| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.7312614259597806, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018281535648994515, | |
| "grad_norm": 2.0932669659411407, | |
| "learning_rate": 0.0, | |
| "loss": 0.5957, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003656307129798903, | |
| "grad_norm": 2.2225871592919857, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 0.5994, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005484460694698354, | |
| "grad_norm": 2.1034495331981367, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.595, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007312614259597806, | |
| "grad_norm": 2.166865863914735, | |
| "learning_rate": 3.2142857142857143e-06, | |
| "loss": 0.6304, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009140767824497258, | |
| "grad_norm": 1.9888391803618566, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.5996, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010968921389396709, | |
| "grad_norm": 1.574113693444916, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 0.5897, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012797074954296161, | |
| "grad_norm": 1.2137552408963244, | |
| "learning_rate": 6.428571428571429e-06, | |
| "loss": 0.6007, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014625228519195612, | |
| "grad_norm": 1.0350502098616132, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.568, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016453382084095063, | |
| "grad_norm": 0.894475082603116, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.5448, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018281535648994516, | |
| "grad_norm": 1.7395859008823984, | |
| "learning_rate": 9.642857142857144e-06, | |
| "loss": 0.5899, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02010968921389397, | |
| "grad_norm": 1.8244614303880249, | |
| "learning_rate": 1.0714285714285714e-05, | |
| "loss": 0.5746, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.021937842778793418, | |
| "grad_norm": 1.5925645409877431, | |
| "learning_rate": 1.1785714285714286e-05, | |
| "loss": 0.5565, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02376599634369287, | |
| "grad_norm": 1.36746956600136, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.5471, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025594149908592323, | |
| "grad_norm": 1.3707686542130966, | |
| "learning_rate": 1.3928571428571429e-05, | |
| "loss": 0.5331, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.027422303473491772, | |
| "grad_norm": 1.242238871065678, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.5699, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.029250457038391225, | |
| "grad_norm": 1.0311550418009305, | |
| "learning_rate": 1.6071428571428572e-05, | |
| "loss": 0.5308, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.031078610603290677, | |
| "grad_norm": 0.7712626361430579, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.4846, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03290676416819013, | |
| "grad_norm": 0.7368030930782065, | |
| "learning_rate": 1.8214285714285712e-05, | |
| "loss": 0.5252, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03473491773308958, | |
| "grad_norm": 0.7569594066617599, | |
| "learning_rate": 1.928571428571429e-05, | |
| "loss": 0.5004, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03656307129798903, | |
| "grad_norm": 0.775121486961269, | |
| "learning_rate": 2.0357142857142858e-05, | |
| "loss": 0.5081, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.038391224862888484, | |
| "grad_norm": 0.6937386237956013, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.4823, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04021937842778794, | |
| "grad_norm": 0.5513659607436744, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.4939, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04204753199268738, | |
| "grad_norm": 0.5026124919107627, | |
| "learning_rate": 2.357142857142857e-05, | |
| "loss": 0.486, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.043875685557586835, | |
| "grad_norm": 0.5955689249641745, | |
| "learning_rate": 2.464285714285714e-05, | |
| "loss": 0.5142, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04570383912248629, | |
| "grad_norm": 0.632394003160485, | |
| "learning_rate": 2.5714285714285714e-05, | |
| "loss": 0.5214, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04753199268738574, | |
| "grad_norm": 0.5309532283880241, | |
| "learning_rate": 2.6785714285714288e-05, | |
| "loss": 0.4774, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04936014625228519, | |
| "grad_norm": 0.4875964236553614, | |
| "learning_rate": 2.7857142857142858e-05, | |
| "loss": 0.5032, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.051188299817184646, | |
| "grad_norm": 0.48155522874845014, | |
| "learning_rate": 2.892857142857143e-05, | |
| "loss": 0.5002, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05301645338208409, | |
| "grad_norm": 0.41845545316273386, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4736, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.054844606946983544, | |
| "grad_norm": 0.426756665035819, | |
| "learning_rate": 2.999972519478696e-05, | |
| "loss": 0.4988, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.056672760511883, | |
| "grad_norm": 0.44740653887205223, | |
| "learning_rate": 2.99989007892169e-05, | |
| "loss": 0.4877, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05850091407678245, | |
| "grad_norm": 0.35965997153522855, | |
| "learning_rate": 2.9997526813496602e-05, | |
| "loss": 0.4791, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0603290676416819, | |
| "grad_norm": 0.3544019526584708, | |
| "learning_rate": 2.9995603317969497e-05, | |
| "loss": 0.4868, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.062157221206581355, | |
| "grad_norm": 0.3939085264133643, | |
| "learning_rate": 2.99931303731138e-05, | |
| "loss": 0.4933, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06398537477148081, | |
| "grad_norm": 0.357457478844048, | |
| "learning_rate": 2.9990108069539932e-05, | |
| "loss": 0.4788, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06581352833638025, | |
| "grad_norm": 0.36148014385057126, | |
| "learning_rate": 2.9986536517987187e-05, | |
| "loss": 0.4861, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06764168190127971, | |
| "grad_norm": 0.3891946625957914, | |
| "learning_rate": 2.998241584931971e-05, | |
| "loss": 0.4728, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06946983546617916, | |
| "grad_norm": 0.3498249933538063, | |
| "learning_rate": 2.9977746214521646e-05, | |
| "loss": 0.4714, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0712979890310786, | |
| "grad_norm": 0.32795069985610614, | |
| "learning_rate": 2.997252778469168e-05, | |
| "loss": 0.4582, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07312614259597806, | |
| "grad_norm": 0.3229786426441325, | |
| "learning_rate": 2.9966760751036697e-05, | |
| "loss": 0.4566, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07495429616087751, | |
| "grad_norm": 0.34189242509234885, | |
| "learning_rate": 2.9960445324864815e-05, | |
| "loss": 0.4669, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07678244972577697, | |
| "grad_norm": 0.3346248011539502, | |
| "learning_rate": 2.995358173757765e-05, | |
| "loss": 0.4719, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07861060329067641, | |
| "grad_norm": 0.3445606995846261, | |
| "learning_rate": 2.994617024066181e-05, | |
| "loss": 0.4657, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.08043875685557587, | |
| "grad_norm": 0.3084736463888063, | |
| "learning_rate": 2.9938211105679677e-05, | |
| "loss": 0.4545, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08226691042047532, | |
| "grad_norm": 0.363723170271056, | |
| "learning_rate": 2.9929704624259508e-05, | |
| "loss": 0.4734, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08409506398537477, | |
| "grad_norm": 0.3434867460486705, | |
| "learning_rate": 2.992065110808469e-05, | |
| "loss": 0.4716, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08592321755027423, | |
| "grad_norm": 0.3091094518128968, | |
| "learning_rate": 2.991105088888234e-05, | |
| "loss": 0.4892, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08775137111517367, | |
| "grad_norm": 0.28634845553939514, | |
| "learning_rate": 2.990090431841117e-05, | |
| "loss": 0.4877, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08957952468007313, | |
| "grad_norm": 0.32229417127358173, | |
| "learning_rate": 2.9890211768448572e-05, | |
| "loss": 0.4665, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09140767824497258, | |
| "grad_norm": 0.2998997014253872, | |
| "learning_rate": 2.9878973630777012e-05, | |
| "loss": 0.4763, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09323583180987204, | |
| "grad_norm": 0.31761959826728753, | |
| "learning_rate": 2.9867190317169665e-05, | |
| "loss": 0.4796, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09506398537477148, | |
| "grad_norm": 0.31856904774743844, | |
| "learning_rate": 2.9854862259375326e-05, | |
| "loss": 0.4795, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09689213893967093, | |
| "grad_norm": 0.3211537093722151, | |
| "learning_rate": 2.9841989909102607e-05, | |
| "loss": 0.4831, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09872029250457039, | |
| "grad_norm": 0.2759276306734268, | |
| "learning_rate": 2.982857373800337e-05, | |
| "loss": 0.4818, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10054844606946983, | |
| "grad_norm": 0.3157463034106083, | |
| "learning_rate": 2.9814614237655445e-05, | |
| "loss": 0.4623, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10237659963436929, | |
| "grad_norm": 0.3032312590655212, | |
| "learning_rate": 2.9800111919544632e-05, | |
| "loss": 0.4858, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10420475319926874, | |
| "grad_norm": 0.3113927848158391, | |
| "learning_rate": 2.9785067315045943e-05, | |
| "loss": 0.4805, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10603290676416818, | |
| "grad_norm": 0.2744027006341311, | |
| "learning_rate": 2.9769480975404143e-05, | |
| "loss": 0.4461, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10786106032906764, | |
| "grad_norm": 0.29230266010233613, | |
| "learning_rate": 2.975335347171356e-05, | |
| "loss": 0.4276, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10968921389396709, | |
| "grad_norm": 0.2889756002721036, | |
| "learning_rate": 2.9736685394897123e-05, | |
| "loss": 0.4591, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11151736745886655, | |
| "grad_norm": 0.3033109334524834, | |
| "learning_rate": 2.9719477355684767e-05, | |
| "loss": 0.4567, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.113345521023766, | |
| "grad_norm": 0.2781677992045063, | |
| "learning_rate": 2.9701729984591003e-05, | |
| "loss": 0.4398, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11517367458866545, | |
| "grad_norm": 0.271555541969376, | |
| "learning_rate": 2.9683443931891837e-05, | |
| "loss": 0.4587, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1170018281535649, | |
| "grad_norm": 0.28462475447605384, | |
| "learning_rate": 2.966461986760096e-05, | |
| "loss": 0.4454, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11882998171846434, | |
| "grad_norm": 0.5503364181994502, | |
| "learning_rate": 2.964525848144517e-05, | |
| "loss": 0.4224, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1206581352833638, | |
| "grad_norm": 0.36156039057533235, | |
| "learning_rate": 2.9625360482839114e-05, | |
| "loss": 0.4832, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12248628884826325, | |
| "grad_norm": 0.28852346241189863, | |
| "learning_rate": 2.9604926600859287e-05, | |
| "loss": 0.4674, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12431444241316271, | |
| "grad_norm": 0.301231419014804, | |
| "learning_rate": 2.958395758421733e-05, | |
| "loss": 0.4497, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12614259597806216, | |
| "grad_norm": 0.2780712443173308, | |
| "learning_rate": 2.9562454201232583e-05, | |
| "loss": 0.477, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12797074954296161, | |
| "grad_norm": 0.32406424984925947, | |
| "learning_rate": 2.9540417239803954e-05, | |
| "loss": 0.4667, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12979890310786105, | |
| "grad_norm": 0.2623604313786073, | |
| "learning_rate": 2.9517847507381022e-05, | |
| "loss": 0.4627, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1316270566727605, | |
| "grad_norm": 0.2977740538239248, | |
| "learning_rate": 2.9494745830934473e-05, | |
| "loss": 0.4599, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13345521023765997, | |
| "grad_norm": 0.3519063323266925, | |
| "learning_rate": 2.9471113056925786e-05, | |
| "loss": 0.4945, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13528336380255943, | |
| "grad_norm": 0.2993177800143668, | |
| "learning_rate": 2.944695005127623e-05, | |
| "loss": 0.4585, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13711151736745886, | |
| "grad_norm": 0.28778657695750315, | |
| "learning_rate": 2.942225769933512e-05, | |
| "loss": 0.4368, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13893967093235832, | |
| "grad_norm": 0.28115223256932165, | |
| "learning_rate": 2.939703690584741e-05, | |
| "loss": 0.4556, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.14076782449725778, | |
| "grad_norm": 0.2665269198819317, | |
| "learning_rate": 2.9371288594920484e-05, | |
| "loss": 0.4334, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1425959780621572, | |
| "grad_norm": 0.24969029459114356, | |
| "learning_rate": 2.934501370999037e-05, | |
| "loss": 0.4243, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14442413162705667, | |
| "grad_norm": 0.28019989404552176, | |
| "learning_rate": 2.93182132137871e-05, | |
| "loss": 0.4465, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14625228519195613, | |
| "grad_norm": 0.3262629362699753, | |
| "learning_rate": 2.9290888088299486e-05, | |
| "loss": 0.4824, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1480804387568556, | |
| "grad_norm": 0.3012249689853169, | |
| "learning_rate": 2.9263039334739127e-05, | |
| "loss": 0.4624, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14990859232175502, | |
| "grad_norm": 0.28055708820894665, | |
| "learning_rate": 2.92346679735037e-05, | |
| "loss": 0.4562, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.15173674588665448, | |
| "grad_norm": 0.27808074144332834, | |
| "learning_rate": 2.9205775044139608e-05, | |
| "loss": 0.4686, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15356489945155394, | |
| "grad_norm": 0.2806140830160609, | |
| "learning_rate": 2.9176361605303867e-05, | |
| "loss": 0.4695, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15539305301645337, | |
| "grad_norm": 0.2705147511752726, | |
| "learning_rate": 2.914642873472531e-05, | |
| "loss": 0.4524, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15722120658135283, | |
| "grad_norm": 0.32806782334241796, | |
| "learning_rate": 2.9115977529165132e-05, | |
| "loss": 0.4625, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.1590493601462523, | |
| "grad_norm": 0.2819528809225014, | |
| "learning_rate": 2.9085009104376663e-05, | |
| "loss": 0.4563, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.16087751371115175, | |
| "grad_norm": 0.3080336756223102, | |
| "learning_rate": 2.905352459506452e-05, | |
| "loss": 0.4369, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16270566727605118, | |
| "grad_norm": 0.30703315538546055, | |
| "learning_rate": 2.9021525154842998e-05, | |
| "loss": 0.4617, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16453382084095064, | |
| "grad_norm": 0.30026528997859425, | |
| "learning_rate": 2.8989011956193834e-05, | |
| "loss": 0.4586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1663619744058501, | |
| "grad_norm": 0.29970528278132424, | |
| "learning_rate": 2.8955986190423225e-05, | |
| "loss": 0.4847, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16819012797074953, | |
| "grad_norm": 0.2943719684747729, | |
| "learning_rate": 2.892244906761819e-05, | |
| "loss": 0.4464, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.170018281535649, | |
| "grad_norm": 0.30807210021130066, | |
| "learning_rate": 2.8888401816602207e-05, | |
| "loss": 0.45, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.17184643510054845, | |
| "grad_norm": 0.2763940481138354, | |
| "learning_rate": 2.885384568489023e-05, | |
| "loss": 0.4343, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1736745886654479, | |
| "grad_norm": 0.3253203044927249, | |
| "learning_rate": 2.881878193864294e-05, | |
| "loss": 0.4659, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17550274223034734, | |
| "grad_norm": 0.266312604175572, | |
| "learning_rate": 2.878321186262037e-05, | |
| "loss": 0.4417, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1773308957952468, | |
| "grad_norm": 0.3137831393989107, | |
| "learning_rate": 2.8747136760134827e-05, | |
| "loss": 0.4528, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.17915904936014626, | |
| "grad_norm": 0.2880929427628198, | |
| "learning_rate": 2.871055795300315e-05, | |
| "loss": 0.4395, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.1809872029250457, | |
| "grad_norm": 0.27062645315880496, | |
| "learning_rate": 2.8673476781498242e-05, | |
| "loss": 0.4306, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18281535648994515, | |
| "grad_norm": 0.28008745353310155, | |
| "learning_rate": 2.8635894604300018e-05, | |
| "loss": 0.4477, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1846435100548446, | |
| "grad_norm": 0.25820871027942605, | |
| "learning_rate": 2.859781279844556e-05, | |
| "loss": 0.4478, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.18647166361974407, | |
| "grad_norm": 0.30634908739409245, | |
| "learning_rate": 2.855923275927871e-05, | |
| "loss": 0.4732, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1882998171846435, | |
| "grad_norm": 0.27737282375395794, | |
| "learning_rate": 2.8520155900398922e-05, | |
| "loss": 0.4347, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.19012797074954296, | |
| "grad_norm": 0.3201019734940999, | |
| "learning_rate": 2.8480583653609457e-05, | |
| "loss": 0.4773, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.19195612431444242, | |
| "grad_norm": 0.25763164370530267, | |
| "learning_rate": 2.844051746886495e-05, | |
| "loss": 0.452, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.19378427787934185, | |
| "grad_norm": 0.26590404580502985, | |
| "learning_rate": 2.8399958814218258e-05, | |
| "loss": 0.4649, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1956124314442413, | |
| "grad_norm": 0.31104547148499023, | |
| "learning_rate": 2.8358909175766674e-05, | |
| "loss": 0.4669, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.19744058500914077, | |
| "grad_norm": 0.3436561078000218, | |
| "learning_rate": 2.831737005759749e-05, | |
| "loss": 0.4818, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.19926873857404023, | |
| "grad_norm": 0.2978747020545755, | |
| "learning_rate": 2.8275342981732868e-05, | |
| "loss": 0.4775, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.20109689213893966, | |
| "grad_norm": 0.3297773258073724, | |
| "learning_rate": 2.8232829488074077e-05, | |
| "loss": 0.4353, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.20292504570383912, | |
| "grad_norm": 0.26736943136912134, | |
| "learning_rate": 2.8189831134345074e-05, | |
| "loss": 0.4402, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.20475319926873858, | |
| "grad_norm": 0.46856659250948707, | |
| "learning_rate": 2.8146349496035426e-05, | |
| "loss": 0.4268, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.20658135283363802, | |
| "grad_norm": 0.27589234154547465, | |
| "learning_rate": 2.8102386166342582e-05, | |
| "loss": 0.4417, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20840950639853748, | |
| "grad_norm": 0.35079551159665306, | |
| "learning_rate": 2.8057942756113504e-05, | |
| "loss": 0.4772, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.21023765996343693, | |
| "grad_norm": 0.2882498554120959, | |
| "learning_rate": 2.8013020893785635e-05, | |
| "loss": 0.454, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.21206581352833637, | |
| "grad_norm": 0.2941664738490589, | |
| "learning_rate": 2.796762222532723e-05, | |
| "loss": 0.4293, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.21389396709323583, | |
| "grad_norm": 0.2702184421809731, | |
| "learning_rate": 2.7921748414177063e-05, | |
| "loss": 0.4687, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21572212065813529, | |
| "grad_norm": 0.3301109570692598, | |
| "learning_rate": 2.787540114118345e-05, | |
| "loss": 0.4345, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.21755027422303475, | |
| "grad_norm": 0.27446129864889174, | |
| "learning_rate": 2.7828582104542694e-05, | |
| "loss": 0.4469, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21937842778793418, | |
| "grad_norm": 0.29213098317806835, | |
| "learning_rate": 2.7781293019736845e-05, | |
| "loss": 0.4395, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22120658135283364, | |
| "grad_norm": 0.29952710363645, | |
| "learning_rate": 2.7733535619470835e-05, | |
| "loss": 0.4482, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2230347349177331, | |
| "grad_norm": 0.33479362279837327, | |
| "learning_rate": 2.7685311653609004e-05, | |
| "loss": 0.4436, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.22486288848263253, | |
| "grad_norm": 0.2943017968120905, | |
| "learning_rate": 2.7636622889110975e-05, | |
| "loss": 0.4751, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.226691042047532, | |
| "grad_norm": 0.30034787188266204, | |
| "learning_rate": 2.758747110996693e-05, | |
| "loss": 0.4646, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22851919561243145, | |
| "grad_norm": 0.28893394314876153, | |
| "learning_rate": 2.7537858117132217e-05, | |
| "loss": 0.451, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2303473491773309, | |
| "grad_norm": 0.2971868357645203, | |
| "learning_rate": 2.7487785728461383e-05, | |
| "loss": 0.4613, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.23217550274223034, | |
| "grad_norm": 0.3404033245887309, | |
| "learning_rate": 2.7437255778641548e-05, | |
| "loss": 0.4685, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2340036563071298, | |
| "grad_norm": 0.27331638590219476, | |
| "learning_rate": 2.7386270119125193e-05, | |
| "loss": 0.448, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23583180987202926, | |
| "grad_norm": 0.272124504795692, | |
| "learning_rate": 2.7334830618062327e-05, | |
| "loss": 0.4333, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2376599634369287, | |
| "grad_norm": 0.29807736925005807, | |
| "learning_rate": 2.728293916023202e-05, | |
| "loss": 0.4372, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23948811700182815, | |
| "grad_norm": 0.31922280763058825, | |
| "learning_rate": 2.7230597646973355e-05, | |
| "loss": 0.4701, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.2413162705667276, | |
| "grad_norm": 0.30586172261152206, | |
| "learning_rate": 2.717780799611576e-05, | |
| "loss": 0.4677, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24314442413162707, | |
| "grad_norm": 0.31340471142779364, | |
| "learning_rate": 2.7124572141908737e-05, | |
| "loss": 0.4643, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2449725776965265, | |
| "grad_norm": 0.2816615247738506, | |
| "learning_rate": 2.707089203495098e-05, | |
| "loss": 0.4503, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24680073126142596, | |
| "grad_norm": 0.33804681353834326, | |
| "learning_rate": 2.701676964211893e-05, | |
| "loss": 0.4874, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24862888482632542, | |
| "grad_norm": 0.30236149555278835, | |
| "learning_rate": 2.696220694649467e-05, | |
| "loss": 0.4311, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.25045703839122485, | |
| "grad_norm": 0.30621062436628155, | |
| "learning_rate": 2.69072059472933e-05, | |
| "loss": 0.4593, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2522851919561243, | |
| "grad_norm": 0.32266904581704525, | |
| "learning_rate": 2.685176865978965e-05, | |
| "loss": 0.4543, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25411334552102377, | |
| "grad_norm": 0.2880721603734785, | |
| "learning_rate": 2.6795897115244478e-05, | |
| "loss": 0.428, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.25594149908592323, | |
| "grad_norm": 0.25365800448080494, | |
| "learning_rate": 2.6739593360830006e-05, | |
| "loss": 0.4113, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2577696526508227, | |
| "grad_norm": 0.3019181957537817, | |
| "learning_rate": 2.668285945955493e-05, | |
| "loss": 0.4664, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2595978062157221, | |
| "grad_norm": 0.3267274278580188, | |
| "learning_rate": 2.6625697490188832e-05, | |
| "loss": 0.4504, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.26142595978062155, | |
| "grad_norm": 0.3069320624243545, | |
| "learning_rate": 2.6568109547185996e-05, | |
| "loss": 0.4782, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.263254113345521, | |
| "grad_norm": 0.2940302664382109, | |
| "learning_rate": 2.651009774060868e-05, | |
| "loss": 0.4678, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.26508226691042047, | |
| "grad_norm": 0.2856321101391868, | |
| "learning_rate": 2.6451664196049802e-05, | |
| "loss": 0.4465, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26691042047531993, | |
| "grad_norm": 0.26787811332236683, | |
| "learning_rate": 2.639281105455505e-05, | |
| "loss": 0.4508, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2687385740402194, | |
| "grad_norm": 0.310781815465313, | |
| "learning_rate": 2.6333540472544442e-05, | |
| "loss": 0.434, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.27056672760511885, | |
| "grad_norm": 0.2831519687807007, | |
| "learning_rate": 2.6273854621733286e-05, | |
| "loss": 0.4506, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.27239488117001825, | |
| "grad_norm": 0.26444215825774453, | |
| "learning_rate": 2.621375568905266e-05, | |
| "loss": 0.4491, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2742230347349177, | |
| "grad_norm": 0.3318439638760175, | |
| "learning_rate": 2.615324587656921e-05, | |
| "loss": 0.4817, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2760511882998172, | |
| "grad_norm": 0.26186010556255945, | |
| "learning_rate": 2.6092327401404538e-05, | |
| "loss": 0.4749, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.27787934186471663, | |
| "grad_norm": 0.28214623332209515, | |
| "learning_rate": 2.6031002495653913e-05, | |
| "loss": 0.4341, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2797074954296161, | |
| "grad_norm": 0.2658749515925076, | |
| "learning_rate": 2.596927340630451e-05, | |
| "loss": 0.4356, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.28153564899451555, | |
| "grad_norm": 0.2919462760948431, | |
| "learning_rate": 2.590714239515306e-05, | |
| "loss": 0.4445, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.283363802559415, | |
| "grad_norm": 0.2729523412693323, | |
| "learning_rate": 2.584461173872301e-05, | |
| "loss": 0.453, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2851919561243144, | |
| "grad_norm": 0.27376503902302884, | |
| "learning_rate": 2.5781683728181066e-05, | |
| "loss": 0.4408, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2870201096892139, | |
| "grad_norm": 0.29825503051650754, | |
| "learning_rate": 2.5718360669253276e-05, | |
| "loss": 0.4534, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.28884826325411334, | |
| "grad_norm": 0.28587305936658125, | |
| "learning_rate": 2.565464488214053e-05, | |
| "loss": 0.4377, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2906764168190128, | |
| "grad_norm": 0.2514941220386994, | |
| "learning_rate": 2.559053870143356e-05, | |
| "loss": 0.4278, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.29250457038391225, | |
| "grad_norm": 0.3064320998896908, | |
| "learning_rate": 2.5526044476027383e-05, | |
| "loss": 0.4606, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2943327239488117, | |
| "grad_norm": 0.2996428461412703, | |
| "learning_rate": 2.5461164569035244e-05, | |
| "loss": 0.4406, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2961608775137112, | |
| "grad_norm": 0.24445658979859197, | |
| "learning_rate": 2.5395901357702032e-05, | |
| "loss": 0.4512, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2979890310786106, | |
| "grad_norm": 0.3020928493183503, | |
| "learning_rate": 2.533025723331718e-05, | |
| "loss": 0.4381, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.29981718464351004, | |
| "grad_norm": 0.259497736171246, | |
| "learning_rate": 2.526423460112703e-05, | |
| "loss": 0.4317, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3016453382084095, | |
| "grad_norm": 0.2845997617061945, | |
| "learning_rate": 2.5197835880246702e-05, | |
| "loss": 0.4508, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.30347349177330896, | |
| "grad_norm": 0.25599143813626735, | |
| "learning_rate": 2.51310635035715e-05, | |
| "loss": 0.4439, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3053016453382084, | |
| "grad_norm": 0.27151337502431655, | |
| "learning_rate": 2.506391991768771e-05, | |
| "loss": 0.4437, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3071297989031079, | |
| "grad_norm": 0.2820431046523887, | |
| "learning_rate": 2.4996407582782987e-05, | |
| "loss": 0.4303, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.30895795246800734, | |
| "grad_norm": 0.2722072935751051, | |
| "learning_rate": 2.4928528972556207e-05, | |
| "loss": 0.4404, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.31078610603290674, | |
| "grad_norm": 0.27234372659701445, | |
| "learning_rate": 2.486028657412683e-05, | |
| "loss": 0.4651, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3126142595978062, | |
| "grad_norm": 0.2887259505587887, | |
| "learning_rate": 2.479168288794377e-05, | |
| "loss": 0.4258, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.31444241316270566, | |
| "grad_norm": 0.2863499537204956, | |
| "learning_rate": 2.472272042769377e-05, | |
| "loss": 0.4417, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3162705667276051, | |
| "grad_norm": 0.28012635605541286, | |
| "learning_rate": 2.4653401720209316e-05, | |
| "loss": 0.4339, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3180987202925046, | |
| "grad_norm": 0.2901255279537208, | |
| "learning_rate": 2.4583729305376014e-05, | |
| "loss": 0.4434, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.31992687385740404, | |
| "grad_norm": 0.28670533171458207, | |
| "learning_rate": 2.451370573603959e-05, | |
| "loss": 0.4465, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3217550274223035, | |
| "grad_norm": 0.29099195398037236, | |
| "learning_rate": 2.4443333577912285e-05, | |
| "loss": 0.4502, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3235831809872029, | |
| "grad_norm": 0.2767320359979587, | |
| "learning_rate": 2.437261540947889e-05, | |
| "loss": 0.4419, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.32541133455210236, | |
| "grad_norm": 0.27043885704043746, | |
| "learning_rate": 2.430155382190225e-05, | |
| "loss": 0.4599, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3272394881170018, | |
| "grad_norm": 0.2730084443234302, | |
| "learning_rate": 2.4230151418928326e-05, | |
| "loss": 0.4414, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3290676416819013, | |
| "grad_norm": 0.26865234353182316, | |
| "learning_rate": 2.415841081679079e-05, | |
| "loss": 0.4566, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.33089579524680074, | |
| "grad_norm": 0.2587412822837069, | |
| "learning_rate": 2.4086334644115176e-05, | |
| "loss": 0.4165, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3327239488117002, | |
| "grad_norm": 0.2576294969452134, | |
| "learning_rate": 2.4013925541822543e-05, | |
| "loss": 0.4389, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.33455210237659966, | |
| "grad_norm": 0.28831507664483796, | |
| "learning_rate": 2.3941186163032736e-05, | |
| "loss": 0.4423, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.33638025594149906, | |
| "grad_norm": 0.3030948383504952, | |
| "learning_rate": 2.3868119172967145e-05, | |
| "loss": 0.4714, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3382084095063985, | |
| "grad_norm": 0.27032107322162624, | |
| "learning_rate": 2.379472724885108e-05, | |
| "loss": 0.421, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.340036563071298, | |
| "grad_norm": 0.2760164862865866, | |
| "learning_rate": 2.3721013079815645e-05, | |
| "loss": 0.4447, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.34186471663619744, | |
| "grad_norm": 0.30959882474810846, | |
| "learning_rate": 2.3646979366799234e-05, | |
| "loss": 0.4609, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3436928702010969, | |
| "grad_norm": 0.2714448004833152, | |
| "learning_rate": 2.3572628822448546e-05, | |
| "loss": 0.4507, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.34552102376599636, | |
| "grad_norm": 0.27266712502337187, | |
| "learning_rate": 2.3497964171019214e-05, | |
| "loss": 0.4365, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.3473491773308958, | |
| "grad_norm": 0.2781660755274924, | |
| "learning_rate": 2.3422988148275963e-05, | |
| "loss": 0.4456, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3491773308957952, | |
| "grad_norm": 0.2555222680884862, | |
| "learning_rate": 2.3347703501392373e-05, | |
| "loss": 0.4539, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.3510054844606947, | |
| "grad_norm": 0.2564207916106098, | |
| "learning_rate": 2.3272112988850237e-05, | |
| "loss": 0.4341, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.35283363802559414, | |
| "grad_norm": 0.3184883384427764, | |
| "learning_rate": 2.319621938033848e-05, | |
| "loss": 0.442, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3546617915904936, | |
| "grad_norm": 0.26891911220862924, | |
| "learning_rate": 2.3120025456651658e-05, | |
| "loss": 0.4526, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.35648994515539306, | |
| "grad_norm": 0.2508425033641723, | |
| "learning_rate": 2.3043534009588108e-05, | |
| "loss": 0.4403, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3583180987202925, | |
| "grad_norm": 0.2844052123597973, | |
| "learning_rate": 2.296674784184761e-05, | |
| "loss": 0.4127, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.360146252285192, | |
| "grad_norm": 0.26258111786205446, | |
| "learning_rate": 2.288966976692873e-05, | |
| "loss": 0.4211, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.3619744058500914, | |
| "grad_norm": 0.2520374791669725, | |
| "learning_rate": 2.2812302609025692e-05, | |
| "loss": 0.4221, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.36380255941499084, | |
| "grad_norm": 0.26337297446508356, | |
| "learning_rate": 2.2734649202924955e-05, | |
| "loss": 0.4522, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3656307129798903, | |
| "grad_norm": 0.2855343544923256, | |
| "learning_rate": 2.265671239390128e-05, | |
| "loss": 0.4523, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36745886654478976, | |
| "grad_norm": 0.28164010246391696, | |
| "learning_rate": 2.2578495037613532e-05, | |
| "loss": 0.4432, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3692870201096892, | |
| "grad_norm": 0.24907385038322546, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.4553, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3711151736745887, | |
| "grad_norm": 0.2627538166376267, | |
| "learning_rate": 2.242123015717343e-05, | |
| "loss": 0.4349, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.37294332723948814, | |
| "grad_norm": 0.31354766743356877, | |
| "learning_rate": 2.2342188395315614e-05, | |
| "loss": 0.4432, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.37477148080438755, | |
| "grad_norm": 0.23675269296963747, | |
| "learning_rate": 2.226287761057165e-05, | |
| "loss": 0.418, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.376599634369287, | |
| "grad_norm": 0.2891554833178968, | |
| "learning_rate": 2.2183300708943803e-05, | |
| "loss": 0.4489, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.37842778793418647, | |
| "grad_norm": 0.26640363819873775, | |
| "learning_rate": 2.2103460606185075e-05, | |
| "loss": 0.4615, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3802559414990859, | |
| "grad_norm": 0.2558864794570339, | |
| "learning_rate": 2.2023360227692323e-05, | |
| "loss": 0.4261, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3820840950639854, | |
| "grad_norm": 0.2699523067538174, | |
| "learning_rate": 2.194300250839908e-05, | |
| "loss": 0.4195, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.38391224862888484, | |
| "grad_norm": 0.27340419517313214, | |
| "learning_rate": 2.1862390392668047e-05, | |
| "loss": 0.4346, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3857404021937843, | |
| "grad_norm": 0.27350160128660483, | |
| "learning_rate": 2.1781526834183172e-05, | |
| "loss": 0.4342, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3875685557586837, | |
| "grad_norm": 0.2708971031355086, | |
| "learning_rate": 2.1700414795841442e-05, | |
| "loss": 0.4273, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.38939670932358317, | |
| "grad_norm": 0.2550932743091051, | |
| "learning_rate": 2.1619057249644323e-05, | |
| "loss": 0.414, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3912248628884826, | |
| "grad_norm": 0.2803194395165915, | |
| "learning_rate": 2.1537457176588854e-05, | |
| "loss": 0.4552, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3930530164533821, | |
| "grad_norm": 0.3055183656958912, | |
| "learning_rate": 2.145561756655844e-05, | |
| "loss": 0.467, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.39488117001828155, | |
| "grad_norm": 0.2559113693108124, | |
| "learning_rate": 2.137354141821326e-05, | |
| "loss": 0.4606, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.396709323583181, | |
| "grad_norm": 0.26418557735551346, | |
| "learning_rate": 2.129123173888045e-05, | |
| "loss": 0.4265, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.39853747714808047, | |
| "grad_norm": 0.28274527019614765, | |
| "learning_rate": 2.120869154444387e-05, | |
| "loss": 0.4446, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.40036563071297987, | |
| "grad_norm": 0.27808247484752874, | |
| "learning_rate": 2.1125923859233613e-05, | |
| "loss": 0.4444, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.40219378427787933, | |
| "grad_norm": 0.27782505030038473, | |
| "learning_rate": 2.1042931715915194e-05, | |
| "loss": 0.4549, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4040219378427788, | |
| "grad_norm": 0.27007523589065596, | |
| "learning_rate": 2.0959718155378434e-05, | |
| "loss": 0.4201, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.40585009140767825, | |
| "grad_norm": 0.24533996656769982, | |
| "learning_rate": 2.0876286226626027e-05, | |
| "loss": 0.4394, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4076782449725777, | |
| "grad_norm": 0.25493787596449796, | |
| "learning_rate": 2.0792638986661835e-05, | |
| "loss": 0.4361, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.40950639853747717, | |
| "grad_norm": 0.24866181426568823, | |
| "learning_rate": 2.0708779500378876e-05, | |
| "loss": 0.422, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.4113345521023766, | |
| "grad_norm": 0.2626340677947448, | |
| "learning_rate": 2.0624710840447002e-05, | |
| "loss": 0.4699, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.41316270566727603, | |
| "grad_norm": 0.2548747162562123, | |
| "learning_rate": 2.054043608720036e-05, | |
| "loss": 0.4381, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.4149908592321755, | |
| "grad_norm": 0.25440840922549496, | |
| "learning_rate": 2.0455958328524484e-05, | |
| "loss": 0.4201, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.41681901279707495, | |
| "grad_norm": 0.26449635353916284, | |
| "learning_rate": 2.0371280659743166e-05, | |
| "loss": 0.4554, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4186471663619744, | |
| "grad_norm": 0.24905974740356548, | |
| "learning_rate": 2.0286406183505042e-05, | |
| "loss": 0.4464, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.42047531992687387, | |
| "grad_norm": 0.2615769980174596, | |
| "learning_rate": 2.0201338009669928e-05, | |
| "loss": 0.4349, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.42230347349177333, | |
| "grad_norm": 0.23359275701455742, | |
| "learning_rate": 2.0116079255194836e-05, | |
| "loss": 0.4081, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.42413162705667273, | |
| "grad_norm": 0.24808773148753901, | |
| "learning_rate": 2.0030633044019785e-05, | |
| "loss": 0.4298, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.4259597806215722, | |
| "grad_norm": 0.24232976420589095, | |
| "learning_rate": 1.9945002506953354e-05, | |
| "loss": 0.4409, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.42778793418647165, | |
| "grad_norm": 0.22567433822272406, | |
| "learning_rate": 1.9859190781557938e-05, | |
| "loss": 0.442, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.4296160877513711, | |
| "grad_norm": 0.2311061708408994, | |
| "learning_rate": 1.9773201012034795e-05, | |
| "loss": 0.421, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.43144424131627057, | |
| "grad_norm": 0.22816213383251369, | |
| "learning_rate": 1.9687036349108852e-05, | |
| "loss": 0.4158, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.43327239488117003, | |
| "grad_norm": 0.2510785001261421, | |
| "learning_rate": 1.9600699949913258e-05, | |
| "loss": 0.4431, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.4351005484460695, | |
| "grad_norm": 0.2598409752211816, | |
| "learning_rate": 1.951419497787367e-05, | |
| "loss": 0.4458, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.4369287020109689, | |
| "grad_norm": 0.24332508968296312, | |
| "learning_rate": 1.942752460259241e-05, | |
| "loss": 0.4349, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.43875685557586835, | |
| "grad_norm": 0.2552207965650322, | |
| "learning_rate": 1.934069199973226e-05, | |
| "loss": 0.451, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4405850091407678, | |
| "grad_norm": 0.3134487399340424, | |
| "learning_rate": 1.9253700350900144e-05, | |
| "loss": 0.4774, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4424131627056673, | |
| "grad_norm": 0.24561994226229733, | |
| "learning_rate": 1.916655284353055e-05, | |
| "loss": 0.4113, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.44424131627056673, | |
| "grad_norm": 0.2441778108075136, | |
| "learning_rate": 1.9079252670768712e-05, | |
| "loss": 0.4164, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4460694698354662, | |
| "grad_norm": 0.24667253648402243, | |
| "learning_rate": 1.8991803031353648e-05, | |
| "loss": 0.4667, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.44789762340036565, | |
| "grad_norm": 0.2532308170425146, | |
| "learning_rate": 1.8904207129500916e-05, | |
| "loss": 0.4507, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.44972577696526506, | |
| "grad_norm": 0.2410145562683754, | |
| "learning_rate": 1.881646817478526e-05, | |
| "loss": 0.4335, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4515539305301645, | |
| "grad_norm": 0.2677011816262894, | |
| "learning_rate": 1.8728589382022958e-05, | |
| "loss": 0.4655, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.453382084095064, | |
| "grad_norm": 0.2813593321239021, | |
| "learning_rate": 1.864057397115406e-05, | |
| "loss": 0.404, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.45521023765996343, | |
| "grad_norm": 0.2335241250466671, | |
| "learning_rate": 1.8552425167124402e-05, | |
| "loss": 0.4455, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.4570383912248629, | |
| "grad_norm": 0.28481078145389616, | |
| "learning_rate": 1.8464146199767426e-05, | |
| "loss": 0.4479, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.45886654478976235, | |
| "grad_norm": 0.29084428436925236, | |
| "learning_rate": 1.837574030368586e-05, | |
| "loss": 0.432, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.4606946983546618, | |
| "grad_norm": 0.2526474847324583, | |
| "learning_rate": 1.8287210718133185e-05, | |
| "loss": 0.4475, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4625228519195612, | |
| "grad_norm": 0.23451195798066396, | |
| "learning_rate": 1.819856068689495e-05, | |
| "loss": 0.4271, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.4643510054844607, | |
| "grad_norm": 0.2857912680872776, | |
| "learning_rate": 1.8109793458169914e-05, | |
| "loss": 0.4127, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.46617915904936014, | |
| "grad_norm": 0.2202114882392875, | |
| "learning_rate": 1.8020912284451047e-05, | |
| "loss": 0.4268, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.4680073126142596, | |
| "grad_norm": 0.2847281184055229, | |
| "learning_rate": 1.7931920422406323e-05, | |
| "loss": 0.4259, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.46983546617915906, | |
| "grad_norm": 0.2598281156255513, | |
| "learning_rate": 1.7842821132759427e-05, | |
| "loss": 0.4514, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.4716636197440585, | |
| "grad_norm": 0.2844988917742255, | |
| "learning_rate": 1.775361768017026e-05, | |
| "loss": 0.4615, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.473491773308958, | |
| "grad_norm": 0.26617898814153196, | |
| "learning_rate": 1.766431333311533e-05, | |
| "loss": 0.4426, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.4753199268738574, | |
| "grad_norm": 0.2428332403956185, | |
| "learning_rate": 1.757491136376799e-05, | |
| "loss": 0.4386, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.47714808043875684, | |
| "grad_norm": 0.2502587869532296, | |
| "learning_rate": 1.7485415047878525e-05, | |
| "loss": 0.4371, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.4789762340036563, | |
| "grad_norm": 0.25465501530170725, | |
| "learning_rate": 1.7395827664654168e-05, | |
| "loss": 0.404, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.48080438756855576, | |
| "grad_norm": 0.27989566357801815, | |
| "learning_rate": 1.7306152496638906e-05, | |
| "loss": 0.4461, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.4826325411334552, | |
| "grad_norm": 0.2671213153187633, | |
| "learning_rate": 1.7216392829593225e-05, | |
| "loss": 0.4461, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.4844606946983547, | |
| "grad_norm": 0.24388873946187056, | |
| "learning_rate": 1.712655195237371e-05, | |
| "loss": 0.4366, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.48628884826325414, | |
| "grad_norm": 0.2621521928469628, | |
| "learning_rate": 1.703663315681256e-05, | |
| "loss": 0.4361, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.48811700182815354, | |
| "grad_norm": 0.23355568735312865, | |
| "learning_rate": 1.6946639737596936e-05, | |
| "loss": 0.4547, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.489945155393053, | |
| "grad_norm": 0.2433323374022999, | |
| "learning_rate": 1.685657499214827e-05, | |
| "loss": 0.4217, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.49177330895795246, | |
| "grad_norm": 0.2835980161687909, | |
| "learning_rate": 1.676644222050144e-05, | |
| "loss": 0.4699, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.4936014625228519, | |
| "grad_norm": 0.2332386437859208, | |
| "learning_rate": 1.6676244725183848e-05, | |
| "loss": 0.4312, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4954296160877514, | |
| "grad_norm": 0.21880840810143093, | |
| "learning_rate": 1.6585985811094412e-05, | |
| "loss": 0.4264, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.49725776965265084, | |
| "grad_norm": 0.26293014355482136, | |
| "learning_rate": 1.6495668785382482e-05, | |
| "loss": 0.4377, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.4990859232175503, | |
| "grad_norm": 0.25742637339428476, | |
| "learning_rate": 1.6405296957326665e-05, | |
| "loss": 0.4443, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5009140767824497, | |
| "grad_norm": 0.2433810121654413, | |
| "learning_rate": 1.631487363821354e-05, | |
| "loss": 0.4496, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.5027422303473492, | |
| "grad_norm": 0.25311836103685337, | |
| "learning_rate": 1.6224402141216387e-05, | |
| "loss": 0.4309, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5045703839122486, | |
| "grad_norm": 0.2363827435920569, | |
| "learning_rate": 1.6133885781273724e-05, | |
| "loss": 0.4444, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.506398537477148, | |
| "grad_norm": 0.2375498739699462, | |
| "learning_rate": 1.6043327874967907e-05, | |
| "loss": 0.4264, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5082266910420475, | |
| "grad_norm": 0.23971752210791406, | |
| "learning_rate": 1.595273174040357e-05, | |
| "loss": 0.4547, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5100548446069469, | |
| "grad_norm": 0.2413334639110406, | |
| "learning_rate": 1.5862100697086037e-05, | |
| "loss": 0.4327, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.5118829981718465, | |
| "grad_norm": 0.26601032650299644, | |
| "learning_rate": 1.5771438065799745e-05, | |
| "loss": 0.4471, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5137111517367459, | |
| "grad_norm": 0.24442841821130334, | |
| "learning_rate": 1.568074716848652e-05, | |
| "loss": 0.4338, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5155393053016454, | |
| "grad_norm": 0.24161461577302404, | |
| "learning_rate": 1.559003132812387e-05, | |
| "loss": 0.4258, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5173674588665448, | |
| "grad_norm": 0.2678319638095495, | |
| "learning_rate": 1.5499293868603244e-05, | |
| "loss": 0.4445, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5191956124314442, | |
| "grad_norm": 0.23988755994528907, | |
| "learning_rate": 1.5408538114608227e-05, | |
| "loss": 0.4708, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5210237659963437, | |
| "grad_norm": 0.23428382288193408, | |
| "learning_rate": 1.531776739149273e-05, | |
| "loss": 0.4623, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5228519195612431, | |
| "grad_norm": 0.2536087816515709, | |
| "learning_rate": 1.522698502515913e-05, | |
| "loss": 0.4261, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5246800731261426, | |
| "grad_norm": 0.24842397791428794, | |
| "learning_rate": 1.513619434193645e-05, | |
| "loss": 0.4366, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.526508226691042, | |
| "grad_norm": 0.2250112860332024, | |
| "learning_rate": 1.5045398668458405e-05, | |
| "loss": 0.4089, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5283363802559415, | |
| "grad_norm": 0.37363010526021784, | |
| "learning_rate": 1.4954601331541598e-05, | |
| "loss": 0.4222, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5301645338208409, | |
| "grad_norm": 0.22688923246924197, | |
| "learning_rate": 1.4863805658063553e-05, | |
| "loss": 0.449, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5319926873857403, | |
| "grad_norm": 0.23042126253173695, | |
| "learning_rate": 1.4773014974840868e-05, | |
| "loss": 0.403, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5338208409506399, | |
| "grad_norm": 0.23434077527141906, | |
| "learning_rate": 1.4682232608507276e-05, | |
| "loss": 0.4234, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5356489945155393, | |
| "grad_norm": 0.2533253906652841, | |
| "learning_rate": 1.4591461885391774e-05, | |
| "loss": 0.4707, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5374771480804388, | |
| "grad_norm": 0.2367196781343478, | |
| "learning_rate": 1.4500706131396758e-05, | |
| "loss": 0.4561, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5393053016453382, | |
| "grad_norm": 0.24155540742179146, | |
| "learning_rate": 1.440996867187613e-05, | |
| "loss": 0.4213, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5411334552102377, | |
| "grad_norm": 0.23888649479779048, | |
| "learning_rate": 1.4319252831513483e-05, | |
| "loss": 0.4539, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5429616087751371, | |
| "grad_norm": 0.2399360717612798, | |
| "learning_rate": 1.4228561934200252e-05, | |
| "loss": 0.4294, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5447897623400365, | |
| "grad_norm": 0.21310241182150114, | |
| "learning_rate": 1.4137899302913967e-05, | |
| "loss": 0.4092, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.546617915904936, | |
| "grad_norm": 0.24287450467753255, | |
| "learning_rate": 1.4047268259596438e-05, | |
| "loss": 0.4393, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5484460694698354, | |
| "grad_norm": 0.3527758266951121, | |
| "learning_rate": 1.3956672125032095e-05, | |
| "loss": 0.4297, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5502742230347349, | |
| "grad_norm": 0.2587107143122098, | |
| "learning_rate": 1.3866114218726276e-05, | |
| "loss": 0.4397, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.5521023765996343, | |
| "grad_norm": 0.21583428645223188, | |
| "learning_rate": 1.3775597858783618e-05, | |
| "loss": 0.4042, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5539305301645339, | |
| "grad_norm": 0.23908697208787028, | |
| "learning_rate": 1.3685126361786459e-05, | |
| "loss": 0.4645, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.5557586837294333, | |
| "grad_norm": 0.22752588531186632, | |
| "learning_rate": 1.359470304267334e-05, | |
| "loss": 0.4293, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5575868372943327, | |
| "grad_norm": 0.23542512484036232, | |
| "learning_rate": 1.3504331214617522e-05, | |
| "loss": 0.404, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5594149908592322, | |
| "grad_norm": 0.25330785799275174, | |
| "learning_rate": 1.3414014188905592e-05, | |
| "loss": 0.4272, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5612431444241316, | |
| "grad_norm": 0.234302364995047, | |
| "learning_rate": 1.3323755274816156e-05, | |
| "loss": 0.4218, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5630712979890311, | |
| "grad_norm": 0.21718003102435057, | |
| "learning_rate": 1.323355777949856e-05, | |
| "loss": 0.435, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5648994515539305, | |
| "grad_norm": 0.23771222899292566, | |
| "learning_rate": 1.3143425007851731e-05, | |
| "loss": 0.4027, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.56672760511883, | |
| "grad_norm": 0.21581964509667692, | |
| "learning_rate": 1.3053360262403065e-05, | |
| "loss": 0.4275, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5685557586837294, | |
| "grad_norm": 0.23676296954191686, | |
| "learning_rate": 1.2963366843187444e-05, | |
| "loss": 0.4573, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5703839122486288, | |
| "grad_norm": 0.22115404157686464, | |
| "learning_rate": 1.2873448047626292e-05, | |
| "loss": 0.4256, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5722120658135283, | |
| "grad_norm": 0.22618312895939494, | |
| "learning_rate": 1.278360717040678e-05, | |
| "loss": 0.4412, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.5740402193784278, | |
| "grad_norm": 0.22530056042606073, | |
| "learning_rate": 1.2693847503361098e-05, | |
| "loss": 0.4396, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.5758683729433273, | |
| "grad_norm": 0.23534013353212493, | |
| "learning_rate": 1.2604172335345831e-05, | |
| "loss": 0.4285, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5776965265082267, | |
| "grad_norm": 0.22290585998943552, | |
| "learning_rate": 1.2514584952121476e-05, | |
| "loss": 0.4307, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5795246800731262, | |
| "grad_norm": 0.2175202235328278, | |
| "learning_rate": 1.2425088636232018e-05, | |
| "loss": 0.4242, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.5813528336380256, | |
| "grad_norm": 0.23770394090188834, | |
| "learning_rate": 1.233568666688467e-05, | |
| "loss": 0.425, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.583180987202925, | |
| "grad_norm": 0.24602097112526436, | |
| "learning_rate": 1.2246382319829742e-05, | |
| "loss": 0.4314, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.5850091407678245, | |
| "grad_norm": 0.21380429110162302, | |
| "learning_rate": 1.2157178867240578e-05, | |
| "loss": 0.4124, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5868372943327239, | |
| "grad_norm": 0.2059051087152771, | |
| "learning_rate": 1.2068079577593677e-05, | |
| "loss": 0.434, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.5886654478976234, | |
| "grad_norm": 0.20782805102967639, | |
| "learning_rate": 1.1979087715548956e-05, | |
| "loss": 0.426, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.5904936014625228, | |
| "grad_norm": 0.2308269667389375, | |
| "learning_rate": 1.1890206541830081e-05, | |
| "loss": 0.4367, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.5923217550274223, | |
| "grad_norm": 0.22433989736933327, | |
| "learning_rate": 1.1801439313105053e-05, | |
| "loss": 0.4137, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.5941499085923218, | |
| "grad_norm": 0.24818633510952665, | |
| "learning_rate": 1.171278928186682e-05, | |
| "loss": 0.4326, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5959780621572212, | |
| "grad_norm": 0.22641907903537842, | |
| "learning_rate": 1.1624259696314142e-05, | |
| "loss": 0.4452, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5978062157221207, | |
| "grad_norm": 0.2502208733018346, | |
| "learning_rate": 1.1535853800232578e-05, | |
| "loss": 0.4219, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.5996343692870201, | |
| "grad_norm": 0.20315718232089036, | |
| "learning_rate": 1.1447574832875602e-05, | |
| "loss": 0.4146, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6014625228519196, | |
| "grad_norm": 0.2704027878651348, | |
| "learning_rate": 1.1359426028845942e-05, | |
| "loss": 0.4388, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.603290676416819, | |
| "grad_norm": 0.2474545966686057, | |
| "learning_rate": 1.1271410617977048e-05, | |
| "loss": 0.4316, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6051188299817185, | |
| "grad_norm": 0.23074391315897977, | |
| "learning_rate": 1.1183531825214744e-05, | |
| "loss": 0.4513, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.6069469835466179, | |
| "grad_norm": 0.25731515715188485, | |
| "learning_rate": 1.1095792870499087e-05, | |
| "loss": 0.4366, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.6087751371115173, | |
| "grad_norm": 0.25634778394355273, | |
| "learning_rate": 1.1008196968646356e-05, | |
| "loss": 0.4718, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.6106032906764168, | |
| "grad_norm": 0.26331951561134626, | |
| "learning_rate": 1.0920747329231289e-05, | |
| "loss": 0.4551, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.6124314442413162, | |
| "grad_norm": 0.3504082117772083, | |
| "learning_rate": 1.0833447156469448e-05, | |
| "loss": 0.4125, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6142595978062158, | |
| "grad_norm": 0.2574920075665294, | |
| "learning_rate": 1.0746299649099855e-05, | |
| "loss": 0.4539, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6160877513711152, | |
| "grad_norm": 0.2292032218757861, | |
| "learning_rate": 1.0659308000267744e-05, | |
| "loss": 0.4305, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6179159049360147, | |
| "grad_norm": 0.26470754729984625, | |
| "learning_rate": 1.0572475397407594e-05, | |
| "loss": 0.4363, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6197440585009141, | |
| "grad_norm": 0.24478744270497613, | |
| "learning_rate": 1.0485805022126331e-05, | |
| "loss": 0.431, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6215722120658135, | |
| "grad_norm": 0.22868825374949903, | |
| "learning_rate": 1.0399300050086748e-05, | |
| "loss": 0.4127, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.623400365630713, | |
| "grad_norm": 0.22844457512096136, | |
| "learning_rate": 1.0312963650891147e-05, | |
| "loss": 0.4248, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6252285191956124, | |
| "grad_norm": 0.20397429178914392, | |
| "learning_rate": 1.0226798987965207e-05, | |
| "loss": 0.397, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6270566727605119, | |
| "grad_norm": 0.24300323205974386, | |
| "learning_rate": 1.014080921844207e-05, | |
| "loss": 0.4192, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6288848263254113, | |
| "grad_norm": 0.23615643791202737, | |
| "learning_rate": 1.0054997493046647e-05, | |
| "loss": 0.4252, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6307129798903108, | |
| "grad_norm": 0.22224915020860084, | |
| "learning_rate": 9.969366955980217e-06, | |
| "loss": 0.442, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6325411334552102, | |
| "grad_norm": 0.21254715342037545, | |
| "learning_rate": 9.883920744805168e-06, | |
| "loss": 0.4269, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6343692870201096, | |
| "grad_norm": 0.2505381298484902, | |
| "learning_rate": 9.798661990330073e-06, | |
| "loss": 0.4491, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6361974405850092, | |
| "grad_norm": 0.222922428077655, | |
| "learning_rate": 9.713593816494955e-06, | |
| "loss": 0.4284, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6380255941499086, | |
| "grad_norm": 0.24435028204482773, | |
| "learning_rate": 9.628719340256842e-06, | |
| "loss": 0.4212, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6398537477148081, | |
| "grad_norm": 0.2141788033954091, | |
| "learning_rate": 9.544041671475522e-06, | |
| "loss": 0.4302, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6416819012797075, | |
| "grad_norm": 0.21973056409704741, | |
| "learning_rate": 9.45956391279964e-06, | |
| "loss": 0.4418, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.643510054844607, | |
| "grad_norm": 0.21401712763189235, | |
| "learning_rate": 9.375289159552999e-06, | |
| "loss": 0.4373, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6453382084095064, | |
| "grad_norm": 0.2390813821307003, | |
| "learning_rate": 9.291220499621128e-06, | |
| "loss": 0.4292, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6471663619744058, | |
| "grad_norm": 0.23488978656838674, | |
| "learning_rate": 9.207361013338166e-06, | |
| "loss": 0.4413, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6489945155393053, | |
| "grad_norm": 0.32109686809085475, | |
| "learning_rate": 9.123713773373977e-06, | |
| "loss": 0.444, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6508226691042047, | |
| "grad_norm": 0.23126269385774315, | |
| "learning_rate": 9.040281844621568e-06, | |
| "loss": 0.4485, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.6526508226691042, | |
| "grad_norm": 0.2312349995045198, | |
| "learning_rate": 8.95706828408481e-06, | |
| "loss": 0.4656, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.6544789762340036, | |
| "grad_norm": 0.22595422455956285, | |
| "learning_rate": 8.874076140766391e-06, | |
| "loss": 0.4322, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6563071297989032, | |
| "grad_norm": 0.2467206671663252, | |
| "learning_rate": 8.791308455556134e-06, | |
| "loss": 0.4647, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6581352833638026, | |
| "grad_norm": 0.20260204616971095, | |
| "learning_rate": 8.708768261119549e-06, | |
| "loss": 0.4034, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.659963436928702, | |
| "grad_norm": 0.21644768273281778, | |
| "learning_rate": 8.626458581786741e-06, | |
| "loss": 0.4254, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6617915904936015, | |
| "grad_norm": 0.23345228716575211, | |
| "learning_rate": 8.54438243344157e-06, | |
| "loss": 0.4298, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6636197440585009, | |
| "grad_norm": 0.23761761296603565, | |
| "learning_rate": 8.462542823411145e-06, | |
| "loss": 0.4656, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.6654478976234004, | |
| "grad_norm": 0.2024404905038723, | |
| "learning_rate": 8.380942750355678e-06, | |
| "loss": 0.4088, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6672760511882998, | |
| "grad_norm": 0.2250242044317574, | |
| "learning_rate": 8.29958520415856e-06, | |
| "loss": 0.4338, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6691042047531993, | |
| "grad_norm": 0.22826787810679472, | |
| "learning_rate": 8.218473165816832e-06, | |
| "loss": 0.4267, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.6709323583180987, | |
| "grad_norm": 0.23447452470751842, | |
| "learning_rate": 8.13760960733195e-06, | |
| "loss": 0.4292, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.6727605118829981, | |
| "grad_norm": 0.2340371506899419, | |
| "learning_rate": 8.056997491600925e-06, | |
| "loss": 0.4288, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6745886654478976, | |
| "grad_norm": 0.21118605658197215, | |
| "learning_rate": 7.976639772307686e-06, | |
| "loss": 0.4453, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.676416819012797, | |
| "grad_norm": 0.23569144130105643, | |
| "learning_rate": 7.896539393814926e-06, | |
| "loss": 0.4344, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6782449725776966, | |
| "grad_norm": 0.23524067199595933, | |
| "learning_rate": 7.816699291056196e-06, | |
| "loss": 0.4279, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.680073126142596, | |
| "grad_norm": 0.23054186478241442, | |
| "learning_rate": 7.737122389428354e-06, | |
| "loss": 0.4535, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6819012797074955, | |
| "grad_norm": 0.22158488037091248, | |
| "learning_rate": 7.657811604684385e-06, | |
| "loss": 0.4406, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6837294332723949, | |
| "grad_norm": 0.21854720152712095, | |
| "learning_rate": 7.578769842826572e-06, | |
| "loss": 0.4169, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.6855575868372943, | |
| "grad_norm": 0.24796845622790753, | |
| "learning_rate": 7.500000000000004e-06, | |
| "loss": 0.4397, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6873857404021938, | |
| "grad_norm": 0.21265082674010502, | |
| "learning_rate": 7.421504962386476e-06, | |
| "loss": 0.4197, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.6892138939670932, | |
| "grad_norm": 0.20579035323970915, | |
| "learning_rate": 7.343287606098719e-06, | |
| "loss": 0.4361, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.6910420475319927, | |
| "grad_norm": 0.26821523496520233, | |
| "learning_rate": 7.265350797075046e-06, | |
| "loss": 0.4271, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.6928702010968921, | |
| "grad_norm": 0.24850263668705072, | |
| "learning_rate": 7.187697390974307e-06, | |
| "loss": 0.4394, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.6946983546617916, | |
| "grad_norm": 0.2368443493234265, | |
| "learning_rate": 7.110330233071275e-06, | |
| "loss": 0.422, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.696526508226691, | |
| "grad_norm": 0.2373300944554724, | |
| "learning_rate": 7.03325215815239e-06, | |
| "loss": 0.4672, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.6983546617915904, | |
| "grad_norm": 0.23091696951892882, | |
| "learning_rate": 6.9564659904118925e-06, | |
| "loss": 0.4475, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.70018281535649, | |
| "grad_norm": 0.22821762753125135, | |
| "learning_rate": 6.879974543348344e-06, | |
| "loss": 0.4526, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7020109689213894, | |
| "grad_norm": 0.22270221536060336, | |
| "learning_rate": 6.803780619661527e-06, | |
| "loss": 0.4334, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.7038391224862889, | |
| "grad_norm": 0.21474970364310128, | |
| "learning_rate": 6.727887011149762e-06, | |
| "loss": 0.41, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7056672760511883, | |
| "grad_norm": 0.20072244573229295, | |
| "learning_rate": 6.652296498607628e-06, | |
| "loss": 0.4242, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7074954296160878, | |
| "grad_norm": 0.22140503089356514, | |
| "learning_rate": 6.577011851724044e-06, | |
| "loss": 0.428, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.7093235831809872, | |
| "grad_norm": 0.22728109637185054, | |
| "learning_rate": 6.502035828980786e-06, | |
| "loss": 0.443, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7111517367458866, | |
| "grad_norm": 0.22775217564080383, | |
| "learning_rate": 6.4273711775514545e-06, | |
| "loss": 0.426, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.7129798903107861, | |
| "grad_norm": 0.2169525200009275, | |
| "learning_rate": 6.353020633200773e-06, | |
| "loss": 0.4324, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7148080438756855, | |
| "grad_norm": 0.21065282217975448, | |
| "learning_rate": 6.278986920184362e-06, | |
| "loss": 0.4402, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.716636197440585, | |
| "grad_norm": 0.23089737840960195, | |
| "learning_rate": 6.2052727511489224e-06, | |
| "loss": 0.435, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.7184643510054844, | |
| "grad_norm": 0.20103080266684545, | |
| "learning_rate": 6.1318808270328595e-06, | |
| "loss": 0.4391, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.720292504570384, | |
| "grad_norm": 0.1819699544299102, | |
| "learning_rate": 6.058813836967268e-06, | |
| "loss": 0.4234, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7221206581352834, | |
| "grad_norm": 0.1970644497845393, | |
| "learning_rate": 5.986074458177454e-06, | |
| "loss": 0.4023, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7239488117001828, | |
| "grad_norm": 0.2207239084699447, | |
| "learning_rate": 5.913665355884823e-06, | |
| "loss": 0.4536, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7257769652650823, | |
| "grad_norm": 0.210996065737019, | |
| "learning_rate": 5.841589183209211e-06, | |
| "loss": 0.4188, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7276051188299817, | |
| "grad_norm": 0.22498686510864818, | |
| "learning_rate": 5.769848581071677e-06, | |
| "loss": 0.4381, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7294332723948812, | |
| "grad_norm": 0.21270916771515924, | |
| "learning_rate": 5.698446178097748e-06, | |
| "loss": 0.4156, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7312614259597806, | |
| "grad_norm": 0.22221640286229932, | |
| "learning_rate": 5.627384590521113e-06, | |
| "loss": 0.4395, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 547, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 263845278842880.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |