| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.09999063758075087, | |
| "global_step": 4272, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.1682242990654204e-06, | |
| "loss": 10.9364, | |
| "theoretical_loss": 20.81281780154715, | |
| "tokens_seen": 65536 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 5.841121495327103e-05, | |
| "loss": 8.9947, | |
| "theoretical_loss": 8.563482664611069, | |
| "tokens_seen": 3276800 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.00011682242990654206, | |
| "loss": 7.0133, | |
| "theoretical_loss": 7.4777587180480305, | |
| "tokens_seen": 6553600 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.00017523364485981307, | |
| "loss": 6.2244, | |
| "theoretical_loss": 6.9337544888949, | |
| "tokens_seen": 9830400 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.00023364485981308412, | |
| "loss": 5.822, | |
| "theoretical_loss": 6.583566228426414, | |
| "tokens_seen": 13107200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.00029205607476635517, | |
| "loss": 5.5642, | |
| "theoretical_loss": 6.330713565116083, | |
| "tokens_seen": 16384000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.00035046728971962614, | |
| "loss": 5.3688, | |
| "theoretical_loss": 6.135529231940326, | |
| "tokens_seen": 19660800 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0004088785046728972, | |
| "loss": 5.1987, | |
| "theoretical_loss": 5.978101583869607, | |
| "tokens_seen": 22937600 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.00046728971962616824, | |
| "loss": 5.1026, | |
| "theoretical_loss": 5.8471173262659235, | |
| "tokens_seen": 26214400 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.000499739928125591, | |
| "loss": 4.9312, | |
| "theoretical_loss": 5.7355768158821245, | |
| "tokens_seen": 29491200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0004991488556837526, | |
| "loss": 4.8598, | |
| "theoretical_loss": 5.638870144071353, | |
| "tokens_seen": 32768000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0004985577832419141, | |
| "loss": 4.7297, | |
| "theoretical_loss": 5.553812381844907, | |
| "tokens_seen": 36044800 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0004979667108000757, | |
| "loss": 4.6466, | |
| "theoretical_loss": 5.478118080556438, | |
| "tokens_seen": 39321600 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004973756383582371, | |
| "loss": 4.5646, | |
| "theoretical_loss": 5.410095959579362, | |
| "tokens_seen": 42598400 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004967845659163987, | |
| "loss": 4.5162, | |
| "theoretical_loss": 5.348462083735834, | |
| "tokens_seen": 45875200 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004961934934745603, | |
| "loss": 4.3809, | |
| "theoretical_loss": 5.292220566937567, | |
| "tokens_seen": 49152000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004956024210327218, | |
| "loss": 4.336, | |
| "theoretical_loss": 5.240584625769978, | |
| "tokens_seen": 52428800 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004950113485908833, | |
| "loss": 4.2829, | |
| "theoretical_loss": 5.192922724525789, | |
| "tokens_seen": 55705600 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004944202761490448, | |
| "loss": 4.209, | |
| "theoretical_loss": 5.1487208633564405, | |
| "tokens_seen": 58982400 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004938292037072064, | |
| "loss": 4.0751, | |
| "theoretical_loss": 5.107555562405102, | |
| "tokens_seen": 62259200 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004932381312653678, | |
| "loss": 3.9696, | |
| "theoretical_loss": 5.069074117143246, | |
| "tokens_seen": 65536000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0004926470588235294, | |
| "loss": 3.9197, | |
| "theoretical_loss": 5.032979909838007, | |
| "tokens_seen": 68812800 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.000492055986381691, | |
| "loss": 3.8378, | |
| "theoretical_loss": 4.999021308224664, | |
| "tokens_seen": 72089600 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0004914649139398525, | |
| "loss": 3.7856, | |
| "theoretical_loss": 4.966983155351962, | |
| "tokens_seen": 75366400 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.000490873841498014, | |
| "loss": 3.6988, | |
| "theoretical_loss": 4.9366801616251355, | |
| "tokens_seen": 78643200 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0004902827690561755, | |
| "loss": 3.6826, | |
| "theoretical_loss": 4.907951713830082, | |
| "tokens_seen": 81920000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0004896916966143371, | |
| "loss": 3.6271, | |
| "theoretical_loss": 4.880657753812926, | |
| "tokens_seen": 85196800 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0004891006241724985, | |
| "loss": 3.617, | |
| "theoretical_loss": 4.854675474481779, | |
| "tokens_seen": 88473600 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0004885095517306601, | |
| "loss": 3.528, | |
| "theoretical_loss": 4.8298966473088125, | |
| "tokens_seen": 91750400 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 0.0004879184792888217, | |
| "loss": 3.5513, | |
| "theoretical_loss": 4.8062254427779205, | |
| "tokens_seen": 95027200 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0004873274068469832, | |
| "loss": 3.4962, | |
| "theoretical_loss": 4.783576639276257, | |
| "tokens_seen": 98304000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.00048673633440514467, | |
| "loss": 3.5334, | |
| "theoretical_loss": 4.761874140772408, | |
| "tokens_seen": 101580800 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0004861452619633062, | |
| "loss": 3.5329, | |
| "theoretical_loss": 4.741049741962473, | |
| "tokens_seen": 104857600 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0004855541895214677, | |
| "loss": 3.5329, | |
| "theoretical_loss": 4.721042093249051, | |
| "tokens_seen": 108134400 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0004849631170796293, | |
| "loss": 3.4926, | |
| "theoretical_loss": 4.701795828231866, | |
| "tokens_seen": 111411200 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0004843720446377908, | |
| "loss": 3.4835, | |
| "theoretical_loss": 4.68326082423593, | |
| "tokens_seen": 114688000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.00048378097219595233, | |
| "loss": 3.4587, | |
| "theoretical_loss": 4.665391572426282, | |
| "tokens_seen": 117964800 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.00048318989975411385, | |
| "loss": 3.4307, | |
| "theoretical_loss": 4.648146638719739, | |
| "tokens_seen": 121241600 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.00048259882731227537, | |
| "loss": 3.443, | |
| "theoretical_loss": 4.631488200339643, | |
| "tokens_seen": 124518400 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0004820077548704369, | |
| "loss": 3.45, | |
| "theoretical_loss": 4.615381645715717, | |
| "tokens_seen": 127795200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00048141668242859847, | |
| "loss": 3.3857, | |
| "theoretical_loss": 4.599795227690505, | |
| "tokens_seen": 131072000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00048082560998676, | |
| "loss": 3.4065, | |
| "theoretical_loss": 4.584699761792674, | |
| "tokens_seen": 134348800 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0004802345375449215, | |
| "loss": 3.3659, | |
| "theoretical_loss": 4.570068362778516, | |
| "tokens_seen": 137625600 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00047964346510308303, | |
| "loss": 3.3781, | |
| "theoretical_loss": 4.555876213804037, | |
| "tokens_seen": 140902400 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00047905239266124455, | |
| "loss": 3.3612, | |
| "theoretical_loss": 4.542100363530799, | |
| "tokens_seen": 144179200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.00047846132021940607, | |
| "loss": 3.3402, | |
| "theoretical_loss": 4.528719547234816, | |
| "tokens_seen": 147456000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0004778702477775676, | |
| "loss": 3.2899, | |
| "theoretical_loss": 4.515714028614996, | |
| "tokens_seen": 150732800 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00047727917533572917, | |
| "loss": 3.3099, | |
| "theoretical_loss": 4.503065459513339, | |
| "tokens_seen": 154009600 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0004766881028938907, | |
| "loss": 3.3162, | |
| "theoretical_loss": 4.4907567551852665, | |
| "tokens_seen": 157286400 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0004760970304520522, | |
| "loss": 3.3036, | |
| "theoretical_loss": 4.478771983111967, | |
| "tokens_seen": 160563200 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "objective/train/docs_used": 104000, | |
| "objective/train/instantaneous_batch_size": 32, | |
| "objective/train/instantaneous_microbatch_size": 32768, | |
| "objective/train/original_loss": 3.2747654914855957, | |
| "objective/train/theoretical_loss": 4.467096263641219, | |
| "objective/train/tokens_used": 184300000, | |
| "theoretical_loss": 4.467096263641219, | |
| "tokens_seen": 163840000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00047550595801021373, | |
| "loss": 3.2837, | |
| "theoretical_loss": 4.467096263641219, | |
| "tokens_seen": 163840000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00047491488556837525, | |
| "loss": 3.2955, | |
| "theoretical_loss": 4.455715680989545, | |
| "tokens_seen": 167116800 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00047432381312653677, | |
| "loss": 3.2608, | |
| "theoretical_loss": 4.44461720334543, | |
| "tokens_seen": 170393600 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00047373274068469835, | |
| "loss": 3.2026, | |
| "theoretical_loss": 4.433788610987646, | |
| "tokens_seen": 173670400 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00047314166824285987, | |
| "loss": 3.2621, | |
| "theoretical_loss": 4.42321843148016, | |
| "tokens_seen": 176947200 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.0004725505958010214, | |
| "loss": 3.1999, | |
| "theoretical_loss": 4.412895881130142, | |
| "tokens_seen": 180224000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0004719595233591829, | |
| "loss": 3.2384, | |
| "theoretical_loss": 4.4028108120020795, | |
| "tokens_seen": 183500800 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.00047136845091734443, | |
| "loss": 3.244, | |
| "theoretical_loss": 4.392953663871862, | |
| "tokens_seen": 186777600 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.00047077737847550595, | |
| "loss": 3.2329, | |
| "theoretical_loss": 4.383315420582533, | |
| "tokens_seen": 190054400 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0004701863060336675, | |
| "loss": 3.2634, | |
| "theoretical_loss": 4.373887570330275, | |
| "tokens_seen": 193331200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.00046959523359182905, | |
| "loss": 3.2143, | |
| "theoretical_loss": 4.364662069466704, | |
| "tokens_seen": 196608000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.00046900416114999057, | |
| "loss": 3.2128, | |
| "theoretical_loss": 4.355631309453283, | |
| "tokens_seen": 199884800 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0004684130887081521, | |
| "loss": 3.1675, | |
| "theoretical_loss": 4.346788086646671, | |
| "tokens_seen": 203161600 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0004678220162663136, | |
| "loss": 3.1967, | |
| "theoretical_loss": 4.33812557463116, | |
| "tokens_seen": 206438400 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.00046723094382447513, | |
| "loss": 3.2042, | |
| "theoretical_loss": 4.329637298846812, | |
| "tokens_seen": 209715200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00046663987138263665, | |
| "loss": 3.1574, | |
| "theoretical_loss": 4.321317113290252, | |
| "tokens_seen": 212992000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.0004660487989407982, | |
| "loss": 3.1317, | |
| "theoretical_loss": 4.3131591790897925, | |
| "tokens_seen": 216268800 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00046545772649895975, | |
| "loss": 3.1829, | |
| "theoretical_loss": 4.305157944778228, | |
| "tokens_seen": 219545600 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00046486665405712127, | |
| "loss": 3.2073, | |
| "theoretical_loss": 4.297308128105687, | |
| "tokens_seen": 222822400 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.0004642755816152828, | |
| "loss": 3.1994, | |
| "theoretical_loss": 4.2896046992515995, | |
| "tokens_seen": 226099200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.0004636845091734443, | |
| "loss": 3.2312, | |
| "theoretical_loss": 4.282042865309616, | |
| "tokens_seen": 229376000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00046309343673160583, | |
| "loss": 3.2006, | |
| "theoretical_loss": 4.274618055932298, | |
| "tokens_seen": 232652800 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.0004625023642897674, | |
| "loss": 3.1756, | |
| "theoretical_loss": 4.267325910033897, | |
| "tokens_seen": 235929600 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0004619112918479289, | |
| "loss": 3.1158, | |
| "theoretical_loss": 4.260162263459744, | |
| "tokens_seen": 239206400 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.00046132021940609044, | |
| "loss": 3.1448, | |
| "theoretical_loss": 4.253123137539814, | |
| "tokens_seen": 242483200 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.00046072914696425197, | |
| "loss": 3.166, | |
| "theoretical_loss": 4.246204728452055, | |
| "tokens_seen": 245760000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0004601380745224135, | |
| "loss": 3.1762, | |
| "theoretical_loss": 4.239403397328261, | |
| "tokens_seen": 249036800 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.000459547002080575, | |
| "loss": 3.1442, | |
| "theoretical_loss": 4.232715661041632, | |
| "tokens_seen": 252313600 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0004589559296387366, | |
| "loss": 3.1504, | |
| "theoretical_loss": 4.226138183620867, | |
| "tokens_seen": 255590400 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0004583648571968981, | |
| "loss": 3.1099, | |
| "theoretical_loss": 4.219667768240775, | |
| "tokens_seen": 258867200 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0004577737847550596, | |
| "loss": 3.114, | |
| "theoretical_loss": 4.213301349743924, | |
| "tokens_seen": 262144000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 0.0004571827123132211, | |
| "loss": 3.0703, | |
| "theoretical_loss": 4.20703598765197, | |
| "tokens_seen": 265420800 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.0004565916398713826, | |
| "loss": 3.0383, | |
| "theoretical_loss": 4.2008688596290025, | |
| "tokens_seen": 268697600 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.00045600056742954413, | |
| "loss": 2.9881, | |
| "theoretical_loss": 4.194797255362549, | |
| "tokens_seen": 271974400 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.0004554094949877057, | |
| "loss": 3.068, | |
| "theoretical_loss": 4.188818570830883, | |
| "tokens_seen": 275251200 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.00045481842254586723, | |
| "loss": 3.0823, | |
| "theoretical_loss": 4.182930302927963, | |
| "tokens_seen": 278528000 | |
| } | |
| ], | |
| "max_steps": 42724, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 1.42878663770112e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |