| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 585, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0, |
| "forget_Q_A_ROUGE": 0.3188276274014988, |
| "step": 0 |
| }, |
| { |
| "epoch": 0.042735042735042736, |
| "grad_norm": 764.0, |
| "learning_rate": 4.3103448275862073e-07, |
| "loss": 14.2, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.08547008547008547, |
| "grad_norm": 652.0, |
| "learning_rate": 8.620689655172415e-07, |
| "loss": 14.0957, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.1282051282051282, |
| "grad_norm": 700.0, |
| "learning_rate": 1.2931034482758623e-06, |
| "loss": 13.9925, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.17094017094017094, |
| "grad_norm": 740.0, |
| "learning_rate": 1.724137931034483e-06, |
| "loss": 13.8799, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.21367521367521367, |
| "grad_norm": 692.0, |
| "learning_rate": 2.1551724137931035e-06, |
| "loss": 13.7433, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.2564102564102564, |
| "grad_norm": 608.0, |
| "learning_rate": 2.5862068965517246e-06, |
| "loss": 13.5505, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.29914529914529914, |
| "grad_norm": 676.0, |
| "learning_rate": 3.017241379310345e-06, |
| "loss": 13.3642, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.3418803418803419, |
| "grad_norm": 760.0, |
| "learning_rate": 3.448275862068966e-06, |
| "loss": 12.9769, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.38461538461538464, |
| "grad_norm": 728.0, |
| "learning_rate": 3.8793103448275865e-06, |
| "loss": 11.7681, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.42735042735042733, |
| "grad_norm": 728.0, |
| "learning_rate": 4.310344827586207e-06, |
| "loss": 10.5377, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.4700854700854701, |
| "grad_norm": 676.0, |
| "learning_rate": 4.741379310344828e-06, |
| "loss": 8.6128, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.5128205128205128, |
| "grad_norm": 664.0, |
| "learning_rate": 5.172413793103449e-06, |
| "loss": 6.0024, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.5555555555555556, |
| "grad_norm": 312.0, |
| "learning_rate": 5.603448275862069e-06, |
| "loss": 2.7774, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.5982905982905983, |
| "grad_norm": 96.5, |
| "learning_rate": 6.03448275862069e-06, |
| "loss": 1.5545, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.6410256410256411, |
| "grad_norm": 79.0, |
| "learning_rate": 6.465517241379311e-06, |
| "loss": 1.4243, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.6837606837606838, |
| "grad_norm": 77.0, |
| "learning_rate": 6.896551724137932e-06, |
| "loss": 1.159, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.7264957264957265, |
| "grad_norm": 81.0, |
| "learning_rate": 7.327586206896552e-06, |
| "loss": 1.0177, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.7692307692307693, |
| "grad_norm": 46.25, |
| "learning_rate": 7.758620689655173e-06, |
| "loss": 0.849, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.811965811965812, |
| "grad_norm": 31.375, |
| "learning_rate": 8.189655172413794e-06, |
| "loss": 0.816, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.8547008547008547, |
| "grad_norm": 39.5, |
| "learning_rate": 8.620689655172414e-06, |
| "loss": 0.728, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.8974358974358975, |
| "grad_norm": 99.0, |
| "learning_rate": 9.051724137931036e-06, |
| "loss": 0.7868, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.9401709401709402, |
| "grad_norm": 26.875, |
| "learning_rate": 9.482758620689655e-06, |
| "loss": 0.5595, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.9829059829059829, |
| "grad_norm": 22.625, |
| "learning_rate": 9.913793103448277e-06, |
| "loss": 0.5535, |
| "step": 115 |
| }, |
| { |
| "epoch": 1.0, |
| "forget_Q_A_ROUGE": 0.18181675542335068, |
| "step": 117 |
| }, |
| { |
| "epoch": 1.0256410256410255, |
| "grad_norm": 17.75, |
| "learning_rate": 9.914712153518125e-06, |
| "loss": 0.5077, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.0683760683760684, |
| "grad_norm": 12.125, |
| "learning_rate": 9.80810234541578e-06, |
| "loss": 0.4655, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.1111111111111112, |
| "grad_norm": 31.25, |
| "learning_rate": 9.701492537313434e-06, |
| "loss": 0.4229, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.1538461538461537, |
| "grad_norm": 24.125, |
| "learning_rate": 9.594882729211089e-06, |
| "loss": 0.4191, |
| "step": 135 |
| }, |
| { |
| "epoch": 1.1965811965811965, |
| "grad_norm": 21.25, |
| "learning_rate": 9.488272921108744e-06, |
| "loss": 0.3993, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.2393162393162394, |
| "grad_norm": 10.75, |
| "learning_rate": 9.381663113006397e-06, |
| "loss": 0.3551, |
| "step": 145 |
| }, |
| { |
| "epoch": 1.282051282051282, |
| "grad_norm": 7.71875, |
| "learning_rate": 9.275053304904051e-06, |
| "loss": 0.3124, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.3247863247863247, |
| "grad_norm": 8.0625, |
| "learning_rate": 9.168443496801706e-06, |
| "loss": 0.3136, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.3675213675213675, |
| "grad_norm": 15.4375, |
| "learning_rate": 9.06183368869936e-06, |
| "loss": 0.3554, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.4102564102564101, |
| "grad_norm": 7.46875, |
| "learning_rate": 8.955223880597016e-06, |
| "loss": 0.2964, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.452991452991453, |
| "grad_norm": 62.0, |
| "learning_rate": 8.84861407249467e-06, |
| "loss": 0.2874, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.4957264957264957, |
| "grad_norm": 13.9375, |
| "learning_rate": 8.742004264392325e-06, |
| "loss": 0.2957, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.5384615384615383, |
| "grad_norm": 8.125, |
| "learning_rate": 8.63539445628998e-06, |
| "loss": 0.3001, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.5811965811965814, |
| "grad_norm": 6.625, |
| "learning_rate": 8.528784648187633e-06, |
| "loss": 0.3171, |
| "step": 185 |
| }, |
| { |
| "epoch": 1.623931623931624, |
| "grad_norm": 6.0625, |
| "learning_rate": 8.42217484008529e-06, |
| "loss": 0.2736, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.6666666666666665, |
| "grad_norm": 7.59375, |
| "learning_rate": 8.315565031982942e-06, |
| "loss": 0.3145, |
| "step": 195 |
| }, |
| { |
| "epoch": 1.7094017094017095, |
| "grad_norm": 13.4375, |
| "learning_rate": 8.208955223880599e-06, |
| "loss": 0.2538, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.7521367521367521, |
| "grad_norm": 18.75, |
| "learning_rate": 8.102345415778252e-06, |
| "loss": 0.2419, |
| "step": 205 |
| }, |
| { |
| "epoch": 1.7948717948717947, |
| "grad_norm": 7.09375, |
| "learning_rate": 7.995735607675907e-06, |
| "loss": 0.2446, |
| "step": 210 |
| }, |
| { |
| "epoch": 1.8376068376068377, |
| "grad_norm": 5.5625, |
| "learning_rate": 7.889125799573561e-06, |
| "loss": 0.2172, |
| "step": 215 |
| }, |
| { |
| "epoch": 1.8803418803418803, |
| "grad_norm": 6.75, |
| "learning_rate": 7.782515991471216e-06, |
| "loss": 0.2354, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.9230769230769231, |
| "grad_norm": 6.28125, |
| "learning_rate": 7.67590618336887e-06, |
| "loss": 0.2031, |
| "step": 225 |
| }, |
| { |
| "epoch": 1.965811965811966, |
| "grad_norm": 7.09375, |
| "learning_rate": 7.569296375266525e-06, |
| "loss": 0.2396, |
| "step": 230 |
| }, |
| { |
| "epoch": 2.0, |
| "forget_Q_A_ROUGE": 0.16664429628381783, |
| "step": 234 |
| }, |
| { |
| "epoch": 2.0085470085470085, |
| "grad_norm": 7.03125, |
| "learning_rate": 7.46268656716418e-06, |
| "loss": 0.2064, |
| "step": 235 |
| }, |
| { |
| "epoch": 2.051282051282051, |
| "grad_norm": 5.5625, |
| "learning_rate": 7.356076759061834e-06, |
| "loss": 0.1929, |
| "step": 240 |
| }, |
| { |
| "epoch": 2.094017094017094, |
| "grad_norm": 6.4375, |
| "learning_rate": 7.249466950959488e-06, |
| "loss": 0.1718, |
| "step": 245 |
| }, |
| { |
| "epoch": 2.1367521367521367, |
| "grad_norm": 6.1875, |
| "learning_rate": 7.1428571428571436e-06, |
| "loss": 0.2016, |
| "step": 250 |
| }, |
| { |
| "epoch": 2.1794871794871793, |
| "grad_norm": 6.0, |
| "learning_rate": 7.0362473347547975e-06, |
| "loss": 0.1856, |
| "step": 255 |
| }, |
| { |
| "epoch": 2.2222222222222223, |
| "grad_norm": 5.0625, |
| "learning_rate": 6.929637526652453e-06, |
| "loss": 0.1694, |
| "step": 260 |
| }, |
| { |
| "epoch": 2.264957264957265, |
| "grad_norm": 5.40625, |
| "learning_rate": 6.823027718550107e-06, |
| "loss": 0.1571, |
| "step": 265 |
| }, |
| { |
| "epoch": 2.3076923076923075, |
| "grad_norm": 7.40625, |
| "learning_rate": 6.7164179104477625e-06, |
| "loss": 0.1763, |
| "step": 270 |
| }, |
| { |
| "epoch": 2.3504273504273505, |
| "grad_norm": 5.5, |
| "learning_rate": 6.609808102345416e-06, |
| "loss": 0.1777, |
| "step": 275 |
| }, |
| { |
| "epoch": 2.393162393162393, |
| "grad_norm": 6.0, |
| "learning_rate": 6.50319829424307e-06, |
| "loss": 0.1754, |
| "step": 280 |
| }, |
| { |
| "epoch": 2.435897435897436, |
| "grad_norm": 5.8125, |
| "learning_rate": 6.396588486140726e-06, |
| "loss": 0.158, |
| "step": 285 |
| }, |
| { |
| "epoch": 2.4786324786324787, |
| "grad_norm": 6.59375, |
| "learning_rate": 6.28997867803838e-06, |
| "loss": 0.1666, |
| "step": 290 |
| }, |
| { |
| "epoch": 2.5213675213675213, |
| "grad_norm": 5.28125, |
| "learning_rate": 6.183368869936035e-06, |
| "loss": 0.1561, |
| "step": 295 |
| }, |
| { |
| "epoch": 2.564102564102564, |
| "grad_norm": 5.25, |
| "learning_rate": 6.076759061833689e-06, |
| "loss": 0.1373, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.606837606837607, |
| "grad_norm": 6.46875, |
| "learning_rate": 5.970149253731343e-06, |
| "loss": 0.1559, |
| "step": 305 |
| }, |
| { |
| "epoch": 2.6495726495726495, |
| "grad_norm": 5.90625, |
| "learning_rate": 5.863539445628999e-06, |
| "loss": 0.1453, |
| "step": 310 |
| }, |
| { |
| "epoch": 2.6923076923076925, |
| "grad_norm": 4.9375, |
| "learning_rate": 5.756929637526653e-06, |
| "loss": 0.153, |
| "step": 315 |
| }, |
| { |
| "epoch": 2.735042735042735, |
| "grad_norm": 5.5, |
| "learning_rate": 5.650319829424308e-06, |
| "loss": 0.1456, |
| "step": 320 |
| }, |
| { |
| "epoch": 2.7777777777777777, |
| "grad_norm": 6.6875, |
| "learning_rate": 5.543710021321962e-06, |
| "loss": 0.1375, |
| "step": 325 |
| }, |
| { |
| "epoch": 2.8205128205128203, |
| "grad_norm": 5.125, |
| "learning_rate": 5.437100213219617e-06, |
| "loss": 0.1433, |
| "step": 330 |
| }, |
| { |
| "epoch": 2.8632478632478633, |
| "grad_norm": 6.34375, |
| "learning_rate": 5.3304904051172716e-06, |
| "loss": 0.1354, |
| "step": 335 |
| }, |
| { |
| "epoch": 2.905982905982906, |
| "grad_norm": 7.5, |
| "learning_rate": 5.2238805970149255e-06, |
| "loss": 0.1252, |
| "step": 340 |
| }, |
| { |
| "epoch": 2.948717948717949, |
| "grad_norm": 5.15625, |
| "learning_rate": 5.11727078891258e-06, |
| "loss": 0.1273, |
| "step": 345 |
| }, |
| { |
| "epoch": 2.9914529914529915, |
| "grad_norm": 4.53125, |
| "learning_rate": 5.010660980810235e-06, |
| "loss": 0.1179, |
| "step": 350 |
| }, |
| { |
| "epoch": 3.0, |
| "forget_Q_A_ROUGE": 0.16380368545856266, |
| "step": 351 |
| }, |
| { |
| "epoch": 3.034188034188034, |
| "grad_norm": 5.4375, |
| "learning_rate": 4.90405117270789e-06, |
| "loss": 0.1379, |
| "step": 355 |
| }, |
| { |
| "epoch": 3.076923076923077, |
| "grad_norm": 5.34375, |
| "learning_rate": 4.797441364605544e-06, |
| "loss": 0.1277, |
| "step": 360 |
| }, |
| { |
| "epoch": 3.1196581196581197, |
| "grad_norm": 5.0, |
| "learning_rate": 4.690831556503198e-06, |
| "loss": 0.1305, |
| "step": 365 |
| }, |
| { |
| "epoch": 3.1623931623931623, |
| "grad_norm": 4.71875, |
| "learning_rate": 4.584221748400853e-06, |
| "loss": 0.118, |
| "step": 370 |
| }, |
| { |
| "epoch": 3.2051282051282053, |
| "grad_norm": 4.84375, |
| "learning_rate": 4.477611940298508e-06, |
| "loss": 0.1176, |
| "step": 375 |
| }, |
| { |
| "epoch": 3.247863247863248, |
| "grad_norm": 4.75, |
| "learning_rate": 4.3710021321961625e-06, |
| "loss": 0.12, |
| "step": 380 |
| }, |
| { |
| "epoch": 3.2905982905982905, |
| "grad_norm": 4.34375, |
| "learning_rate": 4.264392324093816e-06, |
| "loss": 0.1288, |
| "step": 385 |
| }, |
| { |
| "epoch": 3.3333333333333335, |
| "grad_norm": 5.34375, |
| "learning_rate": 4.157782515991471e-06, |
| "loss": 0.1244, |
| "step": 390 |
| }, |
| { |
| "epoch": 3.376068376068376, |
| "grad_norm": 4.4375, |
| "learning_rate": 4.051172707889126e-06, |
| "loss": 0.1189, |
| "step": 395 |
| }, |
| { |
| "epoch": 3.4188034188034186, |
| "grad_norm": 4.71875, |
| "learning_rate": 3.944562899786781e-06, |
| "loss": 0.1073, |
| "step": 400 |
| }, |
| { |
| "epoch": 3.4615384615384617, |
| "grad_norm": 4.6875, |
| "learning_rate": 3.837953091684435e-06, |
| "loss": 0.1056, |
| "step": 405 |
| }, |
| { |
| "epoch": 3.5042735042735043, |
| "grad_norm": 3.96875, |
| "learning_rate": 3.73134328358209e-06, |
| "loss": 0.0968, |
| "step": 410 |
| }, |
| { |
| "epoch": 3.547008547008547, |
| "grad_norm": 4.40625, |
| "learning_rate": 3.624733475479744e-06, |
| "loss": 0.114, |
| "step": 415 |
| }, |
| { |
| "epoch": 3.58974358974359, |
| "grad_norm": 4.34375, |
| "learning_rate": 3.5181236673773987e-06, |
| "loss": 0.0984, |
| "step": 420 |
| }, |
| { |
| "epoch": 3.6324786324786325, |
| "grad_norm": 4.0625, |
| "learning_rate": 3.4115138592750535e-06, |
| "loss": 0.1039, |
| "step": 425 |
| }, |
| { |
| "epoch": 3.6752136752136755, |
| "grad_norm": 5.15625, |
| "learning_rate": 3.304904051172708e-06, |
| "loss": 0.1201, |
| "step": 430 |
| }, |
| { |
| "epoch": 3.717948717948718, |
| "grad_norm": 4.78125, |
| "learning_rate": 3.198294243070363e-06, |
| "loss": 0.1138, |
| "step": 435 |
| }, |
| { |
| "epoch": 3.7606837606837606, |
| "grad_norm": 4.4375, |
| "learning_rate": 3.0916844349680177e-06, |
| "loss": 0.1009, |
| "step": 440 |
| }, |
| { |
| "epoch": 3.8034188034188032, |
| "grad_norm": 5.0, |
| "learning_rate": 2.9850746268656716e-06, |
| "loss": 0.1073, |
| "step": 445 |
| }, |
| { |
| "epoch": 3.8461538461538463, |
| "grad_norm": 3.890625, |
| "learning_rate": 2.8784648187633263e-06, |
| "loss": 0.1065, |
| "step": 450 |
| }, |
| { |
| "epoch": 3.888888888888889, |
| "grad_norm": 5.65625, |
| "learning_rate": 2.771855010660981e-06, |
| "loss": 0.1164, |
| "step": 455 |
| }, |
| { |
| "epoch": 3.931623931623932, |
| "grad_norm": 4.0625, |
| "learning_rate": 2.6652452025586358e-06, |
| "loss": 0.0976, |
| "step": 460 |
| }, |
| { |
| "epoch": 3.9743589743589745, |
| "grad_norm": 4.375, |
| "learning_rate": 2.55863539445629e-06, |
| "loss": 0.1027, |
| "step": 465 |
| }, |
| { |
| "epoch": 4.0, |
| "forget_Q_A_ROUGE": 0.1577463934994216, |
| "step": 468 |
| }, |
| { |
| "epoch": 4.017094017094017, |
| "grad_norm": 5.03125, |
| "learning_rate": 2.452025586353945e-06, |
| "loss": 0.108, |
| "step": 470 |
| }, |
| { |
| "epoch": 4.05982905982906, |
| "grad_norm": 3.53125, |
| "learning_rate": 2.345415778251599e-06, |
| "loss": 0.0954, |
| "step": 475 |
| }, |
| { |
| "epoch": 4.102564102564102, |
| "grad_norm": 4.53125, |
| "learning_rate": 2.238805970149254e-06, |
| "loss": 0.107, |
| "step": 480 |
| }, |
| { |
| "epoch": 4.145299145299146, |
| "grad_norm": 4.625, |
| "learning_rate": 2.132196162046908e-06, |
| "loss": 0.1001, |
| "step": 485 |
| }, |
| { |
| "epoch": 4.188034188034188, |
| "grad_norm": 5.28125, |
| "learning_rate": 2.025586353944563e-06, |
| "loss": 0.109, |
| "step": 490 |
| }, |
| { |
| "epoch": 4.230769230769231, |
| "grad_norm": 4.4375, |
| "learning_rate": 1.9189765458422177e-06, |
| "loss": 0.1133, |
| "step": 495 |
| }, |
| { |
| "epoch": 4.273504273504273, |
| "grad_norm": 4.0, |
| "learning_rate": 1.812366737739872e-06, |
| "loss": 0.092, |
| "step": 500 |
| }, |
| { |
| "epoch": 4.316239316239316, |
| "grad_norm": 4.34375, |
| "learning_rate": 1.7057569296375267e-06, |
| "loss": 0.1034, |
| "step": 505 |
| }, |
| { |
| "epoch": 4.358974358974359, |
| "grad_norm": 4.40625, |
| "learning_rate": 1.5991471215351815e-06, |
| "loss": 0.0986, |
| "step": 510 |
| }, |
| { |
| "epoch": 4.401709401709402, |
| "grad_norm": 4.09375, |
| "learning_rate": 1.4925373134328358e-06, |
| "loss": 0.1168, |
| "step": 515 |
| }, |
| { |
| "epoch": 4.444444444444445, |
| "grad_norm": 4.90625, |
| "learning_rate": 1.3859275053304905e-06, |
| "loss": 0.1033, |
| "step": 520 |
| }, |
| { |
| "epoch": 4.487179487179487, |
| "grad_norm": 4.65625, |
| "learning_rate": 1.279317697228145e-06, |
| "loss": 0.1006, |
| "step": 525 |
| }, |
| { |
| "epoch": 4.52991452991453, |
| "grad_norm": 5.21875, |
| "learning_rate": 1.1727078891257996e-06, |
| "loss": 0.109, |
| "step": 530 |
| }, |
| { |
| "epoch": 4.572649572649572, |
| "grad_norm": 4.40625, |
| "learning_rate": 1.066098081023454e-06, |
| "loss": 0.1129, |
| "step": 535 |
| }, |
| { |
| "epoch": 4.615384615384615, |
| "grad_norm": 5.34375, |
| "learning_rate": 9.594882729211088e-07, |
| "loss": 0.0991, |
| "step": 540 |
| }, |
| { |
| "epoch": 4.6581196581196584, |
| "grad_norm": 4.84375, |
| "learning_rate": 8.528784648187634e-07, |
| "loss": 0.1047, |
| "step": 545 |
| }, |
| { |
| "epoch": 4.700854700854701, |
| "grad_norm": 5.59375, |
| "learning_rate": 7.462686567164179e-07, |
| "loss": 0.1099, |
| "step": 550 |
| }, |
| { |
| "epoch": 4.743589743589744, |
| "grad_norm": 3.984375, |
| "learning_rate": 6.396588486140725e-07, |
| "loss": 0.0981, |
| "step": 555 |
| }, |
| { |
| "epoch": 4.786324786324786, |
| "grad_norm": 3.9375, |
| "learning_rate": 5.33049040511727e-07, |
| "loss": 0.0975, |
| "step": 560 |
| }, |
| { |
| "epoch": 4.829059829059829, |
| "grad_norm": 5.40625, |
| "learning_rate": 4.264392324093817e-07, |
| "loss": 0.1066, |
| "step": 565 |
| }, |
| { |
| "epoch": 4.871794871794872, |
| "grad_norm": 4.53125, |
| "learning_rate": 3.1982942430703626e-07, |
| "loss": 0.1064, |
| "step": 570 |
| }, |
| { |
| "epoch": 4.914529914529915, |
| "grad_norm": 3.953125, |
| "learning_rate": 2.1321961620469084e-07, |
| "loss": 0.0975, |
| "step": 575 |
| }, |
| { |
| "epoch": 4.957264957264957, |
| "grad_norm": 3.96875, |
| "learning_rate": 1.0660980810234542e-07, |
| "loss": 0.1096, |
| "step": 580 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 4.15625, |
| "learning_rate": 0.0, |
| "loss": 0.0957, |
| "step": 585 |
| }, |
| { |
| "epoch": 5.0, |
| "forget_Q_A_ROUGE": 0.16323620599043673, |
| "step": 585 |
| }, |
| { |
| "epoch": 5.0, |
| "step": 585, |
| "total_flos": 0.0, |
| "train_loss": 1.4970051565231421, |
| "train_runtime": 6804.7106, |
| "train_samples_per_second": 5.499, |
| "train_steps_per_second": 0.086 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 585, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
