{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.7094017094017095, "eval_steps": 500, "global_step": 3600, "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 } ], "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": 317299406929920.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }