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": 2.0970451831817627, + "learning_rate": 1.0526315789473685e-07, + "loss": 1.8015528917312622, + "step": 2 + }, + { + "epoch": 0.008456659619450317, + "grad_norm": 0.9556745886802673, + "learning_rate": 3.1578947368421055e-07, + "loss": 2.230623960494995, + "step": 4 + }, + { + "epoch": 0.012684989429175475, + "grad_norm": 7.332006454467773, + "learning_rate": 5.263157894736843e-07, + "loss": 2.2299251556396484, + "step": 6 + }, + { + "epoch": 0.016913319238900635, + "grad_norm": 1.1154991388320923, + "learning_rate": 7.368421052631579e-07, + "loss": 1.994123935699463, + "step": 8 + }, + { + "epoch": 0.021141649048625793, + "grad_norm": 4.7002692222595215, + "learning_rate": 9.473684210526317e-07, + "loss": 2.0223240852355957, + "step": 10 + }, + { + "epoch": 0.02536997885835095, + "grad_norm": 1.8593226671218872, + "learning_rate": 1.1578947368421053e-06, + "loss": 1.6351479291915894, + "step": 12 + }, + { + "epoch": 0.02959830866807611, + "grad_norm": 0.87593674659729, + "learning_rate": 1.3684210526315791e-06, + "loss": 1.6405255794525146, + "step": 14 + }, + { + "epoch": 0.03382663847780127, + "grad_norm": 1.3401015996932983, + "learning_rate": 1.5789473684210526e-06, + "loss": 1.7314574718475342, + "step": 16 + }, + { + "epoch": 0.03805496828752643, + "grad_norm": 1.8513972759246826, + "learning_rate": 1.7894736842105265e-06, + "loss": 1.7740224599838257, + "step": 18 + }, + { + "epoch": 0.042283298097251586, + "grad_norm": 2.61112642288208, + "learning_rate": 2.0000000000000003e-06, + "loss": 1.657949447631836, + "step": 20 + }, + { + "epoch": 0.046511627906976744, + "grad_norm": 2.832732677459717, + "learning_rate": 2.2105263157894738e-06, + "loss": 1.5246399641036987, + "step": 22 + }, + { + "epoch": 0.0507399577167019, + "grad_norm": 1.9506855010986328, + "learning_rate": 2.4210526315789477e-06, + "loss": 1.7635947465896606, + "step": 24 + }, + { + "epoch": 0.05496828752642706, + "grad_norm": 1.782944679260254, + "learning_rate": 2.631578947368421e-06, + "loss": 1.3579049110412598, + "step": 26 + }, + { + "epoch": 0.05919661733615222, + "grad_norm": 0.627625584602356, + "learning_rate": 2.842105263157895e-06, + "loss": 1.6494345664978027, + "step": 28 + }, + { + "epoch": 0.06342494714587738, + "grad_norm": 1.5023951530456543, + "learning_rate": 3.052631578947369e-06, + "loss": 1.2238823175430298, + "step": 30 + }, + { + "epoch": 0.06765327695560254, + "grad_norm": 2.2942779064178467, + "learning_rate": 3.2631578947368423e-06, + "loss": 1.0000499486923218, + "step": 32 + }, + { + "epoch": 0.07188160676532769, + "grad_norm": 0.9715867042541504, + "learning_rate": 3.473684210526316e-06, + "loss": 1.1595420837402344, + "step": 34 + }, + { + "epoch": 0.07610993657505286, + "grad_norm": 0.9334418773651123, + "learning_rate": 3.6842105263157896e-06, + "loss": 1.353593111038208, + "step": 36 + }, + { + "epoch": 0.080338266384778, + "grad_norm": 10.774386405944824, + "learning_rate": 3.894736842105263e-06, + "loss": 1.2713336944580078, + "step": 38 + }, + { + "epoch": 0.08456659619450317, + "grad_norm": 8.507015228271484, + "learning_rate": 4.105263157894737e-06, + "loss": 1.2290350198745728, + "step": 40 + }, + { + "epoch": 0.08879492600422834, + "grad_norm": 4.769407272338867, + "learning_rate": 4.315789473684211e-06, + "loss": 1.6501798629760742, + "step": 42 + }, + { + "epoch": 0.09302325581395349, + "grad_norm": 1.0019969940185547, + "learning_rate": 4.526315789473685e-06, + "loss": 1.6190237998962402, + "step": 44 + }, + { + "epoch": 0.09725158562367865, + "grad_norm": 0.89243084192276, + "learning_rate": 4.736842105263158e-06, + "loss": 1.2771525382995605, + "step": 46 + }, + { + "epoch": 0.1014799154334038, + "grad_norm": 0.7165584564208984, + "learning_rate": 4.947368421052632e-06, + "loss": 1.5023313760757446, + "step": 48 + }, + { + "epoch": 0.10570824524312897, + "grad_norm": 1.4044665098190308, + "learning_rate": 5.157894736842106e-06, + "loss": 1.4827324151992798, + "step": 50 + }, + { + "epoch": 0.10993657505285412, + "grad_norm": 1.7053216695785522, + "learning_rate": 5.36842105263158e-06, + "loss": 0.6851487159729004, + "step": 52 + }, + { + "epoch": 0.11416490486257928, + "grad_norm": 0.4792521893978119, + "learning_rate": 5.578947368421052e-06, + "loss": 1.06097412109375, + "step": 54 + }, + { + "epoch": 0.11839323467230443, + "grad_norm": 0.925794243812561, + "learning_rate": 5.789473684210527e-06, + "loss": 1.3648918867111206, + "step": 56 + }, + { + "epoch": 0.1226215644820296, + "grad_norm": 0.6835237145423889, + "learning_rate": 6e-06, + "loss": 1.386749267578125, + "step": 58 + }, + { + "epoch": 0.12684989429175475, + "grad_norm": 1.7157082557678223, + "learning_rate": 6.2105263157894745e-06, + "loss": 1.3997408151626587, + "step": 60 + }, + { + "epoch": 0.13107822410147993, + "grad_norm": 1.6797003746032715, + "learning_rate": 6.421052631578948e-06, + "loss": 0.7698476314544678, + "step": 62 + }, + { + "epoch": 0.13530655391120508, + "grad_norm": 0.9130530953407288, + "learning_rate": 6.631578947368421e-06, + "loss": 1.0380445718765259, + "step": 64 + }, + { + "epoch": 0.13953488372093023, + "grad_norm": 0.6532490849494934, + "learning_rate": 6.842105263157896e-06, + "loss": 1.3450121879577637, + "step": 66 + }, + { + "epoch": 0.14376321353065538, + "grad_norm": 0.7031619548797607, + "learning_rate": 7.052631578947369e-06, + "loss": 1.3845361471176147, + "step": 68 + }, + { + "epoch": 0.14799154334038056, + "grad_norm": 0.8025256395339966, + "learning_rate": 7.263157894736843e-06, + "loss": 1.352820634841919, + "step": 70 + }, + { + "epoch": 0.1522198731501057, + "grad_norm": 3.4256932735443115, + "learning_rate": 7.473684210526316e-06, + "loss": 1.3450026512145996, + "step": 72 + }, + { + "epoch": 0.15644820295983086, + "grad_norm": 0.9847568869590759, + "learning_rate": 7.68421052631579e-06, + "loss": 0.9240670800209045, + "step": 74 + }, + { + "epoch": 0.160676532769556, + "grad_norm": 1.1269018650054932, + "learning_rate": 7.894736842105265e-06, + "loss": 0.8439642786979675, + "step": 76 + }, + { + "epoch": 0.1649048625792812, + "grad_norm": 1.0185000896453857, + "learning_rate": 8.105263157894736e-06, + "loss": 1.3027528524398804, + "step": 78 + }, + { + "epoch": 0.16913319238900634, + "grad_norm": 1.9112207889556885, + "learning_rate": 8.315789473684212e-06, + "loss": 1.081552267074585, + "step": 80 + }, + { + "epoch": 0.1733615221987315, + "grad_norm": 2.896510124206543, + "learning_rate": 8.526315789473685e-06, + "loss": 1.1126136779785156, + "step": 82 + }, + { + "epoch": 0.17758985200845667, + "grad_norm": 0.6804601550102234, + "learning_rate": 8.736842105263158e-06, + "loss": 0.8543714880943298, + "step": 84 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 1.2680926322937012, + "learning_rate": 8.947368421052632e-06, + "loss": 1.7425063848495483, + "step": 86 + }, + { + "epoch": 0.18604651162790697, + "grad_norm": 0.6533979177474976, + "learning_rate": 9.157894736842105e-06, + "loss": 1.289363980293274, + "step": 88 + }, + { + "epoch": 0.19027484143763213, + "grad_norm": 1.4818757772445679, + "learning_rate": 9.36842105263158e-06, + "loss": 1.0200517177581787, + "step": 90 + }, + { + "epoch": 0.1945031712473573, + "grad_norm": 0.986833930015564, + "learning_rate": 9.578947368421054e-06, + "loss": 1.313726544380188, + "step": 92 + }, + { + "epoch": 0.19873150105708245, + "grad_norm": 0.8675680160522461, + "learning_rate": 9.789473684210527e-06, + "loss": 0.9839202761650085, + "step": 94 + }, + { + "epoch": 0.2029598308668076, + "grad_norm": 1.012480616569519, + "learning_rate": 1e-05, + "loss": 1.298305869102478, + "step": 96 + }, + { + "epoch": 0.20718816067653276, + "grad_norm": 1.0893498659133911, + "learning_rate": 9.999972492846047e-06, + "loss": 1.4132791757583618, + "step": 98 + }, + { + "epoch": 0.21141649048625794, + "grad_norm": 2.9394783973693848, + "learning_rate": 9.999889971720468e-06, + "loss": 1.364189863204956, + "step": 100 + }, + { + "epoch": 0.2156448202959831, + "grad_norm": 0.45424124598503113, + "learning_rate": 9.999752437632121e-06, + "loss": 1.2472785711288452, + "step": 102 + }, + { + "epoch": 0.21987315010570824, + "grad_norm": 0.7732388973236084, + "learning_rate": 9.999559892262412e-06, + "loss": 1.3142788410186768, + "step": 104 + }, + { + "epoch": 0.22410147991543342, + "grad_norm": 0.7232316732406616, + "learning_rate": 9.999312337965291e-06, + "loss": 1.304768443107605, + "step": 106 + }, + { + "epoch": 0.22832980972515857, + "grad_norm": 1.3244025707244873, + "learning_rate": 9.999009777767202e-06, + "loss": 1.0288143157958984, + "step": 108 + }, + { + "epoch": 0.23255813953488372, + "grad_norm": 0.7635775208473206, + "learning_rate": 9.99865221536707e-06, + "loss": 1.2872488498687744, + "step": 110 + }, + { + "epoch": 0.23678646934460887, + "grad_norm": 2.2536940574645996, + "learning_rate": 9.998239655136237e-06, + "loss": 1.3732539415359497, + "step": 112 + }, + { + "epoch": 0.24101479915433405, + "grad_norm": 1.013115406036377, + "learning_rate": 9.997772102118417e-06, + "loss": 1.5521059036254883, + "step": 114 + }, + { + "epoch": 0.2452431289640592, + "grad_norm": 1.0717524290084839, + "learning_rate": 9.997249562029637e-06, + "loss": 1.3396289348602295, + "step": 116 + }, + { + "epoch": 0.24947145877378435, + "grad_norm": 2.3264575004577637, + "learning_rate": 9.996672041258154e-06, + "loss": 1.3412898778915405, + "step": 118 + }, + { + "epoch": 0.2536997885835095, + "grad_norm": 0.8798300623893738, + "learning_rate": 9.996039546864395e-06, + "loss": 0.8074193596839905, + "step": 120 + }, + { + "epoch": 0.25792811839323465, + "grad_norm": 0.6155166029930115, + "learning_rate": 9.995352086580858e-06, + "loss": 1.3092403411865234, + "step": 122 + }, + { + "epoch": 0.26215644820295986, + "grad_norm": 0.6969359517097473, + "learning_rate": 9.994609668812022e-06, + "loss": 1.127609372138977, + "step": 124 + }, + { + "epoch": 0.266384778012685, + "grad_norm": 1.1108065843582153, + "learning_rate": 9.993812302634237e-06, + "loss": 1.3442314863204956, + "step": 126 + }, + { + "epoch": 0.27061310782241016, + "grad_norm": 0.8615521788597107, + "learning_rate": 9.99295999779563e-06, + "loss": 0.95571368932724, + "step": 128 + }, + { + "epoch": 0.2748414376321353, + "grad_norm": 0.9894189834594727, + "learning_rate": 9.99205276471597e-06, + "loss": 0.9644747376441956, + "step": 130 + }, + { + "epoch": 0.27906976744186046, + "grad_norm": 1.4779452085494995, + "learning_rate": 9.991090614486546e-06, + "loss": 1.2963494062423706, + "step": 132 + }, + { + "epoch": 0.2832980972515856, + "grad_norm": 1.1073240041732788, + "learning_rate": 9.990073558870028e-06, + "loss": 0.8963363170623779, + "step": 134 + }, + { + "epoch": 0.28752642706131076, + "grad_norm": 2.0163443088531494, + "learning_rate": 9.989001610300333e-06, + "loss": 1.1175273656845093, + "step": 136 + }, + { + "epoch": 0.2917547568710359, + "grad_norm": 2.5545969009399414, + "learning_rate": 9.987874781882461e-06, + "loss": 1.455842137336731, + "step": 138 + }, + { + "epoch": 0.2959830866807611, + "grad_norm": 0.5453028678894043, + "learning_rate": 9.986693087392341e-06, + "loss": 0.9483217597007751, + "step": 140 + }, + { + "epoch": 0.30021141649048627, + "grad_norm": 0.6908959746360779, + "learning_rate": 9.985456541276667e-06, + "loss": 1.5587284564971924, + "step": 142 + }, + { + "epoch": 0.3044397463002114, + "grad_norm": 2.3970558643341064, + "learning_rate": 9.984165158652708e-06, + "loss": 1.1079477071762085, + "step": 144 + }, + { + "epoch": 0.3086680761099366, + "grad_norm": 0.8545916676521301, + "learning_rate": 9.982818955308135e-06, + "loss": 0.6507930755615234, + "step": 146 + }, + { + "epoch": 0.3128964059196617, + "grad_norm": 0.9650726318359375, + "learning_rate": 9.98141794770083e-06, + "loss": 1.2380722761154175, + "step": 148 + }, + { + "epoch": 0.3171247357293869, + "grad_norm": 0.8409858345985413, + "learning_rate": 9.979962152958669e-06, + "loss": 1.2685638666152954, + "step": 150 + }, + { + "epoch": 0.321353065539112, + "grad_norm": 0.8057692646980286, + "learning_rate": 9.97845158887933e-06, + "loss": 1.1973340511322021, + "step": 152 + }, + { + "epoch": 0.32558139534883723, + "grad_norm": 2.106417417526245, + "learning_rate": 9.976886273930063e-06, + "loss": 0.7574591040611267, + "step": 154 + }, + { + "epoch": 0.3298097251585624, + "grad_norm": 0.8973696231842041, + "learning_rate": 9.975266227247474e-06, + "loss": 1.2414660453796387, + "step": 156 + }, + { + "epoch": 0.33403805496828753, + "grad_norm": 0.6619360446929932, + "learning_rate": 9.973591468637286e-06, + "loss": 1.2720128297805786, + "step": 158 + }, + { + "epoch": 0.3382663847780127, + "grad_norm": 1.33470618724823, + "learning_rate": 9.971862018574094e-06, + "loss": 1.2611830234527588, + "step": 160 + }, + { + "epoch": 0.34249471458773784, + "grad_norm": 3.085461139678955, + "learning_rate": 9.97007789820112e-06, + "loss": 1.0963258743286133, + "step": 162 + }, + { + "epoch": 0.346723044397463, + "grad_norm": 1.3238252401351929, + "learning_rate": 9.968239129329953e-06, + "loss": 1.0119415521621704, + "step": 164 + }, + { + "epoch": 0.35095137420718814, + "grad_norm": 0.9732130765914917, + "learning_rate": 9.96634573444028e-06, + "loss": 0.8724996447563171, + "step": 166 + }, + { + "epoch": 0.35517970401691334, + "grad_norm": 0.667926549911499, + "learning_rate": 9.964397736679615e-06, + "loss": 1.1872084140777588, + "step": 168 + }, + { + "epoch": 0.3594080338266385, + "grad_norm": 0.8903693556785583, + "learning_rate": 9.962395159863014e-06, + "loss": 1.2049840688705444, + "step": 170 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 0.5408523678779602, + "learning_rate": 9.960338028472781e-06, + "loss": 1.038049578666687, + "step": 172 + }, + { + "epoch": 0.3678646934460888, + "grad_norm": 1.0019752979278564, + "learning_rate": 9.958226367658176e-06, + "loss": 1.3585528135299683, + "step": 174 + }, + { + "epoch": 0.37209302325581395, + "grad_norm": 0.8384218215942383, + "learning_rate": 9.9560602032351e-06, + "loss": 1.0646179914474487, + "step": 176 + }, + { + "epoch": 0.3763213530655391, + "grad_norm": 1.0649065971374512, + "learning_rate": 9.953839561685783e-06, + "loss": 0.8854655623435974, + "step": 178 + }, + { + "epoch": 0.38054968287526425, + "grad_norm": 0.7833120226860046, + "learning_rate": 9.95156447015846e-06, + "loss": 1.2828737497329712, + "step": 180 + }, + { + "epoch": 0.38477801268498946, + "grad_norm": 0.7969745993614197, + "learning_rate": 9.94923495646704e-06, + "loss": 1.5253641605377197, + "step": 182 + }, + { + "epoch": 0.3890063424947146, + "grad_norm": 0.7941175103187561, + "learning_rate": 9.946851049090764e-06, + "loss": 1.3306870460510254, + "step": 184 + }, + { + "epoch": 0.39323467230443976, + "grad_norm": 0.7314859628677368, + "learning_rate": 9.944412777173854e-06, + "loss": 1.0905207395553589, + "step": 186 + }, + { + "epoch": 0.3974630021141649, + "grad_norm": 0.8418819904327393, + "learning_rate": 9.941920170525168e-06, + "loss": 0.9973200559616089, + "step": 188 + }, + { + "epoch": 0.40169133192389006, + "grad_norm": 0.8657261729240417, + "learning_rate": 9.939373259617822e-06, + "loss": 0.9982656240463257, + "step": 190 + }, + { + "epoch": 0.4059196617336152, + "grad_norm": 1.027279257774353, + "learning_rate": 9.936772075588825e-06, + "loss": 1.2934733629226685, + "step": 192 + }, + { + "epoch": 0.41014799154334036, + "grad_norm": 0.40911751985549927, + "learning_rate": 9.934116650238697e-06, + "loss": 1.0202374458312988, + "step": 194 + }, + { + "epoch": 0.4143763213530655, + "grad_norm": 0.9035235643386841, + "learning_rate": 9.931407016031078e-06, + "loss": 1.2838034629821777, + "step": 196 + }, + { + "epoch": 0.4186046511627907, + "grad_norm": 0.7726561427116394, + "learning_rate": 9.928643206092338e-06, + "loss": 1.0557918548583984, + "step": 198 + }, + { + "epoch": 0.42283298097251587, + "grad_norm": 0.9403738975524902, + "learning_rate": 9.925825254211161e-06, + "loss": 1.1762772798538208, + "step": 200 + }, + { + "epoch": 0.427061310782241, + "grad_norm": 1.1015313863754272, + "learning_rate": 9.922953194838143e-06, + "loss": 0.5582849979400635, + "step": 202 + }, + { + "epoch": 0.4312896405919662, + "grad_norm": 3.7842440605163574, + "learning_rate": 9.920027063085363e-06, + "loss": 1.2378844022750854, + "step": 204 + }, + { + "epoch": 0.4355179704016913, + "grad_norm": 3.2515435218811035, + "learning_rate": 9.917046894725957e-06, + "loss": 1.3043760061264038, + "step": 206 + }, + { + "epoch": 0.4397463002114165, + "grad_norm": 2.2905519008636475, + "learning_rate": 9.91401272619368e-06, + "loss": 0.7985730171203613, + "step": 208 + }, + { + "epoch": 0.4439746300211416, + "grad_norm": 0.8532588481903076, + "learning_rate": 9.910924594582463e-06, + "loss": 1.2971242666244507, + "step": 210 + }, + { + "epoch": 0.44820295983086683, + "grad_norm": 0.7287648916244507, + "learning_rate": 9.907782537645955e-06, + "loss": 1.2710920572280884, + "step": 212 + }, + { + "epoch": 0.452431289640592, + "grad_norm": 0.9900949001312256, + "learning_rate": 9.904586593797062e-06, + "loss": 1.4209402799606323, + "step": 214 + }, + { + "epoch": 0.45665961945031713, + "grad_norm": 1.7588388919830322, + "learning_rate": 9.901336802107485e-06, + "loss": 0.6507283449172974, + "step": 216 + }, + { + "epoch": 0.4608879492600423, + "grad_norm": 6.916059494018555, + "learning_rate": 9.89803320230723e-06, + "loss": 0.9856141805648804, + "step": 218 + }, + { + "epoch": 0.46511627906976744, + "grad_norm": 0.4474513530731201, + "learning_rate": 9.894675834784136e-06, + "loss": 1.277772068977356, + "step": 220 + }, + { + "epoch": 0.4693446088794926, + "grad_norm": 0.6684972643852234, + "learning_rate": 9.891264740583367e-06, + "loss": 1.2334203720092773, + "step": 222 + }, + { + "epoch": 0.47357293868921774, + "grad_norm": 0.9211118817329407, + "learning_rate": 9.88779996140692e-06, + "loss": 1.3413673639297485, + "step": 224 + }, + { + "epoch": 0.47780126849894294, + "grad_norm": 0.6256110668182373, + "learning_rate": 9.884281539613109e-06, + "loss": 1.2514935731887817, + "step": 226 + }, + { + "epoch": 0.4820295983086681, + "grad_norm": 0.6157636046409607, + "learning_rate": 9.880709518216061e-06, + "loss": 1.2497587203979492, + "step": 228 + }, + { + "epoch": 0.48625792811839325, + "grad_norm": 0.8235659599304199, + "learning_rate": 9.87708394088517e-06, + "loss": 0.9096136689186096, + "step": 230 + }, + { + "epoch": 0.4904862579281184, + "grad_norm": 0.6840317249298096, + "learning_rate": 9.873404851944573e-06, + "loss": 1.2541745901107788, + "step": 232 + }, + { + "epoch": 0.49471458773784355, + "grad_norm": 2.53057599067688, + "learning_rate": 9.869672296372611e-06, + "loss": 0.9377322196960449, + "step": 234 + }, + { + "epoch": 0.4989429175475687, + "grad_norm": 1.7567822933197021, + "learning_rate": 9.86588631980128e-06, + "loss": 0.8832615613937378, + "step": 236 + }, + { + "epoch": 0.5031712473572939, + "grad_norm": 0.97667396068573, + "learning_rate": 9.86204696851566e-06, + "loss": 0.8286206126213074, + "step": 238 + }, + { + "epoch": 0.507399577167019, + "grad_norm": 0.9001383185386658, + "learning_rate": 9.858154289453363e-06, + "loss": 0.9000726342201233, + "step": 240 + }, + { + "epoch": 0.5116279069767442, + "grad_norm": 0.8837960362434387, + "learning_rate": 9.854208330203958e-06, + "loss": 1.309548258781433, + "step": 242 + }, + { + "epoch": 0.5158562367864693, + "grad_norm": 1.1845746040344238, + "learning_rate": 9.850209139008377e-06, + "loss": 1.358070969581604, + "step": 244 + }, + { + "epoch": 0.5200845665961945, + "grad_norm": 2.254845380783081, + "learning_rate": 9.846156764758344e-06, + "loss": 1.2843902111053467, + "step": 246 + }, + { + "epoch": 0.5243128964059197, + "grad_norm": 0.6787340044975281, + "learning_rate": 9.842051256995759e-06, + "loss": 1.2419553995132446, + "step": 248 + }, + { + "epoch": 0.5285412262156448, + "grad_norm": 1.2821767330169678, + "learning_rate": 9.837892665912105e-06, + "loss": 1.1701276302337646, + "step": 250 + }, + { + "epoch": 0.53276955602537, + "grad_norm": 2.022310256958008, + "learning_rate": 9.833681042347827e-06, + "loss": 1.0945854187011719, + "step": 252 + }, + { + "epoch": 0.5369978858350951, + "grad_norm": 1.2969770431518555, + "learning_rate": 9.829416437791721e-06, + "loss": 1.7462306022644043, + "step": 254 + }, + { + "epoch": 0.5412262156448203, + "grad_norm": 0.4596793055534363, + "learning_rate": 9.825098904380284e-06, + "loss": 1.2139190435409546, + "step": 256 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 0.9554147124290466, + "learning_rate": 9.820728494897101e-06, + "loss": 1.3912419080734253, + "step": 258 + }, + { + "epoch": 0.5496828752642706, + "grad_norm": 0.719840407371521, + "learning_rate": 9.816305262772182e-06, + "loss": 1.0096997022628784, + "step": 260 + }, + { + "epoch": 0.5539112050739958, + "grad_norm": 0.4836631119251251, + "learning_rate": 9.811829262081315e-06, + "loss": 1.1987851858139038, + "step": 262 + }, + { + "epoch": 0.5581395348837209, + "grad_norm": 10.481749534606934, + "learning_rate": 9.807300547545411e-06, + "loss": 1.077430248260498, + "step": 264 + }, + { + "epoch": 0.5623678646934461, + "grad_norm": 0.4831627607345581, + "learning_rate": 9.802719174529821e-06, + "loss": 1.4981955289840698, + "step": 266 + }, + { + "epoch": 0.5665961945031712, + "grad_norm": 1.6229199171066284, + "learning_rate": 9.798085199043672e-06, + "loss": 0.5668478608131409, + "step": 268 + }, + { + "epoch": 0.5708245243128964, + "grad_norm": 1.0242642164230347, + "learning_rate": 9.793398677739177e-06, + "loss": 1.0147303342819214, + "step": 270 + }, + { + "epoch": 0.5750528541226215, + "grad_norm": 1.0180574655532837, + "learning_rate": 9.788659667910942e-06, + "loss": 1.20408296585083, + "step": 272 + }, + { + "epoch": 0.5792811839323467, + "grad_norm": 0.6127535700798035, + "learning_rate": 9.783868227495262e-06, + "loss": 0.9585332274436951, + "step": 274 + }, + { + "epoch": 0.5835095137420718, + "grad_norm": 0.6249876618385315, + "learning_rate": 9.779024415069427e-06, + "loss": 0.9284272193908691, + "step": 276 + }, + { + "epoch": 0.587737843551797, + "grad_norm": 0.3046797215938568, + "learning_rate": 9.774128289850987e-06, + "loss": 0.8653221726417542, + "step": 278 + }, + { + "epoch": 0.5919661733615222, + "grad_norm": 0.7098410129547119, + "learning_rate": 9.769179911697037e-06, + "loss": 1.2278540134429932, + "step": 280 + }, + { + "epoch": 0.5961945031712473, + "grad_norm": 0.9684470295906067, + "learning_rate": 9.764179341103493e-06, + "loss": 1.455711841583252, + "step": 282 + }, + { + "epoch": 0.6004228329809725, + "grad_norm": 0.6033283472061157, + "learning_rate": 9.759126639204337e-06, + "loss": 0.9373559355735779, + "step": 284 + }, + { + "epoch": 0.6046511627906976, + "grad_norm": 0.5853375792503357, + "learning_rate": 9.75402186777088e-06, + "loss": 1.2781811952590942, + "step": 286 + }, + { + "epoch": 0.6088794926004228, + "grad_norm": 0.8122193217277527, + "learning_rate": 9.748865089211004e-06, + "loss": 0.4290362596511841, + "step": 288 + }, + { + "epoch": 0.6131078224101479, + "grad_norm": 1.188741683959961, + "learning_rate": 9.743656366568398e-06, + "loss": 0.971450924873352, + "step": 290 + }, + { + "epoch": 0.6173361522198731, + "grad_norm": 1.0113929510116577, + "learning_rate": 9.738395763521792e-06, + "loss": 0.8891305327415466, + "step": 292 + }, + { + "epoch": 0.6215644820295984, + "grad_norm": 0.7812364101409912, + "learning_rate": 9.733083344384164e-06, + "loss": 0.7128481268882751, + "step": 294 + }, + { + "epoch": 0.6257928118393234, + "grad_norm": 2.805624008178711, + "learning_rate": 9.727719174101981e-06, + "loss": 0.7171492576599121, + "step": 296 + }, + { + "epoch": 0.6300211416490487, + "grad_norm": 0.8489518761634827, + "learning_rate": 9.722303318254376e-06, + "loss": 1.3022106885910034, + "step": 298 + }, + { + "epoch": 0.6342494714587738, + "grad_norm": 0.8692296743392944, + "learning_rate": 9.71683584305236e-06, + "loss": 1.251131534576416, + "step": 300 + }, + { + "epoch": 0.638477801268499, + "grad_norm": 0.6883788704872131, + "learning_rate": 9.71131681533802e-06, + "loss": 0.7190396785736084, + "step": 302 + }, + { + "epoch": 0.642706131078224, + "grad_norm": 1.8142250776290894, + "learning_rate": 9.705746302583682e-06, + "loss": 0.8275232315063477, + "step": 304 + }, + { + "epoch": 0.6469344608879493, + "grad_norm": 0.5087741017341614, + "learning_rate": 9.700124372891104e-06, + "loss": 0.9223960041999817, + "step": 306 + }, + { + "epoch": 0.6511627906976745, + "grad_norm": 0.6673185229301453, + "learning_rate": 9.694451094990636e-06, + "loss": 1.2416517734527588, + "step": 308 + }, + { + "epoch": 0.6553911205073996, + "grad_norm": 1.1093357801437378, + "learning_rate": 9.68872653824038e-06, + "loss": 1.2088298797607422, + "step": 310 + }, + { + "epoch": 0.6596194503171248, + "grad_norm": 0.679948627948761, + "learning_rate": 9.68295077262534e-06, + "loss": 1.216339349746704, + "step": 312 + }, + { + "epoch": 0.6638477801268499, + "grad_norm": 1.10062837600708, + "learning_rate": 9.677123868756573e-06, + "loss": 1.239007830619812, + "step": 314 + }, + { + "epoch": 0.6680761099365751, + "grad_norm": 0.5273158550262451, + "learning_rate": 9.671245897870318e-06, + "loss": 1.0844545364379883, + "step": 316 + }, + { + "epoch": 0.6723044397463002, + "grad_norm": 1.866042137145996, + "learning_rate": 9.665316931827131e-06, + "loss": 0.774948239326477, + "step": 318 + }, + { + "epoch": 0.6765327695560254, + "grad_norm": 0.7016245722770691, + "learning_rate": 9.659337043111006e-06, + "loss": 1.2477519512176514, + "step": 320 + }, + { + "epoch": 0.6807610993657506, + "grad_norm": 0.8634307980537415, + "learning_rate": 9.653306304828483e-06, + "loss": 1.3163447380065918, + "step": 322 + }, + { + "epoch": 0.6849894291754757, + "grad_norm": 0.8132389187812805, + "learning_rate": 9.647224790707762e-06, + "loss": 1.0836232900619507, + "step": 324 + }, + { + "epoch": 0.6892177589852009, + "grad_norm": 0.8350993990898132, + "learning_rate": 9.641092575097795e-06, + "loss": 1.2737476825714111, + "step": 326 + }, + { + "epoch": 0.693446088794926, + "grad_norm": 1.1009629964828491, + "learning_rate": 9.63490973296738e-06, + "loss": 1.1934752464294434, + "step": 328 + }, + { + "epoch": 0.6976744186046512, + "grad_norm": 2.0710482597351074, + "learning_rate": 9.62867633990425e-06, + "loss": 0.8600136637687683, + "step": 330 + }, + { + "epoch": 0.7019027484143763, + "grad_norm": 0.8433480262756348, + "learning_rate": 9.622392472114136e-06, + "loss": 1.2586069107055664, + "step": 332 + }, + { + "epoch": 0.7061310782241015, + "grad_norm": 1.0908113718032837, + "learning_rate": 9.61605820641985e-06, + "loss": 1.2059975862503052, + "step": 334 + }, + { + "epoch": 0.7103594080338267, + "grad_norm": 0.6778643727302551, + "learning_rate": 9.60967362026033e-06, + "loss": 1.2187244892120361, + "step": 336 + }, + { + "epoch": 0.7145877378435518, + "grad_norm": 1.1395102739334106, + "learning_rate": 9.60323879168971e-06, + "loss": 1.3507561683654785, + "step": 338 + }, + { + "epoch": 0.718816067653277, + "grad_norm": 1.0423520803451538, + "learning_rate": 9.596753799376355e-06, + "loss": 1.2596522569656372, + "step": 340 + }, + { + "epoch": 0.7230443974630021, + "grad_norm": 1.8989545106887817, + "learning_rate": 9.5902187226019e-06, + "loss": 1.108319640159607, + "step": 342 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 4.0071797370910645, + "learning_rate": 9.583633641260285e-06, + "loss": 0.8814038634300232, + "step": 344 + }, + { + "epoch": 0.7315010570824524, + "grad_norm": 2.5225908756256104, + "learning_rate": 9.576998635856774e-06, + "loss": 0.8677898049354553, + "step": 346 + }, + { + "epoch": 0.7357293868921776, + "grad_norm": 1.2581987380981445, + "learning_rate": 9.570313787506974e-06, + "loss": 0.9733175039291382, + "step": 348 + }, + { + "epoch": 0.7399577167019028, + "grad_norm": 0.7074206471443176, + "learning_rate": 9.563579177935844e-06, + "loss": 1.307414174079895, + "step": 350 + }, + { + "epoch": 0.7441860465116279, + "grad_norm": 1.035905361175537, + "learning_rate": 9.556794889476687e-06, + "loss": 1.1537634134292603, + "step": 352 + }, + { + "epoch": 0.7484143763213531, + "grad_norm": 1.163731575012207, + "learning_rate": 9.549961005070161e-06, + "loss": 0.6972360610961914, + "step": 354 + }, + { + "epoch": 0.7526427061310782, + "grad_norm": 1.5618109703063965, + "learning_rate": 9.543077608263245e-06, + "loss": 1.0250357389450073, + "step": 356 + }, + { + "epoch": 0.7568710359408034, + "grad_norm": 0.5351282954216003, + "learning_rate": 9.536144783208229e-06, + "loss": 1.2143995761871338, + "step": 358 + }, + { + "epoch": 0.7610993657505285, + "grad_norm": 1.0775883197784424, + "learning_rate": 9.529162614661687e-06, + "loss": 0.7594963908195496, + "step": 360 + }, + { + "epoch": 0.7653276955602537, + "grad_norm": 0.7120230793952942, + "learning_rate": 9.522131187983433e-06, + "loss": 1.0900413990020752, + "step": 362 + }, + { + "epoch": 0.7695560253699789, + "grad_norm": 1.272334098815918, + "learning_rate": 9.515050589135485e-06, + "loss": 1.0598536729812622, + "step": 364 + }, + { + "epoch": 0.773784355179704, + "grad_norm": 0.7742589712142944, + "learning_rate": 9.507920904681006e-06, + "loss": 1.2683708667755127, + "step": 366 + }, + { + "epoch": 0.7780126849894292, + "grad_norm": 0.7138161659240723, + "learning_rate": 9.500742221783257e-06, + "loss": 1.213201880455017, + "step": 368 + }, + { + "epoch": 0.7822410147991543, + "grad_norm": 1.2583744525909424, + "learning_rate": 9.493514628204516e-06, + "loss": 0.8842138051986694, + "step": 370 + }, + { + "epoch": 0.7864693446088795, + "grad_norm": 0.5896151065826416, + "learning_rate": 9.48623821230502e-06, + "loss": 1.1948623657226562, + "step": 372 + }, + { + "epoch": 0.7906976744186046, + "grad_norm": 0.5596579313278198, + "learning_rate": 9.478913063041878e-06, + "loss": 1.2149494886398315, + "step": 374 + }, + { + "epoch": 0.7949260042283298, + "grad_norm": 0.80394047498703, + "learning_rate": 9.471539269967982e-06, + "loss": 1.300500750541687, + "step": 376 + }, + { + "epoch": 0.7991543340380549, + "grad_norm": 0.6019960045814514, + "learning_rate": 9.464116923230913e-06, + "loss": 0.9638695120811462, + "step": 378 + }, + { + "epoch": 0.8033826638477801, + "grad_norm": 1.139461636543274, + "learning_rate": 9.456646113571844e-06, + "loss": 1.2993332147598267, + "step": 380 + }, + { + "epoch": 0.8076109936575053, + "grad_norm": 1.0045840740203857, + "learning_rate": 9.449126932324424e-06, + "loss": 1.1740542650222778, + "step": 382 + }, + { + "epoch": 0.8118393234672304, + "grad_norm": 0.8173121213912964, + "learning_rate": 9.441559471413665e-06, + "loss": 0.961137592792511, + "step": 384 + }, + { + "epoch": 0.8160676532769556, + "grad_norm": 0.7342875599861145, + "learning_rate": 9.433943823354816e-06, + "loss": 0.9515892267227173, + "step": 386 + }, + { + "epoch": 0.8202959830866807, + "grad_norm": 0.5617311596870422, + "learning_rate": 9.426280081252232e-06, + "loss": 1.481034755706787, + "step": 388 + }, + { + "epoch": 0.8245243128964059, + "grad_norm": 1.1547036170959473, + "learning_rate": 9.418568338798245e-06, + "loss": 1.1565872430801392, + "step": 390 + }, + { + "epoch": 0.828752642706131, + "grad_norm": 3.4435670375823975, + "learning_rate": 9.410808690271999e-06, + "loss": 0.9369468092918396, + "step": 392 + }, + { + "epoch": 0.8329809725158562, + "grad_norm": 1.6313234567642212, + "learning_rate": 9.40300123053832e-06, + "loss": 1.2493743896484375, + "step": 394 + }, + { + "epoch": 0.8372093023255814, + "grad_norm": 0.21652188897132874, + "learning_rate": 9.395146055046535e-06, + "loss": 0.762942910194397, + "step": 396 + }, + { + "epoch": 0.8414376321353065, + "grad_norm": 3.157466173171997, + "learning_rate": 9.387243259829323e-06, + "loss": 0.6721683144569397, + "step": 398 + }, + { + "epoch": 0.8456659619450317, + "grad_norm": 1.6490358114242554, + "learning_rate": 9.37929294150153e-06, + "loss": 1.1250858306884766, + "step": 400 + }, + { + "epoch": 0.8498942917547568, + "grad_norm": 0.7184737920761108, + "learning_rate": 9.371295197258992e-06, + "loss": 0.9447345733642578, + "step": 402 + }, + { + "epoch": 0.854122621564482, + "grad_norm": 2.5029988288879395, + "learning_rate": 9.363250124877345e-06, + "loss": 0.8158255219459534, + "step": 404 + }, + { + "epoch": 0.8583509513742071, + "grad_norm": 2.717517614364624, + "learning_rate": 9.35515782271083e-06, + "loss": 0.9980059266090393, + "step": 406 + }, + { + "epoch": 0.8625792811839323, + "grad_norm": 0.4281213581562042, + "learning_rate": 9.347018389691093e-06, + "loss": 1.021666169166565, + "step": 408 + }, + { + "epoch": 0.8668076109936576, + "grad_norm": 1.0150679349899292, + "learning_rate": 9.338831925325974e-06, + "loss": 0.905613362789154, + "step": 410 + }, + { + "epoch": 0.8710359408033826, + "grad_norm": 1.5314679145812988, + "learning_rate": 9.330598529698287e-06, + "loss": 1.1555581092834473, + "step": 412 + }, + { + "epoch": 0.8752642706131079, + "grad_norm": 1.700905442237854, + "learning_rate": 9.322318303464603e-06, + "loss": 1.011090874671936, + "step": 414 + }, + { + "epoch": 0.879492600422833, + "grad_norm": 1.0510984659194946, + "learning_rate": 9.313991347854015e-06, + "loss": 1.1004329919815063, + "step": 416 + }, + { + "epoch": 0.8837209302325582, + "grad_norm": 0.7148085832595825, + "learning_rate": 9.305617764666897e-06, + "loss": 1.2213408946990967, + "step": 418 + }, + { + "epoch": 0.8879492600422833, + "grad_norm": 0.59690260887146, + "learning_rate": 9.297197656273672e-06, + "loss": 1.2020318508148193, + "step": 420 + }, + { + "epoch": 0.8921775898520085, + "grad_norm": 2.1009275913238525, + "learning_rate": 9.288731125613544e-06, + "loss": 0.9478926658630371, + "step": 422 + }, + { + "epoch": 0.8964059196617337, + "grad_norm": 0.446461945772171, + "learning_rate": 9.280218276193255e-06, + "loss": 1.2342218160629272, + "step": 424 + }, + { + "epoch": 0.9006342494714588, + "grad_norm": 0.4907039403915405, + "learning_rate": 9.271659212085807e-06, + "loss": 1.0680934190750122, + "step": 426 + }, + { + "epoch": 0.904862579281184, + "grad_norm": 1.4402127265930176, + "learning_rate": 9.263054037929199e-06, + "loss": 0.8719913959503174, + "step": 428 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 0.5996546745300293, + "learning_rate": 9.254402858925142e-06, + "loss": 1.0905053615570068, + "step": 430 + }, + { + "epoch": 0.9133192389006343, + "grad_norm": 2.596825361251831, + "learning_rate": 9.245705780837774e-06, + "loss": 1.2620584964752197, + "step": 432 + }, + { + "epoch": 0.9175475687103594, + "grad_norm": 2.524081230163574, + "learning_rate": 9.236962909992368e-06, + "loss": 0.6456525921821594, + "step": 434 + }, + { + "epoch": 0.9217758985200846, + "grad_norm": 2.527529001235962, + "learning_rate": 9.228174353274037e-06, + "loss": 0.37112611532211304, + "step": 436 + }, + { + "epoch": 0.9260042283298098, + "grad_norm": 0.7277577519416809, + "learning_rate": 9.219340218126413e-06, + "loss": 0.9164543151855469, + "step": 438 + }, + { + "epoch": 0.9302325581395349, + "grad_norm": 0.3515982925891876, + "learning_rate": 9.210460612550348e-06, + "loss": 1.1527671813964844, + "step": 440 + }, + { + "epoch": 0.9344608879492601, + "grad_norm": 1.6336746215820312, + "learning_rate": 9.20153564510259e-06, + "loss": 0.7140083909034729, + "step": 442 + }, + { + "epoch": 0.9386892177589852, + "grad_norm": 0.6215425729751587, + "learning_rate": 9.19256542489445e-06, + "loss": 1.2639110088348389, + "step": 444 + }, + { + "epoch": 0.9429175475687104, + "grad_norm": 0.6592838764190674, + "learning_rate": 9.183550061590476e-06, + "loss": 1.096636176109314, + "step": 446 + }, + { + "epoch": 0.9471458773784355, + "grad_norm": 3.0701558589935303, + "learning_rate": 9.174489665407103e-06, + "loss": 1.1096676588058472, + "step": 448 + }, + { + "epoch": 0.9513742071881607, + "grad_norm": 1.196798324584961, + "learning_rate": 9.165384347111317e-06, + "loss": 0.7902498245239258, + "step": 450 + }, + { + "epoch": 0.9556025369978859, + "grad_norm": 0.37907394766807556, + "learning_rate": 9.156234218019291e-06, + "loss": 0.9096274375915527, + "step": 452 + }, + { + "epoch": 0.959830866807611, + "grad_norm": 3.4193501472473145, + "learning_rate": 9.147039389995029e-06, + "loss": 1.3308203220367432, + "step": 454 + }, + { + "epoch": 0.9640591966173362, + "grad_norm": 1.7397775650024414, + "learning_rate": 9.137799975448998e-06, + "loss": 0.6499373912811279, + "step": 456 + }, + { + "epoch": 0.9682875264270613, + "grad_norm": 0.7532008290290833, + "learning_rate": 9.128516087336755e-06, + "loss": 0.8834115266799927, + "step": 458 + }, + { + "epoch": 0.9725158562367865, + "grad_norm": 0.6991927027702332, + "learning_rate": 9.119187839157559e-06, + "loss": 1.2351363897323608, + "step": 460 + }, + { + "epoch": 0.9767441860465116, + "grad_norm": 0.512692928314209, + "learning_rate": 9.109815344952995e-06, + "loss": 1.2225384712219238, + "step": 462 + }, + { + "epoch": 0.9809725158562368, + "grad_norm": 2.8695507049560547, + "learning_rate": 9.100398719305566e-06, + "loss": 0.7745890617370605, + "step": 464 + }, + { + "epoch": 0.985200845665962, + "grad_norm": 0.5231290459632874, + "learning_rate": 9.090938077337305e-06, + "loss": 1.2137622833251953, + "step": 466 + }, + { + "epoch": 0.9894291754756871, + "grad_norm": 1.8364567756652832, + "learning_rate": 9.081433534708363e-06, + "loss": 0.983952522277832, + "step": 468 + }, + { + "epoch": 0.9936575052854123, + "grad_norm": 0.9315457940101624, + "learning_rate": 9.071885207615591e-06, + "loss": 1.2546133995056152, + "step": 470 + }, + { + "epoch": 0.9978858350951374, + "grad_norm": 0.4838855564594269, + "learning_rate": 9.062293212791123e-06, + "loss": 1.2597992420196533, + "step": 472 + }, + { + "epoch": 1.0021141649048626, + "grad_norm": 1.116201639175415, + "learning_rate": 9.052657667500951e-06, + "loss": 0.9216406345367432, + "step": 474 + }, + { + "epoch": 1.0063424947145878, + "grad_norm": 1.7844558954238892, + "learning_rate": 9.042978689543487e-06, + "loss": 1.146926999092102, + "step": 476 + }, + { + "epoch": 1.0105708245243128, + "grad_norm": 2.6244568824768066, + "learning_rate": 9.033256397248124e-06, + "loss": 0.42595845460891724, + "step": 478 + }, + { + "epoch": 1.014799154334038, + "grad_norm": 6.244502067565918, + "learning_rate": 9.02349090947379e-06, + "loss": 0.7117490768432617, + "step": 480 + }, + { + "epoch": 1.0190274841437632, + "grad_norm": 1.4214531183242798, + "learning_rate": 9.013682345607502e-06, + "loss": 0.6555283069610596, + "step": 482 + }, + { + "epoch": 1.0232558139534884, + "grad_norm": 0.7880099415779114, + "learning_rate": 9.003830825562887e-06, + "loss": 1.034474492073059, + "step": 484 + }, + { + "epoch": 1.0274841437632136, + "grad_norm": 0.8250473737716675, + "learning_rate": 8.99393646977874e-06, + "loss": 1.057166576385498, + "step": 486 + }, + { + "epoch": 1.0317124735729386, + "grad_norm": 1.0640653371810913, + "learning_rate": 8.983999399217536e-06, + "loss": 0.8447081446647644, + "step": 488 + }, + { + "epoch": 1.0359408033826638, + "grad_norm": 2.257565975189209, + "learning_rate": 8.974019735363952e-06, + "loss": 0.778891384601593, + "step": 490 + }, + { + "epoch": 1.040169133192389, + "grad_norm": 1.4785358905792236, + "learning_rate": 8.963997600223387e-06, + "loss": 0.7628078460693359, + "step": 492 + }, + { + "epoch": 1.0443974630021142, + "grad_norm": 0.577157735824585, + "learning_rate": 8.953933116320474e-06, + "loss": 1.1460522413253784, + "step": 494 + }, + { + "epoch": 1.0486257928118394, + "grad_norm": 0.4649570882320404, + "learning_rate": 8.943826406697569e-06, + "loss": 0.9545917510986328, + "step": 496 + }, + { + "epoch": 1.0528541226215644, + "grad_norm": 0.729181170463562, + "learning_rate": 8.93367759491326e-06, + "loss": 0.7581544518470764, + "step": 498 + }, + { + "epoch": 1.0570824524312896, + "grad_norm": 0.5703451633453369, + "learning_rate": 8.923486805040845e-06, + "loss": 0.8621613383293152, + "step": 500 + }, + { + "epoch": 1.0613107822410148, + "grad_norm": 0.48987746238708496, + "learning_rate": 8.913254161666828e-06, + "loss": 1.1343525648117065, + "step": 502 + }, + { + "epoch": 1.06553911205074, + "grad_norm": 0.5839459300041199, + "learning_rate": 8.902979789889382e-06, + "loss": 1.0410970449447632, + "step": 504 + }, + { + "epoch": 1.069767441860465, + "grad_norm": 1.7165319919586182, + "learning_rate": 8.892663815316833e-06, + "loss": 0.8130745887756348, + "step": 506 + }, + { + "epoch": 1.0739957716701902, + "grad_norm": 0.5348541140556335, + "learning_rate": 8.882306364066114e-06, + "loss": 0.783044159412384, + "step": 508 + }, + { + "epoch": 1.0782241014799154, + "grad_norm": 1.005440354347229, + "learning_rate": 8.871907562761226e-06, + "loss": 0.9042244553565979, + "step": 510 + }, + { + "epoch": 1.0824524312896406, + "grad_norm": 0.7983127236366272, + "learning_rate": 8.861467538531692e-06, + "loss": 1.0373883247375488, + "step": 512 + }, + { + "epoch": 1.0866807610993658, + "grad_norm": 0.4987649917602539, + "learning_rate": 8.850986419011006e-06, + "loss": 1.104258418083191, + "step": 514 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 0.6088579893112183, + "learning_rate": 8.840464332335062e-06, + "loss": 0.8750536441802979, + "step": 516 + }, + { + "epoch": 1.095137420718816, + "grad_norm": 0.8371177315711975, + "learning_rate": 8.829901407140599e-06, + "loss": 0.8649636507034302, + "step": 518 + }, + { + "epoch": 1.0993657505285412, + "grad_norm": 0.6117671132087708, + "learning_rate": 8.81929777256362e-06, + "loss": 0.6066982746124268, + "step": 520 + }, + { + "epoch": 1.1035940803382664, + "grad_norm": 0.4654117822647095, + "learning_rate": 8.808653558237818e-06, + "loss": 1.2836765050888062, + "step": 522 + }, + { + "epoch": 1.1078224101479917, + "grad_norm": 1.2938016653060913, + "learning_rate": 8.797968894292992e-06, + "loss": 0.7479973435401917, + "step": 524 + }, + { + "epoch": 1.1120507399577166, + "grad_norm": 0.8367666602134705, + "learning_rate": 8.787243911353447e-06, + "loss": 1.3182156085968018, + "step": 526 + }, + { + "epoch": 1.1162790697674418, + "grad_norm": 0.41761136054992676, + "learning_rate": 8.776478740536409e-06, + "loss": 0.8329712152481079, + "step": 528 + }, + { + "epoch": 1.120507399577167, + "grad_norm": 3.907430648803711, + "learning_rate": 8.76567351345042e-06, + "loss": 0.4749341607093811, + "step": 530 + }, + { + "epoch": 1.1247357293868923, + "grad_norm": 0.5947129726409912, + "learning_rate": 8.754828362193719e-06, + "loss": 1.0100417137145996, + "step": 532 + }, + { + "epoch": 1.1289640591966172, + "grad_norm": 0.670958936214447, + "learning_rate": 8.743943419352638e-06, + "loss": 0.9090619683265686, + "step": 534 + }, + { + "epoch": 1.1331923890063424, + "grad_norm": 0.4388600289821625, + "learning_rate": 8.733018817999977e-06, + "loss": 0.8354282379150391, + "step": 536 + }, + { + "epoch": 1.1374207188160677, + "grad_norm": 1.8613966703414917, + "learning_rate": 8.722054691693374e-06, + "loss": 0.6932773590087891, + "step": 538 + }, + { + "epoch": 1.1416490486257929, + "grad_norm": 1.7173889875411987, + "learning_rate": 8.711051174473682e-06, + "loss": 0.730790376663208, + "step": 540 + }, + { + "epoch": 1.145877378435518, + "grad_norm": 0.5330702066421509, + "learning_rate": 8.700008400863316e-06, + "loss": 1.05195152759552, + "step": 542 + }, + { + "epoch": 1.150105708245243, + "grad_norm": 0.6400938630104065, + "learning_rate": 8.688926505864624e-06, + "loss": 0.734889566898346, + "step": 544 + }, + { + "epoch": 1.1543340380549683, + "grad_norm": 0.47678709030151367, + "learning_rate": 8.67780562495822e-06, + "loss": 1.1482857465744019, + "step": 546 + }, + { + "epoch": 1.1585623678646935, + "grad_norm": 1.6454354524612427, + "learning_rate": 8.666645894101346e-06, + "loss": 1.2353707551956177, + "step": 548 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 0.7140674591064453, + "learning_rate": 8.655447449726189e-06, + "loss": 0.7772641777992249, + "step": 550 + }, + { + "epoch": 1.1670190274841437, + "grad_norm": 0.6186874508857727, + "learning_rate": 8.644210428738237e-06, + "loss": 1.0311760902404785, + "step": 552 + }, + { + "epoch": 1.1712473572938689, + "grad_norm": 0.7900508046150208, + "learning_rate": 8.632934968514582e-06, + "loss": 1.1485145092010498, + "step": 554 + }, + { + "epoch": 1.175475687103594, + "grad_norm": 0.7557849884033203, + "learning_rate": 8.621621206902256e-06, + "loss": 1.0291169881820679, + "step": 556 + }, + { + "epoch": 1.1797040169133193, + "grad_norm": 0.5482131838798523, + "learning_rate": 8.61026928221654e-06, + "loss": 1.056830883026123, + "step": 558 + }, + { + "epoch": 1.1839323467230445, + "grad_norm": 0.5620715618133545, + "learning_rate": 8.598879333239274e-06, + "loss": 0.9988406300544739, + "step": 560 + }, + { + "epoch": 1.1881606765327695, + "grad_norm": 0.5934463739395142, + "learning_rate": 8.587451499217161e-06, + "loss": 0.6829384565353394, + "step": 562 + }, + { + "epoch": 1.1923890063424947, + "grad_norm": 0.6856687664985657, + "learning_rate": 8.575985919860065e-06, + "loss": 1.0374443531036377, + "step": 564 + }, + { + "epoch": 1.1966173361522199, + "grad_norm": 1.6381278038024902, + "learning_rate": 8.564482735339296e-06, + "loss": 0.866679310798645, + "step": 566 + }, + { + "epoch": 1.200845665961945, + "grad_norm": 0.7943741083145142, + "learning_rate": 8.552942086285909e-06, + "loss": 1.200244426727295, + "step": 568 + }, + { + "epoch": 1.20507399577167, + "grad_norm": 1.8203288316726685, + "learning_rate": 8.541364113788974e-06, + "loss": 1.166493535041809, + "step": 570 + }, + { + "epoch": 1.2093023255813953, + "grad_norm": 0.8953076601028442, + "learning_rate": 8.529748959393855e-06, + "loss": 1.0119144916534424, + "step": 572 + }, + { + "epoch": 1.2135306553911205, + "grad_norm": 0.6146345138549805, + "learning_rate": 8.518096765100485e-06, + "loss": 0.7799152135848999, + "step": 574 + }, + { + "epoch": 1.2177589852008457, + "grad_norm": 0.8124692440032959, + "learning_rate": 8.50640767336162e-06, + "loss": 0.7795466184616089, + "step": 576 + }, + { + "epoch": 1.221987315010571, + "grad_norm": 1.1723228693008423, + "learning_rate": 8.494681827081096e-06, + "loss": 0.9221915602684021, + "step": 578 + }, + { + "epoch": 1.226215644820296, + "grad_norm": 1.1240726709365845, + "learning_rate": 8.482919369612103e-06, + "loss": 0.7833749651908875, + "step": 580 + }, + { + "epoch": 1.230443974630021, + "grad_norm": 0.5680848956108093, + "learning_rate": 8.471120444755405e-06, + "loss": 1.0706194639205933, + "step": 582 + }, + { + "epoch": 1.2346723044397463, + "grad_norm": 1.5214192867279053, + "learning_rate": 8.4592851967576e-06, + "loss": 0.8856230974197388, + "step": 584 + }, + { + "epoch": 1.2389006342494715, + "grad_norm": 1.302047610282898, + "learning_rate": 8.447413770309349e-06, + "loss": 0.8525838851928711, + "step": 586 + }, + { + "epoch": 1.2431289640591967, + "grad_norm": 0.780870795249939, + "learning_rate": 8.435506310543607e-06, + "loss": 0.9184660911560059, + "step": 588 + }, + { + "epoch": 1.2473572938689217, + "grad_norm": 0.3887440264225006, + "learning_rate": 8.42356296303386e-06, + "loss": 0.626317024230957, + "step": 590 + }, + { + "epoch": 1.251585623678647, + "grad_norm": 0.3237205743789673, + "learning_rate": 8.411583873792325e-06, + "loss": 0.8544030785560608, + "step": 592 + }, + { + "epoch": 1.255813953488372, + "grad_norm": 1.5150576829910278, + "learning_rate": 8.399569189268182e-06, + "loss": 1.0616247653961182, + "step": 594 + }, + { + "epoch": 1.2600422832980973, + "grad_norm": 0.805270254611969, + "learning_rate": 8.387519056345778e-06, + "loss": 0.6954144835472107, + "step": 596 + }, + { + "epoch": 1.2642706131078225, + "grad_norm": 1.1619280576705933, + "learning_rate": 8.375433622342824e-06, + "loss": 1.1006245613098145, + "step": 598 + }, + { + "epoch": 1.2684989429175475, + "grad_norm": 1.2017593383789062, + "learning_rate": 8.363313035008613e-06, + "loss": 1.2584965229034424, + "step": 600 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 0.5600460767745972, + "learning_rate": 8.351157442522187e-06, + "loss": 0.9120579361915588, + "step": 602 + }, + { + "epoch": 1.276955602536998, + "grad_norm": 1.291706919670105, + "learning_rate": 8.338966993490554e-06, + "loss": 1.0825564861297607, + "step": 604 + }, + { + "epoch": 1.2811839323467231, + "grad_norm": 1.2752633094787598, + "learning_rate": 8.326741836946852e-06, + "loss": 0.7856518030166626, + "step": 606 + }, + { + "epoch": 1.285412262156448, + "grad_norm": 0.6531786322593689, + "learning_rate": 8.314482122348523e-06, + "loss": 1.089545726776123, + "step": 608 + }, + { + "epoch": 1.2896405919661733, + "grad_norm": 0.6609779596328735, + "learning_rate": 8.30218799957551e-06, + "loss": 0.7517077922821045, + "step": 610 + }, + { + "epoch": 1.2938689217758985, + "grad_norm": 0.8231602311134338, + "learning_rate": 8.289859618928404e-06, + "loss": 0.6350244283676147, + "step": 612 + }, + { + "epoch": 1.2980972515856237, + "grad_norm": 0.5657440423965454, + "learning_rate": 8.277497131126608e-06, + "loss": 0.9947375059127808, + "step": 614 + }, + { + "epoch": 1.302325581395349, + "grad_norm": 1.450269341468811, + "learning_rate": 8.265100687306503e-06, + "loss": 0.5063680410385132, + "step": 616 + }, + { + "epoch": 1.306553911205074, + "grad_norm": 2.1341493129730225, + "learning_rate": 8.252670439019595e-06, + "loss": 0.8950817584991455, + "step": 618 + }, + { + "epoch": 1.3107822410147991, + "grad_norm": 0.5161105394363403, + "learning_rate": 8.240206538230663e-06, + "loss": 1.1180918216705322, + "step": 620 + }, + { + "epoch": 1.3150105708245243, + "grad_norm": 0.567069411277771, + "learning_rate": 8.227709137315905e-06, + "loss": 1.0267692804336548, + "step": 622 + }, + { + "epoch": 1.3192389006342495, + "grad_norm": 6.294133186340332, + "learning_rate": 8.215178389061064e-06, + "loss": 0.6222748756408691, + "step": 624 + }, + { + "epoch": 1.3234672304439745, + "grad_norm": 0.5237641334533691, + "learning_rate": 8.202614446659572e-06, + "loss": 1.1326453685760498, + "step": 626 + }, + { + "epoch": 1.3276955602536997, + "grad_norm": 0.9689418077468872, + "learning_rate": 8.190017463710675e-06, + "loss": 1.197157382965088, + "step": 628 + }, + { + "epoch": 1.331923890063425, + "grad_norm": 0.9072657227516174, + "learning_rate": 8.177387594217548e-06, + "loss": 0.9704796075820923, + "step": 630 + }, + { + "epoch": 1.3361522198731501, + "grad_norm": 1.4275466203689575, + "learning_rate": 8.164724992585421e-06, + "loss": 0.5456193089485168, + "step": 632 + }, + { + "epoch": 1.3403805496828753, + "grad_norm": 0.8675680756568909, + "learning_rate": 8.152029813619685e-06, + "loss": 0.6797899603843689, + "step": 634 + }, + { + "epoch": 1.3446088794926006, + "grad_norm": 0.457328200340271, + "learning_rate": 8.139302212524006e-06, + "loss": 0.7478649020195007, + "step": 636 + }, + { + "epoch": 1.3488372093023255, + "grad_norm": 2.034100294113159, + "learning_rate": 8.126542344898419e-06, + "loss": 1.0299596786499023, + "step": 638 + }, + { + "epoch": 1.3530655391120507, + "grad_norm": 0.8121411204338074, + "learning_rate": 8.113750366737431e-06, + "loss": 0.7044599056243896, + "step": 640 + }, + { + "epoch": 1.357293868921776, + "grad_norm": 0.44053447246551514, + "learning_rate": 8.100926434428116e-06, + "loss": 1.1919891834259033, + "step": 642 + }, + { + "epoch": 1.361522198731501, + "grad_norm": 1.0241997241973877, + "learning_rate": 8.0880707047482e-06, + "loss": 1.0898381471633911, + "step": 644 + }, + { + "epoch": 1.3657505285412261, + "grad_norm": 3.370863676071167, + "learning_rate": 8.075183334864139e-06, + "loss": 0.7127438187599182, + "step": 646 + }, + { + "epoch": 1.3699788583509513, + "grad_norm": 1.7245151996612549, + "learning_rate": 8.062264482329209e-06, + "loss": 0.6995323300361633, + "step": 648 + }, + { + "epoch": 1.3742071881606766, + "grad_norm": 2.1740646362304688, + "learning_rate": 8.049314305081575e-06, + "loss": 0.7619326710700989, + "step": 650 + }, + { + "epoch": 1.3784355179704018, + "grad_norm": 0.6445834040641785, + "learning_rate": 8.036332961442355e-06, + "loss": 0.7136253118515015, + "step": 652 + }, + { + "epoch": 1.382663847780127, + "grad_norm": 0.5504980683326721, + "learning_rate": 8.023320610113691e-06, + "loss": 1.1639782190322876, + "step": 654 + }, + { + "epoch": 1.386892177589852, + "grad_norm": 0.5555628538131714, + "learning_rate": 8.010277410176801e-06, + "loss": 1.018182635307312, + "step": 656 + }, + { + "epoch": 1.3911205073995772, + "grad_norm": 2.1361212730407715, + "learning_rate": 7.997203521090053e-06, + "loss": 0.8057630658149719, + "step": 658 + }, + { + "epoch": 1.3953488372093024, + "grad_norm": 0.5145853161811829, + "learning_rate": 7.984099102686986e-06, + "loss": 0.9387868642807007, + "step": 660 + }, + { + "epoch": 1.3995771670190273, + "grad_norm": 0.4842035472393036, + "learning_rate": 7.970964315174384e-06, + "loss": 1.1107807159423828, + "step": 662 + }, + { + "epoch": 1.4038054968287526, + "grad_norm": 0.16610600054264069, + "learning_rate": 7.9577993191303e-06, + "loss": 0.9138076305389404, + "step": 664 + }, + { + "epoch": 1.4080338266384778, + "grad_norm": 2.3093957901000977, + "learning_rate": 7.944604275502102e-06, + "loss": 1.401366949081421, + "step": 666 + }, + { + "epoch": 1.412262156448203, + "grad_norm": 2.1093337535858154, + "learning_rate": 7.931379345604495e-06, + "loss": 1.0819278955459595, + "step": 668 + }, + { + "epoch": 1.4164904862579282, + "grad_norm": 0.6838366389274597, + "learning_rate": 7.918124691117563e-06, + "loss": 0.6194279193878174, + "step": 670 + }, + { + "epoch": 1.4207188160676534, + "grad_norm": 0.44672903418540955, + "learning_rate": 7.904840474084783e-06, + "loss": 1.051002025604248, + "step": 672 + }, + { + "epoch": 1.4249471458773784, + "grad_norm": 0.4658559560775757, + "learning_rate": 7.891526856911046e-06, + "loss": 0.5866225361824036, + "step": 674 + }, + { + "epoch": 1.4291754756871036, + "grad_norm": 1.0386725664138794, + "learning_rate": 7.878184002360664e-06, + "loss": 0.7329708933830261, + "step": 676 + }, + { + "epoch": 1.4334038054968288, + "grad_norm": 0.5810735821723938, + "learning_rate": 7.864812073555402e-06, + "loss": 1.32469642162323, + "step": 678 + }, + { + "epoch": 1.437632135306554, + "grad_norm": 1.0721173286437988, + "learning_rate": 7.851411233972461e-06, + "loss": 1.163739562034607, + "step": 680 + }, + { + "epoch": 1.441860465116279, + "grad_norm": 1.159287929534912, + "learning_rate": 7.837981647442486e-06, + "loss": 1.1509932279586792, + "step": 682 + }, + { + "epoch": 1.4460887949260042, + "grad_norm": 0.7381364703178406, + "learning_rate": 7.82452347814757e-06, + "loss": 1.1009894609451294, + "step": 684 + }, + { + "epoch": 1.4503171247357294, + "grad_norm": 0.9864975214004517, + "learning_rate": 7.811036890619239e-06, + "loss": 1.373154640197754, + "step": 686 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 0.509519636631012, + "learning_rate": 7.797522049736441e-06, + "loss": 1.028425931930542, + "step": 688 + }, + { + "epoch": 1.4587737843551798, + "grad_norm": 1.9247252941131592, + "learning_rate": 7.78397912072354e-06, + "loss": 0.391171395778656, + "step": 690 + }, + { + "epoch": 1.463002114164905, + "grad_norm": 0.44884008169174194, + "learning_rate": 7.770408269148281e-06, + "loss": 1.1188896894454956, + "step": 692 + }, + { + "epoch": 1.46723044397463, + "grad_norm": 5.2882184982299805, + "learning_rate": 7.756809660919778e-06, + "loss": 0.5647270083427429, + "step": 694 + }, + { + "epoch": 1.4714587737843552, + "grad_norm": 2.669262647628784, + "learning_rate": 7.74318346228648e-06, + "loss": 1.050774335861206, + "step": 696 + }, + { + "epoch": 1.4756871035940804, + "grad_norm": 0.45423516631126404, + "learning_rate": 7.72952983983414e-06, + "loss": 1.0962177515029907, + "step": 698 + }, + { + "epoch": 1.4799154334038054, + "grad_norm": 2.6267011165618896, + "learning_rate": 7.715848960483776e-06, + "loss": 0.980463981628418, + "step": 700 + }, + { + "epoch": 1.4841437632135306, + "grad_norm": 0.4617670476436615, + "learning_rate": 7.702140991489637e-06, + "loss": 1.0503666400909424, + "step": 702 + }, + { + "epoch": 1.4883720930232558, + "grad_norm": 0.4600563943386078, + "learning_rate": 7.688406100437149e-06, + "loss": 1.0178788900375366, + "step": 704 + }, + { + "epoch": 1.492600422832981, + "grad_norm": 0.5003790855407715, + "learning_rate": 7.674644455240878e-06, + "loss": 0.7319499850273132, + "step": 706 + }, + { + "epoch": 1.4968287526427062, + "grad_norm": 0.33892884850502014, + "learning_rate": 7.660856224142456e-06, + "loss": 0.22760850191116333, + "step": 708 + }, + { + "epoch": 1.5010570824524314, + "grad_norm": 0.7289655804634094, + "learning_rate": 7.647041575708555e-06, + "loss": 0.6383700370788574, + "step": 710 + }, + { + "epoch": 1.5052854122621564, + "grad_norm": 2.811466693878174, + "learning_rate": 7.633200678828804e-06, + "loss": 0.5055187940597534, + "step": 712 + }, + { + "epoch": 1.5095137420718816, + "grad_norm": 0.5600261092185974, + "learning_rate": 7.6193337027137215e-06, + "loss": 0.8390191197395325, + "step": 714 + }, + { + "epoch": 1.5137420718816068, + "grad_norm": 0.4209119379520416, + "learning_rate": 7.605440816892667e-06, + "loss": 1.0698416233062744, + "step": 716 + }, + { + "epoch": 1.5179704016913318, + "grad_norm": 0.5382829904556274, + "learning_rate": 7.591522191211746e-06, + "loss": 0.9513751864433289, + "step": 718 + }, + { + "epoch": 1.522198731501057, + "grad_norm": 0.21538308262825012, + "learning_rate": 7.577577995831755e-06, + "loss": 0.8911153674125671, + "step": 720 + }, + { + "epoch": 1.5264270613107822, + "grad_norm": 1.1509976387023926, + "learning_rate": 7.563608401226081e-06, + "loss": 0.8212535977363586, + "step": 722 + }, + { + "epoch": 1.5306553911205074, + "grad_norm": 0.39896026253700256, + "learning_rate": 7.549613578178632e-06, + "loss": 0.8824007511138916, + "step": 724 + }, + { + "epoch": 1.5348837209302326, + "grad_norm": 0.8111864924430847, + "learning_rate": 7.535593697781741e-06, + "loss": 0.8548102974891663, + "step": 726 + }, + { + "epoch": 1.5391120507399578, + "grad_norm": 0.7156527042388916, + "learning_rate": 7.52154893143408e-06, + "loss": 0.6663705706596375, + "step": 728 + }, + { + "epoch": 1.543340380549683, + "grad_norm": 0.8205687999725342, + "learning_rate": 7.50747945083856e-06, + "loss": 0.8241264224052429, + "step": 730 + }, + { + "epoch": 1.547568710359408, + "grad_norm": 0.43496695160865784, + "learning_rate": 7.493385428000234e-06, + "loss": 1.0368667840957642, + "step": 732 + }, + { + "epoch": 1.5517970401691332, + "grad_norm": 0.6925005316734314, + "learning_rate": 7.479267035224191e-06, + "loss": 0.33902376890182495, + "step": 734 + }, + { + "epoch": 1.5560253699788582, + "grad_norm": 0.6592570543289185, + "learning_rate": 7.465124445113459e-06, + "loss": 0.971444308757782, + "step": 736 + }, + { + "epoch": 1.5602536997885834, + "grad_norm": 0.26948875188827515, + "learning_rate": 7.450957830566878e-06, + "loss": 0.7512494325637817, + "step": 738 + }, + { + "epoch": 1.5644820295983086, + "grad_norm": 0.6098513007164001, + "learning_rate": 7.436767364777009e-06, + "loss": 1.140318751335144, + "step": 740 + }, + { + "epoch": 1.5687103594080338, + "grad_norm": 0.9478453993797302, + "learning_rate": 7.422553221227993e-06, + "loss": 0.7520893812179565, + "step": 742 + }, + { + "epoch": 1.572938689217759, + "grad_norm": 0.7493337392807007, + "learning_rate": 7.4083155736934484e-06, + "loss": 1.0719984769821167, + "step": 744 + }, + { + "epoch": 1.5771670190274842, + "grad_norm": 0.8848110437393188, + "learning_rate": 7.394054596234336e-06, + "loss": 1.1771358251571655, + "step": 746 + }, + { + "epoch": 1.5813953488372094, + "grad_norm": 0.6681188941001892, + "learning_rate": 7.379770463196831e-06, + "loss": 0.9307011365890503, + "step": 748 + }, + { + "epoch": 1.5856236786469344, + "grad_norm": 0.4746161997318268, + "learning_rate": 7.365463349210203e-06, + "loss": 0.06105047091841698, + "step": 750 + }, + { + "epoch": 1.5898520084566596, + "grad_norm": 0.6364957094192505, + "learning_rate": 7.351133429184666e-06, + "loss": 0.8398118019104004, + "step": 752 + }, + { + "epoch": 1.5940803382663846, + "grad_norm": 0.946345865726471, + "learning_rate": 7.336780878309251e-06, + "loss": 1.1541061401367188, + "step": 754 + }, + { + "epoch": 1.5983086680761098, + "grad_norm": 0.6361474394798279, + "learning_rate": 7.322405872049657e-06, + "loss": 1.108107328414917, + "step": 756 + }, + { + "epoch": 1.602536997885835, + "grad_norm": 0.4905734956264496, + "learning_rate": 7.308008586146111e-06, + "loss": 1.1289057731628418, + "step": 758 + }, + { + "epoch": 1.6067653276955602, + "grad_norm": 0.7073136568069458, + "learning_rate": 7.293589196611221e-06, + "loss": 1.1330186128616333, + "step": 760 + }, + { + "epoch": 1.6109936575052854, + "grad_norm": 0.6589964032173157, + "learning_rate": 7.279147879727811e-06, + "loss": 1.1067560911178589, + "step": 762 + }, + { + "epoch": 1.6152219873150107, + "grad_norm": 0.8479374051094055, + "learning_rate": 7.264684812046787e-06, + "loss": 1.0230149030685425, + "step": 764 + }, + { + "epoch": 1.6194503171247359, + "grad_norm": 1.6476147174835205, + "learning_rate": 7.250200170384961e-06, + "loss": 0.8196643590927124, + "step": 766 + }, + { + "epoch": 1.6236786469344608, + "grad_norm": 0.46958765387535095, + "learning_rate": 7.235694131822894e-06, + "loss": 1.0976183414459229, + "step": 768 + }, + { + "epoch": 1.627906976744186, + "grad_norm": 0.7259538769721985, + "learning_rate": 7.221166873702738e-06, + "loss": 1.053260326385498, + "step": 770 + }, + { + "epoch": 1.6321353065539113, + "grad_norm": 0.30915775895118713, + "learning_rate": 7.206618573626058e-06, + "loss": 1.0872668027877808, + "step": 772 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 0.392713725566864, + "learning_rate": 7.19204940945167e-06, + "loss": 0.8378170132637024, + "step": 774 + }, + { + "epoch": 1.6405919661733614, + "grad_norm": 1.2388817071914673, + "learning_rate": 7.177459559293456e-06, + "loss": 1.107264757156372, + "step": 776 + }, + { + "epoch": 1.6448202959830867, + "grad_norm": 0.6137820482254028, + "learning_rate": 7.162849201518198e-06, + "loss": 0.6787929534912109, + "step": 778 + }, + { + "epoch": 1.6490486257928119, + "grad_norm": 2.9237804412841797, + "learning_rate": 7.1482185147433905e-06, + "loss": 1.1160000562667847, + "step": 780 + }, + { + "epoch": 1.653276955602537, + "grad_norm": 0.4147196412086487, + "learning_rate": 7.133567677835055e-06, + "loss": 0.808272123336792, + "step": 782 + }, + { + "epoch": 1.6575052854122623, + "grad_norm": 0.6064695119857788, + "learning_rate": 7.11889686990556e-06, + "loss": 1.194222331047058, + "step": 784 + }, + { + "epoch": 1.6617336152219875, + "grad_norm": 1.2987606525421143, + "learning_rate": 7.104206270311423e-06, + "loss": 0.45863011479377747, + "step": 786 + }, + { + "epoch": 1.6659619450317125, + "grad_norm": 0.5215296745300293, + "learning_rate": 7.089496058651133e-06, + "loss": 1.1097348928451538, + "step": 788 + }, + { + "epoch": 1.6701902748414377, + "grad_norm": 0.6435741782188416, + "learning_rate": 7.0747664147629335e-06, + "loss": 0.7349862456321716, + "step": 790 + }, + { + "epoch": 1.6744186046511627, + "grad_norm": 0.3204418122768402, + "learning_rate": 7.060017518722637e-06, + "loss": 0.4114014208316803, + "step": 792 + }, + { + "epoch": 1.6786469344608879, + "grad_norm": 0.6571648120880127, + "learning_rate": 7.0452495508414276e-06, + "loss": 1.0704959630966187, + "step": 794 + }, + { + "epoch": 1.682875264270613, + "grad_norm": 0.4121459722518921, + "learning_rate": 7.030462691663641e-06, + "loss": 0.6129454970359802, + "step": 796 + }, + { + "epoch": 1.6871035940803383, + "grad_norm": 0.8095541000366211, + "learning_rate": 7.015657121964575e-06, + "loss": 1.1843053102493286, + "step": 798 + }, + { + "epoch": 1.6913319238900635, + "grad_norm": 0.4884234666824341, + "learning_rate": 7.000833022748267e-06, + "loss": 1.0753151178359985, + "step": 800 + }, + { + "epoch": 1.6955602536997887, + "grad_norm": 1.4372299909591675, + "learning_rate": 6.985990575245283e-06, + "loss": 0.5634129047393799, + "step": 802 + }, + { + "epoch": 1.699788583509514, + "grad_norm": 0.6573989391326904, + "learning_rate": 6.97112996091051e-06, + "loss": 1.0538238286972046, + "step": 804 + }, + { + "epoch": 1.7040169133192389, + "grad_norm": 0.8409382104873657, + "learning_rate": 6.956251361420926e-06, + "loss": 0.8009997606277466, + "step": 806 + }, + { + "epoch": 1.708245243128964, + "grad_norm": 0.7150518298149109, + "learning_rate": 6.94135495867339e-06, + "loss": 0.6996553540229797, + "step": 808 + }, + { + "epoch": 1.712473572938689, + "grad_norm": 0.8264674544334412, + "learning_rate": 6.926440934782407e-06, + "loss": 1.0933129787445068, + "step": 810 + }, + { + "epoch": 1.7167019027484143, + "grad_norm": 1.176154613494873, + "learning_rate": 6.9115094720779155e-06, + "loss": 1.291683316230774, + "step": 812 + }, + { + "epoch": 1.7209302325581395, + "grad_norm": 0.8074145317077637, + "learning_rate": 6.896560753103041e-06, + "loss": 1.0416253805160522, + "step": 814 + }, + { + "epoch": 1.7251585623678647, + "grad_norm": 0.7769083976745605, + "learning_rate": 6.881594960611881e-06, + "loss": 1.125274896621704, + "step": 816 + }, + { + "epoch": 1.72938689217759, + "grad_norm": 0.5327866673469543, + "learning_rate": 6.866612277567261e-06, + "loss": 0.9173880815505981, + "step": 818 + }, + { + "epoch": 1.733615221987315, + "grad_norm": 2.439627170562744, + "learning_rate": 6.851612887138501e-06, + "loss": 1.015201210975647, + "step": 820 + }, + { + "epoch": 1.7378435517970403, + "grad_norm": 0.5043546557426453, + "learning_rate": 6.836596972699175e-06, + "loss": 0.6561012268066406, + "step": 822 + }, + { + "epoch": 1.7420718816067653, + "grad_norm": 0.259673535823822, + "learning_rate": 6.82156471782487e-06, + "loss": 0.6422268152236938, + "step": 824 + }, + { + "epoch": 1.7463002114164905, + "grad_norm": 0.7639176845550537, + "learning_rate": 6.806516306290942e-06, + "loss": 0.8537571430206299, + "step": 826 + }, + { + "epoch": 1.7505285412262155, + "grad_norm": 0.3980555236339569, + "learning_rate": 6.791451922070265e-06, + "loss": 0.9388514757156372, + "step": 828 + }, + { + "epoch": 1.7547568710359407, + "grad_norm": 2.035597562789917, + "learning_rate": 6.77637174933099e-06, + "loss": 0.7623142004013062, + "step": 830 + }, + { + "epoch": 1.758985200845666, + "grad_norm": 0.5751479268074036, + "learning_rate": 6.7612759724342895e-06, + "loss": 1.1322904825210571, + "step": 832 + }, + { + "epoch": 1.763213530655391, + "grad_norm": 0.9078547358512878, + "learning_rate": 6.746164775932096e-06, + "loss": 0.7828465104103088, + "step": 834 + }, + { + "epoch": 1.7674418604651163, + "grad_norm": 0.4219997227191925, + "learning_rate": 6.731038344564862e-06, + "loss": 0.7831478714942932, + "step": 836 + }, + { + "epoch": 1.7716701902748415, + "grad_norm": 0.5556386709213257, + "learning_rate": 6.715896863259286e-06, + "loss": 1.1161986589431763, + "step": 838 + }, + { + "epoch": 1.7758985200845667, + "grad_norm": 0.5635927319526672, + "learning_rate": 6.700740517126061e-06, + "loss": 0.83119136095047, + "step": 840 + }, + { + "epoch": 1.7801268498942917, + "grad_norm": 1.3156684637069702, + "learning_rate": 6.685569491457608e-06, + "loss": 0.37486332654953003, + "step": 842 + }, + { + "epoch": 1.784355179704017, + "grad_norm": 0.8130014538764954, + "learning_rate": 6.670383971725809e-06, + "loss": 0.7733681797981262, + "step": 844 + }, + { + "epoch": 1.7885835095137421, + "grad_norm": 0.8940642476081848, + "learning_rate": 6.655184143579746e-06, + "loss": 0.9664357900619507, + "step": 846 + }, + { + "epoch": 1.792811839323467, + "grad_norm": 0.5689818263053894, + "learning_rate": 6.639970192843423e-06, + "loss": 0.7395713329315186, + "step": 848 + }, + { + "epoch": 1.7970401691331923, + "grad_norm": 0.8605613708496094, + "learning_rate": 6.624742305513502e-06, + "loss": 0.7102298140525818, + "step": 850 + }, + { + "epoch": 1.8012684989429175, + "grad_norm": 0.5446686744689941, + "learning_rate": 6.609500667757022e-06, + "loss": 0.9291532039642334, + "step": 852 + }, + { + "epoch": 1.8054968287526427, + "grad_norm": 0.7106825709342957, + "learning_rate": 6.594245465909127e-06, + "loss": 0.8192090392112732, + "step": 854 + }, + { + "epoch": 1.809725158562368, + "grad_norm": 0.6058146357536316, + "learning_rate": 6.578976886470791e-06, + "loss": 1.069812536239624, + "step": 856 + }, + { + "epoch": 1.8139534883720931, + "grad_norm": 0.5561621785163879, + "learning_rate": 6.56369511610653e-06, + "loss": 0.6693546175956726, + "step": 858 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 0.5408707857131958, + "learning_rate": 6.548400341642128e-06, + "loss": 0.998047411441803, + "step": 860 + }, + { + "epoch": 1.8224101479915433, + "grad_norm": 0.3591640889644623, + "learning_rate": 6.533092750062345e-06, + "loss": 1.049836277961731, + "step": 862 + }, + { + "epoch": 1.8266384778012685, + "grad_norm": 1.5112040042877197, + "learning_rate": 6.5177725285086416e-06, + "loss": 0.4735381603240967, + "step": 864 + }, + { + "epoch": 1.8308668076109935, + "grad_norm": 0.6427939534187317, + "learning_rate": 6.5024398642768785e-06, + "loss": 0.7486822605133057, + "step": 866 + }, + { + "epoch": 1.8350951374207187, + "grad_norm": 0.6862894296646118, + "learning_rate": 6.487094944815036e-06, + "loss": 1.083625078201294, + "step": 868 + }, + { + "epoch": 1.839323467230444, + "grad_norm": 0.45636817812919617, + "learning_rate": 6.47173795772092e-06, + "loss": 1.10338294506073, + "step": 870 + }, + { + "epoch": 1.8435517970401691, + "grad_norm": 0.21469919383525848, + "learning_rate": 6.456369090739869e-06, + "loss": 0.7604748606681824, + "step": 872 + }, + { + "epoch": 1.8477801268498943, + "grad_norm": 0.5490983724594116, + "learning_rate": 6.440988531762454e-06, + "loss": 1.0832877159118652, + "step": 874 + }, + { + "epoch": 1.8520084566596196, + "grad_norm": 4.500369071960449, + "learning_rate": 6.42559646882219e-06, + "loss": 0.8788917660713196, + "step": 876 + }, + { + "epoch": 1.8562367864693448, + "grad_norm": 1.3689038753509521, + "learning_rate": 6.410193090093227e-06, + "loss": 0.2849124073982239, + "step": 878 + }, + { + "epoch": 1.8604651162790697, + "grad_norm": 1.2711057662963867, + "learning_rate": 6.394778583888064e-06, + "loss": 0.6833288073539734, + "step": 880 + }, + { + "epoch": 1.864693446088795, + "grad_norm": 0.43423914909362793, + "learning_rate": 6.379353138655227e-06, + "loss": 1.08613121509552, + "step": 882 + }, + { + "epoch": 1.86892177589852, + "grad_norm": 0.2996130585670471, + "learning_rate": 6.3639169429769805e-06, + "loss": 0.7295659184455872, + "step": 884 + }, + { + "epoch": 1.8731501057082451, + "grad_norm": 3.8660871982574463, + "learning_rate": 6.348470185567025e-06, + "loss": 1.1010972261428833, + "step": 886 + }, + { + "epoch": 1.8773784355179703, + "grad_norm": 0.6088366508483887, + "learning_rate": 6.333013055268169e-06, + "loss": 1.1118422746658325, + "step": 888 + }, + { + "epoch": 1.8816067653276956, + "grad_norm": 0.6090267300605774, + "learning_rate": 6.317545741050044e-06, + "loss": 0.6268060207366943, + "step": 890 + }, + { + "epoch": 1.8858350951374208, + "grad_norm": 0.5162058472633362, + "learning_rate": 6.3020684320067805e-06, + "loss": 0.8153216242790222, + "step": 892 + }, + { + "epoch": 1.890063424947146, + "grad_norm": 3.825115919113159, + "learning_rate": 6.2865813173546976e-06, + "loss": 1.1050728559494019, + "step": 894 + }, + { + "epoch": 1.8942917547568712, + "grad_norm": 2.846707344055176, + "learning_rate": 6.271084586429994e-06, + "loss": 1.3872066736221313, + "step": 896 + }, + { + "epoch": 1.8985200845665962, + "grad_norm": 0.445870041847229, + "learning_rate": 6.255578428686433e-06, + "loss": 0.9879752993583679, + "step": 898 + }, + { + "epoch": 1.9027484143763214, + "grad_norm": 0.8437865376472473, + "learning_rate": 6.240063033693023e-06, + "loss": 0.47253939509391785, + "step": 900 + }, + { + "epoch": 1.9069767441860463, + "grad_norm": 0.5850438475608826, + "learning_rate": 6.2245385911316994e-06, + "loss": 1.0632604360580444, + "step": 902 + }, + { + "epoch": 1.9112050739957716, + "grad_norm": 1.3044452667236328, + "learning_rate": 6.2090052907950096e-06, + "loss": 0.5071629881858826, + "step": 904 + }, + { + "epoch": 1.9154334038054968, + "grad_norm": 0.466347336769104, + "learning_rate": 6.193463322583792e-06, + "loss": 0.8526642918586731, + "step": 906 + }, + { + "epoch": 1.919661733615222, + "grad_norm": 1.3794184923171997, + "learning_rate": 6.177912876504851e-06, + "loss": 1.1799172163009644, + "step": 908 + }, + { + "epoch": 1.9238900634249472, + "grad_norm": 0.4012717306613922, + "learning_rate": 6.16235414266864e-06, + "loss": 0.7791982293128967, + "step": 910 + }, + { + "epoch": 1.9281183932346724, + "grad_norm": 1.2676597833633423, + "learning_rate": 6.146787311286927e-06, + "loss": 1.0873719453811646, + "step": 912 + }, + { + "epoch": 1.9323467230443976, + "grad_norm": 0.5995278358459473, + "learning_rate": 6.131212572670483e-06, + "loss": 0.7078754901885986, + "step": 914 + }, + { + "epoch": 1.9365750528541226, + "grad_norm": 0.41933539509773254, + "learning_rate": 6.115630117226741e-06, + "loss": 1.125902533531189, + "step": 916 + }, + { + "epoch": 1.9408033826638478, + "grad_norm": 0.35379597544670105, + "learning_rate": 6.1000401354574855e-06, + "loss": 0.6423243284225464, + "step": 918 + }, + { + "epoch": 1.945031712473573, + "grad_norm": 0.6899601817131042, + "learning_rate": 6.084442817956502e-06, + "loss": 0.899864673614502, + "step": 920 + }, + { + "epoch": 1.949260042283298, + "grad_norm": 0.4957478940486908, + "learning_rate": 6.0688383554072625e-06, + "loss": 0.9267314076423645, + "step": 922 + }, + { + "epoch": 1.9534883720930232, + "grad_norm": 0.6722287535667419, + "learning_rate": 6.053226938580596e-06, + "loss": 1.0104131698608398, + "step": 924 + }, + { + "epoch": 1.9577167019027484, + "grad_norm": 0.23449595272541046, + "learning_rate": 6.03760875833234e-06, + "loss": 0.8867968320846558, + "step": 926 + }, + { + "epoch": 1.9619450317124736, + "grad_norm": 0.9708488583564758, + "learning_rate": 6.021984005601027e-06, + "loss": 1.233751654624939, + "step": 928 + }, + { + "epoch": 1.9661733615221988, + "grad_norm": 0.67732834815979, + "learning_rate": 6.006352871405534e-06, + "loss": 0.7823659181594849, + "step": 930 + }, + { + "epoch": 1.970401691331924, + "grad_norm": 0.5030365586280823, + "learning_rate": 5.990715546842755e-06, + "loss": 1.223224401473999, + "step": 932 + }, + { + "epoch": 1.9746300211416492, + "grad_norm": 0.9906087517738342, + "learning_rate": 5.9750722230852694e-06, + "loss": 0.8600039482116699, + "step": 934 + }, + { + "epoch": 1.9788583509513742, + "grad_norm": 0.7407654523849487, + "learning_rate": 5.959423091378992e-06, + "loss": 0.7119872570037842, + "step": 936 + }, + { + "epoch": 1.9830866807610994, + "grad_norm": 0.6591317653656006, + "learning_rate": 5.943768343040845e-06, + "loss": 0.8544420003890991, + "step": 938 + }, + { + "epoch": 1.9873150105708244, + "grad_norm": 0.43776705861091614, + "learning_rate": 5.928108169456419e-06, + "loss": 1.0439858436584473, + "step": 940 + }, + { + "epoch": 1.9915433403805496, + "grad_norm": 0.3629160523414612, + "learning_rate": 5.912442762077624e-06, + "loss": 1.064499020576477, + "step": 942 + }, + { + "epoch": 1.9957716701902748, + "grad_norm": 0.5995646119117737, + "learning_rate": 5.89677231242036e-06, + "loss": 1.0699841976165771, + "step": 944 + }, + { + "epoch": 2.0, + "grad_norm": 0.8705946207046509, + "learning_rate": 5.8810970120621714e-06, + "loss": 0.4486532509326935, + "step": 946 + }, + { + "epoch": 2.004228329809725, + "grad_norm": 0.7225093245506287, + "learning_rate": 5.8654170526399014e-06, + "loss": 0.6647614240646362, + "step": 948 + }, + { + "epoch": 2.0084566596194504, + "grad_norm": 0.45483696460723877, + "learning_rate": 5.849732625847354e-06, + "loss": 0.9562214016914368, + "step": 950 + }, + { + "epoch": 2.0126849894291756, + "grad_norm": 1.091545820236206, + "learning_rate": 5.834043923432947e-06, + "loss": 0.8385214805603027, + "step": 952 + }, + { + "epoch": 2.016913319238901, + "grad_norm": 0.39615076780319214, + "learning_rate": 5.818351137197373e-06, + "loss": 0.9980201125144958, + "step": 954 + }, + { + "epoch": 2.0211416490486256, + "grad_norm": 1.4876010417938232, + "learning_rate": 5.802654458991245e-06, + "loss": 1.0496366024017334, + "step": 956 + }, + { + "epoch": 2.025369978858351, + "grad_norm": 0.5281734466552734, + "learning_rate": 5.786954080712763e-06, + "loss": 0.6850022077560425, + "step": 958 + }, + { + "epoch": 2.029598308668076, + "grad_norm": 0.5857702493667603, + "learning_rate": 5.7712501943053576e-06, + "loss": 0.9232272505760193, + "step": 960 + }, + { + "epoch": 2.033826638477801, + "grad_norm": 0.9430474638938904, + "learning_rate": 5.755542991755352e-06, + "loss": 0.8339846134185791, + "step": 962 + }, + { + "epoch": 2.0380549682875264, + "grad_norm": 0.9053657054901123, + "learning_rate": 5.739832665089604e-06, + "loss": 0.7434718608856201, + "step": 964 + }, + { + "epoch": 2.0422832980972516, + "grad_norm": 0.48105019330978394, + "learning_rate": 5.724119406373171e-06, + "loss": 0.6355252861976624, + "step": 966 + }, + { + "epoch": 2.046511627906977, + "grad_norm": 1.1304575204849243, + "learning_rate": 5.708403407706954e-06, + "loss": 0.7226954698562622, + "step": 968 + }, + { + "epoch": 2.050739957716702, + "grad_norm": 0.5827345252037048, + "learning_rate": 5.692684861225349e-06, + "loss": 0.9803634285926819, + "step": 970 + }, + { + "epoch": 2.0549682875264272, + "grad_norm": 0.9064778089523315, + "learning_rate": 5.676963959093904e-06, + "loss": 0.4202920198440552, + "step": 972 + }, + { + "epoch": 2.059196617336152, + "grad_norm": 0.489486426115036, + "learning_rate": 5.66124089350696e-06, + "loss": 0.9203523993492126, + "step": 974 + }, + { + "epoch": 2.063424947145877, + "grad_norm": 0.5507452487945557, + "learning_rate": 5.645515856685316e-06, + "loss": 0.9069164395332336, + "step": 976 + }, + { + "epoch": 2.0676532769556024, + "grad_norm": 0.46944448351860046, + "learning_rate": 5.629789040873862e-06, + "loss": 0.9797834157943726, + "step": 978 + }, + { + "epoch": 2.0718816067653276, + "grad_norm": 7.794976711273193, + "learning_rate": 5.614060638339238e-06, + "loss": 0.6233423352241516, + "step": 980 + }, + { + "epoch": 2.076109936575053, + "grad_norm": 0.9501297473907471, + "learning_rate": 5.598330841367487e-06, + "loss": 0.6793166995048523, + "step": 982 + }, + { + "epoch": 2.080338266384778, + "grad_norm": 0.5520645380020142, + "learning_rate": 5.582599842261693e-06, + "loss": 0.6733763813972473, + "step": 984 + }, + { + "epoch": 2.0845665961945032, + "grad_norm": 0.6170113682746887, + "learning_rate": 5.566867833339644e-06, + "loss": 0.814887285232544, + "step": 986 + }, + { + "epoch": 2.0887949260042284, + "grad_norm": 0.4768456518650055, + "learning_rate": 5.551135006931467e-06, + "loss": 0.9522485136985779, + "step": 988 + }, + { + "epoch": 2.0930232558139537, + "grad_norm": 0.7555639147758484, + "learning_rate": 5.535401555377287e-06, + "loss": 0.8823084831237793, + "step": 990 + }, + { + "epoch": 2.097251585623679, + "grad_norm": 2.535158395767212, + "learning_rate": 5.51966767102487e-06, + "loss": 0.5443646311759949, + "step": 992 + }, + { + "epoch": 2.1014799154334036, + "grad_norm": 2.706549644470215, + "learning_rate": 5.503933546227273e-06, + "loss": 0.6737141013145447, + "step": 994 + }, + { + "epoch": 2.105708245243129, + "grad_norm": 0.4050784409046173, + "learning_rate": 5.4881993733404916e-06, + "loss": 0.6984460353851318, + "step": 996 + }, + { + "epoch": 2.109936575052854, + "grad_norm": 0.5158626437187195, + "learning_rate": 5.472465344721113e-06, + "loss": 0.8758757710456848, + "step": 998 + }, + { + "epoch": 2.1141649048625792, + "grad_norm": 0.486257404088974, + "learning_rate": 5.456731652723959e-06, + "loss": 0.9399458765983582, + "step": 1000 + }, + { + "epoch": 2.1183932346723044, + "grad_norm": 1.2317125797271729, + "learning_rate": 5.440998489699731e-06, + "loss": 0.2101350724697113, + "step": 1002 + }, + { + "epoch": 2.1226215644820297, + "grad_norm": 0.6579080820083618, + "learning_rate": 5.425266047992672e-06, + "loss": 0.6576781868934631, + "step": 1004 + }, + { + "epoch": 2.126849894291755, + "grad_norm": 2.037614107131958, + "learning_rate": 5.4095345199382e-06, + "loss": 0.928475022315979, + "step": 1006 + }, + { + "epoch": 2.13107822410148, + "grad_norm": 1.618047833442688, + "learning_rate": 5.393804097860568e-06, + "loss": 0.8470478653907776, + "step": 1008 + }, + { + "epoch": 2.1353065539112053, + "grad_norm": 0.44501993060112, + "learning_rate": 5.3780749740705065e-06, + "loss": 0.6117417216300964, + "step": 1010 + }, + { + "epoch": 2.13953488372093, + "grad_norm": 0.653949499130249, + "learning_rate": 5.362347340862869e-06, + "loss": 0.5081562995910645, + "step": 1012 + }, + { + "epoch": 2.1437632135306552, + "grad_norm": 0.7473922371864319, + "learning_rate": 5.3466213905142935e-06, + "loss": 0.5856119990348816, + "step": 1014 + }, + { + "epoch": 2.1479915433403804, + "grad_norm": 0.39573782682418823, + "learning_rate": 5.3308973152808395e-06, + "loss": 0.7530453205108643, + "step": 1016 + }, + { + "epoch": 2.1522198731501057, + "grad_norm": 0.48412540555000305, + "learning_rate": 5.315175307395644e-06, + "loss": 0.8841409087181091, + "step": 1018 + }, + { + "epoch": 2.156448202959831, + "grad_norm": 0.8938184380531311, + "learning_rate": 5.299455559066571e-06, + "loss": 0.7354257106781006, + "step": 1020 + }, + { + "epoch": 2.160676532769556, + "grad_norm": 0.5679708123207092, + "learning_rate": 5.283738262473859e-06, + "loss": 0.6463567018508911, + "step": 1022 + }, + { + "epoch": 2.1649048625792813, + "grad_norm": 0.39241480827331543, + "learning_rate": 5.268023609767772e-06, + "loss": 0.8817542791366577, + "step": 1024 + }, + { + "epoch": 2.1691331923890065, + "grad_norm": 0.5357868075370789, + "learning_rate": 5.2523117930662545e-06, + "loss": 0.9386540651321411, + "step": 1026 + }, + { + "epoch": 2.1733615221987317, + "grad_norm": 0.4471728801727295, + "learning_rate": 5.2366030044525785e-06, + "loss": 0.944185197353363, + "step": 1028 + }, + { + "epoch": 2.177589852008457, + "grad_norm": 0.4940197169780731, + "learning_rate": 5.220897435972992e-06, + "loss": 0.9098086357116699, + "step": 1030 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 0.5487205386161804, + "learning_rate": 5.205195279634385e-06, + "loss": 0.9451584815979004, + "step": 1032 + }, + { + "epoch": 2.186046511627907, + "grad_norm": 0.5860447287559509, + "learning_rate": 5.1894967274019245e-06, + "loss": 1.023360252380371, + "step": 1034 + }, + { + "epoch": 2.190274841437632, + "grad_norm": 0.42879801988601685, + "learning_rate": 5.173801971196716e-06, + "loss": 0.6617995500564575, + "step": 1036 + }, + { + "epoch": 2.1945031712473573, + "grad_norm": 2.418081045150757, + "learning_rate": 5.158111202893464e-06, + "loss": 0.2913988530635834, + "step": 1038 + }, + { + "epoch": 2.1987315010570825, + "grad_norm": 0.7089012861251831, + "learning_rate": 5.142424614318112e-06, + "loss": 0.9743428230285645, + "step": 1040 + }, + { + "epoch": 2.2029598308668077, + "grad_norm": 1.3846787214279175, + "learning_rate": 5.12674239724551e-06, + "loss": 0.32029417157173157, + "step": 1042 + }, + { + "epoch": 2.207188160676533, + "grad_norm": 0.37406569719314575, + "learning_rate": 5.111064743397061e-06, + "loss": 0.9631579518318176, + "step": 1044 + }, + { + "epoch": 2.211416490486258, + "grad_norm": 0.4590833783149719, + "learning_rate": 5.09539184443838e-06, + "loss": 0.49289312958717346, + "step": 1046 + }, + { + "epoch": 2.2156448202959833, + "grad_norm": 1.1451623439788818, + "learning_rate": 5.079723891976956e-06, + "loss": 0.787111759185791, + "step": 1048 + }, + { + "epoch": 2.219873150105708, + "grad_norm": 0.7867758274078369, + "learning_rate": 5.064061077559803e-06, + "loss": 0.5016126036643982, + "step": 1050 + }, + { + "epoch": 2.2241014799154333, + "grad_norm": 0.4926295280456543, + "learning_rate": 5.048403592671116e-06, + "loss": 0.7997788190841675, + "step": 1052 + }, + { + "epoch": 2.2283298097251585, + "grad_norm": 0.5615513324737549, + "learning_rate": 5.032751628729943e-06, + "loss": 0.9547005891799927, + "step": 1054 + }, + { + "epoch": 2.2325581395348837, + "grad_norm": 0.37540164589881897, + "learning_rate": 5.017105377087829e-06, + "loss": 0.9538481831550598, + "step": 1056 + }, + { + "epoch": 2.236786469344609, + "grad_norm": 0.1772485077381134, + "learning_rate": 5.0014650290264854e-06, + "loss": 0.6759481430053711, + "step": 1058 + }, + { + "epoch": 2.241014799154334, + "grad_norm": 0.65249103307724, + "learning_rate": 4.9858307757554536e-06, + "loss": 0.7906907796859741, + "step": 1060 + }, + { + "epoch": 2.2452431289640593, + "grad_norm": 0.43567049503326416, + "learning_rate": 4.97020280840976e-06, + "loss": 0.89081209897995, + "step": 1062 + }, + { + "epoch": 2.2494714587737845, + "grad_norm": 0.8811886310577393, + "learning_rate": 4.954581318047582e-06, + "loss": 0.8416023850440979, + "step": 1064 + }, + { + "epoch": 2.2536997885835097, + "grad_norm": 0.8375807404518127, + "learning_rate": 4.938966495647918e-06, + "loss": 0.4984022378921509, + "step": 1066 + }, + { + "epoch": 2.2579281183932345, + "grad_norm": 0.6824806332588196, + "learning_rate": 4.923358532108244e-06, + "loss": 0.7385876178741455, + "step": 1068 + }, + { + "epoch": 2.2621564482029597, + "grad_norm": 1.5460448265075684, + "learning_rate": 4.907757618242186e-06, + "loss": 0.663308322429657, + "step": 1070 + }, + { + "epoch": 2.266384778012685, + "grad_norm": 0.4971848726272583, + "learning_rate": 4.892163944777181e-06, + "loss": 0.940421462059021, + "step": 1072 + }, + { + "epoch": 2.27061310782241, + "grad_norm": 0.3010624945163727, + "learning_rate": 4.876577702352155e-06, + "loss": 0.6079798340797424, + "step": 1074 + }, + { + "epoch": 2.2748414376321353, + "grad_norm": 0.7594307065010071, + "learning_rate": 4.860999081515183e-06, + "loss": 0.6558487415313721, + "step": 1076 + }, + { + "epoch": 2.2790697674418605, + "grad_norm": 2.838071346282959, + "learning_rate": 4.845428272721161e-06, + "loss": 0.22952839732170105, + "step": 1078 + }, + { + "epoch": 2.2832980972515857, + "grad_norm": 0.3779852092266083, + "learning_rate": 4.8298654663294865e-06, + "loss": 0.11567100882530212, + "step": 1080 + }, + { + "epoch": 2.287526427061311, + "grad_norm": 0.16041956841945648, + "learning_rate": 4.814310852601717e-06, + "loss": 0.48160320520401, + "step": 1082 + }, + { + "epoch": 2.291754756871036, + "grad_norm": 0.6174498200416565, + "learning_rate": 4.7987646216992545e-06, + "loss": 0.3124338686466217, + "step": 1084 + }, + { + "epoch": 2.295983086680761, + "grad_norm": 0.36216387152671814, + "learning_rate": 4.783226963681018e-06, + "loss": 0.18971526622772217, + "step": 1086 + }, + { + "epoch": 2.300211416490486, + "grad_norm": 0.9076690673828125, + "learning_rate": 4.76769806850112e-06, + "loss": 0.6296039819717407, + "step": 1088 + }, + { + "epoch": 2.3044397463002113, + "grad_norm": 0.45322856307029724, + "learning_rate": 4.752178126006544e-06, + "loss": 0.8020249605178833, + "step": 1090 + }, + { + "epoch": 2.3086680761099365, + "grad_norm": 0.8528197407722473, + "learning_rate": 4.736667325934819e-06, + "loss": 0.6081079244613647, + "step": 1092 + }, + { + "epoch": 2.3128964059196617, + "grad_norm": 0.6823382377624512, + "learning_rate": 4.721165857911709e-06, + "loss": 0.9147766828536987, + "step": 1094 + }, + { + "epoch": 2.317124735729387, + "grad_norm": 0.7384828925132751, + "learning_rate": 4.705673911448891e-06, + "loss": 0.8753478527069092, + "step": 1096 + }, + { + "epoch": 2.321353065539112, + "grad_norm": 0.6871476769447327, + "learning_rate": 4.690191675941631e-06, + "loss": 0.9042140245437622, + "step": 1098 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 5.743633270263672, + "learning_rate": 4.674719340666477e-06, + "loss": 0.44159966707229614, + "step": 1100 + }, + { + "epoch": 2.3298097251585626, + "grad_norm": 0.7584041357040405, + "learning_rate": 4.659257094778949e-06, + "loss": 0.6251407861709595, + "step": 1102 + }, + { + "epoch": 2.3340380549682873, + "grad_norm": 0.5574578642845154, + "learning_rate": 4.643805127311212e-06, + "loss": 0.9184064865112305, + "step": 1104 + }, + { + "epoch": 2.3382663847780125, + "grad_norm": 0.8222340941429138, + "learning_rate": 4.628363627169774e-06, + "loss": 0.8462425470352173, + "step": 1106 + }, + { + "epoch": 2.3424947145877377, + "grad_norm": 0.496639609336853, + "learning_rate": 4.612932783133182e-06, + "loss": 0.8468197584152222, + "step": 1108 + }, + { + "epoch": 2.346723044397463, + "grad_norm": 0.5885112881660461, + "learning_rate": 4.5975127838497036e-06, + "loss": 1.0569361448287964, + "step": 1110 + }, + { + "epoch": 2.350951374207188, + "grad_norm": 0.44990208745002747, + "learning_rate": 4.582103817835021e-06, + "loss": 0.6253592371940613, + "step": 1112 + }, + { + "epoch": 2.3551797040169133, + "grad_norm": 1.1458122730255127, + "learning_rate": 4.566706073469938e-06, + "loss": 0.7695745825767517, + "step": 1114 + }, + { + "epoch": 2.3594080338266386, + "grad_norm": 0.5391276478767395, + "learning_rate": 4.551319738998067e-06, + "loss": 0.576178789138794, + "step": 1116 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 1.3838157653808594, + "learning_rate": 4.535945002523524e-06, + "loss": 1.1193486452102661, + "step": 1118 + }, + { + "epoch": 2.367864693446089, + "grad_norm": 0.4469482898712158, + "learning_rate": 4.5205820520086415e-06, + "loss": 0.9312993288040161, + "step": 1120 + }, + { + "epoch": 2.3720930232558137, + "grad_norm": 0.4894274175167084, + "learning_rate": 4.505231075271662e-06, + "loss": 0.7663894295692444, + "step": 1122 + }, + { + "epoch": 2.376321353065539, + "grad_norm": 0.700397253036499, + "learning_rate": 4.489892259984446e-06, + "loss": 0.6583611965179443, + "step": 1124 + }, + { + "epoch": 2.380549682875264, + "grad_norm": 1.4910742044448853, + "learning_rate": 4.474565793670167e-06, + "loss": 0.9124075174331665, + "step": 1126 + }, + { + "epoch": 2.3847780126849893, + "grad_norm": 2.0670645236968994, + "learning_rate": 4.459251863701038e-06, + "loss": 0.6279240250587463, + "step": 1128 + }, + { + "epoch": 2.3890063424947146, + "grad_norm": 0.7884926795959473, + "learning_rate": 4.443950657296004e-06, + "loss": 0.7046691179275513, + "step": 1130 + }, + { + "epoch": 2.3932346723044398, + "grad_norm": 0.7733293771743774, + "learning_rate": 4.4286623615184585e-06, + "loss": 0.5494440793991089, + "step": 1132 + }, + { + "epoch": 2.397463002114165, + "grad_norm": 0.45177239179611206, + "learning_rate": 4.413387163273964e-06, + "loss": 0.9340068101882935, + "step": 1134 + }, + { + "epoch": 2.40169133192389, + "grad_norm": 2.670645236968994, + "learning_rate": 4.3981252493079506e-06, + "loss": 0.6311877965927124, + "step": 1136 + }, + { + "epoch": 2.4059196617336154, + "grad_norm": 0.7255088090896606, + "learning_rate": 4.382876806203453e-06, + "loss": 0.7587140202522278, + "step": 1138 + }, + { + "epoch": 2.41014799154334, + "grad_norm": 0.407471626996994, + "learning_rate": 4.3676420203788125e-06, + "loss": 0.9154732823371887, + "step": 1140 + }, + { + "epoch": 2.4143763213530653, + "grad_norm": 0.4105207622051239, + "learning_rate": 4.352421078085405e-06, + "loss": 0.929002583026886, + "step": 1142 + }, + { + "epoch": 2.4186046511627906, + "grad_norm": 0.7631443738937378, + "learning_rate": 4.337214165405368e-06, + "loss": 0.45309212803840637, + "step": 1144 + }, + { + "epoch": 2.4228329809725158, + "grad_norm": 1.3268755674362183, + "learning_rate": 4.322021468249316e-06, + "loss": 0.7121126055717468, + "step": 1146 + }, + { + "epoch": 2.427061310782241, + "grad_norm": 0.3983864188194275, + "learning_rate": 4.306843172354076e-06, + "loss": 0.7074201703071594, + "step": 1148 + }, + { + "epoch": 2.431289640591966, + "grad_norm": 1.0654891729354858, + "learning_rate": 4.291679463280416e-06, + "loss": 0.46396470069885254, + "step": 1150 + }, + { + "epoch": 2.4355179704016914, + "grad_norm": 0.6108584403991699, + "learning_rate": 4.276530526410768e-06, + "loss": 0.601567804813385, + "step": 1152 + }, + { + "epoch": 2.4397463002114166, + "grad_norm": 0.6544230580329895, + "learning_rate": 4.261396546946972e-06, + "loss": 0.48921969532966614, + "step": 1154 + }, + { + "epoch": 2.443974630021142, + "grad_norm": 0.6133583784103394, + "learning_rate": 4.246277709908009e-06, + "loss": 0.819734513759613, + "step": 1156 + }, + { + "epoch": 2.448202959830867, + "grad_norm": 0.4919082224369049, + "learning_rate": 4.231174200127736e-06, + "loss": 0.5521842837333679, + "step": 1158 + }, + { + "epoch": 2.452431289640592, + "grad_norm": 0.43918028473854065, + "learning_rate": 4.216086202252624e-06, + "loss": 0.9051367044448853, + "step": 1160 + }, + { + "epoch": 2.456659619450317, + "grad_norm": 1.2088602781295776, + "learning_rate": 4.201013900739513e-06, + "loss": 0.8928288221359253, + "step": 1162 + }, + { + "epoch": 2.460887949260042, + "grad_norm": 0.5467002391815186, + "learning_rate": 4.185957479853344e-06, + "loss": 0.6880935430526733, + "step": 1164 + }, + { + "epoch": 2.4651162790697674, + "grad_norm": 0.6876582503318787, + "learning_rate": 4.1709171236649096e-06, + "loss": 1.0135117769241333, + "step": 1166 + }, + { + "epoch": 2.4693446088794926, + "grad_norm": 0.3953462243080139, + "learning_rate": 4.155893016048607e-06, + "loss": 1.2272512912750244, + "step": 1168 + }, + { + "epoch": 2.473572938689218, + "grad_norm": 0.5110251903533936, + "learning_rate": 4.140885340680189e-06, + "loss": 0.654666006565094, + "step": 1170 + }, + { + "epoch": 2.477801268498943, + "grad_norm": 0.8853302597999573, + "learning_rate": 4.125894281034517e-06, + "loss": 0.7475524544715881, + "step": 1172 + }, + { + "epoch": 2.482029598308668, + "grad_norm": 0.3335699141025543, + "learning_rate": 4.110920020383316e-06, + "loss": 0.6792094111442566, + "step": 1174 + }, + { + "epoch": 2.4862579281183934, + "grad_norm": 0.4461216628551483, + "learning_rate": 4.0959627417929386e-06, + "loss": 0.6548753380775452, + "step": 1176 + }, + { + "epoch": 2.4904862579281186, + "grad_norm": 1.2208702564239502, + "learning_rate": 4.081022628122127e-06, + "loss": 0.7311459183692932, + "step": 1178 + }, + { + "epoch": 2.4947145877378434, + "grad_norm": 1.5550645589828491, + "learning_rate": 4.06609986201977e-06, + "loss": 0.5898609757423401, + "step": 1180 + }, + { + "epoch": 2.4989429175475686, + "grad_norm": 0.5158658623695374, + "learning_rate": 4.051194625922681e-06, + "loss": 0.41105133295059204, + "step": 1182 + }, + { + "epoch": 2.503171247357294, + "grad_norm": 0.45216265320777893, + "learning_rate": 4.036307102053359e-06, + "loss": 0.1280100792646408, + "step": 1184 + }, + { + "epoch": 2.507399577167019, + "grad_norm": 0.5537520051002502, + "learning_rate": 4.021437472417763e-06, + "loss": 0.8554248809814453, + "step": 1186 + }, + { + "epoch": 2.511627906976744, + "grad_norm": 1.033736228942871, + "learning_rate": 4.006585918803091e-06, + "loss": 0.6770380735397339, + "step": 1188 + }, + { + "epoch": 2.5158562367864694, + "grad_norm": 1.3594704866409302, + "learning_rate": 3.991752622775551e-06, + "loss": 0.6196041107177734, + "step": 1190 + }, + { + "epoch": 2.5200845665961946, + "grad_norm": 0.40698787569999695, + "learning_rate": 3.97693776567815e-06, + "loss": 0.8643728494644165, + "step": 1192 + }, + { + "epoch": 2.52431289640592, + "grad_norm": 0.44846105575561523, + "learning_rate": 3.962141528628465e-06, + "loss": 0.5829656720161438, + "step": 1194 + }, + { + "epoch": 2.528541226215645, + "grad_norm": 3.294137954711914, + "learning_rate": 3.947364092516437e-06, + "loss": 0.7900627255439758, + "step": 1196 + }, + { + "epoch": 2.53276955602537, + "grad_norm": 0.3768402636051178, + "learning_rate": 3.932605638002163e-06, + "loss": 0.901138961315155, + "step": 1198 + }, + { + "epoch": 2.536997885835095, + "grad_norm": 0.666864275932312, + "learning_rate": 3.917866345513677e-06, + "loss": 0.6684475541114807, + "step": 1200 + }, + { + "epoch": 2.54122621564482, + "grad_norm": 0.6581912636756897, + "learning_rate": 3.903146395244749e-06, + "loss": 0.8979118466377258, + "step": 1202 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 0.9961763024330139, + "learning_rate": 3.888445967152691e-06, + "loss": 0.30583012104034424, + "step": 1204 + }, + { + "epoch": 2.5496828752642706, + "grad_norm": 0.6817377805709839, + "learning_rate": 3.873765240956138e-06, + "loss": 0.9645211696624756, + "step": 1206 + }, + { + "epoch": 2.553911205073996, + "grad_norm": 1.19502592086792, + "learning_rate": 3.859104396132863e-06, + "loss": 1.0073546171188354, + "step": 1208 + }, + { + "epoch": 2.558139534883721, + "grad_norm": 0.5000209212303162, + "learning_rate": 3.84446361191759e-06, + "loss": 0.9796168208122253, + "step": 1210 + }, + { + "epoch": 2.5623678646934462, + "grad_norm": 2.498603582382202, + "learning_rate": 3.829843067299788e-06, + "loss": 0.6499704122543335, + "step": 1212 + }, + { + "epoch": 2.5665961945031714, + "grad_norm": 0.5283212661743164, + "learning_rate": 3.8152429410214834e-06, + "loss": 0.8957065343856812, + "step": 1214 + }, + { + "epoch": 2.570824524312896, + "grad_norm": 1.1246083974838257, + "learning_rate": 3.8006634115750886e-06, + "loss": 0.5845621824264526, + "step": 1216 + }, + { + "epoch": 2.5750528541226214, + "grad_norm": 1.1732431650161743, + "learning_rate": 3.7861046572012105e-06, + "loss": 0.5534276366233826, + "step": 1218 + }, + { + "epoch": 2.5792811839323466, + "grad_norm": 1.440114140510559, + "learning_rate": 3.771566855886467e-06, + "loss": 0.4712372124195099, + "step": 1220 + }, + { + "epoch": 2.583509513742072, + "grad_norm": 1.4238877296447754, + "learning_rate": 3.7570501853613227e-06, + "loss": 0.5444931983947754, + "step": 1222 + }, + { + "epoch": 2.587737843551797, + "grad_norm": 1.0960928201675415, + "learning_rate": 3.742554823097905e-06, + "loss": 0.5549309253692627, + "step": 1224 + }, + { + "epoch": 2.5919661733615222, + "grad_norm": 0.2020740658044815, + "learning_rate": 3.7280809463078436e-06, + "loss": 0.3822484016418457, + "step": 1226 + }, + { + "epoch": 2.5961945031712474, + "grad_norm": 0.3880596160888672, + "learning_rate": 3.7136287319400944e-06, + "loss": 0.6528080701828003, + "step": 1228 + }, + { + "epoch": 2.6004228329809727, + "grad_norm": 2.4288744926452637, + "learning_rate": 3.6991983566787847e-06, + "loss": 0.7461565732955933, + "step": 1230 + }, + { + "epoch": 2.604651162790698, + "grad_norm": 0.7945540547370911, + "learning_rate": 3.6847899969410515e-06, + "loss": 0.5631217360496521, + "step": 1232 + }, + { + "epoch": 2.6088794926004226, + "grad_norm": 1.3215166330337524, + "learning_rate": 3.6704038288748812e-06, + "loss": 0.6149224638938904, + "step": 1234 + }, + { + "epoch": 2.613107822410148, + "grad_norm": 0.4192475378513336, + "learning_rate": 3.6560400283569543e-06, + "loss": 0.7575697302818298, + "step": 1236 + }, + { + "epoch": 2.617336152219873, + "grad_norm": 1.3973437547683716, + "learning_rate": 3.641698770990507e-06, + "loss": 0.8324413895606995, + "step": 1238 + }, + { + "epoch": 2.6215644820295982, + "grad_norm": 0.4712655544281006, + "learning_rate": 3.6273802321031738e-06, + "loss": 0.6536996960639954, + "step": 1240 + }, + { + "epoch": 2.6257928118393234, + "grad_norm": 0.4749317169189453, + "learning_rate": 3.6130845867448384e-06, + "loss": 0.7839347124099731, + "step": 1242 + }, + { + "epoch": 2.6300211416490487, + "grad_norm": 0.5369474291801453, + "learning_rate": 3.598812009685514e-06, + "loss": 0.9155866503715515, + "step": 1244 + }, + { + "epoch": 2.634249471458774, + "grad_norm": 1.1646672487258911, + "learning_rate": 3.584562675413189e-06, + "loss": 0.6453189849853516, + "step": 1246 + }, + { + "epoch": 2.638477801268499, + "grad_norm": 0.7816519737243652, + "learning_rate": 3.5703367581316985e-06, + "loss": 0.48041459918022156, + "step": 1248 + }, + { + "epoch": 2.6427061310782243, + "grad_norm": 0.5937077403068542, + "learning_rate": 3.5561344317585957e-06, + "loss": 0.9281110763549805, + "step": 1250 + }, + { + "epoch": 2.646934460887949, + "grad_norm": 1.2447341680526733, + "learning_rate": 3.5419558699230305e-06, + "loss": 0.4202560484409332, + "step": 1252 + }, + { + "epoch": 2.6511627906976747, + "grad_norm": 1.3562244176864624, + "learning_rate": 3.52780124596361e-06, + "loss": 0.5819510221481323, + "step": 1254 + }, + { + "epoch": 2.6553911205073994, + "grad_norm": 0.33719658851623535, + "learning_rate": 3.5136707329263055e-06, + "loss": 0.9201833009719849, + "step": 1256 + }, + { + "epoch": 2.6596194503171247, + "grad_norm": 0.7231669425964355, + "learning_rate": 3.499564503562315e-06, + "loss": 0.21925120055675507, + "step": 1258 + }, + { + "epoch": 2.66384778012685, + "grad_norm": 0.3642725348472595, + "learning_rate": 3.4854827303259575e-06, + "loss": 0.47189339995384216, + "step": 1260 + }, + { + "epoch": 2.668076109936575, + "grad_norm": 1.0332915782928467, + "learning_rate": 3.47142558537257e-06, + "loss": 0.6024248003959656, + "step": 1262 + }, + { + "epoch": 2.6723044397463003, + "grad_norm": 1.0206177234649658, + "learning_rate": 3.4573932405563966e-06, + "loss": 0.6105507016181946, + "step": 1264 + }, + { + "epoch": 2.6765327695560255, + "grad_norm": 1.067035436630249, + "learning_rate": 3.4433858674284894e-06, + "loss": 0.316377192735672, + "step": 1266 + }, + { + "epoch": 2.6807610993657507, + "grad_norm": 0.4384140074253082, + "learning_rate": 3.429403637234614e-06, + "loss": 0.750786304473877, + "step": 1268 + }, + { + "epoch": 2.6849894291754755, + "grad_norm": 0.7142227292060852, + "learning_rate": 3.4154467209131515e-06, + "loss": 0.8820692300796509, + "step": 1270 + }, + { + "epoch": 2.689217758985201, + "grad_norm": 1.4679869413375854, + "learning_rate": 3.401515289093011e-06, + "loss": 0.6552101969718933, + "step": 1272 + }, + { + "epoch": 2.693446088794926, + "grad_norm": 0.5077518820762634, + "learning_rate": 3.3876095120915433e-06, + "loss": 0.3963017761707306, + "step": 1274 + }, + { + "epoch": 2.697674418604651, + "grad_norm": 0.6193831562995911, + "learning_rate": 3.3737295599124592e-06, + "loss": 0.3977617025375366, + "step": 1276 + }, + { + "epoch": 2.7019027484143763, + "grad_norm": 0.5356454849243164, + "learning_rate": 3.3598756022437507e-06, + "loss": 0.8895797729492188, + "step": 1278 + }, + { + "epoch": 2.7061310782241015, + "grad_norm": 0.9390637278556824, + "learning_rate": 3.3460478084556157e-06, + "loss": 0.6869379878044128, + "step": 1280 + }, + { + "epoch": 2.7103594080338267, + "grad_norm": 0.6072316765785217, + "learning_rate": 3.33224634759839e-06, + "loss": 1.0019521713256836, + "step": 1282 + }, + { + "epoch": 2.714587737843552, + "grad_norm": 0.2173154503107071, + "learning_rate": 3.318471388400476e-06, + "loss": 0.6600397229194641, + "step": 1284 + }, + { + "epoch": 2.718816067653277, + "grad_norm": 0.13677340745925903, + "learning_rate": 3.3047230992662834e-06, + "loss": 0.9050495624542236, + "step": 1286 + }, + { + "epoch": 2.723044397463002, + "grad_norm": 0.35973480343818665, + "learning_rate": 3.291001648274175e-06, + "loss": 0.4734467566013336, + "step": 1288 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 0.5269657373428345, + "learning_rate": 3.277307203174395e-06, + "loss": 0.8756322264671326, + "step": 1290 + }, + { + "epoch": 2.7315010570824523, + "grad_norm": 0.4811011850833893, + "learning_rate": 3.263639931387039e-06, + "loss": 0.9484678506851196, + "step": 1292 + }, + { + "epoch": 2.7357293868921775, + "grad_norm": 0.9223408102989197, + "learning_rate": 3.2500000000000015e-06, + "loss": 0.9398493766784668, + "step": 1294 + }, + { + "epoch": 2.7399577167019027, + "grad_norm": 0.4348272979259491, + "learning_rate": 3.236387575766916e-06, + "loss": 0.2903934717178345, + "step": 1296 + }, + { + "epoch": 2.744186046511628, + "grad_norm": 2.80344557762146, + "learning_rate": 3.222802825105138e-06, + "loss": 0.5232171416282654, + "step": 1298 + }, + { + "epoch": 2.748414376321353, + "grad_norm": 0.4914420545101166, + "learning_rate": 3.209245914093708e-06, + "loss": 0.6540259718894958, + "step": 1300 + }, + { + "epoch": 2.7526427061310783, + "grad_norm": 0.6232413649559021, + "learning_rate": 3.1957170084713053e-06, + "loss": 0.7272835969924927, + "step": 1302 + }, + { + "epoch": 2.7568710359408035, + "grad_norm": 0.9709661602973938, + "learning_rate": 3.182216273634235e-06, + "loss": 0.5396301746368408, + "step": 1304 + }, + { + "epoch": 2.7610993657505283, + "grad_norm": 1.5130153894424438, + "learning_rate": 3.1687438746344086e-06, + "loss": 0.76607346534729, + "step": 1306 + }, + { + "epoch": 2.765327695560254, + "grad_norm": 1.499550700187683, + "learning_rate": 3.1552999761773166e-06, + "loss": 0.7811456918716431, + "step": 1308 + }, + { + "epoch": 2.7695560253699787, + "grad_norm": 0.4561913311481476, + "learning_rate": 3.141884742620013e-06, + "loss": 0.9177801609039307, + "step": 1310 + }, + { + "epoch": 2.773784355179704, + "grad_norm": 0.5267754197120667, + "learning_rate": 3.1284983379691247e-06, + "loss": 0.765741765499115, + "step": 1312 + }, + { + "epoch": 2.778012684989429, + "grad_norm": 0.4474433660507202, + "learning_rate": 3.1151409258788235e-06, + "loss": 0.46959996223449707, + "step": 1314 + }, + { + "epoch": 2.7822410147991543, + "grad_norm": 0.5307146310806274, + "learning_rate": 3.101812669648838e-06, + "loss": 0.8912549614906311, + "step": 1316 + }, + { + "epoch": 2.7864693446088795, + "grad_norm": 0.1605590134859085, + "learning_rate": 3.0885137322224546e-06, + "loss": 0.5218514800071716, + "step": 1318 + }, + { + "epoch": 2.7906976744186047, + "grad_norm": 1.1033434867858887, + "learning_rate": 3.075244276184529e-06, + "loss": 0.8091227412223816, + "step": 1320 + }, + { + "epoch": 2.79492600422833, + "grad_norm": 1.2507354021072388, + "learning_rate": 3.0620044637594936e-06, + "loss": 0.30721786618232727, + "step": 1322 + }, + { + "epoch": 2.7991543340380547, + "grad_norm": 0.4536409080028534, + "learning_rate": 3.0487944568093685e-06, + "loss": 0.9971051812171936, + "step": 1324 + }, + { + "epoch": 2.8033826638477803, + "grad_norm": 1.4768913984298706, + "learning_rate": 3.035614416831801e-06, + "loss": 0.21633179485797882, + "step": 1326 + }, + { + "epoch": 2.807610993657505, + "grad_norm": 1.2097821235656738, + "learning_rate": 3.0224645049580746e-06, + "loss": 0.3117099702358246, + "step": 1328 + }, + { + "epoch": 2.8118393234672303, + "grad_norm": 0.46743783354759216, + "learning_rate": 3.0093448819511416e-06, + "loss": 0.5532126426696777, + "step": 1330 + }, + { + "epoch": 2.8160676532769555, + "grad_norm": 0.44059836864471436, + "learning_rate": 2.9962557082036677e-06, + "loss": 0.8996551036834717, + "step": 1332 + }, + { + "epoch": 2.8202959830866807, + "grad_norm": 0.49607768654823303, + "learning_rate": 2.9831971437360606e-06, + "loss": 0.6198577284812927, + "step": 1334 + }, + { + "epoch": 2.824524312896406, + "grad_norm": 0.6772405505180359, + "learning_rate": 2.9701693481945132e-06, + "loss": 0.9710506200790405, + "step": 1336 + }, + { + "epoch": 2.828752642706131, + "grad_norm": 0.4741802215576172, + "learning_rate": 2.957172480849064e-06, + "loss": 0.9453369379043579, + "step": 1338 + }, + { + "epoch": 2.8329809725158563, + "grad_norm": 0.4236464202404022, + "learning_rate": 2.9442067005916375e-06, + "loss": 0.674446702003479, + "step": 1340 + }, + { + "epoch": 2.8372093023255816, + "grad_norm": 0.34691157937049866, + "learning_rate": 2.931272165934105e-06, + "loss": 1.0024499893188477, + "step": 1342 + }, + { + "epoch": 2.8414376321353068, + "grad_norm": 0.6685620546340942, + "learning_rate": 2.918369035006351e-06, + "loss": 0.5898784399032593, + "step": 1344 + }, + { + "epoch": 2.8456659619450315, + "grad_norm": 1.3949618339538574, + "learning_rate": 2.9054974655543335e-06, + "loss": 0.5850456357002258, + "step": 1346 + }, + { + "epoch": 2.8498942917547567, + "grad_norm": 0.5828827023506165, + "learning_rate": 2.89265761493816e-06, + "loss": 0.8425498008728027, + "step": 1348 + }, + { + "epoch": 2.854122621564482, + "grad_norm": 1.5113717317581177, + "learning_rate": 2.8798496401301644e-06, + "loss": 0.5016194581985474, + "step": 1350 + }, + { + "epoch": 2.858350951374207, + "grad_norm": 0.8801661729812622, + "learning_rate": 2.8670736977129843e-06, + "loss": 0.8054553270339966, + "step": 1352 + }, + { + "epoch": 2.8625792811839323, + "grad_norm": 0.4619808793067932, + "learning_rate": 2.854329943877647e-06, + "loss": 0.9055795669555664, + "step": 1354 + }, + { + "epoch": 2.8668076109936576, + "grad_norm": 0.6821467876434326, + "learning_rate": 2.841618534421664e-06, + "loss": 0.2950429320335388, + "step": 1356 + }, + { + "epoch": 2.8710359408033828, + "grad_norm": 1.3057900667190552, + "learning_rate": 2.8289396247471227e-06, + "loss": 0.9481149911880493, + "step": 1358 + }, + { + "epoch": 2.875264270613108, + "grad_norm": 1.2999907732009888, + "learning_rate": 2.816293369858788e-06, + "loss": 0.7217288017272949, + "step": 1360 + }, + { + "epoch": 2.879492600422833, + "grad_norm": 11.962328910827637, + "learning_rate": 2.803679924362206e-06, + "loss": 0.7072125673294067, + "step": 1362 + }, + { + "epoch": 2.883720930232558, + "grad_norm": 0.5187685489654541, + "learning_rate": 2.7910994424618165e-06, + "loss": 0.995893657207489, + "step": 1364 + }, + { + "epoch": 2.887949260042283, + "grad_norm": 1.9199581146240234, + "learning_rate": 2.778552077959064e-06, + "loss": 0.7320709228515625, + "step": 1366 + }, + { + "epoch": 2.8921775898520083, + "grad_norm": 0.6116196513175964, + "learning_rate": 2.766037984250521e-06, + "loss": 0.9193503856658936, + "step": 1368 + }, + { + "epoch": 2.8964059196617336, + "grad_norm": 0.4953746795654297, + "learning_rate": 2.753557314326011e-06, + "loss": 0.9296433329582214, + "step": 1370 + }, + { + "epoch": 2.9006342494714588, + "grad_norm": 1.113531231880188, + "learning_rate": 2.7411102207667373e-06, + "loss": 0.9421595335006714, + "step": 1372 + }, + { + "epoch": 2.904862579281184, + "grad_norm": 0.4699031412601471, + "learning_rate": 2.728696855743418e-06, + "loss": 0.7320655584335327, + "step": 1374 + }, + { + "epoch": 2.909090909090909, + "grad_norm": 0.4174327552318573, + "learning_rate": 2.7163173710144323e-06, + "loss": 0.9137600064277649, + "step": 1376 + }, + { + "epoch": 2.9133192389006344, + "grad_norm": 1.570842981338501, + "learning_rate": 2.703971917923951e-06, + "loss": 0.5102354288101196, + "step": 1378 + }, + { + "epoch": 2.9175475687103596, + "grad_norm": 0.7760786414146423, + "learning_rate": 2.6916606474000965e-06, + "loss": 0.6727113723754883, + "step": 1380 + }, + { + "epoch": 2.9217758985200843, + "grad_norm": 0.6817591190338135, + "learning_rate": 2.679383709953103e-06, + "loss": 0.9822676181793213, + "step": 1382 + }, + { + "epoch": 2.92600422832981, + "grad_norm": 1.9100477695465088, + "learning_rate": 2.6671412556734578e-06, + "loss": 0.7030581831932068, + "step": 1384 + }, + { + "epoch": 2.9302325581395348, + "grad_norm": 0.7380037307739258, + "learning_rate": 2.6549334342300844e-06, + "loss": 0.5955653786659241, + "step": 1386 + }, + { + "epoch": 2.93446088794926, + "grad_norm": 0.5143433809280396, + "learning_rate": 2.642760394868503e-06, + "loss": 0.9238606691360474, + "step": 1388 + }, + { + "epoch": 2.938689217758985, + "grad_norm": 0.23438040912151337, + "learning_rate": 2.6306222864090157e-06, + "loss": 0.7150290012359619, + "step": 1390 + }, + { + "epoch": 2.9429175475687104, + "grad_norm": 0.592908501625061, + "learning_rate": 2.61851925724487e-06, + "loss": 0.8776857256889343, + "step": 1392 + }, + { + "epoch": 2.9471458773784356, + "grad_norm": 0.25268545746803284, + "learning_rate": 2.606451455340459e-06, + "loss": 0.22425909340381622, + "step": 1394 + }, + { + "epoch": 2.951374207188161, + "grad_norm": 0.4203673005104065, + "learning_rate": 2.5944190282295155e-06, + "loss": 0.4790744483470917, + "step": 1396 + }, + { + "epoch": 2.955602536997886, + "grad_norm": 0.9801953434944153, + "learning_rate": 2.5824221230132895e-06, + "loss": 0.8205451965332031, + "step": 1398 + }, + { + "epoch": 2.9598308668076108, + "grad_norm": 0.5639576315879822, + "learning_rate": 2.5704608863587676e-06, + "loss": 0.9858054518699646, + "step": 1400 + }, + { + "epoch": 2.9640591966173364, + "grad_norm": 0.9544370174407959, + "learning_rate": 2.558535464496877e-06, + "loss": 0.7061154842376709, + "step": 1402 + }, + { + "epoch": 2.968287526427061, + "grad_norm": 0.3250182569026947, + "learning_rate": 2.5466460032206872e-06, + "loss": 0.4353572726249695, + "step": 1404 + }, + { + "epoch": 2.9725158562367864, + "grad_norm": 1.3899668455123901, + "learning_rate": 2.53479264788364e-06, + "loss": 0.8007202744483948, + "step": 1406 + }, + { + "epoch": 2.9767441860465116, + "grad_norm": 0.46989011764526367, + "learning_rate": 2.522975543397769e-06, + "loss": 0.7221790552139282, + "step": 1408 + }, + { + "epoch": 2.980972515856237, + "grad_norm": 1.3654996156692505, + "learning_rate": 2.511194834231922e-06, + "loss": 0.6193009614944458, + "step": 1410 + }, + { + "epoch": 2.985200845665962, + "grad_norm": 0.6509087085723877, + "learning_rate": 2.49945066441e-06, + "loss": 0.9009639024734497, + "step": 1412 + }, + { + "epoch": 2.989429175475687, + "grad_norm": 0.33879584074020386, + "learning_rate": 2.4877431775092004e-06, + "loss": 0.6752570271492004, + "step": 1414 + }, + { + "epoch": 2.9936575052854124, + "grad_norm": 1.8351242542266846, + "learning_rate": 2.4760725166582544e-06, + "loss": 0.7708001732826233, + "step": 1416 + }, + { + "epoch": 2.997885835095137, + "grad_norm": 0.4075961410999298, + "learning_rate": 2.464438824535675e-06, + "loss": 0.5021908283233643, + "step": 1418 + }, + { + "epoch": 3.0021141649048624, + "grad_norm": 0.49236583709716797, + "learning_rate": 2.452842243368028e-06, + "loss": 0.8916373252868652, + "step": 1420 + }, + { + "epoch": 3.0063424947145876, + "grad_norm": 0.5360115766525269, + "learning_rate": 2.4412829149281735e-06, + "loss": 0.4697900712490082, + "step": 1422 + }, + { + "epoch": 3.010570824524313, + "grad_norm": 1.1159920692443848, + "learning_rate": 2.4297609805335472e-06, + "loss": 0.7837058305740356, + "step": 1424 + }, + { + "epoch": 3.014799154334038, + "grad_norm": 1.4505268335342407, + "learning_rate": 2.4182765810444263e-06, + "loss": 0.4084257185459137, + "step": 1426 + }, + { + "epoch": 3.019027484143763, + "grad_norm": 0.4365761876106262, + "learning_rate": 2.4068298568622083e-06, + "loss": 0.47668635845184326, + "step": 1428 + }, + { + "epoch": 3.0232558139534884, + "grad_norm": 0.7739666104316711, + "learning_rate": 2.395420947927694e-06, + "loss": 0.6339631080627441, + "step": 1430 + }, + { + "epoch": 3.0274841437632136, + "grad_norm": 0.5527088046073914, + "learning_rate": 2.3840499937193804e-06, + "loss": 0.840812087059021, + "step": 1432 + }, + { + "epoch": 3.031712473572939, + "grad_norm": 0.4348757266998291, + "learning_rate": 2.3727171332517496e-06, + "loss": 0.7406386733055115, + "step": 1434 + }, + { + "epoch": 3.035940803382664, + "grad_norm": 1.594217300415039, + "learning_rate": 2.361422505073575e-06, + "loss": 0.4531082808971405, + "step": 1436 + }, + { + "epoch": 3.040169133192389, + "grad_norm": 0.21405461430549622, + "learning_rate": 2.3501662472662233e-06, + "loss": 0.07939190417528152, + "step": 1438 + }, + { + "epoch": 3.044397463002114, + "grad_norm": 0.48771965503692627, + "learning_rate": 2.3389484974419685e-06, + "loss": 0.7782604098320007, + "step": 1440 + }, + { + "epoch": 3.048625792811839, + "grad_norm": 0.4234258830547333, + "learning_rate": 2.327769392742309e-06, + "loss": 0.5754956603050232, + "step": 1442 + }, + { + "epoch": 3.0528541226215644, + "grad_norm": 1.5171079635620117, + "learning_rate": 2.316629069836291e-06, + "loss": 0.14752614498138428, + "step": 1444 + }, + { + "epoch": 3.0570824524312896, + "grad_norm": 1.5549007654190063, + "learning_rate": 2.305527664918838e-06, + "loss": 0.4007459878921509, + "step": 1446 + }, + { + "epoch": 3.061310782241015, + "grad_norm": 12.22429084777832, + "learning_rate": 2.294465313709085e-06, + "loss": 0.5347014665603638, + "step": 1448 + }, + { + "epoch": 3.06553911205074, + "grad_norm": 0.4916377067565918, + "learning_rate": 2.2834421514487205e-06, + "loss": 0.4656146466732025, + "step": 1450 + }, + { + "epoch": 3.0697674418604652, + "grad_norm": 0.9038828015327454, + "learning_rate": 2.27245831290033e-06, + "loss": 0.5444592237472534, + "step": 1452 + }, + { + "epoch": 3.0739957716701904, + "grad_norm": 0.7440728545188904, + "learning_rate": 2.261513932345755e-06, + "loss": 0.7768815755844116, + "step": 1454 + }, + { + "epoch": 3.0782241014799157, + "grad_norm": 0.2805899679660797, + "learning_rate": 2.250609143584443e-06, + "loss": 0.12607897818088531, + "step": 1456 + }, + { + "epoch": 3.0824524312896404, + "grad_norm": 0.3441605865955353, + "learning_rate": 2.239744079931818e-06, + "loss": 0.5026092529296875, + "step": 1458 + }, + { + "epoch": 3.0866807610993656, + "grad_norm": 0.6418730616569519, + "learning_rate": 2.228918874217649e-06, + "loss": 0.7241525650024414, + "step": 1460 + }, + { + "epoch": 3.090909090909091, + "grad_norm": 0.4679485559463501, + "learning_rate": 2.2181336587844244e-06, + "loss": 0.7635135054588318, + "step": 1462 + }, + { + "epoch": 3.095137420718816, + "grad_norm": 0.5657840371131897, + "learning_rate": 2.207388565485736e-06, + "loss": 0.4499300718307495, + "step": 1464 + }, + { + "epoch": 3.0993657505285412, + "grad_norm": 1.5068707466125488, + "learning_rate": 2.196683725684666e-06, + "loss": 0.585332453250885, + "step": 1466 + }, + { + "epoch": 3.1035940803382664, + "grad_norm": 0.39527320861816406, + "learning_rate": 2.1860192702521826e-06, + "loss": 0.8004910349845886, + "step": 1468 + }, + { + "epoch": 3.1078224101479917, + "grad_norm": 0.6444796919822693, + "learning_rate": 2.1753953295655354e-06, + "loss": 0.8160009384155273, + "step": 1470 + }, + { + "epoch": 3.112050739957717, + "grad_norm": 1.7423079013824463, + "learning_rate": 2.164812033506673e-06, + "loss": 0.3933432102203369, + "step": 1472 + }, + { + "epoch": 3.116279069767442, + "grad_norm": 0.5129163861274719, + "learning_rate": 2.1542695114606367e-06, + "loss": 0.8576091527938843, + "step": 1474 + }, + { + "epoch": 3.120507399577167, + "grad_norm": 0.4182908833026886, + "learning_rate": 2.1437678923139944e-06, + "loss": 0.7476510405540466, + "step": 1476 + }, + { + "epoch": 3.124735729386892, + "grad_norm": 1.8172433376312256, + "learning_rate": 2.1333073044532628e-06, + "loss": 0.1788163185119629, + "step": 1478 + }, + { + "epoch": 3.1289640591966172, + "grad_norm": 0.7242061495780945, + "learning_rate": 2.122887875763327e-06, + "loss": 0.8786553740501404, + "step": 1480 + }, + { + "epoch": 3.1331923890063424, + "grad_norm": 0.4321635961532593, + "learning_rate": 2.1125097336258897e-06, + "loss": 0.8074304461479187, + "step": 1482 + }, + { + "epoch": 3.1374207188160677, + "grad_norm": 0.25225207209587097, + "learning_rate": 2.1021730049179117e-06, + "loss": 0.17656999826431274, + "step": 1484 + }, + { + "epoch": 3.141649048625793, + "grad_norm": 0.612185537815094, + "learning_rate": 2.09187781601005e-06, + "loss": 0.6785790920257568, + "step": 1486 + }, + { + "epoch": 3.145877378435518, + "grad_norm": 0.4040587544441223, + "learning_rate": 2.0816242927651254e-06, + "loss": 0.9036017656326294, + "step": 1488 + }, + { + "epoch": 3.1501057082452433, + "grad_norm": 1.8786982297897339, + "learning_rate": 2.071412560536581e-06, + "loss": 0.5357755422592163, + "step": 1490 + }, + { + "epoch": 3.1543340380549685, + "grad_norm": 1.7735508680343628, + "learning_rate": 2.0612427441669475e-06, + "loss": 0.5528465509414673, + "step": 1492 + }, + { + "epoch": 3.1585623678646932, + "grad_norm": 0.4385274648666382, + "learning_rate": 2.051114967986311e-06, + "loss": 0.39290380477905273, + "step": 1494 + }, + { + "epoch": 3.1627906976744184, + "grad_norm": 0.4222433567047119, + "learning_rate": 2.0410293558108074e-06, + "loss": 0.5798162817955017, + "step": 1496 + }, + { + "epoch": 3.1670190274841437, + "grad_norm": 0.4350104033946991, + "learning_rate": 2.0309860309410986e-06, + "loss": 0.8166102766990662, + "step": 1498 + }, + { + "epoch": 3.171247357293869, + "grad_norm": 0.7142273187637329, + "learning_rate": 2.0209851161608613e-06, + "loss": 0.8961682319641113, + "step": 1500 + }, + { + "epoch": 3.175475687103594, + "grad_norm": 0.7135263085365295, + "learning_rate": 2.011026733735301e-06, + "loss": 0.21471692621707916, + "step": 1502 + }, + { + "epoch": 3.1797040169133193, + "grad_norm": 0.8476817607879639, + "learning_rate": 2.001111005409643e-06, + "loss": 0.40848714113235474, + "step": 1504 + }, + { + "epoch": 3.1839323467230445, + "grad_norm": 0.5775097608566284, + "learning_rate": 1.9912380524076503e-06, + "loss": 0.582668125629425, + "step": 1506 + }, + { + "epoch": 3.1881606765327697, + "grad_norm": 0.8448871374130249, + "learning_rate": 1.981407995430136e-06, + "loss": 0.27065908908843994, + "step": 1508 + }, + { + "epoch": 3.192389006342495, + "grad_norm": 1.8615596294403076, + "learning_rate": 1.9716209546535e-06, + "loss": 0.19515842199325562, + "step": 1510 + }, + { + "epoch": 3.1966173361522197, + "grad_norm": 0.13960838317871094, + "learning_rate": 1.9618770497282487e-06, + "loss": 0.08642781525850296, + "step": 1512 + }, + { + "epoch": 3.200845665961945, + "grad_norm": 0.9059937000274658, + "learning_rate": 1.9521763997775294e-06, + "loss": 0.6139066219329834, + "step": 1514 + }, + { + "epoch": 3.20507399577167, + "grad_norm": 0.7437857985496521, + "learning_rate": 1.942519123395689e-06, + "loss": 0.7973853349685669, + "step": 1516 + }, + { + "epoch": 3.2093023255813953, + "grad_norm": 0.5043436288833618, + "learning_rate": 1.932905338646812e-06, + "loss": 0.45158812403678894, + "step": 1518 + }, + { + "epoch": 3.2135306553911205, + "grad_norm": 0.42636993527412415, + "learning_rate": 1.9233351630632747e-06, + "loss": 0.6722981929779053, + "step": 1520 + }, + { + "epoch": 3.2177589852008457, + "grad_norm": 0.6057955622673035, + "learning_rate": 1.9138087136443227e-06, + "loss": 0.7198382019996643, + "step": 1522 + }, + { + "epoch": 3.221987315010571, + "grad_norm": 1.310678482055664, + "learning_rate": 1.904326106854626e-06, + "loss": 0.5877378582954407, + "step": 1524 + }, + { + "epoch": 3.226215644820296, + "grad_norm": 0.6297647356987, + "learning_rate": 1.8948874586228622e-06, + "loss": 0.7359797358512878, + "step": 1526 + }, + { + "epoch": 3.2304439746300213, + "grad_norm": 1.018224835395813, + "learning_rate": 1.8854928843402987e-06, + "loss": 0.10719530284404755, + "step": 1528 + }, + { + "epoch": 3.234672304439746, + "grad_norm": 1.3252662420272827, + "learning_rate": 1.8761424988593796e-06, + "loss": 0.6593045592308044, + "step": 1530 + }, + { + "epoch": 3.2389006342494713, + "grad_norm": 0.8495606780052185, + "learning_rate": 1.8668364164923248e-06, + "loss": 0.6540462970733643, + "step": 1532 + }, + { + "epoch": 3.2431289640591965, + "grad_norm": 0.49407264590263367, + "learning_rate": 1.8575747510097294e-06, + "loss": 0.752964973449707, + "step": 1534 + }, + { + "epoch": 3.2473572938689217, + "grad_norm": 0.6842861771583557, + "learning_rate": 1.8483576156391747e-06, + "loss": 0.4958687722682953, + "step": 1536 + }, + { + "epoch": 3.251585623678647, + "grad_norm": 0.42404472827911377, + "learning_rate": 1.8391851230638443e-06, + "loss": 0.4690820872783661, + "step": 1538 + }, + { + "epoch": 3.255813953488372, + "grad_norm": 0.5655159950256348, + "learning_rate": 1.8300573854211452e-06, + "loss": 0.45219868421554565, + "step": 1540 + }, + { + "epoch": 3.2600422832980973, + "grad_norm": 1.4673653841018677, + "learning_rate": 1.8209745143013363e-06, + "loss": 0.7019790410995483, + "step": 1542 + }, + { + "epoch": 3.2642706131078225, + "grad_norm": 0.46494677662849426, + "learning_rate": 1.811936620746167e-06, + "loss": 0.8156011700630188, + "step": 1544 + }, + { + "epoch": 3.2684989429175477, + "grad_norm": 0.9783400297164917, + "learning_rate": 1.8029438152475159e-06, + "loss": 0.6611717343330383, + "step": 1546 + }, + { + "epoch": 3.2727272727272725, + "grad_norm": 0.7217451930046082, + "learning_rate": 1.7939962077460448e-06, + "loss": 0.21558183431625366, + "step": 1548 + }, + { + "epoch": 3.276955602536998, + "grad_norm": 1.123657464981079, + "learning_rate": 1.785093907629849e-06, + "loss": 0.5833519697189331, + "step": 1550 + }, + { + "epoch": 3.281183932346723, + "grad_norm": 0.4905749559402466, + "learning_rate": 1.7762370237331233e-06, + "loss": 0.7861868739128113, + "step": 1552 + }, + { + "epoch": 3.285412262156448, + "grad_norm": 0.03765023127198219, + "learning_rate": 1.7674256643348348e-06, + "loss": 0.07963511347770691, + "step": 1554 + }, + { + "epoch": 3.2896405919661733, + "grad_norm": 0.4575358033180237, + "learning_rate": 1.7586599371573887e-06, + "loss": 0.7299337983131409, + "step": 1556 + }, + { + "epoch": 3.2938689217758985, + "grad_norm": 0.48414161801338196, + "learning_rate": 1.7499399493653225e-06, + "loss": 0.42319798469543457, + "step": 1558 + }, + { + "epoch": 3.2980972515856237, + "grad_norm": 1.563560128211975, + "learning_rate": 1.7412658075639921e-06, + "loss": 0.8052334189414978, + "step": 1560 + }, + { + "epoch": 3.302325581395349, + "grad_norm": 7.0854291915893555, + "learning_rate": 1.7326376177982635e-06, + "loss": 0.5969507694244385, + "step": 1562 + }, + { + "epoch": 3.306553911205074, + "grad_norm": 1.6520373821258545, + "learning_rate": 1.7240554855512225e-06, + "loss": 0.3964461088180542, + "step": 1564 + }, + { + "epoch": 3.3107822410147993, + "grad_norm": 0.7322857975959778, + "learning_rate": 1.7155195157428878e-06, + "loss": 0.4578491747379303, + "step": 1566 + }, + { + "epoch": 3.3150105708245245, + "grad_norm": 0.13350917398929596, + "learning_rate": 1.707029812728916e-06, + "loss": 0.06760585308074951, + "step": 1568 + }, + { + "epoch": 3.3192389006342493, + "grad_norm": 0.4783739447593689, + "learning_rate": 1.6985864802993366e-06, + "loss": 0.44442859292030334, + "step": 1570 + }, + { + "epoch": 3.3234672304439745, + "grad_norm": 1.5828505754470825, + "learning_rate": 1.690189621677284e-06, + "loss": 0.5903767943382263, + "step": 1572 + }, + { + "epoch": 3.3276955602536997, + "grad_norm": 1.0708613395690918, + "learning_rate": 1.681839339517728e-06, + "loss": 0.1087808609008789, + "step": 1574 + }, + { + "epoch": 3.331923890063425, + "grad_norm": 2.037487030029297, + "learning_rate": 1.67353573590622e-06, + "loss": 0.24053213000297546, + "step": 1576 + }, + { + "epoch": 3.33615221987315, + "grad_norm": 1.0798691511154175, + "learning_rate": 1.6652789123576513e-06, + "loss": 0.43288537859916687, + "step": 1578 + }, + { + "epoch": 3.3403805496828753, + "grad_norm": 0.34591177105903625, + "learning_rate": 1.6570689698150094e-06, + "loss": 0.456515908241272, + "step": 1580 + }, + { + "epoch": 3.3446088794926006, + "grad_norm": 0.5088324546813965, + "learning_rate": 1.6489060086481373e-06, + "loss": 0.839871346950531, + "step": 1582 + }, + { + "epoch": 3.3488372093023258, + "grad_norm": 0.8262994289398193, + "learning_rate": 1.6407901286525152e-06, + "loss": 0.5351448655128479, + "step": 1584 + }, + { + "epoch": 3.353065539112051, + "grad_norm": 0.48140838742256165, + "learning_rate": 1.6327214290480382e-06, + "loss": 0.5584917664527893, + "step": 1586 + }, + { + "epoch": 3.3572938689217757, + "grad_norm": 1.4282252788543701, + "learning_rate": 1.6247000084778006e-06, + "loss": 0.8573344349861145, + "step": 1588 + }, + { + "epoch": 3.361522198731501, + "grad_norm": 0.512168824672699, + "learning_rate": 1.6167259650068887e-06, + "loss": 0.31845611333847046, + "step": 1590 + }, + { + "epoch": 3.365750528541226, + "grad_norm": 0.6738854050636292, + "learning_rate": 1.6087993961211904e-06, + "loss": 0.6181653738021851, + "step": 1592 + }, + { + "epoch": 3.3699788583509513, + "grad_norm": 0.3983571529388428, + "learning_rate": 1.6009203987261957e-06, + "loss": 0.2663561701774597, + "step": 1594 + }, + { + "epoch": 3.3742071881606766, + "grad_norm": 0.41974449157714844, + "learning_rate": 1.593089069145809e-06, + "loss": 0.7632195353507996, + "step": 1596 + }, + { + "epoch": 3.3784355179704018, + "grad_norm": 1.6651194095611572, + "learning_rate": 1.5853055031211844e-06, + "loss": 0.5548272132873535, + "step": 1598 + }, + { + "epoch": 3.382663847780127, + "grad_norm": 0.4871348738670349, + "learning_rate": 1.577569795809544e-06, + "loss": 0.2500655949115753, + "step": 1600 + }, + { + "epoch": 3.386892177589852, + "grad_norm": 0.6851543188095093, + "learning_rate": 1.5698820417830137e-06, + "loss": 0.5866137742996216, + "step": 1602 + }, + { + "epoch": 3.3911205073995774, + "grad_norm": 0.697858989238739, + "learning_rate": 1.5622423350274784e-06, + "loss": 0.8283005952835083, + "step": 1604 + }, + { + "epoch": 3.395348837209302, + "grad_norm": 0.5530416965484619, + "learning_rate": 1.5546507689414217e-06, + "loss": 0.877478837966919, + "step": 1606 + }, + { + "epoch": 3.3995771670190273, + "grad_norm": 1.7913435697555542, + "learning_rate": 1.5471074363347894e-06, + "loss": 0.4317421615123749, + "step": 1608 + }, + { + "epoch": 3.4038054968287526, + "grad_norm": 0.49932125210762024, + "learning_rate": 1.539612429427852e-06, + "loss": 0.8128949999809265, + "step": 1610 + }, + { + "epoch": 3.4080338266384778, + "grad_norm": 0.6740911602973938, + "learning_rate": 1.5321658398500808e-06, + "loss": 0.6489126086235046, + "step": 1612 + }, + { + "epoch": 3.412262156448203, + "grad_norm": 0.800811767578125, + "learning_rate": 1.5247677586390255e-06, + "loss": 0.5815157294273376, + "step": 1614 + }, + { + "epoch": 3.416490486257928, + "grad_norm": 0.8403217196464539, + "learning_rate": 1.5174182762392007e-06, + "loss": 0.7984689474105835, + "step": 1616 + }, + { + "epoch": 3.4207188160676534, + "grad_norm": 0.42191073298454285, + "learning_rate": 1.5101174825009815e-06, + "loss": 0.8098374009132385, + "step": 1618 + }, + { + "epoch": 3.4249471458773786, + "grad_norm": 0.44885221123695374, + "learning_rate": 1.5028654666795042e-06, + "loss": 0.7655125260353088, + "step": 1620 + }, + { + "epoch": 3.429175475687104, + "grad_norm": 2.0987799167633057, + "learning_rate": 1.495662317433576e-06, + "loss": 0.44010061025619507, + "step": 1622 + }, + { + "epoch": 3.4334038054968286, + "grad_norm": 1.0775363445281982, + "learning_rate": 1.4885081228245909e-06, + "loss": 0.7359012365341187, + "step": 1624 + }, + { + "epoch": 3.4376321353065538, + "grad_norm": 0.553027868270874, + "learning_rate": 1.4814029703154515e-06, + "loss": 0.5531986951828003, + "step": 1626 + }, + { + "epoch": 3.441860465116279, + "grad_norm": 0.4667411148548126, + "learning_rate": 1.4743469467695017e-06, + "loss": 0.8393853902816772, + "step": 1628 + }, + { + "epoch": 3.446088794926004, + "grad_norm": 0.9733679294586182, + "learning_rate": 1.4673401384494651e-06, + "loss": 0.3643987476825714, + "step": 1630 + }, + { + "epoch": 3.4503171247357294, + "grad_norm": 0.9814094305038452, + "learning_rate": 1.4603826310163874e-06, + "loss": 0.6899866461753845, + "step": 1632 + }, + { + "epoch": 3.4545454545454546, + "grad_norm": 1.868240475654602, + "learning_rate": 1.453474509528593e-06, + "loss": 0.45263615250587463, + "step": 1634 + }, + { + "epoch": 3.45877378435518, + "grad_norm": 0.41196560859680176, + "learning_rate": 1.446615858440642e-06, + "loss": 0.7376907467842102, + "step": 1636 + }, + { + "epoch": 3.463002114164905, + "grad_norm": 0.11359843611717224, + "learning_rate": 1.4398067616022993e-06, + "loss": 0.2635776996612549, + "step": 1638 + }, + { + "epoch": 3.46723044397463, + "grad_norm": 0.45712724328041077, + "learning_rate": 1.43304730225751e-06, + "loss": 0.4622797966003418, + "step": 1640 + }, + { + "epoch": 3.471458773784355, + "grad_norm": 0.48741745948791504, + "learning_rate": 1.4263375630433802e-06, + "loss": 0.08808047324419022, + "step": 1642 + }, + { + "epoch": 3.47568710359408, + "grad_norm": 1.0607576370239258, + "learning_rate": 1.4196776259891665e-06, + "loss": 0.7692833542823792, + "step": 1644 + }, + { + "epoch": 3.4799154334038054, + "grad_norm": 0.7436749339103699, + "learning_rate": 1.4130675725152762e-06, + "loss": 0.635974645614624, + "step": 1646 + }, + { + "epoch": 3.4841437632135306, + "grad_norm": 0.09212988615036011, + "learning_rate": 1.4065074834322677e-06, + "loss": 0.4444892704486847, + "step": 1648 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 1.2450265884399414, + "learning_rate": 1.3999974389398669e-06, + "loss": 0.43537190556526184, + "step": 1650 + }, + { + "epoch": 3.492600422832981, + "grad_norm": 0.8162301778793335, + "learning_rate": 1.3935375186259817e-06, + "loss": 0.4205917418003082, + "step": 1652 + }, + { + "epoch": 3.496828752642706, + "grad_norm": 1.5210318565368652, + "learning_rate": 1.3871278014657332e-06, + "loss": 0.6245290637016296, + "step": 1654 + }, + { + "epoch": 3.5010570824524314, + "grad_norm": 0.7625864148139954, + "learning_rate": 1.3807683658204915e-06, + "loss": 0.7488887906074524, + "step": 1656 + }, + { + "epoch": 3.5052854122621566, + "grad_norm": 0.13264936208724976, + "learning_rate": 1.3744592894369093e-06, + "loss": 0.4216771125793457, + "step": 1658 + }, + { + "epoch": 3.5095137420718814, + "grad_norm": 0.523833155632019, + "learning_rate": 1.3682006494459806e-06, + "loss": 0.5845276713371277, + "step": 1660 + }, + { + "epoch": 3.513742071881607, + "grad_norm": 1.078977108001709, + "learning_rate": 1.361992522362096e-06, + "loss": 0.49822723865509033, + "step": 1662 + }, + { + "epoch": 3.517970401691332, + "grad_norm": 0.45897868275642395, + "learning_rate": 1.355834984082101e-06, + "loss": 0.7767106294631958, + "step": 1664 + }, + { + "epoch": 3.522198731501057, + "grad_norm": 0.9096782803535461, + "learning_rate": 1.3497281098843746e-06, + "loss": 0.763041079044342, + "step": 1666 + }, + { + "epoch": 3.526427061310782, + "grad_norm": 0.5180878639221191, + "learning_rate": 1.34367197442791e-06, + "loss": 0.06131262704730034, + "step": 1668 + }, + { + "epoch": 3.5306553911205074, + "grad_norm": 1.4329968690872192, + "learning_rate": 1.3376666517513931e-06, + "loss": 0.6485794186592102, + "step": 1670 + }, + { + "epoch": 3.5348837209302326, + "grad_norm": 0.6127661466598511, + "learning_rate": 1.3317122152723078e-06, + "loss": 0.4568518400192261, + "step": 1672 + }, + { + "epoch": 3.539112050739958, + "grad_norm": 0.5207194685935974, + "learning_rate": 1.3258087377860323e-06, + "loss": 0.6887850761413574, + "step": 1674 + }, + { + "epoch": 3.543340380549683, + "grad_norm": 0.915372371673584, + "learning_rate": 1.3199562914649519e-06, + "loss": 0.4163953363895416, + "step": 1676 + }, + { + "epoch": 3.547568710359408, + "grad_norm": 0.4881134331226349, + "learning_rate": 1.314154947857571e-06, + "loss": 0.7600098252296448, + "step": 1678 + }, + { + "epoch": 3.5517970401691334, + "grad_norm": 0.41513165831565857, + "learning_rate": 1.3084047778876487e-06, + "loss": 0.4089525640010834, + "step": 1680 + }, + { + "epoch": 3.556025369978858, + "grad_norm": 0.5687722563743591, + "learning_rate": 1.3027058518533228e-06, + "loss": 0.428451806306839, + "step": 1682 + }, + { + "epoch": 3.5602536997885834, + "grad_norm": 0.6440165638923645, + "learning_rate": 1.2970582394262516e-06, + "loss": 0.4509245455265045, + "step": 1684 + }, + { + "epoch": 3.5644820295983086, + "grad_norm": 0.774116575717926, + "learning_rate": 1.291462009650768e-06, + "loss": 0.06551752984523773, + "step": 1686 + }, + { + "epoch": 3.568710359408034, + "grad_norm": 0.8542079329490662, + "learning_rate": 1.2859172309430285e-06, + "loss": 0.5117007493972778, + "step": 1688 + }, + { + "epoch": 3.572938689217759, + "grad_norm": 0.12220869213342667, + "learning_rate": 1.280423971090181e-06, + "loss": 0.2072942852973938, + "step": 1690 + }, + { + "epoch": 3.5771670190274842, + "grad_norm": 0.5463467240333557, + "learning_rate": 1.2749822972495333e-06, + "loss": 0.3216285705566406, + "step": 1692 + }, + { + "epoch": 3.5813953488372094, + "grad_norm": 3.1010591983795166, + "learning_rate": 1.2695922759477356e-06, + "loss": 0.5064961314201355, + "step": 1694 + }, + { + "epoch": 3.585623678646934, + "grad_norm": 1.5337337255477905, + "learning_rate": 1.2642539730799622e-06, + "loss": 0.27248242497444153, + "step": 1696 + }, + { + "epoch": 3.58985200845666, + "grad_norm": 1.0044714212417603, + "learning_rate": 1.2589674539091115e-06, + "loss": 0.7509093880653381, + "step": 1698 + }, + { + "epoch": 3.5940803382663846, + "grad_norm": 0.4929828941822052, + "learning_rate": 1.2537327830650042e-06, + "loss": 0.881374716758728, + "step": 1700 + }, + { + "epoch": 3.59830866807611, + "grad_norm": 1.4503240585327148, + "learning_rate": 1.248550024543594e-06, + "loss": 0.6183159351348877, + "step": 1702 + }, + { + "epoch": 3.602536997885835, + "grad_norm": 1.029107928276062, + "learning_rate": 1.2434192417061862e-06, + "loss": 0.19822384417057037, + "step": 1704 + }, + { + "epoch": 3.6067653276955602, + "grad_norm": 1.5719152688980103, + "learning_rate": 1.2383404972786625e-06, + "loss": 0.4333816170692444, + "step": 1706 + }, + { + "epoch": 3.6109936575052854, + "grad_norm": 3.715841054916382, + "learning_rate": 1.2333138533507139e-06, + "loss": 0.6227849721908569, + "step": 1708 + }, + { + "epoch": 3.6152219873150107, + "grad_norm": 0.8299279808998108, + "learning_rate": 1.2283393713750816e-06, + "loss": 0.39489880204200745, + "step": 1710 + }, + { + "epoch": 3.619450317124736, + "grad_norm": 1.2052602767944336, + "learning_rate": 1.2234171121668069e-06, + "loss": 0.40268248319625854, + "step": 1712 + }, + { + "epoch": 3.6236786469344606, + "grad_norm": 0.7987406849861145, + "learning_rate": 1.2185471359024855e-06, + "loss": 0.28327876329421997, + "step": 1714 + }, + { + "epoch": 3.6279069767441863, + "grad_norm": 2.5767972469329834, + "learning_rate": 1.2137295021195341e-06, + "loss": 0.5251152515411377, + "step": 1716 + }, + { + "epoch": 3.632135306553911, + "grad_norm": 0.2198551744222641, + "learning_rate": 1.2089642697154613e-06, + "loss": 0.4255867898464203, + "step": 1718 + }, + { + "epoch": 3.6363636363636362, + "grad_norm": 0.9753926992416382, + "learning_rate": 1.2042514969471475e-06, + "loss": 0.2627681493759155, + "step": 1720 + }, + { + "epoch": 3.6405919661733614, + "grad_norm": 0.49667197465896606, + "learning_rate": 1.1995912414301329e-06, + "loss": 0.6128677725791931, + "step": 1722 + }, + { + "epoch": 3.6448202959830867, + "grad_norm": 0.6341894865036011, + "learning_rate": 1.1949835601379144e-06, + "loss": 0.5457699298858643, + "step": 1724 + }, + { + "epoch": 3.649048625792812, + "grad_norm": 0.4836573004722595, + "learning_rate": 1.1904285094012463e-06, + "loss": 0.8371782898902893, + "step": 1726 + }, + { + "epoch": 3.653276955602537, + "grad_norm": 0.4500487446784973, + "learning_rate": 1.185926144907454e-06, + "loss": 0.8152235746383667, + "step": 1728 + }, + { + "epoch": 3.6575052854122623, + "grad_norm": 2.2880728244781494, + "learning_rate": 1.1814765216997522e-06, + "loss": 0.6790116429328918, + "step": 1730 + }, + { + "epoch": 3.6617336152219875, + "grad_norm": 0.9643025994300842, + "learning_rate": 1.1770796941765727e-06, + "loss": 0.6797136068344116, + "step": 1732 + }, + { + "epoch": 3.6659619450317127, + "grad_norm": 0.9523893594741821, + "learning_rate": 1.1727357160908978e-06, + "loss": 0.6945903897285461, + "step": 1734 + }, + { + "epoch": 3.6701902748414374, + "grad_norm": 0.4832755923271179, + "learning_rate": 1.1684446405496052e-06, + "loss": 0.7960948944091797, + "step": 1736 + }, + { + "epoch": 3.6744186046511627, + "grad_norm": 1.012386441230774, + "learning_rate": 1.1642065200128179e-06, + "loss": 0.07248520851135254, + "step": 1738 + }, + { + "epoch": 3.678646934460888, + "grad_norm": 0.5523192882537842, + "learning_rate": 1.1600214062932613e-06, + "loss": 0.36555737257003784, + "step": 1740 + }, + { + "epoch": 3.682875264270613, + "grad_norm": 0.8524115085601807, + "learning_rate": 1.1558893505556315e-06, + "loss": 0.19492213428020477, + "step": 1742 + }, + { + "epoch": 3.6871035940803383, + "grad_norm": 0.46189063787460327, + "learning_rate": 1.1518104033159715e-06, + "loss": 0.7509604692459106, + "step": 1744 + }, + { + "epoch": 3.6913319238900635, + "grad_norm": 0.14885587990283966, + "learning_rate": 1.1477846144410493e-06, + "loss": 0.39124566316604614, + "step": 1746 + }, + { + "epoch": 3.6955602536997887, + "grad_norm": 3.6966612339019775, + "learning_rate": 1.1438120331477506e-06, + "loss": 0.2900627851486206, + "step": 1748 + }, + { + "epoch": 3.699788583509514, + "grad_norm": 0.8838816285133362, + "learning_rate": 1.1398927080024796e-06, + "loss": 0.9213196039199829, + "step": 1750 + }, + { + "epoch": 3.704016913319239, + "grad_norm": 0.5080142021179199, + "learning_rate": 1.1360266869205593e-06, + "loss": 0.2801329791545868, + "step": 1752 + }, + { + "epoch": 3.708245243128964, + "grad_norm": 0.6076149344444275, + "learning_rate": 1.1322140171656507e-06, + "loss": 0.4164304733276367, + "step": 1754 + }, + { + "epoch": 3.712473572938689, + "grad_norm": 0.5222283601760864, + "learning_rate": 1.1284547453491734e-06, + "loss": 0.5692883729934692, + "step": 1756 + }, + { + "epoch": 3.7167019027484143, + "grad_norm": 0.7487146258354187, + "learning_rate": 1.1247489174297372e-06, + "loss": 0.3346913456916809, + "step": 1758 + }, + { + "epoch": 3.7209302325581395, + "grad_norm": 0.10256338119506836, + "learning_rate": 1.1210965787125743e-06, + "loss": 0.3705770969390869, + "step": 1760 + }, + { + "epoch": 3.7251585623678647, + "grad_norm": 0.8462153077125549, + "learning_rate": 1.1174977738489955e-06, + "loss": 0.7353273630142212, + "step": 1762 + }, + { + "epoch": 3.72938689217759, + "grad_norm": 0.5337866544723511, + "learning_rate": 1.113952546835835e-06, + "loss": 0.4264756441116333, + "step": 1764 + }, + { + "epoch": 3.733615221987315, + "grad_norm": 1.753082513809204, + "learning_rate": 1.1104609410149174e-06, + "loss": 0.6098082065582275, + "step": 1766 + }, + { + "epoch": 3.7378435517970403, + "grad_norm": 0.5783655047416687, + "learning_rate": 1.1070229990725262e-06, + "loss": 0.6269397139549255, + "step": 1768 + }, + { + "epoch": 3.7420718816067655, + "grad_norm": 0.12172426283359528, + "learning_rate": 1.1036387630388841e-06, + "loss": 0.1750326007604599, + "step": 1770 + }, + { + "epoch": 3.7463002114164903, + "grad_norm": 0.9122955799102783, + "learning_rate": 1.100308274287636e-06, + "loss": 0.425304651260376, + "step": 1772 + }, + { + "epoch": 3.7505285412262155, + "grad_norm": 1.2123184204101562, + "learning_rate": 1.0970315735353434e-06, + "loss": 0.8994868993759155, + "step": 1774 + }, + { + "epoch": 3.7547568710359407, + "grad_norm": 0.6799372434616089, + "learning_rate": 1.0938087008409909e-06, + "loss": 0.7489012479782104, + "step": 1776 + }, + { + "epoch": 3.758985200845666, + "grad_norm": 0.42199215292930603, + "learning_rate": 1.0906396956054919e-06, + "loss": 0.8365127444267273, + "step": 1778 + }, + { + "epoch": 3.763213530655391, + "grad_norm": 1.1166168451309204, + "learning_rate": 1.0875245965712077e-06, + "loss": 0.8776699304580688, + "step": 1780 + }, + { + "epoch": 3.7674418604651163, + "grad_norm": 0.7865721583366394, + "learning_rate": 1.0844634418214763e-06, + "loss": 0.6536276340484619, + "step": 1782 + }, + { + "epoch": 3.7716701902748415, + "grad_norm": 0.9009637236595154, + "learning_rate": 1.0814562687801448e-06, + "loss": 0.13780006766319275, + "step": 1784 + }, + { + "epoch": 3.7758985200845667, + "grad_norm": 0.9369373321533203, + "learning_rate": 1.0785031142111103e-06, + "loss": 0.5146877765655518, + "step": 1786 + }, + { + "epoch": 3.780126849894292, + "grad_norm": 0.9798800945281982, + "learning_rate": 1.075604014217875e-06, + "loss": 0.4336180090904236, + "step": 1788 + }, + { + "epoch": 3.7843551797040167, + "grad_norm": 0.9238928556442261, + "learning_rate": 1.0727590042431009e-06, + "loss": 0.0971670001745224, + "step": 1790 + }, + { + "epoch": 3.7885835095137423, + "grad_norm": 0.16246433556079865, + "learning_rate": 1.0699681190681777e-06, + "loss": 0.6635580658912659, + "step": 1792 + }, + { + "epoch": 3.792811839323467, + "grad_norm": 0.9584662914276123, + "learning_rate": 1.067231392812798e-06, + "loss": 0.7299457788467407, + "step": 1794 + }, + { + "epoch": 3.7970401691331923, + "grad_norm": 0.7439003586769104, + "learning_rate": 1.0645488589345396e-06, + "loss": 0.4258429706096649, + "step": 1796 + }, + { + "epoch": 3.8012684989429175, + "grad_norm": 3.486553192138672, + "learning_rate": 1.0619205502284574e-06, + "loss": 0.6950887441635132, + "step": 1798 + }, + { + "epoch": 3.8054968287526427, + "grad_norm": 0.473550021648407, + "learning_rate": 1.0593464988266811e-06, + "loss": 0.8141056299209595, + "step": 1800 + }, + { + "epoch": 3.809725158562368, + "grad_norm": 0.7902247905731201, + "learning_rate": 1.0568267361980225e-06, + "loss": 0.47772112488746643, + "step": 1802 + }, + { + "epoch": 3.813953488372093, + "grad_norm": 1.7617413997650146, + "learning_rate": 1.0543612931475935e-06, + "loss": 0.8090075254440308, + "step": 1804 + }, + { + "epoch": 3.8181818181818183, + "grad_norm": 0.5060732960700989, + "learning_rate": 1.0519501998164247e-06, + "loss": 0.833554744720459, + "step": 1806 + }, + { + "epoch": 3.822410147991543, + "grad_norm": 1.0836882591247559, + "learning_rate": 1.049593485681101e-06, + "loss": 0.12246610224246979, + "step": 1808 + }, + { + "epoch": 3.8266384778012688, + "grad_norm": 0.7263925075531006, + "learning_rate": 1.0472911795533997e-06, + "loss": 0.8387416005134583, + "step": 1810 + }, + { + "epoch": 3.8308668076109935, + "grad_norm": 1.3152225017547607, + "learning_rate": 1.045043309579938e-06, + "loss": 0.7521619200706482, + "step": 1812 + }, + { + "epoch": 3.8350951374207187, + "grad_norm": 0.522961437702179, + "learning_rate": 1.0428499032418297e-06, + "loss": 0.626377522945404, + "step": 1814 + }, + { + "epoch": 3.839323467230444, + "grad_norm": 1.2180602550506592, + "learning_rate": 1.0407109873543484e-06, + "loss": 0.46516966819763184, + "step": 1816 + }, + { + "epoch": 3.843551797040169, + "grad_norm": 3.62945818901062, + "learning_rate": 1.0386265880665996e-06, + "loss": 0.4338098466396332, + "step": 1818 + }, + { + "epoch": 3.8477801268498943, + "grad_norm": 0.4239257574081421, + "learning_rate": 1.036596730861203e-06, + "loss": 0.42830467224121094, + "step": 1820 + }, + { + "epoch": 3.8520084566596196, + "grad_norm": 0.26361945271492004, + "learning_rate": 1.0346214405539783e-06, + "loss": 0.5324674844741821, + "step": 1822 + }, + { + "epoch": 3.8562367864693448, + "grad_norm": 0.49896594882011414, + "learning_rate": 1.0327007412936419e-06, + "loss": 0.49197185039520264, + "step": 1824 + }, + { + "epoch": 3.8604651162790695, + "grad_norm": 0.6829991936683655, + "learning_rate": 1.0308346565615158e-06, + "loss": 0.41304147243499756, + "step": 1826 + }, + { + "epoch": 3.864693446088795, + "grad_norm": 0.5371273756027222, + "learning_rate": 1.0290232091712337e-06, + "loss": 0.8393287658691406, + "step": 1828 + }, + { + "epoch": 3.86892177589852, + "grad_norm": 0.5705682635307312, + "learning_rate": 1.0272664212684687e-06, + "loss": 0.8210970163345337, + "step": 1830 + }, + { + "epoch": 3.873150105708245, + "grad_norm": 0.8254919648170471, + "learning_rate": 1.025564314330659e-06, + "loss": 0.4440838396549225, + "step": 1832 + }, + { + "epoch": 3.8773784355179703, + "grad_norm": 0.5358226299285889, + "learning_rate": 1.0239169091667447e-06, + "loss": 0.40913182497024536, + "step": 1834 + }, + { + "epoch": 3.8816067653276956, + "grad_norm": 0.16638346016407013, + "learning_rate": 1.022324225916917e-06, + "loss": 0.4094028174877167, + "step": 1836 + }, + { + "epoch": 3.8858350951374208, + "grad_norm": 0.09130455553531647, + "learning_rate": 1.0207862840523672e-06, + "loss": 0.41299504041671753, + "step": 1838 + }, + { + "epoch": 3.890063424947146, + "grad_norm": 0.7351621389389038, + "learning_rate": 1.0193031023750534e-06, + "loss": 0.6642860174179077, + "step": 1840 + }, + { + "epoch": 3.894291754756871, + "grad_norm": 0.5131246447563171, + "learning_rate": 1.017874699017468e-06, + "loss": 0.7378905415534973, + "step": 1842 + }, + { + "epoch": 3.898520084566596, + "grad_norm": 0.7365778684616089, + "learning_rate": 1.0165010914424142e-06, + "loss": 0.45080387592315674, + "step": 1844 + }, + { + "epoch": 3.9027484143763216, + "grad_norm": 0.7749892473220825, + "learning_rate": 1.015182296442798e-06, + "loss": 0.43491557240486145, + "step": 1846 + }, + { + "epoch": 3.9069767441860463, + "grad_norm": 1.5793519020080566, + "learning_rate": 1.0139183301414176e-06, + "loss": 0.07734798640012741, + "step": 1848 + }, + { + "epoch": 3.9112050739957716, + "grad_norm": 0.6272465586662292, + "learning_rate": 1.0127092079907682e-06, + "loss": 0.7316566705703735, + "step": 1850 + }, + { + "epoch": 3.9154334038054968, + "grad_norm": 0.21347488462924957, + "learning_rate": 1.0115549447728551e-06, + "loss": 0.44612327218055725, + "step": 1852 + }, + { + "epoch": 3.919661733615222, + "grad_norm": 0.4135867953300476, + "learning_rate": 1.0104555545990088e-06, + "loss": 0.7245345115661621, + "step": 1854 + }, + { + "epoch": 3.923890063424947, + "grad_norm": 0.5526317358016968, + "learning_rate": 1.009411050909716e-06, + "loss": 0.39463159441947937, + "step": 1856 + }, + { + "epoch": 3.9281183932346724, + "grad_norm": 0.772693932056427, + "learning_rate": 1.008421446474455e-06, + "loss": 0.6087676286697388, + "step": 1858 + }, + { + "epoch": 3.9323467230443976, + "grad_norm": 1.7492222785949707, + "learning_rate": 1.0074867533915367e-06, + "loss": 0.26687589287757874, + "step": 1860 + }, + { + "epoch": 3.9365750528541223, + "grad_norm": 0.9755433201789856, + "learning_rate": 1.0066069830879593e-06, + "loss": 0.5821423530578613, + "step": 1862 + }, + { + "epoch": 3.940803382663848, + "grad_norm": 0.6028100252151489, + "learning_rate": 1.0057821463192687e-06, + "loss": 0.7636744379997253, + "step": 1864 + }, + { + "epoch": 3.9450317124735728, + "grad_norm": 0.4274957776069641, + "learning_rate": 1.005012253169426e-06, + "loss": 0.662148118019104, + "step": 1866 + }, + { + "epoch": 3.949260042283298, + "grad_norm": 0.029392855241894722, + "learning_rate": 1.0042973130506839e-06, + "loss": 0.0017496227519586682, + "step": 1868 + }, + { + "epoch": 3.953488372093023, + "grad_norm": 0.5853798389434814, + "learning_rate": 1.003637334703472e-06, + "loss": 0.8540186882019043, + "step": 1870 + }, + { + "epoch": 3.9577167019027484, + "grad_norm": 0.5099213719367981, + "learning_rate": 1.0030323261962922e-06, + "loss": 0.5266919136047363, + "step": 1872 + }, + { + "epoch": 3.9619450317124736, + "grad_norm": 0.8958723545074463, + "learning_rate": 1.002482294925615e-06, + "loss": 0.3962560296058655, + "step": 1874 + }, + { + "epoch": 3.966173361522199, + "grad_norm": 0.45622918009757996, + "learning_rate": 1.0019872476157952e-06, + "loss": 0.7846705913543701, + "step": 1876 + }, + { + "epoch": 3.970401691331924, + "grad_norm": 0.7758667469024658, + "learning_rate": 1.0015471903189834e-06, + "loss": 0.0849628672003746, + "step": 1878 + }, + { + "epoch": 3.974630021141649, + "grad_norm": 0.5936379432678223, + "learning_rate": 1.0011621284150583e-06, + "loss": 0.38633227348327637, + "step": 1880 + }, + { + "epoch": 3.9788583509513744, + "grad_norm": 0.7230780124664307, + "learning_rate": 1.000832066611556e-06, + "loss": 0.6690049767494202, + "step": 1882 + }, + { + "epoch": 3.983086680761099, + "grad_norm": 0.588121771812439, + "learning_rate": 1.0005570089436137e-06, + "loss": 0.890869677066803, + "step": 1884 + }, + { + "epoch": 3.9873150105708244, + "grad_norm": 1.179721713066101, + "learning_rate": 1.0003369587739234e-06, + "loss": 0.5820091366767883, + "step": 1886 + }, + { + "epoch": 3.9915433403805496, + "grad_norm": 0.440155029296875, + "learning_rate": 1.0001719187926854e-06, + "loss": 0.7724121809005737, + "step": 1888 + }, + { + "epoch": 3.995771670190275, + "grad_norm": 0.4935588240623474, + "learning_rate": 1.0000618910175811e-06, + "loss": 0.7828623056411743, + "step": 1890 + }, + { + "epoch": 4.0, + "grad_norm": 0.7185866832733154, + "learning_rate": 1.000006876793743e-06, + "loss": 0.42325642704963684, + "step": 1892 + }, + { + "epoch": 4.0, + "step": 1892, + "total_flos": 3.554237146892075e+18, + "train_loss": 0.8299213015755204, + "train_runtime": 23513.1561, + "train_samples_per_second": 2.414, + "train_steps_per_second": 0.08 + } + ], + "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": 3.554237146892075e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}