + "grad_norm": 1.1217502355575562, + "learning_rate": 8.30218799957551e-06, + "loss": 0.7269242405891418, + "step": 610 + }, + { + "epoch": 1.2938689217758985, + "grad_norm": 1.0048211812973022, + "learning_rate": 8.289859618928404e-06, + "loss": 0.6328207850456238, + "step": 612 + }, + { + "epoch": 1.2980972515856237, + "grad_norm": 1.3996940851211548, + "learning_rate": 8.277497131126608e-06, + "loss": 0.8515206575393677, + "step": 614 + }, + { + "epoch": 1.302325581395349, + "grad_norm": 1.8583265542984009, + "learning_rate": 8.265100687306503e-06, + "loss": 0.5192343592643738, + "step": 616 + }, + { + "epoch": 1.306553911205074, + "grad_norm": 2.725501537322998, + "learning_rate": 8.252670439019595e-06, + "loss": 0.9076979160308838, + "step": 618 + }, + { + "epoch": 1.3107822410147991, + "grad_norm": 0.9787073731422424, + "learning_rate": 8.240206538230663e-06, + "loss": 1.1246055364608765, + "step": 620 + }, + { + "epoch": 1.3150105708245243, + "grad_norm": 1.082661509513855, + "learning_rate": 8.227709137315905e-06, + "loss": 1.0403695106506348, + "step": 622 + }, + { + "epoch": 1.3192389006342495, + "grad_norm": 2.8779456615448, + "learning_rate": 8.215178389061064e-06, + "loss": 0.6083824038505554, + "step": 624 + }, + { + "epoch": 1.3234672304439745, + "grad_norm": 1.0220073461532593, + "learning_rate": 8.202614446659572e-06, + "loss": 1.1074095964431763, + "step": 626 + }, + { + "epoch": 1.3276955602536997, + "grad_norm": 1.632725715637207, + "learning_rate": 8.190017463710675e-06, + "loss": 1.194908857345581, + "step": 628 + }, + { + "epoch": 1.331923890063425, + "grad_norm": 1.0367897748947144, + "learning_rate": 8.177387594217548e-06, + "loss": 0.972278356552124, + "step": 630 + }, + { + "epoch": 1.3361522198731501, + "grad_norm": 1.3404940366744995, + "learning_rate": 8.164724992585421e-06, + "loss": 0.6174381375312805, + "step": 632 + }, + { + "epoch": 1.3403805496828753, + "grad_norm": 3.324969530105591, + "learning_rate": 8.152029813619685e-06, + "loss": 0.7120758295059204, + "step": 634 + }, + { + "epoch": 1.3446088794926006, + "grad_norm": 0.7900609970092773, + "learning_rate": 8.139302212524006e-06, + "loss": 0.7126297354698181, + "step": 636 + }, + { + "epoch": 1.3488372093023255, + "grad_norm": 2.6332106590270996, + "learning_rate": 8.126542344898419e-06, + "loss": 1.003250002861023, + "step": 638 + }, + { + "epoch": 1.3530655391120507, + "grad_norm": 0.4052017033100128, + "learning_rate": 8.113750366737431e-06, + "loss": 0.7192285656929016, + "step": 640 + }, + { + "epoch": 1.357293868921776, + "grad_norm": 0.8239710330963135, + "learning_rate": 8.100926434428116e-06, + "loss": 1.1896562576293945, + "step": 642 + }, + { + "epoch": 1.361522198731501, + "grad_norm": 1.1646950244903564, + "learning_rate": 8.0880707047482e-06, + "loss": 1.043108582496643, + "step": 644 + }, + { + "epoch": 1.3657505285412261, + "grad_norm": 1.558267593383789, + "learning_rate": 8.075183334864139e-06, + "loss": 0.732648491859436, + "step": 646 + }, + { + "epoch": 1.3699788583509513, + "grad_norm": 2.3034307956695557, + "learning_rate": 8.062264482329209e-06, + "loss": 0.6875412464141846, + "step": 648 + }, + { + "epoch": 1.3742071881606766, + "grad_norm": 1.7112483978271484, + "learning_rate": 8.049314305081575e-06, + "loss": 0.7501391172409058, + "step": 650 + }, + { + "epoch": 1.3784355179704018, + "grad_norm": 1.4458472728729248, + "learning_rate": 8.036332961442355e-06, + "loss": 0.7558017373085022, + "step": 652 + }, + { + "epoch": 1.382663847780127, + "grad_norm": 2.3118174076080322, + "learning_rate": 8.023320610113691e-06, + "loss": 1.1631048917770386, + "step": 654 + }, + { + "epoch": 1.386892177589852, + "grad_norm": 0.9235996007919312, + "learning_rate": 8.010277410176801e-06, + "loss": 1.033468246459961, + "step": 656 + }, + { + "epoch": 1.3911205073995772, + "grad_norm": 3.6433463096618652, + "learning_rate": 7.997203521090053e-06, + "loss": 0.822655200958252, + "step": 658 + }, + { + "epoch": 1.3953488372093024, + "grad_norm": 1.0611681938171387, + "learning_rate": 7.984099102686986e-06, + "loss": 0.9356414079666138, + "step": 660 + }, + { + "epoch": 1.3995771670190273, + "grad_norm": 0.9820582270622253, + "learning_rate": 7.970964315174384e-06, + "loss": 1.1282542943954468, + "step": 662 + }, + { + "epoch": 1.4038054968287526, + "grad_norm": 0.2741073966026306, + "learning_rate": 7.9577993191303e-06, + "loss": 0.8920847773551941, + "step": 664 + }, + { + "epoch": 1.4080338266384778, + "grad_norm": 2.034482955932617, + "learning_rate": 7.944604275502102e-06, + "loss": 1.3741508722305298, + "step": 666 + }, + { + "epoch": 1.412262156448203, + "grad_norm": 1.2096502780914307, + "learning_rate": 7.931379345604495e-06, + "loss": 1.074036717414856, + "step": 668 + }, + { + "epoch": 1.4164904862579282, + "grad_norm": 0.6002820730209351, + "learning_rate": 7.918124691117563e-06, + "loss": 0.6389409899711609, + "step": 670 + }, + { + "epoch": 1.4207188160676534, + "grad_norm": 5.750682353973389, + "learning_rate": 7.904840474084783e-06, + "loss": 1.0348693132400513, + "step": 672 + }, + { + "epoch": 1.4249471458773784, + "grad_norm": 0.7815237045288086, + "learning_rate": 7.891526856911046e-06, + "loss": 0.5709630250930786, + "step": 674 + }, + { + "epoch": 1.4291754756871036, + "grad_norm": 1.744208812713623, + "learning_rate": 7.878184002360664e-06, + "loss": 0.8444862365722656, + "step": 676 + }, + { + "epoch": 1.4334038054968288, + "grad_norm": 1.0372453927993774, + "learning_rate": 7.864812073555402e-06, + "loss": 1.2322757244110107, + "step": 678 + }, + { + "epoch": 1.437632135306554, + "grad_norm": 1.1494144201278687, + "learning_rate": 7.851411233972461e-06, + "loss": 1.1701394319534302, + "step": 680 + }, + { + "epoch": 1.441860465116279, + "grad_norm": 1.4448765516281128, + "learning_rate": 7.837981647442486e-06, + "loss": 1.1230388879776, + "step": 682 + }, + { + "epoch": 1.4460887949260042, + "grad_norm": 0.7794700860977173, + "learning_rate": 7.82452347814757e-06, + "loss": 1.0422362089157104, + "step": 684 + }, + { + "epoch": 1.4503171247357294, + "grad_norm": 6.159451961517334, + "learning_rate": 7.811036890619239e-06, + "loss": 1.4155765771865845, + "step": 686 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 1.3153562545776367, + "learning_rate": 7.797522049736441e-06, + "loss": 1.0144433975219727, + "step": 688 + }, + { + "epoch": 1.4587737843551798, + "grad_norm": 1.6745977401733398, + "learning_rate": 7.78397912072354e-06, + "loss": 0.4233197271823883, + "step": 690 + }, + { + "epoch": 1.463002114164905, + "grad_norm": 0.7401508092880249, + "learning_rate": 7.770408269148281e-06, + "loss": 1.1300601959228516, + "step": 692 + }, + { + "epoch": 1.46723044397463, + "grad_norm": 0.8936570286750793, + "learning_rate": 7.756809660919778e-06, + "loss": 0.5549295544624329, + "step": 694 + }, + { + "epoch": 1.4714587737843552, + "grad_norm": 2.4305765628814697, + "learning_rate": 7.74318346228648e-06, + "loss": 1.066399335861206, + "step": 696 + }, + { + "epoch": 1.4756871035940804, + "grad_norm": 1.2228820323944092, + "learning_rate": 7.72952983983414e-06, + "loss": 1.1287052631378174, + "step": 698 + }, + { + "epoch": 1.4799154334038054, + "grad_norm": 1.2910040616989136, + "learning_rate": 7.715848960483776e-06, + "loss": 0.9531479477882385, + "step": 700 + }, + { + "epoch": 1.4841437632135306, + "grad_norm": 1.1233556270599365, + "learning_rate": 7.702140991489637e-06, + "loss": 1.0696746110916138, + "step": 702 + }, + { + "epoch": 1.4883720930232558, + "grad_norm": 0.9682443141937256, + "learning_rate": 7.688406100437149e-06, + "loss": 0.9746364951133728, + "step": 704 + }, + { + "epoch": 1.492600422832981, + "grad_norm": 1.0571790933609009, + "learning_rate": 7.674644455240878e-06, + "loss": 0.771663248538971, + "step": 706 + }, + { + "epoch": 1.4968287526427062, + "grad_norm": 0.570301353931427, + "learning_rate": 7.660856224142456e-06, + "loss": 0.26455774903297424, + "step": 708 + }, + { + "epoch": 1.5010570824524314, + "grad_norm": 1.3108025789260864, + "learning_rate": 7.647041575708555e-06, + "loss": 0.5812461376190186, + "step": 710 + }, + { + "epoch": 1.5052854122621564, + "grad_norm": 2.110165596008301, + "learning_rate": 7.633200678828804e-06, + "loss": 0.6101235151290894, + "step": 712 + }, + { + "epoch": 1.5095137420718816, + "grad_norm": 1.8807343244552612, + "learning_rate": 7.6193337027137215e-06, + "loss": 0.9574325680732727, + "step": 714 + }, + { + "epoch": 1.5137420718816068, + "grad_norm": 0.7067351341247559, + "learning_rate": 7.605440816892667e-06, + "loss": 1.0287755727767944, + "step": 716 + }, + { + "epoch": 1.5179704016913318, + "grad_norm": 0.9900739192962646, + "learning_rate": 7.591522191211746e-06, + "loss": 0.9733884334564209, + "step": 718 + }, + { + "epoch": 1.522198731501057, + "grad_norm": 0.4219542145729065, + "learning_rate": 7.577577995831755e-06, + "loss": 0.8731890320777893, + "step": 720 + }, + { + "epoch": 1.5264270613107822, + "grad_norm": 0.9988309741020203, + "learning_rate": 7.563608401226081e-06, + "loss": 0.8608997464179993, + "step": 722 + }, + { + "epoch": 1.5306553911205074, + "grad_norm": 0.7088948488235474, + "learning_rate": 7.549613578178632e-06, + "loss": 1.0110266208648682, + "step": 724 + }, + { + "epoch": 1.5348837209302326, + "grad_norm": 1.178407073020935, + "learning_rate": 7.535593697781741e-06, + "loss": 0.8564161062240601, + "step": 726 + }, + { + "epoch": 1.5391120507399578, + "grad_norm": 2.1982457637786865, + "learning_rate": 7.52154893143408e-06, + "loss": 0.680258572101593, + "step": 728 + }, + { + "epoch": 1.543340380549683, + "grad_norm": 1.0706956386566162, + "learning_rate": 7.50747945083856e-06, + "loss": 0.8102659583091736, + "step": 730 + }, + { + "epoch": 1.547568710359408, + "grad_norm": 0.7179248929023743, + "learning_rate": 7.493385428000234e-06, + "loss": 1.0346068143844604, + "step": 732 + }, + { + "epoch": 1.5517970401691332, + "grad_norm": 0.5437570810317993, + "learning_rate": 7.479267035224191e-06, + "loss": 0.23715020716190338, + "step": 734 + }, + { + "epoch": 1.5560253699788582, + "grad_norm": 1.0202397108078003, + "learning_rate": 7.465124445113459e-06, + "loss": 1.0141202211380005, + "step": 736 + }, + { + "epoch": 1.5602536997885834, + "grad_norm": 0.3884512782096863, + "learning_rate": 7.450957830566878e-06, + "loss": 0.7230914831161499, + "step": 738 + }, + { + "epoch": 1.5644820295983086, + "grad_norm": 0.7907395362854004, + "learning_rate": 7.436767364777009e-06, + "loss": 1.0852365493774414, + "step": 740 + }, + { + "epoch": 1.5687103594080338, + "grad_norm": 5.782058238983154, + "learning_rate": 7.422553221227993e-06, + "loss": 0.7407255172729492, + "step": 742 + }, + { + "epoch": 1.572938689217759, + "grad_norm": 1.5651192665100098, + "learning_rate": 7.4083155736934484e-06, + "loss": 1.097520351409912, + "step": 744 + }, + { + "epoch": 1.5771670190274842, + "grad_norm": 1.5031559467315674, + "learning_rate": 7.394054596234336e-06, + "loss": 1.1895091533660889, + "step": 746 + }, + { + "epoch": 1.5813953488372094, + "grad_norm": 0.9550702571868896, + "learning_rate": 7.379770463196831e-06, + "loss": 0.919118344783783, + "step": 748 + }, + { + "epoch": 1.5856236786469344, + "grad_norm": 0.47285619378089905, + "learning_rate": 7.365463349210203e-06, + "loss": 0.05800021439790726, + "step": 750 + }, + { + "epoch": 1.5898520084566596, + "grad_norm": 1.1555415391921997, + "learning_rate": 7.351133429184666e-06, + "loss": 0.8398565053939819, + "step": 752 + }, + { + "epoch": 1.5940803382663846, + "grad_norm": 1.6445722579956055, + "learning_rate": 7.336780878309251e-06, + "loss": 1.162068486213684, + "step": 754 + }, + { + "epoch": 1.5983086680761098, + "grad_norm": 0.6990768313407898, + "learning_rate": 7.322405872049657e-06, + "loss": 1.1003162860870361, + "step": 756 + }, + { + "epoch": 1.602536997885835, + "grad_norm": 1.2485425472259521, + "learning_rate": 7.308008586146111e-06, + "loss": 1.0943663120269775, + "step": 758 + }, + { + "epoch": 1.6067653276955602, + "grad_norm": 0.9572765231132507, + "learning_rate": 7.293589196611221e-06, + "loss": 1.0679699182510376, + "step": 760 + }, + { + "epoch": 1.6109936575052854, + "grad_norm": 1.4260057210922241, + "learning_rate": 7.279147879727811e-06, + "loss": 1.1025015115737915, + "step": 762 + }, + { + "epoch": 1.6152219873150107, + "grad_norm": 1.420556664466858, + "learning_rate": 7.264684812046787e-06, + "loss": 0.991265594959259, + "step": 764 + }, + { + "epoch": 1.6194503171247359, + "grad_norm": 2.367309808731079, + "learning_rate": 7.250200170384961e-06, + "loss": 0.8480443954467773, + "step": 766 + }, + { + "epoch": 1.6236786469344608, + "grad_norm": 1.0605480670928955, + "learning_rate": 7.235694131822894e-06, + "loss": 1.1355984210968018, + "step": 768 + }, + { + "epoch": 1.627906976744186, + "grad_norm": 1.5513904094696045, + "learning_rate": 7.221166873702738e-06, + "loss": 1.063872218132019, + "step": 770 + }, + { + "epoch": 1.6321353065539113, + "grad_norm": 0.9517943859100342, + "learning_rate": 7.206618573626058e-06, + "loss": 1.0710200071334839, + "step": 772 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 1.2050000429153442, + "learning_rate": 7.19204940945167e-06, + "loss": 0.8725743889808655, + "step": 774 + }, + { + "epoch": 1.6405919661733614, + "grad_norm": 1.1810718774795532, + "learning_rate": 7.177459559293456e-06, + "loss": 1.1080224514007568, + "step": 776 + }, + { + "epoch": 1.6448202959830867, + "grad_norm": 1.5239593982696533, + "learning_rate": 7.162849201518198e-06, + "loss": 0.68960040807724, + "step": 778 + }, + { + "epoch": 1.6490486257928119, + "grad_norm": 0.8254817128181458, + "learning_rate": 7.1482185147433905e-06, + "loss": 1.0867798328399658, + "step": 780 + }, + { + "epoch": 1.653276955602537, + "grad_norm": 1.45622718334198, + "learning_rate": 7.133567677835055e-06, + "loss": 0.8074498772621155, + "step": 782 + }, + { + "epoch": 1.6575052854122623, + "grad_norm": 1.1257047653198242, + "learning_rate": 7.11889686990556e-06, + "loss": 1.1889450550079346, + "step": 784 + }, + { + "epoch": 1.6617336152219875, + "grad_norm": 1.2921396493911743, + "learning_rate": 7.104206270311423e-06, + "loss": 0.5138803720474243, + "step": 786 + }, + { + "epoch": 1.6659619450317125, + "grad_norm": 1.067741870880127, + "learning_rate": 7.089496058651133e-06, + "loss": 1.0809941291809082, + "step": 788 + }, + { + "epoch": 1.6701902748414377, + "grad_norm": 1.2185288667678833, + "learning_rate": 7.0747664147629335e-06, + "loss": 0.7209405899047852, + "step": 790 + }, + { + "epoch": 1.6744186046511627, + "grad_norm": 0.6701375842094421, + "learning_rate": 7.060017518722637e-06, + "loss": 0.3997204601764679, + "step": 792 + }, + { + "epoch": 1.6786469344608879, + "grad_norm": 1.2024998664855957, + "learning_rate": 7.0452495508414276e-06, + "loss": 1.048186182975769, + "step": 794 + }, + { + "epoch": 1.682875264270613, + "grad_norm": 0.5171627402305603, + "learning_rate": 7.030462691663641e-06, + "loss": 0.6009624600410461, + "step": 796 + }, + { + "epoch": 1.6871035940803383, + "grad_norm": 0.7701303958892822, + "learning_rate": 7.015657121964575e-06, + "loss": 1.1767216920852661, + "step": 798 + }, + { + "epoch": 1.6913319238900635, + "grad_norm": 0.7986422181129456, + "learning_rate": 7.000833022748267e-06, + "loss": 1.0685911178588867, + "step": 800 + }, + { + "epoch": 1.6955602536997887, + "grad_norm": 1.9579163789749146, + "learning_rate": 6.985990575245283e-06, + "loss": 0.6119225025177002, + "step": 802 + }, + { + "epoch": 1.699788583509514, + "grad_norm": 1.2787272930145264, + "learning_rate": 6.97112996091051e-06, + "loss": 1.03071129322052, + "step": 804 + }, + { + "epoch": 1.7040169133192389, + "grad_norm": 3.8576347827911377, + "learning_rate": 6.956251361420926e-06, + "loss": 0.7931581139564514, + "step": 806 + }, + { + "epoch": 1.708245243128964, + "grad_norm": 1.3824182748794556, + "learning_rate": 6.94135495867339e-06, + "loss": 0.7694694399833679, + "step": 808 + }, + { + "epoch": 1.712473572938689, + "grad_norm": 1.6486917734146118, + "learning_rate": 6.926440934782407e-06, + "loss": 1.099272608757019, + "step": 810 + }, + { + "epoch": 1.7167019027484143, + "grad_norm": 1.8737192153930664, + "learning_rate": 6.9115094720779155e-06, + "loss": 1.2829896211624146, + "step": 812 + }, + { + "epoch": 1.7209302325581395, + "grad_norm": 2.0316646099090576, + "learning_rate": 6.896560753103041e-06, + "loss": 1.0311399698257446, + "step": 814 + }, + { + "epoch": 1.7251585623678647, + "grad_norm": 1.0013114213943481, + "learning_rate": 6.881594960611881e-06, + "loss": 1.1290788650512695, + "step": 816 + }, + { + "epoch": 1.72938689217759, + "grad_norm": 0.8890924453735352, + "learning_rate": 6.866612277567261e-06, + "loss": 0.8836404085159302, + "step": 818 + }, + { + "epoch": 1.733615221987315, + "grad_norm": 6.471798896789551, + "learning_rate": 6.851612887138501e-06, + "loss": 0.9929148554801941, + "step": 820 + }, + { + "epoch": 1.7378435517970403, + "grad_norm": 1.1810201406478882, + "learning_rate": 6.836596972699175e-06, + "loss": 0.6458325982093811, + "step": 822 + }, + { + "epoch": 1.7420718816067653, + "grad_norm": 0.791428804397583, + "learning_rate": 6.82156471782487e-06, + "loss": 0.6101911067962646, + "step": 824 + }, + { + "epoch": 1.7463002114164905, + "grad_norm": 3.275075912475586, + "learning_rate": 6.806516306290942e-06, + "loss": 0.8959284424781799, + "step": 826 + }, + { + "epoch": 1.7505285412262155, + "grad_norm": 0.8022454977035522, + "learning_rate": 6.791451922070265e-06, + "loss": 0.9668054580688477, + "step": 828 + }, + { + "epoch": 1.7547568710359407, + "grad_norm": 2.344561815261841, + "learning_rate": 6.77637174933099e-06, + "loss": 0.7571685910224915, + "step": 830 + }, + { + "epoch": 1.758985200845666, + "grad_norm": 1.389305830001831, + "learning_rate": 6.7612759724342895e-06, + "loss": 1.1416950225830078, + "step": 832 + }, + { + "epoch": 1.763213530655391, + "grad_norm": 1.9522590637207031, + "learning_rate": 6.746164775932096e-06, + "loss": 0.7826639413833618, + "step": 834 + }, + { + "epoch": 1.7674418604651163, + "grad_norm": 0.9116184115409851, + "learning_rate": 6.731038344564862e-06, + "loss": 0.8100551962852478, + "step": 836 + }, + { + "epoch": 1.7716701902748415, + "grad_norm": 0.8732422590255737, + "learning_rate": 6.715896863259286e-06, + "loss": 1.0896806716918945, + "step": 838 + }, + { + "epoch": 1.7758985200845667, + "grad_norm": 1.0957021713256836, + "learning_rate": 6.700740517126061e-06, + "loss": 0.8103594183921814, + "step": 840 + }, + { + "epoch": 1.7801268498942917, + "grad_norm": 0.9628067016601562, + "learning_rate": 6.685569491457608e-06, + "loss": 0.35287776589393616, + "step": 842 + }, + { + "epoch": 1.784355179704017, + "grad_norm": 1.035485863685608, + "learning_rate": 6.670383971725809e-06, + "loss": 0.7141310572624207, + "step": 844 + }, + { + "epoch": 1.7885835095137421, + "grad_norm": 1.6569527387619019, + "learning_rate": 6.655184143579746e-06, + "loss": 0.9505995512008667, + "step": 846 + }, + { + "epoch": 1.792811839323467, + "grad_norm": 5.72770881652832, + "learning_rate": 6.639970192843423e-06, + "loss": 0.7124165892601013, + "step": 848 + }, + { + "epoch": 1.7970401691331923, + "grad_norm": 1.8030927181243896, + "learning_rate": 6.624742305513502e-06, + "loss": 0.743495523929596, + "step": 850 + }, + { + "epoch": 1.8012684989429175, + "grad_norm": 0.9843243956565857, + "learning_rate": 6.609500667757022e-06, + "loss": 0.9492723345756531, + "step": 852 + }, + { + "epoch": 1.8054968287526427, + "grad_norm": 0.5774413347244263, + "learning_rate": 6.594245465909127e-06, + "loss": 0.7611253261566162, + "step": 854 + }, + { + "epoch": 1.809725158562368, + "grad_norm": 1.112433671951294, + "learning_rate": 6.578976886470791e-06, + "loss": 1.0239204168319702, + "step": 856 + }, + { + "epoch": 1.8139534883720931, + "grad_norm": 0.8444233536720276, + "learning_rate": 6.56369511610653e-06, + "loss": 0.6033104658126831, + "step": 858 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 0.9478866457939148, + "learning_rate": 6.548400341642128e-06, + "loss": 1.0592596530914307, + "step": 860 + }, + { + "epoch": 1.8224101479915433, + "grad_norm": 0.8479084968566895, + "learning_rate": 6.533092750062345e-06, + "loss": 1.0542442798614502, + "step": 862 + }, + { + "epoch": 1.8266384778012685, + "grad_norm": 0.9617928266525269, + "learning_rate": 6.5177725285086416e-06, + "loss": 0.4782564342021942, + "step": 864 + }, + { + "epoch": 1.8308668076109935, + "grad_norm": 0.8183021545410156, + "learning_rate": 6.5024398642768785e-06, + "loss": 0.7408366203308105, + "step": 866 + }, + { + "epoch": 1.8350951374207187, + "grad_norm": 0.8027344942092896, + "learning_rate": 6.487094944815036e-06, + "loss": 1.0805916786193848, + "step": 868 + }, + { + "epoch": 1.839323467230444, + "grad_norm": 0.7138349413871765, + "learning_rate": 6.47173795772092e-06, + "loss": 1.0949031114578247, + "step": 870 + }, + { + "epoch": 1.8435517970401691, + "grad_norm": 0.3666520118713379, + "learning_rate": 6.456369090739869e-06, + "loss": 0.7219202518463135, + "step": 872 + }, + { + "epoch": 1.8477801268498943, + "grad_norm": 0.8815957307815552, + "learning_rate": 6.440988531762454e-06, + "loss": 1.0463707447052002, + "step": 874 + }, + { + "epoch": 1.8520084566596196, + "grad_norm": 4.529870510101318, + "learning_rate": 6.42559646882219e-06, + "loss": 0.8893094658851624, + "step": 876 + }, + { + "epoch": 1.8562367864693448, + "grad_norm": 1.5743250846862793, + "learning_rate": 6.410193090093227e-06, + "loss": 0.32593920826911926, + "step": 878 + }, + { + "epoch": 1.8604651162790697, + "grad_norm": 1.5475246906280518, + "learning_rate": 6.394778583888064e-06, + "loss": 0.683530867099762, + "step": 880 + }, + { + "epoch": 1.864693446088795, + "grad_norm": 1.3281298875808716, + "learning_rate": 6.379353138655227e-06, + "loss": 1.0770466327667236, + "step": 882 + }, + { + "epoch": 1.86892177589852, + "grad_norm": 0.7169983983039856, + "learning_rate": 6.3639169429769805e-06, + "loss": 0.7519739866256714, + "step": 884 + }, + { + "epoch": 1.8731501057082451, + "grad_norm": 1.4231908321380615, + "learning_rate": 6.348470185567025e-06, + "loss": 1.128954291343689, + "step": 886 + }, + { + "epoch": 1.8773784355179703, + "grad_norm": 0.7575823068618774, + "learning_rate": 6.333013055268169e-06, + "loss": 1.1132702827453613, + "step": 888 + }, + { + "epoch": 1.8816067653276956, + "grad_norm": 1.213281273841858, + "learning_rate": 6.317545741050044e-06, + "loss": 0.579425573348999, + "step": 890 + }, + { + "epoch": 1.8858350951374208, + "grad_norm": 2.6736762523651123, + "learning_rate": 6.3020684320067805e-06, + "loss": 0.8137954473495483, + "step": 892 + }, + { + "epoch": 1.890063424947146, + "grad_norm": 0.8506104946136475, + "learning_rate": 6.2865813173546976e-06, + "loss": 1.07477867603302, + "step": 894 + }, + { + "epoch": 1.8942917547568712, + "grad_norm": 2.8496999740600586, + "learning_rate": 6.271084586429994e-06, + "loss": 1.381723165512085, + "step": 896 + }, + { + "epoch": 1.8985200845665962, + "grad_norm": 1.0546894073486328, + "learning_rate": 6.255578428686433e-06, + "loss": 1.0150777101516724, + "step": 898 + }, + { + "epoch": 1.9027484143763214, + "grad_norm": 1.9049346446990967, + "learning_rate": 6.240063033693023e-06, + "loss": 0.5043352842330933, + "step": 900 + }, + { + "epoch": 1.9069767441860463, + "grad_norm": 0.6271165609359741, + "learning_rate": 6.2245385911316994e-06, + "loss": 1.0388513803482056, + "step": 902 + }, + { + "epoch": 1.9112050739957716, + "grad_norm": 1.6291722059249878, + "learning_rate": 6.2090052907950096e-06, + "loss": 0.5244297981262207, + "step": 904 + }, + { + "epoch": 1.9154334038054968, + "grad_norm": 1.0104281902313232, + "learning_rate": 6.193463322583792e-06, + "loss": 0.8711696863174438, + "step": 906 + }, + { + "epoch": 1.919661733615222, + "grad_norm": 1.5567266941070557, + "learning_rate": 6.177912876504851e-06, + "loss": 1.2081165313720703, + "step": 908 + }, + { + "epoch": 1.9238900634249472, + "grad_norm": 0.7997756004333496, + "learning_rate": 6.16235414266864e-06, + "loss": 0.7782785892486572, + "step": 910 + }, + { + "epoch": 1.9281183932346724, + "grad_norm": 1.074521541595459, + "learning_rate": 6.146787311286927e-06, + "loss": 1.010573148727417, + "step": 912 + }, + { + "epoch": 1.9323467230443976, + "grad_norm": 1.1430543661117554, + "learning_rate": 6.131212572670483e-06, + "loss": 0.8374419212341309, + "step": 914 + }, + { + "epoch": 1.9365750528541226, + "grad_norm": 0.8561668992042542, + "learning_rate": 6.115630117226741e-06, + "loss": 1.0769456624984741, + "step": 916 + }, + { + "epoch": 1.9408033826638478, + "grad_norm": 1.3352936506271362, + "learning_rate": 6.1000401354574855e-06, + "loss": 0.6508890390396118, + "step": 918 + }, + { + "epoch": 1.945031712473573, + "grad_norm": 1.5549503564834595, + "learning_rate": 6.084442817956502e-06, + "loss": 0.9323084354400635, + "step": 920 + }, + { + "epoch": 1.949260042283298, + "grad_norm": 1.8166402578353882, + "learning_rate": 6.0688383554072625e-06, + "loss": 0.9818851351737976, + "step": 922 + }, + { + "epoch": 1.9534883720930232, + "grad_norm": 0.7917953133583069, + "learning_rate": 6.053226938580596e-06, + "loss": 0.9979840517044067, + "step": 924 + }, + { + "epoch": 1.9577167019027484, + "grad_norm": 0.8893802762031555, + "learning_rate": 6.03760875833234e-06, + "loss": 0.8576878309249878, + "step": 926 + }, + { + "epoch": 1.9619450317124736, + "grad_norm": 1.1501379013061523, + "learning_rate": 6.021984005601027e-06, + "loss": 1.098677158355713, + "step": 928 + }, + { + "epoch": 1.9661733615221988, + "grad_norm": 0.8236111402511597, + "learning_rate": 6.006352871405534e-06, + "loss": 0.8075918555259705, + "step": 930 + }, + { + "epoch": 1.970401691331924, + "grad_norm": 0.9850919842720032, + "learning_rate": 5.990715546842755e-06, + "loss": 1.2170652151107788, + "step": 932 + }, + { + "epoch": 1.9746300211416492, + "grad_norm": 1.1624720096588135, + "learning_rate": 5.9750722230852694e-06, + "loss": 0.8378693461418152, + "step": 934 + }, + { + "epoch": 1.9788583509513742, + "grad_norm": 1.2153650522232056, + "learning_rate": 5.959423091378992e-06, + "loss": 0.6704339981079102, + "step": 936 + }, + { + "epoch": 1.9830866807610994, + "grad_norm": 4.025911808013916, + "learning_rate": 5.943768343040845e-06, + "loss": 0.8145028352737427, + "step": 938 + }, + { + "epoch": 1.9873150105708244, + "grad_norm": 0.8239069581031799, + "learning_rate": 5.928108169456419e-06, + "loss": 1.1152684688568115, + "step": 940 + }, + { + "epoch": 1.9915433403805496, + "grad_norm": 0.8116897344589233, + "learning_rate": 5.912442762077624e-06, + "loss": 1.0400241613388062, + "step": 942 + }, + { + "epoch": 1.9957716701902748, + "grad_norm": 1.0503615140914917, + "learning_rate": 5.89677231242036e-06, + "loss": 1.0948755741119385, + "step": 944 + }, + { + "epoch": 2.0, + "grad_norm": 1.0883324146270752, + "learning_rate": 5.8810970120621714e-06, + "loss": 0.45772236585617065, + "step": 946 + }, + { + "epoch": 2.004228329809725, + "grad_norm": 5.065436363220215, + "learning_rate": 5.8654170526399014e-06, + "loss": 0.7291642427444458, + "step": 948 + }, + { + "epoch": 2.0084566596194504, + "grad_norm": 0.7658670544624329, + "learning_rate": 5.849732625847354e-06, + "loss": 0.9443368911743164, + "step": 950 + }, + { + "epoch": 2.0126849894291756, + "grad_norm": 1.391258955001831, + "learning_rate": 5.834043923432947e-06, + "loss": 0.8817076086997986, + "step": 952 + }, + { + "epoch": 2.016913319238901, + "grad_norm": 1.4591436386108398, + "learning_rate": 5.818351137197373e-06, + "loss": 1.1189929246902466, + "step": 954 + }, + { + "epoch": 2.0211416490486256, + "grad_norm": 2.659818410873413, + "learning_rate": 5.802654458991245e-06, + "loss": 1.1417254209518433, + "step": 956 + }, + { + "epoch": 2.025369978858351, + "grad_norm": 1.5345180034637451, + "learning_rate": 5.786954080712763e-06, + "loss": 0.6548402309417725, + "step": 958 + }, + { + "epoch": 2.029598308668076, + "grad_norm": 1.0098344087600708, + "learning_rate": 5.7712501943053576e-06, + "loss": 0.9253395795822144, + "step": 960 + }, + { + "epoch": 2.033826638477801, + "grad_norm": 2.0480527877807617, + "learning_rate": 5.755542991755352e-06, + "loss": 0.8580033779144287, + "step": 962 + }, + { + "epoch": 2.0380549682875264, + "grad_norm": 2.5995616912841797, + "learning_rate": 5.739832665089604e-06, + "loss": 0.7364559769630432, + "step": 964 + }, + { + "epoch": 2.0422832980972516, + "grad_norm": 0.9442983269691467, + "learning_rate": 5.724119406373171e-06, + "loss": 0.607407808303833, + "step": 966 + }, + { + "epoch": 2.046511627906977, + "grad_norm": 3.280029535293579, + "learning_rate": 5.708403407706954e-06, + "loss": 0.7525492906570435, + "step": 968 + }, + { + "epoch": 2.050739957716702, + "grad_norm": 1.4883980751037598, + "learning_rate": 5.692684861225349e-06, + "loss": 0.9570878148078918, + "step": 970 + }, + { + "epoch": 2.0549682875264272, + "grad_norm": 2.017801284790039, + "learning_rate": 5.676963959093904e-06, + "loss": 0.5232957601547241, + "step": 972 + }, + { + "epoch": 2.059196617336152, + "grad_norm": 0.9809946417808533, + "learning_rate": 5.66124089350696e-06, + "loss": 0.9056997299194336, + "step": 974 + }, + { + "epoch": 2.063424947145877, + "grad_norm": 0.7657989263534546, + "learning_rate": 5.645515856685316e-06, + "loss": 0.9179431200027466, + "step": 976 + }, + { + "epoch": 2.0676532769556024, + "grad_norm": 0.9482935070991516, + "learning_rate": 5.629789040873862e-06, + "loss": 0.979408323764801, + "step": 978 + }, + { + "epoch": 2.0718816067653276, + "grad_norm": 2.098449945449829, + "learning_rate": 5.614060638339238e-06, + "loss": 0.7373460531234741, + "step": 980 + }, + { + "epoch": 2.076109936575053, + "grad_norm": 2.790154218673706, + "learning_rate": 5.598330841367487e-06, + "loss": 0.6838187575340271, + "step": 982 + }, + { + "epoch": 2.080338266384778, + "grad_norm": 1.1686556339263916, + "learning_rate": 5.582599842261693e-06, + "loss": 0.6918951272964478, + "step": 984 + }, + { + "epoch": 2.0845665961945032, + "grad_norm": 2.0469274520874023, + "learning_rate": 5.566867833339644e-06, + "loss": 0.8548566102981567, + "step": 986 + }, + { + "epoch": 2.0887949260042284, + "grad_norm": 1.033456802368164, + "learning_rate": 5.551135006931467e-06, + "loss": 0.9515693783760071, + "step": 988 + }, + { + "epoch": 2.0930232558139537, + "grad_norm": 2.8752598762512207, + "learning_rate": 5.535401555377287e-06, + "loss": 0.9892008304595947, + "step": 990 + }, + { + "epoch": 2.097251585623679, + "grad_norm": 4.230721950531006, + "learning_rate": 5.51966767102487e-06, + "loss": 0.5742906928062439, + "step": 992 + }, + { + "epoch": 2.1014799154334036, + "grad_norm": 2.057455062866211, + "learning_rate": 5.503933546227273e-06, + "loss": 0.6247528791427612, + "step": 994 + }, + { + "epoch": 2.105708245243129, + "grad_norm": 1.6091196537017822, + "learning_rate": 5.4881993733404916e-06, + "loss": 0.7494651675224304, + "step": 996 + }, + { + "epoch": 2.109936575052854, + "grad_norm": 1.0004706382751465, + "learning_rate": 5.472465344721113e-06, + "loss": 0.8864080905914307, + "step": 998 + }, + { + "epoch": 2.1141649048625792, + "grad_norm": 0.9034138917922974, + "learning_rate": 5.456731652723959e-06, + "loss": 0.9213463068008423, + "step": 1000 + }, + { + "epoch": 2.1183932346723044, + "grad_norm": 0.7893651127815247, + "learning_rate": 5.440998489699731e-06, + "loss": 0.23825736343860626, + "step": 1002 + }, + { + "epoch": 2.1226215644820297, + "grad_norm": 1.727262020111084, + "learning_rate": 5.425266047992672e-06, + "loss": 0.6444735527038574, + "step": 1004 + }, + { + "epoch": 2.126849894291755, + "grad_norm": 1.7172200679779053, + "learning_rate": 5.4095345199382e-06, + "loss": 0.9326093196868896, + "step": 1006 + }, + { + "epoch": 2.13107822410148, + "grad_norm": 1.124970555305481, + "learning_rate": 5.393804097860568e-06, + "loss": 0.8703416585922241, + "step": 1008 + }, + { + "epoch": 2.1353065539112053, + "grad_norm": 0.8798812627792358, + "learning_rate": 5.3780749740705065e-06, + "loss": 0.565904974937439, + "step": 1010 + }, + { + "epoch": 2.13953488372093, + "grad_norm": 0.4790954291820526, + "learning_rate": 5.362347340862869e-06, + "loss": 0.4800865054130554, + "step": 1012 + }, + { + "epoch": 2.1437632135306552, + "grad_norm": 5.961706161499023, + "learning_rate": 5.3466213905142935e-06, + "loss": 0.6189907193183899, + "step": 1014 + }, + { + "epoch": 2.1479915433403804, + "grad_norm": 0.9315617680549622, + "learning_rate": 5.3308973152808395e-06, + "loss": 0.779002845287323, + "step": 1016 + }, + { + "epoch": 2.1522198731501057, + "grad_norm": 1.0128082036972046, + "learning_rate": 5.315175307395644e-06, + "loss": 0.9131028652191162, + "step": 1018 + }, + { + "epoch": 2.156448202959831, + "grad_norm": 2.6575233936309814, + "learning_rate": 5.299455559066571e-06, + "loss": 0.7087488174438477, + "step": 1020 + }, + { + "epoch": 2.160676532769556, + "grad_norm": 0.9301908612251282, + "learning_rate": 5.283738262473859e-06, + "loss": 0.6987777352333069, + "step": 1022 + }, + { + "epoch": 2.1649048625792813, + "grad_norm": 1.260840892791748, + "learning_rate": 5.268023609767772e-06, + "loss": 0.8693367838859558, + "step": 1024 + }, + { + "epoch": 2.1691331923890065, + "grad_norm": 0.8487909436225891, + "learning_rate": 5.2523117930662545e-06, + "loss": 0.9945523142814636, + "step": 1026 + }, + { + "epoch": 2.1733615221987317, + "grad_norm": 1.1430577039718628, + "learning_rate": 5.2366030044525785e-06, + "loss": 0.9062778949737549, + "step": 1028 + }, + { + "epoch": 2.177589852008457, + "grad_norm": 1.0969204902648926, + "learning_rate": 5.220897435972992e-06, + "loss": 0.8596665263175964, + "step": 1030 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 1.7309213876724243, + "learning_rate": 5.205195279634385e-06, + "loss": 0.9404520988464355, + "step": 1032 + }, + { + "epoch": 2.186046511627907, + "grad_norm": 0.9761110544204712, + "learning_rate": 5.1894967274019245e-06, + "loss": 1.0071791410446167, + "step": 1034 + }, + { + "epoch": 2.190274841437632, + "grad_norm": 1.6570290327072144, + "learning_rate": 5.173801971196716e-06, + "loss": 0.6287547945976257, + "step": 1036 + }, + { + "epoch": 2.1945031712473573, + "grad_norm": 1.7801464796066284, + "learning_rate": 5.158111202893464e-06, + "loss": 0.3434582054615021, + "step": 1038 + }, + { + "epoch": 2.1987315010570825, + "grad_norm": 1.918614149093628, + "learning_rate": 5.142424614318112e-06, + "loss": 0.963275134563446, + "step": 1040 + }, + { + "epoch": 2.2029598308668077, + "grad_norm": 6.645174503326416, + "learning_rate": 5.12674239724551e-06, + "loss": 0.3326680660247803, + "step": 1042 + }, + { + "epoch": 2.207188160676533, + "grad_norm": 1.5196832418441772, + "learning_rate": 5.111064743397061e-06, + "loss": 0.9226099252700806, + "step": 1044 + }, + { + "epoch": 2.211416490486258, + "grad_norm": 1.1765769720077515, + "learning_rate": 5.09539184443838e-06, + "loss": 0.47022897005081177, + "step": 1046 + }, + { + "epoch": 2.2156448202959833, + "grad_norm": 3.335059642791748, + "learning_rate": 5.079723891976956e-06, + "loss": 0.842235267162323, + "step": 1048 + }, + { + "epoch": 2.219873150105708, + "grad_norm": 1.475996732711792, + "learning_rate": 5.064061077559803e-06, + "loss": 0.5239929556846619, + "step": 1050 + }, + { + "epoch": 2.2241014799154333, + "grad_norm": 1.1232885122299194, + "learning_rate": 5.048403592671116e-06, + "loss": 0.8301411271095276, + "step": 1052 + }, + { + "epoch": 2.2283298097251585, + "grad_norm": 2.056427240371704, + "learning_rate": 5.032751628729943e-06, + "loss": 0.9392002820968628, + "step": 1054 + }, + { + "epoch": 2.2325581395348837, + "grad_norm": 0.9355063438415527, + "learning_rate": 5.017105377087829e-06, + "loss": 0.8929227590560913, + "step": 1056 + }, + { + "epoch": 2.236786469344609, + "grad_norm": 0.3439593017101288, + "learning_rate": 5.0014650290264854e-06, + "loss": 0.6246748566627502, + "step": 1058 + }, + { + "epoch": 2.241014799154334, + "grad_norm": 1.0592577457427979, + "learning_rate": 4.9858307757554536e-06, + "loss": 0.737403392791748, + "step": 1060 + }, + { + "epoch": 2.2452431289640593, + "grad_norm": 1.1584621667861938, + "learning_rate": 4.97020280840976e-06, + "loss": 0.911266565322876, + "step": 1062 + }, + { + "epoch": 2.2494714587737845, + "grad_norm": 2.0557973384857178, + "learning_rate": 4.954581318047582e-06, + "loss": 0.8674002289772034, + "step": 1064 + }, + { + "epoch": 2.2536997885835097, + "grad_norm": 1.5586539506912231, + "learning_rate": 4.938966495647918e-06, + "loss": 0.5012240409851074, + "step": 1066 + }, + { + "epoch": 2.2579281183932345, + "grad_norm": 2.3500895500183105, + "learning_rate": 4.923358532108244e-06, + "loss": 0.7392663359642029, + "step": 1068 + }, + { + "epoch": 2.2621564482029597, + "grad_norm": 2.458981513977051, + "learning_rate": 4.907757618242186e-06, + "loss": 0.699637770652771, + "step": 1070 + }, + { + "epoch": 2.266384778012685, + "grad_norm": 1.054579496383667, + "learning_rate": 4.892163944777181e-06, + "loss": 0.9284424781799316, + "step": 1072 + }, + { + "epoch": 2.27061310782241, + "grad_norm": 0.6797510385513306, + "learning_rate": 4.876577702352155e-06, + "loss": 0.6487488746643066, + "step": 1074 + }, + { + "epoch": 2.2748414376321353, + "grad_norm": 1.76516592502594, + "learning_rate": 4.860999081515183e-06, + "loss": 0.6723374128341675, + "step": 1076 + }, + { + "epoch": 2.2790697674418605, + "grad_norm": 6.648386478424072, + "learning_rate": 4.845428272721161e-06, + "loss": 0.20806623995304108, + "step": 1078 + }, + { + "epoch": 2.2832980972515857, + "grad_norm": 1.0107231140136719, + "learning_rate": 4.8298654663294865e-06, + "loss": 0.14377908408641815, + "step": 1080 + }, + { + "epoch": 2.287526427061311, + "grad_norm": 0.4066459834575653, + "learning_rate": 4.814310852601717e-06, + "loss": 0.46209123730659485, + "step": 1082 + }, + { + "epoch": 2.291754756871036, + "grad_norm": 1.234228491783142, + "learning_rate": 4.7987646216992545e-06, + "loss": 0.3461954593658447, + "step": 1084 + }, + { + "epoch": 2.295983086680761, + "grad_norm": 2.384023427963257, + "learning_rate": 4.783226963681018e-06, + "loss": 0.15366527438163757, + "step": 1086 + }, + { + "epoch": 2.300211416490486, + "grad_norm": 1.5009163618087769, + "learning_rate": 4.76769806850112e-06, + "loss": 0.6414340734481812, + "step": 1088 + }, + { + "epoch": 2.3044397463002113, + "grad_norm": 5.656238555908203, + "learning_rate": 4.752178126006544e-06, + "loss": 0.8150467276573181, + "step": 1090 + }, + { + "epoch": 2.3086680761099365, + "grad_norm": 1.2662545442581177, + "learning_rate": 4.736667325934819e-06, + "loss": 0.5979265570640564, + "step": 1092 + }, + { + "epoch": 2.3128964059196617, + "grad_norm": 1.0322731733322144, + "learning_rate": 4.721165857911709e-06, + "loss": 0.931814968585968, + "step": 1094 + }, + { + "epoch": 2.317124735729387, + "grad_norm": 2.8270702362060547, + "learning_rate": 4.705673911448891e-06, + "loss": 0.844617486000061, + "step": 1096 + }, + { + "epoch": 2.321353065539112, + "grad_norm": 1.0264586210250854, + "learning_rate": 4.690191675941631e-06, + "loss": 0.8917978405952454, + "step": 1098 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 4.049459934234619, + "learning_rate": 4.674719340666477e-06, + "loss": 0.49443334341049194, + "step": 1100 + }, + { + "epoch": 2.3298097251585626, + "grad_norm": 1.3870093822479248, + "learning_rate": 4.659257094778949e-06, + "loss": 0.6575090289115906, + "step": 1102 + }, + { + "epoch": 2.3340380549682873, + "grad_norm": 0.8333762288093567, + "learning_rate": 4.643805127311212e-06, + "loss": 0.9781821370124817, + "step": 1104 + }, + { + "epoch": 2.3382663847780125, + "grad_norm": 1.036359190940857, + "learning_rate": 4.628363627169774e-06, + "loss": 0.8609423637390137, + "step": 1106 + }, + { + "epoch": 2.3424947145877377, + "grad_norm": 1.107252836227417, + "learning_rate": 4.612932783133182e-06, + "loss": 0.834477961063385, + "step": 1108 + }, + { + "epoch": 2.346723044397463, + "grad_norm": 3.99241304397583, + "learning_rate": 4.5975127838497036e-06, + "loss": 0.9674366116523743, + "step": 1110 + }, + { + "epoch": 2.350951374207188, + "grad_norm": 0.4219377934932709, + "learning_rate": 4.582103817835021e-06, + "loss": 0.566524863243103, + "step": 1112 + }, + { + "epoch": 2.3551797040169133, + "grad_norm": 3.2431130409240723, + "learning_rate": 4.566706073469938e-06, + "loss": 0.7969292998313904, + "step": 1114 + }, + { + "epoch": 2.3594080338266386, + "grad_norm": 1.186369776725769, + "learning_rate": 4.551319738998067e-06, + "loss": 0.5491366982460022, + "step": 1116 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 3.0447516441345215, + "learning_rate": 4.535945002523524e-06, + "loss": 1.1090550422668457, + "step": 1118 + }, + { + "epoch": 2.367864693446089, + "grad_norm": 1.0337697267532349, + "learning_rate": 4.5205820520086415e-06, + "loss": 0.8828186392784119, + "step": 1120 + }, + { + "epoch": 2.3720930232558137, + "grad_norm": 1.4636832475662231, + "learning_rate": 4.505231075271662e-06, + "loss": 0.7621970772743225, + "step": 1122 + }, + { + "epoch": 2.376321353065539, + "grad_norm": 5.581032752990723, + "learning_rate": 4.489892259984446e-06, + "loss": 0.6078689098358154, + "step": 1124 + }, + { + "epoch": 2.380549682875264, + "grad_norm": 3.288517713546753, + "learning_rate": 4.474565793670167e-06, + "loss": 0.9828864336013794, + "step": 1126 + }, + { + "epoch": 2.3847780126849893, + "grad_norm": 1.8655986785888672, + "learning_rate": 4.459251863701038e-06, + "loss": 0.6074872016906738, + "step": 1128 + }, + { + "epoch": 2.3890063424947146, + "grad_norm": 1.2905350923538208, + "learning_rate": 4.443950657296004e-06, + "loss": 0.7135834097862244, + "step": 1130 + }, + { + "epoch": 2.3932346723044398, + "grad_norm": 1.9687215089797974, + "learning_rate": 4.4286623615184585e-06, + "loss": 0.6180700659751892, + "step": 1132 + }, + { + "epoch": 2.397463002114165, + "grad_norm": 1.1242769956588745, + "learning_rate": 4.413387163273964e-06, + "loss": 0.941765546798706, + "step": 1134 + }, + { + "epoch": 2.40169133192389, + "grad_norm": 3.4840548038482666, + "learning_rate": 4.3981252493079506e-06, + "loss": 0.633338987827301, + "step": 1136 + }, + { + "epoch": 2.4059196617336154, + "grad_norm": 2.4682834148406982, + "learning_rate": 4.382876806203453e-06, + "loss": 0.7694275975227356, + "step": 1138 + }, + { + "epoch": 2.41014799154334, + "grad_norm": 1.0901868343353271, + "learning_rate": 4.3676420203788125e-06, + "loss": 0.8488864302635193, + "step": 1140 + }, + { + "epoch": 2.4143763213530653, + "grad_norm": 1.1759238243103027, + "learning_rate": 4.352421078085405e-06, + "loss": 0.989240288734436, + "step": 1142 + }, + { + "epoch": 2.4186046511627906, + "grad_norm": 1.373141884803772, + "learning_rate": 4.337214165405368e-06, + "loss": 0.5142699480056763, + "step": 1144 + }, + { + "epoch": 2.4228329809725158, + "grad_norm": 3.2051053047180176, + "learning_rate": 4.322021468249316e-06, + "loss": 0.7031230330467224, + "step": 1146 + }, + { + "epoch": 2.427061310782241, + "grad_norm": 0.7673659920692444, + "learning_rate": 4.306843172354076e-06, + "loss": 0.6546354293823242, + "step": 1148 + }, + { + "epoch": 2.431289640591966, + "grad_norm": 1.1194285154342651, + "learning_rate": 4.291679463280416e-06, + "loss": 0.5144006013870239, + "step": 1150 + }, + { + "epoch": 2.4355179704016914, + "grad_norm": 0.9611324667930603, + "learning_rate": 4.276530526410768e-06, + "loss": 0.5544730424880981, + "step": 1152 + }, + { + "epoch": 2.4397463002114166, + "grad_norm": 0.556056797504425, + "learning_rate": 4.261396546946972e-06, + "loss": 0.46504560112953186, + "step": 1154 + }, + { + "epoch": 2.443974630021142, + "grad_norm": 0.8544995784759521, + "learning_rate": 4.246277709908009e-06, + "loss": 0.919431209564209, + "step": 1156 + }, + { + "epoch": 2.448202959830867, + "grad_norm": 0.9173510670661926, + "learning_rate": 4.231174200127736e-06, + "loss": 0.4963262975215912, + "step": 1158 + }, + { + "epoch": 2.452431289640592, + "grad_norm": 0.7897346615791321, + "learning_rate": 4.216086202252624e-06, + "loss": 0.9528647065162659, + "step": 1160 + }, + { + "epoch": 2.456659619450317, + "grad_norm": 0.9443305730819702, + "learning_rate": 4.201013900739513e-06, + "loss": 0.861652135848999, + "step": 1162 + }, + { + "epoch": 2.460887949260042, + "grad_norm": 1.0425554513931274, + "learning_rate": 4.185957479853344e-06, + "loss": 0.7021121382713318, + "step": 1164 + }, + { + "epoch": 2.4651162790697674, + "grad_norm": 2.089914083480835, + "learning_rate": 4.1709171236649096e-06, + "loss": 1.0287466049194336, + "step": 1166 + }, + { + "epoch": 2.4693446088794926, + "grad_norm": 1.4080147743225098, + "learning_rate": 4.155893016048607e-06, + "loss": 1.1282658576965332, + "step": 1168 + }, + { + "epoch": 2.473572938689218, + "grad_norm": 1.826158881187439, + "learning_rate": 4.140885340680189e-06, + "loss": 0.6561502814292908, + "step": 1170 + }, + { + "epoch": 2.477801268498943, + "grad_norm": 2.83551287651062, + "learning_rate": 4.125894281034517e-06, + "loss": 0.7542917132377625, + "step": 1172 + }, + { + "epoch": 2.482029598308668, + "grad_norm": 2.118927001953125, + "learning_rate": 4.110920020383316e-06, + "loss": 0.6499738097190857, + "step": 1174 + }, + { + "epoch": 2.4862579281183934, + "grad_norm": 0.8958560228347778, + "learning_rate": 4.0959627417929386e-06, + "loss": 0.6449573040008545, + "step": 1176 + }, + { + "epoch": 2.4904862579281186, + "grad_norm": 1.7721506357192993, + "learning_rate": 4.081022628122127e-06, + "loss": 0.7136943936347961, + "step": 1178 + }, + { + "epoch": 2.4947145877378434, + "grad_norm": 1.8657937049865723, + "learning_rate": 4.06609986201977e-06, + "loss": 0.6413277387619019, + "step": 1180 + }, + { + "epoch": 2.4989429175475686, + "grad_norm": 0.5630549192428589, + "learning_rate": 4.051194625922681e-06, + "loss": 0.3792707026004791, + "step": 1182 + }, + { + "epoch": 2.503171247357294, + "grad_norm": 1.5422683954238892, + "learning_rate": 4.036307102053359e-06, + "loss": 0.16814373433589935, + "step": 1184 + }, + { + "epoch": 2.507399577167019, + "grad_norm": 1.139864206314087, + "learning_rate": 4.021437472417763e-06, + "loss": 0.8461377024650574, + "step": 1186 + }, + { + "epoch": 2.511627906976744, + "grad_norm": 2.5258994102478027, + "learning_rate": 4.006585918803091e-06, + "loss": 0.7749978303909302, + "step": 1188 + }, + { + "epoch": 2.5158562367864694, + "grad_norm": 2.815574884414673, + "learning_rate": 3.991752622775551e-06, + "loss": 0.6539379358291626, + "step": 1190 + }, + { + "epoch": 2.5200845665961946, + "grad_norm": 1.038074254989624, + "learning_rate": 3.97693776567815e-06, + "loss": 0.850276529788971, + "step": 1192 + }, + { + "epoch": 2.52431289640592, + "grad_norm": 0.8128182291984558, + "learning_rate": 3.962141528628465e-06, + "loss": 0.5508623719215393, + "step": 1194 + }, + { + "epoch": 2.528541226215645, + "grad_norm": 3.49111270904541, + "learning_rate": 3.947364092516437e-06, + "loss": 0.8177246451377869, + "step": 1196 + }, + { + "epoch": 2.53276955602537, + "grad_norm": 1.2982627153396606, + "learning_rate": 3.932605638002163e-06, + "loss": 0.87604159116745, + "step": 1198 + }, + { + "epoch": 2.536997885835095, + "grad_norm": 1.8900153636932373, + "learning_rate": 3.917866345513677e-06, + "loss": 0.6850084066390991, + "step": 1200 + }, + { + "epoch": 2.54122621564482, + "grad_norm": 1.0354785919189453, + "learning_rate": 3.903146395244749e-06, + "loss": 0.8515037894248962, + "step": 1202 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 3.221052885055542, + "learning_rate": 3.888445967152691e-06, + "loss": 0.3199499249458313, + "step": 1204 + }, + { + "epoch": 2.5496828752642706, + "grad_norm": 1.9412068128585815, + "learning_rate": 3.873765240956138e-06, + "loss": 0.9329533576965332, + "step": 1206 + }, + { + "epoch": 2.553911205073996, + "grad_norm": 1.1236411333084106, + "learning_rate": 3.859104396132863e-06, + "loss": 0.9284213781356812, + "step": 1208 + }, + { + "epoch": 2.558139534883721, + "grad_norm": 1.356223464012146, + "learning_rate": 3.84446361191759e-06, + "loss": 0.9393802285194397, + "step": 1210 + }, + { + "epoch": 2.5623678646934462, + "grad_norm": 1.9813258647918701, + "learning_rate": 3.829843067299788e-06, + "loss": 0.64601069688797, + "step": 1212 + }, + { + "epoch": 2.5665961945031714, + "grad_norm": 1.3942888975143433, + "learning_rate": 3.8152429410214834e-06, + "loss": 0.9045993685722351, + "step": 1214 + }, + { + "epoch": 2.570824524312896, + "grad_norm": 2.0219480991363525, + "learning_rate": 3.8006634115750886e-06, + "loss": 0.5468924641609192, + "step": 1216 + }, + { + "epoch": 2.5750528541226214, + "grad_norm": 1.2288607358932495, + "learning_rate": 3.7861046572012105e-06, + "loss": 0.4949299991130829, + "step": 1218 + }, + { + "epoch": 2.5792811839323466, + "grad_norm": 1.714388370513916, + "learning_rate": 3.771566855886467e-06, + "loss": 0.4390013515949249, + "step": 1220 + }, + { + "epoch": 2.583509513742072, + "grad_norm": 0.7751480340957642, + "learning_rate": 3.7570501853613227e-06, + "loss": 0.48840633034706116, + "step": 1222 + }, + { + "epoch": 2.587737843551797, + "grad_norm": 2.781423568725586, + "learning_rate": 3.742554823097905e-06, + "loss": 0.550344705581665, + "step": 1224 + }, + { + "epoch": 2.5919661733615222, + "grad_norm": 0.4711180627346039, + "learning_rate": 3.7280809463078436e-06, + "loss": 0.34713348746299744, + "step": 1226 + }, + { + "epoch": 2.5961945031712474, + "grad_norm": 1.165834903717041, + "learning_rate": 3.7136287319400944e-06, + "loss": 0.6826978921890259, + "step": 1228 + }, + { + "epoch": 2.6004228329809727, + "grad_norm": 4.001968860626221, + "learning_rate": 3.6991983566787847e-06, + "loss": 0.722435474395752, + "step": 1230 + }, + { + "epoch": 2.604651162790698, + "grad_norm": 2.9324543476104736, + "learning_rate": 3.6847899969410515e-06, + "loss": 0.5727831125259399, + "step": 1232 + }, + { + "epoch": 2.6088794926004226, + "grad_norm": 2.9382364749908447, + "learning_rate": 3.6704038288748812e-06, + "loss": 0.6767183542251587, + "step": 1234 + }, + { + "epoch": 2.613107822410148, + "grad_norm": 1.4194291830062866, + "learning_rate": 3.6560400283569543e-06, + "loss": 0.7603731155395508, + "step": 1236 + }, + { + "epoch": 2.617336152219873, + "grad_norm": 1.4806771278381348, + "learning_rate": 3.641698770990507e-06, + "loss": 0.795750081539154, + "step": 1238 + }, + { + "epoch": 2.6215644820295982, + "grad_norm": 1.2420991659164429, + "learning_rate": 3.6273802321031738e-06, + "loss": 0.6428196430206299, + "step": 1240 + }, + { + "epoch": 2.6257928118393234, + "grad_norm": 2.319997787475586, + "learning_rate": 3.6130845867448384e-06, + "loss": 0.7555776834487915, + "step": 1242 + }, + { + "epoch": 2.6300211416490487, + "grad_norm": 1.153585433959961, + "learning_rate": 3.598812009685514e-06, + "loss": 0.8799882531166077, + "step": 1244 + }, + { + "epoch": 2.634249471458774, + "grad_norm": 2.791447162628174, + "learning_rate": 3.584562675413189e-06, + "loss": 0.6038925051689148, + "step": 1246 + }, + { + "epoch": 2.638477801268499, + "grad_norm": 2.799598217010498, + "learning_rate": 3.5703367581316985e-06, + "loss": 0.4906109869480133, + "step": 1248 + }, + { + "epoch": 2.6427061310782243, + "grad_norm": 1.9363486766815186, + "learning_rate": 3.5561344317585957e-06, + "loss": 0.88217693567276, + "step": 1250 + }, + { + "epoch": 2.646934460887949, + "grad_norm": 2.91445255279541, + "learning_rate": 3.5419558699230305e-06, + "loss": 0.49428752064704895, + "step": 1252 + }, + { + "epoch": 2.6511627906976747, + "grad_norm": 3.5309841632843018, + "learning_rate": 3.52780124596361e-06, + "loss": 0.592266857624054, + "step": 1254 + }, + { + "epoch": 2.6553911205073994, + "grad_norm": 0.7691072821617126, + "learning_rate": 3.5136707329263055e-06, + "loss": 0.899162232875824, + "step": 1256 + }, + { + "epoch": 2.6596194503171247, + "grad_norm": 1.1038845777511597, + "learning_rate": 3.499564503562315e-06, + "loss": 0.23337708413600922, + "step": 1258 + }, + { + "epoch": 2.66384778012685, + "grad_norm": 1.3218424320220947, + "learning_rate": 3.4854827303259575e-06, + "loss": 0.4749159812927246, + "step": 1260 + }, + { + "epoch": 2.668076109936575, + "grad_norm": 4.631411075592041, + "learning_rate": 3.47142558537257e-06, + "loss": 0.5858217477798462, + "step": 1262 + }, + { + "epoch": 2.6723044397463003, + "grad_norm": 4.841278076171875, + "learning_rate": 3.4573932405563966e-06, + "loss": 0.5743645429611206, + "step": 1264 + }, + { + "epoch": 2.6765327695560255, + "grad_norm": 1.7203001976013184, + "learning_rate": 3.4433858674284894e-06, + "loss": 0.30530625581741333, + "step": 1266 + }, + { + "epoch": 2.6807610993657507, + "grad_norm": 1.229124903678894, + "learning_rate": 3.429403637234614e-06, + "loss": 0.6572251319885254, + "step": 1268 + }, + { + "epoch": 2.6849894291754755, + "grad_norm": 1.1437455415725708, + "learning_rate": 3.4154467209131515e-06, + "loss": 0.8553752899169922, + "step": 1270 + }, + { + "epoch": 2.689217758985201, + "grad_norm": 2.189396381378174, + "learning_rate": 3.401515289093011e-06, + "loss": 0.6791430711746216, + "step": 1272 + }, + { + "epoch": 2.693446088794926, + "grad_norm": 1.061129093170166, + "learning_rate": 3.3876095120915433e-06, + "loss": 0.3111809194087982, + "step": 1274 + }, + { + "epoch": 2.697674418604651, + "grad_norm": 1.8097518682479858, + "learning_rate": 3.3737295599124592e-06, + "loss": 0.4056800901889801, + "step": 1276 + }, + { + "epoch": 2.7019027484143763, + "grad_norm": 2.302802801132202, + "learning_rate": 3.3598756022437507e-06, + "loss": 0.8088371753692627, + "step": 1278 + }, + { + "epoch": 2.7061310782241015, + "grad_norm": 1.361583948135376, + "learning_rate": 3.3460478084556157e-06, + "loss": 0.578730583190918, + "step": 1280 + }, + { + "epoch": 2.7103594080338267, + "grad_norm": 0.9261577129364014, + "learning_rate": 3.33224634759839e-06, + "loss": 1.0114036798477173, + "step": 1282 + }, + { + "epoch": 2.714587737843552, + "grad_norm": 0.8526695966720581, + "learning_rate": 3.318471388400476e-06, + "loss": 0.6606082320213318, + "step": 1284 + }, + { + "epoch": 2.718816067653277, + "grad_norm": 0.25753316283226013, + "learning_rate": 3.3047230992662834e-06, + "loss": 0.9243538975715637, + "step": 1286 + }, + { + "epoch": 2.723044397463002, + "grad_norm": 0.9987777471542358, + "learning_rate": 3.291001648274175e-06, + "loss": 0.448833167552948, + "step": 1288 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 1.4990894794464111, + "learning_rate": 3.277307203174395e-06, + "loss": 0.8386439085006714, + "step": 1290 + }, + { + "epoch": 2.7315010570824523, + "grad_norm": 0.9797271490097046, + "learning_rate": 3.263639931387039e-06, + "loss": 0.9070420861244202, + "step": 1292 + }, + { + "epoch": 2.7357293868921775, + "grad_norm": 1.4759173393249512, + "learning_rate": 3.2500000000000015e-06, + "loss": 0.8921560645103455, + "step": 1294 + }, + { + "epoch": 2.7399577167019027, + "grad_norm": 0.5459781885147095, + "learning_rate": 3.236387575766916e-06, + "loss": 0.28667134046554565, + "step": 1296 + }, + { + "epoch": 2.744186046511628, + "grad_norm": 1.2515898942947388, + "learning_rate": 3.222802825105138e-06, + "loss": 0.5158063173294067, + "step": 1298 + }, + { + "epoch": 2.748414376321353, + "grad_norm": 1.4339126348495483, + "learning_rate": 3.209245914093708e-06, + "loss": 0.6157616972923279, + "step": 1300 + }, + { + "epoch": 2.7526427061310783, + "grad_norm": 1.2328945398330688, + "learning_rate": 3.1957170084713053e-06, + "loss": 0.7125469446182251, + "step": 1302 + }, + { + "epoch": 2.7568710359408035, + "grad_norm": 0.6484391093254089, + "learning_rate": 3.182216273634235e-06, + "loss": 0.6020433902740479, + "step": 1304 + }, + { + "epoch": 2.7610993657505283, + "grad_norm": 1.7392611503601074, + "learning_rate": 3.1687438746344086e-06, + "loss": 0.8029919266700745, + "step": 1306 + }, + { + "epoch": 2.765327695560254, + "grad_norm": 1.6291215419769287, + "learning_rate": 3.1552999761773166e-06, + "loss": 0.7188165187835693, + "step": 1308 + }, + { + "epoch": 2.7695560253699787, + "grad_norm": 0.8530371785163879, + "learning_rate": 3.141884742620013e-06, + "loss": 0.8661715388298035, + "step": 1310 + }, + { + "epoch": 2.773784355179704, + "grad_norm": 1.1567661762237549, + "learning_rate": 3.1284983379691247e-06, + "loss": 0.7825019359588623, + "step": 1312 + }, + { + "epoch": 2.778012684989429, + "grad_norm": 0.6033177971839905, + "learning_rate": 3.1151409258788235e-06, + "loss": 0.5006937980651855, + "step": 1314 + }, + { + "epoch": 2.7822410147991543, + "grad_norm": 1.095625877380371, + "learning_rate": 3.101812669648838e-06, + "loss": 0.8159549832344055, + "step": 1316 + }, + { + "epoch": 2.7864693446088795, + "grad_norm": 0.5257118940353394, + "learning_rate": 3.0885137322224546e-06, + "loss": 0.49076685309410095, + "step": 1318 + }, + { + "epoch": 2.7906976744186047, + "grad_norm": 1.881252408027649, + "learning_rate": 3.075244276184529e-06, + "loss": 0.8093648552894592, + "step": 1320 + }, + { + "epoch": 2.79492600422833, + "grad_norm": 3.261385440826416, + "learning_rate": 3.0620044637594936e-06, + "loss": 0.299016535282135, + "step": 1322 + }, + { + "epoch": 2.7991543340380547, + "grad_norm": 0.9506343603134155, + "learning_rate": 3.0487944568093685e-06, + "loss": 0.9361976981163025, + "step": 1324 + }, + { + "epoch": 2.8033826638477803, + "grad_norm": 3.1485157012939453, + "learning_rate": 3.035614416831801e-06, + "loss": 0.22837141156196594, + "step": 1326 + }, + { + "epoch": 2.807610993657505, + "grad_norm": 2.317355155944824, + "learning_rate": 3.0224645049580746e-06, + "loss": 0.37786638736724854, + "step": 1328 + }, + { + "epoch": 2.8118393234672303, + "grad_norm": 0.8846312761306763, + "learning_rate": 3.0093448819511416e-06, + "loss": 0.5180804133415222, + "step": 1330 + }, + { + "epoch": 2.8160676532769555, + "grad_norm": 1.2129065990447998, + "learning_rate": 2.9962557082036677e-06, + "loss": 0.9457061290740967, + "step": 1332 + }, + { + "epoch": 2.8202959830866807, + "grad_norm": 1.0315744876861572, + "learning_rate": 2.9831971437360606e-06, + "loss": 0.6278344988822937, + "step": 1334 + }, + { + "epoch": 2.824524312896406, + "grad_norm": 0.9843349456787109, + "learning_rate": 2.9701693481945132e-06, + "loss": 0.9171178936958313, + "step": 1336 + }, + { + "epoch": 2.828752642706131, + "grad_norm": 1.1731762886047363, + "learning_rate": 2.957172480849064e-06, + "loss": 0.9074237942695618, + "step": 1338 + }, + { + "epoch": 2.8329809725158563, + "grad_norm": 0.9682512283325195, + "learning_rate": 2.9442067005916375e-06, + "loss": 0.7521849274635315, + "step": 1340 + }, + { + "epoch": 2.8372093023255816, + "grad_norm": 0.7146301865577698, + "learning_rate": 2.931272165934105e-06, + "loss": 0.8462032079696655, + "step": 1342 + }, + { + "epoch": 2.8414376321353068, + "grad_norm": 1.4611376523971558, + "learning_rate": 2.918369035006351e-06, + "loss": 0.6046394109725952, + "step": 1344 + }, + { + "epoch": 2.8456659619450315, + "grad_norm": 1.9952713251113892, + "learning_rate": 2.9054974655543335e-06, + "loss": 0.6248216032981873, + "step": 1346 + }, + { + "epoch": 2.8498942917547567, + "grad_norm": 1.0322167873382568, + "learning_rate": 2.89265761493816e-06, + "loss": 0.7867326736450195, + "step": 1348 + }, + { + "epoch": 2.854122621564482, + "grad_norm": 4.295095443725586, + "learning_rate": 2.8798496401301644e-06, + "loss": 0.5031394362449646, + "step": 1350 + }, + { + "epoch": 2.858350951374207, + "grad_norm": 1.4924100637435913, + "learning_rate": 2.8670736977129843e-06, + "loss": 0.8665858507156372, + "step": 1352 + }, + { + "epoch": 2.8625792811839323, + "grad_norm": 0.7207593321800232, + "learning_rate": 2.854329943877647e-06, + "loss": 0.8522268533706665, + "step": 1354 + }, + { + "epoch": 2.8668076109936576, + "grad_norm": 1.5618963241577148, + "learning_rate": 2.841618534421664e-06, + "loss": 0.3809153437614441, + "step": 1356 + }, + { + "epoch": 2.8710359408033828, + "grad_norm": 1.1964012384414673, + "learning_rate": 2.8289396247471227e-06, + "loss": 0.9071000218391418, + "step": 1358 + }, + { + "epoch": 2.875264270613108, + "grad_norm": 2.203662872314453, + "learning_rate": 2.816293369858788e-06, + "loss": 0.6387331485748291, + "step": 1360 + }, + { + "epoch": 2.879492600422833, + "grad_norm": 3.9178061485290527, + "learning_rate": 2.803679924362206e-06, + "loss": 0.6241015791893005, + "step": 1362 + }, + { + "epoch": 2.883720930232558, + "grad_norm": 1.9866626262664795, + "learning_rate": 2.7910994424618165e-06, + "loss": 0.9976719617843628, + "step": 1364 + }, + { + "epoch": 2.887949260042283, + "grad_norm": 2.222538709640503, + "learning_rate": 2.778552077959064e-06, + "loss": 0.8397321105003357, + "step": 1366 + }, + { + "epoch": 2.8921775898520083, + "grad_norm": 1.3759814500808716, + "learning_rate": 2.766037984250521e-06, + "loss": 0.8696737289428711, + "step": 1368 + }, + { + "epoch": 2.8964059196617336, + "grad_norm": 0.9230353236198425, + "learning_rate": 2.753557314326011e-06, + "loss": 0.8795881867408752, + "step": 1370 + }, + { + "epoch": 2.9006342494714588, + "grad_norm": 0.7405307292938232, + "learning_rate": 2.7411102207667373e-06, + "loss": 0.9262542128562927, + "step": 1372 + }, + { + "epoch": 2.904862579281184, + "grad_norm": 1.2151702642440796, + "learning_rate": 2.728696855743418e-06, + "loss": 0.7794469594955444, + "step": 1374 + }, + { + "epoch": 2.909090909090909, + "grad_norm": 0.9104248881340027, + "learning_rate": 2.7163173710144323e-06, + "loss": 0.8876684308052063, + "step": 1376 + }, + { + "epoch": 2.9133192389006344, + "grad_norm": 2.51352596282959, + "learning_rate": 2.703971917923951e-06, + "loss": 0.6242119073867798, + "step": 1378 + }, + { + "epoch": 2.9175475687103596, + "grad_norm": 0.8816441297531128, + "learning_rate": 2.6916606474000965e-06, + "loss": 0.6980829834938049, + "step": 1380 + }, + { + "epoch": 2.9217758985200843, + "grad_norm": 1.3975627422332764, + "learning_rate": 2.679383709953103e-06, + "loss": 0.8792454600334167, + "step": 1382 + }, + { + "epoch": 2.92600422832981, + "grad_norm": 2.8038578033447266, + "learning_rate": 2.6671412556734578e-06, + "loss": 0.7361038327217102, + "step": 1384 + }, + { + "epoch": 2.9302325581395348, + "grad_norm": 0.8007720708847046, + "learning_rate": 2.6549334342300844e-06, + "loss": 0.5931589007377625, + "step": 1386 + }, + { + "epoch": 2.93446088794926, + "grad_norm": 1.0942860841751099, + "learning_rate": 2.642760394868503e-06, + "loss": 0.8650692105293274, + "step": 1388 + }, + { + "epoch": 2.938689217758985, + "grad_norm": 0.31571274995803833, + "learning_rate": 2.6306222864090157e-06, + "loss": 0.6512575149536133, + "step": 1390 + }, + { + "epoch": 2.9429175475687104, + "grad_norm": 1.7085269689559937, + "learning_rate": 2.61851925724487e-06, + "loss": 0.8907392621040344, + "step": 1392 + }, + { + "epoch": 2.9471458773784356, + "grad_norm": 0.400570809841156, + "learning_rate": 2.606451455340459e-06, + "loss": 0.23060639202594757, + "step": 1394 + }, + { + "epoch": 2.951374207188161, + "grad_norm": 1.0295450687408447, + "learning_rate": 2.5944190282295155e-06, + "loss": 0.47475987672805786, + "step": 1396 + }, + { + "epoch": 2.955602536997886, + "grad_norm": 1.5369654893875122, + "learning_rate": 2.5824221230132895e-06, + "loss": 0.8710322380065918, + "step": 1398 + }, + { + "epoch": 2.9598308668076108, + "grad_norm": 0.8360358476638794, + "learning_rate": 2.5704608863587676e-06, + "loss": 1.2614446878433228, + "step": 1400 + }, + { + "epoch": 2.9640591966173364, + "grad_norm": 1.9678583145141602, + "learning_rate": 2.558535464496877e-06, + "loss": 0.6945706605911255, + "step": 1402 + }, + { + "epoch": 2.968287526427061, + "grad_norm": 1.1185355186462402, + "learning_rate": 2.5466460032206872e-06, + "loss": 0.41613972187042236, + "step": 1404 + }, + { + "epoch": 2.9725158562367864, + "grad_norm": 2.029306173324585, + "learning_rate": 2.53479264788364e-06, + "loss": 0.8905559778213501, + "step": 1406 + }, + { + "epoch": 2.9767441860465116, + "grad_norm": 0.7787075638771057, + "learning_rate": 2.522975543397769e-06, + "loss": 0.6910480856895447, + "step": 1408 + }, + { + "epoch": 2.980972515856237, + "grad_norm": 2.2516300678253174, + "learning_rate": 2.511194834231922e-06, + "loss": 0.621972382068634, + "step": 1410 + }, + { + "epoch": 2.985200845665962, + "grad_norm": 0.9599463939666748, + "learning_rate": 2.49945066441e-06, + "loss": 0.8999648094177246, + "step": 1412 + }, + { + "epoch": 2.989429175475687, + "grad_norm": 0.8515106439590454, + "learning_rate": 2.4877431775092004e-06, + "loss": 0.6651572585105896, + "step": 1414 + }, + { + "epoch": 2.9936575052854124, + "grad_norm": 2.2343332767486572, + "learning_rate": 2.4760725166582544e-06, + "loss": 0.7343695759773254, + "step": 1416 + }, + { + "epoch": 2.997885835095137, + "grad_norm": 0.7534525990486145, + "learning_rate": 2.464438824535675e-06, + "loss": 0.45403918623924255, + "step": 1418 + }, + { + "epoch": 3.0021141649048624, + "grad_norm": 0.6817217469215393, + "learning_rate": 2.452842243368028e-06, + "loss": 0.8473333716392517, + "step": 1420 + }, + { + "epoch": 3.0063424947145876, + "grad_norm": 1.2478125095367432, + "learning_rate": 2.4412829149281735e-06, + "loss": 0.501265823841095, + "step": 1422 + }, + { + "epoch": 3.010570824524313, + "grad_norm": 1.965147614479065, + "learning_rate": 2.4297609805335472e-06, + "loss": 0.8796722292900085, + "step": 1424 + }, + { + "epoch": 3.014799154334038, + "grad_norm": 5.721529006958008, + "learning_rate": 2.4182765810444263e-06, + "loss": 0.43021053075790405, + "step": 1426 + }, + { + "epoch": 3.019027484143763, + "grad_norm": 4.68020486831665, + "learning_rate": 2.4068298568622083e-06, + "loss": 0.5905739665031433, + "step": 1428 + }, + { + "epoch": 3.0232558139534884, + "grad_norm": 1.287234902381897, + "learning_rate": 2.395420947927694e-06, + "loss": 0.6827127933502197, + "step": 1430 + }, + { + "epoch": 3.0274841437632136, + "grad_norm": 1.3199462890625, + "learning_rate": 2.3840499937193804e-06, + "loss": 0.8912433981895447, + "step": 1432 + }, + { + "epoch": 3.031712473572939, + "grad_norm": 1.7283967733383179, + "learning_rate": 2.3727171332517496e-06, + "loss": 0.6121829152107239, + "step": 1434 + }, + { + "epoch": 3.035940803382664, + "grad_norm": 3.5385262966156006, + "learning_rate": 2.361422505073575e-06, + "loss": 0.5448174476623535, + "step": 1436 + }, + { + "epoch": 3.040169133192389, + "grad_norm": 0.5468651652336121, + "learning_rate": 2.3501662472662233e-06, + "loss": 0.12289305776357651, + "step": 1438 + }, + { + "epoch": 3.044397463002114, + "grad_norm": 1.0555592775344849, + "learning_rate": 2.3389484974419685e-06, + "loss": 0.7503781914710999, + "step": 1440 + }, + { + "epoch": 3.048625792811839, + "grad_norm": 0.9246349334716797, + "learning_rate": 2.327769392742309e-06, + "loss": 0.49217045307159424, + "step": 1442 + }, + { + "epoch": 3.0528541226215644, + "grad_norm": 2.1136772632598877, + "learning_rate": 2.316629069836291e-06, + "loss": 0.12871970236301422, + "step": 1444 + }, + { + "epoch": 3.0570824524312896, + "grad_norm": 1.807900309562683, + "learning_rate": 2.305527664918838e-06, + "loss": 0.4163699746131897, + "step": 1446 + }, + { + "epoch": 3.061310782241015, + "grad_norm": 2.190528631210327, + "learning_rate": 2.294465313709085e-06, + "loss": 0.510996401309967, + "step": 1448 + }, + { + "epoch": 3.06553911205074, + "grad_norm": 0.8874261379241943, + "learning_rate": 2.2834421514487205e-06, + "loss": 0.4852120578289032, + "step": 1450 + }, + { + "epoch": 3.0697674418604652, + "grad_norm": 2.309600353240967, + "learning_rate": 2.27245831290033e-06, + "loss": 0.5598742961883545, + "step": 1452 + }, + { + "epoch": 3.0739957716701904, + "grad_norm": 2.3264451026916504, + "learning_rate": 2.261513932345755e-06, + "loss": 0.7019772529602051, + "step": 1454 + }, + { + "epoch": 3.0782241014799157, + "grad_norm": 0.3824637234210968, + "learning_rate": 2.250609143584443e-06, + "loss": 0.15948505699634552, + "step": 1456 + }, + { + "epoch": 3.0824524312896404, + "grad_norm": 1.0107927322387695, + "learning_rate": 2.239744079931818e-06, + "loss": 0.5560399889945984, + "step": 1458 + }, + { + "epoch": 3.0866807610993656, + "grad_norm": 1.0550543069839478, + "learning_rate": 2.228918874217649e-06, + "loss": 0.6576963663101196, + "step": 1460 + }, + { + "epoch": 3.090909090909091, + "grad_norm": 1.275876760482788, + "learning_rate": 2.2181336587844244e-06, + "loss": 0.7855117321014404, + "step": 1462 + }, + { + "epoch": 3.095137420718816, + "grad_norm": 0.7812337875366211, + "learning_rate": 2.207388565485736e-06, + "loss": 0.48513537645339966, + "step": 1464 + }, + { + "epoch": 3.0993657505285412, + "grad_norm": 1.1482244729995728, + "learning_rate": 2.196683725684666e-06, + "loss": 0.587722897529602, + "step": 1466 + }, + { + "epoch": 3.1035940803382664, + "grad_norm": 1.0105695724487305, + "learning_rate": 2.1860192702521826e-06, + "loss": 0.7577872276306152, + "step": 1468 + }, + { + "epoch": 3.1078224101479917, + "grad_norm": 0.9446401000022888, + "learning_rate": 2.1753953295655354e-06, + "loss": 0.7079393863677979, + "step": 1470 + }, + { + "epoch": 3.112050739957717, + "grad_norm": 1.5413912534713745, + "learning_rate": 2.164812033506673e-06, + "loss": 0.43389561772346497, + "step": 1472 + }, + { + "epoch": 3.116279069767442, + "grad_norm": 0.9413207173347473, + "learning_rate": 2.1542695114606367e-06, + "loss": 0.8424919843673706, + "step": 1474 + }, + { + "epoch": 3.120507399577167, + "grad_norm": 1.3415789604187012, + "learning_rate": 2.1437678923139944e-06, + "loss": 0.6936777830123901, + "step": 1476 + }, + { + "epoch": 3.124735729386892, + "grad_norm": 1.8507732152938843, + "learning_rate": 2.1333073044532628e-06, + "loss": 0.2213955521583557, + "step": 1478 + }, + { + "epoch": 3.1289640591966172, + "grad_norm": 0.769049882888794, + "learning_rate": 2.122887875763327e-06, + "loss": 0.7384406924247742, + "step": 1480 + }, + { + "epoch": 3.1331923890063424, + "grad_norm": 1.0787291526794434, + "learning_rate": 2.1125097336258897e-06, + "loss": 0.7775143980979919, + "step": 1482 + }, + { + "epoch": 3.1374207188160677, + "grad_norm": 0.8996830582618713, + "learning_rate": 2.1021730049179117e-06, + "loss": 0.20124602317810059, + "step": 1484 + }, + { + "epoch": 3.141649048625793, + "grad_norm": 1.2917189598083496, + "learning_rate": 2.09187781601005e-06, + "loss": 0.7534149289131165, + "step": 1486 + }, + { + "epoch": 3.145877378435518, + "grad_norm": 0.7089277505874634, + "learning_rate": 2.0816242927651254e-06, + "loss": 0.8691098093986511, + "step": 1488 + }, + { + "epoch": 3.1501057082452433, + "grad_norm": 2.7333178520202637, + "learning_rate": 2.071412560536581e-06, + "loss": 0.6103125214576721, + "step": 1490 + }, + { + "epoch": 3.1543340380549685, + "grad_norm": 2.3667232990264893, + "learning_rate": 2.0612427441669475e-06, + "loss": 0.7145513892173767, + "step": 1492 + }, + { + "epoch": 3.1585623678646932, + "grad_norm": 0.8767846822738647, + "learning_rate": 2.051114967986311e-06, + "loss": 0.34956127405166626, + "step": 1494 + }, + { + "epoch": 3.1627906976744184, + "grad_norm": 0.2730441689491272, + "learning_rate": 2.0410293558108074e-06, + "loss": 0.5361222624778748, + "step": 1496 + }, + { + "epoch": 3.1670190274841437, + "grad_norm": 1.6024144887924194, + "learning_rate": 2.0309860309410986e-06, + "loss": 0.7061223983764648, + "step": 1498 + }, + { + "epoch": 3.171247357293869, + "grad_norm": 1.080323338508606, + "learning_rate": 2.0209851161608613e-06, + "loss": 0.8353047370910645, + "step": 1500 + }, + { + "epoch": 3.175475687103594, + "grad_norm": 6.807703495025635, + "learning_rate": 2.011026733735301e-06, + "loss": 0.2627289891242981, + "step": 1502 + }, + { + "epoch": 3.1797040169133193, + "grad_norm": 3.5282793045043945, + "learning_rate": 2.001111005409643e-06, + "loss": 0.4190889298915863, + "step": 1504 + }, + { + "epoch": 3.1839323467230445, + "grad_norm": 1.270559310913086, + "learning_rate": 1.9912380524076503e-06, + "loss": 0.5806607007980347, + "step": 1506 + }, + { + "epoch": 3.1881606765327697, + "grad_norm": 3.3433632850646973, + "learning_rate": 1.981407995430136e-06, + "loss": 0.3371882140636444, + "step": 1508 + }, + { + "epoch": 3.192389006342495, + "grad_norm": 3.1091907024383545, + "learning_rate": 1.9716209546535e-06, + "loss": 0.19397024810314178, + "step": 1510 + }, + { + "epoch": 3.1966173361522197, + "grad_norm": 0.2143598198890686, + "learning_rate": 1.9618770497282487e-06, + "loss": 0.051398925483226776, + "step": 1512 + }, + { + "epoch": 3.200845665961945, + "grad_norm": 1.5491254329681396, + "learning_rate": 1.9521763997775294e-06, + "loss": 0.6001431941986084, + "step": 1514 + }, + { + "epoch": 3.20507399577167, + "grad_norm": 1.3003274202346802, + "learning_rate": 1.942519123395689e-06, + "loss": 0.7325740456581116, + "step": 1516 + }, + { + "epoch": 3.2093023255813953, + "grad_norm": 1.1942628622055054, + "learning_rate": 1.932905338646812e-06, + "loss": 0.4662987291812897, + "step": 1518 + }, + { + "epoch": 3.2135306553911205, + "grad_norm": 0.8495181798934937, + "learning_rate": 1.9233351630632747e-06, + "loss": 0.6702975630760193, + "step": 1520 + }, + { + "epoch": 3.2177589852008457, + "grad_norm": 1.3906527757644653, + "learning_rate": 1.9138087136443227e-06, + "loss": 0.7408310770988464, + "step": 1522 + }, + { + "epoch": 3.221987315010571, + "grad_norm": 2.631319999694824, + "learning_rate": 1.904326106854626e-06, + "loss": 0.7249854207038879, + "step": 1524 + }, + { + "epoch": 3.226215644820296, + "grad_norm": 0.9974054098129272, + "learning_rate": 1.8948874586228622e-06, + "loss": 0.737725555896759, + "step": 1526 + }, + { + "epoch": 3.2304439746300213, + "grad_norm": 5.671351909637451, + "learning_rate": 1.8854928843402987e-06, + "loss": 0.09608987718820572, + "step": 1528 + }, + { + "epoch": 3.234672304439746, + "grad_norm": 1.8938038349151611, + "learning_rate": 1.8761424988593796e-06, + "loss": 0.7253928780555725, + "step": 1530 + }, + { + "epoch": 3.2389006342494713, + "grad_norm": 5.95370626449585, + "learning_rate": 1.8668364164923248e-06, + "loss": 0.6535470485687256, + "step": 1532 + }, + { + "epoch": 3.2431289640591965, + "grad_norm": 1.9531099796295166, + "learning_rate": 1.8575747510097294e-06, + "loss": 0.7161535024642944, + "step": 1534 + }, + { + "epoch": 3.2473572938689217, + "grad_norm": 1.2146064043045044, + "learning_rate": 1.8483576156391747e-06, + "loss": 0.5191117525100708, + "step": 1536 + }, + { + "epoch": 3.251585623678647, + "grad_norm": 1.3939259052276611, + "learning_rate": 1.8391851230638443e-06, + "loss": 0.5700780749320984, + "step": 1538 + }, + { + "epoch": 3.255813953488372, + "grad_norm": 1.2802196741104126, + "learning_rate": 1.8300573854211452e-06, + "loss": 0.45202237367630005, + "step": 1540 + }, + { + "epoch": 3.2600422832980973, + "grad_norm": 2.324915885925293, + "learning_rate": 1.8209745143013363e-06, + "loss": 0.9876203536987305, + "step": 1542 + }, + { + "epoch": 3.2642706131078225, + "grad_norm": 0.9092972874641418, + "learning_rate": 1.811936620746167e-06, + "loss": 0.7452791333198547, + "step": 1544 + }, + { + "epoch": 3.2684989429175477, + "grad_norm": 2.8554110527038574, + "learning_rate": 1.8029438152475159e-06, + "loss": 0.6718047261238098, + "step": 1546 + }, + { + "epoch": 3.2727272727272725, + "grad_norm": 1.1308796405792236, + "learning_rate": 1.7939962077460448e-06, + "loss": 0.1404690146446228, + "step": 1548 + }, + { + "epoch": 3.276955602536998, + "grad_norm": 1.478325605392456, + "learning_rate": 1.785093907629849e-06, + "loss": 0.6233620047569275, + "step": 1550 + }, + { + "epoch": 3.281183932346723, + "grad_norm": 2.3967654705047607, + "learning_rate": 1.7762370237331233e-06, + "loss": 0.7281405329704285, + "step": 1552 + }, + { + "epoch": 3.285412262156448, + "grad_norm": 0.08398935198783875, + "learning_rate": 1.7674256643348348e-06, + "loss": 0.0760122686624527, + "step": 1554 + }, + { + "epoch": 3.2896405919661733, + "grad_norm": 1.0803419351577759, + "learning_rate": 1.7586599371573887e-06, + "loss": 0.6622000336647034, + "step": 1556 + }, + { + "epoch": 3.2938689217758985, + "grad_norm": 1.0871472358703613, + "learning_rate": 1.7499399493653225e-06, + "loss": 0.44463807344436646, + "step": 1558 + }, + { + "epoch": 3.2980972515856237, + "grad_norm": 2.569706439971924, + "learning_rate": 1.7412658075639921e-06, + "loss": 0.8966039419174194, + "step": 1560 + }, + { + "epoch": 3.302325581395349, + "grad_norm": 5.8961639404296875, + "learning_rate": 1.7326376177982635e-06, + "loss": 0.5105495452880859, + "step": 1562 + }, + { + "epoch": 3.306553911205074, + "grad_norm": 2.01412034034729, + "learning_rate": 1.7240554855512225e-06, + "loss": 0.3557579219341278, + "step": 1564 + }, + { + "epoch": 3.3107822410147993, + "grad_norm": 1.2747447490692139, + "learning_rate": 1.7155195157428878e-06, + "loss": 0.5155658721923828, + "step": 1566 + }, + { + "epoch": 3.3150105708245245, + "grad_norm": 0.31796056032180786, + "learning_rate": 1.707029812728916e-06, + "loss": 0.0953366681933403, + "step": 1568 + }, + { + "epoch": 3.3192389006342493, + "grad_norm": 1.8880454301834106, + "learning_rate": 1.6985864802993366e-06, + "loss": 0.4693029224872589, + "step": 1570 + }, + { + "epoch": 3.3234672304439745, + "grad_norm": 2.811009645462036, + "learning_rate": 1.690189621677284e-06, + "loss": 0.525675356388092, + "step": 1572 + }, + { + "epoch": 3.3276955602536997, + "grad_norm": 3.7383172512054443, + "learning_rate": 1.681839339517728e-06, + "loss": 0.16659285128116608, + "step": 1574 + }, + { + "epoch": 3.331923890063425, + "grad_norm": 2.164754629135132, + "learning_rate": 1.67353573590622e-06, + "loss": 0.2484208196401596, + "step": 1576 + }, + { + "epoch": 3.33615221987315, + "grad_norm": 2.468291759490967, + "learning_rate": 1.6652789123576513e-06, + "loss": 0.5150903463363647, + "step": 1578 + }, + { + "epoch": 3.3403805496828753, + "grad_norm": 1.3167822360992432, + "learning_rate": 1.6570689698150094e-06, + "loss": 0.4254363477230072, + "step": 1580 + }, + { + "epoch": 3.3446088794926006, + "grad_norm": 0.9184785485267639, + "learning_rate": 1.6489060086481373e-06, + "loss": 0.7732759118080139, + "step": 1582 + }, + { + "epoch": 3.3488372093023258, + "grad_norm": 2.406330108642578, + "learning_rate": 1.6407901286525152e-06, + "loss": 0.5360360741615295, + "step": 1584 + }, + { + "epoch": 3.353065539112051, + "grad_norm": 1.332894206047058, + "learning_rate": 1.6327214290480382e-06, + "loss": 0.5299188494682312, + "step": 1586 + }, + { + "epoch": 3.3572938689217757, + "grad_norm": 3.176917314529419, + "learning_rate": 1.6247000084778006e-06, + "loss": 0.9168533086776733, + "step": 1588 + }, + { + "epoch": 3.361522198731501, + "grad_norm": 1.164693832397461, + "learning_rate": 1.6167259650068887e-06, + "loss": 0.22458931803703308, + "step": 1590 + }, + { + "epoch": 3.365750528541226, + "grad_norm": 0.9778388142585754, + "learning_rate": 1.6087993961211904e-06, + "loss": 0.6299681663513184, + "step": 1592 + }, + { + "epoch": 3.3699788583509513, + "grad_norm": 4.3390703201293945, + "learning_rate": 1.6009203987261957e-06, + "loss": 0.27036425471305847, + "step": 1594 + }, + { + "epoch": 3.3742071881606766, + "grad_norm": 0.9561054706573486, + "learning_rate": 1.593089069145809e-06, + "loss": 0.7483205795288086, + "step": 1596 + }, + { + "epoch": 3.3784355179704018, + "grad_norm": 2.2535130977630615, + "learning_rate": 1.5853055031211844e-06, + "loss": 0.4829716980457306, + "step": 1598 + }, + { + "epoch": 3.382663847780127, + "grad_norm": 0.9475092887878418, + "learning_rate": 1.577569795809544e-06, + "loss": 0.20053090155124664, + "step": 1600 + }, + { + "epoch": 3.386892177589852, + "grad_norm": 1.653645634651184, + "learning_rate": 1.5698820417830137e-06, + "loss": 0.49854278564453125, + "step": 1602 + }, + { + "epoch": 3.3911205073995774, + "grad_norm": 1.2782807350158691, + "learning_rate": 1.5622423350274784e-06, + "loss": 0.7943662405014038, + "step": 1604 + }, + { + "epoch": 3.395348837209302, + "grad_norm": 0.913489580154419, + "learning_rate": 1.5546507689414217e-06, + "loss": 0.7943485379219055, + "step": 1606 + }, + { + "epoch": 3.3995771670190273, + "grad_norm": 7.427999496459961, + "learning_rate": 1.5471074363347894e-06, + "loss": 0.42399805784225464, + "step": 1608 + }, + { + "epoch": 3.4038054968287526, + "grad_norm": 0.8594838380813599, + "learning_rate": 1.539612429427852e-06, + "loss": 0.7737821340560913, + "step": 1610 + }, + { + "epoch": 3.4080338266384778, + "grad_norm": 1.0548810958862305, + "learning_rate": 1.5321658398500808e-06, + "loss": 0.6136525273323059, + "step": 1612 + }, + { + "epoch": 3.412262156448203, + "grad_norm": 1.8661061525344849, + "learning_rate": 1.5247677586390255e-06, + "loss": 0.6268908381462097, + "step": 1614 + }, + { + "epoch": 3.416490486257928, + "grad_norm": 0.8926048278808594, + "learning_rate": 1.5174182762392007e-06, + "loss": 0.7578631043434143, + "step": 1616 + }, + { + "epoch": 3.4207188160676534, + "grad_norm": 0.8501190543174744, + "learning_rate": 1.5101174825009815e-06, + "loss": 0.7076795101165771, + "step": 1618 + }, + { + "epoch": 3.4249471458773786, + "grad_norm": 1.1074150800704956, + "learning_rate": 1.5028654666795042e-06, + "loss": 0.7425435781478882, + "step": 1620 + }, + { + "epoch": 3.429175475687104, + "grad_norm": 1.8880587816238403, + "learning_rate": 1.495662317433576e-06, + "loss": 0.5129321217536926, + "step": 1622 + }, + { + "epoch": 3.4334038054968286, + "grad_norm": 0.9060251712799072, + "learning_rate": 1.4885081228245909e-06, + "loss": 0.7459352612495422, + "step": 1624 + }, + { + "epoch": 3.4376321353065538, + "grad_norm": 1.29948091506958, + "learning_rate": 1.4814029703154515e-06, + "loss": 0.49240386486053467, + "step": 1626 + }, + { + "epoch": 3.441860465116279, + "grad_norm": 0.8913044929504395, + "learning_rate": 1.4743469467695017e-06, + "loss": 0.8139543533325195, + "step": 1628 + }, + { + "epoch": 3.446088794926004, + "grad_norm": 1.787123203277588, + "learning_rate": 1.4673401384494651e-06, + "loss": 0.43955478072166443, + "step": 1630 + }, + { + "epoch": 3.4503171247357294, + "grad_norm": 0.9750705361366272, + "learning_rate": 1.4603826310163874e-06, + "loss": 0.5946288108825684, + "step": 1632 + }, + { + "epoch": 3.4545454545454546, + "grad_norm": 2.027559518814087, + "learning_rate": 1.453474509528593e-06, + "loss": 0.4718847870826721, + "step": 1634 + }, + { + "epoch": 3.45877378435518, + "grad_norm": 0.8577848076820374, + "learning_rate": 1.446615858440642e-06, + "loss": 0.6591483354568481, + "step": 1636 + }, + { + "epoch": 3.463002114164905, + "grad_norm": 0.14572234451770782, + "learning_rate": 1.4398067616022993e-06, + "loss": 0.28346192836761475, + "step": 1638 + }, + { + "epoch": 3.46723044397463, + "grad_norm": 1.3129976987838745, + "learning_rate": 1.43304730225751e-06, + "loss": 0.44907644391059875, + "step": 1640 + }, + { + "epoch": 3.471458773784355, + "grad_norm": 0.5489000678062439, + "learning_rate": 1.4263375630433802e-06, + "loss": 0.1286095380783081, + "step": 1642 + }, + { + "epoch": 3.47568710359408, + "grad_norm": 2.6421170234680176, + "learning_rate": 1.4196776259891665e-06, + "loss": 0.7149624228477478, + "step": 1644 + }, + { + "epoch": 3.4799154334038054, + "grad_norm": 1.500959873199463, + "learning_rate": 1.4130675725152762e-06, + "loss": 0.6911309957504272, + "step": 1646 + }, + { + "epoch": 3.4841437632135306, + "grad_norm": 0.13725422322750092, + "learning_rate": 1.4065074834322677e-06, + "loss": 0.42130085825920105, + "step": 1648 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 0.9783065915107727, + "learning_rate": 1.3999974389398669e-06, + "loss": 0.3836827278137207, + "step": 1650 + }, + { + "epoch": 3.492600422832981, + "grad_norm": 4.479634761810303, + "learning_rate": 1.3935375186259817e-06, + "loss": 0.4610128104686737, + "step": 1652 + }, + { + "epoch": 3.496828752642706, + "grad_norm": 2.4473862648010254, + "learning_rate": 1.3871278014657332e-06, + "loss": 0.6460088491439819, + "step": 1654 + }, + { + "epoch": 3.5010570824524314, + "grad_norm": 1.2623950242996216, + "learning_rate": 1.3807683658204915e-06, + "loss": 0.7506681680679321, + "step": 1656 + }, + { + "epoch": 3.5052854122621566, + "grad_norm": 0.2533678412437439, + "learning_rate": 1.3744592894369093e-06, + "loss": 0.5305776596069336, + "step": 1658 + }, + { + "epoch": 3.5095137420718814, + "grad_norm": 0.8349683880805969, + "learning_rate": 1.3682006494459806e-06, + "loss": 0.5295114517211914, + "step": 1660 + }, + { + "epoch": 3.513742071881607, + "grad_norm": 1.6330463886260986, + "learning_rate": 1.361992522362096e-06, + "loss": 0.4738442003726959, + "step": 1662 + }, + { + "epoch": 3.517970401691332, + "grad_norm": 1.064609169960022, + "learning_rate": 1.355834984082101e-06, + "loss": 0.8023744821548462, + "step": 1664 + }, + { + "epoch": 3.522198731501057, + "grad_norm": 2.5046560764312744, + "learning_rate": 1.3497281098843746e-06, + "loss": 0.803248941898346, + "step": 1666 + }, + { + "epoch": 3.526427061310782, + "grad_norm": 0.34333527088165283, + "learning_rate": 1.34367197442791e-06, + "loss": 0.12590767443180084, + "step": 1668 + }, + { + "epoch": 3.5306553911205074, + "grad_norm": 0.4158186614513397, + "learning_rate": 1.3376666517513931e-06, + "loss": 0.5759648084640503, + "step": 1670 + }, + { + "epoch": 3.5348837209302326, + "grad_norm": 4.645848751068115, + "learning_rate": 1.3317122152723078e-06, + "loss": 0.4734373092651367, + "step": 1672 + }, + { + "epoch": 3.539112050739958, + "grad_norm": 1.015625, + "learning_rate": 1.3258087377860323e-06, + "loss": 0.7225991487503052, + "step": 1674 + }, + { + "epoch": 3.543340380549683, + "grad_norm": 4.84229850769043, + "learning_rate": 1.3199562914649519e-06, + "loss": 0.4575788676738739, + "step": 1676 + }, + { + "epoch": 3.547568710359408, + "grad_norm": 1.0021495819091797, + "learning_rate": 1.314154947857571e-06, + "loss": 0.7831419110298157, + "step": 1678 + }, + { + "epoch": 3.5517970401691334, + "grad_norm": 1.008239984512329, + "learning_rate": 1.3084047778876487e-06, + "loss": 0.4187844395637512, + "step": 1680 + }, + { + "epoch": 3.556025369978858, + "grad_norm": 0.8473500609397888, + "learning_rate": 1.3027058518533228e-06, + "loss": 0.4580603241920471, + "step": 1682 + }, + { + "epoch": 3.5602536997885834, + "grad_norm": 3.1289286613464355, + "learning_rate": 1.2970582394262516e-06, + "loss": 0.5064759850502014, + "step": 1684 + }, + { + "epoch": 3.5644820295983086, + "grad_norm": 2.5335259437561035, + "learning_rate": 1.291462009650768e-06, + "loss": 0.14471778273582458, + "step": 1686 + }, + { + "epoch": 3.568710359408034, + "grad_norm": 1.1093417406082153, + "learning_rate": 1.2859172309430285e-06, + "loss": 0.5130281448364258, + "step": 1688 + }, + { + "epoch": 3.572938689217759, + "grad_norm": 0.20470449328422546, + "learning_rate": 1.280423971090181e-06, + "loss": 0.16852088272571564, + "step": 1690 + }, + { + "epoch": 3.5771670190274842, + "grad_norm": 2.513256549835205, + "learning_rate": 1.2749822972495333e-06, + "loss": 0.3525696098804474, + "step": 1692 + }, + { + "epoch": 3.5813953488372094, + "grad_norm": 1.6751922369003296, + "learning_rate": 1.2695922759477356e-06, + "loss": 0.4803304672241211, + "step": 1694 + }, + { + "epoch": 3.585623678646934, + "grad_norm": 2.707200288772583, + "learning_rate": 1.2642539730799622e-06, + "loss": 0.36848849058151245, + "step": 1696 + }, + { + "epoch": 3.58985200845666, + "grad_norm": 1.1222119331359863, + "learning_rate": 1.2589674539091115e-06, + "loss": 0.765342116355896, + "step": 1698 + }, + { + "epoch": 3.5940803382663846, + "grad_norm": 1.6508913040161133, + "learning_rate": 1.2537327830650042e-06, + "loss": 0.882461428642273, + "step": 1700 + }, + { + "epoch": 3.59830866807611, + "grad_norm": 3.4180219173431396, + "learning_rate": 1.248550024543594e-06, + "loss": 0.6884484887123108, + "step": 1702 + }, + { + "epoch": 3.602536997885835, + "grad_norm": 2.266618490219116, + "learning_rate": 1.2434192417061862e-06, + "loss": 0.2033189982175827, + "step": 1704 + }, + { + "epoch": 3.6067653276955602, + "grad_norm": 1.6902961730957031, + "learning_rate": 1.2383404972786625e-06, + "loss": 0.38487309217453003, + "step": 1706 + }, + { + "epoch": 3.6109936575052854, + "grad_norm": 2.4988107681274414, + "learning_rate": 1.2333138533507139e-06, + "loss": 0.6411733031272888, + "step": 1708 + }, + { + "epoch": 3.6152219873150107, + "grad_norm": 2.259953498840332, + "learning_rate": 1.2283393713750816e-06, + "loss": 0.5524290204048157, + "step": 1710 + }, + { + "epoch": 3.619450317124736, + "grad_norm": 1.1606485843658447, + "learning_rate": 1.2234171121668069e-06, + "loss": 0.4868001639842987, + "step": 1712 + }, + { + "epoch": 3.6236786469344606, + "grad_norm": 0.389941543340683, + "learning_rate": 1.2185471359024855e-06, + "loss": 0.2655688226222992, + "step": 1714 + }, + { + "epoch": 3.6279069767441863, + "grad_norm": 6.926545143127441, + "learning_rate": 1.2137295021195341e-06, + "loss": 0.6682550311088562, + "step": 1716 + }, + { + "epoch": 3.632135306553911, + "grad_norm": 0.2540845274925232, + "learning_rate": 1.2089642697154613e-06, + "loss": 0.43157482147216797, + "step": 1718 + }, + { + "epoch": 3.6363636363636362, + "grad_norm": 2.847259998321533, + "learning_rate": 1.2042514969471475e-06, + "loss": 0.44277024269104004, + "step": 1720 + }, + { + "epoch": 3.6405919661733614, + "grad_norm": 1.932730793952942, + "learning_rate": 1.1995912414301329e-06, + "loss": 0.7075636982917786, + "step": 1722 + }, + { + "epoch": 3.6448202959830867, + "grad_norm": 0.8972611427307129, + "learning_rate": 1.1949835601379144e-06, + "loss": 0.5166904926300049, + "step": 1724 + }, + { + "epoch": 3.649048625792812, + "grad_norm": 1.2340290546417236, + "learning_rate": 1.1904285094012463e-06, + "loss": 0.8118146657943726, + "step": 1726 + }, + { + "epoch": 3.653276955602537, + "grad_norm": 0.9943589568138123, + "learning_rate": 1.185926144907454e-06, + "loss": 0.8302825689315796, + "step": 1728 + }, + { + "epoch": 3.6575052854122623, + "grad_norm": 2.4265196323394775, + "learning_rate": 1.1814765216997522e-06, + "loss": 0.6259387731552124, + "step": 1730 + }, + { + "epoch": 3.6617336152219875, + "grad_norm": 1.3011783361434937, + "learning_rate": 1.1770796941765727e-06, + "loss": 0.7334924936294556, + "step": 1732 + }, + { + "epoch": 3.6659619450317127, + "grad_norm": 1.2644360065460205, + "learning_rate": 1.1727357160908978e-06, + "loss": 0.7298177480697632, + "step": 1734 + }, + { + "epoch": 3.6701902748414374, + "grad_norm": 0.9330185651779175, + "learning_rate": 1.1684446405496052e-06, + "loss": 0.7507150769233704, + "step": 1736 + }, + { + "epoch": 3.6744186046511627, + "grad_norm": 2.3698582649230957, + "learning_rate": 1.1642065200128179e-06, + "loss": 0.12819845974445343, + "step": 1738 + }, + { + "epoch": 3.678646934460888, + "grad_norm": 1.1524574756622314, + "learning_rate": 1.1600214062932613e-06, + "loss": 0.45083189010620117, + "step": 1740 + }, + { + "epoch": 3.682875264270613, + "grad_norm": 1.659613847732544, + "learning_rate": 1.1558893505556315e-06, + "loss": 0.24585355818271637, + "step": 1742 + }, + { + "epoch": 3.6871035940803383, + "grad_norm": 0.8318360447883606, + "learning_rate": 1.1518104033159715e-06, + "loss": 0.7956206798553467, + "step": 1744 + }, + { + "epoch": 3.6913319238900635, + "grad_norm": 0.2553507387638092, + "learning_rate": 1.1477846144410493e-06, + "loss": 0.33877405524253845, + "step": 1746 + }, + { + "epoch": 3.6955602536997887, + "grad_norm": 3.4053497314453125, + "learning_rate": 1.1438120331477506e-06, + "loss": 0.28586113452911377, + "step": 1748 + }, + { + "epoch": 3.699788583509514, + "grad_norm": 1.3830000162124634, + "learning_rate": 1.1398927080024796e-06, + "loss": 0.9693983793258667, + "step": 1750 + }, + { + "epoch": 3.704016913319239, + "grad_norm": 1.2605457305908203, + "learning_rate": 1.1360266869205593e-06, + "loss": 0.3419749438762665, + "step": 1752 + }, + { + "epoch": 3.708245243128964, + "grad_norm": 1.4828147888183594, + "learning_rate": 1.1322140171656507e-06, + "loss": 0.49363499879837036, + "step": 1754 + }, + { + "epoch": 3.712473572938689, + "grad_norm": 0.9981492161750793, + "learning_rate": 1.1284547453491734e-06, + "loss": 0.5495156645774841, + "step": 1756 + }, + { + "epoch": 3.7167019027484143, + "grad_norm": 1.9682940244674683, + "learning_rate": 1.1247489174297372e-06, + "loss": 0.3875855803489685, + "step": 1758 + }, + { + "epoch": 3.7209302325581395, + "grad_norm": 0.21128438413143158, + "learning_rate": 1.1210965787125743e-06, + "loss": 0.3480808734893799, + "step": 1760 + }, + { + "epoch": 3.7251585623678647, + "grad_norm": 1.0908961296081543, + "learning_rate": 1.1174977738489955e-06, + "loss": 0.7193361520767212, + "step": 1762 + }, + { + "epoch": 3.72938689217759, + "grad_norm": 0.9672446846961975, + "learning_rate": 1.113952546835835e-06, + "loss": 0.43024134635925293, + "step": 1764 + }, + { + "epoch": 3.733615221987315, + "grad_norm": 1.4315491914749146, + "learning_rate": 1.1104609410149174e-06, + "loss": 0.6568301320075989, + "step": 1766 + }, + { + "epoch": 3.7378435517970403, + "grad_norm": 1.2984766960144043, + "learning_rate": 1.1070229990725262e-06, + "loss": 0.6367031335830688, + "step": 1768 + }, + { + "epoch": 3.7420718816067655, + "grad_norm": 0.3808463513851166, + "learning_rate": 1.1036387630388841e-06, + "loss": 0.15784475207328796, + "step": 1770 + }, + { + "epoch": 3.7463002114164903, + "grad_norm": 1.1987873315811157, + "learning_rate": 1.100308274287636e-06, + "loss": 0.3977484703063965, + "step": 1772 + }, + { + "epoch": 3.7505285412262155, + "grad_norm": 2.901031732559204, + "learning_rate": 1.0970315735353434e-06, + "loss": 0.9629654288291931, + "step": 1774 + }, + { + "epoch": 3.7547568710359407, + "grad_norm": 1.2292399406433105, + "learning_rate": 1.0938087008409909e-06, + "loss": 0.7654396891593933, + "step": 1776 + }, + { + "epoch": 3.758985200845666, + "grad_norm": 1.3905978202819824, + "learning_rate": 1.0906396956054919e-06, + "loss": 0.7578558921813965, + "step": 1778 + }, + { + "epoch": 3.763213530655391, + "grad_norm": 1.8507306575775146, + "learning_rate": 1.0875245965712077e-06, + "loss": 0.8205432295799255, + "step": 1780 + }, + { + "epoch": 3.7674418604651163, + "grad_norm": 2.31853985786438, + "learning_rate": 1.0844634418214763e-06, + "loss": 0.6655327677726746, + "step": 1782 + }, + { + "epoch": 3.7716701902748415, + "grad_norm": 4.334614276885986, + "learning_rate": 1.0814562687801448e-06, + "loss": 0.11606885492801666, + "step": 1784 + }, + { + "epoch": 3.7758985200845667, + "grad_norm": 1.465037226676941, + "learning_rate": 1.0785031142111103e-06, + "loss": 0.5141998529434204, + "step": 1786 + }, + { + "epoch": 3.780126849894292, + "grad_norm": 74.19483184814453, + "learning_rate": 1.075604014217875e-06, + "loss": 0.4380958676338196, + "step": 1788 + }, + { + "epoch": 3.7843551797040167, + "grad_norm": 2.460592746734619, + "learning_rate": 1.0727590042431009e-06, + "loss": 0.14853282272815704, + "step": 1790 + }, + { + "epoch": 3.7885835095137423, + "grad_norm": 0.42049047350883484, + "learning_rate": 1.0699681190681777e-06, + "loss": 0.6219359040260315, + "step": 1792 + }, + { + "epoch": 3.792811839323467, + "grad_norm": 0.9896020293235779, + "learning_rate": 1.067231392812798e-06, + "loss": 0.7149208188056946, + "step": 1794 + }, + { + "epoch": 3.7970401691331923, + "grad_norm": 1.430011510848999, + "learning_rate": 1.0645488589345396e-06, + "loss": 0.4595027267932892, + "step": 1796 + }, + { + "epoch": 3.8012684989429175, + "grad_norm": 2.8460967540740967, + "learning_rate": 1.0619205502284574e-06, + "loss": 0.6715288758277893, + "step": 1798 + }, + { + "epoch": 3.8054968287526427, + "grad_norm": 1.2419524192810059, + "learning_rate": 1.0593464988266811e-06, + "loss": 0.7421883344650269, + "step": 1800 + }, + { + "epoch": 3.809725158562368, + "grad_norm": 1.353837251663208, + "learning_rate": 1.0568267361980225e-06, + "loss": 0.5135189294815063, + "step": 1802 + }, + { + "epoch": 3.813953488372093, + "grad_norm": 1.2597968578338623, + "learning_rate": 1.0543612931475935e-06, + "loss": 0.7062614560127258, + "step": 1804 + }, + { + "epoch": 3.8181818181818183, + "grad_norm": 1.2221826314926147, + "learning_rate": 1.0519501998164247e-06, + "loss": 0.7496445775032043, + "step": 1806 + }, + { + "epoch": 3.822410147991543, + "grad_norm": 1.7406678199768066, + "learning_rate": 1.049593485681101e-06, + "loss": 0.1540011167526245, + "step": 1808 + }, + { + "epoch": 3.8266384778012688, + "grad_norm": 1.5595349073410034, + "learning_rate": 1.0472911795533997e-06, + "loss": 0.8456817269325256, + "step": 1810 + }, + { + "epoch": 3.8308668076109935, + "grad_norm": 1.3645883798599243, + "learning_rate": 1.045043309579938e-06, + "loss": 0.8958571553230286, + "step": 1812 + }, + { + "epoch": 3.8350951374207187, + "grad_norm": 0.9170961976051331, + "learning_rate": 1.0428499032418297e-06, + "loss": 0.5953072309494019, + "step": 1814 + }, + { + "epoch": 3.839323467230444, + "grad_norm": 2.077226400375366, + "learning_rate": 1.0407109873543484e-06, + "loss": 0.5082079768180847, + "step": 1816 + }, + { + "epoch": 3.843551797040169, + "grad_norm": 2.8861138820648193, + "learning_rate": 1.0386265880665996e-06, + "loss": 0.45712658762931824, + "step": 1818 + }, + { + "epoch": 3.8477801268498943, + "grad_norm": 2.39978289604187, + "learning_rate": 1.036596730861203e-06, + "loss": 0.43675118684768677, + "step": 1820 + }, + { + "epoch": 3.8520084566596196, + "grad_norm": 0.540006160736084, + "learning_rate": 1.0346214405539783e-06, + "loss": 0.4738779366016388, + "step": 1822 + }, + { + "epoch": 3.8562367864693448, + "grad_norm": 0.9755803942680359, + "learning_rate": 1.0327007412936419e-06, + "loss": 0.47678855061531067, + "step": 1824 + }, + { + "epoch": 3.8604651162790695, + "grad_norm": 2.5850977897644043, + "learning_rate": 1.0308346565615158e-06, + "loss": 0.4452104866504669, + "step": 1826 + }, + { + "epoch": 3.864693446088795, + "grad_norm": 0.8845793008804321, + "learning_rate": 1.0290232091712337e-06, + "loss": 0.7840569615364075, + "step": 1828 + }, + { + "epoch": 3.86892177589852, + "grad_norm": 1.2019164562225342, + "learning_rate": 1.0272664212684687e-06, + "loss": 0.7792611122131348, + "step": 1830 + }, + { + "epoch": 3.873150105708245, + "grad_norm": 1.1950770616531372, + "learning_rate": 1.025564314330659e-06, + "loss": 0.4541883170604706, + "step": 1832 + }, + { + "epoch": 3.8773784355179703, + "grad_norm": 1.4915878772735596, + "learning_rate": 1.0239169091667447e-06, + "loss": 0.39403173327445984, + "step": 1834 + }, + { + "epoch": 3.8816067653276956, + "grad_norm": 0.3940155804157257, + "learning_rate": 1.022324225916917e-06, + "loss": 0.3465203046798706, + "step": 1836 + }, + { + "epoch": 3.8858350951374208, + "grad_norm": 0.11024730652570724, + "learning_rate": 1.0207862840523672e-06, + "loss": 0.37984877824783325, + "step": 1838 + }, + { + "epoch": 3.890063424947146, + "grad_norm": 1.5385187864303589, + "learning_rate": 1.0193031023750534e-06, + "loss": 0.6312057971954346, + "step": 1840 + }, + { + "epoch": 3.894291754756871, + "grad_norm": 1.3971006870269775, + "learning_rate": 1.017874699017468e-06, + "loss": 0.6595142483711243, + "step": 1842 + }, + { + "epoch": 3.898520084566596, + "grad_norm": 1.2199381589889526, + "learning_rate": 1.0165010914424142e-06, + "loss": 0.4192304313182831, + "step": 1844 + }, + { + "epoch": 3.9027484143763216, + "grad_norm": 7.076661586761475, + "learning_rate": 1.015182296442798e-06, + "loss": 0.4752802848815918, + "step": 1846 + }, + { + "epoch": 3.9069767441860463, + "grad_norm": 1.413932204246521, + "learning_rate": 1.0139183301414176e-06, + "loss": 0.11056225001811981, + "step": 1848 + }, + { + "epoch": 3.9112050739957716, + "grad_norm": 1.1691763401031494, + "learning_rate": 1.0127092079907682e-06, + "loss": 0.7014780044555664, + "step": 1850 + }, + { + "epoch": 3.9154334038054968, + "grad_norm": 0.31561478972435, + "learning_rate": 1.0115549447728551e-06, + "loss": 0.4741799533367157, + "step": 1852 + }, + { + "epoch": 3.919661733615222, + "grad_norm": 0.8321080207824707, + "learning_rate": 1.0104555545990088e-06, + "loss": 0.6726140975952148, + "step": 1854 + }, + { + "epoch": 3.923890063424947, + "grad_norm": 2.538691759109497, + "learning_rate": 1.009411050909716e-06, + "loss": 0.37050968408584595, + "step": 1856 + }, + { + "epoch": 3.9281183932346724, + "grad_norm": 3.0525474548339844, + "learning_rate": 1.008421446474455e-06, + "loss": 0.6116885542869568, + "step": 1858 + }, + { + "epoch": 3.9323467230443976, + "grad_norm": 4.915146827697754, + "learning_rate": 1.0074867533915367e-06, + "loss": 0.2633376121520996, + "step": 1860 + }, + { + "epoch": 3.9365750528541223, + "grad_norm": 2.6079750061035156, + "learning_rate": 1.0066069830879593e-06, + "loss": 0.7564821839332581, + "step": 1862 + }, + { + "epoch": 3.940803382663848, + "grad_norm": 1.023239254951477, + "learning_rate": 1.0057821463192687e-06, + "loss": 0.7867894172668457, + "step": 1864 + }, + { + "epoch": 3.9450317124735728, + "grad_norm": 0.9829970002174377, + "learning_rate": 1.005012253169426e-06, + "loss": 0.6965225338935852, + "step": 1866 + }, + { + "epoch": 3.949260042283298, + "grad_norm": 0.016823846846818924, + "learning_rate": 1.0042973130506839e-06, + "loss": 0.0010633495403453708, + "step": 1868 + }, + { + "epoch": 3.953488372093023, + "grad_norm": 0.9858480095863342, + "learning_rate": 1.003637334703472e-06, + "loss": 0.8069336414337158, + "step": 1870 + }, + { + "epoch": 3.9577167019027484, + "grad_norm": 2.66190767288208, + "learning_rate": 1.0030323261962922e-06, + "loss": 0.5338172316551208, + "step": 1872 + }, + { + "epoch": 3.9619450317124736, + "grad_norm": 1.0303117036819458, + "learning_rate": 1.002482294925615e-06, + "loss": 0.38796931505203247, + "step": 1874 + }, + { + "epoch": 3.966173361522199, + "grad_norm": 1.3312749862670898, + "learning_rate": 1.0019872476157952e-06, + "loss": 0.7693368792533875, + "step": 1876 + }, + { + "epoch": 3.970401691331924, + "grad_norm": 3.147160530090332, + "learning_rate": 1.0015471903189834e-06, + "loss": 0.16337493062019348, + "step": 1878 + }, + { + "epoch": 3.974630021141649, + "grad_norm": 1.2491008043289185, + "learning_rate": 1.0011621284150583e-06, + "loss": 0.3781284689903259, + "step": 1880 + }, + { + "epoch": 3.9788583509513744, + "grad_norm": 1.9320218563079834, + "learning_rate": 1.000832066611556e-06, + "loss": 0.685487687587738, + "step": 1882 + }, + { + "epoch": 3.983086680761099, + "grad_norm": 1.3038630485534668, + "learning_rate": 1.0005570089436137e-06, + "loss": 0.839418351650238, + "step": 1884 + }, + { + "epoch": 3.9873150105708244, + "grad_norm": 0.9190782904624939, + "learning_rate": 1.0003369587739234e-06, + "loss": 0.5125820636749268, + "step": 1886 + }, + { + "epoch": 3.9915433403805496, + "grad_norm": 0.7910130620002747, + "learning_rate": 1.0001719187926854e-06, + "loss": 0.7280488610267639, + "step": 1888 + }, + { + "epoch": 3.995771670190275, + "grad_norm": 1.1570515632629395, + "learning_rate": 1.0000618910175811e-06, + "loss": 0.7407705783843994, + "step": 1890 + }, + { + "epoch": 4.0, + "grad_norm": 0.9717517495155334, + "learning_rate": 1.000006876793743e-06, + "loss": 0.4140605926513672, + "step": 1892 + }, + { + "epoch": 4.0, + "step": 1892, + "total_flos": 3.5845266747435254e+18, + "train_loss": 0.8296794885006264, + "train_runtime": 8766.5014, + "train_samples_per_second": 6.475, + "train_steps_per_second": 0.216 + } + ], + "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 +}