| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.3169845594913716, | |
| "eval_steps": 500, | |
| "global_step": 4350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009082652134423252, | |
| "grad_norm": 2.5449585914611816, | |
| "learning_rate": 4.9848622464426284e-05, | |
| "loss": 8.7191, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.018165304268846504, | |
| "grad_norm": 2.3144371509552, | |
| "learning_rate": 4.969724492885256e-05, | |
| "loss": 7.4698, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.027247956403269755, | |
| "grad_norm": 2.304499626159668, | |
| "learning_rate": 4.954586739327884e-05, | |
| "loss": 6.5589, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03633060853769301, | |
| "grad_norm": 2.272608757019043, | |
| "learning_rate": 4.9394489857705115e-05, | |
| "loss": 6.2425, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.045413260672116255, | |
| "grad_norm": 2.46329402923584, | |
| "learning_rate": 4.9243112322131396e-05, | |
| "loss": 6.1459, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05449591280653951, | |
| "grad_norm": 1.8283530473709106, | |
| "learning_rate": 4.909173478655768e-05, | |
| "loss": 5.969, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.06357856494096276, | |
| "grad_norm": 2.1723110675811768, | |
| "learning_rate": 4.894035725098395e-05, | |
| "loss": 6.008, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.07266121707538602, | |
| "grad_norm": 2.5368807315826416, | |
| "learning_rate": 4.878897971541024e-05, | |
| "loss": 5.8783, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08174386920980926, | |
| "grad_norm": 2.3222858905792236, | |
| "learning_rate": 4.8637602179836515e-05, | |
| "loss": 5.825, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.09082652134423251, | |
| "grad_norm": 2.557065010070801, | |
| "learning_rate": 4.8486224644262796e-05, | |
| "loss": 5.76, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.09990917347865577, | |
| "grad_norm": 2.4016597270965576, | |
| "learning_rate": 4.833484710868907e-05, | |
| "loss": 5.7039, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.10899182561307902, | |
| "grad_norm": 2.6895477771759033, | |
| "learning_rate": 4.818346957311535e-05, | |
| "loss": 5.5843, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.11807447774750227, | |
| "grad_norm": 2.741234064102173, | |
| "learning_rate": 4.8032092037541634e-05, | |
| "loss": 5.6376, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1271571298819255, | |
| "grad_norm": 2.8266804218292236, | |
| "learning_rate": 4.788071450196791e-05, | |
| "loss": 5.5649, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.1362397820163488, | |
| "grad_norm": 2.792654275894165, | |
| "learning_rate": 4.772933696639419e-05, | |
| "loss": 5.3651, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.14532243415077203, | |
| "grad_norm": 2.7088894844055176, | |
| "learning_rate": 4.757795943082047e-05, | |
| "loss": 5.4921, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.15440508628519528, | |
| "grad_norm": 2.627201795578003, | |
| "learning_rate": 4.7426581895246746e-05, | |
| "loss": 5.461, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.16348773841961853, | |
| "grad_norm": 2.6373610496520996, | |
| "learning_rate": 4.727520435967303e-05, | |
| "loss": 5.3973, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.17257039055404177, | |
| "grad_norm": 2.772226095199585, | |
| "learning_rate": 4.71238268240993e-05, | |
| "loss": 5.3618, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.18165304268846502, | |
| "grad_norm": 2.6005172729492188, | |
| "learning_rate": 4.6972449288525583e-05, | |
| "loss": 5.4365, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1907356948228883, | |
| "grad_norm": 4.7815260887146, | |
| "learning_rate": 4.6821071752951865e-05, | |
| "loss": 5.3225, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.19981834695731154, | |
| "grad_norm": 2.5871763229370117, | |
| "learning_rate": 4.6669694217378146e-05, | |
| "loss": 5.3615, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.2089009990917348, | |
| "grad_norm": 2.686840534210205, | |
| "learning_rate": 4.651831668180443e-05, | |
| "loss": 5.3201, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.21798365122615804, | |
| "grad_norm": 2.6963067054748535, | |
| "learning_rate": 4.63669391462307e-05, | |
| "loss": 5.1972, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.22706630336058128, | |
| "grad_norm": 2.9284744262695312, | |
| "learning_rate": 4.6215561610656984e-05, | |
| "loss": 5.3031, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.23614895549500453, | |
| "grad_norm": 2.7302122116088867, | |
| "learning_rate": 4.606418407508326e-05, | |
| "loss": 5.2057, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2452316076294278, | |
| "grad_norm": 2.5760107040405273, | |
| "learning_rate": 4.591280653950954e-05, | |
| "loss": 5.1767, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.254314259763851, | |
| "grad_norm": 2.9804234504699707, | |
| "learning_rate": 4.576142900393582e-05, | |
| "loss": 5.1875, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.2633969118982743, | |
| "grad_norm": 3.311448812484741, | |
| "learning_rate": 4.5610051468362096e-05, | |
| "loss": 5.0712, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.2724795640326976, | |
| "grad_norm": 2.67448091506958, | |
| "learning_rate": 4.545867393278838e-05, | |
| "loss": 5.1241, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2815622161671208, | |
| "grad_norm": 2.8352444171905518, | |
| "learning_rate": 4.530729639721465e-05, | |
| "loss": 5.1732, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.29064486830154407, | |
| "grad_norm": 2.5969910621643066, | |
| "learning_rate": 4.515591886164093e-05, | |
| "loss": 5.0828, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.2997275204359673, | |
| "grad_norm": 2.8792121410369873, | |
| "learning_rate": 4.5004541326067215e-05, | |
| "loss": 5.0844, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.30881017257039056, | |
| "grad_norm": 2.9506993293762207, | |
| "learning_rate": 4.485316379049349e-05, | |
| "loss": 5.1764, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.3178928247048138, | |
| "grad_norm": 2.8818390369415283, | |
| "learning_rate": 4.470178625491977e-05, | |
| "loss": 5.0663, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.32697547683923706, | |
| "grad_norm": 3.128511667251587, | |
| "learning_rate": 4.4550408719346046e-05, | |
| "loss": 5.1026, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.33605812897366033, | |
| "grad_norm": 3.0155856609344482, | |
| "learning_rate": 4.4399031183772334e-05, | |
| "loss": 5.0686, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.34514078110808355, | |
| "grad_norm": 2.811448097229004, | |
| "learning_rate": 4.424765364819861e-05, | |
| "loss": 5.0351, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.3542234332425068, | |
| "grad_norm": 2.9916000366210938, | |
| "learning_rate": 4.409627611262489e-05, | |
| "loss": 5.1651, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.36330608537693004, | |
| "grad_norm": 2.9689950942993164, | |
| "learning_rate": 4.394489857705117e-05, | |
| "loss": 5.1457, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.3723887375113533, | |
| "grad_norm": 2.7896862030029297, | |
| "learning_rate": 4.3793521041477446e-05, | |
| "loss": 5.0049, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.3814713896457766, | |
| "grad_norm": 2.790712833404541, | |
| "learning_rate": 4.364214350590373e-05, | |
| "loss": 4.9943, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.3905540417801998, | |
| "grad_norm": 2.9977900981903076, | |
| "learning_rate": 4.349076597033e-05, | |
| "loss": 4.996, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.3996366939146231, | |
| "grad_norm": 3.504183530807495, | |
| "learning_rate": 4.333938843475628e-05, | |
| "loss": 4.9611, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.4087193460490463, | |
| "grad_norm": 2.737821578979492, | |
| "learning_rate": 4.3188010899182565e-05, | |
| "loss": 4.9541, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.4178019981834696, | |
| "grad_norm": 3.0585217475891113, | |
| "learning_rate": 4.303663336360884e-05, | |
| "loss": 4.9014, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.4268846503178928, | |
| "grad_norm": 3.004413604736328, | |
| "learning_rate": 4.288525582803512e-05, | |
| "loss": 4.9703, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.4359673024523161, | |
| "grad_norm": 2.9328274726867676, | |
| "learning_rate": 4.27338782924614e-05, | |
| "loss": 4.9637, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.44504995458673935, | |
| "grad_norm": 2.93721604347229, | |
| "learning_rate": 4.258250075688768e-05, | |
| "loss": 4.8024, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.45413260672116257, | |
| "grad_norm": 3.0333001613616943, | |
| "learning_rate": 4.243112322131396e-05, | |
| "loss": 4.8555, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.46321525885558584, | |
| "grad_norm": 3.3445775508880615, | |
| "learning_rate": 4.227974568574024e-05, | |
| "loss": 4.8035, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.47229791099000906, | |
| "grad_norm": 2.9364359378814697, | |
| "learning_rate": 4.212836815016652e-05, | |
| "loss": 4.9296, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.48138056312443234, | |
| "grad_norm": 2.755453586578369, | |
| "learning_rate": 4.1976990614592796e-05, | |
| "loss": 4.8051, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.4904632152588556, | |
| "grad_norm": 3.0365066528320312, | |
| "learning_rate": 4.182561307901908e-05, | |
| "loss": 4.7833, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.49954586739327883, | |
| "grad_norm": 3.2632575035095215, | |
| "learning_rate": 4.167423554344536e-05, | |
| "loss": 4.837, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.508628519527702, | |
| "grad_norm": 3.310817003250122, | |
| "learning_rate": 4.152285800787163e-05, | |
| "loss": 4.7417, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.5177111716621253, | |
| "grad_norm": 3.121156692504883, | |
| "learning_rate": 4.1371480472297915e-05, | |
| "loss": 4.7791, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.5267938237965486, | |
| "grad_norm": 3.200591564178467, | |
| "learning_rate": 4.122010293672419e-05, | |
| "loss": 4.8619, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.5358764759309719, | |
| "grad_norm": 3.1420202255249023, | |
| "learning_rate": 4.106872540115047e-05, | |
| "loss": 4.7576, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.5449591280653951, | |
| "grad_norm": 3.2239160537719727, | |
| "learning_rate": 4.091734786557675e-05, | |
| "loss": 4.7767, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5540417801998183, | |
| "grad_norm": 2.9624414443969727, | |
| "learning_rate": 4.076597033000303e-05, | |
| "loss": 4.8608, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.5631244323342416, | |
| "grad_norm": 3.14367938041687, | |
| "learning_rate": 4.061459279442931e-05, | |
| "loss": 4.7909, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.5722070844686649, | |
| "grad_norm": 3.664564371109009, | |
| "learning_rate": 4.046321525885558e-05, | |
| "loss": 4.7325, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.5812897366030881, | |
| "grad_norm": 2.9251296520233154, | |
| "learning_rate": 4.0311837723281864e-05, | |
| "loss": 4.8017, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.5903723887375113, | |
| "grad_norm": 2.8796215057373047, | |
| "learning_rate": 4.0160460187708146e-05, | |
| "loss": 4.7124, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5994550408719346, | |
| "grad_norm": 3.0257513523101807, | |
| "learning_rate": 4.000908265213443e-05, | |
| "loss": 4.7311, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.6085376930063578, | |
| "grad_norm": 3.096799612045288, | |
| "learning_rate": 3.985770511656071e-05, | |
| "loss": 4.6568, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.6176203451407811, | |
| "grad_norm": 3.1430232524871826, | |
| "learning_rate": 3.970632758098698e-05, | |
| "loss": 4.6451, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.6267029972752044, | |
| "grad_norm": 3.0216684341430664, | |
| "learning_rate": 3.9554950045413265e-05, | |
| "loss": 4.6565, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.6357856494096276, | |
| "grad_norm": 3.0199525356292725, | |
| "learning_rate": 3.940357250983954e-05, | |
| "loss": 4.6988, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.6448683015440508, | |
| "grad_norm": 2.9998953342437744, | |
| "learning_rate": 3.925219497426582e-05, | |
| "loss": 4.6654, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.6539509536784741, | |
| "grad_norm": 3.15533447265625, | |
| "learning_rate": 3.91008174386921e-05, | |
| "loss": 4.616, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.6630336058128974, | |
| "grad_norm": 2.8745930194854736, | |
| "learning_rate": 3.894943990311838e-05, | |
| "loss": 4.649, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.6721162579473207, | |
| "grad_norm": 3.0759665966033936, | |
| "learning_rate": 3.879806236754466e-05, | |
| "loss": 4.6054, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.6811989100817438, | |
| "grad_norm": 3.0508482456207275, | |
| "learning_rate": 3.864668483197093e-05, | |
| "loss": 4.4922, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6902815622161671, | |
| "grad_norm": 2.9260127544403076, | |
| "learning_rate": 3.8495307296397214e-05, | |
| "loss": 4.6469, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.6993642143505904, | |
| "grad_norm": 2.924952268600464, | |
| "learning_rate": 3.8343929760823496e-05, | |
| "loss": 4.6164, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.7084468664850136, | |
| "grad_norm": 3.056288480758667, | |
| "learning_rate": 3.819255222524977e-05, | |
| "loss": 4.5877, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.7175295186194369, | |
| "grad_norm": 4.257227420806885, | |
| "learning_rate": 3.804117468967605e-05, | |
| "loss": 4.6301, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.7266121707538601, | |
| "grad_norm": 3.282137155532837, | |
| "learning_rate": 3.788979715410233e-05, | |
| "loss": 4.4623, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7356948228882834, | |
| "grad_norm": 2.945059299468994, | |
| "learning_rate": 3.7738419618528615e-05, | |
| "loss": 4.6267, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.7447774750227066, | |
| "grad_norm": 3.1374645233154297, | |
| "learning_rate": 3.7587042082954896e-05, | |
| "loss": 4.6835, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.7538601271571299, | |
| "grad_norm": 3.21016001701355, | |
| "learning_rate": 3.743566454738117e-05, | |
| "loss": 4.5581, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.7629427792915532, | |
| "grad_norm": 2.8072383403778076, | |
| "learning_rate": 3.728428701180745e-05, | |
| "loss": 4.571, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.7720254314259763, | |
| "grad_norm": 2.9735002517700195, | |
| "learning_rate": 3.713290947623373e-05, | |
| "loss": 4.5013, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7811080835603996, | |
| "grad_norm": 3.182706832885742, | |
| "learning_rate": 3.698153194066001e-05, | |
| "loss": 4.534, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.7901907356948229, | |
| "grad_norm": 2.958193778991699, | |
| "learning_rate": 3.683015440508629e-05, | |
| "loss": 4.5697, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.7992733878292462, | |
| "grad_norm": 2.950946569442749, | |
| "learning_rate": 3.6678776869512564e-05, | |
| "loss": 4.6066, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.8083560399636693, | |
| "grad_norm": 2.9701859951019287, | |
| "learning_rate": 3.6527399333938846e-05, | |
| "loss": 4.5934, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.8174386920980926, | |
| "grad_norm": 3.2177681922912598, | |
| "learning_rate": 3.637602179836512e-05, | |
| "loss": 4.5418, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8265213442325159, | |
| "grad_norm": 2.7435505390167236, | |
| "learning_rate": 3.62246442627914e-05, | |
| "loss": 4.5485, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.8356039963669392, | |
| "grad_norm": 3.4409849643707275, | |
| "learning_rate": 3.607326672721768e-05, | |
| "loss": 4.4268, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.8446866485013624, | |
| "grad_norm": 3.803256034851074, | |
| "learning_rate": 3.592188919164396e-05, | |
| "loss": 4.5643, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.8537693006357856, | |
| "grad_norm": 3.0399341583251953, | |
| "learning_rate": 3.5770511656070246e-05, | |
| "loss": 4.4783, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.8628519527702089, | |
| "grad_norm": 2.9948980808258057, | |
| "learning_rate": 3.561913412049652e-05, | |
| "loss": 4.4929, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.8719346049046321, | |
| "grad_norm": 3.400299549102783, | |
| "learning_rate": 3.54677565849228e-05, | |
| "loss": 4.4803, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.8810172570390554, | |
| "grad_norm": 2.9282257556915283, | |
| "learning_rate": 3.531637904934908e-05, | |
| "loss": 4.4554, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.8900999091734787, | |
| "grad_norm": 2.957598924636841, | |
| "learning_rate": 3.516500151377536e-05, | |
| "loss": 4.5324, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.8991825613079019, | |
| "grad_norm": 2.9992153644561768, | |
| "learning_rate": 3.501362397820164e-05, | |
| "loss": 4.508, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.9082652134423251, | |
| "grad_norm": 3.1509618759155273, | |
| "learning_rate": 3.4862246442627914e-05, | |
| "loss": 4.4265, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.9173478655767484, | |
| "grad_norm": 3.027726888656616, | |
| "learning_rate": 3.4710868907054196e-05, | |
| "loss": 4.4979, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.9264305177111717, | |
| "grad_norm": 3.0711803436279297, | |
| "learning_rate": 3.455949137148047e-05, | |
| "loss": 4.4946, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.935513169845595, | |
| "grad_norm": 2.982269287109375, | |
| "learning_rate": 3.440811383590675e-05, | |
| "loss": 4.3433, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.9445958219800181, | |
| "grad_norm": 2.9734480381011963, | |
| "learning_rate": 3.425673630033303e-05, | |
| "loss": 4.453, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.9536784741144414, | |
| "grad_norm": 2.985030174255371, | |
| "learning_rate": 3.410535876475931e-05, | |
| "loss": 4.3705, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9627611262488647, | |
| "grad_norm": 3.1812829971313477, | |
| "learning_rate": 3.395398122918559e-05, | |
| "loss": 4.3414, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.971843778383288, | |
| "grad_norm": 3.415923595428467, | |
| "learning_rate": 3.380260369361187e-05, | |
| "loss": 4.522, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.9809264305177112, | |
| "grad_norm": 3.176737070083618, | |
| "learning_rate": 3.3651226158038145e-05, | |
| "loss": 4.4112, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.9900090826521344, | |
| "grad_norm": 3.1306254863739014, | |
| "learning_rate": 3.3499848622464433e-05, | |
| "loss": 4.5104, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.9990917347865577, | |
| "grad_norm": 3.216395616531372, | |
| "learning_rate": 3.334847108689071e-05, | |
| "loss": 4.3244, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.008174386920981, | |
| "grad_norm": 3.1889307498931885, | |
| "learning_rate": 3.319709355131699e-05, | |
| "loss": 4.3521, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 1.017257039055404, | |
| "grad_norm": 2.8001787662506104, | |
| "learning_rate": 3.3045716015743264e-05, | |
| "loss": 4.3047, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 1.0263396911898275, | |
| "grad_norm": 3.5796685218811035, | |
| "learning_rate": 3.2894338480169546e-05, | |
| "loss": 4.1921, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 1.0354223433242506, | |
| "grad_norm": 3.725538730621338, | |
| "learning_rate": 3.274296094459583e-05, | |
| "loss": 4.3203, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 1.044504995458674, | |
| "grad_norm": 2.9058167934417725, | |
| "learning_rate": 3.25915834090221e-05, | |
| "loss": 4.385, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.0535876475930972, | |
| "grad_norm": 3.120119333267212, | |
| "learning_rate": 3.244020587344838e-05, | |
| "loss": 4.2883, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 1.0626702997275204, | |
| "grad_norm": 3.230036735534668, | |
| "learning_rate": 3.228882833787466e-05, | |
| "loss": 4.3602, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 1.0717529518619437, | |
| "grad_norm": 3.482921600341797, | |
| "learning_rate": 3.213745080230094e-05, | |
| "loss": 4.3984, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 1.080835603996367, | |
| "grad_norm": 3.0121572017669678, | |
| "learning_rate": 3.198607326672722e-05, | |
| "loss": 4.3864, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 1.0899182561307903, | |
| "grad_norm": 3.277411460876465, | |
| "learning_rate": 3.1834695731153495e-05, | |
| "loss": 4.2294, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.0990009082652135, | |
| "grad_norm": 3.0383167266845703, | |
| "learning_rate": 3.168331819557978e-05, | |
| "loss": 4.2759, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 1.1080835603996366, | |
| "grad_norm": 3.3026745319366455, | |
| "learning_rate": 3.153194066000605e-05, | |
| "loss": 4.3093, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 1.11716621253406, | |
| "grad_norm": 2.954747200012207, | |
| "learning_rate": 3.138056312443234e-05, | |
| "loss": 4.2476, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 1.1262488646684832, | |
| "grad_norm": 3.2137765884399414, | |
| "learning_rate": 3.1229185588858614e-05, | |
| "loss": 4.2858, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 1.1353315168029066, | |
| "grad_norm": 3.4028799533843994, | |
| "learning_rate": 3.1077808053284896e-05, | |
| "loss": 4.3652, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.1444141689373297, | |
| "grad_norm": 3.0039563179016113, | |
| "learning_rate": 3.092643051771118e-05, | |
| "loss": 4.4106, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 1.1534968210717529, | |
| "grad_norm": 2.973820209503174, | |
| "learning_rate": 3.077505298213745e-05, | |
| "loss": 4.1827, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 1.1625794732061763, | |
| "grad_norm": 2.99037766456604, | |
| "learning_rate": 3.062367544656373e-05, | |
| "loss": 4.3092, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 1.1716621253405994, | |
| "grad_norm": 3.181398391723633, | |
| "learning_rate": 3.047229791099001e-05, | |
| "loss": 4.417, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 1.1807447774750228, | |
| "grad_norm": 3.1933484077453613, | |
| "learning_rate": 3.032092037541629e-05, | |
| "loss": 4.2361, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.189827429609446, | |
| "grad_norm": 3.4427855014801025, | |
| "learning_rate": 3.0169542839842567e-05, | |
| "loss": 4.2687, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 1.1989100817438691, | |
| "grad_norm": 3.0683298110961914, | |
| "learning_rate": 3.001816530426885e-05, | |
| "loss": 4.2748, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 1.2079927338782925, | |
| "grad_norm": 3.044698715209961, | |
| "learning_rate": 2.9866787768695127e-05, | |
| "loss": 4.2671, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 1.2170753860127157, | |
| "grad_norm": 3.1354904174804688, | |
| "learning_rate": 2.9715410233121405e-05, | |
| "loss": 4.2635, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 1.226158038147139, | |
| "grad_norm": 3.282745361328125, | |
| "learning_rate": 2.9564032697547683e-05, | |
| "loss": 4.3544, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.2352406902815622, | |
| "grad_norm": 3.369798183441162, | |
| "learning_rate": 2.941265516197396e-05, | |
| "loss": 4.1993, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 1.2443233424159854, | |
| "grad_norm": 3.395785331726074, | |
| "learning_rate": 2.9261277626400242e-05, | |
| "loss": 4.1131, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 1.2534059945504088, | |
| "grad_norm": 3.500697135925293, | |
| "learning_rate": 2.9109900090826524e-05, | |
| "loss": 4.192, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 1.262488646684832, | |
| "grad_norm": 2.94278621673584, | |
| "learning_rate": 2.8958522555252805e-05, | |
| "loss": 4.2863, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 1.2715712988192553, | |
| "grad_norm": 3.3217315673828125, | |
| "learning_rate": 2.8807145019679083e-05, | |
| "loss": 4.1763, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.2806539509536785, | |
| "grad_norm": 3.232830762863159, | |
| "learning_rate": 2.865576748410536e-05, | |
| "loss": 4.2595, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 1.2897366030881017, | |
| "grad_norm": 3.3042378425598145, | |
| "learning_rate": 2.850438994853164e-05, | |
| "loss": 4.2393, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 1.298819255222525, | |
| "grad_norm": 3.83151912689209, | |
| "learning_rate": 2.835301241295792e-05, | |
| "loss": 4.3005, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 1.3079019073569482, | |
| "grad_norm": 3.245086431503296, | |
| "learning_rate": 2.82016348773842e-05, | |
| "loss": 4.205, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 1.3169845594913716, | |
| "grad_norm": 3.4392285346984863, | |
| "learning_rate": 2.8050257341810477e-05, | |
| "loss": 4.1964, | |
| "step": 4350 | |
| } | |
| ], | |
| "logging_steps": 30, | |
| "max_steps": 9909, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 30, | |
| "total_flos": 1136555016192000.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |