{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 1042, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003838771593090211, "grad_norm": 11.820537567138672, "learning_rate": 3.7735849056603773e-06, "loss": 6.5959, "step": 2 }, { "epoch": 0.007677543186180422, "grad_norm": 9.131757736206055, "learning_rate": 7.547169811320755e-06, "loss": 6.3324, "step": 4 }, { "epoch": 0.011516314779270634, "grad_norm": 9.551312446594238, "learning_rate": 1.1320754716981132e-05, "loss": 6.0616, "step": 6 }, { "epoch": 0.015355086372360844, "grad_norm": 6.00274133682251, "learning_rate": 1.509433962264151e-05, "loss": 5.1431, "step": 8 }, { "epoch": 0.019193857965451054, "grad_norm": 15.785564422607422, "learning_rate": 1.8867924528301888e-05, "loss": 5.0109, "step": 10 }, { "epoch": 0.023032629558541268, "grad_norm": 4.418276786804199, "learning_rate": 2.2641509433962265e-05, "loss": 4.9469, "step": 12 }, { "epoch": 0.026871401151631478, "grad_norm": 5.366930961608887, "learning_rate": 2.641509433962264e-05, "loss": 4.5625, "step": 14 }, { "epoch": 0.030710172744721688, "grad_norm": 5.949010372161865, "learning_rate": 3.018867924528302e-05, "loss": 4.5685, "step": 16 }, { "epoch": 0.0345489443378119, "grad_norm": 3.487248659133911, "learning_rate": 3.39622641509434e-05, "loss": 4.3738, "step": 18 }, { "epoch": 0.03838771593090211, "grad_norm": 4.582718372344971, "learning_rate": 3.7735849056603776e-05, "loss": 4.4494, "step": 20 }, { "epoch": 0.04222648752399232, "grad_norm": 3.4489753246307373, "learning_rate": 4.150943396226415e-05, "loss": 4.2253, "step": 22 }, { "epoch": 0.046065259117082535, "grad_norm": 2.974649429321289, "learning_rate": 4.528301886792453e-05, "loss": 4.1242, "step": 24 }, { "epoch": 0.04990403071017274, "grad_norm": 3.6813883781433105, "learning_rate": 4.9056603773584906e-05, "loss": 4.0906, "step": 26 }, { "epoch": 0.053742802303262956, "grad_norm": 3.122687578201294, "learning_rate": 5.283018867924528e-05, "loss": 4.1557, "step": 28 }, { "epoch": 0.05758157389635317, "grad_norm": 2.735628128051758, "learning_rate": 5.660377358490566e-05, "loss": 3.7832, "step": 30 }, { "epoch": 0.061420345489443376, "grad_norm": 2.757296562194824, "learning_rate": 6.037735849056604e-05, "loss": 4.0969, "step": 32 }, { "epoch": 0.06525911708253358, "grad_norm": 3.5924015045166016, "learning_rate": 6.415094339622641e-05, "loss": 4.0404, "step": 34 }, { "epoch": 0.0690978886756238, "grad_norm": 3.1863701343536377, "learning_rate": 6.79245283018868e-05, "loss": 3.9686, "step": 36 }, { "epoch": 0.07293666026871401, "grad_norm": 3.469578981399536, "learning_rate": 7.169811320754717e-05, "loss": 4.0311, "step": 38 }, { "epoch": 0.07677543186180422, "grad_norm": 3.0389885902404785, "learning_rate": 7.547169811320755e-05, "loss": 3.8048, "step": 40 }, { "epoch": 0.08061420345489444, "grad_norm": 3.441258430480957, "learning_rate": 7.924528301886794e-05, "loss": 3.9566, "step": 42 }, { "epoch": 0.08445297504798464, "grad_norm": 3.2453179359436035, "learning_rate": 8.30188679245283e-05, "loss": 3.9577, "step": 44 }, { "epoch": 0.08829174664107485, "grad_norm": 2.931682586669922, "learning_rate": 8.679245283018869e-05, "loss": 3.871, "step": 46 }, { "epoch": 0.09213051823416507, "grad_norm": 2.4610602855682373, "learning_rate": 9.056603773584906e-05, "loss": 3.6341, "step": 48 }, { "epoch": 0.09596928982725528, "grad_norm": 3.078601598739624, "learning_rate": 9.433962264150944e-05, "loss": 3.9301, "step": 50 }, { "epoch": 0.09980806142034548, "grad_norm": 3.0668110847473145, "learning_rate": 9.811320754716981e-05, "loss": 3.9914, "step": 52 }, { "epoch": 0.1036468330134357, "grad_norm": 3.0697929859161377, "learning_rate": 9.999974774092106e-05, "loss": 3.6566, "step": 54 }, { "epoch": 0.10748560460652591, "grad_norm": 2.7237517833709717, "learning_rate": 9.999772968356182e-05, "loss": 3.6941, "step": 56 }, { "epoch": 0.11132437619961612, "grad_norm": 3.0039498805999756, "learning_rate": 9.999369365029487e-05, "loss": 3.7698, "step": 58 }, { "epoch": 0.11516314779270634, "grad_norm": 2.9920670986175537, "learning_rate": 9.998763980401998e-05, "loss": 3.7305, "step": 60 }, { "epoch": 0.11900191938579655, "grad_norm": 2.935713291168213, "learning_rate": 9.997956838907852e-05, "loss": 3.7833, "step": 62 }, { "epoch": 0.12284069097888675, "grad_norm": 3.031791925430298, "learning_rate": 9.996947973124372e-05, "loss": 3.8708, "step": 64 }, { "epoch": 0.12667946257197696, "grad_norm": 3.4206488132476807, "learning_rate": 9.995737423770745e-05, "loss": 3.5967, "step": 66 }, { "epoch": 0.13051823416506717, "grad_norm": 3.457770824432373, "learning_rate": 9.994325239706377e-05, "loss": 3.6511, "step": 68 }, { "epoch": 0.1343570057581574, "grad_norm": 3.2217350006103516, "learning_rate": 9.992711477928926e-05, "loss": 3.5577, "step": 70 }, { "epoch": 0.1381957773512476, "grad_norm": 2.7584404945373535, "learning_rate": 9.990896203571995e-05, "loss": 3.8195, "step": 72 }, { "epoch": 0.1420345489443378, "grad_norm": 2.4847779273986816, "learning_rate": 9.988879489902509e-05, "loss": 3.6329, "step": 74 }, { "epoch": 0.14587332053742802, "grad_norm": 3.222179412841797, "learning_rate": 9.986661418317759e-05, "loss": 3.6526, "step": 76 }, { "epoch": 0.14971209213051823, "grad_norm": 2.882899045944214, "learning_rate": 9.984242078342107e-05, "loss": 3.7584, "step": 78 }, { "epoch": 0.15355086372360843, "grad_norm": 2.8948464393615723, "learning_rate": 9.981621567623385e-05, "loss": 3.5789, "step": 80 }, { "epoch": 0.15738963531669867, "grad_norm": 3.1748478412628174, "learning_rate": 9.978799991928944e-05, "loss": 3.6618, "step": 82 }, { "epoch": 0.16122840690978887, "grad_norm": 2.8642570972442627, "learning_rate": 9.97577746514139e-05, "loss": 3.6284, "step": 84 }, { "epoch": 0.16506717850287908, "grad_norm": 2.8989341259002686, "learning_rate": 9.972554109253989e-05, "loss": 3.6497, "step": 86 }, { "epoch": 0.1689059500959693, "grad_norm": 2.4885776042938232, "learning_rate": 9.969130054365736e-05, "loss": 3.5491, "step": 88 }, { "epoch": 0.1727447216890595, "grad_norm": 2.85013484954834, "learning_rate": 9.965505438676115e-05, "loss": 3.6339, "step": 90 }, { "epoch": 0.1765834932821497, "grad_norm": 2.275810718536377, "learning_rate": 9.961680408479507e-05, "loss": 3.4046, "step": 92 }, { "epoch": 0.18042226487523993, "grad_norm": 2.4338936805725098, "learning_rate": 9.957655118159304e-05, "loss": 3.5078, "step": 94 }, { "epoch": 0.18426103646833014, "grad_norm": 2.5379140377044678, "learning_rate": 9.953429730181653e-05, "loss": 3.457, "step": 96 }, { "epoch": 0.18809980806142035, "grad_norm": 2.46872878074646, "learning_rate": 9.949004415088927e-05, "loss": 3.465, "step": 98 }, { "epoch": 0.19193857965451055, "grad_norm": 2.2635891437530518, "learning_rate": 9.944379351492818e-05, "loss": 3.3926, "step": 100 }, { "epoch": 0.19577735124760076, "grad_norm": 2.4858951568603516, "learning_rate": 9.939554726067141e-05, "loss": 3.5257, "step": 102 }, { "epoch": 0.19961612284069097, "grad_norm": 2.854139566421509, "learning_rate": 9.934530733540294e-05, "loss": 3.5232, "step": 104 }, { "epoch": 0.2034548944337812, "grad_norm": 3.0548629760742188, "learning_rate": 9.929307576687403e-05, "loss": 3.528, "step": 106 }, { "epoch": 0.2072936660268714, "grad_norm": 3.436591863632202, "learning_rate": 9.923885466322135e-05, "loss": 3.5877, "step": 108 }, { "epoch": 0.21113243761996162, "grad_norm": 2.600471258163452, "learning_rate": 9.918264621288187e-05, "loss": 3.5579, "step": 110 }, { "epoch": 0.21497120921305182, "grad_norm": 2.838399887084961, "learning_rate": 9.912445268450458e-05, "loss": 3.5284, "step": 112 }, { "epoch": 0.21880998080614203, "grad_norm": 2.9568042755126953, "learning_rate": 9.906427642685889e-05, "loss": 3.4821, "step": 114 }, { "epoch": 0.22264875239923224, "grad_norm": 2.8753232955932617, "learning_rate": 9.900211986873985e-05, "loss": 3.4921, "step": 116 }, { "epoch": 0.22648752399232247, "grad_norm": 2.559448719024658, "learning_rate": 9.89379855188701e-05, "loss": 3.414, "step": 118 }, { "epoch": 0.23032629558541268, "grad_norm": 2.7654104232788086, "learning_rate": 9.887187596579866e-05, "loss": 3.4106, "step": 120 }, { "epoch": 0.23416506717850288, "grad_norm": 3.026230812072754, "learning_rate": 9.880379387779637e-05, "loss": 3.341, "step": 122 }, { "epoch": 0.2380038387715931, "grad_norm": 2.73810076713562, "learning_rate": 9.873374200274826e-05, "loss": 3.3992, "step": 124 }, { "epoch": 0.2418426103646833, "grad_norm": 2.587951898574829, "learning_rate": 9.866172316804266e-05, "loss": 3.4407, "step": 126 }, { "epoch": 0.2456813819577735, "grad_norm": 2.8860104084014893, "learning_rate": 9.8587740280457e-05, "loss": 3.2615, "step": 128 }, { "epoch": 0.2495201535508637, "grad_norm": 2.9032886028289795, "learning_rate": 9.851179632604057e-05, "loss": 3.386, "step": 130 }, { "epoch": 0.2533589251439539, "grad_norm": 2.958479881286621, "learning_rate": 9.843389436999396e-05, "loss": 3.3967, "step": 132 }, { "epoch": 0.2571976967370441, "grad_norm": 2.822604179382324, "learning_rate": 9.835403755654536e-05, "loss": 3.455, "step": 134 }, { "epoch": 0.26103646833013433, "grad_norm": 2.626711368560791, "learning_rate": 9.827222910882358e-05, "loss": 3.4964, "step": 136 }, { "epoch": 0.2648752399232246, "grad_norm": 3.292914390563965, "learning_rate": 9.818847232872815e-05, "loss": 3.3882, "step": 138 }, { "epoch": 0.2687140115163148, "grad_norm": 2.6831047534942627, "learning_rate": 9.81027705967958e-05, "loss": 3.3436, "step": 140 }, { "epoch": 0.272552783109405, "grad_norm": 2.936622381210327, "learning_rate": 9.801512737206422e-05, "loss": 3.4679, "step": 142 }, { "epoch": 0.2763915547024952, "grad_norm": 3.395770788192749, "learning_rate": 9.792554619193235e-05, "loss": 3.238, "step": 144 }, { "epoch": 0.2802303262955854, "grad_norm": 2.6605355739593506, "learning_rate": 9.783403067201762e-05, "loss": 3.1712, "step": 146 }, { "epoch": 0.2840690978886756, "grad_norm": 2.5874388217926025, "learning_rate": 9.774058450601002e-05, "loss": 3.4306, "step": 148 }, { "epoch": 0.28790786948176583, "grad_norm": 2.590623378753662, "learning_rate": 9.764521146552309e-05, "loss": 3.3745, "step": 150 }, { "epoch": 0.29174664107485604, "grad_norm": 2.959683895111084, "learning_rate": 9.754791539994152e-05, "loss": 3.3248, "step": 152 }, { "epoch": 0.29558541266794625, "grad_norm": 3.0726540088653564, "learning_rate": 9.744870023626599e-05, "loss": 3.2341, "step": 154 }, { "epoch": 0.29942418426103645, "grad_norm": 2.616612434387207, "learning_rate": 9.734756997895448e-05, "loss": 3.3041, "step": 156 }, { "epoch": 0.30326295585412666, "grad_norm": 2.782773494720459, "learning_rate": 9.724452870976085e-05, "loss": 3.3366, "step": 158 }, { "epoch": 0.30710172744721687, "grad_norm": 2.423705816268921, "learning_rate": 9.713958058756985e-05, "loss": 3.0726, "step": 160 }, { "epoch": 0.31094049904030713, "grad_norm": 3.2293317317962646, "learning_rate": 9.703272984822946e-05, "loss": 3.3771, "step": 162 }, { "epoch": 0.31477927063339733, "grad_norm": 2.3854119777679443, "learning_rate": 9.69239808043799e-05, "loss": 3.2408, "step": 164 }, { "epoch": 0.31861804222648754, "grad_norm": 2.4507429599761963, "learning_rate": 9.681333784527944e-05, "loss": 3.2886, "step": 166 }, { "epoch": 0.32245681381957775, "grad_norm": 3.1059365272521973, "learning_rate": 9.67008054366274e-05, "loss": 3.2542, "step": 168 }, { "epoch": 0.32629558541266795, "grad_norm": 2.4190452098846436, "learning_rate": 9.658638812038379e-05, "loss": 3.3978, "step": 170 }, { "epoch": 0.33013435700575816, "grad_norm": 2.705996036529541, "learning_rate": 9.647009051458604e-05, "loss": 3.2259, "step": 172 }, { "epoch": 0.33397312859884837, "grad_norm": 2.388340711593628, "learning_rate": 9.635191731316262e-05, "loss": 3.1787, "step": 174 }, { "epoch": 0.3378119001919386, "grad_norm": 2.475062608718872, "learning_rate": 9.623187328574357e-05, "loss": 3.284, "step": 176 }, { "epoch": 0.3416506717850288, "grad_norm": 2.7308733463287354, "learning_rate": 9.6109963277468e-05, "loss": 3.259, "step": 178 }, { "epoch": 0.345489443378119, "grad_norm": 2.640472650527954, "learning_rate": 9.598619220878853e-05, "loss": 3.2578, "step": 180 }, { "epoch": 0.3493282149712092, "grad_norm": 2.5652854442596436, "learning_rate": 9.586056507527266e-05, "loss": 3.1727, "step": 182 }, { "epoch": 0.3531669865642994, "grad_norm": 2.9988858699798584, "learning_rate": 9.573308694740121e-05, "loss": 3.2553, "step": 184 }, { "epoch": 0.3570057581573896, "grad_norm": 2.611365556716919, "learning_rate": 9.560376297036362e-05, "loss": 3.2819, "step": 186 }, { "epoch": 0.36084452975047987, "grad_norm": 2.2860095500946045, "learning_rate": 9.547259836385031e-05, "loss": 3.1743, "step": 188 }, { "epoch": 0.3646833013435701, "grad_norm": 2.598378896713257, "learning_rate": 9.533959842184195e-05, "loss": 3.1716, "step": 190 }, { "epoch": 0.3685220729366603, "grad_norm": 1.9929189682006836, "learning_rate": 9.520476851239588e-05, "loss": 3.2953, "step": 192 }, { "epoch": 0.3723608445297505, "grad_norm": 2.3581299781799316, "learning_rate": 9.506811407742937e-05, "loss": 3.294, "step": 194 }, { "epoch": 0.3761996161228407, "grad_norm": 2.4338443279266357, "learning_rate": 9.492964063250001e-05, "loss": 3.124, "step": 196 }, { "epoch": 0.3800383877159309, "grad_norm": 2.2041661739349365, "learning_rate": 9.478935376658307e-05, "loss": 3.2005, "step": 198 }, { "epoch": 0.3838771593090211, "grad_norm": 3.0383594036102295, "learning_rate": 9.4647259141846e-05, "loss": 3.224, "step": 200 }, { "epoch": 0.3877159309021113, "grad_norm": 2.622857093811035, "learning_rate": 9.450336249341976e-05, "loss": 3.3196, "step": 202 }, { "epoch": 0.3915547024952015, "grad_norm": 2.5729992389678955, "learning_rate": 9.435766962916748e-05, "loss": 3.2492, "step": 204 }, { "epoch": 0.39539347408829173, "grad_norm": 2.683148145675659, "learning_rate": 9.421018642944997e-05, "loss": 3.2415, "step": 206 }, { "epoch": 0.39923224568138194, "grad_norm": 2.756338119506836, "learning_rate": 9.406091884688836e-05, "loss": 3.1518, "step": 208 }, { "epoch": 0.40307101727447214, "grad_norm": 2.5508527755737305, "learning_rate": 9.390987290612395e-05, "loss": 3.2104, "step": 210 }, { "epoch": 0.4069097888675624, "grad_norm": 2.478585958480835, "learning_rate": 9.375705470357493e-05, "loss": 3.1351, "step": 212 }, { "epoch": 0.4107485604606526, "grad_norm": 2.756169319152832, "learning_rate": 9.360247040719039e-05, "loss": 3.1652, "step": 214 }, { "epoch": 0.4145873320537428, "grad_norm": 2.579590320587158, "learning_rate": 9.344612625620132e-05, "loss": 3.1559, "step": 216 }, { "epoch": 0.418426103646833, "grad_norm": 2.307854413986206, "learning_rate": 9.32880285608689e-05, "loss": 3.117, "step": 218 }, { "epoch": 0.42226487523992323, "grad_norm": 2.215238332748413, "learning_rate": 9.312818370222962e-05, "loss": 3.0537, "step": 220 }, { "epoch": 0.42610364683301344, "grad_norm": 3.0493693351745605, "learning_rate": 9.296659813183794e-05, "loss": 3.1795, "step": 222 }, { "epoch": 0.42994241842610365, "grad_norm": 2.6957015991210938, "learning_rate": 9.280327837150571e-05, "loss": 3.1738, "step": 224 }, { "epoch": 0.43378119001919385, "grad_norm": 2.6832878589630127, "learning_rate": 9.26382310130391e-05, "loss": 3.1244, "step": 226 }, { "epoch": 0.43761996161228406, "grad_norm": 2.5195510387420654, "learning_rate": 9.247146271797245e-05, "loss": 3.1738, "step": 228 }, { "epoch": 0.44145873320537427, "grad_norm": 2.3536553382873535, "learning_rate": 9.23029802172994e-05, "loss": 3.0517, "step": 230 }, { "epoch": 0.44529750479846447, "grad_norm": 2.653780698776245, "learning_rate": 9.213279031120129e-05, "loss": 2.8809, "step": 232 }, { "epoch": 0.4491362763915547, "grad_norm": 2.6917688846588135, "learning_rate": 9.19608998687726e-05, "loss": 3.2078, "step": 234 }, { "epoch": 0.45297504798464494, "grad_norm": 2.325704574584961, "learning_rate": 9.178731582774381e-05, "loss": 3.1038, "step": 236 }, { "epoch": 0.45681381957773515, "grad_norm": 2.4601848125457764, "learning_rate": 9.161204519420126e-05, "loss": 3.082, "step": 238 }, { "epoch": 0.46065259117082535, "grad_norm": 2.084491491317749, "learning_rate": 9.143509504230449e-05, "loss": 3.1871, "step": 240 }, { "epoch": 0.46449136276391556, "grad_norm": 2.4091813564300537, "learning_rate": 9.125647251400067e-05, "loss": 3.0972, "step": 242 }, { "epoch": 0.46833013435700577, "grad_norm": 2.4480786323547363, "learning_rate": 9.107618481873632e-05, "loss": 3.0268, "step": 244 }, { "epoch": 0.472168905950096, "grad_norm": 2.4696197509765625, "learning_rate": 9.089423923316636e-05, "loss": 3.1253, "step": 246 }, { "epoch": 0.4760076775431862, "grad_norm": 3.0479440689086914, "learning_rate": 9.07106431008604e-05, "loss": 3.0147, "step": 248 }, { "epoch": 0.4798464491362764, "grad_norm": 2.204630136489868, "learning_rate": 9.052540383200633e-05, "loss": 3.0069, "step": 250 }, { "epoch": 0.4836852207293666, "grad_norm": 2.3515751361846924, "learning_rate": 9.033852890311128e-05, "loss": 3.1641, "step": 252 }, { "epoch": 0.4875239923224568, "grad_norm": 2.63689923286438, "learning_rate": 9.01500258566998e-05, "loss": 3.0906, "step": 254 }, { "epoch": 0.491362763915547, "grad_norm": 2.479234457015991, "learning_rate": 8.995990230100949e-05, "loss": 3.0146, "step": 256 }, { "epoch": 0.4952015355086372, "grad_norm": 2.147664785385132, "learning_rate": 8.976816590968387e-05, "loss": 3.1342, "step": 258 }, { "epoch": 0.4990403071017274, "grad_norm": 2.3102312088012695, "learning_rate": 8.957482442146272e-05, "loss": 3.0833, "step": 260 }, { "epoch": 0.5028790786948176, "grad_norm": 2.3604090213775635, "learning_rate": 8.937988563986963e-05, "loss": 2.937, "step": 262 }, { "epoch": 0.5067178502879078, "grad_norm": 2.4375572204589844, "learning_rate": 8.918335743289718e-05, "loss": 3.1493, "step": 264 }, { "epoch": 0.510556621880998, "grad_norm": 2.279297113418579, "learning_rate": 8.898524773268926e-05, "loss": 3.1373, "step": 266 }, { "epoch": 0.5143953934740882, "grad_norm": 2.5170693397521973, "learning_rate": 8.8785564535221e-05, "loss": 3.0832, "step": 268 }, { "epoch": 0.5182341650671785, "grad_norm": 2.407562732696533, "learning_rate": 8.858431589997597e-05, "loss": 3.0379, "step": 270 }, { "epoch": 0.5220729366602687, "grad_norm": 2.3649191856384277, "learning_rate": 8.838150994962093e-05, "loss": 3.1197, "step": 272 }, { "epoch": 0.525911708253359, "grad_norm": 2.5898306369781494, "learning_rate": 8.817715486967803e-05, "loss": 3.1413, "step": 274 }, { "epoch": 0.5297504798464492, "grad_norm": 3.036756992340088, "learning_rate": 8.797125890819428e-05, "loss": 3.0454, "step": 276 }, { "epoch": 0.5335892514395394, "grad_norm": 1.9929702281951904, "learning_rate": 8.776383037540888e-05, "loss": 3.0799, "step": 278 }, { "epoch": 0.5374280230326296, "grad_norm": 1.9196736812591553, "learning_rate": 8.755487764341757e-05, "loss": 2.997, "step": 280 }, { "epoch": 0.5412667946257198, "grad_norm": 2.313257932662964, "learning_rate": 8.734440914583486e-05, "loss": 3.0764, "step": 282 }, { "epoch": 0.54510556621881, "grad_norm": 2.452481508255005, "learning_rate": 8.713243337745365e-05, "loss": 3.0441, "step": 284 }, { "epoch": 0.5489443378119002, "grad_norm": 2.9498703479766846, "learning_rate": 8.691895889390227e-05, "loss": 3.0539, "step": 286 }, { "epoch": 0.5527831094049904, "grad_norm": 1.9671512842178345, "learning_rate": 8.670399431129926e-05, "loss": 2.8646, "step": 288 }, { "epoch": 0.5566218809980806, "grad_norm": 2.5226361751556396, "learning_rate": 8.648754830590552e-05, "loss": 3.184, "step": 290 }, { "epoch": 0.5604606525911708, "grad_norm": 2.4315929412841797, "learning_rate": 8.626962961377423e-05, "loss": 3.09, "step": 292 }, { "epoch": 0.564299424184261, "grad_norm": 2.2799346446990967, "learning_rate": 8.605024703039816e-05, "loss": 3.0375, "step": 294 }, { "epoch": 0.5681381957773513, "grad_norm": 2.298910617828369, "learning_rate": 8.582940941035476e-05, "loss": 3.0763, "step": 296 }, { "epoch": 0.5719769673704415, "grad_norm": 2.156235694885254, "learning_rate": 8.560712566694862e-05, "loss": 3.1509, "step": 298 }, { "epoch": 0.5758157389635317, "grad_norm": 2.3200299739837646, "learning_rate": 8.538340477185192e-05, "loss": 3.0537, "step": 300 }, { "epoch": 0.5796545105566219, "grad_norm": 2.545861005783081, "learning_rate": 8.515825575474221e-05, "loss": 3.0689, "step": 302 }, { "epoch": 0.5834932821497121, "grad_norm": 2.5010151863098145, "learning_rate": 8.493168770293792e-05, "loss": 3.0776, "step": 304 }, { "epoch": 0.5873320537428023, "grad_norm": 2.808166027069092, "learning_rate": 8.47037097610317e-05, "loss": 3.0201, "step": 306 }, { "epoch": 0.5911708253358925, "grad_norm": 2.355896234512329, "learning_rate": 8.447433113052123e-05, "loss": 3.0196, "step": 308 }, { "epoch": 0.5950095969289827, "grad_norm": 2.5034689903259277, "learning_rate": 8.42435610694379e-05, "loss": 3.0153, "step": 310 }, { "epoch": 0.5988483685220729, "grad_norm": 2.3233392238616943, "learning_rate": 8.401140889197305e-05, "loss": 2.9212, "step": 312 }, { "epoch": 0.6026871401151631, "grad_norm": 2.48526930809021, "learning_rate": 8.377788396810223e-05, "loss": 3.0123, "step": 314 }, { "epoch": 0.6065259117082533, "grad_norm": 3.0702834129333496, "learning_rate": 8.354299572320679e-05, "loss": 2.9393, "step": 316 }, { "epoch": 0.6103646833013435, "grad_norm": 2.031675100326538, "learning_rate": 8.330675363769356e-05, "loss": 3.0262, "step": 318 }, { "epoch": 0.6142034548944337, "grad_norm": 2.444798469543457, "learning_rate": 8.306916724661224e-05, "loss": 2.9865, "step": 320 }, { "epoch": 0.6180422264875239, "grad_norm": 2.245131254196167, "learning_rate": 8.283024613927054e-05, "loss": 3.0233, "step": 322 }, { "epoch": 0.6218809980806143, "grad_norm": 2.5943551063537598, "learning_rate": 8.258999995884705e-05, "loss": 2.897, "step": 324 }, { "epoch": 0.6257197696737045, "grad_norm": 2.6721882820129395, "learning_rate": 8.234843840200217e-05, "loss": 2.9885, "step": 326 }, { "epoch": 0.6295585412667947, "grad_norm": 2.2072489261627197, "learning_rate": 8.210557121848663e-05, "loss": 2.9196, "step": 328 }, { "epoch": 0.6333973128598849, "grad_norm": 2.7257487773895264, "learning_rate": 8.186140821074801e-05, "loss": 3.1456, "step": 330 }, { "epoch": 0.6372360844529751, "grad_norm": 2.159874200820923, "learning_rate": 8.161595923353516e-05, "loss": 2.9988, "step": 332 }, { "epoch": 0.6410748560460653, "grad_norm": 2.6628921031951904, "learning_rate": 8.136923419350031e-05, "loss": 3.0058, "step": 334 }, { "epoch": 0.6449136276391555, "grad_norm": 2.2145583629608154, "learning_rate": 8.112124304879937e-05, "loss": 2.9581, "step": 336 }, { "epoch": 0.6487523992322457, "grad_norm": 2.1396119594573975, "learning_rate": 8.087199580868996e-05, "loss": 2.9831, "step": 338 }, { "epoch": 0.6525911708253359, "grad_norm": 2.82460880279541, "learning_rate": 8.062150253312735e-05, "loss": 3.0879, "step": 340 }, { "epoch": 0.6564299424184261, "grad_norm": 2.2683770656585693, "learning_rate": 8.03697733323585e-05, "loss": 3.0195, "step": 342 }, { "epoch": 0.6602687140115163, "grad_norm": 2.508700132369995, "learning_rate": 8.011681836651402e-05, "loss": 2.9775, "step": 344 }, { "epoch": 0.6641074856046065, "grad_norm": 2.38877534866333, "learning_rate": 7.986264784519801e-05, "loss": 2.9442, "step": 346 }, { "epoch": 0.6679462571976967, "grad_norm": 2.2707676887512207, "learning_rate": 7.960727202707605e-05, "loss": 3.0087, "step": 348 }, { "epoch": 0.6717850287907869, "grad_norm": 2.509172201156616, "learning_rate": 7.935070121946115e-05, "loss": 3.0074, "step": 350 }, { "epoch": 0.6756238003838771, "grad_norm": 2.358182907104492, "learning_rate": 7.909294577789766e-05, "loss": 2.9645, "step": 352 }, { "epoch": 0.6794625719769674, "grad_norm": 2.6792430877685547, "learning_rate": 7.883401610574336e-05, "loss": 3.048, "step": 354 }, { "epoch": 0.6833013435700576, "grad_norm": 2.112786054611206, "learning_rate": 7.857392265374963e-05, "loss": 3.0222, "step": 356 }, { "epoch": 0.6871401151631478, "grad_norm": 2.4168145656585693, "learning_rate": 7.83126759196395e-05, "loss": 2.9943, "step": 358 }, { "epoch": 0.690978886756238, "grad_norm": 2.427769899368286, "learning_rate": 7.805028644768407e-05, "loss": 2.9786, "step": 360 }, { "epoch": 0.6948176583493282, "grad_norm": 2.3763539791107178, "learning_rate": 7.778676482827686e-05, "loss": 3.0007, "step": 362 }, { "epoch": 0.6986564299424184, "grad_norm": 1.8985060453414917, "learning_rate": 7.752212169750641e-05, "loss": 2.9787, "step": 364 }, { "epoch": 0.7024952015355086, "grad_norm": 2.2809348106384277, "learning_rate": 7.725636773672693e-05, "loss": 3.0175, "step": 366 }, { "epoch": 0.7063339731285988, "grad_norm": 2.4346015453338623, "learning_rate": 7.69895136721273e-05, "loss": 2.8707, "step": 368 }, { "epoch": 0.710172744721689, "grad_norm": 2.3430464267730713, "learning_rate": 7.672157027429803e-05, "loss": 2.9778, "step": 370 }, { "epoch": 0.7140115163147792, "grad_norm": 2.2653815746307373, "learning_rate": 7.645254835779657e-05, "loss": 3.0468, "step": 372 }, { "epoch": 0.7178502879078695, "grad_norm": 2.409104585647583, "learning_rate": 7.618245878071091e-05, "loss": 3.048, "step": 374 }, { "epoch": 0.7216890595009597, "grad_norm": 2.285231113433838, "learning_rate": 7.591131244422118e-05, "loss": 2.9787, "step": 376 }, { "epoch": 0.72552783109405, "grad_norm": 2.4652438163757324, "learning_rate": 7.563912029215983e-05, "loss": 2.9152, "step": 378 }, { "epoch": 0.7293666026871402, "grad_norm": 2.0631844997406006, "learning_rate": 7.536589331056976e-05, "loss": 3.0042, "step": 380 }, { "epoch": 0.7332053742802304, "grad_norm": 2.3582980632781982, "learning_rate": 7.509164252726107e-05, "loss": 3.0296, "step": 382 }, { "epoch": 0.7370441458733206, "grad_norm": 2.07973051071167, "learning_rate": 7.481637901136577e-05, "loss": 2.9678, "step": 384 }, { "epoch": 0.7408829174664108, "grad_norm": 2.2521328926086426, "learning_rate": 7.454011387289127e-05, "loss": 2.9467, "step": 386 }, { "epoch": 0.744721689059501, "grad_norm": 2.1111714839935303, "learning_rate": 7.42628582622717e-05, "loss": 2.7933, "step": 388 }, { "epoch": 0.7485604606525912, "grad_norm": 2.1656925678253174, "learning_rate": 7.398462336991801e-05, "loss": 2.8266, "step": 390 }, { "epoch": 0.7523992322456814, "grad_norm": 2.116373300552368, "learning_rate": 7.370542042576634e-05, "loss": 2.9002, "step": 392 }, { "epoch": 0.7562380038387716, "grad_norm": 2.533595323562622, "learning_rate": 7.342526069882465e-05, "loss": 3.035, "step": 394 }, { "epoch": 0.7600767754318618, "grad_norm": 2.229055404663086, "learning_rate": 7.314415549671795e-05, "loss": 2.9734, "step": 396 }, { "epoch": 0.763915547024952, "grad_norm": 2.5049266815185547, "learning_rate": 7.286211616523193e-05, "loss": 3.0225, "step": 398 }, { "epoch": 0.7677543186180422, "grad_norm": 2.2834606170654297, "learning_rate": 7.257915408785498e-05, "loss": 3.05, "step": 400 }, { "epoch": 0.7715930902111324, "grad_norm": 2.4472157955169678, "learning_rate": 7.229528068531882e-05, "loss": 2.9768, "step": 402 }, { "epoch": 0.7754318618042226, "grad_norm": 2.1033549308776855, "learning_rate": 7.201050741513736e-05, "loss": 2.7954, "step": 404 }, { "epoch": 0.7792706333973128, "grad_norm": 2.0633785724639893, "learning_rate": 7.17248457711445e-05, "loss": 2.9633, "step": 406 }, { "epoch": 0.783109404990403, "grad_norm": 2.160759925842285, "learning_rate": 7.14383072830301e-05, "loss": 2.9663, "step": 408 }, { "epoch": 0.7869481765834933, "grad_norm": 2.1165664196014404, "learning_rate": 7.115090351587454e-05, "loss": 2.9459, "step": 410 }, { "epoch": 0.7907869481765835, "grad_norm": 1.9940203428268433, "learning_rate": 7.086264606968214e-05, "loss": 2.9373, "step": 412 }, { "epoch": 0.7946257197696737, "grad_norm": 2.3808934688568115, "learning_rate": 7.05735465789128e-05, "loss": 3.0784, "step": 414 }, { "epoch": 0.7984644913627639, "grad_norm": 2.2294881343841553, "learning_rate": 7.028361671201245e-05, "loss": 2.9189, "step": 416 }, { "epoch": 0.8023032629558541, "grad_norm": 2.3160479068756104, "learning_rate": 6.99928681709422e-05, "loss": 2.8899, "step": 418 }, { "epoch": 0.8061420345489443, "grad_norm": 2.3562097549438477, "learning_rate": 6.970131269070592e-05, "loss": 3.0433, "step": 420 }, { "epoch": 0.8099808061420346, "grad_norm": 2.365830421447754, "learning_rate": 6.940896203887659e-05, "loss": 2.9003, "step": 422 }, { "epoch": 0.8138195777351248, "grad_norm": 2.432309865951538, "learning_rate": 6.911582801512147e-05, "loss": 2.9901, "step": 424 }, { "epoch": 0.817658349328215, "grad_norm": 2.517677068710327, "learning_rate": 6.882192245072569e-05, "loss": 2.9469, "step": 426 }, { "epoch": 0.8214971209213052, "grad_norm": 2.3116304874420166, "learning_rate": 6.852725720811488e-05, "loss": 2.909, "step": 428 }, { "epoch": 0.8253358925143954, "grad_norm": 2.257664442062378, "learning_rate": 6.823184418037624e-05, "loss": 2.8977, "step": 430 }, { "epoch": 0.8291746641074856, "grad_norm": 2.4709112644195557, "learning_rate": 6.793569529077864e-05, "loss": 2.8863, "step": 432 }, { "epoch": 0.8330134357005758, "grad_norm": 2.5286576747894287, "learning_rate": 6.76388224922913e-05, "loss": 2.8794, "step": 434 }, { "epoch": 0.836852207293666, "grad_norm": 2.203023910522461, "learning_rate": 6.734123776710137e-05, "loss": 2.9274, "step": 436 }, { "epoch": 0.8406909788867563, "grad_norm": 2.301381826400757, "learning_rate": 6.704295312613037e-05, "loss": 2.829, "step": 438 }, { "epoch": 0.8445297504798465, "grad_norm": 2.1027846336364746, "learning_rate": 6.674398060854931e-05, "loss": 2.8862, "step": 440 }, { "epoch": 0.8483685220729367, "grad_norm": 2.1715035438537598, "learning_rate": 6.644433228129287e-05, "loss": 2.9345, "step": 442 }, { "epoch": 0.8522072936660269, "grad_norm": 2.4000754356384277, "learning_rate": 6.614402023857232e-05, "loss": 2.9769, "step": 444 }, { "epoch": 0.8560460652591171, "grad_norm": 2.0978007316589355, "learning_rate": 6.584305660138734e-05, "loss": 2.9879, "step": 446 }, { "epoch": 0.8598848368522073, "grad_norm": 2.213789224624634, "learning_rate": 6.554145351703689e-05, "loss": 2.9495, "step": 448 }, { "epoch": 0.8637236084452975, "grad_norm": 2.410149335861206, "learning_rate": 6.523922315862887e-05, "loss": 2.8043, "step": 450 }, { "epoch": 0.8675623800383877, "grad_norm": 2.469250440597534, "learning_rate": 6.493637772458879e-05, "loss": 2.9194, "step": 452 }, { "epoch": 0.8714011516314779, "grad_norm": 2.257237195968628, "learning_rate": 6.463292943816746e-05, "loss": 3.008, "step": 454 }, { "epoch": 0.8752399232245681, "grad_norm": 2.0850181579589844, "learning_rate": 6.432889054694764e-05, "loss": 2.8868, "step": 456 }, { "epoch": 0.8790786948176583, "grad_norm": 2.2789504528045654, "learning_rate": 6.402427332234965e-05, "loss": 2.8147, "step": 458 }, { "epoch": 0.8829174664107485, "grad_norm": 2.0366861820220947, "learning_rate": 6.371909005913618e-05, "loss": 2.9317, "step": 460 }, { "epoch": 0.8867562380038387, "grad_norm": 2.035132884979248, "learning_rate": 6.341335307491595e-05, "loss": 2.8763, "step": 462 }, { "epoch": 0.8905950095969289, "grad_norm": 2.3231277465820312, "learning_rate": 6.310707470964668e-05, "loss": 2.8904, "step": 464 }, { "epoch": 0.8944337811900192, "grad_norm": 2.257615566253662, "learning_rate": 6.280026732513689e-05, "loss": 2.8535, "step": 466 }, { "epoch": 0.8982725527831094, "grad_norm": 2.0579357147216797, "learning_rate": 6.249294330454705e-05, "loss": 2.9112, "step": 468 }, { "epoch": 0.9021113243761996, "grad_norm": 2.0222885608673096, "learning_rate": 6.21851150518898e-05, "loss": 2.9029, "step": 470 }, { "epoch": 0.9059500959692899, "grad_norm": 2.360452651977539, "learning_rate": 6.187679499152919e-05, "loss": 2.9357, "step": 472 }, { "epoch": 0.9097888675623801, "grad_norm": 2.136718511581421, "learning_rate": 6.15679955676794e-05, "loss": 2.8546, "step": 474 }, { "epoch": 0.9136276391554703, "grad_norm": 2.203782081604004, "learning_rate": 6.125872924390226e-05, "loss": 2.8119, "step": 476 }, { "epoch": 0.9174664107485605, "grad_norm": 1.986873745918274, "learning_rate": 6.094900850260439e-05, "loss": 2.9073, "step": 478 }, { "epoch": 0.9213051823416507, "grad_norm": 2.3785810470581055, "learning_rate": 6.0638845844533254e-05, "loss": 2.9591, "step": 480 }, { "epoch": 0.9251439539347409, "grad_norm": 2.0311119556427, "learning_rate": 6.032825378827273e-05, "loss": 2.9071, "step": 482 }, { "epoch": 0.9289827255278311, "grad_norm": 2.1780667304992676, "learning_rate": 6.0017244869737746e-05, "loss": 2.8813, "step": 484 }, { "epoch": 0.9328214971209213, "grad_norm": 1.8662631511688232, "learning_rate": 5.970583164166838e-05, "loss": 2.8877, "step": 486 }, { "epoch": 0.9366602687140115, "grad_norm": 2.1670994758605957, "learning_rate": 5.9394026673123157e-05, "loss": 2.9122, "step": 488 }, { "epoch": 0.9404990403071017, "grad_norm": 1.9980396032333374, "learning_rate": 5.908184254897182e-05, "loss": 2.8776, "step": 490 }, { "epoch": 0.944337811900192, "grad_norm": 2.135673761367798, "learning_rate": 5.8769291869387335e-05, "loss": 2.9557, "step": 492 }, { "epoch": 0.9481765834932822, "grad_norm": 2.370652198791504, "learning_rate": 5.845638724933728e-05, "loss": 2.8789, "step": 494 }, { "epoch": 0.9520153550863724, "grad_norm": 2.0871102809906006, "learning_rate": 5.814314131807486e-05, "loss": 2.9034, "step": 496 }, { "epoch": 0.9558541266794626, "grad_norm": 1.9916837215423584, "learning_rate": 5.7829566718628946e-05, "loss": 2.8419, "step": 498 }, { "epoch": 0.9596928982725528, "grad_norm": 2.123903512954712, "learning_rate": 5.751567610729398e-05, "loss": 2.8393, "step": 500 }, { "epoch": 0.963531669865643, "grad_norm": 2.122802257537842, "learning_rate": 5.7201482153119015e-05, "loss": 2.8053, "step": 502 }, { "epoch": 0.9673704414587332, "grad_norm": 2.0287554264068604, "learning_rate": 5.6886997537396494e-05, "loss": 2.7966, "step": 504 }, { "epoch": 0.9712092130518234, "grad_norm": 2.3084447383880615, "learning_rate": 5.6572234953150305e-05, "loss": 2.9208, "step": 506 }, { "epoch": 0.9750479846449136, "grad_norm": 2.002812623977661, "learning_rate": 5.6257207104623524e-05, "loss": 2.8547, "step": 508 }, { "epoch": 0.9788867562380038, "grad_norm": 2.2170443534851074, "learning_rate": 5.594192670676568e-05, "loss": 2.7785, "step": 510 }, { "epoch": 0.982725527831094, "grad_norm": 2.0384042263031006, "learning_rate": 5.5626406484719506e-05, "loss": 2.8527, "step": 512 }, { "epoch": 0.9865642994241842, "grad_norm": 2.0327658653259277, "learning_rate": 5.5310659173307366e-05, "loss": 2.8564, "step": 514 }, { "epoch": 0.9904030710172744, "grad_norm": 2.0325183868408203, "learning_rate": 5.499469751651728e-05, "loss": 2.913, "step": 516 }, { "epoch": 0.9942418426103646, "grad_norm": 2.0311434268951416, "learning_rate": 5.467853426698853e-05, "loss": 2.9032, "step": 518 }, { "epoch": 0.9980806142034548, "grad_norm": 1.8845828771591187, "learning_rate": 5.43621821854969e-05, "loss": 2.8354, "step": 520 }, { "epoch": 1.0019193857965452, "grad_norm": 1.9669525623321533, "learning_rate": 5.4045654040439776e-05, "loss": 2.8744, "step": 522 }, { "epoch": 1.0057581573896353, "grad_norm": 1.953308343887329, "learning_rate": 5.3728962607320654e-05, "loss": 2.6639, "step": 524 }, { "epoch": 1.0095969289827256, "grad_norm": 1.8666753768920898, "learning_rate": 5.341212066823355e-05, "loss": 2.6542, "step": 526 }, { "epoch": 1.0134357005758157, "grad_norm": 1.8802012205123901, "learning_rate": 5.309514101134715e-05, "loss": 2.6397, "step": 528 }, { "epoch": 1.017274472168906, "grad_norm": 2.428213357925415, "learning_rate": 5.2778036430388554e-05, "loss": 2.4209, "step": 530 }, { "epoch": 1.021113243761996, "grad_norm": 2.179196834564209, "learning_rate": 5.246081972412702e-05, "loss": 2.7459, "step": 532 }, { "epoch": 1.0249520153550864, "grad_norm": 1.9748698472976685, "learning_rate": 5.214350369585731e-05, "loss": 2.5832, "step": 534 }, { "epoch": 1.0287907869481765, "grad_norm": 2.1484270095825195, "learning_rate": 5.182610115288295e-05, "loss": 2.6624, "step": 536 }, { "epoch": 1.0326295585412668, "grad_norm": 2.1627635955810547, "learning_rate": 5.150862490599934e-05, "loss": 2.6716, "step": 538 }, { "epoch": 1.036468330134357, "grad_norm": 1.800939679145813, "learning_rate": 5.119108776897664e-05, "loss": 2.6806, "step": 540 }, { "epoch": 1.0403071017274472, "grad_norm": 2.092155694961548, "learning_rate": 5.0873502558042665e-05, "loss": 2.5325, "step": 542 }, { "epoch": 1.0441458733205373, "grad_norm": 2.218632936477661, "learning_rate": 5.055588209136548e-05, "loss": 2.7483, "step": 544 }, { "epoch": 1.0479846449136276, "grad_norm": 1.7875282764434814, "learning_rate": 5.0238239188536215e-05, "loss": 2.4977, "step": 546 }, { "epoch": 1.051823416506718, "grad_norm": 2.2173759937286377, "learning_rate": 4.992058667005149e-05, "loss": 2.7608, "step": 548 }, { "epoch": 1.055662188099808, "grad_norm": 1.9905651807785034, "learning_rate": 4.9602937356796014e-05, "loss": 2.5805, "step": 550 }, { "epoch": 1.0595009596928984, "grad_norm": 2.1195759773254395, "learning_rate": 4.92853040695252e-05, "loss": 2.6454, "step": 552 }, { "epoch": 1.0633397312859885, "grad_norm": 2.1215133666992188, "learning_rate": 4.896769962834754e-05, "loss": 2.5693, "step": 554 }, { "epoch": 1.0671785028790788, "grad_norm": 2.111732244491577, "learning_rate": 4.865013685220733e-05, "loss": 2.5877, "step": 556 }, { "epoch": 1.0710172744721689, "grad_norm": 2.1130542755126953, "learning_rate": 4.83326285583672e-05, "loss": 2.7098, "step": 558 }, { "epoch": 1.0748560460652592, "grad_norm": 2.14615535736084, "learning_rate": 4.801518756189074e-05, "loss": 2.5208, "step": 560 }, { "epoch": 1.0786948176583493, "grad_norm": 2.0096733570098877, "learning_rate": 4.76978266751254e-05, "loss": 2.6982, "step": 562 }, { "epoch": 1.0825335892514396, "grad_norm": 1.9184404611587524, "learning_rate": 4.7380558707185284e-05, "loss": 2.3776, "step": 564 }, { "epoch": 1.0863723608445297, "grad_norm": 2.4721930027008057, "learning_rate": 4.7063396463434104e-05, "loss": 2.6396, "step": 566 }, { "epoch": 1.09021113243762, "grad_norm": 2.181126594543457, "learning_rate": 4.674635274496847e-05, "loss": 2.7188, "step": 568 }, { "epoch": 1.0940499040307101, "grad_norm": 1.7754825353622437, "learning_rate": 4.642944034810113e-05, "loss": 2.6317, "step": 570 }, { "epoch": 1.0978886756238004, "grad_norm": 2.0271716117858887, "learning_rate": 4.61126720638445e-05, "loss": 2.4098, "step": 572 }, { "epoch": 1.1017274472168905, "grad_norm": 1.935907006263733, "learning_rate": 4.5796060677394444e-05, "loss": 2.5833, "step": 574 }, { "epoch": 1.1055662188099808, "grad_norm": 2.4716265201568604, "learning_rate": 4.547961896761419e-05, "loss": 2.5757, "step": 576 }, { "epoch": 1.109404990403071, "grad_norm": 1.9446744918823242, "learning_rate": 4.51633597065186e-05, "loss": 2.5716, "step": 578 }, { "epoch": 1.1132437619961613, "grad_norm": 2.058483839035034, "learning_rate": 4.484729565875865e-05, "loss": 2.571, "step": 580 }, { "epoch": 1.1170825335892514, "grad_norm": 1.9952272176742554, "learning_rate": 4.4531439581106295e-05, "loss": 2.5598, "step": 582 }, { "epoch": 1.1209213051823417, "grad_norm": 2.0618462562561035, "learning_rate": 4.421580422193943e-05, "loss": 2.6121, "step": 584 }, { "epoch": 1.1247600767754318, "grad_norm": 1.943116307258606, "learning_rate": 4.390040232072755e-05, "loss": 2.5964, "step": 586 }, { "epoch": 1.128598848368522, "grad_norm": 1.9002699851989746, "learning_rate": 4.358524660751746e-05, "loss": 2.4376, "step": 588 }, { "epoch": 1.1324376199616122, "grad_norm": 1.9998996257781982, "learning_rate": 4.327034980241941e-05, "loss": 2.6728, "step": 590 }, { "epoch": 1.1362763915547025, "grad_norm": 2.234518051147461, "learning_rate": 4.2955724615093844e-05, "loss": 2.7121, "step": 592 }, { "epoch": 1.1401151631477928, "grad_norm": 2.0190210342407227, "learning_rate": 4.264138374423835e-05, "loss": 2.405, "step": 594 }, { "epoch": 1.143953934740883, "grad_norm": 2.088113307952881, "learning_rate": 4.232733987707505e-05, "loss": 2.7097, "step": 596 }, { "epoch": 1.147792706333973, "grad_norm": 1.9300215244293213, "learning_rate": 4.2013605688838656e-05, "loss": 2.4712, "step": 598 }, { "epoch": 1.1516314779270633, "grad_norm": 2.0667788982391357, "learning_rate": 4.1700193842264814e-05, "loss": 2.5939, "step": 600 }, { "epoch": 1.1554702495201536, "grad_norm": 1.9524165391921997, "learning_rate": 4.138711698707899e-05, "loss": 2.5213, "step": 602 }, { "epoch": 1.1593090211132437, "grad_norm": 2.1112565994262695, "learning_rate": 4.107438775948598e-05, "loss": 2.5185, "step": 604 }, { "epoch": 1.163147792706334, "grad_norm": 1.9430617094039917, "learning_rate": 4.076201878165985e-05, "loss": 2.5257, "step": 606 }, { "epoch": 1.1669865642994242, "grad_norm": 1.9651576280593872, "learning_rate": 4.0450022661234496e-05, "loss": 2.6055, "step": 608 }, { "epoch": 1.1708253358925145, "grad_norm": 2.1704463958740234, "learning_rate": 4.0138411990794786e-05, "loss": 2.625, "step": 610 }, { "epoch": 1.1746641074856046, "grad_norm": 2.089526891708374, "learning_rate": 3.982719934736832e-05, "loss": 2.4952, "step": 612 }, { "epoch": 1.1785028790786949, "grad_norm": 2.162886381149292, "learning_rate": 3.9516397291917754e-05, "loss": 2.5855, "step": 614 }, { "epoch": 1.182341650671785, "grad_norm": 1.9656240940093994, "learning_rate": 3.920601836883389e-05, "loss": 2.3817, "step": 616 }, { "epoch": 1.1861804222648753, "grad_norm": 2.060208559036255, "learning_rate": 3.889607510542936e-05, "loss": 2.4928, "step": 618 }, { "epoch": 1.1900191938579654, "grad_norm": 2.0905065536499023, "learning_rate": 3.858658001143289e-05, "loss": 2.6324, "step": 620 }, { "epoch": 1.1938579654510557, "grad_norm": 2.0947153568267822, "learning_rate": 3.82775455784846e-05, "loss": 2.4893, "step": 622 }, { "epoch": 1.1976967370441458, "grad_norm": 2.033522129058838, "learning_rate": 3.796898427963163e-05, "loss": 2.6109, "step": 624 }, { "epoch": 1.2015355086372361, "grad_norm": 2.082169771194458, "learning_rate": 3.7660908568824805e-05, "loss": 2.7201, "step": 626 }, { "epoch": 1.2053742802303262, "grad_norm": 1.9847004413604736, "learning_rate": 3.735333088041596e-05, "loss": 2.5707, "step": 628 }, { "epoch": 1.2092130518234165, "grad_norm": 1.8889490365982056, "learning_rate": 3.704626362865612e-05, "loss": 2.6231, "step": 630 }, { "epoch": 1.2130518234165066, "grad_norm": 2.0888187885284424, "learning_rate": 3.673971920719431e-05, "loss": 2.5646, "step": 632 }, { "epoch": 1.216890595009597, "grad_norm": 2.108222246170044, "learning_rate": 3.643370998857748e-05, "loss": 2.589, "step": 634 }, { "epoch": 1.220729366602687, "grad_norm": 2.094292402267456, "learning_rate": 3.612824832375109e-05, "loss": 2.3536, "step": 636 }, { "epoch": 1.2245681381957774, "grad_norm": 1.698121428489685, "learning_rate": 3.5823346541560495e-05, "loss": 2.3707, "step": 638 }, { "epoch": 1.2284069097888675, "grad_norm": 2.193507671356201, "learning_rate": 3.551901694825352e-05, "loss": 2.5158, "step": 640 }, { "epoch": 1.2322456813819578, "grad_norm": 1.975661277770996, "learning_rate": 3.521527182698365e-05, "loss": 2.4902, "step": 642 }, { "epoch": 1.236084452975048, "grad_norm": 1.9669182300567627, "learning_rate": 3.491212343731428e-05, "loss": 2.6996, "step": 644 }, { "epoch": 1.2399232245681382, "grad_norm": 2.053800344467163, "learning_rate": 3.4609584014723904e-05, "loss": 2.546, "step": 646 }, { "epoch": 1.2437619961612283, "grad_norm": 2.161032199859619, "learning_rate": 3.430766577011231e-05, "loss": 2.5643, "step": 648 }, { "epoch": 1.2476007677543186, "grad_norm": 1.9639465808868408, "learning_rate": 3.400638088930766e-05, "loss": 2.5846, "step": 650 }, { "epoch": 1.251439539347409, "grad_norm": 2.097055196762085, "learning_rate": 3.370574153257474e-05, "loss": 2.7445, "step": 652 }, { "epoch": 1.255278310940499, "grad_norm": 2.1084017753601074, "learning_rate": 3.340575983412412e-05, "loss": 2.5068, "step": 654 }, { "epoch": 1.2591170825335891, "grad_norm": 1.9084813594818115, "learning_rate": 3.3106447901622324e-05, "loss": 2.6058, "step": 656 }, { "epoch": 1.2629558541266794, "grad_norm": 1.8602763414382935, "learning_rate": 3.280781781570328e-05, "loss": 2.4718, "step": 658 }, { "epoch": 1.2667946257197698, "grad_norm": 2.2032907009124756, "learning_rate": 3.250988162948067e-05, "loss": 2.6737, "step": 660 }, { "epoch": 1.2706333973128598, "grad_norm": 1.9347671270370483, "learning_rate": 3.221265136806139e-05, "loss": 2.5647, "step": 662 }, { "epoch": 1.2744721689059502, "grad_norm": 2.094642400741577, "learning_rate": 3.1916139028060315e-05, "loss": 2.5481, "step": 664 }, { "epoch": 1.2783109404990403, "grad_norm": 2.167880058288574, "learning_rate": 3.162035657711604e-05, "loss": 2.6761, "step": 666 }, { "epoch": 1.2821497120921306, "grad_norm": 1.755005121231079, "learning_rate": 3.132531595340779e-05, "loss": 2.6924, "step": 668 }, { "epoch": 1.2859884836852207, "grad_norm": 2.031616449356079, "learning_rate": 3.1031029065173704e-05, "loss": 2.6972, "step": 670 }, { "epoch": 1.289827255278311, "grad_norm": 1.9664857387542725, "learning_rate": 3.073750779023014e-05, "loss": 2.64, "step": 672 }, { "epoch": 1.293666026871401, "grad_norm": 1.9512957334518433, "learning_rate": 3.0444763975492208e-05, "loss": 2.7087, "step": 674 }, { "epoch": 1.2975047984644914, "grad_norm": 1.8838787078857422, "learning_rate": 3.0152809436495734e-05, "loss": 2.3816, "step": 676 }, { "epoch": 1.3013435700575815, "grad_norm": 2.011364221572876, "learning_rate": 2.9861655956920286e-05, "loss": 2.4251, "step": 678 }, { "epoch": 1.3051823416506718, "grad_norm": 1.9579607248306274, "learning_rate": 2.95713152881136e-05, "loss": 2.6128, "step": 680 }, { "epoch": 1.309021113243762, "grad_norm": 1.9573185443878174, "learning_rate": 2.9281799148617262e-05, "loss": 2.5842, "step": 682 }, { "epoch": 1.3128598848368522, "grad_norm": 1.9739083051681519, "learning_rate": 2.8993119223693755e-05, "loss": 2.6536, "step": 684 }, { "epoch": 1.3166986564299425, "grad_norm": 2.0725173950195312, "learning_rate": 2.8705287164854755e-05, "loss": 2.6563, "step": 686 }, { "epoch": 1.3205374280230326, "grad_norm": 1.9268181324005127, "learning_rate": 2.8418314589391027e-05, "loss": 2.5909, "step": 688 }, { "epoch": 1.3243761996161227, "grad_norm": 2.035675525665283, "learning_rate": 2.8132213079903336e-05, "loss": 2.6599, "step": 690 }, { "epoch": 1.328214971209213, "grad_norm": 1.8536115884780884, "learning_rate": 2.7846994183835074e-05, "loss": 2.684, "step": 692 }, { "epoch": 1.3320537428023034, "grad_norm": 2.036652088165283, "learning_rate": 2.7562669413006148e-05, "loss": 2.4546, "step": 694 }, { "epoch": 1.3358925143953935, "grad_norm": 1.9004546403884888, "learning_rate": 2.7279250243148412e-05, "loss": 2.472, "step": 696 }, { "epoch": 1.3397312859884836, "grad_norm": 1.8974250555038452, "learning_rate": 2.6996748113442394e-05, "loss": 2.4129, "step": 698 }, { "epoch": 1.3435700575815739, "grad_norm": 2.1649343967437744, "learning_rate": 2.6715174426055662e-05, "loss": 2.6387, "step": 700 }, { "epoch": 1.3474088291746642, "grad_norm": 1.8146731853485107, "learning_rate": 2.6434540545682584e-05, "loss": 2.4887, "step": 702 }, { "epoch": 1.3512476007677543, "grad_norm": 2.0081663131713867, "learning_rate": 2.6154857799085643e-05, "loss": 2.4244, "step": 704 }, { "epoch": 1.3550863723608444, "grad_norm": 1.8678373098373413, "learning_rate": 2.5876137474638323e-05, "loss": 2.6523, "step": 706 }, { "epoch": 1.3589251439539347, "grad_norm": 2.149458169937134, "learning_rate": 2.55983908218694e-05, "loss": 2.635, "step": 708 }, { "epoch": 1.362763915547025, "grad_norm": 1.8371909856796265, "learning_rate": 2.5321629051008976e-05, "loss": 2.4454, "step": 710 }, { "epoch": 1.3666026871401151, "grad_norm": 1.733880639076233, "learning_rate": 2.5045863332535945e-05, "loss": 2.3082, "step": 712 }, { "epoch": 1.3704414587332054, "grad_norm": 1.7992384433746338, "learning_rate": 2.4771104796727273e-05, "loss": 2.5468, "step": 714 }, { "epoch": 1.3742802303262955, "grad_norm": 1.8576358556747437, "learning_rate": 2.449736453320854e-05, "loss": 2.6041, "step": 716 }, { "epoch": 1.3781190019193859, "grad_norm": 1.9266445636749268, "learning_rate": 2.4224653590506607e-05, "loss": 2.5576, "step": 718 }, { "epoch": 1.381957773512476, "grad_norm": 1.8108105659484863, "learning_rate": 2.3952982975603496e-05, "loss": 2.4815, "step": 720 }, { "epoch": 1.3857965451055663, "grad_norm": 1.8349239826202393, "learning_rate": 2.3682363653492178e-05, "loss": 2.5743, "step": 722 }, { "epoch": 1.3896353166986564, "grad_norm": 2.0299675464630127, "learning_rate": 2.3412806546734055e-05, "loss": 2.5463, "step": 724 }, { "epoch": 1.3934740882917467, "grad_norm": 2.3662514686584473, "learning_rate": 2.3144322535018127e-05, "loss": 2.5726, "step": 726 }, { "epoch": 1.3973128598848368, "grad_norm": 1.9060813188552856, "learning_rate": 2.2876922454721694e-05, "loss": 2.4306, "step": 728 }, { "epoch": 1.401151631477927, "grad_norm": 1.8330024480819702, "learning_rate": 2.261061709847327e-05, "loss": 2.5845, "step": 730 }, { "epoch": 1.4049904030710172, "grad_norm": 2.0324208736419678, "learning_rate": 2.2345417214716748e-05, "loss": 2.573, "step": 732 }, { "epoch": 1.4088291746641075, "grad_norm": 1.9950467348098755, "learning_rate": 2.208133350727764e-05, "loss": 2.5313, "step": 734 }, { "epoch": 1.4126679462571978, "grad_norm": 2.015324354171753, "learning_rate": 2.1818376634931153e-05, "loss": 2.663, "step": 736 }, { "epoch": 1.416506717850288, "grad_norm": 2.071497917175293, "learning_rate": 2.1556557210971847e-05, "loss": 2.5707, "step": 738 }, { "epoch": 1.420345489443378, "grad_norm": 1.9036519527435303, "learning_rate": 2.129588580278533e-05, "loss": 2.5151, "step": 740 }, { "epoch": 1.4241842610364683, "grad_norm": 1.898582100868225, "learning_rate": 2.103637293142175e-05, "loss": 2.4412, "step": 742 }, { "epoch": 1.4280230326295587, "grad_norm": 1.951640248298645, "learning_rate": 2.077802907117119e-05, "loss": 2.6103, "step": 744 }, { "epoch": 1.4318618042226487, "grad_norm": 1.9598643779754639, "learning_rate": 2.0520864649140765e-05, "loss": 2.5161, "step": 746 }, { "epoch": 1.4357005758157388, "grad_norm": 1.943354606628418, "learning_rate": 2.0264890044833995e-05, "loss": 2.6389, "step": 748 }, { "epoch": 1.4395393474088292, "grad_norm": 1.9923816919326782, "learning_rate": 2.0010115589731613e-05, "loss": 2.4627, "step": 750 }, { "epoch": 1.4433781190019195, "grad_norm": 2.0418148040771484, "learning_rate": 1.9756551566874837e-05, "loss": 2.5058, "step": 752 }, { "epoch": 1.4472168905950096, "grad_norm": 2.014249324798584, "learning_rate": 1.9504208210450127e-05, "loss": 2.2585, "step": 754 }, { "epoch": 1.4510556621880997, "grad_norm": 1.9214091300964355, "learning_rate": 1.9253095705376218e-05, "loss": 2.681, "step": 756 }, { "epoch": 1.45489443378119, "grad_norm": 29.086570739746094, "learning_rate": 1.9003224186893e-05, "loss": 2.542, "step": 758 }, { "epoch": 1.4587332053742803, "grad_norm": 2.7266669273376465, "learning_rate": 1.875460374015253e-05, "loss": 2.5041, "step": 760 }, { "epoch": 1.4625719769673704, "grad_norm": 1.799980878829956, "learning_rate": 1.850724439981186e-05, "loss": 2.4972, "step": 762 }, { "epoch": 1.4664107485604607, "grad_norm": 2.1353044509887695, "learning_rate": 1.82611561496281e-05, "loss": 2.6476, "step": 764 }, { "epoch": 1.4702495201535508, "grad_norm": 1.9255475997924805, "learning_rate": 1.801634892205545e-05, "loss": 2.4671, "step": 766 }, { "epoch": 1.4740882917466411, "grad_norm": 1.7508608102798462, "learning_rate": 1.7772832597844285e-05, "loss": 2.4502, "step": 768 }, { "epoch": 1.4779270633397312, "grad_norm": 1.9461318254470825, "learning_rate": 1.753061700564243e-05, "loss": 2.6253, "step": 770 }, { "epoch": 1.4817658349328215, "grad_norm": 1.8493483066558838, "learning_rate": 1.728971192159836e-05, "loss": 2.5212, "step": 772 }, { "epoch": 1.4856046065259116, "grad_norm": 1.83624267578125, "learning_rate": 1.705012706896668e-05, "loss": 2.4159, "step": 774 }, { "epoch": 1.489443378119002, "grad_norm": 1.761613368988037, "learning_rate": 1.6811872117715673e-05, "loss": 2.5352, "step": 776 }, { "epoch": 1.493282149712092, "grad_norm": 1.8338181972503662, "learning_rate": 1.6574956684137045e-05, "loss": 2.575, "step": 778 }, { "epoch": 1.4971209213051824, "grad_norm": 1.831794023513794, "learning_rate": 1.633939033045766e-05, "loss": 2.5734, "step": 780 }, { "epoch": 1.5009596928982725, "grad_norm": 1.8109732866287231, "learning_rate": 1.610518256445382e-05, "loss": 2.5865, "step": 782 }, { "epoch": 1.5047984644913628, "grad_norm": 1.97743558883667, "learning_rate": 1.5872342839067306e-05, "loss": 2.4631, "step": 784 }, { "epoch": 1.508637236084453, "grad_norm": 1.7919214963912964, "learning_rate": 1.5640880552023955e-05, "loss": 2.6693, "step": 786 }, { "epoch": 1.5124760076775432, "grad_norm": 1.7384066581726074, "learning_rate": 1.541080504545433e-05, "loss": 2.4759, "step": 788 }, { "epoch": 1.5163147792706333, "grad_norm": 1.898398518562317, "learning_rate": 1.5182125605516707e-05, "loss": 2.6413, "step": 790 }, { "epoch": 1.5201535508637236, "grad_norm": 1.8213528394699097, "learning_rate": 1.4954851462022118e-05, "loss": 2.5277, "step": 792 }, { "epoch": 1.523992322456814, "grad_norm": 1.7524523735046387, "learning_rate": 1.4728991788062052e-05, "loss": 2.3249, "step": 794 }, { "epoch": 1.527831094049904, "grad_norm": 1.9670718908309937, "learning_rate": 1.4504555699638034e-05, "loss": 2.3528, "step": 796 }, { "epoch": 1.5316698656429941, "grad_norm": 1.92782723903656, "learning_rate": 1.4281552255293739e-05, "loss": 2.5678, "step": 798 }, { "epoch": 1.5355086372360844, "grad_norm": 1.8225903511047363, "learning_rate": 1.405999045574945e-05, "loss": 2.4719, "step": 800 }, { "epoch": 1.5393474088291748, "grad_norm": 1.6242417097091675, "learning_rate": 1.3839879243538678e-05, "loss": 2.2137, "step": 802 }, { "epoch": 1.5431861804222649, "grad_norm": 1.7999491691589355, "learning_rate": 1.362122750264727e-05, "loss": 2.5464, "step": 804 }, { "epoch": 1.547024952015355, "grad_norm": 2.130880355834961, "learning_rate": 1.3404044058154835e-05, "loss": 2.5688, "step": 806 }, { "epoch": 1.5508637236084453, "grad_norm": 2.0367865562438965, "learning_rate": 1.3188337675878609e-05, "loss": 2.363, "step": 808 }, { "epoch": 1.5547024952015356, "grad_norm": 1.763808250427246, "learning_rate": 1.2974117062019503e-05, "loss": 2.6023, "step": 810 }, { "epoch": 1.5585412667946257, "grad_norm": 3.7647054195404053, "learning_rate": 1.2761390862810907e-05, "loss": 2.5587, "step": 812 }, { "epoch": 1.5623800383877158, "grad_norm": 1.8591704368591309, "learning_rate": 1.2550167664169566e-05, "loss": 2.6078, "step": 814 }, { "epoch": 1.566218809980806, "grad_norm": 1.7693856954574585, "learning_rate": 1.2340455991349093e-05, "loss": 2.5316, "step": 816 }, { "epoch": 1.5700575815738964, "grad_norm": 1.9136343002319336, "learning_rate": 1.2132264308595875e-05, "loss": 2.5943, "step": 818 }, { "epoch": 1.5738963531669867, "grad_norm": 1.8187044858932495, "learning_rate": 1.1925601018807497e-05, "loss": 2.6485, "step": 820 }, { "epoch": 1.5777351247600768, "grad_norm": 1.899484634399414, "learning_rate": 1.1720474463193443e-05, "loss": 2.5813, "step": 822 }, { "epoch": 1.581573896353167, "grad_norm": 1.7205506563186646, "learning_rate": 1.1516892920938626e-05, "loss": 2.5842, "step": 824 }, { "epoch": 1.5854126679462572, "grad_norm": 2.0414419174194336, "learning_rate": 1.131486460886908e-05, "loss": 2.5694, "step": 826 }, { "epoch": 1.5892514395393476, "grad_norm": 1.8457170724868774, "learning_rate": 1.1114397681120387e-05, "loss": 2.5559, "step": 828 }, { "epoch": 1.5930902111324377, "grad_norm": 1.9401755332946777, "learning_rate": 1.0915500228808523e-05, "loss": 2.5666, "step": 830 }, { "epoch": 1.5969289827255277, "grad_norm": 2.234301805496216, "learning_rate": 1.0718180279703371e-05, "loss": 2.4235, "step": 832 }, { "epoch": 1.600767754318618, "grad_norm": 1.9198181629180908, "learning_rate": 1.0522445797904607e-05, "loss": 2.5202, "step": 834 }, { "epoch": 1.6046065259117084, "grad_norm": 2.0166404247283936, "learning_rate": 1.0328304683520306e-05, "loss": 2.5605, "step": 836 }, { "epoch": 1.6084452975047985, "grad_norm": 1.7829755544662476, "learning_rate": 1.0135764772348106e-05, "loss": 2.5584, "step": 838 }, { "epoch": 1.6122840690978886, "grad_norm": 1.8501259088516235, "learning_rate": 9.944833835558887e-06, "loss": 2.62, "step": 840 }, { "epoch": 1.616122840690979, "grad_norm": 1.8745450973510742, "learning_rate": 9.755519579383204e-06, "loss": 2.59, "step": 842 }, { "epoch": 1.6199616122840692, "grad_norm": 1.8195083141326904, "learning_rate": 9.567829644800141e-06, "loss": 2.6145, "step": 844 }, { "epoch": 1.6238003838771593, "grad_norm": 1.72378408908844, "learning_rate": 9.381771607229e-06, "loss": 2.5924, "step": 846 }, { "epoch": 1.6276391554702494, "grad_norm": 2.0315263271331787, "learning_rate": 9.197352976223495e-06, "loss": 2.534, "step": 848 }, { "epoch": 1.6314779270633397, "grad_norm": 1.7934647798538208, "learning_rate": 9.014581195168725e-06, "loss": 2.5265, "step": 850 }, { "epoch": 1.63531669865643, "grad_norm": 1.8798456192016602, "learning_rate": 8.83346364098061e-06, "loss": 2.4499, "step": 852 }, { "epoch": 1.6391554702495201, "grad_norm": 1.756456732749939, "learning_rate": 8.654007623808335e-06, "loss": 2.6371, "step": 854 }, { "epoch": 1.6429942418426102, "grad_norm": 1.8136521577835083, "learning_rate": 8.476220386739153e-06, "loss": 2.3707, "step": 856 }, { "epoch": 1.6468330134357005, "grad_norm": 1.8355995416641235, "learning_rate": 8.30010910550611e-06, "loss": 2.2622, "step": 858 }, { "epoch": 1.6506717850287909, "grad_norm": 1.716508388519287, "learning_rate": 8.125680888198396e-06, "loss": 2.6421, "step": 860 }, { "epoch": 1.654510556621881, "grad_norm": 1.8933945894241333, "learning_rate": 7.95294277497452e-06, "loss": 2.6159, "step": 862 }, { "epoch": 1.658349328214971, "grad_norm": 1.8038040399551392, "learning_rate": 7.781901737778013e-06, "loss": 2.4949, "step": 864 }, { "epoch": 1.6621880998080614, "grad_norm": 1.9260976314544678, "learning_rate": 7.612564680056233e-06, "loss": 2.4853, "step": 866 }, { "epoch": 1.6660268714011517, "grad_norm": 1.903968095779419, "learning_rate": 7.4449384364815464e-06, "loss": 2.4522, "step": 868 }, { "epoch": 1.669865642994242, "grad_norm": 1.855387806892395, "learning_rate": 7.2790297726755716e-06, "loss": 2.4027, "step": 870 }, { "epoch": 1.673704414587332, "grad_norm": 1.8166786432266235, "learning_rate": 7.114845384936108e-06, "loss": 2.4708, "step": 872 }, { "epoch": 1.6775431861804222, "grad_norm": 1.9112259149551392, "learning_rate": 6.9523918999668255e-06, "loss": 2.4039, "step": 874 }, { "epoch": 1.6813819577735125, "grad_norm": 1.8451069593429565, "learning_rate": 6.7916758746098155e-06, "loss": 2.4916, "step": 876 }, { "epoch": 1.6852207293666028, "grad_norm": 1.9163312911987305, "learning_rate": 6.632703795580946e-06, "loss": 2.5985, "step": 878 }, { "epoch": 1.689059500959693, "grad_norm": 1.9589265584945679, "learning_rate": 6.475482079208112e-06, "loss": 2.4371, "step": 880 }, { "epoch": 1.692898272552783, "grad_norm": 1.7410019636154175, "learning_rate": 6.320017071172113e-06, "loss": 2.3609, "step": 882 }, { "epoch": 1.6967370441458733, "grad_norm": 1.7839226722717285, "learning_rate": 6.1663150462507034e-06, "loss": 2.4713, "step": 884 }, { "epoch": 1.7005758157389637, "grad_norm": 2.5581276416778564, "learning_rate": 6.014382208065234e-06, "loss": 2.5357, "step": 886 }, { "epoch": 1.7044145873320538, "grad_norm": 1.7004822492599487, "learning_rate": 5.864224688830283e-06, "loss": 2.3301, "step": 888 }, { "epoch": 1.7082533589251438, "grad_norm": 1.6520910263061523, "learning_rate": 5.715848549106145e-06, "loss": 2.3737, "step": 890 }, { "epoch": 1.7120921305182342, "grad_norm": 1.77219820022583, "learning_rate": 5.569259777554287e-06, "loss": 2.5417, "step": 892 }, { "epoch": 1.7159309021113245, "grad_norm": 1.8998414278030396, "learning_rate": 5.424464290695496e-06, "loss": 2.5243, "step": 894 }, { "epoch": 1.7197696737044146, "grad_norm": 1.8390710353851318, "learning_rate": 5.281467932671253e-06, "loss": 2.4847, "step": 896 }, { "epoch": 1.7236084452975047, "grad_norm": 1.7345848083496094, "learning_rate": 5.140276475007711e-06, "loss": 2.6227, "step": 898 }, { "epoch": 1.727447216890595, "grad_norm": 1.7888513803482056, "learning_rate": 5.000895616382828e-06, "loss": 2.4572, "step": 900 }, { "epoch": 1.7312859884836853, "grad_norm": 1.8349541425704956, "learning_rate": 4.863330982396319e-06, "loss": 2.435, "step": 902 }, { "epoch": 1.7351247600767754, "grad_norm": 1.7568241357803345, "learning_rate": 4.727588125342669e-06, "loss": 2.3567, "step": 904 }, { "epoch": 1.7389635316698655, "grad_norm": 1.8719154596328735, "learning_rate": 4.593672523986936e-06, "loss": 2.4425, "step": 906 }, { "epoch": 1.7428023032629558, "grad_norm": 1.8745254278182983, "learning_rate": 4.461589583343678e-06, "loss": 2.4608, "step": 908 }, { "epoch": 1.7466410748560461, "grad_norm": 1.7677924633026123, "learning_rate": 4.331344634458812e-06, "loss": 2.39, "step": 910 }, { "epoch": 1.7504798464491362, "grad_norm": 1.7285584211349487, "learning_rate": 4.2029429341943984e-06, "loss": 2.4978, "step": 912 }, { "epoch": 1.7543186180422263, "grad_norm": 1.7370328903198242, "learning_rate": 4.076389665016522e-06, "loss": 2.5214, "step": 914 }, { "epoch": 1.7581573896353166, "grad_norm": 1.7510559558868408, "learning_rate": 3.951689934786068e-06, "loss": 2.4772, "step": 916 }, { "epoch": 1.761996161228407, "grad_norm": 1.6975350379943848, "learning_rate": 3.828848776552596e-06, "loss": 2.4112, "step": 918 }, { "epoch": 1.7658349328214973, "grad_norm": 1.8429741859436035, "learning_rate": 3.707871148351183e-06, "loss": 2.5379, "step": 920 }, { "epoch": 1.7696737044145874, "grad_norm": 1.8487157821655273, "learning_rate": 3.5887619330023435e-06, "loss": 2.5535, "step": 922 }, { "epoch": 1.7735124760076775, "grad_norm": 1.7282496690750122, "learning_rate": 3.4715259379148656e-06, "loss": 2.4999, "step": 924 }, { "epoch": 1.7773512476007678, "grad_norm": 1.7872588634490967, "learning_rate": 3.3561678948919095e-06, "loss": 2.3744, "step": 926 }, { "epoch": 1.781190019193858, "grad_norm": 1.9310446977615356, "learning_rate": 3.2426924599399056e-06, "loss": 2.3609, "step": 928 }, { "epoch": 1.7850287907869482, "grad_norm": 1.6719673871994019, "learning_rate": 3.131104213080688e-06, "loss": 2.5972, "step": 930 }, { "epoch": 1.7888675623800383, "grad_norm": 1.8342926502227783, "learning_rate": 3.021407658166636e-06, "loss": 2.4144, "step": 932 }, { "epoch": 1.7927063339731286, "grad_norm": 1.7356289625167847, "learning_rate": 2.913607222698905e-06, "loss": 2.4736, "step": 934 }, { "epoch": 1.796545105566219, "grad_norm": 1.7824493646621704, "learning_rate": 2.807707257648662e-06, "loss": 2.4878, "step": 936 }, { "epoch": 1.800383877159309, "grad_norm": 1.7247017621994019, "learning_rate": 2.703712037281564e-06, "loss": 2.5778, "step": 938 }, { "epoch": 1.8042226487523991, "grad_norm": 1.6570117473602295, "learning_rate": 2.601625758985182e-06, "loss": 2.5897, "step": 940 }, { "epoch": 1.8080614203454894, "grad_norm": 1.7718125581741333, "learning_rate": 2.5014525430995916e-06, "loss": 2.4034, "step": 942 }, { "epoch": 1.8119001919385798, "grad_norm": 1.9651246070861816, "learning_rate": 2.4031964327511313e-06, "loss": 2.5621, "step": 944 }, { "epoch": 1.8157389635316699, "grad_norm": 1.9165457487106323, "learning_rate": 2.3068613936891137e-06, "loss": 2.5763, "step": 946 }, { "epoch": 1.81957773512476, "grad_norm": 1.5878627300262451, "learning_rate": 2.2124513141258574e-06, "loss": 2.5116, "step": 948 }, { "epoch": 1.8234165067178503, "grad_norm": 1.7749850749969482, "learning_rate": 2.1199700045797077e-06, "loss": 2.5202, "step": 950 }, { "epoch": 1.8272552783109406, "grad_norm": 1.7712241411209106, "learning_rate": 2.0294211977212317e-06, "loss": 2.5894, "step": 952 }, { "epoch": 1.8310940499040307, "grad_norm": 1.7715948820114136, "learning_rate": 1.9408085482225945e-06, "loss": 2.6296, "step": 954 }, { "epoch": 1.8349328214971208, "grad_norm": 1.7218071222305298, "learning_rate": 1.8541356326100433e-06, "loss": 2.4942, "step": 956 }, { "epoch": 1.838771593090211, "grad_norm": 1.6863584518432617, "learning_rate": 1.7694059491195013e-06, "loss": 2.5955, "step": 958 }, { "epoch": 1.8426103646833014, "grad_norm": 1.718625545501709, "learning_rate": 1.6866229175554748e-06, "loss": 2.4151, "step": 960 }, { "epoch": 1.8464491362763915, "grad_norm": 1.6728721857070923, "learning_rate": 1.6057898791529303e-06, "loss": 2.4021, "step": 962 }, { "epoch": 1.8502879078694816, "grad_norm": 1.6578346490859985, "learning_rate": 1.526910096442491e-06, "loss": 2.2934, "step": 964 }, { "epoch": 1.854126679462572, "grad_norm": 1.642450213432312, "learning_rate": 1.4499867531187373e-06, "loss": 2.2958, "step": 966 }, { "epoch": 1.8579654510556622, "grad_norm": 1.908416509628296, "learning_rate": 1.3750229539117144e-06, "loss": 2.4597, "step": 968 }, { "epoch": 1.8618042226487526, "grad_norm": 1.6803356409072876, "learning_rate": 1.3020217244616273e-06, "loss": 2.4595, "step": 970 }, { "epoch": 1.8656429942418427, "grad_norm": 1.7671664953231812, "learning_rate": 1.2309860111967053e-06, "loss": 2.5175, "step": 972 }, { "epoch": 1.8694817658349328, "grad_norm": 1.7206978797912598, "learning_rate": 1.1619186812142857e-06, "loss": 2.5618, "step": 974 }, { "epoch": 1.873320537428023, "grad_norm": 1.7256447076797485, "learning_rate": 1.094822522165101e-06, "loss": 2.6376, "step": 976 }, { "epoch": 1.8771593090211134, "grad_norm": 1.7528103590011597, "learning_rate": 1.0297002421407797e-06, "loss": 2.5412, "step": 978 }, { "epoch": 1.8809980806142035, "grad_norm": 1.7252179384231567, "learning_rate": 9.665544695645012e-07, "loss": 2.5049, "step": 980 }, { "epoch": 1.8848368522072936, "grad_norm": 1.7561063766479492, "learning_rate": 9.053877530849464e-07, "loss": 2.4125, "step": 982 }, { "epoch": 1.888675623800384, "grad_norm": 1.961869716644287, "learning_rate": 8.462025614734192e-07, "loss": 2.5765, "step": 984 }, { "epoch": 1.8925143953934742, "grad_norm": 1.6624488830566406, "learning_rate": 7.890012835242044e-07, "loss": 2.5783, "step": 986 }, { "epoch": 1.8963531669865643, "grad_norm": 1.7556750774383545, "learning_rate": 7.337862279581332e-07, "loss": 2.5598, "step": 988 }, { "epoch": 1.9001919385796544, "grad_norm": 1.8471143245697021, "learning_rate": 6.805596233294576e-07, "loss": 2.504, "step": 990 }, { "epoch": 1.9040307101727447, "grad_norm": 1.8926904201507568, "learning_rate": 6.293236179358175e-07, "loss": 2.6038, "step": 992 }, { "epoch": 1.907869481765835, "grad_norm": 1.6474649906158447, "learning_rate": 5.800802797316152e-07, "loss": 2.3219, "step": 994 }, { "epoch": 1.9117082533589251, "grad_norm": 1.7350983619689941, "learning_rate": 5.328315962444874e-07, "loss": 2.6234, "step": 996 }, { "epoch": 1.9155470249520152, "grad_norm": 1.6887606382369995, "learning_rate": 4.875794744951423e-07, "loss": 2.2119, "step": 998 }, { "epoch": 1.9193857965451055, "grad_norm": 3.0028889179229736, "learning_rate": 4.4432574092032053e-07, "loss": 2.3565, "step": 1000 }, { "epoch": 1.9232245681381959, "grad_norm": 1.760825276374817, "learning_rate": 4.0307214129914896e-07, "loss": 2.5502, "step": 1002 }, { "epoch": 1.927063339731286, "grad_norm": 1.753866195678711, "learning_rate": 3.6382034068263015e-07, "loss": 2.4625, "step": 1004 }, { "epoch": 1.930902111324376, "grad_norm": 1.8474925756454468, "learning_rate": 3.265719233264575e-07, "loss": 2.5873, "step": 1006 }, { "epoch": 1.9347408829174664, "grad_norm": 1.8238581418991089, "learning_rate": 2.913283926270771e-07, "loss": 2.5435, "step": 1008 }, { "epoch": 1.9385796545105567, "grad_norm": 1.8236713409423828, "learning_rate": 2.580911710609923e-07, "loss": 2.6754, "step": 1010 }, { "epoch": 1.9424184261036468, "grad_norm": 1.667476773262024, "learning_rate": 2.2686160012735936e-07, "loss": 2.3779, "step": 1012 }, { "epoch": 1.9462571976967369, "grad_norm": 1.635819673538208, "learning_rate": 1.9764094029385284e-07, "loss": 2.3852, "step": 1014 }, { "epoch": 1.9500959692898272, "grad_norm": 1.6991413831710815, "learning_rate": 1.704303709457733e-07, "loss": 2.3626, "step": 1016 }, { "epoch": 1.9539347408829175, "grad_norm": 1.8592206239700317, "learning_rate": 1.452309903384519e-07, "loss": 2.4013, "step": 1018 }, { "epoch": 1.9577735124760078, "grad_norm": 1.7539414167404175, "learning_rate": 1.220438155529302e-07, "loss": 2.5873, "step": 1020 }, { "epoch": 1.961612284069098, "grad_norm": 1.7539663314819336, "learning_rate": 1.0086978245490985e-07, "loss": 2.5274, "step": 1022 }, { "epoch": 1.965451055662188, "grad_norm": 1.8652434349060059, "learning_rate": 8.170974565696599e-08, "loss": 2.5122, "step": 1024 }, { "epoch": 1.9692898272552783, "grad_norm": 1.7227452993392944, "learning_rate": 6.456447848406932e-08, "loss": 2.5307, "step": 1026 }, { "epoch": 1.9731285988483687, "grad_norm": 4.939391136169434, "learning_rate": 4.94346729423556e-08, "loss": 2.5132, "step": 1028 }, { "epoch": 1.9769673704414588, "grad_norm": 1.815574288368225, "learning_rate": 3.632093969121453e-08, "loss": 2.5618, "step": 1030 }, { "epoch": 1.9808061420345489, "grad_norm": 1.6887080669403076, "learning_rate": 2.522380801863733e-08, "loss": 2.4683, "step": 1032 }, { "epoch": 1.9846449136276392, "grad_norm": 1.627553105354309, "learning_rate": 1.6143725819850463e-08, "loss": 2.2912, "step": 1034 }, { "epoch": 1.9884836852207295, "grad_norm": 1.7982923984527588, "learning_rate": 9.081059579235662e-09, "loss": 2.4592, "step": 1036 }, { "epoch": 1.9923224568138196, "grad_norm": 1.8788052797317505, "learning_rate": 4.036094355541753e-09, "loss": 2.508, "step": 1038 }, { "epoch": 1.9961612284069097, "grad_norm": 1.7438900470733643, "learning_rate": 1.0090337703771991e-09, "loss": 2.5036, "step": 1040 }, { "epoch": 2.0, "grad_norm": 2.2197341918945312, "learning_rate": 0.0, "loss": 2.2714, "step": 1042 } ], "logging_steps": 2, "max_steps": 1042, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 521, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.762124413890724e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }