| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.96969696969697, | |
| "eval_steps": 500, | |
| "global_step": 615, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 80.8431859261654, | |
| "learning_rate": 4.0650406504065046e-07, | |
| "loss": 2.64, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 58.95328571668908, | |
| "learning_rate": 8.130081300813009e-07, | |
| "loss": 2.5295, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 30.259656389849923, | |
| "learning_rate": 1.2195121951219514e-06, | |
| "loss": 2.3595, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 19.031626660884115, | |
| "learning_rate": 1.6260162601626018e-06, | |
| "loss": 2.2046, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 14.285496207217582, | |
| "learning_rate": 2.0325203252032523e-06, | |
| "loss": 2.0114, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 11.705575031693748, | |
| "learning_rate": 2.4390243902439027e-06, | |
| "loss": 1.9542, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 10.031255451818282, | |
| "learning_rate": 2.845528455284553e-06, | |
| "loss": 1.9027, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 9.87064503889936, | |
| "learning_rate": 3.2520325203252037e-06, | |
| "loss": 1.8376, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 9.463342602070677, | |
| "learning_rate": 3.6585365853658537e-06, | |
| "loss": 1.8335, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 8.14740265971335, | |
| "learning_rate": 4.0650406504065046e-06, | |
| "loss": 1.7531, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 8.579660709279775, | |
| "learning_rate": 4.471544715447155e-06, | |
| "loss": 1.746, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 7.764594425498635, | |
| "learning_rate": 4.8780487804878055e-06, | |
| "loss": 1.7795, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 7.818967776047545, | |
| "learning_rate": 5.2845528455284555e-06, | |
| "loss": 1.7845, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 8.571992272728625, | |
| "learning_rate": 5.691056910569106e-06, | |
| "loss": 1.7176, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 7.157751637340014, | |
| "learning_rate": 6.0975609756097564e-06, | |
| "loss": 1.6894, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 7.776730945646382, | |
| "learning_rate": 6.504065040650407e-06, | |
| "loss": 1.6733, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 7.538790813241334, | |
| "learning_rate": 6.910569105691057e-06, | |
| "loss": 1.6661, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 7.237611731257021, | |
| "learning_rate": 7.317073170731707e-06, | |
| "loss": 1.6413, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 8.713375291094371, | |
| "learning_rate": 7.723577235772358e-06, | |
| "loss": 1.6712, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 8.148761177726017, | |
| "learning_rate": 8.130081300813009e-06, | |
| "loss": 1.6491, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 7.670084788858041, | |
| "learning_rate": 8.536585365853658e-06, | |
| "loss": 1.6661, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 8.193767782896925, | |
| "learning_rate": 8.94308943089431e-06, | |
| "loss": 1.6061, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 7.327023272280465, | |
| "learning_rate": 9.34959349593496e-06, | |
| "loss": 1.6177, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 7.6240834480577995, | |
| "learning_rate": 9.756097560975611e-06, | |
| "loss": 1.7007, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "grad_norm": 7.022808705113093, | |
| "learning_rate": 9.959349593495936e-06, | |
| "loss": 1.5861, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0505050505050506, | |
| "grad_norm": 7.831382750074179, | |
| "learning_rate": 9.857723577235772e-06, | |
| "loss": 1.397, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 7.482634887493619, | |
| "learning_rate": 9.756097560975611e-06, | |
| "loss": 1.3207, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1313131313131313, | |
| "grad_norm": 7.77405709671628, | |
| "learning_rate": 9.654471544715448e-06, | |
| "loss": 1.4016, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1717171717171717, | |
| "grad_norm": 8.316039462095251, | |
| "learning_rate": 9.552845528455286e-06, | |
| "loss": 1.4044, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 8.220962058193507, | |
| "learning_rate": 9.451219512195122e-06, | |
| "loss": 1.4525, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2525252525252526, | |
| "grad_norm": 8.628523413754008, | |
| "learning_rate": 9.34959349593496e-06, | |
| "loss": 1.3725, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.2929292929292928, | |
| "grad_norm": 9.049067713299571, | |
| "learning_rate": 9.247967479674797e-06, | |
| "loss": 1.3988, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 8.79216448432911, | |
| "learning_rate": 9.146341463414635e-06, | |
| "loss": 1.3923, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.3737373737373737, | |
| "grad_norm": 7.749843189602475, | |
| "learning_rate": 9.044715447154472e-06, | |
| "loss": 1.3965, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4141414141414141, | |
| "grad_norm": 8.10742211898071, | |
| "learning_rate": 8.94308943089431e-06, | |
| "loss": 1.3586, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 8.61374938206157, | |
| "learning_rate": 8.841463414634148e-06, | |
| "loss": 1.3455, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.494949494949495, | |
| "grad_norm": 7.806546434692009, | |
| "learning_rate": 8.739837398373985e-06, | |
| "loss": 1.3164, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5353535353535355, | |
| "grad_norm": 8.055069468513308, | |
| "learning_rate": 8.638211382113821e-06, | |
| "loss": 1.3719, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 7.286077733666312, | |
| "learning_rate": 8.536585365853658e-06, | |
| "loss": 1.331, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6161616161616161, | |
| "grad_norm": 7.722391890399128, | |
| "learning_rate": 8.434959349593497e-06, | |
| "loss": 1.347, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6565656565656566, | |
| "grad_norm": 7.963303799143797, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.2988, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 8.363381324799755, | |
| "learning_rate": 8.23170731707317e-06, | |
| "loss": 1.3731, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.7373737373737375, | |
| "grad_norm": 8.650050591837509, | |
| "learning_rate": 8.130081300813009e-06, | |
| "loss": 1.3554, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 7.821382086934233, | |
| "learning_rate": 8.028455284552846e-06, | |
| "loss": 1.3257, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 7.593249783363984, | |
| "learning_rate": 7.926829268292685e-06, | |
| "loss": 1.2994, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.8585858585858586, | |
| "grad_norm": 8.265959609765153, | |
| "learning_rate": 7.82520325203252e-06, | |
| "loss": 1.258, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.898989898989899, | |
| "grad_norm": 7.47340446063849, | |
| "learning_rate": 7.723577235772358e-06, | |
| "loss": 1.3744, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 7.8636893423505505, | |
| "learning_rate": 7.621951219512196e-06, | |
| "loss": 1.2867, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.9797979797979797, | |
| "grad_norm": 8.66108025838036, | |
| "learning_rate": 7.520325203252034e-06, | |
| "loss": 1.3423, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.0202020202020203, | |
| "grad_norm": 7.253266730457967, | |
| "learning_rate": 7.41869918699187e-06, | |
| "loss": 1.1248, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 7.304801005516647, | |
| "learning_rate": 7.317073170731707e-06, | |
| "loss": 0.8695, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.101010101010101, | |
| "grad_norm": 8.148533408280995, | |
| "learning_rate": 7.215447154471545e-06, | |
| "loss": 0.8396, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.1414141414141414, | |
| "grad_norm": 9.202743834871297, | |
| "learning_rate": 7.113821138211383e-06, | |
| "loss": 0.8835, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 9.248143695051853, | |
| "learning_rate": 7.01219512195122e-06, | |
| "loss": 0.769, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 8.877400578549704, | |
| "learning_rate": 6.910569105691057e-06, | |
| "loss": 0.8883, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.2626262626262625, | |
| "grad_norm": 9.858300545714043, | |
| "learning_rate": 6.808943089430895e-06, | |
| "loss": 0.8892, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 8.940708831871842, | |
| "learning_rate": 6.707317073170733e-06, | |
| "loss": 0.8526, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.3434343434343434, | |
| "grad_norm": 9.167915788662723, | |
| "learning_rate": 6.60569105691057e-06, | |
| "loss": 0.9309, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.3838383838383836, | |
| "grad_norm": 8.435170262522817, | |
| "learning_rate": 6.504065040650407e-06, | |
| "loss": 0.8693, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 9.022959586969035, | |
| "learning_rate": 6.402439024390244e-06, | |
| "loss": 0.8659, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.4646464646464645, | |
| "grad_norm": 8.705530302904208, | |
| "learning_rate": 6.300813008130082e-06, | |
| "loss": 0.9076, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.505050505050505, | |
| "grad_norm": 8.449067406312437, | |
| "learning_rate": 6.199186991869919e-06, | |
| "loss": 0.8896, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 8.962552587001122, | |
| "learning_rate": 6.0975609756097564e-06, | |
| "loss": 0.8568, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.5858585858585856, | |
| "grad_norm": 9.26680724967832, | |
| "learning_rate": 5.995934959349594e-06, | |
| "loss": 0.8707, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.6262626262626263, | |
| "grad_norm": 9.852323988179384, | |
| "learning_rate": 5.894308943089432e-06, | |
| "loss": 0.9007, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 9.473031970955077, | |
| "learning_rate": 5.792682926829269e-06, | |
| "loss": 0.8907, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.707070707070707, | |
| "grad_norm": 8.423216825316242, | |
| "learning_rate": 5.691056910569106e-06, | |
| "loss": 0.8408, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.7474747474747474, | |
| "grad_norm": 8.772022355651819, | |
| "learning_rate": 5.589430894308944e-06, | |
| "loss": 0.8791, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 9.456726517429484, | |
| "learning_rate": 5.487804878048781e-06, | |
| "loss": 0.9048, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.8282828282828283, | |
| "grad_norm": 8.932119436113132, | |
| "learning_rate": 5.386178861788618e-06, | |
| "loss": 0.9458, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.8686868686868685, | |
| "grad_norm": 9.076984191036512, | |
| "learning_rate": 5.2845528455284555e-06, | |
| "loss": 0.887, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 9.228482912276068, | |
| "learning_rate": 5.182926829268293e-06, | |
| "loss": 0.9164, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.9494949494949494, | |
| "grad_norm": 9.39115941622314, | |
| "learning_rate": 5.081300813008131e-06, | |
| "loss": 0.9146, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.98989898989899, | |
| "grad_norm": 9.590905474617363, | |
| "learning_rate": 4.979674796747968e-06, | |
| "loss": 0.8975, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.0303030303030303, | |
| "grad_norm": 8.147437990412763, | |
| "learning_rate": 4.8780487804878055e-06, | |
| "loss": 0.5845, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.0707070707070705, | |
| "grad_norm": 8.966178510928351, | |
| "learning_rate": 4.776422764227643e-06, | |
| "loss": 0.4348, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 9.411638909598917, | |
| "learning_rate": 4.67479674796748e-06, | |
| "loss": 0.4666, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.1515151515151514, | |
| "grad_norm": 8.787258663179166, | |
| "learning_rate": 4.573170731707318e-06, | |
| "loss": 0.4412, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.191919191919192, | |
| "grad_norm": 8.448686917354301, | |
| "learning_rate": 4.471544715447155e-06, | |
| "loss": 0.4644, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.2323232323232323, | |
| "grad_norm": 10.794449599661947, | |
| "learning_rate": 4.369918699186992e-06, | |
| "loss": 0.4697, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 9.903329877706229, | |
| "learning_rate": 4.268292682926829e-06, | |
| "loss": 0.4451, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.313131313131313, | |
| "grad_norm": 10.77521515420911, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.4332, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.3535353535353534, | |
| "grad_norm": 9.850095186977901, | |
| "learning_rate": 4.0650406504065046e-06, | |
| "loss": 0.4571, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.393939393939394, | |
| "grad_norm": 9.719974950746256, | |
| "learning_rate": 3.963414634146342e-06, | |
| "loss": 0.4651, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.4343434343434343, | |
| "grad_norm": 8.909903135988007, | |
| "learning_rate": 3.861788617886179e-06, | |
| "loss": 0.4412, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.474747474747475, | |
| "grad_norm": 9.513088350689786, | |
| "learning_rate": 3.760162601626017e-06, | |
| "loss": 0.4713, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.515151515151515, | |
| "grad_norm": 8.770653506187902, | |
| "learning_rate": 3.6585365853658537e-06, | |
| "loss": 0.4325, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 10.698789680572704, | |
| "learning_rate": 3.5569105691056914e-06, | |
| "loss": 0.456, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.595959595959596, | |
| "grad_norm": 9.719819699090596, | |
| "learning_rate": 3.4552845528455287e-06, | |
| "loss": 0.4673, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 10.503333896340955, | |
| "learning_rate": 3.3536585365853664e-06, | |
| "loss": 0.4571, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.676767676767677, | |
| "grad_norm": 9.503329386063092, | |
| "learning_rate": 3.2520325203252037e-06, | |
| "loss": 0.4389, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 3.717171717171717, | |
| "grad_norm": 10.009998037929371, | |
| "learning_rate": 3.150406504065041e-06, | |
| "loss": 0.4638, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.757575757575758, | |
| "grad_norm": 9.165548124514086, | |
| "learning_rate": 3.0487804878048782e-06, | |
| "loss": 0.4911, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.797979797979798, | |
| "grad_norm": 8.906764540289775, | |
| "learning_rate": 2.947154471544716e-06, | |
| "loss": 0.4568, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.8383838383838382, | |
| "grad_norm": 9.034724254956192, | |
| "learning_rate": 2.845528455284553e-06, | |
| "loss": 0.4336, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.878787878787879, | |
| "grad_norm": 11.166876327596686, | |
| "learning_rate": 2.7439024390243905e-06, | |
| "loss": 0.4499, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.919191919191919, | |
| "grad_norm": 9.217158341597267, | |
| "learning_rate": 2.6422764227642278e-06, | |
| "loss": 0.421, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 3.9595959595959593, | |
| "grad_norm": 10.35588055954636, | |
| "learning_rate": 2.5406504065040655e-06, | |
| "loss": 0.461, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 9.702783400257974, | |
| "learning_rate": 2.4390243902439027e-06, | |
| "loss": 0.4693, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.040404040404041, | |
| "grad_norm": 7.785172054818246, | |
| "learning_rate": 2.33739837398374e-06, | |
| "loss": 0.2154, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.08080808080808, | |
| "grad_norm": 8.930594851346704, | |
| "learning_rate": 2.2357723577235773e-06, | |
| "loss": 0.1967, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.121212121212121, | |
| "grad_norm": 10.107028846989817, | |
| "learning_rate": 2.1341463414634146e-06, | |
| "loss": 0.1844, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.161616161616162, | |
| "grad_norm": 7.826371662630265, | |
| "learning_rate": 2.0325203252032523e-06, | |
| "loss": 0.19, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.202020202020202, | |
| "grad_norm": 7.454368070701538, | |
| "learning_rate": 1.9308943089430896e-06, | |
| "loss": 0.204, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.242424242424242, | |
| "grad_norm": 10.555419961914236, | |
| "learning_rate": 1.8292682926829268e-06, | |
| "loss": 0.1838, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.282828282828283, | |
| "grad_norm": 7.50612949709692, | |
| "learning_rate": 1.7276422764227643e-06, | |
| "loss": 0.2054, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.3232323232323235, | |
| "grad_norm": 7.8740842398900055, | |
| "learning_rate": 1.6260162601626018e-06, | |
| "loss": 0.1933, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 7.239232049440118, | |
| "learning_rate": 1.5243902439024391e-06, | |
| "loss": 0.195, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.404040404040404, | |
| "grad_norm": 8.52719981728453, | |
| "learning_rate": 1.4227642276422766e-06, | |
| "loss": 0.2201, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 6.870928125369453, | |
| "learning_rate": 1.3211382113821139e-06, | |
| "loss": 0.1876, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.484848484848484, | |
| "grad_norm": 7.742011223629616, | |
| "learning_rate": 1.2195121951219514e-06, | |
| "loss": 0.1811, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 4.525252525252525, | |
| "grad_norm": 7.506513525908142, | |
| "learning_rate": 1.1178861788617887e-06, | |
| "loss": 0.1805, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.565656565656566, | |
| "grad_norm": 7.401202667774116, | |
| "learning_rate": 1.0162601626016261e-06, | |
| "loss": 0.1951, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 4.606060606060606, | |
| "grad_norm": 8.135929523728391, | |
| "learning_rate": 9.146341463414634e-07, | |
| "loss": 0.1856, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.646464646464646, | |
| "grad_norm": 9.307497468880753, | |
| "learning_rate": 8.130081300813009e-07, | |
| "loss": 0.1807, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 4.686868686868687, | |
| "grad_norm": 8.362839951501785, | |
| "learning_rate": 7.113821138211383e-07, | |
| "loss": 0.1653, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 9.06145198185612, | |
| "learning_rate": 6.097560975609757e-07, | |
| "loss": 0.1747, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 4.767676767676767, | |
| "grad_norm": 7.9663701451927516, | |
| "learning_rate": 5.081300813008131e-07, | |
| "loss": 0.1922, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.808080808080808, | |
| "grad_norm": 8.512067102195044, | |
| "learning_rate": 4.0650406504065046e-07, | |
| "loss": 0.1837, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 4.848484848484849, | |
| "grad_norm": 7.294652883363056, | |
| "learning_rate": 3.0487804878048784e-07, | |
| "loss": 0.1877, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 8.551601154075781, | |
| "learning_rate": 2.0325203252032523e-07, | |
| "loss": 0.1851, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 4.929292929292929, | |
| "grad_norm": 8.034884715447463, | |
| "learning_rate": 1.0162601626016261e-07, | |
| "loss": 0.187, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.96969696969697, | |
| "grad_norm": 7.727841885701067, | |
| "learning_rate": 0.0, | |
| "loss": 0.1957, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 4.96969696969697, | |
| "step": 615, | |
| "total_flos": 849370300416.0, | |
| "train_loss": 0.9517717417662706, | |
| "train_runtime": 734.4559, | |
| "train_samples_per_second": 26.959, | |
| "train_steps_per_second": 0.837 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 615, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 849370300416.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |