| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 714, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04201680672268908, | |
| "grad_norm": 7.36784279593653, | |
| "learning_rate": 3.4722222222222224e-06, | |
| "loss": 0.9203, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08403361344537816, | |
| "grad_norm": 2.6817007224142193, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.8488, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12605042016806722, | |
| "grad_norm": 1.1173297473773625, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 0.6786, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16806722689075632, | |
| "grad_norm": 0.5432821597326414, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.6023, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21008403361344538, | |
| "grad_norm": 0.4253948970109724, | |
| "learning_rate": 1.736111111111111e-05, | |
| "loss": 0.5601, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.25210084033613445, | |
| "grad_norm": 0.31150432991851523, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.5233, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 0.32050134592917545, | |
| "learning_rate": 2.4305555555555558e-05, | |
| "loss": 0.506, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.33613445378151263, | |
| "grad_norm": 0.27149516786646455, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.4902, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.37815126050420167, | |
| "grad_norm": 0.256388828476645, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.466, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.42016806722689076, | |
| "grad_norm": 0.2604693465205412, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 0.4627, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.46218487394957986, | |
| "grad_norm": 0.23787838722566912, | |
| "learning_rate": 3.8194444444444444e-05, | |
| "loss": 0.443, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5042016806722689, | |
| "grad_norm": 0.23524988298322033, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.4347, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5462184873949579, | |
| "grad_norm": 0.720408504919723, | |
| "learning_rate": 4.5138888888888894e-05, | |
| "loss": 0.4325, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.15502732828730537, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 0.4092, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6302521008403361, | |
| "grad_norm": 0.1681572620191539, | |
| "learning_rate": 4.976635514018692e-05, | |
| "loss": 0.4144, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6722689075630253, | |
| "grad_norm": 0.14289664425952017, | |
| "learning_rate": 4.937694704049845e-05, | |
| "loss": 0.4101, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.1423836892341792, | |
| "learning_rate": 4.898753894080997e-05, | |
| "loss": 0.4028, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7563025210084033, | |
| "grad_norm": 0.13921521879920523, | |
| "learning_rate": 4.85981308411215e-05, | |
| "loss": 0.3961, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7983193277310925, | |
| "grad_norm": 0.1320808306944742, | |
| "learning_rate": 4.820872274143303e-05, | |
| "loss": 0.4057, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8403361344537815, | |
| "grad_norm": 0.13511566705435527, | |
| "learning_rate": 4.781931464174455e-05, | |
| "loss": 0.3982, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 0.12059831677330138, | |
| "learning_rate": 4.742990654205608e-05, | |
| "loss": 0.3926, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9243697478991597, | |
| "grad_norm": 0.14025999739783343, | |
| "learning_rate": 4.7040498442367604e-05, | |
| "loss": 0.3863, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9663865546218487, | |
| "grad_norm": 0.13908375301311737, | |
| "learning_rate": 4.665109034267913e-05, | |
| "loss": 0.3985, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0084033613445378, | |
| "grad_norm": 0.1402995196474956, | |
| "learning_rate": 4.6261682242990654e-05, | |
| "loss": 0.3844, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.050420168067227, | |
| "grad_norm": 0.14896376645944973, | |
| "learning_rate": 4.5872274143302186e-05, | |
| "loss": 0.3711, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.092436974789916, | |
| "grad_norm": 0.1330752274361008, | |
| "learning_rate": 4.548286604361371e-05, | |
| "loss": 0.3627, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.134453781512605, | |
| "grad_norm": 0.12286078605816185, | |
| "learning_rate": 4.5093457943925236e-05, | |
| "loss": 0.3627, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.1300040241362218, | |
| "learning_rate": 4.470404984423676e-05, | |
| "loss": 0.3586, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2184873949579833, | |
| "grad_norm": 0.1212676350760631, | |
| "learning_rate": 4.431464174454829e-05, | |
| "loss": 0.3653, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2605042016806722, | |
| "grad_norm": 0.13672573392631662, | |
| "learning_rate": 4.392523364485982e-05, | |
| "loss": 0.3601, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3025210084033614, | |
| "grad_norm": 0.12866697257558982, | |
| "learning_rate": 4.353582554517134e-05, | |
| "loss": 0.3646, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3445378151260505, | |
| "grad_norm": 0.12251791003688832, | |
| "learning_rate": 4.314641744548287e-05, | |
| "loss": 0.3554, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3865546218487395, | |
| "grad_norm": 0.128140623459501, | |
| "learning_rate": 4.27570093457944e-05, | |
| "loss": 0.3597, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.12517942956515188, | |
| "learning_rate": 4.236760124610592e-05, | |
| "loss": 0.3596, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 0.1264068992533434, | |
| "learning_rate": 4.197819314641745e-05, | |
| "loss": 0.3523, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.5126050420168067, | |
| "grad_norm": 0.12844150586086567, | |
| "learning_rate": 4.1588785046728974e-05, | |
| "loss": 0.3611, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.5546218487394958, | |
| "grad_norm": 0.1463320479548455, | |
| "learning_rate": 4.11993769470405e-05, | |
| "loss": 0.3544, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.596638655462185, | |
| "grad_norm": 0.1387410614743581, | |
| "learning_rate": 4.0809968847352024e-05, | |
| "loss": 0.3506, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.638655462184874, | |
| "grad_norm": 0.11966374174214205, | |
| "learning_rate": 4.0420560747663556e-05, | |
| "loss": 0.3547, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.680672268907563, | |
| "grad_norm": 0.13041047515525458, | |
| "learning_rate": 4.003115264797508e-05, | |
| "loss": 0.3551, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7226890756302522, | |
| "grad_norm": 0.4175865240152316, | |
| "learning_rate": 3.9641744548286606e-05, | |
| "loss": 0.359, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 0.12341307943053885, | |
| "learning_rate": 3.925233644859813e-05, | |
| "loss": 0.3509, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8067226890756303, | |
| "grad_norm": 0.11938805910700397, | |
| "learning_rate": 3.886292834890966e-05, | |
| "loss": 0.353, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8487394957983194, | |
| "grad_norm": 0.11924644124050648, | |
| "learning_rate": 3.847352024922119e-05, | |
| "loss": 0.3543, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8907563025210083, | |
| "grad_norm": 0.1192046521417864, | |
| "learning_rate": 3.808411214953271e-05, | |
| "loss": 0.3459, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9327731092436975, | |
| "grad_norm": 0.11198889670651223, | |
| "learning_rate": 3.769470404984424e-05, | |
| "loss": 0.3526, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9747899159663866, | |
| "grad_norm": 0.11064392419307578, | |
| "learning_rate": 3.730529595015577e-05, | |
| "loss": 0.3481, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.0168067226890756, | |
| "grad_norm": 0.14955656421111635, | |
| "learning_rate": 3.691588785046729e-05, | |
| "loss": 0.334, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0588235294117645, | |
| "grad_norm": 0.1352971990563673, | |
| "learning_rate": 3.652647975077882e-05, | |
| "loss": 0.3192, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.100840336134454, | |
| "grad_norm": 0.16599609164840282, | |
| "learning_rate": 3.6137071651090344e-05, | |
| "loss": 0.3164, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.13656518038762977, | |
| "learning_rate": 3.574766355140187e-05, | |
| "loss": 0.3124, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.184873949579832, | |
| "grad_norm": 0.13474455104744834, | |
| "learning_rate": 3.5358255451713394e-05, | |
| "loss": 0.3119, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.226890756302521, | |
| "grad_norm": 0.12821245064355621, | |
| "learning_rate": 3.4968847352024926e-05, | |
| "loss": 0.3168, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.26890756302521, | |
| "grad_norm": 0.13307269371542046, | |
| "learning_rate": 3.457943925233645e-05, | |
| "loss": 0.3192, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.310924369747899, | |
| "grad_norm": 0.14021981893226773, | |
| "learning_rate": 3.4190031152647976e-05, | |
| "loss": 0.3159, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.13437639361334078, | |
| "learning_rate": 3.38006230529595e-05, | |
| "loss": 0.3123, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.3949579831932772, | |
| "grad_norm": 0.1326899305498554, | |
| "learning_rate": 3.341121495327103e-05, | |
| "loss": 0.3177, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.4369747899159666, | |
| "grad_norm": 0.12293267360959077, | |
| "learning_rate": 3.302180685358255e-05, | |
| "loss": 0.3159, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4789915966386555, | |
| "grad_norm": 0.12467837886737294, | |
| "learning_rate": 3.263239875389408e-05, | |
| "loss": 0.3152, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.5210084033613445, | |
| "grad_norm": 0.1245503385101852, | |
| "learning_rate": 3.224299065420561e-05, | |
| "loss": 0.3172, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5630252100840334, | |
| "grad_norm": 0.12135305650404841, | |
| "learning_rate": 3.185358255451714e-05, | |
| "loss": 0.3131, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.6050420168067228, | |
| "grad_norm": 0.1196869916735729, | |
| "learning_rate": 3.146417445482866e-05, | |
| "loss": 0.3131, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.6470588235294117, | |
| "grad_norm": 0.1422717014600059, | |
| "learning_rate": 3.107476635514019e-05, | |
| "loss": 0.3122, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.689075630252101, | |
| "grad_norm": 0.12049177418433844, | |
| "learning_rate": 3.0685358255451714e-05, | |
| "loss": 0.3118, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.73109243697479, | |
| "grad_norm": 0.1218866413247397, | |
| "learning_rate": 3.0295950155763243e-05, | |
| "loss": 0.3093, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.773109243697479, | |
| "grad_norm": 0.12128404985597208, | |
| "learning_rate": 2.9906542056074764e-05, | |
| "loss": 0.3149, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.815126050420168, | |
| "grad_norm": 0.12165521196292105, | |
| "learning_rate": 2.9517133956386296e-05, | |
| "loss": 0.3152, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.11968176603091883, | |
| "learning_rate": 2.9127725856697818e-05, | |
| "loss": 0.3065, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.899159663865546, | |
| "grad_norm": 0.13174397154227724, | |
| "learning_rate": 2.873831775700935e-05, | |
| "loss": 0.3161, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 0.12078966649058866, | |
| "learning_rate": 2.834890965732087e-05, | |
| "loss": 0.3143, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.9831932773109244, | |
| "grad_norm": 0.12196146741006934, | |
| "learning_rate": 2.79595015576324e-05, | |
| "loss": 0.32, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.0252100840336134, | |
| "grad_norm": 0.22204914838949716, | |
| "learning_rate": 2.7570093457943924e-05, | |
| "loss": 0.2892, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.0672268907563027, | |
| "grad_norm": 0.17224639019450644, | |
| "learning_rate": 2.7180685358255453e-05, | |
| "loss": 0.2803, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.1092436974789917, | |
| "grad_norm": 0.12584678266855343, | |
| "learning_rate": 2.6791277258566978e-05, | |
| "loss": 0.2762, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.1512605042016806, | |
| "grad_norm": 0.14018543572816114, | |
| "learning_rate": 2.6401869158878506e-05, | |
| "loss": 0.2791, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.19327731092437, | |
| "grad_norm": 0.12941623347853976, | |
| "learning_rate": 2.601246105919003e-05, | |
| "loss": 0.2749, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.235294117647059, | |
| "grad_norm": 0.12194288998302227, | |
| "learning_rate": 2.562305295950156e-05, | |
| "loss": 0.2803, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.277310924369748, | |
| "grad_norm": 0.12446866234019027, | |
| "learning_rate": 2.5233644859813084e-05, | |
| "loss": 0.2805, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.3193277310924367, | |
| "grad_norm": 0.12351722292105845, | |
| "learning_rate": 2.4844236760124613e-05, | |
| "loss": 0.2771, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.361344537815126, | |
| "grad_norm": 0.12653348786860177, | |
| "learning_rate": 2.4454828660436138e-05, | |
| "loss": 0.2781, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.403361344537815, | |
| "grad_norm": 0.13084716393704046, | |
| "learning_rate": 2.4065420560747666e-05, | |
| "loss": 0.2782, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.4453781512605044, | |
| "grad_norm": 0.12386845332574656, | |
| "learning_rate": 2.367601246105919e-05, | |
| "loss": 0.2789, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.4873949579831933, | |
| "grad_norm": 0.12966375682019396, | |
| "learning_rate": 2.3286604361370716e-05, | |
| "loss": 0.2859, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.1244203255643183, | |
| "learning_rate": 2.2897196261682244e-05, | |
| "loss": 0.2808, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.12741940430874554, | |
| "learning_rate": 2.250778816199377e-05, | |
| "loss": 0.2787, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.6134453781512605, | |
| "grad_norm": 0.13192944689673486, | |
| "learning_rate": 2.2118380062305298e-05, | |
| "loss": 0.2844, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.6554621848739495, | |
| "grad_norm": 0.12331528955859135, | |
| "learning_rate": 2.1728971962616822e-05, | |
| "loss": 0.2825, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.697478991596639, | |
| "grad_norm": 0.13313354832203048, | |
| "learning_rate": 2.133956386292835e-05, | |
| "loss": 0.2816, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.7394957983193278, | |
| "grad_norm": 0.13549454769327543, | |
| "learning_rate": 2.0950155763239876e-05, | |
| "loss": 0.2817, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.7815126050420167, | |
| "grad_norm": 0.11803733674337212, | |
| "learning_rate": 2.05607476635514e-05, | |
| "loss": 0.28, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.8235294117647056, | |
| "grad_norm": 0.12012685963323684, | |
| "learning_rate": 2.017133956386293e-05, | |
| "loss": 0.2832, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 3.865546218487395, | |
| "grad_norm": 0.11875060997630844, | |
| "learning_rate": 1.9781931464174454e-05, | |
| "loss": 0.2806, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.907563025210084, | |
| "grad_norm": 0.13503311291347644, | |
| "learning_rate": 1.9392523364485982e-05, | |
| "loss": 0.2857, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 3.9495798319327733, | |
| "grad_norm": 0.13899315092229172, | |
| "learning_rate": 1.9003115264797507e-05, | |
| "loss": 0.2819, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.991596638655462, | |
| "grad_norm": 0.12553466656260837, | |
| "learning_rate": 1.8613707165109036e-05, | |
| "loss": 0.2781, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.033613445378151, | |
| "grad_norm": 0.2047499500117637, | |
| "learning_rate": 1.822429906542056e-05, | |
| "loss": 0.2559, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.07563025210084, | |
| "grad_norm": 0.16423755258674735, | |
| "learning_rate": 1.7834890965732086e-05, | |
| "loss": 0.2558, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.117647058823529, | |
| "grad_norm": 0.1348676429167651, | |
| "learning_rate": 1.7445482866043614e-05, | |
| "loss": 0.2527, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.159663865546219, | |
| "grad_norm": 0.14533312788167388, | |
| "learning_rate": 1.705607476635514e-05, | |
| "loss": 0.251, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.201680672268908, | |
| "grad_norm": 0.12637560985053983, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.2519, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.243697478991597, | |
| "grad_norm": 0.12843775741491584, | |
| "learning_rate": 1.6277258566978192e-05, | |
| "loss": 0.2497, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.12903589040675842, | |
| "learning_rate": 1.588785046728972e-05, | |
| "loss": 0.2493, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.3277310924369745, | |
| "grad_norm": 0.1265086519312201, | |
| "learning_rate": 1.5498442367601246e-05, | |
| "loss": 0.2478, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.369747899159664, | |
| "grad_norm": 0.12474752965681195, | |
| "learning_rate": 1.5109034267912772e-05, | |
| "loss": 0.2552, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.411764705882353, | |
| "grad_norm": 0.12276341594269388, | |
| "learning_rate": 1.4719626168224299e-05, | |
| "loss": 0.2566, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.453781512605042, | |
| "grad_norm": 0.12799056315008822, | |
| "learning_rate": 1.4330218068535826e-05, | |
| "loss": 0.2544, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.495798319327731, | |
| "grad_norm": 0.1235959721069884, | |
| "learning_rate": 1.394080996884735e-05, | |
| "loss": 0.2573, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.53781512605042, | |
| "grad_norm": 0.12737316017685743, | |
| "learning_rate": 1.3551401869158877e-05, | |
| "loss": 0.251, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.579831932773109, | |
| "grad_norm": 0.12455494002416401, | |
| "learning_rate": 1.3161993769470404e-05, | |
| "loss": 0.2472, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.621848739495798, | |
| "grad_norm": 0.12200067774607738, | |
| "learning_rate": 1.277258566978193e-05, | |
| "loss": 0.2512, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.663865546218488, | |
| "grad_norm": 0.14205343587283586, | |
| "learning_rate": 1.2383177570093459e-05, | |
| "loss": 0.2518, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.12216097165620664, | |
| "learning_rate": 1.1993769470404986e-05, | |
| "loss": 0.252, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.7478991596638656, | |
| "grad_norm": 0.12513180450823716, | |
| "learning_rate": 1.160436137071651e-05, | |
| "loss": 0.2585, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 4.7899159663865545, | |
| "grad_norm": 0.12156552612032653, | |
| "learning_rate": 1.1214953271028037e-05, | |
| "loss": 0.2543, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.831932773109243, | |
| "grad_norm": 0.12936785318700886, | |
| "learning_rate": 1.0825545171339564e-05, | |
| "loss": 0.2511, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 4.873949579831933, | |
| "grad_norm": 0.12602274784345757, | |
| "learning_rate": 1.043613707165109e-05, | |
| "loss": 0.2514, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.915966386554622, | |
| "grad_norm": 0.16613194488301022, | |
| "learning_rate": 1.0046728971962617e-05, | |
| "loss": 0.2507, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 4.957983193277311, | |
| "grad_norm": 0.12043993253574951, | |
| "learning_rate": 9.657320872274144e-06, | |
| "loss": 0.2524, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.1359154617198902, | |
| "learning_rate": 9.267912772585669e-06, | |
| "loss": 0.2525, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 5.042016806722689, | |
| "grad_norm": 0.1645526234919548, | |
| "learning_rate": 8.878504672897196e-06, | |
| "loss": 0.2349, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.084033613445378, | |
| "grad_norm": 0.13565331505255587, | |
| "learning_rate": 8.489096573208722e-06, | |
| "loss": 0.226, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 5.126050420168067, | |
| "grad_norm": 0.12112957276832616, | |
| "learning_rate": 8.099688473520249e-06, | |
| "loss": 0.2323, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.168067226890757, | |
| "grad_norm": 0.13228103496620836, | |
| "learning_rate": 7.710280373831776e-06, | |
| "loss": 0.2333, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 5.2100840336134455, | |
| "grad_norm": 0.12755621341157677, | |
| "learning_rate": 7.3208722741433015e-06, | |
| "loss": 0.2281, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.2521008403361344, | |
| "grad_norm": 0.13414945242875928, | |
| "learning_rate": 6.931464174454828e-06, | |
| "loss": 0.2321, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 5.294117647058823, | |
| "grad_norm": 0.12052434335169325, | |
| "learning_rate": 6.542056074766355e-06, | |
| "loss": 0.2349, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.336134453781512, | |
| "grad_norm": 0.1251627138575762, | |
| "learning_rate": 6.1526479750778815e-06, | |
| "loss": 0.2343, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 5.378151260504202, | |
| "grad_norm": 0.1269368352207133, | |
| "learning_rate": 5.763239875389408e-06, | |
| "loss": 0.2336, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.420168067226891, | |
| "grad_norm": 0.12304767267974673, | |
| "learning_rate": 5.373831775700935e-06, | |
| "loss": 0.2279, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 5.46218487394958, | |
| "grad_norm": 0.12126938306506918, | |
| "learning_rate": 4.9844236760124615e-06, | |
| "loss": 0.2325, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.504201680672269, | |
| "grad_norm": 0.12430095329916237, | |
| "learning_rate": 4.595015576323987e-06, | |
| "loss": 0.2308, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 5.546218487394958, | |
| "grad_norm": 0.12206854649940266, | |
| "learning_rate": 4.205607476635514e-06, | |
| "loss": 0.2303, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.588235294117647, | |
| "grad_norm": 0.12336264199284398, | |
| "learning_rate": 3.816199376947041e-06, | |
| "loss": 0.2339, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 5.630252100840336, | |
| "grad_norm": 0.11965482655647582, | |
| "learning_rate": 3.426791277258567e-06, | |
| "loss": 0.2303, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.6722689075630255, | |
| "grad_norm": 0.11594681549175202, | |
| "learning_rate": 3.0373831775700936e-06, | |
| "loss": 0.2288, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 0.1194663204669251, | |
| "learning_rate": 2.64797507788162e-06, | |
| "loss": 0.2334, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.756302521008403, | |
| "grad_norm": 0.11767398574784753, | |
| "learning_rate": 2.2585669781931465e-06, | |
| "loss": 0.231, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 5.798319327731092, | |
| "grad_norm": 0.11390297708321843, | |
| "learning_rate": 1.8691588785046728e-06, | |
| "loss": 0.2344, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.840336134453781, | |
| "grad_norm": 0.12442153502768896, | |
| "learning_rate": 1.4797507788161994e-06, | |
| "loss": 0.2334, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.11827825723525778, | |
| "learning_rate": 1.0903426791277259e-06, | |
| "loss": 0.2334, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.92436974789916, | |
| "grad_norm": 0.12132650089651566, | |
| "learning_rate": 7.009345794392523e-07, | |
| "loss": 0.2351, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 5.966386554621849, | |
| "grad_norm": 0.11532246430168183, | |
| "learning_rate": 3.1152647975077885e-07, | |
| "loss": 0.2337, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 714, | |
| "total_flos": 2.5527473493612954e+17, | |
| "train_loss": 0.3212856304745714, | |
| "train_runtime": 7845.5794, | |
| "train_samples_per_second": 1.451, | |
| "train_steps_per_second": 0.091 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 714, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.5527473493612954e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |