| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 951, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010515247108307045, | |
| "grad_norm": 9.591360424317704, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "loss": 1.5585, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005257623554153523, | |
| "grad_norm": 9.792129596772575, | |
| "learning_rate": 1.0416666666666667e-06, | |
| "loss": 1.5717, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010515247108307046, | |
| "grad_norm": 3.5907937606092246, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "loss": 1.54, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015772870662460567, | |
| "grad_norm": 2.305935396467309, | |
| "learning_rate": 3.125e-06, | |
| "loss": 1.5164, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02103049421661409, | |
| "grad_norm": 1.3430988630406788, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.4398, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.026288117770767613, | |
| "grad_norm": 1.3395602302159622, | |
| "learning_rate": 5.208333333333334e-06, | |
| "loss": 1.4374, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.031545741324921134, | |
| "grad_norm": 0.927990494487892, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.3952, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03680336487907466, | |
| "grad_norm": 0.817869963107523, | |
| "learning_rate": 7.291666666666667e-06, | |
| "loss": 1.3688, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04206098843322818, | |
| "grad_norm": 0.8099795395904227, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.3543, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0473186119873817, | |
| "grad_norm": 0.8196989977897423, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 1.4065, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.052576235541535225, | |
| "grad_norm": 0.7891645835706297, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 1.3719, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05783385909568875, | |
| "grad_norm": 0.7730662990445142, | |
| "learning_rate": 1.1458333333333333e-05, | |
| "loss": 1.3534, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06309148264984227, | |
| "grad_norm": 0.8436428208666454, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.3484, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0683491062039958, | |
| "grad_norm": 0.930633139226441, | |
| "learning_rate": 1.3541666666666668e-05, | |
| "loss": 1.3339, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07360672975814932, | |
| "grad_norm": 1.2035603215763095, | |
| "learning_rate": 1.4583333333333333e-05, | |
| "loss": 1.3577, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07886435331230283, | |
| "grad_norm": 0.9963124148010385, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 1.3504, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08412197686645637, | |
| "grad_norm": 0.7705000485956546, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.3316, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08937960042060988, | |
| "grad_norm": 0.892949375851743, | |
| "learning_rate": 1.7708333333333335e-05, | |
| "loss": 1.3289, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0946372239747634, | |
| "grad_norm": 0.909849658521525, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 1.3311, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09989484752891693, | |
| "grad_norm": 0.8255721604143701, | |
| "learning_rate": 1.979166666666667e-05, | |
| "loss": 1.3224, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10515247108307045, | |
| "grad_norm": 0.8253137907707107, | |
| "learning_rate": 1.9998919935516768e-05, | |
| "loss": 1.3291, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11041009463722397, | |
| "grad_norm": 0.8322397994562853, | |
| "learning_rate": 1.999453257340926e-05, | |
| "loss": 1.321, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1156677181913775, | |
| "grad_norm": 0.8169700972120059, | |
| "learning_rate": 1.9986771889316172e-05, | |
| "loss": 1.3145, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12092534174553102, | |
| "grad_norm": 0.8054378165910604, | |
| "learning_rate": 1.9975640502598243e-05, | |
| "loss": 1.3537, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12618296529968454, | |
| "grad_norm": 0.8851185792466449, | |
| "learning_rate": 1.9961142170284762e-05, | |
| "loss": 1.3081, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13144058885383805, | |
| "grad_norm": 0.7983608745240434, | |
| "learning_rate": 1.9943281785805483e-05, | |
| "loss": 1.3235, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1366982124079916, | |
| "grad_norm": 0.8101924460502825, | |
| "learning_rate": 1.9922065377339037e-05, | |
| "loss": 1.3234, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14195583596214512, | |
| "grad_norm": 0.8181918475547033, | |
| "learning_rate": 1.98975001057783e-05, | |
| "loss": 1.3275, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.14721345951629863, | |
| "grad_norm": 0.8204094295073951, | |
| "learning_rate": 1.986959426231349e-05, | |
| "loss": 1.3191, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15247108307045215, | |
| "grad_norm": 0.8193214439419285, | |
| "learning_rate": 1.983835726563373e-05, | |
| "loss": 1.3151, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.15772870662460567, | |
| "grad_norm": 0.8008496299830612, | |
| "learning_rate": 1.9803799658748096e-05, | |
| "loss": 1.3173, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16298633017875921, | |
| "grad_norm": 0.8976369540265376, | |
| "learning_rate": 1.976593310542718e-05, | |
| "loss": 1.3193, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.16824395373291273, | |
| "grad_norm": 0.8219320759522418, | |
| "learning_rate": 1.9724770386266363e-05, | |
| "loss": 1.3074, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17350157728706625, | |
| "grad_norm": 0.7917666416050304, | |
| "learning_rate": 1.968032539437215e-05, | |
| "loss": 1.3229, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.17875920084121977, | |
| "grad_norm": 0.8468899498707804, | |
| "learning_rate": 1.963261313067302e-05, | |
| "loss": 1.3053, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18401682439537329, | |
| "grad_norm": 0.7826255604907222, | |
| "learning_rate": 1.958164969885636e-05, | |
| "loss": 1.2994, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1892744479495268, | |
| "grad_norm": 0.7854651229191463, | |
| "learning_rate": 1.9527452299933192e-05, | |
| "loss": 1.2933, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19453207150368035, | |
| "grad_norm": 0.8335518596689603, | |
| "learning_rate": 1.9470039226432562e-05, | |
| "loss": 1.3053, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.19978969505783387, | |
| "grad_norm": 0.7925970741738505, | |
| "learning_rate": 1.9409429856227487e-05, | |
| "loss": 1.3118, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20504731861198738, | |
| "grad_norm": 0.9022667633817576, | |
| "learning_rate": 1.934564464599461e-05, | |
| "loss": 1.3006, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2103049421661409, | |
| "grad_norm": 0.7846381821366918, | |
| "learning_rate": 1.9278705124309724e-05, | |
| "loss": 1.3019, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21556256572029442, | |
| "grad_norm": 0.8523088106258838, | |
| "learning_rate": 1.9208633884381528e-05, | |
| "loss": 1.3096, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22082018927444794, | |
| "grad_norm": 0.7773410227959319, | |
| "learning_rate": 1.913545457642601e-05, | |
| "loss": 1.3292, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.22607781282860148, | |
| "grad_norm": 0.7700372006553495, | |
| "learning_rate": 1.9059191899684154e-05, | |
| "loss": 1.3039, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.231335436382755, | |
| "grad_norm": 0.888501634972992, | |
| "learning_rate": 1.8979871594085482e-05, | |
| "loss": 1.2877, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23659305993690852, | |
| "grad_norm": 1.1002462859379447, | |
| "learning_rate": 1.8897520431560435e-05, | |
| "loss": 1.3015, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24185068349106204, | |
| "grad_norm": 0.8737204056440837, | |
| "learning_rate": 1.881216620700437e-05, | |
| "loss": 1.3115, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.24710830704521555, | |
| "grad_norm": 0.789475981373434, | |
| "learning_rate": 1.872383772889634e-05, | |
| "loss": 1.3046, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.25236593059936907, | |
| "grad_norm": 0.7857837110627846, | |
| "learning_rate": 1.863256480957574e-05, | |
| "loss": 1.314, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2576235541535226, | |
| "grad_norm": 0.8735939515101088, | |
| "learning_rate": 1.853837825518014e-05, | |
| "loss": 1.2965, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2628811777076761, | |
| "grad_norm": 0.8274285162573444, | |
| "learning_rate": 1.844130985524771e-05, | |
| "loss": 1.2847, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.26813880126182965, | |
| "grad_norm": 0.9655358070442068, | |
| "learning_rate": 1.83413923719877e-05, | |
| "loss": 1.3033, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2733964248159832, | |
| "grad_norm": 0.784933865330852, | |
| "learning_rate": 1.8238659529222672e-05, | |
| "loss": 1.2964, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2786540483701367, | |
| "grad_norm": 0.8072089928871935, | |
| "learning_rate": 1.813314600100612e-05, | |
| "loss": 1.3, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.28391167192429023, | |
| "grad_norm": 0.7600500534210821, | |
| "learning_rate": 1.802488739991941e-05, | |
| "loss": 1.2897, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2891692954784437, | |
| "grad_norm": 0.7716489846603405, | |
| "learning_rate": 1.7913920265051947e-05, | |
| "loss": 1.2994, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.29442691903259727, | |
| "grad_norm": 0.8439416659266935, | |
| "learning_rate": 1.7800282049668593e-05, | |
| "loss": 1.3146, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2996845425867508, | |
| "grad_norm": 0.7413366975663128, | |
| "learning_rate": 1.7684011108568593e-05, | |
| "loss": 1.3157, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3049421661409043, | |
| "grad_norm": 0.9324094146103826, | |
| "learning_rate": 1.7565146685140168e-05, | |
| "loss": 1.2944, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.31019978969505785, | |
| "grad_norm": 0.7497525070350954, | |
| "learning_rate": 1.7443728898115228e-05, | |
| "loss": 1.3041, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.31545741324921134, | |
| "grad_norm": 0.7274267632257158, | |
| "learning_rate": 1.7319798728028617e-05, | |
| "loss": 1.2855, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3207150368033649, | |
| "grad_norm": 0.7306525895599777, | |
| "learning_rate": 1.7193398003386514e-05, | |
| "loss": 1.2967, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.32597266035751843, | |
| "grad_norm": 0.7718504682998596, | |
| "learning_rate": 1.7064569386548586e-05, | |
| "loss": 1.3116, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3312302839116719, | |
| "grad_norm": 0.7002164572114049, | |
| "learning_rate": 1.6933356359328756e-05, | |
| "loss": 1.2812, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.33648790746582546, | |
| "grad_norm": 0.757112323690967, | |
| "learning_rate": 1.679980320831934e-05, | |
| "loss": 1.2654, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.34174553101997895, | |
| "grad_norm": 0.7509591301953, | |
| "learning_rate": 1.6663955009943603e-05, | |
| "loss": 1.2755, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3470031545741325, | |
| "grad_norm": 0.7896423206563724, | |
| "learning_rate": 1.6525857615241686e-05, | |
| "loss": 1.291, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.352260778128286, | |
| "grad_norm": 0.8538357892614057, | |
| "learning_rate": 1.6385557634395138e-05, | |
| "loss": 1.3, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.35751840168243953, | |
| "grad_norm": 0.8257603179264333, | |
| "learning_rate": 1.624310242099518e-05, | |
| "loss": 1.2825, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3627760252365931, | |
| "grad_norm": 0.7506565139625698, | |
| "learning_rate": 1.609854005606009e-05, | |
| "loss": 1.2903, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.36803364879074657, | |
| "grad_norm": 0.7962631813094616, | |
| "learning_rate": 1.5951919331807052e-05, | |
| "loss": 1.32, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3732912723449001, | |
| "grad_norm": 0.7846004527594455, | |
| "learning_rate": 1.5803289735183952e-05, | |
| "loss": 1.3094, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3785488958990536, | |
| "grad_norm": 0.7529193912570731, | |
| "learning_rate": 1.565270143116672e-05, | |
| "loss": 1.3097, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.38380651945320715, | |
| "grad_norm": 0.8450472817355228, | |
| "learning_rate": 1.5500205245827814e-05, | |
| "loss": 1.2954, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.3890641430073607, | |
| "grad_norm": 0.7884988874909659, | |
| "learning_rate": 1.5345852649181555e-05, | |
| "loss": 1.2774, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3943217665615142, | |
| "grad_norm": 0.7181038933510216, | |
| "learning_rate": 1.5189695737812153e-05, | |
| "loss": 1.2788, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.39957939011566773, | |
| "grad_norm": 0.7398759340722635, | |
| "learning_rate": 1.503178721729022e-05, | |
| "loss": 1.2825, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4048370136698212, | |
| "grad_norm": 0.6946338532175111, | |
| "learning_rate": 1.4872180384383772e-05, | |
| "loss": 1.2945, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.41009463722397477, | |
| "grad_norm": 0.7312230396618198, | |
| "learning_rate": 1.4710929109069674e-05, | |
| "loss": 1.2774, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4153522607781283, | |
| "grad_norm": 0.685269793573933, | |
| "learning_rate": 1.4548087816351616e-05, | |
| "loss": 1.2691, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4206098843322818, | |
| "grad_norm": 0.7367025121344211, | |
| "learning_rate": 1.4383711467890776e-05, | |
| "loss": 1.3029, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42586750788643535, | |
| "grad_norm": 0.954670307639884, | |
| "learning_rate": 1.4217855543455323e-05, | |
| "loss": 1.2846, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.43112513144058884, | |
| "grad_norm": 0.7884436143443709, | |
| "learning_rate": 1.4050576022195084e-05, | |
| "loss": 1.2686, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4363827549947424, | |
| "grad_norm": 0.6923648258189327, | |
| "learning_rate": 1.3881929363747628e-05, | |
| "loss": 1.2717, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4416403785488959, | |
| "grad_norm": 0.7131098366130528, | |
| "learning_rate": 1.3711972489182208e-05, | |
| "loss": 1.2968, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4468980021030494, | |
| "grad_norm": 0.7320310034272568, | |
| "learning_rate": 1.3540762761787938e-05, | |
| "loss": 1.2829, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.45215562565720296, | |
| "grad_norm": 0.7327893476336677, | |
| "learning_rate": 1.3368357967712726e-05, | |
| "loss": 1.2877, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.45741324921135645, | |
| "grad_norm": 0.6970014096675695, | |
| "learning_rate": 1.3194816296459483e-05, | |
| "loss": 1.2871, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.46267087276551, | |
| "grad_norm": 0.7014388522833548, | |
| "learning_rate": 1.302019632124619e-05, | |
| "loss": 1.2897, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4679284963196635, | |
| "grad_norm": 0.6970153763179125, | |
| "learning_rate": 1.2844556979236463e-05, | |
| "loss": 1.2714, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.47318611987381703, | |
| "grad_norm": 0.7162117080997548, | |
| "learning_rate": 1.2667957551647263e-05, | |
| "loss": 1.2705, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4784437434279706, | |
| "grad_norm": 0.7402201632812351, | |
| "learning_rate": 1.24904576437405e-05, | |
| "loss": 1.2654, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.48370136698212407, | |
| "grad_norm": 1.0567871681807486, | |
| "learning_rate": 1.2312117164705267e-05, | |
| "loss": 1.2784, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4889589905362776, | |
| "grad_norm": 0.7383322032141622, | |
| "learning_rate": 1.213299630743747e-05, | |
| "loss": 1.2574, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4942166140904311, | |
| "grad_norm": 0.7402615331947396, | |
| "learning_rate": 1.1953155528223728e-05, | |
| "loss": 1.2861, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.49947423764458465, | |
| "grad_norm": 0.7073342447220505, | |
| "learning_rate": 1.1772655526336367e-05, | |
| "loss": 1.2899, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5047318611987381, | |
| "grad_norm": 0.6966270957094003, | |
| "learning_rate": 1.1591557223546394e-05, | |
| "loss": 1.2607, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5099894847528917, | |
| "grad_norm": 0.6995026613750213, | |
| "learning_rate": 1.1409921743561383e-05, | |
| "loss": 1.285, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5152471083070452, | |
| "grad_norm": 0.677915660036085, | |
| "learning_rate": 1.1227810391395199e-05, | |
| "loss": 1.2787, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5205047318611987, | |
| "grad_norm": 0.694058172582696, | |
| "learning_rate": 1.1045284632676535e-05, | |
| "loss": 1.2817, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5257623554153522, | |
| "grad_norm": 0.6977571213673606, | |
| "learning_rate": 1.0862406072903224e-05, | |
| "loss": 1.269, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5310199789695058, | |
| "grad_norm": 0.709669977534319, | |
| "learning_rate": 1.067923643664936e-05, | |
| "loss": 1.2569, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5362776025236593, | |
| "grad_norm": 0.700607260206098, | |
| "learning_rate": 1.0495837546732224e-05, | |
| "loss": 1.2722, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5415352260778128, | |
| "grad_norm": 0.6965062537462676, | |
| "learning_rate": 1.031227130334604e-05, | |
| "loss": 1.2689, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5467928496319664, | |
| "grad_norm": 0.6975551930599063, | |
| "learning_rate": 1.0128599663169629e-05, | |
| "loss": 1.3171, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5520504731861199, | |
| "grad_norm": 0.6835553883087109, | |
| "learning_rate": 9.944884618454996e-06, | |
| "loss": 1.2616, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5573080967402734, | |
| "grad_norm": 0.6932270302779978, | |
| "learning_rate": 9.761188176103902e-06, | |
| "loss": 1.2842, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.562565720294427, | |
| "grad_norm": 0.7523295705383632, | |
| "learning_rate": 9.577572336739491e-06, | |
| "loss": 1.276, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5678233438485805, | |
| "grad_norm": 0.7410406370549468, | |
| "learning_rate": 9.394099073780066e-06, | |
| "loss": 1.2451, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.573080967402734, | |
| "grad_norm": 0.6626519082062838, | |
| "learning_rate": 9.210830312521991e-06, | |
| "loss": 1.2505, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5783385909568874, | |
| "grad_norm": 0.678178932637677, | |
| "learning_rate": 9.027827909238902e-06, | |
| "loss": 1.2884, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.583596214511041, | |
| "grad_norm": 0.6964235915302415, | |
| "learning_rate": 8.84515363030414e-06, | |
| "loss": 1.2674, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5888538380651945, | |
| "grad_norm": 0.7295327929832226, | |
| "learning_rate": 8.662869131343607e-06, | |
| "loss": 1.2606, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.594111461619348, | |
| "grad_norm": 0.6671644394075068, | |
| "learning_rate": 8.481035936425928e-06, | |
| "loss": 1.2631, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5993690851735016, | |
| "grad_norm": 0.6918129382316475, | |
| "learning_rate": 8.299715417297072e-06, | |
| "loss": 1.2733, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6046267087276551, | |
| "grad_norm": 0.7648575778497364, | |
| "learning_rate": 8.118968772666338e-06, | |
| "loss": 1.2768, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6098843322818086, | |
| "grad_norm": 0.6660976751002976, | |
| "learning_rate": 7.938857007550797e-06, | |
| "loss": 1.2712, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6151419558359621, | |
| "grad_norm": 0.8206761231157318, | |
| "learning_rate": 7.759440912685043e-06, | |
| "loss": 1.2629, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6203995793901157, | |
| "grad_norm": 0.6781893460495839, | |
| "learning_rate": 7.580781044003324e-06, | |
| "loss": 1.2928, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6256572029442692, | |
| "grad_norm": 0.7043738811990014, | |
| "learning_rate": 7.402937702200905e-06, | |
| "loss": 1.2565, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6309148264984227, | |
| "grad_norm": 0.6676378108729303, | |
| "learning_rate": 7.225970912381557e-06, | |
| "loss": 1.2441, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6361724500525763, | |
| "grad_norm": 0.7086319165657909, | |
| "learning_rate": 7.04994040379809e-06, | |
| "loss": 1.2526, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6414300736067298, | |
| "grad_norm": 0.6584513029407596, | |
| "learning_rate": 6.874905589692734e-06, | |
| "loss": 1.2689, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6466876971608833, | |
| "grad_norm": 0.6556498134167102, | |
| "learning_rate": 6.700925547244173e-06, | |
| "loss": 1.254, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6519453207150369, | |
| "grad_norm": 0.7317497625816267, | |
| "learning_rate": 6.528058997627995e-06, | |
| "loss": 1.2773, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6572029442691903, | |
| "grad_norm": 0.6943495622984553, | |
| "learning_rate": 6.356364286197341e-06, | |
| "loss": 1.2774, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6624605678233438, | |
| "grad_norm": 0.6557703327216338, | |
| "learning_rate": 6.18589936279034e-06, | |
| "loss": 1.2561, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6677181913774973, | |
| "grad_norm": 0.7479106302240468, | |
| "learning_rate": 6.016721762171098e-06, | |
| "loss": 1.2592, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6729758149316509, | |
| "grad_norm": 0.6501670860792287, | |
| "learning_rate": 5.848888584610727e-06, | |
| "loss": 1.2647, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6782334384858044, | |
| "grad_norm": 0.6750656043866252, | |
| "learning_rate": 5.6824564766150724e-06, | |
| "loss": 1.2687, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6834910620399579, | |
| "grad_norm": 0.707162362118668, | |
| "learning_rate": 5.51748161180554e-06, | |
| "loss": 1.2658, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6887486855941115, | |
| "grad_norm": 0.638182422688773, | |
| "learning_rate": 5.354019671959601e-06, | |
| "loss": 1.2618, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.694006309148265, | |
| "grad_norm": 0.6799258857804185, | |
| "learning_rate": 5.192125828217203e-06, | |
| "loss": 1.265, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6992639327024185, | |
| "grad_norm": 0.6614260729055945, | |
| "learning_rate": 5.0318547224596525e-06, | |
| "loss": 1.2726, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.704521556256572, | |
| "grad_norm": 0.6571261971890797, | |
| "learning_rate": 4.873260448867004e-06, | |
| "loss": 1.2635, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7097791798107256, | |
| "grad_norm": 0.6434154288185444, | |
| "learning_rate": 4.716396535660412e-06, | |
| "loss": 1.2571, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7150368033648791, | |
| "grad_norm": 0.6507843588939136, | |
| "learning_rate": 4.5613159270354455e-06, | |
| "loss": 1.2768, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7202944269190326, | |
| "grad_norm": 0.6888038425645421, | |
| "learning_rate": 4.408070965292534e-06, | |
| "loss": 1.2729, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7255520504731862, | |
| "grad_norm": 0.6548677910438372, | |
| "learning_rate": 4.256713373170565e-06, | |
| "loss": 1.2754, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7308096740273397, | |
| "grad_norm": 0.6924489060189015, | |
| "learning_rate": 4.107294236389603e-06, | |
| "loss": 1.2428, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7360672975814931, | |
| "grad_norm": 0.675766971656018, | |
| "learning_rate": 3.959863986408593e-06, | |
| "loss": 1.2422, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7413249211356467, | |
| "grad_norm": 0.6759780214552317, | |
| "learning_rate": 3.8144723834039076e-06, | |
| "loss": 1.2596, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7465825446898002, | |
| "grad_norm": 0.6577848370025432, | |
| "learning_rate": 3.671168499474449e-06, | |
| "loss": 1.2653, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7518401682439537, | |
| "grad_norm": 0.6816662338694881, | |
| "learning_rate": 3.5300007020789997e-06, | |
| "loss": 1.2612, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7570977917981072, | |
| "grad_norm": 0.6412644189429695, | |
| "learning_rate": 3.3910166377113894e-06, | |
| "loss": 1.2606, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7623554153522608, | |
| "grad_norm": 0.6489342949048617, | |
| "learning_rate": 3.2542632158190135e-06, | |
| "loss": 1.2499, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7676130389064143, | |
| "grad_norm": 0.6305858451193405, | |
| "learning_rate": 3.119786592970102e-06, | |
| "loss": 1.2368, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7728706624605678, | |
| "grad_norm": 0.6601632070842128, | |
| "learning_rate": 2.9876321572751143e-06, | |
| "loss": 1.2826, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7781282860147214, | |
| "grad_norm": 0.6512679140886882, | |
| "learning_rate": 2.8578445130674835e-06, | |
| "loss": 1.2509, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7833859095688749, | |
| "grad_norm": 0.662446849135675, | |
| "learning_rate": 2.7304674658489104e-06, | |
| "loss": 1.2593, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.7886435331230284, | |
| "grad_norm": 0.6503725043711072, | |
| "learning_rate": 2.6055440075042793e-06, | |
| "loss": 1.2696, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7939011566771819, | |
| "grad_norm": 0.6349030168998883, | |
| "learning_rate": 2.4831163017911687e-06, | |
| "loss": 1.2458, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.7991587802313355, | |
| "grad_norm": 0.6296499888770365, | |
| "learning_rate": 2.3632256701088817e-06, | |
| "loss": 1.2581, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.804416403785489, | |
| "grad_norm": 0.6477912073667947, | |
| "learning_rate": 2.2459125775517854e-06, | |
| "loss": 1.2614, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8096740273396424, | |
| "grad_norm": 0.6340147585135972, | |
| "learning_rate": 2.1312166192516593e-06, | |
| "loss": 1.2707, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.814931650893796, | |
| "grad_norm": 0.6535117628009195, | |
| "learning_rate": 2.019176507013677e-06, | |
| "loss": 1.2586, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8201892744479495, | |
| "grad_norm": 0.6467399510691909, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 1.2548, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.825446898002103, | |
| "grad_norm": 0.6237075846859721, | |
| "learning_rate": 1.8032141732190722e-06, | |
| "loss": 1.26, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8307045215562566, | |
| "grad_norm": 0.6374095389974708, | |
| "learning_rate": 1.6993648425638797e-06, | |
| "loss": 1.2605, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8359621451104101, | |
| "grad_norm": 0.6326077447290254, | |
| "learning_rate": 1.5983171151717924e-06, | |
| "loss": 1.2701, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8412197686645636, | |
| "grad_norm": 0.6299244980807773, | |
| "learning_rate": 1.5001050963416718e-06, | |
| "loss": 1.2427, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8464773922187171, | |
| "grad_norm": 0.6281934013360307, | |
| "learning_rate": 1.404761934273291e-06, | |
| "loss": 1.2635, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8517350157728707, | |
| "grad_norm": 0.6374294536071755, | |
| "learning_rate": 1.3123198088792577e-06, | |
| "loss": 1.2717, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8569926393270242, | |
| "grad_norm": 0.6298699771941692, | |
| "learning_rate": 1.222809920923761e-06, | |
| "loss": 1.2499, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8622502628811777, | |
| "grad_norm": 0.6359419744147026, | |
| "learning_rate": 1.1362624814917843e-06, | |
| "loss": 1.2552, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8675078864353313, | |
| "grad_norm": 0.6285813274381151, | |
| "learning_rate": 1.0527067017923654e-06, | |
| "loss": 1.2737, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8727655099894848, | |
| "grad_norm": 0.6210713562517961, | |
| "learning_rate": 9.721707832993232e-07, | |
| "loss": 1.2663, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8780231335436383, | |
| "grad_norm": 0.6645416418213763, | |
| "learning_rate": 8.946819082327829e-07, | |
| "loss": 1.2581, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8832807570977917, | |
| "grad_norm": 0.6251751866944497, | |
| "learning_rate": 8.202662303847298e-07, | |
| "loss": 1.2519, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8885383806519453, | |
| "grad_norm": 0.6431790207111582, | |
| "learning_rate": 7.48948866291661e-07, | |
| "loss": 1.2614, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.8937960042060988, | |
| "grad_norm": 0.6384407054635708, | |
| "learning_rate": 6.80753886757336e-07, | |
| "loss": 1.2722, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8990536277602523, | |
| "grad_norm": 0.631430226459749, | |
| "learning_rate": 6.157043087284797e-07, | |
| "loss": 1.2587, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9043112513144059, | |
| "grad_norm": 0.631843611779411, | |
| "learning_rate": 5.538220875261736e-07, | |
| "loss": 1.2481, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9095688748685594, | |
| "grad_norm": 0.6250494334708813, | |
| "learning_rate": 4.951281094355708e-07, | |
| "loss": 1.2552, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9148264984227129, | |
| "grad_norm": 0.6300532606909163, | |
| "learning_rate": 4.396421846564236e-07, | |
| "loss": 1.2536, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9200841219768665, | |
| "grad_norm": 0.6200012582073006, | |
| "learning_rate": 3.8738304061681107e-07, | |
| "loss": 1.2694, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.92534174553102, | |
| "grad_norm": 0.6324403821172889, | |
| "learning_rate": 3.3836831565231877e-07, | |
| "loss": 1.2545, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9305993690851735, | |
| "grad_norm": 0.6325686488791483, | |
| "learning_rate": 2.926145530528002e-07, | |
| "loss": 1.2531, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.935856992639327, | |
| "grad_norm": 0.6373731914514896, | |
| "learning_rate": 2.501371954787479e-07, | |
| "loss": 1.2648, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9411146161934806, | |
| "grad_norm": 0.6255593953506046, | |
| "learning_rate": 2.109505797491318e-07, | |
| "loss": 1.2392, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.9463722397476341, | |
| "grad_norm": 0.6353907042408413, | |
| "learning_rate": 1.7506793200248507e-07, | |
| "loss": 1.2633, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9516298633017876, | |
| "grad_norm": 0.6307207707434458, | |
| "learning_rate": 1.4250136323285868e-07, | |
| "loss": 1.2603, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9568874868559412, | |
| "grad_norm": 0.6277705175750871, | |
| "learning_rate": 1.1326186520215888e-07, | |
| "loss": 1.2453, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9621451104100947, | |
| "grad_norm": 0.6031775390358434, | |
| "learning_rate": 8.735930673024806e-08, | |
| "loss": 1.2544, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9674027339642481, | |
| "grad_norm": 0.6189714576376613, | |
| "learning_rate": 6.480243036404598e-08, | |
| "loss": 1.2661, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9726603575184016, | |
| "grad_norm": 0.6219922382476483, | |
| "learning_rate": 4.5598849426777833e-08, | |
| "loss": 1.2418, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9779179810725552, | |
| "grad_norm": 0.6188912441406357, | |
| "learning_rate": 2.9755045448351948e-08, | |
| "loss": 1.2471, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9831756046267087, | |
| "grad_norm": 0.6241442557644146, | |
| "learning_rate": 1.7276365977730858e-08, | |
| "loss": 1.2591, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9884332281808622, | |
| "grad_norm": 0.6148942917091785, | |
| "learning_rate": 8.167022778045042e-09, | |
| "loss": 1.2783, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9936908517350158, | |
| "grad_norm": 0.6142845224019251, | |
| "learning_rate": 2.430090405054486e-09, | |
| "loss": 1.2311, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.9989484752891693, | |
| "grad_norm": 0.6318778333420784, | |
| "learning_rate": 6.750516943321295e-11, | |
| "loss": 1.2488, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.266036868095398, | |
| "eval_runtime": 32.6481, | |
| "eval_samples_per_second": 412.213, | |
| "eval_steps_per_second": 6.463, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 951, | |
| "total_flos": 101091718987776.0, | |
| "train_loss": 1.289618753484371, | |
| "train_runtime": 1098.6048, | |
| "train_samples_per_second": 110.718, | |
| "train_steps_per_second": 0.866 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 951, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 101091718987776.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |