| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.96879875195008, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0998439937597504, | |
| "grad_norm": 0.4140388733942774, | |
| "learning_rate": 0.0, | |
| "loss": 1.9149, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1996879875195008, | |
| "grad_norm": 4.125320422125423, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0842, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.2995319812792512, | |
| "grad_norm": 4.114561571796027, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0845, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3993759750390016, | |
| "grad_norm": 7.814135082802521, | |
| "learning_rate": 9.94949494949495e-05, | |
| "loss": 2.2529, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.49921996879875197, | |
| "grad_norm": 4.959578007928057, | |
| "learning_rate": 9.8989898989899e-05, | |
| "loss": 2.1802, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.5990639625585024, | |
| "grad_norm": 4.020980884960991, | |
| "learning_rate": 9.848484848484849e-05, | |
| "loss": 2.0996, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.6989079563182528, | |
| "grad_norm": 2.913148956691379, | |
| "learning_rate": 9.797979797979798e-05, | |
| "loss": 2.0739, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7987519500780031, | |
| "grad_norm": 1.5904378694728827, | |
| "learning_rate": 9.747474747474747e-05, | |
| "loss": 2.0438, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.8985959438377535, | |
| "grad_norm": 1.1120971271683253, | |
| "learning_rate": 9.696969696969698e-05, | |
| "loss": 2.023, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9984399375975039, | |
| "grad_norm": 1.4857926676190432, | |
| "learning_rate": 9.646464646464647e-05, | |
| "loss": 2.0256, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0982839313572543, | |
| "grad_norm": 1.7514837650567026, | |
| "learning_rate": 9.595959595959596e-05, | |
| "loss": 2.0297, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.1981279251170047, | |
| "grad_norm": 1.3684063083659384, | |
| "learning_rate": 9.545454545454546e-05, | |
| "loss": 1.9984, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2979719188767551, | |
| "grad_norm": 0.8391319984906789, | |
| "learning_rate": 9.494949494949495e-05, | |
| "loss": 1.9938, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.3978159126365055, | |
| "grad_norm": 0.7680152250335479, | |
| "learning_rate": 9.444444444444444e-05, | |
| "loss": 1.9984, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.497659906396256, | |
| "grad_norm": 1.1427900590537006, | |
| "learning_rate": 9.393939393939395e-05, | |
| "loss": 2.0059, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.5975039001560063, | |
| "grad_norm": 1.2852588832884364, | |
| "learning_rate": 9.343434343434344e-05, | |
| "loss": 2.0057, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6973478939157567, | |
| "grad_norm": 0.8509981577726656, | |
| "learning_rate": 9.292929292929293e-05, | |
| "loss": 1.97, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.797191887675507, | |
| "grad_norm": 0.4374249257660765, | |
| "learning_rate": 9.242424242424242e-05, | |
| "loss": 1.9852, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8970358814352575, | |
| "grad_norm": 1.006945747433108, | |
| "learning_rate": 9.191919191919192e-05, | |
| "loss": 1.9736, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 1.9968798751950079, | |
| "grad_norm": 1.1714326150671521, | |
| "learning_rate": 9.141414141414141e-05, | |
| "loss": 1.9866, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0967238689547583, | |
| "grad_norm": 0.6697915843016325, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 1.9669, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.1965678627145087, | |
| "grad_norm": 0.43542954442572934, | |
| "learning_rate": 9.040404040404041e-05, | |
| "loss": 1.9596, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.296411856474259, | |
| "grad_norm": 0.8895989581186896, | |
| "learning_rate": 8.98989898989899e-05, | |
| "loss": 1.9777, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.3962558502340094, | |
| "grad_norm": 0.748473401890919, | |
| "learning_rate": 8.93939393939394e-05, | |
| "loss": 1.9828, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.49609984399376, | |
| "grad_norm": 0.4762840239068188, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 1.9863, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.5959438377535102, | |
| "grad_norm": 0.4634914120924797, | |
| "learning_rate": 8.83838383838384e-05, | |
| "loss": 1.9728, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.6957878315132606, | |
| "grad_norm": 0.576721007312459, | |
| "learning_rate": 8.787878787878789e-05, | |
| "loss": 1.9813, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.795631825273011, | |
| "grad_norm": 0.4717088615288276, | |
| "learning_rate": 8.737373737373738e-05, | |
| "loss": 1.9709, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.8954758190327614, | |
| "grad_norm": 0.5243076095653101, | |
| "learning_rate": 8.686868686868688e-05, | |
| "loss": 1.9889, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 2.995319812792512, | |
| "grad_norm": 0.35563844256479116, | |
| "learning_rate": 8.636363636363637e-05, | |
| "loss": 1.969, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.095163806552262, | |
| "grad_norm": 0.5040313506272054, | |
| "learning_rate": 8.585858585858586e-05, | |
| "loss": 1.9701, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.1950078003120126, | |
| "grad_norm": 0.5293887294443628, | |
| "learning_rate": 8.535353535353535e-05, | |
| "loss": 1.9774, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.294851794071763, | |
| "grad_norm": 0.33336016676106733, | |
| "learning_rate": 8.484848484848486e-05, | |
| "loss": 1.9702, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.3946957878315134, | |
| "grad_norm": 0.5156182664373749, | |
| "learning_rate": 8.434343434343435e-05, | |
| "loss": 1.9552, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.4945397815912638, | |
| "grad_norm": 0.410792592829029, | |
| "learning_rate": 8.383838383838384e-05, | |
| "loss": 1.9642, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.594383775351014, | |
| "grad_norm": 0.40267682408922495, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 1.9688, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.6942277691107646, | |
| "grad_norm": 0.3869359148412346, | |
| "learning_rate": 8.282828282828283e-05, | |
| "loss": 1.9733, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.794071762870515, | |
| "grad_norm": 0.37728712869432585, | |
| "learning_rate": 8.232323232323233e-05, | |
| "loss": 1.9648, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.8939157566302653, | |
| "grad_norm": 0.3922418131207954, | |
| "learning_rate": 8.181818181818183e-05, | |
| "loss": 1.9689, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 3.9937597503900157, | |
| "grad_norm": 0.26353046722639645, | |
| "learning_rate": 8.131313131313132e-05, | |
| "loss": 1.9727, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.093603744149766, | |
| "grad_norm": 0.3911091474488452, | |
| "learning_rate": 8.080808080808081e-05, | |
| "loss": 1.9631, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.1934477379095165, | |
| "grad_norm": 0.33402240826623614, | |
| "learning_rate": 8.03030303030303e-05, | |
| "loss": 1.9665, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.2932917316692665, | |
| "grad_norm": 0.34654808232868395, | |
| "learning_rate": 7.97979797979798e-05, | |
| "loss": 1.9646, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.393135725429017, | |
| "grad_norm": 0.3031078864703629, | |
| "learning_rate": 7.92929292929293e-05, | |
| "loss": 1.9693, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.492979719188767, | |
| "grad_norm": 0.35342072957234116, | |
| "learning_rate": 7.878787878787879e-05, | |
| "loss": 1.9688, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.592823712948518, | |
| "grad_norm": 0.3918161921811716, | |
| "learning_rate": 7.828282828282829e-05, | |
| "loss": 1.9609, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.692667706708268, | |
| "grad_norm": 0.24995683506017796, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 1.9515, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.792511700468019, | |
| "grad_norm": 0.3308078104166398, | |
| "learning_rate": 7.727272727272727e-05, | |
| "loss": 1.9607, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.892355694227769, | |
| "grad_norm": 0.3130926472973521, | |
| "learning_rate": 7.676767676767676e-05, | |
| "loss": 1.9699, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 4.99219968798752, | |
| "grad_norm": 0.30892356920484393, | |
| "learning_rate": 7.626262626262627e-05, | |
| "loss": 1.9645, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.09204368174727, | |
| "grad_norm": 0.2804202715276883, | |
| "learning_rate": 7.575757575757576e-05, | |
| "loss": 1.9569, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.1918876755070205, | |
| "grad_norm": 0.2789049399636327, | |
| "learning_rate": 7.525252525252525e-05, | |
| "loss": 1.9585, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.29173166926677, | |
| "grad_norm": 0.2906929505804403, | |
| "learning_rate": 7.474747474747475e-05, | |
| "loss": 1.9565, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.391575663026521, | |
| "grad_norm": 0.2033727950080347, | |
| "learning_rate": 7.424242424242424e-05, | |
| "loss": 1.9755, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.491419656786271, | |
| "grad_norm": 0.31364461369416163, | |
| "learning_rate": 7.373737373737373e-05, | |
| "loss": 1.9647, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.591263650546022, | |
| "grad_norm": 0.2531087381531638, | |
| "learning_rate": 7.323232323232324e-05, | |
| "loss": 1.9578, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.691107644305772, | |
| "grad_norm": 0.23764498764830225, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 1.9617, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.790951638065523, | |
| "grad_norm": 0.24888591334854687, | |
| "learning_rate": 7.222222222222222e-05, | |
| "loss": 1.963, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.890795631825273, | |
| "grad_norm": 0.2647075657405339, | |
| "learning_rate": 7.171717171717171e-05, | |
| "loss": 1.9685, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 5.990639625585024, | |
| "grad_norm": 0.27820470985704615, | |
| "learning_rate": 7.121212121212121e-05, | |
| "loss": 1.9654, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.090483619344774, | |
| "grad_norm": 0.20068946885468097, | |
| "learning_rate": 7.07070707070707e-05, | |
| "loss": 1.9667, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.190327613104524, | |
| "grad_norm": 0.25026234630326394, | |
| "learning_rate": 7.020202020202021e-05, | |
| "loss": 1.9542, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.290171606864274, | |
| "grad_norm": 0.22856925269883635, | |
| "learning_rate": 6.96969696969697e-05, | |
| "loss": 1.9573, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.390015600624025, | |
| "grad_norm": 0.2392183076591563, | |
| "learning_rate": 6.91919191919192e-05, | |
| "loss": 1.9647, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.489859594383775, | |
| "grad_norm": 0.20384525102843132, | |
| "learning_rate": 6.86868686868687e-05, | |
| "loss": 1.9628, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.589703588143526, | |
| "grad_norm": 0.23941897200051984, | |
| "learning_rate": 6.818181818181818e-05, | |
| "loss": 1.9667, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.689547581903276, | |
| "grad_norm": 0.20375278551444306, | |
| "learning_rate": 6.767676767676769e-05, | |
| "loss": 1.9572, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.789391575663027, | |
| "grad_norm": 0.20727005267599333, | |
| "learning_rate": 6.717171717171718e-05, | |
| "loss": 1.9581, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.889235569422777, | |
| "grad_norm": 0.22300809533132504, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.9693, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 6.9890795631825275, | |
| "grad_norm": 0.21742318730398613, | |
| "learning_rate": 6.616161616161617e-05, | |
| "loss": 1.9656, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0889235569422775, | |
| "grad_norm": 0.20343822391223, | |
| "learning_rate": 6.565656565656566e-05, | |
| "loss": 1.9656, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.188767550702028, | |
| "grad_norm": 0.2364200066637671, | |
| "learning_rate": 6.515151515151516e-05, | |
| "loss": 1.95, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.288611544461778, | |
| "grad_norm": 0.18261048615751524, | |
| "learning_rate": 6.464646464646466e-05, | |
| "loss": 1.9538, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.388455538221529, | |
| "grad_norm": 0.24533487813163474, | |
| "learning_rate": 6.414141414141415e-05, | |
| "loss": 1.952, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.488299531981279, | |
| "grad_norm": 0.2539612930496735, | |
| "learning_rate": 6.363636363636364e-05, | |
| "loss": 1.9598, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.58814352574103, | |
| "grad_norm": 0.2991457603613546, | |
| "learning_rate": 6.313131313131313e-05, | |
| "loss": 1.9733, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.68798751950078, | |
| "grad_norm": 0.2209105824741669, | |
| "learning_rate": 6.262626262626264e-05, | |
| "loss": 1.9531, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.787831513260531, | |
| "grad_norm": 0.31698310918964695, | |
| "learning_rate": 6.212121212121213e-05, | |
| "loss": 1.964, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.887675507020281, | |
| "grad_norm": 0.17584276182725114, | |
| "learning_rate": 6.161616161616162e-05, | |
| "loss": 1.9644, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 7.9875195007800315, | |
| "grad_norm": 0.29996919622824225, | |
| "learning_rate": 6.111111111111112e-05, | |
| "loss": 1.9555, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.087363494539781, | |
| "grad_norm": 0.2406502202676367, | |
| "learning_rate": 6.060606060606061e-05, | |
| "loss": 1.9554, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.187207488299531, | |
| "grad_norm": 0.32705142732170855, | |
| "learning_rate": 6.01010101010101e-05, | |
| "loss": 1.9517, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.287051482059283, | |
| "grad_norm": 0.27249925952338305, | |
| "learning_rate": 5.959595959595959e-05, | |
| "loss": 1.9474, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.386895475819033, | |
| "grad_norm": 0.29448831027669287, | |
| "learning_rate": 5.90909090909091e-05, | |
| "loss": 1.9459, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.486739469578783, | |
| "grad_norm": 0.29998154037028857, | |
| "learning_rate": 5.858585858585859e-05, | |
| "loss": 1.9606, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.586583463338533, | |
| "grad_norm": 0.23153724936859055, | |
| "learning_rate": 5.808080808080808e-05, | |
| "loss": 1.9598, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.686427457098285, | |
| "grad_norm": 0.22081595887056477, | |
| "learning_rate": 5.757575757575758e-05, | |
| "loss": 1.9586, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.786271450858035, | |
| "grad_norm": 0.19177670537863922, | |
| "learning_rate": 5.707070707070707e-05, | |
| "loss": 1.9715, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.886115444617785, | |
| "grad_norm": 0.25725928107907137, | |
| "learning_rate": 5.6565656565656563e-05, | |
| "loss": 1.9602, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 8.985959438377535, | |
| "grad_norm": 0.26044371305524344, | |
| "learning_rate": 5.606060606060606e-05, | |
| "loss": 1.9607, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.085803432137286, | |
| "grad_norm": 0.23728151561491595, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 1.9588, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.185647425897036, | |
| "grad_norm": 0.20354348868729488, | |
| "learning_rate": 5.5050505050505056e-05, | |
| "loss": 1.9492, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.285491419656786, | |
| "grad_norm": 0.18672087839741056, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 1.9457, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.385335413416536, | |
| "grad_norm": 0.1939858201242329, | |
| "learning_rate": 5.4040404040404044e-05, | |
| "loss": 1.9453, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.485179407176288, | |
| "grad_norm": 0.19172060706771135, | |
| "learning_rate": 5.353535353535354e-05, | |
| "loss": 1.958, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.585023400936038, | |
| "grad_norm": 0.1837920882880991, | |
| "learning_rate": 5.303030303030303e-05, | |
| "loss": 1.9577, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.684867394695788, | |
| "grad_norm": 0.2162949878555464, | |
| "learning_rate": 5.2525252525252536e-05, | |
| "loss": 1.9622, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.784711388455538, | |
| "grad_norm": 0.19325381586186333, | |
| "learning_rate": 5.2020202020202026e-05, | |
| "loss": 1.9433, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.88455538221529, | |
| "grad_norm": 0.2018142831658023, | |
| "learning_rate": 5.151515151515152e-05, | |
| "loss": 1.9605, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.98439937597504, | |
| "grad_norm": 0.176671565601027, | |
| "learning_rate": 5.101010101010101e-05, | |
| "loss": 1.9578, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.08424336973479, | |
| "grad_norm": 0.2117788085352089, | |
| "learning_rate": 5.050505050505051e-05, | |
| "loss": 1.9478, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.18408736349454, | |
| "grad_norm": 0.1816135304249716, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9423, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.283931357254291, | |
| "grad_norm": 0.2680310363226074, | |
| "learning_rate": 4.94949494949495e-05, | |
| "loss": 1.9519, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.383775351014041, | |
| "grad_norm": 0.17934299698555412, | |
| "learning_rate": 4.898989898989899e-05, | |
| "loss": 1.9625, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.48361934477379, | |
| "grad_norm": 0.19786074542682824, | |
| "learning_rate": 4.848484848484849e-05, | |
| "loss": 1.95, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.58346333853354, | |
| "grad_norm": 0.17490489580858018, | |
| "learning_rate": 4.797979797979798e-05, | |
| "loss": 1.9513, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.683307332293293, | |
| "grad_norm": 0.224513887757472, | |
| "learning_rate": 4.7474747474747476e-05, | |
| "loss": 1.9499, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.783151326053042, | |
| "grad_norm": 0.16993980203530532, | |
| "learning_rate": 4.696969696969697e-05, | |
| "loss": 1.944, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.882995319812792, | |
| "grad_norm": 0.18436224376063975, | |
| "learning_rate": 4.6464646464646464e-05, | |
| "loss": 1.9494, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 10.982839313572542, | |
| "grad_norm": 0.1858801208244774, | |
| "learning_rate": 4.595959595959596e-05, | |
| "loss": 1.9504, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.082683307332294, | |
| "grad_norm": 0.21397140437157122, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 1.9467, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.182527301092044, | |
| "grad_norm": 0.16934479906947233, | |
| "learning_rate": 4.494949494949495e-05, | |
| "loss": 1.9475, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.282371294851794, | |
| "grad_norm": 0.17710883760980722, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.9418, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.382215288611544, | |
| "grad_norm": 0.2278025006688675, | |
| "learning_rate": 4.3939393939393944e-05, | |
| "loss": 1.9456, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.482059282371296, | |
| "grad_norm": 0.18727166458531316, | |
| "learning_rate": 4.343434343434344e-05, | |
| "loss": 1.9408, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.581903276131046, | |
| "grad_norm": 0.17348080665741175, | |
| "learning_rate": 4.292929292929293e-05, | |
| "loss": 1.9367, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.681747269890796, | |
| "grad_norm": 0.21559975863343248, | |
| "learning_rate": 4.242424242424243e-05, | |
| "loss": 1.9509, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.781591263650546, | |
| "grad_norm": 0.20515384184563593, | |
| "learning_rate": 4.191919191919192e-05, | |
| "loss": 1.9503, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.881435257410295, | |
| "grad_norm": 0.17579996751101729, | |
| "learning_rate": 4.141414141414142e-05, | |
| "loss": 1.9443, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 11.981279251170047, | |
| "grad_norm": 0.1870399234707776, | |
| "learning_rate": 4.0909090909090915e-05, | |
| "loss": 1.9507, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.081123244929797, | |
| "grad_norm": 0.2323975590399996, | |
| "learning_rate": 4.0404040404040405e-05, | |
| "loss": 1.9486, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.180967238689547, | |
| "grad_norm": 0.17332911391024705, | |
| "learning_rate": 3.98989898989899e-05, | |
| "loss": 1.9441, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.280811232449299, | |
| "grad_norm": 0.23886491083540215, | |
| "learning_rate": 3.939393939393939e-05, | |
| "loss": 1.9489, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.380655226209049, | |
| "grad_norm": 0.192192583869745, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 1.936, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.480499219968799, | |
| "grad_norm": 0.24070020033146947, | |
| "learning_rate": 3.838383838383838e-05, | |
| "loss": 1.9363, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.580343213728549, | |
| "grad_norm": 0.17061145664967614, | |
| "learning_rate": 3.787878787878788e-05, | |
| "loss": 1.947, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.680187207488299, | |
| "grad_norm": 0.20420044689274344, | |
| "learning_rate": 3.7373737373737376e-05, | |
| "loss": 1.9462, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.78003120124805, | |
| "grad_norm": 0.16640664781155742, | |
| "learning_rate": 3.686868686868687e-05, | |
| "loss": 1.9404, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.8798751950078, | |
| "grad_norm": 0.17534875646136103, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.9441, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 12.97971918876755, | |
| "grad_norm": 0.1881647742956635, | |
| "learning_rate": 3.5858585858585855e-05, | |
| "loss": 1.9452, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0795631825273, | |
| "grad_norm": 0.21130090774448568, | |
| "learning_rate": 3.535353535353535e-05, | |
| "loss": 1.938, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 13.179407176287052, | |
| "grad_norm": 0.19012207225624486, | |
| "learning_rate": 3.484848484848485e-05, | |
| "loss": 1.93, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 13.279251170046802, | |
| "grad_norm": 0.19535583015453165, | |
| "learning_rate": 3.434343434343435e-05, | |
| "loss": 1.9423, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 13.379095163806552, | |
| "grad_norm": 0.1972934873185412, | |
| "learning_rate": 3.3838383838383844e-05, | |
| "loss": 1.9449, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 13.478939157566302, | |
| "grad_norm": 0.21172258190614646, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.9423, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 13.578783151326054, | |
| "grad_norm": 0.20243808248600392, | |
| "learning_rate": 3.282828282828283e-05, | |
| "loss": 1.9454, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 13.678627145085803, | |
| "grad_norm": 0.29468220957824104, | |
| "learning_rate": 3.232323232323233e-05, | |
| "loss": 1.9329, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 13.778471138845553, | |
| "grad_norm": 0.1852836649334086, | |
| "learning_rate": 3.181818181818182e-05, | |
| "loss": 1.9397, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 13.878315132605305, | |
| "grad_norm": 0.17635021846243693, | |
| "learning_rate": 3.131313131313132e-05, | |
| "loss": 1.9414, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 13.978159126365055, | |
| "grad_norm": 0.1837620268343685, | |
| "learning_rate": 3.080808080808081e-05, | |
| "loss": 1.9265, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 14.078003120124805, | |
| "grad_norm": 0.1851416429157977, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 1.938, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 14.177847113884555, | |
| "grad_norm": 0.18177436704033564, | |
| "learning_rate": 2.9797979797979796e-05, | |
| "loss": 1.9338, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 14.277691107644305, | |
| "grad_norm": 0.20249599488147646, | |
| "learning_rate": 2.9292929292929294e-05, | |
| "loss": 1.943, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 14.377535101404057, | |
| "grad_norm": 0.1914943764672633, | |
| "learning_rate": 2.878787878787879e-05, | |
| "loss": 1.9381, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 14.477379095163807, | |
| "grad_norm": 0.18144339446500468, | |
| "learning_rate": 2.8282828282828282e-05, | |
| "loss": 1.9493, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 14.577223088923557, | |
| "grad_norm": 0.22871591479507436, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 1.9394, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 14.677067082683307, | |
| "grad_norm": 0.2409736531836878, | |
| "learning_rate": 2.7272727272727273e-05, | |
| "loss": 1.9363, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 14.776911076443058, | |
| "grad_norm": 0.21702411701682794, | |
| "learning_rate": 2.676767676767677e-05, | |
| "loss": 1.9324, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 14.876755070202808, | |
| "grad_norm": 0.186963824720383, | |
| "learning_rate": 2.6262626262626268e-05, | |
| "loss": 1.9254, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 14.976599063962558, | |
| "grad_norm": 0.20551876684974787, | |
| "learning_rate": 2.575757575757576e-05, | |
| "loss": 1.9385, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.076443057722308, | |
| "grad_norm": 0.17794734645273458, | |
| "learning_rate": 2.5252525252525256e-05, | |
| "loss": 1.935, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 15.17628705148206, | |
| "grad_norm": 0.19787955354426204, | |
| "learning_rate": 2.474747474747475e-05, | |
| "loss": 1.9286, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 15.27613104524181, | |
| "grad_norm": 0.21663975391838738, | |
| "learning_rate": 2.4242424242424244e-05, | |
| "loss": 1.9274, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 15.37597503900156, | |
| "grad_norm": 0.19056508068402894, | |
| "learning_rate": 2.3737373737373738e-05, | |
| "loss": 1.9328, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 15.47581903276131, | |
| "grad_norm": 0.20643529046597323, | |
| "learning_rate": 2.3232323232323232e-05, | |
| "loss": 1.9374, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 15.575663026521061, | |
| "grad_norm": 0.17428582721990332, | |
| "learning_rate": 2.272727272727273e-05, | |
| "loss": 1.9435, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 15.675507020280811, | |
| "grad_norm": 0.17915807350384474, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 1.9342, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 15.775351014040561, | |
| "grad_norm": 0.17934386940217817, | |
| "learning_rate": 2.171717171717172e-05, | |
| "loss": 1.9252, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 15.875195007800311, | |
| "grad_norm": 0.16971494417624172, | |
| "learning_rate": 2.1212121212121215e-05, | |
| "loss": 1.9333, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 15.975039001560063, | |
| "grad_norm": 0.1710725442382166, | |
| "learning_rate": 2.070707070707071e-05, | |
| "loss": 1.9397, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 16.07488299531981, | |
| "grad_norm": 0.16048331708079347, | |
| "learning_rate": 2.0202020202020203e-05, | |
| "loss": 1.9354, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 16.174726989079563, | |
| "grad_norm": 0.2209212482793433, | |
| "learning_rate": 1.9696969696969697e-05, | |
| "loss": 1.9572, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 16.274570982839315, | |
| "grad_norm": 0.17292517371584637, | |
| "learning_rate": 1.919191919191919e-05, | |
| "loss": 1.9384, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 16.374414976599063, | |
| "grad_norm": 0.1756696399704993, | |
| "learning_rate": 1.8686868686868688e-05, | |
| "loss": 1.9287, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 16.474258970358814, | |
| "grad_norm": 0.193814973934712, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 1.9285, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 16.574102964118566, | |
| "grad_norm": 0.21108116449806094, | |
| "learning_rate": 1.7676767676767676e-05, | |
| "loss": 1.9249, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 16.673946957878314, | |
| "grad_norm": 0.164152325154632, | |
| "learning_rate": 1.7171717171717173e-05, | |
| "loss": 1.9335, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 16.773790951638066, | |
| "grad_norm": 0.1934976757474289, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.9344, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 16.873634945397814, | |
| "grad_norm": 0.17861559674997443, | |
| "learning_rate": 1.6161616161616165e-05, | |
| "loss": 1.9315, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 16.973478939157566, | |
| "grad_norm": 0.16812713496720977, | |
| "learning_rate": 1.565656565656566e-05, | |
| "loss": 1.9278, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 17.073322932917318, | |
| "grad_norm": 0.19243202935397094, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 1.939, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 17.173166926677066, | |
| "grad_norm": 0.16546322416204856, | |
| "learning_rate": 1.4646464646464647e-05, | |
| "loss": 1.9295, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 17.273010920436818, | |
| "grad_norm": 0.19615095413628908, | |
| "learning_rate": 1.4141414141414141e-05, | |
| "loss": 1.9357, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 17.37285491419657, | |
| "grad_norm": 0.16562858231287156, | |
| "learning_rate": 1.3636363636363637e-05, | |
| "loss": 1.9372, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 17.472698907956318, | |
| "grad_norm": 0.1755423564949021, | |
| "learning_rate": 1.3131313131313134e-05, | |
| "loss": 1.9208, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 17.57254290171607, | |
| "grad_norm": 0.16572591523274388, | |
| "learning_rate": 1.2626262626262628e-05, | |
| "loss": 1.9196, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 17.672386895475817, | |
| "grad_norm": 0.16066050812369387, | |
| "learning_rate": 1.2121212121212122e-05, | |
| "loss": 1.9379, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 17.77223088923557, | |
| "grad_norm": 0.18230307180057742, | |
| "learning_rate": 1.1616161616161616e-05, | |
| "loss": 1.9344, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 17.87207488299532, | |
| "grad_norm": 0.16147840026521357, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 1.9249, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 17.97191887675507, | |
| "grad_norm": 0.17234298543336798, | |
| "learning_rate": 1.0606060606060607e-05, | |
| "loss": 1.9341, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 18.07176287051482, | |
| "grad_norm": 0.16952419332241464, | |
| "learning_rate": 1.0101010101010101e-05, | |
| "loss": 1.9382, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 18.171606864274573, | |
| "grad_norm": 0.17503197241676455, | |
| "learning_rate": 9.595959595959595e-06, | |
| "loss": 1.9277, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 18.27145085803432, | |
| "grad_norm": 0.16018657280969506, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 1.9259, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 18.371294851794072, | |
| "grad_norm": 0.16577134954028483, | |
| "learning_rate": 8.585858585858587e-06, | |
| "loss": 1.9391, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 18.47113884555382, | |
| "grad_norm": 0.1758462044127833, | |
| "learning_rate": 8.080808080808082e-06, | |
| "loss": 1.9316, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 18.570982839313572, | |
| "grad_norm": 0.16928715932805172, | |
| "learning_rate": 7.5757575757575764e-06, | |
| "loss": 1.9218, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 18.670826833073324, | |
| "grad_norm": 0.16185874983512785, | |
| "learning_rate": 7.0707070707070704e-06, | |
| "loss": 1.9244, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 18.770670826833072, | |
| "grad_norm": 0.16445906712178507, | |
| "learning_rate": 6.565656565656567e-06, | |
| "loss": 1.9425, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 18.870514820592824, | |
| "grad_norm": 0.16313460189322437, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 1.9336, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 18.970358814352576, | |
| "grad_norm": 0.15990081630753986, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 1.9178, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 19.070202808112324, | |
| "grad_norm": 0.16547636636850527, | |
| "learning_rate": 5.050505050505051e-06, | |
| "loss": 1.9281, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 19.170046801872076, | |
| "grad_norm": 0.1625270231867559, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 1.9348, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 19.269890795631824, | |
| "grad_norm": 0.16385675767663568, | |
| "learning_rate": 4.040404040404041e-06, | |
| "loss": 1.9305, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 19.369734789391575, | |
| "grad_norm": 0.16718542619114216, | |
| "learning_rate": 3.5353535353535352e-06, | |
| "loss": 1.9376, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 19.469578783151327, | |
| "grad_norm": 0.16595125072244407, | |
| "learning_rate": 3.0303030303030305e-06, | |
| "loss": 1.9264, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 19.569422776911075, | |
| "grad_norm": 0.16912445317015737, | |
| "learning_rate": 2.5252525252525253e-06, | |
| "loss": 1.9252, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 19.669266770670827, | |
| "grad_norm": 0.15257787442711698, | |
| "learning_rate": 2.0202020202020206e-06, | |
| "loss": 1.9312, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 19.76911076443058, | |
| "grad_norm": 0.17270934725449602, | |
| "learning_rate": 1.5151515151515152e-06, | |
| "loss": 1.9241, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 19.868954758190327, | |
| "grad_norm": 0.16771403116909167, | |
| "learning_rate": 1.0101010101010103e-06, | |
| "loss": 1.9342, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 19.96879875195008, | |
| "grad_norm": 0.17132458008674775, | |
| "learning_rate": 5.050505050505052e-07, | |
| "loss": 1.9347, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |