{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.222244381904602, "learning_rate": 3.0000000000000004e-07, "loss": 1.4835, "step": 10 }, { "grad_norm": 1.012047529220581, "learning_rate": 6.333333333333333e-07, "loss": 1.5052, "step": 20 }, { "grad_norm": 1.0012941360473633, "learning_rate": 9.666666666666668e-07, "loss": 1.4629, "step": 30 }, { "grad_norm": 0.929564356803894, "learning_rate": 1.3e-06, "loss": 1.4549, "step": 40 }, { "grad_norm": 1.164833426475525, "learning_rate": 1.6333333333333333e-06, "loss": 1.4464, "step": 50 }, { "grad_norm": 1.8519965410232544, "learning_rate": 1.9666666666666668e-06, "loss": 1.381, "step": 60 }, { "grad_norm": 2.005781650543213, "learning_rate": 2.3e-06, "loss": 1.2721, "step": 70 }, { "grad_norm": 1.7347328662872314, "learning_rate": 2.6333333333333337e-06, "loss": 1.2604, "step": 80 }, { "grad_norm": 2.4137401580810547, "learning_rate": 2.966666666666667e-06, "loss": 1.205, "step": 90 }, { "grad_norm": 2.0290517807006836, "learning_rate": 3.3e-06, "loss": 1.2439, "step": 100 }, { "grad_norm": 1.6361143589019775, "learning_rate": 3.633333333333334e-06, "loss": 1.1711, "step": 110 }, { "grad_norm": 1.6367464065551758, "learning_rate": 3.966666666666667e-06, "loss": 1.1723, "step": 120 }, { "grad_norm": 2.0098633766174316, "learning_rate": 4.2999999999999995e-06, "loss": 1.143, "step": 130 }, { "grad_norm": 2.668653964996338, "learning_rate": 4.633333333333334e-06, "loss": 1.158, "step": 140 }, { "grad_norm": 2.122525215148926, "learning_rate": 4.966666666666667e-06, "loss": 1.1478, "step": 150 }, { "grad_norm": 2.1425347328186035, "learning_rate": 5.3e-06, "loss": 1.1302, "step": 160 }, { "grad_norm": 2.0650012493133545, "learning_rate": 5.633333333333333e-06, "loss": 1.1246, "step": 170 }, { "grad_norm": 1.6851080656051636, "learning_rate": 5.9666666666666666e-06, "loss": 1.1116, "step": 180 }, { "grad_norm": 1.7610397338867188, "learning_rate": 6.300000000000001e-06, "loss": 1.1126, "step": 190 }, { "grad_norm": 1.8885430097579956, "learning_rate": 6.633333333333333e-06, "loss": 1.1239, "step": 200 }, { "grad_norm": 2.2790918350219727, "learning_rate": 6.966666666666667e-06, "loss": 1.1025, "step": 210 }, { "grad_norm": 2.064957857131958, "learning_rate": 7.2999999999999996e-06, "loss": 1.1113, "step": 220 }, { "grad_norm": 2.0368270874023438, "learning_rate": 7.633333333333334e-06, "loss": 1.1304, "step": 230 }, { "grad_norm": 1.4885128736495972, "learning_rate": 7.966666666666666e-06, "loss": 1.1222, "step": 240 }, { "grad_norm": 1.6127281188964844, "learning_rate": 8.3e-06, "loss": 1.1015, "step": 250 }, { "grad_norm": 1.5937691926956177, "learning_rate": 8.633333333333334e-06, "loss": 1.1053, "step": 260 }, { "grad_norm": 2.0972695350646973, "learning_rate": 8.966666666666668e-06, "loss": 1.1024, "step": 270 }, { "grad_norm": 2.5940635204315186, "learning_rate": 9.3e-06, "loss": 1.1039, "step": 280 }, { "grad_norm": 1.7880648374557495, "learning_rate": 9.633333333333335e-06, "loss": 1.0814, "step": 290 }, { "grad_norm": 2.851484537124634, "learning_rate": 9.966666666666667e-06, "loss": 1.0986, "step": 300 }, { "grad_norm": 2.311448097229004, "learning_rate": 1.03e-05, "loss": 1.0829, "step": 310 }, { "grad_norm": 1.655263900756836, "learning_rate": 1.0633333333333334e-05, "loss": 1.1, "step": 320 }, { "grad_norm": 1.829301118850708, "learning_rate": 1.0966666666666666e-05, "loss": 1.0539, "step": 330 }, { "grad_norm": 1.975527286529541, "learning_rate": 1.13e-05, "loss": 1.1154, "step": 340 }, { "grad_norm": 1.4539953470230103, "learning_rate": 1.1633333333333334e-05, "loss": 1.0992, "step": 350 }, { "grad_norm": 1.9147703647613525, "learning_rate": 1.1966666666666668e-05, "loss": 1.0653, "step": 360 }, { "grad_norm": 2.3578264713287354, "learning_rate": 1.23e-05, "loss": 1.0728, "step": 370 }, { "grad_norm": 1.6201541423797607, "learning_rate": 1.2633333333333333e-05, "loss": 1.0749, "step": 380 }, { "grad_norm": 1.6634047031402588, "learning_rate": 1.2966666666666669e-05, "loss": 1.0491, "step": 390 }, { "grad_norm": 1.9533647298812866, "learning_rate": 1.3300000000000001e-05, "loss": 1.071, "step": 400 }, { "grad_norm": 2.070437431335449, "learning_rate": 1.3633333333333334e-05, "loss": 1.0624, "step": 410 }, { "grad_norm": 2.3643250465393066, "learning_rate": 1.3966666666666666e-05, "loss": 1.0421, "step": 420 }, { "grad_norm": 2.0097718238830566, "learning_rate": 1.43e-05, "loss": 1.0051, "step": 430 }, { "grad_norm": 2.537001609802246, "learning_rate": 1.4633333333333334e-05, "loss": 1.0085, "step": 440 }, { "grad_norm": 1.653346300125122, "learning_rate": 1.4966666666666668e-05, "loss": 1.0129, "step": 450 }, { "grad_norm": 2.319932222366333, "learning_rate": 1.53e-05, "loss": 0.9997, "step": 460 }, { "grad_norm": 1.9820159673690796, "learning_rate": 1.563333333333333e-05, "loss": 0.9799, "step": 470 }, { "grad_norm": 2.0603318214416504, "learning_rate": 1.5966666666666667e-05, "loss": 0.9837, "step": 480 }, { "grad_norm": 2.2650177478790283, "learning_rate": 1.63e-05, "loss": 0.9719, "step": 490 }, { "grad_norm": 2.6378493309020996, "learning_rate": 1.6633333333333336e-05, "loss": 0.9023, "step": 500 }, { "grad_norm": 2.2493607997894287, "learning_rate": 1.6966666666666668e-05, "loss": 0.9292, "step": 510 }, { "grad_norm": 2.414140224456787, "learning_rate": 1.73e-05, "loss": 0.8731, "step": 520 }, { "grad_norm": 2.558781862258911, "learning_rate": 1.7633333333333336e-05, "loss": 0.8442, "step": 530 }, { "grad_norm": 2.6610965728759766, "learning_rate": 1.796666666666667e-05, "loss": 0.851, "step": 540 }, { "grad_norm": 2.4322686195373535, "learning_rate": 1.83e-05, "loss": 0.819, "step": 550 }, { "grad_norm": 2.274527072906494, "learning_rate": 1.8633333333333333e-05, "loss": 0.8261, "step": 560 }, { "grad_norm": 3.046337366104126, "learning_rate": 1.896666666666667e-05, "loss": 0.7404, "step": 570 }, { "grad_norm": 2.64178729057312, "learning_rate": 1.93e-05, "loss": 0.8198, "step": 580 }, { "grad_norm": 2.559373140335083, "learning_rate": 1.9633333333333334e-05, "loss": 0.7595, "step": 590 }, { "grad_norm": 2.7414515018463135, "learning_rate": 1.9966666666666666e-05, "loss": 0.7556, "step": 600 }, { "grad_norm": 2.7549822330474854, "learning_rate": 2.0300000000000002e-05, "loss": 0.7432, "step": 610 }, { "grad_norm": 3.193422317504883, "learning_rate": 2.0633333333333335e-05, "loss": 0.6808, "step": 620 }, { "grad_norm": 3.2745299339294434, "learning_rate": 2.0966666666666667e-05, "loss": 0.6572, "step": 630 }, { "grad_norm": 2.595874071121216, "learning_rate": 2.13e-05, "loss": 0.6352, "step": 640 }, { "grad_norm": 2.7696001529693604, "learning_rate": 2.1633333333333332e-05, "loss": 0.6357, "step": 650 }, { "grad_norm": 3.119767427444458, "learning_rate": 2.1966666666666668e-05, "loss": 0.6183, "step": 660 }, { "grad_norm": 2.9950528144836426, "learning_rate": 2.23e-05, "loss": 0.5874, "step": 670 }, { "grad_norm": 2.688882350921631, "learning_rate": 2.2633333333333336e-05, "loss": 0.6057, "step": 680 }, { "grad_norm": 3.2295682430267334, "learning_rate": 2.2966666666666668e-05, "loss": 0.5628, "step": 690 }, { "grad_norm": 2.627350091934204, "learning_rate": 2.3300000000000004e-05, "loss": 0.5887, "step": 700 }, { "grad_norm": 2.341167449951172, "learning_rate": 2.3633333333333336e-05, "loss": 0.5389, "step": 710 }, { "grad_norm": 3.1809232234954834, "learning_rate": 2.396666666666667e-05, "loss": 0.5223, "step": 720 }, { "grad_norm": 2.4792721271514893, "learning_rate": 2.43e-05, "loss": 0.5554, "step": 730 }, { "grad_norm": 2.609240770339966, "learning_rate": 2.4633333333333334e-05, "loss": 0.5423, "step": 740 }, { "grad_norm": 2.8362863063812256, "learning_rate": 2.496666666666667e-05, "loss": 0.5445, "step": 750 }, { "grad_norm": 2.2577967643737793, "learning_rate": 2.5300000000000002e-05, "loss": 0.516, "step": 760 }, { "grad_norm": 2.512136459350586, "learning_rate": 2.5633333333333338e-05, "loss": 0.5044, "step": 770 }, { "grad_norm": 2.6710665225982666, "learning_rate": 2.5966666666666667e-05, "loss": 0.5022, "step": 780 }, { "grad_norm": 3.10082745552063, "learning_rate": 2.6300000000000002e-05, "loss": 0.4713, "step": 790 }, { "grad_norm": 2.564647912979126, "learning_rate": 2.663333333333333e-05, "loss": 0.5081, "step": 800 }, { "grad_norm": 2.426943063735962, "learning_rate": 2.6966666666666667e-05, "loss": 0.4955, "step": 810 }, { "grad_norm": 3.015472888946533, "learning_rate": 2.7300000000000003e-05, "loss": 0.4701, "step": 820 }, { "grad_norm": 2.102405071258545, "learning_rate": 2.7633333333333332e-05, "loss": 0.4324, "step": 830 }, { "grad_norm": 2.773131847381592, "learning_rate": 2.7966666666666668e-05, "loss": 0.4612, "step": 840 }, { "grad_norm": 3.5257821083068848, "learning_rate": 2.83e-05, "loss": 0.4955, "step": 850 }, { "grad_norm": 2.9996273517608643, "learning_rate": 2.8633333333333336e-05, "loss": 0.4544, "step": 860 }, { "grad_norm": 2.172367811203003, "learning_rate": 2.8966666666666668e-05, "loss": 0.4772, "step": 870 }, { "grad_norm": 2.6747634410858154, "learning_rate": 2.93e-05, "loss": 0.4211, "step": 880 }, { "grad_norm": 1.99396812915802, "learning_rate": 2.9633333333333336e-05, "loss": 0.4148, "step": 890 }, { "grad_norm": 2.381593704223633, "learning_rate": 2.9966666666666672e-05, "loss": 0.3793, "step": 900 }, { "grad_norm": 2.8189287185668945, "learning_rate": 3.03e-05, "loss": 0.3965, "step": 910 }, { "grad_norm": 2.3474156856536865, "learning_rate": 3.063333333333334e-05, "loss": 0.3985, "step": 920 }, { "grad_norm": 2.237523078918457, "learning_rate": 3.096666666666666e-05, "loss": 0.3639, "step": 930 }, { "grad_norm": 3.592775344848633, "learning_rate": 3.13e-05, "loss": 0.3808, "step": 940 }, { "grad_norm": 3.1955320835113525, "learning_rate": 3.1633333333333334e-05, "loss": 0.3709, "step": 950 }, { "grad_norm": 2.37263560295105, "learning_rate": 3.196666666666667e-05, "loss": 0.3616, "step": 960 }, { "grad_norm": 2.3976762294769287, "learning_rate": 3.2300000000000006e-05, "loss": 0.3926, "step": 970 }, { "grad_norm": 2.7282660007476807, "learning_rate": 3.263333333333333e-05, "loss": 0.336, "step": 980 }, { "grad_norm": 2.566795825958252, "learning_rate": 3.296666666666667e-05, "loss": 0.3335, "step": 990 }, { "grad_norm": 3.930429697036743, "learning_rate": 3.33e-05, "loss": 0.3059, "step": 1000 }, { "grad_norm": 2.1986560821533203, "learning_rate": 3.3633333333333335e-05, "loss": 0.3245, "step": 1010 }, { "grad_norm": 2.560429811477661, "learning_rate": 3.396666666666667e-05, "loss": 0.3333, "step": 1020 }, { "grad_norm": 2.289466142654419, "learning_rate": 3.430000000000001e-05, "loss": 0.321, "step": 1030 }, { "grad_norm": 2.389036178588867, "learning_rate": 3.463333333333333e-05, "loss": 0.3276, "step": 1040 }, { "grad_norm": 3.1922836303710938, "learning_rate": 3.496666666666667e-05, "loss": 0.3479, "step": 1050 }, { "grad_norm": 2.1941797733306885, "learning_rate": 3.53e-05, "loss": 0.2967, "step": 1060 }, { "grad_norm": 3.215322256088257, "learning_rate": 3.563333333333334e-05, "loss": 0.3087, "step": 1070 }, { "grad_norm": 2.018450975418091, "learning_rate": 3.596666666666667e-05, "loss": 0.2466, "step": 1080 }, { "grad_norm": 2.1012258529663086, "learning_rate": 3.63e-05, "loss": 0.3475, "step": 1090 }, { "grad_norm": 2.211383581161499, "learning_rate": 3.6633333333333334e-05, "loss": 0.3196, "step": 1100 }, { "grad_norm": 2.2385342121124268, "learning_rate": 3.6966666666666666e-05, "loss": 0.3228, "step": 1110 }, { "grad_norm": 2.66436767578125, "learning_rate": 3.73e-05, "loss": 0.3097, "step": 1120 }, { "grad_norm": 2.174654722213745, "learning_rate": 3.763333333333334e-05, "loss": 0.2989, "step": 1130 }, { "grad_norm": 2.3299593925476074, "learning_rate": 3.796666666666667e-05, "loss": 0.249, "step": 1140 }, { "grad_norm": 1.6545144319534302, "learning_rate": 3.83e-05, "loss": 0.2651, "step": 1150 }, { "grad_norm": 2.2612054347991943, "learning_rate": 3.8633333333333335e-05, "loss": 0.2577, "step": 1160 }, { "grad_norm": 2.637275218963623, "learning_rate": 3.896666666666667e-05, "loss": 0.2755, "step": 1170 }, { "grad_norm": 1.8262406587600708, "learning_rate": 3.9300000000000007e-05, "loss": 0.2364, "step": 1180 }, { "grad_norm": 1.678684115409851, "learning_rate": 3.963333333333333e-05, "loss": 0.2531, "step": 1190 }, { "grad_norm": 2.1345136165618896, "learning_rate": 3.996666666666667e-05, "loss": 0.2817, "step": 1200 }, { "grad_norm": 1.6149238348007202, "learning_rate": 4.0300000000000004e-05, "loss": 0.2649, "step": 1210 }, { "grad_norm": 1.926194667816162, "learning_rate": 4.0633333333333336e-05, "loss": 0.2252, "step": 1220 }, { "grad_norm": 1.9099068641662598, "learning_rate": 4.096666666666667e-05, "loss": 0.237, "step": 1230 }, { "grad_norm": 2.5640950202941895, "learning_rate": 4.13e-05, "loss": 0.2279, "step": 1240 }, { "grad_norm": 2.5290775299072266, "learning_rate": 4.1633333333333333e-05, "loss": 0.2587, "step": 1250 }, { "grad_norm": 2.1171393394470215, "learning_rate": 4.196666666666667e-05, "loss": 0.2302, "step": 1260 }, { "grad_norm": 2.5540144443511963, "learning_rate": 4.23e-05, "loss": 0.2264, "step": 1270 }, { "grad_norm": 2.303940534591675, "learning_rate": 4.263333333333334e-05, "loss": 0.226, "step": 1280 }, { "grad_norm": 2.216931104660034, "learning_rate": 4.296666666666666e-05, "loss": 0.1835, "step": 1290 }, { "grad_norm": 3.2159268856048584, "learning_rate": 4.33e-05, "loss": 0.2378, "step": 1300 }, { "grad_norm": 1.8429259061813354, "learning_rate": 4.3633333333333335e-05, "loss": 0.2007, "step": 1310 }, { "grad_norm": 2.367526054382324, "learning_rate": 4.396666666666667e-05, "loss": 0.2429, "step": 1320 }, { "grad_norm": 1.8085427284240723, "learning_rate": 4.43e-05, "loss": 0.1704, "step": 1330 }, { "grad_norm": 2.0785200595855713, "learning_rate": 4.463333333333334e-05, "loss": 0.2059, "step": 1340 }, { "grad_norm": 1.4840792417526245, "learning_rate": 4.496666666666667e-05, "loss": 0.1863, "step": 1350 }, { "grad_norm": 2.150986433029175, "learning_rate": 4.53e-05, "loss": 0.2264, "step": 1360 }, { "grad_norm": 2.0212745666503906, "learning_rate": 4.5633333333333336e-05, "loss": 0.1968, "step": 1370 }, { "grad_norm": 1.990193486213684, "learning_rate": 4.596666666666667e-05, "loss": 0.1915, "step": 1380 }, { "grad_norm": 2.2553508281707764, "learning_rate": 4.630000000000001e-05, "loss": 0.2165, "step": 1390 }, { "grad_norm": 2.3329432010650635, "learning_rate": 4.663333333333333e-05, "loss": 0.2206, "step": 1400 }, { "grad_norm": 1.6625498533248901, "learning_rate": 4.696666666666667e-05, "loss": 0.184, "step": 1410 }, { "grad_norm": 1.8406716585159302, "learning_rate": 4.73e-05, "loss": 0.2193, "step": 1420 }, { "grad_norm": 2.5656871795654297, "learning_rate": 4.763333333333334e-05, "loss": 0.2175, "step": 1430 }, { "grad_norm": 2.0796287059783936, "learning_rate": 4.796666666666667e-05, "loss": 0.2206, "step": 1440 }, { "grad_norm": 1.2147517204284668, "learning_rate": 4.83e-05, "loss": 0.1849, "step": 1450 }, { "grad_norm": 1.89616858959198, "learning_rate": 4.8633333333333334e-05, "loss": 0.2116, "step": 1460 }, { "grad_norm": 1.6964056491851807, "learning_rate": 4.8966666666666667e-05, "loss": 0.184, "step": 1470 }, { "grad_norm": 1.8681403398513794, "learning_rate": 4.93e-05, "loss": 0.1789, "step": 1480 }, { "grad_norm": 1.6255918741226196, "learning_rate": 4.963333333333334e-05, "loss": 0.2129, "step": 1490 }, { "grad_norm": 2.009958267211914, "learning_rate": 4.996666666666667e-05, "loss": 0.1868, "step": 1500 }, { "grad_norm": 1.8057329654693604, "learning_rate": 5.03e-05, "loss": 0.2084, "step": 1510 }, { "grad_norm": 1.8598825931549072, "learning_rate": 5.0633333333333335e-05, "loss": 0.1954, "step": 1520 }, { "grad_norm": 1.648774266242981, "learning_rate": 5.0966666666666674e-05, "loss": 0.1859, "step": 1530 }, { "grad_norm": 1.7454534769058228, "learning_rate": 5.130000000000001e-05, "loss": 0.2349, "step": 1540 }, { "grad_norm": 1.1933789253234863, "learning_rate": 5.163333333333333e-05, "loss": 0.2381, "step": 1550 }, { "grad_norm": 1.8393455743789673, "learning_rate": 5.196666666666667e-05, "loss": 0.225, "step": 1560 }, { "grad_norm": 1.3674732446670532, "learning_rate": 5.2300000000000004e-05, "loss": 0.1871, "step": 1570 }, { "grad_norm": 1.7382276058197021, "learning_rate": 5.2633333333333336e-05, "loss": 0.2061, "step": 1580 }, { "grad_norm": 2.1424033641815186, "learning_rate": 5.296666666666666e-05, "loss": 0.1956, "step": 1590 }, { "grad_norm": 2.0484142303466797, "learning_rate": 5.330000000000001e-05, "loss": 0.1712, "step": 1600 }, { "grad_norm": 1.9689604043960571, "learning_rate": 5.3633333333333334e-05, "loss": 0.1916, "step": 1610 }, { "grad_norm": 2.4375762939453125, "learning_rate": 5.3966666666666666e-05, "loss": 0.2075, "step": 1620 }, { "grad_norm": 1.437064528465271, "learning_rate": 5.4300000000000005e-05, "loss": 0.1822, "step": 1630 }, { "grad_norm": 1.8707637786865234, "learning_rate": 5.463333333333334e-05, "loss": 0.204, "step": 1640 }, { "grad_norm": 2.2073628902435303, "learning_rate": 5.496666666666666e-05, "loss": 0.2281, "step": 1650 }, { "grad_norm": 1.5257865190505981, "learning_rate": 5.530000000000001e-05, "loss": 0.1796, "step": 1660 }, { "grad_norm": 1.420162320137024, "learning_rate": 5.5633333333333335e-05, "loss": 0.1714, "step": 1670 }, { "grad_norm": 1.9622031450271606, "learning_rate": 5.596666666666667e-05, "loss": 0.2156, "step": 1680 }, { "grad_norm": 1.5459860563278198, "learning_rate": 5.63e-05, "loss": 0.2043, "step": 1690 }, { "grad_norm": 2.0909063816070557, "learning_rate": 5.663333333333334e-05, "loss": 0.2233, "step": 1700 }, { "grad_norm": 1.5182287693023682, "learning_rate": 5.696666666666667e-05, "loss": 0.1667, "step": 1710 }, { "grad_norm": 1.2821073532104492, "learning_rate": 5.73e-05, "loss": 0.1888, "step": 1720 }, { "grad_norm": 1.5797686576843262, "learning_rate": 5.7633333333333336e-05, "loss": 0.2294, "step": 1730 }, { "grad_norm": 1.7305408716201782, "learning_rate": 5.796666666666667e-05, "loss": 0.1926, "step": 1740 }, { "grad_norm": 1.5775530338287354, "learning_rate": 5.83e-05, "loss": 0.192, "step": 1750 }, { "grad_norm": 1.6236004829406738, "learning_rate": 5.863333333333334e-05, "loss": 0.1936, "step": 1760 }, { "grad_norm": 2.172481060028076, "learning_rate": 5.896666666666667e-05, "loss": 0.2078, "step": 1770 }, { "grad_norm": 1.6984221935272217, "learning_rate": 5.93e-05, "loss": 0.1858, "step": 1780 }, { "grad_norm": 1.5792903900146484, "learning_rate": 5.9633333333333344e-05, "loss": 0.197, "step": 1790 }, { "grad_norm": 1.403723120689392, "learning_rate": 5.996666666666667e-05, "loss": 0.1908, "step": 1800 }, { "grad_norm": 1.2004247903823853, "learning_rate": 6.03e-05, "loss": 0.1748, "step": 1810 }, { "grad_norm": 1.473433017730713, "learning_rate": 6.063333333333333e-05, "loss": 0.2053, "step": 1820 }, { "grad_norm": 1.6193299293518066, "learning_rate": 6.0966666666666674e-05, "loss": 0.1712, "step": 1830 }, { "grad_norm": 1.1509960889816284, "learning_rate": 6.13e-05, "loss": 0.1734, "step": 1840 }, { "grad_norm": 1.506272792816162, "learning_rate": 6.163333333333333e-05, "loss": 0.1734, "step": 1850 }, { "grad_norm": 1.1984503269195557, "learning_rate": 6.196666666666668e-05, "loss": 0.2313, "step": 1860 }, { "grad_norm": 1.4665018320083618, "learning_rate": 6.23e-05, "loss": 0.2038, "step": 1870 }, { "grad_norm": 1.464088797569275, "learning_rate": 6.263333333333333e-05, "loss": 0.187, "step": 1880 }, { "grad_norm": 1.184190034866333, "learning_rate": 6.296666666666667e-05, "loss": 0.1651, "step": 1890 }, { "grad_norm": 1.2844144105911255, "learning_rate": 6.330000000000001e-05, "loss": 0.1706, "step": 1900 }, { "grad_norm": 1.2222779989242554, "learning_rate": 6.363333333333334e-05, "loss": 0.1969, "step": 1910 }, { "grad_norm": 1.4896514415740967, "learning_rate": 6.396666666666667e-05, "loss": 0.1752, "step": 1920 }, { "grad_norm": 1.7769428491592407, "learning_rate": 6.43e-05, "loss": 0.2344, "step": 1930 }, { "grad_norm": 1.4942071437835693, "learning_rate": 6.463333333333334e-05, "loss": 0.197, "step": 1940 }, { "grad_norm": 1.2786412239074707, "learning_rate": 6.496666666666667e-05, "loss": 0.1744, "step": 1950 }, { "grad_norm": 1.8109725713729858, "learning_rate": 6.53e-05, "loss": 0.1752, "step": 1960 }, { "grad_norm": 1.9360380172729492, "learning_rate": 6.563333333333333e-05, "loss": 0.1942, "step": 1970 }, { "grad_norm": 1.924268126487732, "learning_rate": 6.596666666666667e-05, "loss": 0.1732, "step": 1980 }, { "grad_norm": 1.8698092699050903, "learning_rate": 6.630000000000001e-05, "loss": 0.1909, "step": 1990 }, { "grad_norm": 1.3741596937179565, "learning_rate": 6.663333333333333e-05, "loss": 0.1787, "step": 2000 }, { "grad_norm": 1.2522257566452026, "learning_rate": 6.696666666666666e-05, "loss": 0.1674, "step": 2010 }, { "grad_norm": 1.3649369478225708, "learning_rate": 6.730000000000001e-05, "loss": 0.204, "step": 2020 }, { "grad_norm": 1.262458324432373, "learning_rate": 6.763333333333334e-05, "loss": 0.2262, "step": 2030 }, { "grad_norm": 1.5564751625061035, "learning_rate": 6.796666666666666e-05, "loss": 0.2134, "step": 2040 }, { "grad_norm": 1.458883285522461, "learning_rate": 6.83e-05, "loss": 0.1578, "step": 2050 }, { "grad_norm": 1.6852613687515259, "learning_rate": 6.863333333333334e-05, "loss": 0.2018, "step": 2060 }, { "grad_norm": 1.6998298168182373, "learning_rate": 6.896666666666667e-05, "loss": 0.1586, "step": 2070 }, { "grad_norm": 1.543869137763977, "learning_rate": 6.93e-05, "loss": 0.1877, "step": 2080 }, { "grad_norm": 1.3899046182632446, "learning_rate": 6.963333333333334e-05, "loss": 0.1984, "step": 2090 }, { "grad_norm": 1.4346624612808228, "learning_rate": 6.996666666666667e-05, "loss": 0.2021, "step": 2100 }, { "grad_norm": 1.0902423858642578, "learning_rate": 7.03e-05, "loss": 0.1936, "step": 2110 }, { "grad_norm": 1.1684595346450806, "learning_rate": 7.063333333333333e-05, "loss": 0.2134, "step": 2120 }, { "grad_norm": 1.4347823858261108, "learning_rate": 7.096666666666667e-05, "loss": 0.2082, "step": 2130 }, { "grad_norm": 1.6328954696655273, "learning_rate": 7.13e-05, "loss": 0.2174, "step": 2140 }, { "grad_norm": 1.0969256162643433, "learning_rate": 7.163333333333334e-05, "loss": 0.2084, "step": 2150 }, { "grad_norm": 1.1617375612258911, "learning_rate": 7.196666666666668e-05, "loss": 0.1894, "step": 2160 }, { "grad_norm": 1.148428201675415, "learning_rate": 7.23e-05, "loss": 0.1699, "step": 2170 }, { "grad_norm": 1.5584297180175781, "learning_rate": 7.263333333333334e-05, "loss": 0.2061, "step": 2180 }, { "grad_norm": 1.3074815273284912, "learning_rate": 7.296666666666667e-05, "loss": 0.1521, "step": 2190 }, { "grad_norm": 1.2664787769317627, "learning_rate": 7.33e-05, "loss": 0.1687, "step": 2200 }, { "grad_norm": 1.1734790802001953, "learning_rate": 7.363333333333334e-05, "loss": 0.1831, "step": 2210 }, { "grad_norm": 1.7685962915420532, "learning_rate": 7.396666666666667e-05, "loss": 0.226, "step": 2220 }, { "grad_norm": 1.8596911430358887, "learning_rate": 7.43e-05, "loss": 0.1848, "step": 2230 }, { "grad_norm": 1.215700626373291, "learning_rate": 7.463333333333334e-05, "loss": 0.1479, "step": 2240 }, { "grad_norm": 1.4416563510894775, "learning_rate": 7.496666666666667e-05, "loss": 0.1676, "step": 2250 }, { "grad_norm": 1.835780143737793, "learning_rate": 7.53e-05, "loss": 0.219, "step": 2260 }, { "grad_norm": 1.3527381420135498, "learning_rate": 7.563333333333333e-05, "loss": 0.2285, "step": 2270 }, { "grad_norm": 2.009838581085205, "learning_rate": 7.596666666666668e-05, "loss": 0.2037, "step": 2280 }, { "grad_norm": 1.3150686025619507, "learning_rate": 7.630000000000001e-05, "loss": 0.17, "step": 2290 }, { "grad_norm": 1.1283451318740845, "learning_rate": 7.663333333333333e-05, "loss": 0.2081, "step": 2300 }, { "grad_norm": 1.556504487991333, "learning_rate": 7.696666666666668e-05, "loss": 0.1702, "step": 2310 }, { "grad_norm": 1.0037450790405273, "learning_rate": 7.730000000000001e-05, "loss": 0.1898, "step": 2320 }, { "grad_norm": 1.3868718147277832, "learning_rate": 7.763333333333334e-05, "loss": 0.1549, "step": 2330 }, { "grad_norm": 1.3610843420028687, "learning_rate": 7.796666666666666e-05, "loss": 0.197, "step": 2340 }, { "grad_norm": 1.6211886405944824, "learning_rate": 7.83e-05, "loss": 0.1675, "step": 2350 }, { "grad_norm": 1.3890984058380127, "learning_rate": 7.863333333333334e-05, "loss": 0.1564, "step": 2360 }, { "grad_norm": 1.1569788455963135, "learning_rate": 7.896666666666667e-05, "loss": 0.2139, "step": 2370 }, { "grad_norm": 1.6783568859100342, "learning_rate": 7.93e-05, "loss": 0.2068, "step": 2380 }, { "grad_norm": 1.3619967699050903, "learning_rate": 7.963333333333334e-05, "loss": 0.1821, "step": 2390 }, { "grad_norm": 1.6114445924758911, "learning_rate": 7.996666666666667e-05, "loss": 0.2069, "step": 2400 }, { "grad_norm": 1.4113259315490723, "learning_rate": 8.030000000000001e-05, "loss": 0.2566, "step": 2410 }, { "grad_norm": 2.002521514892578, "learning_rate": 8.063333333333333e-05, "loss": 0.1811, "step": 2420 }, { "grad_norm": 1.5538158416748047, "learning_rate": 8.096666666666667e-05, "loss": 0.173, "step": 2430 }, { "grad_norm": 1.3147146701812744, "learning_rate": 8.13e-05, "loss": 0.1851, "step": 2440 }, { "grad_norm": 1.4110885858535767, "learning_rate": 8.163333333333334e-05, "loss": 0.2114, "step": 2450 }, { "grad_norm": 1.358155608177185, "learning_rate": 8.196666666666668e-05, "loss": 0.1755, "step": 2460 }, { "grad_norm": 1.671450138092041, "learning_rate": 8.23e-05, "loss": 0.2408, "step": 2470 }, { "grad_norm": 1.097693681716919, "learning_rate": 8.263333333333334e-05, "loss": 0.1558, "step": 2480 }, { "grad_norm": 1.1174920797348022, "learning_rate": 8.296666666666667e-05, "loss": 0.1564, "step": 2490 }, { "grad_norm": 0.992544949054718, "learning_rate": 8.33e-05, "loss": 0.1629, "step": 2500 }, { "grad_norm": 1.6821738481521606, "learning_rate": 8.363333333333334e-05, "loss": 0.1954, "step": 2510 }, { "grad_norm": 0.9200698137283325, "learning_rate": 8.396666666666667e-05, "loss": 0.1719, "step": 2520 }, { "grad_norm": 1.1778998374938965, "learning_rate": 8.43e-05, "loss": 0.1593, "step": 2530 }, { "grad_norm": 1.315022349357605, "learning_rate": 8.463333333333335e-05, "loss": 0.1352, "step": 2540 }, { "grad_norm": 1.3193336725234985, "learning_rate": 8.496666666666667e-05, "loss": 0.1713, "step": 2550 }, { "grad_norm": 1.1842132806777954, "learning_rate": 8.53e-05, "loss": 0.1838, "step": 2560 }, { "grad_norm": 1.4930399656295776, "learning_rate": 8.563333333333333e-05, "loss": 0.1905, "step": 2570 }, { "grad_norm": 1.2738417387008667, "learning_rate": 8.596666666666668e-05, "loss": 0.1543, "step": 2580 }, { "grad_norm": 1.4034218788146973, "learning_rate": 8.63e-05, "loss": 0.183, "step": 2590 }, { "grad_norm": 1.2637975215911865, "learning_rate": 8.663333333333333e-05, "loss": 0.1833, "step": 2600 }, { "grad_norm": 1.1489046812057495, "learning_rate": 8.696666666666668e-05, "loss": 0.1683, "step": 2610 }, { "grad_norm": 1.2608699798583984, "learning_rate": 8.730000000000001e-05, "loss": 0.1742, "step": 2620 }, { "grad_norm": 1.0893968343734741, "learning_rate": 8.763333333333334e-05, "loss": 0.1692, "step": 2630 }, { "grad_norm": 1.2186992168426514, "learning_rate": 8.796666666666667e-05, "loss": 0.1966, "step": 2640 }, { "grad_norm": 0.9464085698127747, "learning_rate": 8.83e-05, "loss": 0.1421, "step": 2650 }, { "grad_norm": 1.3399287462234497, "learning_rate": 8.863333333333334e-05, "loss": 0.1432, "step": 2660 }, { "grad_norm": 1.1298305988311768, "learning_rate": 8.896666666666667e-05, "loss": 0.1371, "step": 2670 }, { "grad_norm": 1.1170662641525269, "learning_rate": 8.93e-05, "loss": 0.17, "step": 2680 }, { "grad_norm": 0.9084122776985168, "learning_rate": 8.963333333333333e-05, "loss": 0.1615, "step": 2690 }, { "grad_norm": 0.831368088722229, "learning_rate": 8.996666666666667e-05, "loss": 0.2024, "step": 2700 }, { "grad_norm": 1.70795738697052, "learning_rate": 9.030000000000001e-05, "loss": 0.1876, "step": 2710 }, { "grad_norm": 1.2434526681900024, "learning_rate": 9.063333333333333e-05, "loss": 0.2008, "step": 2720 }, { "grad_norm": 1.146267294883728, "learning_rate": 9.096666666666666e-05, "loss": 0.1595, "step": 2730 }, { "grad_norm": 1.4028877019882202, "learning_rate": 9.130000000000001e-05, "loss": 0.163, "step": 2740 }, { "grad_norm": 1.067257046699524, "learning_rate": 9.163333333333334e-05, "loss": 0.1433, "step": 2750 }, { "grad_norm": 1.0510531663894653, "learning_rate": 9.196666666666666e-05, "loss": 0.1727, "step": 2760 }, { "grad_norm": 0.9621952772140503, "learning_rate": 9.230000000000001e-05, "loss": 0.162, "step": 2770 }, { "grad_norm": 1.2241133451461792, "learning_rate": 9.263333333333334e-05, "loss": 0.1726, "step": 2780 }, { "grad_norm": 1.0653706789016724, "learning_rate": 9.296666666666667e-05, "loss": 0.1413, "step": 2790 }, { "grad_norm": 0.9794275164604187, "learning_rate": 9.33e-05, "loss": 0.1702, "step": 2800 }, { "grad_norm": 0.8832419514656067, "learning_rate": 9.363333333333334e-05, "loss": 0.1636, "step": 2810 }, { "grad_norm": 1.6451667547225952, "learning_rate": 9.396666666666667e-05, "loss": 0.1796, "step": 2820 }, { "grad_norm": 1.0571362972259521, "learning_rate": 9.43e-05, "loss": 0.1573, "step": 2830 }, { "grad_norm": 1.220323085784912, "learning_rate": 9.463333333333333e-05, "loss": 0.1798, "step": 2840 }, { "grad_norm": 1.749788522720337, "learning_rate": 9.496666666666667e-05, "loss": 0.1512, "step": 2850 }, { "grad_norm": 1.0703591108322144, "learning_rate": 9.53e-05, "loss": 0.1407, "step": 2860 }, { "grad_norm": 1.4793922901153564, "learning_rate": 9.563333333333334e-05, "loss": 0.1784, "step": 2870 }, { "grad_norm": 1.2584705352783203, "learning_rate": 9.596666666666668e-05, "loss": 0.1803, "step": 2880 }, { "grad_norm": 0.9460548162460327, "learning_rate": 9.63e-05, "loss": 0.2105, "step": 2890 }, { "grad_norm": 0.9070188403129578, "learning_rate": 9.663333333333334e-05, "loss": 0.1635, "step": 2900 }, { "grad_norm": 0.9544190764427185, "learning_rate": 9.696666666666667e-05, "loss": 0.2146, "step": 2910 }, { "grad_norm": 1.3899562358856201, "learning_rate": 9.730000000000001e-05, "loss": 0.1512, "step": 2920 }, { "grad_norm": 0.9413778781890869, "learning_rate": 9.763333333333334e-05, "loss": 0.1759, "step": 2930 }, { "grad_norm": 0.7617103457450867, "learning_rate": 9.796666666666667e-05, "loss": 0.1601, "step": 2940 }, { "grad_norm": 1.0962172746658325, "learning_rate": 9.83e-05, "loss": 0.1633, "step": 2950 }, { "grad_norm": 1.0399597883224487, "learning_rate": 9.863333333333334e-05, "loss": 0.153, "step": 2960 }, { "grad_norm": 1.00263249874115, "learning_rate": 9.896666666666667e-05, "loss": 0.1485, "step": 2970 }, { "grad_norm": 0.8617696166038513, "learning_rate": 9.93e-05, "loss": 0.1694, "step": 2980 }, { "grad_norm": 1.0281848907470703, "learning_rate": 9.963333333333333e-05, "loss": 0.1609, "step": 2990 }, { "grad_norm": 0.9312397241592407, "learning_rate": 9.996666666666668e-05, "loss": 0.1601, "step": 3000 }, { "grad_norm": 1.0480811595916748, "learning_rate": 9.999999384858465e-05, "loss": 0.1664, "step": 3010 }, { "grad_norm": 1.3004205226898193, "learning_rate": 9.999997258443473e-05, "loss": 0.1582, "step": 3020 }, { "grad_norm": 1.066149115562439, "learning_rate": 9.999993613161331e-05, "loss": 0.1452, "step": 3030 }, { "grad_norm": 1.5506360530853271, "learning_rate": 9.999988449013146e-05, "loss": 0.1962, "step": 3040 }, { "grad_norm": 1.0999921560287476, "learning_rate": 9.99998176600049e-05, "loss": 0.16, "step": 3050 }, { "grad_norm": 1.336068034172058, "learning_rate": 9.999973564125389e-05, "loss": 0.1808, "step": 3060 }, { "grad_norm": 1.207828164100647, "learning_rate": 9.999963843390335e-05, "loss": 0.1715, "step": 3070 }, { "grad_norm": 0.8893445730209351, "learning_rate": 9.999952603798282e-05, "loss": 0.1521, "step": 3080 }, { "grad_norm": 1.1680717468261719, "learning_rate": 9.999939845352646e-05, "loss": 0.1607, "step": 3090 }, { "grad_norm": 1.1402702331542969, "learning_rate": 9.999925568057298e-05, "loss": 0.1846, "step": 3100 }, { "grad_norm": 1.4598394632339478, "learning_rate": 9.999909771916578e-05, "loss": 0.1978, "step": 3110 }, { "grad_norm": 0.8026534914970398, "learning_rate": 9.999892456935285e-05, "loss": 0.1503, "step": 3120 }, { "grad_norm": 0.946467399597168, "learning_rate": 9.999873623118679e-05, "loss": 0.14, "step": 3130 }, { "grad_norm": 1.1587375402450562, "learning_rate": 9.999853270472479e-05, "loss": 0.1574, "step": 3140 }, { "grad_norm": 1.057289481163025, "learning_rate": 9.999831399002871e-05, "loss": 0.1434, "step": 3150 }, { "grad_norm": 0.9652853012084961, "learning_rate": 9.999808008716494e-05, "loss": 0.1786, "step": 3160 }, { "grad_norm": 1.1347383260726929, "learning_rate": 9.999783099620459e-05, "loss": 0.1717, "step": 3170 }, { "grad_norm": 0.9869375824928284, "learning_rate": 9.999756671722328e-05, "loss": 0.1849, "step": 3180 }, { "grad_norm": 0.9262442588806152, "learning_rate": 9.99972872503013e-05, "loss": 0.1621, "step": 3190 }, { "grad_norm": 0.9116818904876709, "learning_rate": 9.999699259552359e-05, "loss": 0.1441, "step": 3200 }, { "grad_norm": 0.9979340434074402, "learning_rate": 9.99966827529796e-05, "loss": 0.1374, "step": 3210 }, { "grad_norm": 0.9715573787689209, "learning_rate": 9.999635772276348e-05, "loss": 0.1581, "step": 3220 }, { "grad_norm": 1.2980350255966187, "learning_rate": 9.999601750497396e-05, "loss": 0.1608, "step": 3230 }, { "grad_norm": 1.2748390436172485, "learning_rate": 9.99956620997144e-05, "loss": 0.1835, "step": 3240 }, { "grad_norm": 1.0427790880203247, "learning_rate": 9.999529150709275e-05, "loss": 0.2091, "step": 3250 }, { "grad_norm": 0.9368738532066345, "learning_rate": 9.999490572722158e-05, "loss": 0.2542, "step": 3260 }, { "grad_norm": 0.9978147149085999, "learning_rate": 9.99945047602181e-05, "loss": 0.1914, "step": 3270 }, { "grad_norm": 1.2050979137420654, "learning_rate": 9.99940886062041e-05, "loss": 0.2013, "step": 3280 }, { "grad_norm": 1.110849380493164, "learning_rate": 9.999365726530599e-05, "loss": 0.1835, "step": 3290 }, { "grad_norm": 1.2743464708328247, "learning_rate": 9.999321073765481e-05, "loss": 0.1873, "step": 3300 }, { "grad_norm": 1.2437820434570312, "learning_rate": 9.99927490233862e-05, "loss": 0.1581, "step": 3310 }, { "grad_norm": 0.9865133762359619, "learning_rate": 9.999227212264043e-05, "loss": 0.167, "step": 3320 }, { "grad_norm": 1.5543365478515625, "learning_rate": 9.999178003556236e-05, "loss": 0.1537, "step": 3330 }, { "grad_norm": 1.0876944065093994, "learning_rate": 9.999127276230146e-05, "loss": 0.135, "step": 3340 }, { "grad_norm": 1.004517674446106, "learning_rate": 9.999075030301184e-05, "loss": 0.1403, "step": 3350 }, { "grad_norm": 1.7812495231628418, "learning_rate": 9.999021265785221e-05, "loss": 0.1533, "step": 3360 }, { "grad_norm": 1.080330491065979, "learning_rate": 9.998965982698589e-05, "loss": 0.1561, "step": 3370 }, { "grad_norm": 1.0812004804611206, "learning_rate": 9.998909181058082e-05, "loss": 0.1582, "step": 3380 }, { "grad_norm": 0.9739291667938232, "learning_rate": 9.998850860880953e-05, "loss": 0.1569, "step": 3390 }, { "grad_norm": 1.1558170318603516, "learning_rate": 9.998791022184922e-05, "loss": 0.1613, "step": 3400 }, { "grad_norm": 1.4165440797805786, "learning_rate": 9.99872966498816e-05, "loss": 0.159, "step": 3410 }, { "grad_norm": 1.2623534202575684, "learning_rate": 9.998666789309313e-05, "loss": 0.2283, "step": 3420 }, { "grad_norm": 1.2889786958694458, "learning_rate": 9.998602395167475e-05, "loss": 0.186, "step": 3430 }, { "grad_norm": 1.1344654560089111, "learning_rate": 9.998536482582213e-05, "loss": 0.1648, "step": 3440 }, { "grad_norm": 1.2598683834075928, "learning_rate": 9.998469051573544e-05, "loss": 0.1769, "step": 3450 }, { "grad_norm": 1.1623541116714478, "learning_rate": 9.998400102161954e-05, "loss": 0.1666, "step": 3460 }, { "grad_norm": 1.093680739402771, "learning_rate": 9.998329634368388e-05, "loss": 0.1884, "step": 3470 }, { "grad_norm": 1.0887603759765625, "learning_rate": 9.998257648214253e-05, "loss": 0.1379, "step": 3480 }, { "grad_norm": 1.3493725061416626, "learning_rate": 9.998184143721417e-05, "loss": 0.1631, "step": 3490 }, { "grad_norm": 1.2648531198501587, "learning_rate": 9.998109120912206e-05, "loss": 0.1783, "step": 3500 }, { "grad_norm": 1.3045872449874878, "learning_rate": 9.998032579809411e-05, "loss": 0.1686, "step": 3510 }, { "grad_norm": 1.5290392637252808, "learning_rate": 9.997954520436286e-05, "loss": 0.1714, "step": 3520 }, { "grad_norm": 0.8979008793830872, "learning_rate": 9.997874942816538e-05, "loss": 0.1594, "step": 3530 }, { "grad_norm": 1.2230463027954102, "learning_rate": 9.997793846974345e-05, "loss": 0.1585, "step": 3540 }, { "grad_norm": 1.278425931930542, "learning_rate": 9.997711232934341e-05, "loss": 0.1724, "step": 3550 }, { "grad_norm": 1.1165704727172852, "learning_rate": 9.99762710072162e-05, "loss": 0.1632, "step": 3560 }, { "grad_norm": 1.2317882776260376, "learning_rate": 9.997541450361743e-05, "loss": 0.142, "step": 3570 }, { "grad_norm": 1.2039647102355957, "learning_rate": 9.997454281880723e-05, "loss": 0.177, "step": 3580 }, { "grad_norm": 1.1840258836746216, "learning_rate": 9.997365595305044e-05, "loss": 0.1642, "step": 3590 }, { "grad_norm": 1.1801154613494873, "learning_rate": 9.997275390661644e-05, "loss": 0.1817, "step": 3600 }, { "grad_norm": 1.1376960277557373, "learning_rate": 9.997183667977926e-05, "loss": 0.1519, "step": 3610 }, { "grad_norm": 0.8367563486099243, "learning_rate": 9.997090427281752e-05, "loss": 0.1397, "step": 3620 }, { "grad_norm": 1.020606517791748, "learning_rate": 9.996995668601448e-05, "loss": 0.1472, "step": 3630 }, { "grad_norm": 1.5083249807357788, "learning_rate": 9.996899391965798e-05, "loss": 0.191, "step": 3640 }, { "grad_norm": 0.8535178303718567, "learning_rate": 9.996801597404048e-05, "loss": 0.1416, "step": 3650 }, { "grad_norm": 1.8607498407363892, "learning_rate": 9.996702284945905e-05, "loss": 0.203, "step": 3660 }, { "grad_norm": 1.324091911315918, "learning_rate": 9.996601454621539e-05, "loss": 0.1799, "step": 3670 }, { "grad_norm": 1.2586108446121216, "learning_rate": 9.996499106461577e-05, "loss": 0.1513, "step": 3680 }, { "grad_norm": 1.3203495740890503, "learning_rate": 9.996395240497112e-05, "loss": 0.2023, "step": 3690 }, { "grad_norm": 0.8253712058067322, "learning_rate": 9.996289856759696e-05, "loss": 0.1693, "step": 3700 }, { "grad_norm": 1.1873981952667236, "learning_rate": 9.996182955281342e-05, "loss": 0.1601, "step": 3710 }, { "grad_norm": 0.8111411333084106, "learning_rate": 9.996074536094519e-05, "loss": 0.2266, "step": 3720 }, { "grad_norm": 0.9365240931510925, "learning_rate": 9.995964599232168e-05, "loss": 0.1392, "step": 3730 }, { "grad_norm": 0.9519729614257812, "learning_rate": 9.995853144727683e-05, "loss": 0.144, "step": 3740 }, { "grad_norm": 0.8570987582206726, "learning_rate": 9.99574017261492e-05, "loss": 0.1139, "step": 3750 }, { "grad_norm": 1.0131452083587646, "learning_rate": 9.995625682928198e-05, "loss": 0.1424, "step": 3760 }, { "grad_norm": 0.8465255498886108, "learning_rate": 9.995509675702295e-05, "loss": 0.1573, "step": 3770 }, { "grad_norm": 1.2648824453353882, "learning_rate": 9.995392150972451e-05, "loss": 0.15, "step": 3780 }, { "grad_norm": 0.9554521441459656, "learning_rate": 9.995273108774366e-05, "loss": 0.1561, "step": 3790 }, { "grad_norm": 1.054016351699829, "learning_rate": 9.995152549144205e-05, "loss": 0.1326, "step": 3800 }, { "grad_norm": 1.0674822330474854, "learning_rate": 9.995030472118587e-05, "loss": 0.1607, "step": 3810 }, { "grad_norm": 1.0807862281799316, "learning_rate": 9.9949068777346e-05, "loss": 0.1841, "step": 3820 }, { "grad_norm": 1.1157522201538086, "learning_rate": 9.994781766029786e-05, "loss": 0.1742, "step": 3830 }, { "grad_norm": 0.9788591265678406, "learning_rate": 9.994655137042151e-05, "loss": 0.1578, "step": 3840 }, { "grad_norm": 0.994210958480835, "learning_rate": 9.99452699081016e-05, "loss": 0.1358, "step": 3850 }, { "grad_norm": 1.063533902168274, "learning_rate": 9.994397327372743e-05, "loss": 0.1212, "step": 3860 }, { "grad_norm": 1.3037290573120117, "learning_rate": 9.994266146769286e-05, "loss": 0.1465, "step": 3870 }, { "grad_norm": 1.3289448022842407, "learning_rate": 9.994133449039642e-05, "loss": 0.1593, "step": 3880 }, { "grad_norm": 0.9148763418197632, "learning_rate": 9.993999234224118e-05, "loss": 0.165, "step": 3890 }, { "grad_norm": 0.9643988013267517, "learning_rate": 9.993863502363485e-05, "loss": 0.1383, "step": 3900 }, { "grad_norm": 0.9145092368125916, "learning_rate": 9.993726253498976e-05, "loss": 0.1683, "step": 3910 }, { "grad_norm": 0.7874583005905151, "learning_rate": 9.993587487672282e-05, "loss": 0.1698, "step": 3920 }, { "grad_norm": 0.856353223323822, "learning_rate": 9.993447204925558e-05, "loss": 0.1193, "step": 3930 }, { "grad_norm": 1.4501980543136597, "learning_rate": 9.993305405301416e-05, "loss": 0.1497, "step": 3940 }, { "grad_norm": 0.647318422794342, "learning_rate": 9.993162088842935e-05, "loss": 0.1515, "step": 3950 }, { "grad_norm": 0.8091772198677063, "learning_rate": 9.993017255593646e-05, "loss": 0.1669, "step": 3960 }, { "grad_norm": 0.8301352858543396, "learning_rate": 9.992870905597548e-05, "loss": 0.1427, "step": 3970 }, { "grad_norm": 0.9804688692092896, "learning_rate": 9.9927230388991e-05, "loss": 0.1256, "step": 3980 }, { "grad_norm": 1.361372947692871, "learning_rate": 9.992573655543215e-05, "loss": 0.1799, "step": 3990 }, { "grad_norm": 0.9998408555984497, "learning_rate": 9.992422755575277e-05, "loss": 0.1451, "step": 4000 }, { "grad_norm": 0.919647753238678, "learning_rate": 9.992270339041123e-05, "loss": 0.1777, "step": 4010 }, { "grad_norm": 1.0538612604141235, "learning_rate": 9.992116405987053e-05, "loss": 0.21, "step": 4020 }, { "grad_norm": 0.9023498892784119, "learning_rate": 9.991960956459828e-05, "loss": 0.1236, "step": 4030 }, { "grad_norm": 1.0260676145553589, "learning_rate": 9.991803990506669e-05, "loss": 0.1589, "step": 4040 }, { "grad_norm": 0.9463359713554382, "learning_rate": 9.991645508175258e-05, "loss": 0.1131, "step": 4050 }, { "grad_norm": 1.0193088054656982, "learning_rate": 9.99148550951374e-05, "loss": 0.1268, "step": 4060 }, { "grad_norm": 1.0368623733520508, "learning_rate": 9.991323994570716e-05, "loss": 0.1315, "step": 4070 }, { "grad_norm": 0.7464827299118042, "learning_rate": 9.99116096339525e-05, "loss": 0.1596, "step": 4080 }, { "grad_norm": 0.9863251447677612, "learning_rate": 9.990996416036869e-05, "loss": 0.1397, "step": 4090 }, { "grad_norm": 0.9800474047660828, "learning_rate": 9.990830352545555e-05, "loss": 0.1674, "step": 4100 }, { "grad_norm": 0.7018178701400757, "learning_rate": 9.990662772971756e-05, "loss": 0.1244, "step": 4110 }, { "grad_norm": 0.9006850719451904, "learning_rate": 9.990493677366376e-05, "loss": 0.1267, "step": 4120 }, { "grad_norm": 1.188688039779663, "learning_rate": 9.990323065780786e-05, "loss": 0.1364, "step": 4130 }, { "grad_norm": 0.7225694060325623, "learning_rate": 9.990150938266808e-05, "loss": 0.13, "step": 4140 }, { "grad_norm": 1.199400782585144, "learning_rate": 9.989977294876733e-05, "loss": 0.128, "step": 4150 }, { "grad_norm": 0.7754482626914978, "learning_rate": 9.989802135663308e-05, "loss": 0.1365, "step": 4160 }, { "grad_norm": 0.6429449915885925, "learning_rate": 9.989625460679743e-05, "loss": 0.1123, "step": 4170 }, { "grad_norm": 0.9452517628669739, "learning_rate": 9.989447269979706e-05, "loss": 0.1609, "step": 4180 }, { "grad_norm": 1.0701370239257812, "learning_rate": 9.989267563617328e-05, "loss": 0.1355, "step": 4190 }, { "grad_norm": 1.23720121383667, "learning_rate": 9.989086341647198e-05, "loss": 0.1818, "step": 4200 }, { "grad_norm": 0.8172138929367065, "learning_rate": 9.988903604124366e-05, "loss": 0.174, "step": 4210 }, { "grad_norm": 0.7904132604598999, "learning_rate": 9.988719351104343e-05, "loss": 0.1673, "step": 4220 }, { "grad_norm": 1.2381430864334106, "learning_rate": 9.9885335826431e-05, "loss": 0.139, "step": 4230 }, { "grad_norm": 1.1177663803100586, "learning_rate": 9.988346298797071e-05, "loss": 0.1383, "step": 4240 }, { "grad_norm": 0.9486089944839478, "learning_rate": 9.988157499623146e-05, "loss": 0.1575, "step": 4250 }, { "grad_norm": 0.6075055599212646, "learning_rate": 9.987967185178677e-05, "loss": 0.1417, "step": 4260 }, { "grad_norm": 0.9821386933326721, "learning_rate": 9.987775355521476e-05, "loss": 0.1191, "step": 4270 }, { "grad_norm": 0.9779589772224426, "learning_rate": 9.987582010709817e-05, "loss": 0.1101, "step": 4280 }, { "grad_norm": 0.7849007844924927, "learning_rate": 9.987387150802431e-05, "loss": 0.1613, "step": 4290 }, { "grad_norm": 0.8891391158103943, "learning_rate": 9.987190775858517e-05, "loss": 0.1382, "step": 4300 }, { "grad_norm": 0.9614422917366028, "learning_rate": 9.98699288593772e-05, "loss": 0.1288, "step": 4310 }, { "grad_norm": 0.8840270042419434, "learning_rate": 9.986793481100161e-05, "loss": 0.1415, "step": 4320 }, { "grad_norm": 1.069704294204712, "learning_rate": 9.986592561406412e-05, "loss": 0.1469, "step": 4330 }, { "grad_norm": 1.3655332326889038, "learning_rate": 9.986390126917503e-05, "loss": 0.1309, "step": 4340 }, { "grad_norm": 1.384142279624939, "learning_rate": 9.986186177694933e-05, "loss": 0.154, "step": 4350 }, { "grad_norm": 0.8843961954116821, "learning_rate": 9.985980713800656e-05, "loss": 0.179, "step": 4360 }, { "grad_norm": 0.8386816382408142, "learning_rate": 9.985773735297084e-05, "loss": 0.1293, "step": 4370 }, { "grad_norm": 1.5717746019363403, "learning_rate": 9.985565242247092e-05, "loss": 0.1311, "step": 4380 }, { "grad_norm": 0.7808408141136169, "learning_rate": 9.985355234714016e-05, "loss": 0.1455, "step": 4390 }, { "grad_norm": 0.9247413873672485, "learning_rate": 9.985143712761652e-05, "loss": 0.1313, "step": 4400 }, { "grad_norm": 1.0769366025924683, "learning_rate": 9.984930676454252e-05, "loss": 0.1386, "step": 4410 }, { "grad_norm": 0.8255804777145386, "learning_rate": 9.984716125856532e-05, "loss": 0.1256, "step": 4420 }, { "grad_norm": 0.9621076583862305, "learning_rate": 9.984500061033667e-05, "loss": 0.1337, "step": 4430 }, { "grad_norm": 0.9425228238105774, "learning_rate": 9.984282482051293e-05, "loss": 0.1319, "step": 4440 }, { "grad_norm": 0.9578531384468079, "learning_rate": 9.9840633889755e-05, "loss": 0.1228, "step": 4450 }, { "grad_norm": 1.1858348846435547, "learning_rate": 9.983842781872848e-05, "loss": 0.163, "step": 4460 }, { "grad_norm": 1.1145365238189697, "learning_rate": 9.98362066081035e-05, "loss": 0.1384, "step": 4470 }, { "grad_norm": 0.9318302869796753, "learning_rate": 9.983397025855479e-05, "loss": 0.1278, "step": 4480 }, { "grad_norm": 1.0129115581512451, "learning_rate": 9.983171877076171e-05, "loss": 0.144, "step": 4490 }, { "grad_norm": 0.91416335105896, "learning_rate": 9.98294521454082e-05, "loss": 0.1451, "step": 4500 }, { "grad_norm": 0.9135456681251526, "learning_rate": 9.98271703831828e-05, "loss": 0.1464, "step": 4510 }, { "grad_norm": 1.0569438934326172, "learning_rate": 9.982487348477865e-05, "loss": 0.1307, "step": 4520 }, { "grad_norm": 0.8796835541725159, "learning_rate": 9.982256145089347e-05, "loss": 0.121, "step": 4530 }, { "grad_norm": 0.9559353590011597, "learning_rate": 9.982023428222962e-05, "loss": 0.1227, "step": 4540 }, { "grad_norm": 0.8963595628738403, "learning_rate": 9.981789197949403e-05, "loss": 0.1185, "step": 4550 }, { "grad_norm": 0.6858589053153992, "learning_rate": 9.98155345433982e-05, "loss": 0.1324, "step": 4560 }, { "grad_norm": 0.9900723099708557, "learning_rate": 9.981316197465831e-05, "loss": 0.1557, "step": 4570 }, { "grad_norm": 1.1511123180389404, "learning_rate": 9.981077427399504e-05, "loss": 0.1465, "step": 4580 }, { "grad_norm": 0.9604876041412354, "learning_rate": 9.980837144213371e-05, "loss": 0.111, "step": 4590 }, { "grad_norm": 0.9627408981323242, "learning_rate": 9.980595347980426e-05, "loss": 0.1139, "step": 4600 }, { "grad_norm": 0.9142653942108154, "learning_rate": 9.980352038774119e-05, "loss": 0.129, "step": 4610 }, { "grad_norm": 0.5698005557060242, "learning_rate": 9.98010721666836e-05, "loss": 0.1314, "step": 4620 }, { "grad_norm": 0.9347764253616333, "learning_rate": 9.979860881737523e-05, "loss": 0.1247, "step": 4630 }, { "grad_norm": 0.9923544526100159, "learning_rate": 9.979613034056434e-05, "loss": 0.1031, "step": 4640 }, { "grad_norm": 1.2839694023132324, "learning_rate": 9.979363673700386e-05, "loss": 0.1467, "step": 4650 }, { "grad_norm": 0.913835346698761, "learning_rate": 9.979112800745124e-05, "loss": 0.1366, "step": 4660 }, { "grad_norm": 1.010643720626831, "learning_rate": 9.978860415266861e-05, "loss": 0.1468, "step": 4670 }, { "grad_norm": 0.8059553503990173, "learning_rate": 9.978606517342262e-05, "loss": 0.1081, "step": 4680 }, { "grad_norm": 0.7374359369277954, "learning_rate": 9.978351107048456e-05, "loss": 0.1238, "step": 4690 }, { "grad_norm": 0.9672989845275879, "learning_rate": 9.978094184463029e-05, "loss": 0.146, "step": 4700 }, { "grad_norm": 1.0221590995788574, "learning_rate": 9.977835749664029e-05, "loss": 0.1393, "step": 4710 }, { "grad_norm": 1.0615794658660889, "learning_rate": 9.97757580272996e-05, "loss": 0.1538, "step": 4720 }, { "grad_norm": 0.7199716567993164, "learning_rate": 9.977314343739786e-05, "loss": 0.1385, "step": 4730 }, { "grad_norm": 0.7785969376564026, "learning_rate": 9.977051372772934e-05, "loss": 0.1315, "step": 4740 }, { "grad_norm": 0.8504855632781982, "learning_rate": 9.976786889909286e-05, "loss": 0.1378, "step": 4750 }, { "grad_norm": 0.7418860197067261, "learning_rate": 9.976520895229185e-05, "loss": 0.1407, "step": 4760 }, { "grad_norm": 0.7254473567008972, "learning_rate": 9.976253388813433e-05, "loss": 0.1426, "step": 4770 }, { "grad_norm": 0.855712890625, "learning_rate": 9.975984370743293e-05, "loss": 0.1392, "step": 4780 }, { "grad_norm": 1.009682059288025, "learning_rate": 9.975713841100485e-05, "loss": 0.1491, "step": 4790 }, { "grad_norm": 1.0224807262420654, "learning_rate": 9.975441799967187e-05, "loss": 0.1212, "step": 4800 }, { "grad_norm": 0.7044382691383362, "learning_rate": 9.975168247426039e-05, "loss": 0.1186, "step": 4810 }, { "grad_norm": 0.8178209066390991, "learning_rate": 9.974893183560139e-05, "loss": 0.1177, "step": 4820 }, { "grad_norm": 0.8926230072975159, "learning_rate": 9.974616608453045e-05, "loss": 0.1744, "step": 4830 }, { "grad_norm": 1.0079408884048462, "learning_rate": 9.974338522188772e-05, "loss": 0.1632, "step": 4840 }, { "grad_norm": 1.280489444732666, "learning_rate": 9.974058924851797e-05, "loss": 0.146, "step": 4850 }, { "grad_norm": 0.9153552651405334, "learning_rate": 9.973777816527051e-05, "loss": 0.1434, "step": 4860 }, { "grad_norm": 0.9673671126365662, "learning_rate": 9.973495197299931e-05, "loss": 0.148, "step": 4870 }, { "grad_norm": 0.739969789981842, "learning_rate": 9.973211067256287e-05, "loss": 0.1368, "step": 4880 }, { "grad_norm": 1.0197093486785889, "learning_rate": 9.97292542648243e-05, "loss": 0.1168, "step": 4890 }, { "grad_norm": 0.849457323551178, "learning_rate": 9.972638275065131e-05, "loss": 0.1211, "step": 4900 }, { "grad_norm": 0.8156745433807373, "learning_rate": 9.972349613091621e-05, "loss": 0.1295, "step": 4910 }, { "grad_norm": 0.8606939315795898, "learning_rate": 9.972059440649584e-05, "loss": 0.1273, "step": 4920 }, { "grad_norm": 1.0865449905395508, "learning_rate": 9.971767757827168e-05, "loss": 0.1237, "step": 4930 }, { "grad_norm": 0.9661386013031006, "learning_rate": 9.971474564712982e-05, "loss": 0.1112, "step": 4940 }, { "grad_norm": 0.6129279136657715, "learning_rate": 9.971179861396084e-05, "loss": 0.1435, "step": 4950 }, { "grad_norm": 0.9329882860183716, "learning_rate": 9.970883647966003e-05, "loss": 0.1421, "step": 4960 }, { "grad_norm": 1.0164085626602173, "learning_rate": 9.970585924512717e-05, "loss": 0.1286, "step": 4970 }, { "grad_norm": 0.9328160285949707, "learning_rate": 9.970286691126669e-05, "loss": 0.1236, "step": 4980 }, { "grad_norm": 0.7896714806556702, "learning_rate": 9.969985947898756e-05, "loss": 0.1155, "step": 4990 }, { "grad_norm": 0.7160114645957947, "learning_rate": 9.969683694920337e-05, "loss": 0.1066, "step": 5000 }, { "grad_norm": 0.7038450837135315, "learning_rate": 9.969379932283228e-05, "loss": 0.1042, "step": 5010 }, { "grad_norm": 1.1386280059814453, "learning_rate": 9.969074660079704e-05, "loss": 0.1345, "step": 5020 }, { "grad_norm": 0.8809119462966919, "learning_rate": 9.968767878402501e-05, "loss": 0.127, "step": 5030 }, { "grad_norm": 0.7871602773666382, "learning_rate": 9.968459587344808e-05, "loss": 0.1045, "step": 5040 }, { "grad_norm": 0.994079053401947, "learning_rate": 9.968149787000278e-05, "loss": 0.117, "step": 5050 }, { "grad_norm": 0.7957148551940918, "learning_rate": 9.967838477463018e-05, "loss": 0.1173, "step": 5060 }, { "grad_norm": 1.1894958019256592, "learning_rate": 9.967525658827597e-05, "loss": 0.1316, "step": 5070 }, { "grad_norm": 0.7502481341362, "learning_rate": 9.967211331189042e-05, "loss": 0.1097, "step": 5080 }, { "grad_norm": 0.971845269203186, "learning_rate": 9.966895494642834e-05, "loss": 0.1493, "step": 5090 }, { "grad_norm": 0.9532384872436523, "learning_rate": 9.96657814928492e-05, "loss": 0.1118, "step": 5100 }, { "grad_norm": 0.8255453109741211, "learning_rate": 9.966259295211697e-05, "loss": 0.1507, "step": 5110 }, { "grad_norm": 1.1289563179016113, "learning_rate": 9.965938932520028e-05, "loss": 0.1176, "step": 5120 }, { "grad_norm": 1.0836944580078125, "learning_rate": 9.965617061307229e-05, "loss": 0.1383, "step": 5130 }, { "grad_norm": 0.7325627207756042, "learning_rate": 9.965293681671077e-05, "loss": 0.1397, "step": 5140 }, { "grad_norm": 0.7156515717506409, "learning_rate": 9.964968793709804e-05, "loss": 0.1624, "step": 5150 }, { "grad_norm": 0.7346403002738953, "learning_rate": 9.964642397522106e-05, "loss": 0.1265, "step": 5160 }, { "grad_norm": 1.3100234270095825, "learning_rate": 9.96431449320713e-05, "loss": 0.1459, "step": 5170 }, { "grad_norm": 0.7381737232208252, "learning_rate": 9.963985080864486e-05, "loss": 0.1166, "step": 5180 }, { "grad_norm": 0.9554561972618103, "learning_rate": 9.96365416059424e-05, "loss": 0.1635, "step": 5190 }, { "grad_norm": 0.8282257318496704, "learning_rate": 9.963321732496919e-05, "loss": 0.1154, "step": 5200 }, { "grad_norm": 1.1672813892364502, "learning_rate": 9.962987796673506e-05, "loss": 0.1131, "step": 5210 }, { "grad_norm": 1.0208369493484497, "learning_rate": 9.962652353225438e-05, "loss": 0.152, "step": 5220 }, { "grad_norm": 0.9838911294937134, "learning_rate": 9.962315402254619e-05, "loss": 0.1219, "step": 5230 }, { "grad_norm": 0.8798495531082153, "learning_rate": 9.9619769438634e-05, "loss": 0.1126, "step": 5240 }, { "grad_norm": 0.6979522705078125, "learning_rate": 9.9616369781546e-05, "loss": 0.1259, "step": 5250 }, { "grad_norm": 0.7725344896316528, "learning_rate": 9.961295505231491e-05, "loss": 0.1389, "step": 5260 }, { "grad_norm": 0.8738031983375549, "learning_rate": 9.960952525197804e-05, "loss": 0.1255, "step": 5270 }, { "grad_norm": 0.7463611364364624, "learning_rate": 9.960608038157724e-05, "loss": 0.1621, "step": 5280 }, { "grad_norm": 1.4266141653060913, "learning_rate": 9.960262044215901e-05, "loss": 0.1589, "step": 5290 }, { "grad_norm": 1.0959333181381226, "learning_rate": 9.959914543477435e-05, "loss": 0.1313, "step": 5300 }, { "grad_norm": 0.924167275428772, "learning_rate": 9.959565536047892e-05, "loss": 0.1431, "step": 5310 }, { "grad_norm": 0.9277066588401794, "learning_rate": 9.959215022033288e-05, "loss": 0.13, "step": 5320 }, { "grad_norm": 0.8857890963554382, "learning_rate": 9.9588630015401e-05, "loss": 0.1479, "step": 5330 }, { "grad_norm": 0.7727829217910767, "learning_rate": 9.958509474675264e-05, "loss": 0.1527, "step": 5340 }, { "grad_norm": 1.104321002960205, "learning_rate": 9.958154441546171e-05, "loss": 0.1766, "step": 5350 }, { "grad_norm": 0.885715901851654, "learning_rate": 9.957797902260673e-05, "loss": 0.1662, "step": 5360 }, { "grad_norm": 0.8713344931602478, "learning_rate": 9.957439856927073e-05, "loss": 0.1218, "step": 5370 }, { "grad_norm": 0.7636784315109253, "learning_rate": 9.957080305654139e-05, "loss": 0.1124, "step": 5380 }, { "grad_norm": 1.330620288848877, "learning_rate": 9.956719248551092e-05, "loss": 0.1267, "step": 5390 }, { "grad_norm": 0.9822734594345093, "learning_rate": 9.956356685727612e-05, "loss": 0.1185, "step": 5400 }, { "grad_norm": 1.0304569005966187, "learning_rate": 9.955992617293836e-05, "loss": 0.1225, "step": 5410 }, { "grad_norm": 0.841296911239624, "learning_rate": 9.955627043360358e-05, "loss": 0.1447, "step": 5420 }, { "grad_norm": 0.8232971429824829, "learning_rate": 9.955259964038231e-05, "loss": 0.1232, "step": 5430 }, { "grad_norm": 0.6890894770622253, "learning_rate": 9.954891379438962e-05, "loss": 0.1701, "step": 5440 }, { "grad_norm": 0.7422409653663635, "learning_rate": 9.954521289674519e-05, "loss": 0.1357, "step": 5450 }, { "grad_norm": 0.8701598048210144, "learning_rate": 9.954149694857325e-05, "loss": 0.1201, "step": 5460 }, { "grad_norm": 1.08188796043396, "learning_rate": 9.953776595100258e-05, "loss": 0.1172, "step": 5470 }, { "grad_norm": 0.899721622467041, "learning_rate": 9.95340199051666e-05, "loss": 0.1258, "step": 5480 }, { "grad_norm": 0.9479206800460815, "learning_rate": 9.953025881220325e-05, "loss": 0.1241, "step": 5490 }, { "grad_norm": 0.7628172039985657, "learning_rate": 9.952648267325504e-05, "loss": 0.1372, "step": 5500 }, { "grad_norm": 0.8991031050682068, "learning_rate": 9.952269148946905e-05, "loss": 0.1192, "step": 5510 }, { "grad_norm": 0.8127363920211792, "learning_rate": 9.951888526199697e-05, "loss": 0.1235, "step": 5520 }, { "grad_norm": 1.0728341341018677, "learning_rate": 9.951506399199501e-05, "loss": 0.134, "step": 5530 }, { "grad_norm": 0.917293906211853, "learning_rate": 9.951122768062399e-05, "loss": 0.133, "step": 5540 }, { "grad_norm": 0.7466785311698914, "learning_rate": 9.950737632904927e-05, "loss": 0.1378, "step": 5550 }, { "grad_norm": 0.6054357886314392, "learning_rate": 9.950350993844077e-05, "loss": 0.1225, "step": 5560 }, { "grad_norm": 0.8560188412666321, "learning_rate": 9.949962850997303e-05, "loss": 0.1316, "step": 5570 }, { "grad_norm": 0.9763703942298889, "learning_rate": 9.949573204482512e-05, "loss": 0.121, "step": 5580 }, { "grad_norm": 0.8033960461616516, "learning_rate": 9.949182054418064e-05, "loss": 0.1206, "step": 5590 }, { "grad_norm": 1.0750477313995361, "learning_rate": 9.948789400922787e-05, "loss": 0.142, "step": 5600 }, { "grad_norm": 0.9058061242103577, "learning_rate": 9.948395244115953e-05, "loss": 0.1258, "step": 5610 }, { "grad_norm": 0.9833430051803589, "learning_rate": 9.9479995841173e-05, "loss": 0.1755, "step": 5620 }, { "grad_norm": 1.0079678297042847, "learning_rate": 9.947602421047017e-05, "loss": 0.1515, "step": 5630 }, { "grad_norm": 1.0489954948425293, "learning_rate": 9.947203755025753e-05, "loss": 0.1493, "step": 5640 }, { "grad_norm": 1.0249375104904175, "learning_rate": 9.946803586174611e-05, "loss": 0.1203, "step": 5650 }, { "grad_norm": 0.7811262011528015, "learning_rate": 9.946401914615151e-05, "loss": 0.1154, "step": 5660 }, { "grad_norm": 0.8241831064224243, "learning_rate": 9.945998740469394e-05, "loss": 0.129, "step": 5670 }, { "grad_norm": 0.8479894995689392, "learning_rate": 9.945594063859809e-05, "loss": 0.1501, "step": 5680 }, { "grad_norm": 0.7012414932250977, "learning_rate": 9.94518788490933e-05, "loss": 0.1422, "step": 5690 }, { "grad_norm": 0.7136008143424988, "learning_rate": 9.944780203741341e-05, "loss": 0.1193, "step": 5700 }, { "grad_norm": 1.0829662084579468, "learning_rate": 9.944371020479686e-05, "loss": 0.1305, "step": 5710 }, { "grad_norm": 1.067197561264038, "learning_rate": 9.943960335248662e-05, "loss": 0.135, "step": 5720 }, { "grad_norm": 0.8285940885543823, "learning_rate": 9.943548148173027e-05, "loss": 0.128, "step": 5730 }, { "grad_norm": 1.1760265827178955, "learning_rate": 9.943134459377992e-05, "loss": 0.1093, "step": 5740 }, { "grad_norm": 0.9475869536399841, "learning_rate": 9.942719268989222e-05, "loss": 0.1254, "step": 5750 }, { "grad_norm": 0.6616998910903931, "learning_rate": 9.942302577132844e-05, "loss": 0.0947, "step": 5760 }, { "grad_norm": 0.6061694622039795, "learning_rate": 9.941884383935438e-05, "loss": 0.1163, "step": 5770 }, { "grad_norm": 0.7529020309448242, "learning_rate": 9.941464689524039e-05, "loss": 0.1012, "step": 5780 }, { "grad_norm": 0.8589833974838257, "learning_rate": 9.941043494026139e-05, "loss": 0.1396, "step": 5790 }, { "grad_norm": 0.7501980066299438, "learning_rate": 9.940620797569685e-05, "loss": 0.1425, "step": 5800 }, { "grad_norm": 0.7800480127334595, "learning_rate": 9.940196600283082e-05, "loss": 0.1369, "step": 5810 }, { "grad_norm": 1.241601586341858, "learning_rate": 9.939770902295192e-05, "loss": 0.1162, "step": 5820 }, { "grad_norm": 0.8299298286437988, "learning_rate": 9.939343703735329e-05, "loss": 0.1295, "step": 5830 }, { "grad_norm": 0.7308331727981567, "learning_rate": 9.938915004733264e-05, "loss": 0.1255, "step": 5840 }, { "grad_norm": 0.7459497451782227, "learning_rate": 9.938484805419224e-05, "loss": 0.0857, "step": 5850 }, { "grad_norm": 0.7228475213050842, "learning_rate": 9.938053105923894e-05, "loss": 0.0872, "step": 5860 }, { "grad_norm": 0.7020463943481445, "learning_rate": 9.937619906378413e-05, "loss": 0.1054, "step": 5870 }, { "grad_norm": 0.7421414256095886, "learning_rate": 9.937185206914374e-05, "loss": 0.0911, "step": 5880 }, { "grad_norm": 0.8640044331550598, "learning_rate": 9.936749007663829e-05, "loss": 0.0982, "step": 5890 }, { "grad_norm": 0.8889316916465759, "learning_rate": 9.93631130875928e-05, "loss": 0.0948, "step": 5900 }, { "grad_norm": 0.6557901501655579, "learning_rate": 9.935872110333692e-05, "loss": 0.1423, "step": 5910 }, { "grad_norm": 0.7453651428222656, "learning_rate": 9.935431412520484e-05, "loss": 0.1029, "step": 5920 }, { "grad_norm": 0.6869390606880188, "learning_rate": 9.934989215453523e-05, "loss": 0.0968, "step": 5930 }, { "grad_norm": 0.5722059607505798, "learning_rate": 9.934545519267139e-05, "loss": 0.1222, "step": 5940 }, { "grad_norm": 0.7072400450706482, "learning_rate": 9.934100324096117e-05, "loss": 0.114, "step": 5950 }, { "grad_norm": 0.8183084726333618, "learning_rate": 9.933653630075692e-05, "loss": 0.1149, "step": 5960 }, { "grad_norm": 0.8453921675682068, "learning_rate": 9.93320543734156e-05, "loss": 0.1223, "step": 5970 }, { "grad_norm": 0.9003261923789978, "learning_rate": 9.932755746029871e-05, "loss": 0.1595, "step": 5980 }, { "grad_norm": 0.9525597095489502, "learning_rate": 9.932304556277228e-05, "loss": 0.1591, "step": 5990 }, { "grad_norm": 0.8420159816741943, "learning_rate": 9.93185186822069e-05, "loss": 0.1247, "step": 6000 }, { "grad_norm": 0.7618709206581116, "learning_rate": 9.931397681997773e-05, "loss": 0.1222, "step": 6010 }, { "grad_norm": 0.7840919494628906, "learning_rate": 9.930941997746446e-05, "loss": 0.0974, "step": 6020 }, { "grad_norm": 0.6769354343414307, "learning_rate": 9.930484815605134e-05, "loss": 0.1303, "step": 6030 }, { "grad_norm": 0.7383579611778259, "learning_rate": 9.930026135712717e-05, "loss": 0.1045, "step": 6040 }, { "grad_norm": 0.7059565782546997, "learning_rate": 9.92956595820853e-05, "loss": 0.1387, "step": 6050 }, { "grad_norm": 0.7741958498954773, "learning_rate": 9.929104283232362e-05, "loss": 0.1133, "step": 6060 }, { "grad_norm": 0.6355868577957153, "learning_rate": 9.92864111092446e-05, "loss": 0.1168, "step": 6070 }, { "grad_norm": 0.8663500547409058, "learning_rate": 9.92817644142552e-05, "loss": 0.1246, "step": 6080 }, { "grad_norm": 1.0226209163665771, "learning_rate": 9.927710274876698e-05, "loss": 0.116, "step": 6090 }, { "grad_norm": 0.8183728456497192, "learning_rate": 9.927242611419603e-05, "loss": 0.124, "step": 6100 }, { "grad_norm": 0.779615044593811, "learning_rate": 9.926773451196301e-05, "loss": 0.1435, "step": 6110 }, { "grad_norm": 0.6714034676551819, "learning_rate": 9.926302794349306e-05, "loss": 0.1109, "step": 6120 }, { "grad_norm": 0.7265905737876892, "learning_rate": 9.925830641021594e-05, "loss": 0.0837, "step": 6130 }, { "grad_norm": 0.9711920619010925, "learning_rate": 9.925356991356593e-05, "loss": 0.1142, "step": 6140 }, { "grad_norm": 0.8419471979141235, "learning_rate": 9.924881845498184e-05, "loss": 0.1054, "step": 6150 }, { "grad_norm": 0.6812721490859985, "learning_rate": 9.924405203590705e-05, "loss": 0.1372, "step": 6160 }, { "grad_norm": 0.8523169755935669, "learning_rate": 9.923927065778946e-05, "loss": 0.1125, "step": 6170 }, { "grad_norm": 0.8600168228149414, "learning_rate": 9.923447432208154e-05, "loss": 0.1495, "step": 6180 }, { "grad_norm": 0.7704392075538635, "learning_rate": 9.922966303024027e-05, "loss": 0.1341, "step": 6190 }, { "grad_norm": 0.8249804973602295, "learning_rate": 9.922483678372721e-05, "loss": 0.1224, "step": 6200 }, { "grad_norm": 0.9248133301734924, "learning_rate": 9.921999558400845e-05, "loss": 0.108, "step": 6210 }, { "grad_norm": 0.6618738174438477, "learning_rate": 9.92151394325546e-05, "loss": 0.1062, "step": 6220 }, { "grad_norm": 0.7889024019241333, "learning_rate": 9.921026833084084e-05, "loss": 0.1323, "step": 6230 }, { "grad_norm": 0.6196190714836121, "learning_rate": 9.920538228034689e-05, "loss": 0.0917, "step": 6240 }, { "grad_norm": 0.9636349678039551, "learning_rate": 9.920048128255699e-05, "loss": 0.0893, "step": 6250 }, { "grad_norm": 0.9858922958374023, "learning_rate": 9.919556533895995e-05, "loss": 0.1294, "step": 6260 }, { "grad_norm": 0.615280270576477, "learning_rate": 9.919063445104907e-05, "loss": 0.1101, "step": 6270 }, { "grad_norm": 1.0676531791687012, "learning_rate": 9.918568862032227e-05, "loss": 0.138, "step": 6280 }, { "grad_norm": 0.7578985095024109, "learning_rate": 9.918072784828194e-05, "loss": 0.098, "step": 6290 }, { "grad_norm": 0.7284042239189148, "learning_rate": 9.917575213643501e-05, "loss": 0.1052, "step": 6300 }, { "grad_norm": 0.7983983755111694, "learning_rate": 9.917076148629302e-05, "loss": 0.1083, "step": 6310 }, { "grad_norm": 0.6150665283203125, "learning_rate": 9.916575589937196e-05, "loss": 0.1176, "step": 6320 }, { "grad_norm": 0.705398678779602, "learning_rate": 9.916073537719239e-05, "loss": 0.1018, "step": 6330 }, { "grad_norm": 0.7584618330001831, "learning_rate": 9.915569992127944e-05, "loss": 0.0963, "step": 6340 }, { "grad_norm": 0.9380655288696289, "learning_rate": 9.915064953316273e-05, "loss": 0.1253, "step": 6350 }, { "grad_norm": 0.8246062994003296, "learning_rate": 9.914558421437645e-05, "loss": 0.1279, "step": 6360 }, { "grad_norm": 0.7723625302314758, "learning_rate": 9.914050396645929e-05, "loss": 0.1254, "step": 6370 }, { "grad_norm": 1.0259419679641724, "learning_rate": 9.913540879095452e-05, "loss": 0.129, "step": 6380 }, { "grad_norm": 0.7918911576271057, "learning_rate": 9.913029868940987e-05, "loss": 0.1101, "step": 6390 }, { "grad_norm": 0.7570639848709106, "learning_rate": 9.912517366337772e-05, "loss": 0.1283, "step": 6400 }, { "grad_norm": 0.8720314502716064, "learning_rate": 9.912003371441487e-05, "loss": 0.1081, "step": 6410 }, { "grad_norm": 0.6999388337135315, "learning_rate": 9.911487884408271e-05, "loss": 0.1105, "step": 6420 }, { "grad_norm": 1.096037745475769, "learning_rate": 9.910970905394719e-05, "loss": 0.1155, "step": 6430 }, { "grad_norm": 0.8850686550140381, "learning_rate": 9.91045243455787e-05, "loss": 0.1675, "step": 6440 }, { "grad_norm": 0.8697589039802551, "learning_rate": 9.909932472055225e-05, "loss": 0.1343, "step": 6450 }, { "grad_norm": 0.8529297709465027, "learning_rate": 9.909411018044734e-05, "loss": 0.1214, "step": 6460 }, { "grad_norm": 0.7143011689186096, "learning_rate": 9.908888072684802e-05, "loss": 0.128, "step": 6470 }, { "grad_norm": 0.6020171642303467, "learning_rate": 9.908363636134285e-05, "loss": 0.1011, "step": 6480 }, { "grad_norm": 1.0660290718078613, "learning_rate": 9.907837708552493e-05, "loss": 0.1138, "step": 6490 }, { "grad_norm": 0.6946395039558411, "learning_rate": 9.90731029009919e-05, "loss": 0.1427, "step": 6500 }, { "grad_norm": 0.7659786939620972, "learning_rate": 9.906781380934589e-05, "loss": 0.1137, "step": 6510 }, { "grad_norm": 0.8675088882446289, "learning_rate": 9.906250981219362e-05, "loss": 0.1139, "step": 6520 }, { "grad_norm": 0.760338306427002, "learning_rate": 9.905719091114628e-05, "loss": 0.1111, "step": 6530 }, { "grad_norm": 1.1241726875305176, "learning_rate": 9.905185710781964e-05, "loss": 0.1303, "step": 6540 }, { "grad_norm": 1.0017364025115967, "learning_rate": 9.904650840383392e-05, "loss": 0.1108, "step": 6550 }, { "grad_norm": 0.9613134860992432, "learning_rate": 9.904114480081397e-05, "loss": 0.1038, "step": 6560 }, { "grad_norm": 0.9528111815452576, "learning_rate": 9.903576630038906e-05, "loss": 0.1291, "step": 6570 }, { "grad_norm": 0.9474925398826599, "learning_rate": 9.903037290419309e-05, "loss": 0.1086, "step": 6580 }, { "grad_norm": 1.0386041402816772, "learning_rate": 9.902496461386439e-05, "loss": 0.1151, "step": 6590 }, { "grad_norm": 0.9770430326461792, "learning_rate": 9.901954143104588e-05, "loss": 0.1034, "step": 6600 }, { "grad_norm": 0.5697273015975952, "learning_rate": 9.901410335738496e-05, "loss": 0.1253, "step": 6610 }, { "grad_norm": 0.7451756596565247, "learning_rate": 9.900865039453358e-05, "loss": 0.1053, "step": 6620 }, { "grad_norm": 0.5815768241882324, "learning_rate": 9.900318254414821e-05, "loss": 0.1469, "step": 6630 }, { "grad_norm": 0.7967324256896973, "learning_rate": 9.899769980788985e-05, "loss": 0.1389, "step": 6640 }, { "grad_norm": 1.329135775566101, "learning_rate": 9.899220218742398e-05, "loss": 0.1259, "step": 6650 }, { "grad_norm": 1.109104871749878, "learning_rate": 9.898668968442066e-05, "loss": 0.1304, "step": 6660 }, { "grad_norm": 0.7789966464042664, "learning_rate": 9.898116230055443e-05, "loss": 0.1207, "step": 6670 }, { "grad_norm": 0.9893991947174072, "learning_rate": 9.897562003750437e-05, "loss": 0.1077, "step": 6680 }, { "grad_norm": 0.8767921924591064, "learning_rate": 9.897006289695407e-05, "loss": 0.1165, "step": 6690 }, { "grad_norm": 1.0071903467178345, "learning_rate": 9.896449088059164e-05, "loss": 0.1161, "step": 6700 }, { "grad_norm": 0.7219248414039612, "learning_rate": 9.89589039901097e-05, "loss": 0.1093, "step": 6710 }, { "grad_norm": 0.7403485774993896, "learning_rate": 9.895330222720542e-05, "loss": 0.119, "step": 6720 }, { "grad_norm": 0.7000433206558228, "learning_rate": 9.894768559358047e-05, "loss": 0.1053, "step": 6730 }, { "grad_norm": 0.8420553207397461, "learning_rate": 9.894205409094101e-05, "loss": 0.1217, "step": 6740 }, { "grad_norm": 1.1088379621505737, "learning_rate": 9.893640772099777e-05, "loss": 0.1397, "step": 6750 }, { "grad_norm": 0.9918035268783569, "learning_rate": 9.893074648546595e-05, "loss": 0.1685, "step": 6760 }, { "grad_norm": 0.7741139531135559, "learning_rate": 9.892507038606528e-05, "loss": 0.1118, "step": 6770 }, { "grad_norm": 0.6796315312385559, "learning_rate": 9.891937942452003e-05, "loss": 0.0974, "step": 6780 }, { "grad_norm": 0.7580564618110657, "learning_rate": 9.891367360255895e-05, "loss": 0.1182, "step": 6790 }, { "grad_norm": 0.801645815372467, "learning_rate": 9.890795292191532e-05, "loss": 0.096, "step": 6800 }, { "grad_norm": 1.0171194076538086, "learning_rate": 9.890221738432694e-05, "loss": 0.1132, "step": 6810 }, { "grad_norm": 0.8044463992118835, "learning_rate": 9.88964669915361e-05, "loss": 0.0944, "step": 6820 }, { "grad_norm": 0.665590763092041, "learning_rate": 9.889070174528963e-05, "loss": 0.104, "step": 6830 }, { "grad_norm": 0.9238048791885376, "learning_rate": 9.888492164733883e-05, "loss": 0.1001, "step": 6840 }, { "grad_norm": 0.9833086133003235, "learning_rate": 9.88791266994396e-05, "loss": 0.1041, "step": 6850 }, { "grad_norm": 0.9032684564590454, "learning_rate": 9.887331690335223e-05, "loss": 0.1269, "step": 6860 }, { "grad_norm": 0.6378113031387329, "learning_rate": 9.886749226084163e-05, "loss": 0.0993, "step": 6870 }, { "grad_norm": 0.6693492531776428, "learning_rate": 9.886165277367714e-05, "loss": 0.1184, "step": 6880 }, { "grad_norm": 0.6442267894744873, "learning_rate": 9.885579844363265e-05, "loss": 0.1018, "step": 6890 }, { "grad_norm": 0.841157078742981, "learning_rate": 9.884992927248656e-05, "loss": 0.0905, "step": 6900 }, { "grad_norm": 0.8268330097198486, "learning_rate": 9.884404526202178e-05, "loss": 0.1074, "step": 6910 }, { "grad_norm": 0.8828208446502686, "learning_rate": 9.883814641402568e-05, "loss": 0.0937, "step": 6920 }, { "grad_norm": 0.7945942282676697, "learning_rate": 9.88322327302902e-05, "loss": 0.1393, "step": 6930 }, { "grad_norm": 0.625950276851654, "learning_rate": 9.882630421261176e-05, "loss": 0.1033, "step": 6940 }, { "grad_norm": 0.9508978128433228, "learning_rate": 9.88203608627913e-05, "loss": 0.1288, "step": 6950 }, { "grad_norm": 0.7300418019294739, "learning_rate": 9.881440268263422e-05, "loss": 0.1044, "step": 6960 }, { "grad_norm": 0.7552957534790039, "learning_rate": 9.880842967395048e-05, "loss": 0.1114, "step": 6970 }, { "grad_norm": 0.8447068929672241, "learning_rate": 9.880244183855452e-05, "loss": 0.1084, "step": 6980 }, { "grad_norm": 0.8511649370193481, "learning_rate": 9.879643917826527e-05, "loss": 0.1065, "step": 6990 }, { "grad_norm": 0.649890661239624, "learning_rate": 9.87904216949062e-05, "loss": 0.0845, "step": 7000 }, { "grad_norm": 0.8641259670257568, "learning_rate": 9.878438939030526e-05, "loss": 0.1043, "step": 7010 }, { "grad_norm": 0.8796404600143433, "learning_rate": 9.877834226629489e-05, "loss": 0.1019, "step": 7020 }, { "grad_norm": 0.6948674917221069, "learning_rate": 9.877228032471206e-05, "loss": 0.107, "step": 7030 }, { "grad_norm": 0.748417854309082, "learning_rate": 9.876620356739823e-05, "loss": 0.1195, "step": 7040 }, { "grad_norm": 0.8125472664833069, "learning_rate": 9.876011199619935e-05, "loss": 0.1334, "step": 7050 }, { "grad_norm": 0.7498759031295776, "learning_rate": 9.875400561296589e-05, "loss": 0.1166, "step": 7060 }, { "grad_norm": 0.8788738250732422, "learning_rate": 9.874788441955278e-05, "loss": 0.1161, "step": 7070 }, { "grad_norm": 0.9708063006401062, "learning_rate": 9.874174841781951e-05, "loss": 0.1067, "step": 7080 }, { "grad_norm": 0.8202360272407532, "learning_rate": 9.873559760963003e-05, "loss": 0.124, "step": 7090 }, { "grad_norm": 0.788311243057251, "learning_rate": 9.872943199685278e-05, "loss": 0.1321, "step": 7100 }, { "grad_norm": 0.7417576909065247, "learning_rate": 9.872325158136071e-05, "loss": 0.0874, "step": 7110 }, { "grad_norm": 0.6714224815368652, "learning_rate": 9.871705636503128e-05, "loss": 0.1072, "step": 7120 }, { "grad_norm": 1.0182090997695923, "learning_rate": 9.871084634974641e-05, "loss": 0.0981, "step": 7130 }, { "grad_norm": 0.7195924520492554, "learning_rate": 9.870462153739257e-05, "loss": 0.1283, "step": 7140 }, { "grad_norm": 0.8587167263031006, "learning_rate": 9.869838192986067e-05, "loss": 0.1368, "step": 7150 }, { "grad_norm": 0.5441277623176575, "learning_rate": 9.869212752904616e-05, "loss": 0.1165, "step": 7160 }, { "grad_norm": 1.0662858486175537, "learning_rate": 9.868585833684894e-05, "loss": 0.1331, "step": 7170 }, { "grad_norm": 1.023278832435608, "learning_rate": 9.867957435517342e-05, "loss": 0.1112, "step": 7180 }, { "grad_norm": 0.9826900362968445, "learning_rate": 9.867327558592854e-05, "loss": 0.1211, "step": 7190 }, { "grad_norm": 0.6636654734611511, "learning_rate": 9.866696203102766e-05, "loss": 0.1154, "step": 7200 }, { "grad_norm": 0.8370222449302673, "learning_rate": 9.86606336923887e-05, "loss": 0.109, "step": 7210 }, { "grad_norm": 0.7623265981674194, "learning_rate": 9.865429057193403e-05, "loss": 0.1186, "step": 7220 }, { "grad_norm": 0.9802122712135315, "learning_rate": 9.864793267159053e-05, "loss": 0.1095, "step": 7230 }, { "grad_norm": 0.9360064268112183, "learning_rate": 9.864155999328957e-05, "loss": 0.1415, "step": 7240 }, { "grad_norm": 0.6270813941955566, "learning_rate": 9.8635172538967e-05, "loss": 0.0923, "step": 7250 }, { "grad_norm": 0.6241870522499084, "learning_rate": 9.862877031056312e-05, "loss": 0.084, "step": 7260 }, { "grad_norm": 0.6779041886329651, "learning_rate": 9.862235331002279e-05, "loss": 0.0956, "step": 7270 }, { "grad_norm": 0.879892885684967, "learning_rate": 9.861592153929533e-05, "loss": 0.1016, "step": 7280 }, { "grad_norm": 0.8421424627304077, "learning_rate": 9.860947500033455e-05, "loss": 0.1187, "step": 7290 }, { "grad_norm": 0.857740581035614, "learning_rate": 9.86030136950987e-05, "loss": 0.1323, "step": 7300 }, { "grad_norm": 0.6553378105163574, "learning_rate": 9.85965376255506e-05, "loss": 0.1121, "step": 7310 }, { "grad_norm": 0.8015048503875732, "learning_rate": 9.859004679365747e-05, "loss": 0.1012, "step": 7320 }, { "grad_norm": 1.2515732049942017, "learning_rate": 9.858354120139108e-05, "loss": 0.1431, "step": 7330 }, { "grad_norm": 0.9536264538764954, "learning_rate": 9.857702085072764e-05, "loss": 0.1177, "step": 7340 }, { "grad_norm": 1.1854591369628906, "learning_rate": 9.857048574364787e-05, "loss": 0.1247, "step": 7350 }, { "grad_norm": 0.9534831643104553, "learning_rate": 9.856393588213698e-05, "loss": 0.1317, "step": 7360 }, { "grad_norm": 0.9556472897529602, "learning_rate": 9.855737126818458e-05, "loss": 0.1101, "step": 7370 }, { "grad_norm": 0.7689210176467896, "learning_rate": 9.855079190378491e-05, "loss": 0.0946, "step": 7380 }, { "grad_norm": 0.9300874471664429, "learning_rate": 9.854419779093655e-05, "loss": 0.1183, "step": 7390 }, { "grad_norm": 0.6947682499885559, "learning_rate": 9.853758893164264e-05, "loss": 0.0801, "step": 7400 }, { "grad_norm": 0.8127409219741821, "learning_rate": 9.853096532791078e-05, "loss": 0.1165, "step": 7410 }, { "grad_norm": 0.8505898118019104, "learning_rate": 9.852432698175304e-05, "loss": 0.1024, "step": 7420 }, { "grad_norm": 0.9204187989234924, "learning_rate": 9.851767389518597e-05, "loss": 0.1103, "step": 7430 }, { "grad_norm": 0.8442255854606628, "learning_rate": 9.85110060702306e-05, "loss": 0.1175, "step": 7440 }, { "grad_norm": 0.754696786403656, "learning_rate": 9.850432350891245e-05, "loss": 0.1129, "step": 7450 }, { "grad_norm": 0.8107526898384094, "learning_rate": 9.84976262132615e-05, "loss": 0.1, "step": 7460 }, { "grad_norm": 0.9241111874580383, "learning_rate": 9.849091418531222e-05, "loss": 0.116, "step": 7470 }, { "grad_norm": 0.5956346392631531, "learning_rate": 9.848418742710353e-05, "loss": 0.1077, "step": 7480 }, { "grad_norm": 0.7641017436981201, "learning_rate": 9.847744594067885e-05, "loss": 0.106, "step": 7490 }, { "grad_norm": 0.7367520928382874, "learning_rate": 9.847068972808607e-05, "loss": 0.1391, "step": 7500 }, { "grad_norm": 0.7757737636566162, "learning_rate": 9.846391879137756e-05, "loss": 0.111, "step": 7510 }, { "grad_norm": 0.4815749228000641, "learning_rate": 9.845713313261012e-05, "loss": 0.0961, "step": 7520 }, { "grad_norm": 0.7629364728927612, "learning_rate": 9.845033275384505e-05, "loss": 0.1066, "step": 7530 }, { "grad_norm": 1.031085729598999, "learning_rate": 9.844351765714818e-05, "loss": 0.1381, "step": 7540 }, { "grad_norm": 0.8858757019042969, "learning_rate": 9.843668784458971e-05, "loss": 0.1267, "step": 7550 }, { "grad_norm": 0.7585353851318359, "learning_rate": 9.842984331824437e-05, "loss": 0.1202, "step": 7560 }, { "grad_norm": 0.5858237147331238, "learning_rate": 9.842298408019133e-05, "loss": 0.1248, "step": 7570 }, { "grad_norm": 0.7364085912704468, "learning_rate": 9.841611013251429e-05, "loss": 0.1013, "step": 7580 }, { "grad_norm": 0.6320050954818726, "learning_rate": 9.840922147730133e-05, "loss": 0.146, "step": 7590 }, { "grad_norm": 0.7129645943641663, "learning_rate": 9.840231811664506e-05, "loss": 0.124, "step": 7600 }, { "grad_norm": 0.8051250576972961, "learning_rate": 9.839540005264252e-05, "loss": 0.1174, "step": 7610 }, { "grad_norm": 0.6920487880706787, "learning_rate": 9.838846728739527e-05, "loss": 0.0992, "step": 7620 }, { "grad_norm": 0.8447760343551636, "learning_rate": 9.838151982300927e-05, "loss": 0.1216, "step": 7630 }, { "grad_norm": 0.7768735885620117, "learning_rate": 9.8374557661595e-05, "loss": 0.1175, "step": 7640 }, { "grad_norm": 0.8881555199623108, "learning_rate": 9.836758080526735e-05, "loss": 0.1112, "step": 7650 }, { "grad_norm": 0.7657982110977173, "learning_rate": 9.836058925614575e-05, "loss": 0.0952, "step": 7660 }, { "grad_norm": 0.7047959566116333, "learning_rate": 9.8353583016354e-05, "loss": 0.0876, "step": 7670 }, { "grad_norm": 0.7588182687759399, "learning_rate": 9.834656208802044e-05, "loss": 0.0985, "step": 7680 }, { "grad_norm": 0.5253893733024597, "learning_rate": 9.833952647327784e-05, "loss": 0.1326, "step": 7690 }, { "grad_norm": 0.9323515892028809, "learning_rate": 9.833247617426342e-05, "loss": 0.1157, "step": 7700 }, { "grad_norm": 0.7865415811538696, "learning_rate": 9.832541119311889e-05, "loss": 0.1016, "step": 7710 }, { "grad_norm": 1.0611966848373413, "learning_rate": 9.83183315319904e-05, "loss": 0.1055, "step": 7720 }, { "grad_norm": 0.8075302839279175, "learning_rate": 9.831123719302855e-05, "loss": 0.1117, "step": 7730 }, { "grad_norm": 0.9165618419647217, "learning_rate": 9.830412817838842e-05, "loss": 0.0962, "step": 7740 }, { "grad_norm": 0.8608669638633728, "learning_rate": 9.829700449022956e-05, "loss": 0.1035, "step": 7750 }, { "grad_norm": 0.6636890172958374, "learning_rate": 9.828986613071593e-05, "loss": 0.1035, "step": 7760 }, { "grad_norm": 0.6492105722427368, "learning_rate": 9.828271310201601e-05, "loss": 0.1125, "step": 7770 }, { "grad_norm": 0.7295095324516296, "learning_rate": 9.827554540630268e-05, "loss": 0.1001, "step": 7780 }, { "grad_norm": 0.7297043204307556, "learning_rate": 9.826836304575329e-05, "loss": 0.1003, "step": 7790 }, { "grad_norm": 0.7780761122703552, "learning_rate": 9.826116602254966e-05, "loss": 0.1245, "step": 7800 }, { "grad_norm": 1.0564063787460327, "learning_rate": 9.825395433887805e-05, "loss": 0.0989, "step": 7810 }, { "grad_norm": 0.6251338124275208, "learning_rate": 9.824672799692917e-05, "loss": 0.1043, "step": 7820 }, { "grad_norm": 0.8175916075706482, "learning_rate": 9.823948699889823e-05, "loss": 0.097, "step": 7830 }, { "grad_norm": 1.0313719511032104, "learning_rate": 9.823223134698483e-05, "loss": 0.1347, "step": 7840 }, { "grad_norm": 0.7975802421569824, "learning_rate": 9.822496104339303e-05, "loss": 0.1059, "step": 7850 }, { "grad_norm": 0.9363609552383423, "learning_rate": 9.821767609033138e-05, "loss": 0.107, "step": 7860 }, { "grad_norm": 1.1952927112579346, "learning_rate": 9.821037649001284e-05, "loss": 0.1259, "step": 7870 }, { "grad_norm": 0.9573523998260498, "learning_rate": 9.820306224465486e-05, "loss": 0.1237, "step": 7880 }, { "grad_norm": 0.816676139831543, "learning_rate": 9.819573335647928e-05, "loss": 0.1253, "step": 7890 }, { "grad_norm": 0.6842550039291382, "learning_rate": 9.818838982771246e-05, "loss": 0.1131, "step": 7900 }, { "grad_norm": 0.6100413799285889, "learning_rate": 9.818103166058514e-05, "loss": 0.1294, "step": 7910 }, { "grad_norm": 0.6504509449005127, "learning_rate": 9.817365885733254e-05, "loss": 0.1114, "step": 7920 }, { "grad_norm": 0.7172811627388, "learning_rate": 9.816627142019434e-05, "loss": 0.0975, "step": 7930 }, { "grad_norm": 0.9627309441566467, "learning_rate": 9.815886935141463e-05, "loss": 0.1208, "step": 7940 }, { "grad_norm": 0.8311769962310791, "learning_rate": 9.8151452653242e-05, "loss": 0.107, "step": 7950 }, { "grad_norm": 0.64766925573349, "learning_rate": 9.814402132792939e-05, "loss": 0.1151, "step": 7960 }, { "grad_norm": 0.7653962969779968, "learning_rate": 9.813657537773428e-05, "loss": 0.1363, "step": 7970 }, { "grad_norm": 1.1558232307434082, "learning_rate": 9.812911480491854e-05, "loss": 0.1193, "step": 7980 }, { "grad_norm": 1.3010278940200806, "learning_rate": 9.81216396117485e-05, "loss": 0.1183, "step": 7990 }, { "grad_norm": 1.183657169342041, "learning_rate": 9.811414980049491e-05, "loss": 0.1316, "step": 8000 }, { "grad_norm": 0.6862872242927551, "learning_rate": 9.810664537343301e-05, "loss": 0.1018, "step": 8010 }, { "grad_norm": 0.8607950210571289, "learning_rate": 9.809912633284243e-05, "loss": 0.1246, "step": 8020 }, { "grad_norm": 1.1492270231246948, "learning_rate": 9.809159268100725e-05, "loss": 0.1339, "step": 8030 }, { "grad_norm": 0.9316725730895996, "learning_rate": 9.808404442021599e-05, "loss": 0.1114, "step": 8040 }, { "grad_norm": 0.7595432996749878, "learning_rate": 9.807648155276163e-05, "loss": 0.115, "step": 8050 }, { "grad_norm": 0.7480480670928955, "learning_rate": 9.806890408094156e-05, "loss": 0.1396, "step": 8060 }, { "grad_norm": 0.7056934237480164, "learning_rate": 9.806131200705761e-05, "loss": 0.0907, "step": 8070 }, { "grad_norm": 0.7574305534362793, "learning_rate": 9.805370533341605e-05, "loss": 0.1021, "step": 8080 }, { "grad_norm": 0.8656883835792542, "learning_rate": 9.804608406232762e-05, "loss": 0.1227, "step": 8090 }, { "grad_norm": 0.7076953649520874, "learning_rate": 9.803844819610741e-05, "loss": 0.1239, "step": 8100 }, { "grad_norm": 0.8441457748413086, "learning_rate": 9.803079773707504e-05, "loss": 0.1399, "step": 8110 }, { "grad_norm": 0.8106648921966553, "learning_rate": 9.802313268755447e-05, "loss": 0.1154, "step": 8120 }, { "grad_norm": 0.6870909929275513, "learning_rate": 9.801545304987419e-05, "loss": 0.0982, "step": 8130 }, { "grad_norm": 0.6994389295578003, "learning_rate": 9.800775882636704e-05, "loss": 0.1038, "step": 8140 }, { "grad_norm": 0.8145632147789001, "learning_rate": 9.800005001937034e-05, "loss": 0.1226, "step": 8150 }, { "grad_norm": 0.8438630104064941, "learning_rate": 9.79923266312258e-05, "loss": 0.1078, "step": 8160 }, { "grad_norm": 0.8690328001976013, "learning_rate": 9.79845886642796e-05, "loss": 0.1181, "step": 8170 }, { "grad_norm": 0.9680644869804382, "learning_rate": 9.797683612088233e-05, "loss": 0.1158, "step": 8180 }, { "grad_norm": 0.6235941648483276, "learning_rate": 9.796906900338898e-05, "loss": 0.1067, "step": 8190 }, { "grad_norm": 0.9172364473342896, "learning_rate": 9.796128731415903e-05, "loss": 0.1004, "step": 8200 }, { "grad_norm": 0.8048495650291443, "learning_rate": 9.795349105555634e-05, "loss": 0.0913, "step": 8210 }, { "grad_norm": 1.0056560039520264, "learning_rate": 9.794568022994922e-05, "loss": 0.1289, "step": 8220 }, { "grad_norm": 0.7413055896759033, "learning_rate": 9.793785483971034e-05, "loss": 0.1201, "step": 8230 }, { "grad_norm": 0.8831872940063477, "learning_rate": 9.793001488721691e-05, "loss": 0.1144, "step": 8240 }, { "grad_norm": 0.7004793882369995, "learning_rate": 9.792216037485047e-05, "loss": 0.11, "step": 8250 }, { "grad_norm": 0.6825798749923706, "learning_rate": 9.791429130499704e-05, "loss": 0.1133, "step": 8260 }, { "grad_norm": 1.2186864614486694, "learning_rate": 9.790640768004698e-05, "loss": 0.1168, "step": 8270 }, { "grad_norm": 0.7828971147537231, "learning_rate": 9.789850950239518e-05, "loss": 0.1196, "step": 8280 }, { "grad_norm": 0.7505565285682678, "learning_rate": 9.789059677444089e-05, "loss": 0.1167, "step": 8290 }, { "grad_norm": 0.9029496312141418, "learning_rate": 9.788266949858776e-05, "loss": 0.126, "step": 8300 }, { "grad_norm": 1.1484657526016235, "learning_rate": 9.787472767724392e-05, "loss": 0.1305, "step": 8310 }, { "grad_norm": 0.9438452124595642, "learning_rate": 9.786677131282185e-05, "loss": 0.1017, "step": 8320 }, { "grad_norm": 0.9191779494285583, "learning_rate": 9.785880040773853e-05, "loss": 0.1092, "step": 8330 }, { "grad_norm": 0.9350516200065613, "learning_rate": 9.785081496441527e-05, "loss": 0.0988, "step": 8340 }, { "grad_norm": 0.872738242149353, "learning_rate": 9.784281498527785e-05, "loss": 0.0979, "step": 8350 }, { "grad_norm": 0.8212643265724182, "learning_rate": 9.783480047275646e-05, "loss": 0.0998, "step": 8360 }, { "grad_norm": 0.4782106578350067, "learning_rate": 9.78267714292857e-05, "loss": 0.1052, "step": 8370 }, { "grad_norm": 0.5874379873275757, "learning_rate": 9.781872785730454e-05, "loss": 0.0894, "step": 8380 }, { "grad_norm": 0.712817370891571, "learning_rate": 9.781066975925646e-05, "loss": 0.1047, "step": 8390 }, { "grad_norm": 0.5315495729446411, "learning_rate": 9.780259713758928e-05, "loss": 0.1175, "step": 8400 }, { "grad_norm": 0.49772459268569946, "learning_rate": 9.779450999475524e-05, "loss": 0.0974, "step": 8410 }, { "grad_norm": 0.509315013885498, "learning_rate": 9.7786408333211e-05, "loss": 0.1154, "step": 8420 }, { "grad_norm": 0.7482325434684753, "learning_rate": 9.777829215541764e-05, "loss": 0.1142, "step": 8430 }, { "grad_norm": 0.5238484740257263, "learning_rate": 9.777016146384064e-05, "loss": 0.1035, "step": 8440 }, { "grad_norm": 0.7791712284088135, "learning_rate": 9.776201626094988e-05, "loss": 0.1139, "step": 8450 }, { "grad_norm": 0.6068872213363647, "learning_rate": 9.775385654921965e-05, "loss": 0.1269, "step": 8460 }, { "grad_norm": 0.8338019847869873, "learning_rate": 9.774568233112868e-05, "loss": 0.1175, "step": 8470 }, { "grad_norm": 0.5465388298034668, "learning_rate": 9.773749360916007e-05, "loss": 0.124, "step": 8480 }, { "grad_norm": 0.8134889602661133, "learning_rate": 9.772929038580134e-05, "loss": 0.1149, "step": 8490 }, { "grad_norm": 0.8839403390884399, "learning_rate": 9.772107266354439e-05, "loss": 0.1248, "step": 8500 }, { "grad_norm": 0.9107663035392761, "learning_rate": 9.77128404448856e-05, "loss": 0.1132, "step": 8510 }, { "grad_norm": 0.9390890598297119, "learning_rate": 9.770459373232565e-05, "loss": 0.098, "step": 8520 }, { "grad_norm": 0.7055835127830505, "learning_rate": 9.769633252836969e-05, "loss": 0.1052, "step": 8530 }, { "grad_norm": 0.6554982662200928, "learning_rate": 9.768805683552724e-05, "loss": 0.1037, "step": 8540 }, { "grad_norm": 0.7291069030761719, "learning_rate": 9.767976665631228e-05, "loss": 0.1019, "step": 8550 }, { "grad_norm": 0.5723056197166443, "learning_rate": 9.767146199324311e-05, "loss": 0.0975, "step": 8560 }, { "grad_norm": 0.7135376334190369, "learning_rate": 9.766314284884249e-05, "loss": 0.1136, "step": 8570 }, { "grad_norm": 0.8131445646286011, "learning_rate": 9.765480922563752e-05, "loss": 0.1167, "step": 8580 }, { "grad_norm": 1.0183868408203125, "learning_rate": 9.764646112615978e-05, "loss": 0.1149, "step": 8590 }, { "grad_norm": 0.8575623035430908, "learning_rate": 9.763809855294517e-05, "loss": 0.0968, "step": 8600 }, { "grad_norm": 0.9951133728027344, "learning_rate": 9.762972150853404e-05, "loss": 0.0984, "step": 8610 }, { "grad_norm": 0.6580617427825928, "learning_rate": 9.762132999547111e-05, "loss": 0.1044, "step": 8620 }, { "grad_norm": 0.7755747437477112, "learning_rate": 9.761292401630549e-05, "loss": 0.153, "step": 8630 }, { "grad_norm": 1.0263614654541016, "learning_rate": 9.76045035735907e-05, "loss": 0.1103, "step": 8640 }, { "grad_norm": 0.5946556329727173, "learning_rate": 9.759606866988464e-05, "loss": 0.0926, "step": 8650 }, { "grad_norm": 1.0723477602005005, "learning_rate": 9.758761930774963e-05, "loss": 0.1103, "step": 8660 }, { "grad_norm": 1.063755750656128, "learning_rate": 9.757915548975235e-05, "loss": 0.0991, "step": 8670 }, { "grad_norm": 0.8034469485282898, "learning_rate": 9.757067721846389e-05, "loss": 0.1104, "step": 8680 }, { "grad_norm": 0.6719763278961182, "learning_rate": 9.756218449645971e-05, "loss": 0.0923, "step": 8690 }, { "grad_norm": 0.8090630173683167, "learning_rate": 9.75536773263197e-05, "loss": 0.0898, "step": 8700 }, { "grad_norm": 0.9589013457298279, "learning_rate": 9.75451557106281e-05, "loss": 0.106, "step": 8710 }, { "grad_norm": 1.0118097066879272, "learning_rate": 9.753661965197354e-05, "loss": 0.093, "step": 8720 }, { "grad_norm": 0.9715580940246582, "learning_rate": 9.752806915294908e-05, "loss": 0.0994, "step": 8730 }, { "grad_norm": 0.7380316853523254, "learning_rate": 9.75195042161521e-05, "loss": 0.1037, "step": 8740 }, { "grad_norm": 0.7176957726478577, "learning_rate": 9.751092484418442e-05, "loss": 0.0985, "step": 8750 }, { "grad_norm": 0.9365617036819458, "learning_rate": 9.750233103965224e-05, "loss": 0.101, "step": 8760 }, { "grad_norm": 0.6797965168952942, "learning_rate": 9.749372280516611e-05, "loss": 0.0971, "step": 8770 }, { "grad_norm": 1.1935135126113892, "learning_rate": 9.748510014334097e-05, "loss": 0.0868, "step": 8780 }, { "grad_norm": 0.566483199596405, "learning_rate": 9.747646305679621e-05, "loss": 0.1259, "step": 8790 }, { "grad_norm": 0.5371443033218384, "learning_rate": 9.74678115481555e-05, "loss": 0.0835, "step": 8800 }, { "grad_norm": 0.7220647931098938, "learning_rate": 9.745914562004696e-05, "loss": 0.108, "step": 8810 }, { "grad_norm": 0.5437669157981873, "learning_rate": 9.745046527510307e-05, "loss": 0.106, "step": 8820 }, { "grad_norm": 0.9463948011398315, "learning_rate": 9.744177051596068e-05, "loss": 0.1141, "step": 8830 }, { "grad_norm": 0.7946024537086487, "learning_rate": 9.743306134526105e-05, "loss": 0.1205, "step": 8840 }, { "grad_norm": 0.7284457683563232, "learning_rate": 9.742433776564977e-05, "loss": 0.1283, "step": 8850 }, { "grad_norm": 0.7483442425727844, "learning_rate": 9.741559977977683e-05, "loss": 0.0967, "step": 8860 }, { "grad_norm": 0.6880325675010681, "learning_rate": 9.740684739029661e-05, "loss": 0.1104, "step": 8870 }, { "grad_norm": 1.0334876775741577, "learning_rate": 9.739808059986789e-05, "loss": 0.1096, "step": 8880 }, { "grad_norm": 0.6204767227172852, "learning_rate": 9.738929941115373e-05, "loss": 0.114, "step": 8890 }, { "grad_norm": 1.0482783317565918, "learning_rate": 9.738050382682167e-05, "loss": 0.1204, "step": 8900 }, { "grad_norm": 0.7804141044616699, "learning_rate": 9.737169384954355e-05, "loss": 0.119, "step": 8910 }, { "grad_norm": 0.5896562933921814, "learning_rate": 9.736286948199562e-05, "loss": 0.0884, "step": 8920 }, { "grad_norm": 0.6594409346580505, "learning_rate": 9.735403072685848e-05, "loss": 0.1202, "step": 8930 }, { "grad_norm": 0.5982496738433838, "learning_rate": 9.734517758681712e-05, "loss": 0.088, "step": 8940 }, { "grad_norm": 1.2333959341049194, "learning_rate": 9.733631006456088e-05, "loss": 0.0963, "step": 8950 }, { "grad_norm": 0.9001926779747009, "learning_rate": 9.732742816278348e-05, "loss": 0.1123, "step": 8960 }, { "grad_norm": 0.8633454442024231, "learning_rate": 9.731853188418302e-05, "loss": 0.1218, "step": 8970 }, { "grad_norm": 0.7519991993904114, "learning_rate": 9.730962123146194e-05, "loss": 0.1058, "step": 8980 }, { "grad_norm": 1.022164225578308, "learning_rate": 9.730069620732709e-05, "loss": 0.1156, "step": 8990 }, { "grad_norm": 0.8937515020370483, "learning_rate": 9.72917568144896e-05, "loss": 0.0851, "step": 9000 }, { "grad_norm": 0.9746062755584717, "learning_rate": 9.728280305566509e-05, "loss": 0.1341, "step": 9010 }, { "grad_norm": 1.0622831583023071, "learning_rate": 9.727383493357343e-05, "loss": 0.1345, "step": 9020 }, { "grad_norm": 0.8984041213989258, "learning_rate": 9.726485245093891e-05, "loss": 0.095, "step": 9030 }, { "grad_norm": 0.6220551133155823, "learning_rate": 9.725585561049018e-05, "loss": 0.1155, "step": 9040 }, { "grad_norm": 1.0553345680236816, "learning_rate": 9.724684441496022e-05, "loss": 0.0951, "step": 9050 }, { "grad_norm": 1.0393060445785522, "learning_rate": 9.72378188670864e-05, "loss": 0.1149, "step": 9060 }, { "grad_norm": 0.761735200881958, "learning_rate": 9.722877896961047e-05, "loss": 0.1258, "step": 9070 }, { "grad_norm": 1.2190347909927368, "learning_rate": 9.721972472527848e-05, "loss": 0.1292, "step": 9080 }, { "grad_norm": 0.7198113799095154, "learning_rate": 9.721065613684089e-05, "loss": 0.1296, "step": 9090 }, { "grad_norm": 0.6735286116600037, "learning_rate": 9.72015732070525e-05, "loss": 0.0977, "step": 9100 }, { "grad_norm": 0.8916217684745789, "learning_rate": 9.719247593867244e-05, "loss": 0.0883, "step": 9110 }, { "grad_norm": 0.665476381778717, "learning_rate": 9.718336433446423e-05, "loss": 0.0966, "step": 9120 }, { "grad_norm": 0.626725435256958, "learning_rate": 9.717423839719574e-05, "loss": 0.1018, "step": 9130 }, { "grad_norm": 0.5901679396629333, "learning_rate": 9.71650981296392e-05, "loss": 0.1054, "step": 9140 }, { "grad_norm": 0.7043330669403076, "learning_rate": 9.715594353457118e-05, "loss": 0.11, "step": 9150 }, { "grad_norm": 1.2628742456436157, "learning_rate": 9.714677461477257e-05, "loss": 0.1329, "step": 9160 }, { "grad_norm": 0.6864972710609436, "learning_rate": 9.713759137302869e-05, "loss": 0.1037, "step": 9170 }, { "grad_norm": 0.8240109086036682, "learning_rate": 9.712839381212914e-05, "loss": 0.1182, "step": 9180 }, { "grad_norm": 0.7997499108314514, "learning_rate": 9.71191819348679e-05, "loss": 0.11, "step": 9190 }, { "grad_norm": 0.9548439979553223, "learning_rate": 9.710995574404331e-05, "loss": 0.1008, "step": 9200 }, { "grad_norm": 0.9049219489097595, "learning_rate": 9.710071524245802e-05, "loss": 0.1177, "step": 9210 }, { "grad_norm": 0.9184724688529968, "learning_rate": 9.709146043291906e-05, "loss": 0.1151, "step": 9220 }, { "grad_norm": 0.6249889135360718, "learning_rate": 9.70821913182378e-05, "loss": 0.0785, "step": 9230 }, { "grad_norm": 0.7518088817596436, "learning_rate": 9.707290790122995e-05, "loss": 0.1179, "step": 9240 }, { "grad_norm": 0.7891756892204285, "learning_rate": 9.706361018471557e-05, "loss": 0.0986, "step": 9250 }, { "grad_norm": 0.7314367294311523, "learning_rate": 9.705429817151906e-05, "loss": 0.1582, "step": 9260 }, { "grad_norm": 0.6632882952690125, "learning_rate": 9.704497186446917e-05, "loss": 0.1103, "step": 9270 }, { "grad_norm": 0.8936657905578613, "learning_rate": 9.703563126639896e-05, "loss": 0.1239, "step": 9280 }, { "grad_norm": 0.7972515225410461, "learning_rate": 9.70262763801459e-05, "loss": 0.1011, "step": 9290 }, { "grad_norm": 0.8339150547981262, "learning_rate": 9.701690720855171e-05, "loss": 0.1015, "step": 9300 }, { "grad_norm": 1.2597222328186035, "learning_rate": 9.700752375446253e-05, "loss": 0.1127, "step": 9310 }, { "grad_norm": 1.1444292068481445, "learning_rate": 9.69981260207288e-05, "loss": 0.1041, "step": 9320 }, { "grad_norm": 0.6418647170066833, "learning_rate": 9.698871401020529e-05, "loss": 0.1122, "step": 9330 }, { "grad_norm": 1.1342511177062988, "learning_rate": 9.697928772575112e-05, "loss": 0.1104, "step": 9340 }, { "grad_norm": 0.655710756778717, "learning_rate": 9.696984717022976e-05, "loss": 0.092, "step": 9350 }, { "grad_norm": 0.6765821576118469, "learning_rate": 9.6960392346509e-05, "loss": 0.0949, "step": 9360 }, { "grad_norm": 1.0132502317428589, "learning_rate": 9.695092325746097e-05, "loss": 0.0991, "step": 9370 }, { "grad_norm": 0.7932825088500977, "learning_rate": 9.694143990596211e-05, "loss": 0.1271, "step": 9380 }, { "grad_norm": 0.7939724922180176, "learning_rate": 9.693194229489325e-05, "loss": 0.0964, "step": 9390 }, { "grad_norm": 1.1609553098678589, "learning_rate": 9.692243042713944e-05, "loss": 0.1106, "step": 9400 }, { "grad_norm": 0.56732177734375, "learning_rate": 9.691290430559022e-05, "loss": 0.0998, "step": 9410 }, { "grad_norm": 0.8257636427879333, "learning_rate": 9.690336393313932e-05, "loss": 0.1192, "step": 9420 }, { "grad_norm": 0.7910316586494446, "learning_rate": 9.689380931268487e-05, "loss": 0.0976, "step": 9430 }, { "grad_norm": 0.6685461401939392, "learning_rate": 9.688424044712932e-05, "loss": 0.0938, "step": 9440 }, { "grad_norm": 0.7772121429443359, "learning_rate": 9.687465733937942e-05, "loss": 0.1055, "step": 9450 }, { "grad_norm": 0.6659524440765381, "learning_rate": 9.686505999234627e-05, "loss": 0.1298, "step": 9460 }, { "grad_norm": 0.7583153247833252, "learning_rate": 9.685544840894529e-05, "loss": 0.1217, "step": 9470 }, { "grad_norm": 0.6153129935264587, "learning_rate": 9.684582259209624e-05, "loss": 0.0831, "step": 9480 }, { "grad_norm": 0.8033234477043152, "learning_rate": 9.683618254472317e-05, "loss": 0.0992, "step": 9490 }, { "grad_norm": 0.8578560948371887, "learning_rate": 9.682652826975449e-05, "loss": 0.1022, "step": 9500 }, { "grad_norm": 0.8264133334159851, "learning_rate": 9.681685977012291e-05, "loss": 0.1044, "step": 9510 }, { "grad_norm": 1.0219799280166626, "learning_rate": 9.680717704876546e-05, "loss": 0.1132, "step": 9520 }, { "grad_norm": 0.7277815341949463, "learning_rate": 9.679748010862349e-05, "loss": 0.1087, "step": 9530 }, { "grad_norm": 1.2879306077957153, "learning_rate": 9.678776895264267e-05, "loss": 0.1167, "step": 9540 }, { "grad_norm": 0.7438074350357056, "learning_rate": 9.6778043583773e-05, "loss": 0.1314, "step": 9550 }, { "grad_norm": 0.9440454244613647, "learning_rate": 9.67683040049688e-05, "loss": 0.1095, "step": 9560 }, { "grad_norm": 0.767657458782196, "learning_rate": 9.675855021918869e-05, "loss": 0.0907, "step": 9570 }, { "grad_norm": 0.7848324179649353, "learning_rate": 9.674878222939561e-05, "loss": 0.1022, "step": 9580 }, { "grad_norm": 0.8740163445472717, "learning_rate": 9.673900003855681e-05, "loss": 0.1052, "step": 9590 }, { "grad_norm": 0.7634250521659851, "learning_rate": 9.672920364964389e-05, "loss": 0.1127, "step": 9600 }, { "grad_norm": 0.995190441608429, "learning_rate": 9.671939306563269e-05, "loss": 0.1079, "step": 9610 }, { "grad_norm": 0.7369033098220825, "learning_rate": 9.670956828950345e-05, "loss": 0.101, "step": 9620 }, { "grad_norm": 0.8552210330963135, "learning_rate": 9.669972932424065e-05, "loss": 0.0802, "step": 9630 }, { "grad_norm": 0.6591954827308655, "learning_rate": 9.668987617283312e-05, "loss": 0.1154, "step": 9640 }, { "grad_norm": 0.8721870183944702, "learning_rate": 9.668000883827397e-05, "loss": 0.1099, "step": 9650 }, { "grad_norm": 0.8117443919181824, "learning_rate": 9.667012732356067e-05, "loss": 0.112, "step": 9660 }, { "grad_norm": 0.6947233080863953, "learning_rate": 9.666023163169493e-05, "loss": 0.0933, "step": 9670 }, { "grad_norm": 1.174641489982605, "learning_rate": 9.665032176568281e-05, "loss": 0.0992, "step": 9680 }, { "grad_norm": 0.8493891358375549, "learning_rate": 9.664039772853469e-05, "loss": 0.1088, "step": 9690 }, { "grad_norm": 1.043845772743225, "learning_rate": 9.663045952326518e-05, "loss": 0.0846, "step": 9700 }, { "grad_norm": 0.7277197241783142, "learning_rate": 9.662050715289328e-05, "loss": 0.0864, "step": 9710 }, { "grad_norm": 0.785610556602478, "learning_rate": 9.661054062044226e-05, "loss": 0.0794, "step": 9720 }, { "grad_norm": 0.8552690148353577, "learning_rate": 9.660055992893968e-05, "loss": 0.1071, "step": 9730 }, { "grad_norm": 0.6572514772415161, "learning_rate": 9.659056508141739e-05, "loss": 0.0859, "step": 9740 }, { "grad_norm": 0.7947938442230225, "learning_rate": 9.658055608091161e-05, "loss": 0.1094, "step": 9750 }, { "grad_norm": 0.694635808467865, "learning_rate": 9.657053293046276e-05, "loss": 0.1056, "step": 9760 }, { "grad_norm": 0.7915071249008179, "learning_rate": 9.656049563311564e-05, "loss": 0.1267, "step": 9770 }, { "grad_norm": 0.7630674242973328, "learning_rate": 9.655044419191929e-05, "loss": 0.1135, "step": 9780 }, { "grad_norm": 0.5738704800605774, "learning_rate": 9.654037860992711e-05, "loss": 0.0917, "step": 9790 }, { "grad_norm": 0.6831696629524231, "learning_rate": 9.653029889019672e-05, "loss": 0.0763, "step": 9800 }, { "grad_norm": 0.9226415157318115, "learning_rate": 9.65202050357901e-05, "loss": 0.0967, "step": 9810 }, { "grad_norm": 0.7662410140037537, "learning_rate": 9.651009704977347e-05, "loss": 0.1134, "step": 9820 }, { "grad_norm": 0.8188552260398865, "learning_rate": 9.649997493521738e-05, "loss": 0.1212, "step": 9830 }, { "grad_norm": 1.0373748540878296, "learning_rate": 9.64898386951967e-05, "loss": 0.098, "step": 9840 }, { "grad_norm": 0.8655717372894287, "learning_rate": 9.647968833279049e-05, "loss": 0.1204, "step": 9850 }, { "grad_norm": 0.758171021938324, "learning_rate": 9.646952385108218e-05, "loss": 0.0902, "step": 9860 }, { "grad_norm": 0.8766511082649231, "learning_rate": 9.645934525315951e-05, "loss": 0.101, "step": 9870 }, { "grad_norm": 0.5935706496238708, "learning_rate": 9.644915254211442e-05, "loss": 0.0927, "step": 9880 }, { "grad_norm": 0.7341682314872742, "learning_rate": 9.643894572104321e-05, "loss": 0.106, "step": 9890 }, { "grad_norm": 0.5879935026168823, "learning_rate": 9.642872479304644e-05, "loss": 0.0915, "step": 9900 }, { "grad_norm": 0.7321476340293884, "learning_rate": 9.641848976122895e-05, "loss": 0.107, "step": 9910 }, { "grad_norm": 1.024524450302124, "learning_rate": 9.64082406286999e-05, "loss": 0.1112, "step": 9920 }, { "grad_norm": 0.8128095269203186, "learning_rate": 9.639797739857269e-05, "loss": 0.1158, "step": 9930 }, { "grad_norm": 0.792219877243042, "learning_rate": 9.638770007396498e-05, "loss": 0.1137, "step": 9940 }, { "grad_norm": 0.881206750869751, "learning_rate": 9.63774086579988e-05, "loss": 0.1042, "step": 9950 }, { "grad_norm": 1.0010178089141846, "learning_rate": 9.63671031538004e-05, "loss": 0.1045, "step": 9960 }, { "grad_norm": 0.6570562124252319, "learning_rate": 9.635678356450031e-05, "loss": 0.0951, "step": 9970 }, { "grad_norm": 1.050619125366211, "learning_rate": 9.634644989323336e-05, "loss": 0.1252, "step": 9980 }, { "grad_norm": 0.6521835327148438, "learning_rate": 9.633610214313861e-05, "loss": 0.0956, "step": 9990 }, { "grad_norm": 0.7554346919059753, "learning_rate": 9.632574031735951e-05, "loss": 0.1286, "step": 10000 }, { "grad_norm": 0.6573777794837952, "learning_rate": 9.631536441904364e-05, "loss": 0.1128, "step": 10010 }, { "grad_norm": 0.5520334839820862, "learning_rate": 9.630497445134293e-05, "loss": 0.0851, "step": 10020 }, { "grad_norm": 0.8502277731895447, "learning_rate": 9.62945704174136e-05, "loss": 0.1017, "step": 10030 }, { "grad_norm": 1.000646948814392, "learning_rate": 9.628415232041612e-05, "loss": 0.1001, "step": 10040 }, { "grad_norm": 0.6626638174057007, "learning_rate": 9.627372016351524e-05, "loss": 0.0927, "step": 10050 }, { "grad_norm": 0.7191066145896912, "learning_rate": 9.626327394987995e-05, "loss": 0.077, "step": 10060 }, { "grad_norm": 0.7356355786323547, "learning_rate": 9.625281368268355e-05, "loss": 0.0767, "step": 10070 }, { "grad_norm": 0.5464274883270264, "learning_rate": 9.624233936510357e-05, "loss": 0.091, "step": 10080 }, { "grad_norm": 0.6635950207710266, "learning_rate": 9.623185100032187e-05, "loss": 0.1247, "step": 10090 }, { "grad_norm": 0.7459904551506042, "learning_rate": 9.62213485915245e-05, "loss": 0.0903, "step": 10100 }, { "grad_norm": 0.6214340329170227, "learning_rate": 9.621083214190186e-05, "loss": 0.0851, "step": 10110 }, { "grad_norm": 0.5836806893348694, "learning_rate": 9.62003016546485e-05, "loss": 0.0802, "step": 10120 }, { "grad_norm": 0.7187566757202148, "learning_rate": 9.618975713296339e-05, "loss": 0.113, "step": 10130 }, { "grad_norm": 0.6632165312767029, "learning_rate": 9.61791985800496e-05, "loss": 0.1018, "step": 10140 }, { "grad_norm": 0.7439195513725281, "learning_rate": 9.616862599911458e-05, "loss": 0.1287, "step": 10150 }, { "grad_norm": 1.0786354541778564, "learning_rate": 9.615803939337e-05, "loss": 0.101, "step": 10160 }, { "grad_norm": 0.76736980676651, "learning_rate": 9.614743876603178e-05, "loss": 0.1021, "step": 10170 }, { "grad_norm": 0.8249619603157043, "learning_rate": 9.613682412032013e-05, "loss": 0.0979, "step": 10180 }, { "grad_norm": 0.6653015613555908, "learning_rate": 9.612619545945947e-05, "loss": 0.0914, "step": 10190 }, { "grad_norm": 0.8967289328575134, "learning_rate": 9.611555278667852e-05, "loss": 0.1019, "step": 10200 }, { "grad_norm": 0.681297779083252, "learning_rate": 9.610489610521024e-05, "loss": 0.0941, "step": 10210 }, { "grad_norm": 0.9086214303970337, "learning_rate": 9.609422541829187e-05, "loss": 0.1174, "step": 10220 }, { "grad_norm": 0.7292980551719666, "learning_rate": 9.608354072916486e-05, "loss": 0.127, "step": 10230 }, { "grad_norm": 0.9860081672668457, "learning_rate": 9.607284204107493e-05, "loss": 0.1286, "step": 10240 }, { "grad_norm": 0.7455892562866211, "learning_rate": 9.606212935727208e-05, "loss": 0.1232, "step": 10250 }, { "grad_norm": 0.6556670665740967, "learning_rate": 9.605140268101052e-05, "loss": 0.1106, "step": 10260 }, { "grad_norm": 0.5136135220527649, "learning_rate": 9.604066201554875e-05, "loss": 0.094, "step": 10270 }, { "grad_norm": 0.9899166226387024, "learning_rate": 9.60299073641495e-05, "loss": 0.0875, "step": 10280 }, { "grad_norm": 0.6962774395942688, "learning_rate": 9.601913873007974e-05, "loss": 0.0799, "step": 10290 }, { "grad_norm": 0.6558870673179626, "learning_rate": 9.60083561166107e-05, "loss": 0.0779, "step": 10300 }, { "grad_norm": 0.5851113796234131, "learning_rate": 9.599755952701783e-05, "loss": 0.1148, "step": 10310 }, { "grad_norm": 0.7624752521514893, "learning_rate": 9.598674896458089e-05, "loss": 0.1021, "step": 10320 }, { "grad_norm": 0.6769101023674011, "learning_rate": 9.597592443258383e-05, "loss": 0.0995, "step": 10330 }, { "grad_norm": 0.7726882696151733, "learning_rate": 9.596508593431483e-05, "loss": 0.0945, "step": 10340 }, { "grad_norm": 0.7333389520645142, "learning_rate": 9.59542334730664e-05, "loss": 0.0783, "step": 10350 }, { "grad_norm": 0.5222392082214355, "learning_rate": 9.594336705213516e-05, "loss": 0.0891, "step": 10360 }, { "grad_norm": 0.7344526052474976, "learning_rate": 9.593248667482208e-05, "loss": 0.1067, "step": 10370 }, { "grad_norm": 0.8879686594009399, "learning_rate": 9.592159234443233e-05, "loss": 0.1326, "step": 10380 }, { "grad_norm": 0.6837398409843445, "learning_rate": 9.59106840642753e-05, "loss": 0.0826, "step": 10390 }, { "grad_norm": 0.6671168804168701, "learning_rate": 9.589976183766467e-05, "loss": 0.0849, "step": 10400 }, { "grad_norm": 0.7444174885749817, "learning_rate": 9.58888256679183e-05, "loss": 0.0929, "step": 10410 }, { "grad_norm": 0.641899585723877, "learning_rate": 9.587787555835832e-05, "loss": 0.0968, "step": 10420 }, { "grad_norm": 0.6754711270332336, "learning_rate": 9.586691151231107e-05, "loss": 0.0943, "step": 10430 }, { "grad_norm": 0.7559735774993896, "learning_rate": 9.585593353310715e-05, "loss": 0.0997, "step": 10440 }, { "grad_norm": 0.90575110912323, "learning_rate": 9.58449416240814e-05, "loss": 0.1034, "step": 10450 }, { "grad_norm": 0.7931468486785889, "learning_rate": 9.583393578857283e-05, "loss": 0.1032, "step": 10460 }, { "grad_norm": 0.6272667050361633, "learning_rate": 9.582291602992474e-05, "loss": 0.0932, "step": 10470 }, { "grad_norm": 0.7082711458206177, "learning_rate": 9.581188235148466e-05, "loss": 0.081, "step": 10480 }, { "grad_norm": 0.6655589938163757, "learning_rate": 9.58008347566043e-05, "loss": 0.1146, "step": 10490 }, { "grad_norm": 0.7848440408706665, "learning_rate": 9.578977324863965e-05, "loss": 0.1046, "step": 10500 }, { "grad_norm": 0.8168988823890686, "learning_rate": 9.577869783095089e-05, "loss": 0.0829, "step": 10510 }, { "grad_norm": 0.7919855713844299, "learning_rate": 9.576760850690245e-05, "loss": 0.1143, "step": 10520 }, { "grad_norm": 0.6539299488067627, "learning_rate": 9.575650527986298e-05, "loss": 0.097, "step": 10530 }, { "grad_norm": 0.8023370504379272, "learning_rate": 9.574538815320531e-05, "loss": 0.0915, "step": 10540 }, { "grad_norm": 0.9734065532684326, "learning_rate": 9.573425713030656e-05, "loss": 0.0989, "step": 10550 }, { "grad_norm": 0.8131512403488159, "learning_rate": 9.572311221454806e-05, "loss": 0.1194, "step": 10560 }, { "grad_norm": 1.0002405643463135, "learning_rate": 9.57119534093153e-05, "loss": 0.1114, "step": 10570 }, { "grad_norm": 0.6860536336898804, "learning_rate": 9.570078071799806e-05, "loss": 0.101, "step": 10580 }, { "grad_norm": 0.7358897924423218, "learning_rate": 9.568959414399028e-05, "loss": 0.1168, "step": 10590 }, { "grad_norm": 0.5913292169570923, "learning_rate": 9.567839369069018e-05, "loss": 0.0956, "step": 10600 }, { "grad_norm": 0.7161828875541687, "learning_rate": 9.566717936150013e-05, "loss": 0.089, "step": 10610 }, { "grad_norm": 0.8986417055130005, "learning_rate": 9.565595115982678e-05, "loss": 0.081, "step": 10620 }, { "grad_norm": 1.0625656843185425, "learning_rate": 9.564470908908094e-05, "loss": 0.1072, "step": 10630 }, { "grad_norm": 0.7088438868522644, "learning_rate": 9.563345315267764e-05, "loss": 0.0941, "step": 10640 }, { "grad_norm": 0.6803935766220093, "learning_rate": 9.562218335403616e-05, "loss": 0.1232, "step": 10650 }, { "grad_norm": 0.7972970008850098, "learning_rate": 9.561089969657999e-05, "loss": 0.1135, "step": 10660 }, { "grad_norm": 0.6893370151519775, "learning_rate": 9.559960218373673e-05, "loss": 0.0988, "step": 10670 }, { "grad_norm": 0.8587032556533813, "learning_rate": 9.558829081893836e-05, "loss": 0.0955, "step": 10680 }, { "grad_norm": 0.7228196263313293, "learning_rate": 9.55769656056209e-05, "loss": 0.0778, "step": 10690 }, { "grad_norm": 0.8533850908279419, "learning_rate": 9.556562654722469e-05, "loss": 0.0981, "step": 10700 }, { "grad_norm": 0.8784194588661194, "learning_rate": 9.555427364719422e-05, "loss": 0.0857, "step": 10710 }, { "grad_norm": 1.032192349433899, "learning_rate": 9.55429069089782e-05, "loss": 0.1083, "step": 10720 }, { "grad_norm": 0.7604865431785583, "learning_rate": 9.553152633602956e-05, "loss": 0.0792, "step": 10730 }, { "grad_norm": 0.638853132724762, "learning_rate": 9.552013193180543e-05, "loss": 0.0822, "step": 10740 }, { "grad_norm": 0.7586537003517151, "learning_rate": 9.550872369976707e-05, "loss": 0.0898, "step": 10750 }, { "grad_norm": 0.7991253137588501, "learning_rate": 9.549730164338007e-05, "loss": 0.0816, "step": 10760 }, { "grad_norm": 0.8468170166015625, "learning_rate": 9.548586576611408e-05, "loss": 0.1073, "step": 10770 }, { "grad_norm": 1.1445173025131226, "learning_rate": 9.54744160714431e-05, "loss": 0.0926, "step": 10780 }, { "grad_norm": 0.7529821395874023, "learning_rate": 9.546295256284516e-05, "loss": 0.1287, "step": 10790 }, { "grad_norm": 0.7761058211326599, "learning_rate": 9.545147524380265e-05, "loss": 0.1031, "step": 10800 }, { "grad_norm": 0.774732232093811, "learning_rate": 9.543998411780201e-05, "loss": 0.0862, "step": 10810 }, { "grad_norm": 0.6033479571342468, "learning_rate": 9.542847918833397e-05, "loss": 0.0868, "step": 10820 }, { "grad_norm": 0.6946994066238403, "learning_rate": 9.541696045889343e-05, "loss": 0.096, "step": 10830 }, { "grad_norm": 0.7863333225250244, "learning_rate": 9.540542793297947e-05, "loss": 0.0996, "step": 10840 }, { "grad_norm": 0.8448349833488464, "learning_rate": 9.539388161409537e-05, "loss": 0.088, "step": 10850 }, { "grad_norm": 0.9593817591667175, "learning_rate": 9.538232150574857e-05, "loss": 0.1048, "step": 10860 }, { "grad_norm": 0.7169553637504578, "learning_rate": 9.537074761145076e-05, "loss": 0.1262, "step": 10870 }, { "grad_norm": 0.8398162722587585, "learning_rate": 9.535915993471778e-05, "loss": 0.1026, "step": 10880 }, { "grad_norm": 0.8202531337738037, "learning_rate": 9.534755847906964e-05, "loss": 0.0691, "step": 10890 }, { "grad_norm": 0.6409807801246643, "learning_rate": 9.533594324803057e-05, "loss": 0.0861, "step": 10900 }, { "grad_norm": 0.6975145936012268, "learning_rate": 9.532431424512895e-05, "loss": 0.0809, "step": 10910 }, { "grad_norm": 0.6850444078445435, "learning_rate": 9.531267147389741e-05, "loss": 0.0739, "step": 10920 }, { "grad_norm": 0.6534452438354492, "learning_rate": 9.530101493787266e-05, "loss": 0.0938, "step": 10930 }, { "grad_norm": 0.7110395431518555, "learning_rate": 9.528934464059571e-05, "loss": 0.0813, "step": 10940 }, { "grad_norm": 0.657096266746521, "learning_rate": 9.527766058561163e-05, "loss": 0.0812, "step": 10950 }, { "grad_norm": 0.7273487448692322, "learning_rate": 9.526596277646976e-05, "loss": 0.1035, "step": 10960 }, { "grad_norm": 0.7807024121284485, "learning_rate": 9.525425121672358e-05, "loss": 0.0774, "step": 10970 }, { "grad_norm": 0.9129526019096375, "learning_rate": 9.524252590993074e-05, "loss": 0.0837, "step": 10980 }, { "grad_norm": 0.9988386631011963, "learning_rate": 9.523078685965309e-05, "loss": 0.099, "step": 10990 }, { "grad_norm": 1.2114473581314087, "learning_rate": 9.521903406945664e-05, "loss": 0.1074, "step": 11000 }, { "grad_norm": 0.4958730936050415, "learning_rate": 9.520726754291158e-05, "loss": 0.1096, "step": 11010 }, { "grad_norm": 0.6309148073196411, "learning_rate": 9.519548728359227e-05, "loss": 0.1054, "step": 11020 }, { "grad_norm": 0.6931028366088867, "learning_rate": 9.518369329507726e-05, "loss": 0.0953, "step": 11030 }, { "grad_norm": 0.6584204435348511, "learning_rate": 9.51718855809492e-05, "loss": 0.0821, "step": 11040 }, { "grad_norm": 0.6908401846885681, "learning_rate": 9.516006414479502e-05, "loss": 0.083, "step": 11050 }, { "grad_norm": 0.697361946105957, "learning_rate": 9.514822899020572e-05, "loss": 0.0973, "step": 11060 }, { "grad_norm": 0.7282133102416992, "learning_rate": 9.513638012077654e-05, "loss": 0.0828, "step": 11070 }, { "grad_norm": 0.6271811127662659, "learning_rate": 9.512451754010683e-05, "loss": 0.0823, "step": 11080 }, { "grad_norm": 0.4980693757534027, "learning_rate": 9.511264125180013e-05, "loss": 0.0866, "step": 11090 }, { "grad_norm": 0.7858424782752991, "learning_rate": 9.510075125946414e-05, "loss": 0.0897, "step": 11100 }, { "grad_norm": 0.9361438751220703, "learning_rate": 9.508884756671075e-05, "loss": 0.0876, "step": 11110 }, { "grad_norm": 0.7217935919761658, "learning_rate": 9.507693017715596e-05, "loss": 0.0881, "step": 11120 }, { "grad_norm": 0.7381381988525391, "learning_rate": 9.506499909441997e-05, "loss": 0.0928, "step": 11130 }, { "grad_norm": 0.6492237448692322, "learning_rate": 9.505305432212713e-05, "loss": 0.0854, "step": 11140 }, { "grad_norm": 0.6891205906867981, "learning_rate": 9.504109586390595e-05, "loss": 0.1064, "step": 11150 }, { "grad_norm": 1.054131269454956, "learning_rate": 9.502912372338908e-05, "loss": 0.1178, "step": 11160 }, { "grad_norm": 0.6536109447479248, "learning_rate": 9.501713790421335e-05, "loss": 0.0873, "step": 11170 }, { "grad_norm": 0.5720517635345459, "learning_rate": 9.500513841001974e-05, "loss": 0.09, "step": 11180 }, { "grad_norm": 0.5467464327812195, "learning_rate": 9.499312524445336e-05, "loss": 0.0916, "step": 11190 }, { "grad_norm": 0.6595163345336914, "learning_rate": 9.498109841116351e-05, "loss": 0.1145, "step": 11200 }, { "grad_norm": 1.068716287612915, "learning_rate": 9.496905791380363e-05, "loss": 0.0988, "step": 11210 }, { "grad_norm": 0.5233677625656128, "learning_rate": 9.495700375603129e-05, "loss": 0.0953, "step": 11220 }, { "grad_norm": 0.6327134966850281, "learning_rate": 9.494493594150822e-05, "loss": 0.0809, "step": 11230 }, { "grad_norm": 0.7379282116889954, "learning_rate": 9.493285447390032e-05, "loss": 0.0994, "step": 11240 }, { "grad_norm": 0.7382003664970398, "learning_rate": 9.492075935687761e-05, "loss": 0.0824, "step": 11250 }, { "grad_norm": 0.7641199827194214, "learning_rate": 9.490865059411427e-05, "loss": 0.084, "step": 11260 }, { "grad_norm": 0.7192661762237549, "learning_rate": 9.489652818928863e-05, "loss": 0.0908, "step": 11270 }, { "grad_norm": 0.6236059069633484, "learning_rate": 9.488439214608315e-05, "loss": 0.0835, "step": 11280 }, { "grad_norm": 1.1402857303619385, "learning_rate": 9.487224246818444e-05, "loss": 0.0883, "step": 11290 }, { "grad_norm": 0.5234330892562866, "learning_rate": 9.486007915928325e-05, "loss": 0.1269, "step": 11300 }, { "grad_norm": 0.6561239957809448, "learning_rate": 9.484790222307448e-05, "loss": 0.1118, "step": 11310 }, { "grad_norm": 0.7074316740036011, "learning_rate": 9.483571166325716e-05, "loss": 0.0981, "step": 11320 }, { "grad_norm": 0.9431788325309753, "learning_rate": 9.482350748353444e-05, "loss": 0.1082, "step": 11330 }, { "grad_norm": 0.9797177910804749, "learning_rate": 9.481128968761363e-05, "loss": 0.0996, "step": 11340 }, { "grad_norm": 0.5875690579414368, "learning_rate": 9.479905827920621e-05, "loss": 0.0918, "step": 11350 }, { "grad_norm": 0.6747369170188904, "learning_rate": 9.478681326202773e-05, "loss": 0.1065, "step": 11360 }, { "grad_norm": 0.8234943151473999, "learning_rate": 9.477455463979791e-05, "loss": 0.0842, "step": 11370 }, { "grad_norm": 0.7487121224403381, "learning_rate": 9.476228241624059e-05, "loss": 0.1052, "step": 11380 }, { "grad_norm": 0.4524573087692261, "learning_rate": 9.474999659508374e-05, "loss": 0.0779, "step": 11390 }, { "grad_norm": 0.7738951444625854, "learning_rate": 9.47376971800595e-05, "loss": 0.0897, "step": 11400 }, { "grad_norm": 0.5518584251403809, "learning_rate": 9.472538417490409e-05, "loss": 0.0754, "step": 11410 }, { "grad_norm": 0.7581973075866699, "learning_rate": 9.471305758335784e-05, "loss": 0.1286, "step": 11420 }, { "grad_norm": 0.6669503450393677, "learning_rate": 9.47007174091653e-05, "loss": 0.084, "step": 11430 }, { "grad_norm": 0.6982443928718567, "learning_rate": 9.468836365607507e-05, "loss": 0.0925, "step": 11440 }, { "grad_norm": 0.648003101348877, "learning_rate": 9.467599632783988e-05, "loss": 0.1008, "step": 11450 }, { "grad_norm": 0.8954724073410034, "learning_rate": 9.466361542821662e-05, "loss": 0.1023, "step": 11460 }, { "grad_norm": 0.8644186854362488, "learning_rate": 9.465122096096625e-05, "loss": 0.1013, "step": 11470 }, { "grad_norm": 0.8532023429870605, "learning_rate": 9.463881292985391e-05, "loss": 0.1179, "step": 11480 }, { "grad_norm": 0.8164891600608826, "learning_rate": 9.462639133864881e-05, "loss": 0.0929, "step": 11490 }, { "grad_norm": 0.5538641810417175, "learning_rate": 9.461395619112432e-05, "loss": 0.0657, "step": 11500 }, { "grad_norm": 0.68805992603302, "learning_rate": 9.460150749105791e-05, "loss": 0.0968, "step": 11510 }, { "grad_norm": 0.6488337516784668, "learning_rate": 9.458904524223116e-05, "loss": 0.0843, "step": 11520 }, { "grad_norm": 0.5386496782302856, "learning_rate": 9.457656944842976e-05, "loss": 0.0812, "step": 11530 }, { "grad_norm": 0.9095367789268494, "learning_rate": 9.456408011344353e-05, "loss": 0.0952, "step": 11540 }, { "grad_norm": 0.8111530542373657, "learning_rate": 9.455157724106643e-05, "loss": 0.1101, "step": 11550 }, { "grad_norm": 0.6562020182609558, "learning_rate": 9.453906083509647e-05, "loss": 0.0979, "step": 11560 }, { "grad_norm": 0.7189951539039612, "learning_rate": 9.45265308993358e-05, "loss": 0.0889, "step": 11570 }, { "grad_norm": 0.8035237193107605, "learning_rate": 9.451398743759071e-05, "loss": 0.0909, "step": 11580 }, { "grad_norm": 0.8381906151771545, "learning_rate": 9.450143045367156e-05, "loss": 0.118, "step": 11590 }, { "grad_norm": 0.741667628288269, "learning_rate": 9.448885995139283e-05, "loss": 0.0818, "step": 11600 }, { "grad_norm": 0.8736934065818787, "learning_rate": 9.44762759345731e-05, "loss": 0.1017, "step": 11610 }, { "grad_norm": 0.8974677920341492, "learning_rate": 9.446367840703509e-05, "loss": 0.0869, "step": 11620 }, { "grad_norm": 0.9719070792198181, "learning_rate": 9.445106737260556e-05, "loss": 0.0864, "step": 11630 }, { "grad_norm": 0.7326439023017883, "learning_rate": 9.443844283511543e-05, "loss": 0.0908, "step": 11640 }, { "grad_norm": 0.8059300780296326, "learning_rate": 9.442580479839968e-05, "loss": 0.1033, "step": 11650 }, { "grad_norm": 0.7172240614891052, "learning_rate": 9.441315326629745e-05, "loss": 0.112, "step": 11660 }, { "grad_norm": 1.0920814275741577, "learning_rate": 9.44004882426519e-05, "loss": 0.1126, "step": 11670 }, { "grad_norm": 0.8529742956161499, "learning_rate": 9.438780973131037e-05, "loss": 0.0907, "step": 11680 }, { "grad_norm": 0.599350094795227, "learning_rate": 9.437511773612423e-05, "loss": 0.0892, "step": 11690 }, { "grad_norm": 1.1071550846099854, "learning_rate": 9.436241226094896e-05, "loss": 0.0971, "step": 11700 }, { "grad_norm": 0.6581785082817078, "learning_rate": 9.434969330964418e-05, "loss": 0.1047, "step": 11710 }, { "grad_norm": 0.7210835218429565, "learning_rate": 9.433696088607356e-05, "loss": 0.0968, "step": 11720 }, { "grad_norm": 0.6785051226615906, "learning_rate": 9.432421499410486e-05, "loss": 0.1022, "step": 11730 }, { "grad_norm": 0.8757756352424622, "learning_rate": 9.431145563760998e-05, "loss": 0.0977, "step": 11740 }, { "grad_norm": 0.5695882439613342, "learning_rate": 9.429868282046484e-05, "loss": 0.1014, "step": 11750 }, { "grad_norm": 0.7351513504981995, "learning_rate": 9.428589654654951e-05, "loss": 0.098, "step": 11760 }, { "grad_norm": 0.791091799736023, "learning_rate": 9.42730968197481e-05, "loss": 0.1055, "step": 11770 }, { "grad_norm": 0.770745575428009, "learning_rate": 9.426028364394883e-05, "loss": 0.0802, "step": 11780 }, { "grad_norm": 0.7891179323196411, "learning_rate": 9.424745702304402e-05, "loss": 0.1045, "step": 11790 }, { "grad_norm": 0.7478852868080139, "learning_rate": 9.423461696093006e-05, "loss": 0.1147, "step": 11800 }, { "grad_norm": 0.8623188734054565, "learning_rate": 9.422176346150741e-05, "loss": 0.1112, "step": 11810 }, { "grad_norm": 0.8101369142532349, "learning_rate": 9.420889652868063e-05, "loss": 0.1074, "step": 11820 }, { "grad_norm": 0.7258525490760803, "learning_rate": 9.419601616635836e-05, "loss": 0.1045, "step": 11830 }, { "grad_norm": 1.261204719543457, "learning_rate": 9.418312237845331e-05, "loss": 0.1046, "step": 11840 }, { "grad_norm": 0.6967996954917908, "learning_rate": 9.417021516888225e-05, "loss": 0.0763, "step": 11850 }, { "grad_norm": 0.7688332200050354, "learning_rate": 9.415729454156608e-05, "loss": 0.0967, "step": 11860 }, { "grad_norm": 0.7036973237991333, "learning_rate": 9.414436050042973e-05, "loss": 0.0807, "step": 11870 }, { "grad_norm": 0.5775085687637329, "learning_rate": 9.413141304940223e-05, "loss": 0.105, "step": 11880 }, { "grad_norm": 0.6145880818367004, "learning_rate": 9.411845219241666e-05, "loss": 0.0901, "step": 11890 }, { "grad_norm": 0.7364181876182556, "learning_rate": 9.410547793341021e-05, "loss": 0.095, "step": 11900 }, { "grad_norm": 0.9047418236732483, "learning_rate": 9.409249027632408e-05, "loss": 0.094, "step": 11910 }, { "grad_norm": 0.7381111979484558, "learning_rate": 9.407948922510362e-05, "loss": 0.0877, "step": 11920 }, { "grad_norm": 0.8154175281524658, "learning_rate": 9.406647478369817e-05, "loss": 0.0862, "step": 11930 }, { "grad_norm": 0.9642165303230286, "learning_rate": 9.405344695606118e-05, "loss": 0.1165, "step": 11940 }, { "grad_norm": 0.9045552611351013, "learning_rate": 9.404040574615018e-05, "loss": 0.1145, "step": 11950 }, { "grad_norm": 0.6971715092658997, "learning_rate": 9.402735115792674e-05, "loss": 0.1007, "step": 11960 }, { "grad_norm": 0.5973456501960754, "learning_rate": 9.401428319535649e-05, "loss": 0.0978, "step": 11970 }, { "grad_norm": 0.8174471855163574, "learning_rate": 9.400120186240912e-05, "loss": 0.1206, "step": 11980 }, { "grad_norm": 0.7644558548927307, "learning_rate": 9.398810716305844e-05, "loss": 0.1065, "step": 11990 }, { "grad_norm": 0.5804036259651184, "learning_rate": 9.397499910128222e-05, "loss": 0.1383, "step": 12000 }, { "grad_norm": 0.5370171070098877, "learning_rate": 9.396187768106237e-05, "loss": 0.1125, "step": 12010 }, { "grad_norm": 0.6834402084350586, "learning_rate": 9.394874290638482e-05, "loss": 0.0919, "step": 12020 }, { "grad_norm": 0.7506278157234192, "learning_rate": 9.393559478123959e-05, "loss": 0.1027, "step": 12030 }, { "grad_norm": 0.664983868598938, "learning_rate": 9.39224333096207e-05, "loss": 0.0848, "step": 12040 }, { "grad_norm": 0.6101111769676208, "learning_rate": 9.390925849552629e-05, "loss": 0.0917, "step": 12050 }, { "grad_norm": 0.6596612930297852, "learning_rate": 9.389607034295849e-05, "loss": 0.0925, "step": 12060 }, { "grad_norm": 0.5931715369224548, "learning_rate": 9.388286885592355e-05, "loss": 0.0868, "step": 12070 }, { "grad_norm": 0.6413942575454712, "learning_rate": 9.386965403843168e-05, "loss": 0.1084, "step": 12080 }, { "grad_norm": 0.7025232911109924, "learning_rate": 9.385642589449726e-05, "loss": 0.1128, "step": 12090 }, { "grad_norm": 0.643615186214447, "learning_rate": 9.38431844281386e-05, "loss": 0.0925, "step": 12100 }, { "grad_norm": 0.6669476628303528, "learning_rate": 9.38299296433781e-05, "loss": 0.0893, "step": 12110 }, { "grad_norm": 0.7479439377784729, "learning_rate": 9.381666154424226e-05, "loss": 0.0672, "step": 12120 }, { "grad_norm": 0.8121823668479919, "learning_rate": 9.380338013476157e-05, "loss": 0.0847, "step": 12130 }, { "grad_norm": 0.7104564905166626, "learning_rate": 9.379008541897054e-05, "loss": 0.0958, "step": 12140 }, { "grad_norm": 0.7040308117866516, "learning_rate": 9.377677740090777e-05, "loss": 0.087, "step": 12150 }, { "grad_norm": 0.7171938419342041, "learning_rate": 9.376345608461588e-05, "loss": 0.0914, "step": 12160 }, { "grad_norm": 0.5592693090438843, "learning_rate": 9.375012147414155e-05, "loss": 0.1052, "step": 12170 }, { "grad_norm": 0.6269899606704712, "learning_rate": 9.373677357353545e-05, "loss": 0.0947, "step": 12180 }, { "grad_norm": 0.6814196705818176, "learning_rate": 9.372341238685237e-05, "loss": 0.1126, "step": 12190 }, { "grad_norm": 0.7435757517814636, "learning_rate": 9.371003791815102e-05, "loss": 0.1038, "step": 12200 }, { "grad_norm": 0.8596900105476379, "learning_rate": 9.369665017149429e-05, "loss": 0.0755, "step": 12210 }, { "grad_norm": 0.7678030729293823, "learning_rate": 9.368324915094895e-05, "loss": 0.0811, "step": 12220 }, { "grad_norm": 0.6313386559486389, "learning_rate": 9.366983486058591e-05, "loss": 0.0843, "step": 12230 }, { "grad_norm": 0.7105013728141785, "learning_rate": 9.365640730448009e-05, "loss": 0.1039, "step": 12240 }, { "grad_norm": 0.8899890780448914, "learning_rate": 9.36429664867104e-05, "loss": 0.0827, "step": 12250 }, { "grad_norm": 0.8067851662635803, "learning_rate": 9.362951241135982e-05, "loss": 0.0839, "step": 12260 }, { "grad_norm": 0.7258862853050232, "learning_rate": 9.361604508251534e-05, "loss": 0.0917, "step": 12270 }, { "grad_norm": 0.5870862007141113, "learning_rate": 9.360256450426799e-05, "loss": 0.105, "step": 12280 }, { "grad_norm": 0.8460150957107544, "learning_rate": 9.358907068071279e-05, "loss": 0.1206, "step": 12290 }, { "grad_norm": 0.6859310865402222, "learning_rate": 9.357556361594882e-05, "loss": 0.1017, "step": 12300 }, { "grad_norm": 0.8634790182113647, "learning_rate": 9.356204331407917e-05, "loss": 0.1112, "step": 12310 }, { "grad_norm": 0.7588804364204407, "learning_rate": 9.354850977921094e-05, "loss": 0.123, "step": 12320 }, { "grad_norm": 0.7255669832229614, "learning_rate": 9.353496301545529e-05, "loss": 0.1114, "step": 12330 }, { "grad_norm": 0.757771909236908, "learning_rate": 9.352140302692733e-05, "loss": 0.0886, "step": 12340 }, { "grad_norm": 0.7333433628082275, "learning_rate": 9.350782981774627e-05, "loss": 0.0971, "step": 12350 }, { "grad_norm": 0.7229361534118652, "learning_rate": 9.349424339203526e-05, "loss": 0.0853, "step": 12360 }, { "grad_norm": 0.5824282169342041, "learning_rate": 9.34806437539215e-05, "loss": 0.1131, "step": 12370 }, { "grad_norm": 1.0077650547027588, "learning_rate": 9.346703090753622e-05, "loss": 0.1152, "step": 12380 }, { "grad_norm": 0.5740765929222107, "learning_rate": 9.345340485701461e-05, "loss": 0.1017, "step": 12390 }, { "grad_norm": 0.8538813591003418, "learning_rate": 9.343976560649595e-05, "loss": 0.0727, "step": 12400 }, { "grad_norm": 0.7789905667304993, "learning_rate": 9.342611316012344e-05, "loss": 0.09, "step": 12410 }, { "grad_norm": 0.7512446641921997, "learning_rate": 9.341244752204437e-05, "loss": 0.0988, "step": 12420 }, { "grad_norm": 0.5552610754966736, "learning_rate": 9.339876869640995e-05, "loss": 0.0781, "step": 12430 }, { "grad_norm": 0.9592671394348145, "learning_rate": 9.33850766873755e-05, "loss": 0.0826, "step": 12440 }, { "grad_norm": 0.7656723856925964, "learning_rate": 9.337137149910028e-05, "loss": 0.1046, "step": 12450 }, { "grad_norm": 0.6845630407333374, "learning_rate": 9.335765313574753e-05, "loss": 0.0754, "step": 12460 }, { "grad_norm": 0.4599151313304901, "learning_rate": 9.334392160148457e-05, "loss": 0.107, "step": 12470 }, { "grad_norm": 0.6621633172035217, "learning_rate": 9.333017690048264e-05, "loss": 0.0961, "step": 12480 }, { "grad_norm": 0.6777192950248718, "learning_rate": 9.331641903691706e-05, "loss": 0.1026, "step": 12490 }, { "grad_norm": 0.760710597038269, "learning_rate": 9.330264801496707e-05, "loss": 0.1134, "step": 12500 }, { "grad_norm": 0.6771512031555176, "learning_rate": 9.328886383881594e-05, "loss": 0.0781, "step": 12510 }, { "grad_norm": 0.6761650443077087, "learning_rate": 9.327506651265095e-05, "loss": 0.0878, "step": 12520 }, { "grad_norm": 0.7068939208984375, "learning_rate": 9.326125604066338e-05, "loss": 0.0828, "step": 12530 }, { "grad_norm": 0.5633059144020081, "learning_rate": 9.324743242704847e-05, "loss": 0.0899, "step": 12540 }, { "grad_norm": 0.7120540738105774, "learning_rate": 9.323359567600546e-05, "loss": 0.0773, "step": 12550 }, { "grad_norm": 0.7388077974319458, "learning_rate": 9.321974579173761e-05, "loss": 0.0995, "step": 12560 }, { "grad_norm": 0.8399096131324768, "learning_rate": 9.320588277845213e-05, "loss": 0.0876, "step": 12570 }, { "grad_norm": 0.6308977603912354, "learning_rate": 9.319200664036026e-05, "loss": 0.0882, "step": 12580 }, { "grad_norm": 0.6828495860099792, "learning_rate": 9.31781173816772e-05, "loss": 0.1011, "step": 12590 }, { "grad_norm": 0.8642390370368958, "learning_rate": 9.316421500662212e-05, "loss": 0.0819, "step": 12600 }, { "grad_norm": 0.603715717792511, "learning_rate": 9.31502995194182e-05, "loss": 0.0788, "step": 12610 }, { "grad_norm": 0.6993553042411804, "learning_rate": 9.31363709242926e-05, "loss": 0.0855, "step": 12620 }, { "grad_norm": 0.5295499563217163, "learning_rate": 9.312242922547647e-05, "loss": 0.0783, "step": 12630 }, { "grad_norm": 0.810971200466156, "learning_rate": 9.310847442720492e-05, "loss": 0.0797, "step": 12640 }, { "grad_norm": 0.5641019344329834, "learning_rate": 9.309450653371706e-05, "loss": 0.0981, "step": 12650 }, { "grad_norm": 0.6747065782546997, "learning_rate": 9.308052554925595e-05, "loss": 0.0843, "step": 12660 }, { "grad_norm": 0.533088207244873, "learning_rate": 9.306653147806867e-05, "loss": 0.1102, "step": 12670 }, { "grad_norm": 0.9245460033416748, "learning_rate": 9.305252432440622e-05, "loss": 0.0915, "step": 12680 }, { "grad_norm": 0.7131408452987671, "learning_rate": 9.303850409252361e-05, "loss": 0.09, "step": 12690 }, { "grad_norm": 0.7798769474029541, "learning_rate": 9.302447078667985e-05, "loss": 0.0831, "step": 12700 }, { "grad_norm": 0.7114702463150024, "learning_rate": 9.301042441113783e-05, "loss": 0.1032, "step": 12710 }, { "grad_norm": 0.744471549987793, "learning_rate": 9.299636497016451e-05, "loss": 0.0825, "step": 12720 }, { "grad_norm": 0.5510916113853455, "learning_rate": 9.298229246803076e-05, "loss": 0.0728, "step": 12730 }, { "grad_norm": 0.5863403081893921, "learning_rate": 9.296820690901144e-05, "loss": 0.0855, "step": 12740 }, { "grad_norm": 0.639465868473053, "learning_rate": 9.295410829738539e-05, "loss": 0.0778, "step": 12750 }, { "grad_norm": 0.7104703783988953, "learning_rate": 9.293999663743535e-05, "loss": 0.0757, "step": 12760 }, { "grad_norm": 0.5872398018836975, "learning_rate": 9.292587193344813e-05, "loss": 0.0814, "step": 12770 }, { "grad_norm": 0.9481039047241211, "learning_rate": 9.291173418971437e-05, "loss": 0.0926, "step": 12780 }, { "grad_norm": 0.7811170220375061, "learning_rate": 9.28975834105288e-05, "loss": 0.0803, "step": 12790 }, { "grad_norm": 0.7204156517982483, "learning_rate": 9.288341960019004e-05, "loss": 0.1099, "step": 12800 }, { "grad_norm": 0.7946172952651978, "learning_rate": 9.286924276300067e-05, "loss": 0.0825, "step": 12810 }, { "grad_norm": 0.6493054032325745, "learning_rate": 9.285505290326726e-05, "loss": 0.0846, "step": 12820 }, { "grad_norm": 0.6419131755828857, "learning_rate": 9.284085002530027e-05, "loss": 0.1174, "step": 12830 }, { "grad_norm": 0.6142742037773132, "learning_rate": 9.282663413341422e-05, "loss": 0.0703, "step": 12840 }, { "grad_norm": 0.905866265296936, "learning_rate": 9.281240523192747e-05, "loss": 0.0817, "step": 12850 }, { "grad_norm": 0.7930115461349487, "learning_rate": 9.279816332516242e-05, "loss": 0.0831, "step": 12860 }, { "grad_norm": 0.9548645615577698, "learning_rate": 9.278390841744536e-05, "loss": 0.0964, "step": 12870 }, { "grad_norm": 0.49563175439834595, "learning_rate": 9.276964051310658e-05, "loss": 0.0802, "step": 12880 }, { "grad_norm": 0.8132713437080383, "learning_rate": 9.275535961648027e-05, "loss": 0.0816, "step": 12890 }, { "grad_norm": 0.9097430109977722, "learning_rate": 9.274106573190459e-05, "loss": 0.0986, "step": 12900 }, { "grad_norm": 0.9130073189735413, "learning_rate": 9.272675886372168e-05, "loss": 0.0941, "step": 12910 }, { "grad_norm": 0.5098109841346741, "learning_rate": 9.271243901627754e-05, "loss": 0.0767, "step": 12920 }, { "grad_norm": 0.8090039491653442, "learning_rate": 9.269810619392219e-05, "loss": 0.0797, "step": 12930 }, { "grad_norm": 0.7035952210426331, "learning_rate": 9.268376040100955e-05, "loss": 0.0856, "step": 12940 }, { "grad_norm": 0.7929232120513916, "learning_rate": 9.266940164189752e-05, "loss": 0.0806, "step": 12950 }, { "grad_norm": 0.6968507170677185, "learning_rate": 9.265502992094787e-05, "loss": 0.0785, "step": 12960 }, { "grad_norm": 0.7355188727378845, "learning_rate": 9.264064524252638e-05, "loss": 0.0944, "step": 12970 }, { "grad_norm": 1.0819852352142334, "learning_rate": 9.262624761100271e-05, "loss": 0.1096, "step": 12980 }, { "grad_norm": 0.7223683595657349, "learning_rate": 9.261183703075051e-05, "loss": 0.0845, "step": 12990 }, { "grad_norm": 0.6536768078804016, "learning_rate": 9.259741350614733e-05, "loss": 0.0735, "step": 13000 }, { "grad_norm": 0.6277204155921936, "learning_rate": 9.258297704157464e-05, "loss": 0.069, "step": 13010 }, { "grad_norm": 0.6662781834602356, "learning_rate": 9.256852764141786e-05, "loss": 0.087, "step": 13020 }, { "grad_norm": 0.5892080664634705, "learning_rate": 9.255406531006634e-05, "loss": 0.1041, "step": 13030 }, { "grad_norm": 0.8801383376121521, "learning_rate": 9.253959005191335e-05, "loss": 0.0931, "step": 13040 }, { "grad_norm": 0.5701315402984619, "learning_rate": 9.25251018713561e-05, "loss": 0.0888, "step": 13050 }, { "grad_norm": 0.6929526925086975, "learning_rate": 9.251060077279571e-05, "loss": 0.0858, "step": 13060 }, { "grad_norm": 0.778118371963501, "learning_rate": 9.249608676063724e-05, "loss": 0.1022, "step": 13070 }, { "grad_norm": 0.9318115711212158, "learning_rate": 9.248155983928964e-05, "loss": 0.084, "step": 13080 }, { "grad_norm": 0.9019510746002197, "learning_rate": 9.246702001316583e-05, "loss": 0.073, "step": 13090 }, { "grad_norm": 0.9644911289215088, "learning_rate": 9.245246728668262e-05, "loss": 0.0865, "step": 13100 }, { "grad_norm": 0.7470705509185791, "learning_rate": 9.243790166426073e-05, "loss": 0.0903, "step": 13110 }, { "grad_norm": 0.7738308906555176, "learning_rate": 9.242332315032484e-05, "loss": 0.0876, "step": 13120 }, { "grad_norm": 0.7532604336738586, "learning_rate": 9.240873174930349e-05, "loss": 0.0881, "step": 13130 }, { "grad_norm": 0.7296620607376099, "learning_rate": 9.239412746562917e-05, "loss": 0.1083, "step": 13140 }, { "grad_norm": 0.6293517351150513, "learning_rate": 9.237951030373828e-05, "loss": 0.08, "step": 13150 }, { "grad_norm": 0.5246747136116028, "learning_rate": 9.236488026807113e-05, "loss": 0.0923, "step": 13160 }, { "grad_norm": 1.0018202066421509, "learning_rate": 9.235023736307193e-05, "loss": 0.124, "step": 13170 }, { "grad_norm": 0.5998507738113403, "learning_rate": 9.233558159318881e-05, "loss": 0.0898, "step": 13180 }, { "grad_norm": 1.0709987878799438, "learning_rate": 9.232091296287382e-05, "loss": 0.1108, "step": 13190 }, { "grad_norm": 0.683846116065979, "learning_rate": 9.230623147658288e-05, "loss": 0.1187, "step": 13200 }, { "grad_norm": 0.9294993877410889, "learning_rate": 9.229153713877586e-05, "loss": 0.0804, "step": 13210 }, { "grad_norm": 0.5271837711334229, "learning_rate": 9.227682995391649e-05, "loss": 0.0646, "step": 13220 }, { "grad_norm": 0.4563562870025635, "learning_rate": 9.226210992647243e-05, "loss": 0.0822, "step": 13230 }, { "grad_norm": 0.5289033055305481, "learning_rate": 9.224737706091525e-05, "loss": 0.066, "step": 13240 }, { "grad_norm": 0.5639054179191589, "learning_rate": 9.223263136172039e-05, "loss": 0.0849, "step": 13250 }, { "grad_norm": 0.9231234192848206, "learning_rate": 9.22178728333672e-05, "loss": 0.0815, "step": 13260 }, { "grad_norm": 0.5771026611328125, "learning_rate": 9.220310148033897e-05, "loss": 0.0747, "step": 13270 }, { "grad_norm": 0.97658771276474, "learning_rate": 9.21883173071228e-05, "loss": 0.0702, "step": 13280 }, { "grad_norm": 0.628731906414032, "learning_rate": 9.217352031820976e-05, "loss": 0.0785, "step": 13290 }, { "grad_norm": 0.6783809065818787, "learning_rate": 9.215871051809477e-05, "loss": 0.1363, "step": 13300 }, { "grad_norm": 0.682443380355835, "learning_rate": 9.214388791127666e-05, "loss": 0.0788, "step": 13310 }, { "grad_norm": 0.6722790002822876, "learning_rate": 9.212905250225814e-05, "loss": 0.0723, "step": 13320 }, { "grad_norm": 0.7347409725189209, "learning_rate": 9.211420429554583e-05, "loss": 0.0836, "step": 13330 }, { "grad_norm": 0.693667471408844, "learning_rate": 9.209934329565022e-05, "loss": 0.0666, "step": 13340 }, { "grad_norm": 0.7125821113586426, "learning_rate": 9.208446950708568e-05, "loss": 0.0892, "step": 13350 }, { "grad_norm": 0.5338683128356934, "learning_rate": 9.20695829343705e-05, "loss": 0.0908, "step": 13360 }, { "grad_norm": 0.566724419593811, "learning_rate": 9.205468358202678e-05, "loss": 0.0842, "step": 13370 }, { "grad_norm": 0.6255565881729126, "learning_rate": 9.203977145458059e-05, "loss": 0.0875, "step": 13380 }, { "grad_norm": 0.8019883036613464, "learning_rate": 9.202484655656182e-05, "loss": 0.0992, "step": 13390 }, { "grad_norm": 0.8315973877906799, "learning_rate": 9.200990889250427e-05, "loss": 0.083, "step": 13400 }, { "grad_norm": 0.5661808848381042, "learning_rate": 9.19949584669456e-05, "loss": 0.0834, "step": 13410 }, { "grad_norm": 0.8004333972930908, "learning_rate": 9.197999528442738e-05, "loss": 0.0817, "step": 13420 }, { "grad_norm": 0.8358808159828186, "learning_rate": 9.196501934949499e-05, "loss": 0.1106, "step": 13430 }, { "grad_norm": 0.538553774356842, "learning_rate": 9.195003066669776e-05, "loss": 0.0781, "step": 13440 }, { "grad_norm": 0.6841035485267639, "learning_rate": 9.193502924058884e-05, "loss": 0.0842, "step": 13450 }, { "grad_norm": 0.5630517601966858, "learning_rate": 9.192001507572526e-05, "loss": 0.0973, "step": 13460 }, { "grad_norm": 1.062947392463684, "learning_rate": 9.190498817666793e-05, "loss": 0.1258, "step": 13470 }, { "grad_norm": 0.48851555585861206, "learning_rate": 9.188994854798163e-05, "loss": 0.0924, "step": 13480 }, { "grad_norm": 0.8924106955528259, "learning_rate": 9.187489619423499e-05, "loss": 0.097, "step": 13490 }, { "grad_norm": 0.6327501535415649, "learning_rate": 9.185983112000056e-05, "loss": 0.0876, "step": 13500 }, { "grad_norm": 0.5920225381851196, "learning_rate": 9.184475332985464e-05, "loss": 0.0873, "step": 13510 }, { "grad_norm": 0.5247073769569397, "learning_rate": 9.182966282837754e-05, "loss": 0.0831, "step": 13520 }, { "grad_norm": 0.640635073184967, "learning_rate": 9.18145596201533e-05, "loss": 0.0768, "step": 13530 }, { "grad_norm": 0.6074507236480713, "learning_rate": 9.179944370976991e-05, "loss": 0.0737, "step": 13540 }, { "grad_norm": 0.5550722479820251, "learning_rate": 9.178431510181918e-05, "loss": 0.0781, "step": 13550 }, { "grad_norm": 0.6812068819999695, "learning_rate": 9.176917380089675e-05, "loss": 0.0915, "step": 13560 }, { "grad_norm": 0.6056456565856934, "learning_rate": 9.175401981160219e-05, "loss": 0.0825, "step": 13570 }, { "grad_norm": 0.5434857606887817, "learning_rate": 9.173885313853885e-05, "loss": 0.0662, "step": 13580 }, { "grad_norm": 0.7675435543060303, "learning_rate": 9.172367378631398e-05, "loss": 0.0756, "step": 13590 }, { "grad_norm": 0.6757180690765381, "learning_rate": 9.170848175953866e-05, "loss": 0.1123, "step": 13600 }, { "grad_norm": 0.7729046940803528, "learning_rate": 9.169327706282784e-05, "loss": 0.0811, "step": 13610 }, { "grad_norm": 0.566502571105957, "learning_rate": 9.167805970080029e-05, "loss": 0.0807, "step": 13620 }, { "grad_norm": 0.6569561958312988, "learning_rate": 9.166282967807864e-05, "loss": 0.0794, "step": 13630 }, { "grad_norm": 0.7487124800682068, "learning_rate": 9.16475869992894e-05, "loss": 0.1025, "step": 13640 }, { "grad_norm": 0.7384321093559265, "learning_rate": 9.163233166906284e-05, "loss": 0.0806, "step": 13650 }, { "grad_norm": 0.7520129084587097, "learning_rate": 9.161706369203317e-05, "loss": 0.0855, "step": 13660 }, { "grad_norm": 0.46042320132255554, "learning_rate": 9.16017830728384e-05, "loss": 0.087, "step": 13670 }, { "grad_norm": 0.6656346917152405, "learning_rate": 9.158648981612035e-05, "loss": 0.083, "step": 13680 }, { "grad_norm": 0.723717212677002, "learning_rate": 9.157118392652472e-05, "loss": 0.0973, "step": 13690 }, { "grad_norm": 0.5322484374046326, "learning_rate": 9.155586540870104e-05, "loss": 0.0768, "step": 13700 }, { "grad_norm": 0.7509561777114868, "learning_rate": 9.154053426730267e-05, "loss": 0.0779, "step": 13710 }, { "grad_norm": 0.830671489238739, "learning_rate": 9.15251905069868e-05, "loss": 0.0688, "step": 13720 }, { "grad_norm": 0.5118680596351624, "learning_rate": 9.150983413241446e-05, "loss": 0.1062, "step": 13730 }, { "grad_norm": 0.7752504944801331, "learning_rate": 9.149446514825051e-05, "loss": 0.0818, "step": 13740 }, { "grad_norm": 0.833045482635498, "learning_rate": 9.147908355916365e-05, "loss": 0.0894, "step": 13750 }, { "grad_norm": 0.5712637901306152, "learning_rate": 9.146368936982642e-05, "loss": 0.1168, "step": 13760 }, { "grad_norm": 0.6801044940948486, "learning_rate": 9.144828258491511e-05, "loss": 0.085, "step": 13770 }, { "grad_norm": 0.608809232711792, "learning_rate": 9.143286320910996e-05, "loss": 0.0848, "step": 13780 }, { "grad_norm": 0.40890276432037354, "learning_rate": 9.141743124709491e-05, "loss": 0.0698, "step": 13790 }, { "grad_norm": 0.8297561407089233, "learning_rate": 9.140198670355784e-05, "loss": 0.1015, "step": 13800 }, { "grad_norm": 0.49456894397735596, "learning_rate": 9.138652958319034e-05, "loss": 0.0713, "step": 13810 }, { "grad_norm": 0.5950100421905518, "learning_rate": 9.137105989068791e-05, "loss": 0.0682, "step": 13820 }, { "grad_norm": 1.3823782205581665, "learning_rate": 9.135557763074983e-05, "loss": 0.0901, "step": 13830 }, { "grad_norm": 0.640444815158844, "learning_rate": 9.13400828080792e-05, "loss": 0.097, "step": 13840 }, { "grad_norm": 0.8177790641784668, "learning_rate": 9.132457542738292e-05, "loss": 0.0874, "step": 13850 }, { "grad_norm": 0.962632417678833, "learning_rate": 9.130905549337174e-05, "loss": 0.0964, "step": 13860 }, { "grad_norm": 0.6307457089424133, "learning_rate": 9.129352301076021e-05, "loss": 0.0718, "step": 13870 }, { "grad_norm": 0.7034139633178711, "learning_rate": 9.127797798426668e-05, "loss": 0.0712, "step": 13880 }, { "grad_norm": 0.7015705704689026, "learning_rate": 9.126242041861333e-05, "loss": 0.0824, "step": 13890 }, { "grad_norm": 0.41452279686927795, "learning_rate": 9.124685031852611e-05, "loss": 0.0624, "step": 13900 }, { "grad_norm": 0.7805535197257996, "learning_rate": 9.123126768873482e-05, "loss": 0.0859, "step": 13910 }, { "grad_norm": 0.8315103054046631, "learning_rate": 9.121567253397308e-05, "loss": 0.0982, "step": 13920 }, { "grad_norm": 0.5124301314353943, "learning_rate": 9.120006485897824e-05, "loss": 0.0873, "step": 13930 }, { "grad_norm": 0.5962291359901428, "learning_rate": 9.118444466849152e-05, "loss": 0.0873, "step": 13940 }, { "grad_norm": 0.6490293741226196, "learning_rate": 9.116881196725793e-05, "loss": 0.08, "step": 13950 }, { "grad_norm": 0.6422462463378906, "learning_rate": 9.115316676002627e-05, "loss": 0.1119, "step": 13960 }, { "grad_norm": 0.9056212306022644, "learning_rate": 9.113750905154911e-05, "loss": 0.0746, "step": 13970 }, { "grad_norm": 0.81647127866745, "learning_rate": 9.112183884658289e-05, "loss": 0.0806, "step": 13980 }, { "grad_norm": 0.6363530158996582, "learning_rate": 9.11061561498878e-05, "loss": 0.0823, "step": 13990 }, { "grad_norm": 0.614475429058075, "learning_rate": 9.109046096622779e-05, "loss": 0.0746, "step": 14000 }, { "grad_norm": 0.5450493097305298, "learning_rate": 9.107475330037069e-05, "loss": 0.0792, "step": 14010 }, { "grad_norm": 0.4911443889141083, "learning_rate": 9.105903315708806e-05, "loss": 0.0838, "step": 14020 }, { "grad_norm": 0.679205060005188, "learning_rate": 9.104330054115524e-05, "loss": 0.0761, "step": 14030 }, { "grad_norm": 0.6924609541893005, "learning_rate": 9.102755545735141e-05, "loss": 0.0795, "step": 14040 }, { "grad_norm": 0.48030105233192444, "learning_rate": 9.10117979104595e-05, "loss": 0.0993, "step": 14050 }, { "grad_norm": 0.7834540009498596, "learning_rate": 9.099602790526624e-05, "loss": 0.09, "step": 14060 }, { "grad_norm": 0.9683969616889954, "learning_rate": 9.098024544656212e-05, "loss": 0.0871, "step": 14070 }, { "grad_norm": 0.6448737382888794, "learning_rate": 9.096445053914148e-05, "loss": 0.0995, "step": 14080 }, { "grad_norm": 0.7810444831848145, "learning_rate": 9.094864318780236e-05, "loss": 0.0843, "step": 14090 }, { "grad_norm": 0.6535317301750183, "learning_rate": 9.093282339734663e-05, "loss": 0.0865, "step": 14100 }, { "grad_norm": 0.7554954290390015, "learning_rate": 9.091699117257992e-05, "loss": 0.0721, "step": 14110 }, { "grad_norm": 0.8503060340881348, "learning_rate": 9.090114651831163e-05, "loss": 0.08, "step": 14120 }, { "grad_norm": 0.7234598398208618, "learning_rate": 9.088528943935497e-05, "loss": 0.0788, "step": 14130 }, { "grad_norm": 0.8897728323936462, "learning_rate": 9.086941994052689e-05, "loss": 0.0855, "step": 14140 }, { "grad_norm": 0.6327710151672363, "learning_rate": 9.085353802664813e-05, "loss": 0.1049, "step": 14150 }, { "grad_norm": 0.850387454032898, "learning_rate": 9.08376437025432e-05, "loss": 0.1154, "step": 14160 }, { "grad_norm": 0.732766330242157, "learning_rate": 9.082173697304035e-05, "loss": 0.0957, "step": 14170 }, { "grad_norm": 0.7535841464996338, "learning_rate": 9.080581784297166e-05, "loss": 0.0911, "step": 14180 }, { "grad_norm": 0.6978484988212585, "learning_rate": 9.078988631717291e-05, "loss": 0.0657, "step": 14190 }, { "grad_norm": 0.7204114198684692, "learning_rate": 9.077394240048369e-05, "loss": 0.0715, "step": 14200 }, { "grad_norm": 0.9088853001594543, "learning_rate": 9.075798609774736e-05, "loss": 0.1188, "step": 14210 }, { "grad_norm": 0.8932197093963623, "learning_rate": 9.0742017413811e-05, "loss": 0.0868, "step": 14220 }, { "grad_norm": 0.7942870259284973, "learning_rate": 9.072603635352548e-05, "loss": 0.0715, "step": 14230 }, { "grad_norm": 0.6009909510612488, "learning_rate": 9.071004292174541e-05, "loss": 0.0719, "step": 14240 }, { "grad_norm": 1.1526480913162231, "learning_rate": 9.06940371233292e-05, "loss": 0.0743, "step": 14250 }, { "grad_norm": 0.6539580821990967, "learning_rate": 9.067801896313898e-05, "loss": 0.0735, "step": 14260 }, { "grad_norm": 0.7255939245223999, "learning_rate": 9.066198844604064e-05, "loss": 0.103, "step": 14270 }, { "grad_norm": 0.40224915742874146, "learning_rate": 9.06459455769038e-05, "loss": 0.0992, "step": 14280 }, { "grad_norm": 0.7444795370101929, "learning_rate": 9.062989036060193e-05, "loss": 0.1177, "step": 14290 }, { "grad_norm": 0.6141385436058044, "learning_rate": 9.061382280201212e-05, "loss": 0.09, "step": 14300 }, { "grad_norm": 0.6710569858551025, "learning_rate": 9.059774290601528e-05, "loss": 0.0754, "step": 14310 }, { "grad_norm": 1.2142858505249023, "learning_rate": 9.058165067749606e-05, "loss": 0.0973, "step": 14320 }, { "grad_norm": 0.8379811644554138, "learning_rate": 9.056554612134288e-05, "loss": 0.085, "step": 14330 }, { "grad_norm": 0.7146070599555969, "learning_rate": 9.054942924244785e-05, "loss": 0.0738, "step": 14340 }, { "grad_norm": 0.6498344540596008, "learning_rate": 9.053330004570686e-05, "loss": 0.075, "step": 14350 }, { "grad_norm": 1.0048224925994873, "learning_rate": 9.051715853601955e-05, "loss": 0.0829, "step": 14360 }, { "grad_norm": 0.681780219078064, "learning_rate": 9.050100471828926e-05, "loss": 0.1003, "step": 14370 }, { "grad_norm": 0.6607049703598022, "learning_rate": 9.048483859742311e-05, "loss": 0.0896, "step": 14380 }, { "grad_norm": 0.8337885141372681, "learning_rate": 9.046866017833193e-05, "loss": 0.0789, "step": 14390 }, { "grad_norm": 0.8476659059524536, "learning_rate": 9.045246946593029e-05, "loss": 0.0866, "step": 14400 }, { "grad_norm": 0.6116016507148743, "learning_rate": 9.043626646513652e-05, "loss": 0.0789, "step": 14410 }, { "grad_norm": 0.767636775970459, "learning_rate": 9.042005118087267e-05, "loss": 0.0904, "step": 14420 }, { "grad_norm": 0.6838800311088562, "learning_rate": 9.040382361806448e-05, "loss": 0.084, "step": 14430 }, { "grad_norm": 0.6415536403656006, "learning_rate": 9.038758378164148e-05, "loss": 0.0828, "step": 14440 }, { "grad_norm": 0.7130551934242249, "learning_rate": 9.037133167653691e-05, "loss": 0.0698, "step": 14450 }, { "grad_norm": 0.5113057494163513, "learning_rate": 9.035506730768771e-05, "loss": 0.0818, "step": 14460 }, { "grad_norm": 0.5870665311813354, "learning_rate": 9.033879068003458e-05, "loss": 0.0734, "step": 14470 }, { "grad_norm": 0.9636129140853882, "learning_rate": 9.032250179852193e-05, "loss": 0.0837, "step": 14480 }, { "grad_norm": 0.8131876587867737, "learning_rate": 9.030620066809787e-05, "loss": 0.0929, "step": 14490 }, { "grad_norm": 0.7087332606315613, "learning_rate": 9.028988729371428e-05, "loss": 0.0905, "step": 14500 }, { "grad_norm": 0.7984201312065125, "learning_rate": 9.027356168032673e-05, "loss": 0.0815, "step": 14510 }, { "grad_norm": 0.7489101886749268, "learning_rate": 9.02572238328945e-05, "loss": 0.0701, "step": 14520 }, { "grad_norm": 0.7411248087882996, "learning_rate": 9.02408737563806e-05, "loss": 0.0765, "step": 14530 }, { "grad_norm": 0.6409339904785156, "learning_rate": 9.022451145575174e-05, "loss": 0.0895, "step": 14540 }, { "grad_norm": 0.70404452085495, "learning_rate": 9.02081369359784e-05, "loss": 0.0787, "step": 14550 }, { "grad_norm": 0.6011647582054138, "learning_rate": 9.019175020203465e-05, "loss": 0.0967, "step": 14560 }, { "grad_norm": 0.5208867788314819, "learning_rate": 9.017535125889842e-05, "loss": 0.072, "step": 14570 }, { "grad_norm": 0.588434100151062, "learning_rate": 9.015894011155124e-05, "loss": 0.0931, "step": 14580 }, { "grad_norm": 0.6117406487464905, "learning_rate": 9.014251676497838e-05, "loss": 0.0635, "step": 14590 }, { "grad_norm": 0.5336824059486389, "learning_rate": 9.012608122416884e-05, "loss": 0.0751, "step": 14600 }, { "grad_norm": 0.6029449105262756, "learning_rate": 9.010963349411529e-05, "loss": 0.0869, "step": 14610 }, { "grad_norm": 0.9228412508964539, "learning_rate": 9.00931735798141e-05, "loss": 0.0772, "step": 14620 }, { "grad_norm": 0.6989905834197998, "learning_rate": 9.00767014862654e-05, "loss": 0.0841, "step": 14630 }, { "grad_norm": 0.590175449848175, "learning_rate": 9.006021721847295e-05, "loss": 0.0731, "step": 14640 }, { "grad_norm": 0.5279282331466675, "learning_rate": 9.004372078144423e-05, "loss": 0.0745, "step": 14650 }, { "grad_norm": 0.7974891662597656, "learning_rate": 9.002721218019043e-05, "loss": 0.0753, "step": 14660 }, { "grad_norm": 0.6597141623497009, "learning_rate": 9.001069141972642e-05, "loss": 0.0704, "step": 14670 }, { "grad_norm": 0.7246265411376953, "learning_rate": 8.99941585050708e-05, "loss": 0.0798, "step": 14680 }, { "grad_norm": 0.5742676258087158, "learning_rate": 8.997761344124578e-05, "loss": 0.0747, "step": 14690 }, { "grad_norm": 0.6701658368110657, "learning_rate": 8.996105623327737e-05, "loss": 0.06, "step": 14700 }, { "grad_norm": 0.5434255599975586, "learning_rate": 8.994448688619517e-05, "loss": 0.0757, "step": 14710 }, { "grad_norm": 0.6456254124641418, "learning_rate": 8.992790540503253e-05, "loss": 0.0708, "step": 14720 }, { "grad_norm": 0.5873963832855225, "learning_rate": 8.991131179482648e-05, "loss": 0.0675, "step": 14730 }, { "grad_norm": 0.7048465013504028, "learning_rate": 8.989470606061768e-05, "loss": 0.0741, "step": 14740 }, { "grad_norm": 0.5942285060882568, "learning_rate": 8.987808820745056e-05, "loss": 0.0773, "step": 14750 }, { "grad_norm": 0.7103257775306702, "learning_rate": 8.986145824037315e-05, "loss": 0.0771, "step": 14760 }, { "grad_norm": 0.6235703825950623, "learning_rate": 8.984481616443721e-05, "loss": 0.0814, "step": 14770 }, { "grad_norm": 0.896221399307251, "learning_rate": 8.982816198469815e-05, "loss": 0.0801, "step": 14780 }, { "grad_norm": 0.6758143305778503, "learning_rate": 8.98114957062151e-05, "loss": 0.0887, "step": 14790 }, { "grad_norm": 0.5887151956558228, "learning_rate": 8.97948173340508e-05, "loss": 0.0829, "step": 14800 }, { "grad_norm": 0.6915661692619324, "learning_rate": 8.977812687327172e-05, "loss": 0.0778, "step": 14810 }, { "grad_norm": 0.615849494934082, "learning_rate": 8.976142432894798e-05, "loss": 0.0866, "step": 14820 }, { "grad_norm": 0.5834874510765076, "learning_rate": 8.974470970615336e-05, "loss": 0.0751, "step": 14830 }, { "grad_norm": 0.9119956493377686, "learning_rate": 8.972798300996534e-05, "loss": 0.1101, "step": 14840 }, { "grad_norm": 0.8748375773429871, "learning_rate": 8.971124424546504e-05, "loss": 0.1001, "step": 14850 }, { "grad_norm": 0.5121074914932251, "learning_rate": 8.969449341773724e-05, "loss": 0.0927, "step": 14860 }, { "grad_norm": 0.8171043992042542, "learning_rate": 8.967773053187042e-05, "loss": 0.0987, "step": 14870 }, { "grad_norm": 0.518514096736908, "learning_rate": 8.966095559295668e-05, "loss": 0.0874, "step": 14880 }, { "grad_norm": 0.7648111581802368, "learning_rate": 8.964416860609184e-05, "loss": 0.091, "step": 14890 }, { "grad_norm": 0.7435362339019775, "learning_rate": 8.962736957637532e-05, "loss": 0.0739, "step": 14900 }, { "grad_norm": 0.8371986150741577, "learning_rate": 8.96105585089102e-05, "loss": 0.0863, "step": 14910 }, { "grad_norm": 1.5179905891418457, "learning_rate": 8.959373540880329e-05, "loss": 0.0915, "step": 14920 }, { "grad_norm": 0.6685963869094849, "learning_rate": 8.957690028116495e-05, "loss": 0.0906, "step": 14930 }, { "grad_norm": 0.7360060214996338, "learning_rate": 8.956005313110928e-05, "loss": 0.0892, "step": 14940 }, { "grad_norm": 0.5546371340751648, "learning_rate": 8.9543193963754e-05, "loss": 0.0939, "step": 14950 }, { "grad_norm": 0.8017341494560242, "learning_rate": 8.952632278422048e-05, "loss": 0.0832, "step": 14960 }, { "grad_norm": 0.5199772715568542, "learning_rate": 8.95094395976337e-05, "loss": 0.085, "step": 14970 }, { "grad_norm": 0.613749086856842, "learning_rate": 8.949254440912239e-05, "loss": 0.0787, "step": 14980 }, { "grad_norm": 0.6388752460479736, "learning_rate": 8.94756372238188e-05, "loss": 0.0881, "step": 14990 }, { "grad_norm": 0.7779206037521362, "learning_rate": 8.945871804685892e-05, "loss": 0.0882, "step": 15000 }, { "grad_norm": 0.4765072762966156, "learning_rate": 8.944178688338236e-05, "loss": 0.0772, "step": 15010 }, { "grad_norm": 0.5790134072303772, "learning_rate": 8.942484373853233e-05, "loss": 0.089, "step": 15020 }, { "grad_norm": 0.7389034032821655, "learning_rate": 8.940788861745572e-05, "loss": 0.0761, "step": 15030 }, { "grad_norm": 0.7463007569313049, "learning_rate": 8.939092152530308e-05, "loss": 0.0724, "step": 15040 }, { "grad_norm": 0.4929518699645996, "learning_rate": 8.937394246722853e-05, "loss": 0.0687, "step": 15050 }, { "grad_norm": 0.6583940982818604, "learning_rate": 8.935695144838984e-05, "loss": 0.0767, "step": 15060 }, { "grad_norm": 0.5424224734306335, "learning_rate": 8.933994847394849e-05, "loss": 0.0656, "step": 15070 }, { "grad_norm": 0.5765616297721863, "learning_rate": 8.932293354906949e-05, "loss": 0.074, "step": 15080 }, { "grad_norm": 0.4681607186794281, "learning_rate": 8.930590667892153e-05, "loss": 0.076, "step": 15090 }, { "grad_norm": 0.7484533786773682, "learning_rate": 8.928886786867696e-05, "loss": 0.0857, "step": 15100 }, { "grad_norm": 0.5902478694915771, "learning_rate": 8.927181712351168e-05, "loss": 0.0691, "step": 15110 }, { "grad_norm": 0.38611888885498047, "learning_rate": 8.925475444860527e-05, "loss": 0.0813, "step": 15120 }, { "grad_norm": 0.5640222430229187, "learning_rate": 8.923767984914092e-05, "loss": 0.0875, "step": 15130 }, { "grad_norm": 0.5513148307800293, "learning_rate": 8.922059333030545e-05, "loss": 0.089, "step": 15140 }, { "grad_norm": 0.5088780522346497, "learning_rate": 8.920349489728928e-05, "loss": 0.0606, "step": 15150 }, { "grad_norm": 0.6808351874351501, "learning_rate": 8.918638455528646e-05, "loss": 0.1047, "step": 15160 }, { "grad_norm": 0.5443179607391357, "learning_rate": 8.916926230949468e-05, "loss": 0.0941, "step": 15170 }, { "grad_norm": 0.6029876470565796, "learning_rate": 8.915212816511522e-05, "loss": 0.0706, "step": 15180 }, { "grad_norm": 0.555137574672699, "learning_rate": 8.913498212735296e-05, "loss": 0.0995, "step": 15190 }, { "grad_norm": 0.5320030450820923, "learning_rate": 8.911782420141643e-05, "loss": 0.0741, "step": 15200 }, { "grad_norm": 0.814183235168457, "learning_rate": 8.910065439251775e-05, "loss": 0.0965, "step": 15210 }, { "grad_norm": 0.6163471937179565, "learning_rate": 8.908347270587268e-05, "loss": 0.0687, "step": 15220 }, { "grad_norm": 0.7425543665885925, "learning_rate": 8.906627914670054e-05, "loss": 0.093, "step": 15230 }, { "grad_norm": 0.592629075050354, "learning_rate": 8.904907372022427e-05, "loss": 0.079, "step": 15240 }, { "grad_norm": 0.651028037071228, "learning_rate": 8.903185643167042e-05, "loss": 0.086, "step": 15250 }, { "grad_norm": 0.6396533846855164, "learning_rate": 8.901462728626919e-05, "loss": 0.073, "step": 15260 }, { "grad_norm": 0.7235170006752014, "learning_rate": 8.899738628925429e-05, "loss": 0.0677, "step": 15270 }, { "grad_norm": 0.445266455411911, "learning_rate": 8.898013344586312e-05, "loss": 0.0796, "step": 15280 }, { "grad_norm": 0.733781635761261, "learning_rate": 8.896286876133661e-05, "loss": 0.0696, "step": 15290 }, { "grad_norm": 0.45712530612945557, "learning_rate": 8.894559224091933e-05, "loss": 0.0856, "step": 15300 }, { "grad_norm": 0.6579089760780334, "learning_rate": 8.892830388985942e-05, "loss": 0.0733, "step": 15310 }, { "grad_norm": 0.8214157819747925, "learning_rate": 8.891100371340864e-05, "loss": 0.0774, "step": 15320 }, { "grad_norm": 0.6052621006965637, "learning_rate": 8.889369171682231e-05, "loss": 0.0907, "step": 15330 }, { "grad_norm": 0.6585767865180969, "learning_rate": 8.887636790535936e-05, "loss": 0.0784, "step": 15340 }, { "grad_norm": 0.5159403085708618, "learning_rate": 8.885903228428231e-05, "loss": 0.0768, "step": 15350 }, { "grad_norm": 0.7467994093894958, "learning_rate": 8.884168485885727e-05, "loss": 0.0832, "step": 15360 }, { "grad_norm": 0.7513036727905273, "learning_rate": 8.882432563435393e-05, "loss": 0.0851, "step": 15370 }, { "grad_norm": 0.4361516535282135, "learning_rate": 8.880695461604556e-05, "loss": 0.0803, "step": 15380 }, { "grad_norm": 0.9872533679008484, "learning_rate": 8.878957180920901e-05, "loss": 0.0737, "step": 15390 }, { "grad_norm": 0.6491416692733765, "learning_rate": 8.877217721912473e-05, "loss": 0.0639, "step": 15400 }, { "grad_norm": 0.6731883883476257, "learning_rate": 8.875477085107673e-05, "loss": 0.0697, "step": 15410 }, { "grad_norm": 0.5851972699165344, "learning_rate": 8.87373527103526e-05, "loss": 0.1054, "step": 15420 }, { "grad_norm": 0.7783461213111877, "learning_rate": 8.871992280224353e-05, "loss": 0.0747, "step": 15430 }, { "grad_norm": 0.558448851108551, "learning_rate": 8.870248113204422e-05, "loss": 0.085, "step": 15440 }, { "grad_norm": 0.8717674612998962, "learning_rate": 8.868502770505306e-05, "loss": 0.0931, "step": 15450 }, { "grad_norm": 0.8820386528968811, "learning_rate": 8.86675625265719e-05, "loss": 0.0891, "step": 15460 }, { "grad_norm": 0.5633804798126221, "learning_rate": 8.865008560190618e-05, "loss": 0.0852, "step": 15470 }, { "grad_norm": 0.7777029275894165, "learning_rate": 8.863259693636496e-05, "loss": 0.0831, "step": 15480 }, { "grad_norm": 0.7264232635498047, "learning_rate": 8.861509653526083e-05, "loss": 0.0741, "step": 15490 }, { "grad_norm": 0.6767601370811462, "learning_rate": 8.859758440390993e-05, "loss": 0.0811, "step": 15500 }, { "grad_norm": 0.7023001313209534, "learning_rate": 8.858006054763202e-05, "loss": 0.0816, "step": 15510 }, { "grad_norm": 0.7060182690620422, "learning_rate": 8.856252497175035e-05, "loss": 0.0855, "step": 15520 }, { "grad_norm": 0.8245953321456909, "learning_rate": 8.854497768159178e-05, "loss": 0.0913, "step": 15530 }, { "grad_norm": 0.6905299425125122, "learning_rate": 8.852741868248671e-05, "loss": 0.0787, "step": 15540 }, { "grad_norm": 0.7701488733291626, "learning_rate": 8.85098479797691e-05, "loss": 0.0845, "step": 15550 }, { "grad_norm": 0.9803156852722168, "learning_rate": 8.849226557877646e-05, "loss": 0.0909, "step": 15560 }, { "grad_norm": 0.669246256351471, "learning_rate": 8.84746714848499e-05, "loss": 0.0928, "step": 15570 }, { "grad_norm": 0.7853661775588989, "learning_rate": 8.845706570333397e-05, "loss": 0.0785, "step": 15580 }, { "grad_norm": 0.666107714176178, "learning_rate": 8.84394482395769e-05, "loss": 0.0645, "step": 15590 }, { "grad_norm": 0.5902101993560791, "learning_rate": 8.842181909893038e-05, "loss": 0.0782, "step": 15600 }, { "grad_norm": 0.6087527871131897, "learning_rate": 8.840417828674969e-05, "loss": 0.0859, "step": 15610 }, { "grad_norm": 0.6436321139335632, "learning_rate": 8.838652580839364e-05, "loss": 0.0729, "step": 15620 }, { "grad_norm": 0.8665592670440674, "learning_rate": 8.836886166922458e-05, "loss": 0.0796, "step": 15630 }, { "grad_norm": 1.0131323337554932, "learning_rate": 8.835118587460844e-05, "loss": 0.0729, "step": 15640 }, { "grad_norm": 0.609743595123291, "learning_rate": 8.83334984299146e-05, "loss": 0.0702, "step": 15650 }, { "grad_norm": 0.7393342852592468, "learning_rate": 8.83157993405161e-05, "loss": 0.0687, "step": 15660 }, { "grad_norm": 0.4499025344848633, "learning_rate": 8.829808861178943e-05, "loss": 0.064, "step": 15670 }, { "grad_norm": 0.6272403001785278, "learning_rate": 8.828036624911464e-05, "loss": 0.0641, "step": 15680 }, { "grad_norm": 0.716651439666748, "learning_rate": 8.826263225787532e-05, "loss": 0.0771, "step": 15690 }, { "grad_norm": 0.5882294178009033, "learning_rate": 8.824488664345858e-05, "loss": 0.0727, "step": 15700 }, { "grad_norm": 0.47754257917404175, "learning_rate": 8.822712941125508e-05, "loss": 0.0866, "step": 15710 }, { "grad_norm": 0.601530909538269, "learning_rate": 8.820936056665898e-05, "loss": 0.0813, "step": 15720 }, { "grad_norm": 0.6453710794448853, "learning_rate": 8.819158011506801e-05, "loss": 0.0837, "step": 15730 }, { "grad_norm": 0.612797737121582, "learning_rate": 8.81737880618834e-05, "loss": 0.0956, "step": 15740 }, { "grad_norm": 0.44861873984336853, "learning_rate": 8.815598441250987e-05, "loss": 0.0806, "step": 15750 }, { "grad_norm": 0.7971729040145874, "learning_rate": 8.813816917235576e-05, "loss": 0.0685, "step": 15760 }, { "grad_norm": 0.6147456765174866, "learning_rate": 8.812034234683282e-05, "loss": 0.0754, "step": 15770 }, { "grad_norm": 1.0782697200775146, "learning_rate": 8.810250394135637e-05, "loss": 0.0721, "step": 15780 }, { "grad_norm": 0.44394397735595703, "learning_rate": 8.808465396134529e-05, "loss": 0.0943, "step": 15790 }, { "grad_norm": 0.6933355331420898, "learning_rate": 8.806679241222189e-05, "loss": 0.082, "step": 15800 }, { "grad_norm": 0.8242666721343994, "learning_rate": 8.804891929941203e-05, "loss": 0.0969, "step": 15810 }, { "grad_norm": 0.6833471655845642, "learning_rate": 8.803103462834514e-05, "loss": 0.0766, "step": 15820 }, { "grad_norm": 0.8619239926338196, "learning_rate": 8.801313840445408e-05, "loss": 0.1097, "step": 15830 }, { "grad_norm": 0.9687374830245972, "learning_rate": 8.799523063317524e-05, "loss": 0.1086, "step": 15840 }, { "grad_norm": 0.7711012363433838, "learning_rate": 8.797731131994854e-05, "loss": 0.0804, "step": 15850 }, { "grad_norm": 0.6858108043670654, "learning_rate": 8.795938047021739e-05, "loss": 0.0774, "step": 15860 }, { "grad_norm": 0.6531695127487183, "learning_rate": 8.794143808942872e-05, "loss": 0.0712, "step": 15870 }, { "grad_norm": 0.5386190414428711, "learning_rate": 8.792348418303296e-05, "loss": 0.0924, "step": 15880 }, { "grad_norm": 1.6214221715927124, "learning_rate": 8.790551875648398e-05, "loss": 0.0851, "step": 15890 }, { "grad_norm": 0.8181710243225098, "learning_rate": 8.788754181523926e-05, "loss": 0.0771, "step": 15900 }, { "grad_norm": 1.1955864429473877, "learning_rate": 8.78695533647597e-05, "loss": 0.0693, "step": 15910 }, { "grad_norm": 0.5966773629188538, "learning_rate": 8.785155341050972e-05, "loss": 0.0745, "step": 15920 }, { "grad_norm": 0.6820032000541687, "learning_rate": 8.783354195795721e-05, "loss": 0.0749, "step": 15930 }, { "grad_norm": 0.7229114174842834, "learning_rate": 8.78155190125736e-05, "loss": 0.0747, "step": 15940 }, { "grad_norm": 0.5228366255760193, "learning_rate": 8.779748457983378e-05, "loss": 0.0773, "step": 15950 }, { "grad_norm": 0.818385660648346, "learning_rate": 8.777943866521612e-05, "loss": 0.0749, "step": 15960 }, { "grad_norm": 0.7685229778289795, "learning_rate": 8.77613812742025e-05, "loss": 0.0756, "step": 15970 }, { "grad_norm": 0.4971092641353607, "learning_rate": 8.774331241227829e-05, "loss": 0.0804, "step": 15980 }, { "grad_norm": 0.7898941040039062, "learning_rate": 8.772523208493232e-05, "loss": 0.0926, "step": 15990 }, { "grad_norm": 0.6726735830307007, "learning_rate": 8.770714029765692e-05, "loss": 0.0727, "step": 16000 }, { "grad_norm": 0.5289822816848755, "learning_rate": 8.768903705594789e-05, "loss": 0.0806, "step": 16010 }, { "grad_norm": 0.7347676753997803, "learning_rate": 8.767092236530453e-05, "loss": 0.0734, "step": 16020 }, { "grad_norm": 0.7808020114898682, "learning_rate": 8.76527962312296e-05, "loss": 0.0807, "step": 16030 }, { "grad_norm": 0.675324022769928, "learning_rate": 8.763465865922934e-05, "loss": 0.069, "step": 16040 }, { "grad_norm": 0.881462812423706, "learning_rate": 8.761650965481347e-05, "loss": 0.0679, "step": 16050 }, { "grad_norm": 0.4231560230255127, "learning_rate": 8.759834922349516e-05, "loss": 0.0631, "step": 16060 }, { "grad_norm": 0.4521190822124481, "learning_rate": 8.758017737079108e-05, "loss": 0.0727, "step": 16070 }, { "grad_norm": 0.6768683195114136, "learning_rate": 8.756199410222137e-05, "loss": 0.0663, "step": 16080 }, { "grad_norm": 0.5408718585968018, "learning_rate": 8.754379942330963e-05, "loss": 0.0804, "step": 16090 }, { "grad_norm": 0.554100751876831, "learning_rate": 8.75255933395829e-05, "loss": 0.0816, "step": 16100 }, { "grad_norm": 0.6292282342910767, "learning_rate": 8.750737585657171e-05, "loss": 0.0714, "step": 16110 }, { "grad_norm": 0.4018365442752838, "learning_rate": 8.748914697981008e-05, "loss": 0.0618, "step": 16120 }, { "grad_norm": 0.4635210931301117, "learning_rate": 8.747090671483542e-05, "loss": 0.0806, "step": 16130 }, { "grad_norm": 0.5481959581375122, "learning_rate": 8.745265506718869e-05, "loss": 0.0629, "step": 16140 }, { "grad_norm": 0.6921036243438721, "learning_rate": 8.74343920424142e-05, "loss": 0.0784, "step": 16150 }, { "grad_norm": 0.5743776559829712, "learning_rate": 8.741611764605982e-05, "loss": 0.0955, "step": 16160 }, { "grad_norm": 0.42062151432037354, "learning_rate": 8.739783188367682e-05, "loss": 0.0807, "step": 16170 }, { "grad_norm": 0.8702240586280823, "learning_rate": 8.737953476081991e-05, "loss": 0.0803, "step": 16180 }, { "grad_norm": 0.9090121388435364, "learning_rate": 8.73612262830473e-05, "loss": 0.1033, "step": 16190 }, { "grad_norm": 0.48359668254852295, "learning_rate": 8.734290645592061e-05, "loss": 0.0675, "step": 16200 }, { "grad_norm": 0.5271120667457581, "learning_rate": 8.732457528500493e-05, "loss": 0.0975, "step": 16210 }, { "grad_norm": 0.5132089257240295, "learning_rate": 8.730623277586875e-05, "loss": 0.0695, "step": 16220 }, { "grad_norm": 0.5291849374771118, "learning_rate": 8.72878789340841e-05, "loss": 0.0786, "step": 16230 }, { "grad_norm": 0.539274275302887, "learning_rate": 8.726951376522635e-05, "loss": 0.0771, "step": 16240 }, { "grad_norm": 0.5870311260223389, "learning_rate": 8.725113727487435e-05, "loss": 0.0681, "step": 16250 }, { "grad_norm": 0.5986999273300171, "learning_rate": 8.723274946861042e-05, "loss": 0.1023, "step": 16260 }, { "grad_norm": 0.6201570630073547, "learning_rate": 8.721435035202026e-05, "loss": 0.0909, "step": 16270 }, { "grad_norm": 0.6415302157402039, "learning_rate": 8.719593993069306e-05, "loss": 0.0682, "step": 16280 }, { "grad_norm": 0.5721144080162048, "learning_rate": 8.717751821022139e-05, "loss": 0.0661, "step": 16290 }, { "grad_norm": 0.6462592482566833, "learning_rate": 8.715908519620134e-05, "loss": 0.0752, "step": 16300 }, { "grad_norm": 0.515206515789032, "learning_rate": 8.71406408942323e-05, "loss": 0.095, "step": 16310 }, { "grad_norm": 0.6133238673210144, "learning_rate": 8.712218530991723e-05, "loss": 0.0669, "step": 16320 }, { "grad_norm": 0.5345750451087952, "learning_rate": 8.710371844886241e-05, "loss": 0.0722, "step": 16330 }, { "grad_norm": 0.8842787146568298, "learning_rate": 8.708524031667758e-05, "loss": 0.0747, "step": 16340 }, { "grad_norm": 0.5640265345573425, "learning_rate": 8.706675091897592e-05, "loss": 0.0631, "step": 16350 }, { "grad_norm": 0.4899756610393524, "learning_rate": 8.704825026137404e-05, "loss": 0.067, "step": 16360 }, { "grad_norm": 0.481983482837677, "learning_rate": 8.702973834949192e-05, "loss": 0.0563, "step": 16370 }, { "grad_norm": 0.4718632698059082, "learning_rate": 8.701121518895301e-05, "loss": 0.0866, "step": 16380 }, { "grad_norm": 0.5543570518493652, "learning_rate": 8.699268078538414e-05, "loss": 0.0748, "step": 16390 }, { "grad_norm": 0.6297997832298279, "learning_rate": 8.69741351444156e-05, "loss": 0.0713, "step": 16400 }, { "grad_norm": 0.7032830715179443, "learning_rate": 8.695557827168101e-05, "loss": 0.0522, "step": 16410 }, { "grad_norm": 0.6138125658035278, "learning_rate": 8.693701017281753e-05, "loss": 0.0614, "step": 16420 }, { "grad_norm": 0.7066869139671326, "learning_rate": 8.691843085346563e-05, "loss": 0.0822, "step": 16430 }, { "grad_norm": 0.5254324078559875, "learning_rate": 8.689984031926919e-05, "loss": 0.0687, "step": 16440 }, { "grad_norm": 0.5761610865592957, "learning_rate": 8.688123857587555e-05, "loss": 0.0862, "step": 16450 }, { "grad_norm": 0.8494483828544617, "learning_rate": 8.686262562893544e-05, "loss": 0.0937, "step": 16460 }, { "grad_norm": 0.8932580947875977, "learning_rate": 8.684400148410294e-05, "loss": 0.1006, "step": 16470 }, { "grad_norm": 0.5922924280166626, "learning_rate": 8.682536614703562e-05, "loss": 0.073, "step": 16480 }, { "grad_norm": 0.8841191530227661, "learning_rate": 8.680671962339437e-05, "loss": 0.0777, "step": 16490 }, { "grad_norm": 0.6917209029197693, "learning_rate": 8.678806191884352e-05, "loss": 0.0736, "step": 16500 }, { "grad_norm": 0.5594883561134338, "learning_rate": 8.67693930390508e-05, "loss": 0.0561, "step": 16510 }, { "grad_norm": 0.4432305097579956, "learning_rate": 8.67507129896873e-05, "loss": 0.0748, "step": 16520 }, { "grad_norm": 0.5119044780731201, "learning_rate": 8.673202177642757e-05, "loss": 0.0807, "step": 16530 }, { "grad_norm": 0.9201231598854065, "learning_rate": 8.671331940494945e-05, "loss": 0.0664, "step": 16540 }, { "grad_norm": 0.595582127571106, "learning_rate": 8.669460588093427e-05, "loss": 0.081, "step": 16550 }, { "grad_norm": 1.2145352363586426, "learning_rate": 8.667588121006667e-05, "loss": 0.0891, "step": 16560 }, { "grad_norm": 0.5108336806297302, "learning_rate": 8.665714539803475e-05, "loss": 0.0626, "step": 16570 }, { "grad_norm": 0.6679067015647888, "learning_rate": 8.663839845052993e-05, "loss": 0.0705, "step": 16580 }, { "grad_norm": 0.39172354340553284, "learning_rate": 8.661964037324703e-05, "loss": 0.0701, "step": 16590 }, { "grad_norm": 0.4909581243991852, "learning_rate": 8.660087117188427e-05, "loss": 0.0739, "step": 16600 }, { "grad_norm": 0.6120092272758484, "learning_rate": 8.658209085214325e-05, "loss": 0.0621, "step": 16610 }, { "grad_norm": 0.8182880878448486, "learning_rate": 8.656329941972891e-05, "loss": 0.07, "step": 16620 }, { "grad_norm": 0.576850414276123, "learning_rate": 8.654449688034963e-05, "loss": 0.0631, "step": 16630 }, { "grad_norm": 0.5903739333152771, "learning_rate": 8.652568323971706e-05, "loss": 0.0549, "step": 16640 }, { "grad_norm": 0.6267764568328857, "learning_rate": 8.650685850354636e-05, "loss": 0.0613, "step": 16650 }, { "grad_norm": 0.4555777907371521, "learning_rate": 8.648802267755593e-05, "loss": 0.0576, "step": 16660 }, { "grad_norm": 0.559704601764679, "learning_rate": 8.646917576746764e-05, "loss": 0.0473, "step": 16670 }, { "grad_norm": 0.677584707736969, "learning_rate": 8.645031777900666e-05, "loss": 0.0627, "step": 16680 }, { "grad_norm": 0.7032754421234131, "learning_rate": 8.643144871790154e-05, "loss": 0.0822, "step": 16690 }, { "grad_norm": 0.7149921655654907, "learning_rate": 8.641256858988424e-05, "loss": 0.0673, "step": 16700 }, { "grad_norm": 0.4891035556793213, "learning_rate": 8.639367740069e-05, "loss": 0.0642, "step": 16710 }, { "grad_norm": 1.0654314756393433, "learning_rate": 8.63747751560575e-05, "loss": 0.0731, "step": 16720 }, { "grad_norm": 0.5154175758361816, "learning_rate": 8.635586186172871e-05, "loss": 0.0542, "step": 16730 }, { "grad_norm": 0.6592024564743042, "learning_rate": 8.633693752344902e-05, "loss": 0.0697, "step": 16740 }, { "grad_norm": 0.5057939291000366, "learning_rate": 8.631800214696713e-05, "loss": 0.0768, "step": 16750 }, { "grad_norm": 0.7284467220306396, "learning_rate": 8.629905573803511e-05, "loss": 0.0624, "step": 16760 }, { "grad_norm": 0.9201344847679138, "learning_rate": 8.628009830240839e-05, "loss": 0.0679, "step": 16770 }, { "grad_norm": 0.6193267107009888, "learning_rate": 8.626112984584571e-05, "loss": 0.0937, "step": 16780 }, { "grad_norm": 0.4901164770126343, "learning_rate": 8.62421503741092e-05, "loss": 0.058, "step": 16790 }, { "grad_norm": 0.9918043613433838, "learning_rate": 8.622315989296432e-05, "loss": 0.0764, "step": 16800 }, { "grad_norm": 0.7140311598777771, "learning_rate": 8.62041584081799e-05, "loss": 0.0725, "step": 16810 }, { "grad_norm": 0.573899507522583, "learning_rate": 8.618514592552807e-05, "loss": 0.068, "step": 16820 }, { "grad_norm": 0.5240370631217957, "learning_rate": 8.616612245078431e-05, "loss": 0.0564, "step": 16830 }, { "grad_norm": 0.5910142064094543, "learning_rate": 8.614708798972746e-05, "loss": 0.0529, "step": 16840 }, { "grad_norm": 0.7755529880523682, "learning_rate": 8.61280425481397e-05, "loss": 0.0553, "step": 16850 }, { "grad_norm": 0.6251140236854553, "learning_rate": 8.61089861318065e-05, "loss": 0.0647, "step": 16860 }, { "grad_norm": 0.6221529245376587, "learning_rate": 8.608991874651673e-05, "loss": 0.0742, "step": 16870 }, { "grad_norm": 0.8759548664093018, "learning_rate": 8.607084039806255e-05, "loss": 0.0648, "step": 16880 }, { "grad_norm": 0.7200204730033875, "learning_rate": 8.605175109223944e-05, "loss": 0.0688, "step": 16890 }, { "grad_norm": 0.581908106803894, "learning_rate": 8.603265083484624e-05, "loss": 0.057, "step": 16900 }, { "grad_norm": 0.7926521897315979, "learning_rate": 8.60135396316851e-05, "loss": 0.0696, "step": 16910 }, { "grad_norm": 0.4720735251903534, "learning_rate": 8.599441748856152e-05, "loss": 0.0549, "step": 16920 }, { "grad_norm": 0.6586021780967712, "learning_rate": 8.597528441128427e-05, "loss": 0.0717, "step": 16930 }, { "grad_norm": 0.5167269110679626, "learning_rate": 8.595614040566549e-05, "loss": 0.0659, "step": 16940 }, { "grad_norm": 0.7019161581993103, "learning_rate": 8.593698547752063e-05, "loss": 0.0588, "step": 16950 }, { "grad_norm": 0.46432796120643616, "learning_rate": 8.591781963266843e-05, "loss": 0.0642, "step": 16960 }, { "grad_norm": 0.6976434588432312, "learning_rate": 8.5898642876931e-05, "loss": 0.0723, "step": 16970 }, { "grad_norm": 0.8828538060188293, "learning_rate": 8.587945521613369e-05, "loss": 0.0668, "step": 16980 }, { "grad_norm": 0.7667357325553894, "learning_rate": 8.586025665610524e-05, "loss": 0.0812, "step": 16990 }, { "grad_norm": 0.804612398147583, "learning_rate": 8.584104720267765e-05, "loss": 0.0854, "step": 17000 }, { "grad_norm": 0.9079546928405762, "learning_rate": 8.582182686168625e-05, "loss": 0.0669, "step": 17010 }, { "grad_norm": 0.7596332430839539, "learning_rate": 8.580259563896967e-05, "loss": 0.0654, "step": 17020 }, { "grad_norm": 0.5833030939102173, "learning_rate": 8.578335354036983e-05, "loss": 0.065, "step": 17030 }, { "grad_norm": 0.8439614772796631, "learning_rate": 8.576410057173201e-05, "loss": 0.0528, "step": 17040 }, { "grad_norm": 0.46203508973121643, "learning_rate": 8.574483673890474e-05, "loss": 0.0662, "step": 17050 }, { "grad_norm": 0.5786876678466797, "learning_rate": 8.572556204773983e-05, "loss": 0.0753, "step": 17060 }, { "grad_norm": 0.5380387306213379, "learning_rate": 8.570627650409246e-05, "loss": 0.059, "step": 17070 }, { "grad_norm": 0.5256463289260864, "learning_rate": 8.568698011382107e-05, "loss": 0.0813, "step": 17080 }, { "grad_norm": 0.7211951613426208, "learning_rate": 8.566767288278738e-05, "loss": 0.0716, "step": 17090 }, { "grad_norm": 0.5966086387634277, "learning_rate": 8.56483548168564e-05, "loss": 0.0735, "step": 17100 }, { "grad_norm": 0.6849271655082703, "learning_rate": 8.562902592189648e-05, "loss": 0.0636, "step": 17110 }, { "grad_norm": 0.5233376622200012, "learning_rate": 8.560968620377921e-05, "loss": 0.0781, "step": 17120 }, { "grad_norm": 0.8318071961402893, "learning_rate": 8.559033566837951e-05, "loss": 0.1196, "step": 17130 }, { "grad_norm": 0.6128283739089966, "learning_rate": 8.557097432157551e-05, "loss": 0.0775, "step": 17140 }, { "grad_norm": 0.5240698456764221, "learning_rate": 8.555160216924872e-05, "loss": 0.0691, "step": 17150 }, { "grad_norm": 0.6884455680847168, "learning_rate": 8.55322192172839e-05, "loss": 0.0775, "step": 17160 }, { "grad_norm": 0.7922484874725342, "learning_rate": 8.551282547156902e-05, "loss": 0.0875, "step": 17170 }, { "grad_norm": 0.5729776620864868, "learning_rate": 8.549342093799544e-05, "loss": 0.0663, "step": 17180 }, { "grad_norm": 0.4551405608654022, "learning_rate": 8.547400562245773e-05, "loss": 0.0532, "step": 17190 }, { "grad_norm": 0.6674889922142029, "learning_rate": 8.545457953085374e-05, "loss": 0.0617, "step": 17200 }, { "grad_norm": 0.43570271134376526, "learning_rate": 8.543514266908463e-05, "loss": 0.0501, "step": 17210 }, { "grad_norm": 0.8082042336463928, "learning_rate": 8.541569504305478e-05, "loss": 0.0668, "step": 17220 }, { "grad_norm": 0.554106593132019, "learning_rate": 8.539623665867187e-05, "loss": 0.0773, "step": 17230 }, { "grad_norm": 0.8417918682098389, "learning_rate": 8.537676752184685e-05, "loss": 0.0826, "step": 17240 }, { "grad_norm": 0.5863898992538452, "learning_rate": 8.53572876384939e-05, "loss": 0.049, "step": 17250 }, { "grad_norm": 0.5864114165306091, "learning_rate": 8.533779701453056e-05, "loss": 0.0672, "step": 17260 }, { "grad_norm": 0.5393058061599731, "learning_rate": 8.53182956558775e-05, "loss": 0.0681, "step": 17270 }, { "grad_norm": 0.9175775051116943, "learning_rate": 8.529878356845877e-05, "loss": 0.0807, "step": 17280 }, { "grad_norm": 0.42230066657066345, "learning_rate": 8.527926075820158e-05, "loss": 0.0729, "step": 17290 }, { "grad_norm": 0.7051886320114136, "learning_rate": 8.525972723103648e-05, "loss": 0.071, "step": 17300 }, { "grad_norm": 0.4933657646179199, "learning_rate": 8.524018299289722e-05, "loss": 0.0681, "step": 17310 }, { "grad_norm": 0.4506986141204834, "learning_rate": 8.522062804972083e-05, "loss": 0.0654, "step": 17320 }, { "grad_norm": 0.43063050508499146, "learning_rate": 8.520106240744759e-05, "loss": 0.0607, "step": 17330 }, { "grad_norm": 0.5426685810089111, "learning_rate": 8.518148607202102e-05, "loss": 0.0571, "step": 17340 }, { "grad_norm": 0.5293642282485962, "learning_rate": 8.51618990493879e-05, "loss": 0.0523, "step": 17350 }, { "grad_norm": 0.6224992275238037, "learning_rate": 8.514230134549823e-05, "loss": 0.0593, "step": 17360 }, { "grad_norm": 0.5100343823432922, "learning_rate": 8.51226929663053e-05, "loss": 0.0615, "step": 17370 }, { "grad_norm": 0.6257824897766113, "learning_rate": 8.51030739177656e-05, "loss": 0.0569, "step": 17380 }, { "grad_norm": 0.7002766132354736, "learning_rate": 8.508344420583889e-05, "loss": 0.0624, "step": 17390 }, { "grad_norm": 0.6955626606941223, "learning_rate": 8.506380383648816e-05, "loss": 0.0632, "step": 17400 }, { "grad_norm": 0.6293712258338928, "learning_rate": 8.504415281567963e-05, "loss": 0.0709, "step": 17410 }, { "grad_norm": 0.4944801330566406, "learning_rate": 8.502449114938275e-05, "loss": 0.0611, "step": 17420 }, { "grad_norm": 0.5467088222503662, "learning_rate": 8.500481884357025e-05, "loss": 0.065, "step": 17430 }, { "grad_norm": 0.8950826525688171, "learning_rate": 8.498513590421801e-05, "loss": 0.1201, "step": 17440 }, { "grad_norm": 0.7947143316268921, "learning_rate": 8.496544233730522e-05, "loss": 0.0653, "step": 17450 }, { "grad_norm": 0.5524119138717651, "learning_rate": 8.494573814881426e-05, "loss": 0.0596, "step": 17460 }, { "grad_norm": 0.8807410597801208, "learning_rate": 8.492602334473074e-05, "loss": 0.063, "step": 17470 }, { "grad_norm": 0.37819644808769226, "learning_rate": 8.49062979310435e-05, "loss": 0.0606, "step": 17480 }, { "grad_norm": 0.47545522451400757, "learning_rate": 8.488656191374458e-05, "loss": 0.0606, "step": 17490 }, { "grad_norm": 0.5076805949211121, "learning_rate": 8.48668152988293e-05, "loss": 0.0831, "step": 17500 }, { "grad_norm": 0.4736587107181549, "learning_rate": 8.484705809229612e-05, "loss": 0.0581, "step": 17510 }, { "grad_norm": 0.47099408507347107, "learning_rate": 8.482729030014677e-05, "loss": 0.0753, "step": 17520 }, { "grad_norm": 0.7128864526748657, "learning_rate": 8.48075119283862e-05, "loss": 0.0796, "step": 17530 }, { "grad_norm": 0.5990363359451294, "learning_rate": 8.478772298302254e-05, "loss": 0.0717, "step": 17540 }, { "grad_norm": 0.5745400786399841, "learning_rate": 8.476792347006716e-05, "loss": 0.0617, "step": 17550 }, { "grad_norm": 0.43533918261528015, "learning_rate": 8.474811339553462e-05, "loss": 0.0926, "step": 17560 }, { "grad_norm": 0.6064461469650269, "learning_rate": 8.47282927654427e-05, "loss": 0.069, "step": 17570 }, { "grad_norm": 0.6536003351211548, "learning_rate": 8.470846158581238e-05, "loss": 0.0677, "step": 17580 }, { "grad_norm": 0.7067998051643372, "learning_rate": 8.468861986266787e-05, "loss": 0.0608, "step": 17590 }, { "grad_norm": 0.6265163421630859, "learning_rate": 8.466876760203654e-05, "loss": 0.0577, "step": 17600 }, { "grad_norm": 0.5794499516487122, "learning_rate": 8.464890480994898e-05, "loss": 0.074, "step": 17610 }, { "grad_norm": 0.9129511713981628, "learning_rate": 8.462903149243899e-05, "loss": 0.083, "step": 17620 }, { "grad_norm": 0.578751266002655, "learning_rate": 8.460914765554357e-05, "loss": 0.0652, "step": 17630 }, { "grad_norm": 0.7998288869857788, "learning_rate": 8.458925330530288e-05, "loss": 0.0733, "step": 17640 }, { "grad_norm": 0.5034704804420471, "learning_rate": 8.456934844776032e-05, "loss": 0.0688, "step": 17650 }, { "grad_norm": 0.4672757089138031, "learning_rate": 8.454943308896246e-05, "loss": 0.0603, "step": 17660 }, { "grad_norm": 0.6411296129226685, "learning_rate": 8.452950723495905e-05, "loss": 0.0629, "step": 17670 }, { "grad_norm": 0.9404449462890625, "learning_rate": 8.450957089180303e-05, "loss": 0.0646, "step": 17680 }, { "grad_norm": 0.6943418383598328, "learning_rate": 8.448962406555055e-05, "loss": 0.0611, "step": 17690 }, { "grad_norm": 0.8999578952789307, "learning_rate": 8.446966676226093e-05, "loss": 0.0734, "step": 17700 }, { "grad_norm": 0.47619786858558655, "learning_rate": 8.444969898799667e-05, "loss": 0.0626, "step": 17710 }, { "grad_norm": 0.4973874092102051, "learning_rate": 8.442972074882343e-05, "loss": 0.0662, "step": 17720 }, { "grad_norm": 0.8655118346214294, "learning_rate": 8.44097320508101e-05, "loss": 0.0633, "step": 17730 }, { "grad_norm": 0.7083456516265869, "learning_rate": 8.43897329000287e-05, "loss": 0.0697, "step": 17740 }, { "grad_norm": 0.6689200401306152, "learning_rate": 8.436972330255448e-05, "loss": 0.076, "step": 17750 }, { "grad_norm": 0.566824197769165, "learning_rate": 8.434970326446579e-05, "loss": 0.0602, "step": 17760 }, { "grad_norm": 0.7593927979469299, "learning_rate": 8.432967279184418e-05, "loss": 0.0966, "step": 17770 }, { "grad_norm": 0.5609734654426575, "learning_rate": 8.430963189077441e-05, "loss": 0.0638, "step": 17780 }, { "grad_norm": 0.5321109890937805, "learning_rate": 8.428958056734437e-05, "loss": 0.0626, "step": 17790 }, { "grad_norm": 0.6016504168510437, "learning_rate": 8.426951882764513e-05, "loss": 0.0668, "step": 17800 }, { "grad_norm": 0.8467772603034973, "learning_rate": 8.424944667777089e-05, "loss": 0.0783, "step": 17810 }, { "grad_norm": 0.5836473107337952, "learning_rate": 8.422936412381905e-05, "loss": 0.0626, "step": 17820 }, { "grad_norm": 1.0035412311553955, "learning_rate": 8.420927117189017e-05, "loss": 0.0922, "step": 17830 }, { "grad_norm": 0.6643530130386353, "learning_rate": 8.418916782808795e-05, "loss": 0.0801, "step": 17840 }, { "grad_norm": 0.5385020971298218, "learning_rate": 8.416905409851926e-05, "loss": 0.0505, "step": 17850 }, { "grad_norm": 0.6855332851409912, "learning_rate": 8.41489299892941e-05, "loss": 0.0713, "step": 17860 }, { "grad_norm": 0.6654563546180725, "learning_rate": 8.412879550652566e-05, "loss": 0.0672, "step": 17870 }, { "grad_norm": 0.4734596610069275, "learning_rate": 8.410865065633029e-05, "loss": 0.0807, "step": 17880 }, { "grad_norm": 0.48210451006889343, "learning_rate": 8.408849544482742e-05, "loss": 0.0749, "step": 17890 }, { "grad_norm": 0.43486806750297546, "learning_rate": 8.406832987813968e-05, "loss": 0.052, "step": 17900 }, { "grad_norm": 0.49333319067955017, "learning_rate": 8.404815396239286e-05, "loss": 0.0496, "step": 17910 }, { "grad_norm": 0.5720226168632507, "learning_rate": 8.402796770371587e-05, "loss": 0.0813, "step": 17920 }, { "grad_norm": 0.4573301076889038, "learning_rate": 8.400777110824071e-05, "loss": 0.0674, "step": 17930 }, { "grad_norm": 0.6681380867958069, "learning_rate": 8.398756418210263e-05, "loss": 0.0544, "step": 17940 }, { "grad_norm": 0.4722687304019928, "learning_rate": 8.396734693143993e-05, "loss": 0.0691, "step": 17950 }, { "grad_norm": 0.6925555467605591, "learning_rate": 8.39471193623941e-05, "loss": 0.0599, "step": 17960 }, { "grad_norm": 0.6087562441825867, "learning_rate": 8.392688148110974e-05, "loss": 0.1018, "step": 17970 }, { "grad_norm": 0.6598284244537354, "learning_rate": 8.390663329373456e-05, "loss": 0.0851, "step": 17980 }, { "grad_norm": 0.7061864733695984, "learning_rate": 8.388637480641944e-05, "loss": 0.083, "step": 17990 }, { "grad_norm": 0.6059969663619995, "learning_rate": 8.386610602531837e-05, "loss": 0.0613, "step": 18000 }, { "grad_norm": 0.6364942789077759, "learning_rate": 8.384582695658847e-05, "loss": 0.0587, "step": 18010 }, { "grad_norm": 0.6109086275100708, "learning_rate": 8.382553760638999e-05, "loss": 0.0617, "step": 18020 }, { "grad_norm": 0.5726967453956604, "learning_rate": 8.380523798088631e-05, "loss": 0.0653, "step": 18030 }, { "grad_norm": 0.5761200189590454, "learning_rate": 8.378492808624389e-05, "loss": 0.083, "step": 18040 }, { "grad_norm": 0.7875831723213196, "learning_rate": 8.376460792863237e-05, "loss": 0.0988, "step": 18050 }, { "grad_norm": 0.4263255298137665, "learning_rate": 8.374427751422444e-05, "loss": 0.07, "step": 18060 }, { "grad_norm": 0.5171959400177002, "learning_rate": 8.3723936849196e-05, "loss": 0.0702, "step": 18070 }, { "grad_norm": 0.3713339865207672, "learning_rate": 8.370358593972595e-05, "loss": 0.0618, "step": 18080 }, { "grad_norm": 0.7346791625022888, "learning_rate": 8.36832247919964e-05, "loss": 0.0684, "step": 18090 }, { "grad_norm": 0.6378305554389954, "learning_rate": 8.36628534121925e-05, "loss": 0.0655, "step": 18100 }, { "grad_norm": 0.4749068021774292, "learning_rate": 8.364247180650254e-05, "loss": 0.0812, "step": 18110 }, { "grad_norm": 0.6720812320709229, "learning_rate": 8.362207998111794e-05, "loss": 0.0656, "step": 18120 }, { "grad_norm": 0.5318296551704407, "learning_rate": 8.360167794223318e-05, "loss": 0.0579, "step": 18130 }, { "grad_norm": 0.5039643049240112, "learning_rate": 8.358126569604586e-05, "loss": 0.0654, "step": 18140 }, { "grad_norm": 0.7703626155853271, "learning_rate": 8.356084324875668e-05, "loss": 0.0544, "step": 18150 }, { "grad_norm": 0.6006534695625305, "learning_rate": 8.354041060656945e-05, "loss": 0.0629, "step": 18160 }, { "grad_norm": 0.46297207474708557, "learning_rate": 8.351996777569106e-05, "loss": 0.087, "step": 18170 }, { "grad_norm": 0.7560280561447144, "learning_rate": 8.349951476233148e-05, "loss": 0.0732, "step": 18180 }, { "grad_norm": 0.6163004636764526, "learning_rate": 8.347905157270386e-05, "loss": 0.0826, "step": 18190 }, { "grad_norm": 0.6500102281570435, "learning_rate": 8.345857821302432e-05, "loss": 0.069, "step": 18200 }, { "grad_norm": 0.4958086311817169, "learning_rate": 8.343809468951213e-05, "loss": 0.0685, "step": 18210 }, { "grad_norm": 0.825712263584137, "learning_rate": 8.341760100838965e-05, "loss": 0.06, "step": 18220 }, { "grad_norm": 0.49840372800827026, "learning_rate": 8.339709717588233e-05, "loss": 0.0579, "step": 18230 }, { "grad_norm": 0.678253173828125, "learning_rate": 8.33765831982187e-05, "loss": 0.0767, "step": 18240 }, { "grad_norm": 0.6380348205566406, "learning_rate": 8.335605908163035e-05, "loss": 0.0888, "step": 18250 }, { "grad_norm": 0.808606743812561, "learning_rate": 8.333552483235196e-05, "loss": 0.0592, "step": 18260 }, { "grad_norm": 0.5053474307060242, "learning_rate": 8.33149804566213e-05, "loss": 0.0588, "step": 18270 }, { "grad_norm": 0.6803098320960999, "learning_rate": 8.329442596067921e-05, "loss": 0.0695, "step": 18280 }, { "grad_norm": 0.9858689308166504, "learning_rate": 8.32738613507696e-05, "loss": 0.0945, "step": 18290 }, { "grad_norm": 0.48628324270248413, "learning_rate": 8.325328663313946e-05, "loss": 0.0612, "step": 18300 }, { "grad_norm": 0.7307419776916504, "learning_rate": 8.323270181403884e-05, "loss": 0.0588, "step": 18310 }, { "grad_norm": 0.6902833580970764, "learning_rate": 8.321210689972086e-05, "loss": 0.0687, "step": 18320 }, { "grad_norm": 0.4986927807331085, "learning_rate": 8.319150189644174e-05, "loss": 0.0654, "step": 18330 }, { "grad_norm": 0.5118420124053955, "learning_rate": 8.31708868104607e-05, "loss": 0.0578, "step": 18340 }, { "grad_norm": 0.48217159509658813, "learning_rate": 8.315026164804007e-05, "loss": 0.0632, "step": 18350 }, { "grad_norm": 0.509125828742981, "learning_rate": 8.312962641544524e-05, "loss": 0.0615, "step": 18360 }, { "grad_norm": 0.6923204064369202, "learning_rate": 8.310898111894465e-05, "loss": 0.0555, "step": 18370 }, { "grad_norm": 0.38137853145599365, "learning_rate": 8.308832576480977e-05, "loss": 0.0475, "step": 18380 }, { "grad_norm": 0.4411822259426117, "learning_rate": 8.306766035931519e-05, "loss": 0.0665, "step": 18390 }, { "grad_norm": 0.5363131761550903, "learning_rate": 8.304698490873847e-05, "loss": 0.0611, "step": 18400 }, { "grad_norm": 0.9374459981918335, "learning_rate": 8.30262994193603e-05, "loss": 0.067, "step": 18410 }, { "grad_norm": 0.7523533701896667, "learning_rate": 8.300560389746438e-05, "loss": 0.0802, "step": 18420 }, { "grad_norm": 0.43294790387153625, "learning_rate": 8.298489834933745e-05, "loss": 0.0768, "step": 18430 }, { "grad_norm": 0.4928245544433594, "learning_rate": 8.296418278126934e-05, "loss": 0.067, "step": 18440 }, { "grad_norm": 0.571965754032135, "learning_rate": 8.294345719955284e-05, "loss": 0.0588, "step": 18450 }, { "grad_norm": 0.5943648815155029, "learning_rate": 8.29227216104839e-05, "loss": 0.0662, "step": 18460 }, { "grad_norm": 0.6745360493659973, "learning_rate": 8.290197602036137e-05, "loss": 0.0526, "step": 18470 }, { "grad_norm": 0.6642513275146484, "learning_rate": 8.288122043548725e-05, "loss": 0.0788, "step": 18480 }, { "grad_norm": 0.37787124514579773, "learning_rate": 8.286045486216657e-05, "loss": 0.0507, "step": 18490 }, { "grad_norm": 0.7522961497306824, "learning_rate": 8.283967930670733e-05, "loss": 0.0637, "step": 18500 }, { "grad_norm": 0.6092494130134583, "learning_rate": 8.281889377542058e-05, "loss": 0.0655, "step": 18510 }, { "grad_norm": 0.4919734299182892, "learning_rate": 8.279809827462045e-05, "loss": 0.0541, "step": 18520 }, { "grad_norm": 0.5007995367050171, "learning_rate": 8.277729281062402e-05, "loss": 0.0535, "step": 18530 }, { "grad_norm": 0.5525718927383423, "learning_rate": 8.27564773897515e-05, "loss": 0.0725, "step": 18540 }, { "grad_norm": 0.6609475612640381, "learning_rate": 8.273565201832602e-05, "loss": 0.0601, "step": 18550 }, { "grad_norm": 0.9602296352386475, "learning_rate": 8.27148167026738e-05, "loss": 0.075, "step": 18560 }, { "grad_norm": 0.555499792098999, "learning_rate": 8.269397144912405e-05, "loss": 0.0859, "step": 18570 }, { "grad_norm": 1.1970312595367432, "learning_rate": 8.267311626400899e-05, "loss": 0.0909, "step": 18580 }, { "grad_norm": 0.5892061591148376, "learning_rate": 8.26522511536639e-05, "loss": 0.0745, "step": 18590 }, { "grad_norm": 0.5672520399093628, "learning_rate": 8.263137612442706e-05, "loss": 0.0801, "step": 18600 }, { "grad_norm": 0.5429905652999878, "learning_rate": 8.261049118263971e-05, "loss": 0.056, "step": 18610 }, { "grad_norm": 0.6268524527549744, "learning_rate": 8.258959633464619e-05, "loss": 0.0579, "step": 18620 }, { "grad_norm": 0.45909443497657776, "learning_rate": 8.256869158679377e-05, "loss": 0.0789, "step": 18630 }, { "grad_norm": 0.6589524149894714, "learning_rate": 8.254777694543278e-05, "loss": 0.0818, "step": 18640 }, { "grad_norm": 0.6721994876861572, "learning_rate": 8.252685241691651e-05, "loss": 0.0729, "step": 18650 }, { "grad_norm": 0.5163321495056152, "learning_rate": 8.250591800760133e-05, "loss": 0.057, "step": 18660 }, { "grad_norm": 0.5781989693641663, "learning_rate": 8.248497372384649e-05, "loss": 0.0755, "step": 18670 }, { "grad_norm": 0.5432996153831482, "learning_rate": 8.246401957201437e-05, "loss": 0.0612, "step": 18680 }, { "grad_norm": 0.5878350138664246, "learning_rate": 8.244305555847027e-05, "loss": 0.0633, "step": 18690 }, { "grad_norm": 0.42337363958358765, "learning_rate": 8.24220816895825e-05, "loss": 0.0728, "step": 18700 }, { "grad_norm": 0.6126437783241272, "learning_rate": 8.240109797172237e-05, "loss": 0.0583, "step": 18710 }, { "grad_norm": 0.574105441570282, "learning_rate": 8.238010441126416e-05, "loss": 0.068, "step": 18720 }, { "grad_norm": 0.7330495119094849, "learning_rate": 8.23591010145852e-05, "loss": 0.0686, "step": 18730 }, { "grad_norm": 0.81648850440979, "learning_rate": 8.233808778806571e-05, "loss": 0.0946, "step": 18740 }, { "grad_norm": 0.5739789009094238, "learning_rate": 8.231706473808903e-05, "loss": 0.0724, "step": 18750 }, { "grad_norm": 0.5576875805854797, "learning_rate": 8.229603187104133e-05, "loss": 0.0706, "step": 18760 }, { "grad_norm": 0.5675524473190308, "learning_rate": 8.22749891933119e-05, "loss": 0.0739, "step": 18770 }, { "grad_norm": 0.5155162811279297, "learning_rate": 8.225393671129291e-05, "loss": 0.0797, "step": 18780 }, { "grad_norm": 0.6653082370758057, "learning_rate": 8.223287443137957e-05, "loss": 0.0664, "step": 18790 }, { "grad_norm": 0.49437177181243896, "learning_rate": 8.221180235997004e-05, "loss": 0.074, "step": 18800 }, { "grad_norm": 0.7094817161560059, "learning_rate": 8.219072050346544e-05, "loss": 0.078, "step": 18810 }, { "grad_norm": 0.704899787902832, "learning_rate": 8.216962886826992e-05, "loss": 0.0486, "step": 18820 }, { "grad_norm": 0.5956948399543762, "learning_rate": 8.214852746079054e-05, "loss": 0.0524, "step": 18830 }, { "grad_norm": 0.7884605526924133, "learning_rate": 8.212741628743732e-05, "loss": 0.0647, "step": 18840 }, { "grad_norm": 0.6870574951171875, "learning_rate": 8.210629535462333e-05, "loss": 0.0646, "step": 18850 }, { "grad_norm": 0.6332570910453796, "learning_rate": 8.208516466876453e-05, "loss": 0.0616, "step": 18860 }, { "grad_norm": 0.4096418619155884, "learning_rate": 8.206402423627986e-05, "loss": 0.0478, "step": 18870 }, { "grad_norm": 0.6498658657073975, "learning_rate": 8.204287406359124e-05, "loss": 0.0653, "step": 18880 }, { "grad_norm": 0.5986441373825073, "learning_rate": 8.20217141571235e-05, "loss": 0.0669, "step": 18890 }, { "grad_norm": 0.576495349407196, "learning_rate": 8.200054452330449e-05, "loss": 0.0536, "step": 18900 }, { "grad_norm": 0.5487883687019348, "learning_rate": 8.197936516856499e-05, "loss": 0.0559, "step": 18910 }, { "grad_norm": 0.4580700695514679, "learning_rate": 8.195817609933871e-05, "loss": 0.0625, "step": 18920 }, { "grad_norm": 0.4507739841938019, "learning_rate": 8.193697732206233e-05, "loss": 0.0771, "step": 18930 }, { "grad_norm": 0.6927520632743835, "learning_rate": 8.19157688431755e-05, "loss": 0.0906, "step": 18940 }, { "grad_norm": 0.6201183795928955, "learning_rate": 8.189455066912077e-05, "loss": 0.0722, "step": 18950 }, { "grad_norm": 0.6751024127006531, "learning_rate": 8.187332280634369e-05, "loss": 0.0651, "step": 18960 }, { "grad_norm": 0.4933755695819855, "learning_rate": 8.18520852612927e-05, "loss": 0.0477, "step": 18970 }, { "grad_norm": 0.6107763648033142, "learning_rate": 8.183083804041921e-05, "loss": 0.0641, "step": 18980 }, { "grad_norm": 0.31146836280822754, "learning_rate": 8.180958115017757e-05, "loss": 0.0542, "step": 18990 }, { "grad_norm": 0.5889636874198914, "learning_rate": 8.178831459702505e-05, "loss": 0.0597, "step": 19000 }, { "grad_norm": 0.5478366017341614, "learning_rate": 8.17670383874219e-05, "loss": 0.0805, "step": 19010 }, { "grad_norm": 0.6656046509742737, "learning_rate": 8.174575252783124e-05, "loss": 0.0682, "step": 19020 }, { "grad_norm": 0.6721125841140747, "learning_rate": 8.172445702471914e-05, "loss": 0.061, "step": 19030 }, { "grad_norm": 0.6977675557136536, "learning_rate": 8.170315188455466e-05, "loss": 0.0819, "step": 19040 }, { "grad_norm": 0.4132765233516693, "learning_rate": 8.168183711380969e-05, "loss": 0.0673, "step": 19050 }, { "grad_norm": 1.0105555057525635, "learning_rate": 8.166051271895913e-05, "loss": 0.0701, "step": 19060 }, { "grad_norm": 0.6602125763893127, "learning_rate": 8.163917870648075e-05, "loss": 0.0579, "step": 19070 }, { "grad_norm": 0.778460681438446, "learning_rate": 8.161783508285526e-05, "loss": 0.081, "step": 19080 }, { "grad_norm": 0.8374301195144653, "learning_rate": 8.159648185456628e-05, "loss": 0.0785, "step": 19090 }, { "grad_norm": 0.6137996912002563, "learning_rate": 8.157511902810038e-05, "loss": 0.0637, "step": 19100 }, { "grad_norm": 0.7191548347473145, "learning_rate": 8.155374660994701e-05, "loss": 0.0956, "step": 19110 }, { "grad_norm": 0.450888454914093, "learning_rate": 8.153236460659857e-05, "loss": 0.0647, "step": 19120 }, { "grad_norm": 0.8217523694038391, "learning_rate": 8.151097302455031e-05, "loss": 0.0777, "step": 19130 }, { "grad_norm": 0.6399559378623962, "learning_rate": 8.148957187030044e-05, "loss": 0.0787, "step": 19140 }, { "grad_norm": 0.48173025250434875, "learning_rate": 8.146816115035006e-05, "loss": 0.0677, "step": 19150 }, { "grad_norm": 0.8886949419975281, "learning_rate": 8.14467408712032e-05, "loss": 0.0683, "step": 19160 }, { "grad_norm": 0.6398653388023376, "learning_rate": 8.142531103936678e-05, "loss": 0.0561, "step": 19170 }, { "grad_norm": 0.4960240423679352, "learning_rate": 8.14038716613506e-05, "loss": 0.0635, "step": 19180 }, { "grad_norm": 0.4594174325466156, "learning_rate": 8.138242274366736e-05, "loss": 0.0623, "step": 19190 }, { "grad_norm": 0.5771855115890503, "learning_rate": 8.136096429283271e-05, "loss": 0.0637, "step": 19200 }, { "grad_norm": 0.6383109092712402, "learning_rate": 8.133949631536515e-05, "loss": 0.0561, "step": 19210 }, { "grad_norm": 0.6541150212287903, "learning_rate": 8.131801881778607e-05, "loss": 0.0614, "step": 19220 }, { "grad_norm": 0.5137127637863159, "learning_rate": 8.129653180661978e-05, "loss": 0.0784, "step": 19230 }, { "grad_norm": 0.5858141183853149, "learning_rate": 8.127503528839346e-05, "loss": 0.0745, "step": 19240 }, { "grad_norm": 0.5953584909439087, "learning_rate": 8.125352926963721e-05, "loss": 0.0806, "step": 19250 }, { "grad_norm": 0.43331074714660645, "learning_rate": 8.123201375688395e-05, "loss": 0.0536, "step": 19260 }, { "grad_norm": 0.655847430229187, "learning_rate": 8.121048875666954e-05, "loss": 0.0633, "step": 19270 }, { "grad_norm": 0.8043959736824036, "learning_rate": 8.118895427553274e-05, "loss": 0.0751, "step": 19280 }, { "grad_norm": 0.7112107276916504, "learning_rate": 8.116741032001511e-05, "loss": 0.0878, "step": 19290 }, { "grad_norm": 0.7805179357528687, "learning_rate": 8.114585689666114e-05, "loss": 0.0601, "step": 19300 }, { "grad_norm": 0.6103417277336121, "learning_rate": 8.112429401201821e-05, "loss": 0.0635, "step": 19310 }, { "grad_norm": 0.5599119067192078, "learning_rate": 8.110272167263656e-05, "loss": 0.0726, "step": 19320 }, { "grad_norm": 0.6411116719245911, "learning_rate": 8.108113988506929e-05, "loss": 0.0623, "step": 19330 }, { "grad_norm": 0.9190870523452759, "learning_rate": 8.105954865587235e-05, "loss": 0.0809, "step": 19340 }, { "grad_norm": 0.9967173337936401, "learning_rate": 8.103794799160463e-05, "loss": 0.0745, "step": 19350 }, { "grad_norm": 0.6203922629356384, "learning_rate": 8.101633789882781e-05, "loss": 0.0649, "step": 19360 }, { "grad_norm": 0.6873630285263062, "learning_rate": 8.099471838410648e-05, "loss": 0.0651, "step": 19370 }, { "grad_norm": 0.674561619758606, "learning_rate": 8.097308945400806e-05, "loss": 0.075, "step": 19380 }, { "grad_norm": 0.3614344596862793, "learning_rate": 8.095145111510288e-05, "loss": 0.0648, "step": 19390 }, { "grad_norm": 0.5239447355270386, "learning_rate": 8.092980337396406e-05, "loss": 0.0624, "step": 19400 }, { "grad_norm": 0.5203665494918823, "learning_rate": 8.090814623716763e-05, "loss": 0.0669, "step": 19410 }, { "grad_norm": 0.6145167946815491, "learning_rate": 8.088647971129246e-05, "loss": 0.066, "step": 19420 }, { "grad_norm": 0.5463963150978088, "learning_rate": 8.086480380292026e-05, "loss": 0.0564, "step": 19430 }, { "grad_norm": 0.6699617505073547, "learning_rate": 8.084311851863562e-05, "loss": 0.0716, "step": 19440 }, { "grad_norm": 0.520832896232605, "learning_rate": 8.082142386502591e-05, "loss": 0.0743, "step": 19450 }, { "grad_norm": 0.5786413550376892, "learning_rate": 8.079971984868145e-05, "loss": 0.0838, "step": 19460 }, { "grad_norm": 0.4693431258201599, "learning_rate": 8.077800647619532e-05, "loss": 0.0737, "step": 19470 }, { "grad_norm": 0.7193844318389893, "learning_rate": 8.075628375416345e-05, "loss": 0.0741, "step": 19480 }, { "grad_norm": 0.5414043068885803, "learning_rate": 8.073455168918464e-05, "loss": 0.0724, "step": 19490 }, { "grad_norm": 0.45611873269081116, "learning_rate": 8.071281028786055e-05, "loss": 0.0546, "step": 19500 }, { "grad_norm": 0.647841215133667, "learning_rate": 8.069105955679562e-05, "loss": 0.0714, "step": 19510 }, { "grad_norm": 0.5243976712226868, "learning_rate": 8.066929950259713e-05, "loss": 0.0651, "step": 19520 }, { "grad_norm": 0.6211372017860413, "learning_rate": 8.064753013187522e-05, "loss": 0.0653, "step": 19530 }, { "grad_norm": 0.6762186288833618, "learning_rate": 8.062575145124289e-05, "loss": 0.0632, "step": 19540 }, { "grad_norm": 0.7572778463363647, "learning_rate": 8.060396346731587e-05, "loss": 0.0778, "step": 19550 }, { "grad_norm": 0.6068366765975952, "learning_rate": 8.058216618671281e-05, "loss": 0.0706, "step": 19560 }, { "grad_norm": 0.6597235798835754, "learning_rate": 8.056035961605514e-05, "loss": 0.0759, "step": 19570 }, { "grad_norm": 0.6520105004310608, "learning_rate": 8.05385437619671e-05, "loss": 0.07, "step": 19580 }, { "grad_norm": 0.5805871486663818, "learning_rate": 8.05167186310758e-05, "loss": 0.0654, "step": 19590 }, { "grad_norm": 0.7118812203407288, "learning_rate": 8.049488423001113e-05, "loss": 0.0604, "step": 19600 }, { "grad_norm": 0.6779130697250366, "learning_rate": 8.047304056540581e-05, "loss": 0.0778, "step": 19610 }, { "grad_norm": 0.4550801217556, "learning_rate": 8.045118764389534e-05, "loss": 0.0686, "step": 19620 }, { "grad_norm": 0.48796918988227844, "learning_rate": 8.042932547211809e-05, "loss": 0.0858, "step": 19630 }, { "grad_norm": 0.5265882611274719, "learning_rate": 8.04074540567152e-05, "loss": 0.0699, "step": 19640 }, { "grad_norm": 0.611876368522644, "learning_rate": 8.038557340433063e-05, "loss": 0.072, "step": 19650 }, { "grad_norm": 0.6198515892028809, "learning_rate": 8.036368352161115e-05, "loss": 0.076, "step": 19660 }, { "grad_norm": 0.8479505181312561, "learning_rate": 8.034178441520633e-05, "loss": 0.0816, "step": 19670 }, { "grad_norm": 0.6344677805900574, "learning_rate": 8.031987609176852e-05, "loss": 0.0758, "step": 19680 }, { "grad_norm": 0.40476110577583313, "learning_rate": 8.02979585579529e-05, "loss": 0.0832, "step": 19690 }, { "grad_norm": 0.5971135497093201, "learning_rate": 8.027603182041745e-05, "loss": 0.0651, "step": 19700 }, { "grad_norm": 0.520113468170166, "learning_rate": 8.025409588582292e-05, "loss": 0.0583, "step": 19710 }, { "grad_norm": 0.6155751347541809, "learning_rate": 8.023215076083288e-05, "loss": 0.0657, "step": 19720 }, { "grad_norm": 0.7323774695396423, "learning_rate": 8.021019645211367e-05, "loss": 0.0763, "step": 19730 }, { "grad_norm": 0.41758260130882263, "learning_rate": 8.018823296633441e-05, "loss": 0.0606, "step": 19740 }, { "grad_norm": 0.5553961396217346, "learning_rate": 8.016626031016708e-05, "loss": 0.0674, "step": 19750 }, { "grad_norm": 0.7639159560203552, "learning_rate": 8.014427849028636e-05, "loss": 0.0704, "step": 19760 }, { "grad_norm": 0.6271940469741821, "learning_rate": 8.012228751336974e-05, "loss": 0.051, "step": 19770 }, { "grad_norm": 0.563106894493103, "learning_rate": 8.01002873860975e-05, "loss": 0.0535, "step": 19780 }, { "grad_norm": 0.6345698833465576, "learning_rate": 8.00782781151527e-05, "loss": 0.0542, "step": 19790 }, { "grad_norm": 0.6944315433502197, "learning_rate": 8.005625970722119e-05, "loss": 0.0586, "step": 19800 }, { "grad_norm": 0.5382458567619324, "learning_rate": 8.003423216899158e-05, "loss": 0.0604, "step": 19810 }, { "grad_norm": 0.4393470883369446, "learning_rate": 8.001219550715522e-05, "loss": 0.0603, "step": 19820 }, { "grad_norm": 0.6979131102561951, "learning_rate": 7.999014972840632e-05, "loss": 0.0695, "step": 19830 }, { "grad_norm": 0.535869300365448, "learning_rate": 7.996809483944174e-05, "loss": 0.0702, "step": 19840 }, { "grad_norm": 0.443145215511322, "learning_rate": 7.994603084696124e-05, "loss": 0.0601, "step": 19850 }, { "grad_norm": 0.8903296589851379, "learning_rate": 7.992395775766724e-05, "loss": 0.0612, "step": 19860 }, { "grad_norm": 0.8179593682289124, "learning_rate": 7.990187557826497e-05, "loss": 0.0597, "step": 19870 }, { "grad_norm": 0.3709045350551605, "learning_rate": 7.987978431546242e-05, "loss": 0.0676, "step": 19880 }, { "grad_norm": 0.4735274612903595, "learning_rate": 7.985768397597031e-05, "loss": 0.0602, "step": 19890 }, { "grad_norm": 0.57332843542099, "learning_rate": 7.983557456650216e-05, "loss": 0.0652, "step": 19900 }, { "grad_norm": 0.4740515649318695, "learning_rate": 7.981345609377422e-05, "loss": 0.0591, "step": 19910 }, { "grad_norm": 0.5203772783279419, "learning_rate": 7.97913285645055e-05, "loss": 0.0711, "step": 19920 }, { "grad_norm": 0.4434860348701477, "learning_rate": 7.976919198541776e-05, "loss": 0.0603, "step": 19930 }, { "grad_norm": 0.6067339181900024, "learning_rate": 7.974704636323548e-05, "loss": 0.0718, "step": 19940 }, { "grad_norm": 0.7181215286254883, "learning_rate": 7.972489170468597e-05, "loss": 0.0794, "step": 19950 }, { "grad_norm": 0.574401319026947, "learning_rate": 7.970272801649918e-05, "loss": 0.05, "step": 19960 }, { "grad_norm": 0.5872327089309692, "learning_rate": 7.96805553054079e-05, "loss": 0.0803, "step": 19970 }, { "grad_norm": 0.6229599714279175, "learning_rate": 7.965837357814756e-05, "loss": 0.0656, "step": 19980 }, { "grad_norm": 0.6714332103729248, "learning_rate": 7.963618284145643e-05, "loss": 0.0545, "step": 19990 }, { "grad_norm": 0.9501820802688599, "learning_rate": 7.961398310207544e-05, "loss": 0.0829, "step": 20000 }, { "grad_norm": 0.5668588280677795, "learning_rate": 7.95917743667483e-05, "loss": 0.0686, "step": 20010 }, { "grad_norm": 0.6862373352050781, "learning_rate": 7.956955664222144e-05, "loss": 0.0527, "step": 20020 }, { "grad_norm": 0.339362233877182, "learning_rate": 7.954732993524399e-05, "loss": 0.0553, "step": 20030 }, { "grad_norm": 0.5800458192825317, "learning_rate": 7.952509425256786e-05, "loss": 0.0588, "step": 20040 }, { "grad_norm": 0.5669829249382019, "learning_rate": 7.950284960094767e-05, "loss": 0.0526, "step": 20050 }, { "grad_norm": 0.58428955078125, "learning_rate": 7.948059598714076e-05, "loss": 0.0612, "step": 20060 }, { "grad_norm": 0.5849586725234985, "learning_rate": 7.945833341790717e-05, "loss": 0.0776, "step": 20070 }, { "grad_norm": 0.47728267312049866, "learning_rate": 7.94360619000097e-05, "loss": 0.0663, "step": 20080 }, { "grad_norm": 0.9112526178359985, "learning_rate": 7.941378144021381e-05, "loss": 0.0707, "step": 20090 }, { "grad_norm": 0.6670194864273071, "learning_rate": 7.939149204528777e-05, "loss": 0.059, "step": 20100 }, { "grad_norm": 0.3600434362888336, "learning_rate": 7.936919372200246e-05, "loss": 0.082, "step": 20110 }, { "grad_norm": 0.6665999889373779, "learning_rate": 7.934688647713158e-05, "loss": 0.0623, "step": 20120 }, { "grad_norm": 0.48716050386428833, "learning_rate": 7.932457031745143e-05, "loss": 0.0551, "step": 20130 }, { "grad_norm": 0.6418197751045227, "learning_rate": 7.930224524974108e-05, "loss": 0.0697, "step": 20140 }, { "grad_norm": 0.6847913861274719, "learning_rate": 7.927991128078232e-05, "loss": 0.07, "step": 20150 }, { "grad_norm": 0.5159211158752441, "learning_rate": 7.925756841735958e-05, "loss": 0.068, "step": 20160 }, { "grad_norm": 0.6247290372848511, "learning_rate": 7.923521666626008e-05, "loss": 0.0628, "step": 20170 }, { "grad_norm": 0.5600696802139282, "learning_rate": 7.921285603427366e-05, "loss": 0.0767, "step": 20180 }, { "grad_norm": 0.6243577003479004, "learning_rate": 7.91904865281929e-05, "loss": 0.0627, "step": 20190 }, { "grad_norm": 1.0532207489013672, "learning_rate": 7.916810815481307e-05, "loss": 0.0739, "step": 20200 }, { "grad_norm": 0.5657014846801758, "learning_rate": 7.914572092093211e-05, "loss": 0.0492, "step": 20210 }, { "grad_norm": 0.4783635437488556, "learning_rate": 7.912332483335068e-05, "loss": 0.0487, "step": 20220 }, { "grad_norm": 0.714059054851532, "learning_rate": 7.910091989887213e-05, "loss": 0.0619, "step": 20230 }, { "grad_norm": 0.4823620319366455, "learning_rate": 7.907850612430248e-05, "loss": 0.0701, "step": 20240 }, { "grad_norm": 0.5529698133468628, "learning_rate": 7.905608351645044e-05, "loss": 0.0593, "step": 20250 }, { "grad_norm": 0.8669242262840271, "learning_rate": 7.90336520821274e-05, "loss": 0.0804, "step": 20260 }, { "grad_norm": 0.28970927000045776, "learning_rate": 7.901121182814746e-05, "loss": 0.0609, "step": 20270 }, { "grad_norm": 0.5400761365890503, "learning_rate": 7.898876276132736e-05, "loss": 0.0587, "step": 20280 }, { "grad_norm": 0.9014671444892883, "learning_rate": 7.896630488848654e-05, "loss": 0.0619, "step": 20290 }, { "grad_norm": 0.44458651542663574, "learning_rate": 7.89438382164471e-05, "loss": 0.061, "step": 20300 }, { "grad_norm": 0.5250499248504639, "learning_rate": 7.892136275203383e-05, "loss": 0.0511, "step": 20310 }, { "grad_norm": 0.5958737730979919, "learning_rate": 7.889887850207418e-05, "loss": 0.0729, "step": 20320 }, { "grad_norm": 0.4993458092212677, "learning_rate": 7.887638547339827e-05, "loss": 0.0592, "step": 20330 }, { "grad_norm": 0.5700199604034424, "learning_rate": 7.885388367283891e-05, "loss": 0.0619, "step": 20340 }, { "grad_norm": 0.8299131989479065, "learning_rate": 7.88313731072315e-05, "loss": 0.0819, "step": 20350 }, { "grad_norm": 0.5654445290565491, "learning_rate": 7.88088537834142e-05, "loss": 0.0778, "step": 20360 }, { "grad_norm": 0.4813821613788605, "learning_rate": 7.878632570822778e-05, "loss": 0.0604, "step": 20370 }, { "grad_norm": 0.7978344559669495, "learning_rate": 7.876378888851567e-05, "loss": 0.0855, "step": 20380 }, { "grad_norm": 0.5135444402694702, "learning_rate": 7.874124333112396e-05, "loss": 0.0777, "step": 20390 }, { "grad_norm": 0.6503910422325134, "learning_rate": 7.871868904290138e-05, "loss": 0.0515, "step": 20400 }, { "grad_norm": 0.5409901142120361, "learning_rate": 7.869612603069935e-05, "loss": 0.0629, "step": 20410 }, { "grad_norm": 0.4738769829273224, "learning_rate": 7.867355430137192e-05, "loss": 0.0577, "step": 20420 }, { "grad_norm": 0.6137699484825134, "learning_rate": 7.865097386177577e-05, "loss": 0.0645, "step": 20430 }, { "grad_norm": 0.46835389733314514, "learning_rate": 7.862838471877023e-05, "loss": 0.0695, "step": 20440 }, { "grad_norm": 0.4380277991294861, "learning_rate": 7.860578687921731e-05, "loss": 0.0582, "step": 20450 }, { "grad_norm": 0.6357768774032593, "learning_rate": 7.858318034998164e-05, "loss": 0.0777, "step": 20460 }, { "grad_norm": 0.7096210718154907, "learning_rate": 7.856056513793046e-05, "loss": 0.0703, "step": 20470 }, { "grad_norm": 0.40856656432151794, "learning_rate": 7.85379412499337e-05, "loss": 0.0665, "step": 20480 }, { "grad_norm": 0.46648237109184265, "learning_rate": 7.851530869286389e-05, "loss": 0.0616, "step": 20490 }, { "grad_norm": 0.5851783156394958, "learning_rate": 7.849266747359619e-05, "loss": 0.0598, "step": 20500 }, { "grad_norm": 0.6117192506790161, "learning_rate": 7.847001759900843e-05, "loss": 0.0649, "step": 20510 }, { "grad_norm": 0.763060986995697, "learning_rate": 7.844735907598102e-05, "loss": 0.0529, "step": 20520 }, { "grad_norm": 0.8467499017715454, "learning_rate": 7.842469191139703e-05, "loss": 0.0745, "step": 20530 }, { "grad_norm": 0.6126044988632202, "learning_rate": 7.840201611214215e-05, "loss": 0.0662, "step": 20540 }, { "grad_norm": 0.8515439629554749, "learning_rate": 7.837933168510469e-05, "loss": 0.0721, "step": 20550 }, { "grad_norm": 0.8726083636283875, "learning_rate": 7.835663863717559e-05, "loss": 0.0796, "step": 20560 }, { "grad_norm": 0.4639666974544525, "learning_rate": 7.833393697524838e-05, "loss": 0.0928, "step": 20570 }, { "grad_norm": 1.1313575506210327, "learning_rate": 7.831122670621922e-05, "loss": 0.063, "step": 20580 }, { "grad_norm": 0.606310248374939, "learning_rate": 7.82885078369869e-05, "loss": 0.0614, "step": 20590 }, { "grad_norm": 0.8246224522590637, "learning_rate": 7.826578037445283e-05, "loss": 0.0713, "step": 20600 }, { "grad_norm": 0.46208450198173523, "learning_rate": 7.824304432552097e-05, "loss": 0.0487, "step": 20610 }, { "grad_norm": 0.8883548974990845, "learning_rate": 7.822029969709798e-05, "loss": 0.0592, "step": 20620 }, { "grad_norm": 0.5551581382751465, "learning_rate": 7.819754649609306e-05, "loss": 0.0794, "step": 20630 }, { "grad_norm": 0.5803718566894531, "learning_rate": 7.817478472941802e-05, "loss": 0.0631, "step": 20640 }, { "grad_norm": 0.562574565410614, "learning_rate": 7.815201440398727e-05, "loss": 0.0585, "step": 20650 }, { "grad_norm": 0.5350598096847534, "learning_rate": 7.812923552671789e-05, "loss": 0.0635, "step": 20660 }, { "grad_norm": 0.7343052625656128, "learning_rate": 7.810644810452945e-05, "loss": 0.0538, "step": 20670 }, { "grad_norm": 0.6354627013206482, "learning_rate": 7.808365214434417e-05, "loss": 0.0707, "step": 20680 }, { "grad_norm": 0.6608133912086487, "learning_rate": 7.80608476530869e-05, "loss": 0.0637, "step": 20690 }, { "grad_norm": 0.9635444283485413, "learning_rate": 7.8038034637685e-05, "loss": 0.0756, "step": 20700 }, { "grad_norm": 0.6170450448989868, "learning_rate": 7.801521310506848e-05, "loss": 0.0728, "step": 20710 }, { "grad_norm": 1.2503132820129395, "learning_rate": 7.799238306216994e-05, "loss": 0.0587, "step": 20720 }, { "grad_norm": 0.8225238919258118, "learning_rate": 7.796954451592448e-05, "loss": 0.0735, "step": 20730 }, { "grad_norm": 0.674641489982605, "learning_rate": 7.794669747326992e-05, "loss": 0.0594, "step": 20740 }, { "grad_norm": 1.2062830924987793, "learning_rate": 7.792384194114654e-05, "loss": 0.0686, "step": 20750 }, { "grad_norm": 0.4762920141220093, "learning_rate": 7.790097792649729e-05, "loss": 0.0637, "step": 20760 }, { "grad_norm": 0.7139250636100769, "learning_rate": 7.787810543626762e-05, "loss": 0.0643, "step": 20770 }, { "grad_norm": 0.518744945526123, "learning_rate": 7.785522447740558e-05, "loss": 0.0533, "step": 20780 }, { "grad_norm": 0.8154358267784119, "learning_rate": 7.783233505686182e-05, "loss": 0.0658, "step": 20790 }, { "grad_norm": 0.39006441831588745, "learning_rate": 7.780943718158955e-05, "loss": 0.06, "step": 20800 }, { "grad_norm": 0.7183142304420471, "learning_rate": 7.778653085854453e-05, "loss": 0.0606, "step": 20810 }, { "grad_norm": 0.8365603685379028, "learning_rate": 7.77636160946851e-05, "loss": 0.0669, "step": 20820 }, { "grad_norm": 0.7717646360397339, "learning_rate": 7.774069289697215e-05, "loss": 0.0716, "step": 20830 }, { "grad_norm": 0.32567068934440613, "learning_rate": 7.771776127236913e-05, "loss": 0.0631, "step": 20840 }, { "grad_norm": 0.4502370357513428, "learning_rate": 7.769482122784212e-05, "loss": 0.065, "step": 20850 }, { "grad_norm": 0.6545519232749939, "learning_rate": 7.767187277035963e-05, "loss": 0.0533, "step": 20860 }, { "grad_norm": 0.630182683467865, "learning_rate": 7.764891590689285e-05, "loss": 0.0534, "step": 20870 }, { "grad_norm": 0.8816582560539246, "learning_rate": 7.762595064441542e-05, "loss": 0.0705, "step": 20880 }, { "grad_norm": 0.750651478767395, "learning_rate": 7.760297698990362e-05, "loss": 0.059, "step": 20890 }, { "grad_norm": 0.961555540561676, "learning_rate": 7.757999495033623e-05, "loss": 0.1049, "step": 20900 }, { "grad_norm": 0.6030517220497131, "learning_rate": 7.755700453269456e-05, "loss": 0.0566, "step": 20910 }, { "grad_norm": 0.7312986254692078, "learning_rate": 7.753400574396254e-05, "loss": 0.0702, "step": 20920 }, { "grad_norm": 0.5318598747253418, "learning_rate": 7.751099859112655e-05, "loss": 0.0566, "step": 20930 }, { "grad_norm": 0.5575147867202759, "learning_rate": 7.748798308117557e-05, "loss": 0.0622, "step": 20940 }, { "grad_norm": 0.35770612955093384, "learning_rate": 7.746495922110112e-05, "loss": 0.0602, "step": 20950 }, { "grad_norm": 0.7593085765838623, "learning_rate": 7.744192701789723e-05, "loss": 0.0638, "step": 20960 }, { "grad_norm": 0.4585343897342682, "learning_rate": 7.741888647856046e-05, "loss": 0.0506, "step": 20970 }, { "grad_norm": 0.4732387661933899, "learning_rate": 7.739583761008994e-05, "loss": 0.0576, "step": 20980 }, { "grad_norm": 0.6933939456939697, "learning_rate": 7.73727804194873e-05, "loss": 0.0565, "step": 20990 }, { "grad_norm": 0.4018535315990448, "learning_rate": 7.734971491375671e-05, "loss": 0.0578, "step": 21000 }, { "grad_norm": 0.696403980255127, "learning_rate": 7.732664109990485e-05, "loss": 0.0585, "step": 21010 }, { "grad_norm": 0.6325119733810425, "learning_rate": 7.730355898494095e-05, "loss": 0.0899, "step": 21020 }, { "grad_norm": 0.6901130080223083, "learning_rate": 7.728046857587673e-05, "loss": 0.0658, "step": 21030 }, { "grad_norm": 0.4403780400753021, "learning_rate": 7.725736987972647e-05, "loss": 0.0573, "step": 21040 }, { "grad_norm": 0.5555588006973267, "learning_rate": 7.723426290350691e-05, "loss": 0.0792, "step": 21050 }, { "grad_norm": 0.5893208980560303, "learning_rate": 7.721114765423736e-05, "loss": 0.0628, "step": 21060 }, { "grad_norm": 0.44194120168685913, "learning_rate": 7.718802413893963e-05, "loss": 0.0636, "step": 21070 }, { "grad_norm": 0.5001569986343384, "learning_rate": 7.716489236463802e-05, "loss": 0.0531, "step": 21080 }, { "grad_norm": 1.0765501260757446, "learning_rate": 7.714175233835936e-05, "loss": 0.0572, "step": 21090 }, { "grad_norm": 0.8363808989524841, "learning_rate": 7.711860406713299e-05, "loss": 0.0488, "step": 21100 }, { "grad_norm": 0.45851877331733704, "learning_rate": 7.70954475579907e-05, "loss": 0.0526, "step": 21110 }, { "grad_norm": 0.8536600470542908, "learning_rate": 7.707228281796688e-05, "loss": 0.06, "step": 21120 }, { "grad_norm": 0.511414647102356, "learning_rate": 7.704910985409833e-05, "loss": 0.069, "step": 21130 }, { "grad_norm": 0.46532920002937317, "learning_rate": 7.702592867342439e-05, "loss": 0.0534, "step": 21140 }, { "grad_norm": 0.5052671432495117, "learning_rate": 7.700273928298691e-05, "loss": 0.0533, "step": 21150 }, { "grad_norm": 0.39087581634521484, "learning_rate": 7.697954168983021e-05, "loss": 0.0603, "step": 21160 }, { "grad_norm": 0.46330881118774414, "learning_rate": 7.695633590100109e-05, "loss": 0.066, "step": 21170 }, { "grad_norm": 0.6162739992141724, "learning_rate": 7.693312192354886e-05, "loss": 0.0638, "step": 21180 }, { "grad_norm": 0.809948742389679, "learning_rate": 7.690989976452532e-05, "loss": 0.0862, "step": 21190 }, { "grad_norm": 0.48015296459198, "learning_rate": 7.688666943098475e-05, "loss": 0.0736, "step": 21200 }, { "grad_norm": 0.5047924518585205, "learning_rate": 7.686343092998389e-05, "loss": 0.0967, "step": 21210 }, { "grad_norm": 0.5895740389823914, "learning_rate": 7.684018426858202e-05, "loss": 0.0675, "step": 21220 }, { "grad_norm": 0.7608733177185059, "learning_rate": 7.681692945384084e-05, "loss": 0.0878, "step": 21230 }, { "grad_norm": 0.8969094753265381, "learning_rate": 7.679366649282456e-05, "loss": 0.0736, "step": 21240 }, { "grad_norm": 2.2248101234436035, "learning_rate": 7.677039539259983e-05, "loss": 0.0538, "step": 21250 }, { "grad_norm": 0.5092384219169617, "learning_rate": 7.674711616023581e-05, "loss": 0.0522, "step": 21260 }, { "grad_norm": 0.5350058674812317, "learning_rate": 7.672382880280413e-05, "loss": 0.0602, "step": 21270 }, { "grad_norm": 0.6471090912818909, "learning_rate": 7.670053332737885e-05, "loss": 0.0725, "step": 21280 }, { "grad_norm": 0.4601537883281708, "learning_rate": 7.667722974103654e-05, "loss": 0.0737, "step": 21290 }, { "grad_norm": 0.5123391151428223, "learning_rate": 7.66539180508562e-05, "loss": 0.0558, "step": 21300 }, { "grad_norm": 1.346625804901123, "learning_rate": 7.663059826391932e-05, "loss": 0.091, "step": 21310 }, { "grad_norm": 0.7422065138816833, "learning_rate": 7.660727038730981e-05, "loss": 0.0618, "step": 21320 }, { "grad_norm": 0.823847234249115, "learning_rate": 7.65839344281141e-05, "loss": 0.0728, "step": 21330 }, { "grad_norm": 0.6151600480079651, "learning_rate": 7.656059039342101e-05, "loss": 0.0664, "step": 21340 }, { "grad_norm": 1.2082171440124512, "learning_rate": 7.653723829032187e-05, "loss": 0.0584, "step": 21350 }, { "grad_norm": 0.5234292149543762, "learning_rate": 7.65138781259104e-05, "loss": 0.059, "step": 21360 }, { "grad_norm": 0.5280361175537109, "learning_rate": 7.649050990728279e-05, "loss": 0.0614, "step": 21370 }, { "grad_norm": 0.6193971633911133, "learning_rate": 7.646713364153774e-05, "loss": 0.0455, "step": 21380 }, { "grad_norm": 0.745047390460968, "learning_rate": 7.64437493357763e-05, "loss": 0.0565, "step": 21390 }, { "grad_norm": 0.6174812912940979, "learning_rate": 7.642035699710202e-05, "loss": 0.1099, "step": 21400 }, { "grad_norm": 0.49863025546073914, "learning_rate": 7.639695663262089e-05, "loss": 0.0755, "step": 21410 }, { "grad_norm": 0.6252360939979553, "learning_rate": 7.637354824944128e-05, "loss": 0.0608, "step": 21420 }, { "grad_norm": 0.5703227519989014, "learning_rate": 7.635013185467408e-05, "loss": 0.0675, "step": 21430 }, { "grad_norm": 0.6604714393615723, "learning_rate": 7.632670745543256e-05, "loss": 0.0619, "step": 21440 }, { "grad_norm": 0.6179224848747253, "learning_rate": 7.630327505883242e-05, "loss": 0.0701, "step": 21450 }, { "grad_norm": 0.5444523692131042, "learning_rate": 7.627983467199182e-05, "loss": 0.0505, "step": 21460 }, { "grad_norm": 0.7161325812339783, "learning_rate": 7.625638630203132e-05, "loss": 0.0648, "step": 21470 }, { "grad_norm": 0.7694377899169922, "learning_rate": 7.623292995607394e-05, "loss": 0.0543, "step": 21480 }, { "grad_norm": 0.7086231708526611, "learning_rate": 7.620946564124507e-05, "loss": 0.0559, "step": 21490 }, { "grad_norm": 0.44977498054504395, "learning_rate": 7.618599336467256e-05, "loss": 0.0545, "step": 21500 }, { "grad_norm": 0.5079180598258972, "learning_rate": 7.616251313348666e-05, "loss": 0.0692, "step": 21510 }, { "grad_norm": 0.5396995544433594, "learning_rate": 7.613902495482005e-05, "loss": 0.0523, "step": 21520 }, { "grad_norm": 0.42701515555381775, "learning_rate": 7.611552883580784e-05, "loss": 0.0546, "step": 21530 }, { "grad_norm": 0.6835952401161194, "learning_rate": 7.609202478358748e-05, "loss": 0.0729, "step": 21540 }, { "grad_norm": 0.5871222615242004, "learning_rate": 7.606851280529895e-05, "loss": 0.0691, "step": 21550 }, { "grad_norm": 0.6436841487884521, "learning_rate": 7.604499290808449e-05, "loss": 0.0583, "step": 21560 }, { "grad_norm": 0.5134782195091248, "learning_rate": 7.602146509908888e-05, "loss": 0.0539, "step": 21570 }, { "grad_norm": 0.6337180733680725, "learning_rate": 7.599792938545921e-05, "loss": 0.0646, "step": 21580 }, { "grad_norm": 0.6117178797721863, "learning_rate": 7.597438577434506e-05, "loss": 0.0905, "step": 21590 }, { "grad_norm": 0.5419978499412537, "learning_rate": 7.595083427289831e-05, "loss": 0.0572, "step": 21600 }, { "grad_norm": 0.6587945818901062, "learning_rate": 7.59272748882733e-05, "loss": 0.0543, "step": 21610 }, { "grad_norm": 0.7564509510993958, "learning_rate": 7.590370762762675e-05, "loss": 0.0707, "step": 21620 }, { "grad_norm": 0.6786878705024719, "learning_rate": 7.588013249811777e-05, "loss": 0.0706, "step": 21630 }, { "grad_norm": 0.5269999504089355, "learning_rate": 7.585654950690786e-05, "loss": 0.0669, "step": 21640 }, { "grad_norm": 0.4214574098587036, "learning_rate": 7.583295866116091e-05, "loss": 0.0697, "step": 21650 }, { "grad_norm": 0.615185022354126, "learning_rate": 7.580935996804321e-05, "loss": 0.0705, "step": 21660 }, { "grad_norm": 0.8362227082252502, "learning_rate": 7.57857534347234e-05, "loss": 0.0793, "step": 21670 }, { "grad_norm": 0.5807210803031921, "learning_rate": 7.576213906837254e-05, "loss": 0.0496, "step": 21680 }, { "grad_norm": 0.3588697016239166, "learning_rate": 7.573851687616403e-05, "loss": 0.0572, "step": 21690 }, { "grad_norm": 0.6698682308197021, "learning_rate": 7.571488686527368e-05, "loss": 0.0545, "step": 21700 }, { "grad_norm": 0.44200459122657776, "learning_rate": 7.569124904287968e-05, "loss": 0.0467, "step": 21710 }, { "grad_norm": 0.5566723942756653, "learning_rate": 7.566760341616254e-05, "loss": 0.0466, "step": 21720 }, { "grad_norm": 0.4567437171936035, "learning_rate": 7.564394999230519e-05, "loss": 0.0484, "step": 21730 }, { "grad_norm": 0.2979373037815094, "learning_rate": 7.562028877849294e-05, "loss": 0.0559, "step": 21740 }, { "grad_norm": 0.5117394328117371, "learning_rate": 7.559661978191341e-05, "loss": 0.0711, "step": 21750 }, { "grad_norm": 0.4928600490093231, "learning_rate": 7.557294300975664e-05, "loss": 0.0504, "step": 21760 }, { "grad_norm": 0.7672922015190125, "learning_rate": 7.554925846921499e-05, "loss": 0.0619, "step": 21770 }, { "grad_norm": 0.5811465978622437, "learning_rate": 7.552556616748321e-05, "loss": 0.0651, "step": 21780 }, { "grad_norm": 0.5485804080963135, "learning_rate": 7.550186611175838e-05, "loss": 0.0749, "step": 21790 }, { "grad_norm": 0.44895726442337036, "learning_rate": 7.547815830923998e-05, "loss": 0.0533, "step": 21800 }, { "grad_norm": 0.622140109539032, "learning_rate": 7.54544427671298e-05, "loss": 0.0628, "step": 21810 }, { "grad_norm": 0.6589682102203369, "learning_rate": 7.543071949263198e-05, "loss": 0.0482, "step": 21820 }, { "grad_norm": 1.3402364253997803, "learning_rate": 7.540698849295305e-05, "loss": 0.0562, "step": 21830 }, { "grad_norm": 0.4765397906303406, "learning_rate": 7.538324977530183e-05, "loss": 0.0616, "step": 21840 }, { "grad_norm": 0.5997915267944336, "learning_rate": 7.535950334688955e-05, "loss": 0.0673, "step": 21850 }, { "grad_norm": 0.9118224382400513, "learning_rate": 7.533574921492972e-05, "loss": 0.0707, "step": 21860 }, { "grad_norm": 0.7229516506195068, "learning_rate": 7.531198738663824e-05, "loss": 0.0647, "step": 21870 }, { "grad_norm": 0.571185827255249, "learning_rate": 7.528821786923333e-05, "loss": 0.086, "step": 21880 }, { "grad_norm": 0.7283005118370056, "learning_rate": 7.52644406699355e-05, "loss": 0.0638, "step": 21890 }, { "grad_norm": 0.6562538743019104, "learning_rate": 7.524065579596766e-05, "loss": 0.0624, "step": 21900 }, { "grad_norm": 0.722911536693573, "learning_rate": 7.521686325455506e-05, "loss": 0.0637, "step": 21910 }, { "grad_norm": 0.8511524200439453, "learning_rate": 7.51930630529252e-05, "loss": 0.0523, "step": 21920 }, { "grad_norm": 0.6022292971611023, "learning_rate": 7.516925519830797e-05, "loss": 0.0437, "step": 21930 }, { "grad_norm": 0.6224909424781799, "learning_rate": 7.514543969793557e-05, "loss": 0.0451, "step": 21940 }, { "grad_norm": 0.6935185194015503, "learning_rate": 7.512161655904251e-05, "loss": 0.0598, "step": 21950 }, { "grad_norm": 0.4341409504413605, "learning_rate": 7.509778578886563e-05, "loss": 0.0428, "step": 21960 }, { "grad_norm": 0.618579626083374, "learning_rate": 7.507394739464412e-05, "loss": 0.0641, "step": 21970 }, { "grad_norm": 0.47974008321762085, "learning_rate": 7.50501013836194e-05, "loss": 0.0598, "step": 21980 }, { "grad_norm": 0.5133688449859619, "learning_rate": 7.50262477630353e-05, "loss": 0.0495, "step": 21990 }, { "grad_norm": 0.6017083525657654, "learning_rate": 7.500238654013794e-05, "loss": 0.0674, "step": 22000 }, { "grad_norm": 0.48420244455337524, "learning_rate": 7.497851772217566e-05, "loss": 0.0668, "step": 22010 }, { "grad_norm": 1.075354814529419, "learning_rate": 7.495464131639924e-05, "loss": 0.0634, "step": 22020 }, { "grad_norm": 0.5185866951942444, "learning_rate": 7.493075733006166e-05, "loss": 0.0754, "step": 22030 }, { "grad_norm": 0.6124272346496582, "learning_rate": 7.490686577041828e-05, "loss": 0.0588, "step": 22040 }, { "grad_norm": 0.739165186882019, "learning_rate": 7.488296664472668e-05, "loss": 0.0615, "step": 22050 }, { "grad_norm": 0.7199192643165588, "learning_rate": 7.485905996024682e-05, "loss": 0.0634, "step": 22060 }, { "grad_norm": 0.6482220888137817, "learning_rate": 7.483514572424093e-05, "loss": 0.0562, "step": 22070 }, { "grad_norm": 0.7920705080032349, "learning_rate": 7.481122394397349e-05, "loss": 0.0668, "step": 22080 }, { "grad_norm": 0.6245090961456299, "learning_rate": 7.478729462671131e-05, "loss": 0.0574, "step": 22090 }, { "grad_norm": 0.6638580560684204, "learning_rate": 7.47633577797235e-05, "loss": 0.0562, "step": 22100 }, { "grad_norm": 0.9033961296081543, "learning_rate": 7.473941341028144e-05, "loss": 0.059, "step": 22110 }, { "grad_norm": 0.7820343375205994, "learning_rate": 7.471546152565879e-05, "loss": 0.0736, "step": 22120 }, { "grad_norm": 0.5199718475341797, "learning_rate": 7.46915021331315e-05, "loss": 0.0469, "step": 22130 }, { "grad_norm": 0.44470852613449097, "learning_rate": 7.466753523997778e-05, "loss": 0.0462, "step": 22140 }, { "grad_norm": 0.5695542693138123, "learning_rate": 7.464356085347819e-05, "loss": 0.0569, "step": 22150 }, { "grad_norm": 0.5201715230941772, "learning_rate": 7.461957898091548e-05, "loss": 0.0681, "step": 22160 }, { "grad_norm": 0.5096570253372192, "learning_rate": 7.459558962957473e-05, "loss": 0.0463, "step": 22170 }, { "grad_norm": 0.5711193084716797, "learning_rate": 7.457159280674326e-05, "loss": 0.0623, "step": 22180 }, { "grad_norm": 0.47156932950019836, "learning_rate": 7.454758851971066e-05, "loss": 0.0537, "step": 22190 }, { "grad_norm": 0.6078714728355408, "learning_rate": 7.45235767757688e-05, "loss": 0.0704, "step": 22200 }, { "grad_norm": 0.44604313373565674, "learning_rate": 7.449955758221183e-05, "loss": 0.0457, "step": 22210 }, { "grad_norm": 0.5499972701072693, "learning_rate": 7.447553094633615e-05, "loss": 0.075, "step": 22220 }, { "grad_norm": 0.6310759782791138, "learning_rate": 7.445149687544039e-05, "loss": 0.0573, "step": 22230 }, { "grad_norm": 0.46254122257232666, "learning_rate": 7.44274553768255e-05, "loss": 0.0624, "step": 22240 }, { "grad_norm": 0.4614972174167633, "learning_rate": 7.440340645779464e-05, "loss": 0.0769, "step": 22250 }, { "grad_norm": 0.5663689970970154, "learning_rate": 7.437935012565322e-05, "loss": 0.0699, "step": 22260 }, { "grad_norm": 0.5812203884124756, "learning_rate": 7.435528638770893e-05, "loss": 0.0658, "step": 22270 }, { "grad_norm": 0.541157066822052, "learning_rate": 7.433121525127171e-05, "loss": 0.0642, "step": 22280 }, { "grad_norm": 0.5497801303863525, "learning_rate": 7.430713672365371e-05, "loss": 0.0487, "step": 22290 }, { "grad_norm": 0.4277400076389313, "learning_rate": 7.428305081216938e-05, "loss": 0.0515, "step": 22300 }, { "grad_norm": 0.5659223794937134, "learning_rate": 7.425895752413536e-05, "loss": 0.0539, "step": 22310 }, { "grad_norm": 0.6132288575172424, "learning_rate": 7.423485686687057e-05, "loss": 0.055, "step": 22320 }, { "grad_norm": 0.4747675359249115, "learning_rate": 7.421074884769616e-05, "loss": 0.0623, "step": 22330 }, { "grad_norm": 0.5667241215705872, "learning_rate": 7.418663347393548e-05, "loss": 0.0592, "step": 22340 }, { "grad_norm": 1.070845603942871, "learning_rate": 7.416251075291418e-05, "loss": 0.076, "step": 22350 }, { "grad_norm": 0.5942496061325073, "learning_rate": 7.413838069196007e-05, "loss": 0.0486, "step": 22360 }, { "grad_norm": 0.6100648641586304, "learning_rate": 7.411424329840324e-05, "loss": 0.0737, "step": 22370 }, { "grad_norm": 0.4567524492740631, "learning_rate": 7.409009857957601e-05, "loss": 0.0562, "step": 22380 }, { "grad_norm": 0.5584650039672852, "learning_rate": 7.40659465428129e-05, "loss": 0.0654, "step": 22390 }, { "grad_norm": 0.6291924715042114, "learning_rate": 7.404178719545063e-05, "loss": 0.0555, "step": 22400 }, { "grad_norm": 0.5501847267150879, "learning_rate": 7.401762054482822e-05, "loss": 0.0593, "step": 22410 }, { "grad_norm": 0.7451338171958923, "learning_rate": 7.39934465982868e-05, "loss": 0.0707, "step": 22420 }, { "grad_norm": 0.6617288589477539, "learning_rate": 7.396926536316984e-05, "loss": 0.0464, "step": 22430 }, { "grad_norm": 0.7332164645195007, "learning_rate": 7.394507684682293e-05, "loss": 0.062, "step": 22440 }, { "grad_norm": 0.6260108947753906, "learning_rate": 7.392088105659393e-05, "loss": 0.0786, "step": 22450 }, { "grad_norm": 0.576114296913147, "learning_rate": 7.389667799983284e-05, "loss": 0.0522, "step": 22460 }, { "grad_norm": 0.6154944896697998, "learning_rate": 7.387246768389193e-05, "loss": 0.0639, "step": 22470 }, { "grad_norm": 0.4755021035671234, "learning_rate": 7.384825011612563e-05, "loss": 0.0521, "step": 22480 }, { "grad_norm": 0.6971496939659119, "learning_rate": 7.382402530389066e-05, "loss": 0.0614, "step": 22490 }, { "grad_norm": 0.7233211994171143, "learning_rate": 7.379979325454582e-05, "loss": 0.0694, "step": 22500 }, { "grad_norm": 0.6242825388908386, "learning_rate": 7.37755539754522e-05, "loss": 0.0497, "step": 22510 }, { "grad_norm": 0.8691065907478333, "learning_rate": 7.375130747397302e-05, "loss": 0.0722, "step": 22520 }, { "grad_norm": 0.5622522830963135, "learning_rate": 7.372705375747377e-05, "loss": 0.0752, "step": 22530 }, { "grad_norm": 0.40388238430023193, "learning_rate": 7.370279283332205e-05, "loss": 0.06, "step": 22540 }, { "grad_norm": 0.4409674406051636, "learning_rate": 7.36785247088877e-05, "loss": 0.0623, "step": 22550 }, { "grad_norm": 0.8743299841880798, "learning_rate": 7.365424939154275e-05, "loss": 0.104, "step": 22560 }, { "grad_norm": 0.9204819202423096, "learning_rate": 7.362996688866138e-05, "loss": 0.0684, "step": 22570 }, { "grad_norm": 0.7912854552268982, "learning_rate": 7.360567720761999e-05, "loss": 0.0504, "step": 22580 }, { "grad_norm": 0.5970836877822876, "learning_rate": 7.358138035579711e-05, "loss": 0.0758, "step": 22590 }, { "grad_norm": 0.4361727833747864, "learning_rate": 7.355707634057354e-05, "loss": 0.0563, "step": 22600 }, { "grad_norm": 0.7367427349090576, "learning_rate": 7.353276516933215e-05, "loss": 0.0671, "step": 22610 }, { "grad_norm": 0.5984386205673218, "learning_rate": 7.350844684945806e-05, "loss": 0.0616, "step": 22620 }, { "grad_norm": 0.6879570484161377, "learning_rate": 7.348412138833851e-05, "loss": 0.055, "step": 22630 }, { "grad_norm": 0.633961021900177, "learning_rate": 7.345978879336295e-05, "loss": 0.054, "step": 22640 }, { "grad_norm": 0.544555127620697, "learning_rate": 7.343544907192296e-05, "loss": 0.0486, "step": 22650 }, { "grad_norm": 0.6522403359413147, "learning_rate": 7.341110223141235e-05, "loss": 0.0622, "step": 22660 }, { "grad_norm": 0.5448827147483826, "learning_rate": 7.3386748279227e-05, "loss": 0.0539, "step": 22670 }, { "grad_norm": 1.2335478067398071, "learning_rate": 7.336238722276501e-05, "loss": 0.0631, "step": 22680 }, { "grad_norm": 0.461569607257843, "learning_rate": 7.333801906942663e-05, "loss": 0.076, "step": 22690 }, { "grad_norm": 0.4746069610118866, "learning_rate": 7.331364382661428e-05, "loss": 0.0777, "step": 22700 }, { "grad_norm": 0.708051323890686, "learning_rate": 7.328926150173248e-05, "loss": 0.0569, "step": 22710 }, { "grad_norm": 0.6057626008987427, "learning_rate": 7.326487210218795e-05, "loss": 0.0519, "step": 22720 }, { "grad_norm": 0.53667151927948, "learning_rate": 7.324047563538955e-05, "loss": 0.0523, "step": 22730 }, { "grad_norm": 0.5430928468704224, "learning_rate": 7.321607210874828e-05, "loss": 0.0406, "step": 22740 }, { "grad_norm": 0.6292001605033875, "learning_rate": 7.31916615296773e-05, "loss": 0.0591, "step": 22750 }, { "grad_norm": 0.4401807487010956, "learning_rate": 7.316724390559188e-05, "loss": 0.0583, "step": 22760 }, { "grad_norm": 0.4050765633583069, "learning_rate": 7.314281924390946e-05, "loss": 0.0598, "step": 22770 }, { "grad_norm": 0.5324440598487854, "learning_rate": 7.311838755204959e-05, "loss": 0.051, "step": 22780 }, { "grad_norm": 0.8713562488555908, "learning_rate": 7.3093948837434e-05, "loss": 0.078, "step": 22790 }, { "grad_norm": 0.5382103323936462, "learning_rate": 7.306950310748651e-05, "loss": 0.0699, "step": 22800 }, { "grad_norm": 0.5336849689483643, "learning_rate": 7.304505036963311e-05, "loss": 0.0427, "step": 22810 }, { "grad_norm": 0.49037986993789673, "learning_rate": 7.302059063130186e-05, "loss": 0.066, "step": 22820 }, { "grad_norm": 0.5710689425468445, "learning_rate": 7.2996123899923e-05, "loss": 0.0828, "step": 22830 }, { "grad_norm": 0.5461347103118896, "learning_rate": 7.297165018292886e-05, "loss": 0.0546, "step": 22840 }, { "grad_norm": 0.5008642077445984, "learning_rate": 7.294716948775396e-05, "loss": 0.046, "step": 22850 }, { "grad_norm": 0.5083521604537964, "learning_rate": 7.292268182183484e-05, "loss": 0.0408, "step": 22860 }, { "grad_norm": 0.4168052077293396, "learning_rate": 7.28981871926102e-05, "loss": 0.0526, "step": 22870 }, { "grad_norm": 0.7892974019050598, "learning_rate": 7.28736856075209e-05, "loss": 0.0572, "step": 22880 }, { "grad_norm": 0.7438310384750366, "learning_rate": 7.284917707400985e-05, "loss": 0.0469, "step": 22890 }, { "grad_norm": 0.5782968997955322, "learning_rate": 7.282466159952212e-05, "loss": 0.0593, "step": 22900 }, { "grad_norm": 0.6121670603752136, "learning_rate": 7.280013919150483e-05, "loss": 0.0836, "step": 22910 }, { "grad_norm": 0.5560021996498108, "learning_rate": 7.277560985740728e-05, "loss": 0.058, "step": 22920 }, { "grad_norm": 0.5192553400993347, "learning_rate": 7.275107360468079e-05, "loss": 0.0498, "step": 22930 }, { "grad_norm": 0.6599103212356567, "learning_rate": 7.272653044077885e-05, "loss": 0.069, "step": 22940 }, { "grad_norm": 0.5535059571266174, "learning_rate": 7.270198037315703e-05, "loss": 0.0497, "step": 22950 }, { "grad_norm": 0.4799918830394745, "learning_rate": 7.267742340927297e-05, "loss": 0.0525, "step": 22960 }, { "grad_norm": 0.5990359783172607, "learning_rate": 7.265285955658645e-05, "loss": 0.0591, "step": 22970 }, { "grad_norm": 0.7672076225280762, "learning_rate": 7.26282888225593e-05, "loss": 0.0771, "step": 22980 }, { "grad_norm": 0.5284526348114014, "learning_rate": 7.260371121465548e-05, "loss": 0.0458, "step": 22990 }, { "grad_norm": 0.5512226223945618, "learning_rate": 7.2579126740341e-05, "loss": 0.0617, "step": 23000 }, { "grad_norm": 0.5009942054748535, "learning_rate": 7.2554535407084e-05, "loss": 0.0525, "step": 23010 }, { "grad_norm": 0.4977067708969116, "learning_rate": 7.252993722235464e-05, "loss": 0.0585, "step": 23020 }, { "grad_norm": 0.4272769093513489, "learning_rate": 7.250533219362523e-05, "loss": 0.0596, "step": 23030 }, { "grad_norm": 0.5230686068534851, "learning_rate": 7.248072032837012e-05, "loss": 0.0598, "step": 23040 }, { "grad_norm": 0.748413622379303, "learning_rate": 7.245610163406575e-05, "loss": 0.0531, "step": 23050 }, { "grad_norm": 0.49253788590431213, "learning_rate": 7.243147611819061e-05, "loss": 0.0589, "step": 23060 }, { "grad_norm": 0.5531520843505859, "learning_rate": 7.240684378822531e-05, "loss": 0.0564, "step": 23070 }, { "grad_norm": 1.0401819944381714, "learning_rate": 7.238220465165248e-05, "loss": 0.0618, "step": 23080 }, { "grad_norm": 0.6023280024528503, "learning_rate": 7.235755871595684e-05, "loss": 0.0634, "step": 23090 }, { "grad_norm": 0.6016688942909241, "learning_rate": 7.233290598862517e-05, "loss": 0.0599, "step": 23100 }, { "grad_norm": 0.7289943099021912, "learning_rate": 7.230824647714635e-05, "loss": 0.0644, "step": 23110 }, { "grad_norm": 0.41173407435417175, "learning_rate": 7.228358018901124e-05, "loss": 0.0496, "step": 23120 }, { "grad_norm": 0.48549631237983704, "learning_rate": 7.225890713171286e-05, "loss": 0.0666, "step": 23130 }, { "grad_norm": 0.6523394584655762, "learning_rate": 7.223422731274618e-05, "loss": 0.0578, "step": 23140 }, { "grad_norm": 0.7873654961585999, "learning_rate": 7.220954073960832e-05, "loss": 0.0665, "step": 23150 }, { "grad_norm": 0.7277244925498962, "learning_rate": 7.218484741979838e-05, "loss": 0.0704, "step": 23160 }, { "grad_norm": 0.6773493885993958, "learning_rate": 7.216014736081756e-05, "loss": 0.0619, "step": 23170 }, { "grad_norm": 0.5856652855873108, "learning_rate": 7.213544057016906e-05, "loss": 0.0641, "step": 23180 }, { "grad_norm": 0.5316565036773682, "learning_rate": 7.211072705535819e-05, "loss": 0.0557, "step": 23190 }, { "grad_norm": 0.6943445205688477, "learning_rate": 7.208600682389224e-05, "loss": 0.0582, "step": 23200 }, { "grad_norm": 0.6145713329315186, "learning_rate": 7.206127988328055e-05, "loss": 0.0858, "step": 23210 }, { "grad_norm": 0.7975818514823914, "learning_rate": 7.203654624103453e-05, "loss": 0.0735, "step": 23220 }, { "grad_norm": 0.6933742165565491, "learning_rate": 7.201180590466761e-05, "loss": 0.0701, "step": 23230 }, { "grad_norm": 0.822693943977356, "learning_rate": 7.198705888169523e-05, "loss": 0.0666, "step": 23240 }, { "grad_norm": 0.5908656716346741, "learning_rate": 7.196230517963491e-05, "loss": 0.0663, "step": 23250 }, { "grad_norm": 0.5700899958610535, "learning_rate": 7.193754480600615e-05, "loss": 0.0903, "step": 23260 }, { "grad_norm": 0.5663282871246338, "learning_rate": 7.19127777683305e-05, "loss": 0.0616, "step": 23270 }, { "grad_norm": 0.7291322946548462, "learning_rate": 7.188800407413156e-05, "loss": 0.0725, "step": 23280 }, { "grad_norm": 0.6335467100143433, "learning_rate": 7.186322373093489e-05, "loss": 0.0536, "step": 23290 }, { "grad_norm": 0.5177226662635803, "learning_rate": 7.18384367462681e-05, "loss": 0.0549, "step": 23300 }, { "grad_norm": 0.5797964334487915, "learning_rate": 7.181364312766085e-05, "loss": 0.0511, "step": 23310 }, { "grad_norm": 0.4681622087955475, "learning_rate": 7.178884288264477e-05, "loss": 0.0571, "step": 23320 }, { "grad_norm": 0.5806000232696533, "learning_rate": 7.176403601875353e-05, "loss": 0.0556, "step": 23330 }, { "grad_norm": 0.612890899181366, "learning_rate": 7.173922254352279e-05, "loss": 0.0443, "step": 23340 }, { "grad_norm": 0.6950223445892334, "learning_rate": 7.171440246449024e-05, "loss": 0.0624, "step": 23350 }, { "grad_norm": 0.42768287658691406, "learning_rate": 7.168957578919555e-05, "loss": 0.053, "step": 23360 }, { "grad_norm": 0.515734851360321, "learning_rate": 7.16647425251804e-05, "loss": 0.0542, "step": 23370 }, { "grad_norm": 0.5062239766120911, "learning_rate": 7.163990267998852e-05, "loss": 0.0648, "step": 23380 }, { "grad_norm": 0.5049260854721069, "learning_rate": 7.161505626116556e-05, "loss": 0.0636, "step": 23390 }, { "grad_norm": 0.725287914276123, "learning_rate": 7.159020327625923e-05, "loss": 0.0503, "step": 23400 }, { "grad_norm": 0.6656132936477661, "learning_rate": 7.15653437328192e-05, "loss": 0.0589, "step": 23410 }, { "grad_norm": 0.579143762588501, "learning_rate": 7.154047763839713e-05, "loss": 0.0496, "step": 23420 }, { "grad_norm": 0.5399107336997986, "learning_rate": 7.15156050005467e-05, "loss": 0.0758, "step": 23430 }, { "grad_norm": 0.6511619687080383, "learning_rate": 7.149072582682357e-05, "loss": 0.0606, "step": 23440 }, { "grad_norm": 0.5054323077201843, "learning_rate": 7.146584012478535e-05, "loss": 0.0486, "step": 23450 }, { "grad_norm": 0.511099636554718, "learning_rate": 7.144094790199169e-05, "loss": 0.0525, "step": 23460 }, { "grad_norm": 0.5293551087379456, "learning_rate": 7.141604916600415e-05, "loss": 0.0548, "step": 23470 }, { "grad_norm": 0.5884762406349182, "learning_rate": 7.139114392438635e-05, "loss": 0.0535, "step": 23480 }, { "grad_norm": 0.8295354247093201, "learning_rate": 7.136623218470382e-05, "loss": 0.0531, "step": 23490 }, { "grad_norm": 0.6178086996078491, "learning_rate": 7.13413139545241e-05, "loss": 0.0746, "step": 23500 }, { "grad_norm": 0.44516444206237793, "learning_rate": 7.131638924141668e-05, "loss": 0.0516, "step": 23510 }, { "grad_norm": 0.32751142978668213, "learning_rate": 7.129145805295304e-05, "loss": 0.047, "step": 23520 }, { "grad_norm": 0.6766202449798584, "learning_rate": 7.126652039670661e-05, "loss": 0.0618, "step": 23530 }, { "grad_norm": 0.5749101638793945, "learning_rate": 7.124157628025278e-05, "loss": 0.0453, "step": 23540 }, { "grad_norm": 0.5106420516967773, "learning_rate": 7.121662571116894e-05, "loss": 0.0409, "step": 23550 }, { "grad_norm": 0.7380993366241455, "learning_rate": 7.119166869703441e-05, "loss": 0.0488, "step": 23560 }, { "grad_norm": 1.0580840110778809, "learning_rate": 7.116670524543044e-05, "loss": 0.0527, "step": 23570 }, { "grad_norm": 0.8132922649383545, "learning_rate": 7.114173536394032e-05, "loss": 0.0576, "step": 23580 }, { "grad_norm": 0.7815749645233154, "learning_rate": 7.111675906014917e-05, "loss": 0.0534, "step": 23590 }, { "grad_norm": 0.5981679558753967, "learning_rate": 7.109177634164421e-05, "loss": 0.0624, "step": 23600 }, { "grad_norm": 0.6269130110740662, "learning_rate": 7.106678721601449e-05, "loss": 0.0812, "step": 23610 }, { "grad_norm": 0.5284999012947083, "learning_rate": 7.104179169085103e-05, "loss": 0.0594, "step": 23620 }, { "grad_norm": 0.5614046454429626, "learning_rate": 7.101678977374683e-05, "loss": 0.0555, "step": 23630 }, { "grad_norm": 0.4941069185733795, "learning_rate": 7.099178147229685e-05, "loss": 0.0512, "step": 23640 }, { "grad_norm": 0.5853854417800903, "learning_rate": 7.096676679409789e-05, "loss": 0.064, "step": 23650 }, { "grad_norm": 0.6492719054222107, "learning_rate": 7.094174574674877e-05, "loss": 0.0889, "step": 23660 }, { "grad_norm": 0.5051380395889282, "learning_rate": 7.091671833785025e-05, "loss": 0.0505, "step": 23670 }, { "grad_norm": 0.41975149512290955, "learning_rate": 7.089168457500493e-05, "loss": 0.0494, "step": 23680 }, { "grad_norm": 0.4954460859298706, "learning_rate": 7.086664446581747e-05, "loss": 0.0517, "step": 23690 }, { "grad_norm": 0.5253528952598572, "learning_rate": 7.084159801789438e-05, "loss": 0.0613, "step": 23700 }, { "grad_norm": 0.5908346176147461, "learning_rate": 7.081654523884411e-05, "loss": 0.0723, "step": 23710 }, { "grad_norm": 0.6540372371673584, "learning_rate": 7.0791486136277e-05, "loss": 0.0555, "step": 23720 }, { "grad_norm": 0.7219387888908386, "learning_rate": 7.07664207178054e-05, "loss": 0.0563, "step": 23730 }, { "grad_norm": 0.5400617718696594, "learning_rate": 7.074134899104345e-05, "loss": 0.0569, "step": 23740 }, { "grad_norm": 0.5123521685600281, "learning_rate": 7.071627096360735e-05, "loss": 0.0714, "step": 23750 }, { "grad_norm": 0.602677583694458, "learning_rate": 7.069118664311511e-05, "loss": 0.0479, "step": 23760 }, { "grad_norm": 0.6164265275001526, "learning_rate": 7.06660960371867e-05, "loss": 0.052, "step": 23770 }, { "grad_norm": 0.41990649700164795, "learning_rate": 7.064099915344396e-05, "loss": 0.0525, "step": 23780 }, { "grad_norm": 0.5815451741218567, "learning_rate": 7.061589599951066e-05, "loss": 0.0576, "step": 23790 }, { "grad_norm": 1.1809544563293457, "learning_rate": 7.05907865830125e-05, "loss": 0.0642, "step": 23800 }, { "grad_norm": 0.7189044952392578, "learning_rate": 7.056567091157703e-05, "loss": 0.0539, "step": 23810 }, { "grad_norm": 0.7840951681137085, "learning_rate": 7.054054899283375e-05, "loss": 0.0541, "step": 23820 }, { "grad_norm": 0.4645540714263916, "learning_rate": 7.051542083441403e-05, "loss": 0.0452, "step": 23830 }, { "grad_norm": 0.5143914222717285, "learning_rate": 7.049028644395113e-05, "loss": 0.056, "step": 23840 }, { "grad_norm": 0.4473858177661896, "learning_rate": 7.046514582908024e-05, "loss": 0.0522, "step": 23850 }, { "grad_norm": 0.4700939357280731, "learning_rate": 7.043999899743838e-05, "loss": 0.0677, "step": 23860 }, { "grad_norm": 0.8319818377494812, "learning_rate": 7.041484595666451e-05, "loss": 0.0522, "step": 23870 }, { "grad_norm": 0.5935612916946411, "learning_rate": 7.038968671439948e-05, "loss": 0.0574, "step": 23880 }, { "grad_norm": 0.46966513991355896, "learning_rate": 7.036452127828596e-05, "loss": 0.0483, "step": 23890 }, { "grad_norm": 0.36710870265960693, "learning_rate": 7.033934965596859e-05, "loss": 0.044, "step": 23900 }, { "grad_norm": 0.5158824324607849, "learning_rate": 7.031417185509381e-05, "loss": 0.0457, "step": 23910 }, { "grad_norm": 0.6124054789543152, "learning_rate": 7.028898788331e-05, "loss": 0.0438, "step": 23920 }, { "grad_norm": 0.8121569156646729, "learning_rate": 7.026379774826736e-05, "loss": 0.0564, "step": 23930 }, { "grad_norm": 0.2874642610549927, "learning_rate": 7.0238601457618e-05, "loss": 0.0545, "step": 23940 }, { "grad_norm": 0.6230431199073792, "learning_rate": 7.02133990190159e-05, "loss": 0.0541, "step": 23950 }, { "grad_norm": 0.5302366614341736, "learning_rate": 7.018819044011687e-05, "loss": 0.0593, "step": 23960 }, { "grad_norm": 0.3814146816730499, "learning_rate": 7.016297572857863e-05, "loss": 0.0537, "step": 23970 }, { "grad_norm": 0.5078024864196777, "learning_rate": 7.013775489206072e-05, "loss": 0.0569, "step": 23980 }, { "grad_norm": 0.5555856227874756, "learning_rate": 7.01125279382246e-05, "loss": 0.068, "step": 23990 }, { "grad_norm": 0.7167142629623413, "learning_rate": 7.008729487473351e-05, "loss": 0.0661, "step": 24000 }, { "grad_norm": 0.7481728792190552, "learning_rate": 7.006205570925263e-05, "loss": 0.0579, "step": 24010 }, { "grad_norm": 0.7390414476394653, "learning_rate": 7.003681044944892e-05, "loss": 0.0612, "step": 24020 }, { "grad_norm": 0.6549589037895203, "learning_rate": 7.001155910299126e-05, "loss": 0.0534, "step": 24030 }, { "grad_norm": 0.42850568890571594, "learning_rate": 6.99863016775503e-05, "loss": 0.0518, "step": 24040 }, { "grad_norm": 0.41891658306121826, "learning_rate": 6.996103818079859e-05, "loss": 0.062, "step": 24050 }, { "grad_norm": 0.4556024670600891, "learning_rate": 6.993576862041054e-05, "loss": 0.0491, "step": 24060 }, { "grad_norm": 0.6450091004371643, "learning_rate": 6.991049300406235e-05, "loss": 0.0541, "step": 24070 }, { "grad_norm": 0.7948674559593201, "learning_rate": 6.988521133943209e-05, "loss": 0.0574, "step": 24080 }, { "grad_norm": 0.5657199621200562, "learning_rate": 6.985992363419966e-05, "loss": 0.0492, "step": 24090 }, { "grad_norm": 0.5626711249351501, "learning_rate": 6.983462989604682e-05, "loss": 0.0651, "step": 24100 }, { "grad_norm": 0.6662314534187317, "learning_rate": 6.980933013265709e-05, "loss": 0.055, "step": 24110 }, { "grad_norm": 0.447500616312027, "learning_rate": 6.978402435171592e-05, "loss": 0.0571, "step": 24120 }, { "grad_norm": 0.5572739839553833, "learning_rate": 6.975871256091052e-05, "loss": 0.0641, "step": 24130 }, { "grad_norm": 0.4873763918876648, "learning_rate": 6.973339476792995e-05, "loss": 0.0888, "step": 24140 }, { "grad_norm": 0.5260321497917175, "learning_rate": 6.970807098046505e-05, "loss": 0.0424, "step": 24150 }, { "grad_norm": 0.6365775465965271, "learning_rate": 6.968274120620858e-05, "loss": 0.0606, "step": 24160 }, { "grad_norm": 0.43027934432029724, "learning_rate": 6.965740545285499e-05, "loss": 0.0488, "step": 24170 }, { "grad_norm": 0.755134105682373, "learning_rate": 6.963206372810068e-05, "loss": 0.0549, "step": 24180 }, { "grad_norm": 0.6273596286773682, "learning_rate": 6.960671603964375e-05, "loss": 0.0532, "step": 24190 }, { "grad_norm": 0.9214679002761841, "learning_rate": 6.958136239518418e-05, "loss": 0.0722, "step": 24200 }, { "grad_norm": 0.6939124464988708, "learning_rate": 6.955600280242371e-05, "loss": 0.0628, "step": 24210 }, { "grad_norm": 0.4797099530696869, "learning_rate": 6.953063726906596e-05, "loss": 0.0546, "step": 24220 }, { "grad_norm": 0.6089573502540588, "learning_rate": 6.950526580281626e-05, "loss": 0.0475, "step": 24230 }, { "grad_norm": 0.5858430862426758, "learning_rate": 6.947988841138184e-05, "loss": 0.0436, "step": 24240 }, { "grad_norm": 0.7384172081947327, "learning_rate": 6.945450510247165e-05, "loss": 0.0555, "step": 24250 }, { "grad_norm": 0.47493863105773926, "learning_rate": 6.942911588379647e-05, "loss": 0.0566, "step": 24260 }, { "grad_norm": 0.5377970337867737, "learning_rate": 6.940372076306888e-05, "loss": 0.0536, "step": 24270 }, { "grad_norm": 0.6137767434120178, "learning_rate": 6.937831974800326e-05, "loss": 0.0525, "step": 24280 }, { "grad_norm": 0.5103579759597778, "learning_rate": 6.935291284631574e-05, "loss": 0.0454, "step": 24290 }, { "grad_norm": 0.6186220645904541, "learning_rate": 6.932750006572428e-05, "loss": 0.0575, "step": 24300 }, { "grad_norm": 0.38857126235961914, "learning_rate": 6.930208141394863e-05, "loss": 0.0516, "step": 24310 }, { "grad_norm": 0.7305081486701965, "learning_rate": 6.927665689871026e-05, "loss": 0.0587, "step": 24320 }, { "grad_norm": 0.6951887011528015, "learning_rate": 6.925122652773253e-05, "loss": 0.0479, "step": 24330 }, { "grad_norm": 0.5678209662437439, "learning_rate": 6.922579030874046e-05, "loss": 0.0636, "step": 24340 }, { "grad_norm": 0.3421013057231903, "learning_rate": 6.920034824946093e-05, "loss": 0.0516, "step": 24350 }, { "grad_norm": 0.7197185158729553, "learning_rate": 6.917490035762255e-05, "loss": 0.0804, "step": 24360 }, { "grad_norm": 0.4601966142654419, "learning_rate": 6.914944664095573e-05, "loss": 0.0448, "step": 24370 }, { "grad_norm": 0.5496897101402283, "learning_rate": 6.912398710719264e-05, "loss": 0.0575, "step": 24380 }, { "grad_norm": 0.4726774990558624, "learning_rate": 6.90985217640672e-05, "loss": 0.082, "step": 24390 }, { "grad_norm": 0.46996667981147766, "learning_rate": 6.90730506193151e-05, "loss": 0.0497, "step": 24400 }, { "grad_norm": 0.6684935092926025, "learning_rate": 6.904757368067384e-05, "loss": 0.0911, "step": 24410 }, { "grad_norm": 0.9196884632110596, "learning_rate": 6.90220909558826e-05, "loss": 0.0476, "step": 24420 }, { "grad_norm": 0.42894265055656433, "learning_rate": 6.899660245268237e-05, "loss": 0.0639, "step": 24430 }, { "grad_norm": 0.5802112817764282, "learning_rate": 6.897110817881592e-05, "loss": 0.0595, "step": 24440 }, { "grad_norm": 0.6172434091567993, "learning_rate": 6.894560814202769e-05, "loss": 0.0559, "step": 24450 }, { "grad_norm": 0.5008143782615662, "learning_rate": 6.892010235006394e-05, "loss": 0.0558, "step": 24460 }, { "grad_norm": 0.9409613609313965, "learning_rate": 6.889459081067264e-05, "loss": 0.0536, "step": 24470 }, { "grad_norm": 0.42016664147377014, "learning_rate": 6.886907353160356e-05, "loss": 0.0504, "step": 24480 }, { "grad_norm": 0.5653290152549744, "learning_rate": 6.884355052060814e-05, "loss": 0.0792, "step": 24490 }, { "grad_norm": 0.6034054160118103, "learning_rate": 6.88180217854396e-05, "loss": 0.0586, "step": 24500 }, { "grad_norm": 0.49284523725509644, "learning_rate": 6.87924873338529e-05, "loss": 0.0695, "step": 24510 }, { "grad_norm": 0.6137353777885437, "learning_rate": 6.876694717360475e-05, "loss": 0.0642, "step": 24520 }, { "grad_norm": 0.8912748098373413, "learning_rate": 6.874140131245355e-05, "loss": 0.0618, "step": 24530 }, { "grad_norm": 0.5038899183273315, "learning_rate": 6.871584975815948e-05, "loss": 0.0427, "step": 24540 }, { "grad_norm": 0.5360999703407288, "learning_rate": 6.86902925184844e-05, "loss": 0.0462, "step": 24550 }, { "grad_norm": 0.7507320642471313, "learning_rate": 6.866472960119195e-05, "loss": 0.0597, "step": 24560 }, { "grad_norm": 0.5579708218574524, "learning_rate": 6.863916101404748e-05, "loss": 0.0633, "step": 24570 }, { "grad_norm": 0.6383521556854248, "learning_rate": 6.8613586764818e-05, "loss": 0.0563, "step": 24580 }, { "grad_norm": 0.5031806826591492, "learning_rate": 6.858800686127233e-05, "loss": 0.0498, "step": 24590 }, { "grad_norm": 0.4940131604671478, "learning_rate": 6.856242131118097e-05, "loss": 0.0475, "step": 24600 }, { "grad_norm": 0.5977458357810974, "learning_rate": 6.853683012231614e-05, "loss": 0.054, "step": 24610 }, { "grad_norm": 0.37641364336013794, "learning_rate": 6.851123330245173e-05, "loss": 0.0491, "step": 24620 }, { "grad_norm": 0.5477995276451111, "learning_rate": 6.848563085936343e-05, "loss": 0.0558, "step": 24630 }, { "grad_norm": 0.5451090335845947, "learning_rate": 6.846002280082853e-05, "loss": 0.064, "step": 24640 }, { "grad_norm": 0.66241055727005, "learning_rate": 6.843440913462614e-05, "loss": 0.0651, "step": 24650 }, { "grad_norm": 0.4568520486354828, "learning_rate": 6.840878986853698e-05, "loss": 0.0467, "step": 24660 }, { "grad_norm": 0.6684854030609131, "learning_rate": 6.838316501034352e-05, "loss": 0.0654, "step": 24670 }, { "grad_norm": 0.559455394744873, "learning_rate": 6.83575345678299e-05, "loss": 0.0514, "step": 24680 }, { "grad_norm": 0.4629857838153839, "learning_rate": 6.833189854878196e-05, "loss": 0.0462, "step": 24690 }, { "grad_norm": 0.9468343257904053, "learning_rate": 6.83062569609873e-05, "loss": 0.055, "step": 24700 }, { "grad_norm": 0.5696774125099182, "learning_rate": 6.828060981223512e-05, "loss": 0.0706, "step": 24710 }, { "grad_norm": 0.8120402097702026, "learning_rate": 6.825495711031634e-05, "loss": 0.0549, "step": 24720 }, { "grad_norm": 0.40232622623443604, "learning_rate": 6.822929886302359e-05, "loss": 0.0579, "step": 24730 }, { "grad_norm": 0.451998233795166, "learning_rate": 6.820363507815116e-05, "loss": 0.0516, "step": 24740 }, { "grad_norm": 0.482476145029068, "learning_rate": 6.817796576349501e-05, "loss": 0.0582, "step": 24750 }, { "grad_norm": 0.708594799041748, "learning_rate": 6.815229092685285e-05, "loss": 0.0571, "step": 24760 }, { "grad_norm": 0.5441133975982666, "learning_rate": 6.812661057602399e-05, "loss": 0.0555, "step": 24770 }, { "grad_norm": 0.8034444451332092, "learning_rate": 6.810092471880943e-05, "loss": 0.0647, "step": 24780 }, { "grad_norm": 0.6887726783752441, "learning_rate": 6.807523336301187e-05, "loss": 0.0686, "step": 24790 }, { "grad_norm": 0.6186716556549072, "learning_rate": 6.804953651643566e-05, "loss": 0.0495, "step": 24800 }, { "grad_norm": 0.4178203344345093, "learning_rate": 6.802383418688685e-05, "loss": 0.0585, "step": 24810 }, { "grad_norm": 0.4642360508441925, "learning_rate": 6.799812638217309e-05, "loss": 0.0508, "step": 24820 }, { "grad_norm": 0.4678548574447632, "learning_rate": 6.797241311010373e-05, "loss": 0.0405, "step": 24830 }, { "grad_norm": 0.7131872773170471, "learning_rate": 6.794669437848982e-05, "loss": 0.0514, "step": 24840 }, { "grad_norm": 0.5072141885757446, "learning_rate": 6.792097019514402e-05, "loss": 0.0622, "step": 24850 }, { "grad_norm": 0.6935041546821594, "learning_rate": 6.789524056788064e-05, "loss": 0.0726, "step": 24860 }, { "grad_norm": 0.4279196560382843, "learning_rate": 6.786950550451567e-05, "loss": 0.0714, "step": 24870 }, { "grad_norm": 0.8718283176422119, "learning_rate": 6.784376501286676e-05, "loss": 0.0647, "step": 24880 }, { "grad_norm": 0.6166393160820007, "learning_rate": 6.781801910075316e-05, "loss": 0.0564, "step": 24890 }, { "grad_norm": 0.3907325267791748, "learning_rate": 6.779226777599581e-05, "loss": 0.0587, "step": 24900 }, { "grad_norm": 0.5420920848846436, "learning_rate": 6.776651104641729e-05, "loss": 0.0431, "step": 24910 }, { "grad_norm": 0.5075820088386536, "learning_rate": 6.774074891984183e-05, "loss": 0.0682, "step": 24920 }, { "grad_norm": 0.5730043053627014, "learning_rate": 6.771498140409526e-05, "loss": 0.0611, "step": 24930 }, { "grad_norm": 0.7330323457717896, "learning_rate": 6.768920850700506e-05, "loss": 0.0592, "step": 24940 }, { "grad_norm": 0.4732847809791565, "learning_rate": 6.766343023640039e-05, "loss": 0.0492, "step": 24950 }, { "grad_norm": 0.5660962462425232, "learning_rate": 6.763764660011198e-05, "loss": 0.0544, "step": 24960 }, { "grad_norm": 0.5731611847877502, "learning_rate": 6.761185760597223e-05, "loss": 0.0452, "step": 24970 }, { "grad_norm": 0.5201441049575806, "learning_rate": 6.758606326181515e-05, "loss": 0.0607, "step": 24980 }, { "grad_norm": 0.4192039966583252, "learning_rate": 6.75602635754764e-05, "loss": 0.0699, "step": 24990 }, { "grad_norm": 0.6533830165863037, "learning_rate": 6.75344585547932e-05, "loss": 0.0538, "step": 25000 }, { "grad_norm": 0.7759191393852234, "learning_rate": 6.750864820760449e-05, "loss": 0.0671, "step": 25010 }, { "grad_norm": 0.5090858936309814, "learning_rate": 6.748283254175072e-05, "loss": 0.0758, "step": 25020 }, { "grad_norm": 0.5074305534362793, "learning_rate": 6.745701156507404e-05, "loss": 0.0686, "step": 25030 }, { "grad_norm": 0.5265708565711975, "learning_rate": 6.743118528541818e-05, "loss": 0.0492, "step": 25040 }, { "grad_norm": 0.5358762145042419, "learning_rate": 6.740535371062846e-05, "loss": 0.0556, "step": 25050 }, { "grad_norm": 0.5845099687576294, "learning_rate": 6.737951684855185e-05, "loss": 0.0508, "step": 25060 }, { "grad_norm": 0.6273651719093323, "learning_rate": 6.735367470703691e-05, "loss": 0.0648, "step": 25070 }, { "grad_norm": 0.35784822702407837, "learning_rate": 6.732782729393379e-05, "loss": 0.0598, "step": 25080 }, { "grad_norm": 0.8129458427429199, "learning_rate": 6.730197461709425e-05, "loss": 0.0662, "step": 25090 }, { "grad_norm": 0.8542622327804565, "learning_rate": 6.727611668437164e-05, "loss": 0.0519, "step": 25100 }, { "grad_norm": 0.47352418303489685, "learning_rate": 6.725025350362094e-05, "loss": 0.0478, "step": 25110 }, { "grad_norm": 0.7320395112037659, "learning_rate": 6.72243850826987e-05, "loss": 0.0517, "step": 25120 }, { "grad_norm": 0.7161941528320312, "learning_rate": 6.719851142946305e-05, "loss": 0.0616, "step": 25130 }, { "grad_norm": 0.6836283802986145, "learning_rate": 6.717263255177372e-05, "loss": 0.0646, "step": 25140 }, { "grad_norm": 0.9618385434150696, "learning_rate": 6.714674845749205e-05, "loss": 0.0587, "step": 25150 }, { "grad_norm": 0.5941851139068604, "learning_rate": 6.712085915448092e-05, "loss": 0.0604, "step": 25160 }, { "grad_norm": 0.559992253780365, "learning_rate": 6.709496465060486e-05, "loss": 0.0697, "step": 25170 }, { "grad_norm": 0.6452124714851379, "learning_rate": 6.706906495372987e-05, "loss": 0.0546, "step": 25180 }, { "grad_norm": 0.6262770891189575, "learning_rate": 6.704316007172365e-05, "loss": 0.044, "step": 25190 }, { "grad_norm": 0.4543875753879547, "learning_rate": 6.701725001245539e-05, "loss": 0.0527, "step": 25200 }, { "grad_norm": 0.673397958278656, "learning_rate": 6.699133478379588e-05, "loss": 0.0558, "step": 25210 }, { "grad_norm": 0.540380597114563, "learning_rate": 6.69654143936175e-05, "loss": 0.0571, "step": 25220 }, { "grad_norm": 0.6219075918197632, "learning_rate": 6.693948884979419e-05, "loss": 0.046, "step": 25230 }, { "grad_norm": 0.4136693775653839, "learning_rate": 6.691355816020142e-05, "loss": 0.0556, "step": 25240 }, { "grad_norm": 0.7652283906936646, "learning_rate": 6.688762233271624e-05, "loss": 0.0482, "step": 25250 }, { "grad_norm": 0.46926555037498474, "learning_rate": 6.68616813752173e-05, "loss": 0.0475, "step": 25260 }, { "grad_norm": 0.5494879484176636, "learning_rate": 6.683573529558477e-05, "loss": 0.0662, "step": 25270 }, { "grad_norm": 0.578008770942688, "learning_rate": 6.680978410170037e-05, "loss": 0.069, "step": 25280 }, { "grad_norm": 0.5659987926483154, "learning_rate": 6.678382780144741e-05, "loss": 0.0618, "step": 25290 }, { "grad_norm": 0.711338996887207, "learning_rate": 6.675786640271071e-05, "loss": 0.0521, "step": 25300 }, { "grad_norm": 0.6702269911766052, "learning_rate": 6.673189991337665e-05, "loss": 0.046, "step": 25310 }, { "grad_norm": 0.6797541975975037, "learning_rate": 6.670592834133317e-05, "loss": 0.077, "step": 25320 }, { "grad_norm": 0.32852116227149963, "learning_rate": 6.667995169446979e-05, "loss": 0.0592, "step": 25330 }, { "grad_norm": 0.6796600818634033, "learning_rate": 6.665396998067747e-05, "loss": 0.0482, "step": 25340 }, { "grad_norm": 0.6226774454116821, "learning_rate": 6.66279832078488e-05, "loss": 0.0527, "step": 25350 }, { "grad_norm": 0.24596881866455078, "learning_rate": 6.660199138387786e-05, "loss": 0.0558, "step": 25360 }, { "grad_norm": 0.6362400054931641, "learning_rate": 6.65759945166603e-05, "loss": 0.0507, "step": 25370 }, { "grad_norm": 0.4544506371021271, "learning_rate": 6.654999261409326e-05, "loss": 0.0759, "step": 25380 }, { "grad_norm": 0.5204898118972778, "learning_rate": 6.652398568407544e-05, "loss": 0.0604, "step": 25390 }, { "grad_norm": 0.7384818196296692, "learning_rate": 6.649797373450707e-05, "loss": 0.0577, "step": 25400 }, { "grad_norm": 0.3817245364189148, "learning_rate": 6.647195677328988e-05, "loss": 0.045, "step": 25410 }, { "grad_norm": 0.608283281326294, "learning_rate": 6.644593480832712e-05, "loss": 0.0513, "step": 25420 }, { "grad_norm": 0.4197302460670471, "learning_rate": 6.641990784752363e-05, "loss": 0.0525, "step": 25430 }, { "grad_norm": 0.5935376286506653, "learning_rate": 6.639387589878566e-05, "loss": 0.0507, "step": 25440 }, { "grad_norm": 0.5956436395645142, "learning_rate": 6.636783897002103e-05, "loss": 0.0471, "step": 25450 }, { "grad_norm": 0.7242895364761353, "learning_rate": 6.63417970691391e-05, "loss": 0.0525, "step": 25460 }, { "grad_norm": 0.5682616233825684, "learning_rate": 6.63157502040507e-05, "loss": 0.0579, "step": 25470 }, { "grad_norm": 0.5417889356613159, "learning_rate": 6.628969838266819e-05, "loss": 0.0566, "step": 25480 }, { "grad_norm": 0.8867918848991394, "learning_rate": 6.626364161290541e-05, "loss": 0.0436, "step": 25490 }, { "grad_norm": 0.35261914134025574, "learning_rate": 6.623757990267774e-05, "loss": 0.036, "step": 25500 }, { "grad_norm": 0.5881075263023376, "learning_rate": 6.621151325990201e-05, "loss": 0.0612, "step": 25510 }, { "grad_norm": 0.8486085534095764, "learning_rate": 6.618544169249657e-05, "loss": 0.0625, "step": 25520 }, { "grad_norm": 0.6705623865127563, "learning_rate": 6.615936520838133e-05, "loss": 0.073, "step": 25530 }, { "grad_norm": 0.8110930323600769, "learning_rate": 6.613328381547759e-05, "loss": 0.0604, "step": 25540 }, { "grad_norm": 0.5052363872528076, "learning_rate": 6.610719752170821e-05, "loss": 0.0492, "step": 25550 }, { "grad_norm": 0.43563902378082275, "learning_rate": 6.60811063349975e-05, "loss": 0.052, "step": 25560 }, { "grad_norm": 0.8226012587547302, "learning_rate": 6.605501026327127e-05, "loss": 0.0724, "step": 25570 }, { "grad_norm": 0.5338396430015564, "learning_rate": 6.602890931445685e-05, "loss": 0.0469, "step": 25580 }, { "grad_norm": 0.7599382400512695, "learning_rate": 6.6002803496483e-05, "loss": 0.0565, "step": 25590 }, { "grad_norm": 0.4567371904850006, "learning_rate": 6.597669281727997e-05, "loss": 0.0473, "step": 25600 }, { "grad_norm": 0.5538306832313538, "learning_rate": 6.595057728477949e-05, "loss": 0.0597, "step": 25610 }, { "grad_norm": 0.4264487028121948, "learning_rate": 6.59244569069148e-05, "loss": 0.0595, "step": 25620 }, { "grad_norm": 0.4067345857620239, "learning_rate": 6.589833169162054e-05, "loss": 0.0517, "step": 25630 }, { "grad_norm": 0.4668246805667877, "learning_rate": 6.587220164683291e-05, "loss": 0.0588, "step": 25640 }, { "grad_norm": 0.490334153175354, "learning_rate": 6.58460667804895e-05, "loss": 0.0537, "step": 25650 }, { "grad_norm": 0.9343883395195007, "learning_rate": 6.581992710052938e-05, "loss": 0.0624, "step": 25660 }, { "grad_norm": 0.9722798466682434, "learning_rate": 6.579378261489311e-05, "loss": 0.0438, "step": 25670 }, { "grad_norm": 0.6649230122566223, "learning_rate": 6.576763333152268e-05, "loss": 0.0397, "step": 25680 }, { "grad_norm": 0.6932309865951538, "learning_rate": 6.574147925836159e-05, "loss": 0.0419, "step": 25690 }, { "grad_norm": 0.5260859131813049, "learning_rate": 6.571532040335472e-05, "loss": 0.0509, "step": 25700 }, { "grad_norm": 0.6756572723388672, "learning_rate": 6.568915677444845e-05, "loss": 0.0637, "step": 25710 }, { "grad_norm": 0.48992985486984253, "learning_rate": 6.56629883795906e-05, "loss": 0.0477, "step": 25720 }, { "grad_norm": 0.4619085490703583, "learning_rate": 6.563681522673043e-05, "loss": 0.0446, "step": 25730 }, { "grad_norm": 0.6191496253013611, "learning_rate": 6.561063732381867e-05, "loss": 0.0398, "step": 25740 }, { "grad_norm": 0.3703993558883667, "learning_rate": 6.558445467880745e-05, "loss": 0.0396, "step": 25750 }, { "grad_norm": 0.8298439979553223, "learning_rate": 6.55582672996504e-05, "loss": 0.0384, "step": 25760 }, { "grad_norm": 0.41026973724365234, "learning_rate": 6.553207519430253e-05, "loss": 0.0457, "step": 25770 }, { "grad_norm": 0.5091546773910522, "learning_rate": 6.550587837072032e-05, "loss": 0.0528, "step": 25780 }, { "grad_norm": 0.45014914870262146, "learning_rate": 6.547967683686166e-05, "loss": 0.0626, "step": 25790 }, { "grad_norm": 0.39597272872924805, "learning_rate": 6.545347060068591e-05, "loss": 0.0555, "step": 25800 }, { "grad_norm": 0.6292934417724609, "learning_rate": 6.542725967015382e-05, "loss": 0.0543, "step": 25810 }, { "grad_norm": 0.9757052063941956, "learning_rate": 6.540104405322757e-05, "loss": 0.057, "step": 25820 }, { "grad_norm": 0.44212475419044495, "learning_rate": 6.537482375787077e-05, "loss": 0.0638, "step": 25830 }, { "grad_norm": 0.390840619802475, "learning_rate": 6.534859879204845e-05, "loss": 0.0436, "step": 25840 }, { "grad_norm": 0.40948525071144104, "learning_rate": 6.532236916372709e-05, "loss": 0.0404, "step": 25850 }, { "grad_norm": 0.6587900519371033, "learning_rate": 6.529613488087454e-05, "loss": 0.0521, "step": 25860 }, { "grad_norm": 1.0213860273361206, "learning_rate": 6.526989595146009e-05, "loss": 0.0503, "step": 25870 }, { "grad_norm": 0.628628134727478, "learning_rate": 6.524365238345441e-05, "loss": 0.0517, "step": 25880 }, { "grad_norm": 0.40473100543022156, "learning_rate": 6.521740418482964e-05, "loss": 0.0466, "step": 25890 }, { "grad_norm": 0.47377699613571167, "learning_rate": 6.519115136355925e-05, "loss": 0.0334, "step": 25900 }, { "grad_norm": 0.35126447677612305, "learning_rate": 6.51648939276182e-05, "loss": 0.0416, "step": 25910 }, { "grad_norm": 0.4451252818107605, "learning_rate": 6.513863188498277e-05, "loss": 0.0415, "step": 25920 }, { "grad_norm": 0.478104829788208, "learning_rate": 6.511236524363068e-05, "loss": 0.0441, "step": 25930 }, { "grad_norm": 0.6401453614234924, "learning_rate": 6.508609401154104e-05, "loss": 0.0513, "step": 25940 }, { "grad_norm": 0.5286593437194824, "learning_rate": 6.505981819669439e-05, "loss": 0.044, "step": 25950 }, { "grad_norm": 0.7869622111320496, "learning_rate": 6.503353780707258e-05, "loss": 0.0558, "step": 25960 }, { "grad_norm": 0.3579389154911041, "learning_rate": 6.500725285065895e-05, "loss": 0.0478, "step": 25970 }, { "grad_norm": 1.2111321687698364, "learning_rate": 6.498096333543813e-05, "loss": 0.0378, "step": 25980 }, { "grad_norm": 0.3816555142402649, "learning_rate": 6.49546692693962e-05, "loss": 0.0584, "step": 25990 }, { "grad_norm": 0.7042136788368225, "learning_rate": 6.492837066052059e-05, "loss": 0.0448, "step": 26000 }, { "grad_norm": 0.8461688756942749, "learning_rate": 6.490206751680014e-05, "loss": 0.0536, "step": 26010 }, { "grad_norm": 0.9557523727416992, "learning_rate": 6.487575984622505e-05, "loss": 0.0833, "step": 26020 }, { "grad_norm": 0.7358940243721008, "learning_rate": 6.484944765678689e-05, "loss": 0.0617, "step": 26030 }, { "grad_norm": 0.6401054859161377, "learning_rate": 6.482313095647861e-05, "loss": 0.0675, "step": 26040 }, { "grad_norm": 0.5705169439315796, "learning_rate": 6.479680975329451e-05, "loss": 0.0657, "step": 26050 }, { "grad_norm": 0.5558307766914368, "learning_rate": 6.477048405523031e-05, "loss": 0.0535, "step": 26060 }, { "grad_norm": 1.497488260269165, "learning_rate": 6.474415387028304e-05, "loss": 0.0491, "step": 26070 }, { "grad_norm": 0.48786553740501404, "learning_rate": 6.471781920645114e-05, "loss": 0.049, "step": 26080 }, { "grad_norm": 0.5650164484977722, "learning_rate": 6.469148007173434e-05, "loss": 0.0565, "step": 26090 }, { "grad_norm": 0.7046695947647095, "learning_rate": 6.466513647413381e-05, "loss": 0.0513, "step": 26100 }, { "grad_norm": 0.4844236671924591, "learning_rate": 6.463878842165203e-05, "loss": 0.0426, "step": 26110 }, { "grad_norm": 0.5335492491722107, "learning_rate": 6.461243592229286e-05, "loss": 0.0623, "step": 26120 }, { "grad_norm": 0.7614229321479797, "learning_rate": 6.458607898406146e-05, "loss": 0.0456, "step": 26130 }, { "grad_norm": 0.5982668399810791, "learning_rate": 6.455971761496439e-05, "loss": 0.0779, "step": 26140 }, { "grad_norm": 0.5720628499984741, "learning_rate": 6.453335182300953e-05, "loss": 0.0458, "step": 26150 }, { "grad_norm": 0.5444460511207581, "learning_rate": 6.450698161620612e-05, "loss": 0.0424, "step": 26160 }, { "grad_norm": 0.9780800342559814, "learning_rate": 6.448060700256473e-05, "loss": 0.0503, "step": 26170 }, { "grad_norm": 0.6989696025848389, "learning_rate": 6.445422799009726e-05, "loss": 0.0644, "step": 26180 }, { "grad_norm": 0.472922146320343, "learning_rate": 6.442784458681699e-05, "loss": 0.0538, "step": 26190 }, { "grad_norm": 0.30800801515579224, "learning_rate": 6.440145680073847e-05, "loss": 0.0537, "step": 26200 }, { "grad_norm": 0.645792543888092, "learning_rate": 6.437506463987762e-05, "loss": 0.0487, "step": 26210 }, { "grad_norm": 0.6812556982040405, "learning_rate": 6.434866811225168e-05, "loss": 0.0631, "step": 26220 }, { "grad_norm": 0.3974437415599823, "learning_rate": 6.432226722587923e-05, "loss": 0.0549, "step": 26230 }, { "grad_norm": 0.5480038523674011, "learning_rate": 6.429586198878015e-05, "loss": 0.0415, "step": 26240 }, { "grad_norm": 0.5114845633506775, "learning_rate": 6.426945240897566e-05, "loss": 0.0662, "step": 26250 }, { "grad_norm": 0.4430543780326843, "learning_rate": 6.424303849448829e-05, "loss": 0.0436, "step": 26260 }, { "grad_norm": 0.7821565866470337, "learning_rate": 6.42166202533419e-05, "loss": 0.0572, "step": 26270 }, { "grad_norm": 0.5669312477111816, "learning_rate": 6.419019769356164e-05, "loss": 0.0492, "step": 26280 }, { "grad_norm": 0.590433657169342, "learning_rate": 6.416377082317398e-05, "loss": 0.0451, "step": 26290 }, { "grad_norm": 0.4635560214519501, "learning_rate": 6.413733965020674e-05, "loss": 0.0535, "step": 26300 }, { "grad_norm": 0.45392441749572754, "learning_rate": 6.411090418268896e-05, "loss": 0.0459, "step": 26310 }, { "grad_norm": 0.5825352668762207, "learning_rate": 6.408446442865109e-05, "loss": 0.0448, "step": 26320 }, { "grad_norm": 0.9085395932197571, "learning_rate": 6.405802039612479e-05, "loss": 0.0597, "step": 26330 }, { "grad_norm": 0.6026832461357117, "learning_rate": 6.403157209314308e-05, "loss": 0.0386, "step": 26340 }, { "grad_norm": 0.42646345496177673, "learning_rate": 6.400511952774024e-05, "loss": 0.0532, "step": 26350 }, { "grad_norm": 0.5104751586914062, "learning_rate": 6.397866270795187e-05, "loss": 0.0395, "step": 26360 }, { "grad_norm": 0.5264269709587097, "learning_rate": 6.395220164181489e-05, "loss": 0.0539, "step": 26370 }, { "grad_norm": 0.9683710336685181, "learning_rate": 6.39257363373674e-05, "loss": 0.0507, "step": 26380 }, { "grad_norm": 0.46455687284469604, "learning_rate": 6.389926680264892e-05, "loss": 0.0421, "step": 26390 }, { "grad_norm": 0.650915801525116, "learning_rate": 6.387279304570017e-05, "loss": 0.0559, "step": 26400 }, { "grad_norm": 0.6951446533203125, "learning_rate": 6.384631507456319e-05, "loss": 0.0409, "step": 26410 }, { "grad_norm": 0.5365628600120544, "learning_rate": 6.381983289728126e-05, "loss": 0.0364, "step": 26420 }, { "grad_norm": 0.6582127809524536, "learning_rate": 6.3793346521899e-05, "loss": 0.0463, "step": 26430 }, { "grad_norm": 0.3815669119358063, "learning_rate": 6.376685595646226e-05, "loss": 0.0401, "step": 26440 }, { "grad_norm": 0.6646010279655457, "learning_rate": 6.374036120901816e-05, "loss": 0.0604, "step": 26450 }, { "grad_norm": 0.6213977932929993, "learning_rate": 6.371386228761514e-05, "loss": 0.0547, "step": 26460 }, { "grad_norm": 0.8648363351821899, "learning_rate": 6.368735920030283e-05, "loss": 0.076, "step": 26470 }, { "grad_norm": 0.5278541445732117, "learning_rate": 6.366085195513218e-05, "loss": 0.0509, "step": 26480 }, { "grad_norm": 0.7754632830619812, "learning_rate": 6.363434056015543e-05, "loss": 0.0485, "step": 26490 }, { "grad_norm": 0.5194865465164185, "learning_rate": 6.360782502342599e-05, "loss": 0.0493, "step": 26500 }, { "grad_norm": 0.8167347311973572, "learning_rate": 6.358130535299862e-05, "loss": 0.0449, "step": 26510 }, { "grad_norm": 0.46010321378707886, "learning_rate": 6.355478155692926e-05, "loss": 0.0589, "step": 26520 }, { "grad_norm": 0.6249545216560364, "learning_rate": 6.352825364327517e-05, "loss": 0.0468, "step": 26530 }, { "grad_norm": 0.5290242433547974, "learning_rate": 6.350172162009482e-05, "loss": 0.0503, "step": 26540 }, { "grad_norm": 0.6198933124542236, "learning_rate": 6.347518549544793e-05, "loss": 0.043, "step": 26550 }, { "grad_norm": 0.6957939863204956, "learning_rate": 6.344864527739547e-05, "loss": 0.0458, "step": 26560 }, { "grad_norm": 0.37421244382858276, "learning_rate": 6.342210097399966e-05, "loss": 0.0364, "step": 26570 }, { "grad_norm": 0.3402997851371765, "learning_rate": 6.339555259332398e-05, "loss": 0.0486, "step": 26580 }, { "grad_norm": 0.4888652265071869, "learning_rate": 6.33690001434331e-05, "loss": 0.0452, "step": 26590 }, { "grad_norm": 0.3975067138671875, "learning_rate": 6.334244363239296e-05, "loss": 0.0454, "step": 26600 }, { "grad_norm": 0.7072888612747192, "learning_rate": 6.331588306827073e-05, "loss": 0.0462, "step": 26610 }, { "grad_norm": 0.49163612723350525, "learning_rate": 6.328931845913483e-05, "loss": 0.0668, "step": 26620 }, { "grad_norm": 0.48121756315231323, "learning_rate": 6.326274981305484e-05, "loss": 0.0556, "step": 26630 }, { "grad_norm": 0.6370738744735718, "learning_rate": 6.323617713810166e-05, "loss": 0.0568, "step": 26640 }, { "grad_norm": 0.6003077626228333, "learning_rate": 6.320960044234734e-05, "loss": 0.0465, "step": 26650 }, { "grad_norm": 0.4492085576057434, "learning_rate": 6.318301973386518e-05, "loss": 0.0484, "step": 26660 }, { "grad_norm": 0.41157034039497375, "learning_rate": 6.315643502072971e-05, "loss": 0.0544, "step": 26670 }, { "grad_norm": 0.42601278424263, "learning_rate": 6.312984631101667e-05, "loss": 0.0476, "step": 26680 }, { "grad_norm": 0.9350765347480774, "learning_rate": 6.310325361280297e-05, "loss": 0.0525, "step": 26690 }, { "grad_norm": 0.48427075147628784, "learning_rate": 6.30766569341668e-05, "loss": 0.0416, "step": 26700 }, { "grad_norm": 0.5001838207244873, "learning_rate": 6.305005628318753e-05, "loss": 0.0413, "step": 26710 }, { "grad_norm": 0.6705949902534485, "learning_rate": 6.302345166794572e-05, "loss": 0.0523, "step": 26720 }, { "grad_norm": 0.6110866069793701, "learning_rate": 6.299684309652316e-05, "loss": 0.0577, "step": 26730 }, { "grad_norm": 0.7373630404472351, "learning_rate": 6.297023057700283e-05, "loss": 0.0621, "step": 26740 }, { "grad_norm": 0.6954562664031982, "learning_rate": 6.294361411746891e-05, "loss": 0.0453, "step": 26750 }, { "grad_norm": 0.8297215700149536, "learning_rate": 6.291699372600677e-05, "loss": 0.0638, "step": 26760 }, { "grad_norm": 0.8897647261619568, "learning_rate": 6.2890369410703e-05, "loss": 0.0629, "step": 26770 }, { "grad_norm": 0.6355799436569214, "learning_rate": 6.286374117964534e-05, "loss": 0.0609, "step": 26780 }, { "grad_norm": 0.41643232107162476, "learning_rate": 6.283710904092277e-05, "loss": 0.0669, "step": 26790 }, { "grad_norm": 0.5418443083763123, "learning_rate": 6.281047300262542e-05, "loss": 0.0631, "step": 26800 }, { "grad_norm": 0.7048614621162415, "learning_rate": 6.278383307284461e-05, "loss": 0.0442, "step": 26810 }, { "grad_norm": 0.4688805937767029, "learning_rate": 6.275718925967284e-05, "loss": 0.0437, "step": 26820 }, { "grad_norm": 0.4237678349018097, "learning_rate": 6.273054157120382e-05, "loss": 0.0562, "step": 26830 }, { "grad_norm": 0.6794803142547607, "learning_rate": 6.270389001553238e-05, "loss": 0.052, "step": 26840 }, { "grad_norm": 0.6293498277664185, "learning_rate": 6.26772346007546e-05, "loss": 0.0442, "step": 26850 }, { "grad_norm": 0.4142550528049469, "learning_rate": 6.265057533496767e-05, "loss": 0.0569, "step": 26860 }, { "grad_norm": 0.3687520921230316, "learning_rate": 6.262391222626997e-05, "loss": 0.0451, "step": 26870 }, { "grad_norm": 0.5099179744720459, "learning_rate": 6.259724528276106e-05, "loss": 0.0443, "step": 26880 }, { "grad_norm": 0.497857928276062, "learning_rate": 6.257057451254162e-05, "loss": 0.0568, "step": 26890 }, { "grad_norm": 0.7208967208862305, "learning_rate": 6.254389992371357e-05, "loss": 0.0711, "step": 26900 }, { "grad_norm": 0.4865054190158844, "learning_rate": 6.25172215243799e-05, "loss": 0.0692, "step": 26910 }, { "grad_norm": 0.3448513150215149, "learning_rate": 6.249053932264486e-05, "loss": 0.0558, "step": 26920 }, { "grad_norm": 0.6026396155357361, "learning_rate": 6.246385332661376e-05, "loss": 0.0616, "step": 26930 }, { "grad_norm": 0.36857959628105164, "learning_rate": 6.24371635443931e-05, "loss": 0.0573, "step": 26940 }, { "grad_norm": 0.46171170473098755, "learning_rate": 6.241046998409054e-05, "loss": 0.0433, "step": 26950 }, { "grad_norm": 0.6680497527122498, "learning_rate": 6.238377265381489e-05, "loss": 0.0636, "step": 26960 }, { "grad_norm": 0.64412921667099, "learning_rate": 6.235707156167607e-05, "loss": 0.0571, "step": 26970 }, { "grad_norm": 0.6407986879348755, "learning_rate": 6.233036671578519e-05, "loss": 0.0448, "step": 26980 }, { "grad_norm": 0.6036216020584106, "learning_rate": 6.230365812425445e-05, "loss": 0.0554, "step": 26990 }, { "grad_norm": 0.7851673364639282, "learning_rate": 6.227694579519724e-05, "loss": 0.0438, "step": 27000 }, { "grad_norm": 0.6917878985404968, "learning_rate": 6.225022973672805e-05, "loss": 0.0597, "step": 27010 }, { "grad_norm": 0.4476471245288849, "learning_rate": 6.222350995696253e-05, "loss": 0.0637, "step": 27020 }, { "grad_norm": 1.175206184387207, "learning_rate": 6.21967864640174e-05, "loss": 0.0675, "step": 27030 }, { "grad_norm": 0.5797032713890076, "learning_rate": 6.217005926601059e-05, "loss": 0.0517, "step": 27040 }, { "grad_norm": 0.3674978017807007, "learning_rate": 6.214332837106111e-05, "loss": 0.062, "step": 27050 }, { "grad_norm": 0.6306955814361572, "learning_rate": 6.21165937872891e-05, "loss": 0.0518, "step": 27060 }, { "grad_norm": 0.44783633947372437, "learning_rate": 6.208985552281582e-05, "loss": 0.058, "step": 27070 }, { "grad_norm": 0.48780912160873413, "learning_rate": 6.206311358576364e-05, "loss": 0.0436, "step": 27080 }, { "grad_norm": 1.0909770727157593, "learning_rate": 6.203636798425608e-05, "loss": 0.0559, "step": 27090 }, { "grad_norm": 0.3614930510520935, "learning_rate": 6.20096187264177e-05, "loss": 0.0419, "step": 27100 }, { "grad_norm": 0.4935089349746704, "learning_rate": 6.198286582037425e-05, "loss": 0.0552, "step": 27110 }, { "grad_norm": 0.5776824951171875, "learning_rate": 6.195610927425256e-05, "loss": 0.0452, "step": 27120 }, { "grad_norm": 0.4593889117240906, "learning_rate": 6.192934909618056e-05, "loss": 0.0356, "step": 27130 }, { "grad_norm": 0.7283263802528381, "learning_rate": 6.190258529428728e-05, "loss": 0.0355, "step": 27140 }, { "grad_norm": 0.49573156237602234, "learning_rate": 6.187581787670285e-05, "loss": 0.0392, "step": 27150 }, { "grad_norm": 0.47711285948753357, "learning_rate": 6.184904685155852e-05, "loss": 0.0478, "step": 27160 }, { "grad_norm": 0.5649611353874207, "learning_rate": 6.18222722269866e-05, "loss": 0.0549, "step": 27170 }, { "grad_norm": 0.3371294438838959, "learning_rate": 6.179549401112053e-05, "loss": 0.0418, "step": 27180 }, { "grad_norm": 0.3149394094944, "learning_rate": 6.176871221209482e-05, "loss": 0.0717, "step": 27190 }, { "grad_norm": 0.3265933096408844, "learning_rate": 6.174192683804508e-05, "loss": 0.0584, "step": 27200 }, { "grad_norm": 0.5850622057914734, "learning_rate": 6.1715137897108e-05, "loss": 0.0528, "step": 27210 }, { "grad_norm": 0.5617479085922241, "learning_rate": 6.168834539742134e-05, "loss": 0.0475, "step": 27220 }, { "grad_norm": 0.5786117911338806, "learning_rate": 6.166154934712397e-05, "loss": 0.0518, "step": 27230 }, { "grad_norm": 0.296686589717865, "learning_rate": 6.163474975435581e-05, "loss": 0.0787, "step": 27240 }, { "grad_norm": 0.6714733839035034, "learning_rate": 6.160794662725787e-05, "loss": 0.0515, "step": 27250 }, { "grad_norm": 0.5251820087432861, "learning_rate": 6.158113997397222e-05, "loss": 0.0447, "step": 27260 }, { "grad_norm": 0.42421093583106995, "learning_rate": 6.155432980264205e-05, "loss": 0.0521, "step": 27270 }, { "grad_norm": 1.3097147941589355, "learning_rate": 6.152751612141156e-05, "loss": 0.0508, "step": 27280 }, { "grad_norm": 0.6054821610450745, "learning_rate": 6.150069893842602e-05, "loss": 0.0491, "step": 27290 }, { "grad_norm": 0.4075199067592621, "learning_rate": 6.147387826183182e-05, "loss": 0.0618, "step": 27300 }, { "grad_norm": 0.45043250918388367, "learning_rate": 6.144705409977635e-05, "loss": 0.0615, "step": 27310 }, { "grad_norm": 0.5871935486793518, "learning_rate": 6.142022646040808e-05, "loss": 0.0637, "step": 27320 }, { "grad_norm": 0.26343557238578796, "learning_rate": 6.139339535187653e-05, "loss": 0.0475, "step": 27330 }, { "grad_norm": 0.5230186581611633, "learning_rate": 6.136656078233232e-05, "loss": 0.0507, "step": 27340 }, { "grad_norm": 0.5407661199569702, "learning_rate": 6.133972275992707e-05, "loss": 0.0483, "step": 27350 }, { "grad_norm": 0.55363529920578, "learning_rate": 6.131288129281342e-05, "loss": 0.0542, "step": 27360 }, { "grad_norm": 0.592853844165802, "learning_rate": 6.128603638914516e-05, "loss": 0.0506, "step": 27370 }, { "grad_norm": 0.7811775803565979, "learning_rate": 6.125918805707704e-05, "loss": 0.0596, "step": 27380 }, { "grad_norm": 0.6248492002487183, "learning_rate": 6.123233630476485e-05, "loss": 0.0488, "step": 27390 }, { "grad_norm": 0.5945422053337097, "learning_rate": 6.120548114036547e-05, "loss": 0.0533, "step": 27400 }, { "grad_norm": 0.6072249412536621, "learning_rate": 6.117862257203679e-05, "loss": 0.054, "step": 27410 }, { "grad_norm": 0.38005030155181885, "learning_rate": 6.115176060793771e-05, "loss": 0.0513, "step": 27420 }, { "grad_norm": 0.5119679570198059, "learning_rate": 6.112489525622822e-05, "loss": 0.0429, "step": 27430 }, { "grad_norm": 0.6087066531181335, "learning_rate": 6.109802652506928e-05, "loss": 0.0477, "step": 27440 }, { "grad_norm": 0.5674102306365967, "learning_rate": 6.107115442262291e-05, "loss": 0.0447, "step": 27450 }, { "grad_norm": 0.334622323513031, "learning_rate": 6.104427895705214e-05, "loss": 0.041, "step": 27460 }, { "grad_norm": 0.6312388181686401, "learning_rate": 6.101740013652103e-05, "loss": 0.0455, "step": 27470 }, { "grad_norm": 1.1533949375152588, "learning_rate": 6.099051796919465e-05, "loss": 0.0547, "step": 27480 }, { "grad_norm": 0.47193044424057007, "learning_rate": 6.096363246323911e-05, "loss": 0.0598, "step": 27490 }, { "grad_norm": 0.5781168937683105, "learning_rate": 6.0936743626821504e-05, "loss": 0.0378, "step": 27500 }, { "grad_norm": 0.5292292237281799, "learning_rate": 6.090985146810996e-05, "loss": 0.0596, "step": 27510 }, { "grad_norm": 0.40928757190704346, "learning_rate": 6.088295599527357e-05, "loss": 0.0898, "step": 27520 }, { "grad_norm": 0.42069512605667114, "learning_rate": 6.085605721648252e-05, "loss": 0.044, "step": 27530 }, { "grad_norm": 0.6767619252204895, "learning_rate": 6.082915513990792e-05, "loss": 0.0446, "step": 27540 }, { "grad_norm": 0.6870677471160889, "learning_rate": 6.080224977372192e-05, "loss": 0.0565, "step": 27550 }, { "grad_norm": 0.5652996897697449, "learning_rate": 6.0775341126097666e-05, "loss": 0.0474, "step": 27560 }, { "grad_norm": 0.5764406323432922, "learning_rate": 6.074842920520926e-05, "loss": 0.0543, "step": 27570 }, { "grad_norm": 0.9620420336723328, "learning_rate": 6.072151401923186e-05, "loss": 0.06, "step": 27580 }, { "grad_norm": 0.3439388871192932, "learning_rate": 6.069459557634159e-05, "loss": 0.0503, "step": 27590 }, { "grad_norm": 1.220280647277832, "learning_rate": 6.066767388471557e-05, "loss": 0.0523, "step": 27600 }, { "grad_norm": 0.6571613550186157, "learning_rate": 6.064074895253188e-05, "loss": 0.0637, "step": 27610 }, { "grad_norm": 0.5906733274459839, "learning_rate": 6.061382078796961e-05, "loss": 0.0442, "step": 27620 }, { "grad_norm": 0.4652944803237915, "learning_rate": 6.0586889399208814e-05, "loss": 0.0405, "step": 27630 }, { "grad_norm": 0.7141004800796509, "learning_rate": 6.0559954794430565e-05, "loss": 0.0563, "step": 27640 }, { "grad_norm": 0.4648074209690094, "learning_rate": 6.053301698181687e-05, "loss": 0.056, "step": 27650 }, { "grad_norm": 0.5958178043365479, "learning_rate": 6.0506075969550725e-05, "loss": 0.0507, "step": 27660 }, { "grad_norm": 0.4739116132259369, "learning_rate": 6.047913176581609e-05, "loss": 0.0676, "step": 27670 }, { "grad_norm": 0.7090047001838684, "learning_rate": 6.0452184378797904e-05, "loss": 0.0472, "step": 27680 }, { "grad_norm": 0.31723490357398987, "learning_rate": 6.042523381668209e-05, "loss": 0.0385, "step": 27690 }, { "grad_norm": 0.5512009263038635, "learning_rate": 6.03982800876555e-05, "loss": 0.0583, "step": 27700 }, { "grad_norm": 0.4197803735733032, "learning_rate": 6.0371323199905975e-05, "loss": 0.0479, "step": 27710 }, { "grad_norm": 1.0223640203475952, "learning_rate": 6.03443631616223e-05, "loss": 0.0443, "step": 27720 }, { "grad_norm": 0.6659941673278809, "learning_rate": 6.031739998099421e-05, "loss": 0.0596, "step": 27730 }, { "grad_norm": 0.6242811679840088, "learning_rate": 6.029043366621243e-05, "loss": 0.0453, "step": 27740 }, { "grad_norm": 0.6059478521347046, "learning_rate": 6.0263464225468615e-05, "loss": 0.0418, "step": 27750 }, { "grad_norm": 0.5843249559402466, "learning_rate": 6.023649166695534e-05, "loss": 0.0344, "step": 27760 }, { "grad_norm": 0.6667364239692688, "learning_rate": 6.0209515998866186e-05, "loss": 0.0556, "step": 27770 }, { "grad_norm": 0.8191359043121338, "learning_rate": 6.018253722939563e-05, "loss": 0.0783, "step": 27780 }, { "grad_norm": 0.3993068337440491, "learning_rate": 6.015555536673914e-05, "loss": 0.0494, "step": 27790 }, { "grad_norm": 0.5027754902839661, "learning_rate": 6.0128570419093054e-05, "loss": 0.0416, "step": 27800 }, { "grad_norm": 0.5351691842079163, "learning_rate": 6.010158239465471e-05, "loss": 0.0503, "step": 27810 }, { "grad_norm": 0.465078741312027, "learning_rate": 6.007459130162235e-05, "loss": 0.0461, "step": 27820 }, { "grad_norm": 0.37942537665367126, "learning_rate": 6.004759714819516e-05, "loss": 0.0636, "step": 27830 }, { "grad_norm": 0.37315696477890015, "learning_rate": 6.002059994257323e-05, "loss": 0.0393, "step": 27840 }, { "grad_norm": 0.46679243445396423, "learning_rate": 5.999359969295764e-05, "loss": 0.0461, "step": 27850 }, { "grad_norm": 0.53382807970047, "learning_rate": 5.9966596407550314e-05, "loss": 0.0531, "step": 27860 }, { "grad_norm": 0.5924631953239441, "learning_rate": 5.993959009455416e-05, "loss": 0.0517, "step": 27870 }, { "grad_norm": 0.32366082072257996, "learning_rate": 5.991258076217298e-05, "loss": 0.0478, "step": 27880 }, { "grad_norm": 0.6506527662277222, "learning_rate": 5.988556841861147e-05, "loss": 0.054, "step": 27890 }, { "grad_norm": 0.7148016691207886, "learning_rate": 5.985855307207531e-05, "loss": 0.0501, "step": 27900 }, { "grad_norm": 0.4502241015434265, "learning_rate": 5.9831534730771e-05, "loss": 0.0471, "step": 27910 }, { "grad_norm": 0.8125535249710083, "learning_rate": 5.980451340290605e-05, "loss": 0.0457, "step": 27920 }, { "grad_norm": 0.5243474841117859, "learning_rate": 5.97774890966888e-05, "loss": 0.0632, "step": 27930 }, { "grad_norm": 0.46517258882522583, "learning_rate": 5.975046182032851e-05, "loss": 0.0685, "step": 27940 }, { "grad_norm": 0.5294235348701477, "learning_rate": 5.972343158203537e-05, "loss": 0.0449, "step": 27950 }, { "grad_norm": 0.49802061915397644, "learning_rate": 5.969639839002045e-05, "loss": 0.0594, "step": 27960 }, { "grad_norm": 0.6366544365882874, "learning_rate": 5.966936225249572e-05, "loss": 0.0547, "step": 27970 }, { "grad_norm": 0.3248496949672699, "learning_rate": 5.9642323177674044e-05, "loss": 0.0404, "step": 27980 }, { "grad_norm": 0.4137669503688812, "learning_rate": 5.9615281173769154e-05, "loss": 0.0469, "step": 27990 }, { "grad_norm": 0.38169237971305847, "learning_rate": 5.958823624899574e-05, "loss": 0.044, "step": 28000 }, { "grad_norm": 0.3743448555469513, "learning_rate": 5.956118841156933e-05, "loss": 0.0434, "step": 28010 }, { "grad_norm": 0.5178070068359375, "learning_rate": 5.953413766970631e-05, "loss": 0.0563, "step": 28020 }, { "grad_norm": 0.6947453618049622, "learning_rate": 5.9507084031624e-05, "loss": 0.0529, "step": 28030 }, { "grad_norm": 0.3340243101119995, "learning_rate": 5.948002750554058e-05, "loss": 0.0528, "step": 28040 }, { "grad_norm": 0.5293406248092651, "learning_rate": 5.9452968099675124e-05, "loss": 0.045, "step": 28050 }, { "grad_norm": 0.9207466244697571, "learning_rate": 5.9425905822247527e-05, "loss": 0.0516, "step": 28060 }, { "grad_norm": 0.6574786901473999, "learning_rate": 5.939884068147864e-05, "loss": 0.0422, "step": 28070 }, { "grad_norm": 0.7019681930541992, "learning_rate": 5.937177268559011e-05, "loss": 0.0444, "step": 28080 }, { "grad_norm": 0.9057872891426086, "learning_rate": 5.934470184280448e-05, "loss": 0.037, "step": 28090 }, { "grad_norm": 0.3771449625492096, "learning_rate": 5.931762816134516e-05, "loss": 0.0395, "step": 28100 }, { "grad_norm": 0.8187233805656433, "learning_rate": 5.9290551649436434e-05, "loss": 0.0488, "step": 28110 }, { "grad_norm": 0.453836053609848, "learning_rate": 5.9263472315303416e-05, "loss": 0.046, "step": 28120 }, { "grad_norm": 0.40679025650024414, "learning_rate": 5.9236390167172096e-05, "loss": 0.0566, "step": 28130 }, { "grad_norm": 0.5478391051292419, "learning_rate": 5.920930521326932e-05, "loss": 0.046, "step": 28140 }, { "grad_norm": 0.7036687135696411, "learning_rate": 5.918221746182276e-05, "loss": 0.0488, "step": 28150 }, { "grad_norm": 0.7727155685424805, "learning_rate": 5.9155126921061e-05, "loss": 0.0356, "step": 28160 }, { "grad_norm": 1.1507319211959839, "learning_rate": 5.91280335992134e-05, "loss": 0.0605, "step": 28170 }, { "grad_norm": 0.6343638896942139, "learning_rate": 5.91009375045102e-05, "loss": 0.0448, "step": 28180 }, { "grad_norm": 0.9651554822921753, "learning_rate": 5.9073838645182476e-05, "loss": 0.055, "step": 28190 }, { "grad_norm": 0.4208143353462219, "learning_rate": 5.904673702946217e-05, "loss": 0.0536, "step": 28200 }, { "grad_norm": 0.5893574953079224, "learning_rate": 5.9019632665582004e-05, "loss": 0.0429, "step": 28210 }, { "grad_norm": 0.522995114326477, "learning_rate": 5.899252556177559e-05, "loss": 0.0779, "step": 28220 }, { "grad_norm": 0.3622923791408539, "learning_rate": 5.896541572627735e-05, "loss": 0.0347, "step": 28230 }, { "grad_norm": 0.7087518572807312, "learning_rate": 5.893830316732253e-05, "loss": 0.0514, "step": 28240 }, { "grad_norm": 0.773018479347229, "learning_rate": 5.8911187893147214e-05, "loss": 0.0513, "step": 28250 }, { "grad_norm": 0.47548240423202515, "learning_rate": 5.888406991198828e-05, "loss": 0.0609, "step": 28260 }, { "grad_norm": 0.6146911382675171, "learning_rate": 5.885694923208349e-05, "loss": 0.0562, "step": 28270 }, { "grad_norm": 0.5946487784385681, "learning_rate": 5.882982586167138e-05, "loss": 0.0496, "step": 28280 }, { "grad_norm": 0.38903650641441345, "learning_rate": 5.880269980899131e-05, "loss": 0.0434, "step": 28290 }, { "grad_norm": 0.4401216506958008, "learning_rate": 5.8775571082283465e-05, "loss": 0.0488, "step": 28300 }, { "grad_norm": 1.0827083587646484, "learning_rate": 5.8748439689788824e-05, "loss": 0.0719, "step": 28310 }, { "grad_norm": 0.9260937571525574, "learning_rate": 5.87213056397492e-05, "loss": 0.0588, "step": 28320 }, { "grad_norm": 0.3648529648780823, "learning_rate": 5.869416894040719e-05, "loss": 0.0412, "step": 28330 }, { "grad_norm": 0.7809105515480042, "learning_rate": 5.866702960000621e-05, "loss": 0.0442, "step": 28340 }, { "grad_norm": 0.5126382112503052, "learning_rate": 5.863988762679048e-05, "loss": 0.0461, "step": 28350 }, { "grad_norm": 0.35122138261795044, "learning_rate": 5.8612743029005e-05, "loss": 0.0425, "step": 28360 }, { "grad_norm": 0.566070556640625, "learning_rate": 5.858559581489561e-05, "loss": 0.0495, "step": 28370 }, { "grad_norm": 0.4531247913837433, "learning_rate": 5.85584459927089e-05, "loss": 0.0474, "step": 28380 }, { "grad_norm": 0.33255070447921753, "learning_rate": 5.853129357069227e-05, "loss": 0.0418, "step": 28390 }, { "grad_norm": 0.6547325849533081, "learning_rate": 5.8504138557093913e-05, "loss": 0.0421, "step": 28400 }, { "grad_norm": 0.6337282061576843, "learning_rate": 5.8476980960162784e-05, "loss": 0.0454, "step": 28410 }, { "grad_norm": 0.5202630162239075, "learning_rate": 5.844982078814868e-05, "loss": 0.0458, "step": 28420 }, { "grad_norm": 0.5601754784584045, "learning_rate": 5.842265804930211e-05, "loss": 0.0594, "step": 28430 }, { "grad_norm": 0.3354588449001312, "learning_rate": 5.839549275187444e-05, "loss": 0.0392, "step": 28440 }, { "grad_norm": 1.0386630296707153, "learning_rate": 5.836832490411771e-05, "loss": 0.0516, "step": 28450 }, { "grad_norm": 0.4716501533985138, "learning_rate": 5.834115451428485e-05, "loss": 0.0346, "step": 28460 }, { "grad_norm": 0.4385857582092285, "learning_rate": 5.831398159062946e-05, "loss": 0.0402, "step": 28470 }, { "grad_norm": 0.7740780115127563, "learning_rate": 5.828680614140599e-05, "loss": 0.0493, "step": 28480 }, { "grad_norm": 0.5476470589637756, "learning_rate": 5.825962817486962e-05, "loss": 0.0586, "step": 28490 }, { "grad_norm": 0.5468902587890625, "learning_rate": 5.823244769927629e-05, "loss": 0.0404, "step": 28500 }, { "grad_norm": 0.4868328869342804, "learning_rate": 5.8205264722882716e-05, "loss": 0.0413, "step": 28510 }, { "grad_norm": 0.43765634298324585, "learning_rate": 5.817807925394636e-05, "loss": 0.0458, "step": 28520 }, { "grad_norm": 0.5496204495429993, "learning_rate": 5.815089130072546e-05, "loss": 0.048, "step": 28530 }, { "grad_norm": 0.7153398394584656, "learning_rate": 5.8123700871479e-05, "loss": 0.0503, "step": 28540 }, { "grad_norm": 0.5480192303657532, "learning_rate": 5.809650797446671e-05, "loss": 0.0378, "step": 28550 }, { "grad_norm": 0.6713896989822388, "learning_rate": 5.806931261794907e-05, "loss": 0.0551, "step": 28560 }, { "grad_norm": 0.6506978273391724, "learning_rate": 5.804211481018731e-05, "loss": 0.0476, "step": 28570 }, { "grad_norm": 0.39016959071159363, "learning_rate": 5.801491455944341e-05, "loss": 0.0499, "step": 28580 }, { "grad_norm": 0.555509626865387, "learning_rate": 5.79877118739801e-05, "loss": 0.051, "step": 28590 }, { "grad_norm": 0.4500708281993866, "learning_rate": 5.7960506762060816e-05, "loss": 0.0604, "step": 28600 }, { "grad_norm": 0.4963693916797638, "learning_rate": 5.793329923194977e-05, "loss": 0.0373, "step": 28610 }, { "grad_norm": 0.5741287469863892, "learning_rate": 5.790608929191187e-05, "loss": 0.0422, "step": 28620 }, { "grad_norm": 0.35621407628059387, "learning_rate": 5.78788769502128e-05, "loss": 0.0688, "step": 28630 }, { "grad_norm": 0.444844126701355, "learning_rate": 5.785166221511894e-05, "loss": 0.0324, "step": 28640 }, { "grad_norm": 0.42735564708709717, "learning_rate": 5.7824445094897415e-05, "loss": 0.042, "step": 28650 }, { "grad_norm": 0.4531297981739044, "learning_rate": 5.7797225597816065e-05, "loss": 0.0411, "step": 28660 }, { "grad_norm": 0.4194629490375519, "learning_rate": 5.777000373214345e-05, "loss": 0.0425, "step": 28670 }, { "grad_norm": 0.4268917739391327, "learning_rate": 5.774277950614885e-05, "loss": 0.0558, "step": 28680 }, { "grad_norm": 0.474202960729599, "learning_rate": 5.771555292810227e-05, "loss": 0.0521, "step": 28690 }, { "grad_norm": 0.7537068724632263, "learning_rate": 5.768832400627444e-05, "loss": 0.0616, "step": 28700 }, { "grad_norm": 0.6362098455429077, "learning_rate": 5.7661092748936775e-05, "loss": 0.0408, "step": 28710 }, { "grad_norm": 0.4772559404373169, "learning_rate": 5.76338591643614e-05, "loss": 0.0436, "step": 28720 }, { "grad_norm": 0.6606199741363525, "learning_rate": 5.760662326082118e-05, "loss": 0.0497, "step": 28730 }, { "grad_norm": 0.48456981778144836, "learning_rate": 5.757938504658965e-05, "loss": 0.0514, "step": 28740 }, { "grad_norm": 0.4781826436519623, "learning_rate": 5.755214452994107e-05, "loss": 0.0492, "step": 28750 }, { "grad_norm": 0.7126055955886841, "learning_rate": 5.752490171915039e-05, "loss": 0.0408, "step": 28760 }, { "grad_norm": 0.692227303981781, "learning_rate": 5.749765662249324e-05, "loss": 0.0468, "step": 28770 }, { "grad_norm": 0.45610126852989197, "learning_rate": 5.747040924824596e-05, "loss": 0.0645, "step": 28780 }, { "grad_norm": 0.31918051838874817, "learning_rate": 5.7443159604685613e-05, "loss": 0.0539, "step": 28790 }, { "grad_norm": 0.33529898524284363, "learning_rate": 5.74159077000899e-05, "loss": 0.0471, "step": 28800 }, { "grad_norm": 0.3058365285396576, "learning_rate": 5.7388653542737235e-05, "loss": 0.0357, "step": 28810 }, { "grad_norm": 0.431659996509552, "learning_rate": 5.736139714090672e-05, "loss": 0.0459, "step": 28820 }, { "grad_norm": 0.5711402893066406, "learning_rate": 5.73341385028781e-05, "loss": 0.0445, "step": 28830 }, { "grad_norm": 0.5332794189453125, "learning_rate": 5.7306877636931855e-05, "loss": 0.0417, "step": 28840 }, { "grad_norm": 0.6222005486488342, "learning_rate": 5.7279614551349125e-05, "loss": 0.0464, "step": 28850 }, { "grad_norm": 0.6387922167778015, "learning_rate": 5.725234925441169e-05, "loss": 0.0554, "step": 28860 }, { "grad_norm": 0.45238959789276123, "learning_rate": 5.7225081754402044e-05, "loss": 0.0462, "step": 28870 }, { "grad_norm": 0.39935967326164246, "learning_rate": 5.7197812059603326e-05, "loss": 0.0417, "step": 28880 }, { "grad_norm": 0.44747602939605713, "learning_rate": 5.717054017829934e-05, "loss": 0.05, "step": 28890 }, { "grad_norm": 0.42614486813545227, "learning_rate": 5.7143266118774584e-05, "loss": 0.0341, "step": 28900 }, { "grad_norm": 0.34862014651298523, "learning_rate": 5.711598988931418e-05, "loss": 0.0544, "step": 28910 }, { "grad_norm": 0.3351622223854065, "learning_rate": 5.7088711498203954e-05, "loss": 0.0458, "step": 28920 }, { "grad_norm": 0.5065330266952515, "learning_rate": 5.706143095373033e-05, "loss": 0.0509, "step": 28930 }, { "grad_norm": 0.4568069577217102, "learning_rate": 5.703414826418042e-05, "loss": 0.0562, "step": 28940 }, { "grad_norm": 0.41859325766563416, "learning_rate": 5.7006863437842007e-05, "loss": 0.0487, "step": 28950 }, { "grad_norm": 0.5239872336387634, "learning_rate": 5.697957648300348e-05, "loss": 0.0472, "step": 28960 }, { "grad_norm": 0.2917393147945404, "learning_rate": 5.695228740795391e-05, "loss": 0.0516, "step": 28970 }, { "grad_norm": 0.4768933057785034, "learning_rate": 5.6924996220982985e-05, "loss": 0.0473, "step": 28980 }, { "grad_norm": 0.5044600963592529, "learning_rate": 5.6897702930381045e-05, "loss": 0.0567, "step": 28990 }, { "grad_norm": 0.4411458671092987, "learning_rate": 5.687040754443908e-05, "loss": 0.0494, "step": 29000 }, { "grad_norm": 0.46734529733657837, "learning_rate": 5.6843110071448725e-05, "loss": 0.0378, "step": 29010 }, { "grad_norm": 0.6243112683296204, "learning_rate": 5.6815810519702194e-05, "loss": 0.0459, "step": 29020 }, { "grad_norm": 0.29147157073020935, "learning_rate": 5.6788508897492396e-05, "loss": 0.039, "step": 29030 }, { "grad_norm": 0.67236328125, "learning_rate": 5.676120521311282e-05, "loss": 0.0424, "step": 29040 }, { "grad_norm": 0.4706496298313141, "learning_rate": 5.6733899474857634e-05, "loss": 0.0859, "step": 29050 }, { "grad_norm": 0.47983524203300476, "learning_rate": 5.670659169102157e-05, "loss": 0.0374, "step": 29060 }, { "grad_norm": 0.6299968957901001, "learning_rate": 5.6679281869900044e-05, "loss": 0.0442, "step": 29070 }, { "grad_norm": 0.488033652305603, "learning_rate": 5.6651970019789045e-05, "loss": 0.0489, "step": 29080 }, { "grad_norm": 0.46796441078186035, "learning_rate": 5.662465614898519e-05, "loss": 0.0416, "step": 29090 }, { "grad_norm": 0.4996657967567444, "learning_rate": 5.6597340265785695e-05, "loss": 0.0426, "step": 29100 }, { "grad_norm": 0.517728865146637, "learning_rate": 5.657002237848843e-05, "loss": 0.0383, "step": 29110 }, { "grad_norm": 0.40569400787353516, "learning_rate": 5.654270249539183e-05, "loss": 0.0502, "step": 29120 }, { "grad_norm": 0.5829028487205505, "learning_rate": 5.651538062479498e-05, "loss": 0.0352, "step": 29130 }, { "grad_norm": 0.5900128483772278, "learning_rate": 5.648805677499751e-05, "loss": 0.0416, "step": 29140 }, { "grad_norm": 0.5193242430686951, "learning_rate": 5.646073095429969e-05, "loss": 0.0606, "step": 29150 }, { "grad_norm": 0.4805523455142975, "learning_rate": 5.643340317100241e-05, "loss": 0.0476, "step": 29160 }, { "grad_norm": 0.30915597081184387, "learning_rate": 5.64060734334071e-05, "loss": 0.0546, "step": 29170 }, { "grad_norm": 1.0748512744903564, "learning_rate": 5.637874174981583e-05, "loss": 0.0588, "step": 29180 }, { "grad_norm": 0.48100683093070984, "learning_rate": 5.635140812853124e-05, "loss": 0.0426, "step": 29190 }, { "grad_norm": 0.5736052393913269, "learning_rate": 5.6324072577856544e-05, "loss": 0.0657, "step": 29200 }, { "grad_norm": 0.6761820912361145, "learning_rate": 5.629673510609559e-05, "loss": 0.0366, "step": 29210 }, { "grad_norm": 0.4073341190814972, "learning_rate": 5.626939572155276e-05, "loss": 0.0476, "step": 29220 }, { "grad_norm": 0.47883111238479614, "learning_rate": 5.6242054432533054e-05, "loss": 0.0423, "step": 29230 }, { "grad_norm": 0.6947328448295593, "learning_rate": 5.621471124734201e-05, "loss": 0.0443, "step": 29240 }, { "grad_norm": 0.49338191747665405, "learning_rate": 5.6187366174285794e-05, "loss": 0.0322, "step": 29250 }, { "grad_norm": 0.6055486798286438, "learning_rate": 5.616001922167109e-05, "loss": 0.0573, "step": 29260 }, { "grad_norm": 0.637097179889679, "learning_rate": 5.61326703978052e-05, "loss": 0.0444, "step": 29270 }, { "grad_norm": 0.3937479853630066, "learning_rate": 5.6105319710995964e-05, "loss": 0.0435, "step": 29280 }, { "grad_norm": 0.3567149341106415, "learning_rate": 5.60779671695518e-05, "loss": 0.0454, "step": 29290 }, { "grad_norm": 0.5988240838050842, "learning_rate": 5.6050612781781684e-05, "loss": 0.0567, "step": 29300 }, { "grad_norm": 0.2949821949005127, "learning_rate": 5.602325655599516e-05, "loss": 0.0448, "step": 29310 }, { "grad_norm": 0.6386544108390808, "learning_rate": 5.599589850050234e-05, "loss": 0.0571, "step": 29320 }, { "grad_norm": 0.5841313600540161, "learning_rate": 5.5968538623613874e-05, "loss": 0.053, "step": 29330 }, { "grad_norm": 0.4923174977302551, "learning_rate": 5.594117693364095e-05, "loss": 0.0453, "step": 29340 }, { "grad_norm": 0.5597190260887146, "learning_rate": 5.591381343889535e-05, "loss": 0.0373, "step": 29350 }, { "grad_norm": 0.516710638999939, "learning_rate": 5.5886448147689355e-05, "loss": 0.0404, "step": 29360 }, { "grad_norm": 0.4878368675708771, "learning_rate": 5.585908106833585e-05, "loss": 0.0541, "step": 29370 }, { "grad_norm": 0.6990131139755249, "learning_rate": 5.5831712209148226e-05, "loss": 0.0431, "step": 29380 }, { "grad_norm": 0.2052794247865677, "learning_rate": 5.58043415784404e-05, "loss": 0.0462, "step": 29390 }, { "grad_norm": 0.2740967571735382, "learning_rate": 5.577696918452686e-05, "loss": 0.0568, "step": 29400 }, { "grad_norm": 0.6467986702919006, "learning_rate": 5.5749595035722604e-05, "loss": 0.0512, "step": 29410 }, { "grad_norm": 0.9879193305969238, "learning_rate": 5.5722219140343193e-05, "loss": 0.0639, "step": 29420 }, { "grad_norm": 0.5881907939910889, "learning_rate": 5.56948415067047e-05, "loss": 0.0634, "step": 29430 }, { "grad_norm": 0.836418867111206, "learning_rate": 5.5667462143123704e-05, "loss": 0.0507, "step": 29440 }, { "grad_norm": 0.55141681432724, "learning_rate": 5.564008105791737e-05, "loss": 0.0444, "step": 29450 }, { "grad_norm": 0.6084246635437012, "learning_rate": 5.5612698259403316e-05, "loss": 0.046, "step": 29460 }, { "grad_norm": 0.7089601755142212, "learning_rate": 5.5585313755899724e-05, "loss": 0.052, "step": 29470 }, { "grad_norm": 0.3542991578578949, "learning_rate": 5.5557927555725285e-05, "loss": 0.0444, "step": 29480 }, { "grad_norm": 0.5446370840072632, "learning_rate": 5.55305396671992e-05, "loss": 0.0475, "step": 29490 }, { "grad_norm": 0.3811882436275482, "learning_rate": 5.55031500986412e-05, "loss": 0.0527, "step": 29500 }, { "grad_norm": 0.6721628904342651, "learning_rate": 5.547575885837149e-05, "loss": 0.0451, "step": 29510 }, { "grad_norm": 0.4808589816093445, "learning_rate": 5.5448365954710825e-05, "loss": 0.0477, "step": 29520 }, { "grad_norm": 0.5398595929145813, "learning_rate": 5.5420971395980446e-05, "loss": 0.0428, "step": 29530 }, { "grad_norm": 0.49802327156066895, "learning_rate": 5.539357519050209e-05, "loss": 0.0445, "step": 29540 }, { "grad_norm": 0.3130445182323456, "learning_rate": 5.536617734659799e-05, "loss": 0.0508, "step": 29550 }, { "grad_norm": 0.5406243801116943, "learning_rate": 5.533877787259091e-05, "loss": 0.0426, "step": 29560 }, { "grad_norm": 0.6143929958343506, "learning_rate": 5.5311376776804044e-05, "loss": 0.0492, "step": 29570 }, { "grad_norm": 0.520130455493927, "learning_rate": 5.528397406756118e-05, "loss": 0.0357, "step": 29580 }, { "grad_norm": 0.40757036209106445, "learning_rate": 5.525656975318652e-05, "loss": 0.0353, "step": 29590 }, { "grad_norm": 0.7571279406547546, "learning_rate": 5.522916384200474e-05, "loss": 0.0527, "step": 29600 }, { "grad_norm": 0.36050981283187866, "learning_rate": 5.520175634234106e-05, "loss": 0.0385, "step": 29610 }, { "grad_norm": 0.5266242623329163, "learning_rate": 5.517434726252113e-05, "loss": 0.0446, "step": 29620 }, { "grad_norm": 0.6603888273239136, "learning_rate": 5.514693661087113e-05, "loss": 0.0596, "step": 29630 }, { "grad_norm": 0.5026331543922424, "learning_rate": 5.511952439571769e-05, "loss": 0.0443, "step": 29640 }, { "grad_norm": 0.6046909093856812, "learning_rate": 5.509211062538791e-05, "loss": 0.0662, "step": 29650 }, { "grad_norm": 0.38254791498184204, "learning_rate": 5.506469530820939e-05, "loss": 0.0329, "step": 29660 }, { "grad_norm": 0.42042431235313416, "learning_rate": 5.503727845251014e-05, "loss": 0.093, "step": 29670 }, { "grad_norm": 0.514650821685791, "learning_rate": 5.50098600666187e-05, "loss": 0.0489, "step": 29680 }, { "grad_norm": 0.3905046582221985, "learning_rate": 5.498244015886406e-05, "loss": 0.0428, "step": 29690 }, { "grad_norm": 0.40683096647262573, "learning_rate": 5.495501873757565e-05, "loss": 0.0875, "step": 29700 }, { "grad_norm": 0.36437419056892395, "learning_rate": 5.492759581108336e-05, "loss": 0.0397, "step": 29710 }, { "grad_norm": 0.5174524784088135, "learning_rate": 5.490017138771759e-05, "loss": 0.0443, "step": 29720 }, { "grad_norm": 0.5131310224533081, "learning_rate": 5.487274547580912e-05, "loss": 0.0495, "step": 29730 }, { "grad_norm": 0.47613465785980225, "learning_rate": 5.484531808368923e-05, "loss": 0.0444, "step": 29740 }, { "grad_norm": 0.6132310032844543, "learning_rate": 5.4817889219689656e-05, "loss": 0.0407, "step": 29750 }, { "grad_norm": 0.7424232363700867, "learning_rate": 5.4790458892142536e-05, "loss": 0.0541, "step": 29760 }, { "grad_norm": 0.5310806632041931, "learning_rate": 5.476302710938048e-05, "loss": 0.0398, "step": 29770 }, { "grad_norm": 0.6294403076171875, "learning_rate": 5.473559387973657e-05, "loss": 0.0376, "step": 29780 }, { "grad_norm": 0.6408951878547668, "learning_rate": 5.470815921154425e-05, "loss": 0.0452, "step": 29790 }, { "grad_norm": 0.595008373260498, "learning_rate": 5.468072311313749e-05, "loss": 0.0435, "step": 29800 }, { "grad_norm": 0.6181927919387817, "learning_rate": 5.465328559285063e-05, "loss": 0.0491, "step": 29810 }, { "grad_norm": 0.6031759977340698, "learning_rate": 5.462584665901849e-05, "loss": 0.05, "step": 29820 }, { "grad_norm": 0.49639591574668884, "learning_rate": 5.4598406319976235e-05, "loss": 0.0464, "step": 29830 }, { "grad_norm": 0.6455107927322388, "learning_rate": 5.457096458405958e-05, "loss": 0.0483, "step": 29840 }, { "grad_norm": 0.6525421738624573, "learning_rate": 5.454352145960457e-05, "loss": 0.0501, "step": 29850 }, { "grad_norm": 0.5259671211242676, "learning_rate": 5.4516076954947715e-05, "loss": 0.045, "step": 29860 }, { "grad_norm": 0.5554500222206116, "learning_rate": 5.448863107842591e-05, "loss": 0.0532, "step": 29870 }, { "grad_norm": 0.6778960227966309, "learning_rate": 5.446118383837651e-05, "loss": 0.06, "step": 29880 }, { "grad_norm": 0.5066270232200623, "learning_rate": 5.443373524313722e-05, "loss": 0.0513, "step": 29890 }, { "grad_norm": 0.9555484652519226, "learning_rate": 5.440628530104626e-05, "loss": 0.0929, "step": 29900 }, { "grad_norm": 0.5758371353149414, "learning_rate": 5.4378834020442146e-05, "loss": 0.0386, "step": 29910 }, { "grad_norm": 0.571020245552063, "learning_rate": 5.4351381409663884e-05, "loss": 0.0555, "step": 29920 }, { "grad_norm": 0.42429959774017334, "learning_rate": 5.432392747705084e-05, "loss": 0.0377, "step": 29930 }, { "grad_norm": 0.5518330931663513, "learning_rate": 5.429647223094278e-05, "loss": 0.0512, "step": 29940 }, { "grad_norm": 0.5533165335655212, "learning_rate": 5.4269015679679924e-05, "loss": 0.0485, "step": 29950 }, { "grad_norm": 0.4797215759754181, "learning_rate": 5.424155783160281e-05, "loss": 0.049, "step": 29960 }, { "grad_norm": 0.35990265011787415, "learning_rate": 5.4214098695052415e-05, "loss": 0.0419, "step": 29970 }, { "grad_norm": 0.45405513048171997, "learning_rate": 5.418663827837012e-05, "loss": 0.0508, "step": 29980 }, { "grad_norm": 0.47870251536369324, "learning_rate": 5.415917658989763e-05, "loss": 0.0512, "step": 29990 }, { "grad_norm": 0.6768524646759033, "learning_rate": 5.413171363797713e-05, "loss": 0.0642, "step": 30000 }, { "grad_norm": 0.6085295677185059, "learning_rate": 5.4104249430951116e-05, "loss": 0.0451, "step": 30010 }, { "grad_norm": 0.24764467775821686, "learning_rate": 5.4076783977162494e-05, "loss": 0.0356, "step": 30020 }, { "grad_norm": 0.48662978410720825, "learning_rate": 5.4049317284954525e-05, "loss": 0.0377, "step": 30030 }, { "grad_norm": 0.5079306960105896, "learning_rate": 5.4021849362670884e-05, "loss": 0.0487, "step": 30040 }, { "grad_norm": 0.546108067035675, "learning_rate": 5.3994380218655604e-05, "loss": 0.0396, "step": 30050 }, { "grad_norm": 0.617510974407196, "learning_rate": 5.396690986125309e-05, "loss": 0.0537, "step": 30060 }, { "grad_norm": 0.5585018992424011, "learning_rate": 5.3939438298808075e-05, "loss": 0.079, "step": 30070 }, { "grad_norm": 0.587240993976593, "learning_rate": 5.3911965539665744e-05, "loss": 0.0375, "step": 30080 }, { "grad_norm": 0.4209630489349365, "learning_rate": 5.388449159217156e-05, "loss": 0.0516, "step": 30090 }, { "grad_norm": 0.5277723073959351, "learning_rate": 5.3857016464671385e-05, "loss": 0.0647, "step": 30100 }, { "grad_norm": 0.7279467582702637, "learning_rate": 5.382954016551146e-05, "loss": 0.0543, "step": 30110 }, { "grad_norm": 0.43830686807632446, "learning_rate": 5.380206270303835e-05, "loss": 0.0475, "step": 30120 }, { "grad_norm": 0.4427991211414337, "learning_rate": 5.377458408559897e-05, "loss": 0.0366, "step": 30130 }, { "grad_norm": 0.3690774738788605, "learning_rate": 5.374710432154061e-05, "loss": 0.0549, "step": 30140 }, { "grad_norm": 0.46286845207214355, "learning_rate": 5.3719623419210886e-05, "loss": 0.0384, "step": 30150 }, { "grad_norm": 0.45609381794929504, "learning_rate": 5.3692141386957786e-05, "loss": 0.0453, "step": 30160 }, { "grad_norm": 0.7620368003845215, "learning_rate": 5.3664658233129616e-05, "loss": 0.0544, "step": 30170 }, { "grad_norm": 0.49176979064941406, "learning_rate": 5.363717396607504e-05, "loss": 0.049, "step": 30180 }, { "grad_norm": 0.4408179223537445, "learning_rate": 5.360968859414305e-05, "loss": 0.0472, "step": 30190 }, { "grad_norm": 0.32698044180870056, "learning_rate": 5.358220212568295e-05, "loss": 0.0337, "step": 30200 }, { "grad_norm": 0.5472429394721985, "learning_rate": 5.355471456904444e-05, "loss": 0.0415, "step": 30210 }, { "grad_norm": 0.350976824760437, "learning_rate": 5.3527225932577495e-05, "loss": 0.0503, "step": 30220 }, { "grad_norm": 0.5846177935600281, "learning_rate": 5.349973622463246e-05, "loss": 0.0449, "step": 30230 }, { "grad_norm": 0.5259000658988953, "learning_rate": 5.3472245453559956e-05, "loss": 0.0453, "step": 30240 }, { "grad_norm": 0.4502507448196411, "learning_rate": 5.3444753627710955e-05, "loss": 0.0519, "step": 30250 }, { "grad_norm": 0.4078930616378784, "learning_rate": 5.341726075543676e-05, "loss": 0.0449, "step": 30260 }, { "grad_norm": 0.5265432596206665, "learning_rate": 5.338976684508898e-05, "loss": 0.0313, "step": 30270 }, { "grad_norm": 0.6089781522750854, "learning_rate": 5.336227190501953e-05, "loss": 0.0329, "step": 30280 }, { "grad_norm": 0.5901477932929993, "learning_rate": 5.3334775943580664e-05, "loss": 0.0487, "step": 30290 }, { "grad_norm": 0.5097755789756775, "learning_rate": 5.330727896912491e-05, "loss": 0.0381, "step": 30300 }, { "grad_norm": 0.43254879117012024, "learning_rate": 5.327978099000511e-05, "loss": 0.0413, "step": 30310 }, { "grad_norm": 0.6177141666412354, "learning_rate": 5.3252282014574465e-05, "loss": 0.054, "step": 30320 }, { "grad_norm": 0.6027111411094666, "learning_rate": 5.322478205118641e-05, "loss": 0.0508, "step": 30330 }, { "grad_norm": 0.5219056010246277, "learning_rate": 5.3197281108194704e-05, "loss": 0.0668, "step": 30340 }, { "grad_norm": 0.5540003776550293, "learning_rate": 5.316977919395342e-05, "loss": 0.0478, "step": 30350 }, { "grad_norm": 0.4433039724826813, "learning_rate": 5.314227631681691e-05, "loss": 0.0424, "step": 30360 }, { "grad_norm": 0.4535791575908661, "learning_rate": 5.311477248513982e-05, "loss": 0.0481, "step": 30370 }, { "grad_norm": 0.39547064900398254, "learning_rate": 5.30872677072771e-05, "loss": 0.0412, "step": 30380 }, { "grad_norm": 0.7736260294914246, "learning_rate": 5.3059761991583954e-05, "loss": 0.0758, "step": 30390 }, { "grad_norm": 0.42199820280075073, "learning_rate": 5.303225534641592e-05, "loss": 0.0436, "step": 30400 }, { "grad_norm": 0.5278431177139282, "learning_rate": 5.300474778012875e-05, "loss": 0.0449, "step": 30410 }, { "grad_norm": 0.4406886696815491, "learning_rate": 5.297723930107855e-05, "loss": 0.0464, "step": 30420 }, { "grad_norm": 0.5915679335594177, "learning_rate": 5.294972991762167e-05, "loss": 0.0411, "step": 30430 }, { "grad_norm": 0.9509955644607544, "learning_rate": 5.292221963811472e-05, "loss": 0.0537, "step": 30440 }, { "grad_norm": 0.4540323317050934, "learning_rate": 5.28947084709146e-05, "loss": 0.0373, "step": 30450 }, { "grad_norm": 0.7745444178581238, "learning_rate": 5.2867196424378465e-05, "loss": 0.0379, "step": 30460 }, { "grad_norm": 0.40161824226379395, "learning_rate": 5.2839683506863765e-05, "loss": 0.039, "step": 30470 }, { "grad_norm": 0.33712324500083923, "learning_rate": 5.281216972672821e-05, "loss": 0.0405, "step": 30480 }, { "grad_norm": 0.5096070170402527, "learning_rate": 5.278465509232973e-05, "loss": 0.0375, "step": 30490 }, { "grad_norm": 0.4384007751941681, "learning_rate": 5.275713961202655e-05, "loss": 0.0813, "step": 30500 }, { "grad_norm": 0.42343106865882874, "learning_rate": 5.2729623294177165e-05, "loss": 0.0398, "step": 30510 }, { "grad_norm": 0.5128863453865051, "learning_rate": 5.270210614714028e-05, "loss": 0.0449, "step": 30520 }, { "grad_norm": 0.5226965546607971, "learning_rate": 5.267458817927491e-05, "loss": 0.0461, "step": 30530 }, { "grad_norm": 0.4567364454269409, "learning_rate": 5.264706939894026e-05, "loss": 0.0414, "step": 30540 }, { "grad_norm": 0.45287632942199707, "learning_rate": 5.261954981449584e-05, "loss": 0.0408, "step": 30550 }, { "grad_norm": 0.4568250775337219, "learning_rate": 5.2592029434301324e-05, "loss": 0.0383, "step": 30560 }, { "grad_norm": 0.5492931008338928, "learning_rate": 5.256450826671672e-05, "loss": 0.0637, "step": 30570 }, { "grad_norm": 0.47479355335235596, "learning_rate": 5.253698632010221e-05, "loss": 0.0537, "step": 30580 }, { "grad_norm": 0.8779582977294922, "learning_rate": 5.2509463602818246e-05, "loss": 0.0438, "step": 30590 }, { "grad_norm": 0.6169043779373169, "learning_rate": 5.248194012322549e-05, "loss": 0.0475, "step": 30600 }, { "grad_norm": 0.36040446162223816, "learning_rate": 5.245441588968486e-05, "loss": 0.059, "step": 30610 }, { "grad_norm": 0.5364298820495605, "learning_rate": 5.242689091055748e-05, "loss": 0.0402, "step": 30620 }, { "grad_norm": 0.5196921825408936, "learning_rate": 5.239936519420473e-05, "loss": 0.0428, "step": 30630 }, { "grad_norm": 0.499859094619751, "learning_rate": 5.2371838748988175e-05, "loss": 0.0292, "step": 30640 }, { "grad_norm": 0.9483524560928345, "learning_rate": 5.234431158326965e-05, "loss": 0.0455, "step": 30650 }, { "grad_norm": 0.5595124959945679, "learning_rate": 5.231678370541115e-05, "loss": 0.0396, "step": 30660 }, { "grad_norm": 0.35027098655700684, "learning_rate": 5.228925512377495e-05, "loss": 0.0394, "step": 30670 }, { "grad_norm": 0.538703441619873, "learning_rate": 5.2261725846723465e-05, "loss": 0.0518, "step": 30680 }, { "grad_norm": 0.5503232479095459, "learning_rate": 5.22341958826194e-05, "loss": 0.0453, "step": 30690 }, { "grad_norm": 0.39670267701148987, "learning_rate": 5.22066652398256e-05, "loss": 0.0555, "step": 30700 }, { "grad_norm": 0.4095088243484497, "learning_rate": 5.2179133926705185e-05, "loss": 0.0369, "step": 30710 }, { "grad_norm": 0.3595002591609955, "learning_rate": 5.215160195162141e-05, "loss": 0.046, "step": 30720 }, { "grad_norm": 0.7226818799972534, "learning_rate": 5.212406932293776e-05, "loss": 0.0581, "step": 30730 }, { "grad_norm": 0.43391454219818115, "learning_rate": 5.209653604901795e-05, "loss": 0.0342, "step": 30740 }, { "grad_norm": 0.48148444294929504, "learning_rate": 5.206900213822584e-05, "loss": 0.0481, "step": 30750 }, { "grad_norm": 0.3950108289718628, "learning_rate": 5.204146759892551e-05, "loss": 0.0456, "step": 30760 }, { "grad_norm": 0.7580825686454773, "learning_rate": 5.2013932439481216e-05, "loss": 0.0376, "step": 30770 }, { "grad_norm": 0.8801895380020142, "learning_rate": 5.198639666825743e-05, "loss": 0.0513, "step": 30780 }, { "grad_norm": 0.5909494161605835, "learning_rate": 5.195886029361877e-05, "loss": 0.0534, "step": 30790 }, { "grad_norm": 0.47489356994628906, "learning_rate": 5.193132332393009e-05, "loss": 0.0442, "step": 30800 }, { "grad_norm": 0.42561134696006775, "learning_rate": 5.1903785767556376e-05, "loss": 0.0521, "step": 30810 }, { "grad_norm": 0.5609509944915771, "learning_rate": 5.187624763286282e-05, "loss": 0.0374, "step": 30820 }, { "grad_norm": 0.6644941568374634, "learning_rate": 5.184870892821475e-05, "loss": 0.0478, "step": 30830 }, { "grad_norm": 0.47598329186439514, "learning_rate": 5.182116966197773e-05, "loss": 0.0463, "step": 30840 }, { "grad_norm": 0.42285850644111633, "learning_rate": 5.1793629842517466e-05, "loss": 0.0534, "step": 30850 }, { "grad_norm": 0.5468785762786865, "learning_rate": 5.17660894781998e-05, "loss": 0.0689, "step": 30860 }, { "grad_norm": 0.4846310317516327, "learning_rate": 5.173854857739079e-05, "loss": 0.0448, "step": 30870 }, { "grad_norm": 0.5192315578460693, "learning_rate": 5.171100714845661e-05, "loss": 0.0534, "step": 30880 }, { "grad_norm": 0.5028902292251587, "learning_rate": 5.1683465199763646e-05, "loss": 0.0575, "step": 30890 }, { "grad_norm": 0.7394407987594604, "learning_rate": 5.16559227396784e-05, "loss": 0.043, "step": 30900 }, { "grad_norm": 0.4298951327800751, "learning_rate": 5.1628379776567556e-05, "loss": 0.0379, "step": 30910 }, { "grad_norm": 0.5765790343284607, "learning_rate": 5.160083631879792e-05, "loss": 0.0585, "step": 30920 }, { "grad_norm": 0.7637831568717957, "learning_rate": 5.1573292374736484e-05, "loss": 0.0457, "step": 30930 }, { "grad_norm": 0.22086670994758606, "learning_rate": 5.1545747952750356e-05, "loss": 0.0388, "step": 30940 }, { "grad_norm": 0.4890054166316986, "learning_rate": 5.151820306120682e-05, "loss": 0.0352, "step": 30950 }, { "grad_norm": 0.8838196992874146, "learning_rate": 5.149065770847328e-05, "loss": 0.0506, "step": 30960 }, { "grad_norm": 0.5263840556144714, "learning_rate": 5.1463111902917297e-05, "loss": 0.0487, "step": 30970 }, { "grad_norm": 0.6265696883201599, "learning_rate": 5.143556565290654e-05, "loss": 0.0437, "step": 30980 }, { "grad_norm": 0.5152264833450317, "learning_rate": 5.140801896680882e-05, "loss": 0.0404, "step": 30990 }, { "grad_norm": 0.6909259557723999, "learning_rate": 5.1380471852992144e-05, "loss": 0.0552, "step": 31000 }, { "grad_norm": 0.8410260677337646, "learning_rate": 5.135292431982457e-05, "loss": 0.0609, "step": 31010 }, { "grad_norm": 0.370138943195343, "learning_rate": 5.1325376375674294e-05, "loss": 0.043, "step": 31020 }, { "grad_norm": 0.5028444528579712, "learning_rate": 5.129782802890968e-05, "loss": 0.0411, "step": 31030 }, { "grad_norm": 0.8278388381004333, "learning_rate": 5.127027928789916e-05, "loss": 0.0452, "step": 31040 }, { "grad_norm": 0.41557222604751587, "learning_rate": 5.124273016101135e-05, "loss": 0.0414, "step": 31050 }, { "grad_norm": 0.587500810623169, "learning_rate": 5.121518065661492e-05, "loss": 0.0467, "step": 31060 }, { "grad_norm": 0.4424954056739807, "learning_rate": 5.11876307830787e-05, "loss": 0.0467, "step": 31070 }, { "grad_norm": 0.2678564190864563, "learning_rate": 5.1160080548771596e-05, "loss": 0.0353, "step": 31080 }, { "grad_norm": 0.699196457862854, "learning_rate": 5.1132529962062656e-05, "loss": 0.0471, "step": 31090 }, { "grad_norm": 0.31958842277526855, "learning_rate": 5.110497903132101e-05, "loss": 0.0497, "step": 31100 }, { "grad_norm": 0.7003856897354126, "learning_rate": 5.107742776491592e-05, "loss": 0.0628, "step": 31110 }, { "grad_norm": 0.6048109531402588, "learning_rate": 5.104987617121673e-05, "loss": 0.0377, "step": 31120 }, { "grad_norm": 0.7227010130882263, "learning_rate": 5.102232425859287e-05, "loss": 0.0505, "step": 31130 }, { "grad_norm": 0.4313138723373413, "learning_rate": 5.09947720354139e-05, "loss": 0.047, "step": 31140 }, { "grad_norm": 0.5776186585426331, "learning_rate": 5.096721951004942e-05, "loss": 0.0451, "step": 31150 }, { "grad_norm": 0.48154014348983765, "learning_rate": 5.0939666690869227e-05, "loss": 0.0662, "step": 31160 }, { "grad_norm": 0.8493842482566833, "learning_rate": 5.0912113586243096e-05, "loss": 0.0473, "step": 31170 }, { "grad_norm": 0.4482273459434509, "learning_rate": 5.0884560204540935e-05, "loss": 0.046, "step": 31180 }, { "grad_norm": 0.4537147879600525, "learning_rate": 5.0857006554132736e-05, "loss": 0.037, "step": 31190 }, { "grad_norm": 0.431426078081131, "learning_rate": 5.0829452643388575e-05, "loss": 0.0437, "step": 31200 }, { "grad_norm": 0.4036746025085449, "learning_rate": 5.08018984806786e-05, "loss": 0.0426, "step": 31210 }, { "grad_norm": 0.588718593120575, "learning_rate": 5.0774344074373036e-05, "loss": 0.0467, "step": 31220 }, { "grad_norm": 0.7467973232269287, "learning_rate": 5.07467894328422e-05, "loss": 0.0497, "step": 31230 }, { "grad_norm": 0.5506489872932434, "learning_rate": 5.0719234564456454e-05, "loss": 0.0437, "step": 31240 }, { "grad_norm": 0.839850127696991, "learning_rate": 5.0691679477586216e-05, "loss": 0.0498, "step": 31250 }, { "grad_norm": 0.6625369191169739, "learning_rate": 5.0664124180602035e-05, "loss": 0.0522, "step": 31260 }, { "grad_norm": 0.5055615305900574, "learning_rate": 5.063656868187447e-05, "loss": 0.0426, "step": 31270 }, { "grad_norm": 0.4062305986881256, "learning_rate": 5.060901298977413e-05, "loss": 0.0368, "step": 31280 }, { "grad_norm": 0.4682692289352417, "learning_rate": 5.0581457112671725e-05, "loss": 0.0429, "step": 31290 }, { "grad_norm": 0.34335586428642273, "learning_rate": 5.0553901058938016e-05, "loss": 0.0365, "step": 31300 }, { "grad_norm": 0.636486828327179, "learning_rate": 5.052634483694377e-05, "loss": 0.0461, "step": 31310 }, { "grad_norm": 0.6692928671836853, "learning_rate": 5.049878845505988e-05, "loss": 0.0542, "step": 31320 }, { "grad_norm": 0.5973361134529114, "learning_rate": 5.047123192165721e-05, "loss": 0.0418, "step": 31330 }, { "grad_norm": 0.6623035669326782, "learning_rate": 5.0443675245106735e-05, "loss": 0.0394, "step": 31340 }, { "grad_norm": 0.49334055185317993, "learning_rate": 5.0416118433779426e-05, "loss": 0.0354, "step": 31350 }, { "grad_norm": 0.38997310400009155, "learning_rate": 5.038856149604633e-05, "loss": 0.0467, "step": 31360 }, { "grad_norm": 0.4743114411830902, "learning_rate": 5.03610044402785e-05, "loss": 0.0422, "step": 31370 }, { "grad_norm": 0.5639479756355286, "learning_rate": 5.033344727484707e-05, "loss": 0.0512, "step": 31380 }, { "grad_norm": 0.415780633687973, "learning_rate": 5.030589000812315e-05, "loss": 0.0417, "step": 31390 }, { "grad_norm": 0.7355579733848572, "learning_rate": 5.027833264847793e-05, "loss": 0.0426, "step": 31400 }, { "grad_norm": 0.5290705561637878, "learning_rate": 5.025077520428258e-05, "loss": 0.0441, "step": 31410 }, { "grad_norm": 0.528566300868988, "learning_rate": 5.022321768390837e-05, "loss": 0.0552, "step": 31420 }, { "grad_norm": 0.7829656600952148, "learning_rate": 5.0195660095726516e-05, "loss": 0.0446, "step": 31430 }, { "grad_norm": 0.4684756398200989, "learning_rate": 5.016810244810829e-05, "loss": 0.0429, "step": 31440 }, { "grad_norm": 0.6997575759887695, "learning_rate": 5.0140544749424976e-05, "loss": 0.0532, "step": 31450 }, { "grad_norm": 0.5098220109939575, "learning_rate": 5.0112987008047874e-05, "loss": 0.0546, "step": 31460 }, { "grad_norm": 0.4451812505722046, "learning_rate": 5.008542923234831e-05, "loss": 0.0449, "step": 31470 }, { "grad_norm": 0.6815372705459595, "learning_rate": 5.00578714306976e-05, "loss": 0.0351, "step": 31480 }, { "grad_norm": 0.4889306426048279, "learning_rate": 5.0030313611467084e-05, "loss": 0.0532, "step": 31490 }, { "grad_norm": 0.5313901901245117, "learning_rate": 5.0002755783028074e-05, "loss": 0.0459, "step": 31500 }, { "grad_norm": 0.5193480849266052, "learning_rate": 4.997519795375194e-05, "loss": 0.041, "step": 31510 }, { "grad_norm": 0.6226868629455566, "learning_rate": 4.9947640132010016e-05, "loss": 0.0461, "step": 31520 }, { "grad_norm": 0.3753402531147003, "learning_rate": 4.9920082326173625e-05, "loss": 0.0363, "step": 31530 }, { "grad_norm": 0.4474104642868042, "learning_rate": 4.9892524544614114e-05, "loss": 0.0572, "step": 31540 }, { "grad_norm": 0.4711892306804657, "learning_rate": 4.986496679570283e-05, "loss": 0.0501, "step": 31550 }, { "grad_norm": 0.42222127318382263, "learning_rate": 4.983740908781105e-05, "loss": 0.0363, "step": 31560 }, { "grad_norm": 0.5641884207725525, "learning_rate": 4.9809851429310116e-05, "loss": 0.0687, "step": 31570 }, { "grad_norm": 0.6553277969360352, "learning_rate": 4.9782293828571275e-05, "loss": 0.0569, "step": 31580 }, { "grad_norm": 0.6417326331138611, "learning_rate": 4.9754736293965846e-05, "loss": 0.0485, "step": 31590 }, { "grad_norm": 0.3893803656101227, "learning_rate": 4.972717883386502e-05, "loss": 0.0422, "step": 31600 }, { "grad_norm": 0.4310058057308197, "learning_rate": 4.9699621456640075e-05, "loss": 0.0472, "step": 31610 }, { "grad_norm": 0.4187227189540863, "learning_rate": 4.9672064170662214e-05, "loss": 0.0467, "step": 31620 }, { "grad_norm": 0.5257022380828857, "learning_rate": 4.9644506984302583e-05, "loss": 0.0406, "step": 31630 }, { "grad_norm": 0.5866872072219849, "learning_rate": 4.9616949905932356e-05, "loss": 0.0477, "step": 31640 }, { "grad_norm": 0.5029380321502686, "learning_rate": 4.9589392943922615e-05, "loss": 0.0421, "step": 31650 }, { "grad_norm": 0.6301547884941101, "learning_rate": 4.956183610664447e-05, "loss": 0.0433, "step": 31660 }, { "grad_norm": 0.550909698009491, "learning_rate": 4.9534279402468945e-05, "loss": 0.0327, "step": 31670 }, { "grad_norm": 0.5856013894081116, "learning_rate": 4.9506722839767036e-05, "loss": 0.0374, "step": 31680 }, { "grad_norm": 0.8412038087844849, "learning_rate": 4.947916642690972e-05, "loss": 0.0446, "step": 31690 }, { "grad_norm": 0.7750670313835144, "learning_rate": 4.9451610172267874e-05, "loss": 0.0426, "step": 31700 }, { "grad_norm": 0.5852132439613342, "learning_rate": 4.9424054084212376e-05, "loss": 0.0411, "step": 31710 }, { "grad_norm": 0.6317058205604553, "learning_rate": 4.939649817111407e-05, "loss": 0.041, "step": 31720 }, { "grad_norm": 0.74818354845047, "learning_rate": 4.936894244134365e-05, "loss": 0.0587, "step": 31730 }, { "grad_norm": 0.4321196675300598, "learning_rate": 4.9341386903271886e-05, "loss": 0.0421, "step": 31740 }, { "grad_norm": 0.7674451470375061, "learning_rate": 4.931383156526936e-05, "loss": 0.0693, "step": 31750 }, { "grad_norm": 0.6126833558082581, "learning_rate": 4.92862764357067e-05, "loss": 0.0438, "step": 31760 }, { "grad_norm": 0.4491739869117737, "learning_rate": 4.925872152295443e-05, "loss": 0.0448, "step": 31770 }, { "grad_norm": 0.6607751846313477, "learning_rate": 4.923116683538296e-05, "loss": 0.0448, "step": 31780 }, { "grad_norm": 0.41370347142219543, "learning_rate": 4.920361238136273e-05, "loss": 0.036, "step": 31790 }, { "grad_norm": 0.5320568084716797, "learning_rate": 4.9176058169264014e-05, "loss": 0.0378, "step": 31800 }, { "grad_norm": 0.5961381793022156, "learning_rate": 4.9148504207457074e-05, "loss": 0.0413, "step": 31810 }, { "grad_norm": 0.5718346238136292, "learning_rate": 4.912095050431208e-05, "loss": 0.0352, "step": 31820 }, { "grad_norm": 0.5496881008148193, "learning_rate": 4.909339706819911e-05, "loss": 0.0501, "step": 31830 }, { "grad_norm": 0.6305028796195984, "learning_rate": 4.906584390748819e-05, "loss": 0.0561, "step": 31840 }, { "grad_norm": 0.5697990655899048, "learning_rate": 4.9038291030549195e-05, "loss": 0.0514, "step": 31850 }, { "grad_norm": 0.489263653755188, "learning_rate": 4.9010738445751995e-05, "loss": 0.0477, "step": 31860 }, { "grad_norm": 0.39732587337493896, "learning_rate": 4.8983186161466364e-05, "loss": 0.0459, "step": 31870 }, { "grad_norm": 0.606695294380188, "learning_rate": 4.89556341860619e-05, "loss": 0.0508, "step": 31880 }, { "grad_norm": 0.47914862632751465, "learning_rate": 4.892808252790822e-05, "loss": 0.0528, "step": 31890 }, { "grad_norm": 0.6052632927894592, "learning_rate": 4.890053119537475e-05, "loss": 0.0393, "step": 31900 }, { "grad_norm": 0.46952202916145325, "learning_rate": 4.887298019683087e-05, "loss": 0.0409, "step": 31910 }, { "grad_norm": 0.37084582448005676, "learning_rate": 4.884542954064587e-05, "loss": 0.0397, "step": 31920 }, { "grad_norm": 0.3942593038082123, "learning_rate": 4.881787923518887e-05, "loss": 0.0379, "step": 31930 }, { "grad_norm": 0.6087676882743835, "learning_rate": 4.879032928882896e-05, "loss": 0.0558, "step": 31940 }, { "grad_norm": 0.4553002715110779, "learning_rate": 4.876277970993505e-05, "loss": 0.0363, "step": 31950 }, { "grad_norm": 0.4371059536933899, "learning_rate": 4.873523050687602e-05, "loss": 0.0359, "step": 31960 }, { "grad_norm": 0.4531308710575104, "learning_rate": 4.870768168802056e-05, "loss": 0.0481, "step": 31970 }, { "grad_norm": 0.448386013507843, "learning_rate": 4.868013326173728e-05, "loss": 0.0469, "step": 31980 }, { "grad_norm": 0.5740199089050293, "learning_rate": 4.865258523639468e-05, "loss": 0.0457, "step": 31990 }, { "grad_norm": 0.4377584755420685, "learning_rate": 4.862503762036109e-05, "loss": 0.0395, "step": 32000 }, { "grad_norm": 0.7124433517456055, "learning_rate": 4.859749042200478e-05, "loss": 0.0432, "step": 32010 }, { "grad_norm": 0.6669111251831055, "learning_rate": 4.856994364969384e-05, "loss": 0.035, "step": 32020 }, { "grad_norm": 0.41109806299209595, "learning_rate": 4.854239731179625e-05, "loss": 0.0407, "step": 32030 }, { "grad_norm": 0.6526665687561035, "learning_rate": 4.85148514166799e-05, "loss": 0.0334, "step": 32040 }, { "grad_norm": 0.6140821576118469, "learning_rate": 4.8487305972712456e-05, "loss": 0.0434, "step": 32050 }, { "grad_norm": 0.39830541610717773, "learning_rate": 4.8459760988261526e-05, "loss": 0.034, "step": 32060 }, { "grad_norm": 0.5372427701950073, "learning_rate": 4.843221647169453e-05, "loss": 0.036, "step": 32070 }, { "grad_norm": 0.3468855917453766, "learning_rate": 4.840467243137878e-05, "loss": 0.0465, "step": 32080 }, { "grad_norm": 0.3784947097301483, "learning_rate": 4.837712887568143e-05, "loss": 0.0343, "step": 32090 }, { "grad_norm": 0.39902180433273315, "learning_rate": 4.8349585812969464e-05, "loss": 0.0447, "step": 32100 }, { "grad_norm": 0.8048713803291321, "learning_rate": 4.8322043251609775e-05, "loss": 0.037, "step": 32110 }, { "grad_norm": 0.47104766964912415, "learning_rate": 4.8294501199969015e-05, "loss": 0.0457, "step": 32120 }, { "grad_norm": 0.5254197716712952, "learning_rate": 4.826695966641376e-05, "loss": 0.0382, "step": 32130 }, { "grad_norm": 0.5957725644111633, "learning_rate": 4.823941865931043e-05, "loss": 0.0485, "step": 32140 }, { "grad_norm": 0.4614100754261017, "learning_rate": 4.82118781870252e-05, "loss": 0.0367, "step": 32150 }, { "grad_norm": 0.5752915740013123, "learning_rate": 4.8184338257924185e-05, "loss": 0.0438, "step": 32160 }, { "grad_norm": 0.5633243322372437, "learning_rate": 4.815679888037324e-05, "loss": 0.0489, "step": 32170 }, { "grad_norm": 0.5334414839744568, "learning_rate": 4.8129260062738135e-05, "loss": 0.0439, "step": 32180 }, { "grad_norm": 0.5843597650527954, "learning_rate": 4.810172181338445e-05, "loss": 0.0483, "step": 32190 }, { "grad_norm": 0.714172899723053, "learning_rate": 4.807418414067753e-05, "loss": 0.0437, "step": 32200 }, { "grad_norm": 0.6472302675247192, "learning_rate": 4.804664705298264e-05, "loss": 0.0465, "step": 32210 }, { "grad_norm": 0.5824316143989563, "learning_rate": 4.80191105586648e-05, "loss": 0.0385, "step": 32220 }, { "grad_norm": 0.4743952453136444, "learning_rate": 4.799157466608886e-05, "loss": 0.0515, "step": 32230 }, { "grad_norm": 0.42546677589416504, "learning_rate": 4.796403938361951e-05, "loss": 0.0427, "step": 32240 }, { "grad_norm": 0.5097740888595581, "learning_rate": 4.793650471962123e-05, "loss": 0.0441, "step": 32250 }, { "grad_norm": 0.4996458888053894, "learning_rate": 4.790897068245835e-05, "loss": 0.057, "step": 32260 }, { "grad_norm": 0.35188665986061096, "learning_rate": 4.7881437280494954e-05, "loss": 0.0312, "step": 32270 }, { "grad_norm": 0.4392443597316742, "learning_rate": 4.7853904522094965e-05, "loss": 0.0392, "step": 32280 }, { "grad_norm": 0.5389516949653625, "learning_rate": 4.782637241562215e-05, "loss": 0.0433, "step": 32290 }, { "grad_norm": 0.34419578313827515, "learning_rate": 4.779884096943997e-05, "loss": 0.0467, "step": 32300 }, { "grad_norm": 0.6657148599624634, "learning_rate": 4.777131019191182e-05, "loss": 0.0467, "step": 32310 }, { "grad_norm": 0.3786627948284149, "learning_rate": 4.774378009140076e-05, "loss": 0.0351, "step": 32320 }, { "grad_norm": 0.559378981590271, "learning_rate": 4.7716250676269735e-05, "loss": 0.0395, "step": 32330 }, { "grad_norm": 0.3654731810092926, "learning_rate": 4.7688721954881485e-05, "loss": 0.0534, "step": 32340 }, { "grad_norm": 1.0541050434112549, "learning_rate": 4.7661193935598446e-05, "loss": 0.0385, "step": 32350 }, { "grad_norm": 0.38436636328697205, "learning_rate": 4.763366662678296e-05, "loss": 0.0289, "step": 32360 }, { "grad_norm": 0.5011301040649414, "learning_rate": 4.7606140036797064e-05, "loss": 0.0343, "step": 32370 }, { "grad_norm": 0.632178544998169, "learning_rate": 4.7578614174002614e-05, "loss": 0.0389, "step": 32380 }, { "grad_norm": 0.4959447979927063, "learning_rate": 4.755108904676125e-05, "loss": 0.039, "step": 32390 }, { "grad_norm": 0.5392149090766907, "learning_rate": 4.752356466343436e-05, "loss": 0.0411, "step": 32400 }, { "grad_norm": 0.434416800737381, "learning_rate": 4.7496041032383174e-05, "loss": 0.06, "step": 32410 }, { "grad_norm": 0.4159741699695587, "learning_rate": 4.746851816196858e-05, "loss": 0.0463, "step": 32420 }, { "grad_norm": 0.5375228524208069, "learning_rate": 4.744099606055135e-05, "loss": 0.0458, "step": 32430 }, { "grad_norm": 0.33283811807632446, "learning_rate": 4.741347473649193e-05, "loss": 0.0245, "step": 32440 }, { "grad_norm": 0.625566840171814, "learning_rate": 4.738595419815058e-05, "loss": 0.0472, "step": 32450 }, { "grad_norm": 0.5259466171264648, "learning_rate": 4.7358434453887365e-05, "loss": 0.0545, "step": 32460 }, { "grad_norm": 0.4210875928401947, "learning_rate": 4.7330915512061976e-05, "loss": 0.0494, "step": 32470 }, { "grad_norm": 0.4968997836112976, "learning_rate": 4.730339738103402e-05, "loss": 0.0413, "step": 32480 }, { "grad_norm": 0.5630151033401489, "learning_rate": 4.727588006916271e-05, "loss": 0.0476, "step": 32490 }, { "grad_norm": 0.6079937815666199, "learning_rate": 4.724836358480711e-05, "loss": 0.041, "step": 32500 }, { "grad_norm": 0.4426727294921875, "learning_rate": 4.722084793632601e-05, "loss": 0.0388, "step": 32510 }, { "grad_norm": 0.8449121117591858, "learning_rate": 4.719333313207792e-05, "loss": 0.052, "step": 32520 }, { "grad_norm": 0.6206060647964478, "learning_rate": 4.716581918042114e-05, "loss": 0.0515, "step": 32530 }, { "grad_norm": 0.44937944412231445, "learning_rate": 4.7138306089713636e-05, "loss": 0.0399, "step": 32540 }, { "grad_norm": 0.35724127292633057, "learning_rate": 4.7110793868313183e-05, "loss": 0.0534, "step": 32550 }, { "grad_norm": 0.5649448037147522, "learning_rate": 4.708328252457729e-05, "loss": 0.0277, "step": 32560 }, { "grad_norm": 0.9239749312400818, "learning_rate": 4.7055772066863135e-05, "loss": 0.052, "step": 32570 }, { "grad_norm": 0.427839994430542, "learning_rate": 4.702826250352771e-05, "loss": 0.0416, "step": 32580 }, { "grad_norm": 0.443754643201828, "learning_rate": 4.7000753842927653e-05, "loss": 0.0313, "step": 32590 }, { "grad_norm": 0.472173273563385, "learning_rate": 4.6973246093419384e-05, "loss": 0.0375, "step": 32600 }, { "grad_norm": 0.5175598859786987, "learning_rate": 4.694573926335906e-05, "loss": 0.0315, "step": 32610 }, { "grad_norm": 0.29985785484313965, "learning_rate": 4.6918233361102476e-05, "loss": 0.0502, "step": 32620 }, { "grad_norm": 0.6386865377426147, "learning_rate": 4.689072839500525e-05, "loss": 0.038, "step": 32630 }, { "grad_norm": 0.492000550031662, "learning_rate": 4.6863224373422635e-05, "loss": 0.0426, "step": 32640 }, { "grad_norm": 0.4469883143901825, "learning_rate": 4.683572130470962e-05, "loss": 0.0636, "step": 32650 }, { "grad_norm": 0.4484051764011383, "learning_rate": 4.680821919722094e-05, "loss": 0.0349, "step": 32660 }, { "grad_norm": 0.506921112537384, "learning_rate": 4.6780718059310975e-05, "loss": 0.0366, "step": 32670 }, { "grad_norm": 0.6829898953437805, "learning_rate": 4.675321789933389e-05, "loss": 0.038, "step": 32680 }, { "grad_norm": 0.4877735376358032, "learning_rate": 4.6725718725643464e-05, "loss": 0.0347, "step": 32690 }, { "grad_norm": 0.3909919261932373, "learning_rate": 4.669822054659323e-05, "loss": 0.0348, "step": 32700 }, { "grad_norm": 0.504152774810791, "learning_rate": 4.667072337053644e-05, "loss": 0.0571, "step": 32710 }, { "grad_norm": 0.39130598306655884, "learning_rate": 4.6643227205825965e-05, "loss": 0.0333, "step": 32720 }, { "grad_norm": 0.428996205329895, "learning_rate": 4.6615732060814454e-05, "loss": 0.0417, "step": 32730 }, { "grad_norm": 0.4373539388179779, "learning_rate": 4.658823794385417e-05, "loss": 0.051, "step": 32740 }, { "grad_norm": 0.6607068777084351, "learning_rate": 4.6560744863297115e-05, "loss": 0.0551, "step": 32750 }, { "grad_norm": 0.7316754460334778, "learning_rate": 4.653325282749498e-05, "loss": 0.034, "step": 32760 }, { "grad_norm": 0.4268401563167572, "learning_rate": 4.6505761844799075e-05, "loss": 0.0422, "step": 32770 }, { "grad_norm": 0.5403668284416199, "learning_rate": 4.647827192356048e-05, "loss": 0.0492, "step": 32780 }, { "grad_norm": 0.6083895564079285, "learning_rate": 4.645078307212989e-05, "loss": 0.0503, "step": 32790 }, { "grad_norm": 0.5737928152084351, "learning_rate": 4.642329529885768e-05, "loss": 0.0432, "step": 32800 }, { "grad_norm": 0.7532700896263123, "learning_rate": 4.639580861209393e-05, "loss": 0.0576, "step": 32810 }, { "grad_norm": 0.6221258640289307, "learning_rate": 4.636832302018835e-05, "loss": 0.0445, "step": 32820 }, { "grad_norm": 0.44872021675109863, "learning_rate": 4.6340838531490365e-05, "loss": 0.039, "step": 32830 }, { "grad_norm": 0.4370836019515991, "learning_rate": 4.6313355154349e-05, "loss": 0.0411, "step": 32840 }, { "grad_norm": 0.612576961517334, "learning_rate": 4.6285872897113025e-05, "loss": 0.0461, "step": 32850 }, { "grad_norm": 0.6107694506645203, "learning_rate": 4.625839176813077e-05, "loss": 0.0453, "step": 32860 }, { "grad_norm": 0.4275030791759491, "learning_rate": 4.623091177575031e-05, "loss": 0.0406, "step": 32870 }, { "grad_norm": 0.5657408237457275, "learning_rate": 4.620343292831936e-05, "loss": 0.0408, "step": 32880 }, { "grad_norm": 0.6410797238349915, "learning_rate": 4.6175955234185206e-05, "loss": 0.0327, "step": 32890 }, { "grad_norm": 0.36859676241874695, "learning_rate": 4.614847870169492e-05, "loss": 0.0378, "step": 32900 }, { "grad_norm": 0.4250585436820984, "learning_rate": 4.612100333919509e-05, "loss": 0.0407, "step": 32910 }, { "grad_norm": 0.4264342486858368, "learning_rate": 4.609352915503202e-05, "loss": 0.0397, "step": 32920 }, { "grad_norm": 0.4222140312194824, "learning_rate": 4.606605615755166e-05, "loss": 0.0419, "step": 32930 }, { "grad_norm": 0.40639728307724, "learning_rate": 4.6038584355099576e-05, "loss": 0.0368, "step": 32940 }, { "grad_norm": 0.8495758175849915, "learning_rate": 4.6011113756020964e-05, "loss": 0.0369, "step": 32950 }, { "grad_norm": 0.4629928171634674, "learning_rate": 4.598364436866066e-05, "loss": 0.0498, "step": 32960 }, { "grad_norm": 0.4818340241909027, "learning_rate": 4.595617620136316e-05, "loss": 0.0464, "step": 32970 }, { "grad_norm": 0.48565730452537537, "learning_rate": 4.592870926247257e-05, "loss": 0.0534, "step": 32980 }, { "grad_norm": 0.3874340057373047, "learning_rate": 4.5901243560332594e-05, "loss": 0.0389, "step": 32990 }, { "grad_norm": 0.5258336067199707, "learning_rate": 4.587377910328662e-05, "loss": 0.0417, "step": 33000 }, { "grad_norm": 0.4670795202255249, "learning_rate": 4.5846315899677586e-05, "loss": 0.0391, "step": 33010 }, { "grad_norm": 0.4179302752017975, "learning_rate": 4.5818853957848114e-05, "loss": 0.0343, "step": 33020 }, { "grad_norm": 0.32086238265037537, "learning_rate": 4.579139328614043e-05, "loss": 0.0432, "step": 33030 }, { "grad_norm": 0.2607830762863159, "learning_rate": 4.576393389289633e-05, "loss": 0.0415, "step": 33040 }, { "grad_norm": 1.296952486038208, "learning_rate": 4.573647578645728e-05, "loss": 0.0439, "step": 33050 }, { "grad_norm": 0.5792878270149231, "learning_rate": 4.57090189751643e-05, "loss": 0.0452, "step": 33060 }, { "grad_norm": 0.6021563410758972, "learning_rate": 4.568156346735806e-05, "loss": 0.0314, "step": 33070 }, { "grad_norm": 0.6423443555831909, "learning_rate": 4.565410927137882e-05, "loss": 0.0433, "step": 33080 }, { "grad_norm": 0.4685280919075012, "learning_rate": 4.562665639556644e-05, "loss": 0.0399, "step": 33090 }, { "grad_norm": 0.3975920081138611, "learning_rate": 4.559920484826037e-05, "loss": 0.0498, "step": 33100 }, { "grad_norm": 0.38118457794189453, "learning_rate": 4.5571754637799665e-05, "loss": 0.0433, "step": 33110 }, { "grad_norm": 0.49927976727485657, "learning_rate": 4.554430577252298e-05, "loss": 0.0493, "step": 33120 }, { "grad_norm": 0.8706644177436829, "learning_rate": 4.551685826076858e-05, "loss": 0.0434, "step": 33130 }, { "grad_norm": 0.600502073764801, "learning_rate": 4.5489412110874246e-05, "loss": 0.0358, "step": 33140 }, { "grad_norm": 0.2931182086467743, "learning_rate": 4.5461967331177444e-05, "loss": 0.0585, "step": 33150 }, { "grad_norm": 0.30276110768318176, "learning_rate": 4.5434523930015115e-05, "loss": 0.0418, "step": 33160 }, { "grad_norm": 0.29743674397468567, "learning_rate": 4.540708191572388e-05, "loss": 0.0353, "step": 33170 }, { "grad_norm": 0.6821106672286987, "learning_rate": 4.537964129663991e-05, "loss": 0.0353, "step": 33180 }, { "grad_norm": 0.5301848649978638, "learning_rate": 4.535220208109889e-05, "loss": 0.037, "step": 33190 }, { "grad_norm": 0.4436086118221283, "learning_rate": 4.5324764277436194e-05, "loss": 0.0425, "step": 33200 }, { "grad_norm": 0.4817410409450531, "learning_rate": 4.529732789398664e-05, "loss": 0.0389, "step": 33210 }, { "grad_norm": 0.5834327936172485, "learning_rate": 4.526989293908472e-05, "loss": 0.0393, "step": 33220 }, { "grad_norm": 0.4578525424003601, "learning_rate": 4.524245942106442e-05, "loss": 0.0304, "step": 33230 }, { "grad_norm": 0.6279627680778503, "learning_rate": 4.5215027348259345e-05, "loss": 0.0316, "step": 33240 }, { "grad_norm": 0.6764660477638245, "learning_rate": 4.5187596729002616e-05, "loss": 0.0422, "step": 33250 }, { "grad_norm": 0.4874812066555023, "learning_rate": 4.516016757162693e-05, "loss": 0.0426, "step": 33260 }, { "grad_norm": 0.58433598279953, "learning_rate": 4.513273988446457e-05, "loss": 0.0349, "step": 33270 }, { "grad_norm": 0.902131974697113, "learning_rate": 4.5105313675847296e-05, "loss": 0.0519, "step": 33280 }, { "grad_norm": 0.3594360053539276, "learning_rate": 4.5077888954106495e-05, "loss": 0.0508, "step": 33290 }, { "grad_norm": 0.5136241316795349, "learning_rate": 4.505046572757309e-05, "loss": 0.0382, "step": 33300 }, { "grad_norm": 0.7847499847412109, "learning_rate": 4.502304400457749e-05, "loss": 0.0476, "step": 33310 }, { "grad_norm": 0.4373922348022461, "learning_rate": 4.499562379344973e-05, "loss": 0.045, "step": 33320 }, { "grad_norm": 0.679299533367157, "learning_rate": 4.4968205102519306e-05, "loss": 0.0376, "step": 33330 }, { "grad_norm": 0.47906702756881714, "learning_rate": 4.494078794011532e-05, "loss": 0.0768, "step": 33340 }, { "grad_norm": 0.46065428853034973, "learning_rate": 4.491337231456639e-05, "loss": 0.0332, "step": 33350 }, { "grad_norm": 0.733274519443512, "learning_rate": 4.4885958234200634e-05, "loss": 0.0405, "step": 33360 }, { "grad_norm": 0.5538462400436401, "learning_rate": 4.485854570734575e-05, "loss": 0.0442, "step": 33370 }, { "grad_norm": 0.38740360736846924, "learning_rate": 4.483113474232891e-05, "loss": 0.0434, "step": 33380 }, { "grad_norm": 0.49184468388557434, "learning_rate": 4.480372534747688e-05, "loss": 0.0356, "step": 33390 }, { "grad_norm": 0.4791405498981476, "learning_rate": 4.477631753111588e-05, "loss": 0.0323, "step": 33400 }, { "grad_norm": 0.5636563897132874, "learning_rate": 4.4748911301571686e-05, "loss": 0.0482, "step": 33410 }, { "grad_norm": 0.6698736548423767, "learning_rate": 4.472150666716961e-05, "loss": 0.0407, "step": 33420 }, { "grad_norm": 0.6591954827308655, "learning_rate": 4.469410363623442e-05, "loss": 0.0471, "step": 33430 }, { "grad_norm": 0.6624734401702881, "learning_rate": 4.466670221709044e-05, "loss": 0.0416, "step": 33440 }, { "grad_norm": 0.3356441557407379, "learning_rate": 4.463930241806154e-05, "loss": 0.0316, "step": 33450 }, { "grad_norm": 0.3679370880126953, "learning_rate": 4.4611904247471006e-05, "loss": 0.0638, "step": 33460 }, { "grad_norm": 0.37470072507858276, "learning_rate": 4.458450771364171e-05, "loss": 0.0317, "step": 33470 }, { "grad_norm": 0.3148155212402344, "learning_rate": 4.4557112824895965e-05, "loss": 0.0341, "step": 33480 }, { "grad_norm": 0.7730985879898071, "learning_rate": 4.452971958955563e-05, "loss": 0.056, "step": 33490 }, { "grad_norm": 0.3856586813926697, "learning_rate": 4.450232801594208e-05, "loss": 0.0469, "step": 33500 }, { "grad_norm": 0.5106663107872009, "learning_rate": 4.447493811237609e-05, "loss": 0.0403, "step": 33510 }, { "grad_norm": 0.6850948333740234, "learning_rate": 4.444754988717804e-05, "loss": 0.0483, "step": 33520 }, { "grad_norm": 0.5654472708702087, "learning_rate": 4.442016334866771e-05, "loss": 0.0478, "step": 33530 }, { "grad_norm": 0.6600124835968018, "learning_rate": 4.4392778505164445e-05, "loss": 0.0318, "step": 33540 }, { "grad_norm": 0.42543745040893555, "learning_rate": 4.436539536498702e-05, "loss": 0.0365, "step": 33550 }, { "grad_norm": 0.5127907395362854, "learning_rate": 4.433801393645369e-05, "loss": 0.0424, "step": 33560 }, { "grad_norm": 0.6018719673156738, "learning_rate": 4.431063422788226e-05, "loss": 0.0393, "step": 33570 }, { "grad_norm": 0.3573082387447357, "learning_rate": 4.428325624758991e-05, "loss": 0.0345, "step": 33580 }, { "grad_norm": 0.5456638932228088, "learning_rate": 4.4255880003893366e-05, "loss": 0.0475, "step": 33590 }, { "grad_norm": 0.5197299122810364, "learning_rate": 4.422850550510884e-05, "loss": 0.0396, "step": 33600 }, { "grad_norm": 0.49885478615760803, "learning_rate": 4.4201132759551934e-05, "loss": 0.03, "step": 33610 }, { "grad_norm": 0.6779634356498718, "learning_rate": 4.4173761775537804e-05, "loss": 0.046, "step": 33620 }, { "grad_norm": 0.6372296214103699, "learning_rate": 4.414639256138099e-05, "loss": 0.045, "step": 33630 }, { "grad_norm": 0.43374142050743103, "learning_rate": 4.411902512539557e-05, "loss": 0.0372, "step": 33640 }, { "grad_norm": 0.5364989042282104, "learning_rate": 4.4091659475895044e-05, "loss": 0.0432, "step": 33650 }, { "grad_norm": 0.2686699330806732, "learning_rate": 4.406429562119235e-05, "loss": 0.0412, "step": 33660 }, { "grad_norm": 0.3757244348526001, "learning_rate": 4.4036933569599945e-05, "loss": 0.0402, "step": 33670 }, { "grad_norm": 0.6740881204605103, "learning_rate": 4.400957332942965e-05, "loss": 0.0359, "step": 33680 }, { "grad_norm": 0.44094613194465637, "learning_rate": 4.3982214908992844e-05, "loss": 0.0343, "step": 33690 }, { "grad_norm": 0.5767551064491272, "learning_rate": 4.3954858316600235e-05, "loss": 0.0409, "step": 33700 }, { "grad_norm": 0.353031188249588, "learning_rate": 4.392750356056205e-05, "loss": 0.0329, "step": 33710 }, { "grad_norm": 0.5685975551605225, "learning_rate": 4.390015064918798e-05, "loss": 0.0739, "step": 33720 }, { "grad_norm": 0.3845091164112091, "learning_rate": 4.387279959078705e-05, "loss": 0.0423, "step": 33730 }, { "grad_norm": 0.4317370653152466, "learning_rate": 4.384545039366786e-05, "loss": 0.0366, "step": 33740 }, { "grad_norm": 0.3885256350040436, "learning_rate": 4.381810306613831e-05, "loss": 0.0414, "step": 33750 }, { "grad_norm": 0.399370014667511, "learning_rate": 4.3790757616505826e-05, "loss": 0.0403, "step": 33760 }, { "grad_norm": 0.37536415457725525, "learning_rate": 4.376341405307725e-05, "loss": 0.0506, "step": 33770 }, { "grad_norm": 0.5060846209526062, "learning_rate": 4.37360723841588e-05, "loss": 0.0351, "step": 33780 }, { "grad_norm": 0.5463724136352539, "learning_rate": 4.370873261805619e-05, "loss": 0.0373, "step": 33790 }, { "grad_norm": 0.29660823941230774, "learning_rate": 4.368139476307449e-05, "loss": 0.0293, "step": 33800 }, { "grad_norm": 0.7239187359809875, "learning_rate": 4.365405882751822e-05, "loss": 0.0352, "step": 33810 }, { "grad_norm": 0.4708261787891388, "learning_rate": 4.3626724819691326e-05, "loss": 0.0359, "step": 33820 }, { "grad_norm": 0.8051251173019409, "learning_rate": 4.359939274789715e-05, "loss": 0.0486, "step": 33830 }, { "grad_norm": 0.42022785544395447, "learning_rate": 4.357206262043848e-05, "loss": 0.0381, "step": 33840 }, { "grad_norm": 0.5566877126693726, "learning_rate": 4.354473444561745e-05, "loss": 0.0389, "step": 33850 }, { "grad_norm": 0.7898592948913574, "learning_rate": 4.3517408231735644e-05, "loss": 0.0416, "step": 33860 }, { "grad_norm": 0.6134763360023499, "learning_rate": 4.3490083987094086e-05, "loss": 0.0424, "step": 33870 }, { "grad_norm": 0.646737277507782, "learning_rate": 4.34627617199931e-05, "loss": 0.0373, "step": 33880 }, { "grad_norm": 0.501086950302124, "learning_rate": 4.3435441438732526e-05, "loss": 0.046, "step": 33890 }, { "grad_norm": 0.35699182748794556, "learning_rate": 4.340812315161149e-05, "loss": 0.047, "step": 33900 }, { "grad_norm": 0.3361864984035492, "learning_rate": 4.338080686692859e-05, "loss": 0.0367, "step": 33910 }, { "grad_norm": 0.5456534624099731, "learning_rate": 4.3353492592981816e-05, "loss": 0.0371, "step": 33920 }, { "grad_norm": 0.5255162119865417, "learning_rate": 4.3326180338068485e-05, "loss": 0.044, "step": 33930 }, { "grad_norm": 0.6976553797721863, "learning_rate": 4.3298870110485356e-05, "loss": 0.0459, "step": 33940 }, { "grad_norm": 1.3126593828201294, "learning_rate": 4.3271561918528567e-05, "loss": 0.0435, "step": 33950 }, { "grad_norm": 0.49609753489494324, "learning_rate": 4.324425577049359e-05, "loss": 0.0409, "step": 33960 }, { "grad_norm": 0.5023862719535828, "learning_rate": 4.321695167467535e-05, "loss": 0.0513, "step": 33970 }, { "grad_norm": 0.56548011302948, "learning_rate": 4.3189649639368093e-05, "loss": 0.0502, "step": 33980 }, { "grad_norm": 0.5253924131393433, "learning_rate": 4.316234967286547e-05, "loss": 0.0547, "step": 33990 }, { "grad_norm": 0.8525822758674622, "learning_rate": 4.313505178346046e-05, "loss": 0.0404, "step": 34000 }, { "grad_norm": 0.41452938318252563, "learning_rate": 4.3107755979445465e-05, "loss": 0.0463, "step": 34010 }, { "grad_norm": 0.6486383676528931, "learning_rate": 4.308046226911224e-05, "loss": 0.0603, "step": 34020 }, { "grad_norm": 0.5499621629714966, "learning_rate": 4.305317066075185e-05, "loss": 0.0615, "step": 34030 }, { "grad_norm": 0.6390924453735352, "learning_rate": 4.302588116265482e-05, "loss": 0.0543, "step": 34040 }, { "grad_norm": 0.7983718514442444, "learning_rate": 4.299859378311094e-05, "loss": 0.0496, "step": 34050 }, { "grad_norm": 0.4232882261276245, "learning_rate": 4.2971308530409424e-05, "loss": 0.0375, "step": 34060 }, { "grad_norm": 0.4229515790939331, "learning_rate": 4.2944025412838765e-05, "loss": 0.0412, "step": 34070 }, { "grad_norm": 0.4418099820613861, "learning_rate": 4.291674443868689e-05, "loss": 0.0324, "step": 34080 }, { "grad_norm": 0.86976557970047, "learning_rate": 4.288946561624104e-05, "loss": 0.0416, "step": 34090 }, { "grad_norm": 0.47029852867126465, "learning_rate": 4.2862188953787794e-05, "loss": 0.0321, "step": 34100 }, { "grad_norm": 0.43386587500572205, "learning_rate": 4.283491445961308e-05, "loss": 0.0359, "step": 34110 }, { "grad_norm": 0.8348512649536133, "learning_rate": 4.2807642142002155e-05, "loss": 0.0554, "step": 34120 }, { "grad_norm": 0.9963071942329407, "learning_rate": 4.278037200923966e-05, "loss": 0.0369, "step": 34130 }, { "grad_norm": 0.33705437183380127, "learning_rate": 4.275310406960953e-05, "loss": 0.0391, "step": 34140 }, { "grad_norm": 0.6387983560562134, "learning_rate": 4.272583833139502e-05, "loss": 0.0491, "step": 34150 }, { "grad_norm": 0.461124062538147, "learning_rate": 4.2698574802878794e-05, "loss": 0.0382, "step": 34160 }, { "grad_norm": 0.36047428846359253, "learning_rate": 4.2671313492342734e-05, "loss": 0.0374, "step": 34170 }, { "grad_norm": 0.4120961129665375, "learning_rate": 4.264405440806813e-05, "loss": 0.0491, "step": 34180 }, { "grad_norm": 0.3624422252178192, "learning_rate": 4.26167975583356e-05, "loss": 0.0424, "step": 34190 }, { "grad_norm": 0.37305864691734314, "learning_rate": 4.2589542951425e-05, "loss": 0.0325, "step": 34200 }, { "grad_norm": 0.4030051529407501, "learning_rate": 4.2562290595615615e-05, "loss": 0.0317, "step": 34210 }, { "grad_norm": 0.5924611687660217, "learning_rate": 4.2535040499185946e-05, "loss": 0.0378, "step": 34220 }, { "grad_norm": 0.6618282794952393, "learning_rate": 4.250779267041387e-05, "loss": 0.0354, "step": 34230 }, { "grad_norm": 0.5151879787445068, "learning_rate": 4.248054711757657e-05, "loss": 0.0358, "step": 34240 }, { "grad_norm": 0.4624594748020172, "learning_rate": 4.245330384895052e-05, "loss": 0.0344, "step": 34250 }, { "grad_norm": 0.4662933647632599, "learning_rate": 4.242606287281151e-05, "loss": 0.035, "step": 34260 }, { "grad_norm": 0.4485674500465393, "learning_rate": 4.2398824197434595e-05, "loss": 0.0366, "step": 34270 }, { "grad_norm": 0.33288419246673584, "learning_rate": 4.23715878310942e-05, "loss": 0.0301, "step": 34280 }, { "grad_norm": 0.4260443150997162, "learning_rate": 4.234435378206402e-05, "loss": 0.0399, "step": 34290 }, { "grad_norm": 0.31092432141304016, "learning_rate": 4.2317122058617006e-05, "loss": 0.0346, "step": 34300 }, { "grad_norm": 0.24140334129333496, "learning_rate": 4.2289892669025485e-05, "loss": 0.0362, "step": 34310 }, { "grad_norm": 0.4858108162879944, "learning_rate": 4.226266562156097e-05, "loss": 0.0447, "step": 34320 }, { "grad_norm": 0.4222678542137146, "learning_rate": 4.223544092449435e-05, "loss": 0.0465, "step": 34330 }, { "grad_norm": 0.3594076633453369, "learning_rate": 4.2208218586095784e-05, "loss": 0.0319, "step": 34340 }, { "grad_norm": 0.5032150149345398, "learning_rate": 4.218099861463466e-05, "loss": 0.0373, "step": 34350 }, { "grad_norm": 0.5078218579292297, "learning_rate": 4.215378101837972e-05, "loss": 0.0436, "step": 34360 }, { "grad_norm": 0.6222392320632935, "learning_rate": 4.2126565805598937e-05, "loss": 0.0364, "step": 34370 }, { "grad_norm": 0.4366213381290436, "learning_rate": 4.209935298455957e-05, "loss": 0.0312, "step": 34380 }, { "grad_norm": 0.350903183221817, "learning_rate": 4.207214256352817e-05, "loss": 0.0387, "step": 34390 }, { "grad_norm": 0.4707132875919342, "learning_rate": 4.2044934550770524e-05, "loss": 0.0468, "step": 34400 }, { "grad_norm": 0.3576429486274719, "learning_rate": 4.201772895455174e-05, "loss": 0.0306, "step": 34410 }, { "grad_norm": 0.5341459512710571, "learning_rate": 4.199052578313613e-05, "loss": 0.0448, "step": 34420 }, { "grad_norm": 0.40598636865615845, "learning_rate": 4.1963325044787294e-05, "loss": 0.0313, "step": 34430 }, { "grad_norm": 0.7308167815208435, "learning_rate": 4.193612674776814e-05, "loss": 0.0427, "step": 34440 }, { "grad_norm": 0.5919090509414673, "learning_rate": 4.1908930900340745e-05, "loss": 0.0389, "step": 34450 }, { "grad_norm": 0.4191787838935852, "learning_rate": 4.1881737510766536e-05, "loss": 0.0384, "step": 34460 }, { "grad_norm": 0.6103390455245972, "learning_rate": 4.185454658730609e-05, "loss": 0.029, "step": 34470 }, { "grad_norm": 0.5214332938194275, "learning_rate": 4.1827358138219355e-05, "loss": 0.0414, "step": 34480 }, { "grad_norm": 0.37041693925857544, "learning_rate": 4.1800172171765404e-05, "loss": 0.0642, "step": 34490 }, { "grad_norm": 0.366913378238678, "learning_rate": 4.177298869620264e-05, "loss": 0.0305, "step": 34500 }, { "grad_norm": 0.48156243562698364, "learning_rate": 4.1745807719788705e-05, "loss": 0.0335, "step": 34510 }, { "grad_norm": 0.5191292762756348, "learning_rate": 4.1718629250780445e-05, "loss": 0.0341, "step": 34520 }, { "grad_norm": 0.3484908938407898, "learning_rate": 4.1691453297433956e-05, "loss": 0.0318, "step": 34530 }, { "grad_norm": 0.40339699387550354, "learning_rate": 4.166427986800457e-05, "loss": 0.0291, "step": 34540 }, { "grad_norm": 0.6719741821289062, "learning_rate": 4.163710897074688e-05, "loss": 0.0418, "step": 34550 }, { "grad_norm": 0.269720196723938, "learning_rate": 4.1609940613914686e-05, "loss": 0.0343, "step": 34560 }, { "grad_norm": 0.301864892244339, "learning_rate": 4.1582774805760996e-05, "loss": 0.0474, "step": 34570 }, { "grad_norm": 0.36265140771865845, "learning_rate": 4.155561155453809e-05, "loss": 0.0446, "step": 34580 }, { "grad_norm": 0.47974035143852234, "learning_rate": 4.15284508684974e-05, "loss": 0.0418, "step": 34590 }, { "grad_norm": 0.6638578176498413, "learning_rate": 4.1501292755889675e-05, "loss": 0.0364, "step": 34600 }, { "grad_norm": 0.49888813495635986, "learning_rate": 4.1474137224964833e-05, "loss": 0.0453, "step": 34610 }, { "grad_norm": 0.5045255422592163, "learning_rate": 4.144698428397197e-05, "loss": 0.0378, "step": 34620 }, { "grad_norm": 0.4403499662876129, "learning_rate": 4.1419833941159466e-05, "loss": 0.0321, "step": 34630 }, { "grad_norm": 0.3094809353351593, "learning_rate": 4.1392686204774846e-05, "loss": 0.0298, "step": 34640 }, { "grad_norm": 0.34781280159950256, "learning_rate": 4.13655410830649e-05, "loss": 0.0323, "step": 34650 }, { "grad_norm": 0.32630160450935364, "learning_rate": 4.1338398584275594e-05, "loss": 0.0269, "step": 34660 }, { "grad_norm": 0.5586895942687988, "learning_rate": 4.1311258716652104e-05, "loss": 0.0445, "step": 34670 }, { "grad_norm": 0.4449036121368408, "learning_rate": 4.128412148843881e-05, "loss": 0.0345, "step": 34680 }, { "grad_norm": 0.3635903596878052, "learning_rate": 4.125698690787926e-05, "loss": 0.0398, "step": 34690 }, { "grad_norm": 0.5829980969429016, "learning_rate": 4.1229854983216245e-05, "loss": 0.033, "step": 34700 }, { "grad_norm": 0.5253975987434387, "learning_rate": 4.120272572269175e-05, "loss": 0.0503, "step": 34710 }, { "grad_norm": 0.7263593673706055, "learning_rate": 4.117559913454687e-05, "loss": 0.048, "step": 34720 }, { "grad_norm": 0.355692982673645, "learning_rate": 4.114847522702201e-05, "loss": 0.0355, "step": 34730 }, { "grad_norm": 0.5307477116584778, "learning_rate": 4.112135400835664e-05, "loss": 0.045, "step": 34740 }, { "grad_norm": 1.1029529571533203, "learning_rate": 4.109423548678949e-05, "loss": 0.0438, "step": 34750 }, { "grad_norm": 0.42104506492614746, "learning_rate": 4.106711967055848e-05, "loss": 0.0333, "step": 34760 }, { "grad_norm": 0.6269537210464478, "learning_rate": 4.1040006567900636e-05, "loss": 0.0451, "step": 34770 }, { "grad_norm": 0.5581393241882324, "learning_rate": 4.101289618705224e-05, "loss": 0.0426, "step": 34780 }, { "grad_norm": 0.29503965377807617, "learning_rate": 4.0985788536248675e-05, "loss": 0.0363, "step": 34790 }, { "grad_norm": 0.4311757981777191, "learning_rate": 4.095868362372454e-05, "loss": 0.0436, "step": 34800 }, { "grad_norm": 0.46256670355796814, "learning_rate": 4.0931581457713614e-05, "loss": 0.0335, "step": 34810 }, { "grad_norm": 0.4423195719718933, "learning_rate": 4.09044820464488e-05, "loss": 0.0325, "step": 34820 }, { "grad_norm": 0.6409271955490112, "learning_rate": 4.087738539816219e-05, "loss": 0.0386, "step": 34830 }, { "grad_norm": 0.7602289915084839, "learning_rate": 4.085029152108501e-05, "loss": 0.0508, "step": 34840 }, { "grad_norm": 0.5704853534698486, "learning_rate": 4.0823200423447714e-05, "loss": 0.0413, "step": 34850 }, { "grad_norm": 0.6176941990852356, "learning_rate": 4.079611211347981e-05, "loss": 0.0325, "step": 34860 }, { "grad_norm": 0.27535173296928406, "learning_rate": 4.076902659941002e-05, "loss": 0.0341, "step": 34870 }, { "grad_norm": 0.5813370943069458, "learning_rate": 4.074194388946624e-05, "loss": 0.0532, "step": 34880 }, { "grad_norm": 0.5457409620285034, "learning_rate": 4.071486399187545e-05, "loss": 0.0377, "step": 34890 }, { "grad_norm": 0.46621814370155334, "learning_rate": 4.0687786914863836e-05, "loss": 0.0564, "step": 34900 }, { "grad_norm": 0.5874105095863342, "learning_rate": 4.0660712666656666e-05, "loss": 0.0434, "step": 34910 }, { "grad_norm": 0.5145257115364075, "learning_rate": 4.0633641255478394e-05, "loss": 0.0368, "step": 34920 }, { "grad_norm": 0.3208734095096588, "learning_rate": 4.0606572689552624e-05, "loss": 0.0331, "step": 34930 }, { "grad_norm": 0.5635885000228882, "learning_rate": 4.0579506977102036e-05, "loss": 0.0348, "step": 34940 }, { "grad_norm": 0.47405001521110535, "learning_rate": 4.055244412634849e-05, "loss": 0.0397, "step": 34950 }, { "grad_norm": 0.40256500244140625, "learning_rate": 4.052538414551298e-05, "loss": 0.0472, "step": 34960 }, { "grad_norm": 0.4160650372505188, "learning_rate": 4.0498327042815596e-05, "loss": 0.0444, "step": 34970 }, { "grad_norm": 0.45230692625045776, "learning_rate": 4.047127282647559e-05, "loss": 0.039, "step": 34980 }, { "grad_norm": 0.46957194805145264, "learning_rate": 4.04442215047113e-05, "loss": 0.0403, "step": 34990 }, { "grad_norm": 0.36001187562942505, "learning_rate": 4.041717308574023e-05, "loss": 0.0281, "step": 35000 }, { "grad_norm": 0.534615695476532, "learning_rate": 4.039012757777893e-05, "loss": 0.0362, "step": 35010 }, { "grad_norm": 0.7568759322166443, "learning_rate": 4.036308498904314e-05, "loss": 0.043, "step": 35020 }, { "grad_norm": 0.32158103585243225, "learning_rate": 4.033604532774771e-05, "loss": 0.0327, "step": 35030 }, { "grad_norm": 0.4685046076774597, "learning_rate": 4.030900860210652e-05, "loss": 0.0524, "step": 35040 }, { "grad_norm": 0.5895735621452332, "learning_rate": 4.028197482033266e-05, "loss": 0.0396, "step": 35050 }, { "grad_norm": 0.6776599884033203, "learning_rate": 4.0254943990638246e-05, "loss": 0.0448, "step": 35060 }, { "grad_norm": 0.32722097635269165, "learning_rate": 4.022791612123454e-05, "loss": 0.0438, "step": 35070 }, { "grad_norm": 0.4095350205898285, "learning_rate": 4.020089122033192e-05, "loss": 0.0455, "step": 35080 }, { "grad_norm": 0.8176259994506836, "learning_rate": 4.01738692961398e-05, "loss": 0.0339, "step": 35090 }, { "grad_norm": 0.6366498470306396, "learning_rate": 4.014685035686675e-05, "loss": 0.0482, "step": 35100 }, { "grad_norm": 0.4608025550842285, "learning_rate": 4.011983441072039e-05, "loss": 0.0491, "step": 35110 }, { "grad_norm": 0.6306338310241699, "learning_rate": 4.0092821465907485e-05, "loss": 0.0548, "step": 35120 }, { "grad_norm": 0.4071965515613556, "learning_rate": 4.006581153063383e-05, "loss": 0.0369, "step": 35130 }, { "grad_norm": 0.6793075203895569, "learning_rate": 4.003880461310432e-05, "loss": 0.0579, "step": 35140 }, { "grad_norm": 0.5597251057624817, "learning_rate": 4.001180072152298e-05, "loss": 0.0364, "step": 35150 }, { "grad_norm": 0.750381588935852, "learning_rate": 3.998479986409285e-05, "loss": 0.0441, "step": 35160 }, { "grad_norm": 0.38337960839271545, "learning_rate": 3.995780204901607e-05, "loss": 0.0364, "step": 35170 }, { "grad_norm": 0.6178902983665466, "learning_rate": 3.993080728449391e-05, "loss": 0.0412, "step": 35180 }, { "grad_norm": 0.6007737517356873, "learning_rate": 3.990381557872661e-05, "loss": 0.0319, "step": 35190 }, { "grad_norm": 0.5504187345504761, "learning_rate": 3.987682693991359e-05, "loss": 0.0734, "step": 35200 }, { "grad_norm": 0.6206884980201721, "learning_rate": 3.9849841376253226e-05, "loss": 0.0427, "step": 35210 }, { "grad_norm": 0.6355217695236206, "learning_rate": 3.982285889594306e-05, "loss": 0.0415, "step": 35220 }, { "grad_norm": 0.48703110218048096, "learning_rate": 3.9795879507179665e-05, "loss": 0.0477, "step": 35230 }, { "grad_norm": 0.5384232401847839, "learning_rate": 3.9768903218158634e-05, "loss": 0.0404, "step": 35240 }, { "grad_norm": 0.780945360660553, "learning_rate": 3.974193003707468e-05, "loss": 0.0367, "step": 35250 }, { "grad_norm": 0.5305658578872681, "learning_rate": 3.971495997212152e-05, "loss": 0.0339, "step": 35260 }, { "grad_norm": 0.40313053131103516, "learning_rate": 3.9687993031491985e-05, "loss": 0.047, "step": 35270 }, { "grad_norm": 0.5113972425460815, "learning_rate": 3.966102922337787e-05, "loss": 0.0424, "step": 35280 }, { "grad_norm": 0.46285194158554077, "learning_rate": 3.963406855597009e-05, "loss": 0.0385, "step": 35290 }, { "grad_norm": 0.6544591784477234, "learning_rate": 3.960711103745861e-05, "loss": 0.0312, "step": 35300 }, { "grad_norm": 0.5590714812278748, "learning_rate": 3.958015667603237e-05, "loss": 0.0323, "step": 35310 }, { "grad_norm": 0.7959566116333008, "learning_rate": 3.955320547987943e-05, "loss": 0.0704, "step": 35320 }, { "grad_norm": 0.7576131224632263, "learning_rate": 3.952625745718681e-05, "loss": 0.0337, "step": 35330 }, { "grad_norm": 0.5620718598365784, "learning_rate": 3.949931261614064e-05, "loss": 0.0528, "step": 35340 }, { "grad_norm": 0.4490267038345337, "learning_rate": 3.947237096492605e-05, "loss": 0.0497, "step": 35350 }, { "grad_norm": 0.5344087481498718, "learning_rate": 3.944543251172719e-05, "loss": 0.0412, "step": 35360 }, { "grad_norm": 0.4333003759384155, "learning_rate": 3.941849726472725e-05, "loss": 0.0406, "step": 35370 }, { "grad_norm": 0.44530531764030457, "learning_rate": 3.939156523210846e-05, "loss": 0.0319, "step": 35380 }, { "grad_norm": 0.2426714450120926, "learning_rate": 3.9364636422052046e-05, "loss": 0.0544, "step": 35390 }, { "grad_norm": 0.4878357946872711, "learning_rate": 3.933771084273828e-05, "loss": 0.037, "step": 35400 }, { "grad_norm": 0.5130521059036255, "learning_rate": 3.931078850234643e-05, "loss": 0.0298, "step": 35410 }, { "grad_norm": 0.5277429223060608, "learning_rate": 3.928386940905483e-05, "loss": 0.0433, "step": 35420 }, { "grad_norm": 0.3751724064350128, "learning_rate": 3.925695357104073e-05, "loss": 0.0281, "step": 35430 }, { "grad_norm": 0.4688996374607086, "learning_rate": 3.923004099648049e-05, "loss": 0.0395, "step": 35440 }, { "grad_norm": 0.6461942791938782, "learning_rate": 3.920313169354944e-05, "loss": 0.0346, "step": 35450 }, { "grad_norm": 0.33302873373031616, "learning_rate": 3.9176225670421897e-05, "loss": 0.0302, "step": 35460 }, { "grad_norm": 0.2606893479824066, "learning_rate": 3.9149322935271224e-05, "loss": 0.027, "step": 35470 }, { "grad_norm": 0.33909282088279724, "learning_rate": 3.9122423496269725e-05, "loss": 0.061, "step": 35480 }, { "grad_norm": 0.550664484500885, "learning_rate": 3.909552736158877e-05, "loss": 0.0389, "step": 35490 }, { "grad_norm": 0.44409510493278503, "learning_rate": 3.90686345393987e-05, "loss": 0.0645, "step": 35500 }, { "grad_norm": 0.4767763614654541, "learning_rate": 3.9041745037868816e-05, "loss": 0.0375, "step": 35510 }, { "grad_norm": 0.5499173998832703, "learning_rate": 3.9014858865167465e-05, "loss": 0.0471, "step": 35520 }, { "grad_norm": 0.790185809135437, "learning_rate": 3.8987976029461935e-05, "loss": 0.0369, "step": 35530 }, { "grad_norm": 0.852082371711731, "learning_rate": 3.896109653891853e-05, "loss": 0.0497, "step": 35540 }, { "grad_norm": 0.4939943850040436, "learning_rate": 3.893422040170254e-05, "loss": 0.0498, "step": 35550 }, { "grad_norm": 0.3372114300727844, "learning_rate": 3.8907347625978207e-05, "loss": 0.0421, "step": 35560 }, { "grad_norm": 0.6447075605392456, "learning_rate": 3.88804782199088e-05, "loss": 0.0335, "step": 35570 }, { "grad_norm": 0.3817063271999359, "learning_rate": 3.8853612191656495e-05, "loss": 0.0389, "step": 35580 }, { "grad_norm": 0.5334768295288086, "learning_rate": 3.88267495493825e-05, "loss": 0.0407, "step": 35590 }, { "grad_norm": 0.4299335777759552, "learning_rate": 3.8799890301247004e-05, "loss": 0.0397, "step": 35600 }, { "grad_norm": 0.4389583468437195, "learning_rate": 3.8773034455409096e-05, "loss": 0.0372, "step": 35610 }, { "grad_norm": 0.4824255704879761, "learning_rate": 3.8746182020026904e-05, "loss": 0.0572, "step": 35620 }, { "grad_norm": 0.6972919702529907, "learning_rate": 3.871933300325745e-05, "loss": 0.0389, "step": 35630 }, { "grad_norm": 0.4838930666446686, "learning_rate": 3.869248741325679e-05, "loss": 0.0475, "step": 35640 }, { "grad_norm": 0.38354572653770447, "learning_rate": 3.866564525817992e-05, "loss": 0.0394, "step": 35650 }, { "grad_norm": 0.3601970672607422, "learning_rate": 3.8638806546180725e-05, "loss": 0.0399, "step": 35660 }, { "grad_norm": 0.4782958924770355, "learning_rate": 3.861197128541213e-05, "loss": 0.03, "step": 35670 }, { "grad_norm": 0.3808307647705078, "learning_rate": 3.858513948402599e-05, "loss": 0.0337, "step": 35680 }, { "grad_norm": 0.4959695637226105, "learning_rate": 3.8558311150173077e-05, "loss": 0.0397, "step": 35690 }, { "grad_norm": 0.2992154359817505, "learning_rate": 3.853148629200312e-05, "loss": 0.0443, "step": 35700 }, { "grad_norm": 0.3561832010746002, "learning_rate": 3.850466491766482e-05, "loss": 0.0428, "step": 35710 }, { "grad_norm": 0.4696481227874756, "learning_rate": 3.847784703530583e-05, "loss": 0.0404, "step": 35720 }, { "grad_norm": 0.5594174861907959, "learning_rate": 3.845103265307266e-05, "loss": 0.0382, "step": 35730 }, { "grad_norm": 0.6463385224342346, "learning_rate": 3.842422177911086e-05, "loss": 0.0407, "step": 35740 }, { "grad_norm": 0.4118404984474182, "learning_rate": 3.8397414421564826e-05, "loss": 0.0392, "step": 35750 }, { "grad_norm": 0.6409322023391724, "learning_rate": 3.8370610588577935e-05, "loss": 0.0417, "step": 35760 }, { "grad_norm": 0.6822523474693298, "learning_rate": 3.834381028829251e-05, "loss": 0.0357, "step": 35770 }, { "grad_norm": 0.8366168141365051, "learning_rate": 3.8317013528849745e-05, "loss": 0.0469, "step": 35780 }, { "grad_norm": 0.6267874240875244, "learning_rate": 3.8290220318389815e-05, "loss": 0.0376, "step": 35790 }, { "grad_norm": 0.4441753923892975, "learning_rate": 3.8263430665051746e-05, "loss": 0.0435, "step": 35800 }, { "grad_norm": 0.4377976059913635, "learning_rate": 3.8236644576973554e-05, "loss": 0.0422, "step": 35810 }, { "grad_norm": 0.7961519956588745, "learning_rate": 3.820986206229217e-05, "loss": 0.0602, "step": 35820 }, { "grad_norm": 0.6134255528450012, "learning_rate": 3.8183083129143384e-05, "loss": 0.0387, "step": 35830 }, { "grad_norm": 0.663566529750824, "learning_rate": 3.815630778566193e-05, "loss": 0.0367, "step": 35840 }, { "grad_norm": 1.166221261024475, "learning_rate": 3.812953603998145e-05, "loss": 0.0538, "step": 35850 }, { "grad_norm": 0.6088013052940369, "learning_rate": 3.8102767900234504e-05, "loss": 0.0404, "step": 35860 }, { "grad_norm": 0.644365668296814, "learning_rate": 3.807600337455256e-05, "loss": 0.0434, "step": 35870 }, { "grad_norm": 0.42038506269454956, "learning_rate": 3.804924247106593e-05, "loss": 0.038, "step": 35880 }, { "grad_norm": 0.4092133939266205, "learning_rate": 3.8022485197903925e-05, "loss": 0.0612, "step": 35890 }, { "grad_norm": 0.368439257144928, "learning_rate": 3.799573156319464e-05, "loss": 0.0427, "step": 35900 }, { "grad_norm": 0.5081048011779785, "learning_rate": 3.796898157506515e-05, "loss": 0.0419, "step": 35910 }, { "grad_norm": 0.5636425614356995, "learning_rate": 3.794223524164143e-05, "loss": 0.05, "step": 35920 }, { "grad_norm": 0.498939573764801, "learning_rate": 3.7915492571048245e-05, "loss": 0.0309, "step": 35930 }, { "grad_norm": 0.39499396085739136, "learning_rate": 3.788875357140937e-05, "loss": 0.0284, "step": 35940 }, { "grad_norm": 0.44927239418029785, "learning_rate": 3.786201825084736e-05, "loss": 0.0528, "step": 35950 }, { "grad_norm": 0.6221327781677246, "learning_rate": 3.783528661748372e-05, "loss": 0.0447, "step": 35960 }, { "grad_norm": 0.42917126417160034, "learning_rate": 3.780855867943882e-05, "loss": 0.0351, "step": 35970 }, { "grad_norm": 0.34450986981391907, "learning_rate": 3.778183444483189e-05, "loss": 0.0289, "step": 35980 }, { "grad_norm": 0.4515971541404724, "learning_rate": 3.775511392178108e-05, "loss": 0.0592, "step": 35990 }, { "grad_norm": 0.39163216948509216, "learning_rate": 3.772839711840332e-05, "loss": 0.0389, "step": 36000 }, { "grad_norm": 0.6460850238800049, "learning_rate": 3.7701684042814515e-05, "loss": 0.0418, "step": 36010 }, { "grad_norm": 0.3338065445423126, "learning_rate": 3.76749747031294e-05, "loss": 0.0582, "step": 36020 }, { "grad_norm": 0.6181204319000244, "learning_rate": 3.764826910746152e-05, "loss": 0.0447, "step": 36030 }, { "grad_norm": 1.0368599891662598, "learning_rate": 3.762156726392338e-05, "loss": 0.0428, "step": 36040 }, { "grad_norm": 0.40973901748657227, "learning_rate": 3.759486918062625e-05, "loss": 0.0494, "step": 36050 }, { "grad_norm": 0.6093835234642029, "learning_rate": 3.756817486568033e-05, "loss": 0.0389, "step": 36060 }, { "grad_norm": 0.8395344614982605, "learning_rate": 3.7541484327194654e-05, "loss": 0.0397, "step": 36070 }, { "grad_norm": 0.4756925404071808, "learning_rate": 3.751479757327707e-05, "loss": 0.0423, "step": 36080 }, { "grad_norm": 0.6878725290298462, "learning_rate": 3.7488114612034345e-05, "loss": 0.0409, "step": 36090 }, { "grad_norm": 0.33930063247680664, "learning_rate": 3.7461435451572044e-05, "loss": 0.033, "step": 36100 }, { "grad_norm": 0.6108651757240295, "learning_rate": 3.743476009999459e-05, "loss": 0.031, "step": 36110 }, { "grad_norm": 0.413141131401062, "learning_rate": 3.7408088565405245e-05, "loss": 0.0278, "step": 36120 }, { "grad_norm": 0.6426069736480713, "learning_rate": 3.738142085590612e-05, "loss": 0.0474, "step": 36130 }, { "grad_norm": 0.5070937275886536, "learning_rate": 3.7354756979598194e-05, "loss": 0.0381, "step": 36140 }, { "grad_norm": 0.472383052110672, "learning_rate": 3.7328096944581187e-05, "loss": 0.0436, "step": 36150 }, { "grad_norm": 0.7454865574836731, "learning_rate": 3.730144075895377e-05, "loss": 0.0411, "step": 36160 }, { "grad_norm": 0.8532199859619141, "learning_rate": 3.727478843081335e-05, "loss": 0.0442, "step": 36170 }, { "grad_norm": 0.5071873664855957, "learning_rate": 3.72481399682562e-05, "loss": 0.0454, "step": 36180 }, { "grad_norm": 0.5540546774864197, "learning_rate": 3.722149537937747e-05, "loss": 0.0434, "step": 36190 }, { "grad_norm": 0.46604427695274353, "learning_rate": 3.7194854672271015e-05, "loss": 0.0456, "step": 36200 }, { "grad_norm": 0.5445098876953125, "learning_rate": 3.7168217855029644e-05, "loss": 0.0342, "step": 36210 }, { "grad_norm": 0.5617196559906006, "learning_rate": 3.7141584935744856e-05, "loss": 0.0406, "step": 36220 }, { "grad_norm": 0.6255719661712646, "learning_rate": 3.7114955922507055e-05, "loss": 0.0397, "step": 36230 }, { "grad_norm": 0.3332878053188324, "learning_rate": 3.708833082340545e-05, "loss": 0.0407, "step": 36240 }, { "grad_norm": 0.4757959246635437, "learning_rate": 3.7061709646528034e-05, "loss": 0.0432, "step": 36250 }, { "grad_norm": 0.6716545820236206, "learning_rate": 3.7035092399961604e-05, "loss": 0.0434, "step": 36260 }, { "grad_norm": 0.5078366994857788, "learning_rate": 3.700847909179177e-05, "loss": 0.0295, "step": 36270 }, { "grad_norm": 0.22461435198783875, "learning_rate": 3.698186973010297e-05, "loss": 0.0332, "step": 36280 }, { "grad_norm": 0.3703479766845703, "learning_rate": 3.695526432297844e-05, "loss": 0.0389, "step": 36290 }, { "grad_norm": 0.5885752439498901, "learning_rate": 3.692866287850017e-05, "loss": 0.0465, "step": 36300 }, { "grad_norm": 0.5012041926383972, "learning_rate": 3.6902065404749006e-05, "loss": 0.0378, "step": 36310 }, { "grad_norm": 0.44043445587158203, "learning_rate": 3.6875471909804516e-05, "loss": 0.0437, "step": 36320 }, { "grad_norm": 0.7226671576499939, "learning_rate": 3.6848882401745135e-05, "loss": 0.0482, "step": 36330 }, { "grad_norm": 0.37463897466659546, "learning_rate": 3.682229688864806e-05, "loss": 0.0388, "step": 36340 }, { "grad_norm": 0.752318799495697, "learning_rate": 3.6795715378589235e-05, "loss": 0.038, "step": 36350 }, { "grad_norm": 0.480761855840683, "learning_rate": 3.676913787964345e-05, "loss": 0.0344, "step": 36360 }, { "grad_norm": 0.45751598477363586, "learning_rate": 3.674256439988423e-05, "loss": 0.0451, "step": 36370 }, { "grad_norm": 0.1899375468492508, "learning_rate": 3.6715994947383904e-05, "loss": 0.0313, "step": 36380 }, { "grad_norm": 0.3035205602645874, "learning_rate": 3.668942953021357e-05, "loss": 0.0305, "step": 36390 }, { "grad_norm": 1.0922273397445679, "learning_rate": 3.66628681564431e-05, "loss": 0.0507, "step": 36400 }, { "grad_norm": 0.5281892418861389, "learning_rate": 3.663631083414114e-05, "loss": 0.0326, "step": 36410 }, { "grad_norm": 0.46076345443725586, "learning_rate": 3.660975757137509e-05, "loss": 0.0436, "step": 36420 }, { "grad_norm": 0.4553043246269226, "learning_rate": 3.658320837621114e-05, "loss": 0.0283, "step": 36430 }, { "grad_norm": 0.5728567242622375, "learning_rate": 3.655666325671426e-05, "loss": 0.0336, "step": 36440 }, { "grad_norm": 0.5397246479988098, "learning_rate": 3.65301222209481e-05, "loss": 0.0338, "step": 36450 }, { "grad_norm": 0.7096982598304749, "learning_rate": 3.650358527697519e-05, "loss": 0.0362, "step": 36460 }, { "grad_norm": 0.6267794370651245, "learning_rate": 3.64770524328567e-05, "loss": 0.0296, "step": 36470 }, { "grad_norm": 0.5663777589797974, "learning_rate": 3.645052369665265e-05, "loss": 0.0341, "step": 36480 }, { "grad_norm": 0.6474859714508057, "learning_rate": 3.6423999076421724e-05, "loss": 0.0406, "step": 36490 }, { "grad_norm": 0.49592459201812744, "learning_rate": 3.639747858022142e-05, "loss": 0.0409, "step": 36500 }, { "grad_norm": 0.7013306617736816, "learning_rate": 3.637096221610799e-05, "loss": 0.03, "step": 36510 }, { "grad_norm": 0.6206126809120178, "learning_rate": 3.634444999213638e-05, "loss": 0.0403, "step": 36520 }, { "grad_norm": 0.47482404112815857, "learning_rate": 3.6317941916360296e-05, "loss": 0.0475, "step": 36530 }, { "grad_norm": 0.5896309614181519, "learning_rate": 3.629143799683221e-05, "loss": 0.0413, "step": 36540 }, { "grad_norm": 0.6249061226844788, "learning_rate": 3.626493824160331e-05, "loss": 0.0334, "step": 36550 }, { "grad_norm": 0.4629449248313904, "learning_rate": 3.623844265872352e-05, "loss": 0.0363, "step": 36560 }, { "grad_norm": 0.3160385489463806, "learning_rate": 3.621195125624149e-05, "loss": 0.0414, "step": 36570 }, { "grad_norm": 0.45499372482299805, "learning_rate": 3.618546404220463e-05, "loss": 0.0313, "step": 36580 }, { "grad_norm": 0.4565986394882202, "learning_rate": 3.615898102465903e-05, "loss": 0.04, "step": 36590 }, { "grad_norm": 0.43148714303970337, "learning_rate": 3.6132502211649544e-05, "loss": 0.0366, "step": 36600 }, { "grad_norm": 0.5444623827934265, "learning_rate": 3.610602761121975e-05, "loss": 0.0323, "step": 36610 }, { "grad_norm": 0.6924699544906616, "learning_rate": 3.6079557231411897e-05, "loss": 0.0578, "step": 36620 }, { "grad_norm": 0.516751229763031, "learning_rate": 3.6053091080267035e-05, "loss": 0.0346, "step": 36630 }, { "grad_norm": 0.45363423228263855, "learning_rate": 3.602662916582483e-05, "loss": 0.0393, "step": 36640 }, { "grad_norm": 0.49048152565956116, "learning_rate": 3.600017149612375e-05, "loss": 0.0426, "step": 36650 }, { "grad_norm": 0.5517372488975525, "learning_rate": 3.5973718079200935e-05, "loss": 0.03, "step": 36660 }, { "grad_norm": 0.8699166178703308, "learning_rate": 3.5947268923092216e-05, "loss": 0.0404, "step": 36670 }, { "grad_norm": 0.49410441517829895, "learning_rate": 3.592082403583216e-05, "loss": 0.0365, "step": 36680 }, { "grad_norm": 0.4552825093269348, "learning_rate": 3.5894383425454004e-05, "loss": 0.0338, "step": 36690 }, { "grad_norm": 0.37196341156959534, "learning_rate": 3.586794709998975e-05, "loss": 0.0304, "step": 36700 }, { "grad_norm": 0.7613407373428345, "learning_rate": 3.584151506747002e-05, "loss": 0.0363, "step": 36710 }, { "grad_norm": 0.552055299282074, "learning_rate": 3.581508733592418e-05, "loss": 0.0406, "step": 36720 }, { "grad_norm": 0.41419801115989685, "learning_rate": 3.5788663913380297e-05, "loss": 0.0384, "step": 36730 }, { "grad_norm": 0.4507136046886444, "learning_rate": 3.576224480786506e-05, "loss": 0.0439, "step": 36740 }, { "grad_norm": 0.5505679845809937, "learning_rate": 3.573583002740393e-05, "loss": 0.0379, "step": 36750 }, { "grad_norm": 0.5975388884544373, "learning_rate": 3.570941958002103e-05, "loss": 0.0348, "step": 36760 }, { "grad_norm": 0.7224553227424622, "learning_rate": 3.568301347373912e-05, "loss": 0.055, "step": 36770 }, { "grad_norm": 0.31540000438690186, "learning_rate": 3.5656611716579726e-05, "loss": 0.0309, "step": 36780 }, { "grad_norm": 0.6900232434272766, "learning_rate": 3.5630214316562946e-05, "loss": 0.0418, "step": 36790 }, { "grad_norm": 0.37736976146698, "learning_rate": 3.560382128170766e-05, "loss": 0.0452, "step": 36800 }, { "grad_norm": 0.5187094211578369, "learning_rate": 3.5577432620031374e-05, "loss": 0.0317, "step": 36810 }, { "grad_norm": 0.40798836946487427, "learning_rate": 3.5551048339550216e-05, "loss": 0.0435, "step": 36820 }, { "grad_norm": 0.33837223052978516, "learning_rate": 3.55246684482791e-05, "loss": 0.0386, "step": 36830 }, { "grad_norm": 0.39311614632606506, "learning_rate": 3.5498292954231496e-05, "loss": 0.0404, "step": 36840 }, { "grad_norm": 0.6154251098632812, "learning_rate": 3.54719218654196e-05, "loss": 0.0473, "step": 36850 }, { "grad_norm": 0.5034754276275635, "learning_rate": 3.544555518985425e-05, "loss": 0.0317, "step": 36860 }, { "grad_norm": 0.22090058028697968, "learning_rate": 3.541919293554494e-05, "loss": 0.0293, "step": 36870 }, { "grad_norm": 0.5976190567016602, "learning_rate": 3.539283511049985e-05, "loss": 0.0284, "step": 36880 }, { "grad_norm": 0.5017393231391907, "learning_rate": 3.5366481722725755e-05, "loss": 0.0348, "step": 36890 }, { "grad_norm": 0.388324499130249, "learning_rate": 3.534013278022816e-05, "loss": 0.0631, "step": 36900 }, { "grad_norm": 0.3779468238353729, "learning_rate": 3.531378829101113e-05, "loss": 0.0316, "step": 36910 }, { "grad_norm": 0.5184879302978516, "learning_rate": 3.528744826307746e-05, "loss": 0.0462, "step": 36920 }, { "grad_norm": 0.5177839398384094, "learning_rate": 3.5261112704428554e-05, "loss": 0.0418, "step": 36930 }, { "grad_norm": 0.35757264494895935, "learning_rate": 3.523478162306443e-05, "loss": 0.0445, "step": 36940 }, { "grad_norm": 0.682052493095398, "learning_rate": 3.520845502698381e-05, "loss": 0.0443, "step": 36950 }, { "grad_norm": 0.5091152191162109, "learning_rate": 3.5182132924184005e-05, "loss": 0.0353, "step": 36960 }, { "grad_norm": 0.5525805354118347, "learning_rate": 3.5155815322660966e-05, "loss": 0.0341, "step": 36970 }, { "grad_norm": 0.42029401659965515, "learning_rate": 3.512950223040931e-05, "loss": 0.036, "step": 36980 }, { "grad_norm": 0.475737065076828, "learning_rate": 3.5103193655422216e-05, "loss": 0.0421, "step": 36990 }, { "grad_norm": 0.5715951323509216, "learning_rate": 3.5076889605691596e-05, "loss": 0.0489, "step": 37000 }, { "grad_norm": 0.344247967004776, "learning_rate": 3.505059008920787e-05, "loss": 0.051, "step": 37010 }, { "grad_norm": 0.4963536560535431, "learning_rate": 3.502429511396016e-05, "loss": 0.0297, "step": 37020 }, { "grad_norm": 0.547787070274353, "learning_rate": 3.4998004687936196e-05, "loss": 0.0337, "step": 37030 }, { "grad_norm": 0.639954686164856, "learning_rate": 3.497171881912229e-05, "loss": 0.045, "step": 37040 }, { "grad_norm": 0.4853415787220001, "learning_rate": 3.494543751550342e-05, "loss": 0.0422, "step": 37050 }, { "grad_norm": 0.4324636459350586, "learning_rate": 3.491916078506313e-05, "loss": 0.0421, "step": 37060 }, { "grad_norm": 0.28254517912864685, "learning_rate": 3.489288863578361e-05, "loss": 0.0363, "step": 37070 }, { "grad_norm": 0.5417099595069885, "learning_rate": 3.4866621075645646e-05, "loss": 0.0348, "step": 37080 }, { "grad_norm": 0.405921995639801, "learning_rate": 3.4840358112628614e-05, "loss": 0.0393, "step": 37090 }, { "grad_norm": 0.29808345437049866, "learning_rate": 3.481409975471053e-05, "loss": 0.0424, "step": 37100 }, { "grad_norm": 0.491036593914032, "learning_rate": 3.4787846009867986e-05, "loss": 0.0467, "step": 37110 }, { "grad_norm": 0.24628667533397675, "learning_rate": 3.476159688607615e-05, "loss": 0.0583, "step": 37120 }, { "grad_norm": 0.40729281306266785, "learning_rate": 3.4735352391308854e-05, "loss": 0.0317, "step": 37130 }, { "grad_norm": 0.4753563404083252, "learning_rate": 3.4709112533538446e-05, "loss": 0.0316, "step": 37140 }, { "grad_norm": 0.5873309969902039, "learning_rate": 3.4682877320735934e-05, "loss": 0.0268, "step": 37150 }, { "grad_norm": 0.6003186702728271, "learning_rate": 3.465664676087085e-05, "loss": 0.0462, "step": 37160 }, { "grad_norm": 0.3334944248199463, "learning_rate": 3.463042086191136e-05, "loss": 0.0347, "step": 37170 }, { "grad_norm": 0.601223349571228, "learning_rate": 3.460419963182423e-05, "loss": 0.0357, "step": 37180 }, { "grad_norm": 0.8893754482269287, "learning_rate": 3.457798307857473e-05, "loss": 0.0436, "step": 37190 }, { "grad_norm": 0.4418339729309082, "learning_rate": 3.455177121012678e-05, "loss": 0.0371, "step": 37200 }, { "grad_norm": 0.4990105926990509, "learning_rate": 3.452556403444285e-05, "loss": 0.0384, "step": 37210 }, { "grad_norm": 0.31164485216140747, "learning_rate": 3.4499361559483975e-05, "loss": 0.0297, "step": 37220 }, { "grad_norm": 0.5329342484474182, "learning_rate": 3.44731637932098e-05, "loss": 0.0452, "step": 37230 }, { "grad_norm": 0.7134444713592529, "learning_rate": 3.44469707435785e-05, "loss": 0.056, "step": 37240 }, { "grad_norm": 0.5121646523475647, "learning_rate": 3.4420782418546835e-05, "loss": 0.0655, "step": 37250 }, { "grad_norm": 0.44677677750587463, "learning_rate": 3.439459882607012e-05, "loss": 0.0583, "step": 37260 }, { "grad_norm": 0.2896929681301117, "learning_rate": 3.436841997410225e-05, "loss": 0.0257, "step": 37270 }, { "grad_norm": 0.2932506501674652, "learning_rate": 3.434224587059567e-05, "loss": 0.0438, "step": 37280 }, { "grad_norm": 0.48774436116218567, "learning_rate": 3.431607652350136e-05, "loss": 0.0369, "step": 37290 }, { "grad_norm": 0.7303705215454102, "learning_rate": 3.428991194076891e-05, "loss": 0.0348, "step": 37300 }, { "grad_norm": 0.3438964784145355, "learning_rate": 3.4263752130346394e-05, "loss": 0.0458, "step": 37310 }, { "grad_norm": 0.4551008939743042, "learning_rate": 3.4237597100180515e-05, "loss": 0.0378, "step": 37320 }, { "grad_norm": 0.36409321427345276, "learning_rate": 3.4211446858216427e-05, "loss": 0.0345, "step": 37330 }, { "grad_norm": 0.7232042551040649, "learning_rate": 3.4185301412397915e-05, "loss": 0.0449, "step": 37340 }, { "grad_norm": 0.5003844499588013, "learning_rate": 3.415916077066729e-05, "loss": 0.0249, "step": 37350 }, { "grad_norm": 0.9237574338912964, "learning_rate": 3.413302494096535e-05, "loss": 0.0353, "step": 37360 }, { "grad_norm": 0.4013870060443878, "learning_rate": 3.410689393123151e-05, "loss": 0.0373, "step": 37370 }, { "grad_norm": 1.1412829160690308, "learning_rate": 3.408076774940364e-05, "loss": 0.0333, "step": 37380 }, { "grad_norm": 0.3197697699069977, "learning_rate": 3.40546464034182e-05, "loss": 0.0237, "step": 37390 }, { "grad_norm": 0.6311700940132141, "learning_rate": 3.4028529901210185e-05, "loss": 0.038, "step": 37400 }, { "grad_norm": 0.48976534605026245, "learning_rate": 3.4002418250713086e-05, "loss": 0.0359, "step": 37410 }, { "grad_norm": 0.2903216779232025, "learning_rate": 3.3976311459858936e-05, "loss": 0.0308, "step": 37420 }, { "grad_norm": 0.5429863929748535, "learning_rate": 3.395020953657826e-05, "loss": 0.0335, "step": 37430 }, { "grad_norm": 0.6823787093162537, "learning_rate": 3.3924112488800165e-05, "loss": 0.0463, "step": 37440 }, { "grad_norm": 0.5102013349533081, "learning_rate": 3.389802032445225e-05, "loss": 0.0639, "step": 37450 }, { "grad_norm": 0.30712956190109253, "learning_rate": 3.38719330514606e-05, "loss": 0.0516, "step": 37460 }, { "grad_norm": 0.331039160490036, "learning_rate": 3.3845850677749866e-05, "loss": 0.0289, "step": 37470 }, { "grad_norm": 0.686764121055603, "learning_rate": 3.3819773211243157e-05, "loss": 0.0395, "step": 37480 }, { "grad_norm": 0.29968565702438354, "learning_rate": 3.379370065986213e-05, "loss": 0.0279, "step": 37490 }, { "grad_norm": 0.6171908974647522, "learning_rate": 3.3767633031526955e-05, "loss": 0.0345, "step": 37500 }, { "grad_norm": 0.6258488297462463, "learning_rate": 3.374157033415626e-05, "loss": 0.045, "step": 37510 }, { "grad_norm": 0.306361585855484, "learning_rate": 3.371551257566723e-05, "loss": 0.0411, "step": 37520 }, { "grad_norm": 0.34279191493988037, "learning_rate": 3.36894597639755e-05, "loss": 0.051, "step": 37530 }, { "grad_norm": 0.8306016325950623, "learning_rate": 3.366341190699523e-05, "loss": 0.0524, "step": 37540 }, { "grad_norm": 0.6777400970458984, "learning_rate": 3.36373690126391e-05, "loss": 0.033, "step": 37550 }, { "grad_norm": 0.3606766164302826, "learning_rate": 3.3611331088818234e-05, "loss": 0.0631, "step": 37560 }, { "grad_norm": 0.521713376045227, "learning_rate": 3.3585298143442265e-05, "loss": 0.0352, "step": 37570 }, { "grad_norm": 0.3708246052265167, "learning_rate": 3.35592701844193e-05, "loss": 0.0365, "step": 37580 }, { "grad_norm": 0.4539405405521393, "learning_rate": 3.353324721965596e-05, "loss": 0.0348, "step": 37590 }, { "grad_norm": 0.5557113289833069, "learning_rate": 3.350722925705736e-05, "loss": 0.0452, "step": 37600 }, { "grad_norm": 0.5853052735328674, "learning_rate": 3.348121630452703e-05, "loss": 0.041, "step": 37610 }, { "grad_norm": 0.7424765825271606, "learning_rate": 3.3455208369967044e-05, "loss": 0.0314, "step": 37620 }, { "grad_norm": 0.47490766644477844, "learning_rate": 3.34292054612779e-05, "loss": 0.0355, "step": 37630 }, { "grad_norm": 0.373816579580307, "learning_rate": 3.340320758635861e-05, "loss": 0.0401, "step": 37640 }, { "grad_norm": 0.6972861886024475, "learning_rate": 3.337721475310666e-05, "loss": 0.0427, "step": 37650 }, { "grad_norm": 0.582054615020752, "learning_rate": 3.335122696941795e-05, "loss": 0.0331, "step": 37660 }, { "grad_norm": 0.37694311141967773, "learning_rate": 3.332524424318692e-05, "loss": 0.0366, "step": 37670 }, { "grad_norm": 0.683347761631012, "learning_rate": 3.32992665823064e-05, "loss": 0.0386, "step": 37680 }, { "grad_norm": 0.48949578404426575, "learning_rate": 3.327329399466774e-05, "loss": 0.0381, "step": 37690 }, { "grad_norm": 0.7916682958602905, "learning_rate": 3.324732648816072e-05, "loss": 0.04, "step": 37700 }, { "grad_norm": 0.3802902400493622, "learning_rate": 3.322136407067358e-05, "loss": 0.0385, "step": 37710 }, { "grad_norm": 0.4802035093307495, "learning_rate": 3.3195406750093036e-05, "loss": 0.0394, "step": 37720 }, { "grad_norm": 0.4376014173030853, "learning_rate": 3.3169454534304205e-05, "loss": 0.0327, "step": 37730 }, { "grad_norm": 0.4546603858470917, "learning_rate": 3.3143507431190725e-05, "loss": 0.0476, "step": 37740 }, { "grad_norm": 0.2778938114643097, "learning_rate": 3.311756544863459e-05, "loss": 0.0278, "step": 37750 }, { "grad_norm": 0.49383458495140076, "learning_rate": 3.309162859451633e-05, "loss": 0.0275, "step": 37760 }, { "grad_norm": 0.5883743762969971, "learning_rate": 3.306569687671487e-05, "loss": 0.0376, "step": 37770 }, { "grad_norm": 0.22028058767318726, "learning_rate": 3.303977030310756e-05, "loss": 0.0432, "step": 37780 }, { "grad_norm": 0.5455597043037415, "learning_rate": 3.3013848881570245e-05, "loss": 0.0362, "step": 37790 }, { "grad_norm": 0.39561834931373596, "learning_rate": 3.298793261997712e-05, "loss": 0.0474, "step": 37800 }, { "grad_norm": 0.5328279137611389, "learning_rate": 3.2962021526200893e-05, "loss": 0.0301, "step": 37810 }, { "grad_norm": 0.38247764110565186, "learning_rate": 3.293611560811268e-05, "loss": 0.0378, "step": 37820 }, { "grad_norm": 0.5323994755744934, "learning_rate": 3.291021487358199e-05, "loss": 0.0397, "step": 37830 }, { "grad_norm": 0.7155994176864624, "learning_rate": 3.28843193304768e-05, "loss": 0.0403, "step": 37840 }, { "grad_norm": 0.2829228937625885, "learning_rate": 3.2858428986663456e-05, "loss": 0.0314, "step": 37850 }, { "grad_norm": 0.5405501127243042, "learning_rate": 3.283254385000681e-05, "loss": 0.0395, "step": 37860 }, { "grad_norm": 0.3923173248767853, "learning_rate": 3.2806663928370076e-05, "loss": 0.0409, "step": 37870 }, { "grad_norm": 0.7412986159324646, "learning_rate": 3.278078922961485e-05, "loss": 0.0358, "step": 37880 }, { "grad_norm": 0.5098429322242737, "learning_rate": 3.275491976160123e-05, "loss": 0.0337, "step": 37890 }, { "grad_norm": 0.4839889407157898, "learning_rate": 3.2729055532187645e-05, "loss": 0.0319, "step": 37900 }, { "grad_norm": 0.4939863681793213, "learning_rate": 3.270319654923097e-05, "loss": 0.0412, "step": 37910 }, { "grad_norm": 0.519672691822052, "learning_rate": 3.2677342820586506e-05, "loss": 0.0411, "step": 37920 }, { "grad_norm": 0.5838947296142578, "learning_rate": 3.2651494354107905e-05, "loss": 0.0408, "step": 37930 }, { "grad_norm": 0.47715237736701965, "learning_rate": 3.2625651157647266e-05, "loss": 0.0432, "step": 37940 }, { "grad_norm": 0.5462238788604736, "learning_rate": 3.259981323905505e-05, "loss": 0.0329, "step": 37950 }, { "grad_norm": 0.5172702670097351, "learning_rate": 3.257398060618014e-05, "loss": 0.0413, "step": 37960 }, { "grad_norm": 0.2603242099285126, "learning_rate": 3.254815326686983e-05, "loss": 0.0349, "step": 37970 }, { "grad_norm": 0.553459107875824, "learning_rate": 3.2522331228969774e-05, "loss": 0.0408, "step": 37980 }, { "grad_norm": 0.49473080039024353, "learning_rate": 3.2496514500324006e-05, "loss": 0.0469, "step": 37990 }, { "grad_norm": 0.30973851680755615, "learning_rate": 3.247070308877498e-05, "loss": 0.0461, "step": 38000 }, { "grad_norm": 0.709671676158905, "learning_rate": 3.2444897002163515e-05, "loss": 0.0462, "step": 38010 }, { "grad_norm": 0.27766644954681396, "learning_rate": 3.241909624832885e-05, "loss": 0.0295, "step": 38020 }, { "grad_norm": 0.523536741733551, "learning_rate": 3.239330083510852e-05, "loss": 0.0398, "step": 38030 }, { "grad_norm": 0.29828107357025146, "learning_rate": 3.236751077033855e-05, "loss": 0.036, "step": 38040 }, { "grad_norm": 0.2581343650817871, "learning_rate": 3.234172606185322e-05, "loss": 0.0231, "step": 38050 }, { "grad_norm": 0.6004844903945923, "learning_rate": 3.231594671748528e-05, "loss": 0.034, "step": 38060 }, { "grad_norm": 0.8013452291488647, "learning_rate": 3.2290172745065815e-05, "loss": 0.0445, "step": 38070 }, { "grad_norm": 0.26177603006362915, "learning_rate": 3.226440415242426e-05, "loss": 0.0324, "step": 38080 }, { "grad_norm": 0.42776528000831604, "learning_rate": 3.223864094738846e-05, "loss": 0.0305, "step": 38090 }, { "grad_norm": 0.24418683350086212, "learning_rate": 3.221288313778456e-05, "loss": 0.0318, "step": 38100 }, { "grad_norm": 0.31149616837501526, "learning_rate": 3.2187130731437125e-05, "loss": 0.0246, "step": 38110 }, { "grad_norm": 0.3922146260738373, "learning_rate": 3.216138373616905e-05, "loss": 0.0377, "step": 38120 }, { "grad_norm": 0.41752681136131287, "learning_rate": 3.21356421598016e-05, "loss": 0.046, "step": 38130 }, { "grad_norm": 0.36297452449798584, "learning_rate": 3.210990601015438e-05, "loss": 0.0367, "step": 38140 }, { "grad_norm": 0.4669826626777649, "learning_rate": 3.208417529504535e-05, "loss": 0.0369, "step": 38150 }, { "grad_norm": 0.2661777138710022, "learning_rate": 3.205845002229084e-05, "loss": 0.029, "step": 38160 }, { "grad_norm": 0.3672369718551636, "learning_rate": 3.203273019970547e-05, "loss": 0.0402, "step": 38170 }, { "grad_norm": 0.42537522315979004, "learning_rate": 3.200701583510227e-05, "loss": 0.0338, "step": 38180 }, { "grad_norm": 0.5030924677848816, "learning_rate": 3.198130693629261e-05, "loss": 0.0485, "step": 38190 }, { "grad_norm": 0.5355584025382996, "learning_rate": 3.195560351108612e-05, "loss": 0.0321, "step": 38200 }, { "grad_norm": 0.3659297823905945, "learning_rate": 3.1929905567290865e-05, "loss": 0.0336, "step": 38210 }, { "grad_norm": 0.3336765468120575, "learning_rate": 3.1904213112713164e-05, "loss": 0.0304, "step": 38220 }, { "grad_norm": 1.0683648586273193, "learning_rate": 3.187852615515774e-05, "loss": 0.032, "step": 38230 }, { "grad_norm": 0.43668755888938904, "learning_rate": 3.1852844702427606e-05, "loss": 0.0313, "step": 38240 }, { "grad_norm": 0.1817004233598709, "learning_rate": 3.18271687623241e-05, "loss": 0.0375, "step": 38250 }, { "grad_norm": 0.453838586807251, "learning_rate": 3.1801498342646896e-05, "loss": 0.0517, "step": 38260 }, { "grad_norm": 0.4551329016685486, "learning_rate": 3.177583345119398e-05, "loss": 0.0304, "step": 38270 }, { "grad_norm": 0.4194262623786926, "learning_rate": 3.17501740957617e-05, "loss": 0.0372, "step": 38280 }, { "grad_norm": 0.3446890711784363, "learning_rate": 3.172452028414467e-05, "loss": 0.0377, "step": 38290 }, { "grad_norm": 0.4569678008556366, "learning_rate": 3.169887202413583e-05, "loss": 0.0254, "step": 38300 }, { "grad_norm": 0.4935709834098816, "learning_rate": 3.167322932352646e-05, "loss": 0.0363, "step": 38310 }, { "grad_norm": 0.44788867235183716, "learning_rate": 3.164759219010613e-05, "loss": 0.031, "step": 38320 }, { "grad_norm": 0.44960999488830566, "learning_rate": 3.1621960631662725e-05, "loss": 0.0258, "step": 38330 }, { "grad_norm": 0.475160151720047, "learning_rate": 3.159633465598245e-05, "loss": 0.0392, "step": 38340 }, { "grad_norm": 0.3549397587776184, "learning_rate": 3.1570714270849767e-05, "loss": 0.0293, "step": 38350 }, { "grad_norm": 0.5525527000427246, "learning_rate": 3.1545099484047516e-05, "loss": 0.0348, "step": 38360 }, { "grad_norm": 0.6578903198242188, "learning_rate": 3.151949030335674e-05, "loss": 0.0659, "step": 38370 }, { "grad_norm": 0.485774964094162, "learning_rate": 3.149388673655687e-05, "loss": 0.0363, "step": 38380 }, { "grad_norm": 0.6545245051383972, "learning_rate": 3.146828879142559e-05, "loss": 0.0398, "step": 38390 }, { "grad_norm": 0.3230476379394531, "learning_rate": 3.1442696475738866e-05, "loss": 0.0338, "step": 38400 }, { "grad_norm": 0.41321104764938354, "learning_rate": 3.141710979727098e-05, "loss": 0.0411, "step": 38410 }, { "grad_norm": 0.5649597644805908, "learning_rate": 3.139152876379447e-05, "loss": 0.0333, "step": 38420 }, { "grad_norm": 0.43149879574775696, "learning_rate": 3.1365953383080214e-05, "loss": 0.0241, "step": 38430 }, { "grad_norm": 0.3327277600765228, "learning_rate": 3.134038366289731e-05, "loss": 0.0306, "step": 38440 }, { "grad_norm": 0.610876739025116, "learning_rate": 3.131481961101317e-05, "loss": 0.0384, "step": 38450 }, { "grad_norm": 0.45153892040252686, "learning_rate": 3.128926123519349e-05, "loss": 0.0476, "step": 38460 }, { "grad_norm": 0.28594687581062317, "learning_rate": 3.1263708543202194e-05, "loss": 0.0342, "step": 38470 }, { "grad_norm": 0.5460287928581238, "learning_rate": 3.123816154280155e-05, "loss": 0.043, "step": 38480 }, { "grad_norm": 0.5816922783851624, "learning_rate": 3.121262024175207e-05, "loss": 0.0383, "step": 38490 }, { "grad_norm": 0.6521205306053162, "learning_rate": 3.118708464781248e-05, "loss": 0.0343, "step": 38500 }, { "grad_norm": 0.30205440521240234, "learning_rate": 3.116155476873987e-05, "loss": 0.0301, "step": 38510 }, { "grad_norm": 0.3455589711666107, "learning_rate": 3.11360306122895e-05, "loss": 0.0258, "step": 38520 }, { "grad_norm": 0.42846500873565674, "learning_rate": 3.1110512186214975e-05, "loss": 0.0252, "step": 38530 }, { "grad_norm": 0.3831903338432312, "learning_rate": 3.1084999498268095e-05, "loss": 0.025, "step": 38540 }, { "grad_norm": 1.1522718667984009, "learning_rate": 3.1059492556198934e-05, "loss": 0.0517, "step": 38550 }, { "grad_norm": 0.6280884742736816, "learning_rate": 3.103399136775586e-05, "loss": 0.0483, "step": 38560 }, { "grad_norm": 0.6579577326774597, "learning_rate": 3.100849594068541e-05, "loss": 0.0426, "step": 38570 }, { "grad_norm": 0.5057029128074646, "learning_rate": 3.0983006282732484e-05, "loss": 0.0402, "step": 38580 }, { "grad_norm": 1.0610332489013672, "learning_rate": 3.0957522401640116e-05, "loss": 0.0353, "step": 38590 }, { "grad_norm": 0.35305076837539673, "learning_rate": 3.0932044305149645e-05, "loss": 0.0302, "step": 38600 }, { "grad_norm": 0.2569689154624939, "learning_rate": 3.090657200100068e-05, "loss": 0.0351, "step": 38610 }, { "grad_norm": 0.4651903808116913, "learning_rate": 3.088110549693099e-05, "loss": 0.0387, "step": 38620 }, { "grad_norm": 0.3910926580429077, "learning_rate": 3.085564480067667e-05, "loss": 0.0381, "step": 38630 }, { "grad_norm": 0.5013723969459534, "learning_rate": 3.0830189919971955e-05, "loss": 0.042, "step": 38640 }, { "grad_norm": 0.4101734459400177, "learning_rate": 3.080474086254939e-05, "loss": 0.0441, "step": 38650 }, { "grad_norm": 0.4915618896484375, "learning_rate": 3.077929763613975e-05, "loss": 0.0643, "step": 38660 }, { "grad_norm": 0.49587711691856384, "learning_rate": 3.075386024847198e-05, "loss": 0.0309, "step": 38670 }, { "grad_norm": 0.6373248100280762, "learning_rate": 3.072842870727331e-05, "loss": 0.062, "step": 38680 }, { "grad_norm": 0.3812299966812134, "learning_rate": 3.070300302026916e-05, "loss": 0.0383, "step": 38690 }, { "grad_norm": 0.4240989089012146, "learning_rate": 3.067758319518318e-05, "loss": 0.0457, "step": 38700 }, { "grad_norm": 0.4877869784832001, "learning_rate": 3.065216923973725e-05, "loss": 0.03, "step": 38710 }, { "grad_norm": 0.4501579701900482, "learning_rate": 3.062676116165145e-05, "loss": 0.0285, "step": 38720 }, { "grad_norm": 0.3439289927482605, "learning_rate": 3.06013589686441e-05, "loss": 0.0293, "step": 38730 }, { "grad_norm": 0.4514710605144501, "learning_rate": 3.05759626684317e-05, "loss": 0.0354, "step": 38740 }, { "grad_norm": 0.40754997730255127, "learning_rate": 3.055057226872896e-05, "loss": 0.0389, "step": 38750 }, { "grad_norm": 0.37831804156303406, "learning_rate": 3.052518777724887e-05, "loss": 0.0552, "step": 38760 }, { "grad_norm": 0.6805821657180786, "learning_rate": 3.04998092017025e-05, "loss": 0.0392, "step": 38770 }, { "grad_norm": 0.42452186346054077, "learning_rate": 3.0474436549799246e-05, "loss": 0.037, "step": 38780 }, { "grad_norm": 0.44019660353660583, "learning_rate": 3.044906982924661e-05, "loss": 0.0287, "step": 38790 }, { "grad_norm": 0.6853305697441101, "learning_rate": 3.0423709047750337e-05, "loss": 0.0428, "step": 38800 }, { "grad_norm": 0.7470676898956299, "learning_rate": 3.03983542130144e-05, "loss": 0.0319, "step": 38810 }, { "grad_norm": 0.5965067148208618, "learning_rate": 3.0373005332740877e-05, "loss": 0.0332, "step": 38820 }, { "grad_norm": 0.4476011395454407, "learning_rate": 3.034766241463013e-05, "loss": 0.038, "step": 38830 }, { "grad_norm": 0.3564385771751404, "learning_rate": 3.032232546638064e-05, "loss": 0.0348, "step": 38840 }, { "grad_norm": 0.8926910758018494, "learning_rate": 3.0296994495689114e-05, "loss": 0.0397, "step": 38850 }, { "grad_norm": 0.8318261504173279, "learning_rate": 3.0271669510250444e-05, "loss": 0.0601, "step": 38860 }, { "grad_norm": 0.5761263966560364, "learning_rate": 3.024635051775766e-05, "loss": 0.0291, "step": 38870 }, { "grad_norm": 0.40981581807136536, "learning_rate": 3.022103752590205e-05, "loss": 0.0399, "step": 38880 }, { "grad_norm": 0.46171510219573975, "learning_rate": 3.0195730542372992e-05, "loss": 0.0349, "step": 38890 }, { "grad_norm": 0.589672327041626, "learning_rate": 3.0170429574858084e-05, "loss": 0.0404, "step": 38900 }, { "grad_norm": 0.5907676815986633, "learning_rate": 3.0145134631043127e-05, "loss": 0.0411, "step": 38910 }, { "grad_norm": 0.4519406259059906, "learning_rate": 3.0119845718612018e-05, "loss": 0.0203, "step": 38920 }, { "grad_norm": 0.3929522931575775, "learning_rate": 3.009456284524688e-05, "loss": 0.0372, "step": 38930 }, { "grad_norm": 0.4534853994846344, "learning_rate": 3.0069286018627967e-05, "loss": 0.0338, "step": 38940 }, { "grad_norm": 0.356660932302475, "learning_rate": 3.0044015246433743e-05, "loss": 0.0259, "step": 38950 }, { "grad_norm": 0.5562078952789307, "learning_rate": 3.0018750536340755e-05, "loss": 0.0373, "step": 38960 }, { "grad_norm": 0.5737535953521729, "learning_rate": 2.999349189602378e-05, "loss": 0.0405, "step": 38970 }, { "grad_norm": 0.809333860874176, "learning_rate": 2.9968239333155733e-05, "loss": 0.0613, "step": 38980 }, { "grad_norm": 0.5862447619438171, "learning_rate": 2.994299285540767e-05, "loss": 0.0325, "step": 38990 }, { "grad_norm": 0.45949822664260864, "learning_rate": 2.9917752470448813e-05, "loss": 0.0309, "step": 39000 }, { "grad_norm": 0.3673809766769409, "learning_rate": 2.9892518185946495e-05, "loss": 0.0416, "step": 39010 }, { "grad_norm": 0.5415927767753601, "learning_rate": 2.986729000956624e-05, "loss": 0.0363, "step": 39020 }, { "grad_norm": 0.6848320960998535, "learning_rate": 2.9842067948971736e-05, "loss": 0.0296, "step": 39030 }, { "grad_norm": 0.5794480443000793, "learning_rate": 2.9816852011824727e-05, "loss": 0.0346, "step": 39040 }, { "grad_norm": 0.5805789828300476, "learning_rate": 2.979164220578519e-05, "loss": 0.0346, "step": 39050 }, { "grad_norm": 0.6936026215553284, "learning_rate": 2.9766438538511165e-05, "loss": 0.0361, "step": 39060 }, { "grad_norm": 0.35671231150627136, "learning_rate": 2.9741241017658873e-05, "loss": 0.0312, "step": 39070 }, { "grad_norm": 0.39720773696899414, "learning_rate": 2.971604965088267e-05, "loss": 0.0281, "step": 39080 }, { "grad_norm": 0.43086400628089905, "learning_rate": 2.9690864445835008e-05, "loss": 0.0388, "step": 39090 }, { "grad_norm": 0.5174016952514648, "learning_rate": 2.966568541016651e-05, "loss": 0.0368, "step": 39100 }, { "grad_norm": 0.39563360810279846, "learning_rate": 2.9640512551525867e-05, "loss": 0.0394, "step": 39110 }, { "grad_norm": 0.8423597812652588, "learning_rate": 2.961534587755995e-05, "loss": 0.0413, "step": 39120 }, { "grad_norm": 0.5570090413093567, "learning_rate": 2.959018539591375e-05, "loss": 0.0354, "step": 39130 }, { "grad_norm": 1.1126611232757568, "learning_rate": 2.9565031114230325e-05, "loss": 0.0363, "step": 39140 }, { "grad_norm": 0.36193498969078064, "learning_rate": 2.9539883040150895e-05, "loss": 0.0343, "step": 39150 }, { "grad_norm": 0.39540278911590576, "learning_rate": 2.9514741181314774e-05, "loss": 0.0363, "step": 39160 }, { "grad_norm": 0.24871380627155304, "learning_rate": 2.94896055453594e-05, "loss": 0.0325, "step": 39170 }, { "grad_norm": 0.2559897005558014, "learning_rate": 2.9464476139920332e-05, "loss": 0.0299, "step": 39180 }, { "grad_norm": 0.3694121539592743, "learning_rate": 2.9439352972631186e-05, "loss": 0.0365, "step": 39190 }, { "grad_norm": 0.7053818702697754, "learning_rate": 2.9414236051123757e-05, "loss": 0.0288, "step": 39200 }, { "grad_norm": 0.2852344810962677, "learning_rate": 2.938912538302785e-05, "loss": 0.0328, "step": 39210 }, { "grad_norm": 0.4591030776500702, "learning_rate": 2.9364020975971464e-05, "loss": 0.0314, "step": 39220 }, { "grad_norm": 0.5980330109596252, "learning_rate": 2.9338922837580657e-05, "loss": 0.0405, "step": 39230 }, { "grad_norm": 0.4263152778148651, "learning_rate": 2.931383097547955e-05, "loss": 0.0301, "step": 39240 }, { "grad_norm": 0.3906710743904114, "learning_rate": 2.928874539729043e-05, "loss": 0.0437, "step": 39250 }, { "grad_norm": 0.3005788326263428, "learning_rate": 2.926366611063358e-05, "loss": 0.0237, "step": 39260 }, { "grad_norm": 0.3905152976512909, "learning_rate": 2.9238593123127463e-05, "loss": 0.0231, "step": 39270 }, { "grad_norm": 0.19994650781154633, "learning_rate": 2.9213526442388583e-05, "loss": 0.0274, "step": 39280 }, { "grad_norm": 0.5549001097679138, "learning_rate": 2.9188466076031545e-05, "loss": 0.03, "step": 39290 }, { "grad_norm": 0.575748085975647, "learning_rate": 2.9163412031669012e-05, "loss": 0.0332, "step": 39300 }, { "grad_norm": 0.3554759919643402, "learning_rate": 2.913836431691175e-05, "loss": 0.0283, "step": 39310 }, { "grad_norm": 0.34452152252197266, "learning_rate": 2.9113322939368583e-05, "loss": 0.0296, "step": 39320 }, { "grad_norm": 0.4786737561225891, "learning_rate": 2.9088287906646427e-05, "loss": 0.0248, "step": 39330 }, { "grad_norm": 0.36717480421066284, "learning_rate": 2.906325922635024e-05, "loss": 0.0321, "step": 39340 }, { "grad_norm": 0.4912073314189911, "learning_rate": 2.903823690608313e-05, "loss": 0.0354, "step": 39350 }, { "grad_norm": 0.3139062821865082, "learning_rate": 2.9013220953446174e-05, "loss": 0.0265, "step": 39360 }, { "grad_norm": 0.3937898874282837, "learning_rate": 2.8988211376038564e-05, "loss": 0.0417, "step": 39370 }, { "grad_norm": 0.3436187505722046, "learning_rate": 2.8963208181457564e-05, "loss": 0.0316, "step": 39380 }, { "grad_norm": 0.5002281069755554, "learning_rate": 2.8938211377298453e-05, "loss": 0.0335, "step": 39390 }, { "grad_norm": 0.5759468078613281, "learning_rate": 2.8913220971154652e-05, "loss": 0.0363, "step": 39400 }, { "grad_norm": 0.6471039056777954, "learning_rate": 2.888823697061753e-05, "loss": 0.0349, "step": 39410 }, { "grad_norm": 0.30453619360923767, "learning_rate": 2.8863259383276618e-05, "loss": 0.0291, "step": 39420 }, { "grad_norm": 0.5541936755180359, "learning_rate": 2.8838288216719395e-05, "loss": 0.0312, "step": 39430 }, { "grad_norm": 0.5824520587921143, "learning_rate": 2.8813323478531484e-05, "loss": 0.0331, "step": 39440 }, { "grad_norm": 0.2038504183292389, "learning_rate": 2.8788365176296496e-05, "loss": 0.0322, "step": 39450 }, { "grad_norm": 0.47603926062583923, "learning_rate": 2.876341331759611e-05, "loss": 0.0355, "step": 39460 }, { "grad_norm": 0.28565362095832825, "learning_rate": 2.8738467910010036e-05, "loss": 0.0381, "step": 39470 }, { "grad_norm": 0.8344178795814514, "learning_rate": 2.8713528961116032e-05, "loss": 0.0272, "step": 39480 }, { "grad_norm": 0.24089571833610535, "learning_rate": 2.8688596478489875e-05, "loss": 0.0251, "step": 39490 }, { "grad_norm": 0.1961323767900467, "learning_rate": 2.8663670469705434e-05, "loss": 0.0283, "step": 39500 }, { "grad_norm": 0.4808233976364136, "learning_rate": 2.8638750942334546e-05, "loss": 0.0368, "step": 39510 }, { "grad_norm": 0.41645097732543945, "learning_rate": 2.8613837903947115e-05, "loss": 0.0413, "step": 39520 }, { "grad_norm": 0.822510302066803, "learning_rate": 2.858893136211106e-05, "loss": 0.0373, "step": 39530 }, { "grad_norm": 0.5985407829284668, "learning_rate": 2.8564031324392315e-05, "loss": 0.0357, "step": 39540 }, { "grad_norm": 0.5173429250717163, "learning_rate": 2.85391377983549e-05, "loss": 0.0447, "step": 39550 }, { "grad_norm": 0.5918527245521545, "learning_rate": 2.851425079156075e-05, "loss": 0.0392, "step": 39560 }, { "grad_norm": 0.7385585904121399, "learning_rate": 2.848937031156994e-05, "loss": 0.0534, "step": 39570 }, { "grad_norm": 0.3852526843547821, "learning_rate": 2.846449636594044e-05, "loss": 0.0372, "step": 39580 }, { "grad_norm": 0.4834969937801361, "learning_rate": 2.843962896222836e-05, "loss": 0.0309, "step": 39590 }, { "grad_norm": 0.3895624279975891, "learning_rate": 2.8414768107987722e-05, "loss": 0.0357, "step": 39600 }, { "grad_norm": 0.5051502585411072, "learning_rate": 2.838991381077061e-05, "loss": 0.043, "step": 39610 }, { "grad_norm": 0.5782694220542908, "learning_rate": 2.83650660781271e-05, "loss": 0.0279, "step": 39620 }, { "grad_norm": 0.2784908413887024, "learning_rate": 2.8340224917605285e-05, "loss": 0.0365, "step": 39630 }, { "grad_norm": 0.660693883895874, "learning_rate": 2.831539033675122e-05, "loss": 0.0302, "step": 39640 }, { "grad_norm": 0.4164259731769562, "learning_rate": 2.8290562343109038e-05, "loss": 0.0321, "step": 39650 }, { "grad_norm": 1.5536161661148071, "learning_rate": 2.826574094422082e-05, "loss": 0.0423, "step": 39660 }, { "grad_norm": 0.33081963658332825, "learning_rate": 2.8240926147626645e-05, "loss": 0.0312, "step": 39670 }, { "grad_norm": 0.3941747546195984, "learning_rate": 2.8216117960864586e-05, "loss": 0.0314, "step": 39680 }, { "grad_norm": 0.5756046772003174, "learning_rate": 2.8191316391470703e-05, "loss": 0.0344, "step": 39690 }, { "grad_norm": 0.8035621643066406, "learning_rate": 2.816652144697911e-05, "loss": 0.0578, "step": 39700 }, { "grad_norm": 0.5188665390014648, "learning_rate": 2.8141733134921783e-05, "loss": 0.0328, "step": 39710 }, { "grad_norm": 0.5653707981109619, "learning_rate": 2.811695146282884e-05, "loss": 0.0579, "step": 39720 }, { "grad_norm": 0.29278796911239624, "learning_rate": 2.8092176438228212e-05, "loss": 0.0283, "step": 39730 }, { "grad_norm": 0.58804851770401, "learning_rate": 2.806740806864598e-05, "loss": 0.0334, "step": 39740 }, { "grad_norm": 0.4142441153526306, "learning_rate": 2.804264636160604e-05, "loss": 0.0514, "step": 39750 }, { "grad_norm": 0.4865744113922119, "learning_rate": 2.8017891324630402e-05, "loss": 0.0357, "step": 39760 }, { "grad_norm": 0.6571547389030457, "learning_rate": 2.7993142965238976e-05, "loss": 0.0473, "step": 39770 }, { "grad_norm": 0.7525398135185242, "learning_rate": 2.7968401290949665e-05, "loss": 0.0338, "step": 39780 }, { "grad_norm": 0.3100613057613373, "learning_rate": 2.7943666309278328e-05, "loss": 0.0429, "step": 39790 }, { "grad_norm": 0.6443923711776733, "learning_rate": 2.7918938027738783e-05, "loss": 0.0664, "step": 39800 }, { "grad_norm": 0.5832898616790771, "learning_rate": 2.789421645384287e-05, "loss": 0.0373, "step": 39810 }, { "grad_norm": 0.6823177933692932, "learning_rate": 2.786950159510032e-05, "loss": 0.034, "step": 39820 }, { "grad_norm": 0.4811200201511383, "learning_rate": 2.7844793459018876e-05, "loss": 0.0378, "step": 39830 }, { "grad_norm": 0.344714492559433, "learning_rate": 2.7820092053104195e-05, "loss": 0.0322, "step": 39840 }, { "grad_norm": 0.5152150988578796, "learning_rate": 2.7795397384859933e-05, "loss": 0.036, "step": 39850 }, { "grad_norm": 0.45834723114967346, "learning_rate": 2.7770709461787638e-05, "loss": 0.0395, "step": 39860 }, { "grad_norm": 0.6595157384872437, "learning_rate": 2.7746028291386915e-05, "loss": 0.0256, "step": 39870 }, { "grad_norm": 0.5428246259689331, "learning_rate": 2.772135388115519e-05, "loss": 0.0358, "step": 39880 }, { "grad_norm": 0.370323121547699, "learning_rate": 2.7696686238587945e-05, "loss": 0.0398, "step": 39890 }, { "grad_norm": 0.606974720954895, "learning_rate": 2.7672025371178505e-05, "loss": 0.0346, "step": 39900 }, { "grad_norm": 0.4131876528263092, "learning_rate": 2.7647371286418238e-05, "loss": 0.0315, "step": 39910 }, { "grad_norm": 0.5374128818511963, "learning_rate": 2.762272399179639e-05, "loss": 0.0335, "step": 39920 }, { "grad_norm": 0.5124195218086243, "learning_rate": 2.7598083494800154e-05, "loss": 0.0309, "step": 39930 }, { "grad_norm": 0.5355081558227539, "learning_rate": 2.7573449802914664e-05, "loss": 0.031, "step": 39940 }, { "grad_norm": 0.40756145119667053, "learning_rate": 2.7548822923622964e-05, "loss": 0.0289, "step": 39950 }, { "grad_norm": 0.334888756275177, "learning_rate": 2.752420286440609e-05, "loss": 0.0367, "step": 39960 }, { "grad_norm": 0.3317798674106598, "learning_rate": 2.749958963274295e-05, "loss": 0.0282, "step": 39970 }, { "grad_norm": 0.23725828528404236, "learning_rate": 2.747498323611039e-05, "loss": 0.0263, "step": 39980 }, { "grad_norm": 0.34295010566711426, "learning_rate": 2.7450383681983184e-05, "loss": 0.0277, "step": 39990 }, { "grad_norm": 0.4157719612121582, "learning_rate": 2.742579097783403e-05, "loss": 0.0331, "step": 40000 }, { "grad_norm": 0.6753816604614258, "learning_rate": 2.7401205131133512e-05, "loss": 0.0337, "step": 40010 }, { "grad_norm": 0.36788854002952576, "learning_rate": 2.7376626149350238e-05, "loss": 0.031, "step": 40020 }, { "grad_norm": 0.5723915696144104, "learning_rate": 2.735205403995056e-05, "loss": 0.0314, "step": 40030 }, { "grad_norm": 0.6214800477027893, "learning_rate": 2.7327488810398917e-05, "loss": 0.0383, "step": 40040 }, { "grad_norm": 0.45090681314468384, "learning_rate": 2.7302930468157507e-05, "loss": 0.0398, "step": 40050 }, { "grad_norm": 0.4406585991382599, "learning_rate": 2.727837902068655e-05, "loss": 0.0341, "step": 40060 }, { "grad_norm": 0.28812047839164734, "learning_rate": 2.7253834475444123e-05, "loss": 0.0521, "step": 40070 }, { "grad_norm": 0.4660494327545166, "learning_rate": 2.7229296839886204e-05, "loss": 0.0335, "step": 40080 }, { "grad_norm": 0.5344631671905518, "learning_rate": 2.720476612146668e-05, "loss": 0.0294, "step": 40090 }, { "grad_norm": 0.4858464300632477, "learning_rate": 2.7180242327637317e-05, "loss": 0.0322, "step": 40100 }, { "grad_norm": 0.37158769369125366, "learning_rate": 2.7155725465847826e-05, "loss": 0.0269, "step": 40110 }, { "grad_norm": 0.33418846130371094, "learning_rate": 2.713121554354578e-05, "loss": 0.0302, "step": 40120 }, { "grad_norm": 0.36649054288864136, "learning_rate": 2.7106712568176628e-05, "loss": 0.0416, "step": 40130 }, { "grad_norm": 0.6007020473480225, "learning_rate": 2.708221654718374e-05, "loss": 0.0435, "step": 40140 }, { "grad_norm": 0.6609060168266296, "learning_rate": 2.7057727488008357e-05, "loss": 0.0359, "step": 40150 }, { "grad_norm": 0.3143988847732544, "learning_rate": 2.703324539808961e-05, "loss": 0.0281, "step": 40160 }, { "grad_norm": 0.474293977022171, "learning_rate": 2.7008770284864505e-05, "loss": 0.0403, "step": 40170 }, { "grad_norm": 0.32112517952919006, "learning_rate": 2.6984302155767916e-05, "loss": 0.0301, "step": 40180 }, { "grad_norm": 0.3248065412044525, "learning_rate": 2.6959841018232683e-05, "loss": 0.0245, "step": 40190 }, { "grad_norm": 0.3003247082233429, "learning_rate": 2.693538687968937e-05, "loss": 0.0318, "step": 40200 }, { "grad_norm": 0.45282647013664246, "learning_rate": 2.6910939747566556e-05, "loss": 0.032, "step": 40210 }, { "grad_norm": 0.5218245387077332, "learning_rate": 2.6886499629290607e-05, "loss": 0.0282, "step": 40220 }, { "grad_norm": 0.24531392753124237, "learning_rate": 2.6862066532285802e-05, "loss": 0.0253, "step": 40230 }, { "grad_norm": 0.7336336970329285, "learning_rate": 2.6837640463974262e-05, "loss": 0.0437, "step": 40240 }, { "grad_norm": 0.7865092158317566, "learning_rate": 2.681322143177596e-05, "loss": 0.0301, "step": 40250 }, { "grad_norm": 0.3275908827781677, "learning_rate": 2.678880944310882e-05, "loss": 0.0321, "step": 40260 }, { "grad_norm": 0.590843141078949, "learning_rate": 2.6764404505388474e-05, "loss": 0.0304, "step": 40270 }, { "grad_norm": 0.38291552662849426, "learning_rate": 2.6740006626028558e-05, "loss": 0.0343, "step": 40280 }, { "grad_norm": 0.6236094832420349, "learning_rate": 2.671561581244048e-05, "loss": 0.0433, "step": 40290 }, { "grad_norm": 0.25420835614204407, "learning_rate": 2.6691232072033536e-05, "loss": 0.0311, "step": 40300 }, { "grad_norm": 0.33720341324806213, "learning_rate": 2.6666855412214852e-05, "loss": 0.0336, "step": 40310 }, { "grad_norm": 0.44904306530952454, "learning_rate": 2.664248584038942e-05, "loss": 0.0377, "step": 40320 }, { "grad_norm": 0.4345734417438507, "learning_rate": 2.6618123363960047e-05, "loss": 0.0509, "step": 40330 }, { "grad_norm": 0.5332270264625549, "learning_rate": 2.659376799032748e-05, "loss": 0.031, "step": 40340 }, { "grad_norm": 0.3727763593196869, "learning_rate": 2.6569419726890145e-05, "loss": 0.0306, "step": 40350 }, { "grad_norm": 0.44891709089279175, "learning_rate": 2.654507858104447e-05, "loss": 0.0452, "step": 40360 }, { "grad_norm": 0.2955271005630493, "learning_rate": 2.652074456018463e-05, "loss": 0.0263, "step": 40370 }, { "grad_norm": 0.6146302223205566, "learning_rate": 2.6496417671702646e-05, "loss": 0.0252, "step": 40380 }, { "grad_norm": 0.55799800157547, "learning_rate": 2.6472097922988427e-05, "loss": 0.0348, "step": 40390 }, { "grad_norm": 0.32802119851112366, "learning_rate": 2.6447785321429607e-05, "loss": 0.0299, "step": 40400 }, { "grad_norm": 0.5946486592292786, "learning_rate": 2.6423479874411784e-05, "loss": 0.0317, "step": 40410 }, { "grad_norm": 0.5653437972068787, "learning_rate": 2.6399181589318234e-05, "loss": 0.0267, "step": 40420 }, { "grad_norm": 0.45059338212013245, "learning_rate": 2.6374890473530188e-05, "loss": 0.0391, "step": 40430 }, { "grad_norm": 0.34612128138542175, "learning_rate": 2.635060653442664e-05, "loss": 0.0357, "step": 40440 }, { "grad_norm": 0.516845703125, "learning_rate": 2.6326329779384395e-05, "loss": 0.0289, "step": 40450 }, { "grad_norm": 0.7040181756019592, "learning_rate": 2.63020602157781e-05, "loss": 0.0542, "step": 40460 }, { "grad_norm": 0.5565535426139832, "learning_rate": 2.62777978509802e-05, "loss": 0.0231, "step": 40470 }, { "grad_norm": 0.35090959072113037, "learning_rate": 2.6253542692360954e-05, "loss": 0.0338, "step": 40480 }, { "grad_norm": 0.33186593651771545, "learning_rate": 2.6229294747288458e-05, "loss": 0.0322, "step": 40490 }, { "grad_norm": 0.34502550959587097, "learning_rate": 2.6205054023128596e-05, "loss": 0.0312, "step": 40500 }, { "grad_norm": 0.41349801421165466, "learning_rate": 2.6180820527245043e-05, "loss": 0.03, "step": 40510 }, { "grad_norm": 0.2611384093761444, "learning_rate": 2.6156594266999313e-05, "loss": 0.0354, "step": 40520 }, { "grad_norm": 0.5049289464950562, "learning_rate": 2.6132375249750672e-05, "loss": 0.0301, "step": 40530 }, { "grad_norm": 0.48017674684524536, "learning_rate": 2.6108163482856286e-05, "loss": 0.0285, "step": 40540 }, { "grad_norm": 0.5470272302627563, "learning_rate": 2.6083958973670964e-05, "loss": 0.0278, "step": 40550 }, { "grad_norm": 0.4544198215007782, "learning_rate": 2.6059761729547483e-05, "loss": 0.0275, "step": 40560 }, { "grad_norm": 0.46393805742263794, "learning_rate": 2.603557175783624e-05, "loss": 0.0423, "step": 40570 }, { "grad_norm": 0.3483615219593048, "learning_rate": 2.601138906588559e-05, "loss": 0.0339, "step": 40580 }, { "grad_norm": 0.3821447491645813, "learning_rate": 2.598721366104152e-05, "loss": 0.0278, "step": 40590 }, { "grad_norm": 0.316663920879364, "learning_rate": 2.5963045550647945e-05, "loss": 0.0307, "step": 40600 }, { "grad_norm": 0.4867692291736603, "learning_rate": 2.5938884742046466e-05, "loss": 0.0375, "step": 40610 }, { "grad_norm": 0.5348402857780457, "learning_rate": 2.5914731242576507e-05, "loss": 0.0349, "step": 40620 }, { "grad_norm": 0.36297011375427246, "learning_rate": 2.5890585059575268e-05, "loss": 0.026, "step": 40630 }, { "grad_norm": 0.4123227298259735, "learning_rate": 2.5866446200377688e-05, "loss": 0.037, "step": 40640 }, { "grad_norm": 0.6136342883110046, "learning_rate": 2.5842314672316566e-05, "loss": 0.039, "step": 40650 }, { "grad_norm": 0.5361261963844299, "learning_rate": 2.581819048272239e-05, "loss": 0.0316, "step": 40660 }, { "grad_norm": 0.45978161692619324, "learning_rate": 2.5794073638923478e-05, "loss": 0.037, "step": 40670 }, { "grad_norm": 0.3270227909088135, "learning_rate": 2.576996414824586e-05, "loss": 0.0318, "step": 40680 }, { "grad_norm": 0.3248434066772461, "learning_rate": 2.574586201801339e-05, "loss": 0.0393, "step": 40690 }, { "grad_norm": 0.4508345127105713, "learning_rate": 2.572176725554762e-05, "loss": 0.032, "step": 40700 }, { "grad_norm": 0.4168180823326111, "learning_rate": 2.5697679868167966e-05, "loss": 0.0453, "step": 40710 }, { "grad_norm": 0.5330479741096497, "learning_rate": 2.5673599863191468e-05, "loss": 0.0231, "step": 40720 }, { "grad_norm": 0.44645121693611145, "learning_rate": 2.564952724793306e-05, "loss": 0.0357, "step": 40730 }, { "grad_norm": 0.7082310318946838, "learning_rate": 2.5625462029705306e-05, "loss": 0.0475, "step": 40740 }, { "grad_norm": 0.3990601599216461, "learning_rate": 2.5601404215818624e-05, "loss": 0.0276, "step": 40750 }, { "grad_norm": 0.45656055212020874, "learning_rate": 2.5577353813581144e-05, "loss": 0.0598, "step": 40760 }, { "grad_norm": 0.15181921422481537, "learning_rate": 2.5553310830298733e-05, "loss": 0.0305, "step": 40770 }, { "grad_norm": 0.3696686923503876, "learning_rate": 2.5529275273275012e-05, "loss": 0.0279, "step": 40780 }, { "grad_norm": 0.38237929344177246, "learning_rate": 2.550524714981133e-05, "loss": 0.029, "step": 40790 }, { "grad_norm": 0.4086657762527466, "learning_rate": 2.5481226467206837e-05, "loss": 0.0355, "step": 40800 }, { "grad_norm": 0.4550177752971649, "learning_rate": 2.5457213232758365e-05, "loss": 0.0262, "step": 40810 }, { "grad_norm": 0.4459066689014435, "learning_rate": 2.5433207453760498e-05, "loss": 0.0317, "step": 40820 }, { "grad_norm": 0.25317320227622986, "learning_rate": 2.5409209137505552e-05, "loss": 0.0316, "step": 40830 }, { "grad_norm": 0.4824562072753906, "learning_rate": 2.5385218291283597e-05, "loss": 0.0254, "step": 40840 }, { "grad_norm": 0.40943726897239685, "learning_rate": 2.5361234922382383e-05, "loss": 0.0315, "step": 40850 }, { "grad_norm": 0.26318636536598206, "learning_rate": 2.533725903808749e-05, "loss": 0.03, "step": 40860 }, { "grad_norm": 0.3698282539844513, "learning_rate": 2.5313290645682085e-05, "loss": 0.0381, "step": 40870 }, { "grad_norm": 0.4231104850769043, "learning_rate": 2.52893297524472e-05, "loss": 0.0373, "step": 40880 }, { "grad_norm": 0.6555312275886536, "learning_rate": 2.526537636566145e-05, "loss": 0.0427, "step": 40890 }, { "grad_norm": 0.32468369603157043, "learning_rate": 2.5241430492601305e-05, "loss": 0.0277, "step": 40900 }, { "grad_norm": 0.5849882960319519, "learning_rate": 2.5217492140540867e-05, "loss": 0.0624, "step": 40910 }, { "grad_norm": 0.2744714915752411, "learning_rate": 2.5193561316751967e-05, "loss": 0.043, "step": 40920 }, { "grad_norm": 0.6480282545089722, "learning_rate": 2.516963802850416e-05, "loss": 0.0281, "step": 40930 }, { "grad_norm": 0.3798324465751648, "learning_rate": 2.5145722283064698e-05, "loss": 0.0446, "step": 40940 }, { "grad_norm": 0.6081671714782715, "learning_rate": 2.5121814087698602e-05, "loss": 0.0336, "step": 40950 }, { "grad_norm": 0.4441062808036804, "learning_rate": 2.509791344966848e-05, "loss": 0.0354, "step": 40960 }, { "grad_norm": 0.3694341480731964, "learning_rate": 2.5074020376234768e-05, "loss": 0.0259, "step": 40970 }, { "grad_norm": 0.26401007175445557, "learning_rate": 2.5050134874655534e-05, "loss": 0.0241, "step": 40980 }, { "grad_norm": 0.4121363162994385, "learning_rate": 2.5026256952186566e-05, "loss": 0.029, "step": 40990 }, { "grad_norm": 0.5427940487861633, "learning_rate": 2.5002386616081335e-05, "loss": 0.0286, "step": 41000 }, { "grad_norm": 0.4902186691761017, "learning_rate": 2.497852387359103e-05, "loss": 0.0281, "step": 41010 }, { "grad_norm": 0.3973935842514038, "learning_rate": 2.4954668731964496e-05, "loss": 0.0364, "step": 41020 }, { "grad_norm": 0.4844845235347748, "learning_rate": 2.4930821198448364e-05, "loss": 0.0319, "step": 41030 }, { "grad_norm": 0.4274172782897949, "learning_rate": 2.4906981280286796e-05, "loss": 0.0258, "step": 41040 }, { "grad_norm": 0.3083692491054535, "learning_rate": 2.488314898472179e-05, "loss": 0.0333, "step": 41050 }, { "grad_norm": 0.2862100899219513, "learning_rate": 2.485932431899295e-05, "loss": 0.0352, "step": 41060 }, { "grad_norm": 0.1913919448852539, "learning_rate": 2.4835507290337584e-05, "loss": 0.0201, "step": 41070 }, { "grad_norm": 0.2654266059398651, "learning_rate": 2.4811697905990672e-05, "loss": 0.0231, "step": 41080 }, { "grad_norm": 0.5670341849327087, "learning_rate": 2.4787896173184854e-05, "loss": 0.0321, "step": 41090 }, { "grad_norm": 0.3353419601917267, "learning_rate": 2.4764102099150534e-05, "loss": 0.0284, "step": 41100 }, { "grad_norm": 0.5153481960296631, "learning_rate": 2.4740315691115644e-05, "loss": 0.0371, "step": 41110 }, { "grad_norm": 0.5056431889533997, "learning_rate": 2.4716536956305918e-05, "loss": 0.0317, "step": 41120 }, { "grad_norm": 0.4119923412799835, "learning_rate": 2.4692765901944697e-05, "loss": 0.0378, "step": 41130 }, { "grad_norm": 0.6088529229164124, "learning_rate": 2.4669002535253e-05, "loss": 0.0221, "step": 41140 }, { "grad_norm": 0.635035514831543, "learning_rate": 2.46452468634495e-05, "loss": 0.0371, "step": 41150 }, { "grad_norm": 0.46150320768356323, "learning_rate": 2.462149889375055e-05, "loss": 0.0167, "step": 41160 }, { "grad_norm": 0.6090131998062134, "learning_rate": 2.459775863337014e-05, "loss": 0.042, "step": 41170 }, { "grad_norm": 0.4692269563674927, "learning_rate": 2.4574026089519985e-05, "loss": 0.0288, "step": 41180 }, { "grad_norm": 0.4220476448535919, "learning_rate": 2.4550301269409333e-05, "loss": 0.0277, "step": 41190 }, { "grad_norm": 0.4675445854663849, "learning_rate": 2.4526584180245216e-05, "loss": 0.0378, "step": 41200 }, { "grad_norm": 0.41650712490081787, "learning_rate": 2.4502874829232236e-05, "loss": 0.028, "step": 41210 }, { "grad_norm": 0.3814327120780945, "learning_rate": 2.447917322357267e-05, "loss": 0.045, "step": 41220 }, { "grad_norm": 0.3784825801849365, "learning_rate": 2.4455479370466443e-05, "loss": 0.0333, "step": 41230 }, { "grad_norm": 0.6223308444023132, "learning_rate": 2.4431793277111097e-05, "loss": 0.0344, "step": 41240 }, { "grad_norm": 0.4984799325466156, "learning_rate": 2.4408114950701905e-05, "loss": 0.0392, "step": 41250 }, { "grad_norm": 0.36221134662628174, "learning_rate": 2.4384444398431634e-05, "loss": 0.0259, "step": 41260 }, { "grad_norm": 0.430918425321579, "learning_rate": 2.4360781627490837e-05, "loss": 0.0269, "step": 41270 }, { "grad_norm": 0.5086457133293152, "learning_rate": 2.433712664506762e-05, "loss": 0.0281, "step": 41280 }, { "grad_norm": 0.22712871432304382, "learning_rate": 2.431347945834774e-05, "loss": 0.0247, "step": 41290 }, { "grad_norm": 0.48943331837654114, "learning_rate": 2.428984007451458e-05, "loss": 0.0306, "step": 41300 }, { "grad_norm": 0.6624179482460022, "learning_rate": 2.426620850074917e-05, "loss": 0.0326, "step": 41310 }, { "grad_norm": 0.2738364040851593, "learning_rate": 2.424258474423014e-05, "loss": 0.0321, "step": 41320 }, { "grad_norm": 0.45142561197280884, "learning_rate": 2.421896881213382e-05, "loss": 0.0277, "step": 41330 }, { "grad_norm": 0.1914617121219635, "learning_rate": 2.419536071163402e-05, "loss": 0.0178, "step": 41340 }, { "grad_norm": 0.6006987690925598, "learning_rate": 2.417176044990233e-05, "loss": 0.0288, "step": 41350 }, { "grad_norm": 0.6585381627082825, "learning_rate": 2.4148168034107855e-05, "loss": 0.0308, "step": 41360 }, { "grad_norm": 0.36868447065353394, "learning_rate": 2.4124583471417355e-05, "loss": 0.0243, "step": 41370 }, { "grad_norm": 0.5440490245819092, "learning_rate": 2.41010067689952e-05, "loss": 0.0332, "step": 41380 }, { "grad_norm": 0.3460240066051483, "learning_rate": 2.4077437934003338e-05, "loss": 0.0284, "step": 41390 }, { "grad_norm": 1.4714927673339844, "learning_rate": 2.405387697360143e-05, "loss": 0.0396, "step": 41400 }, { "grad_norm": 0.6761341691017151, "learning_rate": 2.4030323894946595e-05, "loss": 0.0285, "step": 41410 }, { "grad_norm": 0.4491843581199646, "learning_rate": 2.40067787051937e-05, "loss": 0.0299, "step": 41420 }, { "grad_norm": 0.31816941499710083, "learning_rate": 2.3983241411495087e-05, "loss": 0.0305, "step": 41430 }, { "grad_norm": 0.4234989881515503, "learning_rate": 2.3959712021000823e-05, "loss": 0.0226, "step": 41440 }, { "grad_norm": 0.47303879261016846, "learning_rate": 2.3936190540858495e-05, "loss": 0.0238, "step": 41450 }, { "grad_norm": 0.4345490634441376, "learning_rate": 2.39126769782133e-05, "loss": 0.0495, "step": 41460 }, { "grad_norm": 0.31350094079971313, "learning_rate": 2.388917134020805e-05, "loss": 0.0258, "step": 41470 }, { "grad_norm": 0.2642328441143036, "learning_rate": 2.3865673633983128e-05, "loss": 0.03, "step": 41480 }, { "grad_norm": 0.43323925137519836, "learning_rate": 2.3842183866676492e-05, "loss": 0.0345, "step": 41490 }, { "grad_norm": 0.42029261589050293, "learning_rate": 2.381870204542377e-05, "loss": 0.0276, "step": 41500 }, { "grad_norm": 0.2725900113582611, "learning_rate": 2.379522817735808e-05, "loss": 0.0286, "step": 41510 }, { "grad_norm": 0.3380485773086548, "learning_rate": 2.377176226961018e-05, "loss": 0.0343, "step": 41520 }, { "grad_norm": 0.4490673243999481, "learning_rate": 2.3748304329308384e-05, "loss": 0.0413, "step": 41530 }, { "grad_norm": 0.5482943058013916, "learning_rate": 2.372485436357858e-05, "loss": 0.0417, "step": 41540 }, { "grad_norm": 0.29528212547302246, "learning_rate": 2.3701412379544296e-05, "loss": 0.0268, "step": 41550 }, { "grad_norm": 0.5359946489334106, "learning_rate": 2.367797838432653e-05, "loss": 0.0393, "step": 41560 }, { "grad_norm": 0.3741709291934967, "learning_rate": 2.3654552385043967e-05, "loss": 0.0277, "step": 41570 }, { "grad_norm": 0.5586929321289062, "learning_rate": 2.3631134388812742e-05, "loss": 0.0327, "step": 41580 }, { "grad_norm": 0.5538291335105896, "learning_rate": 2.3607724402746684e-05, "loss": 0.0321, "step": 41590 }, { "grad_norm": 0.2065950632095337, "learning_rate": 2.35843224339571e-05, "loss": 0.034, "step": 41600 }, { "grad_norm": 0.32850727438926697, "learning_rate": 2.3560928489552897e-05, "loss": 0.0226, "step": 41610 }, { "grad_norm": 0.39350625872612, "learning_rate": 2.353754257664053e-05, "loss": 0.0267, "step": 41620 }, { "grad_norm": 0.5191361904144287, "learning_rate": 2.3514164702324037e-05, "loss": 0.0353, "step": 41630 }, { "grad_norm": 0.2607240378856659, "learning_rate": 2.3490794873704963e-05, "loss": 0.0404, "step": 41640 }, { "grad_norm": 0.534485399723053, "learning_rate": 2.3467433097882496e-05, "loss": 0.0237, "step": 41650 }, { "grad_norm": 0.4243718087673187, "learning_rate": 2.34440793819533e-05, "loss": 0.0236, "step": 41660 }, { "grad_norm": 0.43499359488487244, "learning_rate": 2.3420733733011617e-05, "loss": 0.0306, "step": 41670 }, { "grad_norm": 0.45277684926986694, "learning_rate": 2.3397396158149243e-05, "loss": 0.0365, "step": 41680 }, { "grad_norm": 0.5170884728431702, "learning_rate": 2.3374066664455498e-05, "loss": 0.0413, "step": 41690 }, { "grad_norm": 0.3572273254394531, "learning_rate": 2.3350745259017315e-05, "loss": 0.0359, "step": 41700 }, { "grad_norm": 0.36591920256614685, "learning_rate": 2.332743194891906e-05, "loss": 0.0384, "step": 41710 }, { "grad_norm": 0.3374359607696533, "learning_rate": 2.330412674124276e-05, "loss": 0.0373, "step": 41720 }, { "grad_norm": 0.4022904634475708, "learning_rate": 2.328082964306786e-05, "loss": 0.0284, "step": 41730 }, { "grad_norm": 0.43808096647262573, "learning_rate": 2.325754066147145e-05, "loss": 0.0368, "step": 41740 }, { "grad_norm": 0.4862787425518036, "learning_rate": 2.32342598035281e-05, "loss": 0.035, "step": 41750 }, { "grad_norm": 0.36624282598495483, "learning_rate": 2.321098707630991e-05, "loss": 0.0271, "step": 41760 }, { "grad_norm": 0.5607618093490601, "learning_rate": 2.318772248688652e-05, "loss": 0.0408, "step": 41770 }, { "grad_norm": 0.3834022581577301, "learning_rate": 2.3164466042325107e-05, "loss": 0.0299, "step": 41780 }, { "grad_norm": 0.43804019689559937, "learning_rate": 2.3141217749690353e-05, "loss": 0.0294, "step": 41790 }, { "grad_norm": 0.5250070095062256, "learning_rate": 2.3117977616044466e-05, "loss": 0.0377, "step": 41800 }, { "grad_norm": 0.6545979976654053, "learning_rate": 2.309474564844722e-05, "loss": 0.0276, "step": 41810 }, { "grad_norm": 0.5754119753837585, "learning_rate": 2.307152185395585e-05, "loss": 0.0328, "step": 41820 }, { "grad_norm": 0.45681530237197876, "learning_rate": 2.3048306239625144e-05, "loss": 0.0311, "step": 41830 }, { "grad_norm": 0.42480501532554626, "learning_rate": 2.3025098812507378e-05, "loss": 0.0278, "step": 41840 }, { "grad_norm": 0.4658699333667755, "learning_rate": 2.3001899579652366e-05, "loss": 0.0261, "step": 41850 }, { "grad_norm": 0.37295836210250854, "learning_rate": 2.2978708548107393e-05, "loss": 0.0326, "step": 41860 }, { "grad_norm": 0.47790297865867615, "learning_rate": 2.2955525724917348e-05, "loss": 0.0348, "step": 41870 }, { "grad_norm": 0.4667282998561859, "learning_rate": 2.2932351117124477e-05, "loss": 0.0325, "step": 41880 }, { "grad_norm": 1.0030291080474854, "learning_rate": 2.29091847317687e-05, "loss": 0.0372, "step": 41890 }, { "grad_norm": 0.42533987760543823, "learning_rate": 2.2886026575887277e-05, "loss": 0.0341, "step": 41900 }, { "grad_norm": 0.5051378011703491, "learning_rate": 2.2862876656515094e-05, "loss": 0.0353, "step": 41910 }, { "grad_norm": 0.5131561756134033, "learning_rate": 2.2839734980684464e-05, "loss": 0.0345, "step": 41920 }, { "grad_norm": 0.5251510739326477, "learning_rate": 2.281660155542522e-05, "loss": 0.039, "step": 41930 }, { "grad_norm": 0.47162094712257385, "learning_rate": 2.279347638776469e-05, "loss": 0.0318, "step": 41940 }, { "grad_norm": 0.3592148423194885, "learning_rate": 2.2770359484727665e-05, "loss": 0.0254, "step": 41950 }, { "grad_norm": 0.4844387471675873, "learning_rate": 2.27472508533365e-05, "loss": 0.03, "step": 41960 }, { "grad_norm": 0.6508751511573792, "learning_rate": 2.2724150500610948e-05, "loss": 0.0392, "step": 41970 }, { "grad_norm": 0.5338071584701538, "learning_rate": 2.2701058433568302e-05, "loss": 0.0252, "step": 41980 }, { "grad_norm": 0.4675242602825165, "learning_rate": 2.2677974659223318e-05, "loss": 0.0334, "step": 41990 }, { "grad_norm": 0.3070293068885803, "learning_rate": 2.2654899184588235e-05, "loss": 0.0325, "step": 42000 }, { "grad_norm": 0.3776853084564209, "learning_rate": 2.2631832016672756e-05, "loss": 0.0413, "step": 42010 }, { "grad_norm": 0.47952690720558167, "learning_rate": 2.2608773162484127e-05, "loss": 0.0271, "step": 42020 }, { "grad_norm": 0.6470311880111694, "learning_rate": 2.2585722629026958e-05, "loss": 0.0283, "step": 42030 }, { "grad_norm": 0.3499886095523834, "learning_rate": 2.2562680423303457e-05, "loss": 0.0256, "step": 42040 }, { "grad_norm": 0.34697428345680237, "learning_rate": 2.2539646552313165e-05, "loss": 0.0307, "step": 42050 }, { "grad_norm": 0.500162661075592, "learning_rate": 2.251662102305322e-05, "loss": 0.0465, "step": 42060 }, { "grad_norm": 0.37330448627471924, "learning_rate": 2.2493603842518152e-05, "loss": 0.0375, "step": 42070 }, { "grad_norm": 0.48441770672798157, "learning_rate": 2.2470595017699974e-05, "loss": 0.0255, "step": 42080 }, { "grad_norm": 0.472691148519516, "learning_rate": 2.244759455558816e-05, "loss": 0.0323, "step": 42090 }, { "grad_norm": 0.3230453431606293, "learning_rate": 2.2424602463169614e-05, "loss": 0.0308, "step": 42100 }, { "grad_norm": 0.3836064040660858, "learning_rate": 2.2401618747428776e-05, "loss": 0.04, "step": 42110 }, { "grad_norm": 0.4281107783317566, "learning_rate": 2.237864341534747e-05, "loss": 0.0429, "step": 42120 }, { "grad_norm": 0.35864001512527466, "learning_rate": 2.2355676473904998e-05, "loss": 0.0282, "step": 42130 }, { "grad_norm": 0.34417158365249634, "learning_rate": 2.2332717930078108e-05, "loss": 0.0301, "step": 42140 }, { "grad_norm": 0.5795924663543701, "learning_rate": 2.2309767790840992e-05, "loss": 0.0289, "step": 42150 }, { "grad_norm": 0.2787536084651947, "learning_rate": 2.228682606316529e-05, "loss": 0.0261, "step": 42160 }, { "grad_norm": 0.3736748993396759, "learning_rate": 2.2263892754020138e-05, "loss": 0.0262, "step": 42170 }, { "grad_norm": 1.517612099647522, "learning_rate": 2.2240967870372004e-05, "loss": 0.0396, "step": 42180 }, { "grad_norm": 0.3304959237575531, "learning_rate": 2.2218051419184933e-05, "loss": 0.0352, "step": 42190 }, { "grad_norm": 0.38547006249427795, "learning_rate": 2.219514340742026e-05, "loss": 0.0344, "step": 42200 }, { "grad_norm": 0.6071175336837769, "learning_rate": 2.2172243842036898e-05, "loss": 0.0299, "step": 42210 }, { "grad_norm": 0.5414180159568787, "learning_rate": 2.2149352729991107e-05, "loss": 0.0292, "step": 42220 }, { "grad_norm": 0.7101613283157349, "learning_rate": 2.2126470078236605e-05, "loss": 0.0267, "step": 42230 }, { "grad_norm": 0.27313685417175293, "learning_rate": 2.2103595893724533e-05, "loss": 0.0306, "step": 42240 }, { "grad_norm": 0.2954351603984833, "learning_rate": 2.208073018340345e-05, "loss": 0.0352, "step": 42250 }, { "grad_norm": 0.34008342027664185, "learning_rate": 2.2057872954219405e-05, "loss": 0.032, "step": 42260 }, { "grad_norm": 0.23903392255306244, "learning_rate": 2.203502421311575e-05, "loss": 0.0258, "step": 42270 }, { "grad_norm": 0.5192579627037048, "learning_rate": 2.2012183967033388e-05, "loss": 0.0334, "step": 42280 }, { "grad_norm": 0.3782702088356018, "learning_rate": 2.198935222291056e-05, "loss": 0.0305, "step": 42290 }, { "grad_norm": 0.45544713735580444, "learning_rate": 2.1966528987682948e-05, "loss": 0.0293, "step": 42300 }, { "grad_norm": 0.6488160490989685, "learning_rate": 2.194371426828365e-05, "loss": 0.0378, "step": 42310 }, { "grad_norm": 0.3810856342315674, "learning_rate": 2.192090807164317e-05, "loss": 0.0265, "step": 42320 }, { "grad_norm": 0.449133962392807, "learning_rate": 2.1898110404689422e-05, "loss": 0.0402, "step": 42330 }, { "grad_norm": 0.49893125891685486, "learning_rate": 2.1875321274347776e-05, "loss": 0.0273, "step": 42340 }, { "grad_norm": 0.3581354320049286, "learning_rate": 2.18525406875409e-05, "loss": 0.0287, "step": 42350 }, { "grad_norm": 0.3872610032558441, "learning_rate": 2.1829768651188997e-05, "loss": 0.029, "step": 42360 }, { "grad_norm": 0.4160659909248352, "learning_rate": 2.180700517220958e-05, "loss": 0.0339, "step": 42370 }, { "grad_norm": 0.7114486694335938, "learning_rate": 2.1784250257517603e-05, "loss": 0.0345, "step": 42380 }, { "grad_norm": 0.4687311053276062, "learning_rate": 2.1761503914025406e-05, "loss": 0.0279, "step": 42390 }, { "grad_norm": 0.41904664039611816, "learning_rate": 2.1738766148642705e-05, "loss": 0.0263, "step": 42400 }, { "grad_norm": 0.34561264514923096, "learning_rate": 2.1716036968276683e-05, "loss": 0.0319, "step": 42410 }, { "grad_norm": 0.3574770987033844, "learning_rate": 2.1693316379831808e-05, "loss": 0.0269, "step": 42420 }, { "grad_norm": 0.40141037106513977, "learning_rate": 2.1670604390210037e-05, "loss": 0.0295, "step": 42430 }, { "grad_norm": 0.3712688088417053, "learning_rate": 2.1647901006310656e-05, "loss": 0.0288, "step": 42440 }, { "grad_norm": 0.3512985110282898, "learning_rate": 2.1625206235030353e-05, "loss": 0.035, "step": 42450 }, { "grad_norm": 0.4140358567237854, "learning_rate": 2.160252008326321e-05, "loss": 0.03, "step": 42460 }, { "grad_norm": 0.46427997946739197, "learning_rate": 2.157984255790067e-05, "loss": 0.0336, "step": 42470 }, { "grad_norm": 0.45302119851112366, "learning_rate": 2.1557173665831553e-05, "loss": 0.0272, "step": 42480 }, { "grad_norm": 0.5088571310043335, "learning_rate": 2.153451341394212e-05, "loss": 0.0285, "step": 42490 }, { "grad_norm": 0.4095351994037628, "learning_rate": 2.151186180911589e-05, "loss": 0.025, "step": 42500 }, { "grad_norm": 0.4361305236816406, "learning_rate": 2.1489218858233877e-05, "loss": 0.0381, "step": 42510 }, { "grad_norm": 0.3529554009437561, "learning_rate": 2.1466584568174392e-05, "loss": 0.0279, "step": 42520 }, { "grad_norm": 0.40075570344924927, "learning_rate": 2.1443958945813132e-05, "loss": 0.0413, "step": 42530 }, { "grad_norm": 0.25232768058776855, "learning_rate": 2.1421341998023163e-05, "loss": 0.03, "step": 42540 }, { "grad_norm": 0.2593213617801666, "learning_rate": 2.139873373167491e-05, "loss": 0.0276, "step": 42550 }, { "grad_norm": 0.2955894470214844, "learning_rate": 2.13761341536362e-05, "loss": 0.0306, "step": 42560 }, { "grad_norm": 0.4307028651237488, "learning_rate": 2.1353543270772136e-05, "loss": 0.0343, "step": 42570 }, { "grad_norm": 0.28702133893966675, "learning_rate": 2.1330961089945297e-05, "loss": 0.031, "step": 42580 }, { "grad_norm": 0.5277664661407471, "learning_rate": 2.130838761801548e-05, "loss": 0.0287, "step": 42590 }, { "grad_norm": 0.4585001766681671, "learning_rate": 2.1285822861839966e-05, "loss": 0.0328, "step": 42600 }, { "grad_norm": 0.4857662320137024, "learning_rate": 2.126326682827331e-05, "loss": 0.0246, "step": 42610 }, { "grad_norm": 0.5237287282943726, "learning_rate": 2.124071952416744e-05, "loss": 0.0333, "step": 42620 }, { "grad_norm": 0.20351402461528778, "learning_rate": 2.1218180956371634e-05, "loss": 0.024, "step": 42630 }, { "grad_norm": 0.19375935196876526, "learning_rate": 2.119565113173252e-05, "loss": 0.0321, "step": 42640 }, { "grad_norm": 0.4731431007385254, "learning_rate": 2.1173130057094033e-05, "loss": 0.0343, "step": 42650 }, { "grad_norm": 0.6174805164337158, "learning_rate": 2.115061773929753e-05, "loss": 0.0347, "step": 42660 }, { "grad_norm": 0.4970632791519165, "learning_rate": 2.1128114185181623e-05, "loss": 0.0326, "step": 42670 }, { "grad_norm": 0.38808998465538025, "learning_rate": 2.1105619401582317e-05, "loss": 0.0255, "step": 42680 }, { "grad_norm": 0.3003081679344177, "learning_rate": 2.1083133395332928e-05, "loss": 0.0352, "step": 42690 }, { "grad_norm": 0.3941520154476166, "learning_rate": 2.1060656173264082e-05, "loss": 0.0309, "step": 42700 }, { "grad_norm": 0.4210740327835083, "learning_rate": 2.103818774220383e-05, "loss": 0.0325, "step": 42710 }, { "grad_norm": 0.4830666780471802, "learning_rate": 2.1015728108977412e-05, "loss": 0.038, "step": 42720 }, { "grad_norm": 0.3823695778846741, "learning_rate": 2.0993277280407548e-05, "loss": 0.0324, "step": 42730 }, { "grad_norm": 0.314777135848999, "learning_rate": 2.0970835263314132e-05, "loss": 0.0207, "step": 42740 }, { "grad_norm": 0.27718591690063477, "learning_rate": 2.094840206451451e-05, "loss": 0.0298, "step": 42750 }, { "grad_norm": 0.4666004180908203, "learning_rate": 2.0925977690823273e-05, "loss": 0.0284, "step": 42760 }, { "grad_norm": 0.3933756947517395, "learning_rate": 2.0903562149052364e-05, "loss": 0.031, "step": 42770 }, { "grad_norm": 0.8081509470939636, "learning_rate": 2.0881155446011025e-05, "loss": 0.0378, "step": 42780 }, { "grad_norm": 0.4493977129459381, "learning_rate": 2.0858757588505823e-05, "loss": 0.0442, "step": 42790 }, { "grad_norm": 0.40763723850250244, "learning_rate": 2.0836368583340622e-05, "loss": 0.0429, "step": 42800 }, { "grad_norm": 0.5007255673408508, "learning_rate": 2.081398843731664e-05, "loss": 0.03, "step": 42810 }, { "grad_norm": 0.3251894414424896, "learning_rate": 2.0791617157232357e-05, "loss": 0.061, "step": 42820 }, { "grad_norm": 0.28802788257598877, "learning_rate": 2.0769254749883576e-05, "loss": 0.0346, "step": 42830 }, { "grad_norm": 0.2742372453212738, "learning_rate": 2.0746901222063415e-05, "loss": 0.0441, "step": 42840 }, { "grad_norm": 0.38805103302001953, "learning_rate": 2.072455658056226e-05, "loss": 0.0306, "step": 42850 }, { "grad_norm": 0.3910493850708008, "learning_rate": 2.0702220832167873e-05, "loss": 0.025, "step": 42860 }, { "grad_norm": 0.7356910109519958, "learning_rate": 2.0679893983665205e-05, "loss": 0.0322, "step": 42870 }, { "grad_norm": 0.5705873966217041, "learning_rate": 2.0657576041836622e-05, "loss": 0.0261, "step": 42880 }, { "grad_norm": 0.7983815670013428, "learning_rate": 2.0635267013461666e-05, "loss": 0.0414, "step": 42890 }, { "grad_norm": 0.36299219727516174, "learning_rate": 2.061296690531728e-05, "loss": 0.0512, "step": 42900 }, { "grad_norm": 0.7681049704551697, "learning_rate": 2.0590675724177622e-05, "loss": 0.0309, "step": 42910 }, { "grad_norm": 0.5487278699874878, "learning_rate": 2.0568393476814167e-05, "loss": 0.0286, "step": 42920 }, { "grad_norm": 0.5856730341911316, "learning_rate": 2.0546120169995685e-05, "loss": 0.0408, "step": 42930 }, { "grad_norm": 0.5033236145973206, "learning_rate": 2.0523855810488214e-05, "loss": 0.0376, "step": 42940 }, { "grad_norm": 0.22543621063232422, "learning_rate": 2.050160040505505e-05, "loss": 0.0238, "step": 42950 }, { "grad_norm": 0.23816676437854767, "learning_rate": 2.0479353960456843e-05, "loss": 0.0316, "step": 42960 }, { "grad_norm": 0.48588910698890686, "learning_rate": 2.0457116483451456e-05, "loss": 0.0264, "step": 42970 }, { "grad_norm": 0.44034847617149353, "learning_rate": 2.0434887980794043e-05, "loss": 0.0357, "step": 42980 }, { "grad_norm": 0.45739269256591797, "learning_rate": 2.0412668459237043e-05, "loss": 0.0215, "step": 42990 }, { "grad_norm": 0.2218288779258728, "learning_rate": 2.039045792553016e-05, "loss": 0.0355, "step": 43000 }, { "grad_norm": 0.40457093715667725, "learning_rate": 2.036825638642036e-05, "loss": 0.0429, "step": 43010 }, { "grad_norm": 0.48795241117477417, "learning_rate": 2.0346063848651868e-05, "loss": 0.0217, "step": 43020 }, { "grad_norm": 0.30527743697166443, "learning_rate": 2.0323880318966254e-05, "loss": 0.0233, "step": 43030 }, { "grad_norm": 0.23575179278850555, "learning_rate": 2.030170580410221e-05, "loss": 0.0286, "step": 43040 }, { "grad_norm": 0.5548266172409058, "learning_rate": 2.0279540310795837e-05, "loss": 0.0278, "step": 43050 }, { "grad_norm": 0.384634792804718, "learning_rate": 2.0257383845780365e-05, "loss": 0.0299, "step": 43060 }, { "grad_norm": 0.34258732199668884, "learning_rate": 2.0235236415786384e-05, "loss": 0.0198, "step": 43070 }, { "grad_norm": 0.6583092212677002, "learning_rate": 2.021309802754169e-05, "loss": 0.0304, "step": 43080 }, { "grad_norm": 0.5001470446586609, "learning_rate": 2.0190968687771332e-05, "loss": 0.0248, "step": 43090 }, { "grad_norm": 0.47910740971565247, "learning_rate": 2.016884840319763e-05, "loss": 0.0406, "step": 43100 }, { "grad_norm": 0.4061281383037567, "learning_rate": 2.0146737180540122e-05, "loss": 0.0239, "step": 43110 }, { "grad_norm": 0.7665784955024719, "learning_rate": 2.012463502651564e-05, "loss": 0.0303, "step": 43120 }, { "grad_norm": 0.6753829121589661, "learning_rate": 2.0102541947838228e-05, "loss": 0.0285, "step": 43130 }, { "grad_norm": 0.6652630567550659, "learning_rate": 2.0080457951219173e-05, "loss": 0.0445, "step": 43140 }, { "grad_norm": 0.4343574047088623, "learning_rate": 2.0058383043367017e-05, "loss": 0.026, "step": 43150 }, { "grad_norm": 0.7323869466781616, "learning_rate": 2.0036317230987528e-05, "loss": 0.0326, "step": 43160 }, { "grad_norm": 0.47264668345451355, "learning_rate": 2.0014260520783696e-05, "loss": 0.0365, "step": 43170 }, { "grad_norm": 0.5175274610519409, "learning_rate": 1.9992212919455834e-05, "loss": 0.0361, "step": 43180 }, { "grad_norm": 0.44098204374313354, "learning_rate": 1.9970174433701333e-05, "loss": 0.0284, "step": 43190 }, { "grad_norm": 0.25733476877212524, "learning_rate": 1.9948145070214992e-05, "loss": 0.0287, "step": 43200 }, { "grad_norm": 0.4899895191192627, "learning_rate": 1.9926124835688663e-05, "loss": 0.0277, "step": 43210 }, { "grad_norm": 0.5593078136444092, "learning_rate": 1.9904113736811576e-05, "loss": 0.0385, "step": 43220 }, { "grad_norm": 0.6690559983253479, "learning_rate": 1.9882111780270096e-05, "loss": 0.0279, "step": 43230 }, { "grad_norm": 0.27625909447669983, "learning_rate": 1.986011897274784e-05, "loss": 0.0247, "step": 43240 }, { "grad_norm": 0.46372878551483154, "learning_rate": 1.983813532092565e-05, "loss": 0.0364, "step": 43250 }, { "grad_norm": 0.33916759490966797, "learning_rate": 1.981616083148155e-05, "loss": 0.0251, "step": 43260 }, { "grad_norm": 0.3854941427707672, "learning_rate": 1.9794195511090845e-05, "loss": 0.0283, "step": 43270 }, { "grad_norm": 0.38625261187553406, "learning_rate": 1.977223936642601e-05, "loss": 0.0292, "step": 43280 }, { "grad_norm": 0.3114204406738281, "learning_rate": 1.975029240415674e-05, "loss": 0.0244, "step": 43290 }, { "grad_norm": 0.4419417083263397, "learning_rate": 1.9728354630949936e-05, "loss": 0.0242, "step": 43300 }, { "grad_norm": 0.36972519755363464, "learning_rate": 1.9706426053469716e-05, "loss": 0.0292, "step": 43310 }, { "grad_norm": 0.2223391979932785, "learning_rate": 1.9684506678377396e-05, "loss": 0.0236, "step": 43320 }, { "grad_norm": 0.38802334666252136, "learning_rate": 1.9662596512331544e-05, "loss": 0.0271, "step": 43330 }, { "grad_norm": 0.37018513679504395, "learning_rate": 1.964069556198782e-05, "loss": 0.0317, "step": 43340 }, { "grad_norm": 0.3503495454788208, "learning_rate": 1.9618803833999232e-05, "loss": 0.0262, "step": 43350 }, { "grad_norm": 0.39489126205444336, "learning_rate": 1.9596921335015838e-05, "loss": 0.0255, "step": 43360 }, { "grad_norm": 0.2669942378997803, "learning_rate": 1.957504807168501e-05, "loss": 0.0263, "step": 43370 }, { "grad_norm": 0.5093821883201599, "learning_rate": 1.9553184050651253e-05, "loss": 0.0331, "step": 43380 }, { "grad_norm": 0.31171131134033203, "learning_rate": 1.953132927855628e-05, "loss": 0.0345, "step": 43390 }, { "grad_norm": 0.8020413517951965, "learning_rate": 1.9509483762038995e-05, "loss": 0.0334, "step": 43400 }, { "grad_norm": 0.4051588773727417, "learning_rate": 1.9487647507735467e-05, "loss": 0.0248, "step": 43410 }, { "grad_norm": 0.40721818804740906, "learning_rate": 1.9465820522279032e-05, "loss": 0.026, "step": 43420 }, { "grad_norm": 0.47173359990119934, "learning_rate": 1.9444002812300078e-05, "loss": 0.0303, "step": 43430 }, { "grad_norm": 0.25707313418388367, "learning_rate": 1.94221943844263e-05, "loss": 0.0269, "step": 43440 }, { "grad_norm": 0.3687748610973358, "learning_rate": 1.9400395245282515e-05, "loss": 0.0283, "step": 43450 }, { "grad_norm": 0.6099533438682556, "learning_rate": 1.937860540149071e-05, "loss": 0.0269, "step": 43460 }, { "grad_norm": 0.5101149678230286, "learning_rate": 1.9356824859670082e-05, "loss": 0.025, "step": 43470 }, { "grad_norm": 0.42287716269493103, "learning_rate": 1.9335053626436967e-05, "loss": 0.0335, "step": 43480 }, { "grad_norm": 0.47315695881843567, "learning_rate": 1.9313291708404885e-05, "loss": 0.0296, "step": 43490 }, { "grad_norm": 0.46261319518089294, "learning_rate": 1.9291539112184587e-05, "loss": 0.0237, "step": 43500 }, { "grad_norm": 0.35037949681282043, "learning_rate": 1.9269795844383854e-05, "loss": 0.0255, "step": 43510 }, { "grad_norm": 0.7548481822013855, "learning_rate": 1.9248061911607777e-05, "loss": 0.0484, "step": 43520 }, { "grad_norm": 0.5287097692489624, "learning_rate": 1.9226337320458538e-05, "loss": 0.0398, "step": 43530 }, { "grad_norm": 0.3643955588340759, "learning_rate": 1.9204622077535488e-05, "loss": 0.0318, "step": 43540 }, { "grad_norm": 0.2539442181587219, "learning_rate": 1.9182916189435147e-05, "loss": 0.0251, "step": 43550 }, { "grad_norm": 0.602540135383606, "learning_rate": 1.916121966275117e-05, "loss": 0.0585, "step": 43560 }, { "grad_norm": 0.5726788640022278, "learning_rate": 1.9139532504074443e-05, "loss": 0.0272, "step": 43570 }, { "grad_norm": 0.44920238852500916, "learning_rate": 1.9117854719992885e-05, "loss": 0.0363, "step": 43580 }, { "grad_norm": 0.39201581478118896, "learning_rate": 1.9096186317091687e-05, "loss": 0.0298, "step": 43590 }, { "grad_norm": 0.4887314736843109, "learning_rate": 1.9074527301953116e-05, "loss": 0.0382, "step": 43600 }, { "grad_norm": 0.5298827886581421, "learning_rate": 1.9052877681156607e-05, "loss": 0.0264, "step": 43610 }, { "grad_norm": 0.33133843541145325, "learning_rate": 1.903123746127875e-05, "loss": 0.0298, "step": 43620 }, { "grad_norm": 0.2614663243293762, "learning_rate": 1.900960664889327e-05, "loss": 0.0229, "step": 43630 }, { "grad_norm": 0.44466036558151245, "learning_rate": 1.8987985250571015e-05, "loss": 0.0304, "step": 43640 }, { "grad_norm": 0.7808799743652344, "learning_rate": 1.8966373272880054e-05, "loss": 0.0304, "step": 43650 }, { "grad_norm": 0.6144794225692749, "learning_rate": 1.8944770722385462e-05, "loss": 0.0285, "step": 43660 }, { "grad_norm": 0.4580744802951813, "learning_rate": 1.8923177605649576e-05, "loss": 0.0353, "step": 43670 }, { "grad_norm": 0.28906041383743286, "learning_rate": 1.8901593929231802e-05, "loss": 0.0226, "step": 43680 }, { "grad_norm": 0.47165411710739136, "learning_rate": 1.8880019699688684e-05, "loss": 0.0314, "step": 43690 }, { "grad_norm": 0.4256569743156433, "learning_rate": 1.8858454923573904e-05, "loss": 0.0196, "step": 43700 }, { "grad_norm": 0.313406378030777, "learning_rate": 1.8836899607438253e-05, "loss": 0.0427, "step": 43710 }, { "grad_norm": 0.4438249468803406, "learning_rate": 1.8815353757829723e-05, "loss": 0.0219, "step": 43720 }, { "grad_norm": 0.24158628284931183, "learning_rate": 1.879381738129331e-05, "loss": 0.0272, "step": 43730 }, { "grad_norm": 0.3160351812839508, "learning_rate": 1.8772290484371236e-05, "loss": 0.0229, "step": 43740 }, { "grad_norm": 0.49367108941078186, "learning_rate": 1.8750773073602795e-05, "loss": 0.0217, "step": 43750 }, { "grad_norm": 0.722385823726654, "learning_rate": 1.8729265155524405e-05, "loss": 0.0253, "step": 43760 }, { "grad_norm": 0.20274117588996887, "learning_rate": 1.8707766736669607e-05, "loss": 0.0235, "step": 43770 }, { "grad_norm": 0.47201448678970337, "learning_rate": 1.8686277823569055e-05, "loss": 0.0337, "step": 43780 }, { "grad_norm": 0.3168877363204956, "learning_rate": 1.8664798422750484e-05, "loss": 0.0204, "step": 43790 }, { "grad_norm": 0.4119708836078644, "learning_rate": 1.8643328540738832e-05, "loss": 0.0345, "step": 43800 }, { "grad_norm": 0.5475504398345947, "learning_rate": 1.862186818405601e-05, "loss": 0.0345, "step": 43810 }, { "grad_norm": 0.41527777910232544, "learning_rate": 1.8600417359221156e-05, "loss": 0.0393, "step": 43820 }, { "grad_norm": 0.7753853797912598, "learning_rate": 1.8578976072750454e-05, "loss": 0.041, "step": 43830 }, { "grad_norm": 0.4524478614330292, "learning_rate": 1.8557544331157194e-05, "loss": 0.0303, "step": 43840 }, { "grad_norm": 0.4273028075695038, "learning_rate": 1.8536122140951785e-05, "loss": 0.0302, "step": 43850 }, { "grad_norm": 0.26376962661743164, "learning_rate": 1.8514709508641688e-05, "loss": 0.0293, "step": 43860 }, { "grad_norm": 0.568450927734375, "learning_rate": 1.8493306440731555e-05, "loss": 0.0291, "step": 43870 }, { "grad_norm": 0.5207457542419434, "learning_rate": 1.8471912943723013e-05, "loss": 0.0315, "step": 43880 }, { "grad_norm": 0.49222320318222046, "learning_rate": 1.8450529024114894e-05, "loss": 0.0188, "step": 43890 }, { "grad_norm": 0.38675349950790405, "learning_rate": 1.842915468840301e-05, "loss": 0.0245, "step": 43900 }, { "grad_norm": 0.3722575008869171, "learning_rate": 1.840778994308037e-05, "loss": 0.0292, "step": 43910 }, { "grad_norm": 0.46473515033721924, "learning_rate": 1.8386434794637004e-05, "loss": 0.0251, "step": 43920 }, { "grad_norm": 0.4467141628265381, "learning_rate": 1.8365089249560034e-05, "loss": 0.0376, "step": 43930 }, { "grad_norm": 0.4918926954269409, "learning_rate": 1.8343753314333683e-05, "loss": 0.0307, "step": 43940 }, { "grad_norm": 0.2819015383720398, "learning_rate": 1.8322426995439236e-05, "loss": 0.0181, "step": 43950 }, { "grad_norm": 0.2993682622909546, "learning_rate": 1.8301110299355058e-05, "loss": 0.0319, "step": 43960 }, { "grad_norm": 0.22255763411521912, "learning_rate": 1.8279803232556625e-05, "loss": 0.0207, "step": 43970 }, { "grad_norm": 0.4034425616264343, "learning_rate": 1.8258505801516444e-05, "loss": 0.0348, "step": 43980 }, { "grad_norm": 0.49523380398750305, "learning_rate": 1.8237218012704117e-05, "loss": 0.0325, "step": 43990 }, { "grad_norm": 0.37580224871635437, "learning_rate": 1.821593987258631e-05, "loss": 0.0243, "step": 44000 }, { "grad_norm": 0.6154166460037231, "learning_rate": 1.8194671387626744e-05, "loss": 0.0288, "step": 44010 }, { "grad_norm": 0.43531179428100586, "learning_rate": 1.8173412564286276e-05, "loss": 0.0237, "step": 44020 }, { "grad_norm": 0.5690121054649353, "learning_rate": 1.8152163409022697e-05, "loss": 0.0361, "step": 44030 }, { "grad_norm": 0.39584463834762573, "learning_rate": 1.8130923928291023e-05, "loss": 0.0304, "step": 44040 }, { "grad_norm": 0.6485760807991028, "learning_rate": 1.8109694128543163e-05, "loss": 0.0395, "step": 44050 }, { "grad_norm": 0.5527846217155457, "learning_rate": 1.8088474016228237e-05, "loss": 0.0217, "step": 44060 }, { "grad_norm": 0.32029539346694946, "learning_rate": 1.8067263597792328e-05, "loss": 0.0383, "step": 44070 }, { "grad_norm": 0.1701509654521942, "learning_rate": 1.80460628796786e-05, "loss": 0.0191, "step": 44080 }, { "grad_norm": 0.5514042973518372, "learning_rate": 1.8024871868327276e-05, "loss": 0.0334, "step": 44090 }, { "grad_norm": 0.21002301573753357, "learning_rate": 1.8003690570175608e-05, "loss": 0.0315, "step": 44100 }, { "grad_norm": 0.6211374402046204, "learning_rate": 1.7982518991657943e-05, "loss": 0.028, "step": 44110 }, { "grad_norm": 0.6284967064857483, "learning_rate": 1.7961357139205643e-05, "loss": 0.0302, "step": 44120 }, { "grad_norm": 0.5624222159385681, "learning_rate": 1.7940205019247108e-05, "loss": 0.0297, "step": 44130 }, { "grad_norm": 0.5102939605712891, "learning_rate": 1.79190626382078e-05, "loss": 0.0272, "step": 44140 }, { "grad_norm": 0.4420676827430725, "learning_rate": 1.7897930002510215e-05, "loss": 0.0219, "step": 44150 }, { "grad_norm": 0.3569242060184479, "learning_rate": 1.787680711857387e-05, "loss": 0.0203, "step": 44160 }, { "grad_norm": 0.5339128971099854, "learning_rate": 1.7855693992815398e-05, "loss": 0.0369, "step": 44170 }, { "grad_norm": 0.3618077039718628, "learning_rate": 1.7834590631648328e-05, "loss": 0.0446, "step": 44180 }, { "grad_norm": 0.3960779905319214, "learning_rate": 1.7813497041483384e-05, "loss": 0.0273, "step": 44190 }, { "grad_norm": 0.3679331839084625, "learning_rate": 1.779241322872817e-05, "loss": 0.0296, "step": 44200 }, { "grad_norm": 0.33742302656173706, "learning_rate": 1.777133919978744e-05, "loss": 0.0261, "step": 44210 }, { "grad_norm": 0.5854662656784058, "learning_rate": 1.7750274961062912e-05, "loss": 0.0464, "step": 44220 }, { "grad_norm": 0.41024863719940186, "learning_rate": 1.772922051895335e-05, "loss": 0.0467, "step": 44230 }, { "grad_norm": 0.6207701563835144, "learning_rate": 1.770817587985453e-05, "loss": 0.0343, "step": 44240 }, { "grad_norm": 1.081870436668396, "learning_rate": 1.7687141050159246e-05, "loss": 0.0488, "step": 44250 }, { "grad_norm": 0.3247911334037781, "learning_rate": 1.7666116036257375e-05, "loss": 0.0234, "step": 44260 }, { "grad_norm": 0.5191687345504761, "learning_rate": 1.764510084453569e-05, "loss": 0.0264, "step": 44270 }, { "grad_norm": 0.37516888976097107, "learning_rate": 1.76240954813781e-05, "loss": 0.0276, "step": 44280 }, { "grad_norm": 0.31884706020355225, "learning_rate": 1.7603099953165476e-05, "loss": 0.0229, "step": 44290 }, { "grad_norm": 0.6282243728637695, "learning_rate": 1.7582114266275683e-05, "loss": 0.036, "step": 44300 }, { "grad_norm": 0.6314430832862854, "learning_rate": 1.756113842708364e-05, "loss": 0.031, "step": 44310 }, { "grad_norm": 0.4791548252105713, "learning_rate": 1.7540172441961245e-05, "loss": 0.0254, "step": 44320 }, { "grad_norm": 0.29043814539909363, "learning_rate": 1.7519216317277387e-05, "loss": 0.0219, "step": 44330 }, { "grad_norm": 0.31375521421432495, "learning_rate": 1.7498270059398046e-05, "loss": 0.03, "step": 44340 }, { "grad_norm": 0.8539177179336548, "learning_rate": 1.7477333674686062e-05, "loss": 0.0265, "step": 44350 }, { "grad_norm": 0.5587623715400696, "learning_rate": 1.745640716950142e-05, "loss": 0.0264, "step": 44360 }, { "grad_norm": 0.3341250717639923, "learning_rate": 1.7435490550201017e-05, "loss": 0.0201, "step": 44370 }, { "grad_norm": 0.41702163219451904, "learning_rate": 1.7414583823138762e-05, "loss": 0.0366, "step": 44380 }, { "grad_norm": 0.3734339773654938, "learning_rate": 1.739368699466558e-05, "loss": 0.0283, "step": 44390 }, { "grad_norm": 0.25246456265449524, "learning_rate": 1.737280007112935e-05, "loss": 0.0329, "step": 44400 }, { "grad_norm": 1.3164572715759277, "learning_rate": 1.735192305887502e-05, "loss": 0.0374, "step": 44410 }, { "grad_norm": 0.5067208409309387, "learning_rate": 1.733105596424441e-05, "loss": 0.0315, "step": 44420 }, { "grad_norm": 0.5510420799255371, "learning_rate": 1.7310198793576437e-05, "loss": 0.0311, "step": 44430 }, { "grad_norm": 0.42779022455215454, "learning_rate": 1.7289351553206952e-05, "loss": 0.0251, "step": 44440 }, { "grad_norm": 0.17471542954444885, "learning_rate": 1.7268514249468788e-05, "loss": 0.0192, "step": 44450 }, { "grad_norm": 0.2797345817089081, "learning_rate": 1.7247686888691765e-05, "loss": 0.0235, "step": 44460 }, { "grad_norm": 0.3228563070297241, "learning_rate": 1.7226869477202694e-05, "loss": 0.0346, "step": 44470 }, { "grad_norm": 0.42577919363975525, "learning_rate": 1.7206062021325336e-05, "loss": 0.0313, "step": 44480 }, { "grad_norm": 0.30547332763671875, "learning_rate": 1.7185264527380502e-05, "loss": 0.0351, "step": 44490 }, { "grad_norm": 0.6817706227302551, "learning_rate": 1.716447700168584e-05, "loss": 0.03, "step": 44500 }, { "grad_norm": 0.23904238641262054, "learning_rate": 1.714369945055611e-05, "loss": 0.0231, "step": 44510 }, { "grad_norm": 0.6139406561851501, "learning_rate": 1.7122931880302968e-05, "loss": 0.0431, "step": 44520 }, { "grad_norm": 0.6592464447021484, "learning_rate": 1.710217429723505e-05, "loss": 0.03, "step": 44530 }, { "grad_norm": 0.4694228768348694, "learning_rate": 1.7081426707657972e-05, "loss": 0.029, "step": 44540 }, { "grad_norm": 0.430245041847229, "learning_rate": 1.7060689117874275e-05, "loss": 0.0293, "step": 44550 }, { "grad_norm": 0.45842671394348145, "learning_rate": 1.703996153418354e-05, "loss": 0.0356, "step": 44560 }, { "grad_norm": 0.508072555065155, "learning_rate": 1.7019243962882205e-05, "loss": 0.0304, "step": 44570 }, { "grad_norm": 0.558794379234314, "learning_rate": 1.6998536410263754e-05, "loss": 0.0264, "step": 44580 }, { "grad_norm": 0.2933914065361023, "learning_rate": 1.6977838882618596e-05, "loss": 0.0308, "step": 44590 }, { "grad_norm": 0.48243236541748047, "learning_rate": 1.6957151386234088e-05, "loss": 0.0275, "step": 44600 }, { "grad_norm": 0.4959612488746643, "learning_rate": 1.6936473927394536e-05, "loss": 0.0281, "step": 44610 }, { "grad_norm": 0.30000820755958557, "learning_rate": 1.6915806512381222e-05, "loss": 0.0236, "step": 44620 }, { "grad_norm": 0.42512279748916626, "learning_rate": 1.6895149147472344e-05, "loss": 0.0347, "step": 44630 }, { "grad_norm": 0.309927761554718, "learning_rate": 1.6874501838943073e-05, "loss": 0.0331, "step": 44640 }, { "grad_norm": 0.39877545833587646, "learning_rate": 1.6853864593065506e-05, "loss": 0.0316, "step": 44650 }, { "grad_norm": 0.423211008310318, "learning_rate": 1.683323741610871e-05, "loss": 0.0277, "step": 44660 }, { "grad_norm": 0.2973417043685913, "learning_rate": 1.6812620314338674e-05, "loss": 0.0422, "step": 44670 }, { "grad_norm": 0.4226797819137573, "learning_rate": 1.6792013294018326e-05, "loss": 0.0269, "step": 44680 }, { "grad_norm": 0.7099420428276062, "learning_rate": 1.6771416361407526e-05, "loss": 0.0373, "step": 44690 }, { "grad_norm": 0.5893145799636841, "learning_rate": 1.675082952276308e-05, "loss": 0.0365, "step": 44700 }, { "grad_norm": 0.4953434467315674, "learning_rate": 1.6730252784338757e-05, "loss": 0.0262, "step": 44710 }, { "grad_norm": 0.3463135063648224, "learning_rate": 1.6709686152385166e-05, "loss": 0.0269, "step": 44720 }, { "grad_norm": 0.2397233247756958, "learning_rate": 1.668912963314998e-05, "loss": 0.0286, "step": 44730 }, { "grad_norm": 0.25203871726989746, "learning_rate": 1.6668583232877653e-05, "loss": 0.05, "step": 44740 }, { "grad_norm": 0.3179108798503876, "learning_rate": 1.6648046957809698e-05, "loss": 0.0223, "step": 44750 }, { "grad_norm": 0.3989589810371399, "learning_rate": 1.6627520814184462e-05, "loss": 0.0277, "step": 44760 }, { "grad_norm": 0.5731974244117737, "learning_rate": 1.660700480823726e-05, "loss": 0.0273, "step": 44770 }, { "grad_norm": 0.45078811049461365, "learning_rate": 1.65864989462003e-05, "loss": 0.0194, "step": 44780 }, { "grad_norm": 0.42916885018348694, "learning_rate": 1.656600323430273e-05, "loss": 0.024, "step": 44790 }, { "grad_norm": 0.4880368113517761, "learning_rate": 1.654551767877059e-05, "loss": 0.0268, "step": 44800 }, { "grad_norm": 0.8778945207595825, "learning_rate": 1.6525042285826874e-05, "loss": 0.0408, "step": 44810 }, { "grad_norm": 0.39231738448143005, "learning_rate": 1.6504577061691468e-05, "loss": 0.032, "step": 44820 }, { "grad_norm": 0.4435780644416809, "learning_rate": 1.6484122012581143e-05, "loss": 0.0357, "step": 44830 }, { "grad_norm": 0.42098113894462585, "learning_rate": 1.6463677144709623e-05, "loss": 0.0287, "step": 44840 }, { "grad_norm": 0.5213115215301514, "learning_rate": 1.6443242464287493e-05, "loss": 0.0288, "step": 44850 }, { "grad_norm": 0.4625130295753479, "learning_rate": 1.642281797752232e-05, "loss": 0.0274, "step": 44860 }, { "grad_norm": 0.3803657591342926, "learning_rate": 1.6402403690618456e-05, "loss": 0.0281, "step": 44870 }, { "grad_norm": 0.32505476474761963, "learning_rate": 1.6381999609777295e-05, "loss": 0.0213, "step": 44880 }, { "grad_norm": 0.6972144246101379, "learning_rate": 1.6361605741196983e-05, "loss": 0.0351, "step": 44890 }, { "grad_norm": 0.40997254848480225, "learning_rate": 1.63412220910727e-05, "loss": 0.0264, "step": 44900 }, { "grad_norm": 0.5206490159034729, "learning_rate": 1.6320848665596433e-05, "loss": 0.0288, "step": 44910 }, { "grad_norm": 0.504517674446106, "learning_rate": 1.6300485470957095e-05, "loss": 0.0303, "step": 44920 }, { "grad_norm": 0.6883845329284668, "learning_rate": 1.6280132513340483e-05, "loss": 0.0271, "step": 44930 }, { "grad_norm": 0.5400803685188293, "learning_rate": 1.62597897989293e-05, "loss": 0.0289, "step": 44940 }, { "grad_norm": 0.32713502645492554, "learning_rate": 1.623945733390309e-05, "loss": 0.0263, "step": 44950 }, { "grad_norm": 0.4863600432872772, "learning_rate": 1.6219135124438374e-05, "loss": 0.0306, "step": 44960 }, { "grad_norm": 0.2418343722820282, "learning_rate": 1.6198823176708465e-05, "loss": 0.0312, "step": 44970 }, { "grad_norm": 0.6584184169769287, "learning_rate": 1.6178521496883613e-05, "loss": 0.0284, "step": 44980 }, { "grad_norm": 0.4607936441898346, "learning_rate": 1.6158230091130926e-05, "loss": 0.0289, "step": 44990 }, { "grad_norm": 0.46406397223472595, "learning_rate": 1.613794896561438e-05, "loss": 0.0381, "step": 45000 }, { "grad_norm": 0.36672985553741455, "learning_rate": 1.6117678126494894e-05, "loss": 0.0322, "step": 45010 }, { "grad_norm": 0.3486793041229248, "learning_rate": 1.6097417579930153e-05, "loss": 0.0187, "step": 45020 }, { "grad_norm": 0.2528093755245209, "learning_rate": 1.6077167332074834e-05, "loss": 0.0275, "step": 45030 }, { "grad_norm": 0.42279890179634094, "learning_rate": 1.605692738908037e-05, "loss": 0.0479, "step": 45040 }, { "grad_norm": 0.5154488682746887, "learning_rate": 1.6036697757095176e-05, "loss": 0.0449, "step": 45050 }, { "grad_norm": 0.361527681350708, "learning_rate": 1.6016478442264428e-05, "loss": 0.0269, "step": 45060 }, { "grad_norm": 0.22593113780021667, "learning_rate": 1.599626945073026e-05, "loss": 0.0186, "step": 45070 }, { "grad_norm": 0.3791545629501343, "learning_rate": 1.597607078863162e-05, "loss": 0.0273, "step": 45080 }, { "grad_norm": 0.5040034651756287, "learning_rate": 1.595588246210432e-05, "loss": 0.041, "step": 45090 }, { "grad_norm": 0.7036536931991577, "learning_rate": 1.5935704477281048e-05, "loss": 0.0332, "step": 45100 }, { "grad_norm": 0.4878656566143036, "learning_rate": 1.5915536840291323e-05, "loss": 0.0344, "step": 45110 }, { "grad_norm": 0.6917540431022644, "learning_rate": 1.5895379557261576e-05, "loss": 0.0423, "step": 45120 }, { "grad_norm": 0.2579532563686371, "learning_rate": 1.5875232634315033e-05, "loss": 0.0234, "step": 45130 }, { "grad_norm": 0.38234561681747437, "learning_rate": 1.5855096077571812e-05, "loss": 0.024, "step": 45140 }, { "grad_norm": 0.19119587540626526, "learning_rate": 1.5834969893148855e-05, "loss": 0.0197, "step": 45150 }, { "grad_norm": 0.41034623980522156, "learning_rate": 1.581485408715997e-05, "loss": 0.0316, "step": 45160 }, { "grad_norm": 0.5193532705307007, "learning_rate": 1.5794748665715785e-05, "loss": 0.0239, "step": 45170 }, { "grad_norm": 0.259685754776001, "learning_rate": 1.5774653634923857e-05, "loss": 0.021, "step": 45180 }, { "grad_norm": 0.46618905663490295, "learning_rate": 1.575456900088845e-05, "loss": 0.0325, "step": 45190 }, { "grad_norm": 0.5543297529220581, "learning_rate": 1.5734494769710816e-05, "loss": 0.0236, "step": 45200 }, { "grad_norm": 0.5546193718910217, "learning_rate": 1.5714430947488912e-05, "loss": 0.0237, "step": 45210 }, { "grad_norm": 0.3220064043998718, "learning_rate": 1.5694377540317645e-05, "loss": 0.0288, "step": 45220 }, { "grad_norm": 0.47835224866867065, "learning_rate": 1.5674334554288694e-05, "loss": 0.0276, "step": 45230 }, { "grad_norm": 0.25898799300193787, "learning_rate": 1.5654301995490582e-05, "loss": 0.0232, "step": 45240 }, { "grad_norm": 0.27787986397743225, "learning_rate": 1.5634279870008685e-05, "loss": 0.0237, "step": 45250 }, { "grad_norm": 0.2914179265499115, "learning_rate": 1.5614268183925174e-05, "loss": 0.0324, "step": 45260 }, { "grad_norm": 1.1144505739212036, "learning_rate": 1.5594266943319097e-05, "loss": 0.0421, "step": 45270 }, { "grad_norm": 0.4229932725429535, "learning_rate": 1.5574276154266294e-05, "loss": 0.0279, "step": 45280 }, { "grad_norm": 0.3288373351097107, "learning_rate": 1.5554295822839437e-05, "loss": 0.0205, "step": 45290 }, { "grad_norm": 0.573180615901947, "learning_rate": 1.5534325955108025e-05, "loss": 0.0305, "step": 45300 }, { "grad_norm": 0.29670190811157227, "learning_rate": 1.5514366557138373e-05, "loss": 0.0227, "step": 45310 }, { "grad_norm": 0.5037210583686829, "learning_rate": 1.5494417634993602e-05, "loss": 0.0228, "step": 45320 }, { "grad_norm": 0.46284234523773193, "learning_rate": 1.547447919473372e-05, "loss": 0.0175, "step": 45330 }, { "grad_norm": 0.47124287486076355, "learning_rate": 1.5454551242415434e-05, "loss": 0.0329, "step": 45340 }, { "grad_norm": 0.5531671643257141, "learning_rate": 1.543463378409239e-05, "loss": 0.0321, "step": 45350 }, { "grad_norm": 0.5203455686569214, "learning_rate": 1.541472682581493e-05, "loss": 0.0356, "step": 45360 }, { "grad_norm": 0.442242294549942, "learning_rate": 1.5394830373630298e-05, "loss": 0.0308, "step": 45370 }, { "grad_norm": 0.6268431544303894, "learning_rate": 1.5374944433582506e-05, "loss": 0.0212, "step": 45380 }, { "grad_norm": 0.5281916260719299, "learning_rate": 1.5355069011712375e-05, "loss": 0.0313, "step": 45390 }, { "grad_norm": 0.4202180504798889, "learning_rate": 1.5335204114057526e-05, "loss": 0.0286, "step": 45400 }, { "grad_norm": 0.512858510017395, "learning_rate": 1.5315349746652387e-05, "loss": 0.026, "step": 45410 }, { "grad_norm": 0.5511190891265869, "learning_rate": 1.5295505915528212e-05, "loss": 0.0485, "step": 45420 }, { "grad_norm": 0.37128493189811707, "learning_rate": 1.5275672626713024e-05, "loss": 0.0244, "step": 45430 }, { "grad_norm": 0.4582222104072571, "learning_rate": 1.5255849886231643e-05, "loss": 0.0249, "step": 45440 }, { "grad_norm": 0.36070066690444946, "learning_rate": 1.523603770010571e-05, "loss": 0.0242, "step": 45450 }, { "grad_norm": 0.5225266218185425, "learning_rate": 1.521623607435363e-05, "loss": 0.0279, "step": 45460 }, { "grad_norm": 0.48468491435050964, "learning_rate": 1.5196445014990612e-05, "loss": 0.0246, "step": 45470 }, { "grad_norm": 0.3907381296157837, "learning_rate": 1.5176664528028672e-05, "loss": 0.0327, "step": 45480 }, { "grad_norm": 0.5258684158325195, "learning_rate": 1.5156894619476574e-05, "loss": 0.0322, "step": 45490 }, { "grad_norm": 0.3842531442642212, "learning_rate": 1.5137135295339938e-05, "loss": 0.0329, "step": 45500 }, { "grad_norm": 0.42263925075531006, "learning_rate": 1.5117386561621073e-05, "loss": 0.0328, "step": 45510 }, { "grad_norm": 0.3458612263202667, "learning_rate": 1.5097648424319167e-05, "loss": 0.0368, "step": 45520 }, { "grad_norm": 0.4522940516471863, "learning_rate": 1.5077920889430119e-05, "loss": 0.0236, "step": 45530 }, { "grad_norm": 0.5239602327346802, "learning_rate": 1.5058203962946644e-05, "loss": 0.0314, "step": 45540 }, { "grad_norm": 0.4348854124546051, "learning_rate": 1.503849765085822e-05, "loss": 0.0297, "step": 45550 }, { "grad_norm": 0.38158807158470154, "learning_rate": 1.501880195915109e-05, "loss": 0.0306, "step": 45560 }, { "grad_norm": 0.24762491881847382, "learning_rate": 1.499911689380833e-05, "loss": 0.0203, "step": 45570 }, { "grad_norm": 0.3674316108226776, "learning_rate": 1.4979442460809683e-05, "loss": 0.0279, "step": 45580 }, { "grad_norm": 0.3135519027709961, "learning_rate": 1.4959778666131763e-05, "loss": 0.021, "step": 45590 }, { "grad_norm": 0.27973735332489014, "learning_rate": 1.4940125515747905e-05, "loss": 0.0177, "step": 45600 }, { "grad_norm": 0.26777973771095276, "learning_rate": 1.4920483015628211e-05, "loss": 0.0252, "step": 45610 }, { "grad_norm": 0.41231274604797363, "learning_rate": 1.490085117173956e-05, "loss": 0.0327, "step": 45620 }, { "grad_norm": 0.6516594290733337, "learning_rate": 1.488122999004558e-05, "loss": 0.0335, "step": 45630 }, { "grad_norm": 0.25686168670654297, "learning_rate": 1.486161947650666e-05, "loss": 0.0301, "step": 45640 }, { "grad_norm": 0.33323898911476135, "learning_rate": 1.4842019637079995e-05, "loss": 0.0203, "step": 45650 }, { "grad_norm": 0.38054388761520386, "learning_rate": 1.482243047771944e-05, "loss": 0.0284, "step": 45660 }, { "grad_norm": 0.21020649373531342, "learning_rate": 1.4802852004375712e-05, "loss": 0.0232, "step": 45670 }, { "grad_norm": 0.35861682891845703, "learning_rate": 1.4783284222996218e-05, "loss": 0.0233, "step": 45680 }, { "grad_norm": 0.34548690915107727, "learning_rate": 1.4763727139525135e-05, "loss": 0.0462, "step": 45690 }, { "grad_norm": 0.5508509874343872, "learning_rate": 1.4744180759903392e-05, "loss": 0.0386, "step": 45700 }, { "grad_norm": 0.23326654732227325, "learning_rate": 1.4724645090068635e-05, "loss": 0.0269, "step": 45710 }, { "grad_norm": 0.3453269898891449, "learning_rate": 1.4705120135955341e-05, "loss": 0.0206, "step": 45720 }, { "grad_norm": 0.404219388961792, "learning_rate": 1.4685605903494614e-05, "loss": 0.0261, "step": 45730 }, { "grad_norm": 0.231295645236969, "learning_rate": 1.46661023986144e-05, "loss": 0.0307, "step": 45740 }, { "grad_norm": 0.37527996301651, "learning_rate": 1.4646609627239344e-05, "loss": 0.026, "step": 45750 }, { "grad_norm": 0.47955945134162903, "learning_rate": 1.4627127595290835e-05, "loss": 0.032, "step": 45760 }, { "grad_norm": 0.45889514684677124, "learning_rate": 1.460765630868699e-05, "loss": 0.027, "step": 45770 }, { "grad_norm": 0.3020695149898529, "learning_rate": 1.4588195773342678e-05, "loss": 0.0269, "step": 45780 }, { "grad_norm": 0.287733793258667, "learning_rate": 1.4568745995169485e-05, "loss": 0.0384, "step": 45790 }, { "grad_norm": 0.6733306050300598, "learning_rate": 1.4549306980075778e-05, "loss": 0.0405, "step": 45800 }, { "grad_norm": 0.7050702571868896, "learning_rate": 1.4529878733966557e-05, "loss": 0.0301, "step": 45810 }, { "grad_norm": 0.34896838665008545, "learning_rate": 1.4510461262743658e-05, "loss": 0.0248, "step": 45820 }, { "grad_norm": 0.4016813039779663, "learning_rate": 1.4491054572305585e-05, "loss": 0.0303, "step": 45830 }, { "grad_norm": 0.5119677782058716, "learning_rate": 1.4471658668547566e-05, "loss": 0.0235, "step": 45840 }, { "grad_norm": 0.39475613832473755, "learning_rate": 1.4452273557361579e-05, "loss": 0.0264, "step": 45850 }, { "grad_norm": 0.2716691493988037, "learning_rate": 1.4432899244636282e-05, "loss": 0.0234, "step": 45860 }, { "grad_norm": 0.2244853973388672, "learning_rate": 1.4413535736257134e-05, "loss": 0.0252, "step": 45870 }, { "grad_norm": 0.5197407007217407, "learning_rate": 1.439418303810619e-05, "loss": 0.0258, "step": 45880 }, { "grad_norm": 0.30174753069877625, "learning_rate": 1.4374841156062352e-05, "loss": 0.0216, "step": 45890 }, { "grad_norm": 0.22975896298885345, "learning_rate": 1.4355510096001112e-05, "loss": 0.0501, "step": 45900 }, { "grad_norm": 0.3994704782962799, "learning_rate": 1.4336189863794786e-05, "loss": 0.0243, "step": 45910 }, { "grad_norm": 0.45735105872154236, "learning_rate": 1.4316880465312327e-05, "loss": 0.0252, "step": 45920 }, { "grad_norm": 0.4684562683105469, "learning_rate": 1.4297581906419426e-05, "loss": 0.0232, "step": 45930 }, { "grad_norm": 0.42440861463546753, "learning_rate": 1.4278294192978475e-05, "loss": 0.0246, "step": 45940 }, { "grad_norm": 0.3221494257450104, "learning_rate": 1.4259017330848574e-05, "loss": 0.029, "step": 45950 }, { "grad_norm": 0.2769522964954376, "learning_rate": 1.4239751325885498e-05, "loss": 0.0337, "step": 45960 }, { "grad_norm": 0.6264215111732483, "learning_rate": 1.4220496183941795e-05, "loss": 0.0278, "step": 45970 }, { "grad_norm": 0.32736095786094666, "learning_rate": 1.4201251910866648e-05, "loss": 0.024, "step": 45980 }, { "grad_norm": 0.29677337408065796, "learning_rate": 1.4182018512505957e-05, "loss": 0.0251, "step": 45990 }, { "grad_norm": 0.6740269064903259, "learning_rate": 1.4162795994702327e-05, "loss": 0.0277, "step": 46000 }, { "grad_norm": 0.2114415317773819, "learning_rate": 1.4143584363295032e-05, "loss": 0.0303, "step": 46010 }, { "grad_norm": 0.27816611528396606, "learning_rate": 1.4124383624120101e-05, "loss": 0.0265, "step": 46020 }, { "grad_norm": 0.40624862909317017, "learning_rate": 1.4105193783010151e-05, "loss": 0.0253, "step": 46030 }, { "grad_norm": 0.17732498049736023, "learning_rate": 1.4086014845794621e-05, "loss": 0.0208, "step": 46040 }, { "grad_norm": 0.4460781514644623, "learning_rate": 1.4066846818299489e-05, "loss": 0.0338, "step": 46050 }, { "grad_norm": 0.4827834963798523, "learning_rate": 1.4047689706347555e-05, "loss": 0.0456, "step": 46060 }, { "grad_norm": 0.324177086353302, "learning_rate": 1.402854351575822e-05, "loss": 0.0344, "step": 46070 }, { "grad_norm": 0.3490133583545685, "learning_rate": 1.4009408252347588e-05, "loss": 0.0195, "step": 46080 }, { "grad_norm": 0.37198394536972046, "learning_rate": 1.399028392192846e-05, "loss": 0.0418, "step": 46090 }, { "grad_norm": 0.7763420939445496, "learning_rate": 1.397117053031029e-05, "loss": 0.0293, "step": 46100 }, { "grad_norm": 0.49905410408973694, "learning_rate": 1.3952068083299213e-05, "loss": 0.0293, "step": 46110 }, { "grad_norm": 0.22850818932056427, "learning_rate": 1.3932976586698082e-05, "loss": 0.0231, "step": 46120 }, { "grad_norm": 0.39397552609443665, "learning_rate": 1.3913896046306363e-05, "loss": 0.0221, "step": 46130 }, { "grad_norm": 0.3935571014881134, "learning_rate": 1.389482646792023e-05, "loss": 0.0267, "step": 46140 }, { "grad_norm": 0.3236275911331177, "learning_rate": 1.387576785733251e-05, "loss": 0.02, "step": 46150 }, { "grad_norm": 0.39713266491889954, "learning_rate": 1.3856720220332703e-05, "loss": 0.024, "step": 46160 }, { "grad_norm": 0.5618557929992676, "learning_rate": 1.383768356270701e-05, "loss": 0.0249, "step": 46170 }, { "grad_norm": 0.8634275794029236, "learning_rate": 1.3818657890238207e-05, "loss": 0.0252, "step": 46180 }, { "grad_norm": 0.4631584584712982, "learning_rate": 1.3799643208705859e-05, "loss": 0.0332, "step": 46190 }, { "grad_norm": 0.3605622351169586, "learning_rate": 1.3780639523886058e-05, "loss": 0.0273, "step": 46200 }, { "grad_norm": 0.44188231229782104, "learning_rate": 1.3761646841551668e-05, "loss": 0.0515, "step": 46210 }, { "grad_norm": 0.42424657940864563, "learning_rate": 1.3742665167472146e-05, "loss": 0.0293, "step": 46220 }, { "grad_norm": 0.33987224102020264, "learning_rate": 1.372369450741363e-05, "loss": 0.0259, "step": 46230 }, { "grad_norm": 0.3427579998970032, "learning_rate": 1.3704734867138901e-05, "loss": 0.0387, "step": 46240 }, { "grad_norm": 0.5666027665138245, "learning_rate": 1.36857862524074e-05, "loss": 0.0229, "step": 46250 }, { "grad_norm": 0.30280476808547974, "learning_rate": 1.3666848668975213e-05, "loss": 0.0223, "step": 46260 }, { "grad_norm": 0.27260148525238037, "learning_rate": 1.3647922122595063e-05, "loss": 0.0242, "step": 46270 }, { "grad_norm": 0.23437441885471344, "learning_rate": 1.3629006619016366e-05, "loss": 0.0197, "step": 46280 }, { "grad_norm": 0.5548337697982788, "learning_rate": 1.3610102163985139e-05, "loss": 0.0567, "step": 46290 }, { "grad_norm": 0.3058606684207916, "learning_rate": 1.3591208763244057e-05, "loss": 0.0208, "step": 46300 }, { "grad_norm": 0.5111193656921387, "learning_rate": 1.3572326422532428e-05, "loss": 0.0275, "step": 46310 }, { "grad_norm": 0.5326345562934875, "learning_rate": 1.355345514758622e-05, "loss": 0.0312, "step": 46320 }, { "grad_norm": 0.2854044437408447, "learning_rate": 1.3534594944138007e-05, "loss": 0.0256, "step": 46330 }, { "grad_norm": 0.3486456871032715, "learning_rate": 1.3515745817917069e-05, "loss": 0.0315, "step": 46340 }, { "grad_norm": 0.426004558801651, "learning_rate": 1.3496907774649208e-05, "loss": 0.0225, "step": 46350 }, { "grad_norm": 0.4423982799053192, "learning_rate": 1.3478080820056987e-05, "loss": 0.0262, "step": 46360 }, { "grad_norm": 0.7323771715164185, "learning_rate": 1.3459264959859474e-05, "loss": 0.0419, "step": 46370 }, { "grad_norm": 0.4205777645111084, "learning_rate": 1.3440460199772487e-05, "loss": 0.0219, "step": 46380 }, { "grad_norm": 0.31265464425086975, "learning_rate": 1.3421666545508382e-05, "loss": 0.0218, "step": 46390 }, { "grad_norm": 0.3532252013683319, "learning_rate": 1.3402884002776194e-05, "loss": 0.0266, "step": 46400 }, { "grad_norm": 0.42495036125183105, "learning_rate": 1.3384112577281555e-05, "loss": 0.0291, "step": 46410 }, { "grad_norm": 0.27651315927505493, "learning_rate": 1.3365352274726711e-05, "loss": 0.0275, "step": 46420 }, { "grad_norm": 0.15493784844875336, "learning_rate": 1.3346603100810578e-05, "loss": 0.0244, "step": 46430 }, { "grad_norm": 0.38865411281585693, "learning_rate": 1.3327865061228645e-05, "loss": 0.0283, "step": 46440 }, { "grad_norm": 0.28269779682159424, "learning_rate": 1.330913816167304e-05, "loss": 0.0274, "step": 46450 }, { "grad_norm": 0.5686526894569397, "learning_rate": 1.3290422407832492e-05, "loss": 0.0246, "step": 46460 }, { "grad_norm": 0.5695643424987793, "learning_rate": 1.3271717805392354e-05, "loss": 0.0338, "step": 46470 }, { "grad_norm": 0.441814661026001, "learning_rate": 1.3253024360034582e-05, "loss": 0.0341, "step": 46480 }, { "grad_norm": 0.4092037081718445, "learning_rate": 1.323434207743779e-05, "loss": 0.0276, "step": 46490 }, { "grad_norm": 0.3713624179363251, "learning_rate": 1.3215670963277105e-05, "loss": 0.0226, "step": 46500 }, { "grad_norm": 0.335060179233551, "learning_rate": 1.3197011023224376e-05, "loss": 0.0334, "step": 46510 }, { "grad_norm": 0.22189736366271973, "learning_rate": 1.3178362262947941e-05, "loss": 0.0165, "step": 46520 }, { "grad_norm": 0.7853794693946838, "learning_rate": 1.3159724688112845e-05, "loss": 0.0296, "step": 46530 }, { "grad_norm": 0.7793488502502441, "learning_rate": 1.3141098304380683e-05, "loss": 0.0365, "step": 46540 }, { "grad_norm": 0.3264680504798889, "learning_rate": 1.3122483117409651e-05, "loss": 0.0256, "step": 46550 }, { "grad_norm": 0.4791078269481659, "learning_rate": 1.3103879132854552e-05, "loss": 0.0344, "step": 46560 }, { "grad_norm": 0.7287379503250122, "learning_rate": 1.3085286356366771e-05, "loss": 0.0484, "step": 46570 }, { "grad_norm": 0.448066383600235, "learning_rate": 1.3066704793594337e-05, "loss": 0.0233, "step": 46580 }, { "grad_norm": 0.3356742262840271, "learning_rate": 1.3048134450181816e-05, "loss": 0.0294, "step": 46590 }, { "grad_norm": 0.4889694154262543, "learning_rate": 1.3029575331770394e-05, "loss": 0.0357, "step": 46600 }, { "grad_norm": 0.3984782099723816, "learning_rate": 1.3011027443997837e-05, "loss": 0.0287, "step": 46610 }, { "grad_norm": 0.5172067880630493, "learning_rate": 1.2992490792498507e-05, "loss": 0.0384, "step": 46620 }, { "grad_norm": 0.167162224650383, "learning_rate": 1.297396538290333e-05, "loss": 0.0225, "step": 46630 }, { "grad_norm": 0.4208340644836426, "learning_rate": 1.2955451220839888e-05, "loss": 0.0285, "step": 46640 }, { "grad_norm": 0.652481198310852, "learning_rate": 1.2936948311932223e-05, "loss": 0.0258, "step": 46650 }, { "grad_norm": 0.5432580709457397, "learning_rate": 1.2918456661801104e-05, "loss": 0.0293, "step": 46660 }, { "grad_norm": 0.42307931184768677, "learning_rate": 1.2899976276063736e-05, "loss": 0.0324, "step": 46670 }, { "grad_norm": 0.22972524166107178, "learning_rate": 1.2881507160334022e-05, "loss": 0.0228, "step": 46680 }, { "grad_norm": 0.27489352226257324, "learning_rate": 1.286304932022238e-05, "loss": 0.023, "step": 46690 }, { "grad_norm": 0.3439152240753174, "learning_rate": 1.2844602761335806e-05, "loss": 0.0336, "step": 46700 }, { "grad_norm": 0.35248374938964844, "learning_rate": 1.2826167489277885e-05, "loss": 0.0217, "step": 46710 }, { "grad_norm": 0.49511948227882385, "learning_rate": 1.2807743509648745e-05, "loss": 0.0269, "step": 46720 }, { "grad_norm": 0.2837083637714386, "learning_rate": 1.2789330828045149e-05, "loss": 0.0406, "step": 46730 }, { "grad_norm": 0.4584297835826874, "learning_rate": 1.2770929450060332e-05, "loss": 0.0281, "step": 46740 }, { "grad_norm": 0.4527847468852997, "learning_rate": 1.2752539381284184e-05, "loss": 0.0257, "step": 46750 }, { "grad_norm": 0.493476927280426, "learning_rate": 1.273416062730311e-05, "loss": 0.0392, "step": 46760 }, { "grad_norm": 0.4047684669494629, "learning_rate": 1.2715793193700088e-05, "loss": 0.0231, "step": 46770 }, { "grad_norm": 0.4635102152824402, "learning_rate": 1.2697437086054664e-05, "loss": 0.0427, "step": 46780 }, { "grad_norm": 0.3316252827644348, "learning_rate": 1.2679092309942937e-05, "loss": 0.0243, "step": 46790 }, { "grad_norm": 0.3021501898765564, "learning_rate": 1.266075887093755e-05, "loss": 0.0434, "step": 46800 }, { "grad_norm": 0.24341553449630737, "learning_rate": 1.2642436774607757e-05, "loss": 0.0386, "step": 46810 }, { "grad_norm": 0.39484837651252747, "learning_rate": 1.2624126026519278e-05, "loss": 0.0278, "step": 46820 }, { "grad_norm": 0.5125201940536499, "learning_rate": 1.2605826632234474e-05, "loss": 0.0259, "step": 46830 }, { "grad_norm": 0.3657539188861847, "learning_rate": 1.2587538597312198e-05, "loss": 0.0184, "step": 46840 }, { "grad_norm": 0.36664214730262756, "learning_rate": 1.2569261927307884e-05, "loss": 0.0372, "step": 46850 }, { "grad_norm": 0.4305526614189148, "learning_rate": 1.2550996627773493e-05, "loss": 0.0268, "step": 46860 }, { "grad_norm": 0.4296928942203522, "learning_rate": 1.2532742704257527e-05, "loss": 0.03, "step": 46870 }, { "grad_norm": 0.43034878373146057, "learning_rate": 1.2514500162305087e-05, "loss": 0.0244, "step": 46880 }, { "grad_norm": 0.4150432050228119, "learning_rate": 1.2496269007457728e-05, "loss": 0.0308, "step": 46890 }, { "grad_norm": 0.31794995069503784, "learning_rate": 1.2478049245253625e-05, "loss": 0.0325, "step": 46900 }, { "grad_norm": 0.3426400423049927, "learning_rate": 1.2459840881227459e-05, "loss": 0.0278, "step": 46910 }, { "grad_norm": 0.394623726606369, "learning_rate": 1.2441643920910435e-05, "loss": 0.0276, "step": 46920 }, { "grad_norm": 0.3296925723552704, "learning_rate": 1.2423458369830322e-05, "loss": 0.0217, "step": 46930 }, { "grad_norm": 0.3774091601371765, "learning_rate": 1.2405284233511406e-05, "loss": 0.0267, "step": 46940 }, { "grad_norm": 0.1950535923242569, "learning_rate": 1.2387121517474487e-05, "loss": 0.0239, "step": 46950 }, { "grad_norm": 0.3137427568435669, "learning_rate": 1.2368970227236975e-05, "loss": 0.0219, "step": 46960 }, { "grad_norm": 0.34891414642333984, "learning_rate": 1.2350830368312688e-05, "loss": 0.0258, "step": 46970 }, { "grad_norm": 0.3012612462043762, "learning_rate": 1.2332701946212083e-05, "loss": 0.0255, "step": 46980 }, { "grad_norm": 0.33097293972969055, "learning_rate": 1.2314584966442077e-05, "loss": 0.0288, "step": 46990 }, { "grad_norm": 0.23562322556972504, "learning_rate": 1.2296479434506136e-05, "loss": 0.0507, "step": 47000 }, { "grad_norm": 0.4063968360424042, "learning_rate": 1.2278385355904232e-05, "loss": 0.0323, "step": 47010 }, { "grad_norm": 0.2506423890590668, "learning_rate": 1.2260302736132867e-05, "loss": 0.0189, "step": 47020 }, { "grad_norm": 0.6109652519226074, "learning_rate": 1.2242231580685098e-05, "loss": 0.0396, "step": 47030 }, { "grad_norm": 0.4134526252746582, "learning_rate": 1.2224171895050413e-05, "loss": 0.0358, "step": 47040 }, { "grad_norm": 0.40684521198272705, "learning_rate": 1.2206123684714903e-05, "loss": 0.023, "step": 47050 }, { "grad_norm": 0.402122437953949, "learning_rate": 1.2188086955161132e-05, "loss": 0.027, "step": 47060 }, { "grad_norm": 0.407728374004364, "learning_rate": 1.2170061711868175e-05, "loss": 0.0299, "step": 47070 }, { "grad_norm": 0.4712682366371155, "learning_rate": 1.215204796031163e-05, "loss": 0.0274, "step": 47080 }, { "grad_norm": 0.5195026993751526, "learning_rate": 1.2134045705963599e-05, "loss": 0.0232, "step": 47090 }, { "grad_norm": 0.2911660671234131, "learning_rate": 1.2116054954292689e-05, "loss": 0.0264, "step": 47100 }, { "grad_norm": 0.21715913712978363, "learning_rate": 1.2098075710764011e-05, "loss": 0.028, "step": 47110 }, { "grad_norm": 0.4006708860397339, "learning_rate": 1.2080107980839183e-05, "loss": 0.0231, "step": 47120 }, { "grad_norm": 0.3524590730667114, "learning_rate": 1.2062151769976343e-05, "loss": 0.0234, "step": 47130 }, { "grad_norm": 0.5090892314910889, "learning_rate": 1.204420708363011e-05, "loss": 0.0238, "step": 47140 }, { "grad_norm": 0.4550182521343231, "learning_rate": 1.2026273927251597e-05, "loss": 0.0236, "step": 47150 }, { "grad_norm": 0.35153934359550476, "learning_rate": 1.2008352306288424e-05, "loss": 0.0311, "step": 47160 }, { "grad_norm": 0.17472366988658905, "learning_rate": 1.1990442226184695e-05, "loss": 0.0253, "step": 47170 }, { "grad_norm": 0.5150336027145386, "learning_rate": 1.1972543692381066e-05, "loss": 0.0438, "step": 47180 }, { "grad_norm": 0.46265947818756104, "learning_rate": 1.1954656710314576e-05, "loss": 0.0273, "step": 47190 }, { "grad_norm": 0.4169263243675232, "learning_rate": 1.1936781285418875e-05, "loss": 0.0228, "step": 47200 }, { "grad_norm": 0.4316943883895874, "learning_rate": 1.1918917423123993e-05, "loss": 0.035, "step": 47210 }, { "grad_norm": 0.3001340925693512, "learning_rate": 1.1901065128856537e-05, "loss": 0.0159, "step": 47220 }, { "grad_norm": 0.4896334707736969, "learning_rate": 1.1883224408039551e-05, "loss": 0.0235, "step": 47230 }, { "grad_norm": 0.4895731508731842, "learning_rate": 1.1865395266092578e-05, "loss": 0.0269, "step": 47240 }, { "grad_norm": 0.20763975381851196, "learning_rate": 1.1847577708431633e-05, "loss": 0.031, "step": 47250 }, { "grad_norm": 0.5905879139900208, "learning_rate": 1.1829771740469225e-05, "loss": 0.0282, "step": 47260 }, { "grad_norm": 0.38774439692497253, "learning_rate": 1.1811977367614324e-05, "loss": 0.0235, "step": 47270 }, { "grad_norm": 0.4580650329589844, "learning_rate": 1.1794194595272412e-05, "loss": 0.0306, "step": 47280 }, { "grad_norm": 0.530802309513092, "learning_rate": 1.1776423428845423e-05, "loss": 0.0223, "step": 47290 }, { "grad_norm": 0.3160569667816162, "learning_rate": 1.1758663873731756e-05, "loss": 0.0285, "step": 47300 }, { "grad_norm": 0.8402906656265259, "learning_rate": 1.1740915935326302e-05, "loss": 0.0254, "step": 47310 }, { "grad_norm": 0.3031849265098572, "learning_rate": 1.1723179619020396e-05, "loss": 0.0304, "step": 47320 }, { "grad_norm": 0.30796417593955994, "learning_rate": 1.1705454930201914e-05, "loss": 0.0183, "step": 47330 }, { "grad_norm": 0.42016151547431946, "learning_rate": 1.1687741874255087e-05, "loss": 0.0259, "step": 47340 }, { "grad_norm": 0.41812077164649963, "learning_rate": 1.1670040456560728e-05, "loss": 0.0228, "step": 47350 }, { "grad_norm": 0.28818216919898987, "learning_rate": 1.1652350682496005e-05, "loss": 0.0285, "step": 47360 }, { "grad_norm": 0.35109955072402954, "learning_rate": 1.163467255743465e-05, "loss": 0.0277, "step": 47370 }, { "grad_norm": 0.8139612674713135, "learning_rate": 1.1617006086746796e-05, "loss": 0.0276, "step": 47380 }, { "grad_norm": 0.32310500741004944, "learning_rate": 1.1599351275799047e-05, "loss": 0.0313, "step": 47390 }, { "grad_norm": 0.5955843925476074, "learning_rate": 1.1581708129954466e-05, "loss": 0.025, "step": 47400 }, { "grad_norm": 0.4636073410511017, "learning_rate": 1.1564076654572587e-05, "loss": 0.0225, "step": 47410 }, { "grad_norm": 0.4740544557571411, "learning_rate": 1.1546456855009358e-05, "loss": 0.0217, "step": 47420 }, { "grad_norm": 0.4778388738632202, "learning_rate": 1.1528848736617248e-05, "loss": 0.0212, "step": 47430 }, { "grad_norm": 0.25613582134246826, "learning_rate": 1.1511252304745112e-05, "loss": 0.0294, "step": 47440 }, { "grad_norm": 0.34307578206062317, "learning_rate": 1.1493667564738297e-05, "loss": 0.0334, "step": 47450 }, { "grad_norm": 0.5334523320198059, "learning_rate": 1.1476094521938574e-05, "loss": 0.0296, "step": 47460 }, { "grad_norm": 0.6569293141365051, "learning_rate": 1.1458533181684167e-05, "loss": 0.0315, "step": 47470 }, { "grad_norm": 0.3624248802661896, "learning_rate": 1.1440983549309753e-05, "loss": 0.016, "step": 47480 }, { "grad_norm": 0.2949807941913605, "learning_rate": 1.1423445630146434e-05, "loss": 0.0235, "step": 47490 }, { "grad_norm": 0.6273857951164246, "learning_rate": 1.1405919429521799e-05, "loss": 0.0271, "step": 47500 }, { "grad_norm": 0.47681036591529846, "learning_rate": 1.1388404952759802e-05, "loss": 0.025, "step": 47510 }, { "grad_norm": 0.2524499297142029, "learning_rate": 1.1370902205180923e-05, "loss": 0.027, "step": 47520 }, { "grad_norm": 0.2824949026107788, "learning_rate": 1.1353411192101987e-05, "loss": 0.0274, "step": 47530 }, { "grad_norm": 0.4000931680202484, "learning_rate": 1.133593191883634e-05, "loss": 0.0184, "step": 47540 }, { "grad_norm": 0.6154193878173828, "learning_rate": 1.1318464390693711e-05, "loss": 0.0241, "step": 47550 }, { "grad_norm": 0.4283350706100464, "learning_rate": 1.1301008612980257e-05, "loss": 0.0208, "step": 47560 }, { "grad_norm": 0.4990621507167816, "learning_rate": 1.128356459099863e-05, "loss": 0.0267, "step": 47570 }, { "grad_norm": 0.43033283948898315, "learning_rate": 1.1266132330047802e-05, "loss": 0.0312, "step": 47580 }, { "grad_norm": 0.45241984724998474, "learning_rate": 1.1248711835423281e-05, "loss": 0.038, "step": 47590 }, { "grad_norm": 0.26415377855300903, "learning_rate": 1.123130311241693e-05, "loss": 0.0207, "step": 47600 }, { "grad_norm": 0.44974198937416077, "learning_rate": 1.1213906166317068e-05, "loss": 0.0258, "step": 47610 }, { "grad_norm": 0.30115386843681335, "learning_rate": 1.1196521002408427e-05, "loss": 0.028, "step": 47620 }, { "grad_norm": 0.8395899534225464, "learning_rate": 1.1179147625972159e-05, "loss": 0.0263, "step": 47630 }, { "grad_norm": 0.2498077154159546, "learning_rate": 1.1161786042285822e-05, "loss": 0.0229, "step": 47640 }, { "grad_norm": 1.4968862533569336, "learning_rate": 1.1144436256623447e-05, "loss": 0.0263, "step": 47650 }, { "grad_norm": 0.3268860876560211, "learning_rate": 1.1127098274255392e-05, "loss": 0.0281, "step": 47660 }, { "grad_norm": 0.2927423119544983, "learning_rate": 1.1109772100448512e-05, "loss": 0.0217, "step": 47670 }, { "grad_norm": 0.3278455138206482, "learning_rate": 1.1092457740466033e-05, "loss": 0.0347, "step": 47680 }, { "grad_norm": 0.2862137258052826, "learning_rate": 1.10751551995676e-05, "loss": 0.0174, "step": 47690 }, { "grad_norm": 0.2734987437725067, "learning_rate": 1.1057864483009262e-05, "loss": 0.0226, "step": 47700 }, { "grad_norm": 0.2596791982650757, "learning_rate": 1.1040585596043473e-05, "loss": 0.0228, "step": 47710 }, { "grad_norm": 0.2766079604625702, "learning_rate": 1.1023318543919148e-05, "loss": 0.0191, "step": 47720 }, { "grad_norm": 0.32277992367744446, "learning_rate": 1.10060633318815e-05, "loss": 0.0353, "step": 47730 }, { "grad_norm": 0.32471761107444763, "learning_rate": 1.0988819965172248e-05, "loss": 0.0238, "step": 47740 }, { "grad_norm": 0.3810557425022125, "learning_rate": 1.0971588449029462e-05, "loss": 0.024, "step": 47750 }, { "grad_norm": 0.500877320766449, "learning_rate": 1.095436878868762e-05, "loss": 0.0228, "step": 47760 }, { "grad_norm": 0.24943701922893524, "learning_rate": 1.0937160989377598e-05, "loss": 0.0245, "step": 47770 }, { "grad_norm": 0.2841101884841919, "learning_rate": 1.0919965056326676e-05, "loss": 0.0355, "step": 47780 }, { "grad_norm": 0.4023680090904236, "learning_rate": 1.0902780994758504e-05, "loss": 0.0247, "step": 47790 }, { "grad_norm": 0.3704265356063843, "learning_rate": 1.0885608809893193e-05, "loss": 0.0252, "step": 47800 }, { "grad_norm": 0.4168379008769989, "learning_rate": 1.0868448506947142e-05, "loss": 0.0297, "step": 47810 }, { "grad_norm": 0.44078943133354187, "learning_rate": 1.0851300091133243e-05, "loss": 0.0187, "step": 47820 }, { "grad_norm": 0.25190502405166626, "learning_rate": 1.083416356766071e-05, "loss": 0.0276, "step": 47830 }, { "grad_norm": 0.31747207045555115, "learning_rate": 1.0817038941735175e-05, "loss": 0.0322, "step": 47840 }, { "grad_norm": 0.4250316619873047, "learning_rate": 1.0799926218558642e-05, "loss": 0.0255, "step": 47850 }, { "grad_norm": 0.2629362940788269, "learning_rate": 1.0782825403329488e-05, "loss": 0.024, "step": 47860 }, { "grad_norm": 0.5845738053321838, "learning_rate": 1.076573650124254e-05, "loss": 0.0262, "step": 47870 }, { "grad_norm": 0.26554951071739197, "learning_rate": 1.0748659517488891e-05, "loss": 0.0168, "step": 47880 }, { "grad_norm": 0.39044174551963806, "learning_rate": 1.0731594457256138e-05, "loss": 0.0226, "step": 47890 }, { "grad_norm": 0.45103320479393005, "learning_rate": 1.0714541325728139e-05, "loss": 0.0218, "step": 47900 }, { "grad_norm": 0.602997899055481, "learning_rate": 1.0697500128085231e-05, "loss": 0.0308, "step": 47910 }, { "grad_norm": 0.26389846205711365, "learning_rate": 1.0680470869504055e-05, "loss": 0.0322, "step": 47920 }, { "grad_norm": 0.6064987778663635, "learning_rate": 1.066345355515766e-05, "loss": 0.0292, "step": 47930 }, { "grad_norm": 0.5063713192939758, "learning_rate": 1.0646448190215453e-05, "loss": 0.0305, "step": 47940 }, { "grad_norm": 0.48903709650039673, "learning_rate": 1.0629454779843217e-05, "loss": 0.0304, "step": 47950 }, { "grad_norm": 0.3209734261035919, "learning_rate": 1.0612473329203082e-05, "loss": 0.0193, "step": 47960 }, { "grad_norm": 0.43438291549682617, "learning_rate": 1.0595503843453596e-05, "loss": 0.032, "step": 47970 }, { "grad_norm": 0.4154350161552429, "learning_rate": 1.0578546327749634e-05, "loss": 0.0328, "step": 47980 }, { "grad_norm": 0.35428211092948914, "learning_rate": 1.0561600787242425e-05, "loss": 0.0318, "step": 47990 }, { "grad_norm": 0.47870779037475586, "learning_rate": 1.0544667227079591e-05, "loss": 0.0243, "step": 48000 }, { "grad_norm": 0.49620187282562256, "learning_rate": 1.0527745652405085e-05, "loss": 0.0245, "step": 48010 }, { "grad_norm": 0.3310307562351227, "learning_rate": 1.051083606835927e-05, "loss": 0.0198, "step": 48020 }, { "grad_norm": 0.21949701011180878, "learning_rate": 1.049393848007878e-05, "loss": 0.0306, "step": 48030 }, { "grad_norm": 0.5682628750801086, "learning_rate": 1.0477052892696709e-05, "loss": 0.0236, "step": 48040 }, { "grad_norm": 0.3630443811416626, "learning_rate": 1.0460179311342394e-05, "loss": 0.0251, "step": 48050 }, { "grad_norm": 0.3625478446483612, "learning_rate": 1.0443317741141634e-05, "loss": 0.0257, "step": 48060 }, { "grad_norm": 0.3499285876750946, "learning_rate": 1.0426468187216514e-05, "loss": 0.0202, "step": 48070 }, { "grad_norm": 0.5383820533752441, "learning_rate": 1.0409630654685477e-05, "loss": 0.0265, "step": 48080 }, { "grad_norm": 1.078401803970337, "learning_rate": 1.039280514866332e-05, "loss": 0.0296, "step": 48090 }, { "grad_norm": 0.4632052779197693, "learning_rate": 1.0375991674261198e-05, "loss": 0.0342, "step": 48100 }, { "grad_norm": 0.6270240545272827, "learning_rate": 1.0359190236586575e-05, "loss": 0.0186, "step": 48110 }, { "grad_norm": 0.3406328558921814, "learning_rate": 1.0342400840743322e-05, "loss": 0.0252, "step": 48120 }, { "grad_norm": 0.24792149662971497, "learning_rate": 1.0325623491831593e-05, "loss": 0.0186, "step": 48130 }, { "grad_norm": 0.33666616678237915, "learning_rate": 1.0308858194947906e-05, "loss": 0.0249, "step": 48140 }, { "grad_norm": 0.26080837845802307, "learning_rate": 1.0292104955185111e-05, "loss": 0.0376, "step": 48150 }, { "grad_norm": 0.279622882604599, "learning_rate": 1.0275363777632396e-05, "loss": 0.0284, "step": 48160 }, { "grad_norm": 0.2237253189086914, "learning_rate": 1.0258634667375321e-05, "loss": 0.0236, "step": 48170 }, { "grad_norm": 0.49202829599380493, "learning_rate": 1.02419176294957e-05, "loss": 0.0334, "step": 48180 }, { "grad_norm": 0.7952461242675781, "learning_rate": 1.0225212669071782e-05, "loss": 0.0386, "step": 48190 }, { "grad_norm": 0.5401216149330139, "learning_rate": 1.0208519791178029e-05, "loss": 0.0239, "step": 48200 }, { "grad_norm": 0.2864031195640564, "learning_rate": 1.019183900088535e-05, "loss": 0.0249, "step": 48210 }, { "grad_norm": 0.30128028988838196, "learning_rate": 1.0175170303260906e-05, "loss": 0.0234, "step": 48220 }, { "grad_norm": 0.6385669708251953, "learning_rate": 1.0158513703368206e-05, "loss": 0.0208, "step": 48230 }, { "grad_norm": 0.25324827432632446, "learning_rate": 1.0141869206267095e-05, "loss": 0.0209, "step": 48240 }, { "grad_norm": 0.31019800901412964, "learning_rate": 1.0125236817013723e-05, "loss": 0.038, "step": 48250 }, { "grad_norm": 3.4052555561065674, "learning_rate": 1.010861654066056e-05, "loss": 0.0353, "step": 48260 }, { "grad_norm": 0.34037429094314575, "learning_rate": 1.0092008382256434e-05, "loss": 0.023, "step": 48270 }, { "grad_norm": 0.5312777757644653, "learning_rate": 1.0075412346846458e-05, "loss": 0.0375, "step": 48280 }, { "grad_norm": 0.21876981854438782, "learning_rate": 1.0058828439472056e-05, "loss": 0.0216, "step": 48290 }, { "grad_norm": 0.4537816643714905, "learning_rate": 1.0042256665170996e-05, "loss": 0.0347, "step": 48300 }, { "grad_norm": 0.3645816743373871, "learning_rate": 1.0025697028977332e-05, "loss": 0.0189, "step": 48310 }, { "grad_norm": 0.7193545699119568, "learning_rate": 1.0009149535921454e-05, "loss": 0.0347, "step": 48320 }, { "grad_norm": 0.2982308268547058, "learning_rate": 9.992614191030031e-06, "loss": 0.0218, "step": 48330 }, { "grad_norm": 0.36046987771987915, "learning_rate": 9.976090999326115e-06, "loss": 0.0309, "step": 48340 }, { "grad_norm": 0.3751660883426666, "learning_rate": 9.959579965828952e-06, "loss": 0.0234, "step": 48350 }, { "grad_norm": 0.45029690861701965, "learning_rate": 9.943081095554218e-06, "loss": 0.0274, "step": 48360 }, { "grad_norm": 0.22826911509037018, "learning_rate": 9.926594393513783e-06, "loss": 0.0177, "step": 48370 }, { "grad_norm": 0.8955129384994507, "learning_rate": 9.910119864715906e-06, "loss": 0.0232, "step": 48380 }, { "grad_norm": 0.449083149433136, "learning_rate": 9.8936575141651e-06, "loss": 0.02, "step": 48390 }, { "grad_norm": 0.2567359209060669, "learning_rate": 9.877207346862194e-06, "loss": 0.0211, "step": 48400 }, { "grad_norm": 0.4737868010997772, "learning_rate": 9.860769367804312e-06, "loss": 0.0243, "step": 48410 }, { "grad_norm": 0.4433847963809967, "learning_rate": 9.844343581984877e-06, "loss": 0.0265, "step": 48420 }, { "grad_norm": 0.33279338479042053, "learning_rate": 9.82792999439362e-06, "loss": 0.0311, "step": 48430 }, { "grad_norm": 0.35632479190826416, "learning_rate": 9.811528610016546e-06, "loss": 0.035, "step": 48440 }, { "grad_norm": 0.5059942603111267, "learning_rate": 9.79513943383597e-06, "loss": 0.0213, "step": 48450 }, { "grad_norm": 0.29740220308303833, "learning_rate": 9.778762470830489e-06, "loss": 0.0356, "step": 48460 }, { "grad_norm": 0.421627014875412, "learning_rate": 9.762397725974982e-06, "loss": 0.0196, "step": 48470 }, { "grad_norm": 0.2529224753379822, "learning_rate": 9.746045204240622e-06, "loss": 0.0228, "step": 48480 }, { "grad_norm": 0.3033066391944885, "learning_rate": 9.729704910594917e-06, "loss": 0.0272, "step": 48490 }, { "grad_norm": 0.3662887513637543, "learning_rate": 9.713376850001554e-06, "loss": 0.0293, "step": 48500 }, { "grad_norm": 0.4607449769973755, "learning_rate": 9.697061027420622e-06, "loss": 0.0202, "step": 48510 }, { "grad_norm": 0.4112381935119629, "learning_rate": 9.680757447808385e-06, "loss": 0.0224, "step": 48520 }, { "grad_norm": 0.3141767680644989, "learning_rate": 9.664466116117488e-06, "loss": 0.0244, "step": 48530 }, { "grad_norm": 0.565269410610199, "learning_rate": 9.64818703729678e-06, "loss": 0.0339, "step": 48540 }, { "grad_norm": 0.4649280905723572, "learning_rate": 9.631920216291423e-06, "loss": 0.0256, "step": 48550 }, { "grad_norm": 0.48664212226867676, "learning_rate": 9.615665658042849e-06, "loss": 0.0288, "step": 48560 }, { "grad_norm": 0.5189189314842224, "learning_rate": 9.599423367488747e-06, "loss": 0.0226, "step": 48570 }, { "grad_norm": 0.3643021285533905, "learning_rate": 9.583193349563124e-06, "loss": 0.0485, "step": 48580 }, { "grad_norm": 0.46803778409957886, "learning_rate": 9.566975609196216e-06, "loss": 0.0241, "step": 48590 }, { "grad_norm": 0.4529305100440979, "learning_rate": 9.550770151314548e-06, "loss": 0.0261, "step": 48600 }, { "grad_norm": 0.24838000535964966, "learning_rate": 9.53457698084091e-06, "loss": 0.026, "step": 48610 }, { "grad_norm": 0.6354901194572449, "learning_rate": 9.518396102694355e-06, "loss": 0.0424, "step": 48620 }, { "grad_norm": 0.27740219235420227, "learning_rate": 9.502227521790198e-06, "loss": 0.0284, "step": 48630 }, { "grad_norm": 0.4060554802417755, "learning_rate": 9.486071243040063e-06, "loss": 0.0263, "step": 48640 }, { "grad_norm": 0.22459913790225983, "learning_rate": 9.469927271351747e-06, "loss": 0.017, "step": 48650 }, { "grad_norm": 0.29918748140335083, "learning_rate": 9.453795611629419e-06, "loss": 0.0257, "step": 48660 }, { "grad_norm": 0.42930373549461365, "learning_rate": 9.437676268773399e-06, "loss": 0.0197, "step": 48670 }, { "grad_norm": 0.3009907901287079, "learning_rate": 9.421569247680357e-06, "loss": 0.023, "step": 48680 }, { "grad_norm": 0.4948289692401886, "learning_rate": 9.40547455324316e-06, "loss": 0.0294, "step": 48690 }, { "grad_norm": 0.6274747252464294, "learning_rate": 9.389392190350965e-06, "loss": 0.0296, "step": 48700 }, { "grad_norm": 0.36035817861557007, "learning_rate": 9.373322163889153e-06, "loss": 0.036, "step": 48710 }, { "grad_norm": 0.48472046852111816, "learning_rate": 9.357264478739375e-06, "loss": 0.0184, "step": 48720 }, { "grad_norm": 0.40793225169181824, "learning_rate": 9.341219139779567e-06, "loss": 0.0302, "step": 48730 }, { "grad_norm": 0.43112510442733765, "learning_rate": 9.325186151883824e-06, "loss": 0.018, "step": 48740 }, { "grad_norm": 0.49004948139190674, "learning_rate": 9.30916551992258e-06, "loss": 0.0254, "step": 48750 }, { "grad_norm": 0.364414244890213, "learning_rate": 9.293157248762479e-06, "loss": 0.0199, "step": 48760 }, { "grad_norm": 0.6712944507598877, "learning_rate": 9.2771613432664e-06, "loss": 0.0304, "step": 48770 }, { "grad_norm": 0.4494985342025757, "learning_rate": 9.261177808293481e-06, "loss": 0.0233, "step": 48780 }, { "grad_norm": 0.4345663785934448, "learning_rate": 9.245206648699096e-06, "loss": 0.0283, "step": 48790 }, { "grad_norm": 0.30332523584365845, "learning_rate": 9.22924786933485e-06, "loss": 0.0259, "step": 48800 }, { "grad_norm": 0.23571360111236572, "learning_rate": 9.213301475048642e-06, "loss": 0.0271, "step": 48810 }, { "grad_norm": 0.2682186961174011, "learning_rate": 9.197367470684504e-06, "loss": 0.0234, "step": 48820 }, { "grad_norm": 0.33662575483322144, "learning_rate": 9.181445861082816e-06, "loss": 0.0243, "step": 48830 }, { "grad_norm": 0.7080529928207397, "learning_rate": 9.16553665108012e-06, "loss": 0.0272, "step": 48840 }, { "grad_norm": 0.6239464282989502, "learning_rate": 9.149639845509223e-06, "loss": 0.0343, "step": 48850 }, { "grad_norm": 0.47076767683029175, "learning_rate": 9.133755449199144e-06, "loss": 0.0264, "step": 48860 }, { "grad_norm": 0.44772934913635254, "learning_rate": 9.117883466975135e-06, "loss": 0.0226, "step": 48870 }, { "grad_norm": 0.33078479766845703, "learning_rate": 9.10202390365873e-06, "loss": 0.0222, "step": 48880 }, { "grad_norm": 0.35475584864616394, "learning_rate": 9.086176764067583e-06, "loss": 0.0314, "step": 48890 }, { "grad_norm": 0.4427739977836609, "learning_rate": 9.070342053015684e-06, "loss": 0.0338, "step": 48900 }, { "grad_norm": 0.3793254792690277, "learning_rate": 9.054519775313187e-06, "loss": 0.0387, "step": 48910 }, { "grad_norm": 0.4732681214809418, "learning_rate": 9.038709935766476e-06, "loss": 0.0374, "step": 48920 }, { "grad_norm": 0.41074860095977783, "learning_rate": 9.02291253917817e-06, "loss": 0.0236, "step": 48930 }, { "grad_norm": 0.4773959815502167, "learning_rate": 9.007127590347091e-06, "loss": 0.0315, "step": 48940 }, { "grad_norm": 0.5445898771286011, "learning_rate": 8.991355094068288e-06, "loss": 0.0236, "step": 48950 }, { "grad_norm": 0.5966605544090271, "learning_rate": 8.975595055133062e-06, "loss": 0.0309, "step": 48960 }, { "grad_norm": 0.33238619565963745, "learning_rate": 8.959847478328848e-06, "loss": 0.0281, "step": 48970 }, { "grad_norm": 0.3907148540019989, "learning_rate": 8.944112368439378e-06, "loss": 0.0256, "step": 48980 }, { "grad_norm": 0.7593218684196472, "learning_rate": 8.928389730244552e-06, "loss": 0.0302, "step": 48990 }, { "grad_norm": 0.5140994191169739, "learning_rate": 8.912679568520494e-06, "loss": 0.0296, "step": 49000 }, { "grad_norm": 0.346245676279068, "learning_rate": 8.896981888039534e-06, "loss": 0.0251, "step": 49010 }, { "grad_norm": 0.39265432953834534, "learning_rate": 8.881296693570201e-06, "loss": 0.0147, "step": 49020 }, { "grad_norm": 0.8069069385528564, "learning_rate": 8.865623989877281e-06, "loss": 0.0459, "step": 49030 }, { "grad_norm": 0.29310697317123413, "learning_rate": 8.849963781721681e-06, "loss": 0.0212, "step": 49040 }, { "grad_norm": 0.4273284077644348, "learning_rate": 8.834316073860588e-06, "loss": 0.0313, "step": 49050 }, { "grad_norm": 0.5416485071182251, "learning_rate": 8.818680871047357e-06, "loss": 0.0263, "step": 49060 }, { "grad_norm": 0.3041563928127289, "learning_rate": 8.803058178031549e-06, "loss": 0.0292, "step": 49070 }, { "grad_norm": 0.38567137718200684, "learning_rate": 8.787447999558922e-06, "loss": 0.0244, "step": 49080 }, { "grad_norm": 0.6285576224327087, "learning_rate": 8.77185034037144e-06, "loss": 0.0303, "step": 49090 }, { "grad_norm": 0.341379314661026, "learning_rate": 8.756265205207259e-06, "loss": 0.0176, "step": 49100 }, { "grad_norm": 0.4142002761363983, "learning_rate": 8.740692598800732e-06, "loss": 0.0253, "step": 49110 }, { "grad_norm": 0.3174487054347992, "learning_rate": 8.72513252588239e-06, "loss": 0.0217, "step": 49120 }, { "grad_norm": 0.5599806308746338, "learning_rate": 8.709584991178998e-06, "loss": 0.0253, "step": 49130 }, { "grad_norm": 0.5394163727760315, "learning_rate": 8.694049999413479e-06, "loss": 0.0376, "step": 49140 }, { "grad_norm": 0.9098063707351685, "learning_rate": 8.678527555304945e-06, "loss": 0.0327, "step": 49150 }, { "grad_norm": 0.30645689368247986, "learning_rate": 8.663017663568712e-06, "loss": 0.0308, "step": 49160 }, { "grad_norm": 0.4661237299442291, "learning_rate": 8.647520328916259e-06, "loss": 0.0297, "step": 49170 }, { "grad_norm": 0.363598108291626, "learning_rate": 8.632035556055307e-06, "loss": 0.0273, "step": 49180 }, { "grad_norm": 0.3839073181152344, "learning_rate": 8.616563349689672e-06, "loss": 0.029, "step": 49190 }, { "grad_norm": 0.854802131652832, "learning_rate": 8.601103714519448e-06, "loss": 0.0203, "step": 49200 }, { "grad_norm": 0.32631608843803406, "learning_rate": 8.58565665524082e-06, "loss": 0.0424, "step": 49210 }, { "grad_norm": 0.4678082764148712, "learning_rate": 8.570222176546222e-06, "loss": 0.0242, "step": 49220 }, { "grad_norm": 0.22036035358905792, "learning_rate": 8.554800283124242e-06, "loss": 0.0209, "step": 49230 }, { "grad_norm": 0.39862558245658875, "learning_rate": 8.539390979659639e-06, "loss": 0.0229, "step": 49240 }, { "grad_norm": 0.5953883528709412, "learning_rate": 8.523994270833352e-06, "loss": 0.0462, "step": 49250 }, { "grad_norm": 0.5453844666481018, "learning_rate": 8.5086101613225e-06, "loss": 0.0339, "step": 49260 }, { "grad_norm": 0.3186790645122528, "learning_rate": 8.493238655800346e-06, "loss": 0.0341, "step": 49270 }, { "grad_norm": 0.3146928548812866, "learning_rate": 8.47787975893638e-06, "loss": 0.0237, "step": 49280 }, { "grad_norm": 0.35057252645492554, "learning_rate": 8.462533475396211e-06, "loss": 0.0211, "step": 49290 }, { "grad_norm": 0.7834388613700867, "learning_rate": 8.447199809841643e-06, "loss": 0.0265, "step": 49300 }, { "grad_norm": 0.24240006506443024, "learning_rate": 8.431878766930635e-06, "loss": 0.0308, "step": 49310 }, { "grad_norm": 0.5964083671569824, "learning_rate": 8.416570351317304e-06, "loss": 0.0262, "step": 49320 }, { "grad_norm": 0.5065566301345825, "learning_rate": 8.401274567651973e-06, "loss": 0.03, "step": 49330 }, { "grad_norm": 0.29717737436294556, "learning_rate": 8.385991420581058e-06, "loss": 0.0265, "step": 49340 }, { "grad_norm": 0.38433414697647095, "learning_rate": 8.370720914747215e-06, "loss": 0.02, "step": 49350 }, { "grad_norm": 0.4173760414123535, "learning_rate": 8.355463054789181e-06, "loss": 0.0263, "step": 49360 }, { "grad_norm": 0.3631126582622528, "learning_rate": 8.340217845341919e-06, "loss": 0.0237, "step": 49370 }, { "grad_norm": 0.40685248374938965, "learning_rate": 8.324985291036514e-06, "loss": 0.0289, "step": 49380 }, { "grad_norm": 0.5810068249702454, "learning_rate": 8.309765396500213e-06, "loss": 0.0271, "step": 49390 }, { "grad_norm": 0.3602977395057678, "learning_rate": 8.294558166356419e-06, "loss": 0.0262, "step": 49400 }, { "grad_norm": 0.3364430367946625, "learning_rate": 8.279363605224683e-06, "loss": 0.0306, "step": 49410 }, { "grad_norm": 0.5607398748397827, "learning_rate": 8.264181717720704e-06, "loss": 0.0336, "step": 49420 }, { "grad_norm": 0.7707685232162476, "learning_rate": 8.249012508456361e-06, "loss": 0.0314, "step": 49430 }, { "grad_norm": 0.6954454183578491, "learning_rate": 8.233855982039646e-06, "loss": 0.0333, "step": 49440 }, { "grad_norm": 0.3519316017627716, "learning_rate": 8.218712143074708e-06, "loss": 0.0289, "step": 49450 }, { "grad_norm": 0.5635219216346741, "learning_rate": 8.203580996161858e-06, "loss": 0.0333, "step": 49460 }, { "grad_norm": 0.553232729434967, "learning_rate": 8.188462545897512e-06, "loss": 0.0256, "step": 49470 }, { "grad_norm": 0.43845394253730774, "learning_rate": 8.173356796874304e-06, "loss": 0.0292, "step": 49480 }, { "grad_norm": 0.3840910494327545, "learning_rate": 8.158263753680906e-06, "loss": 0.0361, "step": 49490 }, { "grad_norm": 0.5067612528800964, "learning_rate": 8.143183420902239e-06, "loss": 0.0183, "step": 49500 }, { "grad_norm": 0.756027102470398, "learning_rate": 8.128115803119258e-06, "loss": 0.0306, "step": 49510 }, { "grad_norm": 0.42868903279304504, "learning_rate": 8.11306090490916e-06, "loss": 0.0308, "step": 49520 }, { "grad_norm": 0.5031067728996277, "learning_rate": 8.098018730845169e-06, "loss": 0.0222, "step": 49530 }, { "grad_norm": 0.308216392993927, "learning_rate": 8.082989285496745e-06, "loss": 0.0318, "step": 49540 }, { "grad_norm": 0.436741441488266, "learning_rate": 8.067972573429416e-06, "loss": 0.0313, "step": 49550 }, { "grad_norm": 0.4757033884525299, "learning_rate": 8.052968599204874e-06, "loss": 0.0285, "step": 49560 }, { "grad_norm": 0.40989741683006287, "learning_rate": 8.037977367380922e-06, "loss": 0.0186, "step": 49570 }, { "grad_norm": 0.3936166763305664, "learning_rate": 8.022998882511495e-06, "loss": 0.0202, "step": 49580 }, { "grad_norm": 0.5815455317497253, "learning_rate": 8.008033149146677e-06, "loss": 0.0213, "step": 49590 }, { "grad_norm": 0.1984187215566635, "learning_rate": 7.993080171832656e-06, "loss": 0.0235, "step": 49600 }, { "grad_norm": 0.30706197023391724, "learning_rate": 7.978139955111752e-06, "loss": 0.0192, "step": 49610 }, { "grad_norm": 0.2970035970211029, "learning_rate": 7.9632125035224e-06, "loss": 0.0242, "step": 49620 }, { "grad_norm": 0.37719136476516724, "learning_rate": 7.948297821599177e-06, "loss": 0.0277, "step": 49630 }, { "grad_norm": 0.1701265275478363, "learning_rate": 7.933395913872755e-06, "loss": 0.0224, "step": 49640 }, { "grad_norm": 0.6663338541984558, "learning_rate": 7.918506784869972e-06, "loss": 0.0351, "step": 49650 }, { "grad_norm": 0.2363569736480713, "learning_rate": 7.903630439113707e-06, "loss": 0.0169, "step": 49660 }, { "grad_norm": 0.43038344383239746, "learning_rate": 7.888766881123044e-06, "loss": 0.02, "step": 49670 }, { "grad_norm": 0.7997317910194397, "learning_rate": 7.873916115413099e-06, "loss": 0.0324, "step": 49680 }, { "grad_norm": 0.46335816383361816, "learning_rate": 7.85907814649518e-06, "loss": 0.0213, "step": 49690 }, { "grad_norm": 0.3627558648586273, "learning_rate": 7.844252978876649e-06, "loss": 0.0216, "step": 49700 }, { "grad_norm": 0.33350706100463867, "learning_rate": 7.829440617061001e-06, "loss": 0.0295, "step": 49710 }, { "grad_norm": 0.369144082069397, "learning_rate": 7.814641065547851e-06, "loss": 0.0153, "step": 49720 }, { "grad_norm": 0.5500075817108154, "learning_rate": 7.79985432883289e-06, "loss": 0.0214, "step": 49730 }, { "grad_norm": 0.42377522587776184, "learning_rate": 7.78508041140797e-06, "loss": 0.0346, "step": 49740 }, { "grad_norm": 0.48618391156196594, "learning_rate": 7.770319317760993e-06, "loss": 0.0262, "step": 49750 }, { "grad_norm": 0.43093082308769226, "learning_rate": 7.755571052376004e-06, "loss": 0.0247, "step": 49760 }, { "grad_norm": 0.2845686674118042, "learning_rate": 7.740835619733128e-06, "loss": 0.0301, "step": 49770 }, { "grad_norm": 0.35350245237350464, "learning_rate": 7.726113024308601e-06, "loss": 0.0236, "step": 49780 }, { "grad_norm": 0.4634191393852234, "learning_rate": 7.711403270574746e-06, "loss": 0.0274, "step": 49790 }, { "grad_norm": 0.3775688409805298, "learning_rate": 7.696706363000039e-06, "loss": 0.0322, "step": 49800 }, { "grad_norm": 0.5351905822753906, "learning_rate": 7.682022306048959e-06, "loss": 0.0232, "step": 49810 }, { "grad_norm": 0.43599480390548706, "learning_rate": 7.667351104182186e-06, "loss": 0.0221, "step": 49820 }, { "grad_norm": 0.5332363247871399, "learning_rate": 7.652692761856395e-06, "loss": 0.0222, "step": 49830 }, { "grad_norm": 0.22457805275917053, "learning_rate": 7.63804728352444e-06, "loss": 0.0233, "step": 49840 }, { "grad_norm": 0.3673788607120514, "learning_rate": 7.623414673635215e-06, "loss": 0.0345, "step": 49850 }, { "grad_norm": 0.5624489784240723, "learning_rate": 7.608794936633723e-06, "loss": 0.024, "step": 49860 }, { "grad_norm": 0.5175144672393799, "learning_rate": 7.594188076961056e-06, "loss": 0.0219, "step": 49870 }, { "grad_norm": 0.45995038747787476, "learning_rate": 7.579594099054382e-06, "loss": 0.0266, "step": 49880 }, { "grad_norm": 0.3159193992614746, "learning_rate": 7.565013007346983e-06, "loss": 0.0212, "step": 49890 }, { "grad_norm": 0.14383386075496674, "learning_rate": 7.5504448062682035e-06, "loss": 0.0283, "step": 49900 }, { "grad_norm": 0.31020069122314453, "learning_rate": 7.53588950024347e-06, "loss": 0.0172, "step": 49910 }, { "grad_norm": 0.42112329602241516, "learning_rate": 7.5213470936943145e-06, "loss": 0.0193, "step": 49920 }, { "grad_norm": 0.4171891212463379, "learning_rate": 7.506817591038323e-06, "loss": 0.023, "step": 49930 }, { "grad_norm": 0.3876083195209503, "learning_rate": 7.492300996689183e-06, "loss": 0.0358, "step": 49940 }, { "grad_norm": 0.4977351129055023, "learning_rate": 7.477797315056645e-06, "loss": 0.0318, "step": 49950 }, { "grad_norm": 0.5487213134765625, "learning_rate": 7.463306550546539e-06, "loss": 0.0203, "step": 49960 }, { "grad_norm": 0.5872397422790527, "learning_rate": 7.448828707560812e-06, "loss": 0.0276, "step": 49970 }, { "grad_norm": 0.4300227761268616, "learning_rate": 7.4343637904974e-06, "loss": 0.0504, "step": 49980 }, { "grad_norm": 0.3312123417854309, "learning_rate": 7.419911803750401e-06, "loss": 0.0296, "step": 49990 }, { "grad_norm": 0.5960912704467773, "learning_rate": 7.405472751709935e-06, "loss": 0.0212, "step": 50000 }, { "grad_norm": 0.600159227848053, "learning_rate": 7.3910466387622e-06, "loss": 0.0311, "step": 50010 }, { "grad_norm": 0.43314462900161743, "learning_rate": 7.3766334692894735e-06, "loss": 0.0257, "step": 50020 }, { "grad_norm": 0.44191810488700867, "learning_rate": 7.3622332476700865e-06, "loss": 0.0205, "step": 50030 }, { "grad_norm": 0.18571363389492035, "learning_rate": 7.347845978278472e-06, "loss": 0.0253, "step": 50040 }, { "grad_norm": 0.3011883795261383, "learning_rate": 7.333471665485065e-06, "loss": 0.0186, "step": 50050 }, { "grad_norm": 0.6800824403762817, "learning_rate": 7.31911031365643e-06, "loss": 0.042, "step": 50060 }, { "grad_norm": 0.5583611130714417, "learning_rate": 7.304761927155157e-06, "loss": 0.0261, "step": 50070 }, { "grad_norm": 0.2361316829919815, "learning_rate": 7.29042651033991e-06, "loss": 0.0264, "step": 50080 }, { "grad_norm": 0.30315759778022766, "learning_rate": 7.276104067565409e-06, "loss": 0.0261, "step": 50090 }, { "grad_norm": 0.2831912934780121, "learning_rate": 7.261794603182431e-06, "loss": 0.0248, "step": 50100 }, { "grad_norm": 0.5362198948860168, "learning_rate": 7.24749812153781e-06, "loss": 0.0232, "step": 50110 }, { "grad_norm": 0.38333451747894287, "learning_rate": 7.2332146269744605e-06, "loss": 0.0313, "step": 50120 }, { "grad_norm": 0.4872433543205261, "learning_rate": 7.218944123831295e-06, "loss": 0.025, "step": 50130 }, { "grad_norm": 0.3772129416465759, "learning_rate": 7.204686616443351e-06, "loss": 0.0284, "step": 50140 }, { "grad_norm": 0.2507937252521515, "learning_rate": 7.190442109141665e-06, "loss": 0.022, "step": 50150 }, { "grad_norm": 0.3018864691257477, "learning_rate": 7.176210606253347e-06, "loss": 0.0287, "step": 50160 }, { "grad_norm": 0.4469027519226074, "learning_rate": 7.161992112101551e-06, "loss": 0.0214, "step": 50170 }, { "grad_norm": 0.6086962223052979, "learning_rate": 7.147786631005465e-06, "loss": 0.0277, "step": 50180 }, { "grad_norm": 0.34979483485221863, "learning_rate": 7.1335941672803775e-06, "loss": 0.0258, "step": 50190 }, { "grad_norm": 0.27279379963874817, "learning_rate": 7.1194147252375384e-06, "loss": 0.0315, "step": 50200 }, { "grad_norm": 0.29648786783218384, "learning_rate": 7.10524830918432e-06, "loss": 0.0212, "step": 50210 }, { "grad_norm": 0.49933773279190063, "learning_rate": 7.091094923424097e-06, "loss": 0.0241, "step": 50220 }, { "grad_norm": 0.3822304904460907, "learning_rate": 7.0769545722562894e-06, "loss": 0.0272, "step": 50230 }, { "grad_norm": 0.4401043951511383, "learning_rate": 7.0628272599763675e-06, "loss": 0.0292, "step": 50240 }, { "grad_norm": 0.43393439054489136, "learning_rate": 7.048712990875828e-06, "loss": 0.0431, "step": 50250 }, { "grad_norm": 0.18261171877384186, "learning_rate": 7.034611769242216e-06, "loss": 0.0172, "step": 50260 }, { "grad_norm": 0.29982540011405945, "learning_rate": 7.02052359935913e-06, "loss": 0.0249, "step": 50270 }, { "grad_norm": 0.5565882921218872, "learning_rate": 7.006448485506145e-06, "loss": 0.0267, "step": 50280 }, { "grad_norm": 0.35434022545814514, "learning_rate": 6.992386431958942e-06, "loss": 0.0223, "step": 50290 }, { "grad_norm": 0.42475906014442444, "learning_rate": 6.978337442989197e-06, "loss": 0.0196, "step": 50300 }, { "grad_norm": 0.3792400062084198, "learning_rate": 6.964301522864608e-06, "loss": 0.0235, "step": 50310 }, { "grad_norm": 0.37136369943618774, "learning_rate": 6.950278675848926e-06, "loss": 0.019, "step": 50320 }, { "grad_norm": 0.40304869413375854, "learning_rate": 6.9362689062019145e-06, "loss": 0.0216, "step": 50330 }, { "grad_norm": 0.3319258689880371, "learning_rate": 6.922272218179393e-06, "loss": 0.0185, "step": 50340 }, { "grad_norm": 0.48028913140296936, "learning_rate": 6.908288616033148e-06, "loss": 0.0347, "step": 50350 }, { "grad_norm": 0.17545555531978607, "learning_rate": 6.894318104011077e-06, "loss": 0.0199, "step": 50360 }, { "grad_norm": 0.45978501439094543, "learning_rate": 6.880360686357007e-06, "loss": 0.0307, "step": 50370 }, { "grad_norm": 0.5215505361557007, "learning_rate": 6.8664163673108575e-06, "loss": 0.0214, "step": 50380 }, { "grad_norm": 0.4404781758785248, "learning_rate": 6.85248515110854e-06, "loss": 0.0214, "step": 50390 }, { "grad_norm": 0.25518137216567993, "learning_rate": 6.838567041981992e-06, "loss": 0.0174, "step": 50400 }, { "grad_norm": 0.30784928798675537, "learning_rate": 6.8246620441591634e-06, "loss": 0.0267, "step": 50410 }, { "grad_norm": 0.28752973675727844, "learning_rate": 6.8107701618640275e-06, "loss": 0.0217, "step": 50420 }, { "grad_norm": 0.36609524488449097, "learning_rate": 6.796891399316557e-06, "loss": 0.0271, "step": 50430 }, { "grad_norm": 0.4948812425136566, "learning_rate": 6.7830257607327804e-06, "loss": 0.0191, "step": 50440 }, { "grad_norm": 0.5821401476860046, "learning_rate": 6.7691732503247e-06, "loss": 0.0332, "step": 50450 }, { "grad_norm": 0.5319361090660095, "learning_rate": 6.755333872300346e-06, "loss": 0.0312, "step": 50460 }, { "grad_norm": 0.35084739327430725, "learning_rate": 6.741507630863747e-06, "loss": 0.0207, "step": 50470 }, { "grad_norm": 0.4653893709182739, "learning_rate": 6.727694530214945e-06, "loss": 0.0192, "step": 50480 }, { "grad_norm": 0.48621320724487305, "learning_rate": 6.713894574550028e-06, "loss": 0.0253, "step": 50490 }, { "grad_norm": 0.305934339761734, "learning_rate": 6.700107768061015e-06, "loss": 0.018, "step": 50500 }, { "grad_norm": 0.19992901384830475, "learning_rate": 6.686334114936016e-06, "loss": 0.0233, "step": 50510 }, { "grad_norm": 0.2331731766462326, "learning_rate": 6.672573619359063e-06, "loss": 0.0203, "step": 50520 }, { "grad_norm": 0.2215956449508667, "learning_rate": 6.658826285510256e-06, "loss": 0.0137, "step": 50530 }, { "grad_norm": 0.41134580969810486, "learning_rate": 6.645092117565666e-06, "loss": 0.03, "step": 50540 }, { "grad_norm": 0.6159445643424988, "learning_rate": 6.631371119697371e-06, "loss": 0.028, "step": 50550 }, { "grad_norm": 0.28311896324157715, "learning_rate": 6.6176632960734505e-06, "loss": 0.021, "step": 50560 }, { "grad_norm": 0.3406846821308136, "learning_rate": 6.603968650857978e-06, "loss": 0.0189, "step": 50570 }, { "grad_norm": 0.4195484220981598, "learning_rate": 6.5902871882110085e-06, "loss": 0.0206, "step": 50580 }, { "grad_norm": 0.19241562485694885, "learning_rate": 6.576618912288635e-06, "loss": 0.0283, "step": 50590 }, { "grad_norm": 0.2965633273124695, "learning_rate": 6.562963827242913e-06, "loss": 0.0234, "step": 50600 }, { "grad_norm": 0.3581605851650238, "learning_rate": 6.549321937221886e-06, "loss": 0.0221, "step": 50610 }, { "grad_norm": 0.6992197036743164, "learning_rate": 6.5356932463696064e-06, "loss": 0.0295, "step": 50620 }, { "grad_norm": 0.41429612040519714, "learning_rate": 6.522077758826101e-06, "loss": 0.0207, "step": 50630 }, { "grad_norm": 0.3529225289821625, "learning_rate": 6.5084754787274275e-06, "loss": 0.0258, "step": 50640 }, { "grad_norm": 0.43208032846450806, "learning_rate": 6.494886410205553e-06, "loss": 0.0388, "step": 50650 }, { "grad_norm": 0.39940106868743896, "learning_rate": 6.481310557388521e-06, "loss": 0.0171, "step": 50660 }, { "grad_norm": 0.6251272559165955, "learning_rate": 6.46774792440028e-06, "loss": 0.0273, "step": 50670 }, { "grad_norm": 0.4841625392436981, "learning_rate": 6.4541985153608206e-06, "loss": 0.0187, "step": 50680 }, { "grad_norm": 0.381478488445282, "learning_rate": 6.4406623343861e-06, "loss": 0.0175, "step": 50690 }, { "grad_norm": 0.31051915884017944, "learning_rate": 6.427139385588038e-06, "loss": 0.0229, "step": 50700 }, { "grad_norm": 0.42131102085113525, "learning_rate": 6.413629673074561e-06, "loss": 0.0211, "step": 50710 }, { "grad_norm": 0.5039548873901367, "learning_rate": 6.400133200949554e-06, "loss": 0.0183, "step": 50720 }, { "grad_norm": 0.5012532472610474, "learning_rate": 6.386649973312897e-06, "loss": 0.0344, "step": 50730 }, { "grad_norm": 0.44857993721961975, "learning_rate": 6.37317999426042e-06, "loss": 0.021, "step": 50740 }, { "grad_norm": 0.6485424637794495, "learning_rate": 6.359723267883977e-06, "loss": 0.0261, "step": 50750 }, { "grad_norm": 0.15428178012371063, "learning_rate": 6.346279798271343e-06, "loss": 0.0145, "step": 50760 }, { "grad_norm": 0.29954442381858826, "learning_rate": 6.332849589506301e-06, "loss": 0.0235, "step": 50770 }, { "grad_norm": 0.29200509190559387, "learning_rate": 6.319432645668588e-06, "loss": 0.0265, "step": 50780 }, { "grad_norm": 0.5849834084510803, "learning_rate": 6.306028970833922e-06, "loss": 0.0225, "step": 50790 }, { "grad_norm": 0.6403789520263672, "learning_rate": 6.2926385690739665e-06, "loss": 0.0292, "step": 50800 }, { "grad_norm": 0.4022590219974518, "learning_rate": 6.279261444456413e-06, "loss": 0.0319, "step": 50810 }, { "grad_norm": 0.4706575572490692, "learning_rate": 6.265897601044829e-06, "loss": 0.0224, "step": 50820 }, { "grad_norm": 0.2602421045303345, "learning_rate": 6.2525470428988434e-06, "loss": 0.0136, "step": 50830 }, { "grad_norm": 0.39175814390182495, "learning_rate": 6.239209774073962e-06, "loss": 0.0211, "step": 50840 }, { "grad_norm": 0.37078729271888733, "learning_rate": 6.225885798621728e-06, "loss": 0.0245, "step": 50850 }, { "grad_norm": 0.27803802490234375, "learning_rate": 6.2125751205895925e-06, "loss": 0.027, "step": 50860 }, { "grad_norm": 1.2686492204666138, "learning_rate": 6.199277744020998e-06, "loss": 0.0299, "step": 50870 }, { "grad_norm": 0.2786586880683899, "learning_rate": 6.185993672955337e-06, "loss": 0.0172, "step": 50880 }, { "grad_norm": 0.2762529253959656, "learning_rate": 6.172722911427947e-06, "loss": 0.0423, "step": 50890 }, { "grad_norm": 0.3332161009311676, "learning_rate": 6.159465463470149e-06, "loss": 0.0216, "step": 50900 }, { "grad_norm": 0.21727503836154938, "learning_rate": 6.146221333109204e-06, "loss": 0.0211, "step": 50910 }, { "grad_norm": 0.4441205561161041, "learning_rate": 6.132990524368326e-06, "loss": 0.0221, "step": 50920 }, { "grad_norm": 0.29594600200653076, "learning_rate": 6.119773041266685e-06, "loss": 0.0299, "step": 50930 }, { "grad_norm": 0.33684301376342773, "learning_rate": 6.106568887819402e-06, "loss": 0.0189, "step": 50940 }, { "grad_norm": 0.42706504464149475, "learning_rate": 6.093378068037548e-06, "loss": 0.0339, "step": 50950 }, { "grad_norm": 0.81296306848526, "learning_rate": 6.080200585928164e-06, "loss": 0.0308, "step": 50960 }, { "grad_norm": 0.2810746431350708, "learning_rate": 6.06703644549419e-06, "loss": 0.0249, "step": 50970 }, { "grad_norm": 0.5165730118751526, "learning_rate": 6.053885650734576e-06, "loss": 0.0282, "step": 50980 }, { "grad_norm": 0.44397738575935364, "learning_rate": 6.040748205644153e-06, "loss": 0.0282, "step": 50990 }, { "grad_norm": 0.4543829560279846, "learning_rate": 6.0276241142137646e-06, "loss": 0.0393, "step": 51000 }, { "grad_norm": 0.568985641002655, "learning_rate": 6.014513380430142e-06, "loss": 0.0393, "step": 51010 }, { "grad_norm": 0.2594394385814667, "learning_rate": 6.001416008275984e-06, "loss": 0.0201, "step": 51020 }, { "grad_norm": 2.096857786178589, "learning_rate": 5.988332001729929e-06, "loss": 0.0245, "step": 51030 }, { "grad_norm": 0.9351502656936646, "learning_rate": 5.975261364766543e-06, "loss": 0.0284, "step": 51040 }, { "grad_norm": 0.6476553678512573, "learning_rate": 5.962204101356356e-06, "loss": 0.0297, "step": 51050 }, { "grad_norm": 0.3234396278858185, "learning_rate": 5.94916021546581e-06, "loss": 0.0268, "step": 51060 }, { "grad_norm": 0.3457709550857544, "learning_rate": 5.936129711057298e-06, "loss": 0.0215, "step": 51070 }, { "grad_norm": 0.5550572872161865, "learning_rate": 5.923112592089142e-06, "loss": 0.0283, "step": 51080 }, { "grad_norm": 0.8011764287948608, "learning_rate": 5.9101088625155954e-06, "loss": 0.025, "step": 51090 }, { "grad_norm": 0.4515610933303833, "learning_rate": 5.897118526286843e-06, "loss": 0.024, "step": 51100 }, { "grad_norm": 0.5303567051887512, "learning_rate": 5.884141587349035e-06, "loss": 0.0237, "step": 51110 }, { "grad_norm": 0.27854758501052856, "learning_rate": 5.871178049644177e-06, "loss": 0.0254, "step": 51120 }, { "grad_norm": 0.39999836683273315, "learning_rate": 5.858227917110293e-06, "loss": 0.019, "step": 51130 }, { "grad_norm": 0.36905917525291443, "learning_rate": 5.845291193681252e-06, "loss": 0.0229, "step": 51140 }, { "grad_norm": 0.24433720111846924, "learning_rate": 5.832367883286921e-06, "loss": 0.021, "step": 51150 }, { "grad_norm": 0.27378538250923157, "learning_rate": 5.81945798985305e-06, "loss": 0.0241, "step": 51160 }, { "grad_norm": 0.5394129753112793, "learning_rate": 5.806561517301306e-06, "loss": 0.0232, "step": 51170 }, { "grad_norm": 0.3026033043861389, "learning_rate": 5.793678469549335e-06, "loss": 0.0288, "step": 51180 }, { "grad_norm": 0.3154599964618683, "learning_rate": 5.780808850510627e-06, "loss": 0.0183, "step": 51190 }, { "grad_norm": 0.3567982017993927, "learning_rate": 5.767952664094673e-06, "loss": 0.0343, "step": 51200 }, { "grad_norm": 0.256614089012146, "learning_rate": 5.755109914206791e-06, "loss": 0.0254, "step": 51210 }, { "grad_norm": 0.2774348258972168, "learning_rate": 5.7422806047483125e-06, "loss": 0.0213, "step": 51220 }, { "grad_norm": 0.7253158092498779, "learning_rate": 5.72946473961643e-06, "loss": 0.0326, "step": 51230 }, { "grad_norm": 0.49018749594688416, "learning_rate": 5.716662322704264e-06, "loss": 0.0198, "step": 51240 }, { "grad_norm": 0.48297375440597534, "learning_rate": 5.703873357900852e-06, "loss": 0.0181, "step": 51250 }, { "grad_norm": 0.45372042059898376, "learning_rate": 5.691097849091143e-06, "loss": 0.0246, "step": 51260 }, { "grad_norm": 0.36673063039779663, "learning_rate": 5.678335800155982e-06, "loss": 0.0279, "step": 51270 }, { "grad_norm": 0.4411379396915436, "learning_rate": 5.665587214972174e-06, "loss": 0.0186, "step": 51280 }, { "grad_norm": 0.7353156208992004, "learning_rate": 5.652852097412386e-06, "loss": 0.0201, "step": 51290 }, { "grad_norm": 0.5483447909355164, "learning_rate": 5.640130451345216e-06, "loss": 0.0343, "step": 51300 }, { "grad_norm": 0.2733210325241089, "learning_rate": 5.627422280635159e-06, "loss": 0.0194, "step": 51310 }, { "grad_norm": 0.18306118249893188, "learning_rate": 5.61472758914261e-06, "loss": 0.018, "step": 51320 }, { "grad_norm": 0.5306231379508972, "learning_rate": 5.602046380723918e-06, "loss": 0.0325, "step": 51330 }, { "grad_norm": 0.2830785810947418, "learning_rate": 5.5893786592312535e-06, "loss": 0.0248, "step": 51340 }, { "grad_norm": 0.335852712392807, "learning_rate": 5.576724428512775e-06, "loss": 0.0175, "step": 51350 }, { "grad_norm": 0.23473772406578064, "learning_rate": 5.56408369241247e-06, "loss": 0.0215, "step": 51360 }, { "grad_norm": 0.2232757806777954, "learning_rate": 5.5514564547702875e-06, "loss": 0.0185, "step": 51370 }, { "grad_norm": 0.32665470242500305, "learning_rate": 5.538842719422038e-06, "loss": 0.0184, "step": 51380 }, { "grad_norm": 0.4267584979534149, "learning_rate": 5.52624249019944e-06, "loss": 0.0213, "step": 51390 }, { "grad_norm": 0.39115825295448303, "learning_rate": 5.51365577093011e-06, "loss": 0.021, "step": 51400 }, { "grad_norm": 0.5123347640037537, "learning_rate": 5.501082565437565e-06, "loss": 0.02, "step": 51410 }, { "grad_norm": 0.40581822395324707, "learning_rate": 5.488522877541202e-06, "loss": 0.0249, "step": 51420 }, { "grad_norm": 0.26291313767433167, "learning_rate": 5.475976711056341e-06, "loss": 0.0217, "step": 51430 }, { "grad_norm": 0.2941860854625702, "learning_rate": 5.463444069794166e-06, "loss": 0.0224, "step": 51440 }, { "grad_norm": 0.38324496150016785, "learning_rate": 5.4509249575617594e-06, "loss": 0.0333, "step": 51450 }, { "grad_norm": 0.2496955692768097, "learning_rate": 5.438419378162107e-06, "loss": 0.0217, "step": 51460 }, { "grad_norm": 0.32230451703071594, "learning_rate": 5.425927335394054e-06, "loss": 0.0232, "step": 51470 }, { "grad_norm": 0.48606032133102417, "learning_rate": 5.413448833052387e-06, "loss": 0.0179, "step": 51480 }, { "grad_norm": 1.2687914371490479, "learning_rate": 5.400983874927701e-06, "loss": 0.0236, "step": 51490 }, { "grad_norm": 0.5389137268066406, "learning_rate": 5.388532464806567e-06, "loss": 0.0206, "step": 51500 }, { "grad_norm": 0.484541654586792, "learning_rate": 5.3760946064713546e-06, "loss": 0.0231, "step": 51510 }, { "grad_norm": 0.33314377069473267, "learning_rate": 5.363670303700386e-06, "loss": 0.0219, "step": 51520 }, { "grad_norm": 0.421798437833786, "learning_rate": 5.351259560267824e-06, "loss": 0.0326, "step": 51530 }, { "grad_norm": 0.541917622089386, "learning_rate": 5.338862379943721e-06, "loss": 0.0201, "step": 51540 }, { "grad_norm": 0.197288379073143, "learning_rate": 5.326478766494025e-06, "loss": 0.017, "step": 51550 }, { "grad_norm": 0.3512822985649109, "learning_rate": 5.3141087236805385e-06, "loss": 0.0226, "step": 51560 }, { "grad_norm": 0.44830521941185, "learning_rate": 5.3017522552609615e-06, "loss": 0.0273, "step": 51570 }, { "grad_norm": 0.29946860671043396, "learning_rate": 5.289409364988851e-06, "loss": 0.0186, "step": 51580 }, { "grad_norm": 0.43558359146118164, "learning_rate": 5.277080056613671e-06, "loss": 0.0226, "step": 51590 }, { "grad_norm": 0.4765932261943817, "learning_rate": 5.264764333880729e-06, "loss": 0.0246, "step": 51600 }, { "grad_norm": 0.7219030857086182, "learning_rate": 5.252462200531216e-06, "loss": 0.0232, "step": 51610 }, { "grad_norm": 0.278778612613678, "learning_rate": 5.240173660302194e-06, "loss": 0.0175, "step": 51620 }, { "grad_norm": 0.5416694283485413, "learning_rate": 5.2278987169266044e-06, "loss": 0.0174, "step": 51630 }, { "grad_norm": 0.14574265480041504, "learning_rate": 5.215637374133231e-06, "loss": 0.0147, "step": 51640 }, { "grad_norm": 0.3622649908065796, "learning_rate": 5.203389635646782e-06, "loss": 0.0291, "step": 51650 }, { "grad_norm": 0.20143835246562958, "learning_rate": 5.191155505187756e-06, "loss": 0.0197, "step": 51660 }, { "grad_norm": 0.47954028844833374, "learning_rate": 5.1789349864726e-06, "loss": 0.0331, "step": 51670 }, { "grad_norm": 0.3143613636493683, "learning_rate": 5.16672808321354e-06, "loss": 0.0294, "step": 51680 }, { "grad_norm": 0.38865506649017334, "learning_rate": 5.154534799118749e-06, "loss": 0.0318, "step": 51690 }, { "grad_norm": 0.23062625527381897, "learning_rate": 5.142355137892207e-06, "loss": 0.0186, "step": 51700 }, { "grad_norm": 0.6277095079421997, "learning_rate": 5.130189103233779e-06, "loss": 0.018, "step": 51710 }, { "grad_norm": 0.358280748128891, "learning_rate": 5.118036698839179e-06, "loss": 0.0178, "step": 51720 }, { "grad_norm": 0.6021629571914673, "learning_rate": 5.105897928399983e-06, "loss": 0.026, "step": 51730 }, { "grad_norm": 0.34830841422080994, "learning_rate": 5.0937727956036464e-06, "loss": 0.0269, "step": 51740 }, { "grad_norm": 0.39894309639930725, "learning_rate": 5.081661304133456e-06, "loss": 0.0305, "step": 51750 }, { "grad_norm": 0.4034663438796997, "learning_rate": 5.069563457668558e-06, "loss": 0.0301, "step": 51760 }, { "grad_norm": 0.39972230792045593, "learning_rate": 5.0574792598839624e-06, "loss": 0.0225, "step": 51770 }, { "grad_norm": 0.19977086782455444, "learning_rate": 5.0454087144505276e-06, "loss": 0.0269, "step": 51780 }, { "grad_norm": 0.46915507316589355, "learning_rate": 5.0333518250349655e-06, "loss": 0.0237, "step": 51790 }, { "grad_norm": 0.21871349215507507, "learning_rate": 5.021308595299856e-06, "loss": 0.017, "step": 51800 }, { "grad_norm": 0.12841562926769257, "learning_rate": 5.009279028903585e-06, "loss": 0.0241, "step": 51810 }, { "grad_norm": 0.45912089943885803, "learning_rate": 4.997263129500452e-06, "loss": 0.0276, "step": 51820 }, { "grad_norm": 0.19196148216724396, "learning_rate": 4.985260900740535e-06, "loss": 0.0298, "step": 51830 }, { "grad_norm": 0.4850276708602905, "learning_rate": 4.973272346269814e-06, "loss": 0.027, "step": 51840 }, { "grad_norm": 0.21753139793872833, "learning_rate": 4.961297469730097e-06, "loss": 0.0252, "step": 51850 }, { "grad_norm": 0.35957515239715576, "learning_rate": 4.949336274759031e-06, "loss": 0.0263, "step": 51860 }, { "grad_norm": 0.37513816356658936, "learning_rate": 4.9373887649901144e-06, "loss": 0.0232, "step": 51870 }, { "grad_norm": 0.3688625991344452, "learning_rate": 4.925454944052666e-06, "loss": 0.0198, "step": 51880 }, { "grad_norm": 0.6594316363334656, "learning_rate": 4.913534815571891e-06, "loss": 0.0251, "step": 51890 }, { "grad_norm": 0.3482719659805298, "learning_rate": 4.901628383168805e-06, "loss": 0.0194, "step": 51900 }, { "grad_norm": 0.4589790105819702, "learning_rate": 4.8897356504602585e-06, "loss": 0.0306, "step": 51910 }, { "grad_norm": 0.28353193402290344, "learning_rate": 4.877856621058957e-06, "loss": 0.0319, "step": 51920 }, { "grad_norm": 0.430541455745697, "learning_rate": 4.86599129857343e-06, "loss": 0.0244, "step": 51930 }, { "grad_norm": 0.3001212477684021, "learning_rate": 4.85413968660805e-06, "loss": 0.0235, "step": 51940 }, { "grad_norm": 0.5724522471427917, "learning_rate": 4.842301788763031e-06, "loss": 0.0226, "step": 51950 }, { "grad_norm": 0.2979969382286072, "learning_rate": 4.830477608634393e-06, "loss": 0.0357, "step": 51960 }, { "grad_norm": 0.33818426728248596, "learning_rate": 4.818667149814049e-06, "loss": 0.0208, "step": 51970 }, { "grad_norm": 0.30510154366493225, "learning_rate": 4.806870415889664e-06, "loss": 0.0269, "step": 51980 }, { "grad_norm": 0.18396900594234467, "learning_rate": 4.795087410444798e-06, "loss": 0.0243, "step": 51990 }, { "grad_norm": 0.335938423871994, "learning_rate": 4.783318137058807e-06, "loss": 0.0335, "step": 52000 }, { "grad_norm": 0.31107470393180847, "learning_rate": 4.771562599306895e-06, "loss": 0.0259, "step": 52010 }, { "grad_norm": 0.48065105080604553, "learning_rate": 4.759820800760073e-06, "loss": 0.0437, "step": 52020 }, { "grad_norm": 0.60529625415802, "learning_rate": 4.7480927449851834e-06, "loss": 0.0246, "step": 52030 }, { "grad_norm": 0.5353135466575623, "learning_rate": 4.73637843554493e-06, "loss": 0.0271, "step": 52040 }, { "grad_norm": 0.1995772272348404, "learning_rate": 4.724677875997774e-06, "loss": 0.0195, "step": 52050 }, { "grad_norm": 0.2898903787136078, "learning_rate": 4.712991069898065e-06, "loss": 0.0218, "step": 52060 }, { "grad_norm": 0.4630117118358612, "learning_rate": 4.7013180207959305e-06, "loss": 0.0367, "step": 52070 }, { "grad_norm": 0.3087904453277588, "learning_rate": 4.6896587322373395e-06, "loss": 0.0244, "step": 52080 }, { "grad_norm": 0.506603479385376, "learning_rate": 4.678013207764081e-06, "loss": 0.0302, "step": 52090 }, { "grad_norm": 0.6856614947319031, "learning_rate": 4.666381450913748e-06, "loss": 0.0262, "step": 52100 }, { "grad_norm": 0.4897254407405853, "learning_rate": 4.654763465219752e-06, "loss": 0.0255, "step": 52110 }, { "grad_norm": 0.3104914426803589, "learning_rate": 4.643159254211371e-06, "loss": 0.0259, "step": 52120 }, { "grad_norm": 0.42159026861190796, "learning_rate": 4.631568821413606e-06, "loss": 0.0374, "step": 52130 }, { "grad_norm": 0.3537905216217041, "learning_rate": 4.619992170347359e-06, "loss": 0.0351, "step": 52140 }, { "grad_norm": 0.5249487161636353, "learning_rate": 4.608429304529305e-06, "loss": 0.0255, "step": 52150 }, { "grad_norm": 0.26823174953460693, "learning_rate": 4.596880227471928e-06, "loss": 0.0477, "step": 52160 }, { "grad_norm": 0.4309803545475006, "learning_rate": 4.585344942683539e-06, "loss": 0.0251, "step": 52170 }, { "grad_norm": 0.34559303522109985, "learning_rate": 4.573823453668241e-06, "loss": 0.0279, "step": 52180 }, { "grad_norm": 0.3353012502193451, "learning_rate": 4.562315763925995e-06, "loss": 0.0253, "step": 52190 }, { "grad_norm": 0.6594244837760925, "learning_rate": 4.5508218769524825e-06, "loss": 0.0304, "step": 52200 }, { "grad_norm": 0.427995890378952, "learning_rate": 4.539341796239277e-06, "loss": 0.029, "step": 52210 }, { "grad_norm": 0.3170609772205353, "learning_rate": 4.527875525273717e-06, "loss": 0.0204, "step": 52220 }, { "grad_norm": 0.33059823513031006, "learning_rate": 4.51642306753895e-06, "loss": 0.0225, "step": 52230 }, { "grad_norm": 0.4113222062587738, "learning_rate": 4.5049844265139306e-06, "loss": 0.0259, "step": 52240 }, { "grad_norm": 0.40988215804100037, "learning_rate": 4.4935596056734144e-06, "loss": 0.0299, "step": 52250 }, { "grad_norm": 0.44306454062461853, "learning_rate": 4.482148608487957e-06, "loss": 0.0263, "step": 52260 }, { "grad_norm": 0.29644131660461426, "learning_rate": 4.4707514384239365e-06, "loss": 0.0262, "step": 52270 }, { "grad_norm": 0.26794692873954773, "learning_rate": 4.459368098943484e-06, "loss": 0.0181, "step": 52280 }, { "grad_norm": 0.5061569213867188, "learning_rate": 4.447998593504582e-06, "loss": 0.0367, "step": 52290 }, { "grad_norm": 0.47447121143341064, "learning_rate": 4.4366429255609744e-06, "loss": 0.0248, "step": 52300 }, { "grad_norm": 0.3293200731277466, "learning_rate": 4.425301098562212e-06, "loss": 0.0239, "step": 52310 }, { "grad_norm": 0.3980064392089844, "learning_rate": 4.413973115953651e-06, "loss": 0.0214, "step": 52320 }, { "grad_norm": 0.3249003291130066, "learning_rate": 4.402658981176416e-06, "loss": 0.0224, "step": 52330 }, { "grad_norm": 0.632286548614502, "learning_rate": 4.391358697667475e-06, "loss": 0.0313, "step": 52340 }, { "grad_norm": 0.23370622098445892, "learning_rate": 4.3800722688595195e-06, "loss": 0.019, "step": 52350 }, { "grad_norm": 0.3027536869049072, "learning_rate": 4.368799698181097e-06, "loss": 0.03, "step": 52360 }, { "grad_norm": 0.29077252745628357, "learning_rate": 4.357540989056486e-06, "loss": 0.0291, "step": 52370 }, { "grad_norm": 0.8525390625, "learning_rate": 4.346296144905815e-06, "loss": 0.0337, "step": 52380 }, { "grad_norm": 0.35716450214385986, "learning_rate": 4.335065169144958e-06, "loss": 0.027, "step": 52390 }, { "grad_norm": 0.3917486369609833, "learning_rate": 4.323848065185593e-06, "loss": 0.0216, "step": 52400 }, { "grad_norm": 0.5371189117431641, "learning_rate": 4.31264483643517e-06, "loss": 0.0193, "step": 52410 }, { "grad_norm": 0.4170370101928711, "learning_rate": 4.301455486296946e-06, "loss": 0.033, "step": 52420 }, { "grad_norm": 0.3459438979625702, "learning_rate": 4.290280018169935e-06, "loss": 0.0202, "step": 52430 }, { "grad_norm": 0.34946659207344055, "learning_rate": 4.27911843544897e-06, "loss": 0.0235, "step": 52440 }, { "grad_norm": 0.450019508600235, "learning_rate": 4.2679707415246294e-06, "loss": 0.0324, "step": 52450 }, { "grad_norm": 0.2802029550075531, "learning_rate": 4.256836939783299e-06, "loss": 0.0306, "step": 52460 }, { "grad_norm": 0.3235459625720978, "learning_rate": 4.245717033607127e-06, "loss": 0.0223, "step": 52470 }, { "grad_norm": 0.613115131855011, "learning_rate": 4.234611026374035e-06, "loss": 0.0239, "step": 52480 }, { "grad_norm": 0.40036481618881226, "learning_rate": 4.2235189214577694e-06, "loss": 0.0135, "step": 52490 }, { "grad_norm": 0.3958970904350281, "learning_rate": 4.212440722227779e-06, "loss": 0.0331, "step": 52500 }, { "grad_norm": 0.20152056217193604, "learning_rate": 4.201376432049364e-06, "loss": 0.0255, "step": 52510 }, { "grad_norm": 0.627839207649231, "learning_rate": 4.1903260542835275e-06, "loss": 0.0342, "step": 52520 }, { "grad_norm": 0.2405000925064087, "learning_rate": 4.1792895922871114e-06, "loss": 0.0221, "step": 52530 }, { "grad_norm": 0.583418607711792, "learning_rate": 4.168267049412694e-06, "loss": 0.019, "step": 52540 }, { "grad_norm": 0.44060030579566956, "learning_rate": 4.157258429008626e-06, "loss": 0.0443, "step": 52550 }, { "grad_norm": 0.45697838068008423, "learning_rate": 4.146263734419043e-06, "loss": 0.0278, "step": 52560 }, { "grad_norm": 0.2811497747898102, "learning_rate": 4.135282968983839e-06, "loss": 0.0365, "step": 52570 }, { "grad_norm": 0.41185638308525085, "learning_rate": 4.124316136038675e-06, "loss": 0.049, "step": 52580 }, { "grad_norm": 0.3502779006958008, "learning_rate": 4.1133632389149965e-06, "loss": 0.0258, "step": 52590 }, { "grad_norm": 0.38708412647247314, "learning_rate": 4.102424280939998e-06, "loss": 0.0301, "step": 52600 }, { "grad_norm": 0.2557217478752136, "learning_rate": 4.091499265436649e-06, "loss": 0.0307, "step": 52610 }, { "grad_norm": 0.42742106318473816, "learning_rate": 4.080588195723684e-06, "loss": 0.0423, "step": 52620 }, { "grad_norm": 0.42375513911247253, "learning_rate": 4.069691075115578e-06, "loss": 0.026, "step": 52630 }, { "grad_norm": 0.26651644706726074, "learning_rate": 4.0588079069226235e-06, "loss": 0.0252, "step": 52640 }, { "grad_norm": 0.36448386311531067, "learning_rate": 4.0479386944508034e-06, "loss": 0.0384, "step": 52650 }, { "grad_norm": 0.22565419971942902, "learning_rate": 4.037083441001932e-06, "loss": 0.0253, "step": 52660 }, { "grad_norm": 0.5006480813026428, "learning_rate": 4.026242149873516e-06, "loss": 0.0247, "step": 52670 }, { "grad_norm": 0.5439955592155457, "learning_rate": 4.015414824358871e-06, "loss": 0.031, "step": 52680 }, { "grad_norm": 0.42639854550361633, "learning_rate": 4.004601467747054e-06, "loss": 0.0294, "step": 52690 }, { "grad_norm": 0.5724897384643555, "learning_rate": 3.993802083322873e-06, "loss": 0.0248, "step": 52700 }, { "grad_norm": 0.36632564663887024, "learning_rate": 3.9830166743668906e-06, "loss": 0.0265, "step": 52710 }, { "grad_norm": 0.18222226202487946, "learning_rate": 3.9722452441554425e-06, "loss": 0.0277, "step": 52720 }, { "grad_norm": 1.5518816709518433, "learning_rate": 3.961487795960584e-06, "loss": 0.0246, "step": 52730 }, { "grad_norm": 0.3922550082206726, "learning_rate": 3.95074433305016e-06, "loss": 0.0259, "step": 52740 }, { "grad_norm": 0.5633955001831055, "learning_rate": 3.940014858687752e-06, "loss": 0.0251, "step": 52750 }, { "grad_norm": 0.3884914219379425, "learning_rate": 3.929299376132689e-06, "loss": 0.0215, "step": 52760 }, { "grad_norm": 0.9809042811393738, "learning_rate": 3.918597888640047e-06, "loss": 0.0332, "step": 52770 }, { "grad_norm": 0.34374043345451355, "learning_rate": 3.907910399460657e-06, "loss": 0.0166, "step": 52780 }, { "grad_norm": 0.237521231174469, "learning_rate": 3.8972369118410956e-06, "loss": 0.0301, "step": 52790 }, { "grad_norm": 0.38620007038116455, "learning_rate": 3.88657742902368e-06, "loss": 0.0184, "step": 52800 }, { "grad_norm": 0.40703749656677246, "learning_rate": 3.875931954246504e-06, "loss": 0.0221, "step": 52810 }, { "grad_norm": 0.3224847912788391, "learning_rate": 3.865300490743351e-06, "loss": 0.018, "step": 52820 }, { "grad_norm": 0.3270570635795593, "learning_rate": 3.854683041743806e-06, "loss": 0.0139, "step": 52830 }, { "grad_norm": 0.7316077947616577, "learning_rate": 3.844079610473139e-06, "loss": 0.0334, "step": 52840 }, { "grad_norm": 0.22634820640087128, "learning_rate": 3.833490200152423e-06, "loss": 0.025, "step": 52850 }, { "grad_norm": 0.3895188271999359, "learning_rate": 3.822914813998424e-06, "loss": 0.0215, "step": 52860 }, { "grad_norm": 0.2803262770175934, "learning_rate": 3.812353455223666e-06, "loss": 0.0289, "step": 52870 }, { "grad_norm": 0.3535382151603699, "learning_rate": 3.8018061270364225e-06, "loss": 0.0237, "step": 52880 }, { "grad_norm": 0.3669419586658478, "learning_rate": 3.7912728326406688e-06, "loss": 0.0286, "step": 52890 }, { "grad_norm": 0.422476589679718, "learning_rate": 3.7807535752361732e-06, "loss": 0.0211, "step": 52900 }, { "grad_norm": 0.5113397240638733, "learning_rate": 3.7702483580183855e-06, "loss": 0.0214, "step": 52910 }, { "grad_norm": 0.705388605594635, "learning_rate": 3.759757184178525e-06, "loss": 0.0306, "step": 52920 }, { "grad_norm": 0.4978915750980377, "learning_rate": 3.7492800569035312e-06, "loss": 0.0296, "step": 52930 }, { "grad_norm": 0.4447348415851593, "learning_rate": 3.7388169793760754e-06, "loss": 0.0298, "step": 52940 }, { "grad_norm": 0.2504100203514099, "learning_rate": 3.728367954774553e-06, "loss": 0.0205, "step": 52950 }, { "grad_norm": 0.6491823196411133, "learning_rate": 3.7179329862731317e-06, "loss": 0.0333, "step": 52960 }, { "grad_norm": 0.4590783417224884, "learning_rate": 3.707512077041647e-06, "loss": 0.0262, "step": 52970 }, { "grad_norm": 0.61114901304245, "learning_rate": 3.6971052302457288e-06, "loss": 0.0211, "step": 52980 }, { "grad_norm": 0.3780655562877655, "learning_rate": 3.6867124490466698e-06, "loss": 0.0153, "step": 52990 }, { "grad_norm": 0.36034074425697327, "learning_rate": 3.6763337366015393e-06, "loss": 0.0257, "step": 53000 }, { "grad_norm": 0.5613618493080139, "learning_rate": 3.665969096063121e-06, "loss": 0.0365, "step": 53010 }, { "grad_norm": 0.26440849900245667, "learning_rate": 3.6556185305799074e-06, "loss": 0.018, "step": 53020 }, { "grad_norm": 0.1209363266825676, "learning_rate": 3.645282043296133e-06, "loss": 0.021, "step": 53030 }, { "grad_norm": 0.49286743998527527, "learning_rate": 3.6349596373517427e-06, "loss": 0.0284, "step": 53040 }, { "grad_norm": 0.34709858894348145, "learning_rate": 3.6246513158824215e-06, "loss": 0.0237, "step": 53050 }, { "grad_norm": 0.5182263255119324, "learning_rate": 3.6143570820195593e-06, "loss": 0.0229, "step": 53060 }, { "grad_norm": 0.370491087436676, "learning_rate": 3.6040769388902773e-06, "loss": 0.0233, "step": 53070 }, { "grad_norm": 0.37828126549720764, "learning_rate": 3.593810889617405e-06, "loss": 0.027, "step": 53080 }, { "grad_norm": 0.2545153498649597, "learning_rate": 3.5835589373194978e-06, "loss": 0.0158, "step": 53090 }, { "grad_norm": 0.3830292224884033, "learning_rate": 3.5733210851108257e-06, "loss": 0.0257, "step": 53100 }, { "grad_norm": 0.6146647930145264, "learning_rate": 3.5630973361014008e-06, "loss": 0.0283, "step": 53110 }, { "grad_norm": 0.2464008927345276, "learning_rate": 3.552887693396889e-06, "loss": 0.0238, "step": 53120 }, { "grad_norm": 0.9542460441589355, "learning_rate": 3.542692160098754e-06, "loss": 0.0201, "step": 53130 }, { "grad_norm": 0.35338571667671204, "learning_rate": 3.5325107393040846e-06, "loss": 0.0205, "step": 53140 }, { "grad_norm": 0.2694293260574341, "learning_rate": 3.522343434105757e-06, "loss": 0.0253, "step": 53150 }, { "grad_norm": 0.36319950222969055, "learning_rate": 3.512190247592323e-06, "loss": 0.0295, "step": 53160 }, { "grad_norm": 0.7032114863395691, "learning_rate": 3.502051182848054e-06, "loss": 0.0268, "step": 53170 }, { "grad_norm": 0.45373037457466125, "learning_rate": 3.4919262429529308e-06, "loss": 0.0211, "step": 53180 }, { "grad_norm": 0.7599675059318542, "learning_rate": 3.4818154309826325e-06, "loss": 0.0359, "step": 53190 }, { "grad_norm": 0.1902616322040558, "learning_rate": 3.4717187500085734e-06, "loss": 0.0227, "step": 53200 }, { "grad_norm": 0.3070521056652069, "learning_rate": 3.46163620309784e-06, "loss": 0.0297, "step": 53210 }, { "grad_norm": 0.34494227170944214, "learning_rate": 3.4515677933132595e-06, "loss": 0.0202, "step": 53220 }, { "grad_norm": 0.3538302183151245, "learning_rate": 3.4415135237133466e-06, "loss": 0.024, "step": 53230 }, { "grad_norm": 0.9005641937255859, "learning_rate": 3.4314733973523196e-06, "loss": 0.0414, "step": 53240 }, { "grad_norm": 0.38303184509277344, "learning_rate": 3.4214474172800993e-06, "loss": 0.033, "step": 53250 }, { "grad_norm": 0.6725941896438599, "learning_rate": 3.411435586542322e-06, "loss": 0.017, "step": 53260 }, { "grad_norm": 0.3551807403564453, "learning_rate": 3.4014379081802995e-06, "loss": 0.0219, "step": 53270 }, { "grad_norm": 0.3551011085510254, "learning_rate": 3.391454385231102e-06, "loss": 0.0192, "step": 53280 }, { "grad_norm": 0.7353115081787109, "learning_rate": 3.3814850207274095e-06, "loss": 0.0246, "step": 53290 }, { "grad_norm": 0.598024845123291, "learning_rate": 3.371529817697694e-06, "loss": 0.0237, "step": 53300 }, { "grad_norm": 0.3500533401966095, "learning_rate": 3.3615887791660585e-06, "loss": 0.0221, "step": 53310 }, { "grad_norm": 0.4780050814151764, "learning_rate": 3.3516619081523426e-06, "loss": 0.0269, "step": 53320 }, { "grad_norm": 0.5484272241592407, "learning_rate": 3.3417492076720567e-06, "loss": 0.0178, "step": 53330 }, { "grad_norm": 0.36914071440696716, "learning_rate": 3.3318506807364147e-06, "loss": 0.0276, "step": 53340 }, { "grad_norm": 0.2474418580532074, "learning_rate": 3.3219663303523553e-06, "loss": 0.0198, "step": 53350 }, { "grad_norm": 0.42720332741737366, "learning_rate": 3.3120961595224374e-06, "loss": 0.0203, "step": 53360 }, { "grad_norm": 0.29994213581085205, "learning_rate": 3.302240171245002e-06, "loss": 0.0208, "step": 53370 }, { "grad_norm": 0.607571542263031, "learning_rate": 3.29239836851401e-06, "loss": 0.0324, "step": 53380 }, { "grad_norm": 0.3957028388977051, "learning_rate": 3.2825707543191588e-06, "loss": 0.0382, "step": 53390 }, { "grad_norm": 0.5244816541671753, "learning_rate": 3.272757331645804e-06, "loss": 0.035, "step": 53400 }, { "grad_norm": 0.6346034407615662, "learning_rate": 3.2629581034750166e-06, "loss": 0.0316, "step": 53410 }, { "grad_norm": 0.3238421082496643, "learning_rate": 3.2531730727835218e-06, "loss": 0.0224, "step": 53420 }, { "grad_norm": 0.24530647695064545, "learning_rate": 3.2434022425437914e-06, "loss": 0.0179, "step": 53430 }, { "grad_norm": 0.45621055364608765, "learning_rate": 3.233645615723907e-06, "loss": 0.0223, "step": 53440 }, { "grad_norm": 0.42336931824684143, "learning_rate": 3.2239031952876918e-06, "loss": 0.017, "step": 53450 }, { "grad_norm": 0.28987133502960205, "learning_rate": 3.21417498419464e-06, "loss": 0.018, "step": 53460 }, { "grad_norm": 0.2926830053329468, "learning_rate": 3.204460985399921e-06, "loss": 0.0245, "step": 53470 }, { "grad_norm": 0.5567307472229004, "learning_rate": 3.1947612018543903e-06, "loss": 0.024, "step": 53480 }, { "grad_norm": 0.32730966806411743, "learning_rate": 3.1850756365045753e-06, "loss": 0.0203, "step": 53490 }, { "grad_norm": 0.43329691886901855, "learning_rate": 3.175404292292722e-06, "loss": 0.0364, "step": 53500 }, { "grad_norm": 0.22679178416728973, "learning_rate": 3.1657471721566965e-06, "loss": 0.0228, "step": 53510 }, { "grad_norm": 0.7102733254432678, "learning_rate": 3.1561042790300977e-06, "loss": 0.0295, "step": 53520 }, { "grad_norm": 0.39346906542778015, "learning_rate": 3.1464756158421816e-06, "loss": 0.0252, "step": 53530 }, { "grad_norm": 0.3371201455593109, "learning_rate": 3.136861185517875e-06, "loss": 0.0285, "step": 53540 }, { "grad_norm": 0.7647133469581604, "learning_rate": 3.127260990977798e-06, "loss": 0.0255, "step": 53550 }, { "grad_norm": 0.4030882716178894, "learning_rate": 3.1176750351382235e-06, "loss": 0.0195, "step": 53560 }, { "grad_norm": 0.20895080268383026, "learning_rate": 3.1081033209111153e-06, "loss": 0.0268, "step": 53570 }, { "grad_norm": 0.46047243475914, "learning_rate": 3.0985458512041155e-06, "loss": 0.0185, "step": 53580 }, { "grad_norm": 0.2594375014305115, "learning_rate": 3.089002628920512e-06, "loss": 0.0215, "step": 53590 }, { "grad_norm": 0.45650139451026917, "learning_rate": 3.079473656959303e-06, "loss": 0.022, "step": 53600 }, { "grad_norm": 0.5003728270530701, "learning_rate": 3.0699589382151393e-06, "loss": 0.0249, "step": 53610 }, { "grad_norm": 0.3556647002696991, "learning_rate": 3.060458475578326e-06, "loss": 0.0266, "step": 53620 }, { "grad_norm": 0.3046031594276428, "learning_rate": 3.0509722719348656e-06, "loss": 0.0246, "step": 53630 }, { "grad_norm": 0.526162326335907, "learning_rate": 3.041500330166408e-06, "loss": 0.0245, "step": 53640 }, { "grad_norm": 0.4890054762363434, "learning_rate": 3.032042653150291e-06, "loss": 0.019, "step": 53650 }, { "grad_norm": 0.3572906255722046, "learning_rate": 3.0225992437594887e-06, "loss": 0.0218, "step": 53660 }, { "grad_norm": 0.4825665056705475, "learning_rate": 3.0131701048626783e-06, "loss": 0.0281, "step": 53670 }, { "grad_norm": 0.7189273834228516, "learning_rate": 3.003755239324163e-06, "loss": 0.0303, "step": 53680 }, { "grad_norm": 0.40918606519699097, "learning_rate": 2.9943546500039553e-06, "loss": 0.0295, "step": 53690 }, { "grad_norm": 0.6569284200668335, "learning_rate": 2.9849683397576877e-06, "loss": 0.0198, "step": 53700 }, { "grad_norm": 0.27463653683662415, "learning_rate": 2.975596311436679e-06, "loss": 0.0272, "step": 53710 }, { "grad_norm": 0.5160747170448303, "learning_rate": 2.966238567887902e-06, "loss": 0.0185, "step": 53720 }, { "grad_norm": 0.9647044539451599, "learning_rate": 2.9568951119539943e-06, "loss": 0.0286, "step": 53730 }, { "grad_norm": 0.5444865226745605, "learning_rate": 2.947565946473241e-06, "loss": 0.0222, "step": 53740 }, { "grad_norm": 0.2585434913635254, "learning_rate": 2.9382510742796188e-06, "loss": 0.0204, "step": 53750 }, { "grad_norm": 0.5030410289764404, "learning_rate": 2.9289504982027204e-06, "loss": 0.0258, "step": 53760 }, { "grad_norm": 0.37314361333847046, "learning_rate": 2.9196642210678248e-06, "loss": 0.0241, "step": 53770 }, { "grad_norm": 0.15809552371501923, "learning_rate": 2.910392245695853e-06, "loss": 0.019, "step": 53780 }, { "grad_norm": 0.2914482653141022, "learning_rate": 2.9011345749033804e-06, "loss": 0.0225, "step": 53790 }, { "grad_norm": 0.46923673152923584, "learning_rate": 2.8918912115026677e-06, "loss": 0.0233, "step": 53800 }, { "grad_norm": 0.4648592174053192, "learning_rate": 2.8826621583015743e-06, "loss": 0.0296, "step": 53810 }, { "grad_norm": 0.4122743308544159, "learning_rate": 2.8734474181036643e-06, "loss": 0.0234, "step": 53820 }, { "grad_norm": 0.4576255679130554, "learning_rate": 2.864246993708114e-06, "loss": 0.024, "step": 53830 }, { "grad_norm": 0.8758717179298401, "learning_rate": 2.8550608879097884e-06, "loss": 0.0336, "step": 53840 }, { "grad_norm": 0.5616545081138611, "learning_rate": 2.845889103499172e-06, "loss": 0.0248, "step": 53850 }, { "grad_norm": 0.4807929992675781, "learning_rate": 2.8367316432624138e-06, "loss": 0.0241, "step": 53860 }, { "grad_norm": 0.4720830023288727, "learning_rate": 2.8275885099813105e-06, "loss": 0.0369, "step": 53870 }, { "grad_norm": 0.7832702398300171, "learning_rate": 2.8184597064332963e-06, "loss": 0.0254, "step": 53880 }, { "grad_norm": 0.5888053178787231, "learning_rate": 2.809345235391464e-06, "loss": 0.0278, "step": 53890 }, { "grad_norm": 0.6954466104507446, "learning_rate": 2.80024509962456e-06, "loss": 0.0288, "step": 53900 }, { "grad_norm": 0.21591192483901978, "learning_rate": 2.7911593018969563e-06, "loss": 0.0178, "step": 53910 }, { "grad_norm": 0.36767056584358215, "learning_rate": 2.7820878449686838e-06, "loss": 0.0274, "step": 53920 }, { "grad_norm": 0.3563261926174164, "learning_rate": 2.7730307315953995e-06, "loss": 0.0288, "step": 53930 }, { "grad_norm": 0.3176560699939728, "learning_rate": 2.763987964528425e-06, "loss": 0.0263, "step": 53940 }, { "grad_norm": 0.49521738290786743, "learning_rate": 2.754959546514718e-06, "loss": 0.0201, "step": 53950 }, { "grad_norm": 0.18599702417850494, "learning_rate": 2.7459454802968576e-06, "loss": 0.0216, "step": 53960 }, { "grad_norm": 0.8156259059906006, "learning_rate": 2.7369457686131028e-06, "loss": 0.0307, "step": 53970 }, { "grad_norm": 0.21555981040000916, "learning_rate": 2.7279604141973004e-06, "loss": 0.0174, "step": 53980 }, { "grad_norm": 0.37982115149497986, "learning_rate": 2.7189894197789946e-06, "loss": 0.0206, "step": 53990 }, { "grad_norm": 0.4089737832546234, "learning_rate": 2.7100327880833055e-06, "loss": 0.0261, "step": 54000 }, { "grad_norm": 0.2988273799419403, "learning_rate": 2.70109052183104e-06, "loss": 0.0148, "step": 54010 }, { "grad_norm": 0.4612543284893036, "learning_rate": 2.69216262373862e-06, "loss": 0.0199, "step": 54020 }, { "grad_norm": 0.2773697078227997, "learning_rate": 2.6832490965181036e-06, "loss": 0.021, "step": 54030 }, { "grad_norm": 0.2944963276386261, "learning_rate": 2.6743499428771857e-06, "loss": 0.0309, "step": 54040 }, { "grad_norm": 0.48041218519210815, "learning_rate": 2.6654651655191875e-06, "loss": 0.0275, "step": 54050 }, { "grad_norm": 0.48712342977523804, "learning_rate": 2.6565947671430836e-06, "loss": 0.0231, "step": 54060 }, { "grad_norm": 0.3076384365558624, "learning_rate": 2.647738750443457e-06, "loss": 0.0221, "step": 54070 }, { "grad_norm": 0.29562580585479736, "learning_rate": 2.6388971181105393e-06, "loss": 0.0357, "step": 54080 }, { "grad_norm": 0.35227587819099426, "learning_rate": 2.630069872830171e-06, "loss": 0.0207, "step": 54090 }, { "grad_norm": 0.5705763101577759, "learning_rate": 2.6212570172838514e-06, "loss": 0.0242, "step": 54100 }, { "grad_norm": 0.5381326079368591, "learning_rate": 2.6124585541486778e-06, "loss": 0.0202, "step": 54110 }, { "grad_norm": 0.4769427478313446, "learning_rate": 2.6036744860974127e-06, "loss": 0.0179, "step": 54120 }, { "grad_norm": 0.5157383680343628, "learning_rate": 2.594904815798399e-06, "loss": 0.0239, "step": 54130 }, { "grad_norm": 0.27692991495132446, "learning_rate": 2.586149545915656e-06, "loss": 0.0199, "step": 54140 }, { "grad_norm": 0.3932849168777466, "learning_rate": 2.577408679108778e-06, "loss": 0.0244, "step": 54150 }, { "grad_norm": 0.2899329662322998, "learning_rate": 2.5686822180330306e-06, "loss": 0.0323, "step": 54160 }, { "grad_norm": 0.5288882255554199, "learning_rate": 2.5599701653392703e-06, "loss": 0.0272, "step": 54170 }, { "grad_norm": 0.30620014667510986, "learning_rate": 2.551272523673992e-06, "loss": 0.0216, "step": 54180 }, { "grad_norm": 0.37671276926994324, "learning_rate": 2.542589295679315e-06, "loss": 0.0193, "step": 54190 }, { "grad_norm": 0.39986154437065125, "learning_rate": 2.5339204839929575e-06, "loss": 0.0205, "step": 54200 }, { "grad_norm": 0.6355361938476562, "learning_rate": 2.525266091248296e-06, "loss": 0.0191, "step": 54210 }, { "grad_norm": 0.4563184082508087, "learning_rate": 2.5166261200743e-06, "loss": 0.0282, "step": 54220 }, { "grad_norm": 0.5952759385108948, "learning_rate": 2.5080005730955646e-06, "loss": 0.0296, "step": 54230 }, { "grad_norm": 0.700932502746582, "learning_rate": 2.499389452932299e-06, "loss": 0.0278, "step": 54240 }, { "grad_norm": 0.5671364665031433, "learning_rate": 2.4907927622003336e-06, "loss": 0.0297, "step": 54250 }, { "grad_norm": 0.3301407992839813, "learning_rate": 2.4822105035111177e-06, "loss": 0.0279, "step": 54260 }, { "grad_norm": 0.625255286693573, "learning_rate": 2.4736426794717273e-06, "loss": 0.0349, "step": 54270 }, { "grad_norm": 0.6095210909843445, "learning_rate": 2.4650892926848135e-06, "loss": 0.0309, "step": 54280 }, { "grad_norm": 0.40012112259864807, "learning_rate": 2.456550345748704e-06, "loss": 0.0235, "step": 54290 }, { "grad_norm": 0.37648165225982666, "learning_rate": 2.4480258412572733e-06, "loss": 0.0285, "step": 54300 }, { "grad_norm": 0.6464018821716309, "learning_rate": 2.4395157818000612e-06, "loss": 0.0245, "step": 54310 }, { "grad_norm": 0.390261173248291, "learning_rate": 2.431020169962189e-06, "loss": 0.0238, "step": 54320 }, { "grad_norm": 0.5278199911117554, "learning_rate": 2.422539008324409e-06, "loss": 0.0213, "step": 54330 }, { "grad_norm": 0.2605428397655487, "learning_rate": 2.414072299463066e-06, "loss": 0.0221, "step": 54340 }, { "grad_norm": 0.34059345722198486, "learning_rate": 2.4056200459501186e-06, "loss": 0.0153, "step": 54350 }, { "grad_norm": 0.550759494304657, "learning_rate": 2.397182250353147e-06, "loss": 0.0399, "step": 54360 }, { "grad_norm": 0.2502182424068451, "learning_rate": 2.388758915235334e-06, "loss": 0.0255, "step": 54370 }, { "grad_norm": 0.28125426173210144, "learning_rate": 2.380350043155455e-06, "loss": 0.0252, "step": 54380 }, { "grad_norm": 0.6769407987594604, "learning_rate": 2.371955636667911e-06, "loss": 0.0293, "step": 54390 }, { "grad_norm": 0.5207740068435669, "learning_rate": 2.3635756983227008e-06, "loss": 0.0298, "step": 54400 }, { "grad_norm": 0.46756696701049805, "learning_rate": 2.3552102306654278e-06, "loss": 0.0245, "step": 54410 }, { "grad_norm": 0.6088024973869324, "learning_rate": 2.346859236237292e-06, "loss": 0.0226, "step": 54420 }, { "grad_norm": 0.22884386777877808, "learning_rate": 2.3385227175751145e-06, "loss": 0.0173, "step": 54430 }, { "grad_norm": 0.4033292233943939, "learning_rate": 2.330200677211314e-06, "loss": 0.0191, "step": 54440 }, { "grad_norm": 0.36258116364479065, "learning_rate": 2.3218931176738847e-06, "loss": 0.0199, "step": 54450 }, { "grad_norm": 0.22714966535568237, "learning_rate": 2.313600041486469e-06, "loss": 0.0222, "step": 54460 }, { "grad_norm": 0.4843306541442871, "learning_rate": 2.3053214511682743e-06, "loss": 0.0263, "step": 54470 }, { "grad_norm": 0.1495589166879654, "learning_rate": 2.2970573492341163e-06, "loss": 0.0191, "step": 54480 }, { "grad_norm": 0.7029808759689331, "learning_rate": 2.288807738194415e-06, "loss": 0.0397, "step": 54490 }, { "grad_norm": 0.453774094581604, "learning_rate": 2.2805726205551768e-06, "loss": 0.0213, "step": 54500 }, { "grad_norm": 0.325932115316391, "learning_rate": 2.272351998818023e-06, "loss": 0.019, "step": 54510 }, { "grad_norm": 0.3762201964855194, "learning_rate": 2.2641458754801505e-06, "loss": 0.0171, "step": 54520 }, { "grad_norm": 0.18290701508522034, "learning_rate": 2.2559542530343756e-06, "loss": 0.0185, "step": 54530 }, { "grad_norm": 0.2945437729358673, "learning_rate": 2.247777133969087e-06, "loss": 0.0222, "step": 54540 }, { "grad_norm": 0.2863091826438904, "learning_rate": 2.2396145207682795e-06, "loss": 0.0206, "step": 54550 }, { "grad_norm": 0.5283860564231873, "learning_rate": 2.231466415911543e-06, "loss": 0.0225, "step": 54560 }, { "grad_norm": 0.4151541292667389, "learning_rate": 2.2233328218740524e-06, "loss": 0.0228, "step": 54570 }, { "grad_norm": 0.36157315969467163, "learning_rate": 2.215213741126576e-06, "loss": 0.0126, "step": 54580 }, { "grad_norm": 0.40964475274086, "learning_rate": 2.2071091761354912e-06, "loss": 0.0148, "step": 54590 }, { "grad_norm": 0.4338223934173584, "learning_rate": 2.1990191293627337e-06, "loss": 0.0177, "step": 54600 }, { "grad_norm": 0.28726035356521606, "learning_rate": 2.1909436032658548e-06, "loss": 0.0216, "step": 54610 }, { "grad_norm": 0.4765681028366089, "learning_rate": 2.1828826002979806e-06, "loss": 0.0244, "step": 54620 }, { "grad_norm": 0.2843599319458008, "learning_rate": 2.1748361229078362e-06, "loss": 0.0266, "step": 54630 }, { "grad_norm": 0.3621889650821686, "learning_rate": 2.1668041735397327e-06, "loss": 0.0202, "step": 54640 }, { "grad_norm": 0.23099343478679657, "learning_rate": 2.1587867546335514e-06, "loss": 0.0228, "step": 54650 }, { "grad_norm": 0.5134407877922058, "learning_rate": 2.1507838686247894e-06, "loss": 0.021, "step": 54660 }, { "grad_norm": 0.38324692845344543, "learning_rate": 2.1427955179444957e-06, "loss": 0.0215, "step": 54670 }, { "grad_norm": 0.4055512249469757, "learning_rate": 2.13482170501933e-06, "loss": 0.0213, "step": 54680 }, { "grad_norm": 0.6119892597198486, "learning_rate": 2.1268624322715202e-06, "loss": 0.037, "step": 54690 }, { "grad_norm": 0.5173929333686829, "learning_rate": 2.118917702118889e-06, "loss": 0.0254, "step": 54700 }, { "grad_norm": 0.6945585608482361, "learning_rate": 2.1109875169748327e-06, "loss": 0.0275, "step": 54710 }, { "grad_norm": 0.3919891119003296, "learning_rate": 2.103071879248336e-06, "loss": 0.0389, "step": 54720 }, { "grad_norm": 0.26779890060424805, "learning_rate": 2.0951707913439478e-06, "loss": 0.0257, "step": 54730 }, { "grad_norm": 0.3276280462741852, "learning_rate": 2.0872842556618255e-06, "loss": 0.0212, "step": 54740 }, { "grad_norm": 0.8517383337020874, "learning_rate": 2.079412274597686e-06, "loss": 0.0232, "step": 54750 }, { "grad_norm": 0.33495911955833435, "learning_rate": 2.0715548505428284e-06, "loss": 0.0222, "step": 54760 }, { "grad_norm": 0.39214104413986206, "learning_rate": 2.0637119858841258e-06, "loss": 0.0241, "step": 54770 }, { "grad_norm": 0.3341820240020752, "learning_rate": 2.055883683004034e-06, "loss": 0.0189, "step": 54780 }, { "grad_norm": 0.5136025547981262, "learning_rate": 2.0480699442806006e-06, "loss": 0.018, "step": 54790 }, { "grad_norm": 0.32831716537475586, "learning_rate": 2.0402707720874105e-06, "loss": 0.0191, "step": 54800 }, { "grad_norm": 0.29567238688468933, "learning_rate": 2.032486168793668e-06, "loss": 0.0294, "step": 54810 }, { "grad_norm": 0.33152326941490173, "learning_rate": 2.024716136764104e-06, "loss": 0.0255, "step": 54820 }, { "grad_norm": 0.23883432149887085, "learning_rate": 2.0169606783590757e-06, "loss": 0.0222, "step": 54830 }, { "grad_norm": 0.46788984537124634, "learning_rate": 2.0092197959344638e-06, "loss": 0.0233, "step": 54840 }, { "grad_norm": 0.5893208384513855, "learning_rate": 2.0014934918417606e-06, "loss": 0.0267, "step": 54850 }, { "grad_norm": 0.5264681577682495, "learning_rate": 1.99378176842801e-06, "loss": 0.0311, "step": 54860 }, { "grad_norm": 0.5102050304412842, "learning_rate": 1.9860846280358224e-06, "loss": 0.0307, "step": 54870 }, { "grad_norm": 0.6069658994674683, "learning_rate": 1.978402073003394e-06, "loss": 0.0319, "step": 54880 }, { "grad_norm": 0.24502398073673248, "learning_rate": 1.9707341056644736e-06, "loss": 0.0218, "step": 54890 }, { "grad_norm": 0.3279726505279541, "learning_rate": 1.963080728348399e-06, "loss": 0.0231, "step": 54900 }, { "grad_norm": 0.29956305027008057, "learning_rate": 1.955441943380054e-06, "loss": 0.0253, "step": 54910 }, { "grad_norm": 0.3095155656337738, "learning_rate": 1.947817753079906e-06, "loss": 0.0243, "step": 54920 }, { "grad_norm": 0.6560242772102356, "learning_rate": 1.9402081597639785e-06, "loss": 0.0225, "step": 54930 }, { "grad_norm": 0.6858740448951721, "learning_rate": 1.9326131657438683e-06, "loss": 0.025, "step": 54940 }, { "grad_norm": 0.2818698287010193, "learning_rate": 1.925032773326724e-06, "loss": 0.0244, "step": 54950 }, { "grad_norm": 0.4976808726787567, "learning_rate": 1.9174669848152916e-06, "loss": 0.0264, "step": 54960 }, { "grad_norm": 0.45232415199279785, "learning_rate": 1.9099158025078336e-06, "loss": 0.0237, "step": 54970 }, { "grad_norm": 0.6029608249664307, "learning_rate": 1.90237922869822e-06, "loss": 0.03, "step": 54980 }, { "grad_norm": 0.3132363259792328, "learning_rate": 1.8948572656758367e-06, "loss": 0.0252, "step": 54990 }, { "grad_norm": 0.4986538589000702, "learning_rate": 1.8873499157256834e-06, "loss": 0.0361, "step": 55000 }, { "grad_norm": 0.4291965663433075, "learning_rate": 1.879857181128286e-06, "loss": 0.0304, "step": 55010 }, { "grad_norm": 0.6300171613693237, "learning_rate": 1.8723790641597349e-06, "loss": 0.0262, "step": 55020 }, { "grad_norm": 0.3238847553730011, "learning_rate": 1.8649155670916906e-06, "loss": 0.0213, "step": 55030 }, { "grad_norm": 0.29393044114112854, "learning_rate": 1.8574666921913565e-06, "loss": 0.0317, "step": 55040 }, { "grad_norm": 0.25900998711586, "learning_rate": 1.8500324417215166e-06, "loss": 0.0373, "step": 55050 }, { "grad_norm": 0.2977149784564972, "learning_rate": 1.8426128179404977e-06, "loss": 0.0253, "step": 55060 }, { "grad_norm": 0.32882434129714966, "learning_rate": 1.8352078231021807e-06, "loss": 0.0268, "step": 55070 }, { "grad_norm": 0.4061976969242096, "learning_rate": 1.8278174594560049e-06, "loss": 0.0233, "step": 55080 }, { "grad_norm": 0.2946261167526245, "learning_rate": 1.82044172924698e-06, "loss": 0.0249, "step": 55090 }, { "grad_norm": 0.4270544648170471, "learning_rate": 1.813080634715636e-06, "loss": 0.0289, "step": 55100 }, { "grad_norm": 0.3443894386291504, "learning_rate": 1.8057341780981119e-06, "loss": 0.0179, "step": 55110 }, { "grad_norm": 0.3694981634616852, "learning_rate": 1.7984023616260338e-06, "loss": 0.02, "step": 55120 }, { "grad_norm": 0.24870502948760986, "learning_rate": 1.7910851875266421e-06, "loss": 0.018, "step": 55130 }, { "grad_norm": 0.6639658212661743, "learning_rate": 1.7837826580226757e-06, "loss": 0.0247, "step": 55140 }, { "grad_norm": 0.7568885087966919, "learning_rate": 1.7764947753324656e-06, "loss": 0.0248, "step": 55150 }, { "grad_norm": 0.2841445207595825, "learning_rate": 1.7692215416698799e-06, "loss": 0.026, "step": 55160 }, { "grad_norm": 0.7387504577636719, "learning_rate": 1.7619629592443233e-06, "loss": 0.0236, "step": 55170 }, { "grad_norm": 0.47757476568222046, "learning_rate": 1.7547190302607709e-06, "loss": 0.0195, "step": 55180 }, { "grad_norm": 0.19166088104248047, "learning_rate": 1.747489756919729e-06, "loss": 0.0222, "step": 55190 }, { "grad_norm": 0.683109700679779, "learning_rate": 1.7402751414172802e-06, "loss": 0.0319, "step": 55200 }, { "grad_norm": 0.3104842007160187, "learning_rate": 1.7330751859450044e-06, "loss": 0.0252, "step": 55210 }, { "grad_norm": 0.29491230845451355, "learning_rate": 1.7258898926900801e-06, "loss": 0.0245, "step": 55220 }, { "grad_norm": 0.27189144492149353, "learning_rate": 1.7187192638352002e-06, "loss": 0.0205, "step": 55230 }, { "grad_norm": 0.35510265827178955, "learning_rate": 1.7115633015586163e-06, "loss": 0.0282, "step": 55240 }, { "grad_norm": 0.9453732371330261, "learning_rate": 1.7044220080341178e-06, "loss": 0.0198, "step": 55250 }, { "grad_norm": 0.2661549150943756, "learning_rate": 1.6972953854310414e-06, "loss": 0.0253, "step": 55260 }, { "grad_norm": 0.2937701642513275, "learning_rate": 1.690183435914261e-06, "loss": 0.0175, "step": 55270 }, { "grad_norm": 0.35584646463394165, "learning_rate": 1.6830861616442206e-06, "loss": 0.0195, "step": 55280 }, { "grad_norm": 0.6164565086364746, "learning_rate": 1.6760035647768568e-06, "loss": 0.03, "step": 55290 }, { "grad_norm": 0.3161877989768982, "learning_rate": 1.6689356474636985e-06, "loss": 0.035, "step": 55300 }, { "grad_norm": 0.6210693120956421, "learning_rate": 1.6618824118517784e-06, "loss": 0.0265, "step": 55310 }, { "grad_norm": 0.15116728842258453, "learning_rate": 1.654843860083688e-06, "loss": 0.0467, "step": 55320 }, { "grad_norm": 0.705089807510376, "learning_rate": 1.647819994297556e-06, "loss": 0.0345, "step": 55330 }, { "grad_norm": 0.12311479449272156, "learning_rate": 1.6408108166270363e-06, "loss": 0.0264, "step": 55340 }, { "grad_norm": 0.23658554255962372, "learning_rate": 1.6338163292013486e-06, "loss": 0.0164, "step": 55350 }, { "grad_norm": 0.4203987121582031, "learning_rate": 1.6268365341452208e-06, "loss": 0.0323, "step": 55360 }, { "grad_norm": 0.3740260601043701, "learning_rate": 1.6198714335789345e-06, "loss": 0.0157, "step": 55370 }, { "grad_norm": 0.18516714870929718, "learning_rate": 1.612921029618303e-06, "loss": 0.0171, "step": 55380 }, { "grad_norm": 0.49176836013793945, "learning_rate": 1.6059853243746815e-06, "loss": 0.0271, "step": 55390 }, { "grad_norm": 0.35327067971229553, "learning_rate": 1.5990643199549404e-06, "loss": 0.0226, "step": 55400 }, { "grad_norm": 0.360816091299057, "learning_rate": 1.5921580184615147e-06, "loss": 0.0197, "step": 55410 }, { "grad_norm": 0.6630728244781494, "learning_rate": 1.5852664219923374e-06, "loss": 0.0179, "step": 55420 }, { "grad_norm": 0.28540167212486267, "learning_rate": 1.5783895326409172e-06, "loss": 0.0216, "step": 55430 }, { "grad_norm": 0.303911030292511, "learning_rate": 1.5715273524962438e-06, "loss": 0.0217, "step": 55440 }, { "grad_norm": 0.6674954891204834, "learning_rate": 1.564679883642889e-06, "loss": 0.0256, "step": 55450 }, { "grad_norm": 0.6988134384155273, "learning_rate": 1.5578471281609276e-06, "loss": 0.026, "step": 55460 }, { "grad_norm": 0.46084174513816833, "learning_rate": 1.551029088125966e-06, "loss": 0.0223, "step": 55470 }, { "grad_norm": 0.30574795603752136, "learning_rate": 1.544225765609142e-06, "loss": 0.0296, "step": 55480 }, { "grad_norm": 0.37319669127464294, "learning_rate": 1.53743716267713e-06, "loss": 0.0225, "step": 55490 }, { "grad_norm": 0.3892769515514374, "learning_rate": 1.5306632813921361e-06, "loss": 0.0201, "step": 55500 }, { "grad_norm": 0.37570393085479736, "learning_rate": 1.52390412381187e-06, "loss": 0.0216, "step": 55510 }, { "grad_norm": 0.4311186671257019, "learning_rate": 1.5171596919895948e-06, "loss": 0.0234, "step": 55520 }, { "grad_norm": 0.6938761472702026, "learning_rate": 1.5104299879740936e-06, "loss": 0.0317, "step": 55530 }, { "grad_norm": 0.5882732272148132, "learning_rate": 1.5037150138096701e-06, "loss": 0.0267, "step": 55540 }, { "grad_norm": 0.550614058971405, "learning_rate": 1.4970147715361538e-06, "loss": 0.0289, "step": 55550 }, { "grad_norm": 0.41412046551704407, "learning_rate": 1.4903292631889054e-06, "loss": 0.0195, "step": 55560 }, { "grad_norm": 0.36844155192375183, "learning_rate": 1.483658490798795e-06, "loss": 0.0252, "step": 55570 }, { "grad_norm": 0.2973672151565552, "learning_rate": 1.4770024563922457e-06, "loss": 0.0305, "step": 55580 }, { "grad_norm": 0.6966583132743835, "learning_rate": 1.4703611619911628e-06, "loss": 0.0193, "step": 55590 }, { "grad_norm": 0.47342127561569214, "learning_rate": 1.4637346096130155e-06, "loss": 0.0207, "step": 55600 }, { "grad_norm": 0.8037583231925964, "learning_rate": 1.4571228012707662e-06, "loss": 0.0224, "step": 55610 }, { "grad_norm": 0.3331577479839325, "learning_rate": 1.4505257389729132e-06, "loss": 0.0335, "step": 55620 }, { "grad_norm": 0.3786936104297638, "learning_rate": 1.4439434247234596e-06, "loss": 0.023, "step": 55630 }, { "grad_norm": 0.5588763952255249, "learning_rate": 1.4373758605219445e-06, "loss": 0.0221, "step": 55640 }, { "grad_norm": 0.35370296239852905, "learning_rate": 1.4308230483634333e-06, "loss": 0.0267, "step": 55650 }, { "grad_norm": 0.5352171659469604, "learning_rate": 1.4242849902384724e-06, "loss": 0.0184, "step": 55660 }, { "grad_norm": 0.6615639925003052, "learning_rate": 1.4177616881331734e-06, "loss": 0.0267, "step": 55670 }, { "grad_norm": 0.55434650182724, "learning_rate": 1.4112531440291233e-06, "loss": 0.0307, "step": 55680 }, { "grad_norm": 0.12962603569030762, "learning_rate": 1.4047593599034624e-06, "loss": 0.0204, "step": 55690 }, { "grad_norm": 0.5960956811904907, "learning_rate": 1.3982803377288246e-06, "loss": 0.0202, "step": 55700 }, { "grad_norm": 0.3592130243778229, "learning_rate": 1.3918160794733681e-06, "loss": 0.0274, "step": 55710 }, { "grad_norm": 0.3167349398136139, "learning_rate": 1.3853665871007615e-06, "loss": 0.0219, "step": 55720 }, { "grad_norm": 0.2337118536233902, "learning_rate": 1.378931862570193e-06, "loss": 0.0179, "step": 55730 }, { "grad_norm": 0.3344493508338928, "learning_rate": 1.372511907836349e-06, "loss": 0.0265, "step": 55740 }, { "grad_norm": 0.504852294921875, "learning_rate": 1.3661067248494586e-06, "loss": 0.022, "step": 55750 }, { "grad_norm": 0.4075110852718353, "learning_rate": 1.3597163155552429e-06, "loss": 0.0534, "step": 55760 }, { "grad_norm": 0.3423860967159271, "learning_rate": 1.3533406818949434e-06, "loss": 0.0226, "step": 55770 }, { "grad_norm": 0.3468983471393585, "learning_rate": 1.3469798258053002e-06, "loss": 0.0235, "step": 55780 }, { "grad_norm": 0.5018434524536133, "learning_rate": 1.3406337492185672e-06, "loss": 0.019, "step": 55790 }, { "grad_norm": 0.3302702009677887, "learning_rate": 1.3343024540625414e-06, "loss": 0.0279, "step": 55800 }, { "grad_norm": 0.7797746658325195, "learning_rate": 1.3279859422604735e-06, "loss": 0.0224, "step": 55810 }, { "grad_norm": 0.4964252710342407, "learning_rate": 1.3216842157311781e-06, "loss": 0.0258, "step": 55820 }, { "grad_norm": 0.3383857011795044, "learning_rate": 1.3153972763889355e-06, "loss": 0.0252, "step": 55830 }, { "grad_norm": 0.633790135383606, "learning_rate": 1.3091251261435566e-06, "loss": 0.0229, "step": 55840 }, { "grad_norm": 0.26179754734039307, "learning_rate": 1.3028677669003564e-06, "loss": 0.0215, "step": 55850 }, { "grad_norm": 0.29648876190185547, "learning_rate": 1.2966252005601587e-06, "loss": 0.0271, "step": 55860 }, { "grad_norm": 0.2638907730579376, "learning_rate": 1.2903974290192855e-06, "loss": 0.0251, "step": 55870 }, { "grad_norm": 0.1782074272632599, "learning_rate": 1.284184454169568e-06, "loss": 0.0246, "step": 55880 }, { "grad_norm": 0.3466293513774872, "learning_rate": 1.2779862778983464e-06, "loss": 0.0247, "step": 55890 }, { "grad_norm": 0.3209560215473175, "learning_rate": 1.2718029020884647e-06, "loss": 0.0256, "step": 55900 }, { "grad_norm": 0.2476983666419983, "learning_rate": 1.2656343286182703e-06, "loss": 0.0196, "step": 55910 }, { "grad_norm": 0.2742520272731781, "learning_rate": 1.2594805593616088e-06, "loss": 0.0401, "step": 55920 }, { "grad_norm": 0.45482730865478516, "learning_rate": 1.2533415961878404e-06, "loss": 0.0245, "step": 55930 }, { "grad_norm": 0.3201555013656616, "learning_rate": 1.2472174409618009e-06, "loss": 0.0205, "step": 55940 }, { "grad_norm": 0.3277834951877594, "learning_rate": 1.2411080955438747e-06, "loss": 0.0184, "step": 55950 }, { "grad_norm": 0.7753558158874512, "learning_rate": 1.235013561789894e-06, "loss": 0.0201, "step": 55960 }, { "grad_norm": 0.4912470877170563, "learning_rate": 1.2289338415512385e-06, "loss": 0.0223, "step": 55970 }, { "grad_norm": 0.2982291281223297, "learning_rate": 1.222868936674748e-06, "loss": 0.0226, "step": 55980 }, { "grad_norm": 0.4748316705226898, "learning_rate": 1.2168188490027876e-06, "loss": 0.023, "step": 55990 }, { "grad_norm": 0.24616914987564087, "learning_rate": 1.2107835803732204e-06, "loss": 0.0265, "step": 56000 }, { "grad_norm": 0.38188794255256653, "learning_rate": 1.2047631326193964e-06, "loss": 0.0229, "step": 56010 }, { "grad_norm": 0.35126352310180664, "learning_rate": 1.1987575075701696e-06, "loss": 0.0206, "step": 56020 }, { "grad_norm": 0.5014798641204834, "learning_rate": 1.1927667070498916e-06, "loss": 0.0345, "step": 56030 }, { "grad_norm": 0.5228205323219299, "learning_rate": 1.1867907328784067e-06, "loss": 0.0282, "step": 56040 }, { "grad_norm": 0.4363141357898712, "learning_rate": 1.1808295868710518e-06, "loss": 0.0225, "step": 56050 }, { "grad_norm": 0.5075777769088745, "learning_rate": 1.174883270838678e-06, "loss": 0.0266, "step": 56060 }, { "grad_norm": 0.2852615714073181, "learning_rate": 1.1689517865876187e-06, "loss": 0.0215, "step": 56070 }, { "grad_norm": 0.30555012822151184, "learning_rate": 1.1630351359196933e-06, "loss": 0.0253, "step": 56080 }, { "grad_norm": 0.3338826894760132, "learning_rate": 1.1571333206322255e-06, "loss": 0.0233, "step": 56090 }, { "grad_norm": 0.30191582441329956, "learning_rate": 1.1512463425180365e-06, "loss": 0.0195, "step": 56100 }, { "grad_norm": 0.31393060088157654, "learning_rate": 1.145374203365429e-06, "loss": 0.0264, "step": 56110 }, { "grad_norm": 0.31367143988609314, "learning_rate": 1.1395169049582155e-06, "loss": 0.0196, "step": 56120 }, { "grad_norm": 0.2803199291229248, "learning_rate": 1.1336744490756722e-06, "loss": 0.0293, "step": 56130 }, { "grad_norm": 0.3240787982940674, "learning_rate": 1.1278468374925967e-06, "loss": 0.0195, "step": 56140 }, { "grad_norm": 0.22824452817440033, "learning_rate": 1.12203407197925e-06, "loss": 0.0213, "step": 56150 }, { "grad_norm": 0.3659059703350067, "learning_rate": 1.116236154301409e-06, "loss": 0.0187, "step": 56160 }, { "grad_norm": 0.609093189239502, "learning_rate": 1.11045308622032e-06, "loss": 0.0192, "step": 56170 }, { "grad_norm": 0.3387521207332611, "learning_rate": 1.1046848694927337e-06, "loss": 0.0254, "step": 56180 }, { "grad_norm": 0.34094685316085815, "learning_rate": 1.098931505870876e-06, "loss": 0.0264, "step": 56190 }, { "grad_norm": 0.25737079977989197, "learning_rate": 1.0931929971024657e-06, "loss": 0.0156, "step": 56200 }, { "grad_norm": 0.3665902614593506, "learning_rate": 1.0874693449307193e-06, "loss": 0.0194, "step": 56210 }, { "grad_norm": 0.3138740360736847, "learning_rate": 1.081760551094324e-06, "loss": 0.0224, "step": 56220 }, { "grad_norm": 0.6203614473342896, "learning_rate": 1.0760666173274592e-06, "loss": 0.0257, "step": 56230 }, { "grad_norm": 0.5183113813400269, "learning_rate": 1.0703875453597967e-06, "loss": 0.0221, "step": 56240 }, { "grad_norm": 0.3392391502857208, "learning_rate": 1.0647233369164845e-06, "loss": 0.0207, "step": 56250 }, { "grad_norm": 0.22200360894203186, "learning_rate": 1.0590739937181625e-06, "loss": 0.0171, "step": 56260 }, { "grad_norm": 0.7096726894378662, "learning_rate": 1.053439517480953e-06, "loss": 0.0265, "step": 56270 }, { "grad_norm": 0.44529062509536743, "learning_rate": 1.047819909916453e-06, "loss": 0.0221, "step": 56280 }, { "grad_norm": 0.13778239488601685, "learning_rate": 1.0422151727317697e-06, "loss": 0.0226, "step": 56290 }, { "grad_norm": 0.3475687503814697, "learning_rate": 1.0366253076294462e-06, "loss": 0.0495, "step": 56300 }, { "grad_norm": 0.4621337354183197, "learning_rate": 1.031050316307558e-06, "loss": 0.03, "step": 56310 }, { "grad_norm": 0.48173195123672485, "learning_rate": 1.0254902004596333e-06, "loss": 0.0311, "step": 56320 }, { "grad_norm": 0.5077344179153442, "learning_rate": 1.0199449617746882e-06, "loss": 0.0221, "step": 56330 }, { "grad_norm": 0.22584843635559082, "learning_rate": 1.0144146019372247e-06, "loss": 0.019, "step": 56340 }, { "grad_norm": 0.28258460760116577, "learning_rate": 1.0088991226272048e-06, "loss": 0.0393, "step": 56350 }, { "grad_norm": 0.4514908492565155, "learning_rate": 1.003398525520105e-06, "loss": 0.0202, "step": 56360 }, { "grad_norm": 0.2838647663593292, "learning_rate": 9.979128122868552e-07, "loss": 0.0357, "step": 56370 }, { "grad_norm": 0.389616996049881, "learning_rate": 9.924419845938614e-07, "loss": 0.0362, "step": 56380 }, { "grad_norm": 0.5331401228904724, "learning_rate": 9.869860441030276e-07, "loss": 0.0209, "step": 56390 }, { "grad_norm": 0.387500524520874, "learning_rate": 9.815449924717169e-07, "loss": 0.0288, "step": 56400 }, { "grad_norm": 0.22138813138008118, "learning_rate": 9.761188313527791e-07, "loss": 0.0194, "step": 56410 }, { "grad_norm": 0.12184477597475052, "learning_rate": 9.70707562394546e-07, "loss": 0.0157, "step": 56420 }, { "grad_norm": 0.6847772002220154, "learning_rate": 9.653111872408027e-07, "loss": 0.0208, "step": 56430 }, { "grad_norm": 0.3645419478416443, "learning_rate": 9.599297075308434e-07, "loss": 0.0281, "step": 56440 }, { "grad_norm": 0.41616323590278625, "learning_rate": 9.545631248994048e-07, "loss": 0.0207, "step": 56450 }, { "grad_norm": 0.1874409168958664, "learning_rate": 9.492114409767217e-07, "loss": 0.0315, "step": 56460 }, { "grad_norm": 0.2449888288974762, "learning_rate": 9.438746573884938e-07, "loss": 0.0219, "step": 56470 }, { "grad_norm": 0.37077274918556213, "learning_rate": 9.385527757558909e-07, "loss": 0.0253, "step": 56480 }, { "grad_norm": 0.3512401878833771, "learning_rate": 9.332457976955644e-07, "loss": 0.0246, "step": 56490 }, { "grad_norm": 0.3631170094013214, "learning_rate": 9.279537248196247e-07, "loss": 0.0219, "step": 56500 }, { "grad_norm": 0.34523314237594604, "learning_rate": 9.2267655873568e-07, "loss": 0.0247, "step": 56510 }, { "grad_norm": 0.30596107244491577, "learning_rate": 9.174143010467762e-07, "loss": 0.0224, "step": 56520 }, { "grad_norm": 0.19713059067726135, "learning_rate": 9.12166953351462e-07, "loss": 0.019, "step": 56530 }, { "grad_norm": 0.5589758157730103, "learning_rate": 9.069345172437404e-07, "loss": 0.0156, "step": 56540 }, { "grad_norm": 0.3503514528274536, "learning_rate": 9.017169943130843e-07, "loss": 0.0301, "step": 56550 }, { "grad_norm": 0.39986753463745117, "learning_rate": 8.965143861444425e-07, "loss": 0.0184, "step": 56560 }, { "grad_norm": 0.2201024889945984, "learning_rate": 8.91326694318223e-07, "loss": 0.0175, "step": 56570 }, { "grad_norm": 0.1550266146659851, "learning_rate": 8.861539204103098e-07, "loss": 0.015, "step": 56580 }, { "grad_norm": 0.6291810274124146, "learning_rate": 8.809960659920735e-07, "loss": 0.0272, "step": 56590 }, { "grad_norm": 0.330841988325119, "learning_rate": 8.758531326303055e-07, "loss": 0.0218, "step": 56600 }, { "grad_norm": 0.2524101734161377, "learning_rate": 8.707251218873169e-07, "loss": 0.0194, "step": 56610 }, { "grad_norm": 0.31094127893447876, "learning_rate": 8.656120353208507e-07, "loss": 0.024, "step": 56620 }, { "grad_norm": 0.26676055788993835, "learning_rate": 8.605138744841312e-07, "loss": 0.0219, "step": 56630 }, { "grad_norm": 0.3156278729438782, "learning_rate": 8.554306409258417e-07, "loss": 0.0196, "step": 56640 }, { "grad_norm": 0.4700492322444916, "learning_rate": 8.503623361901358e-07, "loss": 0.0218, "step": 56650 }, { "grad_norm": 0.55814528465271, "learning_rate": 8.453089618166377e-07, "loss": 0.032, "step": 56660 }, { "grad_norm": 0.3355361223220825, "learning_rate": 8.402705193404137e-07, "loss": 0.0198, "step": 56670 }, { "grad_norm": 0.4723540246486664, "learning_rate": 8.352470102920174e-07, "loss": 0.0147, "step": 56680 }, { "grad_norm": 0.5556483268737793, "learning_rate": 8.302384361974669e-07, "loss": 0.0197, "step": 56690 }, { "grad_norm": 0.48099878430366516, "learning_rate": 8.252447985782231e-07, "loss": 0.0213, "step": 56700 }, { "grad_norm": 0.3845166265964508, "learning_rate": 8.202660989512279e-07, "loss": 0.0183, "step": 56710 }, { "grad_norm": 0.4772244393825531, "learning_rate": 8.153023388288772e-07, "loss": 0.0204, "step": 56720 }, { "grad_norm": 0.22163651883602142, "learning_rate": 8.103535197190204e-07, "loss": 0.0249, "step": 56730 }, { "grad_norm": 0.24585109949111938, "learning_rate": 8.054196431249993e-07, "loss": 0.0223, "step": 56740 }, { "grad_norm": 0.5674407482147217, "learning_rate": 8.005007105455709e-07, "loss": 0.0188, "step": 56750 }, { "grad_norm": 0.32637107372283936, "learning_rate": 7.95596723474995e-07, "loss": 0.0351, "step": 56760 }, { "grad_norm": 0.37127071619033813, "learning_rate": 7.907076834029692e-07, "loss": 0.0229, "step": 56770 }, { "grad_norm": 0.29846930503845215, "learning_rate": 7.858335918146498e-07, "loss": 0.0276, "step": 56780 }, { "grad_norm": 0.2815841734409332, "learning_rate": 7.809744501906635e-07, "loss": 0.0202, "step": 56790 }, { "grad_norm": 0.42195624113082886, "learning_rate": 7.761302600070797e-07, "loss": 0.0226, "step": 56800 }, { "grad_norm": 0.4435247778892517, "learning_rate": 7.713010227354545e-07, "loss": 0.0241, "step": 56810 }, { "grad_norm": 0.4129710793495178, "learning_rate": 7.664867398427589e-07, "loss": 0.0228, "step": 56820 }, { "grad_norm": 0.17795515060424805, "learning_rate": 7.616874127914619e-07, "loss": 0.0243, "step": 56830 }, { "grad_norm": 0.1747518926858902, "learning_rate": 7.569030430394641e-07, "loss": 0.0279, "step": 56840 }, { "grad_norm": 0.24271142482757568, "learning_rate": 7.521336320401306e-07, "loss": 0.0197, "step": 56850 }, { "grad_norm": 0.42234209179878235, "learning_rate": 7.473791812422915e-07, "loss": 0.0195, "step": 56860 }, { "grad_norm": 0.7452623248100281, "learning_rate": 7.426396920902134e-07, "loss": 0.0308, "step": 56870 }, { "grad_norm": 0.5264500975608826, "learning_rate": 7.379151660236283e-07, "loss": 0.0232, "step": 56880 }, { "grad_norm": 0.4543452858924866, "learning_rate": 7.332056044777324e-07, "loss": 0.0267, "step": 56890 }, { "grad_norm": 0.30336955189704895, "learning_rate": 7.285110088831537e-07, "loss": 0.0237, "step": 56900 }, { "grad_norm": 0.19872847199440002, "learning_rate": 7.238313806659902e-07, "loss": 0.0173, "step": 56910 }, { "grad_norm": 0.49673017859458923, "learning_rate": 7.191667212477993e-07, "loss": 0.0226, "step": 56920 }, { "grad_norm": 0.32020944356918335, "learning_rate": 7.145170320455697e-07, "loss": 0.0164, "step": 56930 }, { "grad_norm": 0.41849014163017273, "learning_rate": 7.098823144717604e-07, "loss": 0.0226, "step": 56940 }, { "grad_norm": 0.2386217564344406, "learning_rate": 7.052625699342674e-07, "loss": 0.037, "step": 56950 }, { "grad_norm": 0.34145796298980713, "learning_rate": 7.006577998364628e-07, "loss": 0.0244, "step": 56960 }, { "grad_norm": 0.5131716728210449, "learning_rate": 6.96068005577133e-07, "loss": 0.0229, "step": 56970 }, { "grad_norm": 0.27744320034980774, "learning_rate": 6.914931885505627e-07, "loss": 0.025, "step": 56980 }, { "grad_norm": 0.24832698702812195, "learning_rate": 6.869333501464347e-07, "loss": 0.029, "step": 56990 }, { "grad_norm": 0.2525058686733246, "learning_rate": 6.823884917499246e-07, "loss": 0.0193, "step": 57000 }, { "grad_norm": 0.4676513075828552, "learning_rate": 6.778586147416278e-07, "loss": 0.0237, "step": 57010 }, { "grad_norm": 0.34163591265678406, "learning_rate": 6.733437204976156e-07, "loss": 0.0153, "step": 57020 }, { "grad_norm": 0.236243337392807, "learning_rate": 6.688438103893857e-07, "loss": 0.023, "step": 57030 }, { "grad_norm": 0.31259989738464355, "learning_rate": 6.64358885783889e-07, "loss": 0.0251, "step": 57040 }, { "grad_norm": 0.6813659071922302, "learning_rate": 6.598889480435299e-07, "loss": 0.0299, "step": 57050 }, { "grad_norm": 0.5120483040809631, "learning_rate": 6.554339985261615e-07, "loss": 0.0304, "step": 57060 }, { "grad_norm": 0.4823845326900482, "learning_rate": 6.509940385850733e-07, "loss": 0.0219, "step": 57070 }, { "grad_norm": 0.3797207474708557, "learning_rate": 6.465690695690141e-07, "loss": 0.0265, "step": 57080 }, { "grad_norm": 0.5648642778396606, "learning_rate": 6.421590928221699e-07, "loss": 0.0239, "step": 57090 }, { "grad_norm": 0.3990775942802429, "learning_rate": 6.377641096841691e-07, "loss": 0.0188, "step": 57100 }, { "grad_norm": 0.2710058391094208, "learning_rate": 6.333841214901048e-07, "loss": 0.0217, "step": 57110 }, { "grad_norm": 0.37186816334724426, "learning_rate": 6.290191295704906e-07, "loss": 0.0225, "step": 57120 }, { "grad_norm": 0.39731454849243164, "learning_rate": 6.246691352513046e-07, "loss": 0.0474, "step": 57130 }, { "grad_norm": 0.5997823476791382, "learning_rate": 6.203341398539452e-07, "loss": 0.0262, "step": 57140 }, { "grad_norm": 0.3259073793888092, "learning_rate": 6.160141446952872e-07, "loss": 0.0358, "step": 57150 }, { "grad_norm": 0.5412655472755432, "learning_rate": 6.11709151087625e-07, "loss": 0.0176, "step": 57160 }, { "grad_norm": 0.3314434885978699, "learning_rate": 6.074191603386958e-07, "loss": 0.0235, "step": 57170 }, { "grad_norm": 0.21369192004203796, "learning_rate": 6.031441737516907e-07, "loss": 0.0179, "step": 57180 }, { "grad_norm": 0.4960324466228485, "learning_rate": 5.988841926252431e-07, "loss": 0.0273, "step": 57190 }, { "grad_norm": 0.4055357873439789, "learning_rate": 5.946392182534066e-07, "loss": 0.0168, "step": 57200 }, { "grad_norm": 0.25729966163635254, "learning_rate": 5.90409251925711e-07, "loss": 0.0332, "step": 57210 }, { "grad_norm": 0.31043386459350586, "learning_rate": 5.861942949270949e-07, "loss": 0.0289, "step": 57220 }, { "grad_norm": 0.5796815156936646, "learning_rate": 5.819943485379564e-07, "loss": 0.02, "step": 57230 }, { "grad_norm": 0.3944372236728668, "learning_rate": 5.778094140341306e-07, "loss": 0.0304, "step": 57240 }, { "grad_norm": 0.5753656625747681, "learning_rate": 5.736394926868893e-07, "loss": 0.0214, "step": 57250 }, { "grad_norm": 0.5333942770957947, "learning_rate": 5.694845857629416e-07, "loss": 0.0269, "step": 57260 }, { "grad_norm": 0.4079539477825165, "learning_rate": 5.653446945244334e-07, "loss": 0.0237, "step": 57270 }, { "grad_norm": 0.1459839940071106, "learning_rate": 5.612198202289698e-07, "loss": 0.0195, "step": 57280 }, { "grad_norm": 0.5037033557891846, "learning_rate": 5.571099641295596e-07, "loss": 0.0203, "step": 57290 }, { "grad_norm": 0.41038310527801514, "learning_rate": 5.530151274746875e-07, "loss": 0.0198, "step": 57300 }, { "grad_norm": 0.36293619871139526, "learning_rate": 5.489353115082418e-07, "loss": 0.0216, "step": 57310 }, { "grad_norm": 0.5453184247016907, "learning_rate": 5.4487051746957e-07, "loss": 0.0334, "step": 57320 }, { "grad_norm": 0.2760009169578552, "learning_rate": 5.408207465934511e-07, "loss": 0.0226, "step": 57330 }, { "grad_norm": 0.664505124092102, "learning_rate": 5.3678600011009e-07, "loss": 0.0265, "step": 57340 }, { "grad_norm": 0.4057667553424835, "learning_rate": 5.327662792451449e-07, "loss": 0.0296, "step": 57350 }, { "grad_norm": 0.35644975304603577, "learning_rate": 5.287615852196947e-07, "loss": 0.024, "step": 57360 }, { "grad_norm": 0.29775089025497437, "learning_rate": 5.247719192502665e-07, "loss": 0.0281, "step": 57370 }, { "grad_norm": 0.5043103098869324, "learning_rate": 5.207972825488128e-07, "loss": 0.0174, "step": 57380 }, { "grad_norm": 0.4233013987541199, "learning_rate": 5.168376763227178e-07, "loss": 0.0216, "step": 57390 }, { "grad_norm": 0.43769070506095886, "learning_rate": 5.12893101774814e-07, "loss": 0.026, "step": 57400 }, { "grad_norm": 0.2904359996318817, "learning_rate": 5.089635601033483e-07, "loss": 0.0182, "step": 57410 }, { "grad_norm": 0.36585116386413574, "learning_rate": 5.050490525020213e-07, "loss": 0.0184, "step": 57420 }, { "grad_norm": 0.392522394657135, "learning_rate": 5.011495801599541e-07, "loss": 0.0213, "step": 57430 }, { "grad_norm": 0.2968052625656128, "learning_rate": 4.972651442616994e-07, "loss": 0.0353, "step": 57440 }, { "grad_norm": 0.4200272858142853, "learning_rate": 4.933957459872574e-07, "loss": 0.0266, "step": 57450 }, { "grad_norm": 0.24227969348430634, "learning_rate": 4.895413865120324e-07, "loss": 0.0521, "step": 57460 }, { "grad_norm": 0.7099030017852783, "learning_rate": 4.857020670068935e-07, "loss": 0.0197, "step": 57470 }, { "grad_norm": 0.15394853055477142, "learning_rate": 4.818777886381132e-07, "loss": 0.0172, "step": 57480 }, { "grad_norm": 0.8188257813453674, "learning_rate": 4.780685525674122e-07, "loss": 0.0279, "step": 57490 }, { "grad_norm": 0.49103251099586487, "learning_rate": 4.7427435995193724e-07, "loss": 0.0259, "step": 57500 }, { "grad_norm": 0.16626934707164764, "learning_rate": 4.7049521194425515e-07, "loss": 0.0356, "step": 57510 }, { "grad_norm": 0.5336379408836365, "learning_rate": 4.667311096923754e-07, "loss": 0.0235, "step": 57520 }, { "grad_norm": 0.41958752274513245, "learning_rate": 4.6298205433973895e-07, "loss": 0.0279, "step": 57530 }, { "grad_norm": 0.46347150206565857, "learning_rate": 4.5924804702520696e-07, "loss": 0.0204, "step": 57540 }, { "grad_norm": 0.33572113513946533, "learning_rate": 4.5552908888306655e-07, "loss": 0.0204, "step": 57550 }, { "grad_norm": 0.4263055622577667, "learning_rate": 4.5182518104304185e-07, "loss": 0.0244, "step": 57560 }, { "grad_norm": 0.4134148359298706, "learning_rate": 4.4813632463028277e-07, "loss": 0.0227, "step": 57570 }, { "grad_norm": 0.1464596837759018, "learning_rate": 4.444625207653763e-07, "loss": 0.0267, "step": 57580 }, { "grad_norm": 0.28686150908470154, "learning_rate": 4.4080377056430753e-07, "loss": 0.0262, "step": 57590 }, { "grad_norm": 0.5221957564353943, "learning_rate": 4.371600751385263e-07, "loss": 0.0269, "step": 57600 }, { "grad_norm": 0.46036458015441895, "learning_rate": 4.3353143559487495e-07, "loss": 0.024, "step": 57610 }, { "grad_norm": 0.3246321976184845, "learning_rate": 4.2991785303565513e-07, "loss": 0.0257, "step": 57620 }, { "grad_norm": 0.4070093333721161, "learning_rate": 4.2631932855856647e-07, "loss": 0.0188, "step": 57630 }, { "grad_norm": 0.2740495502948761, "learning_rate": 4.2273586325674576e-07, "loss": 0.025, "step": 57640 }, { "grad_norm": 0.35179734230041504, "learning_rate": 4.1916745821876103e-07, "loss": 0.0273, "step": 57650 }, { "grad_norm": 0.2532016336917877, "learning_rate": 4.156141145285897e-07, "loss": 0.021, "step": 57660 }, { "grad_norm": 0.41349923610687256, "learning_rate": 4.1207583326566267e-07, "loss": 0.0251, "step": 57670 }, { "grad_norm": 0.5598593354225159, "learning_rate": 4.085526155047925e-07, "loss": 0.0226, "step": 57680 }, { "grad_norm": 0.6206179857254028, "learning_rate": 4.050444623162564e-07, "loss": 0.026, "step": 57690 }, { "grad_norm": 0.47809934616088867, "learning_rate": 4.015513747657351e-07, "loss": 0.0311, "step": 57700 }, { "grad_norm": 0.2496323436498642, "learning_rate": 3.9807335391433554e-07, "loss": 0.0243, "step": 57710 }, { "grad_norm": 0.44998571276664734, "learning_rate": 3.946104008185847e-07, "loss": 0.0217, "step": 57720 }, { "grad_norm": 0.72134929895401, "learning_rate": 3.9116251653044113e-07, "loss": 0.0244, "step": 57730 }, { "grad_norm": 0.7068828344345093, "learning_rate": 3.877297020972781e-07, "loss": 0.0284, "step": 57740 }, { "grad_norm": 0.20914529263973236, "learning_rate": 3.8431195856190036e-07, "loss": 0.0311, "step": 57750 }, { "grad_norm": 0.26468223333358765, "learning_rate": 3.8090928696251635e-07, "loss": 0.0219, "step": 57760 }, { "grad_norm": 0.3769214153289795, "learning_rate": 3.77521688332777e-07, "loss": 0.0184, "step": 57770 }, { "grad_norm": 0.3629470467567444, "learning_rate": 3.741491637017425e-07, "loss": 0.0237, "step": 57780 }, { "grad_norm": 0.4475114643573761, "learning_rate": 3.707917140939043e-07, "loss": 0.0204, "step": 57790 }, { "grad_norm": 0.3775034248828888, "learning_rate": 3.6744934052915235e-07, "loss": 0.0166, "step": 57800 }, { "grad_norm": 0.3625791668891907, "learning_rate": 3.641220440228188e-07, "loss": 0.0256, "step": 57810 }, { "grad_norm": 0.39736640453338623, "learning_rate": 3.608098255856562e-07, "loss": 0.0271, "step": 57820 }, { "grad_norm": 0.3751695156097412, "learning_rate": 3.575126862238154e-07, "loss": 0.021, "step": 57830 }, { "grad_norm": 0.3062500059604645, "learning_rate": 3.5423062693888955e-07, "loss": 0.0262, "step": 57840 }, { "grad_norm": 0.29355746507644653, "learning_rate": 3.509636487278756e-07, "loss": 0.0244, "step": 57850 }, { "grad_norm": 0.3399164080619812, "learning_rate": 3.4771175258320186e-07, "loss": 0.0281, "step": 57860 }, { "grad_norm": 0.6622183322906494, "learning_rate": 3.4447493949270047e-07, "loss": 0.0235, "step": 57870 }, { "grad_norm": 0.5541637539863586, "learning_rate": 3.4125321043964045e-07, "loss": 0.0231, "step": 57880 }, { "grad_norm": 0.34389394521713257, "learning_rate": 3.380465664026833e-07, "loss": 0.0268, "step": 57890 }, { "grad_norm": 0.14113502204418182, "learning_rate": 3.348550083559388e-07, "loss": 0.0262, "step": 57900 }, { "grad_norm": 0.4239943325519562, "learning_rate": 3.316785372689091e-07, "loss": 0.0326, "step": 57910 }, { "grad_norm": 0.2519177496433258, "learning_rate": 3.285171541065224e-07, "loss": 0.0217, "step": 57920 }, { "grad_norm": 0.4562682807445526, "learning_rate": 3.253708598291272e-07, "loss": 0.027, "step": 57930 }, { "grad_norm": 0.27829039096832275, "learning_rate": 3.2223965539248116e-07, "loss": 0.0188, "step": 57940 }, { "grad_norm": 0.23771744966506958, "learning_rate": 3.1912354174776227e-07, "loss": 0.0191, "step": 57950 }, { "grad_norm": 0.17619802057743073, "learning_rate": 3.1602251984155786e-07, "loss": 0.0192, "step": 57960 }, { "grad_norm": 0.3471243679523468, "learning_rate": 3.1293659061589207e-07, "loss": 0.0166, "step": 57970 }, { "grad_norm": 0.35565635561943054, "learning_rate": 3.098657550081707e-07, "loss": 0.0268, "step": 57980 }, { "grad_norm": 0.36116281151771545, "learning_rate": 3.068100139512475e-07, "loss": 0.0208, "step": 57990 }, { "grad_norm": 0.4543524384498596, "learning_rate": 3.037693683733689e-07, "loss": 0.0313, "step": 58000 }, { "grad_norm": 0.25459152460098267, "learning_rate": 3.007438191982015e-07, "loss": 0.0185, "step": 58010 }, { "grad_norm": 0.3488805890083313, "learning_rate": 2.9773336734482684e-07, "loss": 0.0163, "step": 58020 }, { "grad_norm": 0.6302743554115295, "learning_rate": 2.9473801372774667e-07, "loss": 0.0257, "step": 58030 }, { "grad_norm": 0.17898699641227722, "learning_rate": 2.91757759256861e-07, "loss": 0.0207, "step": 58040 }, { "grad_norm": 0.308952659368515, "learning_rate": 2.887926048375067e-07, "loss": 0.0217, "step": 58050 }, { "grad_norm": 0.20356805622577667, "learning_rate": 2.858425513704022e-07, "loss": 0.0199, "step": 58060 }, { "grad_norm": 0.452865868806839, "learning_rate": 2.8290759975170834e-07, "loss": 0.0213, "step": 58070 }, { "grad_norm": 0.1813037246465683, "learning_rate": 2.799877508729787e-07, "loss": 0.02, "step": 58080 }, { "grad_norm": 0.6826667189598083, "learning_rate": 2.770830056211926e-07, "loss": 0.0197, "step": 58090 }, { "grad_norm": 0.3682693541049957, "learning_rate": 2.741933648787331e-07, "loss": 0.0253, "step": 58100 }, { "grad_norm": 0.33776313066482544, "learning_rate": 2.7131882952339263e-07, "loss": 0.0239, "step": 58110 }, { "grad_norm": 0.4532252252101898, "learning_rate": 2.684594004283836e-07, "loss": 0.0289, "step": 58120 }, { "grad_norm": 0.6123549342155457, "learning_rate": 2.6561507846232234e-07, "loss": 0.0219, "step": 58130 }, { "grad_norm": 0.2411874681711197, "learning_rate": 2.6278586448924005e-07, "loss": 0.021, "step": 58140 }, { "grad_norm": 0.1451924443244934, "learning_rate": 2.5997175936857685e-07, "loss": 0.0176, "step": 58150 }, { "grad_norm": 0.4230889678001404, "learning_rate": 2.57172763955188e-07, "loss": 0.0259, "step": 58160 }, { "grad_norm": 0.44367364048957825, "learning_rate": 2.543888790993265e-07, "loss": 0.0239, "step": 58170 }, { "grad_norm": 0.4188757538795471, "learning_rate": 2.5162010564666607e-07, "loss": 0.0166, "step": 58180 }, { "grad_norm": 0.7136043310165405, "learning_rate": 2.488664444382893e-07, "loss": 0.026, "step": 58190 }, { "grad_norm": 0.3375810384750366, "learning_rate": 2.461278963106828e-07, "loss": 0.0258, "step": 58200 }, { "grad_norm": 0.43348363041877747, "learning_rate": 2.434044620957421e-07, "loss": 0.0163, "step": 58210 }, { "grad_norm": 0.26845410466194153, "learning_rate": 2.406961426207832e-07, "loss": 0.0164, "step": 58220 }, { "grad_norm": 0.8230626583099365, "learning_rate": 2.380029387085203e-07, "loss": 0.0205, "step": 58230 }, { "grad_norm": 0.2671978175640106, "learning_rate": 2.353248511770767e-07, "loss": 0.0201, "step": 58240 }, { "grad_norm": 0.48881086707115173, "learning_rate": 2.3266188083997942e-07, "loss": 0.0267, "step": 58250 }, { "grad_norm": 0.41928595304489136, "learning_rate": 2.3001402850617027e-07, "loss": 0.0232, "step": 58260 }, { "grad_norm": 0.3761276602745056, "learning_rate": 2.2738129498000581e-07, "loss": 0.0174, "step": 58270 }, { "grad_norm": 0.245551198720932, "learning_rate": 2.2476368106122414e-07, "loss": 0.0238, "step": 58280 }, { "grad_norm": 0.397356241941452, "learning_rate": 2.2216118754500582e-07, "loss": 0.0195, "step": 58290 }, { "grad_norm": 0.33498120307922363, "learning_rate": 2.1957381522190735e-07, "loss": 0.0189, "step": 58300 }, { "grad_norm": 0.23605285584926605, "learning_rate": 2.1700156487790558e-07, "loss": 0.0301, "step": 58310 }, { "grad_norm": 0.3663305640220642, "learning_rate": 2.1444443729439213e-07, "loss": 0.0225, "step": 58320 }, { "grad_norm": 0.5954736471176147, "learning_rate": 2.1190243324814007e-07, "loss": 0.0235, "step": 58330 }, { "grad_norm": 0.41660356521606445, "learning_rate": 2.0937555351135395e-07, "loss": 0.0331, "step": 58340 }, { "grad_norm": 0.5904051661491394, "learning_rate": 2.0686379885162532e-07, "loss": 0.0176, "step": 58350 }, { "grad_norm": 0.3953811526298523, "learning_rate": 2.0436717003196604e-07, "loss": 0.0178, "step": 58360 }, { "grad_norm": 0.19811765849590302, "learning_rate": 2.0188566781078057e-07, "loss": 0.0105, "step": 58370 }, { "grad_norm": 0.3012900948524475, "learning_rate": 1.994192929418881e-07, "loss": 0.0298, "step": 58380 }, { "grad_norm": 0.4455496668815613, "learning_rate": 1.9696804617451158e-07, "loss": 0.0283, "step": 58390 }, { "grad_norm": 0.37105464935302734, "learning_rate": 1.9453192825326093e-07, "loss": 0.0245, "step": 58400 }, { "grad_norm": 0.8953291177749634, "learning_rate": 1.9211093991817753e-07, "loss": 0.0262, "step": 58410 }, { "grad_norm": 0.5062271952629089, "learning_rate": 1.8970508190468973e-07, "loss": 0.0226, "step": 58420 }, { "grad_norm": 0.41742706298828125, "learning_rate": 1.8731435494362958e-07, "loss": 0.0365, "step": 58430 }, { "grad_norm": 0.3567653298377991, "learning_rate": 1.849387597612495e-07, "loss": 0.0286, "step": 58440 }, { "grad_norm": 0.2158624678850174, "learning_rate": 1.8257829707917228e-07, "loss": 0.0239, "step": 58450 }, { "grad_norm": 0.49696335196495056, "learning_rate": 1.8023296761446317e-07, "loss": 0.0194, "step": 58460 }, { "grad_norm": 0.23134703934192657, "learning_rate": 1.7790277207956341e-07, "loss": 0.0432, "step": 58470 }, { "grad_norm": 0.33594751358032227, "learning_rate": 1.7558771118232343e-07, "loss": 0.0205, "step": 58480 }, { "grad_norm": 0.6887562870979309, "learning_rate": 1.7328778562599734e-07, "loss": 0.0244, "step": 58490 }, { "grad_norm": 0.4914855360984802, "learning_rate": 1.7100299610924298e-07, "loss": 0.0183, "step": 58500 }, { "grad_norm": 0.268344908952713, "learning_rate": 1.6873334332612733e-07, "loss": 0.0266, "step": 58510 }, { "grad_norm": 0.5228548049926758, "learning_rate": 1.6647882796609894e-07, "loss": 0.031, "step": 58520 }, { "grad_norm": 0.3604435622692108, "learning_rate": 1.6423945071402102e-07, "loss": 0.0211, "step": 58530 }, { "grad_norm": 0.23637336492538452, "learning_rate": 1.6201521225016614e-07, "loss": 0.0268, "step": 58540 }, { "grad_norm": 0.27040529251098633, "learning_rate": 1.598061132501938e-07, "loss": 0.032, "step": 58550 }, { "grad_norm": 0.36819732189178467, "learning_rate": 1.576121543851672e-07, "loss": 0.0179, "step": 58560 }, { "grad_norm": 0.5365504622459412, "learning_rate": 1.5543333632155876e-07, "loss": 0.0316, "step": 58570 }, { "grad_norm": 0.4415118992328644, "learning_rate": 1.5326965972123352e-07, "loss": 0.0204, "step": 58580 }, { "grad_norm": 0.5892996788024902, "learning_rate": 1.5112112524146016e-07, "loss": 0.0222, "step": 58590 }, { "grad_norm": 0.25041329860687256, "learning_rate": 1.4898773353489992e-07, "loss": 0.0123, "step": 58600 }, { "grad_norm": 0.2286396473646164, "learning_rate": 1.4686948524962884e-07, "loss": 0.0246, "step": 58610 }, { "grad_norm": 0.28987205028533936, "learning_rate": 1.4476638102911556e-07, "loss": 0.0257, "step": 58620 }, { "grad_norm": 0.3088963031768799, "learning_rate": 1.4267842151222123e-07, "loss": 0.0198, "step": 58630 }, { "grad_norm": 0.28364771604537964, "learning_rate": 1.4060560733321626e-07, "loss": 0.017, "step": 58640 }, { "grad_norm": 0.15668796002864838, "learning_rate": 1.385479391217692e-07, "loss": 0.0112, "step": 58650 }, { "grad_norm": 0.4399748146533966, "learning_rate": 1.365054175029412e-07, "loss": 0.029, "step": 58660 }, { "grad_norm": 0.4385770559310913, "learning_rate": 1.3447804309719702e-07, "loss": 0.0258, "step": 58670 }, { "grad_norm": 0.2591712474822998, "learning_rate": 1.3246581652040512e-07, "loss": 0.0259, "step": 58680 }, { "grad_norm": 0.46393391489982605, "learning_rate": 1.3046873838381546e-07, "loss": 0.0294, "step": 58690 }, { "grad_norm": 0.3348217010498047, "learning_rate": 1.2848680929409828e-07, "loss": 0.027, "step": 58700 }, { "grad_norm": 0.3716055154800415, "learning_rate": 1.2652002985331091e-07, "loss": 0.0213, "step": 58710 }, { "grad_norm": 0.27868330478668213, "learning_rate": 1.2456840065889764e-07, "loss": 0.0265, "step": 58720 }, { "grad_norm": 0.16759537160396576, "learning_rate": 1.226319223037231e-07, "loss": 0.0203, "step": 58730 }, { "grad_norm": 0.4992171823978424, "learning_rate": 1.2071059537603902e-07, "loss": 0.0267, "step": 58740 }, { "grad_norm": 0.3736114501953125, "learning_rate": 1.1880442045948403e-07, "loss": 0.0179, "step": 58750 }, { "grad_norm": 0.519278347492218, "learning_rate": 1.1691339813311164e-07, "loss": 0.0283, "step": 58760 }, { "grad_norm": 0.3952038884162903, "learning_rate": 1.1503752897136233e-07, "loss": 0.0172, "step": 58770 }, { "grad_norm": 0.1269531399011612, "learning_rate": 1.1317681354407472e-07, "loss": 0.0177, "step": 58780 }, { "grad_norm": 0.35917630791664124, "learning_rate": 1.1133125241649111e-07, "loss": 0.0302, "step": 58790 }, { "grad_norm": 0.24702712893486023, "learning_rate": 1.0950084614922973e-07, "loss": 0.0214, "step": 58800 }, { "grad_norm": 0.5853767991065979, "learning_rate": 1.0768559529834021e-07, "loss": 0.0247, "step": 58810 }, { "grad_norm": 0.6142809987068176, "learning_rate": 1.0588550041522594e-07, "loss": 0.024, "step": 58820 }, { "grad_norm": 0.4549104571342468, "learning_rate": 1.0410056204672169e-07, "loss": 0.0254, "step": 58830 }, { "grad_norm": 0.27260151505470276, "learning_rate": 1.0233078073504376e-07, "loss": 0.0151, "step": 58840 }, { "grad_norm": 0.3364139795303345, "learning_rate": 1.0057615701780654e-07, "loss": 0.018, "step": 58850 }, { "grad_norm": 0.29252421855926514, "learning_rate": 9.883669142801144e-08, "loss": 0.0292, "step": 58860 }, { "grad_norm": 0.43150076270103455, "learning_rate": 9.711238449406356e-08, "loss": 0.0179, "step": 58870 }, { "grad_norm": 0.6032550930976868, "learning_rate": 9.540323673976614e-08, "loss": 0.03, "step": 58880 }, { "grad_norm": 0.291092187166214, "learning_rate": 9.370924868430942e-08, "loss": 0.0227, "step": 58890 }, { "grad_norm": 0.18667954206466675, "learning_rate": 9.203042084228175e-08, "loss": 0.0189, "step": 58900 }, { "grad_norm": 0.08880016207695007, "learning_rate": 9.036675372366965e-08, "loss": 0.0163, "step": 58910 }, { "grad_norm": 0.4689936339855194, "learning_rate": 8.871824783385218e-08, "loss": 0.0246, "step": 58920 }, { "grad_norm": 0.3507569432258606, "learning_rate": 8.7084903673601e-08, "loss": 0.0237, "step": 58930 }, { "grad_norm": 0.44138574600219727, "learning_rate": 8.546672173908032e-08, "loss": 0.0213, "step": 58940 }, { "grad_norm": 0.25038251280784607, "learning_rate": 8.386370252185249e-08, "loss": 0.0353, "step": 58950 }, { "grad_norm": 0.3275665044784546, "learning_rate": 8.227584650887243e-08, "loss": 0.0172, "step": 58960 }, { "grad_norm": 0.7144222259521484, "learning_rate": 8.070315418249319e-08, "loss": 0.0223, "step": 58970 }, { "grad_norm": 0.30246293544769287, "learning_rate": 7.914562602044929e-08, "loss": 0.0253, "step": 58980 }, { "grad_norm": 0.8604696989059448, "learning_rate": 7.76032624958789e-08, "loss": 0.0279, "step": 58990 }, { "grad_norm": 0.4715009033679962, "learning_rate": 7.607606407731282e-08, "loss": 0.0265, "step": 59000 }, { "grad_norm": 0.2516164481639862, "learning_rate": 7.45640312286744e-08, "loss": 0.0228, "step": 59010 }, { "grad_norm": 0.2565489709377289, "learning_rate": 7.306716440927952e-08, "loss": 0.0212, "step": 59020 }, { "grad_norm": 0.41689640283584595, "learning_rate": 7.15854640738367e-08, "loss": 0.0226, "step": 59030 }, { "grad_norm": 0.3080771565437317, "learning_rate": 7.011893067244701e-08, "loss": 0.0275, "step": 59040 }, { "grad_norm": 0.3204699754714966, "learning_rate": 6.866756465060408e-08, "loss": 0.022, "step": 59050 }, { "grad_norm": 0.2407829761505127, "learning_rate": 6.723136644918859e-08, "loss": 0.0218, "step": 59060 }, { "grad_norm": 0.4729510545730591, "learning_rate": 6.581033650449042e-08, "loss": 0.0202, "step": 59070 }, { "grad_norm": 0.2744436264038086, "learning_rate": 6.440447524817539e-08, "loss": 0.0427, "step": 59080 }, { "grad_norm": 0.43921345472335815, "learning_rate": 6.301378310730743e-08, "loss": 0.0182, "step": 59090 }, { "grad_norm": 0.36004194617271423, "learning_rate": 6.163826050434307e-08, "loss": 0.0244, "step": 59100 }, { "grad_norm": 0.34685105085372925, "learning_rate": 6.027790785713139e-08, "loss": 0.0269, "step": 59110 }, { "grad_norm": 0.5420806407928467, "learning_rate": 5.89327255789085e-08, "loss": 0.0238, "step": 59120 }, { "grad_norm": 0.23130932450294495, "learning_rate": 5.7602714078303085e-08, "loss": 0.0167, "step": 59130 }, { "grad_norm": 0.5884978771209717, "learning_rate": 5.628787375934197e-08, "loss": 0.0288, "step": 59140 }, { "grad_norm": 0.5535768866539001, "learning_rate": 5.498820502143898e-08, "loss": 0.0324, "step": 59150 }, { "grad_norm": 0.4761885106563568, "learning_rate": 5.370370825939497e-08, "loss": 0.026, "step": 59160 }, { "grad_norm": 0.5707082152366638, "learning_rate": 5.243438386340893e-08, "loss": 0.021, "step": 59170 }, { "grad_norm": 0.3473092317581177, "learning_rate": 5.118023221907242e-08, "loss": 0.0243, "step": 59180 }, { "grad_norm": 0.25615251064300537, "learning_rate": 4.994125370735292e-08, "loss": 0.0172, "step": 59190 }, { "grad_norm": 0.2671779692173004, "learning_rate": 4.871744870462713e-08, "loss": 0.0181, "step": 59200 }, { "grad_norm": 0.3555522561073303, "learning_rate": 4.7508817582658794e-08, "loss": 0.0195, "step": 59210 }, { "grad_norm": 0.20484638214111328, "learning_rate": 4.631536070858755e-08, "loss": 0.0158, "step": 59220 }, { "grad_norm": 0.19663231074810028, "learning_rate": 4.513707844495674e-08, "loss": 0.0274, "step": 59230 }, { "grad_norm": 0.3510832190513611, "learning_rate": 4.3973971149702255e-08, "loss": 0.0243, "step": 59240 }, { "grad_norm": 0.33772724866867065, "learning_rate": 4.2826039176147025e-08, "loss": 0.0347, "step": 59250 }, { "grad_norm": 0.2535938024520874, "learning_rate": 4.169328287299545e-08, "loss": 0.0162, "step": 59260 }, { "grad_norm": 0.552007794380188, "learning_rate": 4.057570258435006e-08, "loss": 0.0184, "step": 59270 }, { "grad_norm": 0.502922773361206, "learning_rate": 3.947329864970595e-08, "loss": 0.0258, "step": 59280 }, { "grad_norm": 0.30219873785972595, "learning_rate": 3.8386071403939686e-08, "loss": 0.0235, "step": 59290 }, { "grad_norm": 0.541185200214386, "learning_rate": 3.731402117733152e-08, "loss": 0.0213, "step": 59300 }, { "grad_norm": 0.5127459168434143, "learning_rate": 3.625714829552651e-08, "loss": 0.0205, "step": 59310 }, { "grad_norm": 0.49367716908454895, "learning_rate": 3.5215453079590065e-08, "loss": 0.0347, "step": 59320 }, { "grad_norm": 0.6430655121803284, "learning_rate": 3.4188935845952396e-08, "loss": 0.0311, "step": 59330 }, { "grad_norm": 0.326445996761322, "learning_rate": 3.3177596906447396e-08, "loss": 0.019, "step": 59340 }, { "grad_norm": 0.31558603048324585, "learning_rate": 3.218143656829043e-08, "loss": 0.0323, "step": 59350 }, { "grad_norm": 0.2903711199760437, "learning_rate": 3.120045513408387e-08, "loss": 0.019, "step": 59360 }, { "grad_norm": 0.38126176595687866, "learning_rate": 3.02346529018338e-08, "loss": 0.0229, "step": 59370 }, { "grad_norm": 0.1930529624223709, "learning_rate": 2.9284030164922204e-08, "loss": 0.0249, "step": 59380 }, { "grad_norm": 0.3200541138648987, "learning_rate": 2.8348587212123634e-08, "loss": 0.0165, "step": 59390 }, { "grad_norm": 0.6996000409126282, "learning_rate": 2.7428324327594125e-08, "loss": 0.0267, "step": 59400 }, { "grad_norm": 0.2653576731681824, "learning_rate": 2.6523241790893383e-08, "loss": 0.0202, "step": 59410 }, { "grad_norm": 0.2954314053058624, "learning_rate": 2.563333987695704e-08, "loss": 0.0261, "step": 59420 }, { "grad_norm": 0.329713374376297, "learning_rate": 2.4758618856118852e-08, "loss": 0.0235, "step": 59430 }, { "grad_norm": 0.49916765093803406, "learning_rate": 2.3899078994088497e-08, "loss": 0.0211, "step": 59440 }, { "grad_norm": 0.43243327736854553, "learning_rate": 2.3054720551973775e-08, "loss": 0.0228, "step": 59450 }, { "grad_norm": 3.8010976314544678, "learning_rate": 2.222554378627506e-08, "loss": 0.047, "step": 59460 }, { "grad_norm": 0.238407164812088, "learning_rate": 2.1411548948868653e-08, "loss": 0.0195, "step": 59470 }, { "grad_norm": 0.3894789516925812, "learning_rate": 2.0612736287023426e-08, "loss": 0.0225, "step": 59480 }, { "grad_norm": 0.32540011405944824, "learning_rate": 1.9829106043400826e-08, "loss": 0.0211, "step": 59490 }, { "grad_norm": 0.35909155011177063, "learning_rate": 1.9060658456043768e-08, "loss": 0.0243, "step": 59500 }, { "grad_norm": 0.20258358120918274, "learning_rate": 1.830739375838775e-08, "loss": 0.0167, "step": 59510 }, { "grad_norm": 0.2966870069503784, "learning_rate": 1.7569312179260832e-08, "loss": 0.0253, "step": 59520 }, { "grad_norm": 0.3472457528114319, "learning_rate": 1.684641394286146e-08, "loss": 0.0287, "step": 59530 }, { "grad_norm": 0.302550345659256, "learning_rate": 1.6138699268797296e-08, "loss": 0.0185, "step": 59540 }, { "grad_norm": 0.43368321657180786, "learning_rate": 1.5446168372046376e-08, "loss": 0.0297, "step": 59550 }, { "grad_norm": 0.7065489888191223, "learning_rate": 1.4768821462984861e-08, "loss": 0.027, "step": 59560 }, { "grad_norm": 0.6503373980522156, "learning_rate": 1.4106658747370383e-08, "loss": 0.026, "step": 59570 }, { "grad_norm": 0.6310417652130127, "learning_rate": 1.3459680426353149e-08, "loss": 0.0234, "step": 59580 }, { "grad_norm": 0.13608171045780182, "learning_rate": 1.2827886696464841e-08, "loss": 0.0205, "step": 59590 }, { "grad_norm": 0.5266280770301819, "learning_rate": 1.2211277749635264e-08, "loss": 0.0185, "step": 59600 }, { "grad_norm": 0.2531123757362366, "learning_rate": 1.1609853773164592e-08, "loss": 0.0163, "step": 59610 }, { "grad_norm": 0.37950456142425537, "learning_rate": 1.1023614949751127e-08, "loss": 0.0217, "step": 59620 }, { "grad_norm": 0.35216474533081055, "learning_rate": 1.0452561457485744e-08, "loss": 0.0247, "step": 59630 }, { "grad_norm": 0.3499719798564911, "learning_rate": 9.896693469829688e-09, "loss": 0.0204, "step": 59640 }, { "grad_norm": 0.4023189842700958, "learning_rate": 9.35601115564788e-09, "loss": 0.0323, "step": 59650 }, { "grad_norm": 0.7048588395118713, "learning_rate": 8.830514679186719e-09, "loss": 0.0232, "step": 59660 }, { "grad_norm": 0.7264280915260315, "learning_rate": 8.320204200074066e-09, "loss": 0.0369, "step": 59670 }, { "grad_norm": 0.20969878137111664, "learning_rate": 7.825079873324814e-09, "loss": 0.0237, "step": 59680 }, { "grad_norm": 0.3051105737686157, "learning_rate": 7.345141849351977e-09, "loss": 0.0313, "step": 59690 }, { "grad_norm": 0.3841957747936249, "learning_rate": 6.880390273944493e-09, "loss": 0.0262, "step": 59700 }, { "grad_norm": 0.3433675467967987, "learning_rate": 6.4308252882838704e-09, "loss": 0.0238, "step": 59710 }, { "grad_norm": 0.6329560279846191, "learning_rate": 5.9964470289386455e-09, "loss": 0.031, "step": 59720 }, { "grad_norm": 0.5124959945678711, "learning_rate": 5.577255627853273e-09, "loss": 0.0196, "step": 59730 }, { "grad_norm": 0.29777348041534424, "learning_rate": 5.173251212370334e-09, "loss": 0.0188, "step": 59740 }, { "grad_norm": 0.25789037346839905, "learning_rate": 4.784433905219432e-09, "loss": 0.0196, "step": 59750 }, { "grad_norm": 0.32734090089797974, "learning_rate": 4.4108038245060934e-09, "loss": 0.0272, "step": 59760 }, { "grad_norm": 0.8150956630706787, "learning_rate": 4.05236108373952e-09, "loss": 0.0212, "step": 59770 }, { "grad_norm": 0.324966162443161, "learning_rate": 3.7091057917937324e-09, "loss": 0.0186, "step": 59780 }, { "grad_norm": 0.3555181622505188, "learning_rate": 3.381038052946428e-09, "loss": 0.0458, "step": 59790 }, { "grad_norm": 0.43258893489837646, "learning_rate": 3.0681579668623283e-09, "loss": 0.0273, "step": 59800 }, { "grad_norm": 0.38406285643577576, "learning_rate": 2.7704656285709727e-09, "loss": 0.0187, "step": 59810 }, { "grad_norm": 0.47081851959228516, "learning_rate": 2.4879611285166803e-09, "loss": 0.0263, "step": 59820 }, { "grad_norm": 0.5169332027435303, "learning_rate": 2.2206445525085883e-09, "loss": 0.0229, "step": 59830 }, { "grad_norm": 0.5385672450065613, "learning_rate": 1.9685159817595112e-09, "loss": 0.0268, "step": 59840 }, { "grad_norm": 0.23740355670452118, "learning_rate": 1.7315754928470817e-09, "loss": 0.0201, "step": 59850 }, { "grad_norm": 0.36418765783309937, "learning_rate": 1.5098231577581611e-09, "loss": 0.0367, "step": 59860 }, { "grad_norm": 0.3220703601837158, "learning_rate": 1.3032590438499804e-09, "loss": 0.0195, "step": 59870 }, { "grad_norm": 0.2946290373802185, "learning_rate": 1.1118832138723444e-09, "loss": 0.0299, "step": 59880 }, { "grad_norm": 0.3982000946998596, "learning_rate": 9.356957259620825e-10, "loss": 0.0234, "step": 59890 }, { "grad_norm": 0.48682117462158203, "learning_rate": 7.74696633637495e-10, "loss": 0.0172, "step": 59900 }, { "grad_norm": 0.4116099178791046, "learning_rate": 6.288859858039064e-10, "loss": 0.0254, "step": 59910 }, { "grad_norm": 0.37491029500961304, "learning_rate": 4.982638267647665e-10, "loss": 0.0205, "step": 59920 }, { "grad_norm": 0.18031257390975952, "learning_rate": 3.8283019618834406e-10, "loss": 0.0218, "step": 59930 }, { "grad_norm": 0.27237093448638916, "learning_rate": 2.825851291410331e-10, "loss": 0.0209, "step": 59940 }, { "grad_norm": 0.5414235591888428, "learning_rate": 1.975286560873535e-10, "loss": 0.024, "step": 59950 }, { "grad_norm": 0.38309502601623535, "learning_rate": 1.2766080285109284e-10, "loss": 0.0238, "step": 59960 }, { "grad_norm": 0.3393029272556305, "learning_rate": 7.298159065971533e-11, "loss": 0.0271, "step": 59970 }, { "grad_norm": 0.25325140357017517, "learning_rate": 3.349103612770854e-11, "loss": 0.0162, "step": 59980 }, { "grad_norm": 0.4099539816379547, "learning_rate": 9.189151245481142e-12, "loss": 0.0188, "step": 59990 }, { "grad_norm": 0.6824719309806824, "learning_rate": 7.594340156735769e-14, "loss": 0.0268, "step": 60000 } ], "logging_steps": 10, "max_steps": 60000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "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": 4, "trial_name": null, "trial_params": null }