| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 2166, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0004616805170821791, | |
| "grad_norm": 15.672144611335952, | |
| "learning_rate": 9.216589861751152e-08, | |
| "loss": 1.3168, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0023084025854108957, | |
| "grad_norm": 14.680930201512727, | |
| "learning_rate": 4.608294930875577e-07, | |
| "loss": 1.2513, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0046168051708217915, | |
| "grad_norm": 8.024031806354767, | |
| "learning_rate": 9.216589861751154e-07, | |
| "loss": 1.1965, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.006925207756232687, | |
| "grad_norm": 5.48716759630272, | |
| "learning_rate": 1.382488479262673e-06, | |
| "loss": 1.0985, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.009233610341643583, | |
| "grad_norm": 5.068857847099135, | |
| "learning_rate": 1.8433179723502307e-06, | |
| "loss": 1.0416, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.011542012927054479, | |
| "grad_norm": 5.784416390233076, | |
| "learning_rate": 2.3041474654377884e-06, | |
| "loss": 1.0414, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.013850415512465374, | |
| "grad_norm": 5.088059220204017, | |
| "learning_rate": 2.764976958525346e-06, | |
| "loss": 1.0977, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.016158818097876268, | |
| "grad_norm": 4.831013532516624, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 1.0864, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.018467220683287166, | |
| "grad_norm": 5.01522341202905, | |
| "learning_rate": 3.6866359447004615e-06, | |
| "loss": 1.0875, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02077562326869806, | |
| "grad_norm": 5.2743430965797495, | |
| "learning_rate": 4.147465437788019e-06, | |
| "loss": 1.0582, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.023084025854108958, | |
| "grad_norm": 4.868797420697276, | |
| "learning_rate": 4.608294930875577e-06, | |
| "loss": 1.0706, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.025392428439519853, | |
| "grad_norm": 4.866881384682284, | |
| "learning_rate": 5.0691244239631346e-06, | |
| "loss": 1.0874, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.027700831024930747, | |
| "grad_norm": 4.942215298185941, | |
| "learning_rate": 5.529953917050692e-06, | |
| "loss": 1.0765, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.030009233610341645, | |
| "grad_norm": 4.510624758005134, | |
| "learning_rate": 5.9907834101382485e-06, | |
| "loss": 1.0651, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.032317636195752536, | |
| "grad_norm": 4.961160516804359, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 1.0826, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03462603878116344, | |
| "grad_norm": 4.999318871040395, | |
| "learning_rate": 6.912442396313365e-06, | |
| "loss": 1.0966, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03693444136657433, | |
| "grad_norm": 4.639315170945839, | |
| "learning_rate": 7.373271889400923e-06, | |
| "loss": 1.0934, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.039242843951985226, | |
| "grad_norm": 4.79620290699333, | |
| "learning_rate": 7.83410138248848e-06, | |
| "loss": 1.0932, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04155124653739612, | |
| "grad_norm": 4.957993386602933, | |
| "learning_rate": 8.294930875576038e-06, | |
| "loss": 1.1032, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.043859649122807015, | |
| "grad_norm": 4.669607524515842, | |
| "learning_rate": 8.755760368663595e-06, | |
| "loss": 1.0875, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.046168051708217916, | |
| "grad_norm": 4.602861109332021, | |
| "learning_rate": 9.216589861751153e-06, | |
| "loss": 1.0809, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04847645429362881, | |
| "grad_norm": 4.548726172146522, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 1.1218, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.050784856879039705, | |
| "grad_norm": 4.724149571099335, | |
| "learning_rate": 1.0138248847926269e-05, | |
| "loss": 1.1007, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0530932594644506, | |
| "grad_norm": 5.309010432204349, | |
| "learning_rate": 1.0599078341013826e-05, | |
| "loss": 1.1368, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.055401662049861494, | |
| "grad_norm": 4.839305137997795, | |
| "learning_rate": 1.1059907834101385e-05, | |
| "loss": 1.1055, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05771006463527239, | |
| "grad_norm": 4.5796294161615565, | |
| "learning_rate": 1.152073732718894e-05, | |
| "loss": 1.1155, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06001846722068329, | |
| "grad_norm": 4.706959812240538, | |
| "learning_rate": 1.1981566820276497e-05, | |
| "loss": 1.1387, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.062326869806094184, | |
| "grad_norm": 4.492165983938348, | |
| "learning_rate": 1.2442396313364056e-05, | |
| "loss": 1.1733, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.06463527239150507, | |
| "grad_norm": 4.746032213098828, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 1.1375, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06694367497691597, | |
| "grad_norm": 4.713817907356248, | |
| "learning_rate": 1.3364055299539171e-05, | |
| "loss": 1.158, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.06925207756232687, | |
| "grad_norm": 4.342905646572964, | |
| "learning_rate": 1.382488479262673e-05, | |
| "loss": 1.1607, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07156048014773776, | |
| "grad_norm": 4.502102336400582, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 1.1382, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.07386888273314866, | |
| "grad_norm": 4.300393542300411, | |
| "learning_rate": 1.4746543778801846e-05, | |
| "loss": 1.1518, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07617728531855955, | |
| "grad_norm": 4.400546990325483, | |
| "learning_rate": 1.5207373271889403e-05, | |
| "loss": 1.1436, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.07848568790397045, | |
| "grad_norm": 4.77590791643038, | |
| "learning_rate": 1.566820276497696e-05, | |
| "loss": 1.2173, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08079409048938135, | |
| "grad_norm": 4.32969974114785, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 1.1654, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.08310249307479224, | |
| "grad_norm": 5.285074448558262, | |
| "learning_rate": 1.6589861751152075e-05, | |
| "loss": 1.2185, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.08541089566020314, | |
| "grad_norm": 6.312179413881035, | |
| "learning_rate": 1.705069124423963e-05, | |
| "loss": 1.2063, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.08771929824561403, | |
| "grad_norm": 4.351482667809684, | |
| "learning_rate": 1.751152073732719e-05, | |
| "loss": 1.1814, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09002770083102493, | |
| "grad_norm": 4.468079454686115, | |
| "learning_rate": 1.7972350230414748e-05, | |
| "loss": 1.2058, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.09233610341643583, | |
| "grad_norm": 5.51425273025908, | |
| "learning_rate": 1.8433179723502307e-05, | |
| "loss": 1.1646, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09464450600184672, | |
| "grad_norm": 4.661323669253999, | |
| "learning_rate": 1.8894009216589862e-05, | |
| "loss": 1.1689, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.09695290858725762, | |
| "grad_norm": 726.888849011745, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 1.8785, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.09926131117266851, | |
| "grad_norm": 5.867844131835615, | |
| "learning_rate": 1.981566820276498e-05, | |
| "loss": 1.2661, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.10156971375807941, | |
| "grad_norm": 6.2684199277472015, | |
| "learning_rate": 1.9999883080288618e-05, | |
| "loss": 1.2545, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.1038781163434903, | |
| "grad_norm": 5.426004523317811, | |
| "learning_rate": 1.999916858084231e-05, | |
| "loss": 1.2259, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1061865189289012, | |
| "grad_norm": 4.617593739028291, | |
| "learning_rate": 1.999780458369908e-05, | |
| "loss": 1.177, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1084949215143121, | |
| "grad_norm": 4.412649452769939, | |
| "learning_rate": 1.9995791177457598e-05, | |
| "loss": 1.2127, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.11080332409972299, | |
| "grad_norm": 4.3422685444059965, | |
| "learning_rate": 1.9993128492899012e-05, | |
| "loss": 1.2398, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.11311172668513389, | |
| "grad_norm": 4.837187367612426, | |
| "learning_rate": 1.9989816702978447e-05, | |
| "loss": 1.2189, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.11542012927054478, | |
| "grad_norm": 4.203624655101154, | |
| "learning_rate": 1.998585602281378e-05, | |
| "loss": 1.1641, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.11772853185595568, | |
| "grad_norm": 4.172078683953242, | |
| "learning_rate": 1.9981246709671668e-05, | |
| "loss": 1.217, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.12003693444136658, | |
| "grad_norm": 4.445815868978626, | |
| "learning_rate": 1.9975989062950828e-05, | |
| "loss": 1.2198, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.12234533702677747, | |
| "grad_norm": 4.5591861583880045, | |
| "learning_rate": 1.9970083424162598e-05, | |
| "loss": 1.2971, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.12465373961218837, | |
| "grad_norm": 8.794456155689286, | |
| "learning_rate": 1.9963530176908752e-05, | |
| "loss": 1.2543, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.12696214219759927, | |
| "grad_norm": 4.296337355363852, | |
| "learning_rate": 1.9956329746856583e-05, | |
| "loss": 1.1902, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.12927054478301014, | |
| "grad_norm": 4.210037276606183, | |
| "learning_rate": 1.9948482601711245e-05, | |
| "loss": 1.2119, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.13157894736842105, | |
| "grad_norm": 4.525819047133829, | |
| "learning_rate": 1.9939989251185386e-05, | |
| "loss": 1.2267, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.13388734995383195, | |
| "grad_norm": 4.617114491344834, | |
| "learning_rate": 1.993085024696604e-05, | |
| "loss": 1.253, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.13619575253924285, | |
| "grad_norm": 4.306884827311129, | |
| "learning_rate": 1.992106618267878e-05, | |
| "loss": 1.2968, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.13850415512465375, | |
| "grad_norm": 4.06156778374873, | |
| "learning_rate": 1.9910637693849166e-05, | |
| "loss": 1.2523, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14081255771006462, | |
| "grad_norm": 4.1725194301873225, | |
| "learning_rate": 1.9899565457861463e-05, | |
| "loss": 1.2465, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.14312096029547552, | |
| "grad_norm": 6.1582094432239005, | |
| "learning_rate": 1.988785019391465e-05, | |
| "loss": 1.2893, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.14542936288088643, | |
| "grad_norm": 4.468419480755536, | |
| "learning_rate": 1.987549266297568e-05, | |
| "loss": 1.2684, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.14773776546629733, | |
| "grad_norm": 4.663314719431853, | |
| "learning_rate": 1.986249366773009e-05, | |
| "loss": 1.2472, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.15004616805170823, | |
| "grad_norm": 4.557295583444763, | |
| "learning_rate": 1.9848854052529822e-05, | |
| "loss": 1.2856, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.1523545706371191, | |
| "grad_norm": 4.128322557091226, | |
| "learning_rate": 1.9834574703338406e-05, | |
| "loss": 1.2717, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.15466297322253, | |
| "grad_norm": 4.265562249971871, | |
| "learning_rate": 1.9819656547673393e-05, | |
| "loss": 1.2614, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.1569713758079409, | |
| "grad_norm": 4.189648461283852, | |
| "learning_rate": 1.9804100554546127e-05, | |
| "loss": 1.2221, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.1592797783933518, | |
| "grad_norm": 4.753781028146877, | |
| "learning_rate": 1.9787907734398785e-05, | |
| "loss": 1.2641, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.1615881809787627, | |
| "grad_norm": 4.624571477954883, | |
| "learning_rate": 1.9771079139038765e-05, | |
| "loss": 1.3082, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.16389658356417358, | |
| "grad_norm": 4.51192227446534, | |
| "learning_rate": 1.9753615861570338e-05, | |
| "loss": 1.3116, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.16620498614958448, | |
| "grad_norm": 4.392313959090253, | |
| "learning_rate": 1.9735519036323656e-05, | |
| "loss": 1.2304, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.16851338873499538, | |
| "grad_norm": 4.979240339208881, | |
| "learning_rate": 1.9716789838781095e-05, | |
| "loss": 1.2682, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.17082179132040629, | |
| "grad_norm": 4.96937836441046, | |
| "learning_rate": 1.9697429485500862e-05, | |
| "loss": 1.3054, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1731301939058172, | |
| "grad_norm": 3.935739346153204, | |
| "learning_rate": 1.9677439234038004e-05, | |
| "loss": 1.2704, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 4.366123456450803, | |
| "learning_rate": 1.96568203828627e-05, | |
| "loss": 1.236, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.17774699907663896, | |
| "grad_norm": 4.003638705307624, | |
| "learning_rate": 1.963557427127594e-05, | |
| "loss": 1.2134, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.18005540166204986, | |
| "grad_norm": 4.711836278485082, | |
| "learning_rate": 1.9613702279322518e-05, | |
| "loss": 1.2424, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.18236380424746076, | |
| "grad_norm": 4.7756346414851345, | |
| "learning_rate": 1.95912058277014e-05, | |
| "loss": 1.2513, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.18467220683287167, | |
| "grad_norm": 4.055556447653374, | |
| "learning_rate": 1.9568086377673422e-05, | |
| "loss": 1.2305, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.18698060941828254, | |
| "grad_norm": 3.9870929086001605, | |
| "learning_rate": 1.9544345430966398e-05, | |
| "loss": 1.2766, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.18928901200369344, | |
| "grad_norm": 4.3683569271591525, | |
| "learning_rate": 1.951998452967756e-05, | |
| "loss": 1.2701, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.19159741458910434, | |
| "grad_norm": 4.282177503327308, | |
| "learning_rate": 1.9495005256173398e-05, | |
| "loss": 1.2173, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.19390581717451524, | |
| "grad_norm": 4.122228465513596, | |
| "learning_rate": 1.9469409232986876e-05, | |
| "loss": 1.293, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.19621421975992612, | |
| "grad_norm": 4.391730062186428, | |
| "learning_rate": 1.9443198122712036e-05, | |
| "loss": 1.3013, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.19852262234533702, | |
| "grad_norm": 4.2533205751093, | |
| "learning_rate": 1.9416373627896002e-05, | |
| "loss": 1.2478, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.20083102493074792, | |
| "grad_norm": 4.982151398275928, | |
| "learning_rate": 1.9388937490928402e-05, | |
| "loss": 1.289, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.20313942751615882, | |
| "grad_norm": 4.254393940238592, | |
| "learning_rate": 1.9360891493928186e-05, | |
| "loss": 1.2773, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.20544783010156972, | |
| "grad_norm": 4.812233488623846, | |
| "learning_rate": 1.933223745862786e-05, | |
| "loss": 1.2571, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.2077562326869806, | |
| "grad_norm": 4.193819364681046, | |
| "learning_rate": 1.930297724625516e-05, | |
| "loss": 1.3167, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2100646352723915, | |
| "grad_norm": 4.318967687699199, | |
| "learning_rate": 1.9273112757412165e-05, | |
| "loss": 1.2578, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.2123730378578024, | |
| "grad_norm": 4.021438837096732, | |
| "learning_rate": 1.9242645931951833e-05, | |
| "loss": 1.2703, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2146814404432133, | |
| "grad_norm": 3.9988355301981344, | |
| "learning_rate": 1.921157874885199e-05, | |
| "loss": 1.2702, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.2169898430286242, | |
| "grad_norm": 3.866018897785007, | |
| "learning_rate": 1.91799132260868e-05, | |
| "loss": 1.2651, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.21929824561403508, | |
| "grad_norm": 4.228145732575894, | |
| "learning_rate": 1.9147651420495696e-05, | |
| "loss": 1.2429, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.22160664819944598, | |
| "grad_norm": 4.16044625111994, | |
| "learning_rate": 1.9114795427649735e-05, | |
| "loss": 1.2263, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.22391505078485688, | |
| "grad_norm": 3.7071606709047678, | |
| "learning_rate": 1.9081347381715535e-05, | |
| "loss": 1.2592, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.22622345337026778, | |
| "grad_norm": 4.093983584879632, | |
| "learning_rate": 1.904730945531661e-05, | |
| "loss": 1.2819, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.22853185595567868, | |
| "grad_norm": 4.247421291613911, | |
| "learning_rate": 1.901268385939226e-05, | |
| "loss": 1.3118, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.23084025854108955, | |
| "grad_norm": 4.088704419142061, | |
| "learning_rate": 1.8977472843053962e-05, | |
| "loss": 1.2529, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23314866112650046, | |
| "grad_norm": 3.9526614218286698, | |
| "learning_rate": 1.8941678693439272e-05, | |
| "loss": 1.2254, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.23545706371191136, | |
| "grad_norm": 3.767319095108075, | |
| "learning_rate": 1.8905303735563274e-05, | |
| "loss": 1.2705, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.23776546629732226, | |
| "grad_norm": 4.1464464034097, | |
| "learning_rate": 1.886835033216755e-05, | |
| "loss": 1.2841, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.24007386888273316, | |
| "grad_norm": 4.154511161776497, | |
| "learning_rate": 1.88308208835667e-05, | |
| "loss": 1.2715, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.24238227146814403, | |
| "grad_norm": 4.815166096996458, | |
| "learning_rate": 1.8792717827492446e-05, | |
| "loss": 1.3034, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.24469067405355494, | |
| "grad_norm": 22.245546847367528, | |
| "learning_rate": 1.8754043638935283e-05, | |
| "loss": 1.2532, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.24699907663896584, | |
| "grad_norm": 4.177323522295811, | |
| "learning_rate": 1.871480082998371e-05, | |
| "loss": 1.2501, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.24930747922437674, | |
| "grad_norm": 3.9426463777773346, | |
| "learning_rate": 1.867499194966106e-05, | |
| "loss": 1.2683, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2516158818097876, | |
| "grad_norm": 3.912690873331932, | |
| "learning_rate": 1.8634619583759933e-05, | |
| "loss": 1.2874, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.25392428439519854, | |
| "grad_norm": 3.972529239438344, | |
| "learning_rate": 1.8593686354674223e-05, | |
| "loss": 1.2698, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2562326869806094, | |
| "grad_norm": 3.958572886167977, | |
| "learning_rate": 1.8552194921228793e-05, | |
| "loss": 1.2293, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.2585410895660203, | |
| "grad_norm": 3.7553829117034767, | |
| "learning_rate": 1.851014797850676e-05, | |
| "loss": 1.2818, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2608494921514312, | |
| "grad_norm": 4.352268879736511, | |
| "learning_rate": 1.8467548257674453e-05, | |
| "loss": 1.2552, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 5.014139215739045, | |
| "learning_rate": 1.8424398525803983e-05, | |
| "loss": 1.2228, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.265466297322253, | |
| "grad_norm": 4.192590762422093, | |
| "learning_rate": 1.8380701585693526e-05, | |
| "loss": 1.2526, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.2677746999076639, | |
| "grad_norm": 4.209340122955672, | |
| "learning_rate": 1.8336460275685267e-05, | |
| "loss": 1.2681, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.27008310249307477, | |
| "grad_norm": 3.801129619164067, | |
| "learning_rate": 1.8291677469481025e-05, | |
| "loss": 1.2623, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.2723915050784857, | |
| "grad_norm": 5.60448449703679, | |
| "learning_rate": 1.8246356075955594e-05, | |
| "loss": 1.2778, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.27469990766389657, | |
| "grad_norm": 3.8415685450636143, | |
| "learning_rate": 1.820049903896782e-05, | |
| "loss": 1.2546, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.2770083102493075, | |
| "grad_norm": 3.766423848242755, | |
| "learning_rate": 1.8154109337169326e-05, | |
| "loss": 1.2994, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2793167128347184, | |
| "grad_norm": 3.8445299977202363, | |
| "learning_rate": 1.8107189983811094e-05, | |
| "loss": 1.2779, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.28162511542012925, | |
| "grad_norm": 4.20182793655244, | |
| "learning_rate": 1.8059744026547713e-05, | |
| "loss": 1.2794, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2839335180055402, | |
| "grad_norm": 3.6927184982852554, | |
| "learning_rate": 1.8011774547239403e-05, | |
| "loss": 1.2217, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.28624192059095105, | |
| "grad_norm": 3.906241578603264, | |
| "learning_rate": 1.796328466175186e-05, | |
| "loss": 1.3162, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.288550323176362, | |
| "grad_norm": 3.7221850266429675, | |
| "learning_rate": 1.791427751975385e-05, | |
| "loss": 1.2591, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.29085872576177285, | |
| "grad_norm": 4.11815775927983, | |
| "learning_rate": 1.786475630451262e-05, | |
| "loss": 1.2572, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2931671283471837, | |
| "grad_norm": 3.8995508626898454, | |
| "learning_rate": 1.781472423268713e-05, | |
| "loss": 1.2604, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.29547553093259465, | |
| "grad_norm": 4.5219499712986035, | |
| "learning_rate": 1.776418455411913e-05, | |
| "loss": 1.298, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.29778393351800553, | |
| "grad_norm": 4.5899598168207785, | |
| "learning_rate": 1.7713140551622032e-05, | |
| "loss": 1.2664, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.30009233610341646, | |
| "grad_norm": 4.641570078800192, | |
| "learning_rate": 1.7661595540767714e-05, | |
| "loss": 1.2689, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.30240073868882733, | |
| "grad_norm": 4.383087991217795, | |
| "learning_rate": 1.7609552869671126e-05, | |
| "loss": 1.2551, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.3047091412742382, | |
| "grad_norm": 3.9687899547292576, | |
| "learning_rate": 1.7557015918772822e-05, | |
| "loss": 1.2379, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.30701754385964913, | |
| "grad_norm": 4.133840300932013, | |
| "learning_rate": 1.750398810061939e-05, | |
| "loss": 1.2779, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.30932594644506, | |
| "grad_norm": 3.84778329275165, | |
| "learning_rate": 1.745047285964179e-05, | |
| "loss": 1.2306, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.31163434903047094, | |
| "grad_norm": 4.054603771464119, | |
| "learning_rate": 1.7396473671931597e-05, | |
| "loss": 1.2089, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.3139427516158818, | |
| "grad_norm": 4.013882196193361, | |
| "learning_rate": 1.7341994045015245e-05, | |
| "loss": 1.2225, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3162511542012927, | |
| "grad_norm": 4.076399340438248, | |
| "learning_rate": 1.7287037517626174e-05, | |
| "loss": 1.3166, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.3185595567867036, | |
| "grad_norm": 3.991144267549364, | |
| "learning_rate": 1.7231607659474972e-05, | |
| "loss": 1.2706, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.3208679593721145, | |
| "grad_norm": 3.592102167186549, | |
| "learning_rate": 1.7175708071017503e-05, | |
| "loss": 1.2066, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3231763619575254, | |
| "grad_norm": 4.2490266329322655, | |
| "learning_rate": 1.7119342383221055e-05, | |
| "loss": 1.3011, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3254847645429363, | |
| "grad_norm": 3.7487591296204266, | |
| "learning_rate": 1.7062514257328474e-05, | |
| "loss": 1.2587, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.32779316712834716, | |
| "grad_norm": 3.6111287365523466, | |
| "learning_rate": 1.7005227384620336e-05, | |
| "loss": 1.2626, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3301015697137581, | |
| "grad_norm": 3.8624035554609892, | |
| "learning_rate": 1.6947485486175223e-05, | |
| "loss": 1.266, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.33240997229916897, | |
| "grad_norm": 4.191574332500623, | |
| "learning_rate": 1.688929231262797e-05, | |
| "loss": 1.2275, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3347183748845799, | |
| "grad_norm": 3.931766819485826, | |
| "learning_rate": 1.683065164392606e-05, | |
| "loss": 1.2525, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.33702677746999077, | |
| "grad_norm": 3.8224846577065685, | |
| "learning_rate": 1.6771567289084122e-05, | |
| "loss": 1.228, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.33933518005540164, | |
| "grad_norm": 3.7975499971303024, | |
| "learning_rate": 1.6712043085936473e-05, | |
| "loss": 1.2121, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.34164358264081257, | |
| "grad_norm": 3.7233983105114326, | |
| "learning_rate": 1.6652082900887858e-05, | |
| "loss": 1.2439, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.34395198522622344, | |
| "grad_norm": 4.0496534376278674, | |
| "learning_rate": 1.6591690628662305e-05, | |
| "loss": 1.3064, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.3462603878116344, | |
| "grad_norm": 4.397682055950332, | |
| "learning_rate": 1.6530870192050134e-05, | |
| "loss": 1.2433, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.34856879039704525, | |
| "grad_norm": 3.999160650641557, | |
| "learning_rate": 1.6469625541653152e-05, | |
| "loss": 1.2117, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 4.475385002364299, | |
| "learning_rate": 1.6407960655628055e-05, | |
| "loss": 1.203, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.35318559556786705, | |
| "grad_norm": 3.5042875341184416, | |
| "learning_rate": 1.6345879539428e-05, | |
| "loss": 1.2567, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.3554939981532779, | |
| "grad_norm": 3.678612416780679, | |
| "learning_rate": 1.6283386225542467e-05, | |
| "loss": 1.2276, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.35780240073868885, | |
| "grad_norm": 5.063348081613382, | |
| "learning_rate": 1.622048477323529e-05, | |
| "loss": 1.2297, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.3601108033240997, | |
| "grad_norm": 4.04397764374825, | |
| "learning_rate": 1.6157179268281007e-05, | |
| "loss": 1.2498, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.3624192059095106, | |
| "grad_norm": 3.7786600086660553, | |
| "learning_rate": 1.6093473822699467e-05, | |
| "loss": 1.2156, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.36472760849492153, | |
| "grad_norm": 3.726670143436363, | |
| "learning_rate": 1.6029372574488732e-05, | |
| "loss": 1.248, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3670360110803324, | |
| "grad_norm": 3.6023664901819115, | |
| "learning_rate": 1.5964879687356286e-05, | |
| "loss": 1.2762, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.36934441366574333, | |
| "grad_norm": 3.684618843127009, | |
| "learning_rate": 1.589999935044859e-05, | |
| "loss": 1.2269, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3716528162511542, | |
| "grad_norm": 3.6119834291134465, | |
| "learning_rate": 1.5834735778078968e-05, | |
| "loss": 1.2078, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.3739612188365651, | |
| "grad_norm": 3.66332363718426, | |
| "learning_rate": 1.5769093209453876e-05, | |
| "loss": 1.2713, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.376269621421976, | |
| "grad_norm": 4.137676249046753, | |
| "learning_rate": 1.5703075908397523e-05, | |
| "loss": 1.2816, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.3785780240073869, | |
| "grad_norm": 3.8481468093108475, | |
| "learning_rate": 1.563668816307494e-05, | |
| "loss": 1.2203, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.3808864265927978, | |
| "grad_norm": 3.7158307301305156, | |
| "learning_rate": 1.556993428571342e-05, | |
| "loss": 1.2163, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.3831948291782087, | |
| "grad_norm": 3.851222452502614, | |
| "learning_rate": 1.550281861232243e-05, | |
| "loss": 1.243, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.38550323176361956, | |
| "grad_norm": 3.6817891692377978, | |
| "learning_rate": 1.5435345502411956e-05, | |
| "loss": 1.2821, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.3878116343490305, | |
| "grad_norm": 3.9683025462284998, | |
| "learning_rate": 1.536751933870934e-05, | |
| "loss": 1.2019, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.39012003693444136, | |
| "grad_norm": 3.94265762295689, | |
| "learning_rate": 1.5299344526874576e-05, | |
| "loss": 1.2774, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.39242843951985223, | |
| "grad_norm": 4.123641725136207, | |
| "learning_rate": 1.5230825495214184e-05, | |
| "loss": 1.2352, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.39473684210526316, | |
| "grad_norm": 3.9570109790957653, | |
| "learning_rate": 1.5161966694393516e-05, | |
| "loss": 1.215, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.39704524469067404, | |
| "grad_norm": 3.6427091867450714, | |
| "learning_rate": 1.5092772597147707e-05, | |
| "loss": 1.2202, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.39935364727608497, | |
| "grad_norm": 3.8425754107191796, | |
| "learning_rate": 1.5023247697991114e-05, | |
| "loss": 1.2432, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.40166204986149584, | |
| "grad_norm": 3.759319372367797, | |
| "learning_rate": 1.4953396512925398e-05, | |
| "loss": 1.1838, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.4039704524469067, | |
| "grad_norm": 3.872324982369786, | |
| "learning_rate": 1.4883223579146167e-05, | |
| "loss": 1.2331, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.40627885503231764, | |
| "grad_norm": 3.8616658245003435, | |
| "learning_rate": 1.4812733454748283e-05, | |
| "loss": 1.2277, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4085872576177285, | |
| "grad_norm": 3.5624714154298163, | |
| "learning_rate": 1.4741930718429772e-05, | |
| "loss": 1.2051, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.41089566020313945, | |
| "grad_norm": 3.6961173549363924, | |
| "learning_rate": 1.4670819969194416e-05, | |
| "loss": 1.2309, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4132040627885503, | |
| "grad_norm": 3.5654510220296847, | |
| "learning_rate": 1.4599405826053039e-05, | |
| "loss": 1.1884, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.4155124653739612, | |
| "grad_norm": 4.205884899208378, | |
| "learning_rate": 1.4527692927723465e-05, | |
| "loss": 1.2223, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4178208679593721, | |
| "grad_norm": 3.9431786244545997, | |
| "learning_rate": 1.4455685932329204e-05, | |
| "loss": 1.2389, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.420129270544783, | |
| "grad_norm": 3.579703652121505, | |
| "learning_rate": 1.4383389517096899e-05, | |
| "loss": 1.2429, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.4224376731301939, | |
| "grad_norm": 3.7807582830713105, | |
| "learning_rate": 1.4310808378052506e-05, | |
| "loss": 1.1874, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.4247460757156048, | |
| "grad_norm": 3.9020463886513914, | |
| "learning_rate": 1.4237947229716262e-05, | |
| "loss": 1.2587, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.42705447830101567, | |
| "grad_norm": 3.7663448915088633, | |
| "learning_rate": 1.4164810804796464e-05, | |
| "loss": 1.184, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.4293628808864266, | |
| "grad_norm": 3.7907471270783937, | |
| "learning_rate": 1.409140385388203e-05, | |
| "loss": 1.2445, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.4316712834718375, | |
| "grad_norm": 3.791543245723202, | |
| "learning_rate": 1.4017731145133955e-05, | |
| "loss": 1.2527, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.4339796860572484, | |
| "grad_norm": 3.8566751713668666, | |
| "learning_rate": 1.3943797463975575e-05, | |
| "loss": 1.2048, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.4362880886426593, | |
| "grad_norm": 3.943257567360323, | |
| "learning_rate": 1.3869607612781733e-05, | |
| "loss": 1.2773, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 3.53206021655625, | |
| "learning_rate": 1.3795166410566834e-05, | |
| "loss": 1.2066, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.4409048938134811, | |
| "grad_norm": 3.8322607840339504, | |
| "learning_rate": 1.372047869267184e-05, | |
| "loss": 1.2104, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.44321329639889195, | |
| "grad_norm": 4.982802180271467, | |
| "learning_rate": 1.364554931045018e-05, | |
| "loss": 1.2782, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.4455216989843029, | |
| "grad_norm": 4.121927772157904, | |
| "learning_rate": 1.3570383130952627e-05, | |
| "loss": 1.2221, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.44783010156971376, | |
| "grad_norm": 3.5401426054616674, | |
| "learning_rate": 1.349498503661116e-05, | |
| "loss": 1.249, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.45013850415512463, | |
| "grad_norm": 3.8347876039826647, | |
| "learning_rate": 1.3419359924921833e-05, | |
| "loss": 1.2736, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.45244690674053556, | |
| "grad_norm": 4.86416192250325, | |
| "learning_rate": 1.3343512708126642e-05, | |
| "loss": 1.2032, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.45475530932594643, | |
| "grad_norm": 3.8508803970513004, | |
| "learning_rate": 1.326744831289447e-05, | |
| "loss": 1.2465, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.45706371191135736, | |
| "grad_norm": 3.276661833625774, | |
| "learning_rate": 1.3191171680001048e-05, | |
| "loss": 1.1905, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.45937211449676824, | |
| "grad_norm": 3.6488550777243933, | |
| "learning_rate": 1.3114687764008048e-05, | |
| "loss": 1.1991, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.4616805170821791, | |
| "grad_norm": 3.9637997706000223, | |
| "learning_rate": 1.3038001532941249e-05, | |
| "loss": 1.1994, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46398891966759004, | |
| "grad_norm": 3.7798295608326447, | |
| "learning_rate": 1.2961117967967844e-05, | |
| "loss": 1.2327, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.4662973222530009, | |
| "grad_norm": 3.742363753899004, | |
| "learning_rate": 1.2884042063072881e-05, | |
| "loss": 1.2415, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.46860572483841184, | |
| "grad_norm": 4.00995610689072, | |
| "learning_rate": 1.280677882473488e-05, | |
| "loss": 1.2449, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.4709141274238227, | |
| "grad_norm": 3.7802768150285284, | |
| "learning_rate": 1.272933327160063e-05, | |
| "loss": 1.2055, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.4732225300092336, | |
| "grad_norm": 3.979719082398227, | |
| "learning_rate": 1.2651710434159223e-05, | |
| "loss": 1.1452, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.4755309325946445, | |
| "grad_norm": 3.7987734509998012, | |
| "learning_rate": 1.2573915354415274e-05, | |
| "loss": 1.2266, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.4778393351800554, | |
| "grad_norm": 3.4449265105850344, | |
| "learning_rate": 1.2495953085561426e-05, | |
| "loss": 1.1678, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.4801477377654663, | |
| "grad_norm": 4.703831538180476, | |
| "learning_rate": 1.241782869165012e-05, | |
| "loss": 1.1893, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.4824561403508772, | |
| "grad_norm": 3.56138065098868, | |
| "learning_rate": 1.2339547247264658e-05, | |
| "loss": 1.2285, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.48476454293628807, | |
| "grad_norm": 3.8664090630676147, | |
| "learning_rate": 1.2261113837189587e-05, | |
| "loss": 1.1995, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.487072945521699, | |
| "grad_norm": 3.6587622685467553, | |
| "learning_rate": 1.2182533556080402e-05, | |
| "loss": 1.2456, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.48938134810710987, | |
| "grad_norm": 3.4219623018934615, | |
| "learning_rate": 1.2103811508132642e-05, | |
| "loss": 1.1904, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.4916897506925208, | |
| "grad_norm": 3.91141223990254, | |
| "learning_rate": 1.2024952806750321e-05, | |
| "loss": 1.1811, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.4939981532779317, | |
| "grad_norm": 3.707066130468398, | |
| "learning_rate": 1.1945962574213814e-05, | |
| "loss": 1.212, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.49630655586334255, | |
| "grad_norm": 3.5782501836947653, | |
| "learning_rate": 1.1866845941347118e-05, | |
| "loss": 1.2255, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.4986149584487535, | |
| "grad_norm": 4.303350644777213, | |
| "learning_rate": 1.1787608047184583e-05, | |
| "loss": 1.1376, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5009233610341643, | |
| "grad_norm": 3.419543860379626, | |
| "learning_rate": 1.1708254038637115e-05, | |
| "loss": 1.1872, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.5032317636195752, | |
| "grad_norm": 3.586294780528409, | |
| "learning_rate": 1.1628789070157836e-05, | |
| "loss": 1.2114, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5055401662049861, | |
| "grad_norm": 3.6647616517214496, | |
| "learning_rate": 1.1549218303407305e-05, | |
| "loss": 1.2088, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.5078485687903971, | |
| "grad_norm": 3.6209405687157794, | |
| "learning_rate": 1.1469546906918219e-05, | |
| "loss": 1.1535, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.510156971375808, | |
| "grad_norm": 3.4760951984933777, | |
| "learning_rate": 1.1389780055759689e-05, | |
| "loss": 1.1692, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.5124653739612188, | |
| "grad_norm": 3.523587148397925, | |
| "learning_rate": 1.1309922931201114e-05, | |
| "loss": 1.1795, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5147737765466297, | |
| "grad_norm": 3.399747435026194, | |
| "learning_rate": 1.1229980720375609e-05, | |
| "loss": 1.1913, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.5170821791320406, | |
| "grad_norm": 3.802970464768176, | |
| "learning_rate": 1.114995861594308e-05, | |
| "loss": 1.1692, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5193905817174516, | |
| "grad_norm": 3.571347595436078, | |
| "learning_rate": 1.1069861815752944e-05, | |
| "loss": 1.1575, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.5216989843028624, | |
| "grad_norm": 3.702241350827994, | |
| "learning_rate": 1.0989695522506486e-05, | |
| "loss": 1.1776, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.5240073868882733, | |
| "grad_norm": 4.396145181294285, | |
| "learning_rate": 1.0909464943418926e-05, | |
| "loss": 1.2055, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 3.402649511273165, | |
| "learning_rate": 1.0829175289881188e-05, | |
| "loss": 1.2024, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.528624192059095, | |
| "grad_norm": 3.321901777095843, | |
| "learning_rate": 1.074883177712138e-05, | |
| "loss": 1.1317, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.530932594644506, | |
| "grad_norm": 4.575011114858196, | |
| "learning_rate": 1.0668439623866043e-05, | |
| "loss": 1.1516, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.5332409972299169, | |
| "grad_norm": 3.428811319179132, | |
| "learning_rate": 1.0588004052001177e-05, | |
| "loss": 1.1326, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.5355493998153278, | |
| "grad_norm": 3.758823500740248, | |
| "learning_rate": 1.0507530286233042e-05, | |
| "loss": 1.1523, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.5378578024007387, | |
| "grad_norm": 3.828420445656179, | |
| "learning_rate": 1.0427023553748792e-05, | |
| "loss": 1.215, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.5401662049861495, | |
| "grad_norm": 3.872474623427253, | |
| "learning_rate": 1.0346489083876928e-05, | |
| "loss": 1.1798, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.5424746075715605, | |
| "grad_norm": 4.343223419966708, | |
| "learning_rate": 1.0265932107747656e-05, | |
| "loss": 1.1964, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.5447830101569714, | |
| "grad_norm": 3.4458152638291533, | |
| "learning_rate": 1.0185357857953064e-05, | |
| "loss": 1.188, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.5470914127423823, | |
| "grad_norm": 3.3343026801443765, | |
| "learning_rate": 1.0104771568207266e-05, | |
| "loss": 1.1524, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.5493998153277931, | |
| "grad_norm": 3.8325280372919774, | |
| "learning_rate": 1.0024178473006418e-05, | |
| "loss": 1.1445, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.551708217913204, | |
| "grad_norm": 3.913934401934443, | |
| "learning_rate": 9.943583807288746e-06, | |
| "loss": 1.1497, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.554016620498615, | |
| "grad_norm": 3.8771337742661585, | |
| "learning_rate": 9.862992806094473e-06, | |
| "loss": 1.1584, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5563250230840259, | |
| "grad_norm": 3.385706053842486, | |
| "learning_rate": 9.782410704225793e-06, | |
| "loss": 1.133, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.5586334256694367, | |
| "grad_norm": 3.228558572718497, | |
| "learning_rate": 9.701842735906855e-06, | |
| "loss": 1.1714, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.5609418282548476, | |
| "grad_norm": 3.376489834368575, | |
| "learning_rate": 9.621294134443747e-06, | |
| "loss": 1.1782, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.5632502308402585, | |
| "grad_norm": 4.101023970778267, | |
| "learning_rate": 9.54077013188459e-06, | |
| "loss": 1.1679, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.5655586334256695, | |
| "grad_norm": 3.459693677788322, | |
| "learning_rate": 9.460275958679674e-06, | |
| "loss": 1.2272, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.5678670360110804, | |
| "grad_norm": 3.5741244509053556, | |
| "learning_rate": 9.379816843341715e-06, | |
| "loss": 1.1679, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.5701754385964912, | |
| "grad_norm": 14.959841662019736, | |
| "learning_rate": 9.299398012106246e-06, | |
| "loss": 1.1557, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.5724838411819021, | |
| "grad_norm": 3.479142794568544, | |
| "learning_rate": 9.219024688592136e-06, | |
| "loss": 1.191, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.574792243767313, | |
| "grad_norm": 3.4791994405128195, | |
| "learning_rate": 9.138702093462286e-06, | |
| "loss": 1.1632, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.577100646352724, | |
| "grad_norm": 3.378297795269278, | |
| "learning_rate": 9.058435444084543e-06, | |
| "loss": 1.2058, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5794090489381348, | |
| "grad_norm": 3.3312286796444948, | |
| "learning_rate": 8.978229954192775e-06, | |
| "loss": 1.2072, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.5817174515235457, | |
| "grad_norm": 3.2936946867277497, | |
| "learning_rate": 8.898090833548226e-06, | |
| "loss": 1.1479, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.5840258541089566, | |
| "grad_norm": 3.5657195698986306, | |
| "learning_rate": 8.818023287601117e-06, | |
| "loss": 1.1579, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.5863342566943675, | |
| "grad_norm": 3.85534125869907, | |
| "learning_rate": 8.738032517152523e-06, | |
| "loss": 1.1748, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5886426592797784, | |
| "grad_norm": 3.3807308381583585, | |
| "learning_rate": 8.658123718016548e-06, | |
| "loss": 1.1365, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.5909510618651893, | |
| "grad_norm": 3.75547737356039, | |
| "learning_rate": 8.578302080682844e-06, | |
| "loss": 1.1657, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5932594644506002, | |
| "grad_norm": 3.334058557259955, | |
| "learning_rate": 8.498572789979446e-06, | |
| "loss": 1.1653, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.5955678670360111, | |
| "grad_norm": 3.596795067568704, | |
| "learning_rate": 8.418941024735997e-06, | |
| "loss": 1.1909, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.5978762696214219, | |
| "grad_norm": 3.754106205642103, | |
| "learning_rate": 8.33941195744737e-06, | |
| "loss": 1.1595, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.6001846722068329, | |
| "grad_norm": 3.3575559431036988, | |
| "learning_rate": 8.259990753937662e-06, | |
| "loss": 1.1378, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6024930747922438, | |
| "grad_norm": 4.011372021010383, | |
| "learning_rate": 8.18068257302466e-06, | |
| "loss": 1.1832, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.6048014773776547, | |
| "grad_norm": 3.404379906541828, | |
| "learning_rate": 8.101492566184757e-06, | |
| "loss": 1.1592, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6071098799630655, | |
| "grad_norm": 3.498132121516362, | |
| "learning_rate": 8.022425877218321e-06, | |
| "loss": 1.1591, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.6094182825484764, | |
| "grad_norm": 3.523586580045349, | |
| "learning_rate": 7.943487641915595e-06, | |
| "loss": 1.1525, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.6117266851338874, | |
| "grad_norm": 3.6229726894839858, | |
| "learning_rate": 7.864682987723082e-06, | |
| "loss": 1.1618, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 3.696989469787097, | |
| "learning_rate": 7.78601703341051e-06, | |
| "loss": 1.1824, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.6163434903047091, | |
| "grad_norm": 3.567967173775001, | |
| "learning_rate": 7.70749488873833e-06, | |
| "loss": 1.1792, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.61865189289012, | |
| "grad_norm": 3.399928397766497, | |
| "learning_rate": 7.629121654125808e-06, | |
| "loss": 1.1438, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.6209602954755309, | |
| "grad_norm": 3.6344006441397414, | |
| "learning_rate": 7.550902420319742e-06, | |
| "loss": 1.1591, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.6232686980609419, | |
| "grad_norm": 3.538106316840523, | |
| "learning_rate": 7.472842268063776e-06, | |
| "loss": 1.1311, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.6255771006463527, | |
| "grad_norm": 3.661558906665894, | |
| "learning_rate": 7.394946267768381e-06, | |
| "loss": 1.1621, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.6278855032317636, | |
| "grad_norm": 3.6197107279149954, | |
| "learning_rate": 7.317219479181517e-06, | |
| "loss": 1.1028, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.6301939058171745, | |
| "grad_norm": 3.4094252840241355, | |
| "learning_rate": 7.23966695105996e-06, | |
| "loss": 1.119, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.6325023084025854, | |
| "grad_norm": 3.4085855538144467, | |
| "learning_rate": 7.162293720841378e-06, | |
| "loss": 1.1438, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.6348107109879964, | |
| "grad_norm": 4.073406312500022, | |
| "learning_rate": 7.085104814317101e-06, | |
| "loss": 1.1729, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.6371191135734072, | |
| "grad_norm": 3.572178264074241, | |
| "learning_rate": 7.008105245305699e-06, | |
| "loss": 1.1661, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.6394275161588181, | |
| "grad_norm": 3.81528951625221, | |
| "learning_rate": 6.931300015327274e-06, | |
| "loss": 1.1571, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.641735918744229, | |
| "grad_norm": 3.2846636335941763, | |
| "learning_rate": 6.854694113278614e-06, | |
| "loss": 1.154, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.6440443213296398, | |
| "grad_norm": 3.2544013227776007, | |
| "learning_rate": 6.7782925151091224e-06, | |
| "loss": 1.0823, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.6463527239150508, | |
| "grad_norm": 3.482450898904014, | |
| "learning_rate": 6.702100183497613e-06, | |
| "loss": 1.1803, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.6486611265004617, | |
| "grad_norm": 3.412256349030684, | |
| "learning_rate": 6.62612206752995e-06, | |
| "loss": 1.1643, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.6509695290858726, | |
| "grad_norm": 3.75196899322532, | |
| "learning_rate": 6.550363102377588e-06, | |
| "loss": 1.1117, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.6532779316712835, | |
| "grad_norm": 3.3485189294016258, | |
| "learning_rate": 6.474828208976998e-06, | |
| "loss": 1.1466, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.6555863342566943, | |
| "grad_norm": 3.4421443761863104, | |
| "learning_rate": 6.3995222937100455e-06, | |
| "loss": 1.1468, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.6578947368421053, | |
| "grad_norm": 3.4653107797221683, | |
| "learning_rate": 6.324450248085265e-06, | |
| "loss": 1.1418, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.6602031394275162, | |
| "grad_norm": 3.450235228111911, | |
| "learning_rate": 6.249616948420161e-06, | |
| "loss": 1.1393, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.6625115420129271, | |
| "grad_norm": 3.648594332616919, | |
| "learning_rate": 6.175027255524446e-06, | |
| "loss": 1.1263, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.6648199445983379, | |
| "grad_norm": 3.50804118935427, | |
| "learning_rate": 6.100686014384315e-06, | |
| "loss": 1.1497, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.6671283471837488, | |
| "grad_norm": 3.407303145023877, | |
| "learning_rate": 6.026598053847743e-06, | |
| "loss": 1.1217, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.6694367497691598, | |
| "grad_norm": 3.6049741156075426, | |
| "learning_rate": 5.952768186310813e-06, | |
| "loss": 1.2134, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.6717451523545707, | |
| "grad_norm": 3.347553717603198, | |
| "learning_rate": 5.879201207405136e-06, | |
| "loss": 1.1189, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.6740535549399815, | |
| "grad_norm": 3.7624263901785087, | |
| "learning_rate": 5.805901895686344e-06, | |
| "loss": 1.1217, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.6763619575253924, | |
| "grad_norm": 3.6359056480115193, | |
| "learning_rate": 5.732875012323712e-06, | |
| "loss": 1.1275, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.6786703601108033, | |
| "grad_norm": 3.5660085050284946, | |
| "learning_rate": 5.660125300790873e-06, | |
| "loss": 1.153, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.6809787626962143, | |
| "grad_norm": 3.4188915438262946, | |
| "learning_rate": 5.58765748655772e-06, | |
| "loss": 1.126, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.6832871652816251, | |
| "grad_norm": 3.7409360766713995, | |
| "learning_rate": 5.5154762767834605e-06, | |
| "loss": 1.1312, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.685595567867036, | |
| "grad_norm": 3.5176838276710787, | |
| "learning_rate": 5.443586360010859e-06, | |
| "loss": 1.118, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.6879039704524469, | |
| "grad_norm": 3.940112737940071, | |
| "learning_rate": 5.3719924058616975e-06, | |
| "loss": 1.1084, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.6902123730378578, | |
| "grad_norm": 3.5725656073039516, | |
| "learning_rate": 5.30069906473345e-06, | |
| "loss": 1.1462, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.6925207756232687, | |
| "grad_norm": 3.585328985764251, | |
| "learning_rate": 5.2297109674972166e-06, | |
| "loss": 1.1275, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6948291782086796, | |
| "grad_norm": 3.7150630899084276, | |
| "learning_rate": 5.159032725196946e-06, | |
| "loss": 1.1573, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.6971375807940905, | |
| "grad_norm": 3.4991531847637893, | |
| "learning_rate": 5.088668928749891e-06, | |
| "loss": 1.1339, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.6994459833795014, | |
| "grad_norm": 3.337315702796277, | |
| "learning_rate": 5.0186241486484245e-06, | |
| "loss": 1.1121, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 3.166495977462906, | |
| "learning_rate": 4.948902934663158e-06, | |
| "loss": 1.1207, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.7040627885503232, | |
| "grad_norm": 3.269883473204096, | |
| "learning_rate": 4.879509815547413e-06, | |
| "loss": 1.1067, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.7063711911357341, | |
| "grad_norm": 3.2549683491943138, | |
| "learning_rate": 4.810449298743051e-06, | |
| "loss": 1.0858, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.708679593721145, | |
| "grad_norm": 3.673192940396545, | |
| "learning_rate": 4.741725870087693e-06, | |
| "loss": 1.1674, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.7109879963065558, | |
| "grad_norm": 3.295243146355197, | |
| "learning_rate": 4.673343993523347e-06, | |
| "loss": 1.1087, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.7132963988919667, | |
| "grad_norm": 3.4162872942710867, | |
| "learning_rate": 4.605308110806436e-06, | |
| "loss": 1.1224, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.7156048014773777, | |
| "grad_norm": 3.3989883160652865, | |
| "learning_rate": 4.537622641219309e-06, | |
| "loss": 1.1307, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.7179132040627886, | |
| "grad_norm": 3.2956559559454663, | |
| "learning_rate": 4.47029198128316e-06, | |
| "loss": 1.0944, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.7202216066481995, | |
| "grad_norm": 3.3797718456778765, | |
| "learning_rate": 4.403320504472463e-06, | |
| "loss": 1.1426, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.7225300092336103, | |
| "grad_norm": 3.1832339015639826, | |
| "learning_rate": 4.336712560930891e-06, | |
| "loss": 1.1223, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.7248384118190212, | |
| "grad_norm": 3.40273969921815, | |
| "learning_rate": 4.270472477188755e-06, | |
| "loss": 1.1151, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.7271468144044322, | |
| "grad_norm": 3.32953363908172, | |
| "learning_rate": 4.204604555881967e-06, | |
| "loss": 1.1055, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.7294552169898431, | |
| "grad_norm": 3.363759228103856, | |
| "learning_rate": 4.139113075472565e-06, | |
| "loss": 1.15, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.7317636195752539, | |
| "grad_norm": 3.5692625205390214, | |
| "learning_rate": 4.074002289970801e-06, | |
| "loss": 1.1249, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.7340720221606648, | |
| "grad_norm": 3.6298117912857895, | |
| "learning_rate": 4.009276428658836e-06, | |
| "loss": 1.0911, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.7363804247460757, | |
| "grad_norm": 3.501911680130801, | |
| "learning_rate": 3.944939695816005e-06, | |
| "loss": 1.0591, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.7386888273314867, | |
| "grad_norm": 3.314254645913856, | |
| "learning_rate": 3.8809962704457375e-06, | |
| "loss": 1.122, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.7409972299168975, | |
| "grad_norm": 3.56145944415269, | |
| "learning_rate": 3.81745030600411e-06, | |
| "loss": 1.1036, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.7433056325023084, | |
| "grad_norm": 3.4910849192084235, | |
| "learning_rate": 3.75430593013006e-06, | |
| "loss": 1.1353, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.7456140350877193, | |
| "grad_norm": 3.325715619787326, | |
| "learning_rate": 3.6915672443772644e-06, | |
| "loss": 1.1538, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.7479224376731302, | |
| "grad_norm": 3.5950013679874724, | |
| "learning_rate": 3.62923832394774e-06, | |
| "loss": 1.0909, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.7502308402585411, | |
| "grad_norm": 3.1524005532212334, | |
| "learning_rate": 3.56732321742712e-06, | |
| "loss": 1.1125, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.752539242843952, | |
| "grad_norm": 3.6760451234626124, | |
| "learning_rate": 3.5058259465216828e-06, | |
| "loss": 1.1039, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.7548476454293629, | |
| "grad_norm": 3.341546948891595, | |
| "learning_rate": 3.444750505797123e-06, | |
| "loss": 1.0531, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.7571560480147738, | |
| "grad_norm": 3.35627649123262, | |
| "learning_rate": 3.384100862419096e-06, | |
| "loss": 1.0931, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.7594644506001846, | |
| "grad_norm": 3.6221419833131527, | |
| "learning_rate": 3.3238809558955054e-06, | |
| "loss": 1.0797, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.7617728531855956, | |
| "grad_norm": 3.3487671296828267, | |
| "learning_rate": 3.2640946978206266e-06, | |
| "loss": 1.0812, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.7640812557710065, | |
| "grad_norm": 3.441031645390376, | |
| "learning_rate": 3.2047459716210306e-06, | |
| "loss": 1.1155, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.7663896583564174, | |
| "grad_norm": 3.4825057106301096, | |
| "learning_rate": 3.145838632303325e-06, | |
| "loss": 1.096, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.7686980609418282, | |
| "grad_norm": 3.4525699686491875, | |
| "learning_rate": 3.087376506203763e-06, | |
| "loss": 1.145, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.7710064635272391, | |
| "grad_norm": 3.2639030957505715, | |
| "learning_rate": 3.0293633907396903e-06, | |
| "loss": 1.0711, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.7733148661126501, | |
| "grad_norm": 3.247147491878351, | |
| "learning_rate": 2.971803054162903e-06, | |
| "loss": 1.0367, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.775623268698061, | |
| "grad_norm": 3.3628039668359824, | |
| "learning_rate": 2.914699235314855e-06, | |
| "loss": 1.1311, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.7779316712834718, | |
| "grad_norm": 3.294560766749018, | |
| "learning_rate": 2.858055643383818e-06, | |
| "loss": 1.1303, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.7802400738688827, | |
| "grad_norm": 3.252460881051861, | |
| "learning_rate": 2.8018759576639478e-06, | |
| "loss": 1.0894, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.7825484764542936, | |
| "grad_norm": 3.6541818791755083, | |
| "learning_rate": 2.7461638273162895e-06, | |
| "loss": 1.1416, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.7848568790397045, | |
| "grad_norm": 3.3018114290440286, | |
| "learning_rate": 2.6909228711317526e-06, | |
| "loss": 1.0898, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.7871652816251155, | |
| "grad_norm": 3.5110479717681704, | |
| "learning_rate": 2.6361566772960466e-06, | |
| "loss": 1.0887, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 3.469571849173682, | |
| "learning_rate": 2.5818688031566132e-06, | |
| "loss": 1.0182, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.7917820867959372, | |
| "grad_norm": 3.761287355693432, | |
| "learning_rate": 2.5280627749915544e-06, | |
| "loss": 1.1246, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.7940904893813481, | |
| "grad_norm": 3.7171990367681866, | |
| "learning_rate": 2.4747420877805905e-06, | |
| "loss": 1.1008, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.796398891966759, | |
| "grad_norm": 3.583342537171837, | |
| "learning_rate": 2.421910204978033e-06, | |
| "loss": 1.092, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.7987072945521699, | |
| "grad_norm": 3.3105866570237343, | |
| "learning_rate": 2.369570558287819e-06, | |
| "loss": 1.0495, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.8010156971375808, | |
| "grad_norm": 3.453250654565143, | |
| "learning_rate": 2.3177265474406084e-06, | |
| "loss": 1.0952, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.8033240997229917, | |
| "grad_norm": 3.2111312681793294, | |
| "learning_rate": 2.2663815399729495e-06, | |
| "loss": 1.0756, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.8056325023084026, | |
| "grad_norm": 3.398739502823191, | |
| "learning_rate": 2.215538871008538e-06, | |
| "loss": 1.0855, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.8079409048938134, | |
| "grad_norm": 3.4089573083048883, | |
| "learning_rate": 2.1652018430415923e-06, | |
| "loss": 1.0707, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.8102493074792244, | |
| "grad_norm": 3.7996382043873744, | |
| "learning_rate": 2.115373725722326e-06, | |
| "loss": 1.1419, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.8125577100646353, | |
| "grad_norm": 3.4303103622199203, | |
| "learning_rate": 2.066057755644587e-06, | |
| "loss": 1.1101, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.8148661126500462, | |
| "grad_norm": 3.3758394994097363, | |
| "learning_rate": 2.0172571361356007e-06, | |
| "loss": 1.0975, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.817174515235457, | |
| "grad_norm": 3.2901551940425673, | |
| "learning_rate": 1.9689750370479134e-06, | |
| "loss": 1.0797, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.8194829178208679, | |
| "grad_norm": 3.661068632899665, | |
| "learning_rate": 1.921214594553488e-06, | |
| "loss": 1.1287, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.8217913204062789, | |
| "grad_norm": 3.5442080312978415, | |
| "learning_rate": 1.8739789109399954e-06, | |
| "loss": 1.1514, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.8240997229916898, | |
| "grad_norm": 3.3534741257777325, | |
| "learning_rate": 1.8272710544093019e-06, | |
| "loss": 1.0824, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.8264081255771006, | |
| "grad_norm": 3.570055818522298, | |
| "learning_rate": 1.7810940588781811e-06, | |
| "loss": 1.1313, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.8287165281625115, | |
| "grad_norm": 3.3907592881825352, | |
| "learning_rate": 1.7354509237812334e-06, | |
| "loss": 1.0458, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.8310249307479224, | |
| "grad_norm": 3.7660635086416794, | |
| "learning_rate": 1.690344613876066e-06, | |
| "loss": 1.109, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 20.624336407323348, | |
| "learning_rate": 1.64577805905072e-06, | |
| "loss": 1.0872, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.8356417359187442, | |
| "grad_norm": 3.38434013035599, | |
| "learning_rate": 1.601754154133347e-06, | |
| "loss": 1.0943, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.8379501385041551, | |
| "grad_norm": 3.3282071197431318, | |
| "learning_rate": 1.558275758704183e-06, | |
| "loss": 1.0983, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.840258541089566, | |
| "grad_norm": 3.4156960745203286, | |
| "learning_rate": 1.5153456969098013e-06, | |
| "loss": 1.0381, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.8425669436749769, | |
| "grad_norm": 3.3418973703656274, | |
| "learning_rate": 1.4729667572796735e-06, | |
| "loss": 1.1452, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.8448753462603878, | |
| "grad_norm": 3.333897377962453, | |
| "learning_rate": 1.431141692545036e-06, | |
| "loss": 1.1076, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.8471837488457987, | |
| "grad_norm": 3.402941306050666, | |
| "learning_rate": 1.389873219460085e-06, | |
| "loss": 1.0869, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.8494921514312096, | |
| "grad_norm": 3.3313186519496423, | |
| "learning_rate": 1.349164018625513e-06, | |
| "loss": 1.0765, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.8518005540166205, | |
| "grad_norm": 3.6011720414080566, | |
| "learning_rate": 1.3090167343143911e-06, | |
| "loss": 1.0846, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.8541089566020313, | |
| "grad_norm": 3.629326020817196, | |
| "learning_rate": 1.2694339743004037e-06, | |
| "loss": 1.1088, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.8564173591874423, | |
| "grad_norm": 3.6305906598709767, | |
| "learning_rate": 1.2304183096884626e-06, | |
| "loss": 1.0875, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.8587257617728532, | |
| "grad_norm": 3.35865168543221, | |
| "learning_rate": 1.1919722747477024e-06, | |
| "loss": 1.1143, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.8610341643582641, | |
| "grad_norm": 3.3889339992199177, | |
| "learning_rate": 1.1540983667468686e-06, | |
| "loss": 1.0916, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 0.863342566943675, | |
| "grad_norm": 3.3133014347890324, | |
| "learning_rate": 1.1167990457920985e-06, | |
| "loss": 1.0877, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.8656509695290858, | |
| "grad_norm": 3.415023896862017, | |
| "learning_rate": 1.0800767346671347e-06, | |
| "loss": 1.0284, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.8679593721144968, | |
| "grad_norm": 3.322962975732958, | |
| "learning_rate": 1.043933818675944e-06, | |
| "loss": 1.0782, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.8702677746999077, | |
| "grad_norm": 3.583896655771928, | |
| "learning_rate": 1.008372645487785e-06, | |
| "loss": 1.08, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 0.8725761772853186, | |
| "grad_norm": 3.3057678718948726, | |
| "learning_rate": 9.733955249847183e-07, | |
| "loss": 1.1034, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.8748845798707294, | |
| "grad_norm": 3.4387092657320997, | |
| "learning_rate": 9.390047291115567e-07, | |
| "loss": 1.0915, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 3.8029482282950324, | |
| "learning_rate": 9.052024917282987e-07, | |
| "loss": 1.057, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.8795013850415513, | |
| "grad_norm": 3.3990790831971465, | |
| "learning_rate": 8.719910084650262e-07, | |
| "loss": 1.0725, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 0.8818097876269622, | |
| "grad_norm": 3.262416726762208, | |
| "learning_rate": 8.393724365792866e-07, | |
| "loss": 1.1028, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.884118190212373, | |
| "grad_norm": 3.551691283783414, | |
| "learning_rate": 8.073488948159691e-07, | |
| "loss": 1.0546, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 0.8864265927977839, | |
| "grad_norm": 3.5211563130144197, | |
| "learning_rate": 7.759224632696793e-07, | |
| "loss": 1.1024, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.8887349953831948, | |
| "grad_norm": 3.5958803804208976, | |
| "learning_rate": 7.450951832496233e-07, | |
| "loss": 1.0698, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.8910433979686058, | |
| "grad_norm": 4.107811963680795, | |
| "learning_rate": 7.148690571470251e-07, | |
| "loss": 1.0613, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.8933518005540166, | |
| "grad_norm": 3.6280688174940416, | |
| "learning_rate": 6.852460483050494e-07, | |
| "loss": 1.0987, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 0.8956602031394275, | |
| "grad_norm": 3.4197153407779055, | |
| "learning_rate": 6.562280808912768e-07, | |
| "loss": 1.081, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.8979686057248384, | |
| "grad_norm": 3.3975321682078494, | |
| "learning_rate": 6.278170397727179e-07, | |
| "loss": 1.0881, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 0.9002770083102493, | |
| "grad_norm": 3.385824657440924, | |
| "learning_rate": 6.000147703933845e-07, | |
| "loss": 1.0725, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.9025854108956602, | |
| "grad_norm": 3.733106691108023, | |
| "learning_rate": 5.728230786544153e-07, | |
| "loss": 1.0886, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 0.9048938134810711, | |
| "grad_norm": 3.3831515124529288, | |
| "learning_rate": 5.46243730796776e-07, | |
| "loss": 1.0854, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.907202216066482, | |
| "grad_norm": 3.4106013139907065, | |
| "learning_rate": 5.202784532865302e-07, | |
| "loss": 1.114, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 0.9095106186518929, | |
| "grad_norm": 3.130011381325973, | |
| "learning_rate": 4.949289327026952e-07, | |
| "loss": 1.0873, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.9118190212373037, | |
| "grad_norm": 3.3600219468750394, | |
| "learning_rate": 4.7019681562769816e-07, | |
| "loss": 1.0689, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.9141274238227147, | |
| "grad_norm": 3.379655670825615, | |
| "learning_rate": 4.460837085404113e-07, | |
| "loss": 1.0874, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.9164358264081256, | |
| "grad_norm": 3.324809563310868, | |
| "learning_rate": 4.225911777118097e-07, | |
| "loss": 1.0894, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 0.9187442289935365, | |
| "grad_norm": 3.4668181744618196, | |
| "learning_rate": 3.9972074910323066e-07, | |
| "loss": 1.0896, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.9210526315789473, | |
| "grad_norm": 3.4175120046363276, | |
| "learning_rate": 3.7747390826725736e-07, | |
| "loss": 1.0608, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 0.9233610341643582, | |
| "grad_norm": 3.365932789028912, | |
| "learning_rate": 3.5585210025122166e-07, | |
| "loss": 1.0465, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9256694367497692, | |
| "grad_norm": 3.3721429412301442, | |
| "learning_rate": 3.3485672950334447e-07, | |
| "loss": 1.0782, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 0.9279778393351801, | |
| "grad_norm": 3.402893452692765, | |
| "learning_rate": 3.1448915978150365e-07, | |
| "loss": 1.0575, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.930286241920591, | |
| "grad_norm": 3.3246351042614606, | |
| "learning_rate": 2.947507140646588e-07, | |
| "loss": 1.093, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 0.9325946445060018, | |
| "grad_norm": 3.42392243323848, | |
| "learning_rate": 2.756426744669105e-07, | |
| "loss": 1.0709, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.9349030470914127, | |
| "grad_norm": 3.3870385964627565, | |
| "learning_rate": 2.57166282154222e-07, | |
| "loss": 1.0944, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.9372114496768237, | |
| "grad_norm": 3.4345800654530128, | |
| "learning_rate": 2.393227372638018e-07, | |
| "loss": 1.0829, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.9395198522622346, | |
| "grad_norm": 3.2304527099741094, | |
| "learning_rate": 2.221131988261438e-07, | |
| "loss": 1.0663, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 0.9418282548476454, | |
| "grad_norm": 3.4212248154000324, | |
| "learning_rate": 2.055387846897472e-07, | |
| "loss": 1.0608, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.9441366574330563, | |
| "grad_norm": 3.3424495231710356, | |
| "learning_rate": 1.8960057144850163e-07, | |
| "loss": 1.0513, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 0.9464450600184672, | |
| "grad_norm": 8.42913586604929, | |
| "learning_rate": 1.742995943717607e-07, | |
| "loss": 1.0698, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.9487534626038782, | |
| "grad_norm": 4.03605470816158, | |
| "learning_rate": 1.5963684733709462e-07, | |
| "loss": 1.0787, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 0.951061865189289, | |
| "grad_norm": 3.572766321551919, | |
| "learning_rate": 1.4561328276573415e-07, | |
| "loss": 1.0625, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.9533702677746999, | |
| "grad_norm": 3.213406555168112, | |
| "learning_rate": 1.3222981156070126e-07, | |
| "loss": 1.0861, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 0.9556786703601108, | |
| "grad_norm": 3.216022210724082, | |
| "learning_rate": 1.1948730304764622e-07, | |
| "loss": 1.0572, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.9579870729455217, | |
| "grad_norm": 3.8142801195990237, | |
| "learning_rate": 1.073865849183786e-07, | |
| "loss": 1.1151, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.9602954755309326, | |
| "grad_norm": 3.2011503381896413, | |
| "learning_rate": 9.592844317710238e-08, | |
| "loss": 1.0585, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.9626038781163435, | |
| "grad_norm": 3.3780038857652226, | |
| "learning_rate": 8.511362208936447e-08, | |
| "loss": 1.0591, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 3.3212612452494295, | |
| "learning_rate": 7.494282413371135e-08, | |
| "loss": 1.0787, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.9672206832871653, | |
| "grad_norm": 3.797857330316498, | |
| "learning_rate": 6.541670995605321e-08, | |
| "loss": 1.0859, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 0.9695290858725761, | |
| "grad_norm": 3.153773745189338, | |
| "learning_rate": 5.653589832675943e-08, | |
| "loss": 1.0983, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.9718374884579871, | |
| "grad_norm": 3.4652822167549906, | |
| "learning_rate": 4.830096610045854e-08, | |
| "loss": 1.0713, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 0.974145891043398, | |
| "grad_norm": 3.6601967632905796, | |
| "learning_rate": 4.071244817857589e-08, | |
| "loss": 1.1118, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.9764542936288089, | |
| "grad_norm": 3.135385406063897, | |
| "learning_rate": 3.3770837474584874e-08, | |
| "loss": 1.072, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 0.9787626962142197, | |
| "grad_norm": 3.4884714571677784, | |
| "learning_rate": 2.747658488199023e-08, | |
| "loss": 1.0738, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.9810710987996306, | |
| "grad_norm": 3.7803263448925706, | |
| "learning_rate": 2.1830099245040427e-08, | |
| "loss": 1.0549, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.9833795013850416, | |
| "grad_norm": 3.3045019552603585, | |
| "learning_rate": 1.683174733216997e-08, | |
| "loss": 1.1129, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.9856879039704525, | |
| "grad_norm": 3.2212668180182784, | |
| "learning_rate": 1.248185381217848e-08, | |
| "loss": 1.0777, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 0.9879963065558633, | |
| "grad_norm": 3.324768260260177, | |
| "learning_rate": 8.780701233139789e-09, | |
| "loss": 1.0503, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.9903047091412742, | |
| "grad_norm": 3.214869100486745, | |
| "learning_rate": 5.728530004051047e-09, | |
| "loss": 1.0367, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 0.9926131117266851, | |
| "grad_norm": 3.3583215428853666, | |
| "learning_rate": 3.325538379211901e-09, | |
| "loss": 1.0554, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.9949215143120961, | |
| "grad_norm": 4.075445923312751, | |
| "learning_rate": 1.5718824453525572e-09, | |
| "loss": 1.1222, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 0.997229916897507, | |
| "grad_norm": 3.3599147688364903, | |
| "learning_rate": 4.676761114941197e-10, | |
| "loss": 1.0646, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.9995383194829178, | |
| "grad_norm": 3.4496692841727543, | |
| "learning_rate": 1.2991101545622998e-11, | |
| "loss": 1.1038, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.1177629232406616, | |
| "eval_runtime": 1154.8442, | |
| "eval_samples_per_second": 26.579, | |
| "eval_steps_per_second": 0.831, | |
| "step": 2166 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 2166, | |
| "total_flos": 113379083550720.0, | |
| "train_loss": 1.171416565762112, | |
| "train_runtime": 11018.7418, | |
| "train_samples_per_second": 6.29, | |
| "train_steps_per_second": 0.197 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2166, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 113379083550720.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |