| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.481637567730283, |
| "eval_steps": 200, |
| "global_step": 2000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0012040939193257074, |
| "grad_norm": 1.2087849378585815, |
| "learning_rate": 1.9975915221579964e-05, |
| "loss": 1.7501, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.002408187838651415, |
| "grad_norm": 1.0137544870376587, |
| "learning_rate": 1.9951830443159926e-05, |
| "loss": 1.3294, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.003612281757977122, |
| "grad_norm": 0.8906710147857666, |
| "learning_rate": 1.9927745664739885e-05, |
| "loss": 1.4793, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.00481637567730283, |
| "grad_norm": 0.8175792098045349, |
| "learning_rate": 1.9903660886319848e-05, |
| "loss": 1.4537, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.006020469596628537, |
| "grad_norm": 0.8326422572135925, |
| "learning_rate": 1.987957610789981e-05, |
| "loss": 1.5579, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.007224563515954244, |
| "grad_norm": 0.8402993083000183, |
| "learning_rate": 1.985549132947977e-05, |
| "loss": 1.4146, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.008428657435279952, |
| "grad_norm": 0.9211377501487732, |
| "learning_rate": 1.983140655105973e-05, |
| "loss": 1.4986, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.00963275135460566, |
| "grad_norm": 0.7729477286338806, |
| "learning_rate": 1.9807321772639694e-05, |
| "loss": 1.3949, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.010836845273931367, |
| "grad_norm": 0.8256559371948242, |
| "learning_rate": 1.9783236994219656e-05, |
| "loss": 1.2517, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.012040939193257074, |
| "grad_norm": 0.7404232025146484, |
| "learning_rate": 1.9759152215799615e-05, |
| "loss": 1.4761, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.013245033112582781, |
| "grad_norm": 0.773309588432312, |
| "learning_rate": 1.9735067437379577e-05, |
| "loss": 1.227, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.014449127031908489, |
| "grad_norm": 0.7535562515258789, |
| "learning_rate": 1.971098265895954e-05, |
| "loss": 1.3443, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.015653220951234198, |
| "grad_norm": 0.8561655282974243, |
| "learning_rate": 1.9686897880539502e-05, |
| "loss": 1.3229, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.016857314870559904, |
| "grad_norm": 0.7888396978378296, |
| "learning_rate": 1.966281310211946e-05, |
| "loss": 1.4274, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.018061408789885613, |
| "grad_norm": 0.7886033654212952, |
| "learning_rate": 1.9638728323699423e-05, |
| "loss": 1.4469, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.01926550270921132, |
| "grad_norm": 0.8100034594535828, |
| "learning_rate": 1.9614643545279385e-05, |
| "loss": 1.314, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.020469596628537028, |
| "grad_norm": 0.7786366939544678, |
| "learning_rate": 1.9590558766859348e-05, |
| "loss": 1.2631, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.021673690547862733, |
| "grad_norm": 0.8259329795837402, |
| "learning_rate": 1.9566473988439307e-05, |
| "loss": 1.4006, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.022877784467188442, |
| "grad_norm": 0.7402159571647644, |
| "learning_rate": 1.954238921001927e-05, |
| "loss": 1.416, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.024081878386514148, |
| "grad_norm": 0.8619882464408875, |
| "learning_rate": 1.951830443159923e-05, |
| "loss": 1.2583, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.025285972305839857, |
| "grad_norm": 0.7436647415161133, |
| "learning_rate": 1.949421965317919e-05, |
| "loss": 1.3527, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.026490066225165563, |
| "grad_norm": 0.7261799573898315, |
| "learning_rate": 1.9470134874759156e-05, |
| "loss": 1.4234, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.027694160144491272, |
| "grad_norm": 0.8648309707641602, |
| "learning_rate": 1.9446050096339115e-05, |
| "loss": 1.2541, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.028898254063816978, |
| "grad_norm": 0.7091889381408691, |
| "learning_rate": 1.9421965317919077e-05, |
| "loss": 1.2473, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.030102347983142687, |
| "grad_norm": 0.7450091242790222, |
| "learning_rate": 1.939788053949904e-05, |
| "loss": 1.2555, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.031306441902468396, |
| "grad_norm": 0.8190895318984985, |
| "learning_rate": 1.9373795761079e-05, |
| "loss": 1.2874, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.0325105358217941, |
| "grad_norm": 0.7458487749099731, |
| "learning_rate": 1.934971098265896e-05, |
| "loss": 1.4139, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.03371462974111981, |
| "grad_norm": 0.806698203086853, |
| "learning_rate": 1.9325626204238923e-05, |
| "loss": 1.4061, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.034918723660445516, |
| "grad_norm": 0.7941407561302185, |
| "learning_rate": 1.9301541425818882e-05, |
| "loss": 1.2819, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.036122817579771226, |
| "grad_norm": 0.6824349164962769, |
| "learning_rate": 1.9277456647398845e-05, |
| "loss": 1.3025, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.03732691149909693, |
| "grad_norm": 0.8056986927986145, |
| "learning_rate": 1.9253371868978807e-05, |
| "loss": 1.1617, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.03853100541842264, |
| "grad_norm": 0.7486716508865356, |
| "learning_rate": 1.922928709055877e-05, |
| "loss": 1.3523, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.039735099337748346, |
| "grad_norm": 0.7773532867431641, |
| "learning_rate": 1.920520231213873e-05, |
| "loss": 1.3013, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.040939193257074055, |
| "grad_norm": 0.7498481869697571, |
| "learning_rate": 1.918111753371869e-05, |
| "loss": 1.2563, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.04214328717639976, |
| "grad_norm": 0.7634468078613281, |
| "learning_rate": 1.9157032755298653e-05, |
| "loss": 1.2365, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.04334738109572547, |
| "grad_norm": 0.6561328172683716, |
| "learning_rate": 1.9132947976878615e-05, |
| "loss": 1.3058, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.044551475015051176, |
| "grad_norm": 0.7443628907203674, |
| "learning_rate": 1.9108863198458578e-05, |
| "loss": 1.3264, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.045755568934376885, |
| "grad_norm": 0.6697310209274292, |
| "learning_rate": 1.9084778420038536e-05, |
| "loss": 1.1029, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.04695966285370259, |
| "grad_norm": 0.7119889259338379, |
| "learning_rate": 1.90606936416185e-05, |
| "loss": 1.2503, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.048163756773028296, |
| "grad_norm": 0.8120700120925903, |
| "learning_rate": 1.903660886319846e-05, |
| "loss": 1.246, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.048163756773028296, |
| "eval_loss": 1.2944953441619873, |
| "eval_runtime": 10.5151, |
| "eval_samples_per_second": 9.51, |
| "eval_steps_per_second": 1.236, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.049367850692354005, |
| "grad_norm": 0.7876719832420349, |
| "learning_rate": 1.901252408477842e-05, |
| "loss": 1.2097, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.050571944611679714, |
| "grad_norm": 0.7367354035377502, |
| "learning_rate": 1.8988439306358382e-05, |
| "loss": 1.3332, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.05177603853100542, |
| "grad_norm": 0.7384558320045471, |
| "learning_rate": 1.8964354527938345e-05, |
| "loss": 1.3897, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.052980132450331126, |
| "grad_norm": 0.8261380195617676, |
| "learning_rate": 1.8940269749518304e-05, |
| "loss": 1.2677, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.054184226369656835, |
| "grad_norm": 0.7237825989723206, |
| "learning_rate": 1.891618497109827e-05, |
| "loss": 1.2919, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.055388320288982544, |
| "grad_norm": 0.7296762466430664, |
| "learning_rate": 1.889210019267823e-05, |
| "loss": 1.2977, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.056592414208308246, |
| "grad_norm": 0.8224279880523682, |
| "learning_rate": 1.886801541425819e-05, |
| "loss": 1.3751, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.057796508127633955, |
| "grad_norm": 0.7233418822288513, |
| "learning_rate": 1.8843930635838153e-05, |
| "loss": 1.293, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.059000602046959665, |
| "grad_norm": 0.7015657424926758, |
| "learning_rate": 1.8819845857418112e-05, |
| "loss": 1.3093, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.060204695966285374, |
| "grad_norm": 0.8122280240058899, |
| "learning_rate": 1.8795761078998074e-05, |
| "loss": 1.3979, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.061408789885611076, |
| "grad_norm": 0.7582533955574036, |
| "learning_rate": 1.8771676300578037e-05, |
| "loss": 1.3147, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.06261288380493679, |
| "grad_norm": 0.746967077255249, |
| "learning_rate": 1.8747591522158e-05, |
| "loss": 1.1785, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.0638169777242625, |
| "grad_norm": 0.7588161826133728, |
| "learning_rate": 1.8723506743737958e-05, |
| "loss": 1.3648, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.0650210716435882, |
| "grad_norm": 0.696869432926178, |
| "learning_rate": 1.869942196531792e-05, |
| "loss": 1.2854, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.06622516556291391, |
| "grad_norm": 0.8297455906867981, |
| "learning_rate": 1.8675337186897883e-05, |
| "loss": 1.2011, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.06742925948223961, |
| "grad_norm": 0.7183229327201843, |
| "learning_rate": 1.8651252408477845e-05, |
| "loss": 1.1944, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.06863335340156532, |
| "grad_norm": 0.7092033624649048, |
| "learning_rate": 1.8627167630057804e-05, |
| "loss": 1.3296, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.06983744732089103, |
| "grad_norm": 0.7024344801902771, |
| "learning_rate": 1.8603082851637766e-05, |
| "loss": 1.2349, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.07104154124021674, |
| "grad_norm": 0.7504032254219055, |
| "learning_rate": 1.857899807321773e-05, |
| "loss": 1.1986, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.07224563515954245, |
| "grad_norm": 0.7584034204483032, |
| "learning_rate": 1.855491329479769e-05, |
| "loss": 1.2998, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.07344972907886815, |
| "grad_norm": 0.7563251852989197, |
| "learning_rate": 1.853082851637765e-05, |
| "loss": 1.3703, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.07465382299819386, |
| "grad_norm": 0.728660523891449, |
| "learning_rate": 1.8506743737957612e-05, |
| "loss": 1.2916, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.07585791691751957, |
| "grad_norm": 0.6596492528915405, |
| "learning_rate": 1.8482658959537575e-05, |
| "loss": 1.1821, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.07706201083684527, |
| "grad_norm": 0.7137461304664612, |
| "learning_rate": 1.8458574181117533e-05, |
| "loss": 1.1982, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.07826610475617098, |
| "grad_norm": 0.7411230206489563, |
| "learning_rate": 1.8434489402697496e-05, |
| "loss": 1.2711, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.07947019867549669, |
| "grad_norm": 0.7566368579864502, |
| "learning_rate": 1.8410404624277458e-05, |
| "loss": 1.1603, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.0806742925948224, |
| "grad_norm": 0.7235690355300903, |
| "learning_rate": 1.838631984585742e-05, |
| "loss": 1.1993, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.08187838651414811, |
| "grad_norm": 0.7648451328277588, |
| "learning_rate": 1.8362235067437383e-05, |
| "loss": 1.134, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.08308248043347381, |
| "grad_norm": 0.740710973739624, |
| "learning_rate": 1.8338150289017342e-05, |
| "loss": 1.1533, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.08428657435279951, |
| "grad_norm": 0.7610028982162476, |
| "learning_rate": 1.8314065510597304e-05, |
| "loss": 1.3126, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.08549066827212523, |
| "grad_norm": 0.6956422924995422, |
| "learning_rate": 1.8289980732177266e-05, |
| "loss": 1.2272, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.08669476219145093, |
| "grad_norm": 0.6914700269699097, |
| "learning_rate": 1.8265895953757225e-05, |
| "loss": 1.3049, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.08789885611077664, |
| "grad_norm": 0.7015173435211182, |
| "learning_rate": 1.8241811175337188e-05, |
| "loss": 1.3569, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.08910295003010235, |
| "grad_norm": 0.6985088586807251, |
| "learning_rate": 1.821772639691715e-05, |
| "loss": 1.2307, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.09030704394942805, |
| "grad_norm": 0.7313199043273926, |
| "learning_rate": 1.8193641618497112e-05, |
| "loss": 1.1783, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.09151113786875377, |
| "grad_norm": 0.7657830715179443, |
| "learning_rate": 1.8169556840077075e-05, |
| "loss": 1.2228, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.09271523178807947, |
| "grad_norm": 0.8051377534866333, |
| "learning_rate": 1.8145472061657034e-05, |
| "loss": 1.25, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.09391932570740517, |
| "grad_norm": 0.8114392161369324, |
| "learning_rate": 1.8121387283236996e-05, |
| "loss": 1.2347, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.09512341962673089, |
| "grad_norm": 0.7304120659828186, |
| "learning_rate": 1.809730250481696e-05, |
| "loss": 1.2437, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.09632751354605659, |
| "grad_norm": 0.8418581485748291, |
| "learning_rate": 1.8073217726396917e-05, |
| "loss": 1.3255, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.09632751354605659, |
| "eval_loss": 1.255526065826416, |
| "eval_runtime": 10.4675, |
| "eval_samples_per_second": 9.553, |
| "eval_steps_per_second": 1.242, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.0975316074653823, |
| "grad_norm": 0.7186573147773743, |
| "learning_rate": 1.804913294797688e-05, |
| "loss": 1.2859, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.09873570138470801, |
| "grad_norm": 0.7544155716896057, |
| "learning_rate": 1.8025048169556842e-05, |
| "loss": 1.2771, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.09993979530403371, |
| "grad_norm": 0.700529158115387, |
| "learning_rate": 1.8000963391136804e-05, |
| "loss": 1.2696, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.10114388922335943, |
| "grad_norm": 0.6894243955612183, |
| "learning_rate": 1.7976878612716763e-05, |
| "loss": 1.2556, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.10234798314268513, |
| "grad_norm": 0.8769268989562988, |
| "learning_rate": 1.7952793834296726e-05, |
| "loss": 1.2832, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.10355207706201083, |
| "grad_norm": 0.7283303141593933, |
| "learning_rate": 1.7928709055876688e-05, |
| "loss": 1.3163, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.10475617098133655, |
| "grad_norm": 0.6984531283378601, |
| "learning_rate": 1.7904624277456647e-05, |
| "loss": 1.1873, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.10596026490066225, |
| "grad_norm": 0.7543785572052002, |
| "learning_rate": 1.7880539499036613e-05, |
| "loss": 1.2617, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.10716435881998795, |
| "grad_norm": 0.7481861710548401, |
| "learning_rate": 1.785645472061657e-05, |
| "loss": 1.1792, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.10836845273931367, |
| "grad_norm": 0.7607905268669128, |
| "learning_rate": 1.7832369942196534e-05, |
| "loss": 1.1987, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.10957254665863937, |
| "grad_norm": 0.7427639365196228, |
| "learning_rate": 1.7808285163776496e-05, |
| "loss": 1.2142, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.11077664057796509, |
| "grad_norm": 0.7348979711532593, |
| "learning_rate": 1.7784200385356455e-05, |
| "loss": 1.2727, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.11198073449729079, |
| "grad_norm": 0.7192914485931396, |
| "learning_rate": 1.7760115606936417e-05, |
| "loss": 1.2593, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.11318482841661649, |
| "grad_norm": 0.6672971248626709, |
| "learning_rate": 1.773603082851638e-05, |
| "loss": 1.1905, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.11438892233594221, |
| "grad_norm": 0.7585046291351318, |
| "learning_rate": 1.771194605009634e-05, |
| "loss": 1.3045, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.11559301625526791, |
| "grad_norm": 0.7200559973716736, |
| "learning_rate": 1.76878612716763e-05, |
| "loss": 1.1416, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.11679711017459361, |
| "grad_norm": 0.7220486998558044, |
| "learning_rate": 1.7663776493256263e-05, |
| "loss": 1.2749, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.11800120409391933, |
| "grad_norm": 0.750998318195343, |
| "learning_rate": 1.7639691714836226e-05, |
| "loss": 1.2823, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.11920529801324503, |
| "grad_norm": 0.705752432346344, |
| "learning_rate": 1.7615606936416188e-05, |
| "loss": 1.1662, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.12040939193257075, |
| "grad_norm": 0.7431082129478455, |
| "learning_rate": 1.7591522157996147e-05, |
| "loss": 1.203, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.12161348585189645, |
| "grad_norm": 0.7106168866157532, |
| "learning_rate": 1.756743737957611e-05, |
| "loss": 1.162, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.12281757977122215, |
| "grad_norm": 0.7027086615562439, |
| "learning_rate": 1.754335260115607e-05, |
| "loss": 1.2636, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.12402167369054787, |
| "grad_norm": 0.774591863155365, |
| "learning_rate": 1.7519267822736034e-05, |
| "loss": 1.2517, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.12522576760987358, |
| "grad_norm": 0.6927476525306702, |
| "learning_rate": 1.7495183044315993e-05, |
| "loss": 1.1396, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.12642986152919927, |
| "grad_norm": 0.6912556886672974, |
| "learning_rate": 1.7471098265895955e-05, |
| "loss": 1.1658, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.127633955448525, |
| "grad_norm": 0.6708123683929443, |
| "learning_rate": 1.7447013487475918e-05, |
| "loss": 1.1288, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.1288380493678507, |
| "grad_norm": 0.6942213177680969, |
| "learning_rate": 1.7422928709055877e-05, |
| "loss": 1.1174, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.1300421432871764, |
| "grad_norm": 0.7080583572387695, |
| "learning_rate": 1.739884393063584e-05, |
| "loss": 1.2279, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.1312462372065021, |
| "grad_norm": 0.8131110668182373, |
| "learning_rate": 1.73747591522158e-05, |
| "loss": 1.2525, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.13245033112582782, |
| "grad_norm": 0.7450209259986877, |
| "learning_rate": 1.735067437379576e-05, |
| "loss": 1.1506, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.1336544250451535, |
| "grad_norm": 0.7066523432731628, |
| "learning_rate": 1.7326589595375726e-05, |
| "loss": 1.2383, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.13485851896447923, |
| "grad_norm": 0.7041836977005005, |
| "learning_rate": 1.7302504816955685e-05, |
| "loss": 1.1911, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.13606261288380495, |
| "grad_norm": 0.672095000743866, |
| "learning_rate": 1.7278420038535647e-05, |
| "loss": 1.1839, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.13726670680313063, |
| "grad_norm": 0.7928294539451599, |
| "learning_rate": 1.725433526011561e-05, |
| "loss": 1.3347, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.13847080072245635, |
| "grad_norm": 0.7465523481369019, |
| "learning_rate": 1.723025048169557e-05, |
| "loss": 1.134, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.13967489464178207, |
| "grad_norm": 0.7117074131965637, |
| "learning_rate": 1.720616570327553e-05, |
| "loss": 1.3538, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.14087898856110775, |
| "grad_norm": 0.7291346192359924, |
| "learning_rate": 1.7182080924855493e-05, |
| "loss": 1.1339, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.14208308248043347, |
| "grad_norm": 0.7633514404296875, |
| "learning_rate": 1.7157996146435455e-05, |
| "loss": 1.2007, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.1432871763997592, |
| "grad_norm": 0.7896651029586792, |
| "learning_rate": 1.7133911368015418e-05, |
| "loss": 1.2656, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.1444912703190849, |
| "grad_norm": 0.7436801791191101, |
| "learning_rate": 1.7109826589595377e-05, |
| "loss": 1.28, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.1444912703190849, |
| "eval_loss": 1.23267662525177, |
| "eval_runtime": 10.4568, |
| "eval_samples_per_second": 9.563, |
| "eval_steps_per_second": 1.243, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.1456953642384106, |
| "grad_norm": 0.7123108506202698, |
| "learning_rate": 1.708574181117534e-05, |
| "loss": 1.1687, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.1468994581577363, |
| "grad_norm": 0.7362635731697083, |
| "learning_rate": 1.70616570327553e-05, |
| "loss": 1.2471, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.14810355207706202, |
| "grad_norm": 0.6770139336585999, |
| "learning_rate": 1.703757225433526e-05, |
| "loss": 1.149, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.1493076459963877, |
| "grad_norm": 0.7292876243591309, |
| "learning_rate": 1.7013487475915223e-05, |
| "loss": 1.2957, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.15051173991571343, |
| "grad_norm": 0.7105087637901306, |
| "learning_rate": 1.6989402697495185e-05, |
| "loss": 1.1462, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.15171583383503914, |
| "grad_norm": 0.7218934297561646, |
| "learning_rate": 1.6965317919075147e-05, |
| "loss": 1.2181, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.15291992775436483, |
| "grad_norm": 0.7794796824455261, |
| "learning_rate": 1.6941233140655106e-05, |
| "loss": 1.174, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.15412402167369055, |
| "grad_norm": 0.6424360871315002, |
| "learning_rate": 1.691714836223507e-05, |
| "loss": 1.2149, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.15532811559301626, |
| "grad_norm": 0.7393337488174438, |
| "learning_rate": 1.689306358381503e-05, |
| "loss": 1.2259, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.15653220951234195, |
| "grad_norm": 0.7363811731338501, |
| "learning_rate": 1.686897880539499e-05, |
| "loss": 1.2087, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.15773630343166767, |
| "grad_norm": 0.7336999773979187, |
| "learning_rate": 1.6844894026974952e-05, |
| "loss": 1.2526, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.15894039735099338, |
| "grad_norm": 0.7329117655754089, |
| "learning_rate": 1.6820809248554915e-05, |
| "loss": 1.233, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.16014449127031907, |
| "grad_norm": 0.7724847197532654, |
| "learning_rate": 1.6796724470134877e-05, |
| "loss": 1.2506, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.1613485851896448, |
| "grad_norm": 0.7213071584701538, |
| "learning_rate": 1.677263969171484e-05, |
| "loss": 1.2021, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.1625526791089705, |
| "grad_norm": 0.6397632360458374, |
| "learning_rate": 1.6748554913294798e-05, |
| "loss": 1.1909, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.16375677302829622, |
| "grad_norm": 0.6840227842330933, |
| "learning_rate": 1.672447013487476e-05, |
| "loss": 1.1838, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.1649608669476219, |
| "grad_norm": 0.7317889332771301, |
| "learning_rate": 1.6700385356454723e-05, |
| "loss": 1.169, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.16616496086694763, |
| "grad_norm": 0.7134217619895935, |
| "learning_rate": 1.6676300578034682e-05, |
| "loss": 1.2711, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.16736905478627334, |
| "grad_norm": 0.7525413036346436, |
| "learning_rate": 1.6652215799614644e-05, |
| "loss": 1.2183, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.16857314870559903, |
| "grad_norm": 0.7243443131446838, |
| "learning_rate": 1.6628131021194607e-05, |
| "loss": 1.2585, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.16977724262492475, |
| "grad_norm": 0.7554174661636353, |
| "learning_rate": 1.660404624277457e-05, |
| "loss": 1.1662, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.17098133654425046, |
| "grad_norm": 0.6472408175468445, |
| "learning_rate": 1.657996146435453e-05, |
| "loss": 1.1471, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.17218543046357615, |
| "grad_norm": 0.7514950037002563, |
| "learning_rate": 1.655587668593449e-05, |
| "loss": 1.157, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.17338952438290187, |
| "grad_norm": 0.6952060461044312, |
| "learning_rate": 1.6531791907514452e-05, |
| "loss": 1.2981, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.17459361830222758, |
| "grad_norm": 0.832417905330658, |
| "learning_rate": 1.6507707129094415e-05, |
| "loss": 1.2457, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.17579771222155327, |
| "grad_norm": 0.6825481653213501, |
| "learning_rate": 1.6483622350674374e-05, |
| "loss": 1.2632, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.177001806140879, |
| "grad_norm": 0.8267092704772949, |
| "learning_rate": 1.6459537572254336e-05, |
| "loss": 1.0368, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.1782059000602047, |
| "grad_norm": 0.7782288789749146, |
| "learning_rate": 1.64354527938343e-05, |
| "loss": 1.201, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.1794099939795304, |
| "grad_norm": 0.7548500299453735, |
| "learning_rate": 1.641136801541426e-05, |
| "loss": 1.1831, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.1806140878988561, |
| "grad_norm": 0.6344689726829529, |
| "learning_rate": 1.638728323699422e-05, |
| "loss": 1.1985, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.18181818181818182, |
| "grad_norm": 0.7326412796974182, |
| "learning_rate": 1.6363198458574182e-05, |
| "loss": 1.3051, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.18302227573750754, |
| "grad_norm": 0.6460621953010559, |
| "learning_rate": 1.6339113680154144e-05, |
| "loss": 1.1524, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.18422636965683323, |
| "grad_norm": 0.6921172142028809, |
| "learning_rate": 1.6315028901734103e-05, |
| "loss": 1.29, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.18543046357615894, |
| "grad_norm": 0.7872111201286316, |
| "learning_rate": 1.629094412331407e-05, |
| "loss": 1.1187, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.18663455749548466, |
| "grad_norm": 0.7098067402839661, |
| "learning_rate": 1.6266859344894028e-05, |
| "loss": 1.2006, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.18783865141481035, |
| "grad_norm": 0.7701388001441956, |
| "learning_rate": 1.624277456647399e-05, |
| "loss": 1.1609, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.18904274533413606, |
| "grad_norm": 0.7027246952056885, |
| "learning_rate": 1.6218689788053953e-05, |
| "loss": 1.2083, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.19024683925346178, |
| "grad_norm": 0.7161267995834351, |
| "learning_rate": 1.619460500963391e-05, |
| "loss": 1.131, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.19145093317278747, |
| "grad_norm": 0.8034841418266296, |
| "learning_rate": 1.6170520231213874e-05, |
| "loss": 1.2107, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.19265502709211318, |
| "grad_norm": 0.7797010540962219, |
| "learning_rate": 1.6146435452793836e-05, |
| "loss": 1.2931, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.19265502709211318, |
| "eval_loss": 1.2178008556365967, |
| "eval_runtime": 10.4492, |
| "eval_samples_per_second": 9.57, |
| "eval_steps_per_second": 1.244, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.1938591210114389, |
| "grad_norm": 0.6586723923683167, |
| "learning_rate": 1.6122350674373795e-05, |
| "loss": 1.2007, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.1950632149307646, |
| "grad_norm": 0.7271411418914795, |
| "learning_rate": 1.6098265895953758e-05, |
| "loss": 1.1049, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.1962673088500903, |
| "grad_norm": 0.743730902671814, |
| "learning_rate": 1.607418111753372e-05, |
| "loss": 1.1815, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.19747140276941602, |
| "grad_norm": 0.7163010835647583, |
| "learning_rate": 1.6050096339113682e-05, |
| "loss": 1.1635, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.1986754966887417, |
| "grad_norm": 0.7846854329109192, |
| "learning_rate": 1.6026011560693645e-05, |
| "loss": 1.1576, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.19987959060806743, |
| "grad_norm": 0.755066990852356, |
| "learning_rate": 1.6001926782273604e-05, |
| "loss": 1.2626, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.20108368452739314, |
| "grad_norm": 0.6981719732284546, |
| "learning_rate": 1.5977842003853566e-05, |
| "loss": 1.1883, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.20228777844671886, |
| "grad_norm": 0.712697446346283, |
| "learning_rate": 1.5953757225433528e-05, |
| "loss": 1.1539, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.20349187236604455, |
| "grad_norm": 0.8146225810050964, |
| "learning_rate": 1.592967244701349e-05, |
| "loss": 1.161, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.20469596628537026, |
| "grad_norm": 0.7796756625175476, |
| "learning_rate": 1.590558766859345e-05, |
| "loss": 1.1536, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.20590006020469598, |
| "grad_norm": 0.7026313543319702, |
| "learning_rate": 1.5881502890173412e-05, |
| "loss": 1.2206, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.20710415412402167, |
| "grad_norm": 0.6918179988861084, |
| "learning_rate": 1.5857418111753374e-05, |
| "loss": 1.0569, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.20830824804334738, |
| "grad_norm": 0.7711547613143921, |
| "learning_rate": 1.5833333333333333e-05, |
| "loss": 1.1826, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.2095123419626731, |
| "grad_norm": 0.6604825854301453, |
| "learning_rate": 1.5809248554913295e-05, |
| "loss": 1.1743, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.2107164358819988, |
| "grad_norm": 0.7115917205810547, |
| "learning_rate": 1.5785163776493258e-05, |
| "loss": 1.2815, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.2119205298013245, |
| "grad_norm": 0.69574373960495, |
| "learning_rate": 1.5761078998073217e-05, |
| "loss": 1.2131, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.21312462372065022, |
| "grad_norm": 0.6801597476005554, |
| "learning_rate": 1.5736994219653182e-05, |
| "loss": 1.3022, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.2143287176399759, |
| "grad_norm": 0.7042235136032104, |
| "learning_rate": 1.571290944123314e-05, |
| "loss": 1.1936, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.21553281155930162, |
| "grad_norm": 0.6610032916069031, |
| "learning_rate": 1.5688824662813104e-05, |
| "loss": 1.1972, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.21673690547862734, |
| "grad_norm": 0.8218846321105957, |
| "learning_rate": 1.5664739884393066e-05, |
| "loss": 1.2951, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.21794099939795303, |
| "grad_norm": 0.7200918197631836, |
| "learning_rate": 1.5640655105973025e-05, |
| "loss": 1.1256, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.21914509331727874, |
| "grad_norm": 0.6200802326202393, |
| "learning_rate": 1.5616570327552987e-05, |
| "loss": 1.2071, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.22034918723660446, |
| "grad_norm": 0.7431606650352478, |
| "learning_rate": 1.559248554913295e-05, |
| "loss": 1.2569, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.22155328115593018, |
| "grad_norm": 0.802537202835083, |
| "learning_rate": 1.5568400770712912e-05, |
| "loss": 1.1342, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.22275737507525586, |
| "grad_norm": 0.7078563570976257, |
| "learning_rate": 1.5544315992292874e-05, |
| "loss": 1.1928, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.22396146899458158, |
| "grad_norm": 0.6626926064491272, |
| "learning_rate": 1.5520231213872833e-05, |
| "loss": 1.1295, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.2251655629139073, |
| "grad_norm": 0.664715588092804, |
| "learning_rate": 1.5496146435452796e-05, |
| "loss": 1.1616, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.22636965683323299, |
| "grad_norm": 0.7077795267105103, |
| "learning_rate": 1.5472061657032758e-05, |
| "loss": 1.24, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.2275737507525587, |
| "grad_norm": 0.6813915371894836, |
| "learning_rate": 1.5447976878612717e-05, |
| "loss": 1.1807, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.22877784467188442, |
| "grad_norm": 0.7245080471038818, |
| "learning_rate": 1.542389210019268e-05, |
| "loss": 1.1737, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.2299819385912101, |
| "grad_norm": 0.7851933836936951, |
| "learning_rate": 1.539980732177264e-05, |
| "loss": 1.0702, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.23118603251053582, |
| "grad_norm": 0.7153802514076233, |
| "learning_rate": 1.5375722543352604e-05, |
| "loss": 1.1302, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.23239012642986154, |
| "grad_norm": 0.8112071752548218, |
| "learning_rate": 1.5351637764932563e-05, |
| "loss": 1.1591, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.23359422034918723, |
| "grad_norm": 0.7842305898666382, |
| "learning_rate": 1.5327552986512525e-05, |
| "loss": 1.1695, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.23479831426851294, |
| "grad_norm": 0.6949911713600159, |
| "learning_rate": 1.5303468208092487e-05, |
| "loss": 1.1726, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.23600240818783866, |
| "grad_norm": 0.7541187405586243, |
| "learning_rate": 1.5279383429672446e-05, |
| "loss": 1.2209, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.23720650210716435, |
| "grad_norm": 0.7082781791687012, |
| "learning_rate": 1.525529865125241e-05, |
| "loss": 1.2685, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.23841059602649006, |
| "grad_norm": 0.6862682104110718, |
| "learning_rate": 1.5231213872832371e-05, |
| "loss": 1.1569, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.23961468994581578, |
| "grad_norm": 0.836452066898346, |
| "learning_rate": 1.5207129094412332e-05, |
| "loss": 1.2158, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.2408187838651415, |
| "grad_norm": 0.8637534379959106, |
| "learning_rate": 1.5183044315992294e-05, |
| "loss": 1.1891, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.2408187838651415, |
| "eval_loss": 1.2066378593444824, |
| "eval_runtime": 10.4478, |
| "eval_samples_per_second": 9.571, |
| "eval_steps_per_second": 1.244, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.24202287778446718, |
| "grad_norm": 0.8397049903869629, |
| "learning_rate": 1.5158959537572255e-05, |
| "loss": 1.2199, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.2432269717037929, |
| "grad_norm": 0.7865377068519592, |
| "learning_rate": 1.5134874759152215e-05, |
| "loss": 1.2102, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.24443106562311862, |
| "grad_norm": 0.6602733731269836, |
| "learning_rate": 1.511078998073218e-05, |
| "loss": 1.1512, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.2456351595424443, |
| "grad_norm": 0.69524747133255, |
| "learning_rate": 1.508670520231214e-05, |
| "loss": 1.2387, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.24683925346177002, |
| "grad_norm": 0.6567270755767822, |
| "learning_rate": 1.50626204238921e-05, |
| "loss": 1.159, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.24804334738109574, |
| "grad_norm": 0.6918655633926392, |
| "learning_rate": 1.5038535645472063e-05, |
| "loss": 1.236, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.24924744130042142, |
| "grad_norm": 0.748435378074646, |
| "learning_rate": 1.5014450867052024e-05, |
| "loss": 1.1422, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.25045153521974717, |
| "grad_norm": 0.8211308717727661, |
| "learning_rate": 1.4990366088631986e-05, |
| "loss": 1.1775, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.25165562913907286, |
| "grad_norm": 0.6989510655403137, |
| "learning_rate": 1.4966281310211948e-05, |
| "loss": 1.2473, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.25285972305839854, |
| "grad_norm": 0.6981227993965149, |
| "learning_rate": 1.4942196531791909e-05, |
| "loss": 1.1887, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.2540638169777243, |
| "grad_norm": 0.6409928202629089, |
| "learning_rate": 1.4918111753371871e-05, |
| "loss": 1.1323, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.25526791089705, |
| "grad_norm": 0.8026551604270935, |
| "learning_rate": 1.4894026974951832e-05, |
| "loss": 1.2205, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.25647200481637566, |
| "grad_norm": 0.7427927255630493, |
| "learning_rate": 1.4869942196531793e-05, |
| "loss": 1.3051, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.2576760987357014, |
| "grad_norm": 0.7286926507949829, |
| "learning_rate": 1.4845857418111755e-05, |
| "loss": 1.1952, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.2588801926550271, |
| "grad_norm": 0.717553436756134, |
| "learning_rate": 1.4821772639691716e-05, |
| "loss": 1.1395, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.2600842865743528, |
| "grad_norm": 0.6549174189567566, |
| "learning_rate": 1.4797687861271676e-05, |
| "loss": 1.1677, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.26128838049367853, |
| "grad_norm": 0.7400743365287781, |
| "learning_rate": 1.477360308285164e-05, |
| "loss": 1.239, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.2624924744130042, |
| "grad_norm": 0.8405967950820923, |
| "learning_rate": 1.4749518304431601e-05, |
| "loss": 1.1379, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.2636965683323299, |
| "grad_norm": 0.7308910489082336, |
| "learning_rate": 1.4725433526011562e-05, |
| "loss": 1.1709, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.26490066225165565, |
| "grad_norm": 0.7904026508331299, |
| "learning_rate": 1.4701348747591524e-05, |
| "loss": 1.2348, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.26610475617098134, |
| "grad_norm": 0.7531107664108276, |
| "learning_rate": 1.4677263969171484e-05, |
| "loss": 1.1702, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.267308850090307, |
| "grad_norm": 0.7167130708694458, |
| "learning_rate": 1.4653179190751445e-05, |
| "loss": 1.1417, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.26851294400963277, |
| "grad_norm": 0.7529981732368469, |
| "learning_rate": 1.4629094412331407e-05, |
| "loss": 1.3126, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.26971703792895846, |
| "grad_norm": 0.7192036509513855, |
| "learning_rate": 1.4605009633911368e-05, |
| "loss": 1.2149, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.27092113184828415, |
| "grad_norm": 0.7200658321380615, |
| "learning_rate": 1.458092485549133e-05, |
| "loss": 1.1958, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.2721252257676099, |
| "grad_norm": 0.7063027024269104, |
| "learning_rate": 1.4556840077071293e-05, |
| "loss": 1.1266, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.2733293196869356, |
| "grad_norm": 0.7414741516113281, |
| "learning_rate": 1.4532755298651253e-05, |
| "loss": 1.2046, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.27453341360626127, |
| "grad_norm": 0.7233819365501404, |
| "learning_rate": 1.4508670520231216e-05, |
| "loss": 1.3376, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.275737507525587, |
| "grad_norm": 0.737486720085144, |
| "learning_rate": 1.4484585741811176e-05, |
| "loss": 1.2529, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.2769416014449127, |
| "grad_norm": 0.6850925087928772, |
| "learning_rate": 1.4460500963391137e-05, |
| "loss": 1.1784, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.2781456953642384, |
| "grad_norm": 0.6742652058601379, |
| "learning_rate": 1.4436416184971101e-05, |
| "loss": 1.0412, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.27934978928356413, |
| "grad_norm": 0.7630188465118408, |
| "learning_rate": 1.4412331406551062e-05, |
| "loss": 1.2421, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.2805538832028898, |
| "grad_norm": 0.7039279341697693, |
| "learning_rate": 1.4388246628131022e-05, |
| "loss": 1.2076, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.2817579771222155, |
| "grad_norm": 0.6968697905540466, |
| "learning_rate": 1.4364161849710985e-05, |
| "loss": 1.2709, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.28296207104154125, |
| "grad_norm": 0.757686972618103, |
| "learning_rate": 1.4340077071290945e-05, |
| "loss": 1.2237, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.28416616496086694, |
| "grad_norm": 0.70340895652771, |
| "learning_rate": 1.4315992292870906e-05, |
| "loss": 1.1979, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.28537025888019263, |
| "grad_norm": 0.7078365683555603, |
| "learning_rate": 1.4291907514450868e-05, |
| "loss": 1.1801, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.2865743527995184, |
| "grad_norm": 0.671120822429657, |
| "learning_rate": 1.4267822736030829e-05, |
| "loss": 1.0744, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.28777844671884406, |
| "grad_norm": 0.6995368599891663, |
| "learning_rate": 1.424373795761079e-05, |
| "loss": 1.1057, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.2889825406381698, |
| "grad_norm": 0.8009471297264099, |
| "learning_rate": 1.4219653179190754e-05, |
| "loss": 1.2852, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.2889825406381698, |
| "eval_loss": 1.1969135999679565, |
| "eval_runtime": 10.4824, |
| "eval_samples_per_second": 9.54, |
| "eval_steps_per_second": 1.24, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.2901866345574955, |
| "grad_norm": 0.8135602474212646, |
| "learning_rate": 1.4195568400770714e-05, |
| "loss": 1.1475, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.2913907284768212, |
| "grad_norm": 0.6879997849464417, |
| "learning_rate": 1.4171483622350675e-05, |
| "loss": 1.2054, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.2925948223961469, |
| "grad_norm": 0.774534285068512, |
| "learning_rate": 1.4147398843930637e-05, |
| "loss": 1.1133, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.2937989163154726, |
| "grad_norm": 0.7712222933769226, |
| "learning_rate": 1.4123314065510598e-05, |
| "loss": 1.2367, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.2950030102347983, |
| "grad_norm": 0.7632959485054016, |
| "learning_rate": 1.4099229287090558e-05, |
| "loss": 1.2142, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.29620710415412405, |
| "grad_norm": 0.6896396279335022, |
| "learning_rate": 1.4075144508670523e-05, |
| "loss": 1.2692, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.29741119807344973, |
| "grad_norm": 0.7135001420974731, |
| "learning_rate": 1.4051059730250483e-05, |
| "loss": 1.1922, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.2986152919927754, |
| "grad_norm": 0.7016868591308594, |
| "learning_rate": 1.4026974951830444e-05, |
| "loss": 1.2775, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.29981938591210117, |
| "grad_norm": 0.7863327860832214, |
| "learning_rate": 1.4002890173410406e-05, |
| "loss": 1.193, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.30102347983142685, |
| "grad_norm": 0.7253124713897705, |
| "learning_rate": 1.3978805394990367e-05, |
| "loss": 1.1557, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.30222757375075254, |
| "grad_norm": 0.7214757204055786, |
| "learning_rate": 1.3954720616570329e-05, |
| "loss": 1.2212, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.3034316676700783, |
| "grad_norm": 0.7036615014076233, |
| "learning_rate": 1.393063583815029e-05, |
| "loss": 1.3103, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.304635761589404, |
| "grad_norm": 0.7152957320213318, |
| "learning_rate": 1.390655105973025e-05, |
| "loss": 1.3316, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.30583985550872966, |
| "grad_norm": 0.7779828906059265, |
| "learning_rate": 1.3882466281310214e-05, |
| "loss": 1.2773, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.3070439494280554, |
| "grad_norm": 0.7995044589042664, |
| "learning_rate": 1.3858381502890175e-05, |
| "loss": 1.297, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.3082480433473811, |
| "grad_norm": 0.7020440697669983, |
| "learning_rate": 1.3834296724470136e-05, |
| "loss": 1.1593, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.3094521372667068, |
| "grad_norm": 0.7052685022354126, |
| "learning_rate": 1.3810211946050098e-05, |
| "loss": 1.1415, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.3106562311860325, |
| "grad_norm": 0.7097790837287903, |
| "learning_rate": 1.3786127167630059e-05, |
| "loss": 1.0992, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.3118603251053582, |
| "grad_norm": 0.7684621214866638, |
| "learning_rate": 1.376204238921002e-05, |
| "loss": 1.2256, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.3130644190246839, |
| "grad_norm": 0.746864914894104, |
| "learning_rate": 1.3737957610789983e-05, |
| "loss": 1.2155, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.31426851294400965, |
| "grad_norm": 0.733629584312439, |
| "learning_rate": 1.3713872832369944e-05, |
| "loss": 1.1287, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.31547260686333534, |
| "grad_norm": 0.6500353217124939, |
| "learning_rate": 1.3689788053949905e-05, |
| "loss": 1.267, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.316676700782661, |
| "grad_norm": 0.7301894426345825, |
| "learning_rate": 1.3665703275529867e-05, |
| "loss": 1.2167, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.31788079470198677, |
| "grad_norm": 0.677039623260498, |
| "learning_rate": 1.3641618497109828e-05, |
| "loss": 1.1037, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.31908488862131246, |
| "grad_norm": 0.7031332850456238, |
| "learning_rate": 1.3617533718689788e-05, |
| "loss": 1.1252, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.32028898254063815, |
| "grad_norm": 0.8084255456924438, |
| "learning_rate": 1.359344894026975e-05, |
| "loss": 1.1728, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.3214930764599639, |
| "grad_norm": 0.6915790438652039, |
| "learning_rate": 1.3569364161849711e-05, |
| "loss": 1.1552, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.3226971703792896, |
| "grad_norm": 0.742514431476593, |
| "learning_rate": 1.3545279383429672e-05, |
| "loss": 1.1406, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.32390126429861527, |
| "grad_norm": 0.6682133674621582, |
| "learning_rate": 1.3521194605009636e-05, |
| "loss": 1.1642, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.325105358217941, |
| "grad_norm": 0.7231415510177612, |
| "learning_rate": 1.3497109826589597e-05, |
| "loss": 1.253, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.3263094521372667, |
| "grad_norm": 0.7634385824203491, |
| "learning_rate": 1.3473025048169559e-05, |
| "loss": 1.2276, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.32751354605659244, |
| "grad_norm": 0.6695632934570312, |
| "learning_rate": 1.344894026974952e-05, |
| "loss": 1.1864, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.32871763997591813, |
| "grad_norm": 0.7134613394737244, |
| "learning_rate": 1.342485549132948e-05, |
| "loss": 1.1097, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.3299217338952438, |
| "grad_norm": 0.6643304228782654, |
| "learning_rate": 1.3400770712909442e-05, |
| "loss": 1.2438, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.33112582781456956, |
| "grad_norm": 0.6821620464324951, |
| "learning_rate": 1.3376685934489405e-05, |
| "loss": 1.2137, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.33232992173389525, |
| "grad_norm": 0.7436670660972595, |
| "learning_rate": 1.3352601156069365e-05, |
| "loss": 1.0745, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.33353401565322094, |
| "grad_norm": 0.7283557653427124, |
| "learning_rate": 1.3328516377649328e-05, |
| "loss": 1.04, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.3347381095725467, |
| "grad_norm": 0.6368799805641174, |
| "learning_rate": 1.3304431599229288e-05, |
| "loss": 1.1829, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.33594220349187237, |
| "grad_norm": 0.8481898307800293, |
| "learning_rate": 1.3280346820809249e-05, |
| "loss": 1.0901, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.33714629741119806, |
| "grad_norm": 0.7311575412750244, |
| "learning_rate": 1.3256262042389211e-05, |
| "loss": 1.1135, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.33714629741119806, |
| "eval_loss": 1.1897846460342407, |
| "eval_runtime": 10.4886, |
| "eval_samples_per_second": 9.534, |
| "eval_steps_per_second": 1.239, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.3383503913305238, |
| "grad_norm": 0.7002428770065308, |
| "learning_rate": 1.3232177263969172e-05, |
| "loss": 1.1384, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.3395544852498495, |
| "grad_norm": 0.7509122490882874, |
| "learning_rate": 1.3208092485549133e-05, |
| "loss": 1.1667, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.3407585791691752, |
| "grad_norm": 0.7156943082809448, |
| "learning_rate": 1.3184007707129097e-05, |
| "loss": 1.0665, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.3419626730885009, |
| "grad_norm": 0.8622729182243347, |
| "learning_rate": 1.3159922928709057e-05, |
| "loss": 1.1503, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.3431667670078266, |
| "grad_norm": 0.7578678131103516, |
| "learning_rate": 1.3135838150289018e-05, |
| "loss": 1.0761, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.3443708609271523, |
| "grad_norm": 0.7222859263420105, |
| "learning_rate": 1.311175337186898e-05, |
| "loss": 1.2214, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.34557495484647804, |
| "grad_norm": 0.7335066795349121, |
| "learning_rate": 1.3087668593448941e-05, |
| "loss": 1.1857, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.34677904876580373, |
| "grad_norm": 0.6828716397285461, |
| "learning_rate": 1.3063583815028902e-05, |
| "loss": 1.2078, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.3479831426851294, |
| "grad_norm": 0.8298027515411377, |
| "learning_rate": 1.3039499036608864e-05, |
| "loss": 1.2375, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.34918723660445516, |
| "grad_norm": 0.7282932996749878, |
| "learning_rate": 1.3015414258188825e-05, |
| "loss": 1.1631, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.35039133052378085, |
| "grad_norm": 0.6626558303833008, |
| "learning_rate": 1.2991329479768787e-05, |
| "loss": 1.1374, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.35159542444310654, |
| "grad_norm": 0.8166629672050476, |
| "learning_rate": 1.296724470134875e-05, |
| "loss": 1.2521, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.3527995183624323, |
| "grad_norm": 0.688026487827301, |
| "learning_rate": 1.294315992292871e-05, |
| "loss": 1.1008, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.354003612281758, |
| "grad_norm": 0.7461696267127991, |
| "learning_rate": 1.2919075144508672e-05, |
| "loss": 1.0932, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.35520770620108366, |
| "grad_norm": 0.7331961989402771, |
| "learning_rate": 1.2894990366088633e-05, |
| "loss": 1.0613, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.3564118001204094, |
| "grad_norm": 0.7345579862594604, |
| "learning_rate": 1.2870905587668594e-05, |
| "loss": 1.3048, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.3576158940397351, |
| "grad_norm": 0.8133093118667603, |
| "learning_rate": 1.2846820809248558e-05, |
| "loss": 1.2435, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.3588199879590608, |
| "grad_norm": 0.7404915690422058, |
| "learning_rate": 1.2822736030828518e-05, |
| "loss": 1.1767, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.3600240818783865, |
| "grad_norm": 0.8186049461364746, |
| "learning_rate": 1.2798651252408479e-05, |
| "loss": 1.0789, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.3612281757977122, |
| "grad_norm": 0.8047255277633667, |
| "learning_rate": 1.2774566473988441e-05, |
| "loss": 1.1343, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.3624322697170379, |
| "grad_norm": 0.6876709461212158, |
| "learning_rate": 1.2750481695568402e-05, |
| "loss": 1.0948, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.36363636363636365, |
| "grad_norm": 0.7112250328063965, |
| "learning_rate": 1.2726396917148362e-05, |
| "loss": 1.1495, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.36484045755568933, |
| "grad_norm": 0.6787546873092651, |
| "learning_rate": 1.2702312138728325e-05, |
| "loss": 1.2422, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.3660445514750151, |
| "grad_norm": 0.677456796169281, |
| "learning_rate": 1.2678227360308285e-05, |
| "loss": 1.2153, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.36724864539434077, |
| "grad_norm": 0.6819108724594116, |
| "learning_rate": 1.2654142581888246e-05, |
| "loss": 1.0196, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.36845273931366646, |
| "grad_norm": 0.7156027555465698, |
| "learning_rate": 1.263005780346821e-05, |
| "loss": 1.2116, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.3696568332329922, |
| "grad_norm": 0.7217147350311279, |
| "learning_rate": 1.260597302504817e-05, |
| "loss": 1.1021, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.3708609271523179, |
| "grad_norm": 0.6850091814994812, |
| "learning_rate": 1.2581888246628131e-05, |
| "loss": 1.1054, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.3720650210716436, |
| "grad_norm": 0.7076934576034546, |
| "learning_rate": 1.2557803468208094e-05, |
| "loss": 1.1533, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.3732691149909693, |
| "grad_norm": 0.7233012318611145, |
| "learning_rate": 1.2533718689788054e-05, |
| "loss": 1.2163, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.374473208910295, |
| "grad_norm": 0.8073462843894958, |
| "learning_rate": 1.2509633911368015e-05, |
| "loss": 1.2245, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.3756773028296207, |
| "grad_norm": 0.72356116771698, |
| "learning_rate": 1.2485549132947979e-05, |
| "loss": 1.1593, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.37688139674894644, |
| "grad_norm": 0.645554780960083, |
| "learning_rate": 1.246146435452794e-05, |
| "loss": 1.1967, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.37808549066827213, |
| "grad_norm": 0.6589043140411377, |
| "learning_rate": 1.2437379576107902e-05, |
| "loss": 1.2093, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.3792895845875978, |
| "grad_norm": 0.6996996402740479, |
| "learning_rate": 1.2413294797687863e-05, |
| "loss": 1.1498, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.38049367850692356, |
| "grad_norm": 0.6371117234230042, |
| "learning_rate": 1.2389210019267823e-05, |
| "loss": 1.1499, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.38169777242624925, |
| "grad_norm": 0.6780497431755066, |
| "learning_rate": 1.2365125240847786e-05, |
| "loss": 1.1191, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.38290186634557494, |
| "grad_norm": 0.7249863743782043, |
| "learning_rate": 1.2341040462427746e-05, |
| "loss": 1.2254, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.3841059602649007, |
| "grad_norm": 0.6600310206413269, |
| "learning_rate": 1.2316955684007707e-05, |
| "loss": 1.0585, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.38531005418422637, |
| "grad_norm": 0.6883255839347839, |
| "learning_rate": 1.2292870905587671e-05, |
| "loss": 1.1318, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.38531005418422637, |
| "eval_loss": 1.1825724840164185, |
| "eval_runtime": 10.5009, |
| "eval_samples_per_second": 9.523, |
| "eval_steps_per_second": 1.238, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.38651414810355206, |
| "grad_norm": 0.7173585295677185, |
| "learning_rate": 1.2268786127167632e-05, |
| "loss": 1.284, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.3877182420228778, |
| "grad_norm": 0.7643466591835022, |
| "learning_rate": 1.2244701348747592e-05, |
| "loss": 1.3699, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.3889223359422035, |
| "grad_norm": 0.690148115158081, |
| "learning_rate": 1.2220616570327555e-05, |
| "loss": 1.1335, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.3901264298615292, |
| "grad_norm": 0.6676989197731018, |
| "learning_rate": 1.2196531791907515e-05, |
| "loss": 1.1887, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.3913305237808549, |
| "grad_norm": 1.0724517107009888, |
| "learning_rate": 1.2172447013487476e-05, |
| "loss": 1.2212, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.3925346177001806, |
| "grad_norm": 0.6685481071472168, |
| "learning_rate": 1.214836223506744e-05, |
| "loss": 1.0551, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.3937387116195063, |
| "grad_norm": 0.7440733313560486, |
| "learning_rate": 1.21242774566474e-05, |
| "loss": 1.1565, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.39494280553883204, |
| "grad_norm": 0.7820375561714172, |
| "learning_rate": 1.2100192678227361e-05, |
| "loss": 1.0733, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.39614689945815773, |
| "grad_norm": 0.7016002535820007, |
| "learning_rate": 1.2076107899807323e-05, |
| "loss": 1.1914, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.3973509933774834, |
| "grad_norm": 0.8267973065376282, |
| "learning_rate": 1.2052023121387284e-05, |
| "loss": 1.1687, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.39855508729680916, |
| "grad_norm": 0.7492110729217529, |
| "learning_rate": 1.2027938342967245e-05, |
| "loss": 1.2462, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.39975918121613485, |
| "grad_norm": 0.6964198350906372, |
| "learning_rate": 1.2003853564547207e-05, |
| "loss": 1.1537, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.40096327513546054, |
| "grad_norm": 0.7141043543815613, |
| "learning_rate": 1.1979768786127168e-05, |
| "loss": 1.1106, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.4021673690547863, |
| "grad_norm": 0.7486653327941895, |
| "learning_rate": 1.1955684007707128e-05, |
| "loss": 1.1255, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.40337146297411197, |
| "grad_norm": 0.68989098072052, |
| "learning_rate": 1.1931599229287092e-05, |
| "loss": 1.1491, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.4045755568934377, |
| "grad_norm": 0.7117523550987244, |
| "learning_rate": 1.1907514450867053e-05, |
| "loss": 1.2564, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.4057796508127634, |
| "grad_norm": 0.7256306409835815, |
| "learning_rate": 1.1883429672447015e-05, |
| "loss": 1.2544, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.4069837447320891, |
| "grad_norm": 0.6354299187660217, |
| "learning_rate": 1.1859344894026976e-05, |
| "loss": 1.1011, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.40818783865141484, |
| "grad_norm": 0.7653043270111084, |
| "learning_rate": 1.1835260115606937e-05, |
| "loss": 1.1881, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.4093919325707405, |
| "grad_norm": 0.6969402432441711, |
| "learning_rate": 1.1811175337186899e-05, |
| "loss": 1.2509, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.4105960264900662, |
| "grad_norm": 0.7759403586387634, |
| "learning_rate": 1.1787090558766861e-05, |
| "loss": 1.1777, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.41180012040939196, |
| "grad_norm": 0.8177191019058228, |
| "learning_rate": 1.1763005780346822e-05, |
| "loss": 1.2677, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.41300421432871764, |
| "grad_norm": 0.6352664828300476, |
| "learning_rate": 1.1738921001926784e-05, |
| "loss": 1.107, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.41420830824804333, |
| "grad_norm": 0.7194890975952148, |
| "learning_rate": 1.1714836223506745e-05, |
| "loss": 1.1646, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.4154124021673691, |
| "grad_norm": 0.6629904508590698, |
| "learning_rate": 1.1690751445086706e-05, |
| "loss": 1.1938, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.41661649608669477, |
| "grad_norm": 0.6937288641929626, |
| "learning_rate": 1.1666666666666668e-05, |
| "loss": 1.1254, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.41782059000602045, |
| "grad_norm": 0.6822580099105835, |
| "learning_rate": 1.1642581888246629e-05, |
| "loss": 1.1878, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.4190246839253462, |
| "grad_norm": 0.8951644897460938, |
| "learning_rate": 1.161849710982659e-05, |
| "loss": 1.0153, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.4202287778446719, |
| "grad_norm": 0.7985068559646606, |
| "learning_rate": 1.1594412331406553e-05, |
| "loss": 1.2309, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.4214328717639976, |
| "grad_norm": 0.7742875814437866, |
| "learning_rate": 1.1570327552986514e-05, |
| "loss": 1.2868, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.4226369656833233, |
| "grad_norm": 0.7290802001953125, |
| "learning_rate": 1.1546242774566474e-05, |
| "loss": 1.1529, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.423841059602649, |
| "grad_norm": 0.7852470278739929, |
| "learning_rate": 1.1522157996146437e-05, |
| "loss": 1.1758, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.4250451535219747, |
| "grad_norm": 0.6864736676216125, |
| "learning_rate": 1.1498073217726397e-05, |
| "loss": 1.173, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.42624924744130044, |
| "grad_norm": 0.6752791404724121, |
| "learning_rate": 1.1473988439306358e-05, |
| "loss": 1.101, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.4274533413606261, |
| "grad_norm": 0.6282265186309814, |
| "learning_rate": 1.144990366088632e-05, |
| "loss": 1.0218, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.4286574352799518, |
| "grad_norm": 0.7045950889587402, |
| "learning_rate": 1.1425818882466281e-05, |
| "loss": 1.2667, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.42986152919927756, |
| "grad_norm": 0.7136130928993225, |
| "learning_rate": 1.1401734104046245e-05, |
| "loss": 1.2445, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.43106562311860325, |
| "grad_norm": 0.7001535296440125, |
| "learning_rate": 1.1377649325626206e-05, |
| "loss": 1.2076, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.43226971703792894, |
| "grad_norm": 0.7172293066978455, |
| "learning_rate": 1.1353564547206166e-05, |
| "loss": 1.0828, |
| "step": 1795 |
| }, |
| { |
| "epoch": 0.4334738109572547, |
| "grad_norm": 0.7160560488700867, |
| "learning_rate": 1.1329479768786129e-05, |
| "loss": 1.1676, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.4334738109572547, |
| "eval_loss": 1.1778274774551392, |
| "eval_runtime": 10.4751, |
| "eval_samples_per_second": 9.546, |
| "eval_steps_per_second": 1.241, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.43467790487658037, |
| "grad_norm": 0.7364829778671265, |
| "learning_rate": 1.130539499036609e-05, |
| "loss": 1.1337, |
| "step": 1805 |
| }, |
| { |
| "epoch": 0.43588199879590606, |
| "grad_norm": 0.7415775656700134, |
| "learning_rate": 1.128131021194605e-05, |
| "loss": 1.1882, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.4370860927152318, |
| "grad_norm": 0.7825621366500854, |
| "learning_rate": 1.1257225433526014e-05, |
| "loss": 1.1627, |
| "step": 1815 |
| }, |
| { |
| "epoch": 0.4382901866345575, |
| "grad_norm": 0.7862880229949951, |
| "learning_rate": 1.1233140655105975e-05, |
| "loss": 1.2287, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.4394942805538832, |
| "grad_norm": 0.7740750908851624, |
| "learning_rate": 1.1209055876685935e-05, |
| "loss": 1.1488, |
| "step": 1825 |
| }, |
| { |
| "epoch": 0.4406983744732089, |
| "grad_norm": 0.8551940321922302, |
| "learning_rate": 1.1184971098265898e-05, |
| "loss": 1.1894, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.4419024683925346, |
| "grad_norm": 0.7718561291694641, |
| "learning_rate": 1.1160886319845858e-05, |
| "loss": 1.1825, |
| "step": 1835 |
| }, |
| { |
| "epoch": 0.44310656231186035, |
| "grad_norm": 0.7665084600448608, |
| "learning_rate": 1.1136801541425819e-05, |
| "loss": 1.0931, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.44431065623118604, |
| "grad_norm": 0.7155455350875854, |
| "learning_rate": 1.1112716763005781e-05, |
| "loss": 1.2208, |
| "step": 1845 |
| }, |
| { |
| "epoch": 0.44551475015051173, |
| "grad_norm": 0.6576226353645325, |
| "learning_rate": 1.1088631984585742e-05, |
| "loss": 1.1294, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.4467188440698375, |
| "grad_norm": 0.6880900859832764, |
| "learning_rate": 1.1064547206165703e-05, |
| "loss": 1.0646, |
| "step": 1855 |
| }, |
| { |
| "epoch": 0.44792293798916316, |
| "grad_norm": 0.633608341217041, |
| "learning_rate": 1.1040462427745667e-05, |
| "loss": 1.0871, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.44912703190848885, |
| "grad_norm": 0.6906594634056091, |
| "learning_rate": 1.1016377649325627e-05, |
| "loss": 1.1449, |
| "step": 1865 |
| }, |
| { |
| "epoch": 0.4503311258278146, |
| "grad_norm": 0.7329970598220825, |
| "learning_rate": 1.0992292870905588e-05, |
| "loss": 1.1158, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.4515352197471403, |
| "grad_norm": 0.7347983121871948, |
| "learning_rate": 1.096820809248555e-05, |
| "loss": 1.1252, |
| "step": 1875 |
| }, |
| { |
| "epoch": 0.45273931366646597, |
| "grad_norm": 0.7085864543914795, |
| "learning_rate": 1.094412331406551e-05, |
| "loss": 1.1535, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.4539434075857917, |
| "grad_norm": 0.7692695260047913, |
| "learning_rate": 1.0920038535645471e-05, |
| "loss": 1.1632, |
| "step": 1885 |
| }, |
| { |
| "epoch": 0.4551475015051174, |
| "grad_norm": 0.7081477046012878, |
| "learning_rate": 1.0895953757225436e-05, |
| "loss": 1.1269, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.4563515954244431, |
| "grad_norm": 0.663478434085846, |
| "learning_rate": 1.0871868978805396e-05, |
| "loss": 1.1256, |
| "step": 1895 |
| }, |
| { |
| "epoch": 0.45755568934376883, |
| "grad_norm": 0.7210192084312439, |
| "learning_rate": 1.0847784200385358e-05, |
| "loss": 1.1887, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.4587597832630945, |
| "grad_norm": 0.7571423649787903, |
| "learning_rate": 1.0823699421965319e-05, |
| "loss": 1.1451, |
| "step": 1905 |
| }, |
| { |
| "epoch": 0.4599638771824202, |
| "grad_norm": 0.79608154296875, |
| "learning_rate": 1.079961464354528e-05, |
| "loss": 1.1874, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.46116797110174595, |
| "grad_norm": 0.857712984085083, |
| "learning_rate": 1.0775529865125242e-05, |
| "loss": 1.1669, |
| "step": 1915 |
| }, |
| { |
| "epoch": 0.46237206502107164, |
| "grad_norm": 0.644782304763794, |
| "learning_rate": 1.0751445086705203e-05, |
| "loss": 1.1627, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.46357615894039733, |
| "grad_norm": 0.6408756971359253, |
| "learning_rate": 1.0727360308285163e-05, |
| "loss": 1.069, |
| "step": 1925 |
| }, |
| { |
| "epoch": 0.4647802528597231, |
| "grad_norm": 0.6929676532745361, |
| "learning_rate": 1.0703275529865127e-05, |
| "loss": 1.1435, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.46598434677904876, |
| "grad_norm": 0.7538990378379822, |
| "learning_rate": 1.0679190751445088e-05, |
| "loss": 1.2604, |
| "step": 1935 |
| }, |
| { |
| "epoch": 0.46718844069837445, |
| "grad_norm": 0.6711621880531311, |
| "learning_rate": 1.0655105973025049e-05, |
| "loss": 1.1923, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.4683925346177002, |
| "grad_norm": 0.6384220123291016, |
| "learning_rate": 1.0631021194605011e-05, |
| "loss": 1.1269, |
| "step": 1945 |
| }, |
| { |
| "epoch": 0.4695966285370259, |
| "grad_norm": 0.682374119758606, |
| "learning_rate": 1.0606936416184972e-05, |
| "loss": 1.1268, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.4708007224563516, |
| "grad_norm": 0.7387133240699768, |
| "learning_rate": 1.0582851637764932e-05, |
| "loss": 1.1043, |
| "step": 1955 |
| }, |
| { |
| "epoch": 0.4720048163756773, |
| "grad_norm": 0.6632955074310303, |
| "learning_rate": 1.0558766859344896e-05, |
| "loss": 1.1148, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.473208910295003, |
| "grad_norm": 0.6747167706489563, |
| "learning_rate": 1.0534682080924857e-05, |
| "loss": 1.1788, |
| "step": 1965 |
| }, |
| { |
| "epoch": 0.4744130042143287, |
| "grad_norm": 0.6498105525970459, |
| "learning_rate": 1.0510597302504818e-05, |
| "loss": 1.2199, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.47561709813365444, |
| "grad_norm": 0.72532057762146, |
| "learning_rate": 1.048651252408478e-05, |
| "loss": 0.9953, |
| "step": 1975 |
| }, |
| { |
| "epoch": 0.4768211920529801, |
| "grad_norm": 0.7126530408859253, |
| "learning_rate": 1.046242774566474e-05, |
| "loss": 1.1014, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.4780252859723058, |
| "grad_norm": 0.8094469904899597, |
| "learning_rate": 1.0438342967244701e-05, |
| "loss": 1.0567, |
| "step": 1985 |
| }, |
| { |
| "epoch": 0.47922937989163156, |
| "grad_norm": 0.7181186676025391, |
| "learning_rate": 1.0414258188824664e-05, |
| "loss": 1.1077, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.48043347381095725, |
| "grad_norm": 0.6791627407073975, |
| "learning_rate": 1.0390173410404624e-05, |
| "loss": 1.1272, |
| "step": 1995 |
| }, |
| { |
| "epoch": 0.481637567730283, |
| "grad_norm": 0.7202663421630859, |
| "learning_rate": 1.0366088631984585e-05, |
| "loss": 1.139, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.481637567730283, |
| "eval_loss": 1.1731120347976685, |
| "eval_runtime": 10.4779, |
| "eval_samples_per_second": 9.544, |
| "eval_steps_per_second": 1.241, |
| "step": 2000 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 4152, |
| "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": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 9755901083039616.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
