| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 288, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006944444444444444, | |
| "grad_norm": 0.0634646986292456, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.1218, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013888888888888888, | |
| "grad_norm": 0.0859646592063227, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 0.1441, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020833333333333332, | |
| "grad_norm": 0.08311493152111031, | |
| "learning_rate": 1.0344827586206897e-05, | |
| "loss": 0.1517, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 0.14507101606184478, | |
| "learning_rate": 1.3793103448275863e-05, | |
| "loss": 0.184, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.034722222222222224, | |
| "grad_norm": 0.049880571226207794, | |
| "learning_rate": 1.7241379310344828e-05, | |
| "loss": 0.0937, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.041666666666666664, | |
| "grad_norm": 0.13472688960866697, | |
| "learning_rate": 2.0689655172413793e-05, | |
| "loss": 0.1932, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04861111111111111, | |
| "grad_norm": 0.063128720217521, | |
| "learning_rate": 2.413793103448276e-05, | |
| "loss": 0.1163, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 0.05791338456495497, | |
| "learning_rate": 2.7586206896551727e-05, | |
| "loss": 0.0997, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 0.03814512155396182, | |
| "learning_rate": 3.103448275862069e-05, | |
| "loss": 0.0691, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06944444444444445, | |
| "grad_norm": 0.07278628165772846, | |
| "learning_rate": 3.4482758620689657e-05, | |
| "loss": 0.1096, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0763888888888889, | |
| "grad_norm": 0.06072973165332569, | |
| "learning_rate": 3.793103448275862e-05, | |
| "loss": 0.0915, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 0.05086161703424197, | |
| "learning_rate": 4.1379310344827587e-05, | |
| "loss": 0.0758, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09027777777777778, | |
| "grad_norm": 0.06108424147364289, | |
| "learning_rate": 4.482758620689655e-05, | |
| "loss": 0.085, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09722222222222222, | |
| "grad_norm": 0.08171893276043111, | |
| "learning_rate": 4.827586206896552e-05, | |
| "loss": 0.1039, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "grad_norm": 0.10806188024280473, | |
| "learning_rate": 5.172413793103449e-05, | |
| "loss": 0.1153, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.08264842267025586, | |
| "learning_rate": 5.517241379310345e-05, | |
| "loss": 0.096, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11805555555555555, | |
| "grad_norm": 0.11707563126376766, | |
| "learning_rate": 5.862068965517241e-05, | |
| "loss": 0.1187, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.04624996354545763, | |
| "learning_rate": 6.206896551724138e-05, | |
| "loss": 0.0613, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13194444444444445, | |
| "grad_norm": 0.12096307356119897, | |
| "learning_rate": 6.551724137931034e-05, | |
| "loss": 0.112, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 0.11329767467249009, | |
| "learning_rate": 6.896551724137931e-05, | |
| "loss": 0.1131, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14583333333333334, | |
| "grad_norm": 0.07076834653741017, | |
| "learning_rate": 7.241379310344828e-05, | |
| "loss": 0.0754, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1527777777777778, | |
| "grad_norm": 0.07631833003946596, | |
| "learning_rate": 7.586206896551724e-05, | |
| "loss": 0.0476, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1597222222222222, | |
| "grad_norm": 0.09111429978286256, | |
| "learning_rate": 7.931034482758621e-05, | |
| "loss": 0.0778, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 0.09638107817903875, | |
| "learning_rate": 8.275862068965517e-05, | |
| "loss": 0.0749, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1736111111111111, | |
| "grad_norm": 0.08873808756623787, | |
| "learning_rate": 8.620689655172413e-05, | |
| "loss": 0.0692, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.18055555555555555, | |
| "grad_norm": 0.1193395529072999, | |
| "learning_rate": 8.96551724137931e-05, | |
| "loss": 0.1005, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 0.07825819960868853, | |
| "learning_rate": 9.310344827586207e-05, | |
| "loss": 0.0796, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 0.09882721733405724, | |
| "learning_rate": 9.655172413793105e-05, | |
| "loss": 0.0808, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2013888888888889, | |
| "grad_norm": 0.07496690540223322, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0589, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 0.06753626384871714, | |
| "learning_rate": 9.999632180371776e-05, | |
| "loss": 0.0885, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2152777777777778, | |
| "grad_norm": 0.11227962647329344, | |
| "learning_rate": 9.998528775603611e-05, | |
| "loss": 0.0859, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.04228956053702444, | |
| "learning_rate": 9.99668994803708e-05, | |
| "loss": 0.0482, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.22916666666666666, | |
| "grad_norm": 0.049938706422113184, | |
| "learning_rate": 9.994115968214932e-05, | |
| "loss": 0.0575, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2361111111111111, | |
| "grad_norm": 0.054468426952476535, | |
| "learning_rate": 9.990807214841287e-05, | |
| "loss": 0.0661, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24305555555555555, | |
| "grad_norm": 0.048987738819091665, | |
| "learning_rate": 9.986764174725919e-05, | |
| "loss": 0.0556, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.05029249306240408, | |
| "learning_rate": 9.981987442712633e-05, | |
| "loss": 0.0546, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2569444444444444, | |
| "grad_norm": 0.07920414870467878, | |
| "learning_rate": 9.976477721591745e-05, | |
| "loss": 0.0704, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2638888888888889, | |
| "grad_norm": 0.04856426595651698, | |
| "learning_rate": 9.97023582199669e-05, | |
| "loss": 0.0627, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2708333333333333, | |
| "grad_norm": 0.03338189886854429, | |
| "learning_rate": 9.963262662284736e-05, | |
| "loss": 0.0353, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 0.05814625123252708, | |
| "learning_rate": 9.955559268401893e-05, | |
| "loss": 0.0562, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2847222222222222, | |
| "grad_norm": 0.09810680657571984, | |
| "learning_rate": 9.947126773731948e-05, | |
| "loss": 0.0774, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2916666666666667, | |
| "grad_norm": 0.08137711658725671, | |
| "learning_rate": 9.937966418929726e-05, | |
| "loss": 0.0688, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2986111111111111, | |
| "grad_norm": 0.04984221097640415, | |
| "learning_rate": 9.928079551738543e-05, | |
| "loss": 0.0631, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 0.045586138765645726, | |
| "learning_rate": 9.917467626791925e-05, | |
| "loss": 0.0602, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 0.04555416039001976, | |
| "learning_rate": 9.90613220539959e-05, | |
| "loss": 0.0603, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3194444444444444, | |
| "grad_norm": 0.06929704690081513, | |
| "learning_rate": 9.89407495531773e-05, | |
| "loss": 0.05, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3263888888888889, | |
| "grad_norm": 0.05071419561253368, | |
| "learning_rate": 9.881297650503641e-05, | |
| "loss": 0.0666, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.05861571605506496, | |
| "learning_rate": 9.867802170854724e-05, | |
| "loss": 0.0576, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3402777777777778, | |
| "grad_norm": 0.05602477728042172, | |
| "learning_rate": 9.853590501931904e-05, | |
| "loss": 0.0714, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3472222222222222, | |
| "grad_norm": 0.0600399573347184, | |
| "learning_rate": 9.838664734667495e-05, | |
| "loss": 0.0557, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3541666666666667, | |
| "grad_norm": 0.04310502477291791, | |
| "learning_rate": 9.82302706505756e-05, | |
| "loss": 0.0608, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 0.04208369800984692, | |
| "learning_rate": 9.806679793838829e-05, | |
| "loss": 0.0457, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3680555555555556, | |
| "grad_norm": 0.04390617615736525, | |
| "learning_rate": 9.78962532615019e-05, | |
| "loss": 0.0538, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.040829606477323165, | |
| "learning_rate": 9.771866171178831e-05, | |
| "loss": 0.0535, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3819444444444444, | |
| "grad_norm": 0.04189168396894191, | |
| "learning_rate": 9.753404941791062e-05, | |
| "loss": 0.0533, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 0.048795880713950655, | |
| "learning_rate": 9.734244354147895e-05, | |
| "loss": 0.0558, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3958333333333333, | |
| "grad_norm": 0.04594331706981098, | |
| "learning_rate": 9.714387227305422e-05, | |
| "loss": 0.059, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.4027777777777778, | |
| "grad_norm": 0.059333907325760896, | |
| "learning_rate": 9.693836482800044e-05, | |
| "loss": 0.0666, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4097222222222222, | |
| "grad_norm": 0.04504422858266263, | |
| "learning_rate": 9.672595144218646e-05, | |
| "loss": 0.0553, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 0.06138639267675264, | |
| "learning_rate": 9.650666336753728e-05, | |
| "loss": 0.0736, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4236111111111111, | |
| "grad_norm": 0.054120807804378494, | |
| "learning_rate": 9.628053286743619e-05, | |
| "loss": 0.0658, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4305555555555556, | |
| "grad_norm": 0.05215983401668381, | |
| "learning_rate": 9.604759321197773e-05, | |
| "loss": 0.0597, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 0.05923618865174348, | |
| "learning_rate": 9.580787867307293e-05, | |
| "loss": 0.0661, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.06697666067838456, | |
| "learning_rate": 9.55614245194068e-05, | |
| "loss": 0.0684, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4513888888888889, | |
| "grad_norm": 0.04541593572547064, | |
| "learning_rate": 9.530826701124939e-05, | |
| "loss": 0.0547, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4583333333333333, | |
| "grad_norm": 0.05700134510715612, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.0596, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4652777777777778, | |
| "grad_norm": 0.041366318222701697, | |
| "learning_rate": 9.478199189831183e-05, | |
| "loss": 0.0411, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 0.043353520617435734, | |
| "learning_rate": 9.450895172325822e-05, | |
| "loss": 0.0548, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4791666666666667, | |
| "grad_norm": 0.05496280723006238, | |
| "learning_rate": 9.422936304177439e-05, | |
| "loss": 0.0531, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4861111111111111, | |
| "grad_norm": 0.057292146852494466, | |
| "learning_rate": 9.39432669891423e-05, | |
| "loss": 0.0658, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4930555555555556, | |
| "grad_norm": 0.044098969405188045, | |
| "learning_rate": 9.365070565805941e-05, | |
| "loss": 0.0511, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.10907567767546891, | |
| "learning_rate": 9.335172209244575e-05, | |
| "loss": 0.0547, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5069444444444444, | |
| "grad_norm": 0.05193284317496665, | |
| "learning_rate": 9.304636028111094e-05, | |
| "loss": 0.0674, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5138888888888888, | |
| "grad_norm": 0.06799161021928798, | |
| "learning_rate": 9.273466515128209e-05, | |
| "loss": 0.067, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "grad_norm": 0.05626247552701963, | |
| "learning_rate": 9.241668256199392e-05, | |
| "loss": 0.0501, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5277777777777778, | |
| "grad_norm": 0.04578630890465305, | |
| "learning_rate": 9.209245929734156e-05, | |
| "loss": 0.0509, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5347222222222222, | |
| "grad_norm": 0.0451951125591856, | |
| "learning_rate": 9.176204305959726e-05, | |
| "loss": 0.0465, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5416666666666666, | |
| "grad_norm": 0.04052385478349237, | |
| "learning_rate": 9.142548246219212e-05, | |
| "loss": 0.0454, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5486111111111112, | |
| "grad_norm": 0.04695475481560532, | |
| "learning_rate": 9.108282702256365e-05, | |
| "loss": 0.0598, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.05974716728603661, | |
| "learning_rate": 9.073412715487044e-05, | |
| "loss": 0.0632, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.04594564477538813, | |
| "learning_rate": 9.037943416257474e-05, | |
| "loss": 0.059, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5694444444444444, | |
| "grad_norm": 0.04581771185786344, | |
| "learning_rate": 9.001880023089441e-05, | |
| "loss": 0.0578, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5763888888888888, | |
| "grad_norm": 0.03808060898920002, | |
| "learning_rate": 8.965227841912489e-05, | |
| "loss": 0.0435, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 0.06582783981663642, | |
| "learning_rate": 8.927992265283282e-05, | |
| "loss": 0.058, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5902777777777778, | |
| "grad_norm": 0.0409049374684989, | |
| "learning_rate": 8.890178771592199e-05, | |
| "loss": 0.045, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5972222222222222, | |
| "grad_norm": 0.056258831948584206, | |
| "learning_rate": 8.851792924257317e-05, | |
| "loss": 0.0709, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6041666666666666, | |
| "grad_norm": 0.04580764978743741, | |
| "learning_rate": 8.812840370905873e-05, | |
| "loss": 0.0525, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 0.043263482428543644, | |
| "learning_rate": 8.773326842543347e-05, | |
| "loss": 0.0464, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6180555555555556, | |
| "grad_norm": 0.059777623657141676, | |
| "learning_rate": 8.733258152710262e-05, | |
| "loss": 0.0649, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.049405025547198826, | |
| "learning_rate": 8.692640196626858e-05, | |
| "loss": 0.0575, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6319444444444444, | |
| "grad_norm": 0.054678965104638884, | |
| "learning_rate": 8.651478950325737e-05, | |
| "loss": 0.0694, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6388888888888888, | |
| "grad_norm": 0.05031387124367363, | |
| "learning_rate": 8.609780469772623e-05, | |
| "loss": 0.0581, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6458333333333334, | |
| "grad_norm": 0.043219919789406584, | |
| "learning_rate": 8.567550889975362e-05, | |
| "loss": 0.0489, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6527777777777778, | |
| "grad_norm": 0.04843201217718916, | |
| "learning_rate": 8.524796424081292e-05, | |
| "loss": 0.0597, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6597222222222222, | |
| "grad_norm": 0.0522833112433358, | |
| "learning_rate": 8.481523362463111e-05, | |
| "loss": 0.0609, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.04658288241167051, | |
| "learning_rate": 8.437738071793394e-05, | |
| "loss": 0.0461, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6736111111111112, | |
| "grad_norm": 0.05192918762269953, | |
| "learning_rate": 8.393446994107877e-05, | |
| "loss": 0.0675, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6805555555555556, | |
| "grad_norm": 0.04625340723515702, | |
| "learning_rate": 8.348656645857649e-05, | |
| "loss": 0.0501, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6875, | |
| "grad_norm": 0.04545506022482319, | |
| "learning_rate": 8.303373616950408e-05, | |
| "loss": 0.0465, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 0.048499611503797835, | |
| "learning_rate": 8.257604569780897e-05, | |
| "loss": 0.0538, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7013888888888888, | |
| "grad_norm": 0.04849174145205094, | |
| "learning_rate": 8.21135623825068e-05, | |
| "loss": 0.0532, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7083333333333334, | |
| "grad_norm": 0.046809029835309476, | |
| "learning_rate": 8.164635426777404e-05, | |
| "loss": 0.0584, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7152777777777778, | |
| "grad_norm": 0.04338551366678789, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.0486, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 0.04288423757366809, | |
| "learning_rate": 8.069803928235689e-05, | |
| "loss": 0.0512, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7291666666666666, | |
| "grad_norm": 0.04297925782588076, | |
| "learning_rate": 8.021707193521865e-05, | |
| "loss": 0.0466, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7361111111111112, | |
| "grad_norm": 0.04840548556064357, | |
| "learning_rate": 7.973165881521434e-05, | |
| "loss": 0.0614, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7430555555555556, | |
| "grad_norm": 0.05089661493349612, | |
| "learning_rate": 7.924187134013323e-05, | |
| "loss": 0.0468, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.043594180551687345, | |
| "learning_rate": 7.874778157135415e-05, | |
| "loss": 0.0522, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7569444444444444, | |
| "grad_norm": 0.07921359503782698, | |
| "learning_rate": 7.824946220324312e-05, | |
| "loss": 0.0639, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7638888888888888, | |
| "grad_norm": 0.046303003192425585, | |
| "learning_rate": 7.774698655245802e-05, | |
| "loss": 0.054, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7708333333333334, | |
| "grad_norm": 0.0443323814040479, | |
| "learning_rate": 7.724042854716169e-05, | |
| "loss": 0.0492, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.04970508540718419, | |
| "learning_rate": 7.6729862716145e-05, | |
| "loss": 0.0562, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7847222222222222, | |
| "grad_norm": 0.09180702014526151, | |
| "learning_rate": 7.621536417786159e-05, | |
| "loss": 0.069, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7916666666666666, | |
| "grad_norm": 0.047409146676109855, | |
| "learning_rate": 7.56970086293759e-05, | |
| "loss": 0.057, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7986111111111112, | |
| "grad_norm": 0.05564719359814639, | |
| "learning_rate": 7.5174872335226e-05, | |
| "loss": 0.0545, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8055555555555556, | |
| "grad_norm": 0.044385293233721056, | |
| "learning_rate": 7.464903211620291e-05, | |
| "loss": 0.0533, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8125, | |
| "grad_norm": 0.048373968528083365, | |
| "learning_rate": 7.411956533804818e-05, | |
| "loss": 0.0567, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8194444444444444, | |
| "grad_norm": 0.06631096530721073, | |
| "learning_rate": 7.358654990007122e-05, | |
| "loss": 0.0634, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8263888888888888, | |
| "grad_norm": 0.03919679279316613, | |
| "learning_rate": 7.305006422368811e-05, | |
| "loss": 0.0396, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.0612342230223103, | |
| "learning_rate": 7.251018724088367e-05, | |
| "loss": 0.0598, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8402777777777778, | |
| "grad_norm": 0.04761240051837147, | |
| "learning_rate": 7.196699838259834e-05, | |
| "loss": 0.0542, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8472222222222222, | |
| "grad_norm": 0.055239940867233474, | |
| "learning_rate": 7.142057756704168e-05, | |
| "loss": 0.064, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8541666666666666, | |
| "grad_norm": 0.05625030436845944, | |
| "learning_rate": 7.087100518793421e-05, | |
| "loss": 0.0552, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8611111111111112, | |
| "grad_norm": 0.04917209401012176, | |
| "learning_rate": 7.031836210267915e-05, | |
| "loss": 0.0519, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8680555555555556, | |
| "grad_norm": 0.045923821351934915, | |
| "learning_rate": 6.976272962046619e-05, | |
| "loss": 0.0544, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.055983725032457854, | |
| "learning_rate": 6.920418949030856e-05, | |
| "loss": 0.0684, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8819444444444444, | |
| "grad_norm": 0.09704535662409224, | |
| "learning_rate": 6.864282388901544e-05, | |
| "loss": 0.0737, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.044657916384289814, | |
| "learning_rate": 6.807871540910154e-05, | |
| "loss": 0.0515, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8958333333333334, | |
| "grad_norm": 0.07059390531279443, | |
| "learning_rate": 6.751194704663543e-05, | |
| "loss": 0.0574, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.9027777777777778, | |
| "grad_norm": 0.04502070349330495, | |
| "learning_rate": 6.694260218902844e-05, | |
| "loss": 0.0445, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9097222222222222, | |
| "grad_norm": 0.06605902723986466, | |
| "learning_rate": 6.637076460276613e-05, | |
| "loss": 0.0569, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 0.051118257758994964, | |
| "learning_rate": 6.57965184210838e-05, | |
| "loss": 0.0551, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9236111111111112, | |
| "grad_norm": 0.04518774034822205, | |
| "learning_rate": 6.521994813158834e-05, | |
| "loss": 0.0557, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9305555555555556, | |
| "grad_norm": 0.04284945779426791, | |
| "learning_rate": 6.464113856382752e-05, | |
| "loss": 0.041, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.04754862748835347, | |
| "learning_rate": 6.406017487680937e-05, | |
| "loss": 0.0522, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 0.04459511286211968, | |
| "learning_rate": 6.347714254647284e-05, | |
| "loss": 0.0486, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9513888888888888, | |
| "grad_norm": 0.06113290499228477, | |
| "learning_rate": 6.28921273531119e-05, | |
| "loss": 0.0593, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9583333333333334, | |
| "grad_norm": 0.056574006950422875, | |
| "learning_rate": 6.230521536875494e-05, | |
| "loss": 0.0603, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9652777777777778, | |
| "grad_norm": 0.04419993037171285, | |
| "learning_rate": 6.171649294450113e-05, | |
| "loss": 0.0422, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 0.058933736730721914, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.0628, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9791666666666666, | |
| "grad_norm": 0.043997579067719075, | |
| "learning_rate": 6.0533963499786314e-05, | |
| "loss": 0.0416, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9861111111111112, | |
| "grad_norm": 0.0787901330317747, | |
| "learning_rate": 5.994033046234162e-05, | |
| "loss": 0.0663, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9930555555555556, | |
| "grad_norm": 0.056572258685848587, | |
| "learning_rate": 5.934523492543489e-05, | |
| "loss": 0.0598, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.06836555724673425, | |
| "learning_rate": 5.874876444419377e-05, | |
| "loss": 0.0553, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.0069444444444444, | |
| "grad_norm": 0.05528468814497019, | |
| "learning_rate": 5.8151006776038544e-05, | |
| "loss": 0.0487, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0138888888888888, | |
| "grad_norm": 0.04442209097541711, | |
| "learning_rate": 5.75520498677705e-05, | |
| "loss": 0.0496, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0208333333333333, | |
| "grad_norm": 0.04265736671883575, | |
| "learning_rate": 5.6951981842632585e-05, | |
| "loss": 0.0447, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0277777777777777, | |
| "grad_norm": 0.06716393769395008, | |
| "learning_rate": 5.6350890987343944e-05, | |
| "loss": 0.0574, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0347222222222223, | |
| "grad_norm": 0.04619919754230532, | |
| "learning_rate": 5.574886573911056e-05, | |
| "loss": 0.0484, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0416666666666667, | |
| "grad_norm": 0.04088308722298511, | |
| "learning_rate": 5.5145994672613624e-05, | |
| "loss": 0.0424, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0486111111111112, | |
| "grad_norm": 0.07106013370406222, | |
| "learning_rate": 5.4542366486977756e-05, | |
| "loss": 0.0506, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 0.06544088043365288, | |
| "learning_rate": 5.39380699927209e-05, | |
| "loss": 0.0527, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0625, | |
| "grad_norm": 0.055005258429162714, | |
| "learning_rate": 5.3333194098687764e-05, | |
| "loss": 0.0511, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0694444444444444, | |
| "grad_norm": 0.05657332977154827, | |
| "learning_rate": 5.272782779896898e-05, | |
| "loss": 0.0543, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0763888888888888, | |
| "grad_norm": 0.08633216171192509, | |
| "learning_rate": 5.212206015980742e-05, | |
| "loss": 0.0595, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0833333333333333, | |
| "grad_norm": 0.04344217768745669, | |
| "learning_rate": 5.151598030649425e-05, | |
| "loss": 0.0396, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0902777777777777, | |
| "grad_norm": 0.04446718541450083, | |
| "learning_rate": 5.0909677410255985e-05, | |
| "loss": 0.0436, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0972222222222223, | |
| "grad_norm": 0.05936952558342409, | |
| "learning_rate": 5.030324067513499e-05, | |
| "loss": 0.0514, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.1041666666666667, | |
| "grad_norm": 0.06506431319350331, | |
| "learning_rate": 4.969675932486503e-05, | |
| "loss": 0.0472, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.04763084580847737, | |
| "learning_rate": 4.9090322589744027e-05, | |
| "loss": 0.0436, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1180555555555556, | |
| "grad_norm": 0.06531630358906274, | |
| "learning_rate": 4.848401969350577e-05, | |
| "loss": 0.0547, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 0.06486635891777948, | |
| "learning_rate": 4.78779398401926e-05, | |
| "loss": 0.0454, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.1319444444444444, | |
| "grad_norm": 0.06325850216814397, | |
| "learning_rate": 4.7272172201031054e-05, | |
| "loss": 0.0481, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.1388888888888888, | |
| "grad_norm": 0.05964254119159708, | |
| "learning_rate": 4.666680590131225e-05, | |
| "loss": 0.0463, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.1458333333333333, | |
| "grad_norm": 0.05342845760927094, | |
| "learning_rate": 4.606193000727913e-05, | |
| "loss": 0.0506, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.1527777777777777, | |
| "grad_norm": 0.05377115554688926, | |
| "learning_rate": 4.545763351302224e-05, | |
| "loss": 0.0474, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.1597222222222223, | |
| "grad_norm": 0.05055891815765679, | |
| "learning_rate": 4.485400532738638e-05, | |
| "loss": 0.0454, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.05497839217223033, | |
| "learning_rate": 4.425113426088945e-05, | |
| "loss": 0.0517, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.1736111111111112, | |
| "grad_norm": 0.07316550289620528, | |
| "learning_rate": 4.364910901265606e-05, | |
| "loss": 0.0538, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.1805555555555556, | |
| "grad_norm": 0.047990097455734924, | |
| "learning_rate": 4.3048018157367433e-05, | |
| "loss": 0.0413, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1875, | |
| "grad_norm": 0.060217032523026584, | |
| "learning_rate": 4.244795013222951e-05, | |
| "loss": 0.0571, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1944444444444444, | |
| "grad_norm": 0.04698866045082108, | |
| "learning_rate": 4.184899322396147e-05, | |
| "loss": 0.0406, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.2013888888888888, | |
| "grad_norm": 0.048369004985907946, | |
| "learning_rate": 4.125123555580624e-05, | |
| "loss": 0.0492, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.2083333333333333, | |
| "grad_norm": 0.05659619196658486, | |
| "learning_rate": 4.0654765074565124e-05, | |
| "loss": 0.057, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2152777777777777, | |
| "grad_norm": 0.057459855223361965, | |
| "learning_rate": 4.005966953765839e-05, | |
| "loss": 0.0479, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.04572437211261691, | |
| "learning_rate": 3.94660365002137e-05, | |
| "loss": 0.0395, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2291666666666667, | |
| "grad_norm": 0.048219072573456936, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0391, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2361111111111112, | |
| "grad_norm": 0.06136131829899375, | |
| "learning_rate": 3.8283507055498886e-05, | |
| "loss": 0.043, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2430555555555556, | |
| "grad_norm": 0.059005391153620405, | |
| "learning_rate": 3.769478463124507e-05, | |
| "loss": 0.0449, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.0746524056732719, | |
| "learning_rate": 3.7107872646888116e-05, | |
| "loss": 0.0563, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2569444444444444, | |
| "grad_norm": 0.06512798583168254, | |
| "learning_rate": 3.652285745352717e-05, | |
| "loss": 0.0652, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2638888888888888, | |
| "grad_norm": 0.05118809031893748, | |
| "learning_rate": 3.5939825123190635e-05, | |
| "loss": 0.0439, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2708333333333333, | |
| "grad_norm": 0.0491373328570806, | |
| "learning_rate": 3.5358861436172485e-05, | |
| "loss": 0.0396, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.2777777777777777, | |
| "grad_norm": 0.045715529417524076, | |
| "learning_rate": 3.4780051868411675e-05, | |
| "loss": 0.0349, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2847222222222223, | |
| "grad_norm": 0.05499812650132431, | |
| "learning_rate": 3.4203481578916194e-05, | |
| "loss": 0.045, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2916666666666667, | |
| "grad_norm": 0.05444819130310788, | |
| "learning_rate": 3.362923539723389e-05, | |
| "loss": 0.054, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2986111111111112, | |
| "grad_norm": 0.05755054434891386, | |
| "learning_rate": 3.305739781097157e-05, | |
| "loss": 0.0462, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.3055555555555556, | |
| "grad_norm": 0.056782562230561044, | |
| "learning_rate": 3.248805295336458e-05, | |
| "loss": 0.049, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.3125, | |
| "grad_norm": 0.05059433419507378, | |
| "learning_rate": 3.1921284590898456e-05, | |
| "loss": 0.0398, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3194444444444444, | |
| "grad_norm": 0.06312033575652291, | |
| "learning_rate": 3.135717611098458e-05, | |
| "loss": 0.0622, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3263888888888888, | |
| "grad_norm": 0.06684980111755691, | |
| "learning_rate": 3.079581050969146e-05, | |
| "loss": 0.0628, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.09404520793015296, | |
| "learning_rate": 3.023727037953382e-05, | |
| "loss": 0.0538, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3402777777777777, | |
| "grad_norm": 0.055119231303708986, | |
| "learning_rate": 2.9681637897320868e-05, | |
| "loss": 0.0494, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3472222222222223, | |
| "grad_norm": 0.06806024002262281, | |
| "learning_rate": 2.912899481206582e-05, | |
| "loss": 0.0494, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3541666666666667, | |
| "grad_norm": 0.05634926402516711, | |
| "learning_rate": 2.8579422432958312e-05, | |
| "loss": 0.0474, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3611111111111112, | |
| "grad_norm": 0.062090271560473254, | |
| "learning_rate": 2.803300161740166e-05, | |
| "loss": 0.0522, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3680555555555556, | |
| "grad_norm": 0.05050749410112152, | |
| "learning_rate": 2.748981275911633e-05, | |
| "loss": 0.0393, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.05546645285466708, | |
| "learning_rate": 2.6949935776311896e-05, | |
| "loss": 0.0395, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3819444444444444, | |
| "grad_norm": 0.05531341419035151, | |
| "learning_rate": 2.6413450099928783e-05, | |
| "loss": 0.0445, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "grad_norm": 0.057587022878495084, | |
| "learning_rate": 2.5880434661951823e-05, | |
| "loss": 0.046, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3888888888888888, | |
| "eval_loss": 0.05843019485473633, | |
| "eval_runtime": 2.8965, | |
| "eval_samples_per_second": 2.071, | |
| "eval_steps_per_second": 0.69, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3958333333333333, | |
| "grad_norm": 0.06587462746896457, | |
| "learning_rate": 2.5350967883797096e-05, | |
| "loss": 0.0592, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.4027777777777777, | |
| "grad_norm": 0.05623019516285661, | |
| "learning_rate": 2.4825127664774006e-05, | |
| "loss": 0.0492, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.4097222222222223, | |
| "grad_norm": 0.06970774002767344, | |
| "learning_rate": 2.4302991370624107e-05, | |
| "loss": 0.0599, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.4166666666666667, | |
| "grad_norm": 0.050591438460209116, | |
| "learning_rate": 2.3784635822138424e-05, | |
| "loss": 0.0405, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.4236111111111112, | |
| "grad_norm": 0.06087979461292365, | |
| "learning_rate": 2.327013728385502e-05, | |
| "loss": 0.0522, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4305555555555556, | |
| "grad_norm": 0.05750045979367977, | |
| "learning_rate": 2.2759571452838324e-05, | |
| "loss": 0.0498, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.4375, | |
| "grad_norm": 0.06261633186118712, | |
| "learning_rate": 2.225301344754199e-05, | |
| "loss": 0.0516, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.04778739915939838, | |
| "learning_rate": 2.17505377967569e-05, | |
| "loss": 0.0356, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.4513888888888888, | |
| "grad_norm": 0.06838727020740598, | |
| "learning_rate": 2.1252218428645846e-05, | |
| "loss": 0.0553, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.4583333333333333, | |
| "grad_norm": 0.05699327263249455, | |
| "learning_rate": 2.075812865986677e-05, | |
| "loss": 0.0436, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4652777777777777, | |
| "grad_norm": 0.05755232280782536, | |
| "learning_rate": 2.026834118478567e-05, | |
| "loss": 0.0531, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.4722222222222223, | |
| "grad_norm": 0.06672702654987289, | |
| "learning_rate": 1.978292806478134e-05, | |
| "loss": 0.0529, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.4791666666666667, | |
| "grad_norm": 0.05318360817731425, | |
| "learning_rate": 1.9301960717643118e-05, | |
| "loss": 0.0408, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.4861111111111112, | |
| "grad_norm": 0.05852145674739832, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0532, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.4930555555555556, | |
| "grad_norm": 0.04695844857372584, | |
| "learning_rate": 1.8353645732225976e-05, | |
| "loss": 0.0357, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.06015704770880845, | |
| "learning_rate": 1.7886437617493205e-05, | |
| "loss": 0.0458, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.5069444444444444, | |
| "grad_norm": 0.06607871507117989, | |
| "learning_rate": 1.7423954302191047e-05, | |
| "loss": 0.065, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.5138888888888888, | |
| "grad_norm": 0.07091336376792805, | |
| "learning_rate": 1.6966263830495936e-05, | |
| "loss": 0.0363, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5208333333333335, | |
| "grad_norm": 0.05916059675448067, | |
| "learning_rate": 1.6513433541423528e-05, | |
| "loss": 0.0504, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5277777777777777, | |
| "grad_norm": 0.06409972590419019, | |
| "learning_rate": 1.606553005892125e-05, | |
| "loss": 0.0393, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5347222222222223, | |
| "grad_norm": 0.06555802647292505, | |
| "learning_rate": 1.5622619282066082e-05, | |
| "loss": 0.0482, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.5416666666666665, | |
| "grad_norm": 0.05144622081409977, | |
| "learning_rate": 1.5184766375368915e-05, | |
| "loss": 0.037, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.5486111111111112, | |
| "grad_norm": 0.06265878609345442, | |
| "learning_rate": 1.4752035759187106e-05, | |
| "loss": 0.0536, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.05834542354430463, | |
| "learning_rate": 1.4324491100246385e-05, | |
| "loss": 0.0461, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 0.059108211105270696, | |
| "learning_rate": 1.3902195302273779e-05, | |
| "loss": 0.0461, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5694444444444444, | |
| "grad_norm": 0.06072177930930135, | |
| "learning_rate": 1.348521049674264e-05, | |
| "loss": 0.0448, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5763888888888888, | |
| "grad_norm": 0.06147140484331682, | |
| "learning_rate": 1.3073598033731426e-05, | |
| "loss": 0.0475, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.5833333333333335, | |
| "grad_norm": 0.05617171004497398, | |
| "learning_rate": 1.2667418472897386e-05, | |
| "loss": 0.0432, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.5902777777777777, | |
| "grad_norm": 0.06266225996920342, | |
| "learning_rate": 1.2266731574566536e-05, | |
| "loss": 0.0446, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5972222222222223, | |
| "grad_norm": 0.054440007451986294, | |
| "learning_rate": 1.1871596290941278e-05, | |
| "loss": 0.0403, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.6041666666666665, | |
| "grad_norm": 0.05771513928537823, | |
| "learning_rate": 1.1482070757426856e-05, | |
| "loss": 0.0424, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.6111111111111112, | |
| "grad_norm": 0.07317554509540891, | |
| "learning_rate": 1.1098212284078036e-05, | |
| "loss": 0.0486, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6180555555555556, | |
| "grad_norm": 0.0592998028315768, | |
| "learning_rate": 1.0720077347167202e-05, | |
| "loss": 0.0452, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 0.05734391978777635, | |
| "learning_rate": 1.0347721580875126e-05, | |
| "loss": 0.0418, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.6319444444444444, | |
| "grad_norm": 0.05781410001692482, | |
| "learning_rate": 9.981199769105604e-06, | |
| "loss": 0.0488, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6388888888888888, | |
| "grad_norm": 0.06084611217062411, | |
| "learning_rate": 9.620565837425271e-06, | |
| "loss": 0.0544, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6458333333333335, | |
| "grad_norm": 0.05484483522551538, | |
| "learning_rate": 9.26587284512957e-06, | |
| "loss": 0.0418, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6527777777777777, | |
| "grad_norm": 0.061075107734742555, | |
| "learning_rate": 8.917172977436356e-06, | |
| "loss": 0.0509, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.6597222222222223, | |
| "grad_norm": 0.056522690917951715, | |
| "learning_rate": 8.574517537807897e-06, | |
| "loss": 0.0493, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.06145420768545982, | |
| "learning_rate": 8.237956940402757e-06, | |
| "loss": 0.0538, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.6736111111111112, | |
| "grad_norm": 0.07712832225863908, | |
| "learning_rate": 7.907540702658456e-06, | |
| "loss": 0.0417, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.6805555555555556, | |
| "grad_norm": 0.1210354286083926, | |
| "learning_rate": 7.583317438006093e-06, | |
| "loss": 0.0648, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.6875, | |
| "grad_norm": 0.056155382622722974, | |
| "learning_rate": 7.265334848717931e-06, | |
| "loss": 0.0409, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6944444444444444, | |
| "grad_norm": 0.06052338165886097, | |
| "learning_rate": 6.953639718889076e-06, | |
| "loss": 0.0425, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.7013888888888888, | |
| "grad_norm": 0.06942550868803853, | |
| "learning_rate": 6.648277907554235e-06, | |
| "loss": 0.0483, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.7083333333333335, | |
| "grad_norm": 0.06102534205447137, | |
| "learning_rate": 6.349294341940593e-06, | |
| "loss": 0.0461, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.7152777777777777, | |
| "grad_norm": 0.06664766533657307, | |
| "learning_rate": 6.056733010857712e-06, | |
| "loss": 0.0538, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7222222222222223, | |
| "grad_norm": 0.055246592628772316, | |
| "learning_rate": 5.770636958225617e-06, | |
| "loss": 0.0455, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7291666666666665, | |
| "grad_norm": 0.061169981220161505, | |
| "learning_rate": 5.491048276741784e-06, | |
| "loss": 0.0444, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.7361111111111112, | |
| "grad_norm": 0.07327155244232343, | |
| "learning_rate": 5.218008101688171e-06, | |
| "loss": 0.0607, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7430555555555556, | |
| "grad_norm": 0.06145818807633177, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0445, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.05297976039230036, | |
| "learning_rate": 4.691732988750614e-06, | |
| "loss": 0.0388, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.7569444444444444, | |
| "grad_norm": 0.054054619462972, | |
| "learning_rate": 4.43857548059321e-06, | |
| "loss": 0.0391, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.7638888888888888, | |
| "grad_norm": 0.05798999532961248, | |
| "learning_rate": 4.192121326927073e-06, | |
| "loss": 0.0432, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7708333333333335, | |
| "grad_norm": 0.05728064735233047, | |
| "learning_rate": 3.952406788022267e-06, | |
| "loss": 0.044, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.05548812892517234, | |
| "learning_rate": 3.7194671325638198e-06, | |
| "loss": 0.042, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7847222222222223, | |
| "grad_norm": 0.05776655459801612, | |
| "learning_rate": 3.493336632462718e-06, | |
| "loss": 0.0446, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7916666666666665, | |
| "grad_norm": 0.06297495738200812, | |
| "learning_rate": 3.274048557813553e-06, | |
| "loss": 0.0375, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7986111111111112, | |
| "grad_norm": 0.0571294673264027, | |
| "learning_rate": 3.061635171999566e-06, | |
| "loss": 0.0458, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.8055555555555556, | |
| "grad_norm": 0.05711653877850546, | |
| "learning_rate": 2.85612772694579e-06, | |
| "loss": 0.043, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.8125, | |
| "grad_norm": 0.056535662892683275, | |
| "learning_rate": 2.657556458521049e-06, | |
| "loss": 0.0422, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8194444444444444, | |
| "grad_norm": 0.05861262152488929, | |
| "learning_rate": 2.4659505820893826e-06, | |
| "loss": 0.049, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.8263888888888888, | |
| "grad_norm": 0.057433042780234714, | |
| "learning_rate": 2.2813382882116986e-06, | |
| "loss": 0.0437, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.05389278487965654, | |
| "learning_rate": 2.1037467384981026e-06, | |
| "loss": 0.0426, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8402777777777777, | |
| "grad_norm": 0.05557228450230491, | |
| "learning_rate": 1.933202061611722e-06, | |
| "loss": 0.0458, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8472222222222223, | |
| "grad_norm": 0.05917014990003017, | |
| "learning_rate": 1.769729349424415e-06, | |
| "loss": 0.0531, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.8541666666666665, | |
| "grad_norm": 0.057944432503720486, | |
| "learning_rate": 1.6133526533250565e-06, | |
| "loss": 0.0425, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8611111111111112, | |
| "grad_norm": 0.06092440732902568, | |
| "learning_rate": 1.4640949806809523e-06, | |
| "loss": 0.0492, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8680555555555556, | |
| "grad_norm": 0.05900717359821016, | |
| "learning_rate": 1.3219782914527634e-06, | |
| "loss": 0.0452, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 0.06389890797796334, | |
| "learning_rate": 1.1870234949636073e-06, | |
| "loss": 0.0549, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8819444444444444, | |
| "grad_norm": 0.05962740065521039, | |
| "learning_rate": 1.0592504468227127e-06, | |
| "loss": 0.0463, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.07174635089821045, | |
| "learning_rate": 9.386779460041017e-07, | |
| "loss": 0.0522, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.8958333333333335, | |
| "grad_norm": 0.06082782248412747, | |
| "learning_rate": 8.253237320807461e-07, | |
| "loss": 0.0475, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.9027777777777777, | |
| "grad_norm": 0.0601520271544199, | |
| "learning_rate": 7.192044826145771e-07, | |
| "loss": 0.0498, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.9097222222222223, | |
| "grad_norm": 0.05825976235647901, | |
| "learning_rate": 6.20335810702749e-07, | |
| "loss": 0.0422, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.9166666666666665, | |
| "grad_norm": 0.05517949333399056, | |
| "learning_rate": 5.287322626805202e-07, | |
| "loss": 0.0433, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9236111111111112, | |
| "grad_norm": 0.06763262497028635, | |
| "learning_rate": 4.4440731598107686e-07, | |
| "loss": 0.0499, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9305555555555556, | |
| "grad_norm": 0.16139138859406038, | |
| "learning_rate": 3.673733771526466e-07, | |
| "loss": 0.0664, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9375, | |
| "grad_norm": 0.056764953267334266, | |
| "learning_rate": 2.976417800331144e-07, | |
| "loss": 0.0395, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9444444444444444, | |
| "grad_norm": 0.06087509888868891, | |
| "learning_rate": 2.352227840825394e-07, | |
| "loss": 0.0456, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9513888888888888, | |
| "grad_norm": 0.07013826621193169, | |
| "learning_rate": 1.8012557287367392e-07, | |
| "loss": 0.0449, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9583333333333335, | |
| "grad_norm": 0.06122401225786666, | |
| "learning_rate": 1.3235825274081625e-07, | |
| "loss": 0.0433, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.9652777777777777, | |
| "grad_norm": 0.062035625059352666, | |
| "learning_rate": 9.19278515871369e-08, | |
| "loss": 0.0516, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.9722222222222223, | |
| "grad_norm": 0.05458425287261523, | |
| "learning_rate": 5.8840317850683555e-08, | |
| "loss": 0.0392, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9791666666666665, | |
| "grad_norm": 0.0574212811670188, | |
| "learning_rate": 3.310051962920335e-08, | |
| "loss": 0.0471, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9861111111111112, | |
| "grad_norm": 0.05924979972086086, | |
| "learning_rate": 1.471224396389359e-08, | |
| "loss": 0.0542, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9930555555555556, | |
| "grad_norm": 0.05752784210946029, | |
| "learning_rate": 3.6781962822529657e-09, | |
| "loss": 0.0474, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.0627018913384721, | |
| "learning_rate": 0.0, | |
| "loss": 0.0448, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 288, | |
| "total_flos": 547031730880512.0, | |
| "train_loss": 0.056736280779457755, | |
| "train_runtime": 1071.8886, | |
| "train_samples_per_second": 1.069, | |
| "train_steps_per_second": 0.269 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 288, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 547031730880512.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |