{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 19.942103570279833, "eval_steps": 500, "global_step": 62000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0321646831778707, "grad_norm": 1.2931797504425049, "learning_rate": 4.992039240913477e-05, "loss": 1.7938, "step": 100 }, { "epoch": 0.0643293663557414, "grad_norm": 1.2261773347854614, "learning_rate": 4.98399807011901e-05, "loss": 1.4694, "step": 200 }, { "epoch": 0.0964940495336121, "grad_norm": 1.1642584800720215, "learning_rate": 4.975956899324542e-05, "loss": 1.3725, "step": 300 }, { "epoch": 0.1286587327114828, "grad_norm": 1.1396938562393188, "learning_rate": 4.967915728530074e-05, "loss": 1.3302, "step": 400 }, { "epoch": 0.16082341588935348, "grad_norm": 1.0271061658859253, "learning_rate": 4.959874557735607e-05, "loss": 1.2981, "step": 500 }, { "epoch": 0.1929880990672242, "grad_norm": 1.0276116132736206, "learning_rate": 4.951833386941139e-05, "loss": 1.2702, "step": 600 }, { "epoch": 0.22515278224509488, "grad_norm": 0.9277088046073914, "learning_rate": 4.943792216146671e-05, "loss": 1.2354, "step": 700 }, { "epoch": 0.2573174654229656, "grad_norm": 0.8870278596878052, "learning_rate": 4.935751045352204e-05, "loss": 1.2262, "step": 800 }, { "epoch": 0.28948214860083626, "grad_norm": 0.9166188836097717, "learning_rate": 4.927709874557736e-05, "loss": 1.2211, "step": 900 }, { "epoch": 0.32164683177870695, "grad_norm": 0.9699078798294067, "learning_rate": 4.9196687037632685e-05, "loss": 1.1927, "step": 1000 }, { "epoch": 0.3538115149565777, "grad_norm": 0.8429589867591858, "learning_rate": 4.9116275329688006e-05, "loss": 1.1822, "step": 1100 }, { "epoch": 0.3859761981344484, "grad_norm": 0.921406090259552, "learning_rate": 4.9035863621743327e-05, "loss": 1.1754, "step": 1200 }, { "epoch": 0.4181408813123191, "grad_norm": 0.9048774242401123, "learning_rate": 4.895545191379865e-05, "loss": 1.1757, "step": 1300 }, { "epoch": 0.45030556449018977, "grad_norm": 0.8832495808601379, "learning_rate": 4.8875040205853975e-05, "loss": 1.156, "step": 1400 }, { "epoch": 0.48247024766806046, "grad_norm": 1.008846640586853, "learning_rate": 4.87946284979093e-05, "loss": 1.1307, "step": 1500 }, { "epoch": 0.5146349308459311, "grad_norm": 0.9244402647018433, "learning_rate": 4.871421678996462e-05, "loss": 1.1405, "step": 1600 }, { "epoch": 0.5467996140238018, "grad_norm": 0.8623345494270325, "learning_rate": 4.8633805082019944e-05, "loss": 1.117, "step": 1700 }, { "epoch": 0.5789642972016725, "grad_norm": 0.8600289821624756, "learning_rate": 4.8553393374075265e-05, "loss": 1.1364, "step": 1800 }, { "epoch": 0.6111289803795432, "grad_norm": 0.8522710800170898, "learning_rate": 4.847298166613059e-05, "loss": 1.1196, "step": 1900 }, { "epoch": 0.6432936635574139, "grad_norm": 0.8213996291160583, "learning_rate": 4.839256995818591e-05, "loss": 1.0908, "step": 2000 }, { "epoch": 0.6754583467352847, "grad_norm": 0.8777344226837158, "learning_rate": 4.8312158250241234e-05, "loss": 1.0863, "step": 2100 }, { "epoch": 0.7076230299131554, "grad_norm": 0.8957362174987793, "learning_rate": 4.823174654229656e-05, "loss": 1.1033, "step": 2200 }, { "epoch": 0.7397877130910261, "grad_norm": 0.8989800810813904, "learning_rate": 4.815133483435189e-05, "loss": 1.1242, "step": 2300 }, { "epoch": 0.7719523962688968, "grad_norm": 0.8211784362792969, "learning_rate": 4.807092312640721e-05, "loss": 1.0763, "step": 2400 }, { "epoch": 0.8041170794467675, "grad_norm": 0.8380767107009888, "learning_rate": 4.799051141846253e-05, "loss": 1.115, "step": 2500 }, { "epoch": 0.8362817626246382, "grad_norm": 0.8601460456848145, "learning_rate": 4.791009971051785e-05, "loss": 1.0721, "step": 2600 }, { "epoch": 0.8684464458025088, "grad_norm": 0.7620469927787781, "learning_rate": 4.782968800257317e-05, "loss": 1.0923, "step": 2700 }, { "epoch": 0.9006111289803795, "grad_norm": 0.7965347170829773, "learning_rate": 4.77492762946285e-05, "loss": 1.0701, "step": 2800 }, { "epoch": 0.9327758121582502, "grad_norm": 0.8240953087806702, "learning_rate": 4.766886458668383e-05, "loss": 1.0559, "step": 2900 }, { "epoch": 0.9649404953361209, "grad_norm": 0.8731254935264587, "learning_rate": 4.758845287873915e-05, "loss": 1.0658, "step": 3000 }, { "epoch": 0.9971051785139916, "grad_norm": 0.8962804675102234, "learning_rate": 4.750804117079447e-05, "loss": 1.0566, "step": 3100 }, { "epoch": 1.0292698616918623, "grad_norm": 0.8816267251968384, "learning_rate": 4.7427629462849796e-05, "loss": 0.9773, "step": 3200 }, { "epoch": 1.061434544869733, "grad_norm": 0.8294162154197693, "learning_rate": 4.7348021871984565e-05, "loss": 0.9802, "step": 3300 }, { "epoch": 1.0935992280476037, "grad_norm": 0.789862334728241, "learning_rate": 4.7267610164039886e-05, "loss": 1.0045, "step": 3400 }, { "epoch": 1.1257639112254745, "grad_norm": 0.9009740948677063, "learning_rate": 4.718719845609521e-05, "loss": 0.9923, "step": 3500 }, { "epoch": 1.157928594403345, "grad_norm": 0.9688478112220764, "learning_rate": 4.7106786748150534e-05, "loss": 0.9924, "step": 3600 }, { "epoch": 1.1900932775812159, "grad_norm": 0.9292545318603516, "learning_rate": 4.7026375040205855e-05, "loss": 0.9685, "step": 3700 }, { "epoch": 1.2222579607590864, "grad_norm": 0.8641929030418396, "learning_rate": 4.6945963332261175e-05, "loss": 0.9894, "step": 3800 }, { "epoch": 1.2544226439369572, "grad_norm": 0.9717555046081543, "learning_rate": 4.68655516243165e-05, "loss": 0.9758, "step": 3900 }, { "epoch": 1.286587327114828, "grad_norm": 0.8688477277755737, "learning_rate": 4.678513991637183e-05, "loss": 0.962, "step": 4000 }, { "epoch": 1.3187520102926986, "grad_norm": 0.9667224287986755, "learning_rate": 4.670472820842715e-05, "loss": 0.9509, "step": 4100 }, { "epoch": 1.3509166934705692, "grad_norm": 0.9459453225135803, "learning_rate": 4.662431650048247e-05, "loss": 0.9901, "step": 4200 }, { "epoch": 1.38308137664844, "grad_norm": 0.7960919141769409, "learning_rate": 4.654390479253779e-05, "loss": 0.9448, "step": 4300 }, { "epoch": 1.4152460598263108, "grad_norm": 0.9302664399147034, "learning_rate": 4.646349308459312e-05, "loss": 0.9773, "step": 4400 }, { "epoch": 1.4474107430041814, "grad_norm": 0.8294656276702881, "learning_rate": 4.638308137664844e-05, "loss": 0.9551, "step": 4500 }, { "epoch": 1.4795754261820522, "grad_norm": 0.8560650944709778, "learning_rate": 4.630266966870376e-05, "loss": 0.979, "step": 4600 }, { "epoch": 1.5117401093599228, "grad_norm": 0.896445631980896, "learning_rate": 4.622225796075909e-05, "loss": 0.9666, "step": 4700 }, { "epoch": 1.5439047925377936, "grad_norm": 0.7956883907318115, "learning_rate": 4.614184625281442e-05, "loss": 0.9691, "step": 4800 }, { "epoch": 1.5760694757156641, "grad_norm": 0.8232746124267578, "learning_rate": 4.606143454486974e-05, "loss": 0.9729, "step": 4900 }, { "epoch": 1.608234158893535, "grad_norm": 0.9488968849182129, "learning_rate": 4.598102283692506e-05, "loss": 0.9719, "step": 5000 }, { "epoch": 1.6403988420714057, "grad_norm": 0.8127655386924744, "learning_rate": 4.590061112898038e-05, "loss": 0.9576, "step": 5100 }, { "epoch": 1.6725635252492763, "grad_norm": 0.8098297119140625, "learning_rate": 4.582019942103571e-05, "loss": 0.9717, "step": 5200 }, { "epoch": 1.704728208427147, "grad_norm": 0.798922598361969, "learning_rate": 4.573978771309103e-05, "loss": 0.946, "step": 5300 }, { "epoch": 1.7368928916050177, "grad_norm": 0.827904224395752, "learning_rate": 4.5659376005146355e-05, "loss": 0.994, "step": 5400 }, { "epoch": 1.7690575747828885, "grad_norm": 0.8054550290107727, "learning_rate": 4.5578964297201676e-05, "loss": 0.9506, "step": 5500 }, { "epoch": 1.801222257960759, "grad_norm": 0.787251889705658, "learning_rate": 4.5498552589256996e-05, "loss": 0.937, "step": 5600 }, { "epoch": 1.8333869411386297, "grad_norm": 0.8783938884735107, "learning_rate": 4.5418140881312324e-05, "loss": 0.9576, "step": 5700 }, { "epoch": 1.8655516243165005, "grad_norm": 0.7947699427604675, "learning_rate": 4.5337729173367645e-05, "loss": 0.9586, "step": 5800 }, { "epoch": 1.8977163074943713, "grad_norm": 0.8789839744567871, "learning_rate": 4.5257317465422966e-05, "loss": 0.9852, "step": 5900 }, { "epoch": 1.9298809906722418, "grad_norm": 0.7798435091972351, "learning_rate": 4.5176905757478286e-05, "loss": 0.9536, "step": 6000 }, { "epoch": 1.9620456738501124, "grad_norm": 0.8084624409675598, "learning_rate": 4.5096494049533614e-05, "loss": 0.9886, "step": 6100 }, { "epoch": 1.9942103570279834, "grad_norm": 0.7713519334793091, "learning_rate": 4.501608234158894e-05, "loss": 0.9737, "step": 6200 }, { "epoch": 2.026375040205854, "grad_norm": 0.9505152106285095, "learning_rate": 4.493567063364426e-05, "loss": 0.8649, "step": 6300 }, { "epoch": 2.0585397233837246, "grad_norm": 1.0130715370178223, "learning_rate": 4.485525892569958e-05, "loss": 0.8337, "step": 6400 }, { "epoch": 2.090704406561595, "grad_norm": 0.8994264006614685, "learning_rate": 4.4774847217754904e-05, "loss": 0.8414, "step": 6500 }, { "epoch": 2.122869089739466, "grad_norm": 0.8061904311180115, "learning_rate": 4.469443550981023e-05, "loss": 0.8375, "step": 6600 }, { "epoch": 2.1550337729173368, "grad_norm": 0.8046594262123108, "learning_rate": 4.461402380186556e-05, "loss": 0.857, "step": 6700 }, { "epoch": 2.1871984560952074, "grad_norm": 0.8733768463134766, "learning_rate": 4.453441621100032e-05, "loss": 0.8693, "step": 6800 }, { "epoch": 2.219363139273078, "grad_norm": 0.8263835906982422, "learning_rate": 4.445400450305565e-05, "loss": 0.8408, "step": 6900 }, { "epoch": 2.251527822450949, "grad_norm": 0.7763977646827698, "learning_rate": 4.437359279511097e-05, "loss": 0.87, "step": 7000 }, { "epoch": 2.2836925056288195, "grad_norm": 0.8745999932289124, "learning_rate": 4.429318108716629e-05, "loss": 0.8538, "step": 7100 }, { "epoch": 2.31585718880669, "grad_norm": 0.8505412936210632, "learning_rate": 4.421276937922162e-05, "loss": 0.8431, "step": 7200 }, { "epoch": 2.348021871984561, "grad_norm": 0.8500067591667175, "learning_rate": 4.4132357671276945e-05, "loss": 0.841, "step": 7300 }, { "epoch": 2.3801865551624317, "grad_norm": 0.9537181854248047, "learning_rate": 4.4051945963332266e-05, "loss": 0.8473, "step": 7400 }, { "epoch": 2.4123512383403023, "grad_norm": 0.889831006526947, "learning_rate": 4.3971534255387586e-05, "loss": 0.8653, "step": 7500 }, { "epoch": 2.444515921518173, "grad_norm": 0.9562848210334778, "learning_rate": 4.389112254744291e-05, "loss": 0.8451, "step": 7600 }, { "epoch": 2.476680604696044, "grad_norm": 0.869602382183075, "learning_rate": 4.3810710839498235e-05, "loss": 0.8589, "step": 7700 }, { "epoch": 2.5088452878739145, "grad_norm": 0.9119632244110107, "learning_rate": 4.3730299131553555e-05, "loss": 0.8468, "step": 7800 }, { "epoch": 2.541009971051785, "grad_norm": 0.8096926212310791, "learning_rate": 4.364988742360888e-05, "loss": 0.8495, "step": 7900 }, { "epoch": 2.573174654229656, "grad_norm": 0.9040408134460449, "learning_rate": 4.3569475715664204e-05, "loss": 0.8627, "step": 8000 }, { "epoch": 2.6053393374075267, "grad_norm": 0.7748663425445557, "learning_rate": 4.3489064007719524e-05, "loss": 0.8717, "step": 8100 }, { "epoch": 2.6375040205853972, "grad_norm": 0.9192063212394714, "learning_rate": 4.340865229977485e-05, "loss": 0.8845, "step": 8200 }, { "epoch": 2.669668703763268, "grad_norm": 0.7993588447570801, "learning_rate": 4.332824059183017e-05, "loss": 0.8686, "step": 8300 }, { "epoch": 2.7018333869411384, "grad_norm": 0.9893975257873535, "learning_rate": 4.3247828883885494e-05, "loss": 0.8655, "step": 8400 }, { "epoch": 2.7339980701190094, "grad_norm": 0.6633499264717102, "learning_rate": 4.3167417175940814e-05, "loss": 0.8764, "step": 8500 }, { "epoch": 2.76616275329688, "grad_norm": 0.9091640114784241, "learning_rate": 4.308700546799614e-05, "loss": 0.8751, "step": 8600 }, { "epoch": 2.7983274364747506, "grad_norm": 0.820894181728363, "learning_rate": 4.300659376005147e-05, "loss": 0.875, "step": 8700 }, { "epoch": 2.8304921196526216, "grad_norm": 0.7544598579406738, "learning_rate": 4.292618205210679e-05, "loss": 0.8731, "step": 8800 }, { "epoch": 2.862656802830492, "grad_norm": 0.8050575256347656, "learning_rate": 4.284577034416211e-05, "loss": 0.8787, "step": 8900 }, { "epoch": 2.8948214860083628, "grad_norm": 0.8130388855934143, "learning_rate": 4.276535863621743e-05, "loss": 0.8566, "step": 9000 }, { "epoch": 2.9269861691862333, "grad_norm": 0.861630380153656, "learning_rate": 4.268494692827276e-05, "loss": 0.8776, "step": 9100 }, { "epoch": 2.9591508523641044, "grad_norm": 0.8607050776481628, "learning_rate": 4.260453522032809e-05, "loss": 0.8568, "step": 9200 }, { "epoch": 2.991315535541975, "grad_norm": 0.8430234789848328, "learning_rate": 4.252412351238341e-05, "loss": 0.8723, "step": 9300 }, { "epoch": 3.0234802187198455, "grad_norm": 0.7705909013748169, "learning_rate": 4.244371180443873e-05, "loss": 0.7587, "step": 9400 }, { "epoch": 3.055644901897716, "grad_norm": 1.0884994268417358, "learning_rate": 4.236330009649405e-05, "loss": 0.7223, "step": 9500 }, { "epoch": 3.087809585075587, "grad_norm": 0.9366634488105774, "learning_rate": 4.2282888388549377e-05, "loss": 0.7462, "step": 9600 }, { "epoch": 3.1199742682534577, "grad_norm": 0.8941793441772461, "learning_rate": 4.22024766806047e-05, "loss": 0.7351, "step": 9700 }, { "epoch": 3.1521389514313283, "grad_norm": 0.8178213834762573, "learning_rate": 4.212206497266002e-05, "loss": 0.7465, "step": 9800 }, { "epoch": 3.1843036346091993, "grad_norm": 1.0509120225906372, "learning_rate": 4.2041653264715346e-05, "loss": 0.7434, "step": 9900 }, { "epoch": 3.21646831778707, "grad_norm": 0.9800424575805664, "learning_rate": 4.1961241556770666e-05, "loss": 0.7469, "step": 10000 }, { "epoch": 3.2486330009649405, "grad_norm": 0.9956560134887695, "learning_rate": 4.1880829848825994e-05, "loss": 0.7559, "step": 10100 }, { "epoch": 3.280797684142811, "grad_norm": 0.877349317073822, "learning_rate": 4.1800418140881315e-05, "loss": 0.7538, "step": 10200 }, { "epoch": 3.312962367320682, "grad_norm": 0.935783326625824, "learning_rate": 4.1720006432936635e-05, "loss": 0.7359, "step": 10300 }, { "epoch": 3.3451270504985526, "grad_norm": 1.0015981197357178, "learning_rate": 4.1639594724991956e-05, "loss": 0.7665, "step": 10400 }, { "epoch": 3.377291733676423, "grad_norm": 0.9721918702125549, "learning_rate": 4.1559183017047284e-05, "loss": 0.7776, "step": 10500 }, { "epoch": 3.409456416854294, "grad_norm": 0.8627845644950867, "learning_rate": 4.147877130910261e-05, "loss": 0.7536, "step": 10600 }, { "epoch": 3.441621100032165, "grad_norm": 0.9311351180076599, "learning_rate": 4.139835960115793e-05, "loss": 0.7614, "step": 10700 }, { "epoch": 3.4737857832100354, "grad_norm": 1.148760437965393, "learning_rate": 4.131794789321325e-05, "loss": 0.7663, "step": 10800 }, { "epoch": 3.505950466387906, "grad_norm": 0.7732231020927429, "learning_rate": 4.123753618526858e-05, "loss": 0.742, "step": 10900 }, { "epoch": 3.538115149565777, "grad_norm": 1.0244348049163818, "learning_rate": 4.11571244773239e-05, "loss": 0.7509, "step": 11000 }, { "epoch": 3.5702798327436476, "grad_norm": 0.8451341390609741, "learning_rate": 4.107671276937922e-05, "loss": 0.755, "step": 11100 }, { "epoch": 3.602444515921518, "grad_norm": 0.9181256890296936, "learning_rate": 4.099630106143454e-05, "loss": 0.7556, "step": 11200 }, { "epoch": 3.6346091990993887, "grad_norm": 0.9541407227516174, "learning_rate": 4.091588935348987e-05, "loss": 0.7595, "step": 11300 }, { "epoch": 3.6667738822772593, "grad_norm": 0.9781204462051392, "learning_rate": 4.08354776455452e-05, "loss": 0.7621, "step": 11400 }, { "epoch": 3.6989385654551303, "grad_norm": 0.9283120632171631, "learning_rate": 4.075506593760052e-05, "loss": 0.7531, "step": 11500 }, { "epoch": 3.731103248633001, "grad_norm": 1.1551364660263062, "learning_rate": 4.067465422965584e-05, "loss": 0.7578, "step": 11600 }, { "epoch": 3.763267931810872, "grad_norm": 0.9374954104423523, "learning_rate": 4.059424252171116e-05, "loss": 0.7624, "step": 11700 }, { "epoch": 3.7954326149887425, "grad_norm": 0.9222885966300964, "learning_rate": 4.051383081376649e-05, "loss": 0.7703, "step": 11800 }, { "epoch": 3.827597298166613, "grad_norm": 0.9635455012321472, "learning_rate": 4.043341910582181e-05, "loss": 0.7714, "step": 11900 }, { "epoch": 3.8597619813444837, "grad_norm": 0.8468224406242371, "learning_rate": 4.0353007397877136e-05, "loss": 0.7624, "step": 12000 }, { "epoch": 3.8919266645223543, "grad_norm": 1.1616623401641846, "learning_rate": 4.0272595689932457e-05, "loss": 0.7553, "step": 12100 }, { "epoch": 3.9240913477002253, "grad_norm": 0.9972304105758667, "learning_rate": 4.019218398198778e-05, "loss": 0.7446, "step": 12200 }, { "epoch": 3.956256030878096, "grad_norm": 0.9060898423194885, "learning_rate": 4.0111772274043105e-05, "loss": 0.7645, "step": 12300 }, { "epoch": 3.9884207140559664, "grad_norm": 0.876514732837677, "learning_rate": 4.0031360566098426e-05, "loss": 0.7724, "step": 12400 }, { "epoch": 4.0205853972338375, "grad_norm": 0.9708064198493958, "learning_rate": 3.9950948858153746e-05, "loss": 0.6719, "step": 12500 }, { "epoch": 4.052750080411708, "grad_norm": 1.1206649541854858, "learning_rate": 3.987053715020907e-05, "loss": 0.6307, "step": 12600 }, { "epoch": 4.084914763589579, "grad_norm": 0.9474230408668518, "learning_rate": 3.9790125442264395e-05, "loss": 0.626, "step": 12700 }, { "epoch": 4.117079446767449, "grad_norm": 0.938599169254303, "learning_rate": 3.970971373431972e-05, "loss": 0.6249, "step": 12800 }, { "epoch": 4.14924412994532, "grad_norm": 1.1760433912277222, "learning_rate": 3.962930202637504e-05, "loss": 0.6387, "step": 12900 }, { "epoch": 4.18140881312319, "grad_norm": 1.1726824045181274, "learning_rate": 3.9548890318430364e-05, "loss": 0.6458, "step": 13000 }, { "epoch": 4.213573496301062, "grad_norm": 0.961887538433075, "learning_rate": 3.9468478610485685e-05, "loss": 0.6361, "step": 13100 }, { "epoch": 4.245738179478932, "grad_norm": 1.0183879137039185, "learning_rate": 3.938806690254101e-05, "loss": 0.6288, "step": 13200 }, { "epoch": 4.277902862656803, "grad_norm": 1.2366642951965332, "learning_rate": 3.930765519459634e-05, "loss": 0.6378, "step": 13300 }, { "epoch": 4.3100675458346736, "grad_norm": 0.9713104963302612, "learning_rate": 3.922724348665166e-05, "loss": 0.6389, "step": 13400 }, { "epoch": 4.342232229012544, "grad_norm": 0.9488447904586792, "learning_rate": 3.914683177870698e-05, "loss": 0.6491, "step": 13500 }, { "epoch": 4.374396912190415, "grad_norm": 1.1355348825454712, "learning_rate": 3.90664200707623e-05, "loss": 0.659, "step": 13600 }, { "epoch": 4.406561595368285, "grad_norm": 1.10310959815979, "learning_rate": 3.898600836281763e-05, "loss": 0.6617, "step": 13700 }, { "epoch": 4.438726278546156, "grad_norm": 1.0456990003585815, "learning_rate": 3.890559665487295e-05, "loss": 0.6541, "step": 13800 }, { "epoch": 4.470890961724027, "grad_norm": 1.2041975259780884, "learning_rate": 3.882518494692827e-05, "loss": 0.6553, "step": 13900 }, { "epoch": 4.503055644901898, "grad_norm": 0.860864520072937, "learning_rate": 3.87447732389836e-05, "loss": 0.6497, "step": 14000 }, { "epoch": 4.5352203280797685, "grad_norm": 1.1725019216537476, "learning_rate": 3.8664361531038926e-05, "loss": 0.6533, "step": 14100 }, { "epoch": 4.567385011257639, "grad_norm": 1.0299824476242065, "learning_rate": 3.858394982309425e-05, "loss": 0.6472, "step": 14200 }, { "epoch": 4.59954969443551, "grad_norm": 1.1528712511062622, "learning_rate": 3.850353811514957e-05, "loss": 0.6508, "step": 14300 }, { "epoch": 4.63171437761338, "grad_norm": 1.106142520904541, "learning_rate": 3.842312640720489e-05, "loss": 0.655, "step": 14400 }, { "epoch": 4.663879060791251, "grad_norm": 1.019434928894043, "learning_rate": 3.8342714699260216e-05, "loss": 0.6419, "step": 14500 }, { "epoch": 4.696043743969122, "grad_norm": 1.1405712366104126, "learning_rate": 3.826230299131554e-05, "loss": 0.6535, "step": 14600 }, { "epoch": 4.728208427146993, "grad_norm": 1.1748415231704712, "learning_rate": 3.8181891283370864e-05, "loss": 0.6609, "step": 14700 }, { "epoch": 4.760373110324863, "grad_norm": 1.0870500802993774, "learning_rate": 3.810228369250563e-05, "loss": 0.6582, "step": 14800 }, { "epoch": 4.792537793502734, "grad_norm": 1.2334747314453125, "learning_rate": 3.8021871984560954e-05, "loss": 0.655, "step": 14900 }, { "epoch": 4.824702476680605, "grad_norm": 1.0857813358306885, "learning_rate": 3.7941460276616274e-05, "loss": 0.6568, "step": 15000 }, { "epoch": 4.856867159858475, "grad_norm": 1.1353741884231567, "learning_rate": 3.7861048568671595e-05, "loss": 0.6652, "step": 15100 }, { "epoch": 4.889031843036346, "grad_norm": 1.1487704515457153, "learning_rate": 3.778063686072692e-05, "loss": 0.6696, "step": 15200 }, { "epoch": 4.921196526214217, "grad_norm": 1.0605279207229614, "learning_rate": 3.770022515278225e-05, "loss": 0.6561, "step": 15300 }, { "epoch": 4.953361209392088, "grad_norm": 1.0910602807998657, "learning_rate": 3.761981344483757e-05, "loss": 0.6794, "step": 15400 }, { "epoch": 4.985525892569958, "grad_norm": 1.147965431213379, "learning_rate": 3.753940173689289e-05, "loss": 0.6717, "step": 15500 }, { "epoch": 5.017690575747829, "grad_norm": 1.1085681915283203, "learning_rate": 3.745899002894821e-05, "loss": 0.5856, "step": 15600 }, { "epoch": 5.0498552589256995, "grad_norm": 1.1322296857833862, "learning_rate": 3.737857832100354e-05, "loss": 0.5282, "step": 15700 }, { "epoch": 5.08201994210357, "grad_norm": 1.0307468175888062, "learning_rate": 3.729816661305887e-05, "loss": 0.538, "step": 15800 }, { "epoch": 5.114184625281441, "grad_norm": 1.3094079494476318, "learning_rate": 3.7218559022193636e-05, "loss": 0.5269, "step": 15900 }, { "epoch": 5.146349308459311, "grad_norm": 0.9686407446861267, "learning_rate": 3.713814731424896e-05, "loss": 0.5398, "step": 16000 }, { "epoch": 5.178513991637183, "grad_norm": 1.1006598472595215, "learning_rate": 3.705773560630428e-05, "loss": 0.5419, "step": 16100 }, { "epoch": 5.210678674815053, "grad_norm": 1.0708500146865845, "learning_rate": 3.69773238983596e-05, "loss": 0.5339, "step": 16200 }, { "epoch": 5.242843357992924, "grad_norm": 0.994500994682312, "learning_rate": 3.6896912190414926e-05, "loss": 0.5446, "step": 16300 }, { "epoch": 5.2750080411707945, "grad_norm": 1.0130538940429688, "learning_rate": 3.6816500482470254e-05, "loss": 0.5457, "step": 16400 }, { "epoch": 5.307172724348665, "grad_norm": 1.1695197820663452, "learning_rate": 3.6736088774525574e-05, "loss": 0.5556, "step": 16500 }, { "epoch": 5.339337407526536, "grad_norm": 1.2486307621002197, "learning_rate": 3.6655677066580895e-05, "loss": 0.5478, "step": 16600 }, { "epoch": 5.371502090704406, "grad_norm": 1.3196483850479126, "learning_rate": 3.6575265358636216e-05, "loss": 0.5479, "step": 16700 }, { "epoch": 5.403666773882278, "grad_norm": 1.1621615886688232, "learning_rate": 3.6494853650691544e-05, "loss": 0.5507, "step": 16800 }, { "epoch": 5.435831457060148, "grad_norm": 1.1101694107055664, "learning_rate": 3.6414441942746864e-05, "loss": 0.5541, "step": 16900 }, { "epoch": 5.467996140238019, "grad_norm": 1.0300599336624146, "learning_rate": 3.633403023480219e-05, "loss": 0.5595, "step": 17000 }, { "epoch": 5.500160823415889, "grad_norm": 1.133117914199829, "learning_rate": 3.625361852685751e-05, "loss": 0.5461, "step": 17100 }, { "epoch": 5.53232550659376, "grad_norm": 1.1690670251846313, "learning_rate": 3.617320681891283e-05, "loss": 0.5648, "step": 17200 }, { "epoch": 5.564490189771631, "grad_norm": 0.9958963990211487, "learning_rate": 3.609279511096816e-05, "loss": 0.5679, "step": 17300 }, { "epoch": 5.596654872949501, "grad_norm": 1.1210240125656128, "learning_rate": 3.601238340302348e-05, "loss": 0.5478, "step": 17400 }, { "epoch": 5.628819556127372, "grad_norm": 1.1824806928634644, "learning_rate": 3.59319716950788e-05, "loss": 0.5611, "step": 17500 }, { "epoch": 5.660984239305243, "grad_norm": 1.201159119606018, "learning_rate": 3.585155998713413e-05, "loss": 0.5513, "step": 17600 }, { "epoch": 5.693148922483114, "grad_norm": 1.264073371887207, "learning_rate": 3.577114827918945e-05, "loss": 0.5693, "step": 17700 }, { "epoch": 5.725313605660984, "grad_norm": 1.2534087896347046, "learning_rate": 3.569073657124478e-05, "loss": 0.5673, "step": 17800 }, { "epoch": 5.757478288838855, "grad_norm": 1.2071319818496704, "learning_rate": 3.56103248633001e-05, "loss": 0.5592, "step": 17900 }, { "epoch": 5.7896429720167255, "grad_norm": 1.3951480388641357, "learning_rate": 3.552991315535542e-05, "loss": 0.5576, "step": 18000 }, { "epoch": 5.821807655194596, "grad_norm": 1.1245003938674927, "learning_rate": 3.544950144741074e-05, "loss": 0.5608, "step": 18100 }, { "epoch": 5.853972338372467, "grad_norm": 1.0020116567611694, "learning_rate": 3.536908973946607e-05, "loss": 0.5676, "step": 18200 }, { "epoch": 5.886137021550338, "grad_norm": 1.2194671630859375, "learning_rate": 3.5288678031521396e-05, "loss": 0.5619, "step": 18300 }, { "epoch": 5.918301704728209, "grad_norm": 1.162885308265686, "learning_rate": 3.5208266323576716e-05, "loss": 0.5673, "step": 18400 }, { "epoch": 5.950466387906079, "grad_norm": 1.1328020095825195, "learning_rate": 3.5128658732711485e-05, "loss": 0.5744, "step": 18500 }, { "epoch": 5.98263107108395, "grad_norm": 1.0173712968826294, "learning_rate": 3.5048247024766806e-05, "loss": 0.5727, "step": 18600 }, { "epoch": 6.0147957542618204, "grad_norm": 1.096663475036621, "learning_rate": 3.4967835316822127e-05, "loss": 0.5169, "step": 18700 }, { "epoch": 6.046960437439691, "grad_norm": 1.217439889907837, "learning_rate": 3.4887423608877454e-05, "loss": 0.4442, "step": 18800 }, { "epoch": 6.079125120617562, "grad_norm": 1.2011640071868896, "learning_rate": 3.480701190093278e-05, "loss": 0.4432, "step": 18900 }, { "epoch": 6.111289803795432, "grad_norm": 1.1648433208465576, "learning_rate": 3.47266001929881e-05, "loss": 0.4448, "step": 19000 }, { "epoch": 6.143454486973304, "grad_norm": 1.0002985000610352, "learning_rate": 3.464618848504342e-05, "loss": 0.4422, "step": 19100 }, { "epoch": 6.175619170151174, "grad_norm": 1.227947473526001, "learning_rate": 3.4565776777098744e-05, "loss": 0.4448, "step": 19200 }, { "epoch": 6.207783853329045, "grad_norm": 1.3668509721755981, "learning_rate": 3.448536506915407e-05, "loss": 0.458, "step": 19300 }, { "epoch": 6.239948536506915, "grad_norm": 1.1402783393859863, "learning_rate": 3.440495336120939e-05, "loss": 0.4489, "step": 19400 }, { "epoch": 6.272113219684786, "grad_norm": 1.2867486476898193, "learning_rate": 3.432454165326472e-05, "loss": 0.4611, "step": 19500 }, { "epoch": 6.3042779028626565, "grad_norm": 1.3049463033676147, "learning_rate": 3.424412994532004e-05, "loss": 0.4554, "step": 19600 }, { "epoch": 6.336442586040527, "grad_norm": 1.2762079238891602, "learning_rate": 3.416371823737536e-05, "loss": 0.4651, "step": 19700 }, { "epoch": 6.368607269218399, "grad_norm": 1.0911743640899658, "learning_rate": 3.408330652943069e-05, "loss": 0.46, "step": 19800 }, { "epoch": 6.400771952396269, "grad_norm": 1.2043310403823853, "learning_rate": 3.400289482148601e-05, "loss": 0.4689, "step": 19900 }, { "epoch": 6.43293663557414, "grad_norm": 1.1901593208312988, "learning_rate": 3.392248311354133e-05, "loss": 0.4689, "step": 20000 }, { "epoch": 6.46510131875201, "grad_norm": 1.3777788877487183, "learning_rate": 3.384207140559666e-05, "loss": 0.4646, "step": 20100 }, { "epoch": 6.497266001929881, "grad_norm": 1.2720263004302979, "learning_rate": 3.376165969765198e-05, "loss": 0.4684, "step": 20200 }, { "epoch": 6.5294306851077515, "grad_norm": 1.0634431838989258, "learning_rate": 3.3681247989707306e-05, "loss": 0.4699, "step": 20300 }, { "epoch": 6.561595368285622, "grad_norm": 1.3379625082015991, "learning_rate": 3.360083628176263e-05, "loss": 0.4807, "step": 20400 }, { "epoch": 6.593760051463493, "grad_norm": 1.379611849784851, "learning_rate": 3.352042457381795e-05, "loss": 0.4743, "step": 20500 }, { "epoch": 6.625924734641364, "grad_norm": 1.2262407541275024, "learning_rate": 3.344001286587327e-05, "loss": 0.4699, "step": 20600 }, { "epoch": 6.658089417819235, "grad_norm": 1.2290821075439453, "learning_rate": 3.3360405275008044e-05, "loss": 0.4728, "step": 20700 }, { "epoch": 6.690254100997105, "grad_norm": 1.286399006843567, "learning_rate": 3.3279993567063365e-05, "loss": 0.4811, "step": 20800 }, { "epoch": 6.722418784174976, "grad_norm": 1.3331542015075684, "learning_rate": 3.319958185911869e-05, "loss": 0.4795, "step": 20900 }, { "epoch": 6.754583467352846, "grad_norm": 0.9238734841346741, "learning_rate": 3.311917015117401e-05, "loss": 0.4719, "step": 21000 }, { "epoch": 6.786748150530717, "grad_norm": 1.0958573818206787, "learning_rate": 3.3038758443229334e-05, "loss": 0.4787, "step": 21100 }, { "epoch": 6.818912833708588, "grad_norm": 1.3021626472473145, "learning_rate": 3.2958346735284655e-05, "loss": 0.481, "step": 21200 }, { "epoch": 6.851077516886459, "grad_norm": 1.2894419431686401, "learning_rate": 3.287793502733998e-05, "loss": 0.4799, "step": 21300 }, { "epoch": 6.88324220006433, "grad_norm": 1.2276169061660767, "learning_rate": 3.279752331939531e-05, "loss": 0.4791, "step": 21400 }, { "epoch": 6.9154068832422, "grad_norm": 1.2167209386825562, "learning_rate": 3.271711161145063e-05, "loss": 0.4777, "step": 21500 }, { "epoch": 6.947571566420071, "grad_norm": 1.0899590253829956, "learning_rate": 3.263669990350595e-05, "loss": 0.4807, "step": 21600 }, { "epoch": 6.979736249597941, "grad_norm": 1.0014961957931519, "learning_rate": 3.255628819556127e-05, "loss": 0.4809, "step": 21700 }, { "epoch": 7.011900932775812, "grad_norm": 1.2910642623901367, "learning_rate": 3.24758764876166e-05, "loss": 0.4473, "step": 21800 }, { "epoch": 7.0440656159536825, "grad_norm": 1.2702795267105103, "learning_rate": 3.239546477967192e-05, "loss": 0.3631, "step": 21900 }, { "epoch": 7.076230299131554, "grad_norm": 1.1640880107879639, "learning_rate": 3.231505307172725e-05, "loss": 0.3648, "step": 22000 }, { "epoch": 7.108394982309425, "grad_norm": 1.3022829294204712, "learning_rate": 3.223464136378257e-05, "loss": 0.372, "step": 22100 }, { "epoch": 7.140559665487295, "grad_norm": 1.1453783512115479, "learning_rate": 3.215422965583789e-05, "loss": 0.3713, "step": 22200 }, { "epoch": 7.172724348665166, "grad_norm": 1.2497502565383911, "learning_rate": 3.207381794789322e-05, "loss": 0.3744, "step": 22300 }, { "epoch": 7.204889031843036, "grad_norm": 1.0671898126602173, "learning_rate": 3.199340623994854e-05, "loss": 0.3818, "step": 22400 }, { "epoch": 7.237053715020907, "grad_norm": 1.3146332502365112, "learning_rate": 3.191299453200386e-05, "loss": 0.3818, "step": 22500 }, { "epoch": 7.2692183981987775, "grad_norm": 1.155218482017517, "learning_rate": 3.1832582824059186e-05, "loss": 0.3849, "step": 22600 }, { "epoch": 7.301383081376648, "grad_norm": 1.493849515914917, "learning_rate": 3.175217111611451e-05, "loss": 0.3914, "step": 22700 }, { "epoch": 7.3335477645545195, "grad_norm": 1.4199835062026978, "learning_rate": 3.1671759408169834e-05, "loss": 0.3877, "step": 22800 }, { "epoch": 7.36571244773239, "grad_norm": 1.3320192098617554, "learning_rate": 3.1591347700225155e-05, "loss": 0.3898, "step": 22900 }, { "epoch": 7.397877130910261, "grad_norm": 1.2950165271759033, "learning_rate": 3.1510935992280476e-05, "loss": 0.3915, "step": 23000 }, { "epoch": 7.430041814088131, "grad_norm": 1.2923767566680908, "learning_rate": 3.1430524284335796e-05, "loss": 0.3866, "step": 23100 }, { "epoch": 7.462206497266002, "grad_norm": 1.2137949466705322, "learning_rate": 3.1350112576391124e-05, "loss": 0.3848, "step": 23200 }, { "epoch": 7.494371180443872, "grad_norm": 1.4630000591278076, "learning_rate": 3.126970086844645e-05, "loss": 0.3965, "step": 23300 }, { "epoch": 7.526535863621743, "grad_norm": 1.2107799053192139, "learning_rate": 3.118928916050177e-05, "loss": 0.3922, "step": 23400 }, { "epoch": 7.558700546799614, "grad_norm": 1.1131380796432495, "learning_rate": 3.110887745255709e-05, "loss": 0.3892, "step": 23500 }, { "epoch": 7.590865229977485, "grad_norm": 1.2040220499038696, "learning_rate": 3.1028465744612414e-05, "loss": 0.3931, "step": 23600 }, { "epoch": 7.623029913155356, "grad_norm": 1.1216152906417847, "learning_rate": 3.094805403666774e-05, "loss": 0.3901, "step": 23700 }, { "epoch": 7.655194596333226, "grad_norm": 1.2603133916854858, "learning_rate": 3.086764232872306e-05, "loss": 0.3976, "step": 23800 }, { "epoch": 7.687359279511097, "grad_norm": 1.2045493125915527, "learning_rate": 3.078723062077838e-05, "loss": 0.3983, "step": 23900 }, { "epoch": 7.719523962688967, "grad_norm": 1.3943126201629639, "learning_rate": 3.070681891283371e-05, "loss": 0.3905, "step": 24000 }, { "epoch": 7.751688645866838, "grad_norm": 1.204542875289917, "learning_rate": 3.062640720488904e-05, "loss": 0.3996, "step": 24100 }, { "epoch": 7.7838533290447085, "grad_norm": 1.377310037612915, "learning_rate": 3.054599549694436e-05, "loss": 0.3996, "step": 24200 }, { "epoch": 7.81601801222258, "grad_norm": 1.1735265254974365, "learning_rate": 3.046558378899968e-05, "loss": 0.4043, "step": 24300 }, { "epoch": 7.8481826954004505, "grad_norm": 1.2886712551116943, "learning_rate": 3.0385172081055004e-05, "loss": 0.4006, "step": 24400 }, { "epoch": 7.880347378578321, "grad_norm": 1.1405675411224365, "learning_rate": 3.0304760373110324e-05, "loss": 0.4031, "step": 24500 }, { "epoch": 7.912512061756192, "grad_norm": 1.6100174188613892, "learning_rate": 3.022434866516565e-05, "loss": 0.3999, "step": 24600 }, { "epoch": 7.944676744934062, "grad_norm": 1.123673677444458, "learning_rate": 3.0143936957220976e-05, "loss": 0.4096, "step": 24700 }, { "epoch": 7.976841428111933, "grad_norm": 1.3668479919433594, "learning_rate": 3.006432936635574e-05, "loss": 0.4084, "step": 24800 }, { "epoch": 8.009006111289803, "grad_norm": 1.191008448600769, "learning_rate": 2.9983917658411066e-05, "loss": 0.3732, "step": 24900 }, { "epoch": 8.041170794467675, "grad_norm": 1.3444154262542725, "learning_rate": 2.9903505950466386e-05, "loss": 0.3036, "step": 25000 }, { "epoch": 8.073335477645545, "grad_norm": 1.2217003107070923, "learning_rate": 2.9823094242521714e-05, "loss": 0.3044, "step": 25100 }, { "epoch": 8.105500160823416, "grad_norm": 1.3392952680587769, "learning_rate": 2.9742682534577038e-05, "loss": 0.3051, "step": 25200 }, { "epoch": 8.137664844001286, "grad_norm": 1.4353630542755127, "learning_rate": 2.966227082663236e-05, "loss": 0.3116, "step": 25300 }, { "epoch": 8.169829527179157, "grad_norm": 1.2575156688690186, "learning_rate": 2.9581859118687683e-05, "loss": 0.3154, "step": 25400 }, { "epoch": 8.201994210357029, "grad_norm": 1.276118278503418, "learning_rate": 2.9501447410743004e-05, "loss": 0.3137, "step": 25500 }, { "epoch": 8.234158893534898, "grad_norm": 1.2896833419799805, "learning_rate": 2.9421035702798328e-05, "loss": 0.3132, "step": 25600 }, { "epoch": 8.26632357671277, "grad_norm": 1.390541672706604, "learning_rate": 2.934062399485365e-05, "loss": 0.3174, "step": 25700 }, { "epoch": 8.29848825989064, "grad_norm": 1.0846691131591797, "learning_rate": 2.926021228690898e-05, "loss": 0.3258, "step": 25800 }, { "epoch": 8.330652943068511, "grad_norm": 1.1695352792739868, "learning_rate": 2.91798005789643e-05, "loss": 0.3188, "step": 25900 }, { "epoch": 8.36281762624638, "grad_norm": 1.36068856716156, "learning_rate": 2.909938887101962e-05, "loss": 0.3233, "step": 26000 }, { "epoch": 8.394982309424252, "grad_norm": 1.087651252746582, "learning_rate": 2.9018977163074945e-05, "loss": 0.3195, "step": 26100 }, { "epoch": 8.427146992602124, "grad_norm": 1.1580371856689453, "learning_rate": 2.8938565455130266e-05, "loss": 0.3187, "step": 26200 }, { "epoch": 8.459311675779993, "grad_norm": 1.2364060878753662, "learning_rate": 2.885815374718559e-05, "loss": 0.3225, "step": 26300 }, { "epoch": 8.491476358957865, "grad_norm": 1.0226277112960815, "learning_rate": 2.877774203924091e-05, "loss": 0.3213, "step": 26400 }, { "epoch": 8.523641042135734, "grad_norm": 1.1340548992156982, "learning_rate": 2.8697330331296242e-05, "loss": 0.3137, "step": 26500 }, { "epoch": 8.555805725313606, "grad_norm": 1.1343051195144653, "learning_rate": 2.8616918623351563e-05, "loss": 0.3219, "step": 26600 }, { "epoch": 8.587970408491476, "grad_norm": 1.0772162675857544, "learning_rate": 2.8536506915406887e-05, "loss": 0.329, "step": 26700 }, { "epoch": 8.620135091669347, "grad_norm": 1.3438891172409058, "learning_rate": 2.8456899324541652e-05, "loss": 0.3299, "step": 26800 }, { "epoch": 8.652299774847219, "grad_norm": 1.4512920379638672, "learning_rate": 2.837648761659698e-05, "loss": 0.3348, "step": 26900 }, { "epoch": 8.684464458025088, "grad_norm": 1.2788150310516357, "learning_rate": 2.8296075908652304e-05, "loss": 0.3338, "step": 27000 }, { "epoch": 8.71662914120296, "grad_norm": 1.142204999923706, "learning_rate": 2.8215664200707625e-05, "loss": 0.3309, "step": 27100 }, { "epoch": 8.74879382438083, "grad_norm": 1.4097461700439453, "learning_rate": 2.813525249276295e-05, "loss": 0.3326, "step": 27200 }, { "epoch": 8.7809585075587, "grad_norm": 1.1304296255111694, "learning_rate": 2.805484078481827e-05, "loss": 0.333, "step": 27300 }, { "epoch": 8.81312319073657, "grad_norm": 1.2657917737960815, "learning_rate": 2.7974429076873594e-05, "loss": 0.3383, "step": 27400 }, { "epoch": 8.845287873914442, "grad_norm": 1.2562810182571411, "learning_rate": 2.7894017368928914e-05, "loss": 0.3331, "step": 27500 }, { "epoch": 8.877452557092312, "grad_norm": 1.2090494632720947, "learning_rate": 2.7813605660984242e-05, "loss": 0.3335, "step": 27600 }, { "epoch": 8.909617240270183, "grad_norm": 1.055282711982727, "learning_rate": 2.7733193953039566e-05, "loss": 0.3345, "step": 27700 }, { "epoch": 8.941781923448055, "grad_norm": 1.3975212574005127, "learning_rate": 2.7652782245094887e-05, "loss": 0.3374, "step": 27800 }, { "epoch": 8.973946606625924, "grad_norm": 1.4486628770828247, "learning_rate": 2.757237053715021e-05, "loss": 0.3353, "step": 27900 }, { "epoch": 9.006111289803796, "grad_norm": 1.0647599697113037, "learning_rate": 2.7491958829205532e-05, "loss": 0.3258, "step": 28000 }, { "epoch": 9.038275972981666, "grad_norm": 1.1231446266174316, "learning_rate": 2.7411547121260856e-05, "loss": 0.2498, "step": 28100 }, { "epoch": 9.070440656159537, "grad_norm": 1.1911687850952148, "learning_rate": 2.7331135413316177e-05, "loss": 0.2507, "step": 28200 }, { "epoch": 9.102605339337407, "grad_norm": 1.1804819107055664, "learning_rate": 2.7250723705371508e-05, "loss": 0.2538, "step": 28300 }, { "epoch": 9.134770022515278, "grad_norm": 1.2449610233306885, "learning_rate": 2.7170311997426828e-05, "loss": 0.254, "step": 28400 }, { "epoch": 9.16693470569315, "grad_norm": 1.3183308839797974, "learning_rate": 2.708990028948215e-05, "loss": 0.2569, "step": 28500 }, { "epoch": 9.19909938887102, "grad_norm": 1.2077155113220215, "learning_rate": 2.7009488581537473e-05, "loss": 0.264, "step": 28600 }, { "epoch": 9.23126407204889, "grad_norm": 1.2201523780822754, "learning_rate": 2.6929076873592794e-05, "loss": 0.2618, "step": 28700 }, { "epoch": 9.26342875522676, "grad_norm": 1.010611653327942, "learning_rate": 2.6848665165648118e-05, "loss": 0.2586, "step": 28800 }, { "epoch": 9.295593438404632, "grad_norm": 1.4196604490280151, "learning_rate": 2.676825345770344e-05, "loss": 0.2645, "step": 28900 }, { "epoch": 9.327758121582502, "grad_norm": 1.0224769115447998, "learning_rate": 2.668784174975877e-05, "loss": 0.2677, "step": 29000 }, { "epoch": 9.359922804760373, "grad_norm": 1.0215541124343872, "learning_rate": 2.6608234158893535e-05, "loss": 0.2662, "step": 29100 }, { "epoch": 9.392087487938245, "grad_norm": 1.1498533487319946, "learning_rate": 2.652782245094886e-05, "loss": 0.2647, "step": 29200 }, { "epoch": 9.424252171116114, "grad_norm": 1.1814836263656616, "learning_rate": 2.644741074300418e-05, "loss": 0.2681, "step": 29300 }, { "epoch": 9.456416854293986, "grad_norm": 1.3575364351272583, "learning_rate": 2.6366999035059508e-05, "loss": 0.2659, "step": 29400 }, { "epoch": 9.488581537471855, "grad_norm": 1.4364129304885864, "learning_rate": 2.6286587327114832e-05, "loss": 0.2701, "step": 29500 }, { "epoch": 9.520746220649727, "grad_norm": 1.4229729175567627, "learning_rate": 2.6206175619170152e-05, "loss": 0.2744, "step": 29600 }, { "epoch": 9.552910903827597, "grad_norm": 1.2050132751464844, "learning_rate": 2.6125763911225477e-05, "loss": 0.2678, "step": 29700 }, { "epoch": 9.585075587005468, "grad_norm": 1.360245704650879, "learning_rate": 2.6045352203280797e-05, "loss": 0.2668, "step": 29800 }, { "epoch": 9.61724027018334, "grad_norm": 1.2902683019638062, "learning_rate": 2.596494049533612e-05, "loss": 0.2749, "step": 29900 }, { "epoch": 9.64940495336121, "grad_norm": 1.3184953927993774, "learning_rate": 2.5884528787391442e-05, "loss": 0.2729, "step": 30000 }, { "epoch": 9.68156963653908, "grad_norm": 1.1212196350097656, "learning_rate": 2.580411707944677e-05, "loss": 0.2711, "step": 30100 }, { "epoch": 9.71373431971695, "grad_norm": 1.1325451135635376, "learning_rate": 2.5723705371502094e-05, "loss": 0.2739, "step": 30200 }, { "epoch": 9.745899002894822, "grad_norm": 1.2344636917114258, "learning_rate": 2.5643293663557415e-05, "loss": 0.276, "step": 30300 }, { "epoch": 9.778063686072691, "grad_norm": 1.3600369691848755, "learning_rate": 2.556288195561274e-05, "loss": 0.2788, "step": 30400 }, { "epoch": 9.810228369250563, "grad_norm": 1.3456294536590576, "learning_rate": 2.548247024766806e-05, "loss": 0.2768, "step": 30500 }, { "epoch": 9.842393052428434, "grad_norm": 1.316645622253418, "learning_rate": 2.5402058539723384e-05, "loss": 0.2756, "step": 30600 }, { "epoch": 9.874557735606304, "grad_norm": 1.4233603477478027, "learning_rate": 2.5321646831778705e-05, "loss": 0.2767, "step": 30700 }, { "epoch": 9.906722418784176, "grad_norm": 1.3680518865585327, "learning_rate": 2.5241235123834036e-05, "loss": 0.2746, "step": 30800 }, { "epoch": 9.938887101962045, "grad_norm": 1.4857224225997925, "learning_rate": 2.5160823415889356e-05, "loss": 0.2775, "step": 30900 }, { "epoch": 9.971051785139917, "grad_norm": 1.3762352466583252, "learning_rate": 2.508041170794468e-05, "loss": 0.2806, "step": 31000 }, { "epoch": 10.003216468317786, "grad_norm": 0.9549392461776733, "learning_rate": 2.5e-05, "loss": 0.2723, "step": 31100 }, { "epoch": 10.035381151495658, "grad_norm": 1.2942768335342407, "learning_rate": 2.4919588292055322e-05, "loss": 0.2038, "step": 31200 }, { "epoch": 10.067545834673528, "grad_norm": 1.1192700862884521, "learning_rate": 2.483917658411065e-05, "loss": 0.208, "step": 31300 }, { "epoch": 10.099710517851399, "grad_norm": 1.0974397659301758, "learning_rate": 2.475876487616597e-05, "loss": 0.2106, "step": 31400 }, { "epoch": 10.13187520102927, "grad_norm": 1.3536230325698853, "learning_rate": 2.4678353168221294e-05, "loss": 0.2099, "step": 31500 }, { "epoch": 10.16403988420714, "grad_norm": 1.1772140264511108, "learning_rate": 2.4597941460276615e-05, "loss": 0.2205, "step": 31600 }, { "epoch": 10.196204567385012, "grad_norm": 1.053501844406128, "learning_rate": 2.4517529752331943e-05, "loss": 0.2118, "step": 31700 }, { "epoch": 10.228369250562881, "grad_norm": 1.1771047115325928, "learning_rate": 2.4437118044387263e-05, "loss": 0.2169, "step": 31800 }, { "epoch": 10.260533933740753, "grad_norm": 1.249473214149475, "learning_rate": 2.4356706336442588e-05, "loss": 0.2184, "step": 31900 }, { "epoch": 10.292698616918623, "grad_norm": 1.2319327592849731, "learning_rate": 2.4276294628497912e-05, "loss": 0.2186, "step": 32000 }, { "epoch": 10.324863300096494, "grad_norm": 1.2124056816101074, "learning_rate": 2.419668703763268e-05, "loss": 0.2165, "step": 32100 }, { "epoch": 10.357027983274365, "grad_norm": 1.0940320491790771, "learning_rate": 2.4116275329688005e-05, "loss": 0.2196, "step": 32200 }, { "epoch": 10.389192666452235, "grad_norm": 1.3821940422058105, "learning_rate": 2.4035863621743325e-05, "loss": 0.2201, "step": 32300 }, { "epoch": 10.421357349630107, "grad_norm": 1.4705811738967896, "learning_rate": 2.3955451913798653e-05, "loss": 0.2237, "step": 32400 }, { "epoch": 10.453522032807976, "grad_norm": 0.9724491834640503, "learning_rate": 2.3875040205853974e-05, "loss": 0.2202, "step": 32500 }, { "epoch": 10.485686715985848, "grad_norm": 1.2271063327789307, "learning_rate": 2.3794628497909298e-05, "loss": 0.2248, "step": 32600 }, { "epoch": 10.517851399163717, "grad_norm": 1.049058437347412, "learning_rate": 2.371421678996462e-05, "loss": 0.2261, "step": 32700 }, { "epoch": 10.550016082341589, "grad_norm": 1.324881911277771, "learning_rate": 2.3633805082019943e-05, "loss": 0.2241, "step": 32800 }, { "epoch": 10.58218076551946, "grad_norm": 1.2529538869857788, "learning_rate": 2.3553393374075267e-05, "loss": 0.2206, "step": 32900 }, { "epoch": 10.61434544869733, "grad_norm": 1.1270591020584106, "learning_rate": 2.3472981666130588e-05, "loss": 0.2264, "step": 33000 }, { "epoch": 10.646510131875202, "grad_norm": 1.256332278251648, "learning_rate": 2.3392569958185915e-05, "loss": 0.2227, "step": 33100 }, { "epoch": 10.678674815053071, "grad_norm": 1.2265748977661133, "learning_rate": 2.3312158250241236e-05, "loss": 0.227, "step": 33200 }, { "epoch": 10.710839498230943, "grad_norm": 1.367353081703186, "learning_rate": 2.323174654229656e-05, "loss": 0.2312, "step": 33300 }, { "epoch": 10.743004181408812, "grad_norm": 1.185103416442871, "learning_rate": 2.315133483435188e-05, "loss": 0.2298, "step": 33400 }, { "epoch": 10.775168864586684, "grad_norm": 1.3100920915603638, "learning_rate": 2.307092312640721e-05, "loss": 0.2265, "step": 33500 }, { "epoch": 10.807333547764555, "grad_norm": 1.1452741622924805, "learning_rate": 2.299051141846253e-05, "loss": 0.2296, "step": 33600 }, { "epoch": 10.839498230942425, "grad_norm": 1.3622289896011353, "learning_rate": 2.2910099710517853e-05, "loss": 0.2318, "step": 33700 }, { "epoch": 10.871662914120297, "grad_norm": 1.2668938636779785, "learning_rate": 2.2829688002573177e-05, "loss": 0.227, "step": 33800 }, { "epoch": 10.903827597298166, "grad_norm": 1.139683723449707, "learning_rate": 2.2749276294628498e-05, "loss": 0.23, "step": 33900 }, { "epoch": 10.935992280476038, "grad_norm": 1.293151617050171, "learning_rate": 2.2668864586683822e-05, "loss": 0.2325, "step": 34000 }, { "epoch": 10.968156963653907, "grad_norm": 1.6205676794052124, "learning_rate": 2.2588452878739143e-05, "loss": 0.2348, "step": 34100 }, { "epoch": 11.000321646831779, "grad_norm": 0.9748951196670532, "learning_rate": 2.2508845287873915e-05, "loss": 0.2312, "step": 34200 }, { "epoch": 11.03248633000965, "grad_norm": 1.0078290700912476, "learning_rate": 2.242843357992924e-05, "loss": 0.1733, "step": 34300 }, { "epoch": 11.06465101318752, "grad_norm": 1.0165376663208008, "learning_rate": 2.2348021871984564e-05, "loss": 0.1744, "step": 34400 }, { "epoch": 11.096815696365391, "grad_norm": 1.2080796957015991, "learning_rate": 2.2267610164039884e-05, "loss": 0.1752, "step": 34500 }, { "epoch": 11.128980379543261, "grad_norm": 1.0622116327285767, "learning_rate": 2.218719845609521e-05, "loss": 0.1746, "step": 34600 }, { "epoch": 11.161145062721133, "grad_norm": 1.2261426448822021, "learning_rate": 2.2106786748150533e-05, "loss": 0.1792, "step": 34700 }, { "epoch": 11.193309745899002, "grad_norm": 1.2450960874557495, "learning_rate": 2.2026375040205853e-05, "loss": 0.1769, "step": 34800 }, { "epoch": 11.225474429076874, "grad_norm": 1.2167742252349854, "learning_rate": 2.194596333226118e-05, "loss": 0.1807, "step": 34900 }, { "epoch": 11.257639112254743, "grad_norm": 1.187880277633667, "learning_rate": 2.18655516243165e-05, "loss": 0.1799, "step": 35000 }, { "epoch": 11.289803795432615, "grad_norm": 1.1898218393325806, "learning_rate": 2.1785139916371826e-05, "loss": 0.1778, "step": 35100 }, { "epoch": 11.321968478610486, "grad_norm": 1.105224609375, "learning_rate": 2.1705532325506595e-05, "loss": 0.1821, "step": 35200 }, { "epoch": 11.354133161788356, "grad_norm": 1.1716156005859375, "learning_rate": 2.1625120617561915e-05, "loss": 0.1817, "step": 35300 }, { "epoch": 11.386297844966228, "grad_norm": 0.952513575553894, "learning_rate": 2.1544708909617243e-05, "loss": 0.1848, "step": 35400 }, { "epoch": 11.418462528144097, "grad_norm": 1.2010760307312012, "learning_rate": 2.1464297201672564e-05, "loss": 0.1813, "step": 35500 }, { "epoch": 11.450627211321969, "grad_norm": 1.2307448387145996, "learning_rate": 2.1383885493727888e-05, "loss": 0.1842, "step": 35600 }, { "epoch": 11.482791894499838, "grad_norm": 1.3126641511917114, "learning_rate": 2.1303473785783212e-05, "loss": 0.1849, "step": 35700 }, { "epoch": 11.51495657767771, "grad_norm": 1.1805635690689087, "learning_rate": 2.1223062077838536e-05, "loss": 0.1862, "step": 35800 }, { "epoch": 11.547121260855581, "grad_norm": 1.1975961923599243, "learning_rate": 2.1142650369893857e-05, "loss": 0.1897, "step": 35900 }, { "epoch": 11.579285944033451, "grad_norm": 1.1553934812545776, "learning_rate": 2.106223866194918e-05, "loss": 0.1842, "step": 36000 }, { "epoch": 11.611450627211322, "grad_norm": 1.3459019660949707, "learning_rate": 2.0981826954004505e-05, "loss": 0.1879, "step": 36100 }, { "epoch": 11.643615310389192, "grad_norm": 1.1215953826904297, "learning_rate": 2.0901415246059826e-05, "loss": 0.1882, "step": 36200 }, { "epoch": 11.675779993567064, "grad_norm": 1.3726115226745605, "learning_rate": 2.082100353811515e-05, "loss": 0.1888, "step": 36300 }, { "epoch": 11.707944676744933, "grad_norm": 1.240684151649475, "learning_rate": 2.0740591830170474e-05, "loss": 0.1866, "step": 36400 }, { "epoch": 11.740109359922805, "grad_norm": 1.291636347770691, "learning_rate": 2.06601801222258e-05, "loss": 0.1888, "step": 36500 }, { "epoch": 11.772274043100676, "grad_norm": 1.172780156135559, "learning_rate": 2.057976841428112e-05, "loss": 0.1934, "step": 36600 }, { "epoch": 11.804438726278546, "grad_norm": 1.1808074712753296, "learning_rate": 2.0499356706336443e-05, "loss": 0.1912, "step": 36700 }, { "epoch": 11.836603409456417, "grad_norm": 1.2383092641830444, "learning_rate": 2.0418944998391767e-05, "loss": 0.1911, "step": 36800 }, { "epoch": 11.868768092634287, "grad_norm": 1.0541281700134277, "learning_rate": 2.033853329044709e-05, "loss": 0.1948, "step": 36900 }, { "epoch": 11.900932775812159, "grad_norm": 1.151892066001892, "learning_rate": 2.0258121582502412e-05, "loss": 0.1905, "step": 37000 }, { "epoch": 11.933097458990028, "grad_norm": 1.2144666910171509, "learning_rate": 2.0177709874557736e-05, "loss": 0.1923, "step": 37100 }, { "epoch": 11.9652621421679, "grad_norm": 1.1869224309921265, "learning_rate": 2.009729816661306e-05, "loss": 0.1878, "step": 37200 }, { "epoch": 11.99742682534577, "grad_norm": 1.4024932384490967, "learning_rate": 2.001688645866838e-05, "loss": 0.1954, "step": 37300 }, { "epoch": 12.029591508523641, "grad_norm": 1.0352271795272827, "learning_rate": 1.993647475072371e-05, "loss": 0.1495, "step": 37400 }, { "epoch": 12.061756191701512, "grad_norm": 1.0901597738265991, "learning_rate": 1.985606304277903e-05, "loss": 0.1474, "step": 37500 }, { "epoch": 12.093920874879382, "grad_norm": 0.9055734872817993, "learning_rate": 1.9775651334834354e-05, "loss": 0.1494, "step": 37600 }, { "epoch": 12.126085558057254, "grad_norm": 1.3142735958099365, "learning_rate": 1.9695239626889675e-05, "loss": 0.1494, "step": 37700 }, { "epoch": 12.158250241235123, "grad_norm": 1.0741150379180908, "learning_rate": 1.9614827918945e-05, "loss": 0.1496, "step": 37800 }, { "epoch": 12.190414924412995, "grad_norm": 1.259490728378296, "learning_rate": 1.9534416211000323e-05, "loss": 0.1496, "step": 37900 }, { "epoch": 12.222579607590864, "grad_norm": 1.353095531463623, "learning_rate": 1.9454004503055644e-05, "loss": 0.1503, "step": 38000 }, { "epoch": 12.254744290768736, "grad_norm": 1.0767805576324463, "learning_rate": 1.937359279511097e-05, "loss": 0.1513, "step": 38100 }, { "epoch": 12.286908973946607, "grad_norm": 1.162312388420105, "learning_rate": 1.9293181087166292e-05, "loss": 0.153, "step": 38200 }, { "epoch": 12.319073657124477, "grad_norm": 1.1145883798599243, "learning_rate": 1.9212769379221616e-05, "loss": 0.1561, "step": 38300 }, { "epoch": 12.351238340302348, "grad_norm": 1.1558126211166382, "learning_rate": 1.9132357671276937e-05, "loss": 0.1535, "step": 38400 }, { "epoch": 12.383403023480218, "grad_norm": 0.838792085647583, "learning_rate": 1.9051945963332264e-05, "loss": 0.1518, "step": 38500 }, { "epoch": 12.41556770665809, "grad_norm": 1.1334800720214844, "learning_rate": 1.8971534255387585e-05, "loss": 0.1522, "step": 38600 }, { "epoch": 12.44773238983596, "grad_norm": 1.1585274934768677, "learning_rate": 1.889112254744291e-05, "loss": 0.1535, "step": 38700 }, { "epoch": 12.47989707301383, "grad_norm": 1.1669962406158447, "learning_rate": 1.8810710839498233e-05, "loss": 0.1556, "step": 38800 }, { "epoch": 12.512061756191702, "grad_norm": 1.205310344696045, "learning_rate": 1.8730299131553554e-05, "loss": 0.1585, "step": 38900 }, { "epoch": 12.544226439369572, "grad_norm": 0.9617523550987244, "learning_rate": 1.864988742360888e-05, "loss": 0.1571, "step": 39000 }, { "epoch": 12.576391122547443, "grad_norm": 1.259700894355774, "learning_rate": 1.8569475715664203e-05, "loss": 0.1563, "step": 39100 }, { "epoch": 12.608555805725313, "grad_norm": 1.0180591344833374, "learning_rate": 1.8489064007719527e-05, "loss": 0.1553, "step": 39200 }, { "epoch": 12.640720488903185, "grad_norm": 1.2915213108062744, "learning_rate": 1.8408652299774847e-05, "loss": 0.1599, "step": 39300 }, { "epoch": 12.672885172081054, "grad_norm": 1.1494709253311157, "learning_rate": 1.832824059183017e-05, "loss": 0.1575, "step": 39400 }, { "epoch": 12.705049855258926, "grad_norm": 1.1564290523529053, "learning_rate": 1.8247828883885496e-05, "loss": 0.1563, "step": 39500 }, { "epoch": 12.737214538436797, "grad_norm": 1.2692680358886719, "learning_rate": 1.8167417175940816e-05, "loss": 0.1577, "step": 39600 }, { "epoch": 12.769379221614667, "grad_norm": 1.0520470142364502, "learning_rate": 1.808780958507559e-05, "loss": 0.1603, "step": 39700 }, { "epoch": 12.801543904792538, "grad_norm": 1.1396217346191406, "learning_rate": 1.800739787713091e-05, "loss": 0.1601, "step": 39800 }, { "epoch": 12.833708587970408, "grad_norm": 1.1824222803115845, "learning_rate": 1.7926986169186237e-05, "loss": 0.1597, "step": 39900 }, { "epoch": 12.86587327114828, "grad_norm": 1.0760473012924194, "learning_rate": 1.7846574461241558e-05, "loss": 0.1557, "step": 40000 }, { "epoch": 12.89803795432615, "grad_norm": 1.2268940210342407, "learning_rate": 1.7766162753296882e-05, "loss": 0.1574, "step": 40100 }, { "epoch": 12.93020263750402, "grad_norm": 1.4867218732833862, "learning_rate": 1.7685751045352203e-05, "loss": 0.1599, "step": 40200 }, { "epoch": 12.962367320681892, "grad_norm": 1.271653413772583, "learning_rate": 1.7605339337407527e-05, "loss": 0.1577, "step": 40300 }, { "epoch": 12.994532003859762, "grad_norm": 1.1374136209487915, "learning_rate": 1.752492762946285e-05, "loss": 0.1612, "step": 40400 }, { "epoch": 13.026696687037633, "grad_norm": 1.023424744606018, "learning_rate": 1.744451592151817e-05, "loss": 0.1305, "step": 40500 }, { "epoch": 13.058861370215503, "grad_norm": 1.096927523612976, "learning_rate": 1.73641042135735e-05, "loss": 0.1237, "step": 40600 }, { "epoch": 13.091026053393374, "grad_norm": 1.0655758380889893, "learning_rate": 1.7284496622708268e-05, "loss": 0.1263, "step": 40700 }, { "epoch": 13.123190736571244, "grad_norm": 1.0864534378051758, "learning_rate": 1.7204084914763592e-05, "loss": 0.1234, "step": 40800 }, { "epoch": 13.155355419749116, "grad_norm": 1.1215620040893555, "learning_rate": 1.7123673206818913e-05, "loss": 0.1271, "step": 40900 }, { "epoch": 13.187520102926985, "grad_norm": 1.2142025232315063, "learning_rate": 1.7043261498874237e-05, "loss": 0.1266, "step": 41000 }, { "epoch": 13.219684786104857, "grad_norm": 1.1056804656982422, "learning_rate": 1.696284979092956e-05, "loss": 0.1296, "step": 41100 }, { "epoch": 13.251849469282728, "grad_norm": 0.9434877038002014, "learning_rate": 1.6882438082984882e-05, "loss": 0.1288, "step": 41200 }, { "epoch": 13.284014152460598, "grad_norm": 0.9666481018066406, "learning_rate": 1.6802026375040206e-05, "loss": 0.1297, "step": 41300 }, { "epoch": 13.31617883563847, "grad_norm": 1.1036096811294556, "learning_rate": 1.672161466709553e-05, "loss": 0.1267, "step": 41400 }, { "epoch": 13.348343518816339, "grad_norm": 1.0307798385620117, "learning_rate": 1.6641202959150854e-05, "loss": 0.1297, "step": 41500 }, { "epoch": 13.38050820199421, "grad_norm": 1.0009799003601074, "learning_rate": 1.6560791251206175e-05, "loss": 0.1286, "step": 41600 }, { "epoch": 13.41267288517208, "grad_norm": 1.098833680152893, "learning_rate": 1.6480379543261503e-05, "loss": 0.131, "step": 41700 }, { "epoch": 13.444837568349952, "grad_norm": 1.3332812786102295, "learning_rate": 1.6399967835316823e-05, "loss": 0.1311, "step": 41800 }, { "epoch": 13.477002251527823, "grad_norm": 1.133938193321228, "learning_rate": 1.6319556127372148e-05, "loss": 0.1291, "step": 41900 }, { "epoch": 13.509166934705693, "grad_norm": 1.1072946786880493, "learning_rate": 1.6239144419427468e-05, "loss": 0.1311, "step": 42000 }, { "epoch": 13.541331617883564, "grad_norm": 1.088657021522522, "learning_rate": 1.6158732711482792e-05, "loss": 0.1301, "step": 42100 }, { "epoch": 13.573496301061434, "grad_norm": 1.2645858526229858, "learning_rate": 1.6078321003538117e-05, "loss": 0.13, "step": 42200 }, { "epoch": 13.605660984239305, "grad_norm": 1.117727518081665, "learning_rate": 1.5997909295593437e-05, "loss": 0.1316, "step": 42300 }, { "epoch": 13.637825667417175, "grad_norm": 1.268072485923767, "learning_rate": 1.5917497587648765e-05, "loss": 0.1292, "step": 42400 }, { "epoch": 13.669990350595047, "grad_norm": 1.0627697706222534, "learning_rate": 1.5837085879704086e-05, "loss": 0.1325, "step": 42500 }, { "epoch": 13.702155033772918, "grad_norm": 1.2271528244018555, "learning_rate": 1.575667417175941e-05, "loss": 0.1326, "step": 42600 }, { "epoch": 13.734319716950788, "grad_norm": 1.1797391176223755, "learning_rate": 1.567626246381473e-05, "loss": 0.1361, "step": 42700 }, { "epoch": 13.76648440012866, "grad_norm": 1.1041834354400635, "learning_rate": 1.5595850755870055e-05, "loss": 0.1362, "step": 42800 }, { "epoch": 13.798649083306529, "grad_norm": 1.1098655462265015, "learning_rate": 1.551543904792538e-05, "loss": 0.1349, "step": 42900 }, { "epoch": 13.8308137664844, "grad_norm": 1.1673643589019775, "learning_rate": 1.54350273399807e-05, "loss": 0.1372, "step": 43000 }, { "epoch": 13.86297844966227, "grad_norm": 1.078134298324585, "learning_rate": 1.5354615632036027e-05, "loss": 0.1326, "step": 43100 }, { "epoch": 13.895143132840142, "grad_norm": 1.3180980682373047, "learning_rate": 1.5274203924091348e-05, "loss": 0.1319, "step": 43200 }, { "epoch": 13.927307816018013, "grad_norm": 1.16948664188385, "learning_rate": 1.519379221614667e-05, "loss": 0.1316, "step": 43300 }, { "epoch": 13.959472499195883, "grad_norm": 1.0479273796081543, "learning_rate": 1.5113380508201993e-05, "loss": 0.1361, "step": 43400 }, { "epoch": 13.991637182373754, "grad_norm": 1.2957446575164795, "learning_rate": 1.5032968800257319e-05, "loss": 0.1348, "step": 43500 }, { "epoch": 14.023801865551624, "grad_norm": 0.9477380514144897, "learning_rate": 1.4952557092312641e-05, "loss": 0.1132, "step": 43600 }, { "epoch": 14.055966548729495, "grad_norm": 1.090921401977539, "learning_rate": 1.4872145384367964e-05, "loss": 0.1055, "step": 43700 }, { "epoch": 14.088131231907365, "grad_norm": 0.9108441472053528, "learning_rate": 1.479173367642329e-05, "loss": 0.107, "step": 43800 }, { "epoch": 14.120295915085237, "grad_norm": 0.9658478498458862, "learning_rate": 1.4711321968478612e-05, "loss": 0.1074, "step": 43900 }, { "epoch": 14.152460598263108, "grad_norm": 1.125514030456543, "learning_rate": 1.4630910260533934e-05, "loss": 0.1069, "step": 44000 }, { "epoch": 14.184625281440978, "grad_norm": 1.137123942375183, "learning_rate": 1.4550498552589258e-05, "loss": 0.108, "step": 44100 }, { "epoch": 14.21678996461885, "grad_norm": 1.2043941020965576, "learning_rate": 1.4470086844644581e-05, "loss": 0.1086, "step": 44200 }, { "epoch": 14.248954647796719, "grad_norm": 1.0090394020080566, "learning_rate": 1.4389675136699903e-05, "loss": 0.1088, "step": 44300 }, { "epoch": 14.28111933097459, "grad_norm": 1.0625391006469727, "learning_rate": 1.4309263428755226e-05, "loss": 0.1098, "step": 44400 }, { "epoch": 14.31328401415246, "grad_norm": 0.9002700448036194, "learning_rate": 1.4228851720810552e-05, "loss": 0.109, "step": 44500 }, { "epoch": 14.345448697330331, "grad_norm": 1.1506325006484985, "learning_rate": 1.4148440012865874e-05, "loss": 0.1095, "step": 44600 }, { "epoch": 14.377613380508201, "grad_norm": 0.9931826591491699, "learning_rate": 1.4068028304921197e-05, "loss": 0.1105, "step": 44700 }, { "epoch": 14.409778063686073, "grad_norm": 1.258078932762146, "learning_rate": 1.398761659697652e-05, "loss": 0.1107, "step": 44800 }, { "epoch": 14.441942746863944, "grad_norm": 1.1100337505340576, "learning_rate": 1.3907204889031843e-05, "loss": 0.1119, "step": 44900 }, { "epoch": 14.474107430041814, "grad_norm": 1.0397074222564697, "learning_rate": 1.3826793181087166e-05, "loss": 0.1104, "step": 45000 }, { "epoch": 14.506272113219685, "grad_norm": 1.0517898797988892, "learning_rate": 1.3746381473142488e-05, "loss": 0.1132, "step": 45100 }, { "epoch": 14.538436796397555, "grad_norm": 0.9665665030479431, "learning_rate": 1.3665969765197814e-05, "loss": 0.1124, "step": 45200 }, { "epoch": 14.570601479575426, "grad_norm": 1.032689094543457, "learning_rate": 1.3585558057253136e-05, "loss": 0.1108, "step": 45300 }, { "epoch": 14.602766162753296, "grad_norm": 1.0238643884658813, "learning_rate": 1.3505146349308459e-05, "loss": 0.1109, "step": 45400 }, { "epoch": 14.634930845931168, "grad_norm": 0.9571881890296936, "learning_rate": 1.3424734641363785e-05, "loss": 0.1124, "step": 45500 }, { "epoch": 14.667095529109039, "grad_norm": 0.9984620809555054, "learning_rate": 1.3345127050498553e-05, "loss": 0.1108, "step": 45600 }, { "epoch": 14.699260212286909, "grad_norm": 1.1483553647994995, "learning_rate": 1.3264715342553876e-05, "loss": 0.1127, "step": 45700 }, { "epoch": 14.73142489546478, "grad_norm": 1.1123391389846802, "learning_rate": 1.3184303634609198e-05, "loss": 0.1125, "step": 45800 }, { "epoch": 14.76358957864265, "grad_norm": 1.0833643674850464, "learning_rate": 1.3103891926664524e-05, "loss": 0.1124, "step": 45900 }, { "epoch": 14.795754261820521, "grad_norm": 0.8864045143127441, "learning_rate": 1.3023480218719847e-05, "loss": 0.1117, "step": 46000 }, { "epoch": 14.827918944998391, "grad_norm": 1.0828707218170166, "learning_rate": 1.2943068510775169e-05, "loss": 0.1135, "step": 46100 }, { "epoch": 14.860083628176262, "grad_norm": 0.9754586815834045, "learning_rate": 1.2862656802830492e-05, "loss": 0.1146, "step": 46200 }, { "epoch": 14.892248311354134, "grad_norm": 1.0156996250152588, "learning_rate": 1.2782245094885817e-05, "loss": 0.1117, "step": 46300 }, { "epoch": 14.924412994532004, "grad_norm": 1.1564407348632812, "learning_rate": 1.270183338694114e-05, "loss": 0.113, "step": 46400 }, { "epoch": 14.956577677709875, "grad_norm": 1.151871919631958, "learning_rate": 1.2621421678996462e-05, "loss": 0.1156, "step": 46500 }, { "epoch": 14.988742360887745, "grad_norm": 1.0806056261062622, "learning_rate": 1.2541009971051786e-05, "loss": 0.1115, "step": 46600 }, { "epoch": 15.020907044065616, "grad_norm": 1.0419082641601562, "learning_rate": 1.2460598263107109e-05, "loss": 0.0983, "step": 46700 }, { "epoch": 15.053071727243486, "grad_norm": 1.059793472290039, "learning_rate": 1.2380186555162431e-05, "loss": 0.0914, "step": 46800 }, { "epoch": 15.085236410421357, "grad_norm": 0.9557967782020569, "learning_rate": 1.2299774847217756e-05, "loss": 0.0898, "step": 46900 }, { "epoch": 15.117401093599229, "grad_norm": 0.9494596719741821, "learning_rate": 1.2219363139273078e-05, "loss": 0.0923, "step": 47000 }, { "epoch": 15.149565776777099, "grad_norm": 0.9562474489212036, "learning_rate": 1.2138951431328402e-05, "loss": 0.0951, "step": 47100 }, { "epoch": 15.18173045995497, "grad_norm": 0.9924389123916626, "learning_rate": 1.2058539723383726e-05, "loss": 0.0924, "step": 47200 }, { "epoch": 15.21389514313284, "grad_norm": 0.9415188431739807, "learning_rate": 1.1978128015439049e-05, "loss": 0.0946, "step": 47300 }, { "epoch": 15.246059826310711, "grad_norm": 1.003299593925476, "learning_rate": 1.1897716307494373e-05, "loss": 0.0938, "step": 47400 }, { "epoch": 15.278224509488581, "grad_norm": 0.977337121963501, "learning_rate": 1.1817304599549694e-05, "loss": 0.0949, "step": 47500 }, { "epoch": 15.310389192666452, "grad_norm": 0.9806488156318665, "learning_rate": 1.1736892891605018e-05, "loss": 0.0949, "step": 47600 }, { "epoch": 15.342553875844324, "grad_norm": 0.9360681176185608, "learning_rate": 1.1656481183660342e-05, "loss": 0.0927, "step": 47700 }, { "epoch": 15.374718559022194, "grad_norm": 1.0371599197387695, "learning_rate": 1.1576069475715664e-05, "loss": 0.0928, "step": 47800 }, { "epoch": 15.406883242200065, "grad_norm": 1.0855216979980469, "learning_rate": 1.1495657767770989e-05, "loss": 0.0937, "step": 47900 }, { "epoch": 15.439047925377935, "grad_norm": 0.9102591276168823, "learning_rate": 1.1415246059826311e-05, "loss": 0.0954, "step": 48000 }, { "epoch": 15.471212608555806, "grad_norm": 0.8920603394508362, "learning_rate": 1.1334834351881635e-05, "loss": 0.0944, "step": 48100 }, { "epoch": 15.503377291733676, "grad_norm": 1.0157232284545898, "learning_rate": 1.1254422643936958e-05, "loss": 0.0941, "step": 48200 }, { "epoch": 15.535541974911547, "grad_norm": 1.0315204858779907, "learning_rate": 1.1174010935992282e-05, "loss": 0.0943, "step": 48300 }, { "epoch": 15.567706658089417, "grad_norm": 0.9432384371757507, "learning_rate": 1.109440334512705e-05, "loss": 0.0942, "step": 48400 }, { "epoch": 15.599871341267288, "grad_norm": 1.0601836442947388, "learning_rate": 1.1013991637182375e-05, "loss": 0.0959, "step": 48500 }, { "epoch": 15.63203602444516, "grad_norm": 1.0262197256088257, "learning_rate": 1.0933579929237697e-05, "loss": 0.0959, "step": 48600 }, { "epoch": 15.66420070762303, "grad_norm": 1.1089009046554565, "learning_rate": 1.0853168221293021e-05, "loss": 0.0927, "step": 48700 }, { "epoch": 15.696365390800901, "grad_norm": 0.9688426852226257, "learning_rate": 1.0772756513348344e-05, "loss": 0.0943, "step": 48800 }, { "epoch": 15.72853007397877, "grad_norm": 0.9896754026412964, "learning_rate": 1.0692344805403668e-05, "loss": 0.0943, "step": 48900 }, { "epoch": 15.760694757156642, "grad_norm": 1.1106398105621338, "learning_rate": 1.0611933097458992e-05, "loss": 0.0978, "step": 49000 }, { "epoch": 15.792859440334512, "grad_norm": 0.9831110835075378, "learning_rate": 1.0531521389514314e-05, "loss": 0.0968, "step": 49100 }, { "epoch": 15.825024123512383, "grad_norm": 0.9822130799293518, "learning_rate": 1.0451109681569637e-05, "loss": 0.0961, "step": 49200 }, { "epoch": 15.857188806690255, "grad_norm": 0.9292415380477905, "learning_rate": 1.037069797362496e-05, "loss": 0.0954, "step": 49300 }, { "epoch": 15.889353489868125, "grad_norm": 1.1434463262557983, "learning_rate": 1.0290286265680284e-05, "loss": 0.0971, "step": 49400 }, { "epoch": 15.921518173045996, "grad_norm": 0.9617388844490051, "learning_rate": 1.0209874557735606e-05, "loss": 0.0958, "step": 49500 }, { "epoch": 15.953682856223866, "grad_norm": 1.0446373224258423, "learning_rate": 1.012946284979093e-05, "loss": 0.0946, "step": 49600 }, { "epoch": 15.985847539401737, "grad_norm": 1.1061015129089355, "learning_rate": 1.0049051141846254e-05, "loss": 0.0983, "step": 49700 }, { "epoch": 16.018012222579607, "grad_norm": 0.9545947909355164, "learning_rate": 9.968639433901577e-06, "loss": 0.0845, "step": 49800 }, { "epoch": 16.05017690575748, "grad_norm": 0.9615588188171387, "learning_rate": 9.888227725956901e-06, "loss": 0.0786, "step": 49900 }, { "epoch": 16.08234158893535, "grad_norm": 0.9855136275291443, "learning_rate": 9.807816018012223e-06, "loss": 0.0802, "step": 50000 }, { "epoch": 16.11450627211322, "grad_norm": 1.1269737482070923, "learning_rate": 9.727404310067546e-06, "loss": 0.0775, "step": 50100 }, { "epoch": 16.14667095529109, "grad_norm": 0.8828549385070801, "learning_rate": 9.64699260212287e-06, "loss": 0.0805, "step": 50200 }, { "epoch": 16.17883563846896, "grad_norm": 0.9752237200737, "learning_rate": 9.566580894178192e-06, "loss": 0.0791, "step": 50300 }, { "epoch": 16.211000321646832, "grad_norm": 1.014918565750122, "learning_rate": 9.486169186233517e-06, "loss": 0.0802, "step": 50400 }, { "epoch": 16.243165004824704, "grad_norm": 0.9082208871841431, "learning_rate": 9.405757478288839e-06, "loss": 0.0805, "step": 50500 }, { "epoch": 16.27532968800257, "grad_norm": 0.793496310710907, "learning_rate": 9.325345770344163e-06, "loss": 0.0791, "step": 50600 }, { "epoch": 16.307494371180443, "grad_norm": 0.922303318977356, "learning_rate": 9.244934062399486e-06, "loss": 0.0823, "step": 50700 }, { "epoch": 16.339659054358314, "grad_norm": 0.9956395626068115, "learning_rate": 9.16452235445481e-06, "loss": 0.0808, "step": 50800 }, { "epoch": 16.371823737536186, "grad_norm": 1.1846128702163696, "learning_rate": 9.084110646510132e-06, "loss": 0.0811, "step": 50900 }, { "epoch": 16.403988420714057, "grad_norm": 0.8967270255088806, "learning_rate": 9.003698938565455e-06, "loss": 0.0812, "step": 51000 }, { "epoch": 16.436153103891925, "grad_norm": 1.0155181884765625, "learning_rate": 8.923287230620779e-06, "loss": 0.0809, "step": 51100 }, { "epoch": 16.468317787069797, "grad_norm": 1.014799952507019, "learning_rate": 8.842875522676101e-06, "loss": 0.0819, "step": 51200 }, { "epoch": 16.50048247024767, "grad_norm": 1.1782442331314087, "learning_rate": 8.763267931810872e-06, "loss": 0.0815, "step": 51300 }, { "epoch": 16.53264715342554, "grad_norm": 0.9318787455558777, "learning_rate": 8.682856223866196e-06, "loss": 0.0814, "step": 51400 }, { "epoch": 16.56481183660341, "grad_norm": 0.9740914702415466, "learning_rate": 8.60244451592152e-06, "loss": 0.0809, "step": 51500 }, { "epoch": 16.59697651978128, "grad_norm": 1.0369396209716797, "learning_rate": 8.522032807976842e-06, "loss": 0.0808, "step": 51600 }, { "epoch": 16.62914120295915, "grad_norm": 0.8972563743591309, "learning_rate": 8.441621100032165e-06, "loss": 0.079, "step": 51700 }, { "epoch": 16.661305886137022, "grad_norm": 1.1047544479370117, "learning_rate": 8.361209392087487e-06, "loss": 0.0803, "step": 51800 }, { "epoch": 16.693470569314893, "grad_norm": 1.0711828470230103, "learning_rate": 8.280797684142811e-06, "loss": 0.0832, "step": 51900 }, { "epoch": 16.72563525249276, "grad_norm": 0.8830587267875671, "learning_rate": 8.200385976198136e-06, "loss": 0.0815, "step": 52000 }, { "epoch": 16.757799935670633, "grad_norm": 0.8716908693313599, "learning_rate": 8.119974268253458e-06, "loss": 0.0808, "step": 52100 }, { "epoch": 16.789964618848504, "grad_norm": 1.0468723773956299, "learning_rate": 8.039562560308782e-06, "loss": 0.0825, "step": 52200 }, { "epoch": 16.822129302026376, "grad_norm": 1.0085501670837402, "learning_rate": 7.959150852364105e-06, "loss": 0.0817, "step": 52300 }, { "epoch": 16.854293985204247, "grad_norm": 0.8367465138435364, "learning_rate": 7.878739144419429e-06, "loss": 0.0819, "step": 52400 }, { "epoch": 16.886458668382115, "grad_norm": 0.9274358749389648, "learning_rate": 7.798327436474751e-06, "loss": 0.0818, "step": 52500 }, { "epoch": 16.918623351559987, "grad_norm": 1.1087638139724731, "learning_rate": 7.717915728530074e-06, "loss": 0.083, "step": 52600 }, { "epoch": 16.950788034737858, "grad_norm": 1.0249075889587402, "learning_rate": 7.637504020585398e-06, "loss": 0.0818, "step": 52700 }, { "epoch": 16.98295271791573, "grad_norm": 0.9565846920013428, "learning_rate": 7.55709231264072e-06, "loss": 0.0802, "step": 52800 }, { "epoch": 17.015117401093597, "grad_norm": 0.8502324819564819, "learning_rate": 7.4766806046960445e-06, "loss": 0.0762, "step": 52900 }, { "epoch": 17.04728208427147, "grad_norm": 0.998939037322998, "learning_rate": 7.396268896751367e-06, "loss": 0.0693, "step": 53000 }, { "epoch": 17.07944676744934, "grad_norm": 0.8634834289550781, "learning_rate": 7.31585718880669e-06, "loss": 0.069, "step": 53100 }, { "epoch": 17.111611450627212, "grad_norm": 0.8927751779556274, "learning_rate": 7.2354454808620144e-06, "loss": 0.0686, "step": 53200 }, { "epoch": 17.143776133805083, "grad_norm": 0.7095975875854492, "learning_rate": 7.155033772917337e-06, "loss": 0.0688, "step": 53300 }, { "epoch": 17.17594081698295, "grad_norm": 1.0496543645858765, "learning_rate": 7.074622064972661e-06, "loss": 0.0709, "step": 53400 }, { "epoch": 17.208105500160823, "grad_norm": 0.8392481803894043, "learning_rate": 6.9942103570279835e-06, "loss": 0.0692, "step": 53500 }, { "epoch": 17.240270183338694, "grad_norm": 0.9044252634048462, "learning_rate": 6.913798649083307e-06, "loss": 0.0706, "step": 53600 }, { "epoch": 17.272434866516566, "grad_norm": 0.8854852914810181, "learning_rate": 6.833386941138629e-06, "loss": 0.0693, "step": 53700 }, { "epoch": 17.304599549694437, "grad_norm": 0.8078619837760925, "learning_rate": 6.752975233193953e-06, "loss": 0.0702, "step": 53800 }, { "epoch": 17.336764232872305, "grad_norm": 0.9859098196029663, "learning_rate": 6.672563525249277e-06, "loss": 0.07, "step": 53900 }, { "epoch": 17.368928916050177, "grad_norm": 1.025367259979248, "learning_rate": 6.592151817304599e-06, "loss": 0.0699, "step": 54000 }, { "epoch": 17.401093599228048, "grad_norm": 1.0151135921478271, "learning_rate": 6.511740109359923e-06, "loss": 0.0704, "step": 54100 }, { "epoch": 17.43325828240592, "grad_norm": 1.0093625783920288, "learning_rate": 6.431328401415246e-06, "loss": 0.0694, "step": 54200 }, { "epoch": 17.465422965583787, "grad_norm": 0.8962597846984863, "learning_rate": 6.35091669347057e-06, "loss": 0.0703, "step": 54300 }, { "epoch": 17.49758764876166, "grad_norm": 0.8843083381652832, "learning_rate": 6.270504985525893e-06, "loss": 0.0709, "step": 54400 }, { "epoch": 17.52975233193953, "grad_norm": 1.1030328273773193, "learning_rate": 6.190093277581216e-06, "loss": 0.0694, "step": 54500 }, { "epoch": 17.5619170151174, "grad_norm": 0.9267283082008362, "learning_rate": 6.109681569636539e-06, "loss": 0.0708, "step": 54600 }, { "epoch": 17.594081698295273, "grad_norm": 1.0376532077789307, "learning_rate": 6.029269861691863e-06, "loss": 0.0708, "step": 54700 }, { "epoch": 17.62624638147314, "grad_norm": 0.8866332173347473, "learning_rate": 5.9488581537471864e-06, "loss": 0.0709, "step": 54800 }, { "epoch": 17.658411064651013, "grad_norm": 0.9702731966972351, "learning_rate": 5.868446445802509e-06, "loss": 0.0711, "step": 54900 }, { "epoch": 17.690575747828884, "grad_norm": 0.904787540435791, "learning_rate": 5.788034737857832e-06, "loss": 0.0719, "step": 55000 }, { "epoch": 17.722740431006756, "grad_norm": 0.8797966241836548, "learning_rate": 5.7076230299131555e-06, "loss": 0.0712, "step": 55100 }, { "epoch": 17.754905114184623, "grad_norm": 0.9212404489517212, "learning_rate": 5.628015439047925e-06, "loss": 0.0687, "step": 55200 }, { "epoch": 17.787069797362495, "grad_norm": 0.9659159779548645, "learning_rate": 5.547603731103248e-06, "loss": 0.0716, "step": 55300 }, { "epoch": 17.819234480540366, "grad_norm": 0.9781568050384521, "learning_rate": 5.4671920231585725e-06, "loss": 0.0712, "step": 55400 }, { "epoch": 17.851399163718238, "grad_norm": 0.9969457387924194, "learning_rate": 5.386780315213896e-06, "loss": 0.0716, "step": 55500 }, { "epoch": 17.88356384689611, "grad_norm": 0.8759910464286804, "learning_rate": 5.306368607269218e-06, "loss": 0.0687, "step": 55600 }, { "epoch": 17.915728530073977, "grad_norm": 0.9174026250839233, "learning_rate": 5.225956899324542e-06, "loss": 0.0705, "step": 55700 }, { "epoch": 17.94789321325185, "grad_norm": 1.0490466356277466, "learning_rate": 5.145545191379865e-06, "loss": 0.0701, "step": 55800 }, { "epoch": 17.98005789642972, "grad_norm": 0.8875726461410522, "learning_rate": 5.065133483435188e-06, "loss": 0.0713, "step": 55900 }, { "epoch": 18.01222257960759, "grad_norm": 0.7829542756080627, "learning_rate": 4.984721775490512e-06, "loss": 0.068, "step": 56000 }, { "epoch": 18.044387262785463, "grad_norm": 0.9281951785087585, "learning_rate": 4.904310067545835e-06, "loss": 0.0615, "step": 56100 }, { "epoch": 18.07655194596333, "grad_norm": 0.8894643783569336, "learning_rate": 4.823898359601158e-06, "loss": 0.0608, "step": 56200 }, { "epoch": 18.108716629141203, "grad_norm": 0.7356093525886536, "learning_rate": 4.743486651656481e-06, "loss": 0.0616, "step": 56300 }, { "epoch": 18.140881312319074, "grad_norm": 0.7402681708335876, "learning_rate": 4.663074943711805e-06, "loss": 0.0616, "step": 56400 }, { "epoch": 18.173045995496945, "grad_norm": 0.8778253793716431, "learning_rate": 4.582663235767128e-06, "loss": 0.0613, "step": 56500 }, { "epoch": 18.205210678674813, "grad_norm": 0.7177903652191162, "learning_rate": 4.502251527822451e-06, "loss": 0.061, "step": 56600 }, { "epoch": 18.237375361852685, "grad_norm": 1.0053707361221313, "learning_rate": 4.421839819877775e-06, "loss": 0.0614, "step": 56700 }, { "epoch": 18.269540045030556, "grad_norm": 0.8634065985679626, "learning_rate": 4.341428111933098e-06, "loss": 0.0609, "step": 56800 }, { "epoch": 18.301704728208428, "grad_norm": 0.8410573601722717, "learning_rate": 4.261016403988421e-06, "loss": 0.0618, "step": 56900 }, { "epoch": 18.3338694113863, "grad_norm": 0.9490267038345337, "learning_rate": 4.180604696043744e-06, "loss": 0.0617, "step": 57000 }, { "epoch": 18.366034094564167, "grad_norm": 0.8697332739830017, "learning_rate": 4.100192988099068e-06, "loss": 0.0617, "step": 57100 }, { "epoch": 18.39819877774204, "grad_norm": 1.0434635877609253, "learning_rate": 4.019781280154391e-06, "loss": 0.0613, "step": 57200 }, { "epoch": 18.43036346091991, "grad_norm": 1.034525990486145, "learning_rate": 3.9393695722097144e-06, "loss": 0.0631, "step": 57300 }, { "epoch": 18.46252814409778, "grad_norm": 0.8523834347724915, "learning_rate": 3.858957864265037e-06, "loss": 0.0619, "step": 57400 }, { "epoch": 18.494692827275653, "grad_norm": 0.9277452826499939, "learning_rate": 3.77854615632036e-06, "loss": 0.0619, "step": 57500 }, { "epoch": 18.52685751045352, "grad_norm": 0.8697579503059387, "learning_rate": 3.6981344483756835e-06, "loss": 0.0609, "step": 57600 }, { "epoch": 18.559022193631392, "grad_norm": 1.1431069374084473, "learning_rate": 3.6177227404310072e-06, "loss": 0.0627, "step": 57700 }, { "epoch": 18.591186876809264, "grad_norm": 0.8121923804283142, "learning_rate": 3.5373110324863305e-06, "loss": 0.0611, "step": 57800 }, { "epoch": 18.623351559987135, "grad_norm": 0.8631033301353455, "learning_rate": 3.4568993245416534e-06, "loss": 0.0643, "step": 57900 }, { "epoch": 18.655516243165003, "grad_norm": 0.8957691788673401, "learning_rate": 3.3764876165969767e-06, "loss": 0.0641, "step": 58000 }, { "epoch": 18.687680926342875, "grad_norm": 0.8032559156417847, "learning_rate": 3.2960759086522996e-06, "loss": 0.0628, "step": 58100 }, { "epoch": 18.719845609520746, "grad_norm": 0.8172631859779358, "learning_rate": 3.215664200707623e-06, "loss": 0.061, "step": 58200 }, { "epoch": 18.752010292698618, "grad_norm": 0.8193603157997131, "learning_rate": 3.1352524927629466e-06, "loss": 0.0624, "step": 58300 }, { "epoch": 18.78417497587649, "grad_norm": 0.8261486887931824, "learning_rate": 3.0548407848182695e-06, "loss": 0.0601, "step": 58400 }, { "epoch": 18.816339659054357, "grad_norm": 0.9643902778625488, "learning_rate": 2.9744290768735932e-06, "loss": 0.0618, "step": 58500 }, { "epoch": 18.84850434223223, "grad_norm": 0.919653058052063, "learning_rate": 2.894017368928916e-06, "loss": 0.0621, "step": 58600 }, { "epoch": 18.8806690254101, "grad_norm": 0.99927818775177, "learning_rate": 2.8136056609842394e-06, "loss": 0.0626, "step": 58700 }, { "epoch": 18.91283370858797, "grad_norm": 1.0389022827148438, "learning_rate": 2.7331939530395627e-06, "loss": 0.0607, "step": 58800 }, { "epoch": 18.944998391765843, "grad_norm": 0.8360564708709717, "learning_rate": 2.652782245094886e-06, "loss": 0.0621, "step": 58900 }, { "epoch": 18.97716307494371, "grad_norm": 0.7946717143058777, "learning_rate": 2.572370537150209e-06, "loss": 0.0613, "step": 59000 }, { "epoch": 19.009327758121582, "grad_norm": 0.8464102149009705, "learning_rate": 2.4919588292055326e-06, "loss": 0.0598, "step": 59100 }, { "epoch": 19.041492441299454, "grad_norm": 0.8916983008384705, "learning_rate": 2.4115471212608555e-06, "loss": 0.0553, "step": 59200 }, { "epoch": 19.073657124477325, "grad_norm": 0.8545793294906616, "learning_rate": 2.3311354133161792e-06, "loss": 0.0561, "step": 59300 }, { "epoch": 19.105821807655193, "grad_norm": 0.8248264789581299, "learning_rate": 2.2515278224509488e-06, "loss": 0.0551, "step": 59400 }, { "epoch": 19.137986490833065, "grad_norm": 0.9491187334060669, "learning_rate": 2.1711161145062725e-06, "loss": 0.0569, "step": 59500 }, { "epoch": 19.170151174010936, "grad_norm": 0.7597662806510925, "learning_rate": 2.0907044065615954e-06, "loss": 0.0554, "step": 59600 }, { "epoch": 19.202315857188808, "grad_norm": 0.7805320024490356, "learning_rate": 2.0102926986169187e-06, "loss": 0.0567, "step": 59700 }, { "epoch": 19.23448054036668, "grad_norm": 0.7836379408836365, "learning_rate": 1.929880990672242e-06, "loss": 0.0551, "step": 59800 }, { "epoch": 19.266645223544547, "grad_norm": 0.8051148056983948, "learning_rate": 1.849469282727565e-06, "loss": 0.0565, "step": 59900 }, { "epoch": 19.29880990672242, "grad_norm": 0.8301971554756165, "learning_rate": 1.7690575747828884e-06, "loss": 0.0567, "step": 60000 }, { "epoch": 19.33097458990029, "grad_norm": 0.8996140956878662, "learning_rate": 1.6886458668382117e-06, "loss": 0.056, "step": 60100 }, { "epoch": 19.36313927307816, "grad_norm": 0.8115219473838806, "learning_rate": 1.608234158893535e-06, "loss": 0.0575, "step": 60200 }, { "epoch": 19.39530395625603, "grad_norm": 0.8823007345199585, "learning_rate": 1.5278224509488583e-06, "loss": 0.0552, "step": 60300 }, { "epoch": 19.4274686394339, "grad_norm": 0.7915894389152527, "learning_rate": 1.4474107430041816e-06, "loss": 0.0561, "step": 60400 }, { "epoch": 19.459633322611772, "grad_norm": 0.7833816409111023, "learning_rate": 1.3669990350595047e-06, "loss": 0.0563, "step": 60500 }, { "epoch": 19.491798005789644, "grad_norm": 0.9442763924598694, "learning_rate": 1.286587327114828e-06, "loss": 0.0583, "step": 60600 }, { "epoch": 19.523962688967515, "grad_norm": 0.8519619107246399, "learning_rate": 1.2061756191701513e-06, "loss": 0.0561, "step": 60700 }, { "epoch": 19.556127372145383, "grad_norm": 0.8903068900108337, "learning_rate": 1.1257639112254744e-06, "loss": 0.0564, "step": 60800 }, { "epoch": 19.588292055323254, "grad_norm": 0.9718756079673767, "learning_rate": 1.0453522032807977e-06, "loss": 0.0556, "step": 60900 }, { "epoch": 19.620456738501126, "grad_norm": 0.7963365316390991, "learning_rate": 9.64940495336121e-07, "loss": 0.0561, "step": 61000 }, { "epoch": 19.652621421678997, "grad_norm": 0.7003450989723206, "learning_rate": 8.845287873914442e-07, "loss": 0.0558, "step": 61100 }, { "epoch": 19.68478610485687, "grad_norm": 0.8856065273284912, "learning_rate": 8.041170794467675e-07, "loss": 0.0558, "step": 61200 }, { "epoch": 19.716950788034737, "grad_norm": 0.6978211998939514, "learning_rate": 7.237053715020908e-07, "loss": 0.0547, "step": 61300 }, { "epoch": 19.74911547121261, "grad_norm": 0.6877977848052979, "learning_rate": 6.440977806368608e-07, "loss": 0.0566, "step": 61400 }, { "epoch": 19.78128015439048, "grad_norm": 0.9814041256904602, "learning_rate": 5.636860726921841e-07, "loss": 0.0544, "step": 61500 }, { "epoch": 19.81344483756835, "grad_norm": 0.8647675514221191, "learning_rate": 4.832743647475073e-07, "loss": 0.0561, "step": 61600 }, { "epoch": 19.84560952074622, "grad_norm": 0.8660107851028442, "learning_rate": 4.028626568028305e-07, "loss": 0.0552, "step": 61700 }, { "epoch": 19.87777420392409, "grad_norm": 0.7724593281745911, "learning_rate": 3.224509488581538e-07, "loss": 0.0563, "step": 61800 }, { "epoch": 19.909938887101962, "grad_norm": 0.8536872863769531, "learning_rate": 2.42039240913477e-07, "loss": 0.0557, "step": 61900 }, { "epoch": 19.942103570279833, "grad_norm": 0.8342153429985046, "learning_rate": 1.6162753296880025e-07, "loss": 0.0561, "step": 62000 } ], "logging_steps": 100, "max_steps": 62180, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 5.396914765824e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }