| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 90669, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005514563963427412, | |
| "grad_norm": 4.689566135406494, | |
| "learning_rate": 4.972427180182863e-05, | |
| "loss": 6.5185, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.011029127926854823, | |
| "grad_norm": 4.601550102233887, | |
| "learning_rate": 4.944854360365726e-05, | |
| "loss": 5.5109, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.016543691890282236, | |
| "grad_norm": 4.204834461212158, | |
| "learning_rate": 4.917281540548589e-05, | |
| "loss": 5.1513, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.022058255853709647, | |
| "grad_norm": 4.12762451171875, | |
| "learning_rate": 4.889708720731452e-05, | |
| "loss": 4.9528, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.027572819817137058, | |
| "grad_norm": 4.254393100738525, | |
| "learning_rate": 4.862135900914315e-05, | |
| "loss": 4.7961, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.03308738378056447, | |
| "grad_norm": 4.859085559844971, | |
| "learning_rate": 4.834563081097178e-05, | |
| "loss": 4.7004, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.03860194774399188, | |
| "grad_norm": 4.5257368087768555, | |
| "learning_rate": 4.806990261280041e-05, | |
| "loss": 4.5864, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.044116511707419294, | |
| "grad_norm": 4.193604946136475, | |
| "learning_rate": 4.779417441462904e-05, | |
| "loss": 4.5161, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.04963107567084671, | |
| "grad_norm": 4.394799709320068, | |
| "learning_rate": 4.7518446216457665e-05, | |
| "loss": 4.4474, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.055145639634274116, | |
| "grad_norm": 4.351430892944336, | |
| "learning_rate": 4.72427180182863e-05, | |
| "loss": 4.3782, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.06066020359770153, | |
| "grad_norm": 4.718796730041504, | |
| "learning_rate": 4.696698982011493e-05, | |
| "loss": 4.3183, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.06617476756112894, | |
| "grad_norm": 4.186001300811768, | |
| "learning_rate": 4.6691261621943555e-05, | |
| "loss": 4.2597, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.07168933152455635, | |
| "grad_norm": 4.420439720153809, | |
| "learning_rate": 4.641553342377218e-05, | |
| "loss": 4.2321, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.07720389548798376, | |
| "grad_norm": 4.082899570465088, | |
| "learning_rate": 4.613980522560081e-05, | |
| "loss": 4.1731, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.08271845945141118, | |
| "grad_norm": 4.149295806884766, | |
| "learning_rate": 4.5864077027429445e-05, | |
| "loss": 4.1479, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.08823302341483859, | |
| "grad_norm": 4.364389419555664, | |
| "learning_rate": 4.558834882925807e-05, | |
| "loss": 4.1119, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.093747587378266, | |
| "grad_norm": 4.409417629241943, | |
| "learning_rate": 4.53126206310867e-05, | |
| "loss": 4.0806, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.09926215134169342, | |
| "grad_norm": 4.639771938323975, | |
| "learning_rate": 4.5036892432915335e-05, | |
| "loss": 4.0431, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.10477671530512082, | |
| "grad_norm": 4.332629203796387, | |
| "learning_rate": 4.476116423474396e-05, | |
| "loss": 4.0352, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.11029127926854823, | |
| "grad_norm": 4.82522439956665, | |
| "learning_rate": 4.448543603657259e-05, | |
| "loss": 3.9858, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.11580584323197565, | |
| "grad_norm": 4.305941104888916, | |
| "learning_rate": 4.420970783840122e-05, | |
| "loss": 3.9731, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.12132040719540306, | |
| "grad_norm": 4.728514194488525, | |
| "learning_rate": 4.393397964022985e-05, | |
| "loss": 3.9417, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.12683497115883047, | |
| "grad_norm": 4.233896732330322, | |
| "learning_rate": 4.365825144205848e-05, | |
| "loss": 3.9396, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.1323495351222579, | |
| "grad_norm": 4.335183143615723, | |
| "learning_rate": 4.338252324388711e-05, | |
| "loss": 3.9091, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.13786409908568528, | |
| "grad_norm": 4.590264797210693, | |
| "learning_rate": 4.3106795045715735e-05, | |
| "loss": 3.8935, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.1433786630491127, | |
| "grad_norm": 4.33479642868042, | |
| "learning_rate": 4.283106684754436e-05, | |
| "loss": 3.8875, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.14889322701254012, | |
| "grad_norm": 4.2722697257995605, | |
| "learning_rate": 4.2555338649373e-05, | |
| "loss": 3.8566, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.15440779097596752, | |
| "grad_norm": 4.284050464630127, | |
| "learning_rate": 4.2279610451201625e-05, | |
| "loss": 3.8433, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.15992235493939494, | |
| "grad_norm": 4.086195945739746, | |
| "learning_rate": 4.200388225303025e-05, | |
| "loss": 3.8381, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.16543691890282236, | |
| "grad_norm": 4.229586124420166, | |
| "learning_rate": 4.172815405485889e-05, | |
| "loss": 3.8148, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.17095148286624975, | |
| "grad_norm": 4.43237829208374, | |
| "learning_rate": 4.1452425856687515e-05, | |
| "loss": 3.7945, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.17646604682967718, | |
| "grad_norm": 4.232430934906006, | |
| "learning_rate": 4.117669765851614e-05, | |
| "loss": 3.8078, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.1819806107931046, | |
| "grad_norm": 5.106810092926025, | |
| "learning_rate": 4.090096946034477e-05, | |
| "loss": 3.7684, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.187495174756532, | |
| "grad_norm": 4.939910411834717, | |
| "learning_rate": 4.0625241262173405e-05, | |
| "loss": 3.7634, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.1930097387199594, | |
| "grad_norm": 4.215509414672852, | |
| "learning_rate": 4.034951306400203e-05, | |
| "loss": 3.7487, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.19852430268338683, | |
| "grad_norm": 4.279122829437256, | |
| "learning_rate": 4.007378486583066e-05, | |
| "loss": 3.7697, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.20403886664681423, | |
| "grad_norm": 4.503846168518066, | |
| "learning_rate": 3.979805666765929e-05, | |
| "loss": 3.7488, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.20955343061024165, | |
| "grad_norm": 3.935098648071289, | |
| "learning_rate": 3.9522328469487916e-05, | |
| "loss": 3.7347, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.21506799457366907, | |
| "grad_norm": 4.217621326446533, | |
| "learning_rate": 3.924660027131655e-05, | |
| "loss": 3.7063, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.22058255853709646, | |
| "grad_norm": 4.404201030731201, | |
| "learning_rate": 3.897087207314518e-05, | |
| "loss": 3.6878, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.22609712250052388, | |
| "grad_norm": 4.507588863372803, | |
| "learning_rate": 3.869514387497381e-05, | |
| "loss": 3.7021, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.2316116864639513, | |
| "grad_norm": 4.7501220703125, | |
| "learning_rate": 3.841941567680244e-05, | |
| "loss": 3.6856, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.2371262504273787, | |
| "grad_norm": 4.5834879875183105, | |
| "learning_rate": 3.814368747863107e-05, | |
| "loss": 3.6799, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.24264081439080612, | |
| "grad_norm": 4.500739574432373, | |
| "learning_rate": 3.7867959280459695e-05, | |
| "loss": 3.6695, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.24815537835423354, | |
| "grad_norm": 4.357424736022949, | |
| "learning_rate": 3.759223108228832e-05, | |
| "loss": 3.6648, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.25366994231766093, | |
| "grad_norm": 4.667726039886475, | |
| "learning_rate": 3.731650288411696e-05, | |
| "loss": 3.6623, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.25918450628108836, | |
| "grad_norm": 4.472695827484131, | |
| "learning_rate": 3.7040774685945585e-05, | |
| "loss": 3.6336, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.2646990702445158, | |
| "grad_norm": 4.226781368255615, | |
| "learning_rate": 3.676504648777421e-05, | |
| "loss": 3.6362, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.2702136342079432, | |
| "grad_norm": 4.575997829437256, | |
| "learning_rate": 3.648931828960284e-05, | |
| "loss": 3.6146, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.27572819817137056, | |
| "grad_norm": 4.648991584777832, | |
| "learning_rate": 3.621359009143147e-05, | |
| "loss": 3.6352, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.281242762134798, | |
| "grad_norm": 4.165131092071533, | |
| "learning_rate": 3.59378618932601e-05, | |
| "loss": 3.6165, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.2867573260982254, | |
| "grad_norm": 4.220915794372559, | |
| "learning_rate": 3.566213369508873e-05, | |
| "loss": 3.6067, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.29227189006165283, | |
| "grad_norm": 4.650350093841553, | |
| "learning_rate": 3.5386405496917365e-05, | |
| "loss": 3.5922, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.29778645402508025, | |
| "grad_norm": 4.175040245056152, | |
| "learning_rate": 3.511067729874599e-05, | |
| "loss": 3.6063, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.30330101798850767, | |
| "grad_norm": 4.40975284576416, | |
| "learning_rate": 3.483494910057462e-05, | |
| "loss": 3.5945, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.30881558195193504, | |
| "grad_norm": 3.9698615074157715, | |
| "learning_rate": 3.455922090240325e-05, | |
| "loss": 3.5887, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.31433014591536246, | |
| "grad_norm": 4.441317081451416, | |
| "learning_rate": 3.4283492704231876e-05, | |
| "loss": 3.5741, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.3198447098787899, | |
| "grad_norm": 4.244263648986816, | |
| "learning_rate": 3.400776450606051e-05, | |
| "loss": 3.5688, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.3253592738422173, | |
| "grad_norm": 4.017004013061523, | |
| "learning_rate": 3.373203630788914e-05, | |
| "loss": 3.568, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.3308738378056447, | |
| "grad_norm": 4.664565563201904, | |
| "learning_rate": 3.3456308109717765e-05, | |
| "loss": 3.5571, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.33638840176907214, | |
| "grad_norm": 4.073508262634277, | |
| "learning_rate": 3.318057991154639e-05, | |
| "loss": 3.5189, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.3419029657324995, | |
| "grad_norm": 4.424101829528809, | |
| "learning_rate": 3.290485171337502e-05, | |
| "loss": 3.554, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.34741752969592693, | |
| "grad_norm": 4.302523136138916, | |
| "learning_rate": 3.2629123515203655e-05, | |
| "loss": 3.5388, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.35293209365935435, | |
| "grad_norm": 4.329090118408203, | |
| "learning_rate": 3.235339531703228e-05, | |
| "loss": 3.5485, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.3584466576227818, | |
| "grad_norm": 4.2849531173706055, | |
| "learning_rate": 3.207766711886092e-05, | |
| "loss": 3.5388, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.3639612215862092, | |
| "grad_norm": 4.334972381591797, | |
| "learning_rate": 3.1801938920689545e-05, | |
| "loss": 3.5225, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.3694757855496366, | |
| "grad_norm": 4.848361492156982, | |
| "learning_rate": 3.152621072251817e-05, | |
| "loss": 3.5226, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.374990349513064, | |
| "grad_norm": 4.436476230621338, | |
| "learning_rate": 3.12504825243468e-05, | |
| "loss": 3.5046, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.3805049134764914, | |
| "grad_norm": 4.017549991607666, | |
| "learning_rate": 3.097475432617543e-05, | |
| "loss": 3.5091, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.3860194774399188, | |
| "grad_norm": 4.507646083831787, | |
| "learning_rate": 3.069902612800406e-05, | |
| "loss": 3.4951, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.39153404140334624, | |
| "grad_norm": 4.1406989097595215, | |
| "learning_rate": 3.042329792983269e-05, | |
| "loss": 3.496, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.39704860536677367, | |
| "grad_norm": 4.320881366729736, | |
| "learning_rate": 3.0147569731661318e-05, | |
| "loss": 3.5025, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.4025631693302011, | |
| "grad_norm": 4.030999183654785, | |
| "learning_rate": 2.9871841533489946e-05, | |
| "loss": 3.4974, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.40807773329362845, | |
| "grad_norm": 4.489917755126953, | |
| "learning_rate": 2.9596113335318577e-05, | |
| "loss": 3.4867, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.4135922972570559, | |
| "grad_norm": 4.384711742401123, | |
| "learning_rate": 2.9320385137147204e-05, | |
| "loss": 3.4753, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.4191068612204833, | |
| "grad_norm": 4.64800500869751, | |
| "learning_rate": 2.9044656938975836e-05, | |
| "loss": 3.4703, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.4246214251839107, | |
| "grad_norm": 4.490517616271973, | |
| "learning_rate": 2.876892874080447e-05, | |
| "loss": 3.471, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.43013598914733814, | |
| "grad_norm": 4.496025085449219, | |
| "learning_rate": 2.8493200542633098e-05, | |
| "loss": 3.4579, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.43565055311076556, | |
| "grad_norm": 4.765578746795654, | |
| "learning_rate": 2.8217472344461725e-05, | |
| "loss": 3.4646, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.4411651170741929, | |
| "grad_norm": 4.592626094818115, | |
| "learning_rate": 2.7941744146290356e-05, | |
| "loss": 3.4656, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.44667968103762035, | |
| "grad_norm": 4.292928695678711, | |
| "learning_rate": 2.7666015948118984e-05, | |
| "loss": 3.4533, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.45219424500104777, | |
| "grad_norm": 4.014820098876953, | |
| "learning_rate": 2.7390287749947612e-05, | |
| "loss": 3.4576, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.4577088089644752, | |
| "grad_norm": 4.129273891448975, | |
| "learning_rate": 2.7114559551776243e-05, | |
| "loss": 3.4509, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.4632233729279026, | |
| "grad_norm": 4.679018497467041, | |
| "learning_rate": 2.683883135360487e-05, | |
| "loss": 3.4382, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.46873793689133, | |
| "grad_norm": 4.382132053375244, | |
| "learning_rate": 2.6563103155433498e-05, | |
| "loss": 3.4435, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.4742525008547574, | |
| "grad_norm": 4.3672380447387695, | |
| "learning_rate": 2.628737495726213e-05, | |
| "loss": 3.4398, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.4797670648181848, | |
| "grad_norm": 4.159623622894287, | |
| "learning_rate": 2.6011646759090757e-05, | |
| "loss": 3.4529, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.48528162878161224, | |
| "grad_norm": 4.100943565368652, | |
| "learning_rate": 2.573591856091939e-05, | |
| "loss": 3.411, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.49079619274503966, | |
| "grad_norm": 4.237346649169922, | |
| "learning_rate": 2.5460190362748023e-05, | |
| "loss": 3.4153, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.4963107567084671, | |
| "grad_norm": 4.697793960571289, | |
| "learning_rate": 2.518446216457665e-05, | |
| "loss": 3.4153, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.5018253206718944, | |
| "grad_norm": 4.274381160736084, | |
| "learning_rate": 2.4908733966405278e-05, | |
| "loss": 3.4228, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.5073398846353219, | |
| "grad_norm": 4.421125411987305, | |
| "learning_rate": 2.463300576823391e-05, | |
| "loss": 3.3963, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.5128544485987493, | |
| "grad_norm": 4.356249809265137, | |
| "learning_rate": 2.4357277570062537e-05, | |
| "loss": 3.4186, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.5183690125621767, | |
| "grad_norm": 4.516757965087891, | |
| "learning_rate": 2.4081549371891164e-05, | |
| "loss": 3.4222, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.5238835765256041, | |
| "grad_norm": 5.137631416320801, | |
| "learning_rate": 2.3805821173719796e-05, | |
| "loss": 3.4121, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.5293981404890316, | |
| "grad_norm": 4.224301338195801, | |
| "learning_rate": 2.3530092975548423e-05, | |
| "loss": 3.3854, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.534912704452459, | |
| "grad_norm": 4.442605972290039, | |
| "learning_rate": 2.3254364777377054e-05, | |
| "loss": 3.3857, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.5404272684158864, | |
| "grad_norm": 4.190525531768799, | |
| "learning_rate": 2.2978636579205685e-05, | |
| "loss": 3.3958, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.5459418323793138, | |
| "grad_norm": 4.089470386505127, | |
| "learning_rate": 2.2702908381034313e-05, | |
| "loss": 3.4038, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 0.5514563963427411, | |
| "grad_norm": 4.37148380279541, | |
| "learning_rate": 2.242718018286294e-05, | |
| "loss": 3.3908, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.5569709603061685, | |
| "grad_norm": 4.484643936157227, | |
| "learning_rate": 2.2151451984691572e-05, | |
| "loss": 3.392, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 0.562485524269596, | |
| "grad_norm": 4.480875492095947, | |
| "learning_rate": 2.18757237865202e-05, | |
| "loss": 3.38, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.5680000882330234, | |
| "grad_norm": 4.415227890014648, | |
| "learning_rate": 2.159999558834883e-05, | |
| "loss": 3.4014, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 0.5735146521964508, | |
| "grad_norm": 4.461363315582275, | |
| "learning_rate": 2.132426739017746e-05, | |
| "loss": 3.3768, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.5790292161598782, | |
| "grad_norm": 4.364917755126953, | |
| "learning_rate": 2.104853919200609e-05, | |
| "loss": 3.3769, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 0.5845437801233057, | |
| "grad_norm": 4.509827613830566, | |
| "learning_rate": 2.0772810993834717e-05, | |
| "loss": 3.3567, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.5900583440867331, | |
| "grad_norm": 4.165256023406982, | |
| "learning_rate": 2.0497082795663348e-05, | |
| "loss": 3.3649, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 0.5955729080501605, | |
| "grad_norm": 4.39963436126709, | |
| "learning_rate": 2.0221354597491976e-05, | |
| "loss": 3.3688, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.6010874720135879, | |
| "grad_norm": 4.492909908294678, | |
| "learning_rate": 1.9945626399320607e-05, | |
| "loss": 3.3654, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 0.6066020359770153, | |
| "grad_norm": 4.136989593505859, | |
| "learning_rate": 1.9669898201149238e-05, | |
| "loss": 3.3588, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.6121165999404428, | |
| "grad_norm": 4.091104030609131, | |
| "learning_rate": 1.9394170002977866e-05, | |
| "loss": 3.3714, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 0.6176311639038701, | |
| "grad_norm": 4.557612895965576, | |
| "learning_rate": 1.9118441804806493e-05, | |
| "loss": 3.3783, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.6231457278672975, | |
| "grad_norm": 4.4669718742370605, | |
| "learning_rate": 1.8842713606635124e-05, | |
| "loss": 3.3576, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 0.6286602918307249, | |
| "grad_norm": 4.214612007141113, | |
| "learning_rate": 1.8566985408463752e-05, | |
| "loss": 3.3769, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.6341748557941523, | |
| "grad_norm": 4.079827785491943, | |
| "learning_rate": 1.8291257210292383e-05, | |
| "loss": 3.3584, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 0.6396894197575798, | |
| "grad_norm": 4.0199713706970215, | |
| "learning_rate": 1.8015529012121014e-05, | |
| "loss": 3.3593, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.6452039837210072, | |
| "grad_norm": 4.746074199676514, | |
| "learning_rate": 1.7739800813949642e-05, | |
| "loss": 3.3417, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 0.6507185476844346, | |
| "grad_norm": 4.219590187072754, | |
| "learning_rate": 1.746407261577827e-05, | |
| "loss": 3.3518, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.656233111647862, | |
| "grad_norm": 4.15669584274292, | |
| "learning_rate": 1.71883444176069e-05, | |
| "loss": 3.3401, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 0.6617476756112894, | |
| "grad_norm": 4.129217147827148, | |
| "learning_rate": 1.6912616219435532e-05, | |
| "loss": 3.3504, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.6672622395747169, | |
| "grad_norm": 4.176223278045654, | |
| "learning_rate": 1.663688802126416e-05, | |
| "loss": 3.3429, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 0.6727768035381443, | |
| "grad_norm": 3.982861042022705, | |
| "learning_rate": 1.636115982309279e-05, | |
| "loss": 3.3239, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 0.6782913675015716, | |
| "grad_norm": 4.495360851287842, | |
| "learning_rate": 1.6085431624921418e-05, | |
| "loss": 3.334, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 0.683805931464999, | |
| "grad_norm": 4.5026679039001465, | |
| "learning_rate": 1.5809703426750046e-05, | |
| "loss": 3.3434, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 0.6893204954284264, | |
| "grad_norm": 4.469930648803711, | |
| "learning_rate": 1.5533975228578677e-05, | |
| "loss": 3.3085, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 0.6948350593918539, | |
| "grad_norm": 4.942314147949219, | |
| "learning_rate": 1.5258247030407308e-05, | |
| "loss": 3.3169, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 0.7003496233552813, | |
| "grad_norm": 4.131747245788574, | |
| "learning_rate": 1.4982518832235937e-05, | |
| "loss": 3.3242, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 0.7058641873187087, | |
| "grad_norm": 4.662265777587891, | |
| "learning_rate": 1.4706790634064565e-05, | |
| "loss": 3.3455, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 0.7113787512821361, | |
| "grad_norm": 4.53313684463501, | |
| "learning_rate": 1.4431062435893195e-05, | |
| "loss": 3.3093, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 0.7168933152455635, | |
| "grad_norm": 4.306014537811279, | |
| "learning_rate": 1.4155334237721824e-05, | |
| "loss": 3.3146, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 0.722407879208991, | |
| "grad_norm": 4.205687999725342, | |
| "learning_rate": 1.3879606039550452e-05, | |
| "loss": 3.3182, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 0.7279224431724184, | |
| "grad_norm": 4.351266384124756, | |
| "learning_rate": 1.3603877841379084e-05, | |
| "loss": 3.3363, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 0.7334370071358458, | |
| "grad_norm": 4.580765724182129, | |
| "learning_rate": 1.3328149643207714e-05, | |
| "loss": 3.2988, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 0.7389515710992732, | |
| "grad_norm": 4.511965274810791, | |
| "learning_rate": 1.3052421445036341e-05, | |
| "loss": 3.3153, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 0.7444661350627005, | |
| "grad_norm": 4.1504950523376465, | |
| "learning_rate": 1.277669324686497e-05, | |
| "loss": 3.3193, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 0.749980699026128, | |
| "grad_norm": 4.668148994445801, | |
| "learning_rate": 1.25009650486936e-05, | |
| "loss": 3.2916, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 0.7554952629895554, | |
| "grad_norm": 4.106932163238525, | |
| "learning_rate": 1.222523685052223e-05, | |
| "loss": 3.3268, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 0.7610098269529828, | |
| "grad_norm": 4.127325057983398, | |
| "learning_rate": 1.1949508652350859e-05, | |
| "loss": 3.3036, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 0.7665243909164102, | |
| "grad_norm": 4.444930076599121, | |
| "learning_rate": 1.1673780454179488e-05, | |
| "loss": 3.3024, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 0.7720389548798376, | |
| "grad_norm": 4.07660436630249, | |
| "learning_rate": 1.1398052256008118e-05, | |
| "loss": 3.3053, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.7775535188432651, | |
| "grad_norm": 4.658594131469727, | |
| "learning_rate": 1.1122324057836747e-05, | |
| "loss": 3.3048, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 0.7830680828066925, | |
| "grad_norm": 4.439772129058838, | |
| "learning_rate": 1.0846595859665378e-05, | |
| "loss": 3.2943, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 0.7885826467701199, | |
| "grad_norm": 4.101642608642578, | |
| "learning_rate": 1.0570867661494006e-05, | |
| "loss": 3.2914, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 0.7940972107335473, | |
| "grad_norm": 4.650053024291992, | |
| "learning_rate": 1.0295139463322635e-05, | |
| "loss": 3.2977, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 0.7996117746969748, | |
| "grad_norm": 4.005575180053711, | |
| "learning_rate": 1.0019411265151266e-05, | |
| "loss": 3.2971, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 0.8051263386604022, | |
| "grad_norm": 4.499767780303955, | |
| "learning_rate": 9.743683066979894e-06, | |
| "loss": 3.3002, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 0.8106409026238295, | |
| "grad_norm": 4.143964767456055, | |
| "learning_rate": 9.467954868808523e-06, | |
| "loss": 3.2879, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 0.8161554665872569, | |
| "grad_norm": 4.027842998504639, | |
| "learning_rate": 9.192226670637155e-06, | |
| "loss": 3.2974, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 0.8216700305506843, | |
| "grad_norm": 4.330503463745117, | |
| "learning_rate": 8.916498472465782e-06, | |
| "loss": 3.2971, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 0.8271845945141117, | |
| "grad_norm": 4.108890056610107, | |
| "learning_rate": 8.640770274294412e-06, | |
| "loss": 3.2951, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 0.8326991584775392, | |
| "grad_norm": 4.396561622619629, | |
| "learning_rate": 8.365042076123043e-06, | |
| "loss": 3.3018, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 0.8382137224409666, | |
| "grad_norm": 4.230642795562744, | |
| "learning_rate": 8.08931387795167e-06, | |
| "loss": 3.2832, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 0.843728286404394, | |
| "grad_norm": 4.438147068023682, | |
| "learning_rate": 7.8135856797803e-06, | |
| "loss": 3.2845, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 0.8492428503678214, | |
| "grad_norm": 4.5078325271606445, | |
| "learning_rate": 7.53785748160893e-06, | |
| "loss": 3.2977, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 0.8547574143312489, | |
| "grad_norm": 4.344171524047852, | |
| "learning_rate": 7.262129283437559e-06, | |
| "loss": 3.2785, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 0.8602719782946763, | |
| "grad_norm": 4.636903762817383, | |
| "learning_rate": 6.986401085266188e-06, | |
| "loss": 3.2784, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 0.8657865422581037, | |
| "grad_norm": 4.955584526062012, | |
| "learning_rate": 6.710672887094818e-06, | |
| "loss": 3.2676, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 0.8713011062215311, | |
| "grad_norm": 4.11828088760376, | |
| "learning_rate": 6.4349446889234475e-06, | |
| "loss": 3.2901, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 0.8768156701849584, | |
| "grad_norm": 4.916619777679443, | |
| "learning_rate": 6.159216490752077e-06, | |
| "loss": 3.2964, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 0.8823302341483859, | |
| "grad_norm": 4.217592239379883, | |
| "learning_rate": 5.883488292580705e-06, | |
| "loss": 3.2923, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.8878447981118133, | |
| "grad_norm": 4.360821723937988, | |
| "learning_rate": 5.607760094409336e-06, | |
| "loss": 3.2807, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 0.8933593620752407, | |
| "grad_norm": 4.062866687774658, | |
| "learning_rate": 5.332031896237965e-06, | |
| "loss": 3.2936, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 0.8988739260386681, | |
| "grad_norm": 4.373243808746338, | |
| "learning_rate": 5.0563036980665936e-06, | |
| "loss": 3.2592, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 0.9043884900020955, | |
| "grad_norm": 4.527072906494141, | |
| "learning_rate": 4.780575499895224e-06, | |
| "loss": 3.2754, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 0.909903053965523, | |
| "grad_norm": 4.206862926483154, | |
| "learning_rate": 4.504847301723853e-06, | |
| "loss": 3.2609, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 0.9154176179289504, | |
| "grad_norm": 4.414552688598633, | |
| "learning_rate": 4.229119103552482e-06, | |
| "loss": 3.2936, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 0.9209321818923778, | |
| "grad_norm": 4.722365856170654, | |
| "learning_rate": 3.953390905381112e-06, | |
| "loss": 3.2733, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 0.9264467458558052, | |
| "grad_norm": 4.194797515869141, | |
| "learning_rate": 3.6776627072097413e-06, | |
| "loss": 3.272, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 0.9319613098192326, | |
| "grad_norm": 4.476502895355225, | |
| "learning_rate": 3.4019345090383703e-06, | |
| "loss": 3.286, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 0.93747587378266, | |
| "grad_norm": 4.418745994567871, | |
| "learning_rate": 3.1262063108669997e-06, | |
| "loss": 3.2611, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 0.9429904377460874, | |
| "grad_norm": 4.2887797355651855, | |
| "learning_rate": 2.8504781126956295e-06, | |
| "loss": 3.2749, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 0.9485050017095148, | |
| "grad_norm": 4.276843547821045, | |
| "learning_rate": 2.574749914524259e-06, | |
| "loss": 3.2739, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 0.9540195656729422, | |
| "grad_norm": 4.448587417602539, | |
| "learning_rate": 2.299021716352888e-06, | |
| "loss": 3.2756, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 0.9595341296363696, | |
| "grad_norm": 4.589288711547852, | |
| "learning_rate": 2.0232935181815176e-06, | |
| "loss": 3.2734, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 0.9650486935997971, | |
| "grad_norm": 4.560390472412109, | |
| "learning_rate": 1.7475653200101468e-06, | |
| "loss": 3.2652, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 0.9705632575632245, | |
| "grad_norm": 4.633021831512451, | |
| "learning_rate": 1.4718371218387762e-06, | |
| "loss": 3.2831, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 0.9760778215266519, | |
| "grad_norm": 4.438389301300049, | |
| "learning_rate": 1.1961089236674058e-06, | |
| "loss": 3.2584, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 0.9815923854900793, | |
| "grad_norm": 4.40911340713501, | |
| "learning_rate": 9.203807254960352e-07, | |
| "loss": 3.2831, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 0.9871069494535067, | |
| "grad_norm": 4.270487308502197, | |
| "learning_rate": 6.446525273246646e-07, | |
| "loss": 3.2702, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 0.9926215134169342, | |
| "grad_norm": 4.149001598358154, | |
| "learning_rate": 3.689243291532939e-07, | |
| "loss": 3.2504, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.9981360773803616, | |
| "grad_norm": 4.047245502471924, | |
| "learning_rate": 9.319613098192327e-08, | |
| "loss": 3.2622, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 90669, | |
| "total_flos": 2.4124793739319296e+16, | |
| "train_loss": 3.571538051929227, | |
| "train_runtime": 3663.4284, | |
| "train_samples_per_second": 197.997, | |
| "train_steps_per_second": 24.75 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 90669, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.4124793739319296e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |