diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,6357 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 500, + "global_step": 1804, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.004434589800443459, + "grad_norm": 4.369466304779053, + "learning_rate": 1.098901098901099e-07, + "loss": 1.865166187286377, + "step": 2 + }, + { + "epoch": 0.008869179600886918, + "grad_norm": 6.864917278289795, + "learning_rate": 3.296703296703297e-07, + "loss": 2.130244493484497, + "step": 4 + }, + { + "epoch": 0.013303769401330377, + "grad_norm": 3.4614903926849365, + "learning_rate": 5.494505494505495e-07, + "loss": 1.9028818607330322, + "step": 6 + }, + { + "epoch": 0.017738359201773836, + "grad_norm": 1.170404076576233, + "learning_rate": 7.692307692307694e-07, + "loss": 1.815584659576416, + "step": 8 + }, + { + "epoch": 0.022172949002217297, + "grad_norm": 4.44560432434082, + "learning_rate": 9.890109890109891e-07, + "loss": 1.616060733795166, + "step": 10 + }, + { + "epoch": 0.026607538802660754, + "grad_norm": 3.004225730895996, + "learning_rate": 1.2087912087912089e-06, + "loss": 2.0138046741485596, + "step": 12 + }, + { + "epoch": 0.031042128603104215, + "grad_norm": 6.603450298309326, + "learning_rate": 1.4285714285714286e-06, + "loss": 1.5741312503814697, + "step": 14 + }, + { + "epoch": 0.03547671840354767, + "grad_norm": 3.626671552658081, + "learning_rate": 1.6483516483516484e-06, + "loss": 1.1557213068008423, + "step": 16 + }, + { + "epoch": 0.03991130820399113, + "grad_norm": 1.9218900203704834, + "learning_rate": 1.8681318681318684e-06, + "loss": 1.167083978652954, + "step": 18 + }, + { + "epoch": 0.04434589800443459, + "grad_norm": 27.988502502441406, + "learning_rate": 2.0879120879120883e-06, + "loss": 1.2604156732559204, + "step": 20 + }, + { + "epoch": 0.04878048780487805, + "grad_norm": 1.9937968254089355, + "learning_rate": 2.307692307692308e-06, + "loss": 1.300764799118042, + "step": 22 + }, + { + "epoch": 0.05321507760532151, + "grad_norm": 1.050702691078186, + "learning_rate": 2.5274725274725274e-06, + "loss": 1.5187550783157349, + "step": 24 + }, + { + "epoch": 0.057649667405764965, + "grad_norm": 0.8804575204849243, + "learning_rate": 2.7472527472527476e-06, + "loss": 1.2839621305465698, + "step": 26 + }, + { + "epoch": 0.06208425720620843, + "grad_norm": 1.4461461305618286, + "learning_rate": 2.9670329670329673e-06, + "loss": 1.0896079540252686, + "step": 28 + }, + { + "epoch": 0.06651884700665188, + "grad_norm": 1.1735917329788208, + "learning_rate": 3.1868131868131867e-06, + "loss": 1.2101552486419678, + "step": 30 + }, + { + "epoch": 0.07095343680709534, + "grad_norm": 1.792823076248169, + "learning_rate": 3.406593406593407e-06, + "loss": 1.1142032146453857, + "step": 32 + }, + { + "epoch": 0.07538802660753881, + "grad_norm": 0.8939670920372009, + "learning_rate": 3.6263736263736266e-06, + "loss": 1.4834434986114502, + "step": 34 + }, + { + "epoch": 0.07982261640798226, + "grad_norm": 5.579226970672607, + "learning_rate": 3.846153846153847e-06, + "loss": 1.2127963304519653, + "step": 36 + }, + { + "epoch": 0.08425720620842572, + "grad_norm": 11.267181396484375, + "learning_rate": 4.065934065934066e-06, + "loss": 1.535402774810791, + "step": 38 + }, + { + "epoch": 0.08869179600886919, + "grad_norm": 1.9921720027923584, + "learning_rate": 4.2857142857142855e-06, + "loss": 0.7790605425834656, + "step": 40 + }, + { + "epoch": 0.09312638580931264, + "grad_norm": 1.5248279571533203, + "learning_rate": 4.505494505494506e-06, + "loss": 1.3986105918884277, + "step": 42 + }, + { + "epoch": 0.0975609756097561, + "grad_norm": 0.9806777238845825, + "learning_rate": 4.725274725274726e-06, + "loss": 1.3404041528701782, + "step": 44 + }, + { + "epoch": 0.10199556541019955, + "grad_norm": 1.6693533658981323, + "learning_rate": 4.945054945054946e-06, + "loss": 1.3681906461715698, + "step": 46 + }, + { + "epoch": 0.10643015521064302, + "grad_norm": 2.5410993099212646, + "learning_rate": 5.164835164835166e-06, + "loss": 1.6145933866500854, + "step": 48 + }, + { + "epoch": 0.11086474501108648, + "grad_norm": 1.414282202720642, + "learning_rate": 5.384615384615385e-06, + "loss": 1.4454115629196167, + "step": 50 + }, + { + "epoch": 0.11529933481152993, + "grad_norm": 1.7944461107254028, + "learning_rate": 5.604395604395605e-06, + "loss": 1.3703702688217163, + "step": 52 + }, + { + "epoch": 0.1197339246119734, + "grad_norm": 1.0293529033660889, + "learning_rate": 5.824175824175825e-06, + "loss": 1.368741750717163, + "step": 54 + }, + { + "epoch": 0.12416851441241686, + "grad_norm": 2.983600616455078, + "learning_rate": 6.043956043956044e-06, + "loss": 1.1135094165802002, + "step": 56 + }, + { + "epoch": 0.1286031042128603, + "grad_norm": 1.0177669525146484, + "learning_rate": 6.2637362637362645e-06, + "loss": 1.3418025970458984, + "step": 58 + }, + { + "epoch": 0.13303769401330376, + "grad_norm": 0.7596590518951416, + "learning_rate": 6.483516483516485e-06, + "loss": 1.3871124982833862, + "step": 60 + }, + { + "epoch": 0.13747228381374724, + "grad_norm": 1.77037513256073, + "learning_rate": 6.703296703296703e-06, + "loss": 1.336590051651001, + "step": 62 + }, + { + "epoch": 0.1419068736141907, + "grad_norm": 1.6162333488464355, + "learning_rate": 6.923076923076923e-06, + "loss": 1.3942538499832153, + "step": 64 + }, + { + "epoch": 0.14634146341463414, + "grad_norm": 3.2967028617858887, + "learning_rate": 7.1428571428571436e-06, + "loss": 1.4562139511108398, + "step": 66 + }, + { + "epoch": 0.15077605321507762, + "grad_norm": 0.9108039140701294, + "learning_rate": 7.362637362637364e-06, + "loss": 1.324294924736023, + "step": 68 + }, + { + "epoch": 0.15521064301552107, + "grad_norm": 3.4899282455444336, + "learning_rate": 7.582417582417583e-06, + "loss": 1.3148343563079834, + "step": 70 + }, + { + "epoch": 0.15964523281596452, + "grad_norm": 1.4243505001068115, + "learning_rate": 7.802197802197802e-06, + "loss": 1.3413867950439453, + "step": 72 + }, + { + "epoch": 0.164079822616408, + "grad_norm": 0.9133158922195435, + "learning_rate": 8.021978021978023e-06, + "loss": 1.2512249946594238, + "step": 74 + }, + { + "epoch": 0.16851441241685144, + "grad_norm": 1.0660244226455688, + "learning_rate": 8.241758241758243e-06, + "loss": 1.3262264728546143, + "step": 76 + }, + { + "epoch": 0.1729490022172949, + "grad_norm": 1.9149298667907715, + "learning_rate": 8.461538461538462e-06, + "loss": 1.3764103651046753, + "step": 78 + }, + { + "epoch": 0.17738359201773837, + "grad_norm": 1.6447206735610962, + "learning_rate": 8.681318681318681e-06, + "loss": 1.3489444255828857, + "step": 80 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 1.179309368133545, + "learning_rate": 8.9010989010989e-06, + "loss": 0.8915775418281555, + "step": 82 + }, + { + "epoch": 0.18625277161862527, + "grad_norm": 1.079959511756897, + "learning_rate": 9.120879120879122e-06, + "loss": 1.0530604124069214, + "step": 84 + }, + { + "epoch": 0.19068736141906872, + "grad_norm": 1.5468087196350098, + "learning_rate": 9.340659340659341e-06, + "loss": 1.0116887092590332, + "step": 86 + }, + { + "epoch": 0.1951219512195122, + "grad_norm": 1.0850789546966553, + "learning_rate": 9.560439560439562e-06, + "loss": 1.2681747674942017, + "step": 88 + }, + { + "epoch": 0.19955654101995565, + "grad_norm": 3.765213966369629, + "learning_rate": 9.780219780219781e-06, + "loss": 1.398827314376831, + "step": 90 + }, + { + "epoch": 0.2039911308203991, + "grad_norm": 1.0919721126556396, + "learning_rate": 1e-05, + "loss": 1.2438007593154907, + "step": 92 + }, + { + "epoch": 0.20842572062084258, + "grad_norm": 1.347847819328308, + "learning_rate": 9.99996972898091e-06, + "loss": 1.2643123865127563, + "step": 94 + }, + { + "epoch": 0.21286031042128603, + "grad_norm": 2.1960842609405518, + "learning_rate": 9.999878916330893e-06, + "loss": 1.8062665462493896, + "step": 96 + }, + { + "epoch": 0.21729490022172948, + "grad_norm": 1.4442038536071777, + "learning_rate": 9.999727563271727e-06, + "loss": 0.9751254320144653, + "step": 98 + }, + { + "epoch": 0.22172949002217296, + "grad_norm": 1.2550246715545654, + "learning_rate": 9.999515671839682e-06, + "loss": 1.1938117742538452, + "step": 100 + }, + { + "epoch": 0.2261640798226164, + "grad_norm": 1.0004876852035522, + "learning_rate": 9.999243244885499e-06, + "loss": 1.3510534763336182, + "step": 102 + }, + { + "epoch": 0.23059866962305986, + "grad_norm": 1.2936842441558838, + "learning_rate": 9.998910286074355e-06, + "loss": 1.0911916494369507, + "step": 104 + }, + { + "epoch": 0.23503325942350334, + "grad_norm": 1.3704265356063843, + "learning_rate": 9.998516799885806e-06, + "loss": 1.3209298849105835, + "step": 106 + }, + { + "epoch": 0.2394678492239468, + "grad_norm": 1.4554086923599243, + "learning_rate": 9.998062791613729e-06, + "loss": 1.513480305671692, + "step": 108 + }, + { + "epoch": 0.24390243902439024, + "grad_norm": 1.170039176940918, + "learning_rate": 9.997548267366255e-06, + "loss": 1.0687693357467651, + "step": 110 + }, + { + "epoch": 0.24833702882483372, + "grad_norm": 3.054081678390503, + "learning_rate": 9.996973234065685e-06, + "loss": 1.310200810432434, + "step": 112 + }, + { + "epoch": 0.25277161862527714, + "grad_norm": 1.4177216291427612, + "learning_rate": 9.996337699448392e-06, + "loss": 0.8176467418670654, + "step": 114 + }, + { + "epoch": 0.2572062084257206, + "grad_norm": 1.1188538074493408, + "learning_rate": 9.995641672064726e-06, + "loss": 1.3813865184783936, + "step": 116 + }, + { + "epoch": 0.2616407982261641, + "grad_norm": 0.9215674996376038, + "learning_rate": 9.994885161278885e-06, + "loss": 1.1077316999435425, + "step": 118 + }, + { + "epoch": 0.2660753880266075, + "grad_norm": 1.6228851079940796, + "learning_rate": 9.994068177268807e-06, + "loss": 1.2811754941940308, + "step": 120 + }, + { + "epoch": 0.270509977827051, + "grad_norm": 1.9863255023956299, + "learning_rate": 9.993190731026024e-06, + "loss": 1.1968728303909302, + "step": 122 + }, + { + "epoch": 0.2749445676274945, + "grad_norm": 1.8475158214569092, + "learning_rate": 9.992252834355503e-06, + "loss": 1.5891046524047852, + "step": 124 + }, + { + "epoch": 0.2793791574279379, + "grad_norm": 1.485600471496582, + "learning_rate": 9.99125449987551e-06, + "loss": 0.9440798163414001, + "step": 126 + }, + { + "epoch": 0.2838137472283814, + "grad_norm": 1.168931484222412, + "learning_rate": 9.990195741017422e-06, + "loss": 1.21555495262146, + "step": 128 + }, + { + "epoch": 0.28824833702882485, + "grad_norm": 1.28008234500885, + "learning_rate": 9.989076572025554e-06, + "loss": 0.9523183703422546, + "step": 130 + }, + { + "epoch": 0.2926829268292683, + "grad_norm": 3.5025172233581543, + "learning_rate": 9.987897007956968e-06, + "loss": 0.9987781047821045, + "step": 132 + }, + { + "epoch": 0.29711751662971175, + "grad_norm": 2.9309399127960205, + "learning_rate": 9.986657064681267e-06, + "loss": 0.9768642783164978, + "step": 134 + }, + { + "epoch": 0.30155210643015523, + "grad_norm": 6.351615905761719, + "learning_rate": 9.98535675888038e-06, + "loss": 1.2585439682006836, + "step": 136 + }, + { + "epoch": 0.30598669623059865, + "grad_norm": 5.213984489440918, + "learning_rate": 9.983996108048345e-06, + "loss": 0.7967538833618164, + "step": 138 + }, + { + "epoch": 0.31042128603104213, + "grad_norm": 1.3048728704452515, + "learning_rate": 9.982575130491068e-06, + "loss": 1.1146520376205444, + "step": 140 + }, + { + "epoch": 0.3148558758314856, + "grad_norm": 0.8629128336906433, + "learning_rate": 9.981093845326079e-06, + "loss": 0.9639315009117126, + "step": 142 + }, + { + "epoch": 0.31929046563192903, + "grad_norm": 1.43756902217865, + "learning_rate": 9.979552272482268e-06, + "loss": 0.9869639873504639, + "step": 144 + }, + { + "epoch": 0.3237250554323725, + "grad_norm": 2.2892510890960693, + "learning_rate": 9.977950432699629e-06, + "loss": 0.9267846345901489, + "step": 146 + }, + { + "epoch": 0.328159645232816, + "grad_norm": 2.10282301902771, + "learning_rate": 9.976288347528972e-06, + "loss": 1.4183735847473145, + "step": 148 + }, + { + "epoch": 0.3325942350332594, + "grad_norm": 2.502537488937378, + "learning_rate": 9.974566039331634e-06, + "loss": 1.1835788488388062, + "step": 150 + }, + { + "epoch": 0.3370288248337029, + "grad_norm": 2.951907157897949, + "learning_rate": 9.972783531279184e-06, + "loss": 1.0829112529754639, + "step": 152 + }, + { + "epoch": 0.34146341463414637, + "grad_norm": 1.5924861431121826, + "learning_rate": 9.970940847353103e-06, + "loss": 1.2782995700836182, + "step": 154 + }, + { + "epoch": 0.3458980044345898, + "grad_norm": 3.426605701446533, + "learning_rate": 9.969038012344465e-06, + "loss": 1.2216734886169434, + "step": 156 + }, + { + "epoch": 0.35033259423503327, + "grad_norm": 2.055274724960327, + "learning_rate": 9.967075051853609e-06, + "loss": 1.3016668558120728, + "step": 158 + }, + { + "epoch": 0.35476718403547675, + "grad_norm": 0.5714741945266724, + "learning_rate": 9.965051992289782e-06, + "loss": 1.135823130607605, + "step": 160 + }, + { + "epoch": 0.35920177383592017, + "grad_norm": 4.934542655944824, + "learning_rate": 9.962968860870798e-06, + "loss": 0.8752337098121643, + "step": 162 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.104079008102417, + "learning_rate": 9.96082568562266e-06, + "loss": 0.904317319393158, + "step": 164 + }, + { + "epoch": 0.36807095343680707, + "grad_norm": 1.8156555891036987, + "learning_rate": 9.958622495379193e-06, + "loss": 1.284702181816101, + "step": 166 + }, + { + "epoch": 0.37250554323725055, + "grad_norm": 1.5921517610549927, + "learning_rate": 9.956359319781642e-06, + "loss": 1.2042418718338013, + "step": 168 + }, + { + "epoch": 0.376940133037694, + "grad_norm": 1.4465107917785645, + "learning_rate": 9.954036189278292e-06, + "loss": 1.2029085159301758, + "step": 170 + }, + { + "epoch": 0.38137472283813745, + "grad_norm": 2.0260727405548096, + "learning_rate": 9.951653135124045e-06, + "loss": 0.7804557681083679, + "step": 172 + }, + { + "epoch": 0.3858093126385809, + "grad_norm": 3.316241979598999, + "learning_rate": 9.94921018938e-06, + "loss": 1.6452889442443848, + "step": 174 + }, + { + "epoch": 0.3902439024390244, + "grad_norm": 1.2219425439834595, + "learning_rate": 9.946707384913027e-06, + "loss": 1.2721954584121704, + "step": 176 + }, + { + "epoch": 0.3946784922394678, + "grad_norm": 1.0764427185058594, + "learning_rate": 9.944144755395321e-06, + "loss": 1.325971245765686, + "step": 178 + }, + { + "epoch": 0.3991130820399113, + "grad_norm": 0.585249662399292, + "learning_rate": 9.941522335303955e-06, + "loss": 1.0814615488052368, + "step": 180 + }, + { + "epoch": 0.4035476718403548, + "grad_norm": 1.7605838775634766, + "learning_rate": 9.938840159920406e-06, + "loss": 1.2727550268173218, + "step": 182 + }, + { + "epoch": 0.4079822616407982, + "grad_norm": 0.8892675638198853, + "learning_rate": 9.93609826533009e-06, + "loss": 0.8603274822235107, + "step": 184 + }, + { + "epoch": 0.4124168514412417, + "grad_norm": 1.1300294399261475, + "learning_rate": 9.933296688421872e-06, + "loss": 0.986240565776825, + "step": 186 + }, + { + "epoch": 0.41685144124168516, + "grad_norm": 1.141814947128296, + "learning_rate": 9.930435466887564e-06, + "loss": 0.99045330286026, + "step": 188 + }, + { + "epoch": 0.4212860310421286, + "grad_norm": 2.3094711303710938, + "learning_rate": 9.927514639221433e-06, + "loss": 1.0205762386322021, + "step": 190 + }, + { + "epoch": 0.42572062084257206, + "grad_norm": 3.4740211963653564, + "learning_rate": 9.92453424471967e-06, + "loss": 0.8767862319946289, + "step": 192 + }, + { + "epoch": 0.43015521064301554, + "grad_norm": 1.3851454257965088, + "learning_rate": 9.921494323479862e-06, + "loss": 1.306305170059204, + "step": 194 + }, + { + "epoch": 0.43458980044345896, + "grad_norm": 2.2007641792297363, + "learning_rate": 9.918394916400465e-06, + "loss": 1.5621771812438965, + "step": 196 + }, + { + "epoch": 0.43902439024390244, + "grad_norm": 4.152230262756348, + "learning_rate": 9.915236065180235e-06, + "loss": 1.2867047786712646, + "step": 198 + }, + { + "epoch": 0.4434589800443459, + "grad_norm": 0.7970097064971924, + "learning_rate": 9.912017812317684e-06, + "loss": 1.1508140563964844, + "step": 200 + }, + { + "epoch": 0.44789356984478934, + "grad_norm": 1.0820010900497437, + "learning_rate": 9.908740201110497e-06, + "loss": 1.2738271951675415, + "step": 202 + }, + { + "epoch": 0.4523281596452328, + "grad_norm": 0.833768367767334, + "learning_rate": 9.905403275654951e-06, + "loss": 1.229623556137085, + "step": 204 + }, + { + "epoch": 0.4567627494456763, + "grad_norm": 1.3860057592391968, + "learning_rate": 9.902007080845336e-06, + "loss": 1.0840635299682617, + "step": 206 + }, + { + "epoch": 0.4611973392461197, + "grad_norm": 1.5478756427764893, + "learning_rate": 9.898551662373325e-06, + "loss": 1.0657180547714233, + "step": 208 + }, + { + "epoch": 0.4656319290465632, + "grad_norm": 1.0406877994537354, + "learning_rate": 9.895037066727382e-06, + "loss": 0.7027549743652344, + "step": 210 + }, + { + "epoch": 0.4700665188470067, + "grad_norm": 4.616578578948975, + "learning_rate": 9.891463341192124e-06, + "loss": 0.8683266043663025, + "step": 212 + }, + { + "epoch": 0.4745011086474501, + "grad_norm": 0.9766786098480225, + "learning_rate": 9.88783053384769e-06, + "loss": 1.1040089130401611, + "step": 214 + }, + { + "epoch": 0.4789356984478936, + "grad_norm": 1.0400160551071167, + "learning_rate": 9.884138693569095e-06, + "loss": 1.1147682666778564, + "step": 216 + }, + { + "epoch": 0.48337028824833705, + "grad_norm": 1.1635631322860718, + "learning_rate": 9.88038787002557e-06, + "loss": 1.1477664709091187, + "step": 218 + }, + { + "epoch": 0.4878048780487805, + "grad_norm": 3.343935489654541, + "learning_rate": 9.876578113679891e-06, + "loss": 1.24171781539917, + "step": 220 + }, + { + "epoch": 0.49223946784922396, + "grad_norm": 1.407214879989624, + "learning_rate": 9.872709475787708e-06, + "loss": 1.2408087253570557, + "step": 222 + }, + { + "epoch": 0.49667405764966743, + "grad_norm": 0.8646455407142639, + "learning_rate": 9.868782008396848e-06, + "loss": 1.2195643186569214, + "step": 224 + }, + { + "epoch": 0.5011086474501109, + "grad_norm": 1.0053114891052246, + "learning_rate": 9.864795764346615e-06, + "loss": 1.2099742889404297, + "step": 226 + }, + { + "epoch": 0.5055432372505543, + "grad_norm": 1.0781813859939575, + "learning_rate": 9.860750797267085e-06, + "loss": 1.2584586143493652, + "step": 228 + }, + { + "epoch": 0.5099778270509978, + "grad_norm": 2.1706690788269043, + "learning_rate": 9.856647161578384e-06, + "loss": 1.8491344451904297, + "step": 230 + }, + { + "epoch": 0.5144124168514412, + "grad_norm": 1.5501409769058228, + "learning_rate": 9.852484912489946e-06, + "loss": 0.8673834204673767, + "step": 232 + }, + { + "epoch": 0.5188470066518847, + "grad_norm": 1.1405872106552124, + "learning_rate": 9.848264105999783e-06, + "loss": 1.287527322769165, + "step": 234 + }, + { + "epoch": 0.5232815964523282, + "grad_norm": 1.5920919179916382, + "learning_rate": 9.843984798893722e-06, + "loss": 0.9661246538162231, + "step": 236 + }, + { + "epoch": 0.5277161862527716, + "grad_norm": 1.4155439138412476, + "learning_rate": 9.839647048744645e-06, + "loss": 1.0282375812530518, + "step": 238 + }, + { + "epoch": 0.532150776053215, + "grad_norm": 1.3770828247070312, + "learning_rate": 9.83525091391172e-06, + "loss": 1.2707240581512451, + "step": 240 + }, + { + "epoch": 0.5365853658536586, + "grad_norm": 1.1959309577941895, + "learning_rate": 9.8307964535396e-06, + "loss": 1.3354653120040894, + "step": 242 + }, + { + "epoch": 0.541019955654102, + "grad_norm": 5.0889363288879395, + "learning_rate": 9.826283727557644e-06, + "loss": 0.9864997863769531, + "step": 244 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 0.7636584639549255, + "learning_rate": 9.821712796679106e-06, + "loss": 1.2685621976852417, + "step": 246 + }, + { + "epoch": 0.549889135254989, + "grad_norm": 1.781443476676941, + "learning_rate": 9.817083722400309e-06, + "loss": 1.4938619136810303, + "step": 248 + }, + { + "epoch": 0.5543237250554324, + "grad_norm": 2.2715258598327637, + "learning_rate": 9.812396566999832e-06, + "loss": 1.2405654191970825, + "step": 250 + }, + { + "epoch": 0.5587583148558758, + "grad_norm": 30.650293350219727, + "learning_rate": 9.807651393537659e-06, + "loss": 0.6915596127510071, + "step": 252 + }, + { + "epoch": 0.5631929046563193, + "grad_norm": 4.594031810760498, + "learning_rate": 9.802848265854343e-06, + "loss": 1.4810152053833008, + "step": 254 + }, + { + "epoch": 0.5676274944567627, + "grad_norm": 8.177288055419922, + "learning_rate": 9.797987248570137e-06, + "loss": 1.3323289155960083, + "step": 256 + }, + { + "epoch": 0.5720620842572062, + "grad_norm": 1.162073016166687, + "learning_rate": 9.793068407084125e-06, + "loss": 1.2933639287948608, + "step": 258 + }, + { + "epoch": 0.5764966740576497, + "grad_norm": 1.2364484071731567, + "learning_rate": 9.78809180757335e-06, + "loss": 1.2908434867858887, + "step": 260 + }, + { + "epoch": 0.5809312638580931, + "grad_norm": 2.779216766357422, + "learning_rate": 9.783057516991921e-06, + "loss": 0.7582840919494629, + "step": 262 + }, + { + "epoch": 0.5853658536585366, + "grad_norm": 1.085245966911316, + "learning_rate": 9.777965603070106e-06, + "loss": 1.3473342657089233, + "step": 264 + }, + { + "epoch": 0.5898004434589801, + "grad_norm": 1.5546146631240845, + "learning_rate": 9.772816134313424e-06, + "loss": 1.0741627216339111, + "step": 266 + }, + { + "epoch": 0.5942350332594235, + "grad_norm": 1.2348523139953613, + "learning_rate": 9.76760918000173e-06, + "loss": 1.5512841939926147, + "step": 268 + }, + { + "epoch": 0.5986696230598669, + "grad_norm": 2.3876912593841553, + "learning_rate": 9.762344810188276e-06, + "loss": 1.1929247379302979, + "step": 270 + }, + { + "epoch": 0.6031042128603105, + "grad_norm": 1.362686038017273, + "learning_rate": 9.757023095698766e-06, + "loss": 1.2366266250610352, + "step": 272 + }, + { + "epoch": 0.6075388026607539, + "grad_norm": 1.314932107925415, + "learning_rate": 9.751644108130405e-06, + "loss": 1.230374813079834, + "step": 274 + }, + { + "epoch": 0.6119733924611973, + "grad_norm": 1.0009852647781372, + "learning_rate": 9.746207919850951e-06, + "loss": 1.230873942375183, + "step": 276 + }, + { + "epoch": 0.6164079822616408, + "grad_norm": 2.000821590423584, + "learning_rate": 9.740714603997712e-06, + "loss": 1.241438627243042, + "step": 278 + }, + { + "epoch": 0.6208425720620843, + "grad_norm": 1.346825122833252, + "learning_rate": 9.735164234476588e-06, + "loss": 1.3042198419570923, + "step": 280 + }, + { + "epoch": 0.6252771618625277, + "grad_norm": 1.8797986507415771, + "learning_rate": 9.729556885961064e-06, + "loss": 0.9092460870742798, + "step": 282 + }, + { + "epoch": 0.6297117516629712, + "grad_norm": 0.9287862777709961, + "learning_rate": 9.72389263389121e-06, + "loss": 1.269888162612915, + "step": 284 + }, + { + "epoch": 0.6341463414634146, + "grad_norm": 1.3805606365203857, + "learning_rate": 9.718171554472662e-06, + "loss": 1.3446077108383179, + "step": 286 + }, + { + "epoch": 0.6385809312638581, + "grad_norm": 1.057990312576294, + "learning_rate": 9.712393724675597e-06, + "loss": 1.259419322013855, + "step": 288 + }, + { + "epoch": 0.6430155210643016, + "grad_norm": 1.940826654434204, + "learning_rate": 9.706559222233704e-06, + "loss": 1.2493295669555664, + "step": 290 + }, + { + "epoch": 0.647450110864745, + "grad_norm": 1.3292601108551025, + "learning_rate": 9.700668125643132e-06, + "loss": 1.3664789199829102, + "step": 292 + }, + { + "epoch": 0.6518847006651884, + "grad_norm": 1.5756752490997314, + "learning_rate": 9.694720514161437e-06, + "loss": 0.9199124574661255, + "step": 294 + }, + { + "epoch": 0.656319290465632, + "grad_norm": 0.7564427256584167, + "learning_rate": 9.688716467806508e-06, + "loss": 1.0202033519744873, + "step": 296 + }, + { + "epoch": 0.6607538802660754, + "grad_norm": 0.8535528182983398, + "learning_rate": 9.682656067355505e-06, + "loss": 1.2138911485671997, + "step": 298 + }, + { + "epoch": 0.6651884700665188, + "grad_norm": 1.1474356651306152, + "learning_rate": 9.67653939434376e-06, + "loss": 1.2619802951812744, + "step": 300 + }, + { + "epoch": 0.6696230598669624, + "grad_norm": 1.2519088983535767, + "learning_rate": 9.670366531063686e-06, + "loss": 1.2084006071090698, + "step": 302 + }, + { + "epoch": 0.6740576496674058, + "grad_norm": 1.4693647623062134, + "learning_rate": 9.664137560563663e-06, + "loss": 1.2800395488739014, + "step": 304 + }, + { + "epoch": 0.6784922394678492, + "grad_norm": 0.9270733594894409, + "learning_rate": 9.657852566646929e-06, + "loss": 1.2464381456375122, + "step": 306 + }, + { + "epoch": 0.6829268292682927, + "grad_norm": 0.7677431106567383, + "learning_rate": 9.651511633870451e-06, + "loss": 0.8917385339736938, + "step": 308 + }, + { + "epoch": 0.6873614190687362, + "grad_norm": 0.7328879833221436, + "learning_rate": 9.645114847543781e-06, + "loss": 1.248028039932251, + "step": 310 + }, + { + "epoch": 0.6917960088691796, + "grad_norm": 0.6110285520553589, + "learning_rate": 9.638662293727916e-06, + "loss": 1.2161647081375122, + "step": 312 + }, + { + "epoch": 0.6962305986696231, + "grad_norm": 0.7204345464706421, + "learning_rate": 9.632154059234137e-06, + "loss": 1.2376904487609863, + "step": 314 + }, + { + "epoch": 0.7006651884700665, + "grad_norm": 0.8166297078132629, + "learning_rate": 9.625590231622837e-06, + "loss": 1.3569447994232178, + "step": 316 + }, + { + "epoch": 0.70509977827051, + "grad_norm": 0.9921203255653381, + "learning_rate": 9.618970899202354e-06, + "loss": 1.0280705690383911, + "step": 318 + }, + { + "epoch": 0.7095343680709535, + "grad_norm": 2.89858078956604, + "learning_rate": 9.612296151027765e-06, + "loss": 0.9978334903717041, + "step": 320 + }, + { + "epoch": 0.7139689578713969, + "grad_norm": 1.732686996459961, + "learning_rate": 9.605566076899714e-06, + "loss": 0.9863637685775757, + "step": 322 + }, + { + "epoch": 0.7184035476718403, + "grad_norm": 9.696839332580566, + "learning_rate": 9.598780767363174e-06, + "loss": 1.1041550636291504, + "step": 324 + }, + { + "epoch": 0.7228381374722838, + "grad_norm": 2.4485864639282227, + "learning_rate": 9.591940313706248e-06, + "loss": 1.1134647130966187, + "step": 326 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 0.39739716053009033, + "learning_rate": 9.585044807958942e-06, + "loss": 0.8426170349121094, + "step": 328 + }, + { + "epoch": 0.7317073170731707, + "grad_norm": 1.39254629611969, + "learning_rate": 9.578094342891915e-06, + "loss": 0.8422537446022034, + "step": 330 + }, + { + "epoch": 0.7361419068736141, + "grad_norm": 1.350459337234497, + "learning_rate": 9.571089012015237e-06, + "loss": 1.4164326190948486, + "step": 332 + }, + { + "epoch": 0.7405764966740577, + "grad_norm": 1.751500129699707, + "learning_rate": 9.564028909577132e-06, + "loss": 1.2159979343414307, + "step": 334 + }, + { + "epoch": 0.7450110864745011, + "grad_norm": 1.3023916482925415, + "learning_rate": 9.55691413056271e-06, + "loss": 1.2543408870697021, + "step": 336 + }, + { + "epoch": 0.7494456762749445, + "grad_norm": 2.021757125854492, + "learning_rate": 9.54974477069269e-06, + "loss": 0.7136483788490295, + "step": 338 + }, + { + "epoch": 0.753880266075388, + "grad_norm": 2.4734063148498535, + "learning_rate": 9.542520926422105e-06, + "loss": 0.3343808054924011, + "step": 340 + }, + { + "epoch": 0.7583148558758315, + "grad_norm": 4.090439796447754, + "learning_rate": 9.535242694939011e-06, + "loss": 0.8977051973342896, + "step": 342 + }, + { + "epoch": 0.7627494456762749, + "grad_norm": 1.0740253925323486, + "learning_rate": 9.527910174163179e-06, + "loss": 1.317376732826233, + "step": 344 + }, + { + "epoch": 0.7671840354767184, + "grad_norm": 0.9744492173194885, + "learning_rate": 9.520523462744776e-06, + "loss": 1.249900221824646, + "step": 346 + }, + { + "epoch": 0.7716186252771619, + "grad_norm": 1.6449717283248901, + "learning_rate": 9.51308266006304e-06, + "loss": 1.1062456369400024, + "step": 348 + }, + { + "epoch": 0.7760532150776053, + "grad_norm": 1.0315587520599365, + "learning_rate": 9.505587866224939e-06, + "loss": 1.2459303140640259, + "step": 350 + }, + { + "epoch": 0.7804878048780488, + "grad_norm": 1.7614156007766724, + "learning_rate": 9.498039182063828e-06, + "loss": 1.1719058752059937, + "step": 352 + }, + { + "epoch": 0.7849223946784922, + "grad_norm": 3.0436220169067383, + "learning_rate": 9.49043670913809e-06, + "loss": 1.0381566286087036, + "step": 354 + }, + { + "epoch": 0.7893569844789357, + "grad_norm": 6.369256019592285, + "learning_rate": 9.48278054972977e-06, + "loss": 1.2438908815383911, + "step": 356 + }, + { + "epoch": 0.7937915742793792, + "grad_norm": 3.2349908351898193, + "learning_rate": 9.475070806843202e-06, + "loss": 1.323697566986084, + "step": 358 + }, + { + "epoch": 0.7982261640798226, + "grad_norm": 1.123029351234436, + "learning_rate": 9.467307584203619e-06, + "loss": 1.2511515617370605, + "step": 360 + }, + { + "epoch": 0.802660753880266, + "grad_norm": 1.8244556188583374, + "learning_rate": 9.459490986255756e-06, + "loss": 0.6574705839157104, + "step": 362 + }, + { + "epoch": 0.8070953436807096, + "grad_norm": 1.0850750207901, + "learning_rate": 9.451621118162453e-06, + "loss": 1.3474435806274414, + "step": 364 + }, + { + "epoch": 0.811529933481153, + "grad_norm": 1.0580451488494873, + "learning_rate": 9.443698085803235e-06, + "loss": 1.2190296649932861, + "step": 366 + }, + { + "epoch": 0.8159645232815964, + "grad_norm": 1.2732537984848022, + "learning_rate": 9.435721995772884e-06, + "loss": 1.001318335533142, + "step": 368 + }, + { + "epoch": 0.8203991130820399, + "grad_norm": 3.1396641731262207, + "learning_rate": 9.42769295538001e-06, + "loss": 0.8487209677696228, + "step": 370 + }, + { + "epoch": 0.8248337028824834, + "grad_norm": 1.9196490049362183, + "learning_rate": 9.419611072645608e-06, + "loss": 1.2449017763137817, + "step": 372 + }, + { + "epoch": 0.8292682926829268, + "grad_norm": 0.878728449344635, + "learning_rate": 9.4114764563016e-06, + "loss": 1.2437989711761475, + "step": 374 + }, + { + "epoch": 0.8337028824833703, + "grad_norm": 1.0601072311401367, + "learning_rate": 9.403289215789373e-06, + "loss": 1.205723524093628, + "step": 376 + }, + { + "epoch": 0.8381374722838137, + "grad_norm": 6.103841304779053, + "learning_rate": 9.395049461258318e-06, + "loss": 1.246256709098816, + "step": 378 + }, + { + "epoch": 0.8425720620842572, + "grad_norm": 2.8056368827819824, + "learning_rate": 9.386757303564323e-06, + "loss": 0.7605912089347839, + "step": 380 + }, + { + "epoch": 0.8470066518847007, + "grad_norm": 1.2943381071090698, + "learning_rate": 9.37841285426831e-06, + "loss": 1.3198047876358032, + "step": 382 + }, + { + "epoch": 0.8514412416851441, + "grad_norm": 1.5381852388381958, + "learning_rate": 9.370016225634719e-06, + "loss": 1.1507357358932495, + "step": 384 + }, + { + "epoch": 0.8558758314855875, + "grad_norm": 4.027266025543213, + "learning_rate": 9.361567530629988e-06, + "loss": 1.253443956375122, + "step": 386 + }, + { + "epoch": 0.8603104212860311, + "grad_norm": 0.5111842155456543, + "learning_rate": 9.353066882921063e-06, + "loss": 1.1759639978408813, + "step": 388 + }, + { + "epoch": 0.8647450110864745, + "grad_norm": 0.8026149868965149, + "learning_rate": 9.344514396873837e-06, + "loss": 1.2608635425567627, + "step": 390 + }, + { + "epoch": 0.8691796008869179, + "grad_norm": 1.007558822631836, + "learning_rate": 9.335910187551628e-06, + "loss": 0.9073533415794373, + "step": 392 + }, + { + "epoch": 0.8736141906873615, + "grad_norm": 1.0619909763336182, + "learning_rate": 9.327254370713636e-06, + "loss": 1.2199063301086426, + "step": 394 + }, + { + "epoch": 0.8780487804878049, + "grad_norm": 0.797447681427002, + "learning_rate": 9.31854706281336e-06, + "loss": 1.2162882089614868, + "step": 396 + }, + { + "epoch": 0.8824833702882483, + "grad_norm": 1.5606045722961426, + "learning_rate": 9.309788380997069e-06, + "loss": 1.208269476890564, + "step": 398 + }, + { + "epoch": 0.8869179600886918, + "grad_norm": 0.7051414847373962, + "learning_rate": 9.30097844310219e-06, + "loss": 0.5201588869094849, + "step": 400 + }, + { + "epoch": 0.8913525498891353, + "grad_norm": 1.7164850234985352, + "learning_rate": 9.292117367655749e-06, + "loss": 1.0428240299224854, + "step": 402 + }, + { + "epoch": 0.8957871396895787, + "grad_norm": 0.5142672657966614, + "learning_rate": 9.283205273872757e-06, + "loss": 0.9249513745307922, + "step": 404 + }, + { + "epoch": 0.9002217294900222, + "grad_norm": 2.530928373336792, + "learning_rate": 9.274242281654621e-06, + "loss": 1.2720247507095337, + "step": 406 + }, + { + "epoch": 0.9046563192904656, + "grad_norm": 2.1848092079162598, + "learning_rate": 9.265228511587525e-06, + "loss": 1.130611538887024, + "step": 408 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 1.4184147119522095, + "learning_rate": 9.2561640849408e-06, + "loss": 1.1104357242584229, + "step": 410 + }, + { + "epoch": 0.9135254988913526, + "grad_norm": 2.9592013359069824, + "learning_rate": 9.247049123665306e-06, + "loss": 1.2363438606262207, + "step": 412 + }, + { + "epoch": 0.917960088691796, + "grad_norm": 1.4257206916809082, + "learning_rate": 9.237883750391786e-06, + "loss": 1.2583258152008057, + "step": 414 + }, + { + "epoch": 0.9223946784922394, + "grad_norm": 4.5751776695251465, + "learning_rate": 9.228668088429212e-06, + "loss": 0.9869507551193237, + "step": 416 + }, + { + "epoch": 0.926829268292683, + "grad_norm": 0.9538375735282898, + "learning_rate": 9.219402261763129e-06, + "loss": 1.2694066762924194, + "step": 418 + }, + { + "epoch": 0.9312638580931264, + "grad_norm": 1.5408073663711548, + "learning_rate": 9.210086395053992e-06, + "loss": 0.7773984670639038, + "step": 420 + }, + { + "epoch": 0.9356984478935698, + "grad_norm": 3.4732754230499268, + "learning_rate": 9.200720613635476e-06, + "loss": 1.4664820432662964, + "step": 422 + }, + { + "epoch": 0.9401330376940134, + "grad_norm": 2.461186408996582, + "learning_rate": 9.191305043512806e-06, + "loss": 1.0693186521530151, + "step": 424 + }, + { + "epoch": 0.9445676274944568, + "grad_norm": 3.3022894859313965, + "learning_rate": 9.181839811361048e-06, + "loss": 1.2006292343139648, + "step": 426 + }, + { + "epoch": 0.9490022172949002, + "grad_norm": 0.8173018097877502, + "learning_rate": 9.172325044523413e-06, + "loss": 1.100250005722046, + "step": 428 + }, + { + "epoch": 0.9534368070953437, + "grad_norm": 1.222411870956421, + "learning_rate": 9.16276087100954e-06, + "loss": 0.6654449701309204, + "step": 430 + }, + { + "epoch": 0.9578713968957872, + "grad_norm": 1.2987111806869507, + "learning_rate": 9.153147419493774e-06, + "loss": 1.2298423051834106, + "step": 432 + }, + { + "epoch": 0.9623059866962306, + "grad_norm": 2.3424274921417236, + "learning_rate": 9.143484819313441e-06, + "loss": 1.1017597913742065, + "step": 434 + }, + { + "epoch": 0.9667405764966741, + "grad_norm": 1.5677627325057983, + "learning_rate": 9.133773200467095e-06, + "loss": 1.2761000394821167, + "step": 436 + }, + { + "epoch": 0.9711751662971175, + "grad_norm": 1.3605282306671143, + "learning_rate": 9.12401269361278e-06, + "loss": 1.2378090620040894, + "step": 438 + }, + { + "epoch": 0.975609756097561, + "grad_norm": 6.334904193878174, + "learning_rate": 9.114203430066273e-06, + "loss": 0.9209675788879395, + "step": 440 + }, + { + "epoch": 0.9800443458980045, + "grad_norm": 1.7632060050964355, + "learning_rate": 9.104345541799304e-06, + "loss": 1.0448412895202637, + "step": 442 + }, + { + "epoch": 0.9844789356984479, + "grad_norm": 1.790408730506897, + "learning_rate": 9.094439161437797e-06, + "loss": 1.2859784364700317, + "step": 444 + }, + { + "epoch": 0.9889135254988913, + "grad_norm": 1.3654117584228516, + "learning_rate": 9.084484422260079e-06, + "loss": 1.2079429626464844, + "step": 446 + }, + { + "epoch": 0.9933481152993349, + "grad_norm": 1.1531587839126587, + "learning_rate": 9.074481458195077e-06, + "loss": 1.2449597120285034, + "step": 448 + }, + { + "epoch": 0.9977827050997783, + "grad_norm": 0.8101630806922913, + "learning_rate": 9.064430403820538e-06, + "loss": 1.0602502822875977, + "step": 450 + }, + { + "epoch": 1.0022172949002217, + "grad_norm": 3.6886227130889893, + "learning_rate": 9.054331394361195e-06, + "loss": 1.0475915670394897, + "step": 452 + }, + { + "epoch": 1.0066518847006651, + "grad_norm": 1.2973512411117554, + "learning_rate": 9.044184565686963e-06, + "loss": 1.2166190147399902, + "step": 454 + }, + { + "epoch": 1.0110864745011086, + "grad_norm": 1.0704854726791382, + "learning_rate": 9.033990054311108e-06, + "loss": 1.214512586593628, + "step": 456 + }, + { + "epoch": 1.0155210643015522, + "grad_norm": 5.711384296417236, + "learning_rate": 9.023747997388409e-06, + "loss": 0.8170270919799805, + "step": 458 + }, + { + "epoch": 1.0199556541019956, + "grad_norm": 0.5732322931289673, + "learning_rate": 9.013458532713303e-06, + "loss": 0.7988295555114746, + "step": 460 + }, + { + "epoch": 1.024390243902439, + "grad_norm": 1.6274175643920898, + "learning_rate": 9.003121798718055e-06, + "loss": 0.520905077457428, + "step": 462 + }, + { + "epoch": 1.0288248337028825, + "grad_norm": 1.794394612312317, + "learning_rate": 8.992737934470875e-06, + "loss": 0.7766972184181213, + "step": 464 + }, + { + "epoch": 1.033259423503326, + "grad_norm": 1.3309742212295532, + "learning_rate": 8.982307079674051e-06, + "loss": 0.6682877540588379, + "step": 466 + }, + { + "epoch": 1.0376940133037693, + "grad_norm": 1.3699984550476074, + "learning_rate": 8.971829374662075e-06, + "loss": 0.9524274468421936, + "step": 468 + }, + { + "epoch": 1.042128603104213, + "grad_norm": 1.1349494457244873, + "learning_rate": 8.961304960399746e-06, + "loss": 0.610037624835968, + "step": 470 + }, + { + "epoch": 1.0465631929046564, + "grad_norm": 1.1969398260116577, + "learning_rate": 8.950733978480295e-06, + "loss": 0.9021300077438354, + "step": 472 + }, + { + "epoch": 1.0509977827050998, + "grad_norm": 1.6304041147232056, + "learning_rate": 8.940116571123442e-06, + "loss": 1.0323752164840698, + "step": 474 + }, + { + "epoch": 1.0554323725055432, + "grad_norm": 1.383133888244629, + "learning_rate": 8.929452881173522e-06, + "loss": 1.06901216506958, + "step": 476 + }, + { + "epoch": 1.0598669623059866, + "grad_norm": 1.9538545608520508, + "learning_rate": 8.91874305209754e-06, + "loss": 1.087662696838379, + "step": 478 + }, + { + "epoch": 1.06430155210643, + "grad_norm": 0.9558976888656616, + "learning_rate": 8.907987227983244e-06, + "loss": 0.628630518913269, + "step": 480 + }, + { + "epoch": 1.0687361419068737, + "grad_norm": 0.9924697875976562, + "learning_rate": 8.897185553537199e-06, + "loss": 1.079522728919983, + "step": 482 + }, + { + "epoch": 1.0731707317073171, + "grad_norm": 0.4461686313152313, + "learning_rate": 8.886338174082818e-06, + "loss": 0.9315227270126343, + "step": 484 + }, + { + "epoch": 1.0776053215077606, + "grad_norm": 0.8930976986885071, + "learning_rate": 8.875445235558429e-06, + "loss": 1.030774474143982, + "step": 486 + }, + { + "epoch": 1.082039911308204, + "grad_norm": 0.9663419723510742, + "learning_rate": 8.864506884515298e-06, + "loss": 0.775393009185791, + "step": 488 + }, + { + "epoch": 1.0864745011086474, + "grad_norm": 1.8761645555496216, + "learning_rate": 8.853523268115662e-06, + "loss": 0.878838062286377, + "step": 490 + }, + { + "epoch": 1.0909090909090908, + "grad_norm": 1.990370512008667, + "learning_rate": 8.84249453413075e-06, + "loss": 1.1431140899658203, + "step": 492 + }, + { + "epoch": 1.0953436807095343, + "grad_norm": 4.04485559463501, + "learning_rate": 8.831420830938787e-06, + "loss": 1.0441185235977173, + "step": 494 + }, + { + "epoch": 1.099778270509978, + "grad_norm": 3.171431303024292, + "learning_rate": 8.820302307523012e-06, + "loss": 0.8327910304069519, + "step": 496 + }, + { + "epoch": 1.1042128603104213, + "grad_norm": 0.8679511547088623, + "learning_rate": 8.809139113469664e-06, + "loss": 0.9644469618797302, + "step": 498 + }, + { + "epoch": 1.1086474501108647, + "grad_norm": 1.1949224472045898, + "learning_rate": 8.797931398965968e-06, + "loss": 0.6914905905723572, + "step": 500 + }, + { + "epoch": 1.1130820399113082, + "grad_norm": 1.3626770973205566, + "learning_rate": 8.78667931479812e-06, + "loss": 1.0657753944396973, + "step": 502 + }, + { + "epoch": 1.1175166297117516, + "grad_norm": 1.2040343284606934, + "learning_rate": 8.775383012349255e-06, + "loss": 0.8900744915008545, + "step": 504 + }, + { + "epoch": 1.1219512195121952, + "grad_norm": 2.972325325012207, + "learning_rate": 8.764042643597413e-06, + "loss": 0.9030492305755615, + "step": 506 + }, + { + "epoch": 1.1263858093126387, + "grad_norm": 2.5267043113708496, + "learning_rate": 8.75265836111349e-06, + "loss": 0.8307876586914062, + "step": 508 + }, + { + "epoch": 1.130820399113082, + "grad_norm": 1.421155333518982, + "learning_rate": 8.741230318059188e-06, + "loss": 1.1410354375839233, + "step": 510 + }, + { + "epoch": 1.1352549889135255, + "grad_norm": 1.9269262552261353, + "learning_rate": 8.72975866818496e-06, + "loss": 0.5328395962715149, + "step": 512 + }, + { + "epoch": 1.139689578713969, + "grad_norm": 1.2638076543807983, + "learning_rate": 8.718243565827927e-06, + "loss": 0.44444531202316284, + "step": 514 + }, + { + "epoch": 1.1441241685144123, + "grad_norm": 1.3325319290161133, + "learning_rate": 8.706685165909817e-06, + "loss": 0.9025118350982666, + "step": 516 + }, + { + "epoch": 1.1485587583148558, + "grad_norm": 1.31704843044281, + "learning_rate": 8.695083623934872e-06, + "loss": 1.0184192657470703, + "step": 518 + }, + { + "epoch": 1.1529933481152994, + "grad_norm": 0.9442891478538513, + "learning_rate": 8.683439095987758e-06, + "loss": 1.075165033340454, + "step": 520 + }, + { + "epoch": 1.1574279379157428, + "grad_norm": 2.553899049758911, + "learning_rate": 8.671751738731464e-06, + "loss": 0.545891284942627, + "step": 522 + }, + { + "epoch": 1.1618625277161863, + "grad_norm": 1.0405535697937012, + "learning_rate": 8.660021709405197e-06, + "loss": 0.9202799201011658, + "step": 524 + }, + { + "epoch": 1.1662971175166297, + "grad_norm": 1.757952332496643, + "learning_rate": 8.648249165822265e-06, + "loss": 1.147822380065918, + "step": 526 + }, + { + "epoch": 1.170731707317073, + "grad_norm": 0.8222943544387817, + "learning_rate": 8.636434266367956e-06, + "loss": 0.6507425904273987, + "step": 528 + }, + { + "epoch": 1.1751662971175167, + "grad_norm": 0.7118900418281555, + "learning_rate": 8.624577169997394e-06, + "loss": 0.8483346104621887, + "step": 530 + }, + { + "epoch": 1.1796008869179602, + "grad_norm": 0.7820613384246826, + "learning_rate": 8.612678036233428e-06, + "loss": 0.9108617305755615, + "step": 532 + }, + { + "epoch": 1.1840354767184036, + "grad_norm": 1.182815670967102, + "learning_rate": 8.600737025164454e-06, + "loss": 1.0241936445236206, + "step": 534 + }, + { + "epoch": 1.188470066518847, + "grad_norm": 0.36487239599227905, + "learning_rate": 8.588754297442288e-06, + "loss": 0.7436657547950745, + "step": 536 + }, + { + "epoch": 1.1929046563192904, + "grad_norm": 0.996915876865387, + "learning_rate": 8.576730014279982e-06, + "loss": 1.047356128692627, + "step": 538 + }, + { + "epoch": 1.1973392461197339, + "grad_norm": 0.9109376072883606, + "learning_rate": 8.564664337449677e-06, + "loss": 0.6739315986633301, + "step": 540 + }, + { + "epoch": 1.2017738359201773, + "grad_norm": 1.8329638242721558, + "learning_rate": 8.552557429280407e-06, + "loss": 0.6435797214508057, + "step": 542 + }, + { + "epoch": 1.206208425720621, + "grad_norm": 1.6585458517074585, + "learning_rate": 8.540409452655927e-06, + "loss": 1.0112941265106201, + "step": 544 + }, + { + "epoch": 1.2106430155210643, + "grad_norm": 0.4716590940952301, + "learning_rate": 8.528220571012518e-06, + "loss": 0.7702317237854004, + "step": 546 + }, + { + "epoch": 1.2150776053215078, + "grad_norm": 1.1359302997589111, + "learning_rate": 8.51599094833679e-06, + "loss": 1.208444356918335, + "step": 548 + }, + { + "epoch": 1.2195121951219512, + "grad_norm": 1.7960258722305298, + "learning_rate": 8.503720749163472e-06, + "loss": 0.6087892055511475, + "step": 550 + }, + { + "epoch": 1.2239467849223946, + "grad_norm": 1.232657551765442, + "learning_rate": 8.491410138573201e-06, + "loss": 1.0305365324020386, + "step": 552 + }, + { + "epoch": 1.2283813747228383, + "grad_norm": 1.60177481174469, + "learning_rate": 8.479059282190298e-06, + "loss": 0.7768429517745972, + "step": 554 + }, + { + "epoch": 1.2328159645232817, + "grad_norm": 0.3955824077129364, + "learning_rate": 8.466668346180548e-06, + "loss": 0.7353140115737915, + "step": 556 + }, + { + "epoch": 1.237250554323725, + "grad_norm": 2.8532001972198486, + "learning_rate": 8.454237497248956e-06, + "loss": 0.6795215010643005, + "step": 558 + }, + { + "epoch": 1.2416851441241685, + "grad_norm": 2.764127016067505, + "learning_rate": 8.441766902637506e-06, + "loss": 1.2555410861968994, + "step": 560 + }, + { + "epoch": 1.246119733924612, + "grad_norm": 2.7635672092437744, + "learning_rate": 8.429256730122909e-06, + "loss": 0.9878131747245789, + "step": 562 + }, + { + "epoch": 1.2505543237250554, + "grad_norm": 0.5536887645721436, + "learning_rate": 8.416707148014358e-06, + "loss": 0.6607757806777954, + "step": 564 + }, + { + "epoch": 1.2549889135254988, + "grad_norm": 2.1173784732818604, + "learning_rate": 8.404118325151245e-06, + "loss": 0.9502476453781128, + "step": 566 + }, + { + "epoch": 1.2594235033259422, + "grad_norm": 1.0017777681350708, + "learning_rate": 8.391490430900902e-06, + "loss": 0.8671638369560242, + "step": 568 + }, + { + "epoch": 1.2638580931263859, + "grad_norm": 7.101733207702637, + "learning_rate": 8.378823635156319e-06, + "loss": 1.1196980476379395, + "step": 570 + }, + { + "epoch": 1.2682926829268293, + "grad_norm": 0.7950246930122375, + "learning_rate": 8.366118108333861e-06, + "loss": 0.7276540398597717, + "step": 572 + }, + { + "epoch": 1.2727272727272727, + "grad_norm": 2.6005003452301025, + "learning_rate": 8.353374021370967e-06, + "loss": 1.2642651796340942, + "step": 574 + }, + { + "epoch": 1.2771618625277161, + "grad_norm": 1.111757516860962, + "learning_rate": 8.340591545723861e-06, + "loss": 0.9772793650627136, + "step": 576 + }, + { + "epoch": 1.2815964523281598, + "grad_norm": 1.9483938217163086, + "learning_rate": 8.327770853365238e-06, + "loss": 1.144858479499817, + "step": 578 + }, + { + "epoch": 1.2860310421286032, + "grad_norm": 1.3433754444122314, + "learning_rate": 8.314912116781954e-06, + "loss": 1.0109100341796875, + "step": 580 + }, + { + "epoch": 1.2904656319290466, + "grad_norm": 3.104252815246582, + "learning_rate": 8.302015508972702e-06, + "loss": 1.3024400472640991, + "step": 582 + }, + { + "epoch": 1.29490022172949, + "grad_norm": 0.7278043031692505, + "learning_rate": 8.289081203445686e-06, + "loss": 0.6852630376815796, + "step": 584 + }, + { + "epoch": 1.2993348115299335, + "grad_norm": 0.9191624522209167, + "learning_rate": 8.276109374216286e-06, + "loss": 0.7753503322601318, + "step": 586 + }, + { + "epoch": 1.3037694013303769, + "grad_norm": 2.7969019412994385, + "learning_rate": 8.263100195804722e-06, + "loss": 0.5715383291244507, + "step": 588 + }, + { + "epoch": 1.3082039911308203, + "grad_norm": 3.702993392944336, + "learning_rate": 8.250053843233704e-06, + "loss": 1.0089302062988281, + "step": 590 + }, + { + "epoch": 1.3126385809312637, + "grad_norm": 0.8864567279815674, + "learning_rate": 8.236970492026063e-06, + "loss": 0.9880189895629883, + "step": 592 + }, + { + "epoch": 1.3170731707317074, + "grad_norm": 0.795543372631073, + "learning_rate": 8.223850318202415e-06, + "loss": 1.063581109046936, + "step": 594 + }, + { + "epoch": 1.3215077605321508, + "grad_norm": 2.8183748722076416, + "learning_rate": 8.210693498278773e-06, + "loss": 1.100631833076477, + "step": 596 + }, + { + "epoch": 1.3259423503325942, + "grad_norm": 0.9428039789199829, + "learning_rate": 8.197500209264181e-06, + "loss": 1.1460652351379395, + "step": 598 + }, + { + "epoch": 1.3303769401330376, + "grad_norm": 2.286390781402588, + "learning_rate": 8.18427062865833e-06, + "loss": 0.4320100247859955, + "step": 600 + }, + { + "epoch": 1.3348115299334813, + "grad_norm": 2.263921022415161, + "learning_rate": 8.171004934449166e-06, + "loss": 0.9320815205574036, + "step": 602 + }, + { + "epoch": 1.3392461197339247, + "grad_norm": 2.219259023666382, + "learning_rate": 8.157703305110508e-06, + "loss": 0.8366701602935791, + "step": 604 + }, + { + "epoch": 1.3436807095343681, + "grad_norm": 0.45662444829940796, + "learning_rate": 8.144365919599632e-06, + "loss": 0.7252450585365295, + "step": 606 + }, + { + "epoch": 1.3481152993348116, + "grad_norm": 3.525024890899658, + "learning_rate": 8.130992957354872e-06, + "loss": 1.1409834623336792, + "step": 608 + }, + { + "epoch": 1.352549889135255, + "grad_norm": 1.893685221672058, + "learning_rate": 8.117584598293204e-06, + "loss": 1.1217989921569824, + "step": 610 + }, + { + "epoch": 1.3569844789356984, + "grad_norm": 19.174959182739258, + "learning_rate": 8.104141022807824e-06, + "loss": 0.6694931983947754, + "step": 612 + }, + { + "epoch": 1.3614190687361418, + "grad_norm": 0.9400426745414734, + "learning_rate": 8.090662411765726e-06, + "loss": 1.087446689605713, + "step": 614 + }, + { + "epoch": 1.3658536585365852, + "grad_norm": 0.36289501190185547, + "learning_rate": 8.077148946505258e-06, + "loss": 0.6697701215744019, + "step": 616 + }, + { + "epoch": 1.370288248337029, + "grad_norm": 0.8704696893692017, + "learning_rate": 8.063600808833698e-06, + "loss": 0.8936916589736938, + "step": 618 + }, + { + "epoch": 1.3747228381374723, + "grad_norm": 0.25890159606933594, + "learning_rate": 8.050018181024788e-06, + "loss": 0.08715429157018661, + "step": 620 + }, + { + "epoch": 1.3791574279379157, + "grad_norm": 0.6486095190048218, + "learning_rate": 8.036401245816306e-06, + "loss": 0.991193413734436, + "step": 622 + }, + { + "epoch": 1.3835920177383592, + "grad_norm": 1.103645920753479, + "learning_rate": 8.022750186407586e-06, + "loss": 0.943727970123291, + "step": 624 + }, + { + "epoch": 1.3880266075388026, + "grad_norm": 2.296745538711548, + "learning_rate": 8.009065186457061e-06, + "loss": 0.9769091606140137, + "step": 626 + }, + { + "epoch": 1.3924611973392462, + "grad_norm": 2.5007238388061523, + "learning_rate": 7.995346430079799e-06, + "loss": 0.976685106754303, + "step": 628 + }, + { + "epoch": 1.3968957871396896, + "grad_norm": 1.3621313571929932, + "learning_rate": 7.981594101845012e-06, + "loss": 1.103744387626648, + "step": 630 + }, + { + "epoch": 1.401330376940133, + "grad_norm": 0.9651110172271729, + "learning_rate": 7.967808386773591e-06, + "loss": 0.7250331044197083, + "step": 632 + }, + { + "epoch": 1.4057649667405765, + "grad_norm": 1.190347671508789, + "learning_rate": 7.953989470335592e-06, + "loss": 0.9121408462524414, + "step": 634 + }, + { + "epoch": 1.41019955654102, + "grad_norm": 0.8217507004737854, + "learning_rate": 7.940137538447769e-06, + "loss": 1.0396496057510376, + "step": 636 + }, + { + "epoch": 1.4146341463414633, + "grad_norm": 7.098917484283447, + "learning_rate": 7.92625277747105e-06, + "loss": 0.8050640225410461, + "step": 638 + }, + { + "epoch": 1.4190687361419068, + "grad_norm": 1.0582122802734375, + "learning_rate": 7.912335374208043e-06, + "loss": 0.6821240186691284, + "step": 640 + }, + { + "epoch": 1.4235033259423504, + "grad_norm": 1.3478533029556274, + "learning_rate": 7.898385515900517e-06, + "loss": 0.8159670829772949, + "step": 642 + }, + { + "epoch": 1.4279379157427938, + "grad_norm": 1.2756816148757935, + "learning_rate": 7.884403390226883e-06, + "loss": 0.7149834036827087, + "step": 644 + }, + { + "epoch": 1.4323725055432373, + "grad_norm": 0.8082099556922913, + "learning_rate": 7.870389185299672e-06, + "loss": 1.2317858934402466, + "step": 646 + }, + { + "epoch": 1.4368070953436807, + "grad_norm": 0.9018802046775818, + "learning_rate": 7.856343089663002e-06, + "loss": 1.0831472873687744, + "step": 648 + }, + { + "epoch": 1.441241685144124, + "grad_norm": 1.004470944404602, + "learning_rate": 7.842265292290039e-06, + "loss": 1.1530892848968506, + "step": 650 + }, + { + "epoch": 1.4456762749445677, + "grad_norm": 2.8486859798431396, + "learning_rate": 7.828155982580465e-06, + "loss": 0.8907480835914612, + "step": 652 + }, + { + "epoch": 1.4501108647450112, + "grad_norm": 3.3158531188964844, + "learning_rate": 7.814015350357912e-06, + "loss": 1.065255880355835, + "step": 654 + }, + { + "epoch": 1.4545454545454546, + "grad_norm": 2.0944631099700928, + "learning_rate": 7.799843585867426e-06, + "loss": 0.9423489570617676, + "step": 656 + }, + { + "epoch": 1.458980044345898, + "grad_norm": 1.330924391746521, + "learning_rate": 7.785640879772897e-06, + "loss": 0.797570526599884, + "step": 658 + }, + { + "epoch": 1.4634146341463414, + "grad_norm": 9.296934127807617, + "learning_rate": 7.771407423154498e-06, + "loss": 1.0860297679901123, + "step": 660 + }, + { + "epoch": 1.4678492239467849, + "grad_norm": 1.0748159885406494, + "learning_rate": 7.757143407506111e-06, + "loss": 1.0053012371063232, + "step": 662 + }, + { + "epoch": 1.4722838137472283, + "grad_norm": 1.2716701030731201, + "learning_rate": 7.742849024732754e-06, + "loss": 0.7912719249725342, + "step": 664 + }, + { + "epoch": 1.476718403547672, + "grad_norm": 0.5906314849853516, + "learning_rate": 7.728524467148e-06, + "loss": 0.5228022933006287, + "step": 666 + }, + { + "epoch": 1.4811529933481153, + "grad_norm": 12.187054634094238, + "learning_rate": 7.714169927471379e-06, + "loss": 0.5079471468925476, + "step": 668 + }, + { + "epoch": 1.4855875831485588, + "grad_norm": 1.7716224193572998, + "learning_rate": 7.699785598825805e-06, + "loss": 0.903016209602356, + "step": 670 + }, + { + "epoch": 1.4900221729490022, + "grad_norm": 2.2982170581817627, + "learning_rate": 7.68537167473496e-06, + "loss": 1.0076159238815308, + "step": 672 + }, + { + "epoch": 1.4944567627494456, + "grad_norm": 0.5563678741455078, + "learning_rate": 7.670928349120699e-06, + "loss": 0.6532096266746521, + "step": 674 + }, + { + "epoch": 1.4988913525498893, + "grad_norm": 0.7698723077774048, + "learning_rate": 7.656455816300434e-06, + "loss": 0.7211450934410095, + "step": 676 + }, + { + "epoch": 1.5033259423503327, + "grad_norm": 1.3669579029083252, + "learning_rate": 7.641954270984532e-06, + "loss": 1.0425605773925781, + "step": 678 + }, + { + "epoch": 1.507760532150776, + "grad_norm": 0.6051411628723145, + "learning_rate": 7.627423908273683e-06, + "loss": 0.5795204639434814, + "step": 680 + }, + { + "epoch": 1.5121951219512195, + "grad_norm": 1.3261148929595947, + "learning_rate": 7.61286492365628e-06, + "loss": 0.5524274110794067, + "step": 682 + }, + { + "epoch": 1.516629711751663, + "grad_norm": 2.459198474884033, + "learning_rate": 7.598277513005793e-06, + "loss": 1.0282782316207886, + "step": 684 + }, + { + "epoch": 1.5210643015521064, + "grad_norm": 0.8493002653121948, + "learning_rate": 7.583661872578124e-06, + "loss": 0.961588978767395, + "step": 686 + }, + { + "epoch": 1.5254988913525498, + "grad_norm": 1.2923295497894287, + "learning_rate": 7.569018199008976e-06, + "loss": 0.5933117270469666, + "step": 688 + }, + { + "epoch": 1.5299334811529932, + "grad_norm": 1.0922276973724365, + "learning_rate": 7.554346689311205e-06, + "loss": 0.4136204123497009, + "step": 690 + }, + { + "epoch": 1.5343680709534369, + "grad_norm": 0.827107310295105, + "learning_rate": 7.539647540872165e-06, + "loss": 0.724818766117096, + "step": 692 + }, + { + "epoch": 1.5388026607538803, + "grad_norm": 10.415434837341309, + "learning_rate": 7.5249209514510595e-06, + "loss": 0.9134396314620972, + "step": 694 + }, + { + "epoch": 1.5432372505543237, + "grad_norm": 1.308587908744812, + "learning_rate": 7.510167119176273e-06, + "loss": 0.5761569738388062, + "step": 696 + }, + { + "epoch": 1.5476718403547673, + "grad_norm": 0.7201342582702637, + "learning_rate": 7.49538624254271e-06, + "loss": 1.1129498481750488, + "step": 698 + }, + { + "epoch": 1.5521064301552108, + "grad_norm": 0.7750831842422485, + "learning_rate": 7.48057852040913e-06, + "loss": 1.0308855772018433, + "step": 700 + }, + { + "epoch": 1.5565410199556542, + "grad_norm": 2.1023459434509277, + "learning_rate": 7.465744151995458e-06, + "loss": 1.0691447257995605, + "step": 702 + }, + { + "epoch": 1.5609756097560976, + "grad_norm": 0.7603687047958374, + "learning_rate": 7.450883336880116e-06, + "loss": 0.8402605652809143, + "step": 704 + }, + { + "epoch": 1.565410199556541, + "grad_norm": 2.021561861038208, + "learning_rate": 7.435996274997337e-06, + "loss": 0.43568840622901917, + "step": 706 + }, + { + "epoch": 1.5698447893569845, + "grad_norm": 0.7875716090202332, + "learning_rate": 7.421083166634466e-06, + "loss": 1.1289795637130737, + "step": 708 + }, + { + "epoch": 1.5742793791574279, + "grad_norm": 3.5039725303649902, + "learning_rate": 7.40614421242928e-06, + "loss": 0.9707282781600952, + "step": 710 + }, + { + "epoch": 1.5787139689578713, + "grad_norm": 1.424716591835022, + "learning_rate": 7.391179613367272e-06, + "loss": 0.6375144720077515, + "step": 712 + }, + { + "epoch": 1.5831485587583147, + "grad_norm": 0.6600134968757629, + "learning_rate": 7.37618957077896e-06, + "loss": 0.9794768691062927, + "step": 714 + }, + { + "epoch": 1.5875831485587582, + "grad_norm": 1.2803093194961548, + "learning_rate": 7.361174286337175e-06, + "loss": 1.127861738204956, + "step": 716 + }, + { + "epoch": 1.5920177383592018, + "grad_norm": 3.5021018981933594, + "learning_rate": 7.346133962054341e-06, + "loss": 1.2201976776123047, + "step": 718 + }, + { + "epoch": 1.5964523281596452, + "grad_norm": 1.4853342771530151, + "learning_rate": 7.33106880027977e-06, + "loss": 0.5727948546409607, + "step": 720 + }, + { + "epoch": 1.6008869179600886, + "grad_norm": 0.3484899401664734, + "learning_rate": 7.315979003696927e-06, + "loss": 0.5201311111450195, + "step": 722 + }, + { + "epoch": 1.6053215077605323, + "grad_norm": 0.6640962958335876, + "learning_rate": 7.300864775320708e-06, + "loss": 1.0324370861053467, + "step": 724 + }, + { + "epoch": 1.6097560975609757, + "grad_norm": 0.6866331696510315, + "learning_rate": 7.285726318494717e-06, + "loss": 1.0323240756988525, + "step": 726 + }, + { + "epoch": 1.6141906873614191, + "grad_norm": 0.8838914632797241, + "learning_rate": 7.2705638368885105e-06, + "loss": 0.708541989326477, + "step": 728 + }, + { + "epoch": 1.6186252771618626, + "grad_norm": 0.6660059690475464, + "learning_rate": 7.255377534494875e-06, + "loss": 1.0379067659378052, + "step": 730 + }, + { + "epoch": 1.623059866962306, + "grad_norm": 0.61259925365448, + "learning_rate": 7.240167615627082e-06, + "loss": 1.0389152765274048, + "step": 732 + }, + { + "epoch": 1.6274944567627494, + "grad_norm": 0.6811412572860718, + "learning_rate": 7.224934284916127e-06, + "loss": 0.9704182744026184, + "step": 734 + }, + { + "epoch": 1.6319290465631928, + "grad_norm": 0.36625775694847107, + "learning_rate": 7.209677747307982e-06, + "loss": 0.6105785965919495, + "step": 736 + }, + { + "epoch": 1.6363636363636362, + "grad_norm": 0.6093775033950806, + "learning_rate": 7.194398208060848e-06, + "loss": 1.1938239336013794, + "step": 738 + }, + { + "epoch": 1.6407982261640797, + "grad_norm": 0.6593520045280457, + "learning_rate": 7.179095872742378e-06, + "loss": 1.0288550853729248, + "step": 740 + }, + { + "epoch": 1.6452328159645233, + "grad_norm": 2.0731823444366455, + "learning_rate": 7.16377094722692e-06, + "loss": 1.107767939567566, + "step": 742 + }, + { + "epoch": 1.6496674057649667, + "grad_norm": 1.3170353174209595, + "learning_rate": 7.148423637692748e-06, + "loss": 0.8513399362564087, + "step": 744 + }, + { + "epoch": 1.6541019955654102, + "grad_norm": 0.9616439342498779, + "learning_rate": 7.133054150619282e-06, + "loss": 0.9469323754310608, + "step": 746 + }, + { + "epoch": 1.6585365853658538, + "grad_norm": 1.0744072198867798, + "learning_rate": 7.117662692784318e-06, + "loss": 0.6854807734489441, + "step": 748 + }, + { + "epoch": 1.6629711751662972, + "grad_norm": 0.8329254984855652, + "learning_rate": 7.102249471261241e-06, + "loss": 1.0346763134002686, + "step": 750 + }, + { + "epoch": 1.6674057649667406, + "grad_norm": 1.6085177659988403, + "learning_rate": 7.0868146934162365e-06, + "loss": 0.9381062388420105, + "step": 752 + }, + { + "epoch": 1.671840354767184, + "grad_norm": 0.6383267641067505, + "learning_rate": 7.071358566905507e-06, + "loss": 0.6153408885002136, + "step": 754 + }, + { + "epoch": 1.6762749445676275, + "grad_norm": 0.7473850846290588, + "learning_rate": 7.055881299672476e-06, + "loss": 0.9577728509902954, + "step": 756 + }, + { + "epoch": 1.680709534368071, + "grad_norm": 0.7026475071907043, + "learning_rate": 7.040383099944988e-06, + "loss": 0.9313692450523376, + "step": 758 + }, + { + "epoch": 1.6851441241685143, + "grad_norm": 13.826199531555176, + "learning_rate": 7.02486417623251e-06, + "loss": 0.8703269362449646, + "step": 760 + }, + { + "epoch": 1.6895787139689578, + "grad_norm": 1.2666261196136475, + "learning_rate": 7.009324737323325e-06, + "loss": 1.1479053497314453, + "step": 762 + }, + { + "epoch": 1.6940133037694012, + "grad_norm": 1.4821723699569702, + "learning_rate": 6.993764992281722e-06, + "loss": 1.0106499195098877, + "step": 764 + }, + { + "epoch": 1.6984478935698448, + "grad_norm": 2.4803268909454346, + "learning_rate": 6.978185150445187e-06, + "loss": 0.9789531230926514, + "step": 766 + }, + { + "epoch": 1.7028824833702882, + "grad_norm": 1.579527497291565, + "learning_rate": 6.96258542142158e-06, + "loss": 0.6890674233436584, + "step": 768 + }, + { + "epoch": 1.7073170731707317, + "grad_norm": 1.3111584186553955, + "learning_rate": 6.946966015086321e-06, + "loss": 1.0892305374145508, + "step": 770 + }, + { + "epoch": 1.7117516629711753, + "grad_norm": 3.6061043739318848, + "learning_rate": 6.931327141579565e-06, + "loss": 1.0220389366149902, + "step": 772 + }, + { + "epoch": 1.7161862527716187, + "grad_norm": 0.8147404789924622, + "learning_rate": 6.915669011303374e-06, + "loss": 0.9009866118431091, + "step": 774 + }, + { + "epoch": 1.7206208425720622, + "grad_norm": 0.8462866544723511, + "learning_rate": 6.899991834918884e-06, + "loss": 1.1094865798950195, + "step": 776 + }, + { + "epoch": 1.7250554323725056, + "grad_norm": 0.7154016494750977, + "learning_rate": 6.884295823343479e-06, + "loss": 1.0449351072311401, + "step": 778 + }, + { + "epoch": 1.729490022172949, + "grad_norm": 1.2640376091003418, + "learning_rate": 6.868581187747941e-06, + "loss": 0.6298045516014099, + "step": 780 + }, + { + "epoch": 1.7339246119733924, + "grad_norm": 0.6618129014968872, + "learning_rate": 6.852848139553619e-06, + "loss": 0.9618121981620789, + "step": 782 + }, + { + "epoch": 1.7383592017738358, + "grad_norm": 0.4092734158039093, + "learning_rate": 6.837096890429582e-06, + "loss": 0.6855942606925964, + "step": 784 + }, + { + "epoch": 1.7427937915742793, + "grad_norm": 0.6238394379615784, + "learning_rate": 6.821327652289768e-06, + "loss": 0.7413522005081177, + "step": 786 + }, + { + "epoch": 1.7472283813747227, + "grad_norm": 1.0907618999481201, + "learning_rate": 6.8055406372901344e-06, + "loss": 0.6723021268844604, + "step": 788 + }, + { + "epoch": 1.7516629711751663, + "grad_norm": 3.868008613586426, + "learning_rate": 6.789736057825812e-06, + "loss": 0.6211203932762146, + "step": 790 + }, + { + "epoch": 1.7560975609756098, + "grad_norm": 0.46340852975845337, + "learning_rate": 6.77391412652823e-06, + "loss": 0.6684106588363647, + "step": 792 + }, + { + "epoch": 1.7605321507760532, + "grad_norm": 0.6993800401687622, + "learning_rate": 6.758075056262271e-06, + "loss": 0.8443524837493896, + "step": 794 + }, + { + "epoch": 1.7649667405764968, + "grad_norm": 0.9882437586784363, + "learning_rate": 6.742219060123403e-06, + "loss": 0.7883599400520325, + "step": 796 + }, + { + "epoch": 1.7694013303769403, + "grad_norm": 0.7495700716972351, + "learning_rate": 6.7263463514348095e-06, + "loss": 0.9688935875892639, + "step": 798 + }, + { + "epoch": 1.7738359201773837, + "grad_norm": 0.7344732880592346, + "learning_rate": 6.710457143744519e-06, + "loss": 1.0964137315750122, + "step": 800 + }, + { + "epoch": 1.778270509977827, + "grad_norm": 1.3385992050170898, + "learning_rate": 6.6945516508225325e-06, + "loss": 0.9830183982849121, + "step": 802 + }, + { + "epoch": 1.7827050997782705, + "grad_norm": 1.4036580324172974, + "learning_rate": 6.678630086657959e-06, + "loss": 1.1154334545135498, + "step": 804 + }, + { + "epoch": 1.787139689578714, + "grad_norm": 1.4150375127792358, + "learning_rate": 6.662692665456115e-06, + "loss": 0.7647712230682373, + "step": 806 + }, + { + "epoch": 1.7915742793791574, + "grad_norm": 1.0048019886016846, + "learning_rate": 6.646739601635661e-06, + "loss": 1.0117907524108887, + "step": 808 + }, + { + "epoch": 1.7960088691796008, + "grad_norm": 0.8310498595237732, + "learning_rate": 6.6307711098257074e-06, + "loss": 0.5957526564598083, + "step": 810 + }, + { + "epoch": 1.8004434589800442, + "grad_norm": 1.8356163501739502, + "learning_rate": 6.6147874048629294e-06, + "loss": 0.6535270810127258, + "step": 812 + }, + { + "epoch": 1.8048780487804879, + "grad_norm": 2.6936192512512207, + "learning_rate": 6.598788701788677e-06, + "loss": 1.2300708293914795, + "step": 814 + }, + { + "epoch": 1.8093126385809313, + "grad_norm": 1.4539717435836792, + "learning_rate": 6.582775215846082e-06, + "loss": 0.9219212532043457, + "step": 816 + }, + { + "epoch": 1.8137472283813747, + "grad_norm": 1.1613236665725708, + "learning_rate": 6.566747162477164e-06, + "loss": 0.4706512689590454, + "step": 818 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 0.9962729811668396, + "learning_rate": 6.5507047573199235e-06, + "loss": 1.0331782102584839, + "step": 820 + }, + { + "epoch": 1.8226164079822618, + "grad_norm": 1.4629065990447998, + "learning_rate": 6.5346482162054526e-06, + "loss": 1.0425474643707275, + "step": 822 + }, + { + "epoch": 1.8270509977827052, + "grad_norm": 0.9501014351844788, + "learning_rate": 6.518577755155024e-06, + "loss": 1.1319053173065186, + "step": 824 + }, + { + "epoch": 1.8314855875831486, + "grad_norm": 2.260728120803833, + "learning_rate": 6.502493590377184e-06, + "loss": 0.9275112748146057, + "step": 826 + }, + { + "epoch": 1.835920177383592, + "grad_norm": 1.412589192390442, + "learning_rate": 6.48639593826485e-06, + "loss": 0.8285998106002808, + "step": 828 + }, + { + "epoch": 1.8403547671840355, + "grad_norm": 0.8061251640319824, + "learning_rate": 6.4702850153923915e-06, + "loss": 0.9728037714958191, + "step": 830 + }, + { + "epoch": 1.8447893569844789, + "grad_norm": 3.3840270042419434, + "learning_rate": 6.45416103851272e-06, + "loss": 0.9926842451095581, + "step": 832 + }, + { + "epoch": 1.8492239467849223, + "grad_norm": 0.75401771068573, + "learning_rate": 6.438024224554378e-06, + "loss": 0.9589172005653381, + "step": 834 + }, + { + "epoch": 1.8536585365853657, + "grad_norm": 0.8730403184890747, + "learning_rate": 6.421874790618608e-06, + "loss": 0.9601749181747437, + "step": 836 + }, + { + "epoch": 1.8580931263858091, + "grad_norm": 7.889377593994141, + "learning_rate": 6.405712953976444e-06, + "loss": 0.9879204034805298, + "step": 838 + }, + { + "epoch": 1.8625277161862528, + "grad_norm": 1.303679347038269, + "learning_rate": 6.389538932065783e-06, + "loss": 1.1837224960327148, + "step": 840 + }, + { + "epoch": 1.8669623059866962, + "grad_norm": 2.8485257625579834, + "learning_rate": 6.373352942488455e-06, + "loss": 0.7557274103164673, + "step": 842 + }, + { + "epoch": 1.8713968957871396, + "grad_norm": 1.183737874031067, + "learning_rate": 6.357155203007307e-06, + "loss": 0.7274500131607056, + "step": 844 + }, + { + "epoch": 1.8758314855875833, + "grad_norm": 1.8683289289474487, + "learning_rate": 6.340945931543263e-06, + "loss": 0.8478662371635437, + "step": 846 + }, + { + "epoch": 1.8802660753880267, + "grad_norm": 2.6959264278411865, + "learning_rate": 6.324725346172399e-06, + "loss": 0.8910980820655823, + "step": 848 + }, + { + "epoch": 1.8847006651884701, + "grad_norm": 1.2880501747131348, + "learning_rate": 6.308493665123e-06, + "loss": 1.2397483587265015, + "step": 850 + }, + { + "epoch": 1.8891352549889135, + "grad_norm": 2.6425297260284424, + "learning_rate": 6.2922511067726365e-06, + "loss": 0.6860370635986328, + "step": 852 + }, + { + "epoch": 1.893569844789357, + "grad_norm": 1.1020407676696777, + "learning_rate": 6.2759978896452155e-06, + "loss": 1.0302581787109375, + "step": 854 + }, + { + "epoch": 1.8980044345898004, + "grad_norm": 1.1593939065933228, + "learning_rate": 6.259734232408047e-06, + "loss": 0.9292960166931152, + "step": 856 + }, + { + "epoch": 1.9024390243902438, + "grad_norm": 0.899362325668335, + "learning_rate": 6.2434603538688975e-06, + "loss": 0.726291835308075, + "step": 858 + }, + { + "epoch": 1.9068736141906872, + "grad_norm": 2.7879903316497803, + "learning_rate": 6.2271764729730525e-06, + "loss": 0.7539620995521545, + "step": 860 + }, + { + "epoch": 1.9113082039911307, + "grad_norm": 2.1077184677124023, + "learning_rate": 6.210882808800366e-06, + "loss": 1.0768671035766602, + "step": 862 + }, + { + "epoch": 1.9157427937915743, + "grad_norm": 0.9502280354499817, + "learning_rate": 6.19457958056231e-06, + "loss": 0.7523056864738464, + "step": 864 + }, + { + "epoch": 1.9201773835920177, + "grad_norm": 1.5630991458892822, + "learning_rate": 6.178267007599034e-06, + "loss": 0.6150217652320862, + "step": 866 + }, + { + "epoch": 1.9246119733924612, + "grad_norm": 1.0712734460830688, + "learning_rate": 6.161945309376409e-06, + "loss": 1.0981650352478027, + "step": 868 + }, + { + "epoch": 1.9290465631929048, + "grad_norm": 0.8349732160568237, + "learning_rate": 6.145614705483075e-06, + "loss": 1.0270382165908813, + "step": 870 + }, + { + "epoch": 1.9334811529933482, + "grad_norm": 0.5500652194023132, + "learning_rate": 6.129275415627485e-06, + "loss": 0.7593640685081482, + "step": 872 + }, + { + "epoch": 1.9379157427937916, + "grad_norm": 1.514795184135437, + "learning_rate": 6.11292765963495e-06, + "loss": 0.6513434052467346, + "step": 874 + }, + { + "epoch": 1.942350332594235, + "grad_norm": 0.9090774655342102, + "learning_rate": 6.09657165744469e-06, + "loss": 0.6522112488746643, + "step": 876 + }, + { + "epoch": 1.9467849223946785, + "grad_norm": 0.8675610423088074, + "learning_rate": 6.080207629106859e-06, + "loss": 0.740065336227417, + "step": 878 + }, + { + "epoch": 1.951219512195122, + "grad_norm": 1.5948195457458496, + "learning_rate": 6.063835794779598e-06, + "loss": 0.8240799307823181, + "step": 880 + }, + { + "epoch": 1.9556541019955653, + "grad_norm": 3.2498433589935303, + "learning_rate": 6.047456374726067e-06, + "loss": 0.5737804174423218, + "step": 882 + }, + { + "epoch": 1.9600886917960088, + "grad_norm": 0.8827385902404785, + "learning_rate": 6.031069589311481e-06, + "loss": 0.7567857503890991, + "step": 884 + }, + { + "epoch": 1.9645232815964522, + "grad_norm": 1.2192054986953735, + "learning_rate": 6.01467565900015e-06, + "loss": 0.9545295238494873, + "step": 886 + }, + { + "epoch": 1.9689578713968958, + "grad_norm": 0.7741813659667969, + "learning_rate": 5.99827480435251e-06, + "loss": 0.9531633257865906, + "step": 888 + }, + { + "epoch": 1.9733924611973392, + "grad_norm": 0.8368670344352722, + "learning_rate": 5.981867246022149e-06, + "loss": 1.0802265405654907, + "step": 890 + }, + { + "epoch": 1.9778270509977827, + "grad_norm": 1.4058384895324707, + "learning_rate": 5.965453204752855e-06, + "loss": 1.0436638593673706, + "step": 892 + }, + { + "epoch": 1.9822616407982263, + "grad_norm": 1.1106218099594116, + "learning_rate": 5.949032901375627e-06, + "loss": 1.013320803642273, + "step": 894 + }, + { + "epoch": 1.9866962305986697, + "grad_norm": 5.496564865112305, + "learning_rate": 5.932606556805719e-06, + "loss": 0.8702308535575867, + "step": 896 + }, + { + "epoch": 1.9911308203991132, + "grad_norm": 0.7846924066543579, + "learning_rate": 5.916174392039659e-06, + "loss": 0.9531219005584717, + "step": 898 + }, + { + "epoch": 1.9955654101995566, + "grad_norm": 1.4032496213912964, + "learning_rate": 5.899736628152284e-06, + "loss": 0.7455130815505981, + "step": 900 + }, + { + "epoch": 2.0, + "grad_norm": 0.851958155632019, + "learning_rate": 5.88329348629375e-06, + "loss": 0.9732981324195862, + "step": 902 + }, + { + "epoch": 2.0044345898004434, + "grad_norm": 0.8989659547805786, + "learning_rate": 5.8668451876865736e-06, + "loss": 0.6848430633544922, + "step": 904 + }, + { + "epoch": 2.008869179600887, + "grad_norm": 0.9057672023773193, + "learning_rate": 5.850391953622652e-06, + "loss": 0.7127602100372314, + "step": 906 + }, + { + "epoch": 2.0133037694013303, + "grad_norm": 0.8921846747398376, + "learning_rate": 5.8339340054602775e-06, + "loss": 0.851362943649292, + "step": 908 + }, + { + "epoch": 2.0177383592017737, + "grad_norm": 0.8631959557533264, + "learning_rate": 5.817471564621169e-06, + "loss": 0.7096256017684937, + "step": 910 + }, + { + "epoch": 2.022172949002217, + "grad_norm": 1.8130816221237183, + "learning_rate": 5.801004852587485e-06, + "loss": 0.5044680237770081, + "step": 912 + }, + { + "epoch": 2.0266075388026605, + "grad_norm": 1.5005384683609009, + "learning_rate": 5.784534090898849e-06, + "loss": 0.5559933185577393, + "step": 914 + }, + { + "epoch": 2.0310421286031044, + "grad_norm": 1.5633256435394287, + "learning_rate": 5.768059501149369e-06, + "loss": 0.6281445622444153, + "step": 916 + }, + { + "epoch": 2.035476718403548, + "grad_norm": 1.614864706993103, + "learning_rate": 5.751581304984657e-06, + "loss": 0.6711671948432922, + "step": 918 + }, + { + "epoch": 2.0399113082039912, + "grad_norm": 1.0288786888122559, + "learning_rate": 5.735099724098838e-06, + "loss": 0.5363720655441284, + "step": 920 + }, + { + "epoch": 2.0443458980044347, + "grad_norm": 6.116771221160889, + "learning_rate": 5.718614980231582e-06, + "loss": 0.5760122537612915, + "step": 922 + }, + { + "epoch": 2.048780487804878, + "grad_norm": 0.6346896290779114, + "learning_rate": 5.702127295165107e-06, + "loss": 0.1978059560060501, + "step": 924 + }, + { + "epoch": 2.0532150776053215, + "grad_norm": 1.3448708057403564, + "learning_rate": 5.685636890721205e-06, + "loss": 0.8290249705314636, + "step": 926 + }, + { + "epoch": 2.057649667405765, + "grad_norm": 2.6560189723968506, + "learning_rate": 5.669143988758253e-06, + "loss": 0.5688458681106567, + "step": 928 + }, + { + "epoch": 2.0620842572062084, + "grad_norm": 0.2830648124217987, + "learning_rate": 5.652648811168228e-06, + "loss": 0.4656969904899597, + "step": 930 + }, + { + "epoch": 2.066518847006652, + "grad_norm": 2.9565589427948, + "learning_rate": 5.636151579873726e-06, + "loss": 0.5648703575134277, + "step": 932 + }, + { + "epoch": 2.070953436807095, + "grad_norm": 0.8213310837745667, + "learning_rate": 5.619652516824967e-06, + "loss": 0.6612739562988281, + "step": 934 + }, + { + "epoch": 2.0753880266075386, + "grad_norm": 1.1572480201721191, + "learning_rate": 5.603151843996822e-06, + "loss": 0.7529350519180298, + "step": 936 + }, + { + "epoch": 2.079822616407982, + "grad_norm": 1.7981088161468506, + "learning_rate": 5.586649783385813e-06, + "loss": 0.5522570610046387, + "step": 938 + }, + { + "epoch": 2.084257206208426, + "grad_norm": 1.3611361980438232, + "learning_rate": 5.570146557007141e-06, + "loss": 0.48875561356544495, + "step": 940 + }, + { + "epoch": 2.0886917960088693, + "grad_norm": 1.683354139328003, + "learning_rate": 5.553642386891683e-06, + "loss": 0.811143159866333, + "step": 942 + }, + { + "epoch": 2.0931263858093128, + "grad_norm": 1.7977195978164673, + "learning_rate": 5.537137495083018e-06, + "loss": 0.45032718777656555, + "step": 944 + }, + { + "epoch": 2.097560975609756, + "grad_norm": 1.0323246717453003, + "learning_rate": 5.5206321036344304e-06, + "loss": 0.703310489654541, + "step": 946 + }, + { + "epoch": 2.1019955654101996, + "grad_norm": 1.2375712394714355, + "learning_rate": 5.504126434605932e-06, + "loss": 0.7233847379684448, + "step": 948 + }, + { + "epoch": 2.106430155210643, + "grad_norm": 1.348511815071106, + "learning_rate": 5.487620710061262e-06, + "loss": 0.4606630802154541, + "step": 950 + }, + { + "epoch": 2.1108647450110865, + "grad_norm": 1.3615354299545288, + "learning_rate": 5.471115152064916e-06, + "loss": 0.5963136553764343, + "step": 952 + }, + { + "epoch": 2.11529933481153, + "grad_norm": 3.1463687419891357, + "learning_rate": 5.454609982679138e-06, + "loss": 0.5811668038368225, + "step": 954 + }, + { + "epoch": 2.1197339246119733, + "grad_norm": 1.2822504043579102, + "learning_rate": 5.4381054239609525e-06, + "loss": 0.7643156051635742, + "step": 956 + }, + { + "epoch": 2.1241685144124167, + "grad_norm": 1.20485258102417, + "learning_rate": 5.421601697959164e-06, + "loss": 0.5839754343032837, + "step": 958 + }, + { + "epoch": 2.12860310421286, + "grad_norm": 2.490534782409668, + "learning_rate": 5.405099026711374e-06, + "loss": 0.5331164598464966, + "step": 960 + }, + { + "epoch": 2.1330376940133036, + "grad_norm": 3.638983726501465, + "learning_rate": 5.388597632240994e-06, + "loss": 0.40006810426712036, + "step": 962 + }, + { + "epoch": 2.1374722838137474, + "grad_norm": 6.238394737243652, + "learning_rate": 5.372097736554261e-06, + "loss": 0.6906276941299438, + "step": 964 + }, + { + "epoch": 2.141906873614191, + "grad_norm": 0.839712917804718, + "learning_rate": 5.35559956163724e-06, + "loss": 0.6319162249565125, + "step": 966 + }, + { + "epoch": 2.1463414634146343, + "grad_norm": 0.7551127672195435, + "learning_rate": 5.339103329452856e-06, + "loss": 0.41347965598106384, + "step": 968 + }, + { + "epoch": 2.1507760532150777, + "grad_norm": 0.8068141937255859, + "learning_rate": 5.322609261937887e-06, + "loss": 0.5399714112281799, + "step": 970 + }, + { + "epoch": 2.155210643015521, + "grad_norm": 1.4304999113082886, + "learning_rate": 5.306117580999993e-06, + "loss": 0.4987761676311493, + "step": 972 + }, + { + "epoch": 2.1596452328159645, + "grad_norm": 1.4152289628982544, + "learning_rate": 5.289628508514725e-06, + "loss": 0.7639641165733337, + "step": 974 + }, + { + "epoch": 2.164079822616408, + "grad_norm": 1.2822636365890503, + "learning_rate": 5.2731422663225385e-06, + "loss": 0.7585563659667969, + "step": 976 + }, + { + "epoch": 2.1685144124168514, + "grad_norm": 2.3158578872680664, + "learning_rate": 5.256659076225813e-06, + "loss": 0.3914712071418762, + "step": 978 + }, + { + "epoch": 2.172949002217295, + "grad_norm": 1.916115164756775, + "learning_rate": 5.240179159985866e-06, + "loss": 0.8097031712532043, + "step": 980 + }, + { + "epoch": 2.1773835920177382, + "grad_norm": 1.6557332277297974, + "learning_rate": 5.2237027393199645e-06, + "loss": 0.37278667092323303, + "step": 982 + }, + { + "epoch": 2.1818181818181817, + "grad_norm": 1.2614706754684448, + "learning_rate": 5.207230035898356e-06, + "loss": 0.2516429126262665, + "step": 984 + }, + { + "epoch": 2.186252771618625, + "grad_norm": 1.9887027740478516, + "learning_rate": 5.190761271341268e-06, + "loss": 0.659031331539154, + "step": 986 + }, + { + "epoch": 2.1906873614190685, + "grad_norm": 3.152275323867798, + "learning_rate": 5.174296667215939e-06, + "loss": 0.28620240092277527, + "step": 988 + }, + { + "epoch": 2.1951219512195124, + "grad_norm": 0.960594654083252, + "learning_rate": 5.157836445033636e-06, + "loss": 0.702060878276825, + "step": 990 + }, + { + "epoch": 2.199556541019956, + "grad_norm": 2.6779539585113525, + "learning_rate": 5.141380826246667e-06, + "loss": 0.9569138288497925, + "step": 992 + }, + { + "epoch": 2.203991130820399, + "grad_norm": 1.3519855737686157, + "learning_rate": 5.124930032245415e-06, + "loss": 0.5782943964004517, + "step": 994 + }, + { + "epoch": 2.2084257206208426, + "grad_norm": 1.4382355213165283, + "learning_rate": 5.108484284355339e-06, + "loss": 0.7067066431045532, + "step": 996 + }, + { + "epoch": 2.212860310421286, + "grad_norm": 1.1686522960662842, + "learning_rate": 5.0920438038340194e-06, + "loss": 0.6596247553825378, + "step": 998 + }, + { + "epoch": 2.2172949002217295, + "grad_norm": 1.4124189615249634, + "learning_rate": 5.075608811868169e-06, + "loss": 0.6456693410873413, + "step": 1000 + }, + { + "epoch": 2.221729490022173, + "grad_norm": 4.106961727142334, + "learning_rate": 5.059179529570657e-06, + "loss": 0.3303482234477997, + "step": 1002 + }, + { + "epoch": 2.2261640798226163, + "grad_norm": 0.9121060371398926, + "learning_rate": 5.042756177977534e-06, + "loss": 0.6765180230140686, + "step": 1004 + }, + { + "epoch": 2.2305986696230597, + "grad_norm": 1.4012293815612793, + "learning_rate": 5.026338978045062e-06, + "loss": 0.5389603972434998, + "step": 1006 + }, + { + "epoch": 2.235033259423503, + "grad_norm": 0.968221127986908, + "learning_rate": 5.009928150646741e-06, + "loss": 0.6899822950363159, + "step": 1008 + }, + { + "epoch": 2.2394678492239466, + "grad_norm": 1.1386181116104126, + "learning_rate": 4.993523916570334e-06, + "loss": 0.6064386367797852, + "step": 1010 + }, + { + "epoch": 2.2439024390243905, + "grad_norm": 1.1713107824325562, + "learning_rate": 4.977126496514902e-06, + "loss": 0.6847143769264221, + "step": 1012 + }, + { + "epoch": 2.248337028824834, + "grad_norm": 1.6454007625579834, + "learning_rate": 4.960736111087827e-06, + "loss": 0.6909704804420471, + "step": 1014 + }, + { + "epoch": 2.2527716186252773, + "grad_norm": 0.9556133151054382, + "learning_rate": 4.9443529808018545e-06, + "loss": 0.8876364231109619, + "step": 1016 + }, + { + "epoch": 2.2572062084257207, + "grad_norm": 1.2127729654312134, + "learning_rate": 4.927977326072115e-06, + "loss": 0.36061155796051025, + "step": 1018 + }, + { + "epoch": 2.261640798226164, + "grad_norm": 0.925119936466217, + "learning_rate": 4.911609367213168e-06, + "loss": 0.8120240569114685, + "step": 1020 + }, + { + "epoch": 2.2660753880266076, + "grad_norm": 0.3137432932853699, + "learning_rate": 4.895249324436035e-06, + "loss": 0.4972486197948456, + "step": 1022 + }, + { + "epoch": 2.270509977827051, + "grad_norm": 1.6952173709869385, + "learning_rate": 4.8788974178452316e-06, + "loss": 0.9327743053436279, + "step": 1024 + }, + { + "epoch": 2.2749445676274944, + "grad_norm": 0.680587112903595, + "learning_rate": 4.86255386743582e-06, + "loss": 0.524124026298523, + "step": 1026 + }, + { + "epoch": 2.279379157427938, + "grad_norm": 1.0751386880874634, + "learning_rate": 4.846218893090426e-06, + "loss": 0.7662097215652466, + "step": 1028 + }, + { + "epoch": 2.2838137472283813, + "grad_norm": 0.8584238290786743, + "learning_rate": 4.829892714576307e-06, + "loss": 0.7474344968795776, + "step": 1030 + }, + { + "epoch": 2.2882483370288247, + "grad_norm": 2.667839765548706, + "learning_rate": 4.813575551542381e-06, + "loss": 0.6120243072509766, + "step": 1032 + }, + { + "epoch": 2.292682926829268, + "grad_norm": 2.76969313621521, + "learning_rate": 4.7972676235162714e-06, + "loss": 0.7319304347038269, + "step": 1034 + }, + { + "epoch": 2.2971175166297115, + "grad_norm": 1.307244896888733, + "learning_rate": 4.780969149901354e-06, + "loss": 0.7238577604293823, + "step": 1036 + }, + { + "epoch": 2.3015521064301554, + "grad_norm": 2.190412998199463, + "learning_rate": 4.764680349973812e-06, + "loss": 0.732725977897644, + "step": 1038 + }, + { + "epoch": 2.305986696230599, + "grad_norm": 1.9367870092391968, + "learning_rate": 4.748401442879674e-06, + "loss": 0.6513870358467102, + "step": 1040 + }, + { + "epoch": 2.3104212860310422, + "grad_norm": 1.6586291790008545, + "learning_rate": 4.732132647631881e-06, + "loss": 0.9142364263534546, + "step": 1042 + }, + { + "epoch": 2.3148558758314857, + "grad_norm": 0.5376819968223572, + "learning_rate": 4.715874183107324e-06, + "loss": 0.6070502996444702, + "step": 1044 + }, + { + "epoch": 2.319290465631929, + "grad_norm": 1.9359222650527954, + "learning_rate": 4.699626268043911e-06, + "loss": 0.7449045181274414, + "step": 1046 + }, + { + "epoch": 2.3237250554323725, + "grad_norm": 1.128960132598877, + "learning_rate": 4.683389121037618e-06, + "loss": 0.6669731736183167, + "step": 1048 + }, + { + "epoch": 2.328159645232816, + "grad_norm": 0.9983499646186829, + "learning_rate": 4.667162960539552e-06, + "loss": 0.7283903360366821, + "step": 1050 + }, + { + "epoch": 2.3325942350332594, + "grad_norm": 2.680569887161255, + "learning_rate": 4.650948004853006e-06, + "loss": 0.614159107208252, + "step": 1052 + }, + { + "epoch": 2.337028824833703, + "grad_norm": 2.4752540588378906, + "learning_rate": 4.634744472130529e-06, + "loss": 0.4821033477783203, + "step": 1054 + }, + { + "epoch": 2.341463414634146, + "grad_norm": 0.9671631455421448, + "learning_rate": 4.618552580370988e-06, + "loss": 0.6265279054641724, + "step": 1056 + }, + { + "epoch": 2.3458980044345896, + "grad_norm": 0.278334379196167, + "learning_rate": 4.6023725474166324e-06, + "loss": 0.3953332304954529, + "step": 1058 + }, + { + "epoch": 2.3503325942350335, + "grad_norm": 1.0874994993209839, + "learning_rate": 4.586204590950169e-06, + "loss": 0.8931505084037781, + "step": 1060 + }, + { + "epoch": 2.354767184035477, + "grad_norm": 3.5059680938720703, + "learning_rate": 4.570048928491824e-06, + "loss": 0.37978875637054443, + "step": 1062 + }, + { + "epoch": 2.3592017738359203, + "grad_norm": 1.128374695777893, + "learning_rate": 4.5539057773964316e-06, + "loss": 0.4617552161216736, + "step": 1064 + }, + { + "epoch": 2.3636363636363638, + "grad_norm": 2.8418750762939453, + "learning_rate": 4.537775354850496e-06, + "loss": 0.6248428821563721, + "step": 1066 + }, + { + "epoch": 2.368070953436807, + "grad_norm": 1.0952420234680176, + "learning_rate": 4.5216578778692725e-06, + "loss": 0.7397058010101318, + "step": 1068 + }, + { + "epoch": 2.3725055432372506, + "grad_norm": 1.1010463237762451, + "learning_rate": 4.5055535632938526e-06, + "loss": 0.6921043395996094, + "step": 1070 + }, + { + "epoch": 2.376940133037694, + "grad_norm": 1.206845998764038, + "learning_rate": 4.489462627788242e-06, + "loss": 0.8171138167381287, + "step": 1072 + }, + { + "epoch": 2.3813747228381374, + "grad_norm": 0.2211584895849228, + "learning_rate": 4.473385287836448e-06, + "loss": 0.034517209976911545, + "step": 1074 + }, + { + "epoch": 2.385809312638581, + "grad_norm": 1.39876389503479, + "learning_rate": 4.457321759739567e-06, + "loss": 0.4455287456512451, + "step": 1076 + }, + { + "epoch": 2.3902439024390243, + "grad_norm": 0.5344479084014893, + "learning_rate": 4.4412722596128686e-06, + "loss": 0.2573848068714142, + "step": 1078 + }, + { + "epoch": 2.3946784922394677, + "grad_norm": 3.3447608947753906, + "learning_rate": 4.425237003382903e-06, + "loss": 0.6614237427711487, + "step": 1080 + }, + { + "epoch": 2.399113082039911, + "grad_norm": 0.2269127368927002, + "learning_rate": 4.409216206784577e-06, + "loss": 0.4690076410770416, + "step": 1082 + }, + { + "epoch": 2.4035476718403546, + "grad_norm": 0.7799621820449829, + "learning_rate": 4.393210085358265e-06, + "loss": 0.5260664224624634, + "step": 1084 + }, + { + "epoch": 2.4079822616407984, + "grad_norm": 0.9041287899017334, + "learning_rate": 4.3772188544469016e-06, + "loss": 0.71802818775177, + "step": 1086 + }, + { + "epoch": 2.412416851441242, + "grad_norm": 0.303363174200058, + "learning_rate": 4.3612427291930915e-06, + "loss": 0.20643645524978638, + "step": 1088 + }, + { + "epoch": 2.4168514412416853, + "grad_norm": 2.20377779006958, + "learning_rate": 4.345281924536208e-06, + "loss": 0.7628622651100159, + "step": 1090 + }, + { + "epoch": 2.4212860310421287, + "grad_norm": 1.9586279392242432, + "learning_rate": 4.329336655209505e-06, + "loss": 0.6242840886116028, + "step": 1092 + }, + { + "epoch": 2.425720620842572, + "grad_norm": 2.349154233932495, + "learning_rate": 4.31340713573723e-06, + "loss": 0.4106002748012543, + "step": 1094 + }, + { + "epoch": 2.4301552106430155, + "grad_norm": 4.811046600341797, + "learning_rate": 4.297493580431732e-06, + "loss": 0.4525107741355896, + "step": 1096 + }, + { + "epoch": 2.434589800443459, + "grad_norm": 1.139159917831421, + "learning_rate": 4.281596203390582e-06, + "loss": 0.43235841393470764, + "step": 1098 + }, + { + "epoch": 2.4390243902439024, + "grad_norm": 2.029642343521118, + "learning_rate": 4.265715218493695e-06, + "loss": 0.6632136702537537, + "step": 1100 + }, + { + "epoch": 2.443458980044346, + "grad_norm": 0.7346834540367126, + "learning_rate": 4.249850839400446e-06, + "loss": 0.8561656475067139, + "step": 1102 + }, + { + "epoch": 2.4478935698447892, + "grad_norm": 1.7342432737350464, + "learning_rate": 4.2340032795468e-06, + "loss": 0.3280484080314636, + "step": 1104 + }, + { + "epoch": 2.4523281596452327, + "grad_norm": 1.2294172048568726, + "learning_rate": 4.218172752142442e-06, + "loss": 0.8532360792160034, + "step": 1106 + }, + { + "epoch": 2.4567627494456765, + "grad_norm": 2.7901058197021484, + "learning_rate": 4.202359470167903e-06, + "loss": 0.6427351236343384, + "step": 1108 + }, + { + "epoch": 2.4611973392461195, + "grad_norm": 1.3930901288986206, + "learning_rate": 4.186563646371696e-06, + "loss": 0.7979812622070312, + "step": 1110 + }, + { + "epoch": 2.4656319290465634, + "grad_norm": 0.6308720111846924, + "learning_rate": 4.170785493267463e-06, + "loss": 0.5055820345878601, + "step": 1112 + }, + { + "epoch": 2.470066518847007, + "grad_norm": 1.5060687065124512, + "learning_rate": 4.155025223131102e-06, + "loss": 0.7073782086372375, + "step": 1114 + }, + { + "epoch": 2.47450110864745, + "grad_norm": 1.2661110162734985, + "learning_rate": 4.139283047997919e-06, + "loss": 0.2692304849624634, + "step": 1116 + }, + { + "epoch": 2.4789356984478936, + "grad_norm": 0.8920889496803284, + "learning_rate": 4.123559179659771e-06, + "loss": 0.6808326840400696, + "step": 1118 + }, + { + "epoch": 2.483370288248337, + "grad_norm": 1.1632484197616577, + "learning_rate": 4.107853829662224e-06, + "loss": 0.6864634156227112, + "step": 1120 + }, + { + "epoch": 2.4878048780487805, + "grad_norm": 1.0989418029785156, + "learning_rate": 4.0921672093017e-06, + "loss": 0.6558045148849487, + "step": 1122 + }, + { + "epoch": 2.492239467849224, + "grad_norm": 2.8047521114349365, + "learning_rate": 4.076499529622636e-06, + "loss": 0.8816790580749512, + "step": 1124 + }, + { + "epoch": 2.4966740576496673, + "grad_norm": 0.8200728297233582, + "learning_rate": 4.0608510014146455e-06, + "loss": 0.7856003642082214, + "step": 1126 + }, + { + "epoch": 2.5011086474501107, + "grad_norm": 1.6482559442520142, + "learning_rate": 4.045221835209684e-06, + "loss": 0.5203614234924316, + "step": 1128 + }, + { + "epoch": 2.505543237250554, + "grad_norm": 1.5657063722610474, + "learning_rate": 4.02961224127921e-06, + "loss": 0.5360028147697449, + "step": 1130 + }, + { + "epoch": 2.5099778270509976, + "grad_norm": 1.1746268272399902, + "learning_rate": 4.014022429631368e-06, + "loss": 0.6573871970176697, + "step": 1132 + }, + { + "epoch": 2.5144124168514415, + "grad_norm": 1.0907959938049316, + "learning_rate": 3.998452610008147e-06, + "loss": 0.3955955505371094, + "step": 1134 + }, + { + "epoch": 2.5188470066518844, + "grad_norm": 0.861308217048645, + "learning_rate": 3.982902991882578e-06, + "loss": 0.7564470767974854, + "step": 1136 + }, + { + "epoch": 2.5232815964523283, + "grad_norm": 0.7468408346176147, + "learning_rate": 3.967373784455896e-06, + "loss": 0.6149483919143677, + "step": 1138 + }, + { + "epoch": 2.5277161862527717, + "grad_norm": 0.710243284702301, + "learning_rate": 3.951865196654738e-06, + "loss": 0.8047510385513306, + "step": 1140 + }, + { + "epoch": 2.532150776053215, + "grad_norm": 2.8993804454803467, + "learning_rate": 3.936377437128329e-06, + "loss": 0.41506367921829224, + "step": 1142 + }, + { + "epoch": 2.5365853658536586, + "grad_norm": 0.2720961272716522, + "learning_rate": 3.920910714245679e-06, + "loss": 0.44911813735961914, + "step": 1144 + }, + { + "epoch": 2.541019955654102, + "grad_norm": 0.7515284419059753, + "learning_rate": 3.905465236092771e-06, + "loss": 0.7769864201545715, + "step": 1146 + }, + { + "epoch": 2.5454545454545454, + "grad_norm": 0.7892196178436279, + "learning_rate": 3.890041210469765e-06, + "loss": 0.7185046076774597, + "step": 1148 + }, + { + "epoch": 2.549889135254989, + "grad_norm": 0.7974555492401123, + "learning_rate": 3.8746388448882055e-06, + "loss": 0.6233813166618347, + "step": 1150 + }, + { + "epoch": 2.5543237250554323, + "grad_norm": 0.8632022142410278, + "learning_rate": 3.859258346568228e-06, + "loss": 0.44528669118881226, + "step": 1152 + }, + { + "epoch": 2.5587583148558757, + "grad_norm": 0.21908880770206451, + "learning_rate": 3.843899922435767e-06, + "loss": 0.2759099304676056, + "step": 1154 + }, + { + "epoch": 2.5631929046563195, + "grad_norm": 0.6866997480392456, + "learning_rate": 3.8285637791197815e-06, + "loss": 0.5508578419685364, + "step": 1156 + }, + { + "epoch": 2.5676274944567625, + "grad_norm": 2.519130229949951, + "learning_rate": 3.8132501229494635e-06, + "loss": 0.5549399852752686, + "step": 1158 + }, + { + "epoch": 2.5720620842572064, + "grad_norm": 2.8596558570861816, + "learning_rate": 3.7979591599514696e-06, + "loss": 0.6041897535324097, + "step": 1160 + }, + { + "epoch": 2.57649667405765, + "grad_norm": 0.7755621671676636, + "learning_rate": 3.782691095847151e-06, + "loss": 0.7620603442192078, + "step": 1162 + }, + { + "epoch": 2.5809312638580932, + "grad_norm": 0.9750792980194092, + "learning_rate": 3.767446136049775e-06, + "loss": 0.939260721206665, + "step": 1164 + }, + { + "epoch": 2.5853658536585367, + "grad_norm": 1.4371932744979858, + "learning_rate": 3.752224485661775e-06, + "loss": 0.4603317975997925, + "step": 1166 + }, + { + "epoch": 2.58980044345898, + "grad_norm": 1.9136115312576294, + "learning_rate": 3.7370263494719805e-06, + "loss": 0.9075461626052856, + "step": 1168 + }, + { + "epoch": 2.5942350332594235, + "grad_norm": 1.7198249101638794, + "learning_rate": 3.721851931952869e-06, + "loss": 0.8085947632789612, + "step": 1170 + }, + { + "epoch": 2.598669623059867, + "grad_norm": 2.1932644844055176, + "learning_rate": 3.706701437257808e-06, + "loss": 0.28652137517929077, + "step": 1172 + }, + { + "epoch": 2.6031042128603104, + "grad_norm": 0.7438820600509644, + "learning_rate": 3.691575069218314e-06, + "loss": 0.5113945007324219, + "step": 1174 + }, + { + "epoch": 2.6075388026607538, + "grad_norm": 0.860270619392395, + "learning_rate": 3.676473031341313e-06, + "loss": 0.6308207511901855, + "step": 1176 + }, + { + "epoch": 2.611973392461197, + "grad_norm": 1.0259983539581299, + "learning_rate": 3.661395526806395e-06, + "loss": 0.4299200773239136, + "step": 1178 + }, + { + "epoch": 2.6164079822616406, + "grad_norm": 2.737011194229126, + "learning_rate": 3.6463427584630806e-06, + "loss": 0.5992394089698792, + "step": 1180 + }, + { + "epoch": 2.6208425720620845, + "grad_norm": 0.8241181373596191, + "learning_rate": 3.631314928828099e-06, + "loss": 0.7023974061012268, + "step": 1182 + }, + { + "epoch": 2.6252771618625275, + "grad_norm": 0.6770392656326294, + "learning_rate": 3.616312240082659e-06, + "loss": 0.7103127241134644, + "step": 1184 + }, + { + "epoch": 2.6297117516629713, + "grad_norm": 1.7923256158828735, + "learning_rate": 3.601334894069728e-06, + "loss": 0.8160955309867859, + "step": 1186 + }, + { + "epoch": 2.6341463414634148, + "grad_norm": 2.411703586578369, + "learning_rate": 3.5863830922913147e-06, + "loss": 0.8449252247810364, + "step": 1188 + }, + { + "epoch": 2.638580931263858, + "grad_norm": 0.7744673490524292, + "learning_rate": 3.5714570359057676e-06, + "loss": 0.4491943418979645, + "step": 1190 + }, + { + "epoch": 2.6430155210643016, + "grad_norm": 2.5860702991485596, + "learning_rate": 3.556556925725061e-06, + "loss": 0.5912431478500366, + "step": 1192 + }, + { + "epoch": 2.647450110864745, + "grad_norm": 3.4122977256774902, + "learning_rate": 3.5416829622120875e-06, + "loss": 0.5446506142616272, + "step": 1194 + }, + { + "epoch": 2.6518847006651884, + "grad_norm": 1.6620879173278809, + "learning_rate": 3.526835345477978e-06, + "loss": 0.6308864951133728, + "step": 1196 + }, + { + "epoch": 2.656319290465632, + "grad_norm": 0.7548476457595825, + "learning_rate": 3.5120142752793907e-06, + "loss": 0.21907749772071838, + "step": 1198 + }, + { + "epoch": 2.6607538802660753, + "grad_norm": 3.578019380569458, + "learning_rate": 3.4972199510158393e-06, + "loss": 0.9215325117111206, + "step": 1200 + }, + { + "epoch": 2.6651884700665187, + "grad_norm": 0.7092931270599365, + "learning_rate": 3.4824525717269975e-06, + "loss": 0.8297696709632874, + "step": 1202 + }, + { + "epoch": 2.6696230598669626, + "grad_norm": 0.9896040558815002, + "learning_rate": 3.4677123360900342e-06, + "loss": 0.38038522005081177, + "step": 1204 + }, + { + "epoch": 2.6740576496674056, + "grad_norm": 2.033034563064575, + "learning_rate": 3.4529994424169233e-06, + "loss": 0.5968571901321411, + "step": 1206 + }, + { + "epoch": 2.6784922394678494, + "grad_norm": 2.7922661304473877, + "learning_rate": 3.4383140886517953e-06, + "loss": 0.7829728722572327, + "step": 1208 + }, + { + "epoch": 2.682926829268293, + "grad_norm": 1.3708895444869995, + "learning_rate": 3.423656472368262e-06, + "loss": 0.4674299359321594, + "step": 1210 + }, + { + "epoch": 2.6873614190687363, + "grad_norm": 1.8650954961776733, + "learning_rate": 3.409026790766756e-06, + "loss": 0.25274747610092163, + "step": 1212 + }, + { + "epoch": 2.6917960088691797, + "grad_norm": 2.8033974170684814, + "learning_rate": 3.394425240671891e-06, + "loss": 0.4370385706424713, + "step": 1214 + }, + { + "epoch": 2.696230598669623, + "grad_norm": 2.1162285804748535, + "learning_rate": 3.379852018529799e-06, + "loss": 0.5205950736999512, + "step": 1216 + }, + { + "epoch": 2.7006651884700665, + "grad_norm": 1.2731273174285889, + "learning_rate": 3.3653073204054942e-06, + "loss": 0.5236338973045349, + "step": 1218 + }, + { + "epoch": 2.70509977827051, + "grad_norm": 1.8009387254714966, + "learning_rate": 3.3507913419802403e-06, + "loss": 0.7941880822181702, + "step": 1220 + }, + { + "epoch": 2.7095343680709534, + "grad_norm": 1.7554905414581299, + "learning_rate": 3.336304278548903e-06, + "loss": 0.7005539536476135, + "step": 1222 + }, + { + "epoch": 2.713968957871397, + "grad_norm": 1.5743012428283691, + "learning_rate": 3.321846325017342e-06, + "loss": 0.7519204616546631, + "step": 1224 + }, + { + "epoch": 2.7184035476718402, + "grad_norm": 3.2630934715270996, + "learning_rate": 3.3074176758997744e-06, + "loss": 0.37882906198501587, + "step": 1226 + }, + { + "epoch": 2.7228381374722836, + "grad_norm": 0.9136330485343933, + "learning_rate": 3.2930185253161574e-06, + "loss": 0.8159320950508118, + "step": 1228 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 1.3966491222381592, + "learning_rate": 3.2786490669895883e-06, + "loss": 0.6657707095146179, + "step": 1230 + }, + { + "epoch": 2.7317073170731705, + "grad_norm": 0.7242875099182129, + "learning_rate": 3.2643094942436865e-06, + "loss": 0.7183330655097961, + "step": 1232 + }, + { + "epoch": 2.7361419068736144, + "grad_norm": 2.514469623565674, + "learning_rate": 3.2500000000000015e-06, + "loss": 0.4106196463108063, + "step": 1234 + }, + { + "epoch": 2.740576496674058, + "grad_norm": 1.027852177619934, + "learning_rate": 3.2357207767754063e-06, + "loss": 0.6766651272773743, + "step": 1236 + }, + { + "epoch": 2.745011086474501, + "grad_norm": 1.0152777433395386, + "learning_rate": 3.221472016679521e-06, + "loss": 0.47056448459625244, + "step": 1238 + }, + { + "epoch": 2.7494456762749446, + "grad_norm": 0.7261103987693787, + "learning_rate": 3.2072539114121188e-06, + "loss": 0.46467670798301697, + "step": 1240 + }, + { + "epoch": 2.753880266075388, + "grad_norm": 0.899706244468689, + "learning_rate": 3.193066652260547e-06, + "loss": 0.8382993340492249, + "step": 1242 + }, + { + "epoch": 2.7583148558758315, + "grad_norm": 0.9996641874313354, + "learning_rate": 3.1789104300971603e-06, + "loss": 0.7458208203315735, + "step": 1244 + }, + { + "epoch": 2.762749445676275, + "grad_norm": 0.36415642499923706, + "learning_rate": 3.164785435376745e-06, + "loss": 0.23638926446437836, + "step": 1246 + }, + { + "epoch": 2.7671840354767183, + "grad_norm": 1.592871904373169, + "learning_rate": 3.1506918581339583e-06, + "loss": 0.47278061509132385, + "step": 1248 + }, + { + "epoch": 2.7716186252771617, + "grad_norm": 1.0408028364181519, + "learning_rate": 3.136629887980781e-06, + "loss": 0.5122473835945129, + "step": 1250 + }, + { + "epoch": 2.776053215077605, + "grad_norm": 0.9867532849311829, + "learning_rate": 3.122599714103949e-06, + "loss": 0.8818725347518921, + "step": 1252 + }, + { + "epoch": 2.7804878048780486, + "grad_norm": 1.422898769378662, + "learning_rate": 3.1086015252624257e-06, + "loss": 0.8071056604385376, + "step": 1254 + }, + { + "epoch": 2.7849223946784925, + "grad_norm": 1.5934648513793945, + "learning_rate": 3.0946355097848535e-06, + "loss": 0.7926267385482788, + "step": 1256 + }, + { + "epoch": 2.7893569844789354, + "grad_norm": 0.5067124962806702, + "learning_rate": 3.0807018555670153e-06, + "loss": 0.13775405287742615, + "step": 1258 + }, + { + "epoch": 2.7937915742793793, + "grad_norm": 1.111436128616333, + "learning_rate": 3.0668007500693216e-06, + "loss": 0.7149184346199036, + "step": 1260 + }, + { + "epoch": 2.7982261640798227, + "grad_norm": 2.614287853240967, + "learning_rate": 3.0529323803142697e-06, + "loss": 0.5375425815582275, + "step": 1262 + }, + { + "epoch": 2.802660753880266, + "grad_norm": 1.180129885673523, + "learning_rate": 3.0390969328839464e-06, + "loss": 0.48145541548728943, + "step": 1264 + }, + { + "epoch": 2.8070953436807096, + "grad_norm": 1.0282138586044312, + "learning_rate": 3.0252945939175004e-06, + "loss": 0.7159358859062195, + "step": 1266 + }, + { + "epoch": 2.811529933481153, + "grad_norm": 3.18182635307312, + "learning_rate": 3.0115255491086537e-06, + "loss": 0.8956208229064941, + "step": 1268 + }, + { + "epoch": 2.8159645232815964, + "grad_norm": 0.7246714234352112, + "learning_rate": 2.9977899837031895e-06, + "loss": 0.7317441701889038, + "step": 1270 + }, + { + "epoch": 2.82039911308204, + "grad_norm": 1.5273503065109253, + "learning_rate": 2.984088082496469e-06, + "loss": 0.2754761874675751, + "step": 1272 + }, + { + "epoch": 2.8248337028824833, + "grad_norm": 0.9987423419952393, + "learning_rate": 2.970420029830946e-06, + "loss": 0.4880000650882721, + "step": 1274 + }, + { + "epoch": 2.8292682926829267, + "grad_norm": 0.7692528367042542, + "learning_rate": 2.9567860095936775e-06, + "loss": 0.9671233892440796, + "step": 1276 + }, + { + "epoch": 2.8337028824833705, + "grad_norm": 0.9331156015396118, + "learning_rate": 2.9431862052138545e-06, + "loss": 0.8612651824951172, + "step": 1278 + }, + { + "epoch": 2.8381374722838135, + "grad_norm": 0.9668262600898743, + "learning_rate": 2.929620799660343e-06, + "loss": 0.3867911994457245, + "step": 1280 + }, + { + "epoch": 2.8425720620842574, + "grad_norm": 2.516414165496826, + "learning_rate": 2.916089975439207e-06, + "loss": 0.47361209988594055, + "step": 1282 + }, + { + "epoch": 2.847006651884701, + "grad_norm": 0.9498918056488037, + "learning_rate": 2.9025939145912655e-06, + "loss": 0.4672809839248657, + "step": 1284 + }, + { + "epoch": 2.8514412416851442, + "grad_norm": 0.8674972653388977, + "learning_rate": 2.8891327986896345e-06, + "loss": 0.8502570390701294, + "step": 1286 + }, + { + "epoch": 2.8558758314855877, + "grad_norm": 2.823939085006714, + "learning_rate": 2.875706808837292e-06, + "loss": 0.1998748630285263, + "step": 1288 + }, + { + "epoch": 2.860310421286031, + "grad_norm": 1.9392937421798706, + "learning_rate": 2.862316125664636e-06, + "loss": 0.8196284770965576, + "step": 1290 + }, + { + "epoch": 2.8647450110864745, + "grad_norm": 8.883353233337402, + "learning_rate": 2.848960929327053e-06, + "loss": 0.7396450042724609, + "step": 1292 + }, + { + "epoch": 2.869179600886918, + "grad_norm": 1.4906433820724487, + "learning_rate": 2.8356413995025044e-06, + "loss": 0.717079758644104, + "step": 1294 + }, + { + "epoch": 2.8736141906873613, + "grad_norm": 0.8910164833068848, + "learning_rate": 2.8223577153890934e-06, + "loss": 0.7069391012191772, + "step": 1296 + }, + { + "epoch": 2.8780487804878048, + "grad_norm": 3.354166030883789, + "learning_rate": 2.8091100557026702e-06, + "loss": 0.4838540852069855, + "step": 1298 + }, + { + "epoch": 2.882483370288248, + "grad_norm": 0.841211199760437, + "learning_rate": 2.795898598674415e-06, + "loss": 0.6780248284339905, + "step": 1300 + }, + { + "epoch": 2.8869179600886916, + "grad_norm": 1.9939450025558472, + "learning_rate": 2.782723522048444e-06, + "loss": 0.2901532053947449, + "step": 1302 + }, + { + "epoch": 2.8913525498891355, + "grad_norm": 1.6043404340744019, + "learning_rate": 2.7695850030794293e-06, + "loss": 0.7313271760940552, + "step": 1304 + }, + { + "epoch": 2.8957871396895785, + "grad_norm": 2.5174219608306885, + "learning_rate": 2.7564832185301915e-06, + "loss": 0.6357086896896362, + "step": 1306 + }, + { + "epoch": 2.9002217294900223, + "grad_norm": 0.29980650544166565, + "learning_rate": 2.7434183446693397e-06, + "loss": 0.2599072754383087, + "step": 1308 + }, + { + "epoch": 2.9046563192904657, + "grad_norm": 1.3467071056365967, + "learning_rate": 2.730390557268897e-06, + "loss": 0.33962565660476685, + "step": 1310 + }, + { + "epoch": 2.909090909090909, + "grad_norm": 0.9841307401657104, + "learning_rate": 2.7174000316019277e-06, + "loss": 0.6833657622337341, + "step": 1312 + }, + { + "epoch": 2.9135254988913526, + "grad_norm": 1.2662699222564697, + "learning_rate": 2.704446942440191e-06, + "loss": 0.7377205491065979, + "step": 1314 + }, + { + "epoch": 2.917960088691796, + "grad_norm": 2.137340784072876, + "learning_rate": 2.6915314640517755e-06, + "loss": 0.6367099285125732, + "step": 1316 + }, + { + "epoch": 2.9223946784922394, + "grad_norm": 1.5056116580963135, + "learning_rate": 2.6786537701987703e-06, + "loss": 0.8529772162437439, + "step": 1318 + }, + { + "epoch": 2.926829268292683, + "grad_norm": 1.2501715421676636, + "learning_rate": 2.665814034134916e-06, + "loss": 0.5121623873710632, + "step": 1320 + }, + { + "epoch": 2.9312638580931263, + "grad_norm": 1.0894711017608643, + "learning_rate": 2.6530124286032755e-06, + "loss": 0.7275009155273438, + "step": 1322 + }, + { + "epoch": 2.9356984478935697, + "grad_norm": 3.4206230640411377, + "learning_rate": 2.640249125833915e-06, + "loss": 0.7550503611564636, + "step": 1324 + }, + { + "epoch": 2.9401330376940136, + "grad_norm": 0.8373146653175354, + "learning_rate": 2.6275242975415804e-06, + "loss": 0.8755195736885071, + "step": 1326 + }, + { + "epoch": 2.9445676274944566, + "grad_norm": 1.1842820644378662, + "learning_rate": 2.614838114923394e-06, + "loss": 0.7242047786712646, + "step": 1328 + }, + { + "epoch": 2.9490022172949004, + "grad_norm": 0.9061135053634644, + "learning_rate": 2.6021907486565447e-06, + "loss": 0.7160645127296448, + "step": 1330 + }, + { + "epoch": 2.953436807095344, + "grad_norm": 1.9043796062469482, + "learning_rate": 2.589582368895992e-06, + "loss": 0.7991137504577637, + "step": 1332 + }, + { + "epoch": 2.9578713968957873, + "grad_norm": 0.7662011384963989, + "learning_rate": 2.577013145272185e-06, + "loss": 0.42025405168533325, + "step": 1334 + }, + { + "epoch": 2.9623059866962307, + "grad_norm": 0.9441994428634644, + "learning_rate": 2.564483246888772e-06, + "loss": 0.6747118830680847, + "step": 1336 + }, + { + "epoch": 2.966740576496674, + "grad_norm": 1.0354760885238647, + "learning_rate": 2.5519928423203266e-06, + "loss": 0.7479525208473206, + "step": 1338 + }, + { + "epoch": 2.9711751662971175, + "grad_norm": 0.9520703554153442, + "learning_rate": 2.539542099610084e-06, + "loss": 0.5493751168251038, + "step": 1340 + }, + { + "epoch": 2.975609756097561, + "grad_norm": 2.1422736644744873, + "learning_rate": 2.5271311862676727e-06, + "loss": 0.5847128033638, + "step": 1342 + }, + { + "epoch": 2.9800443458980044, + "grad_norm": 0.5576608777046204, + "learning_rate": 2.514760269266871e-06, + "loss": 0.1638008952140808, + "step": 1344 + }, + { + "epoch": 2.984478935698448, + "grad_norm": 1.6136894226074219, + "learning_rate": 2.50242951504335e-06, + "loss": 0.5379747748374939, + "step": 1346 + }, + { + "epoch": 2.988913525498891, + "grad_norm": 0.31877100467681885, + "learning_rate": 2.490139089492443e-06, + "loss": 0.42808306217193604, + "step": 1348 + }, + { + "epoch": 2.9933481152993346, + "grad_norm": 2.9408724308013916, + "learning_rate": 2.4778891579669067e-06, + "loss": 0.42848342657089233, + "step": 1350 + }, + { + "epoch": 2.9977827050997785, + "grad_norm": 1.016264796257019, + "learning_rate": 2.4656798852747023e-06, + "loss": 0.5269397497177124, + "step": 1352 + }, + { + "epoch": 3.002217294900222, + "grad_norm": 1.1126813888549805, + "learning_rate": 2.453511435676777e-06, + "loss": 0.557039737701416, + "step": 1354 + }, + { + "epoch": 3.0066518847006654, + "grad_norm": 1.4688986539840698, + "learning_rate": 2.441383972884848e-06, + "loss": 0.2774271070957184, + "step": 1356 + }, + { + "epoch": 3.011086474501109, + "grad_norm": 1.5381768941879272, + "learning_rate": 2.4292976600592095e-06, + "loss": 0.45656082034111023, + "step": 1358 + }, + { + "epoch": 3.015521064301552, + "grad_norm": 0.8177427649497986, + "learning_rate": 2.4172526598065304e-06, + "loss": 0.5471428036689758, + "step": 1360 + }, + { + "epoch": 3.0199556541019956, + "grad_norm": 1.1968615055084229, + "learning_rate": 2.4052491341776686e-06, + "loss": 0.3854435086250305, + "step": 1362 + }, + { + "epoch": 3.024390243902439, + "grad_norm": 0.8419128656387329, + "learning_rate": 2.393287244665494e-06, + "loss": 0.4754073917865753, + "step": 1364 + }, + { + "epoch": 3.0288248337028825, + "grad_norm": 0.12323111295700073, + "learning_rate": 2.3813671522027094e-06, + "loss": 0.22163067758083344, + "step": 1366 + }, + { + "epoch": 3.033259423503326, + "grad_norm": 1.2366307973861694, + "learning_rate": 2.369489017159692e-06, + "loss": 0.20418155193328857, + "step": 1368 + }, + { + "epoch": 3.0376940133037693, + "grad_norm": 2.100736141204834, + "learning_rate": 2.357652999342334e-06, + "loss": 0.2910291850566864, + "step": 1370 + }, + { + "epoch": 3.0421286031042127, + "grad_norm": 4.004879951477051, + "learning_rate": 2.345859257989886e-06, + "loss": 0.47884419560432434, + "step": 1372 + }, + { + "epoch": 3.046563192904656, + "grad_norm": 1.1280056238174438, + "learning_rate": 2.334107951772826e-06, + "loss": 0.15896357595920563, + "step": 1374 + }, + { + "epoch": 3.0509977827050996, + "grad_norm": 3.1216893196105957, + "learning_rate": 2.3223992387907137e-06, + "loss": 0.23939193785190582, + "step": 1376 + }, + { + "epoch": 3.0554323725055434, + "grad_norm": 1.0042521953582764, + "learning_rate": 2.3107332765700733e-06, + "loss": 0.1343676894903183, + "step": 1378 + }, + { + "epoch": 3.059866962305987, + "grad_norm": 0.09163781255483627, + "learning_rate": 2.2991102220622647e-06, + "loss": 0.16645547747612, + "step": 1380 + }, + { + "epoch": 3.0643015521064303, + "grad_norm": 2.4105827808380127, + "learning_rate": 2.2875302316413807e-06, + "loss": 0.1257064789533615, + "step": 1382 + }, + { + "epoch": 3.0687361419068737, + "grad_norm": 1.94559645652771, + "learning_rate": 2.275993461102138e-06, + "loss": 0.3437502384185791, + "step": 1384 + }, + { + "epoch": 3.073170731707317, + "grad_norm": 0.2491624653339386, + "learning_rate": 2.2645000656577793e-06, + "loss": 0.1117410808801651, + "step": 1386 + }, + { + "epoch": 3.0776053215077606, + "grad_norm": 0.1076858639717102, + "learning_rate": 2.2530501999379932e-06, + "loss": 0.17695897817611694, + "step": 1388 + }, + { + "epoch": 3.082039911308204, + "grad_norm": 0.8501622080802917, + "learning_rate": 2.2416440179868236e-06, + "loss": 0.3754989504814148, + "step": 1390 + }, + { + "epoch": 3.0864745011086474, + "grad_norm": 1.7676734924316406, + "learning_rate": 2.230281673260605e-06, + "loss": 0.15238375961780548, + "step": 1392 + }, + { + "epoch": 3.090909090909091, + "grad_norm": 6.170825958251953, + "learning_rate": 2.218963318625895e-06, + "loss": 0.22011463344097137, + "step": 1394 + }, + { + "epoch": 3.0953436807095343, + "grad_norm": 2.093130350112915, + "learning_rate": 2.2076891063574167e-06, + "loss": 0.5483108162879944, + "step": 1396 + }, + { + "epoch": 3.0997782705099777, + "grad_norm": 0.9943727850914001, + "learning_rate": 2.196459188136014e-06, + "loss": 0.39382025599479675, + "step": 1398 + }, + { + "epoch": 3.104212860310421, + "grad_norm": 0.4041390120983124, + "learning_rate": 2.1852737150466064e-06, + "loss": 0.37452182173728943, + "step": 1400 + }, + { + "epoch": 3.1086474501108645, + "grad_norm": 0.41770124435424805, + "learning_rate": 2.174132837576156e-06, + "loss": 0.1812993884086609, + "step": 1402 + }, + { + "epoch": 3.1130820399113084, + "grad_norm": 3.0523788928985596, + "learning_rate": 2.1630367056116496e-06, + "loss": 0.3220471739768982, + "step": 1404 + }, + { + "epoch": 3.117516629711752, + "grad_norm": 2.8227522373199463, + "learning_rate": 2.1519854684380724e-06, + "loss": 0.5665332674980164, + "step": 1406 + }, + { + "epoch": 3.1219512195121952, + "grad_norm": 1.215554118156433, + "learning_rate": 2.1409792747364103e-06, + "loss": 0.6941906809806824, + "step": 1408 + }, + { + "epoch": 3.1263858093126387, + "grad_norm": 0.9595901370048523, + "learning_rate": 2.1300182725816378e-06, + "loss": 0.34386202692985535, + "step": 1410 + }, + { + "epoch": 3.130820399113082, + "grad_norm": 1.20671808719635, + "learning_rate": 2.1191026094407386e-06, + "loss": 0.40727710723876953, + "step": 1412 + }, + { + "epoch": 3.1352549889135255, + "grad_norm": 0.9301721453666687, + "learning_rate": 2.1082324321707075e-06, + "loss": 0.24828168749809265, + "step": 1414 + }, + { + "epoch": 3.139689578713969, + "grad_norm": 0.29069337248802185, + "learning_rate": 2.0974078870165882e-06, + "loss": 0.02273000217974186, + "step": 1416 + }, + { + "epoch": 3.1441241685144123, + "grad_norm": 1.1463552713394165, + "learning_rate": 2.086629119609499e-06, + "loss": 0.4149464964866638, + "step": 1418 + }, + { + "epoch": 3.1485587583148558, + "grad_norm": 0.8834893107414246, + "learning_rate": 2.0758962749646716e-06, + "loss": 0.47038036584854126, + "step": 1420 + }, + { + "epoch": 3.152993348115299, + "grad_norm": 3.7323062419891357, + "learning_rate": 2.065209497479502e-06, + "loss": 0.34102943539619446, + "step": 1422 + }, + { + "epoch": 3.1574279379157426, + "grad_norm": 0.32775676250457764, + "learning_rate": 2.0545689309316138e-06, + "loss": 0.06301730126142502, + "step": 1424 + }, + { + "epoch": 3.1618625277161865, + "grad_norm": 1.2714399099349976, + "learning_rate": 2.043974718476911e-06, + "loss": 0.4001501798629761, + "step": 1426 + }, + { + "epoch": 3.16629711751663, + "grad_norm": 1.573042631149292, + "learning_rate": 2.033427002647668e-06, + "loss": 0.37112024426460266, + "step": 1428 + }, + { + "epoch": 3.1707317073170733, + "grad_norm": 1.6631622314453125, + "learning_rate": 2.0229259253505946e-06, + "loss": 0.2901914417743683, + "step": 1430 + }, + { + "epoch": 3.1751662971175167, + "grad_norm": 1.1970975399017334, + "learning_rate": 2.012471627864943e-06, + "loss": 0.29324379563331604, + "step": 1432 + }, + { + "epoch": 3.17960088691796, + "grad_norm": 1.13413667678833, + "learning_rate": 2.0020642508405984e-06, + "loss": 0.3578546643257141, + "step": 1434 + }, + { + "epoch": 3.1840354767184036, + "grad_norm": 2.5074386596679688, + "learning_rate": 1.9917039342961837e-06, + "loss": 0.0840587466955185, + "step": 1436 + }, + { + "epoch": 3.188470066518847, + "grad_norm": 0.9287826418876648, + "learning_rate": 1.9813908176171857e-06, + "loss": 0.27323436737060547, + "step": 1438 + }, + { + "epoch": 3.1929046563192904, + "grad_norm": 2.366941213607788, + "learning_rate": 1.97112503955407e-06, + "loss": 0.43178707361221313, + "step": 1440 + }, + { + "epoch": 3.197339246119734, + "grad_norm": 1.5955890417099, + "learning_rate": 1.9609067382204224e-06, + "loss": 0.47529783844947815, + "step": 1442 + }, + { + "epoch": 3.2017738359201773, + "grad_norm": 1.0811793804168701, + "learning_rate": 1.950736051091084e-06, + "loss": 0.4838941693305969, + "step": 1444 + }, + { + "epoch": 3.2062084257206207, + "grad_norm": 1.8544070720672607, + "learning_rate": 1.9406131150003036e-06, + "loss": 0.37333235144615173, + "step": 1446 + }, + { + "epoch": 3.210643015521064, + "grad_norm": 7.789588451385498, + "learning_rate": 1.930538066139904e-06, + "loss": 0.1161646842956543, + "step": 1448 + }, + { + "epoch": 3.2150776053215075, + "grad_norm": 0.8985663056373596, + "learning_rate": 1.9205110400574368e-06, + "loss": 0.5245546698570251, + "step": 1450 + }, + { + "epoch": 3.2195121951219514, + "grad_norm": 1.785593867301941, + "learning_rate": 1.910532171654367e-06, + "loss": 0.23958845436573029, + "step": 1452 + }, + { + "epoch": 3.223946784922395, + "grad_norm": 1.4889003038406372, + "learning_rate": 1.9006015951842587e-06, + "loss": 0.3209075331687927, + "step": 1454 + }, + { + "epoch": 3.2283813747228383, + "grad_norm": 3.653893232345581, + "learning_rate": 1.8907194442509642e-06, + "loss": 0.43527886271476746, + "step": 1456 + }, + { + "epoch": 3.2328159645232817, + "grad_norm": 2.4832544326782227, + "learning_rate": 1.8808858518068312e-06, + "loss": 0.2721869647502899, + "step": 1458 + }, + { + "epoch": 3.237250554323725, + "grad_norm": 1.3052841424942017, + "learning_rate": 1.8711009501509087e-06, + "loss": 0.44211310148239136, + "step": 1460 + }, + { + "epoch": 3.2416851441241685, + "grad_norm": 1.3836771249771118, + "learning_rate": 1.8613648709271732e-06, + "loss": 0.26933524012565613, + "step": 1462 + }, + { + "epoch": 3.246119733924612, + "grad_norm": 3.1456692218780518, + "learning_rate": 1.8516777451227552e-06, + "loss": 0.39066338539123535, + "step": 1464 + }, + { + "epoch": 3.2505543237250554, + "grad_norm": 1.4456260204315186, + "learning_rate": 1.842039703066172e-06, + "loss": 0.47105956077575684, + "step": 1466 + }, + { + "epoch": 3.254988913525499, + "grad_norm": 0.20494961738586426, + "learning_rate": 1.8324508744255842e-06, + "loss": 0.039846230298280716, + "step": 1468 + }, + { + "epoch": 3.259423503325942, + "grad_norm": 2.348210573196411, + "learning_rate": 1.8229113882070398e-06, + "loss": 0.310930460691452, + "step": 1470 + }, + { + "epoch": 3.2638580931263856, + "grad_norm": 1.5028283596038818, + "learning_rate": 1.8134213727527504e-06, + "loss": 0.420907586812973, + "step": 1472 + }, + { + "epoch": 3.2682926829268295, + "grad_norm": 0.636045515537262, + "learning_rate": 1.803980955739354e-06, + "loss": 0.08325402438640594, + "step": 1474 + }, + { + "epoch": 3.2727272727272725, + "grad_norm": 0.7118128538131714, + "learning_rate": 1.7945902641762027e-06, + "loss": 0.3166371285915375, + "step": 1476 + }, + { + "epoch": 3.2771618625277164, + "grad_norm": 1.0103552341461182, + "learning_rate": 1.785249424403654e-06, + "loss": 0.4005421996116638, + "step": 1478 + }, + { + "epoch": 3.2815964523281598, + "grad_norm": 0.3991997539997101, + "learning_rate": 1.7759585620913723e-06, + "loss": 0.27822190523147583, + "step": 1480 + }, + { + "epoch": 3.286031042128603, + "grad_norm": 0.9653980731964111, + "learning_rate": 1.7667178022366294e-06, + "loss": 0.4057963192462921, + "step": 1482 + }, + { + "epoch": 3.2904656319290466, + "grad_norm": 8.63097095489502, + "learning_rate": 1.757527269162636e-06, + "loss": 0.327663779258728, + "step": 1484 + }, + { + "epoch": 3.29490022172949, + "grad_norm": 0.9608578681945801, + "learning_rate": 1.7483870865168585e-06, + "loss": 0.3130677342414856, + "step": 1486 + }, + { + "epoch": 3.2993348115299335, + "grad_norm": 2.4934980869293213, + "learning_rate": 1.739297377269361e-06, + "loss": 0.18216699361801147, + "step": 1488 + }, + { + "epoch": 3.303769401330377, + "grad_norm": 2.835766553878784, + "learning_rate": 1.730258263711149e-06, + "loss": 0.1408637911081314, + "step": 1490 + }, + { + "epoch": 3.3082039911308203, + "grad_norm": 1.3449338674545288, + "learning_rate": 1.7212698674525246e-06, + "loss": 0.3613136410713196, + "step": 1492 + }, + { + "epoch": 3.3126385809312637, + "grad_norm": 1.2930020093917847, + "learning_rate": 1.7123323094214485e-06, + "loss": 0.3133498728275299, + "step": 1494 + }, + { + "epoch": 3.317073170731707, + "grad_norm": 2.178227663040161, + "learning_rate": 1.7034457098619176e-06, + "loss": 0.329238623380661, + "step": 1496 + }, + { + "epoch": 3.3215077605321506, + "grad_norm": 0.9758614301681519, + "learning_rate": 1.6946101883323435e-06, + "loss": 0.5130437016487122, + "step": 1498 + }, + { + "epoch": 3.3259423503325944, + "grad_norm": 0.9882797002792358, + "learning_rate": 1.6858258637039421e-06, + "loss": 0.3322954475879669, + "step": 1500 + }, + { + "epoch": 3.330376940133038, + "grad_norm": 1.1139705181121826, + "learning_rate": 1.677092854159142e-06, + "loss": 0.37578579783439636, + "step": 1502 + }, + { + "epoch": 3.3348115299334813, + "grad_norm": 1.0699121952056885, + "learning_rate": 1.6684112771899858e-06, + "loss": 0.3910093903541565, + "step": 1504 + }, + { + "epoch": 3.3392461197339247, + "grad_norm": 0.19254817068576813, + "learning_rate": 1.6597812495965537e-06, + "loss": 0.17204663157463074, + "step": 1506 + }, + { + "epoch": 3.343680709534368, + "grad_norm": 1.2403380870819092, + "learning_rate": 1.651202887485394e-06, + "loss": 0.21663136780261993, + "step": 1508 + }, + { + "epoch": 3.3481152993348116, + "grad_norm": 2.0894668102264404, + "learning_rate": 1.6426763062679553e-06, + "loss": 0.47299203276634216, + "step": 1510 + }, + { + "epoch": 3.352549889135255, + "grad_norm": 0.15021318197250366, + "learning_rate": 1.63420162065904e-06, + "loss": 0.20796214044094086, + "step": 1512 + }, + { + "epoch": 3.3569844789356984, + "grad_norm": 1.074012041091919, + "learning_rate": 1.625778944675257e-06, + "loss": 0.4941790997982025, + "step": 1514 + }, + { + "epoch": 3.361419068736142, + "grad_norm": 6.77968168258667, + "learning_rate": 1.6174083916334877e-06, + "loss": 0.3023959994316101, + "step": 1516 + }, + { + "epoch": 3.3658536585365852, + "grad_norm": 1.5364946126937866, + "learning_rate": 1.609090074149366e-06, + "loss": 0.30802658200263977, + "step": 1518 + }, + { + "epoch": 3.3702882483370287, + "grad_norm": 2.5270133018493652, + "learning_rate": 1.6008241041357535e-06, + "loss": 0.5083972811698914, + "step": 1520 + }, + { + "epoch": 3.374722838137472, + "grad_norm": 0.16544599831104279, + "learning_rate": 1.5926105928012486e-06, + "loss": 0.2425152212381363, + "step": 1522 + }, + { + "epoch": 3.3791574279379155, + "grad_norm": 1.5787891149520874, + "learning_rate": 1.5844496506486734e-06, + "loss": 0.5767493844032288, + "step": 1524 + }, + { + "epoch": 3.3835920177383594, + "grad_norm": 24.345401763916016, + "learning_rate": 1.576341387473601e-06, + "loss": 0.3439426124095917, + "step": 1526 + }, + { + "epoch": 3.388026607538803, + "grad_norm": 1.3219624757766724, + "learning_rate": 1.568285912362872e-06, + "loss": 0.21882264316082, + "step": 1528 + }, + { + "epoch": 3.3924611973392462, + "grad_norm": 0.47698870301246643, + "learning_rate": 1.5602833336931242e-06, + "loss": 0.209283709526062, + "step": 1530 + }, + { + "epoch": 3.3968957871396896, + "grad_norm": 0.9787905812263489, + "learning_rate": 1.552333759129344e-06, + "loss": 0.09470443427562714, + "step": 1532 + }, + { + "epoch": 3.401330376940133, + "grad_norm": 1.2374992370605469, + "learning_rate": 1.5444372956234062e-06, + "loss": 0.3461211919784546, + "step": 1534 + }, + { + "epoch": 3.4057649667405765, + "grad_norm": 1.2397676706314087, + "learning_rate": 1.5365940494126424e-06, + "loss": 0.46922361850738525, + "step": 1536 + }, + { + "epoch": 3.41019955654102, + "grad_norm": 1.3618249893188477, + "learning_rate": 1.5288041260184132e-06, + "loss": 0.3622947037220001, + "step": 1538 + }, + { + "epoch": 3.4146341463414633, + "grad_norm": 1.0507458448410034, + "learning_rate": 1.5210676302446801e-06, + "loss": 0.39398759603500366, + "step": 1540 + }, + { + "epoch": 3.4190687361419068, + "grad_norm": 1.3180983066558838, + "learning_rate": 1.5133846661766058e-06, + "loss": 0.3517080545425415, + "step": 1542 + }, + { + "epoch": 3.42350332594235, + "grad_norm": 1.1892797946929932, + "learning_rate": 1.5057553371791461e-06, + "loss": 0.3794390559196472, + "step": 1544 + }, + { + "epoch": 3.4279379157427936, + "grad_norm": 34.94332504272461, + "learning_rate": 1.4981797458956624e-06, + "loss": 0.10474438220262527, + "step": 1546 + }, + { + "epoch": 3.4323725055432375, + "grad_norm": 0.027800027281045914, + "learning_rate": 1.490657994246542e-06, + "loss": 0.0768439918756485, + "step": 1548 + }, + { + "epoch": 3.436807095343681, + "grad_norm": 1.083954095840454, + "learning_rate": 1.4831901834278212e-06, + "loss": 0.42769551277160645, + "step": 1550 + }, + { + "epoch": 3.4412416851441243, + "grad_norm": 1.0862568616867065, + "learning_rate": 1.4757764139098332e-06, + "loss": 0.3436740040779114, + "step": 1552 + }, + { + "epoch": 3.4456762749445677, + "grad_norm": 0.10507642477750778, + "learning_rate": 1.468416785435847e-06, + "loss": 0.29862260818481445, + "step": 1554 + }, + { + "epoch": 3.450110864745011, + "grad_norm": 1.1534576416015625, + "learning_rate": 1.461111397020732e-06, + "loss": 0.4363846778869629, + "step": 1556 + }, + { + "epoch": 3.4545454545454546, + "grad_norm": 1.3515652418136597, + "learning_rate": 1.4538603469496215e-06, + "loss": 0.4772418439388275, + "step": 1558 + }, + { + "epoch": 3.458980044345898, + "grad_norm": 0.9060417413711548, + "learning_rate": 1.4466637327765937e-06, + "loss": 0.5036817789077759, + "step": 1560 + }, + { + "epoch": 3.4634146341463414, + "grad_norm": 1.3972153663635254, + "learning_rate": 1.4395216513233584e-06, + "loss": 0.0940362960100174, + "step": 1562 + }, + { + "epoch": 3.467849223946785, + "grad_norm": 0.9557284712791443, + "learning_rate": 1.4324341986779527e-06, + "loss": 0.6766175627708435, + "step": 1564 + }, + { + "epoch": 3.4722838137472283, + "grad_norm": 0.8374095559120178, + "learning_rate": 1.4254014701934481e-06, + "loss": 0.21213091909885406, + "step": 1566 + }, + { + "epoch": 3.4767184035476717, + "grad_norm": 1.0304781198501587, + "learning_rate": 1.4184235604866725e-06, + "loss": 0.3508773744106293, + "step": 1568 + }, + { + "epoch": 3.481152993348115, + "grad_norm": 2.0591909885406494, + "learning_rate": 1.4115005634369296e-06, + "loss": 0.34335634112358093, + "step": 1570 + }, + { + "epoch": 3.4855875831485585, + "grad_norm": 0.2248295098543167, + "learning_rate": 1.4046325721847443e-06, + "loss": 0.3259221315383911, + "step": 1572 + }, + { + "epoch": 3.4900221729490024, + "grad_norm": 1.0798416137695312, + "learning_rate": 1.397819679130601e-06, + "loss": 0.5986089110374451, + "step": 1574 + }, + { + "epoch": 3.494456762749446, + "grad_norm": 3.3881709575653076, + "learning_rate": 1.3910619759337074e-06, + "loss": 0.26687368750572205, + "step": 1576 + }, + { + "epoch": 3.4988913525498893, + "grad_norm": 0.6919270157814026, + "learning_rate": 1.3843595535107587e-06, + "loss": 0.23883309960365295, + "step": 1578 + }, + { + "epoch": 3.5033259423503327, + "grad_norm": 1.5870050191879272, + "learning_rate": 1.377712502034712e-06, + "loss": 0.5271674394607544, + "step": 1580 + }, + { + "epoch": 3.507760532150776, + "grad_norm": 1.3132179975509644, + "learning_rate": 1.3711209109335793e-06, + "loss": 0.48455068469047546, + "step": 1582 + }, + { + "epoch": 3.5121951219512195, + "grad_norm": 0.9779510498046875, + "learning_rate": 1.3645848688892162e-06, + "loss": 0.2770904302597046, + "step": 1584 + }, + { + "epoch": 3.516629711751663, + "grad_norm": 1.488776445388794, + "learning_rate": 1.3581044638361373e-06, + "loss": 0.2443387657403946, + "step": 1586 + }, + { + "epoch": 3.5210643015521064, + "grad_norm": 1.3030271530151367, + "learning_rate": 1.3516797829603256e-06, + "loss": 0.44791534543037415, + "step": 1588 + }, + { + "epoch": 3.52549889135255, + "grad_norm": 2.1599605083465576, + "learning_rate": 1.3453109126980643e-06, + "loss": 0.19074156880378723, + "step": 1590 + }, + { + "epoch": 3.529933481152993, + "grad_norm": 1.5765827894210815, + "learning_rate": 1.3389979387347743e-06, + "loss": 0.23592326045036316, + "step": 1592 + }, + { + "epoch": 3.5343680709534366, + "grad_norm": 1.3213473558425903, + "learning_rate": 1.332740946003857e-06, + "loss": 0.4785956144332886, + "step": 1594 + }, + { + "epoch": 3.5388026607538805, + "grad_norm": 1.8811619281768799, + "learning_rate": 1.3265400186855548e-06, + "loss": 0.07135710120201111, + "step": 1596 + }, + { + "epoch": 3.5432372505543235, + "grad_norm": 3.579979181289673, + "learning_rate": 1.320395240205819e-06, + "loss": 0.4762045741081238, + "step": 1598 + }, + { + "epoch": 3.5476718403547673, + "grad_norm": 1.0917787551879883, + "learning_rate": 1.3143066932351856e-06, + "loss": 0.2512458860874176, + "step": 1600 + }, + { + "epoch": 3.5521064301552108, + "grad_norm": 1.354770302772522, + "learning_rate": 1.308274459687665e-06, + "loss": 0.2989339232444763, + "step": 1602 + }, + { + "epoch": 3.556541019955654, + "grad_norm": 0.9256249070167542, + "learning_rate": 1.3022986207196367e-06, + "loss": 0.5872430205345154, + "step": 1604 + }, + { + "epoch": 3.5609756097560976, + "grad_norm": 1.0018621683120728, + "learning_rate": 1.2963792567287617e-06, + "loss": 0.5670958757400513, + "step": 1606 + }, + { + "epoch": 3.565410199556541, + "grad_norm": 0.1671750247478485, + "learning_rate": 1.290516447352899e-06, + "loss": 0.034741759300231934, + "step": 1608 + }, + { + "epoch": 3.5698447893569845, + "grad_norm": 1.0988103151321411, + "learning_rate": 1.2847102714690308e-06, + "loss": 0.3017559051513672, + "step": 1610 + }, + { + "epoch": 3.574279379157428, + "grad_norm": 2.024768352508545, + "learning_rate": 1.2789608071922076e-06, + "loss": 0.09087596833705902, + "step": 1612 + }, + { + "epoch": 3.5787139689578713, + "grad_norm": 2.462632656097412, + "learning_rate": 1.2732681318744923e-06, + "loss": 0.3814306855201721, + "step": 1614 + }, + { + "epoch": 3.5831485587583147, + "grad_norm": 1.586369276046753, + "learning_rate": 1.2676323221039236e-06, + "loss": 0.7159979939460754, + "step": 1616 + }, + { + "epoch": 3.587583148558758, + "grad_norm": 3.622847557067871, + "learning_rate": 1.2620534537034795e-06, + "loss": 0.24280951917171478, + "step": 1618 + }, + { + "epoch": 3.5920177383592016, + "grad_norm": 2.6759791374206543, + "learning_rate": 1.2565316017300635e-06, + "loss": 0.40450718998908997, + "step": 1620 + }, + { + "epoch": 3.5964523281596454, + "grad_norm": 0.979674220085144, + "learning_rate": 1.2510668404734924e-06, + "loss": 0.3859134614467621, + "step": 1622 + }, + { + "epoch": 3.6008869179600884, + "grad_norm": 1.1429744958877563, + "learning_rate": 1.2456592434554963e-06, + "loss": 0.4199633002281189, + "step": 1624 + }, + { + "epoch": 3.6053215077605323, + "grad_norm": 10.887653350830078, + "learning_rate": 1.2403088834287282e-06, + "loss": 0.11880503594875336, + "step": 1626 + }, + { + "epoch": 3.6097560975609757, + "grad_norm": 1.1529648303985596, + "learning_rate": 1.2350158323757903e-06, + "loss": 0.3755669593811035, + "step": 1628 + }, + { + "epoch": 3.614190687361419, + "grad_norm": 2.1950061321258545, + "learning_rate": 1.229780161508259e-06, + "loss": 0.30650004744529724, + "step": 1630 + }, + { + "epoch": 3.6186252771618626, + "grad_norm": 1.1274499893188477, + "learning_rate": 1.2246019412657319e-06, + "loss": 0.49355947971343994, + "step": 1632 + }, + { + "epoch": 3.623059866962306, + "grad_norm": 1.7049529552459717, + "learning_rate": 1.2194812413148756e-06, + "loss": 0.49852749705314636, + "step": 1634 + }, + { + "epoch": 3.6274944567627494, + "grad_norm": 1.2080436944961548, + "learning_rate": 1.214418130548495e-06, + "loss": 0.3345094621181488, + "step": 1636 + }, + { + "epoch": 3.631929046563193, + "grad_norm": 2.6900432109832764, + "learning_rate": 1.2094126770845986e-06, + "loss": 0.38614609837532043, + "step": 1638 + }, + { + "epoch": 3.6363636363636362, + "grad_norm": 0.9296088814735413, + "learning_rate": 1.2044649482654876e-06, + "loss": 0.5239455103874207, + "step": 1640 + }, + { + "epoch": 3.6407982261640797, + "grad_norm": 0.42654258012771606, + "learning_rate": 1.1995750106568496e-06, + "loss": 0.24880681931972504, + "step": 1642 + }, + { + "epoch": 3.6452328159645235, + "grad_norm": 0.9114068746566772, + "learning_rate": 1.1947429300468575e-06, + "loss": 0.42800384759902954, + "step": 1644 + }, + { + "epoch": 3.6496674057649665, + "grad_norm": 0.22093704342842102, + "learning_rate": 1.1899687714452932e-06, + "loss": 0.24181388318538666, + "step": 1646 + }, + { + "epoch": 3.6541019955654104, + "grad_norm": 2.1429035663604736, + "learning_rate": 1.1852525990826658e-06, + "loss": 0.16454415023326874, + "step": 1648 + }, + { + "epoch": 3.658536585365854, + "grad_norm": 1.6719216108322144, + "learning_rate": 1.1805944764093484e-06, + "loss": 0.34569817781448364, + "step": 1650 + }, + { + "epoch": 3.662971175166297, + "grad_norm": 2.202148199081421, + "learning_rate": 1.1759944660947301e-06, + "loss": 0.44416671991348267, + "step": 1652 + }, + { + "epoch": 3.6674057649667406, + "grad_norm": 1.0591462850570679, + "learning_rate": 1.171452630026365e-06, + "loss": 0.1786390095949173, + "step": 1654 + }, + { + "epoch": 3.671840354767184, + "grad_norm": 1.1581944227218628, + "learning_rate": 1.1669690293091452e-06, + "loss": 0.4971603453159332, + "step": 1656 + }, + { + "epoch": 3.6762749445676275, + "grad_norm": 2.9780452251434326, + "learning_rate": 1.1625437242644772e-06, + "loss": 0.28031206130981445, + "step": 1658 + }, + { + "epoch": 3.680709534368071, + "grad_norm": 0.9189567565917969, + "learning_rate": 1.1581767744294682e-06, + "loss": 0.325800359249115, + "step": 1660 + }, + { + "epoch": 3.6851441241685143, + "grad_norm": 0.1943116933107376, + "learning_rate": 1.1538682385561286e-06, + "loss": 0.2070523202419281, + "step": 1662 + }, + { + "epoch": 3.6895787139689578, + "grad_norm": 1.0158610343933105, + "learning_rate": 1.1496181746105784e-06, + "loss": 0.23334433138370514, + "step": 1664 + }, + { + "epoch": 3.694013303769401, + "grad_norm": 0.8135693073272705, + "learning_rate": 1.1454266397722707e-06, + "loss": 0.40261054039001465, + "step": 1666 + }, + { + "epoch": 3.6984478935698446, + "grad_norm": 1.025733470916748, + "learning_rate": 1.1412936904332181e-06, + "loss": 0.3658636212348938, + "step": 1668 + }, + { + "epoch": 3.7028824833702885, + "grad_norm": 1.5712085962295532, + "learning_rate": 1.1372193821972379e-06, + "loss": 0.4164735674858093, + "step": 1670 + }, + { + "epoch": 3.7073170731707314, + "grad_norm": 1.251022458076477, + "learning_rate": 1.1332037698792033e-06, + "loss": 0.32476893067359924, + "step": 1672 + }, + { + "epoch": 3.7117516629711753, + "grad_norm": 1.0968090295791626, + "learning_rate": 1.1292469075043026e-06, + "loss": 0.5027676224708557, + "step": 1674 + }, + { + "epoch": 3.7161862527716187, + "grad_norm": 1.1787587404251099, + "learning_rate": 1.1253488483073177e-06, + "loss": 0.4444116950035095, + "step": 1676 + }, + { + "epoch": 3.720620842572062, + "grad_norm": 1.1607528924942017, + "learning_rate": 1.1215096447319038e-06, + "loss": 0.41037842631340027, + "step": 1678 + }, + { + "epoch": 3.7250554323725056, + "grad_norm": 2.8341286182403564, + "learning_rate": 1.117729348429884e-06, + "loss": 0.19149872660636902, + "step": 1680 + }, + { + "epoch": 3.729490022172949, + "grad_norm": 1.3583526611328125, + "learning_rate": 1.114008010260558e-06, + "loss": 0.4621056318283081, + "step": 1682 + }, + { + "epoch": 3.7339246119733924, + "grad_norm": 0.825949490070343, + "learning_rate": 1.1103456802900134e-06, + "loss": 0.18589909374713898, + "step": 1684 + }, + { + "epoch": 3.738359201773836, + "grad_norm": 2.031763792037964, + "learning_rate": 1.1067424077904555e-06, + "loss": 0.3091331422328949, + "step": 1686 + }, + { + "epoch": 3.7427937915742793, + "grad_norm": 0.24129338562488556, + "learning_rate": 1.103198241239542e-06, + "loss": 0.07289690524339676, + "step": 1688 + }, + { + "epoch": 3.7472283813747227, + "grad_norm": 1.1850625276565552, + "learning_rate": 1.0997132283197324e-06, + "loss": 0.5156506896018982, + "step": 1690 + }, + { + "epoch": 3.7516629711751666, + "grad_norm": 2.6560683250427246, + "learning_rate": 1.0962874159176454e-06, + "loss": 0.5042511820793152, + "step": 1692 + }, + { + "epoch": 3.7560975609756095, + "grad_norm": 1.1266878843307495, + "learning_rate": 1.0929208501234286e-06, + "loss": 0.5055519938468933, + "step": 1694 + }, + { + "epoch": 3.7605321507760534, + "grad_norm": 1.4622995853424072, + "learning_rate": 1.0896135762301393e-06, + "loss": 0.46531805396080017, + "step": 1696 + }, + { + "epoch": 3.764966740576497, + "grad_norm": 1.787441372871399, + "learning_rate": 1.0863656387331328e-06, + "loss": 0.21703627705574036, + "step": 1698 + }, + { + "epoch": 3.7694013303769403, + "grad_norm": 1.8496445417404175, + "learning_rate": 1.0831770813294668e-06, + "loss": 0.3597804009914398, + "step": 1700 + }, + { + "epoch": 3.7738359201773837, + "grad_norm": 0.731442928314209, + "learning_rate": 1.0800479469173101e-06, + "loss": 0.6957812309265137, + "step": 1702 + }, + { + "epoch": 3.778270509977827, + "grad_norm": 0.28417858481407166, + "learning_rate": 1.076978277595369e-06, + "loss": 0.05468475818634033, + "step": 1704 + }, + { + "epoch": 3.7827050997782705, + "grad_norm": 0.22050651907920837, + "learning_rate": 1.0739681146623185e-06, + "loss": 0.26601287722587585, + "step": 1706 + }, + { + "epoch": 3.787139689578714, + "grad_norm": 1.0851820707321167, + "learning_rate": 1.0710174986162471e-06, + "loss": 0.2386590987443924, + "step": 1708 + }, + { + "epoch": 3.7915742793791574, + "grad_norm": 0.8813753724098206, + "learning_rate": 1.0681264691541127e-06, + "loss": 0.6136298775672913, + "step": 1710 + }, + { + "epoch": 3.796008869179601, + "grad_norm": 2.6262285709381104, + "learning_rate": 1.0652950651712072e-06, + "loss": 0.2965908348560333, + "step": 1712 + }, + { + "epoch": 3.800443458980044, + "grad_norm": 2.0973598957061768, + "learning_rate": 1.0625233247606348e-06, + "loss": 0.2669585049152374, + "step": 1714 + }, + { + "epoch": 3.8048780487804876, + "grad_norm": 1.7458832263946533, + "learning_rate": 1.059811285212799e-06, + "loss": 0.5059341192245483, + "step": 1716 + }, + { + "epoch": 3.8093126385809315, + "grad_norm": 1.2826629877090454, + "learning_rate": 1.0571589830149e-06, + "loss": 0.27312329411506653, + "step": 1718 + }, + { + "epoch": 3.8137472283813745, + "grad_norm": 3.2964842319488525, + "learning_rate": 1.054566453850444e-06, + "loss": 0.3187982738018036, + "step": 1720 + }, + { + "epoch": 3.8181818181818183, + "grad_norm": 2.765103340148926, + "learning_rate": 1.0520337325987649e-06, + "loss": 0.5375199913978577, + "step": 1722 + }, + { + "epoch": 3.8226164079822618, + "grad_norm": 5.0600361824035645, + "learning_rate": 1.049560853334553e-06, + "loss": 0.4722135066986084, + "step": 1724 + }, + { + "epoch": 3.827050997782705, + "grad_norm": 1.3732205629348755, + "learning_rate": 1.0471478493273976e-06, + "loss": 0.4118424654006958, + "step": 1726 + }, + { + "epoch": 3.8314855875831486, + "grad_norm": 1.0661771297454834, + "learning_rate": 1.0447947530413389e-06, + "loss": 0.2754386365413666, + "step": 1728 + }, + { + "epoch": 3.835920177383592, + "grad_norm": 1.2804239988327026, + "learning_rate": 1.042501596134431e-06, + "loss": 0.2026994377374649, + "step": 1730 + }, + { + "epoch": 3.8403547671840355, + "grad_norm": 1.4092806577682495, + "learning_rate": 1.0402684094583173e-06, + "loss": 0.4652438163757324, + "step": 1732 + }, + { + "epoch": 3.844789356984479, + "grad_norm": 4.014774322509766, + "learning_rate": 1.0380952230578125e-06, + "loss": 0.379792720079422, + "step": 1734 + }, + { + "epoch": 3.8492239467849223, + "grad_norm": 1.469107985496521, + "learning_rate": 1.0359820661705042e-06, + "loss": 0.3514306843280792, + "step": 1736 + }, + { + "epoch": 3.8536585365853657, + "grad_norm": 1.1409002542495728, + "learning_rate": 1.0339289672263519e-06, + "loss": 0.44202250242233276, + "step": 1738 + }, + { + "epoch": 3.858093126385809, + "grad_norm": 0.8285048604011536, + "learning_rate": 1.0319359538473107e-06, + "loss": 0.23279811441898346, + "step": 1740 + }, + { + "epoch": 3.8625277161862526, + "grad_norm": 1.168976902961731, + "learning_rate": 1.0300030528469564e-06, + "loss": 0.20990443229675293, + "step": 1742 + }, + { + "epoch": 3.8669623059866964, + "grad_norm": 1.5003247261047363, + "learning_rate": 1.0281302902301254e-06, + "loss": 0.4064357280731201, + "step": 1744 + }, + { + "epoch": 3.8713968957871394, + "grad_norm": 0.8279268741607666, + "learning_rate": 1.026317691192567e-06, + "loss": 0.4411630630493164, + "step": 1746 + }, + { + "epoch": 3.8758314855875833, + "grad_norm": 1.1625535488128662, + "learning_rate": 1.0245652801205999e-06, + "loss": 0.2272336483001709, + "step": 1748 + }, + { + "epoch": 3.8802660753880267, + "grad_norm": 12.195842742919922, + "learning_rate": 1.0228730805907891e-06, + "loss": 0.3394715189933777, + "step": 1750 + }, + { + "epoch": 3.88470066518847, + "grad_norm": 3.731518507003784, + "learning_rate": 1.0212411153696247e-06, + "loss": 0.34466421604156494, + "step": 1752 + }, + { + "epoch": 3.8891352549889135, + "grad_norm": 1.2109719514846802, + "learning_rate": 1.019669406413218e-06, + "loss": 0.3292272388935089, + "step": 1754 + }, + { + "epoch": 3.893569844789357, + "grad_norm": 3.1141412258148193, + "learning_rate": 1.0181579748670054e-06, + "loss": 0.3200131058692932, + "step": 1756 + }, + { + "epoch": 3.8980044345898004, + "grad_norm": 0.9708091020584106, + "learning_rate": 1.0167068410654643e-06, + "loss": 0.4250810444355011, + "step": 1758 + }, + { + "epoch": 3.902439024390244, + "grad_norm": 0.030687185004353523, + "learning_rate": 1.0153160245318384e-06, + "loss": 0.002097110729664564, + "step": 1760 + }, + { + "epoch": 3.9068736141906872, + "grad_norm": 1.3904731273651123, + "learning_rate": 1.0139855439778766e-06, + "loss": 0.08549664914608002, + "step": 1762 + }, + { + "epoch": 3.9113082039911307, + "grad_norm": 0.10882235318422318, + "learning_rate": 1.0127154173035787e-06, + "loss": 0.28717437386512756, + "step": 1764 + }, + { + "epoch": 3.9157427937915745, + "grad_norm": 1.5540599822998047, + "learning_rate": 1.0115056615969584e-06, + "loss": 0.2964329421520233, + "step": 1766 + }, + { + "epoch": 3.9201773835920175, + "grad_norm": 1.1777206659317017, + "learning_rate": 1.0103562931338105e-06, + "loss": 0.619647741317749, + "step": 1768 + }, + { + "epoch": 3.9246119733924614, + "grad_norm": 2.035280704498291, + "learning_rate": 1.009267327377492e-06, + "loss": 0.48704907298088074, + "step": 1770 + }, + { + "epoch": 3.929046563192905, + "grad_norm": 0.8825568556785583, + "learning_rate": 1.008238778978716e-06, + "loss": 0.18097802996635437, + "step": 1772 + }, + { + "epoch": 3.933481152993348, + "grad_norm": 0.8677657246589661, + "learning_rate": 1.0072706617753528e-06, + "loss": 0.39700010418891907, + "step": 1774 + }, + { + "epoch": 3.9379157427937916, + "grad_norm": 1.6046772003173828, + "learning_rate": 1.0063629887922441e-06, + "loss": 0.5214118361473083, + "step": 1776 + }, + { + "epoch": 3.942350332594235, + "grad_norm": 0.15087999403476715, + "learning_rate": 1.0055157722410279e-06, + "loss": 0.03213101252913475, + "step": 1778 + }, + { + "epoch": 3.9467849223946785, + "grad_norm": 3.6360549926757812, + "learning_rate": 1.0047290235199753e-06, + "loss": 0.32588061690330505, + "step": 1780 + }, + { + "epoch": 3.951219512195122, + "grad_norm": 0.1825929582118988, + "learning_rate": 1.0040027532138351e-06, + "loss": 0.22568024694919586, + "step": 1782 + }, + { + "epoch": 3.9556541019955653, + "grad_norm": 0.905255138874054, + "learning_rate": 1.0033369710936928e-06, + "loss": 0.3658754527568817, + "step": 1784 + }, + { + "epoch": 3.9600886917960088, + "grad_norm": 0.8106067180633545, + "learning_rate": 1.0027316861168388e-06, + "loss": 0.357939213514328, + "step": 1786 + }, + { + "epoch": 3.964523281596452, + "grad_norm": 0.7470359206199646, + "learning_rate": 1.0021869064266472e-06, + "loss": 0.30442333221435547, + "step": 1788 + }, + { + "epoch": 3.9689578713968956, + "grad_norm": 0.6582420468330383, + "learning_rate": 1.0017026393524684e-06, + "loss": 0.23345550894737244, + "step": 1790 + }, + { + "epoch": 3.9733924611973395, + "grad_norm": 1.1834862232208252, + "learning_rate": 1.0012788914095275e-06, + "loss": 0.5164616703987122, + "step": 1792 + }, + { + "epoch": 3.9778270509977824, + "grad_norm": 0.7614670395851135, + "learning_rate": 1.0009156682988395e-06, + "loss": 0.36194929480552673, + "step": 1794 + }, + { + "epoch": 3.9822616407982263, + "grad_norm": 1.1429847478866577, + "learning_rate": 1.0006129749071298e-06, + "loss": 0.3539275825023651, + "step": 1796 + }, + { + "epoch": 3.9866962305986697, + "grad_norm": 1.9152874946594238, + "learning_rate": 1.00037081530677e-06, + "loss": 0.4812496304512024, + "step": 1798 + }, + { + "epoch": 3.991130820399113, + "grad_norm": 4.806298732757568, + "learning_rate": 1.0001891927557255e-06, + "loss": 0.4314287304878235, + "step": 1800 + }, + { + "epoch": 3.9955654101995566, + "grad_norm": 1.1714129447937012, + "learning_rate": 1.0000681096975056e-06, + "loss": 0.37641072273254395, + "step": 1802 + }, + { + "epoch": 4.0, + "grad_norm": 0.90379798412323, + "learning_rate": 1.0000075677611364e-06, + "loss": 0.12868885695934296, + "step": 1804 + }, + { + "epoch": 4.0, + "step": 1804, + "total_flos": 3.4175049861232067e+18, + "train_loss": 0.7562091423920304, + "train_runtime": 9085.3052, + "train_samples_per_second": 5.957, + "train_steps_per_second": 0.199 + } + ], + "logging_steps": 2, + "max_steps": 1804, + "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": 3.4175049861232067e+18, + "train_batch_size": 3, + "trial_name": null, + "trial_params": null +}