| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.5836035201482168, |
| "eval_steps": 500, |
| "global_step": 315, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.009263547938860583, |
| "grad_norm": 13.104758262634277, |
| "learning_rate": 2.162496240601504e-06, |
| "loss": 0.506, |
| "mean_token_accuracy": 0.8712682545185089, |
| "num_tokens": 873838.0, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.018527095877721167, |
| "grad_norm": 3.3287055492401123, |
| "learning_rate": 4.865616541353383e-06, |
| "loss": 0.3595, |
| "mean_token_accuracy": 0.9212990075349807, |
| "num_tokens": 1737742.0, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.02779064381658175, |
| "grad_norm": 1.2767220735549927, |
| "learning_rate": 7.568736842105265e-06, |
| "loss": 0.252, |
| "mean_token_accuracy": 0.9290035903453827, |
| "num_tokens": 2607865.0, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.037054191755442334, |
| "grad_norm": 1.0932143926620483, |
| "learning_rate": 1.0271857142857143e-05, |
| "loss": 0.2665, |
| "mean_token_accuracy": 0.9247954428195954, |
| "num_tokens": 3469048.0, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.04631773969430292, |
| "grad_norm": 1.009112000465393, |
| "learning_rate": 1.2974977443609024e-05, |
| "loss": 0.2555, |
| "mean_token_accuracy": 0.9292442113161087, |
| "num_tokens": 4332146.0, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.0555812876331635, |
| "grad_norm": 1.0787107944488525, |
| "learning_rate": 1.5678097744360906e-05, |
| "loss": 0.249, |
| "mean_token_accuracy": 0.932613629102707, |
| "num_tokens": 5193234.0, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.06484483557202408, |
| "grad_norm": 0.9236277341842651, |
| "learning_rate": 1.8381218045112784e-05, |
| "loss": 0.2596, |
| "mean_token_accuracy": 0.9296454936265945, |
| "num_tokens": 6058966.0, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.07410838351088467, |
| "grad_norm": 0.9933350682258606, |
| "learning_rate": 1.891469721679475e-05, |
| "loss": 0.2209, |
| "mean_token_accuracy": 0.936568307876587, |
| "num_tokens": 6923299.0, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.08337193144974525, |
| "grad_norm": 0.8766371011734009, |
| "learning_rate": 1.8885695766218317e-05, |
| "loss": 0.2094, |
| "mean_token_accuracy": 0.9389084279537201, |
| "num_tokens": 7781156.0, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.09263547938860583, |
| "grad_norm": 0.967535138130188, |
| "learning_rate": 1.8834482377020943e-05, |
| "loss": 0.2342, |
| "mean_token_accuracy": 0.931867802143097, |
| "num_tokens": 8642107.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.10189902732746642, |
| "grad_norm": 0.9646473526954651, |
| "learning_rate": 1.8761218185472167e-05, |
| "loss": 0.2099, |
| "mean_token_accuracy": 0.941218912601471, |
| "num_tokens": 9498418.0, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.111162575266327, |
| "grad_norm": 0.9485169053077698, |
| "learning_rate": 1.8666133707823747e-05, |
| "loss": 0.2399, |
| "mean_token_accuracy": 0.934317621588707, |
| "num_tokens": 10356782.0, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.12042612320518759, |
| "grad_norm": 0.797245979309082, |
| "learning_rate": 1.8549528115020233e-05, |
| "loss": 0.2192, |
| "mean_token_accuracy": 0.9368351072072982, |
| "num_tokens": 11215585.0, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.12968967114404817, |
| "grad_norm": 0.9420742988586426, |
| "learning_rate": 1.8411768291396496e-05, |
| "loss": 0.2195, |
| "mean_token_accuracy": 0.9353754937648773, |
| "num_tokens": 12076168.0, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.13895321908290875, |
| "grad_norm": 0.8384352922439575, |
| "learning_rate": 1.8253287680323885e-05, |
| "loss": 0.206, |
| "mean_token_accuracy": 0.9410769432783127, |
| "num_tokens": 12944191.0, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.14821676702176934, |
| "grad_norm": 0.6907543540000916, |
| "learning_rate": 1.8074584920437013e-05, |
| "loss": 0.1883, |
| "mean_token_accuracy": 0.9415487110614776, |
| "num_tokens": 13817659.0, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.15748031496062992, |
| "grad_norm": 1.0261110067367554, |
| "learning_rate": 1.7876222276732135e-05, |
| "loss": 0.2363, |
| "mean_token_accuracy": 0.9339721232652665, |
| "num_tokens": 14677271.0, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.1667438628994905, |
| "grad_norm": 0.8936851620674133, |
| "learning_rate": 1.7658823871473464e-05, |
| "loss": 0.1956, |
| "mean_token_accuracy": 0.9425812482833862, |
| "num_tokens": 15547415.0, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.17600741083835109, |
| "grad_norm": 0.9527491331100464, |
| "learning_rate": 1.74230737204736e-05, |
| "loss": 0.205, |
| "mean_token_accuracy": 0.940983098745346, |
| "num_tokens": 16415759.0, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.18527095877721167, |
| "grad_norm": 0.9818997383117676, |
| "learning_rate": 1.7169713580926744e-05, |
| "loss": 0.2084, |
| "mean_token_accuracy": 0.9393031448125839, |
| "num_tokens": 17287335.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.19453450671607225, |
| "grad_norm": 0.8504663705825806, |
| "learning_rate": 1.6899540617566137e-05, |
| "loss": 0.1974, |
| "mean_token_accuracy": 0.9433632969856263, |
| "num_tokens": 18153892.0, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.20379805465493284, |
| "grad_norm": 0.8879144787788391, |
| "learning_rate": 1.661340489448888e-05, |
| "loss": 0.2012, |
| "mean_token_accuracy": 0.9429984986782074, |
| "num_tokens": 19020559.0, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.21306160259379342, |
| "grad_norm": 0.8018971681594849, |
| "learning_rate": 1.6312206700539794e-05, |
| "loss": 0.1947, |
| "mean_token_accuracy": 0.9416427314281464, |
| "num_tokens": 19873885.0, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.222325150532654, |
| "grad_norm": 0.7375516295433044, |
| "learning_rate": 1.5996893716669644e-05, |
| "loss": 0.1823, |
| "mean_token_accuracy": 0.9444915145635605, |
| "num_tokens": 20746036.0, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.2315886984715146, |
| "grad_norm": 0.8536906242370605, |
| "learning_rate": 1.5668458034180252e-05, |
| "loss": 0.1996, |
| "mean_token_accuracy": 0.9446161150932312, |
| "num_tokens": 21599195.0, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.24085224641037517, |
| "grad_norm": 0.7917492389678955, |
| "learning_rate": 1.532793303323823e-05, |
| "loss": 0.1797, |
| "mean_token_accuracy": 0.9451331734657288, |
| "num_tokens": 22464999.0, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.25011579434923575, |
| "grad_norm": 0.8580403923988342, |
| "learning_rate": 1.4976390131478662e-05, |
| "loss": 0.1976, |
| "mean_token_accuracy": 0.9438081711530686, |
| "num_tokens": 23329248.0, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.25937934228809634, |
| "grad_norm": 0.8116864562034607, |
| "learning_rate": 1.4614935412928825e-05, |
| "loss": 0.1991, |
| "mean_token_accuracy": 0.9409991979599, |
| "num_tokens": 24179228.0, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.2686428902269569, |
| "grad_norm": 0.65013188123703, |
| "learning_rate": 1.4244706147858645e-05, |
| "loss": 0.1867, |
| "mean_token_accuracy": 0.9426582366228103, |
| "num_tokens": 25047486.0, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.2779064381658175, |
| "grad_norm": 0.8328827023506165, |
| "learning_rate": 1.3866867214507666e-05, |
| "loss": 0.1786, |
| "mean_token_accuracy": 0.9469089567661285, |
| "num_tokens": 25913450.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.2871699861046781, |
| "grad_norm": 0.9690430164337158, |
| "learning_rate": 1.3482607433947159e-05, |
| "loss": 0.1952, |
| "mean_token_accuracy": 0.9430926829576493, |
| "num_tokens": 26781613.0, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.29643353404353867, |
| "grad_norm": 0.7961167693138123, |
| "learning_rate": 1.3093135829609184e-05, |
| "loss": 0.191, |
| "mean_token_accuracy": 0.9435102790594101, |
| "num_tokens": 27635586.0, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.30569708198239925, |
| "grad_norm": 0.7861698269844055, |
| "learning_rate": 1.2699677823251611e-05, |
| "loss": 0.1921, |
| "mean_token_accuracy": 0.943246129155159, |
| "num_tokens": 28496268.0, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.31496062992125984, |
| "grad_norm": 0.9173203110694885, |
| "learning_rate": 1.2303471379327878e-05, |
| "loss": 0.2045, |
| "mean_token_accuracy": 0.9452885925769806, |
| "num_tokens": 29354500.0, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.3242241778601204, |
| "grad_norm": 0.830530047416687, |
| "learning_rate": 1.1905763109892843e-05, |
| "loss": 0.1825, |
| "mean_token_accuracy": 0.9460207551717759, |
| "num_tokens": 30227174.0, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.333487725798981, |
| "grad_norm": 0.6210633516311646, |
| "learning_rate": 1.1507804352300028e-05, |
| "loss": 0.167, |
| "mean_token_accuracy": 0.9506563901901245, |
| "num_tokens": 31094821.0, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.3427512737378416, |
| "grad_norm": 0.7468693256378174, |
| "learning_rate": 1.1110847232031383e-05, |
| "loss": 0.178, |
| "mean_token_accuracy": 0.946472841501236, |
| "num_tokens": 31961760.0, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.35201482167670217, |
| "grad_norm": 0.6815229058265686, |
| "learning_rate": 1.0716140723047263e-05, |
| "loss": 0.1672, |
| "mean_token_accuracy": 0.9525276869535446, |
| "num_tokens": 32824679.0, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.36127836961556276, |
| "grad_norm": 1.0449384450912476, |
| "learning_rate": 1.0324926718052282e-05, |
| "loss": 0.1762, |
| "mean_token_accuracy": 0.9472320646047592, |
| "num_tokens": 33687921.0, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.37054191755442334, |
| "grad_norm": 0.8821719288825989, |
| "learning_rate": 9.938436121041386e-06, |
| "loss": 0.1831, |
| "mean_token_accuracy": 0.9451801896095275, |
| "num_tokens": 34537679.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.3798054654932839, |
| "grad_norm": 0.9379377365112305, |
| "learning_rate": 9.557884974420524e-06, |
| "loss": 0.1814, |
| "mean_token_accuracy": 0.9483665764331818, |
| "num_tokens": 35405475.0, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.3890690134321445, |
| "grad_norm": 0.8580158352851868, |
| "learning_rate": 9.18447063288733e-06, |
| "loss": 0.1787, |
| "mean_token_accuracy": 0.9478948295116425, |
| "num_tokens": 36256130.0, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.3983325613710051, |
| "grad_norm": 0.8793883323669434, |
| "learning_rate": 8.819367996110269e-06, |
| "loss": 0.1613, |
| "mean_token_accuracy": 0.9521415024995804, |
| "num_tokens": 37127526.0, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.4075961093098657, |
| "grad_norm": 0.8262807130813599, |
| "learning_rate": 8.463725812059604e-06, |
| "loss": 0.1753, |
| "mean_token_accuracy": 0.9494184970855712, |
| "num_tokens": 37991629.0, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.41685965724872626, |
| "grad_norm": 0.7080918550491333, |
| "learning_rate": 8.118663062621298e-06, |
| "loss": 0.171, |
| "mean_token_accuracy": 0.9474876701831818, |
| "num_tokens": 38862547.0, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.42612320518758684, |
| "grad_norm": 1.2623839378356934, |
| "learning_rate": 7.785265442866e-06, |
| "loss": 0.1745, |
| "mean_token_accuracy": 0.9472187548875809, |
| "num_tokens": 39730487.0, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.4353867531264474, |
| "grad_norm": 0.8184564113616943, |
| "learning_rate": 7.464581945050706e-06, |
| "loss": 0.1844, |
| "mean_token_accuracy": 0.9464471131563187, |
| "num_tokens": 40578272.0, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.444650301065308, |
| "grad_norm": 1.069190263748169, |
| "learning_rate": 7.157621558101005e-06, |
| "loss": 0.1816, |
| "mean_token_accuracy": 0.9481162607669831, |
| "num_tokens": 41440335.0, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.4539138490041686, |
| "grad_norm": 0.9806954860687256, |
| "learning_rate": 6.86535009295865e-06, |
| "loss": 0.1616, |
| "mean_token_accuracy": 0.9513800472021103, |
| "num_tokens": 42300428.0, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.4631773969430292, |
| "grad_norm": 0.8368266224861145, |
| "learning_rate": 6.588687143783003e-06, |
| "loss": 0.1802, |
| "mean_token_accuracy": 0.947989496588707, |
| "num_tokens": 43158864.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.47244094488188976, |
| "grad_norm": 0.7476319074630737, |
| "learning_rate": 6.328503194567604e-06, |
| "loss": 0.1456, |
| "mean_token_accuracy": 0.9543960303068161, |
| "num_tokens": 44025265.0, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.48170449282075034, |
| "grad_norm": 0.8824567794799805, |
| "learning_rate": 6.085616880275477e-06, |
| "loss": 0.1416, |
| "mean_token_accuracy": 0.9571489185094834, |
| "num_tokens": 44894229.0, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.4909680407596109, |
| "grad_norm": 0.8804921507835388, |
| "learning_rate": 5.860792411110664e-06, |
| "loss": 0.1814, |
| "mean_token_accuracy": 0.9470549017190933, |
| "num_tokens": 45759005.0, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.5002315886984715, |
| "grad_norm": 0.7593424916267395, |
| "learning_rate": 5.654737168030201e-06, |
| "loss": 0.1615, |
| "mean_token_accuracy": 0.9508086234331131, |
| "num_tokens": 46619012.0, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.5094951366373321, |
| "grad_norm": 1.1181437969207764, |
| "learning_rate": 5.468099477061918e-06, |
| "loss": 0.1871, |
| "mean_token_accuracy": 0.9448368161916733, |
| "num_tokens": 47480812.0, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.5187586845761927, |
| "grad_norm": 0.8610813021659851, |
| "learning_rate": 5.301466569430894e-06, |
| "loss": 0.1662, |
| "mean_token_accuracy": 0.9499164700508118, |
| "num_tokens": 48342621.0, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.5280222325150533, |
| "grad_norm": 0.9683968424797058, |
| "learning_rate": 5.155362733912775e-06, |
| "loss": 0.1667, |
| "mean_token_accuracy": 0.9505377233028411, |
| "num_tokens": 49196606.0, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.5372857804539138, |
| "grad_norm": 0.9030554890632629, |
| "learning_rate": 5.030247667227314e-06, |
| "loss": 0.1619, |
| "mean_token_accuracy": 0.9480885207653046, |
| "num_tokens": 50060321.0, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.5465493283927745, |
| "grad_norm": 0.8542042374610901, |
| "learning_rate": 4.926515027662422e-06, |
| "loss": 0.1609, |
| "mean_token_accuracy": 0.9516505330801011, |
| "num_tokens": 50928245.0, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.555812876331635, |
| "grad_norm": 0.9249922037124634, |
| "learning_rate": 4.844491196479547e-06, |
| "loss": 0.1499, |
| "mean_token_accuracy": 0.9537470370531083, |
| "num_tokens": 51805269.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5650764242704956, |
| "grad_norm": 1.025533676147461, |
| "learning_rate": 4.784434250997487e-06, |
| "loss": 0.1542, |
| "mean_token_accuracy": 0.9532197445631028, |
| "num_tokens": 52662752.0, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.5743399722093562, |
| "grad_norm": 0.8454712629318237, |
| "learning_rate": 4.746533152585658e-06, |
| "loss": 0.1454, |
| "mean_token_accuracy": 0.956709685921669, |
| "num_tokens": 53526046.0, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.5836035201482168, |
| "grad_norm": 1.0178312063217163, |
| "learning_rate": 4.730907152121721e-06, |
| "loss": 0.158, |
| "mean_token_accuracy": 0.9514861524105072, |
| "num_tokens": 54397820.0, |
| "step": 315 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 315, |
| "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": 3.0040080593505485e+18, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
