diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,21034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 18.844221105527637, + "eval_steps": 500, + "global_step": 30000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00628140703517588, + "grad_norm": 9.363000869750977, + "learning_rate": 6.000000000000001e-07, + "loss": 0.918, + "step": 10 + }, + { + "epoch": 0.01256281407035176, + "grad_norm": 9.200325965881348, + "learning_rate": 1.2666666666666667e-06, + "loss": 0.9444, + "step": 20 + }, + { + "epoch": 0.018844221105527637, + "grad_norm": 7.444415092468262, + "learning_rate": 1.9333333333333336e-06, + "loss": 0.8031, + "step": 30 + }, + { + "epoch": 0.02512562814070352, + "grad_norm": 3.9234108924865723, + "learning_rate": 2.6e-06, + "loss": 0.5813, + "step": 40 + }, + { + "epoch": 0.031407035175879394, + "grad_norm": 2.597147226333618, + "learning_rate": 3.2666666666666666e-06, + "loss": 0.407, + "step": 50 + }, + { + "epoch": 0.03768844221105527, + "grad_norm": 2.7692222595214844, + "learning_rate": 3.9333333333333335e-06, + "loss": 0.3455, + "step": 60 + }, + { + "epoch": 0.04396984924623116, + "grad_norm": 1.085033893585205, + "learning_rate": 4.6e-06, + "loss": 0.2315, + "step": 70 + }, + { + "epoch": 0.05025125628140704, + "grad_norm": 0.9910252690315247, + "learning_rate": 5.266666666666667e-06, + "loss": 0.2096, + "step": 80 + }, + { + "epoch": 0.056532663316582916, + "grad_norm": 1.1305606365203857, + "learning_rate": 5.933333333333334e-06, + "loss": 0.1691, + "step": 90 + }, + { + "epoch": 0.06281407035175879, + "grad_norm": 1.2873859405517578, + "learning_rate": 6.6e-06, + "loss": 0.165, + "step": 100 + }, + { + "epoch": 0.06909547738693467, + "grad_norm": 1.5585795640945435, + "learning_rate": 7.266666666666668e-06, + "loss": 0.157, + "step": 110 + }, + { + "epoch": 0.07537688442211055, + "grad_norm": 0.9218589067459106, + "learning_rate": 7.933333333333334e-06, + "loss": 0.1343, + "step": 120 + }, + { + "epoch": 0.08165829145728644, + "grad_norm": 0.9817622900009155, + "learning_rate": 8.599999999999999e-06, + "loss": 0.1294, + "step": 130 + }, + { + "epoch": 0.08793969849246232, + "grad_norm": 0.9538266062736511, + "learning_rate": 9.266666666666667e-06, + "loss": 0.1164, + "step": 140 + }, + { + "epoch": 0.0942211055276382, + "grad_norm": 1.0492459535598755, + "learning_rate": 9.933333333333334e-06, + "loss": 0.1174, + "step": 150 + }, + { + "epoch": 0.10050251256281408, + "grad_norm": 1.4293967485427856, + "learning_rate": 1.06e-05, + "loss": 0.107, + "step": 160 + }, + { + "epoch": 0.10678391959798995, + "grad_norm": 1.2728917598724365, + "learning_rate": 1.1266666666666667e-05, + "loss": 0.105, + "step": 170 + }, + { + "epoch": 0.11306532663316583, + "grad_norm": 1.2511484622955322, + "learning_rate": 1.1933333333333333e-05, + "loss": 0.104, + "step": 180 + }, + { + "epoch": 0.11934673366834171, + "grad_norm": 0.7011515498161316, + "learning_rate": 1.2600000000000001e-05, + "loss": 0.0936, + "step": 190 + }, + { + "epoch": 0.12562814070351758, + "grad_norm": 0.7602205276489258, + "learning_rate": 1.3266666666666666e-05, + "loss": 0.0972, + "step": 200 + }, + { + "epoch": 0.13190954773869346, + "grad_norm": 0.7420077323913574, + "learning_rate": 1.3933333333333334e-05, + "loss": 0.0995, + "step": 210 + }, + { + "epoch": 0.13819095477386933, + "grad_norm": 0.6092446446418762, + "learning_rate": 1.4599999999999999e-05, + "loss": 0.095, + "step": 220 + }, + { + "epoch": 0.1444723618090452, + "grad_norm": 0.6789249181747437, + "learning_rate": 1.5266666666666667e-05, + "loss": 0.0839, + "step": 230 + }, + { + "epoch": 0.1507537688442211, + "grad_norm": 0.913944661617279, + "learning_rate": 1.5933333333333332e-05, + "loss": 0.0879, + "step": 240 + }, + { + "epoch": 0.157035175879397, + "grad_norm": 0.8573445677757263, + "learning_rate": 1.66e-05, + "loss": 0.0845, + "step": 250 + }, + { + "epoch": 0.16331658291457288, + "grad_norm": 1.0646635293960571, + "learning_rate": 1.726666666666667e-05, + "loss": 0.0838, + "step": 260 + }, + { + "epoch": 0.16959798994974876, + "grad_norm": 1.1676357984542847, + "learning_rate": 1.7933333333333337e-05, + "loss": 0.0889, + "step": 270 + }, + { + "epoch": 0.17587939698492464, + "grad_norm": 0.9000412225723267, + "learning_rate": 1.86e-05, + "loss": 0.0787, + "step": 280 + }, + { + "epoch": 0.18216080402010051, + "grad_norm": 0.9253113269805908, + "learning_rate": 1.926666666666667e-05, + "loss": 0.0856, + "step": 290 + }, + { + "epoch": 0.1884422110552764, + "grad_norm": 0.9942055344581604, + "learning_rate": 1.9933333333333334e-05, + "loss": 0.0735, + "step": 300 + }, + { + "epoch": 0.19472361809045227, + "grad_norm": 1.0392463207244873, + "learning_rate": 2.06e-05, + "loss": 0.0833, + "step": 310 + }, + { + "epoch": 0.20100502512562815, + "grad_norm": 0.6066302061080933, + "learning_rate": 2.1266666666666667e-05, + "loss": 0.075, + "step": 320 + }, + { + "epoch": 0.20728643216080403, + "grad_norm": 0.825148344039917, + "learning_rate": 2.1933333333333332e-05, + "loss": 0.0751, + "step": 330 + }, + { + "epoch": 0.2135678391959799, + "grad_norm": 0.6892251968383789, + "learning_rate": 2.26e-05, + "loss": 0.0719, + "step": 340 + }, + { + "epoch": 0.2198492462311558, + "grad_norm": 0.8216244578361511, + "learning_rate": 2.326666666666667e-05, + "loss": 0.0676, + "step": 350 + }, + { + "epoch": 0.22613065326633167, + "grad_norm": 1.0715235471725464, + "learning_rate": 2.3933333333333337e-05, + "loss": 0.0709, + "step": 360 + }, + { + "epoch": 0.23241206030150754, + "grad_norm": 0.9699908494949341, + "learning_rate": 2.46e-05, + "loss": 0.0714, + "step": 370 + }, + { + "epoch": 0.23869346733668342, + "grad_norm": 0.8490661382675171, + "learning_rate": 2.5266666666666666e-05, + "loss": 0.0686, + "step": 380 + }, + { + "epoch": 0.2449748743718593, + "grad_norm": 0.9879564046859741, + "learning_rate": 2.5933333333333338e-05, + "loss": 0.0756, + "step": 390 + }, + { + "epoch": 0.25125628140703515, + "grad_norm": 0.8191818594932556, + "learning_rate": 2.6600000000000003e-05, + "loss": 0.0745, + "step": 400 + }, + { + "epoch": 0.25753768844221103, + "grad_norm": 1.0145955085754395, + "learning_rate": 2.7266666666666668e-05, + "loss": 0.0758, + "step": 410 + }, + { + "epoch": 0.2638190954773869, + "grad_norm": 0.9888347387313843, + "learning_rate": 2.7933333333333332e-05, + "loss": 0.0746, + "step": 420 + }, + { + "epoch": 0.2701005025125628, + "grad_norm": 1.241927146911621, + "learning_rate": 2.86e-05, + "loss": 0.0675, + "step": 430 + }, + { + "epoch": 0.27638190954773867, + "grad_norm": 0.6055115461349487, + "learning_rate": 2.926666666666667e-05, + "loss": 0.063, + "step": 440 + }, + { + "epoch": 0.28266331658291455, + "grad_norm": 0.7960091829299927, + "learning_rate": 2.9933333333333337e-05, + "loss": 0.0617, + "step": 450 + }, + { + "epoch": 0.2889447236180904, + "grad_norm": 0.9290257692337036, + "learning_rate": 3.06e-05, + "loss": 0.068, + "step": 460 + }, + { + "epoch": 0.2952261306532663, + "grad_norm": 1.0968607664108276, + "learning_rate": 3.126666666666666e-05, + "loss": 0.0713, + "step": 470 + }, + { + "epoch": 0.3015075376884422, + "grad_norm": 1.0905603170394897, + "learning_rate": 3.1933333333333335e-05, + "loss": 0.0685, + "step": 480 + }, + { + "epoch": 0.30778894472361806, + "grad_norm": 0.7062813639640808, + "learning_rate": 3.26e-05, + "loss": 0.0652, + "step": 490 + }, + { + "epoch": 0.314070351758794, + "grad_norm": 0.5975621938705444, + "learning_rate": 3.326666666666667e-05, + "loss": 0.0644, + "step": 500 + }, + { + "epoch": 0.3203517587939699, + "grad_norm": 1.0010426044464111, + "learning_rate": 3.3933333333333336e-05, + "loss": 0.0695, + "step": 510 + }, + { + "epoch": 0.32663316582914576, + "grad_norm": 0.7888609170913696, + "learning_rate": 3.46e-05, + "loss": 0.0623, + "step": 520 + }, + { + "epoch": 0.33291457286432163, + "grad_norm": 0.5556163787841797, + "learning_rate": 3.526666666666667e-05, + "loss": 0.0647, + "step": 530 + }, + { + "epoch": 0.3391959798994975, + "grad_norm": 0.6649987101554871, + "learning_rate": 3.593333333333334e-05, + "loss": 0.0623, + "step": 540 + }, + { + "epoch": 0.3454773869346734, + "grad_norm": 0.47039103507995605, + "learning_rate": 3.66e-05, + "loss": 0.0604, + "step": 550 + }, + { + "epoch": 0.35175879396984927, + "grad_norm": 0.6532632112503052, + "learning_rate": 3.726666666666667e-05, + "loss": 0.0615, + "step": 560 + }, + { + "epoch": 0.35804020100502515, + "grad_norm": 0.5929540395736694, + "learning_rate": 3.793333333333334e-05, + "loss": 0.0555, + "step": 570 + }, + { + "epoch": 0.36432160804020103, + "grad_norm": 0.7575441002845764, + "learning_rate": 3.86e-05, + "loss": 0.0622, + "step": 580 + }, + { + "epoch": 0.3706030150753769, + "grad_norm": 1.1008965969085693, + "learning_rate": 3.926666666666667e-05, + "loss": 0.0668, + "step": 590 + }, + { + "epoch": 0.3768844221105528, + "grad_norm": 0.658219575881958, + "learning_rate": 3.993333333333333e-05, + "loss": 0.0613, + "step": 600 + }, + { + "epoch": 0.38316582914572866, + "grad_norm": 1.0479565858840942, + "learning_rate": 4.0600000000000004e-05, + "loss": 0.066, + "step": 610 + }, + { + "epoch": 0.38944723618090454, + "grad_norm": 0.662824809551239, + "learning_rate": 4.126666666666667e-05, + "loss": 0.0627, + "step": 620 + }, + { + "epoch": 0.3957286432160804, + "grad_norm": 0.895227313041687, + "learning_rate": 4.1933333333333334e-05, + "loss": 0.0498, + "step": 630 + }, + { + "epoch": 0.4020100502512563, + "grad_norm": 0.8258369565010071, + "learning_rate": 4.26e-05, + "loss": 0.0551, + "step": 640 + }, + { + "epoch": 0.4082914572864322, + "grad_norm": 0.6400461792945862, + "learning_rate": 4.3266666666666664e-05, + "loss": 0.0584, + "step": 650 + }, + { + "epoch": 0.41457286432160806, + "grad_norm": 0.9474914073944092, + "learning_rate": 4.3933333333333335e-05, + "loss": 0.0586, + "step": 660 + }, + { + "epoch": 0.42085427135678394, + "grad_norm": 0.8004749417304993, + "learning_rate": 4.46e-05, + "loss": 0.0628, + "step": 670 + }, + { + "epoch": 0.4271356783919598, + "grad_norm": 0.7029049396514893, + "learning_rate": 4.526666666666667e-05, + "loss": 0.0611, + "step": 680 + }, + { + "epoch": 0.4334170854271357, + "grad_norm": 0.4970827102661133, + "learning_rate": 4.5933333333333336e-05, + "loss": 0.0542, + "step": 690 + }, + { + "epoch": 0.4396984924623116, + "grad_norm": 0.6710481643676758, + "learning_rate": 4.660000000000001e-05, + "loss": 0.0569, + "step": 700 + }, + { + "epoch": 0.44597989949748745, + "grad_norm": 0.7602967619895935, + "learning_rate": 4.726666666666667e-05, + "loss": 0.0636, + "step": 710 + }, + { + "epoch": 0.45226130653266333, + "grad_norm": 0.7425283193588257, + "learning_rate": 4.793333333333334e-05, + "loss": 0.0525, + "step": 720 + }, + { + "epoch": 0.4585427135678392, + "grad_norm": 0.39933669567108154, + "learning_rate": 4.86e-05, + "loss": 0.0589, + "step": 730 + }, + { + "epoch": 0.4648241206030151, + "grad_norm": 0.5321558117866516, + "learning_rate": 4.926666666666667e-05, + "loss": 0.0469, + "step": 740 + }, + { + "epoch": 0.47110552763819097, + "grad_norm": 0.7330015301704407, + "learning_rate": 4.993333333333334e-05, + "loss": 0.0555, + "step": 750 + }, + { + "epoch": 0.47738693467336685, + "grad_norm": 0.6829407215118408, + "learning_rate": 5.0600000000000003e-05, + "loss": 0.0529, + "step": 760 + }, + { + "epoch": 0.4836683417085427, + "grad_norm": 0.7489681839942932, + "learning_rate": 5.1266666666666675e-05, + "loss": 0.0559, + "step": 770 + }, + { + "epoch": 0.4899497487437186, + "grad_norm": 0.6877656579017639, + "learning_rate": 5.193333333333333e-05, + "loss": 0.0501, + "step": 780 + }, + { + "epoch": 0.4962311557788945, + "grad_norm": 0.6515507698059082, + "learning_rate": 5.2600000000000005e-05, + "loss": 0.052, + "step": 790 + }, + { + "epoch": 0.5025125628140703, + "grad_norm": 0.43258044123649597, + "learning_rate": 5.326666666666666e-05, + "loss": 0.0513, + "step": 800 + }, + { + "epoch": 0.5087939698492462, + "grad_norm": 0.5532262325286865, + "learning_rate": 5.3933333333333334e-05, + "loss": 0.0485, + "step": 810 + }, + { + "epoch": 0.5150753768844221, + "grad_norm": 0.6552866101264954, + "learning_rate": 5.4600000000000006e-05, + "loss": 0.0553, + "step": 820 + }, + { + "epoch": 0.5213567839195979, + "grad_norm": 0.6157335638999939, + "learning_rate": 5.5266666666666664e-05, + "loss": 0.0518, + "step": 830 + }, + { + "epoch": 0.5276381909547738, + "grad_norm": 0.724036693572998, + "learning_rate": 5.5933333333333335e-05, + "loss": 0.0527, + "step": 840 + }, + { + "epoch": 0.5339195979899497, + "grad_norm": 1.0251004695892334, + "learning_rate": 5.66e-05, + "loss": 0.0654, + "step": 850 + }, + { + "epoch": 0.5402010050251256, + "grad_norm": 0.6391533017158508, + "learning_rate": 5.726666666666667e-05, + "loss": 0.0587, + "step": 860 + }, + { + "epoch": 0.5464824120603015, + "grad_norm": 0.6915408968925476, + "learning_rate": 5.7933333333333337e-05, + "loss": 0.0536, + "step": 870 + }, + { + "epoch": 0.5527638190954773, + "grad_norm": 0.7550404071807861, + "learning_rate": 5.86e-05, + "loss": 0.0477, + "step": 880 + }, + { + "epoch": 0.5590452261306532, + "grad_norm": 0.5762304663658142, + "learning_rate": 5.926666666666667e-05, + "loss": 0.0504, + "step": 890 + }, + { + "epoch": 0.5653266331658291, + "grad_norm": 0.5115172863006592, + "learning_rate": 5.9933333333333345e-05, + "loss": 0.0555, + "step": 900 + }, + { + "epoch": 0.571608040201005, + "grad_norm": 0.534849226474762, + "learning_rate": 6.06e-05, + "loss": 0.0484, + "step": 910 + }, + { + "epoch": 0.5778894472361809, + "grad_norm": 0.8149421811103821, + "learning_rate": 6.126666666666667e-05, + "loss": 0.0549, + "step": 920 + }, + { + "epoch": 0.5841708542713567, + "grad_norm": 0.7693353891372681, + "learning_rate": 6.193333333333333e-05, + "loss": 0.0561, + "step": 930 + }, + { + "epoch": 0.5904522613065326, + "grad_norm": 0.758346438407898, + "learning_rate": 6.26e-05, + "loss": 0.0527, + "step": 940 + }, + { + "epoch": 0.5967336683417085, + "grad_norm": 0.6447525024414062, + "learning_rate": 6.326666666666667e-05, + "loss": 0.0582, + "step": 950 + }, + { + "epoch": 0.6030150753768844, + "grad_norm": 0.6217967867851257, + "learning_rate": 6.393333333333333e-05, + "loss": 0.0457, + "step": 960 + }, + { + "epoch": 0.6092964824120602, + "grad_norm": 0.5512794256210327, + "learning_rate": 6.460000000000001e-05, + "loss": 0.0463, + "step": 970 + }, + { + "epoch": 0.6155778894472361, + "grad_norm": 0.717473030090332, + "learning_rate": 6.526666666666666e-05, + "loss": 0.0525, + "step": 980 + }, + { + "epoch": 0.621859296482412, + "grad_norm": 0.49841171503067017, + "learning_rate": 6.593333333333334e-05, + "loss": 0.051, + "step": 990 + }, + { + "epoch": 0.628140703517588, + "grad_norm": 0.6388250589370728, + "learning_rate": 6.66e-05, + "loss": 0.0521, + "step": 1000 + }, + { + "epoch": 0.6344221105527639, + "grad_norm": 0.6067572832107544, + "learning_rate": 6.726666666666667e-05, + "loss": 0.0499, + "step": 1010 + }, + { + "epoch": 0.6407035175879398, + "grad_norm": 0.7342574000358582, + "learning_rate": 6.793333333333334e-05, + "loss": 0.0509, + "step": 1020 + }, + { + "epoch": 0.6469849246231156, + "grad_norm": 0.7634968161582947, + "learning_rate": 6.860000000000001e-05, + "loss": 0.0479, + "step": 1030 + }, + { + "epoch": 0.6532663316582915, + "grad_norm": 0.707443356513977, + "learning_rate": 6.926666666666667e-05, + "loss": 0.0501, + "step": 1040 + }, + { + "epoch": 0.6595477386934674, + "grad_norm": 0.7503585815429688, + "learning_rate": 6.993333333333334e-05, + "loss": 0.0486, + "step": 1050 + }, + { + "epoch": 0.6658291457286433, + "grad_norm": 0.5248116850852966, + "learning_rate": 7.06e-05, + "loss": 0.0612, + "step": 1060 + }, + { + "epoch": 0.6721105527638191, + "grad_norm": 0.8077252507209778, + "learning_rate": 7.126666666666667e-05, + "loss": 0.0488, + "step": 1070 + }, + { + "epoch": 0.678391959798995, + "grad_norm": 0.43453532457351685, + "learning_rate": 7.193333333333334e-05, + "loss": 0.0506, + "step": 1080 + }, + { + "epoch": 0.6846733668341709, + "grad_norm": 0.4878779947757721, + "learning_rate": 7.26e-05, + "loss": 0.0427, + "step": 1090 + }, + { + "epoch": 0.6909547738693468, + "grad_norm": 0.32926174998283386, + "learning_rate": 7.326666666666667e-05, + "loss": 0.047, + "step": 1100 + }, + { + "epoch": 0.6972361809045227, + "grad_norm": 0.43741583824157715, + "learning_rate": 7.393333333333333e-05, + "loss": 0.0466, + "step": 1110 + }, + { + "epoch": 0.7035175879396985, + "grad_norm": 0.6949731707572937, + "learning_rate": 7.46e-05, + "loss": 0.0477, + "step": 1120 + }, + { + "epoch": 0.7097989949748744, + "grad_norm": 0.7657146453857422, + "learning_rate": 7.526666666666668e-05, + "loss": 0.0531, + "step": 1130 + }, + { + "epoch": 0.7160804020100503, + "grad_norm": 0.3714277148246765, + "learning_rate": 7.593333333333334e-05, + "loss": 0.0529, + "step": 1140 + }, + { + "epoch": 0.7223618090452262, + "grad_norm": 0.36182713508605957, + "learning_rate": 7.66e-05, + "loss": 0.0465, + "step": 1150 + }, + { + "epoch": 0.7286432160804021, + "grad_norm": 0.6097015142440796, + "learning_rate": 7.726666666666667e-05, + "loss": 0.0472, + "step": 1160 + }, + { + "epoch": 0.7349246231155779, + "grad_norm": 0.5928237438201904, + "learning_rate": 7.793333333333333e-05, + "loss": 0.049, + "step": 1170 + }, + { + "epoch": 0.7412060301507538, + "grad_norm": 0.4496796131134033, + "learning_rate": 7.860000000000001e-05, + "loss": 0.0458, + "step": 1180 + }, + { + "epoch": 0.7474874371859297, + "grad_norm": 0.6489536762237549, + "learning_rate": 7.926666666666666e-05, + "loss": 0.0475, + "step": 1190 + }, + { + "epoch": 0.7537688442211056, + "grad_norm": 0.7015570402145386, + "learning_rate": 7.993333333333334e-05, + "loss": 0.0502, + "step": 1200 + }, + { + "epoch": 0.7600502512562815, + "grad_norm": 0.6302024126052856, + "learning_rate": 8.060000000000001e-05, + "loss": 0.0493, + "step": 1210 + }, + { + "epoch": 0.7663316582914573, + "grad_norm": 0.7938515543937683, + "learning_rate": 8.126666666666667e-05, + "loss": 0.0451, + "step": 1220 + }, + { + "epoch": 0.7726130653266332, + "grad_norm": 0.75667804479599, + "learning_rate": 8.193333333333334e-05, + "loss": 0.0484, + "step": 1230 + }, + { + "epoch": 0.7788944723618091, + "grad_norm": 0.8149307370185852, + "learning_rate": 8.26e-05, + "loss": 0.044, + "step": 1240 + }, + { + "epoch": 0.785175879396985, + "grad_norm": 0.7652427554130554, + "learning_rate": 8.326666666666667e-05, + "loss": 0.0493, + "step": 1250 + }, + { + "epoch": 0.7914572864321608, + "grad_norm": 0.4265637993812561, + "learning_rate": 8.393333333333335e-05, + "loss": 0.0479, + "step": 1260 + }, + { + "epoch": 0.7977386934673367, + "grad_norm": 0.4759034514427185, + "learning_rate": 8.46e-05, + "loss": 0.0504, + "step": 1270 + }, + { + "epoch": 0.8040201005025126, + "grad_norm": 0.5574660301208496, + "learning_rate": 8.526666666666667e-05, + "loss": 0.0521, + "step": 1280 + }, + { + "epoch": 0.8103015075376885, + "grad_norm": 0.8064752817153931, + "learning_rate": 8.593333333333333e-05, + "loss": 0.0582, + "step": 1290 + }, + { + "epoch": 0.8165829145728644, + "grad_norm": 0.5513815879821777, + "learning_rate": 8.66e-05, + "loss": 0.0549, + "step": 1300 + }, + { + "epoch": 0.8228643216080402, + "grad_norm": 0.5721848607063293, + "learning_rate": 8.726666666666667e-05, + "loss": 0.045, + "step": 1310 + }, + { + "epoch": 0.8291457286432161, + "grad_norm": 0.7926440238952637, + "learning_rate": 8.793333333333333e-05, + "loss": 0.0565, + "step": 1320 + }, + { + "epoch": 0.835427135678392, + "grad_norm": 0.7132392525672913, + "learning_rate": 8.86e-05, + "loss": 0.0518, + "step": 1330 + }, + { + "epoch": 0.8417085427135679, + "grad_norm": 0.4522932469844818, + "learning_rate": 8.926666666666668e-05, + "loss": 0.0487, + "step": 1340 + }, + { + "epoch": 0.8479899497487438, + "grad_norm": 0.7226340770721436, + "learning_rate": 8.993333333333334e-05, + "loss": 0.0442, + "step": 1350 + }, + { + "epoch": 0.8542713567839196, + "grad_norm": 0.5771322846412659, + "learning_rate": 9.06e-05, + "loss": 0.0527, + "step": 1360 + }, + { + "epoch": 0.8605527638190955, + "grad_norm": 0.6682405471801758, + "learning_rate": 9.126666666666667e-05, + "loss": 0.0464, + "step": 1370 + }, + { + "epoch": 0.8668341708542714, + "grad_norm": 0.32375437021255493, + "learning_rate": 9.193333333333334e-05, + "loss": 0.0486, + "step": 1380 + }, + { + "epoch": 0.8731155778894473, + "grad_norm": 0.6192367076873779, + "learning_rate": 9.260000000000001e-05, + "loss": 0.0453, + "step": 1390 + }, + { + "epoch": 0.8793969849246231, + "grad_norm": 0.7526844143867493, + "learning_rate": 9.326666666666667e-05, + "loss": 0.0514, + "step": 1400 + }, + { + "epoch": 0.885678391959799, + "grad_norm": 0.635410726070404, + "learning_rate": 9.393333333333334e-05, + "loss": 0.045, + "step": 1410 + }, + { + "epoch": 0.8919597989949749, + "grad_norm": 0.3692018389701843, + "learning_rate": 9.46e-05, + "loss": 0.05, + "step": 1420 + }, + { + "epoch": 0.8982412060301508, + "grad_norm": 0.9415658712387085, + "learning_rate": 9.526666666666667e-05, + "loss": 0.0499, + "step": 1430 + }, + { + "epoch": 0.9045226130653267, + "grad_norm": 0.5795928239822388, + "learning_rate": 9.593333333333334e-05, + "loss": 0.0472, + "step": 1440 + }, + { + "epoch": 0.9108040201005025, + "grad_norm": 0.3640764355659485, + "learning_rate": 9.66e-05, + "loss": 0.048, + "step": 1450 + }, + { + "epoch": 0.9170854271356784, + "grad_norm": 0.687280535697937, + "learning_rate": 9.726666666666667e-05, + "loss": 0.0464, + "step": 1460 + }, + { + "epoch": 0.9233668341708543, + "grad_norm": 0.6024627089500427, + "learning_rate": 9.793333333333333e-05, + "loss": 0.0488, + "step": 1470 + }, + { + "epoch": 0.9296482412060302, + "grad_norm": 0.7639083862304688, + "learning_rate": 9.86e-05, + "loss": 0.0539, + "step": 1480 + }, + { + "epoch": 0.9359296482412061, + "grad_norm": 0.6096637845039368, + "learning_rate": 9.926666666666668e-05, + "loss": 0.0448, + "step": 1490 + }, + { + "epoch": 0.9422110552763819, + "grad_norm": 0.6019980907440186, + "learning_rate": 9.993333333333334e-05, + "loss": 0.057, + "step": 1500 + }, + { + "epoch": 0.9484924623115578, + "grad_norm": 0.5511391162872314, + "learning_rate": 9.999997539434007e-05, + "loss": 0.045, + "step": 1510 + }, + { + "epoch": 0.9547738693467337, + "grad_norm": 0.6483543515205383, + "learning_rate": 9.999989033776898e-05, + "loss": 0.0474, + "step": 1520 + }, + { + "epoch": 0.9610552763819096, + "grad_norm": 0.36361801624298096, + "learning_rate": 9.999974452661641e-05, + "loss": 0.046, + "step": 1530 + }, + { + "epoch": 0.9673366834170855, + "grad_norm": 0.6088789701461792, + "learning_rate": 9.999953796105959e-05, + "loss": 0.0506, + "step": 1540 + }, + { + "epoch": 0.9736180904522613, + "grad_norm": 0.35969704389572144, + "learning_rate": 9.999927064134949e-05, + "loss": 0.0473, + "step": 1550 + }, + { + "epoch": 0.9798994974874372, + "grad_norm": 0.5269743800163269, + "learning_rate": 9.999894256781095e-05, + "loss": 0.0476, + "step": 1560 + }, + { + "epoch": 0.9861809045226131, + "grad_norm": 0.6000348925590515, + "learning_rate": 9.99985537408426e-05, + "loss": 0.044, + "step": 1570 + }, + { + "epoch": 0.992462311557789, + "grad_norm": 0.5426010489463806, + "learning_rate": 9.999810416091688e-05, + "loss": 0.0427, + "step": 1580 + }, + { + "epoch": 0.9987437185929648, + "grad_norm": 0.3993867337703705, + "learning_rate": 9.99975938285801e-05, + "loss": 0.0477, + "step": 1590 + }, + { + "epoch": 1.0050251256281406, + "grad_norm": 0.49689656496047974, + "learning_rate": 9.999702274445236e-05, + "loss": 0.0408, + "step": 1600 + }, + { + "epoch": 1.0113065326633166, + "grad_norm": 0.6569452285766602, + "learning_rate": 9.999639090922756e-05, + "loss": 0.0458, + "step": 1610 + }, + { + "epoch": 1.0175879396984924, + "grad_norm": 0.6446689367294312, + "learning_rate": 9.999569832367346e-05, + "loss": 0.0413, + "step": 1620 + }, + { + "epoch": 1.0238693467336684, + "grad_norm": 0.593984842300415, + "learning_rate": 9.999494498863162e-05, + "loss": 0.0421, + "step": 1630 + }, + { + "epoch": 1.0301507537688441, + "grad_norm": 0.43230894207954407, + "learning_rate": 9.99941309050174e-05, + "loss": 0.0437, + "step": 1640 + }, + { + "epoch": 1.0364321608040201, + "grad_norm": 0.7140646576881409, + "learning_rate": 9.999325607381999e-05, + "loss": 0.0501, + "step": 1650 + }, + { + "epoch": 1.0427135678391959, + "grad_norm": 0.5066368579864502, + "learning_rate": 9.999232049610238e-05, + "loss": 0.0399, + "step": 1660 + }, + { + "epoch": 1.0489949748743719, + "grad_norm": 0.43402788043022156, + "learning_rate": 9.999132417300142e-05, + "loss": 0.0421, + "step": 1670 + }, + { + "epoch": 1.0552763819095476, + "grad_norm": 0.6031082272529602, + "learning_rate": 9.99902671057277e-05, + "loss": 0.0426, + "step": 1680 + }, + { + "epoch": 1.0615577889447236, + "grad_norm": 0.5212262868881226, + "learning_rate": 9.998914929556569e-05, + "loss": 0.0447, + "step": 1690 + }, + { + "epoch": 1.0678391959798994, + "grad_norm": 0.6960680484771729, + "learning_rate": 9.998797074387361e-05, + "loss": 0.0428, + "step": 1700 + }, + { + "epoch": 1.0741206030150754, + "grad_norm": 0.8098211288452148, + "learning_rate": 9.99867314520835e-05, + "loss": 0.0474, + "step": 1710 + }, + { + "epoch": 1.0804020100502512, + "grad_norm": 0.523608386516571, + "learning_rate": 9.998543142170126e-05, + "loss": 0.0471, + "step": 1720 + }, + { + "epoch": 1.0866834170854272, + "grad_norm": 0.4902782738208771, + "learning_rate": 9.99840706543065e-05, + "loss": 0.0405, + "step": 1730 + }, + { + "epoch": 1.092964824120603, + "grad_norm": 0.36833465099334717, + "learning_rate": 9.998264915155274e-05, + "loss": 0.0428, + "step": 1740 + }, + { + "epoch": 1.099246231155779, + "grad_norm": 0.43600478768348694, + "learning_rate": 9.998116691516718e-05, + "loss": 0.0453, + "step": 1750 + }, + { + "epoch": 1.1055276381909547, + "grad_norm": 0.5431431531906128, + "learning_rate": 9.997962394695091e-05, + "loss": 0.0457, + "step": 1760 + }, + { + "epoch": 1.1118090452261307, + "grad_norm": 0.36344292759895325, + "learning_rate": 9.997802024877875e-05, + "loss": 0.043, + "step": 1770 + }, + { + "epoch": 1.1180904522613064, + "grad_norm": 0.35352572798728943, + "learning_rate": 9.99763558225994e-05, + "loss": 0.044, + "step": 1780 + }, + { + "epoch": 1.1243718592964824, + "grad_norm": 0.5377246141433716, + "learning_rate": 9.997463067043526e-05, + "loss": 0.0457, + "step": 1790 + }, + { + "epoch": 1.1306532663316582, + "grad_norm": 0.4468127489089966, + "learning_rate": 9.997284479438253e-05, + "loss": 0.042, + "step": 1800 + }, + { + "epoch": 1.1369346733668342, + "grad_norm": 0.33781641721725464, + "learning_rate": 9.997099819661127e-05, + "loss": 0.0461, + "step": 1810 + }, + { + "epoch": 1.1432160804020102, + "grad_norm": 0.6892178058624268, + "learning_rate": 9.996909087936524e-05, + "loss": 0.0442, + "step": 1820 + }, + { + "epoch": 1.149497487437186, + "grad_norm": 0.73509681224823, + "learning_rate": 9.996712284496201e-05, + "loss": 0.0438, + "step": 1830 + }, + { + "epoch": 1.1557788944723617, + "grad_norm": 0.6080371737480164, + "learning_rate": 9.996509409579293e-05, + "loss": 0.0454, + "step": 1840 + }, + { + "epoch": 1.1620603015075377, + "grad_norm": 0.6413587331771851, + "learning_rate": 9.996300463432312e-05, + "loss": 0.0461, + "step": 1850 + }, + { + "epoch": 1.1683417085427137, + "grad_norm": 0.6291435956954956, + "learning_rate": 9.996085446309148e-05, + "loss": 0.0466, + "step": 1860 + }, + { + "epoch": 1.1746231155778895, + "grad_norm": 0.4369809329509735, + "learning_rate": 9.995864358471066e-05, + "loss": 0.0442, + "step": 1870 + }, + { + "epoch": 1.1809045226130652, + "grad_norm": 0.348025381565094, + "learning_rate": 9.99563720018671e-05, + "loss": 0.0405, + "step": 1880 + }, + { + "epoch": 1.1871859296482412, + "grad_norm": 0.6925215721130371, + "learning_rate": 9.995403971732098e-05, + "loss": 0.0368, + "step": 1890 + }, + { + "epoch": 1.1934673366834172, + "grad_norm": 0.3654880225658417, + "learning_rate": 9.995164673390625e-05, + "loss": 0.0452, + "step": 1900 + }, + { + "epoch": 1.199748743718593, + "grad_norm": 0.4873250424861908, + "learning_rate": 9.994919305453059e-05, + "loss": 0.046, + "step": 1910 + }, + { + "epoch": 1.2060301507537687, + "grad_norm": 0.6772422194480896, + "learning_rate": 9.994667868217548e-05, + "loss": 0.0474, + "step": 1920 + }, + { + "epoch": 1.2123115577889447, + "grad_norm": 0.473649263381958, + "learning_rate": 9.99441036198961e-05, + "loss": 0.0465, + "step": 1930 + }, + { + "epoch": 1.2185929648241207, + "grad_norm": 0.49097567796707153, + "learning_rate": 9.99414678708214e-05, + "loss": 0.0434, + "step": 1940 + }, + { + "epoch": 1.2248743718592965, + "grad_norm": 0.7209442257881165, + "learning_rate": 9.993877143815407e-05, + "loss": 0.0446, + "step": 1950 + }, + { + "epoch": 1.2311557788944723, + "grad_norm": 0.6855082511901855, + "learning_rate": 9.993601432517053e-05, + "loss": 0.0388, + "step": 1960 + }, + { + "epoch": 1.2374371859296482, + "grad_norm": 0.4037642776966095, + "learning_rate": 9.993319653522091e-05, + "loss": 0.0454, + "step": 1970 + }, + { + "epoch": 1.2437185929648242, + "grad_norm": 0.5641746520996094, + "learning_rate": 9.993031807172911e-05, + "loss": 0.0471, + "step": 1980 + }, + { + "epoch": 1.25, + "grad_norm": 0.7517688870429993, + "learning_rate": 9.992737893819273e-05, + "loss": 0.0448, + "step": 1990 + }, + { + "epoch": 1.2562814070351758, + "grad_norm": 0.6314425468444824, + "learning_rate": 9.992437913818312e-05, + "loss": 0.0412, + "step": 2000 + }, + { + "epoch": 1.2625628140703518, + "grad_norm": 0.4326089024543762, + "learning_rate": 9.992131867534526e-05, + "loss": 0.0409, + "step": 2010 + }, + { + "epoch": 1.2688442211055277, + "grad_norm": 0.42544814944267273, + "learning_rate": 9.991819755339796e-05, + "loss": 0.042, + "step": 2020 + }, + { + "epoch": 1.2751256281407035, + "grad_norm": 0.457621693611145, + "learning_rate": 9.991501577613365e-05, + "loss": 0.0494, + "step": 2030 + }, + { + "epoch": 1.2814070351758793, + "grad_norm": 0.7491344809532166, + "learning_rate": 9.99117733474185e-05, + "loss": 0.0476, + "step": 2040 + }, + { + "epoch": 1.2876884422110553, + "grad_norm": 0.4226253032684326, + "learning_rate": 9.990847027119234e-05, + "loss": 0.0387, + "step": 2050 + }, + { + "epoch": 1.2939698492462313, + "grad_norm": 0.3292161226272583, + "learning_rate": 9.990510655146877e-05, + "loss": 0.0382, + "step": 2060 + }, + { + "epoch": 1.300251256281407, + "grad_norm": 0.3090316951274872, + "learning_rate": 9.990168219233496e-05, + "loss": 0.0392, + "step": 2070 + }, + { + "epoch": 1.3065326633165828, + "grad_norm": 0.348040372133255, + "learning_rate": 9.989819719795188e-05, + "loss": 0.0392, + "step": 2080 + }, + { + "epoch": 1.3128140703517588, + "grad_norm": 0.3757565915584564, + "learning_rate": 9.989465157255412e-05, + "loss": 0.0379, + "step": 2090 + }, + { + "epoch": 1.3190954773869348, + "grad_norm": 0.38036489486694336, + "learning_rate": 9.989104532044994e-05, + "loss": 0.0357, + "step": 2100 + }, + { + "epoch": 1.3253768844221105, + "grad_norm": 0.4913692772388458, + "learning_rate": 9.988737844602128e-05, + "loss": 0.0413, + "step": 2110 + }, + { + "epoch": 1.3316582914572863, + "grad_norm": 0.3998088240623474, + "learning_rate": 9.988365095372372e-05, + "loss": 0.0382, + "step": 2120 + }, + { + "epoch": 1.3379396984924623, + "grad_norm": 0.5522318482398987, + "learning_rate": 9.987986284808654e-05, + "loss": 0.0425, + "step": 2130 + }, + { + "epoch": 1.3442211055276383, + "grad_norm": 0.8188551068305969, + "learning_rate": 9.987601413371264e-05, + "loss": 0.0463, + "step": 2140 + }, + { + "epoch": 1.350502512562814, + "grad_norm": 0.675620973110199, + "learning_rate": 9.987210481527855e-05, + "loss": 0.0508, + "step": 2150 + }, + { + "epoch": 1.3567839195979898, + "grad_norm": 0.5800641179084778, + "learning_rate": 9.98681348975345e-05, + "loss": 0.0387, + "step": 2160 + }, + { + "epoch": 1.3630653266331658, + "grad_norm": 0.41477686166763306, + "learning_rate": 9.986410438530427e-05, + "loss": 0.0407, + "step": 2170 + }, + { + "epoch": 1.3693467336683418, + "grad_norm": 0.5505456328392029, + "learning_rate": 9.986001328348534e-05, + "loss": 0.0408, + "step": 2180 + }, + { + "epoch": 1.3756281407035176, + "grad_norm": 0.6489054560661316, + "learning_rate": 9.985586159704878e-05, + "loss": 0.0438, + "step": 2190 + }, + { + "epoch": 1.3819095477386933, + "grad_norm": 0.5654657483100891, + "learning_rate": 9.985164933103929e-05, + "loss": 0.0434, + "step": 2200 + }, + { + "epoch": 1.3881909547738693, + "grad_norm": 0.6295890808105469, + "learning_rate": 9.984737649057513e-05, + "loss": 0.0371, + "step": 2210 + }, + { + "epoch": 1.3944723618090453, + "grad_norm": 0.5740566849708557, + "learning_rate": 9.984304308084827e-05, + "loss": 0.0356, + "step": 2220 + }, + { + "epoch": 1.400753768844221, + "grad_norm": 0.33665406703948975, + "learning_rate": 9.983864910712416e-05, + "loss": 0.0348, + "step": 2230 + }, + { + "epoch": 1.4070351758793969, + "grad_norm": 0.4991978406906128, + "learning_rate": 9.98341945747419e-05, + "loss": 0.0412, + "step": 2240 + }, + { + "epoch": 1.4133165829145728, + "grad_norm": 0.3742690086364746, + "learning_rate": 9.98296794891142e-05, + "loss": 0.0388, + "step": 2250 + }, + { + "epoch": 1.4195979899497488, + "grad_norm": 0.47544077038764954, + "learning_rate": 9.982510385572725e-05, + "loss": 0.0375, + "step": 2260 + }, + { + "epoch": 1.4258793969849246, + "grad_norm": 0.515170693397522, + "learning_rate": 9.982046768014094e-05, + "loss": 0.0408, + "step": 2270 + }, + { + "epoch": 1.4321608040201004, + "grad_norm": 0.4256803095340729, + "learning_rate": 9.981577096798863e-05, + "loss": 0.0397, + "step": 2280 + }, + { + "epoch": 1.4384422110552764, + "grad_norm": 0.5829164385795593, + "learning_rate": 9.981101372497727e-05, + "loss": 0.0421, + "step": 2290 + }, + { + "epoch": 1.4447236180904524, + "grad_norm": 0.35821732878685, + "learning_rate": 9.980619595688737e-05, + "loss": 0.0403, + "step": 2300 + }, + { + "epoch": 1.4510050251256281, + "grad_norm": 0.4592154622077942, + "learning_rate": 9.980131766957295e-05, + "loss": 0.0426, + "step": 2310 + }, + { + "epoch": 1.457286432160804, + "grad_norm": 0.41883620619773865, + "learning_rate": 9.979637886896163e-05, + "loss": 0.036, + "step": 2320 + }, + { + "epoch": 1.4635678391959799, + "grad_norm": 0.6471456289291382, + "learning_rate": 9.979137956105447e-05, + "loss": 0.0433, + "step": 2330 + }, + { + "epoch": 1.4698492462311559, + "grad_norm": 0.2853199541568756, + "learning_rate": 9.978631975192613e-05, + "loss": 0.0398, + "step": 2340 + }, + { + "epoch": 1.4761306532663316, + "grad_norm": 0.40830931067466736, + "learning_rate": 9.978119944772475e-05, + "loss": 0.0386, + "step": 2350 + }, + { + "epoch": 1.4824120603015074, + "grad_norm": 0.6145526766777039, + "learning_rate": 9.977601865467197e-05, + "loss": 0.0425, + "step": 2360 + }, + { + "epoch": 1.4886934673366834, + "grad_norm": 0.6156154870986938, + "learning_rate": 9.977077737906297e-05, + "loss": 0.04, + "step": 2370 + }, + { + "epoch": 1.4949748743718594, + "grad_norm": 0.3813233971595764, + "learning_rate": 9.976547562726636e-05, + "loss": 0.0375, + "step": 2380 + }, + { + "epoch": 1.5012562814070352, + "grad_norm": 0.37265413999557495, + "learning_rate": 9.976011340572429e-05, + "loss": 0.0439, + "step": 2390 + }, + { + "epoch": 1.507537688442211, + "grad_norm": 0.5291724801063538, + "learning_rate": 9.975469072095237e-05, + "loss": 0.0414, + "step": 2400 + }, + { + "epoch": 1.513819095477387, + "grad_norm": 0.3292425572872162, + "learning_rate": 9.974920757953965e-05, + "loss": 0.0351, + "step": 2410 + }, + { + "epoch": 1.520100502512563, + "grad_norm": 0.542716920375824, + "learning_rate": 9.97436639881487e-05, + "loss": 0.0378, + "step": 2420 + }, + { + "epoch": 1.5263819095477387, + "grad_norm": 0.35756492614746094, + "learning_rate": 9.973805995351545e-05, + "loss": 0.0315, + "step": 2430 + }, + { + "epoch": 1.5326633165829144, + "grad_norm": 0.582633912563324, + "learning_rate": 9.973239548244939e-05, + "loss": 0.0442, + "step": 2440 + }, + { + "epoch": 1.5389447236180904, + "grad_norm": 0.3422170579433441, + "learning_rate": 9.972667058183333e-05, + "loss": 0.0351, + "step": 2450 + }, + { + "epoch": 1.5452261306532664, + "grad_norm": 0.4990876019001007, + "learning_rate": 9.972088525862362e-05, + "loss": 0.0425, + "step": 2460 + }, + { + "epoch": 1.5515075376884422, + "grad_norm": 0.6402345299720764, + "learning_rate": 9.971503951984995e-05, + "loss": 0.0403, + "step": 2470 + }, + { + "epoch": 1.557788944723618, + "grad_norm": 0.24171121418476105, + "learning_rate": 9.970913337261543e-05, + "loss": 0.0376, + "step": 2480 + }, + { + "epoch": 1.564070351758794, + "grad_norm": 0.28990092873573303, + "learning_rate": 9.97031668240966e-05, + "loss": 0.0318, + "step": 2490 + }, + { + "epoch": 1.57035175879397, + "grad_norm": 0.47528430819511414, + "learning_rate": 9.969713988154339e-05, + "loss": 0.0375, + "step": 2500 + }, + { + "epoch": 1.5766331658291457, + "grad_norm": 0.45832890272140503, + "learning_rate": 9.969105255227906e-05, + "loss": 0.0365, + "step": 2510 + }, + { + "epoch": 1.5829145728643215, + "grad_norm": 0.4669021666049957, + "learning_rate": 9.968490484370035e-05, + "loss": 0.0333, + "step": 2520 + }, + { + "epoch": 1.5891959798994975, + "grad_norm": 0.3869768977165222, + "learning_rate": 9.967869676327726e-05, + "loss": 0.0326, + "step": 2530 + }, + { + "epoch": 1.5954773869346734, + "grad_norm": 0.3955910801887512, + "learning_rate": 9.96724283185532e-05, + "loss": 0.0361, + "step": 2540 + }, + { + "epoch": 1.6017587939698492, + "grad_norm": 0.40126585960388184, + "learning_rate": 9.966609951714494e-05, + "loss": 0.0436, + "step": 2550 + }, + { + "epoch": 1.608040201005025, + "grad_norm": 0.3339569866657257, + "learning_rate": 9.965971036674255e-05, + "loss": 0.0366, + "step": 2560 + }, + { + "epoch": 1.614321608040201, + "grad_norm": 0.2697726786136627, + "learning_rate": 9.965326087510947e-05, + "loss": 0.0377, + "step": 2570 + }, + { + "epoch": 1.620603015075377, + "grad_norm": 0.40801364183425903, + "learning_rate": 9.964675105008244e-05, + "loss": 0.0377, + "step": 2580 + }, + { + "epoch": 1.6268844221105527, + "grad_norm": 0.38616862893104553, + "learning_rate": 9.964018089957147e-05, + "loss": 0.0375, + "step": 2590 + }, + { + "epoch": 1.6331658291457285, + "grad_norm": 0.34115949273109436, + "learning_rate": 9.963355043155996e-05, + "loss": 0.0345, + "step": 2600 + }, + { + "epoch": 1.6394472361809045, + "grad_norm": 0.5191960334777832, + "learning_rate": 9.962685965410455e-05, + "loss": 0.0336, + "step": 2610 + }, + { + "epoch": 1.6457286432160805, + "grad_norm": 0.6063273549079895, + "learning_rate": 9.962010857533514e-05, + "loss": 0.0363, + "step": 2620 + }, + { + "epoch": 1.6520100502512562, + "grad_norm": 0.30862191319465637, + "learning_rate": 9.961329720345493e-05, + "loss": 0.0364, + "step": 2630 + }, + { + "epoch": 1.658291457286432, + "grad_norm": 0.3568628132343292, + "learning_rate": 9.96064255467404e-05, + "loss": 0.0319, + "step": 2640 + }, + { + "epoch": 1.664572864321608, + "grad_norm": 0.5367453694343567, + "learning_rate": 9.959949361354126e-05, + "loss": 0.038, + "step": 2650 + }, + { + "epoch": 1.670854271356784, + "grad_norm": 0.32836663722991943, + "learning_rate": 9.959250141228045e-05, + "loss": 0.0452, + "step": 2660 + }, + { + "epoch": 1.6771356783919598, + "grad_norm": 0.48674583435058594, + "learning_rate": 9.958544895145414e-05, + "loss": 0.0412, + "step": 2670 + }, + { + "epoch": 1.6834170854271355, + "grad_norm": 0.18520933389663696, + "learning_rate": 9.957833623963177e-05, + "loss": 0.0382, + "step": 2680 + }, + { + "epoch": 1.6896984924623115, + "grad_norm": 0.47650912404060364, + "learning_rate": 9.957116328545593e-05, + "loss": 0.0378, + "step": 2690 + }, + { + "epoch": 1.6959798994974875, + "grad_norm": 0.6011577844619751, + "learning_rate": 9.956393009764244e-05, + "loss": 0.0399, + "step": 2700 + }, + { + "epoch": 1.7022613065326633, + "grad_norm": 0.4028601348400116, + "learning_rate": 9.955663668498032e-05, + "loss": 0.0335, + "step": 2710 + }, + { + "epoch": 1.708542713567839, + "grad_norm": 0.6205360889434814, + "learning_rate": 9.954928305633173e-05, + "loss": 0.0366, + "step": 2720 + }, + { + "epoch": 1.714824120603015, + "grad_norm": 0.6682284474372864, + "learning_rate": 9.954186922063204e-05, + "loss": 0.0368, + "step": 2730 + }, + { + "epoch": 1.721105527638191, + "grad_norm": 0.3651426136493683, + "learning_rate": 9.953439518688974e-05, + "loss": 0.0408, + "step": 2740 + }, + { + "epoch": 1.7273869346733668, + "grad_norm": 0.506019651889801, + "learning_rate": 9.952686096418652e-05, + "loss": 0.0382, + "step": 2750 + }, + { + "epoch": 1.7336683417085426, + "grad_norm": 0.40663158893585205, + "learning_rate": 9.951926656167714e-05, + "loss": 0.0362, + "step": 2760 + }, + { + "epoch": 1.7399497487437185, + "grad_norm": 0.37695392966270447, + "learning_rate": 9.951161198858953e-05, + "loss": 0.033, + "step": 2770 + }, + { + "epoch": 1.7462311557788945, + "grad_norm": 0.485268235206604, + "learning_rate": 9.950389725422471e-05, + "loss": 0.0386, + "step": 2780 + }, + { + "epoch": 1.7525125628140703, + "grad_norm": 0.5020673871040344, + "learning_rate": 9.949612236795682e-05, + "loss": 0.0355, + "step": 2790 + }, + { + "epoch": 1.758793969849246, + "grad_norm": 0.45358121395111084, + "learning_rate": 9.948828733923305e-05, + "loss": 0.0369, + "step": 2800 + }, + { + "epoch": 1.765075376884422, + "grad_norm": 0.3243740499019623, + "learning_rate": 9.948039217757374e-05, + "loss": 0.0356, + "step": 2810 + }, + { + "epoch": 1.771356783919598, + "grad_norm": 0.31892311573028564, + "learning_rate": 9.947243689257225e-05, + "loss": 0.032, + "step": 2820 + }, + { + "epoch": 1.7776381909547738, + "grad_norm": 0.6157580018043518, + "learning_rate": 9.946442149389497e-05, + "loss": 0.0365, + "step": 2830 + }, + { + "epoch": 1.7839195979899496, + "grad_norm": 0.2818092405796051, + "learning_rate": 9.945634599128139e-05, + "loss": 0.0372, + "step": 2840 + }, + { + "epoch": 1.7902010050251256, + "grad_norm": 0.46666038036346436, + "learning_rate": 9.944821039454402e-05, + "loss": 0.0378, + "step": 2850 + }, + { + "epoch": 1.7964824120603016, + "grad_norm": 0.45188435912132263, + "learning_rate": 9.944001471356835e-05, + "loss": 0.0374, + "step": 2860 + }, + { + "epoch": 1.8027638190954773, + "grad_norm": 0.2433045506477356, + "learning_rate": 9.94317589583129e-05, + "loss": 0.0346, + "step": 2870 + }, + { + "epoch": 1.809045226130653, + "grad_norm": 0.424797385931015, + "learning_rate": 9.942344313880922e-05, + "loss": 0.0344, + "step": 2880 + }, + { + "epoch": 1.815326633165829, + "grad_norm": 0.41626161336898804, + "learning_rate": 9.941506726516179e-05, + "loss": 0.0355, + "step": 2890 + }, + { + "epoch": 1.821608040201005, + "grad_norm": 0.5162527561187744, + "learning_rate": 9.94066313475481e-05, + "loss": 0.038, + "step": 2900 + }, + { + "epoch": 1.8278894472361809, + "grad_norm": 0.5474295616149902, + "learning_rate": 9.939813539621857e-05, + "loss": 0.0467, + "step": 2910 + }, + { + "epoch": 1.8341708542713566, + "grad_norm": 0.602220356464386, + "learning_rate": 9.93895794214966e-05, + "loss": 0.0376, + "step": 2920 + }, + { + "epoch": 1.8404522613065326, + "grad_norm": 0.37179726362228394, + "learning_rate": 9.938096343377852e-05, + "loss": 0.0398, + "step": 2930 + }, + { + "epoch": 1.8467336683417086, + "grad_norm": 0.3597119152545929, + "learning_rate": 9.937228744353353e-05, + "loss": 0.0353, + "step": 2940 + }, + { + "epoch": 1.8530150753768844, + "grad_norm": 0.3195698857307434, + "learning_rate": 9.936355146130379e-05, + "loss": 0.0341, + "step": 2950 + }, + { + "epoch": 1.8592964824120601, + "grad_norm": 0.4336487948894501, + "learning_rate": 9.935475549770435e-05, + "loss": 0.0433, + "step": 2960 + }, + { + "epoch": 1.8655778894472361, + "grad_norm": 0.5204407572746277, + "learning_rate": 9.934589956342315e-05, + "loss": 0.0403, + "step": 2970 + }, + { + "epoch": 1.8718592964824121, + "grad_norm": 0.5571388006210327, + "learning_rate": 9.933698366922093e-05, + "loss": 0.0359, + "step": 2980 + }, + { + "epoch": 1.8781407035175879, + "grad_norm": 0.37410813570022583, + "learning_rate": 9.93280078259314e-05, + "loss": 0.0312, + "step": 2990 + }, + { + "epoch": 1.8844221105527639, + "grad_norm": 0.3582216799259186, + "learning_rate": 9.931897204446104e-05, + "loss": 0.0322, + "step": 3000 + }, + { + "epoch": 1.8907035175879399, + "grad_norm": 0.3906169533729553, + "learning_rate": 9.930987633578915e-05, + "loss": 0.0312, + "step": 3010 + }, + { + "epoch": 1.8969849246231156, + "grad_norm": 0.3865334391593933, + "learning_rate": 9.93007207109679e-05, + "loss": 0.0368, + "step": 3020 + }, + { + "epoch": 1.9032663316582914, + "grad_norm": 0.37340402603149414, + "learning_rate": 9.929150518112224e-05, + "loss": 0.0317, + "step": 3030 + }, + { + "epoch": 1.9095477386934674, + "grad_norm": 0.2541466951370239, + "learning_rate": 9.928222975744991e-05, + "loss": 0.0319, + "step": 3040 + }, + { + "epoch": 1.9158291457286434, + "grad_norm": 0.3276437520980835, + "learning_rate": 9.92728944512214e-05, + "loss": 0.0337, + "step": 3050 + }, + { + "epoch": 1.9221105527638191, + "grad_norm": 0.3015574514865875, + "learning_rate": 9.926349927378001e-05, + "loss": 0.0373, + "step": 3060 + }, + { + "epoch": 1.928391959798995, + "grad_norm": 0.49897170066833496, + "learning_rate": 9.925404423654174e-05, + "loss": 0.0385, + "step": 3070 + }, + { + "epoch": 1.934673366834171, + "grad_norm": 0.5457289814949036, + "learning_rate": 9.924452935099537e-05, + "loss": 0.0333, + "step": 3080 + }, + { + "epoch": 1.940954773869347, + "grad_norm": 0.458911269903183, + "learning_rate": 9.92349546287024e-05, + "loss": 0.0371, + "step": 3090 + }, + { + "epoch": 1.9472361809045227, + "grad_norm": 0.4206310510635376, + "learning_rate": 9.9225320081297e-05, + "loss": 0.0342, + "step": 3100 + }, + { + "epoch": 1.9535175879396984, + "grad_norm": 0.5345395803451538, + "learning_rate": 9.921562572048606e-05, + "loss": 0.0466, + "step": 3110 + }, + { + "epoch": 1.9597989949748744, + "grad_norm": 0.725232720375061, + "learning_rate": 9.920587155804913e-05, + "loss": 0.0393, + "step": 3120 + }, + { + "epoch": 1.9660804020100504, + "grad_norm": 0.4959801435470581, + "learning_rate": 9.919605760583845e-05, + "loss": 0.0366, + "step": 3130 + }, + { + "epoch": 1.9723618090452262, + "grad_norm": 0.2679874002933502, + "learning_rate": 9.91861838757789e-05, + "loss": 0.039, + "step": 3140 + }, + { + "epoch": 1.978643216080402, + "grad_norm": 0.6184228658676147, + "learning_rate": 9.917625037986798e-05, + "loss": 0.0353, + "step": 3150 + }, + { + "epoch": 1.984924623115578, + "grad_norm": 0.44256457686424255, + "learning_rate": 9.916625713017583e-05, + "loss": 0.0337, + "step": 3160 + }, + { + "epoch": 1.991206030150754, + "grad_norm": 0.45629993081092834, + "learning_rate": 9.915620413884519e-05, + "loss": 0.0419, + "step": 3170 + }, + { + "epoch": 1.9974874371859297, + "grad_norm": 0.3704904019832611, + "learning_rate": 9.914609141809139e-05, + "loss": 0.036, + "step": 3180 + }, + { + "epoch": 2.0037688442211055, + "grad_norm": 0.36662808060646057, + "learning_rate": 9.913591898020235e-05, + "loss": 0.0336, + "step": 3190 + }, + { + "epoch": 2.0100502512562812, + "grad_norm": 0.35551607608795166, + "learning_rate": 9.912568683753853e-05, + "loss": 0.0448, + "step": 3200 + }, + { + "epoch": 2.0163316582914574, + "grad_norm": 0.45012369751930237, + "learning_rate": 9.911539500253295e-05, + "loss": 0.0359, + "step": 3210 + }, + { + "epoch": 2.022613065326633, + "grad_norm": 0.569905698299408, + "learning_rate": 9.910504348769118e-05, + "loss": 0.0372, + "step": 3220 + }, + { + "epoch": 2.028894472361809, + "grad_norm": 0.26649320125579834, + "learning_rate": 9.909463230559127e-05, + "loss": 0.0357, + "step": 3230 + }, + { + "epoch": 2.0351758793969847, + "grad_norm": 0.4068954586982727, + "learning_rate": 9.908416146888376e-05, + "loss": 0.0314, + "step": 3240 + }, + { + "epoch": 2.041457286432161, + "grad_norm": 0.6464240550994873, + "learning_rate": 9.907363099029175e-05, + "loss": 0.0325, + "step": 3250 + }, + { + "epoch": 2.0477386934673367, + "grad_norm": 0.5090212821960449, + "learning_rate": 9.906304088261073e-05, + "loss": 0.0398, + "step": 3260 + }, + { + "epoch": 2.0540201005025125, + "grad_norm": 0.47795116901397705, + "learning_rate": 9.905239115870872e-05, + "loss": 0.0371, + "step": 3270 + }, + { + "epoch": 2.0603015075376883, + "grad_norm": 0.5819860100746155, + "learning_rate": 9.90416818315261e-05, + "loss": 0.0336, + "step": 3280 + }, + { + "epoch": 2.0665829145728645, + "grad_norm": 0.3318374752998352, + "learning_rate": 9.903091291407573e-05, + "loss": 0.0365, + "step": 3290 + }, + { + "epoch": 2.0728643216080402, + "grad_norm": 0.4634975790977478, + "learning_rate": 9.902008441944286e-05, + "loss": 0.0305, + "step": 3300 + }, + { + "epoch": 2.079145728643216, + "grad_norm": 0.5309159159660339, + "learning_rate": 9.900919636078512e-05, + "loss": 0.0381, + "step": 3310 + }, + { + "epoch": 2.0854271356783918, + "grad_norm": 0.3075733780860901, + "learning_rate": 9.899824875133255e-05, + "loss": 0.0297, + "step": 3320 + }, + { + "epoch": 2.091708542713568, + "grad_norm": 0.5231513381004333, + "learning_rate": 9.898724160438749e-05, + "loss": 0.0333, + "step": 3330 + }, + { + "epoch": 2.0979899497487438, + "grad_norm": 0.3576298654079437, + "learning_rate": 9.89761749333247e-05, + "loss": 0.0357, + "step": 3340 + }, + { + "epoch": 2.1042713567839195, + "grad_norm": 0.3550163507461548, + "learning_rate": 9.896504875159122e-05, + "loss": 0.0347, + "step": 3350 + }, + { + "epoch": 2.1105527638190953, + "grad_norm": 0.7416375279426575, + "learning_rate": 9.89538630727064e-05, + "loss": 0.0366, + "step": 3360 + }, + { + "epoch": 2.1168341708542715, + "grad_norm": 0.24025285243988037, + "learning_rate": 9.894261791026189e-05, + "loss": 0.0413, + "step": 3370 + }, + { + "epoch": 2.1231155778894473, + "grad_norm": 0.36895090341567993, + "learning_rate": 9.893131327792165e-05, + "loss": 0.0348, + "step": 3380 + }, + { + "epoch": 2.129396984924623, + "grad_norm": 0.44980618357658386, + "learning_rate": 9.891994918942182e-05, + "loss": 0.0384, + "step": 3390 + }, + { + "epoch": 2.135678391959799, + "grad_norm": 0.41477423906326294, + "learning_rate": 9.890852565857092e-05, + "loss": 0.0417, + "step": 3400 + }, + { + "epoch": 2.141959798994975, + "grad_norm": 0.31709566712379456, + "learning_rate": 9.889704269924954e-05, + "loss": 0.0331, + "step": 3410 + }, + { + "epoch": 2.148241206030151, + "grad_norm": 0.44892793893814087, + "learning_rate": 9.888550032541059e-05, + "loss": 0.0382, + "step": 3420 + }, + { + "epoch": 2.1545226130653266, + "grad_norm": 0.3267815113067627, + "learning_rate": 9.887389855107916e-05, + "loss": 0.0348, + "step": 3430 + }, + { + "epoch": 2.1608040201005023, + "grad_norm": 0.38743704557418823, + "learning_rate": 9.886223739035248e-05, + "loss": 0.0303, + "step": 3440 + }, + { + "epoch": 2.1670854271356785, + "grad_norm": 0.4421914517879486, + "learning_rate": 9.885051685739997e-05, + "loss": 0.0352, + "step": 3450 + }, + { + "epoch": 2.1733668341708543, + "grad_norm": 0.3349846601486206, + "learning_rate": 9.883873696646316e-05, + "loss": 0.0393, + "step": 3460 + }, + { + "epoch": 2.17964824120603, + "grad_norm": 0.39491546154022217, + "learning_rate": 9.882689773185575e-05, + "loss": 0.0355, + "step": 3470 + }, + { + "epoch": 2.185929648241206, + "grad_norm": 0.3810426592826843, + "learning_rate": 9.881499916796353e-05, + "loss": 0.0346, + "step": 3480 + }, + { + "epoch": 2.192211055276382, + "grad_norm": 0.31053629517555237, + "learning_rate": 9.880304128924434e-05, + "loss": 0.0343, + "step": 3490 + }, + { + "epoch": 2.198492462311558, + "grad_norm": 0.5048114657402039, + "learning_rate": 9.879102411022817e-05, + "loss": 0.0407, + "step": 3500 + }, + { + "epoch": 2.2047738693467336, + "grad_norm": 0.343868613243103, + "learning_rate": 9.877894764551703e-05, + "loss": 0.0364, + "step": 3510 + }, + { + "epoch": 2.2110552763819094, + "grad_norm": 0.38671329617500305, + "learning_rate": 9.876681190978494e-05, + "loss": 0.0368, + "step": 3520 + }, + { + "epoch": 2.2173366834170856, + "grad_norm": 0.45966246724128723, + "learning_rate": 9.875461691777797e-05, + "loss": 0.0305, + "step": 3530 + }, + { + "epoch": 2.2236180904522613, + "grad_norm": 0.6345669627189636, + "learning_rate": 9.874236268431417e-05, + "loss": 0.0326, + "step": 3540 + }, + { + "epoch": 2.229899497487437, + "grad_norm": 0.35427990555763245, + "learning_rate": 9.873004922428361e-05, + "loss": 0.0406, + "step": 3550 + }, + { + "epoch": 2.236180904522613, + "grad_norm": 0.42104846239089966, + "learning_rate": 9.871767655264829e-05, + "loss": 0.0364, + "step": 3560 + }, + { + "epoch": 2.242462311557789, + "grad_norm": 0.3619641959667206, + "learning_rate": 9.87052446844422e-05, + "loss": 0.0394, + "step": 3570 + }, + { + "epoch": 2.248743718592965, + "grad_norm": 0.47864192724227905, + "learning_rate": 9.869275363477122e-05, + "loss": 0.0346, + "step": 3580 + }, + { + "epoch": 2.2550251256281406, + "grad_norm": 0.42718935012817383, + "learning_rate": 9.868020341881312e-05, + "loss": 0.0378, + "step": 3590 + }, + { + "epoch": 2.2613065326633164, + "grad_norm": 0.2940632104873657, + "learning_rate": 9.866759405181765e-05, + "loss": 0.037, + "step": 3600 + }, + { + "epoch": 2.2675879396984926, + "grad_norm": 0.2504681646823883, + "learning_rate": 9.865492554910633e-05, + "loss": 0.0332, + "step": 3610 + }, + { + "epoch": 2.2738693467336684, + "grad_norm": 0.4826835095882416, + "learning_rate": 9.864219792607262e-05, + "loss": 0.0362, + "step": 3620 + }, + { + "epoch": 2.280150753768844, + "grad_norm": 0.28798359632492065, + "learning_rate": 9.862941119818177e-05, + "loss": 0.0321, + "step": 3630 + }, + { + "epoch": 2.2864321608040203, + "grad_norm": 0.6196483373641968, + "learning_rate": 9.861656538097086e-05, + "loss": 0.0373, + "step": 3640 + }, + { + "epoch": 2.292713567839196, + "grad_norm": 0.3851224184036255, + "learning_rate": 9.860366049004877e-05, + "loss": 0.0347, + "step": 3650 + }, + { + "epoch": 2.298994974874372, + "grad_norm": 0.4099680185317993, + "learning_rate": 9.859069654109615e-05, + "loss": 0.036, + "step": 3660 + }, + { + "epoch": 2.3052763819095476, + "grad_norm": 0.37255412340164185, + "learning_rate": 9.857767354986545e-05, + "loss": 0.036, + "step": 3670 + }, + { + "epoch": 2.3115577889447234, + "grad_norm": 0.5716639161109924, + "learning_rate": 9.856459153218078e-05, + "loss": 0.0356, + "step": 3680 + }, + { + "epoch": 2.3178391959798996, + "grad_norm": 0.6833359003067017, + "learning_rate": 9.855145050393808e-05, + "loss": 0.0356, + "step": 3690 + }, + { + "epoch": 2.3241206030150754, + "grad_norm": 0.668482780456543, + "learning_rate": 9.85382504811049e-05, + "loss": 0.0326, + "step": 3700 + }, + { + "epoch": 2.330402010050251, + "grad_norm": 0.3588888645172119, + "learning_rate": 9.852499147972054e-05, + "loss": 0.0401, + "step": 3710 + }, + { + "epoch": 2.3366834170854274, + "grad_norm": 0.34679996967315674, + "learning_rate": 9.851167351589592e-05, + "loss": 0.0343, + "step": 3720 + }, + { + "epoch": 2.342964824120603, + "grad_norm": 0.3339659869670868, + "learning_rate": 9.849829660581363e-05, + "loss": 0.0344, + "step": 3730 + }, + { + "epoch": 2.349246231155779, + "grad_norm": 0.40834543108940125, + "learning_rate": 9.848486076572787e-05, + "loss": 0.0344, + "step": 3740 + }, + { + "epoch": 2.3555276381909547, + "grad_norm": 0.22233933210372925, + "learning_rate": 9.847136601196446e-05, + "loss": 0.0317, + "step": 3750 + }, + { + "epoch": 2.3618090452261304, + "grad_norm": 0.28656330704689026, + "learning_rate": 9.845781236092078e-05, + "loss": 0.0328, + "step": 3760 + }, + { + "epoch": 2.3680904522613067, + "grad_norm": 0.4446433186531067, + "learning_rate": 9.844419982906583e-05, + "loss": 0.0342, + "step": 3770 + }, + { + "epoch": 2.3743718592964824, + "grad_norm": 0.4535566568374634, + "learning_rate": 9.843052843294008e-05, + "loss": 0.0369, + "step": 3780 + }, + { + "epoch": 2.380653266331658, + "grad_norm": 0.36285385489463806, + "learning_rate": 9.841679818915559e-05, + "loss": 0.0295, + "step": 3790 + }, + { + "epoch": 2.3869346733668344, + "grad_norm": 0.6404628157615662, + "learning_rate": 9.840300911439591e-05, + "loss": 0.0369, + "step": 3800 + }, + { + "epoch": 2.39321608040201, + "grad_norm": 0.35256797075271606, + "learning_rate": 9.838916122541603e-05, + "loss": 0.0293, + "step": 3810 + }, + { + "epoch": 2.399497487437186, + "grad_norm": 0.5223060250282288, + "learning_rate": 9.837525453904246e-05, + "loss": 0.0386, + "step": 3820 + }, + { + "epoch": 2.4057788944723617, + "grad_norm": 0.346017450094223, + "learning_rate": 9.836128907217314e-05, + "loss": 0.0391, + "step": 3830 + }, + { + "epoch": 2.4120603015075375, + "grad_norm": 0.2839776873588562, + "learning_rate": 9.834726484177743e-05, + "loss": 0.0301, + "step": 3840 + }, + { + "epoch": 2.4183417085427137, + "grad_norm": 0.4568172097206116, + "learning_rate": 9.833318186489609e-05, + "loss": 0.0327, + "step": 3850 + }, + { + "epoch": 2.4246231155778895, + "grad_norm": 0.42910876870155334, + "learning_rate": 9.831904015864126e-05, + "loss": 0.0349, + "step": 3860 + }, + { + "epoch": 2.430904522613065, + "grad_norm": 0.3084694445133209, + "learning_rate": 9.830483974019645e-05, + "loss": 0.0346, + "step": 3870 + }, + { + "epoch": 2.4371859296482414, + "grad_norm": 0.34692493081092834, + "learning_rate": 9.82905806268165e-05, + "loss": 0.0369, + "step": 3880 + }, + { + "epoch": 2.443467336683417, + "grad_norm": 0.44404682517051697, + "learning_rate": 9.82762628358276e-05, + "loss": 0.0332, + "step": 3890 + }, + { + "epoch": 2.449748743718593, + "grad_norm": 0.2709495425224304, + "learning_rate": 9.826188638462718e-05, + "loss": 0.0403, + "step": 3900 + }, + { + "epoch": 2.4560301507537687, + "grad_norm": 0.645540714263916, + "learning_rate": 9.824745129068402e-05, + "loss": 0.034, + "step": 3910 + }, + { + "epoch": 2.4623115577889445, + "grad_norm": 0.31420671939849854, + "learning_rate": 9.82329575715381e-05, + "loss": 0.0334, + "step": 3920 + }, + { + "epoch": 2.4685929648241207, + "grad_norm": 0.42626070976257324, + "learning_rate": 9.821840524480066e-05, + "loss": 0.0353, + "step": 3930 + }, + { + "epoch": 2.4748743718592965, + "grad_norm": 0.24827630817890167, + "learning_rate": 9.820379432815414e-05, + "loss": 0.0349, + "step": 3940 + }, + { + "epoch": 2.4811557788944723, + "grad_norm": 0.2967325747013092, + "learning_rate": 9.81891248393522e-05, + "loss": 0.0355, + "step": 3950 + }, + { + "epoch": 2.4874371859296485, + "grad_norm": 0.4242379367351532, + "learning_rate": 9.817439679621963e-05, + "loss": 0.0341, + "step": 3960 + }, + { + "epoch": 2.4937185929648242, + "grad_norm": 0.4138161242008209, + "learning_rate": 9.815961021665243e-05, + "loss": 0.0362, + "step": 3970 + }, + { + "epoch": 2.5, + "grad_norm": 0.46482565999031067, + "learning_rate": 9.814476511861763e-05, + "loss": 0.033, + "step": 3980 + }, + { + "epoch": 2.5062814070351758, + "grad_norm": 0.34451529383659363, + "learning_rate": 9.812986152015348e-05, + "loss": 0.0335, + "step": 3990 + }, + { + "epoch": 2.5125628140703515, + "grad_norm": 0.40582263469696045, + "learning_rate": 9.811489943936922e-05, + "loss": 0.0329, + "step": 4000 + }, + { + "epoch": 2.5188442211055277, + "grad_norm": 0.4663388133049011, + "learning_rate": 9.809987889444522e-05, + "loss": 0.0316, + "step": 4010 + }, + { + "epoch": 2.5251256281407035, + "grad_norm": 0.3309445381164551, + "learning_rate": 9.808479990363282e-05, + "loss": 0.0317, + "step": 4020 + }, + { + "epoch": 2.5314070351758793, + "grad_norm": 0.4379842281341553, + "learning_rate": 9.806966248525445e-05, + "loss": 0.0377, + "step": 4030 + }, + { + "epoch": 2.5376884422110555, + "grad_norm": 0.29708901047706604, + "learning_rate": 9.805446665770348e-05, + "loss": 0.0322, + "step": 4040 + }, + { + "epoch": 2.5439698492462313, + "grad_norm": 0.3605194091796875, + "learning_rate": 9.803921243944429e-05, + "loss": 0.0324, + "step": 4050 + }, + { + "epoch": 2.550251256281407, + "grad_norm": 0.47992756962776184, + "learning_rate": 9.802389984901218e-05, + "loss": 0.0391, + "step": 4060 + }, + { + "epoch": 2.556532663316583, + "grad_norm": 0.30806195735931396, + "learning_rate": 9.80085289050134e-05, + "loss": 0.0363, + "step": 4070 + }, + { + "epoch": 2.5628140703517586, + "grad_norm": 0.2861126959323883, + "learning_rate": 9.799309962612508e-05, + "loss": 0.0373, + "step": 4080 + }, + { + "epoch": 2.569095477386935, + "grad_norm": 0.3013077974319458, + "learning_rate": 9.797761203109527e-05, + "loss": 0.0304, + "step": 4090 + }, + { + "epoch": 2.5753768844221105, + "grad_norm": 0.43744799494743347, + "learning_rate": 9.796206613874283e-05, + "loss": 0.0386, + "step": 4100 + }, + { + "epoch": 2.5816582914572863, + "grad_norm": 0.6361374855041504, + "learning_rate": 9.794646196795754e-05, + "loss": 0.0362, + "step": 4110 + }, + { + "epoch": 2.5879396984924625, + "grad_norm": 0.32843080163002014, + "learning_rate": 9.793079953769987e-05, + "loss": 0.0339, + "step": 4120 + }, + { + "epoch": 2.5942211055276383, + "grad_norm": 0.3396553099155426, + "learning_rate": 9.79150788670012e-05, + "loss": 0.0331, + "step": 4130 + }, + { + "epoch": 2.600502512562814, + "grad_norm": 0.4812799394130707, + "learning_rate": 9.78992999749636e-05, + "loss": 0.0344, + "step": 4140 + }, + { + "epoch": 2.60678391959799, + "grad_norm": 0.35393401980400085, + "learning_rate": 9.788346288075994e-05, + "loss": 0.0355, + "step": 4150 + }, + { + "epoch": 2.6130653266331656, + "grad_norm": 0.3586488962173462, + "learning_rate": 9.786756760363373e-05, + "loss": 0.0263, + "step": 4160 + }, + { + "epoch": 2.619346733668342, + "grad_norm": 0.40881749987602234, + "learning_rate": 9.78516141628993e-05, + "loss": 0.0377, + "step": 4170 + }, + { + "epoch": 2.6256281407035176, + "grad_norm": 0.5561473965644836, + "learning_rate": 9.783560257794154e-05, + "loss": 0.0357, + "step": 4180 + }, + { + "epoch": 2.6319095477386933, + "grad_norm": 0.5046454668045044, + "learning_rate": 9.781953286821603e-05, + "loss": 0.0406, + "step": 4190 + }, + { + "epoch": 2.6381909547738696, + "grad_norm": 0.4121033251285553, + "learning_rate": 9.780340505324901e-05, + "loss": 0.0327, + "step": 4200 + }, + { + "epoch": 2.6444723618090453, + "grad_norm": 0.31073909997940063, + "learning_rate": 9.778721915263727e-05, + "loss": 0.0294, + "step": 4210 + }, + { + "epoch": 2.650753768844221, + "grad_norm": 0.39934393763542175, + "learning_rate": 9.777097518604824e-05, + "loss": 0.035, + "step": 4220 + }, + { + "epoch": 2.657035175879397, + "grad_norm": 0.3092843294143677, + "learning_rate": 9.775467317321984e-05, + "loss": 0.0316, + "step": 4230 + }, + { + "epoch": 2.6633165829145726, + "grad_norm": 0.23820720613002777, + "learning_rate": 9.773831313396055e-05, + "loss": 0.0328, + "step": 4240 + }, + { + "epoch": 2.669597989949749, + "grad_norm": 0.46945682168006897, + "learning_rate": 9.77218950881494e-05, + "loss": 0.0344, + "step": 4250 + }, + { + "epoch": 2.6758793969849246, + "grad_norm": 0.3173041045665741, + "learning_rate": 9.770541905573583e-05, + "loss": 0.039, + "step": 4260 + }, + { + "epoch": 2.6821608040201004, + "grad_norm": 0.4258037209510803, + "learning_rate": 9.768888505673976e-05, + "loss": 0.0332, + "step": 4270 + }, + { + "epoch": 2.6884422110552766, + "grad_norm": 0.36937326192855835, + "learning_rate": 9.767229311125162e-05, + "loss": 0.0319, + "step": 4280 + }, + { + "epoch": 2.6947236180904524, + "grad_norm": 0.3560713529586792, + "learning_rate": 9.765564323943211e-05, + "loss": 0.0341, + "step": 4290 + }, + { + "epoch": 2.701005025125628, + "grad_norm": 0.30429401993751526, + "learning_rate": 9.763893546151244e-05, + "loss": 0.0319, + "step": 4300 + }, + { + "epoch": 2.707286432160804, + "grad_norm": 0.4624983072280884, + "learning_rate": 9.762216979779412e-05, + "loss": 0.0292, + "step": 4310 + }, + { + "epoch": 2.7135678391959797, + "grad_norm": 0.45035749673843384, + "learning_rate": 9.760534626864902e-05, + "loss": 0.0296, + "step": 4320 + }, + { + "epoch": 2.719849246231156, + "grad_norm": 0.2929127812385559, + "learning_rate": 9.758846489451931e-05, + "loss": 0.0292, + "step": 4330 + }, + { + "epoch": 2.7261306532663316, + "grad_norm": 0.18873991072177887, + "learning_rate": 9.757152569591748e-05, + "loss": 0.0287, + "step": 4340 + }, + { + "epoch": 2.7324120603015074, + "grad_norm": 0.43994206190109253, + "learning_rate": 9.75545286934262e-05, + "loss": 0.0344, + "step": 4350 + }, + { + "epoch": 2.7386934673366836, + "grad_norm": 0.6771456599235535, + "learning_rate": 9.753747390769847e-05, + "loss": 0.0324, + "step": 4360 + }, + { + "epoch": 2.7449748743718594, + "grad_norm": 0.3539806306362152, + "learning_rate": 9.752036135945744e-05, + "loss": 0.0356, + "step": 4370 + }, + { + "epoch": 2.751256281407035, + "grad_norm": 0.426303893327713, + "learning_rate": 9.750319106949649e-05, + "loss": 0.0312, + "step": 4380 + }, + { + "epoch": 2.757537688442211, + "grad_norm": 0.4279865324497223, + "learning_rate": 9.748596305867913e-05, + "loss": 0.0338, + "step": 4390 + }, + { + "epoch": 2.7638190954773867, + "grad_norm": 0.22335122525691986, + "learning_rate": 9.746867734793903e-05, + "loss": 0.0375, + "step": 4400 + }, + { + "epoch": 2.770100502512563, + "grad_norm": 0.3346351981163025, + "learning_rate": 9.745133395827993e-05, + "loss": 0.0312, + "step": 4410 + }, + { + "epoch": 2.7763819095477387, + "grad_norm": 0.3312491178512573, + "learning_rate": 9.743393291077572e-05, + "loss": 0.0352, + "step": 4420 + }, + { + "epoch": 2.7826633165829144, + "grad_norm": 0.3554938733577728, + "learning_rate": 9.741647422657028e-05, + "loss": 0.0323, + "step": 4430 + }, + { + "epoch": 2.7889447236180906, + "grad_norm": 0.2258182018995285, + "learning_rate": 9.739895792687758e-05, + "loss": 0.0337, + "step": 4440 + }, + { + "epoch": 2.7952261306532664, + "grad_norm": 0.40879562497138977, + "learning_rate": 9.738138403298157e-05, + "loss": 0.0327, + "step": 4450 + }, + { + "epoch": 2.801507537688442, + "grad_norm": 0.35315436124801636, + "learning_rate": 9.736375256623619e-05, + "loss": 0.0295, + "step": 4460 + }, + { + "epoch": 2.807788944723618, + "grad_norm": 0.3434261679649353, + "learning_rate": 9.734606354806533e-05, + "loss": 0.0317, + "step": 4470 + }, + { + "epoch": 2.8140703517587937, + "grad_norm": 0.6161462664604187, + "learning_rate": 9.73283169999628e-05, + "loss": 0.0316, + "step": 4480 + }, + { + "epoch": 2.82035175879397, + "grad_norm": 0.37313491106033325, + "learning_rate": 9.731051294349238e-05, + "loss": 0.0337, + "step": 4490 + }, + { + "epoch": 2.8266331658291457, + "grad_norm": 0.5472199320793152, + "learning_rate": 9.729265140028762e-05, + "loss": 0.0341, + "step": 4500 + }, + { + "epoch": 2.8329145728643215, + "grad_norm": 0.241695836186409, + "learning_rate": 9.727473239205201e-05, + "loss": 0.0354, + "step": 4510 + }, + { + "epoch": 2.8391959798994977, + "grad_norm": 0.6691701412200928, + "learning_rate": 9.725675594055883e-05, + "loss": 0.0375, + "step": 4520 + }, + { + "epoch": 2.8454773869346734, + "grad_norm": 0.40983229875564575, + "learning_rate": 9.723872206765116e-05, + "loss": 0.0355, + "step": 4530 + }, + { + "epoch": 2.851758793969849, + "grad_norm": 0.32241731882095337, + "learning_rate": 9.722063079524185e-05, + "loss": 0.0315, + "step": 4540 + }, + { + "epoch": 2.858040201005025, + "grad_norm": 0.2689880430698395, + "learning_rate": 9.720248214531351e-05, + "loss": 0.032, + "step": 4550 + }, + { + "epoch": 2.8643216080402008, + "grad_norm": 0.4152458608150482, + "learning_rate": 9.718427613991848e-05, + "loss": 0.0266, + "step": 4560 + }, + { + "epoch": 2.870603015075377, + "grad_norm": 0.3515586256980896, + "learning_rate": 9.716601280117873e-05, + "loss": 0.0302, + "step": 4570 + }, + { + "epoch": 2.8768844221105527, + "grad_norm": 0.4979972541332245, + "learning_rate": 9.714769215128596e-05, + "loss": 0.0314, + "step": 4580 + }, + { + "epoch": 2.8831658291457285, + "grad_norm": 0.35863181948661804, + "learning_rate": 9.712931421250152e-05, + "loss": 0.0349, + "step": 4590 + }, + { + "epoch": 2.8894472361809047, + "grad_norm": 0.31198757886886597, + "learning_rate": 9.711087900715627e-05, + "loss": 0.0291, + "step": 4600 + }, + { + "epoch": 2.8957286432160805, + "grad_norm": 0.41221681237220764, + "learning_rate": 9.709238655765078e-05, + "loss": 0.0353, + "step": 4610 + }, + { + "epoch": 2.9020100502512562, + "grad_norm": 0.4487650990486145, + "learning_rate": 9.707383688645511e-05, + "loss": 0.0344, + "step": 4620 + }, + { + "epoch": 2.908291457286432, + "grad_norm": 0.40229862928390503, + "learning_rate": 9.705523001610883e-05, + "loss": 0.0303, + "step": 4630 + }, + { + "epoch": 2.914572864321608, + "grad_norm": 0.34695741534233093, + "learning_rate": 9.703656596922107e-05, + "loss": 0.033, + "step": 4640 + }, + { + "epoch": 2.920854271356784, + "grad_norm": 0.35474902391433716, + "learning_rate": 9.70178447684704e-05, + "loss": 0.0338, + "step": 4650 + }, + { + "epoch": 2.9271356783919598, + "grad_norm": 0.31470850110054016, + "learning_rate": 9.699906643660483e-05, + "loss": 0.0456, + "step": 4660 + }, + { + "epoch": 2.9334170854271355, + "grad_norm": 0.390476793050766, + "learning_rate": 9.698023099644185e-05, + "loss": 0.0316, + "step": 4670 + }, + { + "epoch": 2.9396984924623117, + "grad_norm": 0.4800759553909302, + "learning_rate": 9.696133847086823e-05, + "loss": 0.03, + "step": 4680 + }, + { + "epoch": 2.9459798994974875, + "grad_norm": 0.32381752133369446, + "learning_rate": 9.694238888284022e-05, + "loss": 0.0309, + "step": 4690 + }, + { + "epoch": 2.9522613065326633, + "grad_norm": 0.5208995938301086, + "learning_rate": 9.692338225538333e-05, + "loss": 0.0406, + "step": 4700 + }, + { + "epoch": 2.958542713567839, + "grad_norm": 0.35063716769218445, + "learning_rate": 9.690431861159241e-05, + "loss": 0.0322, + "step": 4710 + }, + { + "epoch": 2.964824120603015, + "grad_norm": 0.4030611217021942, + "learning_rate": 9.688519797463161e-05, + "loss": 0.0348, + "step": 4720 + }, + { + "epoch": 2.971105527638191, + "grad_norm": 0.5864812731742859, + "learning_rate": 9.686602036773426e-05, + "loss": 0.0333, + "step": 4730 + }, + { + "epoch": 2.977386934673367, + "grad_norm": 0.2874599099159241, + "learning_rate": 9.684678581420302e-05, + "loss": 0.0327, + "step": 4740 + }, + { + "epoch": 2.9836683417085426, + "grad_norm": 0.32226550579071045, + "learning_rate": 9.682749433740962e-05, + "loss": 0.035, + "step": 4750 + }, + { + "epoch": 2.9899497487437188, + "grad_norm": 0.24282562732696533, + "learning_rate": 9.680814596079507e-05, + "loss": 0.0322, + "step": 4760 + }, + { + "epoch": 2.9962311557788945, + "grad_norm": 0.4657776355743408, + "learning_rate": 9.678874070786945e-05, + "loss": 0.0348, + "step": 4770 + }, + { + "epoch": 3.0025125628140703, + "grad_norm": 0.4828051030635834, + "learning_rate": 9.676927860221199e-05, + "loss": 0.036, + "step": 4780 + }, + { + "epoch": 3.008793969849246, + "grad_norm": 0.3510207235813141, + "learning_rate": 9.674975966747097e-05, + "loss": 0.0284, + "step": 4790 + }, + { + "epoch": 3.0150753768844223, + "grad_norm": 0.3650939464569092, + "learning_rate": 9.673018392736374e-05, + "loss": 0.0347, + "step": 4800 + }, + { + "epoch": 3.021356783919598, + "grad_norm": 0.4062788188457489, + "learning_rate": 9.671055140567667e-05, + "loss": 0.0285, + "step": 4810 + }, + { + "epoch": 3.027638190954774, + "grad_norm": 0.30405086278915405, + "learning_rate": 9.669086212626511e-05, + "loss": 0.0287, + "step": 4820 + }, + { + "epoch": 3.0339195979899496, + "grad_norm": 0.30499446392059326, + "learning_rate": 9.667111611305341e-05, + "loss": 0.0292, + "step": 4830 + }, + { + "epoch": 3.040201005025126, + "grad_norm": 0.41859257221221924, + "learning_rate": 9.665131339003486e-05, + "loss": 0.0425, + "step": 4840 + }, + { + "epoch": 3.0464824120603016, + "grad_norm": 0.3500816524028778, + "learning_rate": 9.663145398127158e-05, + "loss": 0.0312, + "step": 4850 + }, + { + "epoch": 3.0527638190954773, + "grad_norm": 0.46941953897476196, + "learning_rate": 9.661153791089467e-05, + "loss": 0.0286, + "step": 4860 + }, + { + "epoch": 3.059045226130653, + "grad_norm": 0.3775452971458435, + "learning_rate": 9.659156520310402e-05, + "loss": 0.0306, + "step": 4870 + }, + { + "epoch": 3.0653266331658293, + "grad_norm": 0.38154470920562744, + "learning_rate": 9.657153588216835e-05, + "loss": 0.0346, + "step": 4880 + }, + { + "epoch": 3.071608040201005, + "grad_norm": 0.3604390621185303, + "learning_rate": 9.655144997242516e-05, + "loss": 0.0272, + "step": 4890 + }, + { + "epoch": 3.077889447236181, + "grad_norm": 0.2850189507007599, + "learning_rate": 9.653130749828075e-05, + "loss": 0.0312, + "step": 4900 + }, + { + "epoch": 3.0841708542713566, + "grad_norm": 0.426397442817688, + "learning_rate": 9.65111084842101e-05, + "loss": 0.0335, + "step": 4910 + }, + { + "epoch": 3.090452261306533, + "grad_norm": 0.33880549669265747, + "learning_rate": 9.649085295475695e-05, + "loss": 0.0272, + "step": 4920 + }, + { + "epoch": 3.0967336683417086, + "grad_norm": 0.3479103744029999, + "learning_rate": 9.647054093453365e-05, + "loss": 0.0317, + "step": 4930 + }, + { + "epoch": 3.1030150753768844, + "grad_norm": 0.42636290192604065, + "learning_rate": 9.645017244822123e-05, + "loss": 0.0271, + "step": 4940 + }, + { + "epoch": 3.10929648241206, + "grad_norm": 0.5864477753639221, + "learning_rate": 9.642974752056931e-05, + "loss": 0.0327, + "step": 4950 + }, + { + "epoch": 3.1155778894472363, + "grad_norm": 0.22826074063777924, + "learning_rate": 9.640926617639613e-05, + "loss": 0.0275, + "step": 4960 + }, + { + "epoch": 3.121859296482412, + "grad_norm": 0.39270836114883423, + "learning_rate": 9.638872844058843e-05, + "loss": 0.034, + "step": 4970 + }, + { + "epoch": 3.128140703517588, + "grad_norm": 0.3437449336051941, + "learning_rate": 9.63681343381015e-05, + "loss": 0.0385, + "step": 4980 + }, + { + "epoch": 3.1344221105527637, + "grad_norm": 0.3804841637611389, + "learning_rate": 9.634748389395914e-05, + "loss": 0.0301, + "step": 4990 + }, + { + "epoch": 3.14070351758794, + "grad_norm": 0.268151193857193, + "learning_rate": 9.632677713325353e-05, + "loss": 0.0305, + "step": 5000 + }, + { + "epoch": 3.1469849246231156, + "grad_norm": 0.4932478666305542, + "learning_rate": 9.63060140811454e-05, + "loss": 0.0311, + "step": 5010 + }, + { + "epoch": 3.1532663316582914, + "grad_norm": 0.4123835563659668, + "learning_rate": 9.628519476286379e-05, + "loss": 0.0333, + "step": 5020 + }, + { + "epoch": 3.159547738693467, + "grad_norm": 0.2757653295993805, + "learning_rate": 9.626431920370612e-05, + "loss": 0.0317, + "step": 5030 + }, + { + "epoch": 3.1658291457286434, + "grad_norm": 0.270212322473526, + "learning_rate": 9.624338742903819e-05, + "loss": 0.0293, + "step": 5040 + }, + { + "epoch": 3.172110552763819, + "grad_norm": 0.3482019305229187, + "learning_rate": 9.622239946429406e-05, + "loss": 0.0303, + "step": 5050 + }, + { + "epoch": 3.178391959798995, + "grad_norm": 0.41004347801208496, + "learning_rate": 9.620135533497609e-05, + "loss": 0.0291, + "step": 5060 + }, + { + "epoch": 3.1846733668341707, + "grad_norm": 0.3257509469985962, + "learning_rate": 9.61802550666549e-05, + "loss": 0.0243, + "step": 5070 + }, + { + "epoch": 3.190954773869347, + "grad_norm": 0.2406129688024521, + "learning_rate": 9.615909868496928e-05, + "loss": 0.028, + "step": 5080 + }, + { + "epoch": 3.1972361809045227, + "grad_norm": 0.2754453718662262, + "learning_rate": 9.613788621562622e-05, + "loss": 0.0281, + "step": 5090 + }, + { + "epoch": 3.2035175879396984, + "grad_norm": 0.43377286195755005, + "learning_rate": 9.611661768440093e-05, + "loss": 0.0362, + "step": 5100 + }, + { + "epoch": 3.209798994974874, + "grad_norm": 0.38642656803131104, + "learning_rate": 9.609529311713661e-05, + "loss": 0.0341, + "step": 5110 + }, + { + "epoch": 3.2160804020100504, + "grad_norm": 0.2840151786804199, + "learning_rate": 9.607391253974466e-05, + "loss": 0.0389, + "step": 5120 + }, + { + "epoch": 3.222361809045226, + "grad_norm": 0.3441442847251892, + "learning_rate": 9.605247597820448e-05, + "loss": 0.0308, + "step": 5130 + }, + { + "epoch": 3.228643216080402, + "grad_norm": 0.326242059469223, + "learning_rate": 9.603098345856354e-05, + "loss": 0.0292, + "step": 5140 + }, + { + "epoch": 3.2349246231155777, + "grad_norm": 0.31250330805778503, + "learning_rate": 9.600943500693725e-05, + "loss": 0.0267, + "step": 5150 + }, + { + "epoch": 3.241206030150754, + "grad_norm": 0.30340665578842163, + "learning_rate": 9.598783064950902e-05, + "loss": 0.0346, + "step": 5160 + }, + { + "epoch": 3.2474874371859297, + "grad_norm": 0.306669682264328, + "learning_rate": 9.596617041253018e-05, + "loss": 0.0223, + "step": 5170 + }, + { + "epoch": 3.2537688442211055, + "grad_norm": 0.34116870164871216, + "learning_rate": 9.594445432231996e-05, + "loss": 0.0324, + "step": 5180 + }, + { + "epoch": 3.2600502512562812, + "grad_norm": 0.3403201103210449, + "learning_rate": 9.592268240526547e-05, + "loss": 0.0336, + "step": 5190 + }, + { + "epoch": 3.2663316582914574, + "grad_norm": 0.4742276966571808, + "learning_rate": 9.590085468782162e-05, + "loss": 0.0382, + "step": 5200 + }, + { + "epoch": 3.272613065326633, + "grad_norm": 0.353008896112442, + "learning_rate": 9.587897119651116e-05, + "loss": 0.0335, + "step": 5210 + }, + { + "epoch": 3.278894472361809, + "grad_norm": 0.5106473565101624, + "learning_rate": 9.585703195792459e-05, + "loss": 0.0339, + "step": 5220 + }, + { + "epoch": 3.2851758793969847, + "grad_norm": 0.3447539210319519, + "learning_rate": 9.583503699872016e-05, + "loss": 0.0362, + "step": 5230 + }, + { + "epoch": 3.291457286432161, + "grad_norm": 0.434509813785553, + "learning_rate": 9.581298634562381e-05, + "loss": 0.0297, + "step": 5240 + }, + { + "epoch": 3.2977386934673367, + "grad_norm": 0.6209575533866882, + "learning_rate": 9.579088002542917e-05, + "loss": 0.0298, + "step": 5250 + }, + { + "epoch": 3.3040201005025125, + "grad_norm": 0.4804987609386444, + "learning_rate": 9.57687180649975e-05, + "loss": 0.0353, + "step": 5260 + }, + { + "epoch": 3.3103015075376883, + "grad_norm": 0.42006680369377136, + "learning_rate": 9.574650049125768e-05, + "loss": 0.0384, + "step": 5270 + }, + { + "epoch": 3.3165829145728645, + "grad_norm": 0.4436415433883667, + "learning_rate": 9.572422733120614e-05, + "loss": 0.0348, + "step": 5280 + }, + { + "epoch": 3.3228643216080402, + "grad_norm": 0.37437883019447327, + "learning_rate": 9.570189861190689e-05, + "loss": 0.0332, + "step": 5290 + }, + { + "epoch": 3.329145728643216, + "grad_norm": 0.47929689288139343, + "learning_rate": 9.56795143604914e-05, + "loss": 0.0299, + "step": 5300 + }, + { + "epoch": 3.3354271356783918, + "grad_norm": 0.2935585081577301, + "learning_rate": 9.565707460415869e-05, + "loss": 0.0338, + "step": 5310 + }, + { + "epoch": 3.341708542713568, + "grad_norm": 0.33489277958869934, + "learning_rate": 9.563457937017515e-05, + "loss": 0.0345, + "step": 5320 + }, + { + "epoch": 3.3479899497487438, + "grad_norm": 0.21881592273712158, + "learning_rate": 9.56120286858746e-05, + "loss": 0.0286, + "step": 5330 + }, + { + "epoch": 3.3542713567839195, + "grad_norm": 0.42485570907592773, + "learning_rate": 9.558942257865829e-05, + "loss": 0.0305, + "step": 5340 + }, + { + "epoch": 3.3605527638190953, + "grad_norm": 0.3356097936630249, + "learning_rate": 9.556676107599472e-05, + "loss": 0.032, + "step": 5350 + }, + { + "epoch": 3.3668341708542715, + "grad_norm": 0.376590371131897, + "learning_rate": 9.554404420541978e-05, + "loss": 0.0357, + "step": 5360 + }, + { + "epoch": 3.3731155778894473, + "grad_norm": 0.2474142462015152, + "learning_rate": 9.55212719945366e-05, + "loss": 0.0311, + "step": 5370 + }, + { + "epoch": 3.379396984924623, + "grad_norm": 0.32987862825393677, + "learning_rate": 9.549844447101559e-05, + "loss": 0.0299, + "step": 5380 + }, + { + "epoch": 3.385678391959799, + "grad_norm": 0.31293362379074097, + "learning_rate": 9.547556166259433e-05, + "loss": 0.0283, + "step": 5390 + }, + { + "epoch": 3.391959798994975, + "grad_norm": 0.4200120270252228, + "learning_rate": 9.545262359707756e-05, + "loss": 0.0323, + "step": 5400 + }, + { + "epoch": 3.398241206030151, + "grad_norm": 0.3345884382724762, + "learning_rate": 9.542963030233724e-05, + "loss": 0.032, + "step": 5410 + }, + { + "epoch": 3.4045226130653266, + "grad_norm": 0.37696489691734314, + "learning_rate": 9.540658180631237e-05, + "loss": 0.0262, + "step": 5420 + }, + { + "epoch": 3.4108040201005023, + "grad_norm": 0.39309564232826233, + "learning_rate": 9.538347813700904e-05, + "loss": 0.028, + "step": 5430 + }, + { + "epoch": 3.4170854271356785, + "grad_norm": 0.3277542293071747, + "learning_rate": 9.536031932250036e-05, + "loss": 0.0292, + "step": 5440 + }, + { + "epoch": 3.4233668341708543, + "grad_norm": 0.3717183768749237, + "learning_rate": 9.533710539092653e-05, + "loss": 0.0353, + "step": 5450 + }, + { + "epoch": 3.42964824120603, + "grad_norm": 0.306644082069397, + "learning_rate": 9.531383637049464e-05, + "loss": 0.0323, + "step": 5460 + }, + { + "epoch": 3.435929648241206, + "grad_norm": 0.47923892736434937, + "learning_rate": 9.529051228947875e-05, + "loss": 0.0342, + "step": 5470 + }, + { + "epoch": 3.442211055276382, + "grad_norm": 0.36502864956855774, + "learning_rate": 9.52671331762198e-05, + "loss": 0.0276, + "step": 5480 + }, + { + "epoch": 3.448492462311558, + "grad_norm": 0.2177933007478714, + "learning_rate": 9.524369905912565e-05, + "loss": 0.0276, + "step": 5490 + }, + { + "epoch": 3.4547738693467336, + "grad_norm": 0.3741821050643921, + "learning_rate": 9.522020996667092e-05, + "loss": 0.0292, + "step": 5500 + }, + { + "epoch": 3.4610552763819094, + "grad_norm": 0.40517130494117737, + "learning_rate": 9.519666592739709e-05, + "loss": 0.0302, + "step": 5510 + }, + { + "epoch": 3.4673366834170856, + "grad_norm": 0.4100216031074524, + "learning_rate": 9.517306696991241e-05, + "loss": 0.0265, + "step": 5520 + }, + { + "epoch": 3.4736180904522613, + "grad_norm": 0.3412381410598755, + "learning_rate": 9.51494131228918e-05, + "loss": 0.0291, + "step": 5530 + }, + { + "epoch": 3.479899497487437, + "grad_norm": 0.3763032853603363, + "learning_rate": 9.512570441507695e-05, + "loss": 0.0409, + "step": 5540 + }, + { + "epoch": 3.4861809045226133, + "grad_norm": 0.36994245648384094, + "learning_rate": 9.510194087527615e-05, + "loss": 0.028, + "step": 5550 + }, + { + "epoch": 3.492462311557789, + "grad_norm": 0.34395429491996765, + "learning_rate": 9.507812253236435e-05, + "loss": 0.0253, + "step": 5560 + }, + { + "epoch": 3.498743718592965, + "grad_norm": 0.2330450415611267, + "learning_rate": 9.505424941528309e-05, + "loss": 0.0269, + "step": 5570 + }, + { + "epoch": 3.5050251256281406, + "grad_norm": 0.23572692275047302, + "learning_rate": 9.503032155304046e-05, + "loss": 0.0291, + "step": 5580 + }, + { + "epoch": 3.5113065326633164, + "grad_norm": 0.3665548861026764, + "learning_rate": 9.500633897471106e-05, + "loss": 0.0317, + "step": 5590 + }, + { + "epoch": 3.5175879396984926, + "grad_norm": 0.3363001048564911, + "learning_rate": 9.498230170943596e-05, + "loss": 0.0292, + "step": 5600 + }, + { + "epoch": 3.5238693467336684, + "grad_norm": 0.36650410294532776, + "learning_rate": 9.495820978642275e-05, + "loss": 0.0336, + "step": 5610 + }, + { + "epoch": 3.530150753768844, + "grad_norm": 0.49274924397468567, + "learning_rate": 9.493406323494535e-05, + "loss": 0.0323, + "step": 5620 + }, + { + "epoch": 3.5364321608040203, + "grad_norm": 0.3658444583415985, + "learning_rate": 9.490986208434413e-05, + "loss": 0.0309, + "step": 5630 + }, + { + "epoch": 3.542713567839196, + "grad_norm": 0.42133796215057373, + "learning_rate": 9.488560636402577e-05, + "loss": 0.0348, + "step": 5640 + }, + { + "epoch": 3.548994974874372, + "grad_norm": 0.47544535994529724, + "learning_rate": 9.486129610346321e-05, + "loss": 0.0317, + "step": 5650 + }, + { + "epoch": 3.5552763819095476, + "grad_norm": 0.3424922227859497, + "learning_rate": 9.483693133219576e-05, + "loss": 0.0284, + "step": 5660 + }, + { + "epoch": 3.5615577889447234, + "grad_norm": 0.44671934843063354, + "learning_rate": 9.481251207982888e-05, + "loss": 0.0363, + "step": 5670 + }, + { + "epoch": 3.5678391959798996, + "grad_norm": 0.294164776802063, + "learning_rate": 9.47880383760343e-05, + "loss": 0.0311, + "step": 5680 + }, + { + "epoch": 3.5741206030150754, + "grad_norm": 0.2692391872406006, + "learning_rate": 9.476351025054983e-05, + "loss": 0.0297, + "step": 5690 + }, + { + "epoch": 3.580402010050251, + "grad_norm": 0.39461398124694824, + "learning_rate": 9.473892773317952e-05, + "loss": 0.0285, + "step": 5700 + }, + { + "epoch": 3.5866834170854274, + "grad_norm": 0.39258649945259094, + "learning_rate": 9.471429085379338e-05, + "loss": 0.0285, + "step": 5710 + }, + { + "epoch": 3.592964824120603, + "grad_norm": 0.5298832654953003, + "learning_rate": 9.468959964232757e-05, + "loss": 0.0355, + "step": 5720 + }, + { + "epoch": 3.599246231155779, + "grad_norm": 0.3074873685836792, + "learning_rate": 9.466485412878425e-05, + "loss": 0.0365, + "step": 5730 + }, + { + "epoch": 3.6055276381909547, + "grad_norm": 0.2667095363140106, + "learning_rate": 9.464005434323154e-05, + "loss": 0.0297, + "step": 5740 + }, + { + "epoch": 3.6118090452261304, + "grad_norm": 0.33825770020484924, + "learning_rate": 9.461520031580352e-05, + "loss": 0.0306, + "step": 5750 + }, + { + "epoch": 3.6180904522613067, + "grad_norm": 0.31162863969802856, + "learning_rate": 9.459029207670019e-05, + "loss": 0.0276, + "step": 5760 + }, + { + "epoch": 3.6243718592964824, + "grad_norm": 0.37730512022972107, + "learning_rate": 9.456532965618737e-05, + "loss": 0.03, + "step": 5770 + }, + { + "epoch": 3.630653266331658, + "grad_norm": 0.4943307340145111, + "learning_rate": 9.454031308459681e-05, + "loss": 0.0322, + "step": 5780 + }, + { + "epoch": 3.6369346733668344, + "grad_norm": 0.28119415044784546, + "learning_rate": 9.451524239232595e-05, + "loss": 0.0333, + "step": 5790 + }, + { + "epoch": 3.64321608040201, + "grad_norm": 0.45504650473594666, + "learning_rate": 9.449011760983809e-05, + "loss": 0.0274, + "step": 5800 + }, + { + "epoch": 3.649497487437186, + "grad_norm": 0.5633286237716675, + "learning_rate": 9.446493876766218e-05, + "loss": 0.0321, + "step": 5810 + }, + { + "epoch": 3.6557788944723617, + "grad_norm": 0.296716570854187, + "learning_rate": 9.44397058963929e-05, + "loss": 0.0273, + "step": 5820 + }, + { + "epoch": 3.6620603015075375, + "grad_norm": 0.31040364503860474, + "learning_rate": 9.441441902669056e-05, + "loss": 0.0357, + "step": 5830 + }, + { + "epoch": 3.6683417085427137, + "grad_norm": 0.4219314157962799, + "learning_rate": 9.43890781892811e-05, + "loss": 0.0319, + "step": 5840 + }, + { + "epoch": 3.6746231155778895, + "grad_norm": 0.32672610878944397, + "learning_rate": 9.436368341495603e-05, + "loss": 0.0379, + "step": 5850 + }, + { + "epoch": 3.680904522613065, + "grad_norm": 0.4703666567802429, + "learning_rate": 9.43382347345724e-05, + "loss": 0.0357, + "step": 5860 + }, + { + "epoch": 3.6871859296482414, + "grad_norm": 0.3000369668006897, + "learning_rate": 9.431273217905273e-05, + "loss": 0.0296, + "step": 5870 + }, + { + "epoch": 3.693467336683417, + "grad_norm": 0.5552109479904175, + "learning_rate": 9.428717577938504e-05, + "loss": 0.0322, + "step": 5880 + }, + { + "epoch": 3.699748743718593, + "grad_norm": 0.33412104845046997, + "learning_rate": 9.426156556662276e-05, + "loss": 0.0379, + "step": 5890 + }, + { + "epoch": 3.7060301507537687, + "grad_norm": 0.40116897225379944, + "learning_rate": 9.423590157188474e-05, + "loss": 0.0319, + "step": 5900 + }, + { + "epoch": 3.7123115577889445, + "grad_norm": 0.44303998351097107, + "learning_rate": 9.421018382635513e-05, + "loss": 0.0306, + "step": 5910 + }, + { + "epoch": 3.7185929648241207, + "grad_norm": 0.4027128517627716, + "learning_rate": 9.418441236128343e-05, + "loss": 0.0338, + "step": 5920 + }, + { + "epoch": 3.7248743718592965, + "grad_norm": 0.48801830410957336, + "learning_rate": 9.41585872079844e-05, + "loss": 0.0337, + "step": 5930 + }, + { + "epoch": 3.7311557788944723, + "grad_norm": 0.4391435980796814, + "learning_rate": 9.413270839783802e-05, + "loss": 0.0286, + "step": 5940 + }, + { + "epoch": 3.7374371859296485, + "grad_norm": 0.4125594198703766, + "learning_rate": 9.41067759622895e-05, + "loss": 0.033, + "step": 5950 + }, + { + "epoch": 3.7437185929648242, + "grad_norm": 0.5943503379821777, + "learning_rate": 9.408078993284917e-05, + "loss": 0.0298, + "step": 5960 + }, + { + "epoch": 3.75, + "grad_norm": 0.2922727167606354, + "learning_rate": 9.405475034109254e-05, + "loss": 0.0294, + "step": 5970 + }, + { + "epoch": 3.7562814070351758, + "grad_norm": 0.3908199667930603, + "learning_rate": 9.402865721866015e-05, + "loss": 0.0286, + "step": 5980 + }, + { + "epoch": 3.7625628140703515, + "grad_norm": 0.4014753997325897, + "learning_rate": 9.400251059725762e-05, + "loss": 0.029, + "step": 5990 + }, + { + "epoch": 3.7688442211055277, + "grad_norm": 0.33277627825737, + "learning_rate": 9.397631050865554e-05, + "loss": 0.031, + "step": 6000 + }, + { + "epoch": 3.7751256281407035, + "grad_norm": 0.3413005769252777, + "learning_rate": 9.395005698468949e-05, + "loss": 0.0303, + "step": 6010 + }, + { + "epoch": 3.7814070351758793, + "grad_norm": 0.45078206062316895, + "learning_rate": 9.392375005726e-05, + "loss": 0.0276, + "step": 6020 + }, + { + "epoch": 3.7876884422110555, + "grad_norm": 0.3183113932609558, + "learning_rate": 9.389738975833242e-05, + "loss": 0.0293, + "step": 6030 + }, + { + "epoch": 3.7939698492462313, + "grad_norm": 0.3508162796497345, + "learning_rate": 9.387097611993707e-05, + "loss": 0.0273, + "step": 6040 + }, + { + "epoch": 3.800251256281407, + "grad_norm": 0.3887679874897003, + "learning_rate": 9.384450917416894e-05, + "loss": 0.0301, + "step": 6050 + }, + { + "epoch": 3.806532663316583, + "grad_norm": 0.2992834150791168, + "learning_rate": 9.381798895318792e-05, + "loss": 0.0321, + "step": 6060 + }, + { + "epoch": 3.8128140703517586, + "grad_norm": 0.48060840368270874, + "learning_rate": 9.379141548921853e-05, + "loss": 0.0282, + "step": 6070 + }, + { + "epoch": 3.819095477386935, + "grad_norm": 0.49912846088409424, + "learning_rate": 9.376478881455009e-05, + "loss": 0.0314, + "step": 6080 + }, + { + "epoch": 3.8253768844221105, + "grad_norm": 0.26678764820098877, + "learning_rate": 9.373810896153647e-05, + "loss": 0.0285, + "step": 6090 + }, + { + "epoch": 3.8316582914572863, + "grad_norm": 0.3090830445289612, + "learning_rate": 9.371137596259623e-05, + "loss": 0.0258, + "step": 6100 + }, + { + "epoch": 3.8379396984924625, + "grad_norm": 0.3382675349712372, + "learning_rate": 9.368458985021248e-05, + "loss": 0.027, + "step": 6110 + }, + { + "epoch": 3.8442211055276383, + "grad_norm": 0.4056253731250763, + "learning_rate": 9.365775065693287e-05, + "loss": 0.027, + "step": 6120 + }, + { + "epoch": 3.850502512562814, + "grad_norm": 0.39629247784614563, + "learning_rate": 9.363085841536957e-05, + "loss": 0.0267, + "step": 6130 + }, + { + "epoch": 3.85678391959799, + "grad_norm": 0.2716200053691864, + "learning_rate": 9.360391315819917e-05, + "loss": 0.0296, + "step": 6140 + }, + { + "epoch": 3.8630653266331656, + "grad_norm": 0.3804766535758972, + "learning_rate": 9.357691491816269e-05, + "loss": 0.0247, + "step": 6150 + }, + { + "epoch": 3.869346733668342, + "grad_norm": 0.3813430368900299, + "learning_rate": 9.354986372806557e-05, + "loss": 0.031, + "step": 6160 + }, + { + "epoch": 3.8756281407035176, + "grad_norm": 0.3719359338283539, + "learning_rate": 9.352275962077752e-05, + "loss": 0.0294, + "step": 6170 + }, + { + "epoch": 3.8819095477386933, + "grad_norm": 0.3748750388622284, + "learning_rate": 9.349560262923262e-05, + "loss": 0.0315, + "step": 6180 + }, + { + "epoch": 3.8881909547738696, + "grad_norm": 0.3115510642528534, + "learning_rate": 9.346839278642913e-05, + "loss": 0.034, + "step": 6190 + }, + { + "epoch": 3.8944723618090453, + "grad_norm": 0.3216722309589386, + "learning_rate": 9.344113012542963e-05, + "loss": 0.0349, + "step": 6200 + }, + { + "epoch": 3.900753768844221, + "grad_norm": 0.4627048075199127, + "learning_rate": 9.341381467936079e-05, + "loss": 0.0278, + "step": 6210 + }, + { + "epoch": 3.907035175879397, + "grad_norm": 0.24978110194206238, + "learning_rate": 9.338644648141346e-05, + "loss": 0.03, + "step": 6220 + }, + { + "epoch": 3.9133165829145726, + "grad_norm": 0.4483942985534668, + "learning_rate": 9.335902556484257e-05, + "loss": 0.0284, + "step": 6230 + }, + { + "epoch": 3.919597989949749, + "grad_norm": 0.45611536502838135, + "learning_rate": 9.333155196296712e-05, + "loss": 0.0248, + "step": 6240 + }, + { + "epoch": 3.9258793969849246, + "grad_norm": 0.38677483797073364, + "learning_rate": 9.330402570917016e-05, + "loss": 0.0319, + "step": 6250 + }, + { + "epoch": 3.9321608040201004, + "grad_norm": 0.3776909410953522, + "learning_rate": 9.327644683689865e-05, + "loss": 0.0261, + "step": 6260 + }, + { + "epoch": 3.9384422110552766, + "grad_norm": 0.24702589213848114, + "learning_rate": 9.324881537966354e-05, + "loss": 0.0315, + "step": 6270 + }, + { + "epoch": 3.9447236180904524, + "grad_norm": 0.39510923624038696, + "learning_rate": 9.322113137103964e-05, + "loss": 0.0306, + "step": 6280 + }, + { + "epoch": 3.951005025125628, + "grad_norm": 0.30005618929862976, + "learning_rate": 9.319339484466564e-05, + "loss": 0.0334, + "step": 6290 + }, + { + "epoch": 3.957286432160804, + "grad_norm": 0.5173377394676208, + "learning_rate": 9.316560583424403e-05, + "loss": 0.0299, + "step": 6300 + }, + { + "epoch": 3.9635678391959797, + "grad_norm": 0.3946971297264099, + "learning_rate": 9.313776437354108e-05, + "loss": 0.0277, + "step": 6310 + }, + { + "epoch": 3.969849246231156, + "grad_norm": 0.38305577635765076, + "learning_rate": 9.31098704963868e-05, + "loss": 0.0267, + "step": 6320 + }, + { + "epoch": 3.9761306532663316, + "grad_norm": 0.3092247545719147, + "learning_rate": 9.308192423667485e-05, + "loss": 0.0289, + "step": 6330 + }, + { + "epoch": 3.9824120603015074, + "grad_norm": 0.37768909335136414, + "learning_rate": 9.305392562836262e-05, + "loss": 0.0276, + "step": 6340 + }, + { + "epoch": 3.9886934673366836, + "grad_norm": 0.43444138765335083, + "learning_rate": 9.3025874705471e-05, + "loss": 0.0258, + "step": 6350 + }, + { + "epoch": 3.9949748743718594, + "grad_norm": 0.3566616177558899, + "learning_rate": 9.299777150208457e-05, + "loss": 0.0278, + "step": 6360 + }, + { + "epoch": 4.001256281407035, + "grad_norm": 0.4797225296497345, + "learning_rate": 9.296961605235133e-05, + "loss": 0.0347, + "step": 6370 + }, + { + "epoch": 4.007537688442211, + "grad_norm": 0.4299781620502472, + "learning_rate": 9.29414083904828e-05, + "loss": 0.0302, + "step": 6380 + }, + { + "epoch": 4.013819095477387, + "grad_norm": 0.25682976841926575, + "learning_rate": 9.291314855075397e-05, + "loss": 0.0306, + "step": 6390 + }, + { + "epoch": 4.0201005025125625, + "grad_norm": 0.3892262279987335, + "learning_rate": 9.288483656750322e-05, + "loss": 0.0316, + "step": 6400 + }, + { + "epoch": 4.026381909547739, + "grad_norm": 0.5177354216575623, + "learning_rate": 9.285647247513225e-05, + "loss": 0.0335, + "step": 6410 + }, + { + "epoch": 4.032663316582915, + "grad_norm": 0.42021608352661133, + "learning_rate": 9.282805630810614e-05, + "loss": 0.0313, + "step": 6420 + }, + { + "epoch": 4.038944723618091, + "grad_norm": 0.5041647553443909, + "learning_rate": 9.279958810095317e-05, + "loss": 0.0275, + "step": 6430 + }, + { + "epoch": 4.045226130653266, + "grad_norm": 0.5505565404891968, + "learning_rate": 9.277106788826494e-05, + "loss": 0.0292, + "step": 6440 + }, + { + "epoch": 4.051507537688442, + "grad_norm": 0.3862675726413727, + "learning_rate": 9.274249570469618e-05, + "loss": 0.0289, + "step": 6450 + }, + { + "epoch": 4.057788944723618, + "grad_norm": 0.36249589920043945, + "learning_rate": 9.271387158496476e-05, + "loss": 0.0279, + "step": 6460 + }, + { + "epoch": 4.064070351758794, + "grad_norm": 0.470560759305954, + "learning_rate": 9.268519556385173e-05, + "loss": 0.0284, + "step": 6470 + }, + { + "epoch": 4.0703517587939695, + "grad_norm": 0.3406115174293518, + "learning_rate": 9.265646767620113e-05, + "loss": 0.0319, + "step": 6480 + }, + { + "epoch": 4.076633165829146, + "grad_norm": 0.38485053181648254, + "learning_rate": 9.262768795692006e-05, + "loss": 0.0275, + "step": 6490 + }, + { + "epoch": 4.082914572864322, + "grad_norm": 0.519453763961792, + "learning_rate": 9.259885644097861e-05, + "loss": 0.0274, + "step": 6500 + }, + { + "epoch": 4.089195979899498, + "grad_norm": 0.2815948724746704, + "learning_rate": 9.256997316340976e-05, + "loss": 0.0259, + "step": 6510 + }, + { + "epoch": 4.0954773869346734, + "grad_norm": 0.30149516463279724, + "learning_rate": 9.254103815930943e-05, + "loss": 0.0288, + "step": 6520 + }, + { + "epoch": 4.101758793969849, + "grad_norm": 0.30567777156829834, + "learning_rate": 9.251205146383636e-05, + "loss": 0.0299, + "step": 6530 + }, + { + "epoch": 4.108040201005025, + "grad_norm": 0.35401734709739685, + "learning_rate": 9.248301311221216e-05, + "loss": 0.0293, + "step": 6540 + }, + { + "epoch": 4.114321608040201, + "grad_norm": 0.5033503770828247, + "learning_rate": 9.245392313972115e-05, + "loss": 0.0258, + "step": 6550 + }, + { + "epoch": 4.1206030150753765, + "grad_norm": 0.31380781531333923, + "learning_rate": 9.242478158171037e-05, + "loss": 0.0304, + "step": 6560 + }, + { + "epoch": 4.126884422110553, + "grad_norm": 0.29039663076400757, + "learning_rate": 9.239558847358957e-05, + "loss": 0.0271, + "step": 6570 + }, + { + "epoch": 4.133165829145729, + "grad_norm": 0.42805659770965576, + "learning_rate": 9.236634385083114e-05, + "loss": 0.0332, + "step": 6580 + }, + { + "epoch": 4.139447236180905, + "grad_norm": 0.35286349058151245, + "learning_rate": 9.233704774897005e-05, + "loss": 0.0267, + "step": 6590 + }, + { + "epoch": 4.1457286432160805, + "grad_norm": 0.2522640526294708, + "learning_rate": 9.230770020360382e-05, + "loss": 0.0263, + "step": 6600 + }, + { + "epoch": 4.152010050251256, + "grad_norm": 0.3191537857055664, + "learning_rate": 9.227830125039247e-05, + "loss": 0.0284, + "step": 6610 + }, + { + "epoch": 4.158291457286432, + "grad_norm": 0.3477436304092407, + "learning_rate": 9.224885092505853e-05, + "loss": 0.0284, + "step": 6620 + }, + { + "epoch": 4.164572864321608, + "grad_norm": 0.2587297856807709, + "learning_rate": 9.221934926338691e-05, + "loss": 0.0284, + "step": 6630 + }, + { + "epoch": 4.1708542713567835, + "grad_norm": 0.34130948781967163, + "learning_rate": 9.21897963012249e-05, + "loss": 0.0273, + "step": 6640 + }, + { + "epoch": 4.17713567839196, + "grad_norm": 0.5816400051116943, + "learning_rate": 9.216019207448217e-05, + "loss": 0.0255, + "step": 6650 + }, + { + "epoch": 4.183417085427136, + "grad_norm": 0.26632973551750183, + "learning_rate": 9.21305366191306e-05, + "loss": 0.0294, + "step": 6660 + }, + { + "epoch": 4.189698492462312, + "grad_norm": 0.2191411405801773, + "learning_rate": 9.21008299712044e-05, + "loss": 0.0279, + "step": 6670 + }, + { + "epoch": 4.1959798994974875, + "grad_norm": 0.26030150055885315, + "learning_rate": 9.207107216679993e-05, + "loss": 0.0292, + "step": 6680 + }, + { + "epoch": 4.202261306532663, + "grad_norm": 0.1794740855693817, + "learning_rate": 9.204126324207574e-05, + "loss": 0.0205, + "step": 6690 + }, + { + "epoch": 4.208542713567839, + "grad_norm": 0.2536580264568329, + "learning_rate": 9.201140323325248e-05, + "loss": 0.0292, + "step": 6700 + }, + { + "epoch": 4.214824120603015, + "grad_norm": 0.27463024854660034, + "learning_rate": 9.198149217661287e-05, + "loss": 0.024, + "step": 6710 + }, + { + "epoch": 4.221105527638191, + "grad_norm": 0.40730518102645874, + "learning_rate": 9.195153010850167e-05, + "loss": 0.029, + "step": 6720 + }, + { + "epoch": 4.227386934673367, + "grad_norm": 0.3181063234806061, + "learning_rate": 9.192151706532562e-05, + "loss": 0.0269, + "step": 6730 + }, + { + "epoch": 4.233668341708543, + "grad_norm": 0.40531036257743835, + "learning_rate": 9.189145308355337e-05, + "loss": 0.0359, + "step": 6740 + }, + { + "epoch": 4.239949748743719, + "grad_norm": 0.2365943044424057, + "learning_rate": 9.186133819971555e-05, + "loss": 0.0249, + "step": 6750 + }, + { + "epoch": 4.2462311557788945, + "grad_norm": 0.28128600120544434, + "learning_rate": 9.183117245040455e-05, + "loss": 0.0242, + "step": 6760 + }, + { + "epoch": 4.25251256281407, + "grad_norm": 0.38492926955223083, + "learning_rate": 9.180095587227458e-05, + "loss": 0.0288, + "step": 6770 + }, + { + "epoch": 4.258793969849246, + "grad_norm": 0.5706233382225037, + "learning_rate": 9.177068850204167e-05, + "loss": 0.0267, + "step": 6780 + }, + { + "epoch": 4.265075376884422, + "grad_norm": 0.39232656359672546, + "learning_rate": 9.174037037648351e-05, + "loss": 0.0286, + "step": 6790 + }, + { + "epoch": 4.271356783919598, + "grad_norm": 0.3997446894645691, + "learning_rate": 9.171000153243947e-05, + "loss": 0.0299, + "step": 6800 + }, + { + "epoch": 4.277638190954774, + "grad_norm": 0.39612311124801636, + "learning_rate": 9.167958200681057e-05, + "loss": 0.0287, + "step": 6810 + }, + { + "epoch": 4.28391959798995, + "grad_norm": 0.35515686869621277, + "learning_rate": 9.164911183655943e-05, + "loss": 0.0268, + "step": 6820 + }, + { + "epoch": 4.290201005025126, + "grad_norm": 0.2370511293411255, + "learning_rate": 9.161859105871013e-05, + "loss": 0.021, + "step": 6830 + }, + { + "epoch": 4.296482412060302, + "grad_norm": 0.3025975823402405, + "learning_rate": 9.158801971034832e-05, + "loss": 0.0212, + "step": 6840 + }, + { + "epoch": 4.302763819095477, + "grad_norm": 0.40096548199653625, + "learning_rate": 9.155739782862107e-05, + "loss": 0.0246, + "step": 6850 + }, + { + "epoch": 4.309045226130653, + "grad_norm": 0.41571828722953796, + "learning_rate": 9.152672545073687e-05, + "loss": 0.0261, + "step": 6860 + }, + { + "epoch": 4.315326633165829, + "grad_norm": 0.4207042455673218, + "learning_rate": 9.149600261396551e-05, + "loss": 0.0317, + "step": 6870 + }, + { + "epoch": 4.321608040201005, + "grad_norm": 0.38248777389526367, + "learning_rate": 9.146522935563817e-05, + "loss": 0.0332, + "step": 6880 + }, + { + "epoch": 4.327889447236181, + "grad_norm": 0.3359278738498688, + "learning_rate": 9.143440571314723e-05, + "loss": 0.0275, + "step": 6890 + }, + { + "epoch": 4.334170854271357, + "grad_norm": 0.42426225543022156, + "learning_rate": 9.140353172394636e-05, + "loss": 0.0296, + "step": 6900 + }, + { + "epoch": 4.340452261306533, + "grad_norm": 0.3157092034816742, + "learning_rate": 9.137260742555032e-05, + "loss": 0.0309, + "step": 6910 + }, + { + "epoch": 4.346733668341709, + "grad_norm": 0.2495763599872589, + "learning_rate": 9.13416328555351e-05, + "loss": 0.0342, + "step": 6920 + }, + { + "epoch": 4.353015075376884, + "grad_norm": 0.36803483963012695, + "learning_rate": 9.131060805153769e-05, + "loss": 0.028, + "step": 6930 + }, + { + "epoch": 4.35929648241206, + "grad_norm": 0.36660051345825195, + "learning_rate": 9.127953305125618e-05, + "loss": 0.0266, + "step": 6940 + }, + { + "epoch": 4.365577889447236, + "grad_norm": 0.2811267375946045, + "learning_rate": 9.124840789244958e-05, + "loss": 0.0265, + "step": 6950 + }, + { + "epoch": 4.371859296482412, + "grad_norm": 0.28998008370399475, + "learning_rate": 9.121723261293794e-05, + "loss": 0.0268, + "step": 6960 + }, + { + "epoch": 4.378140703517588, + "grad_norm": 0.3667970895767212, + "learning_rate": 9.118600725060214e-05, + "loss": 0.0248, + "step": 6970 + }, + { + "epoch": 4.384422110552764, + "grad_norm": 0.3147960305213928, + "learning_rate": 9.115473184338392e-05, + "loss": 0.0282, + "step": 6980 + }, + { + "epoch": 4.39070351758794, + "grad_norm": 0.2640340030193329, + "learning_rate": 9.11234064292859e-05, + "loss": 0.0297, + "step": 6990 + }, + { + "epoch": 4.396984924623116, + "grad_norm": 0.264731764793396, + "learning_rate": 9.109203104637137e-05, + "loss": 0.0264, + "step": 7000 + }, + { + "epoch": 4.403266331658291, + "grad_norm": 0.32718175649642944, + "learning_rate": 9.10606057327644e-05, + "loss": 0.0297, + "step": 7010 + }, + { + "epoch": 4.409547738693467, + "grad_norm": 0.34609454870224, + "learning_rate": 9.10291305266497e-05, + "loss": 0.0232, + "step": 7020 + }, + { + "epoch": 4.415829145728643, + "grad_norm": 0.3056018352508545, + "learning_rate": 9.099760546627261e-05, + "loss": 0.0243, + "step": 7030 + }, + { + "epoch": 4.422110552763819, + "grad_norm": 0.3392927944660187, + "learning_rate": 9.096603058993907e-05, + "loss": 0.0288, + "step": 7040 + }, + { + "epoch": 4.428391959798995, + "grad_norm": 0.5688233971595764, + "learning_rate": 9.093440593601552e-05, + "loss": 0.0318, + "step": 7050 + }, + { + "epoch": 4.434673366834171, + "grad_norm": 0.2772183418273926, + "learning_rate": 9.090273154292889e-05, + "loss": 0.0284, + "step": 7060 + }, + { + "epoch": 4.440954773869347, + "grad_norm": 0.25148868560791016, + "learning_rate": 9.087100744916656e-05, + "loss": 0.0274, + "step": 7070 + }, + { + "epoch": 4.447236180904523, + "grad_norm": 0.4513634145259857, + "learning_rate": 9.08392336932763e-05, + "loss": 0.0231, + "step": 7080 + }, + { + "epoch": 4.453517587939698, + "grad_norm": 0.22159737348556519, + "learning_rate": 9.080741031386619e-05, + "loss": 0.0255, + "step": 7090 + }, + { + "epoch": 4.459798994974874, + "grad_norm": 0.2512880563735962, + "learning_rate": 9.077553734960468e-05, + "loss": 0.0272, + "step": 7100 + }, + { + "epoch": 4.46608040201005, + "grad_norm": 0.2888525128364563, + "learning_rate": 9.074361483922041e-05, + "loss": 0.0291, + "step": 7110 + }, + { + "epoch": 4.472361809045226, + "grad_norm": 0.37154054641723633, + "learning_rate": 9.071164282150224e-05, + "loss": 0.0231, + "step": 7120 + }, + { + "epoch": 4.478643216080402, + "grad_norm": 0.37089043855667114, + "learning_rate": 9.067962133529919e-05, + "loss": 0.0275, + "step": 7130 + }, + { + "epoch": 4.484924623115578, + "grad_norm": 0.19012053310871124, + "learning_rate": 9.064755041952036e-05, + "loss": 0.0257, + "step": 7140 + }, + { + "epoch": 4.491206030150754, + "grad_norm": 0.2934897840023041, + "learning_rate": 9.061543011313497e-05, + "loss": 0.0312, + "step": 7150 + }, + { + "epoch": 4.49748743718593, + "grad_norm": 0.23064038157463074, + "learning_rate": 9.05832604551722e-05, + "loss": 0.0275, + "step": 7160 + }, + { + "epoch": 4.5037688442211055, + "grad_norm": 0.24207785725593567, + "learning_rate": 9.055104148472124e-05, + "loss": 0.0258, + "step": 7170 + }, + { + "epoch": 4.510050251256281, + "grad_norm": 0.23824989795684814, + "learning_rate": 9.051877324093114e-05, + "loss": 0.029, + "step": 7180 + }, + { + "epoch": 4.516331658291457, + "grad_norm": 0.33671101927757263, + "learning_rate": 9.04864557630109e-05, + "loss": 0.0326, + "step": 7190 + }, + { + "epoch": 4.522613065326633, + "grad_norm": 0.40902507305145264, + "learning_rate": 9.045408909022928e-05, + "loss": 0.0322, + "step": 7200 + }, + { + "epoch": 4.528894472361809, + "grad_norm": 0.35681191086769104, + "learning_rate": 9.042167326191484e-05, + "loss": 0.0265, + "step": 7210 + }, + { + "epoch": 4.535175879396985, + "grad_norm": 0.47573965787887573, + "learning_rate": 9.038920831745587e-05, + "loss": 0.0288, + "step": 7220 + }, + { + "epoch": 4.541457286432161, + "grad_norm": 0.3065149486064911, + "learning_rate": 9.035669429630036e-05, + "loss": 0.0308, + "step": 7230 + }, + { + "epoch": 4.547738693467337, + "grad_norm": 0.3286792039871216, + "learning_rate": 9.032413123795588e-05, + "loss": 0.0279, + "step": 7240 + }, + { + "epoch": 4.5540201005025125, + "grad_norm": 0.5413084030151367, + "learning_rate": 9.029151918198962e-05, + "loss": 0.0246, + "step": 7250 + }, + { + "epoch": 4.560301507537688, + "grad_norm": 0.2910110056400299, + "learning_rate": 9.025885816802833e-05, + "loss": 0.028, + "step": 7260 + }, + { + "epoch": 4.566582914572864, + "grad_norm": 0.40109527111053467, + "learning_rate": 9.02261482357582e-05, + "loss": 0.03, + "step": 7270 + }, + { + "epoch": 4.572864321608041, + "grad_norm": 0.21857570111751556, + "learning_rate": 9.019338942492485e-05, + "loss": 0.0227, + "step": 7280 + }, + { + "epoch": 4.5791457286432165, + "grad_norm": 0.24216684699058533, + "learning_rate": 9.01605817753334e-05, + "loss": 0.0226, + "step": 7290 + }, + { + "epoch": 4.585427135678392, + "grad_norm": 0.19824083149433136, + "learning_rate": 9.012772532684818e-05, + "loss": 0.027, + "step": 7300 + }, + { + "epoch": 4.591708542713568, + "grad_norm": 0.3478552997112274, + "learning_rate": 9.00948201193929e-05, + "loss": 0.0271, + "step": 7310 + }, + { + "epoch": 4.597989949748744, + "grad_norm": 0.30900657176971436, + "learning_rate": 9.006186619295047e-05, + "loss": 0.026, + "step": 7320 + }, + { + "epoch": 4.6042713567839195, + "grad_norm": 0.4580268859863281, + "learning_rate": 9.002886358756305e-05, + "loss": 0.0301, + "step": 7330 + }, + { + "epoch": 4.610552763819095, + "grad_norm": 0.29740196466445923, + "learning_rate": 8.999581234333188e-05, + "loss": 0.0264, + "step": 7340 + }, + { + "epoch": 4.616834170854271, + "grad_norm": 0.2530321180820465, + "learning_rate": 8.996271250041734e-05, + "loss": 0.0275, + "step": 7350 + }, + { + "epoch": 4.623115577889447, + "grad_norm": 0.37428829073905945, + "learning_rate": 8.99295640990389e-05, + "loss": 0.0238, + "step": 7360 + }, + { + "epoch": 4.6293969849246235, + "grad_norm": 0.25250789523124695, + "learning_rate": 8.989636717947494e-05, + "loss": 0.0228, + "step": 7370 + }, + { + "epoch": 4.635678391959799, + "grad_norm": 0.33328962326049805, + "learning_rate": 8.986312178206291e-05, + "loss": 0.0311, + "step": 7380 + }, + { + "epoch": 4.641959798994975, + "grad_norm": 0.5096357464790344, + "learning_rate": 8.982982794719904e-05, + "loss": 0.03, + "step": 7390 + }, + { + "epoch": 4.648241206030151, + "grad_norm": 0.35762646794319153, + "learning_rate": 8.979648571533852e-05, + "loss": 0.0274, + "step": 7400 + }, + { + "epoch": 4.6545226130653266, + "grad_norm": 0.27665406465530396, + "learning_rate": 8.976309512699528e-05, + "loss": 0.0252, + "step": 7410 + }, + { + "epoch": 4.660804020100502, + "grad_norm": 0.3409089148044586, + "learning_rate": 8.972965622274205e-05, + "loss": 0.026, + "step": 7420 + }, + { + "epoch": 4.667085427135678, + "grad_norm": 0.2520487308502197, + "learning_rate": 8.969616904321026e-05, + "loss": 0.0251, + "step": 7430 + }, + { + "epoch": 4.673366834170855, + "grad_norm": 0.31582289934158325, + "learning_rate": 8.966263362908998e-05, + "loss": 0.029, + "step": 7440 + }, + { + "epoch": 4.6796482412060305, + "grad_norm": 0.32974308729171753, + "learning_rate": 8.962905002112989e-05, + "loss": 0.0273, + "step": 7450 + }, + { + "epoch": 4.685929648241206, + "grad_norm": 0.32970190048217773, + "learning_rate": 8.959541826013724e-05, + "loss": 0.0245, + "step": 7460 + }, + { + "epoch": 4.692211055276382, + "grad_norm": 0.2563774287700653, + "learning_rate": 8.956173838697781e-05, + "loss": 0.0284, + "step": 7470 + }, + { + "epoch": 4.698492462311558, + "grad_norm": 0.20923945307731628, + "learning_rate": 8.952801044257581e-05, + "loss": 0.0239, + "step": 7480 + }, + { + "epoch": 4.704773869346734, + "grad_norm": 0.3112412095069885, + "learning_rate": 8.949423446791387e-05, + "loss": 0.0259, + "step": 7490 + }, + { + "epoch": 4.711055276381909, + "grad_norm": 0.2621079385280609, + "learning_rate": 8.9460410504033e-05, + "loss": 0.0285, + "step": 7500 + }, + { + "epoch": 4.717336683417085, + "grad_norm": 0.40748852491378784, + "learning_rate": 8.942653859203248e-05, + "loss": 0.0258, + "step": 7510 + }, + { + "epoch": 4.723618090452261, + "grad_norm": 0.25703927874565125, + "learning_rate": 8.93926187730699e-05, + "loss": 0.0262, + "step": 7520 + }, + { + "epoch": 4.7298994974874375, + "grad_norm": 0.457692414522171, + "learning_rate": 8.935865108836102e-05, + "loss": 0.0276, + "step": 7530 + }, + { + "epoch": 4.736180904522613, + "grad_norm": 0.32499468326568604, + "learning_rate": 8.932463557917981e-05, + "loss": 0.0272, + "step": 7540 + }, + { + "epoch": 4.742462311557789, + "grad_norm": 0.25809821486473083, + "learning_rate": 8.929057228685829e-05, + "loss": 0.0227, + "step": 7550 + }, + { + "epoch": 4.748743718592965, + "grad_norm": 0.4343753457069397, + "learning_rate": 8.925646125278658e-05, + "loss": 0.0248, + "step": 7560 + }, + { + "epoch": 4.755025125628141, + "grad_norm": 0.27087312936782837, + "learning_rate": 8.92223025184128e-05, + "loss": 0.0267, + "step": 7570 + }, + { + "epoch": 4.761306532663316, + "grad_norm": 0.32985416054725647, + "learning_rate": 8.918809612524303e-05, + "loss": 0.0289, + "step": 7580 + }, + { + "epoch": 4.767587939698492, + "grad_norm": 0.29161617159843445, + "learning_rate": 8.91538421148413e-05, + "loss": 0.0239, + "step": 7590 + }, + { + "epoch": 4.773869346733669, + "grad_norm": 0.5662843585014343, + "learning_rate": 8.91195405288294e-05, + "loss": 0.0287, + "step": 7600 + }, + { + "epoch": 4.780150753768845, + "grad_norm": 0.26126715540885925, + "learning_rate": 8.908519140888705e-05, + "loss": 0.0232, + "step": 7610 + }, + { + "epoch": 4.78643216080402, + "grad_norm": 0.3081778287887573, + "learning_rate": 8.905079479675164e-05, + "loss": 0.025, + "step": 7620 + }, + { + "epoch": 4.792713567839196, + "grad_norm": 0.34764108061790466, + "learning_rate": 8.901635073421831e-05, + "loss": 0.0278, + "step": 7630 + }, + { + "epoch": 4.798994974874372, + "grad_norm": 0.2897902727127075, + "learning_rate": 8.898185926313983e-05, + "loss": 0.0279, + "step": 7640 + }, + { + "epoch": 4.805276381909548, + "grad_norm": 0.4881150722503662, + "learning_rate": 8.894732042542659e-05, + "loss": 0.0286, + "step": 7650 + }, + { + "epoch": 4.811557788944723, + "grad_norm": 0.4248841106891632, + "learning_rate": 8.891273426304654e-05, + "loss": 0.0273, + "step": 7660 + }, + { + "epoch": 4.817839195979899, + "grad_norm": 0.495546817779541, + "learning_rate": 8.887810081802515e-05, + "loss": 0.0275, + "step": 7670 + }, + { + "epoch": 4.824120603015075, + "grad_norm": 0.19292764365673065, + "learning_rate": 8.884342013244529e-05, + "loss": 0.0239, + "step": 7680 + }, + { + "epoch": 4.830402010050252, + "grad_norm": 0.32270967960357666, + "learning_rate": 8.880869224844725e-05, + "loss": 0.0258, + "step": 7690 + }, + { + "epoch": 4.836683417085427, + "grad_norm": 0.202042818069458, + "learning_rate": 8.877391720822875e-05, + "loss": 0.0239, + "step": 7700 + }, + { + "epoch": 4.842964824120603, + "grad_norm": 0.25800806283950806, + "learning_rate": 8.873909505404468e-05, + "loss": 0.0213, + "step": 7710 + }, + { + "epoch": 4.849246231155779, + "grad_norm": 0.18559080362319946, + "learning_rate": 8.870422582820726e-05, + "loss": 0.023, + "step": 7720 + }, + { + "epoch": 4.855527638190955, + "grad_norm": 0.34898465871810913, + "learning_rate": 8.866930957308588e-05, + "loss": 0.0258, + "step": 7730 + }, + { + "epoch": 4.86180904522613, + "grad_norm": 0.3318503797054291, + "learning_rate": 8.86343463311071e-05, + "loss": 0.0258, + "step": 7740 + }, + { + "epoch": 4.868090452261306, + "grad_norm": 0.21038712561130524, + "learning_rate": 8.859933614475452e-05, + "loss": 0.022, + "step": 7750 + }, + { + "epoch": 4.874371859296483, + "grad_norm": 0.2756730318069458, + "learning_rate": 8.85642790565689e-05, + "loss": 0.0245, + "step": 7760 + }, + { + "epoch": 4.880653266331659, + "grad_norm": 0.2682247459888458, + "learning_rate": 8.852917510914782e-05, + "loss": 0.022, + "step": 7770 + }, + { + "epoch": 4.886934673366834, + "grad_norm": 0.19023752212524414, + "learning_rate": 8.84940243451459e-05, + "loss": 0.0248, + "step": 7780 + }, + { + "epoch": 4.89321608040201, + "grad_norm": 0.2874515950679779, + "learning_rate": 8.845882680727469e-05, + "loss": 0.0241, + "step": 7790 + }, + { + "epoch": 4.899497487437186, + "grad_norm": 0.3211495280265808, + "learning_rate": 8.842358253830246e-05, + "loss": 0.0243, + "step": 7800 + }, + { + "epoch": 4.905778894472362, + "grad_norm": 0.5323463082313538, + "learning_rate": 8.838829158105434e-05, + "loss": 0.0208, + "step": 7810 + }, + { + "epoch": 4.9120603015075375, + "grad_norm": 0.3491944968700409, + "learning_rate": 8.835295397841217e-05, + "loss": 0.0315, + "step": 7820 + }, + { + "epoch": 4.918341708542713, + "grad_norm": 0.26018664240837097, + "learning_rate": 8.831756977331447e-05, + "loss": 0.0306, + "step": 7830 + }, + { + "epoch": 4.924623115577889, + "grad_norm": 0.290791779756546, + "learning_rate": 8.828213900875638e-05, + "loss": 0.0249, + "step": 7840 + }, + { + "epoch": 4.930904522613066, + "grad_norm": 0.2629106640815735, + "learning_rate": 8.824666172778964e-05, + "loss": 0.0285, + "step": 7850 + }, + { + "epoch": 4.937185929648241, + "grad_norm": 0.35335010290145874, + "learning_rate": 8.821113797352245e-05, + "loss": 0.0219, + "step": 7860 + }, + { + "epoch": 4.943467336683417, + "grad_norm": 0.28409820795059204, + "learning_rate": 8.817556778911956e-05, + "loss": 0.0209, + "step": 7870 + }, + { + "epoch": 4.949748743718593, + "grad_norm": 0.19491563737392426, + "learning_rate": 8.81399512178021e-05, + "loss": 0.023, + "step": 7880 + }, + { + "epoch": 4.956030150753769, + "grad_norm": 0.31930771470069885, + "learning_rate": 8.810428830284752e-05, + "loss": 0.0297, + "step": 7890 + }, + { + "epoch": 4.9623115577889445, + "grad_norm": 0.38047343492507935, + "learning_rate": 8.806857908758967e-05, + "loss": 0.0264, + "step": 7900 + }, + { + "epoch": 4.96859296482412, + "grad_norm": 0.3519040048122406, + "learning_rate": 8.80328236154186e-05, + "loss": 0.0307, + "step": 7910 + }, + { + "epoch": 4.974874371859297, + "grad_norm": 0.42882952094078064, + "learning_rate": 8.799702192978056e-05, + "loss": 0.0274, + "step": 7920 + }, + { + "epoch": 4.981155778894473, + "grad_norm": 0.347683310508728, + "learning_rate": 8.7961174074178e-05, + "loss": 0.0257, + "step": 7930 + }, + { + "epoch": 4.9874371859296485, + "grad_norm": 0.4148564636707306, + "learning_rate": 8.792528009216942e-05, + "loss": 0.0271, + "step": 7940 + }, + { + "epoch": 4.993718592964824, + "grad_norm": 0.6030399203300476, + "learning_rate": 8.788934002736943e-05, + "loss": 0.0312, + "step": 7950 + }, + { + "epoch": 5.0, + "grad_norm": 1.04879629611969, + "learning_rate": 8.785335392344857e-05, + "loss": 0.0347, + "step": 7960 + }, + { + "epoch": 5.006281407035176, + "grad_norm": 0.45246580243110657, + "learning_rate": 8.781732182413336e-05, + "loss": 0.035, + "step": 7970 + }, + { + "epoch": 5.0125628140703515, + "grad_norm": 0.36123958230018616, + "learning_rate": 8.778124377320618e-05, + "loss": 0.031, + "step": 7980 + }, + { + "epoch": 5.018844221105527, + "grad_norm": 0.3607802391052246, + "learning_rate": 8.774511981450529e-05, + "loss": 0.0346, + "step": 7990 + }, + { + "epoch": 5.025125628140704, + "grad_norm": 0.2563689351081848, + "learning_rate": 8.770894999192467e-05, + "loss": 0.0273, + "step": 8000 + }, + { + "epoch": 5.03140703517588, + "grad_norm": 0.3936782479286194, + "learning_rate": 8.767273434941412e-05, + "loss": 0.0294, + "step": 8010 + }, + { + "epoch": 5.0376884422110555, + "grad_norm": 0.44221338629722595, + "learning_rate": 8.7636472930979e-05, + "loss": 0.0299, + "step": 8020 + }, + { + "epoch": 5.043969849246231, + "grad_norm": 0.269692599773407, + "learning_rate": 8.76001657806804e-05, + "loss": 0.0254, + "step": 8030 + }, + { + "epoch": 5.050251256281407, + "grad_norm": 0.4202915132045746, + "learning_rate": 8.756381294263491e-05, + "loss": 0.0282, + "step": 8040 + }, + { + "epoch": 5.056532663316583, + "grad_norm": 0.26435768604278564, + "learning_rate": 8.752741446101464e-05, + "loss": 0.0234, + "step": 8050 + }, + { + "epoch": 5.062814070351759, + "grad_norm": 0.22500240802764893, + "learning_rate": 8.749097038004722e-05, + "loss": 0.024, + "step": 8060 + }, + { + "epoch": 5.069095477386934, + "grad_norm": 0.31065401434898376, + "learning_rate": 8.745448074401561e-05, + "loss": 0.0239, + "step": 8070 + }, + { + "epoch": 5.075376884422111, + "grad_norm": 0.32205188274383545, + "learning_rate": 8.741794559725818e-05, + "loss": 0.0243, + "step": 8080 + }, + { + "epoch": 5.081658291457287, + "grad_norm": 0.3522653877735138, + "learning_rate": 8.738136498416856e-05, + "loss": 0.023, + "step": 8090 + }, + { + "epoch": 5.0879396984924625, + "grad_norm": 0.32308831810951233, + "learning_rate": 8.734473894919564e-05, + "loss": 0.0254, + "step": 8100 + }, + { + "epoch": 5.094221105527638, + "grad_norm": 0.22363929450511932, + "learning_rate": 8.730806753684353e-05, + "loss": 0.0223, + "step": 8110 + }, + { + "epoch": 5.100502512562814, + "grad_norm": 0.3351522386074066, + "learning_rate": 8.727135079167144e-05, + "loss": 0.0263, + "step": 8120 + }, + { + "epoch": 5.10678391959799, + "grad_norm": 0.42227211594581604, + "learning_rate": 8.723458875829368e-05, + "loss": 0.0293, + "step": 8130 + }, + { + "epoch": 5.113065326633166, + "grad_norm": 0.3576669692993164, + "learning_rate": 8.719778148137958e-05, + "loss": 0.0257, + "step": 8140 + }, + { + "epoch": 5.119346733668341, + "grad_norm": 0.36446499824523926, + "learning_rate": 8.716092900565346e-05, + "loss": 0.0276, + "step": 8150 + }, + { + "epoch": 5.125628140703517, + "grad_norm": 0.380591481924057, + "learning_rate": 8.712403137589456e-05, + "loss": 0.0283, + "step": 8160 + }, + { + "epoch": 5.131909547738694, + "grad_norm": 0.5622051358222961, + "learning_rate": 8.708708863693697e-05, + "loss": 0.0248, + "step": 8170 + }, + { + "epoch": 5.13819095477387, + "grad_norm": 0.2899767756462097, + "learning_rate": 8.70501008336696e-05, + "loss": 0.0276, + "step": 8180 + }, + { + "epoch": 5.144472361809045, + "grad_norm": 0.3737749457359314, + "learning_rate": 8.701306801103611e-05, + "loss": 0.0298, + "step": 8190 + }, + { + "epoch": 5.150753768844221, + "grad_norm": 0.3837808072566986, + "learning_rate": 8.697599021403489e-05, + "loss": 0.0268, + "step": 8200 + }, + { + "epoch": 5.157035175879397, + "grad_norm": 0.3062860667705536, + "learning_rate": 8.693886748771896e-05, + "loss": 0.022, + "step": 8210 + }, + { + "epoch": 5.163316582914573, + "grad_norm": 0.35778114199638367, + "learning_rate": 8.690169987719593e-05, + "loss": 0.0287, + "step": 8220 + }, + { + "epoch": 5.169597989949748, + "grad_norm": 0.388101190328598, + "learning_rate": 8.686448742762791e-05, + "loss": 0.0223, + "step": 8230 + }, + { + "epoch": 5.175879396984925, + "grad_norm": 0.23058559000492096, + "learning_rate": 8.682723018423161e-05, + "loss": 0.0267, + "step": 8240 + }, + { + "epoch": 5.182160804020101, + "grad_norm": 0.3470137119293213, + "learning_rate": 8.678992819227804e-05, + "loss": 0.024, + "step": 8250 + }, + { + "epoch": 5.188442211055277, + "grad_norm": 0.38234877586364746, + "learning_rate": 8.675258149709265e-05, + "loss": 0.0251, + "step": 8260 + }, + { + "epoch": 5.194723618090452, + "grad_norm": 0.32590577006340027, + "learning_rate": 8.67151901440552e-05, + "loss": 0.0279, + "step": 8270 + }, + { + "epoch": 5.201005025125628, + "grad_norm": 0.22873982787132263, + "learning_rate": 8.667775417859971e-05, + "loss": 0.0323, + "step": 8280 + }, + { + "epoch": 5.207286432160804, + "grad_norm": 0.2541535496711731, + "learning_rate": 8.664027364621441e-05, + "loss": 0.0265, + "step": 8290 + }, + { + "epoch": 5.21356783919598, + "grad_norm": 0.3016241192817688, + "learning_rate": 8.660274859244168e-05, + "loss": 0.0228, + "step": 8300 + }, + { + "epoch": 5.219849246231155, + "grad_norm": 0.33894112706184387, + "learning_rate": 8.656517906287799e-05, + "loss": 0.0212, + "step": 8310 + }, + { + "epoch": 5.226130653266332, + "grad_norm": 0.33989858627319336, + "learning_rate": 8.652756510317387e-05, + "loss": 0.0296, + "step": 8320 + }, + { + "epoch": 5.232412060301508, + "grad_norm": 0.35444310307502747, + "learning_rate": 8.648990675903382e-05, + "loss": 0.0245, + "step": 8330 + }, + { + "epoch": 5.238693467336684, + "grad_norm": 0.3144480884075165, + "learning_rate": 8.64522040762163e-05, + "loss": 0.0269, + "step": 8340 + }, + { + "epoch": 5.244974874371859, + "grad_norm": 0.2558532953262329, + "learning_rate": 8.64144571005336e-05, + "loss": 0.0234, + "step": 8350 + }, + { + "epoch": 5.251256281407035, + "grad_norm": 0.2356215864419937, + "learning_rate": 8.637666587785184e-05, + "loss": 0.025, + "step": 8360 + }, + { + "epoch": 5.257537688442211, + "grad_norm": 0.3928479552268982, + "learning_rate": 8.633883045409096e-05, + "loss": 0.0241, + "step": 8370 + }, + { + "epoch": 5.263819095477387, + "grad_norm": 0.35047316551208496, + "learning_rate": 8.630095087522457e-05, + "loss": 0.0227, + "step": 8380 + }, + { + "epoch": 5.2701005025125625, + "grad_norm": 0.30769819021224976, + "learning_rate": 8.62630271872799e-05, + "loss": 0.026, + "step": 8390 + }, + { + "epoch": 5.276381909547739, + "grad_norm": 0.22329965233802795, + "learning_rate": 8.62250594363378e-05, + "loss": 0.0242, + "step": 8400 + }, + { + "epoch": 5.282663316582915, + "grad_norm": 0.28908371925354004, + "learning_rate": 8.61870476685327e-05, + "loss": 0.0216, + "step": 8410 + }, + { + "epoch": 5.288944723618091, + "grad_norm": 0.30530714988708496, + "learning_rate": 8.614899193005248e-05, + "loss": 0.0221, + "step": 8420 + }, + { + "epoch": 5.295226130653266, + "grad_norm": 0.25850775837898254, + "learning_rate": 8.611089226713842e-05, + "loss": 0.0242, + "step": 8430 + }, + { + "epoch": 5.301507537688442, + "grad_norm": 0.29022642970085144, + "learning_rate": 8.607274872608521e-05, + "loss": 0.0236, + "step": 8440 + }, + { + "epoch": 5.307788944723618, + "grad_norm": 0.32484468817710876, + "learning_rate": 8.603456135324088e-05, + "loss": 0.0227, + "step": 8450 + }, + { + "epoch": 5.314070351758794, + "grad_norm": 0.31057628989219666, + "learning_rate": 8.599633019500665e-05, + "loss": 0.024, + "step": 8460 + }, + { + "epoch": 5.3203517587939695, + "grad_norm": 0.30286943912506104, + "learning_rate": 8.595805529783702e-05, + "loss": 0.0225, + "step": 8470 + }, + { + "epoch": 5.326633165829146, + "grad_norm": 0.4564790427684784, + "learning_rate": 8.591973670823958e-05, + "loss": 0.0219, + "step": 8480 + }, + { + "epoch": 5.332914572864322, + "grad_norm": 0.3121969699859619, + "learning_rate": 8.588137447277503e-05, + "loss": 0.0297, + "step": 8490 + }, + { + "epoch": 5.339195979899498, + "grad_norm": 0.31617996096611023, + "learning_rate": 8.584296863805709e-05, + "loss": 0.0243, + "step": 8500 + }, + { + "epoch": 5.3454773869346734, + "grad_norm": 0.3601621985435486, + "learning_rate": 8.580451925075248e-05, + "loss": 0.0246, + "step": 8510 + }, + { + "epoch": 5.351758793969849, + "grad_norm": 0.3266506791114807, + "learning_rate": 8.576602635758085e-05, + "loss": 0.025, + "step": 8520 + }, + { + "epoch": 5.358040201005025, + "grad_norm": 0.2536793053150177, + "learning_rate": 8.572749000531469e-05, + "loss": 0.0251, + "step": 8530 + }, + { + "epoch": 5.364321608040201, + "grad_norm": 0.36127668619155884, + "learning_rate": 8.568891024077926e-05, + "loss": 0.0217, + "step": 8540 + }, + { + "epoch": 5.3706030150753765, + "grad_norm": 0.27707967162132263, + "learning_rate": 8.565028711085265e-05, + "loss": 0.0247, + "step": 8550 + }, + { + "epoch": 5.376884422110553, + "grad_norm": 0.35620754957199097, + "learning_rate": 8.561162066246561e-05, + "loss": 0.0236, + "step": 8560 + }, + { + "epoch": 5.383165829145729, + "grad_norm": 0.2778279483318329, + "learning_rate": 8.557291094260152e-05, + "loss": 0.0231, + "step": 8570 + }, + { + "epoch": 5.389447236180905, + "grad_norm": 0.23720428347587585, + "learning_rate": 8.553415799829629e-05, + "loss": 0.0261, + "step": 8580 + }, + { + "epoch": 5.3957286432160805, + "grad_norm": 0.34397655725479126, + "learning_rate": 8.549536187663847e-05, + "loss": 0.0232, + "step": 8590 + }, + { + "epoch": 5.402010050251256, + "grad_norm": 0.32463961839675903, + "learning_rate": 8.545652262476897e-05, + "loss": 0.0267, + "step": 8600 + }, + { + "epoch": 5.408291457286432, + "grad_norm": 0.271507203578949, + "learning_rate": 8.541764028988115e-05, + "loss": 0.026, + "step": 8610 + }, + { + "epoch": 5.414572864321608, + "grad_norm": 0.27203211188316345, + "learning_rate": 8.537871491922071e-05, + "loss": 0.0247, + "step": 8620 + }, + { + "epoch": 5.4208542713567835, + "grad_norm": 0.31602606177330017, + "learning_rate": 8.533974656008566e-05, + "loss": 0.0237, + "step": 8630 + }, + { + "epoch": 5.42713567839196, + "grad_norm": 0.2882000505924225, + "learning_rate": 8.530073525982621e-05, + "loss": 0.0234, + "step": 8640 + }, + { + "epoch": 5.433417085427136, + "grad_norm": 0.29579266905784607, + "learning_rate": 8.526168106584476e-05, + "loss": 0.0232, + "step": 8650 + }, + { + "epoch": 5.439698492462312, + "grad_norm": 0.24966979026794434, + "learning_rate": 8.522258402559586e-05, + "loss": 0.0236, + "step": 8660 + }, + { + "epoch": 5.4459798994974875, + "grad_norm": 0.2739429175853729, + "learning_rate": 8.518344418658613e-05, + "loss": 0.0204, + "step": 8670 + }, + { + "epoch": 5.452261306532663, + "grad_norm": 0.47511157393455505, + "learning_rate": 8.51442615963741e-05, + "loss": 0.0267, + "step": 8680 + }, + { + "epoch": 5.458542713567839, + "grad_norm": 0.4640398919582367, + "learning_rate": 8.510503630257034e-05, + "loss": 0.0245, + "step": 8690 + }, + { + "epoch": 5.464824120603015, + "grad_norm": 0.2460966557264328, + "learning_rate": 8.50657683528373e-05, + "loss": 0.0241, + "step": 8700 + }, + { + "epoch": 5.471105527638191, + "grad_norm": 0.379395455121994, + "learning_rate": 8.502645779488922e-05, + "loss": 0.0212, + "step": 8710 + }, + { + "epoch": 5.477386934673367, + "grad_norm": 0.347065806388855, + "learning_rate": 8.498710467649213e-05, + "loss": 0.0235, + "step": 8720 + }, + { + "epoch": 5.483668341708543, + "grad_norm": 0.30544570088386536, + "learning_rate": 8.494770904546381e-05, + "loss": 0.0213, + "step": 8730 + }, + { + "epoch": 5.489949748743719, + "grad_norm": 0.39561107754707336, + "learning_rate": 8.490827094967363e-05, + "loss": 0.0235, + "step": 8740 + }, + { + "epoch": 5.4962311557788945, + "grad_norm": 0.26307958364486694, + "learning_rate": 8.486879043704263e-05, + "loss": 0.0273, + "step": 8750 + }, + { + "epoch": 5.50251256281407, + "grad_norm": 0.5232782363891602, + "learning_rate": 8.482926755554332e-05, + "loss": 0.0274, + "step": 8760 + }, + { + "epoch": 5.508793969849246, + "grad_norm": 0.2863772511482239, + "learning_rate": 8.478970235319975e-05, + "loss": 0.0229, + "step": 8770 + }, + { + "epoch": 5.515075376884422, + "grad_norm": 0.29917481541633606, + "learning_rate": 8.475009487808737e-05, + "loss": 0.0271, + "step": 8780 + }, + { + "epoch": 5.521356783919598, + "grad_norm": 0.4564204812049866, + "learning_rate": 8.471044517833298e-05, + "loss": 0.0258, + "step": 8790 + }, + { + "epoch": 5.527638190954773, + "grad_norm": 0.48759666085243225, + "learning_rate": 8.467075330211474e-05, + "loss": 0.0232, + "step": 8800 + }, + { + "epoch": 5.53391959798995, + "grad_norm": 0.337454229593277, + "learning_rate": 8.463101929766198e-05, + "loss": 0.0288, + "step": 8810 + }, + { + "epoch": 5.540201005025126, + "grad_norm": 0.2580784857273102, + "learning_rate": 8.459124321325528e-05, + "loss": 0.0222, + "step": 8820 + }, + { + "epoch": 5.546482412060302, + "grad_norm": 0.42616650462150574, + "learning_rate": 8.455142509722634e-05, + "loss": 0.0252, + "step": 8830 + }, + { + "epoch": 5.552763819095477, + "grad_norm": 0.42503923177719116, + "learning_rate": 8.451156499795792e-05, + "loss": 0.0253, + "step": 8840 + }, + { + "epoch": 5.559045226130653, + "grad_norm": 0.38046029210090637, + "learning_rate": 8.447166296388381e-05, + "loss": 0.0269, + "step": 8850 + }, + { + "epoch": 5.565326633165829, + "grad_norm": 0.3981180489063263, + "learning_rate": 8.443171904348872e-05, + "loss": 0.0279, + "step": 8860 + }, + { + "epoch": 5.571608040201005, + "grad_norm": 0.3777122497558594, + "learning_rate": 8.43917332853083e-05, + "loss": 0.0248, + "step": 8870 + }, + { + "epoch": 5.577889447236181, + "grad_norm": 0.31962454319000244, + "learning_rate": 8.435170573792901e-05, + "loss": 0.0217, + "step": 8880 + }, + { + "epoch": 5.584170854271357, + "grad_norm": 0.237874373793602, + "learning_rate": 8.431163644998808e-05, + "loss": 0.0219, + "step": 8890 + }, + { + "epoch": 5.590452261306533, + "grad_norm": 0.3126550316810608, + "learning_rate": 8.42715254701735e-05, + "loss": 0.0217, + "step": 8900 + }, + { + "epoch": 5.596733668341709, + "grad_norm": 0.25522738695144653, + "learning_rate": 8.42313728472239e-05, + "loss": 0.0233, + "step": 8910 + }, + { + "epoch": 5.603015075376884, + "grad_norm": 0.3736885190010071, + "learning_rate": 8.419117862992844e-05, + "loss": 0.0221, + "step": 8920 + }, + { + "epoch": 5.60929648241206, + "grad_norm": 0.3164113461971283, + "learning_rate": 8.415094286712694e-05, + "loss": 0.0224, + "step": 8930 + }, + { + "epoch": 5.615577889447236, + "grad_norm": 0.2683207392692566, + "learning_rate": 8.411066560770965e-05, + "loss": 0.0248, + "step": 8940 + }, + { + "epoch": 5.621859296482412, + "grad_norm": 0.3931278586387634, + "learning_rate": 8.407034690061721e-05, + "loss": 0.0277, + "step": 8950 + }, + { + "epoch": 5.628140703517588, + "grad_norm": 0.5444124341011047, + "learning_rate": 8.402998679484067e-05, + "loss": 0.0243, + "step": 8960 + }, + { + "epoch": 5.634422110552764, + "grad_norm": 0.46885979175567627, + "learning_rate": 8.398958533942135e-05, + "loss": 0.0228, + "step": 8970 + }, + { + "epoch": 5.64070351758794, + "grad_norm": 0.31709495186805725, + "learning_rate": 8.394914258345083e-05, + "loss": 0.0195, + "step": 8980 + }, + { + "epoch": 5.646984924623116, + "grad_norm": 0.4893335998058319, + "learning_rate": 8.390865857607089e-05, + "loss": 0.0284, + "step": 8990 + }, + { + "epoch": 5.653266331658291, + "grad_norm": 0.38482367992401123, + "learning_rate": 8.386813336647339e-05, + "loss": 0.0264, + "step": 9000 + }, + { + "epoch": 5.659547738693467, + "grad_norm": 0.25309768319129944, + "learning_rate": 8.38275670039003e-05, + "loss": 0.0228, + "step": 9010 + }, + { + "epoch": 5.665829145728643, + "grad_norm": 0.2234959602355957, + "learning_rate": 8.378695953764356e-05, + "loss": 0.0199, + "step": 9020 + }, + { + "epoch": 5.672110552763819, + "grad_norm": 0.43675699830055237, + "learning_rate": 8.374631101704509e-05, + "loss": 0.0235, + "step": 9030 + }, + { + "epoch": 5.678391959798995, + "grad_norm": 0.21102704107761383, + "learning_rate": 8.370562149149665e-05, + "loss": 0.0195, + "step": 9040 + }, + { + "epoch": 5.684673366834171, + "grad_norm": 0.3263660669326782, + "learning_rate": 8.366489101043989e-05, + "loss": 0.0215, + "step": 9050 + }, + { + "epoch": 5.690954773869347, + "grad_norm": 0.20126082003116608, + "learning_rate": 8.362411962336614e-05, + "loss": 0.0248, + "step": 9060 + }, + { + "epoch": 5.697236180904523, + "grad_norm": 0.2302996814250946, + "learning_rate": 8.358330737981651e-05, + "loss": 0.0194, + "step": 9070 + }, + { + "epoch": 5.703517587939698, + "grad_norm": 0.2895908057689667, + "learning_rate": 8.35424543293817e-05, + "loss": 0.0227, + "step": 9080 + }, + { + "epoch": 5.709798994974874, + "grad_norm": 0.3053527772426605, + "learning_rate": 8.350156052170206e-05, + "loss": 0.0183, + "step": 9090 + }, + { + "epoch": 5.71608040201005, + "grad_norm": 0.3696857988834381, + "learning_rate": 8.346062600646738e-05, + "loss": 0.0239, + "step": 9100 + }, + { + "epoch": 5.722361809045227, + "grad_norm": 0.24292874336242676, + "learning_rate": 8.341965083341697e-05, + "loss": 0.025, + "step": 9110 + }, + { + "epoch": 5.728643216080402, + "grad_norm": 0.3124007284641266, + "learning_rate": 8.337863505233953e-05, + "loss": 0.0272, + "step": 9120 + }, + { + "epoch": 5.734924623115578, + "grad_norm": 0.2142084836959839, + "learning_rate": 8.33375787130731e-05, + "loss": 0.0195, + "step": 9130 + }, + { + "epoch": 5.741206030150754, + "grad_norm": 0.2713933289051056, + "learning_rate": 8.3296481865505e-05, + "loss": 0.0219, + "step": 9140 + }, + { + "epoch": 5.74748743718593, + "grad_norm": 0.2436063289642334, + "learning_rate": 8.325534455957181e-05, + "loss": 0.0227, + "step": 9150 + }, + { + "epoch": 5.7537688442211055, + "grad_norm": 0.42991364002227783, + "learning_rate": 8.321416684525917e-05, + "loss": 0.0282, + "step": 9160 + }, + { + "epoch": 5.760050251256281, + "grad_norm": 0.2963815927505493, + "learning_rate": 8.317294877260193e-05, + "loss": 0.0234, + "step": 9170 + }, + { + "epoch": 5.766331658291457, + "grad_norm": 0.3156718313694, + "learning_rate": 8.313169039168394e-05, + "loss": 0.026, + "step": 9180 + }, + { + "epoch": 5.772613065326633, + "grad_norm": 0.22411596775054932, + "learning_rate": 8.3090391752638e-05, + "loss": 0.0259, + "step": 9190 + }, + { + "epoch": 5.778894472361809, + "grad_norm": 0.34376177191734314, + "learning_rate": 8.304905290564586e-05, + "loss": 0.0238, + "step": 9200 + }, + { + "epoch": 5.785175879396985, + "grad_norm": 0.3075128495693207, + "learning_rate": 8.300767390093813e-05, + "loss": 0.0236, + "step": 9210 + }, + { + "epoch": 5.791457286432161, + "grad_norm": 0.4047025740146637, + "learning_rate": 8.296625478879417e-05, + "loss": 0.0222, + "step": 9220 + }, + { + "epoch": 5.797738693467337, + "grad_norm": 0.22325825691223145, + "learning_rate": 8.292479561954214e-05, + "loss": 0.0247, + "step": 9230 + }, + { + "epoch": 5.8040201005025125, + "grad_norm": 0.3292871117591858, + "learning_rate": 8.288329644355883e-05, + "loss": 0.0223, + "step": 9240 + }, + { + "epoch": 5.810301507537688, + "grad_norm": 0.1788097321987152, + "learning_rate": 8.284175731126965e-05, + "loss": 0.0222, + "step": 9250 + }, + { + "epoch": 5.816582914572864, + "grad_norm": 0.3363478183746338, + "learning_rate": 8.280017827314854e-05, + "loss": 0.024, + "step": 9260 + }, + { + "epoch": 5.822864321608041, + "grad_norm": 0.23464137315750122, + "learning_rate": 8.2758559379718e-05, + "loss": 0.0216, + "step": 9270 + }, + { + "epoch": 5.8291457286432165, + "grad_norm": 0.3152177631855011, + "learning_rate": 8.271690068154887e-05, + "loss": 0.0214, + "step": 9280 + }, + { + "epoch": 5.835427135678392, + "grad_norm": 0.22605597972869873, + "learning_rate": 8.26752022292604e-05, + "loss": 0.0214, + "step": 9290 + }, + { + "epoch": 5.841708542713568, + "grad_norm": 0.2720118463039398, + "learning_rate": 8.263346407352018e-05, + "loss": 0.0231, + "step": 9300 + }, + { + "epoch": 5.847989949748744, + "grad_norm": 0.41102612018585205, + "learning_rate": 8.259168626504395e-05, + "loss": 0.0207, + "step": 9310 + }, + { + "epoch": 5.8542713567839195, + "grad_norm": 0.3458177149295807, + "learning_rate": 8.254986885459572e-05, + "loss": 0.0244, + "step": 9320 + }, + { + "epoch": 5.860552763819095, + "grad_norm": 0.32249176502227783, + "learning_rate": 8.250801189298757e-05, + "loss": 0.0242, + "step": 9330 + }, + { + "epoch": 5.866834170854271, + "grad_norm": 0.42554768919944763, + "learning_rate": 8.246611543107967e-05, + "loss": 0.0245, + "step": 9340 + }, + { + "epoch": 5.873115577889447, + "grad_norm": 0.29772815108299255, + "learning_rate": 8.242417951978018e-05, + "loss": 0.0232, + "step": 9350 + }, + { + "epoch": 5.8793969849246235, + "grad_norm": 0.2990202307701111, + "learning_rate": 8.238220421004517e-05, + "loss": 0.0225, + "step": 9360 + }, + { + "epoch": 5.885678391959799, + "grad_norm": 0.470551460981369, + "learning_rate": 8.234018955287859e-05, + "loss": 0.03, + "step": 9370 + }, + { + "epoch": 5.891959798994975, + "grad_norm": 0.24817430973052979, + "learning_rate": 8.229813559933224e-05, + "loss": 0.023, + "step": 9380 + }, + { + "epoch": 5.898241206030151, + "grad_norm": 0.334698885679245, + "learning_rate": 8.22560424005056e-05, + "loss": 0.027, + "step": 9390 + }, + { + "epoch": 5.9045226130653266, + "grad_norm": 0.304716020822525, + "learning_rate": 8.22139100075459e-05, + "loss": 0.0218, + "step": 9400 + }, + { + "epoch": 5.910804020100502, + "grad_norm": 0.28615304827690125, + "learning_rate": 8.217173847164799e-05, + "loss": 0.0219, + "step": 9410 + }, + { + "epoch": 5.917085427135678, + "grad_norm": 0.5487352013587952, + "learning_rate": 8.212952784405422e-05, + "loss": 0.0226, + "step": 9420 + }, + { + "epoch": 5.923366834170855, + "grad_norm": 0.280885249376297, + "learning_rate": 8.208727817605451e-05, + "loss": 0.0186, + "step": 9430 + }, + { + "epoch": 5.9296482412060305, + "grad_norm": 0.3216884732246399, + "learning_rate": 8.204498951898618e-05, + "loss": 0.0204, + "step": 9440 + }, + { + "epoch": 5.935929648241206, + "grad_norm": 0.3267388343811035, + "learning_rate": 8.200266192423395e-05, + "loss": 0.0299, + "step": 9450 + }, + { + "epoch": 5.942211055276382, + "grad_norm": 0.3486521542072296, + "learning_rate": 8.196029544322981e-05, + "loss": 0.0233, + "step": 9460 + }, + { + "epoch": 5.948492462311558, + "grad_norm": 0.45434606075286865, + "learning_rate": 8.191789012745305e-05, + "loss": 0.0249, + "step": 9470 + }, + { + "epoch": 5.954773869346734, + "grad_norm": 0.36901530623435974, + "learning_rate": 8.187544602843014e-05, + "loss": 0.0247, + "step": 9480 + }, + { + "epoch": 5.961055276381909, + "grad_norm": 0.25977766513824463, + "learning_rate": 8.183296319773466e-05, + "loss": 0.0234, + "step": 9490 + }, + { + "epoch": 5.967336683417085, + "grad_norm": 0.23057210445404053, + "learning_rate": 8.179044168698721e-05, + "loss": 0.0245, + "step": 9500 + }, + { + "epoch": 5.973618090452261, + "grad_norm": 0.3161158263683319, + "learning_rate": 8.174788154785548e-05, + "loss": 0.0244, + "step": 9510 + }, + { + "epoch": 5.9798994974874375, + "grad_norm": 0.2829616963863373, + "learning_rate": 8.170528283205403e-05, + "loss": 0.0239, + "step": 9520 + }, + { + "epoch": 5.986180904522613, + "grad_norm": 0.3354852795600891, + "learning_rate": 8.166264559134435e-05, + "loss": 0.023, + "step": 9530 + }, + { + "epoch": 5.992462311557789, + "grad_norm": 0.3281971216201782, + "learning_rate": 8.161996987753467e-05, + "loss": 0.0264, + "step": 9540 + }, + { + "epoch": 5.998743718592965, + "grad_norm": 0.36705002188682556, + "learning_rate": 8.157725574248e-05, + "loss": 0.0244, + "step": 9550 + }, + { + "epoch": 6.005025125628141, + "grad_norm": 0.2908255457878113, + "learning_rate": 8.153450323808209e-05, + "loss": 0.0234, + "step": 9560 + }, + { + "epoch": 6.011306532663316, + "grad_norm": 0.2724786698818207, + "learning_rate": 8.149171241628923e-05, + "loss": 0.0256, + "step": 9570 + }, + { + "epoch": 6.017587939698492, + "grad_norm": 0.4373242259025574, + "learning_rate": 8.144888332909631e-05, + "loss": 0.0303, + "step": 9580 + }, + { + "epoch": 6.023869346733668, + "grad_norm": 0.3434266746044159, + "learning_rate": 8.14060160285447e-05, + "loss": 0.026, + "step": 9590 + }, + { + "epoch": 6.030150753768845, + "grad_norm": 0.3103485703468323, + "learning_rate": 8.136311056672223e-05, + "loss": 0.0199, + "step": 9600 + }, + { + "epoch": 6.03643216080402, + "grad_norm": 0.22113661468029022, + "learning_rate": 8.132016699576308e-05, + "loss": 0.0252, + "step": 9610 + }, + { + "epoch": 6.042713567839196, + "grad_norm": 0.4767902195453644, + "learning_rate": 8.127718536784771e-05, + "loss": 0.0269, + "step": 9620 + }, + { + "epoch": 6.048994974874372, + "grad_norm": 0.40791305899620056, + "learning_rate": 8.12341657352029e-05, + "loss": 0.0282, + "step": 9630 + }, + { + "epoch": 6.055276381909548, + "grad_norm": 0.34674617648124695, + "learning_rate": 8.119110815010151e-05, + "loss": 0.021, + "step": 9640 + }, + { + "epoch": 6.061557788944723, + "grad_norm": 0.40955623984336853, + "learning_rate": 8.114801266486261e-05, + "loss": 0.0247, + "step": 9650 + }, + { + "epoch": 6.067839195979899, + "grad_norm": 0.4061400294303894, + "learning_rate": 8.110487933185122e-05, + "loss": 0.0221, + "step": 9660 + }, + { + "epoch": 6.074120603015075, + "grad_norm": 0.20677658915519714, + "learning_rate": 8.106170820347847e-05, + "loss": 0.0232, + "step": 9670 + }, + { + "epoch": 6.080402010050252, + "grad_norm": 0.2079462856054306, + "learning_rate": 8.101849933220132e-05, + "loss": 0.0187, + "step": 9680 + }, + { + "epoch": 6.086683417085427, + "grad_norm": 0.3473798334598541, + "learning_rate": 8.097525277052264e-05, + "loss": 0.0214, + "step": 9690 + }, + { + "epoch": 6.092964824120603, + "grad_norm": 0.2941737174987793, + "learning_rate": 8.093196857099106e-05, + "loss": 0.0205, + "step": 9700 + }, + { + "epoch": 6.099246231155779, + "grad_norm": 0.3311646282672882, + "learning_rate": 8.088864678620096e-05, + "loss": 0.0208, + "step": 9710 + }, + { + "epoch": 6.105527638190955, + "grad_norm": 0.5368154048919678, + "learning_rate": 8.084528746879243e-05, + "loss": 0.0257, + "step": 9720 + }, + { + "epoch": 6.11180904522613, + "grad_norm": 0.26474863290786743, + "learning_rate": 8.080189067145108e-05, + "loss": 0.0191, + "step": 9730 + }, + { + "epoch": 6.118090452261306, + "grad_norm": 0.32742124795913696, + "learning_rate": 8.075845644690813e-05, + "loss": 0.0258, + "step": 9740 + }, + { + "epoch": 6.124371859296482, + "grad_norm": 0.3561902344226837, + "learning_rate": 8.07149848479403e-05, + "loss": 0.0216, + "step": 9750 + }, + { + "epoch": 6.130653266331659, + "grad_norm": 0.26944613456726074, + "learning_rate": 8.067147592736962e-05, + "loss": 0.0193, + "step": 9760 + }, + { + "epoch": 6.136934673366834, + "grad_norm": 0.3250126242637634, + "learning_rate": 8.062792973806357e-05, + "loss": 0.0207, + "step": 9770 + }, + { + "epoch": 6.14321608040201, + "grad_norm": 0.2671772241592407, + "learning_rate": 8.058434633293485e-05, + "loss": 0.0234, + "step": 9780 + }, + { + "epoch": 6.149497487437186, + "grad_norm": 0.26812589168548584, + "learning_rate": 8.054072576494143e-05, + "loss": 0.0172, + "step": 9790 + }, + { + "epoch": 6.155778894472362, + "grad_norm": 0.265675812959671, + "learning_rate": 8.04970680870864e-05, + "loss": 0.0256, + "step": 9800 + }, + { + "epoch": 6.1620603015075375, + "grad_norm": 0.37212783098220825, + "learning_rate": 8.045337335241793e-05, + "loss": 0.0197, + "step": 9810 + }, + { + "epoch": 6.168341708542713, + "grad_norm": 0.3685654103755951, + "learning_rate": 8.040964161402932e-05, + "loss": 0.0208, + "step": 9820 + }, + { + "epoch": 6.174623115577889, + "grad_norm": 0.3796579837799072, + "learning_rate": 8.036587292505869e-05, + "loss": 0.0266, + "step": 9830 + }, + { + "epoch": 6.180904522613066, + "grad_norm": 0.31764721870422363, + "learning_rate": 8.032206733868912e-05, + "loss": 0.0215, + "step": 9840 + }, + { + "epoch": 6.187185929648241, + "grad_norm": 0.2919877767562866, + "learning_rate": 8.027822490814858e-05, + "loss": 0.025, + "step": 9850 + }, + { + "epoch": 6.193467336683417, + "grad_norm": 0.29711028933525085, + "learning_rate": 8.023434568670972e-05, + "loss": 0.0288, + "step": 9860 + }, + { + "epoch": 6.199748743718593, + "grad_norm": 0.33261194825172424, + "learning_rate": 8.019042972768992e-05, + "loss": 0.02, + "step": 9870 + }, + { + "epoch": 6.206030150753769, + "grad_norm": 0.28349247574806213, + "learning_rate": 8.014647708445124e-05, + "loss": 0.0246, + "step": 9880 + }, + { + "epoch": 6.2123115577889445, + "grad_norm": 0.21387149393558502, + "learning_rate": 8.010248781040027e-05, + "loss": 0.0209, + "step": 9890 + }, + { + "epoch": 6.21859296482412, + "grad_norm": 0.3471492528915405, + "learning_rate": 8.005846195898814e-05, + "loss": 0.0244, + "step": 9900 + }, + { + "epoch": 6.224874371859297, + "grad_norm": 0.3045746386051178, + "learning_rate": 8.00143995837104e-05, + "loss": 0.0201, + "step": 9910 + }, + { + "epoch": 6.231155778894473, + "grad_norm": 0.25125256180763245, + "learning_rate": 7.997030073810698e-05, + "loss": 0.0264, + "step": 9920 + }, + { + "epoch": 6.2374371859296485, + "grad_norm": 0.29968178272247314, + "learning_rate": 7.992616547576217e-05, + "loss": 0.0165, + "step": 9930 + }, + { + "epoch": 6.243718592964824, + "grad_norm": 0.2851395905017853, + "learning_rate": 7.988199385030445e-05, + "loss": 0.0236, + "step": 9940 + }, + { + "epoch": 6.25, + "grad_norm": 0.17102482914924622, + "learning_rate": 7.98377859154065e-05, + "loss": 0.0226, + "step": 9950 + }, + { + "epoch": 6.256281407035176, + "grad_norm": 0.30839189887046814, + "learning_rate": 7.979354172478516e-05, + "loss": 0.0273, + "step": 9960 + }, + { + "epoch": 6.2625628140703515, + "grad_norm": 0.27903464436531067, + "learning_rate": 7.974926133220127e-05, + "loss": 0.0183, + "step": 9970 + }, + { + "epoch": 6.268844221105527, + "grad_norm": 0.30706506967544556, + "learning_rate": 7.970494479145967e-05, + "loss": 0.0224, + "step": 9980 + }, + { + "epoch": 6.275125628140703, + "grad_norm": 0.37102529406547546, + "learning_rate": 7.966059215640919e-05, + "loss": 0.0236, + "step": 9990 + }, + { + "epoch": 6.28140703517588, + "grad_norm": 0.2910369038581848, + "learning_rate": 7.961620348094241e-05, + "loss": 0.0221, + "step": 10000 + }, + { + "epoch": 6.2876884422110555, + "grad_norm": 0.2697252929210663, + "learning_rate": 7.957177881899577e-05, + "loss": 0.0226, + "step": 10010 + }, + { + "epoch": 6.293969849246231, + "grad_norm": 0.34704941511154175, + "learning_rate": 7.952731822454944e-05, + "loss": 0.0231, + "step": 10020 + }, + { + "epoch": 6.300251256281407, + "grad_norm": 0.2343139350414276, + "learning_rate": 7.948282175162722e-05, + "loss": 0.0222, + "step": 10030 + }, + { + "epoch": 6.306532663316583, + "grad_norm": 0.2664300203323364, + "learning_rate": 7.943828945429652e-05, + "loss": 0.0204, + "step": 10040 + }, + { + "epoch": 6.312814070351759, + "grad_norm": 0.5102452635765076, + "learning_rate": 7.939372138666827e-05, + "loss": 0.023, + "step": 10050 + }, + { + "epoch": 6.319095477386934, + "grad_norm": 0.3121643662452698, + "learning_rate": 7.934911760289692e-05, + "loss": 0.0233, + "step": 10060 + }, + { + "epoch": 6.325376884422111, + "grad_norm": 0.3364714980125427, + "learning_rate": 7.930447815718022e-05, + "loss": 0.0221, + "step": 10070 + }, + { + "epoch": 6.331658291457287, + "grad_norm": 0.212020605802536, + "learning_rate": 7.925980310375934e-05, + "loss": 0.0227, + "step": 10080 + }, + { + "epoch": 6.3379396984924625, + "grad_norm": 0.32262253761291504, + "learning_rate": 7.921509249691865e-05, + "loss": 0.0209, + "step": 10090 + }, + { + "epoch": 6.344221105527638, + "grad_norm": 0.3108358085155487, + "learning_rate": 7.917034639098579e-05, + "loss": 0.0201, + "step": 10100 + }, + { + "epoch": 6.350502512562814, + "grad_norm": 0.29488882422447205, + "learning_rate": 7.912556484033146e-05, + "loss": 0.0206, + "step": 10110 + }, + { + "epoch": 6.35678391959799, + "grad_norm": 0.29313746094703674, + "learning_rate": 7.908074789936952e-05, + "loss": 0.0209, + "step": 10120 + }, + { + "epoch": 6.363065326633166, + "grad_norm": 0.25015318393707275, + "learning_rate": 7.903589562255673e-05, + "loss": 0.0194, + "step": 10130 + }, + { + "epoch": 6.369346733668341, + "grad_norm": 0.22176992893218994, + "learning_rate": 7.899100806439285e-05, + "loss": 0.0204, + "step": 10140 + }, + { + "epoch": 6.375628140703517, + "grad_norm": 0.26033326983451843, + "learning_rate": 7.894608527942049e-05, + "loss": 0.023, + "step": 10150 + }, + { + "epoch": 6.381909547738694, + "grad_norm": 0.32965901494026184, + "learning_rate": 7.89011273222251e-05, + "loss": 0.0215, + "step": 10160 + }, + { + "epoch": 6.38819095477387, + "grad_norm": 0.4177955687046051, + "learning_rate": 7.885613424743482e-05, + "loss": 0.0222, + "step": 10170 + }, + { + "epoch": 6.394472361809045, + "grad_norm": 0.4120129346847534, + "learning_rate": 7.881110610972044e-05, + "loss": 0.0189, + "step": 10180 + }, + { + "epoch": 6.400753768844221, + "grad_norm": 0.33140307664871216, + "learning_rate": 7.876604296379544e-05, + "loss": 0.0217, + "step": 10190 + }, + { + "epoch": 6.407035175879397, + "grad_norm": 0.3909315764904022, + "learning_rate": 7.872094486441579e-05, + "loss": 0.022, + "step": 10200 + }, + { + "epoch": 6.413316582914573, + "grad_norm": 0.23999257385730743, + "learning_rate": 7.867581186637991e-05, + "loss": 0.0245, + "step": 10210 + }, + { + "epoch": 6.419597989949748, + "grad_norm": 0.28001558780670166, + "learning_rate": 7.863064402452866e-05, + "loss": 0.0192, + "step": 10220 + }, + { + "epoch": 6.425879396984925, + "grad_norm": 0.28556978702545166, + "learning_rate": 7.858544139374524e-05, + "loss": 0.0234, + "step": 10230 + }, + { + "epoch": 6.432160804020101, + "grad_norm": 0.35878413915634155, + "learning_rate": 7.854020402895508e-05, + "loss": 0.0206, + "step": 10240 + }, + { + "epoch": 6.438442211055277, + "grad_norm": 0.2771122455596924, + "learning_rate": 7.849493198512587e-05, + "loss": 0.0217, + "step": 10250 + }, + { + "epoch": 6.444723618090452, + "grad_norm": 0.25113940238952637, + "learning_rate": 7.84496253172674e-05, + "loss": 0.0177, + "step": 10260 + }, + { + "epoch": 6.451005025125628, + "grad_norm": 0.23600871860980988, + "learning_rate": 7.840428408043155e-05, + "loss": 0.0241, + "step": 10270 + }, + { + "epoch": 6.457286432160804, + "grad_norm": 0.2896650433540344, + "learning_rate": 7.835890832971218e-05, + "loss": 0.0267, + "step": 10280 + }, + { + "epoch": 6.46356783919598, + "grad_norm": 0.24635863304138184, + "learning_rate": 7.831349812024512e-05, + "loss": 0.0249, + "step": 10290 + }, + { + "epoch": 6.469849246231155, + "grad_norm": 0.253523588180542, + "learning_rate": 7.826805350720807e-05, + "loss": 0.0209, + "step": 10300 + }, + { + "epoch": 6.476130653266331, + "grad_norm": 0.21351641416549683, + "learning_rate": 7.822257454582049e-05, + "loss": 0.0194, + "step": 10310 + }, + { + "epoch": 6.482412060301508, + "grad_norm": 0.26200997829437256, + "learning_rate": 7.817706129134363e-05, + "loss": 0.025, + "step": 10320 + }, + { + "epoch": 6.488693467336684, + "grad_norm": 0.19615308940410614, + "learning_rate": 7.813151379908036e-05, + "loss": 0.0186, + "step": 10330 + }, + { + "epoch": 6.494974874371859, + "grad_norm": 0.24795280396938324, + "learning_rate": 7.808593212437523e-05, + "loss": 0.0225, + "step": 10340 + }, + { + "epoch": 6.501256281407035, + "grad_norm": 0.36242222785949707, + "learning_rate": 7.804031632261421e-05, + "loss": 0.0201, + "step": 10350 + }, + { + "epoch": 6.507537688442211, + "grad_norm": 0.3760211169719696, + "learning_rate": 7.799466644922484e-05, + "loss": 0.0187, + "step": 10360 + }, + { + "epoch": 6.513819095477387, + "grad_norm": 0.4191446900367737, + "learning_rate": 7.794898255967602e-05, + "loss": 0.0216, + "step": 10370 + }, + { + "epoch": 6.5201005025125625, + "grad_norm": 0.35948118567466736, + "learning_rate": 7.790326470947795e-05, + "loss": 0.0205, + "step": 10380 + }, + { + "epoch": 6.526381909547739, + "grad_norm": 0.34078946709632874, + "learning_rate": 7.785751295418217e-05, + "loss": 0.0216, + "step": 10390 + }, + { + "epoch": 6.532663316582915, + "grad_norm": 0.4186576008796692, + "learning_rate": 7.781172734938136e-05, + "loss": 0.022, + "step": 10400 + }, + { + "epoch": 6.538944723618091, + "grad_norm": 0.21483036875724792, + "learning_rate": 7.776590795070933e-05, + "loss": 0.0188, + "step": 10410 + }, + { + "epoch": 6.545226130653266, + "grad_norm": 0.298566609621048, + "learning_rate": 7.772005481384099e-05, + "loss": 0.0223, + "step": 10420 + }, + { + "epoch": 6.551507537688442, + "grad_norm": 0.5134165287017822, + "learning_rate": 7.767416799449223e-05, + "loss": 0.0201, + "step": 10430 + }, + { + "epoch": 6.557788944723618, + "grad_norm": 0.25708889961242676, + "learning_rate": 7.762824754841984e-05, + "loss": 0.0248, + "step": 10440 + }, + { + "epoch": 6.564070351758794, + "grad_norm": 0.372173011302948, + "learning_rate": 7.758229353142152e-05, + "loss": 0.0216, + "step": 10450 + }, + { + "epoch": 6.5703517587939695, + "grad_norm": 0.2393854260444641, + "learning_rate": 7.753630599933572e-05, + "loss": 0.0203, + "step": 10460 + }, + { + "epoch": 6.576633165829145, + "grad_norm": 0.3151538670063019, + "learning_rate": 7.749028500804161e-05, + "loss": 0.0222, + "step": 10470 + }, + { + "epoch": 6.582914572864322, + "grad_norm": 0.2922440767288208, + "learning_rate": 7.744423061345906e-05, + "loss": 0.0218, + "step": 10480 + }, + { + "epoch": 6.589195979899498, + "grad_norm": 0.36818432807922363, + "learning_rate": 7.73981428715485e-05, + "loss": 0.0231, + "step": 10490 + }, + { + "epoch": 6.5954773869346734, + "grad_norm": 0.3946378827095032, + "learning_rate": 7.735202183831085e-05, + "loss": 0.0207, + "step": 10500 + }, + { + "epoch": 6.601758793969849, + "grad_norm": 0.3421030640602112, + "learning_rate": 7.730586756978757e-05, + "loss": 0.0199, + "step": 10510 + }, + { + "epoch": 6.608040201005025, + "grad_norm": 0.2805671989917755, + "learning_rate": 7.72596801220604e-05, + "loss": 0.0257, + "step": 10520 + }, + { + "epoch": 6.614321608040201, + "grad_norm": 0.1997491419315338, + "learning_rate": 7.721345955125147e-05, + "loss": 0.0235, + "step": 10530 + }, + { + "epoch": 6.6206030150753765, + "grad_norm": 0.4932772219181061, + "learning_rate": 7.71672059135231e-05, + "loss": 0.0204, + "step": 10540 + }, + { + "epoch": 6.626884422110553, + "grad_norm": 0.3377569317817688, + "learning_rate": 7.712091926507787e-05, + "loss": 0.0204, + "step": 10550 + }, + { + "epoch": 6.633165829145729, + "grad_norm": 0.24789312481880188, + "learning_rate": 7.70745996621584e-05, + "loss": 0.0204, + "step": 10560 + }, + { + "epoch": 6.639447236180905, + "grad_norm": 0.2870739996433258, + "learning_rate": 7.702824716104735e-05, + "loss": 0.022, + "step": 10570 + }, + { + "epoch": 6.6457286432160805, + "grad_norm": 0.1735752373933792, + "learning_rate": 7.698186181806743e-05, + "loss": 0.0206, + "step": 10580 + }, + { + "epoch": 6.652010050251256, + "grad_norm": 0.26756933331489563, + "learning_rate": 7.693544368958116e-05, + "loss": 0.0214, + "step": 10590 + }, + { + "epoch": 6.658291457286432, + "grad_norm": 0.27158504724502563, + "learning_rate": 7.688899283199096e-05, + "loss": 0.0196, + "step": 10600 + }, + { + "epoch": 6.664572864321608, + "grad_norm": 0.2205105721950531, + "learning_rate": 7.684250930173901e-05, + "loss": 0.0185, + "step": 10610 + }, + { + "epoch": 6.6708542713567835, + "grad_norm": 0.27551865577697754, + "learning_rate": 7.679599315530718e-05, + "loss": 0.0202, + "step": 10620 + }, + { + "epoch": 6.677135678391959, + "grad_norm": 0.344947874546051, + "learning_rate": 7.674944444921695e-05, + "loss": 0.0186, + "step": 10630 + }, + { + "epoch": 6.683417085427136, + "grad_norm": 0.3017767667770386, + "learning_rate": 7.670286324002944e-05, + "loss": 0.0221, + "step": 10640 + }, + { + "epoch": 6.689698492462312, + "grad_norm": 0.21548901498317719, + "learning_rate": 7.665624958434514e-05, + "loss": 0.0169, + "step": 10650 + }, + { + "epoch": 6.6959798994974875, + "grad_norm": 0.4681090712547302, + "learning_rate": 7.66096035388041e-05, + "loss": 0.0211, + "step": 10660 + }, + { + "epoch": 6.702261306532663, + "grad_norm": 0.25293850898742676, + "learning_rate": 7.656292516008563e-05, + "loss": 0.0197, + "step": 10670 + }, + { + "epoch": 6.708542713567839, + "grad_norm": 0.5078480243682861, + "learning_rate": 7.651621450490837e-05, + "loss": 0.0188, + "step": 10680 + }, + { + "epoch": 6.714824120603015, + "grad_norm": 0.2468118965625763, + "learning_rate": 7.646947163003017e-05, + "loss": 0.0211, + "step": 10690 + }, + { + "epoch": 6.721105527638191, + "grad_norm": 0.4678939878940582, + "learning_rate": 7.642269659224803e-05, + "loss": 0.0216, + "step": 10700 + }, + { + "epoch": 6.727386934673367, + "grad_norm": 0.18523606657981873, + "learning_rate": 7.637588944839803e-05, + "loss": 0.0208, + "step": 10710 + }, + { + "epoch": 6.733668341708543, + "grad_norm": 0.3498935103416443, + "learning_rate": 7.632905025535528e-05, + "loss": 0.0219, + "step": 10720 + }, + { + "epoch": 6.739949748743719, + "grad_norm": 0.21040363609790802, + "learning_rate": 7.628217907003378e-05, + "loss": 0.0178, + "step": 10730 + }, + { + "epoch": 6.7462311557788945, + "grad_norm": 0.3820658028125763, + "learning_rate": 7.623527594938649e-05, + "loss": 0.0198, + "step": 10740 + }, + { + "epoch": 6.75251256281407, + "grad_norm": 0.23536895215511322, + "learning_rate": 7.618834095040509e-05, + "loss": 0.0216, + "step": 10750 + }, + { + "epoch": 6.758793969849246, + "grad_norm": 0.22053612768650055, + "learning_rate": 7.614137413012e-05, + "loss": 0.0204, + "step": 10760 + }, + { + "epoch": 6.765075376884422, + "grad_norm": 0.40420326590538025, + "learning_rate": 7.609437554560042e-05, + "loss": 0.0228, + "step": 10770 + }, + { + "epoch": 6.771356783919598, + "grad_norm": 0.22698457539081573, + "learning_rate": 7.604734525395398e-05, + "loss": 0.0251, + "step": 10780 + }, + { + "epoch": 6.777638190954773, + "grad_norm": 0.3210805058479309, + "learning_rate": 7.600028331232699e-05, + "loss": 0.02, + "step": 10790 + }, + { + "epoch": 6.78391959798995, + "grad_norm": 0.2944237291812897, + "learning_rate": 7.595318977790408e-05, + "loss": 0.0202, + "step": 10800 + }, + { + "epoch": 6.790201005025126, + "grad_norm": 0.254542738199234, + "learning_rate": 7.590606470790836e-05, + "loss": 0.0195, + "step": 10810 + }, + { + "epoch": 6.796482412060302, + "grad_norm": 0.21092692017555237, + "learning_rate": 7.585890815960125e-05, + "loss": 0.0202, + "step": 10820 + }, + { + "epoch": 6.802763819095477, + "grad_norm": 0.2160113900899887, + "learning_rate": 7.581172019028238e-05, + "loss": 0.0238, + "step": 10830 + }, + { + "epoch": 6.809045226130653, + "grad_norm": 0.26821550726890564, + "learning_rate": 7.576450085728959e-05, + "loss": 0.0248, + "step": 10840 + }, + { + "epoch": 6.815326633165829, + "grad_norm": 0.18682992458343506, + "learning_rate": 7.571725021799885e-05, + "loss": 0.0203, + "step": 10850 + }, + { + "epoch": 6.821608040201005, + "grad_norm": 0.28786107897758484, + "learning_rate": 7.566996832982409e-05, + "loss": 0.0187, + "step": 10860 + }, + { + "epoch": 6.827889447236181, + "grad_norm": 0.27012181282043457, + "learning_rate": 7.56226552502173e-05, + "loss": 0.0198, + "step": 10870 + }, + { + "epoch": 6.834170854271357, + "grad_norm": 0.33450907468795776, + "learning_rate": 7.557531103666833e-05, + "loss": 0.0177, + "step": 10880 + }, + { + "epoch": 6.840452261306533, + "grad_norm": 0.20785638689994812, + "learning_rate": 7.552793574670485e-05, + "loss": 0.0185, + "step": 10890 + }, + { + "epoch": 6.846733668341709, + "grad_norm": 0.3278399109840393, + "learning_rate": 7.54805294378923e-05, + "loss": 0.0209, + "step": 10900 + }, + { + "epoch": 6.853015075376884, + "grad_norm": 0.24101728200912476, + "learning_rate": 7.543309216783384e-05, + "loss": 0.017, + "step": 10910 + }, + { + "epoch": 6.85929648241206, + "grad_norm": 0.2172003835439682, + "learning_rate": 7.53856239941702e-05, + "loss": 0.017, + "step": 10920 + }, + { + "epoch": 6.865577889447236, + "grad_norm": 0.2605127990245819, + "learning_rate": 7.533812497457972e-05, + "loss": 0.0201, + "step": 10930 + }, + { + "epoch": 6.871859296482412, + "grad_norm": 0.2789131700992584, + "learning_rate": 7.529059516677814e-05, + "loss": 0.0191, + "step": 10940 + }, + { + "epoch": 6.878140703517588, + "grad_norm": 0.170410618185997, + "learning_rate": 7.524303462851872e-05, + "loss": 0.0204, + "step": 10950 + }, + { + "epoch": 6.884422110552764, + "grad_norm": 0.2760780155658722, + "learning_rate": 7.519544341759192e-05, + "loss": 0.017, + "step": 10960 + }, + { + "epoch": 6.89070351758794, + "grad_norm": 0.17949159443378448, + "learning_rate": 7.514782159182562e-05, + "loss": 0.0181, + "step": 10970 + }, + { + "epoch": 6.896984924623116, + "grad_norm": 0.17671114206314087, + "learning_rate": 7.510016920908481e-05, + "loss": 0.0204, + "step": 10980 + }, + { + "epoch": 6.903266331658291, + "grad_norm": 0.31526312232017517, + "learning_rate": 7.505248632727158e-05, + "loss": 0.0238, + "step": 10990 + }, + { + "epoch": 6.909547738693467, + "grad_norm": 0.4044812023639679, + "learning_rate": 7.50047730043252e-05, + "loss": 0.0234, + "step": 11000 + }, + { + "epoch": 6.915829145728643, + "grad_norm": 0.2581842243671417, + "learning_rate": 7.495702929822183e-05, + "loss": 0.0234, + "step": 11010 + }, + { + "epoch": 6.922110552763819, + "grad_norm": 0.27843156456947327, + "learning_rate": 7.490925526697455e-05, + "loss": 0.0168, + "step": 11020 + }, + { + "epoch": 6.928391959798995, + "grad_norm": 0.3051445484161377, + "learning_rate": 7.486145096863334e-05, + "loss": 0.0254, + "step": 11030 + }, + { + "epoch": 6.934673366834171, + "grad_norm": 0.218070387840271, + "learning_rate": 7.481361646128491e-05, + "loss": 0.0185, + "step": 11040 + }, + { + "epoch": 6.940954773869347, + "grad_norm": 0.24008238315582275, + "learning_rate": 7.476575180305271e-05, + "loss": 0.0204, + "step": 11050 + }, + { + "epoch": 6.947236180904523, + "grad_norm": 0.2584075331687927, + "learning_rate": 7.471785705209682e-05, + "loss": 0.0179, + "step": 11060 + }, + { + "epoch": 6.953517587939698, + "grad_norm": 0.33625638484954834, + "learning_rate": 7.466993226661387e-05, + "loss": 0.0194, + "step": 11070 + }, + { + "epoch": 6.959798994974874, + "grad_norm": 0.2942887544631958, + "learning_rate": 7.462197750483699e-05, + "loss": 0.0202, + "step": 11080 + }, + { + "epoch": 6.96608040201005, + "grad_norm": 0.3100837469100952, + "learning_rate": 7.457399282503574e-05, + "loss": 0.0203, + "step": 11090 + }, + { + "epoch": 6.972361809045227, + "grad_norm": 0.3826773166656494, + "learning_rate": 7.452597828551604e-05, + "loss": 0.0215, + "step": 11100 + }, + { + "epoch": 6.978643216080402, + "grad_norm": 0.34749555587768555, + "learning_rate": 7.447793394462006e-05, + "loss": 0.0267, + "step": 11110 + }, + { + "epoch": 6.984924623115578, + "grad_norm": 0.37785300612449646, + "learning_rate": 7.442985986072624e-05, + "loss": 0.0209, + "step": 11120 + }, + { + "epoch": 6.991206030150754, + "grad_norm": 0.41797155141830444, + "learning_rate": 7.438175609224908e-05, + "loss": 0.0205, + "step": 11130 + }, + { + "epoch": 6.99748743718593, + "grad_norm": 0.28801533579826355, + "learning_rate": 7.433362269763924e-05, + "loss": 0.0214, + "step": 11140 + }, + { + "epoch": 7.0037688442211055, + "grad_norm": 0.30195263028144836, + "learning_rate": 7.428545973538329e-05, + "loss": 0.0264, + "step": 11150 + }, + { + "epoch": 7.010050251256281, + "grad_norm": 0.22220826148986816, + "learning_rate": 7.42372672640038e-05, + "loss": 0.0176, + "step": 11160 + }, + { + "epoch": 7.016331658291457, + "grad_norm": 0.22556257247924805, + "learning_rate": 7.418904534205917e-05, + "loss": 0.0207, + "step": 11170 + }, + { + "epoch": 7.022613065326633, + "grad_norm": 0.2554349899291992, + "learning_rate": 7.414079402814356e-05, + "loss": 0.0202, + "step": 11180 + }, + { + "epoch": 7.028894472361809, + "grad_norm": 0.3513859510421753, + "learning_rate": 7.409251338088687e-05, + "loss": 0.0192, + "step": 11190 + }, + { + "epoch": 7.035175879396985, + "grad_norm": 0.21543194353580475, + "learning_rate": 7.404420345895467e-05, + "loss": 0.0223, + "step": 11200 + }, + { + "epoch": 7.041457286432161, + "grad_norm": 0.18087854981422424, + "learning_rate": 7.399586432104804e-05, + "loss": 0.0218, + "step": 11210 + }, + { + "epoch": 7.047738693467337, + "grad_norm": 0.27543503046035767, + "learning_rate": 7.394749602590358e-05, + "loss": 0.0211, + "step": 11220 + }, + { + "epoch": 7.0540201005025125, + "grad_norm": 0.25289031863212585, + "learning_rate": 7.389909863229336e-05, + "loss": 0.023, + "step": 11230 + }, + { + "epoch": 7.060301507537688, + "grad_norm": 0.26854294538497925, + "learning_rate": 7.385067219902477e-05, + "loss": 0.0233, + "step": 11240 + }, + { + "epoch": 7.066582914572864, + "grad_norm": 0.29156437516212463, + "learning_rate": 7.380221678494049e-05, + "loss": 0.0224, + "step": 11250 + }, + { + "epoch": 7.07286432160804, + "grad_norm": 0.5608501434326172, + "learning_rate": 7.37537324489184e-05, + "loss": 0.0249, + "step": 11260 + }, + { + "epoch": 7.0791457286432165, + "grad_norm": 0.37179625034332275, + "learning_rate": 7.370521924987155e-05, + "loss": 0.0203, + "step": 11270 + }, + { + "epoch": 7.085427135678392, + "grad_norm": 0.2241164892911911, + "learning_rate": 7.365667724674806e-05, + "loss": 0.0193, + "step": 11280 + }, + { + "epoch": 7.091708542713568, + "grad_norm": 0.3449963331222534, + "learning_rate": 7.360810649853105e-05, + "loss": 0.0193, + "step": 11290 + }, + { + "epoch": 7.097989949748744, + "grad_norm": 0.3766011595726013, + "learning_rate": 7.355950706423853e-05, + "loss": 0.0207, + "step": 11300 + }, + { + "epoch": 7.1042713567839195, + "grad_norm": 0.34822145104408264, + "learning_rate": 7.351087900292342e-05, + "loss": 0.0189, + "step": 11310 + }, + { + "epoch": 7.110552763819095, + "grad_norm": 0.21538610756397247, + "learning_rate": 7.346222237367339e-05, + "loss": 0.0194, + "step": 11320 + }, + { + "epoch": 7.116834170854271, + "grad_norm": 0.34439536929130554, + "learning_rate": 7.341353723561081e-05, + "loss": 0.0217, + "step": 11330 + }, + { + "epoch": 7.123115577889447, + "grad_norm": 0.30673766136169434, + "learning_rate": 7.336482364789277e-05, + "loss": 0.02, + "step": 11340 + }, + { + "epoch": 7.1293969849246235, + "grad_norm": 0.3130420744419098, + "learning_rate": 7.331608166971082e-05, + "loss": 0.0222, + "step": 11350 + }, + { + "epoch": 7.135678391959799, + "grad_norm": 0.2717101573944092, + "learning_rate": 7.326731136029108e-05, + "loss": 0.0186, + "step": 11360 + }, + { + "epoch": 7.141959798994975, + "grad_norm": 0.27844080328941345, + "learning_rate": 7.321851277889408e-05, + "loss": 0.0208, + "step": 11370 + }, + { + "epoch": 7.148241206030151, + "grad_norm": 0.15523798763751984, + "learning_rate": 7.316968598481469e-05, + "loss": 0.0167, + "step": 11380 + }, + { + "epoch": 7.1545226130653266, + "grad_norm": 0.2124178260564804, + "learning_rate": 7.312083103738206e-05, + "loss": 0.0176, + "step": 11390 + }, + { + "epoch": 7.160804020100502, + "grad_norm": 0.26974987983703613, + "learning_rate": 7.307194799595958e-05, + "loss": 0.0193, + "step": 11400 + }, + { + "epoch": 7.167085427135678, + "grad_norm": 0.28242823481559753, + "learning_rate": 7.302303691994473e-05, + "loss": 0.0178, + "step": 11410 + }, + { + "epoch": 7.173366834170854, + "grad_norm": 0.17147797346115112, + "learning_rate": 7.29740978687691e-05, + "loss": 0.0201, + "step": 11420 + }, + { + "epoch": 7.1796482412060305, + "grad_norm": 0.3264372944831848, + "learning_rate": 7.292513090189825e-05, + "loss": 0.0191, + "step": 11430 + }, + { + "epoch": 7.185929648241206, + "grad_norm": 0.2010350525379181, + "learning_rate": 7.287613607883163e-05, + "loss": 0.018, + "step": 11440 + }, + { + "epoch": 7.192211055276382, + "grad_norm": 0.26256662607192993, + "learning_rate": 7.282711345910263e-05, + "loss": 0.0199, + "step": 11450 + }, + { + "epoch": 7.198492462311558, + "grad_norm": 0.22287245094776154, + "learning_rate": 7.27780631022783e-05, + "loss": 0.0207, + "step": 11460 + }, + { + "epoch": 7.204773869346734, + "grad_norm": 0.33496370911598206, + "learning_rate": 7.272898506795948e-05, + "loss": 0.026, + "step": 11470 + }, + { + "epoch": 7.211055276381909, + "grad_norm": 0.2900213599205017, + "learning_rate": 7.267987941578058e-05, + "loss": 0.0199, + "step": 11480 + }, + { + "epoch": 7.217336683417085, + "grad_norm": 0.26101019978523254, + "learning_rate": 7.263074620540962e-05, + "loss": 0.0163, + "step": 11490 + }, + { + "epoch": 7.223618090452261, + "grad_norm": 0.2825392186641693, + "learning_rate": 7.25815854965481e-05, + "loss": 0.0176, + "step": 11500 + }, + { + "epoch": 7.2298994974874375, + "grad_norm": 0.21983551979064941, + "learning_rate": 7.253239734893089e-05, + "loss": 0.0185, + "step": 11510 + }, + { + "epoch": 7.236180904522613, + "grad_norm": 0.27311447262763977, + "learning_rate": 7.248318182232623e-05, + "loss": 0.019, + "step": 11520 + }, + { + "epoch": 7.242462311557789, + "grad_norm": 0.3223676383495331, + "learning_rate": 7.243393897653565e-05, + "loss": 0.0229, + "step": 11530 + }, + { + "epoch": 7.248743718592965, + "grad_norm": 0.2837064862251282, + "learning_rate": 7.23846688713938e-05, + "loss": 0.0195, + "step": 11540 + }, + { + "epoch": 7.255025125628141, + "grad_norm": 0.2275300770998001, + "learning_rate": 7.233537156676854e-05, + "loss": 0.0147, + "step": 11550 + }, + { + "epoch": 7.261306532663316, + "grad_norm": 0.2057236135005951, + "learning_rate": 7.228604712256076e-05, + "loss": 0.0194, + "step": 11560 + }, + { + "epoch": 7.267587939698492, + "grad_norm": 0.27775657176971436, + "learning_rate": 7.223669559870426e-05, + "loss": 0.0188, + "step": 11570 + }, + { + "epoch": 7.273869346733669, + "grad_norm": 0.24071310460567474, + "learning_rate": 7.218731705516585e-05, + "loss": 0.0166, + "step": 11580 + }, + { + "epoch": 7.280150753768845, + "grad_norm": 0.2318638414144516, + "learning_rate": 7.21379115519451e-05, + "loss": 0.0186, + "step": 11590 + }, + { + "epoch": 7.28643216080402, + "grad_norm": 0.30982524156570435, + "learning_rate": 7.20884791490743e-05, + "loss": 0.0212, + "step": 11600 + }, + { + "epoch": 7.292713567839196, + "grad_norm": 0.2268175333738327, + "learning_rate": 7.203901990661857e-05, + "loss": 0.0198, + "step": 11610 + }, + { + "epoch": 7.298994974874372, + "grad_norm": 0.2666980028152466, + "learning_rate": 7.198953388467549e-05, + "loss": 0.0184, + "step": 11620 + }, + { + "epoch": 7.305276381909548, + "grad_norm": 0.2802163064479828, + "learning_rate": 7.194002114337526e-05, + "loss": 0.0178, + "step": 11630 + }, + { + "epoch": 7.311557788944723, + "grad_norm": 0.26090723276138306, + "learning_rate": 7.189048174288054e-05, + "loss": 0.0219, + "step": 11640 + }, + { + "epoch": 7.317839195979899, + "grad_norm": 0.2720130383968353, + "learning_rate": 7.184091574338636e-05, + "loss": 0.0179, + "step": 11650 + }, + { + "epoch": 7.324120603015075, + "grad_norm": 0.2929893732070923, + "learning_rate": 7.179132320512009e-05, + "loss": 0.0173, + "step": 11660 + }, + { + "epoch": 7.330402010050252, + "grad_norm": 0.38821807503700256, + "learning_rate": 7.174170418834134e-05, + "loss": 0.0188, + "step": 11670 + }, + { + "epoch": 7.336683417085427, + "grad_norm": 0.3499528467655182, + "learning_rate": 7.169205875334189e-05, + "loss": 0.0189, + "step": 11680 + }, + { + "epoch": 7.342964824120603, + "grad_norm": 0.2278987169265747, + "learning_rate": 7.164238696044562e-05, + "loss": 0.0194, + "step": 11690 + }, + { + "epoch": 7.349246231155779, + "grad_norm": 0.21238131821155548, + "learning_rate": 7.159268887000846e-05, + "loss": 0.0205, + "step": 11700 + }, + { + "epoch": 7.355527638190955, + "grad_norm": 0.31879037618637085, + "learning_rate": 7.154296454241827e-05, + "loss": 0.0219, + "step": 11710 + }, + { + "epoch": 7.36180904522613, + "grad_norm": 0.29950833320617676, + "learning_rate": 7.149321403809479e-05, + "loss": 0.0178, + "step": 11720 + }, + { + "epoch": 7.368090452261306, + "grad_norm": 0.28033775091171265, + "learning_rate": 7.144343741748956e-05, + "loss": 0.0151, + "step": 11730 + }, + { + "epoch": 7.374371859296483, + "grad_norm": 0.3173861801624298, + "learning_rate": 7.139363474108589e-05, + "loss": 0.0189, + "step": 11740 + }, + { + "epoch": 7.380653266331659, + "grad_norm": 0.2662088871002197, + "learning_rate": 7.134380606939874e-05, + "loss": 0.0194, + "step": 11750 + }, + { + "epoch": 7.386934673366834, + "grad_norm": 0.3049890398979187, + "learning_rate": 7.12939514629746e-05, + "loss": 0.0164, + "step": 11760 + }, + { + "epoch": 7.39321608040201, + "grad_norm": 0.20428906381130219, + "learning_rate": 7.124407098239155e-05, + "loss": 0.0173, + "step": 11770 + }, + { + "epoch": 7.399497487437186, + "grad_norm": 0.24870134890079498, + "learning_rate": 7.119416468825908e-05, + "loss": 0.0171, + "step": 11780 + }, + { + "epoch": 7.405778894472362, + "grad_norm": 0.3460098206996918, + "learning_rate": 7.114423264121804e-05, + "loss": 0.0165, + "step": 11790 + }, + { + "epoch": 7.4120603015075375, + "grad_norm": 0.24236410856246948, + "learning_rate": 7.109427490194056e-05, + "loss": 0.0202, + "step": 11800 + }, + { + "epoch": 7.418341708542713, + "grad_norm": 0.22189855575561523, + "learning_rate": 7.104429153113001e-05, + "loss": 0.0159, + "step": 11810 + }, + { + "epoch": 7.424623115577889, + "grad_norm": 0.2926773428916931, + "learning_rate": 7.099428258952092e-05, + "loss": 0.0196, + "step": 11820 + }, + { + "epoch": 7.430904522613066, + "grad_norm": 0.263986736536026, + "learning_rate": 7.094424813787883e-05, + "loss": 0.0222, + "step": 11830 + }, + { + "epoch": 7.437185929648241, + "grad_norm": 0.25867733359336853, + "learning_rate": 7.089418823700034e-05, + "loss": 0.0162, + "step": 11840 + }, + { + "epoch": 7.443467336683417, + "grad_norm": 0.2807507812976837, + "learning_rate": 7.084410294771298e-05, + "loss": 0.0221, + "step": 11850 + }, + { + "epoch": 7.449748743718593, + "grad_norm": 0.2612255811691284, + "learning_rate": 7.079399233087504e-05, + "loss": 0.0188, + "step": 11860 + }, + { + "epoch": 7.456030150753769, + "grad_norm": 0.28247377276420593, + "learning_rate": 7.074385644737568e-05, + "loss": 0.0217, + "step": 11870 + }, + { + "epoch": 7.4623115577889445, + "grad_norm": 0.3859853148460388, + "learning_rate": 7.069369535813473e-05, + "loss": 0.0217, + "step": 11880 + }, + { + "epoch": 7.46859296482412, + "grad_norm": 0.3573700487613678, + "learning_rate": 7.06435091241026e-05, + "loss": 0.0168, + "step": 11890 + }, + { + "epoch": 7.474874371859297, + "grad_norm": 0.27910369634628296, + "learning_rate": 7.059329780626034e-05, + "loss": 0.0224, + "step": 11900 + }, + { + "epoch": 7.481155778894473, + "grad_norm": 0.3811277449131012, + "learning_rate": 7.054306146561944e-05, + "loss": 0.0201, + "step": 11910 + }, + { + "epoch": 7.4874371859296485, + "grad_norm": 0.2843306064605713, + "learning_rate": 7.049280016322178e-05, + "loss": 0.0219, + "step": 11920 + }, + { + "epoch": 7.493718592964824, + "grad_norm": 0.2717389762401581, + "learning_rate": 7.044251396013957e-05, + "loss": 0.0188, + "step": 11930 + }, + { + "epoch": 7.5, + "grad_norm": 0.4124199450016022, + "learning_rate": 7.039220291747528e-05, + "loss": 0.0199, + "step": 11940 + }, + { + "epoch": 7.506281407035176, + "grad_norm": 0.23677101731300354, + "learning_rate": 7.034186709636159e-05, + "loss": 0.0172, + "step": 11950 + }, + { + "epoch": 7.5125628140703515, + "grad_norm": 0.28205549716949463, + "learning_rate": 7.029150655796129e-05, + "loss": 0.0173, + "step": 11960 + }, + { + "epoch": 7.518844221105527, + "grad_norm": 0.4060165286064148, + "learning_rate": 7.024112136346712e-05, + "loss": 0.0234, + "step": 11970 + }, + { + "epoch": 7.525125628140703, + "grad_norm": 0.32671141624450684, + "learning_rate": 7.01907115741019e-05, + "loss": 0.0202, + "step": 11980 + }, + { + "epoch": 7.53140703517588, + "grad_norm": 0.27831122279167175, + "learning_rate": 7.014027725111825e-05, + "loss": 0.018, + "step": 11990 + }, + { + "epoch": 7.5376884422110555, + "grad_norm": 0.18938367068767548, + "learning_rate": 7.008981845579864e-05, + "loss": 0.0167, + "step": 12000 + }, + { + "epoch": 7.543969849246231, + "grad_norm": 0.23781350255012512, + "learning_rate": 7.003933524945528e-05, + "loss": 0.0214, + "step": 12010 + }, + { + "epoch": 7.550251256281407, + "grad_norm": 0.20393674075603485, + "learning_rate": 6.998882769342998e-05, + "loss": 0.0194, + "step": 12020 + }, + { + "epoch": 7.556532663316583, + "grad_norm": 0.2923421859741211, + "learning_rate": 6.993829584909423e-05, + "loss": 0.0197, + "step": 12030 + }, + { + "epoch": 7.562814070351759, + "grad_norm": 0.22450369596481323, + "learning_rate": 6.988773977784895e-05, + "loss": 0.0175, + "step": 12040 + }, + { + "epoch": 7.569095477386934, + "grad_norm": 0.25141239166259766, + "learning_rate": 6.983715954112454e-05, + "loss": 0.02, + "step": 12050 + }, + { + "epoch": 7.575376884422111, + "grad_norm": 0.2446281462907791, + "learning_rate": 6.978655520038079e-05, + "loss": 0.0169, + "step": 12060 + }, + { + "epoch": 7.581658291457287, + "grad_norm": 0.4257443845272064, + "learning_rate": 6.97359268171067e-05, + "loss": 0.0256, + "step": 12070 + }, + { + "epoch": 7.5879396984924625, + "grad_norm": 0.2333296537399292, + "learning_rate": 6.968527445282056e-05, + "loss": 0.0169, + "step": 12080 + }, + { + "epoch": 7.594221105527638, + "grad_norm": 0.24716629087924957, + "learning_rate": 6.963459816906977e-05, + "loss": 0.0191, + "step": 12090 + }, + { + "epoch": 7.600502512562814, + "grad_norm": 0.2676331400871277, + "learning_rate": 6.958389802743077e-05, + "loss": 0.0174, + "step": 12100 + }, + { + "epoch": 7.60678391959799, + "grad_norm": 0.2757915258407593, + "learning_rate": 6.953317408950903e-05, + "loss": 0.0174, + "step": 12110 + }, + { + "epoch": 7.613065326633166, + "grad_norm": 0.21483805775642395, + "learning_rate": 6.94824264169389e-05, + "loss": 0.0177, + "step": 12120 + }, + { + "epoch": 7.619346733668341, + "grad_norm": 0.2527216672897339, + "learning_rate": 6.94316550713836e-05, + "loss": 0.0196, + "step": 12130 + }, + { + "epoch": 7.625628140703517, + "grad_norm": 0.22806167602539062, + "learning_rate": 6.938086011453513e-05, + "loss": 0.0161, + "step": 12140 + }, + { + "epoch": 7.631909547738694, + "grad_norm": 0.25554361939430237, + "learning_rate": 6.93300416081141e-05, + "loss": 0.0172, + "step": 12150 + }, + { + "epoch": 7.63819095477387, + "grad_norm": 0.47685807943344116, + "learning_rate": 6.927919961386984e-05, + "loss": 0.0249, + "step": 12160 + }, + { + "epoch": 7.644472361809045, + "grad_norm": 0.27965274453163147, + "learning_rate": 6.922833419358013e-05, + "loss": 0.0184, + "step": 12170 + }, + { + "epoch": 7.650753768844221, + "grad_norm": 0.20891472697257996, + "learning_rate": 6.917744540905124e-05, + "loss": 0.0194, + "step": 12180 + }, + { + "epoch": 7.657035175879397, + "grad_norm": 0.364086389541626, + "learning_rate": 6.912653332211787e-05, + "loss": 0.0199, + "step": 12190 + }, + { + "epoch": 7.663316582914573, + "grad_norm": 0.19420239329338074, + "learning_rate": 6.9075597994643e-05, + "loss": 0.02, + "step": 12200 + }, + { + "epoch": 7.669597989949748, + "grad_norm": 0.2617897689342499, + "learning_rate": 6.902463948851787e-05, + "loss": 0.018, + "step": 12210 + }, + { + "epoch": 7.675879396984925, + "grad_norm": 0.2913472354412079, + "learning_rate": 6.897365786566184e-05, + "loss": 0.02, + "step": 12220 + }, + { + "epoch": 7.682160804020101, + "grad_norm": 0.2914985418319702, + "learning_rate": 6.892265318802242e-05, + "loss": 0.0191, + "step": 12230 + }, + { + "epoch": 7.688442211055277, + "grad_norm": 0.3215749263763428, + "learning_rate": 6.887162551757507e-05, + "loss": 0.0218, + "step": 12240 + }, + { + "epoch": 7.694723618090452, + "grad_norm": 0.3400065302848816, + "learning_rate": 6.882057491632326e-05, + "loss": 0.02, + "step": 12250 + }, + { + "epoch": 7.701005025125628, + "grad_norm": 0.2995882034301758, + "learning_rate": 6.876950144629824e-05, + "loss": 0.0208, + "step": 12260 + }, + { + "epoch": 7.707286432160804, + "grad_norm": 0.3265787661075592, + "learning_rate": 6.87184051695591e-05, + "loss": 0.0176, + "step": 12270 + }, + { + "epoch": 7.71356783919598, + "grad_norm": 0.27934345602989197, + "learning_rate": 6.866728614819268e-05, + "loss": 0.0181, + "step": 12280 + }, + { + "epoch": 7.719849246231155, + "grad_norm": 0.33224862813949585, + "learning_rate": 6.861614444431337e-05, + "loss": 0.02, + "step": 12290 + }, + { + "epoch": 7.726130653266331, + "grad_norm": 0.1601611077785492, + "learning_rate": 6.856498012006318e-05, + "loss": 0.0155, + "step": 12300 + }, + { + "epoch": 7.732412060301508, + "grad_norm": 0.22676153481006622, + "learning_rate": 6.851379323761157e-05, + "loss": 0.015, + "step": 12310 + }, + { + "epoch": 7.738693467336684, + "grad_norm": 0.4766152799129486, + "learning_rate": 6.846258385915545e-05, + "loss": 0.0198, + "step": 12320 + }, + { + "epoch": 7.744974874371859, + "grad_norm": 0.2585076093673706, + "learning_rate": 6.841135204691902e-05, + "loss": 0.0174, + "step": 12330 + }, + { + "epoch": 7.751256281407035, + "grad_norm": 0.371358186006546, + "learning_rate": 6.836009786315377e-05, + "loss": 0.0161, + "step": 12340 + }, + { + "epoch": 7.757537688442211, + "grad_norm": 0.30426084995269775, + "learning_rate": 6.830882137013839e-05, + "loss": 0.0173, + "step": 12350 + }, + { + "epoch": 7.763819095477387, + "grad_norm": 0.24801313877105713, + "learning_rate": 6.825752263017863e-05, + "loss": 0.0174, + "step": 12360 + }, + { + "epoch": 7.7701005025125625, + "grad_norm": 0.23885278403759003, + "learning_rate": 6.820620170560731e-05, + "loss": 0.0169, + "step": 12370 + }, + { + "epoch": 7.776381909547739, + "grad_norm": 0.24543572962284088, + "learning_rate": 6.815485865878418e-05, + "loss": 0.0163, + "step": 12380 + }, + { + "epoch": 7.782663316582915, + "grad_norm": 0.2804911136627197, + "learning_rate": 6.810349355209588e-05, + "loss": 0.0197, + "step": 12390 + }, + { + "epoch": 7.788944723618091, + "grad_norm": 0.23293448984622955, + "learning_rate": 6.805210644795588e-05, + "loss": 0.0186, + "step": 12400 + }, + { + "epoch": 7.795226130653266, + "grad_norm": 0.3576602339744568, + "learning_rate": 6.800069740880433e-05, + "loss": 0.0224, + "step": 12410 + }, + { + "epoch": 7.801507537688442, + "grad_norm": 0.4356251060962677, + "learning_rate": 6.794926649710807e-05, + "loss": 0.0194, + "step": 12420 + }, + { + "epoch": 7.807788944723618, + "grad_norm": 0.3857504427433014, + "learning_rate": 6.789781377536052e-05, + "loss": 0.0203, + "step": 12430 + }, + { + "epoch": 7.814070351758794, + "grad_norm": 0.1841343641281128, + "learning_rate": 6.784633930608158e-05, + "loss": 0.0179, + "step": 12440 + }, + { + "epoch": 7.8203517587939695, + "grad_norm": 0.21040426194667816, + "learning_rate": 6.779484315181759e-05, + "loss": 0.0168, + "step": 12450 + }, + { + "epoch": 7.826633165829145, + "grad_norm": 0.26891517639160156, + "learning_rate": 6.774332537514122e-05, + "loss": 0.0172, + "step": 12460 + }, + { + "epoch": 7.832914572864322, + "grad_norm": 0.2441043108701706, + "learning_rate": 6.769178603865143e-05, + "loss": 0.0212, + "step": 12470 + }, + { + "epoch": 7.839195979899498, + "grad_norm": 0.32549262046813965, + "learning_rate": 6.764022520497337e-05, + "loss": 0.0182, + "step": 12480 + }, + { + "epoch": 7.8454773869346734, + "grad_norm": 0.23284734785556793, + "learning_rate": 6.758864293675833e-05, + "loss": 0.0176, + "step": 12490 + }, + { + "epoch": 7.851758793969849, + "grad_norm": 0.29312676191329956, + "learning_rate": 6.753703929668362e-05, + "loss": 0.018, + "step": 12500 + }, + { + "epoch": 7.858040201005025, + "grad_norm": 0.20800378918647766, + "learning_rate": 6.748541434745253e-05, + "loss": 0.0189, + "step": 12510 + }, + { + "epoch": 7.864321608040201, + "grad_norm": 0.28805795311927795, + "learning_rate": 6.743376815179424e-05, + "loss": 0.0176, + "step": 12520 + }, + { + "epoch": 7.8706030150753765, + "grad_norm": 0.2611129581928253, + "learning_rate": 6.738210077246376e-05, + "loss": 0.0163, + "step": 12530 + }, + { + "epoch": 7.876884422110553, + "grad_norm": 0.41834157705307007, + "learning_rate": 6.733041227224181e-05, + "loss": 0.0178, + "step": 12540 + }, + { + "epoch": 7.883165829145729, + "grad_norm": 0.3410183787345886, + "learning_rate": 6.72787027139348e-05, + "loss": 0.0182, + "step": 12550 + }, + { + "epoch": 7.889447236180905, + "grad_norm": 0.29769906401634216, + "learning_rate": 6.72269721603747e-05, + "loss": 0.0205, + "step": 12560 + }, + { + "epoch": 7.8957286432160805, + "grad_norm": 0.3118989169597626, + "learning_rate": 6.717522067441904e-05, + "loss": 0.0146, + "step": 12570 + }, + { + "epoch": 7.902010050251256, + "grad_norm": 0.2518850266933441, + "learning_rate": 6.712344831895074e-05, + "loss": 0.0216, + "step": 12580 + }, + { + "epoch": 7.908291457286432, + "grad_norm": 0.19940434396266937, + "learning_rate": 6.707165515687811e-05, + "loss": 0.0166, + "step": 12590 + }, + { + "epoch": 7.914572864321608, + "grad_norm": 0.23634549975395203, + "learning_rate": 6.70198412511347e-05, + "loss": 0.0151, + "step": 12600 + }, + { + "epoch": 7.9208542713567835, + "grad_norm": 0.24558775126934052, + "learning_rate": 6.69680066646793e-05, + "loss": 0.0206, + "step": 12610 + }, + { + "epoch": 7.927135678391959, + "grad_norm": 0.23629963397979736, + "learning_rate": 6.691615146049583e-05, + "loss": 0.0185, + "step": 12620 + }, + { + "epoch": 7.933417085427136, + "grad_norm": 0.24404899775981903, + "learning_rate": 6.686427570159324e-05, + "loss": 0.0173, + "step": 12630 + }, + { + "epoch": 7.939698492462312, + "grad_norm": 0.34271422028541565, + "learning_rate": 6.681237945100548e-05, + "loss": 0.0152, + "step": 12640 + }, + { + "epoch": 7.9459798994974875, + "grad_norm": 0.1892630159854889, + "learning_rate": 6.676046277179139e-05, + "loss": 0.0173, + "step": 12650 + }, + { + "epoch": 7.952261306532663, + "grad_norm": 0.2416757196187973, + "learning_rate": 6.670852572703462e-05, + "loss": 0.0163, + "step": 12660 + }, + { + "epoch": 7.958542713567839, + "grad_norm": 0.1865210384130478, + "learning_rate": 6.665656837984359e-05, + "loss": 0.0173, + "step": 12670 + }, + { + "epoch": 7.964824120603015, + "grad_norm": 0.25622275471687317, + "learning_rate": 6.660459079335135e-05, + "loss": 0.0173, + "step": 12680 + }, + { + "epoch": 7.971105527638191, + "grad_norm": 0.2639559805393219, + "learning_rate": 6.655259303071558e-05, + "loss": 0.019, + "step": 12690 + }, + { + "epoch": 7.977386934673367, + "grad_norm": 0.3761753439903259, + "learning_rate": 6.650057515511848e-05, + "loss": 0.0211, + "step": 12700 + }, + { + "epoch": 7.983668341708543, + "grad_norm": 0.20850658416748047, + "learning_rate": 6.644853722976667e-05, + "loss": 0.0183, + "step": 12710 + }, + { + "epoch": 7.989949748743719, + "grad_norm": 0.4166220426559448, + "learning_rate": 6.639647931789113e-05, + "loss": 0.0192, + "step": 12720 + }, + { + "epoch": 7.9962311557788945, + "grad_norm": 0.24724051356315613, + "learning_rate": 6.634440148274713e-05, + "loss": 0.0165, + "step": 12730 + }, + { + "epoch": 8.00251256281407, + "grad_norm": 0.446709543466568, + "learning_rate": 6.629230378761415e-05, + "loss": 0.0183, + "step": 12740 + }, + { + "epoch": 8.008793969849247, + "grad_norm": 0.18123802542686462, + "learning_rate": 6.624018629579581e-05, + "loss": 0.0185, + "step": 12750 + }, + { + "epoch": 8.015075376884422, + "grad_norm": 0.3145778477191925, + "learning_rate": 6.618804907061976e-05, + "loss": 0.0168, + "step": 12760 + }, + { + "epoch": 8.021356783919598, + "grad_norm": 0.4132004976272583, + "learning_rate": 6.613589217543766e-05, + "loss": 0.0165, + "step": 12770 + }, + { + "epoch": 8.027638190954773, + "grad_norm": 0.3214345872402191, + "learning_rate": 6.608371567362504e-05, + "loss": 0.0161, + "step": 12780 + }, + { + "epoch": 8.03391959798995, + "grad_norm": 0.22266437113285065, + "learning_rate": 6.60315196285813e-05, + "loss": 0.0193, + "step": 12790 + }, + { + "epoch": 8.040201005025125, + "grad_norm": 0.35485899448394775, + "learning_rate": 6.597930410372954e-05, + "loss": 0.024, + "step": 12800 + }, + { + "epoch": 8.046482412060302, + "grad_norm": 0.16840597987174988, + "learning_rate": 6.592706916251653e-05, + "loss": 0.0186, + "step": 12810 + }, + { + "epoch": 8.052763819095478, + "grad_norm": 0.22216559946537018, + "learning_rate": 6.587481486841267e-05, + "loss": 0.0165, + "step": 12820 + }, + { + "epoch": 8.059045226130653, + "grad_norm": 0.2586454749107361, + "learning_rate": 6.582254128491184e-05, + "loss": 0.016, + "step": 12830 + }, + { + "epoch": 8.06532663316583, + "grad_norm": 0.32044291496276855, + "learning_rate": 6.577024847553139e-05, + "loss": 0.0202, + "step": 12840 + }, + { + "epoch": 8.071608040201005, + "grad_norm": 0.36288705468177795, + "learning_rate": 6.571793650381201e-05, + "loss": 0.0192, + "step": 12850 + }, + { + "epoch": 8.077889447236181, + "grad_norm": 0.20525488257408142, + "learning_rate": 6.566560543331766e-05, + "loss": 0.0148, + "step": 12860 + }, + { + "epoch": 8.084170854271356, + "grad_norm": 0.1952119618654251, + "learning_rate": 6.561325532763554e-05, + "loss": 0.0162, + "step": 12870 + }, + { + "epoch": 8.090452261306533, + "grad_norm": 0.355580598115921, + "learning_rate": 6.556088625037597e-05, + "loss": 0.0229, + "step": 12880 + }, + { + "epoch": 8.096733668341708, + "grad_norm": 0.2285645306110382, + "learning_rate": 6.55084982651723e-05, + "loss": 0.0206, + "step": 12890 + }, + { + "epoch": 8.103015075376884, + "grad_norm": 0.24861761927604675, + "learning_rate": 6.54560914356809e-05, + "loss": 0.0181, + "step": 12900 + }, + { + "epoch": 8.109296482412061, + "grad_norm": 0.286292165517807, + "learning_rate": 6.540366582558097e-05, + "loss": 0.0158, + "step": 12910 + }, + { + "epoch": 8.115577889447236, + "grad_norm": 0.2285054475069046, + "learning_rate": 6.53512214985746e-05, + "loss": 0.0203, + "step": 12920 + }, + { + "epoch": 8.121859296482413, + "grad_norm": 0.29309093952178955, + "learning_rate": 6.529875851838658e-05, + "loss": 0.0192, + "step": 12930 + }, + { + "epoch": 8.128140703517587, + "grad_norm": 0.20366497337818146, + "learning_rate": 6.52462769487644e-05, + "loss": 0.0167, + "step": 12940 + }, + { + "epoch": 8.134422110552764, + "grad_norm": 0.47723641991615295, + "learning_rate": 6.519377685347807e-05, + "loss": 0.0162, + "step": 12950 + }, + { + "epoch": 8.140703517587939, + "grad_norm": 0.3068526089191437, + "learning_rate": 6.514125829632021e-05, + "loss": 0.0193, + "step": 12960 + }, + { + "epoch": 8.146984924623116, + "grad_norm": 0.3344767987728119, + "learning_rate": 6.508872134110577e-05, + "loss": 0.0175, + "step": 12970 + }, + { + "epoch": 8.153266331658292, + "grad_norm": 0.3658784031867981, + "learning_rate": 6.503616605167213e-05, + "loss": 0.0204, + "step": 12980 + }, + { + "epoch": 8.159547738693467, + "grad_norm": 0.3123801648616791, + "learning_rate": 6.498359249187893e-05, + "loss": 0.0196, + "step": 12990 + }, + { + "epoch": 8.165829145728644, + "grad_norm": 0.2797507345676422, + "learning_rate": 6.493100072560799e-05, + "loss": 0.0155, + "step": 13000 + }, + { + "epoch": 8.172110552763819, + "grad_norm": 0.25348323583602905, + "learning_rate": 6.487839081676327e-05, + "loss": 0.0224, + "step": 13010 + }, + { + "epoch": 8.178391959798995, + "grad_norm": 0.2173084318637848, + "learning_rate": 6.482576282927076e-05, + "loss": 0.0184, + "step": 13020 + }, + { + "epoch": 8.18467336683417, + "grad_norm": 0.23611733317375183, + "learning_rate": 6.477311682707845e-05, + "loss": 0.0182, + "step": 13030 + }, + { + "epoch": 8.190954773869347, + "grad_norm": 0.34175732731819153, + "learning_rate": 6.472045287415616e-05, + "loss": 0.0204, + "step": 13040 + }, + { + "epoch": 8.197236180904522, + "grad_norm": 0.2690509557723999, + "learning_rate": 6.466777103449558e-05, + "loss": 0.0151, + "step": 13050 + }, + { + "epoch": 8.203517587939698, + "grad_norm": 0.17922192811965942, + "learning_rate": 6.461507137211012e-05, + "loss": 0.0174, + "step": 13060 + }, + { + "epoch": 8.209798994974875, + "grad_norm": 0.2531132400035858, + "learning_rate": 6.456235395103483e-05, + "loss": 0.0149, + "step": 13070 + }, + { + "epoch": 8.21608040201005, + "grad_norm": 0.2793917655944824, + "learning_rate": 6.450961883532634e-05, + "loss": 0.0146, + "step": 13080 + }, + { + "epoch": 8.222361809045227, + "grad_norm": 0.22266210615634918, + "learning_rate": 6.445686608906283e-05, + "loss": 0.0171, + "step": 13090 + }, + { + "epoch": 8.228643216080402, + "grad_norm": 0.2960972189903259, + "learning_rate": 6.44040957763438e-05, + "loss": 0.0171, + "step": 13100 + }, + { + "epoch": 8.234924623115578, + "grad_norm": 0.21317829191684723, + "learning_rate": 6.435130796129018e-05, + "loss": 0.0185, + "step": 13110 + }, + { + "epoch": 8.241206030150753, + "grad_norm": 0.18180476129055023, + "learning_rate": 6.429850270804416e-05, + "loss": 0.0153, + "step": 13120 + }, + { + "epoch": 8.24748743718593, + "grad_norm": 0.25482526421546936, + "learning_rate": 6.424568008076909e-05, + "loss": 0.0214, + "step": 13130 + }, + { + "epoch": 8.253768844221106, + "grad_norm": 0.3107151687145233, + "learning_rate": 6.419284014364943e-05, + "loss": 0.0169, + "step": 13140 + }, + { + "epoch": 8.260050251256281, + "grad_norm": 0.28002527356147766, + "learning_rate": 6.41399829608907e-05, + "loss": 0.0156, + "step": 13150 + }, + { + "epoch": 8.266331658291458, + "grad_norm": 0.2773038446903229, + "learning_rate": 6.408710859671938e-05, + "loss": 0.0155, + "step": 13160 + }, + { + "epoch": 8.272613065326633, + "grad_norm": 0.29862871766090393, + "learning_rate": 6.403421711538278e-05, + "loss": 0.0146, + "step": 13170 + }, + { + "epoch": 8.27889447236181, + "grad_norm": 0.3554619252681732, + "learning_rate": 6.398130858114903e-05, + "loss": 0.0166, + "step": 13180 + }, + { + "epoch": 8.285175879396984, + "grad_norm": 0.23079128563404083, + "learning_rate": 6.392838305830701e-05, + "loss": 0.0168, + "step": 13190 + }, + { + "epoch": 8.291457286432161, + "grad_norm": 0.2590559720993042, + "learning_rate": 6.387544061116621e-05, + "loss": 0.0188, + "step": 13200 + }, + { + "epoch": 8.297738693467336, + "grad_norm": 0.34655025601387024, + "learning_rate": 6.382248130405671e-05, + "loss": 0.0219, + "step": 13210 + }, + { + "epoch": 8.304020100502512, + "grad_norm": 0.19056223332881927, + "learning_rate": 6.376950520132905e-05, + "loss": 0.0182, + "step": 13220 + }, + { + "epoch": 8.31030150753769, + "grad_norm": 0.35391637682914734, + "learning_rate": 6.371651236735418e-05, + "loss": 0.0192, + "step": 13230 + }, + { + "epoch": 8.316582914572864, + "grad_norm": 0.2248072326183319, + "learning_rate": 6.366350286652341e-05, + "loss": 0.0196, + "step": 13240 + }, + { + "epoch": 8.32286432160804, + "grad_norm": 0.31620514392852783, + "learning_rate": 6.361047676324827e-05, + "loss": 0.0202, + "step": 13250 + }, + { + "epoch": 8.329145728643216, + "grad_norm": 0.21889477968215942, + "learning_rate": 6.355743412196047e-05, + "loss": 0.0169, + "step": 13260 + }, + { + "epoch": 8.335427135678392, + "grad_norm": 0.2730642855167389, + "learning_rate": 6.350437500711184e-05, + "loss": 0.0238, + "step": 13270 + }, + { + "epoch": 8.341708542713567, + "grad_norm": 0.2883596420288086, + "learning_rate": 6.345129948317418e-05, + "loss": 0.0225, + "step": 13280 + }, + { + "epoch": 8.347989949748744, + "grad_norm": 0.41515636444091797, + "learning_rate": 6.33982076146393e-05, + "loss": 0.0196, + "step": 13290 + }, + { + "epoch": 8.35427135678392, + "grad_norm": 0.2918541431427002, + "learning_rate": 6.334509946601879e-05, + "loss": 0.0243, + "step": 13300 + }, + { + "epoch": 8.360552763819095, + "grad_norm": 0.2272183746099472, + "learning_rate": 6.329197510184406e-05, + "loss": 0.0221, + "step": 13310 + }, + { + "epoch": 8.366834170854272, + "grad_norm": 0.30143028497695923, + "learning_rate": 6.323883458666624e-05, + "loss": 0.0149, + "step": 13320 + }, + { + "epoch": 8.373115577889447, + "grad_norm": 0.20580077171325684, + "learning_rate": 6.318567798505605e-05, + "loss": 0.0176, + "step": 13330 + }, + { + "epoch": 8.379396984924623, + "grad_norm": 0.33833497762680054, + "learning_rate": 6.313250536160378e-05, + "loss": 0.0207, + "step": 13340 + }, + { + "epoch": 8.385678391959798, + "grad_norm": 0.3065182566642761, + "learning_rate": 6.307931678091917e-05, + "loss": 0.0169, + "step": 13350 + }, + { + "epoch": 8.391959798994975, + "grad_norm": 0.30797961354255676, + "learning_rate": 6.302611230763138e-05, + "loss": 0.0175, + "step": 13360 + }, + { + "epoch": 8.39824120603015, + "grad_norm": 0.27342402935028076, + "learning_rate": 6.297289200638887e-05, + "loss": 0.0178, + "step": 13370 + }, + { + "epoch": 8.404522613065327, + "grad_norm": 0.31477847695350647, + "learning_rate": 6.29196559418593e-05, + "loss": 0.0143, + "step": 13380 + }, + { + "epoch": 8.410804020100503, + "grad_norm": 0.2790130078792572, + "learning_rate": 6.286640417872951e-05, + "loss": 0.0199, + "step": 13390 + }, + { + "epoch": 8.417085427135678, + "grad_norm": 0.3239259421825409, + "learning_rate": 6.281313678170542e-05, + "loss": 0.0165, + "step": 13400 + }, + { + "epoch": 8.423366834170855, + "grad_norm": 0.2531527876853943, + "learning_rate": 6.275985381551194e-05, + "loss": 0.0165, + "step": 13410 + }, + { + "epoch": 8.42964824120603, + "grad_norm": 0.30716627836227417, + "learning_rate": 6.270655534489292e-05, + "loss": 0.0201, + "step": 13420 + }, + { + "epoch": 8.435929648241206, + "grad_norm": 0.1749267429113388, + "learning_rate": 6.265324143461098e-05, + "loss": 0.0191, + "step": 13430 + }, + { + "epoch": 8.442211055276381, + "grad_norm": 0.268770307302475, + "learning_rate": 6.259991214944758e-05, + "loss": 0.0203, + "step": 13440 + }, + { + "epoch": 8.448492462311558, + "grad_norm": 0.33017265796661377, + "learning_rate": 6.254656755420283e-05, + "loss": 0.02, + "step": 13450 + }, + { + "epoch": 8.454773869346734, + "grad_norm": 0.30846133828163147, + "learning_rate": 6.249320771369545e-05, + "loss": 0.018, + "step": 13460 + }, + { + "epoch": 8.46105527638191, + "grad_norm": 0.2637775242328644, + "learning_rate": 6.243983269276263e-05, + "loss": 0.0193, + "step": 13470 + }, + { + "epoch": 8.467336683417086, + "grad_norm": 0.2691890001296997, + "learning_rate": 6.238644255626012e-05, + "loss": 0.019, + "step": 13480 + }, + { + "epoch": 8.47361809045226, + "grad_norm": 0.3565405607223511, + "learning_rate": 6.233303736906193e-05, + "loss": 0.0168, + "step": 13490 + }, + { + "epoch": 8.479899497487438, + "grad_norm": 0.1629990190267563, + "learning_rate": 6.227961719606039e-05, + "loss": 0.0208, + "step": 13500 + }, + { + "epoch": 8.486180904522612, + "grad_norm": 0.2013728767633438, + "learning_rate": 6.222618210216608e-05, + "loss": 0.0165, + "step": 13510 + }, + { + "epoch": 8.492462311557789, + "grad_norm": 0.30625662207603455, + "learning_rate": 6.217273215230767e-05, + "loss": 0.0181, + "step": 13520 + }, + { + "epoch": 8.498743718592966, + "grad_norm": 0.299230694770813, + "learning_rate": 6.211926741143188e-05, + "loss": 0.0143, + "step": 13530 + }, + { + "epoch": 8.50502512562814, + "grad_norm": 0.6879630088806152, + "learning_rate": 6.20657879445034e-05, + "loss": 0.0182, + "step": 13540 + }, + { + "epoch": 8.511306532663317, + "grad_norm": 0.3015194833278656, + "learning_rate": 6.201229381650485e-05, + "loss": 0.0206, + "step": 13550 + }, + { + "epoch": 8.517587939698492, + "grad_norm": 0.28048887848854065, + "learning_rate": 6.195878509243661e-05, + "loss": 0.0176, + "step": 13560 + }, + { + "epoch": 8.523869346733669, + "grad_norm": 0.3310387134552002, + "learning_rate": 6.190526183731685e-05, + "loss": 0.0191, + "step": 13570 + }, + { + "epoch": 8.530150753768844, + "grad_norm": 0.2682656943798065, + "learning_rate": 6.185172411618138e-05, + "loss": 0.0195, + "step": 13580 + }, + { + "epoch": 8.53643216080402, + "grad_norm": 0.18786472082138062, + "learning_rate": 6.179817199408355e-05, + "loss": 0.0188, + "step": 13590 + }, + { + "epoch": 8.542713567839195, + "grad_norm": 0.3311130702495575, + "learning_rate": 6.174460553609426e-05, + "loss": 0.0176, + "step": 13600 + }, + { + "epoch": 8.548994974874372, + "grad_norm": 0.2520885765552521, + "learning_rate": 6.16910248073018e-05, + "loss": 0.0193, + "step": 13610 + }, + { + "epoch": 8.555276381909549, + "grad_norm": 0.2331012785434723, + "learning_rate": 6.16374298728118e-05, + "loss": 0.0181, + "step": 13620 + }, + { + "epoch": 8.561557788944723, + "grad_norm": 0.21677501499652863, + "learning_rate": 6.158382079774716e-05, + "loss": 0.0179, + "step": 13630 + }, + { + "epoch": 8.5678391959799, + "grad_norm": 0.5201489329338074, + "learning_rate": 6.153019764724799e-05, + "loss": 0.0206, + "step": 13640 + }, + { + "epoch": 8.574120603015075, + "grad_norm": 0.3474429249763489, + "learning_rate": 6.147656048647144e-05, + "loss": 0.0188, + "step": 13650 + }, + { + "epoch": 8.580402010050252, + "grad_norm": 0.39877063035964966, + "learning_rate": 6.142290938059173e-05, + "loss": 0.0213, + "step": 13660 + }, + { + "epoch": 8.586683417085426, + "grad_norm": 0.26181560754776, + "learning_rate": 6.136924439480001e-05, + "loss": 0.0165, + "step": 13670 + }, + { + "epoch": 8.592964824120603, + "grad_norm": 0.3182620108127594, + "learning_rate": 6.13155655943043e-05, + "loss": 0.0164, + "step": 13680 + }, + { + "epoch": 8.599246231155778, + "grad_norm": 0.20717507600784302, + "learning_rate": 6.126187304432941e-05, + "loss": 0.0188, + "step": 13690 + }, + { + "epoch": 8.605527638190955, + "grad_norm": 0.3112598955631256, + "learning_rate": 6.120816681011684e-05, + "loss": 0.0198, + "step": 13700 + }, + { + "epoch": 8.611809045226131, + "grad_norm": 0.34198275208473206, + "learning_rate": 6.115444695692474e-05, + "loss": 0.0164, + "step": 13710 + }, + { + "epoch": 8.618090452261306, + "grad_norm": 0.13125090301036835, + "learning_rate": 6.110071355002779e-05, + "loss": 0.0183, + "step": 13720 + }, + { + "epoch": 8.624371859296483, + "grad_norm": 0.35711783170700073, + "learning_rate": 6.104696665471714e-05, + "loss": 0.0167, + "step": 13730 + }, + { + "epoch": 8.630653266331658, + "grad_norm": 0.2152659147977829, + "learning_rate": 6.099320633630036e-05, + "loss": 0.0182, + "step": 13740 + }, + { + "epoch": 8.636934673366834, + "grad_norm": 0.22015109658241272, + "learning_rate": 6.093943266010128e-05, + "loss": 0.0173, + "step": 13750 + }, + { + "epoch": 8.64321608040201, + "grad_norm": 0.18440057337284088, + "learning_rate": 6.088564569146e-05, + "loss": 0.0165, + "step": 13760 + }, + { + "epoch": 8.649497487437186, + "grad_norm": 0.21020926535129547, + "learning_rate": 6.083184549573274e-05, + "loss": 0.0162, + "step": 13770 + }, + { + "epoch": 8.655778894472363, + "grad_norm": 0.18129916489124298, + "learning_rate": 6.077803213829184e-05, + "loss": 0.0166, + "step": 13780 + }, + { + "epoch": 8.662060301507537, + "grad_norm": 0.2515626549720764, + "learning_rate": 6.0724205684525594e-05, + "loss": 0.0166, + "step": 13790 + }, + { + "epoch": 8.668341708542714, + "grad_norm": 0.2354903519153595, + "learning_rate": 6.0670366199838215e-05, + "loss": 0.0175, + "step": 13800 + }, + { + "epoch": 8.674623115577889, + "grad_norm": 0.2202025055885315, + "learning_rate": 6.061651374964974e-05, + "loss": 0.0181, + "step": 13810 + }, + { + "epoch": 8.680904522613066, + "grad_norm": 0.1939549744129181, + "learning_rate": 6.0562648399396004e-05, + "loss": 0.0203, + "step": 13820 + }, + { + "epoch": 8.68718592964824, + "grad_norm": 0.2823108434677124, + "learning_rate": 6.050877021452845e-05, + "loss": 0.0176, + "step": 13830 + }, + { + "epoch": 8.693467336683417, + "grad_norm": 0.26838186383247375, + "learning_rate": 6.04548792605142e-05, + "loss": 0.017, + "step": 13840 + }, + { + "epoch": 8.699748743718594, + "grad_norm": 0.2714647650718689, + "learning_rate": 6.04009756028358e-05, + "loss": 0.0166, + "step": 13850 + }, + { + "epoch": 8.706030150753769, + "grad_norm": 0.22688157856464386, + "learning_rate": 6.0347059306991294e-05, + "loss": 0.0174, + "step": 13860 + }, + { + "epoch": 8.712311557788945, + "grad_norm": 0.2905252277851105, + "learning_rate": 6.029313043849407e-05, + "loss": 0.0175, + "step": 13870 + }, + { + "epoch": 8.71859296482412, + "grad_norm": 0.3968747556209564, + "learning_rate": 6.023918906287279e-05, + "loss": 0.0151, + "step": 13880 + }, + { + "epoch": 8.724874371859297, + "grad_norm": 0.21801552176475525, + "learning_rate": 6.01852352456713e-05, + "loss": 0.0152, + "step": 13890 + }, + { + "epoch": 8.731155778894472, + "grad_norm": 0.25278177857398987, + "learning_rate": 6.013126905244858e-05, + "loss": 0.0195, + "step": 13900 + }, + { + "epoch": 8.737437185929648, + "grad_norm": 0.23640722036361694, + "learning_rate": 6.007729054877864e-05, + "loss": 0.0172, + "step": 13910 + }, + { + "epoch": 8.743718592964823, + "grad_norm": 0.35148942470550537, + "learning_rate": 6.0023299800250466e-05, + "loss": 0.0168, + "step": 13920 + }, + { + "epoch": 8.75, + "grad_norm": 0.2123006135225296, + "learning_rate": 5.9969296872467894e-05, + "loss": 0.0185, + "step": 13930 + }, + { + "epoch": 8.756281407035177, + "grad_norm": 0.25853970646858215, + "learning_rate": 5.9915281831049585e-05, + "loss": 0.0156, + "step": 13940 + }, + { + "epoch": 8.762562814070352, + "grad_norm": 0.2356589138507843, + "learning_rate": 5.98612547416289e-05, + "loss": 0.02, + "step": 13950 + }, + { + "epoch": 8.768844221105528, + "grad_norm": 0.2690274715423584, + "learning_rate": 5.9807215669853855e-05, + "loss": 0.0192, + "step": 13960 + }, + { + "epoch": 8.775125628140703, + "grad_norm": 0.23130066692829132, + "learning_rate": 5.9753164681387e-05, + "loss": 0.0154, + "step": 13970 + }, + { + "epoch": 8.78140703517588, + "grad_norm": 0.1657598316669464, + "learning_rate": 5.969910184190539e-05, + "loss": 0.015, + "step": 13980 + }, + { + "epoch": 8.787688442211055, + "grad_norm": 0.1356191784143448, + "learning_rate": 5.9645027217100477e-05, + "loss": 0.0164, + "step": 13990 + }, + { + "epoch": 8.793969849246231, + "grad_norm": 0.2519312798976898, + "learning_rate": 5.959094087267804e-05, + "loss": 0.0147, + "step": 14000 + }, + { + "epoch": 8.800251256281406, + "grad_norm": 0.33603495359420776, + "learning_rate": 5.9536842874358065e-05, + "loss": 0.0139, + "step": 14010 + }, + { + "epoch": 8.806532663316583, + "grad_norm": 0.4023720324039459, + "learning_rate": 5.9482733287874734e-05, + "loss": 0.0186, + "step": 14020 + }, + { + "epoch": 8.81281407035176, + "grad_norm": 0.30967438220977783, + "learning_rate": 5.942861217897631e-05, + "loss": 0.018, + "step": 14030 + }, + { + "epoch": 8.819095477386934, + "grad_norm": 0.22944334149360657, + "learning_rate": 5.9374479613425e-05, + "loss": 0.0153, + "step": 14040 + }, + { + "epoch": 8.825376884422111, + "grad_norm": 0.47633764147758484, + "learning_rate": 5.932033565699704e-05, + "loss": 0.0156, + "step": 14050 + }, + { + "epoch": 8.831658291457286, + "grad_norm": 0.3859303295612335, + "learning_rate": 5.926618037548237e-05, + "loss": 0.0166, + "step": 14060 + }, + { + "epoch": 8.837939698492463, + "grad_norm": 0.2078687697649002, + "learning_rate": 5.9212013834684824e-05, + "loss": 0.0144, + "step": 14070 + }, + { + "epoch": 8.844221105527637, + "grad_norm": 0.4454312026500702, + "learning_rate": 5.9157836100421824e-05, + "loss": 0.0174, + "step": 14080 + }, + { + "epoch": 8.850502512562814, + "grad_norm": 0.25728827714920044, + "learning_rate": 5.910364723852444e-05, + "loss": 0.0149, + "step": 14090 + }, + { + "epoch": 8.85678391959799, + "grad_norm": 0.21139173209667206, + "learning_rate": 5.904944731483723e-05, + "loss": 0.0141, + "step": 14100 + }, + { + "epoch": 8.863065326633166, + "grad_norm": 0.32752904295921326, + "learning_rate": 5.899523639521825e-05, + "loss": 0.0172, + "step": 14110 + }, + { + "epoch": 8.869346733668342, + "grad_norm": 0.25399863719940186, + "learning_rate": 5.894101454553883e-05, + "loss": 0.0129, + "step": 14120 + }, + { + "epoch": 8.875628140703517, + "grad_norm": 0.2041916698217392, + "learning_rate": 5.888678183168368e-05, + "loss": 0.016, + "step": 14130 + }, + { + "epoch": 8.881909547738694, + "grad_norm": 0.38128793239593506, + "learning_rate": 5.883253831955061e-05, + "loss": 0.019, + "step": 14140 + }, + { + "epoch": 8.888190954773869, + "grad_norm": 0.35627999901771545, + "learning_rate": 5.8778284075050625e-05, + "loss": 0.0209, + "step": 14150 + }, + { + "epoch": 8.894472361809045, + "grad_norm": 0.3577904999256134, + "learning_rate": 5.872401916410777e-05, + "loss": 0.0179, + "step": 14160 + }, + { + "epoch": 8.900753768844222, + "grad_norm": 0.2770218253135681, + "learning_rate": 5.8669743652659014e-05, + "loss": 0.0158, + "step": 14170 + }, + { + "epoch": 8.907035175879397, + "grad_norm": 0.2962396442890167, + "learning_rate": 5.86154576066542e-05, + "loss": 0.0201, + "step": 14180 + }, + { + "epoch": 8.913316582914574, + "grad_norm": 0.2305608093738556, + "learning_rate": 5.856116109205602e-05, + "loss": 0.0165, + "step": 14190 + }, + { + "epoch": 8.919597989949748, + "grad_norm": 0.4492342174053192, + "learning_rate": 5.850685417483983e-05, + "loss": 0.0196, + "step": 14200 + }, + { + "epoch": 8.925879396984925, + "grad_norm": 0.1983441859483719, + "learning_rate": 5.8452536920993684e-05, + "loss": 0.0184, + "step": 14210 + }, + { + "epoch": 8.9321608040201, + "grad_norm": 0.2750817537307739, + "learning_rate": 5.839820939651817e-05, + "loss": 0.0166, + "step": 14220 + }, + { + "epoch": 8.938442211055277, + "grad_norm": 0.4153270721435547, + "learning_rate": 5.834387166742632e-05, + "loss": 0.0191, + "step": 14230 + }, + { + "epoch": 8.944723618090451, + "grad_norm": 0.21883109211921692, + "learning_rate": 5.828952379974364e-05, + "loss": 0.0171, + "step": 14240 + }, + { + "epoch": 8.951005025125628, + "grad_norm": 0.20654234290122986, + "learning_rate": 5.8235165859507864e-05, + "loss": 0.0154, + "step": 14250 + }, + { + "epoch": 8.957286432160805, + "grad_norm": 0.2649286389350891, + "learning_rate": 5.818079791276907e-05, + "loss": 0.0186, + "step": 14260 + }, + { + "epoch": 8.96356783919598, + "grad_norm": 0.18446868658065796, + "learning_rate": 5.8126420025589415e-05, + "loss": 0.0175, + "step": 14270 + }, + { + "epoch": 8.969849246231156, + "grad_norm": 0.17436644434928894, + "learning_rate": 5.807203226404313e-05, + "loss": 0.0129, + "step": 14280 + }, + { + "epoch": 8.976130653266331, + "grad_norm": 0.21565717458724976, + "learning_rate": 5.801763469421652e-05, + "loss": 0.0131, + "step": 14290 + }, + { + "epoch": 8.982412060301508, + "grad_norm": 0.16914698481559753, + "learning_rate": 5.796322738220774e-05, + "loss": 0.0135, + "step": 14300 + }, + { + "epoch": 8.988693467336683, + "grad_norm": 0.24431075155735016, + "learning_rate": 5.79088103941268e-05, + "loss": 0.0163, + "step": 14310 + }, + { + "epoch": 8.99497487437186, + "grad_norm": 0.3070754408836365, + "learning_rate": 5.785438379609549e-05, + "loss": 0.0154, + "step": 14320 + }, + { + "epoch": 9.001256281407036, + "grad_norm": 0.353528767824173, + "learning_rate": 5.779994765424724e-05, + "loss": 0.0179, + "step": 14330 + }, + { + "epoch": 9.007537688442211, + "grad_norm": 0.23689502477645874, + "learning_rate": 5.7745502034727114e-05, + "loss": 0.0173, + "step": 14340 + }, + { + "epoch": 9.013819095477388, + "grad_norm": 0.26426365971565247, + "learning_rate": 5.769104700369164e-05, + "loss": 0.0199, + "step": 14350 + }, + { + "epoch": 9.020100502512562, + "grad_norm": 0.32898014783859253, + "learning_rate": 5.7636582627308854e-05, + "loss": 0.0152, + "step": 14360 + }, + { + "epoch": 9.02638190954774, + "grad_norm": 0.27470818161964417, + "learning_rate": 5.7582108971758095e-05, + "loss": 0.0213, + "step": 14370 + }, + { + "epoch": 9.032663316582914, + "grad_norm": 0.2951900362968445, + "learning_rate": 5.752762610322999e-05, + "loss": 0.016, + "step": 14380 + }, + { + "epoch": 9.03894472361809, + "grad_norm": 0.20586568117141724, + "learning_rate": 5.747313408792636e-05, + "loss": 0.0193, + "step": 14390 + }, + { + "epoch": 9.045226130653266, + "grad_norm": 0.2276245504617691, + "learning_rate": 5.741863299206014e-05, + "loss": 0.0162, + "step": 14400 + }, + { + "epoch": 9.051507537688442, + "grad_norm": 0.18262580037117004, + "learning_rate": 5.73641228818553e-05, + "loss": 0.0132, + "step": 14410 + }, + { + "epoch": 9.057788944723619, + "grad_norm": 0.2428569346666336, + "learning_rate": 5.730960382354676e-05, + "loss": 0.0181, + "step": 14420 + }, + { + "epoch": 9.064070351758794, + "grad_norm": 0.23663504421710968, + "learning_rate": 5.725507588338035e-05, + "loss": 0.0143, + "step": 14430 + }, + { + "epoch": 9.07035175879397, + "grad_norm": 0.24259936809539795, + "learning_rate": 5.7200539127612604e-05, + "loss": 0.0108, + "step": 14440 + }, + { + "epoch": 9.076633165829145, + "grad_norm": 0.27219951152801514, + "learning_rate": 5.714599362251087e-05, + "loss": 0.0155, + "step": 14450 + }, + { + "epoch": 9.082914572864322, + "grad_norm": 0.1901514083147049, + "learning_rate": 5.709143943435307e-05, + "loss": 0.02, + "step": 14460 + }, + { + "epoch": 9.089195979899497, + "grad_norm": 0.289776474237442, + "learning_rate": 5.7036876629427646e-05, + "loss": 0.017, + "step": 14470 + }, + { + "epoch": 9.095477386934673, + "grad_norm": 0.20963294804096222, + "learning_rate": 5.698230527403361e-05, + "loss": 0.0183, + "step": 14480 + }, + { + "epoch": 9.10175879396985, + "grad_norm": 0.1948523372411728, + "learning_rate": 5.6927725434480274e-05, + "loss": 0.0196, + "step": 14490 + }, + { + "epoch": 9.108040201005025, + "grad_norm": 0.17459623515605927, + "learning_rate": 5.687313717708728e-05, + "loss": 0.0127, + "step": 14500 + }, + { + "epoch": 9.114321608040202, + "grad_norm": 0.13826018571853638, + "learning_rate": 5.681854056818453e-05, + "loss": 0.0127, + "step": 14510 + }, + { + "epoch": 9.120603015075377, + "grad_norm": 0.2622370719909668, + "learning_rate": 5.676393567411205e-05, + "loss": 0.0162, + "step": 14520 + }, + { + "epoch": 9.126884422110553, + "grad_norm": 0.35966232419013977, + "learning_rate": 5.670932256121991e-05, + "loss": 0.0134, + "step": 14530 + }, + { + "epoch": 9.133165829145728, + "grad_norm": 0.2705182433128357, + "learning_rate": 5.6654701295868216e-05, + "loss": 0.0145, + "step": 14540 + }, + { + "epoch": 9.139447236180905, + "grad_norm": 0.16763058304786682, + "learning_rate": 5.660007194442697e-05, + "loss": 0.0173, + "step": 14550 + }, + { + "epoch": 9.14572864321608, + "grad_norm": 0.24407193064689636, + "learning_rate": 5.6545434573275946e-05, + "loss": 0.0198, + "step": 14560 + }, + { + "epoch": 9.152010050251256, + "grad_norm": 0.21189357340335846, + "learning_rate": 5.649078924880472e-05, + "loss": 0.015, + "step": 14570 + }, + { + "epoch": 9.158291457286433, + "grad_norm": 0.26543092727661133, + "learning_rate": 5.643613603741252e-05, + "loss": 0.0163, + "step": 14580 + }, + { + "epoch": 9.164572864321608, + "grad_norm": 0.24147406220436096, + "learning_rate": 5.6381475005508154e-05, + "loss": 0.0183, + "step": 14590 + }, + { + "epoch": 9.170854271356784, + "grad_norm": 0.3922499418258667, + "learning_rate": 5.632680621950992e-05, + "loss": 0.0138, + "step": 14600 + }, + { + "epoch": 9.17713567839196, + "grad_norm": 0.2842576205730438, + "learning_rate": 5.627212974584555e-05, + "loss": 0.0178, + "step": 14610 + }, + { + "epoch": 9.183417085427136, + "grad_norm": 0.24882568418979645, + "learning_rate": 5.62174456509521e-05, + "loss": 0.0184, + "step": 14620 + }, + { + "epoch": 9.18969849246231, + "grad_norm": 0.31998175382614136, + "learning_rate": 5.616275400127594e-05, + "loss": 0.0161, + "step": 14630 + }, + { + "epoch": 9.195979899497488, + "grad_norm": 0.26329073309898376, + "learning_rate": 5.610805486327254e-05, + "loss": 0.0214, + "step": 14640 + }, + { + "epoch": 9.202261306532664, + "grad_norm": 0.1800401657819748, + "learning_rate": 5.6053348303406536e-05, + "loss": 0.0153, + "step": 14650 + }, + { + "epoch": 9.208542713567839, + "grad_norm": 0.20253024995326996, + "learning_rate": 5.599863438815156e-05, + "loss": 0.0142, + "step": 14660 + }, + { + "epoch": 9.214824120603016, + "grad_norm": 0.2836392819881439, + "learning_rate": 5.594391318399017e-05, + "loss": 0.015, + "step": 14670 + }, + { + "epoch": 9.22110552763819, + "grad_norm": 0.2648107707500458, + "learning_rate": 5.588918475741377e-05, + "loss": 0.0137, + "step": 14680 + }, + { + "epoch": 9.227386934673367, + "grad_norm": 0.28511860966682434, + "learning_rate": 5.5834449174922586e-05, + "loss": 0.0139, + "step": 14690 + }, + { + "epoch": 9.233668341708542, + "grad_norm": 0.24188773334026337, + "learning_rate": 5.57797065030255e-05, + "loss": 0.0148, + "step": 14700 + }, + { + "epoch": 9.239949748743719, + "grad_norm": 0.2602974474430084, + "learning_rate": 5.5724956808240016e-05, + "loss": 0.0176, + "step": 14710 + }, + { + "epoch": 9.246231155778894, + "grad_norm": 0.17097103595733643, + "learning_rate": 5.567020015709219e-05, + "loss": 0.0242, + "step": 14720 + }, + { + "epoch": 9.25251256281407, + "grad_norm": 0.4247480630874634, + "learning_rate": 5.561543661611649e-05, + "loss": 0.0191, + "step": 14730 + }, + { + "epoch": 9.258793969849247, + "grad_norm": 0.4170309007167816, + "learning_rate": 5.556066625185583e-05, + "loss": 0.018, + "step": 14740 + }, + { + "epoch": 9.265075376884422, + "grad_norm": 0.195440873503685, + "learning_rate": 5.550588913086131e-05, + "loss": 0.0191, + "step": 14750 + }, + { + "epoch": 9.271356783919598, + "grad_norm": 0.45482468605041504, + "learning_rate": 5.545110531969234e-05, + "loss": 0.0204, + "step": 14760 + }, + { + "epoch": 9.277638190954773, + "grad_norm": 0.1985183209180832, + "learning_rate": 5.539631488491641e-05, + "loss": 0.0187, + "step": 14770 + }, + { + "epoch": 9.28391959798995, + "grad_norm": 0.23929987847805023, + "learning_rate": 5.534151789310904e-05, + "loss": 0.0172, + "step": 14780 + }, + { + "epoch": 9.290201005025125, + "grad_norm": 0.28998780250549316, + "learning_rate": 5.528671441085376e-05, + "loss": 0.0143, + "step": 14790 + }, + { + "epoch": 9.296482412060302, + "grad_norm": 0.21511884033679962, + "learning_rate": 5.523190450474197e-05, + "loss": 0.0168, + "step": 14800 + }, + { + "epoch": 9.302763819095478, + "grad_norm": 0.19581836462020874, + "learning_rate": 5.5177088241372865e-05, + "loss": 0.0131, + "step": 14810 + }, + { + "epoch": 9.309045226130653, + "grad_norm": 0.21417856216430664, + "learning_rate": 5.512226568735338e-05, + "loss": 0.0155, + "step": 14820 + }, + { + "epoch": 9.31532663316583, + "grad_norm": 0.2455228716135025, + "learning_rate": 5.506743690929809e-05, + "loss": 0.0141, + "step": 14830 + }, + { + "epoch": 9.321608040201005, + "grad_norm": 0.1986345648765564, + "learning_rate": 5.501260197382912e-05, + "loss": 0.0141, + "step": 14840 + }, + { + "epoch": 9.327889447236181, + "grad_norm": 0.2137695699930191, + "learning_rate": 5.49577609475761e-05, + "loss": 0.0139, + "step": 14850 + }, + { + "epoch": 9.334170854271356, + "grad_norm": 0.25377893447875977, + "learning_rate": 5.490291389717603e-05, + "loss": 0.0142, + "step": 14860 + }, + { + "epoch": 9.340452261306533, + "grad_norm": 0.23331530392169952, + "learning_rate": 5.484806088927329e-05, + "loss": 0.0136, + "step": 14870 + }, + { + "epoch": 9.346733668341708, + "grad_norm": 0.23277547955513, + "learning_rate": 5.479320199051942e-05, + "loss": 0.0141, + "step": 14880 + }, + { + "epoch": 9.353015075376884, + "grad_norm": 0.2091362476348877, + "learning_rate": 5.473833726757314e-05, + "loss": 0.013, + "step": 14890 + }, + { + "epoch": 9.359296482412061, + "grad_norm": 0.33981287479400635, + "learning_rate": 5.4683466787100304e-05, + "loss": 0.0156, + "step": 14900 + }, + { + "epoch": 9.365577889447236, + "grad_norm": 0.19682225584983826, + "learning_rate": 5.462859061577369e-05, + "loss": 0.0111, + "step": 14910 + }, + { + "epoch": 9.371859296482413, + "grad_norm": 0.29577234387397766, + "learning_rate": 5.4573708820273026e-05, + "loss": 0.0143, + "step": 14920 + }, + { + "epoch": 9.378140703517587, + "grad_norm": 0.16814082860946655, + "learning_rate": 5.451882146728488e-05, + "loss": 0.0204, + "step": 14930 + }, + { + "epoch": 9.384422110552764, + "grad_norm": 0.20980046689510345, + "learning_rate": 5.446392862350255e-05, + "loss": 0.0132, + "step": 14940 + }, + { + "epoch": 9.390703517587939, + "grad_norm": 0.18771019577980042, + "learning_rate": 5.440903035562603e-05, + "loss": 0.0178, + "step": 14950 + }, + { + "epoch": 9.396984924623116, + "grad_norm": 0.23696446418762207, + "learning_rate": 5.435412673036188e-05, + "loss": 0.0139, + "step": 14960 + }, + { + "epoch": 9.403266331658292, + "grad_norm": 0.1560058295726776, + "learning_rate": 5.429921781442318e-05, + "loss": 0.0119, + "step": 14970 + }, + { + "epoch": 9.409547738693467, + "grad_norm": 0.24617986381053925, + "learning_rate": 5.424430367452945e-05, + "loss": 0.0146, + "step": 14980 + }, + { + "epoch": 9.415829145728644, + "grad_norm": 0.2906186878681183, + "learning_rate": 5.418938437740655e-05, + "loss": 0.0134, + "step": 14990 + }, + { + "epoch": 9.422110552763819, + "grad_norm": 0.2038620561361313, + "learning_rate": 5.413445998978658e-05, + "loss": 0.0153, + "step": 15000 + }, + { + "epoch": 9.428391959798995, + "grad_norm": 0.34731271862983704, + "learning_rate": 5.407953057840789e-05, + "loss": 0.015, + "step": 15010 + }, + { + "epoch": 9.43467336683417, + "grad_norm": 0.22504855692386627, + "learning_rate": 5.4024596210014853e-05, + "loss": 0.0139, + "step": 15020 + }, + { + "epoch": 9.440954773869347, + "grad_norm": 0.22500762343406677, + "learning_rate": 5.3969656951357935e-05, + "loss": 0.0116, + "step": 15030 + }, + { + "epoch": 9.447236180904522, + "grad_norm": 0.29427045583724976, + "learning_rate": 5.3914712869193504e-05, + "loss": 0.016, + "step": 15040 + }, + { + "epoch": 9.453517587939698, + "grad_norm": 0.259273499250412, + "learning_rate": 5.385976403028381e-05, + "loss": 0.0137, + "step": 15050 + }, + { + "epoch": 9.459798994974875, + "grad_norm": 0.1440667062997818, + "learning_rate": 5.3804810501396875e-05, + "loss": 0.0127, + "step": 15060 + }, + { + "epoch": 9.46608040201005, + "grad_norm": 0.17854620516300201, + "learning_rate": 5.374985234930639e-05, + "loss": 0.0122, + "step": 15070 + }, + { + "epoch": 9.472361809045227, + "grad_norm": 0.3640791177749634, + "learning_rate": 5.3694889640791724e-05, + "loss": 0.0179, + "step": 15080 + }, + { + "epoch": 9.478643216080402, + "grad_norm": 0.28167083859443665, + "learning_rate": 5.3639922442637736e-05, + "loss": 0.0124, + "step": 15090 + }, + { + "epoch": 9.484924623115578, + "grad_norm": 0.12923525273799896, + "learning_rate": 5.358495082163476e-05, + "loss": 0.0135, + "step": 15100 + }, + { + "epoch": 9.491206030150753, + "grad_norm": 0.2619512677192688, + "learning_rate": 5.35299748445785e-05, + "loss": 0.015, + "step": 15110 + }, + { + "epoch": 9.49748743718593, + "grad_norm": 0.2646692991256714, + "learning_rate": 5.3474994578269944e-05, + "loss": 0.0195, + "step": 15120 + }, + { + "epoch": 9.503768844221106, + "grad_norm": 0.24682371318340302, + "learning_rate": 5.34200100895153e-05, + "loss": 0.0167, + "step": 15130 + }, + { + "epoch": 9.510050251256281, + "grad_norm": 0.12578897178173065, + "learning_rate": 5.3365021445125916e-05, + "loss": 0.0151, + "step": 15140 + }, + { + "epoch": 9.516331658291458, + "grad_norm": 0.17654597759246826, + "learning_rate": 5.331002871191817e-05, + "loss": 0.0144, + "step": 15150 + }, + { + "epoch": 9.522613065326633, + "grad_norm": 0.28274208307266235, + "learning_rate": 5.3255031956713443e-05, + "loss": 0.0134, + "step": 15160 + }, + { + "epoch": 9.52889447236181, + "grad_norm": 0.30873820185661316, + "learning_rate": 5.320003124633795e-05, + "loss": 0.0122, + "step": 15170 + }, + { + "epoch": 9.535175879396984, + "grad_norm": 0.24974800646305084, + "learning_rate": 5.314502664762274e-05, + "loss": 0.0142, + "step": 15180 + }, + { + "epoch": 9.541457286432161, + "grad_norm": 0.33158090710639954, + "learning_rate": 5.3090018227403604e-05, + "loss": 0.0154, + "step": 15190 + }, + { + "epoch": 9.547738693467338, + "grad_norm": 0.25835588574409485, + "learning_rate": 5.3035006052520955e-05, + "loss": 0.0144, + "step": 15200 + }, + { + "epoch": 9.554020100502512, + "grad_norm": 0.22452090680599213, + "learning_rate": 5.297999018981976e-05, + "loss": 0.0172, + "step": 15210 + }, + { + "epoch": 9.56030150753769, + "grad_norm": 0.23452043533325195, + "learning_rate": 5.2924970706149505e-05, + "loss": 0.0142, + "step": 15220 + }, + { + "epoch": 9.566582914572864, + "grad_norm": 0.34226396679878235, + "learning_rate": 5.2869947668364014e-05, + "loss": 0.0128, + "step": 15230 + }, + { + "epoch": 9.57286432160804, + "grad_norm": 0.284004807472229, + "learning_rate": 5.2814921143321504e-05, + "loss": 0.0171, + "step": 15240 + }, + { + "epoch": 9.579145728643216, + "grad_norm": 0.30896854400634766, + "learning_rate": 5.275989119788436e-05, + "loss": 0.0186, + "step": 15250 + }, + { + "epoch": 9.585427135678392, + "grad_norm": 0.20259976387023926, + "learning_rate": 5.270485789891919e-05, + "loss": 0.0166, + "step": 15260 + }, + { + "epoch": 9.591708542713567, + "grad_norm": 0.2767786383628845, + "learning_rate": 5.264982131329661e-05, + "loss": 0.0144, + "step": 15270 + }, + { + "epoch": 9.597989949748744, + "grad_norm": 0.222272127866745, + "learning_rate": 5.259478150789128e-05, + "loss": 0.0153, + "step": 15280 + }, + { + "epoch": 9.60427135678392, + "grad_norm": 0.2007196694612503, + "learning_rate": 5.253973854958173e-05, + "loss": 0.016, + "step": 15290 + }, + { + "epoch": 9.610552763819095, + "grad_norm": 0.16782639920711517, + "learning_rate": 5.2484692505250376e-05, + "loss": 0.0171, + "step": 15300 + }, + { + "epoch": 9.616834170854272, + "grad_norm": 0.23196326196193695, + "learning_rate": 5.242964344178333e-05, + "loss": 0.0132, + "step": 15310 + }, + { + "epoch": 9.623115577889447, + "grad_norm": 0.337106853723526, + "learning_rate": 5.237459142607041e-05, + "loss": 0.017, + "step": 15320 + }, + { + "epoch": 9.629396984924623, + "grad_norm": 0.2502707839012146, + "learning_rate": 5.231953652500498e-05, + "loss": 0.0194, + "step": 15330 + }, + { + "epoch": 9.635678391959798, + "grad_norm": 0.413240909576416, + "learning_rate": 5.2264478805483976e-05, + "loss": 0.0171, + "step": 15340 + }, + { + "epoch": 9.641959798994975, + "grad_norm": 0.32475629448890686, + "learning_rate": 5.220941833440768e-05, + "loss": 0.0132, + "step": 15350 + }, + { + "epoch": 9.64824120603015, + "grad_norm": 0.2918689548969269, + "learning_rate": 5.215435517867977e-05, + "loss": 0.0139, + "step": 15360 + }, + { + "epoch": 9.654522613065327, + "grad_norm": 0.19334101676940918, + "learning_rate": 5.209928940520719e-05, + "loss": 0.0138, + "step": 15370 + }, + { + "epoch": 9.660804020100503, + "grad_norm": 0.3754654824733734, + "learning_rate": 5.204422108090004e-05, + "loss": 0.0141, + "step": 15380 + }, + { + "epoch": 9.667085427135678, + "grad_norm": 0.3143870532512665, + "learning_rate": 5.19891502726715e-05, + "loss": 0.0167, + "step": 15390 + }, + { + "epoch": 9.673366834170855, + "grad_norm": 0.3357028067111969, + "learning_rate": 5.193407704743782e-05, + "loss": 0.0204, + "step": 15400 + }, + { + "epoch": 9.67964824120603, + "grad_norm": 0.3087024390697479, + "learning_rate": 5.187900147211815e-05, + "loss": 0.0191, + "step": 15410 + }, + { + "epoch": 9.685929648241206, + "grad_norm": 0.20117010176181793, + "learning_rate": 5.182392361363453e-05, + "loss": 0.0142, + "step": 15420 + }, + { + "epoch": 9.692211055276381, + "grad_norm": 0.28502118587493896, + "learning_rate": 5.176884353891172e-05, + "loss": 0.0155, + "step": 15430 + }, + { + "epoch": 9.698492462311558, + "grad_norm": 0.29075515270233154, + "learning_rate": 5.171376131487722e-05, + "loss": 0.0138, + "step": 15440 + }, + { + "epoch": 9.704773869346734, + "grad_norm": 0.2034284621477127, + "learning_rate": 5.165867700846113e-05, + "loss": 0.0154, + "step": 15450 + }, + { + "epoch": 9.71105527638191, + "grad_norm": 0.2906748354434967, + "learning_rate": 5.160359068659606e-05, + "loss": 0.0148, + "step": 15460 + }, + { + "epoch": 9.717336683417086, + "grad_norm": 0.13504572212696075, + "learning_rate": 5.154850241621712e-05, + "loss": 0.0133, + "step": 15470 + }, + { + "epoch": 9.72361809045226, + "grad_norm": 0.2988359332084656, + "learning_rate": 5.149341226426172e-05, + "loss": 0.0177, + "step": 15480 + }, + { + "epoch": 9.729899497487438, + "grad_norm": 0.3093462586402893, + "learning_rate": 5.14383202976696e-05, + "loss": 0.0129, + "step": 15490 + }, + { + "epoch": 9.736180904522612, + "grad_norm": 0.11383581906557083, + "learning_rate": 5.1383226583382684e-05, + "loss": 0.0152, + "step": 15500 + }, + { + "epoch": 9.742462311557789, + "grad_norm": 0.24412290751934052, + "learning_rate": 5.132813118834504e-05, + "loss": 0.014, + "step": 15510 + }, + { + "epoch": 9.748743718592966, + "grad_norm": 0.17528676986694336, + "learning_rate": 5.1273034179502777e-05, + "loss": 0.0123, + "step": 15520 + }, + { + "epoch": 9.75502512562814, + "grad_norm": 0.28518256545066833, + "learning_rate": 5.1217935623803945e-05, + "loss": 0.0151, + "step": 15530 + }, + { + "epoch": 9.761306532663317, + "grad_norm": 0.21241484582424164, + "learning_rate": 5.1162835588198476e-05, + "loss": 0.0184, + "step": 15540 + }, + { + "epoch": 9.767587939698492, + "grad_norm": 0.37531358003616333, + "learning_rate": 5.110773413963813e-05, + "loss": 0.0176, + "step": 15550 + }, + { + "epoch": 9.773869346733669, + "grad_norm": 0.17525920271873474, + "learning_rate": 5.105263134507636e-05, + "loss": 0.0157, + "step": 15560 + }, + { + "epoch": 9.780150753768844, + "grad_norm": 0.2328646332025528, + "learning_rate": 5.099752727146824e-05, + "loss": 0.0129, + "step": 15570 + }, + { + "epoch": 9.78643216080402, + "grad_norm": 0.16415320336818695, + "learning_rate": 5.094242198577042e-05, + "loss": 0.012, + "step": 15580 + }, + { + "epoch": 9.792713567839195, + "grad_norm": 0.24605071544647217, + "learning_rate": 5.088731555494102e-05, + "loss": 0.0131, + "step": 15590 + }, + { + "epoch": 9.798994974874372, + "grad_norm": 0.1994076818227768, + "learning_rate": 5.0832208045939556e-05, + "loss": 0.015, + "step": 15600 + }, + { + "epoch": 9.805276381909549, + "grad_norm": 0.19578798115253448, + "learning_rate": 5.0777099525726844e-05, + "loss": 0.0128, + "step": 15610 + }, + { + "epoch": 9.811557788944723, + "grad_norm": 0.353898286819458, + "learning_rate": 5.072199006126494e-05, + "loss": 0.0137, + "step": 15620 + }, + { + "epoch": 9.8178391959799, + "grad_norm": 0.21526645123958588, + "learning_rate": 5.0666879719517026e-05, + "loss": 0.0127, + "step": 15630 + }, + { + "epoch": 9.824120603015075, + "grad_norm": 0.2506767511367798, + "learning_rate": 5.061176856744737e-05, + "loss": 0.0128, + "step": 15640 + }, + { + "epoch": 9.830402010050252, + "grad_norm": 0.20809641480445862, + "learning_rate": 5.0556656672021205e-05, + "loss": 0.0169, + "step": 15650 + }, + { + "epoch": 9.836683417085426, + "grad_norm": 0.31773126125335693, + "learning_rate": 5.050154410020472e-05, + "loss": 0.0157, + "step": 15660 + }, + { + "epoch": 9.842964824120603, + "grad_norm": 0.18626925349235535, + "learning_rate": 5.044643091896485e-05, + "loss": 0.0153, + "step": 15670 + }, + { + "epoch": 9.849246231155778, + "grad_norm": 0.3519774377346039, + "learning_rate": 5.039131719526932e-05, + "loss": 0.0121, + "step": 15680 + }, + { + "epoch": 9.855527638190955, + "grad_norm": 0.2627323269844055, + "learning_rate": 5.03362029960865e-05, + "loss": 0.0103, + "step": 15690 + }, + { + "epoch": 9.861809045226131, + "grad_norm": 0.22583328187465668, + "learning_rate": 5.028108838838533e-05, + "loss": 0.0135, + "step": 15700 + }, + { + "epoch": 9.868090452261306, + "grad_norm": 0.3446460962295532, + "learning_rate": 5.0225973439135276e-05, + "loss": 0.0142, + "step": 15710 + }, + { + "epoch": 9.874371859296483, + "grad_norm": 0.20017264783382416, + "learning_rate": 5.017085821530617e-05, + "loss": 0.0144, + "step": 15720 + }, + { + "epoch": 9.880653266331658, + "grad_norm": 0.12551522254943848, + "learning_rate": 5.011574278386822e-05, + "loss": 0.0127, + "step": 15730 + }, + { + "epoch": 9.886934673366834, + "grad_norm": 0.3098171651363373, + "learning_rate": 5.006062721179189e-05, + "loss": 0.0123, + "step": 15740 + }, + { + "epoch": 9.89321608040201, + "grad_norm": 0.27495813369750977, + "learning_rate": 5.000551156604777e-05, + "loss": 0.0152, + "step": 15750 + }, + { + "epoch": 9.899497487437186, + "grad_norm": 0.17469124495983124, + "learning_rate": 4.9950395913606594e-05, + "loss": 0.0149, + "step": 15760 + }, + { + "epoch": 9.905778894472363, + "grad_norm": 0.2193002551794052, + "learning_rate": 4.989528032143903e-05, + "loss": 0.0146, + "step": 15770 + }, + { + "epoch": 9.912060301507537, + "grad_norm": 0.2049970179796219, + "learning_rate": 4.984016485651578e-05, + "loss": 0.0148, + "step": 15780 + }, + { + "epoch": 9.918341708542714, + "grad_norm": 0.24872785806655884, + "learning_rate": 4.9785049585807274e-05, + "loss": 0.0155, + "step": 15790 + }, + { + "epoch": 9.924623115577889, + "grad_norm": 0.26323509216308594, + "learning_rate": 4.972993457628381e-05, + "loss": 0.0161, + "step": 15800 + }, + { + "epoch": 9.930904522613066, + "grad_norm": 0.1946573257446289, + "learning_rate": 4.9674819894915306e-05, + "loss": 0.0141, + "step": 15810 + }, + { + "epoch": 9.93718592964824, + "grad_norm": 0.256229043006897, + "learning_rate": 4.9619705608671265e-05, + "loss": 0.0147, + "step": 15820 + }, + { + "epoch": 9.943467336683417, + "grad_norm": 0.3065662086009979, + "learning_rate": 4.956459178452079e-05, + "loss": 0.0157, + "step": 15830 + }, + { + "epoch": 9.949748743718594, + "grad_norm": 0.31413188576698303, + "learning_rate": 4.950947848943235e-05, + "loss": 0.0139, + "step": 15840 + }, + { + "epoch": 9.956030150753769, + "grad_norm": 0.45242130756378174, + "learning_rate": 4.9454365790373805e-05, + "loss": 0.0174, + "step": 15850 + }, + { + "epoch": 9.962311557788945, + "grad_norm": 0.18546338379383087, + "learning_rate": 4.939925375431226e-05, + "loss": 0.0133, + "step": 15860 + }, + { + "epoch": 9.96859296482412, + "grad_norm": 0.29368874430656433, + "learning_rate": 4.9344142448214046e-05, + "loss": 0.014, + "step": 15870 + }, + { + "epoch": 9.974874371859297, + "grad_norm": 0.19083301723003387, + "learning_rate": 4.928903193904461e-05, + "loss": 0.0147, + "step": 15880 + }, + { + "epoch": 9.981155778894472, + "grad_norm": 0.2504326403141022, + "learning_rate": 4.923392229376842e-05, + "loss": 0.0189, + "step": 15890 + }, + { + "epoch": 9.987437185929648, + "grad_norm": 0.25570568442344666, + "learning_rate": 4.9178813579348865e-05, + "loss": 0.0125, + "step": 15900 + }, + { + "epoch": 9.993718592964823, + "grad_norm": 0.17515204846858978, + "learning_rate": 4.9123705862748253e-05, + "loss": 0.0141, + "step": 15910 + }, + { + "epoch": 10.0, + "grad_norm": 0.5857352018356323, + "learning_rate": 4.9068599210927627e-05, + "loss": 0.0197, + "step": 15920 + }, + { + "epoch": 10.006281407035177, + "grad_norm": 0.3368167579174042, + "learning_rate": 4.901349369084681e-05, + "loss": 0.0154, + "step": 15930 + }, + { + "epoch": 10.012562814070352, + "grad_norm": 0.24018648266792297, + "learning_rate": 4.8958389369464156e-05, + "loss": 0.0239, + "step": 15940 + }, + { + "epoch": 10.018844221105528, + "grad_norm": 0.30967438220977783, + "learning_rate": 4.890328631373666e-05, + "loss": 0.0124, + "step": 15950 + }, + { + "epoch": 10.025125628140703, + "grad_norm": 0.3869783580303192, + "learning_rate": 4.88481845906197e-05, + "loss": 0.0164, + "step": 15960 + }, + { + "epoch": 10.03140703517588, + "grad_norm": 0.2890858054161072, + "learning_rate": 4.879308426706707e-05, + "loss": 0.0184, + "step": 15970 + }, + { + "epoch": 10.037688442211055, + "grad_norm": 0.3029031753540039, + "learning_rate": 4.873798541003084e-05, + "loss": 0.0167, + "step": 15980 + }, + { + "epoch": 10.043969849246231, + "grad_norm": 0.3194185495376587, + "learning_rate": 4.868288808646136e-05, + "loss": 0.0143, + "step": 15990 + }, + { + "epoch": 10.050251256281408, + "grad_norm": 0.22216373682022095, + "learning_rate": 4.862779236330705e-05, + "loss": 0.0122, + "step": 16000 + }, + { + "epoch": 10.056532663316583, + "grad_norm": 0.20218525826931, + "learning_rate": 4.8572698307514395e-05, + "loss": 0.0149, + "step": 16010 + }, + { + "epoch": 10.06281407035176, + "grad_norm": 0.2076236456632614, + "learning_rate": 4.85176059860279e-05, + "loss": 0.0121, + "step": 16020 + }, + { + "epoch": 10.069095477386934, + "grad_norm": 0.2436683028936386, + "learning_rate": 4.846251546578989e-05, + "loss": 0.0111, + "step": 16030 + }, + { + "epoch": 10.075376884422111, + "grad_norm": 0.24590761959552765, + "learning_rate": 4.840742681374058e-05, + "loss": 0.0141, + "step": 16040 + }, + { + "epoch": 10.081658291457286, + "grad_norm": 0.271528035402298, + "learning_rate": 4.835234009681787e-05, + "loss": 0.0131, + "step": 16050 + }, + { + "epoch": 10.087939698492463, + "grad_norm": 0.24570222198963165, + "learning_rate": 4.8297255381957296e-05, + "loss": 0.0153, + "step": 16060 + }, + { + "epoch": 10.094221105527637, + "grad_norm": 0.16218437254428864, + "learning_rate": 4.824217273609198e-05, + "loss": 0.0117, + "step": 16070 + }, + { + "epoch": 10.100502512562814, + "grad_norm": 0.2344510555267334, + "learning_rate": 4.8187092226152555e-05, + "loss": 0.0124, + "step": 16080 + }, + { + "epoch": 10.10678391959799, + "grad_norm": 0.2891082167625427, + "learning_rate": 4.8132013919067016e-05, + "loss": 0.0174, + "step": 16090 + }, + { + "epoch": 10.113065326633166, + "grad_norm": 0.3675501346588135, + "learning_rate": 4.807693788176071e-05, + "loss": 0.0134, + "step": 16100 + }, + { + "epoch": 10.119346733668342, + "grad_norm": 0.3280535042285919, + "learning_rate": 4.8021864181156214e-05, + "loss": 0.0123, + "step": 16110 + }, + { + "epoch": 10.125628140703517, + "grad_norm": 0.22342172265052795, + "learning_rate": 4.796679288417326e-05, + "loss": 0.0131, + "step": 16120 + }, + { + "epoch": 10.131909547738694, + "grad_norm": 0.3152918517589569, + "learning_rate": 4.791172405772866e-05, + "loss": 0.015, + "step": 16130 + }, + { + "epoch": 10.138190954773869, + "grad_norm": 0.24409478902816772, + "learning_rate": 4.785665776873625e-05, + "loss": 0.0106, + "step": 16140 + }, + { + "epoch": 10.144472361809045, + "grad_norm": 0.2729201316833496, + "learning_rate": 4.7801594084106763e-05, + "loss": 0.0165, + "step": 16150 + }, + { + "epoch": 10.150753768844222, + "grad_norm": 0.27373647689819336, + "learning_rate": 4.774653307074775e-05, + "loss": 0.0132, + "step": 16160 + }, + { + "epoch": 10.157035175879397, + "grad_norm": 0.1799398809671402, + "learning_rate": 4.769147479556355e-05, + "loss": 0.0177, + "step": 16170 + }, + { + "epoch": 10.163316582914574, + "grad_norm": 0.3248099088668823, + "learning_rate": 4.763641932545515e-05, + "loss": 0.015, + "step": 16180 + }, + { + "epoch": 10.169597989949748, + "grad_norm": 0.2744094729423523, + "learning_rate": 4.758136672732013e-05, + "loss": 0.0164, + "step": 16190 + }, + { + "epoch": 10.175879396984925, + "grad_norm": 0.2694583535194397, + "learning_rate": 4.752631706805261e-05, + "loss": 0.0132, + "step": 16200 + }, + { + "epoch": 10.1821608040201, + "grad_norm": 0.2320755273103714, + "learning_rate": 4.7471270414543114e-05, + "loss": 0.0169, + "step": 16210 + }, + { + "epoch": 10.188442211055277, + "grad_norm": 0.14982864260673523, + "learning_rate": 4.741622683367849e-05, + "loss": 0.0114, + "step": 16220 + }, + { + "epoch": 10.194723618090451, + "grad_norm": 0.19483081996440887, + "learning_rate": 4.736118639234191e-05, + "loss": 0.0164, + "step": 16230 + }, + { + "epoch": 10.201005025125628, + "grad_norm": 0.21885180473327637, + "learning_rate": 4.730614915741267e-05, + "loss": 0.0141, + "step": 16240 + }, + { + "epoch": 10.207286432160805, + "grad_norm": 0.1911638379096985, + "learning_rate": 4.7251115195766234e-05, + "loss": 0.0156, + "step": 16250 + }, + { + "epoch": 10.21356783919598, + "grad_norm": 0.33109524846076965, + "learning_rate": 4.719608457427404e-05, + "loss": 0.012, + "step": 16260 + }, + { + "epoch": 10.219849246231156, + "grad_norm": 0.29765263199806213, + "learning_rate": 4.714105735980347e-05, + "loss": 0.0134, + "step": 16270 + }, + { + "epoch": 10.226130653266331, + "grad_norm": 0.2181636542081833, + "learning_rate": 4.70860336192178e-05, + "loss": 0.0105, + "step": 16280 + }, + { + "epoch": 10.232412060301508, + "grad_norm": 0.2638109624385834, + "learning_rate": 4.7031013419376034e-05, + "loss": 0.0129, + "step": 16290 + }, + { + "epoch": 10.238693467336683, + "grad_norm": 0.22152097523212433, + "learning_rate": 4.6975996827132926e-05, + "loss": 0.015, + "step": 16300 + }, + { + "epoch": 10.24497487437186, + "grad_norm": 0.35555994510650635, + "learning_rate": 4.692098390933883e-05, + "loss": 0.0131, + "step": 16310 + }, + { + "epoch": 10.251256281407034, + "grad_norm": 0.17383532226085663, + "learning_rate": 4.6865974732839615e-05, + "loss": 0.015, + "step": 16320 + }, + { + "epoch": 10.257537688442211, + "grad_norm": 0.17386285960674286, + "learning_rate": 4.6810969364476615e-05, + "loss": 0.0134, + "step": 16330 + }, + { + "epoch": 10.263819095477388, + "grad_norm": 0.19390976428985596, + "learning_rate": 4.675596787108653e-05, + "loss": 0.0143, + "step": 16340 + }, + { + "epoch": 10.270100502512562, + "grad_norm": 0.3752560317516327, + "learning_rate": 4.670097031950139e-05, + "loss": 0.013, + "step": 16350 + }, + { + "epoch": 10.27638190954774, + "grad_norm": 0.25860854983329773, + "learning_rate": 4.664597677654838e-05, + "loss": 0.0127, + "step": 16360 + }, + { + "epoch": 10.282663316582914, + "grad_norm": 0.27008378505706787, + "learning_rate": 4.659098730904986e-05, + "loss": 0.0157, + "step": 16370 + }, + { + "epoch": 10.28894472361809, + "grad_norm": 0.31061089038848877, + "learning_rate": 4.6536001983823204e-05, + "loss": 0.0173, + "step": 16380 + }, + { + "epoch": 10.295226130653266, + "grad_norm": 0.14691859483718872, + "learning_rate": 4.648102086768078e-05, + "loss": 0.0158, + "step": 16390 + }, + { + "epoch": 10.301507537688442, + "grad_norm": 0.2882964313030243, + "learning_rate": 4.6426044027429786e-05, + "loss": 0.0143, + "step": 16400 + }, + { + "epoch": 10.307788944723619, + "grad_norm": 0.24709340929985046, + "learning_rate": 4.637107152987234e-05, + "loss": 0.0122, + "step": 16410 + }, + { + "epoch": 10.314070351758794, + "grad_norm": 0.22374431788921356, + "learning_rate": 4.631610344180515e-05, + "loss": 0.0153, + "step": 16420 + }, + { + "epoch": 10.32035175879397, + "grad_norm": 0.23256413638591766, + "learning_rate": 4.626113983001965e-05, + "loss": 0.0105, + "step": 16430 + }, + { + "epoch": 10.326633165829145, + "grad_norm": 0.15445923805236816, + "learning_rate": 4.620618076130182e-05, + "loss": 0.0113, + "step": 16440 + }, + { + "epoch": 10.332914572864322, + "grad_norm": 0.26278945803642273, + "learning_rate": 4.6151226302432074e-05, + "loss": 0.0139, + "step": 16450 + }, + { + "epoch": 10.339195979899497, + "grad_norm": 0.1875658482313156, + "learning_rate": 4.6096276520185304e-05, + "loss": 0.0112, + "step": 16460 + }, + { + "epoch": 10.345477386934673, + "grad_norm": 0.13035817444324493, + "learning_rate": 4.604133148133066e-05, + "loss": 0.0134, + "step": 16470 + }, + { + "epoch": 10.35175879396985, + "grad_norm": 0.18519185483455658, + "learning_rate": 4.598639125263155e-05, + "loss": 0.0125, + "step": 16480 + }, + { + "epoch": 10.358040201005025, + "grad_norm": 0.2187056690454483, + "learning_rate": 4.593145590084553e-05, + "loss": 0.0158, + "step": 16490 + }, + { + "epoch": 10.364321608040202, + "grad_norm": 0.2463843673467636, + "learning_rate": 4.58765254927242e-05, + "loss": 0.0144, + "step": 16500 + }, + { + "epoch": 10.370603015075377, + "grad_norm": 0.2347671389579773, + "learning_rate": 4.582160009501323e-05, + "loss": 0.012, + "step": 16510 + }, + { + "epoch": 10.376884422110553, + "grad_norm": 0.2918078601360321, + "learning_rate": 4.5766679774452143e-05, + "loss": 0.0114, + "step": 16520 + }, + { + "epoch": 10.383165829145728, + "grad_norm": 0.2752187252044678, + "learning_rate": 4.571176459777431e-05, + "loss": 0.0138, + "step": 16530 + }, + { + "epoch": 10.389447236180905, + "grad_norm": 0.27010178565979004, + "learning_rate": 4.5656854631706846e-05, + "loss": 0.0145, + "step": 16540 + }, + { + "epoch": 10.39572864321608, + "grad_norm": 0.2870396077632904, + "learning_rate": 4.560194994297054e-05, + "loss": 0.0126, + "step": 16550 + }, + { + "epoch": 10.402010050251256, + "grad_norm": 0.11835948377847672, + "learning_rate": 4.554705059827974e-05, + "loss": 0.0144, + "step": 16560 + }, + { + "epoch": 10.408291457286433, + "grad_norm": 0.27479061484336853, + "learning_rate": 4.5492156664342365e-05, + "loss": 0.0137, + "step": 16570 + }, + { + "epoch": 10.414572864321608, + "grad_norm": 0.2548033893108368, + "learning_rate": 4.5437268207859695e-05, + "loss": 0.0153, + "step": 16580 + }, + { + "epoch": 10.420854271356784, + "grad_norm": 0.24864423274993896, + "learning_rate": 4.538238529552641e-05, + "loss": 0.0156, + "step": 16590 + }, + { + "epoch": 10.42713567839196, + "grad_norm": 0.26207035779953003, + "learning_rate": 4.5327507994030394e-05, + "loss": 0.0119, + "step": 16600 + }, + { + "epoch": 10.433417085427136, + "grad_norm": 0.23778694868087769, + "learning_rate": 4.527263637005273e-05, + "loss": 0.014, + "step": 16610 + }, + { + "epoch": 10.43969849246231, + "grad_norm": 0.22349542379379272, + "learning_rate": 4.521777049026766e-05, + "loss": 0.013, + "step": 16620 + }, + { + "epoch": 10.445979899497488, + "grad_norm": 0.2654074728488922, + "learning_rate": 4.5162910421342386e-05, + "loss": 0.0154, + "step": 16630 + }, + { + "epoch": 10.452261306532664, + "grad_norm": 0.21729768812656403, + "learning_rate": 4.510805622993706e-05, + "loss": 0.0142, + "step": 16640 + }, + { + "epoch": 10.458542713567839, + "grad_norm": 0.194333016872406, + "learning_rate": 4.505320798270467e-05, + "loss": 0.0121, + "step": 16650 + }, + { + "epoch": 10.464824120603016, + "grad_norm": 0.2895128130912781, + "learning_rate": 4.499836574629105e-05, + "loss": 0.0147, + "step": 16660 + }, + { + "epoch": 10.47110552763819, + "grad_norm": 0.1655578762292862, + "learning_rate": 4.494352958733465e-05, + "loss": 0.0124, + "step": 16670 + }, + { + "epoch": 10.477386934673367, + "grad_norm": 0.3103717863559723, + "learning_rate": 4.488869957246663e-05, + "loss": 0.0162, + "step": 16680 + }, + { + "epoch": 10.483668341708542, + "grad_norm": 0.23004931211471558, + "learning_rate": 4.4833875768310575e-05, + "loss": 0.0114, + "step": 16690 + }, + { + "epoch": 10.489949748743719, + "grad_norm": 0.31581926345825195, + "learning_rate": 4.47790582414826e-05, + "loss": 0.0141, + "step": 16700 + }, + { + "epoch": 10.496231155778894, + "grad_norm": 0.1604623794555664, + "learning_rate": 4.472424705859114e-05, + "loss": 0.0158, + "step": 16710 + }, + { + "epoch": 10.50251256281407, + "grad_norm": 0.1571303904056549, + "learning_rate": 4.466944228623701e-05, + "loss": 0.0104, + "step": 16720 + }, + { + "epoch": 10.508793969849247, + "grad_norm": 0.1666451394557953, + "learning_rate": 4.461464399101312e-05, + "loss": 0.0157, + "step": 16730 + }, + { + "epoch": 10.515075376884422, + "grad_norm": 0.21544195711612701, + "learning_rate": 4.45598522395046e-05, + "loss": 0.0112, + "step": 16740 + }, + { + "epoch": 10.521356783919598, + "grad_norm": 0.20503343641757965, + "learning_rate": 4.4505067098288577e-05, + "loss": 0.0097, + "step": 16750 + }, + { + "epoch": 10.527638190954773, + "grad_norm": 0.328474223613739, + "learning_rate": 4.445028863393417e-05, + "loss": 0.0135, + "step": 16760 + }, + { + "epoch": 10.53391959798995, + "grad_norm": 0.268358051776886, + "learning_rate": 4.439551691300236e-05, + "loss": 0.0136, + "step": 16770 + }, + { + "epoch": 10.540201005025125, + "grad_norm": 0.2174750119447708, + "learning_rate": 4.434075200204599e-05, + "loss": 0.0153, + "step": 16780 + }, + { + "epoch": 10.546482412060302, + "grad_norm": 0.2137508988380432, + "learning_rate": 4.428599396760956e-05, + "loss": 0.0108, + "step": 16790 + }, + { + "epoch": 10.552763819095478, + "grad_norm": 0.18171638250350952, + "learning_rate": 4.423124287622926e-05, + "loss": 0.0126, + "step": 16800 + }, + { + "epoch": 10.559045226130653, + "grad_norm": 0.28647395968437195, + "learning_rate": 4.417649879443282e-05, + "loss": 0.0141, + "step": 16810 + }, + { + "epoch": 10.56532663316583, + "grad_norm": 0.252716600894928, + "learning_rate": 4.4121761788739444e-05, + "loss": 0.014, + "step": 16820 + }, + { + "epoch": 10.571608040201005, + "grad_norm": 0.18739689886569977, + "learning_rate": 4.4067031925659805e-05, + "loss": 0.012, + "step": 16830 + }, + { + "epoch": 10.577889447236181, + "grad_norm": 0.1873965710401535, + "learning_rate": 4.4012309271695816e-05, + "loss": 0.0128, + "step": 16840 + }, + { + "epoch": 10.584170854271356, + "grad_norm": 0.21827663481235504, + "learning_rate": 4.395759389334067e-05, + "loss": 0.0136, + "step": 16850 + }, + { + "epoch": 10.590452261306533, + "grad_norm": 0.19481195509433746, + "learning_rate": 4.3902885857078685e-05, + "loss": 0.0124, + "step": 16860 + }, + { + "epoch": 10.596733668341708, + "grad_norm": 0.25803473591804504, + "learning_rate": 4.384818522938531e-05, + "loss": 0.0118, + "step": 16870 + }, + { + "epoch": 10.603015075376884, + "grad_norm": 0.4732378423213959, + "learning_rate": 4.379349207672695e-05, + "loss": 0.0153, + "step": 16880 + }, + { + "epoch": 10.609296482412061, + "grad_norm": 0.19650083780288696, + "learning_rate": 4.373880646556098e-05, + "loss": 0.0151, + "step": 16890 + }, + { + "epoch": 10.615577889447236, + "grad_norm": 0.23567551374435425, + "learning_rate": 4.3684128462335533e-05, + "loss": 0.0164, + "step": 16900 + }, + { + "epoch": 10.621859296482413, + "grad_norm": 0.23747503757476807, + "learning_rate": 4.362945813348955e-05, + "loss": 0.0198, + "step": 16910 + }, + { + "epoch": 10.628140703517587, + "grad_norm": 0.27163439989089966, + "learning_rate": 4.357479554545262e-05, + "loss": 0.0111, + "step": 16920 + }, + { + "epoch": 10.634422110552764, + "grad_norm": 0.3426504135131836, + "learning_rate": 4.352014076464499e-05, + "loss": 0.0134, + "step": 16930 + }, + { + "epoch": 10.640703517587939, + "grad_norm": 0.22208461165428162, + "learning_rate": 4.346549385747734e-05, + "loss": 0.0123, + "step": 16940 + }, + { + "epoch": 10.646984924623116, + "grad_norm": 0.2744370102882385, + "learning_rate": 4.34108548903508e-05, + "loss": 0.0152, + "step": 16950 + }, + { + "epoch": 10.653266331658292, + "grad_norm": 0.25182920694351196, + "learning_rate": 4.335622392965689e-05, + "loss": 0.0113, + "step": 16960 + }, + { + "epoch": 10.659547738693467, + "grad_norm": 0.2657153308391571, + "learning_rate": 4.330160104177738e-05, + "loss": 0.0145, + "step": 16970 + }, + { + "epoch": 10.665829145728644, + "grad_norm": 0.34340375661849976, + "learning_rate": 4.32469862930842e-05, + "loss": 0.0166, + "step": 16980 + }, + { + "epoch": 10.672110552763819, + "grad_norm": 0.22754482924938202, + "learning_rate": 4.319237974993946e-05, + "loss": 0.0123, + "step": 16990 + }, + { + "epoch": 10.678391959798995, + "grad_norm": 0.23122213780879974, + "learning_rate": 4.3137781478695236e-05, + "loss": 0.0141, + "step": 17000 + }, + { + "epoch": 10.68467336683417, + "grad_norm": 0.16181117296218872, + "learning_rate": 4.308319154569358e-05, + "loss": 0.0157, + "step": 17010 + }, + { + "epoch": 10.690954773869347, + "grad_norm": 0.27296194434165955, + "learning_rate": 4.302861001726641e-05, + "loss": 0.0131, + "step": 17020 + }, + { + "epoch": 10.697236180904522, + "grad_norm": 0.2444194108247757, + "learning_rate": 4.2974036959735415e-05, + "loss": 0.0133, + "step": 17030 + }, + { + "epoch": 10.703517587939698, + "grad_norm": 0.2695755958557129, + "learning_rate": 4.2919472439412034e-05, + "loss": 0.0134, + "step": 17040 + }, + { + "epoch": 10.709798994974875, + "grad_norm": 0.2353193163871765, + "learning_rate": 4.286491652259729e-05, + "loss": 0.0108, + "step": 17050 + }, + { + "epoch": 10.71608040201005, + "grad_norm": 0.2244122326374054, + "learning_rate": 4.281036927558178e-05, + "loss": 0.0119, + "step": 17060 + }, + { + "epoch": 10.722361809045227, + "grad_norm": 0.1454545557498932, + "learning_rate": 4.275583076464552e-05, + "loss": 0.0113, + "step": 17070 + }, + { + "epoch": 10.728643216080402, + "grad_norm": 0.27379727363586426, + "learning_rate": 4.270130105605794e-05, + "loss": 0.017, + "step": 17080 + }, + { + "epoch": 10.734924623115578, + "grad_norm": 0.2714517116546631, + "learning_rate": 4.264678021607782e-05, + "loss": 0.0125, + "step": 17090 + }, + { + "epoch": 10.741206030150753, + "grad_norm": 0.18454299867153168, + "learning_rate": 4.25922683109531e-05, + "loss": 0.0096, + "step": 17100 + }, + { + "epoch": 10.74748743718593, + "grad_norm": 0.1960168331861496, + "learning_rate": 4.25377654069209e-05, + "loss": 0.0139, + "step": 17110 + }, + { + "epoch": 10.753768844221106, + "grad_norm": 0.424572229385376, + "learning_rate": 4.248327157020737e-05, + "loss": 0.0139, + "step": 17120 + }, + { + "epoch": 10.760050251256281, + "grad_norm": 0.286518931388855, + "learning_rate": 4.242878686702763e-05, + "loss": 0.0135, + "step": 17130 + }, + { + "epoch": 10.766331658291458, + "grad_norm": 0.225290447473526, + "learning_rate": 4.23743113635858e-05, + "loss": 0.0139, + "step": 17140 + }, + { + "epoch": 10.772613065326633, + "grad_norm": 0.2067255824804306, + "learning_rate": 4.2319845126074705e-05, + "loss": 0.0154, + "step": 17150 + }, + { + "epoch": 10.77889447236181, + "grad_norm": 0.21000993251800537, + "learning_rate": 4.226538822067598e-05, + "loss": 0.015, + "step": 17160 + }, + { + "epoch": 10.785175879396984, + "grad_norm": 0.2664952576160431, + "learning_rate": 4.2210940713559896e-05, + "loss": 0.013, + "step": 17170 + }, + { + "epoch": 10.791457286432161, + "grad_norm": 0.22616316378116608, + "learning_rate": 4.21565026708853e-05, + "loss": 0.0127, + "step": 17180 + }, + { + "epoch": 10.797738693467338, + "grad_norm": 0.27437278628349304, + "learning_rate": 4.2102074158799535e-05, + "loss": 0.0146, + "step": 17190 + }, + { + "epoch": 10.804020100502512, + "grad_norm": 0.19943615794181824, + "learning_rate": 4.204765524343841e-05, + "loss": 0.0119, + "step": 17200 + }, + { + "epoch": 10.81030150753769, + "grad_norm": 0.28157493472099304, + "learning_rate": 4.1993245990926036e-05, + "loss": 0.0164, + "step": 17210 + }, + { + "epoch": 10.816582914572864, + "grad_norm": 0.2274855673313141, + "learning_rate": 4.193884646737474e-05, + "loss": 0.0127, + "step": 17220 + }, + { + "epoch": 10.82286432160804, + "grad_norm": 0.27968335151672363, + "learning_rate": 4.1884456738885124e-05, + "loss": 0.0144, + "step": 17230 + }, + { + "epoch": 10.829145728643216, + "grad_norm": 0.23505151271820068, + "learning_rate": 4.1830076871545795e-05, + "loss": 0.0131, + "step": 17240 + }, + { + "epoch": 10.835427135678392, + "grad_norm": 0.24641048908233643, + "learning_rate": 4.177570693143346e-05, + "loss": 0.01, + "step": 17250 + }, + { + "epoch": 10.841708542713567, + "grad_norm": 0.24266445636749268, + "learning_rate": 4.172134698461271e-05, + "loss": 0.0173, + "step": 17260 + }, + { + "epoch": 10.847989949748744, + "grad_norm": 0.23284019529819489, + "learning_rate": 4.166699709713599e-05, + "loss": 0.0139, + "step": 17270 + }, + { + "epoch": 10.85427135678392, + "grad_norm": 0.25439947843551636, + "learning_rate": 4.1612657335043555e-05, + "loss": 0.0137, + "step": 17280 + }, + { + "epoch": 10.860552763819095, + "grad_norm": 0.33706286549568176, + "learning_rate": 4.155832776436331e-05, + "loss": 0.0143, + "step": 17290 + }, + { + "epoch": 10.866834170854272, + "grad_norm": 0.22112047672271729, + "learning_rate": 4.1504008451110845e-05, + "loss": 0.012, + "step": 17300 + }, + { + "epoch": 10.873115577889447, + "grad_norm": 0.18763285875320435, + "learning_rate": 4.1449699461289225e-05, + "loss": 0.0142, + "step": 17310 + }, + { + "epoch": 10.879396984924623, + "grad_norm": 0.17629805207252502, + "learning_rate": 4.139540086088901e-05, + "loss": 0.0136, + "step": 17320 + }, + { + "epoch": 10.885678391959798, + "grad_norm": 0.13851767778396606, + "learning_rate": 4.1341112715888106e-05, + "loss": 0.0097, + "step": 17330 + }, + { + "epoch": 10.891959798994975, + "grad_norm": 0.23669514060020447, + "learning_rate": 4.128683509225172e-05, + "loss": 0.0138, + "step": 17340 + }, + { + "epoch": 10.89824120603015, + "grad_norm": 0.1839110255241394, + "learning_rate": 4.123256805593231e-05, + "loss": 0.0136, + "step": 17350 + }, + { + "epoch": 10.904522613065327, + "grad_norm": 0.21461784839630127, + "learning_rate": 4.117831167286943e-05, + "loss": 0.0127, + "step": 17360 + }, + { + "epoch": 10.910804020100503, + "grad_norm": 0.1422787755727768, + "learning_rate": 4.1124066008989685e-05, + "loss": 0.013, + "step": 17370 + }, + { + "epoch": 10.917085427135678, + "grad_norm": 0.1497839242219925, + "learning_rate": 4.106983113020669e-05, + "loss": 0.0115, + "step": 17380 + }, + { + "epoch": 10.923366834170855, + "grad_norm": 0.3088744878768921, + "learning_rate": 4.101560710242094e-05, + "loss": 0.0121, + "step": 17390 + }, + { + "epoch": 10.92964824120603, + "grad_norm": 0.13388369977474213, + "learning_rate": 4.096139399151971e-05, + "loss": 0.0103, + "step": 17400 + }, + { + "epoch": 10.935929648241206, + "grad_norm": 0.442278653383255, + "learning_rate": 4.090719186337709e-05, + "loss": 0.0109, + "step": 17410 + }, + { + "epoch": 10.942211055276381, + "grad_norm": 0.2337704747915268, + "learning_rate": 4.0853000783853746e-05, + "loss": 0.0132, + "step": 17420 + }, + { + "epoch": 10.948492462311558, + "grad_norm": 0.17455917596817017, + "learning_rate": 4.0798820818796956e-05, + "loss": 0.0148, + "step": 17430 + }, + { + "epoch": 10.954773869346734, + "grad_norm": 0.26373812556266785, + "learning_rate": 4.074465203404048e-05, + "loss": 0.0169, + "step": 17440 + }, + { + "epoch": 10.96105527638191, + "grad_norm": 0.14672744274139404, + "learning_rate": 4.0690494495404495e-05, + "loss": 0.0096, + "step": 17450 + }, + { + "epoch": 10.967336683417086, + "grad_norm": 0.1968153566122055, + "learning_rate": 4.0636348268695535e-05, + "loss": 0.0124, + "step": 17460 + }, + { + "epoch": 10.97361809045226, + "grad_norm": 0.2285657525062561, + "learning_rate": 4.058221341970637e-05, + "loss": 0.0123, + "step": 17470 + }, + { + "epoch": 10.979899497487438, + "grad_norm": 0.20792020857334137, + "learning_rate": 4.0528090014215945e-05, + "loss": 0.012, + "step": 17480 + }, + { + "epoch": 10.986180904522612, + "grad_norm": 0.2560802400112152, + "learning_rate": 4.047397811798929e-05, + "loss": 0.0138, + "step": 17490 + }, + { + "epoch": 10.992462311557789, + "grad_norm": 0.2750023603439331, + "learning_rate": 4.041987779677745e-05, + "loss": 0.0147, + "step": 17500 + }, + { + "epoch": 10.998743718592966, + "grad_norm": 0.22804930806159973, + "learning_rate": 4.036578911631745e-05, + "loss": 0.0085, + "step": 17510 + }, + { + "epoch": 11.00502512562814, + "grad_norm": 0.3036993145942688, + "learning_rate": 4.031171214233211e-05, + "loss": 0.0095, + "step": 17520 + }, + { + "epoch": 11.011306532663317, + "grad_norm": 0.24263349175453186, + "learning_rate": 4.025764694053008e-05, + "loss": 0.0135, + "step": 17530 + }, + { + "epoch": 11.017587939698492, + "grad_norm": 0.3206203281879425, + "learning_rate": 4.020359357660566e-05, + "loss": 0.0135, + "step": 17540 + }, + { + "epoch": 11.023869346733669, + "grad_norm": 0.17668874561786652, + "learning_rate": 4.014955211623875e-05, + "loss": 0.0165, + "step": 17550 + }, + { + "epoch": 11.030150753768844, + "grad_norm": 0.19440071284770966, + "learning_rate": 4.0095522625094895e-05, + "loss": 0.0138, + "step": 17560 + }, + { + "epoch": 11.03643216080402, + "grad_norm": 0.3560439348220825, + "learning_rate": 4.004150516882497e-05, + "loss": 0.0136, + "step": 17570 + }, + { + "epoch": 11.042713567839195, + "grad_norm": 0.29321029782295227, + "learning_rate": 3.99874998130653e-05, + "loss": 0.0127, + "step": 17580 + }, + { + "epoch": 11.048994974874372, + "grad_norm": 0.29089564085006714, + "learning_rate": 3.9933506623437455e-05, + "loss": 0.0148, + "step": 17590 + }, + { + "epoch": 11.055276381909549, + "grad_norm": 0.343657910823822, + "learning_rate": 3.9879525665548276e-05, + "loss": 0.0105, + "step": 17600 + }, + { + "epoch": 11.061557788944723, + "grad_norm": 0.2884354591369629, + "learning_rate": 3.9825557004989704e-05, + "loss": 0.0105, + "step": 17610 + }, + { + "epoch": 11.0678391959799, + "grad_norm": 0.2210513800382614, + "learning_rate": 3.977160070733877e-05, + "loss": 0.0118, + "step": 17620 + }, + { + "epoch": 11.074120603015075, + "grad_norm": 0.1846415400505066, + "learning_rate": 3.971765683815746e-05, + "loss": 0.011, + "step": 17630 + }, + { + "epoch": 11.080402010050252, + "grad_norm": 0.27770766615867615, + "learning_rate": 3.966372546299265e-05, + "loss": 0.0115, + "step": 17640 + }, + { + "epoch": 11.086683417085426, + "grad_norm": 0.333791583776474, + "learning_rate": 3.960980664737604e-05, + "loss": 0.014, + "step": 17650 + }, + { + "epoch": 11.092964824120603, + "grad_norm": 0.2174055278301239, + "learning_rate": 3.955590045682408e-05, + "loss": 0.0117, + "step": 17660 + }, + { + "epoch": 11.099246231155778, + "grad_norm": 0.21008925139904022, + "learning_rate": 3.950200695683788e-05, + "loss": 0.0122, + "step": 17670 + }, + { + "epoch": 11.105527638190955, + "grad_norm": 0.3088732063770294, + "learning_rate": 3.944812621290314e-05, + "loss": 0.0138, + "step": 17680 + }, + { + "epoch": 11.111809045226131, + "grad_norm": 0.2408359944820404, + "learning_rate": 3.939425829049002e-05, + "loss": 0.0122, + "step": 17690 + }, + { + "epoch": 11.118090452261306, + "grad_norm": 0.2245197892189026, + "learning_rate": 3.9340403255053135e-05, + "loss": 0.0108, + "step": 17700 + }, + { + "epoch": 11.124371859296483, + "grad_norm": 0.18725447356700897, + "learning_rate": 3.9286561172031407e-05, + "loss": 0.015, + "step": 17710 + }, + { + "epoch": 11.130653266331658, + "grad_norm": 0.12523557245731354, + "learning_rate": 3.923273210684809e-05, + "loss": 0.0114, + "step": 17720 + }, + { + "epoch": 11.136934673366834, + "grad_norm": 0.3275805711746216, + "learning_rate": 3.9178916124910555e-05, + "loss": 0.0132, + "step": 17730 + }, + { + "epoch": 11.14321608040201, + "grad_norm": 0.284804105758667, + "learning_rate": 3.9125113291610274e-05, + "loss": 0.0134, + "step": 17740 + }, + { + "epoch": 11.149497487437186, + "grad_norm": 0.2645522356033325, + "learning_rate": 3.907132367232278e-05, + "loss": 0.0122, + "step": 17750 + }, + { + "epoch": 11.155778894472363, + "grad_norm": 0.2732006907463074, + "learning_rate": 3.901754733240752e-05, + "loss": 0.0097, + "step": 17760 + }, + { + "epoch": 11.162060301507537, + "grad_norm": 0.24562738835811615, + "learning_rate": 3.896378433720785e-05, + "loss": 0.0126, + "step": 17770 + }, + { + "epoch": 11.168341708542714, + "grad_norm": 0.2367866486310959, + "learning_rate": 3.891003475205086e-05, + "loss": 0.0156, + "step": 17780 + }, + { + "epoch": 11.174623115577889, + "grad_norm": 0.24508962035179138, + "learning_rate": 3.885629864224736e-05, + "loss": 0.0122, + "step": 17790 + }, + { + "epoch": 11.180904522613066, + "grad_norm": 0.27607324719429016, + "learning_rate": 3.880257607309178e-05, + "loss": 0.0133, + "step": 17800 + }, + { + "epoch": 11.18718592964824, + "grad_norm": 0.22839023172855377, + "learning_rate": 3.874886710986213e-05, + "loss": 0.0145, + "step": 17810 + }, + { + "epoch": 11.193467336683417, + "grad_norm": 0.2709227204322815, + "learning_rate": 3.8695171817819834e-05, + "loss": 0.0136, + "step": 17820 + }, + { + "epoch": 11.199748743718594, + "grad_norm": 0.2816867232322693, + "learning_rate": 3.8641490262209774e-05, + "loss": 0.0133, + "step": 17830 + }, + { + "epoch": 11.206030150753769, + "grad_norm": 0.2106635868549347, + "learning_rate": 3.8587822508260084e-05, + "loss": 0.0124, + "step": 17840 + }, + { + "epoch": 11.212311557788945, + "grad_norm": 0.318843275308609, + "learning_rate": 3.853416862118214e-05, + "loss": 0.0141, + "step": 17850 + }, + { + "epoch": 11.21859296482412, + "grad_norm": 0.22315669059753418, + "learning_rate": 3.848052866617049e-05, + "loss": 0.0221, + "step": 17860 + }, + { + "epoch": 11.224874371859297, + "grad_norm": 0.414875864982605, + "learning_rate": 3.84269027084027e-05, + "loss": 0.0112, + "step": 17870 + }, + { + "epoch": 11.231155778894472, + "grad_norm": 0.17001605033874512, + "learning_rate": 3.83732908130394e-05, + "loss": 0.011, + "step": 17880 + }, + { + "epoch": 11.237437185929648, + "grad_norm": 0.29663586616516113, + "learning_rate": 3.831969304522409e-05, + "loss": 0.0147, + "step": 17890 + }, + { + "epoch": 11.243718592964823, + "grad_norm": 0.18227216601371765, + "learning_rate": 3.826610947008312e-05, + "loss": 0.0107, + "step": 17900 + }, + { + "epoch": 11.25, + "grad_norm": 0.2670266032218933, + "learning_rate": 3.821254015272559e-05, + "loss": 0.0141, + "step": 17910 + }, + { + "epoch": 11.256281407035177, + "grad_norm": 0.19229631125926971, + "learning_rate": 3.8158985158243214e-05, + "loss": 0.0108, + "step": 17920 + }, + { + "epoch": 11.262562814070352, + "grad_norm": 0.20246490836143494, + "learning_rate": 3.810544455171044e-05, + "loss": 0.0156, + "step": 17930 + }, + { + "epoch": 11.268844221105528, + "grad_norm": 0.32041358947753906, + "learning_rate": 3.8051918398184114e-05, + "loss": 0.0129, + "step": 17940 + }, + { + "epoch": 11.275125628140703, + "grad_norm": 0.2541782855987549, + "learning_rate": 3.799840676270356e-05, + "loss": 0.015, + "step": 17950 + }, + { + "epoch": 11.28140703517588, + "grad_norm": 0.2743317484855652, + "learning_rate": 3.794490971029048e-05, + "loss": 0.0116, + "step": 17960 + }, + { + "epoch": 11.287688442211055, + "grad_norm": 0.31807729601860046, + "learning_rate": 3.789142730594881e-05, + "loss": 0.0148, + "step": 17970 + }, + { + "epoch": 11.293969849246231, + "grad_norm": 0.39163777232170105, + "learning_rate": 3.783795961466471e-05, + "loss": 0.0162, + "step": 17980 + }, + { + "epoch": 11.300251256281406, + "grad_norm": 0.15449953079223633, + "learning_rate": 3.778450670140651e-05, + "loss": 0.0118, + "step": 17990 + }, + { + "epoch": 11.306532663316583, + "grad_norm": 0.2006523609161377, + "learning_rate": 3.773106863112451e-05, + "loss": 0.0123, + "step": 18000 + }, + { + "epoch": 11.31281407035176, + "grad_norm": 0.2470698356628418, + "learning_rate": 3.7677645468751e-05, + "loss": 0.0146, + "step": 18010 + }, + { + "epoch": 11.319095477386934, + "grad_norm": 0.16177426278591156, + "learning_rate": 3.762423727920018e-05, + "loss": 0.0129, + "step": 18020 + }, + { + "epoch": 11.325376884422111, + "grad_norm": 0.29547518491744995, + "learning_rate": 3.7570844127367995e-05, + "loss": 0.0165, + "step": 18030 + }, + { + "epoch": 11.331658291457286, + "grad_norm": 0.23419690132141113, + "learning_rate": 3.751746607813221e-05, + "loss": 0.0164, + "step": 18040 + }, + { + "epoch": 11.337939698492463, + "grad_norm": 0.18356600403785706, + "learning_rate": 3.746410319635217e-05, + "loss": 0.0133, + "step": 18050 + }, + { + "epoch": 11.344221105527637, + "grad_norm": 0.2306535691022873, + "learning_rate": 3.7410755546868806e-05, + "loss": 0.0124, + "step": 18060 + }, + { + "epoch": 11.350502512562814, + "grad_norm": 0.18863151967525482, + "learning_rate": 3.735742319450454e-05, + "loss": 0.0148, + "step": 18070 + }, + { + "epoch": 11.35678391959799, + "grad_norm": 0.1826995611190796, + "learning_rate": 3.7304106204063184e-05, + "loss": 0.0152, + "step": 18080 + }, + { + "epoch": 11.363065326633166, + "grad_norm": 0.2063819169998169, + "learning_rate": 3.725080464032996e-05, + "loss": 0.0119, + "step": 18090 + }, + { + "epoch": 11.369346733668342, + "grad_norm": 0.19362975656986237, + "learning_rate": 3.719751856807125e-05, + "loss": 0.0102, + "step": 18100 + }, + { + "epoch": 11.375628140703517, + "grad_norm": 0.24095307290554047, + "learning_rate": 3.71442480520347e-05, + "loss": 0.0117, + "step": 18110 + }, + { + "epoch": 11.381909547738694, + "grad_norm": 0.15208971500396729, + "learning_rate": 3.709099315694897e-05, + "loss": 0.0122, + "step": 18120 + }, + { + "epoch": 11.388190954773869, + "grad_norm": 0.287173330783844, + "learning_rate": 3.703775394752378e-05, + "loss": 0.0126, + "step": 18130 + }, + { + "epoch": 11.394472361809045, + "grad_norm": 0.19339196383953094, + "learning_rate": 3.698453048844983e-05, + "loss": 0.0102, + "step": 18140 + }, + { + "epoch": 11.400753768844222, + "grad_norm": 0.23538090288639069, + "learning_rate": 3.693132284439861e-05, + "loss": 0.016, + "step": 18150 + }, + { + "epoch": 11.407035175879397, + "grad_norm": 0.3098812699317932, + "learning_rate": 3.6878131080022414e-05, + "loss": 0.0094, + "step": 18160 + }, + { + "epoch": 11.413316582914574, + "grad_norm": 0.2545091509819031, + "learning_rate": 3.682495525995429e-05, + "loss": 0.015, + "step": 18170 + }, + { + "epoch": 11.419597989949748, + "grad_norm": 0.15865261852741241, + "learning_rate": 3.6771795448807846e-05, + "loss": 0.0104, + "step": 18180 + }, + { + "epoch": 11.425879396984925, + "grad_norm": 0.17288118600845337, + "learning_rate": 3.671865171117724e-05, + "loss": 0.0146, + "step": 18190 + }, + { + "epoch": 11.4321608040201, + "grad_norm": 0.25976699590682983, + "learning_rate": 3.666552411163718e-05, + "loss": 0.0123, + "step": 18200 + }, + { + "epoch": 11.438442211055277, + "grad_norm": 0.20643706619739532, + "learning_rate": 3.661241271474269e-05, + "loss": 0.0104, + "step": 18210 + }, + { + "epoch": 11.444723618090451, + "grad_norm": 0.2169375717639923, + "learning_rate": 3.6559317585029116e-05, + "loss": 0.0087, + "step": 18220 + }, + { + "epoch": 11.451005025125628, + "grad_norm": 0.1982221007347107, + "learning_rate": 3.650623878701204e-05, + "loss": 0.0106, + "step": 18230 + }, + { + "epoch": 11.457286432160805, + "grad_norm": 0.1835186630487442, + "learning_rate": 3.645317638518721e-05, + "loss": 0.0135, + "step": 18240 + }, + { + "epoch": 11.46356783919598, + "grad_norm": 0.2344437539577484, + "learning_rate": 3.640013044403046e-05, + "loss": 0.013, + "step": 18250 + }, + { + "epoch": 11.469849246231156, + "grad_norm": 0.1968843638896942, + "learning_rate": 3.634710102799761e-05, + "loss": 0.013, + "step": 18260 + }, + { + "epoch": 11.476130653266331, + "grad_norm": 0.2233797311782837, + "learning_rate": 3.6294088201524394e-05, + "loss": 0.0099, + "step": 18270 + }, + { + "epoch": 11.482412060301508, + "grad_norm": 0.19558505713939667, + "learning_rate": 3.62410920290264e-05, + "loss": 0.0121, + "step": 18280 + }, + { + "epoch": 11.488693467336683, + "grad_norm": 0.13045237958431244, + "learning_rate": 3.6188112574898954e-05, + "loss": 0.0098, + "step": 18290 + }, + { + "epoch": 11.49497487437186, + "grad_norm": 0.2025410681962967, + "learning_rate": 3.6135149903517115e-05, + "loss": 0.012, + "step": 18300 + }, + { + "epoch": 11.501256281407034, + "grad_norm": 0.2796749770641327, + "learning_rate": 3.608220407923552e-05, + "loss": 0.0147, + "step": 18310 + }, + { + "epoch": 11.507537688442211, + "grad_norm": 0.14447376132011414, + "learning_rate": 3.602927516638833e-05, + "loss": 0.0132, + "step": 18320 + }, + { + "epoch": 11.513819095477388, + "grad_norm": 0.15016844868659973, + "learning_rate": 3.5976363229289165e-05, + "loss": 0.0076, + "step": 18330 + }, + { + "epoch": 11.520100502512562, + "grad_norm": 0.16624099016189575, + "learning_rate": 3.5923468332231e-05, + "loss": 0.0129, + "step": 18340 + }, + { + "epoch": 11.52638190954774, + "grad_norm": 0.22910980880260468, + "learning_rate": 3.587059053948616e-05, + "loss": 0.0085, + "step": 18350 + }, + { + "epoch": 11.532663316582914, + "grad_norm": 0.2765357792377472, + "learning_rate": 3.5817729915306134e-05, + "loss": 0.0099, + "step": 18360 + }, + { + "epoch": 11.53894472361809, + "grad_norm": 0.2592473328113556, + "learning_rate": 3.5764886523921566e-05, + "loss": 0.0111, + "step": 18370 + }, + { + "epoch": 11.545226130653266, + "grad_norm": 0.28097957372665405, + "learning_rate": 3.571206042954214e-05, + "loss": 0.0144, + "step": 18380 + }, + { + "epoch": 11.551507537688442, + "grad_norm": 0.14862437546253204, + "learning_rate": 3.5659251696356566e-05, + "loss": 0.0104, + "step": 18390 + }, + { + "epoch": 11.557788944723619, + "grad_norm": 0.14289364218711853, + "learning_rate": 3.56064603885324e-05, + "loss": 0.0105, + "step": 18400 + }, + { + "epoch": 11.564070351758794, + "grad_norm": 0.13119833171367645, + "learning_rate": 3.5553686570216115e-05, + "loss": 0.0127, + "step": 18410 + }, + { + "epoch": 11.57035175879397, + "grad_norm": 0.15600553154945374, + "learning_rate": 3.550093030553284e-05, + "loss": 0.0166, + "step": 18420 + }, + { + "epoch": 11.576633165829145, + "grad_norm": 0.2371159791946411, + "learning_rate": 3.544819165858642e-05, + "loss": 0.0126, + "step": 18430 + }, + { + "epoch": 11.582914572864322, + "grad_norm": 0.2644501328468323, + "learning_rate": 3.539547069345926e-05, + "loss": 0.0136, + "step": 18440 + }, + { + "epoch": 11.589195979899497, + "grad_norm": 0.24654138088226318, + "learning_rate": 3.5342767474212344e-05, + "loss": 0.0115, + "step": 18450 + }, + { + "epoch": 11.595477386934673, + "grad_norm": 0.17095930874347687, + "learning_rate": 3.529008206488502e-05, + "loss": 0.0119, + "step": 18460 + }, + { + "epoch": 11.60175879396985, + "grad_norm": 0.22720737755298615, + "learning_rate": 3.5237414529495055e-05, + "loss": 0.0146, + "step": 18470 + }, + { + "epoch": 11.608040201005025, + "grad_norm": 0.21828748285770416, + "learning_rate": 3.5184764932038454e-05, + "loss": 0.0104, + "step": 18480 + }, + { + "epoch": 11.614321608040202, + "grad_norm": 0.24841317534446716, + "learning_rate": 3.513213333648945e-05, + "loss": 0.01, + "step": 18490 + }, + { + "epoch": 11.620603015075377, + "grad_norm": 0.2986377477645874, + "learning_rate": 3.5079519806800374e-05, + "loss": 0.0109, + "step": 18500 + }, + { + "epoch": 11.626884422110553, + "grad_norm": 0.20567549765110016, + "learning_rate": 3.502692440690165e-05, + "loss": 0.0119, + "step": 18510 + }, + { + "epoch": 11.633165829145728, + "grad_norm": 0.32484379410743713, + "learning_rate": 3.497434720070165e-05, + "loss": 0.0153, + "step": 18520 + }, + { + "epoch": 11.639447236180905, + "grad_norm": 0.19467034935951233, + "learning_rate": 3.4921788252086616e-05, + "loss": 0.0106, + "step": 18530 + }, + { + "epoch": 11.64572864321608, + "grad_norm": 0.18407899141311646, + "learning_rate": 3.486924762492065e-05, + "loss": 0.0113, + "step": 18540 + }, + { + "epoch": 11.652010050251256, + "grad_norm": 0.24306872487068176, + "learning_rate": 3.4816725383045534e-05, + "loss": 0.0151, + "step": 18550 + }, + { + "epoch": 11.658291457286433, + "grad_norm": 0.18827463686466217, + "learning_rate": 3.476422159028079e-05, + "loss": 0.0099, + "step": 18560 + }, + { + "epoch": 11.664572864321608, + "grad_norm": 0.14529332518577576, + "learning_rate": 3.471173631042345e-05, + "loss": 0.0164, + "step": 18570 + }, + { + "epoch": 11.670854271356784, + "grad_norm": 0.17499157786369324, + "learning_rate": 3.465926960724808e-05, + "loss": 0.0125, + "step": 18580 + }, + { + "epoch": 11.67713567839196, + "grad_norm": 0.18664611876010895, + "learning_rate": 3.460682154450666e-05, + "loss": 0.0112, + "step": 18590 + }, + { + "epoch": 11.683417085427136, + "grad_norm": 0.30024775862693787, + "learning_rate": 3.4554392185928564e-05, + "loss": 0.0131, + "step": 18600 + }, + { + "epoch": 11.68969849246231, + "grad_norm": 0.2874975800514221, + "learning_rate": 3.450198159522037e-05, + "loss": 0.0133, + "step": 18610 + }, + { + "epoch": 11.695979899497488, + "grad_norm": 0.2580103576183319, + "learning_rate": 3.444958983606592e-05, + "loss": 0.0131, + "step": 18620 + }, + { + "epoch": 11.702261306532662, + "grad_norm": 0.21010112762451172, + "learning_rate": 3.439721697212612e-05, + "loss": 0.0087, + "step": 18630 + }, + { + "epoch": 11.708542713567839, + "grad_norm": 0.15232571959495544, + "learning_rate": 3.4344863067038954e-05, + "loss": 0.011, + "step": 18640 + }, + { + "epoch": 11.714824120603016, + "grad_norm": 0.23264461755752563, + "learning_rate": 3.429252818441935e-05, + "loss": 0.0094, + "step": 18650 + }, + { + "epoch": 11.72110552763819, + "grad_norm": 0.2175249457359314, + "learning_rate": 3.4240212387859094e-05, + "loss": 0.0114, + "step": 18660 + }, + { + "epoch": 11.727386934673367, + "grad_norm": 0.25763028860092163, + "learning_rate": 3.418791574092686e-05, + "loss": 0.0127, + "step": 18670 + }, + { + "epoch": 11.733668341708542, + "grad_norm": 0.34997323155403137, + "learning_rate": 3.413563830716796e-05, + "loss": 0.02, + "step": 18680 + }, + { + "epoch": 11.739949748743719, + "grad_norm": 0.288896769285202, + "learning_rate": 3.408338015010445e-05, + "loss": 0.0123, + "step": 18690 + }, + { + "epoch": 11.746231155778894, + "grad_norm": 0.31152597069740295, + "learning_rate": 3.40311413332349e-05, + "loss": 0.0131, + "step": 18700 + }, + { + "epoch": 11.75251256281407, + "grad_norm": 0.16146419942378998, + "learning_rate": 3.3978921920034365e-05, + "loss": 0.0136, + "step": 18710 + }, + { + "epoch": 11.758793969849247, + "grad_norm": 0.18438567221164703, + "learning_rate": 3.392672197395441e-05, + "loss": 0.0134, + "step": 18720 + }, + { + "epoch": 11.765075376884422, + "grad_norm": 0.18482300639152527, + "learning_rate": 3.387454155842287e-05, + "loss": 0.0105, + "step": 18730 + }, + { + "epoch": 11.771356783919598, + "grad_norm": 0.2251647263765335, + "learning_rate": 3.382238073684386e-05, + "loss": 0.0112, + "step": 18740 + }, + { + "epoch": 11.777638190954773, + "grad_norm": 0.18330419063568115, + "learning_rate": 3.377023957259771e-05, + "loss": 0.0106, + "step": 18750 + }, + { + "epoch": 11.78391959798995, + "grad_norm": 0.18595869839191437, + "learning_rate": 3.3718118129040835e-05, + "loss": 0.0103, + "step": 18760 + }, + { + "epoch": 11.790201005025125, + "grad_norm": 0.26770931482315063, + "learning_rate": 3.3666016469505724e-05, + "loss": 0.0102, + "step": 18770 + }, + { + "epoch": 11.796482412060302, + "grad_norm": 0.2716752290725708, + "learning_rate": 3.361393465730079e-05, + "loss": 0.0126, + "step": 18780 + }, + { + "epoch": 11.802763819095478, + "grad_norm": 0.23507891595363617, + "learning_rate": 3.3561872755710366e-05, + "loss": 0.0123, + "step": 18790 + }, + { + "epoch": 11.809045226130653, + "grad_norm": 0.23740562796592712, + "learning_rate": 3.350983082799456e-05, + "loss": 0.0121, + "step": 18800 + }, + { + "epoch": 11.81532663316583, + "grad_norm": 0.25879356265068054, + "learning_rate": 3.34578089373892e-05, + "loss": 0.011, + "step": 18810 + }, + { + "epoch": 11.821608040201005, + "grad_norm": 0.19404757022857666, + "learning_rate": 3.340580714710581e-05, + "loss": 0.0123, + "step": 18820 + }, + { + "epoch": 11.827889447236181, + "grad_norm": 0.20331452786922455, + "learning_rate": 3.3353825520331466e-05, + "loss": 0.0115, + "step": 18830 + }, + { + "epoch": 11.834170854271356, + "grad_norm": 0.19864408671855927, + "learning_rate": 3.330186412022876e-05, + "loss": 0.0103, + "step": 18840 + }, + { + "epoch": 11.840452261306533, + "grad_norm": 0.30444711446762085, + "learning_rate": 3.324992300993568e-05, + "loss": 0.0089, + "step": 18850 + }, + { + "epoch": 11.846733668341708, + "grad_norm": 0.3120463788509369, + "learning_rate": 3.319800225256556e-05, + "loss": 0.0121, + "step": 18860 + }, + { + "epoch": 11.853015075376884, + "grad_norm": 0.20235837996006012, + "learning_rate": 3.314610191120702e-05, + "loss": 0.0101, + "step": 18870 + }, + { + "epoch": 11.859296482412061, + "grad_norm": 0.15556682646274567, + "learning_rate": 3.30942220489239e-05, + "loss": 0.0115, + "step": 18880 + }, + { + "epoch": 11.865577889447236, + "grad_norm": 0.168738454580307, + "learning_rate": 3.3042362728755086e-05, + "loss": 0.0117, + "step": 18890 + }, + { + "epoch": 11.871859296482413, + "grad_norm": 0.13974033296108246, + "learning_rate": 3.299052401371456e-05, + "loss": 0.0113, + "step": 18900 + }, + { + "epoch": 11.878140703517587, + "grad_norm": 0.20316560566425323, + "learning_rate": 3.293870596679125e-05, + "loss": 0.0115, + "step": 18910 + }, + { + "epoch": 11.884422110552764, + "grad_norm": 0.20932678878307343, + "learning_rate": 3.288690865094895e-05, + "loss": 0.0088, + "step": 18920 + }, + { + "epoch": 11.890703517587939, + "grad_norm": 0.2133147269487381, + "learning_rate": 3.283513212912632e-05, + "loss": 0.0142, + "step": 18930 + }, + { + "epoch": 11.896984924623116, + "grad_norm": 0.19184312224388123, + "learning_rate": 3.2783376464236684e-05, + "loss": 0.0109, + "step": 18940 + }, + { + "epoch": 11.903266331658292, + "grad_norm": 0.20613424479961395, + "learning_rate": 3.273164171916806e-05, + "loss": 0.0115, + "step": 18950 + }, + { + "epoch": 11.909547738693467, + "grad_norm": 0.27219995856285095, + "learning_rate": 3.267992795678306e-05, + "loss": 0.0105, + "step": 18960 + }, + { + "epoch": 11.915829145728644, + "grad_norm": 0.270359605550766, + "learning_rate": 3.2628235239918744e-05, + "loss": 0.013, + "step": 18970 + }, + { + "epoch": 11.922110552763819, + "grad_norm": 0.17966631054878235, + "learning_rate": 3.2576563631386695e-05, + "loss": 0.0118, + "step": 18980 + }, + { + "epoch": 11.928391959798995, + "grad_norm": 0.18094632029533386, + "learning_rate": 3.252491319397275e-05, + "loss": 0.0105, + "step": 18990 + }, + { + "epoch": 11.93467336683417, + "grad_norm": 0.2722184658050537, + "learning_rate": 3.247328399043706e-05, + "loss": 0.0091, + "step": 19000 + }, + { + "epoch": 11.940954773869347, + "grad_norm": 0.23857015371322632, + "learning_rate": 3.242167608351399e-05, + "loss": 0.0108, + "step": 19010 + }, + { + "epoch": 11.947236180904522, + "grad_norm": 0.33349233865737915, + "learning_rate": 3.2370089535911986e-05, + "loss": 0.0094, + "step": 19020 + }, + { + "epoch": 11.953517587939698, + "grad_norm": 0.2061135172843933, + "learning_rate": 3.23185244103136e-05, + "loss": 0.0129, + "step": 19030 + }, + { + "epoch": 11.959798994974875, + "grad_norm": 0.1634758710861206, + "learning_rate": 3.22669807693753e-05, + "loss": 0.0141, + "step": 19040 + }, + { + "epoch": 11.96608040201005, + "grad_norm": 0.2941510081291199, + "learning_rate": 3.2215458675727495e-05, + "loss": 0.0113, + "step": 19050 + }, + { + "epoch": 11.972361809045227, + "grad_norm": 0.20273222029209137, + "learning_rate": 3.216395819197438e-05, + "loss": 0.0111, + "step": 19060 + }, + { + "epoch": 11.978643216080402, + "grad_norm": 0.1912243813276291, + "learning_rate": 3.211247938069387e-05, + "loss": 0.0094, + "step": 19070 + }, + { + "epoch": 11.984924623115578, + "grad_norm": 0.1848934441804886, + "learning_rate": 3.206102230443759e-05, + "loss": 0.0104, + "step": 19080 + }, + { + "epoch": 11.991206030150753, + "grad_norm": 0.1969435214996338, + "learning_rate": 3.2009587025730764e-05, + "loss": 0.011, + "step": 19090 + }, + { + "epoch": 11.99748743718593, + "grad_norm": 0.20728200674057007, + "learning_rate": 3.195817360707207e-05, + "loss": 0.01, + "step": 19100 + }, + { + "epoch": 12.003768844221106, + "grad_norm": 0.1915435940027237, + "learning_rate": 3.19067821109337e-05, + "loss": 0.0109, + "step": 19110 + }, + { + "epoch": 12.010050251256281, + "grad_norm": 0.27789872884750366, + "learning_rate": 3.185541259976114e-05, + "loss": 0.0102, + "step": 19120 + }, + { + "epoch": 12.016331658291458, + "grad_norm": 0.17358946800231934, + "learning_rate": 3.180406513597316e-05, + "loss": 0.0094, + "step": 19130 + }, + { + "epoch": 12.022613065326633, + "grad_norm": 0.20915262401103973, + "learning_rate": 3.1752739781961835e-05, + "loss": 0.0113, + "step": 19140 + }, + { + "epoch": 12.02889447236181, + "grad_norm": 0.15823568403720856, + "learning_rate": 3.170143660009228e-05, + "loss": 0.0115, + "step": 19150 + }, + { + "epoch": 12.035175879396984, + "grad_norm": 0.19273583590984344, + "learning_rate": 3.16501556527027e-05, + "loss": 0.0109, + "step": 19160 + }, + { + "epoch": 12.041457286432161, + "grad_norm": 0.20688003301620483, + "learning_rate": 3.1598897002104265e-05, + "loss": 0.0105, + "step": 19170 + }, + { + "epoch": 12.047738693467336, + "grad_norm": 0.27340760827064514, + "learning_rate": 3.154766071058108e-05, + "loss": 0.0126, + "step": 19180 + }, + { + "epoch": 12.054020100502512, + "grad_norm": 0.16221192479133606, + "learning_rate": 3.149644684039008e-05, + "loss": 0.012, + "step": 19190 + }, + { + "epoch": 12.06030150753769, + "grad_norm": 0.39649248123168945, + "learning_rate": 3.144525545376095e-05, + "loss": 0.0103, + "step": 19200 + }, + { + "epoch": 12.066582914572864, + "grad_norm": 0.2077530324459076, + "learning_rate": 3.139408661289603e-05, + "loss": 0.0099, + "step": 19210 + }, + { + "epoch": 12.07286432160804, + "grad_norm": 0.15679386258125305, + "learning_rate": 3.134294037997032e-05, + "loss": 0.008, + "step": 19220 + }, + { + "epoch": 12.079145728643216, + "grad_norm": 0.13558270037174225, + "learning_rate": 3.129181681713127e-05, + "loss": 0.0097, + "step": 19230 + }, + { + "epoch": 12.085427135678392, + "grad_norm": 0.2697191536426544, + "learning_rate": 3.1240715986498855e-05, + "loss": 0.012, + "step": 19240 + }, + { + "epoch": 12.091708542713567, + "grad_norm": 0.1623874306678772, + "learning_rate": 3.1189637950165394e-05, + "loss": 0.0102, + "step": 19250 + }, + { + "epoch": 12.097989949748744, + "grad_norm": 0.22612264752388, + "learning_rate": 3.1138582770195544e-05, + "loss": 0.0113, + "step": 19260 + }, + { + "epoch": 12.10427135678392, + "grad_norm": 0.21459683775901794, + "learning_rate": 3.108755050862615e-05, + "loss": 0.0097, + "step": 19270 + }, + { + "epoch": 12.110552763819095, + "grad_norm": 0.2030550241470337, + "learning_rate": 3.1036541227466204e-05, + "loss": 0.0095, + "step": 19280 + }, + { + "epoch": 12.116834170854272, + "grad_norm": 0.15254774689674377, + "learning_rate": 3.0985554988696786e-05, + "loss": 0.0115, + "step": 19290 + }, + { + "epoch": 12.123115577889447, + "grad_norm": 0.19191886484622955, + "learning_rate": 3.093459185427102e-05, + "loss": 0.0134, + "step": 19300 + }, + { + "epoch": 12.129396984924623, + "grad_norm": 0.15385422110557556, + "learning_rate": 3.0883651886113905e-05, + "loss": 0.0104, + "step": 19310 + }, + { + "epoch": 12.135678391959798, + "grad_norm": 0.1874881088733673, + "learning_rate": 3.0832735146122294e-05, + "loss": 0.01, + "step": 19320 + }, + { + "epoch": 12.141959798994975, + "grad_norm": 0.14504356682300568, + "learning_rate": 3.078184169616485e-05, + "loss": 0.0095, + "step": 19330 + }, + { + "epoch": 12.14824120603015, + "grad_norm": 0.16753926873207092, + "learning_rate": 3.073097159808187e-05, + "loss": 0.0101, + "step": 19340 + }, + { + "epoch": 12.154522613065327, + "grad_norm": 0.1886853575706482, + "learning_rate": 3.068012491368537e-05, + "loss": 0.0103, + "step": 19350 + }, + { + "epoch": 12.160804020100503, + "grad_norm": 0.23865672945976257, + "learning_rate": 3.062930170475885e-05, + "loss": 0.0096, + "step": 19360 + }, + { + "epoch": 12.167085427135678, + "grad_norm": 0.17449991405010223, + "learning_rate": 3.057850203305729e-05, + "loss": 0.0107, + "step": 19370 + }, + { + "epoch": 12.173366834170855, + "grad_norm": 0.21366263926029205, + "learning_rate": 3.052772596030708e-05, + "loss": 0.0102, + "step": 19380 + }, + { + "epoch": 12.17964824120603, + "grad_norm": 0.2085852175951004, + "learning_rate": 3.0476973548205943e-05, + "loss": 0.0105, + "step": 19390 + }, + { + "epoch": 12.185929648241206, + "grad_norm": 0.21372422575950623, + "learning_rate": 3.0426244858422847e-05, + "loss": 0.0105, + "step": 19400 + }, + { + "epoch": 12.192211055276381, + "grad_norm": 0.2539060413837433, + "learning_rate": 3.0375539952597943e-05, + "loss": 0.0098, + "step": 19410 + }, + { + "epoch": 12.198492462311558, + "grad_norm": 0.14869362115859985, + "learning_rate": 3.0324858892342468e-05, + "loss": 0.0105, + "step": 19420 + }, + { + "epoch": 12.204773869346734, + "grad_norm": 0.16969045996665955, + "learning_rate": 3.0274201739238672e-05, + "loss": 0.011, + "step": 19430 + }, + { + "epoch": 12.21105527638191, + "grad_norm": 0.21650969982147217, + "learning_rate": 3.0223568554839786e-05, + "loss": 0.0111, + "step": 19440 + }, + { + "epoch": 12.217336683417086, + "grad_norm": 0.13951748609542847, + "learning_rate": 3.0172959400669886e-05, + "loss": 0.0103, + "step": 19450 + }, + { + "epoch": 12.22361809045226, + "grad_norm": 0.19674861431121826, + "learning_rate": 3.0122374338223902e-05, + "loss": 0.0096, + "step": 19460 + }, + { + "epoch": 12.229899497487438, + "grad_norm": 0.1835966557264328, + "learning_rate": 3.0071813428967427e-05, + "loss": 0.0097, + "step": 19470 + }, + { + "epoch": 12.236180904522612, + "grad_norm": 0.2587984502315521, + "learning_rate": 3.0021276734336746e-05, + "loss": 0.0116, + "step": 19480 + }, + { + "epoch": 12.242462311557789, + "grad_norm": 0.1690255105495453, + "learning_rate": 2.997076431573871e-05, + "loss": 0.0133, + "step": 19490 + }, + { + "epoch": 12.248743718592964, + "grad_norm": 0.20671407878398895, + "learning_rate": 2.9920276234550636e-05, + "loss": 0.0141, + "step": 19500 + }, + { + "epoch": 12.25502512562814, + "grad_norm": 0.22522655129432678, + "learning_rate": 2.986981255212035e-05, + "loss": 0.0087, + "step": 19510 + }, + { + "epoch": 12.261306532663317, + "grad_norm": 0.2359679490327835, + "learning_rate": 2.9819373329765977e-05, + "loss": 0.0106, + "step": 19520 + }, + { + "epoch": 12.267587939698492, + "grad_norm": 0.19972094893455505, + "learning_rate": 2.9768958628775902e-05, + "loss": 0.0096, + "step": 19530 + }, + { + "epoch": 12.273869346733669, + "grad_norm": 0.2805884778499603, + "learning_rate": 2.9718568510408762e-05, + "loss": 0.0115, + "step": 19540 + }, + { + "epoch": 12.280150753768844, + "grad_norm": 0.13650736212730408, + "learning_rate": 2.9668203035893272e-05, + "loss": 0.0107, + "step": 19550 + }, + { + "epoch": 12.28643216080402, + "grad_norm": 0.2557235658168793, + "learning_rate": 2.9617862266428287e-05, + "loss": 0.015, + "step": 19560 + }, + { + "epoch": 12.292713567839195, + "grad_norm": 0.13077989220619202, + "learning_rate": 2.9567546263182556e-05, + "loss": 0.0122, + "step": 19570 + }, + { + "epoch": 12.298994974874372, + "grad_norm": 0.28605228662490845, + "learning_rate": 2.951725508729476e-05, + "loss": 0.0137, + "step": 19580 + }, + { + "epoch": 12.305276381909549, + "grad_norm": 0.17540182173252106, + "learning_rate": 2.946698879987344e-05, + "loss": 0.0117, + "step": 19590 + }, + { + "epoch": 12.311557788944723, + "grad_norm": 0.21988964080810547, + "learning_rate": 2.9416747461996853e-05, + "loss": 0.0086, + "step": 19600 + }, + { + "epoch": 12.3178391959799, + "grad_norm": 0.1874978095293045, + "learning_rate": 2.9366531134712973e-05, + "loss": 0.0091, + "step": 19610 + }, + { + "epoch": 12.324120603015075, + "grad_norm": 0.15955983102321625, + "learning_rate": 2.9316339879039367e-05, + "loss": 0.0081, + "step": 19620 + }, + { + "epoch": 12.330402010050252, + "grad_norm": 0.2018986940383911, + "learning_rate": 2.9266173755963167e-05, + "loss": 0.0084, + "step": 19630 + }, + { + "epoch": 12.336683417085426, + "grad_norm": 0.15243971347808838, + "learning_rate": 2.9216032826440926e-05, + "loss": 0.01, + "step": 19640 + }, + { + "epoch": 12.342964824120603, + "grad_norm": 0.16988235712051392, + "learning_rate": 2.9165917151398592e-05, + "loss": 0.0077, + "step": 19650 + }, + { + "epoch": 12.349246231155778, + "grad_norm": 0.23163069784641266, + "learning_rate": 2.9115826791731428e-05, + "loss": 0.0097, + "step": 19660 + }, + { + "epoch": 12.355527638190955, + "grad_norm": 0.19204659759998322, + "learning_rate": 2.906576180830398e-05, + "loss": 0.0118, + "step": 19670 + }, + { + "epoch": 12.361809045226131, + "grad_norm": 0.3229486346244812, + "learning_rate": 2.9015722261949917e-05, + "loss": 0.0131, + "step": 19680 + }, + { + "epoch": 12.368090452261306, + "grad_norm": 0.3149772584438324, + "learning_rate": 2.8965708213471986e-05, + "loss": 0.0091, + "step": 19690 + }, + { + "epoch": 12.374371859296483, + "grad_norm": 0.30555596947669983, + "learning_rate": 2.8915719723641975e-05, + "loss": 0.0099, + "step": 19700 + }, + { + "epoch": 12.380653266331658, + "grad_norm": 0.2569314241409302, + "learning_rate": 2.8865756853200604e-05, + "loss": 0.0124, + "step": 19710 + }, + { + "epoch": 12.386934673366834, + "grad_norm": 0.1611790955066681, + "learning_rate": 2.8815819662857503e-05, + "loss": 0.0137, + "step": 19720 + }, + { + "epoch": 12.39321608040201, + "grad_norm": 0.32003945112228394, + "learning_rate": 2.876590821329105e-05, + "loss": 0.0105, + "step": 19730 + }, + { + "epoch": 12.399497487437186, + "grad_norm": 0.20420344173908234, + "learning_rate": 2.871602256514836e-05, + "loss": 0.0117, + "step": 19740 + }, + { + "epoch": 12.405778894472363, + "grad_norm": 0.2163122594356537, + "learning_rate": 2.8666162779045203e-05, + "loss": 0.0124, + "step": 19750 + }, + { + "epoch": 12.412060301507537, + "grad_norm": 0.18341253697872162, + "learning_rate": 2.8616328915565904e-05, + "loss": 0.0109, + "step": 19760 + }, + { + "epoch": 12.418341708542714, + "grad_norm": 0.15650993585586548, + "learning_rate": 2.856652103526334e-05, + "loss": 0.0096, + "step": 19770 + }, + { + "epoch": 12.424623115577889, + "grad_norm": 0.28134214878082275, + "learning_rate": 2.8516739198658755e-05, + "loss": 0.0121, + "step": 19780 + }, + { + "epoch": 12.430904522613066, + "grad_norm": 0.17561203241348267, + "learning_rate": 2.846698346624177e-05, + "loss": 0.0123, + "step": 19790 + }, + { + "epoch": 12.43718592964824, + "grad_norm": 0.3331955671310425, + "learning_rate": 2.841725389847032e-05, + "loss": 0.0113, + "step": 19800 + }, + { + "epoch": 12.443467336683417, + "grad_norm": 0.23100942373275757, + "learning_rate": 2.8367550555770506e-05, + "loss": 0.0107, + "step": 19810 + }, + { + "epoch": 12.449748743718594, + "grad_norm": 0.24024233222007751, + "learning_rate": 2.831787349853655e-05, + "loss": 0.0116, + "step": 19820 + }, + { + "epoch": 12.456030150753769, + "grad_norm": 0.2856957018375397, + "learning_rate": 2.8268222787130806e-05, + "loss": 0.0092, + "step": 19830 + }, + { + "epoch": 12.462311557788945, + "grad_norm": 0.22598211467266083, + "learning_rate": 2.821859848188355e-05, + "loss": 0.0139, + "step": 19840 + }, + { + "epoch": 12.46859296482412, + "grad_norm": 0.27873003482818604, + "learning_rate": 2.816900064309299e-05, + "loss": 0.0157, + "step": 19850 + }, + { + "epoch": 12.474874371859297, + "grad_norm": 0.23396456241607666, + "learning_rate": 2.811942933102517e-05, + "loss": 0.0108, + "step": 19860 + }, + { + "epoch": 12.481155778894472, + "grad_norm": 0.24046416580677032, + "learning_rate": 2.806988460591391e-05, + "loss": 0.0136, + "step": 19870 + }, + { + "epoch": 12.487437185929648, + "grad_norm": 0.15696625411510468, + "learning_rate": 2.802036652796074e-05, + "loss": 0.0112, + "step": 19880 + }, + { + "epoch": 12.493718592964823, + "grad_norm": 0.1413506120443344, + "learning_rate": 2.797087515733478e-05, + "loss": 0.0094, + "step": 19890 + }, + { + "epoch": 12.5, + "grad_norm": 0.15791255235671997, + "learning_rate": 2.7921410554172723e-05, + "loss": 0.0111, + "step": 19900 + }, + { + "epoch": 12.506281407035177, + "grad_norm": 0.15810950100421906, + "learning_rate": 2.787197277857871e-05, + "loss": 0.0106, + "step": 19910 + }, + { + "epoch": 12.512562814070352, + "grad_norm": 0.10429465025663376, + "learning_rate": 2.782256189062429e-05, + "loss": 0.0098, + "step": 19920 + }, + { + "epoch": 12.518844221105528, + "grad_norm": 0.2563154399394989, + "learning_rate": 2.777317795034839e-05, + "loss": 0.0101, + "step": 19930 + }, + { + "epoch": 12.525125628140703, + "grad_norm": 0.29390949010849, + "learning_rate": 2.7723821017757112e-05, + "loss": 0.0079, + "step": 19940 + }, + { + "epoch": 12.53140703517588, + "grad_norm": 0.1802944391965866, + "learning_rate": 2.7674491152823822e-05, + "loss": 0.0084, + "step": 19950 + }, + { + "epoch": 12.537688442211055, + "grad_norm": 0.203603133559227, + "learning_rate": 2.7625188415488944e-05, + "loss": 0.0132, + "step": 19960 + }, + { + "epoch": 12.543969849246231, + "grad_norm": 0.16414202749729156, + "learning_rate": 2.7575912865659924e-05, + "loss": 0.0092, + "step": 19970 + }, + { + "epoch": 12.550251256281406, + "grad_norm": 0.2113833874464035, + "learning_rate": 2.7526664563211245e-05, + "loss": 0.0118, + "step": 19980 + }, + { + "epoch": 12.556532663316583, + "grad_norm": 0.24079887568950653, + "learning_rate": 2.7477443567984224e-05, + "loss": 0.0093, + "step": 19990 + }, + { + "epoch": 12.56281407035176, + "grad_norm": 0.27356186509132385, + "learning_rate": 2.7428249939787e-05, + "loss": 0.0117, + "step": 20000 + }, + { + "epoch": 12.569095477386934, + "grad_norm": 0.3238692879676819, + "learning_rate": 2.7379083738394484e-05, + "loss": 0.0102, + "step": 20010 + }, + { + "epoch": 12.575376884422111, + "grad_norm": 0.19072325527668, + "learning_rate": 2.732994502354823e-05, + "loss": 0.0075, + "step": 20020 + }, + { + "epoch": 12.581658291457286, + "grad_norm": 0.18499872088432312, + "learning_rate": 2.72808338549564e-05, + "loss": 0.0099, + "step": 20030 + }, + { + "epoch": 12.587939698492463, + "grad_norm": 0.11942707002162933, + "learning_rate": 2.723175029229374e-05, + "loss": 0.0107, + "step": 20040 + }, + { + "epoch": 12.594221105527637, + "grad_norm": 0.24408091604709625, + "learning_rate": 2.718269439520138e-05, + "loss": 0.0113, + "step": 20050 + }, + { + "epoch": 12.600502512562814, + "grad_norm": 0.22844573855400085, + "learning_rate": 2.7133666223286858e-05, + "loss": 0.009, + "step": 20060 + }, + { + "epoch": 12.60678391959799, + "grad_norm": 0.19148758053779602, + "learning_rate": 2.7084665836124006e-05, + "loss": 0.0113, + "step": 20070 + }, + { + "epoch": 12.613065326633166, + "grad_norm": 0.22013014554977417, + "learning_rate": 2.703569329325296e-05, + "loss": 0.0099, + "step": 20080 + }, + { + "epoch": 12.619346733668342, + "grad_norm": 0.20292291045188904, + "learning_rate": 2.698674865417994e-05, + "loss": 0.0083, + "step": 20090 + }, + { + "epoch": 12.625628140703517, + "grad_norm": 0.2790031433105469, + "learning_rate": 2.6937831978377325e-05, + "loss": 0.0113, + "step": 20100 + }, + { + "epoch": 12.631909547738694, + "grad_norm": 0.22517995536327362, + "learning_rate": 2.6888943325283485e-05, + "loss": 0.0099, + "step": 20110 + }, + { + "epoch": 12.638190954773869, + "grad_norm": 0.14891941845417023, + "learning_rate": 2.6840082754302733e-05, + "loss": 0.0102, + "step": 20120 + }, + { + "epoch": 12.644472361809045, + "grad_norm": 0.19086642563343048, + "learning_rate": 2.679125032480525e-05, + "loss": 0.0086, + "step": 20130 + }, + { + "epoch": 12.650753768844222, + "grad_norm": 0.23904721438884735, + "learning_rate": 2.674244609612708e-05, + "loss": 0.0086, + "step": 20140 + }, + { + "epoch": 12.657035175879397, + "grad_norm": 0.268151193857193, + "learning_rate": 2.6693670127569958e-05, + "loss": 0.0118, + "step": 20150 + }, + { + "epoch": 12.663316582914574, + "grad_norm": 0.2710839509963989, + "learning_rate": 2.6644922478401268e-05, + "loss": 0.0139, + "step": 20160 + }, + { + "epoch": 12.669597989949748, + "grad_norm": 0.2917962074279785, + "learning_rate": 2.6596203207854004e-05, + "loss": 0.0119, + "step": 20170 + }, + { + "epoch": 12.675879396984925, + "grad_norm": 0.1728828102350235, + "learning_rate": 2.654751237512666e-05, + "loss": 0.0087, + "step": 20180 + }, + { + "epoch": 12.6821608040201, + "grad_norm": 0.14040876924991608, + "learning_rate": 2.649885003938323e-05, + "loss": 0.0152, + "step": 20190 + }, + { + "epoch": 12.688442211055277, + "grad_norm": 0.2732071578502655, + "learning_rate": 2.6450216259753e-05, + "loss": 0.0108, + "step": 20200 + }, + { + "epoch": 12.694723618090451, + "grad_norm": 0.2793784737586975, + "learning_rate": 2.640161109533063e-05, + "loss": 0.0134, + "step": 20210 + }, + { + "epoch": 12.701005025125628, + "grad_norm": 0.18674607574939728, + "learning_rate": 2.6353034605175936e-05, + "loss": 0.0107, + "step": 20220 + }, + { + "epoch": 12.707286432160805, + "grad_norm": 0.18014845252037048, + "learning_rate": 2.630448684831398e-05, + "loss": 0.0097, + "step": 20230 + }, + { + "epoch": 12.71356783919598, + "grad_norm": 0.17706726491451263, + "learning_rate": 2.625596788373482e-05, + "loss": 0.0101, + "step": 20240 + }, + { + "epoch": 12.719849246231156, + "grad_norm": 0.23989318311214447, + "learning_rate": 2.620747777039363e-05, + "loss": 0.0104, + "step": 20250 + }, + { + "epoch": 12.726130653266331, + "grad_norm": 0.1676013320684433, + "learning_rate": 2.6159016567210424e-05, + "loss": 0.0076, + "step": 20260 + }, + { + "epoch": 12.732412060301508, + "grad_norm": 0.20459747314453125, + "learning_rate": 2.6110584333070153e-05, + "loss": 0.0115, + "step": 20270 + }, + { + "epoch": 12.738693467336683, + "grad_norm": 0.18310889601707458, + "learning_rate": 2.6062181126822537e-05, + "loss": 0.011, + "step": 20280 + }, + { + "epoch": 12.74497487437186, + "grad_norm": 0.2695169746875763, + "learning_rate": 2.601380700728203e-05, + "loss": 0.0086, + "step": 20290 + }, + { + "epoch": 12.751256281407034, + "grad_norm": 0.14328987896442413, + "learning_rate": 2.596546203322777e-05, + "loss": 0.0081, + "step": 20300 + }, + { + "epoch": 12.757537688442211, + "grad_norm": 0.12292409688234329, + "learning_rate": 2.5917146263403457e-05, + "loss": 0.0098, + "step": 20310 + }, + { + "epoch": 12.763819095477388, + "grad_norm": 0.23216213285923004, + "learning_rate": 2.5868859756517296e-05, + "loss": 0.0093, + "step": 20320 + }, + { + "epoch": 12.770100502512562, + "grad_norm": 0.18020421266555786, + "learning_rate": 2.5820602571241947e-05, + "loss": 0.0083, + "step": 20330 + }, + { + "epoch": 12.77638190954774, + "grad_norm": 0.1879940927028656, + "learning_rate": 2.577237476621442e-05, + "loss": 0.0107, + "step": 20340 + }, + { + "epoch": 12.782663316582914, + "grad_norm": 0.16725970804691315, + "learning_rate": 2.572417640003609e-05, + "loss": 0.0125, + "step": 20350 + }, + { + "epoch": 12.78894472361809, + "grad_norm": 0.16574358940124512, + "learning_rate": 2.5676007531272473e-05, + "loss": 0.0113, + "step": 20360 + }, + { + "epoch": 12.795226130653266, + "grad_norm": 0.20751261711120605, + "learning_rate": 2.5627868218453332e-05, + "loss": 0.0111, + "step": 20370 + }, + { + "epoch": 12.801507537688442, + "grad_norm": 0.20640990138053894, + "learning_rate": 2.5579758520072443e-05, + "loss": 0.01, + "step": 20380 + }, + { + "epoch": 12.807788944723619, + "grad_norm": 0.23280645906925201, + "learning_rate": 2.5531678494587615e-05, + "loss": 0.0073, + "step": 20390 + }, + { + "epoch": 12.814070351758794, + "grad_norm": 0.20260591804981232, + "learning_rate": 2.5483628200420644e-05, + "loss": 0.0104, + "step": 20400 + }, + { + "epoch": 12.82035175879397, + "grad_norm": 0.17827899754047394, + "learning_rate": 2.5435607695957154e-05, + "loss": 0.0069, + "step": 20410 + }, + { + "epoch": 12.826633165829145, + "grad_norm": 0.2619847059249878, + "learning_rate": 2.5387617039546586e-05, + "loss": 0.0127, + "step": 20420 + }, + { + "epoch": 12.832914572864322, + "grad_norm": 0.16657882928848267, + "learning_rate": 2.5339656289502106e-05, + "loss": 0.0083, + "step": 20430 + }, + { + "epoch": 12.839195979899497, + "grad_norm": 0.16552196443080902, + "learning_rate": 2.529172550410056e-05, + "loss": 0.0092, + "step": 20440 + }, + { + "epoch": 12.845477386934673, + "grad_norm": 0.25846096873283386, + "learning_rate": 2.524382474158234e-05, + "loss": 0.0099, + "step": 20450 + }, + { + "epoch": 12.85175879396985, + "grad_norm": 0.17665468156337738, + "learning_rate": 2.5195954060151433e-05, + "loss": 0.0098, + "step": 20460 + }, + { + "epoch": 12.858040201005025, + "grad_norm": 0.12102749943733215, + "learning_rate": 2.5148113517975212e-05, + "loss": 0.0107, + "step": 20470 + }, + { + "epoch": 12.864321608040202, + "grad_norm": 0.14590507745742798, + "learning_rate": 2.5100303173184447e-05, + "loss": 0.0112, + "step": 20480 + }, + { + "epoch": 12.870603015075377, + "grad_norm": 0.2247614860534668, + "learning_rate": 2.5052523083873196e-05, + "loss": 0.0083, + "step": 20490 + }, + { + "epoch": 12.876884422110553, + "grad_norm": 0.156962051987648, + "learning_rate": 2.500477330809881e-05, + "loss": 0.0079, + "step": 20500 + }, + { + "epoch": 12.883165829145728, + "grad_norm": 0.16395951807498932, + "learning_rate": 2.4957053903881733e-05, + "loss": 0.0092, + "step": 20510 + }, + { + "epoch": 12.889447236180905, + "grad_norm": 0.12030437588691711, + "learning_rate": 2.4909364929205576e-05, + "loss": 0.0122, + "step": 20520 + }, + { + "epoch": 12.89572864321608, + "grad_norm": 0.2177876979112625, + "learning_rate": 2.4861706442016923e-05, + "loss": 0.0099, + "step": 20530 + }, + { + "epoch": 12.902010050251256, + "grad_norm": 0.28360480070114136, + "learning_rate": 2.481407850022533e-05, + "loss": 0.0094, + "step": 20540 + }, + { + "epoch": 12.908291457286433, + "grad_norm": 0.14457039535045624, + "learning_rate": 2.4766481161703213e-05, + "loss": 0.0104, + "step": 20550 + }, + { + "epoch": 12.914572864321608, + "grad_norm": 0.19448800384998322, + "learning_rate": 2.4718914484285876e-05, + "loss": 0.0095, + "step": 20560 + }, + { + "epoch": 12.920854271356784, + "grad_norm": 0.20492933690547943, + "learning_rate": 2.467137852577129e-05, + "loss": 0.0109, + "step": 20570 + }, + { + "epoch": 12.92713567839196, + "grad_norm": 0.23589566349983215, + "learning_rate": 2.4623873343920123e-05, + "loss": 0.01, + "step": 20580 + }, + { + "epoch": 12.933417085427136, + "grad_norm": 0.17581722140312195, + "learning_rate": 2.4576398996455656e-05, + "loss": 0.0099, + "step": 20590 + }, + { + "epoch": 12.93969849246231, + "grad_norm": 0.1912926435470581, + "learning_rate": 2.452895554106368e-05, + "loss": 0.0105, + "step": 20600 + }, + { + "epoch": 12.945979899497488, + "grad_norm": 0.23433993756771088, + "learning_rate": 2.4481543035392506e-05, + "loss": 0.0111, + "step": 20610 + }, + { + "epoch": 12.952261306532662, + "grad_norm": 0.21419398486614227, + "learning_rate": 2.4434161537052774e-05, + "loss": 0.0103, + "step": 20620 + }, + { + "epoch": 12.958542713567839, + "grad_norm": 0.09409201890230179, + "learning_rate": 2.4386811103617474e-05, + "loss": 0.0078, + "step": 20630 + }, + { + "epoch": 12.964824120603016, + "grad_norm": 0.2478906661272049, + "learning_rate": 2.4339491792621833e-05, + "loss": 0.0088, + "step": 20640 + }, + { + "epoch": 12.97110552763819, + "grad_norm": 0.19108760356903076, + "learning_rate": 2.4292203661563312e-05, + "loss": 0.0093, + "step": 20650 + }, + { + "epoch": 12.977386934673367, + "grad_norm": 0.24026836454868317, + "learning_rate": 2.424494676790141e-05, + "loss": 0.0096, + "step": 20660 + }, + { + "epoch": 12.983668341708542, + "grad_norm": 0.2522642910480499, + "learning_rate": 2.4197721169057745e-05, + "loss": 0.013, + "step": 20670 + }, + { + "epoch": 12.989949748743719, + "grad_norm": 0.1732054352760315, + "learning_rate": 2.4150526922415857e-05, + "loss": 0.0102, + "step": 20680 + }, + { + "epoch": 12.996231155778894, + "grad_norm": 0.13546256721019745, + "learning_rate": 2.41033640853212e-05, + "loss": 0.0088, + "step": 20690 + }, + { + "epoch": 13.00251256281407, + "grad_norm": 0.22783473134040833, + "learning_rate": 2.405623271508108e-05, + "loss": 0.0097, + "step": 20700 + }, + { + "epoch": 13.008793969849247, + "grad_norm": 0.20083682239055634, + "learning_rate": 2.4009132868964522e-05, + "loss": 0.0095, + "step": 20710 + }, + { + "epoch": 13.015075376884422, + "grad_norm": 0.163008451461792, + "learning_rate": 2.3962064604202327e-05, + "loss": 0.0097, + "step": 20720 + }, + { + "epoch": 13.021356783919598, + "grad_norm": 0.13710036873817444, + "learning_rate": 2.3915027977986858e-05, + "loss": 0.0119, + "step": 20730 + }, + { + "epoch": 13.027638190954773, + "grad_norm": 0.17067471146583557, + "learning_rate": 2.386802304747205e-05, + "loss": 0.0094, + "step": 20740 + }, + { + "epoch": 13.03391959798995, + "grad_norm": 0.15057457983493805, + "learning_rate": 2.382104986977332e-05, + "loss": 0.0099, + "step": 20750 + }, + { + "epoch": 13.040201005025125, + "grad_norm": 0.17020867764949799, + "learning_rate": 2.377410850196749e-05, + "loss": 0.0077, + "step": 20760 + }, + { + "epoch": 13.046482412060302, + "grad_norm": 0.12676210701465607, + "learning_rate": 2.3727199001092797e-05, + "loss": 0.0098, + "step": 20770 + }, + { + "epoch": 13.052763819095478, + "grad_norm": 0.2920474410057068, + "learning_rate": 2.3680321424148677e-05, + "loss": 0.0101, + "step": 20780 + }, + { + "epoch": 13.059045226130653, + "grad_norm": 0.14141082763671875, + "learning_rate": 2.3633475828095793e-05, + "loss": 0.0107, + "step": 20790 + }, + { + "epoch": 13.06532663316583, + "grad_norm": 0.2338477373123169, + "learning_rate": 2.358666226985599e-05, + "loss": 0.0118, + "step": 20800 + }, + { + "epoch": 13.071608040201005, + "grad_norm": 0.19857868552207947, + "learning_rate": 2.353988080631213e-05, + "loss": 0.0121, + "step": 20810 + }, + { + "epoch": 13.077889447236181, + "grad_norm": 0.21104645729064941, + "learning_rate": 2.3493131494308142e-05, + "loss": 0.0083, + "step": 20820 + }, + { + "epoch": 13.084170854271356, + "grad_norm": 0.1960153877735138, + "learning_rate": 2.344641439064881e-05, + "loss": 0.0094, + "step": 20830 + }, + { + "epoch": 13.090452261306533, + "grad_norm": 0.29504308104515076, + "learning_rate": 2.339972955209984e-05, + "loss": 0.0098, + "step": 20840 + }, + { + "epoch": 13.096733668341708, + "grad_norm": 0.1732785999774933, + "learning_rate": 2.335307703538771e-05, + "loss": 0.0085, + "step": 20850 + }, + { + "epoch": 13.103015075376884, + "grad_norm": 0.14724378287792206, + "learning_rate": 2.330645689719962e-05, + "loss": 0.011, + "step": 20860 + }, + { + "epoch": 13.109296482412061, + "grad_norm": 0.19405204057693481, + "learning_rate": 2.3259869194183414e-05, + "loss": 0.0085, + "step": 20870 + }, + { + "epoch": 13.115577889447236, + "grad_norm": 0.18226632475852966, + "learning_rate": 2.321331398294759e-05, + "loss": 0.0087, + "step": 20880 + }, + { + "epoch": 13.121859296482413, + "grad_norm": 0.22831180691719055, + "learning_rate": 2.3166791320061094e-05, + "loss": 0.0092, + "step": 20890 + }, + { + "epoch": 13.128140703517587, + "grad_norm": 0.3081909716129303, + "learning_rate": 2.312030126205335e-05, + "loss": 0.0091, + "step": 20900 + }, + { + "epoch": 13.134422110552764, + "grad_norm": 0.18195979297161102, + "learning_rate": 2.3073843865414162e-05, + "loss": 0.0084, + "step": 20910 + }, + { + "epoch": 13.140703517587939, + "grad_norm": 0.1959652453660965, + "learning_rate": 2.302741918659363e-05, + "loss": 0.0096, + "step": 20920 + }, + { + "epoch": 13.146984924623116, + "grad_norm": 0.4322569966316223, + "learning_rate": 2.2981027282002154e-05, + "loss": 0.011, + "step": 20930 + }, + { + "epoch": 13.153266331658292, + "grad_norm": 0.1459801197052002, + "learning_rate": 2.2934668208010236e-05, + "loss": 0.0066, + "step": 20940 + }, + { + "epoch": 13.159547738693467, + "grad_norm": 0.16677621006965637, + "learning_rate": 2.2888342020948555e-05, + "loss": 0.0107, + "step": 20950 + }, + { + "epoch": 13.165829145728644, + "grad_norm": 0.10833536833524704, + "learning_rate": 2.2842048777107783e-05, + "loss": 0.0095, + "step": 20960 + }, + { + "epoch": 13.172110552763819, + "grad_norm": 0.2098398357629776, + "learning_rate": 2.2795788532738553e-05, + "loss": 0.0097, + "step": 20970 + }, + { + "epoch": 13.178391959798995, + "grad_norm": 0.18453285098075867, + "learning_rate": 2.2749561344051468e-05, + "loss": 0.0106, + "step": 20980 + }, + { + "epoch": 13.18467336683417, + "grad_norm": 0.3253284990787506, + "learning_rate": 2.2703367267216896e-05, + "loss": 0.0108, + "step": 20990 + }, + { + "epoch": 13.190954773869347, + "grad_norm": 0.2251325249671936, + "learning_rate": 2.2657206358364997e-05, + "loss": 0.0122, + "step": 21000 + }, + { + "epoch": 13.197236180904522, + "grad_norm": 0.1767096072435379, + "learning_rate": 2.261107867358563e-05, + "loss": 0.0138, + "step": 21010 + }, + { + "epoch": 13.203517587939698, + "grad_norm": 0.21544905006885529, + "learning_rate": 2.2564984268928263e-05, + "loss": 0.0082, + "step": 21020 + }, + { + "epoch": 13.209798994974875, + "grad_norm": 0.26004186272621155, + "learning_rate": 2.251892320040198e-05, + "loss": 0.008, + "step": 21030 + }, + { + "epoch": 13.21608040201005, + "grad_norm": 0.17255471646785736, + "learning_rate": 2.2472895523975312e-05, + "loss": 0.0094, + "step": 21040 + }, + { + "epoch": 13.222361809045227, + "grad_norm": 0.24310548603534698, + "learning_rate": 2.2426901295576212e-05, + "loss": 0.0077, + "step": 21050 + }, + { + "epoch": 13.228643216080402, + "grad_norm": 0.16579750180244446, + "learning_rate": 2.2380940571092013e-05, + "loss": 0.0098, + "step": 21060 + }, + { + "epoch": 13.234924623115578, + "grad_norm": 0.22338326275348663, + "learning_rate": 2.2335013406369302e-05, + "loss": 0.0099, + "step": 21070 + }, + { + "epoch": 13.241206030150753, + "grad_norm": 0.15400362014770508, + "learning_rate": 2.228911985721397e-05, + "loss": 0.0075, + "step": 21080 + }, + { + "epoch": 13.24748743718593, + "grad_norm": 0.19983460009098053, + "learning_rate": 2.224325997939095e-05, + "loss": 0.0165, + "step": 21090 + }, + { + "epoch": 13.253768844221106, + "grad_norm": 0.13499465584754944, + "learning_rate": 2.219743382862437e-05, + "loss": 0.008, + "step": 21100 + }, + { + "epoch": 13.260050251256281, + "grad_norm": 0.2421891987323761, + "learning_rate": 2.2151641460597294e-05, + "loss": 0.0105, + "step": 21110 + }, + { + "epoch": 13.266331658291458, + "grad_norm": 0.13476026058197021, + "learning_rate": 2.210588293095177e-05, + "loss": 0.008, + "step": 21120 + }, + { + "epoch": 13.272613065326633, + "grad_norm": 0.12455989420413971, + "learning_rate": 2.2060158295288714e-05, + "loss": 0.0085, + "step": 21130 + }, + { + "epoch": 13.27889447236181, + "grad_norm": 0.26662224531173706, + "learning_rate": 2.2014467609167905e-05, + "loss": 0.008, + "step": 21140 + }, + { + "epoch": 13.285175879396984, + "grad_norm": 0.11082585155963898, + "learning_rate": 2.1968810928107806e-05, + "loss": 0.0108, + "step": 21150 + }, + { + "epoch": 13.291457286432161, + "grad_norm": 0.17443419992923737, + "learning_rate": 2.1923188307585606e-05, + "loss": 0.0093, + "step": 21160 + }, + { + "epoch": 13.297738693467336, + "grad_norm": 0.21441295742988586, + "learning_rate": 2.187759980303708e-05, + "loss": 0.0075, + "step": 21170 + }, + { + "epoch": 13.304020100502512, + "grad_norm": 0.10200546681880951, + "learning_rate": 2.1832045469856545e-05, + "loss": 0.0099, + "step": 21180 + }, + { + "epoch": 13.31030150753769, + "grad_norm": 0.19804596900939941, + "learning_rate": 2.1786525363396844e-05, + "loss": 0.0076, + "step": 21190 + }, + { + "epoch": 13.316582914572864, + "grad_norm": 0.1994001418352127, + "learning_rate": 2.1741039538969183e-05, + "loss": 0.0076, + "step": 21200 + }, + { + "epoch": 13.32286432160804, + "grad_norm": 0.2027440369129181, + "learning_rate": 2.169558805184313e-05, + "loss": 0.0093, + "step": 21210 + }, + { + "epoch": 13.329145728643216, + "grad_norm": 0.3110339939594269, + "learning_rate": 2.165017095724651e-05, + "loss": 0.0092, + "step": 21220 + }, + { + "epoch": 13.335427135678392, + "grad_norm": 0.31586289405822754, + "learning_rate": 2.1604788310365405e-05, + "loss": 0.009, + "step": 21230 + }, + { + "epoch": 13.341708542713567, + "grad_norm": 0.15471583604812622, + "learning_rate": 2.155944016634398e-05, + "loss": 0.0094, + "step": 21240 + }, + { + "epoch": 13.347989949748744, + "grad_norm": 0.13440871238708496, + "learning_rate": 2.1514126580284538e-05, + "loss": 0.0107, + "step": 21250 + }, + { + "epoch": 13.35427135678392, + "grad_norm": 0.14488273859024048, + "learning_rate": 2.1468847607247344e-05, + "loss": 0.0097, + "step": 21260 + }, + { + "epoch": 13.360552763819095, + "grad_norm": 0.17321549355983734, + "learning_rate": 2.1423603302250623e-05, + "loss": 0.0096, + "step": 21270 + }, + { + "epoch": 13.366834170854272, + "grad_norm": 0.21315933763980865, + "learning_rate": 2.1378393720270468e-05, + "loss": 0.0143, + "step": 21280 + }, + { + "epoch": 13.373115577889447, + "grad_norm": 0.3002174496650696, + "learning_rate": 2.133321891624076e-05, + "loss": 0.0112, + "step": 21290 + }, + { + "epoch": 13.379396984924623, + "grad_norm": 0.21275845170021057, + "learning_rate": 2.1288078945053193e-05, + "loss": 0.0071, + "step": 21300 + }, + { + "epoch": 13.385678391959798, + "grad_norm": 0.25623393058776855, + "learning_rate": 2.1242973861557064e-05, + "loss": 0.0102, + "step": 21310 + }, + { + "epoch": 13.391959798994975, + "grad_norm": 0.2710457146167755, + "learning_rate": 2.1197903720559304e-05, + "loss": 0.0067, + "step": 21320 + }, + { + "epoch": 13.39824120603015, + "grad_norm": 0.2693053185939789, + "learning_rate": 2.115286857682438e-05, + "loss": 0.0075, + "step": 21330 + }, + { + "epoch": 13.404522613065327, + "grad_norm": 0.2339046746492386, + "learning_rate": 2.1107868485074228e-05, + "loss": 0.0095, + "step": 21340 + }, + { + "epoch": 13.410804020100503, + "grad_norm": 0.13889411091804504, + "learning_rate": 2.1062903499988234e-05, + "loss": 0.0092, + "step": 21350 + }, + { + "epoch": 13.417085427135678, + "grad_norm": 0.187186598777771, + "learning_rate": 2.101797367620308e-05, + "loss": 0.0105, + "step": 21360 + }, + { + "epoch": 13.423366834170855, + "grad_norm": 0.2233499139547348, + "learning_rate": 2.0973079068312713e-05, + "loss": 0.0137, + "step": 21370 + }, + { + "epoch": 13.42964824120603, + "grad_norm": 0.1659775674343109, + "learning_rate": 2.0928219730868358e-05, + "loss": 0.0112, + "step": 21380 + }, + { + "epoch": 13.435929648241206, + "grad_norm": 0.23325271904468536, + "learning_rate": 2.0883395718378302e-05, + "loss": 0.008, + "step": 21390 + }, + { + "epoch": 13.442211055276381, + "grad_norm": 0.21517755091190338, + "learning_rate": 2.083860708530798e-05, + "loss": 0.0086, + "step": 21400 + }, + { + "epoch": 13.448492462311558, + "grad_norm": 0.1797383427619934, + "learning_rate": 2.0793853886079796e-05, + "loss": 0.0114, + "step": 21410 + }, + { + "epoch": 13.454773869346734, + "grad_norm": 0.22263437509536743, + "learning_rate": 2.074913617507309e-05, + "loss": 0.0092, + "step": 21420 + }, + { + "epoch": 13.46105527638191, + "grad_norm": 0.1869269758462906, + "learning_rate": 2.070445400662411e-05, + "loss": 0.0104, + "step": 21430 + }, + { + "epoch": 13.467336683417086, + "grad_norm": 0.25241896510124207, + "learning_rate": 2.0659807435025907e-05, + "loss": 0.0085, + "step": 21440 + }, + { + "epoch": 13.47361809045226, + "grad_norm": 0.2344004213809967, + "learning_rate": 2.061519651452825e-05, + "loss": 0.0114, + "step": 21450 + }, + { + "epoch": 13.479899497487438, + "grad_norm": 0.4753926396369934, + "learning_rate": 2.0570621299337656e-05, + "loss": 0.0115, + "step": 21460 + }, + { + "epoch": 13.486180904522612, + "grad_norm": 0.2021659016609192, + "learning_rate": 2.0526081843617183e-05, + "loss": 0.0103, + "step": 21470 + }, + { + "epoch": 13.492462311557789, + "grad_norm": 0.3413088619709015, + "learning_rate": 2.0481578201486485e-05, + "loss": 0.0094, + "step": 21480 + }, + { + "epoch": 13.498743718592966, + "grad_norm": 0.23444928228855133, + "learning_rate": 2.0437110427021677e-05, + "loss": 0.0086, + "step": 21490 + }, + { + "epoch": 13.50502512562814, + "grad_norm": 0.17304569482803345, + "learning_rate": 2.0392678574255282e-05, + "loss": 0.0088, + "step": 21500 + }, + { + "epoch": 13.511306532663317, + "grad_norm": 0.3122180700302124, + "learning_rate": 2.0348282697176217e-05, + "loss": 0.0092, + "step": 21510 + }, + { + "epoch": 13.517587939698492, + "grad_norm": 0.2453009933233261, + "learning_rate": 2.030392284972964e-05, + "loss": 0.0086, + "step": 21520 + }, + { + "epoch": 13.523869346733669, + "grad_norm": 0.14529870450496674, + "learning_rate": 2.0259599085816973e-05, + "loss": 0.0097, + "step": 21530 + }, + { + "epoch": 13.530150753768844, + "grad_norm": 0.1411176323890686, + "learning_rate": 2.0215311459295755e-05, + "loss": 0.0082, + "step": 21540 + }, + { + "epoch": 13.53643216080402, + "grad_norm": 0.14912313222885132, + "learning_rate": 2.0171060023979606e-05, + "loss": 0.0081, + "step": 21550 + }, + { + "epoch": 13.542713567839195, + "grad_norm": 0.24054986238479614, + "learning_rate": 2.012684483363823e-05, + "loss": 0.0069, + "step": 21560 + }, + { + "epoch": 13.548994974874372, + "grad_norm": 0.10598018765449524, + "learning_rate": 2.0082665941997237e-05, + "loss": 0.0074, + "step": 21570 + }, + { + "epoch": 13.555276381909549, + "grad_norm": 0.1747768372297287, + "learning_rate": 2.0038523402738146e-05, + "loss": 0.0082, + "step": 21580 + }, + { + "epoch": 13.561557788944723, + "grad_norm": 0.13098907470703125, + "learning_rate": 1.99944172694983e-05, + "loss": 0.008, + "step": 21590 + }, + { + "epoch": 13.5678391959799, + "grad_norm": 0.3309729993343353, + "learning_rate": 1.99503475958708e-05, + "loss": 0.0125, + "step": 21600 + }, + { + "epoch": 13.574120603015075, + "grad_norm": 0.3286510407924652, + "learning_rate": 1.9906314435404484e-05, + "loss": 0.007, + "step": 21610 + }, + { + "epoch": 13.580402010050252, + "grad_norm": 0.16493529081344604, + "learning_rate": 1.986231784160378e-05, + "loss": 0.0082, + "step": 21620 + }, + { + "epoch": 13.586683417085426, + "grad_norm": 0.17338161170482635, + "learning_rate": 1.9818357867928693e-05, + "loss": 0.009, + "step": 21630 + }, + { + "epoch": 13.592964824120603, + "grad_norm": 0.15512174367904663, + "learning_rate": 1.977443456779474e-05, + "loss": 0.0089, + "step": 21640 + }, + { + "epoch": 13.599246231155778, + "grad_norm": 0.11736559122800827, + "learning_rate": 1.9730547994572858e-05, + "loss": 0.0079, + "step": 21650 + }, + { + "epoch": 13.605527638190955, + "grad_norm": 0.14767786860466003, + "learning_rate": 1.9686698201589393e-05, + "loss": 0.0087, + "step": 21660 + }, + { + "epoch": 13.611809045226131, + "grad_norm": 0.1636088788509369, + "learning_rate": 1.9642885242125964e-05, + "loss": 0.0082, + "step": 21670 + }, + { + "epoch": 13.618090452261306, + "grad_norm": 0.2228415608406067, + "learning_rate": 1.9599109169419466e-05, + "loss": 0.0079, + "step": 21680 + }, + { + "epoch": 13.624371859296483, + "grad_norm": 0.15854512155056, + "learning_rate": 1.9555370036661948e-05, + "loss": 0.0086, + "step": 21690 + }, + { + "epoch": 13.630653266331658, + "grad_norm": 0.1367972046136856, + "learning_rate": 1.9511667897000574e-05, + "loss": 0.0108, + "step": 21700 + }, + { + "epoch": 13.636934673366834, + "grad_norm": 0.16768968105316162, + "learning_rate": 1.946800280353755e-05, + "loss": 0.0089, + "step": 21710 + }, + { + "epoch": 13.64321608040201, + "grad_norm": 0.18657004833221436, + "learning_rate": 1.9424374809330114e-05, + "loss": 0.0089, + "step": 21720 + }, + { + "epoch": 13.649497487437186, + "grad_norm": 0.1910765916109085, + "learning_rate": 1.9380783967390382e-05, + "loss": 0.0101, + "step": 21730 + }, + { + "epoch": 13.655778894472363, + "grad_norm": 0.1578960120677948, + "learning_rate": 1.933723033068533e-05, + "loss": 0.0099, + "step": 21740 + }, + { + "epoch": 13.662060301507537, + "grad_norm": 0.21472570300102234, + "learning_rate": 1.929371395213674e-05, + "loss": 0.01, + "step": 21750 + }, + { + "epoch": 13.668341708542714, + "grad_norm": 0.2206518054008484, + "learning_rate": 1.9250234884621092e-05, + "loss": 0.01, + "step": 21760 + }, + { + "epoch": 13.674623115577889, + "grad_norm": 0.17132753133773804, + "learning_rate": 1.920679318096959e-05, + "loss": 0.0096, + "step": 21770 + }, + { + "epoch": 13.680904522613066, + "grad_norm": 0.21203091740608215, + "learning_rate": 1.9163388893967982e-05, + "loss": 0.0093, + "step": 21780 + }, + { + "epoch": 13.68718592964824, + "grad_norm": 0.18270601332187653, + "learning_rate": 1.9120022076356575e-05, + "loss": 0.0074, + "step": 21790 + }, + { + "epoch": 13.693467336683417, + "grad_norm": 0.18825258314609528, + "learning_rate": 1.9076692780830114e-05, + "loss": 0.0107, + "step": 21800 + }, + { + "epoch": 13.699748743718594, + "grad_norm": 0.26772090792655945, + "learning_rate": 1.903340106003782e-05, + "loss": 0.0085, + "step": 21810 + }, + { + "epoch": 13.706030150753769, + "grad_norm": 0.2298189401626587, + "learning_rate": 1.8990146966583183e-05, + "loss": 0.0097, + "step": 21820 + }, + { + "epoch": 13.712311557788945, + "grad_norm": 0.20134997367858887, + "learning_rate": 1.8946930553024035e-05, + "loss": 0.0088, + "step": 21830 + }, + { + "epoch": 13.71859296482412, + "grad_norm": 0.26491275429725647, + "learning_rate": 1.8903751871872378e-05, + "loss": 0.0096, + "step": 21840 + }, + { + "epoch": 13.724874371859297, + "grad_norm": 0.30908843874931335, + "learning_rate": 1.8860610975594382e-05, + "loss": 0.0079, + "step": 21850 + }, + { + "epoch": 13.731155778894472, + "grad_norm": 0.5222078561782837, + "learning_rate": 1.8817507916610307e-05, + "loss": 0.0101, + "step": 21860 + }, + { + "epoch": 13.737437185929648, + "grad_norm": 0.29698488116264343, + "learning_rate": 1.8774442747294407e-05, + "loss": 0.0079, + "step": 21870 + }, + { + "epoch": 13.743718592964823, + "grad_norm": 0.22147132456302643, + "learning_rate": 1.8731415519974965e-05, + "loss": 0.008, + "step": 21880 + }, + { + "epoch": 13.75, + "grad_norm": 0.248886376619339, + "learning_rate": 1.86884262869341e-05, + "loss": 0.0096, + "step": 21890 + }, + { + "epoch": 13.756281407035177, + "grad_norm": 0.17622387409210205, + "learning_rate": 1.8645475100407788e-05, + "loss": 0.0084, + "step": 21900 + }, + { + "epoch": 13.762562814070352, + "grad_norm": 0.24811987578868866, + "learning_rate": 1.8602562012585768e-05, + "loss": 0.0091, + "step": 21910 + }, + { + "epoch": 13.768844221105528, + "grad_norm": 0.1494988203048706, + "learning_rate": 1.8559687075611466e-05, + "loss": 0.009, + "step": 21920 + }, + { + "epoch": 13.775125628140703, + "grad_norm": 0.16652090847492218, + "learning_rate": 1.8516850341582017e-05, + "loss": 0.0073, + "step": 21930 + }, + { + "epoch": 13.78140703517588, + "grad_norm": 0.22236508131027222, + "learning_rate": 1.847405186254807e-05, + "loss": 0.0107, + "step": 21940 + }, + { + "epoch": 13.787688442211055, + "grad_norm": 0.2549535036087036, + "learning_rate": 1.843129169051379e-05, + "loss": 0.0084, + "step": 21950 + }, + { + "epoch": 13.793969849246231, + "grad_norm": 0.19973450899124146, + "learning_rate": 1.8388569877436863e-05, + "loss": 0.0087, + "step": 21960 + }, + { + "epoch": 13.800251256281406, + "grad_norm": 0.17261096835136414, + "learning_rate": 1.834588647522828e-05, + "loss": 0.0081, + "step": 21970 + }, + { + "epoch": 13.806532663316583, + "grad_norm": 0.13793139159679413, + "learning_rate": 1.8303241535752435e-05, + "loss": 0.0057, + "step": 21980 + }, + { + "epoch": 13.81281407035176, + "grad_norm": 0.1833457350730896, + "learning_rate": 1.8260635110826934e-05, + "loss": 0.0133, + "step": 21990 + }, + { + "epoch": 13.819095477386934, + "grad_norm": 0.23032629489898682, + "learning_rate": 1.8218067252222597e-05, + "loss": 0.0094, + "step": 22000 + }, + { + "epoch": 13.825376884422111, + "grad_norm": 0.23317153751850128, + "learning_rate": 1.817553801166339e-05, + "loss": 0.0069, + "step": 22010 + }, + { + "epoch": 13.831658291457286, + "grad_norm": 0.2185666412115097, + "learning_rate": 1.8133047440826333e-05, + "loss": 0.01, + "step": 22020 + }, + { + "epoch": 13.837939698492463, + "grad_norm": 0.35848262906074524, + "learning_rate": 1.809059559134151e-05, + "loss": 0.0096, + "step": 22030 + }, + { + "epoch": 13.844221105527637, + "grad_norm": 0.09857897460460663, + "learning_rate": 1.80481825147919e-05, + "loss": 0.0089, + "step": 22040 + }, + { + "epoch": 13.850502512562814, + "grad_norm": 0.24338160455226898, + "learning_rate": 1.8005808262713396e-05, + "loss": 0.0086, + "step": 22050 + }, + { + "epoch": 13.85678391959799, + "grad_norm": 0.24859832227230072, + "learning_rate": 1.7963472886594714e-05, + "loss": 0.0083, + "step": 22060 + }, + { + "epoch": 13.863065326633166, + "grad_norm": 0.12897834181785583, + "learning_rate": 1.7921176437877302e-05, + "loss": 0.0079, + "step": 22070 + }, + { + "epoch": 13.869346733668342, + "grad_norm": 0.17907065153121948, + "learning_rate": 1.7878918967955366e-05, + "loss": 0.0074, + "step": 22080 + }, + { + "epoch": 13.875628140703517, + "grad_norm": 0.17110517621040344, + "learning_rate": 1.7836700528175693e-05, + "loss": 0.0104, + "step": 22090 + }, + { + "epoch": 13.881909547738694, + "grad_norm": 0.150108203291893, + "learning_rate": 1.779452116983769e-05, + "loss": 0.0097, + "step": 22100 + }, + { + "epoch": 13.888190954773869, + "grad_norm": 0.2080538272857666, + "learning_rate": 1.775238094419325e-05, + "loss": 0.0102, + "step": 22110 + }, + { + "epoch": 13.894472361809045, + "grad_norm": 0.2316233515739441, + "learning_rate": 1.7710279902446708e-05, + "loss": 0.0114, + "step": 22120 + }, + { + "epoch": 13.900753768844222, + "grad_norm": 0.18619416654109955, + "learning_rate": 1.7668218095754795e-05, + "loss": 0.0084, + "step": 22130 + }, + { + "epoch": 13.907035175879397, + "grad_norm": 0.2154015302658081, + "learning_rate": 1.7626195575226594e-05, + "loss": 0.0086, + "step": 22140 + }, + { + "epoch": 13.913316582914574, + "grad_norm": 0.2404315322637558, + "learning_rate": 1.758421239192343e-05, + "loss": 0.0089, + "step": 22150 + }, + { + "epoch": 13.919597989949748, + "grad_norm": 0.1807640939950943, + "learning_rate": 1.7542268596858814e-05, + "loss": 0.0099, + "step": 22160 + }, + { + "epoch": 13.925879396984925, + "grad_norm": 0.21466413140296936, + "learning_rate": 1.750036424099841e-05, + "loss": 0.0093, + "step": 22170 + }, + { + "epoch": 13.9321608040201, + "grad_norm": 0.10164029896259308, + "learning_rate": 1.7458499375259956e-05, + "loss": 0.008, + "step": 22180 + }, + { + "epoch": 13.938442211055277, + "grad_norm": 0.14772890508174896, + "learning_rate": 1.7416674050513245e-05, + "loss": 0.0072, + "step": 22190 + }, + { + "epoch": 13.944723618090451, + "grad_norm": 0.11260966956615448, + "learning_rate": 1.737488831757997e-05, + "loss": 0.0087, + "step": 22200 + }, + { + "epoch": 13.951005025125628, + "grad_norm": 0.15969261527061462, + "learning_rate": 1.733314222723373e-05, + "loss": 0.0075, + "step": 22210 + }, + { + "epoch": 13.957286432160805, + "grad_norm": 0.17739297449588776, + "learning_rate": 1.7291435830199954e-05, + "loss": 0.0073, + "step": 22220 + }, + { + "epoch": 13.96356783919598, + "grad_norm": 0.22389495372772217, + "learning_rate": 1.7249769177155878e-05, + "loss": 0.0102, + "step": 22230 + }, + { + "epoch": 13.969849246231156, + "grad_norm": 0.1677085906267166, + "learning_rate": 1.720814231873038e-05, + "loss": 0.0096, + "step": 22240 + }, + { + "epoch": 13.976130653266331, + "grad_norm": 0.19166694581508636, + "learning_rate": 1.716655530550405e-05, + "loss": 0.0092, + "step": 22250 + }, + { + "epoch": 13.982412060301508, + "grad_norm": 0.1913347840309143, + "learning_rate": 1.7125008188009017e-05, + "loss": 0.0092, + "step": 22260 + }, + { + "epoch": 13.988693467336683, + "grad_norm": 0.24561214447021484, + "learning_rate": 1.708350101672894e-05, + "loss": 0.0094, + "step": 22270 + }, + { + "epoch": 13.99497487437186, + "grad_norm": 0.1777213215827942, + "learning_rate": 1.7042033842098957e-05, + "loss": 0.0081, + "step": 22280 + }, + { + "epoch": 14.001256281407036, + "grad_norm": 0.23033060133457184, + "learning_rate": 1.7000606714505567e-05, + "loss": 0.009, + "step": 22290 + }, + { + "epoch": 14.007537688442211, + "grad_norm": 0.1975415199995041, + "learning_rate": 1.695921968428668e-05, + "loss": 0.007, + "step": 22300 + }, + { + "epoch": 14.013819095477388, + "grad_norm": 0.14465877413749695, + "learning_rate": 1.6917872801731416e-05, + "loss": 0.0075, + "step": 22310 + }, + { + "epoch": 14.020100502512562, + "grad_norm": 0.31168895959854126, + "learning_rate": 1.687656611708014e-05, + "loss": 0.0086, + "step": 22320 + }, + { + "epoch": 14.02638190954774, + "grad_norm": 0.1836441606283188, + "learning_rate": 1.6835299680524368e-05, + "loss": 0.0077, + "step": 22330 + }, + { + "epoch": 14.032663316582914, + "grad_norm": 0.21965697407722473, + "learning_rate": 1.679407354220669e-05, + "loss": 0.0106, + "step": 22340 + }, + { + "epoch": 14.03894472361809, + "grad_norm": 0.14024481177330017, + "learning_rate": 1.6752887752220793e-05, + "loss": 0.0089, + "step": 22350 + }, + { + "epoch": 14.045226130653266, + "grad_norm": 0.2062184065580368, + "learning_rate": 1.6711742360611277e-05, + "loss": 0.0091, + "step": 22360 + }, + { + "epoch": 14.051507537688442, + "grad_norm": 0.19950523972511292, + "learning_rate": 1.6670637417373652e-05, + "loss": 0.0062, + "step": 22370 + }, + { + "epoch": 14.057788944723619, + "grad_norm": 0.22793325781822205, + "learning_rate": 1.6629572972454333e-05, + "loss": 0.0101, + "step": 22380 + }, + { + "epoch": 14.064070351758794, + "grad_norm": 0.1525895893573761, + "learning_rate": 1.6588549075750464e-05, + "loss": 0.01, + "step": 22390 + }, + { + "epoch": 14.07035175879397, + "grad_norm": 0.07370323687791824, + "learning_rate": 1.6547565777109976e-05, + "loss": 0.0079, + "step": 22400 + }, + { + "epoch": 14.076633165829145, + "grad_norm": 0.1854361593723297, + "learning_rate": 1.650662312633143e-05, + "loss": 0.0101, + "step": 22410 + }, + { + "epoch": 14.082914572864322, + "grad_norm": 0.3016980290412903, + "learning_rate": 1.6465721173164002e-05, + "loss": 0.0095, + "step": 22420 + }, + { + "epoch": 14.089195979899497, + "grad_norm": 0.22532515227794647, + "learning_rate": 1.6424859967307428e-05, + "loss": 0.0109, + "step": 22430 + }, + { + "epoch": 14.095477386934673, + "grad_norm": 0.13742215931415558, + "learning_rate": 1.6384039558411903e-05, + "loss": 0.0072, + "step": 22440 + }, + { + "epoch": 14.10175879396985, + "grad_norm": 0.09152957797050476, + "learning_rate": 1.634325999607811e-05, + "loss": 0.006, + "step": 22450 + }, + { + "epoch": 14.108040201005025, + "grad_norm": 0.23711921274662018, + "learning_rate": 1.6302521329857045e-05, + "loss": 0.0114, + "step": 22460 + }, + { + "epoch": 14.114321608040202, + "grad_norm": 0.14304648339748383, + "learning_rate": 1.6261823609250027e-05, + "loss": 0.0102, + "step": 22470 + }, + { + "epoch": 14.120603015075377, + "grad_norm": 0.24246278405189514, + "learning_rate": 1.6221166883708627e-05, + "loss": 0.0082, + "step": 22480 + }, + { + "epoch": 14.126884422110553, + "grad_norm": 0.1678002029657364, + "learning_rate": 1.6180551202634603e-05, + "loss": 0.0089, + "step": 22490 + }, + { + "epoch": 14.133165829145728, + "grad_norm": 0.19520977139472961, + "learning_rate": 1.613997661537981e-05, + "loss": 0.008, + "step": 22500 + }, + { + "epoch": 14.139447236180905, + "grad_norm": 0.26928243041038513, + "learning_rate": 1.609944317124624e-05, + "loss": 0.0097, + "step": 22510 + }, + { + "epoch": 14.14572864321608, + "grad_norm": 0.1712978333234787, + "learning_rate": 1.605895091948582e-05, + "loss": 0.0064, + "step": 22520 + }, + { + "epoch": 14.152010050251256, + "grad_norm": 0.2113756686449051, + "learning_rate": 1.6018499909300477e-05, + "loss": 0.0099, + "step": 22530 + }, + { + "epoch": 14.158291457286433, + "grad_norm": 0.20333488285541534, + "learning_rate": 1.597809018984199e-05, + "loss": 0.0091, + "step": 22540 + }, + { + "epoch": 14.164572864321608, + "grad_norm": 0.1569119095802307, + "learning_rate": 1.593772181021196e-05, + "loss": 0.0069, + "step": 22550 + }, + { + "epoch": 14.170854271356784, + "grad_norm": 0.158197820186615, + "learning_rate": 1.5897394819461813e-05, + "loss": 0.0083, + "step": 22560 + }, + { + "epoch": 14.17713567839196, + "grad_norm": 0.2743305563926697, + "learning_rate": 1.5857109266592625e-05, + "loss": 0.0072, + "step": 22570 + }, + { + "epoch": 14.183417085427136, + "grad_norm": 0.173813134431839, + "learning_rate": 1.581686520055514e-05, + "loss": 0.0071, + "step": 22580 + }, + { + "epoch": 14.18969849246231, + "grad_norm": 0.16120190918445587, + "learning_rate": 1.5776662670249704e-05, + "loss": 0.0092, + "step": 22590 + }, + { + "epoch": 14.195979899497488, + "grad_norm": 0.20933988690376282, + "learning_rate": 1.573650172452615e-05, + "loss": 0.0068, + "step": 22600 + }, + { + "epoch": 14.202261306532664, + "grad_norm": 0.16838395595550537, + "learning_rate": 1.5696382412183852e-05, + "loss": 0.0071, + "step": 22610 + }, + { + "epoch": 14.208542713567839, + "grad_norm": 0.1635112166404724, + "learning_rate": 1.565630478197155e-05, + "loss": 0.0072, + "step": 22620 + }, + { + "epoch": 14.214824120603016, + "grad_norm": 0.14624720811843872, + "learning_rate": 1.5616268882587332e-05, + "loss": 0.0094, + "step": 22630 + }, + { + "epoch": 14.22110552763819, + "grad_norm": 0.22855977714061737, + "learning_rate": 1.5576274762678594e-05, + "loss": 0.0067, + "step": 22640 + }, + { + "epoch": 14.227386934673367, + "grad_norm": 0.23409321904182434, + "learning_rate": 1.5536322470841952e-05, + "loss": 0.0101, + "step": 22650 + }, + { + "epoch": 14.233668341708542, + "grad_norm": 0.15358300507068634, + "learning_rate": 1.549641205562324e-05, + "loss": 0.0072, + "step": 22660 + }, + { + "epoch": 14.239949748743719, + "grad_norm": 0.18265032768249512, + "learning_rate": 1.5456543565517334e-05, + "loss": 0.0088, + "step": 22670 + }, + { + "epoch": 14.246231155778894, + "grad_norm": 0.13590024411678314, + "learning_rate": 1.541671704896825e-05, + "loss": 0.0087, + "step": 22680 + }, + { + "epoch": 14.25251256281407, + "grad_norm": 0.19401074945926666, + "learning_rate": 1.5376932554368938e-05, + "loss": 0.0098, + "step": 22690 + }, + { + "epoch": 14.258793969849247, + "grad_norm": 0.21491461992263794, + "learning_rate": 1.5337190130061318e-05, + "loss": 0.0081, + "step": 22700 + }, + { + "epoch": 14.265075376884422, + "grad_norm": 0.17727750539779663, + "learning_rate": 1.5297489824336165e-05, + "loss": 0.0091, + "step": 22710 + }, + { + "epoch": 14.271356783919598, + "grad_norm": 0.1420249342918396, + "learning_rate": 1.5257831685433121e-05, + "loss": 0.0105, + "step": 22720 + }, + { + "epoch": 14.277638190954773, + "grad_norm": 0.3689980208873749, + "learning_rate": 1.521821576154055e-05, + "loss": 0.0095, + "step": 22730 + }, + { + "epoch": 14.28391959798995, + "grad_norm": 0.19894646108150482, + "learning_rate": 1.5178642100795542e-05, + "loss": 0.0089, + "step": 22740 + }, + { + "epoch": 14.290201005025125, + "grad_norm": 0.271245539188385, + "learning_rate": 1.5139110751283819e-05, + "loss": 0.0091, + "step": 22750 + }, + { + "epoch": 14.296482412060302, + "grad_norm": 0.30372968316078186, + "learning_rate": 1.5099621761039683e-05, + "loss": 0.0073, + "step": 22760 + }, + { + "epoch": 14.302763819095478, + "grad_norm": 0.1501666009426117, + "learning_rate": 1.5060175178046016e-05, + "loss": 0.01, + "step": 22770 + }, + { + "epoch": 14.309045226130653, + "grad_norm": 0.15624839067459106, + "learning_rate": 1.5020771050234117e-05, + "loss": 0.01, + "step": 22780 + }, + { + "epoch": 14.31532663316583, + "grad_norm": 0.21079538762569427, + "learning_rate": 1.4981409425483716e-05, + "loss": 0.0066, + "step": 22790 + }, + { + "epoch": 14.321608040201005, + "grad_norm": 0.1543215811252594, + "learning_rate": 1.4942090351622883e-05, + "loss": 0.0119, + "step": 22800 + }, + { + "epoch": 14.327889447236181, + "grad_norm": 0.16055257618427277, + "learning_rate": 1.490281387642804e-05, + "loss": 0.0079, + "step": 22810 + }, + { + "epoch": 14.334170854271356, + "grad_norm": 0.20959055423736572, + "learning_rate": 1.4863580047623765e-05, + "loss": 0.0091, + "step": 22820 + }, + { + "epoch": 14.340452261306533, + "grad_norm": 0.18951891362667084, + "learning_rate": 1.4824388912882897e-05, + "loss": 0.0074, + "step": 22830 + }, + { + "epoch": 14.346733668341708, + "grad_norm": 0.3136639893054962, + "learning_rate": 1.4785240519826344e-05, + "loss": 0.0077, + "step": 22840 + }, + { + "epoch": 14.353015075376884, + "grad_norm": 0.16106289625167847, + "learning_rate": 1.4746134916023097e-05, + "loss": 0.0082, + "step": 22850 + }, + { + "epoch": 14.359296482412061, + "grad_norm": 0.15605080127716064, + "learning_rate": 1.4707072148990141e-05, + "loss": 0.0072, + "step": 22860 + }, + { + "epoch": 14.365577889447236, + "grad_norm": 0.16072869300842285, + "learning_rate": 1.4668052266192423e-05, + "loss": 0.0093, + "step": 22870 + }, + { + "epoch": 14.371859296482413, + "grad_norm": 0.14859811961650848, + "learning_rate": 1.4629075315042795e-05, + "loss": 0.0062, + "step": 22880 + }, + { + "epoch": 14.378140703517587, + "grad_norm": 0.17262373864650726, + "learning_rate": 1.4590141342901925e-05, + "loss": 0.0094, + "step": 22890 + }, + { + "epoch": 14.384422110552764, + "grad_norm": 0.23533426225185394, + "learning_rate": 1.4551250397078253e-05, + "loss": 0.0088, + "step": 22900 + }, + { + "epoch": 14.390703517587939, + "grad_norm": 0.1811661720275879, + "learning_rate": 1.4512402524827945e-05, + "loss": 0.0078, + "step": 22910 + }, + { + "epoch": 14.396984924623116, + "grad_norm": 0.11159911751747131, + "learning_rate": 1.447359777335482e-05, + "loss": 0.0071, + "step": 22920 + }, + { + "epoch": 14.403266331658292, + "grad_norm": 0.1800864040851593, + "learning_rate": 1.4434836189810335e-05, + "loss": 0.0095, + "step": 22930 + }, + { + "epoch": 14.409547738693467, + "grad_norm": 0.26906973123550415, + "learning_rate": 1.4396117821293453e-05, + "loss": 0.0082, + "step": 22940 + }, + { + "epoch": 14.415829145728644, + "grad_norm": 0.2144717574119568, + "learning_rate": 1.4357442714850633e-05, + "loss": 0.0067, + "step": 22950 + }, + { + "epoch": 14.422110552763819, + "grad_norm": 0.21504764258861542, + "learning_rate": 1.43188109174758e-05, + "loss": 0.0079, + "step": 22960 + }, + { + "epoch": 14.428391959798995, + "grad_norm": 0.1592240333557129, + "learning_rate": 1.4280222476110206e-05, + "loss": 0.0068, + "step": 22970 + }, + { + "epoch": 14.43467336683417, + "grad_norm": 0.19210213422775269, + "learning_rate": 1.4241677437642469e-05, + "loss": 0.0061, + "step": 22980 + }, + { + "epoch": 14.440954773869347, + "grad_norm": 0.1279435008764267, + "learning_rate": 1.420317584890844e-05, + "loss": 0.01, + "step": 22990 + }, + { + "epoch": 14.447236180904522, + "grad_norm": 0.15876562893390656, + "learning_rate": 1.4164717756691176e-05, + "loss": 0.0097, + "step": 23000 + }, + { + "epoch": 14.453517587939698, + "grad_norm": 0.12432048469781876, + "learning_rate": 1.4126303207720882e-05, + "loss": 0.0065, + "step": 23010 + }, + { + "epoch": 14.459798994974875, + "grad_norm": 0.1682938188314438, + "learning_rate": 1.408793224867484e-05, + "loss": 0.0092, + "step": 23020 + }, + { + "epoch": 14.46608040201005, + "grad_norm": 0.22940655052661896, + "learning_rate": 1.4049604926177423e-05, + "loss": 0.0076, + "step": 23030 + }, + { + "epoch": 14.472361809045227, + "grad_norm": 0.18770629167556763, + "learning_rate": 1.4011321286799916e-05, + "loss": 0.0085, + "step": 23040 + }, + { + "epoch": 14.478643216080402, + "grad_norm": 0.1644836664199829, + "learning_rate": 1.3973081377060565e-05, + "loss": 0.0085, + "step": 23050 + }, + { + "epoch": 14.484924623115578, + "grad_norm": 0.32297831773757935, + "learning_rate": 1.393488524342445e-05, + "loss": 0.0115, + "step": 23060 + }, + { + "epoch": 14.491206030150753, + "grad_norm": 0.2930419147014618, + "learning_rate": 1.3896732932303486e-05, + "loss": 0.0087, + "step": 23070 + }, + { + "epoch": 14.49748743718593, + "grad_norm": 0.1194843277335167, + "learning_rate": 1.3858624490056304e-05, + "loss": 0.0092, + "step": 23080 + }, + { + "epoch": 14.503768844221106, + "grad_norm": 0.2369421273469925, + "learning_rate": 1.3820559962988299e-05, + "loss": 0.0093, + "step": 23090 + }, + { + "epoch": 14.510050251256281, + "grad_norm": 0.2115481197834015, + "learning_rate": 1.3782539397351418e-05, + "loss": 0.0094, + "step": 23100 + }, + { + "epoch": 14.516331658291458, + "grad_norm": 0.20541080832481384, + "learning_rate": 1.3744562839344266e-05, + "loss": 0.0094, + "step": 23110 + }, + { + "epoch": 14.522613065326633, + "grad_norm": 0.2074899822473526, + "learning_rate": 1.370663033511193e-05, + "loss": 0.0092, + "step": 23120 + }, + { + "epoch": 14.52889447236181, + "grad_norm": 0.2935747802257538, + "learning_rate": 1.3668741930745964e-05, + "loss": 0.0083, + "step": 23130 + }, + { + "epoch": 14.535175879396984, + "grad_norm": 0.12127149850130081, + "learning_rate": 1.3630897672284382e-05, + "loss": 0.0072, + "step": 23140 + }, + { + "epoch": 14.541457286432161, + "grad_norm": 0.19163572788238525, + "learning_rate": 1.3593097605711508e-05, + "loss": 0.0081, + "step": 23150 + }, + { + "epoch": 14.547738693467338, + "grad_norm": 0.14671045541763306, + "learning_rate": 1.3555341776957992e-05, + "loss": 0.0103, + "step": 23160 + }, + { + "epoch": 14.554020100502512, + "grad_norm": 0.3288838267326355, + "learning_rate": 1.3517630231900724e-05, + "loss": 0.0103, + "step": 23170 + }, + { + "epoch": 14.56030150753769, + "grad_norm": 0.09854491055011749, + "learning_rate": 1.3479963016362767e-05, + "loss": 0.0084, + "step": 23180 + }, + { + "epoch": 14.566582914572864, + "grad_norm": 0.13359981775283813, + "learning_rate": 1.3442340176113377e-05, + "loss": 0.0101, + "step": 23190 + }, + { + "epoch": 14.57286432160804, + "grad_norm": 0.14235183596611023, + "learning_rate": 1.340476175686784e-05, + "loss": 0.0085, + "step": 23200 + }, + { + "epoch": 14.579145728643216, + "grad_norm": 0.08123176544904709, + "learning_rate": 1.3367227804287469e-05, + "loss": 0.0077, + "step": 23210 + }, + { + "epoch": 14.585427135678392, + "grad_norm": 0.1029137596487999, + "learning_rate": 1.332973836397956e-05, + "loss": 0.0097, + "step": 23220 + }, + { + "epoch": 14.591708542713567, + "grad_norm": 0.2829362154006958, + "learning_rate": 1.329229348149731e-05, + "loss": 0.0097, + "step": 23230 + }, + { + "epoch": 14.597989949748744, + "grad_norm": 0.20824337005615234, + "learning_rate": 1.3254893202339796e-05, + "loss": 0.0105, + "step": 23240 + }, + { + "epoch": 14.60427135678392, + "grad_norm": 0.17087659239768982, + "learning_rate": 1.3217537571951872e-05, + "loss": 0.0104, + "step": 23250 + }, + { + "epoch": 14.610552763819095, + "grad_norm": 0.2611936926841736, + "learning_rate": 1.3180226635724169e-05, + "loss": 0.0083, + "step": 23260 + }, + { + "epoch": 14.616834170854272, + "grad_norm": 0.2348429560661316, + "learning_rate": 1.314296043899298e-05, + "loss": 0.011, + "step": 23270 + }, + { + "epoch": 14.623115577889447, + "grad_norm": 0.22585374116897583, + "learning_rate": 1.3105739027040248e-05, + "loss": 0.0098, + "step": 23280 + }, + { + "epoch": 14.629396984924623, + "grad_norm": 0.14362797141075134, + "learning_rate": 1.3068562445093491e-05, + "loss": 0.0076, + "step": 23290 + }, + { + "epoch": 14.635678391959798, + "grad_norm": 0.18760131299495697, + "learning_rate": 1.3031430738325778e-05, + "loss": 0.0078, + "step": 23300 + }, + { + "epoch": 14.641959798994975, + "grad_norm": 0.18779003620147705, + "learning_rate": 1.299434395185563e-05, + "loss": 0.0083, + "step": 23310 + }, + { + "epoch": 14.64824120603015, + "grad_norm": 0.22547022998332977, + "learning_rate": 1.295730213074699e-05, + "loss": 0.0086, + "step": 23320 + }, + { + "epoch": 14.654522613065327, + "grad_norm": 0.09857223927974701, + "learning_rate": 1.2920305320009152e-05, + "loss": 0.0076, + "step": 23330 + }, + { + "epoch": 14.660804020100503, + "grad_norm": 0.29169487953186035, + "learning_rate": 1.2883353564596729e-05, + "loss": 0.0088, + "step": 23340 + }, + { + "epoch": 14.667085427135678, + "grad_norm": 0.136124849319458, + "learning_rate": 1.2846446909409604e-05, + "loss": 0.0069, + "step": 23350 + }, + { + "epoch": 14.673366834170855, + "grad_norm": 0.16442176699638367, + "learning_rate": 1.280958539929284e-05, + "loss": 0.0088, + "step": 23360 + }, + { + "epoch": 14.67964824120603, + "grad_norm": 0.12460197508335114, + "learning_rate": 1.277276907903664e-05, + "loss": 0.0078, + "step": 23370 + }, + { + "epoch": 14.685929648241206, + "grad_norm": 0.20493492484092712, + "learning_rate": 1.2735997993376298e-05, + "loss": 0.0069, + "step": 23380 + }, + { + "epoch": 14.692211055276381, + "grad_norm": 0.15360134840011597, + "learning_rate": 1.2699272186992166e-05, + "loss": 0.0071, + "step": 23390 + }, + { + "epoch": 14.698492462311558, + "grad_norm": 0.16647683084011078, + "learning_rate": 1.2662591704509547e-05, + "loss": 0.0083, + "step": 23400 + }, + { + "epoch": 14.704773869346734, + "grad_norm": 0.3013375401496887, + "learning_rate": 1.2625956590498711e-05, + "loss": 0.0112, + "step": 23410 + }, + { + "epoch": 14.71105527638191, + "grad_norm": 0.12205390632152557, + "learning_rate": 1.2589366889474757e-05, + "loss": 0.009, + "step": 23420 + }, + { + "epoch": 14.717336683417086, + "grad_norm": 0.17225202918052673, + "learning_rate": 1.2552822645897621e-05, + "loss": 0.0072, + "step": 23430 + }, + { + "epoch": 14.72361809045226, + "grad_norm": 0.12665458023548126, + "learning_rate": 1.2516323904172e-05, + "loss": 0.0074, + "step": 23440 + }, + { + "epoch": 14.729899497487438, + "grad_norm": 0.11310118436813354, + "learning_rate": 1.2479870708647324e-05, + "loss": 0.0067, + "step": 23450 + }, + { + "epoch": 14.736180904522612, + "grad_norm": 0.14470739662647247, + "learning_rate": 1.2443463103617658e-05, + "loss": 0.0073, + "step": 23460 + }, + { + "epoch": 14.742462311557789, + "grad_norm": 0.18233442306518555, + "learning_rate": 1.240710113332167e-05, + "loss": 0.0062, + "step": 23470 + }, + { + "epoch": 14.748743718592966, + "grad_norm": 0.2774430811405182, + "learning_rate": 1.2370784841942584e-05, + "loss": 0.0071, + "step": 23480 + }, + { + "epoch": 14.75502512562814, + "grad_norm": 0.26177871227264404, + "learning_rate": 1.2334514273608117e-05, + "loss": 0.0089, + "step": 23490 + }, + { + "epoch": 14.761306532663317, + "grad_norm": 0.20972420275211334, + "learning_rate": 1.2298289472390417e-05, + "loss": 0.0085, + "step": 23500 + }, + { + "epoch": 14.767587939698492, + "grad_norm": 0.22828197479248047, + "learning_rate": 1.226211048230606e-05, + "loss": 0.0093, + "step": 23510 + }, + { + "epoch": 14.773869346733669, + "grad_norm": 0.20946834981441498, + "learning_rate": 1.2225977347315921e-05, + "loss": 0.008, + "step": 23520 + }, + { + "epoch": 14.780150753768844, + "grad_norm": 0.2364979088306427, + "learning_rate": 1.2189890111325148e-05, + "loss": 0.0091, + "step": 23530 + }, + { + "epoch": 14.78643216080402, + "grad_norm": 0.18005070090293884, + "learning_rate": 1.2153848818183161e-05, + "loss": 0.0066, + "step": 23540 + }, + { + "epoch": 14.792713567839195, + "grad_norm": 0.1988871991634369, + "learning_rate": 1.2117853511683507e-05, + "loss": 0.0079, + "step": 23550 + }, + { + "epoch": 14.798994974874372, + "grad_norm": 0.1736476719379425, + "learning_rate": 1.2081904235563906e-05, + "loss": 0.0071, + "step": 23560 + }, + { + "epoch": 14.805276381909549, + "grad_norm": 0.21239282190799713, + "learning_rate": 1.20460010335061e-05, + "loss": 0.0071, + "step": 23570 + }, + { + "epoch": 14.811557788944723, + "grad_norm": 0.1585923284292221, + "learning_rate": 1.2010143949135865e-05, + "loss": 0.0072, + "step": 23580 + }, + { + "epoch": 14.8178391959799, + "grad_norm": 0.16761064529418945, + "learning_rate": 1.1974333026022938e-05, + "loss": 0.0066, + "step": 23590 + }, + { + "epoch": 14.824120603015075, + "grad_norm": 0.23883384466171265, + "learning_rate": 1.1938568307680964e-05, + "loss": 0.0082, + "step": 23600 + }, + { + "epoch": 14.830402010050252, + "grad_norm": 0.23068825900554657, + "learning_rate": 1.1902849837567464e-05, + "loss": 0.0095, + "step": 23610 + }, + { + "epoch": 14.836683417085426, + "grad_norm": 0.25647565722465515, + "learning_rate": 1.1867177659083739e-05, + "loss": 0.0088, + "step": 23620 + }, + { + "epoch": 14.842964824120603, + "grad_norm": 0.1751037985086441, + "learning_rate": 1.1831551815574848e-05, + "loss": 0.0108, + "step": 23630 + }, + { + "epoch": 14.849246231155778, + "grad_norm": 0.18252703547477722, + "learning_rate": 1.1795972350329554e-05, + "loss": 0.0083, + "step": 23640 + }, + { + "epoch": 14.855527638190955, + "grad_norm": 0.1709042489528656, + "learning_rate": 1.176043930658024e-05, + "loss": 0.0086, + "step": 23650 + }, + { + "epoch": 14.861809045226131, + "grad_norm": 0.2817991077899933, + "learning_rate": 1.1724952727502942e-05, + "loss": 0.0092, + "step": 23660 + }, + { + "epoch": 14.868090452261306, + "grad_norm": 0.1620395928621292, + "learning_rate": 1.1689512656217178e-05, + "loss": 0.0127, + "step": 23670 + }, + { + "epoch": 14.874371859296483, + "grad_norm": 0.2416924387216568, + "learning_rate": 1.1654119135785962e-05, + "loss": 0.0093, + "step": 23680 + }, + { + "epoch": 14.880653266331658, + "grad_norm": 0.1557752937078476, + "learning_rate": 1.1618772209215794e-05, + "loss": 0.0068, + "step": 23690 + }, + { + "epoch": 14.886934673366834, + "grad_norm": 0.11335250735282898, + "learning_rate": 1.1583471919456506e-05, + "loss": 0.0087, + "step": 23700 + }, + { + "epoch": 14.89321608040201, + "grad_norm": 0.11056298017501831, + "learning_rate": 1.1548218309401266e-05, + "loss": 0.0086, + "step": 23710 + }, + { + "epoch": 14.899497487437186, + "grad_norm": 0.22596877813339233, + "learning_rate": 1.1513011421886555e-05, + "loss": 0.0057, + "step": 23720 + }, + { + "epoch": 14.905778894472363, + "grad_norm": 0.2708052098751068, + "learning_rate": 1.1477851299692056e-05, + "loss": 0.0084, + "step": 23730 + }, + { + "epoch": 14.912060301507537, + "grad_norm": 0.14088724553585052, + "learning_rate": 1.144273798554063e-05, + "loss": 0.0144, + "step": 23740 + }, + { + "epoch": 14.918341708542714, + "grad_norm": 0.20620734989643097, + "learning_rate": 1.1407671522098262e-05, + "loss": 0.0065, + "step": 23750 + }, + { + "epoch": 14.924623115577889, + "grad_norm": 0.23270264267921448, + "learning_rate": 1.1372651951974e-05, + "loss": 0.007, + "step": 23760 + }, + { + "epoch": 14.930904522613066, + "grad_norm": 0.37136101722717285, + "learning_rate": 1.1337679317719952e-05, + "loss": 0.0099, + "step": 23770 + }, + { + "epoch": 14.93718592964824, + "grad_norm": 0.2549695074558258, + "learning_rate": 1.1302753661831151e-05, + "loss": 0.0103, + "step": 23780 + }, + { + "epoch": 14.943467336683417, + "grad_norm": 0.25638848543167114, + "learning_rate": 1.1267875026745561e-05, + "loss": 0.007, + "step": 23790 + }, + { + "epoch": 14.949748743718594, + "grad_norm": 0.12494678795337677, + "learning_rate": 1.1233043454844016e-05, + "loss": 0.0098, + "step": 23800 + }, + { + "epoch": 14.956030150753769, + "grad_norm": 0.1628507822751999, + "learning_rate": 1.1198258988450144e-05, + "loss": 0.0083, + "step": 23810 + }, + { + "epoch": 14.962311557788945, + "grad_norm": 0.15940187871456146, + "learning_rate": 1.1163521669830368e-05, + "loss": 0.0068, + "step": 23820 + }, + { + "epoch": 14.96859296482412, + "grad_norm": 0.21312926709651947, + "learning_rate": 1.1128831541193818e-05, + "loss": 0.0097, + "step": 23830 + }, + { + "epoch": 14.974874371859297, + "grad_norm": 0.18845684826374054, + "learning_rate": 1.1094188644692255e-05, + "loss": 0.0082, + "step": 23840 + }, + { + "epoch": 14.981155778894472, + "grad_norm": 0.14442896842956543, + "learning_rate": 1.1059593022420067e-05, + "loss": 0.0078, + "step": 23850 + }, + { + "epoch": 14.987437185929648, + "grad_norm": 0.22065645456314087, + "learning_rate": 1.1025044716414184e-05, + "loss": 0.0134, + "step": 23860 + }, + { + "epoch": 14.993718592964823, + "grad_norm": 0.26563501358032227, + "learning_rate": 1.0990543768654082e-05, + "loss": 0.0093, + "step": 23870 + }, + { + "epoch": 15.0, + "grad_norm": 0.18131612241268158, + "learning_rate": 1.0956090221061655e-05, + "loss": 0.008, + "step": 23880 + }, + { + "epoch": 15.006281407035177, + "grad_norm": 0.2505190372467041, + "learning_rate": 1.0921684115501208e-05, + "loss": 0.0116, + "step": 23890 + }, + { + "epoch": 15.012562814070352, + "grad_norm": 0.23517662286758423, + "learning_rate": 1.0887325493779405e-05, + "loss": 0.0101, + "step": 23900 + }, + { + "epoch": 15.018844221105528, + "grad_norm": 0.18309162557125092, + "learning_rate": 1.0853014397645207e-05, + "loss": 0.0076, + "step": 23910 + }, + { + "epoch": 15.025125628140703, + "grad_norm": 0.20102734863758087, + "learning_rate": 1.0818750868789829e-05, + "loss": 0.0059, + "step": 23920 + }, + { + "epoch": 15.03140703517588, + "grad_norm": 0.35064417123794556, + "learning_rate": 1.0784534948846703e-05, + "loss": 0.0094, + "step": 23930 + }, + { + "epoch": 15.037688442211055, + "grad_norm": 0.1550222486257553, + "learning_rate": 1.0750366679391393e-05, + "loss": 0.0099, + "step": 23940 + }, + { + "epoch": 15.043969849246231, + "grad_norm": 0.2603715658187866, + "learning_rate": 1.0716246101941557e-05, + "loss": 0.0075, + "step": 23950 + }, + { + "epoch": 15.050251256281408, + "grad_norm": 0.22049209475517273, + "learning_rate": 1.0682173257956934e-05, + "loss": 0.0086, + "step": 23960 + }, + { + "epoch": 15.056532663316583, + "grad_norm": 0.259227454662323, + "learning_rate": 1.064814818883922e-05, + "loss": 0.0078, + "step": 23970 + }, + { + "epoch": 15.06281407035176, + "grad_norm": 0.07906452566385269, + "learning_rate": 1.06141709359321e-05, + "loss": 0.0066, + "step": 23980 + }, + { + "epoch": 15.069095477386934, + "grad_norm": 0.3878621459007263, + "learning_rate": 1.058024154052114e-05, + "loss": 0.0074, + "step": 23990 + }, + { + "epoch": 15.075376884422111, + "grad_norm": 0.15069842338562012, + "learning_rate": 1.0546360043833747e-05, + "loss": 0.0085, + "step": 24000 + }, + { + "epoch": 15.081658291457286, + "grad_norm": 0.19329088926315308, + "learning_rate": 1.0512526487039138e-05, + "loss": 0.0084, + "step": 24010 + }, + { + "epoch": 15.087939698492463, + "grad_norm": 0.14829577505588531, + "learning_rate": 1.0478740911248258e-05, + "loss": 0.0107, + "step": 24020 + }, + { + "epoch": 15.094221105527637, + "grad_norm": 0.18873737752437592, + "learning_rate": 1.0445003357513793e-05, + "loss": 0.0093, + "step": 24030 + }, + { + "epoch": 15.100502512562814, + "grad_norm": 0.11490311473608017, + "learning_rate": 1.0411313866830042e-05, + "loss": 0.009, + "step": 24040 + }, + { + "epoch": 15.10678391959799, + "grad_norm": 0.18201330304145813, + "learning_rate": 1.0377672480132915e-05, + "loss": 0.0079, + "step": 24050 + }, + { + "epoch": 15.113065326633166, + "grad_norm": 0.16868893802165985, + "learning_rate": 1.0344079238299864e-05, + "loss": 0.0063, + "step": 24060 + }, + { + "epoch": 15.119346733668342, + "grad_norm": 0.14949141442775726, + "learning_rate": 1.0310534182149833e-05, + "loss": 0.0078, + "step": 24070 + }, + { + "epoch": 15.125628140703517, + "grad_norm": 0.24981363117694855, + "learning_rate": 1.0277037352443258e-05, + "loss": 0.0065, + "step": 24080 + }, + { + "epoch": 15.131909547738694, + "grad_norm": 0.1972152441740036, + "learning_rate": 1.024358878988193e-05, + "loss": 0.0072, + "step": 24090 + }, + { + "epoch": 15.138190954773869, + "grad_norm": 0.19345630705356598, + "learning_rate": 1.0210188535108995e-05, + "loss": 0.0102, + "step": 24100 + }, + { + "epoch": 15.144472361809045, + "grad_norm": 0.3728492259979248, + "learning_rate": 1.0176836628708935e-05, + "loss": 0.0066, + "step": 24110 + }, + { + "epoch": 15.150753768844222, + "grad_norm": 0.17253273725509644, + "learning_rate": 1.0143533111207454e-05, + "loss": 0.0064, + "step": 24120 + }, + { + "epoch": 15.157035175879397, + "grad_norm": 0.07921869307756424, + "learning_rate": 1.0110278023071446e-05, + "loss": 0.0087, + "step": 24130 + }, + { + "epoch": 15.163316582914574, + "grad_norm": 0.11405930668115616, + "learning_rate": 1.007707140470901e-05, + "loss": 0.0071, + "step": 24140 + }, + { + "epoch": 15.169597989949748, + "grad_norm": 0.11556612700223923, + "learning_rate": 1.0043913296469298e-05, + "loss": 0.007, + "step": 24150 + }, + { + "epoch": 15.175879396984925, + "grad_norm": 0.27301546931266785, + "learning_rate": 1.001080373864255e-05, + "loss": 0.0072, + "step": 24160 + }, + { + "epoch": 15.1821608040201, + "grad_norm": 0.12504318356513977, + "learning_rate": 9.977742771459991e-06, + "loss": 0.0088, + "step": 24170 + }, + { + "epoch": 15.188442211055277, + "grad_norm": 0.1287487894296646, + "learning_rate": 9.944730435093802e-06, + "loss": 0.0075, + "step": 24180 + }, + { + "epoch": 15.194723618090451, + "grad_norm": 0.23840072751045227, + "learning_rate": 9.911766769657115e-06, + "loss": 0.0074, + "step": 24190 + }, + { + "epoch": 15.201005025125628, + "grad_norm": 0.15411531925201416, + "learning_rate": 9.878851815203882e-06, + "loss": 0.0072, + "step": 24200 + }, + { + "epoch": 15.207286432160805, + "grad_norm": 0.2145812064409256, + "learning_rate": 9.845985611728864e-06, + "loss": 0.0068, + "step": 24210 + }, + { + "epoch": 15.21356783919598, + "grad_norm": 0.13352671265602112, + "learning_rate": 9.813168199167606e-06, + "loss": 0.0074, + "step": 24220 + }, + { + "epoch": 15.219849246231156, + "grad_norm": 0.14208027720451355, + "learning_rate": 9.780399617396341e-06, + "loss": 0.0089, + "step": 24230 + }, + { + "epoch": 15.226130653266331, + "grad_norm": 0.19726422429084778, + "learning_rate": 9.747679906232016e-06, + "loss": 0.0081, + "step": 24240 + }, + { + "epoch": 15.232412060301508, + "grad_norm": 0.18080618977546692, + "learning_rate": 9.715009105432138e-06, + "loss": 0.0084, + "step": 24250 + }, + { + "epoch": 15.238693467336683, + "grad_norm": 0.1420256346464157, + "learning_rate": 9.682387254694835e-06, + "loss": 0.008, + "step": 24260 + }, + { + "epoch": 15.24497487437186, + "grad_norm": 0.15970851480960846, + "learning_rate": 9.649814393658723e-06, + "loss": 0.0072, + "step": 24270 + }, + { + "epoch": 15.251256281407034, + "grad_norm": 0.08175116032361984, + "learning_rate": 9.617290561902881e-06, + "loss": 0.0067, + "step": 24280 + }, + { + "epoch": 15.257537688442211, + "grad_norm": 0.22266583144664764, + "learning_rate": 9.584815798946861e-06, + "loss": 0.0067, + "step": 24290 + }, + { + "epoch": 15.263819095477388, + "grad_norm": 0.14154142141342163, + "learning_rate": 9.55239014425055e-06, + "loss": 0.0081, + "step": 24300 + }, + { + "epoch": 15.270100502512562, + "grad_norm": 0.11802668869495392, + "learning_rate": 9.520013637214176e-06, + "loss": 0.0067, + "step": 24310 + }, + { + "epoch": 15.27638190954774, + "grad_norm": 0.2653592824935913, + "learning_rate": 9.48768631717824e-06, + "loss": 0.008, + "step": 24320 + }, + { + "epoch": 15.282663316582914, + "grad_norm": 0.26394879817962646, + "learning_rate": 9.455408223423496e-06, + "loss": 0.0102, + "step": 24330 + }, + { + "epoch": 15.28894472361809, + "grad_norm": 0.09666930884122849, + "learning_rate": 9.423179395170845e-06, + "loss": 0.0057, + "step": 24340 + }, + { + "epoch": 15.295226130653266, + "grad_norm": 0.16300144791603088, + "learning_rate": 9.390999871581391e-06, + "loss": 0.0073, + "step": 24350 + }, + { + "epoch": 15.301507537688442, + "grad_norm": 0.11088357120752335, + "learning_rate": 9.358869691756273e-06, + "loss": 0.0064, + "step": 24360 + }, + { + "epoch": 15.307788944723619, + "grad_norm": 0.18946921825408936, + "learning_rate": 9.326788894736688e-06, + "loss": 0.0068, + "step": 24370 + }, + { + "epoch": 15.314070351758794, + "grad_norm": 0.23581817746162415, + "learning_rate": 9.294757519503811e-06, + "loss": 0.0078, + "step": 24380 + }, + { + "epoch": 15.32035175879397, + "grad_norm": 0.17614004015922546, + "learning_rate": 9.262775604978819e-06, + "loss": 0.0069, + "step": 24390 + }, + { + "epoch": 15.326633165829145, + "grad_norm": 0.17691151797771454, + "learning_rate": 9.230843190022725e-06, + "loss": 0.0091, + "step": 24400 + }, + { + "epoch": 15.332914572864322, + "grad_norm": 0.16105109453201294, + "learning_rate": 9.198960313436445e-06, + "loss": 0.0093, + "step": 24410 + }, + { + "epoch": 15.339195979899497, + "grad_norm": 0.17465798556804657, + "learning_rate": 9.16712701396067e-06, + "loss": 0.0069, + "step": 24420 + }, + { + "epoch": 15.345477386934673, + "grad_norm": 0.19773881137371063, + "learning_rate": 9.135343330275864e-06, + "loss": 0.0088, + "step": 24430 + }, + { + "epoch": 15.35175879396985, + "grad_norm": 0.20766718685626984, + "learning_rate": 9.10360930100218e-06, + "loss": 0.0104, + "step": 24440 + }, + { + "epoch": 15.358040201005025, + "grad_norm": 0.25664401054382324, + "learning_rate": 9.07192496469949e-06, + "loss": 0.0055, + "step": 24450 + }, + { + "epoch": 15.364321608040202, + "grad_norm": 0.1605786681175232, + "learning_rate": 9.040290359867231e-06, + "loss": 0.0075, + "step": 24460 + }, + { + "epoch": 15.370603015075377, + "grad_norm": 0.15057553350925446, + "learning_rate": 9.008705524944438e-06, + "loss": 0.0074, + "step": 24470 + }, + { + "epoch": 15.376884422110553, + "grad_norm": 0.141757532954216, + "learning_rate": 8.977170498309651e-06, + "loss": 0.008, + "step": 24480 + }, + { + "epoch": 15.383165829145728, + "grad_norm": 0.10183164477348328, + "learning_rate": 8.945685318280916e-06, + "loss": 0.0079, + "step": 24490 + }, + { + "epoch": 15.389447236180905, + "grad_norm": 0.1961367428302765, + "learning_rate": 8.914250023115672e-06, + "loss": 0.0065, + "step": 24500 + }, + { + "epoch": 15.39572864321608, + "grad_norm": 0.14510947465896606, + "learning_rate": 8.882864651010798e-06, + "loss": 0.0054, + "step": 24510 + }, + { + "epoch": 15.402010050251256, + "grad_norm": 0.22071616351604462, + "learning_rate": 8.851529240102462e-06, + "loss": 0.0059, + "step": 24520 + }, + { + "epoch": 15.408291457286433, + "grad_norm": 0.19871842861175537, + "learning_rate": 8.820243828466134e-06, + "loss": 0.0062, + "step": 24530 + }, + { + "epoch": 15.414572864321608, + "grad_norm": 0.13148047029972076, + "learning_rate": 8.789008454116566e-06, + "loss": 0.0082, + "step": 24540 + }, + { + "epoch": 15.420854271356784, + "grad_norm": 0.18991614878177643, + "learning_rate": 8.757823155007655e-06, + "loss": 0.0092, + "step": 24550 + }, + { + "epoch": 15.42713567839196, + "grad_norm": 0.16937746107578278, + "learning_rate": 8.72668796903251e-06, + "loss": 0.0078, + "step": 24560 + }, + { + "epoch": 15.433417085427136, + "grad_norm": 0.22226227819919586, + "learning_rate": 8.695602934023301e-06, + "loss": 0.0076, + "step": 24570 + }, + { + "epoch": 15.43969849246231, + "grad_norm": 0.1664494425058365, + "learning_rate": 8.664568087751274e-06, + "loss": 0.0048, + "step": 24580 + }, + { + "epoch": 15.445979899497488, + "grad_norm": 0.1078210100531578, + "learning_rate": 8.633583467926697e-06, + "loss": 0.0059, + "step": 24590 + }, + { + "epoch": 15.452261306532664, + "grad_norm": 0.1698831170797348, + "learning_rate": 8.602649112198796e-06, + "loss": 0.0069, + "step": 24600 + }, + { + "epoch": 15.458542713567839, + "grad_norm": 0.1881914734840393, + "learning_rate": 8.571765058155745e-06, + "loss": 0.0119, + "step": 24610 + }, + { + "epoch": 15.464824120603016, + "grad_norm": 0.17776870727539062, + "learning_rate": 8.540931343324583e-06, + "loss": 0.0103, + "step": 24620 + }, + { + "epoch": 15.47110552763819, + "grad_norm": 0.20906959474086761, + "learning_rate": 8.51014800517117e-06, + "loss": 0.0074, + "step": 24630 + }, + { + "epoch": 15.477386934673367, + "grad_norm": 0.2740943729877472, + "learning_rate": 8.479415081100167e-06, + "loss": 0.0078, + "step": 24640 + }, + { + "epoch": 15.483668341708542, + "grad_norm": 0.23778663575649261, + "learning_rate": 8.448732608454967e-06, + "loss": 0.0069, + "step": 24650 + }, + { + "epoch": 15.489949748743719, + "grad_norm": 0.1284702867269516, + "learning_rate": 8.418100624517688e-06, + "loss": 0.0064, + "step": 24660 + }, + { + "epoch": 15.496231155778894, + "grad_norm": 0.12934795022010803, + "learning_rate": 8.387519166509062e-06, + "loss": 0.0063, + "step": 24670 + }, + { + "epoch": 15.50251256281407, + "grad_norm": 0.2117137908935547, + "learning_rate": 8.356988271588445e-06, + "loss": 0.0074, + "step": 24680 + }, + { + "epoch": 15.508793969849247, + "grad_norm": 0.12615744769573212, + "learning_rate": 8.326507976853765e-06, + "loss": 0.008, + "step": 24690 + }, + { + "epoch": 15.515075376884422, + "grad_norm": 0.1934594064950943, + "learning_rate": 8.296078319341443e-06, + "loss": 0.0088, + "step": 24700 + }, + { + "epoch": 15.521356783919598, + "grad_norm": 0.1561375856399536, + "learning_rate": 8.265699336026383e-06, + "loss": 0.0087, + "step": 24710 + }, + { + "epoch": 15.527638190954773, + "grad_norm": 0.1701745092868805, + "learning_rate": 8.235371063821923e-06, + "loss": 0.0059, + "step": 24720 + }, + { + "epoch": 15.53391959798995, + "grad_norm": 0.20079082250595093, + "learning_rate": 8.205093539579767e-06, + "loss": 0.0076, + "step": 24730 + }, + { + "epoch": 15.540201005025125, + "grad_norm": 0.18819697201251984, + "learning_rate": 8.174866800089963e-06, + "loss": 0.0062, + "step": 24740 + }, + { + "epoch": 15.546482412060302, + "grad_norm": 0.10381215810775757, + "learning_rate": 8.144690882080853e-06, + "loss": 0.0048, + "step": 24750 + }, + { + "epoch": 15.552763819095478, + "grad_norm": 0.1316290944814682, + "learning_rate": 8.114565822219006e-06, + "loss": 0.0068, + "step": 24760 + }, + { + "epoch": 15.559045226130653, + "grad_norm": 0.19433611631393433, + "learning_rate": 8.084491657109233e-06, + "loss": 0.0052, + "step": 24770 + }, + { + "epoch": 15.56532663316583, + "grad_norm": 0.08581721037626266, + "learning_rate": 8.054468423294476e-06, + "loss": 0.0046, + "step": 24780 + }, + { + "epoch": 15.571608040201005, + "grad_norm": 0.13920317590236664, + "learning_rate": 8.024496157255785e-06, + "loss": 0.0075, + "step": 24790 + }, + { + "epoch": 15.577889447236181, + "grad_norm": 0.12991651892662048, + "learning_rate": 7.994574895412293e-06, + "loss": 0.0066, + "step": 24800 + }, + { + "epoch": 15.584170854271356, + "grad_norm": 0.1365784853696823, + "learning_rate": 7.964704674121149e-06, + "loss": 0.0079, + "step": 24810 + }, + { + "epoch": 15.590452261306533, + "grad_norm": 0.17265000939369202, + "learning_rate": 7.934885529677505e-06, + "loss": 0.0074, + "step": 24820 + }, + { + "epoch": 15.596733668341708, + "grad_norm": 0.11120668053627014, + "learning_rate": 7.905117498314412e-06, + "loss": 0.0054, + "step": 24830 + }, + { + "epoch": 15.603015075376884, + "grad_norm": 0.19741210341453552, + "learning_rate": 7.87540061620286e-06, + "loss": 0.006, + "step": 24840 + }, + { + "epoch": 15.609296482412061, + "grad_norm": 0.08985383808612823, + "learning_rate": 7.845734919451647e-06, + "loss": 0.0068, + "step": 24850 + }, + { + "epoch": 15.615577889447236, + "grad_norm": 0.13575077056884766, + "learning_rate": 7.816120444107383e-06, + "loss": 0.0096, + "step": 24860 + }, + { + "epoch": 15.621859296482413, + "grad_norm": 0.23509441316127777, + "learning_rate": 7.786557226154473e-06, + "loss": 0.0076, + "step": 24870 + }, + { + "epoch": 15.628140703517587, + "grad_norm": 0.1140216812491417, + "learning_rate": 7.757045301514998e-06, + "loss": 0.005, + "step": 24880 + }, + { + "epoch": 15.634422110552764, + "grad_norm": 0.22186033427715302, + "learning_rate": 7.727584706048735e-06, + "loss": 0.0064, + "step": 24890 + }, + { + "epoch": 15.640703517587939, + "grad_norm": 0.2359655201435089, + "learning_rate": 7.698175475553077e-06, + "loss": 0.0069, + "step": 24900 + }, + { + "epoch": 15.646984924623116, + "grad_norm": 0.2936978340148926, + "learning_rate": 7.668817645763021e-06, + "loss": 0.0069, + "step": 24910 + }, + { + "epoch": 15.653266331658292, + "grad_norm": 0.17581194639205933, + "learning_rate": 7.639511252351088e-06, + "loss": 0.0051, + "step": 24920 + }, + { + "epoch": 15.659547738693467, + "grad_norm": 0.19635511934757233, + "learning_rate": 7.610256330927324e-06, + "loss": 0.0063, + "step": 24930 + }, + { + "epoch": 15.665829145728644, + "grad_norm": 0.3674083352088928, + "learning_rate": 7.581052917039211e-06, + "loss": 0.0061, + "step": 24940 + }, + { + "epoch": 15.672110552763819, + "grad_norm": 0.20472627878189087, + "learning_rate": 7.551901046171645e-06, + "loss": 0.011, + "step": 24950 + }, + { + "epoch": 15.678391959798995, + "grad_norm": 0.19781334698200226, + "learning_rate": 7.522800753746895e-06, + "loss": 0.0081, + "step": 24960 + }, + { + "epoch": 15.68467336683417, + "grad_norm": 0.0763436108827591, + "learning_rate": 7.493752075124577e-06, + "loss": 0.0073, + "step": 24970 + }, + { + "epoch": 15.690954773869347, + "grad_norm": 0.25284478068351746, + "learning_rate": 7.464755045601557e-06, + "loss": 0.0062, + "step": 24980 + }, + { + "epoch": 15.697236180904522, + "grad_norm": 0.17930731177330017, + "learning_rate": 7.435809700411972e-06, + "loss": 0.0133, + "step": 24990 + }, + { + "epoch": 15.703517587939698, + "grad_norm": 0.1355779618024826, + "learning_rate": 7.40691607472715e-06, + "loss": 0.0056, + "step": 25000 + }, + { + "epoch": 15.709798994974875, + "grad_norm": 0.08358604460954666, + "learning_rate": 7.3780742036555606e-06, + "loss": 0.006, + "step": 25010 + }, + { + "epoch": 15.71608040201005, + "grad_norm": 0.12572267651557922, + "learning_rate": 7.349284122242783e-06, + "loss": 0.0106, + "step": 25020 + }, + { + "epoch": 15.722361809045227, + "grad_norm": 0.22967395186424255, + "learning_rate": 7.320545865471512e-06, + "loss": 0.0072, + "step": 25030 + }, + { + "epoch": 15.728643216080402, + "grad_norm": 0.16845041513442993, + "learning_rate": 7.291859468261425e-06, + "loss": 0.0061, + "step": 25040 + }, + { + "epoch": 15.734924623115578, + "grad_norm": 0.1116507425904274, + "learning_rate": 7.263224965469195e-06, + "loss": 0.0063, + "step": 25050 + }, + { + "epoch": 15.741206030150753, + "grad_norm": 0.1517229825258255, + "learning_rate": 7.234642391888446e-06, + "loss": 0.0065, + "step": 25060 + }, + { + "epoch": 15.74748743718593, + "grad_norm": 0.19908449053764343, + "learning_rate": 7.206111782249697e-06, + "loss": 0.0085, + "step": 25070 + }, + { + "epoch": 15.753768844221106, + "grad_norm": 0.11468596756458282, + "learning_rate": 7.1776331712203385e-06, + "loss": 0.0078, + "step": 25080 + }, + { + "epoch": 15.760050251256281, + "grad_norm": 0.31807446479797363, + "learning_rate": 7.149206593404561e-06, + "loss": 0.0055, + "step": 25090 + }, + { + "epoch": 15.766331658291458, + "grad_norm": 0.2501654028892517, + "learning_rate": 7.120832083343337e-06, + "loss": 0.0063, + "step": 25100 + }, + { + "epoch": 15.772613065326633, + "grad_norm": 0.28322190046310425, + "learning_rate": 7.092509675514369e-06, + "loss": 0.0087, + "step": 25110 + }, + { + "epoch": 15.77889447236181, + "grad_norm": 0.1790676862001419, + "learning_rate": 7.064239404332063e-06, + "loss": 0.0071, + "step": 25120 + }, + { + "epoch": 15.785175879396984, + "grad_norm": 0.23174069821834564, + "learning_rate": 7.03602130414745e-06, + "loss": 0.0099, + "step": 25130 + }, + { + "epoch": 15.791457286432161, + "grad_norm": 0.08796414732933044, + "learning_rate": 7.007855409248198e-06, + "loss": 0.0059, + "step": 25140 + }, + { + "epoch": 15.797738693467338, + "grad_norm": 0.15270297229290009, + "learning_rate": 6.979741753858521e-06, + "loss": 0.0075, + "step": 25150 + }, + { + "epoch": 15.804020100502512, + "grad_norm": 0.179665207862854, + "learning_rate": 6.951680372139158e-06, + "loss": 0.009, + "step": 25160 + }, + { + "epoch": 15.81030150753769, + "grad_norm": 0.16492989659309387, + "learning_rate": 6.923671298187334e-06, + "loss": 0.0091, + "step": 25170 + }, + { + "epoch": 15.816582914572864, + "grad_norm": 0.1613747477531433, + "learning_rate": 6.895714566036704e-06, + "loss": 0.0081, + "step": 25180 + }, + { + "epoch": 15.82286432160804, + "grad_norm": 0.26840609312057495, + "learning_rate": 6.86781020965736e-06, + "loss": 0.0056, + "step": 25190 + }, + { + "epoch": 15.829145728643216, + "grad_norm": 0.08163484930992126, + "learning_rate": 6.839958262955709e-06, + "loss": 0.0069, + "step": 25200 + }, + { + "epoch": 15.835427135678392, + "grad_norm": 0.14107710123062134, + "learning_rate": 6.8121587597744894e-06, + "loss": 0.0058, + "step": 25210 + }, + { + "epoch": 15.841708542713567, + "grad_norm": 0.16026955842971802, + "learning_rate": 6.784411733892732e-06, + "loss": 0.006, + "step": 25220 + }, + { + "epoch": 15.847989949748744, + "grad_norm": 1.786471962928772, + "learning_rate": 6.756717219025666e-06, + "loss": 0.0071, + "step": 25230 + }, + { + "epoch": 15.85427135678392, + "grad_norm": 0.1400795727968216, + "learning_rate": 6.729075248824762e-06, + "loss": 0.0073, + "step": 25240 + }, + { + "epoch": 15.860552763819095, + "grad_norm": 0.21076767146587372, + "learning_rate": 6.701485856877615e-06, + "loss": 0.0051, + "step": 25250 + }, + { + "epoch": 15.866834170854272, + "grad_norm": 0.11991211771965027, + "learning_rate": 6.673949076707925e-06, + "loss": 0.0094, + "step": 25260 + }, + { + "epoch": 15.873115577889447, + "grad_norm": 0.10134661942720413, + "learning_rate": 6.646464941775499e-06, + "loss": 0.0071, + "step": 25270 + }, + { + "epoch": 15.879396984924623, + "grad_norm": 0.2506278157234192, + "learning_rate": 6.619033485476128e-06, + "loss": 0.009, + "step": 25280 + }, + { + "epoch": 15.885678391959798, + "grad_norm": 0.2198912501335144, + "learning_rate": 6.59165474114164e-06, + "loss": 0.0084, + "step": 25290 + }, + { + "epoch": 15.891959798994975, + "grad_norm": 0.17705994844436646, + "learning_rate": 6.564328742039783e-06, + "loss": 0.0062, + "step": 25300 + }, + { + "epoch": 15.89824120603015, + "grad_norm": 0.25665295124053955, + "learning_rate": 6.537055521374219e-06, + "loss": 0.0059, + "step": 25310 + }, + { + "epoch": 15.904522613065327, + "grad_norm": 0.14017704129219055, + "learning_rate": 6.509835112284485e-06, + "loss": 0.0075, + "step": 25320 + }, + { + "epoch": 15.910804020100503, + "grad_norm": 0.26687127351760864, + "learning_rate": 6.482667547845944e-06, + "loss": 0.0081, + "step": 25330 + }, + { + "epoch": 15.917085427135678, + "grad_norm": 0.25596874952316284, + "learning_rate": 6.455552861069736e-06, + "loss": 0.0076, + "step": 25340 + }, + { + "epoch": 15.923366834170855, + "grad_norm": 0.3371482789516449, + "learning_rate": 6.428491084902788e-06, + "loss": 0.0059, + "step": 25350 + }, + { + "epoch": 15.92964824120603, + "grad_norm": 0.28802546858787537, + "learning_rate": 6.401482252227698e-06, + "loss": 0.0077, + "step": 25360 + }, + { + "epoch": 15.935929648241206, + "grad_norm": 0.1493028849363327, + "learning_rate": 6.37452639586274e-06, + "loss": 0.0053, + "step": 25370 + }, + { + "epoch": 15.942211055276381, + "grad_norm": 0.17558130621910095, + "learning_rate": 6.347623548561826e-06, + "loss": 0.0082, + "step": 25380 + }, + { + "epoch": 15.948492462311558, + "grad_norm": 0.1258350908756256, + "learning_rate": 6.320773743014441e-06, + "loss": 0.0083, + "step": 25390 + }, + { + "epoch": 15.954773869346734, + "grad_norm": 0.08432921767234802, + "learning_rate": 6.293977011845648e-06, + "loss": 0.0106, + "step": 25400 + }, + { + "epoch": 15.96105527638191, + "grad_norm": 0.19509227573871613, + "learning_rate": 6.2672333876159835e-06, + "loss": 0.005, + "step": 25410 + }, + { + "epoch": 15.967336683417086, + "grad_norm": 0.2672835886478424, + "learning_rate": 6.2405429028215e-06, + "loss": 0.006, + "step": 25420 + }, + { + "epoch": 15.97361809045226, + "grad_norm": 0.09370110929012299, + "learning_rate": 6.213905589893631e-06, + "loss": 0.0078, + "step": 25430 + }, + { + "epoch": 15.979899497487438, + "grad_norm": 0.1694602370262146, + "learning_rate": 6.187321481199221e-06, + "loss": 0.011, + "step": 25440 + }, + { + "epoch": 15.986180904522612, + "grad_norm": 0.23776738345623016, + "learning_rate": 6.1607906090404895e-06, + "loss": 0.0103, + "step": 25450 + }, + { + "epoch": 15.992462311557789, + "grad_norm": 0.34769970178604126, + "learning_rate": 6.134313005654929e-06, + "loss": 0.0073, + "step": 25460 + }, + { + "epoch": 15.998743718592966, + "grad_norm": 0.1363392025232315, + "learning_rate": 6.107888703215337e-06, + "loss": 0.009, + "step": 25470 + }, + { + "epoch": 16.00502512562814, + "grad_norm": 0.11508668959140778, + "learning_rate": 6.081517733829722e-06, + "loss": 0.0063, + "step": 25480 + }, + { + "epoch": 16.011306532663315, + "grad_norm": 0.16578857600688934, + "learning_rate": 6.055200129541294e-06, + "loss": 0.0049, + "step": 25490 + }, + { + "epoch": 16.017587939698494, + "grad_norm": 0.2514283061027527, + "learning_rate": 6.0289359223284434e-06, + "loss": 0.0058, + "step": 25500 + }, + { + "epoch": 16.02386934673367, + "grad_norm": 0.16690915822982788, + "learning_rate": 6.002725144104648e-06, + "loss": 0.0088, + "step": 25510 + }, + { + "epoch": 16.030150753768844, + "grad_norm": 0.13841427862644196, + "learning_rate": 5.976567826718477e-06, + "loss": 0.0096, + "step": 25520 + }, + { + "epoch": 16.03643216080402, + "grad_norm": 0.5190842151641846, + "learning_rate": 5.9504640019535315e-06, + "loss": 0.0111, + "step": 25530 + }, + { + "epoch": 16.042713567839197, + "grad_norm": 0.23094475269317627, + "learning_rate": 5.924413701528414e-06, + "loss": 0.0078, + "step": 25540 + }, + { + "epoch": 16.048994974874372, + "grad_norm": 0.18236024677753448, + "learning_rate": 5.898416957096703e-06, + "loss": 0.0073, + "step": 25550 + }, + { + "epoch": 16.055276381909547, + "grad_norm": 0.16717974841594696, + "learning_rate": 5.872473800246913e-06, + "loss": 0.0066, + "step": 25560 + }, + { + "epoch": 16.061557788944725, + "grad_norm": 0.09758424758911133, + "learning_rate": 5.846584262502402e-06, + "loss": 0.0059, + "step": 25570 + }, + { + "epoch": 16.0678391959799, + "grad_norm": 0.1952383667230606, + "learning_rate": 5.820748375321411e-06, + "loss": 0.008, + "step": 25580 + }, + { + "epoch": 16.074120603015075, + "grad_norm": 0.1673065423965454, + "learning_rate": 5.794966170096977e-06, + "loss": 0.0104, + "step": 25590 + }, + { + "epoch": 16.08040201005025, + "grad_norm": 0.23456381261348724, + "learning_rate": 5.769237678156897e-06, + "loss": 0.0086, + "step": 25600 + }, + { + "epoch": 16.08668341708543, + "grad_norm": 0.17800025641918182, + "learning_rate": 5.743562930763735e-06, + "loss": 0.007, + "step": 25610 + }, + { + "epoch": 16.092964824120603, + "grad_norm": 0.17952047288417816, + "learning_rate": 5.717941959114726e-06, + "loss": 0.0071, + "step": 25620 + }, + { + "epoch": 16.099246231155778, + "grad_norm": 0.10446257889270782, + "learning_rate": 5.69237479434176e-06, + "loss": 0.0052, + "step": 25630 + }, + { + "epoch": 16.105527638190956, + "grad_norm": 0.1614079475402832, + "learning_rate": 5.666861467511353e-06, + "loss": 0.0076, + "step": 25640 + }, + { + "epoch": 16.11180904522613, + "grad_norm": 0.15766704082489014, + "learning_rate": 5.641402009624591e-06, + "loss": 0.0074, + "step": 25650 + }, + { + "epoch": 16.118090452261306, + "grad_norm": 0.09763055294752121, + "learning_rate": 5.615996451617145e-06, + "loss": 0.0069, + "step": 25660 + }, + { + "epoch": 16.12437185929648, + "grad_norm": 0.15153726935386658, + "learning_rate": 5.590644824359148e-06, + "loss": 0.0062, + "step": 25670 + }, + { + "epoch": 16.13065326633166, + "grad_norm": 0.24505051970481873, + "learning_rate": 5.56534715865521e-06, + "loss": 0.0056, + "step": 25680 + }, + { + "epoch": 16.136934673366834, + "grad_norm": 0.3446964621543884, + "learning_rate": 5.5401034852443965e-06, + "loss": 0.0068, + "step": 25690 + }, + { + "epoch": 16.14321608040201, + "grad_norm": 0.13532233238220215, + "learning_rate": 5.514913834800134e-06, + "loss": 0.0086, + "step": 25700 + }, + { + "epoch": 16.149497487437188, + "grad_norm": 0.1451926976442337, + "learning_rate": 5.489778237930238e-06, + "loss": 0.0082, + "step": 25710 + }, + { + "epoch": 16.155778894472363, + "grad_norm": 0.15522824227809906, + "learning_rate": 5.464696725176827e-06, + "loss": 0.0076, + "step": 25720 + }, + { + "epoch": 16.162060301507537, + "grad_norm": 0.19680117070674896, + "learning_rate": 5.439669327016294e-06, + "loss": 0.0065, + "step": 25730 + }, + { + "epoch": 16.168341708542712, + "grad_norm": 0.11749772727489471, + "learning_rate": 5.4146960738592985e-06, + "loss": 0.0079, + "step": 25740 + }, + { + "epoch": 16.17462311557789, + "grad_norm": 0.15718911588191986, + "learning_rate": 5.389776996050694e-06, + "loss": 0.0093, + "step": 25750 + }, + { + "epoch": 16.180904522613066, + "grad_norm": 0.18338912725448608, + "learning_rate": 5.364912123869492e-06, + "loss": 0.0047, + "step": 25760 + }, + { + "epoch": 16.18718592964824, + "grad_norm": 0.17256124317646027, + "learning_rate": 5.3401014875288864e-06, + "loss": 0.0062, + "step": 25770 + }, + { + "epoch": 16.193467336683415, + "grad_norm": 0.1998785138130188, + "learning_rate": 5.315345117176129e-06, + "loss": 0.0056, + "step": 25780 + }, + { + "epoch": 16.199748743718594, + "grad_norm": 0.21761392056941986, + "learning_rate": 5.2906430428925405e-06, + "loss": 0.006, + "step": 25790 + }, + { + "epoch": 16.20603015075377, + "grad_norm": 0.18251337110996246, + "learning_rate": 5.265995294693471e-06, + "loss": 0.0103, + "step": 25800 + }, + { + "epoch": 16.212311557788944, + "grad_norm": 0.15696652233600616, + "learning_rate": 5.241401902528254e-06, + "loss": 0.0053, + "step": 25810 + }, + { + "epoch": 16.218592964824122, + "grad_norm": 0.25381505489349365, + "learning_rate": 5.216862896280189e-06, + "loss": 0.0052, + "step": 25820 + }, + { + "epoch": 16.224874371859297, + "grad_norm": 0.15843726694583893, + "learning_rate": 5.192378305766471e-06, + "loss": 0.0053, + "step": 25830 + }, + { + "epoch": 16.231155778894472, + "grad_norm": 0.16248852014541626, + "learning_rate": 5.1679481607382065e-06, + "loss": 0.0052, + "step": 25840 + }, + { + "epoch": 16.237437185929647, + "grad_norm": 0.13264839351177216, + "learning_rate": 5.1435724908803105e-06, + "loss": 0.0084, + "step": 25850 + }, + { + "epoch": 16.243718592964825, + "grad_norm": 0.14164654910564423, + "learning_rate": 5.119251325811514e-06, + "loss": 0.0072, + "step": 25860 + }, + { + "epoch": 16.25, + "grad_norm": 0.12810258567333221, + "learning_rate": 5.094984695084348e-06, + "loss": 0.0083, + "step": 25870 + }, + { + "epoch": 16.256281407035175, + "grad_norm": 0.206588014960289, + "learning_rate": 5.070772628185039e-06, + "loss": 0.0055, + "step": 25880 + }, + { + "epoch": 16.262562814070353, + "grad_norm": 0.25744742155075073, + "learning_rate": 5.046615154533535e-06, + "loss": 0.0074, + "step": 25890 + }, + { + "epoch": 16.268844221105528, + "grad_norm": 0.1719508171081543, + "learning_rate": 5.022512303483451e-06, + "loss": 0.0087, + "step": 25900 + }, + { + "epoch": 16.275125628140703, + "grad_norm": 0.137325718998909, + "learning_rate": 4.998464104322015e-06, + "loss": 0.0073, + "step": 25910 + }, + { + "epoch": 16.281407035175878, + "grad_norm": 0.25743401050567627, + "learning_rate": 4.974470586270047e-06, + "loss": 0.0086, + "step": 25920 + }, + { + "epoch": 16.287688442211056, + "grad_norm": 0.26894593238830566, + "learning_rate": 4.950531778481965e-06, + "loss": 0.0063, + "step": 25930 + }, + { + "epoch": 16.29396984924623, + "grad_norm": 0.1339692622423172, + "learning_rate": 4.926647710045651e-06, + "loss": 0.0049, + "step": 25940 + }, + { + "epoch": 16.300251256281406, + "grad_norm": 0.13080450892448425, + "learning_rate": 4.9028184099825125e-06, + "loss": 0.0063, + "step": 25950 + }, + { + "epoch": 16.306532663316585, + "grad_norm": 0.2640838921070099, + "learning_rate": 4.879043907247382e-06, + "loss": 0.0079, + "step": 25960 + }, + { + "epoch": 16.31281407035176, + "grad_norm": 0.09775843471288681, + "learning_rate": 4.8553242307285414e-06, + "loss": 0.005, + "step": 25970 + }, + { + "epoch": 16.319095477386934, + "grad_norm": 0.0863003358244896, + "learning_rate": 4.8316594092476185e-06, + "loss": 0.0065, + "step": 25980 + }, + { + "epoch": 16.32537688442211, + "grad_norm": 0.1571255475282669, + "learning_rate": 4.808049471559617e-06, + "loss": 0.0069, + "step": 25990 + }, + { + "epoch": 16.331658291457288, + "grad_norm": 0.15241871774196625, + "learning_rate": 4.784494446352833e-06, + "loss": 0.0069, + "step": 26000 + }, + { + "epoch": 16.337939698492463, + "grad_norm": 0.20740200579166412, + "learning_rate": 4.760994362248833e-06, + "loss": 0.0077, + "step": 26010 + }, + { + "epoch": 16.344221105527637, + "grad_norm": 0.2225024700164795, + "learning_rate": 4.737549247802442e-06, + "loss": 0.0073, + "step": 26020 + }, + { + "epoch": 16.350502512562812, + "grad_norm": 0.12289492040872574, + "learning_rate": 4.714159131501689e-06, + "loss": 0.0055, + "step": 26030 + }, + { + "epoch": 16.35678391959799, + "grad_norm": 0.14923076331615448, + "learning_rate": 4.690824041767766e-06, + "loss": 0.0094, + "step": 26040 + }, + { + "epoch": 16.363065326633166, + "grad_norm": 0.18526186048984528, + "learning_rate": 4.6675440069550105e-06, + "loss": 0.0061, + "step": 26050 + }, + { + "epoch": 16.36934673366834, + "grad_norm": 0.23702625930309296, + "learning_rate": 4.64431905535086e-06, + "loss": 0.0055, + "step": 26060 + }, + { + "epoch": 16.37562814070352, + "grad_norm": 0.17731088399887085, + "learning_rate": 4.621149215175808e-06, + "loss": 0.0044, + "step": 26070 + }, + { + "epoch": 16.381909547738694, + "grad_norm": 0.08954228460788727, + "learning_rate": 4.5980345145834154e-06, + "loss": 0.0077, + "step": 26080 + }, + { + "epoch": 16.38819095477387, + "grad_norm": 0.21448710560798645, + "learning_rate": 4.574974981660213e-06, + "loss": 0.0066, + "step": 26090 + }, + { + "epoch": 16.394472361809044, + "grad_norm": 0.07934491336345673, + "learning_rate": 4.551970644425707e-06, + "loss": 0.0077, + "step": 26100 + }, + { + "epoch": 16.400753768844222, + "grad_norm": 0.12047973275184631, + "learning_rate": 4.529021530832328e-06, + "loss": 0.0079, + "step": 26110 + }, + { + "epoch": 16.407035175879397, + "grad_norm": 0.2353978455066681, + "learning_rate": 4.5061276687654306e-06, + "loss": 0.0097, + "step": 26120 + }, + { + "epoch": 16.41331658291457, + "grad_norm": 0.1489868015050888, + "learning_rate": 4.4832890860431965e-06, + "loss": 0.0067, + "step": 26130 + }, + { + "epoch": 16.41959798994975, + "grad_norm": 0.10002940893173218, + "learning_rate": 4.460505810416682e-06, + "loss": 0.0064, + "step": 26140 + }, + { + "epoch": 16.425879396984925, + "grad_norm": 0.08586425334215164, + "learning_rate": 4.4377778695696984e-06, + "loss": 0.0053, + "step": 26150 + }, + { + "epoch": 16.4321608040201, + "grad_norm": 0.18994610011577606, + "learning_rate": 4.415105291118843e-06, + "loss": 0.0065, + "step": 26160 + }, + { + "epoch": 16.438442211055275, + "grad_norm": 0.13901755213737488, + "learning_rate": 4.392488102613435e-06, + "loss": 0.0074, + "step": 26170 + }, + { + "epoch": 16.444723618090453, + "grad_norm": 0.1574508398771286, + "learning_rate": 4.369926331535473e-06, + "loss": 0.0065, + "step": 26180 + }, + { + "epoch": 16.451005025125628, + "grad_norm": 0.15778610110282898, + "learning_rate": 4.347420005299668e-06, + "loss": 0.0087, + "step": 26190 + }, + { + "epoch": 16.457286432160803, + "grad_norm": 0.23142878711223602, + "learning_rate": 4.324969151253317e-06, + "loss": 0.0051, + "step": 26200 + }, + { + "epoch": 16.46356783919598, + "grad_norm": 0.20028924942016602, + "learning_rate": 4.302573796676312e-06, + "loss": 0.0055, + "step": 26210 + }, + { + "epoch": 16.469849246231156, + "grad_norm": 0.19725173711776733, + "learning_rate": 4.2802339687811385e-06, + "loss": 0.0084, + "step": 26220 + }, + { + "epoch": 16.47613065326633, + "grad_norm": 0.15638677775859833, + "learning_rate": 4.257949694712771e-06, + "loss": 0.0071, + "step": 26230 + }, + { + "epoch": 16.482412060301506, + "grad_norm": 0.22199900448322296, + "learning_rate": 4.235721001548726e-06, + "loss": 0.0079, + "step": 26240 + }, + { + "epoch": 16.488693467336685, + "grad_norm": 0.08355461061000824, + "learning_rate": 4.21354791629896e-06, + "loss": 0.0054, + "step": 26250 + }, + { + "epoch": 16.49497487437186, + "grad_norm": 0.1285744458436966, + "learning_rate": 4.191430465905843e-06, + "loss": 0.0064, + "step": 26260 + }, + { + "epoch": 16.501256281407034, + "grad_norm": 0.24992157518863678, + "learning_rate": 4.169368677244184e-06, + "loss": 0.0088, + "step": 26270 + }, + { + "epoch": 16.507537688442213, + "grad_norm": 0.17267563939094543, + "learning_rate": 4.147362577121122e-06, + "loss": 0.0081, + "step": 26280 + }, + { + "epoch": 16.513819095477388, + "grad_norm": 0.14897999167442322, + "learning_rate": 4.125412192276157e-06, + "loss": 0.0048, + "step": 26290 + }, + { + "epoch": 16.520100502512562, + "grad_norm": 0.16491813957691193, + "learning_rate": 4.10351754938107e-06, + "loss": 0.0075, + "step": 26300 + }, + { + "epoch": 16.526381909547737, + "grad_norm": 0.19591006636619568, + "learning_rate": 4.081678675039913e-06, + "loss": 0.0057, + "step": 26310 + }, + { + "epoch": 16.532663316582916, + "grad_norm": 0.1836249679327011, + "learning_rate": 4.0598955957889876e-06, + "loss": 0.0063, + "step": 26320 + }, + { + "epoch": 16.53894472361809, + "grad_norm": 0.1280887871980667, + "learning_rate": 4.038168338096776e-06, + "loss": 0.0064, + "step": 26330 + }, + { + "epoch": 16.545226130653266, + "grad_norm": 0.1596054583787918, + "learning_rate": 4.016496928363944e-06, + "loss": 0.0058, + "step": 26340 + }, + { + "epoch": 16.551507537688444, + "grad_norm": 0.12449732422828674, + "learning_rate": 3.994881392923317e-06, + "loss": 0.0075, + "step": 26350 + }, + { + "epoch": 16.55778894472362, + "grad_norm": 0.1359025239944458, + "learning_rate": 3.9733217580397945e-06, + "loss": 0.0067, + "step": 26360 + }, + { + "epoch": 16.564070351758794, + "grad_norm": 0.12694570422172546, + "learning_rate": 3.951818049910366e-06, + "loss": 0.0063, + "step": 26370 + }, + { + "epoch": 16.57035175879397, + "grad_norm": 0.19445101916790009, + "learning_rate": 3.930370294664071e-06, + "loss": 0.0083, + "step": 26380 + }, + { + "epoch": 16.576633165829147, + "grad_norm": 0.2000349462032318, + "learning_rate": 3.908978518361938e-06, + "loss": 0.0066, + "step": 26390 + }, + { + "epoch": 16.582914572864322, + "grad_norm": 0.09491092711687088, + "learning_rate": 3.887642746997017e-06, + "loss": 0.0069, + "step": 26400 + }, + { + "epoch": 16.589195979899497, + "grad_norm": 0.1539655327796936, + "learning_rate": 3.866363006494255e-06, + "loss": 0.005, + "step": 26410 + }, + { + "epoch": 16.59547738693467, + "grad_norm": 0.15679632127285004, + "learning_rate": 3.8451393227105725e-06, + "loss": 0.0053, + "step": 26420 + }, + { + "epoch": 16.60175879396985, + "grad_norm": 0.16239243745803833, + "learning_rate": 3.823971721434727e-06, + "loss": 0.0076, + "step": 26430 + }, + { + "epoch": 16.608040201005025, + "grad_norm": 0.15958023071289062, + "learning_rate": 3.8028602283873505e-06, + "loss": 0.007, + "step": 26440 + }, + { + "epoch": 16.6143216080402, + "grad_norm": 0.22681252658367157, + "learning_rate": 3.781804869220912e-06, + "loss": 0.0049, + "step": 26450 + }, + { + "epoch": 16.62060301507538, + "grad_norm": 0.08722344040870667, + "learning_rate": 3.760805669519646e-06, + "loss": 0.0109, + "step": 26460 + }, + { + "epoch": 16.626884422110553, + "grad_norm": 0.14116171002388, + "learning_rate": 3.7398626547995587e-06, + "loss": 0.0068, + "step": 26470 + }, + { + "epoch": 16.633165829145728, + "grad_norm": 0.21690021455287933, + "learning_rate": 3.7189758505083995e-06, + "loss": 0.0054, + "step": 26480 + }, + { + "epoch": 16.639447236180903, + "grad_norm": 0.12882182002067566, + "learning_rate": 3.6981452820255837e-06, + "loss": 0.005, + "step": 26490 + }, + { + "epoch": 16.64572864321608, + "grad_norm": 0.11476683616638184, + "learning_rate": 3.6773709746622363e-06, + "loss": 0.0053, + "step": 26500 + }, + { + "epoch": 16.652010050251256, + "grad_norm": 0.13834761083126068, + "learning_rate": 3.6566529536610995e-06, + "loss": 0.0072, + "step": 26510 + }, + { + "epoch": 16.65829145728643, + "grad_norm": 0.2981722354888916, + "learning_rate": 3.635991244196513e-06, + "loss": 0.0046, + "step": 26520 + }, + { + "epoch": 16.66457286432161, + "grad_norm": 0.22664038836956024, + "learning_rate": 3.615385871374405e-06, + "loss": 0.009, + "step": 26530 + }, + { + "epoch": 16.670854271356784, + "grad_norm": 0.1676933318376541, + "learning_rate": 3.5948368602322367e-06, + "loss": 0.0068, + "step": 26540 + }, + { + "epoch": 16.67713567839196, + "grad_norm": 0.23804248869419098, + "learning_rate": 3.5743442357390187e-06, + "loss": 0.0085, + "step": 26550 + }, + { + "epoch": 16.683417085427134, + "grad_norm": 0.264737069606781, + "learning_rate": 3.5539080227951937e-06, + "loss": 0.0065, + "step": 26560 + }, + { + "epoch": 16.689698492462313, + "grad_norm": 0.1040617972612381, + "learning_rate": 3.533528246232709e-06, + "loss": 0.0075, + "step": 26570 + }, + { + "epoch": 16.695979899497488, + "grad_norm": 0.16781878471374512, + "learning_rate": 3.5132049308149117e-06, + "loss": 0.0061, + "step": 26580 + }, + { + "epoch": 16.702261306532662, + "grad_norm": 0.1874973624944687, + "learning_rate": 3.4929381012365425e-06, + "loss": 0.0086, + "step": 26590 + }, + { + "epoch": 16.70854271356784, + "grad_norm": 0.10712270438671112, + "learning_rate": 3.472727782123697e-06, + "loss": 0.0062, + "step": 26600 + }, + { + "epoch": 16.714824120603016, + "grad_norm": 0.2084541618824005, + "learning_rate": 3.4525739980338425e-06, + "loss": 0.007, + "step": 26610 + }, + { + "epoch": 16.72110552763819, + "grad_norm": 0.12348504364490509, + "learning_rate": 3.432476773455712e-06, + "loss": 0.0072, + "step": 26620 + }, + { + "epoch": 16.727386934673365, + "grad_norm": 0.26043209433555603, + "learning_rate": 3.4124361328093378e-06, + "loss": 0.0056, + "step": 26630 + }, + { + "epoch": 16.733668341708544, + "grad_norm": 0.17123505473136902, + "learning_rate": 3.392452100445975e-06, + "loss": 0.0057, + "step": 26640 + }, + { + "epoch": 16.73994974874372, + "grad_norm": 0.16309408843517303, + "learning_rate": 3.3725247006481163e-06, + "loss": 0.0074, + "step": 26650 + }, + { + "epoch": 16.746231155778894, + "grad_norm": 0.20002205669879913, + "learning_rate": 3.3526539576294313e-06, + "loss": 0.014, + "step": 26660 + }, + { + "epoch": 16.75251256281407, + "grad_norm": 0.10679427534341812, + "learning_rate": 3.332839895534745e-06, + "loss": 0.0067, + "step": 26670 + }, + { + "epoch": 16.758793969849247, + "grad_norm": 0.16640935838222504, + "learning_rate": 3.3130825384400155e-06, + "loss": 0.0065, + "step": 26680 + }, + { + "epoch": 16.765075376884422, + "grad_norm": 0.2365013062953949, + "learning_rate": 3.293381910352278e-06, + "loss": 0.0067, + "step": 26690 + }, + { + "epoch": 16.771356783919597, + "grad_norm": 0.13385744392871857, + "learning_rate": 3.273738035209678e-06, + "loss": 0.0099, + "step": 26700 + }, + { + "epoch": 16.777638190954775, + "grad_norm": 0.11509007960557938, + "learning_rate": 3.254150936881356e-06, + "loss": 0.0052, + "step": 26710 + }, + { + "epoch": 16.78391959798995, + "grad_norm": 0.1979365199804306, + "learning_rate": 3.234620639167496e-06, + "loss": 0.0072, + "step": 26720 + }, + { + "epoch": 16.790201005025125, + "grad_norm": 0.11373197287321091, + "learning_rate": 3.2151471657992482e-06, + "loss": 0.0084, + "step": 26730 + }, + { + "epoch": 16.7964824120603, + "grad_norm": 0.09649147838354111, + "learning_rate": 3.195730540438718e-06, + "loss": 0.0051, + "step": 26740 + }, + { + "epoch": 16.80276381909548, + "grad_norm": 0.21292780339717865, + "learning_rate": 3.1763707866789336e-06, + "loss": 0.007, + "step": 26750 + }, + { + "epoch": 16.809045226130653, + "grad_norm": 0.12464366108179092, + "learning_rate": 3.1570679280438165e-06, + "loss": 0.0058, + "step": 26760 + }, + { + "epoch": 16.815326633165828, + "grad_norm": 0.08573342114686966, + "learning_rate": 3.1378219879881767e-06, + "loss": 0.007, + "step": 26770 + }, + { + "epoch": 16.821608040201006, + "grad_norm": 0.23342382907867432, + "learning_rate": 3.11863298989763e-06, + "loss": 0.0068, + "step": 26780 + }, + { + "epoch": 16.82788944723618, + "grad_norm": 0.10594352334737778, + "learning_rate": 3.0995009570886303e-06, + "loss": 0.0054, + "step": 26790 + }, + { + "epoch": 16.834170854271356, + "grad_norm": 0.23619455099105835, + "learning_rate": 3.0804259128083936e-06, + "loss": 0.0075, + "step": 26800 + }, + { + "epoch": 16.84045226130653, + "grad_norm": 0.10237495601177216, + "learning_rate": 3.06140788023489e-06, + "loss": 0.0063, + "step": 26810 + }, + { + "epoch": 16.84673366834171, + "grad_norm": 0.14859092235565186, + "learning_rate": 3.042446882476846e-06, + "loss": 0.0069, + "step": 26820 + }, + { + "epoch": 16.853015075376884, + "grad_norm": 0.105214424431324, + "learning_rate": 3.023542942573643e-06, + "loss": 0.0049, + "step": 26830 + }, + { + "epoch": 16.85929648241206, + "grad_norm": 0.13950201869010925, + "learning_rate": 3.004696083495351e-06, + "loss": 0.0072, + "step": 26840 + }, + { + "epoch": 16.865577889447238, + "grad_norm": 0.20687682926654816, + "learning_rate": 2.98590632814269e-06, + "loss": 0.0089, + "step": 26850 + }, + { + "epoch": 16.871859296482413, + "grad_norm": 0.1909753531217575, + "learning_rate": 2.9671736993469702e-06, + "loss": 0.0058, + "step": 26860 + }, + { + "epoch": 16.878140703517587, + "grad_norm": 0.2131378948688507, + "learning_rate": 2.9484982198701216e-06, + "loss": 0.0087, + "step": 26870 + }, + { + "epoch": 16.884422110552762, + "grad_norm": 0.24551810324192047, + "learning_rate": 2.929879912404604e-06, + "loss": 0.0085, + "step": 26880 + }, + { + "epoch": 16.89070351758794, + "grad_norm": 0.14596037566661835, + "learning_rate": 2.9113187995734147e-06, + "loss": 0.0075, + "step": 26890 + }, + { + "epoch": 16.896984924623116, + "grad_norm": 0.19911986589431763, + "learning_rate": 2.8928149039300524e-06, + "loss": 0.0067, + "step": 26900 + }, + { + "epoch": 16.90326633165829, + "grad_norm": 0.16665054857730865, + "learning_rate": 2.8743682479584975e-06, + "loss": 0.0053, + "step": 26910 + }, + { + "epoch": 16.90954773869347, + "grad_norm": 0.1604067087173462, + "learning_rate": 2.8559788540731824e-06, + "loss": 0.007, + "step": 26920 + }, + { + "epoch": 16.915829145728644, + "grad_norm": 0.08772148191928864, + "learning_rate": 2.8376467446189493e-06, + "loss": 0.006, + "step": 26930 + }, + { + "epoch": 16.92211055276382, + "grad_norm": 0.11479980498552322, + "learning_rate": 2.8193719418710406e-06, + "loss": 0.0048, + "step": 26940 + }, + { + "epoch": 16.928391959798994, + "grad_norm": 0.18683061003684998, + "learning_rate": 2.8011544680350667e-06, + "loss": 0.0092, + "step": 26950 + }, + { + "epoch": 16.934673366834172, + "grad_norm": 0.10781048983335495, + "learning_rate": 2.7829943452469753e-06, + "loss": 0.0084, + "step": 26960 + }, + { + "epoch": 16.940954773869347, + "grad_norm": 0.18485447764396667, + "learning_rate": 2.7648915955730213e-06, + "loss": 0.0053, + "step": 26970 + }, + { + "epoch": 16.94723618090452, + "grad_norm": 0.11769340187311172, + "learning_rate": 2.746846241009765e-06, + "loss": 0.0059, + "step": 26980 + }, + { + "epoch": 16.9535175879397, + "grad_norm": 0.1902921497821808, + "learning_rate": 2.7288583034839945e-06, + "loss": 0.0061, + "step": 26990 + }, + { + "epoch": 16.959798994974875, + "grad_norm": 0.21465308964252472, + "learning_rate": 2.7109278048527752e-06, + "loss": 0.01, + "step": 27000 + }, + { + "epoch": 16.96608040201005, + "grad_norm": 0.12692305445671082, + "learning_rate": 2.6930547669033413e-06, + "loss": 0.0055, + "step": 27010 + }, + { + "epoch": 16.972361809045225, + "grad_norm": 0.29463475942611694, + "learning_rate": 2.675239211353109e-06, + "loss": 0.0079, + "step": 27020 + }, + { + "epoch": 16.978643216080403, + "grad_norm": 0.23902346193790436, + "learning_rate": 2.6574811598496786e-06, + "loss": 0.0063, + "step": 27030 + }, + { + "epoch": 16.984924623115578, + "grad_norm": 0.1303122639656067, + "learning_rate": 2.6397806339707456e-06, + "loss": 0.0048, + "step": 27040 + }, + { + "epoch": 16.991206030150753, + "grad_norm": 0.14880459010601044, + "learning_rate": 2.6221376552241217e-06, + "loss": 0.0052, + "step": 27050 + }, + { + "epoch": 16.99748743718593, + "grad_norm": 0.09494416415691376, + "learning_rate": 2.604552245047681e-06, + "loss": 0.007, + "step": 27060 + }, + { + "epoch": 17.003768844221106, + "grad_norm": 0.15154646337032318, + "learning_rate": 2.587024424809359e-06, + "loss": 0.0057, + "step": 27070 + }, + { + "epoch": 17.01005025125628, + "grad_norm": 0.17682980000972748, + "learning_rate": 2.5695542158071186e-06, + "loss": 0.0077, + "step": 27080 + }, + { + "epoch": 17.016331658291456, + "grad_norm": 0.09535016119480133, + "learning_rate": 2.5521416392689068e-06, + "loss": 0.0077, + "step": 27090 + }, + { + "epoch": 17.022613065326635, + "grad_norm": 0.17054226994514465, + "learning_rate": 2.5347867163526384e-06, + "loss": 0.0066, + "step": 27100 + }, + { + "epoch": 17.02889447236181, + "grad_norm": 0.19506265223026276, + "learning_rate": 2.517489468146189e-06, + "loss": 0.0099, + "step": 27110 + }, + { + "epoch": 17.035175879396984, + "grad_norm": 0.234808087348938, + "learning_rate": 2.500249915667341e-06, + "loss": 0.0096, + "step": 27120 + }, + { + "epoch": 17.04145728643216, + "grad_norm": 0.18437589704990387, + "learning_rate": 2.4830680798637817e-06, + "loss": 0.0052, + "step": 27130 + }, + { + "epoch": 17.047738693467338, + "grad_norm": 0.11160774528980255, + "learning_rate": 2.465943981613056e-06, + "loss": 0.0068, + "step": 27140 + }, + { + "epoch": 17.054020100502512, + "grad_norm": 0.23293474316596985, + "learning_rate": 2.448877641722569e-06, + "loss": 0.008, + "step": 27150 + }, + { + "epoch": 17.060301507537687, + "grad_norm": 0.19176483154296875, + "learning_rate": 2.431869080929522e-06, + "loss": 0.0071, + "step": 27160 + }, + { + "epoch": 17.066582914572866, + "grad_norm": 0.24945424497127533, + "learning_rate": 2.4149183199009216e-06, + "loss": 0.0087, + "step": 27170 + }, + { + "epoch": 17.07286432160804, + "grad_norm": 0.11020611971616745, + "learning_rate": 2.3980253792335427e-06, + "loss": 0.0048, + "step": 27180 + }, + { + "epoch": 17.079145728643216, + "grad_norm": 0.25345098972320557, + "learning_rate": 2.381190279453899e-06, + "loss": 0.0098, + "step": 27190 + }, + { + "epoch": 17.08542713567839, + "grad_norm": 0.16273361444473267, + "learning_rate": 2.364413041018232e-06, + "loss": 0.0092, + "step": 27200 + }, + { + "epoch": 17.09170854271357, + "grad_norm": 0.14581908285617828, + "learning_rate": 2.347693684312463e-06, + "loss": 0.0078, + "step": 27210 + }, + { + "epoch": 17.097989949748744, + "grad_norm": 0.09183979034423828, + "learning_rate": 2.331032229652186e-06, + "loss": 0.004, + "step": 27220 + }, + { + "epoch": 17.10427135678392, + "grad_norm": 0.13825593888759613, + "learning_rate": 2.314428697282628e-06, + "loss": 0.0062, + "step": 27230 + }, + { + "epoch": 17.110552763819097, + "grad_norm": 0.07496145367622375, + "learning_rate": 2.297883107378673e-06, + "loss": 0.0043, + "step": 27240 + }, + { + "epoch": 17.116834170854272, + "grad_norm": 0.14740310609340668, + "learning_rate": 2.2813954800447513e-06, + "loss": 0.0056, + "step": 27250 + }, + { + "epoch": 17.123115577889447, + "grad_norm": 0.14972737431526184, + "learning_rate": 2.2649658353148974e-06, + "loss": 0.0057, + "step": 27260 + }, + { + "epoch": 17.12939698492462, + "grad_norm": 0.22126655280590057, + "learning_rate": 2.2485941931526645e-06, + "loss": 0.0069, + "step": 27270 + }, + { + "epoch": 17.1356783919598, + "grad_norm": 0.0594448521733284, + "learning_rate": 2.232280573451151e-06, + "loss": 0.0058, + "step": 27280 + }, + { + "epoch": 17.141959798994975, + "grad_norm": 0.17575815320014954, + "learning_rate": 2.216024996032945e-06, + "loss": 0.0077, + "step": 27290 + }, + { + "epoch": 17.14824120603015, + "grad_norm": 0.07905078679323196, + "learning_rate": 2.1998274806501074e-06, + "loss": 0.0078, + "step": 27300 + }, + { + "epoch": 17.15452261306533, + "grad_norm": 0.26970353722572327, + "learning_rate": 2.183688046984139e-06, + "loss": 0.0072, + "step": 27310 + }, + { + "epoch": 17.160804020100503, + "grad_norm": 0.08259978890419006, + "learning_rate": 2.167606714645981e-06, + "loss": 0.0074, + "step": 27320 + }, + { + "epoch": 17.167085427135678, + "grad_norm": 0.13270756602287292, + "learning_rate": 2.151583503175958e-06, + "loss": 0.0061, + "step": 27330 + }, + { + "epoch": 17.173366834170853, + "grad_norm": 0.326485812664032, + "learning_rate": 2.1356184320437957e-06, + "loss": 0.0086, + "step": 27340 + }, + { + "epoch": 17.17964824120603, + "grad_norm": 0.21641714870929718, + "learning_rate": 2.119711520648554e-06, + "loss": 0.0053, + "step": 27350 + }, + { + "epoch": 17.185929648241206, + "grad_norm": 0.1555628925561905, + "learning_rate": 2.103862788318628e-06, + "loss": 0.0073, + "step": 27360 + }, + { + "epoch": 17.19221105527638, + "grad_norm": 0.17456355690956116, + "learning_rate": 2.088072254311729e-06, + "loss": 0.0087, + "step": 27370 + }, + { + "epoch": 17.198492462311556, + "grad_norm": 0.14001718163490295, + "learning_rate": 2.0723399378148435e-06, + "loss": 0.0076, + "step": 27380 + }, + { + "epoch": 17.204773869346734, + "grad_norm": 0.15439637005329132, + "learning_rate": 2.0566658579442065e-06, + "loss": 0.0062, + "step": 27390 + }, + { + "epoch": 17.21105527638191, + "grad_norm": 0.24933287501335144, + "learning_rate": 2.0410500337453176e-06, + "loss": 0.0076, + "step": 27400 + }, + { + "epoch": 17.217336683417084, + "grad_norm": 0.14670556783676147, + "learning_rate": 2.0254924841928645e-06, + "loss": 0.0079, + "step": 27410 + }, + { + "epoch": 17.223618090452263, + "grad_norm": 0.13927485048770905, + "learning_rate": 2.009993228190754e-06, + "loss": 0.0085, + "step": 27420 + }, + { + "epoch": 17.229899497487438, + "grad_norm": 0.097551129758358, + "learning_rate": 1.9945522845720323e-06, + "loss": 0.0052, + "step": 27430 + }, + { + "epoch": 17.236180904522612, + "grad_norm": 0.14993266761302948, + "learning_rate": 1.9791696720988963e-06, + "loss": 0.0079, + "step": 27440 + }, + { + "epoch": 17.242462311557787, + "grad_norm": 0.2073020339012146, + "learning_rate": 1.9638454094626833e-06, + "loss": 0.0063, + "step": 27450 + }, + { + "epoch": 17.248743718592966, + "grad_norm": 0.18664003908634186, + "learning_rate": 1.9485795152838104e-06, + "loss": 0.0059, + "step": 27460 + }, + { + "epoch": 17.25502512562814, + "grad_norm": 0.21333898603916168, + "learning_rate": 1.933372008111778e-06, + "loss": 0.007, + "step": 27470 + }, + { + "epoch": 17.261306532663315, + "grad_norm": 0.12153317034244537, + "learning_rate": 1.918222906425143e-06, + "loss": 0.0108, + "step": 27480 + }, + { + "epoch": 17.267587939698494, + "grad_norm": 0.17056246101856232, + "learning_rate": 1.903132228631488e-06, + "loss": 0.0076, + "step": 27490 + }, + { + "epoch": 17.27386934673367, + "grad_norm": 0.20078368484973907, + "learning_rate": 1.8880999930674215e-06, + "loss": 0.0083, + "step": 27500 + }, + { + "epoch": 17.280150753768844, + "grad_norm": 0.2632201910018921, + "learning_rate": 1.8731262179985166e-06, + "loss": 0.0081, + "step": 27510 + }, + { + "epoch": 17.28643216080402, + "grad_norm": 0.19306138157844543, + "learning_rate": 1.8582109216193244e-06, + "loss": 0.0061, + "step": 27520 + }, + { + "epoch": 17.292713567839197, + "grad_norm": 0.19993025064468384, + "learning_rate": 1.8433541220533368e-06, + "loss": 0.0082, + "step": 27530 + }, + { + "epoch": 17.298994974874372, + "grad_norm": 0.31495529413223267, + "learning_rate": 1.8285558373529577e-06, + "loss": 0.0062, + "step": 27540 + }, + { + "epoch": 17.305276381909547, + "grad_norm": 0.1742606908082962, + "learning_rate": 1.8138160854995145e-06, + "loss": 0.0062, + "step": 27550 + }, + { + "epoch": 17.311557788944725, + "grad_norm": 0.1084170937538147, + "learning_rate": 1.7991348844031864e-06, + "loss": 0.0043, + "step": 27560 + }, + { + "epoch": 17.3178391959799, + "grad_norm": 0.1582164466381073, + "learning_rate": 1.78451225190302e-06, + "loss": 0.0062, + "step": 27570 + }, + { + "epoch": 17.324120603015075, + "grad_norm": 0.25446414947509766, + "learning_rate": 1.7699482057668914e-06, + "loss": 0.0075, + "step": 27580 + }, + { + "epoch": 17.33040201005025, + "grad_norm": 0.21719922125339508, + "learning_rate": 1.7554427636914895e-06, + "loss": 0.0053, + "step": 27590 + }, + { + "epoch": 17.33668341708543, + "grad_norm": 0.09698139876127243, + "learning_rate": 1.7409959433022871e-06, + "loss": 0.0048, + "step": 27600 + }, + { + "epoch": 17.342964824120603, + "grad_norm": 0.07788410037755966, + "learning_rate": 1.7266077621535481e-06, + "loss": 0.0047, + "step": 27610 + }, + { + "epoch": 17.349246231155778, + "grad_norm": 0.11927897483110428, + "learning_rate": 1.7122782377282597e-06, + "loss": 0.0066, + "step": 27620 + }, + { + "epoch": 17.355527638190956, + "grad_norm": 0.13954825699329376, + "learning_rate": 1.6980073874381497e-06, + "loss": 0.0056, + "step": 27630 + }, + { + "epoch": 17.36180904522613, + "grad_norm": 0.14202164113521576, + "learning_rate": 1.6837952286236413e-06, + "loss": 0.0056, + "step": 27640 + }, + { + "epoch": 17.368090452261306, + "grad_norm": 0.1293388456106186, + "learning_rate": 1.6696417785538488e-06, + "loss": 0.0075, + "step": 27650 + }, + { + "epoch": 17.37437185929648, + "grad_norm": 0.1332010179758072, + "learning_rate": 1.6555470544265538e-06, + "loss": 0.0068, + "step": 27660 + }, + { + "epoch": 17.38065326633166, + "grad_norm": 0.0950307846069336, + "learning_rate": 1.6415110733681738e-06, + "loss": 0.0081, + "step": 27670 + }, + { + "epoch": 17.386934673366834, + "grad_norm": 0.06528349965810776, + "learning_rate": 1.6275338524337436e-06, + "loss": 0.007, + "step": 27680 + }, + { + "epoch": 17.39321608040201, + "grad_norm": 0.2808053195476532, + "learning_rate": 1.6136154086069056e-06, + "loss": 0.0079, + "step": 27690 + }, + { + "epoch": 17.399497487437188, + "grad_norm": 0.1333058476448059, + "learning_rate": 1.5997557587998868e-06, + "loss": 0.0053, + "step": 27700 + }, + { + "epoch": 17.405778894472363, + "grad_norm": 0.12153434753417969, + "learning_rate": 1.5859549198534606e-06, + "loss": 0.0068, + "step": 27710 + }, + { + "epoch": 17.412060301507537, + "grad_norm": 0.09190189093351364, + "learning_rate": 1.572212908536963e-06, + "loss": 0.0049, + "step": 27720 + }, + { + "epoch": 17.418341708542712, + "grad_norm": 0.09856624901294708, + "learning_rate": 1.5585297415482202e-06, + "loss": 0.0066, + "step": 27730 + }, + { + "epoch": 17.42462311557789, + "grad_norm": 0.16106554865837097, + "learning_rate": 1.5449054355135717e-06, + "loss": 0.0056, + "step": 27740 + }, + { + "epoch": 17.430904522613066, + "grad_norm": 0.24775199592113495, + "learning_rate": 1.5313400069878415e-06, + "loss": 0.0093, + "step": 27750 + }, + { + "epoch": 17.43718592964824, + "grad_norm": 0.17308726906776428, + "learning_rate": 1.5178334724542887e-06, + "loss": 0.0067, + "step": 27760 + }, + { + "epoch": 17.443467336683415, + "grad_norm": 0.15941742062568665, + "learning_rate": 1.504385848324641e-06, + "loss": 0.0069, + "step": 27770 + }, + { + "epoch": 17.449748743718594, + "grad_norm": 0.20731601119041443, + "learning_rate": 1.490997150939033e-06, + "loss": 0.0074, + "step": 27780 + }, + { + "epoch": 17.45603015075377, + "grad_norm": 0.09693028032779694, + "learning_rate": 1.4776673965659793e-06, + "loss": 0.0049, + "step": 27790 + }, + { + "epoch": 17.462311557788944, + "grad_norm": 0.12705643475055695, + "learning_rate": 1.4643966014023957e-06, + "loss": 0.0058, + "step": 27800 + }, + { + "epoch": 17.468592964824122, + "grad_norm": 0.1498476266860962, + "learning_rate": 1.4511847815735503e-06, + "loss": 0.0059, + "step": 27810 + }, + { + "epoch": 17.474874371859297, + "grad_norm": 0.14018553495407104, + "learning_rate": 1.438031953133051e-06, + "loss": 0.0075, + "step": 27820 + }, + { + "epoch": 17.481155778894472, + "grad_norm": 0.2627798616886139, + "learning_rate": 1.4249381320628197e-06, + "loss": 0.0068, + "step": 27830 + }, + { + "epoch": 17.487437185929647, + "grad_norm": 0.43465185165405273, + "learning_rate": 1.4119033342730902e-06, + "loss": 0.0086, + "step": 27840 + }, + { + "epoch": 17.493718592964825, + "grad_norm": 0.33725589513778687, + "learning_rate": 1.3989275756023656e-06, + "loss": 0.006, + "step": 27850 + }, + { + "epoch": 17.5, + "grad_norm": 0.20212982594966888, + "learning_rate": 1.386010871817417e-06, + "loss": 0.0066, + "step": 27860 + }, + { + "epoch": 17.506281407035175, + "grad_norm": 0.1582840234041214, + "learning_rate": 1.3731532386132616e-06, + "loss": 0.0056, + "step": 27870 + }, + { + "epoch": 17.512562814070353, + "grad_norm": 0.13995033502578735, + "learning_rate": 1.360354691613136e-06, + "loss": 0.0049, + "step": 27880 + }, + { + "epoch": 17.518844221105528, + "grad_norm": 0.15379869937896729, + "learning_rate": 1.3476152463684776e-06, + "loss": 0.0052, + "step": 27890 + }, + { + "epoch": 17.525125628140703, + "grad_norm": 0.14645810425281525, + "learning_rate": 1.3349349183589155e-06, + "loss": 0.0069, + "step": 27900 + }, + { + "epoch": 17.531407035175878, + "grad_norm": 0.12568466365337372, + "learning_rate": 1.3223137229922356e-06, + "loss": 0.0066, + "step": 27910 + }, + { + "epoch": 17.537688442211056, + "grad_norm": 0.16633038222789764, + "learning_rate": 1.3097516756043981e-06, + "loss": 0.0058, + "step": 27920 + }, + { + "epoch": 17.54396984924623, + "grad_norm": 0.09484007954597473, + "learning_rate": 1.29724879145946e-06, + "loss": 0.007, + "step": 27930 + }, + { + "epoch": 17.550251256281406, + "grad_norm": 0.14168323576450348, + "learning_rate": 1.284805085749613e-06, + "loss": 0.0079, + "step": 27940 + }, + { + "epoch": 17.556532663316585, + "grad_norm": 0.13008199632167816, + "learning_rate": 1.2724205735951288e-06, + "loss": 0.0065, + "step": 27950 + }, + { + "epoch": 17.56281407035176, + "grad_norm": 0.15203717350959778, + "learning_rate": 1.2600952700443591e-06, + "loss": 0.0084, + "step": 27960 + }, + { + "epoch": 17.569095477386934, + "grad_norm": 0.1610209345817566, + "learning_rate": 1.247829190073707e-06, + "loss": 0.0064, + "step": 27970 + }, + { + "epoch": 17.57537688442211, + "grad_norm": 0.13040009140968323, + "learning_rate": 1.2356223485876173e-06, + "loss": 0.006, + "step": 27980 + }, + { + "epoch": 17.581658291457288, + "grad_norm": 0.13261644542217255, + "learning_rate": 1.2234747604185526e-06, + "loss": 0.0042, + "step": 27990 + }, + { + "epoch": 17.587939698492463, + "grad_norm": 0.11774436384439468, + "learning_rate": 1.2113864403269836e-06, + "loss": 0.0065, + "step": 28000 + }, + { + "epoch": 17.594221105527637, + "grad_norm": 0.2139647901058197, + "learning_rate": 1.1993574030013555e-06, + "loss": 0.0047, + "step": 28010 + }, + { + "epoch": 17.600502512562812, + "grad_norm": 0.08927455544471741, + "learning_rate": 1.1873876630580816e-06, + "loss": 0.0056, + "step": 28020 + }, + { + "epoch": 17.60678391959799, + "grad_norm": 0.23126229643821716, + "learning_rate": 1.1754772350415278e-06, + "loss": 0.0073, + "step": 28030 + }, + { + "epoch": 17.613065326633166, + "grad_norm": 0.14410418272018433, + "learning_rate": 1.163626133423984e-06, + "loss": 0.0077, + "step": 28040 + }, + { + "epoch": 17.61934673366834, + "grad_norm": 0.21467958390712738, + "learning_rate": 1.1518343726056591e-06, + "loss": 0.0079, + "step": 28050 + }, + { + "epoch": 17.62562814070352, + "grad_norm": 0.11036968976259232, + "learning_rate": 1.1401019669146474e-06, + "loss": 0.0068, + "step": 28060 + }, + { + "epoch": 17.631909547738694, + "grad_norm": 0.1723591834306717, + "learning_rate": 1.128428930606934e-06, + "loss": 0.0057, + "step": 28070 + }, + { + "epoch": 17.63819095477387, + "grad_norm": 0.11529406160116196, + "learning_rate": 1.1168152778663621e-06, + "loss": 0.0053, + "step": 28080 + }, + { + "epoch": 17.644472361809044, + "grad_norm": 0.19840645790100098, + "learning_rate": 1.10526102280461e-06, + "loss": 0.0062, + "step": 28090 + }, + { + "epoch": 17.650753768844222, + "grad_norm": 0.19137519598007202, + "learning_rate": 1.0937661794611864e-06, + "loss": 0.005, + "step": 28100 + }, + { + "epoch": 17.657035175879397, + "grad_norm": 0.1524164378643036, + "learning_rate": 1.082330761803413e-06, + "loss": 0.0082, + "step": 28110 + }, + { + "epoch": 17.66331658291457, + "grad_norm": 0.11807109415531158, + "learning_rate": 1.0709547837263966e-06, + "loss": 0.0071, + "step": 28120 + }, + { + "epoch": 17.66959798994975, + "grad_norm": 0.14275571703910828, + "learning_rate": 1.059638259053025e-06, + "loss": 0.006, + "step": 28130 + }, + { + "epoch": 17.675879396984925, + "grad_norm": 0.20063506066799164, + "learning_rate": 1.0483812015339434e-06, + "loss": 0.005, + "step": 28140 + }, + { + "epoch": 17.6821608040201, + "grad_norm": 0.19718168675899506, + "learning_rate": 1.037183624847543e-06, + "loss": 0.0035, + "step": 28150 + }, + { + "epoch": 17.688442211055275, + "grad_norm": 0.10679283738136292, + "learning_rate": 1.0260455425999238e-06, + "loss": 0.0055, + "step": 28160 + }, + { + "epoch": 17.694723618090453, + "grad_norm": 0.12524540722370148, + "learning_rate": 1.0149669683249208e-06, + "loss": 0.0051, + "step": 28170 + }, + { + "epoch": 17.701005025125628, + "grad_norm": 0.20479629933834076, + "learning_rate": 1.003947915484027e-06, + "loss": 0.006, + "step": 28180 + }, + { + "epoch": 17.707286432160803, + "grad_norm": 0.10765405744314194, + "learning_rate": 9.92988397466449e-07, + "loss": 0.0051, + "step": 28190 + }, + { + "epoch": 17.71356783919598, + "grad_norm": 0.14588141441345215, + "learning_rate": 9.820884275890286e-07, + "loss": 0.0068, + "step": 28200 + }, + { + "epoch": 17.719849246231156, + "grad_norm": 0.13059480488300323, + "learning_rate": 9.712480190962546e-07, + "loss": 0.0063, + "step": 28210 + }, + { + "epoch": 17.72613065326633, + "grad_norm": 0.08463677763938904, + "learning_rate": 9.604671851602464e-07, + "loss": 0.0068, + "step": 28220 + }, + { + "epoch": 17.732412060301506, + "grad_norm": 0.10322614759206772, + "learning_rate": 9.497459388807306e-07, + "loss": 0.0053, + "step": 28230 + }, + { + "epoch": 17.738693467336685, + "grad_norm": 0.2695980966091156, + "learning_rate": 9.390842932850364e-07, + "loss": 0.0084, + "step": 28240 + }, + { + "epoch": 17.74497487437186, + "grad_norm": 0.16168110072612762, + "learning_rate": 9.28482261328073e-07, + "loss": 0.0079, + "step": 28250 + }, + { + "epoch": 17.751256281407034, + "grad_norm": 0.11593770235776901, + "learning_rate": 9.179398558923025e-07, + "loss": 0.0069, + "step": 28260 + }, + { + "epoch": 17.757537688442213, + "grad_norm": 0.21430358290672302, + "learning_rate": 9.074570897877388e-07, + "loss": 0.0062, + "step": 28270 + }, + { + "epoch": 17.763819095477388, + "grad_norm": 0.11505214869976044, + "learning_rate": 8.970339757519375e-07, + "loss": 0.0067, + "step": 28280 + }, + { + "epoch": 17.770100502512562, + "grad_norm": 0.10474999994039536, + "learning_rate": 8.866705264499619e-07, + "loss": 0.0056, + "step": 28290 + }, + { + "epoch": 17.776381909547737, + "grad_norm": 0.10137923806905746, + "learning_rate": 8.763667544743836e-07, + "loss": 0.0051, + "step": 28300 + }, + { + "epoch": 17.782663316582916, + "grad_norm": 0.07817348837852478, + "learning_rate": 8.661226723452543e-07, + "loss": 0.0045, + "step": 28310 + }, + { + "epoch": 17.78894472361809, + "grad_norm": 0.11985107511281967, + "learning_rate": 8.559382925101001e-07, + "loss": 0.0053, + "step": 28320 + }, + { + "epoch": 17.795226130653266, + "grad_norm": 0.12586714327335358, + "learning_rate": 8.458136273438943e-07, + "loss": 0.0063, + "step": 28330 + }, + { + "epoch": 17.801507537688444, + "grad_norm": 0.3606361448764801, + "learning_rate": 8.357486891490795e-07, + "loss": 0.0088, + "step": 28340 + }, + { + "epoch": 17.80778894472362, + "grad_norm": 0.1316625326871872, + "learning_rate": 8.257434901554895e-07, + "loss": 0.0072, + "step": 28350 + }, + { + "epoch": 17.814070351758794, + "grad_norm": 0.14042818546295166, + "learning_rate": 8.157980425203937e-07, + "loss": 0.005, + "step": 28360 + }, + { + "epoch": 17.82035175879397, + "grad_norm": 0.17309603095054626, + "learning_rate": 8.059123583284367e-07, + "loss": 0.007, + "step": 28370 + }, + { + "epoch": 17.826633165829147, + "grad_norm": 0.1516924798488617, + "learning_rate": 7.960864495916654e-07, + "loss": 0.0064, + "step": 28380 + }, + { + "epoch": 17.832914572864322, + "grad_norm": 0.10025578737258911, + "learning_rate": 7.863203282494847e-07, + "loss": 0.0074, + "step": 28390 + }, + { + "epoch": 17.839195979899497, + "grad_norm": 0.1148303970694542, + "learning_rate": 7.766140061686522e-07, + "loss": 0.0064, + "step": 28400 + }, + { + "epoch": 17.84547738693467, + "grad_norm": 0.20660251379013062, + "learning_rate": 7.669674951432615e-07, + "loss": 0.0089, + "step": 28410 + }, + { + "epoch": 17.85175879396985, + "grad_norm": 0.2036808580160141, + "learning_rate": 7.573808068947363e-07, + "loss": 0.0049, + "step": 28420 + }, + { + "epoch": 17.858040201005025, + "grad_norm": 0.11376306414604187, + "learning_rate": 7.478539530718087e-07, + "loss": 0.0067, + "step": 28430 + }, + { + "epoch": 17.8643216080402, + "grad_norm": 0.1357523798942566, + "learning_rate": 7.383869452504965e-07, + "loss": 0.008, + "step": 28440 + }, + { + "epoch": 17.87060301507538, + "grad_norm": 0.11328667402267456, + "learning_rate": 7.289797949341204e-07, + "loss": 0.0061, + "step": 28450 + }, + { + "epoch": 17.876884422110553, + "grad_norm": 0.1712828129529953, + "learning_rate": 7.196325135532423e-07, + "loss": 0.0096, + "step": 28460 + }, + { + "epoch": 17.883165829145728, + "grad_norm": 0.18172721564769745, + "learning_rate": 7.10345112465699e-07, + "loss": 0.0057, + "step": 28470 + }, + { + "epoch": 17.889447236180903, + "grad_norm": 0.158191978931427, + "learning_rate": 7.011176029565525e-07, + "loss": 0.0082, + "step": 28480 + }, + { + "epoch": 17.89572864321608, + "grad_norm": 0.29038453102111816, + "learning_rate": 6.919499962381004e-07, + "loss": 0.0063, + "step": 28490 + }, + { + "epoch": 17.902010050251256, + "grad_norm": 0.1755780428647995, + "learning_rate": 6.828423034498488e-07, + "loss": 0.0064, + "step": 28500 + }, + { + "epoch": 17.90829145728643, + "grad_norm": 0.1357237696647644, + "learning_rate": 6.737945356585007e-07, + "loss": 0.0071, + "step": 28510 + }, + { + "epoch": 17.91457286432161, + "grad_norm": 0.09247785806655884, + "learning_rate": 6.648067038579508e-07, + "loss": 0.0058, + "step": 28520 + }, + { + "epoch": 17.920854271356784, + "grad_norm": 0.23100236058235168, + "learning_rate": 6.558788189692578e-07, + "loss": 0.0057, + "step": 28530 + }, + { + "epoch": 17.92713567839196, + "grad_norm": 0.13841266930103302, + "learning_rate": 6.470108918406492e-07, + "loss": 0.006, + "step": 28540 + }, + { + "epoch": 17.933417085427134, + "grad_norm": 0.1735645979642868, + "learning_rate": 6.382029332474893e-07, + "loss": 0.0062, + "step": 28550 + }, + { + "epoch": 17.939698492462313, + "grad_norm": 0.18910877406597137, + "learning_rate": 6.294549538922778e-07, + "loss": 0.007, + "step": 28560 + }, + { + "epoch": 17.945979899497488, + "grad_norm": 0.1468404233455658, + "learning_rate": 6.207669644046344e-07, + "loss": 0.0064, + "step": 28570 + }, + { + "epoch": 17.952261306532662, + "grad_norm": 0.1430070847272873, + "learning_rate": 6.121389753412865e-07, + "loss": 0.0047, + "step": 28580 + }, + { + "epoch": 17.95854271356784, + "grad_norm": 0.13665567338466644, + "learning_rate": 6.035709971860592e-07, + "loss": 0.0056, + "step": 28590 + }, + { + "epoch": 17.964824120603016, + "grad_norm": 0.22900241613388062, + "learning_rate": 5.950630403498469e-07, + "loss": 0.0073, + "step": 28600 + }, + { + "epoch": 17.97110552763819, + "grad_norm": 0.11574529111385345, + "learning_rate": 5.866151151706189e-07, + "loss": 0.0061, + "step": 28610 + }, + { + "epoch": 17.977386934673365, + "grad_norm": 0.14552676677703857, + "learning_rate": 5.782272319134086e-07, + "loss": 0.0063, + "step": 28620 + }, + { + "epoch": 17.983668341708544, + "grad_norm": 0.09723411500453949, + "learning_rate": 5.698994007702796e-07, + "loss": 0.0085, + "step": 28630 + }, + { + "epoch": 17.98994974874372, + "grad_norm": 0.2545289397239685, + "learning_rate": 5.616316318603321e-07, + "loss": 0.0078, + "step": 28640 + }, + { + "epoch": 17.996231155778894, + "grad_norm": 0.13836827874183655, + "learning_rate": 5.534239352296799e-07, + "loss": 0.0065, + "step": 28650 + }, + { + "epoch": 18.002512562814072, + "grad_norm": 0.07939734309911728, + "learning_rate": 5.452763208514621e-07, + "loss": 0.0071, + "step": 28660 + }, + { + "epoch": 18.008793969849247, + "grad_norm": 0.13539274036884308, + "learning_rate": 5.371887986257873e-07, + "loss": 0.0049, + "step": 28670 + }, + { + "epoch": 18.015075376884422, + "grad_norm": 0.2377186268568039, + "learning_rate": 5.291613783797611e-07, + "loss": 0.007, + "step": 28680 + }, + { + "epoch": 18.021356783919597, + "grad_norm": 0.20826736092567444, + "learning_rate": 5.211940698674534e-07, + "loss": 0.0083, + "step": 28690 + }, + { + "epoch": 18.027638190954775, + "grad_norm": 0.24656341969966888, + "learning_rate": 5.132868827698978e-07, + "loss": 0.0071, + "step": 28700 + }, + { + "epoch": 18.03391959798995, + "grad_norm": 0.11674555391073227, + "learning_rate": 5.054398266950755e-07, + "loss": 0.0047, + "step": 28710 + }, + { + "epoch": 18.040201005025125, + "grad_norm": 0.18100731074810028, + "learning_rate": 4.976529111778872e-07, + "loss": 0.006, + "step": 28720 + }, + { + "epoch": 18.0464824120603, + "grad_norm": 0.15892614424228668, + "learning_rate": 4.899261456801862e-07, + "loss": 0.0055, + "step": 28730 + }, + { + "epoch": 18.05276381909548, + "grad_norm": 0.12166585773229599, + "learning_rate": 4.822595395907126e-07, + "loss": 0.0064, + "step": 28740 + }, + { + "epoch": 18.059045226130653, + "grad_norm": 0.1044788807630539, + "learning_rate": 4.7465310222510886e-07, + "loss": 0.0051, + "step": 28750 + }, + { + "epoch": 18.065326633165828, + "grad_norm": 0.39492928981781006, + "learning_rate": 4.6710684282593175e-07, + "loss": 0.0046, + "step": 28760 + }, + { + "epoch": 18.071608040201006, + "grad_norm": 0.1567634791135788, + "learning_rate": 4.5962077056257993e-07, + "loss": 0.0057, + "step": 28770 + }, + { + "epoch": 18.07788944723618, + "grad_norm": 0.09841161221265793, + "learning_rate": 4.521948945313492e-07, + "loss": 0.0101, + "step": 28780 + }, + { + "epoch": 18.084170854271356, + "grad_norm": 0.22830891609191895, + "learning_rate": 4.4482922375537195e-07, + "loss": 0.0076, + "step": 28790 + }, + { + "epoch": 18.09045226130653, + "grad_norm": 0.09455195814371109, + "learning_rate": 4.375237671846333e-07, + "loss": 0.0072, + "step": 28800 + }, + { + "epoch": 18.09673366834171, + "grad_norm": 0.1464862823486328, + "learning_rate": 4.302785336959547e-07, + "loss": 0.0071, + "step": 28810 + }, + { + "epoch": 18.103015075376884, + "grad_norm": 0.1865178644657135, + "learning_rate": 4.2309353209297744e-07, + "loss": 0.0053, + "step": 28820 + }, + { + "epoch": 18.10929648241206, + "grad_norm": 0.2174028754234314, + "learning_rate": 4.159687711061566e-07, + "loss": 0.005, + "step": 28830 + }, + { + "epoch": 18.115577889447238, + "grad_norm": 0.10583844780921936, + "learning_rate": 4.0890425939275055e-07, + "loss": 0.0042, + "step": 28840 + }, + { + "epoch": 18.121859296482413, + "grad_norm": 0.15250837802886963, + "learning_rate": 4.0190000553679827e-07, + "loss": 0.0059, + "step": 28850 + }, + { + "epoch": 18.128140703517587, + "grad_norm": 0.18485210835933685, + "learning_rate": 3.9495601804913627e-07, + "loss": 0.0045, + "step": 28860 + }, + { + "epoch": 18.134422110552762, + "grad_norm": 0.25837287306785583, + "learning_rate": 3.880723053673652e-07, + "loss": 0.0077, + "step": 28870 + }, + { + "epoch": 18.14070351758794, + "grad_norm": 0.16366513073444366, + "learning_rate": 3.812488758558386e-07, + "loss": 0.0062, + "step": 28880 + }, + { + "epoch": 18.146984924623116, + "grad_norm": 0.2013944685459137, + "learning_rate": 3.744857378056743e-07, + "loss": 0.0075, + "step": 28890 + }, + { + "epoch": 18.15326633165829, + "grad_norm": 0.08087635785341263, + "learning_rate": 3.677828994347154e-07, + "loss": 0.0042, + "step": 28900 + }, + { + "epoch": 18.15954773869347, + "grad_norm": 0.38147228956222534, + "learning_rate": 3.61140368887547e-07, + "loss": 0.0068, + "step": 28910 + }, + { + "epoch": 18.165829145728644, + "grad_norm": 0.08277861028909683, + "learning_rate": 3.545581542354681e-07, + "loss": 0.0045, + "step": 28920 + }, + { + "epoch": 18.17211055276382, + "grad_norm": 0.10835427790880203, + "learning_rate": 3.480362634764922e-07, + "loss": 0.006, + "step": 28930 + }, + { + "epoch": 18.178391959798994, + "grad_norm": 0.21344228088855743, + "learning_rate": 3.4157470453533014e-07, + "loss": 0.0072, + "step": 28940 + }, + { + "epoch": 18.184673366834172, + "grad_norm": 0.1102982610464096, + "learning_rate": 3.3517348526339034e-07, + "loss": 0.0054, + "step": 28950 + }, + { + "epoch": 18.190954773869347, + "grad_norm": 0.1867319941520691, + "learning_rate": 3.288326134387454e-07, + "loss": 0.0035, + "step": 28960 + }, + { + "epoch": 18.19723618090452, + "grad_norm": 0.0828981101512909, + "learning_rate": 3.225520967661655e-07, + "loss": 0.0046, + "step": 28970 + }, + { + "epoch": 18.2035175879397, + "grad_norm": 0.09847334027290344, + "learning_rate": 3.163319428770628e-07, + "loss": 0.0056, + "step": 28980 + }, + { + "epoch": 18.209798994974875, + "grad_norm": 0.15965382754802704, + "learning_rate": 3.1017215932951374e-07, + "loss": 0.007, + "step": 28990 + }, + { + "epoch": 18.21608040201005, + "grad_norm": 0.12046926468610764, + "learning_rate": 3.040727536082366e-07, + "loss": 0.0057, + "step": 29000 + }, + { + "epoch": 18.222361809045225, + "grad_norm": 0.16416752338409424, + "learning_rate": 2.980337331245864e-07, + "loss": 0.0094, + "step": 29010 + }, + { + "epoch": 18.228643216080403, + "grad_norm": 0.21736080944538116, + "learning_rate": 2.9205510521653213e-07, + "loss": 0.0077, + "step": 29020 + }, + { + "epoch": 18.234924623115578, + "grad_norm": 0.24882303178310394, + "learning_rate": 2.86136877148685e-07, + "loss": 0.0073, + "step": 29030 + }, + { + "epoch": 18.241206030150753, + "grad_norm": 0.15103134512901306, + "learning_rate": 2.8027905611223704e-07, + "loss": 0.0049, + "step": 29040 + }, + { + "epoch": 18.247487437185928, + "grad_norm": 0.323935329914093, + "learning_rate": 2.7448164922500573e-07, + "loss": 0.0056, + "step": 29050 + }, + { + "epoch": 18.253768844221106, + "grad_norm": 0.12751802802085876, + "learning_rate": 2.687446635313784e-07, + "loss": 0.0049, + "step": 29060 + }, + { + "epoch": 18.26005025125628, + "grad_norm": 0.19040335714817047, + "learning_rate": 2.630681060023343e-07, + "loss": 0.0059, + "step": 29070 + }, + { + "epoch": 18.266331658291456, + "grad_norm": 0.14866675436496735, + "learning_rate": 2.5745198353542833e-07, + "loss": 0.0066, + "step": 29080 + }, + { + "epoch": 18.272613065326635, + "grad_norm": 0.13203470408916473, + "learning_rate": 2.518963029547794e-07, + "loss": 0.0049, + "step": 29090 + }, + { + "epoch": 18.27889447236181, + "grad_norm": 0.17947359383106232, + "learning_rate": 2.464010710110598e-07, + "loss": 0.0091, + "step": 29100 + }, + { + "epoch": 18.285175879396984, + "grad_norm": 0.19403871893882751, + "learning_rate": 2.4096629438150054e-07, + "loss": 0.0067, + "step": 29110 + }, + { + "epoch": 18.29145728643216, + "grad_norm": 0.17872844636440277, + "learning_rate": 2.3559197966985802e-07, + "loss": 0.0074, + "step": 29120 + }, + { + "epoch": 18.297738693467338, + "grad_norm": 0.27938979864120483, + "learning_rate": 2.3027813340644188e-07, + "loss": 0.007, + "step": 29130 + }, + { + "epoch": 18.304020100502512, + "grad_norm": 0.2715089023113251, + "learning_rate": 2.2502476204807055e-07, + "loss": 0.0061, + "step": 29140 + }, + { + "epoch": 18.310301507537687, + "grad_norm": 0.2602129578590393, + "learning_rate": 2.1983187197808786e-07, + "loss": 0.005, + "step": 29150 + }, + { + "epoch": 18.316582914572866, + "grad_norm": 0.13536551594734192, + "learning_rate": 2.1469946950634644e-07, + "loss": 0.0054, + "step": 29160 + }, + { + "epoch": 18.32286432160804, + "grad_norm": 0.1295653134584427, + "learning_rate": 2.096275608691911e-07, + "loss": 0.0056, + "step": 29170 + }, + { + "epoch": 18.329145728643216, + "grad_norm": 0.1381261795759201, + "learning_rate": 2.046161522294754e-07, + "loss": 0.0053, + "step": 29180 + }, + { + "epoch": 18.33542713567839, + "grad_norm": 0.15450970828533173, + "learning_rate": 1.9966524967653944e-07, + "loss": 0.0064, + "step": 29190 + }, + { + "epoch": 18.34170854271357, + "grad_norm": 0.15897521376609802, + "learning_rate": 1.9477485922618222e-07, + "loss": 0.0054, + "step": 29200 + }, + { + "epoch": 18.347989949748744, + "grad_norm": 0.2164761871099472, + "learning_rate": 1.899449868207004e-07, + "loss": 0.0069, + "step": 29210 + }, + { + "epoch": 18.35427135678392, + "grad_norm": 0.12174555659294128, + "learning_rate": 1.851756383288439e-07, + "loss": 0.0083, + "step": 29220 + }, + { + "epoch": 18.360552763819097, + "grad_norm": 0.13597124814987183, + "learning_rate": 1.8046681954581035e-07, + "loss": 0.0054, + "step": 29230 + }, + { + "epoch": 18.366834170854272, + "grad_norm": 0.2730105221271515, + "learning_rate": 1.7581853619327294e-07, + "loss": 0.0078, + "step": 29240 + }, + { + "epoch": 18.373115577889447, + "grad_norm": 0.17662861943244934, + "learning_rate": 1.7123079391932472e-07, + "loss": 0.0063, + "step": 29250 + }, + { + "epoch": 18.37939698492462, + "grad_norm": 0.1380220651626587, + "learning_rate": 1.6670359829850657e-07, + "loss": 0.0055, + "step": 29260 + }, + { + "epoch": 18.3856783919598, + "grad_norm": 0.11557086557149887, + "learning_rate": 1.6223695483179037e-07, + "loss": 0.0051, + "step": 29270 + }, + { + "epoch": 18.391959798994975, + "grad_norm": 0.23970536887645721, + "learning_rate": 1.5783086894656795e-07, + "loss": 0.0084, + "step": 29280 + }, + { + "epoch": 18.39824120603015, + "grad_norm": 0.1585719734430313, + "learning_rate": 1.5348534599665121e-07, + "loss": 0.0059, + "step": 29290 + }, + { + "epoch": 18.40452261306533, + "grad_norm": 0.3307650089263916, + "learning_rate": 1.4920039126225527e-07, + "loss": 0.0049, + "step": 29300 + }, + { + "epoch": 18.410804020100503, + "grad_norm": 0.0947142168879509, + "learning_rate": 1.449760099500097e-07, + "loss": 0.0072, + "step": 29310 + }, + { + "epoch": 18.417085427135678, + "grad_norm": 0.132417693734169, + "learning_rate": 1.4081220719293075e-07, + "loss": 0.0052, + "step": 29320 + }, + { + "epoch": 18.423366834170853, + "grad_norm": 0.14502517879009247, + "learning_rate": 1.3670898805043797e-07, + "loss": 0.006, + "step": 29330 + }, + { + "epoch": 18.42964824120603, + "grad_norm": 0.13710470497608185, + "learning_rate": 1.326663575083209e-07, + "loss": 0.0057, + "step": 29340 + }, + { + "epoch": 18.435929648241206, + "grad_norm": 0.1276075392961502, + "learning_rate": 1.2868432047876688e-07, + "loss": 0.0069, + "step": 29350 + }, + { + "epoch": 18.44221105527638, + "grad_norm": 0.3342688977718353, + "learning_rate": 1.2476288180032213e-07, + "loss": 0.0059, + "step": 29360 + }, + { + "epoch": 18.448492462311556, + "grad_norm": 0.14222240447998047, + "learning_rate": 1.209020462379029e-07, + "loss": 0.0081, + "step": 29370 + }, + { + "epoch": 18.454773869346734, + "grad_norm": 0.05781752988696098, + "learning_rate": 1.1710181848278435e-07, + "loss": 0.0052, + "step": 29380 + }, + { + "epoch": 18.46105527638191, + "grad_norm": 0.10229113698005676, + "learning_rate": 1.133622031526116e-07, + "loss": 0.0045, + "step": 29390 + }, + { + "epoch": 18.467336683417084, + "grad_norm": 0.24169184267520905, + "learning_rate": 1.0968320479136095e-07, + "loss": 0.0074, + "step": 29400 + }, + { + "epoch": 18.473618090452263, + "grad_norm": 0.11586479842662811, + "learning_rate": 1.0606482786936767e-07, + "loss": 0.0074, + "step": 29410 + }, + { + "epoch": 18.479899497487438, + "grad_norm": 0.13327839970588684, + "learning_rate": 1.0250707678329808e-07, + "loss": 0.0073, + "step": 29420 + }, + { + "epoch": 18.486180904522612, + "grad_norm": 0.11354170739650726, + "learning_rate": 9.900995585615525e-08, + "loss": 0.005, + "step": 29430 + }, + { + "epoch": 18.492462311557787, + "grad_norm": 0.11315594613552094, + "learning_rate": 9.55734693372734e-08, + "loss": 0.0063, + "step": 29440 + }, + { + "epoch": 18.498743718592966, + "grad_norm": 0.10254649817943573, + "learning_rate": 9.219762140231236e-08, + "loss": 0.0061, + "step": 29450 + }, + { + "epoch": 18.50502512562814, + "grad_norm": 0.10047972947359085, + "learning_rate": 8.888241615322978e-08, + "loss": 0.0042, + "step": 29460 + }, + { + "epoch": 18.511306532663315, + "grad_norm": 0.20406900346279144, + "learning_rate": 8.562785761833114e-08, + "loss": 0.0073, + "step": 29470 + }, + { + "epoch": 18.517587939698494, + "grad_norm": 0.09750339388847351, + "learning_rate": 8.243394975219753e-08, + "loss": 0.0066, + "step": 29480 + }, + { + "epoch": 18.52386934673367, + "grad_norm": 0.15511460602283478, + "learning_rate": 7.930069643573568e-08, + "loss": 0.0044, + "step": 29490 + }, + { + "epoch": 18.530150753768844, + "grad_norm": 0.17475847899913788, + "learning_rate": 7.622810147614456e-08, + "loss": 0.0074, + "step": 29500 + }, + { + "epoch": 18.53643216080402, + "grad_norm": 0.09517867863178253, + "learning_rate": 7.321616860690995e-08, + "loss": 0.005, + "step": 29510 + }, + { + "epoch": 18.542713567839197, + "grad_norm": 0.1267365962266922, + "learning_rate": 7.026490148782095e-08, + "loss": 0.0052, + "step": 29520 + }, + { + "epoch": 18.548994974874372, + "grad_norm": 0.14582936465740204, + "learning_rate": 6.737430370494236e-08, + "loss": 0.0077, + "step": 29530 + }, + { + "epoch": 18.555276381909547, + "grad_norm": 0.17223408818244934, + "learning_rate": 6.454437877062569e-08, + "loss": 0.0079, + "step": 29540 + }, + { + "epoch": 18.561557788944725, + "grad_norm": 0.1441621333360672, + "learning_rate": 6.177513012349256e-08, + "loss": 0.0063, + "step": 29550 + }, + { + "epoch": 18.5678391959799, + "grad_norm": 0.24988897144794464, + "learning_rate": 5.9066561128445775e-08, + "loss": 0.0046, + "step": 29560 + }, + { + "epoch": 18.574120603015075, + "grad_norm": 0.2458711862564087, + "learning_rate": 5.6418675076641556e-08, + "loss": 0.0069, + "step": 29570 + }, + { + "epoch": 18.58040201005025, + "grad_norm": 0.12190630286931992, + "learning_rate": 5.383147518552845e-08, + "loss": 0.0065, + "step": 29580 + }, + { + "epoch": 18.58668341708543, + "grad_norm": 0.27268433570861816, + "learning_rate": 5.1304964598786196e-08, + "loss": 0.0046, + "step": 29590 + }, + { + "epoch": 18.592964824120603, + "grad_norm": 0.1632656306028366, + "learning_rate": 4.883914638636467e-08, + "loss": 0.0092, + "step": 29600 + }, + { + "epoch": 18.599246231155778, + "grad_norm": 0.21346358954906464, + "learning_rate": 4.643402354446713e-08, + "loss": 0.0066, + "step": 29610 + }, + { + "epoch": 18.605527638190956, + "grad_norm": 0.15229511260986328, + "learning_rate": 4.4089598995544766e-08, + "loss": 0.0065, + "step": 29620 + }, + { + "epoch": 18.61180904522613, + "grad_norm": 0.26550284028053284, + "learning_rate": 4.180587558829663e-08, + "loss": 0.0051, + "step": 29630 + }, + { + "epoch": 18.618090452261306, + "grad_norm": 0.44738510251045227, + "learning_rate": 3.958285609765855e-08, + "loss": 0.006, + "step": 29640 + }, + { + "epoch": 18.62437185929648, + "grad_norm": 0.2252192199230194, + "learning_rate": 3.74205432248087e-08, + "loss": 0.0047, + "step": 29650 + }, + { + "epoch": 18.63065326633166, + "grad_norm": 0.10328078269958496, + "learning_rate": 3.531893959716204e-08, + "loss": 0.0051, + "step": 29660 + }, + { + "epoch": 18.636934673366834, + "grad_norm": 0.2012810856103897, + "learning_rate": 3.327804776837029e-08, + "loss": 0.0069, + "step": 29670 + }, + { + "epoch": 18.64321608040201, + "grad_norm": 0.08085604757070541, + "learning_rate": 3.129787021829977e-08, + "loss": 0.0068, + "step": 29680 + }, + { + "epoch": 18.649497487437188, + "grad_norm": 0.12689168751239777, + "learning_rate": 2.9378409353059133e-08, + "loss": 0.0075, + "step": 29690 + }, + { + "epoch": 18.655778894472363, + "grad_norm": 0.09985805302858353, + "learning_rate": 2.7519667504971593e-08, + "loss": 0.0063, + "step": 29700 + }, + { + "epoch": 18.662060301507537, + "grad_norm": 0.1520301252603531, + "learning_rate": 2.572164693258605e-08, + "loss": 0.0049, + "step": 29710 + }, + { + "epoch": 18.668341708542712, + "grad_norm": 0.18411098420619965, + "learning_rate": 2.3984349820665996e-08, + "loss": 0.0058, + "step": 29720 + }, + { + "epoch": 18.67462311557789, + "grad_norm": 0.15979772806167603, + "learning_rate": 2.2307778280189485e-08, + "loss": 0.0064, + "step": 29730 + }, + { + "epoch": 18.680904522613066, + "grad_norm": 0.13281023502349854, + "learning_rate": 2.0691934348354704e-08, + "loss": 0.007, + "step": 29740 + }, + { + "epoch": 18.68718592964824, + "grad_norm": 0.2212098240852356, + "learning_rate": 1.9136819988557763e-08, + "loss": 0.0061, + "step": 29750 + }, + { + "epoch": 18.693467336683415, + "grad_norm": 0.1470957249403, + "learning_rate": 1.7642437090414908e-08, + "loss": 0.0058, + "step": 29760 + }, + { + "epoch": 18.699748743718594, + "grad_norm": 0.16389614343643188, + "learning_rate": 1.6208787469734755e-08, + "loss": 0.008, + "step": 29770 + }, + { + "epoch": 18.70603015075377, + "grad_norm": 0.20274977385997772, + "learning_rate": 1.483587286854604e-08, + "loss": 0.0057, + "step": 29780 + }, + { + "epoch": 18.712311557788944, + "grad_norm": 0.08999701589345932, + "learning_rate": 1.3523694955064336e-08, + "loss": 0.0054, + "step": 29790 + }, + { + "epoch": 18.718592964824122, + "grad_norm": 0.09353580325841904, + "learning_rate": 1.2272255323708681e-08, + "loss": 0.0067, + "step": 29800 + }, + { + "epoch": 18.724874371859297, + "grad_norm": 0.12401217222213745, + "learning_rate": 1.1081555495096042e-08, + "loss": 0.01, + "step": 29810 + }, + { + "epoch": 18.731155778894472, + "grad_norm": 0.14299121499061584, + "learning_rate": 9.951596916041306e-09, + "loss": 0.0063, + "step": 29820 + }, + { + "epoch": 18.737437185929647, + "grad_norm": 0.19786103069782257, + "learning_rate": 8.882380959551739e-09, + "loss": 0.0076, + "step": 29830 + }, + { + "epoch": 18.743718592964825, + "grad_norm": 0.057169489562511444, + "learning_rate": 7.873908924821428e-09, + "loss": 0.0042, + "step": 29840 + }, + { + "epoch": 18.75, + "grad_norm": 0.17132189869880676, + "learning_rate": 6.926182037242379e-09, + "loss": 0.009, + "step": 29850 + }, + { + "epoch": 18.756281407035175, + "grad_norm": 0.11860886961221695, + "learning_rate": 6.039201448387876e-09, + "loss": 0.0074, + "step": 29860 + }, + { + "epoch": 18.762562814070353, + "grad_norm": 0.1588948667049408, + "learning_rate": 5.212968236029125e-09, + "loss": 0.0059, + "step": 29870 + }, + { + "epoch": 18.768844221105528, + "grad_norm": 0.07842261344194412, + "learning_rate": 4.447483404118602e-09, + "loss": 0.0094, + "step": 29880 + }, + { + "epoch": 18.775125628140703, + "grad_norm": 0.2275543361902237, + "learning_rate": 3.742747882784503e-09, + "loss": 0.0073, + "step": 29890 + }, + { + "epoch": 18.781407035175878, + "grad_norm": 0.11445245146751404, + "learning_rate": 3.0987625283473987e-09, + "loss": 0.0041, + "step": 29900 + }, + { + "epoch": 18.787688442211056, + "grad_norm": 0.1638801544904709, + "learning_rate": 2.515528123320232e-09, + "loss": 0.0067, + "step": 29910 + }, + { + "epoch": 18.79396984924623, + "grad_norm": 0.059369660913944244, + "learning_rate": 1.9930453763750133e-09, + "loss": 0.0052, + "step": 29920 + }, + { + "epoch": 18.800251256281406, + "grad_norm": 0.07346530258655548, + "learning_rate": 1.5313149223872281e-09, + "loss": 0.0064, + "step": 29930 + }, + { + "epoch": 18.806532663316585, + "grad_norm": 0.2043532431125641, + "learning_rate": 1.1303373224025305e-09, + "loss": 0.005, + "step": 29940 + }, + { + "epoch": 18.81281407035176, + "grad_norm": 0.32668429613113403, + "learning_rate": 7.901130636367438e-10, + "loss": 0.0064, + "step": 29950 + }, + { + "epoch": 18.819095477386934, + "grad_norm": 0.14015771448612213, + "learning_rate": 5.106425595036158e-10, + "loss": 0.0065, + "step": 29960 + }, + { + "epoch": 18.82537688442211, + "grad_norm": 0.17479540407657623, + "learning_rate": 2.9192614958706286e-10, + "loss": 0.0068, + "step": 29970 + }, + { + "epoch": 18.831658291457288, + "grad_norm": 0.19859153032302856, + "learning_rate": 1.3396409964117064e-10, + "loss": 0.0071, + "step": 29980 + }, + { + "epoch": 18.837939698492463, + "grad_norm": 0.09940286725759506, + "learning_rate": 3.6756601606846574e-11, + "loss": 0.0049, + "step": 29990 + }, + { + "epoch": 18.844221105527637, + "grad_norm": 0.12079217284917831, + "learning_rate": 3.0377360626943076e-13, + "loss": 0.005, + "step": 30000 + } + ], + "logging_steps": 10, + "max_steps": 30000, + "num_input_tokens_seen": 0, + "num_train_epochs": 19, + "save_steps": 5000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}