| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.996101364522417, | |
| "eval_steps": 50, | |
| "global_step": 512, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 2.7497, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 2.7052, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5.769230769230769e-05, | |
| "loss": 2.6213, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 2.6018, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.615384615384617e-05, | |
| "loss": 2.5872, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.998328666948438e-05, | |
| "loss": 2.5665, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.991540791356342e-05, | |
| "loss": 2.5494, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.979538999730047e-05, | |
| "loss": 2.5143, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.962335828546048e-05, | |
| "loss": 2.441, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.939949247384046e-05, | |
| "loss": 2.4432, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_accuracy": 0.5032714541605602, | |
| "eval_loss": 2.3850491046905518, | |
| "eval_runtime": 14.1442, | |
| "eval_samples_per_second": 21.21, | |
| "eval_steps_per_second": 10.605, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 9.912402640156811e-05, | |
| "loss": 2.4841, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.879724780684519e-05, | |
| "loss": 2.4121, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.84194980263903e-05, | |
| "loss": 2.4448, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 9.799117163889559e-05, | |
| "loss": 2.4297, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.751271605286941e-05, | |
| "loss": 2.4581, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 2.435, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 9.640746820959684e-05, | |
| "loss": 2.3842, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.57818304394503e-05, | |
| "loss": 2.3872, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 9.51083712390519e-05, | |
| "loss": 2.426, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.438779407049281e-05, | |
| "loss": 2.3655, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_accuracy": 0.5129278619117408, | |
| "eval_loss": 2.3123531341552734, | |
| "eval_runtime": 14.1415, | |
| "eval_samples_per_second": 21.214, | |
| "eval_steps_per_second": 10.607, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.362085161295769e-05, | |
| "loss": 2.402, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 9.280834497651334e-05, | |
| "loss": 2.3523, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 9.195112286530873e-05, | |
| "loss": 2.3354, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.105008069106093e-05, | |
| "loss": 2.37, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 9.01061596377522e-05, | |
| "loss": 2.3597, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 8.912034567851599e-05, | |
| "loss": 2.3646, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 8.809366854573831e-05, | |
| "loss": 2.3601, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 8.702720065545024e-05, | |
| "loss": 2.3424, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 8.592205598713539e-05, | |
| "loss": 2.3362, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 8.47793889201221e-05, | |
| "loss": 2.374, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "eval_accuracy": 0.5214769581501384, | |
| "eval_loss": 2.258837938308716, | |
| "eval_runtime": 14.3574, | |
| "eval_samples_per_second": 20.895, | |
| "eval_steps_per_second": 10.448, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 8.360039302777612e-05, | |
| "loss": 2.3383, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 8.238629983075294e-05, | |
| "loss": 2.319, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 8.113837751061246e-05, | |
| "loss": 2.3517, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 7.985792958513931e-05, | |
| "loss": 2.307, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 7.854629354675291e-05, | |
| "loss": 2.2915, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 7.720483946542914e-05, | |
| "loss": 2.3198, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.583496855759316e-05, | |
| "loss": 2.3304, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 7.443811172247821e-05, | |
| "loss": 2.2926, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.30157280474793e-05, | |
| "loss": 2.2753, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.156930328406268e-05, | |
| "loss": 2.3558, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_accuracy": 0.5291418335775933, | |
| "eval_loss": 2.2132091522216797, | |
| "eval_runtime": 14.0805, | |
| "eval_samples_per_second": 21.306, | |
| "eval_steps_per_second": 10.653, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 7.01003482958237e-05, | |
| "loss": 2.3026, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 6.861039748031351e-05, | |
| "loss": 2.3327, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 2.3047, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 6.557375398802123e-05, | |
| "loss": 2.2849, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6.403023323847695e-05, | |
| "loss": 2.2972, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 6.247205720289907e-05, | |
| "loss": 2.3069, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 6.09008534747213e-05, | |
| "loss": 2.3305, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.9318263255459116e-05, | |
| "loss": 2.2845, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5.772593964039203e-05, | |
| "loss": 2.2698, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5.6125545891822274e-05, | |
| "loss": 2.2677, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_accuracy": 0.5348330890734408, | |
| "eval_loss": 2.182823896408081, | |
| "eval_runtime": 14.0998, | |
| "eval_samples_per_second": 21.277, | |
| "eval_steps_per_second": 10.638, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5.451875370171341e-05, | |
| "loss": 2.2776, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5.290724144552379e-05, | |
| "loss": 2.1158, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 5.129269242905882e-05, | |
| "loss": 2.0758, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 4.967679313017303e-05, | |
| "loss": 2.053, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 4.806123143715916e-05, | |
| "loss": 2.0824, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 4.6447694885663514e-05, | |
| "loss": 2.062, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 4.4837868895969936e-05, | |
| "loss": 2.0986, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 4.323343501249346e-05, | |
| "loss": 2.0749, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 4.1636069147322246e-05, | |
| "loss": 2.0365, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 4.004743982964298e-05, | |
| "loss": 2.0701, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.5372854583943983, | |
| "eval_loss": 2.178772211074829, | |
| "eval_runtime": 14.087, | |
| "eval_samples_per_second": 21.296, | |
| "eval_steps_per_second": 10.648, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 3.846920646287799e-05, | |
| "loss": 2.0554, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 3.6903017591354706e-05, | |
| "loss": 2.0432, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 3.535050917831797e-05, | |
| "loss": 2.0663, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 3.381330289708396e-05, | |
| "loss": 2.0673, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 3.2293004437120624e-05, | |
| "loss": 2.0106, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 3.079120182682412e-05, | |
| "loss": 2.0177, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 2.9309463774743046e-05, | |
| "loss": 2.0314, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2.7849338030983257e-05, | |
| "loss": 2.06, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 2.641234977050484e-05, | |
| "loss": 2.0765, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.0766, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "eval_accuracy": 0.5397508549096238, | |
| "eval_loss": 2.1673190593719482, | |
| "eval_runtime": 14.2515, | |
| "eval_samples_per_second": 21.05, | |
| "eval_steps_per_second": 10.525, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2.361376399001592e-05, | |
| "loss": 2.0481, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2.225508973396016e-05, | |
| "loss": 2.0246, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2.0925396435598664e-05, | |
| "loss": 2.0563, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 1.9626073026625818e-05, | |
| "loss": 2.0373, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 1.835847671585526e-05, | |
| "loss": 2.0495, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 1.7123931571546827e-05, | |
| "loss": 2.0612, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.592372713835055e-05, | |
| "loss": 2.0543, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 1.4759117090312197e-05, | |
| "loss": 2.0719, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 1.3631317921347563e-05, | |
| "loss": 2.0468, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 1.25415076745532e-05, | |
| "loss": 2.0669, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "eval_accuracy": 0.5402182055039896, | |
| "eval_loss": 2.165104866027832, | |
| "eval_runtime": 14.1201, | |
| "eval_samples_per_second": 21.246, | |
| "eval_steps_per_second": 10.623, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 1.1490824711681025e-05, | |
| "loss": 2.0491, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 1.0480366524062042e-05, | |
| "loss": 2.0686, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 9.511188586221376e-06, | |
| "loss": 2.0816, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 8.584303253381847e-06, | |
| "loss": 2.0808, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 7.700678704007947e-06, | |
| "loss": 2.0244, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 6.861237928494579e-06, | |
| "loss": 2.0715, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 6.066857765057055e-06, | |
| "loss": 2.0629, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 5.318367983829392e-06, | |
| "loss": 2.0738, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 4.616550420127563e-06, | |
| "loss": 2.0449, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 3.962138157783085e-06, | |
| "loss": 2.0314, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "eval_accuracy": 0.5405943657384791, | |
| "eval_loss": 2.1640877723693848, | |
| "eval_runtime": 14.1591, | |
| "eval_samples_per_second": 21.188, | |
| "eval_steps_per_second": 10.594, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 3.3558147633999728e-06, | |
| "loss": 2.0818, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 2.798213572335001e-06, | |
| "loss": 2.0991, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 2.2899170271469428e-06, | |
| "loss": 2.0317, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.8314560692059835e-06, | |
| "loss": 2.0518, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 1.4233095840986753e-06, | |
| "loss": 2.013, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 1.0659039014077944e-06, | |
| "loss": 2.0317, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 2.038, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 5.047548650136513e-07, | |
| "loss": 2.0414, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 3.0159765977250673e-07, | |
| "loss": 2.0509, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 1.503529416103988e-07, | |
| "loss": 2.0281, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_accuracy": 0.5406627585083863, | |
| "eval_loss": 2.163918972015381, | |
| "eval_runtime": 14.1546, | |
| "eval_samples_per_second": 21.194, | |
| "eval_steps_per_second": 10.597, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 5.11786932613223e-08, | |
| "loss": 2.0471, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 4.178507228136397e-09, | |
| "loss": 2.055, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 512, | |
| "total_flos": 4.1654970530503066e+17, | |
| "train_loss": 2.223597173579037, | |
| "train_runtime": 8040.0291, | |
| "train_samples_per_second": 4.084, | |
| "train_steps_per_second": 0.064, | |
| "train_tokens_per_second": 8363.146 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 512, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "total_flos": 4.1654970530503066e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |