| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.97907949790795, | |
| "eval_steps": 500, | |
| "global_step": 357, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008368200836820083, | |
| "grad_norm": 1.1835436820983887, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 1.0482, | |
| "mean_token_accuracy": 0.8137692213058472, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04184100418410042, | |
| "grad_norm": 1.143718957901001, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 1.1531, | |
| "mean_token_accuracy": 0.7920060083270073, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08368200836820083, | |
| "grad_norm": 0.6482775807380676, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 0.9534, | |
| "mean_token_accuracy": 0.8159618139266968, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12552301255230125, | |
| "grad_norm": 0.5585402250289917, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 0.8057, | |
| "mean_token_accuracy": 0.8243314325809479, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16736401673640167, | |
| "grad_norm": 49.08249282836914, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 0.7266, | |
| "mean_token_accuracy": 0.8348599970340729, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.20920502092050208, | |
| "grad_norm": 0.2956574857234955, | |
| "learning_rate": 0.0001388888888888889, | |
| "loss": 0.5231, | |
| "mean_token_accuracy": 0.8722780823707581, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2510460251046025, | |
| "grad_norm": 0.33143389225006104, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 0.4544, | |
| "mean_token_accuracy": 0.8879955649375916, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2928870292887029, | |
| "grad_norm": 0.35626646876335144, | |
| "learning_rate": 0.00019444444444444446, | |
| "loss": 0.3863, | |
| "mean_token_accuracy": 0.9022103250026703, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.33472803347280333, | |
| "grad_norm": 0.26618582010269165, | |
| "learning_rate": 0.00019992338316586133, | |
| "loss": 0.3502, | |
| "mean_token_accuracy": 0.9123009383678437, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.37656903765690375, | |
| "grad_norm": 0.19758877158164978, | |
| "learning_rate": 0.0001996123284790336, | |
| "loss": 0.33, | |
| "mean_token_accuracy": 0.9169199109077454, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.41841004184100417, | |
| "grad_norm": 0.17389552295207977, | |
| "learning_rate": 0.00019906279151130337, | |
| "loss": 0.3597, | |
| "mean_token_accuracy": 0.9097823560237884, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4602510460251046, | |
| "grad_norm": 0.18199238181114197, | |
| "learning_rate": 0.000198276087911361, | |
| "loss": 0.331, | |
| "mean_token_accuracy": 0.91509507894516, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.502092050209205, | |
| "grad_norm": 0.17495596408843994, | |
| "learning_rate": 0.0001972541011294959, | |
| "loss": 0.3322, | |
| "mean_token_accuracy": 0.9161225736141205, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5439330543933054, | |
| "grad_norm": 0.15940091013908386, | |
| "learning_rate": 0.00019599927790842073, | |
| "loss": 0.3483, | |
| "mean_token_accuracy": 0.9118252336978913, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5857740585774058, | |
| "grad_norm": 0.1603671759366989, | |
| "learning_rate": 0.0001945146224255145, | |
| "loss": 0.3215, | |
| "mean_token_accuracy": 0.9172553181648254, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6276150627615062, | |
| "grad_norm": 0.16840051114559174, | |
| "learning_rate": 0.00019280368910050942, | |
| "loss": 0.3131, | |
| "mean_token_accuracy": 0.9191941678524017, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6694560669456067, | |
| "grad_norm": 0.16477249562740326, | |
| "learning_rate": 0.0001908705740858402, | |
| "loss": 0.3337, | |
| "mean_token_accuracy": 0.9134519279003144, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7112970711297071, | |
| "grad_norm": 0.18231800198554993, | |
| "learning_rate": 0.0001887199054600286, | |
| "loss": 0.288, | |
| "mean_token_accuracy": 0.9250746846199036, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7531380753138075, | |
| "grad_norm": 0.14658205211162567, | |
| "learning_rate": 0.00018635683214758214, | |
| "loss": 0.3291, | |
| "mean_token_accuracy": 0.9156664907932281, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7949790794979079, | |
| "grad_norm": 0.14870484173297882, | |
| "learning_rate": 0.0001837870115919327, | |
| "loss": 0.3056, | |
| "mean_token_accuracy": 0.920881724357605, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8368200836820083, | |
| "grad_norm": 0.1470540314912796, | |
| "learning_rate": 0.0001810165962109283, | |
| "loss": 0.3065, | |
| "mean_token_accuracy": 0.9206145107746124, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8786610878661087, | |
| "grad_norm": 0.1503264307975769, | |
| "learning_rate": 0.00017805221866730458, | |
| "loss": 0.3204, | |
| "mean_token_accuracy": 0.9173631191253662, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9205020920502092, | |
| "grad_norm": 0.15052708983421326, | |
| "learning_rate": 0.0001749009759893999, | |
| "loss": 0.3065, | |
| "mean_token_accuracy": 0.9213023066520691, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9623430962343096, | |
| "grad_norm": 0.13865287601947784, | |
| "learning_rate": 0.00017157041258013074, | |
| "loss": 0.2935, | |
| "mean_token_accuracy": 0.9245201706886291, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.2014499455690384, | |
| "learning_rate": 0.0001680685021549063, | |
| "loss": 0.3392, | |
| "mean_token_accuracy": 0.9127676884333292, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3295583426952362, | |
| "eval_mean_token_accuracy": 0.9158889651298523, | |
| "eval_runtime": 188.907, | |
| "eval_samples_per_second": 1.906, | |
| "eval_steps_per_second": 0.238, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0418410041841004, | |
| "grad_norm": 0.14540784060955048, | |
| "learning_rate": 0.00016440362865172373, | |
| "loss": 0.2842, | |
| "mean_token_accuracy": 0.9250402748584747, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0836820083682008, | |
| "grad_norm": 0.15603987872600555, | |
| "learning_rate": 0.00016058456615914817, | |
| "loss": 0.2592, | |
| "mean_token_accuracy": 0.9306007921695709, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1255230125523012, | |
| "grad_norm": 0.20846955478191376, | |
| "learning_rate": 0.00015662045791023173, | |
| "loss": 0.2977, | |
| "mean_token_accuracy": 0.9211509883403778, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1673640167364017, | |
| "grad_norm": 0.1669706404209137, | |
| "learning_rate": 0.00015252079439266177, | |
| "loss": 0.3082, | |
| "mean_token_accuracy": 0.9184517621994018, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.209205020920502, | |
| "grad_norm": 0.20693209767341614, | |
| "learning_rate": 0.00014829539062754596, | |
| "loss": 0.2919, | |
| "mean_token_accuracy": 0.9223342299461365, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2510460251046025, | |
| "grad_norm": 0.20154564082622528, | |
| "learning_rate": 0.00014395436267123016, | |
| "loss": 0.2819, | |
| "mean_token_accuracy": 0.9249872386455535, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.292887029288703, | |
| "grad_norm": 0.19385474920272827, | |
| "learning_rate": 0.00013950810339640688, | |
| "loss": 0.286, | |
| "mean_token_accuracy": 0.9242500007152558, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3347280334728033, | |
| "grad_norm": 0.18098486959934235, | |
| "learning_rate": 0.00013496725761049638, | |
| "loss": 0.2797, | |
| "mean_token_accuracy": 0.9268094718456268, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3765690376569037, | |
| "grad_norm": 0.1658986210823059, | |
| "learning_rate": 0.00013034269657086992, | |
| "loss": 0.2535, | |
| "mean_token_accuracy": 0.931826388835907, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4184100418410042, | |
| "grad_norm": 0.16022320091724396, | |
| "learning_rate": 0.0001256454919579284, | |
| "loss": 0.2564, | |
| "mean_token_accuracy": 0.9323311626911164, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4602510460251046, | |
| "grad_norm": 0.1953819990158081, | |
| "learning_rate": 0.00012088688936834704, | |
| "loss": 0.2459, | |
| "mean_token_accuracy": 0.9334764361381531, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.502092050209205, | |
| "grad_norm": 0.18393699824810028, | |
| "learning_rate": 0.00011607828139194683, | |
| "loss": 0.286, | |
| "mean_token_accuracy": 0.9224025547504425, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.5439330543933054, | |
| "grad_norm": 0.21144607663154602, | |
| "learning_rate": 0.00011123118033664876, | |
| "loss": 0.2604, | |
| "mean_token_accuracy": 0.9307535707950592, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5857740585774058, | |
| "grad_norm": 0.18700766563415527, | |
| "learning_rate": 0.00010635719066681063, | |
| "loss": 0.2579, | |
| "mean_token_accuracy": 0.9308933317661285, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6276150627615062, | |
| "grad_norm": 0.20680415630340576, | |
| "learning_rate": 0.00010146798122093166, | |
| "loss": 0.2745, | |
| "mean_token_accuracy": 0.9262791752815247, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6694560669456067, | |
| "grad_norm": 0.2134338617324829, | |
| "learning_rate": 9.657525727523897e-05, | |
| "loss": 0.2838, | |
| "mean_token_accuracy": 0.9231680512428284, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.711297071129707, | |
| "grad_norm": 0.18799664080142975, | |
| "learning_rate": 9.169073252003779e-05, | |
| "loss": 0.2694, | |
| "mean_token_accuracy": 0.9275272071361542, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.7531380753138075, | |
| "grad_norm": 0.18628208339214325, | |
| "learning_rate": 8.682610101591814e-05, | |
| "loss": 0.2518, | |
| "mean_token_accuracy": 0.9325793802738189, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.794979079497908, | |
| "grad_norm": 0.18667584657669067, | |
| "learning_rate": 8.19930091969567e-05, | |
| "loss": 0.2671, | |
| "mean_token_accuracy": 0.9276439845561981, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8368200836820083, | |
| "grad_norm": 0.2169790118932724, | |
| "learning_rate": 7.720302798794153e-05, | |
| "loss": 0.2608, | |
| "mean_token_accuracy": 0.9305820167064667, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8786610878661087, | |
| "grad_norm": 0.20957927405834198, | |
| "learning_rate": 7.246762510237403e-05, | |
| "loss": 0.2592, | |
| "mean_token_accuracy": 0.9301635086536407, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9205020920502092, | |
| "grad_norm": 0.21752670407295227, | |
| "learning_rate": 6.779813758756943e-05, | |
| "loss": 0.2683, | |
| "mean_token_accuracy": 0.9270337641239166, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9623430962343096, | |
| "grad_norm": 0.21591052412986755, | |
| "learning_rate": 6.320574468258555e-05, | |
| "loss": 0.2555, | |
| "mean_token_accuracy": 0.930656349658966, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.2959473729133606, | |
| "learning_rate": 5.870144105396118e-05, | |
| "loss": 0.253, | |
| "mean_token_accuracy": 0.9325528542200724, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.3456202745437622, | |
| "eval_mean_token_accuracy": 0.914831633037991, | |
| "eval_runtime": 188.7768, | |
| "eval_samples_per_second": 1.907, | |
| "eval_steps_per_second": 0.238, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0418410041841004, | |
| "grad_norm": 0.24203084409236908, | |
| "learning_rate": 5.429601047334022e-05, | |
| "loss": 0.1868, | |
| "mean_token_accuracy": 0.9483842253684998, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.083682008368201, | |
| "grad_norm": 0.2632400393486023, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.2226, | |
| "mean_token_accuracy": 0.9376705825328827, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.1255230125523012, | |
| "grad_norm": 0.24455517530441284, | |
| "learning_rate": 4.58236947300939e-05, | |
| "loss": 0.2038, | |
| "mean_token_accuracy": 0.9432938873767853, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.1673640167364017, | |
| "grad_norm": 0.27484190464019775, | |
| "learning_rate": 4.17770931730606e-05, | |
| "loss": 0.207, | |
| "mean_token_accuracy": 0.9414038896560669, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.209205020920502, | |
| "grad_norm": 0.22127795219421387, | |
| "learning_rate": 3.786988331415211e-05, | |
| "loss": 0.2139, | |
| "mean_token_accuracy": 0.939983355998993, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.2510460251046025, | |
| "grad_norm": 0.2719532549381256, | |
| "learning_rate": 3.41114194203889e-05, | |
| "loss": 0.2091, | |
| "mean_token_accuracy": 0.9410282075405121, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.292887029288703, | |
| "grad_norm": 0.27510541677474976, | |
| "learning_rate": 3.0510699645470984e-05, | |
| "loss": 0.204, | |
| "mean_token_accuracy": 0.9429362654685974, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.3347280334728033, | |
| "grad_norm": 0.2817744016647339, | |
| "learning_rate": 2.7076344487261697e-05, | |
| "loss": 0.2278, | |
| "mean_token_accuracy": 0.9364924848079681, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.3765690376569037, | |
| "grad_norm": 0.26610061526298523, | |
| "learning_rate": 2.381657614941858e-05, | |
| "loss": 0.1996, | |
| "mean_token_accuracy": 0.9442806720733643, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.418410041841004, | |
| "grad_norm": 0.2731967568397522, | |
| "learning_rate": 2.073919885658223e-05, | |
| "loss": 0.2217, | |
| "mean_token_accuracy": 0.9379707038402557, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4602510460251046, | |
| "grad_norm": 0.2731280028820038, | |
| "learning_rate": 1.7851580170250304e-05, | |
| "loss": 0.1916, | |
| "mean_token_accuracy": 0.9465275406837463, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.502092050209205, | |
| "grad_norm": 0.2769322097301483, | |
| "learning_rate": 1.5160633350068509e-05, | |
| "loss": 0.232, | |
| "mean_token_accuracy": 0.9358445882797242, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5439330543933054, | |
| "grad_norm": 0.27492755651474, | |
| "learning_rate": 1.2672800802767715e-05, | |
| "loss": 0.2094, | |
| "mean_token_accuracy": 0.9408862173557282, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.585774058577406, | |
| "grad_norm": 0.309391051530838, | |
| "learning_rate": 1.0394038658371574e-05, | |
| "loss": 0.1979, | |
| "mean_token_accuracy": 0.9446548879146576, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.6276150627615062, | |
| "grad_norm": 0.2758495509624481, | |
| "learning_rate": 8.329802510601559e-06, | |
| "loss": 0.2048, | |
| "mean_token_accuracy": 0.9433695197105407, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.6694560669456067, | |
| "grad_norm": 0.2817879617214203, | |
| "learning_rate": 6.485034355617747e-06, | |
| "loss": 0.1977, | |
| "mean_token_accuracy": 0.944527405500412, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.711297071129707, | |
| "grad_norm": 0.2818117141723633, | |
| "learning_rate": 4.864150760365771e-06, | |
| "loss": 0.2205, | |
| "mean_token_accuracy": 0.9380866944789886, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.7531380753138075, | |
| "grad_norm": 0.2880750894546509, | |
| "learning_rate": 3.471032288855869e-06, | |
| "loss": 0.2078, | |
| "mean_token_accuracy": 0.942375785112381, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.794979079497908, | |
| "grad_norm": 0.26675674319267273, | |
| "learning_rate": 2.3090142116888646e-06, | |
| "loss": 0.2037, | |
| "mean_token_accuracy": 0.9432387411594391, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.8368200836820083, | |
| "grad_norm": 0.2919078469276428, | |
| "learning_rate": 1.3808785210711606e-06, | |
| "loss": 0.2217, | |
| "mean_token_accuracy": 0.9386197745800018, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8786610878661087, | |
| "grad_norm": 0.27360963821411133, | |
| "learning_rate": 6.888472704359661e-07, | |
| "loss": 0.2101, | |
| "mean_token_accuracy": 0.9409088790416718, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.920502092050209, | |
| "grad_norm": 0.2891521453857422, | |
| "learning_rate": 2.3457725461607517e-07, | |
| "loss": 0.2033, | |
| "mean_token_accuracy": 0.9430344641208649, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.9623430962343096, | |
| "grad_norm": 0.2698313593864441, | |
| "learning_rate": 1.915604330464671e-08, | |
| "loss": 0.1824, | |
| "mean_token_accuracy": 0.9485956370830536, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.97907949790795, | |
| "eval_loss": 0.37856200337409973, | |
| "eval_mean_token_accuracy": 0.9150180731500898, | |
| "eval_runtime": 188.9862, | |
| "eval_samples_per_second": 1.905, | |
| "eval_steps_per_second": 0.238, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.97907949790795, | |
| "step": 357, | |
| "total_flos": 2.503141107915817e+17, | |
| "train_loss": 0.3060124111609633, | |
| "train_runtime": 5934.9405, | |
| "train_samples_per_second": 0.483, | |
| "train_steps_per_second": 0.06 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 357, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.503141107915817e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |