| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.564102564102564, |
| "eval_steps": 500, |
| "global_step": 5400, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.004748338081671415, |
| "grad_norm": 17.248645403121134, |
| "learning_rate": 8.547008547008548e-07, |
| "loss": 8.2157, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.00949667616334283, |
| "grad_norm": 16.203890965350034, |
| "learning_rate": 1.8043684710351376e-06, |
| "loss": 8.1071, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.014245014245014245, |
| "grad_norm": 8.493709023324136, |
| "learning_rate": 2.7540360873694207e-06, |
| "loss": 7.5503, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01899335232668566, |
| "grad_norm": 4.353558153638053, |
| "learning_rate": 3.7037037037037037e-06, |
| "loss": 6.8809, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.023741690408357077, |
| "grad_norm": 2.9132436951385614, |
| "learning_rate": 4.653371320037987e-06, |
| "loss": 6.4378, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.02849002849002849, |
| "grad_norm": 2.833505754111567, |
| "learning_rate": 5.60303893637227e-06, |
| "loss": 6.1219, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.03323836657169991, |
| "grad_norm": 4.030480311977539, |
| "learning_rate": 6.5527065527065525e-06, |
| "loss": 5.7708, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.03798670465337132, |
| "grad_norm": 4.0022029355083495, |
| "learning_rate": 7.5023741690408365e-06, |
| "loss": 5.2108, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.042735042735042736, |
| "grad_norm": 5.980918003098927, |
| "learning_rate": 8.452041785375119e-06, |
| "loss": 4.5973, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.04748338081671415, |
| "grad_norm": 4.152083496320786, |
| "learning_rate": 9.401709401709402e-06, |
| "loss": 4.1132, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.052231718898385564, |
| "grad_norm": 4.254138168293581, |
| "learning_rate": 1.0351377018043685e-05, |
| "loss": 3.7444, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.05698005698005698, |
| "grad_norm": 4.471875030151703, |
| "learning_rate": 1.1301044634377968e-05, |
| "loss": 3.4974, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.06172839506172839, |
| "grad_norm": 7.773025633603853, |
| "learning_rate": 1.2250712250712251e-05, |
| "loss": 3.2938, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.06647673314339982, |
| "grad_norm": 7.604774713019628, |
| "learning_rate": 1.3200379867046534e-05, |
| "loss": 3.1239, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.07122507122507123, |
| "grad_norm": 6.717865327866952, |
| "learning_rate": 1.4150047483380819e-05, |
| "loss": 2.9924, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.07597340930674264, |
| "grad_norm": 7.517202713477607, |
| "learning_rate": 1.50997150997151e-05, |
| "loss": 2.8674, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.08072174738841406, |
| "grad_norm": 7.174708019589993, |
| "learning_rate": 1.604938271604938e-05, |
| "loss": 2.7716, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.08547008547008547, |
| "grad_norm": 9.240996998590784, |
| "learning_rate": 1.6999050332383668e-05, |
| "loss": 2.6567, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.09021842355175688, |
| "grad_norm": 15.526460622107155, |
| "learning_rate": 1.794871794871795e-05, |
| "loss": 2.5825, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.0949667616334283, |
| "grad_norm": 15.199569893771704, |
| "learning_rate": 1.889838556505223e-05, |
| "loss": 2.5226, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.09971509971509972, |
| "grad_norm": 14.290463569494644, |
| "learning_rate": 1.9848053181386517e-05, |
| "loss": 2.4532, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.10446343779677113, |
| "grad_norm": 7.6270682138007935, |
| "learning_rate": 2.07977207977208e-05, |
| "loss": 2.3895, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.10921177587844255, |
| "grad_norm": 10.674148327286458, |
| "learning_rate": 2.174738841405508e-05, |
| "loss": 2.3529, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.11396011396011396, |
| "grad_norm": 9.855868748637448, |
| "learning_rate": 2.2697056030389367e-05, |
| "loss": 2.2858, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.11870845204178537, |
| "grad_norm": 7.717891137745923, |
| "learning_rate": 2.364672364672365e-05, |
| "loss": 2.2221, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.12345679012345678, |
| "grad_norm": 17.019721376573923, |
| "learning_rate": 2.459639126305793e-05, |
| "loss": 2.1835, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.1282051282051282, |
| "grad_norm": 21.23562867272186, |
| "learning_rate": 2.5546058879392216e-05, |
| "loss": 2.1316, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.13295346628679963, |
| "grad_norm": 10.920996863882513, |
| "learning_rate": 2.64957264957265e-05, |
| "loss": 2.0652, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.13770180436847104, |
| "grad_norm": 20.344957211794714, |
| "learning_rate": 2.744539411206078e-05, |
| "loss": 2.0128, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.14245014245014245, |
| "grad_norm": 13.419017826969121, |
| "learning_rate": 2.839506172839506e-05, |
| "loss": 1.9634, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.14719848053181386, |
| "grad_norm": 10.214748322650372, |
| "learning_rate": 2.9344729344729345e-05, |
| "loss": 1.8695, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.15194681861348527, |
| "grad_norm": 7.322544825964267, |
| "learning_rate": 3.029439696106363e-05, |
| "loss": 1.7607, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.15669515669515668, |
| "grad_norm": 6.452213529244279, |
| "learning_rate": 3.1244064577397914e-05, |
| "loss": 1.6477, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.16144349477682812, |
| "grad_norm": 7.47327345145978, |
| "learning_rate": 3.2193732193732194e-05, |
| "loss": 1.5519, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.16619183285849953, |
| "grad_norm": 8.197373649001618, |
| "learning_rate": 3.314339981006648e-05, |
| "loss": 1.4507, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.17094017094017094, |
| "grad_norm": 8.160811253827458, |
| "learning_rate": 3.409306742640076e-05, |
| "loss": 1.3809, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.17568850902184235, |
| "grad_norm": 7.9644429009331255, |
| "learning_rate": 3.504273504273504e-05, |
| "loss": 1.3325, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.18043684710351376, |
| "grad_norm": 6.025499510848279, |
| "learning_rate": 3.599240265906933e-05, |
| "loss": 1.2776, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.18518518518518517, |
| "grad_norm": 6.334520446256984, |
| "learning_rate": 3.694207027540361e-05, |
| "loss": 1.2107, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.1899335232668566, |
| "grad_norm": 4.974486502991656, |
| "learning_rate": 3.789173789173789e-05, |
| "loss": 1.1731, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.19468186134852802, |
| "grad_norm": 6.3356011104722745, |
| "learning_rate": 3.884140550807218e-05, |
| "loss": 1.141, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.19943019943019943, |
| "grad_norm": 4.475069523356231, |
| "learning_rate": 3.979107312440646e-05, |
| "loss": 1.1016, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.20417853751187084, |
| "grad_norm": 5.061513544694345, |
| "learning_rate": 4.074074074074074e-05, |
| "loss": 1.0632, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.20892687559354226, |
| "grad_norm": 5.894917124836078, |
| "learning_rate": 4.169040835707503e-05, |
| "loss": 1.0386, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.21367521367521367, |
| "grad_norm": 4.92327172504983, |
| "learning_rate": 4.264007597340931e-05, |
| "loss": 1.0105, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.2184235517568851, |
| "grad_norm": 4.6622262222129685, |
| "learning_rate": 4.358974358974359e-05, |
| "loss": 0.9809, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.22317188983855651, |
| "grad_norm": 4.20310786757654, |
| "learning_rate": 4.453941120607788e-05, |
| "loss": 0.9464, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.22792022792022792, |
| "grad_norm": 4.388997458414462, |
| "learning_rate": 4.548907882241216e-05, |
| "loss": 0.9358, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.23266856600189934, |
| "grad_norm": 4.819602636786309, |
| "learning_rate": 4.643874643874644e-05, |
| "loss": 0.9076, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.23741690408357075, |
| "grad_norm": 5.258121230630652, |
| "learning_rate": 4.738841405508073e-05, |
| "loss": 0.8798, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.24216524216524216, |
| "grad_norm": 5.661296053094604, |
| "learning_rate": 4.833808167141501e-05, |
| "loss": 0.8736, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.24691358024691357, |
| "grad_norm": 3.1947628862609783, |
| "learning_rate": 4.928774928774929e-05, |
| "loss": 0.8544, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.251661918328585, |
| "grad_norm": 3.833365046332591, |
| "learning_rate": 5.0237416904083576e-05, |
| "loss": 0.8249, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.2564102564102564, |
| "grad_norm": 4.05762453093816, |
| "learning_rate": 5.118708452041786e-05, |
| "loss": 0.8087, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.2611585944919278, |
| "grad_norm": 3.8528661470158596, |
| "learning_rate": 5.213675213675214e-05, |
| "loss": 0.7804, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.26590693257359926, |
| "grad_norm": 3.754862594736756, |
| "learning_rate": 5.308641975308642e-05, |
| "loss": 0.763, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.2706552706552707, |
| "grad_norm": 4.776944298014696, |
| "learning_rate": 5.403608736942071e-05, |
| "loss": 0.7421, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.2754036087369421, |
| "grad_norm": 4.880864301113887, |
| "learning_rate": 5.498575498575499e-05, |
| "loss": 0.7389, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.2801519468186135, |
| "grad_norm": 4.490459115690417, |
| "learning_rate": 5.5935422602089274e-05, |
| "loss": 0.7258, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.2849002849002849, |
| "grad_norm": 3.131984303584432, |
| "learning_rate": 5.6885090218423554e-05, |
| "loss": 0.6994, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.2896486229819563, |
| "grad_norm": 3.912135876860296, |
| "learning_rate": 5.7834757834757834e-05, |
| "loss": 0.6759, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.2943969610636277, |
| "grad_norm": 3.2548167681023257, |
| "learning_rate": 5.878442545109212e-05, |
| "loss": 0.6596, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.29914529914529914, |
| "grad_norm": 4.44834915161039, |
| "learning_rate": 5.97340930674264e-05, |
| "loss": 0.6413, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.30389363722697055, |
| "grad_norm": 2.7086867568987545, |
| "learning_rate": 6.068376068376068e-05, |
| "loss": 0.6219, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.30864197530864196, |
| "grad_norm": 3.044330352436113, |
| "learning_rate": 6.163342830009497e-05, |
| "loss": 0.5995, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.31339031339031337, |
| "grad_norm": 2.742378672440347, |
| "learning_rate": 6.258309591642925e-05, |
| "loss": 0.5983, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.3181386514719848, |
| "grad_norm": 3.1530287605674503, |
| "learning_rate": 6.353276353276355e-05, |
| "loss": 0.5694, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.32288698955365625, |
| "grad_norm": 3.077851663993601, |
| "learning_rate": 6.448243114909783e-05, |
| "loss": 0.5557, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.32763532763532766, |
| "grad_norm": 3.6604315251059347, |
| "learning_rate": 6.54320987654321e-05, |
| "loss": 0.5539, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.33238366571699907, |
| "grad_norm": 2.24466549736577, |
| "learning_rate": 6.638176638176638e-05, |
| "loss": 0.5271, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.3371320037986705, |
| "grad_norm": 2.033177624969655, |
| "learning_rate": 6.733143399810066e-05, |
| "loss": 0.516, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.3418803418803419, |
| "grad_norm": 2.363255788796514, |
| "learning_rate": 6.828110161443494e-05, |
| "loss": 0.4967, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.3466286799620133, |
| "grad_norm": 2.0837207484976688, |
| "learning_rate": 6.923076923076924e-05, |
| "loss": 0.4839, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.3513770180436847, |
| "grad_norm": 2.234695826275351, |
| "learning_rate": 7.018043684710352e-05, |
| "loss": 0.4607, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.3561253561253561, |
| "grad_norm": 2.801873920520953, |
| "learning_rate": 7.11301044634378e-05, |
| "loss": 0.461, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.36087369420702753, |
| "grad_norm": 1.9907968806919822, |
| "learning_rate": 7.207977207977208e-05, |
| "loss": 0.4357, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.36562203228869894, |
| "grad_norm": 2.499886388553699, |
| "learning_rate": 7.302943969610636e-05, |
| "loss": 0.4256, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.37037037037037035, |
| "grad_norm": 1.9986161794388067, |
| "learning_rate": 7.397910731244065e-05, |
| "loss": 0.4143, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.37511870845204176, |
| "grad_norm": 1.728896975920789, |
| "learning_rate": 7.492877492877494e-05, |
| "loss": 0.3973, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.3798670465337132, |
| "grad_norm": 1.9719574176621415, |
| "learning_rate": 7.587844254510922e-05, |
| "loss": 0.3937, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.38461538461538464, |
| "grad_norm": 2.240521607884822, |
| "learning_rate": 7.68281101614435e-05, |
| "loss": 0.3774, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.38936372269705605, |
| "grad_norm": 1.9946921567513152, |
| "learning_rate": 7.777777777777778e-05, |
| "loss": 0.3746, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.39411206077872746, |
| "grad_norm": 1.8191712831296742, |
| "learning_rate": 7.872744539411206e-05, |
| "loss": 0.3649, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.39886039886039887, |
| "grad_norm": 1.562981286179522, |
| "learning_rate": 7.967711301044634e-05, |
| "loss": 0.3541, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.4036087369420703, |
| "grad_norm": 1.3239788781887885, |
| "learning_rate": 8.062678062678063e-05, |
| "loss": 0.3427, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.4083570750237417, |
| "grad_norm": 1.5554615167125585, |
| "learning_rate": 8.157644824311491e-05, |
| "loss": 0.3327, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.4131054131054131, |
| "grad_norm": 1.7007522642883415, |
| "learning_rate": 8.25261158594492e-05, |
| "loss": 0.3196, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.4178537511870845, |
| "grad_norm": 1.587943196723503, |
| "learning_rate": 8.347578347578347e-05, |
| "loss": 0.3183, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.4226020892687559, |
| "grad_norm": 1.6063882915726966, |
| "learning_rate": 8.442545109211775e-05, |
| "loss": 0.3073, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.42735042735042733, |
| "grad_norm": 1.138249479733028, |
| "learning_rate": 8.537511870845205e-05, |
| "loss": 0.2943, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.43209876543209874, |
| "grad_norm": 1.4448964722707505, |
| "learning_rate": 8.632478632478634e-05, |
| "loss": 0.2942, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.4368471035137702, |
| "grad_norm": 1.4120258045855991, |
| "learning_rate": 8.727445394112062e-05, |
| "loss": 0.2851, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.4415954415954416, |
| "grad_norm": 1.3012296999414996, |
| "learning_rate": 8.82241215574549e-05, |
| "loss": 0.2764, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.44634377967711303, |
| "grad_norm": 1.4271339311302704, |
| "learning_rate": 8.917378917378918e-05, |
| "loss": 0.2671, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.45109211775878444, |
| "grad_norm": 1.3694712277412053, |
| "learning_rate": 9.012345679012346e-05, |
| "loss": 0.2613, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.45584045584045585, |
| "grad_norm": 1.1953213042525168, |
| "learning_rate": 9.107312440645774e-05, |
| "loss": 0.2601, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.46058879392212726, |
| "grad_norm": 1.3055440465284454, |
| "learning_rate": 9.202279202279203e-05, |
| "loss": 0.2514, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.46533713200379867, |
| "grad_norm": 1.0536182232036395, |
| "learning_rate": 9.297245963912631e-05, |
| "loss": 0.247, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.4700854700854701, |
| "grad_norm": 1.2195241432373358, |
| "learning_rate": 9.392212725546059e-05, |
| "loss": 0.2344, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.4748338081671415, |
| "grad_norm": 1.0959224723274623, |
| "learning_rate": 9.487179487179487e-05, |
| "loss": 0.2396, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.4795821462488129, |
| "grad_norm": 0.8580238248589764, |
| "learning_rate": 9.582146248812915e-05, |
| "loss": 0.2287, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.4843304843304843, |
| "grad_norm": 1.0947613479475533, |
| "learning_rate": 9.677113010446344e-05, |
| "loss": 0.2238, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.4890788224121557, |
| "grad_norm": 1.0156537397126817, |
| "learning_rate": 9.772079772079774e-05, |
| "loss": 0.2234, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.49382716049382713, |
| "grad_norm": 0.9701117287600136, |
| "learning_rate": 9.867046533713202e-05, |
| "loss": 0.2163, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.4985754985754986, |
| "grad_norm": 0.8505350855055186, |
| "learning_rate": 9.96201329534663e-05, |
| "loss": 0.2104, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.50332383665717, |
| "grad_norm": 0.919541770109071, |
| "learning_rate": 9.999990109907572e-05, |
| "loss": 0.2057, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.5080721747388414, |
| "grad_norm": 0.8070940788803623, |
| "learning_rate": 9.999929670595522e-05, |
| "loss": 0.1982, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.5128205128205128, |
| "grad_norm": 0.7862523916413411, |
| "learning_rate": 9.99981428713058e-05, |
| "loss": 0.1961, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.5175688509021842, |
| "grad_norm": 0.6433390539036313, |
| "learning_rate": 9.999643960780693e-05, |
| "loss": 0.1921, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.5223171889838556, |
| "grad_norm": 0.7846731950775512, |
| "learning_rate": 9.999418693417573e-05, |
| "loss": 0.1864, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.5270655270655271, |
| "grad_norm": 0.6576054537041652, |
| "learning_rate": 9.999138487516681e-05, |
| "loss": 0.1819, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.5318138651471985, |
| "grad_norm": 0.7485461445273004, |
| "learning_rate": 9.998803346157195e-05, |
| "loss": 0.1809, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.5365622032288699, |
| "grad_norm": 0.9279248523360117, |
| "learning_rate": 9.99841327302198e-05, |
| "loss": 0.1768, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.5413105413105413, |
| "grad_norm": 0.6039316478015732, |
| "learning_rate": 9.997968272397542e-05, |
| "loss": 0.1708, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.5460588793922128, |
| "grad_norm": 0.5957854675964963, |
| "learning_rate": 9.997468349173985e-05, |
| "loss": 0.1709, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.5508072174738842, |
| "grad_norm": 0.6250421837563528, |
| "learning_rate": 9.99691350884496e-05, |
| "loss": 0.1669, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.5555555555555556, |
| "grad_norm": 0.6211554754758593, |
| "learning_rate": 9.996303757507602e-05, |
| "loss": 0.163, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.560303893637227, |
| "grad_norm": 0.5693317753143289, |
| "learning_rate": 9.995639101862453e-05, |
| "loss": 0.1647, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.5650522317188984, |
| "grad_norm": 0.5341669223258995, |
| "learning_rate": 9.994919549213408e-05, |
| "loss": 0.1618, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.5698005698005698, |
| "grad_norm": 0.6322078334748438, |
| "learning_rate": 9.99414510746762e-05, |
| "loss": 0.1578, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.5745489078822412, |
| "grad_norm": 0.6461469440468907, |
| "learning_rate": 9.993315785135416e-05, |
| "loss": 0.1578, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.5792972459639126, |
| "grad_norm": 0.6703579359938157, |
| "learning_rate": 9.99243159133021e-05, |
| "loss": 0.1556, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.584045584045584, |
| "grad_norm": 0.4522524280791954, |
| "learning_rate": 9.991492535768392e-05, |
| "loss": 0.1533, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.5887939221272555, |
| "grad_norm": 0.6262783617044025, |
| "learning_rate": 9.99049862876923e-05, |
| "loss": 0.1535, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.5935422602089269, |
| "grad_norm": 0.5639085005723359, |
| "learning_rate": 9.989449881254754e-05, |
| "loss": 0.1521, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.5982905982905983, |
| "grad_norm": 0.602521609513433, |
| "learning_rate": 9.988346304749635e-05, |
| "loss": 0.1496, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.6030389363722697, |
| "grad_norm": 0.5984313589183607, |
| "learning_rate": 9.987187911381059e-05, |
| "loss": 0.1483, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.6077872744539411, |
| "grad_norm": 0.5704476522163662, |
| "learning_rate": 9.985974713878592e-05, |
| "loss": 0.1447, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.6125356125356125, |
| "grad_norm": 0.5120246871842873, |
| "learning_rate": 9.984706725574044e-05, |
| "loss": 0.1442, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.6172839506172839, |
| "grad_norm": 0.4423173770677043, |
| "learning_rate": 9.983383960401319e-05, |
| "loss": 0.1413, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.6220322886989553, |
| "grad_norm": 0.5447233964358744, |
| "learning_rate": 9.982006432896266e-05, |
| "loss": 0.1443, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.6267806267806267, |
| "grad_norm": 0.47488686505078226, |
| "learning_rate": 9.980574158196512e-05, |
| "loss": 0.1403, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.6315289648622981, |
| "grad_norm": 0.39994838809351746, |
| "learning_rate": 9.979087152041304e-05, |
| "loss": 0.1375, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.6362773029439696, |
| "grad_norm": 0.4486514324629703, |
| "learning_rate": 9.977545430771332e-05, |
| "loss": 0.1372, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.6410256410256411, |
| "grad_norm": 0.4215625682082093, |
| "learning_rate": 9.975949011328542e-05, |
| "loss": 0.1363, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.6457739791073125, |
| "grad_norm": 0.4556940782689893, |
| "learning_rate": 9.97429791125597e-05, |
| "loss": 0.1359, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.6505223171889839, |
| "grad_norm": 0.4571325450394861, |
| "learning_rate": 9.972592148697528e-05, |
| "loss": 0.134, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.6552706552706553, |
| "grad_norm": 0.45728914817916727, |
| "learning_rate": 9.970831742397815e-05, |
| "loss": 0.1336, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.6600189933523267, |
| "grad_norm": 0.4164566682304999, |
| "learning_rate": 9.969016711701908e-05, |
| "loss": 0.1317, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.6647673314339981, |
| "grad_norm": 0.7369784799950742, |
| "learning_rate": 9.967147076555157e-05, |
| "loss": 0.1327, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.6695156695156695, |
| "grad_norm": 0.42226787634153845, |
| "learning_rate": 9.965222857502951e-05, |
| "loss": 0.1309, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.674264007597341, |
| "grad_norm": 0.4860967833723572, |
| "learning_rate": 9.963244075690505e-05, |
| "loss": 0.1301, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.6790123456790124, |
| "grad_norm": 0.4131434714198591, |
| "learning_rate": 9.961210752862626e-05, |
| "loss": 0.129, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.6837606837606838, |
| "grad_norm": 0.3876119109435019, |
| "learning_rate": 9.959122911363463e-05, |
| "loss": 0.1286, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.6885090218423552, |
| "grad_norm": 0.395758026911608, |
| "learning_rate": 9.956980574136281e-05, |
| "loss": 0.1276, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.6932573599240266, |
| "grad_norm": 0.40507641316691656, |
| "learning_rate": 9.954783764723188e-05, |
| "loss": 0.1264, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.698005698005698, |
| "grad_norm": 0.42457278824168765, |
| "learning_rate": 9.952532507264891e-05, |
| "loss": 0.125, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.7027540360873694, |
| "grad_norm": 0.42342402364455467, |
| "learning_rate": 9.950226826500426e-05, |
| "loss": 0.1246, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.7075023741690408, |
| "grad_norm": 0.4522016176700106, |
| "learning_rate": 9.947866747766882e-05, |
| "loss": 0.1242, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.7122507122507122, |
| "grad_norm": 0.40575444762945795, |
| "learning_rate": 9.945452296999131e-05, |
| "loss": 0.1242, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.7169990503323836, |
| "grad_norm": 0.43773549286702973, |
| "learning_rate": 9.942983500729536e-05, |
| "loss": 0.1231, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.7217473884140551, |
| "grad_norm": 0.43494873154948677, |
| "learning_rate": 9.940460386087667e-05, |
| "loss": 0.1191, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.7264957264957265, |
| "grad_norm": 0.4110550234671372, |
| "learning_rate": 9.937882980799986e-05, |
| "loss": 0.1205, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.7312440645773979, |
| "grad_norm": 0.3611439292921592, |
| "learning_rate": 9.935251313189564e-05, |
| "loss": 0.1197, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.7359924026590693, |
| "grad_norm": 0.36979340008153927, |
| "learning_rate": 9.93256541217576e-05, |
| "loss": 0.1204, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.7407407407407407, |
| "grad_norm": 0.4144817599435258, |
| "learning_rate": 9.929825307273898e-05, |
| "loss": 0.1179, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.7454890788224121, |
| "grad_norm": 0.3740043433204569, |
| "learning_rate": 9.927031028594951e-05, |
| "loss": 0.1178, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.7502374169040835, |
| "grad_norm": 0.3832431167907128, |
| "learning_rate": 9.924182606845204e-05, |
| "loss": 0.1157, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.7549857549857549, |
| "grad_norm": 0.4128745103505223, |
| "learning_rate": 9.921280073325922e-05, |
| "loss": 0.1163, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.7597340930674265, |
| "grad_norm": 0.40692556086462084, |
| "learning_rate": 9.918323459933005e-05, |
| "loss": 0.116, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.7644824311490979, |
| "grad_norm": 0.36393810730007625, |
| "learning_rate": 9.915312799156632e-05, |
| "loss": 0.1147, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.7692307692307693, |
| "grad_norm": 0.3549121719387913, |
| "learning_rate": 9.91224812408091e-05, |
| "loss": 0.1138, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.7739791073124407, |
| "grad_norm": 0.3733601076093646, |
| "learning_rate": 9.909129468383506e-05, |
| "loss": 0.1106, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.7787274453941121, |
| "grad_norm": 0.5206917588367477, |
| "learning_rate": 9.905956866335286e-05, |
| "loss": 0.1132, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.7834757834757835, |
| "grad_norm": 0.34786476356129886, |
| "learning_rate": 9.90273035279992e-05, |
| "loss": 0.113, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.7882241215574549, |
| "grad_norm": 0.3609054171968716, |
| "learning_rate": 9.899449963233523e-05, |
| "loss": 0.1122, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.7929724596391263, |
| "grad_norm": 0.3418839042515417, |
| "learning_rate": 9.896115733684244e-05, |
| "loss": 0.1111, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.7977207977207977, |
| "grad_norm": 0.38927046464571985, |
| "learning_rate": 9.892727700791881e-05, |
| "loss": 0.1098, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.8024691358024691, |
| "grad_norm": 0.3368112428945925, |
| "learning_rate": 9.889285901787482e-05, |
| "loss": 0.1106, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.8072174738841406, |
| "grad_norm": 0.3262329941162413, |
| "learning_rate": 9.885790374492921e-05, |
| "loss": 0.1087, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.811965811965812, |
| "grad_norm": 0.3340210979702793, |
| "learning_rate": 9.882241157320497e-05, |
| "loss": 0.1055, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.8167141500474834, |
| "grad_norm": 0.38106884025967774, |
| "learning_rate": 9.878638289272507e-05, |
| "loss": 0.1104, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.8214624881291548, |
| "grad_norm": 0.3665942399638638, |
| "learning_rate": 9.874981809940808e-05, |
| "loss": 0.1076, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.8262108262108262, |
| "grad_norm": 0.3511637217731337, |
| "learning_rate": 9.871271759506404e-05, |
| "loss": 0.1068, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.8309591642924976, |
| "grad_norm": 0.3290291492915372, |
| "learning_rate": 9.86750817873898e-05, |
| "loss": 0.1014, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.835707502374169, |
| "grad_norm": 0.3073088226054363, |
| "learning_rate": 9.863691108996471e-05, |
| "loss": 0.1031, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.8404558404558404, |
| "grad_norm": 0.3579776907985757, |
| "learning_rate": 9.859820592224599e-05, |
| "loss": 0.1052, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.8452041785375118, |
| "grad_norm": 0.2933581562069798, |
| "learning_rate": 9.855896670956415e-05, |
| "loss": 0.1018, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.8499525166191833, |
| "grad_norm": 0.33129373013252383, |
| "learning_rate": 9.851919388311832e-05, |
| "loss": 0.0998, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.8547008547008547, |
| "grad_norm": 0.32334865919183675, |
| "learning_rate": 9.84788878799715e-05, |
| "loss": 0.1003, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.8594491927825261, |
| "grad_norm": 0.3258084610922373, |
| "learning_rate": 9.843804914304577e-05, |
| "loss": 0.102, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.8641975308641975, |
| "grad_norm": 0.36745967245364125, |
| "learning_rate": 9.839667812111739e-05, |
| "loss": 0.1001, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.8689458689458689, |
| "grad_norm": 0.2677606208520182, |
| "learning_rate": 9.835477526881189e-05, |
| "loss": 0.1, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.8736942070275404, |
| "grad_norm": 0.31378317445026904, |
| "learning_rate": 9.831234104659913e-05, |
| "loss": 0.0978, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.8784425451092118, |
| "grad_norm": 0.3670428106522396, |
| "learning_rate": 9.826937592078811e-05, |
| "loss": 0.0958, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.8831908831908832, |
| "grad_norm": 0.3242528709875007, |
| "learning_rate": 9.822588036352201e-05, |
| "loss": 0.0964, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.8879392212725546, |
| "grad_norm": 0.3023903315211582, |
| "learning_rate": 9.81818548527728e-05, |
| "loss": 0.0948, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.8926875593542261, |
| "grad_norm": 0.319357256810946, |
| "learning_rate": 9.81372998723362e-05, |
| "loss": 0.0945, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.8974358974358975, |
| "grad_norm": 0.27594908092048714, |
| "learning_rate": 9.809221591182623e-05, |
| "loss": 0.0927, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.9021842355175689, |
| "grad_norm": 0.32676988957298875, |
| "learning_rate": 9.804660346666984e-05, |
| "loss": 0.096, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.9069325735992403, |
| "grad_norm": 0.3090093573308265, |
| "learning_rate": 9.800046303810154e-05, |
| "loss": 0.091, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.9116809116809117, |
| "grad_norm": 0.35187675840822397, |
| "learning_rate": 9.795379513315779e-05, |
| "loss": 0.093, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.9164292497625831, |
| "grad_norm": 0.38016324065051443, |
| "learning_rate": 9.790660026467153e-05, |
| "loss": 0.0917, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.9211775878442545, |
| "grad_norm": 0.28521547244557494, |
| "learning_rate": 9.785887895126644e-05, |
| "loss": 0.0907, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.9259259259259259, |
| "grad_norm": 0.3419707364445786, |
| "learning_rate": 9.781063171735134e-05, |
| "loss": 0.0912, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.9306742640075973, |
| "grad_norm": 0.27054996100610457, |
| "learning_rate": 9.776185909311437e-05, |
| "loss": 0.0884, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.9354226020892688, |
| "grad_norm": 0.30445805909595236, |
| "learning_rate": 9.771256161451716e-05, |
| "loss": 0.0877, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.9401709401709402, |
| "grad_norm": 0.31041895763678207, |
| "learning_rate": 9.766273982328899e-05, |
| "loss": 0.0871, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.9449192782526116, |
| "grad_norm": 0.2832133032119395, |
| "learning_rate": 9.761239426692077e-05, |
| "loss": 0.0849, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.949667616334283, |
| "grad_norm": 0.2708401169290125, |
| "learning_rate": 9.756152549865907e-05, |
| "loss": 0.0847, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.9544159544159544, |
| "grad_norm": 0.2613767046635369, |
| "learning_rate": 9.751013407750004e-05, |
| "loss": 0.0813, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.9591642924976258, |
| "grad_norm": 0.26373499137957396, |
| "learning_rate": 9.745822056818324e-05, |
| "loss": 0.0838, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.9639126305792972, |
| "grad_norm": 0.27778152029868663, |
| "learning_rate": 9.740578554118545e-05, |
| "loss": 0.0829, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.9686609686609686, |
| "grad_norm": 0.23689767301606995, |
| "learning_rate": 9.735282957271438e-05, |
| "loss": 0.0797, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.97340930674264, |
| "grad_norm": 0.27072882951414534, |
| "learning_rate": 9.729935324470243e-05, |
| "loss": 0.0794, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.9781576448243114, |
| "grad_norm": 0.2315940320021251, |
| "learning_rate": 9.724535714480012e-05, |
| "loss": 0.0801, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.9829059829059829, |
| "grad_norm": 0.2352379745574888, |
| "learning_rate": 9.719084186636984e-05, |
| "loss": 0.0792, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.9876543209876543, |
| "grad_norm": 0.24192660977031283, |
| "learning_rate": 9.713580800847916e-05, |
| "loss": 0.0777, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.9924026590693258, |
| "grad_norm": 0.22901532341530006, |
| "learning_rate": 9.708025617589431e-05, |
| "loss": 0.0785, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.9971509971509972, |
| "grad_norm": 0.2196235931427345, |
| "learning_rate": 9.702418697907357e-05, |
| "loss": 0.0754, |
| "step": 2100 |
| }, |
| { |
| "epoch": 1.0018993352326686, |
| "grad_norm": 0.23048731147797555, |
| "learning_rate": 9.696760103416049e-05, |
| "loss": 0.0766, |
| "step": 2110 |
| }, |
| { |
| "epoch": 1.00664767331434, |
| "grad_norm": 0.2228824692133708, |
| "learning_rate": 9.69104989629772e-05, |
| "loss": 0.0761, |
| "step": 2120 |
| }, |
| { |
| "epoch": 1.0113960113960114, |
| "grad_norm": 0.22505600692470903, |
| "learning_rate": 9.685288139301749e-05, |
| "loss": 0.0739, |
| "step": 2130 |
| }, |
| { |
| "epoch": 1.0161443494776827, |
| "grad_norm": 0.22627209515135713, |
| "learning_rate": 9.679474895744002e-05, |
| "loss": 0.0742, |
| "step": 2140 |
| }, |
| { |
| "epoch": 1.0208926875593543, |
| "grad_norm": 0.24833587458936565, |
| "learning_rate": 9.673610229506123e-05, |
| "loss": 0.0738, |
| "step": 2150 |
| }, |
| { |
| "epoch": 1.0256410256410255, |
| "grad_norm": 0.23891436986960332, |
| "learning_rate": 9.667694205034842e-05, |
| "loss": 0.0716, |
| "step": 2160 |
| }, |
| { |
| "epoch": 1.030389363722697, |
| "grad_norm": 0.2391418635503859, |
| "learning_rate": 9.661726887341264e-05, |
| "loss": 0.0724, |
| "step": 2170 |
| }, |
| { |
| "epoch": 1.0351377018043684, |
| "grad_norm": 0.19037486990848157, |
| "learning_rate": 9.655708342000158e-05, |
| "loss": 0.071, |
| "step": 2180 |
| }, |
| { |
| "epoch": 1.03988603988604, |
| "grad_norm": 0.22776501997885065, |
| "learning_rate": 9.649638635149225e-05, |
| "loss": 0.0725, |
| "step": 2190 |
| }, |
| { |
| "epoch": 1.0446343779677112, |
| "grad_norm": 0.23138809459804194, |
| "learning_rate": 9.643517833488387e-05, |
| "loss": 0.0725, |
| "step": 2200 |
| }, |
| { |
| "epoch": 1.0493827160493827, |
| "grad_norm": 0.20139355430993858, |
| "learning_rate": 9.637346004279038e-05, |
| "loss": 0.0715, |
| "step": 2210 |
| }, |
| { |
| "epoch": 1.0541310541310542, |
| "grad_norm": 0.21270206852322332, |
| "learning_rate": 9.631123215343323e-05, |
| "loss": 0.0728, |
| "step": 2220 |
| }, |
| { |
| "epoch": 1.0588793922127255, |
| "grad_norm": 0.20528549903188958, |
| "learning_rate": 9.624849535063373e-05, |
| "loss": 0.0712, |
| "step": 2230 |
| }, |
| { |
| "epoch": 1.063627730294397, |
| "grad_norm": 0.1897382940052522, |
| "learning_rate": 9.61852503238057e-05, |
| "loss": 0.0713, |
| "step": 2240 |
| }, |
| { |
| "epoch": 1.0683760683760684, |
| "grad_norm": 0.2234014925732005, |
| "learning_rate": 9.612149776794776e-05, |
| "loss": 0.0712, |
| "step": 2250 |
| }, |
| { |
| "epoch": 1.0731244064577399, |
| "grad_norm": 0.20930775390549886, |
| "learning_rate": 9.605723838363586e-05, |
| "loss": 0.0726, |
| "step": 2260 |
| }, |
| { |
| "epoch": 1.0778727445394112, |
| "grad_norm": 0.19480519273536226, |
| "learning_rate": 9.599247287701539e-05, |
| "loss": 0.0704, |
| "step": 2270 |
| }, |
| { |
| "epoch": 1.0826210826210827, |
| "grad_norm": 0.1862466434584284, |
| "learning_rate": 9.592720195979353e-05, |
| "loss": 0.0683, |
| "step": 2280 |
| }, |
| { |
| "epoch": 1.087369420702754, |
| "grad_norm": 0.22669475744836864, |
| "learning_rate": 9.586142634923141e-05, |
| "loss": 0.069, |
| "step": 2290 |
| }, |
| { |
| "epoch": 1.0921177587844255, |
| "grad_norm": 0.23313464762805208, |
| "learning_rate": 9.579514676813627e-05, |
| "loss": 0.0685, |
| "step": 2300 |
| }, |
| { |
| "epoch": 1.0968660968660968, |
| "grad_norm": 0.18978697732394295, |
| "learning_rate": 9.572836394485345e-05, |
| "loss": 0.0686, |
| "step": 2310 |
| }, |
| { |
| "epoch": 1.1016144349477683, |
| "grad_norm": 0.21821159143827515, |
| "learning_rate": 9.566107861325836e-05, |
| "loss": 0.0705, |
| "step": 2320 |
| }, |
| { |
| "epoch": 1.1063627730294396, |
| "grad_norm": 0.2085391811587992, |
| "learning_rate": 9.559329151274854e-05, |
| "loss": 0.069, |
| "step": 2330 |
| }, |
| { |
| "epoch": 1.1111111111111112, |
| "grad_norm": 0.21344220210497813, |
| "learning_rate": 9.552500338823543e-05, |
| "loss": 0.0665, |
| "step": 2340 |
| }, |
| { |
| "epoch": 1.1158594491927825, |
| "grad_norm": 0.18558671035053925, |
| "learning_rate": 9.54562149901362e-05, |
| "loss": 0.0666, |
| "step": 2350 |
| }, |
| { |
| "epoch": 1.120607787274454, |
| "grad_norm": 0.18194222094420473, |
| "learning_rate": 9.538692707436554e-05, |
| "loss": 0.0682, |
| "step": 2360 |
| }, |
| { |
| "epoch": 1.1253561253561253, |
| "grad_norm": 0.17508086615169313, |
| "learning_rate": 9.531714040232733e-05, |
| "loss": 0.0664, |
| "step": 2370 |
| }, |
| { |
| "epoch": 1.1301044634377968, |
| "grad_norm": 0.18092813445289843, |
| "learning_rate": 9.524685574090627e-05, |
| "loss": 0.0688, |
| "step": 2380 |
| }, |
| { |
| "epoch": 1.134852801519468, |
| "grad_norm": 0.17940386070656392, |
| "learning_rate": 9.517607386245948e-05, |
| "loss": 0.0671, |
| "step": 2390 |
| }, |
| { |
| "epoch": 1.1396011396011396, |
| "grad_norm": 0.18207603099039493, |
| "learning_rate": 9.510479554480793e-05, |
| "loss": 0.0666, |
| "step": 2400 |
| }, |
| { |
| "epoch": 1.144349477682811, |
| "grad_norm": 0.20066035832261697, |
| "learning_rate": 9.503302157122804e-05, |
| "loss": 0.0666, |
| "step": 2410 |
| }, |
| { |
| "epoch": 1.1490978157644824, |
| "grad_norm": 0.16723923818656553, |
| "learning_rate": 9.496075273044293e-05, |
| "loss": 0.0662, |
| "step": 2420 |
| }, |
| { |
| "epoch": 1.1538461538461537, |
| "grad_norm": 0.2011658938828226, |
| "learning_rate": 9.488798981661379e-05, |
| "loss": 0.068, |
| "step": 2430 |
| }, |
| { |
| "epoch": 1.1585944919278253, |
| "grad_norm": 0.20043398022969394, |
| "learning_rate": 9.481473362933122e-05, |
| "loss": 0.0679, |
| "step": 2440 |
| }, |
| { |
| "epoch": 1.1633428300094968, |
| "grad_norm": 0.17742709223664174, |
| "learning_rate": 9.474098497360637e-05, |
| "loss": 0.0653, |
| "step": 2450 |
| }, |
| { |
| "epoch": 1.168091168091168, |
| "grad_norm": 0.19341031326710123, |
| "learning_rate": 9.466674465986211e-05, |
| "loss": 0.0648, |
| "step": 2460 |
| }, |
| { |
| "epoch": 1.1728395061728394, |
| "grad_norm": 0.17908760371671564, |
| "learning_rate": 9.459201350392418e-05, |
| "loss": 0.0665, |
| "step": 2470 |
| }, |
| { |
| "epoch": 1.177587844254511, |
| "grad_norm": 0.21541654966298862, |
| "learning_rate": 9.451679232701212e-05, |
| "loss": 0.0665, |
| "step": 2480 |
| }, |
| { |
| "epoch": 1.1823361823361824, |
| "grad_norm": 0.20145655577101054, |
| "learning_rate": 9.444108195573031e-05, |
| "loss": 0.067, |
| "step": 2490 |
| }, |
| { |
| "epoch": 1.1870845204178537, |
| "grad_norm": 0.17902372874871034, |
| "learning_rate": 9.436488322205892e-05, |
| "loss": 0.0657, |
| "step": 2500 |
| }, |
| { |
| "epoch": 1.1918328584995252, |
| "grad_norm": 0.20936093359608424, |
| "learning_rate": 9.42881969633447e-05, |
| "loss": 0.0685, |
| "step": 2510 |
| }, |
| { |
| "epoch": 1.1965811965811965, |
| "grad_norm": 0.17428770499256205, |
| "learning_rate": 9.42110240222918e-05, |
| "loss": 0.0663, |
| "step": 2520 |
| }, |
| { |
| "epoch": 1.201329534662868, |
| "grad_norm": 0.18156595785306356, |
| "learning_rate": 9.413336524695253e-05, |
| "loss": 0.0654, |
| "step": 2530 |
| }, |
| { |
| "epoch": 1.2060778727445394, |
| "grad_norm": 0.16374433991886897, |
| "learning_rate": 9.405522149071801e-05, |
| "loss": 0.0657, |
| "step": 2540 |
| }, |
| { |
| "epoch": 1.210826210826211, |
| "grad_norm": 0.1683090791229423, |
| "learning_rate": 9.397659361230883e-05, |
| "loss": 0.0647, |
| "step": 2550 |
| }, |
| { |
| "epoch": 1.2155745489078822, |
| "grad_norm": 0.19135168316533313, |
| "learning_rate": 9.389748247576557e-05, |
| "loss": 0.0639, |
| "step": 2560 |
| }, |
| { |
| "epoch": 1.2203228869895537, |
| "grad_norm": 0.1740777181027489, |
| "learning_rate": 9.381788895043933e-05, |
| "loss": 0.0656, |
| "step": 2570 |
| }, |
| { |
| "epoch": 1.225071225071225, |
| "grad_norm": 0.19364478652869332, |
| "learning_rate": 9.373781391098218e-05, |
| "loss": 0.0655, |
| "step": 2580 |
| }, |
| { |
| "epoch": 1.2298195631528965, |
| "grad_norm": 0.18366674584797033, |
| "learning_rate": 9.365725823733753e-05, |
| "loss": 0.0645, |
| "step": 2590 |
| }, |
| { |
| "epoch": 1.2345679012345678, |
| "grad_norm": 0.17450561227147082, |
| "learning_rate": 9.35762228147305e-05, |
| "loss": 0.0647, |
| "step": 2600 |
| }, |
| { |
| "epoch": 1.2393162393162394, |
| "grad_norm": 0.19302134951793246, |
| "learning_rate": 9.349470853365815e-05, |
| "loss": 0.0642, |
| "step": 2610 |
| }, |
| { |
| "epoch": 1.2440645773979107, |
| "grad_norm": 0.1660731917372023, |
| "learning_rate": 9.341271628987968e-05, |
| "loss": 0.0635, |
| "step": 2620 |
| }, |
| { |
| "epoch": 1.2488129154795822, |
| "grad_norm": 0.19282598054436592, |
| "learning_rate": 9.333024698440667e-05, |
| "loss": 0.0655, |
| "step": 2630 |
| }, |
| { |
| "epoch": 1.2535612535612537, |
| "grad_norm": 0.17366361594832658, |
| "learning_rate": 9.324730152349306e-05, |
| "loss": 0.0635, |
| "step": 2640 |
| }, |
| { |
| "epoch": 1.258309591642925, |
| "grad_norm": 0.19007023594974443, |
| "learning_rate": 9.31638808186253e-05, |
| "loss": 0.0652, |
| "step": 2650 |
| }, |
| { |
| "epoch": 1.2630579297245963, |
| "grad_norm": 1.1024382434197253, |
| "learning_rate": 9.307998578651228e-05, |
| "loss": 0.0772, |
| "step": 2660 |
| }, |
| { |
| "epoch": 1.2678062678062678, |
| "grad_norm": 0.21586082771106202, |
| "learning_rate": 9.299561734907528e-05, |
| "loss": 0.07, |
| "step": 2670 |
| }, |
| { |
| "epoch": 1.2725546058879393, |
| "grad_norm": 0.17722272954110585, |
| "learning_rate": 9.291077643343776e-05, |
| "loss": 0.066, |
| "step": 2680 |
| }, |
| { |
| "epoch": 1.2773029439696106, |
| "grad_norm": 0.17539447588198026, |
| "learning_rate": 9.282546397191531e-05, |
| "loss": 0.0645, |
| "step": 2690 |
| }, |
| { |
| "epoch": 1.282051282051282, |
| "grad_norm": 0.18796071657065863, |
| "learning_rate": 9.273968090200533e-05, |
| "loss": 0.0626, |
| "step": 2700 |
| }, |
| { |
| "epoch": 1.2867996201329535, |
| "grad_norm": 0.21102862789072901, |
| "learning_rate": 9.265342816637666e-05, |
| "loss": 0.0646, |
| "step": 2710 |
| }, |
| { |
| "epoch": 1.291547958214625, |
| "grad_norm": 0.18410753354993722, |
| "learning_rate": 9.256670671285934e-05, |
| "loss": 0.0642, |
| "step": 2720 |
| }, |
| { |
| "epoch": 1.2962962962962963, |
| "grad_norm": 0.16105620612088856, |
| "learning_rate": 9.247951749443413e-05, |
| "loss": 0.0643, |
| "step": 2730 |
| }, |
| { |
| "epoch": 1.3010446343779676, |
| "grad_norm": 0.1750774481025004, |
| "learning_rate": 9.239186146922208e-05, |
| "loss": 0.0634, |
| "step": 2740 |
| }, |
| { |
| "epoch": 1.305792972459639, |
| "grad_norm": 0.17664588539201698, |
| "learning_rate": 9.230373960047394e-05, |
| "loss": 0.0627, |
| "step": 2750 |
| }, |
| { |
| "epoch": 1.3105413105413106, |
| "grad_norm": 0.15729300032537766, |
| "learning_rate": 9.22151528565596e-05, |
| "loss": 0.0609, |
| "step": 2760 |
| }, |
| { |
| "epoch": 1.315289648622982, |
| "grad_norm": 0.15926299646583086, |
| "learning_rate": 9.212610221095748e-05, |
| "loss": 0.0617, |
| "step": 2770 |
| }, |
| { |
| "epoch": 1.3200379867046534, |
| "grad_norm": 0.17043048975727534, |
| "learning_rate": 9.203658864224382e-05, |
| "loss": 0.0635, |
| "step": 2780 |
| }, |
| { |
| "epoch": 1.3247863247863247, |
| "grad_norm": 0.16927971254642155, |
| "learning_rate": 9.194661313408187e-05, |
| "loss": 0.0625, |
| "step": 2790 |
| }, |
| { |
| "epoch": 1.3295346628679963, |
| "grad_norm": 0.16948164587255984, |
| "learning_rate": 9.185617667521117e-05, |
| "loss": 0.0643, |
| "step": 2800 |
| }, |
| { |
| "epoch": 1.3342830009496676, |
| "grad_norm": 0.1683741018197293, |
| "learning_rate": 9.176528025943663e-05, |
| "loss": 0.0634, |
| "step": 2810 |
| }, |
| { |
| "epoch": 1.339031339031339, |
| "grad_norm": 0.18132690281549088, |
| "learning_rate": 9.167392488561758e-05, |
| "loss": 0.0615, |
| "step": 2820 |
| }, |
| { |
| "epoch": 1.3437796771130104, |
| "grad_norm": 0.1748082847286973, |
| "learning_rate": 9.158211155765692e-05, |
| "loss": 0.0623, |
| "step": 2830 |
| }, |
| { |
| "epoch": 1.348528015194682, |
| "grad_norm": 0.16640919415411776, |
| "learning_rate": 9.14898412844899e-05, |
| "loss": 0.0603, |
| "step": 2840 |
| }, |
| { |
| "epoch": 1.3532763532763532, |
| "grad_norm": 0.15919435676570565, |
| "learning_rate": 9.13971150800732e-05, |
| "loss": 0.0632, |
| "step": 2850 |
| }, |
| { |
| "epoch": 1.3580246913580247, |
| "grad_norm": 0.144516155129044, |
| "learning_rate": 9.130393396337368e-05, |
| "loss": 0.0621, |
| "step": 2860 |
| }, |
| { |
| "epoch": 1.362773029439696, |
| "grad_norm": 0.16195510823123485, |
| "learning_rate": 9.121029895835731e-05, |
| "loss": 0.0614, |
| "step": 2870 |
| }, |
| { |
| "epoch": 1.3675213675213675, |
| "grad_norm": 0.16437399859015953, |
| "learning_rate": 9.111621109397769e-05, |
| "loss": 0.0622, |
| "step": 2880 |
| }, |
| { |
| "epoch": 1.3722697056030388, |
| "grad_norm": 0.15604474628280193, |
| "learning_rate": 9.102167140416503e-05, |
| "loss": 0.0602, |
| "step": 2890 |
| }, |
| { |
| "epoch": 1.3770180436847104, |
| "grad_norm": 0.16524727699859193, |
| "learning_rate": 9.092668092781454e-05, |
| "loss": 0.0624, |
| "step": 2900 |
| }, |
| { |
| "epoch": 1.381766381766382, |
| "grad_norm": 0.14437628425830004, |
| "learning_rate": 9.083124070877516e-05, |
| "loss": 0.0622, |
| "step": 2910 |
| }, |
| { |
| "epoch": 1.3865147198480532, |
| "grad_norm": 0.15565283165668614, |
| "learning_rate": 9.073535179583806e-05, |
| "loss": 0.0624, |
| "step": 2920 |
| }, |
| { |
| "epoch": 1.3912630579297245, |
| "grad_norm": 0.16363573021258815, |
| "learning_rate": 9.063901524272505e-05, |
| "loss": 0.0621, |
| "step": 2930 |
| }, |
| { |
| "epoch": 1.396011396011396, |
| "grad_norm": 0.16032215880473685, |
| "learning_rate": 9.054223210807709e-05, |
| "loss": 0.0616, |
| "step": 2940 |
| }, |
| { |
| "epoch": 1.4007597340930675, |
| "grad_norm": 0.16751316364149346, |
| "learning_rate": 9.044500345544262e-05, |
| "loss": 0.0617, |
| "step": 2950 |
| }, |
| { |
| "epoch": 1.4055080721747388, |
| "grad_norm": 0.1750365347048612, |
| "learning_rate": 9.034733035326584e-05, |
| "loss": 0.0618, |
| "step": 2960 |
| }, |
| { |
| "epoch": 1.4102564102564101, |
| "grad_norm": 0.18325337506822195, |
| "learning_rate": 9.024921387487504e-05, |
| "loss": 0.0611, |
| "step": 2970 |
| }, |
| { |
| "epoch": 1.4150047483380817, |
| "grad_norm": 0.16325381674672315, |
| "learning_rate": 9.015065509847075e-05, |
| "loss": 0.0627, |
| "step": 2980 |
| }, |
| { |
| "epoch": 1.4197530864197532, |
| "grad_norm": 0.16833321360134446, |
| "learning_rate": 9.005165510711387e-05, |
| "loss": 0.0619, |
| "step": 2990 |
| }, |
| { |
| "epoch": 1.4245014245014245, |
| "grad_norm": 0.15518995882768358, |
| "learning_rate": 8.995221498871387e-05, |
| "loss": 0.0628, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.429249762583096, |
| "grad_norm": 0.1515363714825507, |
| "learning_rate": 8.98523358360167e-05, |
| "loss": 0.0593, |
| "step": 3010 |
| }, |
| { |
| "epoch": 1.4339981006647673, |
| "grad_norm": 0.16210417405045813, |
| "learning_rate": 8.975201874659292e-05, |
| "loss": 0.0606, |
| "step": 3020 |
| }, |
| { |
| "epoch": 1.4387464387464388, |
| "grad_norm": 0.1579262025156676, |
| "learning_rate": 8.965126482282549e-05, |
| "loss": 0.0616, |
| "step": 3030 |
| }, |
| { |
| "epoch": 1.4434947768281101, |
| "grad_norm": 0.13466579123963218, |
| "learning_rate": 8.955007517189784e-05, |
| "loss": 0.0602, |
| "step": 3040 |
| }, |
| { |
| "epoch": 1.4482431149097816, |
| "grad_norm": 0.16751690047819287, |
| "learning_rate": 8.944845090578149e-05, |
| "loss": 0.0579, |
| "step": 3050 |
| }, |
| { |
| "epoch": 1.452991452991453, |
| "grad_norm": 0.16039639069860553, |
| "learning_rate": 8.934639314122401e-05, |
| "loss": 0.0602, |
| "step": 3060 |
| }, |
| { |
| "epoch": 1.4577397910731245, |
| "grad_norm": 0.1465236802201964, |
| "learning_rate": 8.924390299973666e-05, |
| "loss": 0.0592, |
| "step": 3070 |
| }, |
| { |
| "epoch": 1.4624881291547958, |
| "grad_norm": 0.16228491338881668, |
| "learning_rate": 8.91409816075821e-05, |
| "loss": 0.0578, |
| "step": 3080 |
| }, |
| { |
| "epoch": 1.4672364672364673, |
| "grad_norm": 0.15634321028105402, |
| "learning_rate": 8.903763009576195e-05, |
| "loss": 0.0581, |
| "step": 3090 |
| }, |
| { |
| "epoch": 1.4719848053181386, |
| "grad_norm": 0.1394123864334026, |
| "learning_rate": 8.893384960000445e-05, |
| "loss": 0.0594, |
| "step": 3100 |
| }, |
| { |
| "epoch": 1.47673314339981, |
| "grad_norm": 0.18130392750175844, |
| "learning_rate": 8.882964126075192e-05, |
| "loss": 0.0602, |
| "step": 3110 |
| }, |
| { |
| "epoch": 1.4814814814814814, |
| "grad_norm": 0.16461487590742802, |
| "learning_rate": 8.872500622314829e-05, |
| "loss": 0.0594, |
| "step": 3120 |
| }, |
| { |
| "epoch": 1.486229819563153, |
| "grad_norm": 0.15607091377871343, |
| "learning_rate": 8.861994563702638e-05, |
| "loss": 0.0584, |
| "step": 3130 |
| }, |
| { |
| "epoch": 1.4909781576448244, |
| "grad_norm": 0.14944642766529434, |
| "learning_rate": 8.851446065689544e-05, |
| "loss": 0.0609, |
| "step": 3140 |
| }, |
| { |
| "epoch": 1.4957264957264957, |
| "grad_norm": 0.14598220233338433, |
| "learning_rate": 8.840855244192836e-05, |
| "loss": 0.059, |
| "step": 3150 |
| }, |
| { |
| "epoch": 1.500474833808167, |
| "grad_norm": 0.14616760022404598, |
| "learning_rate": 8.83022221559489e-05, |
| "loss": 0.0583, |
| "step": 3160 |
| }, |
| { |
| "epoch": 1.5052231718898386, |
| "grad_norm": 0.14193403749024164, |
| "learning_rate": 8.819547096741904e-05, |
| "loss": 0.0595, |
| "step": 3170 |
| }, |
| { |
| "epoch": 1.50997150997151, |
| "grad_norm": 0.15132543993474756, |
| "learning_rate": 8.8088300049426e-05, |
| "loss": 0.0589, |
| "step": 3180 |
| }, |
| { |
| "epoch": 1.5147198480531814, |
| "grad_norm": 0.134346097177862, |
| "learning_rate": 8.798071057966935e-05, |
| "loss": 0.0577, |
| "step": 3190 |
| }, |
| { |
| "epoch": 1.5194681861348527, |
| "grad_norm": 0.126465871447514, |
| "learning_rate": 8.787270374044824e-05, |
| "loss": 0.059, |
| "step": 3200 |
| }, |
| { |
| "epoch": 1.5242165242165242, |
| "grad_norm": 0.14761437113460957, |
| "learning_rate": 8.776428071864819e-05, |
| "loss": 0.0585, |
| "step": 3210 |
| }, |
| { |
| "epoch": 1.5289648622981957, |
| "grad_norm": 0.15057947914224948, |
| "learning_rate": 8.76554427057282e-05, |
| "loss": 0.0581, |
| "step": 3220 |
| }, |
| { |
| "epoch": 1.533713200379867, |
| "grad_norm": 0.13407438428241547, |
| "learning_rate": 8.754619089770755e-05, |
| "loss": 0.0572, |
| "step": 3230 |
| }, |
| { |
| "epoch": 1.5384615384615383, |
| "grad_norm": 0.1680545900390031, |
| "learning_rate": 8.743652649515275e-05, |
| "loss": 0.06, |
| "step": 3240 |
| }, |
| { |
| "epoch": 1.5432098765432098, |
| "grad_norm": 0.14871133114005317, |
| "learning_rate": 8.73264507031643e-05, |
| "loss": 0.0579, |
| "step": 3250 |
| }, |
| { |
| "epoch": 1.5479582146248814, |
| "grad_norm": 0.13895672801089876, |
| "learning_rate": 8.721596473136342e-05, |
| "loss": 0.0576, |
| "step": 3260 |
| }, |
| { |
| "epoch": 1.5527065527065527, |
| "grad_norm": 0.2146963762601183, |
| "learning_rate": 8.710506979387882e-05, |
| "loss": 0.0597, |
| "step": 3270 |
| }, |
| { |
| "epoch": 1.557454890788224, |
| "grad_norm": 0.15924305974367928, |
| "learning_rate": 8.699376710933336e-05, |
| "loss": 0.0592, |
| "step": 3280 |
| }, |
| { |
| "epoch": 1.5622032288698955, |
| "grad_norm": 0.20318970954479887, |
| "learning_rate": 8.688205790083053e-05, |
| "loss": 0.0597, |
| "step": 3290 |
| }, |
| { |
| "epoch": 1.566951566951567, |
| "grad_norm": 0.1947044816388516, |
| "learning_rate": 8.676994339594118e-05, |
| "loss": 0.0597, |
| "step": 3300 |
| }, |
| { |
| "epoch": 1.5716999050332383, |
| "grad_norm": 0.14076123437862254, |
| "learning_rate": 8.665742482668996e-05, |
| "loss": 0.0587, |
| "step": 3310 |
| }, |
| { |
| "epoch": 1.5764482431149098, |
| "grad_norm": 0.14666227785262584, |
| "learning_rate": 8.654450342954175e-05, |
| "loss": 0.0585, |
| "step": 3320 |
| }, |
| { |
| "epoch": 1.5811965811965814, |
| "grad_norm": 0.15436547330795972, |
| "learning_rate": 8.643118044538808e-05, |
| "loss": 0.0582, |
| "step": 3330 |
| }, |
| { |
| "epoch": 1.5859449192782527, |
| "grad_norm": 0.13924828945146098, |
| "learning_rate": 8.631745711953357e-05, |
| "loss": 0.0572, |
| "step": 3340 |
| }, |
| { |
| "epoch": 1.590693257359924, |
| "grad_norm": 0.13931674073255168, |
| "learning_rate": 8.620333470168214e-05, |
| "loss": 0.0571, |
| "step": 3350 |
| }, |
| { |
| "epoch": 1.5954415954415955, |
| "grad_norm": 0.14548741134561383, |
| "learning_rate": 8.608881444592337e-05, |
| "loss": 0.0583, |
| "step": 3360 |
| }, |
| { |
| "epoch": 1.600189933523267, |
| "grad_norm": 0.1468408963622709, |
| "learning_rate": 8.597389761071862e-05, |
| "loss": 0.058, |
| "step": 3370 |
| }, |
| { |
| "epoch": 1.6049382716049383, |
| "grad_norm": 0.14494373270068772, |
| "learning_rate": 8.58585854588873e-05, |
| "loss": 0.0577, |
| "step": 3380 |
| }, |
| { |
| "epoch": 1.6096866096866096, |
| "grad_norm": 1.027988632662331, |
| "learning_rate": 8.574287925759294e-05, |
| "loss": 0.0663, |
| "step": 3390 |
| }, |
| { |
| "epoch": 1.6144349477682811, |
| "grad_norm": 0.2874561224559185, |
| "learning_rate": 8.562678027832925e-05, |
| "loss": 0.0802, |
| "step": 3400 |
| }, |
| { |
| "epoch": 1.6191832858499526, |
| "grad_norm": 0.1706826516389152, |
| "learning_rate": 8.551028979690621e-05, |
| "loss": 0.0622, |
| "step": 3410 |
| }, |
| { |
| "epoch": 1.623931623931624, |
| "grad_norm": 0.14889846382177033, |
| "learning_rate": 8.539340909343597e-05, |
| "loss": 0.0579, |
| "step": 3420 |
| }, |
| { |
| "epoch": 1.6286799620132952, |
| "grad_norm": 0.12292204670625376, |
| "learning_rate": 8.527613945231885e-05, |
| "loss": 0.0578, |
| "step": 3430 |
| }, |
| { |
| "epoch": 1.6334283000949668, |
| "grad_norm": 0.14933767171181656, |
| "learning_rate": 8.51584821622292e-05, |
| "loss": 0.059, |
| "step": 3440 |
| }, |
| { |
| "epoch": 1.6381766381766383, |
| "grad_norm": 0.15766461753476227, |
| "learning_rate": 8.504043851610122e-05, |
| "loss": 0.0579, |
| "step": 3450 |
| }, |
| { |
| "epoch": 1.6429249762583096, |
| "grad_norm": 0.13951544973677227, |
| "learning_rate": 8.492200981111482e-05, |
| "loss": 0.0575, |
| "step": 3460 |
| }, |
| { |
| "epoch": 1.6476733143399809, |
| "grad_norm": 0.1758053271341649, |
| "learning_rate": 8.480319734868124e-05, |
| "loss": 0.0566, |
| "step": 3470 |
| }, |
| { |
| "epoch": 1.6524216524216524, |
| "grad_norm": 0.13913868291062342, |
| "learning_rate": 8.468400243442888e-05, |
| "loss": 0.0582, |
| "step": 3480 |
| }, |
| { |
| "epoch": 1.657169990503324, |
| "grad_norm": 0.13551544048442887, |
| "learning_rate": 8.456442637818889e-05, |
| "loss": 0.0592, |
| "step": 3490 |
| }, |
| { |
| "epoch": 1.6619183285849952, |
| "grad_norm": 0.12930810620262106, |
| "learning_rate": 8.444447049398076e-05, |
| "loss": 0.0576, |
| "step": 3500 |
| }, |
| { |
| "epoch": 1.6666666666666665, |
| "grad_norm": 0.14350178998909022, |
| "learning_rate": 8.432413609999794e-05, |
| "loss": 0.0556, |
| "step": 3510 |
| }, |
| { |
| "epoch": 1.671415004748338, |
| "grad_norm": 0.1494406735925045, |
| "learning_rate": 8.420342451859329e-05, |
| "loss": 0.0593, |
| "step": 3520 |
| }, |
| { |
| "epoch": 1.6761633428300096, |
| "grad_norm": 0.14383495236320634, |
| "learning_rate": 8.408233707626458e-05, |
| "loss": 0.0572, |
| "step": 3530 |
| }, |
| { |
| "epoch": 1.6809116809116809, |
| "grad_norm": 0.1529269018391572, |
| "learning_rate": 8.39608751036399e-05, |
| "loss": 0.0578, |
| "step": 3540 |
| }, |
| { |
| "epoch": 1.6856600189933522, |
| "grad_norm": 0.1254768675507267, |
| "learning_rate": 8.383903993546311e-05, |
| "loss": 0.0558, |
| "step": 3550 |
| }, |
| { |
| "epoch": 1.690408357075024, |
| "grad_norm": 0.14137028013724545, |
| "learning_rate": 8.371683291057902e-05, |
| "loss": 0.0548, |
| "step": 3560 |
| }, |
| { |
| "epoch": 1.6951566951566952, |
| "grad_norm": 0.1446166303495371, |
| "learning_rate": 8.359425537191883e-05, |
| "loss": 0.0553, |
| "step": 3570 |
| }, |
| { |
| "epoch": 1.6999050332383665, |
| "grad_norm": 0.14496525089142717, |
| "learning_rate": 8.347130866648528e-05, |
| "loss": 0.0563, |
| "step": 3580 |
| }, |
| { |
| "epoch": 1.704653371320038, |
| "grad_norm": 0.15714440754569942, |
| "learning_rate": 8.334799414533791e-05, |
| "loss": 0.0556, |
| "step": 3590 |
| }, |
| { |
| "epoch": 1.7094017094017095, |
| "grad_norm": 0.13117043963470876, |
| "learning_rate": 8.322431316357815e-05, |
| "loss": 0.0578, |
| "step": 3600 |
| }, |
| { |
| "epoch": 1.7141500474833808, |
| "grad_norm": 0.1365276626406426, |
| "learning_rate": 8.310026708033447e-05, |
| "loss": 0.0553, |
| "step": 3610 |
| }, |
| { |
| "epoch": 1.7188983855650521, |
| "grad_norm": 0.13462462401033237, |
| "learning_rate": 8.297585725874743e-05, |
| "loss": 0.0559, |
| "step": 3620 |
| }, |
| { |
| "epoch": 1.7236467236467237, |
| "grad_norm": 0.12972719741121871, |
| "learning_rate": 8.285108506595471e-05, |
| "loss": 0.0569, |
| "step": 3630 |
| }, |
| { |
| "epoch": 1.7283950617283952, |
| "grad_norm": 0.13444324965542687, |
| "learning_rate": 8.272595187307612e-05, |
| "loss": 0.0572, |
| "step": 3640 |
| }, |
| { |
| "epoch": 1.7331433998100665, |
| "grad_norm": 0.1352948751088701, |
| "learning_rate": 8.260045905519842e-05, |
| "loss": 0.0547, |
| "step": 3650 |
| }, |
| { |
| "epoch": 1.7378917378917378, |
| "grad_norm": 0.13712282670839518, |
| "learning_rate": 8.247460799136032e-05, |
| "loss": 0.056, |
| "step": 3660 |
| }, |
| { |
| "epoch": 1.7426400759734093, |
| "grad_norm": 0.14840727583045715, |
| "learning_rate": 8.234840006453734e-05, |
| "loss": 0.0566, |
| "step": 3670 |
| }, |
| { |
| "epoch": 1.7473884140550808, |
| "grad_norm": 0.11754811913273457, |
| "learning_rate": 8.222183666162647e-05, |
| "loss": 0.0547, |
| "step": 3680 |
| }, |
| { |
| "epoch": 1.7521367521367521, |
| "grad_norm": 0.13356139519585572, |
| "learning_rate": 8.209491917343104e-05, |
| "loss": 0.0555, |
| "step": 3690 |
| }, |
| { |
| "epoch": 1.7568850902184234, |
| "grad_norm": 0.14508993625447808, |
| "learning_rate": 8.196764899464551e-05, |
| "loss": 0.0568, |
| "step": 3700 |
| }, |
| { |
| "epoch": 1.761633428300095, |
| "grad_norm": 0.13495442569215396, |
| "learning_rate": 8.184002752383993e-05, |
| "loss": 0.0557, |
| "step": 3710 |
| }, |
| { |
| "epoch": 1.7663817663817665, |
| "grad_norm": 0.13948360671434681, |
| "learning_rate": 8.171205616344478e-05, |
| "loss": 0.0551, |
| "step": 3720 |
| }, |
| { |
| "epoch": 1.7711301044634378, |
| "grad_norm": 0.14658880920142256, |
| "learning_rate": 8.158373631973542e-05, |
| "loss": 0.0562, |
| "step": 3730 |
| }, |
| { |
| "epoch": 1.775878442545109, |
| "grad_norm": 0.12457016142144928, |
| "learning_rate": 8.14550694028167e-05, |
| "loss": 0.0552, |
| "step": 3740 |
| }, |
| { |
| "epoch": 1.7806267806267806, |
| "grad_norm": 0.14490406189414082, |
| "learning_rate": 8.13260568266075e-05, |
| "loss": 0.0579, |
| "step": 3750 |
| }, |
| { |
| "epoch": 1.7853751187084521, |
| "grad_norm": 0.15756920562405344, |
| "learning_rate": 8.119670000882506e-05, |
| "loss": 0.0564, |
| "step": 3760 |
| }, |
| { |
| "epoch": 1.7901234567901234, |
| "grad_norm": 0.1371219405622126, |
| "learning_rate": 8.106700037096955e-05, |
| "loss": 0.0556, |
| "step": 3770 |
| }, |
| { |
| "epoch": 1.7948717948717947, |
| "grad_norm": 0.12215034831314227, |
| "learning_rate": 8.093695933830838e-05, |
| "loss": 0.055, |
| "step": 3780 |
| }, |
| { |
| "epoch": 1.7996201329534662, |
| "grad_norm": 0.13558185829406466, |
| "learning_rate": 8.080657833986047e-05, |
| "loss": 0.0549, |
| "step": 3790 |
| }, |
| { |
| "epoch": 1.8043684710351378, |
| "grad_norm": 0.13756088055437052, |
| "learning_rate": 8.067585880838073e-05, |
| "loss": 0.0567, |
| "step": 3800 |
| }, |
| { |
| "epoch": 1.809116809116809, |
| "grad_norm": 0.13066759711523296, |
| "learning_rate": 8.054480218034413e-05, |
| "loss": 0.0549, |
| "step": 3810 |
| }, |
| { |
| "epoch": 1.8138651471984806, |
| "grad_norm": 0.11737735156248243, |
| "learning_rate": 8.041340989593003e-05, |
| "loss": 0.0533, |
| "step": 3820 |
| }, |
| { |
| "epoch": 1.818613485280152, |
| "grad_norm": 0.13369087417533748, |
| "learning_rate": 8.028168339900625e-05, |
| "loss": 0.0542, |
| "step": 3830 |
| }, |
| { |
| "epoch": 1.8233618233618234, |
| "grad_norm": 0.14154528831038649, |
| "learning_rate": 8.014962413711334e-05, |
| "loss": 0.0566, |
| "step": 3840 |
| }, |
| { |
| "epoch": 1.8281101614434947, |
| "grad_norm": 0.13729260186633008, |
| "learning_rate": 8.001723356144855e-05, |
| "loss": 0.0545, |
| "step": 3850 |
| }, |
| { |
| "epoch": 1.8328584995251662, |
| "grad_norm": 0.12726444289628347, |
| "learning_rate": 7.988451312684997e-05, |
| "loss": 0.0544, |
| "step": 3860 |
| }, |
| { |
| "epoch": 1.8376068376068377, |
| "grad_norm": 0.1457966431471358, |
| "learning_rate": 7.975146429178047e-05, |
| "loss": 0.0539, |
| "step": 3870 |
| }, |
| { |
| "epoch": 1.842355175688509, |
| "grad_norm": 0.13864555167082063, |
| "learning_rate": 7.961808851831172e-05, |
| "loss": 0.0543, |
| "step": 3880 |
| }, |
| { |
| "epoch": 1.8471035137701803, |
| "grad_norm": 0.12438561167954812, |
| "learning_rate": 7.948438727210814e-05, |
| "loss": 0.0559, |
| "step": 3890 |
| }, |
| { |
| "epoch": 1.8518518518518519, |
| "grad_norm": 0.12601228423610955, |
| "learning_rate": 7.935036202241071e-05, |
| "loss": 0.0539, |
| "step": 3900 |
| }, |
| { |
| "epoch": 1.8566001899335234, |
| "grad_norm": 0.13078943788955055, |
| "learning_rate": 7.921601424202091e-05, |
| "loss": 0.0545, |
| "step": 3910 |
| }, |
| { |
| "epoch": 1.8613485280151947, |
| "grad_norm": 0.1288678460120429, |
| "learning_rate": 7.908134540728452e-05, |
| "loss": 0.0545, |
| "step": 3920 |
| }, |
| { |
| "epoch": 1.866096866096866, |
| "grad_norm": 0.13079908087133826, |
| "learning_rate": 7.894635699807533e-05, |
| "loss": 0.0554, |
| "step": 3930 |
| }, |
| { |
| "epoch": 1.8708452041785375, |
| "grad_norm": 0.1272428561784055, |
| "learning_rate": 7.881105049777901e-05, |
| "loss": 0.054, |
| "step": 3940 |
| }, |
| { |
| "epoch": 1.875593542260209, |
| "grad_norm": 0.13740732096430242, |
| "learning_rate": 7.867542739327667e-05, |
| "loss": 0.0553, |
| "step": 3950 |
| }, |
| { |
| "epoch": 1.8803418803418803, |
| "grad_norm": 0.1406312486201405, |
| "learning_rate": 7.853948917492857e-05, |
| "loss": 0.0552, |
| "step": 3960 |
| }, |
| { |
| "epoch": 1.8850902184235516, |
| "grad_norm": 0.11031923209404337, |
| "learning_rate": 7.840323733655778e-05, |
| "loss": 0.0532, |
| "step": 3970 |
| }, |
| { |
| "epoch": 1.8898385565052231, |
| "grad_norm": 0.1927099264028857, |
| "learning_rate": 7.826667337543377e-05, |
| "loss": 0.0538, |
| "step": 3980 |
| }, |
| { |
| "epoch": 1.8945868945868947, |
| "grad_norm": 0.1198535804528265, |
| "learning_rate": 7.812979879225581e-05, |
| "loss": 0.0545, |
| "step": 3990 |
| }, |
| { |
| "epoch": 1.899335232668566, |
| "grad_norm": 0.12721019731521777, |
| "learning_rate": 7.799261509113671e-05, |
| "loss": 0.0535, |
| "step": 4000 |
| }, |
| { |
| "epoch": 1.9040835707502373, |
| "grad_norm": 0.12508154803117758, |
| "learning_rate": 7.78551237795861e-05, |
| "loss": 0.0536, |
| "step": 4010 |
| }, |
| { |
| "epoch": 1.9088319088319088, |
| "grad_norm": 0.11723551332606386, |
| "learning_rate": 7.771732636849399e-05, |
| "loss": 0.0529, |
| "step": 4020 |
| }, |
| { |
| "epoch": 1.9135802469135803, |
| "grad_norm": 0.15219932544419731, |
| "learning_rate": 7.757922437211405e-05, |
| "loss": 0.0535, |
| "step": 4030 |
| }, |
| { |
| "epoch": 1.9183285849952516, |
| "grad_norm": 0.1315779713221699, |
| "learning_rate": 7.744081930804707e-05, |
| "loss": 0.0542, |
| "step": 4040 |
| }, |
| { |
| "epoch": 1.9230769230769231, |
| "grad_norm": 0.11600147713426219, |
| "learning_rate": 7.730211269722427e-05, |
| "loss": 0.0553, |
| "step": 4050 |
| }, |
| { |
| "epoch": 1.9278252611585946, |
| "grad_norm": 0.11652128743882831, |
| "learning_rate": 7.716310606389052e-05, |
| "loss": 0.0545, |
| "step": 4060 |
| }, |
| { |
| "epoch": 1.932573599240266, |
| "grad_norm": 0.13680801266595588, |
| "learning_rate": 7.702380093558765e-05, |
| "loss": 0.0538, |
| "step": 4070 |
| }, |
| { |
| "epoch": 1.9373219373219372, |
| "grad_norm": 0.1227181445186413, |
| "learning_rate": 7.688419884313764e-05, |
| "loss": 0.0539, |
| "step": 4080 |
| }, |
| { |
| "epoch": 1.9420702754036088, |
| "grad_norm": 0.12872608419380774, |
| "learning_rate": 7.674430132062585e-05, |
| "loss": 0.0548, |
| "step": 4090 |
| }, |
| { |
| "epoch": 1.9468186134852803, |
| "grad_norm": 0.12200503175301114, |
| "learning_rate": 7.660410990538403e-05, |
| "loss": 0.0542, |
| "step": 4100 |
| }, |
| { |
| "epoch": 1.9515669515669516, |
| "grad_norm": 0.13368769696036206, |
| "learning_rate": 7.646362613797356e-05, |
| "loss": 0.053, |
| "step": 4110 |
| }, |
| { |
| "epoch": 1.956315289648623, |
| "grad_norm": 0.1356311805363903, |
| "learning_rate": 7.63228515621685e-05, |
| "loss": 0.0558, |
| "step": 4120 |
| }, |
| { |
| "epoch": 1.9610636277302944, |
| "grad_norm": 0.13886294495646934, |
| "learning_rate": 7.618178772493856e-05, |
| "loss": 0.0528, |
| "step": 4130 |
| }, |
| { |
| "epoch": 1.965811965811966, |
| "grad_norm": 0.12184672522260906, |
| "learning_rate": 7.604043617643215e-05, |
| "loss": 0.0548, |
| "step": 4140 |
| }, |
| { |
| "epoch": 1.9705603038936372, |
| "grad_norm": 0.13070115769795196, |
| "learning_rate": 7.589879846995934e-05, |
| "loss": 0.0544, |
| "step": 4150 |
| }, |
| { |
| "epoch": 1.9753086419753085, |
| "grad_norm": 0.11226755795577562, |
| "learning_rate": 7.575687616197477e-05, |
| "loss": 0.054, |
| "step": 4160 |
| }, |
| { |
| "epoch": 1.98005698005698, |
| "grad_norm": 0.12969931416626887, |
| "learning_rate": 7.561467081206061e-05, |
| "loss": 0.053, |
| "step": 4170 |
| }, |
| { |
| "epoch": 1.9848053181386516, |
| "grad_norm": 0.12164796500415757, |
| "learning_rate": 7.547218398290932e-05, |
| "loss": 0.0539, |
| "step": 4180 |
| }, |
| { |
| "epoch": 1.9895536562203229, |
| "grad_norm": 0.13244317157275187, |
| "learning_rate": 7.532941724030654e-05, |
| "loss": 0.0541, |
| "step": 4190 |
| }, |
| { |
| "epoch": 1.9943019943019942, |
| "grad_norm": 0.11105078541758974, |
| "learning_rate": 7.518637215311388e-05, |
| "loss": 0.0528, |
| "step": 4200 |
| }, |
| { |
| "epoch": 1.9990503323836657, |
| "grad_norm": 0.11686176190570292, |
| "learning_rate": 7.504305029325172e-05, |
| "loss": 0.0533, |
| "step": 4210 |
| }, |
| { |
| "epoch": 2.003798670465337, |
| "grad_norm": 0.11985517056752061, |
| "learning_rate": 7.489945323568178e-05, |
| "loss": 0.052, |
| "step": 4220 |
| }, |
| { |
| "epoch": 2.0085470085470085, |
| "grad_norm": 0.11815522523627926, |
| "learning_rate": 7.475558255839006e-05, |
| "loss": 0.0503, |
| "step": 4230 |
| }, |
| { |
| "epoch": 2.01329534662868, |
| "grad_norm": 0.10792283132487078, |
| "learning_rate": 7.461143984236924e-05, |
| "loss": 0.0498, |
| "step": 4240 |
| }, |
| { |
| "epoch": 2.0180436847103516, |
| "grad_norm": 0.1127414167901404, |
| "learning_rate": 7.446702667160154e-05, |
| "loss": 0.0502, |
| "step": 4250 |
| }, |
| { |
| "epoch": 2.022792022792023, |
| "grad_norm": 0.11613353462797621, |
| "learning_rate": 7.432234463304111e-05, |
| "loss": 0.0508, |
| "step": 4260 |
| }, |
| { |
| "epoch": 2.027540360873694, |
| "grad_norm": 0.11572850834831004, |
| "learning_rate": 7.417739531659674e-05, |
| "loss": 0.052, |
| "step": 4270 |
| }, |
| { |
| "epoch": 2.0322886989553655, |
| "grad_norm": 0.11273415395030174, |
| "learning_rate": 7.403218031511428e-05, |
| "loss": 0.0497, |
| "step": 4280 |
| }, |
| { |
| "epoch": 2.037037037037037, |
| "grad_norm": 0.12418278997231333, |
| "learning_rate": 7.388670122435928e-05, |
| "loss": 0.0509, |
| "step": 4290 |
| }, |
| { |
| "epoch": 2.0417853751187085, |
| "grad_norm": 0.10327465007056622, |
| "learning_rate": 7.374095964299928e-05, |
| "loss": 0.0495, |
| "step": 4300 |
| }, |
| { |
| "epoch": 2.04653371320038, |
| "grad_norm": 0.1223594319466957, |
| "learning_rate": 7.359495717258636e-05, |
| "loss": 0.0501, |
| "step": 4310 |
| }, |
| { |
| "epoch": 2.051282051282051, |
| "grad_norm": 0.13026380845413757, |
| "learning_rate": 7.34486954175395e-05, |
| "loss": 0.0515, |
| "step": 4320 |
| }, |
| { |
| "epoch": 2.056030389363723, |
| "grad_norm": 0.15011490540872358, |
| "learning_rate": 7.330217598512695e-05, |
| "loss": 0.0477, |
| "step": 4330 |
| }, |
| { |
| "epoch": 2.060778727445394, |
| "grad_norm": 0.12503768986220248, |
| "learning_rate": 7.315540048544857e-05, |
| "loss": 0.0506, |
| "step": 4340 |
| }, |
| { |
| "epoch": 2.0655270655270654, |
| "grad_norm": 0.10565387471065048, |
| "learning_rate": 7.300837053141815e-05, |
| "loss": 0.0498, |
| "step": 4350 |
| }, |
| { |
| "epoch": 2.0702754036087367, |
| "grad_norm": 0.10700109701778658, |
| "learning_rate": 7.286108773874566e-05, |
| "loss": 0.0496, |
| "step": 4360 |
| }, |
| { |
| "epoch": 2.0750237416904085, |
| "grad_norm": 0.10003352491732063, |
| "learning_rate": 7.271355372591956e-05, |
| "loss": 0.0483, |
| "step": 4370 |
| }, |
| { |
| "epoch": 2.07977207977208, |
| "grad_norm": 0.11144146396030583, |
| "learning_rate": 7.256577011418888e-05, |
| "loss": 0.0496, |
| "step": 4380 |
| }, |
| { |
| "epoch": 2.084520417853751, |
| "grad_norm": 0.10680493099395319, |
| "learning_rate": 7.241773852754556e-05, |
| "loss": 0.0493, |
| "step": 4390 |
| }, |
| { |
| "epoch": 2.0892687559354224, |
| "grad_norm": 0.11974678661099225, |
| "learning_rate": 7.226946059270651e-05, |
| "loss": 0.0502, |
| "step": 4400 |
| }, |
| { |
| "epoch": 2.094017094017094, |
| "grad_norm": 0.12248796357405212, |
| "learning_rate": 7.212093793909573e-05, |
| "loss": 0.0512, |
| "step": 4410 |
| }, |
| { |
| "epoch": 2.0987654320987654, |
| "grad_norm": 0.12590585228798748, |
| "learning_rate": 7.197217219882647e-05, |
| "loss": 0.0503, |
| "step": 4420 |
| }, |
| { |
| "epoch": 2.1035137701804367, |
| "grad_norm": 0.1169637108796757, |
| "learning_rate": 7.182316500668326e-05, |
| "loss": 0.051, |
| "step": 4430 |
| }, |
| { |
| "epoch": 2.1082621082621085, |
| "grad_norm": 0.12360169441454422, |
| "learning_rate": 7.167391800010389e-05, |
| "loss": 0.05, |
| "step": 4440 |
| }, |
| { |
| "epoch": 2.1130104463437798, |
| "grad_norm": 0.12433259182922494, |
| "learning_rate": 7.152443281916151e-05, |
| "loss": 0.0502, |
| "step": 4450 |
| }, |
| { |
| "epoch": 2.117758784425451, |
| "grad_norm": 0.12747063451316107, |
| "learning_rate": 7.137471110654656e-05, |
| "loss": 0.0513, |
| "step": 4460 |
| }, |
| { |
| "epoch": 2.1225071225071224, |
| "grad_norm": 0.11580065014607085, |
| "learning_rate": 7.122475450754869e-05, |
| "loss": 0.0488, |
| "step": 4470 |
| }, |
| { |
| "epoch": 2.127255460588794, |
| "grad_norm": 0.1202935502893037, |
| "learning_rate": 7.107456467003877e-05, |
| "loss": 0.0511, |
| "step": 4480 |
| }, |
| { |
| "epoch": 2.1320037986704654, |
| "grad_norm": 0.12712841677739506, |
| "learning_rate": 7.092414324445068e-05, |
| "loss": 0.0496, |
| "step": 4490 |
| }, |
| { |
| "epoch": 2.1367521367521367, |
| "grad_norm": 0.11145780649446338, |
| "learning_rate": 7.077349188376325e-05, |
| "loss": 0.0499, |
| "step": 4500 |
| }, |
| { |
| "epoch": 2.141500474833808, |
| "grad_norm": 0.12442688712317078, |
| "learning_rate": 7.062261224348203e-05, |
| "loss": 0.0504, |
| "step": 4510 |
| }, |
| { |
| "epoch": 2.1462488129154798, |
| "grad_norm": 0.11841371177218044, |
| "learning_rate": 7.047150598162114e-05, |
| "loss": 0.0498, |
| "step": 4520 |
| }, |
| { |
| "epoch": 2.150997150997151, |
| "grad_norm": 0.11265032815924783, |
| "learning_rate": 7.032017475868505e-05, |
| "loss": 0.0495, |
| "step": 4530 |
| }, |
| { |
| "epoch": 2.1557454890788224, |
| "grad_norm": 0.11061387303547136, |
| "learning_rate": 7.016862023765033e-05, |
| "loss": 0.0487, |
| "step": 4540 |
| }, |
| { |
| "epoch": 2.1604938271604937, |
| "grad_norm": 0.1060970054082258, |
| "learning_rate": 7.001684408394735e-05, |
| "loss": 0.0481, |
| "step": 4550 |
| }, |
| { |
| "epoch": 2.1652421652421654, |
| "grad_norm": 0.1061562833370673, |
| "learning_rate": 6.986484796544203e-05, |
| "loss": 0.0504, |
| "step": 4560 |
| }, |
| { |
| "epoch": 2.1699905033238367, |
| "grad_norm": 0.11636343937270621, |
| "learning_rate": 6.971263355241742e-05, |
| "loss": 0.0506, |
| "step": 4570 |
| }, |
| { |
| "epoch": 2.174738841405508, |
| "grad_norm": 0.11296486565238632, |
| "learning_rate": 6.956020251755548e-05, |
| "loss": 0.0498, |
| "step": 4580 |
| }, |
| { |
| "epoch": 2.1794871794871793, |
| "grad_norm": 0.11632860185711963, |
| "learning_rate": 6.940755653591858e-05, |
| "loss": 0.0486, |
| "step": 4590 |
| }, |
| { |
| "epoch": 2.184235517568851, |
| "grad_norm": 0.11588233164136699, |
| "learning_rate": 6.925469728493114e-05, |
| "loss": 0.0486, |
| "step": 4600 |
| }, |
| { |
| "epoch": 2.1889838556505223, |
| "grad_norm": 0.12552463964575056, |
| "learning_rate": 6.91016264443612e-05, |
| "loss": 0.0506, |
| "step": 4610 |
| }, |
| { |
| "epoch": 2.1937321937321936, |
| "grad_norm": 0.11335949983781556, |
| "learning_rate": 6.894834569630195e-05, |
| "loss": 0.0504, |
| "step": 4620 |
| }, |
| { |
| "epoch": 2.198480531813865, |
| "grad_norm": 0.11135859085870432, |
| "learning_rate": 6.879485672515327e-05, |
| "loss": 0.0505, |
| "step": 4630 |
| }, |
| { |
| "epoch": 2.2032288698955367, |
| "grad_norm": 0.12924950833022672, |
| "learning_rate": 6.864116121760316e-05, |
| "loss": 0.0509, |
| "step": 4640 |
| }, |
| { |
| "epoch": 2.207977207977208, |
| "grad_norm": 0.1164231876355468, |
| "learning_rate": 6.848726086260929e-05, |
| "loss": 0.0511, |
| "step": 4650 |
| }, |
| { |
| "epoch": 2.2127255460588793, |
| "grad_norm": 0.12439021897094324, |
| "learning_rate": 6.833315735138035e-05, |
| "loss": 0.0504, |
| "step": 4660 |
| }, |
| { |
| "epoch": 2.217473884140551, |
| "grad_norm": 0.10918851132484461, |
| "learning_rate": 6.817885237735758e-05, |
| "loss": 0.0505, |
| "step": 4670 |
| }, |
| { |
| "epoch": 2.2222222222222223, |
| "grad_norm": 0.12664032717274012, |
| "learning_rate": 6.802434763619599e-05, |
| "loss": 0.0512, |
| "step": 4680 |
| }, |
| { |
| "epoch": 2.2269705603038936, |
| "grad_norm": 0.12789634241743944, |
| "learning_rate": 6.786964482574594e-05, |
| "loss": 0.0495, |
| "step": 4690 |
| }, |
| { |
| "epoch": 2.231718898385565, |
| "grad_norm": 0.12498057095260956, |
| "learning_rate": 6.771474564603428e-05, |
| "loss": 0.0494, |
| "step": 4700 |
| }, |
| { |
| "epoch": 2.2364672364672367, |
| "grad_norm": 0.10520453609463827, |
| "learning_rate": 6.755965179924582e-05, |
| "loss": 0.0495, |
| "step": 4710 |
| }, |
| { |
| "epoch": 2.241215574548908, |
| "grad_norm": 0.11499787166262901, |
| "learning_rate": 6.740436498970452e-05, |
| "loss": 0.0504, |
| "step": 4720 |
| }, |
| { |
| "epoch": 2.2459639126305793, |
| "grad_norm": 0.1118438057434224, |
| "learning_rate": 6.724888692385485e-05, |
| "loss": 0.0505, |
| "step": 4730 |
| }, |
| { |
| "epoch": 2.2507122507122506, |
| "grad_norm": 0.11101152182098649, |
| "learning_rate": 6.709321931024293e-05, |
| "loss": 0.0492, |
| "step": 4740 |
| }, |
| { |
| "epoch": 2.2554605887939223, |
| "grad_norm": 0.12885302765643578, |
| "learning_rate": 6.693736385949788e-05, |
| "loss": 0.0506, |
| "step": 4750 |
| }, |
| { |
| "epoch": 2.2602089268755936, |
| "grad_norm": 0.1161213304336198, |
| "learning_rate": 6.678132228431294e-05, |
| "loss": 0.05, |
| "step": 4760 |
| }, |
| { |
| "epoch": 2.264957264957265, |
| "grad_norm": 0.10777929444702916, |
| "learning_rate": 6.662509629942661e-05, |
| "loss": 0.048, |
| "step": 4770 |
| }, |
| { |
| "epoch": 2.269705603038936, |
| "grad_norm": 0.1524234079506994, |
| "learning_rate": 6.646868762160399e-05, |
| "loss": 0.0489, |
| "step": 4780 |
| }, |
| { |
| "epoch": 2.274453941120608, |
| "grad_norm": 0.10871071270348015, |
| "learning_rate": 6.631209796961767e-05, |
| "loss": 0.05, |
| "step": 4790 |
| }, |
| { |
| "epoch": 2.2792022792022792, |
| "grad_norm": 0.1389362902619363, |
| "learning_rate": 6.615532906422902e-05, |
| "loss": 0.0503, |
| "step": 4800 |
| }, |
| { |
| "epoch": 2.2839506172839505, |
| "grad_norm": 0.1182558185826198, |
| "learning_rate": 6.599838262816922e-05, |
| "loss": 0.0497, |
| "step": 4810 |
| }, |
| { |
| "epoch": 2.288698955365622, |
| "grad_norm": 0.12473134012146316, |
| "learning_rate": 6.584126038612033e-05, |
| "loss": 0.0498, |
| "step": 4820 |
| }, |
| { |
| "epoch": 2.2934472934472936, |
| "grad_norm": 0.11015962586807758, |
| "learning_rate": 6.568396406469632e-05, |
| "loss": 0.0493, |
| "step": 4830 |
| }, |
| { |
| "epoch": 2.298195631528965, |
| "grad_norm": 0.11522128152541927, |
| "learning_rate": 6.552649539242414e-05, |
| "loss": 0.0494, |
| "step": 4840 |
| }, |
| { |
| "epoch": 2.302943969610636, |
| "grad_norm": 0.11851157339445201, |
| "learning_rate": 6.536885609972467e-05, |
| "loss": 0.0496, |
| "step": 4850 |
| }, |
| { |
| "epoch": 2.3076923076923075, |
| "grad_norm": 0.12367735679385537, |
| "learning_rate": 6.521104791889381e-05, |
| "loss": 0.0499, |
| "step": 4860 |
| }, |
| { |
| "epoch": 2.3124406457739792, |
| "grad_norm": 0.11038723546767162, |
| "learning_rate": 6.505307258408325e-05, |
| "loss": 0.0492, |
| "step": 4870 |
| }, |
| { |
| "epoch": 2.3171889838556505, |
| "grad_norm": 0.13160267156026526, |
| "learning_rate": 6.489493183128163e-05, |
| "loss": 0.0499, |
| "step": 4880 |
| }, |
| { |
| "epoch": 2.321937321937322, |
| "grad_norm": 0.12495190660680677, |
| "learning_rate": 6.473662739829533e-05, |
| "loss": 0.0496, |
| "step": 4890 |
| }, |
| { |
| "epoch": 2.3266856600189936, |
| "grad_norm": 0.11136470406999355, |
| "learning_rate": 6.457816102472942e-05, |
| "loss": 0.0496, |
| "step": 4900 |
| }, |
| { |
| "epoch": 2.331433998100665, |
| "grad_norm": 0.11903045963813727, |
| "learning_rate": 6.44195344519685e-05, |
| "loss": 0.0503, |
| "step": 4910 |
| }, |
| { |
| "epoch": 2.336182336182336, |
| "grad_norm": 0.10930739841179492, |
| "learning_rate": 6.426074942315766e-05, |
| "loss": 0.0487, |
| "step": 4920 |
| }, |
| { |
| "epoch": 2.3409306742640075, |
| "grad_norm": 0.11451265717748751, |
| "learning_rate": 6.410180768318321e-05, |
| "loss": 0.0499, |
| "step": 4930 |
| }, |
| { |
| "epoch": 2.3456790123456788, |
| "grad_norm": 0.1213898603907506, |
| "learning_rate": 6.394271097865355e-05, |
| "loss": 0.0491, |
| "step": 4940 |
| }, |
| { |
| "epoch": 2.3504273504273505, |
| "grad_norm": 0.12234588742924166, |
| "learning_rate": 6.378346105788002e-05, |
| "loss": 0.0488, |
| "step": 4950 |
| }, |
| { |
| "epoch": 2.355175688509022, |
| "grad_norm": 0.10956011406442001, |
| "learning_rate": 6.362405967085765e-05, |
| "loss": 0.0487, |
| "step": 4960 |
| }, |
| { |
| "epoch": 2.359924026590693, |
| "grad_norm": 0.12856280177091386, |
| "learning_rate": 6.346450856924588e-05, |
| "loss": 0.0496, |
| "step": 4970 |
| }, |
| { |
| "epoch": 2.364672364672365, |
| "grad_norm": 0.12203320248173269, |
| "learning_rate": 6.330480950634942e-05, |
| "loss": 0.0476, |
| "step": 4980 |
| }, |
| { |
| "epoch": 2.369420702754036, |
| "grad_norm": 0.114449100814288, |
| "learning_rate": 6.314496423709887e-05, |
| "loss": 0.0481, |
| "step": 4990 |
| }, |
| { |
| "epoch": 2.3741690408357075, |
| "grad_norm": 0.1160440028701934, |
| "learning_rate": 6.298497451803157e-05, |
| "loss": 0.0489, |
| "step": 5000 |
| }, |
| { |
| "epoch": 2.3789173789173788, |
| "grad_norm": 0.11819129519672794, |
| "learning_rate": 6.282484210727213e-05, |
| "loss": 0.0487, |
| "step": 5010 |
| }, |
| { |
| "epoch": 2.3836657169990505, |
| "grad_norm": 0.1008521705129008, |
| "learning_rate": 6.266456876451324e-05, |
| "loss": 0.0471, |
| "step": 5020 |
| }, |
| { |
| "epoch": 2.388414055080722, |
| "grad_norm": 0.10420493671943033, |
| "learning_rate": 6.25041562509963e-05, |
| "loss": 0.0476, |
| "step": 5030 |
| }, |
| { |
| "epoch": 2.393162393162393, |
| "grad_norm": 0.10750401355477318, |
| "learning_rate": 6.234360632949201e-05, |
| "loss": 0.0493, |
| "step": 5040 |
| }, |
| { |
| "epoch": 2.3979107312440644, |
| "grad_norm": 0.12165236164292002, |
| "learning_rate": 6.21829207642811e-05, |
| "loss": 0.0489, |
| "step": 5050 |
| }, |
| { |
| "epoch": 2.402659069325736, |
| "grad_norm": 0.10743996867903316, |
| "learning_rate": 6.202210132113484e-05, |
| "loss": 0.0492, |
| "step": 5060 |
| }, |
| { |
| "epoch": 2.4074074074074074, |
| "grad_norm": 0.11006147439530166, |
| "learning_rate": 6.186114976729573e-05, |
| "loss": 0.049, |
| "step": 5070 |
| }, |
| { |
| "epoch": 2.4121557454890787, |
| "grad_norm": 0.13736066574497494, |
| "learning_rate": 6.170006787145796e-05, |
| "loss": 0.0494, |
| "step": 5080 |
| }, |
| { |
| "epoch": 2.41690408357075, |
| "grad_norm": 0.10381012959225568, |
| "learning_rate": 6.153885740374813e-05, |
| "loss": 0.0492, |
| "step": 5090 |
| }, |
| { |
| "epoch": 2.421652421652422, |
| "grad_norm": 0.11302028895505241, |
| "learning_rate": 6.137752013570566e-05, |
| "loss": 0.0484, |
| "step": 5100 |
| }, |
| { |
| "epoch": 2.426400759734093, |
| "grad_norm": 0.10522975384164665, |
| "learning_rate": 6.121605784026339e-05, |
| "loss": 0.0487, |
| "step": 5110 |
| }, |
| { |
| "epoch": 2.4311490978157644, |
| "grad_norm": 0.10604021592636895, |
| "learning_rate": 6.10544722917281e-05, |
| "loss": 0.0495, |
| "step": 5120 |
| }, |
| { |
| "epoch": 2.435897435897436, |
| "grad_norm": 0.10106488606867245, |
| "learning_rate": 6.089276526576094e-05, |
| "loss": 0.0485, |
| "step": 5130 |
| }, |
| { |
| "epoch": 2.4406457739791074, |
| "grad_norm": 0.12035676767142514, |
| "learning_rate": 6.0730938539358084e-05, |
| "loss": 0.0493, |
| "step": 5140 |
| }, |
| { |
| "epoch": 2.4453941120607787, |
| "grad_norm": 0.1164211028522092, |
| "learning_rate": 6.056899389083097e-05, |
| "loss": 0.0483, |
| "step": 5150 |
| }, |
| { |
| "epoch": 2.45014245014245, |
| "grad_norm": 0.12037931973810435, |
| "learning_rate": 6.0406933099786976e-05, |
| "loss": 0.0497, |
| "step": 5160 |
| }, |
| { |
| "epoch": 2.4548907882241213, |
| "grad_norm": 0.12528161144671687, |
| "learning_rate": 6.02447579471097e-05, |
| "loss": 0.0493, |
| "step": 5170 |
| }, |
| { |
| "epoch": 2.459639126305793, |
| "grad_norm": 0.1004073889039229, |
| "learning_rate": 6.0082470214939514e-05, |
| "loss": 0.048, |
| "step": 5180 |
| }, |
| { |
| "epoch": 2.4643874643874644, |
| "grad_norm": 0.11447945556304559, |
| "learning_rate": 5.9920071686653855e-05, |
| "loss": 0.0487, |
| "step": 5190 |
| }, |
| { |
| "epoch": 2.4691358024691357, |
| "grad_norm": 0.12463475244856546, |
| "learning_rate": 5.9757564146847764e-05, |
| "loss": 0.0488, |
| "step": 5200 |
| }, |
| { |
| "epoch": 2.4738841405508074, |
| "grad_norm": 0.11900023752667457, |
| "learning_rate": 5.959494938131418e-05, |
| "loss": 0.0484, |
| "step": 5210 |
| }, |
| { |
| "epoch": 2.4786324786324787, |
| "grad_norm": 0.1119059802137425, |
| "learning_rate": 5.9432229177024346e-05, |
| "loss": 0.049, |
| "step": 5220 |
| }, |
| { |
| "epoch": 2.48338081671415, |
| "grad_norm": 0.10625933744451875, |
| "learning_rate": 5.926940532210817e-05, |
| "loss": 0.0484, |
| "step": 5230 |
| }, |
| { |
| "epoch": 2.4881291547958213, |
| "grad_norm": 0.11392423214304202, |
| "learning_rate": 5.910647960583457e-05, |
| "loss": 0.0495, |
| "step": 5240 |
| }, |
| { |
| "epoch": 2.492877492877493, |
| "grad_norm": 0.09478513705518692, |
| "learning_rate": 5.8943453818591845e-05, |
| "loss": 0.0486, |
| "step": 5250 |
| }, |
| { |
| "epoch": 2.4976258309591644, |
| "grad_norm": 0.11056971988901111, |
| "learning_rate": 5.878032975186791e-05, |
| "loss": 0.0504, |
| "step": 5260 |
| }, |
| { |
| "epoch": 2.5023741690408356, |
| "grad_norm": 0.12899883149816868, |
| "learning_rate": 5.8617109198230744e-05, |
| "loss": 0.049, |
| "step": 5270 |
| }, |
| { |
| "epoch": 2.5071225071225074, |
| "grad_norm": 0.11831990366789337, |
| "learning_rate": 5.845379395130859e-05, |
| "loss": 0.0482, |
| "step": 5280 |
| }, |
| { |
| "epoch": 2.5118708452041787, |
| "grad_norm": 0.10425149780008909, |
| "learning_rate": 5.829038580577025e-05, |
| "loss": 0.0474, |
| "step": 5290 |
| }, |
| { |
| "epoch": 2.51661918328585, |
| "grad_norm": 0.10955378333660806, |
| "learning_rate": 5.812688655730542e-05, |
| "loss": 0.0482, |
| "step": 5300 |
| }, |
| { |
| "epoch": 2.5213675213675213, |
| "grad_norm": 0.11170253299315429, |
| "learning_rate": 5.796329800260489e-05, |
| "loss": 0.0495, |
| "step": 5310 |
| }, |
| { |
| "epoch": 2.5261158594491926, |
| "grad_norm": 0.12485494338656279, |
| "learning_rate": 5.7799621939340896e-05, |
| "loss": 0.0491, |
| "step": 5320 |
| }, |
| { |
| "epoch": 2.5308641975308643, |
| "grad_norm": 0.112858953113323, |
| "learning_rate": 5.7635860166147246e-05, |
| "loss": 0.0471, |
| "step": 5330 |
| }, |
| { |
| "epoch": 2.5356125356125356, |
| "grad_norm": 0.1076614489045162, |
| "learning_rate": 5.747201448259964e-05, |
| "loss": 0.0487, |
| "step": 5340 |
| }, |
| { |
| "epoch": 2.540360873694207, |
| "grad_norm": 0.1430181478411634, |
| "learning_rate": 5.730808668919584e-05, |
| "loss": 0.048, |
| "step": 5350 |
| }, |
| { |
| "epoch": 2.5451092117758787, |
| "grad_norm": 0.11575456302187916, |
| "learning_rate": 5.714407858733596e-05, |
| "loss": 0.0478, |
| "step": 5360 |
| }, |
| { |
| "epoch": 2.54985754985755, |
| "grad_norm": 0.11470298932399466, |
| "learning_rate": 5.6979991979302585e-05, |
| "loss": 0.0493, |
| "step": 5370 |
| }, |
| { |
| "epoch": 2.5546058879392213, |
| "grad_norm": 0.1008226727049796, |
| "learning_rate": 5.6815828668241e-05, |
| "loss": 0.0473, |
| "step": 5380 |
| }, |
| { |
| "epoch": 2.5593542260208926, |
| "grad_norm": 0.12449574177250117, |
| "learning_rate": 5.6651590458139424e-05, |
| "loss": 0.0486, |
| "step": 5390 |
| }, |
| { |
| "epoch": 2.564102564102564, |
| "grad_norm": 0.1124061683188527, |
| "learning_rate": 5.6487279153809094e-05, |
| "loss": 0.0492, |
| "step": 5400 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10530, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 1800, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 475817556049920.0, |
| "train_batch_size": 128, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
