{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.7021216840707738, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 0.00014999965456413008, "loss": 0.6484, "step": 500 }, { "epoch": 0.02, "learning_rate": 0.0001499986182599296, "loss": 0.5436, "step": 1000 }, { "epoch": 0.03, "learning_rate": 0.00014999689109762647, "loss": 0.5149, "step": 1500 }, { "epoch": 0.03, "learning_rate": 0.00014999447309426713, "loss": 0.4996, "step": 2000 }, { "epoch": 0.04, "learning_rate": 0.00014999136427371623, "loss": 0.4888, "step": 2500 }, { "epoch": 0.05, "learning_rate": 0.0001499875646666566, "loss": 0.4797, "step": 3000 }, { "epoch": 0.06, "learning_rate": 0.00014998307431058887, "loss": 0.4711, "step": 3500 }, { "epoch": 0.07, "learning_rate": 0.00014997789324983099, "loss": 0.463, "step": 4000 }, { "epoch": 0.08, "learning_rate": 0.00014997202153551792, "loss": 0.4545, "step": 4500 }, { "epoch": 0.09, "learning_rate": 0.0001499654592256012, "loss": 0.4466, "step": 5000 }, { "epoch": 0.09, "learning_rate": 0.0001499582063848481, "loss": 0.4398, "step": 5500 }, { "epoch": 0.1, "learning_rate": 0.00014995026308484122, "loss": 0.4337, "step": 6000 }, { "epoch": 0.11, "learning_rate": 0.00014994162940397778, "loss": 0.4286, "step": 6500 }, { "epoch": 0.12, "learning_rate": 0.00014993230542746872, "loss": 0.4242, "step": 7000 }, { "epoch": 0.13, "learning_rate": 0.00014992229124733788, "loss": 0.4203, "step": 7500 }, { "epoch": 0.14, "learning_rate": 0.0001499115869624212, "loss": 0.4173, "step": 8000 }, { "epoch": 0.14, "learning_rate": 0.00014990019267836565, "loss": 0.4141, "step": 8500 }, { "epoch": 0.15, "learning_rate": 0.00014988810850762824, "loss": 0.4116, "step": 9000 }, { "epoch": 0.16, "learning_rate": 0.00014987533456947482, "loss": 0.4095, "step": 9500 }, { "epoch": 0.17, "learning_rate": 0.000149861870989979, "loss": 0.4075, "step": 10000 }, { "epoch": 0.18, "learning_rate": 0.0001498477758876907, "loss": 0.4057, "step": 10500 }, { "epoch": 0.19, "learning_rate": 0.00014983293618814342, "loss": 0.4038, "step": 11000 }, { "epoch": 0.2, "learning_rate": 0.00014981740726570863, "loss": 0.4021, "step": 11500 }, { "epoch": 0.2, "learning_rate": 0.00014980118927365057, "loss": 0.4005, "step": 12000 }, { "epoch": 0.21, "learning_rate": 0.00014978428237203426, "loss": 0.3991, "step": 12500 }, { "epoch": 0.22, "learning_rate": 0.00014976668672772396, "loss": 0.3977, "step": 13000 }, { "epoch": 0.23, "learning_rate": 0.00014974843976988143, "loss": 0.3968, "step": 13500 }, { "epoch": 0.24, "learning_rate": 0.00014972946854455738, "loss": 0.3953, "step": 14000 }, { "epoch": 0.25, "learning_rate": 0.00014970980911752974, "loss": 0.3945, "step": 14500 }, { "epoch": 0.26, "learning_rate": 0.0001496894616828291, "loss": 0.3931, "step": 15000 }, { "epoch": 0.26, "learning_rate": 0.0001496684264412764, "loss": 0.3921, "step": 15500 }, { "epoch": 0.27, "learning_rate": 0.00014964670360048104, "loss": 0.3909, "step": 16000 }, { "epoch": 0.28, "learning_rate": 0.00014962429337483864, "loss": 0.3903, "step": 16500 }, { "epoch": 0.29, "learning_rate": 0.00014960119598552907, "loss": 0.3892, "step": 17000 }, { "epoch": 0.3, "learning_rate": 0.00014957745991457127, "loss": 0.3885, "step": 17500 }, { "epoch": 0.31, "learning_rate": 0.00014955303988623036, "loss": 0.3877, "step": 18000 }, { "epoch": 0.31, "learning_rate": 0.00014952788514615529, "loss": 0.3871, "step": 18500 }, { "epoch": 0.32, "learning_rate": 0.0001495020441939234, "loss": 0.3858, "step": 19000 }, { "epoch": 0.33, "learning_rate": 0.0001494755172845744, "loss": 0.3854, "step": 19500 }, { "epoch": 0.34, "learning_rate": 0.00014944830467991823, "loss": 0.3847, "step": 20000 }, { "epoch": 0.35, "learning_rate": 0.00014942046312846905, "loss": 0.3839, "step": 20500 }, { "epoch": 0.36, "learning_rate": 0.0001493918813157191, "loss": 0.3837, "step": 21000 }, { "epoch": 0.37, "learning_rate": 0.0001493626146331152, "loss": 0.383, "step": 21500 }, { "epoch": 0.37, "learning_rate": 0.00014933266336950772, "loss": 0.3821, "step": 22000 }, { "epoch": 0.38, "learning_rate": 0.00014930208977431846, "loss": 0.382, "step": 22500 }, { "epoch": 0.39, "learning_rate": 0.00014927077160993785, "loss": 0.381, "step": 23000 }, { "epoch": 0.4, "learning_rate": 0.0001492387697710072, "loss": 0.3804, "step": 23500 }, { "epoch": 0.41, "learning_rate": 0.0001492060845733717, "loss": 0.38, "step": 24000 }, { "epoch": 0.42, "learning_rate": 0.00014917278375754084, "loss": 0.3793, "step": 24500 }, { "epoch": 0.43, "learning_rate": 0.00014913880296235724, "loss": 0.3791, "step": 25000 }, { "epoch": 0.43, "learning_rate": 0.00014910407237991193, "loss": 0.3787, "step": 25500 }, { "epoch": 0.44, "learning_rate": 0.00014906865976817066, "loss": 0.3779, "step": 26000 }, { "epoch": 0.45, "learning_rate": 0.00014903256547664165, "loss": 0.3778, "step": 26500 }, { "epoch": 0.46, "learning_rate": 0.0001489958640925119, "loss": 0.377, "step": 27000 }, { "epoch": 0.47, "learning_rate": 0.00014895840887839445, "loss": 0.3768, "step": 27500 }, { "epoch": 0.48, "learning_rate": 0.000148920273072621, "loss": 0.3761, "step": 28000 }, { "epoch": 0.49, "learning_rate": 0.00014888145705157658, "loss": 0.3759, "step": 28500 }, { "epoch": 0.49, "learning_rate": 0.0001488420408682806, "loss": 0.3755, "step": 29000 }, { "epoch": 0.5, "learning_rate": 0.00014880186693119052, "loss": 0.3749, "step": 29500 }, { "epoch": 0.51, "learning_rate": 0.00014876101394745023, "loss": 0.3747, "step": 30000 }, { "epoch": 0.52, "learning_rate": 0.00014871948232026222, "loss": 0.3741, "step": 30500 }, { "epoch": 0.53, "learning_rate": 0.00014867727245952692, "loss": 0.3739, "step": 31000 }, { "epoch": 0.54, "learning_rate": 0.00014863447123337502, "loss": 0.3736, "step": 31500 }, { "epoch": 0.54, "learning_rate": 0.00014859090751637814, "loss": 0.3733, "step": 32000 }, { "epoch": 0.55, "learning_rate": 0.00014854666683481534, "loss": 0.3728, "step": 32500 }, { "epoch": 0.56, "learning_rate": 0.0001485017496253243, "loss": 0.3725, "step": 33000 }, { "epoch": 0.57, "learning_rate": 0.00014845624819224238, "loss": 0.3722, "step": 33500 }, { "epoch": 0.58, "learning_rate": 0.00014840998061432743, "loss": 0.3719, "step": 34000 }, { "epoch": 0.59, "learning_rate": 0.00014836313241655734, "loss": 0.3717, "step": 34500 }, { "epoch": 0.6, "learning_rate": 0.00014831551629313194, "loss": 0.3711, "step": 35000 }, { "epoch": 0.6, "learning_rate": 0.00014826722592314076, "loss": 0.3709, "step": 35500 }, { "epoch": 0.61, "learning_rate": 0.00014821826178319038, "loss": 0.3709, "step": 36000 }, { "epoch": 0.62, "learning_rate": 0.0001481687243030059, "loss": 0.3703, "step": 36500 }, { "epoch": 0.63, "learning_rate": 0.00014811841542465117, "loss": 0.3698, "step": 37000 }, { "epoch": 0.64, "learning_rate": 0.00014806743424503673, "loss": 0.3697, "step": 37500 }, { "epoch": 0.65, "learning_rate": 0.0001480157812673262, "loss": 0.3692, "step": 38000 }, { "epoch": 0.66, "learning_rate": 0.0001479634570013137, "loss": 0.369, "step": 38500 }, { "epoch": 0.66, "learning_rate": 0.00014791046196341849, "loss": 0.3689, "step": 39000 }, { "epoch": 0.67, "learning_rate": 0.0001478567966766803, "loss": 0.3685, "step": 39500 }, { "epoch": 0.68, "learning_rate": 0.0001478024616707538, "loss": 0.3683, "step": 40000 }, { "epoch": 0.69, "learning_rate": 0.00014774756815776608, "loss": 0.368, "step": 40500 }, { "epoch": 0.7, "learning_rate": 0.00014769189666559521, "loss": 0.3679, "step": 41000 }, { "epoch": 0.71, "learning_rate": 0.00014763555708173254, "loss": 0.3676, "step": 41500 }, { "epoch": 0.71, "learning_rate": 0.0001475785499622271, "loss": 0.3669, "step": 42000 }, { "epoch": 0.72, "learning_rate": 0.00014752087586971605, "loss": 0.3673, "step": 42500 }, { "epoch": 0.73, "learning_rate": 0.0001474626527191014, "loss": 0.3668, "step": 43000 }, { "epoch": 0.74, "learning_rate": 0.00014740364772589337, "loss": 0.3666, "step": 43500 }, { "epoch": 0.75, "learning_rate": 0.00014734397748589426, "loss": 0.3665, "step": 44000 }, { "epoch": 0.76, "learning_rate": 0.0001472836425880252, "loss": 0.366, "step": 44500 }, { "epoch": 0.77, "learning_rate": 0.00014722264362776731, "loss": 0.3657, "step": 45000 }, { "epoch": 0.77, "learning_rate": 0.0001471609812071557, "loss": 0.3657, "step": 45500 }, { "epoch": 0.78, "learning_rate": 0.00014709865593477356, "loss": 0.3653, "step": 46000 }, { "epoch": 0.79, "learning_rate": 0.00014703566842574615, "loss": 0.3652, "step": 46500 }, { "epoch": 0.8, "learning_rate": 0.0001469721472598581, "loss": 0.3645, "step": 47000 }, { "epoch": 0.81, "learning_rate": 0.00014690783847039622, "loss": 0.3647, "step": 47500 }, { "epoch": 0.82, "learning_rate": 0.00014684286932758046, "loss": 0.3643, "step": 48000 }, { "epoch": 0.83, "learning_rate": 0.00014677724047262996, "loss": 0.3644, "step": 48500 }, { "epoch": 0.83, "learning_rate": 0.0001467110857864263, "loss": 0.3641, "step": 49000 }, { "epoch": 0.84, "learning_rate": 0.00014664414077306508, "loss": 0.3636, "step": 49500 }, { "epoch": 0.85, "learning_rate": 0.0001465765380089398, "loss": 0.3635, "step": 50000 }, { "epoch": 0.86, "learning_rate": 0.00014650841533628042, "loss": 0.3635, "step": 50500 }, { "epoch": 0.87, "learning_rate": 0.00014643950039089107, "loss": 0.3631, "step": 51000 }, { "epoch": 0.88, "learning_rate": 0.0001463699297144556, "loss": 0.3628, "step": 51500 }, { "epoch": 0.89, "learning_rate": 0.00014629970399360846, "loss": 0.363, "step": 52000 }, { "epoch": 0.89, "learning_rate": 0.00014622882392144917, "loss": 0.3624, "step": 52500 }, { "epoch": 0.9, "learning_rate": 0.00014615729019753547, "loss": 0.3623, "step": 53000 }, { "epoch": 0.91, "learning_rate": 0.00014608510352787625, "loss": 0.3623, "step": 53500 }, { "epoch": 0.92, "learning_rate": 0.0001460122646249249, "loss": 0.3621, "step": 54000 }, { "epoch": 0.93, "learning_rate": 0.00014593892183813694, "loss": 0.3617, "step": 54500 }, { "epoch": 0.94, "learning_rate": 0.00014586478193255307, "loss": 0.3617, "step": 55000 }, { "epoch": 0.94, "learning_rate": 0.00014578999196816224, "loss": 0.3616, "step": 55500 }, { "epoch": 0.95, "learning_rate": 0.00014571455268311119, "loss": 0.361, "step": 56000 }, { "epoch": 0.96, "learning_rate": 0.00014563861764445836, "loss": 0.3613, "step": 56500 }, { "epoch": 0.97, "learning_rate": 0.00014556188325305048, "loss": 0.3607, "step": 57000 }, { "epoch": 0.98, "learning_rate": 0.00014548450179232327, "loss": 0.3606, "step": 57500 }, { "epoch": 0.99, "learning_rate": 0.00014540647402600055, "loss": 0.3607, "step": 58000 }, { "epoch": 1.0, "learning_rate": 0.00014532780072418487, "loss": 0.3602, "step": 58500 }, { "epoch": 1.0, "learning_rate": 0.00014524864194242183, "loss": 0.3607, "step": 59000 }, { "epoch": 1.01, "learning_rate": 0.00014516868119257174, "loss": 0.3599, "step": 59500 }, { "epoch": 1.02, "learning_rate": 0.00014508823910339927, "loss": 0.3599, "step": 60000 }, { "epoch": 1.03, "learning_rate": 0.00014500699405591802, "loss": 0.3599, "step": 60500 }, { "epoch": 1.04, "learning_rate": 0.00014492510741564985, "loss": 0.3595, "step": 61000 }, { "epoch": 1.05, "learning_rate": 0.00014484257999078278, "loss": 0.3593, "step": 61500 }, { "epoch": 1.06, "learning_rate": 0.00014475941259582925, "loss": 0.3595, "step": 62000 }, { "epoch": 1.06, "learning_rate": 0.0001446756060516178, "loss": 0.359, "step": 62500 }, { "epoch": 1.07, "learning_rate": 0.00014459116118528524, "loss": 0.3589, "step": 63000 }, { "epoch": 1.08, "learning_rate": 0.00014450607883026827, "loss": 0.3586, "step": 63500 }, { "epoch": 1.09, "learning_rate": 0.00014442053189911774, "loss": 0.3587, "step": 64000 }, { "epoch": 1.1, "learning_rate": 0.00014433417836295833, "loss": 0.3583, "step": 64500 }, { "epoch": 1.11, "learning_rate": 0.00014424718987443127, "loss": 0.3583, "step": 65000 }, { "epoch": 1.11, "learning_rate": 0.00014415956729207775, "loss": 0.3582, "step": 65500 }, { "epoch": 1.12, "learning_rate": 0.00014407131148069734, "loss": 0.3579, "step": 66000 }, { "epoch": 1.13, "learning_rate": 0.00014398260171818916, "loss": 0.3578, "step": 66500 }, { "epoch": 1.14, "learning_rate": 0.00014389308333022545, "loss": 0.3576, "step": 67000 }, { "epoch": 1.15, "learning_rate": 0.00014380311527004513, "loss": 0.3577, "step": 67500 }, { "epoch": 1.16, "learning_rate": 0.00014371233783246623, "loss": 0.3579, "step": 68000 }, { "epoch": 1.17, "learning_rate": 0.00014362093157983466, "loss": 0.3572, "step": 68500 }, { "epoch": 1.17, "learning_rate": 0.0001435288974142932, "loss": 0.3575, "step": 69000 }, { "epoch": 1.18, "learning_rate": 0.00014343623624418192, "loss": 0.3571, "step": 69500 }, { "epoch": 1.19, "learning_rate": 0.00014334294898402916, "loss": 0.3567, "step": 70000 }, { "epoch": 1.2, "learning_rate": 0.0001432490365545425, "loss": 0.3567, "step": 70500 }, { "epoch": 1.21, "learning_rate": 0.00014315449988259974, "loss": 0.3566, "step": 71000 }, { "epoch": 1.22, "learning_rate": 0.0001430593399012396, "loss": 0.3564, "step": 71500 }, { "epoch": 1.23, "learning_rate": 0.00014296374973485408, "loss": 0.3561, "step": 72000 }, { "epoch": 1.23, "learning_rate": 0.0001428675406248976, "loss": 0.356, "step": 72500 }, { "epoch": 1.24, "learning_rate": 0.00014277051885498018, "loss": 0.3562, "step": 73000 }, { "epoch": 1.25, "learning_rate": 0.00014267287756538312, "loss": 0.3557, "step": 73500 }, { "epoch": 1.26, "learning_rate": 0.00014257461771978656, "loss": 0.3557, "step": 74000 }, { "epoch": 1.27, "learning_rate": 0.0001424757402879755, "loss": 0.3556, "step": 74500 }, { "epoch": 1.28, "learning_rate": 0.0001423762462458303, "loss": 0.3552, "step": 75000 }, { "epoch": 1.29, "learning_rate": 0.00014227613657531694, "loss": 0.3552, "step": 75500 }, { "epoch": 1.29, "learning_rate": 0.00014217541226447747, "loss": 0.3551, "step": 76000 }, { "epoch": 1.3, "learning_rate": 0.00014207427759508945, "loss": 0.3552, "step": 76500 }, { "epoch": 1.31, "learning_rate": 0.000141972328216269, "loss": 0.355, "step": 77000 }, { "epoch": 1.32, "learning_rate": 0.0001418697671955883, "loss": 0.3546, "step": 77500 }, { "epoch": 1.33, "learning_rate": 0.00014176659554528318, "loss": 0.3546, "step": 78000 }, { "epoch": 1.34, "learning_rate": 0.00014166302245385192, "loss": 0.3544, "step": 78500 }, { "epoch": 1.34, "learning_rate": 0.00014155863382125022, "loss": 0.3545, "step": 79000 }, { "epoch": 1.35, "learning_rate": 0.00014145363762978493, "loss": 0.3544, "step": 79500 }, { "epoch": 1.36, "learning_rate": 0.00014134803491572606, "loss": 0.3542, "step": 80000 }, { "epoch": 1.37, "learning_rate": 0.00014124203974129223, "loss": 0.3539, "step": 80500 }, { "epoch": 1.38, "learning_rate": 0.00014113522832260466, "loss": 0.3541, "step": 81000 }, { "epoch": 1.39, "learning_rate": 0.00014102781352389485, "loss": 0.3537, "step": 81500 }, { "epoch": 1.4, "learning_rate": 0.00014091979640530345, "loss": 0.3535, "step": 82000 }, { "epoch": 1.4, "learning_rate": 0.00014081139586900036, "loss": 0.3533, "step": 82500 }, { "epoch": 1.41, "learning_rate": 0.0001407021785141258, "loss": 0.3535, "step": 83000 }, { "epoch": 1.42, "learning_rate": 0.00014059236205325574, "loss": 0.3536, "step": 83500 }, { "epoch": 1.43, "learning_rate": 0.00014048194757023435, "loss": 0.353, "step": 84000 }, { "epoch": 1.44, "learning_rate": 0.00014037115877265014, "loss": 0.353, "step": 84500 }, { "epoch": 1.45, "learning_rate": 0.00014025955271103284, "loss": 0.3529, "step": 85000 }, { "epoch": 1.46, "learning_rate": 0.00014014735191195807, "loss": 0.3528, "step": 85500 }, { "epoch": 1.46, "learning_rate": 0.00014003478366336457, "loss": 0.3527, "step": 86000 }, { "epoch": 1.47, "learning_rate": 0.00013992253468820287, "loss": 0.3559, "step": 86500 }, { "epoch": 1.48, "learning_rate": 0.00013980856343455598, "loss": 0.3566, "step": 87000 }, { "epoch": 1.49, "learning_rate": 0.0001396940018945337, "loss": 0.3554, "step": 87500 }, { "epoch": 1.5, "learning_rate": 0.00013957885119881218, "loss": 0.3552, "step": 88000 }, { "epoch": 1.51, "learning_rate": 0.00013946311248388228, "loss": 0.3548, "step": 88500 }, { "epoch": 1.51, "learning_rate": 0.00013934678689203836, "loss": 0.3547, "step": 89000 }, { "epoch": 1.52, "learning_rate": 0.0001392298755713671, "loss": 0.3547, "step": 89500 }, { "epoch": 1.53, "learning_rate": 0.00013911237967573605, "loss": 0.3541, "step": 90000 }, { "epoch": 1.54, "learning_rate": 0.0001389943003647822, "loss": 0.3539, "step": 90500 }, { "epoch": 1.55, "learning_rate": 0.00013887563880390073, "loss": 0.3536, "step": 91000 }, { "epoch": 1.56, "learning_rate": 0.0001387563961642333, "loss": 0.3537, "step": 91500 }, { "epoch": 1.57, "learning_rate": 0.00013863657362265668, "loss": 0.3549, "step": 92000 }, { "epoch": 1.57, "learning_rate": 0.0001385161723617709, "loss": 0.3539, "step": 92500 }, { "epoch": 1.58, "learning_rate": 0.00013839519356988788, "loss": 0.3535, "step": 93000 }, { "epoch": 1.59, "learning_rate": 0.0001382736384410194, "loss": 0.3537, "step": 93500 }, { "epoch": 1.6, "learning_rate": 0.00013815150817486554, "loss": 0.3527, "step": 94000 }, { "epoch": 1.61, "learning_rate": 0.0001380288039768027, "loss": 0.3529, "step": 94500 }, { "epoch": 1.62, "learning_rate": 0.00013790552705787178, "loss": 0.353, "step": 95000 }, { "epoch": 1.63, "learning_rate": 0.0001377816786347662, "loss": 0.3526, "step": 95500 }, { "epoch": 1.63, "learning_rate": 0.00013765725992981992, "loss": 0.3527, "step": 96000 }, { "epoch": 1.64, "learning_rate": 0.00013753227217099532, "loss": 0.3528, "step": 96500 }, { "epoch": 1.65, "learning_rate": 0.0001374067165918711, "loss": 0.3524, "step": 97000 }, { "epoch": 1.66, "learning_rate": 0.00013728059443163013, "loss": 0.352, "step": 97500 }, { "epoch": 1.67, "learning_rate": 0.00013715390693504724, "loss": 0.3519, "step": 98000 }, { "epoch": 1.68, "learning_rate": 0.00013702665535247687, "loss": 0.3519, "step": 98500 }, { "epoch": 1.69, "learning_rate": 0.00013689884093984076, "loss": 0.3519, "step": 99000 }, { "epoch": 1.69, "learning_rate": 0.00013677072227018223, "loss": 0.3526, "step": 99500 }, { "epoch": 1.7, "learning_rate": 0.00013664178710672204, "loss": 0.3517, "step": 100000 } ], "max_steps": 500000, "num_train_epochs": 9, "total_flos": 0.0, "trial_name": null, "trial_params": null }