| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 4180, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005980861244019139, | |
| "grad_norm": 3.412325382232666, | |
| "learning_rate": 0.0002, | |
| "loss": 3.3895, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.011961722488038277, | |
| "grad_norm": 9.85770034790039, | |
| "learning_rate": 0.0002, | |
| "loss": 2.6708, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.017942583732057416, | |
| "grad_norm": 1.8022898435592651, | |
| "learning_rate": 0.0002, | |
| "loss": 2.4984, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.023923444976076555, | |
| "grad_norm": 1.3909275531768799, | |
| "learning_rate": 0.0002, | |
| "loss": 2.4433, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.029904306220095694, | |
| "grad_norm": 1.3175278902053833, | |
| "learning_rate": 0.0002, | |
| "loss": 2.4068, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03588516746411483, | |
| "grad_norm": 2.1698503494262695, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3572, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.041866028708133975, | |
| "grad_norm": 3.156744956970215, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3665, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04784688995215311, | |
| "grad_norm": 1.0966124534606934, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3631, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05382775119617225, | |
| "grad_norm": 1.5008922815322876, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3375, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.05980861244019139, | |
| "grad_norm": 1.0263694524765015, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3668, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06578947368421052, | |
| "grad_norm": 0.9050750136375427, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2953, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.07177033492822966, | |
| "grad_norm": 1.1184417009353638, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3101, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07775119617224881, | |
| "grad_norm": 1.2090150117874146, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3292, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.08373205741626795, | |
| "grad_norm": 2.347069263458252, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3108, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08971291866028708, | |
| "grad_norm": 1.3362812995910645, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2901, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.09569377990430622, | |
| "grad_norm": 0.906521201133728, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3137, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.10167464114832536, | |
| "grad_norm": 0.8491584658622742, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2915, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.1076555023923445, | |
| "grad_norm": 0.9403386116027832, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2874, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "grad_norm": 0.7675734758377075, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2536, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.11961722488038277, | |
| "grad_norm": 0.8333762288093567, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3332, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1255980861244019, | |
| "grad_norm": 0.8489273190498352, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3112, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.13157894736842105, | |
| "grad_norm": 1.2032957077026367, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2614, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1375598086124402, | |
| "grad_norm": 0.8014360070228577, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3014, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.14354066985645933, | |
| "grad_norm": 0.8756849765777588, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2626, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.14952153110047847, | |
| "grad_norm": 1.0479413270950317, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2487, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.15550239234449761, | |
| "grad_norm": 1.6525335311889648, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2643, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.16148325358851676, | |
| "grad_norm": 0.7974942922592163, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2983, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.1674641148325359, | |
| "grad_norm": 0.8229785561561584, | |
| "learning_rate": 0.0002, | |
| "loss": 2.27, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.173444976076555, | |
| "grad_norm": 0.9374330639839172, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2322, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.17942583732057416, | |
| "grad_norm": 0.8249229788780212, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3029, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.1854066985645933, | |
| "grad_norm": 0.8934934735298157, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2204, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.19138755980861244, | |
| "grad_norm": 0.8451672196388245, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2576, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.19736842105263158, | |
| "grad_norm": 0.8721255660057068, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2357, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.20334928229665072, | |
| "grad_norm": 0.9069824814796448, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2624, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.20933014354066987, | |
| "grad_norm": 0.8029842972755432, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2697, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.215311004784689, | |
| "grad_norm": 3.0026650428771973, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2058, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.22129186602870812, | |
| "grad_norm": 0.8026193380355835, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2537, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 0.770354688167572, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3118, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.2332535885167464, | |
| "grad_norm": 0.822100043296814, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2443, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.23923444976076555, | |
| "grad_norm": 0.8492611050605774, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2295, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2452153110047847, | |
| "grad_norm": 0.7530927658081055, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2321, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.2511961722488038, | |
| "grad_norm": 0.7999204993247986, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2462, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.25717703349282295, | |
| "grad_norm": 0.765783965587616, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2362, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.9569802284240723, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2309, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.26913875598086123, | |
| "grad_norm": 0.8249408602714539, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2433, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.2751196172248804, | |
| "grad_norm": 0.848108172416687, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2013, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.2811004784688995, | |
| "grad_norm": 0.8488432765007019, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2393, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.28708133971291866, | |
| "grad_norm": 0.7786160111427307, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2247, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.2930622009569378, | |
| "grad_norm": 2.2969539165496826, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2523, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.29904306220095694, | |
| "grad_norm": 0.8230640292167664, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2327, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.3050239234449761, | |
| "grad_norm": 0.8330740928649902, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2306, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.31100478468899523, | |
| "grad_norm": 0.8412021994590759, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2299, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.31698564593301437, | |
| "grad_norm": 0.8107555508613586, | |
| "learning_rate": 0.0002, | |
| "loss": 2.264, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.3229665071770335, | |
| "grad_norm": 2.283083200454712, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2085, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.32894736842105265, | |
| "grad_norm": 0.7781470417976379, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2123, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.3349282296650718, | |
| "grad_norm": 0.7660220265388489, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2525, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 0.8373708724975586, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2142, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.34688995215311, | |
| "grad_norm": 0.7727882862091064, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1824, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.35287081339712917, | |
| "grad_norm": 0.9092174768447876, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1939, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.3588516746411483, | |
| "grad_norm": 0.8021971583366394, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2175, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.36483253588516745, | |
| "grad_norm": 0.7922872304916382, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2079, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.3708133971291866, | |
| "grad_norm": 1.0842111110687256, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1879, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.37679425837320574, | |
| "grad_norm": 0.9562531113624573, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1896, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.3827751196172249, | |
| "grad_norm": 0.8320727348327637, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2254, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.388755980861244, | |
| "grad_norm": 0.80451899766922, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2144, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.39473684210526316, | |
| "grad_norm": 0.7288826704025269, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2349, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.4007177033492823, | |
| "grad_norm": 0.8023431897163391, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2037, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.40669856459330145, | |
| "grad_norm": 0.8532123565673828, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2314, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.4126794258373206, | |
| "grad_norm": 0.7851171493530273, | |
| "learning_rate": 0.0002, | |
| "loss": 2.185, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.41866028708133973, | |
| "grad_norm": 0.8515769243240356, | |
| "learning_rate": 0.0002, | |
| "loss": 2.232, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.4246411483253589, | |
| "grad_norm": 0.782311201095581, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1859, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.430622009569378, | |
| "grad_norm": 0.7590478658676147, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1837, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.4366028708133971, | |
| "grad_norm": 0.7843049168586731, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2069, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.44258373205741625, | |
| "grad_norm": 0.7173344492912292, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1989, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4485645933014354, | |
| "grad_norm": 0.8669169545173645, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1941, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.45564603805542, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2045, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4605263157894737, | |
| "grad_norm": 0.8243363499641418, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1819, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.4665071770334928, | |
| "grad_norm": 0.8817090392112732, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1533, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.47248803827751196, | |
| "grad_norm": 0.82022625207901, | |
| "learning_rate": 0.0002, | |
| "loss": 2.243, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.4784688995215311, | |
| "grad_norm": 0.8863716721534729, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1638, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.48444976076555024, | |
| "grad_norm": 0.7413605451583862, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2107, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.4904306220095694, | |
| "grad_norm": 0.8566731810569763, | |
| "learning_rate": 0.0002, | |
| "loss": 2.148, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.4964114832535885, | |
| "grad_norm": 0.9010487794876099, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1828, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.5023923444976076, | |
| "grad_norm": 0.8197215795516968, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1683, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.5083732057416268, | |
| "grad_norm": 0.9159034490585327, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2269, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.5143540669856459, | |
| "grad_norm": 0.8410281538963318, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2111, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.5203349282296651, | |
| "grad_norm": 1.8926668167114258, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1884, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.0006904602050781, | |
| "learning_rate": 0.0002, | |
| "loss": 2.192, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.5322966507177034, | |
| "grad_norm": 0.7826078534126282, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1894, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.5382775119617225, | |
| "grad_norm": 0.8574744462966919, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1854, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.5442583732057417, | |
| "grad_norm": 0.9018279314041138, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1725, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.5502392344497608, | |
| "grad_norm": 1.3028662204742432, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1944, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.55622009569378, | |
| "grad_norm": 0.8321689367294312, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1652, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.562200956937799, | |
| "grad_norm": 0.7999281287193298, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2041, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "grad_norm": 4.9575629234313965, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2154, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.5741626794258373, | |
| "grad_norm": 0.7689957022666931, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1695, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.5801435406698564, | |
| "grad_norm": 0.9012035131454468, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1557, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.5861244019138756, | |
| "grad_norm": 0.8276737928390503, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1906, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.5921052631578947, | |
| "grad_norm": 0.9128056168556213, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1445, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.5980861244019139, | |
| "grad_norm": 0.8623008131980896, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1852, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.604066985645933, | |
| "grad_norm": 0.7866010665893555, | |
| "learning_rate": 0.0002, | |
| "loss": 2.173, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.6100478468899522, | |
| "grad_norm": 0.8097877502441406, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1324, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.6160287081339713, | |
| "grad_norm": 0.8153032660484314, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1271, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.6220095693779905, | |
| "grad_norm": 0.7427578568458557, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1617, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.6279904306220095, | |
| "grad_norm": 0.8749725222587585, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1961, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.6339712918660287, | |
| "grad_norm": 0.7804417610168457, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1782, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.6399521531100478, | |
| "grad_norm": 0.8692734837532043, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1852, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.645933014354067, | |
| "grad_norm": 0.8000411987304688, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1628, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.6519138755980861, | |
| "grad_norm": 0.9027504324913025, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1666, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.6578947368421053, | |
| "grad_norm": 0.8674067854881287, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1394, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.6638755980861244, | |
| "grad_norm": 1.3791645765304565, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1626, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.6698564593301436, | |
| "grad_norm": 0.8177993297576904, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1664, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.6758373205741627, | |
| "grad_norm": 0.8040952682495117, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1603, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.8698276281356812, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2068, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.687799043062201, | |
| "grad_norm": 0.8038722276687622, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1528, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.69377990430622, | |
| "grad_norm": 0.8705615401268005, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1538, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.6997607655502392, | |
| "grad_norm": 0.9985973834991455, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1686, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.7057416267942583, | |
| "grad_norm": 0.7473865747451782, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2257, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.7117224880382775, | |
| "grad_norm": 0.8028366565704346, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1712, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.7177033492822966, | |
| "grad_norm": 0.767857551574707, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1782, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7236842105263158, | |
| "grad_norm": 0.7830066680908203, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1488, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.7296650717703349, | |
| "grad_norm": 0.8238586783409119, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1588, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.7356459330143541, | |
| "grad_norm": 0.7727087140083313, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1398, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.7416267942583732, | |
| "grad_norm": 0.8918077945709229, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1829, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.7476076555023924, | |
| "grad_norm": 1.202504277229309, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1784, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.7535885167464115, | |
| "grad_norm": 0.8316906094551086, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1806, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.7595693779904307, | |
| "grad_norm": 0.7766339182853699, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1577, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.7655502392344498, | |
| "grad_norm": 0.9902828931808472, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1519, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.7715311004784688, | |
| "grad_norm": 0.895126461982727, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1712, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.777511961722488, | |
| "grad_norm": 0.8055546879768372, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1814, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.7834928229665071, | |
| "grad_norm": 0.7867780327796936, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1507, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.8065791726112366, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2211, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.7954545454545454, | |
| "grad_norm": 1.5913640260696411, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2257, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.8014354066985646, | |
| "grad_norm": 0.7849767208099365, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1893, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.8074162679425837, | |
| "grad_norm": 0.7633355855941772, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1956, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.8133971291866029, | |
| "grad_norm": 0.8164528608322144, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2224, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.819377990430622, | |
| "grad_norm": 0.7906235456466675, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1833, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.8253588516746412, | |
| "grad_norm": 0.8774910569190979, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1471, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.8313397129186603, | |
| "grad_norm": 0.8200404644012451, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1554, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 0.8373205741626795, | |
| "grad_norm": 0.7728098630905151, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2009, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8433014354066986, | |
| "grad_norm": 0.7523846626281738, | |
| "learning_rate": 0.0002, | |
| "loss": 2.2052, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 0.8492822966507177, | |
| "grad_norm": 0.8525931239128113, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1811, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.8552631578947368, | |
| "grad_norm": 0.7875164747238159, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1308, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 0.861244019138756, | |
| "grad_norm": 0.7879646420478821, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1665, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.8672248803827751, | |
| "grad_norm": 0.7715153694152832, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1593, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 0.8732057416267942, | |
| "grad_norm": 0.8685998320579529, | |
| "learning_rate": 0.0002, | |
| "loss": 2.18, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.8791866028708134, | |
| "grad_norm": 0.8396874666213989, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1639, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 0.8851674641148325, | |
| "grad_norm": 0.8163192272186279, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1742, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.8911483253588517, | |
| "grad_norm": 0.8341553211212158, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1569, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 0.8971291866028708, | |
| "grad_norm": 0.7632786631584167, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1596, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.90311004784689, | |
| "grad_norm": 0.7861719131469727, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1853, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.8243244886398315, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1695, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.9150717703349283, | |
| "grad_norm": 0.760749876499176, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1501, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 0.9210526315789473, | |
| "grad_norm": 0.9622604250907898, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1256, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.9270334928229665, | |
| "grad_norm": 0.7732083797454834, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1433, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 0.9330143540669856, | |
| "grad_norm": 0.7828539609909058, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1356, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.9389952153110048, | |
| "grad_norm": 0.8860824704170227, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1525, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 0.9449760765550239, | |
| "grad_norm": 0.8569679260253906, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1501, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.9509569377990431, | |
| "grad_norm": 0.7966086864471436, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1484, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 0.9569377990430622, | |
| "grad_norm": 0.7861948609352112, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1461, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.9629186602870813, | |
| "grad_norm": 0.8073152303695679, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1681, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 0.9688995215311005, | |
| "grad_norm": 0.8233998417854309, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1513, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.9748803827751196, | |
| "grad_norm": 0.836236834526062, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1665, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 0.9808612440191388, | |
| "grad_norm": 0.7221957445144653, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1079, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.9868421052631579, | |
| "grad_norm": 0.7149819731712341, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1858, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 0.992822966507177, | |
| "grad_norm": 0.7578993439674377, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1467, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.9988038277511961, | |
| "grad_norm": 1.0370241403579712, | |
| "learning_rate": 0.0002, | |
| "loss": 2.1626, | |
| "step": 4175 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 4180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.194945264893952e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |