| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.999627004848937, | |
| "eval_steps": 500, | |
| "global_step": 670, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.477611940298507e-07, | |
| "loss": 2.637, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.738805970149253e-06, | |
| "loss": 2.6461, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.477611940298506e-06, | |
| "loss": 2.6439, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.4216417910447761e-05, | |
| "loss": 2.642, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.895522388059701e-05, | |
| "loss": 2.6207, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.369402985074627e-05, | |
| "loss": 2.6062, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.8432835820895522e-05, | |
| "loss": 2.5733, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.317164179104477e-05, | |
| "loss": 2.531, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.791044776119402e-05, | |
| "loss": 2.495, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 4.2649253731343286e-05, | |
| "loss": 2.4618, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 4.738805970149254e-05, | |
| "loss": 2.4374, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 5.2126865671641794e-05, | |
| "loss": 2.4205, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 5.6865671641791044e-05, | |
| "loss": 2.3846, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 6.16044776119403e-05, | |
| "loss": 2.3675, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 6.349612195786573e-05, | |
| "loss": 2.3648, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 6.3472426242142e-05, | |
| "loss": 2.3394, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.09912109375, | |
| "learning_rate": 6.342720533773385e-05, | |
| "loss": 2.3241, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 6.336048992919527e-05, | |
| "loss": 2.3123, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 6.327232528613285e-05, | |
| "loss": 2.3201, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 6.316277123248829e-05, | |
| "loss": 2.2999, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 6.303190210594489e-05, | |
| "loss": 2.3114, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 6.287980670748592e-05, | |
| "loss": 2.3002, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 6.270658824113884e-05, | |
| "loss": 2.2732, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 6.251236424394651e-05, | |
| "loss": 2.2867, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 6.229726650621257e-05, | |
| "loss": 2.2718, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 6.20614409820754e-05, | |
| "loss": 2.274, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 6.180504769047129e-05, | |
| "loss": 2.2688, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 6.152826060655387e-05, | |
| "loss": 2.272, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 6.123126754364366e-05, | |
| "loss": 2.2807, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 6.091427002578765e-05, | |
| "loss": 2.2519, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 6.057748315101562e-05, | |
| "loss": 2.2648, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 6.0221135445385774e-05, | |
| "loss": 2.2739, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 5.984546870791885e-05, | |
| "loss": 2.2764, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0751953125, | |
| "learning_rate": 5.945073784652589e-05, | |
| "loss": 2.2514, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 5.9037210705040984e-05, | |
| "loss": 2.2544, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 5.860516788147634e-05, | |
| "loss": 2.2479, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.07763671875, | |
| "learning_rate": 5.815490253762313e-05, | |
| "loss": 2.253, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 5.7686720200127084e-05, | |
| "loss": 2.2528, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 5.7200938553174043e-05, | |
| "loss": 2.253, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 5.669788722292595e-05, | |
| "loss": 2.2571, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 5.617790755385372e-05, | |
| "loss": 2.2666, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 5.5641352377118605e-05, | |
| "loss": 2.2601, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 5.508858577115933e-05, | |
| "loss": 2.2471, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 5.451998281464741e-05, | |
| "loss": 2.2523, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 5.393592933197822e-05, | |
| "loss": 2.244, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 5.333682163147071e-05, | |
| "loss": 2.242, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 5.2723066236453086e-05, | |
| "loss": 2.2519, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 5.209507960941733e-05, | |
| "loss": 2.2521, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 5.145328786942933e-05, | |
| "loss": 2.2525, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 5.0798126502986685e-05, | |
| "loss": 2.2423, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 5.013004006852019e-05, | |
| "loss": 2.2444, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 4.944948189473962e-05, | |
| "loss": 2.2562, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 4.875691377302846e-05, | |
| "loss": 2.2443, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 4.805280564409623e-05, | |
| "loss": 2.2492, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 4.7337635279101233e-05, | |
| "loss": 2.2402, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 4.661188795545985e-05, | |
| "loss": 2.2402, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 4.5876056127562524e-05, | |
| "loss": 2.2464, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 4.5130639092619825e-05, | |
| "loss": 2.2518, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 4.437614265186536e-05, | |
| "loss": 2.2491, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 4.361307876734529e-05, | |
| "loss": 2.2474, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 4.2841965214527606e-05, | |
| "loss": 2.2634, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 4.206332523096655e-05, | |
| "loss": 2.2433, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 4.127768716126082e-05, | |
| "loss": 2.2468, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 4.0485584098546456e-05, | |
| "loss": 2.2399, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 3.968755352276755e-05, | |
| "loss": 2.2431, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 3.888413693597025e-05, | |
| "loss": 2.2309, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 3.8075879494867705e-05, | |
| "loss": 2.2298, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 3.726332964092504e-05, | |
| "loss": 2.2443, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 3.644703872821547e-05, | |
| "loss": 2.2481, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 3.5627560649300175e-05, | |
| "loss": 2.2375, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 3.4805451459385544e-05, | |
| "loss": 2.2368, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 3.398126899901305e-05, | |
| "loss": 2.2359, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 3.31555725155377e-05, | |
| "loss": 2.2472, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 3.232892228365181e-05, | |
| "loss": 2.2511, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 3.15018792252118e-05, | |
| "loss": 2.2273, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 3.067500452862575e-05, | |
| "loss": 2.2391, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 2.984885926806012e-05, | |
| "loss": 2.2424, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 2.9024004022724027e-05, | |
| "loss": 2.2435, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 2.8200998496489373e-05, | |
| "loss": 2.223, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 2.7380401138104845e-05, | |
| "loss": 2.2235, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 2.656276876226166e-05, | |
| "loss": 2.2379, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 2.5748656171768127e-05, | |
| "loss": 2.2232, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 2.4938615781089216e-05, | |
| "loss": 2.2492, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 2.413319724150689e-05, | |
| "loss": 2.2429, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 2.3332947068155364e-05, | |
| "loss": 2.2486, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 2.2538408269184268e-05, | |
| "loss": 2.2415, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 2.1750119977301616e-05, | |
| "loss": 2.2568, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 2.096861708394641e-05, | |
| "loss": 2.235, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 2.0194429876339054e-05, | |
| "loss": 2.2364, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 1.9428083677656066e-05, | |
| "loss": 2.2439, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 1.8670098490573132e-05, | |
| "loss": 2.238, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 1.792098864441825e-05, | |
| "loss": 2.2325, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 1.7181262446174615e-05, | |
| "loss": 2.2342, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 1.6451421835570044e-05, | |
| "loss": 2.2262, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 1.5731962044486608e-05, | |
| "loss": 2.2577, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 1.5023371260922157e-05, | |
| "loss": 2.2481, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.083984375, | |
| "learning_rate": 1.4326130297731294e-05, | |
| "loss": 2.229, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 1.3640712266370778e-05, | |
| "loss": 2.2365, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 1.2967582255870662e-05, | |
| "loss": 2.2276, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 1.2307197017249163e-05, | |
| "loss": 2.245, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 1.166000465358504e-05, | |
| "loss": 2.2327, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 1.1026444315958248e-05, | |
| "loss": 2.2257, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 1.0406945905464832e-05, | |
| "loss": 2.2355, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 9.801929781508377e-06, | |
| "loss": 2.2366, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 9.211806476565995e-06, | |
| "loss": 2.2307, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 8.63697641762235e-06, | |
| "loss": 2.2298, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.0830078125, | |
| "learning_rate": 8.077829654460684e-06, | |
| "loss": 2.24, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 7.534745594995376e-06, | |
| "loss": 2.2349, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 7.00809274782543e-06, | |
| "loss": 2.2522, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 6.498228472183709e-06, | |
| "loss": 2.2253, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 6.00549873545155e-06, | |
| "loss": 2.2319, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 5.530237878403297e-06, | |
| "loss": 2.23, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 5.0727683883399965e-06, | |
| "loss": 2.2366, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 4.63340068026638e-06, | |
| "loss": 2.238, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 4.212432886259269e-06, | |
| "loss": 2.2238, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 3.8101506531707373e-06, | |
| "loss": 2.2308, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 3.426826948802979e-06, | |
| "loss": 2.2305, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 3.0627218766865635e-06, | |
| "loss": 2.2364, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 2.7180824995877015e-06, | |
| "loss": 2.2435, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 2.3931426718643005e-06, | |
| "loss": 2.2347, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 2.0881228807845587e-06, | |
| "loss": 2.2589, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 1.8032300969157738e-06, | |
| "loss": 2.2323, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 1.5386576336848828e-06, | |
| "loss": 2.2285, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 1.2945850162060271e-06, | |
| "loss": 2.2368, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 1.0711778594641267e-06, | |
| "loss": 2.2417, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 8.685877559371869e-07, | |
| "loss": 2.2355, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 6.869521727335145e-07, | |
| "loss": 2.2285, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 5.26394358313689e-07, | |
| "loss": 2.2262, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.08154296875, | |
| "learning_rate": 3.8702325886057816e-07, | |
| "loss": 2.2289, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 2.6893344435409695e-07, | |
| "loss": 2.2428, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.08740234375, | |
| "learning_rate": 1.722050444009702e-07, | |
| "loss": 2.2248, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 9.690369386293054e-08, | |
| "loss": 2.2297, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0849609375, | |
| "learning_rate": 4.308048832030403e-08, | |
| "loss": 2.2344, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 1.0771949401241265e-08, | |
| "loss": 2.2296, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.087890625, | |
| "learning_rate": 0.0, | |
| "loss": 2.2321, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.2356441020965576, | |
| "eval_runtime": 186.994, | |
| "eval_samples_per_second": 25.787, | |
| "eval_steps_per_second": 3.225, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 670, | |
| "total_flos": 3.038889647317975e+17, | |
| "train_loss": 2.275535515884855, | |
| "train_runtime": 10859.5951, | |
| "train_samples_per_second": 7.898, | |
| "train_steps_per_second": 0.062 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 670, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 3.038889647317975e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |