{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.4, "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": 0.7327033281326294, "learning_rate": 1.8e-07, "loss": 1.5122, "step": 10 }, { "grad_norm": 0.71014004945755, "learning_rate": 3.8e-07, "loss": 1.5131, "step": 20 }, { "grad_norm": 0.7012451887130737, "learning_rate": 5.8e-07, "loss": 1.4943, "step": 30 }, { "grad_norm": 0.7859618067741394, "learning_rate": 7.8e-07, "loss": 1.4971, "step": 40 }, { "grad_norm": 0.9699172377586365, "learning_rate": 9.8e-07, "loss": 1.4913, "step": 50 }, { "grad_norm": 0.59827721118927, "learning_rate": 1.18e-06, "loss": 1.4508, "step": 60 }, { "grad_norm": 0.8511502146720886, "learning_rate": 1.3800000000000001e-06, "loss": 1.3875, "step": 70 }, { "grad_norm": 0.9131712913513184, "learning_rate": 1.5800000000000003e-06, "loss": 1.3383, "step": 80 }, { "grad_norm": 0.7141631245613098, "learning_rate": 1.7800000000000001e-06, "loss": 1.2888, "step": 90 }, { "grad_norm": 0.6053668856620789, "learning_rate": 1.98e-06, "loss": 1.2639, "step": 100 }, { "grad_norm": 0.564238965511322, "learning_rate": 2.1800000000000003e-06, "loss": 1.2427, "step": 110 }, { "grad_norm": 0.8382623791694641, "learning_rate": 2.38e-06, "loss": 1.2344, "step": 120 }, { "grad_norm": 0.6189005374908447, "learning_rate": 2.5800000000000003e-06, "loss": 1.2003, "step": 130 }, { "grad_norm": 0.7980174422264099, "learning_rate": 2.78e-06, "loss": 1.1938, "step": 140 }, { "grad_norm": 0.5623437762260437, "learning_rate": 2.9800000000000003e-06, "loss": 1.168, "step": 150 }, { "grad_norm": 0.654207170009613, "learning_rate": 3.1800000000000005e-06, "loss": 1.1561, "step": 160 }, { "grad_norm": 0.7317996025085449, "learning_rate": 3.38e-06, "loss": 1.141, "step": 170 }, { "grad_norm": 0.9571701884269714, "learning_rate": 3.58e-06, "loss": 1.1274, "step": 180 }, { "grad_norm": 0.6726157665252686, "learning_rate": 3.7800000000000002e-06, "loss": 1.1181, "step": 190 }, { "grad_norm": 0.9971309304237366, "learning_rate": 3.98e-06, "loss": 1.1147, "step": 200 }, { "grad_norm": 0.8433402180671692, "learning_rate": 4.18e-06, "loss": 1.0961, "step": 210 }, { "grad_norm": 1.0107462406158447, "learning_rate": 4.38e-06, "loss": 1.0843, "step": 220 }, { "grad_norm": 2.2627816200256348, "learning_rate": 4.58e-06, "loss": 1.0819, "step": 230 }, { "grad_norm": 0.933554470539093, "learning_rate": 4.780000000000001e-06, "loss": 1.0849, "step": 240 }, { "grad_norm": 1.175357699394226, "learning_rate": 4.98e-06, "loss": 1.0758, "step": 250 }, { "grad_norm": 0.8214442133903503, "learning_rate": 5.18e-06, "loss": 1.078, "step": 260 }, { "grad_norm": 0.9311444759368896, "learning_rate": 5.38e-06, "loss": 1.0672, "step": 270 }, { "grad_norm": 1.200514554977417, "learning_rate": 5.580000000000001e-06, "loss": 1.0501, "step": 280 }, { "grad_norm": 1.8605272769927979, "learning_rate": 5.78e-06, "loss": 1.0294, "step": 290 }, { "grad_norm": 0.8741481304168701, "learning_rate": 5.98e-06, "loss": 1.0067, "step": 300 }, { "grad_norm": 1.0533535480499268, "learning_rate": 6.18e-06, "loss": 0.9814, "step": 310 }, { "grad_norm": 1.4344227313995361, "learning_rate": 6.38e-06, "loss": 0.9426, "step": 320 }, { "grad_norm": 1.0919215679168701, "learning_rate": 6.58e-06, "loss": 0.9059, "step": 330 }, { "grad_norm": 1.2169408798217773, "learning_rate": 6.78e-06, "loss": 0.8666, "step": 340 }, { "grad_norm": 1.249167799949646, "learning_rate": 6.98e-06, "loss": 0.8247, "step": 350 }, { "grad_norm": 1.6789575815200806, "learning_rate": 7.180000000000001e-06, "loss": 0.7924, "step": 360 }, { "grad_norm": 1.8729385137557983, "learning_rate": 7.3800000000000005e-06, "loss": 0.7425, "step": 370 }, { "grad_norm": 1.346977949142456, "learning_rate": 7.580000000000001e-06, "loss": 0.7023, "step": 380 }, { "grad_norm": 1.7686227560043335, "learning_rate": 7.78e-06, "loss": 0.6785, "step": 390 }, { "grad_norm": 1.7767062187194824, "learning_rate": 7.98e-06, "loss": 0.645, "step": 400 }, { "grad_norm": 1.7567898035049438, "learning_rate": 8.18e-06, "loss": 0.6136, "step": 410 }, { "grad_norm": 1.5689750909805298, "learning_rate": 8.380000000000001e-06, "loss": 0.5694, "step": 420 }, { "grad_norm": 1.6290255784988403, "learning_rate": 8.580000000000001e-06, "loss": 0.5521, "step": 430 }, { "grad_norm": 1.9847170114517212, "learning_rate": 8.78e-06, "loss": 0.5049, "step": 440 }, { "grad_norm": 1.8207364082336426, "learning_rate": 8.98e-06, "loss": 0.4738, "step": 450 }, { "grad_norm": 1.6240087747573853, "learning_rate": 9.180000000000002e-06, "loss": 0.4397, "step": 460 }, { "grad_norm": 2.0701427459716797, "learning_rate": 9.38e-06, "loss": 0.3952, "step": 470 }, { "grad_norm": 2.043442726135254, "learning_rate": 9.58e-06, "loss": 0.3722, "step": 480 }, { "grad_norm": 2.0447282791137695, "learning_rate": 9.78e-06, "loss": 0.3537, "step": 490 }, { "grad_norm": 2.405467987060547, "learning_rate": 9.980000000000001e-06, "loss": 0.339, "step": 500 }, { "grad_norm": 2.203561305999756, "learning_rate": 1.018e-05, "loss": 0.3118, "step": 510 }, { "grad_norm": 2.4483087062835693, "learning_rate": 1.038e-05, "loss": 0.3006, "step": 520 }, { "grad_norm": 2.0677897930145264, "learning_rate": 1.058e-05, "loss": 0.2795, "step": 530 }, { "grad_norm": 2.5633273124694824, "learning_rate": 1.0780000000000002e-05, "loss": 0.2595, "step": 540 }, { "grad_norm": 2.000044107437134, "learning_rate": 1.098e-05, "loss": 0.227, "step": 550 }, { "grad_norm": 1.801668643951416, "learning_rate": 1.118e-05, "loss": 0.2246, "step": 560 }, { "grad_norm": 1.9397791624069214, "learning_rate": 1.1380000000000001e-05, "loss": 0.2263, "step": 570 }, { "grad_norm": 1.786154866218567, "learning_rate": 1.1580000000000001e-05, "loss": 0.2083, "step": 580 }, { "grad_norm": 2.624748468399048, "learning_rate": 1.178e-05, "loss": 0.2052, "step": 590 }, { "grad_norm": 1.8504701852798462, "learning_rate": 1.198e-05, "loss": 0.2053, "step": 600 }, { "grad_norm": 1.9778707027435303, "learning_rate": 1.2180000000000002e-05, "loss": 0.1927, "step": 610 }, { "grad_norm": 2.013929605484009, "learning_rate": 1.238e-05, "loss": 0.1906, "step": 620 }, { "grad_norm": 1.9572361707687378, "learning_rate": 1.258e-05, "loss": 0.1843, "step": 630 }, { "grad_norm": 2.0028231143951416, "learning_rate": 1.278e-05, "loss": 0.19, "step": 640 }, { "grad_norm": 2.355475664138794, "learning_rate": 1.2980000000000001e-05, "loss": 0.1791, "step": 650 }, { "grad_norm": 1.999730110168457, "learning_rate": 1.3180000000000001e-05, "loss": 0.175, "step": 660 }, { "grad_norm": 1.5819891691207886, "learning_rate": 1.338e-05, "loss": 0.165, "step": 670 }, { "grad_norm": 2.7455027103424072, "learning_rate": 1.358e-05, "loss": 0.1648, "step": 680 }, { "grad_norm": 2.053696632385254, "learning_rate": 1.3780000000000002e-05, "loss": 0.1611, "step": 690 }, { "grad_norm": 1.683611273765564, "learning_rate": 1.3980000000000002e-05, "loss": 0.162, "step": 700 }, { "grad_norm": 2.5456717014312744, "learning_rate": 1.4180000000000001e-05, "loss": 0.1604, "step": 710 }, { "grad_norm": 2.085470199584961, "learning_rate": 1.4380000000000001e-05, "loss": 0.1608, "step": 720 }, { "grad_norm": 2.603590250015259, "learning_rate": 1.4580000000000003e-05, "loss": 0.1576, "step": 730 }, { "grad_norm": 1.8641881942749023, "learning_rate": 1.4779999999999999e-05, "loss": 0.1641, "step": 740 }, { "grad_norm": 1.6837650537490845, "learning_rate": 1.4979999999999999e-05, "loss": 0.1524, "step": 750 }, { "grad_norm": 2.0251355171203613, "learning_rate": 1.518e-05, "loss": 0.1553, "step": 760 }, { "grad_norm": 1.7289551496505737, "learning_rate": 1.538e-05, "loss": 0.1555, "step": 770 }, { "grad_norm": 2.1936397552490234, "learning_rate": 1.558e-05, "loss": 0.1489, "step": 780 }, { "grad_norm": 1.667404294013977, "learning_rate": 1.578e-05, "loss": 0.1389, "step": 790 }, { "grad_norm": 1.951653242111206, "learning_rate": 1.598e-05, "loss": 0.1437, "step": 800 }, { "grad_norm": 1.8602066040039062, "learning_rate": 1.618e-05, "loss": 0.1352, "step": 810 }, { "grad_norm": 2.480349063873291, "learning_rate": 1.6380000000000002e-05, "loss": 0.1386, "step": 820 }, { "grad_norm": 1.3204578161239624, "learning_rate": 1.658e-05, "loss": 0.1339, "step": 830 }, { "grad_norm": 1.469136118888855, "learning_rate": 1.6780000000000002e-05, "loss": 0.1403, "step": 840 }, { "grad_norm": 1.256475806236267, "learning_rate": 1.698e-05, "loss": 0.142, "step": 850 }, { "grad_norm": 1.6318267583847046, "learning_rate": 1.718e-05, "loss": 0.1385, "step": 860 }, { "grad_norm": 2.3485913276672363, "learning_rate": 1.7380000000000003e-05, "loss": 0.1309, "step": 870 }, { "grad_norm": 2.187861204147339, "learning_rate": 1.758e-05, "loss": 0.1344, "step": 880 }, { "grad_norm": 1.5754073858261108, "learning_rate": 1.7780000000000003e-05, "loss": 0.129, "step": 890 }, { "grad_norm": 1.869720697402954, "learning_rate": 1.798e-05, "loss": 0.1335, "step": 900 }, { "grad_norm": 1.5695561170578003, "learning_rate": 1.818e-05, "loss": 0.1188, "step": 910 }, { "grad_norm": 2.159921884536743, "learning_rate": 1.838e-05, "loss": 0.1231, "step": 920 }, { "grad_norm": 1.7697882652282715, "learning_rate": 1.858e-05, "loss": 0.1268, "step": 930 }, { "grad_norm": 1.7314186096191406, "learning_rate": 1.878e-05, "loss": 0.1217, "step": 940 }, { "grad_norm": 1.670783281326294, "learning_rate": 1.898e-05, "loss": 0.1222, "step": 950 }, { "grad_norm": 1.4610756635665894, "learning_rate": 1.918e-05, "loss": 0.1249, "step": 960 }, { "grad_norm": 2.0597431659698486, "learning_rate": 1.938e-05, "loss": 0.1223, "step": 970 }, { "grad_norm": 1.51002037525177, "learning_rate": 1.9580000000000002e-05, "loss": 0.1166, "step": 980 }, { "grad_norm": 1.5389305353164673, "learning_rate": 1.978e-05, "loss": 0.128, "step": 990 }, { "grad_norm": 1.5327554941177368, "learning_rate": 1.9980000000000002e-05, "loss": 0.1121, "step": 1000 }, { "grad_norm": 1.8238202333450317, "learning_rate": 2.0180000000000003e-05, "loss": 0.1184, "step": 1010 }, { "grad_norm": 1.734198808670044, "learning_rate": 2.038e-05, "loss": 0.123, "step": 1020 }, { "grad_norm": 1.1834206581115723, "learning_rate": 2.0580000000000003e-05, "loss": 0.1084, "step": 1030 }, { "grad_norm": 1.5783696174621582, "learning_rate": 2.078e-05, "loss": 0.1046, "step": 1040 }, { "grad_norm": 1.5859957933425903, "learning_rate": 2.098e-05, "loss": 0.1007, "step": 1050 }, { "grad_norm": 1.6463723182678223, "learning_rate": 2.118e-05, "loss": 0.1039, "step": 1060 }, { "grad_norm": 1.7808300256729126, "learning_rate": 2.138e-05, "loss": 0.1139, "step": 1070 }, { "grad_norm": 1.4802111387252808, "learning_rate": 2.158e-05, "loss": 0.1177, "step": 1080 }, { "grad_norm": 1.5730774402618408, "learning_rate": 2.178e-05, "loss": 0.1083, "step": 1090 }, { "grad_norm": 1.6908550262451172, "learning_rate": 2.198e-05, "loss": 0.1022, "step": 1100 }, { "grad_norm": 2.4301140308380127, "learning_rate": 2.218e-05, "loss": 0.1022, "step": 1110 }, { "grad_norm": 1.7893362045288086, "learning_rate": 2.2380000000000003e-05, "loss": 0.1026, "step": 1120 }, { "grad_norm": 1.404384732246399, "learning_rate": 2.258e-05, "loss": 0.1084, "step": 1130 }, { "grad_norm": 1.5309234857559204, "learning_rate": 2.2780000000000002e-05, "loss": 0.1078, "step": 1140 }, { "grad_norm": 1.7546483278274536, "learning_rate": 2.298e-05, "loss": 0.0998, "step": 1150 }, { "grad_norm": 1.3915176391601562, "learning_rate": 2.318e-05, "loss": 0.1023, "step": 1160 }, { "grad_norm": 1.3447935581207275, "learning_rate": 2.3380000000000003e-05, "loss": 0.1006, "step": 1170 }, { "grad_norm": 1.3458402156829834, "learning_rate": 2.358e-05, "loss": 0.102, "step": 1180 }, { "grad_norm": 1.3708909749984741, "learning_rate": 2.3780000000000003e-05, "loss": 0.0962, "step": 1190 }, { "grad_norm": 1.551927924156189, "learning_rate": 2.398e-05, "loss": 0.1046, "step": 1200 }, { "grad_norm": 1.9609512090682983, "learning_rate": 2.418e-05, "loss": 0.092, "step": 1210 }, { "grad_norm": 1.3751682043075562, "learning_rate": 2.438e-05, "loss": 0.0983, "step": 1220 }, { "grad_norm": 1.6424765586853027, "learning_rate": 2.4580000000000002e-05, "loss": 0.0976, "step": 1230 }, { "grad_norm": 1.2670347690582275, "learning_rate": 2.478e-05, "loss": 0.0939, "step": 1240 }, { "grad_norm": 1.4752342700958252, "learning_rate": 2.498e-05, "loss": 0.0916, "step": 1250 }, { "grad_norm": 1.3821603059768677, "learning_rate": 2.5180000000000003e-05, "loss": 0.0955, "step": 1260 }, { "grad_norm": 1.6177419424057007, "learning_rate": 2.5380000000000004e-05, "loss": 0.0939, "step": 1270 }, { "grad_norm": 1.4536430835723877, "learning_rate": 2.5580000000000002e-05, "loss": 0.1027, "step": 1280 }, { "grad_norm": 1.36681067943573, "learning_rate": 2.5779999999999997e-05, "loss": 0.0961, "step": 1290 }, { "grad_norm": 1.2298552989959717, "learning_rate": 2.598e-05, "loss": 0.096, "step": 1300 }, { "grad_norm": 1.657827377319336, "learning_rate": 2.618e-05, "loss": 0.0929, "step": 1310 }, { "grad_norm": 1.1323761940002441, "learning_rate": 2.6379999999999998e-05, "loss": 0.0844, "step": 1320 }, { "grad_norm": 1.2501518726348877, "learning_rate": 2.658e-05, "loss": 0.0961, "step": 1330 }, { "grad_norm": 1.3695417642593384, "learning_rate": 2.678e-05, "loss": 0.0942, "step": 1340 }, { "grad_norm": 1.3569042682647705, "learning_rate": 2.698e-05, "loss": 0.0913, "step": 1350 }, { "grad_norm": 1.5746957063674927, "learning_rate": 2.718e-05, "loss": 0.0936, "step": 1360 }, { "grad_norm": 1.2149338722229004, "learning_rate": 2.738e-05, "loss": 0.0929, "step": 1370 }, { "grad_norm": 1.2978042364120483, "learning_rate": 2.758e-05, "loss": 0.0828, "step": 1380 }, { "grad_norm": 1.3753395080566406, "learning_rate": 2.778e-05, "loss": 0.0873, "step": 1390 }, { "grad_norm": 1.1865711212158203, "learning_rate": 2.798e-05, "loss": 0.0849, "step": 1400 }, { "grad_norm": 1.2263109683990479, "learning_rate": 2.818e-05, "loss": 0.088, "step": 1410 }, { "grad_norm": 1.7374314069747925, "learning_rate": 2.8380000000000003e-05, "loss": 0.0888, "step": 1420 }, { "grad_norm": 1.4374351501464844, "learning_rate": 2.858e-05, "loss": 0.0893, "step": 1430 }, { "grad_norm": 1.2439711093902588, "learning_rate": 2.8780000000000002e-05, "loss": 0.0837, "step": 1440 }, { "grad_norm": 1.3081313371658325, "learning_rate": 2.898e-05, "loss": 0.0867, "step": 1450 }, { "grad_norm": 1.243116855621338, "learning_rate": 2.9180000000000002e-05, "loss": 0.094, "step": 1460 }, { "grad_norm": 1.4147266149520874, "learning_rate": 2.9380000000000003e-05, "loss": 0.0852, "step": 1470 }, { "grad_norm": 1.4877698421478271, "learning_rate": 2.958e-05, "loss": 0.0893, "step": 1480 }, { "grad_norm": 1.6115580797195435, "learning_rate": 2.9780000000000003e-05, "loss": 0.0882, "step": 1490 }, { "grad_norm": 1.5416656732559204, "learning_rate": 2.998e-05, "loss": 0.0824, "step": 1500 }, { "grad_norm": 1.3870471715927124, "learning_rate": 3.0180000000000002e-05, "loss": 0.0894, "step": 1510 }, { "grad_norm": 1.1771296262741089, "learning_rate": 3.0380000000000004e-05, "loss": 0.0812, "step": 1520 }, { "grad_norm": 1.358467698097229, "learning_rate": 3.058e-05, "loss": 0.0875, "step": 1530 }, { "grad_norm": 1.196520447731018, "learning_rate": 3.078e-05, "loss": 0.0874, "step": 1540 }, { "grad_norm": 1.3909105062484741, "learning_rate": 3.0980000000000005e-05, "loss": 0.0848, "step": 1550 }, { "grad_norm": 1.1223912239074707, "learning_rate": 3.118e-05, "loss": 0.0879, "step": 1560 }, { "grad_norm": 0.9514079093933105, "learning_rate": 3.138e-05, "loss": 0.0858, "step": 1570 }, { "grad_norm": 1.7387593984603882, "learning_rate": 3.1580000000000006e-05, "loss": 0.0827, "step": 1580 }, { "grad_norm": 1.0481780767440796, "learning_rate": 3.1780000000000004e-05, "loss": 0.0885, "step": 1590 }, { "grad_norm": 1.4705548286437988, "learning_rate": 3.198e-05, "loss": 0.0942, "step": 1600 }, { "grad_norm": 0.9064429998397827, "learning_rate": 3.218e-05, "loss": 0.0824, "step": 1610 }, { "grad_norm": 1.203469157218933, "learning_rate": 3.238e-05, "loss": 0.0778, "step": 1620 }, { "grad_norm": 1.3097162246704102, "learning_rate": 3.2579999999999996e-05, "loss": 0.0877, "step": 1630 }, { "grad_norm": 1.1492650508880615, "learning_rate": 3.278e-05, "loss": 0.0833, "step": 1640 }, { "grad_norm": 1.8563276529312134, "learning_rate": 3.298e-05, "loss": 0.0861, "step": 1650 }, { "grad_norm": 1.02285897731781, "learning_rate": 3.318e-05, "loss": 0.0825, "step": 1660 }, { "grad_norm": 1.0588065385818481, "learning_rate": 3.338e-05, "loss": 0.0829, "step": 1670 }, { "grad_norm": 1.5409966707229614, "learning_rate": 3.358e-05, "loss": 0.087, "step": 1680 }, { "grad_norm": 1.2057996988296509, "learning_rate": 3.378e-05, "loss": 0.0917, "step": 1690 }, { "grad_norm": 1.2207502126693726, "learning_rate": 3.398e-05, "loss": 0.0887, "step": 1700 }, { "grad_norm": 1.2312177419662476, "learning_rate": 3.418e-05, "loss": 0.079, "step": 1710 }, { "grad_norm": 1.5669586658477783, "learning_rate": 3.438e-05, "loss": 0.0816, "step": 1720 }, { "grad_norm": 1.1478687524795532, "learning_rate": 3.4580000000000004e-05, "loss": 0.0856, "step": 1730 }, { "grad_norm": 1.032617449760437, "learning_rate": 3.478e-05, "loss": 0.0788, "step": 1740 }, { "grad_norm": 1.0875649452209473, "learning_rate": 3.498e-05, "loss": 0.0839, "step": 1750 }, { "grad_norm": 1.0715899467468262, "learning_rate": 3.518e-05, "loss": 0.0777, "step": 1760 }, { "grad_norm": 1.3469873666763306, "learning_rate": 3.5380000000000003e-05, "loss": 0.0779, "step": 1770 }, { "grad_norm": 0.9829103946685791, "learning_rate": 3.558e-05, "loss": 0.079, "step": 1780 }, { "grad_norm": 1.078170895576477, "learning_rate": 3.578e-05, "loss": 0.0777, "step": 1790 }, { "grad_norm": 1.4245727062225342, "learning_rate": 3.5980000000000004e-05, "loss": 0.0787, "step": 1800 }, { "grad_norm": 0.9944266676902771, "learning_rate": 3.618e-05, "loss": 0.071, "step": 1810 }, { "grad_norm": 1.1887671947479248, "learning_rate": 3.638e-05, "loss": 0.0766, "step": 1820 }, { "grad_norm": 1.1121882200241089, "learning_rate": 3.6580000000000006e-05, "loss": 0.0724, "step": 1830 }, { "grad_norm": 1.3346726894378662, "learning_rate": 3.6780000000000004e-05, "loss": 0.0793, "step": 1840 }, { "grad_norm": 1.2907655239105225, "learning_rate": 3.698e-05, "loss": 0.0779, "step": 1850 }, { "grad_norm": 1.2246330976486206, "learning_rate": 3.7180000000000007e-05, "loss": 0.0822, "step": 1860 }, { "grad_norm": 0.8389066457748413, "learning_rate": 3.7380000000000005e-05, "loss": 0.0804, "step": 1870 }, { "grad_norm": 1.0993220806121826, "learning_rate": 3.758e-05, "loss": 0.0751, "step": 1880 }, { "grad_norm": 1.1497172117233276, "learning_rate": 3.778000000000001e-05, "loss": 0.0785, "step": 1890 }, { "grad_norm": 1.1071795225143433, "learning_rate": 3.7980000000000006e-05, "loss": 0.0735, "step": 1900 }, { "grad_norm": 0.751824140548706, "learning_rate": 3.818e-05, "loss": 0.0729, "step": 1910 }, { "grad_norm": 0.9575705528259277, "learning_rate": 3.838e-05, "loss": 0.076, "step": 1920 }, { "grad_norm": 1.254533290863037, "learning_rate": 3.858e-05, "loss": 0.0779, "step": 1930 }, { "grad_norm": 0.8262670040130615, "learning_rate": 3.878e-05, "loss": 0.0744, "step": 1940 }, { "grad_norm": 1.158899188041687, "learning_rate": 3.898e-05, "loss": 0.0736, "step": 1950 }, { "grad_norm": 1.0663623809814453, "learning_rate": 3.918e-05, "loss": 0.0736, "step": 1960 }, { "grad_norm": 1.11249577999115, "learning_rate": 3.938e-05, "loss": 0.0792, "step": 1970 }, { "grad_norm": 1.001099705696106, "learning_rate": 3.958e-05, "loss": 0.07, "step": 1980 }, { "grad_norm": 1.1428520679473877, "learning_rate": 3.978e-05, "loss": 0.0743, "step": 1990 }, { "grad_norm": 1.1269803047180176, "learning_rate": 3.998e-05, "loss": 0.0744, "step": 2000 }, { "grad_norm": 0.8521148562431335, "learning_rate": 4.018e-05, "loss": 0.0706, "step": 2010 }, { "grad_norm": 0.9583781361579895, "learning_rate": 4.038e-05, "loss": 0.078, "step": 2020 }, { "grad_norm": 0.9078081846237183, "learning_rate": 4.058e-05, "loss": 0.0785, "step": 2030 }, { "grad_norm": 0.7499627470970154, "learning_rate": 4.078e-05, "loss": 0.0725, "step": 2040 }, { "grad_norm": 0.8382400870323181, "learning_rate": 4.0980000000000004e-05, "loss": 0.0768, "step": 2050 }, { "grad_norm": 1.2415342330932617, "learning_rate": 4.118e-05, "loss": 0.0763, "step": 2060 }, { "grad_norm": 1.151893973350525, "learning_rate": 4.138e-05, "loss": 0.0749, "step": 2070 }, { "grad_norm": 1.236398458480835, "learning_rate": 4.1580000000000005e-05, "loss": 0.0725, "step": 2080 }, { "grad_norm": 1.2280226945877075, "learning_rate": 4.178e-05, "loss": 0.0772, "step": 2090 }, { "grad_norm": 1.0770632028579712, "learning_rate": 4.198e-05, "loss": 0.0726, "step": 2100 }, { "grad_norm": 1.086588740348816, "learning_rate": 4.2180000000000006e-05, "loss": 0.0666, "step": 2110 }, { "grad_norm": 1.0168242454528809, "learning_rate": 4.2380000000000004e-05, "loss": 0.0697, "step": 2120 }, { "grad_norm": 0.8630361557006836, "learning_rate": 4.258e-05, "loss": 0.0733, "step": 2130 }, { "grad_norm": 0.9190702438354492, "learning_rate": 4.278e-05, "loss": 0.07, "step": 2140 }, { "grad_norm": 0.8139598369598389, "learning_rate": 4.2980000000000005e-05, "loss": 0.0693, "step": 2150 }, { "grad_norm": 0.9934466481208801, "learning_rate": 4.318e-05, "loss": 0.0737, "step": 2160 }, { "grad_norm": 0.7925798892974854, "learning_rate": 4.338e-05, "loss": 0.0682, "step": 2170 }, { "grad_norm": 1.1828385591506958, "learning_rate": 4.3580000000000006e-05, "loss": 0.0713, "step": 2180 }, { "grad_norm": 1.158026933670044, "learning_rate": 4.3780000000000004e-05, "loss": 0.0703, "step": 2190 }, { "grad_norm": 1.0946063995361328, "learning_rate": 4.398e-05, "loss": 0.0799, "step": 2200 }, { "grad_norm": 0.887649416923523, "learning_rate": 4.418000000000001e-05, "loss": 0.074, "step": 2210 }, { "grad_norm": 1.1469734907150269, "learning_rate": 4.438e-05, "loss": 0.0773, "step": 2220 }, { "grad_norm": 0.9358341097831726, "learning_rate": 4.458e-05, "loss": 0.07, "step": 2230 }, { "grad_norm": 0.9340203404426575, "learning_rate": 4.478e-05, "loss": 0.0729, "step": 2240 }, { "grad_norm": 1.0851939916610718, "learning_rate": 4.498e-05, "loss": 0.0709, "step": 2250 }, { "grad_norm": 0.9772363305091858, "learning_rate": 4.518e-05, "loss": 0.075, "step": 2260 }, { "grad_norm": 0.8624616265296936, "learning_rate": 4.538e-05, "loss": 0.0749, "step": 2270 }, { "grad_norm": 1.1534589529037476, "learning_rate": 4.558e-05, "loss": 0.0655, "step": 2280 }, { "grad_norm": 1.1729040145874023, "learning_rate": 4.578e-05, "loss": 0.0698, "step": 2290 }, { "grad_norm": 0.7696219682693481, "learning_rate": 4.5980000000000004e-05, "loss": 0.07, "step": 2300 }, { "grad_norm": 0.8490058183670044, "learning_rate": 4.618e-05, "loss": 0.0717, "step": 2310 }, { "grad_norm": 0.910654604434967, "learning_rate": 4.638e-05, "loss": 0.0723, "step": 2320 }, { "grad_norm": 0.8725091218948364, "learning_rate": 4.6580000000000005e-05, "loss": 0.0754, "step": 2330 }, { "grad_norm": 1.0125480890274048, "learning_rate": 4.678e-05, "loss": 0.0669, "step": 2340 }, { "grad_norm": 0.9392992854118347, "learning_rate": 4.698e-05, "loss": 0.0749, "step": 2350 }, { "grad_norm": 0.8727006316184998, "learning_rate": 4.718e-05, "loss": 0.0719, "step": 2360 }, { "grad_norm": 0.9452412128448486, "learning_rate": 4.7380000000000004e-05, "loss": 0.0684, "step": 2370 }, { "grad_norm": 1.0700864791870117, "learning_rate": 4.758e-05, "loss": 0.07, "step": 2380 }, { "grad_norm": 0.9431473016738892, "learning_rate": 4.778e-05, "loss": 0.0711, "step": 2390 }, { "grad_norm": 0.8899092674255371, "learning_rate": 4.7980000000000005e-05, "loss": 0.0684, "step": 2400 }, { "grad_norm": 0.8410334587097168, "learning_rate": 4.818e-05, "loss": 0.0682, "step": 2410 }, { "grad_norm": 0.7519370317459106, "learning_rate": 4.838e-05, "loss": 0.0807, "step": 2420 }, { "grad_norm": 0.8178788423538208, "learning_rate": 4.8580000000000006e-05, "loss": 0.0744, "step": 2430 }, { "grad_norm": 0.9266697764396667, "learning_rate": 4.8780000000000004e-05, "loss": 0.0725, "step": 2440 }, { "grad_norm": 0.8747384548187256, "learning_rate": 4.898e-05, "loss": 0.0704, "step": 2450 }, { "grad_norm": 0.9026219844818115, "learning_rate": 4.918000000000001e-05, "loss": 0.0627, "step": 2460 }, { "grad_norm": 0.9394896030426025, "learning_rate": 4.9380000000000005e-05, "loss": 0.069, "step": 2470 }, { "grad_norm": 0.9191563725471497, "learning_rate": 4.958e-05, "loss": 0.0651, "step": 2480 }, { "grad_norm": 0.7097572088241577, "learning_rate": 4.978e-05, "loss": 0.0716, "step": 2490 }, { "grad_norm": 1.0233832597732544, "learning_rate": 4.9980000000000006e-05, "loss": 0.0679, "step": 2500 }, { "grad_norm": 1.104649543762207, "learning_rate": 5.0180000000000004e-05, "loss": 0.0681, "step": 2510 }, { "grad_norm": 0.7325562238693237, "learning_rate": 5.038e-05, "loss": 0.0635, "step": 2520 }, { "grad_norm": 0.778907299041748, "learning_rate": 5.058000000000001e-05, "loss": 0.065, "step": 2530 }, { "grad_norm": 0.7638477087020874, "learning_rate": 5.0780000000000005e-05, "loss": 0.0612, "step": 2540 }, { "grad_norm": 0.6208319664001465, "learning_rate": 5.098e-05, "loss": 0.0668, "step": 2550 }, { "grad_norm": 0.9259415864944458, "learning_rate": 5.118000000000001e-05, "loss": 0.0655, "step": 2560 }, { "grad_norm": 0.6604902148246765, "learning_rate": 5.1380000000000006e-05, "loss": 0.0594, "step": 2570 }, { "grad_norm": 1.0834776163101196, "learning_rate": 5.1580000000000004e-05, "loss": 0.0615, "step": 2580 }, { "grad_norm": 0.7458115816116333, "learning_rate": 5.178000000000001e-05, "loss": 0.0574, "step": 2590 }, { "grad_norm": 0.6866340637207031, "learning_rate": 5.198000000000001e-05, "loss": 0.0646, "step": 2600 }, { "grad_norm": 0.7553998827934265, "learning_rate": 5.2180000000000005e-05, "loss": 0.062, "step": 2610 }, { "grad_norm": 0.775993824005127, "learning_rate": 5.238000000000001e-05, "loss": 0.0614, "step": 2620 }, { "grad_norm": 0.6599122285842896, "learning_rate": 5.258000000000001e-05, "loss": 0.0594, "step": 2630 }, { "grad_norm": 1.060158371925354, "learning_rate": 5.2780000000000006e-05, "loss": 0.0634, "step": 2640 }, { "grad_norm": 0.6930858492851257, "learning_rate": 5.2980000000000004e-05, "loss": 0.0643, "step": 2650 }, { "grad_norm": 0.6336960792541504, "learning_rate": 5.318000000000001e-05, "loss": 0.0608, "step": 2660 }, { "grad_norm": 0.8317698836326599, "learning_rate": 5.338000000000001e-05, "loss": 0.0567, "step": 2670 }, { "grad_norm": 1.0220651626586914, "learning_rate": 5.3580000000000005e-05, "loss": 0.0655, "step": 2680 }, { "grad_norm": 0.7170953750610352, "learning_rate": 5.378e-05, "loss": 0.0656, "step": 2690 }, { "grad_norm": 0.7928021550178528, "learning_rate": 5.3979999999999995e-05, "loss": 0.0654, "step": 2700 }, { "grad_norm": 1.0239522457122803, "learning_rate": 5.418e-05, "loss": 0.0675, "step": 2710 }, { "grad_norm": 0.871756374835968, "learning_rate": 5.438e-05, "loss": 0.0683, "step": 2720 }, { "grad_norm": 0.6744424700737, "learning_rate": 5.4579999999999996e-05, "loss": 0.0751, "step": 2730 }, { "grad_norm": 0.8701303005218506, "learning_rate": 5.478e-05, "loss": 0.0695, "step": 2740 }, { "grad_norm": 0.7759307026863098, "learning_rate": 5.498e-05, "loss": 0.0641, "step": 2750 }, { "grad_norm": 0.8982049822807312, "learning_rate": 5.518e-05, "loss": 0.0646, "step": 2760 }, { "grad_norm": 0.974320113658905, "learning_rate": 5.538e-05, "loss": 0.0714, "step": 2770 }, { "grad_norm": 1.170229196548462, "learning_rate": 5.558e-05, "loss": 0.07, "step": 2780 }, { "grad_norm": 0.9780553579330444, "learning_rate": 5.578e-05, "loss": 0.0661, "step": 2790 }, { "grad_norm": 1.034454345703125, "learning_rate": 5.5979999999999996e-05, "loss": 0.0652, "step": 2800 }, { "grad_norm": 0.5880447626113892, "learning_rate": 5.618e-05, "loss": 0.0658, "step": 2810 }, { "grad_norm": 0.5424485206604004, "learning_rate": 5.638e-05, "loss": 0.0632, "step": 2820 }, { "grad_norm": 0.850857138633728, "learning_rate": 5.658e-05, "loss": 0.0588, "step": 2830 }, { "grad_norm": 0.6090867519378662, "learning_rate": 5.678e-05, "loss": 0.06, "step": 2840 }, { "grad_norm": 0.49484601616859436, "learning_rate": 5.698e-05, "loss": 0.0567, "step": 2850 }, { "grad_norm": 0.6222899556159973, "learning_rate": 5.718e-05, "loss": 0.0662, "step": 2860 }, { "grad_norm": 0.6327296495437622, "learning_rate": 5.738e-05, "loss": 0.062, "step": 2870 }, { "grad_norm": 0.5482462048530579, "learning_rate": 5.758e-05, "loss": 0.0568, "step": 2880 }, { "grad_norm": 0.8132043480873108, "learning_rate": 5.778e-05, "loss": 0.0612, "step": 2890 }, { "grad_norm": 0.6558915972709656, "learning_rate": 5.7980000000000004e-05, "loss": 0.057, "step": 2900 }, { "grad_norm": 0.7355269193649292, "learning_rate": 5.818e-05, "loss": 0.0583, "step": 2910 }, { "grad_norm": 0.7380754351615906, "learning_rate": 5.838e-05, "loss": 0.061, "step": 2920 }, { "grad_norm": 0.7111390829086304, "learning_rate": 5.858e-05, "loss": 0.064, "step": 2930 }, { "grad_norm": 0.8267529010772705, "learning_rate": 5.878e-05, "loss": 0.0667, "step": 2940 }, { "grad_norm": 0.8110604882240295, "learning_rate": 5.898e-05, "loss": 0.0636, "step": 2950 }, { "grad_norm": 0.7646610736846924, "learning_rate": 5.918e-05, "loss": 0.0618, "step": 2960 }, { "grad_norm": 0.7454716563224792, "learning_rate": 5.9380000000000004e-05, "loss": 0.0616, "step": 2970 }, { "grad_norm": 0.6766965389251709, "learning_rate": 5.958e-05, "loss": 0.0612, "step": 2980 }, { "grad_norm": 0.9046964645385742, "learning_rate": 5.978e-05, "loss": 0.0576, "step": 2990 }, { "grad_norm": 0.9157920479774475, "learning_rate": 5.9980000000000005e-05, "loss": 0.0574, "step": 3000 }, { "grad_norm": 0.6898856163024902, "learning_rate": 6.018e-05, "loss": 0.0577, "step": 3010 }, { "grad_norm": 0.7178052067756653, "learning_rate": 6.038e-05, "loss": 0.0641, "step": 3020 }, { "grad_norm": 0.7920030355453491, "learning_rate": 6.0580000000000006e-05, "loss": 0.0633, "step": 3030 }, { "grad_norm": 0.7980801463127136, "learning_rate": 6.0780000000000004e-05, "loss": 0.0611, "step": 3040 }, { "grad_norm": 0.8872457146644592, "learning_rate": 6.098e-05, "loss": 0.0568, "step": 3050 }, { "grad_norm": 0.7494392395019531, "learning_rate": 6.118000000000001e-05, "loss": 0.0574, "step": 3060 }, { "grad_norm": 0.6575875282287598, "learning_rate": 6.138e-05, "loss": 0.0548, "step": 3070 }, { "grad_norm": 0.8318756222724915, "learning_rate": 6.158e-05, "loss": 0.0577, "step": 3080 }, { "grad_norm": 0.7445926666259766, "learning_rate": 6.178000000000001e-05, "loss": 0.0636, "step": 3090 }, { "grad_norm": 0.6470189690589905, "learning_rate": 6.198e-05, "loss": 0.065, "step": 3100 }, { "grad_norm": 0.7413876056671143, "learning_rate": 6.218e-05, "loss": 0.0587, "step": 3110 }, { "grad_norm": 0.7423843741416931, "learning_rate": 6.238000000000001e-05, "loss": 0.0586, "step": 3120 }, { "grad_norm": 0.6846850514411926, "learning_rate": 6.258e-05, "loss": 0.061, "step": 3130 }, { "grad_norm": 0.47221237421035767, "learning_rate": 6.278e-05, "loss": 0.0565, "step": 3140 }, { "grad_norm": 0.6895976662635803, "learning_rate": 6.298000000000001e-05, "loss": 0.0579, "step": 3150 }, { "grad_norm": 0.596169650554657, "learning_rate": 6.318e-05, "loss": 0.0552, "step": 3160 }, { "grad_norm": 0.5421828031539917, "learning_rate": 6.338e-05, "loss": 0.0567, "step": 3170 }, { "grad_norm": 0.6699155569076538, "learning_rate": 6.358000000000001e-05, "loss": 0.0545, "step": 3180 }, { "grad_norm": 0.7528892159461975, "learning_rate": 6.378e-05, "loss": 0.0584, "step": 3190 }, { "grad_norm": 0.6975288391113281, "learning_rate": 6.398000000000001e-05, "loss": 0.0633, "step": 3200 }, { "grad_norm": 0.6970722079277039, "learning_rate": 6.418000000000001e-05, "loss": 0.0628, "step": 3210 }, { "grad_norm": 0.4895705580711365, "learning_rate": 6.438e-05, "loss": 0.0578, "step": 3220 }, { "grad_norm": 0.5439069867134094, "learning_rate": 6.458000000000001e-05, "loss": 0.0538, "step": 3230 }, { "grad_norm": 0.5369759798049927, "learning_rate": 6.478000000000001e-05, "loss": 0.0578, "step": 3240 }, { "grad_norm": 0.586226761341095, "learning_rate": 6.498e-05, "loss": 0.0543, "step": 3250 }, { "grad_norm": 0.9654123783111572, "learning_rate": 6.518000000000001e-05, "loss": 0.0633, "step": 3260 }, { "grad_norm": 1.1271116733551025, "learning_rate": 6.538000000000001e-05, "loss": 0.0749, "step": 3270 }, { "grad_norm": 0.6925221085548401, "learning_rate": 6.558e-05, "loss": 0.0711, "step": 3280 }, { "grad_norm": 0.6315380930900574, "learning_rate": 6.578000000000001e-05, "loss": 0.0748, "step": 3290 }, { "grad_norm": 0.6805611252784729, "learning_rate": 6.598e-05, "loss": 0.0629, "step": 3300 }, { "grad_norm": 0.4829985499382019, "learning_rate": 6.618e-05, "loss": 0.0626, "step": 3310 }, { "grad_norm": 0.5802217721939087, "learning_rate": 6.638e-05, "loss": 0.0619, "step": 3320 }, { "grad_norm": 0.6381668448448181, "learning_rate": 6.658e-05, "loss": 0.061, "step": 3330 }, { "grad_norm": 0.805056631565094, "learning_rate": 6.678e-05, "loss": 0.0647, "step": 3340 }, { "grad_norm": 0.5729668140411377, "learning_rate": 6.698e-05, "loss": 0.0611, "step": 3350 }, { "grad_norm": 0.6330121755599976, "learning_rate": 6.718e-05, "loss": 0.0564, "step": 3360 }, { "grad_norm": 0.9729316234588623, "learning_rate": 6.738e-05, "loss": 0.061, "step": 3370 }, { "grad_norm": 0.6738839745521545, "learning_rate": 6.758e-05, "loss": 0.0607, "step": 3380 }, { "grad_norm": 0.6228020191192627, "learning_rate": 6.778e-05, "loss": 0.0581, "step": 3390 }, { "grad_norm": 0.7606655359268188, "learning_rate": 6.798e-05, "loss": 0.066, "step": 3400 }, { "grad_norm": 0.8585534691810608, "learning_rate": 6.818e-05, "loss": 0.0572, "step": 3410 }, { "grad_norm": 0.796058177947998, "learning_rate": 6.838e-05, "loss": 0.0611, "step": 3420 }, { "grad_norm": 0.6467854380607605, "learning_rate": 6.858e-05, "loss": 0.0713, "step": 3430 }, { "grad_norm": 0.6691304445266724, "learning_rate": 6.878e-05, "loss": 0.0693, "step": 3440 }, { "grad_norm": 0.6632699966430664, "learning_rate": 6.898e-05, "loss": 0.0601, "step": 3450 }, { "grad_norm": 0.7652498483657837, "learning_rate": 6.918e-05, "loss": 0.0622, "step": 3460 }, { "grad_norm": 0.508461058139801, "learning_rate": 6.938e-05, "loss": 0.0549, "step": 3470 }, { "grad_norm": 0.5704008340835571, "learning_rate": 6.958e-05, "loss": 0.0611, "step": 3480 }, { "grad_norm": 0.513997495174408, "learning_rate": 6.978e-05, "loss": 0.0525, "step": 3490 }, { "grad_norm": 0.5967794060707092, "learning_rate": 6.998e-05, "loss": 0.0566, "step": 3500 }, { "grad_norm": 0.6005149483680725, "learning_rate": 7.018e-05, "loss": 0.0542, "step": 3510 }, { "grad_norm": 0.682636559009552, "learning_rate": 7.038e-05, "loss": 0.0568, "step": 3520 }, { "grad_norm": 0.6655474305152893, "learning_rate": 7.058e-05, "loss": 0.0594, "step": 3530 }, { "grad_norm": 0.6624171137809753, "learning_rate": 7.078e-05, "loss": 0.0597, "step": 3540 }, { "grad_norm": 0.6817507147789001, "learning_rate": 7.098e-05, "loss": 0.0587, "step": 3550 }, { "grad_norm": 0.4803317189216614, "learning_rate": 7.118e-05, "loss": 0.0563, "step": 3560 }, { "grad_norm": 0.5730146765708923, "learning_rate": 7.138e-05, "loss": 0.0534, "step": 3570 }, { "grad_norm": 0.6176438927650452, "learning_rate": 7.158e-05, "loss": 0.0556, "step": 3580 }, { "grad_norm": 0.7720077633857727, "learning_rate": 7.178000000000001e-05, "loss": 0.0564, "step": 3590 }, { "grad_norm": 0.552908182144165, "learning_rate": 7.198e-05, "loss": 0.0594, "step": 3600 }, { "grad_norm": 0.6944467425346375, "learning_rate": 7.218e-05, "loss": 0.0552, "step": 3610 }, { "grad_norm": 0.4685668349266052, "learning_rate": 7.238000000000001e-05, "loss": 0.0531, "step": 3620 }, { "grad_norm": 0.6087079048156738, "learning_rate": 7.258e-05, "loss": 0.0535, "step": 3630 }, { "grad_norm": 0.6034401655197144, "learning_rate": 7.278e-05, "loss": 0.0518, "step": 3640 }, { "grad_norm": 0.6862146258354187, "learning_rate": 7.298000000000001e-05, "loss": 0.056, "step": 3650 }, { "grad_norm": 0.6068137884140015, "learning_rate": 7.318e-05, "loss": 0.0613, "step": 3660 }, { "grad_norm": 0.6361678242683411, "learning_rate": 7.338e-05, "loss": 0.0592, "step": 3670 }, { "grad_norm": 0.5271837711334229, "learning_rate": 7.358000000000001e-05, "loss": 0.0574, "step": 3680 }, { "grad_norm": 0.6818010807037354, "learning_rate": 7.378e-05, "loss": 0.0611, "step": 3690 }, { "grad_norm": 0.7780253291130066, "learning_rate": 7.398e-05, "loss": 0.0635, "step": 3700 }, { "grad_norm": 0.5439037084579468, "learning_rate": 7.418000000000001e-05, "loss": 0.0567, "step": 3710 }, { "grad_norm": 0.5978808999061584, "learning_rate": 7.438e-05, "loss": 0.0564, "step": 3720 }, { "grad_norm": 0.5650739073753357, "learning_rate": 7.458000000000001e-05, "loss": 0.0539, "step": 3730 }, { "grad_norm": 0.38106438517570496, "learning_rate": 7.478e-05, "loss": 0.0551, "step": 3740 }, { "grad_norm": 0.5229088068008423, "learning_rate": 7.498e-05, "loss": 0.0542, "step": 3750 }, { "grad_norm": 0.5042112469673157, "learning_rate": 7.518000000000001e-05, "loss": 0.0533, "step": 3760 }, { "grad_norm": 0.6412204504013062, "learning_rate": 7.538e-05, "loss": 0.0544, "step": 3770 }, { "grad_norm": 0.5971466302871704, "learning_rate": 7.558e-05, "loss": 0.0499, "step": 3780 }, { "grad_norm": 0.6629607081413269, "learning_rate": 7.578000000000001e-05, "loss": 0.0516, "step": 3790 }, { "grad_norm": 0.6135489344596863, "learning_rate": 7.598e-05, "loss": 0.056, "step": 3800 }, { "grad_norm": 0.4827142655849457, "learning_rate": 7.618e-05, "loss": 0.0494, "step": 3810 }, { "grad_norm": 0.4219382107257843, "learning_rate": 7.638000000000001e-05, "loss": 0.0518, "step": 3820 }, { "grad_norm": 0.6230560541152954, "learning_rate": 7.658e-05, "loss": 0.0529, "step": 3830 }, { "grad_norm": 0.5097956657409668, "learning_rate": 7.678000000000001e-05, "loss": 0.0562, "step": 3840 }, { "grad_norm": 0.7422462701797485, "learning_rate": 7.698000000000001e-05, "loss": 0.0557, "step": 3850 }, { "grad_norm": 0.5605557560920715, "learning_rate": 7.718e-05, "loss": 0.0558, "step": 3860 }, { "grad_norm": 0.6768728494644165, "learning_rate": 7.738000000000001e-05, "loss": 0.0549, "step": 3870 }, { "grad_norm": 0.46962177753448486, "learning_rate": 7.758000000000001e-05, "loss": 0.0566, "step": 3880 }, { "grad_norm": 0.542169451713562, "learning_rate": 7.778e-05, "loss": 0.0525, "step": 3890 }, { "grad_norm": 0.43965619802474976, "learning_rate": 7.798000000000001e-05, "loss": 0.0524, "step": 3900 }, { "grad_norm": 0.5304702520370483, "learning_rate": 7.818000000000001e-05, "loss": 0.0579, "step": 3910 }, { "grad_norm": 0.5975410342216492, "learning_rate": 7.838e-05, "loss": 0.0594, "step": 3920 }, { "grad_norm": 0.5356630086898804, "learning_rate": 7.858000000000001e-05, "loss": 0.0494, "step": 3930 }, { "grad_norm": 0.6033384799957275, "learning_rate": 7.878e-05, "loss": 0.0538, "step": 3940 }, { "grad_norm": 0.553108274936676, "learning_rate": 7.897999999999999e-05, "loss": 0.0519, "step": 3950 }, { "grad_norm": 0.5994779467582703, "learning_rate": 7.918e-05, "loss": 0.054, "step": 3960 }, { "grad_norm": 0.5871763229370117, "learning_rate": 7.938e-05, "loss": 0.0509, "step": 3970 }, { "grad_norm": 0.5651211738586426, "learning_rate": 7.958e-05, "loss": 0.0596, "step": 3980 }, { "grad_norm": 0.7323636412620544, "learning_rate": 7.978e-05, "loss": 0.059, "step": 3990 }, { "grad_norm": 0.5582093596458435, "learning_rate": 7.998e-05, "loss": 0.0535, "step": 4000 }, { "grad_norm": 0.6246515512466431, "learning_rate": 8.018e-05, "loss": 0.0563, "step": 4010 }, { "grad_norm": 0.4885365664958954, "learning_rate": 8.038e-05, "loss": 0.0545, "step": 4020 }, { "grad_norm": 0.6407764554023743, "learning_rate": 8.058e-05, "loss": 0.0551, "step": 4030 }, { "grad_norm": 0.5367677807807922, "learning_rate": 8.078e-05, "loss": 0.0522, "step": 4040 }, { "grad_norm": 0.46138623356819153, "learning_rate": 8.098e-05, "loss": 0.0457, "step": 4050 }, { "grad_norm": 0.5401464104652405, "learning_rate": 8.118e-05, "loss": 0.053, "step": 4060 }, { "grad_norm": 0.8143817186355591, "learning_rate": 8.138e-05, "loss": 0.055, "step": 4070 }, { "grad_norm": 0.547332763671875, "learning_rate": 8.158e-05, "loss": 0.0552, "step": 4080 }, { "grad_norm": 0.39551353454589844, "learning_rate": 8.178e-05, "loss": 0.0522, "step": 4090 }, { "grad_norm": 0.48010990023612976, "learning_rate": 8.198e-05, "loss": 0.0537, "step": 4100 }, { "grad_norm": 0.6067258715629578, "learning_rate": 8.218e-05, "loss": 0.0533, "step": 4110 }, { "grad_norm": 0.5142095685005188, "learning_rate": 8.238000000000001e-05, "loss": 0.0504, "step": 4120 }, { "grad_norm": 0.49233323335647583, "learning_rate": 8.258e-05, "loss": 0.05, "step": 4130 }, { "grad_norm": 0.5436373949050903, "learning_rate": 8.278e-05, "loss": 0.0553, "step": 4140 }, { "grad_norm": 0.5813378095626831, "learning_rate": 8.298000000000001e-05, "loss": 0.0531, "step": 4150 }, { "grad_norm": 0.5976006388664246, "learning_rate": 8.318e-05, "loss": 0.0553, "step": 4160 }, { "grad_norm": 0.4747331440448761, "learning_rate": 8.338e-05, "loss": 0.0509, "step": 4170 }, { "grad_norm": 0.4282720386981964, "learning_rate": 8.358e-05, "loss": 0.0516, "step": 4180 }, { "grad_norm": 0.5323525667190552, "learning_rate": 8.378e-05, "loss": 0.0498, "step": 4190 }, { "grad_norm": 0.5399609804153442, "learning_rate": 8.398e-05, "loss": 0.0523, "step": 4200 }, { "grad_norm": 0.5413873791694641, "learning_rate": 8.418e-05, "loss": 0.0521, "step": 4210 }, { "grad_norm": 0.5721999406814575, "learning_rate": 8.438e-05, "loss": 0.0505, "step": 4220 }, { "grad_norm": 0.32109543681144714, "learning_rate": 8.458e-05, "loss": 0.0461, "step": 4230 }, { "grad_norm": 0.5597385168075562, "learning_rate": 8.478e-05, "loss": 0.0513, "step": 4240 }, { "grad_norm": 0.5460967421531677, "learning_rate": 8.498e-05, "loss": 0.0521, "step": 4250 }, { "grad_norm": 0.5875177383422852, "learning_rate": 8.518000000000001e-05, "loss": 0.0529, "step": 4260 }, { "grad_norm": 0.492908775806427, "learning_rate": 8.538e-05, "loss": 0.0465, "step": 4270 }, { "grad_norm": 0.4367402493953705, "learning_rate": 8.558e-05, "loss": 0.0535, "step": 4280 }, { "grad_norm": 0.48697489500045776, "learning_rate": 8.578000000000001e-05, "loss": 0.0582, "step": 4290 }, { "grad_norm": 0.4602562487125397, "learning_rate": 8.598e-05, "loss": 0.0597, "step": 4300 }, { "grad_norm": 0.5547934174537659, "learning_rate": 8.618e-05, "loss": 0.0604, "step": 4310 }, { "grad_norm": 0.4442889988422394, "learning_rate": 8.638000000000001e-05, "loss": 0.0556, "step": 4320 }, { "grad_norm": 0.43597614765167236, "learning_rate": 8.658e-05, "loss": 0.055, "step": 4330 }, { "grad_norm": 0.5243062376976013, "learning_rate": 8.678e-05, "loss": 0.0532, "step": 4340 }, { "grad_norm": 0.4550265371799469, "learning_rate": 8.698000000000001e-05, "loss": 0.0493, "step": 4350 }, { "grad_norm": 0.4338337779045105, "learning_rate": 8.718e-05, "loss": 0.0493, "step": 4360 }, { "grad_norm": 0.44544923305511475, "learning_rate": 8.738000000000001e-05, "loss": 0.052, "step": 4370 }, { "grad_norm": 0.48967108130455017, "learning_rate": 8.758000000000001e-05, "loss": 0.0518, "step": 4380 }, { "grad_norm": 0.34587642550468445, "learning_rate": 8.778e-05, "loss": 0.0496, "step": 4390 }, { "grad_norm": 0.4918307960033417, "learning_rate": 8.798000000000001e-05, "loss": 0.0537, "step": 4400 }, { "grad_norm": 0.4976046681404114, "learning_rate": 8.818000000000001e-05, "loss": 0.0515, "step": 4410 }, { "grad_norm": 0.3702969551086426, "learning_rate": 8.838e-05, "loss": 0.0474, "step": 4420 }, { "grad_norm": 0.393335223197937, "learning_rate": 8.858000000000001e-05, "loss": 0.0477, "step": 4430 }, { "grad_norm": 0.40824922919273376, "learning_rate": 8.878000000000001e-05, "loss": 0.0462, "step": 4440 }, { "grad_norm": 0.4517180919647217, "learning_rate": 8.898e-05, "loss": 0.0482, "step": 4450 }, { "grad_norm": 0.6099254488945007, "learning_rate": 8.918000000000001e-05, "loss": 0.0451, "step": 4460 }, { "grad_norm": 0.6540037393569946, "learning_rate": 8.938e-05, "loss": 0.049, "step": 4470 }, { "grad_norm": 0.4156933128833771, "learning_rate": 8.958e-05, "loss": 0.0506, "step": 4480 }, { "grad_norm": 0.4851490557193756, "learning_rate": 8.978000000000001e-05, "loss": 0.0484, "step": 4490 }, { "grad_norm": 0.416775643825531, "learning_rate": 8.998e-05, "loss": 0.049, "step": 4500 }, { "grad_norm": 0.5364426970481873, "learning_rate": 9.018000000000001e-05, "loss": 0.0486, "step": 4510 }, { "grad_norm": 0.45595499873161316, "learning_rate": 9.038000000000001e-05, "loss": 0.0484, "step": 4520 }, { "grad_norm": 0.47213858366012573, "learning_rate": 9.058e-05, "loss": 0.0492, "step": 4530 }, { "grad_norm": 0.4608618915081024, "learning_rate": 9.078000000000001e-05, "loss": 0.0479, "step": 4540 }, { "grad_norm": 0.5390095114707947, "learning_rate": 9.098000000000001e-05, "loss": 0.0494, "step": 4550 }, { "grad_norm": 0.4771552085876465, "learning_rate": 9.118e-05, "loss": 0.0517, "step": 4560 }, { "grad_norm": 0.47849521040916443, "learning_rate": 9.138e-05, "loss": 0.0492, "step": 4570 }, { "grad_norm": 0.45128053426742554, "learning_rate": 9.158e-05, "loss": 0.054, "step": 4580 }, { "grad_norm": 0.5492345094680786, "learning_rate": 9.178e-05, "loss": 0.0522, "step": 4590 }, { "grad_norm": 0.4941067099571228, "learning_rate": 9.198e-05, "loss": 0.0482, "step": 4600 }, { "grad_norm": 0.5618913173675537, "learning_rate": 9.218e-05, "loss": 0.0504, "step": 4610 }, { "grad_norm": 0.5912190079689026, "learning_rate": 9.238e-05, "loss": 0.0519, "step": 4620 }, { "grad_norm": 0.4416705071926117, "learning_rate": 9.258e-05, "loss": 0.0502, "step": 4630 }, { "grad_norm": 0.4184553623199463, "learning_rate": 9.278e-05, "loss": 0.0462, "step": 4640 }, { "grad_norm": 0.5338907837867737, "learning_rate": 9.298e-05, "loss": 0.0476, "step": 4650 }, { "grad_norm": 0.5457320809364319, "learning_rate": 9.318e-05, "loss": 0.054, "step": 4660 }, { "grad_norm": 0.40947145223617554, "learning_rate": 9.338e-05, "loss": 0.0534, "step": 4670 }, { "grad_norm": 0.6344555020332336, "learning_rate": 9.358e-05, "loss": 0.0481, "step": 4680 }, { "grad_norm": 0.6770515441894531, "learning_rate": 9.378e-05, "loss": 0.0705, "step": 4690 }, { "grad_norm": 0.3724403381347656, "learning_rate": 9.398e-05, "loss": 0.058, "step": 4700 }, { "grad_norm": 0.3366471230983734, "learning_rate": 9.418e-05, "loss": 0.0527, "step": 4710 }, { "grad_norm": 0.4419296085834503, "learning_rate": 9.438e-05, "loss": 0.056, "step": 4720 }, { "grad_norm": 0.5128366947174072, "learning_rate": 9.458e-05, "loss": 0.055, "step": 4730 }, { "grad_norm": 0.4052845537662506, "learning_rate": 9.478e-05, "loss": 0.0489, "step": 4740 }, { "grad_norm": 0.49691009521484375, "learning_rate": 9.498e-05, "loss": 0.052, "step": 4750 }, { "grad_norm": 0.5064414739608765, "learning_rate": 9.518000000000001e-05, "loss": 0.0505, "step": 4760 }, { "grad_norm": 0.4273929297924042, "learning_rate": 9.538e-05, "loss": 0.0531, "step": 4770 }, { "grad_norm": 0.37186360359191895, "learning_rate": 9.558e-05, "loss": 0.0508, "step": 4780 }, { "grad_norm": 0.46876659989356995, "learning_rate": 9.578000000000001e-05, "loss": 0.0535, "step": 4790 }, { "grad_norm": 0.40781718492507935, "learning_rate": 9.598e-05, "loss": 0.0469, "step": 4800 }, { "grad_norm": 0.3954528868198395, "learning_rate": 9.618e-05, "loss": 0.0533, "step": 4810 }, { "grad_norm": 0.5094861388206482, "learning_rate": 9.638000000000001e-05, "loss": 0.0539, "step": 4820 }, { "grad_norm": 0.4822026491165161, "learning_rate": 9.658e-05, "loss": 0.0537, "step": 4830 }, { "grad_norm": 0.5171741843223572, "learning_rate": 9.678e-05, "loss": 0.0536, "step": 4840 }, { "grad_norm": 0.5116572976112366, "learning_rate": 9.698000000000001e-05, "loss": 0.0519, "step": 4850 }, { "grad_norm": 0.5983439087867737, "learning_rate": 9.718e-05, "loss": 0.0614, "step": 4860 }, { "grad_norm": 0.5761286616325378, "learning_rate": 9.738e-05, "loss": 0.0558, "step": 4870 }, { "grad_norm": 0.6650732755661011, "learning_rate": 9.758000000000001e-05, "loss": 0.0531, "step": 4880 }, { "grad_norm": 0.47995421290397644, "learning_rate": 9.778e-05, "loss": 0.0497, "step": 4890 }, { "grad_norm": 0.7858373522758484, "learning_rate": 9.798000000000001e-05, "loss": 0.0517, "step": 4900 }, { "grad_norm": 0.4860854148864746, "learning_rate": 9.818000000000001e-05, "loss": 0.0549, "step": 4910 }, { "grad_norm": 0.461408793926239, "learning_rate": 9.838e-05, "loss": 0.0563, "step": 4920 }, { "grad_norm": 0.6118635535240173, "learning_rate": 9.858000000000001e-05, "loss": 0.0551, "step": 4930 }, { "grad_norm": 0.47015970945358276, "learning_rate": 9.878e-05, "loss": 0.052, "step": 4940 }, { "grad_norm": 0.5437532663345337, "learning_rate": 9.898e-05, "loss": 0.0568, "step": 4950 }, { "grad_norm": 0.48606041073799133, "learning_rate": 9.918000000000001e-05, "loss": 0.054, "step": 4960 }, { "grad_norm": 0.3774937689304352, "learning_rate": 9.938e-05, "loss": 0.0512, "step": 4970 }, { "grad_norm": 0.3810940682888031, "learning_rate": 9.958e-05, "loss": 0.0501, "step": 4980 }, { "grad_norm": 0.39563870429992676, "learning_rate": 9.978000000000001e-05, "loss": 0.0493, "step": 4990 }, { "grad_norm": 0.55155348777771, "learning_rate": 9.998e-05, "loss": 0.0557, "step": 5000 }, { "grad_norm": 0.4652394950389862, "learning_rate": 9.999999778549045e-05, "loss": 0.0523, "step": 5010 }, { "grad_norm": 0.48009902238845825, "learning_rate": 9.999999013039593e-05, "loss": 0.0538, "step": 5020 }, { "grad_norm": 0.45218977332115173, "learning_rate": 9.999997700737766e-05, "loss": 0.0456, "step": 5030 }, { "grad_norm": 0.4762789309024811, "learning_rate": 9.999995841643709e-05, "loss": 0.0465, "step": 5040 }, { "grad_norm": 0.5468534231185913, "learning_rate": 9.999993435757623e-05, "loss": 0.0509, "step": 5050 }, { "grad_norm": 0.554854154586792, "learning_rate": 9.999990483079773e-05, "loss": 0.0537, "step": 5060 }, { "grad_norm": 0.4641476571559906, "learning_rate": 9.999986983610481e-05, "loss": 0.0493, "step": 5070 }, { "grad_norm": 0.43800032138824463, "learning_rate": 9.99998293735013e-05, "loss": 0.0477, "step": 5080 }, { "grad_norm": 0.45098358392715454, "learning_rate": 9.999978344299161e-05, "loss": 0.0488, "step": 5090 }, { "grad_norm": 0.48698267340660095, "learning_rate": 9.99997320445808e-05, "loss": 0.0489, "step": 5100 }, { "grad_norm": 0.5717257261276245, "learning_rate": 9.999967517827444e-05, "loss": 0.0499, "step": 5110 }, { "grad_norm": 0.33073800802230835, "learning_rate": 9.999961284407879e-05, "loss": 0.0471, "step": 5120 }, { "grad_norm": 0.49221619963645935, "learning_rate": 9.999954504200067e-05, "loss": 0.0471, "step": 5130 }, { "grad_norm": 0.46254655718803406, "learning_rate": 9.999947177204744e-05, "loss": 0.0508, "step": 5140 }, { "grad_norm": 0.5199896693229675, "learning_rate": 9.999939303422718e-05, "loss": 0.0468, "step": 5150 }, { "grad_norm": 0.44831186532974243, "learning_rate": 9.999930882854847e-05, "loss": 0.0495, "step": 5160 }, { "grad_norm": 0.4628754258155823, "learning_rate": 9.999921915502051e-05, "loss": 0.0437, "step": 5170 }, { "grad_norm": 0.37178364396095276, "learning_rate": 9.99991240136531e-05, "loss": 0.0395, "step": 5180 }, { "grad_norm": 0.4137933552265167, "learning_rate": 9.999902340445668e-05, "loss": 0.0471, "step": 5190 }, { "grad_norm": 0.3976815342903137, "learning_rate": 9.999891732744224e-05, "loss": 0.0422, "step": 5200 }, { "grad_norm": 0.44029730558395386, "learning_rate": 9.999880578262135e-05, "loss": 0.0447, "step": 5210 }, { "grad_norm": 0.5281795859336853, "learning_rate": 9.999868877000624e-05, "loss": 0.0417, "step": 5220 }, { "grad_norm": 0.42131924629211426, "learning_rate": 9.99985662896097e-05, "loss": 0.0445, "step": 5230 }, { "grad_norm": 0.4225093424320221, "learning_rate": 9.999843834144513e-05, "loss": 0.0461, "step": 5240 }, { "grad_norm": 0.36050233244895935, "learning_rate": 9.99983049255265e-05, "loss": 0.0447, "step": 5250 }, { "grad_norm": 0.4533052146434784, "learning_rate": 9.999816604186843e-05, "loss": 0.0485, "step": 5260 }, { "grad_norm": 0.3251153230667114, "learning_rate": 9.999802169048609e-05, "loss": 0.048, "step": 5270 }, { "grad_norm": 0.34078481793403625, "learning_rate": 9.999787187139527e-05, "loss": 0.0457, "step": 5280 }, { "grad_norm": 0.35132360458374023, "learning_rate": 9.999771658461234e-05, "loss": 0.0454, "step": 5290 }, { "grad_norm": 0.3401744067668915, "learning_rate": 9.999755583015431e-05, "loss": 0.0483, "step": 5300 }, { "grad_norm": 0.5982875227928162, "learning_rate": 9.999738960803874e-05, "loss": 0.0434, "step": 5310 }, { "grad_norm": 0.511320948600769, "learning_rate": 9.99972179182838e-05, "loss": 0.0494, "step": 5320 }, { "grad_norm": 0.49892789125442505, "learning_rate": 9.99970407609083e-05, "loss": 0.0482, "step": 5330 }, { "grad_norm": 0.36111342906951904, "learning_rate": 9.999685813593159e-05, "loss": 0.0463, "step": 5340 }, { "grad_norm": 0.4559060335159302, "learning_rate": 9.999667004337362e-05, "loss": 0.0447, "step": 5350 }, { "grad_norm": 0.4354434907436371, "learning_rate": 9.9996476483255e-05, "loss": 0.0434, "step": 5360 }, { "grad_norm": 0.36624380946159363, "learning_rate": 9.999627745559688e-05, "loss": 0.0445, "step": 5370 }, { "grad_norm": 0.4005759656429291, "learning_rate": 9.999607296042101e-05, "loss": 0.0462, "step": 5380 }, { "grad_norm": 0.4552546441555023, "learning_rate": 9.99958629977498e-05, "loss": 0.0447, "step": 5390 }, { "grad_norm": 0.6227610111236572, "learning_rate": 9.999564756760615e-05, "loss": 0.0529, "step": 5400 }, { "grad_norm": 0.3645300269126892, "learning_rate": 9.999542667001366e-05, "loss": 0.0524, "step": 5410 }, { "grad_norm": 0.3981894254684448, "learning_rate": 9.999520030499647e-05, "loss": 0.0495, "step": 5420 }, { "grad_norm": 0.5785505771636963, "learning_rate": 9.999496847257936e-05, "loss": 0.0494, "step": 5430 }, { "grad_norm": 0.5509626269340515, "learning_rate": 9.999473117278764e-05, "loss": 0.0471, "step": 5440 }, { "grad_norm": 0.5468688011169434, "learning_rate": 9.999448840564731e-05, "loss": 0.0467, "step": 5450 }, { "grad_norm": 0.441291481256485, "learning_rate": 9.999424017118488e-05, "loss": 0.0473, "step": 5460 }, { "grad_norm": 0.41137465834617615, "learning_rate": 9.999398646942751e-05, "loss": 0.0497, "step": 5470 }, { "grad_norm": 0.39310184121131897, "learning_rate": 9.999372730040296e-05, "loss": 0.0491, "step": 5480 }, { "grad_norm": 0.45481550693511963, "learning_rate": 9.999346266413953e-05, "loss": 0.0499, "step": 5490 }, { "grad_norm": 0.5378975868225098, "learning_rate": 9.99931925606662e-05, "loss": 0.0571, "step": 5500 }, { "grad_norm": 0.6047338247299194, "learning_rate": 9.99929169900125e-05, "loss": 0.0558, "step": 5510 }, { "grad_norm": 0.5091220736503601, "learning_rate": 9.999263595220855e-05, "loss": 0.055, "step": 5520 }, { "grad_norm": 0.4720809757709503, "learning_rate": 9.99923494472851e-05, "loss": 0.0545, "step": 5530 }, { "grad_norm": 0.4258424937725067, "learning_rate": 9.999205747527348e-05, "loss": 0.0531, "step": 5540 }, { "grad_norm": 0.3555290400981903, "learning_rate": 9.999176003620561e-05, "loss": 0.0509, "step": 5550 }, { "grad_norm": 0.46665966510772705, "learning_rate": 9.999145713011405e-05, "loss": 0.051, "step": 5560 }, { "grad_norm": 0.3965032994747162, "learning_rate": 9.999114875703186e-05, "loss": 0.0463, "step": 5570 }, { "grad_norm": 0.5833112597465515, "learning_rate": 9.999083491699281e-05, "loss": 0.0583, "step": 5580 }, { "grad_norm": 0.569023609161377, "learning_rate": 9.999051561003123e-05, "loss": 0.0636, "step": 5590 }, { "grad_norm": 0.46931910514831543, "learning_rate": 9.999019083618202e-05, "loss": 0.0685, "step": 5600 }, { "grad_norm": 0.5379312038421631, "learning_rate": 9.99898605954807e-05, "loss": 0.0568, "step": 5610 }, { "grad_norm": 0.43313419818878174, "learning_rate": 9.998952488796338e-05, "loss": 0.0552, "step": 5620 }, { "grad_norm": 0.49676594138145447, "learning_rate": 9.998918371366676e-05, "loss": 0.0575, "step": 5630 }, { "grad_norm": 0.5204843282699585, "learning_rate": 9.99888370726282e-05, "loss": 0.0479, "step": 5640 }, { "grad_norm": 0.471370130777359, "learning_rate": 9.998848496488556e-05, "loss": 0.0508, "step": 5650 }, { "grad_norm": 0.45952460169792175, "learning_rate": 9.998812739047736e-05, "loss": 0.0458, "step": 5660 }, { "grad_norm": 0.4239512085914612, "learning_rate": 9.99877643494427e-05, "loss": 0.0479, "step": 5670 }, { "grad_norm": 0.4575538635253906, "learning_rate": 9.998739584182128e-05, "loss": 0.0456, "step": 5680 }, { "grad_norm": 0.33665406703948975, "learning_rate": 9.998702186765342e-05, "loss": 0.0472, "step": 5690 }, { "grad_norm": 0.4410259425640106, "learning_rate": 9.998664242698e-05, "loss": 0.0429, "step": 5700 }, { "grad_norm": 0.5990079641342163, "learning_rate": 9.998625751984251e-05, "loss": 0.0515, "step": 5710 }, { "grad_norm": 0.34993186593055725, "learning_rate": 9.998586714628307e-05, "loss": 0.0451, "step": 5720 }, { "grad_norm": 0.8419672846794128, "learning_rate": 9.998547130634432e-05, "loss": 0.0536, "step": 5730 }, { "grad_norm": 0.3735809624195099, "learning_rate": 9.99850700000696e-05, "loss": 0.054, "step": 5740 }, { "grad_norm": 0.529926598072052, "learning_rate": 9.998466322750278e-05, "loss": 0.0492, "step": 5750 }, { "grad_norm": 0.42414039373397827, "learning_rate": 9.998425098868834e-05, "loss": 0.051, "step": 5760 }, { "grad_norm": 0.45567867159843445, "learning_rate": 9.998383328367136e-05, "loss": 0.0511, "step": 5770 }, { "grad_norm": 0.41810259222984314, "learning_rate": 9.99834101124975e-05, "loss": 0.0535, "step": 5780 }, { "grad_norm": 0.36855173110961914, "learning_rate": 9.998298147521309e-05, "loss": 0.0476, "step": 5790 }, { "grad_norm": 0.4812522828578949, "learning_rate": 9.998254737186496e-05, "loss": 0.0498, "step": 5800 }, { "grad_norm": 0.4027993083000183, "learning_rate": 9.99821078025006e-05, "loss": 0.048, "step": 5810 }, { "grad_norm": 0.41386333107948303, "learning_rate": 9.998166276716807e-05, "loss": 0.046, "step": 5820 }, { "grad_norm": 0.431102454662323, "learning_rate": 9.998121226591606e-05, "loss": 0.0458, "step": 5830 }, { "grad_norm": 0.4758017063140869, "learning_rate": 9.998075629879382e-05, "loss": 0.0456, "step": 5840 }, { "grad_norm": 0.5174846649169922, "learning_rate": 9.99802948658512e-05, "loss": 0.0465, "step": 5850 }, { "grad_norm": 0.43040338158607483, "learning_rate": 9.99798279671387e-05, "loss": 0.0467, "step": 5860 }, { "grad_norm": 0.43140262365341187, "learning_rate": 9.997935560270734e-05, "loss": 0.0429, "step": 5870 }, { "grad_norm": 0.4723197817802429, "learning_rate": 9.997887777260879e-05, "loss": 0.0472, "step": 5880 }, { "grad_norm": 0.3241417109966278, "learning_rate": 9.997839447689532e-05, "loss": 0.0449, "step": 5890 }, { "grad_norm": 0.3333324193954468, "learning_rate": 9.997790571561978e-05, "loss": 0.0498, "step": 5900 }, { "grad_norm": 0.39329642057418823, "learning_rate": 9.99774114888356e-05, "loss": 0.0526, "step": 5910 }, { "grad_norm": 0.3293435871601105, "learning_rate": 9.997691179659684e-05, "loss": 0.0519, "step": 5920 }, { "grad_norm": 0.4313287138938904, "learning_rate": 9.997640663895815e-05, "loss": 0.0516, "step": 5930 }, { "grad_norm": 0.47420838475227356, "learning_rate": 9.997589601597477e-05, "loss": 0.052, "step": 5940 }, { "grad_norm": 0.3627326786518097, "learning_rate": 9.997537992770252e-05, "loss": 0.0519, "step": 5950 }, { "grad_norm": 0.39799079298973083, "learning_rate": 9.997485837419788e-05, "loss": 0.0444, "step": 5960 }, { "grad_norm": 0.39310961961746216, "learning_rate": 9.997433135551786e-05, "loss": 0.0465, "step": 5970 }, { "grad_norm": 0.42606884241104126, "learning_rate": 9.997379887172009e-05, "loss": 0.045, "step": 5980 }, { "grad_norm": 0.4562212824821472, "learning_rate": 9.997326092286281e-05, "loss": 0.0453, "step": 5990 }, { "grad_norm": 0.4937647581100464, "learning_rate": 9.997271750900486e-05, "loss": 0.0456, "step": 6000 }, { "grad_norm": 0.39182135462760925, "learning_rate": 9.997216863020565e-05, "loss": 0.0513, "step": 6010 }, { "grad_norm": 0.48334723711013794, "learning_rate": 9.99716142865252e-05, "loss": 0.0469, "step": 6020 }, { "grad_norm": 0.4135109484195709, "learning_rate": 9.997105447802415e-05, "loss": 0.0521, "step": 6030 }, { "grad_norm": 0.3801746964454651, "learning_rate": 9.997048920476373e-05, "loss": 0.047, "step": 6040 }, { "grad_norm": 0.38572952151298523, "learning_rate": 9.996991846680572e-05, "loss": 0.0442, "step": 6050 }, { "grad_norm": 0.31702500581741333, "learning_rate": 9.996934226421257e-05, "loss": 0.043, "step": 6060 }, { "grad_norm": 0.3369702696800232, "learning_rate": 9.996876059704726e-05, "loss": 0.0411, "step": 6070 }, { "grad_norm": 0.47830265760421753, "learning_rate": 9.996817346537343e-05, "loss": 0.0467, "step": 6080 }, { "grad_norm": 0.3039551079273224, "learning_rate": 9.996758086925526e-05, "loss": 0.0434, "step": 6090 }, { "grad_norm": 0.3026672303676605, "learning_rate": 9.996698280875759e-05, "loss": 0.0431, "step": 6100 }, { "grad_norm": 0.4486963450908661, "learning_rate": 9.99663792839458e-05, "loss": 0.0464, "step": 6110 }, { "grad_norm": 0.4063723385334015, "learning_rate": 9.99657702948859e-05, "loss": 0.049, "step": 6120 }, { "grad_norm": 0.48092666268348694, "learning_rate": 9.996515584164448e-05, "loss": 0.0483, "step": 6130 }, { "grad_norm": 0.36563318967819214, "learning_rate": 9.996453592428873e-05, "loss": 0.0449, "step": 6140 }, { "grad_norm": 0.34046703577041626, "learning_rate": 9.996391054288646e-05, "loss": 0.0416, "step": 6150 }, { "grad_norm": 0.28409069776535034, "learning_rate": 9.996327969750605e-05, "loss": 0.0429, "step": 6160 }, { "grad_norm": 0.30896613001823425, "learning_rate": 9.996264338821649e-05, "loss": 0.0402, "step": 6170 }, { "grad_norm": 0.45243898034095764, "learning_rate": 9.996200161508735e-05, "loss": 0.0417, "step": 6180 }, { "grad_norm": 0.46383383870124817, "learning_rate": 9.996135437818885e-05, "loss": 0.0466, "step": 6190 }, { "grad_norm": 0.38687121868133545, "learning_rate": 9.996070167759175e-05, "loss": 0.0427, "step": 6200 }, { "grad_norm": 0.404679536819458, "learning_rate": 9.996004351336743e-05, "loss": 0.0426, "step": 6210 }, { "grad_norm": 0.6573272943496704, "learning_rate": 9.995937988558785e-05, "loss": 0.0448, "step": 6220 }, { "grad_norm": 0.38580065965652466, "learning_rate": 9.995871079432561e-05, "loss": 0.0455, "step": 6230 }, { "grad_norm": 0.4093706011772156, "learning_rate": 9.995803623965389e-05, "loss": 0.05, "step": 6240 }, { "grad_norm": 0.3351023197174072, "learning_rate": 9.995735622164641e-05, "loss": 0.0447, "step": 6250 }, { "grad_norm": 0.3639284074306488, "learning_rate": 9.995667074037758e-05, "loss": 0.0429, "step": 6260 }, { "grad_norm": 0.3530043959617615, "learning_rate": 9.995597979592232e-05, "loss": 0.0402, "step": 6270 }, { "grad_norm": 0.36565449833869934, "learning_rate": 9.995528338835625e-05, "loss": 0.0402, "step": 6280 }, { "grad_norm": 0.296699196100235, "learning_rate": 9.995458151775547e-05, "loss": 0.0435, "step": 6290 }, { "grad_norm": 0.5972246527671814, "learning_rate": 9.995387418419677e-05, "loss": 0.0409, "step": 6300 }, { "grad_norm": 0.29363206028938293, "learning_rate": 9.99531613877575e-05, "loss": 0.043, "step": 6310 }, { "grad_norm": 0.3079931437969208, "learning_rate": 9.995244312851559e-05, "loss": 0.0458, "step": 6320 }, { "grad_norm": 0.384907603263855, "learning_rate": 9.995171940654961e-05, "loss": 0.0461, "step": 6330 }, { "grad_norm": 0.3829537034034729, "learning_rate": 9.995099022193871e-05, "loss": 0.046, "step": 6340 }, { "grad_norm": 0.2939909100532532, "learning_rate": 9.995025557476261e-05, "loss": 0.0503, "step": 6350 }, { "grad_norm": 0.30880603194236755, "learning_rate": 9.994951546510165e-05, "loss": 0.045, "step": 6360 }, { "grad_norm": 0.37984010577201843, "learning_rate": 9.994876989303679e-05, "loss": 0.044, "step": 6370 }, { "grad_norm": 0.4509146213531494, "learning_rate": 9.994801885864955e-05, "loss": 0.0411, "step": 6380 }, { "grad_norm": 0.361045777797699, "learning_rate": 9.994726236202205e-05, "loss": 0.0428, "step": 6390 }, { "grad_norm": 0.35288840532302856, "learning_rate": 9.994650040323704e-05, "loss": 0.044, "step": 6400 }, { "grad_norm": 0.3510708212852478, "learning_rate": 9.994573298237784e-05, "loss": 0.0441, "step": 6410 }, { "grad_norm": 0.3716573715209961, "learning_rate": 9.994496009952837e-05, "loss": 0.0416, "step": 6420 }, { "grad_norm": 0.41743728518486023, "learning_rate": 9.994418175477316e-05, "loss": 0.0428, "step": 6430 }, { "grad_norm": 0.41866374015808105, "learning_rate": 9.994339794819733e-05, "loss": 0.048, "step": 6440 }, { "grad_norm": 0.37289369106292725, "learning_rate": 9.994260867988658e-05, "loss": 0.0444, "step": 6450 }, { "grad_norm": 0.38568946719169617, "learning_rate": 9.994181394992723e-05, "loss": 0.0452, "step": 6460 }, { "grad_norm": 0.31941941380500793, "learning_rate": 9.994101375840618e-05, "loss": 0.0435, "step": 6470 }, { "grad_norm": 0.6113542914390564, "learning_rate": 9.994020810541098e-05, "loss": 0.0453, "step": 6480 }, { "grad_norm": 0.4383511543273926, "learning_rate": 9.99393969910297e-05, "loss": 0.0478, "step": 6490 }, { "grad_norm": 0.34374091029167175, "learning_rate": 9.993858041535104e-05, "loss": 0.0471, "step": 6500 }, { "grad_norm": 0.3596305251121521, "learning_rate": 9.99377583784643e-05, "loss": 0.0472, "step": 6510 }, { "grad_norm": 0.3576115667819977, "learning_rate": 9.993693088045939e-05, "loss": 0.0412, "step": 6520 }, { "grad_norm": 0.4310414493083954, "learning_rate": 9.99360979214268e-05, "loss": 0.0476, "step": 6530 }, { "grad_norm": 0.4244522154331207, "learning_rate": 9.99352595014576e-05, "loss": 0.0487, "step": 6540 }, { "grad_norm": 0.3861454427242279, "learning_rate": 9.993441562064354e-05, "loss": 0.0506, "step": 6550 }, { "grad_norm": 0.3318592309951782, "learning_rate": 9.993356627907685e-05, "loss": 0.0464, "step": 6560 }, { "grad_norm": 0.4420553147792816, "learning_rate": 9.99327114768504e-05, "loss": 0.0456, "step": 6570 }, { "grad_norm": 0.32915547490119934, "learning_rate": 9.99318512140577e-05, "loss": 0.0391, "step": 6580 }, { "grad_norm": 0.3720093071460724, "learning_rate": 9.993098549079284e-05, "loss": 0.0429, "step": 6590 }, { "grad_norm": 0.44665902853012085, "learning_rate": 9.993011430715047e-05, "loss": 0.0443, "step": 6600 }, { "grad_norm": 0.3117353022098541, "learning_rate": 9.992923766322586e-05, "loss": 0.0439, "step": 6610 }, { "grad_norm": 0.32220280170440674, "learning_rate": 9.99283555591149e-05, "loss": 0.0464, "step": 6620 }, { "grad_norm": 0.3665585219860077, "learning_rate": 9.992746799491404e-05, "loss": 0.0515, "step": 6630 }, { "grad_norm": 0.3030692934989929, "learning_rate": 9.992657497072033e-05, "loss": 0.0503, "step": 6640 }, { "grad_norm": 0.32841742038726807, "learning_rate": 9.992567648663147e-05, "loss": 0.0453, "step": 6650 }, { "grad_norm": 0.402860552072525, "learning_rate": 9.992477254274568e-05, "loss": 0.0431, "step": 6660 }, { "grad_norm": 0.3936229944229126, "learning_rate": 9.992386313916183e-05, "loss": 0.0468, "step": 6670 }, { "grad_norm": 0.3430684804916382, "learning_rate": 9.992294827597934e-05, "loss": 0.0458, "step": 6680 }, { "grad_norm": 0.3606080114841461, "learning_rate": 9.992202795329831e-05, "loss": 0.0425, "step": 6690 }, { "grad_norm": 0.32103943824768066, "learning_rate": 9.992110217121936e-05, "loss": 0.0402, "step": 6700 }, { "grad_norm": 0.3335001468658447, "learning_rate": 9.992017092984372e-05, "loss": 0.0413, "step": 6710 }, { "grad_norm": 0.5532493591308594, "learning_rate": 9.991923422927326e-05, "loss": 0.0475, "step": 6720 }, { "grad_norm": 1.0991729497909546, "learning_rate": 9.991829206961037e-05, "loss": 0.0449, "step": 6730 }, { "grad_norm": 0.43903008103370667, "learning_rate": 9.991734445095813e-05, "loss": 0.05, "step": 6740 }, { "grad_norm": 0.29700767993927, "learning_rate": 9.991639137342015e-05, "loss": 0.0536, "step": 6750 }, { "grad_norm": 0.5213426947593689, "learning_rate": 9.991543283710064e-05, "loss": 0.0456, "step": 6760 }, { "grad_norm": 0.43933889269828796, "learning_rate": 9.991446884210445e-05, "loss": 0.0437, "step": 6770 }, { "grad_norm": 0.35694554448127747, "learning_rate": 9.9913499388537e-05, "loss": 0.0437, "step": 6780 }, { "grad_norm": 0.270111620426178, "learning_rate": 9.99125244765043e-05, "loss": 0.0471, "step": 6790 }, { "grad_norm": 0.379349023103714, "learning_rate": 9.991154410611296e-05, "loss": 0.0444, "step": 6800 }, { "grad_norm": 0.5090283155441284, "learning_rate": 9.99105582774702e-05, "loss": 0.044, "step": 6810 }, { "grad_norm": 0.40633219480514526, "learning_rate": 9.990956699068384e-05, "loss": 0.043, "step": 6820 }, { "grad_norm": 0.32212167978286743, "learning_rate": 9.990857024586224e-05, "loss": 0.0421, "step": 6830 }, { "grad_norm": 0.3827524185180664, "learning_rate": 9.990756804311446e-05, "loss": 0.0404, "step": 6840 }, { "grad_norm": 0.3704434633255005, "learning_rate": 9.990656038255006e-05, "loss": 0.0419, "step": 6850 }, { "grad_norm": 0.3431469202041626, "learning_rate": 9.990554726427926e-05, "loss": 0.0419, "step": 6860 }, { "grad_norm": 0.3943329155445099, "learning_rate": 9.990452868841284e-05, "loss": 0.0457, "step": 6870 }, { "grad_norm": 0.34736916422843933, "learning_rate": 9.99035046550622e-05, "loss": 0.0434, "step": 6880 }, { "grad_norm": 0.38886937499046326, "learning_rate": 9.99024751643393e-05, "loss": 0.0388, "step": 6890 }, { "grad_norm": 0.2960921823978424, "learning_rate": 9.990144021635677e-05, "loss": 0.0436, "step": 6900 }, { "grad_norm": 0.35149356722831726, "learning_rate": 9.990039981122775e-05, "loss": 0.0427, "step": 6910 }, { "grad_norm": 0.28116872906684875, "learning_rate": 9.989935394906602e-05, "loss": 0.0395, "step": 6920 }, { "grad_norm": 0.38498544692993164, "learning_rate": 9.989830262998598e-05, "loss": 0.0411, "step": 6930 }, { "grad_norm": 0.32349058985710144, "learning_rate": 9.989724585410259e-05, "loss": 0.0368, "step": 6940 }, { "grad_norm": 0.33998408913612366, "learning_rate": 9.989618362153139e-05, "loss": 0.0408, "step": 6950 }, { "grad_norm": 0.5851848125457764, "learning_rate": 9.989511593238859e-05, "loss": 0.0418, "step": 6960 }, { "grad_norm": 0.5167311429977417, "learning_rate": 9.98940427867909e-05, "loss": 0.0487, "step": 6970 }, { "grad_norm": 0.2998656630516052, "learning_rate": 9.989296418485573e-05, "loss": 0.0459, "step": 6980 }, { "grad_norm": 0.3022063672542572, "learning_rate": 9.989188012670101e-05, "loss": 0.0405, "step": 6990 }, { "grad_norm": 0.45556268095970154, "learning_rate": 9.989079061244528e-05, "loss": 0.0435, "step": 7000 }, { "grad_norm": 0.42451998591423035, "learning_rate": 9.988969564220769e-05, "loss": 0.0414, "step": 7010 }, { "grad_norm": 0.3605329394340515, "learning_rate": 9.988859521610801e-05, "loss": 0.0405, "step": 7020 }, { "grad_norm": 0.40284496545791626, "learning_rate": 9.988748933426656e-05, "loss": 0.0423, "step": 7030 }, { "grad_norm": 0.4341355860233307, "learning_rate": 9.988637799680428e-05, "loss": 0.0402, "step": 7040 }, { "grad_norm": 0.41272783279418945, "learning_rate": 9.98852612038427e-05, "loss": 0.042, "step": 7050 }, { "grad_norm": 0.27617332339286804, "learning_rate": 9.988413895550397e-05, "loss": 0.0388, "step": 7060 }, { "grad_norm": 0.3790992498397827, "learning_rate": 9.98830112519108e-05, "loss": 0.048, "step": 7070 }, { "grad_norm": 0.3599565923213959, "learning_rate": 9.98818780931865e-05, "loss": 0.049, "step": 7080 }, { "grad_norm": 0.40766891837120056, "learning_rate": 9.988073947945502e-05, "loss": 0.0483, "step": 7090 }, { "grad_norm": 0.40758898854255676, "learning_rate": 9.987959541084087e-05, "loss": 0.044, "step": 7100 }, { "grad_norm": 0.38083040714263916, "learning_rate": 9.987844588746915e-05, "loss": 0.0439, "step": 7110 }, { "grad_norm": 0.2860131561756134, "learning_rate": 9.987729090946558e-05, "loss": 0.043, "step": 7120 }, { "grad_norm": 0.5482214093208313, "learning_rate": 9.987613047695647e-05, "loss": 0.0466, "step": 7130 }, { "grad_norm": 0.33388441801071167, "learning_rate": 9.987496459006871e-05, "loss": 0.043, "step": 7140 }, { "grad_norm": 0.31956741213798523, "learning_rate": 9.987379324892982e-05, "loss": 0.0398, "step": 7150 }, { "grad_norm": 0.3215472102165222, "learning_rate": 9.987261645366788e-05, "loss": 0.0419, "step": 7160 }, { "grad_norm": 0.3366938531398773, "learning_rate": 9.987143420441158e-05, "loss": 0.0414, "step": 7170 }, { "grad_norm": 0.28086885809898376, "learning_rate": 9.987024650129022e-05, "loss": 0.041, "step": 7180 }, { "grad_norm": 0.3258735239505768, "learning_rate": 9.986905334443368e-05, "loss": 0.0466, "step": 7190 }, { "grad_norm": 0.49446579813957214, "learning_rate": 9.986785473397245e-05, "loss": 0.0455, "step": 7200 }, { "grad_norm": 0.3158825635910034, "learning_rate": 9.98666506700376e-05, "loss": 0.0442, "step": 7210 }, { "grad_norm": 0.308738648891449, "learning_rate": 9.986544115276081e-05, "loss": 0.04, "step": 7220 }, { "grad_norm": 0.2733738422393799, "learning_rate": 9.986422618227433e-05, "loss": 0.0466, "step": 7230 }, { "grad_norm": 0.3089170753955841, "learning_rate": 9.986300575871106e-05, "loss": 0.0537, "step": 7240 }, { "grad_norm": 0.42219358682632446, "learning_rate": 9.986177988220444e-05, "loss": 0.0448, "step": 7250 }, { "grad_norm": 0.314069539308548, "learning_rate": 9.986054855288856e-05, "loss": 0.0401, "step": 7260 }, { "grad_norm": 0.35659509897232056, "learning_rate": 9.985931177089802e-05, "loss": 0.0393, "step": 7270 }, { "grad_norm": 0.262362003326416, "learning_rate": 9.985806953636814e-05, "loss": 0.0395, "step": 7280 }, { "grad_norm": 0.2821563482284546, "learning_rate": 9.985682184943471e-05, "loss": 0.0396, "step": 7290 }, { "grad_norm": 0.4632941484451294, "learning_rate": 9.98555687102342e-05, "loss": 0.044, "step": 7300 }, { "grad_norm": 0.3713415563106537, "learning_rate": 9.985431011890367e-05, "loss": 0.0415, "step": 7310 }, { "grad_norm": 0.35777199268341064, "learning_rate": 9.985304607558075e-05, "loss": 0.0433, "step": 7320 }, { "grad_norm": 0.4003869891166687, "learning_rate": 9.985177658040364e-05, "loss": 0.0392, "step": 7330 }, { "grad_norm": 0.39461368322372437, "learning_rate": 9.985050163351119e-05, "loss": 0.038, "step": 7340 }, { "grad_norm": 0.2844635546207428, "learning_rate": 9.984922123504286e-05, "loss": 0.0382, "step": 7350 }, { "grad_norm": 0.2547836899757385, "learning_rate": 9.984793538513862e-05, "loss": 0.0461, "step": 7360 }, { "grad_norm": 0.3260858654975891, "learning_rate": 9.984664408393912e-05, "loss": 0.0451, "step": 7370 }, { "grad_norm": 0.3281562328338623, "learning_rate": 9.984534733158556e-05, "loss": 0.04, "step": 7380 }, { "grad_norm": 0.3725183308124542, "learning_rate": 9.984404512821977e-05, "loss": 0.0466, "step": 7390 }, { "grad_norm": 0.37659937143325806, "learning_rate": 9.984273747398411e-05, "loss": 0.0388, "step": 7400 }, { "grad_norm": 0.37474524974823, "learning_rate": 9.984142436902165e-05, "loss": 0.0396, "step": 7410 }, { "grad_norm": 0.30420559644699097, "learning_rate": 9.984010581347596e-05, "loss": 0.0376, "step": 7420 }, { "grad_norm": 0.3311018645763397, "learning_rate": 9.983878180749121e-05, "loss": 0.0374, "step": 7430 }, { "grad_norm": 0.4850529730319977, "learning_rate": 9.983745235121222e-05, "loss": 0.041, "step": 7440 }, { "grad_norm": 0.3405227065086365, "learning_rate": 9.983611744478438e-05, "loss": 0.0415, "step": 7450 }, { "grad_norm": 0.34578779339790344, "learning_rate": 9.983477708835365e-05, "loss": 0.0442, "step": 7460 }, { "grad_norm": 0.48974791169166565, "learning_rate": 9.983343128206664e-05, "loss": 0.0407, "step": 7470 }, { "grad_norm": 0.3289615511894226, "learning_rate": 9.983208002607049e-05, "loss": 0.04, "step": 7480 }, { "grad_norm": 0.2946765124797821, "learning_rate": 9.9830723320513e-05, "loss": 0.036, "step": 7490 }, { "grad_norm": 0.34836238622665405, "learning_rate": 9.982936116554254e-05, "loss": 0.0378, "step": 7500 }, { "grad_norm": 0.28835296630859375, "learning_rate": 9.982799356130803e-05, "loss": 0.0343, "step": 7510 }, { "grad_norm": 0.2969600260257721, "learning_rate": 9.982662050795908e-05, "loss": 0.0397, "step": 7520 }, { "grad_norm": 0.3461100459098816, "learning_rate": 9.982524200564583e-05, "loss": 0.036, "step": 7530 }, { "grad_norm": 0.45183971524238586, "learning_rate": 9.982385805451901e-05, "loss": 0.0364, "step": 7540 }, { "grad_norm": 0.3976472020149231, "learning_rate": 9.982246865472998e-05, "loss": 0.0397, "step": 7550 }, { "grad_norm": 0.3713207542896271, "learning_rate": 9.982107380643069e-05, "loss": 0.0424, "step": 7560 }, { "grad_norm": 0.3515833020210266, "learning_rate": 9.981967350977368e-05, "loss": 0.0394, "step": 7570 }, { "grad_norm": 0.30707886815071106, "learning_rate": 9.981826776491208e-05, "loss": 0.0367, "step": 7580 }, { "grad_norm": 0.28875887393951416, "learning_rate": 9.98168565719996e-05, "loss": 0.0382, "step": 7590 }, { "grad_norm": 0.29953739047050476, "learning_rate": 9.98154399311906e-05, "loss": 0.0381, "step": 7600 }, { "grad_norm": 0.35191428661346436, "learning_rate": 9.981401784263997e-05, "loss": 0.0371, "step": 7610 }, { "grad_norm": 0.3902840316295624, "learning_rate": 9.981259030650326e-05, "loss": 0.0375, "step": 7620 }, { "grad_norm": 0.3622820973396301, "learning_rate": 9.981115732293655e-05, "loss": 0.0394, "step": 7630 }, { "grad_norm": 0.30041900277137756, "learning_rate": 9.980971889209659e-05, "loss": 0.0389, "step": 7640 }, { "grad_norm": 0.2692466378211975, "learning_rate": 9.980827501414064e-05, "loss": 0.0375, "step": 7650 }, { "grad_norm": 0.26088616251945496, "learning_rate": 9.980682568922663e-05, "loss": 0.043, "step": 7660 }, { "grad_norm": 0.261846125125885, "learning_rate": 9.980537091751304e-05, "loss": 0.0405, "step": 7670 }, { "grad_norm": 0.33335351943969727, "learning_rate": 9.980391069915897e-05, "loss": 0.0367, "step": 7680 }, { "grad_norm": 0.3177869915962219, "learning_rate": 9.98024450343241e-05, "loss": 0.0349, "step": 7690 }, { "grad_norm": 0.28887850046157837, "learning_rate": 9.980097392316872e-05, "loss": 0.0385, "step": 7700 }, { "grad_norm": 0.31602832674980164, "learning_rate": 9.97994973658537e-05, "loss": 0.0383, "step": 7710 }, { "grad_norm": 0.38739556074142456, "learning_rate": 9.979801536254054e-05, "loss": 0.0356, "step": 7720 }, { "grad_norm": 0.3721819221973419, "learning_rate": 9.979652791339127e-05, "loss": 0.0365, "step": 7730 }, { "grad_norm": 0.3343510925769806, "learning_rate": 9.97950350185686e-05, "loss": 0.0364, "step": 7740 }, { "grad_norm": 0.2613339424133301, "learning_rate": 9.979353667823574e-05, "loss": 0.0394, "step": 7750 }, { "grad_norm": 0.24123772978782654, "learning_rate": 9.979203289255658e-05, "loss": 0.0353, "step": 7760 }, { "grad_norm": 0.3250539302825928, "learning_rate": 9.979052366169557e-05, "loss": 0.0413, "step": 7770 }, { "grad_norm": 0.25741076469421387, "learning_rate": 9.978900898581775e-05, "loss": 0.0341, "step": 7780 }, { "grad_norm": 0.32393398880958557, "learning_rate": 9.978748886508875e-05, "loss": 0.0351, "step": 7790 }, { "grad_norm": 0.3532504141330719, "learning_rate": 9.978596329967484e-05, "loss": 0.0388, "step": 7800 }, { "grad_norm": 0.31908538937568665, "learning_rate": 9.978443228974284e-05, "loss": 0.0364, "step": 7810 }, { "grad_norm": 0.35865190625190735, "learning_rate": 9.978289583546015e-05, "loss": 0.0342, "step": 7820 }, { "grad_norm": 0.26574432849884033, "learning_rate": 9.978135393699484e-05, "loss": 0.0387, "step": 7830 }, { "grad_norm": 0.4169375002384186, "learning_rate": 9.977980659451548e-05, "loss": 0.035, "step": 7840 }, { "grad_norm": 0.34893181920051575, "learning_rate": 9.977825380819135e-05, "loss": 0.0366, "step": 7850 }, { "grad_norm": 0.3056526780128479, "learning_rate": 9.97766955781922e-05, "loss": 0.0379, "step": 7860 }, { "grad_norm": 0.37551724910736084, "learning_rate": 9.977513190468848e-05, "loss": 0.0351, "step": 7870 }, { "grad_norm": 0.4234655797481537, "learning_rate": 9.977356278785116e-05, "loss": 0.0333, "step": 7880 }, { "grad_norm": 0.3499770164489746, "learning_rate": 9.977198822785184e-05, "loss": 0.0399, "step": 7890 }, { "grad_norm": 0.35572385787963867, "learning_rate": 9.977040822486273e-05, "loss": 0.0393, "step": 7900 }, { "grad_norm": 0.4107483923435211, "learning_rate": 9.97688227790566e-05, "loss": 0.0401, "step": 7910 }, { "grad_norm": 0.31210464239120483, "learning_rate": 9.976723189060684e-05, "loss": 0.0349, "step": 7920 }, { "grad_norm": 0.24909284710884094, "learning_rate": 9.976563555968742e-05, "loss": 0.0361, "step": 7930 }, { "grad_norm": 0.2775442600250244, "learning_rate": 9.976403378647292e-05, "loss": 0.0374, "step": 7940 }, { "grad_norm": 0.36058053374290466, "learning_rate": 9.97624265711385e-05, "loss": 0.0376, "step": 7950 }, { "grad_norm": 0.3358818292617798, "learning_rate": 9.976081391385993e-05, "loss": 0.0398, "step": 7960 }, { "grad_norm": 0.31835833191871643, "learning_rate": 9.975919581481356e-05, "loss": 0.0421, "step": 7970 }, { "grad_norm": 0.5814517140388489, "learning_rate": 9.975757227417634e-05, "loss": 0.0422, "step": 7980 }, { "grad_norm": 0.361295610666275, "learning_rate": 9.975594329212586e-05, "loss": 0.0402, "step": 7990 }, { "grad_norm": 0.35282453894615173, "learning_rate": 9.97543088688402e-05, "loss": 0.0362, "step": 8000 }, { "grad_norm": 0.32858678698539734, "learning_rate": 9.975266900449814e-05, "loss": 0.0319, "step": 8010 }, { "grad_norm": 0.3746180236339569, "learning_rate": 9.975102369927898e-05, "loss": 0.0353, "step": 8020 }, { "grad_norm": 0.41977497935295105, "learning_rate": 9.974937295336269e-05, "loss": 0.0371, "step": 8030 }, { "grad_norm": 0.3215549886226654, "learning_rate": 9.974771676692975e-05, "loss": 0.0381, "step": 8040 }, { "grad_norm": 0.34216850996017456, "learning_rate": 9.974605514016131e-05, "loss": 0.0392, "step": 8050 }, { "grad_norm": 0.2829672694206238, "learning_rate": 9.974438807323907e-05, "loss": 0.0428, "step": 8060 }, { "grad_norm": 0.41252729296684265, "learning_rate": 9.974271556634535e-05, "loss": 0.0383, "step": 8070 }, { "grad_norm": 0.4220689833164215, "learning_rate": 9.974103761966302e-05, "loss": 0.0444, "step": 8080 }, { "grad_norm": 0.37492066621780396, "learning_rate": 9.973935423337563e-05, "loss": 0.0435, "step": 8090 }, { "grad_norm": 0.25259488821029663, "learning_rate": 9.973766540766722e-05, "loss": 0.0412, "step": 8100 }, { "grad_norm": 0.266946405172348, "learning_rate": 9.97359711427225e-05, "loss": 0.0367, "step": 8110 }, { "grad_norm": 0.4091462790966034, "learning_rate": 9.973427143872677e-05, "loss": 0.0366, "step": 8120 }, { "grad_norm": 0.3725566267967224, "learning_rate": 9.973256629586589e-05, "loss": 0.0383, "step": 8130 }, { "grad_norm": 0.39160871505737305, "learning_rate": 9.973085571432632e-05, "loss": 0.0367, "step": 8140 }, { "grad_norm": 0.31658098101615906, "learning_rate": 9.972913969429513e-05, "loss": 0.0377, "step": 8150 }, { "grad_norm": 0.3017365634441376, "learning_rate": 9.972741823596e-05, "loss": 0.0355, "step": 8160 }, { "grad_norm": 0.29577383399009705, "learning_rate": 9.972569133950917e-05, "loss": 0.0373, "step": 8170 }, { "grad_norm": 0.30785679817199707, "learning_rate": 9.972395900513151e-05, "loss": 0.0367, "step": 8180 }, { "grad_norm": 0.2807563543319702, "learning_rate": 9.972222123301645e-05, "loss": 0.0364, "step": 8190 }, { "grad_norm": 0.28085264563560486, "learning_rate": 9.972047802335403e-05, "loss": 0.0361, "step": 8200 }, { "grad_norm": 0.26224270462989807, "learning_rate": 9.971872937633488e-05, "loss": 0.0341, "step": 8210 }, { "grad_norm": 0.29678311944007874, "learning_rate": 9.971697529215024e-05, "loss": 0.0324, "step": 8220 }, { "grad_norm": 0.3179382383823395, "learning_rate": 9.971521577099192e-05, "loss": 0.0349, "step": 8230 }, { "grad_norm": 0.34122326970100403, "learning_rate": 9.971345081305236e-05, "loss": 0.0368, "step": 8240 }, { "grad_norm": 0.3802780508995056, "learning_rate": 9.971168041852456e-05, "loss": 0.0403, "step": 8250 }, { "grad_norm": 0.34532594680786133, "learning_rate": 9.970990458760215e-05, "loss": 0.0364, "step": 8260 }, { "grad_norm": 0.35550743341445923, "learning_rate": 9.970812332047929e-05, "loss": 0.0351, "step": 8270 }, { "grad_norm": 0.5211509466171265, "learning_rate": 9.97063366173508e-05, "loss": 0.0365, "step": 8280 }, { "grad_norm": 0.43288078904151917, "learning_rate": 9.970454447841207e-05, "loss": 0.0426, "step": 8290 }, { "grad_norm": 0.33441707491874695, "learning_rate": 9.970274690385909e-05, "loss": 0.0374, "step": 8300 }, { "grad_norm": 0.3190881311893463, "learning_rate": 9.970094389388844e-05, "loss": 0.0436, "step": 8310 }, { "grad_norm": 0.3559439182281494, "learning_rate": 9.969913544869728e-05, "loss": 0.043, "step": 8320 }, { "grad_norm": 0.3029703199863434, "learning_rate": 9.96973215684834e-05, "loss": 0.0432, "step": 8330 }, { "grad_norm": 0.2638092339038849, "learning_rate": 9.969550225344513e-05, "loss": 0.0387, "step": 8340 }, { "grad_norm": 0.36929193139076233, "learning_rate": 9.969367750378147e-05, "loss": 0.0419, "step": 8350 }, { "grad_norm": 0.3397532105445862, "learning_rate": 9.969184731969194e-05, "loss": 0.0381, "step": 8360 }, { "grad_norm": 0.3380941152572632, "learning_rate": 9.96900117013767e-05, "loss": 0.039, "step": 8370 }, { "grad_norm": 0.3001883327960968, "learning_rate": 9.96881706490365e-05, "loss": 0.0383, "step": 8380 }, { "grad_norm": 0.25531861186027527, "learning_rate": 9.968632416287265e-05, "loss": 0.0419, "step": 8390 }, { "grad_norm": 0.3516106605529785, "learning_rate": 9.96844722430871e-05, "loss": 0.0381, "step": 8400 }, { "grad_norm": 0.3168662488460541, "learning_rate": 9.968261488988235e-05, "loss": 0.0414, "step": 8410 }, { "grad_norm": 0.2949429452419281, "learning_rate": 9.968075210346155e-05, "loss": 0.0357, "step": 8420 }, { "grad_norm": 0.2480984777212143, "learning_rate": 9.967888388402839e-05, "loss": 0.0387, "step": 8430 }, { "grad_norm": 0.3708096742630005, "learning_rate": 9.967701023178717e-05, "loss": 0.04, "step": 8440 }, { "grad_norm": 0.3630291819572449, "learning_rate": 9.967513114694282e-05, "loss": 0.0441, "step": 8450 }, { "grad_norm": 0.3120895028114319, "learning_rate": 9.967324662970079e-05, "loss": 0.0427, "step": 8460 }, { "grad_norm": 0.32662147283554077, "learning_rate": 9.96713566802672e-05, "loss": 0.0375, "step": 8470 }, { "grad_norm": 0.4241991639137268, "learning_rate": 9.966946129884873e-05, "loss": 0.0379, "step": 8480 }, { "grad_norm": 0.33441615104675293, "learning_rate": 9.966756048565265e-05, "loss": 0.0384, "step": 8490 }, { "grad_norm": 0.2909177541732788, "learning_rate": 9.966565424088681e-05, "loss": 0.0381, "step": 8500 }, { "grad_norm": 0.24089492857456207, "learning_rate": 9.96637425647597e-05, "loss": 0.0354, "step": 8510 }, { "grad_norm": 0.33868229389190674, "learning_rate": 9.966182545748038e-05, "loss": 0.0367, "step": 8520 }, { "grad_norm": 0.38544896245002747, "learning_rate": 9.96599029192585e-05, "loss": 0.0369, "step": 8530 }, { "grad_norm": 0.3475954532623291, "learning_rate": 9.965797495030428e-05, "loss": 0.0388, "step": 8540 }, { "grad_norm": 0.3243304491043091, "learning_rate": 9.96560415508286e-05, "loss": 0.0342, "step": 8550 }, { "grad_norm": 0.7492494583129883, "learning_rate": 9.965410272104286e-05, "loss": 0.0388, "step": 8560 }, { "grad_norm": 0.3595585227012634, "learning_rate": 9.96521584611591e-05, "loss": 0.0465, "step": 8570 }, { "grad_norm": 0.31151872873306274, "learning_rate": 9.965020877138994e-05, "loss": 0.0382, "step": 8580 }, { "grad_norm": 0.36602431535720825, "learning_rate": 9.964825365194861e-05, "loss": 0.0397, "step": 8590 }, { "grad_norm": 0.33451688289642334, "learning_rate": 9.96462931030489e-05, "loss": 0.0381, "step": 8600 }, { "grad_norm": 0.4313506782054901, "learning_rate": 9.96443271249052e-05, "loss": 0.0414, "step": 8610 }, { "grad_norm": 0.2977641224861145, "learning_rate": 9.964235571773255e-05, "loss": 0.0429, "step": 8620 }, { "grad_norm": 0.28711560368537903, "learning_rate": 9.96403788817465e-05, "loss": 0.0423, "step": 8630 }, { "grad_norm": 0.3444620966911316, "learning_rate": 9.963839661716325e-05, "loss": 0.0401, "step": 8640 }, { "grad_norm": 0.38503581285476685, "learning_rate": 9.963640892419958e-05, "loss": 0.042, "step": 8650 }, { "grad_norm": 0.4031970500946045, "learning_rate": 9.963441580307286e-05, "loss": 0.0424, "step": 8660 }, { "grad_norm": 0.3895817995071411, "learning_rate": 9.963241725400104e-05, "loss": 0.0461, "step": 8670 }, { "grad_norm": 0.33094656467437744, "learning_rate": 9.963041327720271e-05, "loss": 0.0453, "step": 8680 }, { "grad_norm": 0.2599477469921112, "learning_rate": 9.962840387289697e-05, "loss": 0.0382, "step": 8690 }, { "grad_norm": 0.3108479678630829, "learning_rate": 9.962638904130363e-05, "loss": 0.0351, "step": 8700 }, { "grad_norm": 0.31531035900115967, "learning_rate": 9.962436878264298e-05, "loss": 0.0392, "step": 8710 }, { "grad_norm": 0.3593563735485077, "learning_rate": 9.962234309713598e-05, "loss": 0.039, "step": 8720 }, { "grad_norm": 0.4983099102973938, "learning_rate": 9.962031198500414e-05, "loss": 0.0375, "step": 8730 }, { "grad_norm": 0.30155250430107117, "learning_rate": 9.961827544646958e-05, "loss": 0.0349, "step": 8740 }, { "grad_norm": 0.3195149600505829, "learning_rate": 9.961623348175501e-05, "loss": 0.0352, "step": 8750 }, { "grad_norm": 0.358973890542984, "learning_rate": 9.961418609108377e-05, "loss": 0.0351, "step": 8760 }, { "grad_norm": 0.34750425815582275, "learning_rate": 9.961213327467971e-05, "loss": 0.0416, "step": 8770 }, { "grad_norm": 0.3297937512397766, "learning_rate": 9.961007503276736e-05, "loss": 0.0397, "step": 8780 }, { "grad_norm": 0.29339170455932617, "learning_rate": 9.960801136557179e-05, "loss": 0.0365, "step": 8790 }, { "grad_norm": 0.31847354769706726, "learning_rate": 9.960594227331866e-05, "loss": 0.034, "step": 8800 }, { "grad_norm": 0.3266264498233795, "learning_rate": 9.960386775623429e-05, "loss": 0.0351, "step": 8810 }, { "grad_norm": 0.3881116509437561, "learning_rate": 9.96017878145455e-05, "loss": 0.035, "step": 8820 }, { "grad_norm": 0.4014137089252472, "learning_rate": 9.959970244847977e-05, "loss": 0.0376, "step": 8830 }, { "grad_norm": 0.30747273564338684, "learning_rate": 9.959761165826518e-05, "loss": 0.0395, "step": 8840 }, { "grad_norm": 0.24657438695430756, "learning_rate": 9.959551544413033e-05, "loss": 0.0379, "step": 8850 }, { "grad_norm": 0.3416788876056671, "learning_rate": 9.959341380630448e-05, "loss": 0.0345, "step": 8860 }, { "grad_norm": 0.3308376371860504, "learning_rate": 9.959130674501746e-05, "loss": 0.0367, "step": 8870 }, { "grad_norm": 0.29626962542533875, "learning_rate": 9.958919426049968e-05, "loss": 0.0316, "step": 8880 }, { "grad_norm": 0.4179859757423401, "learning_rate": 9.958707635298219e-05, "loss": 0.0399, "step": 8890 }, { "grad_norm": 0.2961205542087555, "learning_rate": 9.958495302269657e-05, "loss": 0.0404, "step": 8900 }, { "grad_norm": 0.34605273604393005, "learning_rate": 9.958282426987503e-05, "loss": 0.0373, "step": 8910 }, { "grad_norm": 0.42164671421051025, "learning_rate": 9.95806900947504e-05, "loss": 0.0406, "step": 8920 }, { "grad_norm": 0.4203040897846222, "learning_rate": 9.957855049755604e-05, "loss": 0.0447, "step": 8930 }, { "grad_norm": 0.339720755815506, "learning_rate": 9.957640547852593e-05, "loss": 0.0355, "step": 8940 }, { "grad_norm": 0.28102797269821167, "learning_rate": 9.957425503789466e-05, "loss": 0.0348, "step": 8950 }, { "grad_norm": 0.3584672212600708, "learning_rate": 9.957209917589738e-05, "loss": 0.04, "step": 8960 }, { "grad_norm": 0.3205435574054718, "learning_rate": 9.956993789276987e-05, "loss": 0.0413, "step": 8970 }, { "grad_norm": 0.39265552163124084, "learning_rate": 9.956777118874847e-05, "loss": 0.0391, "step": 8980 }, { "grad_norm": 0.2849726378917694, "learning_rate": 9.956559906407016e-05, "loss": 0.0387, "step": 8990 }, { "grad_norm": 0.33931368589401245, "learning_rate": 9.956342151897245e-05, "loss": 0.0388, "step": 9000 }, { "grad_norm": 0.35004332661628723, "learning_rate": 9.956123855369346e-05, "loss": 0.04, "step": 9010 }, { "grad_norm": 0.31862208247184753, "learning_rate": 9.955905016847196e-05, "loss": 0.0408, "step": 9020 }, { "grad_norm": 0.343330979347229, "learning_rate": 9.955685636354723e-05, "loss": 0.0392, "step": 9030 }, { "grad_norm": 0.2658959925174713, "learning_rate": 9.95546571391592e-05, "loss": 0.0344, "step": 9040 }, { "grad_norm": 0.24726195633411407, "learning_rate": 9.955245249554837e-05, "loss": 0.0328, "step": 9050 }, { "grad_norm": 0.3093395531177521, "learning_rate": 9.955024243295582e-05, "loss": 0.0333, "step": 9060 }, { "grad_norm": 0.3366636037826538, "learning_rate": 9.954802695162328e-05, "loss": 0.0344, "step": 9070 }, { "grad_norm": 0.2997897267341614, "learning_rate": 9.954580605179302e-05, "loss": 0.0376, "step": 9080 }, { "grad_norm": 0.24989160895347595, "learning_rate": 9.954357973370788e-05, "loss": 0.0339, "step": 9090 }, { "grad_norm": 0.41685619950294495, "learning_rate": 9.954134799761135e-05, "loss": 0.0349, "step": 9100 }, { "grad_norm": 0.27845439314842224, "learning_rate": 9.953911084374748e-05, "loss": 0.0365, "step": 9110 }, { "grad_norm": 0.2887788712978363, "learning_rate": 9.953686827236093e-05, "loss": 0.0385, "step": 9120 }, { "grad_norm": 0.37617602944374084, "learning_rate": 9.953462028369695e-05, "loss": 0.0354, "step": 9130 }, { "grad_norm": 0.27342483401298523, "learning_rate": 9.953236687800136e-05, "loss": 0.034, "step": 9140 }, { "grad_norm": 0.31782764196395874, "learning_rate": 9.95301080555206e-05, "loss": 0.0386, "step": 9150 }, { "grad_norm": 0.3088516294956207, "learning_rate": 9.952784381650171e-05, "loss": 0.0326, "step": 9160 }, { "grad_norm": 0.2681210935115814, "learning_rate": 9.952557416119226e-05, "loss": 0.0347, "step": 9170 }, { "grad_norm": 0.22724255919456482, "learning_rate": 9.95232990898405e-05, "loss": 0.034, "step": 9180 }, { "grad_norm": 0.2401963621377945, "learning_rate": 9.95210186026952e-05, "loss": 0.0362, "step": 9190 }, { "grad_norm": 0.444059282541275, "learning_rate": 9.951873270000576e-05, "loss": 0.0353, "step": 9200 }, { "grad_norm": 0.27141302824020386, "learning_rate": 9.951644138202216e-05, "loss": 0.0367, "step": 9210 }, { "grad_norm": 0.28152593970298767, "learning_rate": 9.951414464899498e-05, "loss": 0.038, "step": 9220 }, { "grad_norm": 0.4177212417125702, "learning_rate": 9.951184250117538e-05, "loss": 0.0404, "step": 9230 }, { "grad_norm": 0.3016221225261688, "learning_rate": 9.950953493881513e-05, "loss": 0.0388, "step": 9240 }, { "grad_norm": 0.47397592663764954, "learning_rate": 9.950722196216658e-05, "loss": 0.038, "step": 9250 }, { "grad_norm": 0.4197461009025574, "learning_rate": 9.950490357148265e-05, "loss": 0.0454, "step": 9260 }, { "grad_norm": 0.35915055871009827, "learning_rate": 9.950257976701692e-05, "loss": 0.0448, "step": 9270 }, { "grad_norm": 0.274582177400589, "learning_rate": 9.950025054902348e-05, "loss": 0.0387, "step": 9280 }, { "grad_norm": 0.2528107166290283, "learning_rate": 9.949791591775706e-05, "loss": 0.038, "step": 9290 }, { "grad_norm": 0.34214848279953003, "learning_rate": 9.949557587347298e-05, "loss": 0.0401, "step": 9300 }, { "grad_norm": 0.28271329402923584, "learning_rate": 9.949323041642713e-05, "loss": 0.0372, "step": 9310 }, { "grad_norm": 0.33326074481010437, "learning_rate": 9.949087954687602e-05, "loss": 0.0354, "step": 9320 }, { "grad_norm": 0.3144960403442383, "learning_rate": 9.948852326507672e-05, "loss": 0.0357, "step": 9330 }, { "grad_norm": 0.24789448082447052, "learning_rate": 9.948616157128694e-05, "loss": 0.0409, "step": 9340 }, { "grad_norm": 0.3888627290725708, "learning_rate": 9.948379446576493e-05, "loss": 0.0364, "step": 9350 }, { "grad_norm": 0.2675584554672241, "learning_rate": 9.948142194876952e-05, "loss": 0.0343, "step": 9360 }, { "grad_norm": 0.3080196976661682, "learning_rate": 9.947904402056024e-05, "loss": 0.036, "step": 9370 }, { "grad_norm": 0.2542737126350403, "learning_rate": 9.947666068139708e-05, "loss": 0.0345, "step": 9380 }, { "grad_norm": 0.23345474898815155, "learning_rate": 9.947427193154071e-05, "loss": 0.0344, "step": 9390 }, { "grad_norm": 0.2705834209918976, "learning_rate": 9.947187777125233e-05, "loss": 0.0365, "step": 9400 }, { "grad_norm": 0.32384148240089417, "learning_rate": 9.946947820079377e-05, "loss": 0.0359, "step": 9410 }, { "grad_norm": 0.31733977794647217, "learning_rate": 9.946707322042747e-05, "loss": 0.0334, "step": 9420 }, { "grad_norm": 0.25580257177352905, "learning_rate": 9.94646628304164e-05, "loss": 0.037, "step": 9430 }, { "grad_norm": 0.30113905668258667, "learning_rate": 9.946224703102418e-05, "loss": 0.0355, "step": 9440 }, { "grad_norm": 0.34201207756996155, "learning_rate": 9.945982582251498e-05, "loss": 0.0346, "step": 9450 }, { "grad_norm": 0.28740593791007996, "learning_rate": 9.94573992051536e-05, "loss": 0.0372, "step": 9460 }, { "grad_norm": 0.3526862561702728, "learning_rate": 9.94549671792054e-05, "loss": 0.0344, "step": 9470 }, { "grad_norm": 0.3042892515659332, "learning_rate": 9.945252974493635e-05, "loss": 0.0338, "step": 9480 }, { "grad_norm": 0.42658063769340515, "learning_rate": 9.9450086902613e-05, "loss": 0.0334, "step": 9490 }, { "grad_norm": 0.23561705648899078, "learning_rate": 9.944763865250248e-05, "loss": 0.032, "step": 9500 }, { "grad_norm": 0.2701778709888458, "learning_rate": 9.944518499487254e-05, "loss": 0.0357, "step": 9510 }, { "grad_norm": 0.27398625016212463, "learning_rate": 9.944272592999151e-05, "loss": 0.0367, "step": 9520 }, { "grad_norm": 0.25747713446617126, "learning_rate": 9.94402614581283e-05, "loss": 0.0315, "step": 9530 }, { "grad_norm": 0.2884243428707123, "learning_rate": 9.943779157955244e-05, "loss": 0.0357, "step": 9540 }, { "grad_norm": 0.2706860899925232, "learning_rate": 9.943531629453403e-05, "loss": 0.0311, "step": 9550 }, { "grad_norm": 0.30717048048973083, "learning_rate": 9.943283560334375e-05, "loss": 0.0307, "step": 9560 }, { "grad_norm": 0.2281423658132553, "learning_rate": 9.943034950625288e-05, "loss": 0.0332, "step": 9570 }, { "grad_norm": 0.31057843565940857, "learning_rate": 9.942785800353332e-05, "loss": 0.035, "step": 9580 }, { "grad_norm": 0.3019918203353882, "learning_rate": 9.942536109545751e-05, "loss": 0.0363, "step": 9590 }, { "grad_norm": 0.3791530430316925, "learning_rate": 9.942285878229853e-05, "loss": 0.0355, "step": 9600 }, { "grad_norm": 0.23747120797634125, "learning_rate": 9.942035106433001e-05, "loss": 0.0337, "step": 9610 }, { "grad_norm": 0.19965289533138275, "learning_rate": 9.94178379418262e-05, "loss": 0.0344, "step": 9620 }, { "grad_norm": 0.37639620900154114, "learning_rate": 9.941531941506194e-05, "loss": 0.0342, "step": 9630 }, { "grad_norm": 0.7673988938331604, "learning_rate": 9.941279548431263e-05, "loss": 0.0365, "step": 9640 }, { "grad_norm": 0.3935032784938812, "learning_rate": 9.941026614985431e-05, "loss": 0.0439, "step": 9650 }, { "grad_norm": 0.3655143976211548, "learning_rate": 9.940773141196357e-05, "loss": 0.0399, "step": 9660 }, { "grad_norm": 0.3674315810203552, "learning_rate": 9.94051912709176e-05, "loss": 0.0386, "step": 9670 }, { "grad_norm": 0.2960885465145111, "learning_rate": 9.940264572699421e-05, "loss": 0.0377, "step": 9680 }, { "grad_norm": 0.24093791842460632, "learning_rate": 9.940009478047174e-05, "loss": 0.038, "step": 9690 }, { "grad_norm": 0.3550066649913788, "learning_rate": 9.939753843162918e-05, "loss": 0.0369, "step": 9700 }, { "grad_norm": 0.45446231961250305, "learning_rate": 9.939497668074609e-05, "loss": 0.0398, "step": 9710 }, { "grad_norm": 0.2988639771938324, "learning_rate": 9.93924095281026e-05, "loss": 0.0365, "step": 9720 }, { "grad_norm": 0.407010555267334, "learning_rate": 9.938983697397948e-05, "loss": 0.0325, "step": 9730 }, { "grad_norm": 0.3197079300880432, "learning_rate": 9.938725901865805e-05, "loss": 0.0403, "step": 9740 }, { "grad_norm": 0.2763616740703583, "learning_rate": 9.93846756624202e-05, "loss": 0.039, "step": 9750 }, { "grad_norm": 0.24377849698066711, "learning_rate": 9.938208690554849e-05, "loss": 0.0357, "step": 9760 }, { "grad_norm": 0.4227578938007355, "learning_rate": 9.9379492748326e-05, "loss": 0.0389, "step": 9770 }, { "grad_norm": 0.44752630591392517, "learning_rate": 9.937689319103641e-05, "loss": 0.0422, "step": 9780 }, { "grad_norm": 0.30197545886039734, "learning_rate": 9.937428823396404e-05, "loss": 0.0405, "step": 9790 }, { "grad_norm": 0.2608872056007385, "learning_rate": 9.937167787739372e-05, "loss": 0.0334, "step": 9800 }, { "grad_norm": 0.2930588722229004, "learning_rate": 9.936906212161095e-05, "loss": 0.0339, "step": 9810 }, { "grad_norm": 0.2522111237049103, "learning_rate": 9.936644096690176e-05, "loss": 0.0316, "step": 9820 }, { "grad_norm": 0.20018541812896729, "learning_rate": 9.936381441355282e-05, "loss": 0.0347, "step": 9830 }, { "grad_norm": 0.20698875188827515, "learning_rate": 9.936118246185136e-05, "loss": 0.0327, "step": 9840 }, { "grad_norm": 0.3684735894203186, "learning_rate": 9.935854511208518e-05, "loss": 0.0342, "step": 9850 }, { "grad_norm": 0.5849128365516663, "learning_rate": 9.935590236454272e-05, "loss": 0.0361, "step": 9860 }, { "grad_norm": 0.29204699397087097, "learning_rate": 9.935325421951298e-05, "loss": 0.0433, "step": 9870 }, { "grad_norm": 0.37010860443115234, "learning_rate": 9.935060067728557e-05, "loss": 0.0373, "step": 9880 }, { "grad_norm": 0.3323727548122406, "learning_rate": 9.934794173815067e-05, "loss": 0.0375, "step": 9890 }, { "grad_norm": 0.32033973932266235, "learning_rate": 9.934527740239906e-05, "loss": 0.0375, "step": 9900 }, { "grad_norm": 0.3366210460662842, "learning_rate": 9.934260767032209e-05, "loss": 0.0354, "step": 9910 }, { "grad_norm": 0.3304809033870697, "learning_rate": 9.933993254221172e-05, "loss": 0.0375, "step": 9920 }, { "grad_norm": 0.33461102843284607, "learning_rate": 9.933725201836053e-05, "loss": 0.0358, "step": 9930 }, { "grad_norm": 0.2990121841430664, "learning_rate": 9.933456609906162e-05, "loss": 0.0343, "step": 9940 }, { "grad_norm": 0.3271486461162567, "learning_rate": 9.933187478460875e-05, "loss": 0.0356, "step": 9950 }, { "grad_norm": 0.3160102367401123, "learning_rate": 9.93291780752962e-05, "loss": 0.0321, "step": 9960 }, { "grad_norm": 0.29491350054740906, "learning_rate": 9.932647597141893e-05, "loss": 0.0334, "step": 9970 }, { "grad_norm": 0.28795820474624634, "learning_rate": 9.932376847327239e-05, "loss": 0.0343, "step": 9980 }, { "grad_norm": 0.2074001580476761, "learning_rate": 9.932105558115268e-05, "loss": 0.0317, "step": 9990 }, { "grad_norm": 0.30386602878570557, "learning_rate": 9.931833729535651e-05, "loss": 0.0326, "step": 10000 }, { "grad_norm": 0.3496251106262207, "learning_rate": 9.931561361618111e-05, "loss": 0.0325, "step": 10010 }, { "grad_norm": 0.31292054057121277, "learning_rate": 9.931288454392435e-05, "loss": 0.0338, "step": 10020 }, { "grad_norm": 0.3022579848766327, "learning_rate": 9.931015007888467e-05, "loss": 0.0355, "step": 10030 }, { "grad_norm": 0.2947157323360443, "learning_rate": 9.930741022136112e-05, "loss": 0.0347, "step": 10040 }, { "grad_norm": 0.3113752007484436, "learning_rate": 9.930466497165333e-05, "loss": 0.037, "step": 10050 }, { "grad_norm": 0.30652427673339844, "learning_rate": 9.93019143300615e-05, "loss": 0.0343, "step": 10060 }, { "grad_norm": 0.4763837456703186, "learning_rate": 9.929915829688644e-05, "loss": 0.0376, "step": 10070 }, { "grad_norm": 0.269095778465271, "learning_rate": 9.929639687242955e-05, "loss": 0.0425, "step": 10080 }, { "grad_norm": 0.2721669375896454, "learning_rate": 9.929363005699281e-05, "loss": 0.0408, "step": 10090 }, { "grad_norm": 0.3364569842815399, "learning_rate": 9.92908578508788e-05, "loss": 0.0399, "step": 10100 }, { "grad_norm": 0.346690833568573, "learning_rate": 9.928808025439069e-05, "loss": 0.0385, "step": 10110 }, { "grad_norm": 0.38601720333099365, "learning_rate": 9.928529726783223e-05, "loss": 0.0365, "step": 10120 }, { "grad_norm": 0.515877366065979, "learning_rate": 9.928250889150774e-05, "loss": 0.037, "step": 10130 }, { "grad_norm": 0.5292946696281433, "learning_rate": 9.92797151257222e-05, "loss": 0.0364, "step": 10140 }, { "grad_norm": 0.2491820901632309, "learning_rate": 9.927691597078108e-05, "loss": 0.0338, "step": 10150 }, { "grad_norm": 0.22924473881721497, "learning_rate": 9.927411142699053e-05, "loss": 0.0339, "step": 10160 }, { "grad_norm": 0.5223578810691833, "learning_rate": 9.927130149465725e-05, "loss": 0.0413, "step": 10170 }, { "grad_norm": 0.33648326992988586, "learning_rate": 9.92684861740885e-05, "loss": 0.0361, "step": 10180 }, { "grad_norm": 0.3177519142627716, "learning_rate": 9.926566546559217e-05, "loss": 0.0348, "step": 10190 }, { "grad_norm": 0.3657688796520233, "learning_rate": 9.926283936947673e-05, "loss": 0.0337, "step": 10200 }, { "grad_norm": 0.4064166843891144, "learning_rate": 9.926000788605126e-05, "loss": 0.0386, "step": 10210 }, { "grad_norm": 0.33968424797058105, "learning_rate": 9.92571710156254e-05, "loss": 0.039, "step": 10220 }, { "grad_norm": 0.4423878788948059, "learning_rate": 9.925432875850936e-05, "loss": 0.0356, "step": 10230 }, { "grad_norm": 0.3151157796382904, "learning_rate": 9.925148111501396e-05, "loss": 0.0364, "step": 10240 }, { "grad_norm": 0.38225051760673523, "learning_rate": 9.924862808545066e-05, "loss": 0.0373, "step": 10250 }, { "grad_norm": 0.3345944583415985, "learning_rate": 9.924576967013141e-05, "loss": 0.0417, "step": 10260 }, { "grad_norm": 0.32322874665260315, "learning_rate": 9.924290586936887e-05, "loss": 0.0368, "step": 10270 }, { "grad_norm": 0.3218047320842743, "learning_rate": 9.924003668347614e-05, "loss": 0.0369, "step": 10280 }, { "grad_norm": 0.28868597745895386, "learning_rate": 9.923716211276704e-05, "loss": 0.0354, "step": 10290 }, { "grad_norm": 0.3380086421966553, "learning_rate": 9.923428215755594e-05, "loss": 0.0347, "step": 10300 }, { "grad_norm": 0.28165170550346375, "learning_rate": 9.923139681815775e-05, "loss": 0.0349, "step": 10310 }, { "grad_norm": 0.25029340386390686, "learning_rate": 9.922850609488801e-05, "loss": 0.0323, "step": 10320 }, { "grad_norm": 0.3717927932739258, "learning_rate": 9.922560998806287e-05, "loss": 0.0375, "step": 10330 }, { "grad_norm": 0.33960792422294617, "learning_rate": 9.922270849799905e-05, "loss": 0.0392, "step": 10340 }, { "grad_norm": 0.31057432293891907, "learning_rate": 9.92198016250138e-05, "loss": 0.0394, "step": 10350 }, { "grad_norm": 0.24672165513038635, "learning_rate": 9.921688936942506e-05, "loss": 0.0365, "step": 10360 }, { "grad_norm": 0.28437545895576477, "learning_rate": 9.921397173155129e-05, "loss": 0.0377, "step": 10370 }, { "grad_norm": 0.23237629234790802, "learning_rate": 9.921104871171157e-05, "loss": 0.0349, "step": 10380 }, { "grad_norm": 0.2519855797290802, "learning_rate": 9.920812031022554e-05, "loss": 0.0372, "step": 10390 }, { "grad_norm": 0.3281717300415039, "learning_rate": 9.920518652741348e-05, "loss": 0.0409, "step": 10400 }, { "grad_norm": 0.293912410736084, "learning_rate": 9.920224736359618e-05, "loss": 0.0352, "step": 10410 }, { "grad_norm": 0.34688568115234375, "learning_rate": 9.91993028190951e-05, "loss": 0.0343, "step": 10420 }, { "grad_norm": 0.24481041729450226, "learning_rate": 9.919635289423222e-05, "loss": 0.0307, "step": 10430 }, { "grad_norm": 0.30443790555000305, "learning_rate": 9.919339758933015e-05, "loss": 0.033, "step": 10440 }, { "grad_norm": 0.4192357063293457, "learning_rate": 9.919043690471209e-05, "loss": 0.0328, "step": 10450 }, { "grad_norm": 0.2800079584121704, "learning_rate": 9.91874708407018e-05, "loss": 0.0328, "step": 10460 }, { "grad_norm": 0.23265165090560913, "learning_rate": 9.918449939762367e-05, "loss": 0.0315, "step": 10470 }, { "grad_norm": 0.2888496220111847, "learning_rate": 9.91815225758026e-05, "loss": 0.0337, "step": 10480 }, { "grad_norm": 0.3217746317386627, "learning_rate": 9.917854037556419e-05, "loss": 0.0375, "step": 10490 }, { "grad_norm": 0.2637491822242737, "learning_rate": 9.917555279723454e-05, "loss": 0.0395, "step": 10500 }, { "grad_norm": 0.2951022684574127, "learning_rate": 9.917255984114036e-05, "loss": 0.0315, "step": 10510 }, { "grad_norm": 0.3096931278705597, "learning_rate": 9.916956150760896e-05, "loss": 0.034, "step": 10520 }, { "grad_norm": 0.27218419313430786, "learning_rate": 9.916655779696826e-05, "loss": 0.0367, "step": 10530 }, { "grad_norm": 0.2672524154186249, "learning_rate": 9.916354870954671e-05, "loss": 0.0359, "step": 10540 }, { "grad_norm": 0.23815786838531494, "learning_rate": 9.91605342456734e-05, "loss": 0.0335, "step": 10550 }, { "grad_norm": 0.3111903965473175, "learning_rate": 9.915751440567795e-05, "loss": 0.0353, "step": 10560 }, { "grad_norm": 0.2485356479883194, "learning_rate": 9.915448918989066e-05, "loss": 0.033, "step": 10570 }, { "grad_norm": 0.22796525061130524, "learning_rate": 9.915145859864232e-05, "loss": 0.032, "step": 10580 }, { "grad_norm": 0.25789862871170044, "learning_rate": 9.914842263226437e-05, "loss": 0.0306, "step": 10590 }, { "grad_norm": 0.220188707113266, "learning_rate": 9.914538129108882e-05, "loss": 0.0288, "step": 10600 }, { "grad_norm": 0.321157306432724, "learning_rate": 9.914233457544825e-05, "loss": 0.03, "step": 10610 }, { "grad_norm": 0.3170474171638489, "learning_rate": 9.913928248567586e-05, "loss": 0.0348, "step": 10620 }, { "grad_norm": 0.32093560695648193, "learning_rate": 9.913622502210542e-05, "loss": 0.0336, "step": 10630 }, { "grad_norm": 0.31993329524993896, "learning_rate": 9.913316218507128e-05, "loss": 0.0367, "step": 10640 }, { "grad_norm": 0.28486141562461853, "learning_rate": 9.91300939749084e-05, "loss": 0.0347, "step": 10650 }, { "grad_norm": 0.2019234448671341, "learning_rate": 9.91270203919523e-05, "loss": 0.0288, "step": 10660 }, { "grad_norm": 0.4828374683856964, "learning_rate": 9.912394143653912e-05, "loss": 0.0337, "step": 10670 }, { "grad_norm": 0.3158879578113556, "learning_rate": 9.912085710900555e-05, "loss": 0.0364, "step": 10680 }, { "grad_norm": 0.2676694989204407, "learning_rate": 9.911776740968892e-05, "loss": 0.0349, "step": 10690 }, { "grad_norm": 0.28877130150794983, "learning_rate": 9.911467233892709e-05, "loss": 0.0308, "step": 10700 }, { "grad_norm": 0.2983270585536957, "learning_rate": 9.911157189705853e-05, "loss": 0.0327, "step": 10710 }, { "grad_norm": 0.25957635045051575, "learning_rate": 9.910846608442229e-05, "loss": 0.0333, "step": 10720 }, { "grad_norm": 0.3325898051261902, "learning_rate": 9.910535490135805e-05, "loss": 0.0329, "step": 10730 }, { "grad_norm": 0.2724117934703827, "learning_rate": 9.910223834820603e-05, "loss": 0.0336, "step": 10740 }, { "grad_norm": 0.26410529017448425, "learning_rate": 9.909911642530703e-05, "loss": 0.0316, "step": 10750 }, { "grad_norm": 0.2370099276304245, "learning_rate": 9.909598913300249e-05, "loss": 0.0312, "step": 10760 }, { "grad_norm": 0.2247948795557022, "learning_rate": 9.909285647163438e-05, "loss": 0.0289, "step": 10770 }, { "grad_norm": 0.25475481152534485, "learning_rate": 9.908971844154531e-05, "loss": 0.0277, "step": 10780 }, { "grad_norm": 0.20326916873455048, "learning_rate": 9.908657504307843e-05, "loss": 0.0299, "step": 10790 }, { "grad_norm": 0.286348819732666, "learning_rate": 9.908342627657751e-05, "loss": 0.0333, "step": 10800 }, { "grad_norm": 0.2638295292854309, "learning_rate": 9.908027214238689e-05, "loss": 0.0331, "step": 10810 }, { "grad_norm": 0.26968714594841003, "learning_rate": 9.90771126408515e-05, "loss": 0.0316, "step": 10820 }, { "grad_norm": 0.29365912079811096, "learning_rate": 9.907394777231685e-05, "loss": 0.0304, "step": 10830 }, { "grad_norm": 0.3016856610774994, "learning_rate": 9.907077753712905e-05, "loss": 0.0305, "step": 10840 }, { "grad_norm": 0.34789589047431946, "learning_rate": 9.906760193563482e-05, "loss": 0.0328, "step": 10850 }, { "grad_norm": 0.4131518006324768, "learning_rate": 9.906442096818139e-05, "loss": 0.0311, "step": 10860 }, { "grad_norm": 0.2644939124584198, "learning_rate": 9.906123463511665e-05, "loss": 0.0291, "step": 10870 }, { "grad_norm": 0.3533836901187897, "learning_rate": 9.905804293678907e-05, "loss": 0.0315, "step": 10880 }, { "grad_norm": 0.4275108277797699, "learning_rate": 9.905484587354766e-05, "loss": 0.0325, "step": 10890 }, { "grad_norm": 0.2936553359031677, "learning_rate": 9.905164344574205e-05, "loss": 0.034, "step": 10900 }, { "grad_norm": 0.26199930906295776, "learning_rate": 9.904843565372248e-05, "loss": 0.0321, "step": 10910 }, { "grad_norm": 0.27541688084602356, "learning_rate": 9.904522249783972e-05, "loss": 0.0342, "step": 10920 }, { "grad_norm": 0.2065402865409851, "learning_rate": 9.904200397844517e-05, "loss": 0.0314, "step": 10930 }, { "grad_norm": 0.3028777241706848, "learning_rate": 9.903878009589078e-05, "loss": 0.032, "step": 10940 }, { "grad_norm": 0.26472461223602295, "learning_rate": 9.903555085052915e-05, "loss": 0.0303, "step": 10950 }, { "grad_norm": 0.23401661217212677, "learning_rate": 9.903231624271338e-05, "loss": 0.0296, "step": 10960 }, { "grad_norm": 0.22681327164173126, "learning_rate": 9.902907627279724e-05, "loss": 0.0332, "step": 10970 }, { "grad_norm": 0.23116107285022736, "learning_rate": 9.902583094113504e-05, "loss": 0.0308, "step": 10980 }, { "grad_norm": 0.2188740074634552, "learning_rate": 9.902258024808168e-05, "loss": 0.0305, "step": 10990 }, { "grad_norm": 0.29408103227615356, "learning_rate": 9.901932419399264e-05, "loss": 0.0317, "step": 11000 }, { "grad_norm": 0.2688520848751068, "learning_rate": 9.9016062779224e-05, "loss": 0.0356, "step": 11010 }, { "grad_norm": 0.30383455753326416, "learning_rate": 9.901279600413242e-05, "loss": 0.0335, "step": 11020 }, { "grad_norm": 0.27939942479133606, "learning_rate": 9.900952386907518e-05, "loss": 0.0318, "step": 11030 }, { "grad_norm": 0.30425211787223816, "learning_rate": 9.90062463744101e-05, "loss": 0.0305, "step": 11040 }, { "grad_norm": 0.3119353652000427, "learning_rate": 9.900296352049558e-05, "loss": 0.0319, "step": 11050 }, { "grad_norm": 0.2687254846096039, "learning_rate": 9.899967530769065e-05, "loss": 0.032, "step": 11060 }, { "grad_norm": 0.3001071810722351, "learning_rate": 9.899638173635489e-05, "loss": 0.0263, "step": 11070 }, { "grad_norm": 0.41645970940589905, "learning_rate": 9.899308280684849e-05, "loss": 0.0329, "step": 11080 }, { "grad_norm": 0.2868386209011078, "learning_rate": 9.898977851953222e-05, "loss": 0.0313, "step": 11090 }, { "grad_norm": 0.250701367855072, "learning_rate": 9.898646887476741e-05, "loss": 0.0313, "step": 11100 }, { "grad_norm": 0.3529292941093445, "learning_rate": 9.898315387291603e-05, "loss": 0.0293, "step": 11110 }, { "grad_norm": 0.2930152118206024, "learning_rate": 9.89798335143406e-05, "loss": 0.0299, "step": 11120 }, { "grad_norm": 0.3340801000595093, "learning_rate": 9.897650779940419e-05, "loss": 0.0333, "step": 11130 }, { "grad_norm": 0.2663291096687317, "learning_rate": 9.897317672847054e-05, "loss": 0.0292, "step": 11140 }, { "grad_norm": 0.23033404350280762, "learning_rate": 9.89698403019039e-05, "loss": 0.0269, "step": 11150 }, { "grad_norm": 0.3343072235584259, "learning_rate": 9.896649852006917e-05, "loss": 0.0313, "step": 11160 }, { "grad_norm": 0.32473161816596985, "learning_rate": 9.896315138333177e-05, "loss": 0.0314, "step": 11170 }, { "grad_norm": 0.3066188395023346, "learning_rate": 9.895979889205774e-05, "loss": 0.0279, "step": 11180 }, { "grad_norm": 0.3220541179180145, "learning_rate": 9.895644104661372e-05, "loss": 0.0325, "step": 11190 }, { "grad_norm": 0.41424083709716797, "learning_rate": 9.895307784736691e-05, "loss": 0.0326, "step": 11200 }, { "grad_norm": 0.248422771692276, "learning_rate": 9.894970929468512e-05, "loss": 0.0316, "step": 11210 }, { "grad_norm": 0.25167563557624817, "learning_rate": 9.89463353889367e-05, "loss": 0.0278, "step": 11220 }, { "grad_norm": 0.20476922392845154, "learning_rate": 9.894295613049065e-05, "loss": 0.0361, "step": 11230 }, { "grad_norm": 0.2471615970134735, "learning_rate": 9.893957151971649e-05, "loss": 0.0307, "step": 11240 }, { "grad_norm": 0.2505338490009308, "learning_rate": 9.893618155698436e-05, "loss": 0.032, "step": 11250 }, { "grad_norm": 0.22278772294521332, "learning_rate": 9.8932786242665e-05, "loss": 0.032, "step": 11260 }, { "grad_norm": 0.26105818152427673, "learning_rate": 9.89293855771297e-05, "loss": 0.0303, "step": 11270 }, { "grad_norm": 0.23802784085273743, "learning_rate": 9.892597956075036e-05, "loss": 0.0271, "step": 11280 }, { "grad_norm": 0.34186241030693054, "learning_rate": 9.892256819389947e-05, "loss": 0.027, "step": 11290 }, { "grad_norm": 0.323870450258255, "learning_rate": 9.891915147695006e-05, "loss": 0.0284, "step": 11300 }, { "grad_norm": 0.2522684931755066, "learning_rate": 9.891572941027577e-05, "loss": 0.0312, "step": 11310 }, { "grad_norm": 0.3466262221336365, "learning_rate": 9.89123019942509e-05, "loss": 0.029, "step": 11320 }, { "grad_norm": 0.3201492428779602, "learning_rate": 9.89088692292502e-05, "loss": 0.029, "step": 11330 }, { "grad_norm": 0.27645862102508545, "learning_rate": 9.89054311156491e-05, "loss": 0.0333, "step": 11340 }, { "grad_norm": 0.2330683022737503, "learning_rate": 9.890198765382357e-05, "loss": 0.0319, "step": 11350 }, { "grad_norm": 0.2822727560997009, "learning_rate": 9.889853884415021e-05, "loss": 0.0329, "step": 11360 }, { "grad_norm": 0.38239356875419617, "learning_rate": 9.889508468700614e-05, "loss": 0.0326, "step": 11370 }, { "grad_norm": 0.31259867548942566, "learning_rate": 9.889162518276915e-05, "loss": 0.0318, "step": 11380 }, { "grad_norm": 0.25812312960624695, "learning_rate": 9.888816033181752e-05, "loss": 0.0297, "step": 11390 }, { "grad_norm": 0.37295907735824585, "learning_rate": 9.888469013453018e-05, "loss": 0.0328, "step": 11400 }, { "grad_norm": 0.25666487216949463, "learning_rate": 9.888121459128663e-05, "loss": 0.0317, "step": 11410 }, { "grad_norm": 0.28390827775001526, "learning_rate": 9.887773370246693e-05, "loss": 0.0297, "step": 11420 }, { "grad_norm": 0.26649710536003113, "learning_rate": 9.887424746845177e-05, "loss": 0.0297, "step": 11430 }, { "grad_norm": 0.21174398064613342, "learning_rate": 9.887075588962239e-05, "loss": 0.0321, "step": 11440 }, { "grad_norm": 0.28881406784057617, "learning_rate": 9.88672589663606e-05, "loss": 0.0321, "step": 11450 }, { "grad_norm": 0.24677252769470215, "learning_rate": 9.886375669904886e-05, "loss": 0.0318, "step": 11460 }, { "grad_norm": 0.32634687423706055, "learning_rate": 9.886024908807014e-05, "loss": 0.0314, "step": 11470 }, { "grad_norm": 0.27678146958351135, "learning_rate": 9.885673613380806e-05, "loss": 0.0316, "step": 11480 }, { "grad_norm": 0.24722537398338318, "learning_rate": 9.885321783664676e-05, "loss": 0.0293, "step": 11490 }, { "grad_norm": 0.3052874803543091, "learning_rate": 9.884969419697101e-05, "loss": 0.0299, "step": 11500 }, { "grad_norm": 0.332055926322937, "learning_rate": 9.884616521516614e-05, "loss": 0.0344, "step": 11510 }, { "grad_norm": 0.2684774100780487, "learning_rate": 9.88426308916181e-05, "loss": 0.0395, "step": 11520 }, { "grad_norm": 0.2457578331232071, "learning_rate": 9.883909122671335e-05, "loss": 0.0378, "step": 11530 }, { "grad_norm": 0.18144917488098145, "learning_rate": 9.883554622083904e-05, "loss": 0.0368, "step": 11540 }, { "grad_norm": 0.32356634736061096, "learning_rate": 9.88319958743828e-05, "loss": 0.0358, "step": 11550 }, { "grad_norm": 0.3818190395832062, "learning_rate": 9.882844018773291e-05, "loss": 0.0355, "step": 11560 }, { "grad_norm": 0.31049081683158875, "learning_rate": 9.882487916127823e-05, "loss": 0.0338, "step": 11570 }, { "grad_norm": 0.2709570527076721, "learning_rate": 9.882131279540815e-05, "loss": 0.0319, "step": 11580 }, { "grad_norm": 0.26593878865242004, "learning_rate": 9.881774109051271e-05, "loss": 0.0298, "step": 11590 }, { "grad_norm": 0.4379030764102936, "learning_rate": 9.881416404698252e-05, "loss": 0.03, "step": 11600 }, { "grad_norm": 0.31694716215133667, "learning_rate": 9.881058166520873e-05, "loss": 0.0354, "step": 11610 }, { "grad_norm": 0.299419641494751, "learning_rate": 9.880699394558311e-05, "loss": 0.0405, "step": 11620 }, { "grad_norm": 0.32320600748062134, "learning_rate": 9.880340088849801e-05, "loss": 0.0369, "step": 11630 }, { "grad_norm": 0.2606182396411896, "learning_rate": 9.879980249434637e-05, "loss": 0.0335, "step": 11640 }, { "grad_norm": 0.2538188695907593, "learning_rate": 9.879619876352168e-05, "loss": 0.0323, "step": 11650 }, { "grad_norm": 0.2914847135543823, "learning_rate": 9.879258969641809e-05, "loss": 0.0297, "step": 11660 }, { "grad_norm": 0.27977457642555237, "learning_rate": 9.878897529343023e-05, "loss": 0.0309, "step": 11670 }, { "grad_norm": 0.25747808814048767, "learning_rate": 9.878535555495338e-05, "loss": 0.0297, "step": 11680 }, { "grad_norm": 0.21670448780059814, "learning_rate": 9.87817304813834e-05, "loss": 0.0305, "step": 11690 }, { "grad_norm": 0.22538281977176666, "learning_rate": 9.877810007311671e-05, "loss": 0.034, "step": 11700 }, { "grad_norm": 0.23823927342891693, "learning_rate": 9.877446433055035e-05, "loss": 0.029, "step": 11710 }, { "grad_norm": 0.2970329523086548, "learning_rate": 9.877082325408191e-05, "loss": 0.0298, "step": 11720 }, { "grad_norm": 0.25848302245140076, "learning_rate": 9.876717684410954e-05, "loss": 0.0298, "step": 11730 }, { "grad_norm": 0.2403530478477478, "learning_rate": 9.876352510103204e-05, "loss": 0.0302, "step": 11740 }, { "grad_norm": 0.25486406683921814, "learning_rate": 9.875986802524875e-05, "loss": 0.0295, "step": 11750 }, { "grad_norm": 0.24736100435256958, "learning_rate": 9.87562056171596e-05, "loss": 0.0294, "step": 11760 }, { "grad_norm": 0.31311824917793274, "learning_rate": 9.875253787716511e-05, "loss": 0.0286, "step": 11770 }, { "grad_norm": 0.3226577341556549, "learning_rate": 9.874886480566637e-05, "loss": 0.0292, "step": 11780 }, { "grad_norm": 0.3114382028579712, "learning_rate": 9.874518640306507e-05, "loss": 0.0365, "step": 11790 }, { "grad_norm": 0.26887041330337524, "learning_rate": 9.874150266976347e-05, "loss": 0.0338, "step": 11800 }, { "grad_norm": 0.30415070056915283, "learning_rate": 9.873781360616443e-05, "loss": 0.0343, "step": 11810 }, { "grad_norm": 0.2935875654220581, "learning_rate": 9.873411921267137e-05, "loss": 0.0361, "step": 11820 }, { "grad_norm": 0.23091617226600647, "learning_rate": 9.873041948968829e-05, "loss": 0.0327, "step": 11830 }, { "grad_norm": 0.33096012473106384, "learning_rate": 9.872671443761981e-05, "loss": 0.0341, "step": 11840 }, { "grad_norm": 0.26060235500335693, "learning_rate": 9.872300405687109e-05, "loss": 0.0295, "step": 11850 }, { "grad_norm": 0.27902111411094666, "learning_rate": 9.871928834784792e-05, "loss": 0.029, "step": 11860 }, { "grad_norm": 0.2656967043876648, "learning_rate": 9.871556731095661e-05, "loss": 0.0328, "step": 11870 }, { "grad_norm": 0.2607432007789612, "learning_rate": 9.871184094660411e-05, "loss": 0.0297, "step": 11880 }, { "grad_norm": 0.2438676655292511, "learning_rate": 9.870810925519791e-05, "loss": 0.0305, "step": 11890 }, { "grad_norm": 0.2414432168006897, "learning_rate": 9.870437223714612e-05, "loss": 0.0288, "step": 11900 }, { "grad_norm": 0.2457628697156906, "learning_rate": 9.87006298928574e-05, "loss": 0.0316, "step": 11910 }, { "grad_norm": 0.25315025448799133, "learning_rate": 9.869688222274103e-05, "loss": 0.0325, "step": 11920 }, { "grad_norm": 0.264393150806427, "learning_rate": 9.869312922720681e-05, "loss": 0.0341, "step": 11930 }, { "grad_norm": 0.33067765831947327, "learning_rate": 9.868937090666521e-05, "loss": 0.0364, "step": 11940 }, { "grad_norm": 0.2896922528743744, "learning_rate": 9.86856072615272e-05, "loss": 0.0297, "step": 11950 }, { "grad_norm": 0.3400804102420807, "learning_rate": 9.868183829220438e-05, "loss": 0.0342, "step": 11960 }, { "grad_norm": 0.24958032369613647, "learning_rate": 9.867806399910893e-05, "loss": 0.0299, "step": 11970 }, { "grad_norm": 0.23846010863780975, "learning_rate": 9.867428438265356e-05, "loss": 0.0316, "step": 11980 }, { "grad_norm": 0.29436472058296204, "learning_rate": 9.867049944325165e-05, "loss": 0.0294, "step": 11990 }, { "grad_norm": 0.3293205797672272, "learning_rate": 9.86667091813171e-05, "loss": 0.0358, "step": 12000 }, { "grad_norm": 0.29897990822792053, "learning_rate": 9.866291359726438e-05, "loss": 0.0338, "step": 12010 }, { "grad_norm": 0.3164058327674866, "learning_rate": 9.865911269150861e-05, "loss": 0.0334, "step": 12020 }, { "grad_norm": 0.30533385276794434, "learning_rate": 9.865530646446544e-05, "loss": 0.0337, "step": 12030 }, { "grad_norm": 0.2540026605129242, "learning_rate": 9.86514949165511e-05, "loss": 0.033, "step": 12040 }, { "grad_norm": 0.27885428071022034, "learning_rate": 9.864767804818243e-05, "loss": 0.0319, "step": 12050 }, { "grad_norm": 0.2755364179611206, "learning_rate": 9.86438558597768e-05, "loss": 0.0319, "step": 12060 }, { "grad_norm": 0.28007280826568604, "learning_rate": 9.864002835175225e-05, "loss": 0.0303, "step": 12070 }, { "grad_norm": 0.24029065668582916, "learning_rate": 9.863619552452734e-05, "loss": 0.0349, "step": 12080 }, { "grad_norm": 0.2693674862384796, "learning_rate": 9.863235737852119e-05, "loss": 0.0302, "step": 12090 }, { "grad_norm": 0.280762642621994, "learning_rate": 9.862851391415356e-05, "loss": 0.033, "step": 12100 }, { "grad_norm": 0.33224305510520935, "learning_rate": 9.862466513184477e-05, "loss": 0.0295, "step": 12110 }, { "grad_norm": 0.3164673149585724, "learning_rate": 9.86208110320157e-05, "loss": 0.0328, "step": 12120 }, { "grad_norm": 0.31927335262298584, "learning_rate": 9.861695161508784e-05, "loss": 0.0316, "step": 12130 }, { "grad_norm": 0.6258080005645752, "learning_rate": 9.861308688148324e-05, "loss": 0.0319, "step": 12140 }, { "grad_norm": 0.25822246074676514, "learning_rate": 9.860921683162455e-05, "loss": 0.0325, "step": 12150 }, { "grad_norm": 0.27213147282600403, "learning_rate": 9.860534146593499e-05, "loss": 0.0328, "step": 12160 }, { "grad_norm": 0.3196735978126526, "learning_rate": 9.860146078483836e-05, "loss": 0.035, "step": 12170 }, { "grad_norm": 0.2462085783481598, "learning_rate": 9.859757478875905e-05, "loss": 0.033, "step": 12180 }, { "grad_norm": 0.29175594449043274, "learning_rate": 9.859368347812204e-05, "loss": 0.0315, "step": 12190 }, { "grad_norm": 0.2993980050086975, "learning_rate": 9.858978685335285e-05, "loss": 0.0343, "step": 12200 }, { "grad_norm": 0.27669405937194824, "learning_rate": 9.858588491487763e-05, "loss": 0.0332, "step": 12210 }, { "grad_norm": 0.20525409281253815, "learning_rate": 9.858197766312308e-05, "loss": 0.0305, "step": 12220 }, { "grad_norm": 0.34698331356048584, "learning_rate": 9.857806509851649e-05, "loss": 0.0322, "step": 12230 }, { "grad_norm": 0.23810385167598724, "learning_rate": 9.857414722148574e-05, "loss": 0.0335, "step": 12240 }, { "grad_norm": 0.2726283371448517, "learning_rate": 9.857022403245928e-05, "loss": 0.0289, "step": 12250 }, { "grad_norm": 0.25630736351013184, "learning_rate": 9.856629553186615e-05, "loss": 0.0298, "step": 12260 }, { "grad_norm": 0.2010577768087387, "learning_rate": 9.856236172013595e-05, "loss": 0.033, "step": 12270 }, { "grad_norm": 0.2661651074886322, "learning_rate": 9.85584225976989e-05, "loss": 0.0262, "step": 12280 }, { "grad_norm": 0.2828258275985718, "learning_rate": 9.855447816498575e-05, "loss": 0.0261, "step": 12290 }, { "grad_norm": 0.29815995693206787, "learning_rate": 9.855052842242787e-05, "loss": 0.0323, "step": 12300 }, { "grad_norm": 0.327409029006958, "learning_rate": 9.85465733704572e-05, "loss": 0.0308, "step": 12310 }, { "grad_norm": 0.352463960647583, "learning_rate": 9.854261300950624e-05, "loss": 0.0315, "step": 12320 }, { "grad_norm": 0.2436172366142273, "learning_rate": 9.853864734000813e-05, "loss": 0.0307, "step": 12330 }, { "grad_norm": 0.1889030486345291, "learning_rate": 9.85346763623965e-05, "loss": 0.0315, "step": 12340 }, { "grad_norm": 0.29687970876693726, "learning_rate": 9.853070007710564e-05, "loss": 0.032, "step": 12350 }, { "grad_norm": 0.253577321767807, "learning_rate": 9.85267184845704e-05, "loss": 0.0303, "step": 12360 }, { "grad_norm": 0.3075958490371704, "learning_rate": 9.852273158522616e-05, "loss": 0.0304, "step": 12370 }, { "grad_norm": 0.29280850291252136, "learning_rate": 9.851873937950896e-05, "loss": 0.0281, "step": 12380 }, { "grad_norm": 0.30875808000564575, "learning_rate": 9.851474186785537e-05, "loss": 0.0281, "step": 12390 }, { "grad_norm": 0.24993829429149628, "learning_rate": 9.851073905070254e-05, "loss": 0.0308, "step": 12400 }, { "grad_norm": 0.3390289545059204, "learning_rate": 9.850673092848824e-05, "loss": 0.0272, "step": 12410 }, { "grad_norm": 0.2608056366443634, "learning_rate": 9.850271750165077e-05, "loss": 0.0278, "step": 12420 }, { "grad_norm": 0.28345149755477905, "learning_rate": 9.849869877062902e-05, "loss": 0.0306, "step": 12430 }, { "grad_norm": 0.28367021679878235, "learning_rate": 9.849467473586252e-05, "loss": 0.0262, "step": 12440 }, { "grad_norm": 0.30639001727104187, "learning_rate": 9.849064539779127e-05, "loss": 0.0309, "step": 12450 }, { "grad_norm": 0.2922918498516083, "learning_rate": 9.848661075685594e-05, "loss": 0.0331, "step": 12460 }, { "grad_norm": 0.30370137095451355, "learning_rate": 9.848257081349778e-05, "loss": 0.0367, "step": 12470 }, { "grad_norm": 0.25859203934669495, "learning_rate": 9.847852556815856e-05, "loss": 0.0299, "step": 12480 }, { "grad_norm": 0.20588599145412445, "learning_rate": 9.847447502128067e-05, "loss": 0.0332, "step": 12490 }, { "grad_norm": 0.29991424083709717, "learning_rate": 9.847041917330708e-05, "loss": 0.0308, "step": 12500 }, { "grad_norm": 0.2604827582836151, "learning_rate": 9.846635802468132e-05, "loss": 0.0322, "step": 12510 }, { "grad_norm": 0.15874840319156647, "learning_rate": 9.84622915758475e-05, "loss": 0.0269, "step": 12520 }, { "grad_norm": 0.3801437020301819, "learning_rate": 9.845821982725034e-05, "loss": 0.0318, "step": 12530 }, { "grad_norm": 0.4411872923374176, "learning_rate": 9.845414277933514e-05, "loss": 0.0287, "step": 12540 }, { "grad_norm": 0.25526830554008484, "learning_rate": 9.845006043254771e-05, "loss": 0.0314, "step": 12550 }, { "grad_norm": 0.34599804878234863, "learning_rate": 9.844597278733451e-05, "loss": 0.0341, "step": 12560 }, { "grad_norm": 0.2606031000614166, "learning_rate": 9.844187984414259e-05, "loss": 0.0334, "step": 12570 }, { "grad_norm": 0.41043567657470703, "learning_rate": 9.84377816034195e-05, "loss": 0.0375, "step": 12580 }, { "grad_norm": 0.2575817406177521, "learning_rate": 9.843367806561345e-05, "loss": 0.0322, "step": 12590 }, { "grad_norm": 0.24298608303070068, "learning_rate": 9.842956923117317e-05, "loss": 0.0312, "step": 12600 }, { "grad_norm": 0.25791940093040466, "learning_rate": 9.842545510054802e-05, "loss": 0.0304, "step": 12610 }, { "grad_norm": 0.22011728584766388, "learning_rate": 9.842133567418792e-05, "loss": 0.0304, "step": 12620 }, { "grad_norm": 0.24153274297714233, "learning_rate": 9.841721095254333e-05, "loss": 0.031, "step": 12630 }, { "grad_norm": 0.2503680884838104, "learning_rate": 9.841308093606537e-05, "loss": 0.031, "step": 12640 }, { "grad_norm": 0.27490395307540894, "learning_rate": 9.840894562520565e-05, "loss": 0.0302, "step": 12650 }, { "grad_norm": 0.2157500982284546, "learning_rate": 9.840480502041642e-05, "loss": 0.033, "step": 12660 }, { "grad_norm": 0.362918496131897, "learning_rate": 9.840065912215049e-05, "loss": 0.033, "step": 12670 }, { "grad_norm": 0.25107142329216003, "learning_rate": 9.839650793086124e-05, "loss": 0.0293, "step": 12680 }, { "grad_norm": 0.2565593421459198, "learning_rate": 9.839235144700265e-05, "loss": 0.031, "step": 12690 }, { "grad_norm": 0.2755335569381714, "learning_rate": 9.838818967102926e-05, "loss": 0.0305, "step": 12700 }, { "grad_norm": 0.23030316829681396, "learning_rate": 9.83840226033962e-05, "loss": 0.0312, "step": 12710 }, { "grad_norm": 0.31065598130226135, "learning_rate": 9.837985024455918e-05, "loss": 0.0355, "step": 12720 }, { "grad_norm": 0.3443688750267029, "learning_rate": 9.837567259497447e-05, "loss": 0.034, "step": 12730 }, { "grad_norm": 0.2732866704463959, "learning_rate": 9.837148965509894e-05, "loss": 0.0307, "step": 12740 }, { "grad_norm": 0.2308664172887802, "learning_rate": 9.836730142539001e-05, "loss": 0.0307, "step": 12750 }, { "grad_norm": 0.24247901141643524, "learning_rate": 9.836310790630574e-05, "loss": 0.0324, "step": 12760 }, { "grad_norm": 0.24822449684143066, "learning_rate": 9.83589090983047e-05, "loss": 0.0326, "step": 12770 }, { "grad_norm": 0.2904506027698517, "learning_rate": 9.835470500184605e-05, "loss": 0.0302, "step": 12780 }, { "grad_norm": 0.3029264509677887, "learning_rate": 9.835049561738957e-05, "loss": 0.0337, "step": 12790 }, { "grad_norm": 0.31976228952407837, "learning_rate": 9.834628094539558e-05, "loss": 0.0311, "step": 12800 }, { "grad_norm": 0.27661511301994324, "learning_rate": 9.834206098632499e-05, "loss": 0.0285, "step": 12810 }, { "grad_norm": 0.33802783489227295, "learning_rate": 9.833783574063931e-05, "loss": 0.0311, "step": 12820 }, { "grad_norm": 0.2841523587703705, "learning_rate": 9.833360520880058e-05, "loss": 0.0287, "step": 12830 }, { "grad_norm": 0.26713237166404724, "learning_rate": 9.832936939127144e-05, "loss": 0.0286, "step": 12840 }, { "grad_norm": 0.2743699848651886, "learning_rate": 9.832512828851515e-05, "loss": 0.0291, "step": 12850 }, { "grad_norm": 0.21926574409008026, "learning_rate": 9.832088190099546e-05, "loss": 0.0282, "step": 12860 }, { "grad_norm": 0.24252301454544067, "learning_rate": 9.831663022917679e-05, "loss": 0.0304, "step": 12870 }, { "grad_norm": 0.3545571565628052, "learning_rate": 9.831237327352407e-05, "loss": 0.0285, "step": 12880 }, { "grad_norm": 0.25153762102127075, "learning_rate": 9.830811103450286e-05, "loss": 0.0326, "step": 12890 }, { "grad_norm": 0.35110315680503845, "learning_rate": 9.830384351257924e-05, "loss": 0.0286, "step": 12900 }, { "grad_norm": 0.25692468881607056, "learning_rate": 9.829957070821993e-05, "loss": 0.0265, "step": 12910 }, { "grad_norm": 0.2023288607597351, "learning_rate": 9.829529262189218e-05, "loss": 0.0308, "step": 12920 }, { "grad_norm": 0.27651751041412354, "learning_rate": 9.829100925406385e-05, "loss": 0.0346, "step": 12930 }, { "grad_norm": 0.26212576031684875, "learning_rate": 9.828672060520333e-05, "loss": 0.0302, "step": 12940 }, { "grad_norm": 0.2216440737247467, "learning_rate": 9.828242667577966e-05, "loss": 0.0324, "step": 12950 }, { "grad_norm": 0.2867185175418854, "learning_rate": 9.82781274662624e-05, "loss": 0.0286, "step": 12960 }, { "grad_norm": 0.21703961491584778, "learning_rate": 9.82738229771217e-05, "loss": 0.0294, "step": 12970 }, { "grad_norm": 0.1991436779499054, "learning_rate": 9.826951320882829e-05, "loss": 0.0275, "step": 12980 }, { "grad_norm": 0.23399218916893005, "learning_rate": 9.826519816185351e-05, "loss": 0.0264, "step": 12990 }, { "grad_norm": 0.30466488003730774, "learning_rate": 9.826087783666921e-05, "loss": 0.03, "step": 13000 }, { "grad_norm": 0.23817996680736542, "learning_rate": 9.825655223374787e-05, "loss": 0.0272, "step": 13010 }, { "grad_norm": 0.33046042919158936, "learning_rate": 9.825222135356253e-05, "loss": 0.0285, "step": 13020 }, { "grad_norm": 0.2838480770587921, "learning_rate": 9.82478851965868e-05, "loss": 0.0314, "step": 13030 }, { "grad_norm": 0.40331944823265076, "learning_rate": 9.82435437632949e-05, "loss": 0.0347, "step": 13040 }, { "grad_norm": 0.2548849880695343, "learning_rate": 9.823919705416158e-05, "loss": 0.0331, "step": 13050 }, { "grad_norm": 0.3286636471748352, "learning_rate": 9.82348450696622e-05, "loss": 0.0327, "step": 13060 }, { "grad_norm": 0.2439725250005722, "learning_rate": 9.823048781027268e-05, "loss": 0.0282, "step": 13070 }, { "grad_norm": 0.28563544154167175, "learning_rate": 9.822612527646953e-05, "loss": 0.0317, "step": 13080 }, { "grad_norm": 0.2979956567287445, "learning_rate": 9.822175746872984e-05, "loss": 0.032, "step": 13090 }, { "grad_norm": 0.37255826592445374, "learning_rate": 9.821738438753123e-05, "loss": 0.0297, "step": 13100 }, { "grad_norm": 0.27318790555000305, "learning_rate": 9.821300603335196e-05, "loss": 0.031, "step": 13110 }, { "grad_norm": 0.3284663259983063, "learning_rate": 9.820862240667085e-05, "loss": 0.0296, "step": 13120 }, { "grad_norm": 0.30182284116744995, "learning_rate": 9.820423350796726e-05, "loss": 0.0372, "step": 13130 }, { "grad_norm": 0.3393653333187103, "learning_rate": 9.819983933772118e-05, "loss": 0.0342, "step": 13140 }, { "grad_norm": 0.2826792299747467, "learning_rate": 9.819543989641314e-05, "loss": 0.0354, "step": 13150 }, { "grad_norm": 0.2552085220813751, "learning_rate": 9.819103518452423e-05, "loss": 0.0338, "step": 13160 }, { "grad_norm": 0.4268893003463745, "learning_rate": 9.818662520253618e-05, "loss": 0.033, "step": 13170 }, { "grad_norm": 0.3651731610298157, "learning_rate": 9.818220995093126e-05, "loss": 0.0362, "step": 13180 }, { "grad_norm": 0.32722294330596924, "learning_rate": 9.817778943019228e-05, "loss": 0.0361, "step": 13190 }, { "grad_norm": 0.46189454197883606, "learning_rate": 9.81733636408027e-05, "loss": 0.0287, "step": 13200 }, { "grad_norm": 0.24264180660247803, "learning_rate": 9.816893258324649e-05, "loss": 0.0312, "step": 13210 }, { "grad_norm": 0.293254554271698, "learning_rate": 9.816449625800823e-05, "loss": 0.033, "step": 13220 }, { "grad_norm": 0.33683139085769653, "learning_rate": 9.816005466557308e-05, "loss": 0.0332, "step": 13230 }, { "grad_norm": 0.3053472340106964, "learning_rate": 9.815560780642674e-05, "loss": 0.0297, "step": 13240 }, { "grad_norm": 0.30011266469955444, "learning_rate": 9.815115568105555e-05, "loss": 0.0293, "step": 13250 }, { "grad_norm": 0.33604180812835693, "learning_rate": 9.814669828994638e-05, "loss": 0.0319, "step": 13260 }, { "grad_norm": 0.3488183319568634, "learning_rate": 9.814223563358665e-05, "loss": 0.0323, "step": 13270 }, { "grad_norm": 0.31464847922325134, "learning_rate": 9.813776771246443e-05, "loss": 0.0337, "step": 13280 }, { "grad_norm": 0.24002525210380554, "learning_rate": 9.813329452706829e-05, "loss": 0.0364, "step": 13290 }, { "grad_norm": 0.2479318231344223, "learning_rate": 9.812881607788744e-05, "loss": 0.0337, "step": 13300 }, { "grad_norm": 0.39938828349113464, "learning_rate": 9.812433236541163e-05, "loss": 0.0384, "step": 13310 }, { "grad_norm": 0.4021884799003601, "learning_rate": 9.811984339013116e-05, "loss": 0.0425, "step": 13320 }, { "grad_norm": 0.30954745411872864, "learning_rate": 9.811534915253698e-05, "loss": 0.0389, "step": 13330 }, { "grad_norm": 0.4165562093257904, "learning_rate": 9.811084965312056e-05, "loss": 0.0373, "step": 13340 }, { "grad_norm": 0.237814262509346, "learning_rate": 9.810634489237396e-05, "loss": 0.0333, "step": 13350 }, { "grad_norm": 0.30517521500587463, "learning_rate": 9.81018348707898e-05, "loss": 0.0364, "step": 13360 }, { "grad_norm": 0.2958488166332245, "learning_rate": 9.809731958886131e-05, "loss": 0.0318, "step": 13370 }, { "grad_norm": 0.28781241178512573, "learning_rate": 9.809279904708224e-05, "loss": 0.033, "step": 13380 }, { "grad_norm": 0.26611506938934326, "learning_rate": 9.808827324594699e-05, "loss": 0.0317, "step": 13390 }, { "grad_norm": 0.23561981320381165, "learning_rate": 9.808374218595046e-05, "loss": 0.03, "step": 13400 }, { "grad_norm": 0.2920841872692108, "learning_rate": 9.80792058675882e-05, "loss": 0.031, "step": 13410 }, { "grad_norm": 0.21680541336536407, "learning_rate": 9.807466429135627e-05, "loss": 0.0308, "step": 13420 }, { "grad_norm": 0.25628340244293213, "learning_rate": 9.807011745775132e-05, "loss": 0.0316, "step": 13430 }, { "grad_norm": 0.23259133100509644, "learning_rate": 9.806556536727061e-05, "loss": 0.0308, "step": 13440 }, { "grad_norm": 0.3060726821422577, "learning_rate": 9.806100802041193e-05, "loss": 0.0284, "step": 13450 }, { "grad_norm": 0.2954760193824768, "learning_rate": 9.805644541767368e-05, "loss": 0.0283, "step": 13460 }, { "grad_norm": 0.2563876211643219, "learning_rate": 9.805187755955478e-05, "loss": 0.0315, "step": 13470 }, { "grad_norm": 0.1909412294626236, "learning_rate": 9.804730444655483e-05, "loss": 0.0298, "step": 13480 }, { "grad_norm": 0.2604028284549713, "learning_rate": 9.804272607917388e-05, "loss": 0.0288, "step": 13490 }, { "grad_norm": 0.22944574058055878, "learning_rate": 9.803814245791265e-05, "loss": 0.032, "step": 13500 }, { "grad_norm": 0.3260123133659363, "learning_rate": 9.803355358327239e-05, "loss": 0.0348, "step": 13510 }, { "grad_norm": 0.32243216037750244, "learning_rate": 9.802895945575492e-05, "loss": 0.0306, "step": 13520 }, { "grad_norm": 0.22199389338493347, "learning_rate": 9.802436007586266e-05, "loss": 0.0283, "step": 13530 }, { "grad_norm": 0.2128460854291916, "learning_rate": 9.801975544409858e-05, "loss": 0.0294, "step": 13540 }, { "grad_norm": 0.2635265588760376, "learning_rate": 9.801514556096625e-05, "loss": 0.0319, "step": 13550 }, { "grad_norm": 0.2695370614528656, "learning_rate": 9.801053042696977e-05, "loss": 0.0304, "step": 13560 }, { "grad_norm": 0.26869112253189087, "learning_rate": 9.800591004261388e-05, "loss": 0.0305, "step": 13570 }, { "grad_norm": 0.22939139604568481, "learning_rate": 9.800128440840385e-05, "loss": 0.0302, "step": 13580 }, { "grad_norm": 0.298961341381073, "learning_rate": 9.799665352484552e-05, "loss": 0.0335, "step": 13590 }, { "grad_norm": 0.3180226981639862, "learning_rate": 9.799201739244532e-05, "loss": 0.0305, "step": 13600 }, { "grad_norm": 0.30963635444641113, "learning_rate": 9.798737601171025e-05, "loss": 0.029, "step": 13610 }, { "grad_norm": 0.26703596115112305, "learning_rate": 9.79827293831479e-05, "loss": 0.0286, "step": 13620 }, { "grad_norm": 0.2075873613357544, "learning_rate": 9.797807750726638e-05, "loss": 0.0299, "step": 13630 }, { "grad_norm": 0.20777374505996704, "learning_rate": 9.797342038457446e-05, "loss": 0.029, "step": 13640 }, { "grad_norm": 0.302465558052063, "learning_rate": 9.796875801558141e-05, "loss": 0.0295, "step": 13650 }, { "grad_norm": 0.2556585371494293, "learning_rate": 9.79640904007971e-05, "loss": 0.0291, "step": 13660 }, { "grad_norm": 0.24454431235790253, "learning_rate": 9.795941754073199e-05, "loss": 0.027, "step": 13670 }, { "grad_norm": 0.261404424905777, "learning_rate": 9.795473943589705e-05, "loss": 0.0296, "step": 13680 }, { "grad_norm": 0.2783116102218628, "learning_rate": 9.795005608680394e-05, "loss": 0.0277, "step": 13690 }, { "grad_norm": 0.2941468358039856, "learning_rate": 9.794536749396477e-05, "loss": 0.0314, "step": 13700 }, { "grad_norm": 0.2939215898513794, "learning_rate": 9.79406736578923e-05, "loss": 0.0312, "step": 13710 }, { "grad_norm": 0.22574836015701294, "learning_rate": 9.793597457909984e-05, "loss": 0.0261, "step": 13720 }, { "grad_norm": 0.2142631709575653, "learning_rate": 9.793127025810127e-05, "loss": 0.0292, "step": 13730 }, { "grad_norm": 0.19440428912639618, "learning_rate": 9.792656069541104e-05, "loss": 0.0263, "step": 13740 }, { "grad_norm": 0.2398623526096344, "learning_rate": 9.79218458915442e-05, "loss": 0.0292, "step": 13750 }, { "grad_norm": 0.2992594540119171, "learning_rate": 9.791712584701634e-05, "loss": 0.0305, "step": 13760 }, { "grad_norm": 0.32678699493408203, "learning_rate": 9.791240056234364e-05, "loss": 0.028, "step": 13770 }, { "grad_norm": 0.27622243762016296, "learning_rate": 9.790767003804283e-05, "loss": 0.0298, "step": 13780 }, { "grad_norm": 0.27445924282073975, "learning_rate": 9.790293427463126e-05, "loss": 0.0301, "step": 13790 }, { "grad_norm": 0.2923699617385864, "learning_rate": 9.789819327262684e-05, "loss": 0.0274, "step": 13800 }, { "grad_norm": 0.3320182263851166, "learning_rate": 9.7893447032548e-05, "loss": 0.0266, "step": 13810 }, { "grad_norm": 0.33338662981987, "learning_rate": 9.78886955549138e-05, "loss": 0.0303, "step": 13820 }, { "grad_norm": 0.2092846930027008, "learning_rate": 9.788393884024387e-05, "loss": 0.0302, "step": 13830 }, { "grad_norm": 0.25150734186172485, "learning_rate": 9.787917688905836e-05, "loss": 0.0281, "step": 13840 }, { "grad_norm": 0.30129608511924744, "learning_rate": 9.787440970187807e-05, "loss": 0.0274, "step": 13850 }, { "grad_norm": 0.247513547539711, "learning_rate": 9.786963727922429e-05, "loss": 0.0255, "step": 13860 }, { "grad_norm": 0.19644443690776825, "learning_rate": 9.786485962161897e-05, "loss": 0.0305, "step": 13870 }, { "grad_norm": 0.2696179151535034, "learning_rate": 9.786007672958455e-05, "loss": 0.0306, "step": 13880 }, { "grad_norm": 0.2405666708946228, "learning_rate": 9.78552886036441e-05, "loss": 0.0287, "step": 13890 }, { "grad_norm": 0.22654910385608673, "learning_rate": 9.785049524432124e-05, "loss": 0.0259, "step": 13900 }, { "grad_norm": 0.26767444610595703, "learning_rate": 9.784569665214016e-05, "loss": 0.03, "step": 13910 }, { "grad_norm": 0.3031527101993561, "learning_rate": 9.784089282762563e-05, "loss": 0.0273, "step": 13920 }, { "grad_norm": 0.23645278811454773, "learning_rate": 9.7836083771303e-05, "loss": 0.0246, "step": 13930 }, { "grad_norm": 0.24594293534755707, "learning_rate": 9.783126948369817e-05, "loss": 0.0241, "step": 13940 }, { "grad_norm": 0.27141955494880676, "learning_rate": 9.78264499653376e-05, "loss": 0.0274, "step": 13950 }, { "grad_norm": 0.2776170074939728, "learning_rate": 9.782162521674838e-05, "loss": 0.0283, "step": 13960 }, { "grad_norm": 0.2669479250907898, "learning_rate": 9.781679523845812e-05, "loss": 0.0291, "step": 13970 }, { "grad_norm": 0.22734753787517548, "learning_rate": 9.781196003099502e-05, "loss": 0.0264, "step": 13980 }, { "grad_norm": 0.2289205640554428, "learning_rate": 9.780711959488786e-05, "loss": 0.0282, "step": 13990 }, { "grad_norm": 0.23414333164691925, "learning_rate": 9.780227393066599e-05, "loss": 0.0251, "step": 14000 }, { "grad_norm": 0.27117517590522766, "learning_rate": 9.77974230388593e-05, "loss": 0.0269, "step": 14010 }, { "grad_norm": 0.3184029161930084, "learning_rate": 9.779256691999829e-05, "loss": 0.0289, "step": 14020 }, { "grad_norm": 0.17561864852905273, "learning_rate": 9.778770557461403e-05, "loss": 0.0252, "step": 14030 }, { "grad_norm": 0.28779086470603943, "learning_rate": 9.778283900323812e-05, "loss": 0.026, "step": 14040 }, { "grad_norm": 0.3090340793132782, "learning_rate": 9.777796720640277e-05, "loss": 0.029, "step": 14050 }, { "grad_norm": 0.28799736499786377, "learning_rate": 9.777309018464078e-05, "loss": 0.0272, "step": 14060 }, { "grad_norm": 0.23291467130184174, "learning_rate": 9.776820793848547e-05, "loss": 0.023, "step": 14070 }, { "grad_norm": 0.251741886138916, "learning_rate": 9.776332046847075e-05, "loss": 0.0253, "step": 14080 }, { "grad_norm": 0.23076948523521423, "learning_rate": 9.775842777513111e-05, "loss": 0.0252, "step": 14090 }, { "grad_norm": 0.2907038629055023, "learning_rate": 9.775352985900163e-05, "loss": 0.027, "step": 14100 }, { "grad_norm": 0.298289954662323, "learning_rate": 9.774862672061791e-05, "loss": 0.0273, "step": 14110 }, { "grad_norm": 0.23724757134914398, "learning_rate": 9.774371836051616e-05, "loss": 0.0279, "step": 14120 }, { "grad_norm": 0.32800889015197754, "learning_rate": 9.773880477923315e-05, "loss": 0.0266, "step": 14130 }, { "grad_norm": 0.2935485541820526, "learning_rate": 9.773388597730623e-05, "loss": 0.0266, "step": 14140 }, { "grad_norm": 0.30408209562301636, "learning_rate": 9.77289619552733e-05, "loss": 0.0268, "step": 14150 }, { "grad_norm": 0.2030346393585205, "learning_rate": 9.772403271367285e-05, "loss": 0.0281, "step": 14160 }, { "grad_norm": 0.2448008805513382, "learning_rate": 9.771909825304396e-05, "loss": 0.0294, "step": 14170 }, { "grad_norm": 0.19641289114952087, "learning_rate": 9.771415857392619e-05, "loss": 0.0287, "step": 14180 }, { "grad_norm": 0.26247474551200867, "learning_rate": 9.770921367685978e-05, "loss": 0.0288, "step": 14190 }, { "grad_norm": 0.2699918746948242, "learning_rate": 9.770426356238551e-05, "loss": 0.0264, "step": 14200 }, { "grad_norm": 0.314998596906662, "learning_rate": 9.769930823104469e-05, "loss": 0.0324, "step": 14210 }, { "grad_norm": 0.27405720949172974, "learning_rate": 9.769434768337926e-05, "loss": 0.0308, "step": 14220 }, { "grad_norm": 0.3520926237106323, "learning_rate": 9.768938191993164e-05, "loss": 0.0317, "step": 14230 }, { "grad_norm": 0.2502322196960449, "learning_rate": 9.768441094124494e-05, "loss": 0.03, "step": 14240 }, { "grad_norm": 0.3095063865184784, "learning_rate": 9.767943474786275e-05, "loss": 0.0309, "step": 14250 }, { "grad_norm": 0.31748828291893005, "learning_rate": 9.767445334032923e-05, "loss": 0.0289, "step": 14260 }, { "grad_norm": 0.2391621619462967, "learning_rate": 9.766946671918919e-05, "loss": 0.0304, "step": 14270 }, { "grad_norm": 0.18647700548171997, "learning_rate": 9.766447488498796e-05, "loss": 0.0269, "step": 14280 }, { "grad_norm": 0.2270863652229309, "learning_rate": 9.765947783827139e-05, "loss": 0.0283, "step": 14290 }, { "grad_norm": 0.2577766180038452, "learning_rate": 9.765447557958599e-05, "loss": 0.0283, "step": 14300 }, { "grad_norm": 0.19905437529087067, "learning_rate": 9.764946810947879e-05, "loss": 0.031, "step": 14310 }, { "grad_norm": 0.2727477252483368, "learning_rate": 9.764445542849738e-05, "loss": 0.0289, "step": 14320 }, { "grad_norm": 0.30688413977622986, "learning_rate": 9.763943753718998e-05, "loss": 0.0281, "step": 14330 }, { "grad_norm": 0.23842239379882812, "learning_rate": 9.76344144361053e-05, "loss": 0.03, "step": 14340 }, { "grad_norm": 0.21833154559135437, "learning_rate": 9.762938612579269e-05, "loss": 0.0272, "step": 14350 }, { "grad_norm": 0.22243042290210724, "learning_rate": 9.762435260680202e-05, "loss": 0.0252, "step": 14360 }, { "grad_norm": 0.2733312249183655, "learning_rate": 9.761931387968373e-05, "loss": 0.0237, "step": 14370 }, { "grad_norm": 0.2626339793205261, "learning_rate": 9.76142699449889e-05, "loss": 0.0283, "step": 14380 }, { "grad_norm": 0.19455000758171082, "learning_rate": 9.760922080326908e-05, "loss": 0.0292, "step": 14390 }, { "grad_norm": 0.2517252266407013, "learning_rate": 9.760416645507644e-05, "loss": 0.0242, "step": 14400 }, { "grad_norm": 0.24748505651950836, "learning_rate": 9.759910690096375e-05, "loss": 0.0248, "step": 14410 }, { "grad_norm": 0.21888084709644318, "learning_rate": 9.759404214148429e-05, "loss": 0.0222, "step": 14420 }, { "grad_norm": 0.2254614681005478, "learning_rate": 9.758897217719191e-05, "loss": 0.0253, "step": 14430 }, { "grad_norm": 0.25682127475738525, "learning_rate": 9.758389700864113e-05, "loss": 0.0275, "step": 14440 }, { "grad_norm": 0.26598337292671204, "learning_rate": 9.757881663638688e-05, "loss": 0.0261, "step": 14450 }, { "grad_norm": 0.30154910683631897, "learning_rate": 9.757373106098478e-05, "loss": 0.0261, "step": 14460 }, { "grad_norm": 0.25273069739341736, "learning_rate": 9.756864028299097e-05, "loss": 0.0246, "step": 14470 }, { "grad_norm": 0.36628827452659607, "learning_rate": 9.75635443029622e-05, "loss": 0.0256, "step": 14480 }, { "grad_norm": 0.28910377621650696, "learning_rate": 9.755844312145572e-05, "loss": 0.0263, "step": 14490 }, { "grad_norm": 0.28148046135902405, "learning_rate": 9.755333673902941e-05, "loss": 0.0279, "step": 14500 }, { "grad_norm": 0.22082363069057465, "learning_rate": 9.75482251562417e-05, "loss": 0.0273, "step": 14510 }, { "grad_norm": 0.2292354851961136, "learning_rate": 9.754310837365155e-05, "loss": 0.0263, "step": 14520 }, { "grad_norm": 0.2276451587677002, "learning_rate": 9.753798639181856e-05, "loss": 0.0262, "step": 14530 }, { "grad_norm": 0.23344208300113678, "learning_rate": 9.753285921130286e-05, "loss": 0.0251, "step": 14540 }, { "grad_norm": 0.298591285943985, "learning_rate": 9.752772683266512e-05, "loss": 0.027, "step": 14550 }, { "grad_norm": 0.2860519587993622, "learning_rate": 9.752258925646665e-05, "loss": 0.0283, "step": 14560 }, { "grad_norm": 0.2752930521965027, "learning_rate": 9.751744648326926e-05, "loss": 0.026, "step": 14570 }, { "grad_norm": 0.24824205040931702, "learning_rate": 9.751229851363536e-05, "loss": 0.0261, "step": 14580 }, { "grad_norm": 0.1915438175201416, "learning_rate": 9.750714534812793e-05, "loss": 0.0248, "step": 14590 }, { "grad_norm": 0.19697384536266327, "learning_rate": 9.750198698731053e-05, "loss": 0.0257, "step": 14600 }, { "grad_norm": 0.2527920603752136, "learning_rate": 9.749682343174722e-05, "loss": 0.0274, "step": 14610 }, { "grad_norm": 0.23739713430404663, "learning_rate": 9.749165468200272e-05, "loss": 0.0251, "step": 14620 }, { "grad_norm": 0.2883125841617584, "learning_rate": 9.748648073864229e-05, "loss": 0.0242, "step": 14630 }, { "grad_norm": 0.3415689766407013, "learning_rate": 9.748130160223168e-05, "loss": 0.0275, "step": 14640 }, { "grad_norm": 0.19158576428890228, "learning_rate": 9.747611727333734e-05, "loss": 0.0258, "step": 14650 }, { "grad_norm": 0.23102834820747375, "learning_rate": 9.74709277525262e-05, "loss": 0.0255, "step": 14660 }, { "grad_norm": 0.24478214979171753, "learning_rate": 9.746573304036576e-05, "loss": 0.0298, "step": 14670 }, { "grad_norm": 0.2533312439918518, "learning_rate": 9.746053313742412e-05, "loss": 0.0269, "step": 14680 }, { "grad_norm": 0.2794458866119385, "learning_rate": 9.745532804426994e-05, "loss": 0.0282, "step": 14690 }, { "grad_norm": 0.252586305141449, "learning_rate": 9.745011776147242e-05, "loss": 0.0297, "step": 14700 }, { "grad_norm": 0.22728823125362396, "learning_rate": 9.744490228960138e-05, "loss": 0.0256, "step": 14710 }, { "grad_norm": 0.31294262409210205, "learning_rate": 9.743968162922713e-05, "loss": 0.0264, "step": 14720 }, { "grad_norm": 0.23228289186954498, "learning_rate": 9.743445578092064e-05, "loss": 0.0272, "step": 14730 }, { "grad_norm": 0.24923112988471985, "learning_rate": 9.742922474525338e-05, "loss": 0.0252, "step": 14740 }, { "grad_norm": 0.2799650728702545, "learning_rate": 9.742398852279741e-05, "loss": 0.0267, "step": 14750 }, { "grad_norm": 0.23215612769126892, "learning_rate": 9.741874711412535e-05, "loss": 0.0296, "step": 14760 }, { "grad_norm": 0.2055446356534958, "learning_rate": 9.741350051981042e-05, "loss": 0.0254, "step": 14770 }, { "grad_norm": 0.2284316122531891, "learning_rate": 9.740824874042633e-05, "loss": 0.0263, "step": 14780 }, { "grad_norm": 0.25222963094711304, "learning_rate": 9.740299177654746e-05, "loss": 0.0261, "step": 14790 }, { "grad_norm": 0.20729883015155792, "learning_rate": 9.739772962874867e-05, "loss": 0.028, "step": 14800 }, { "grad_norm": 0.18505802750587463, "learning_rate": 9.739246229760541e-05, "loss": 0.0236, "step": 14810 }, { "grad_norm": 0.2638043165206909, "learning_rate": 9.738718978369376e-05, "loss": 0.0245, "step": 14820 }, { "grad_norm": 0.214996799826622, "learning_rate": 9.738191208759025e-05, "loss": 0.026, "step": 14830 }, { "grad_norm": 0.2905028462409973, "learning_rate": 9.73766292098721e-05, "loss": 0.0279, "step": 14840 }, { "grad_norm": 0.2876034677028656, "learning_rate": 9.737134115111699e-05, "loss": 0.0231, "step": 14850 }, { "grad_norm": 0.2178347259759903, "learning_rate": 9.736604791190323e-05, "loss": 0.0254, "step": 14860 }, { "grad_norm": 0.18213436007499695, "learning_rate": 9.73607494928097e-05, "loss": 0.0253, "step": 14870 }, { "grad_norm": 0.28656885027885437, "learning_rate": 9.735544589441581e-05, "loss": 0.0259, "step": 14880 }, { "grad_norm": 0.23305976390838623, "learning_rate": 9.735013711730154e-05, "loss": 0.0263, "step": 14890 }, { "grad_norm": 0.23939120769500732, "learning_rate": 9.734482316204747e-05, "loss": 0.0294, "step": 14900 }, { "grad_norm": 0.2941504120826721, "learning_rate": 9.733950402923473e-05, "loss": 0.0268, "step": 14910 }, { "grad_norm": 0.2255883365869522, "learning_rate": 9.7334179719445e-05, "loss": 0.0255, "step": 14920 }, { "grad_norm": 0.2585432529449463, "learning_rate": 9.732885023326053e-05, "loss": 0.0304, "step": 14930 }, { "grad_norm": 0.2745744287967682, "learning_rate": 9.732351557126418e-05, "loss": 0.028, "step": 14940 }, { "grad_norm": 0.3240102231502533, "learning_rate": 9.731817573403929e-05, "loss": 0.0251, "step": 14950 }, { "grad_norm": 0.20691834390163422, "learning_rate": 9.731283072216985e-05, "loss": 0.024, "step": 14960 }, { "grad_norm": 0.22144129872322083, "learning_rate": 9.730748053624039e-05, "loss": 0.0265, "step": 14970 }, { "grad_norm": 0.2775701582431793, "learning_rate": 9.730212517683598e-05, "loss": 0.025, "step": 14980 }, { "grad_norm": 0.3059388995170593, "learning_rate": 9.729676464454228e-05, "loss": 0.0251, "step": 14990 }, { "grad_norm": 0.24881458282470703, "learning_rate": 9.72913989399455e-05, "loss": 0.0256, "step": 15000 }, { "grad_norm": 0.26439762115478516, "learning_rate": 9.728602806363242e-05, "loss": 0.0293, "step": 15010 }, { "grad_norm": 0.263225257396698, "learning_rate": 9.728065201619043e-05, "loss": 0.0318, "step": 15020 }, { "grad_norm": 0.28166836500167847, "learning_rate": 9.727527079820742e-05, "loss": 0.028, "step": 15030 }, { "grad_norm": 0.34041112661361694, "learning_rate": 9.726988441027186e-05, "loss": 0.028, "step": 15040 }, { "grad_norm": 0.6426969170570374, "learning_rate": 9.726449285297281e-05, "loss": 0.0311, "step": 15050 }, { "grad_norm": 0.28584131598472595, "learning_rate": 9.72590961268999e-05, "loss": 0.0318, "step": 15060 }, { "grad_norm": 0.32489123940467834, "learning_rate": 9.725369423264328e-05, "loss": 0.0339, "step": 15070 }, { "grad_norm": 0.3932313323020935, "learning_rate": 9.72482871707937e-05, "loss": 0.0366, "step": 15080 }, { "grad_norm": 0.22063206136226654, "learning_rate": 9.724287494194247e-05, "loss": 0.0334, "step": 15090 }, { "grad_norm": 0.22293372452259064, "learning_rate": 9.723745754668147e-05, "loss": 0.0303, "step": 15100 }, { "grad_norm": 0.26221221685409546, "learning_rate": 9.723203498560313e-05, "loss": 0.0308, "step": 15110 }, { "grad_norm": 0.2558801472187042, "learning_rate": 9.722660725930046e-05, "loss": 0.0293, "step": 15120 }, { "grad_norm": 0.23808367550373077, "learning_rate": 9.722117436836702e-05, "loss": 0.0278, "step": 15130 }, { "grad_norm": 0.28882837295532227, "learning_rate": 9.721573631339696e-05, "loss": 0.0304, "step": 15140 }, { "grad_norm": 0.3150293827056885, "learning_rate": 9.721029309498494e-05, "loss": 0.03, "step": 15150 }, { "grad_norm": 0.6899219751358032, "learning_rate": 9.720484471372627e-05, "loss": 0.0364, "step": 15160 }, { "grad_norm": 0.34068915247917175, "learning_rate": 9.719939117021673e-05, "loss": 0.0322, "step": 15170 }, { "grad_norm": 0.2614908218383789, "learning_rate": 9.719393246505275e-05, "loss": 0.0301, "step": 15180 }, { "grad_norm": 0.18559610843658447, "learning_rate": 9.718846859883128e-05, "loss": 0.0325, "step": 15190 }, { "grad_norm": 0.31829509139060974, "learning_rate": 9.718299957214982e-05, "loss": 0.0336, "step": 15200 }, { "grad_norm": 0.27056509256362915, "learning_rate": 9.717752538560646e-05, "loss": 0.0327, "step": 15210 }, { "grad_norm": 0.2936606705188751, "learning_rate": 9.717204603979986e-05, "loss": 0.0314, "step": 15220 }, { "grad_norm": 0.2825523614883423, "learning_rate": 9.716656153532922e-05, "loss": 0.0333, "step": 15230 }, { "grad_norm": 0.2044389843940735, "learning_rate": 9.716107187279434e-05, "loss": 0.0278, "step": 15240 }, { "grad_norm": 0.28212350606918335, "learning_rate": 9.715557705279555e-05, "loss": 0.0275, "step": 15250 }, { "grad_norm": 0.30115315318107605, "learning_rate": 9.715007707593372e-05, "loss": 0.0275, "step": 15260 }, { "grad_norm": 0.2772437632083893, "learning_rate": 9.714457194281036e-05, "loss": 0.031, "step": 15270 }, { "grad_norm": 0.25552985072135925, "learning_rate": 9.713906165402751e-05, "loss": 0.0319, "step": 15280 }, { "grad_norm": 0.24487441778182983, "learning_rate": 9.713354621018774e-05, "loss": 0.028, "step": 15290 }, { "grad_norm": 0.2281511425971985, "learning_rate": 9.712802561189422e-05, "loss": 0.0313, "step": 15300 }, { "grad_norm": 0.26994436979293823, "learning_rate": 9.712249985975069e-05, "loss": 0.0303, "step": 15310 }, { "grad_norm": 0.31791800260543823, "learning_rate": 9.71169689543614e-05, "loss": 0.0283, "step": 15320 }, { "grad_norm": 0.28957507014274597, "learning_rate": 9.711143289633123e-05, "loss": 0.0395, "step": 15330 }, { "grad_norm": 0.32951927185058594, "learning_rate": 9.710589168626561e-05, "loss": 0.0392, "step": 15340 }, { "grad_norm": 0.26442602276802063, "learning_rate": 9.710034532477048e-05, "loss": 0.0346, "step": 15350 }, { "grad_norm": 0.2909398674964905, "learning_rate": 9.709479381245239e-05, "loss": 0.0338, "step": 15360 }, { "grad_norm": 0.25230392813682556, "learning_rate": 9.708923714991847e-05, "loss": 0.0334, "step": 15370 }, { "grad_norm": 0.32680997252464294, "learning_rate": 9.708367533777638e-05, "loss": 0.0302, "step": 15380 }, { "grad_norm": 0.2301722913980484, "learning_rate": 9.707810837663431e-05, "loss": 0.0312, "step": 15390 }, { "grad_norm": 0.2425609976053238, "learning_rate": 9.707253626710113e-05, "loss": 0.0304, "step": 15400 }, { "grad_norm": 0.21582132577896118, "learning_rate": 9.706695900978613e-05, "loss": 0.0237, "step": 15410 }, { "grad_norm": 0.2647491693496704, "learning_rate": 9.706137660529926e-05, "loss": 0.0318, "step": 15420 }, { "grad_norm": 0.2833963632583618, "learning_rate": 9.705578905425101e-05, "loss": 0.0296, "step": 15430 }, { "grad_norm": 0.2789071202278137, "learning_rate": 9.705019635725241e-05, "loss": 0.0357, "step": 15440 }, { "grad_norm": 0.28495121002197266, "learning_rate": 9.704459851491508e-05, "loss": 0.0291, "step": 15450 }, { "grad_norm": 0.36387526988983154, "learning_rate": 9.703899552785118e-05, "loss": 0.029, "step": 15460 }, { "grad_norm": 0.2918214201927185, "learning_rate": 9.703338739667346e-05, "loss": 0.0265, "step": 15470 }, { "grad_norm": 0.20762860774993896, "learning_rate": 9.70277741219952e-05, "loss": 0.0285, "step": 15480 }, { "grad_norm": 0.30133581161499023, "learning_rate": 9.702215570443027e-05, "loss": 0.0308, "step": 15490 }, { "grad_norm": 0.2515028417110443, "learning_rate": 9.701653214459309e-05, "loss": 0.0315, "step": 15500 }, { "grad_norm": 0.24528324604034424, "learning_rate": 9.701090344309865e-05, "loss": 0.0325, "step": 15510 }, { "grad_norm": 0.2587517201900482, "learning_rate": 9.700526960056247e-05, "loss": 0.0306, "step": 15520 }, { "grad_norm": 0.2674628496170044, "learning_rate": 9.699963061760068e-05, "loss": 0.0272, "step": 15530 }, { "grad_norm": 0.22572079300880432, "learning_rate": 9.699398649482997e-05, "loss": 0.0281, "step": 15540 }, { "grad_norm": 0.24437536299228668, "learning_rate": 9.698833723286753e-05, "loss": 0.0295, "step": 15550 }, { "grad_norm": 0.2775322496891022, "learning_rate": 9.698268283233118e-05, "loss": 0.0301, "step": 15560 }, { "grad_norm": 0.27205780148506165, "learning_rate": 9.697702329383929e-05, "loss": 0.0293, "step": 15570 }, { "grad_norm": 0.31607958674430847, "learning_rate": 9.697135861801074e-05, "loss": 0.0282, "step": 15580 }, { "grad_norm": 0.21651768684387207, "learning_rate": 9.696568880546505e-05, "loss": 0.0255, "step": 15590 }, { "grad_norm": 0.21868498623371124, "learning_rate": 9.696001385682223e-05, "loss": 0.0261, "step": 15600 }, { "grad_norm": 0.2463313490152359, "learning_rate": 9.695433377270291e-05, "loss": 0.0282, "step": 15610 }, { "grad_norm": 0.23275558650493622, "learning_rate": 9.694864855372824e-05, "loss": 0.0284, "step": 15620 }, { "grad_norm": 0.28463876247406006, "learning_rate": 9.694295820051995e-05, "loss": 0.0262, "step": 15630 }, { "grad_norm": 0.25348785519599915, "learning_rate": 9.693726271370032e-05, "loss": 0.0267, "step": 15640 }, { "grad_norm": 0.21031400561332703, "learning_rate": 9.693156209389221e-05, "loss": 0.0264, "step": 15650 }, { "grad_norm": 0.2101132720708847, "learning_rate": 9.692585634171905e-05, "loss": 0.0269, "step": 15660 }, { "grad_norm": 0.16286510229110718, "learning_rate": 9.692014545780476e-05, "loss": 0.0226, "step": 15670 }, { "grad_norm": 0.25879818201065063, "learning_rate": 9.691442944277393e-05, "loss": 0.0261, "step": 15680 }, { "grad_norm": 0.3095563054084778, "learning_rate": 9.690870829725162e-05, "loss": 0.0295, "step": 15690 }, { "grad_norm": 0.2868935167789459, "learning_rate": 9.69029820218635e-05, "loss": 0.0286, "step": 15700 }, { "grad_norm": 0.21906791627407074, "learning_rate": 9.689725061723579e-05, "loss": 0.0279, "step": 15710 }, { "grad_norm": 0.2331373244524002, "learning_rate": 9.689151408399527e-05, "loss": 0.0289, "step": 15720 }, { "grad_norm": 0.3022160232067108, "learning_rate": 9.688577242276924e-05, "loss": 0.0257, "step": 15730 }, { "grad_norm": 0.17854981124401093, "learning_rate": 9.688002563418566e-05, "loss": 0.0282, "step": 15740 }, { "grad_norm": 0.2069576382637024, "learning_rate": 9.687427371887293e-05, "loss": 0.0266, "step": 15750 }, { "grad_norm": 0.3401699364185333, "learning_rate": 9.686851667746012e-05, "loss": 0.0263, "step": 15760 }, { "grad_norm": 0.2634603679180145, "learning_rate": 9.686275451057677e-05, "loss": 0.0257, "step": 15770 }, { "grad_norm": 0.2228902280330658, "learning_rate": 9.685698721885308e-05, "loss": 0.0267, "step": 15780 }, { "grad_norm": 0.22453714907169342, "learning_rate": 9.68512148029197e-05, "loss": 0.0322, "step": 15790 }, { "grad_norm": 0.2505667507648468, "learning_rate": 9.684543726340791e-05, "loss": 0.0289, "step": 15800 }, { "grad_norm": 0.23990362882614136, "learning_rate": 9.683965460094952e-05, "loss": 0.0263, "step": 15810 }, { "grad_norm": 0.2149142026901245, "learning_rate": 9.683386681617694e-05, "loss": 0.0253, "step": 15820 }, { "grad_norm": 0.2281549572944641, "learning_rate": 9.68280739097231e-05, "loss": 0.0265, "step": 15830 }, { "grad_norm": 0.23760508000850677, "learning_rate": 9.682227588222148e-05, "loss": 0.0242, "step": 15840 }, { "grad_norm": 0.4832364320755005, "learning_rate": 9.681647273430618e-05, "loss": 0.0274, "step": 15850 }, { "grad_norm": 0.2452782243490219, "learning_rate": 9.681066446661182e-05, "loss": 0.0313, "step": 15860 }, { "grad_norm": 0.21269267797470093, "learning_rate": 9.680485107977357e-05, "loss": 0.0282, "step": 15870 }, { "grad_norm": 0.27125439047813416, "learning_rate": 9.679903257442716e-05, "loss": 0.0312, "step": 15880 }, { "grad_norm": 0.19509756565093994, "learning_rate": 9.679320895120891e-05, "loss": 0.0292, "step": 15890 }, { "grad_norm": 0.24789798259735107, "learning_rate": 9.67873802107557e-05, "loss": 0.0271, "step": 15900 }, { "grad_norm": 0.2124495655298233, "learning_rate": 9.67815463537049e-05, "loss": 0.0267, "step": 15910 }, { "grad_norm": 0.244104266166687, "learning_rate": 9.677570738069457e-05, "loss": 0.0274, "step": 15920 }, { "grad_norm": 0.2056456208229065, "learning_rate": 9.676986329236318e-05, "loss": 0.0257, "step": 15930 }, { "grad_norm": 0.28650447726249695, "learning_rate": 9.676401408934987e-05, "loss": 0.0246, "step": 15940 }, { "grad_norm": 0.30517578125, "learning_rate": 9.675815977229428e-05, "loss": 0.027, "step": 15950 }, { "grad_norm": 0.24661733210086823, "learning_rate": 9.675230034183664e-05, "loss": 0.025, "step": 15960 }, { "grad_norm": 0.24215583503246307, "learning_rate": 9.674643579861773e-05, "loss": 0.0274, "step": 15970 }, { "grad_norm": 0.24834980070590973, "learning_rate": 9.674056614327886e-05, "loss": 0.0284, "step": 15980 }, { "grad_norm": 0.19310379028320312, "learning_rate": 9.673469137646198e-05, "loss": 0.0267, "step": 15990 }, { "grad_norm": 0.2246493399143219, "learning_rate": 9.67288114988095e-05, "loss": 0.0262, "step": 16000 }, { "grad_norm": 0.23687219619750977, "learning_rate": 9.672292651096447e-05, "loss": 0.0235, "step": 16010 }, { "grad_norm": 0.2228720486164093, "learning_rate": 9.671703641357042e-05, "loss": 0.0252, "step": 16020 }, { "grad_norm": 0.2190045714378357, "learning_rate": 9.67111412072715e-05, "loss": 0.0245, "step": 16030 }, { "grad_norm": 0.22312642633914948, "learning_rate": 9.670524089271242e-05, "loss": 0.0257, "step": 16040 }, { "grad_norm": 0.34297800064086914, "learning_rate": 9.669933547053842e-05, "loss": 0.0266, "step": 16050 }, { "grad_norm": 0.2649386525154114, "learning_rate": 9.669342494139531e-05, "loss": 0.0278, "step": 16060 }, { "grad_norm": 0.2854766249656677, "learning_rate": 9.668750930592943e-05, "loss": 0.0304, "step": 16070 }, { "grad_norm": 0.1621909886598587, "learning_rate": 9.668158856478775e-05, "loss": 0.0286, "step": 16080 }, { "grad_norm": 0.27694061398506165, "learning_rate": 9.66756627186177e-05, "loss": 0.0296, "step": 16090 }, { "grad_norm": 0.194977805018425, "learning_rate": 9.666973176806737e-05, "loss": 0.0226, "step": 16100 }, { "grad_norm": 0.25906822085380554, "learning_rate": 9.666379571378534e-05, "loss": 0.0244, "step": 16110 }, { "grad_norm": 0.24851535260677338, "learning_rate": 9.665785455642076e-05, "loss": 0.0256, "step": 16120 }, { "grad_norm": 0.2508121430873871, "learning_rate": 9.665190829662337e-05, "loss": 0.0247, "step": 16130 }, { "grad_norm": 0.2476905733346939, "learning_rate": 9.664595693504342e-05, "loss": 0.0242, "step": 16140 }, { "grad_norm": 0.2129795104265213, "learning_rate": 9.664000047233175e-05, "loss": 0.0248, "step": 16150 }, { "grad_norm": 0.22607582807540894, "learning_rate": 9.663403890913976e-05, "loss": 0.0242, "step": 16160 }, { "grad_norm": 0.30994731187820435, "learning_rate": 9.662807224611938e-05, "loss": 0.0267, "step": 16170 }, { "grad_norm": 0.25746646523475647, "learning_rate": 9.662210048392311e-05, "loss": 0.0257, "step": 16180 }, { "grad_norm": 0.2238623946905136, "learning_rate": 9.661612362320405e-05, "loss": 0.0254, "step": 16190 }, { "grad_norm": 0.2293505221605301, "learning_rate": 9.661014166461579e-05, "loss": 0.0264, "step": 16200 }, { "grad_norm": 0.2878328859806061, "learning_rate": 9.66041546088125e-05, "loss": 0.0242, "step": 16210 }, { "grad_norm": 0.18915337324142456, "learning_rate": 9.659816245644895e-05, "loss": 0.0258, "step": 16220 }, { "grad_norm": 0.18288356065750122, "learning_rate": 9.65921652081804e-05, "loss": 0.0249, "step": 16230 }, { "grad_norm": 0.2215806543827057, "learning_rate": 9.658616286466271e-05, "loss": 0.0224, "step": 16240 }, { "grad_norm": 0.23613549768924713, "learning_rate": 9.65801554265523e-05, "loss": 0.0253, "step": 16250 }, { "grad_norm": 0.2403198480606079, "learning_rate": 9.657414289450612e-05, "loss": 0.0258, "step": 16260 }, { "grad_norm": 0.2654426097869873, "learning_rate": 9.656812526918171e-05, "loss": 0.0265, "step": 16270 }, { "grad_norm": 0.23420776426792145, "learning_rate": 9.656210255123712e-05, "loss": 0.0248, "step": 16280 }, { "grad_norm": 0.2552405297756195, "learning_rate": 9.6556074741331e-05, "loss": 0.03, "step": 16290 }, { "grad_norm": 0.26441243290901184, "learning_rate": 9.655004184012256e-05, "loss": 0.0277, "step": 16300 }, { "grad_norm": 0.2506883442401886, "learning_rate": 9.654400384827152e-05, "loss": 0.0268, "step": 16310 }, { "grad_norm": 0.26494112610816956, "learning_rate": 9.653796076643818e-05, "loss": 0.0291, "step": 16320 }, { "grad_norm": 0.21297024190425873, "learning_rate": 9.653191259528344e-05, "loss": 0.027, "step": 16330 }, { "grad_norm": 0.26174500584602356, "learning_rate": 9.65258593354687e-05, "loss": 0.0268, "step": 16340 }, { "grad_norm": 0.2299984246492386, "learning_rate": 9.651980098765591e-05, "loss": 0.026, "step": 16350 }, { "grad_norm": 0.23590925335884094, "learning_rate": 9.651373755250765e-05, "loss": 0.0253, "step": 16360 }, { "grad_norm": 0.2393990308046341, "learning_rate": 9.650766903068697e-05, "loss": 0.0235, "step": 16370 }, { "grad_norm": 0.24090509116649628, "learning_rate": 9.650159542285753e-05, "loss": 0.0228, "step": 16380 }, { "grad_norm": 0.24653209745883942, "learning_rate": 9.649551672968353e-05, "loss": 0.0266, "step": 16390 }, { "grad_norm": 0.20921041071414948, "learning_rate": 9.648943295182973e-05, "loss": 0.0236, "step": 16400 }, { "grad_norm": 0.21684159338474274, "learning_rate": 9.648334408996144e-05, "loss": 0.025, "step": 16410 }, { "grad_norm": 0.22495697438716888, "learning_rate": 9.647725014474452e-05, "loss": 0.0249, "step": 16420 }, { "grad_norm": 0.2434118390083313, "learning_rate": 9.64711511168454e-05, "loss": 0.0246, "step": 16430 }, { "grad_norm": 0.2327684611082077, "learning_rate": 9.646504700693108e-05, "loss": 0.0231, "step": 16440 }, { "grad_norm": 0.21507830917835236, "learning_rate": 9.645893781566907e-05, "loss": 0.0248, "step": 16450 }, { "grad_norm": 0.1756526529788971, "learning_rate": 9.645282354372744e-05, "loss": 0.021, "step": 16460 }, { "grad_norm": 0.2118740975856781, "learning_rate": 9.644670419177491e-05, "loss": 0.0225, "step": 16470 }, { "grad_norm": 0.21245665848255157, "learning_rate": 9.644057976048062e-05, "loss": 0.0256, "step": 16480 }, { "grad_norm": 0.20278362929821014, "learning_rate": 9.643445025051435e-05, "loss": 0.0252, "step": 16490 }, { "grad_norm": 0.26060664653778076, "learning_rate": 9.642831566254641e-05, "loss": 0.0224, "step": 16500 }, { "grad_norm": 0.20143930613994598, "learning_rate": 9.642217599724769e-05, "loss": 0.0228, "step": 16510 }, { "grad_norm": 0.3152583837509155, "learning_rate": 9.64160312552896e-05, "loss": 0.0225, "step": 16520 }, { "grad_norm": 0.24847501516342163, "learning_rate": 9.64098814373441e-05, "loss": 0.0243, "step": 16530 }, { "grad_norm": 0.259521484375, "learning_rate": 9.640372654408374e-05, "loss": 0.0208, "step": 16540 }, { "grad_norm": 0.23502376675605774, "learning_rate": 9.639756657618162e-05, "loss": 0.0226, "step": 16550 }, { "grad_norm": 0.331023633480072, "learning_rate": 9.639140153431138e-05, "loss": 0.0226, "step": 16560 }, { "grad_norm": 0.19155408442020416, "learning_rate": 9.638523141914721e-05, "loss": 0.0225, "step": 16570 }, { "grad_norm": 0.21778947114944458, "learning_rate": 9.637905623136388e-05, "loss": 0.0248, "step": 16580 }, { "grad_norm": 0.21832942962646484, "learning_rate": 9.637287597163669e-05, "loss": 0.0241, "step": 16590 }, { "grad_norm": 0.2743746042251587, "learning_rate": 9.63666906406415e-05, "loss": 0.0279, "step": 16600 }, { "grad_norm": 0.21731489896774292, "learning_rate": 9.636050023905473e-05, "loss": 0.0247, "step": 16610 }, { "grad_norm": 0.2538300156593323, "learning_rate": 9.635430476755336e-05, "loss": 0.0246, "step": 16620 }, { "grad_norm": 0.1931641846895218, "learning_rate": 9.63481042268149e-05, "loss": 0.0232, "step": 16630 }, { "grad_norm": 0.24980224668979645, "learning_rate": 9.634189861751745e-05, "loss": 0.027, "step": 16640 }, { "grad_norm": 0.3097100257873535, "learning_rate": 9.633568794033967e-05, "loss": 0.0241, "step": 16650 }, { "grad_norm": 0.2634294629096985, "learning_rate": 9.63294721959607e-05, "loss": 0.0241, "step": 16660 }, { "grad_norm": 0.32125017046928406, "learning_rate": 9.63232513850603e-05, "loss": 0.0257, "step": 16670 }, { "grad_norm": 0.23252630233764648, "learning_rate": 9.631702550831878e-05, "loss": 0.0224, "step": 16680 }, { "grad_norm": 0.20754414796829224, "learning_rate": 9.631079456641698e-05, "loss": 0.027, "step": 16690 }, { "grad_norm": 0.24008432030677795, "learning_rate": 9.630455856003632e-05, "loss": 0.0235, "step": 16700 }, { "grad_norm": 0.1944286823272705, "learning_rate": 9.629831748985876e-05, "loss": 0.0213, "step": 16710 }, { "grad_norm": 0.275768905878067, "learning_rate": 9.629207135656679e-05, "loss": 0.023, "step": 16720 }, { "grad_norm": 0.2534201443195343, "learning_rate": 9.628582016084353e-05, "loss": 0.0232, "step": 16730 }, { "grad_norm": 0.23831520974636078, "learning_rate": 9.627956390337254e-05, "loss": 0.0221, "step": 16740 }, { "grad_norm": 0.31455108523368835, "learning_rate": 9.627330258483802e-05, "loss": 0.0242, "step": 16750 }, { "grad_norm": 0.22387662529945374, "learning_rate": 9.62670362059247e-05, "loss": 0.0228, "step": 16760 }, { "grad_norm": 0.24699681997299194, "learning_rate": 9.626076476731786e-05, "loss": 0.024, "step": 16770 }, { "grad_norm": 0.22211384773254395, "learning_rate": 9.625448826970336e-05, "loss": 0.0231, "step": 16780 }, { "grad_norm": 0.2400481253862381, "learning_rate": 9.624820671376755e-05, "loss": 0.0224, "step": 16790 }, { "grad_norm": 0.22272437810897827, "learning_rate": 9.62419201001974e-05, "loss": 0.0254, "step": 16800 }, { "grad_norm": 0.22402971982955933, "learning_rate": 9.623562842968037e-05, "loss": 0.0242, "step": 16810 }, { "grad_norm": 0.266009658575058, "learning_rate": 9.622933170290454e-05, "loss": 0.0256, "step": 16820 }, { "grad_norm": 0.24088531732559204, "learning_rate": 9.622302992055849e-05, "loss": 0.0259, "step": 16830 }, { "grad_norm": 0.2812417447566986, "learning_rate": 9.62167230833314e-05, "loss": 0.0254, "step": 16840 }, { "grad_norm": 0.34467360377311707, "learning_rate": 9.621041119191295e-05, "loss": 0.0263, "step": 16850 }, { "grad_norm": 0.2313227504491806, "learning_rate": 9.620409424699342e-05, "loss": 0.0268, "step": 16860 }, { "grad_norm": 0.19525647163391113, "learning_rate": 9.619777224926359e-05, "loss": 0.0263, "step": 16870 }, { "grad_norm": 0.2110292911529541, "learning_rate": 9.619144519941485e-05, "loss": 0.0232, "step": 16880 }, { "grad_norm": 0.23873397707939148, "learning_rate": 9.618511309813912e-05, "loss": 0.0232, "step": 16890 }, { "grad_norm": 0.20154738426208496, "learning_rate": 9.617877594612886e-05, "loss": 0.0268, "step": 16900 }, { "grad_norm": 0.24355962872505188, "learning_rate": 9.617243374407707e-05, "loss": 0.0254, "step": 16910 }, { "grad_norm": 0.27711576223373413, "learning_rate": 9.616608649267736e-05, "loss": 0.0249, "step": 16920 }, { "grad_norm": 0.24620705842971802, "learning_rate": 9.615973419262385e-05, "loss": 0.0259, "step": 16930 }, { "grad_norm": 0.20824487507343292, "learning_rate": 9.615337684461119e-05, "loss": 0.0215, "step": 16940 }, { "grad_norm": 0.21008014678955078, "learning_rate": 9.614701444933465e-05, "loss": 0.0235, "step": 16950 }, { "grad_norm": 0.19341731071472168, "learning_rate": 9.614064700748997e-05, "loss": 0.0247, "step": 16960 }, { "grad_norm": 0.2991916239261627, "learning_rate": 9.613427451977352e-05, "loss": 0.0218, "step": 16970 }, { "grad_norm": 0.25260430574417114, "learning_rate": 9.612789698688216e-05, "loss": 0.0271, "step": 16980 }, { "grad_norm": 0.2970471680164337, "learning_rate": 9.612151440951334e-05, "loss": 0.0302, "step": 16990 }, { "grad_norm": 0.1910911500453949, "learning_rate": 9.611512678836506e-05, "loss": 0.0313, "step": 17000 }, { "grad_norm": 0.28431183099746704, "learning_rate": 9.610873412413584e-05, "loss": 0.0299, "step": 17010 }, { "grad_norm": 0.31066611409187317, "learning_rate": 9.610233641752476e-05, "loss": 0.0298, "step": 17020 }, { "grad_norm": 0.24266979098320007, "learning_rate": 9.609593366923151e-05, "loss": 0.0393, "step": 17030 }, { "grad_norm": 0.26999926567077637, "learning_rate": 9.608952587995625e-05, "loss": 0.0366, "step": 17040 }, { "grad_norm": 0.2572301924228668, "learning_rate": 9.608311305039972e-05, "loss": 0.0314, "step": 17050 }, { "grad_norm": 0.2549716830253601, "learning_rate": 9.607669518126326e-05, "loss": 0.0323, "step": 17060 }, { "grad_norm": 0.2579192817211151, "learning_rate": 9.607027227324866e-05, "loss": 0.0262, "step": 17070 }, { "grad_norm": 0.20931857824325562, "learning_rate": 9.606384432705837e-05, "loss": 0.0288, "step": 17080 }, { "grad_norm": 0.2115527093410492, "learning_rate": 9.60574113433953e-05, "loss": 0.0245, "step": 17090 }, { "grad_norm": 0.2789776921272278, "learning_rate": 9.6050973322963e-05, "loss": 0.0273, "step": 17100 }, { "grad_norm": 0.25645774602890015, "learning_rate": 9.604453026646547e-05, "loss": 0.029, "step": 17110 }, { "grad_norm": 0.22281724214553833, "learning_rate": 9.603808217460735e-05, "loss": 0.0234, "step": 17120 }, { "grad_norm": 0.2144767791032791, "learning_rate": 9.603162904809377e-05, "loss": 0.0259, "step": 17130 }, { "grad_norm": 0.29099318385124207, "learning_rate": 9.602517088763045e-05, "loss": 0.0263, "step": 17140 }, { "grad_norm": 0.22868208587169647, "learning_rate": 9.601870769392365e-05, "loss": 0.0246, "step": 17150 }, { "grad_norm": 0.3086695373058319, "learning_rate": 9.601223946768017e-05, "loss": 0.025, "step": 17160 }, { "grad_norm": 0.2906528115272522, "learning_rate": 9.600576620960734e-05, "loss": 0.0238, "step": 17170 }, { "grad_norm": 0.20298901200294495, "learning_rate": 9.599928792041308e-05, "loss": 0.025, "step": 17180 }, { "grad_norm": 0.24981652200222015, "learning_rate": 9.599280460080587e-05, "loss": 0.0237, "step": 17190 }, { "grad_norm": 0.21809163689613342, "learning_rate": 9.59863162514947e-05, "loss": 0.0246, "step": 17200 }, { "grad_norm": 0.24218051135540009, "learning_rate": 9.597982287318911e-05, "loss": 0.0247, "step": 17210 }, { "grad_norm": 0.22900883853435516, "learning_rate": 9.597332446659923e-05, "loss": 0.0281, "step": 17220 }, { "grad_norm": 0.44423484802246094, "learning_rate": 9.59668210324357e-05, "loss": 0.0256, "step": 17230 }, { "grad_norm": 0.2587476074695587, "learning_rate": 9.596031257140974e-05, "loss": 0.0277, "step": 17240 }, { "grad_norm": 0.4631190299987793, "learning_rate": 9.59537990842331e-05, "loss": 0.0273, "step": 17250 }, { "grad_norm": 0.2644234001636505, "learning_rate": 9.594728057161806e-05, "loss": 0.0257, "step": 17260 }, { "grad_norm": 0.168812558054924, "learning_rate": 9.594075703427752e-05, "loss": 0.0228, "step": 17270 }, { "grad_norm": 0.2568431496620178, "learning_rate": 9.593422847292486e-05, "loss": 0.0262, "step": 17280 }, { "grad_norm": 0.2660306990146637, "learning_rate": 9.592769488827402e-05, "loss": 0.0248, "step": 17290 }, { "grad_norm": 0.255607932806015, "learning_rate": 9.592115628103952e-05, "loss": 0.0285, "step": 17300 }, { "grad_norm": 0.23258627951145172, "learning_rate": 9.591461265193643e-05, "loss": 0.0305, "step": 17310 }, { "grad_norm": 0.2367270290851593, "learning_rate": 9.590806400168032e-05, "loss": 0.0304, "step": 17320 }, { "grad_norm": 0.21801415085792542, "learning_rate": 9.590151033098735e-05, "loss": 0.0272, "step": 17330 }, { "grad_norm": 0.2651680111885071, "learning_rate": 9.589495164057423e-05, "loss": 0.0256, "step": 17340 }, { "grad_norm": 0.23554648458957672, "learning_rate": 9.58883879311582e-05, "loss": 0.0325, "step": 17350 }, { "grad_norm": 0.2520216703414917, "learning_rate": 9.588181920345705e-05, "loss": 0.0288, "step": 17360 }, { "grad_norm": 0.24133636057376862, "learning_rate": 9.587524545818913e-05, "loss": 0.0279, "step": 17370 }, { "grad_norm": 0.21689631044864655, "learning_rate": 9.586866669607335e-05, "loss": 0.0268, "step": 17380 }, { "grad_norm": 0.2359756976366043, "learning_rate": 9.586208291782915e-05, "loss": 0.0263, "step": 17390 }, { "grad_norm": 0.22183099389076233, "learning_rate": 9.58554941241765e-05, "loss": 0.0302, "step": 17400 }, { "grad_norm": 0.1963128000497818, "learning_rate": 9.584890031583596e-05, "loss": 0.025, "step": 17410 }, { "grad_norm": 0.24121898412704468, "learning_rate": 9.584230149352861e-05, "loss": 0.0243, "step": 17420 }, { "grad_norm": 0.25055554509162903, "learning_rate": 9.58356976579761e-05, "loss": 0.0239, "step": 17430 }, { "grad_norm": 0.2936663031578064, "learning_rate": 9.58290888099006e-05, "loss": 0.0222, "step": 17440 }, { "grad_norm": 0.1981295347213745, "learning_rate": 9.582247495002486e-05, "loss": 0.0221, "step": 17450 }, { "grad_norm": 0.30270200967788696, "learning_rate": 9.581585607907214e-05, "loss": 0.0282, "step": 17460 }, { "grad_norm": 0.23906970024108887, "learning_rate": 9.580923219776628e-05, "loss": 0.0248, "step": 17470 }, { "grad_norm": 0.31291452050209045, "learning_rate": 9.580260330683167e-05, "loss": 0.0257, "step": 17480 }, { "grad_norm": 0.23236699402332306, "learning_rate": 9.579596940699322e-05, "loss": 0.0242, "step": 17490 }, { "grad_norm": 0.1880815476179123, "learning_rate": 9.578933049897643e-05, "loss": 0.0223, "step": 17500 }, { "grad_norm": 0.2403341829776764, "learning_rate": 9.578268658350728e-05, "loss": 0.0229, "step": 17510 }, { "grad_norm": 0.27449849247932434, "learning_rate": 9.577603766131235e-05, "loss": 0.0255, "step": 17520 }, { "grad_norm": 0.27091559767723083, "learning_rate": 9.576938373311878e-05, "loss": 0.0257, "step": 17530 }, { "grad_norm": 0.2851676642894745, "learning_rate": 9.576272479965421e-05, "loss": 0.0251, "step": 17540 }, { "grad_norm": 0.2197033315896988, "learning_rate": 9.575606086164687e-05, "loss": 0.0241, "step": 17550 }, { "grad_norm": 0.250887393951416, "learning_rate": 9.57493919198255e-05, "loss": 0.0286, "step": 17560 }, { "grad_norm": 0.2344929724931717, "learning_rate": 9.57427179749194e-05, "loss": 0.028, "step": 17570 }, { "grad_norm": 0.2376689463853836, "learning_rate": 9.573603902765846e-05, "loss": 0.0261, "step": 17580 }, { "grad_norm": 0.25110968947410583, "learning_rate": 9.572935507877304e-05, "loss": 0.0252, "step": 17590 }, { "grad_norm": 0.24709554016590118, "learning_rate": 9.57226661289941e-05, "loss": 0.0261, "step": 17600 }, { "grad_norm": 0.21388502418994904, "learning_rate": 9.571597217905315e-05, "loss": 0.0239, "step": 17610 }, { "grad_norm": 0.1759594976902008, "learning_rate": 9.57092732296822e-05, "loss": 0.0229, "step": 17620 }, { "grad_norm": 0.27239182591438293, "learning_rate": 9.570256928161385e-05, "loss": 0.0262, "step": 17630 }, { "grad_norm": 0.2602698504924774, "learning_rate": 9.569586033558126e-05, "loss": 0.0246, "step": 17640 }, { "grad_norm": 0.2842346429824829, "learning_rate": 9.568914639231807e-05, "loss": 0.0258, "step": 17650 }, { "grad_norm": 0.31904882192611694, "learning_rate": 9.568242745255852e-05, "loss": 0.0255, "step": 17660 }, { "grad_norm": 0.19638396799564362, "learning_rate": 9.567570351703739e-05, "loss": 0.0244, "step": 17670 }, { "grad_norm": 0.3031190037727356, "learning_rate": 9.566897458649001e-05, "loss": 0.0251, "step": 17680 }, { "grad_norm": 0.26040568947792053, "learning_rate": 9.566224066165221e-05, "loss": 0.0289, "step": 17690 }, { "grad_norm": 0.2196435183286667, "learning_rate": 9.565550174326043e-05, "loss": 0.0229, "step": 17700 }, { "grad_norm": 0.2315569967031479, "learning_rate": 9.564875783205162e-05, "loss": 0.0218, "step": 17710 }, { "grad_norm": 0.1784910410642624, "learning_rate": 9.564200892876328e-05, "loss": 0.0219, "step": 17720 }, { "grad_norm": 0.17819926142692566, "learning_rate": 9.563525503413348e-05, "loss": 0.0245, "step": 17730 }, { "grad_norm": 0.19785535335540771, "learning_rate": 9.562849614890079e-05, "loss": 0.02, "step": 17740 }, { "grad_norm": 0.28692057728767395, "learning_rate": 9.562173227380436e-05, "loss": 0.0241, "step": 17750 }, { "grad_norm": 0.2122054249048233, "learning_rate": 9.561496340958389e-05, "loss": 0.0205, "step": 17760 }, { "grad_norm": 0.21305477619171143, "learning_rate": 9.560818955697959e-05, "loss": 0.0222, "step": 17770 }, { "grad_norm": 0.2735503613948822, "learning_rate": 9.560141071673228e-05, "loss": 0.0218, "step": 17780 }, { "grad_norm": 0.23799580335617065, "learning_rate": 9.559462688958323e-05, "loss": 0.0229, "step": 17790 }, { "grad_norm": 0.17807726562023163, "learning_rate": 9.558783807627434e-05, "loss": 0.0216, "step": 17800 }, { "grad_norm": 0.20019780099391937, "learning_rate": 9.558104427754801e-05, "loss": 0.0225, "step": 17810 }, { "grad_norm": 0.28783127665519714, "learning_rate": 9.557424549414722e-05, "loss": 0.0246, "step": 17820 }, { "grad_norm": 0.20014256238937378, "learning_rate": 9.556744172681546e-05, "loss": 0.0228, "step": 17830 }, { "grad_norm": 0.24812497198581696, "learning_rate": 9.556063297629677e-05, "loss": 0.0243, "step": 17840 }, { "grad_norm": 0.22336681187152863, "learning_rate": 9.555381924333578e-05, "loss": 0.0251, "step": 17850 }, { "grad_norm": 0.27520841360092163, "learning_rate": 9.554700052867758e-05, "loss": 0.0232, "step": 17860 }, { "grad_norm": 0.22132886946201324, "learning_rate": 9.554017683306789e-05, "loss": 0.024, "step": 17870 }, { "grad_norm": 0.23015782237052917, "learning_rate": 9.553334815725294e-05, "loss": 0.0219, "step": 17880 }, { "grad_norm": 0.22252096235752106, "learning_rate": 9.552651450197949e-05, "loss": 0.0228, "step": 17890 }, { "grad_norm": 0.2394144982099533, "learning_rate": 9.551967586799486e-05, "loss": 0.024, "step": 17900 }, { "grad_norm": 0.17749622464179993, "learning_rate": 9.551283225604692e-05, "loss": 0.022, "step": 17910 }, { "grad_norm": 0.18634478747844696, "learning_rate": 9.550598366688406e-05, "loss": 0.024, "step": 17920 }, { "grad_norm": 0.20417670905590057, "learning_rate": 9.549913010125526e-05, "loss": 0.0227, "step": 17930 }, { "grad_norm": 0.28448382019996643, "learning_rate": 9.549227155990999e-05, "loss": 0.0224, "step": 17940 }, { "grad_norm": 0.20538149774074554, "learning_rate": 9.548540804359828e-05, "loss": 0.0226, "step": 17950 }, { "grad_norm": 0.23148572444915771, "learning_rate": 9.547853955307077e-05, "loss": 0.0216, "step": 17960 }, { "grad_norm": 0.21126361191272736, "learning_rate": 9.547166608907853e-05, "loss": 0.0239, "step": 17970 }, { "grad_norm": 0.20343679189682007, "learning_rate": 9.546478765237326e-05, "loss": 0.0195, "step": 17980 }, { "grad_norm": 0.23281598091125488, "learning_rate": 9.545790424370715e-05, "loss": 0.0207, "step": 17990 }, { "grad_norm": 0.33129045367240906, "learning_rate": 9.5451015863833e-05, "loss": 0.0242, "step": 18000 }, { "grad_norm": 0.22435641288757324, "learning_rate": 9.544412251350408e-05, "loss": 0.028, "step": 18010 }, { "grad_norm": 0.2541605234146118, "learning_rate": 9.543722419347422e-05, "loss": 0.0236, "step": 18020 }, { "grad_norm": 0.2619258761405945, "learning_rate": 9.543032090449788e-05, "loss": 0.0238, "step": 18030 }, { "grad_norm": 0.20156021416187286, "learning_rate": 9.542341264732992e-05, "loss": 0.024, "step": 18040 }, { "grad_norm": 0.19170555472373962, "learning_rate": 9.541649942272585e-05, "loss": 0.025, "step": 18050 }, { "grad_norm": 0.2880629003047943, "learning_rate": 9.54095812314417e-05, "loss": 0.0218, "step": 18060 }, { "grad_norm": 0.28581979870796204, "learning_rate": 9.540265807423401e-05, "loss": 0.0246, "step": 18070 }, { "grad_norm": 0.2712382376194, "learning_rate": 9.53957299518599e-05, "loss": 0.0242, "step": 18080 }, { "grad_norm": 0.25476035475730896, "learning_rate": 9.5388796865077e-05, "loss": 0.0321, "step": 18090 }, { "grad_norm": 0.35473743081092834, "learning_rate": 9.538185881464353e-05, "loss": 0.0286, "step": 18100 }, { "grad_norm": 0.22540295124053955, "learning_rate": 9.537491580131821e-05, "loss": 0.0274, "step": 18110 }, { "grad_norm": 0.24451731145381927, "learning_rate": 9.53679678258603e-05, "loss": 0.0245, "step": 18120 }, { "grad_norm": 0.2200918048620224, "learning_rate": 9.536101488902966e-05, "loss": 0.0245, "step": 18130 }, { "grad_norm": 0.27746838331222534, "learning_rate": 9.535405699158663e-05, "loss": 0.0301, "step": 18140 }, { "grad_norm": 0.2571420967578888, "learning_rate": 9.53470941342921e-05, "loss": 0.0248, "step": 18150 }, { "grad_norm": 0.22999975085258484, "learning_rate": 9.534012631790756e-05, "loss": 0.0235, "step": 18160 }, { "grad_norm": 0.28011974692344666, "learning_rate": 9.533315354319494e-05, "loss": 0.0304, "step": 18170 }, { "grad_norm": 0.21861916780471802, "learning_rate": 9.532617581091682e-05, "loss": 0.0286, "step": 18180 }, { "grad_norm": 0.2698538601398468, "learning_rate": 9.531919312183629e-05, "loss": 0.0255, "step": 18190 }, { "grad_norm": 0.20297442376613617, "learning_rate": 9.531220547671688e-05, "loss": 0.0238, "step": 18200 }, { "grad_norm": 0.191908061504364, "learning_rate": 9.530521287632285e-05, "loss": 0.0245, "step": 18210 }, { "grad_norm": 0.20754100382328033, "learning_rate": 9.529821532141884e-05, "loss": 0.0287, "step": 18220 }, { "grad_norm": 0.308383584022522, "learning_rate": 9.52912128127701e-05, "loss": 0.0245, "step": 18230 }, { "grad_norm": 0.19864560663700104, "learning_rate": 9.528420535114244e-05, "loss": 0.0246, "step": 18240 }, { "grad_norm": 0.23964634537696838, "learning_rate": 9.527719293730215e-05, "loss": 0.0246, "step": 18250 }, { "grad_norm": 0.3621634840965271, "learning_rate": 9.527017557201611e-05, "loss": 0.025, "step": 18260 }, { "grad_norm": 0.2134091705083847, "learning_rate": 9.526315325605176e-05, "loss": 0.0262, "step": 18270 }, { "grad_norm": 0.29215508699417114, "learning_rate": 9.525612599017699e-05, "loss": 0.0253, "step": 18280 }, { "grad_norm": 0.21514810621738434, "learning_rate": 9.524909377516033e-05, "loss": 0.0255, "step": 18290 }, { "grad_norm": 0.21709778904914856, "learning_rate": 9.524205661177081e-05, "loss": 0.0251, "step": 18300 }, { "grad_norm": 0.23174692690372467, "learning_rate": 9.523501450077801e-05, "loss": 0.0243, "step": 18310 }, { "grad_norm": 0.2482123225927353, "learning_rate": 9.522796744295202e-05, "loss": 0.0235, "step": 18320 }, { "grad_norm": 0.23398804664611816, "learning_rate": 9.522091543906352e-05, "loss": 0.0259, "step": 18330 }, { "grad_norm": 0.1731613278388977, "learning_rate": 9.521385848988369e-05, "loss": 0.0241, "step": 18340 }, { "grad_norm": 0.2569689154624939, "learning_rate": 9.520679659618428e-05, "loss": 0.0221, "step": 18350 }, { "grad_norm": 0.20631517469882965, "learning_rate": 9.519972975873754e-05, "loss": 0.0218, "step": 18360 }, { "grad_norm": 0.20200856029987335, "learning_rate": 9.519265797831633e-05, "loss": 0.0252, "step": 18370 }, { "grad_norm": 0.21890811622142792, "learning_rate": 9.518558125569399e-05, "loss": 0.0261, "step": 18380 }, { "grad_norm": 0.24819837510585785, "learning_rate": 9.517849959164442e-05, "loss": 0.0266, "step": 18390 }, { "grad_norm": 0.22844405472278595, "learning_rate": 9.517141298694205e-05, "loss": 0.0254, "step": 18400 }, { "grad_norm": 0.197190523147583, "learning_rate": 9.516432144236188e-05, "loss": 0.0284, "step": 18410 }, { "grad_norm": 0.20869025588035583, "learning_rate": 9.515722495867941e-05, "loss": 0.0226, "step": 18420 }, { "grad_norm": 0.2726140022277832, "learning_rate": 9.515012353667072e-05, "loss": 0.0256, "step": 18430 }, { "grad_norm": 0.18681257963180542, "learning_rate": 9.51430171771124e-05, "loss": 0.0202, "step": 18440 }, { "grad_norm": 0.19961538910865784, "learning_rate": 9.513590588078159e-05, "loss": 0.0239, "step": 18450 }, { "grad_norm": 0.19282373785972595, "learning_rate": 9.512878964845597e-05, "loss": 0.023, "step": 18460 }, { "grad_norm": 0.2890223264694214, "learning_rate": 9.512166848091377e-05, "loss": 0.022, "step": 18470 }, { "grad_norm": 0.2575094699859619, "learning_rate": 9.511454237893376e-05, "loss": 0.0229, "step": 18480 }, { "grad_norm": 0.19621820747852325, "learning_rate": 9.51074113432952e-05, "loss": 0.0231, "step": 18490 }, { "grad_norm": 0.1817408800125122, "learning_rate": 9.510027537477797e-05, "loss": 0.022, "step": 18500 }, { "grad_norm": 0.25401678681373596, "learning_rate": 9.509313447416242e-05, "loss": 0.024, "step": 18510 }, { "grad_norm": 0.24020782113075256, "learning_rate": 9.508598864222949e-05, "loss": 0.0241, "step": 18520 }, { "grad_norm": 0.2859041392803192, "learning_rate": 9.507883787976062e-05, "loss": 0.0242, "step": 18530 }, { "grad_norm": 0.19928216934204102, "learning_rate": 9.507168218753781e-05, "loss": 0.026, "step": 18540 }, { "grad_norm": 0.20764976739883423, "learning_rate": 9.506452156634362e-05, "loss": 0.0235, "step": 18550 }, { "grad_norm": 0.18357129395008087, "learning_rate": 9.505735601696109e-05, "loss": 0.0223, "step": 18560 }, { "grad_norm": 0.23457825183868408, "learning_rate": 9.505018554017385e-05, "loss": 0.0251, "step": 18570 }, { "grad_norm": 0.25535959005355835, "learning_rate": 9.504301013676604e-05, "loss": 0.021, "step": 18580 }, { "grad_norm": 0.32418984174728394, "learning_rate": 9.503582980752238e-05, "loss": 0.0229, "step": 18590 }, { "grad_norm": 0.2372564673423767, "learning_rate": 9.502864455322809e-05, "loss": 0.0219, "step": 18600 }, { "grad_norm": 0.20284400880336761, "learning_rate": 9.502145437466891e-05, "loss": 0.0234, "step": 18610 }, { "grad_norm": 0.25989609956741333, "learning_rate": 9.501425927263116e-05, "loss": 0.021, "step": 18620 }, { "grad_norm": 0.23050762712955475, "learning_rate": 9.500705924790172e-05, "loss": 0.0233, "step": 18630 }, { "grad_norm": 0.18021099269390106, "learning_rate": 9.499985430126794e-05, "loss": 0.0222, "step": 18640 }, { "grad_norm": 0.28068649768829346, "learning_rate": 9.499264443351775e-05, "loss": 0.0205, "step": 18650 }, { "grad_norm": 0.24038328230381012, "learning_rate": 9.498542964543961e-05, "loss": 0.023, "step": 18660 }, { "grad_norm": 0.20719867944717407, "learning_rate": 9.497820993782252e-05, "loss": 0.0183, "step": 18670 }, { "grad_norm": 0.2121955007314682, "learning_rate": 9.497098531145601e-05, "loss": 0.0235, "step": 18680 }, { "grad_norm": 0.26116228103637695, "learning_rate": 9.496375576713017e-05, "loss": 0.0235, "step": 18690 }, { "grad_norm": 0.203903928399086, "learning_rate": 9.49565213056356e-05, "loss": 0.0205, "step": 18700 }, { "grad_norm": 0.25796887278556824, "learning_rate": 9.494928192776342e-05, "loss": 0.0215, "step": 18710 }, { "grad_norm": 0.24072732031345367, "learning_rate": 9.494203763430538e-05, "loss": 0.0234, "step": 18720 }, { "grad_norm": 0.24226848781108856, "learning_rate": 9.493478842605366e-05, "loss": 0.0192, "step": 18730 }, { "grad_norm": 0.30090653896331787, "learning_rate": 9.492753430380105e-05, "loss": 0.0261, "step": 18740 }, { "grad_norm": 0.20015940070152283, "learning_rate": 9.492027526834083e-05, "loss": 0.0246, "step": 18750 }, { "grad_norm": 0.22694528102874756, "learning_rate": 9.491301132046684e-05, "loss": 0.0212, "step": 18760 }, { "grad_norm": 0.2602463960647583, "learning_rate": 9.490574246097345e-05, "loss": 0.0244, "step": 18770 }, { "grad_norm": 0.21441254019737244, "learning_rate": 9.48984686906556e-05, "loss": 0.0227, "step": 18780 }, { "grad_norm": 0.24215620756149292, "learning_rate": 9.489119001030871e-05, "loss": 0.0232, "step": 18790 }, { "grad_norm": 0.22831502556800842, "learning_rate": 9.488390642072878e-05, "loss": 0.0226, "step": 18800 }, { "grad_norm": 0.26763561367988586, "learning_rate": 9.48766179227123e-05, "loss": 0.0242, "step": 18810 }, { "grad_norm": 0.26833975315093994, "learning_rate": 9.486932451705636e-05, "loss": 0.0244, "step": 18820 }, { "grad_norm": 0.2403983473777771, "learning_rate": 9.486202620455857e-05, "loss": 0.0223, "step": 18830 }, { "grad_norm": 0.266262948513031, "learning_rate": 9.485472298601704e-05, "loss": 0.0253, "step": 18840 }, { "grad_norm": 0.39422324299812317, "learning_rate": 9.484741486223043e-05, "loss": 0.0259, "step": 18850 }, { "grad_norm": 0.311014860868454, "learning_rate": 9.484010183399797e-05, "loss": 0.0244, "step": 18860 }, { "grad_norm": 0.19375020265579224, "learning_rate": 9.483278390211938e-05, "loss": 0.0224, "step": 18870 }, { "grad_norm": 0.24484053254127502, "learning_rate": 9.482546106739496e-05, "loss": 0.0241, "step": 18880 }, { "grad_norm": 0.23880082368850708, "learning_rate": 9.48181333306255e-05, "loss": 0.0214, "step": 18890 }, { "grad_norm": 0.28885552287101746, "learning_rate": 9.481080069261237e-05, "loss": 0.0233, "step": 18900 }, { "grad_norm": 0.23176567256450653, "learning_rate": 9.480346315415745e-05, "loss": 0.0218, "step": 18910 }, { "grad_norm": 0.20703214406967163, "learning_rate": 9.479612071606314e-05, "loss": 0.024, "step": 18920 }, { "grad_norm": 0.1591760218143463, "learning_rate": 9.478877337913244e-05, "loss": 0.0215, "step": 18930 }, { "grad_norm": 0.1980855017900467, "learning_rate": 9.478142114416881e-05, "loss": 0.0204, "step": 18940 }, { "grad_norm": 0.24549995362758636, "learning_rate": 9.47740640119763e-05, "loss": 0.0235, "step": 18950 }, { "grad_norm": 0.2582508325576782, "learning_rate": 9.476670198335947e-05, "loss": 0.0231, "step": 18960 }, { "grad_norm": 0.21709169447422028, "learning_rate": 9.47593350591234e-05, "loss": 0.0242, "step": 18970 }, { "grad_norm": 0.24253539741039276, "learning_rate": 9.475196324007376e-05, "loss": 0.022, "step": 18980 }, { "grad_norm": 0.25800883769989014, "learning_rate": 9.474458652701669e-05, "loss": 0.0251, "step": 18990 }, { "grad_norm": 0.22084079682826996, "learning_rate": 9.473720492075892e-05, "loss": 0.0227, "step": 19000 }, { "grad_norm": 0.3670812249183655, "learning_rate": 9.472981842210768e-05, "loss": 0.0233, "step": 19010 }, { "grad_norm": 0.19825555384159088, "learning_rate": 9.472242703187074e-05, "loss": 0.0205, "step": 19020 }, { "grad_norm": 0.23894159495830536, "learning_rate": 9.471503075085643e-05, "loss": 0.0247, "step": 19030 }, { "grad_norm": 0.1556113064289093, "learning_rate": 9.470762957987359e-05, "loss": 0.0232, "step": 19040 }, { "grad_norm": 0.26331770420074463, "learning_rate": 9.470022351973158e-05, "loss": 0.0323, "step": 19050 }, { "grad_norm": 0.28775233030319214, "learning_rate": 9.469281257124034e-05, "loss": 0.0329, "step": 19060 }, { "grad_norm": 0.2861866056919098, "learning_rate": 9.46853967352103e-05, "loss": 0.0255, "step": 19070 }, { "grad_norm": 0.2811777591705322, "learning_rate": 9.467797601245246e-05, "loss": 0.0277, "step": 19080 }, { "grad_norm": 0.2806764543056488, "learning_rate": 9.467055040377834e-05, "loss": 0.0285, "step": 19090 }, { "grad_norm": 0.24100305140018463, "learning_rate": 9.466311990999999e-05, "loss": 0.0276, "step": 19100 }, { "grad_norm": 0.21970035135746002, "learning_rate": 9.465568453193e-05, "loss": 0.0282, "step": 19110 }, { "grad_norm": 0.2187330424785614, "learning_rate": 9.464824427038148e-05, "loss": 0.0264, "step": 19120 }, { "grad_norm": 0.3178470730781555, "learning_rate": 9.46407991261681e-05, "loss": 0.0249, "step": 19130 }, { "grad_norm": 0.2514428496360779, "learning_rate": 9.463334910010404e-05, "loss": 0.028, "step": 19140 }, { "grad_norm": 0.2373575121164322, "learning_rate": 9.462589419300403e-05, "loss": 0.0286, "step": 19150 }, { "grad_norm": 0.26273685693740845, "learning_rate": 9.461843440568333e-05, "loss": 0.025, "step": 19160 }, { "grad_norm": 0.2199968844652176, "learning_rate": 9.461096973895773e-05, "loss": 0.0213, "step": 19170 }, { "grad_norm": 0.2317422330379486, "learning_rate": 9.460350019364355e-05, "loss": 0.0227, "step": 19180 }, { "grad_norm": 0.2340826541185379, "learning_rate": 9.459602577055764e-05, "loss": 0.0243, "step": 19190 }, { "grad_norm": 0.22598052024841309, "learning_rate": 9.45885464705174e-05, "loss": 0.0284, "step": 19200 }, { "grad_norm": 0.24789363145828247, "learning_rate": 9.458106229434076e-05, "loss": 0.0223, "step": 19210 }, { "grad_norm": 0.15330353379249573, "learning_rate": 9.457357324284617e-05, "loss": 0.0246, "step": 19220 }, { "grad_norm": 0.19569972157478333, "learning_rate": 9.456607931685262e-05, "loss": 0.0234, "step": 19230 }, { "grad_norm": 0.18553955852985382, "learning_rate": 9.455858051717965e-05, "loss": 0.0227, "step": 19240 }, { "grad_norm": 0.23294295370578766, "learning_rate": 9.45510768446473e-05, "loss": 0.0249, "step": 19250 }, { "grad_norm": 0.20965245366096497, "learning_rate": 9.454356830007618e-05, "loss": 0.0229, "step": 19260 }, { "grad_norm": 0.19482097029685974, "learning_rate": 9.45360548842874e-05, "loss": 0.0188, "step": 19270 }, { "grad_norm": 0.17201191186904907, "learning_rate": 9.452853659810261e-05, "loss": 0.0236, "step": 19280 }, { "grad_norm": 0.22186900675296783, "learning_rate": 9.452101344234401e-05, "loss": 0.0225, "step": 19290 }, { "grad_norm": 0.21400852501392365, "learning_rate": 9.451348541783431e-05, "loss": 0.0226, "step": 19300 }, { "grad_norm": 0.3737814426422119, "learning_rate": 9.450595252539678e-05, "loss": 0.0221, "step": 19310 }, { "grad_norm": 0.21442317962646484, "learning_rate": 9.449841476585518e-05, "loss": 0.0236, "step": 19320 }, { "grad_norm": 0.31992870569229126, "learning_rate": 9.449087214003384e-05, "loss": 0.0225, "step": 19330 }, { "grad_norm": 0.22969722747802734, "learning_rate": 9.448332464875765e-05, "loss": 0.0236, "step": 19340 }, { "grad_norm": 0.18242527544498444, "learning_rate": 9.447577229285192e-05, "loss": 0.0216, "step": 19350 }, { "grad_norm": 0.19152985513210297, "learning_rate": 9.446821507314261e-05, "loss": 0.0215, "step": 19360 }, { "grad_norm": 0.28154096007347107, "learning_rate": 9.446065299045617e-05, "loss": 0.0254, "step": 19370 }, { "grad_norm": 0.26516735553741455, "learning_rate": 9.445308604561955e-05, "loss": 0.0227, "step": 19380 }, { "grad_norm": 0.23732760548591614, "learning_rate": 9.444551423946028e-05, "loss": 0.0248, "step": 19390 }, { "grad_norm": 0.28169506788253784, "learning_rate": 9.443793757280638e-05, "loss": 0.0245, "step": 19400 }, { "grad_norm": 0.24010014533996582, "learning_rate": 9.443035604648646e-05, "loss": 0.0247, "step": 19410 }, { "grad_norm": 0.23221731185913086, "learning_rate": 9.44227696613296e-05, "loss": 0.0223, "step": 19420 }, { "grad_norm": 0.19766223430633545, "learning_rate": 9.441517841816542e-05, "loss": 0.0209, "step": 19430 }, { "grad_norm": 0.2303171008825302, "learning_rate": 9.440758231782413e-05, "loss": 0.0258, "step": 19440 }, { "grad_norm": 0.23043619096279144, "learning_rate": 9.439998136113639e-05, "loss": 0.0186, "step": 19450 }, { "grad_norm": 0.5371993780136108, "learning_rate": 9.439237554893344e-05, "loss": 0.0219, "step": 19460 }, { "grad_norm": 0.21061116456985474, "learning_rate": 9.438476488204705e-05, "loss": 0.0219, "step": 19470 }, { "grad_norm": 0.21826571226119995, "learning_rate": 9.43771493613095e-05, "loss": 0.02, "step": 19480 }, { "grad_norm": 0.2528955042362213, "learning_rate": 9.436952898755362e-05, "loss": 0.0238, "step": 19490 }, { "grad_norm": 0.22221356630325317, "learning_rate": 9.436190376161276e-05, "loss": 0.0221, "step": 19500 }, { "grad_norm": 0.21864700317382812, "learning_rate": 9.43542736843208e-05, "loss": 0.0257, "step": 19510 }, { "grad_norm": 0.32535281777381897, "learning_rate": 9.434663875651216e-05, "loss": 0.0236, "step": 19520 }, { "grad_norm": 0.21850691735744476, "learning_rate": 9.433899897902177e-05, "loss": 0.0249, "step": 19530 }, { "grad_norm": 0.22716659307479858, "learning_rate": 9.433135435268511e-05, "loss": 0.0214, "step": 19540 }, { "grad_norm": 0.24109108746051788, "learning_rate": 9.432370487833819e-05, "loss": 0.0238, "step": 19550 }, { "grad_norm": 0.17862063646316528, "learning_rate": 9.431605055681756e-05, "loss": 0.023, "step": 19560 }, { "grad_norm": 0.2501940429210663, "learning_rate": 9.430839138896026e-05, "loss": 0.0243, "step": 19570 }, { "grad_norm": 0.18284861743450165, "learning_rate": 9.43007273756039e-05, "loss": 0.0243, "step": 19580 }, { "grad_norm": 0.2653256058692932, "learning_rate": 9.429305851758658e-05, "loss": 0.0232, "step": 19590 }, { "grad_norm": 0.34059762954711914, "learning_rate": 9.428538481574699e-05, "loss": 0.0254, "step": 19600 }, { "grad_norm": 0.1908068060874939, "learning_rate": 9.42777062709243e-05, "loss": 0.0233, "step": 19610 }, { "grad_norm": 0.17926208674907684, "learning_rate": 9.427002288395821e-05, "loss": 0.0214, "step": 19620 }, { "grad_norm": 0.21472570300102234, "learning_rate": 9.426233465568898e-05, "loss": 0.023, "step": 19630 }, { "grad_norm": 0.23382365703582764, "learning_rate": 9.42546415869574e-05, "loss": 0.0206, "step": 19640 }, { "grad_norm": 0.23441293835639954, "learning_rate": 9.424694367860473e-05, "loss": 0.0236, "step": 19650 }, { "grad_norm": 0.18576912581920624, "learning_rate": 9.423924093147284e-05, "loss": 0.0251, "step": 19660 }, { "grad_norm": 0.23771557211875916, "learning_rate": 9.423153334640407e-05, "loss": 0.026, "step": 19670 }, { "grad_norm": 0.19721437990665436, "learning_rate": 9.42238209242413e-05, "loss": 0.0269, "step": 19680 }, { "grad_norm": 0.2663930654525757, "learning_rate": 9.421610366582798e-05, "loss": 0.023, "step": 19690 }, { "grad_norm": 0.2350468933582306, "learning_rate": 9.420838157200803e-05, "loss": 0.0241, "step": 19700 }, { "grad_norm": 0.21061229705810547, "learning_rate": 9.420065464362594e-05, "loss": 0.0207, "step": 19710 }, { "grad_norm": 0.18582741916179657, "learning_rate": 9.419292288152673e-05, "loss": 0.022, "step": 19720 }, { "grad_norm": 0.3286854326725006, "learning_rate": 9.418518628655588e-05, "loss": 0.0229, "step": 19730 }, { "grad_norm": 0.22465547919273376, "learning_rate": 9.417744485955951e-05, "loss": 0.0212, "step": 19740 }, { "grad_norm": 0.281169056892395, "learning_rate": 9.41696986013842e-05, "loss": 0.0216, "step": 19750 }, { "grad_norm": 0.1737411618232727, "learning_rate": 9.416194751287705e-05, "loss": 0.0239, "step": 19760 }, { "grad_norm": 0.21484427154064178, "learning_rate": 9.415419159488572e-05, "loss": 0.0222, "step": 19770 }, { "grad_norm": 0.2484482377767563, "learning_rate": 9.414643084825837e-05, "loss": 0.0219, "step": 19780 }, { "grad_norm": 0.205261692404747, "learning_rate": 9.413866527384372e-05, "loss": 0.0248, "step": 19790 }, { "grad_norm": 0.2644059658050537, "learning_rate": 9.4130894872491e-05, "loss": 0.0246, "step": 19800 }, { "grad_norm": 0.2964570224285126, "learning_rate": 9.412311964504998e-05, "loss": 0.0257, "step": 19810 }, { "grad_norm": 0.22985579073429108, "learning_rate": 9.411533959237091e-05, "loss": 0.0217, "step": 19820 }, { "grad_norm": 0.3065703511238098, "learning_rate": 9.410755471530464e-05, "loss": 0.025, "step": 19830 }, { "grad_norm": 0.3983549475669861, "learning_rate": 9.40997650147025e-05, "loss": 0.0219, "step": 19840 }, { "grad_norm": 0.21155674755573273, "learning_rate": 9.409197049141637e-05, "loss": 0.0212, "step": 19850 }, { "grad_norm": 0.26723384857177734, "learning_rate": 9.408417114629863e-05, "loss": 0.0233, "step": 19860 }, { "grad_norm": 0.2097216546535492, "learning_rate": 9.40763669802022e-05, "loss": 0.0228, "step": 19870 }, { "grad_norm": 0.21127936244010925, "learning_rate": 9.406855799398056e-05, "loss": 0.0215, "step": 19880 }, { "grad_norm": 0.2368205487728119, "learning_rate": 9.406074418848767e-05, "loss": 0.0222, "step": 19890 }, { "grad_norm": 0.21015091240406036, "learning_rate": 9.405292556457805e-05, "loss": 0.0211, "step": 19900 }, { "grad_norm": 0.17958968877792358, "learning_rate": 9.404510212310671e-05, "loss": 0.0242, "step": 19910 }, { "grad_norm": 0.2580457031726837, "learning_rate": 9.403727386492924e-05, "loss": 0.0221, "step": 19920 }, { "grad_norm": 0.2029513269662857, "learning_rate": 9.40294407909017e-05, "loss": 0.0248, "step": 19930 }, { "grad_norm": 0.21960024535655975, "learning_rate": 9.40216029018807e-05, "loss": 0.0215, "step": 19940 }, { "grad_norm": 0.23210394382476807, "learning_rate": 9.401376019872338e-05, "loss": 0.0224, "step": 19950 }, { "grad_norm": 0.23420706391334534, "learning_rate": 9.400591268228746e-05, "loss": 0.0262, "step": 19960 }, { "grad_norm": 0.21315835416316986, "learning_rate": 9.399806035343106e-05, "loss": 0.0196, "step": 19970 }, { "grad_norm": 0.2454032152891159, "learning_rate": 9.399020321301294e-05, "loss": 0.021, "step": 19980 }, { "grad_norm": 0.3682987689971924, "learning_rate": 9.398234126189234e-05, "loss": 0.0203, "step": 19990 }, { "grad_norm": 0.19246704876422882, "learning_rate": 9.397447450092902e-05, "loss": 0.0226, "step": 20000 }, { "grad_norm": 0.1993972659111023, "learning_rate": 9.39666029309833e-05, "loss": 0.0209, "step": 20010 }, { "grad_norm": 0.18179617822170258, "learning_rate": 9.395872655291596e-05, "loss": 0.0222, "step": 20020 }, { "grad_norm": 0.2612220048904419, "learning_rate": 9.395084536758838e-05, "loss": 0.0249, "step": 20030 }, { "grad_norm": 0.1670226752758026, "learning_rate": 9.394295937586243e-05, "loss": 0.0235, "step": 20040 }, { "grad_norm": 0.21388624608516693, "learning_rate": 9.393506857860052e-05, "loss": 0.0205, "step": 20050 }, { "grad_norm": 0.22102776169776917, "learning_rate": 9.392717297666555e-05, "loss": 0.02, "step": 20060 }, { "grad_norm": 0.22942394018173218, "learning_rate": 9.391927257092101e-05, "loss": 0.0214, "step": 20070 }, { "grad_norm": 0.23049378395080566, "learning_rate": 9.391136736223085e-05, "loss": 0.0227, "step": 20080 }, { "grad_norm": 0.2007390707731247, "learning_rate": 9.390345735145956e-05, "loss": 0.0203, "step": 20090 }, { "grad_norm": 0.21076379716396332, "learning_rate": 9.389554253947219e-05, "loss": 0.0234, "step": 20100 }, { "grad_norm": 0.3157179355621338, "learning_rate": 9.388762292713428e-05, "loss": 0.024, "step": 20110 }, { "grad_norm": 0.24189195036888123, "learning_rate": 9.38796985153119e-05, "loss": 0.0235, "step": 20120 }, { "grad_norm": 0.2401948720216751, "learning_rate": 9.387176930487169e-05, "loss": 0.0227, "step": 20130 }, { "grad_norm": 0.20154733955860138, "learning_rate": 9.386383529668072e-05, "loss": 0.0218, "step": 20140 }, { "grad_norm": 0.18057964742183685, "learning_rate": 9.385589649160669e-05, "loss": 0.0209, "step": 20150 }, { "grad_norm": 0.23764075338840485, "learning_rate": 9.384795289051775e-05, "loss": 0.0241, "step": 20160 }, { "grad_norm": 0.206182599067688, "learning_rate": 9.384000449428261e-05, "loss": 0.0231, "step": 20170 }, { "grad_norm": 0.18606294691562653, "learning_rate": 9.383205130377048e-05, "loss": 0.0224, "step": 20180 }, { "grad_norm": 0.1727084517478943, "learning_rate": 9.382409331985114e-05, "loss": 0.0216, "step": 20190 }, { "grad_norm": 0.1778259575366974, "learning_rate": 9.381613054339482e-05, "loss": 0.0219, "step": 20200 }, { "grad_norm": 0.2131895124912262, "learning_rate": 9.380816297527235e-05, "loss": 0.0244, "step": 20210 }, { "grad_norm": 0.29365625977516174, "learning_rate": 9.380019061635506e-05, "loss": 0.0262, "step": 20220 }, { "grad_norm": 0.1805582046508789, "learning_rate": 9.379221346751474e-05, "loss": 0.0255, "step": 20230 }, { "grad_norm": 0.2364986538887024, "learning_rate": 9.378423152962382e-05, "loss": 0.0237, "step": 20240 }, { "grad_norm": 0.2893465757369995, "learning_rate": 9.377624480355517e-05, "loss": 0.0207, "step": 20250 }, { "grad_norm": 0.22532498836517334, "learning_rate": 9.376825329018219e-05, "loss": 0.0222, "step": 20260 }, { "grad_norm": 0.2293819636106491, "learning_rate": 9.376025699037884e-05, "loss": 0.0213, "step": 20270 }, { "grad_norm": 0.20858393609523773, "learning_rate": 9.37522559050196e-05, "loss": 0.0197, "step": 20280 }, { "grad_norm": 0.2483440637588501, "learning_rate": 9.37442500349794e-05, "loss": 0.02, "step": 20290 }, { "grad_norm": 0.1613052487373352, "learning_rate": 9.373623938113381e-05, "loss": 0.0221, "step": 20300 }, { "grad_norm": 0.21579021215438843, "learning_rate": 9.372822394435883e-05, "loss": 0.0207, "step": 20310 }, { "grad_norm": 0.1767842024564743, "learning_rate": 9.372020372553102e-05, "loss": 0.0187, "step": 20320 }, { "grad_norm": 0.2042369395494461, "learning_rate": 9.371217872552746e-05, "loss": 0.0203, "step": 20330 }, { "grad_norm": 0.21541251242160797, "learning_rate": 9.370414894522576e-05, "loss": 0.0205, "step": 20340 }, { "grad_norm": 0.1782296895980835, "learning_rate": 9.369611438550406e-05, "loss": 0.02, "step": 20350 }, { "grad_norm": 0.21148312091827393, "learning_rate": 9.368807504724095e-05, "loss": 0.022, "step": 20360 }, { "grad_norm": 0.23223572969436646, "learning_rate": 9.368003093131565e-05, "loss": 0.0217, "step": 20370 }, { "grad_norm": 0.20440691709518433, "learning_rate": 9.367198203860785e-05, "loss": 0.0247, "step": 20380 }, { "grad_norm": 0.25028112530708313, "learning_rate": 9.366392836999774e-05, "loss": 0.0205, "step": 20390 }, { "grad_norm": 0.297803670167923, "learning_rate": 9.365586992636607e-05, "loss": 0.0216, "step": 20400 }, { "grad_norm": 0.1656206250190735, "learning_rate": 9.364780670859412e-05, "loss": 0.0188, "step": 20410 }, { "grad_norm": 0.30747121572494507, "learning_rate": 9.363973871756364e-05, "loss": 0.02, "step": 20420 }, { "grad_norm": 0.16705480217933655, "learning_rate": 9.363166595415696e-05, "loss": 0.0197, "step": 20430 }, { "grad_norm": 0.24536150693893433, "learning_rate": 9.362358841925686e-05, "loss": 0.0217, "step": 20440 }, { "grad_norm": 0.20412464439868927, "learning_rate": 9.361550611374674e-05, "loss": 0.0208, "step": 20450 }, { "grad_norm": 0.2442627251148224, "learning_rate": 9.360741903851043e-05, "loss": 0.021, "step": 20460 }, { "grad_norm": 0.22728808224201202, "learning_rate": 9.359932719443236e-05, "loss": 0.021, "step": 20470 }, { "grad_norm": 0.21560899913311005, "learning_rate": 9.35912305823974e-05, "loss": 0.0231, "step": 20480 }, { "grad_norm": 0.22689472138881683, "learning_rate": 9.358312920329101e-05, "loss": 0.0228, "step": 20490 }, { "grad_norm": 0.20354938507080078, "learning_rate": 9.357502305799914e-05, "loss": 0.0204, "step": 20500 }, { "grad_norm": 0.2251058965921402, "learning_rate": 9.356691214740824e-05, "loss": 0.0204, "step": 20510 }, { "grad_norm": 0.230509951710701, "learning_rate": 9.355879647240535e-05, "loss": 0.0199, "step": 20520 }, { "grad_norm": 0.19820566475391388, "learning_rate": 9.355067603387798e-05, "loss": 0.0218, "step": 20530 }, { "grad_norm": 0.3463982939720154, "learning_rate": 9.354255083271412e-05, "loss": 0.0223, "step": 20540 }, { "grad_norm": 0.41795673966407776, "learning_rate": 9.353442086980239e-05, "loss": 0.0219, "step": 20550 }, { "grad_norm": 0.2319822460412979, "learning_rate": 9.352628614603185e-05, "loss": 0.021, "step": 20560 }, { "grad_norm": 0.20989090204238892, "learning_rate": 9.351814666229209e-05, "loss": 0.0227, "step": 20570 }, { "grad_norm": 0.2651185989379883, "learning_rate": 9.351000241947324e-05, "loss": 0.0238, "step": 20580 }, { "grad_norm": 0.17567254602909088, "learning_rate": 9.350185341846594e-05, "loss": 0.0221, "step": 20590 }, { "grad_norm": 0.29587242007255554, "learning_rate": 9.349369966016134e-05, "loss": 0.0223, "step": 20600 }, { "grad_norm": 0.4107820689678192, "learning_rate": 9.348554114545117e-05, "loss": 0.0218, "step": 20610 }, { "grad_norm": 0.2871105372905731, "learning_rate": 9.347737787522758e-05, "loss": 0.0235, "step": 20620 }, { "grad_norm": 0.18723827600479126, "learning_rate": 9.346920985038332e-05, "loss": 0.0229, "step": 20630 }, { "grad_norm": 0.3715973496437073, "learning_rate": 9.346103707181162e-05, "loss": 0.0255, "step": 20640 }, { "grad_norm": 0.247826486825943, "learning_rate": 9.345285954040626e-05, "loss": 0.0435, "step": 20650 }, { "grad_norm": 0.2871733605861664, "learning_rate": 9.34446772570615e-05, "loss": 0.0336, "step": 20660 }, { "grad_norm": 0.19711987674236298, "learning_rate": 9.343649022267214e-05, "loss": 0.0259, "step": 20670 }, { "grad_norm": 0.30458706617355347, "learning_rate": 9.342829843813353e-05, "loss": 0.0258, "step": 20680 }, { "grad_norm": 0.18475162982940674, "learning_rate": 9.342010190434149e-05, "loss": 0.0252, "step": 20690 }, { "grad_norm": 0.19165046513080597, "learning_rate": 9.34119006221924e-05, "loss": 0.0244, "step": 20700 }, { "grad_norm": 0.2440756857395172, "learning_rate": 9.340369459258313e-05, "loss": 0.0272, "step": 20710 }, { "grad_norm": 0.24294456839561462, "learning_rate": 9.339548381641106e-05, "loss": 0.0268, "step": 20720 }, { "grad_norm": 0.24046410620212555, "learning_rate": 9.338726829457413e-05, "loss": 0.0226, "step": 20730 }, { "grad_norm": 0.17870326340198517, "learning_rate": 9.337904802797078e-05, "loss": 0.021, "step": 20740 }, { "grad_norm": 0.24042664468288422, "learning_rate": 9.337082301749993e-05, "loss": 0.0218, "step": 20750 }, { "grad_norm": 0.1799677312374115, "learning_rate": 9.336259326406109e-05, "loss": 0.0194, "step": 20760 }, { "grad_norm": 0.20270612835884094, "learning_rate": 9.335435876855427e-05, "loss": 0.0226, "step": 20770 }, { "grad_norm": 0.21565745770931244, "learning_rate": 9.334611953187994e-05, "loss": 0.0221, "step": 20780 }, { "grad_norm": 0.18403151631355286, "learning_rate": 9.333787555493914e-05, "loss": 0.0209, "step": 20790 }, { "grad_norm": 0.17169922590255737, "learning_rate": 9.332962683863345e-05, "loss": 0.0201, "step": 20800 }, { "grad_norm": 0.2306191623210907, "learning_rate": 9.332137338386489e-05, "loss": 0.0218, "step": 20810 }, { "grad_norm": 0.3000849783420563, "learning_rate": 9.33131151915361e-05, "loss": 0.0198, "step": 20820 }, { "grad_norm": 0.2062983363866806, "learning_rate": 9.330485226255012e-05, "loss": 0.0213, "step": 20830 }, { "grad_norm": 0.18636232614517212, "learning_rate": 9.329658459781061e-05, "loss": 0.0178, "step": 20840 }, { "grad_norm": 0.19569319486618042, "learning_rate": 9.328831219822172e-05, "loss": 0.0211, "step": 20850 }, { "grad_norm": 0.1677950620651245, "learning_rate": 9.328003506468808e-05, "loss": 0.0213, "step": 20860 }, { "grad_norm": 0.16143257915973663, "learning_rate": 9.327175319811488e-05, "loss": 0.0212, "step": 20870 }, { "grad_norm": 0.2022370994091034, "learning_rate": 9.326346659940781e-05, "loss": 0.0192, "step": 20880 }, { "grad_norm": 0.2527800500392914, "learning_rate": 9.325517526947308e-05, "loss": 0.0218, "step": 20890 }, { "grad_norm": 0.19249841570854187, "learning_rate": 9.32468792092174e-05, "loss": 0.0212, "step": 20900 }, { "grad_norm": 0.2056146115064621, "learning_rate": 9.323857841954803e-05, "loss": 0.0224, "step": 20910 }, { "grad_norm": 0.16283977031707764, "learning_rate": 9.323027290137276e-05, "loss": 0.0215, "step": 20920 }, { "grad_norm": 0.21923470497131348, "learning_rate": 9.322196265559981e-05, "loss": 0.02, "step": 20930 }, { "grad_norm": 0.28022268414497375, "learning_rate": 9.321364768313803e-05, "loss": 0.0219, "step": 20940 }, { "grad_norm": 0.22431495785713196, "learning_rate": 9.32053279848967e-05, "loss": 0.0227, "step": 20950 }, { "grad_norm": 0.216247096657753, "learning_rate": 9.319700356178567e-05, "loss": 0.0225, "step": 20960 }, { "grad_norm": 0.1678852140903473, "learning_rate": 9.318867441471527e-05, "loss": 0.0221, "step": 20970 }, { "grad_norm": 0.15887758135795593, "learning_rate": 9.318034054459637e-05, "loss": 0.0215, "step": 20980 }, { "grad_norm": 0.1730155348777771, "learning_rate": 9.317200195234034e-05, "loss": 0.0202, "step": 20990 }, { "grad_norm": 0.2503337860107422, "learning_rate": 9.316365863885909e-05, "loss": 0.0198, "step": 21000 }, { "grad_norm": 0.2323831021785736, "learning_rate": 9.315531060506502e-05, "loss": 0.022, "step": 21010 }, { "grad_norm": 0.23372524976730347, "learning_rate": 9.314695785187108e-05, "loss": 0.0235, "step": 21020 }, { "grad_norm": 0.2702455222606659, "learning_rate": 9.313860038019069e-05, "loss": 0.0217, "step": 21030 }, { "grad_norm": 0.20339550077915192, "learning_rate": 9.313023819093782e-05, "loss": 0.0226, "step": 21040 }, { "grad_norm": 0.24720163643360138, "learning_rate": 9.312187128502695e-05, "loss": 0.0211, "step": 21050 }, { "grad_norm": 0.191057026386261, "learning_rate": 9.311349966337307e-05, "loss": 0.0201, "step": 21060 }, { "grad_norm": 0.15627746284008026, "learning_rate": 9.310512332689169e-05, "loss": 0.0193, "step": 21070 }, { "grad_norm": 0.17290569841861725, "learning_rate": 9.309674227649883e-05, "loss": 0.0219, "step": 21080 }, { "grad_norm": 0.22754867374897003, "learning_rate": 9.308835651311103e-05, "loss": 0.0201, "step": 21090 }, { "grad_norm": 0.20101730525493622, "learning_rate": 9.307996603764533e-05, "loss": 0.0201, "step": 21100 }, { "grad_norm": 0.1910969614982605, "learning_rate": 9.307157085101932e-05, "loss": 0.02, "step": 21110 }, { "grad_norm": 0.19034305214881897, "learning_rate": 9.306317095415109e-05, "loss": 0.0203, "step": 21120 }, { "grad_norm": 0.18735076487064362, "learning_rate": 9.305476634795922e-05, "loss": 0.0211, "step": 21130 }, { "grad_norm": 0.21834036707878113, "learning_rate": 9.304635703336284e-05, "loss": 0.0199, "step": 21140 }, { "grad_norm": 0.22035276889801025, "learning_rate": 9.303794301128157e-05, "loss": 0.0194, "step": 21150 }, { "grad_norm": 0.132268026471138, "learning_rate": 9.302952428263555e-05, "loss": 0.0188, "step": 21160 }, { "grad_norm": 0.16549596190452576, "learning_rate": 9.302110084834545e-05, "loss": 0.0195, "step": 21170 }, { "grad_norm": 0.2779596745967865, "learning_rate": 9.301267270933245e-05, "loss": 0.0191, "step": 21180 }, { "grad_norm": 0.19379039108753204, "learning_rate": 9.300423986651823e-05, "loss": 0.0217, "step": 21190 }, { "grad_norm": 0.43335169553756714, "learning_rate": 9.299580232082501e-05, "loss": 0.0219, "step": 21200 }, { "grad_norm": 0.24005503952503204, "learning_rate": 9.298736007317547e-05, "loss": 0.0266, "step": 21210 }, { "grad_norm": 0.20101094245910645, "learning_rate": 9.297891312449288e-05, "loss": 0.0248, "step": 21220 }, { "grad_norm": 0.20226295292377472, "learning_rate": 9.297046147570094e-05, "loss": 0.0235, "step": 21230 }, { "grad_norm": 0.23895952105522156, "learning_rate": 9.296200512772396e-05, "loss": 0.0249, "step": 21240 }, { "grad_norm": 0.20744016766548157, "learning_rate": 9.295354408148668e-05, "loss": 0.02, "step": 21250 }, { "grad_norm": 0.20162558555603027, "learning_rate": 9.294507833791441e-05, "loss": 0.0205, "step": 21260 }, { "grad_norm": 0.19392669200897217, "learning_rate": 9.293660789793295e-05, "loss": 0.0198, "step": 21270 }, { "grad_norm": 0.2472435086965561, "learning_rate": 9.292813276246858e-05, "loss": 0.0214, "step": 21280 }, { "grad_norm": 0.17778649926185608, "learning_rate": 9.291965293244816e-05, "loss": 0.0207, "step": 21290 }, { "grad_norm": 0.204337477684021, "learning_rate": 9.291116840879904e-05, "loss": 0.0194, "step": 21300 }, { "grad_norm": 0.1728149652481079, "learning_rate": 9.290267919244904e-05, "loss": 0.0239, "step": 21310 }, { "grad_norm": 0.21005772054195404, "learning_rate": 9.289418528432655e-05, "loss": 0.0202, "step": 21320 }, { "grad_norm": 0.30090805888175964, "learning_rate": 9.288568668536045e-05, "loss": 0.0225, "step": 21330 }, { "grad_norm": 0.2978401482105255, "learning_rate": 9.287718339648013e-05, "loss": 0.0224, "step": 21340 }, { "grad_norm": 0.2019490748643875, "learning_rate": 9.28686754186155e-05, "loss": 0.0226, "step": 21350 }, { "grad_norm": 0.16967016458511353, "learning_rate": 9.286016275269698e-05, "loss": 0.0215, "step": 21360 }, { "grad_norm": 0.18470047414302826, "learning_rate": 9.285164539965551e-05, "loss": 0.0175, "step": 21370 }, { "grad_norm": 0.2149476259946823, "learning_rate": 9.284312336042251e-05, "loss": 0.0204, "step": 21380 }, { "grad_norm": 0.16692695021629333, "learning_rate": 9.283459663592996e-05, "loss": 0.02, "step": 21390 }, { "grad_norm": 0.2026868313550949, "learning_rate": 9.282606522711033e-05, "loss": 0.0225, "step": 21400 }, { "grad_norm": 0.20519274473190308, "learning_rate": 9.281752913489657e-05, "loss": 0.0229, "step": 21410 }, { "grad_norm": 0.20941613614559174, "learning_rate": 9.280898836022222e-05, "loss": 0.0229, "step": 21420 }, { "grad_norm": 0.26072263717651367, "learning_rate": 9.280044290402126e-05, "loss": 0.0217, "step": 21430 }, { "grad_norm": 0.2527777850627899, "learning_rate": 9.279189276722821e-05, "loss": 0.0226, "step": 21440 }, { "grad_norm": 0.172867089509964, "learning_rate": 9.278333795077812e-05, "loss": 0.0195, "step": 21450 }, { "grad_norm": 0.24705743789672852, "learning_rate": 9.27747784556065e-05, "loss": 0.0197, "step": 21460 }, { "grad_norm": 0.23305629193782806, "learning_rate": 9.276621428264942e-05, "loss": 0.024, "step": 21470 }, { "grad_norm": 0.1990155577659607, "learning_rate": 9.275764543284345e-05, "loss": 0.0219, "step": 21480 }, { "grad_norm": 0.25153762102127075, "learning_rate": 9.274907190712566e-05, "loss": 0.023, "step": 21490 }, { "grad_norm": 0.22397424280643463, "learning_rate": 9.274049370643363e-05, "loss": 0.0202, "step": 21500 }, { "grad_norm": 0.19916662573814392, "learning_rate": 9.273191083170547e-05, "loss": 0.0201, "step": 21510 }, { "grad_norm": 0.186665877699852, "learning_rate": 9.27233232838798e-05, "loss": 0.0225, "step": 21520 }, { "grad_norm": 0.15942074358463287, "learning_rate": 9.27147310638957e-05, "loss": 0.0198, "step": 21530 }, { "grad_norm": 0.2272411435842514, "learning_rate": 9.270613417269286e-05, "loss": 0.0201, "step": 21540 }, { "grad_norm": 0.26590579748153687, "learning_rate": 9.269753261121138e-05, "loss": 0.0203, "step": 21550 }, { "grad_norm": 0.19009707868099213, "learning_rate": 9.268892638039194e-05, "loss": 0.022, "step": 21560 }, { "grad_norm": 0.18108756840229034, "learning_rate": 9.268031548117569e-05, "loss": 0.0209, "step": 21570 }, { "grad_norm": 0.2120889276266098, "learning_rate": 9.26716999145043e-05, "loss": 0.022, "step": 21580 }, { "grad_norm": 0.1973283439874649, "learning_rate": 9.266307968131998e-05, "loss": 0.0192, "step": 21590 }, { "grad_norm": 0.14105023443698883, "learning_rate": 9.26544547825654e-05, "loss": 0.0206, "step": 21600 }, { "grad_norm": 0.19469669461250305, "learning_rate": 9.264582521918376e-05, "loss": 0.0229, "step": 21610 }, { "grad_norm": 0.38097578287124634, "learning_rate": 9.263719099211881e-05, "loss": 0.0197, "step": 21620 }, { "grad_norm": 0.1909530907869339, "learning_rate": 9.262855210231476e-05, "loss": 0.0204, "step": 21630 }, { "grad_norm": 0.1640319526195526, "learning_rate": 9.261990855071633e-05, "loss": 0.0203, "step": 21640 }, { "grad_norm": 0.19166940450668335, "learning_rate": 9.261126033826878e-05, "loss": 0.0224, "step": 21650 }, { "grad_norm": 0.18261517584323883, "learning_rate": 9.260260746591786e-05, "loss": 0.0192, "step": 21660 }, { "grad_norm": 0.18376019597053528, "learning_rate": 9.259394993460985e-05, "loss": 0.0203, "step": 21670 }, { "grad_norm": 0.21710410714149475, "learning_rate": 9.258528774529151e-05, "loss": 0.0221, "step": 21680 }, { "grad_norm": 0.23863454163074493, "learning_rate": 9.257662089891013e-05, "loss": 0.0203, "step": 21690 }, { "grad_norm": 0.3277105689048767, "learning_rate": 9.25679493964135e-05, "loss": 0.0235, "step": 21700 }, { "grad_norm": 0.17132768034934998, "learning_rate": 9.255927323874994e-05, "loss": 0.0206, "step": 21710 }, { "grad_norm": 0.19619522988796234, "learning_rate": 9.255059242686822e-05, "loss": 0.0225, "step": 21720 }, { "grad_norm": 0.2719091773033142, "learning_rate": 9.254190696171769e-05, "loss": 0.0215, "step": 21730 }, { "grad_norm": 0.18169431388378143, "learning_rate": 9.25332168442482e-05, "loss": 0.02, "step": 21740 }, { "grad_norm": 0.23046819865703583, "learning_rate": 9.252452207541004e-05, "loss": 0.0241, "step": 21750 }, { "grad_norm": 0.2166685312986374, "learning_rate": 9.251582265615409e-05, "loss": 0.0215, "step": 21760 }, { "grad_norm": 0.28037166595458984, "learning_rate": 9.250711858743169e-05, "loss": 0.0205, "step": 21770 }, { "grad_norm": 0.2702064514160156, "learning_rate": 9.24984098701947e-05, "loss": 0.0242, "step": 21780 }, { "grad_norm": 0.22153060138225555, "learning_rate": 9.248969650539552e-05, "loss": 0.0174, "step": 21790 }, { "grad_norm": 0.21100616455078125, "learning_rate": 9.2480978493987e-05, "loss": 0.0219, "step": 21800 }, { "grad_norm": 0.22289443016052246, "learning_rate": 9.247225583692256e-05, "loss": 0.0284, "step": 21810 }, { "grad_norm": 0.24425889551639557, "learning_rate": 9.246352853515607e-05, "loss": 0.0229, "step": 21820 }, { "grad_norm": 0.235299214720726, "learning_rate": 9.245479658964194e-05, "loss": 0.0232, "step": 21830 }, { "grad_norm": 0.17727550864219666, "learning_rate": 9.244606000133507e-05, "loss": 0.0215, "step": 21840 }, { "grad_norm": 0.2327592819929123, "learning_rate": 9.24373187711909e-05, "loss": 0.0241, "step": 21850 }, { "grad_norm": 0.21310767531394958, "learning_rate": 9.242857290016537e-05, "loss": 0.0252, "step": 21860 }, { "grad_norm": 0.2788141071796417, "learning_rate": 9.241982238921488e-05, "loss": 0.0244, "step": 21870 }, { "grad_norm": 0.17589840292930603, "learning_rate": 9.24110672392964e-05, "loss": 0.0225, "step": 21880 }, { "grad_norm": 0.2067006230354309, "learning_rate": 9.240230745136737e-05, "loss": 0.0208, "step": 21890 }, { "grad_norm": 0.17446671426296234, "learning_rate": 9.239354302638575e-05, "loss": 0.018, "step": 21900 }, { "grad_norm": 0.2235146462917328, "learning_rate": 9.238477396531e-05, "loss": 0.0199, "step": 21910 }, { "grad_norm": 0.24204696714878082, "learning_rate": 9.23760002690991e-05, "loss": 0.0194, "step": 21920 }, { "grad_norm": 0.2968597114086151, "learning_rate": 9.236722193871252e-05, "loss": 0.0235, "step": 21930 }, { "grad_norm": 0.24446460604667664, "learning_rate": 9.235843897511023e-05, "loss": 0.0218, "step": 21940 }, { "grad_norm": 0.15477193892002106, "learning_rate": 9.234965137925276e-05, "loss": 0.0217, "step": 21950 }, { "grad_norm": 0.2554436922073364, "learning_rate": 9.234085915210108e-05, "loss": 0.0205, "step": 21960 }, { "grad_norm": 0.242889866232872, "learning_rate": 9.23320622946167e-05, "loss": 0.0224, "step": 21970 }, { "grad_norm": 0.21918216347694397, "learning_rate": 9.232326080776163e-05, "loss": 0.0256, "step": 21980 }, { "grad_norm": 0.2446543574333191, "learning_rate": 9.23144546924984e-05, "loss": 0.0264, "step": 21990 }, { "grad_norm": 0.34579551219940186, "learning_rate": 9.230564394979e-05, "loss": 0.0255, "step": 22000 }, { "grad_norm": 0.16953998804092407, "learning_rate": 9.22968285806e-05, "loss": 0.0219, "step": 22010 }, { "grad_norm": 0.3250987231731415, "learning_rate": 9.228800858589242e-05, "loss": 0.026, "step": 22020 }, { "grad_norm": 0.302810400724411, "learning_rate": 9.227918396663179e-05, "loss": 0.0232, "step": 22030 }, { "grad_norm": 0.25382640957832336, "learning_rate": 9.227035472378319e-05, "loss": 0.0226, "step": 22040 }, { "grad_norm": 0.2399957776069641, "learning_rate": 9.226152085831213e-05, "loss": 0.0214, "step": 22050 }, { "grad_norm": 0.35961687564849854, "learning_rate": 9.22526823711847e-05, "loss": 0.0218, "step": 22060 }, { "grad_norm": 0.2098866105079651, "learning_rate": 9.224383926336745e-05, "loss": 0.0238, "step": 22070 }, { "grad_norm": 0.22382842004299164, "learning_rate": 9.223499153582744e-05, "loss": 0.0233, "step": 22080 }, { "grad_norm": 0.1553637534379959, "learning_rate": 9.222613918953226e-05, "loss": 0.0221, "step": 22090 }, { "grad_norm": 0.20855988562107086, "learning_rate": 9.221728222544999e-05, "loss": 0.0203, "step": 22100 }, { "grad_norm": 0.18145659565925598, "learning_rate": 9.22084206445492e-05, "loss": 0.0222, "step": 22110 }, { "grad_norm": 0.19515933096408844, "learning_rate": 9.2199554447799e-05, "loss": 0.0198, "step": 22120 }, { "grad_norm": 0.21507132053375244, "learning_rate": 9.219068363616897e-05, "loss": 0.0228, "step": 22130 }, { "grad_norm": 0.20309266448020935, "learning_rate": 9.218180821062919e-05, "loss": 0.0205, "step": 22140 }, { "grad_norm": 0.22144140303134918, "learning_rate": 9.21729281721503e-05, "loss": 0.021, "step": 22150 }, { "grad_norm": 0.2778385281562805, "learning_rate": 9.216404352170339e-05, "loss": 0.0217, "step": 22160 }, { "grad_norm": 0.27477824687957764, "learning_rate": 9.215515426026007e-05, "loss": 0.0206, "step": 22170 }, { "grad_norm": 0.2106906920671463, "learning_rate": 9.214626038879246e-05, "loss": 0.0219, "step": 22180 }, { "grad_norm": 0.2335902452468872, "learning_rate": 9.21373619082732e-05, "loss": 0.021, "step": 22190 }, { "grad_norm": 0.25115764141082764, "learning_rate": 9.212845881967535e-05, "loss": 0.0244, "step": 22200 }, { "grad_norm": 0.26908746361732483, "learning_rate": 9.211955112397262e-05, "loss": 0.0254, "step": 22210 }, { "grad_norm": 0.17511612176895142, "learning_rate": 9.211063882213909e-05, "loss": 0.0205, "step": 22220 }, { "grad_norm": 0.21265414357185364, "learning_rate": 9.210172191514942e-05, "loss": 0.0226, "step": 22230 }, { "grad_norm": 0.25208234786987305, "learning_rate": 9.209280040397874e-05, "loss": 0.0204, "step": 22240 }, { "grad_norm": 0.18002949655056, "learning_rate": 9.208387428960268e-05, "loss": 0.0192, "step": 22250 }, { "grad_norm": 0.1412469446659088, "learning_rate": 9.20749435729974e-05, "loss": 0.0182, "step": 22260 }, { "grad_norm": 0.20440128445625305, "learning_rate": 9.206600825513957e-05, "loss": 0.0203, "step": 22270 }, { "grad_norm": 0.17443229258060455, "learning_rate": 9.20570683370063e-05, "loss": 0.0209, "step": 22280 }, { "grad_norm": 0.18869391083717346, "learning_rate": 9.204812381957528e-05, "loss": 0.0178, "step": 22290 }, { "grad_norm": 0.19052380323410034, "learning_rate": 9.203917470382465e-05, "loss": 0.0199, "step": 22300 }, { "grad_norm": 0.21369419991970062, "learning_rate": 9.203022099073309e-05, "loss": 0.0187, "step": 22310 }, { "grad_norm": 0.19143255054950714, "learning_rate": 9.202126268127976e-05, "loss": 0.0182, "step": 22320 }, { "grad_norm": 0.18851788341999054, "learning_rate": 9.20122997764443e-05, "loss": 0.0205, "step": 22330 }, { "grad_norm": 0.21402618288993835, "learning_rate": 9.200333227720692e-05, "loss": 0.0206, "step": 22340 }, { "grad_norm": 0.20725689828395844, "learning_rate": 9.199436018454826e-05, "loss": 0.0213, "step": 22350 }, { "grad_norm": 0.22304004430770874, "learning_rate": 9.198538349944952e-05, "loss": 0.0183, "step": 22360 }, { "grad_norm": 0.18336988985538483, "learning_rate": 9.197640222289234e-05, "loss": 0.0195, "step": 22370 }, { "grad_norm": 0.3124880790710449, "learning_rate": 9.196741635585895e-05, "loss": 0.0186, "step": 22380 }, { "grad_norm": 0.23018261790275574, "learning_rate": 9.195842589933199e-05, "loss": 0.0238, "step": 22390 }, { "grad_norm": 0.1761118471622467, "learning_rate": 9.194943085429466e-05, "loss": 0.0212, "step": 22400 }, { "grad_norm": 0.17367134988307953, "learning_rate": 9.194043122173065e-05, "loss": 0.0213, "step": 22410 }, { "grad_norm": 0.19547255337238312, "learning_rate": 9.193142700262413e-05, "loss": 0.0225, "step": 22420 }, { "grad_norm": 0.21846596896648407, "learning_rate": 9.192241819795979e-05, "loss": 0.0191, "step": 22430 }, { "grad_norm": 0.15139788389205933, "learning_rate": 9.191340480872284e-05, "loss": 0.02, "step": 22440 }, { "grad_norm": 0.27566856145858765, "learning_rate": 9.190438683589895e-05, "loss": 0.0202, "step": 22450 }, { "grad_norm": 0.20727206766605377, "learning_rate": 9.189536428047432e-05, "loss": 0.0214, "step": 22460 }, { "grad_norm": 0.19865189492702484, "learning_rate": 9.188633714343564e-05, "loss": 0.0198, "step": 22470 }, { "grad_norm": 0.16148817539215088, "learning_rate": 9.18773054257701e-05, "loss": 0.0169, "step": 22480 }, { "grad_norm": 0.2243918925523758, "learning_rate": 9.18682691284654e-05, "loss": 0.016, "step": 22490 }, { "grad_norm": 0.19005431234836578, "learning_rate": 9.185922825250974e-05, "loss": 0.0197, "step": 22500 }, { "grad_norm": 0.21002568304538727, "learning_rate": 9.185018279889181e-05, "loss": 0.0199, "step": 22510 }, { "grad_norm": 0.195374995470047, "learning_rate": 9.184113276860082e-05, "loss": 0.023, "step": 22520 }, { "grad_norm": 0.24187560379505157, "learning_rate": 9.183207816262645e-05, "loss": 0.0175, "step": 22530 }, { "grad_norm": 0.21128477156162262, "learning_rate": 9.182301898195891e-05, "loss": 0.0215, "step": 22540 }, { "grad_norm": 0.24554158747196198, "learning_rate": 9.181395522758889e-05, "loss": 0.0203, "step": 22550 }, { "grad_norm": 0.1936846226453781, "learning_rate": 9.180488690050759e-05, "loss": 0.0194, "step": 22560 }, { "grad_norm": 0.21728357672691345, "learning_rate": 9.179581400170671e-05, "loss": 0.0212, "step": 22570 }, { "grad_norm": 0.18319086730480194, "learning_rate": 9.178673653217845e-05, "loss": 0.0193, "step": 22580 }, { "grad_norm": 0.156431183218956, "learning_rate": 9.177765449291551e-05, "loss": 0.0195, "step": 22590 }, { "grad_norm": 0.18035274744033813, "learning_rate": 9.176856788491109e-05, "loss": 0.018, "step": 22600 }, { "grad_norm": 0.21025395393371582, "learning_rate": 9.175947670915887e-05, "loss": 0.0224, "step": 22610 }, { "grad_norm": 0.2848845422267914, "learning_rate": 9.175038096665309e-05, "loss": 0.0216, "step": 22620 }, { "grad_norm": 0.2767973244190216, "learning_rate": 9.17412806583884e-05, "loss": 0.0213, "step": 22630 }, { "grad_norm": 0.27046865224838257, "learning_rate": 9.173217578536002e-05, "loss": 0.0203, "step": 22640 }, { "grad_norm": 0.19701239466667175, "learning_rate": 9.172306634856362e-05, "loss": 0.0207, "step": 22650 }, { "grad_norm": 0.1817750185728073, "learning_rate": 9.171395234899545e-05, "loss": 0.0211, "step": 22660 }, { "grad_norm": 0.14712409675121307, "learning_rate": 9.170483378765214e-05, "loss": 0.0207, "step": 22670 }, { "grad_norm": 0.1857726275920868, "learning_rate": 9.169571066553091e-05, "loss": 0.0188, "step": 22680 }, { "grad_norm": 0.18935349583625793, "learning_rate": 9.168658298362946e-05, "loss": 0.0208, "step": 22690 }, { "grad_norm": 0.19077393412590027, "learning_rate": 9.167745074294598e-05, "loss": 0.019, "step": 22700 }, { "grad_norm": 0.18938378989696503, "learning_rate": 9.166831394447913e-05, "loss": 0.0209, "step": 22710 }, { "grad_norm": 0.18216882646083832, "learning_rate": 9.165917258922812e-05, "loss": 0.0213, "step": 22720 }, { "grad_norm": 0.1495978981256485, "learning_rate": 9.165002667819262e-05, "loss": 0.0189, "step": 22730 }, { "grad_norm": 0.3102926015853882, "learning_rate": 9.164087621237282e-05, "loss": 0.0217, "step": 22740 }, { "grad_norm": 0.20305126905441284, "learning_rate": 9.163172119276942e-05, "loss": 0.0212, "step": 22750 }, { "grad_norm": 0.17341727018356323, "learning_rate": 9.162256162038358e-05, "loss": 0.0208, "step": 22760 }, { "grad_norm": 0.17602814733982086, "learning_rate": 9.161339749621698e-05, "loss": 0.02, "step": 22770 }, { "grad_norm": 0.2197459191083908, "learning_rate": 9.160422882127177e-05, "loss": 0.021, "step": 22780 }, { "grad_norm": 0.19483940303325653, "learning_rate": 9.159505559655069e-05, "loss": 0.0213, "step": 22790 }, { "grad_norm": 0.2195805460214615, "learning_rate": 9.158587782305684e-05, "loss": 0.0174, "step": 22800 }, { "grad_norm": 0.19437983632087708, "learning_rate": 9.157669550179391e-05, "loss": 0.0201, "step": 22810 }, { "grad_norm": 0.1429910957813263, "learning_rate": 9.156750863376609e-05, "loss": 0.0172, "step": 22820 }, { "grad_norm": 0.22534571588039398, "learning_rate": 9.155831721997801e-05, "loss": 0.0165, "step": 22830 }, { "grad_norm": 0.17948707938194275, "learning_rate": 9.154912126143484e-05, "loss": 0.0176, "step": 22840 }, { "grad_norm": 0.19686156511306763, "learning_rate": 9.153992075914224e-05, "loss": 0.0183, "step": 22850 }, { "grad_norm": 0.2223711460828781, "learning_rate": 9.153071571410635e-05, "loss": 0.0181, "step": 22860 }, { "grad_norm": 0.20820651948451996, "learning_rate": 9.152150612733384e-05, "loss": 0.0225, "step": 22870 }, { "grad_norm": 0.20606817305088043, "learning_rate": 9.151229199983184e-05, "loss": 0.022, "step": 22880 }, { "grad_norm": 0.17037218809127808, "learning_rate": 9.150307333260802e-05, "loss": 0.0197, "step": 22890 }, { "grad_norm": 0.2442602664232254, "learning_rate": 9.149385012667048e-05, "loss": 0.0209, "step": 22900 }, { "grad_norm": 0.18019980192184448, "learning_rate": 9.148462238302788e-05, "loss": 0.0194, "step": 22910 }, { "grad_norm": 0.28294670581817627, "learning_rate": 9.147539010268936e-05, "loss": 0.0213, "step": 22920 }, { "grad_norm": 0.16851945221424103, "learning_rate": 9.14661532866645e-05, "loss": 0.0217, "step": 22930 }, { "grad_norm": 0.15878865122795105, "learning_rate": 9.145691193596348e-05, "loss": 0.0197, "step": 22940 }, { "grad_norm": 0.16364529728889465, "learning_rate": 9.144766605159691e-05, "loss": 0.0186, "step": 22950 }, { "grad_norm": 0.2021270990371704, "learning_rate": 9.14384156345759e-05, "loss": 0.0204, "step": 22960 }, { "grad_norm": 0.1690753847360611, "learning_rate": 9.142916068591204e-05, "loss": 0.0185, "step": 22970 }, { "grad_norm": 0.20583704113960266, "learning_rate": 9.141990120661746e-05, "loss": 0.0192, "step": 22980 }, { "grad_norm": 0.220866858959198, "learning_rate": 9.141063719770475e-05, "loss": 0.02, "step": 22990 }, { "grad_norm": 0.125491201877594, "learning_rate": 9.140136866018704e-05, "loss": 0.0186, "step": 23000 }, { "grad_norm": 0.1977064162492752, "learning_rate": 9.139209559507788e-05, "loss": 0.0183, "step": 23010 }, { "grad_norm": 0.21298450231552124, "learning_rate": 9.13828180033914e-05, "loss": 0.0197, "step": 23020 }, { "grad_norm": 0.20182664692401886, "learning_rate": 9.137353588614212e-05, "loss": 0.0196, "step": 23030 }, { "grad_norm": 0.17606589198112488, "learning_rate": 9.136424924434519e-05, "loss": 0.0191, "step": 23040 }, { "grad_norm": 0.2507498264312744, "learning_rate": 9.135495807901615e-05, "loss": 0.0213, "step": 23050 }, { "grad_norm": 0.1805453896522522, "learning_rate": 9.134566239117108e-05, "loss": 0.0189, "step": 23060 }, { "grad_norm": 0.1965552419424057, "learning_rate": 9.13363621818265e-05, "loss": 0.0177, "step": 23070 }, { "grad_norm": 0.1582871824502945, "learning_rate": 9.132705745199953e-05, "loss": 0.0182, "step": 23080 }, { "grad_norm": 0.18104484677314758, "learning_rate": 9.131774820270768e-05, "loss": 0.0182, "step": 23090 }, { "grad_norm": 0.1921590119600296, "learning_rate": 9.130843443496901e-05, "loss": 0.0209, "step": 23100 }, { "grad_norm": 0.20579907298088074, "learning_rate": 9.129911614980206e-05, "loss": 0.0253, "step": 23110 }, { "grad_norm": 0.19558444619178772, "learning_rate": 9.128979334822584e-05, "loss": 0.022, "step": 23120 }, { "grad_norm": 0.1756492704153061, "learning_rate": 9.128046603125992e-05, "loss": 0.0206, "step": 23130 }, { "grad_norm": 0.23087337613105774, "learning_rate": 9.12711341999243e-05, "loss": 0.0193, "step": 23140 }, { "grad_norm": 0.17813585698604584, "learning_rate": 9.12617978552395e-05, "loss": 0.0232, "step": 23150 }, { "grad_norm": 0.2036418467760086, "learning_rate": 9.12524569982265e-05, "loss": 0.02, "step": 23160 }, { "grad_norm": 0.24967432022094727, "learning_rate": 9.124311162990684e-05, "loss": 0.0189, "step": 23170 }, { "grad_norm": 0.21857434511184692, "learning_rate": 9.12337617513025e-05, "loss": 0.0195, "step": 23180 }, { "grad_norm": 0.28165656328201294, "learning_rate": 9.122440736343596e-05, "loss": 0.0191, "step": 23190 }, { "grad_norm": 0.19334420561790466, "learning_rate": 9.12150484673302e-05, "loss": 0.0194, "step": 23200 }, { "grad_norm": 0.18684612214565277, "learning_rate": 9.120568506400873e-05, "loss": 0.0167, "step": 23210 }, { "grad_norm": 0.1825951188802719, "learning_rate": 9.119631715449548e-05, "loss": 0.0192, "step": 23220 }, { "grad_norm": 0.26870742440223694, "learning_rate": 9.118694473981493e-05, "loss": 0.0202, "step": 23230 }, { "grad_norm": 0.23557206988334656, "learning_rate": 9.117756782099203e-05, "loss": 0.0209, "step": 23240 }, { "grad_norm": 0.16136391460895538, "learning_rate": 9.11681863990522e-05, "loss": 0.0195, "step": 23250 }, { "grad_norm": 0.22940325736999512, "learning_rate": 9.115880047502142e-05, "loss": 0.0203, "step": 23260 }, { "grad_norm": 0.20558705925941467, "learning_rate": 9.114941004992609e-05, "loss": 0.0181, "step": 23270 }, { "grad_norm": 0.1955414116382599, "learning_rate": 9.114001512479317e-05, "loss": 0.0207, "step": 23280 }, { "grad_norm": 0.16754591464996338, "learning_rate": 9.113061570065003e-05, "loss": 0.0179, "step": 23290 }, { "grad_norm": 0.22821784019470215, "learning_rate": 9.112121177852459e-05, "loss": 0.0179, "step": 23300 }, { "grad_norm": 0.21072529256343842, "learning_rate": 9.111180335944527e-05, "loss": 0.0187, "step": 23310 }, { "grad_norm": 0.2887144982814789, "learning_rate": 9.110239044444093e-05, "loss": 0.0205, "step": 23320 }, { "grad_norm": 0.2444588840007782, "learning_rate": 9.109297303454099e-05, "loss": 0.0189, "step": 23330 }, { "grad_norm": 0.1803542524576187, "learning_rate": 9.108355113077526e-05, "loss": 0.0215, "step": 23340 }, { "grad_norm": 0.29915663599967957, "learning_rate": 9.107412473417419e-05, "loss": 0.0177, "step": 23350 }, { "grad_norm": 0.19401617348194122, "learning_rate": 9.106469384576858e-05, "loss": 0.0193, "step": 23360 }, { "grad_norm": 0.28838077187538147, "learning_rate": 9.105525846658978e-05, "loss": 0.0208, "step": 23370 }, { "grad_norm": 0.180415078997612, "learning_rate": 9.104581859766965e-05, "loss": 0.0194, "step": 23380 }, { "grad_norm": 0.20154829323291779, "learning_rate": 9.10363742400405e-05, "loss": 0.0187, "step": 23390 }, { "grad_norm": 0.21925552189350128, "learning_rate": 9.102692539473518e-05, "loss": 0.0171, "step": 23400 }, { "grad_norm": 0.18933852016925812, "learning_rate": 9.101747206278697e-05, "loss": 0.0183, "step": 23410 }, { "grad_norm": 0.17961063981056213, "learning_rate": 9.100801424522968e-05, "loss": 0.0208, "step": 23420 }, { "grad_norm": 0.1597280353307724, "learning_rate": 9.099855194309762e-05, "loss": 0.0209, "step": 23430 }, { "grad_norm": 0.21632367372512817, "learning_rate": 9.098908515742554e-05, "loss": 0.0187, "step": 23440 }, { "grad_norm": 0.2155858725309372, "learning_rate": 9.097961388924873e-05, "loss": 0.0188, "step": 23450 }, { "grad_norm": 0.19298377633094788, "learning_rate": 9.097013813960298e-05, "loss": 0.0212, "step": 23460 }, { "grad_norm": 0.16678576171398163, "learning_rate": 9.09606579095245e-05, "loss": 0.0204, "step": 23470 }, { "grad_norm": 0.21577173471450806, "learning_rate": 9.095117320005008e-05, "loss": 0.0194, "step": 23480 }, { "grad_norm": 0.19340699911117554, "learning_rate": 9.094168401221691e-05, "loss": 0.0174, "step": 23490 }, { "grad_norm": 0.24062150716781616, "learning_rate": 9.093219034706273e-05, "loss": 0.0204, "step": 23500 }, { "grad_norm": 0.15552109479904175, "learning_rate": 9.092269220562577e-05, "loss": 0.0188, "step": 23510 }, { "grad_norm": 0.17149735987186432, "learning_rate": 9.09131895889447e-05, "loss": 0.0201, "step": 23520 }, { "grad_norm": 0.19286271929740906, "learning_rate": 9.090368249805873e-05, "loss": 0.0195, "step": 23530 }, { "grad_norm": 0.16929592192173004, "learning_rate": 9.089417093400754e-05, "loss": 0.0231, "step": 23540 }, { "grad_norm": 0.2010231465101242, "learning_rate": 9.088465489783131e-05, "loss": 0.0195, "step": 23550 }, { "grad_norm": 0.1945628821849823, "learning_rate": 9.087513439057068e-05, "loss": 0.0205, "step": 23560 }, { "grad_norm": 0.1729268729686737, "learning_rate": 9.08656094132668e-05, "loss": 0.0204, "step": 23570 }, { "grad_norm": 0.20446574687957764, "learning_rate": 9.085607996696134e-05, "loss": 0.0182, "step": 23580 }, { "grad_norm": 0.20868434011936188, "learning_rate": 9.084654605269639e-05, "loss": 0.0178, "step": 23590 }, { "grad_norm": 0.20912307500839233, "learning_rate": 9.083700767151457e-05, "loss": 0.0187, "step": 23600 }, { "grad_norm": 0.2165943831205368, "learning_rate": 9.082746482445898e-05, "loss": 0.0184, "step": 23610 }, { "grad_norm": 0.20594660937786102, "learning_rate": 9.081791751257325e-05, "loss": 0.0198, "step": 23620 }, { "grad_norm": 0.2135678082704544, "learning_rate": 9.080836573690142e-05, "loss": 0.0199, "step": 23630 }, { "grad_norm": 0.22579820454120636, "learning_rate": 9.079880949848805e-05, "loss": 0.0191, "step": 23640 }, { "grad_norm": 0.16116581857204437, "learning_rate": 9.078924879837822e-05, "loss": 0.0201, "step": 23650 }, { "grad_norm": 0.19458329677581787, "learning_rate": 9.077968363761747e-05, "loss": 0.0225, "step": 23660 }, { "grad_norm": 0.20231975615024567, "learning_rate": 9.077011401725182e-05, "loss": 0.0195, "step": 23670 }, { "grad_norm": 0.24658052623271942, "learning_rate": 9.07605399383278e-05, "loss": 0.0161, "step": 23680 }, { "grad_norm": 0.20147864520549774, "learning_rate": 9.075096140189243e-05, "loss": 0.0173, "step": 23690 }, { "grad_norm": 0.1971912384033203, "learning_rate": 9.074137840899318e-05, "loss": 0.0167, "step": 23700 }, { "grad_norm": 0.16269826889038086, "learning_rate": 9.073179096067804e-05, "loss": 0.0173, "step": 23710 }, { "grad_norm": 0.2024509310722351, "learning_rate": 9.072219905799549e-05, "loss": 0.0189, "step": 23720 }, { "grad_norm": 0.19734355807304382, "learning_rate": 9.071260270199447e-05, "loss": 0.0185, "step": 23730 }, { "grad_norm": 0.14982178807258606, "learning_rate": 9.070300189372441e-05, "loss": 0.0174, "step": 23740 }, { "grad_norm": 0.21441790461540222, "learning_rate": 9.069339663423528e-05, "loss": 0.0193, "step": 23750 }, { "grad_norm": 0.1457238346338272, "learning_rate": 9.068378692457747e-05, "loss": 0.02, "step": 23760 }, { "grad_norm": 0.18482626974582672, "learning_rate": 9.067417276580189e-05, "loss": 0.0172, "step": 23770 }, { "grad_norm": 0.22304488718509674, "learning_rate": 9.066455415895993e-05, "loss": 0.0195, "step": 23780 }, { "grad_norm": 0.21433019638061523, "learning_rate": 9.065493110510346e-05, "loss": 0.0181, "step": 23790 }, { "grad_norm": 0.21713820099830627, "learning_rate": 9.064530360528484e-05, "loss": 0.0175, "step": 23800 }, { "grad_norm": 0.2001868039369583, "learning_rate": 9.063567166055695e-05, "loss": 0.0183, "step": 23810 }, { "grad_norm": 0.2326505482196808, "learning_rate": 9.062603527197308e-05, "loss": 0.0183, "step": 23820 }, { "grad_norm": 0.20727427303791046, "learning_rate": 9.06163944405871e-05, "loss": 0.0197, "step": 23830 }, { "grad_norm": 0.22819718718528748, "learning_rate": 9.060674916745327e-05, "loss": 0.0191, "step": 23840 }, { "grad_norm": 0.19441574811935425, "learning_rate": 9.05970994536264e-05, "loss": 0.0181, "step": 23850 }, { "grad_norm": 0.2140989899635315, "learning_rate": 9.05874453001618e-05, "loss": 0.018, "step": 23860 }, { "grad_norm": 0.17176415026187897, "learning_rate": 9.057778670811517e-05, "loss": 0.0178, "step": 23870 }, { "grad_norm": 0.1854410618543625, "learning_rate": 9.056812367854281e-05, "loss": 0.0164, "step": 23880 }, { "grad_norm": 0.16025076806545258, "learning_rate": 9.055845621250143e-05, "loss": 0.0163, "step": 23890 }, { "grad_norm": 0.16231535375118256, "learning_rate": 9.054878431104825e-05, "loss": 0.02, "step": 23900 }, { "grad_norm": 0.1315878927707672, "learning_rate": 9.0539107975241e-05, "loss": 0.0175, "step": 23910 }, { "grad_norm": 0.21492359042167664, "learning_rate": 9.052942720613784e-05, "loss": 0.0169, "step": 23920 }, { "grad_norm": 0.21582241356372833, "learning_rate": 9.051974200479745e-05, "loss": 0.019, "step": 23930 }, { "grad_norm": 0.21941491961479187, "learning_rate": 9.051005237227901e-05, "loss": 0.0181, "step": 23940 }, { "grad_norm": 0.16780279576778412, "learning_rate": 9.050035830964215e-05, "loss": 0.0202, "step": 23950 }, { "grad_norm": 0.1648034155368805, "learning_rate": 9.049065981794698e-05, "loss": 0.0203, "step": 23960 }, { "grad_norm": 0.20885369181632996, "learning_rate": 9.048095689825414e-05, "loss": 0.0211, "step": 23970 }, { "grad_norm": 0.21405358612537384, "learning_rate": 9.047124955162472e-05, "loss": 0.0191, "step": 23980 }, { "grad_norm": 0.16932456195354462, "learning_rate": 9.046153777912028e-05, "loss": 0.0185, "step": 23990 }, { "grad_norm": 0.21587224304676056, "learning_rate": 9.045182158180292e-05, "loss": 0.0169, "step": 24000 }, { "grad_norm": 0.17148953676223755, "learning_rate": 9.044210096073516e-05, "loss": 0.0186, "step": 24010 }, { "grad_norm": 0.23413561284542084, "learning_rate": 9.043237591698004e-05, "loss": 0.0174, "step": 24020 }, { "grad_norm": 0.19001656770706177, "learning_rate": 9.04226464516011e-05, "loss": 0.0161, "step": 24030 }, { "grad_norm": 0.20503875613212585, "learning_rate": 9.041291256566229e-05, "loss": 0.0158, "step": 24040 }, { "grad_norm": 0.2122953087091446, "learning_rate": 9.040317426022814e-05, "loss": 0.0168, "step": 24050 }, { "grad_norm": 0.22064809501171112, "learning_rate": 9.03934315363636e-05, "loss": 0.0186, "step": 24060 }, { "grad_norm": 0.14466503262519836, "learning_rate": 9.038368439513409e-05, "loss": 0.0201, "step": 24070 }, { "grad_norm": 0.19877545535564423, "learning_rate": 9.03739328376056e-05, "loss": 0.0208, "step": 24080 }, { "grad_norm": 0.26931658387184143, "learning_rate": 9.036417686484451e-05, "loss": 0.0194, "step": 24090 }, { "grad_norm": 0.21453435719013214, "learning_rate": 9.035441647791773e-05, "loss": 0.021, "step": 24100 }, { "grad_norm": 0.18373961746692657, "learning_rate": 9.034465167789263e-05, "loss": 0.018, "step": 24110 }, { "grad_norm": 0.1721847951412201, "learning_rate": 9.033488246583706e-05, "loss": 0.0165, "step": 24120 }, { "grad_norm": 0.2970373332500458, "learning_rate": 9.032510884281941e-05, "loss": 0.0199, "step": 24130 }, { "grad_norm": 0.15794865787029266, "learning_rate": 9.031533080990848e-05, "loss": 0.018, "step": 24140 }, { "grad_norm": 0.21884123980998993, "learning_rate": 9.030554836817358e-05, "loss": 0.0161, "step": 24150 }, { "grad_norm": 0.18679599463939667, "learning_rate": 9.029576151868451e-05, "loss": 0.0206, "step": 24160 }, { "grad_norm": 0.17286401987075806, "learning_rate": 9.028597026251155e-05, "loss": 0.0196, "step": 24170 }, { "grad_norm": 0.18668390810489655, "learning_rate": 9.027617460072547e-05, "loss": 0.0184, "step": 24180 }, { "grad_norm": 0.1954338699579239, "learning_rate": 9.026637453439745e-05, "loss": 0.0175, "step": 24190 }, { "grad_norm": 0.19320900738239288, "learning_rate": 9.025657006459927e-05, "loss": 0.021, "step": 24200 }, { "grad_norm": 0.17037498950958252, "learning_rate": 9.024676119240311e-05, "loss": 0.0189, "step": 24210 }, { "grad_norm": 0.16532640159130096, "learning_rate": 9.023694791888166e-05, "loss": 0.0186, "step": 24220 }, { "grad_norm": 0.17718154191970825, "learning_rate": 9.022713024510808e-05, "loss": 0.0179, "step": 24230 }, { "grad_norm": 0.1748380959033966, "learning_rate": 9.021730817215601e-05, "loss": 0.0167, "step": 24240 }, { "grad_norm": 0.17165683209896088, "learning_rate": 9.02074817010996e-05, "loss": 0.0178, "step": 24250 }, { "grad_norm": 0.18398529291152954, "learning_rate": 9.019765083301342e-05, "loss": 0.0183, "step": 24260 }, { "grad_norm": 0.18418444693088531, "learning_rate": 9.01878155689726e-05, "loss": 0.0169, "step": 24270 }, { "grad_norm": 0.1799398958683014, "learning_rate": 9.017797591005268e-05, "loss": 0.0195, "step": 24280 }, { "grad_norm": 0.22816893458366394, "learning_rate": 9.016813185732972e-05, "loss": 0.0318, "step": 24290 }, { "grad_norm": 0.23606760799884796, "learning_rate": 9.015828341188027e-05, "loss": 0.0266, "step": 24300 }, { "grad_norm": 0.49988555908203125, "learning_rate": 9.01484305747813e-05, "loss": 0.0281, "step": 24310 }, { "grad_norm": 0.31571680307388306, "learning_rate": 9.013857334711033e-05, "loss": 0.0333, "step": 24320 }, { "grad_norm": 0.27277612686157227, "learning_rate": 9.012871172994534e-05, "loss": 0.0372, "step": 24330 }, { "grad_norm": 0.2863160967826843, "learning_rate": 9.011884572436476e-05, "loss": 0.0356, "step": 24340 }, { "grad_norm": 0.272276908159256, "learning_rate": 9.010897533144754e-05, "loss": 0.0298, "step": 24350 }, { "grad_norm": 0.19569823145866394, "learning_rate": 9.009910055227306e-05, "loss": 0.0272, "step": 24360 }, { "grad_norm": 0.20132440328598022, "learning_rate": 9.008922138792124e-05, "loss": 0.0238, "step": 24370 }, { "grad_norm": 0.6692080497741699, "learning_rate": 9.007933783947244e-05, "loss": 0.0251, "step": 24380 }, { "grad_norm": 0.2244579792022705, "learning_rate": 9.006944990800752e-05, "loss": 0.0295, "step": 24390 }, { "grad_norm": 0.23346498608589172, "learning_rate": 9.005955759460779e-05, "loss": 0.0295, "step": 24400 }, { "grad_norm": 0.3910177946090698, "learning_rate": 9.004966090035508e-05, "loss": 0.0293, "step": 24410 }, { "grad_norm": 0.24452562630176544, "learning_rate": 9.003975982633166e-05, "loss": 0.0257, "step": 24420 }, { "grad_norm": 0.25515300035476685, "learning_rate": 9.00298543736203e-05, "loss": 0.024, "step": 24430 }, { "grad_norm": 0.2800231873989105, "learning_rate": 9.001994454330427e-05, "loss": 0.0266, "step": 24440 }, { "grad_norm": 0.17123542726039886, "learning_rate": 9.001003033646727e-05, "loss": 0.0254, "step": 24450 }, { "grad_norm": 0.23488827049732208, "learning_rate": 9.00001117541935e-05, "loss": 0.0199, "step": 24460 }, { "grad_norm": 0.20040598511695862, "learning_rate": 8.999018879756764e-05, "loss": 0.0196, "step": 24470 }, { "grad_norm": 0.28993725776672363, "learning_rate": 8.998026146767487e-05, "loss": 0.0211, "step": 24480 }, { "grad_norm": 0.21195146441459656, "learning_rate": 8.99703297656008e-05, "loss": 0.0186, "step": 24490 }, { "grad_norm": 0.169404998421669, "learning_rate": 8.996039369243156e-05, "loss": 0.0219, "step": 24500 }, { "grad_norm": 0.2495001256465912, "learning_rate": 8.995045324925378e-05, "loss": 0.0223, "step": 24510 }, { "grad_norm": 0.17242707312107086, "learning_rate": 8.994050843715448e-05, "loss": 0.0215, "step": 24520 }, { "grad_norm": 0.20612065494060516, "learning_rate": 8.993055925722121e-05, "loss": 0.0212, "step": 24530 }, { "grad_norm": 0.23739930987358093, "learning_rate": 8.992060571054202e-05, "loss": 0.0185, "step": 24540 }, { "grad_norm": 0.16886290907859802, "learning_rate": 8.991064779820542e-05, "loss": 0.0194, "step": 24550 }, { "grad_norm": 0.16338768601417542, "learning_rate": 8.990068552130036e-05, "loss": 0.0204, "step": 24560 }, { "grad_norm": 0.1455460637807846, "learning_rate": 8.989071888091634e-05, "loss": 0.0186, "step": 24570 }, { "grad_norm": 0.20590592920780182, "learning_rate": 8.988074787814329e-05, "loss": 0.0206, "step": 24580 }, { "grad_norm": 0.17995113134384155, "learning_rate": 8.987077251407158e-05, "loss": 0.0192, "step": 24590 }, { "grad_norm": 0.2184264212846756, "learning_rate": 8.986079278979216e-05, "loss": 0.0191, "step": 24600 }, { "grad_norm": 0.22282692790031433, "learning_rate": 8.985080870639635e-05, "loss": 0.0188, "step": 24610 }, { "grad_norm": 0.2210530936717987, "learning_rate": 8.984082026497603e-05, "loss": 0.0186, "step": 24620 }, { "grad_norm": 0.17924585938453674, "learning_rate": 8.98308274666235e-05, "loss": 0.0187, "step": 24630 }, { "grad_norm": 0.22352373600006104, "learning_rate": 8.982083031243155e-05, "loss": 0.019, "step": 24640 }, { "grad_norm": 0.2461577206850052, "learning_rate": 8.98108288034935e-05, "loss": 0.0214, "step": 24650 }, { "grad_norm": 0.2683808505535126, "learning_rate": 8.980082294090305e-05, "loss": 0.0295, "step": 24660 }, { "grad_norm": 0.24880605936050415, "learning_rate": 8.979081272575443e-05, "loss": 0.0249, "step": 24670 }, { "grad_norm": 0.30156025290489197, "learning_rate": 8.978079815914236e-05, "loss": 0.0235, "step": 24680 }, { "grad_norm": 0.23907193541526794, "learning_rate": 8.977077924216202e-05, "loss": 0.0277, "step": 24690 }, { "grad_norm": 0.32591381669044495, "learning_rate": 8.976075597590905e-05, "loss": 0.0227, "step": 24700 }, { "grad_norm": 0.32324931025505066, "learning_rate": 8.975072836147958e-05, "loss": 0.0245, "step": 24710 }, { "grad_norm": 0.1710014045238495, "learning_rate": 8.974069639997025e-05, "loss": 0.0225, "step": 24720 }, { "grad_norm": 0.16492727398872375, "learning_rate": 8.973066009247808e-05, "loss": 0.0214, "step": 24730 }, { "grad_norm": 0.20712430775165558, "learning_rate": 8.972061944010066e-05, "loss": 0.0203, "step": 24740 }, { "grad_norm": 0.1847788244485855, "learning_rate": 8.971057444393603e-05, "loss": 0.0195, "step": 24750 }, { "grad_norm": 0.26656433939933777, "learning_rate": 8.970052510508268e-05, "loss": 0.0213, "step": 24760 }, { "grad_norm": 0.2267438769340515, "learning_rate": 8.969047142463959e-05, "loss": 0.0235, "step": 24770 }, { "grad_norm": 0.17504039406776428, "learning_rate": 8.968041340370621e-05, "loss": 0.0207, "step": 24780 }, { "grad_norm": 0.1975811868906021, "learning_rate": 8.96703510433825e-05, "loss": 0.0212, "step": 24790 }, { "grad_norm": 0.22448009252548218, "learning_rate": 8.966028434476883e-05, "loss": 0.0231, "step": 24800 }, { "grad_norm": 0.2150687873363495, "learning_rate": 8.96502133089661e-05, "loss": 0.0208, "step": 24810 }, { "grad_norm": 0.16334094107151031, "learning_rate": 8.964013793707564e-05, "loss": 0.0228, "step": 24820 }, { "grad_norm": 0.1624923199415207, "learning_rate": 8.963005823019932e-05, "loss": 0.0213, "step": 24830 }, { "grad_norm": 0.27968958020210266, "learning_rate": 8.961997418943939e-05, "loss": 0.0201, "step": 24840 }, { "grad_norm": 0.1976132094860077, "learning_rate": 8.960988581589865e-05, "loss": 0.0185, "step": 24850 }, { "grad_norm": 0.24763131141662598, "learning_rate": 8.959979311068037e-05, "loss": 0.0256, "step": 24860 }, { "grad_norm": 0.16088810563087463, "learning_rate": 8.958969607488823e-05, "loss": 0.0215, "step": 24870 }, { "grad_norm": 0.16506175696849823, "learning_rate": 8.957959470962647e-05, "loss": 0.0212, "step": 24880 }, { "grad_norm": 0.1547107696533203, "learning_rate": 8.956948901599971e-05, "loss": 0.0197, "step": 24890 }, { "grad_norm": 0.27434173226356506, "learning_rate": 8.955937899511315e-05, "loss": 0.019, "step": 24900 }, { "grad_norm": 0.17371000349521637, "learning_rate": 8.954926464807238e-05, "loss": 0.02, "step": 24910 }, { "grad_norm": 0.2054244875907898, "learning_rate": 8.953914597598347e-05, "loss": 0.0188, "step": 24920 }, { "grad_norm": 0.18849557638168335, "learning_rate": 8.952902297995303e-05, "loss": 0.0179, "step": 24930 }, { "grad_norm": 0.24014008045196533, "learning_rate": 8.951889566108804e-05, "loss": 0.0205, "step": 24940 }, { "grad_norm": 0.1697649359703064, "learning_rate": 8.950876402049606e-05, "loss": 0.0171, "step": 24950 }, { "grad_norm": 0.17505523562431335, "learning_rate": 8.949862805928504e-05, "loss": 0.0199, "step": 24960 }, { "grad_norm": 0.16688695549964905, "learning_rate": 8.948848777856343e-05, "loss": 0.0187, "step": 24970 }, { "grad_norm": 0.281995564699173, "learning_rate": 8.947834317944017e-05, "loss": 0.0191, "step": 24980 }, { "grad_norm": 0.21531787514686584, "learning_rate": 8.946819426302466e-05, "loss": 0.018, "step": 24990 }, { "grad_norm": 0.2536165714263916, "learning_rate": 8.945804103042676e-05, "loss": 0.023, "step": 25000 }, { "grad_norm": 0.24243822693824768, "learning_rate": 8.944788348275681e-05, "loss": 0.0178, "step": 25010 }, { "grad_norm": 0.17125341296195984, "learning_rate": 8.943772162112565e-05, "loss": 0.0183, "step": 25020 }, { "grad_norm": 0.18713174760341644, "learning_rate": 8.942755544664454e-05, "loss": 0.0191, "step": 25030 }, { "grad_norm": 0.24485886096954346, "learning_rate": 8.941738496042525e-05, "loss": 0.0204, "step": 25040 }, { "grad_norm": 0.20697249472141266, "learning_rate": 8.940721016357999e-05, "loss": 0.0169, "step": 25050 }, { "grad_norm": 0.20749913156032562, "learning_rate": 8.939703105722148e-05, "loss": 0.0201, "step": 25060 }, { "grad_norm": 0.18416418135166168, "learning_rate": 8.93868476424629e-05, "loss": 0.0176, "step": 25070 }, { "grad_norm": 0.16547957062721252, "learning_rate": 8.937665992041786e-05, "loss": 0.0223, "step": 25080 }, { "grad_norm": 0.1346752643585205, "learning_rate": 8.93664678922005e-05, "loss": 0.0175, "step": 25090 }, { "grad_norm": 0.171980619430542, "learning_rate": 8.93562715589254e-05, "loss": 0.0178, "step": 25100 }, { "grad_norm": 0.18458189070224762, "learning_rate": 8.934607092170762e-05, "loss": 0.0183, "step": 25110 }, { "grad_norm": 0.17697009444236755, "learning_rate": 8.933586598166266e-05, "loss": 0.0166, "step": 25120 }, { "grad_norm": 0.21538029611110687, "learning_rate": 8.932565673990655e-05, "loss": 0.0176, "step": 25130 }, { "grad_norm": 0.18591859936714172, "learning_rate": 8.931544319755574e-05, "loss": 0.0182, "step": 25140 }, { "grad_norm": 0.17370247840881348, "learning_rate": 8.930522535572718e-05, "loss": 0.0181, "step": 25150 }, { "grad_norm": 0.15735957026481628, "learning_rate": 8.929500321553826e-05, "loss": 0.0176, "step": 25160 }, { "grad_norm": 0.19324122369289398, "learning_rate": 8.928477677810686e-05, "loss": 0.0172, "step": 25170 }, { "grad_norm": 0.16719545423984528, "learning_rate": 8.927454604455137e-05, "loss": 0.0165, "step": 25180 }, { "grad_norm": 0.16889074444770813, "learning_rate": 8.926431101599053e-05, "loss": 0.0167, "step": 25190 }, { "grad_norm": 0.16109523177146912, "learning_rate": 8.925407169354369e-05, "loss": 0.016, "step": 25200 }, { "grad_norm": 0.1818261742591858, "learning_rate": 8.92438280783306e-05, "loss": 0.0178, "step": 25210 }, { "grad_norm": 0.16131806373596191, "learning_rate": 8.923358017147146e-05, "loss": 0.0182, "step": 25220 }, { "grad_norm": 0.18416059017181396, "learning_rate": 8.922332797408697e-05, "loss": 0.0175, "step": 25230 }, { "grad_norm": 0.13681551814079285, "learning_rate": 8.921307148729831e-05, "loss": 0.0185, "step": 25240 }, { "grad_norm": 0.1792910248041153, "learning_rate": 8.920281071222712e-05, "loss": 0.019, "step": 25250 }, { "grad_norm": 0.3277716040611267, "learning_rate": 8.919254564999548e-05, "loss": 0.0175, "step": 25260 }, { "grad_norm": 0.23528048396110535, "learning_rate": 8.918227630172598e-05, "loss": 0.0197, "step": 25270 }, { "grad_norm": 0.15900561213493347, "learning_rate": 8.917200266854165e-05, "loss": 0.0165, "step": 25280 }, { "grad_norm": 0.16127093136310577, "learning_rate": 8.9161724751566e-05, "loss": 0.019, "step": 25290 }, { "grad_norm": 0.1414322406053543, "learning_rate": 8.915144255192302e-05, "loss": 0.0161, "step": 25300 }, { "grad_norm": 0.22282624244689941, "learning_rate": 8.914115607073714e-05, "loss": 0.0174, "step": 25310 }, { "grad_norm": 0.21774619817733765, "learning_rate": 8.913086530913327e-05, "loss": 0.0164, "step": 25320 }, { "grad_norm": 0.2373054474592209, "learning_rate": 8.912057026823681e-05, "loss": 0.019, "step": 25330 }, { "grad_norm": 0.19994430243968964, "learning_rate": 8.91102709491736e-05, "loss": 0.0179, "step": 25340 }, { "grad_norm": 0.1695718914270401, "learning_rate": 8.909996735306996e-05, "loss": 0.0186, "step": 25350 }, { "grad_norm": 0.17600314319133759, "learning_rate": 8.908965948105268e-05, "loss": 0.0163, "step": 25360 }, { "grad_norm": 0.16201342642307281, "learning_rate": 8.907934733424901e-05, "loss": 0.0211, "step": 25370 }, { "grad_norm": 0.21317949891090393, "learning_rate": 8.906903091378666e-05, "loss": 0.0176, "step": 25380 }, { "grad_norm": 0.19392259418964386, "learning_rate": 8.905871022079384e-05, "loss": 0.019, "step": 25390 }, { "grad_norm": 0.26855501532554626, "learning_rate": 8.90483852563992e-05, "loss": 0.0208, "step": 25400 }, { "grad_norm": 0.2349003255367279, "learning_rate": 8.903805602173185e-05, "loss": 0.0201, "step": 25410 }, { "grad_norm": 0.2027229517698288, "learning_rate": 8.902772251792137e-05, "loss": 0.0202, "step": 25420 }, { "grad_norm": 0.27380990982055664, "learning_rate": 8.901738474609786e-05, "loss": 0.0182, "step": 25430 }, { "grad_norm": 0.18903949856758118, "learning_rate": 8.900704270739179e-05, "loss": 0.0167, "step": 25440 }, { "grad_norm": 0.22298772633075714, "learning_rate": 8.89966964029342e-05, "loss": 0.02, "step": 25450 }, { "grad_norm": 0.14020222425460815, "learning_rate": 8.898634583385652e-05, "loss": 0.0164, "step": 25460 }, { "grad_norm": 0.20470286905765533, "learning_rate": 8.897599100129065e-05, "loss": 0.0193, "step": 25470 }, { "grad_norm": 0.18912658095359802, "learning_rate": 8.896563190636903e-05, "loss": 0.0206, "step": 25480 }, { "grad_norm": 0.19447414577007294, "learning_rate": 8.895526855022448e-05, "loss": 0.0175, "step": 25490 }, { "grad_norm": 0.16641290485858917, "learning_rate": 8.894490093399033e-05, "loss": 0.019, "step": 25500 }, { "grad_norm": 0.27245694398880005, "learning_rate": 8.893452905880035e-05, "loss": 0.0169, "step": 25510 }, { "grad_norm": 0.2106790393590927, "learning_rate": 8.892415292578883e-05, "loss": 0.0186, "step": 25520 }, { "grad_norm": 0.22883376479148865, "learning_rate": 8.891377253609046e-05, "loss": 0.0202, "step": 25530 }, { "grad_norm": 0.18804536759853363, "learning_rate": 8.890338789084043e-05, "loss": 0.0178, "step": 25540 }, { "grad_norm": 0.21333622932434082, "learning_rate": 8.88929989911744e-05, "loss": 0.0182, "step": 25550 }, { "grad_norm": 0.184300497174263, "learning_rate": 8.888260583822847e-05, "loss": 0.0185, "step": 25560 }, { "grad_norm": 0.20544733107089996, "learning_rate": 8.887220843313921e-05, "loss": 0.0174, "step": 25570 }, { "grad_norm": 0.28674405813217163, "learning_rate": 8.88618067770437e-05, "loss": 0.0157, "step": 25580 }, { "grad_norm": 0.23437386751174927, "learning_rate": 8.885140087107942e-05, "loss": 0.0193, "step": 25590 }, { "grad_norm": 0.17673349380493164, "learning_rate": 8.884099071638436e-05, "loss": 0.018, "step": 25600 }, { "grad_norm": 0.18431642651557922, "learning_rate": 8.883057631409695e-05, "loss": 0.0162, "step": 25610 }, { "grad_norm": 0.22210678458213806, "learning_rate": 8.882015766535608e-05, "loss": 0.0158, "step": 25620 }, { "grad_norm": 0.22302131354808807, "learning_rate": 8.880973477130115e-05, "loss": 0.0182, "step": 25630 }, { "grad_norm": 0.17676685750484467, "learning_rate": 8.879930763307197e-05, "loss": 0.0172, "step": 25640 }, { "grad_norm": 0.20023015141487122, "learning_rate": 8.878887625180884e-05, "loss": 0.017, "step": 25650 }, { "grad_norm": 0.17251408100128174, "learning_rate": 8.877844062865253e-05, "loss": 0.0159, "step": 25660 }, { "grad_norm": 0.17091676592826843, "learning_rate": 8.876800076474424e-05, "loss": 0.017, "step": 25670 }, { "grad_norm": 0.1764954775571823, "learning_rate": 8.875755666122568e-05, "loss": 0.0185, "step": 25680 }, { "grad_norm": 0.1702713668346405, "learning_rate": 8.8747108319239e-05, "loss": 0.0172, "step": 25690 }, { "grad_norm": 0.16603249311447144, "learning_rate": 8.87366557399268e-05, "loss": 0.0178, "step": 25700 }, { "grad_norm": 0.17040717601776123, "learning_rate": 8.872619892443217e-05, "loss": 0.019, "step": 25710 }, { "grad_norm": 0.18056203424930573, "learning_rate": 8.871573787389865e-05, "loss": 0.0159, "step": 25720 }, { "grad_norm": 0.14421458542346954, "learning_rate": 8.870527258947024e-05, "loss": 0.0163, "step": 25730 }, { "grad_norm": 0.22860029339790344, "learning_rate": 8.869480307229143e-05, "loss": 0.0172, "step": 25740 }, { "grad_norm": 0.22520479559898376, "learning_rate": 8.868432932350712e-05, "loss": 0.0172, "step": 25750 }, { "grad_norm": 0.19720812141895294, "learning_rate": 8.867385134426272e-05, "loss": 0.0176, "step": 25760 }, { "grad_norm": 0.21675735712051392, "learning_rate": 8.866336913570407e-05, "loss": 0.0188, "step": 25770 }, { "grad_norm": 0.19404378533363342, "learning_rate": 8.865288269897751e-05, "loss": 0.0162, "step": 25780 }, { "grad_norm": 0.18671615421772003, "learning_rate": 8.864239203522981e-05, "loss": 0.0211, "step": 25790 }, { "grad_norm": 0.21406146883964539, "learning_rate": 8.863189714560822e-05, "loss": 0.0173, "step": 25800 }, { "grad_norm": 0.2158251851797104, "learning_rate": 8.862139803126043e-05, "loss": 0.0171, "step": 25810 }, { "grad_norm": 0.2574293315410614, "learning_rate": 8.861089469333463e-05, "loss": 0.0197, "step": 25820 }, { "grad_norm": 0.33946749567985535, "learning_rate": 8.860038713297944e-05, "loss": 0.0226, "step": 25830 }, { "grad_norm": 0.2839808166027069, "learning_rate": 8.858987535134394e-05, "loss": 0.0243, "step": 25840 }, { "grad_norm": 0.36731526255607605, "learning_rate": 8.857935934957769e-05, "loss": 0.024, "step": 25850 }, { "grad_norm": 0.23313497006893158, "learning_rate": 8.856883912883071e-05, "loss": 0.0225, "step": 25860 }, { "grad_norm": 0.21831578016281128, "learning_rate": 8.855831469025346e-05, "loss": 0.0228, "step": 25870 }, { "grad_norm": 0.20411697030067444, "learning_rate": 8.854778603499689e-05, "loss": 0.0205, "step": 25880 }, { "grad_norm": 0.20795752108097076, "learning_rate": 8.85372531642124e-05, "loss": 0.023, "step": 25890 }, { "grad_norm": 0.30934619903564453, "learning_rate": 8.852671607905185e-05, "loss": 0.0248, "step": 25900 }, { "grad_norm": 0.2804880142211914, "learning_rate": 8.851617478066754e-05, "loss": 0.0209, "step": 25910 }, { "grad_norm": 0.1417790651321411, "learning_rate": 8.850562927021227e-05, "loss": 0.0194, "step": 25920 }, { "grad_norm": 0.18553560972213745, "learning_rate": 8.849507954883928e-05, "loss": 0.0188, "step": 25930 }, { "grad_norm": 0.1648865044116974, "learning_rate": 8.848452561770226e-05, "loss": 0.0167, "step": 25940 }, { "grad_norm": 0.21381734311580658, "learning_rate": 8.847396747795538e-05, "loss": 0.0179, "step": 25950 }, { "grad_norm": 0.20201225578784943, "learning_rate": 8.846340513075327e-05, "loss": 0.0172, "step": 25960 }, { "grad_norm": 0.2683424651622772, "learning_rate": 8.845283857725099e-05, "loss": 0.0216, "step": 25970 }, { "grad_norm": 0.25063231587409973, "learning_rate": 8.844226781860409e-05, "loss": 0.0213, "step": 25980 }, { "grad_norm": 0.19265183806419373, "learning_rate": 8.84316928559686e-05, "loss": 0.0226, "step": 25990 }, { "grad_norm": 0.15405882894992828, "learning_rate": 8.842111369050094e-05, "loss": 0.018, "step": 26000 }, { "grad_norm": 0.17188657820224762, "learning_rate": 8.841053032335808e-05, "loss": 0.0199, "step": 26010 }, { "grad_norm": 0.3100607693195343, "learning_rate": 8.839994275569735e-05, "loss": 0.0181, "step": 26020 }, { "grad_norm": 0.11745345592498779, "learning_rate": 8.838935098867662e-05, "loss": 0.0162, "step": 26030 }, { "grad_norm": 0.16625455021858215, "learning_rate": 8.837875502345418e-05, "loss": 0.0191, "step": 26040 }, { "grad_norm": 0.17230741679668427, "learning_rate": 8.83681548611888e-05, "loss": 0.0176, "step": 26050 }, { "grad_norm": 0.23303203284740448, "learning_rate": 8.835755050303969e-05, "loss": 0.0224, "step": 26060 }, { "grad_norm": 0.19157390296459198, "learning_rate": 8.834694195016653e-05, "loss": 0.0176, "step": 26070 }, { "grad_norm": 0.25705772638320923, "learning_rate": 8.833632920372942e-05, "loss": 0.0179, "step": 26080 }, { "grad_norm": 0.17647497355937958, "learning_rate": 8.832571226488903e-05, "loss": 0.0178, "step": 26090 }, { "grad_norm": 0.1400136798620224, "learning_rate": 8.831509113480634e-05, "loss": 0.0159, "step": 26100 }, { "grad_norm": 0.1850859671831131, "learning_rate": 8.83044658146429e-05, "loss": 0.0174, "step": 26110 }, { "grad_norm": 0.182755246758461, "learning_rate": 8.829383630556067e-05, "loss": 0.019, "step": 26120 }, { "grad_norm": 0.1966252624988556, "learning_rate": 8.828320260872207e-05, "loss": 0.0161, "step": 26130 }, { "grad_norm": 0.1776714026927948, "learning_rate": 8.827256472529e-05, "loss": 0.0166, "step": 26140 }, { "grad_norm": 0.21574842929840088, "learning_rate": 8.826192265642778e-05, "loss": 0.0168, "step": 26150 }, { "grad_norm": 0.21974003314971924, "learning_rate": 8.825127640329923e-05, "loss": 0.0169, "step": 26160 }, { "grad_norm": 0.20515213906764984, "learning_rate": 8.824062596706861e-05, "loss": 0.0169, "step": 26170 }, { "grad_norm": 0.21104846894741058, "learning_rate": 8.822997134890062e-05, "loss": 0.0157, "step": 26180 }, { "grad_norm": 0.2620253562927246, "learning_rate": 8.821931254996044e-05, "loss": 0.0166, "step": 26190 }, { "grad_norm": 0.26329344511032104, "learning_rate": 8.82086495714137e-05, "loss": 0.0178, "step": 26200 }, { "grad_norm": 0.24321508407592773, "learning_rate": 8.81979824144265e-05, "loss": 0.0161, "step": 26210 }, { "grad_norm": 0.1962619423866272, "learning_rate": 8.818731108016536e-05, "loss": 0.0173, "step": 26220 }, { "grad_norm": 0.19545157253742218, "learning_rate": 8.81766355697973e-05, "loss": 0.0179, "step": 26230 }, { "grad_norm": 0.2329261302947998, "learning_rate": 8.816595588448977e-05, "loss": 0.0177, "step": 26240 }, { "grad_norm": 0.14315398037433624, "learning_rate": 8.81552720254107e-05, "loss": 0.016, "step": 26250 }, { "grad_norm": 0.2113378494977951, "learning_rate": 8.814458399372842e-05, "loss": 0.0213, "step": 26260 }, { "grad_norm": 0.1665688157081604, "learning_rate": 8.813389179061181e-05, "loss": 0.0168, "step": 26270 }, { "grad_norm": 0.15964484214782715, "learning_rate": 8.812319541723012e-05, "loss": 0.0175, "step": 26280 }, { "grad_norm": 0.1892811506986618, "learning_rate": 8.811249487475309e-05, "loss": 0.0166, "step": 26290 }, { "grad_norm": 0.26054584980010986, "learning_rate": 8.810179016435092e-05, "loss": 0.0145, "step": 26300 }, { "grad_norm": 0.1372520625591278, "learning_rate": 8.809108128719428e-05, "loss": 0.0206, "step": 26310 }, { "grad_norm": 0.16403605043888092, "learning_rate": 8.808036824445424e-05, "loss": 0.017, "step": 26320 }, { "grad_norm": 0.19040969014167786, "learning_rate": 8.806965103730238e-05, "loss": 0.0174, "step": 26330 }, { "grad_norm": 0.24405212700366974, "learning_rate": 8.805892966691074e-05, "loss": 0.0244, "step": 26340 }, { "grad_norm": 0.1975875347852707, "learning_rate": 8.804820413445175e-05, "loss": 0.0186, "step": 26350 }, { "grad_norm": 0.1782716065645218, "learning_rate": 8.803747444109837e-05, "loss": 0.0178, "step": 26360 }, { "grad_norm": 0.18098129332065582, "learning_rate": 8.802674058802399e-05, "loss": 0.0184, "step": 26370 }, { "grad_norm": 0.1526300013065338, "learning_rate": 8.801600257640241e-05, "loss": 0.0177, "step": 26380 }, { "grad_norm": 0.3487970232963562, "learning_rate": 8.800526040740795e-05, "loss": 0.019, "step": 26390 }, { "grad_norm": 0.29200008511543274, "learning_rate": 8.799451408221535e-05, "loss": 0.0196, "step": 26400 }, { "grad_norm": 0.2086661458015442, "learning_rate": 8.798376360199982e-05, "loss": 0.018, "step": 26410 }, { "grad_norm": 0.1485031694173813, "learning_rate": 8.797300896793701e-05, "loss": 0.0164, "step": 26420 }, { "grad_norm": 0.14380881190299988, "learning_rate": 8.796225018120302e-05, "loss": 0.0171, "step": 26430 }, { "grad_norm": 0.19378992915153503, "learning_rate": 8.795148724297444e-05, "loss": 0.0179, "step": 26440 }, { "grad_norm": 0.2334037721157074, "learning_rate": 8.794072015442825e-05, "loss": 0.0171, "step": 26450 }, { "grad_norm": 0.1394425928592682, "learning_rate": 8.792994891674198e-05, "loss": 0.0157, "step": 26460 }, { "grad_norm": 0.1988588273525238, "learning_rate": 8.79191735310935e-05, "loss": 0.0186, "step": 26470 }, { "grad_norm": 0.2174750566482544, "learning_rate": 8.790839399866122e-05, "loss": 0.0174, "step": 26480 }, { "grad_norm": 0.23348690569400787, "learning_rate": 8.789761032062397e-05, "loss": 0.0183, "step": 26490 }, { "grad_norm": 0.20369857549667358, "learning_rate": 8.788682249816103e-05, "loss": 0.0186, "step": 26500 }, { "grad_norm": 0.2094798982143402, "learning_rate": 8.787603053245215e-05, "loss": 0.0195, "step": 26510 }, { "grad_norm": 0.19278818368911743, "learning_rate": 8.78652344246775e-05, "loss": 0.0165, "step": 26520 }, { "grad_norm": 0.16255688667297363, "learning_rate": 8.785443417601776e-05, "loss": 0.019, "step": 26530 }, { "grad_norm": 0.18768444657325745, "learning_rate": 8.784362978765401e-05, "loss": 0.016, "step": 26540 }, { "grad_norm": 0.23595035076141357, "learning_rate": 8.783282126076779e-05, "loss": 0.0169, "step": 26550 }, { "grad_norm": 0.16398775577545166, "learning_rate": 8.782200859654112e-05, "loss": 0.0176, "step": 26560 }, { "grad_norm": 0.2022639364004135, "learning_rate": 8.781119179615646e-05, "loss": 0.0197, "step": 26570 }, { "grad_norm": 0.2224925458431244, "learning_rate": 8.780037086079674e-05, "loss": 0.0197, "step": 26580 }, { "grad_norm": 0.14410540461540222, "learning_rate": 8.778954579164527e-05, "loss": 0.0163, "step": 26590 }, { "grad_norm": 0.18651233613491058, "learning_rate": 8.777871658988588e-05, "loss": 0.021, "step": 26600 }, { "grad_norm": 0.2629796862602234, "learning_rate": 8.776788325670285e-05, "loss": 0.0166, "step": 26610 }, { "grad_norm": 0.2677038609981537, "learning_rate": 8.775704579328089e-05, "loss": 0.0157, "step": 26620 }, { "grad_norm": 0.1554105281829834, "learning_rate": 8.774620420080517e-05, "loss": 0.0191, "step": 26630 }, { "grad_norm": 0.212503582239151, "learning_rate": 8.773535848046131e-05, "loss": 0.0202, "step": 26640 }, { "grad_norm": 0.18955282866954803, "learning_rate": 8.772450863343538e-05, "loss": 0.0201, "step": 26650 }, { "grad_norm": 0.2333092838525772, "learning_rate": 8.77136546609139e-05, "loss": 0.0201, "step": 26660 }, { "grad_norm": 0.22241142392158508, "learning_rate": 8.770279656408385e-05, "loss": 0.0186, "step": 26670 }, { "grad_norm": 0.21146468818187714, "learning_rate": 8.769193434413265e-05, "loss": 0.0219, "step": 26680 }, { "grad_norm": 0.36338010430336, "learning_rate": 8.76810680022482e-05, "loss": 0.0245, "step": 26690 }, { "grad_norm": 0.16081373393535614, "learning_rate": 8.767019753961878e-05, "loss": 0.0229, "step": 26700 }, { "grad_norm": 0.19993886351585388, "learning_rate": 8.765932295743321e-05, "loss": 0.0221, "step": 26710 }, { "grad_norm": 0.18058210611343384, "learning_rate": 8.764844425688068e-05, "loss": 0.0204, "step": 26720 }, { "grad_norm": 0.25790685415267944, "learning_rate": 8.763756143915092e-05, "loss": 0.0253, "step": 26730 }, { "grad_norm": 0.22036045789718628, "learning_rate": 8.7626674505434e-05, "loss": 0.019, "step": 26740 }, { "grad_norm": 0.20784583687782288, "learning_rate": 8.761578345692053e-05, "loss": 0.0175, "step": 26750 }, { "grad_norm": 0.17522262036800385, "learning_rate": 8.760488829480156e-05, "loss": 0.0193, "step": 26760 }, { "grad_norm": 0.19660155475139618, "learning_rate": 8.759398902026854e-05, "loss": 0.0187, "step": 26770 }, { "grad_norm": 0.1629495918750763, "learning_rate": 8.758308563451339e-05, "loss": 0.0166, "step": 26780 }, { "grad_norm": 0.18714477121829987, "learning_rate": 8.75721781387285e-05, "loss": 0.0216, "step": 26790 }, { "grad_norm": 0.2608773112297058, "learning_rate": 8.75612665341067e-05, "loss": 0.019, "step": 26800 }, { "grad_norm": 0.1669297218322754, "learning_rate": 8.755035082184126e-05, "loss": 0.0166, "step": 26810 }, { "grad_norm": 0.18969495594501495, "learning_rate": 8.753943100312592e-05, "loss": 0.0169, "step": 26820 }, { "grad_norm": 0.1402960568666458, "learning_rate": 8.752850707915484e-05, "loss": 0.0183, "step": 26830 }, { "grad_norm": 0.21115675568580627, "learning_rate": 8.751757905112264e-05, "loss": 0.0148, "step": 26840 }, { "grad_norm": 0.198400616645813, "learning_rate": 8.75066469202244e-05, "loss": 0.0173, "step": 26850 }, { "grad_norm": 0.15864141285419464, "learning_rate": 8.749571068765567e-05, "loss": 0.0183, "step": 26860 }, { "grad_norm": 0.14736725389957428, "learning_rate": 8.748477035461238e-05, "loss": 0.0164, "step": 26870 }, { "grad_norm": 0.18299269676208496, "learning_rate": 8.747382592229095e-05, "loss": 0.0171, "step": 26880 }, { "grad_norm": 0.17609892785549164, "learning_rate": 8.746287739188828e-05, "loss": 0.0177, "step": 26890 }, { "grad_norm": 0.22105498611927032, "learning_rate": 8.745192476460165e-05, "loss": 0.0189, "step": 26900 }, { "grad_norm": 0.22298836708068848, "learning_rate": 8.744096804162882e-05, "loss": 0.0187, "step": 26910 }, { "grad_norm": 0.1918667107820511, "learning_rate": 8.743000722416804e-05, "loss": 0.017, "step": 26920 }, { "grad_norm": 0.21499332785606384, "learning_rate": 8.741904231341793e-05, "loss": 0.0195, "step": 26930 }, { "grad_norm": 0.14546462893486023, "learning_rate": 8.740807331057762e-05, "loss": 0.0163, "step": 26940 }, { "grad_norm": 0.2468854933977127, "learning_rate": 8.739710021684667e-05, "loss": 0.0174, "step": 26950 }, { "grad_norm": 0.2566671371459961, "learning_rate": 8.738612303342503e-05, "loss": 0.0186, "step": 26960 }, { "grad_norm": 0.3113747239112854, "learning_rate": 8.73751417615132e-05, "loss": 0.0154, "step": 26970 }, { "grad_norm": 0.2600780427455902, "learning_rate": 8.736415640231208e-05, "loss": 0.0186, "step": 26980 }, { "grad_norm": 0.18049481511116028, "learning_rate": 8.735316695702297e-05, "loss": 0.0172, "step": 26990 }, { "grad_norm": 0.24477934837341309, "learning_rate": 8.734217342684769e-05, "loss": 0.0201, "step": 27000 }, { "grad_norm": 0.1602129340171814, "learning_rate": 8.733117581298847e-05, "loss": 0.0168, "step": 27010 }, { "grad_norm": 0.17861293256282806, "learning_rate": 8.732017411664796e-05, "loss": 0.0185, "step": 27020 }, { "grad_norm": 0.16947439312934875, "learning_rate": 8.730916833902936e-05, "loss": 0.0212, "step": 27030 }, { "grad_norm": 0.2775917053222656, "learning_rate": 8.729815848133618e-05, "loss": 0.019, "step": 27040 }, { "grad_norm": 0.28637275099754333, "learning_rate": 8.728714454477247e-05, "loss": 0.0185, "step": 27050 }, { "grad_norm": 0.1809680163860321, "learning_rate": 8.727612653054269e-05, "loss": 0.0186, "step": 27060 }, { "grad_norm": 0.21702368557453156, "learning_rate": 8.726510443985176e-05, "loss": 0.0192, "step": 27070 }, { "grad_norm": 0.18378935754299164, "learning_rate": 8.725407827390503e-05, "loss": 0.0175, "step": 27080 }, { "grad_norm": 0.18816551566123962, "learning_rate": 8.724304803390833e-05, "loss": 0.0179, "step": 27090 }, { "grad_norm": 0.16426566243171692, "learning_rate": 8.723201372106788e-05, "loss": 0.017, "step": 27100 }, { "grad_norm": 0.19400432705879211, "learning_rate": 8.722097533659038e-05, "loss": 0.0174, "step": 27110 }, { "grad_norm": 0.24064292013645172, "learning_rate": 8.720993288168299e-05, "loss": 0.0174, "step": 27120 }, { "grad_norm": 0.1438094526529312, "learning_rate": 8.719888635755327e-05, "loss": 0.0189, "step": 27130 }, { "grad_norm": 0.24968686699867249, "learning_rate": 8.718783576540928e-05, "loss": 0.0167, "step": 27140 }, { "grad_norm": 0.17817488312721252, "learning_rate": 8.717678110645948e-05, "loss": 0.0165, "step": 27150 }, { "grad_norm": 0.18425863981246948, "learning_rate": 8.716572238191279e-05, "loss": 0.0162, "step": 27160 }, { "grad_norm": 0.1386585235595703, "learning_rate": 8.715465959297857e-05, "loss": 0.0173, "step": 27170 }, { "grad_norm": 0.18579529225826263, "learning_rate": 8.714359274086665e-05, "loss": 0.0182, "step": 27180 }, { "grad_norm": 0.339572936296463, "learning_rate": 8.713252182678726e-05, "loss": 0.0181, "step": 27190 }, { "grad_norm": 0.22468021512031555, "learning_rate": 8.712144685195112e-05, "loss": 0.0191, "step": 27200 }, { "grad_norm": 0.16214223206043243, "learning_rate": 8.711036781756936e-05, "loss": 0.0173, "step": 27210 }, { "grad_norm": 0.19382044672966003, "learning_rate": 8.709928472485357e-05, "loss": 0.0168, "step": 27220 }, { "grad_norm": 0.18168365955352783, "learning_rate": 8.708819757501579e-05, "loss": 0.0144, "step": 27230 }, { "grad_norm": 0.1481359899044037, "learning_rate": 8.707710636926846e-05, "loss": 0.0149, "step": 27240 }, { "grad_norm": 0.2284182757139206, "learning_rate": 8.706601110882455e-05, "loss": 0.0177, "step": 27250 }, { "grad_norm": 0.19707649946212769, "learning_rate": 8.705491179489738e-05, "loss": 0.0175, "step": 27260 }, { "grad_norm": 0.21516744792461395, "learning_rate": 8.704380842870077e-05, "loss": 0.0148, "step": 27270 }, { "grad_norm": 0.15736478567123413, "learning_rate": 8.703270101144895e-05, "loss": 0.0168, "step": 27280 }, { "grad_norm": 0.2037106156349182, "learning_rate": 8.702158954435664e-05, "loss": 0.0161, "step": 27290 }, { "grad_norm": 0.187372624874115, "learning_rate": 8.701047402863896e-05, "loss": 0.019, "step": 27300 }, { "grad_norm": 0.1901237815618515, "learning_rate": 8.699935446551148e-05, "loss": 0.0185, "step": 27310 }, { "grad_norm": 0.17579135298728943, "learning_rate": 8.698823085619022e-05, "loss": 0.0184, "step": 27320 }, { "grad_norm": 0.17788930237293243, "learning_rate": 8.697710320189166e-05, "loss": 0.0173, "step": 27330 }, { "grad_norm": 0.40866193175315857, "learning_rate": 8.696597150383268e-05, "loss": 0.0193, "step": 27340 }, { "grad_norm": 0.21228423714637756, "learning_rate": 8.695483576323063e-05, "loss": 0.0181, "step": 27350 }, { "grad_norm": 0.21186316013336182, "learning_rate": 8.69436959813033e-05, "loss": 0.0203, "step": 27360 }, { "grad_norm": 0.2204643040895462, "learning_rate": 8.693255215926892e-05, "loss": 0.0178, "step": 27370 }, { "grad_norm": 0.17354024946689606, "learning_rate": 8.692140429834617e-05, "loss": 0.0172, "step": 27380 }, { "grad_norm": 0.121524378657341, "learning_rate": 8.691025239975415e-05, "loss": 0.0143, "step": 27390 }, { "grad_norm": 0.3109537363052368, "learning_rate": 8.689909646471243e-05, "loss": 0.0173, "step": 27400 }, { "grad_norm": 0.1844123750925064, "learning_rate": 8.688793649444099e-05, "loss": 0.0163, "step": 27410 }, { "grad_norm": 0.17874011397361755, "learning_rate": 8.687677249016029e-05, "loss": 0.015, "step": 27420 }, { "grad_norm": 0.20339298248291016, "learning_rate": 8.686560445309118e-05, "loss": 0.0187, "step": 27430 }, { "grad_norm": 0.18635433912277222, "learning_rate": 8.685443238445499e-05, "loss": 0.0191, "step": 27440 }, { "grad_norm": 0.21051618456840515, "learning_rate": 8.68432562854735e-05, "loss": 0.0187, "step": 27450 }, { "grad_norm": 0.19848985970020294, "learning_rate": 8.683207615736887e-05, "loss": 0.0197, "step": 27460 }, { "grad_norm": 0.17912575602531433, "learning_rate": 8.682089200136379e-05, "loss": 0.0166, "step": 27470 }, { "grad_norm": 0.15220125019550323, "learning_rate": 8.680970381868132e-05, "loss": 0.0155, "step": 27480 }, { "grad_norm": 0.15426501631736755, "learning_rate": 8.679851161054498e-05, "loss": 0.0181, "step": 27490 }, { "grad_norm": 0.18894602358341217, "learning_rate": 8.678731537817873e-05, "loss": 0.015, "step": 27500 }, { "grad_norm": 0.18351256847381592, "learning_rate": 8.677611512280697e-05, "loss": 0.0166, "step": 27510 }, { "grad_norm": 0.27019497752189636, "learning_rate": 8.676491084565457e-05, "loss": 0.0194, "step": 27520 }, { "grad_norm": 0.20773820579051971, "learning_rate": 8.675370254794678e-05, "loss": 0.0151, "step": 27530 }, { "grad_norm": 0.19175651669502258, "learning_rate": 8.674249023090935e-05, "loss": 0.0169, "step": 27540 }, { "grad_norm": 0.17316165566444397, "learning_rate": 8.673127389576843e-05, "loss": 0.0153, "step": 27550 }, { "grad_norm": 0.15306048095226288, "learning_rate": 8.67200535437506e-05, "loss": 0.0195, "step": 27560 }, { "grad_norm": 0.14545056223869324, "learning_rate": 8.670882917608296e-05, "loss": 0.0172, "step": 27570 }, { "grad_norm": 0.20166845619678497, "learning_rate": 8.669760079399292e-05, "loss": 0.0145, "step": 27580 }, { "grad_norm": 0.17482031881809235, "learning_rate": 8.668636839870845e-05, "loss": 0.0158, "step": 27590 }, { "grad_norm": 0.2107315957546234, "learning_rate": 8.667513199145789e-05, "loss": 0.0171, "step": 27600 }, { "grad_norm": 0.19148968160152435, "learning_rate": 8.666389157347002e-05, "loss": 0.0166, "step": 27610 }, { "grad_norm": 0.18357306718826294, "learning_rate": 8.66526471459741e-05, "loss": 0.0182, "step": 27620 }, { "grad_norm": 0.14725956320762634, "learning_rate": 8.66413987101998e-05, "loss": 0.0169, "step": 27630 }, { "grad_norm": 0.1715526282787323, "learning_rate": 8.663014626737723e-05, "loss": 0.0193, "step": 27640 }, { "grad_norm": 0.24839475750923157, "learning_rate": 8.661888981873691e-05, "loss": 0.019, "step": 27650 }, { "grad_norm": 0.1766866147518158, "learning_rate": 8.660762936550988e-05, "loss": 0.0183, "step": 27660 }, { "grad_norm": 0.1834867298603058, "learning_rate": 8.659636490892753e-05, "loss": 0.0155, "step": 27670 }, { "grad_norm": 0.1909785270690918, "learning_rate": 8.658509645022174e-05, "loss": 0.0172, "step": 27680 }, { "grad_norm": 0.14593659341335297, "learning_rate": 8.657382399062481e-05, "loss": 0.0171, "step": 27690 }, { "grad_norm": 0.1807757318019867, "learning_rate": 8.656254753136946e-05, "loss": 0.0154, "step": 27700 }, { "grad_norm": 0.13731740415096283, "learning_rate": 8.655126707368891e-05, "loss": 0.0182, "step": 27710 }, { "grad_norm": 0.18159912526607513, "learning_rate": 8.653998261881672e-05, "loss": 0.0198, "step": 27720 }, { "grad_norm": 0.17683975398540497, "learning_rate": 8.652869416798699e-05, "loss": 0.0164, "step": 27730 }, { "grad_norm": 0.25609201192855835, "learning_rate": 8.651740172243417e-05, "loss": 0.0187, "step": 27740 }, { "grad_norm": 0.15853694081306458, "learning_rate": 8.65061052833932e-05, "loss": 0.0182, "step": 27750 }, { "grad_norm": 0.18841968476772308, "learning_rate": 8.649480485209945e-05, "loss": 0.0179, "step": 27760 }, { "grad_norm": 0.39376744627952576, "learning_rate": 8.64835004297887e-05, "loss": 0.0153, "step": 27770 }, { "grad_norm": 0.16067259013652802, "learning_rate": 8.64721920176972e-05, "loss": 0.0183, "step": 27780 }, { "grad_norm": 0.17770391702651978, "learning_rate": 8.646087961706164e-05, "loss": 0.0174, "step": 27790 }, { "grad_norm": 0.21388082206249237, "learning_rate": 8.644956322911908e-05, "loss": 0.0188, "step": 27800 }, { "grad_norm": 0.17120689153671265, "learning_rate": 8.643824285510709e-05, "loss": 0.0158, "step": 27810 }, { "grad_norm": 0.22924670577049255, "learning_rate": 8.642691849626364e-05, "loss": 0.0189, "step": 27820 }, { "grad_norm": 0.1814754456281662, "learning_rate": 8.641559015382717e-05, "loss": 0.0178, "step": 27830 }, { "grad_norm": 0.13983823359012604, "learning_rate": 8.640425782903649e-05, "loss": 0.0147, "step": 27840 }, { "grad_norm": 0.2628869116306305, "learning_rate": 8.639292152313091e-05, "loss": 0.0149, "step": 27850 }, { "grad_norm": 0.1882711946964264, "learning_rate": 8.638158123735015e-05, "loss": 0.0143, "step": 27860 }, { "grad_norm": 0.20290124416351318, "learning_rate": 8.637023697293436e-05, "loss": 0.0161, "step": 27870 }, { "grad_norm": 0.21037732064723969, "learning_rate": 8.635888873112414e-05, "loss": 0.0183, "step": 27880 }, { "grad_norm": 0.15475308895111084, "learning_rate": 8.634753651316052e-05, "loss": 0.0174, "step": 27890 }, { "grad_norm": 0.15323801338672638, "learning_rate": 8.633618032028496e-05, "loss": 0.0173, "step": 27900 }, { "grad_norm": 0.181695818901062, "learning_rate": 8.632482015373934e-05, "loss": 0.0186, "step": 27910 }, { "grad_norm": 0.20731604099273682, "learning_rate": 8.6313456014766e-05, "loss": 0.0186, "step": 27920 }, { "grad_norm": 0.17972911894321442, "learning_rate": 8.630208790460771e-05, "loss": 0.0151, "step": 27930 }, { "grad_norm": 0.20929597318172455, "learning_rate": 8.629071582450768e-05, "loss": 0.0171, "step": 27940 }, { "grad_norm": 0.16777734458446503, "learning_rate": 8.62793397757095e-05, "loss": 0.0169, "step": 27950 }, { "grad_norm": 0.19726788997650146, "learning_rate": 8.626795975945729e-05, "loss": 0.0188, "step": 27960 }, { "grad_norm": 0.22037145495414734, "learning_rate": 8.625657577699551e-05, "loss": 0.0152, "step": 27970 }, { "grad_norm": 0.1792427897453308, "learning_rate": 8.624518782956914e-05, "loss": 0.0152, "step": 27980 }, { "grad_norm": 0.1762152463197708, "learning_rate": 8.62337959184235e-05, "loss": 0.0185, "step": 27990 }, { "grad_norm": 0.26164934039115906, "learning_rate": 8.622240004480441e-05, "loss": 0.0181, "step": 28000 }, { "grad_norm": 0.17688962817192078, "learning_rate": 8.621100020995814e-05, "loss": 0.0186, "step": 28010 }, { "grad_norm": 0.1845390945672989, "learning_rate": 8.619959641513132e-05, "loss": 0.0166, "step": 28020 }, { "grad_norm": 0.17894592881202698, "learning_rate": 8.618818866157105e-05, "loss": 0.0159, "step": 28030 }, { "grad_norm": 0.17430047690868378, "learning_rate": 8.617677695052487e-05, "loss": 0.0189, "step": 28040 }, { "grad_norm": 0.15820817649364471, "learning_rate": 8.616536128324078e-05, "loss": 0.0168, "step": 28050 }, { "grad_norm": 0.19647082686424255, "learning_rate": 8.615394166096712e-05, "loss": 0.0176, "step": 28060 }, { "grad_norm": 0.15673494338989258, "learning_rate": 8.614251808495279e-05, "loss": 0.0165, "step": 28070 }, { "grad_norm": 0.16894952952861786, "learning_rate": 8.6131090556447e-05, "loss": 0.017, "step": 28080 }, { "grad_norm": 0.18428777158260345, "learning_rate": 8.611965907669947e-05, "loss": 0.018, "step": 28090 }, { "grad_norm": 0.21332338452339172, "learning_rate": 8.610822364696034e-05, "loss": 0.0154, "step": 28100 }, { "grad_norm": 0.20639890432357788, "learning_rate": 8.609678426848015e-05, "loss": 0.0183, "step": 28110 }, { "grad_norm": 0.17167918384075165, "learning_rate": 8.60853409425099e-05, "loss": 0.0172, "step": 28120 }, { "grad_norm": 0.16586117446422577, "learning_rate": 8.607389367030104e-05, "loss": 0.0165, "step": 28130 }, { "grad_norm": 0.19796894490718842, "learning_rate": 8.606244245310538e-05, "loss": 0.018, "step": 28140 }, { "grad_norm": 0.1864924281835556, "learning_rate": 8.605098729217525e-05, "loss": 0.0177, "step": 28150 }, { "grad_norm": 0.22475691139698029, "learning_rate": 8.603952818876335e-05, "loss": 0.0151, "step": 28160 }, { "grad_norm": 0.1849125623703003, "learning_rate": 8.602806514412281e-05, "loss": 0.015, "step": 28170 }, { "grad_norm": 0.19067327678203583, "learning_rate": 8.601659815950726e-05, "loss": 0.0172, "step": 28180 }, { "grad_norm": 0.21439780294895172, "learning_rate": 8.600512723617067e-05, "loss": 0.0165, "step": 28190 }, { "grad_norm": 0.139804869890213, "learning_rate": 8.59936523753675e-05, "loss": 0.0156, "step": 28200 }, { "grad_norm": 0.2260652333498001, "learning_rate": 8.598217357835264e-05, "loss": 0.0169, "step": 28210 }, { "grad_norm": 0.25430208444595337, "learning_rate": 8.597069084638135e-05, "loss": 0.0187, "step": 28220 }, { "grad_norm": 0.22676804661750793, "learning_rate": 8.595920418070939e-05, "loss": 0.0242, "step": 28230 }, { "grad_norm": 0.21902644634246826, "learning_rate": 8.594771358259295e-05, "loss": 0.0213, "step": 28240 }, { "grad_norm": 0.15147534012794495, "learning_rate": 8.593621905328858e-05, "loss": 0.0203, "step": 28250 }, { "grad_norm": 0.19656217098236084, "learning_rate": 8.592472059405333e-05, "loss": 0.0228, "step": 28260 }, { "grad_norm": 0.23587849736213684, "learning_rate": 8.591321820614464e-05, "loss": 0.0197, "step": 28270 }, { "grad_norm": 0.20044980943202972, "learning_rate": 8.590171189082041e-05, "loss": 0.0178, "step": 28280 }, { "grad_norm": 0.21402151882648468, "learning_rate": 8.589020164933894e-05, "loss": 0.0184, "step": 28290 }, { "grad_norm": 0.18161100149154663, "learning_rate": 8.587868748295898e-05, "loss": 0.0169, "step": 28300 }, { "grad_norm": 0.18848709762096405, "learning_rate": 8.586716939293971e-05, "loss": 0.0188, "step": 28310 }, { "grad_norm": 0.1607094258069992, "learning_rate": 8.58556473805407e-05, "loss": 0.0176, "step": 28320 }, { "grad_norm": 0.20792853832244873, "learning_rate": 8.584412144702202e-05, "loss": 0.0167, "step": 28330 }, { "grad_norm": 0.18113566935062408, "learning_rate": 8.58325915936441e-05, "loss": 0.0198, "step": 28340 }, { "grad_norm": 0.17209625244140625, "learning_rate": 8.582105782166783e-05, "loss": 0.0177, "step": 28350 }, { "grad_norm": 0.1471552848815918, "learning_rate": 8.580952013235455e-05, "loss": 0.0192, "step": 28360 }, { "grad_norm": 0.2502193748950958, "learning_rate": 8.579797852696596e-05, "loss": 0.0177, "step": 28370 }, { "grad_norm": 0.1605151891708374, "learning_rate": 8.578643300676428e-05, "loss": 0.0175, "step": 28380 }, { "grad_norm": 0.21447952091693878, "learning_rate": 8.577488357301209e-05, "loss": 0.0203, "step": 28390 }, { "grad_norm": 0.17036956548690796, "learning_rate": 8.576333022697242e-05, "loss": 0.0152, "step": 28400 }, { "grad_norm": 0.1829938143491745, "learning_rate": 8.575177296990873e-05, "loss": 0.0169, "step": 28410 }, { "grad_norm": 0.19097496569156647, "learning_rate": 8.574021180308489e-05, "loss": 0.0145, "step": 28420 }, { "grad_norm": 0.17668718099594116, "learning_rate": 8.572864672776523e-05, "loss": 0.016, "step": 28430 }, { "grad_norm": 0.21190489828586578, "learning_rate": 8.571707774521447e-05, "loss": 0.016, "step": 28440 }, { "grad_norm": 0.25490158796310425, "learning_rate": 8.57055048566978e-05, "loss": 0.0203, "step": 28450 }, { "grad_norm": 0.19059595465660095, "learning_rate": 8.569392806348078e-05, "loss": 0.0186, "step": 28460 }, { "grad_norm": 0.29666051268577576, "learning_rate": 8.568234736682947e-05, "loss": 0.0182, "step": 28470 }, { "grad_norm": 0.4843316376209259, "learning_rate": 8.567076276801029e-05, "loss": 0.0209, "step": 28480 }, { "grad_norm": 0.19854922592639923, "learning_rate": 8.565917426829013e-05, "loss": 0.0189, "step": 28490 }, { "grad_norm": 0.20351313054561615, "learning_rate": 8.564758186893628e-05, "loss": 0.0229, "step": 28500 }, { "grad_norm": 0.16843301057815552, "learning_rate": 8.563598557121649e-05, "loss": 0.0207, "step": 28510 }, { "grad_norm": 0.16186171770095825, "learning_rate": 8.562438537639888e-05, "loss": 0.018, "step": 28520 }, { "grad_norm": 0.2447318136692047, "learning_rate": 8.561278128575206e-05, "loss": 0.0175, "step": 28530 }, { "grad_norm": 0.19316366314888, "learning_rate": 8.5601173300545e-05, "loss": 0.0201, "step": 28540 }, { "grad_norm": 0.17099899053573608, "learning_rate": 8.558956142204717e-05, "loss": 0.016, "step": 28550 }, { "grad_norm": 0.278751403093338, "learning_rate": 8.55779456515284e-05, "loss": 0.0185, "step": 28560 }, { "grad_norm": 0.18541237711906433, "learning_rate": 8.556632599025898e-05, "loss": 0.0191, "step": 28570 }, { "grad_norm": 0.15771663188934326, "learning_rate": 8.555470243950964e-05, "loss": 0.0208, "step": 28580 }, { "grad_norm": 0.17135724425315857, "learning_rate": 8.554307500055148e-05, "loss": 0.0167, "step": 28590 }, { "grad_norm": 0.19786964356899261, "learning_rate": 8.553144367465609e-05, "loss": 0.0212, "step": 28600 }, { "grad_norm": 0.25649258494377136, "learning_rate": 8.551980846309544e-05, "loss": 0.0172, "step": 28610 }, { "grad_norm": 0.26290878653526306, "learning_rate": 8.550816936714193e-05, "loss": 0.018, "step": 28620 }, { "grad_norm": 0.15978798270225525, "learning_rate": 8.549652638806841e-05, "loss": 0.018, "step": 28630 }, { "grad_norm": 0.19304612278938293, "learning_rate": 8.548487952714812e-05, "loss": 0.0173, "step": 28640 }, { "grad_norm": 0.16830630600452423, "learning_rate": 8.547322878565478e-05, "loss": 0.0153, "step": 28650 }, { "grad_norm": 0.17942023277282715, "learning_rate": 8.546157416486245e-05, "loss": 0.0153, "step": 28660 }, { "grad_norm": 0.18648532032966614, "learning_rate": 8.54499156660457e-05, "loss": 0.0187, "step": 28670 }, { "grad_norm": 0.15831537544727325, "learning_rate": 8.543825329047947e-05, "loss": 0.0146, "step": 28680 }, { "grad_norm": 0.2004602551460266, "learning_rate": 8.542658703943913e-05, "loss": 0.0174, "step": 28690 }, { "grad_norm": 0.1770331710577011, "learning_rate": 8.541491691420051e-05, "loss": 0.0163, "step": 28700 }, { "grad_norm": 0.1808597892522812, "learning_rate": 8.54032429160398e-05, "loss": 0.0151, "step": 28710 }, { "grad_norm": 0.1891322284936905, "learning_rate": 8.539156504623369e-05, "loss": 0.017, "step": 28720 }, { "grad_norm": 0.2301267832517624, "learning_rate": 8.537988330605923e-05, "loss": 0.0166, "step": 28730 }, { "grad_norm": 0.2656189799308777, "learning_rate": 8.536819769679393e-05, "loss": 0.0174, "step": 28740 }, { "grad_norm": 0.22136512398719788, "learning_rate": 8.53565082197157e-05, "loss": 0.0187, "step": 28750 }, { "grad_norm": 0.2219599038362503, "learning_rate": 8.534481487610289e-05, "loss": 0.0195, "step": 28760 }, { "grad_norm": 0.1553192287683487, "learning_rate": 8.533311766723428e-05, "loss": 0.0167, "step": 28770 }, { "grad_norm": 0.1995093673467636, "learning_rate": 8.532141659438901e-05, "loss": 0.0172, "step": 28780 }, { "grad_norm": 0.1992115080356598, "learning_rate": 8.530971165884675e-05, "loss": 0.0161, "step": 28790 }, { "grad_norm": 0.16419516503810883, "learning_rate": 8.529800286188752e-05, "loss": 0.0149, "step": 28800 }, { "grad_norm": 0.35896265506744385, "learning_rate": 8.528629020479175e-05, "loss": 0.0172, "step": 28810 }, { "grad_norm": 0.16144129633903503, "learning_rate": 8.527457368884033e-05, "loss": 0.0185, "step": 28820 }, { "grad_norm": 0.19357559084892273, "learning_rate": 8.526285331531458e-05, "loss": 0.0169, "step": 28830 }, { "grad_norm": 0.19457529485225677, "learning_rate": 8.525112908549621e-05, "loss": 0.0178, "step": 28840 }, { "grad_norm": 0.17007127404212952, "learning_rate": 8.523940100066735e-05, "loss": 0.018, "step": 28850 }, { "grad_norm": 0.1922341287136078, "learning_rate": 8.52276690621106e-05, "loss": 0.0173, "step": 28860 }, { "grad_norm": 0.2249765396118164, "learning_rate": 8.521593327110889e-05, "loss": 0.0182, "step": 28870 }, { "grad_norm": 0.19047020375728607, "learning_rate": 8.520419362894569e-05, "loss": 0.0184, "step": 28880 }, { "grad_norm": 0.1794101744890213, "learning_rate": 8.51924501369048e-05, "loss": 0.0148, "step": 28890 }, { "grad_norm": 0.16168418526649475, "learning_rate": 8.518070279627047e-05, "loss": 0.0168, "step": 28900 }, { "grad_norm": 0.14261366426944733, "learning_rate": 8.516895160832737e-05, "loss": 0.0201, "step": 28910 }, { "grad_norm": 0.20108608901500702, "learning_rate": 8.515719657436061e-05, "loss": 0.0161, "step": 28920 }, { "grad_norm": 0.2140648066997528, "learning_rate": 8.514543769565568e-05, "loss": 0.0166, "step": 28930 }, { "grad_norm": 0.1282052844762802, "learning_rate": 8.513367497349853e-05, "loss": 0.0149, "step": 28940 }, { "grad_norm": 0.16978494822978973, "learning_rate": 8.51219084091755e-05, "loss": 0.0165, "step": 28950 }, { "grad_norm": 0.1586761176586151, "learning_rate": 8.511013800397338e-05, "loss": 0.0177, "step": 28960 }, { "grad_norm": 0.20826327800750732, "learning_rate": 8.509836375917937e-05, "loss": 0.0173, "step": 28970 }, { "grad_norm": 0.17886050045490265, "learning_rate": 8.508658567608104e-05, "loss": 0.0168, "step": 28980 }, { "grad_norm": 0.18547089397907257, "learning_rate": 8.507480375596647e-05, "loss": 0.0159, "step": 28990 }, { "grad_norm": 0.1934443712234497, "learning_rate": 8.506301800012408e-05, "loss": 0.0165, "step": 29000 }, { "grad_norm": 0.24039216339588165, "learning_rate": 8.505122840984278e-05, "loss": 0.0167, "step": 29010 }, { "grad_norm": 0.1721821278333664, "learning_rate": 8.503943498641182e-05, "loss": 0.0156, "step": 29020 }, { "grad_norm": 0.15363189578056335, "learning_rate": 8.502763773112095e-05, "loss": 0.0152, "step": 29030 }, { "grad_norm": 0.18986691534519196, "learning_rate": 8.501583664526026e-05, "loss": 0.0172, "step": 29040 }, { "grad_norm": 0.2579951882362366, "learning_rate": 8.500403173012032e-05, "loss": 0.0199, "step": 29050 }, { "grad_norm": 0.26253750920295715, "learning_rate": 8.499222298699211e-05, "loss": 0.0175, "step": 29060 }, { "grad_norm": 0.21762220561504364, "learning_rate": 8.498041041716701e-05, "loss": 0.0173, "step": 29070 }, { "grad_norm": 0.24121713638305664, "learning_rate": 8.496859402193681e-05, "loss": 0.0193, "step": 29080 }, { "grad_norm": 0.12650595605373383, "learning_rate": 8.495677380259374e-05, "loss": 0.0175, "step": 29090 }, { "grad_norm": 0.14304234087467194, "learning_rate": 8.494494976043045e-05, "loss": 0.0179, "step": 29100 }, { "grad_norm": 0.19263702630996704, "learning_rate": 8.493312189673998e-05, "loss": 0.0164, "step": 29110 }, { "grad_norm": 0.19283662736415863, "learning_rate": 8.492129021281584e-05, "loss": 0.0189, "step": 29120 }, { "grad_norm": 0.17692749202251434, "learning_rate": 8.490945470995188e-05, "loss": 0.0187, "step": 29130 }, { "grad_norm": 0.2135516256093979, "learning_rate": 8.489761538944247e-05, "loss": 0.0189, "step": 29140 }, { "grad_norm": 0.20847442746162415, "learning_rate": 8.48857722525823e-05, "loss": 0.0184, "step": 29150 }, { "grad_norm": 0.21341198682785034, "learning_rate": 8.487392530066652e-05, "loss": 0.0173, "step": 29160 }, { "grad_norm": 0.27633532881736755, "learning_rate": 8.486207453499069e-05, "loss": 0.0182, "step": 29170 }, { "grad_norm": 0.16159997880458832, "learning_rate": 8.485021995685082e-05, "loss": 0.018, "step": 29180 }, { "grad_norm": 0.1750599443912506, "learning_rate": 8.483836156754328e-05, "loss": 0.0158, "step": 29190 }, { "grad_norm": 0.15860308706760406, "learning_rate": 8.482649936836491e-05, "loss": 0.0139, "step": 29200 }, { "grad_norm": 0.1976730227470398, "learning_rate": 8.481463336061293e-05, "loss": 0.0173, "step": 29210 }, { "grad_norm": 0.17589694261550903, "learning_rate": 8.480276354558496e-05, "loss": 0.0161, "step": 29220 }, { "grad_norm": 0.14308547973632812, "learning_rate": 8.479088992457913e-05, "loss": 0.0168, "step": 29230 }, { "grad_norm": 0.18092529475688934, "learning_rate": 8.477901249889387e-05, "loss": 0.0173, "step": 29240 }, { "grad_norm": 0.1785929948091507, "learning_rate": 8.47671312698281e-05, "loss": 0.0138, "step": 29250 }, { "grad_norm": 0.17821289598941803, "learning_rate": 8.475524623868112e-05, "loss": 0.0167, "step": 29260 }, { "grad_norm": 0.270394891500473, "learning_rate": 8.474335740675266e-05, "loss": 0.0173, "step": 29270 }, { "grad_norm": 0.16227209568023682, "learning_rate": 8.473146477534289e-05, "loss": 0.0172, "step": 29280 }, { "grad_norm": 0.17720642685890198, "learning_rate": 8.471956834575232e-05, "loss": 0.0176, "step": 29290 }, { "grad_norm": 0.1938062161207199, "learning_rate": 8.470766811928197e-05, "loss": 0.0168, "step": 29300 }, { "grad_norm": 0.18685618042945862, "learning_rate": 8.469576409723323e-05, "loss": 0.0166, "step": 29310 }, { "grad_norm": 0.23410089313983917, "learning_rate": 8.468385628090788e-05, "loss": 0.0178, "step": 29320 }, { "grad_norm": 0.19501982629299164, "learning_rate": 8.467194467160815e-05, "loss": 0.0144, "step": 29330 }, { "grad_norm": 0.22893545031547546, "learning_rate": 8.466002927063667e-05, "loss": 0.0186, "step": 29340 }, { "grad_norm": 0.1897154301404953, "learning_rate": 8.464811007929651e-05, "loss": 0.018, "step": 29350 }, { "grad_norm": 0.1325455904006958, "learning_rate": 8.463618709889114e-05, "loss": 0.0158, "step": 29360 }, { "grad_norm": 0.22257442772388458, "learning_rate": 8.462426033072442e-05, "loss": 0.0174, "step": 29370 }, { "grad_norm": 0.18309760093688965, "learning_rate": 8.461232977610061e-05, "loss": 0.0156, "step": 29380 }, { "grad_norm": 0.15412451326847076, "learning_rate": 8.46003954363245e-05, "loss": 0.0171, "step": 29390 }, { "grad_norm": 0.17947699129581451, "learning_rate": 8.458845731270115e-05, "loss": 0.0156, "step": 29400 }, { "grad_norm": 0.19385306537151337, "learning_rate": 8.45765154065361e-05, "loss": 0.0152, "step": 29410 }, { "grad_norm": 0.19203752279281616, "learning_rate": 8.456456971913532e-05, "loss": 0.0167, "step": 29420 }, { "grad_norm": 0.18967704474925995, "learning_rate": 8.455262025180517e-05, "loss": 0.0187, "step": 29430 }, { "grad_norm": 0.15605156123638153, "learning_rate": 8.454066700585242e-05, "loss": 0.0141, "step": 29440 }, { "grad_norm": 0.24378864467144012, "learning_rate": 8.452870998258423e-05, "loss": 0.0146, "step": 29450 }, { "grad_norm": 0.16728712618350983, "learning_rate": 8.451674918330825e-05, "loss": 0.0167, "step": 29460 }, { "grad_norm": 0.1731930673122406, "learning_rate": 8.450478460933246e-05, "loss": 0.0154, "step": 29470 }, { "grad_norm": 0.19586648046970367, "learning_rate": 8.449281626196532e-05, "loss": 0.0148, "step": 29480 }, { "grad_norm": 0.1707073599100113, "learning_rate": 8.448084414251564e-05, "loss": 0.0167, "step": 29490 }, { "grad_norm": 0.196922168135643, "learning_rate": 8.446886825229271e-05, "loss": 0.0155, "step": 29500 }, { "grad_norm": 0.1186244934797287, "learning_rate": 8.445688859260615e-05, "loss": 0.0161, "step": 29510 }, { "grad_norm": 0.15432408452033997, "learning_rate": 8.444490516476606e-05, "loss": 0.0156, "step": 29520 }, { "grad_norm": 0.17263653874397278, "learning_rate": 8.443291797008293e-05, "loss": 0.0153, "step": 29530 }, { "grad_norm": 0.13788236677646637, "learning_rate": 8.442092700986765e-05, "loss": 0.0163, "step": 29540 }, { "grad_norm": 0.21492062509059906, "learning_rate": 8.440893228543156e-05, "loss": 0.0156, "step": 29550 }, { "grad_norm": 0.21388572454452515, "learning_rate": 8.439693379808638e-05, "loss": 0.0181, "step": 29560 }, { "grad_norm": 0.3163239359855652, "learning_rate": 8.43849315491442e-05, "loss": 0.0225, "step": 29570 }, { "grad_norm": 0.20238924026489258, "learning_rate": 8.437292553991763e-05, "loss": 0.0249, "step": 29580 }, { "grad_norm": 0.20450712740421295, "learning_rate": 8.436091577171959e-05, "loss": 0.0188, "step": 29590 }, { "grad_norm": 0.1649884581565857, "learning_rate": 8.434890224586347e-05, "loss": 0.0186, "step": 29600 }, { "grad_norm": 0.20741991698741913, "learning_rate": 8.433688496366303e-05, "loss": 0.018, "step": 29610 }, { "grad_norm": 0.1679430603981018, "learning_rate": 8.432486392643248e-05, "loss": 0.0182, "step": 29620 }, { "grad_norm": 0.15697717666625977, "learning_rate": 8.431283913548643e-05, "loss": 0.0194, "step": 29630 }, { "grad_norm": 0.15471883118152618, "learning_rate": 8.430081059213985e-05, "loss": 0.018, "step": 29640 }, { "grad_norm": 0.1931433528661728, "learning_rate": 8.428877829770823e-05, "loss": 0.0164, "step": 29650 }, { "grad_norm": 0.15594060719013214, "learning_rate": 8.427674225350735e-05, "loss": 0.0158, "step": 29660 }, { "grad_norm": 0.13913880288600922, "learning_rate": 8.426470246085347e-05, "loss": 0.0158, "step": 29670 }, { "grad_norm": 0.3091944456100464, "learning_rate": 8.425265892106324e-05, "loss": 0.0171, "step": 29680 }, { "grad_norm": 0.19157415628433228, "learning_rate": 8.424061163545374e-05, "loss": 0.0169, "step": 29690 }, { "grad_norm": 0.13676145672798157, "learning_rate": 8.422856060534243e-05, "loss": 0.0168, "step": 29700 }, { "grad_norm": 0.1467711478471756, "learning_rate": 8.421650583204718e-05, "loss": 0.0171, "step": 29710 }, { "grad_norm": 0.14950117468833923, "learning_rate": 8.420444731688633e-05, "loss": 0.0145, "step": 29720 }, { "grad_norm": 0.1932746171951294, "learning_rate": 8.419238506117852e-05, "loss": 0.0152, "step": 29730 }, { "grad_norm": 0.19339387118816376, "learning_rate": 8.418031906624289e-05, "loss": 0.0148, "step": 29740 }, { "grad_norm": 0.19786441326141357, "learning_rate": 8.416824933339898e-05, "loss": 0.0164, "step": 29750 }, { "grad_norm": 0.1732678860425949, "learning_rate": 8.415617586396667e-05, "loss": 0.0159, "step": 29760 }, { "grad_norm": 0.16625866293907166, "learning_rate": 8.414409865926632e-05, "loss": 0.0165, "step": 29770 }, { "grad_norm": 0.1837516725063324, "learning_rate": 8.413201772061867e-05, "loss": 0.0203, "step": 29780 }, { "grad_norm": 0.18587414920330048, "learning_rate": 8.411993304934488e-05, "loss": 0.0179, "step": 29790 }, { "grad_norm": 0.20260903239250183, "learning_rate": 8.410784464676654e-05, "loss": 0.0161, "step": 29800 }, { "grad_norm": 0.1742885857820511, "learning_rate": 8.409575251420556e-05, "loss": 0.0185, "step": 29810 }, { "grad_norm": 0.16956652700901031, "learning_rate": 8.408365665298435e-05, "loss": 0.016, "step": 29820 }, { "grad_norm": 0.15888629853725433, "learning_rate": 8.40715570644257e-05, "loss": 0.016, "step": 29830 }, { "grad_norm": 0.13827019929885864, "learning_rate": 8.40594537498528e-05, "loss": 0.0163, "step": 29840 }, { "grad_norm": 0.16525132954120636, "learning_rate": 8.404734671058924e-05, "loss": 0.0172, "step": 29850 }, { "grad_norm": 0.13175344467163086, "learning_rate": 8.403523594795902e-05, "loss": 0.0164, "step": 29860 }, { "grad_norm": 0.12835396826267242, "learning_rate": 8.402312146328659e-05, "loss": 0.0147, "step": 29870 }, { "grad_norm": 0.1447165608406067, "learning_rate": 8.401100325789675e-05, "loss": 0.0166, "step": 29880 }, { "grad_norm": 0.2652386426925659, "learning_rate": 8.399888133311472e-05, "loss": 0.0183, "step": 29890 }, { "grad_norm": 0.22647452354431152, "learning_rate": 8.398675569026613e-05, "loss": 0.0154, "step": 29900 }, { "grad_norm": 0.14378266036510468, "learning_rate": 8.397462633067705e-05, "loss": 0.0148, "step": 29910 }, { "grad_norm": 0.1785324364900589, "learning_rate": 8.396249325567392e-05, "loss": 0.0158, "step": 29920 }, { "grad_norm": 0.24169696867465973, "learning_rate": 8.395035646658357e-05, "loss": 0.0176, "step": 29930 }, { "grad_norm": 0.18733146786689758, "learning_rate": 8.39382159647333e-05, "loss": 0.0163, "step": 29940 }, { "grad_norm": 0.19404534995555878, "learning_rate": 8.392607175145075e-05, "loss": 0.0163, "step": 29950 }, { "grad_norm": 0.14689168334007263, "learning_rate": 8.3913923828064e-05, "loss": 0.0144, "step": 29960 }, { "grad_norm": 0.22796125710010529, "learning_rate": 8.390177219590152e-05, "loss": 0.0159, "step": 29970 }, { "grad_norm": 0.20174995064735413, "learning_rate": 8.388961685629222e-05, "loss": 0.0174, "step": 29980 }, { "grad_norm": 0.1802302896976471, "learning_rate": 8.387745781056536e-05, "loss": 0.0153, "step": 29990 }, { "grad_norm": 0.15979163348674774, "learning_rate": 8.386529506005065e-05, "loss": 0.0139, "step": 30000 }, { "grad_norm": 0.26494908332824707, "learning_rate": 8.38531286060782e-05, "loss": 0.0181, "step": 30010 }, { "grad_norm": 0.18841779232025146, "learning_rate": 8.384095844997849e-05, "loss": 0.0167, "step": 30020 }, { "grad_norm": 0.2247696816921234, "learning_rate": 8.382878459308245e-05, "loss": 0.0161, "step": 30030 }, { "grad_norm": 0.19535012543201447, "learning_rate": 8.381660703672138e-05, "loss": 0.0168, "step": 30040 }, { "grad_norm": 0.22072258591651917, "learning_rate": 8.380442578222702e-05, "loss": 0.0154, "step": 30050 }, { "grad_norm": 0.2153388410806656, "learning_rate": 8.379224083093146e-05, "loss": 0.0151, "step": 30060 }, { "grad_norm": 0.1701727658510208, "learning_rate": 8.378005218416727e-05, "loss": 0.0178, "step": 30070 }, { "grad_norm": 0.12456992268562317, "learning_rate": 8.376785984326735e-05, "loss": 0.0146, "step": 30080 }, { "grad_norm": 0.2140870988368988, "learning_rate": 8.375566380956506e-05, "loss": 0.0172, "step": 30090 }, { "grad_norm": 0.17022597789764404, "learning_rate": 8.374346408439411e-05, "loss": 0.0195, "step": 30100 }, { "grad_norm": 0.20420688390731812, "learning_rate": 8.373126066908868e-05, "loss": 0.0169, "step": 30110 }, { "grad_norm": 0.21025484800338745, "learning_rate": 8.371905356498326e-05, "loss": 0.0178, "step": 30120 }, { "grad_norm": 0.16355952620506287, "learning_rate": 8.370684277341288e-05, "loss": 0.0185, "step": 30130 }, { "grad_norm": 0.18146780133247375, "learning_rate": 8.369462829571282e-05, "loss": 0.0176, "step": 30140 }, { "grad_norm": 0.18396244943141937, "learning_rate": 8.36824101332189e-05, "loss": 0.0183, "step": 30150 }, { "grad_norm": 0.1750025749206543, "learning_rate": 8.367018828726721e-05, "loss": 0.0149, "step": 30160 }, { "grad_norm": 0.1599869728088379, "learning_rate": 8.365796275919438e-05, "loss": 0.0163, "step": 30170 }, { "grad_norm": 0.14608851075172424, "learning_rate": 8.364573355033734e-05, "loss": 0.0182, "step": 30180 }, { "grad_norm": 0.16775910556316376, "learning_rate": 8.363350066203346e-05, "loss": 0.0143, "step": 30190 }, { "grad_norm": 0.19208791851997375, "learning_rate": 8.362126409562053e-05, "loss": 0.0166, "step": 30200 }, { "grad_norm": 0.37159132957458496, "learning_rate": 8.360902385243667e-05, "loss": 0.0168, "step": 30210 }, { "grad_norm": 0.18910939991474152, "learning_rate": 8.359677993382052e-05, "loss": 0.0146, "step": 30220 }, { "grad_norm": 0.18681477010250092, "learning_rate": 8.358453234111103e-05, "loss": 0.0156, "step": 30230 }, { "grad_norm": 0.24836724996566772, "learning_rate": 8.357228107564756e-05, "loss": 0.018, "step": 30240 }, { "grad_norm": 0.17526789009571075, "learning_rate": 8.356002613876993e-05, "loss": 0.0176, "step": 30250 }, { "grad_norm": 0.16023047268390656, "learning_rate": 8.35477675318183e-05, "loss": 0.0152, "step": 30260 }, { "grad_norm": 0.1511952430009842, "learning_rate": 8.353550525613323e-05, "loss": 0.0179, "step": 30270 }, { "grad_norm": 0.1754503697156906, "learning_rate": 8.352323931305572e-05, "loss": 0.0148, "step": 30280 }, { "grad_norm": 0.17227675020694733, "learning_rate": 8.351096970392717e-05, "loss": 0.0154, "step": 30290 }, { "grad_norm": 0.1947152465581894, "learning_rate": 8.349869643008937e-05, "loss": 0.0178, "step": 30300 }, { "grad_norm": 0.12969863414764404, "learning_rate": 8.348641949288449e-05, "loss": 0.0152, "step": 30310 }, { "grad_norm": 0.1553472876548767, "learning_rate": 8.34741388936551e-05, "loss": 0.0154, "step": 30320 }, { "grad_norm": 0.16220705211162567, "learning_rate": 8.346185463374423e-05, "loss": 0.0182, "step": 30330 }, { "grad_norm": 0.1933119297027588, "learning_rate": 8.344956671449524e-05, "loss": 0.0155, "step": 30340 }, { "grad_norm": 0.15644410252571106, "learning_rate": 8.343727513725192e-05, "loss": 0.0179, "step": 30350 }, { "grad_norm": 0.218192920088768, "learning_rate": 8.342497990335847e-05, "loss": 0.0164, "step": 30360 }, { "grad_norm": 0.1556692272424698, "learning_rate": 8.341268101415946e-05, "loss": 0.0154, "step": 30370 }, { "grad_norm": 0.18056192994117737, "learning_rate": 8.34003784709999e-05, "loss": 0.0146, "step": 30380 }, { "grad_norm": 0.22677238285541534, "learning_rate": 8.338807227522516e-05, "loss": 0.016, "step": 30390 }, { "grad_norm": 0.13335968554019928, "learning_rate": 8.337576242818103e-05, "loss": 0.0186, "step": 30400 }, { "grad_norm": 0.22892381250858307, "learning_rate": 8.336344893121372e-05, "loss": 0.0197, "step": 30410 }, { "grad_norm": 0.21941599249839783, "learning_rate": 8.335113178566977e-05, "loss": 0.0181, "step": 30420 }, { "grad_norm": 0.19374071061611176, "learning_rate": 8.33388109928962e-05, "loss": 0.0155, "step": 30430 }, { "grad_norm": 0.166035458445549, "learning_rate": 8.33264865542404e-05, "loss": 0.0135, "step": 30440 }, { "grad_norm": 0.18606531620025635, "learning_rate": 8.331415847105013e-05, "loss": 0.0183, "step": 30450 }, { "grad_norm": 0.18465837836265564, "learning_rate": 8.330182674467357e-05, "loss": 0.0161, "step": 30460 }, { "grad_norm": 0.16138675808906555, "learning_rate": 8.32894913764593e-05, "loss": 0.0162, "step": 30470 }, { "grad_norm": 0.16312575340270996, "learning_rate": 8.327715236775633e-05, "loss": 0.0155, "step": 30480 }, { "grad_norm": 0.23715467751026154, "learning_rate": 8.326480971991398e-05, "loss": 0.0141, "step": 30490 }, { "grad_norm": 0.15187764167785645, "learning_rate": 8.325246343428206e-05, "loss": 0.0138, "step": 30500 }, { "grad_norm": 0.18151718378067017, "learning_rate": 8.324011351221072e-05, "loss": 0.0135, "step": 30510 }, { "grad_norm": 0.34958329796791077, "learning_rate": 8.322775995505057e-05, "loss": 0.0163, "step": 30520 }, { "grad_norm": 0.19932088255882263, "learning_rate": 8.321540276415254e-05, "loss": 0.0147, "step": 30530 }, { "grad_norm": 0.1585625857114792, "learning_rate": 8.320304194086798e-05, "loss": 0.016, "step": 30540 }, { "grad_norm": 0.20014725625514984, "learning_rate": 8.31906774865487e-05, "loss": 0.0163, "step": 30550 }, { "grad_norm": 0.17686407268047333, "learning_rate": 8.317830940254682e-05, "loss": 0.0171, "step": 30560 }, { "grad_norm": 0.19454334676265717, "learning_rate": 8.316593769021491e-05, "loss": 0.0157, "step": 30570 }, { "grad_norm": 0.31360962986946106, "learning_rate": 8.315356235090592e-05, "loss": 0.0129, "step": 30580 }, { "grad_norm": 0.17946647107601166, "learning_rate": 8.314118338597319e-05, "loss": 0.0152, "step": 30590 }, { "grad_norm": 0.2024093121290207, "learning_rate": 8.312880079677048e-05, "loss": 0.0176, "step": 30600 }, { "grad_norm": 0.17516405880451202, "learning_rate": 8.311641458465191e-05, "loss": 0.0172, "step": 30610 }, { "grad_norm": 0.18330858647823334, "learning_rate": 8.310402475097205e-05, "loss": 0.0177, "step": 30620 }, { "grad_norm": 0.17476509511470795, "learning_rate": 8.309163129708581e-05, "loss": 0.0155, "step": 30630 }, { "grad_norm": 0.15790468454360962, "learning_rate": 8.307923422434852e-05, "loss": 0.0146, "step": 30640 }, { "grad_norm": 0.17972427606582642, "learning_rate": 8.30668335341159e-05, "loss": 0.0142, "step": 30650 }, { "grad_norm": 0.18120770156383514, "learning_rate": 8.305442922774408e-05, "loss": 0.0146, "step": 30660 }, { "grad_norm": 0.1409057378768921, "learning_rate": 8.304202130658959e-05, "loss": 0.0161, "step": 30670 }, { "grad_norm": 0.23433497548103333, "learning_rate": 8.302960977200931e-05, "loss": 0.0169, "step": 30680 }, { "grad_norm": 0.19577433168888092, "learning_rate": 8.301719462536058e-05, "loss": 0.0142, "step": 30690 }, { "grad_norm": 0.14890196919441223, "learning_rate": 8.300477586800108e-05, "loss": 0.0159, "step": 30700 }, { "grad_norm": 0.1312059909105301, "learning_rate": 8.299235350128892e-05, "loss": 0.0182, "step": 30710 }, { "grad_norm": 0.15854108333587646, "learning_rate": 8.297992752658258e-05, "loss": 0.0157, "step": 30720 }, { "grad_norm": 0.1782914251089096, "learning_rate": 8.296749794524094e-05, "loss": 0.0174, "step": 30730 }, { "grad_norm": 0.2129165381193161, "learning_rate": 8.295506475862332e-05, "loss": 0.0139, "step": 30740 }, { "grad_norm": 0.1620725691318512, "learning_rate": 8.294262796808933e-05, "loss": 0.0172, "step": 30750 }, { "grad_norm": 0.1512286216020584, "learning_rate": 8.293018757499909e-05, "loss": 0.0165, "step": 30760 }, { "grad_norm": 0.17268826067447662, "learning_rate": 8.291774358071305e-05, "loss": 0.015, "step": 30770 }, { "grad_norm": 0.12248478084802628, "learning_rate": 8.290529598659205e-05, "loss": 0.015, "step": 30780 }, { "grad_norm": 0.17621992528438568, "learning_rate": 8.289284479399738e-05, "loss": 0.0145, "step": 30790 }, { "grad_norm": 0.2715289294719696, "learning_rate": 8.288039000429064e-05, "loss": 0.0148, "step": 30800 }, { "grad_norm": 0.21631984412670135, "learning_rate": 8.286793161883388e-05, "loss": 0.0147, "step": 30810 }, { "grad_norm": 0.19797539710998535, "learning_rate": 8.285546963898954e-05, "loss": 0.0143, "step": 30820 }, { "grad_norm": 0.17916208505630493, "learning_rate": 8.284300406612044e-05, "loss": 0.0151, "step": 30830 }, { "grad_norm": 0.20537330210208893, "learning_rate": 8.283053490158978e-05, "loss": 0.0134, "step": 30840 }, { "grad_norm": 0.18641479313373566, "learning_rate": 8.28180621467612e-05, "loss": 0.0148, "step": 30850 }, { "grad_norm": 0.16329805552959442, "learning_rate": 8.280558580299867e-05, "loss": 0.0151, "step": 30860 }, { "grad_norm": 0.19668333232402802, "learning_rate": 8.279310587166661e-05, "loss": 0.0121, "step": 30870 }, { "grad_norm": 0.1737566888332367, "learning_rate": 8.278062235412978e-05, "loss": 0.0155, "step": 30880 }, { "grad_norm": 0.1660473346710205, "learning_rate": 8.276813525175339e-05, "loss": 0.0159, "step": 30890 }, { "grad_norm": 0.19628633558750153, "learning_rate": 8.2755644565903e-05, "loss": 0.0185, "step": 30900 }, { "grad_norm": 0.19776377081871033, "learning_rate": 8.274315029794454e-05, "loss": 0.0149, "step": 30910 }, { "grad_norm": 0.15449851751327515, "learning_rate": 8.273065244924443e-05, "loss": 0.0139, "step": 30920 }, { "grad_norm": 0.14520712196826935, "learning_rate": 8.271815102116936e-05, "loss": 0.0156, "step": 30930 }, { "grad_norm": 0.16608451306819916, "learning_rate": 8.270564601508648e-05, "loss": 0.014, "step": 30940 }, { "grad_norm": 0.18448975682258606, "learning_rate": 8.269313743236333e-05, "loss": 0.0176, "step": 30950 }, { "grad_norm": 0.15339182317256927, "learning_rate": 8.268062527436783e-05, "loss": 0.0158, "step": 30960 }, { "grad_norm": 0.1337568163871765, "learning_rate": 8.266810954246828e-05, "loss": 0.012, "step": 30970 }, { "grad_norm": 0.16501113772392273, "learning_rate": 8.265559023803338e-05, "loss": 0.0134, "step": 30980 }, { "grad_norm": 0.15391546487808228, "learning_rate": 8.264306736243223e-05, "loss": 0.0147, "step": 30990 }, { "grad_norm": 0.16111959517002106, "learning_rate": 8.263054091703432e-05, "loss": 0.0144, "step": 31000 }, { "grad_norm": 0.13703390955924988, "learning_rate": 8.26180109032095e-05, "loss": 0.0151, "step": 31010 }, { "grad_norm": 0.1627739816904068, "learning_rate": 8.260547732232807e-05, "loss": 0.0143, "step": 31020 }, { "grad_norm": 0.17614030838012695, "learning_rate": 8.259294017576063e-05, "loss": 0.0163, "step": 31030 }, { "grad_norm": 0.20059141516685486, "learning_rate": 8.258039946487828e-05, "loss": 0.0149, "step": 31040 }, { "grad_norm": 0.2018694430589676, "learning_rate": 8.256785519105241e-05, "loss": 0.0152, "step": 31050 }, { "grad_norm": 0.20354704558849335, "learning_rate": 8.255530735565488e-05, "loss": 0.018, "step": 31060 }, { "grad_norm": 0.14438441395759583, "learning_rate": 8.254275596005787e-05, "loss": 0.0118, "step": 31070 }, { "grad_norm": 0.19342175126075745, "learning_rate": 8.2530201005634e-05, "loss": 0.0164, "step": 31080 }, { "grad_norm": 0.16005538403987885, "learning_rate": 8.251764249375626e-05, "loss": 0.0159, "step": 31090 }, { "grad_norm": 0.2297547310590744, "learning_rate": 8.250508042579803e-05, "loss": 0.0164, "step": 31100 }, { "grad_norm": 0.19554772973060608, "learning_rate": 8.249251480313307e-05, "loss": 0.0154, "step": 31110 }, { "grad_norm": 0.1270834058523178, "learning_rate": 8.247994562713555e-05, "loss": 0.0168, "step": 31120 }, { "grad_norm": 0.17998181283473969, "learning_rate": 8.246737289917998e-05, "loss": 0.0149, "step": 31130 }, { "grad_norm": 0.14658284187316895, "learning_rate": 8.245479662064135e-05, "loss": 0.0144, "step": 31140 }, { "grad_norm": 0.15166844427585602, "learning_rate": 8.244221679289496e-05, "loss": 0.0158, "step": 31150 }, { "grad_norm": 0.15475249290466309, "learning_rate": 8.242963341731652e-05, "loss": 0.0131, "step": 31160 }, { "grad_norm": 0.1646915227174759, "learning_rate": 8.24170464952821e-05, "loss": 0.0154, "step": 31170 }, { "grad_norm": 0.15102925896644592, "learning_rate": 8.240445602816824e-05, "loss": 0.0165, "step": 31180 }, { "grad_norm": 0.19026203453540802, "learning_rate": 8.239186201735179e-05, "loss": 0.018, "step": 31190 }, { "grad_norm": 0.19242146611213684, "learning_rate": 8.237926446420998e-05, "loss": 0.0143, "step": 31200 }, { "grad_norm": 0.12391471117734909, "learning_rate": 8.236666337012052e-05, "loss": 0.014, "step": 31210 }, { "grad_norm": 0.1544303596019745, "learning_rate": 8.23540587364614e-05, "loss": 0.0163, "step": 31220 }, { "grad_norm": 0.15937452018260956, "learning_rate": 8.234145056461107e-05, "loss": 0.0139, "step": 31230 }, { "grad_norm": 0.23647254705429077, "learning_rate": 8.232883885594832e-05, "loss": 0.0138, "step": 31240 }, { "grad_norm": 0.17331752181053162, "learning_rate": 8.231622361185236e-05, "loss": 0.0158, "step": 31250 }, { "grad_norm": 0.18906910717487335, "learning_rate": 8.230360483370278e-05, "loss": 0.0152, "step": 31260 }, { "grad_norm": 0.20502398908138275, "learning_rate": 8.229098252287953e-05, "loss": 0.0168, "step": 31270 }, { "grad_norm": 0.18164314329624176, "learning_rate": 8.2278356680763e-05, "loss": 0.0146, "step": 31280 }, { "grad_norm": 0.22763827443122864, "learning_rate": 8.22657273087339e-05, "loss": 0.0175, "step": 31290 }, { "grad_norm": 0.18696723878383636, "learning_rate": 8.225309440817336e-05, "loss": 0.0138, "step": 31300 }, { "grad_norm": 0.1539563238620758, "learning_rate": 8.224045798046293e-05, "loss": 0.0151, "step": 31310 }, { "grad_norm": 0.24892236292362213, "learning_rate": 8.22278180269845e-05, "loss": 0.0141, "step": 31320 }, { "grad_norm": 0.1557304710149765, "learning_rate": 8.221517454912031e-05, "loss": 0.0176, "step": 31330 }, { "grad_norm": 0.19472208619117737, "learning_rate": 8.220252754825308e-05, "loss": 0.0182, "step": 31340 }, { "grad_norm": 0.17371976375579834, "learning_rate": 8.218987702576586e-05, "loss": 0.0151, "step": 31350 }, { "grad_norm": 0.24540956318378448, "learning_rate": 8.217722298304207e-05, "loss": 0.0144, "step": 31360 }, { "grad_norm": 0.20204849541187286, "learning_rate": 8.216456542146557e-05, "loss": 0.019, "step": 31370 }, { "grad_norm": 0.12114205211400986, "learning_rate": 8.215190434242055e-05, "loss": 0.0136, "step": 31380 }, { "grad_norm": 0.1450125277042389, "learning_rate": 8.213923974729161e-05, "loss": 0.0161, "step": 31390 }, { "grad_norm": 0.154092937707901, "learning_rate": 8.212657163746373e-05, "loss": 0.0142, "step": 31400 }, { "grad_norm": 0.17643529176712036, "learning_rate": 8.211390001432227e-05, "loss": 0.0184, "step": 31410 }, { "grad_norm": 0.1966543197631836, "learning_rate": 8.210122487925297e-05, "loss": 0.0152, "step": 31420 }, { "grad_norm": 0.17242881655693054, "learning_rate": 8.208854623364202e-05, "loss": 0.0163, "step": 31430 }, { "grad_norm": 0.15390175580978394, "learning_rate": 8.207586407887589e-05, "loss": 0.0151, "step": 31440 }, { "grad_norm": 0.12868742644786835, "learning_rate": 8.206317841634148e-05, "loss": 0.0151, "step": 31450 }, { "grad_norm": 0.13852296769618988, "learning_rate": 8.205048924742609e-05, "loss": 0.0126, "step": 31460 }, { "grad_norm": 0.17557910084724426, "learning_rate": 8.203779657351738e-05, "loss": 0.0139, "step": 31470 }, { "grad_norm": 0.15559592843055725, "learning_rate": 8.20251003960034e-05, "loss": 0.0145, "step": 31480 }, { "grad_norm": 0.17601266503334045, "learning_rate": 8.201240071627258e-05, "loss": 0.0145, "step": 31490 }, { "grad_norm": 0.15602700412273407, "learning_rate": 8.199969753571377e-05, "loss": 0.0147, "step": 31500 }, { "grad_norm": 0.22967706620693207, "learning_rate": 8.198699085571615e-05, "loss": 0.0164, "step": 31510 }, { "grad_norm": 0.2002837061882019, "learning_rate": 8.197428067766928e-05, "loss": 0.0157, "step": 31520 }, { "grad_norm": 0.19265441596508026, "learning_rate": 8.196156700296316e-05, "loss": 0.0159, "step": 31530 }, { "grad_norm": 0.15939141809940338, "learning_rate": 8.194884983298814e-05, "loss": 0.0153, "step": 31540 }, { "grad_norm": 0.1576237976551056, "learning_rate": 8.193612916913491e-05, "loss": 0.0167, "step": 31550 }, { "grad_norm": 0.18256130814552307, "learning_rate": 8.192340501279463e-05, "loss": 0.0157, "step": 31560 }, { "grad_norm": 0.20945048332214355, "learning_rate": 8.191067736535876e-05, "loss": 0.0174, "step": 31570 }, { "grad_norm": 0.1720043122768402, "learning_rate": 8.18979462282192e-05, "loss": 0.0157, "step": 31580 }, { "grad_norm": 0.17826807498931885, "learning_rate": 8.188521160276819e-05, "loss": 0.0166, "step": 31590 }, { "grad_norm": 0.17984983325004578, "learning_rate": 8.187247349039837e-05, "loss": 0.0155, "step": 31600 }, { "grad_norm": 0.17282986640930176, "learning_rate": 8.185973189250278e-05, "loss": 0.0145, "step": 31610 }, { "grad_norm": 0.1991591602563858, "learning_rate": 8.184698681047482e-05, "loss": 0.0166, "step": 31620 }, { "grad_norm": 0.1603146493434906, "learning_rate": 8.183423824570827e-05, "loss": 0.0172, "step": 31630 }, { "grad_norm": 0.17735926806926727, "learning_rate": 8.182148619959725e-05, "loss": 0.0149, "step": 31640 }, { "grad_norm": 0.19260947406291962, "learning_rate": 8.180873067353636e-05, "loss": 0.0165, "step": 31650 }, { "grad_norm": 0.21070219576358795, "learning_rate": 8.179597166892052e-05, "loss": 0.0145, "step": 31660 }, { "grad_norm": 0.19795209169387817, "learning_rate": 8.178320918714501e-05, "loss": 0.0159, "step": 31670 }, { "grad_norm": 0.20580974221229553, "learning_rate": 8.177044322960554e-05, "loss": 0.015, "step": 31680 }, { "grad_norm": 0.19451801478862762, "learning_rate": 8.175767379769816e-05, "loss": 0.0155, "step": 31690 }, { "grad_norm": 0.15388616919517517, "learning_rate": 8.174490089281932e-05, "loss": 0.016, "step": 31700 }, { "grad_norm": 0.18346908688545227, "learning_rate": 8.173212451636584e-05, "loss": 0.0142, "step": 31710 }, { "grad_norm": 0.18676967918872833, "learning_rate": 8.171934466973493e-05, "loss": 0.0149, "step": 31720 }, { "grad_norm": 0.19524775445461273, "learning_rate": 8.170656135432418e-05, "loss": 0.0149, "step": 31730 }, { "grad_norm": 0.17607715725898743, "learning_rate": 8.169377457153155e-05, "loss": 0.0191, "step": 31740 }, { "grad_norm": 0.48269927501678467, "learning_rate": 8.168098432275539e-05, "loss": 0.0173, "step": 31750 }, { "grad_norm": 0.1804138869047165, "learning_rate": 8.166819060939442e-05, "loss": 0.0171, "step": 31760 }, { "grad_norm": 0.3432479798793793, "learning_rate": 8.165539343284772e-05, "loss": 0.0272, "step": 31770 }, { "grad_norm": 0.1686263084411621, "learning_rate": 8.16425927945148e-05, "loss": 0.0248, "step": 31780 }, { "grad_norm": 0.19193407893180847, "learning_rate": 8.162978869579551e-05, "loss": 0.0263, "step": 31790 }, { "grad_norm": 0.1944740265607834, "learning_rate": 8.161698113809007e-05, "loss": 0.0217, "step": 31800 }, { "grad_norm": 0.203481525182724, "learning_rate": 8.160417012279911e-05, "loss": 0.018, "step": 31810 }, { "grad_norm": 0.23860688507556915, "learning_rate": 8.159135565132363e-05, "loss": 0.0208, "step": 31820 }, { "grad_norm": 0.11959277093410492, "learning_rate": 8.157853772506498e-05, "loss": 0.0185, "step": 31830 }, { "grad_norm": 0.1534917950630188, "learning_rate": 8.156571634542494e-05, "loss": 0.0186, "step": 31840 }, { "grad_norm": 0.35668665170669556, "learning_rate": 8.15528915138056e-05, "loss": 0.0199, "step": 31850 }, { "grad_norm": 0.24677816033363342, "learning_rate": 8.154006323160949e-05, "loss": 0.0196, "step": 31860 }, { "grad_norm": 0.1993587464094162, "learning_rate": 8.152723150023949e-05, "loss": 0.016, "step": 31870 }, { "grad_norm": 0.2030564546585083, "learning_rate": 8.151439632109886e-05, "loss": 0.0194, "step": 31880 }, { "grad_norm": 0.15859942138195038, "learning_rate": 8.150155769559122e-05, "loss": 0.0205, "step": 31890 }, { "grad_norm": 0.20925109088420868, "learning_rate": 8.148871562512058e-05, "loss": 0.0172, "step": 31900 }, { "grad_norm": 0.2224154770374298, "learning_rate": 8.147587011109136e-05, "loss": 0.0165, "step": 31910 }, { "grad_norm": 0.18425780534744263, "learning_rate": 8.14630211549083e-05, "loss": 0.0146, "step": 31920 }, { "grad_norm": 0.14108781516551971, "learning_rate": 8.145016875797655e-05, "loss": 0.0158, "step": 31930 }, { "grad_norm": 0.16172894835472107, "learning_rate": 8.143731292170164e-05, "loss": 0.0164, "step": 31940 }, { "grad_norm": 0.24833790957927704, "learning_rate": 8.142445364748944e-05, "loss": 0.0189, "step": 31950 }, { "grad_norm": 0.1751924455165863, "learning_rate": 8.141159093674624e-05, "loss": 0.0148, "step": 31960 }, { "grad_norm": 0.23969444632530212, "learning_rate": 8.139872479087869e-05, "loss": 0.0181, "step": 31970 }, { "grad_norm": 0.1859252005815506, "learning_rate": 8.13858552112938e-05, "loss": 0.0168, "step": 31980 }, { "grad_norm": 0.19837401807308197, "learning_rate": 8.137298219939895e-05, "loss": 0.0159, "step": 31990 }, { "grad_norm": 0.1523612141609192, "learning_rate": 8.136010575660196e-05, "loss": 0.0142, "step": 32000 }, { "grad_norm": 0.20317572355270386, "learning_rate": 8.134722588431095e-05, "loss": 0.016, "step": 32010 }, { "grad_norm": 0.1477094292640686, "learning_rate": 8.133434258393444e-05, "loss": 0.0171, "step": 32020 }, { "grad_norm": 0.2100086808204651, "learning_rate": 8.132145585688134e-05, "loss": 0.0176, "step": 32030 }, { "grad_norm": 0.15009231865406036, "learning_rate": 8.130856570456093e-05, "loss": 0.0152, "step": 32040 }, { "grad_norm": 0.19147954881191254, "learning_rate": 8.129567212838283e-05, "loss": 0.0157, "step": 32050 }, { "grad_norm": 0.21285805106163025, "learning_rate": 8.128277512975708e-05, "loss": 0.0143, "step": 32060 }, { "grad_norm": 0.20105478167533875, "learning_rate": 8.126987471009408e-05, "loss": 0.0142, "step": 32070 }, { "grad_norm": 0.27202707529067993, "learning_rate": 8.125697087080459e-05, "loss": 0.0136, "step": 32080 }, { "grad_norm": 0.19778504967689514, "learning_rate": 8.124406361329976e-05, "loss": 0.0154, "step": 32090 }, { "grad_norm": 0.25738462805747986, "learning_rate": 8.123115293899111e-05, "loss": 0.0173, "step": 32100 }, { "grad_norm": 0.27552667260169983, "learning_rate": 8.121823884929055e-05, "loss": 0.016, "step": 32110 }, { "grad_norm": 0.20017032325267792, "learning_rate": 8.12053213456103e-05, "loss": 0.0142, "step": 32120 }, { "grad_norm": 0.12028706073760986, "learning_rate": 8.119240042936303e-05, "loss": 0.0153, "step": 32130 }, { "grad_norm": 0.2107958346605301, "learning_rate": 8.117947610196175e-05, "loss": 0.016, "step": 32140 }, { "grad_norm": 0.2143041044473648, "learning_rate": 8.116654836481982e-05, "loss": 0.0163, "step": 32150 }, { "grad_norm": 0.21609953045845032, "learning_rate": 8.115361721935106e-05, "loss": 0.016, "step": 32160 }, { "grad_norm": 0.19080720841884613, "learning_rate": 8.114068266696953e-05, "loss": 0.0146, "step": 32170 }, { "grad_norm": 0.2126571536064148, "learning_rate": 8.112774470908978e-05, "loss": 0.0148, "step": 32180 }, { "grad_norm": 0.1975114494562149, "learning_rate": 8.111480334712665e-05, "loss": 0.0153, "step": 32190 }, { "grad_norm": 0.2688625156879425, "learning_rate": 8.110185858249542e-05, "loss": 0.0152, "step": 32200 }, { "grad_norm": 0.1518619954586029, "learning_rate": 8.108891041661168e-05, "loss": 0.0134, "step": 32210 }, { "grad_norm": 0.22424966096878052, "learning_rate": 8.107595885089146e-05, "loss": 0.0157, "step": 32220 }, { "grad_norm": 0.19335588812828064, "learning_rate": 8.106300388675108e-05, "loss": 0.0155, "step": 32230 }, { "grad_norm": 0.14676913619041443, "learning_rate": 8.105004552560731e-05, "loss": 0.0142, "step": 32240 }, { "grad_norm": 0.2056763917207718, "learning_rate": 8.103708376887724e-05, "loss": 0.015, "step": 32250 }, { "grad_norm": 0.17469823360443115, "learning_rate": 8.102411861797836e-05, "loss": 0.0148, "step": 32260 }, { "grad_norm": 0.1782679557800293, "learning_rate": 8.101115007432851e-05, "loss": 0.0142, "step": 32270 }, { "grad_norm": 0.17408373951911926, "learning_rate": 8.09981781393459e-05, "loss": 0.0153, "step": 32280 }, { "grad_norm": 0.14171579480171204, "learning_rate": 8.098520281444915e-05, "loss": 0.0131, "step": 32290 }, { "grad_norm": 0.19649513065814972, "learning_rate": 8.09722241010572e-05, "loss": 0.0153, "step": 32300 }, { "grad_norm": 0.19046995043754578, "learning_rate": 8.095924200058939e-05, "loss": 0.0177, "step": 32310 }, { "grad_norm": 0.21270200610160828, "learning_rate": 8.094625651446541e-05, "loss": 0.0158, "step": 32320 }, { "grad_norm": 0.15421302616596222, "learning_rate": 8.093326764410536e-05, "loss": 0.014, "step": 32330 }, { "grad_norm": 0.16651912033557892, "learning_rate": 8.092027539092966e-05, "loss": 0.0148, "step": 32340 }, { "grad_norm": 0.16291497647762299, "learning_rate": 8.090727975635913e-05, "loss": 0.0154, "step": 32350 }, { "grad_norm": 0.17688633501529694, "learning_rate": 8.089428074181497e-05, "loss": 0.0143, "step": 32360 }, { "grad_norm": 0.15947279334068298, "learning_rate": 8.088127834871871e-05, "loss": 0.0133, "step": 32370 }, { "grad_norm": 0.1429196000099182, "learning_rate": 8.086827257849226e-05, "loss": 0.0164, "step": 32380 }, { "grad_norm": 0.14883846044540405, "learning_rate": 8.085526343255795e-05, "loss": 0.0136, "step": 32390 }, { "grad_norm": 0.2642219662666321, "learning_rate": 8.084225091233842e-05, "loss": 0.0168, "step": 32400 }, { "grad_norm": 0.15599234402179718, "learning_rate": 8.082923501925668e-05, "loss": 0.0153, "step": 32410 }, { "grad_norm": 0.13643519580364227, "learning_rate": 8.081621575473617e-05, "loss": 0.0192, "step": 32420 }, { "grad_norm": 0.17255562543869019, "learning_rate": 8.080319312020064e-05, "loss": 0.0155, "step": 32430 }, { "grad_norm": 0.18525002896785736, "learning_rate": 8.079016711707421e-05, "loss": 0.0153, "step": 32440 }, { "grad_norm": 0.1571035385131836, "learning_rate": 8.077713774678139e-05, "loss": 0.0142, "step": 32450 }, { "grad_norm": 0.175939679145813, "learning_rate": 8.076410501074707e-05, "loss": 0.0132, "step": 32460 }, { "grad_norm": 0.20800729095935822, "learning_rate": 8.075106891039647e-05, "loss": 0.0158, "step": 32470 }, { "grad_norm": 0.18984274566173553, "learning_rate": 8.073802944715523e-05, "loss": 0.0131, "step": 32480 }, { "grad_norm": 0.19440309703350067, "learning_rate": 8.072498662244929e-05, "loss": 0.0137, "step": 32490 }, { "grad_norm": 0.15226343274116516, "learning_rate": 8.0711940437705e-05, "loss": 0.0168, "step": 32500 }, { "grad_norm": 0.16336126625537872, "learning_rate": 8.06988908943491e-05, "loss": 0.0155, "step": 32510 }, { "grad_norm": 0.11829130351543427, "learning_rate": 8.068583799380863e-05, "loss": 0.015, "step": 32520 }, { "grad_norm": 0.14615723490715027, "learning_rate": 8.067278173751104e-05, "loss": 0.0168, "step": 32530 }, { "grad_norm": 0.19681070744991302, "learning_rate": 8.065972212688417e-05, "loss": 0.0135, "step": 32540 }, { "grad_norm": 0.17811016738414764, "learning_rate": 8.064665916335618e-05, "loss": 0.0145, "step": 32550 }, { "grad_norm": 0.271340012550354, "learning_rate": 8.063359284835564e-05, "loss": 0.0163, "step": 32560 }, { "grad_norm": 0.23747220635414124, "learning_rate": 8.062052318331142e-05, "loss": 0.0165, "step": 32570 }, { "grad_norm": 0.1534629911184311, "learning_rate": 8.060745016965283e-05, "loss": 0.0162, "step": 32580 }, { "grad_norm": 0.1471238136291504, "learning_rate": 8.059437380880952e-05, "loss": 0.0142, "step": 32590 }, { "grad_norm": 0.1305069625377655, "learning_rate": 8.058129410221146e-05, "loss": 0.0147, "step": 32600 }, { "grad_norm": 0.14852486550807953, "learning_rate": 8.056821105128908e-05, "loss": 0.0159, "step": 32610 }, { "grad_norm": 0.22567546367645264, "learning_rate": 8.05551246574731e-05, "loss": 0.0144, "step": 32620 }, { "grad_norm": 0.18832449615001678, "learning_rate": 8.05420349221946e-05, "loss": 0.0155, "step": 32630 }, { "grad_norm": 0.16330982744693756, "learning_rate": 8.05289418468851e-05, "loss": 0.0141, "step": 32640 }, { "grad_norm": 0.1889074593782425, "learning_rate": 8.051584543297642e-05, "loss": 0.0148, "step": 32650 }, { "grad_norm": 0.15535666048526764, "learning_rate": 8.050274568190074e-05, "loss": 0.0137, "step": 32660 }, { "grad_norm": 0.13887915015220642, "learning_rate": 8.048964259509067e-05, "loss": 0.014, "step": 32670 }, { "grad_norm": 0.15879623591899872, "learning_rate": 8.047653617397914e-05, "loss": 0.0154, "step": 32680 }, { "grad_norm": 0.16474710404872894, "learning_rate": 8.046342641999941e-05, "loss": 0.014, "step": 32690 }, { "grad_norm": 0.16158351302146912, "learning_rate": 8.045031333458517e-05, "loss": 0.0153, "step": 32700 }, { "grad_norm": 0.1568574160337448, "learning_rate": 8.043719691917047e-05, "loss": 0.0145, "step": 32710 }, { "grad_norm": 0.20212547481060028, "learning_rate": 8.042407717518966e-05, "loss": 0.0146, "step": 32720 }, { "grad_norm": 0.16966237127780914, "learning_rate": 8.041095410407751e-05, "loss": 0.0153, "step": 32730 }, { "grad_norm": 0.22541670501232147, "learning_rate": 8.039782770726913e-05, "loss": 0.0148, "step": 32740 }, { "grad_norm": 0.18536075949668884, "learning_rate": 8.038469798620004e-05, "loss": 0.0145, "step": 32750 }, { "grad_norm": 0.17353594303131104, "learning_rate": 8.037156494230604e-05, "loss": 0.0153, "step": 32760 }, { "grad_norm": 0.18191677331924438, "learning_rate": 8.035842857702338e-05, "loss": 0.0161, "step": 32770 }, { "grad_norm": 0.20602431893348694, "learning_rate": 8.03452888917886e-05, "loss": 0.0163, "step": 32780 }, { "grad_norm": 0.1543285995721817, "learning_rate": 8.033214588803866e-05, "loss": 0.0162, "step": 32790 }, { "grad_norm": 0.15285883843898773, "learning_rate": 8.031899956721083e-05, "loss": 0.0133, "step": 32800 }, { "grad_norm": 0.22335182130336761, "learning_rate": 8.030584993074282e-05, "loss": 0.0139, "step": 32810 }, { "grad_norm": 0.1734919548034668, "learning_rate": 8.02926969800726e-05, "loss": 0.0129, "step": 32820 }, { "grad_norm": 0.20559845864772797, "learning_rate": 8.027954071663859e-05, "loss": 0.0121, "step": 32830 }, { "grad_norm": 0.22379522025585175, "learning_rate": 8.026638114187954e-05, "loss": 0.0151, "step": 32840 }, { "grad_norm": 0.23627379536628723, "learning_rate": 8.025321825723456e-05, "loss": 0.0161, "step": 32850 }, { "grad_norm": 0.21184752881526947, "learning_rate": 8.02400520641431e-05, "loss": 0.0129, "step": 32860 }, { "grad_norm": 0.19171567261219025, "learning_rate": 8.022688256404501e-05, "loss": 0.0147, "step": 32870 }, { "grad_norm": 0.17009738087654114, "learning_rate": 8.02137097583805e-05, "loss": 0.0136, "step": 32880 }, { "grad_norm": 0.22765086591243744, "learning_rate": 8.02005336485901e-05, "loss": 0.0154, "step": 32890 }, { "grad_norm": 0.13187077641487122, "learning_rate": 8.018735423611476e-05, "loss": 0.0156, "step": 32900 }, { "grad_norm": 0.17937946319580078, "learning_rate": 8.017417152239574e-05, "loss": 0.0149, "step": 32910 }, { "grad_norm": 0.15052078664302826, "learning_rate": 8.01609855088747e-05, "loss": 0.0122, "step": 32920 }, { "grad_norm": 0.17483708262443542, "learning_rate": 8.014779619699362e-05, "loss": 0.0141, "step": 32930 }, { "grad_norm": 0.1873374879360199, "learning_rate": 8.013460358819489e-05, "loss": 0.013, "step": 32940 }, { "grad_norm": 0.1703493595123291, "learning_rate": 8.01214076839212e-05, "loss": 0.0141, "step": 32950 }, { "grad_norm": 0.15140724182128906, "learning_rate": 8.010820848561565e-05, "loss": 0.0147, "step": 32960 }, { "grad_norm": 0.1784825474023819, "learning_rate": 8.009500599472171e-05, "loss": 0.0152, "step": 32970 }, { "grad_norm": 0.15369434654712677, "learning_rate": 8.008180021268314e-05, "loss": 0.0131, "step": 32980 }, { "grad_norm": 0.25102609395980835, "learning_rate": 8.006859114094414e-05, "loss": 0.0145, "step": 32990 }, { "grad_norm": 0.29085132479667664, "learning_rate": 8.005537878094921e-05, "loss": 0.0143, "step": 33000 }, { "grad_norm": 0.1862347424030304, "learning_rate": 8.004216313414323e-05, "loss": 0.0143, "step": 33010 }, { "grad_norm": 0.15187107026576996, "learning_rate": 8.002894420197149e-05, "loss": 0.015, "step": 33020 }, { "grad_norm": 0.25846317410469055, "learning_rate": 8.001572198587954e-05, "loss": 0.0145, "step": 33030 }, { "grad_norm": 0.1973928064107895, "learning_rate": 8.000249648731338e-05, "loss": 0.0139, "step": 33040 }, { "grad_norm": 0.15751948952674866, "learning_rate": 7.998926770771928e-05, "loss": 0.0144, "step": 33050 }, { "grad_norm": 0.20039138197898865, "learning_rate": 7.997603564854397e-05, "loss": 0.0132, "step": 33060 }, { "grad_norm": 0.21300837397575378, "learning_rate": 7.996280031123448e-05, "loss": 0.013, "step": 33070 }, { "grad_norm": 0.1557224988937378, "learning_rate": 7.994956169723818e-05, "loss": 0.0158, "step": 33080 }, { "grad_norm": 0.15989479422569275, "learning_rate": 7.993631980800285e-05, "loss": 0.0149, "step": 33090 }, { "grad_norm": 0.17827032506465912, "learning_rate": 7.992307464497659e-05, "loss": 0.0149, "step": 33100 }, { "grad_norm": 0.20949572324752808, "learning_rate": 7.990982620960787e-05, "loss": 0.0144, "step": 33110 }, { "grad_norm": 0.12492585182189941, "learning_rate": 7.989657450334554e-05, "loss": 0.013, "step": 33120 }, { "grad_norm": 0.14505071938037872, "learning_rate": 7.988331952763877e-05, "loss": 0.0142, "step": 33130 }, { "grad_norm": 0.15054632723331451, "learning_rate": 7.987006128393709e-05, "loss": 0.0134, "step": 33140 }, { "grad_norm": 0.1911473274230957, "learning_rate": 7.985679977369043e-05, "loss": 0.016, "step": 33150 }, { "grad_norm": 0.14388488233089447, "learning_rate": 7.984353499834902e-05, "loss": 0.0129, "step": 33160 }, { "grad_norm": 0.14675013720989227, "learning_rate": 7.983026695936351e-05, "loss": 0.0156, "step": 33170 }, { "grad_norm": 0.1395319700241089, "learning_rate": 7.981699565818486e-05, "loss": 0.014, "step": 33180 }, { "grad_norm": 0.17663834989070892, "learning_rate": 7.980372109626437e-05, "loss": 0.0131, "step": 33190 }, { "grad_norm": 0.16471198201179504, "learning_rate": 7.979044327505375e-05, "loss": 0.0145, "step": 33200 }, { "grad_norm": 0.19498884677886963, "learning_rate": 7.977716219600506e-05, "loss": 0.0168, "step": 33210 }, { "grad_norm": 0.2967371642589569, "learning_rate": 7.97638778605707e-05, "loss": 0.0157, "step": 33220 }, { "grad_norm": 0.1637161374092102, "learning_rate": 7.975059027020338e-05, "loss": 0.0158, "step": 33230 }, { "grad_norm": 0.13739539682865143, "learning_rate": 7.973729942635623e-05, "loss": 0.0157, "step": 33240 }, { "grad_norm": 0.17723935842514038, "learning_rate": 7.972400533048273e-05, "loss": 0.0159, "step": 33250 }, { "grad_norm": 0.1658935844898224, "learning_rate": 7.97107079840367e-05, "loss": 0.0149, "step": 33260 }, { "grad_norm": 0.1547410637140274, "learning_rate": 7.969740738847231e-05, "loss": 0.0148, "step": 33270 }, { "grad_norm": 0.14714953303337097, "learning_rate": 7.968410354524411e-05, "loss": 0.0116, "step": 33280 }, { "grad_norm": 0.16027608513832092, "learning_rate": 7.967079645580697e-05, "loss": 0.0153, "step": 33290 }, { "grad_norm": 0.1732482612133026, "learning_rate": 7.965748612161612e-05, "loss": 0.0141, "step": 33300 }, { "grad_norm": 0.21139436960220337, "learning_rate": 7.96441725441272e-05, "loss": 0.0152, "step": 33310 }, { "grad_norm": 0.18601001799106598, "learning_rate": 7.963085572479614e-05, "loss": 0.0149, "step": 33320 }, { "grad_norm": 0.15774139761924744, "learning_rate": 7.961753566507924e-05, "loss": 0.0137, "step": 33330 }, { "grad_norm": 0.16732686758041382, "learning_rate": 7.960421236643316e-05, "loss": 0.0141, "step": 33340 }, { "grad_norm": 0.13223132491111755, "learning_rate": 7.959088583031496e-05, "loss": 0.0175, "step": 33350 }, { "grad_norm": 0.1583927869796753, "learning_rate": 7.957755605818194e-05, "loss": 0.0134, "step": 33360 }, { "grad_norm": 0.1755298674106598, "learning_rate": 7.956422305149185e-05, "loss": 0.0128, "step": 33370 }, { "grad_norm": 0.19232334196567535, "learning_rate": 7.95508868117028e-05, "loss": 0.0133, "step": 33380 }, { "grad_norm": 0.19337821006774902, "learning_rate": 7.953754734027318e-05, "loss": 0.0128, "step": 33390 }, { "grad_norm": 0.19248951971530914, "learning_rate": 7.952420463866182e-05, "loss": 0.0125, "step": 33400 }, { "grad_norm": 0.211879163980484, "learning_rate": 7.951085870832782e-05, "loss": 0.0161, "step": 33410 }, { "grad_norm": 0.17556072771549225, "learning_rate": 7.949750955073067e-05, "loss": 0.0144, "step": 33420 }, { "grad_norm": 0.19906118512153625, "learning_rate": 7.948415716733022e-05, "loss": 0.0134, "step": 33430 }, { "grad_norm": 0.2653280794620514, "learning_rate": 7.947080155958669e-05, "loss": 0.0161, "step": 33440 }, { "grad_norm": 0.20398426055908203, "learning_rate": 7.94574427289606e-05, "loss": 0.0144, "step": 33450 }, { "grad_norm": 0.18663163483142853, "learning_rate": 7.944408067691284e-05, "loss": 0.0169, "step": 33460 }, { "grad_norm": 0.15107902884483337, "learning_rate": 7.943071540490473e-05, "loss": 0.015, "step": 33470 }, { "grad_norm": 0.29243019223213196, "learning_rate": 7.94173469143978e-05, "loss": 0.0166, "step": 33480 }, { "grad_norm": 0.22057735919952393, "learning_rate": 7.940397520685406e-05, "loss": 0.0148, "step": 33490 }, { "grad_norm": 0.142525777220726, "learning_rate": 7.939060028373577e-05, "loss": 0.0149, "step": 33500 }, { "grad_norm": 0.1611424833536148, "learning_rate": 7.937722214650565e-05, "loss": 0.0177, "step": 33510 }, { "grad_norm": 0.14813505113124847, "learning_rate": 7.936384079662666e-05, "loss": 0.0143, "step": 33520 }, { "grad_norm": 0.11483567208051682, "learning_rate": 7.93504562355622e-05, "loss": 0.0157, "step": 33530 }, { "grad_norm": 0.1320239007472992, "learning_rate": 7.933706846477599e-05, "loss": 0.0149, "step": 33540 }, { "grad_norm": 0.15639278292655945, "learning_rate": 7.932367748573206e-05, "loss": 0.0155, "step": 33550 }, { "grad_norm": 0.16699790954589844, "learning_rate": 7.931028329989485e-05, "loss": 0.0148, "step": 33560 }, { "grad_norm": 0.18744777143001556, "learning_rate": 7.929688590872913e-05, "loss": 0.0136, "step": 33570 }, { "grad_norm": 0.18321779370307922, "learning_rate": 7.928348531370003e-05, "loss": 0.0157, "step": 33580 }, { "grad_norm": 0.26150253415107727, "learning_rate": 7.927008151627297e-05, "loss": 0.0135, "step": 33590 }, { "grad_norm": 0.17619483172893524, "learning_rate": 7.925667451791383e-05, "loss": 0.0131, "step": 33600 }, { "grad_norm": 0.21822014451026917, "learning_rate": 7.924326432008874e-05, "loss": 0.0139, "step": 33610 }, { "grad_norm": 0.23939873278141022, "learning_rate": 7.922985092426422e-05, "loss": 0.0162, "step": 33620 }, { "grad_norm": 0.13991259038448334, "learning_rate": 7.921643433190717e-05, "loss": 0.0172, "step": 33630 }, { "grad_norm": 0.15614528954029083, "learning_rate": 7.920301454448478e-05, "loss": 0.0148, "step": 33640 }, { "grad_norm": 0.29799237847328186, "learning_rate": 7.918959156346461e-05, "loss": 0.0163, "step": 33650 }, { "grad_norm": 0.11702563613653183, "learning_rate": 7.91761653903146e-05, "loss": 0.0165, "step": 33660 }, { "grad_norm": 0.13458746671676636, "learning_rate": 7.916273602650302e-05, "loss": 0.0154, "step": 33670 }, { "grad_norm": 0.1598140299320221, "learning_rate": 7.914930347349847e-05, "loss": 0.0156, "step": 33680 }, { "grad_norm": 0.17381833493709564, "learning_rate": 7.913586773276992e-05, "loss": 0.0134, "step": 33690 }, { "grad_norm": 0.6591023802757263, "learning_rate": 7.912242880578667e-05, "loss": 0.0214, "step": 33700 }, { "grad_norm": 0.2051892876625061, "learning_rate": 7.910898669401839e-05, "loss": 0.0181, "step": 33710 }, { "grad_norm": 0.18071943521499634, "learning_rate": 7.909554139893511e-05, "loss": 0.0199, "step": 33720 }, { "grad_norm": 0.32397720217704773, "learning_rate": 7.908209292200715e-05, "loss": 0.0162, "step": 33730 }, { "grad_norm": 0.20063862204551697, "learning_rate": 7.906864126470523e-05, "loss": 0.0181, "step": 33740 }, { "grad_norm": 0.21661239862442017, "learning_rate": 7.905518642850041e-05, "loss": 0.0164, "step": 33750 }, { "grad_norm": 0.23716731369495392, "learning_rate": 7.904172841486409e-05, "loss": 0.0158, "step": 33760 }, { "grad_norm": 0.3158196210861206, "learning_rate": 7.902826722526801e-05, "loss": 0.0138, "step": 33770 }, { "grad_norm": 0.28388309478759766, "learning_rate": 7.901480286118427e-05, "loss": 0.0163, "step": 33780 }, { "grad_norm": 0.21246756613254547, "learning_rate": 7.900133532408531e-05, "loss": 0.0159, "step": 33790 }, { "grad_norm": 0.16350339353084564, "learning_rate": 7.898786461544395e-05, "loss": 0.0171, "step": 33800 }, { "grad_norm": 0.16849097609519958, "learning_rate": 7.897439073673325e-05, "loss": 0.0152, "step": 33810 }, { "grad_norm": 0.18358270823955536, "learning_rate": 7.896091368942677e-05, "loss": 0.0144, "step": 33820 }, { "grad_norm": 0.15946216881275177, "learning_rate": 7.894743347499832e-05, "loss": 0.0146, "step": 33830 }, { "grad_norm": 0.13975274562835693, "learning_rate": 7.893395009492203e-05, "loss": 0.0162, "step": 33840 }, { "grad_norm": 0.16669926047325134, "learning_rate": 7.892046355067248e-05, "loss": 0.0145, "step": 33850 }, { "grad_norm": 0.25397077202796936, "learning_rate": 7.890697384372451e-05, "loss": 0.0165, "step": 33860 }, { "grad_norm": 0.12731227278709412, "learning_rate": 7.889348097555336e-05, "loss": 0.0166, "step": 33870 }, { "grad_norm": 0.16239696741104126, "learning_rate": 7.887998494763455e-05, "loss": 0.0134, "step": 33880 }, { "grad_norm": 0.12825369834899902, "learning_rate": 7.886648576144404e-05, "loss": 0.017, "step": 33890 }, { "grad_norm": 0.17131565511226654, "learning_rate": 7.885298341845802e-05, "loss": 0.0162, "step": 33900 }, { "grad_norm": 0.15066289901733398, "learning_rate": 7.883947792015311e-05, "loss": 0.0144, "step": 33910 }, { "grad_norm": 0.1465950906276703, "learning_rate": 7.882596926800628e-05, "loss": 0.0138, "step": 33920 }, { "grad_norm": 0.1256878525018692, "learning_rate": 7.881245746349477e-05, "loss": 0.0139, "step": 33930 }, { "grad_norm": 0.15602804720401764, "learning_rate": 7.879894250809623e-05, "loss": 0.0167, "step": 33940 }, { "grad_norm": 0.20874333381652832, "learning_rate": 7.878542440328865e-05, "loss": 0.016, "step": 33950 }, { "grad_norm": 0.20087984204292297, "learning_rate": 7.877190315055031e-05, "loss": 0.0152, "step": 33960 }, { "grad_norm": 0.3356994092464447, "learning_rate": 7.875837875135991e-05, "loss": 0.0142, "step": 33970 }, { "grad_norm": 0.16900070011615753, "learning_rate": 7.874485120719646e-05, "loss": 0.0144, "step": 33980 }, { "grad_norm": 0.13074363768100739, "learning_rate": 7.873132051953928e-05, "loss": 0.0134, "step": 33990 }, { "grad_norm": 0.21708983182907104, "learning_rate": 7.87177866898681e-05, "loss": 0.0133, "step": 34000 }, { "grad_norm": 0.17895692586898804, "learning_rate": 7.870424971966294e-05, "loss": 0.0153, "step": 34010 }, { "grad_norm": 0.17542827129364014, "learning_rate": 7.869070961040419e-05, "loss": 0.0135, "step": 34020 }, { "grad_norm": 0.4687327742576599, "learning_rate": 7.867716636357257e-05, "loss": 0.0176, "step": 34030 }, { "grad_norm": 0.18599551916122437, "learning_rate": 7.866361998064915e-05, "loss": 0.016, "step": 34040 }, { "grad_norm": 0.1540157049894333, "learning_rate": 7.865007046311534e-05, "loss": 0.0188, "step": 34050 }, { "grad_norm": 0.27765780687332153, "learning_rate": 7.86365178124529e-05, "loss": 0.016, "step": 34060 }, { "grad_norm": 0.1793532371520996, "learning_rate": 7.862296203014394e-05, "loss": 0.0128, "step": 34070 }, { "grad_norm": 0.19933322072029114, "learning_rate": 7.860940311767088e-05, "loss": 0.0151, "step": 34080 }, { "grad_norm": 0.13782159984111786, "learning_rate": 7.85958410765165e-05, "loss": 0.0178, "step": 34090 }, { "grad_norm": 0.15716233849525452, "learning_rate": 7.858227590816394e-05, "loss": 0.0135, "step": 34100 }, { "grad_norm": 0.1408965289592743, "learning_rate": 7.856870761409664e-05, "loss": 0.0148, "step": 34110 }, { "grad_norm": 0.12518630921840668, "learning_rate": 7.855513619579846e-05, "loss": 0.013, "step": 34120 }, { "grad_norm": 0.1453227400779724, "learning_rate": 7.85415616547535e-05, "loss": 0.0131, "step": 34130 }, { "grad_norm": 0.2197207659482956, "learning_rate": 7.852798399244627e-05, "loss": 0.0154, "step": 34140 }, { "grad_norm": 0.17630240321159363, "learning_rate": 7.851440321036161e-05, "loss": 0.0141, "step": 34150 }, { "grad_norm": 0.21952879428863525, "learning_rate": 7.850081930998468e-05, "loss": 0.0139, "step": 34160 }, { "grad_norm": 0.23304876685142517, "learning_rate": 7.8487232292801e-05, "loss": 0.0145, "step": 34170 }, { "grad_norm": 0.1521960198879242, "learning_rate": 7.847364216029642e-05, "loss": 0.0146, "step": 34180 }, { "grad_norm": 0.16541028022766113, "learning_rate": 7.846004891395716e-05, "loss": 0.0131, "step": 34190 }, { "grad_norm": 0.3185485005378723, "learning_rate": 7.844645255526972e-05, "loss": 0.0174, "step": 34200 }, { "grad_norm": 0.14373581111431122, "learning_rate": 7.843285308572101e-05, "loss": 0.0139, "step": 34210 }, { "grad_norm": 0.17573387920856476, "learning_rate": 7.841925050679823e-05, "loss": 0.0128, "step": 34220 }, { "grad_norm": 0.22015602886676788, "learning_rate": 7.840564481998895e-05, "loss": 0.0143, "step": 34230 }, { "grad_norm": 0.17567142844200134, "learning_rate": 7.839203602678105e-05, "loss": 0.013, "step": 34240 }, { "grad_norm": 0.16755737364292145, "learning_rate": 7.837842412866279e-05, "loss": 0.0123, "step": 34250 }, { "grad_norm": 0.26900267601013184, "learning_rate": 7.836480912712272e-05, "loss": 0.0134, "step": 34260 }, { "grad_norm": 0.16157200932502747, "learning_rate": 7.835119102364976e-05, "loss": 0.0134, "step": 34270 }, { "grad_norm": 0.16734035313129425, "learning_rate": 7.83375698197332e-05, "loss": 0.0152, "step": 34280 }, { "grad_norm": 0.15545828640460968, "learning_rate": 7.83239455168626e-05, "loss": 0.0121, "step": 34290 }, { "grad_norm": 0.15603268146514893, "learning_rate": 7.83103181165279e-05, "loss": 0.0152, "step": 34300 }, { "grad_norm": 0.17913790047168732, "learning_rate": 7.829668762021937e-05, "loss": 0.016, "step": 34310 }, { "grad_norm": 0.16922754049301147, "learning_rate": 7.828305402942764e-05, "loss": 0.0146, "step": 34320 }, { "grad_norm": 0.1468348354101181, "learning_rate": 7.826941734564363e-05, "loss": 0.0164, "step": 34330 }, { "grad_norm": 0.19745205342769623, "learning_rate": 7.825577757035865e-05, "loss": 0.0183, "step": 34340 }, { "grad_norm": 0.16385777294635773, "learning_rate": 7.824213470506431e-05, "loss": 0.0178, "step": 34350 }, { "grad_norm": 0.1593979299068451, "learning_rate": 7.822848875125257e-05, "loss": 0.0142, "step": 34360 }, { "grad_norm": 0.1710788756608963, "learning_rate": 7.821483971041576e-05, "loss": 0.0152, "step": 34370 }, { "grad_norm": 0.17817895114421844, "learning_rate": 7.820118758404649e-05, "loss": 0.0168, "step": 34380 }, { "grad_norm": 0.14576783776283264, "learning_rate": 7.818753237363776e-05, "loss": 0.0162, "step": 34390 }, { "grad_norm": 0.166025772690773, "learning_rate": 7.817387408068286e-05, "loss": 0.0167, "step": 34400 }, { "grad_norm": 0.1991625428199768, "learning_rate": 7.816021270667544e-05, "loss": 0.0184, "step": 34410 }, { "grad_norm": 0.2970562279224396, "learning_rate": 7.81465482531095e-05, "loss": 0.0133, "step": 34420 }, { "grad_norm": 0.18807750940322876, "learning_rate": 7.813288072147938e-05, "loss": 0.0139, "step": 34430 }, { "grad_norm": 0.1453501135110855, "learning_rate": 7.811921011327972e-05, "loss": 0.0156, "step": 34440 }, { "grad_norm": 0.1761753261089325, "learning_rate": 7.810553643000549e-05, "loss": 0.0143, "step": 34450 }, { "grad_norm": 0.151766836643219, "learning_rate": 7.809185967315206e-05, "loss": 0.014, "step": 34460 }, { "grad_norm": 0.20370997488498688, "learning_rate": 7.80781798442151e-05, "loss": 0.0136, "step": 34470 }, { "grad_norm": 0.15696024894714355, "learning_rate": 7.806449694469061e-05, "loss": 0.0145, "step": 34480 }, { "grad_norm": 0.2754785418510437, "learning_rate": 7.805081097607492e-05, "loss": 0.0141, "step": 34490 }, { "grad_norm": 0.3351423442363739, "learning_rate": 7.803712193986474e-05, "loss": 0.0162, "step": 34500 }, { "grad_norm": 0.20693613588809967, "learning_rate": 7.802342983755702e-05, "loss": 0.0152, "step": 34510 }, { "grad_norm": 0.2374124825000763, "learning_rate": 7.800973467064918e-05, "loss": 0.0136, "step": 34520 }, { "grad_norm": 0.16519396007061005, "learning_rate": 7.799603644063884e-05, "loss": 0.0146, "step": 34530 }, { "grad_norm": 0.20835110545158386, "learning_rate": 7.798233514902405e-05, "loss": 0.0174, "step": 34540 }, { "grad_norm": 0.1966121941804886, "learning_rate": 7.796863079730318e-05, "loss": 0.0156, "step": 34550 }, { "grad_norm": 0.19138382375240326, "learning_rate": 7.795492338697488e-05, "loss": 0.0139, "step": 34560 }, { "grad_norm": 0.16243568062782288, "learning_rate": 7.794121291953819e-05, "loss": 0.0129, "step": 34570 }, { "grad_norm": 0.13846305012702942, "learning_rate": 7.792749939649246e-05, "loss": 0.0137, "step": 34580 }, { "grad_norm": 0.1795508712530136, "learning_rate": 7.79137828193374e-05, "loss": 0.0172, "step": 34590 }, { "grad_norm": 0.14406399428844452, "learning_rate": 7.790006318957301e-05, "loss": 0.0156, "step": 34600 }, { "grad_norm": 0.15586768090724945, "learning_rate": 7.788634050869965e-05, "loss": 0.0135, "step": 34610 }, { "grad_norm": 0.1513211876153946, "learning_rate": 7.787261477821803e-05, "loss": 0.0145, "step": 34620 }, { "grad_norm": 0.2206203043460846, "learning_rate": 7.785888599962916e-05, "loss": 0.0135, "step": 34630 }, { "grad_norm": 0.2243809849023819, "learning_rate": 7.784515417443439e-05, "loss": 0.0152, "step": 34640 }, { "grad_norm": 0.1543043702840805, "learning_rate": 7.783141930413545e-05, "loss": 0.0147, "step": 34650 }, { "grad_norm": 0.25674423575401306, "learning_rate": 7.78176813902343e-05, "loss": 0.0139, "step": 34660 }, { "grad_norm": 0.20798330008983612, "learning_rate": 7.780394043423336e-05, "loss": 0.0136, "step": 34670 }, { "grad_norm": 0.21307694911956787, "learning_rate": 7.77901964376353e-05, "loss": 0.0137, "step": 34680 }, { "grad_norm": 0.16373321413993835, "learning_rate": 7.777644940194316e-05, "loss": 0.0126, "step": 34690 }, { "grad_norm": 0.17887970805168152, "learning_rate": 7.776269932866023e-05, "loss": 0.0136, "step": 34700 }, { "grad_norm": 0.23417624831199646, "learning_rate": 7.774894621929026e-05, "loss": 0.0126, "step": 34710 }, { "grad_norm": 0.18702717125415802, "learning_rate": 7.773519007533725e-05, "loss": 0.0138, "step": 34720 }, { "grad_norm": 0.1759510338306427, "learning_rate": 7.772143089830556e-05, "loss": 0.0149, "step": 34730 }, { "grad_norm": 0.15721885859966278, "learning_rate": 7.770766868969985e-05, "loss": 0.015, "step": 34740 }, { "grad_norm": 0.18192678689956665, "learning_rate": 7.769390345102518e-05, "loss": 0.0164, "step": 34750 }, { "grad_norm": 0.21939609944820404, "learning_rate": 7.768013518378683e-05, "loss": 0.0164, "step": 34760 }, { "grad_norm": 0.20544004440307617, "learning_rate": 7.766636388949053e-05, "loss": 0.0142, "step": 34770 }, { "grad_norm": 0.1596134603023529, "learning_rate": 7.765258956964229e-05, "loss": 0.0165, "step": 34780 }, { "grad_norm": 0.20868335664272308, "learning_rate": 7.76388122257484e-05, "loss": 0.0154, "step": 34790 }, { "grad_norm": 0.16355514526367188, "learning_rate": 7.762503185931558e-05, "loss": 0.0138, "step": 34800 }, { "grad_norm": 0.16891498863697052, "learning_rate": 7.76112484718508e-05, "loss": 0.0161, "step": 34810 }, { "grad_norm": 0.18612991273403168, "learning_rate": 7.75974620648614e-05, "loss": 0.0145, "step": 34820 }, { "grad_norm": 0.158577561378479, "learning_rate": 7.758367263985503e-05, "loss": 0.0145, "step": 34830 }, { "grad_norm": 0.19887077808380127, "learning_rate": 7.75698801983397e-05, "loss": 0.0128, "step": 34840 }, { "grad_norm": 0.1192111149430275, "learning_rate": 7.755608474182372e-05, "loss": 0.0135, "step": 34850 }, { "grad_norm": 0.15652742981910706, "learning_rate": 7.754228627181574e-05, "loss": 0.0142, "step": 34860 }, { "grad_norm": 0.16001366078853607, "learning_rate": 7.752848478982476e-05, "loss": 0.0138, "step": 34870 }, { "grad_norm": 0.17078272998332977, "learning_rate": 7.751468029736006e-05, "loss": 0.014, "step": 34880 }, { "grad_norm": 0.14812777936458588, "learning_rate": 7.750087279593129e-05, "loss": 0.014, "step": 34890 }, { "grad_norm": 0.15784281492233276, "learning_rate": 7.748706228704843e-05, "loss": 0.0143, "step": 34900 }, { "grad_norm": 0.1338375061750412, "learning_rate": 7.747324877222176e-05, "loss": 0.0147, "step": 34910 }, { "grad_norm": 0.15310783684253693, "learning_rate": 7.745943225296188e-05, "loss": 0.0133, "step": 34920 }, { "grad_norm": 0.14872859418392181, "learning_rate": 7.744561273077981e-05, "loss": 0.0134, "step": 34930 }, { "grad_norm": 0.21659334003925323, "learning_rate": 7.743179020718678e-05, "loss": 0.0141, "step": 34940 }, { "grad_norm": 0.16917960345745087, "learning_rate": 7.741796468369443e-05, "loss": 0.015, "step": 34950 }, { "grad_norm": 0.13248375058174133, "learning_rate": 7.740413616181466e-05, "loss": 0.0127, "step": 34960 }, { "grad_norm": 0.14616520702838898, "learning_rate": 7.739030464305978e-05, "loss": 0.0144, "step": 34970 }, { "grad_norm": 0.21714571118354797, "learning_rate": 7.737647012894235e-05, "loss": 0.0122, "step": 34980 }, { "grad_norm": 0.14599958062171936, "learning_rate": 7.736263262097532e-05, "loss": 0.0131, "step": 34990 }, { "grad_norm": 0.17108109593391418, "learning_rate": 7.734879212067192e-05, "loss": 0.0139, "step": 35000 }, { "grad_norm": 0.1609138399362564, "learning_rate": 7.733494862954573e-05, "loss": 0.0114, "step": 35010 }, { "grad_norm": 0.1982639729976654, "learning_rate": 7.732110214911066e-05, "loss": 0.014, "step": 35020 }, { "grad_norm": 0.38383039832115173, "learning_rate": 7.730725268088094e-05, "loss": 0.0163, "step": 35030 }, { "grad_norm": 0.2096216380596161, "learning_rate": 7.729340022637112e-05, "loss": 0.0131, "step": 35040 }, { "grad_norm": 0.2506135404109955, "learning_rate": 7.727954478709607e-05, "loss": 0.0165, "step": 35050 }, { "grad_norm": 0.1431780904531479, "learning_rate": 7.726568636457103e-05, "loss": 0.0154, "step": 35060 }, { "grad_norm": 0.16545182466506958, "learning_rate": 7.725182496031153e-05, "loss": 0.0151, "step": 35070 }, { "grad_norm": 0.16305659711360931, "learning_rate": 7.723796057583341e-05, "loss": 0.0139, "step": 35080 }, { "grad_norm": 0.2022865265607834, "learning_rate": 7.722409321265287e-05, "loss": 0.0135, "step": 35090 }, { "grad_norm": 0.19079755246639252, "learning_rate": 7.721022287228645e-05, "loss": 0.0142, "step": 35100 }, { "grad_norm": 0.1166379377245903, "learning_rate": 7.719634955625094e-05, "loss": 0.013, "step": 35110 }, { "grad_norm": 0.18558187782764435, "learning_rate": 7.718247326606355e-05, "loss": 0.0144, "step": 35120 }, { "grad_norm": 0.16050985455513, "learning_rate": 7.716859400324175e-05, "loss": 0.0127, "step": 35130 }, { "grad_norm": 0.1435326337814331, "learning_rate": 7.715471176930335e-05, "loss": 0.0121, "step": 35140 }, { "grad_norm": 0.18248535692691803, "learning_rate": 7.714082656576651e-05, "loss": 0.0125, "step": 35150 }, { "grad_norm": 0.13640040159225464, "learning_rate": 7.712693839414968e-05, "loss": 0.0139, "step": 35160 }, { "grad_norm": 0.2126043438911438, "learning_rate": 7.711304725597164e-05, "loss": 0.0114, "step": 35170 }, { "grad_norm": 0.28144416213035583, "learning_rate": 7.709915315275151e-05, "loss": 0.0129, "step": 35180 }, { "grad_norm": 0.16828137636184692, "learning_rate": 7.708525608600876e-05, "loss": 0.0124, "step": 35190 }, { "grad_norm": 0.1506800502538681, "learning_rate": 7.707135605726311e-05, "loss": 0.0142, "step": 35200 }, { "grad_norm": 0.1855095624923706, "learning_rate": 7.705745306803466e-05, "loss": 0.0155, "step": 35210 }, { "grad_norm": 0.20714178681373596, "learning_rate": 7.704354711984383e-05, "loss": 0.0136, "step": 35220 }, { "grad_norm": 0.14834854006767273, "learning_rate": 7.702963821421133e-05, "loss": 0.0146, "step": 35230 }, { "grad_norm": 0.21207574009895325, "learning_rate": 7.701572635265826e-05, "loss": 0.0154, "step": 35240 }, { "grad_norm": 0.16429801285266876, "learning_rate": 7.700181153670596e-05, "loss": 0.0118, "step": 35250 }, { "grad_norm": 0.1362384408712387, "learning_rate": 7.698789376787616e-05, "loss": 0.0126, "step": 35260 }, { "grad_norm": 0.22184564173221588, "learning_rate": 7.697397304769085e-05, "loss": 0.0126, "step": 35270 }, { "grad_norm": 0.15903039276599884, "learning_rate": 7.696004937767241e-05, "loss": 0.0149, "step": 35280 }, { "grad_norm": 0.18856707215309143, "learning_rate": 7.694612275934352e-05, "loss": 0.0128, "step": 35290 }, { "grad_norm": 0.23780091106891632, "learning_rate": 7.693219319422714e-05, "loss": 0.0151, "step": 35300 }, { "grad_norm": 0.14549075067043304, "learning_rate": 7.69182606838466e-05, "loss": 0.0119, "step": 35310 }, { "grad_norm": 0.21003444492816925, "learning_rate": 7.690432522972558e-05, "loss": 0.0129, "step": 35320 }, { "grad_norm": 0.12583290040493011, "learning_rate": 7.689038683338796e-05, "loss": 0.0153, "step": 35330 }, { "grad_norm": 0.11715053766965866, "learning_rate": 7.687644549635808e-05, "loss": 0.0137, "step": 35340 }, { "grad_norm": 0.138309508562088, "learning_rate": 7.686250122016053e-05, "loss": 0.0128, "step": 35350 }, { "grad_norm": 0.15896770358085632, "learning_rate": 7.684855400632022e-05, "loss": 0.0136, "step": 35360 }, { "grad_norm": 0.13216185569763184, "learning_rate": 7.683460385636243e-05, "loss": 0.0147, "step": 35370 }, { "grad_norm": 0.23249955475330353, "learning_rate": 7.68206507718127e-05, "loss": 0.0152, "step": 35380 }, { "grad_norm": 0.18583796918392181, "learning_rate": 7.680669475419693e-05, "loss": 0.0123, "step": 35390 }, { "grad_norm": 0.2042534202337265, "learning_rate": 7.679273580504132e-05, "loss": 0.014, "step": 35400 }, { "grad_norm": 0.17449970543384552, "learning_rate": 7.67787739258724e-05, "loss": 0.0155, "step": 35410 }, { "grad_norm": 0.15031693875789642, "learning_rate": 7.676480911821705e-05, "loss": 0.0141, "step": 35420 }, { "grad_norm": 0.16488878428936005, "learning_rate": 7.675084138360238e-05, "loss": 0.014, "step": 35430 }, { "grad_norm": 0.18393053114414215, "learning_rate": 7.673687072355592e-05, "loss": 0.0148, "step": 35440 }, { "grad_norm": 0.1717911958694458, "learning_rate": 7.67228971396055e-05, "loss": 0.0133, "step": 35450 }, { "grad_norm": 0.21642333269119263, "learning_rate": 7.670892063327922e-05, "loss": 0.013, "step": 35460 }, { "grad_norm": 0.17347677052021027, "learning_rate": 7.669494120610552e-05, "loss": 0.0141, "step": 35470 }, { "grad_norm": 0.1746087372303009, "learning_rate": 7.66809588596132e-05, "loss": 0.0121, "step": 35480 }, { "grad_norm": 0.19100461900234222, "learning_rate": 7.666697359533132e-05, "loss": 0.0126, "step": 35490 }, { "grad_norm": 0.15247568488121033, "learning_rate": 7.665298541478932e-05, "loss": 0.0136, "step": 35500 }, { "grad_norm": 0.1739262342453003, "learning_rate": 7.66389943195169e-05, "loss": 0.0128, "step": 35510 }, { "grad_norm": 0.16885749995708466, "learning_rate": 7.66250003110441e-05, "loss": 0.0157, "step": 35520 }, { "grad_norm": 0.15543551743030548, "learning_rate": 7.66110033909013e-05, "loss": 0.0133, "step": 35530 }, { "grad_norm": 0.26512792706489563, "learning_rate": 7.659700356061918e-05, "loss": 0.0144, "step": 35540 }, { "grad_norm": 0.1761859804391861, "learning_rate": 7.658300082172875e-05, "loss": 0.0135, "step": 35550 }, { "grad_norm": 0.1618853062391281, "learning_rate": 7.65689951757613e-05, "loss": 0.0146, "step": 35560 }, { "grad_norm": 0.17328031361103058, "learning_rate": 7.655498662424849e-05, "loss": 0.0124, "step": 35570 }, { "grad_norm": 0.19559606909751892, "learning_rate": 7.654097516872227e-05, "loss": 0.0152, "step": 35580 }, { "grad_norm": 0.19374454021453857, "learning_rate": 7.65269608107149e-05, "loss": 0.014, "step": 35590 }, { "grad_norm": 0.178702250123024, "learning_rate": 7.6512943551759e-05, "loss": 0.015, "step": 35600 }, { "grad_norm": 0.15228214859962463, "learning_rate": 7.649892339338743e-05, "loss": 0.0166, "step": 35610 }, { "grad_norm": 0.13932929933071136, "learning_rate": 7.648490033713344e-05, "loss": 0.0137, "step": 35620 }, { "grad_norm": 0.20730161666870117, "learning_rate": 7.647087438453058e-05, "loss": 0.0139, "step": 35630 }, { "grad_norm": 0.15678660571575165, "learning_rate": 7.64568455371127e-05, "loss": 0.0138, "step": 35640 }, { "grad_norm": 0.204858660697937, "learning_rate": 7.644281379641396e-05, "loss": 0.0135, "step": 35650 }, { "grad_norm": 0.17764557898044586, "learning_rate": 7.642877916396887e-05, "loss": 0.0147, "step": 35660 }, { "grad_norm": 0.1710328459739685, "learning_rate": 7.641474164131221e-05, "loss": 0.014, "step": 35670 }, { "grad_norm": 0.18283089995384216, "learning_rate": 7.640070122997913e-05, "loss": 0.0142, "step": 35680 }, { "grad_norm": 0.21154682338237762, "learning_rate": 7.638665793150505e-05, "loss": 0.0162, "step": 35690 }, { "grad_norm": 0.20565971732139587, "learning_rate": 7.637261174742574e-05, "loss": 0.0126, "step": 35700 }, { "grad_norm": 0.18475551903247833, "learning_rate": 7.635856267927725e-05, "loss": 0.0143, "step": 35710 }, { "grad_norm": 0.1668654978275299, "learning_rate": 7.634451072859597e-05, "loss": 0.0142, "step": 35720 }, { "grad_norm": 0.12133143097162247, "learning_rate": 7.633045589691863e-05, "loss": 0.0134, "step": 35730 }, { "grad_norm": 0.16160905361175537, "learning_rate": 7.63163981857822e-05, "loss": 0.0134, "step": 35740 }, { "grad_norm": 0.1667315810918808, "learning_rate": 7.630233759672403e-05, "loss": 0.0134, "step": 35750 }, { "grad_norm": 0.17531529068946838, "learning_rate": 7.628827413128177e-05, "loss": 0.013, "step": 35760 }, { "grad_norm": 0.2219262719154358, "learning_rate": 7.627420779099338e-05, "loss": 0.0138, "step": 35770 }, { "grad_norm": 0.18601252138614655, "learning_rate": 7.626013857739711e-05, "loss": 0.0116, "step": 35780 }, { "grad_norm": 0.16295938193798065, "learning_rate": 7.624606649203158e-05, "loss": 0.0134, "step": 35790 }, { "grad_norm": 0.16517972946166992, "learning_rate": 7.623199153643569e-05, "loss": 0.0134, "step": 35800 }, { "grad_norm": 0.20859067142009735, "learning_rate": 7.621791371214863e-05, "loss": 0.0151, "step": 35810 }, { "grad_norm": 0.31960368156433105, "learning_rate": 7.620383302070995e-05, "loss": 0.0169, "step": 35820 }, { "grad_norm": 0.17969003319740295, "learning_rate": 7.61897494636595e-05, "loss": 0.0144, "step": 35830 }, { "grad_norm": 0.2846548855304718, "learning_rate": 7.617566304253739e-05, "loss": 0.0138, "step": 35840 }, { "grad_norm": 0.14043661952018738, "learning_rate": 7.616157375888416e-05, "loss": 0.0136, "step": 35850 }, { "grad_norm": 0.23348429799079895, "learning_rate": 7.614748161424053e-05, "loss": 0.0127, "step": 35860 }, { "grad_norm": 0.21975232660770416, "learning_rate": 7.613338661014763e-05, "loss": 0.0158, "step": 35870 }, { "grad_norm": 0.17442703247070312, "learning_rate": 7.611928874814686e-05, "loss": 0.0138, "step": 35880 }, { "grad_norm": 0.14443786442279816, "learning_rate": 7.610518802977996e-05, "loss": 0.0145, "step": 35890 }, { "grad_norm": 0.14211222529411316, "learning_rate": 7.609108445658893e-05, "loss": 0.0146, "step": 35900 }, { "grad_norm": 0.17363113164901733, "learning_rate": 7.607697803011612e-05, "loss": 0.0144, "step": 35910 }, { "grad_norm": 0.1863177865743637, "learning_rate": 7.606286875190421e-05, "loss": 0.0163, "step": 35920 }, { "grad_norm": 0.14584575593471527, "learning_rate": 7.604875662349617e-05, "loss": 0.0142, "step": 35930 }, { "grad_norm": 0.21464332938194275, "learning_rate": 7.603464164643525e-05, "loss": 0.0117, "step": 35940 }, { "grad_norm": 0.155022531747818, "learning_rate": 7.602052382226507e-05, "loss": 0.0133, "step": 35950 }, { "grad_norm": 0.10530497133731842, "learning_rate": 7.600640315252954e-05, "loss": 0.0112, "step": 35960 }, { "grad_norm": 0.1827317625284195, "learning_rate": 7.599227963877284e-05, "loss": 0.0142, "step": 35970 }, { "grad_norm": 0.15796251595020294, "learning_rate": 7.597815328253954e-05, "loss": 0.0151, "step": 35980 }, { "grad_norm": 0.11618997156620026, "learning_rate": 7.596402408537443e-05, "loss": 0.0122, "step": 35990 }, { "grad_norm": 0.16507765650749207, "learning_rate": 7.59498920488227e-05, "loss": 0.0156, "step": 36000 }, { "grad_norm": 0.2149137556552887, "learning_rate": 7.593575717442979e-05, "loss": 0.0123, "step": 36010 }, { "grad_norm": 0.16816909611225128, "learning_rate": 7.592161946374147e-05, "loss": 0.0143, "step": 36020 }, { "grad_norm": 0.24292226135730743, "learning_rate": 7.590747891830381e-05, "loss": 0.0131, "step": 36030 }, { "grad_norm": 0.19014444947242737, "learning_rate": 7.58933355396632e-05, "loss": 0.0146, "step": 36040 }, { "grad_norm": 0.18765144050121307, "learning_rate": 7.587918932936636e-05, "loss": 0.0132, "step": 36050 }, { "grad_norm": 0.14005272090435028, "learning_rate": 7.586504028896028e-05, "loss": 0.0126, "step": 36060 }, { "grad_norm": 0.16641965508460999, "learning_rate": 7.585088841999228e-05, "loss": 0.014, "step": 36070 }, { "grad_norm": 0.1713034212589264, "learning_rate": 7.583673372400999e-05, "loss": 0.0142, "step": 36080 }, { "grad_norm": 0.11870523542165756, "learning_rate": 7.582257620256134e-05, "loss": 0.0127, "step": 36090 }, { "grad_norm": 0.1844743937253952, "learning_rate": 7.580841585719458e-05, "loss": 0.0129, "step": 36100 }, { "grad_norm": 0.1657913476228714, "learning_rate": 7.579425268945825e-05, "loss": 0.0125, "step": 36110 }, { "grad_norm": 0.1138351634144783, "learning_rate": 7.578008670090127e-05, "loss": 0.012, "step": 36120 }, { "grad_norm": 0.24390225112438202, "learning_rate": 7.576591789307272e-05, "loss": 0.0132, "step": 36130 }, { "grad_norm": 0.11856957525014877, "learning_rate": 7.575174626752216e-05, "loss": 0.0117, "step": 36140 }, { "grad_norm": 0.16595034301280975, "learning_rate": 7.573757182579934e-05, "loss": 0.0127, "step": 36150 }, { "grad_norm": 0.18055158853530884, "learning_rate": 7.572339456945435e-05, "loss": 0.0125, "step": 36160 }, { "grad_norm": 0.23541375994682312, "learning_rate": 7.570921450003762e-05, "loss": 0.0144, "step": 36170 }, { "grad_norm": 0.14896096289157867, "learning_rate": 7.569503161909983e-05, "loss": 0.0142, "step": 36180 }, { "grad_norm": 0.19221647083759308, "learning_rate": 7.568084592819202e-05, "loss": 0.0152, "step": 36190 }, { "grad_norm": 0.24592027068138123, "learning_rate": 7.566665742886551e-05, "loss": 0.0153, "step": 36200 }, { "grad_norm": 0.14635516703128815, "learning_rate": 7.565246612267194e-05, "loss": 0.0159, "step": 36210 }, { "grad_norm": 0.18366561830043793, "learning_rate": 7.563827201116325e-05, "loss": 0.0154, "step": 36220 }, { "grad_norm": 0.2574116885662079, "learning_rate": 7.562407509589164e-05, "loss": 0.0149, "step": 36230 }, { "grad_norm": 0.17217865586280823, "learning_rate": 7.560987537840973e-05, "loss": 0.0163, "step": 36240 }, { "grad_norm": 0.21559540927410126, "learning_rate": 7.559567286027036e-05, "loss": 0.0139, "step": 36250 }, { "grad_norm": 0.33876270055770874, "learning_rate": 7.558146754302668e-05, "loss": 0.0139, "step": 36260 }, { "grad_norm": 0.1350976675748825, "learning_rate": 7.556725942823217e-05, "loss": 0.0169, "step": 36270 }, { "grad_norm": 0.1517421156167984, "learning_rate": 7.555304851744061e-05, "loss": 0.0145, "step": 36280 }, { "grad_norm": 0.14117075502872467, "learning_rate": 7.553883481220608e-05, "loss": 0.016, "step": 36290 }, { "grad_norm": 0.18104122579097748, "learning_rate": 7.552461831408298e-05, "loss": 0.0135, "step": 36300 }, { "grad_norm": 0.21878384053707123, "learning_rate": 7.551039902462599e-05, "loss": 0.0152, "step": 36310 }, { "grad_norm": 0.18610112369060516, "learning_rate": 7.549617694539011e-05, "loss": 0.0148, "step": 36320 }, { "grad_norm": 0.15018214285373688, "learning_rate": 7.548195207793066e-05, "loss": 0.0156, "step": 36330 }, { "grad_norm": 0.22224046289920807, "learning_rate": 7.546772442380323e-05, "loss": 0.0127, "step": 36340 }, { "grad_norm": 0.1613764464855194, "learning_rate": 7.545349398456376e-05, "loss": 0.015, "step": 36350 }, { "grad_norm": 0.14022289216518402, "learning_rate": 7.543926076176845e-05, "loss": 0.0135, "step": 36360 }, { "grad_norm": 0.16321390867233276, "learning_rate": 7.542502475697385e-05, "loss": 0.0158, "step": 36370 }, { "grad_norm": 0.14746184647083282, "learning_rate": 7.541078597173675e-05, "loss": 0.0141, "step": 36380 }, { "grad_norm": 0.1329106241464615, "learning_rate": 7.539654440761431e-05, "loss": 0.0164, "step": 36390 }, { "grad_norm": 0.23554542660713196, "learning_rate": 7.538230006616395e-05, "loss": 0.0166, "step": 36400 }, { "grad_norm": 0.11609901487827301, "learning_rate": 7.536805294894342e-05, "loss": 0.0141, "step": 36410 }, { "grad_norm": 0.23372383415699005, "learning_rate": 7.535380305751076e-05, "loss": 0.0154, "step": 36420 }, { "grad_norm": 0.1687721461057663, "learning_rate": 7.533955039342431e-05, "loss": 0.0131, "step": 36430 }, { "grad_norm": 0.16317442059516907, "learning_rate": 7.532529495824274e-05, "loss": 0.0154, "step": 36440 }, { "grad_norm": 0.20812539756298065, "learning_rate": 7.531103675352497e-05, "loss": 0.0122, "step": 36450 }, { "grad_norm": 0.2488567978143692, "learning_rate": 7.529677578083029e-05, "loss": 0.0156, "step": 36460 }, { "grad_norm": 0.16998103260993958, "learning_rate": 7.528251204171823e-05, "loss": 0.0136, "step": 36470 }, { "grad_norm": 0.18489788472652435, "learning_rate": 7.526824553774867e-05, "loss": 0.017, "step": 36480 }, { "grad_norm": 0.25793489813804626, "learning_rate": 7.525397627048178e-05, "loss": 0.0188, "step": 36490 }, { "grad_norm": 0.19117602705955505, "learning_rate": 7.523970424147802e-05, "loss": 0.0166, "step": 36500 }, { "grad_norm": 0.15652453899383545, "learning_rate": 7.522542945229813e-05, "loss": 0.0145, "step": 36510 }, { "grad_norm": 0.18552935123443604, "learning_rate": 7.521115190450321e-05, "loss": 0.0144, "step": 36520 }, { "grad_norm": 0.1915290653705597, "learning_rate": 7.519687159965462e-05, "loss": 0.017, "step": 36530 }, { "grad_norm": 0.1392955780029297, "learning_rate": 7.518258853931403e-05, "loss": 0.0139, "step": 36540 }, { "grad_norm": 0.18319763243198395, "learning_rate": 7.516830272504342e-05, "loss": 0.0154, "step": 36550 }, { "grad_norm": 0.20542022585868835, "learning_rate": 7.515401415840509e-05, "loss": 0.013, "step": 36560 }, { "grad_norm": 0.24672859907150269, "learning_rate": 7.513972284096155e-05, "loss": 0.0137, "step": 36570 }, { "grad_norm": 0.1768067628145218, "learning_rate": 7.512542877427574e-05, "loss": 0.0144, "step": 36580 }, { "grad_norm": 0.14460507035255432, "learning_rate": 7.51111319599108e-05, "loss": 0.0146, "step": 36590 }, { "grad_norm": 0.15522702038288116, "learning_rate": 7.509683239943024e-05, "loss": 0.0136, "step": 36600 }, { "grad_norm": 0.14394938945770264, "learning_rate": 7.50825300943978e-05, "loss": 0.0115, "step": 36610 }, { "grad_norm": 0.10829632729291916, "learning_rate": 7.50682250463776e-05, "loss": 0.0133, "step": 36620 }, { "grad_norm": 0.23046612739562988, "learning_rate": 7.5053917256934e-05, "loss": 0.0153, "step": 36630 }, { "grad_norm": 0.162908136844635, "learning_rate": 7.503960672763166e-05, "loss": 0.0159, "step": 36640 }, { "grad_norm": 0.15107452869415283, "learning_rate": 7.502529346003559e-05, "loss": 0.0131, "step": 36650 }, { "grad_norm": 0.17311111092567444, "learning_rate": 7.501097745571106e-05, "loss": 0.0144, "step": 36660 }, { "grad_norm": 0.1838325709104538, "learning_rate": 7.499665871622361e-05, "loss": 0.0127, "step": 36670 }, { "grad_norm": 0.14866897463798523, "learning_rate": 7.498233724313917e-05, "loss": 0.0124, "step": 36680 }, { "grad_norm": 0.1935822069644928, "learning_rate": 7.496801303802387e-05, "loss": 0.0122, "step": 36690 }, { "grad_norm": 0.17430052161216736, "learning_rate": 7.495368610244423e-05, "loss": 0.012, "step": 36700 }, { "grad_norm": 0.13921795785427094, "learning_rate": 7.493935643796697e-05, "loss": 0.0145, "step": 36710 }, { "grad_norm": 0.1912461370229721, "learning_rate": 7.49250240461592e-05, "loss": 0.0114, "step": 36720 }, { "grad_norm": 0.17124561965465546, "learning_rate": 7.491068892858826e-05, "loss": 0.0137, "step": 36730 }, { "grad_norm": 0.13221582770347595, "learning_rate": 7.489635108682184e-05, "loss": 0.0136, "step": 36740 }, { "grad_norm": 0.157096728682518, "learning_rate": 7.48820105224279e-05, "loss": 0.014, "step": 36750 }, { "grad_norm": 0.12358475476503372, "learning_rate": 7.48676672369747e-05, "loss": 0.0119, "step": 36760 }, { "grad_norm": 0.18836502730846405, "learning_rate": 7.485332123203079e-05, "loss": 0.0154, "step": 36770 }, { "grad_norm": 0.14108586311340332, "learning_rate": 7.483897250916504e-05, "loss": 0.0132, "step": 36780 }, { "grad_norm": 0.18786625564098358, "learning_rate": 7.48246210699466e-05, "loss": 0.013, "step": 36790 }, { "grad_norm": 0.13494019210338593, "learning_rate": 7.481026691594492e-05, "loss": 0.0121, "step": 36800 }, { "grad_norm": 0.17111259698867798, "learning_rate": 7.479591004872975e-05, "loss": 0.0171, "step": 36810 }, { "grad_norm": 0.1592697948217392, "learning_rate": 7.478155046987114e-05, "loss": 0.0121, "step": 36820 }, { "grad_norm": 0.1986382007598877, "learning_rate": 7.476718818093944e-05, "loss": 0.0128, "step": 36830 }, { "grad_norm": 0.15652500092983246, "learning_rate": 7.475282318350527e-05, "loss": 0.0139, "step": 36840 }, { "grad_norm": 0.17062921822071075, "learning_rate": 7.473845547913957e-05, "loss": 0.0115, "step": 36850 }, { "grad_norm": 0.22098705172538757, "learning_rate": 7.472408506941357e-05, "loss": 0.0137, "step": 36860 }, { "grad_norm": 0.16610488295555115, "learning_rate": 7.470971195589881e-05, "loss": 0.0134, "step": 36870 }, { "grad_norm": 0.12199606746435165, "learning_rate": 7.46953361401671e-05, "loss": 0.0119, "step": 36880 }, { "grad_norm": 0.17811377346515656, "learning_rate": 7.468095762379055e-05, "loss": 0.0127, "step": 36890 }, { "grad_norm": 0.18649369478225708, "learning_rate": 7.466657640834158e-05, "loss": 0.0135, "step": 36900 }, { "grad_norm": 0.1277543157339096, "learning_rate": 7.46521924953929e-05, "loss": 0.0129, "step": 36910 }, { "grad_norm": 0.13554081320762634, "learning_rate": 7.463780588651752e-05, "loss": 0.0131, "step": 36920 }, { "grad_norm": 0.15058016777038574, "learning_rate": 7.462341658328873e-05, "loss": 0.0142, "step": 36930 }, { "grad_norm": 0.31217363476753235, "learning_rate": 7.460902458728012e-05, "loss": 0.015, "step": 36940 }, { "grad_norm": 0.15979540348052979, "learning_rate": 7.459462990006558e-05, "loss": 0.0137, "step": 36950 }, { "grad_norm": 0.18622688949108124, "learning_rate": 7.458023252321929e-05, "loss": 0.0132, "step": 36960 }, { "grad_norm": 0.15892505645751953, "learning_rate": 7.456583245831573e-05, "loss": 0.0134, "step": 36970 }, { "grad_norm": 0.18420949578285217, "learning_rate": 7.455142970692968e-05, "loss": 0.0138, "step": 36980 }, { "grad_norm": 0.11742285639047623, "learning_rate": 7.453702427063615e-05, "loss": 0.0108, "step": 36990 }, { "grad_norm": 0.12982311844825745, "learning_rate": 7.452261615101057e-05, "loss": 0.0141, "step": 37000 }, { "grad_norm": 0.17543798685073853, "learning_rate": 7.450820534962856e-05, "loss": 0.0132, "step": 37010 }, { "grad_norm": 0.24789351224899292, "learning_rate": 7.449379186806604e-05, "loss": 0.0126, "step": 37020 }, { "grad_norm": 0.16996310651302338, "learning_rate": 7.44793757078993e-05, "loss": 0.014, "step": 37030 }, { "grad_norm": 0.14818070828914642, "learning_rate": 7.44649568707048e-05, "loss": 0.0172, "step": 37040 }, { "grad_norm": 0.1614086627960205, "learning_rate": 7.445053535805942e-05, "loss": 0.0124, "step": 37050 }, { "grad_norm": 0.16679537296295166, "learning_rate": 7.443611117154024e-05, "loss": 0.0162, "step": 37060 }, { "grad_norm": 0.16039949655532837, "learning_rate": 7.442168431272469e-05, "loss": 0.0146, "step": 37070 }, { "grad_norm": 0.1859152764081955, "learning_rate": 7.440725478319044e-05, "loss": 0.0139, "step": 37080 }, { "grad_norm": 0.19670726358890533, "learning_rate": 7.439282258451553e-05, "loss": 0.0127, "step": 37090 }, { "grad_norm": 0.15352311730384827, "learning_rate": 7.43783877182782e-05, "loss": 0.0124, "step": 37100 }, { "grad_norm": 0.14559844136238098, "learning_rate": 7.436395018605707e-05, "loss": 0.0113, "step": 37110 }, { "grad_norm": 0.1517454832792282, "learning_rate": 7.434950998943094e-05, "loss": 0.0133, "step": 37120 }, { "grad_norm": 0.24340026080608368, "learning_rate": 7.433506712997902e-05, "loss": 0.0152, "step": 37130 }, { "grad_norm": 0.1756163090467453, "learning_rate": 7.432062160928077e-05, "loss": 0.0108, "step": 37140 }, { "grad_norm": 0.16460594534873962, "learning_rate": 7.430617342891588e-05, "loss": 0.0114, "step": 37150 }, { "grad_norm": 0.2437351793050766, "learning_rate": 7.429172259046442e-05, "loss": 0.0143, "step": 37160 }, { "grad_norm": 0.13230375945568085, "learning_rate": 7.42772690955067e-05, "loss": 0.0106, "step": 37170 }, { "grad_norm": 0.15905487537384033, "learning_rate": 7.426281294562332e-05, "loss": 0.0134, "step": 37180 }, { "grad_norm": 0.12587280571460724, "learning_rate": 7.424835414239519e-05, "loss": 0.0156, "step": 37190 }, { "grad_norm": 0.20829473435878754, "learning_rate": 7.423389268740352e-05, "loss": 0.0149, "step": 37200 }, { "grad_norm": 0.15015091001987457, "learning_rate": 7.42194285822298e-05, "loss": 0.0134, "step": 37210 }, { "grad_norm": 0.1778375655412674, "learning_rate": 7.420496182845576e-05, "loss": 0.013, "step": 37220 }, { "grad_norm": 0.17034897208213806, "learning_rate": 7.419049242766348e-05, "loss": 0.0129, "step": 37230 }, { "grad_norm": 0.2493201196193695, "learning_rate": 7.417602038143533e-05, "loss": 0.0138, "step": 37240 }, { "grad_norm": 0.21361765265464783, "learning_rate": 7.416154569135393e-05, "loss": 0.0141, "step": 37250 }, { "grad_norm": 0.19033464789390564, "learning_rate": 7.414706835900222e-05, "loss": 0.0138, "step": 37260 }, { "grad_norm": 0.23194749653339386, "learning_rate": 7.413258838596345e-05, "loss": 0.014, "step": 37270 }, { "grad_norm": 0.18091347813606262, "learning_rate": 7.411810577382106e-05, "loss": 0.0143, "step": 37280 }, { "grad_norm": 0.24054886400699615, "learning_rate": 7.410362052415889e-05, "loss": 0.0168, "step": 37290 }, { "grad_norm": 0.20172740519046783, "learning_rate": 7.408913263856102e-05, "loss": 0.0168, "step": 37300 }, { "grad_norm": 0.14830879867076874, "learning_rate": 7.407464211861183e-05, "loss": 0.0145, "step": 37310 }, { "grad_norm": 0.22660435736179352, "learning_rate": 7.406014896589597e-05, "loss": 0.0138, "step": 37320 }, { "grad_norm": 0.14226415753364563, "learning_rate": 7.404565318199838e-05, "loss": 0.0139, "step": 37330 }, { "grad_norm": 0.23667998611927032, "learning_rate": 7.403115476850432e-05, "loss": 0.0119, "step": 37340 }, { "grad_norm": 0.11565715074539185, "learning_rate": 7.401665372699932e-05, "loss": 0.0139, "step": 37350 }, { "grad_norm": 0.15629343688488007, "learning_rate": 7.400215005906916e-05, "loss": 0.0128, "step": 37360 }, { "grad_norm": 0.23123912513256073, "learning_rate": 7.398764376629996e-05, "loss": 0.0137, "step": 37370 }, { "grad_norm": 0.1612190157175064, "learning_rate": 7.39731348502781e-05, "loss": 0.0128, "step": 37380 }, { "grad_norm": 0.13371001183986664, "learning_rate": 7.395862331259028e-05, "loss": 0.0131, "step": 37390 }, { "grad_norm": 0.24127526581287384, "learning_rate": 7.394410915482341e-05, "loss": 0.0128, "step": 37400 }, { "grad_norm": 0.24280937016010284, "learning_rate": 7.392959237856476e-05, "loss": 0.0134, "step": 37410 }, { "grad_norm": 0.14868901669979095, "learning_rate": 7.391507298540187e-05, "loss": 0.0142, "step": 37420 }, { "grad_norm": 0.18394802510738373, "learning_rate": 7.390055097692257e-05, "loss": 0.0118, "step": 37430 }, { "grad_norm": 0.20569244027137756, "learning_rate": 7.388602635471494e-05, "loss": 0.0123, "step": 37440 }, { "grad_norm": 0.2013821303844452, "learning_rate": 7.387149912036738e-05, "loss": 0.0155, "step": 37450 }, { "grad_norm": 0.17670662701129913, "learning_rate": 7.385696927546854e-05, "loss": 0.0134, "step": 37460 }, { "grad_norm": 0.12857604026794434, "learning_rate": 7.384243682160743e-05, "loss": 0.0122, "step": 37470 }, { "grad_norm": 0.19905157387256622, "learning_rate": 7.382790176037329e-05, "loss": 0.0124, "step": 37480 }, { "grad_norm": 0.1554027497768402, "learning_rate": 7.381336409335562e-05, "loss": 0.0125, "step": 37490 }, { "grad_norm": 0.2423485517501831, "learning_rate": 7.379882382214426e-05, "loss": 0.0136, "step": 37500 }, { "grad_norm": 0.17867644131183624, "learning_rate": 7.378428094832931e-05, "loss": 0.0167, "step": 37510 }, { "grad_norm": 0.20499496161937714, "learning_rate": 7.376973547350114e-05, "loss": 0.0152, "step": 37520 }, { "grad_norm": 0.158075749874115, "learning_rate": 7.375518739925046e-05, "loss": 0.0142, "step": 37530 }, { "grad_norm": 0.1751418262720108, "learning_rate": 7.374063672716818e-05, "loss": 0.0147, "step": 37540 }, { "grad_norm": 0.15970394015312195, "learning_rate": 7.372608345884558e-05, "loss": 0.0118, "step": 37550 }, { "grad_norm": 0.16944018006324768, "learning_rate": 7.371152759587415e-05, "loss": 0.0145, "step": 37560 }, { "grad_norm": 0.15281669795513153, "learning_rate": 7.369696913984572e-05, "loss": 0.0149, "step": 37570 }, { "grad_norm": 0.189438596367836, "learning_rate": 7.368240809235237e-05, "loss": 0.0132, "step": 37580 }, { "grad_norm": 0.12560302019119263, "learning_rate": 7.366784445498649e-05, "loss": 0.0155, "step": 37590 }, { "grad_norm": 0.17931313812732697, "learning_rate": 7.36532782293407e-05, "loss": 0.0162, "step": 37600 }, { "grad_norm": 0.1677968055009842, "learning_rate": 7.363870941700797e-05, "loss": 0.0139, "step": 37610 }, { "grad_norm": 0.16371503472328186, "learning_rate": 7.362413801958152e-05, "loss": 0.0127, "step": 37620 }, { "grad_norm": 0.14735813438892365, "learning_rate": 7.360956403865486e-05, "loss": 0.0133, "step": 37630 }, { "grad_norm": 0.17424841225147247, "learning_rate": 7.359498747582177e-05, "loss": 0.0123, "step": 37640 }, { "grad_norm": 0.1960277110338211, "learning_rate": 7.358040833267634e-05, "loss": 0.0128, "step": 37650 }, { "grad_norm": 0.1573697179555893, "learning_rate": 7.356582661081288e-05, "loss": 0.0124, "step": 37660 }, { "grad_norm": 0.20335794985294342, "learning_rate": 7.355124231182607e-05, "loss": 0.0117, "step": 37670 }, { "grad_norm": 0.2270636260509491, "learning_rate": 7.353665543731083e-05, "loss": 0.0117, "step": 37680 }, { "grad_norm": 0.1506805419921875, "learning_rate": 7.352206598886229e-05, "loss": 0.0138, "step": 37690 }, { "grad_norm": 0.18438765406608582, "learning_rate": 7.350747396807601e-05, "loss": 0.0128, "step": 37700 }, { "grad_norm": 0.10818417370319366, "learning_rate": 7.349287937654772e-05, "loss": 0.0137, "step": 37710 }, { "grad_norm": 0.1607155203819275, "learning_rate": 7.347828221587345e-05, "loss": 0.012, "step": 37720 }, { "grad_norm": 0.1997402310371399, "learning_rate": 7.346368248764955e-05, "loss": 0.0124, "step": 37730 }, { "grad_norm": 0.16839441657066345, "learning_rate": 7.34490801934726e-05, "loss": 0.016, "step": 37740 }, { "grad_norm": 0.18459630012512207, "learning_rate": 7.343447533493947e-05, "loss": 0.0131, "step": 37750 }, { "grad_norm": 0.12730494141578674, "learning_rate": 7.341986791364739e-05, "loss": 0.0123, "step": 37760 }, { "grad_norm": 0.17995616793632507, "learning_rate": 7.340525793119374e-05, "loss": 0.0127, "step": 37770 }, { "grad_norm": 0.19813182950019836, "learning_rate": 7.33906453891763e-05, "loss": 0.0126, "step": 37780 }, { "grad_norm": 0.1367397904396057, "learning_rate": 7.337603028919303e-05, "loss": 0.0122, "step": 37790 }, { "grad_norm": 0.14637240767478943, "learning_rate": 7.336141263284225e-05, "loss": 0.0128, "step": 37800 }, { "grad_norm": 0.17981308698654175, "learning_rate": 7.334679242172249e-05, "loss": 0.0133, "step": 37810 }, { "grad_norm": 0.2011341005563736, "learning_rate": 7.33321696574326e-05, "loss": 0.0121, "step": 37820 }, { "grad_norm": 0.19956903159618378, "learning_rate": 7.331754434157174e-05, "loss": 0.0147, "step": 37830 }, { "grad_norm": 0.17988507449626923, "learning_rate": 7.330291647573928e-05, "loss": 0.0129, "step": 37840 }, { "grad_norm": 0.22448670864105225, "learning_rate": 7.32882860615349e-05, "loss": 0.0139, "step": 37850 }, { "grad_norm": 0.14932551980018616, "learning_rate": 7.327365310055858e-05, "loss": 0.0124, "step": 37860 }, { "grad_norm": 0.16815683245658875, "learning_rate": 7.325901759441055e-05, "loss": 0.014, "step": 37870 }, { "grad_norm": 0.1958535760641098, "learning_rate": 7.324437954469132e-05, "loss": 0.0146, "step": 37880 }, { "grad_norm": 0.19692550599575043, "learning_rate": 7.32297389530017e-05, "loss": 0.011, "step": 37890 }, { "grad_norm": 0.13475467264652252, "learning_rate": 7.321509582094274e-05, "loss": 0.012, "step": 37900 }, { "grad_norm": 0.15852268040180206, "learning_rate": 7.320045015011582e-05, "loss": 0.0118, "step": 37910 }, { "grad_norm": 0.19638699293136597, "learning_rate": 7.318580194212253e-05, "loss": 0.0143, "step": 37920 }, { "grad_norm": 0.14968055486679077, "learning_rate": 7.317115119856483e-05, "loss": 0.013, "step": 37930 }, { "grad_norm": 0.15126557648181915, "learning_rate": 7.315649792104485e-05, "loss": 0.0118, "step": 37940 }, { "grad_norm": 0.19092170894145966, "learning_rate": 7.314184211116507e-05, "loss": 0.0128, "step": 37950 }, { "grad_norm": 0.13779544830322266, "learning_rate": 7.312718377052823e-05, "loss": 0.0126, "step": 37960 }, { "grad_norm": 0.17790164053440094, "learning_rate": 7.311252290073737e-05, "loss": 0.0132, "step": 37970 }, { "grad_norm": 0.20168687403202057, "learning_rate": 7.309785950339572e-05, "loss": 0.0136, "step": 37980 }, { "grad_norm": 0.17687873542308807, "learning_rate": 7.308319358010692e-05, "loss": 0.012, "step": 37990 }, { "grad_norm": 0.21052032709121704, "learning_rate": 7.306852513247474e-05, "loss": 0.0131, "step": 38000 }, { "grad_norm": 0.18194147944450378, "learning_rate": 7.305385416210336e-05, "loss": 0.0105, "step": 38010 }, { "grad_norm": 0.16025125980377197, "learning_rate": 7.303918067059711e-05, "loss": 0.0127, "step": 38020 }, { "grad_norm": 0.12363617867231369, "learning_rate": 7.302450465956075e-05, "loss": 0.0113, "step": 38030 }, { "grad_norm": 0.2859969735145569, "learning_rate": 7.300982613059914e-05, "loss": 0.0129, "step": 38040 }, { "grad_norm": 0.13502560555934906, "learning_rate": 7.299514508531757e-05, "loss": 0.0124, "step": 38050 }, { "grad_norm": 0.17241758108139038, "learning_rate": 7.298046152532148e-05, "loss": 0.0132, "step": 38060 }, { "grad_norm": 0.17778363823890686, "learning_rate": 7.296577545221668e-05, "loss": 0.0132, "step": 38070 }, { "grad_norm": 0.1341552883386612, "learning_rate": 7.29510868676092e-05, "loss": 0.0143, "step": 38080 }, { "grad_norm": 0.13308465480804443, "learning_rate": 7.293639577310538e-05, "loss": 0.0129, "step": 38090 }, { "grad_norm": 0.18819601833820343, "learning_rate": 7.292170217031179e-05, "loss": 0.0108, "step": 38100 }, { "grad_norm": 0.2673678398132324, "learning_rate": 7.290700606083532e-05, "loss": 0.0122, "step": 38110 }, { "grad_norm": 0.14059267938137054, "learning_rate": 7.28923074462831e-05, "loss": 0.0123, "step": 38120 }, { "grad_norm": 0.16336220502853394, "learning_rate": 7.287760632826255e-05, "loss": 0.0135, "step": 38130 }, { "grad_norm": 0.13402973115444183, "learning_rate": 7.286290270838139e-05, "loss": 0.0106, "step": 38140 }, { "grad_norm": 0.16997399926185608, "learning_rate": 7.284819658824756e-05, "loss": 0.0151, "step": 38150 }, { "grad_norm": 0.20732875168323517, "learning_rate": 7.283348796946929e-05, "loss": 0.0131, "step": 38160 }, { "grad_norm": 0.21800561249256134, "learning_rate": 7.28187768536551e-05, "loss": 0.0127, "step": 38170 }, { "grad_norm": 0.1561170369386673, "learning_rate": 7.280406324241379e-05, "loss": 0.0125, "step": 38180 }, { "grad_norm": 0.14034715294837952, "learning_rate": 7.278934713735441e-05, "loss": 0.0133, "step": 38190 }, { "grad_norm": 0.18140390515327454, "learning_rate": 7.277462854008629e-05, "loss": 0.0116, "step": 38200 }, { "grad_norm": 0.168365478515625, "learning_rate": 7.275990745221904e-05, "loss": 0.0139, "step": 38210 }, { "grad_norm": 0.14289715886116028, "learning_rate": 7.274518387536252e-05, "loss": 0.0123, "step": 38220 }, { "grad_norm": 0.1276896744966507, "learning_rate": 7.273045781112689e-05, "loss": 0.0115, "step": 38230 }, { "grad_norm": 0.1610236018896103, "learning_rate": 7.271572926112257e-05, "loss": 0.0103, "step": 38240 }, { "grad_norm": 0.11440132558345795, "learning_rate": 7.270099822696024e-05, "loss": 0.0106, "step": 38250 }, { "grad_norm": 0.187849760055542, "learning_rate": 7.26862647102509e-05, "loss": 0.0117, "step": 38260 }, { "grad_norm": 0.15036138892173767, "learning_rate": 7.267152871260573e-05, "loss": 0.012, "step": 38270 }, { "grad_norm": 0.16407598555088043, "learning_rate": 7.265679023563629e-05, "loss": 0.0117, "step": 38280 }, { "grad_norm": 0.15358126163482666, "learning_rate": 7.264204928095433e-05, "loss": 0.0113, "step": 38290 }, { "grad_norm": 0.17571990191936493, "learning_rate": 7.262730585017188e-05, "loss": 0.0139, "step": 38300 }, { "grad_norm": 0.1717609465122223, "learning_rate": 7.261255994490131e-05, "loss": 0.0127, "step": 38310 }, { "grad_norm": 0.1532701998949051, "learning_rate": 7.259781156675517e-05, "loss": 0.013, "step": 38320 }, { "grad_norm": 0.10060461610555649, "learning_rate": 7.258306071734633e-05, "loss": 0.0114, "step": 38330 }, { "grad_norm": 0.16118553280830383, "learning_rate": 7.256830739828792e-05, "loss": 0.0143, "step": 38340 }, { "grad_norm": 0.17357484996318817, "learning_rate": 7.255355161119336e-05, "loss": 0.012, "step": 38350 }, { "grad_norm": 0.12843449413776398, "learning_rate": 7.253879335767628e-05, "loss": 0.0123, "step": 38360 }, { "grad_norm": 0.20138847827911377, "learning_rate": 7.252403263935067e-05, "loss": 0.0129, "step": 38370 }, { "grad_norm": 0.29879871010780334, "learning_rate": 7.25092694578307e-05, "loss": 0.0123, "step": 38380 }, { "grad_norm": 0.5854761600494385, "learning_rate": 7.249450381473087e-05, "loss": 0.0179, "step": 38390 }, { "grad_norm": 0.2746860980987549, "learning_rate": 7.247973571166593e-05, "loss": 0.0191, "step": 38400 }, { "grad_norm": 0.18001487851142883, "learning_rate": 7.246496515025089e-05, "loss": 0.0188, "step": 38410 }, { "grad_norm": 0.19778862595558167, "learning_rate": 7.245019213210101e-05, "loss": 0.0169, "step": 38420 }, { "grad_norm": 0.2196112424135208, "learning_rate": 7.24354166588319e-05, "loss": 0.0177, "step": 38430 }, { "grad_norm": 0.17187979817390442, "learning_rate": 7.242063873205935e-05, "loss": 0.0132, "step": 38440 }, { "grad_norm": 0.19568456709384918, "learning_rate": 7.240585835339946e-05, "loss": 0.0171, "step": 38450 }, { "grad_norm": 0.5530155897140503, "learning_rate": 7.239107552446857e-05, "loss": 0.0174, "step": 38460 }, { "grad_norm": 0.19481322169303894, "learning_rate": 7.237629024688334e-05, "loss": 0.0173, "step": 38470 }, { "grad_norm": 0.24114976823329926, "learning_rate": 7.236150252226064e-05, "loss": 0.0167, "step": 38480 }, { "grad_norm": 0.20416225492954254, "learning_rate": 7.234671235221765e-05, "loss": 0.0154, "step": 38490 }, { "grad_norm": 0.12729708850383759, "learning_rate": 7.233191973837179e-05, "loss": 0.0136, "step": 38500 }, { "grad_norm": 0.20742815732955933, "learning_rate": 7.231712468234078e-05, "loss": 0.0139, "step": 38510 }, { "grad_norm": 0.13464808464050293, "learning_rate": 7.230232718574254e-05, "loss": 0.0128, "step": 38520 }, { "grad_norm": 0.1561369150876999, "learning_rate": 7.228752725019535e-05, "loss": 0.0141, "step": 38530 }, { "grad_norm": 0.15545469522476196, "learning_rate": 7.227272487731769e-05, "loss": 0.0133, "step": 38540 }, { "grad_norm": 0.14350710809230804, "learning_rate": 7.225792006872831e-05, "loss": 0.0128, "step": 38550 }, { "grad_norm": 0.2617299556732178, "learning_rate": 7.224311282604628e-05, "loss": 0.014, "step": 38560 }, { "grad_norm": 0.14354346692562103, "learning_rate": 7.222830315089085e-05, "loss": 0.0124, "step": 38570 }, { "grad_norm": 0.14198239147663116, "learning_rate": 7.22134910448816e-05, "loss": 0.0118, "step": 38580 }, { "grad_norm": 0.1702164262533188, "learning_rate": 7.219867650963839e-05, "loss": 0.0141, "step": 38590 }, { "grad_norm": 0.21019332110881805, "learning_rate": 7.21838595467813e-05, "loss": 0.0143, "step": 38600 }, { "grad_norm": 0.15393561124801636, "learning_rate": 7.216904015793067e-05, "loss": 0.015, "step": 38610 }, { "grad_norm": 0.22136428952217102, "learning_rate": 7.215421834470713e-05, "loss": 0.0104, "step": 38620 }, { "grad_norm": 0.17049585282802582, "learning_rate": 7.213939410873158e-05, "loss": 0.0113, "step": 38630 }, { "grad_norm": 0.24327844381332397, "learning_rate": 7.212456745162519e-05, "loss": 0.0126, "step": 38640 }, { "grad_norm": 0.13274461030960083, "learning_rate": 7.210973837500937e-05, "loss": 0.0142, "step": 38650 }, { "grad_norm": 0.15000048279762268, "learning_rate": 7.209490688050578e-05, "loss": 0.0126, "step": 38660 }, { "grad_norm": 0.15008848905563354, "learning_rate": 7.208007296973641e-05, "loss": 0.0138, "step": 38670 }, { "grad_norm": 0.19486097991466522, "learning_rate": 7.206523664432345e-05, "loss": 0.0153, "step": 38680 }, { "grad_norm": 0.14428706467151642, "learning_rate": 7.205039790588939e-05, "loss": 0.0127, "step": 38690 }, { "grad_norm": 0.1662835329771042, "learning_rate": 7.203555675605697e-05, "loss": 0.0136, "step": 38700 }, { "grad_norm": 0.19785715639591217, "learning_rate": 7.202071319644917e-05, "loss": 0.0116, "step": 38710 }, { "grad_norm": 0.14088642597198486, "learning_rate": 7.200586722868932e-05, "loss": 0.0108, "step": 38720 }, { "grad_norm": 0.17044289410114288, "learning_rate": 7.199101885440088e-05, "loss": 0.0147, "step": 38730 }, { "grad_norm": 0.20282958447933197, "learning_rate": 7.197616807520771e-05, "loss": 0.0137, "step": 38740 }, { "grad_norm": 0.17577502131462097, "learning_rate": 7.196131489273381e-05, "loss": 0.0125, "step": 38750 }, { "grad_norm": 0.1937417834997177, "learning_rate": 7.194645930860355e-05, "loss": 0.0142, "step": 38760 }, { "grad_norm": 0.14843717217445374, "learning_rate": 7.19316013244415e-05, "loss": 0.0126, "step": 38770 }, { "grad_norm": 0.13411293923854828, "learning_rate": 7.191674094187248e-05, "loss": 0.0122, "step": 38780 }, { "grad_norm": 0.1685926914215088, "learning_rate": 7.190187816252165e-05, "loss": 0.0129, "step": 38790 }, { "grad_norm": 0.1708688884973526, "learning_rate": 7.188701298801435e-05, "loss": 0.0127, "step": 38800 }, { "grad_norm": 0.25439542531967163, "learning_rate": 7.18721454199762e-05, "loss": 0.0123, "step": 38810 }, { "grad_norm": 0.1719786524772644, "learning_rate": 7.185727546003312e-05, "loss": 0.014, "step": 38820 }, { "grad_norm": 0.11096597462892532, "learning_rate": 7.184240310981126e-05, "loss": 0.0155, "step": 38830 }, { "grad_norm": 0.19907675683498383, "learning_rate": 7.182752837093704e-05, "loss": 0.0133, "step": 38840 }, { "grad_norm": 0.11420128494501114, "learning_rate": 7.181265124503711e-05, "loss": 0.0145, "step": 38850 }, { "grad_norm": 0.1236891821026802, "learning_rate": 7.179777173373847e-05, "loss": 0.0111, "step": 38860 }, { "grad_norm": 0.16370823979377747, "learning_rate": 7.178288983866826e-05, "loss": 0.0154, "step": 38870 }, { "grad_norm": 0.19924458861351013, "learning_rate": 7.176800556145397e-05, "loss": 0.0139, "step": 38880 }, { "grad_norm": 0.212299644947052, "learning_rate": 7.175311890372334e-05, "loss": 0.0152, "step": 38890 }, { "grad_norm": 0.1856020838022232, "learning_rate": 7.17382298671043e-05, "loss": 0.0155, "step": 38900 }, { "grad_norm": 0.2015421986579895, "learning_rate": 7.172333845322515e-05, "loss": 0.0159, "step": 38910 }, { "grad_norm": 0.16424788534641266, "learning_rate": 7.170844466371436e-05, "loss": 0.0142, "step": 38920 }, { "grad_norm": 0.242717444896698, "learning_rate": 7.16935485002007e-05, "loss": 0.0156, "step": 38930 }, { "grad_norm": 0.14786599576473236, "learning_rate": 7.167864996431319e-05, "loss": 0.0137, "step": 38940 }, { "grad_norm": 0.16434744000434875, "learning_rate": 7.166374905768111e-05, "loss": 0.0141, "step": 38950 }, { "grad_norm": 0.16639943420886993, "learning_rate": 7.164884578193401e-05, "loss": 0.0128, "step": 38960 }, { "grad_norm": 0.22223584353923798, "learning_rate": 7.163394013870168e-05, "loss": 0.0112, "step": 38970 }, { "grad_norm": 0.17233040928840637, "learning_rate": 7.161903212961419e-05, "loss": 0.0122, "step": 38980 }, { "grad_norm": 0.1486247032880783, "learning_rate": 7.160412175630183e-05, "loss": 0.0153, "step": 38990 }, { "grad_norm": 0.15860360860824585, "learning_rate": 7.158920902039521e-05, "loss": 0.0131, "step": 39000 }, { "grad_norm": 0.17686346173286438, "learning_rate": 7.157429392352514e-05, "loss": 0.0129, "step": 39010 }, { "grad_norm": 0.16761066019535065, "learning_rate": 7.155937646732275e-05, "loss": 0.0131, "step": 39020 }, { "grad_norm": 0.12566299736499786, "learning_rate": 7.154445665341933e-05, "loss": 0.0135, "step": 39030 }, { "grad_norm": 0.14442552626132965, "learning_rate": 7.152953448344654e-05, "loss": 0.0125, "step": 39040 }, { "grad_norm": 0.16268479824066162, "learning_rate": 7.151460995903624e-05, "loss": 0.0112, "step": 39050 }, { "grad_norm": 0.18115727603435516, "learning_rate": 7.149968308182052e-05, "loss": 0.0108, "step": 39060 }, { "grad_norm": 0.12363219261169434, "learning_rate": 7.14847538534318e-05, "loss": 0.0126, "step": 39070 }, { "grad_norm": 0.21086262166500092, "learning_rate": 7.14698222755027e-05, "loss": 0.0114, "step": 39080 }, { "grad_norm": 0.16744950413703918, "learning_rate": 7.14548883496661e-05, "loss": 0.0125, "step": 39090 }, { "grad_norm": 0.14527428150177002, "learning_rate": 7.143995207755517e-05, "loss": 0.012, "step": 39100 }, { "grad_norm": 0.15613451600074768, "learning_rate": 7.142501346080333e-05, "loss": 0.0115, "step": 39110 }, { "grad_norm": 0.16265900433063507, "learning_rate": 7.141007250104421e-05, "loss": 0.0116, "step": 39120 }, { "grad_norm": 0.14179030060768127, "learning_rate": 7.139512919991176e-05, "loss": 0.0121, "step": 39130 }, { "grad_norm": 0.13128241896629333, "learning_rate": 7.138018355904014e-05, "loss": 0.0129, "step": 39140 }, { "grad_norm": 0.16869071125984192, "learning_rate": 7.13652355800638e-05, "loss": 0.0132, "step": 39150 }, { "grad_norm": 0.13217566907405853, "learning_rate": 7.135028526461743e-05, "loss": 0.0113, "step": 39160 }, { "grad_norm": 0.15560849010944366, "learning_rate": 7.133533261433594e-05, "loss": 0.0118, "step": 39170 }, { "grad_norm": 0.18406037986278534, "learning_rate": 7.132037763085457e-05, "loss": 0.013, "step": 39180 }, { "grad_norm": 0.1409391611814499, "learning_rate": 7.130542031580875e-05, "loss": 0.0114, "step": 39190 }, { "grad_norm": 0.19897246360778809, "learning_rate": 7.12904606708342e-05, "loss": 0.0127, "step": 39200 }, { "grad_norm": 0.17169027030467987, "learning_rate": 7.127549869756687e-05, "loss": 0.0135, "step": 39210 }, { "grad_norm": 0.13699916005134583, "learning_rate": 7.126053439764299e-05, "loss": 0.0132, "step": 39220 }, { "grad_norm": 0.1436539590358734, "learning_rate": 7.124556777269904e-05, "loss": 0.0124, "step": 39230 }, { "grad_norm": 0.136586993932724, "learning_rate": 7.123059882437174e-05, "loss": 0.0112, "step": 39240 }, { "grad_norm": 0.12982960045337677, "learning_rate": 7.121562755429807e-05, "loss": 0.0113, "step": 39250 }, { "grad_norm": 0.16657669842243195, "learning_rate": 7.120065396411528e-05, "loss": 0.0138, "step": 39260 }, { "grad_norm": 0.1968129277229309, "learning_rate": 7.118567805546084e-05, "loss": 0.0119, "step": 39270 }, { "grad_norm": 0.15915830433368683, "learning_rate": 7.117069982997248e-05, "loss": 0.0108, "step": 39280 }, { "grad_norm": 0.15561048686504364, "learning_rate": 7.115571928928825e-05, "loss": 0.01, "step": 39290 }, { "grad_norm": 0.14699098467826843, "learning_rate": 7.114073643504635e-05, "loss": 0.0102, "step": 39300 }, { "grad_norm": 0.1541278064250946, "learning_rate": 7.11257512688853e-05, "loss": 0.0168, "step": 39310 }, { "grad_norm": 0.12313591688871384, "learning_rate": 7.111076379244384e-05, "loss": 0.0123, "step": 39320 }, { "grad_norm": 0.16079770028591156, "learning_rate": 7.109577400736101e-05, "loss": 0.0119, "step": 39330 }, { "grad_norm": 0.18892663717269897, "learning_rate": 7.108078191527605e-05, "loss": 0.013, "step": 39340 }, { "grad_norm": 0.14039872586727142, "learning_rate": 7.106578751782847e-05, "loss": 0.0136, "step": 39350 }, { "grad_norm": 0.13603518903255463, "learning_rate": 7.105079081665803e-05, "loss": 0.014, "step": 39360 }, { "grad_norm": 0.16831062734127045, "learning_rate": 7.103579181340476e-05, "loss": 0.0154, "step": 39370 }, { "grad_norm": 0.19339227676391602, "learning_rate": 7.102079050970893e-05, "loss": 0.0131, "step": 39380 }, { "grad_norm": 0.19787682592868805, "learning_rate": 7.100578690721105e-05, "loss": 0.0127, "step": 39390 }, { "grad_norm": 0.133689746260643, "learning_rate": 7.09907810075519e-05, "loss": 0.0112, "step": 39400 }, { "grad_norm": 0.1761358082294464, "learning_rate": 7.097577281237249e-05, "loss": 0.0138, "step": 39410 }, { "grad_norm": 0.2174062579870224, "learning_rate": 7.09607623233141e-05, "loss": 0.0127, "step": 39420 }, { "grad_norm": 0.1316869705915451, "learning_rate": 7.094574954201827e-05, "loss": 0.0129, "step": 39430 }, { "grad_norm": 0.17686079442501068, "learning_rate": 7.093073447012675e-05, "loss": 0.0121, "step": 39440 }, { "grad_norm": 0.17194317281246185, "learning_rate": 7.09157171092816e-05, "loss": 0.014, "step": 39450 }, { "grad_norm": 0.13741502165794373, "learning_rate": 7.090069746112504e-05, "loss": 0.0127, "step": 39460 }, { "grad_norm": 0.13488167524337769, "learning_rate": 7.088567552729965e-05, "loss": 0.0107, "step": 39470 }, { "grad_norm": 0.14179393649101257, "learning_rate": 7.087065130944818e-05, "loss": 0.0114, "step": 39480 }, { "grad_norm": 0.23669622838497162, "learning_rate": 7.085562480921366e-05, "loss": 0.0116, "step": 39490 }, { "grad_norm": 0.19456267356872559, "learning_rate": 7.084059602823937e-05, "loss": 0.009, "step": 39500 }, { "grad_norm": 0.13965867459774017, "learning_rate": 7.082556496816882e-05, "loss": 0.0118, "step": 39510 }, { "grad_norm": 0.14351066946983337, "learning_rate": 7.081053163064582e-05, "loss": 0.0094, "step": 39520 }, { "grad_norm": 0.11976295709609985, "learning_rate": 7.079549601731434e-05, "loss": 0.0112, "step": 39530 }, { "grad_norm": 0.19469359517097473, "learning_rate": 7.07804581298187e-05, "loss": 0.0135, "step": 39540 }, { "grad_norm": 0.15128467977046967, "learning_rate": 7.07654179698034e-05, "loss": 0.0106, "step": 39550 }, { "grad_norm": 0.16452662646770477, "learning_rate": 7.075037553891321e-05, "loss": 0.0121, "step": 39560 }, { "grad_norm": 0.1444176435470581, "learning_rate": 7.073533083879315e-05, "loss": 0.0149, "step": 39570 }, { "grad_norm": 0.1667730212211609, "learning_rate": 7.072028387108849e-05, "loss": 0.0117, "step": 39580 }, { "grad_norm": 0.15500527620315552, "learning_rate": 7.070523463744472e-05, "loss": 0.0115, "step": 39590 }, { "grad_norm": 0.15239904820919037, "learning_rate": 7.069018313950763e-05, "loss": 0.0134, "step": 39600 }, { "grad_norm": 0.12955304980278015, "learning_rate": 7.067512937892322e-05, "loss": 0.0139, "step": 39610 }, { "grad_norm": 0.15341989696025848, "learning_rate": 7.066007335733775e-05, "loss": 0.0177, "step": 39620 }, { "grad_norm": 0.25467023253440857, "learning_rate": 7.064501507639772e-05, "loss": 0.013, "step": 39630 }, { "grad_norm": 0.14085721969604492, "learning_rate": 7.062995453774987e-05, "loss": 0.0148, "step": 39640 }, { "grad_norm": 0.17961834371089935, "learning_rate": 7.061489174304121e-05, "loss": 0.014, "step": 39650 }, { "grad_norm": 0.23642992973327637, "learning_rate": 7.0599826693919e-05, "loss": 0.0136, "step": 39660 }, { "grad_norm": 0.15302953124046326, "learning_rate": 7.05847593920307e-05, "loss": 0.0131, "step": 39670 }, { "grad_norm": 0.16334033012390137, "learning_rate": 7.056968983902406e-05, "loss": 0.0146, "step": 39680 }, { "grad_norm": 0.1876717507839203, "learning_rate": 7.055461803654706e-05, "loss": 0.0168, "step": 39690 }, { "grad_norm": 0.1883692890405655, "learning_rate": 7.053954398624794e-05, "loss": 0.0143, "step": 39700 }, { "grad_norm": 0.16716636717319489, "learning_rate": 7.052446768977518e-05, "loss": 0.0139, "step": 39710 }, { "grad_norm": 0.17294631898403168, "learning_rate": 7.050938914877748e-05, "loss": 0.013, "step": 39720 }, { "grad_norm": 0.16947326064109802, "learning_rate": 7.049430836490382e-05, "loss": 0.0129, "step": 39730 }, { "grad_norm": 0.21162691712379456, "learning_rate": 7.047922533980342e-05, "loss": 0.0134, "step": 39740 }, { "grad_norm": 0.13447973132133484, "learning_rate": 7.046414007512571e-05, "loss": 0.0104, "step": 39750 }, { "grad_norm": 0.19042612612247467, "learning_rate": 7.044905257252042e-05, "loss": 0.0122, "step": 39760 }, { "grad_norm": 0.13002227246761322, "learning_rate": 7.04339628336375e-05, "loss": 0.0123, "step": 39770 }, { "grad_norm": 0.16663457453250885, "learning_rate": 7.041887086012711e-05, "loss": 0.0132, "step": 39780 }, { "grad_norm": 0.1706714928150177, "learning_rate": 7.040377665363969e-05, "loss": 0.0128, "step": 39790 }, { "grad_norm": 0.22005721926689148, "learning_rate": 7.038868021582594e-05, "loss": 0.0139, "step": 39800 }, { "grad_norm": 0.1377142369747162, "learning_rate": 7.037358154833679e-05, "loss": 0.0122, "step": 39810 }, { "grad_norm": 0.1240050420165062, "learning_rate": 7.035848065282339e-05, "loss": 0.015, "step": 39820 }, { "grad_norm": 0.18358765542507172, "learning_rate": 7.034337753093714e-05, "loss": 0.0119, "step": 39830 }, { "grad_norm": 0.15146715939044952, "learning_rate": 7.032827218432972e-05, "loss": 0.0122, "step": 39840 }, { "grad_norm": 0.15270492434501648, "learning_rate": 7.031316461465302e-05, "loss": 0.0118, "step": 39850 }, { "grad_norm": 0.11666551977396011, "learning_rate": 7.029805482355915e-05, "loss": 0.0098, "step": 39860 }, { "grad_norm": 0.11724576354026794, "learning_rate": 7.028294281270055e-05, "loss": 0.0104, "step": 39870 }, { "grad_norm": 0.18317493796348572, "learning_rate": 7.026782858372982e-05, "loss": 0.0131, "step": 39880 }, { "grad_norm": 0.16186024248600006, "learning_rate": 7.025271213829982e-05, "loss": 0.0118, "step": 39890 }, { "grad_norm": 0.14768345654010773, "learning_rate": 7.023759347806366e-05, "loss": 0.0131, "step": 39900 }, { "grad_norm": 0.15911924839019775, "learning_rate": 7.022247260467472e-05, "loss": 0.0129, "step": 39910 }, { "grad_norm": 0.16965830326080322, "learning_rate": 7.020734951978658e-05, "loss": 0.0111, "step": 39920 }, { "grad_norm": 0.16270877420902252, "learning_rate": 7.019222422505307e-05, "loss": 0.0112, "step": 39930 }, { "grad_norm": 0.12334450334310532, "learning_rate": 7.017709672212827e-05, "loss": 0.0102, "step": 39940 }, { "grad_norm": 0.1777273565530777, "learning_rate": 7.016196701266652e-05, "loss": 0.0111, "step": 39950 }, { "grad_norm": 0.2155030220746994, "learning_rate": 7.014683509832238e-05, "loss": 0.0113, "step": 39960 }, { "grad_norm": 0.17360226809978485, "learning_rate": 7.013170098075063e-05, "loss": 0.0123, "step": 39970 }, { "grad_norm": 0.18351508677005768, "learning_rate": 7.011656466160632e-05, "loss": 0.0139, "step": 39980 }, { "grad_norm": 0.15131127834320068, "learning_rate": 7.010142614254475e-05, "loss": 0.012, "step": 39990 }, { "grad_norm": 0.23295429348945618, "learning_rate": 7.008628542522147e-05, "loss": 0.0142, "step": 40000 }, { "grad_norm": 0.15720653533935547, "learning_rate": 7.007114251129218e-05, "loss": 0.0127, "step": 40010 }, { "grad_norm": 0.16591310501098633, "learning_rate": 7.005599740241293e-05, "loss": 0.012, "step": 40020 }, { "grad_norm": 0.1631501466035843, "learning_rate": 7.004085010023996e-05, "loss": 0.0134, "step": 40030 }, { "grad_norm": 0.31345048546791077, "learning_rate": 7.002570060642976e-05, "loss": 0.0116, "step": 40040 }, { "grad_norm": 0.2787454426288605, "learning_rate": 7.001054892263903e-05, "loss": 0.0125, "step": 40050 }, { "grad_norm": 0.21642276644706726, "learning_rate": 6.999539505052477e-05, "loss": 0.0122, "step": 40060 }, { "grad_norm": 0.14180131256580353, "learning_rate": 6.998023899174418e-05, "loss": 0.013, "step": 40070 }, { "grad_norm": 0.143415629863739, "learning_rate": 6.996508074795467e-05, "loss": 0.0159, "step": 40080 }, { "grad_norm": 0.14701324701309204, "learning_rate": 6.994992032081396e-05, "loss": 0.0125, "step": 40090 }, { "grad_norm": 0.23092038929462433, "learning_rate": 6.993475771197995e-05, "loss": 0.0144, "step": 40100 }, { "grad_norm": 0.22499783337116241, "learning_rate": 6.991959292311082e-05, "loss": 0.0124, "step": 40110 }, { "grad_norm": 0.12217184156179428, "learning_rate": 6.990442595586495e-05, "loss": 0.0119, "step": 40120 }, { "grad_norm": 0.11655492335557938, "learning_rate": 6.988925681190098e-05, "loss": 0.0121, "step": 40130 }, { "grad_norm": 0.17043517529964447, "learning_rate": 6.987408549287778e-05, "loss": 0.012, "step": 40140 }, { "grad_norm": 0.153675839304924, "learning_rate": 6.985891200045449e-05, "loss": 0.0127, "step": 40150 }, { "grad_norm": 0.15722490847110748, "learning_rate": 6.984373633629045e-05, "loss": 0.0145, "step": 40160 }, { "grad_norm": 0.13150285184383392, "learning_rate": 6.98285585020452e-05, "loss": 0.0143, "step": 40170 }, { "grad_norm": 0.16425083577632904, "learning_rate": 6.981337849937864e-05, "loss": 0.0094, "step": 40180 }, { "grad_norm": 0.15775750577449799, "learning_rate": 6.979819632995078e-05, "loss": 0.0126, "step": 40190 }, { "grad_norm": 0.1311500370502472, "learning_rate": 6.978301199542193e-05, "loss": 0.0107, "step": 40200 }, { "grad_norm": 0.16797791421413422, "learning_rate": 6.976782549745263e-05, "loss": 0.0113, "step": 40210 }, { "grad_norm": 0.14684849977493286, "learning_rate": 6.975263683770367e-05, "loss": 0.0116, "step": 40220 }, { "grad_norm": 0.16367007791996002, "learning_rate": 6.973744601783604e-05, "loss": 0.011, "step": 40230 }, { "grad_norm": 0.15555058419704437, "learning_rate": 6.972225303951097e-05, "loss": 0.0139, "step": 40240 }, { "grad_norm": 0.1460108608007431, "learning_rate": 6.970705790438998e-05, "loss": 0.0159, "step": 40250 }, { "grad_norm": 0.1825232207775116, "learning_rate": 6.969186061413477e-05, "loss": 0.0119, "step": 40260 }, { "grad_norm": 0.17891626060009003, "learning_rate": 6.967666117040727e-05, "loss": 0.0119, "step": 40270 }, { "grad_norm": 0.17252816259860992, "learning_rate": 6.966145957486972e-05, "loss": 0.0114, "step": 40280 }, { "grad_norm": 0.1966264247894287, "learning_rate": 6.964625582918449e-05, "loss": 0.0135, "step": 40290 }, { "grad_norm": 0.20580759644508362, "learning_rate": 6.963104993501425e-05, "loss": 0.0145, "step": 40300 }, { "grad_norm": 0.1805620640516281, "learning_rate": 6.961584189402192e-05, "loss": 0.0134, "step": 40310 }, { "grad_norm": 0.16014711558818817, "learning_rate": 6.96006317078706e-05, "loss": 0.0104, "step": 40320 }, { "grad_norm": 0.14926941692829132, "learning_rate": 6.958541937822367e-05, "loss": 0.0117, "step": 40330 }, { "grad_norm": 0.1689773052930832, "learning_rate": 6.957020490674473e-05, "loss": 0.0122, "step": 40340 }, { "grad_norm": 0.16470399498939514, "learning_rate": 6.95549882950976e-05, "loss": 0.0127, "step": 40350 }, { "grad_norm": 0.15852411091327667, "learning_rate": 6.953976954494635e-05, "loss": 0.0124, "step": 40360 }, { "grad_norm": 0.11752522736787796, "learning_rate": 6.952454865795528e-05, "loss": 0.0138, "step": 40370 }, { "grad_norm": 0.2128528505563736, "learning_rate": 6.950932563578892e-05, "loss": 0.011, "step": 40380 }, { "grad_norm": 0.1369456797838211, "learning_rate": 6.949410048011206e-05, "loss": 0.0122, "step": 40390 }, { "grad_norm": 0.20708434283733368, "learning_rate": 6.947887319258966e-05, "loss": 0.0123, "step": 40400 }, { "grad_norm": 0.17804765701293945, "learning_rate": 6.946364377488696e-05, "loss": 0.0135, "step": 40410 }, { "grad_norm": 0.19489149749279022, "learning_rate": 6.944841222866947e-05, "loss": 0.012, "step": 40420 }, { "grad_norm": 0.13446448743343353, "learning_rate": 6.943317855560284e-05, "loss": 0.0133, "step": 40430 }, { "grad_norm": 0.16137222945690155, "learning_rate": 6.941794275735302e-05, "loss": 0.0123, "step": 40440 }, { "grad_norm": 0.2269235998392105, "learning_rate": 6.94027048355862e-05, "loss": 0.0095, "step": 40450 }, { "grad_norm": 0.13452842831611633, "learning_rate": 6.938746479196871e-05, "loss": 0.0102, "step": 40460 }, { "grad_norm": 0.16268561780452728, "learning_rate": 6.937222262816724e-05, "loss": 0.0116, "step": 40470 }, { "grad_norm": 0.11736813932657242, "learning_rate": 6.935697834584865e-05, "loss": 0.0104, "step": 40480 }, { "grad_norm": 0.1500602811574936, "learning_rate": 6.934173194667998e-05, "loss": 0.0116, "step": 40490 }, { "grad_norm": 0.1193365827202797, "learning_rate": 6.93264834323286e-05, "loss": 0.0106, "step": 40500 }, { "grad_norm": 0.16132338345050812, "learning_rate": 6.931123280446206e-05, "loss": 0.0128, "step": 40510 }, { "grad_norm": 0.17726007103919983, "learning_rate": 6.929598006474811e-05, "loss": 0.0113, "step": 40520 }, { "grad_norm": 0.17061790823936462, "learning_rate": 6.92807252148548e-05, "loss": 0.0099, "step": 40530 }, { "grad_norm": 0.1478458195924759, "learning_rate": 6.926546825645037e-05, "loss": 0.0124, "step": 40540 }, { "grad_norm": 0.17997997999191284, "learning_rate": 6.92502091912033e-05, "loss": 0.0111, "step": 40550 }, { "grad_norm": 0.14448721706867218, "learning_rate": 6.923494802078229e-05, "loss": 0.0118, "step": 40560 }, { "grad_norm": 0.14445661008358002, "learning_rate": 6.921968474685628e-05, "loss": 0.0118, "step": 40570 }, { "grad_norm": 0.11578457802534103, "learning_rate": 6.920441937109445e-05, "loss": 0.0109, "step": 40580 }, { "grad_norm": 0.11177108436822891, "learning_rate": 6.91891518951662e-05, "loss": 0.0103, "step": 40590 }, { "grad_norm": 0.10848554223775864, "learning_rate": 6.917388232074114e-05, "loss": 0.0123, "step": 40600 }, { "grad_norm": 0.16733640432357788, "learning_rate": 6.915861064948914e-05, "loss": 0.0132, "step": 40610 }, { "grad_norm": 0.16919976472854614, "learning_rate": 6.914333688308025e-05, "loss": 0.0137, "step": 40620 }, { "grad_norm": 0.1390647292137146, "learning_rate": 6.912806102318487e-05, "loss": 0.011, "step": 40630 }, { "grad_norm": 0.157529816031456, "learning_rate": 6.911278307147347e-05, "loss": 0.0125, "step": 40640 }, { "grad_norm": 0.11969296634197235, "learning_rate": 6.909750302961684e-05, "loss": 0.0097, "step": 40650 }, { "grad_norm": 0.15103785693645477, "learning_rate": 6.9082220899286e-05, "loss": 0.013, "step": 40660 }, { "grad_norm": 0.21722500026226044, "learning_rate": 6.906693668215219e-05, "loss": 0.0099, "step": 40670 }, { "grad_norm": 0.17548559606075287, "learning_rate": 6.905165037988683e-05, "loss": 0.0129, "step": 40680 }, { "grad_norm": 0.15046513080596924, "learning_rate": 6.903636199416164e-05, "loss": 0.0112, "step": 40690 }, { "grad_norm": 0.1444002389907837, "learning_rate": 6.902107152664851e-05, "loss": 0.0132, "step": 40700 }, { "grad_norm": 0.15013934671878815, "learning_rate": 6.90057789790196e-05, "loss": 0.0123, "step": 40710 }, { "grad_norm": 0.14674703776836395, "learning_rate": 6.899048435294728e-05, "loss": 0.0127, "step": 40720 }, { "grad_norm": 0.13543477654457092, "learning_rate": 6.897518765010415e-05, "loss": 0.0137, "step": 40730 }, { "grad_norm": 0.13473190367221832, "learning_rate": 6.895988887216302e-05, "loss": 0.0119, "step": 40740 }, { "grad_norm": 0.13548393547534943, "learning_rate": 6.894458802079694e-05, "loss": 0.0106, "step": 40750 }, { "grad_norm": 0.15313778817653656, "learning_rate": 6.892928509767922e-05, "loss": 0.0139, "step": 40760 }, { "grad_norm": 0.16264602541923523, "learning_rate": 6.891398010448333e-05, "loss": 0.0125, "step": 40770 }, { "grad_norm": 0.16978874802589417, "learning_rate": 6.8898673042883e-05, "loss": 0.0117, "step": 40780 }, { "grad_norm": 0.15236704051494598, "learning_rate": 6.888336391455222e-05, "loss": 0.0111, "step": 40790 }, { "grad_norm": 0.1467570662498474, "learning_rate": 6.886805272116513e-05, "loss": 0.0115, "step": 40800 }, { "grad_norm": 0.17160587012767792, "learning_rate": 6.885273946439617e-05, "loss": 0.0105, "step": 40810 }, { "grad_norm": 0.1508513242006302, "learning_rate": 6.883742414591998e-05, "loss": 0.0121, "step": 40820 }, { "grad_norm": 0.15092594921588898, "learning_rate": 6.88221067674114e-05, "loss": 0.0123, "step": 40830 }, { "grad_norm": 0.15564388036727905, "learning_rate": 6.88067873305455e-05, "loss": 0.0112, "step": 40840 }, { "grad_norm": 0.24234889447689056, "learning_rate": 6.879146583699765e-05, "loss": 0.0108, "step": 40850 }, { "grad_norm": 0.14306001365184784, "learning_rate": 6.877614228844334e-05, "loss": 0.0118, "step": 40860 }, { "grad_norm": 0.14760206639766693, "learning_rate": 6.876081668655832e-05, "loss": 0.0135, "step": 40870 }, { "grad_norm": 0.13872556388378143, "learning_rate": 6.874548903301861e-05, "loss": 0.0137, "step": 40880 }, { "grad_norm": 0.15136747062206268, "learning_rate": 6.873015932950039e-05, "loss": 0.0134, "step": 40890 }, { "grad_norm": 0.1830708533525467, "learning_rate": 6.871482757768012e-05, "loss": 0.0124, "step": 40900 }, { "grad_norm": 0.1985873281955719, "learning_rate": 6.869949377923442e-05, "loss": 0.0131, "step": 40910 }, { "grad_norm": 0.16476580500602722, "learning_rate": 6.868415793584022e-05, "loss": 0.0109, "step": 40920 }, { "grad_norm": 0.19163016974925995, "learning_rate": 6.86688200491746e-05, "loss": 0.0128, "step": 40930 }, { "grad_norm": 0.1695328950881958, "learning_rate": 6.865348012091488e-05, "loss": 0.0118, "step": 40940 }, { "grad_norm": 0.14336182177066803, "learning_rate": 6.86381381527386e-05, "loss": 0.0137, "step": 40950 }, { "grad_norm": 0.12239952385425568, "learning_rate": 6.862279414632359e-05, "loss": 0.0102, "step": 40960 }, { "grad_norm": 0.11618522554636002, "learning_rate": 6.860744810334778e-05, "loss": 0.0126, "step": 40970 }, { "grad_norm": 0.14446184039115906, "learning_rate": 6.859210002548943e-05, "loss": 0.0123, "step": 40980 }, { "grad_norm": 0.21650095283985138, "learning_rate": 6.8576749914427e-05, "loss": 0.0117, "step": 40990 }, { "grad_norm": 0.1975022405385971, "learning_rate": 6.85613977718391e-05, "loss": 0.0132, "step": 41000 }, { "grad_norm": 0.14772410690784454, "learning_rate": 6.854604359940464e-05, "loss": 0.0118, "step": 41010 }, { "grad_norm": 0.12713314592838287, "learning_rate": 6.853068739880276e-05, "loss": 0.0111, "step": 41020 }, { "grad_norm": 0.1348835974931717, "learning_rate": 6.851532917171276e-05, "loss": 0.0124, "step": 41030 }, { "grad_norm": 0.18031899631023407, "learning_rate": 6.849996891981418e-05, "loss": 0.0142, "step": 41040 }, { "grad_norm": 0.17300958931446075, "learning_rate": 6.848460664478684e-05, "loss": 0.0144, "step": 41050 }, { "grad_norm": 0.23751722276210785, "learning_rate": 6.846924234831069e-05, "loss": 0.0121, "step": 41060 }, { "grad_norm": 0.12183482199907303, "learning_rate": 6.845387603206597e-05, "loss": 0.0118, "step": 41070 }, { "grad_norm": 0.18323275446891785, "learning_rate": 6.843850769773312e-05, "loss": 0.0113, "step": 41080 }, { "grad_norm": 0.14538218080997467, "learning_rate": 6.842313734699278e-05, "loss": 0.0119, "step": 41090 }, { "grad_norm": 0.15574343502521515, "learning_rate": 6.840776498152584e-05, "loss": 0.0108, "step": 41100 }, { "grad_norm": 0.1554805040359497, "learning_rate": 6.83923906030134e-05, "loss": 0.013, "step": 41110 }, { "grad_norm": 0.17764759063720703, "learning_rate": 6.837701421313677e-05, "loss": 0.0146, "step": 41120 }, { "grad_norm": 0.2043449580669403, "learning_rate": 6.83616358135775e-05, "loss": 0.011, "step": 41130 }, { "grad_norm": 0.16070184111595154, "learning_rate": 6.834625540601734e-05, "loss": 0.0122, "step": 41140 }, { "grad_norm": 0.13885092735290527, "learning_rate": 6.833087299213829e-05, "loss": 0.0114, "step": 41150 }, { "grad_norm": 0.15190188586711884, "learning_rate": 6.83154885736225e-05, "loss": 0.0105, "step": 41160 }, { "grad_norm": 0.13383829593658447, "learning_rate": 6.830010215215244e-05, "loss": 0.0101, "step": 41170 }, { "grad_norm": 0.16511626541614532, "learning_rate": 6.828471372941072e-05, "loss": 0.0147, "step": 41180 }, { "grad_norm": 0.17419768869876862, "learning_rate": 6.826932330708018e-05, "loss": 0.012, "step": 41190 }, { "grad_norm": 0.14978499710559845, "learning_rate": 6.825393088684393e-05, "loss": 0.0114, "step": 41200 }, { "grad_norm": 0.13268737494945526, "learning_rate": 6.823853647038524e-05, "loss": 0.011, "step": 41210 }, { "grad_norm": 0.13122344017028809, "learning_rate": 6.822314005938762e-05, "loss": 0.0115, "step": 41220 }, { "grad_norm": 0.17940552532672882, "learning_rate": 6.820774165553479e-05, "loss": 0.0112, "step": 41230 }, { "grad_norm": 0.13452838361263275, "learning_rate": 6.819234126051072e-05, "loss": 0.0113, "step": 41240 }, { "grad_norm": 0.14400264620780945, "learning_rate": 6.817693887599956e-05, "loss": 0.0118, "step": 41250 }, { "grad_norm": 0.12012603878974915, "learning_rate": 6.816153450368568e-05, "loss": 0.013, "step": 41260 }, { "grad_norm": 0.1217806488275528, "learning_rate": 6.814612814525372e-05, "loss": 0.0117, "step": 41270 }, { "grad_norm": 0.16704118251800537, "learning_rate": 6.813071980238846e-05, "loss": 0.0126, "step": 41280 }, { "grad_norm": 0.17187686264514923, "learning_rate": 6.811530947677492e-05, "loss": 0.0123, "step": 41290 }, { "grad_norm": 0.17926685512065887, "learning_rate": 6.809989717009839e-05, "loss": 0.0125, "step": 41300 }, { "grad_norm": 0.12714101374149323, "learning_rate": 6.808448288404431e-05, "loss": 0.0122, "step": 41310 }, { "grad_norm": 0.2431280016899109, "learning_rate": 6.806906662029838e-05, "loss": 0.0128, "step": 41320 }, { "grad_norm": 0.21016907691955566, "learning_rate": 6.805364838054647e-05, "loss": 0.0119, "step": 41330 }, { "grad_norm": 0.1551634669303894, "learning_rate": 6.803822816647474e-05, "loss": 0.0107, "step": 41340 }, { "grad_norm": 0.14093440771102905, "learning_rate": 6.802280597976949e-05, "loss": 0.0116, "step": 41350 }, { "grad_norm": 0.12932835519313812, "learning_rate": 6.800738182211727e-05, "loss": 0.0105, "step": 41360 }, { "grad_norm": 0.1555766463279724, "learning_rate": 6.799195569520485e-05, "loss": 0.0117, "step": 41370 }, { "grad_norm": 0.12767904996871948, "learning_rate": 6.79765276007192e-05, "loss": 0.0123, "step": 41380 }, { "grad_norm": 0.15736085176467896, "learning_rate": 6.796109754034753e-05, "loss": 0.0137, "step": 41390 }, { "grad_norm": 0.2203303575515747, "learning_rate": 6.794566551577724e-05, "loss": 0.0137, "step": 41400 }, { "grad_norm": 0.17435257136821747, "learning_rate": 6.793023152869594e-05, "loss": 0.0121, "step": 41410 }, { "grad_norm": 0.23973354697227478, "learning_rate": 6.791479558079147e-05, "loss": 0.0107, "step": 41420 }, { "grad_norm": 0.17454233765602112, "learning_rate": 6.789935767375191e-05, "loss": 0.0124, "step": 41430 }, { "grad_norm": 0.17089012265205383, "learning_rate": 6.78839178092655e-05, "loss": 0.0125, "step": 41440 }, { "grad_norm": 0.1452469378709793, "learning_rate": 6.786847598902072e-05, "loss": 0.0127, "step": 41450 }, { "grad_norm": 0.13570380210876465, "learning_rate": 6.785303221470627e-05, "loss": 0.0104, "step": 41460 }, { "grad_norm": 0.13660408556461334, "learning_rate": 6.783758648801108e-05, "loss": 0.0135, "step": 41470 }, { "grad_norm": 0.11253712326288223, "learning_rate": 6.782213881062422e-05, "loss": 0.0129, "step": 41480 }, { "grad_norm": 0.15531733632087708, "learning_rate": 6.780668918423508e-05, "loss": 0.0124, "step": 41490 }, { "grad_norm": 0.17719650268554688, "learning_rate": 6.779123761053317e-05, "loss": 0.0102, "step": 41500 }, { "grad_norm": 0.13398189842700958, "learning_rate": 6.777578409120826e-05, "loss": 0.012, "step": 41510 }, { "grad_norm": 0.16888052225112915, "learning_rate": 6.776032862795033e-05, "loss": 0.0124, "step": 41520 }, { "grad_norm": 0.13129372894763947, "learning_rate": 6.774487122244956e-05, "loss": 0.0103, "step": 41530 }, { "grad_norm": 0.18335913121700287, "learning_rate": 6.772941187639637e-05, "loss": 0.0108, "step": 41540 }, { "grad_norm": 0.12127657234668732, "learning_rate": 6.771395059148134e-05, "loss": 0.0109, "step": 41550 }, { "grad_norm": 0.1824617236852646, "learning_rate": 6.76984873693953e-05, "loss": 0.0126, "step": 41560 }, { "grad_norm": 0.12377551943063736, "learning_rate": 6.768302221182931e-05, "loss": 0.0115, "step": 41570 }, { "grad_norm": 0.1720828413963318, "learning_rate": 6.766755512047457e-05, "loss": 0.0111, "step": 41580 }, { "grad_norm": 0.13229715824127197, "learning_rate": 6.765208609702259e-05, "loss": 0.0114, "step": 41590 }, { "grad_norm": 0.16830068826675415, "learning_rate": 6.763661514316499e-05, "loss": 0.011, "step": 41600 }, { "grad_norm": 0.18092717230319977, "learning_rate": 6.76211422605937e-05, "loss": 0.0114, "step": 41610 }, { "grad_norm": 0.38597556948661804, "learning_rate": 6.760566745100076e-05, "loss": 0.0125, "step": 41620 }, { "grad_norm": 0.1607353538274765, "learning_rate": 6.75901907160785e-05, "loss": 0.0112, "step": 41630 }, { "grad_norm": 0.16035093367099762, "learning_rate": 6.757471205751943e-05, "loss": 0.011, "step": 41640 }, { "grad_norm": 0.18082496523857117, "learning_rate": 6.75592314770163e-05, "loss": 0.0111, "step": 41650 }, { "grad_norm": 0.1218644380569458, "learning_rate": 6.7543748976262e-05, "loss": 0.0125, "step": 41660 }, { "grad_norm": 0.15907923877239227, "learning_rate": 6.752826455694968e-05, "loss": 0.0124, "step": 41670 }, { "grad_norm": 0.16912415623664856, "learning_rate": 6.751277822077271e-05, "loss": 0.0099, "step": 41680 }, { "grad_norm": 0.1341959536075592, "learning_rate": 6.749728996942465e-05, "loss": 0.0117, "step": 41690 }, { "grad_norm": 0.1829524040222168, "learning_rate": 6.748179980459924e-05, "loss": 0.0123, "step": 41700 }, { "grad_norm": 0.13173137605190277, "learning_rate": 6.746630772799052e-05, "loss": 0.011, "step": 41710 }, { "grad_norm": 0.1538497507572174, "learning_rate": 6.745081374129262e-05, "loss": 0.0115, "step": 41720 }, { "grad_norm": 0.16426412761211395, "learning_rate": 6.743531784619998e-05, "loss": 0.0119, "step": 41730 }, { "grad_norm": 0.19001446664333344, "learning_rate": 6.741982004440719e-05, "loss": 0.012, "step": 41740 }, { "grad_norm": 0.1795348972082138, "learning_rate": 6.740432033760907e-05, "loss": 0.0104, "step": 41750 }, { "grad_norm": 0.165652334690094, "learning_rate": 6.738881872750066e-05, "loss": 0.0115, "step": 41760 }, { "grad_norm": 0.2415633499622345, "learning_rate": 6.737331521577715e-05, "loss": 0.0115, "step": 41770 }, { "grad_norm": 0.20943251252174377, "learning_rate": 6.735780980413403e-05, "loss": 0.0139, "step": 41780 }, { "grad_norm": 0.16513288021087646, "learning_rate": 6.734230249426692e-05, "loss": 0.013, "step": 41790 }, { "grad_norm": 0.2636071741580963, "learning_rate": 6.732679328787168e-05, "loss": 0.0128, "step": 41800 }, { "grad_norm": 0.18963655829429626, "learning_rate": 6.731128218664438e-05, "loss": 0.0122, "step": 41810 }, { "grad_norm": 0.16053515672683716, "learning_rate": 6.729576919228129e-05, "loss": 0.0138, "step": 41820 }, { "grad_norm": 0.14972218871116638, "learning_rate": 6.728025430647888e-05, "loss": 0.0149, "step": 41830 }, { "grad_norm": 0.1678619086742401, "learning_rate": 6.726473753093383e-05, "loss": 0.0128, "step": 41840 }, { "grad_norm": 0.18695107102394104, "learning_rate": 6.724921886734305e-05, "loss": 0.0151, "step": 41850 }, { "grad_norm": 0.43059200048446655, "learning_rate": 6.723369831740363e-05, "loss": 0.0138, "step": 41860 }, { "grad_norm": 0.15419401228427887, "learning_rate": 6.721817588281288e-05, "loss": 0.0124, "step": 41870 }, { "grad_norm": 0.1437041163444519, "learning_rate": 6.720265156526828e-05, "loss": 0.0141, "step": 41880 }, { "grad_norm": 0.18368105590343475, "learning_rate": 6.718712536646758e-05, "loss": 0.013, "step": 41890 }, { "grad_norm": 0.19809240102767944, "learning_rate": 6.71715972881087e-05, "loss": 0.0144, "step": 41900 }, { "grad_norm": 0.16410084068775177, "learning_rate": 6.715606733188974e-05, "loss": 0.0142, "step": 41910 }, { "grad_norm": 0.22624197602272034, "learning_rate": 6.714053549950909e-05, "loss": 0.012, "step": 41920 }, { "grad_norm": 0.14671111106872559, "learning_rate": 6.71250017926652e-05, "loss": 0.0145, "step": 41930 }, { "grad_norm": 0.18643909692764282, "learning_rate": 6.71094662130569e-05, "loss": 0.0133, "step": 41940 }, { "grad_norm": 0.16301937401294708, "learning_rate": 6.709392876238307e-05, "loss": 0.0133, "step": 41950 }, { "grad_norm": 0.1394805759191513, "learning_rate": 6.707838944234289e-05, "loss": 0.0116, "step": 41960 }, { "grad_norm": 0.14903873205184937, "learning_rate": 6.706284825463574e-05, "loss": 0.0143, "step": 41970 }, { "grad_norm": 0.12665283679962158, "learning_rate": 6.704730520096114e-05, "loss": 0.0122, "step": 41980 }, { "grad_norm": 0.15429329872131348, "learning_rate": 6.703176028301888e-05, "loss": 0.01, "step": 41990 }, { "grad_norm": 0.18060524761676788, "learning_rate": 6.701621350250892e-05, "loss": 0.0103, "step": 42000 }, { "grad_norm": 0.13318972289562225, "learning_rate": 6.700066486113144e-05, "loss": 0.0127, "step": 42010 }, { "grad_norm": 0.21633730828762054, "learning_rate": 6.69851143605868e-05, "loss": 0.0112, "step": 42020 }, { "grad_norm": 0.13693007826805115, "learning_rate": 6.696956200257561e-05, "loss": 0.0107, "step": 42030 }, { "grad_norm": 0.14072537422180176, "learning_rate": 6.695400778879863e-05, "loss": 0.0116, "step": 42040 }, { "grad_norm": 0.14910313487052917, "learning_rate": 6.693845172095683e-05, "loss": 0.0116, "step": 42050 }, { "grad_norm": 0.1865144968032837, "learning_rate": 6.692289380075142e-05, "loss": 0.0153, "step": 42060 }, { "grad_norm": 0.12207360565662384, "learning_rate": 6.690733402988379e-05, "loss": 0.016, "step": 42070 }, { "grad_norm": 0.1362762302160263, "learning_rate": 6.689177241005553e-05, "loss": 0.0121, "step": 42080 }, { "grad_norm": 0.18075880408287048, "learning_rate": 6.687620894296844e-05, "loss": 0.0108, "step": 42090 }, { "grad_norm": 0.11923978477716446, "learning_rate": 6.686064363032451e-05, "loss": 0.0105, "step": 42100 }, { "grad_norm": 0.16690900921821594, "learning_rate": 6.684507647382596e-05, "loss": 0.0134, "step": 42110 }, { "grad_norm": 0.14709258079528809, "learning_rate": 6.682950747517513e-05, "loss": 0.0107, "step": 42120 }, { "grad_norm": 0.17016221582889557, "learning_rate": 6.68139366360747e-05, "loss": 0.0111, "step": 42130 }, { "grad_norm": 0.16082005202770233, "learning_rate": 6.679836395822744e-05, "loss": 0.0118, "step": 42140 }, { "grad_norm": 0.15720541775226593, "learning_rate": 6.678278944333633e-05, "loss": 0.0113, "step": 42150 }, { "grad_norm": 0.14798325300216675, "learning_rate": 6.676721309310462e-05, "loss": 0.0123, "step": 42160 }, { "grad_norm": 0.18593506515026093, "learning_rate": 6.67516349092357e-05, "loss": 0.0144, "step": 42170 }, { "grad_norm": 0.09690088033676147, "learning_rate": 6.673605489343315e-05, "loss": 0.0121, "step": 42180 }, { "grad_norm": 0.1614646166563034, "learning_rate": 6.672047304740081e-05, "loss": 0.0114, "step": 42190 }, { "grad_norm": 0.15161731839179993, "learning_rate": 6.670488937284268e-05, "loss": 0.0132, "step": 42200 }, { "grad_norm": 0.1710667908191681, "learning_rate": 6.668930387146294e-05, "loss": 0.0121, "step": 42210 }, { "grad_norm": 0.17188116908073425, "learning_rate": 6.667371654496605e-05, "loss": 0.0124, "step": 42220 }, { "grad_norm": 0.19178365170955658, "learning_rate": 6.665812739505659e-05, "loss": 0.0112, "step": 42230 }, { "grad_norm": 0.13525240123271942, "learning_rate": 6.664253642343935e-05, "loss": 0.0106, "step": 42240 }, { "grad_norm": 0.14418363571166992, "learning_rate": 6.662694363181935e-05, "loss": 0.0108, "step": 42250 }, { "grad_norm": 0.144066721200943, "learning_rate": 6.66113490219018e-05, "loss": 0.0143, "step": 42260 }, { "grad_norm": 0.18283767998218536, "learning_rate": 6.659575259539207e-05, "loss": 0.0119, "step": 42270 }, { "grad_norm": 0.15967057645320892, "learning_rate": 6.65801543539958e-05, "loss": 0.0112, "step": 42280 }, { "grad_norm": 0.1797247678041458, "learning_rate": 6.65645542994188e-05, "loss": 0.0117, "step": 42290 }, { "grad_norm": 0.22546254098415375, "learning_rate": 6.654895243336702e-05, "loss": 0.0103, "step": 42300 }, { "grad_norm": 0.20369243621826172, "learning_rate": 6.65333487575467e-05, "loss": 0.0111, "step": 42310 }, { "grad_norm": 0.14127053320407867, "learning_rate": 6.651774327366422e-05, "loss": 0.011, "step": 42320 }, { "grad_norm": 0.12803703546524048, "learning_rate": 6.650213598342613e-05, "loss": 0.0111, "step": 42330 }, { "grad_norm": 0.1343618929386139, "learning_rate": 6.648652688853931e-05, "loss": 0.0099, "step": 42340 }, { "grad_norm": 0.19729898869991302, "learning_rate": 6.647091599071066e-05, "loss": 0.0116, "step": 42350 }, { "grad_norm": 0.10689252614974976, "learning_rate": 6.645530329164743e-05, "loss": 0.0095, "step": 42360 }, { "grad_norm": 0.24260112643241882, "learning_rate": 6.643968879305694e-05, "loss": 0.0147, "step": 42370 }, { "grad_norm": 0.1491769701242447, "learning_rate": 6.642407249664683e-05, "loss": 0.0128, "step": 42380 }, { "grad_norm": 0.11297577619552612, "learning_rate": 6.640845440412483e-05, "loss": 0.0113, "step": 42390 }, { "grad_norm": 0.1651027500629425, "learning_rate": 6.639283451719893e-05, "loss": 0.0107, "step": 42400 }, { "grad_norm": 0.10285834968090057, "learning_rate": 6.637721283757729e-05, "loss": 0.0117, "step": 42410 }, { "grad_norm": 0.16629892587661743, "learning_rate": 6.636158936696828e-05, "loss": 0.0125, "step": 42420 }, { "grad_norm": 0.15100552141666412, "learning_rate": 6.634596410708047e-05, "loss": 0.013, "step": 42430 }, { "grad_norm": 0.1539221704006195, "learning_rate": 6.633033705962257e-05, "loss": 0.0135, "step": 42440 }, { "grad_norm": 0.16035090386867523, "learning_rate": 6.631470822630359e-05, "loss": 0.012, "step": 42450 }, { "grad_norm": 0.16172046959400177, "learning_rate": 6.629907760883263e-05, "loss": 0.0101, "step": 42460 }, { "grad_norm": 0.18886543810367584, "learning_rate": 6.628344520891905e-05, "loss": 0.0126, "step": 42470 }, { "grad_norm": 0.15474557876586914, "learning_rate": 6.62678110282724e-05, "loss": 0.0111, "step": 42480 }, { "grad_norm": 0.2003946453332901, "learning_rate": 6.625217506860237e-05, "loss": 0.0115, "step": 42490 }, { "grad_norm": 0.21722735464572906, "learning_rate": 6.623653733161892e-05, "loss": 0.0123, "step": 42500 }, { "grad_norm": 0.11916518211364746, "learning_rate": 6.622089781903217e-05, "loss": 0.0099, "step": 42510 }, { "grad_norm": 0.17528696358203888, "learning_rate": 6.620525653255241e-05, "loss": 0.0117, "step": 42520 }, { "grad_norm": 0.17598436772823334, "learning_rate": 6.618961347389015e-05, "loss": 0.0112, "step": 42530 }, { "grad_norm": 0.15274658799171448, "learning_rate": 6.617396864475613e-05, "loss": 0.0122, "step": 42540 }, { "grad_norm": 0.17097900807857513, "learning_rate": 6.61583220468612e-05, "loss": 0.0109, "step": 42550 }, { "grad_norm": 0.18812820315361023, "learning_rate": 6.614267368191645e-05, "loss": 0.0143, "step": 42560 }, { "grad_norm": 0.15878625214099884, "learning_rate": 6.61270235516332e-05, "loss": 0.0126, "step": 42570 }, { "grad_norm": 0.17400890588760376, "learning_rate": 6.611137165772287e-05, "loss": 0.0113, "step": 42580 }, { "grad_norm": 0.14590469002723694, "learning_rate": 6.60957180018972e-05, "loss": 0.0136, "step": 42590 }, { "grad_norm": 0.14649002254009247, "learning_rate": 6.608006258586797e-05, "loss": 0.0114, "step": 42600 }, { "grad_norm": 0.2156181037425995, "learning_rate": 6.606440541134731e-05, "loss": 0.0115, "step": 42610 }, { "grad_norm": 0.1860159933567047, "learning_rate": 6.60487464800474e-05, "loss": 0.01, "step": 42620 }, { "grad_norm": 0.20274974405765533, "learning_rate": 6.603308579368071e-05, "loss": 0.0131, "step": 42630 }, { "grad_norm": 0.2311193197965622, "learning_rate": 6.601742335395987e-05, "loss": 0.013, "step": 42640 }, { "grad_norm": 0.13998542726039886, "learning_rate": 6.600175916259769e-05, "loss": 0.0097, "step": 42650 }, { "grad_norm": 0.13044914603233337, "learning_rate": 6.598609322130718e-05, "loss": 0.0118, "step": 42660 }, { "grad_norm": 0.13192319869995117, "learning_rate": 6.597042553180155e-05, "loss": 0.0123, "step": 42670 }, { "grad_norm": 0.16350960731506348, "learning_rate": 6.59547560957942e-05, "loss": 0.0111, "step": 42680 }, { "grad_norm": 0.23209354281425476, "learning_rate": 6.59390849149987e-05, "loss": 0.0134, "step": 42690 }, { "grad_norm": 0.20085909962654114, "learning_rate": 6.592341199112886e-05, "loss": 0.0126, "step": 42700 }, { "grad_norm": 0.1656549870967865, "learning_rate": 6.590773732589861e-05, "loss": 0.0119, "step": 42710 }, { "grad_norm": 0.15486282110214233, "learning_rate": 6.589206092102213e-05, "loss": 0.0123, "step": 42720 }, { "grad_norm": 0.16191954910755157, "learning_rate": 6.587638277821375e-05, "loss": 0.0123, "step": 42730 }, { "grad_norm": 0.1414187103509903, "learning_rate": 6.586070289918803e-05, "loss": 0.011, "step": 42740 }, { "grad_norm": 0.16113847494125366, "learning_rate": 6.584502128565968e-05, "loss": 0.0131, "step": 42750 }, { "grad_norm": 0.21444746851921082, "learning_rate": 6.582933793934363e-05, "loss": 0.0095, "step": 42760 }, { "grad_norm": 0.16690748929977417, "learning_rate": 6.581365286195498e-05, "loss": 0.0116, "step": 42770 }, { "grad_norm": 0.18102726340293884, "learning_rate": 6.579796605520903e-05, "loss": 0.0122, "step": 42780 }, { "grad_norm": 0.17244963347911835, "learning_rate": 6.578227752082127e-05, "loss": 0.0107, "step": 42790 }, { "grad_norm": 0.3647846579551697, "learning_rate": 6.576658726050735e-05, "loss": 0.0091, "step": 42800 }, { "grad_norm": 0.15726186335086823, "learning_rate": 6.575089527598316e-05, "loss": 0.0121, "step": 42810 }, { "grad_norm": 0.16603732109069824, "learning_rate": 6.573520156896476e-05, "loss": 0.0115, "step": 42820 }, { "grad_norm": 0.20552591979503632, "learning_rate": 6.571950614116835e-05, "loss": 0.011, "step": 42830 }, { "grad_norm": 0.16430048644542694, "learning_rate": 6.570380899431039e-05, "loss": 0.0119, "step": 42840 }, { "grad_norm": 0.14443339407444, "learning_rate": 6.568811013010749e-05, "loss": 0.0132, "step": 42850 }, { "grad_norm": 0.131360724568367, "learning_rate": 6.567240955027645e-05, "loss": 0.0109, "step": 42860 }, { "grad_norm": 0.1537703424692154, "learning_rate": 6.565670725653427e-05, "loss": 0.0113, "step": 42870 }, { "grad_norm": 0.13866755366325378, "learning_rate": 6.564100325059812e-05, "loss": 0.0104, "step": 42880 }, { "grad_norm": 0.15497078001499176, "learning_rate": 6.562529753418538e-05, "loss": 0.0115, "step": 42890 }, { "grad_norm": 0.17748181521892548, "learning_rate": 6.56095901090136e-05, "loss": 0.012, "step": 42900 }, { "grad_norm": 0.15130403637886047, "learning_rate": 6.55938809768005e-05, "loss": 0.0108, "step": 42910 }, { "grad_norm": 0.1358916312456131, "learning_rate": 6.557817013926402e-05, "loss": 0.013, "step": 42920 }, { "grad_norm": 0.1386948674917221, "learning_rate": 6.556245759812227e-05, "loss": 0.0104, "step": 42930 }, { "grad_norm": 0.2138194441795349, "learning_rate": 6.554674335509357e-05, "loss": 0.01, "step": 42940 }, { "grad_norm": 0.15817607939243317, "learning_rate": 6.553102741189638e-05, "loss": 0.011, "step": 42950 }, { "grad_norm": 0.1911112368106842, "learning_rate": 6.55153097702494e-05, "loss": 0.011, "step": 42960 }, { "grad_norm": 0.14271292090415955, "learning_rate": 6.549959043187146e-05, "loss": 0.012, "step": 42970 }, { "grad_norm": 0.1272440403699875, "learning_rate": 6.54838693984816e-05, "loss": 0.012, "step": 42980 }, { "grad_norm": 0.13061846792697906, "learning_rate": 6.546814667179909e-05, "loss": 0.011, "step": 42990 }, { "grad_norm": 0.22465963661670685, "learning_rate": 6.545242225354328e-05, "loss": 0.0124, "step": 43000 }, { "grad_norm": 0.19476106762886047, "learning_rate": 6.543669614543383e-05, "loss": 0.0135, "step": 43010 }, { "grad_norm": 0.19181181490421295, "learning_rate": 6.542096834919049e-05, "loss": 0.013, "step": 43020 }, { "grad_norm": 0.2122710645198822, "learning_rate": 6.540523886653324e-05, "loss": 0.012, "step": 43030 }, { "grad_norm": 0.24891337752342224, "learning_rate": 6.538950769918222e-05, "loss": 0.0141, "step": 43040 }, { "grad_norm": 0.162422314286232, "learning_rate": 6.537377484885779e-05, "loss": 0.0123, "step": 43050 }, { "grad_norm": 0.1876281499862671, "learning_rate": 6.535804031728045e-05, "loss": 0.012, "step": 43060 }, { "grad_norm": 0.2491196244955063, "learning_rate": 6.534230410617092e-05, "loss": 0.0127, "step": 43070 }, { "grad_norm": 0.1487247496843338, "learning_rate": 6.532656621725007e-05, "loss": 0.0116, "step": 43080 }, { "grad_norm": 0.1703576296567917, "learning_rate": 6.531082665223898e-05, "loss": 0.012, "step": 43090 }, { "grad_norm": 0.17458294332027435, "learning_rate": 6.529508541285889e-05, "loss": 0.0112, "step": 43100 }, { "grad_norm": 0.15866516530513763, "learning_rate": 6.527934250083128e-05, "loss": 0.0103, "step": 43110 }, { "grad_norm": 0.13584673404693604, "learning_rate": 6.526359791787775e-05, "loss": 0.0108, "step": 43120 }, { "grad_norm": 0.28107625246047974, "learning_rate": 6.524785166572009e-05, "loss": 0.0117, "step": 43130 }, { "grad_norm": 0.18632619082927704, "learning_rate": 6.523210374608029e-05, "loss": 0.0105, "step": 43140 }, { "grad_norm": 0.13861528038978577, "learning_rate": 6.521635416068054e-05, "loss": 0.0123, "step": 43150 }, { "grad_norm": 0.1392243653535843, "learning_rate": 6.520060291124317e-05, "loss": 0.0123, "step": 43160 }, { "grad_norm": 0.15135900676250458, "learning_rate": 6.518484999949072e-05, "loss": 0.0129, "step": 43170 }, { "grad_norm": 0.14251606166362762, "learning_rate": 6.516909542714591e-05, "loss": 0.0116, "step": 43180 }, { "grad_norm": 0.2090378999710083, "learning_rate": 6.515333919593161e-05, "loss": 0.011, "step": 43190 }, { "grad_norm": 0.14670231938362122, "learning_rate": 6.513758130757094e-05, "loss": 0.0121, "step": 43200 }, { "grad_norm": 0.17404073476791382, "learning_rate": 6.512182176378713e-05, "loss": 0.0106, "step": 43210 }, { "grad_norm": 0.1649545282125473, "learning_rate": 6.510606056630362e-05, "loss": 0.0134, "step": 43220 }, { "grad_norm": 0.17020538449287415, "learning_rate": 6.509029771684405e-05, "loss": 0.0113, "step": 43230 }, { "grad_norm": 0.11443687230348587, "learning_rate": 6.50745332171322e-05, "loss": 0.0114, "step": 43240 }, { "grad_norm": 0.12581659853458405, "learning_rate": 6.505876706889207e-05, "loss": 0.0093, "step": 43250 }, { "grad_norm": 0.20448599755764008, "learning_rate": 6.504299927384781e-05, "loss": 0.0104, "step": 43260 }, { "grad_norm": 0.17449918389320374, "learning_rate": 6.502722983372378e-05, "loss": 0.0127, "step": 43270 }, { "grad_norm": 0.13475514948368073, "learning_rate": 6.501145875024446e-05, "loss": 0.0133, "step": 43280 }, { "grad_norm": 0.1206444576382637, "learning_rate": 6.499568602513459e-05, "loss": 0.0098, "step": 43290 }, { "grad_norm": 0.21430492401123047, "learning_rate": 6.497991166011903e-05, "loss": 0.0107, "step": 43300 }, { "grad_norm": 0.22678494453430176, "learning_rate": 6.49641356569229e-05, "loss": 0.0108, "step": 43310 }, { "grad_norm": 0.18965178728103638, "learning_rate": 6.494835801727135e-05, "loss": 0.012, "step": 43320 }, { "grad_norm": 0.11473917961120605, "learning_rate": 6.493257874288986e-05, "loss": 0.0108, "step": 43330 }, { "grad_norm": 0.16319026052951813, "learning_rate": 6.491679783550399e-05, "loss": 0.0114, "step": 43340 }, { "grad_norm": 0.13989461958408356, "learning_rate": 6.490101529683957e-05, "loss": 0.0113, "step": 43350 }, { "grad_norm": 0.15714703500270844, "learning_rate": 6.488523112862248e-05, "loss": 0.0106, "step": 43360 }, { "grad_norm": 0.16793549060821533, "learning_rate": 6.486944533257891e-05, "loss": 0.0121, "step": 43370 }, { "grad_norm": 0.17875032126903534, "learning_rate": 6.485365791043517e-05, "loss": 0.0125, "step": 43380 }, { "grad_norm": 0.13603103160858154, "learning_rate": 6.483786886391772e-05, "loss": 0.01, "step": 43390 }, { "grad_norm": 0.20367960631847382, "learning_rate": 6.482207819475323e-05, "loss": 0.0099, "step": 43400 }, { "grad_norm": 0.1364472657442093, "learning_rate": 6.480628590466857e-05, "loss": 0.01, "step": 43410 }, { "grad_norm": 0.16880060732364655, "learning_rate": 6.479049199539074e-05, "loss": 0.0104, "step": 43420 }, { "grad_norm": 0.16904525458812714, "learning_rate": 6.477469646864693e-05, "loss": 0.01, "step": 43430 }, { "grad_norm": 0.20255903899669647, "learning_rate": 6.475889932616454e-05, "loss": 0.0104, "step": 43440 }, { "grad_norm": 0.15356196463108063, "learning_rate": 6.474310056967111e-05, "loss": 0.0105, "step": 43450 }, { "grad_norm": 0.15845955908298492, "learning_rate": 6.472730020089437e-05, "loss": 0.0103, "step": 43460 }, { "grad_norm": 0.14569775760173798, "learning_rate": 6.471149822156222e-05, "loss": 0.0109, "step": 43470 }, { "grad_norm": 0.10968679934740067, "learning_rate": 6.469569463340275e-05, "loss": 0.0118, "step": 43480 }, { "grad_norm": 0.2514677345752716, "learning_rate": 6.467988943814418e-05, "loss": 0.0137, "step": 43490 }, { "grad_norm": 0.14225012063980103, "learning_rate": 6.4664082637515e-05, "loss": 0.0129, "step": 43500 }, { "grad_norm": 0.11232155561447144, "learning_rate": 6.464827423324377e-05, "loss": 0.0126, "step": 43510 }, { "grad_norm": 0.19369161128997803, "learning_rate": 6.463246422705931e-05, "loss": 0.0127, "step": 43520 }, { "grad_norm": 0.1422833949327469, "learning_rate": 6.461665262069055e-05, "loss": 0.0134, "step": 43530 }, { "grad_norm": 0.17268291115760803, "learning_rate": 6.460083941586665e-05, "loss": 0.0118, "step": 43540 }, { "grad_norm": 0.1810864508152008, "learning_rate": 6.45850246143169e-05, "loss": 0.014, "step": 43550 }, { "grad_norm": 0.1778060495853424, "learning_rate": 6.456920821777077e-05, "loss": 0.0135, "step": 43560 }, { "grad_norm": 0.15944653749465942, "learning_rate": 6.455339022795796e-05, "loss": 0.0122, "step": 43570 }, { "grad_norm": 0.13689382374286652, "learning_rate": 6.453757064660825e-05, "loss": 0.0115, "step": 43580 }, { "grad_norm": 0.14841283857822418, "learning_rate": 6.452174947545169e-05, "loss": 0.0137, "step": 43590 }, { "grad_norm": 0.12785091996192932, "learning_rate": 6.450592671621842e-05, "loss": 0.0137, "step": 43600 }, { "grad_norm": 0.20196588337421417, "learning_rate": 6.449010237063882e-05, "loss": 0.0138, "step": 43610 }, { "grad_norm": 0.19812721014022827, "learning_rate": 6.447427644044343e-05, "loss": 0.0129, "step": 43620 }, { "grad_norm": 0.1995416134595871, "learning_rate": 6.44584489273629e-05, "loss": 0.0138, "step": 43630 }, { "grad_norm": 0.1770942658185959, "learning_rate": 6.444261983312817e-05, "loss": 0.0109, "step": 43640 }, { "grad_norm": 0.19506342709064484, "learning_rate": 6.442678915947023e-05, "loss": 0.0126, "step": 43650 }, { "grad_norm": 0.09728382527828217, "learning_rate": 6.441095690812032e-05, "loss": 0.0151, "step": 43660 }, { "grad_norm": 0.20399849116802216, "learning_rate": 6.439512308080982e-05, "loss": 0.0108, "step": 43670 }, { "grad_norm": 0.13278643786907196, "learning_rate": 6.43792876792703e-05, "loss": 0.0117, "step": 43680 }, { "grad_norm": 0.19226378202438354, "learning_rate": 6.43634507052335e-05, "loss": 0.0141, "step": 43690 }, { "grad_norm": 0.176580011844635, "learning_rate": 6.434761216043135e-05, "loss": 0.0113, "step": 43700 }, { "grad_norm": 0.2881716191768646, "learning_rate": 6.433177204659589e-05, "loss": 0.0134, "step": 43710 }, { "grad_norm": 0.22929294407367706, "learning_rate": 6.431593036545938e-05, "loss": 0.0138, "step": 43720 }, { "grad_norm": 0.1698792576789856, "learning_rate": 6.430008711875426e-05, "loss": 0.0146, "step": 43730 }, { "grad_norm": 0.1308266669511795, "learning_rate": 6.42842423082131e-05, "loss": 0.0122, "step": 43740 }, { "grad_norm": 0.20025202631950378, "learning_rate": 6.426839593556868e-05, "loss": 0.0125, "step": 43750 }, { "grad_norm": 0.16538849472999573, "learning_rate": 6.425254800255392e-05, "loss": 0.0147, "step": 43760 }, { "grad_norm": 0.09689357876777649, "learning_rate": 6.423669851090196e-05, "loss": 0.0123, "step": 43770 }, { "grad_norm": 0.14954520761966705, "learning_rate": 6.422084746234604e-05, "loss": 0.0118, "step": 43780 }, { "grad_norm": 0.17868410050868988, "learning_rate": 6.420499485861965e-05, "loss": 0.0107, "step": 43790 }, { "grad_norm": 0.19238801300525665, "learning_rate": 6.418914070145636e-05, "loss": 0.0135, "step": 43800 }, { "grad_norm": 0.10704004764556885, "learning_rate": 6.417328499258996e-05, "loss": 0.0116, "step": 43810 }, { "grad_norm": 0.16895410418510437, "learning_rate": 6.415742773375446e-05, "loss": 0.0119, "step": 43820 }, { "grad_norm": 0.5253299474716187, "learning_rate": 6.414156892668393e-05, "loss": 0.0115, "step": 43830 }, { "grad_norm": 0.14781297743320465, "learning_rate": 6.412570857311267e-05, "loss": 0.0143, "step": 43840 }, { "grad_norm": 0.14062368869781494, "learning_rate": 6.410984667477518e-05, "loss": 0.0129, "step": 43850 }, { "grad_norm": 0.17993749678134918, "learning_rate": 6.409398323340607e-05, "loss": 0.0117, "step": 43860 }, { "grad_norm": 0.21318189799785614, "learning_rate": 6.407811825074012e-05, "loss": 0.0134, "step": 43870 }, { "grad_norm": 0.1693698763847351, "learning_rate": 6.406225172851234e-05, "loss": 0.0121, "step": 43880 }, { "grad_norm": 0.20663680136203766, "learning_rate": 6.404638366845786e-05, "loss": 0.0134, "step": 43890 }, { "grad_norm": 0.18470431864261627, "learning_rate": 6.403051407231196e-05, "loss": 0.0116, "step": 43900 }, { "grad_norm": 0.1449466347694397, "learning_rate": 6.401464294181016e-05, "loss": 0.0122, "step": 43910 }, { "grad_norm": 0.11883822083473206, "learning_rate": 6.399877027868808e-05, "loss": 0.0118, "step": 43920 }, { "grad_norm": 0.15647612512111664, "learning_rate": 6.39828960846815e-05, "loss": 0.012, "step": 43930 }, { "grad_norm": 0.19409921765327454, "learning_rate": 6.396702036152644e-05, "loss": 0.0113, "step": 43940 }, { "grad_norm": 0.2116292119026184, "learning_rate": 6.395114311095905e-05, "loss": 0.0108, "step": 43950 }, { "grad_norm": 0.24853597581386566, "learning_rate": 6.393526433471559e-05, "loss": 0.0119, "step": 43960 }, { "grad_norm": 0.16566354036331177, "learning_rate": 6.391938403453262e-05, "loss": 0.0117, "step": 43970 }, { "grad_norm": 0.20420706272125244, "learning_rate": 6.390350221214671e-05, "loss": 0.0127, "step": 43980 }, { "grad_norm": 0.1813725233078003, "learning_rate": 6.38876188692947e-05, "loss": 0.0119, "step": 43990 }, { "grad_norm": 0.22467516362667084, "learning_rate": 6.387173400771358e-05, "loss": 0.0096, "step": 44000 }, { "grad_norm": 0.12912750244140625, "learning_rate": 6.385584762914047e-05, "loss": 0.01, "step": 44010 }, { "grad_norm": 0.16369345784187317, "learning_rate": 6.38399597353127e-05, "loss": 0.0114, "step": 44020 }, { "grad_norm": 0.21818780899047852, "learning_rate": 6.382407032796775e-05, "loss": 0.012, "step": 44030 }, { "grad_norm": 0.1695113480091095, "learning_rate": 6.380817940884325e-05, "loss": 0.012, "step": 44040 }, { "grad_norm": 0.14604440331459045, "learning_rate": 6.379228697967702e-05, "loss": 0.0122, "step": 44050 }, { "grad_norm": 0.16705617308616638, "learning_rate": 6.3776393042207e-05, "loss": 0.0128, "step": 44060 }, { "grad_norm": 0.29825863242149353, "learning_rate": 6.376049759817137e-05, "loss": 0.0129, "step": 44070 }, { "grad_norm": 0.15940696001052856, "learning_rate": 6.37446006493084e-05, "loss": 0.0157, "step": 44080 }, { "grad_norm": 0.1629469096660614, "learning_rate": 6.372870219735656e-05, "loss": 0.0146, "step": 44090 }, { "grad_norm": 0.1554054617881775, "learning_rate": 6.371280224405452e-05, "loss": 0.0125, "step": 44100 }, { "grad_norm": 0.1455163210630417, "learning_rate": 6.369690079114101e-05, "loss": 0.0142, "step": 44110 }, { "grad_norm": 0.1010839194059372, "learning_rate": 6.368099784035504e-05, "loss": 0.0121, "step": 44120 }, { "grad_norm": 0.1442566066980362, "learning_rate": 6.366509339343572e-05, "loss": 0.0124, "step": 44130 }, { "grad_norm": 0.16527801752090454, "learning_rate": 6.364918745212232e-05, "loss": 0.0115, "step": 44140 }, { "grad_norm": 0.1257210522890091, "learning_rate": 6.363328001815431e-05, "loss": 0.0136, "step": 44150 }, { "grad_norm": 0.1305142194032669, "learning_rate": 6.361737109327128e-05, "loss": 0.0142, "step": 44160 }, { "grad_norm": 0.15109305083751678, "learning_rate": 6.360146067921301e-05, "loss": 0.0128, "step": 44170 }, { "grad_norm": 0.15021704137325287, "learning_rate": 6.35855487777195e-05, "loss": 0.0119, "step": 44180 }, { "grad_norm": 0.15041586756706238, "learning_rate": 6.356963539053076e-05, "loss": 0.0108, "step": 44190 }, { "grad_norm": 0.16627460718154907, "learning_rate": 6.355372051938711e-05, "loss": 0.0113, "step": 44200 }, { "grad_norm": 0.10054576396942139, "learning_rate": 6.353780416602894e-05, "loss": 0.0132, "step": 44210 }, { "grad_norm": 0.14336653053760529, "learning_rate": 6.352188633219689e-05, "loss": 0.0122, "step": 44220 }, { "grad_norm": 0.18368007242679596, "learning_rate": 6.350596701963166e-05, "loss": 0.0106, "step": 44230 }, { "grad_norm": 0.20103581249713898, "learning_rate": 6.349004623007419e-05, "loss": 0.0149, "step": 44240 }, { "grad_norm": 0.18628551065921783, "learning_rate": 6.347412396526555e-05, "loss": 0.0115, "step": 44250 }, { "grad_norm": 0.12464535236358643, "learning_rate": 6.345820022694696e-05, "loss": 0.0113, "step": 44260 }, { "grad_norm": 0.13455785810947418, "learning_rate": 6.344227501685984e-05, "loss": 0.0117, "step": 44270 }, { "grad_norm": 0.16719456017017365, "learning_rate": 6.342634833674572e-05, "loss": 0.013, "step": 44280 }, { "grad_norm": 0.1487097591161728, "learning_rate": 6.341042018834635e-05, "loss": 0.0114, "step": 44290 }, { "grad_norm": 0.15573428571224213, "learning_rate": 6.339449057340359e-05, "loss": 0.009, "step": 44300 }, { "grad_norm": 0.1362849324941635, "learning_rate": 6.337855949365945e-05, "loss": 0.0091, "step": 44310 }, { "grad_norm": 0.1136585995554924, "learning_rate": 6.336262695085619e-05, "loss": 0.011, "step": 44320 }, { "grad_norm": 0.15053072571754456, "learning_rate": 6.334669294673612e-05, "loss": 0.0103, "step": 44330 }, { "grad_norm": 0.1504662036895752, "learning_rate": 6.333075748304179e-05, "loss": 0.0095, "step": 44340 }, { "grad_norm": 0.129134863615036, "learning_rate": 6.331482056151585e-05, "loss": 0.0098, "step": 44350 }, { "grad_norm": 0.1453971117734909, "learning_rate": 6.329888218390117e-05, "loss": 0.0109, "step": 44360 }, { "grad_norm": 0.16366083920001984, "learning_rate": 6.328294235194072e-05, "loss": 0.0104, "step": 44370 }, { "grad_norm": 0.17657485604286194, "learning_rate": 6.326700106737765e-05, "loss": 0.0105, "step": 44380 }, { "grad_norm": 0.1790255457162857, "learning_rate": 6.325105833195531e-05, "loss": 0.0105, "step": 44390 }, { "grad_norm": 0.13840852677822113, "learning_rate": 6.323511414741715e-05, "loss": 0.0108, "step": 44400 }, { "grad_norm": 0.12883159518241882, "learning_rate": 6.321916851550678e-05, "loss": 0.0096, "step": 44410 }, { "grad_norm": 0.16053347289562225, "learning_rate": 6.320322143796806e-05, "loss": 0.0123, "step": 44420 }, { "grad_norm": 0.15476034581661224, "learning_rate": 6.318727291654488e-05, "loss": 0.0106, "step": 44430 }, { "grad_norm": 0.1104801818728447, "learning_rate": 6.317132295298134e-05, "loss": 0.0109, "step": 44440 }, { "grad_norm": 0.17256197333335876, "learning_rate": 6.315537154902173e-05, "loss": 0.011, "step": 44450 }, { "grad_norm": 0.1166987419128418, "learning_rate": 6.313941870641048e-05, "loss": 0.0104, "step": 44460 }, { "grad_norm": 0.14385180175304413, "learning_rate": 6.312346442689214e-05, "loss": 0.0109, "step": 44470 }, { "grad_norm": 0.13448630273342133, "learning_rate": 6.310750871221147e-05, "loss": 0.0103, "step": 44480 }, { "grad_norm": 0.12234631180763245, "learning_rate": 6.309155156411335e-05, "loss": 0.0086, "step": 44490 }, { "grad_norm": 0.12749962508678436, "learning_rate": 6.307559298434284e-05, "loss": 0.0116, "step": 44500 }, { "grad_norm": 0.10708039253950119, "learning_rate": 6.305963297464512e-05, "loss": 0.0098, "step": 44510 }, { "grad_norm": 0.1049051433801651, "learning_rate": 6.304367153676561e-05, "loss": 0.0092, "step": 44520 }, { "grad_norm": 0.1312587708234787, "learning_rate": 6.302770867244978e-05, "loss": 0.0124, "step": 44530 }, { "grad_norm": 0.1553749442100525, "learning_rate": 6.301174438344328e-05, "loss": 0.0111, "step": 44540 }, { "grad_norm": 0.21068033576011658, "learning_rate": 6.299577867149202e-05, "loss": 0.0127, "step": 44550 }, { "grad_norm": 0.13720382750034332, "learning_rate": 6.297981153834192e-05, "loss": 0.0108, "step": 44560 }, { "grad_norm": 0.1496523916721344, "learning_rate": 6.296384298573916e-05, "loss": 0.0136, "step": 44570 }, { "grad_norm": 0.13145014643669128, "learning_rate": 6.294787301543001e-05, "loss": 0.0105, "step": 44580 }, { "grad_norm": 0.1812637895345688, "learning_rate": 6.293190162916095e-05, "loss": 0.0097, "step": 44590 }, { "grad_norm": 0.14589713513851166, "learning_rate": 6.291592882867855e-05, "loss": 0.0096, "step": 44600 }, { "grad_norm": 0.15641754865646362, "learning_rate": 6.28999546157296e-05, "loss": 0.0105, "step": 44610 }, { "grad_norm": 0.2239120453596115, "learning_rate": 6.288397899206102e-05, "loss": 0.0107, "step": 44620 }, { "grad_norm": 0.1332228183746338, "learning_rate": 6.286800195941984e-05, "loss": 0.0108, "step": 44630 }, { "grad_norm": 0.10897782444953918, "learning_rate": 6.285202351955334e-05, "loss": 0.0092, "step": 44640 }, { "grad_norm": 0.1504030078649521, "learning_rate": 6.283604367420887e-05, "loss": 0.0112, "step": 44650 }, { "grad_norm": 0.24229952692985535, "learning_rate": 6.282006242513394e-05, "loss": 0.0098, "step": 44660 }, { "grad_norm": 0.1497141569852829, "learning_rate": 6.280407977407628e-05, "loss": 0.0099, "step": 44670 }, { "grad_norm": 0.17216408252716064, "learning_rate": 6.27880957227837e-05, "loss": 0.014, "step": 44680 }, { "grad_norm": 0.22095762193202972, "learning_rate": 6.27721102730042e-05, "loss": 0.0138, "step": 44690 }, { "grad_norm": 0.12345567345619202, "learning_rate": 6.275612342648592e-05, "loss": 0.0117, "step": 44700 }, { "grad_norm": 0.18795888125896454, "learning_rate": 6.274013518497716e-05, "loss": 0.0101, "step": 44710 }, { "grad_norm": 0.21181000769138336, "learning_rate": 6.272414555022636e-05, "loss": 0.0135, "step": 44720 }, { "grad_norm": 0.1437341272830963, "learning_rate": 6.270815452398215e-05, "loss": 0.0106, "step": 44730 }, { "grad_norm": 0.16385799646377563, "learning_rate": 6.269216210799326e-05, "loss": 0.0104, "step": 44740 }, { "grad_norm": 0.2659047245979309, "learning_rate": 6.26761683040086e-05, "loss": 0.0112, "step": 44750 }, { "grad_norm": 0.20586444437503815, "learning_rate": 6.266017311377723e-05, "loss": 0.0103, "step": 44760 }, { "grad_norm": 0.155567467212677, "learning_rate": 6.264417653904839e-05, "loss": 0.0105, "step": 44770 }, { "grad_norm": 0.20591028034687042, "learning_rate": 6.262817858157139e-05, "loss": 0.0101, "step": 44780 }, { "grad_norm": 0.12689870595932007, "learning_rate": 6.261217924309576e-05, "loss": 0.0087, "step": 44790 }, { "grad_norm": 0.1708378791809082, "learning_rate": 6.259617852537118e-05, "loss": 0.0102, "step": 44800 }, { "grad_norm": 0.16022109985351562, "learning_rate": 6.258017643014747e-05, "loss": 0.0125, "step": 44810 }, { "grad_norm": 0.3385406732559204, "learning_rate": 6.256417295917456e-05, "loss": 0.0113, "step": 44820 }, { "grad_norm": 0.12549051642417908, "learning_rate": 6.254816811420258e-05, "loss": 0.0152, "step": 44830 }, { "grad_norm": 0.17516538500785828, "learning_rate": 6.253216189698183e-05, "loss": 0.0134, "step": 44840 }, { "grad_norm": 0.1414332538843155, "learning_rate": 6.251615430926267e-05, "loss": 0.0125, "step": 44850 }, { "grad_norm": 0.11662400513887405, "learning_rate": 6.25001453527957e-05, "loss": 0.0109, "step": 44860 }, { "grad_norm": 0.15576273202896118, "learning_rate": 6.248413502933164e-05, "loss": 0.0105, "step": 44870 }, { "grad_norm": 0.1318574994802475, "learning_rate": 6.246812334062133e-05, "loss": 0.0107, "step": 44880 }, { "grad_norm": 0.16345052421092987, "learning_rate": 6.245211028841579e-05, "loss": 0.0118, "step": 44890 }, { "grad_norm": 0.14987079799175262, "learning_rate": 6.24360958744662e-05, "loss": 0.0146, "step": 44900 }, { "grad_norm": 0.16174162924289703, "learning_rate": 6.242008010052387e-05, "loss": 0.0122, "step": 44910 }, { "grad_norm": 0.6039591431617737, "learning_rate": 6.240406296834024e-05, "loss": 0.0154, "step": 44920 }, { "grad_norm": 0.2077016830444336, "learning_rate": 6.238804447966694e-05, "loss": 0.0157, "step": 44930 }, { "grad_norm": 0.20575588941574097, "learning_rate": 6.237202463625573e-05, "loss": 0.0162, "step": 44940 }, { "grad_norm": 0.18422755599021912, "learning_rate": 6.235600343985848e-05, "loss": 0.014, "step": 44950 }, { "grad_norm": 0.22549112141132355, "learning_rate": 6.233998089222729e-05, "loss": 0.0117, "step": 44960 }, { "grad_norm": 0.14329412579536438, "learning_rate": 6.232395699511433e-05, "loss": 0.0105, "step": 44970 }, { "grad_norm": 0.1812409907579422, "learning_rate": 6.230793175027199e-05, "loss": 0.0106, "step": 44980 }, { "grad_norm": 0.17454244196414948, "learning_rate": 6.22919051594527e-05, "loss": 0.0123, "step": 44990 }, { "grad_norm": 0.15103186666965485, "learning_rate": 6.227587722440917e-05, "loss": 0.0129, "step": 45000 }, { "grad_norm": 0.11053890734910965, "learning_rate": 6.225984794689414e-05, "loss": 0.0118, "step": 45010 }, { "grad_norm": 0.1683194935321808, "learning_rate": 6.22438173286606e-05, "loss": 0.0117, "step": 45020 }, { "grad_norm": 0.13169905543327332, "learning_rate": 6.222778537146159e-05, "loss": 0.0112, "step": 45030 }, { "grad_norm": 0.13584160804748535, "learning_rate": 6.221175207705037e-05, "loss": 0.0135, "step": 45040 }, { "grad_norm": 0.14389042556285858, "learning_rate": 6.219571744718027e-05, "loss": 0.0113, "step": 45050 }, { "grad_norm": 0.12533430755138397, "learning_rate": 6.217968148360487e-05, "loss": 0.0111, "step": 45060 }, { "grad_norm": 0.2049873322248459, "learning_rate": 6.216364418807782e-05, "loss": 0.0117, "step": 45070 }, { "grad_norm": 0.19519105553627014, "learning_rate": 6.214760556235292e-05, "loss": 0.0105, "step": 45080 }, { "grad_norm": 0.13867473602294922, "learning_rate": 6.213156560818414e-05, "loss": 0.0108, "step": 45090 }, { "grad_norm": 0.15223783254623413, "learning_rate": 6.211552432732559e-05, "loss": 0.0119, "step": 45100 }, { "grad_norm": 0.11894300580024719, "learning_rate": 6.20994817215315e-05, "loss": 0.0121, "step": 45110 }, { "grad_norm": 0.16619248688220978, "learning_rate": 6.208343779255628e-05, "loss": 0.0114, "step": 45120 }, { "grad_norm": 0.15960368514060974, "learning_rate": 6.206739254215449e-05, "loss": 0.0112, "step": 45130 }, { "grad_norm": 0.22870683670043945, "learning_rate": 6.205134597208077e-05, "loss": 0.0124, "step": 45140 }, { "grad_norm": 0.18678824603557587, "learning_rate": 6.203529808409e-05, "loss": 0.0104, "step": 45150 }, { "grad_norm": 0.17202621698379517, "learning_rate": 6.201924887993712e-05, "loss": 0.0104, "step": 45160 }, { "grad_norm": 0.16918426752090454, "learning_rate": 6.200319836137725e-05, "loss": 0.0118, "step": 45170 }, { "grad_norm": 0.2707463800907135, "learning_rate": 6.198714653016565e-05, "loss": 0.013, "step": 45180 }, { "grad_norm": 0.1471213847398758, "learning_rate": 6.197109338805774e-05, "loss": 0.0115, "step": 45190 }, { "grad_norm": 0.14735724031925201, "learning_rate": 6.195503893680903e-05, "loss": 0.0128, "step": 45200 }, { "grad_norm": 0.18242354691028595, "learning_rate": 6.193898317817524e-05, "loss": 0.0119, "step": 45210 }, { "grad_norm": 0.09561514854431152, "learning_rate": 6.192292611391221e-05, "loss": 0.011, "step": 45220 }, { "grad_norm": 0.17236776649951935, "learning_rate": 6.190686774577591e-05, "loss": 0.0099, "step": 45230 }, { "grad_norm": 0.1260072886943817, "learning_rate": 6.189080807552245e-05, "loss": 0.0114, "step": 45240 }, { "grad_norm": 0.13695165514945984, "learning_rate": 6.187474710490809e-05, "loss": 0.013, "step": 45250 }, { "grad_norm": 0.15172363817691803, "learning_rate": 6.185868483568926e-05, "loss": 0.0111, "step": 45260 }, { "grad_norm": 0.15261955559253693, "learning_rate": 6.184262126962245e-05, "loss": 0.0113, "step": 45270 }, { "grad_norm": 0.15735365450382233, "learning_rate": 6.182655640846442e-05, "loss": 0.0107, "step": 45280 }, { "grad_norm": 0.23245790600776672, "learning_rate": 6.181049025397196e-05, "loss": 0.0117, "step": 45290 }, { "grad_norm": 0.1498822122812271, "learning_rate": 6.179442280790202e-05, "loss": 0.0122, "step": 45300 }, { "grad_norm": 0.15232296288013458, "learning_rate": 6.177835407201174e-05, "loss": 0.0134, "step": 45310 }, { "grad_norm": 0.164845272898674, "learning_rate": 6.176228404805839e-05, "loss": 0.0107, "step": 45320 }, { "grad_norm": 0.10587316751480103, "learning_rate": 6.174621273779932e-05, "loss": 0.0102, "step": 45330 }, { "grad_norm": 0.18813203275203705, "learning_rate": 6.17301401429921e-05, "loss": 0.0109, "step": 45340 }, { "grad_norm": 0.22165250778198242, "learning_rate": 6.171406626539436e-05, "loss": 0.0112, "step": 45350 }, { "grad_norm": 0.1597631275653839, "learning_rate": 6.169799110676398e-05, "loss": 0.0092, "step": 45360 }, { "grad_norm": 0.14057093858718872, "learning_rate": 6.168191466885885e-05, "loss": 0.0099, "step": 45370 }, { "grad_norm": 0.1749584674835205, "learning_rate": 6.166583695343711e-05, "loss": 0.0111, "step": 45380 }, { "grad_norm": 0.2732115685939789, "learning_rate": 6.164975796225698e-05, "loss": 0.0104, "step": 45390 }, { "grad_norm": 0.2140560895204544, "learning_rate": 6.163367769707683e-05, "loss": 0.0103, "step": 45400 }, { "grad_norm": 0.1255931258201599, "learning_rate": 6.161759615965519e-05, "loss": 0.01, "step": 45410 }, { "grad_norm": 0.11197636276483536, "learning_rate": 6.16015133517507e-05, "loss": 0.0107, "step": 45420 }, { "grad_norm": 0.22593434154987335, "learning_rate": 6.158542927512214e-05, "loss": 0.0106, "step": 45430 }, { "grad_norm": 0.16756287217140198, "learning_rate": 6.156934393152846e-05, "loss": 0.0102, "step": 45440 }, { "grad_norm": 0.18076983094215393, "learning_rate": 6.15532573227287e-05, "loss": 0.0115, "step": 45450 }, { "grad_norm": 0.1679845154285431, "learning_rate": 6.153716945048212e-05, "loss": 0.0135, "step": 45460 }, { "grad_norm": 0.18990640342235565, "learning_rate": 6.152108031654802e-05, "loss": 0.0109, "step": 45470 }, { "grad_norm": 0.13980263471603394, "learning_rate": 6.15049899226859e-05, "loss": 0.0112, "step": 45480 }, { "grad_norm": 0.15383592247962952, "learning_rate": 6.148889827065537e-05, "loss": 0.0102, "step": 45490 }, { "grad_norm": 0.14357666671276093, "learning_rate": 6.147280536221622e-05, "loss": 0.0098, "step": 45500 }, { "grad_norm": 0.12135182321071625, "learning_rate": 6.145671119912832e-05, "loss": 0.0095, "step": 45510 }, { "grad_norm": 0.14981940388679504, "learning_rate": 6.144061578315169e-05, "loss": 0.0126, "step": 45520 }, { "grad_norm": 0.1434866487979889, "learning_rate": 6.142451911604654e-05, "loss": 0.0103, "step": 45530 }, { "grad_norm": 0.15394967794418335, "learning_rate": 6.140842119957315e-05, "loss": 0.0101, "step": 45540 }, { "grad_norm": 0.11566886305809021, "learning_rate": 6.139232203549197e-05, "loss": 0.0124, "step": 45550 }, { "grad_norm": 0.16265909373760223, "learning_rate": 6.137622162556357e-05, "loss": 0.0103, "step": 45560 }, { "grad_norm": 0.20498812198638916, "learning_rate": 6.13601199715487e-05, "loss": 0.0129, "step": 45570 }, { "grad_norm": 0.13742317259311676, "learning_rate": 6.134401707520816e-05, "loss": 0.0114, "step": 45580 }, { "grad_norm": 0.15221978724002838, "learning_rate": 6.132791293830298e-05, "loss": 0.0105, "step": 45590 }, { "grad_norm": 0.20167632400989532, "learning_rate": 6.131180756259428e-05, "loss": 0.0114, "step": 45600 }, { "grad_norm": 0.15198776125907898, "learning_rate": 6.129570094984331e-05, "loss": 0.0109, "step": 45610 }, { "grad_norm": 0.17252300679683685, "learning_rate": 6.127959310181145e-05, "loss": 0.0095, "step": 45620 }, { "grad_norm": 0.14018948376178741, "learning_rate": 6.126348402026026e-05, "loss": 0.0117, "step": 45630 }, { "grad_norm": 0.16787439584732056, "learning_rate": 6.12473737069514e-05, "loss": 0.0128, "step": 45640 }, { "grad_norm": 0.13611848652362823, "learning_rate": 6.123126216364665e-05, "loss": 0.0102, "step": 45650 }, { "grad_norm": 0.12842786312103271, "learning_rate": 6.121514939210797e-05, "loss": 0.0111, "step": 45660 }, { "grad_norm": 0.13165168464183807, "learning_rate": 6.119903539409741e-05, "loss": 0.0097, "step": 45670 }, { "grad_norm": 0.14555513858795166, "learning_rate": 6.118292017137716e-05, "loss": 0.0113, "step": 45680 }, { "grad_norm": 0.28075867891311646, "learning_rate": 6.116680372570959e-05, "loss": 0.0087, "step": 45690 }, { "grad_norm": 0.1276899129152298, "learning_rate": 6.115068605885713e-05, "loss": 0.0097, "step": 45700 }, { "grad_norm": 0.1242280825972557, "learning_rate": 6.113456717258243e-05, "loss": 0.011, "step": 45710 }, { "grad_norm": 0.14301446080207825, "learning_rate": 6.11184470686482e-05, "loss": 0.0089, "step": 45720 }, { "grad_norm": 0.13083040714263916, "learning_rate": 6.110232574881731e-05, "loss": 0.0093, "step": 45730 }, { "grad_norm": 0.1226641833782196, "learning_rate": 6.108620321485277e-05, "loss": 0.0093, "step": 45740 }, { "grad_norm": 0.2990153729915619, "learning_rate": 6.107007946851773e-05, "loss": 0.0095, "step": 45750 }, { "grad_norm": 0.12129583954811096, "learning_rate": 6.105395451157542e-05, "loss": 0.0135, "step": 45760 }, { "grad_norm": 0.16406302154064178, "learning_rate": 6.103782834578928e-05, "loss": 0.0091, "step": 45770 }, { "grad_norm": 0.2711482048034668, "learning_rate": 6.102170097292281e-05, "loss": 0.0117, "step": 45780 }, { "grad_norm": 0.16112200915813446, "learning_rate": 6.1005572394739716e-05, "loss": 0.0094, "step": 45790 }, { "grad_norm": 0.14377668499946594, "learning_rate": 6.0989442613003765e-05, "loss": 0.0104, "step": 45800 }, { "grad_norm": 0.22588811814785004, "learning_rate": 6.097331162947888e-05, "loss": 0.0128, "step": 45810 }, { "grad_norm": 0.15611614286899567, "learning_rate": 6.095717944592914e-05, "loss": 0.011, "step": 45820 }, { "grad_norm": 0.1360735446214676, "learning_rate": 6.094104606411873e-05, "loss": 0.0095, "step": 45830 }, { "grad_norm": 0.17281238734722137, "learning_rate": 6.0924911485811966e-05, "loss": 0.0109, "step": 45840 }, { "grad_norm": 0.34166714549064636, "learning_rate": 6.090877571277331e-05, "loss": 0.0098, "step": 45850 }, { "grad_norm": 0.17578309774398804, "learning_rate": 6.089263874676736e-05, "loss": 0.0107, "step": 45860 }, { "grad_norm": 0.10023539513349533, "learning_rate": 6.0876500589558796e-05, "loss": 0.0101, "step": 45870 }, { "grad_norm": 0.10306305438280106, "learning_rate": 6.086036124291248e-05, "loss": 0.0095, "step": 45880 }, { "grad_norm": 0.12341251969337463, "learning_rate": 6.084422070859339e-05, "loss": 0.0094, "step": 45890 }, { "grad_norm": 0.16030557453632355, "learning_rate": 6.082807898836663e-05, "loss": 0.0116, "step": 45900 }, { "grad_norm": 0.1488502472639084, "learning_rate": 6.081193608399742e-05, "loss": 0.011, "step": 45910 }, { "grad_norm": 0.1332460641860962, "learning_rate": 6.0795791997251164e-05, "loss": 0.0123, "step": 45920 }, { "grad_norm": 0.14789415895938873, "learning_rate": 6.0779646729893294e-05, "loss": 0.0103, "step": 45930 }, { "grad_norm": 0.12185239046812057, "learning_rate": 6.0763500283689476e-05, "loss": 0.0115, "step": 45940 }, { "grad_norm": 0.13251420855522156, "learning_rate": 6.0747352660405455e-05, "loss": 0.0104, "step": 45950 }, { "grad_norm": 0.15943610668182373, "learning_rate": 6.073120386180709e-05, "loss": 0.0089, "step": 45960 }, { "grad_norm": 0.1355944722890854, "learning_rate": 6.0715053889660425e-05, "loss": 0.0116, "step": 45970 }, { "grad_norm": 0.16694624722003937, "learning_rate": 6.069890274573157e-05, "loss": 0.0126, "step": 45980 }, { "grad_norm": 0.11628682166337967, "learning_rate": 6.068275043178679e-05, "loss": 0.0091, "step": 45990 }, { "grad_norm": 0.1083919107913971, "learning_rate": 6.066659694959248e-05, "loss": 0.0112, "step": 46000 }, { "grad_norm": 0.20818817615509033, "learning_rate": 6.0650442300915176e-05, "loss": 0.011, "step": 46010 }, { "grad_norm": 0.23020923137664795, "learning_rate": 6.063428648752152e-05, "loss": 0.0106, "step": 46020 }, { "grad_norm": 0.1857309341430664, "learning_rate": 6.0618129511178266e-05, "loss": 0.01, "step": 46030 }, { "grad_norm": 0.18756186962127686, "learning_rate": 6.060197137365234e-05, "loss": 0.0131, "step": 46040 }, { "grad_norm": 0.25580310821533203, "learning_rate": 6.058581207671077e-05, "loss": 0.0112, "step": 46050 }, { "grad_norm": 0.1464567929506302, "learning_rate": 6.056965162212072e-05, "loss": 0.0102, "step": 46060 }, { "grad_norm": 0.20915377140045166, "learning_rate": 6.055349001164943e-05, "loss": 0.01, "step": 46070 }, { "grad_norm": 0.12614451348781586, "learning_rate": 6.0537327247064347e-05, "loss": 0.011, "step": 46080 }, { "grad_norm": 0.1620287001132965, "learning_rate": 6.0521163330133014e-05, "loss": 0.0093, "step": 46090 }, { "grad_norm": 0.1806933879852295, "learning_rate": 6.050499826262306e-05, "loss": 0.0093, "step": 46100 }, { "grad_norm": 0.13760939240455627, "learning_rate": 6.0488832046302294e-05, "loss": 0.0111, "step": 46110 }, { "grad_norm": 0.15708474814891815, "learning_rate": 6.0472664682938626e-05, "loss": 0.0102, "step": 46120 }, { "grad_norm": 0.13074105978012085, "learning_rate": 6.045649617430009e-05, "loss": 0.0092, "step": 46130 }, { "grad_norm": 0.1617141216993332, "learning_rate": 6.0440326522154866e-05, "loss": 0.009, "step": 46140 }, { "grad_norm": 0.1534774750471115, "learning_rate": 6.0424155728271224e-05, "loss": 0.0104, "step": 46150 }, { "grad_norm": 0.16909627616405487, "learning_rate": 6.040798379441758e-05, "loss": 0.0119, "step": 46160 }, { "grad_norm": 0.15695933997631073, "learning_rate": 6.0391810722362485e-05, "loss": 0.0127, "step": 46170 }, { "grad_norm": 0.13823895156383514, "learning_rate": 6.037563651387458e-05, "loss": 0.0107, "step": 46180 }, { "grad_norm": 0.13337227702140808, "learning_rate": 6.0359461170722666e-05, "loss": 0.0084, "step": 46190 }, { "grad_norm": 0.14854399859905243, "learning_rate": 6.034328469467566e-05, "loss": 0.01, "step": 46200 }, { "grad_norm": 0.18517127633094788, "learning_rate": 6.0327107087502596e-05, "loss": 0.0091, "step": 46210 }, { "grad_norm": 0.15395401418209076, "learning_rate": 6.031092835097262e-05, "loss": 0.0098, "step": 46220 }, { "grad_norm": 0.13706304132938385, "learning_rate": 6.0294748486855024e-05, "loss": 0.01, "step": 46230 }, { "grad_norm": 0.1381223350763321, "learning_rate": 6.0278567496919216e-05, "loss": 0.0112, "step": 46240 }, { "grad_norm": 0.2000565379858017, "learning_rate": 6.026238538293472e-05, "loss": 0.0091, "step": 46250 }, { "grad_norm": 0.14178603887557983, "learning_rate": 6.024620214667118e-05, "loss": 0.0103, "step": 46260 }, { "grad_norm": 0.12702831625938416, "learning_rate": 6.0230017789898384e-05, "loss": 0.0099, "step": 46270 }, { "grad_norm": 0.12446623295545578, "learning_rate": 6.021383231438622e-05, "loss": 0.0109, "step": 46280 }, { "grad_norm": 0.12518438696861267, "learning_rate": 6.0197645721904704e-05, "loss": 0.0101, "step": 46290 }, { "grad_norm": 0.16838054358959198, "learning_rate": 6.0181458014224e-05, "loss": 0.0125, "step": 46300 }, { "grad_norm": 0.13992154598236084, "learning_rate": 6.016526919311435e-05, "loss": 0.0091, "step": 46310 }, { "grad_norm": 0.16094614565372467, "learning_rate": 6.014907926034613e-05, "loss": 0.0114, "step": 46320 }, { "grad_norm": 0.2071073055267334, "learning_rate": 6.0132888217689875e-05, "loss": 0.0105, "step": 46330 }, { "grad_norm": 0.1936027854681015, "learning_rate": 6.01166960669162e-05, "loss": 0.0103, "step": 46340 }, { "grad_norm": 0.22944732010364532, "learning_rate": 6.0100502809795845e-05, "loss": 0.0086, "step": 46350 }, { "grad_norm": 0.1978740692138672, "learning_rate": 6.00843084480997e-05, "loss": 0.0104, "step": 46360 }, { "grad_norm": 0.20518195629119873, "learning_rate": 6.0068112983598736e-05, "loss": 0.0109, "step": 46370 }, { "grad_norm": 0.10419692099094391, "learning_rate": 6.005191641806407e-05, "loss": 0.0097, "step": 46380 }, { "grad_norm": 0.13300231099128723, "learning_rate": 6.003571875326694e-05, "loss": 0.0104, "step": 46390 }, { "grad_norm": 0.16965708136558533, "learning_rate": 6.00195199909787e-05, "loss": 0.0114, "step": 46400 }, { "grad_norm": 0.17482762038707733, "learning_rate": 6.000332013297082e-05, "loss": 0.0114, "step": 46410 }, { "grad_norm": 0.18662263453006744, "learning_rate": 5.998711918101487e-05, "loss": 0.0117, "step": 46420 }, { "grad_norm": 0.14701437950134277, "learning_rate": 5.997091713688261e-05, "loss": 0.0116, "step": 46430 }, { "grad_norm": 0.12761057913303375, "learning_rate": 5.9954714002345836e-05, "loss": 0.0103, "step": 46440 }, { "grad_norm": 0.19954822957515717, "learning_rate": 5.993850977917649e-05, "loss": 0.012, "step": 46450 }, { "grad_norm": 0.15625609457492828, "learning_rate": 5.992230446914667e-05, "loss": 0.0101, "step": 46460 }, { "grad_norm": 0.15555347502231598, "learning_rate": 5.990609807402855e-05, "loss": 0.0152, "step": 46470 }, { "grad_norm": 0.14837688207626343, "learning_rate": 5.988989059559443e-05, "loss": 0.0161, "step": 46480 }, { "grad_norm": 0.16466298699378967, "learning_rate": 5.987368203561675e-05, "loss": 0.0098, "step": 46490 }, { "grad_norm": 0.1834094226360321, "learning_rate": 5.9857472395868055e-05, "loss": 0.0103, "step": 46500 }, { "grad_norm": 0.13773418962955475, "learning_rate": 5.9841261678120983e-05, "loss": 0.0107, "step": 46510 }, { "grad_norm": 0.19914673268795013, "learning_rate": 5.982504988414834e-05, "loss": 0.0107, "step": 46520 }, { "grad_norm": 0.21216078102588654, "learning_rate": 5.9808837015723015e-05, "loss": 0.0103, "step": 46530 }, { "grad_norm": 0.14308704435825348, "learning_rate": 5.9792623074618016e-05, "loss": 0.0146, "step": 46540 }, { "grad_norm": 0.1524103581905365, "learning_rate": 5.977640806260648e-05, "loss": 0.0105, "step": 46550 }, { "grad_norm": 0.1586969941854477, "learning_rate": 5.976019198146166e-05, "loss": 0.0111, "step": 46560 }, { "grad_norm": 0.16363617777824402, "learning_rate": 5.974397483295692e-05, "loss": 0.0101, "step": 46570 }, { "grad_norm": 0.13921143114566803, "learning_rate": 5.972775661886575e-05, "loss": 0.0105, "step": 46580 }, { "grad_norm": 0.11545281857252121, "learning_rate": 5.971153734096173e-05, "loss": 0.01, "step": 46590 }, { "grad_norm": 0.15260645747184753, "learning_rate": 5.969531700101859e-05, "loss": 0.0115, "step": 46600 }, { "grad_norm": 0.14329423010349274, "learning_rate": 5.9679095600810155e-05, "loss": 0.0105, "step": 46610 }, { "grad_norm": 0.15740713477134705, "learning_rate": 5.9662873142110384e-05, "loss": 0.0122, "step": 46620 }, { "grad_norm": 0.143373042345047, "learning_rate": 5.964664962669333e-05, "loss": 0.0094, "step": 46630 }, { "grad_norm": 0.15440957248210907, "learning_rate": 5.9630425056333186e-05, "loss": 0.0112, "step": 46640 }, { "grad_norm": 0.1710168570280075, "learning_rate": 5.961419943280422e-05, "loss": 0.0118, "step": 46650 }, { "grad_norm": 0.17965522408485413, "learning_rate": 5.959797275788087e-05, "loss": 0.0099, "step": 46660 }, { "grad_norm": 0.11558445543050766, "learning_rate": 5.958174503333765e-05, "loss": 0.0092, "step": 46670 }, { "grad_norm": 0.150421142578125, "learning_rate": 5.9565516260949195e-05, "loss": 0.0119, "step": 46680 }, { "grad_norm": 0.12644803524017334, "learning_rate": 5.9549286442490273e-05, "loss": 0.0094, "step": 46690 }, { "grad_norm": 0.16711904108524323, "learning_rate": 5.953305557973572e-05, "loss": 0.0094, "step": 46700 }, { "grad_norm": 0.14502546191215515, "learning_rate": 5.9516823674460564e-05, "loss": 0.0083, "step": 46710 }, { "grad_norm": 0.18404290080070496, "learning_rate": 5.9500590728439875e-05, "loss": 0.0109, "step": 46720 }, { "grad_norm": 0.2489922046661377, "learning_rate": 5.948435674344886e-05, "loss": 0.012, "step": 46730 }, { "grad_norm": 0.14274518191814423, "learning_rate": 5.946812172126285e-05, "loss": 0.0088, "step": 46740 }, { "grad_norm": 0.20280230045318604, "learning_rate": 5.945188566365729e-05, "loss": 0.0119, "step": 46750 }, { "grad_norm": 0.1522025167942047, "learning_rate": 5.943564857240773e-05, "loss": 0.0108, "step": 46760 }, { "grad_norm": 0.1503000259399414, "learning_rate": 5.941941044928981e-05, "loss": 0.0095, "step": 46770 }, { "grad_norm": 0.1509857326745987, "learning_rate": 5.940317129607935e-05, "loss": 0.0138, "step": 46780 }, { "grad_norm": 0.19886133074760437, "learning_rate": 5.9386931114552204e-05, "loss": 0.0092, "step": 46790 }, { "grad_norm": 0.16538013517856598, "learning_rate": 5.937068990648438e-05, "loss": 0.0103, "step": 46800 }, { "grad_norm": 0.16493695974349976, "learning_rate": 5.935444767365199e-05, "loss": 0.011, "step": 46810 }, { "grad_norm": 0.10382591187953949, "learning_rate": 5.933820441783129e-05, "loss": 0.0111, "step": 46820 }, { "grad_norm": 0.15774239599704742, "learning_rate": 5.932196014079857e-05, "loss": 0.0102, "step": 46830 }, { "grad_norm": 0.24624881148338318, "learning_rate": 5.930571484433032e-05, "loss": 0.0105, "step": 46840 }, { "grad_norm": 0.15565699338912964, "learning_rate": 5.928946853020309e-05, "loss": 0.01, "step": 46850 }, { "grad_norm": 0.17165417969226837, "learning_rate": 5.927322120019352e-05, "loss": 0.0095, "step": 46860 }, { "grad_norm": 0.15946204960346222, "learning_rate": 5.925697285607844e-05, "loss": 0.01, "step": 46870 }, { "grad_norm": 0.13391156494617462, "learning_rate": 5.924072349963472e-05, "loss": 0.0114, "step": 46880 }, { "grad_norm": 0.15027640759944916, "learning_rate": 5.922447313263938e-05, "loss": 0.0098, "step": 46890 }, { "grad_norm": 0.20952211320400238, "learning_rate": 5.920822175686952e-05, "loss": 0.0125, "step": 46900 }, { "grad_norm": 0.14093433320522308, "learning_rate": 5.919196937410239e-05, "loss": 0.0114, "step": 46910 }, { "grad_norm": 0.19142359495162964, "learning_rate": 5.917571598611529e-05, "loss": 0.0124, "step": 46920 }, { "grad_norm": 0.17362487316131592, "learning_rate": 5.9159461594685686e-05, "loss": 0.0107, "step": 46930 }, { "grad_norm": 0.23311461508274078, "learning_rate": 5.914320620159114e-05, "loss": 0.0117, "step": 46940 }, { "grad_norm": 0.1376788318157196, "learning_rate": 5.912694980860931e-05, "loss": 0.0113, "step": 46950 }, { "grad_norm": 0.16446378827095032, "learning_rate": 5.9110692417517964e-05, "loss": 0.01, "step": 46960 }, { "grad_norm": 0.16302280128002167, "learning_rate": 5.9094434030095e-05, "loss": 0.0115, "step": 46970 }, { "grad_norm": 0.1383613795042038, "learning_rate": 5.907817464811841e-05, "loss": 0.0112, "step": 46980 }, { "grad_norm": 0.14795707166194916, "learning_rate": 5.906191427336627e-05, "loss": 0.0119, "step": 46990 }, { "grad_norm": 0.14450310170650482, "learning_rate": 5.9045652907616835e-05, "loss": 0.0103, "step": 47000 }, { "grad_norm": 0.11475377529859543, "learning_rate": 5.902939055264838e-05, "loss": 0.0106, "step": 47010 }, { "grad_norm": 0.14264565706253052, "learning_rate": 5.901312721023935e-05, "loss": 0.0112, "step": 47020 }, { "grad_norm": 0.16432934999465942, "learning_rate": 5.8996862882168294e-05, "loss": 0.0109, "step": 47030 }, { "grad_norm": 0.18155069649219513, "learning_rate": 5.8980597570213826e-05, "loss": 0.0086, "step": 47040 }, { "grad_norm": 0.21251508593559265, "learning_rate": 5.896433127615471e-05, "loss": 0.0094, "step": 47050 }, { "grad_norm": 0.1626366525888443, "learning_rate": 5.894806400176981e-05, "loss": 0.0102, "step": 47060 }, { "grad_norm": 0.13239435851573944, "learning_rate": 5.893179574883808e-05, "loss": 0.0105, "step": 47070 }, { "grad_norm": 0.18561813235282898, "learning_rate": 5.8915526519138585e-05, "loss": 0.0131, "step": 47080 }, { "grad_norm": 0.17001546919345856, "learning_rate": 5.889925631445053e-05, "loss": 0.0105, "step": 47090 }, { "grad_norm": 0.1424221396446228, "learning_rate": 5.888298513655318e-05, "loss": 0.0093, "step": 47100 }, { "grad_norm": 0.23693330585956573, "learning_rate": 5.886671298722592e-05, "loss": 0.0102, "step": 47110 }, { "grad_norm": 0.17628033459186554, "learning_rate": 5.885043986824828e-05, "loss": 0.0115, "step": 47120 }, { "grad_norm": 0.15855692327022552, "learning_rate": 5.883416578139982e-05, "loss": 0.0095, "step": 47130 }, { "grad_norm": 0.1733686923980713, "learning_rate": 5.881789072846029e-05, "loss": 0.0109, "step": 47140 }, { "grad_norm": 0.14521433413028717, "learning_rate": 5.8801614711209474e-05, "loss": 0.0107, "step": 47150 }, { "grad_norm": 0.1069309189915657, "learning_rate": 5.878533773142732e-05, "loss": 0.009, "step": 47160 }, { "grad_norm": 0.14342853426933289, "learning_rate": 5.8769059790893834e-05, "loss": 0.0118, "step": 47170 }, { "grad_norm": 0.1458330750465393, "learning_rate": 5.8752780891389145e-05, "loss": 0.01, "step": 47180 }, { "grad_norm": 0.1752462387084961, "learning_rate": 5.873650103469351e-05, "loss": 0.0095, "step": 47190 }, { "grad_norm": 0.14802688360214233, "learning_rate": 5.872022022258725e-05, "loss": 0.0091, "step": 47200 }, { "grad_norm": 0.13367536664009094, "learning_rate": 5.8703938456850814e-05, "loss": 0.0102, "step": 47210 }, { "grad_norm": 0.11703260987997055, "learning_rate": 5.868765573926477e-05, "loss": 0.0086, "step": 47220 }, { "grad_norm": 0.14541368186473846, "learning_rate": 5.867137207160974e-05, "loss": 0.0107, "step": 47230 }, { "grad_norm": 0.14227750897407532, "learning_rate": 5.86550874556665e-05, "loss": 0.0096, "step": 47240 }, { "grad_norm": 0.22189649939537048, "learning_rate": 5.863880189321592e-05, "loss": 0.0108, "step": 47250 }, { "grad_norm": 0.13689692318439484, "learning_rate": 5.862251538603894e-05, "loss": 0.0098, "step": 47260 }, { "grad_norm": 0.12478601187467575, "learning_rate": 5.8606227935916656e-05, "loss": 0.0118, "step": 47270 }, { "grad_norm": 0.1483628898859024, "learning_rate": 5.858993954463021e-05, "loss": 0.0092, "step": 47280 }, { "grad_norm": 0.31914985179901123, "learning_rate": 5.8573650213960896e-05, "loss": 0.0101, "step": 47290 }, { "grad_norm": 0.1723087579011917, "learning_rate": 5.8557359945690084e-05, "loss": 0.0095, "step": 47300 }, { "grad_norm": 0.14097988605499268, "learning_rate": 5.8541068741599246e-05, "loss": 0.0089, "step": 47310 }, { "grad_norm": 0.1515571027994156, "learning_rate": 5.8524776603469985e-05, "loss": 0.0114, "step": 47320 }, { "grad_norm": 0.11951670795679092, "learning_rate": 5.850848353308397e-05, "loss": 0.011, "step": 47330 }, { "grad_norm": 0.07523427903652191, "learning_rate": 5.849218953222297e-05, "loss": 0.0095, "step": 47340 }, { "grad_norm": 0.33364900946617126, "learning_rate": 5.847589460266891e-05, "loss": 0.0101, "step": 47350 }, { "grad_norm": 0.14475977420806885, "learning_rate": 5.845959874620375e-05, "loss": 0.0097, "step": 47360 }, { "grad_norm": 0.1903279572725296, "learning_rate": 5.84433019646096e-05, "loss": 0.0117, "step": 47370 }, { "grad_norm": 0.12359727174043655, "learning_rate": 5.842700425966863e-05, "loss": 0.0112, "step": 47380 }, { "grad_norm": 0.1783244013786316, "learning_rate": 5.841070563316315e-05, "loss": 0.0117, "step": 47390 }, { "grad_norm": 0.160554438829422, "learning_rate": 5.839440608687554e-05, "loss": 0.0101, "step": 47400 }, { "grad_norm": 0.18868021667003632, "learning_rate": 5.837810562258831e-05, "loss": 0.0109, "step": 47410 }, { "grad_norm": 0.14717242121696472, "learning_rate": 5.836180424208405e-05, "loss": 0.0094, "step": 47420 }, { "grad_norm": 0.15308889746665955, "learning_rate": 5.834550194714543e-05, "loss": 0.0116, "step": 47430 }, { "grad_norm": 0.11324921250343323, "learning_rate": 5.832919873955528e-05, "loss": 0.0116, "step": 47440 }, { "grad_norm": 0.1561630368232727, "learning_rate": 5.831289462109648e-05, "loss": 0.0097, "step": 47450 }, { "grad_norm": 0.1990983486175537, "learning_rate": 5.829658959355202e-05, "loss": 0.0095, "step": 47460 }, { "grad_norm": 0.23585258424282074, "learning_rate": 5.8280283658704995e-05, "loss": 0.0107, "step": 47470 }, { "grad_norm": 0.1877458542585373, "learning_rate": 5.826397681833859e-05, "loss": 0.0107, "step": 47480 }, { "grad_norm": 0.14022068679332733, "learning_rate": 5.824766907423612e-05, "loss": 0.0128, "step": 47490 }, { "grad_norm": 0.15572737157344818, "learning_rate": 5.8231360428180945e-05, "loss": 0.0107, "step": 47500 }, { "grad_norm": 0.1561592072248459, "learning_rate": 5.821505088195658e-05, "loss": 0.012, "step": 47510 }, { "grad_norm": 0.15846645832061768, "learning_rate": 5.819874043734661e-05, "loss": 0.0109, "step": 47520 }, { "grad_norm": 0.1360045224428177, "learning_rate": 5.8182429096134695e-05, "loss": 0.0109, "step": 47530 }, { "grad_norm": 0.1403227597475052, "learning_rate": 5.816611686010465e-05, "loss": 0.0093, "step": 47540 }, { "grad_norm": 0.15106357634067535, "learning_rate": 5.814980373104033e-05, "loss": 0.0129, "step": 47550 }, { "grad_norm": 0.16835182905197144, "learning_rate": 5.813348971072572e-05, "loss": 0.0099, "step": 47560 }, { "grad_norm": 0.16414466500282288, "learning_rate": 5.811717480094492e-05, "loss": 0.0095, "step": 47570 }, { "grad_norm": 0.15754126012325287, "learning_rate": 5.810085900348209e-05, "loss": 0.0135, "step": 47580 }, { "grad_norm": 0.19609013199806213, "learning_rate": 5.8084542320121483e-05, "loss": 0.0109, "step": 47590 }, { "grad_norm": 0.17863205075263977, "learning_rate": 5.8068224752647497e-05, "loss": 0.0109, "step": 47600 }, { "grad_norm": 0.15940919518470764, "learning_rate": 5.805190630284457e-05, "loss": 0.0099, "step": 47610 }, { "grad_norm": 0.12691684067249298, "learning_rate": 5.803558697249729e-05, "loss": 0.01, "step": 47620 }, { "grad_norm": 0.141611710190773, "learning_rate": 5.8019266763390276e-05, "loss": 0.0121, "step": 47630 }, { "grad_norm": 0.2511221170425415, "learning_rate": 5.800294567730833e-05, "loss": 0.0103, "step": 47640 }, { "grad_norm": 0.10862916707992554, "learning_rate": 5.798662371603628e-05, "loss": 0.0095, "step": 47650 }, { "grad_norm": 0.14211086928844452, "learning_rate": 5.797030088135904e-05, "loss": 0.0091, "step": 47660 }, { "grad_norm": 0.15488971769809723, "learning_rate": 5.7953977175061705e-05, "loss": 0.01, "step": 47670 }, { "grad_norm": 0.12575629353523254, "learning_rate": 5.793765259892938e-05, "loss": 0.0108, "step": 47680 }, { "grad_norm": 0.14089705049991608, "learning_rate": 5.792132715474729e-05, "loss": 0.0104, "step": 47690 }, { "grad_norm": 0.19569262862205505, "learning_rate": 5.790500084430078e-05, "loss": 0.0098, "step": 47700 }, { "grad_norm": 0.1787155717611313, "learning_rate": 5.7888673669375274e-05, "loss": 0.0087, "step": 47710 }, { "grad_norm": 0.1427692323923111, "learning_rate": 5.787234563175625e-05, "loss": 0.0101, "step": 47720 }, { "grad_norm": 0.1526680886745453, "learning_rate": 5.7856016733229355e-05, "loss": 0.0108, "step": 47730 }, { "grad_norm": 0.13427706062793732, "learning_rate": 5.7839686975580297e-05, "loss": 0.0089, "step": 47740 }, { "grad_norm": 0.12073611468076706, "learning_rate": 5.782335636059484e-05, "loss": 0.0088, "step": 47750 }, { "grad_norm": 0.17112717032432556, "learning_rate": 5.780702489005889e-05, "loss": 0.0092, "step": 47760 }, { "grad_norm": 0.17460529506206512, "learning_rate": 5.779069256575845e-05, "loss": 0.0105, "step": 47770 }, { "grad_norm": 0.15745124220848083, "learning_rate": 5.7774359389479574e-05, "loss": 0.0092, "step": 47780 }, { "grad_norm": 0.22370894253253937, "learning_rate": 5.775802536300845e-05, "loss": 0.0101, "step": 47790 }, { "grad_norm": 0.11625129729509354, "learning_rate": 5.774169048813134e-05, "loss": 0.0095, "step": 47800 }, { "grad_norm": 0.13932958245277405, "learning_rate": 5.77253547666346e-05, "loss": 0.0103, "step": 47810 }, { "grad_norm": 0.1447271704673767, "learning_rate": 5.770901820030465e-05, "loss": 0.0112, "step": 47820 }, { "grad_norm": 0.17356756329536438, "learning_rate": 5.769268079092809e-05, "loss": 0.0093, "step": 47830 }, { "grad_norm": 0.12854979932308197, "learning_rate": 5.767634254029151e-05, "loss": 0.0094, "step": 47840 }, { "grad_norm": 0.14588245749473572, "learning_rate": 5.7660003450181655e-05, "loss": 0.0098, "step": 47850 }, { "grad_norm": 0.14073455333709717, "learning_rate": 5.764366352238534e-05, "loss": 0.0123, "step": 47860 }, { "grad_norm": 0.10494106262922287, "learning_rate": 5.7627322758689474e-05, "loss": 0.0116, "step": 47870 }, { "grad_norm": 0.12752880156040192, "learning_rate": 5.761098116088105e-05, "loss": 0.0103, "step": 47880 }, { "grad_norm": 0.15496574342250824, "learning_rate": 5.759463873074717e-05, "loss": 0.013, "step": 47890 }, { "grad_norm": 0.2274402379989624, "learning_rate": 5.757829547007504e-05, "loss": 0.0133, "step": 47900 }, { "grad_norm": 0.1547900140285492, "learning_rate": 5.756195138065189e-05, "loss": 0.0116, "step": 47910 }, { "grad_norm": 0.21262338757514954, "learning_rate": 5.754560646426511e-05, "loss": 0.011, "step": 47920 }, { "grad_norm": 0.1526826024055481, "learning_rate": 5.752926072270216e-05, "loss": 0.0113, "step": 47930 }, { "grad_norm": 0.15230482816696167, "learning_rate": 5.7512914157750563e-05, "loss": 0.0118, "step": 47940 }, { "grad_norm": 0.1275794804096222, "learning_rate": 5.749656677119798e-05, "loss": 0.0106, "step": 47950 }, { "grad_norm": 0.19639159739017487, "learning_rate": 5.7480218564832125e-05, "loss": 0.0103, "step": 47960 }, { "grad_norm": 0.11991732567548752, "learning_rate": 5.746386954044082e-05, "loss": 0.0123, "step": 47970 }, { "grad_norm": 0.12684009969234467, "learning_rate": 5.744751969981195e-05, "loss": 0.0103, "step": 47980 }, { "grad_norm": 0.13561181724071503, "learning_rate": 5.7431169044733526e-05, "loss": 0.0091, "step": 47990 }, { "grad_norm": 0.15259391069412231, "learning_rate": 5.741481757699364e-05, "loss": 0.0096, "step": 48000 }, { "grad_norm": 0.20717093348503113, "learning_rate": 5.7398465298380434e-05, "loss": 0.0101, "step": 48010 }, { "grad_norm": 0.15287214517593384, "learning_rate": 5.7382112210682193e-05, "loss": 0.0096, "step": 48020 }, { "grad_norm": 0.08599082380533218, "learning_rate": 5.7365758315687266e-05, "loss": 0.0082, "step": 48030 }, { "grad_norm": 0.20130079984664917, "learning_rate": 5.734940361518407e-05, "loss": 0.0099, "step": 48040 }, { "grad_norm": 0.1515551507472992, "learning_rate": 5.733304811096116e-05, "loss": 0.0074, "step": 48050 }, { "grad_norm": 0.15666021406650543, "learning_rate": 5.731669180480713e-05, "loss": 0.0119, "step": 48060 }, { "grad_norm": 0.12471365928649902, "learning_rate": 5.730033469851067e-05, "loss": 0.009, "step": 48070 }, { "grad_norm": 0.14889025688171387, "learning_rate": 5.7283976793860607e-05, "loss": 0.0092, "step": 48080 }, { "grad_norm": 0.16658495366573334, "learning_rate": 5.726761809264577e-05, "loss": 0.0086, "step": 48090 }, { "grad_norm": 0.14664743840694427, "learning_rate": 5.7251258596655155e-05, "loss": 0.0111, "step": 48100 }, { "grad_norm": 0.14432112872600555, "learning_rate": 5.72348983076778e-05, "loss": 0.0111, "step": 48110 }, { "grad_norm": 0.1621984988451004, "learning_rate": 5.7218537227502854e-05, "loss": 0.0104, "step": 48120 }, { "grad_norm": 0.33552098274230957, "learning_rate": 5.720217535791951e-05, "loss": 0.0084, "step": 48130 }, { "grad_norm": 0.13513459265232086, "learning_rate": 5.718581270071711e-05, "loss": 0.0101, "step": 48140 }, { "grad_norm": 0.1714777648448944, "learning_rate": 5.716944925768505e-05, "loss": 0.0122, "step": 48150 }, { "grad_norm": 0.11067783087491989, "learning_rate": 5.7153085030612786e-05, "loss": 0.0089, "step": 48160 }, { "grad_norm": 0.1591605842113495, "learning_rate": 5.713672002128989e-05, "loss": 0.0096, "step": 48170 }, { "grad_norm": 0.15269935131072998, "learning_rate": 5.712035423150602e-05, "loss": 0.0104, "step": 48180 }, { "grad_norm": 0.11407250165939331, "learning_rate": 5.710398766305094e-05, "loss": 0.0101, "step": 48190 }, { "grad_norm": 0.15030699968338013, "learning_rate": 5.708762031771442e-05, "loss": 0.0079, "step": 48200 }, { "grad_norm": 0.2073933184146881, "learning_rate": 5.707125219728642e-05, "loss": 0.0106, "step": 48210 }, { "grad_norm": 0.2272820621728897, "learning_rate": 5.7054883303556905e-05, "loss": 0.0091, "step": 48220 }, { "grad_norm": 0.13554821908473969, "learning_rate": 5.703851363831595e-05, "loss": 0.0096, "step": 48230 }, { "grad_norm": 0.1750679612159729, "learning_rate": 5.702214320335374e-05, "loss": 0.0094, "step": 48240 }, { "grad_norm": 0.15411750972270966, "learning_rate": 5.7005772000460514e-05, "loss": 0.0108, "step": 48250 }, { "grad_norm": 0.12500117719173431, "learning_rate": 5.698940003142656e-05, "loss": 0.0098, "step": 48260 }, { "grad_norm": 0.14482159912586212, "learning_rate": 5.697302729804236e-05, "loss": 0.0108, "step": 48270 }, { "grad_norm": 0.13968393206596375, "learning_rate": 5.695665380209837e-05, "loss": 0.009, "step": 48280 }, { "grad_norm": 0.17135073244571686, "learning_rate": 5.6940279545385165e-05, "loss": 0.0107, "step": 48290 }, { "grad_norm": 0.21812516450881958, "learning_rate": 5.692390452969344e-05, "loss": 0.0105, "step": 48300 }, { "grad_norm": 0.1099877804517746, "learning_rate": 5.690752875681392e-05, "loss": 0.0104, "step": 48310 }, { "grad_norm": 0.14924459159374237, "learning_rate": 5.6891152228537435e-05, "loss": 0.0106, "step": 48320 }, { "grad_norm": 0.11799876391887665, "learning_rate": 5.687477494665492e-05, "loss": 0.0104, "step": 48330 }, { "grad_norm": 0.1148175597190857, "learning_rate": 5.685839691295733e-05, "loss": 0.0093, "step": 48340 }, { "grad_norm": 0.15427984297275543, "learning_rate": 5.6842018129235786e-05, "loss": 0.01, "step": 48350 }, { "grad_norm": 0.15520548820495605, "learning_rate": 5.6825638597281404e-05, "loss": 0.0102, "step": 48360 }, { "grad_norm": 0.14776912331581116, "learning_rate": 5.680925831888546e-05, "loss": 0.0106, "step": 48370 }, { "grad_norm": 0.12725067138671875, "learning_rate": 5.6792877295839274e-05, "loss": 0.0095, "step": 48380 }, { "grad_norm": 0.14388351142406464, "learning_rate": 5.6776495529934224e-05, "loss": 0.0094, "step": 48390 }, { "grad_norm": 0.14912059903144836, "learning_rate": 5.6760113022961824e-05, "loss": 0.0088, "step": 48400 }, { "grad_norm": 0.20279015600681305, "learning_rate": 5.6743729776713617e-05, "loss": 0.0108, "step": 48410 }, { "grad_norm": 0.12753282487392426, "learning_rate": 5.672734579298126e-05, "loss": 0.0092, "step": 48420 }, { "grad_norm": 0.17950958013534546, "learning_rate": 5.671096107355649e-05, "loss": 0.0113, "step": 48430 }, { "grad_norm": 0.1582447588443756, "learning_rate": 5.66945756202311e-05, "loss": 0.0108, "step": 48440 }, { "grad_norm": 0.14194244146347046, "learning_rate": 5.667818943479699e-05, "loss": 0.0117, "step": 48450 }, { "grad_norm": 0.1009523868560791, "learning_rate": 5.666180251904612e-05, "loss": 0.008, "step": 48460 }, { "grad_norm": 0.19486865401268005, "learning_rate": 5.6645414874770555e-05, "loss": 0.0092, "step": 48470 }, { "grad_norm": 0.13310027122497559, "learning_rate": 5.66290265037624e-05, "loss": 0.0096, "step": 48480 }, { "grad_norm": 0.09497922658920288, "learning_rate": 5.661263740781386e-05, "loss": 0.0088, "step": 48490 }, { "grad_norm": 0.1484462320804596, "learning_rate": 5.6596247588717254e-05, "loss": 0.01, "step": 48500 }, { "grad_norm": 0.13964909315109253, "learning_rate": 5.6579857048264926e-05, "loss": 0.0099, "step": 48510 }, { "grad_norm": 0.1748417615890503, "learning_rate": 5.6563465788249314e-05, "loss": 0.0102, "step": 48520 }, { "grad_norm": 0.12639378011226654, "learning_rate": 5.6547073810462956e-05, "loss": 0.0102, "step": 48530 }, { "grad_norm": 0.1326463371515274, "learning_rate": 5.653068111669846e-05, "loss": 0.0083, "step": 48540 }, { "grad_norm": 0.1341038942337036, "learning_rate": 5.651428770874848e-05, "loss": 0.0122, "step": 48550 }, { "grad_norm": 0.18475352227687836, "learning_rate": 5.64978935884058e-05, "loss": 0.012, "step": 48560 }, { "grad_norm": 0.12967447936534882, "learning_rate": 5.6481498757463244e-05, "loss": 0.0098, "step": 48570 }, { "grad_norm": 0.1542513370513916, "learning_rate": 5.646510321771373e-05, "loss": 0.0103, "step": 48580 }, { "grad_norm": 0.24427512288093567, "learning_rate": 5.644870697095024e-05, "loss": 0.0085, "step": 48590 }, { "grad_norm": 0.17167161405086517, "learning_rate": 5.643231001896586e-05, "loss": 0.0109, "step": 48600 }, { "grad_norm": 0.2274307757616043, "learning_rate": 5.6415912363553726e-05, "loss": 0.0104, "step": 48610 }, { "grad_norm": 0.16390861570835114, "learning_rate": 5.639951400650706e-05, "loss": 0.0128, "step": 48620 }, { "grad_norm": 0.1794913113117218, "learning_rate": 5.6383114949619165e-05, "loss": 0.0107, "step": 48630 }, { "grad_norm": 0.17789745330810547, "learning_rate": 5.636671519468342e-05, "loss": 0.0108, "step": 48640 }, { "grad_norm": 0.15075109899044037, "learning_rate": 5.635031474349327e-05, "loss": 0.0117, "step": 48650 }, { "grad_norm": 0.16090360283851624, "learning_rate": 5.6333913597842246e-05, "loss": 0.0103, "step": 48660 }, { "grad_norm": 0.15708822011947632, "learning_rate": 5.6317511759523955e-05, "loss": 0.0092, "step": 48670 }, { "grad_norm": 0.11619319021701813, "learning_rate": 5.630110923033207e-05, "loss": 0.0095, "step": 48680 }, { "grad_norm": 0.19488483667373657, "learning_rate": 5.628470601206036e-05, "loss": 0.0101, "step": 48690 }, { "grad_norm": 0.15755720436573029, "learning_rate": 5.626830210650266e-05, "loss": 0.0111, "step": 48700 }, { "grad_norm": 0.1338784247636795, "learning_rate": 5.625189751545285e-05, "loss": 0.0102, "step": 48710 }, { "grad_norm": 0.16745226085186005, "learning_rate": 5.6235492240704936e-05, "loss": 0.0094, "step": 48720 }, { "grad_norm": 0.13165031373500824, "learning_rate": 5.621908628405296e-05, "loss": 0.0104, "step": 48730 }, { "grad_norm": 0.1263481229543686, "learning_rate": 5.620267964729106e-05, "loss": 0.0084, "step": 48740 }, { "grad_norm": 0.18515640497207642, "learning_rate": 5.618627233221344e-05, "loss": 0.0104, "step": 48750 }, { "grad_norm": 0.17282533645629883, "learning_rate": 5.616986434061438e-05, "loss": 0.0108, "step": 48760 }, { "grad_norm": 0.1517357975244522, "learning_rate": 5.615345567428822e-05, "loss": 0.0094, "step": 48770 }, { "grad_norm": 0.1627376526594162, "learning_rate": 5.613704633502941e-05, "loss": 0.0109, "step": 48780 }, { "grad_norm": 0.13066181540489197, "learning_rate": 5.612063632463245e-05, "loss": 0.0115, "step": 48790 }, { "grad_norm": 0.15644127130508423, "learning_rate": 5.610422564489188e-05, "loss": 0.0125, "step": 48800 }, { "grad_norm": 0.22968627512454987, "learning_rate": 5.608781429760239e-05, "loss": 0.0103, "step": 48810 }, { "grad_norm": 0.12930089235305786, "learning_rate": 5.607140228455866e-05, "loss": 0.0102, "step": 48820 }, { "grad_norm": 0.11334927380084991, "learning_rate": 5.605498960755553e-05, "loss": 0.0102, "step": 48830 }, { "grad_norm": 0.14475278556346893, "learning_rate": 5.603857626838782e-05, "loss": 0.0098, "step": 48840 }, { "grad_norm": 0.11932909488677979, "learning_rate": 5.60221622688505e-05, "loss": 0.0093, "step": 48850 }, { "grad_norm": 0.17966313660144806, "learning_rate": 5.6005747610738565e-05, "loss": 0.0094, "step": 48860 }, { "grad_norm": 0.09883391112089157, "learning_rate": 5.59893322958471e-05, "loss": 0.0075, "step": 48870 }, { "grad_norm": 0.19630302488803864, "learning_rate": 5.5972916325971256e-05, "loss": 0.0112, "step": 48880 }, { "grad_norm": 0.15789766609668732, "learning_rate": 5.595649970290628e-05, "loss": 0.0091, "step": 48890 }, { "grad_norm": 0.14409923553466797, "learning_rate": 5.5940082428447426e-05, "loss": 0.0094, "step": 48900 }, { "grad_norm": 0.12761607766151428, "learning_rate": 5.592366450439012e-05, "loss": 0.0079, "step": 48910 }, { "grad_norm": 0.16457541286945343, "learning_rate": 5.590724593252975e-05, "loss": 0.0087, "step": 48920 }, { "grad_norm": 0.13288424909114838, "learning_rate": 5.589082671466184e-05, "loss": 0.0107, "step": 48930 }, { "grad_norm": 0.1493820995092392, "learning_rate": 5.587440685258199e-05, "loss": 0.0092, "step": 48940 }, { "grad_norm": 0.16718770563602448, "learning_rate": 5.585798634808583e-05, "loss": 0.0088, "step": 48950 }, { "grad_norm": 0.12249676883220673, "learning_rate": 5.584156520296909e-05, "loss": 0.0103, "step": 48960 }, { "grad_norm": 0.15514837205410004, "learning_rate": 5.582514341902757e-05, "loss": 0.008, "step": 48970 }, { "grad_norm": 0.16013683378696442, "learning_rate": 5.580872099805713e-05, "loss": 0.0121, "step": 48980 }, { "grad_norm": 0.15555258095264435, "learning_rate": 5.5792297941853674e-05, "loss": 0.01, "step": 48990 }, { "grad_norm": 0.20800627768039703, "learning_rate": 5.5775874252213247e-05, "loss": 0.0092, "step": 49000 }, { "grad_norm": 0.11794089525938034, "learning_rate": 5.575944993093189e-05, "loss": 0.0111, "step": 49010 }, { "grad_norm": 0.16175366938114166, "learning_rate": 5.574302497980574e-05, "loss": 0.0144, "step": 49020 }, { "grad_norm": 0.16438566148281097, "learning_rate": 5.5726599400631005e-05, "loss": 0.0104, "step": 49030 }, { "grad_norm": 0.14625754952430725, "learning_rate": 5.571017319520399e-05, "loss": 0.0093, "step": 49040 }, { "grad_norm": 0.14082296192646027, "learning_rate": 5.5693746365320984e-05, "loss": 0.0108, "step": 49050 }, { "grad_norm": 0.13902463018894196, "learning_rate": 5.567731891277848e-05, "loss": 0.0116, "step": 49060 }, { "grad_norm": 0.13944323360919952, "learning_rate": 5.566089083937288e-05, "loss": 0.0104, "step": 49070 }, { "grad_norm": 0.14800244569778442, "learning_rate": 5.564446214690079e-05, "loss": 0.01, "step": 49080 }, { "grad_norm": 0.16924786567687988, "learning_rate": 5.56280328371588e-05, "loss": 0.0112, "step": 49090 }, { "grad_norm": 0.1840982735157013, "learning_rate": 5.5611602911943596e-05, "loss": 0.0107, "step": 49100 }, { "grad_norm": 0.1454068422317505, "learning_rate": 5.559517237305194e-05, "loss": 0.0103, "step": 49110 }, { "grad_norm": 0.13923031091690063, "learning_rate": 5.557874122228064e-05, "loss": 0.0104, "step": 49120 }, { "grad_norm": 0.29073309898376465, "learning_rate": 5.55623094614266e-05, "loss": 0.0093, "step": 49130 }, { "grad_norm": 0.13793861865997314, "learning_rate": 5.5545877092286743e-05, "loss": 0.009, "step": 49140 }, { "grad_norm": 0.1314346194267273, "learning_rate": 5.552944411665811e-05, "loss": 0.0111, "step": 49150 }, { "grad_norm": 0.19590818881988525, "learning_rate": 5.5513010536337795e-05, "loss": 0.0124, "step": 49160 }, { "grad_norm": 0.17454273998737335, "learning_rate": 5.5496576353122934e-05, "loss": 0.01, "step": 49170 }, { "grad_norm": 0.16128619015216827, "learning_rate": 5.548014156881074e-05, "loss": 0.0102, "step": 49180 }, { "grad_norm": 0.11566925048828125, "learning_rate": 5.5463706185198494e-05, "loss": 0.0105, "step": 49190 }, { "grad_norm": 0.17274022102355957, "learning_rate": 5.544727020408358e-05, "loss": 0.0122, "step": 49200 }, { "grad_norm": 0.12229954451322556, "learning_rate": 5.5430833627263366e-05, "loss": 0.0114, "step": 49210 }, { "grad_norm": 0.14824217557907104, "learning_rate": 5.5414396456535354e-05, "loss": 0.0082, "step": 49220 }, { "grad_norm": 0.13374535739421844, "learning_rate": 5.539795869369709e-05, "loss": 0.0108, "step": 49230 }, { "grad_norm": 0.21793964505195618, "learning_rate": 5.538152034054618e-05, "loss": 0.0118, "step": 49240 }, { "grad_norm": 0.14616268873214722, "learning_rate": 5.536508139888028e-05, "loss": 0.0105, "step": 49250 }, { "grad_norm": 0.14197272062301636, "learning_rate": 5.534864187049716e-05, "loss": 0.0088, "step": 49260 }, { "grad_norm": 0.1591525822877884, "learning_rate": 5.5332201757194615e-05, "loss": 0.0129, "step": 49270 }, { "grad_norm": 0.14382120966911316, "learning_rate": 5.531576106077049e-05, "loss": 0.0093, "step": 49280 }, { "grad_norm": 0.1134566217660904, "learning_rate": 5.529931978302272e-05, "loss": 0.0111, "step": 49290 }, { "grad_norm": 0.22628536820411682, "learning_rate": 5.528287792574932e-05, "loss": 0.01, "step": 49300 }, { "grad_norm": 0.12327927350997925, "learning_rate": 5.5266435490748324e-05, "loss": 0.0076, "step": 49310 }, { "grad_norm": 0.10254040360450745, "learning_rate": 5.524999247981787e-05, "loss": 0.008, "step": 49320 }, { "grad_norm": 0.15540598332881927, "learning_rate": 5.523354889475613e-05, "loss": 0.0113, "step": 49330 }, { "grad_norm": 0.18487586081027985, "learning_rate": 5.521710473736134e-05, "loss": 0.0098, "step": 49340 }, { "grad_norm": 0.25557494163513184, "learning_rate": 5.5200660009431826e-05, "loss": 0.0089, "step": 49350 }, { "grad_norm": 0.15806758403778076, "learning_rate": 5.518421471276596e-05, "loss": 0.0106, "step": 49360 }, { "grad_norm": 0.18862001597881317, "learning_rate": 5.5167768849162174e-05, "loss": 0.01, "step": 49370 }, { "grad_norm": 0.224599689245224, "learning_rate": 5.515132242041893e-05, "loss": 0.011, "step": 49380 }, { "grad_norm": 0.14534056186676025, "learning_rate": 5.513487542833483e-05, "loss": 0.0086, "step": 49390 }, { "grad_norm": 0.16108448803424835, "learning_rate": 5.5118427874708466e-05, "loss": 0.009, "step": 49400 }, { "grad_norm": 0.15886056423187256, "learning_rate": 5.510197976133853e-05, "loss": 0.0115, "step": 49410 }, { "grad_norm": 0.13295656442642212, "learning_rate": 5.508553109002376e-05, "loss": 0.0103, "step": 49420 }, { "grad_norm": 0.15610802173614502, "learning_rate": 5.5069081862562957e-05, "loss": 0.0102, "step": 49430 }, { "grad_norm": 0.13424672186374664, "learning_rate": 5.5052632080754965e-05, "loss": 0.0108, "step": 49440 }, { "grad_norm": 0.11867436021566391, "learning_rate": 5.503618174639874e-05, "loss": 0.0096, "step": 49450 }, { "grad_norm": 0.14033715426921844, "learning_rate": 5.501973086129325e-05, "loss": 0.0084, "step": 49460 }, { "grad_norm": 0.156142920255661, "learning_rate": 5.500327942723753e-05, "loss": 0.0091, "step": 49470 }, { "grad_norm": 0.11856748163700104, "learning_rate": 5.498682744603071e-05, "loss": 0.0095, "step": 49480 }, { "grad_norm": 0.16402852535247803, "learning_rate": 5.497037491947194e-05, "loss": 0.0115, "step": 49490 }, { "grad_norm": 0.12168565392494202, "learning_rate": 5.4953921849360424e-05, "loss": 0.0091, "step": 49500 }, { "grad_norm": 0.1509040892124176, "learning_rate": 5.493746823749547e-05, "loss": 0.0088, "step": 49510 }, { "grad_norm": 0.12623269855976105, "learning_rate": 5.49210140856764e-05, "loss": 0.009, "step": 49520 }, { "grad_norm": 0.14822499454021454, "learning_rate": 5.4904559395702647e-05, "loss": 0.0091, "step": 49530 }, { "grad_norm": 0.1656200885772705, "learning_rate": 5.488810416937364e-05, "loss": 0.0098, "step": 49540 }, { "grad_norm": 0.16926242411136627, "learning_rate": 5.4871648408488926e-05, "loss": 0.0119, "step": 49550 }, { "grad_norm": 0.14536789059638977, "learning_rate": 5.485519211484807e-05, "loss": 0.0093, "step": 49560 }, { "grad_norm": 0.13245151937007904, "learning_rate": 5.4838735290250676e-05, "loss": 0.0094, "step": 49570 }, { "grad_norm": 0.1426638513803482, "learning_rate": 5.482227793649649e-05, "loss": 0.0099, "step": 49580 }, { "grad_norm": 0.12120257318019867, "learning_rate": 5.480582005538524e-05, "loss": 0.0108, "step": 49590 }, { "grad_norm": 0.11249744147062302, "learning_rate": 5.478936164871671e-05, "loss": 0.0092, "step": 49600 }, { "grad_norm": 0.10826583951711655, "learning_rate": 5.477290271829081e-05, "loss": 0.0095, "step": 49610 }, { "grad_norm": 0.1867247074842453, "learning_rate": 5.475644326590744e-05, "loss": 0.0087, "step": 49620 }, { "grad_norm": 0.12165036797523499, "learning_rate": 5.473998329336658e-05, "loss": 0.0087, "step": 49630 }, { "grad_norm": 0.13711513578891754, "learning_rate": 5.4723522802468286e-05, "loss": 0.0113, "step": 49640 }, { "grad_norm": 0.16207048296928406, "learning_rate": 5.470706179501264e-05, "loss": 0.009, "step": 49650 }, { "grad_norm": 0.11175964027643204, "learning_rate": 5.4690600272799776e-05, "loss": 0.0091, "step": 49660 }, { "grad_norm": 0.142986461520195, "learning_rate": 5.467413823762993e-05, "loss": 0.009, "step": 49670 }, { "grad_norm": 0.14209625124931335, "learning_rate": 5.465767569130335e-05, "loss": 0.0093, "step": 49680 }, { "grad_norm": 0.18220043182373047, "learning_rate": 5.464121263562036e-05, "loss": 0.01, "step": 49690 }, { "grad_norm": 0.17031647264957428, "learning_rate": 5.4624749072381343e-05, "loss": 0.0095, "step": 49700 }, { "grad_norm": 0.12420903891324997, "learning_rate": 5.460828500338672e-05, "loss": 0.0087, "step": 49710 }, { "grad_norm": 0.09434240311384201, "learning_rate": 5.459182043043698e-05, "loss": 0.0083, "step": 49720 }, { "grad_norm": 0.12641729414463043, "learning_rate": 5.457535535533265e-05, "loss": 0.0097, "step": 49730 }, { "grad_norm": 0.12538474798202515, "learning_rate": 5.4558889779874334e-05, "loss": 0.0074, "step": 49740 }, { "grad_norm": 0.12722338736057281, "learning_rate": 5.454242370586269e-05, "loss": 0.0104, "step": 49750 }, { "grad_norm": 0.1289779543876648, "learning_rate": 5.452595713509843e-05, "loss": 0.0089, "step": 49760 }, { "grad_norm": 0.17468667030334473, "learning_rate": 5.4509490069382274e-05, "loss": 0.0084, "step": 49770 }, { "grad_norm": 0.1827162355184555, "learning_rate": 5.449302251051509e-05, "loss": 0.0101, "step": 49780 }, { "grad_norm": 0.1671185940504074, "learning_rate": 5.44765544602977e-05, "loss": 0.0097, "step": 49790 }, { "grad_norm": 0.21987603604793549, "learning_rate": 5.446008592053107e-05, "loss": 0.0096, "step": 49800 }, { "grad_norm": 0.09791955351829529, "learning_rate": 5.4443616893016135e-05, "loss": 0.009, "step": 49810 }, { "grad_norm": 0.10607537627220154, "learning_rate": 5.4427147379553934e-05, "loss": 0.0118, "step": 49820 }, { "grad_norm": 0.13748720288276672, "learning_rate": 5.441067738194556e-05, "loss": 0.0123, "step": 49830 }, { "grad_norm": 0.13777655363082886, "learning_rate": 5.439420690199214e-05, "loss": 0.0097, "step": 49840 }, { "grad_norm": 0.13550208508968353, "learning_rate": 5.437773594149484e-05, "loss": 0.0102, "step": 49850 }, { "grad_norm": 0.10277862101793289, "learning_rate": 5.436126450225495e-05, "loss": 0.0113, "step": 49860 }, { "grad_norm": 0.16985110938549042, "learning_rate": 5.434479258607371e-05, "loss": 0.0105, "step": 49870 }, { "grad_norm": 0.23754803836345673, "learning_rate": 5.43283201947525e-05, "loss": 0.0094, "step": 49880 }, { "grad_norm": 0.10108514875173569, "learning_rate": 5.431184733009268e-05, "loss": 0.0086, "step": 49890 }, { "grad_norm": 0.12293507903814316, "learning_rate": 5.4295373993895736e-05, "loss": 0.0088, "step": 49900 }, { "grad_norm": 0.10285396873950958, "learning_rate": 5.4278900187963157e-05, "loss": 0.0087, "step": 49910 }, { "grad_norm": 0.1496201902627945, "learning_rate": 5.426242591409646e-05, "loss": 0.0114, "step": 49920 }, { "grad_norm": 0.15181118249893188, "learning_rate": 5.42459511740973e-05, "loss": 0.0091, "step": 49930 }, { "grad_norm": 0.16087046265602112, "learning_rate": 5.422947596976729e-05, "loss": 0.0118, "step": 49940 }, { "grad_norm": 0.10326847434043884, "learning_rate": 5.4213000302908134e-05, "loss": 0.0112, "step": 49950 }, { "grad_norm": 0.14584660530090332, "learning_rate": 5.419652417532162e-05, "loss": 0.0085, "step": 49960 }, { "grad_norm": 0.16980746388435364, "learning_rate": 5.4180047588809534e-05, "loss": 0.0086, "step": 49970 }, { "grad_norm": 0.23317524790763855, "learning_rate": 5.4163570545173704e-05, "loss": 0.0084, "step": 49980 }, { "grad_norm": 0.17723076045513153, "learning_rate": 5.414709304621608e-05, "loss": 0.0094, "step": 49990 }, { "grad_norm": 0.16379298269748688, "learning_rate": 5.413061509373858e-05, "loss": 0.0098, "step": 50000 }, { "grad_norm": 0.11041621118783951, "learning_rate": 5.411413668954324e-05, "loss": 0.0099, "step": 50010 }, { "grad_norm": 0.12547127902507782, "learning_rate": 5.4097657835432083e-05, "loss": 0.0103, "step": 50020 }, { "grad_norm": 0.2669958472251892, "learning_rate": 5.408117853320723e-05, "loss": 0.0097, "step": 50030 }, { "grad_norm": 0.20576398074626923, "learning_rate": 5.406469878467084e-05, "loss": 0.0126, "step": 50040 }, { "grad_norm": 0.16620297729969025, "learning_rate": 5.404821859162509e-05, "loss": 0.0119, "step": 50050 }, { "grad_norm": 0.16164235770702362, "learning_rate": 5.403173795587225e-05, "loss": 0.0124, "step": 50060 }, { "grad_norm": 0.17804287374019623, "learning_rate": 5.4015256879214606e-05, "loss": 0.0153, "step": 50070 }, { "grad_norm": 0.1860884130001068, "learning_rate": 5.39987753634545e-05, "loss": 0.0146, "step": 50080 }, { "grad_norm": 0.12982703745365143, "learning_rate": 5.3982293410394335e-05, "loss": 0.0113, "step": 50090 }, { "grad_norm": 0.19327080249786377, "learning_rate": 5.396581102183655e-05, "loss": 0.0115, "step": 50100 }, { "grad_norm": 0.1975776106119156, "learning_rate": 5.3949328199583634e-05, "loss": 0.0125, "step": 50110 }, { "grad_norm": 0.24451743066310883, "learning_rate": 5.393284494543813e-05, "loss": 0.0117, "step": 50120 }, { "grad_norm": 0.12644320726394653, "learning_rate": 5.391636126120262e-05, "loss": 0.0131, "step": 50130 }, { "grad_norm": 0.14634987711906433, "learning_rate": 5.389987714867971e-05, "loss": 0.0111, "step": 50140 }, { "grad_norm": 0.09231393784284592, "learning_rate": 5.3883392609672114e-05, "loss": 0.0117, "step": 50150 }, { "grad_norm": 0.16682930290699005, "learning_rate": 5.3866907645982545e-05, "loss": 0.01, "step": 50160 }, { "grad_norm": 0.1558980643749237, "learning_rate": 5.385042225941375e-05, "loss": 0.0115, "step": 50170 }, { "grad_norm": 0.22776581346988678, "learning_rate": 5.3833936451768585e-05, "loss": 0.0107, "step": 50180 }, { "grad_norm": 0.11792749166488647, "learning_rate": 5.3817450224849884e-05, "loss": 0.0105, "step": 50190 }, { "grad_norm": 0.14873553812503815, "learning_rate": 5.380096358046056e-05, "loss": 0.0125, "step": 50200 }, { "grad_norm": 0.18971696496009827, "learning_rate": 5.378447652040359e-05, "loss": 0.0105, "step": 50210 }, { "grad_norm": 0.4599340260028839, "learning_rate": 5.376798904648195e-05, "loss": 0.0113, "step": 50220 }, { "grad_norm": 0.14052744209766388, "learning_rate": 5.375150116049869e-05, "loss": 0.0126, "step": 50230 }, { "grad_norm": 0.15617333352565765, "learning_rate": 5.373501286425691e-05, "loss": 0.0103, "step": 50240 }, { "grad_norm": 0.1314995288848877, "learning_rate": 5.3718524159559726e-05, "loss": 0.0098, "step": 50250 }, { "grad_norm": 0.1545267552137375, "learning_rate": 5.370203504821034e-05, "loss": 0.0119, "step": 50260 }, { "grad_norm": 0.22550314664840698, "learning_rate": 5.368554553201196e-05, "loss": 0.0123, "step": 50270 }, { "grad_norm": 0.11858770996332169, "learning_rate": 5.3669055612767874e-05, "loss": 0.0119, "step": 50280 }, { "grad_norm": 0.1616525650024414, "learning_rate": 5.3652565292281376e-05, "loss": 0.0104, "step": 50290 }, { "grad_norm": 0.14483393728733063, "learning_rate": 5.3636074572355824e-05, "loss": 0.0102, "step": 50300 }, { "grad_norm": 0.1639111489057541, "learning_rate": 5.361958345479463e-05, "loss": 0.0099, "step": 50310 }, { "grad_norm": 0.14359162747859955, "learning_rate": 5.360309194140123e-05, "loss": 0.01, "step": 50320 }, { "grad_norm": 0.1257173866033554, "learning_rate": 5.3586600033979106e-05, "loss": 0.0114, "step": 50330 }, { "grad_norm": 0.17102761566638947, "learning_rate": 5.357010773433181e-05, "loss": 0.0142, "step": 50340 }, { "grad_norm": 0.15245477855205536, "learning_rate": 5.3553615044262895e-05, "loss": 0.0114, "step": 50350 }, { "grad_norm": 0.23525571823120117, "learning_rate": 5.3537121965575964e-05, "loss": 0.0138, "step": 50360 }, { "grad_norm": 0.23736326396465302, "learning_rate": 5.352062850007471e-05, "loss": 0.0168, "step": 50370 }, { "grad_norm": 0.21852907538414001, "learning_rate": 5.350413464956282e-05, "loss": 0.0136, "step": 50380 }, { "grad_norm": 0.18465328216552734, "learning_rate": 5.348764041584403e-05, "loss": 0.0125, "step": 50390 }, { "grad_norm": 0.18911421298980713, "learning_rate": 5.347114580072212e-05, "loss": 0.0114, "step": 50400 }, { "grad_norm": 0.17417728900909424, "learning_rate": 5.345465080600093e-05, "loss": 0.0105, "step": 50410 }, { "grad_norm": 0.22651562094688416, "learning_rate": 5.3438155433484326e-05, "loss": 0.0103, "step": 50420 }, { "grad_norm": 0.11893864721059799, "learning_rate": 5.3421659684976197e-05, "loss": 0.0099, "step": 50430 }, { "grad_norm": 0.1540326327085495, "learning_rate": 5.340516356228052e-05, "loss": 0.0097, "step": 50440 }, { "grad_norm": 0.15309536457061768, "learning_rate": 5.338866706720128e-05, "loss": 0.0107, "step": 50450 }, { "grad_norm": 0.17715522646903992, "learning_rate": 5.337217020154249e-05, "loss": 0.0109, "step": 50460 }, { "grad_norm": 0.15247443318367004, "learning_rate": 5.335567296710825e-05, "loss": 0.0117, "step": 50470 }, { "grad_norm": 0.09976622462272644, "learning_rate": 5.333917536570265e-05, "loss": 0.011, "step": 50480 }, { "grad_norm": 0.11747092008590698, "learning_rate": 5.332267739912986e-05, "loss": 0.0092, "step": 50490 }, { "grad_norm": 0.13888372480869293, "learning_rate": 5.330617906919405e-05, "loss": 0.0093, "step": 50500 }, { "grad_norm": 0.16528327763080597, "learning_rate": 5.328968037769949e-05, "loss": 0.0109, "step": 50510 }, { "grad_norm": 0.17320826649665833, "learning_rate": 5.327318132645042e-05, "loss": 0.009, "step": 50520 }, { "grad_norm": 0.1704566925764084, "learning_rate": 5.325668191725116e-05, "loss": 0.0094, "step": 50530 }, { "grad_norm": 0.12923599779605865, "learning_rate": 5.324018215190606e-05, "loss": 0.0093, "step": 50540 }, { "grad_norm": 0.16848696768283844, "learning_rate": 5.3223682032219515e-05, "loss": 0.0095, "step": 50550 }, { "grad_norm": 0.20224107801914215, "learning_rate": 5.320718155999595e-05, "loss": 0.0127, "step": 50560 }, { "grad_norm": 0.20558881759643555, "learning_rate": 5.3190680737039835e-05, "loss": 0.0171, "step": 50570 }, { "grad_norm": 0.148173525929451, "learning_rate": 5.317417956515567e-05, "loss": 0.0126, "step": 50580 }, { "grad_norm": 0.22781895101070404, "learning_rate": 5.315767804614798e-05, "loss": 0.0135, "step": 50590 }, { "grad_norm": 0.14164209365844727, "learning_rate": 5.3141176181821395e-05, "loss": 0.012, "step": 50600 }, { "grad_norm": 0.1673586219549179, "learning_rate": 5.31246739739805e-05, "loss": 0.0123, "step": 50610 }, { "grad_norm": 0.1692696064710617, "learning_rate": 5.310817142442994e-05, "loss": 0.0123, "step": 50620 }, { "grad_norm": 0.13368918001651764, "learning_rate": 5.309166853497445e-05, "loss": 0.0098, "step": 50630 }, { "grad_norm": 0.12697327136993408, "learning_rate": 5.307516530741873e-05, "loss": 0.011, "step": 50640 }, { "grad_norm": 0.13649220764636993, "learning_rate": 5.305866174356754e-05, "loss": 0.0112, "step": 50650 }, { "grad_norm": 0.13054512441158295, "learning_rate": 5.304215784522571e-05, "loss": 0.0081, "step": 50660 }, { "grad_norm": 0.14493969082832336, "learning_rate": 5.302565361419808e-05, "loss": 0.0096, "step": 50670 }, { "grad_norm": 0.18453830480575562, "learning_rate": 5.3009149052289507e-05, "loss": 0.0092, "step": 50680 }, { "grad_norm": 0.17820772528648376, "learning_rate": 5.299264416130493e-05, "loss": 0.0089, "step": 50690 }, { "grad_norm": 0.11611409485340118, "learning_rate": 5.297613894304928e-05, "loss": 0.0097, "step": 50700 }, { "grad_norm": 0.12271478772163391, "learning_rate": 5.2959633399327534e-05, "loss": 0.0092, "step": 50710 }, { "grad_norm": 0.1244945377111435, "learning_rate": 5.294312753194476e-05, "loss": 0.0113, "step": 50720 }, { "grad_norm": 0.15634766221046448, "learning_rate": 5.292662134270596e-05, "loss": 0.0103, "step": 50730 }, { "grad_norm": 0.16252383589744568, "learning_rate": 5.291011483341626e-05, "loss": 0.0091, "step": 50740 }, { "grad_norm": 0.1537599414587021, "learning_rate": 5.2893608005880767e-05, "loss": 0.0112, "step": 50750 }, { "grad_norm": 0.1791493445634842, "learning_rate": 5.287710086190467e-05, "loss": 0.0093, "step": 50760 }, { "grad_norm": 0.10792315006256104, "learning_rate": 5.286059340329314e-05, "loss": 0.0109, "step": 50770 }, { "grad_norm": 0.15302714705467224, "learning_rate": 5.284408563185141e-05, "loss": 0.0128, "step": 50780 }, { "grad_norm": 0.14147669076919556, "learning_rate": 5.282757754938476e-05, "loss": 0.0107, "step": 50790 }, { "grad_norm": 0.14600127935409546, "learning_rate": 5.281106915769849e-05, "loss": 0.0112, "step": 50800 }, { "grad_norm": 0.12240811437368393, "learning_rate": 5.2794560458597897e-05, "loss": 0.0078, "step": 50810 }, { "grad_norm": 0.15091492235660553, "learning_rate": 5.27780514538884e-05, "loss": 0.0112, "step": 50820 }, { "grad_norm": 0.13699160516262054, "learning_rate": 5.2761542145375365e-05, "loss": 0.0085, "step": 50830 }, { "grad_norm": 0.17604930698871613, "learning_rate": 5.274503253486421e-05, "loss": 0.0094, "step": 50840 }, { "grad_norm": 0.1167907640337944, "learning_rate": 5.272852262416046e-05, "loss": 0.0092, "step": 50850 }, { "grad_norm": 0.18680036067962646, "learning_rate": 5.2712012415069555e-05, "loss": 0.0112, "step": 50860 }, { "grad_norm": 0.2603208124637604, "learning_rate": 5.269550190939705e-05, "loss": 0.0094, "step": 50870 }, { "grad_norm": 0.17042194306850433, "learning_rate": 5.267899110894852e-05, "loss": 0.0097, "step": 50880 }, { "grad_norm": 0.13991869986057281, "learning_rate": 5.266248001552955e-05, "loss": 0.0092, "step": 50890 }, { "grad_norm": 0.13108573853969574, "learning_rate": 5.264596863094575e-05, "loss": 0.0094, "step": 50900 }, { "grad_norm": 0.16679596900939941, "learning_rate": 5.262945695700282e-05, "loss": 0.0092, "step": 50910 }, { "grad_norm": 0.11846987903118134, "learning_rate": 5.261294499550643e-05, "loss": 0.0103, "step": 50920 }, { "grad_norm": 0.12101039290428162, "learning_rate": 5.25964327482623e-05, "loss": 0.0107, "step": 50930 }, { "grad_norm": 0.2399710714817047, "learning_rate": 5.257992021707617e-05, "loss": 0.0096, "step": 50940 }, { "grad_norm": 0.15203474462032318, "learning_rate": 5.256340740375387e-05, "loss": 0.0099, "step": 50950 }, { "grad_norm": 0.11399316787719727, "learning_rate": 5.254689431010117e-05, "loss": 0.0099, "step": 50960 }, { "grad_norm": 0.1536114662885666, "learning_rate": 5.253038093792395e-05, "loss": 0.0078, "step": 50970 }, { "grad_norm": 0.17755722999572754, "learning_rate": 5.251386728902806e-05, "loss": 0.0097, "step": 50980 }, { "grad_norm": 0.1511172503232956, "learning_rate": 5.2497353365219446e-05, "loss": 0.0102, "step": 50990 }, { "grad_norm": 0.12385831028223038, "learning_rate": 5.2480839168304e-05, "loss": 0.0092, "step": 51000 }, { "grad_norm": 0.15937459468841553, "learning_rate": 5.246432470008773e-05, "loss": 0.0077, "step": 51010 }, { "grad_norm": 0.12301193922758102, "learning_rate": 5.2447809962376616e-05, "loss": 0.0079, "step": 51020 }, { "grad_norm": 0.11306796222925186, "learning_rate": 5.243129495697666e-05, "loss": 0.0086, "step": 51030 }, { "grad_norm": 0.22235213220119476, "learning_rate": 5.241477968569397e-05, "loss": 0.0105, "step": 51040 }, { "grad_norm": 0.13680022954940796, "learning_rate": 5.23982641503346e-05, "loss": 0.01, "step": 51050 }, { "grad_norm": 0.15852443873882294, "learning_rate": 5.238174835270465e-05, "loss": 0.0104, "step": 51060 }, { "grad_norm": 0.131050243973732, "learning_rate": 5.23652322946103e-05, "loss": 0.0086, "step": 51070 }, { "grad_norm": 0.1186608374118805, "learning_rate": 5.234871597785771e-05, "loss": 0.0095, "step": 51080 }, { "grad_norm": 0.13215188682079315, "learning_rate": 5.233219940425306e-05, "loss": 0.0081, "step": 51090 }, { "grad_norm": 0.2025149017572403, "learning_rate": 5.2315682575602576e-05, "loss": 0.0107, "step": 51100 }, { "grad_norm": 0.13798795640468597, "learning_rate": 5.229916549371254e-05, "loss": 0.0103, "step": 51110 }, { "grad_norm": 0.11772400140762329, "learning_rate": 5.228264816038924e-05, "loss": 0.0087, "step": 51120 }, { "grad_norm": 0.2825058400630951, "learning_rate": 5.2266130577438935e-05, "loss": 0.0088, "step": 51130 }, { "grad_norm": 0.10241382569074631, "learning_rate": 5.224961274666802e-05, "loss": 0.0078, "step": 51140 }, { "grad_norm": 0.1597653180360794, "learning_rate": 5.223309466988283e-05, "loss": 0.0091, "step": 51150 }, { "grad_norm": 0.11778134852647781, "learning_rate": 5.221657634888976e-05, "loss": 0.0098, "step": 51160 }, { "grad_norm": 0.16498374938964844, "learning_rate": 5.2200057785495235e-05, "loss": 0.0115, "step": 51170 }, { "grad_norm": 0.16369710862636566, "learning_rate": 5.2183538981505706e-05, "loss": 0.0106, "step": 51180 }, { "grad_norm": 0.18797680735588074, "learning_rate": 5.216701993872762e-05, "loss": 0.0092, "step": 51190 }, { "grad_norm": 0.185142382979393, "learning_rate": 5.215050065896752e-05, "loss": 0.0086, "step": 51200 }, { "grad_norm": 0.20749497413635254, "learning_rate": 5.2133981144031875e-05, "loss": 0.0094, "step": 51210 }, { "grad_norm": 0.12175779044628143, "learning_rate": 5.211746139572727e-05, "loss": 0.0102, "step": 51220 }, { "grad_norm": 0.16151568293571472, "learning_rate": 5.210094141586026e-05, "loss": 0.0096, "step": 51230 }, { "grad_norm": 0.1566953957080841, "learning_rate": 5.2084421206237476e-05, "loss": 0.0097, "step": 51240 }, { "grad_norm": 0.13217629492282867, "learning_rate": 5.20679007686655e-05, "loss": 0.0085, "step": 51250 }, { "grad_norm": 0.1317145973443985, "learning_rate": 5.205138010495102e-05, "loss": 0.0098, "step": 51260 }, { "grad_norm": 0.09482491761445999, "learning_rate": 5.20348592169007e-05, "loss": 0.0074, "step": 51270 }, { "grad_norm": 0.125077486038208, "learning_rate": 5.201833810632123e-05, "loss": 0.0091, "step": 51280 }, { "grad_norm": 0.13674667477607727, "learning_rate": 5.200181677501933e-05, "loss": 0.0079, "step": 51290 }, { "grad_norm": 0.1274593621492386, "learning_rate": 5.1985295224801777e-05, "loss": 0.0085, "step": 51300 }, { "grad_norm": 0.09982812404632568, "learning_rate": 5.196877345747531e-05, "loss": 0.0083, "step": 51310 }, { "grad_norm": 0.1506650149822235, "learning_rate": 5.195225147484674e-05, "loss": 0.0078, "step": 51320 }, { "grad_norm": 0.1934451162815094, "learning_rate": 5.193572927872288e-05, "loss": 0.0095, "step": 51330 }, { "grad_norm": 0.1503307968378067, "learning_rate": 5.191920687091059e-05, "loss": 0.0084, "step": 51340 }, { "grad_norm": 0.15338046848773956, "learning_rate": 5.1902684253216695e-05, "loss": 0.0076, "step": 51350 }, { "grad_norm": 0.16049544513225555, "learning_rate": 5.1886161427448134e-05, "loss": 0.0082, "step": 51360 }, { "grad_norm": 0.1085018590092659, "learning_rate": 5.186963839541179e-05, "loss": 0.0083, "step": 51370 }, { "grad_norm": 0.23392832279205322, "learning_rate": 5.185311515891459e-05, "loss": 0.0105, "step": 51380 }, { "grad_norm": 0.13227450847625732, "learning_rate": 5.1836591719763516e-05, "loss": 0.0078, "step": 51390 }, { "grad_norm": 0.10709100216627121, "learning_rate": 5.182006807976554e-05, "loss": 0.0082, "step": 51400 }, { "grad_norm": 0.1656360924243927, "learning_rate": 5.180354424072763e-05, "loss": 0.0087, "step": 51410 }, { "grad_norm": 0.12138085812330246, "learning_rate": 5.178702020445686e-05, "loss": 0.0089, "step": 51420 }, { "grad_norm": 0.17635655403137207, "learning_rate": 5.177049597276024e-05, "loss": 0.0097, "step": 51430 }, { "grad_norm": 0.13416223227977753, "learning_rate": 5.175397154744482e-05, "loss": 0.0084, "step": 51440 }, { "grad_norm": 0.13653242588043213, "learning_rate": 5.1737446930317736e-05, "loss": 0.0096, "step": 51450 }, { "grad_norm": 0.16256719827651978, "learning_rate": 5.172092212318604e-05, "loss": 0.0092, "step": 51460 }, { "grad_norm": 0.14525887370109558, "learning_rate": 5.170439712785692e-05, "loss": 0.0088, "step": 51470 }, { "grad_norm": 0.14421045780181885, "learning_rate": 5.1687871946137467e-05, "loss": 0.0087, "step": 51480 }, { "grad_norm": 0.11512308567762375, "learning_rate": 5.16713465798349e-05, "loss": 0.0091, "step": 51490 }, { "grad_norm": 0.140392005443573, "learning_rate": 5.165482103075637e-05, "loss": 0.0084, "step": 51500 }, { "grad_norm": 0.12925314903259277, "learning_rate": 5.163829530070909e-05, "loss": 0.0087, "step": 51510 }, { "grad_norm": 0.16374057531356812, "learning_rate": 5.162176939150032e-05, "loss": 0.0077, "step": 51520 }, { "grad_norm": 0.12118705362081528, "learning_rate": 5.160524330493728e-05, "loss": 0.0082, "step": 51530 }, { "grad_norm": 0.23276735842227936, "learning_rate": 5.158871704282724e-05, "loss": 0.0127, "step": 51540 }, { "grad_norm": 0.133916974067688, "learning_rate": 5.15721906069775e-05, "loss": 0.0082, "step": 51550 }, { "grad_norm": 0.13006217777729034, "learning_rate": 5.155566399919536e-05, "loss": 0.0103, "step": 51560 }, { "grad_norm": 0.1516636162996292, "learning_rate": 5.153913722128813e-05, "loss": 0.0083, "step": 51570 }, { "grad_norm": 0.16148844361305237, "learning_rate": 5.1522610275063176e-05, "loss": 0.0096, "step": 51580 }, { "grad_norm": 0.1339043378829956, "learning_rate": 5.150608316232786e-05, "loss": 0.0087, "step": 51590 }, { "grad_norm": 0.09354989975690842, "learning_rate": 5.148955588488955e-05, "loss": 0.0092, "step": 51600 }, { "grad_norm": 0.17848610877990723, "learning_rate": 5.147302844455564e-05, "loss": 0.0098, "step": 51610 }, { "grad_norm": 0.13891850411891937, "learning_rate": 5.145650084313356e-05, "loss": 0.0108, "step": 51620 }, { "grad_norm": 0.23452778160572052, "learning_rate": 5.1439973082430734e-05, "loss": 0.0121, "step": 51630 }, { "grad_norm": 0.1723414808511734, "learning_rate": 5.142344516425461e-05, "loss": 0.0103, "step": 51640 }, { "grad_norm": 0.14001621305942535, "learning_rate": 5.1406917090412673e-05, "loss": 0.01, "step": 51650 }, { "grad_norm": 0.13031552731990814, "learning_rate": 5.1390388862712404e-05, "loss": 0.0089, "step": 51660 }, { "grad_norm": 0.1803794503211975, "learning_rate": 5.137386048296129e-05, "loss": 0.0087, "step": 51670 }, { "grad_norm": 0.13731259107589722, "learning_rate": 5.135733195296686e-05, "loss": 0.0086, "step": 51680 }, { "grad_norm": 0.11544998735189438, "learning_rate": 5.134080327453664e-05, "loss": 0.0093, "step": 51690 }, { "grad_norm": 0.150513157248497, "learning_rate": 5.1324274449478204e-05, "loss": 0.009, "step": 51700 }, { "grad_norm": 0.1391259729862213, "learning_rate": 5.130774547959909e-05, "loss": 0.0087, "step": 51710 }, { "grad_norm": 0.18727146089076996, "learning_rate": 5.129121636670692e-05, "loss": 0.0102, "step": 51720 }, { "grad_norm": 0.1220742017030716, "learning_rate": 5.1274687112609246e-05, "loss": 0.0091, "step": 51730 }, { "grad_norm": 0.12115669250488281, "learning_rate": 5.125815771911372e-05, "loss": 0.0095, "step": 51740 }, { "grad_norm": 0.10248511284589767, "learning_rate": 5.1241628188027956e-05, "loss": 0.0089, "step": 51750 }, { "grad_norm": 0.13434016704559326, "learning_rate": 5.1225098521159585e-05, "loss": 0.0083, "step": 51760 }, { "grad_norm": 0.13184890151023865, "learning_rate": 5.12085687203163e-05, "loss": 0.0091, "step": 51770 }, { "grad_norm": 0.12777221202850342, "learning_rate": 5.119203878730575e-05, "loss": 0.0073, "step": 51780 }, { "grad_norm": 0.12210123986005783, "learning_rate": 5.117550872393564e-05, "loss": 0.0119, "step": 51790 }, { "grad_norm": 0.10709588974714279, "learning_rate": 5.1158978532013645e-05, "loss": 0.0077, "step": 51800 }, { "grad_norm": 0.10622766613960266, "learning_rate": 5.114244821334751e-05, "loss": 0.0077, "step": 51810 }, { "grad_norm": 0.12266853451728821, "learning_rate": 5.112591776974497e-05, "loss": 0.0076, "step": 51820 }, { "grad_norm": 0.14233683049678802, "learning_rate": 5.110938720301374e-05, "loss": 0.0087, "step": 51830 }, { "grad_norm": 0.14648142457008362, "learning_rate": 5.1092856514961595e-05, "loss": 0.0088, "step": 51840 }, { "grad_norm": 0.10563404113054276, "learning_rate": 5.107632570739631e-05, "loss": 0.0095, "step": 51850 }, { "grad_norm": 0.14790871739387512, "learning_rate": 5.1059794782125645e-05, "loss": 0.0086, "step": 51860 }, { "grad_norm": 0.1758023500442505, "learning_rate": 5.104326374095744e-05, "loss": 0.0111, "step": 51870 }, { "grad_norm": 0.15046469867229462, "learning_rate": 5.102673258569947e-05, "loss": 0.0093, "step": 51880 }, { "grad_norm": 0.12001950293779373, "learning_rate": 5.101020131815956e-05, "loss": 0.0076, "step": 51890 }, { "grad_norm": 0.18484915792942047, "learning_rate": 5.099366994014556e-05, "loss": 0.0088, "step": 51900 }, { "grad_norm": 0.1614055633544922, "learning_rate": 5.097713845346532e-05, "loss": 0.0077, "step": 51910 }, { "grad_norm": 0.12835693359375, "learning_rate": 5.096060685992668e-05, "loss": 0.0079, "step": 51920 }, { "grad_norm": 0.1825239211320877, "learning_rate": 5.0944075161337514e-05, "loss": 0.0091, "step": 51930 }, { "grad_norm": 0.20532745122909546, "learning_rate": 5.09275433595057e-05, "loss": 0.0087, "step": 51940 }, { "grad_norm": 0.13319836556911469, "learning_rate": 5.0911011456239157e-05, "loss": 0.0082, "step": 51950 }, { "grad_norm": 0.12502773106098175, "learning_rate": 5.089447945334574e-05, "loss": 0.0107, "step": 51960 }, { "grad_norm": 0.15972156822681427, "learning_rate": 5.087794735263343e-05, "loss": 0.0086, "step": 51970 }, { "grad_norm": 0.11523404717445374, "learning_rate": 5.086141515591011e-05, "loss": 0.0095, "step": 51980 }, { "grad_norm": 0.1381402611732483, "learning_rate": 5.084488286498371e-05, "loss": 0.0079, "step": 51990 }, { "grad_norm": 0.11899339407682419, "learning_rate": 5.08283504816622e-05, "loss": 0.0108, "step": 52000 }, { "grad_norm": 0.11089731007814407, "learning_rate": 5.081181800775353e-05, "loss": 0.0082, "step": 52010 }, { "grad_norm": 0.13146796822547913, "learning_rate": 5.079528544506565e-05, "loss": 0.0078, "step": 52020 }, { "grad_norm": 0.12896326184272766, "learning_rate": 5.077875279540657e-05, "loss": 0.0086, "step": 52030 }, { "grad_norm": 0.1658753603696823, "learning_rate": 5.076222006058424e-05, "loss": 0.0101, "step": 52040 }, { "grad_norm": 0.18614260852336884, "learning_rate": 5.074568724240667e-05, "loss": 0.0075, "step": 52050 }, { "grad_norm": 0.12651845812797546, "learning_rate": 5.072915434268187e-05, "loss": 0.0093, "step": 52060 }, { "grad_norm": 0.1254754215478897, "learning_rate": 5.071262136321785e-05, "loss": 0.0091, "step": 52070 }, { "grad_norm": 0.11989196389913559, "learning_rate": 5.069608830582262e-05, "loss": 0.0061, "step": 52080 }, { "grad_norm": 0.1292385756969452, "learning_rate": 5.067955517230423e-05, "loss": 0.0099, "step": 52090 }, { "grad_norm": 0.1356799155473709, "learning_rate": 5.06630219644707e-05, "loss": 0.0079, "step": 52100 }, { "grad_norm": 0.2877293527126312, "learning_rate": 5.064648868413008e-05, "loss": 0.0092, "step": 52110 }, { "grad_norm": 0.13879166543483734, "learning_rate": 5.062995533309044e-05, "loss": 0.0091, "step": 52120 }, { "grad_norm": 0.22415579855442047, "learning_rate": 5.061342191315983e-05, "loss": 0.0096, "step": 52130 }, { "grad_norm": 0.12323939800262451, "learning_rate": 5.0596888426146325e-05, "loss": 0.0077, "step": 52140 }, { "grad_norm": 0.17038866877555847, "learning_rate": 5.058035487385798e-05, "loss": 0.008, "step": 52150 }, { "grad_norm": 0.15646246075630188, "learning_rate": 5.0563821258102915e-05, "loss": 0.0076, "step": 52160 }, { "grad_norm": 0.09278231859207153, "learning_rate": 5.05472875806892e-05, "loss": 0.0082, "step": 52170 }, { "grad_norm": 0.12407627701759338, "learning_rate": 5.053075384342494e-05, "loss": 0.0079, "step": 52180 }, { "grad_norm": 0.10384286940097809, "learning_rate": 5.0514220048118234e-05, "loss": 0.0074, "step": 52190 }, { "grad_norm": 0.1974959820508957, "learning_rate": 5.049768619657721e-05, "loss": 0.0091, "step": 52200 }, { "grad_norm": 0.1825256049633026, "learning_rate": 5.048115229060996e-05, "loss": 0.0093, "step": 52210 }, { "grad_norm": 0.12499496340751648, "learning_rate": 5.046461833202462e-05, "loss": 0.0094, "step": 52220 }, { "grad_norm": 0.12165132164955139, "learning_rate": 5.044808432262932e-05, "loss": 0.0106, "step": 52230 }, { "grad_norm": 0.19374588131904602, "learning_rate": 5.043155026423219e-05, "loss": 0.0079, "step": 52240 }, { "grad_norm": 0.13688476383686066, "learning_rate": 5.041501615864137e-05, "loss": 0.0083, "step": 52250 }, { "grad_norm": 0.12999145686626434, "learning_rate": 5.039848200766502e-05, "loss": 0.0086, "step": 52260 }, { "grad_norm": 0.14313367009162903, "learning_rate": 5.038194781311125e-05, "loss": 0.0084, "step": 52270 }, { "grad_norm": 0.11289375275373459, "learning_rate": 5.0365413576788264e-05, "loss": 0.0081, "step": 52280 }, { "grad_norm": 0.11761212348937988, "learning_rate": 5.0348879300504194e-05, "loss": 0.0083, "step": 52290 }, { "grad_norm": 0.15422940254211426, "learning_rate": 5.033234498606721e-05, "loss": 0.0086, "step": 52300 }, { "grad_norm": 0.12309601902961731, "learning_rate": 5.0315810635285476e-05, "loss": 0.0086, "step": 52310 }, { "grad_norm": 0.13148169219493866, "learning_rate": 5.0299276249967164e-05, "loss": 0.008, "step": 52320 }, { "grad_norm": 0.1991167813539505, "learning_rate": 5.0282741831920454e-05, "loss": 0.0089, "step": 52330 }, { "grad_norm": 0.13471297919750214, "learning_rate": 5.026620738295351e-05, "loss": 0.0082, "step": 52340 }, { "grad_norm": 0.15027455985546112, "learning_rate": 5.024967290487455e-05, "loss": 0.0083, "step": 52350 }, { "grad_norm": 0.11113619804382324, "learning_rate": 5.023313839949172e-05, "loss": 0.0083, "step": 52360 }, { "grad_norm": 0.1501714289188385, "learning_rate": 5.0216603868613224e-05, "loss": 0.009, "step": 52370 }, { "grad_norm": 0.1554039567708969, "learning_rate": 5.020006931404726e-05, "loss": 0.009, "step": 52380 }, { "grad_norm": 0.13932043313980103, "learning_rate": 5.0183534737602014e-05, "loss": 0.0095, "step": 52390 }, { "grad_norm": 0.22043335437774658, "learning_rate": 5.0167000141085686e-05, "loss": 0.0099, "step": 52400 }, { "grad_norm": 0.18510116636753082, "learning_rate": 5.015046552630647e-05, "loss": 0.0089, "step": 52410 }, { "grad_norm": 0.13367857038974762, "learning_rate": 5.013393089507258e-05, "loss": 0.0098, "step": 52420 }, { "grad_norm": 0.1752949208021164, "learning_rate": 5.011739624919219e-05, "loss": 0.0082, "step": 52430 }, { "grad_norm": 0.11399279534816742, "learning_rate": 5.0100861590473545e-05, "loss": 0.0085, "step": 52440 }, { "grad_norm": 0.13751477003097534, "learning_rate": 5.008432692072482e-05, "loss": 0.0098, "step": 52450 }, { "grad_norm": 0.1382593810558319, "learning_rate": 5.006779224175423e-05, "loss": 0.0075, "step": 52460 }, { "grad_norm": 0.16734760999679565, "learning_rate": 5.005125755536998e-05, "loss": 0.0087, "step": 52470 }, { "grad_norm": 0.14457851648330688, "learning_rate": 5.003472286338029e-05, "loss": 0.0079, "step": 52480 }, { "grad_norm": 0.16663601994514465, "learning_rate": 5.001818816759335e-05, "loss": 0.0093, "step": 52490 }, { "grad_norm": 0.13973134756088257, "learning_rate": 5.000165346981738e-05, "loss": 0.0092, "step": 52500 }, { "grad_norm": 0.11710551381111145, "learning_rate": 4.9985118771860595e-05, "loss": 0.0072, "step": 52510 }, { "grad_norm": 0.14744701981544495, "learning_rate": 4.996858407553119e-05, "loss": 0.0091, "step": 52520 }, { "grad_norm": 0.1378176212310791, "learning_rate": 4.99520493826374e-05, "loss": 0.0083, "step": 52530 }, { "grad_norm": 0.13376298546791077, "learning_rate": 4.9935514694987384e-05, "loss": 0.0097, "step": 52540 }, { "grad_norm": 0.12891091406345367, "learning_rate": 4.991898001438941e-05, "loss": 0.0074, "step": 52550 }, { "grad_norm": 0.17735500633716583, "learning_rate": 4.990244534265166e-05, "loss": 0.0089, "step": 52560 }, { "grad_norm": 0.17989736795425415, "learning_rate": 4.988591068158234e-05, "loss": 0.0101, "step": 52570 }, { "grad_norm": 0.13853512704372406, "learning_rate": 4.986937603298965e-05, "loss": 0.0087, "step": 52580 }, { "grad_norm": 0.1996079683303833, "learning_rate": 4.985284139868181e-05, "loss": 0.0106, "step": 52590 }, { "grad_norm": 0.14629675447940826, "learning_rate": 4.9836306780467e-05, "loss": 0.0093, "step": 52600 }, { "grad_norm": 0.16960039734840393, "learning_rate": 4.981977218015343e-05, "loss": 0.0073, "step": 52610 }, { "grad_norm": 0.12126096338033676, "learning_rate": 4.980323759954932e-05, "loss": 0.0103, "step": 52620 }, { "grad_norm": 0.3263159394264221, "learning_rate": 4.978670304046284e-05, "loss": 0.0087, "step": 52630 }, { "grad_norm": 0.17845933139324188, "learning_rate": 4.9770168504702195e-05, "loss": 0.0111, "step": 52640 }, { "grad_norm": 0.14215785264968872, "learning_rate": 4.975363399407559e-05, "loss": 0.0091, "step": 52650 }, { "grad_norm": 0.15455195307731628, "learning_rate": 4.9737099510391194e-05, "loss": 0.0104, "step": 52660 }, { "grad_norm": 0.123029425740242, "learning_rate": 4.972056505545719e-05, "loss": 0.0104, "step": 52670 }, { "grad_norm": 0.149839386343956, "learning_rate": 4.9704030631081775e-05, "loss": 0.011, "step": 52680 }, { "grad_norm": 0.21975663304328918, "learning_rate": 4.968749623907313e-05, "loss": 0.01, "step": 52690 }, { "grad_norm": 0.15701577067375183, "learning_rate": 4.967096188123943e-05, "loss": 0.0092, "step": 52700 }, { "grad_norm": 0.19945284724235535, "learning_rate": 4.965442755938884e-05, "loss": 0.0083, "step": 52710 }, { "grad_norm": 0.18083614110946655, "learning_rate": 4.963789327532954e-05, "loss": 0.0094, "step": 52720 }, { "grad_norm": 0.20174553990364075, "learning_rate": 4.962135903086968e-05, "loss": 0.0087, "step": 52730 }, { "grad_norm": 0.13803550601005554, "learning_rate": 4.960482482781741e-05, "loss": 0.0083, "step": 52740 }, { "grad_norm": 0.10454288870096207, "learning_rate": 4.958829066798092e-05, "loss": 0.0093, "step": 52750 }, { "grad_norm": 0.30421343445777893, "learning_rate": 4.957175655316835e-05, "loss": 0.0086, "step": 52760 }, { "grad_norm": 0.25200924277305603, "learning_rate": 4.9555222485187834e-05, "loss": 0.0089, "step": 52770 }, { "grad_norm": 0.10792658478021622, "learning_rate": 4.953868846584752e-05, "loss": 0.0087, "step": 52780 }, { "grad_norm": 0.13729768991470337, "learning_rate": 4.9522154496955544e-05, "loss": 0.0078, "step": 52790 }, { "grad_norm": 0.19740736484527588, "learning_rate": 4.950562058032001e-05, "loss": 0.0083, "step": 52800 }, { "grad_norm": 0.1234796941280365, "learning_rate": 4.948908671774909e-05, "loss": 0.0077, "step": 52810 }, { "grad_norm": 0.11734232306480408, "learning_rate": 4.947255291105087e-05, "loss": 0.0102, "step": 52820 }, { "grad_norm": 0.18075522780418396, "learning_rate": 4.945601916203348e-05, "loss": 0.0078, "step": 52830 }, { "grad_norm": 0.11697930097579956, "learning_rate": 4.9439485472505016e-05, "loss": 0.0096, "step": 52840 }, { "grad_norm": 0.17832162976264954, "learning_rate": 4.942295184427356e-05, "loss": 0.0086, "step": 52850 }, { "grad_norm": 0.14494474232196808, "learning_rate": 4.9406418279147234e-05, "loss": 0.0091, "step": 52860 }, { "grad_norm": 0.1467360258102417, "learning_rate": 4.9389884778934116e-05, "loss": 0.0087, "step": 52870 }, { "grad_norm": 0.16573061048984528, "learning_rate": 4.937335134544228e-05, "loss": 0.0078, "step": 52880 }, { "grad_norm": 0.11787626147270203, "learning_rate": 4.935681798047979e-05, "loss": 0.0083, "step": 52890 }, { "grad_norm": 0.20424909889698029, "learning_rate": 4.934028468585473e-05, "loss": 0.0107, "step": 52900 }, { "grad_norm": 0.17097695171833038, "learning_rate": 4.932375146337514e-05, "loss": 0.0076, "step": 52910 }, { "grad_norm": 0.1382947862148285, "learning_rate": 4.930721831484908e-05, "loss": 0.0072, "step": 52920 }, { "grad_norm": 0.16733482480049133, "learning_rate": 4.929068524208456e-05, "loss": 0.0076, "step": 52930 }, { "grad_norm": 0.1958308070898056, "learning_rate": 4.927415224688965e-05, "loss": 0.0118, "step": 52940 }, { "grad_norm": 0.21958909928798676, "learning_rate": 4.9257619331072375e-05, "loss": 0.009, "step": 52950 }, { "grad_norm": 0.2628328502178192, "learning_rate": 4.924108649644072e-05, "loss": 0.009, "step": 52960 }, { "grad_norm": 0.1525522619485855, "learning_rate": 4.922455374480271e-05, "loss": 0.0094, "step": 52970 }, { "grad_norm": 0.12699663639068604, "learning_rate": 4.920802107796634e-05, "loss": 0.008, "step": 52980 }, { "grad_norm": 0.1279401034116745, "learning_rate": 4.9191488497739564e-05, "loss": 0.0081, "step": 52990 }, { "grad_norm": 0.11600655317306519, "learning_rate": 4.917495600593042e-05, "loss": 0.0074, "step": 53000 }, { "grad_norm": 0.2124411165714264, "learning_rate": 4.915842360434684e-05, "loss": 0.0081, "step": 53010 }, { "grad_norm": 0.13779857754707336, "learning_rate": 4.914189129479679e-05, "loss": 0.0085, "step": 53020 }, { "grad_norm": 0.15155860781669617, "learning_rate": 4.912535907908822e-05, "loss": 0.0087, "step": 53030 }, { "grad_norm": 0.15722554922103882, "learning_rate": 4.910882695902907e-05, "loss": 0.0121, "step": 53040 }, { "grad_norm": 0.1828087866306305, "learning_rate": 4.9092294936427234e-05, "loss": 0.0089, "step": 53050 }, { "grad_norm": 0.1304309368133545, "learning_rate": 4.9075763013090685e-05, "loss": 0.0089, "step": 53060 }, { "grad_norm": 0.12878279387950897, "learning_rate": 4.9059231190827305e-05, "loss": 0.0086, "step": 53070 }, { "grad_norm": 0.15545204281806946, "learning_rate": 4.904269947144499e-05, "loss": 0.01, "step": 53080 }, { "grad_norm": 0.18220116198062897, "learning_rate": 4.902616785675161e-05, "loss": 0.0075, "step": 53090 }, { "grad_norm": 0.21348640322685242, "learning_rate": 4.900963634855505e-05, "loss": 0.0083, "step": 53100 }, { "grad_norm": 0.14148637652397156, "learning_rate": 4.8993104948663173e-05, "loss": 0.0088, "step": 53110 }, { "grad_norm": 0.1941441297531128, "learning_rate": 4.8976573658883823e-05, "loss": 0.0113, "step": 53120 }, { "grad_norm": 0.27585485577583313, "learning_rate": 4.896004248102484e-05, "loss": 0.0101, "step": 53130 }, { "grad_norm": 0.17546139657497406, "learning_rate": 4.8943511416894045e-05, "loss": 0.0081, "step": 53140 }, { "grad_norm": 0.15366767346858978, "learning_rate": 4.892698046829926e-05, "loss": 0.007, "step": 53150 }, { "grad_norm": 0.15025237202644348, "learning_rate": 4.8910449637048275e-05, "loss": 0.0079, "step": 53160 }, { "grad_norm": 0.1444062441587448, "learning_rate": 4.8893918924948884e-05, "loss": 0.008, "step": 53170 }, { "grad_norm": 0.09741635620594025, "learning_rate": 4.8877388333808835e-05, "loss": 0.0088, "step": 53180 }, { "grad_norm": 0.20450344681739807, "learning_rate": 4.8860857865435925e-05, "loss": 0.0079, "step": 53190 }, { "grad_norm": 0.1300639510154724, "learning_rate": 4.884432752163789e-05, "loss": 0.0098, "step": 53200 }, { "grad_norm": 0.18042591214179993, "learning_rate": 4.8827797304222466e-05, "loss": 0.0084, "step": 53210 }, { "grad_norm": 0.20503348112106323, "learning_rate": 4.8811267214997366e-05, "loss": 0.0079, "step": 53220 }, { "grad_norm": 0.15907274186611176, "learning_rate": 4.879473725577029e-05, "loss": 0.0088, "step": 53230 }, { "grad_norm": 0.18653038144111633, "learning_rate": 4.877820742834891e-05, "loss": 0.0108, "step": 53240 }, { "grad_norm": 0.13883672654628754, "learning_rate": 4.8761677734540956e-05, "loss": 0.0087, "step": 53250 }, { "grad_norm": 0.1486406773328781, "learning_rate": 4.874514817615407e-05, "loss": 0.0087, "step": 53260 }, { "grad_norm": 0.13410145044326782, "learning_rate": 4.872861875499588e-05, "loss": 0.0105, "step": 53270 }, { "grad_norm": 0.1455197036266327, "learning_rate": 4.871208947287404e-05, "loss": 0.0082, "step": 53280 }, { "grad_norm": 0.17213883996009827, "learning_rate": 4.869556033159615e-05, "loss": 0.0084, "step": 53290 }, { "grad_norm": 0.1476515382528305, "learning_rate": 4.8679031332969835e-05, "loss": 0.0089, "step": 53300 }, { "grad_norm": 0.25494617223739624, "learning_rate": 4.8662502478802635e-05, "loss": 0.0102, "step": 53310 }, { "grad_norm": 0.12396117299795151, "learning_rate": 4.8645973770902176e-05, "loss": 0.0128, "step": 53320 }, { "grad_norm": 0.21994221210479736, "learning_rate": 4.862944521107599e-05, "loss": 0.0093, "step": 53330 }, { "grad_norm": 0.11259909719228745, "learning_rate": 4.861291680113162e-05, "loss": 0.0092, "step": 53340 }, { "grad_norm": 0.13957668840885162, "learning_rate": 4.8596388542876563e-05, "loss": 0.0086, "step": 53350 }, { "grad_norm": 0.1224505752325058, "learning_rate": 4.857986043811836e-05, "loss": 0.0095, "step": 53360 }, { "grad_norm": 0.1089533120393753, "learning_rate": 4.856333248866448e-05, "loss": 0.0086, "step": 53370 }, { "grad_norm": 0.1527041643857956, "learning_rate": 4.85468046963224e-05, "loss": 0.0084, "step": 53380 }, { "grad_norm": 0.22904202342033386, "learning_rate": 4.853027706289957e-05, "loss": 0.0091, "step": 53390 }, { "grad_norm": 0.12077085673809052, "learning_rate": 4.851374959020344e-05, "loss": 0.0095, "step": 53400 }, { "grad_norm": 0.12229885160923004, "learning_rate": 4.849722228004141e-05, "loss": 0.0077, "step": 53410 }, { "grad_norm": 0.17415182292461395, "learning_rate": 4.848069513422091e-05, "loss": 0.009, "step": 53420 }, { "grad_norm": 0.0887325257062912, "learning_rate": 4.8464168154549294e-05, "loss": 0.0073, "step": 53430 }, { "grad_norm": 0.19001488387584686, "learning_rate": 4.844764134283392e-05, "loss": 0.0083, "step": 53440 }, { "grad_norm": 0.16632112860679626, "learning_rate": 4.843111470088218e-05, "loss": 0.0089, "step": 53450 }, { "grad_norm": 0.1395314335823059, "learning_rate": 4.841458823050137e-05, "loss": 0.0072, "step": 53460 }, { "grad_norm": 0.16728857159614563, "learning_rate": 4.8398061933498816e-05, "loss": 0.0082, "step": 53470 }, { "grad_norm": 0.20129776000976562, "learning_rate": 4.83815358116818e-05, "loss": 0.0082, "step": 53480 }, { "grad_norm": 0.18002310395240784, "learning_rate": 4.8365009866857584e-05, "loss": 0.0083, "step": 53490 }, { "grad_norm": 0.17504067718982697, "learning_rate": 4.834848410083342e-05, "loss": 0.0091, "step": 53500 }, { "grad_norm": 0.1223766878247261, "learning_rate": 4.833195851541657e-05, "loss": 0.0088, "step": 53510 }, { "grad_norm": 0.16666661202907562, "learning_rate": 4.831543311241423e-05, "loss": 0.0115, "step": 53520 }, { "grad_norm": 0.15146702527999878, "learning_rate": 4.829890789363359e-05, "loss": 0.0075, "step": 53530 }, { "grad_norm": 0.14256010949611664, "learning_rate": 4.828238286088182e-05, "loss": 0.0086, "step": 53540 }, { "grad_norm": 0.13975003361701965, "learning_rate": 4.826585801596609e-05, "loss": 0.0093, "step": 53550 }, { "grad_norm": 0.17061635851860046, "learning_rate": 4.8249333360693486e-05, "loss": 0.0094, "step": 53560 }, { "grad_norm": 0.15682083368301392, "learning_rate": 4.8232808896871185e-05, "loss": 0.0093, "step": 53570 }, { "grad_norm": 0.2211519032716751, "learning_rate": 4.821628462630624e-05, "loss": 0.0096, "step": 53580 }, { "grad_norm": 0.12786084413528442, "learning_rate": 4.8199760550805735e-05, "loss": 0.0091, "step": 53590 }, { "grad_norm": 0.18433760106563568, "learning_rate": 4.818323667217669e-05, "loss": 0.009, "step": 53600 }, { "grad_norm": 0.13637298345565796, "learning_rate": 4.8166712992226165e-05, "loss": 0.0094, "step": 53610 }, { "grad_norm": 0.1588038206100464, "learning_rate": 4.8150189512761154e-05, "loss": 0.0083, "step": 53620 }, { "grad_norm": 0.09656491130590439, "learning_rate": 4.813366623558862e-05, "loss": 0.0093, "step": 53630 }, { "grad_norm": 0.20549219846725464, "learning_rate": 4.811714316251554e-05, "loss": 0.0081, "step": 53640 }, { "grad_norm": 0.11717838048934937, "learning_rate": 4.810062029534886e-05, "loss": 0.0081, "step": 53650 }, { "grad_norm": 0.14522749185562134, "learning_rate": 4.8084097635895505e-05, "loss": 0.0109, "step": 53660 }, { "grad_norm": 0.15750335156917572, "learning_rate": 4.8067575185962335e-05, "loss": 0.0088, "step": 53670 }, { "grad_norm": 0.1570606380701065, "learning_rate": 4.805105294735625e-05, "loss": 0.0116, "step": 53680 }, { "grad_norm": 0.13579419255256653, "learning_rate": 4.803453092188406e-05, "loss": 0.0103, "step": 53690 }, { "grad_norm": 0.1395365297794342, "learning_rate": 4.801800911135263e-05, "loss": 0.0083, "step": 53700 }, { "grad_norm": 0.25501930713653564, "learning_rate": 4.8001487517568744e-05, "loss": 0.0094, "step": 53710 }, { "grad_norm": 0.14137864112854004, "learning_rate": 4.798496614233918e-05, "loss": 0.009, "step": 53720 }, { "grad_norm": 0.12144065648317337, "learning_rate": 4.796844498747069e-05, "loss": 0.0082, "step": 53730 }, { "grad_norm": 0.2000894993543625, "learning_rate": 4.795192405477e-05, "loss": 0.0097, "step": 53740 }, { "grad_norm": 0.1475680023431778, "learning_rate": 4.79354033460438e-05, "loss": 0.0082, "step": 53750 }, { "grad_norm": 0.16275426745414734, "learning_rate": 4.79188828630988e-05, "loss": 0.0097, "step": 53760 }, { "grad_norm": 0.15471860766410828, "learning_rate": 4.790236260774165e-05, "loss": 0.0102, "step": 53770 }, { "grad_norm": 0.130527064204216, "learning_rate": 4.788584258177896e-05, "loss": 0.0081, "step": 53780 }, { "grad_norm": 0.11743316799402237, "learning_rate": 4.7869322787017356e-05, "loss": 0.0081, "step": 53790 }, { "grad_norm": 0.11894434690475464, "learning_rate": 4.78528032252634e-05, "loss": 0.0092, "step": 53800 }, { "grad_norm": 0.1533333957195282, "learning_rate": 4.783628389832366e-05, "loss": 0.0091, "step": 53810 }, { "grad_norm": 0.13570071756839752, "learning_rate": 4.7819764808004644e-05, "loss": 0.0082, "step": 53820 }, { "grad_norm": 0.13267946243286133, "learning_rate": 4.780324595611289e-05, "loss": 0.0092, "step": 53830 }, { "grad_norm": 0.08971138298511505, "learning_rate": 4.7786727344454854e-05, "loss": 0.0081, "step": 53840 }, { "grad_norm": 0.1130305752158165, "learning_rate": 4.777020897483697e-05, "loss": 0.0089, "step": 53850 }, { "grad_norm": 0.28634408116340637, "learning_rate": 4.775369084906569e-05, "loss": 0.0075, "step": 53860 }, { "grad_norm": 0.10386945307254791, "learning_rate": 4.773717296894742e-05, "loss": 0.0091, "step": 53870 }, { "grad_norm": 0.16886039078235626, "learning_rate": 4.772065533628848e-05, "loss": 0.0079, "step": 53880 }, { "grad_norm": 0.18742763996124268, "learning_rate": 4.7704137952895265e-05, "loss": 0.0096, "step": 53890 }, { "grad_norm": 0.19433698058128357, "learning_rate": 4.768762082057407e-05, "loss": 0.0098, "step": 53900 }, { "grad_norm": 0.1205001100897789, "learning_rate": 4.7671103941131186e-05, "loss": 0.009, "step": 53910 }, { "grad_norm": 0.14608916640281677, "learning_rate": 4.765458731637287e-05, "loss": 0.0093, "step": 53920 }, { "grad_norm": 0.12033751606941223, "learning_rate": 4.763807094810537e-05, "loss": 0.0089, "step": 53930 }, { "grad_norm": 0.26311835646629333, "learning_rate": 4.762155483813487e-05, "loss": 0.0081, "step": 53940 }, { "grad_norm": 0.18387043476104736, "learning_rate": 4.760503898826754e-05, "loss": 0.0078, "step": 53950 }, { "grad_norm": 0.18504296243190765, "learning_rate": 4.758852340030957e-05, "loss": 0.008, "step": 53960 }, { "grad_norm": 0.13418544828891754, "learning_rate": 4.757200807606705e-05, "loss": 0.0108, "step": 53970 }, { "grad_norm": 0.13468438386917114, "learning_rate": 4.755549301734607e-05, "loss": 0.0087, "step": 53980 }, { "grad_norm": 0.17007996141910553, "learning_rate": 4.753897822595271e-05, "loss": 0.009, "step": 53990 }, { "grad_norm": 0.18537850677967072, "learning_rate": 4.7522463703692994e-05, "loss": 0.0085, "step": 54000 }, { "grad_norm": 0.1322212517261505, "learning_rate": 4.75059494523729e-05, "loss": 0.0086, "step": 54010 }, { "grad_norm": 0.268794447183609, "learning_rate": 4.748943547379843e-05, "loss": 0.0082, "step": 54020 }, { "grad_norm": 0.1206914559006691, "learning_rate": 4.7472921769775535e-05, "loss": 0.0095, "step": 54030 }, { "grad_norm": 0.16224747896194458, "learning_rate": 4.745640834211011e-05, "loss": 0.0087, "step": 54040 }, { "grad_norm": 0.13579140603542328, "learning_rate": 4.7439895192608054e-05, "loss": 0.0084, "step": 54050 }, { "grad_norm": 0.12561434507369995, "learning_rate": 4.7423382323075186e-05, "loss": 0.0084, "step": 54060 }, { "grad_norm": 0.22192347049713135, "learning_rate": 4.740686973531737e-05, "loss": 0.0081, "step": 54070 }, { "grad_norm": 0.17936384677886963, "learning_rate": 4.7390357431140375e-05, "loss": 0.008, "step": 54080 }, { "grad_norm": 0.13688510656356812, "learning_rate": 4.737384541234998e-05, "loss": 0.0096, "step": 54090 }, { "grad_norm": 0.14821042120456696, "learning_rate": 4.735733368075189e-05, "loss": 0.0081, "step": 54100 }, { "grad_norm": 0.11018595844507217, "learning_rate": 4.734082223815182e-05, "loss": 0.0096, "step": 54110 }, { "grad_norm": 0.12093260139226913, "learning_rate": 4.7324311086355436e-05, "loss": 0.0097, "step": 54120 }, { "grad_norm": 0.17331887781620026, "learning_rate": 4.730780022716836e-05, "loss": 0.0095, "step": 54130 }, { "grad_norm": 0.19966396689414978, "learning_rate": 4.729128966239619e-05, "loss": 0.009, "step": 54140 }, { "grad_norm": 0.11152906715869904, "learning_rate": 4.7274779393844536e-05, "loss": 0.0087, "step": 54150 }, { "grad_norm": 0.13110323250293732, "learning_rate": 4.725826942331891e-05, "loss": 0.0088, "step": 54160 }, { "grad_norm": 0.17012599110603333, "learning_rate": 4.724175975262481e-05, "loss": 0.009, "step": 54170 }, { "grad_norm": 0.16876357793807983, "learning_rate": 4.722525038356772e-05, "loss": 0.0071, "step": 54180 }, { "grad_norm": 0.14340434968471527, "learning_rate": 4.7208741317953075e-05, "loss": 0.0073, "step": 54190 }, { "grad_norm": 0.17077387869358063, "learning_rate": 4.7192232557586264e-05, "loss": 0.0087, "step": 54200 }, { "grad_norm": 0.16282768547534943, "learning_rate": 4.717572410427269e-05, "loss": 0.008, "step": 54210 }, { "grad_norm": 0.09344840049743652, "learning_rate": 4.71592159598177e-05, "loss": 0.0087, "step": 54220 }, { "grad_norm": 0.21091113984584808, "learning_rate": 4.714270812602657e-05, "loss": 0.0081, "step": 54230 }, { "grad_norm": 0.14364559948444366, "learning_rate": 4.712620060470458e-05, "loss": 0.0095, "step": 54240 }, { "grad_norm": 0.2605628967285156, "learning_rate": 4.7109693397656986e-05, "loss": 0.0095, "step": 54250 }, { "grad_norm": 0.14108490943908691, "learning_rate": 4.709318650668894e-05, "loss": 0.0078, "step": 54260 }, { "grad_norm": 0.1296284794807434, "learning_rate": 4.7076679933605676e-05, "loss": 0.0092, "step": 54270 }, { "grad_norm": 0.1932850182056427, "learning_rate": 4.70601736802123e-05, "loss": 0.0084, "step": 54280 }, { "grad_norm": 0.18658766150474548, "learning_rate": 4.704366774831391e-05, "loss": 0.0087, "step": 54290 }, { "grad_norm": 0.30058228969573975, "learning_rate": 4.702716213971557e-05, "loss": 0.0098, "step": 54300 }, { "grad_norm": 0.14032237231731415, "learning_rate": 4.70106568562223e-05, "loss": 0.0067, "step": 54310 }, { "grad_norm": 0.11197233945131302, "learning_rate": 4.6994151899639105e-05, "loss": 0.0095, "step": 54320 }, { "grad_norm": 0.09527107328176498, "learning_rate": 4.697764727177093e-05, "loss": 0.009, "step": 54330 }, { "grad_norm": 0.2226254791021347, "learning_rate": 4.696114297442272e-05, "loss": 0.0093, "step": 54340 }, { "grad_norm": 0.13407664000988007, "learning_rate": 4.694463900939933e-05, "loss": 0.0111, "step": 54350 }, { "grad_norm": 0.109332375228405, "learning_rate": 4.692813537850564e-05, "loss": 0.0081, "step": 54360 }, { "grad_norm": 0.1221659928560257, "learning_rate": 4.6911632083546445e-05, "loss": 0.0083, "step": 54370 }, { "grad_norm": 0.12448932230472565, "learning_rate": 4.6895129126326525e-05, "loss": 0.0074, "step": 54380 }, { "grad_norm": 0.13516011834144592, "learning_rate": 4.687862650865059e-05, "loss": 0.0087, "step": 54390 }, { "grad_norm": 0.12409903109073639, "learning_rate": 4.686212423232339e-05, "loss": 0.0094, "step": 54400 }, { "grad_norm": 0.1240624189376831, "learning_rate": 4.6845622299149565e-05, "loss": 0.0077, "step": 54410 }, { "grad_norm": 0.1558169424533844, "learning_rate": 4.682912071093374e-05, "loss": 0.0095, "step": 54420 }, { "grad_norm": 0.13365447521209717, "learning_rate": 4.681261946948052e-05, "loss": 0.0083, "step": 54430 }, { "grad_norm": 0.130596324801445, "learning_rate": 4.679611857659443e-05, "loss": 0.0088, "step": 54440 }, { "grad_norm": 0.17748098075389862, "learning_rate": 4.6779618034079975e-05, "loss": 0.009, "step": 54450 }, { "grad_norm": 0.20166341960430145, "learning_rate": 4.6763117843741675e-05, "loss": 0.0094, "step": 54460 }, { "grad_norm": 0.16771823167800903, "learning_rate": 4.674661800738394e-05, "loss": 0.0089, "step": 54470 }, { "grad_norm": 0.13836443424224854, "learning_rate": 4.673011852681117e-05, "loss": 0.01, "step": 54480 }, { "grad_norm": 0.12190518528223038, "learning_rate": 4.671361940382772e-05, "loss": 0.0086, "step": 54490 }, { "grad_norm": 0.13905395567417145, "learning_rate": 4.669712064023791e-05, "loss": 0.0076, "step": 54500 }, { "grad_norm": 0.12310472130775452, "learning_rate": 4.668062223784602e-05, "loss": 0.0092, "step": 54510 }, { "grad_norm": 0.13560950756072998, "learning_rate": 4.6664124198456274e-05, "loss": 0.0086, "step": 54520 }, { "grad_norm": 0.1552044153213501, "learning_rate": 4.6647626523872905e-05, "loss": 0.0083, "step": 54530 }, { "grad_norm": 0.17920422554016113, "learning_rate": 4.6631129215900066e-05, "loss": 0.0074, "step": 54540 }, { "grad_norm": 0.22536906599998474, "learning_rate": 4.661463227634187e-05, "loss": 0.009, "step": 54550 }, { "grad_norm": 0.16955320537090302, "learning_rate": 4.659813570700238e-05, "loss": 0.0089, "step": 54560 }, { "grad_norm": 0.17992551624774933, "learning_rate": 4.658163950968568e-05, "loss": 0.0082, "step": 54570 }, { "grad_norm": 0.14085829257965088, "learning_rate": 4.6565143686195726e-05, "loss": 0.0074, "step": 54580 }, { "grad_norm": 0.12427600473165512, "learning_rate": 4.654864823833649e-05, "loss": 0.0073, "step": 54590 }, { "grad_norm": 0.14620374143123627, "learning_rate": 4.65321531679119e-05, "loss": 0.0096, "step": 54600 }, { "grad_norm": 0.12801988422870636, "learning_rate": 4.6515658476725834e-05, "loss": 0.0078, "step": 54610 }, { "grad_norm": 0.12828479707241058, "learning_rate": 4.649916416658212e-05, "loss": 0.0096, "step": 54620 }, { "grad_norm": 0.15785376727581024, "learning_rate": 4.648267023928455e-05, "loss": 0.0073, "step": 54630 }, { "grad_norm": 0.10478789359331131, "learning_rate": 4.646617669663687e-05, "loss": 0.0099, "step": 54640 }, { "grad_norm": 0.20656746625900269, "learning_rate": 4.644968354044277e-05, "loss": 0.0085, "step": 54650 }, { "grad_norm": 0.15599538385868073, "learning_rate": 4.643319077250596e-05, "loss": 0.0086, "step": 54660 }, { "grad_norm": 0.15768414735794067, "learning_rate": 4.6416698394630055e-05, "loss": 0.0095, "step": 54670 }, { "grad_norm": 0.15458807349205017, "learning_rate": 4.640020640861862e-05, "loss": 0.0108, "step": 54680 }, { "grad_norm": 0.1928025186061859, "learning_rate": 4.6383714816275195e-05, "loss": 0.0084, "step": 54690 }, { "grad_norm": 0.12391006201505661, "learning_rate": 4.636722361940329e-05, "loss": 0.0083, "step": 54700 }, { "grad_norm": 0.17959649860858917, "learning_rate": 4.635073281980632e-05, "loss": 0.0103, "step": 54710 }, { "grad_norm": 0.12756435573101044, "learning_rate": 4.633424241928775e-05, "loss": 0.0073, "step": 54720 }, { "grad_norm": 0.13127486407756805, "learning_rate": 4.6317752419650914e-05, "loss": 0.0101, "step": 54730 }, { "grad_norm": 0.11191241443157196, "learning_rate": 4.630126282269914e-05, "loss": 0.0082, "step": 54740 }, { "grad_norm": 0.2814042568206787, "learning_rate": 4.628477363023569e-05, "loss": 0.0094, "step": 54750 }, { "grad_norm": 0.14699693024158478, "learning_rate": 4.626828484406383e-05, "loss": 0.0085, "step": 54760 }, { "grad_norm": 0.15490594506263733, "learning_rate": 4.625179646598669e-05, "loss": 0.0093, "step": 54770 }, { "grad_norm": 0.15502575039863586, "learning_rate": 4.6235308497807487e-05, "loss": 0.0092, "step": 54780 }, { "grad_norm": 0.1457812339067459, "learning_rate": 4.6218820941329276e-05, "loss": 0.0078, "step": 54790 }, { "grad_norm": 0.13822343945503235, "learning_rate": 4.620233379835513e-05, "loss": 0.0072, "step": 54800 }, { "grad_norm": 0.13268887996673584, "learning_rate": 4.618584707068804e-05, "loss": 0.0091, "step": 54810 }, { "grad_norm": 0.15085838735103607, "learning_rate": 4.616936076013099e-05, "loss": 0.008, "step": 54820 }, { "grad_norm": 0.1815316379070282, "learning_rate": 4.615287486848688e-05, "loss": 0.0088, "step": 54830 }, { "grad_norm": 0.14905527234077454, "learning_rate": 4.6136389397558584e-05, "loss": 0.0105, "step": 54840 }, { "grad_norm": 0.14076676964759827, "learning_rate": 4.6119904349148935e-05, "loss": 0.0076, "step": 54850 }, { "grad_norm": 0.11967730522155762, "learning_rate": 4.6103419725060726e-05, "loss": 0.0079, "step": 54860 }, { "grad_norm": 0.17350368201732635, "learning_rate": 4.608693552709667e-05, "loss": 0.0085, "step": 54870 }, { "grad_norm": 0.16159027814865112, "learning_rate": 4.607045175705947e-05, "loss": 0.009, "step": 54880 }, { "grad_norm": 0.18705713748931885, "learning_rate": 4.605396841675176e-05, "loss": 0.0104, "step": 54890 }, { "grad_norm": 0.10619789361953735, "learning_rate": 4.6037485507976094e-05, "loss": 0.0111, "step": 54900 }, { "grad_norm": 0.23987972736358643, "learning_rate": 4.602100303253509e-05, "loss": 0.0079, "step": 54910 }, { "grad_norm": 0.12155090272426605, "learning_rate": 4.600452099223122e-05, "loss": 0.0098, "step": 54920 }, { "grad_norm": 0.09759223461151123, "learning_rate": 4.598803938886691e-05, "loss": 0.0076, "step": 54930 }, { "grad_norm": 0.1112922802567482, "learning_rate": 4.597155822424459e-05, "loss": 0.0076, "step": 54940 }, { "grad_norm": 0.163605734705925, "learning_rate": 4.59550775001666e-05, "loss": 0.0115, "step": 54950 }, { "grad_norm": 0.17141222953796387, "learning_rate": 4.5938597218435234e-05, "loss": 0.0081, "step": 54960 }, { "grad_norm": 0.12656712532043457, "learning_rate": 4.592211738085279e-05, "loss": 0.0082, "step": 54970 }, { "grad_norm": 0.12285618484020233, "learning_rate": 4.590563798922146e-05, "loss": 0.0085, "step": 54980 }, { "grad_norm": 0.11252815276384354, "learning_rate": 4.5889159045343404e-05, "loss": 0.0068, "step": 54990 }, { "grad_norm": 0.1341620236635208, "learning_rate": 4.5872680551020734e-05, "loss": 0.0095, "step": 55000 }, { "grad_norm": 0.11494985222816467, "learning_rate": 4.585620250805552e-05, "loss": 0.0085, "step": 55010 }, { "grad_norm": 0.14927278459072113, "learning_rate": 4.583972491824975e-05, "loss": 0.0094, "step": 55020 }, { "grad_norm": 0.4107294976711273, "learning_rate": 4.582324778340541e-05, "loss": 0.008, "step": 55030 }, { "grad_norm": 0.16536350548267365, "learning_rate": 4.5806771105324424e-05, "loss": 0.0098, "step": 55040 }, { "grad_norm": 0.1569008082151413, "learning_rate": 4.579029488580865e-05, "loss": 0.0081, "step": 55050 }, { "grad_norm": 0.14782468974590302, "learning_rate": 4.5773819126659876e-05, "loss": 0.008, "step": 55060 }, { "grad_norm": 0.1603502631187439, "learning_rate": 4.575734382967991e-05, "loss": 0.008, "step": 55070 }, { "grad_norm": 0.14376740157604218, "learning_rate": 4.574086899667044e-05, "loss": 0.0091, "step": 55080 }, { "grad_norm": 0.11827529966831207, "learning_rate": 4.5724394629433106e-05, "loss": 0.0096, "step": 55090 }, { "grad_norm": 0.10638505965471268, "learning_rate": 4.570792072976958e-05, "loss": 0.0075, "step": 55100 }, { "grad_norm": 0.14538393914699554, "learning_rate": 4.569144729948138e-05, "loss": 0.0072, "step": 55110 }, { "grad_norm": 0.11779237538576126, "learning_rate": 4.5674974340370033e-05, "loss": 0.0097, "step": 55120 }, { "grad_norm": 0.09380929917097092, "learning_rate": 4.565850185423699e-05, "loss": 0.0074, "step": 55130 }, { "grad_norm": 0.2613351047039032, "learning_rate": 4.564202984288365e-05, "loss": 0.0081, "step": 55140 }, { "grad_norm": 0.1453246772289276, "learning_rate": 4.5625558308111356e-05, "loss": 0.0088, "step": 55150 }, { "grad_norm": 0.11892881244421005, "learning_rate": 4.560908725172145e-05, "loss": 0.007, "step": 55160 }, { "grad_norm": 0.12988488376140594, "learning_rate": 4.559261667551516e-05, "loss": 0.0093, "step": 55170 }, { "grad_norm": 0.1273387372493744, "learning_rate": 4.5576146581293685e-05, "loss": 0.0092, "step": 55180 }, { "grad_norm": 0.13000471889972687, "learning_rate": 4.5559676970858164e-05, "loss": 0.0101, "step": 55190 }, { "grad_norm": 0.1292789727449417, "learning_rate": 4.5543207846009705e-05, "loss": 0.0096, "step": 55200 }, { "grad_norm": 0.12462325394153595, "learning_rate": 4.552673920854933e-05, "loss": 0.0075, "step": 55210 }, { "grad_norm": 0.18447570502758026, "learning_rate": 4.5510271060278006e-05, "loss": 0.0093, "step": 55220 }, { "grad_norm": 0.21876850724220276, "learning_rate": 4.5493803402996704e-05, "loss": 0.0089, "step": 55230 }, { "grad_norm": 0.1714988499879837, "learning_rate": 4.5477336238506295e-05, "loss": 0.008, "step": 55240 }, { "grad_norm": 0.1202186569571495, "learning_rate": 4.546086956860759e-05, "loss": 0.0094, "step": 55250 }, { "grad_norm": 0.14060144126415253, "learning_rate": 4.5444403395101374e-05, "loss": 0.007, "step": 55260 }, { "grad_norm": 0.162745863199234, "learning_rate": 4.542793771978834e-05, "loss": 0.0077, "step": 55270 }, { "grad_norm": 0.09799814224243164, "learning_rate": 4.541147254446916e-05, "loss": 0.0098, "step": 55280 }, { "grad_norm": 0.16501657664775848, "learning_rate": 4.539500787094445e-05, "loss": 0.0082, "step": 55290 }, { "grad_norm": 0.12488280981779099, "learning_rate": 4.537854370101475e-05, "loss": 0.0091, "step": 55300 }, { "grad_norm": 0.13027380406856537, "learning_rate": 4.5362080036480566e-05, "loss": 0.0099, "step": 55310 }, { "grad_norm": 0.16371530294418335, "learning_rate": 4.534561687914235e-05, "loss": 0.0077, "step": 55320 }, { "grad_norm": 0.1340557336807251, "learning_rate": 4.5329154230800466e-05, "loss": 0.0094, "step": 55330 }, { "grad_norm": 0.14965519309043884, "learning_rate": 4.531269209325525e-05, "loss": 0.0074, "step": 55340 }, { "grad_norm": 0.16930530965328217, "learning_rate": 4.529623046830696e-05, "loss": 0.0091, "step": 55350 }, { "grad_norm": 0.16111618280410767, "learning_rate": 4.5279769357755866e-05, "loss": 0.0078, "step": 55360 }, { "grad_norm": 0.1407073736190796, "learning_rate": 4.5263308763402084e-05, "loss": 0.0086, "step": 55370 }, { "grad_norm": 0.11670058965682983, "learning_rate": 4.524684868704574e-05, "loss": 0.0095, "step": 55380 }, { "grad_norm": 0.19707004725933075, "learning_rate": 4.523038913048688e-05, "loss": 0.0105, "step": 55390 }, { "grad_norm": 0.1666460931301117, "learning_rate": 4.521393009552549e-05, "loss": 0.0107, "step": 55400 }, { "grad_norm": 0.17341506481170654, "learning_rate": 4.5197471583961484e-05, "loss": 0.0095, "step": 55410 }, { "grad_norm": 0.20152604579925537, "learning_rate": 4.51810135975948e-05, "loss": 0.0085, "step": 55420 }, { "grad_norm": 0.18134304881095886, "learning_rate": 4.516455613822521e-05, "loss": 0.0077, "step": 55430 }, { "grad_norm": 0.1351657509803772, "learning_rate": 4.514809920765249e-05, "loss": 0.0067, "step": 55440 }, { "grad_norm": 0.15609556436538696, "learning_rate": 4.5131642807676356e-05, "loss": 0.0074, "step": 55450 }, { "grad_norm": 0.13815514743328094, "learning_rate": 4.511518694009644e-05, "loss": 0.008, "step": 55460 }, { "grad_norm": 0.1485927850008011, "learning_rate": 4.509873160671231e-05, "loss": 0.0075, "step": 55470 }, { "grad_norm": 0.1903068721294403, "learning_rate": 4.508227680932356e-05, "loss": 0.0083, "step": 55480 }, { "grad_norm": 0.12280310690402985, "learning_rate": 4.5065822549729605e-05, "loss": 0.009, "step": 55490 }, { "grad_norm": 0.11732923984527588, "learning_rate": 4.504936882972989e-05, "loss": 0.0096, "step": 55500 }, { "grad_norm": 0.139064759016037, "learning_rate": 4.503291565112374e-05, "loss": 0.0079, "step": 55510 }, { "grad_norm": 0.10460760444402695, "learning_rate": 4.501646301571048e-05, "loss": 0.0091, "step": 55520 }, { "grad_norm": 0.16204684972763062, "learning_rate": 4.500001092528933e-05, "loss": 0.0074, "step": 55530 }, { "grad_norm": 0.1414865106344223, "learning_rate": 4.4983559381659455e-05, "loss": 0.0089, "step": 55540 }, { "grad_norm": 0.11219070106744766, "learning_rate": 4.496710838661999e-05, "loss": 0.0077, "step": 55550 }, { "grad_norm": 0.14052142202854156, "learning_rate": 4.495065794196999e-05, "loss": 0.0071, "step": 55560 }, { "grad_norm": 0.23105858266353607, "learning_rate": 4.4934208049508455e-05, "loss": 0.0098, "step": 55570 }, { "grad_norm": 0.15492752194404602, "learning_rate": 4.4917758711034303e-05, "loss": 0.0088, "step": 55580 }, { "grad_norm": 0.21276846528053284, "learning_rate": 4.490130992834642e-05, "loss": 0.0088, "step": 55590 }, { "grad_norm": 0.2553097903728485, "learning_rate": 4.488486170324359e-05, "loss": 0.0088, "step": 55600 }, { "grad_norm": 0.11926216632127762, "learning_rate": 4.486841403752463e-05, "loss": 0.0118, "step": 55610 }, { "grad_norm": 0.16987445950508118, "learning_rate": 4.4851966932988174e-05, "loss": 0.0098, "step": 55620 }, { "grad_norm": 0.13090500235557556, "learning_rate": 4.4835520391432894e-05, "loss": 0.0088, "step": 55630 }, { "grad_norm": 0.13285328447818756, "learning_rate": 4.481907441465733e-05, "loss": 0.0079, "step": 55640 }, { "grad_norm": 0.09951110929250717, "learning_rate": 4.480262900446e-05, "loss": 0.0088, "step": 55650 }, { "grad_norm": 0.14270387589931488, "learning_rate": 4.478618416263933e-05, "loss": 0.0084, "step": 55660 }, { "grad_norm": 0.188887357711792, "learning_rate": 4.476973989099373e-05, "loss": 0.0098, "step": 55670 }, { "grad_norm": 0.122595876455307, "learning_rate": 4.475329619132153e-05, "loss": 0.009, "step": 55680 }, { "grad_norm": 0.12214609980583191, "learning_rate": 4.4736853065420956e-05, "loss": 0.0092, "step": 55690 }, { "grad_norm": 0.11585326492786407, "learning_rate": 4.472041051509023e-05, "loss": 0.0091, "step": 55700 }, { "grad_norm": 0.23567242920398712, "learning_rate": 4.470396854212746e-05, "loss": 0.0085, "step": 55710 }, { "grad_norm": 0.1371084302663803, "learning_rate": 4.468752714833072e-05, "loss": 0.0071, "step": 55720 }, { "grad_norm": 0.10271555930376053, "learning_rate": 4.467108633549803e-05, "loss": 0.0102, "step": 55730 }, { "grad_norm": 0.12908370792865753, "learning_rate": 4.4654646105427335e-05, "loss": 0.0093, "step": 55740 }, { "grad_norm": 0.12675008177757263, "learning_rate": 4.463820645991651e-05, "loss": 0.0088, "step": 55750 }, { "grad_norm": 0.1081404983997345, "learning_rate": 4.462176740076334e-05, "loss": 0.0079, "step": 55760 }, { "grad_norm": 0.2420990765094757, "learning_rate": 4.4605328929765624e-05, "loss": 0.0082, "step": 55770 }, { "grad_norm": 0.08492676913738251, "learning_rate": 4.458889104872102e-05, "loss": 0.008, "step": 55780 }, { "grad_norm": 0.17798198759555817, "learning_rate": 4.457245375942715e-05, "loss": 0.0078, "step": 55790 }, { "grad_norm": 0.13555565476417542, "learning_rate": 4.455601706368157e-05, "loss": 0.0086, "step": 55800 }, { "grad_norm": 0.12931858003139496, "learning_rate": 4.453958096328179e-05, "loss": 0.0066, "step": 55810 }, { "grad_norm": 0.15950334072113037, "learning_rate": 4.452314546002523e-05, "loss": 0.0096, "step": 55820 }, { "grad_norm": 0.16317377984523773, "learning_rate": 4.4506710555709246e-05, "loss": 0.009, "step": 55830 }, { "grad_norm": 0.128897562623024, "learning_rate": 4.449027625213114e-05, "loss": 0.0081, "step": 55840 }, { "grad_norm": 0.19523268938064575, "learning_rate": 4.4473842551088135e-05, "loss": 0.0078, "step": 55850 }, { "grad_norm": 0.1611240655183792, "learning_rate": 4.445740945437738e-05, "loss": 0.0095, "step": 55860 }, { "grad_norm": 0.20732621848583221, "learning_rate": 4.444097696379602e-05, "loss": 0.0086, "step": 55870 }, { "grad_norm": 0.18095289170742035, "learning_rate": 4.442454508114106e-05, "loss": 0.0076, "step": 55880 }, { "grad_norm": 0.1722019761800766, "learning_rate": 4.4408113808209465e-05, "loss": 0.008, "step": 55890 }, { "grad_norm": 0.11163759231567383, "learning_rate": 4.439168314679813e-05, "loss": 0.0073, "step": 55900 }, { "grad_norm": 0.1908641904592514, "learning_rate": 4.43752530987039e-05, "loss": 0.0084, "step": 55910 }, { "grad_norm": 0.14991723001003265, "learning_rate": 4.4358823665723515e-05, "loss": 0.0086, "step": 55920 }, { "grad_norm": 0.16873425245285034, "learning_rate": 4.434239484965371e-05, "loss": 0.0085, "step": 55930 }, { "grad_norm": 0.16603510081768036, "learning_rate": 4.4325966652291103e-05, "loss": 0.0078, "step": 55940 }, { "grad_norm": 0.16371196508407593, "learning_rate": 4.430953907543225e-05, "loss": 0.0087, "step": 55950 }, { "grad_norm": 0.11709810793399811, "learning_rate": 4.4293112120873645e-05, "loss": 0.0099, "step": 55960 }, { "grad_norm": 0.15565267205238342, "learning_rate": 4.427668579041172e-05, "loss": 0.0073, "step": 55970 }, { "grad_norm": 0.22629135847091675, "learning_rate": 4.4260260085842816e-05, "loss": 0.0083, "step": 55980 }, { "grad_norm": 0.1443723887205124, "learning_rate": 4.4243835008963256e-05, "loss": 0.0087, "step": 55990 }, { "grad_norm": 0.15650928020477295, "learning_rate": 4.422741056156925e-05, "loss": 0.0096, "step": 56000 }, { "grad_norm": 0.19543291628360748, "learning_rate": 4.421098674545693e-05, "loss": 0.0093, "step": 56010 }, { "grad_norm": 0.159671351313591, "learning_rate": 4.419456356242241e-05, "loss": 0.008, "step": 56020 }, { "grad_norm": 0.12427517026662827, "learning_rate": 4.4178141014261685e-05, "loss": 0.0074, "step": 56030 }, { "grad_norm": 0.14965389668941498, "learning_rate": 4.416171910277071e-05, "loss": 0.0112, "step": 56040 }, { "grad_norm": 0.13572371006011963, "learning_rate": 4.414529782974533e-05, "loss": 0.0092, "step": 56050 }, { "grad_norm": 0.15503700077533722, "learning_rate": 4.41288771969814e-05, "loss": 0.0085, "step": 56060 }, { "grad_norm": 0.23629188537597656, "learning_rate": 4.411245720627464e-05, "loss": 0.0103, "step": 56070 }, { "grad_norm": 0.09120530635118484, "learning_rate": 4.40960378594207e-05, "loss": 0.0069, "step": 56080 }, { "grad_norm": 0.10889263451099396, "learning_rate": 4.407961915821519e-05, "loss": 0.0103, "step": 56090 }, { "grad_norm": 0.1924630105495453, "learning_rate": 4.406320110445362e-05, "loss": 0.009, "step": 56100 }, { "grad_norm": 0.15323495864868164, "learning_rate": 4.404678369993144e-05, "loss": 0.0092, "step": 56110 }, { "grad_norm": 0.1677304059267044, "learning_rate": 4.403036694644406e-05, "loss": 0.008, "step": 56120 }, { "grad_norm": 0.15650317072868347, "learning_rate": 4.4013950845786764e-05, "loss": 0.0108, "step": 56130 }, { "grad_norm": 0.1357453316450119, "learning_rate": 4.399753539975482e-05, "loss": 0.0089, "step": 56140 }, { "grad_norm": 0.20577308535575867, "learning_rate": 4.398112061014337e-05, "loss": 0.0094, "step": 56150 }, { "grad_norm": 0.12890036404132843, "learning_rate": 4.396470647874753e-05, "loss": 0.0071, "step": 56160 }, { "grad_norm": 0.1243925616145134, "learning_rate": 4.394829300736229e-05, "loss": 0.0075, "step": 56170 }, { "grad_norm": 0.16488097608089447, "learning_rate": 4.393188019778265e-05, "loss": 0.0089, "step": 56180 }, { "grad_norm": 0.13813193142414093, "learning_rate": 4.391546805180347e-05, "loss": 0.0076, "step": 56190 }, { "grad_norm": 0.144423708319664, "learning_rate": 4.389905657121955e-05, "loss": 0.0087, "step": 56200 }, { "grad_norm": 0.20413874089717865, "learning_rate": 4.3882645757825635e-05, "loss": 0.0097, "step": 56210 }, { "grad_norm": 0.13841727375984192, "learning_rate": 4.386623561341637e-05, "loss": 0.0101, "step": 56220 }, { "grad_norm": 0.17178282141685486, "learning_rate": 4.384982613978637e-05, "loss": 0.0088, "step": 56230 }, { "grad_norm": 0.11539863795042038, "learning_rate": 4.383341733873012e-05, "loss": 0.0079, "step": 56240 }, { "grad_norm": 0.132226824760437, "learning_rate": 4.381700921204209e-05, "loss": 0.0067, "step": 56250 }, { "grad_norm": 0.13821108639240265, "learning_rate": 4.380060176151663e-05, "loss": 0.0095, "step": 56260 }, { "grad_norm": 0.14714966714382172, "learning_rate": 4.378419498894805e-05, "loss": 0.0098, "step": 56270 }, { "grad_norm": 0.12421871721744537, "learning_rate": 4.376778889613056e-05, "loss": 0.0068, "step": 56280 }, { "grad_norm": 0.12431035935878754, "learning_rate": 4.375138348485831e-05, "loss": 0.0081, "step": 56290 }, { "grad_norm": 0.15121829509735107, "learning_rate": 4.3734978756925346e-05, "loss": 0.0095, "step": 56300 }, { "grad_norm": 0.19220133125782013, "learning_rate": 4.3718574714125706e-05, "loss": 0.011, "step": 56310 }, { "grad_norm": 0.15669062733650208, "learning_rate": 4.370217135825329e-05, "loss": 0.01, "step": 56320 }, { "grad_norm": 0.1765974909067154, "learning_rate": 4.368576869110194e-05, "loss": 0.0077, "step": 56330 }, { "grad_norm": 0.1712736189365387, "learning_rate": 4.366936671446544e-05, "loss": 0.0084, "step": 56340 }, { "grad_norm": 0.22517487406730652, "learning_rate": 4.3652965430137474e-05, "loss": 0.0085, "step": 56350 }, { "grad_norm": 0.14330749213695526, "learning_rate": 4.3636564839911646e-05, "loss": 0.0095, "step": 56360 }, { "grad_norm": 0.11523118615150452, "learning_rate": 4.3620164945581545e-05, "loss": 0.0097, "step": 56370 }, { "grad_norm": 0.1363823413848877, "learning_rate": 4.360376574894061e-05, "loss": 0.0069, "step": 56380 }, { "grad_norm": 0.19321149587631226, "learning_rate": 4.358736725178224e-05, "loss": 0.0091, "step": 56390 }, { "grad_norm": 0.21145273745059967, "learning_rate": 4.357096945589974e-05, "loss": 0.0112, "step": 56400 }, { "grad_norm": 0.14712437987327576, "learning_rate": 4.355457236308636e-05, "loss": 0.0083, "step": 56410 }, { "grad_norm": 0.13802571594715118, "learning_rate": 4.353817597513526e-05, "loss": 0.0096, "step": 56420 }, { "grad_norm": 0.12873783707618713, "learning_rate": 4.352178029383948e-05, "loss": 0.0087, "step": 56430 }, { "grad_norm": 0.18040847778320312, "learning_rate": 4.3505385320992105e-05, "loss": 0.0098, "step": 56440 }, { "grad_norm": 0.09510919451713562, "learning_rate": 4.348899105838602e-05, "loss": 0.0087, "step": 56450 }, { "grad_norm": 0.13910077512264252, "learning_rate": 4.3472597507814087e-05, "loss": 0.0083, "step": 56460 }, { "grad_norm": 0.11461598426103592, "learning_rate": 4.3456204671069066e-05, "loss": 0.0086, "step": 56470 }, { "grad_norm": 0.1993628293275833, "learning_rate": 4.343981254994367e-05, "loss": 0.0078, "step": 56480 }, { "grad_norm": 0.16674958169460297, "learning_rate": 4.34234211462305e-05, "loss": 0.0073, "step": 56490 }, { "grad_norm": 0.11173785477876663, "learning_rate": 4.340703046172213e-05, "loss": 0.0076, "step": 56500 }, { "grad_norm": 0.13170017302036285, "learning_rate": 4.339064049821097e-05, "loss": 0.0075, "step": 56510 }, { "grad_norm": 0.16674987971782684, "learning_rate": 4.3374251257489446e-05, "loss": 0.0095, "step": 56520 }, { "grad_norm": 0.14419350028038025, "learning_rate": 4.335786274134984e-05, "loss": 0.008, "step": 56530 }, { "grad_norm": 0.18823494017124176, "learning_rate": 4.334147495158438e-05, "loss": 0.01, "step": 56540 }, { "grad_norm": 0.1446455717086792, "learning_rate": 4.33250878899852e-05, "loss": 0.0084, "step": 56550 }, { "grad_norm": 0.12355005741119385, "learning_rate": 4.330870155834437e-05, "loss": 0.0077, "step": 56560 }, { "grad_norm": 0.1505095213651657, "learning_rate": 4.3292315958453884e-05, "loss": 0.0079, "step": 56570 }, { "grad_norm": 0.25558948516845703, "learning_rate": 4.3275931092105646e-05, "loss": 0.0079, "step": 56580 }, { "grad_norm": 0.09374744445085526, "learning_rate": 4.325954696109147e-05, "loss": 0.0086, "step": 56590 }, { "grad_norm": 0.1324072778224945, "learning_rate": 4.32431635672031e-05, "loss": 0.007, "step": 56600 }, { "grad_norm": 0.09391514956951141, "learning_rate": 4.3226780912232214e-05, "loss": 0.0064, "step": 56610 }, { "grad_norm": 0.1480572670698166, "learning_rate": 4.321039899797036e-05, "loss": 0.0092, "step": 56620 }, { "grad_norm": 0.13314616680145264, "learning_rate": 4.319401782620908e-05, "loss": 0.0084, "step": 56630 }, { "grad_norm": 0.1275075525045395, "learning_rate": 4.317763739873978e-05, "loss": 0.0083, "step": 56640 }, { "grad_norm": 0.1827608048915863, "learning_rate": 4.31612577173538e-05, "loss": 0.0089, "step": 56650 }, { "grad_norm": 0.18425342440605164, "learning_rate": 4.314487878384239e-05, "loss": 0.0096, "step": 56660 }, { "grad_norm": 0.11760075390338898, "learning_rate": 4.3128500599996736e-05, "loss": 0.0094, "step": 56670 }, { "grad_norm": 0.12145499140024185, "learning_rate": 4.31121231676079e-05, "loss": 0.0078, "step": 56680 }, { "grad_norm": 0.10344254970550537, "learning_rate": 4.309574648846694e-05, "loss": 0.0068, "step": 56690 }, { "grad_norm": 0.18095940351486206, "learning_rate": 4.3079370564364755e-05, "loss": 0.0076, "step": 56700 }, { "grad_norm": 0.15912236273288727, "learning_rate": 4.306299539709221e-05, "loss": 0.0082, "step": 56710 }, { "grad_norm": 0.1731131672859192, "learning_rate": 4.3046620988440046e-05, "loss": 0.0079, "step": 56720 }, { "grad_norm": 0.10237330943346024, "learning_rate": 4.303024734019897e-05, "loss": 0.0085, "step": 56730 }, { "grad_norm": 0.13806737959384918, "learning_rate": 4.3013874454159565e-05, "loss": 0.0076, "step": 56740 }, { "grad_norm": 0.13284088671207428, "learning_rate": 4.299750233211233e-05, "loss": 0.0079, "step": 56750 }, { "grad_norm": 0.1447003334760666, "learning_rate": 4.2981130975847715e-05, "loss": 0.0083, "step": 56760 }, { "grad_norm": 0.12229117006063461, "learning_rate": 4.2964760387156076e-05, "loss": 0.0077, "step": 56770 }, { "grad_norm": 0.1822391152381897, "learning_rate": 4.294839056782767e-05, "loss": 0.0074, "step": 56780 }, { "grad_norm": 0.11292477697134018, "learning_rate": 4.293202151965266e-05, "loss": 0.0081, "step": 56790 }, { "grad_norm": 0.12716206908226013, "learning_rate": 4.291565324442115e-05, "loss": 0.0079, "step": 56800 }, { "grad_norm": 0.11142362654209137, "learning_rate": 4.2899285743923135e-05, "loss": 0.0081, "step": 56810 }, { "grad_norm": 0.17638473212718964, "learning_rate": 4.288291901994857e-05, "loss": 0.008, "step": 56820 }, { "grad_norm": 0.17486980557441711, "learning_rate": 4.286655307428729e-05, "loss": 0.0072, "step": 56830 }, { "grad_norm": 0.1253647804260254, "learning_rate": 4.285018790872903e-05, "loss": 0.0088, "step": 56840 }, { "grad_norm": 0.1576671451330185, "learning_rate": 4.283382352506347e-05, "loss": 0.0079, "step": 56850 }, { "grad_norm": 0.1409808248281479, "learning_rate": 4.2817459925080205e-05, "loss": 0.008, "step": 56860 }, { "grad_norm": 0.2402142733335495, "learning_rate": 4.28010971105687e-05, "loss": 0.0081, "step": 56870 }, { "grad_norm": 0.12601882219314575, "learning_rate": 4.278473508331841e-05, "loss": 0.0097, "step": 56880 }, { "grad_norm": 0.1719030737876892, "learning_rate": 4.276837384511864e-05, "loss": 0.0085, "step": 56890 }, { "grad_norm": 0.139625683426857, "learning_rate": 4.275201339775864e-05, "loss": 0.009, "step": 56900 }, { "grad_norm": 0.1244533360004425, "learning_rate": 4.273565374302756e-05, "loss": 0.0071, "step": 56910 }, { "grad_norm": 0.09827951341867447, "learning_rate": 4.2719294882714454e-05, "loss": 0.0079, "step": 56920 }, { "grad_norm": 0.1265052855014801, "learning_rate": 4.270293681860831e-05, "loss": 0.0079, "step": 56930 }, { "grad_norm": 0.1001489982008934, "learning_rate": 4.268657955249802e-05, "loss": 0.0085, "step": 56940 }, { "grad_norm": 0.13714691996574402, "learning_rate": 4.267022308617241e-05, "loss": 0.0087, "step": 56950 }, { "grad_norm": 0.08631332963705063, "learning_rate": 4.265386742142018e-05, "loss": 0.0071, "step": 56960 }, { "grad_norm": 0.22056007385253906, "learning_rate": 4.263751256002996e-05, "loss": 0.0088, "step": 56970 }, { "grad_norm": 0.22922064363956451, "learning_rate": 4.26211585037903e-05, "loss": 0.008, "step": 56980 }, { "grad_norm": 0.10947945713996887, "learning_rate": 4.2604805254489656e-05, "loss": 0.0072, "step": 56990 }, { "grad_norm": 0.1340094804763794, "learning_rate": 4.258845281391638e-05, "loss": 0.0092, "step": 57000 }, { "grad_norm": 0.10802310705184937, "learning_rate": 4.257210118385876e-05, "loss": 0.0077, "step": 57010 }, { "grad_norm": 0.157649427652359, "learning_rate": 4.2555750366104994e-05, "loss": 0.0069, "step": 57020 }, { "grad_norm": 0.11540400981903076, "learning_rate": 4.2539400362443184e-05, "loss": 0.0076, "step": 57030 }, { "grad_norm": 0.1109040379524231, "learning_rate": 4.252305117466133e-05, "loss": 0.0096, "step": 57040 }, { "grad_norm": 0.2406684160232544, "learning_rate": 4.2506702804547356e-05, "loss": 0.0072, "step": 57050 }, { "grad_norm": 0.16429714858531952, "learning_rate": 4.249035525388907e-05, "loss": 0.0074, "step": 57060 }, { "grad_norm": 0.1339990794658661, "learning_rate": 4.247400852447427e-05, "loss": 0.0082, "step": 57070 }, { "grad_norm": 0.12704221904277802, "learning_rate": 4.245766261809059e-05, "loss": 0.0082, "step": 57080 }, { "grad_norm": 0.11249211430549622, "learning_rate": 4.244131753652558e-05, "loss": 0.0086, "step": 57090 }, { "grad_norm": 0.11306658387184143, "learning_rate": 4.242497328156672e-05, "loss": 0.0084, "step": 57100 }, { "grad_norm": 0.13571421802043915, "learning_rate": 4.240862985500139e-05, "loss": 0.0078, "step": 57110 }, { "grad_norm": 0.10969724506139755, "learning_rate": 4.2392287258616894e-05, "loss": 0.0081, "step": 57120 }, { "grad_norm": 0.1250598430633545, "learning_rate": 4.2375945494200396e-05, "loss": 0.008, "step": 57130 }, { "grad_norm": 0.26414474844932556, "learning_rate": 4.2359604563539054e-05, "loss": 0.0072, "step": 57140 }, { "grad_norm": 0.1432221680879593, "learning_rate": 4.2343264468419885e-05, "loss": 0.0077, "step": 57150 }, { "grad_norm": 0.14737369120121002, "learning_rate": 4.232692521062979e-05, "loss": 0.0114, "step": 57160 }, { "grad_norm": 0.11233356595039368, "learning_rate": 4.231058679195562e-05, "loss": 0.0079, "step": 57170 }, { "grad_norm": 0.2116428017616272, "learning_rate": 4.229424921418411e-05, "loss": 0.0082, "step": 57180 }, { "grad_norm": 0.2048584371805191, "learning_rate": 4.2277912479101916e-05, "loss": 0.0075, "step": 57190 }, { "grad_norm": 0.11865995079278946, "learning_rate": 4.226157658849562e-05, "loss": 0.0075, "step": 57200 }, { "grad_norm": 0.14546999335289001, "learning_rate": 4.2245241544151664e-05, "loss": 0.0077, "step": 57210 }, { "grad_norm": 0.13758449256420135, "learning_rate": 4.2228907347856425e-05, "loss": 0.008, "step": 57220 }, { "grad_norm": 0.11075084656476974, "learning_rate": 4.2212574001396214e-05, "loss": 0.0114, "step": 57230 }, { "grad_norm": 0.17473946511745453, "learning_rate": 4.21962415065572e-05, "loss": 0.0068, "step": 57240 }, { "grad_norm": 0.12090478837490082, "learning_rate": 4.217990986512548e-05, "loss": 0.0082, "step": 57250 }, { "grad_norm": 0.13643603026866913, "learning_rate": 4.216357907888704e-05, "loss": 0.0095, "step": 57260 }, { "grad_norm": 0.20831875503063202, "learning_rate": 4.2147249149627824e-05, "loss": 0.007, "step": 57270 }, { "grad_norm": 0.1957690417766571, "learning_rate": 4.213092007913364e-05, "loss": 0.0102, "step": 57280 }, { "grad_norm": 0.2576158344745636, "learning_rate": 4.2114591869190205e-05, "loss": 0.0073, "step": 57290 }, { "grad_norm": 0.1045396700501442, "learning_rate": 4.209826452158315e-05, "loss": 0.008, "step": 57300 }, { "grad_norm": 0.10710926353931427, "learning_rate": 4.2081938038098e-05, "loss": 0.0083, "step": 57310 }, { "grad_norm": 0.21811988949775696, "learning_rate": 4.2065612420520185e-05, "loss": 0.0071, "step": 57320 }, { "grad_norm": 0.12045392394065857, "learning_rate": 4.204928767063509e-05, "loss": 0.0072, "step": 57330 }, { "grad_norm": 0.29097801446914673, "learning_rate": 4.2032963790227944e-05, "loss": 0.0094, "step": 57340 }, { "grad_norm": 0.1260826736688614, "learning_rate": 4.2016640781083887e-05, "loss": 0.0082, "step": 57350 }, { "grad_norm": 0.15475508570671082, "learning_rate": 4.2000318644988004e-05, "loss": 0.0104, "step": 57360 }, { "grad_norm": 0.1455080360174179, "learning_rate": 4.198399738372524e-05, "loss": 0.0105, "step": 57370 }, { "grad_norm": 0.13392342627048492, "learning_rate": 4.196767699908045e-05, "loss": 0.0078, "step": 57380 }, { "grad_norm": 0.18224354088306427, "learning_rate": 4.195135749283845e-05, "loss": 0.0087, "step": 57390 }, { "grad_norm": 0.1574099063873291, "learning_rate": 4.1935038866783894e-05, "loss": 0.0096, "step": 57400 }, { "grad_norm": 0.17596174776554108, "learning_rate": 4.1918721122701355e-05, "loss": 0.0101, "step": 57410 }, { "grad_norm": 0.16990616917610168, "learning_rate": 4.190240426237533e-05, "loss": 0.0088, "step": 57420 }, { "grad_norm": 0.12703202664852142, "learning_rate": 4.188608828759019e-05, "loss": 0.0081, "step": 57430 }, { "grad_norm": 0.17506466805934906, "learning_rate": 4.1869773200130246e-05, "loss": 0.0077, "step": 57440 }, { "grad_norm": 0.1403406262397766, "learning_rate": 4.185345900177967e-05, "loss": 0.0093, "step": 57450 }, { "grad_norm": 0.12995514273643494, "learning_rate": 4.183714569432258e-05, "loss": 0.0072, "step": 57460 }, { "grad_norm": 0.1386106312274933, "learning_rate": 4.182083327954295e-05, "loss": 0.0072, "step": 57470 }, { "grad_norm": 0.13008761405944824, "learning_rate": 4.1804521759224714e-05, "loss": 0.0084, "step": 57480 }, { "grad_norm": 0.13499759137630463, "learning_rate": 4.178821113515165e-05, "loss": 0.0081, "step": 57490 }, { "grad_norm": 0.20450080931186676, "learning_rate": 4.1771901409107476e-05, "loss": 0.0075, "step": 57500 }, { "grad_norm": 0.1367075890302658, "learning_rate": 4.175559258287577e-05, "loss": 0.0078, "step": 57510 }, { "grad_norm": 0.1669093519449234, "learning_rate": 4.173928465824008e-05, "loss": 0.0075, "step": 57520 }, { "grad_norm": 0.12626539170742035, "learning_rate": 4.17229776369838e-05, "loss": 0.0077, "step": 57530 }, { "grad_norm": 0.16167698800563812, "learning_rate": 4.170667152089025e-05, "loss": 0.0072, "step": 57540 }, { "grad_norm": 0.16444477438926697, "learning_rate": 4.1690366311742626e-05, "loss": 0.0079, "step": 57550 }, { "grad_norm": 0.13313205540180206, "learning_rate": 4.1674062011324055e-05, "loss": 0.0095, "step": 57560 }, { "grad_norm": 0.11025974154472351, "learning_rate": 4.165775862141752e-05, "loss": 0.0069, "step": 57570 }, { "grad_norm": 0.11792013794183731, "learning_rate": 4.164145614380598e-05, "loss": 0.0064, "step": 57580 }, { "grad_norm": 0.10183433443307877, "learning_rate": 4.162515458027223e-05, "loss": 0.008, "step": 57590 }, { "grad_norm": 0.11408671736717224, "learning_rate": 4.160885393259899e-05, "loss": 0.0074, "step": 57600 }, { "grad_norm": 0.10943051427602768, "learning_rate": 4.159255420256886e-05, "loss": 0.0076, "step": 57610 }, { "grad_norm": 0.18970578908920288, "learning_rate": 4.157625539196436e-05, "loss": 0.0076, "step": 57620 }, { "grad_norm": 0.16139256954193115, "learning_rate": 4.155995750256791e-05, "loss": 0.0066, "step": 57630 }, { "grad_norm": 0.11965356022119522, "learning_rate": 4.15436605361618e-05, "loss": 0.0075, "step": 57640 }, { "grad_norm": 0.11551733314990997, "learning_rate": 4.152736449452827e-05, "loss": 0.0064, "step": 57650 }, { "grad_norm": 0.2766694724559784, "learning_rate": 4.151106937944943e-05, "loss": 0.0075, "step": 57660 }, { "grad_norm": 0.09859995543956757, "learning_rate": 4.149477519270727e-05, "loss": 0.0073, "step": 57670 }, { "grad_norm": 0.14171884953975677, "learning_rate": 4.147848193608371e-05, "loss": 0.0074, "step": 57680 }, { "grad_norm": 0.12579332292079926, "learning_rate": 4.146218961136056e-05, "loss": 0.0073, "step": 57690 }, { "grad_norm": 0.1441735178232193, "learning_rate": 4.14458982203195e-05, "loss": 0.0086, "step": 57700 }, { "grad_norm": 0.13374000787734985, "learning_rate": 4.1429607764742165e-05, "loss": 0.0073, "step": 57710 }, { "grad_norm": 0.19597938656806946, "learning_rate": 4.141331824641003e-05, "loss": 0.0067, "step": 57720 }, { "grad_norm": 0.15156015753746033, "learning_rate": 4.139702966710452e-05, "loss": 0.0076, "step": 57730 }, { "grad_norm": 0.09239346534013748, "learning_rate": 4.13807420286069e-05, "loss": 0.0076, "step": 57740 }, { "grad_norm": 0.11744895577430725, "learning_rate": 4.136445533269837e-05, "loss": 0.0077, "step": 57750 }, { "grad_norm": 0.19858792424201965, "learning_rate": 4.134816958116003e-05, "loss": 0.0075, "step": 57760 }, { "grad_norm": 0.1260213404893875, "learning_rate": 4.133188477577283e-05, "loss": 0.0078, "step": 57770 }, { "grad_norm": 0.19501517713069916, "learning_rate": 4.131560091831771e-05, "loss": 0.0065, "step": 57780 }, { "grad_norm": 0.14094902575016022, "learning_rate": 4.1299318010575405e-05, "loss": 0.0087, "step": 57790 }, { "grad_norm": 0.1262517273426056, "learning_rate": 4.128303605432661e-05, "loss": 0.008, "step": 57800 }, { "grad_norm": 0.10207552462816238, "learning_rate": 4.126675505135188e-05, "loss": 0.007, "step": 57810 }, { "grad_norm": 0.2624349594116211, "learning_rate": 4.125047500343169e-05, "loss": 0.0072, "step": 57820 }, { "grad_norm": 0.176225945353508, "learning_rate": 4.1234195912346376e-05, "loss": 0.0087, "step": 57830 }, { "grad_norm": 0.18865148723125458, "learning_rate": 4.1217917779876235e-05, "loss": 0.0083, "step": 57840 }, { "grad_norm": 0.15993182361125946, "learning_rate": 4.120164060780141e-05, "loss": 0.0097, "step": 57850 }, { "grad_norm": 0.10305193066596985, "learning_rate": 4.1185364397901935e-05, "loss": 0.0078, "step": 57860 }, { "grad_norm": 0.114278145134449, "learning_rate": 4.116908915195776e-05, "loss": 0.0066, "step": 57870 }, { "grad_norm": 0.16119445860385895, "learning_rate": 4.115281487174872e-05, "loss": 0.0068, "step": 57880 }, { "grad_norm": 0.13927239179611206, "learning_rate": 4.113654155905452e-05, "loss": 0.0081, "step": 57890 }, { "grad_norm": 0.14275221526622772, "learning_rate": 4.112026921565483e-05, "loss": 0.0083, "step": 57900 }, { "grad_norm": 0.11016812920570374, "learning_rate": 4.1103997843329147e-05, "loss": 0.0066, "step": 57910 }, { "grad_norm": 0.12059733271598816, "learning_rate": 4.108772744385689e-05, "loss": 0.0081, "step": 57920 }, { "grad_norm": 0.13027557730674744, "learning_rate": 4.1071458019017346e-05, "loss": 0.0075, "step": 57930 }, { "grad_norm": 0.10833876579999924, "learning_rate": 4.1055189570589734e-05, "loss": 0.0075, "step": 57940 }, { "grad_norm": 0.09331879019737244, "learning_rate": 4.103892210035315e-05, "loss": 0.0069, "step": 57950 }, { "grad_norm": 0.11081445962190628, "learning_rate": 4.102265561008656e-05, "loss": 0.0077, "step": 57960 }, { "grad_norm": 0.1498682051897049, "learning_rate": 4.100639010156885e-05, "loss": 0.008, "step": 57970 }, { "grad_norm": 0.14192622900009155, "learning_rate": 4.099012557657882e-05, "loss": 0.0087, "step": 57980 }, { "grad_norm": 0.1708797663450241, "learning_rate": 4.09738620368951e-05, "loss": 0.0082, "step": 57990 }, { "grad_norm": 0.13343743979930878, "learning_rate": 4.095759948429626e-05, "loss": 0.0077, "step": 58000 }, { "grad_norm": 0.11418557167053223, "learning_rate": 4.094133792056075e-05, "loss": 0.0075, "step": 58010 }, { "grad_norm": 0.12896673381328583, "learning_rate": 4.092507734746688e-05, "loss": 0.0079, "step": 58020 }, { "grad_norm": 0.14450496435165405, "learning_rate": 4.090881776679293e-05, "loss": 0.0068, "step": 58030 }, { "grad_norm": 0.18577434122562408, "learning_rate": 4.0892559180316994e-05, "loss": 0.0084, "step": 58040 }, { "grad_norm": 0.10992246121168137, "learning_rate": 4.08763015898171e-05, "loss": 0.0089, "step": 58050 }, { "grad_norm": 0.1344933807849884, "learning_rate": 4.0860044997071145e-05, "loss": 0.0094, "step": 58060 }, { "grad_norm": 0.11658282577991486, "learning_rate": 4.084378940385693e-05, "loss": 0.0089, "step": 58070 }, { "grad_norm": 0.1715407371520996, "learning_rate": 4.082753481195211e-05, "loss": 0.0069, "step": 58080 }, { "grad_norm": 0.10695135593414307, "learning_rate": 4.0811281223134324e-05, "loss": 0.0073, "step": 58090 }, { "grad_norm": 0.11167263239622116, "learning_rate": 4.0795028639181e-05, "loss": 0.011, "step": 58100 }, { "grad_norm": 0.15650010108947754, "learning_rate": 4.077877706186951e-05, "loss": 0.0072, "step": 58110 }, { "grad_norm": 0.13357797265052795, "learning_rate": 4.076252649297709e-05, "loss": 0.0081, "step": 58120 }, { "grad_norm": 0.08861033618450165, "learning_rate": 4.07462769342809e-05, "loss": 0.0081, "step": 58130 }, { "grad_norm": 0.1305462121963501, "learning_rate": 4.073002838755793e-05, "loss": 0.0074, "step": 58140 }, { "grad_norm": 0.14861497282981873, "learning_rate": 4.0713780854585115e-05, "loss": 0.0076, "step": 58150 }, { "grad_norm": 0.1371157467365265, "learning_rate": 4.069753433713927e-05, "loss": 0.0112, "step": 58160 }, { "grad_norm": 0.18287651240825653, "learning_rate": 4.068128883699709e-05, "loss": 0.0085, "step": 58170 }, { "grad_norm": 0.1679590344429016, "learning_rate": 4.0665044355935155e-05, "loss": 0.0082, "step": 58180 }, { "grad_norm": 0.1140279769897461, "learning_rate": 4.064880089572993e-05, "loss": 0.0096, "step": 58190 }, { "grad_norm": 0.1199360340833664, "learning_rate": 4.063255845815779e-05, "loss": 0.0099, "step": 58200 }, { "grad_norm": 0.1225115954875946, "learning_rate": 4.0616317044994945e-05, "loss": 0.0092, "step": 58210 }, { "grad_norm": 0.18170617520809174, "learning_rate": 4.0600076658017585e-05, "loss": 0.008, "step": 58220 }, { "grad_norm": 0.16547349095344543, "learning_rate": 4.0583837299001706e-05, "loss": 0.0082, "step": 58230 }, { "grad_norm": 0.11589524894952774, "learning_rate": 4.056759896972322e-05, "loss": 0.008, "step": 58240 }, { "grad_norm": 0.08561193943023682, "learning_rate": 4.055136167195793e-05, "loss": 0.0098, "step": 58250 }, { "grad_norm": 0.11428587138652802, "learning_rate": 4.053512540748152e-05, "loss": 0.0067, "step": 58260 }, { "grad_norm": 0.10005922615528107, "learning_rate": 4.051889017806955e-05, "loss": 0.0066, "step": 58270 }, { "grad_norm": 0.11534301191568375, "learning_rate": 4.0502655985497516e-05, "loss": 0.0066, "step": 58280 }, { "grad_norm": 0.16261470317840576, "learning_rate": 4.048642283154074e-05, "loss": 0.0074, "step": 58290 }, { "grad_norm": 0.13763022422790527, "learning_rate": 4.0470190717974464e-05, "loss": 0.0091, "step": 58300 }, { "grad_norm": 0.12249232083559036, "learning_rate": 4.045395964657381e-05, "loss": 0.009, "step": 58310 }, { "grad_norm": 0.21006087958812714, "learning_rate": 4.043772961911376e-05, "loss": 0.0073, "step": 58320 }, { "grad_norm": 0.14739283919334412, "learning_rate": 4.042150063736924e-05, "loss": 0.0077, "step": 58330 }, { "grad_norm": 0.10323545336723328, "learning_rate": 4.0405272703114976e-05, "loss": 0.0072, "step": 58340 }, { "grad_norm": 0.09191632270812988, "learning_rate": 4.038904581812569e-05, "loss": 0.0075, "step": 58350 }, { "grad_norm": 0.13485760986804962, "learning_rate": 4.03728199841759e-05, "loss": 0.0078, "step": 58360 }, { "grad_norm": 0.14947625994682312, "learning_rate": 4.0356595203040044e-05, "loss": 0.0081, "step": 58370 }, { "grad_norm": 0.14434964954853058, "learning_rate": 4.0340371476492426e-05, "loss": 0.0072, "step": 58380 }, { "grad_norm": 0.11355685442686081, "learning_rate": 4.032414880630727e-05, "loss": 0.0069, "step": 58390 }, { "grad_norm": 0.1319607049226761, "learning_rate": 4.0307927194258646e-05, "loss": 0.0082, "step": 58400 }, { "grad_norm": 0.15012697875499725, "learning_rate": 4.029170664212054e-05, "loss": 0.009, "step": 58410 }, { "grad_norm": 0.10541466623544693, "learning_rate": 4.0275487151666794e-05, "loss": 0.0076, "step": 58420 }, { "grad_norm": 0.1330631822347641, "learning_rate": 4.025926872467116e-05, "loss": 0.0078, "step": 58430 }, { "grad_norm": 0.09710516035556793, "learning_rate": 4.024305136290725e-05, "loss": 0.0083, "step": 58440 }, { "grad_norm": 0.1148296445608139, "learning_rate": 4.022683506814858e-05, "loss": 0.007, "step": 58450 }, { "grad_norm": 0.11802345514297485, "learning_rate": 4.021061984216853e-05, "loss": 0.0086, "step": 58460 }, { "grad_norm": 0.148759126663208, "learning_rate": 4.019440568674035e-05, "loss": 0.0071, "step": 58470 }, { "grad_norm": 0.08937596529722214, "learning_rate": 4.017819260363724e-05, "loss": 0.0088, "step": 58480 }, { "grad_norm": 0.13577455282211304, "learning_rate": 4.016198059463223e-05, "loss": 0.0064, "step": 58490 }, { "grad_norm": 0.10888323187828064, "learning_rate": 4.0145769661498233e-05, "loss": 0.007, "step": 58500 }, { "grad_norm": 0.16383859515190125, "learning_rate": 4.012955980600804e-05, "loss": 0.0075, "step": 58510 }, { "grad_norm": 0.13680598139762878, "learning_rate": 4.0113351029934355e-05, "loss": 0.0064, "step": 58520 }, { "grad_norm": 0.1580231636762619, "learning_rate": 4.0097143335049714e-05, "loss": 0.0086, "step": 58530 }, { "grad_norm": 0.17196118831634521, "learning_rate": 4.0080936723126606e-05, "loss": 0.01, "step": 58540 }, { "grad_norm": 0.10192453861236572, "learning_rate": 4.006473119593735e-05, "loss": 0.0084, "step": 58550 }, { "grad_norm": 0.116146020591259, "learning_rate": 4.004852675525415e-05, "loss": 0.0071, "step": 58560 }, { "grad_norm": 0.18580086529254913, "learning_rate": 4.00323234028491e-05, "loss": 0.0093, "step": 58570 }, { "grad_norm": 0.14564044773578644, "learning_rate": 4.0016121140494186e-05, "loss": 0.0087, "step": 58580 }, { "grad_norm": 0.16567951440811157, "learning_rate": 3.999991996996122e-05, "loss": 0.0067, "step": 58590 }, { "grad_norm": 0.167337566614151, "learning_rate": 3.9983719893022e-05, "loss": 0.0073, "step": 58600 }, { "grad_norm": 0.14138652384281158, "learning_rate": 3.996752091144812e-05, "loss": 0.0088, "step": 58610 }, { "grad_norm": 0.1599665731191635, "learning_rate": 3.9951323027011055e-05, "loss": 0.0076, "step": 58620 }, { "grad_norm": 0.10104970633983612, "learning_rate": 3.993512624148219e-05, "loss": 0.009, "step": 58630 }, { "grad_norm": 0.152579665184021, "learning_rate": 3.99189305566328e-05, "loss": 0.0071, "step": 58640 }, { "grad_norm": 0.1274402141571045, "learning_rate": 3.990273597423401e-05, "loss": 0.0064, "step": 58650 }, { "grad_norm": 0.10048240423202515, "learning_rate": 3.988654249605681e-05, "loss": 0.0083, "step": 58660 }, { "grad_norm": 0.11564268171787262, "learning_rate": 3.987035012387213e-05, "loss": 0.0088, "step": 58670 }, { "grad_norm": 0.20625168085098267, "learning_rate": 3.9854158859450735e-05, "loss": 0.0093, "step": 58680 }, { "grad_norm": 0.15912652015686035, "learning_rate": 3.983796870456328e-05, "loss": 0.0083, "step": 58690 }, { "grad_norm": 0.1293436884880066, "learning_rate": 3.982177966098028e-05, "loss": 0.0071, "step": 58700 }, { "grad_norm": 0.1552901268005371, "learning_rate": 3.980559173047216e-05, "loss": 0.0071, "step": 58710 }, { "grad_norm": 0.13629360496997833, "learning_rate": 3.9789404914809176e-05, "loss": 0.0073, "step": 58720 }, { "grad_norm": 0.15224657952785492, "learning_rate": 3.9773219215761544e-05, "loss": 0.0069, "step": 58730 }, { "grad_norm": 0.09972003847360611, "learning_rate": 3.975703463509928e-05, "loss": 0.0083, "step": 58740 }, { "grad_norm": 0.13268357515335083, "learning_rate": 3.974085117459231e-05, "loss": 0.007, "step": 58750 }, { "grad_norm": 0.15273840725421906, "learning_rate": 3.972466883601043e-05, "loss": 0.0075, "step": 58760 }, { "grad_norm": 0.1289534717798233, "learning_rate": 3.9708487621123316e-05, "loss": 0.0085, "step": 58770 }, { "grad_norm": 0.12853263318538666, "learning_rate": 3.969230753170051e-05, "loss": 0.008, "step": 58780 }, { "grad_norm": 0.1027332991361618, "learning_rate": 3.967612856951146e-05, "loss": 0.008, "step": 58790 }, { "grad_norm": 0.18642514944076538, "learning_rate": 3.9659950736325476e-05, "loss": 0.0096, "step": 58800 }, { "grad_norm": 0.1569809764623642, "learning_rate": 3.964377403391174e-05, "loss": 0.0087, "step": 58810 }, { "grad_norm": 0.09862605482339859, "learning_rate": 3.9627598464039303e-05, "loss": 0.0081, "step": 58820 }, { "grad_norm": 0.1269240528345108, "learning_rate": 3.9611424028477096e-05, "loss": 0.0072, "step": 58830 }, { "grad_norm": 0.14055678248405457, "learning_rate": 3.959525072899394e-05, "loss": 0.0082, "step": 58840 }, { "grad_norm": 0.2176840752363205, "learning_rate": 3.9579078567358506e-05, "loss": 0.0086, "step": 58850 }, { "grad_norm": 0.19189396500587463, "learning_rate": 3.956290754533939e-05, "loss": 0.0062, "step": 58860 }, { "grad_norm": 0.10504741221666336, "learning_rate": 3.954673766470501e-05, "loss": 0.0056, "step": 58870 }, { "grad_norm": 0.15382792055606842, "learning_rate": 3.9530568927223674e-05, "loss": 0.0082, "step": 58880 }, { "grad_norm": 0.13892944157123566, "learning_rate": 3.95144013346636e-05, "loss": 0.0099, "step": 58890 }, { "grad_norm": 0.16997933387756348, "learning_rate": 3.949823488879282e-05, "loss": 0.0068, "step": 58900 }, { "grad_norm": 0.09370630234479904, "learning_rate": 3.9482069591379276e-05, "loss": 0.0079, "step": 58910 }, { "grad_norm": 0.14252574741840363, "learning_rate": 3.9465905444190784e-05, "loss": 0.0074, "step": 58920 }, { "grad_norm": 0.12073177844285965, "learning_rate": 3.944974244899505e-05, "loss": 0.0088, "step": 58930 }, { "grad_norm": 0.11046915501356125, "learning_rate": 3.9433580607559616e-05, "loss": 0.0086, "step": 58940 }, { "grad_norm": 0.17524491250514984, "learning_rate": 3.941741992165193e-05, "loss": 0.0085, "step": 58950 }, { "grad_norm": 0.12313022464513779, "learning_rate": 3.940126039303928e-05, "loss": 0.0071, "step": 58960 }, { "grad_norm": 0.1608356386423111, "learning_rate": 3.938510202348886e-05, "loss": 0.0064, "step": 58970 }, { "grad_norm": 0.13422057032585144, "learning_rate": 3.93689448147677e-05, "loss": 0.0067, "step": 58980 }, { "grad_norm": 0.12122396379709244, "learning_rate": 3.935278876864278e-05, "loss": 0.0072, "step": 58990 }, { "grad_norm": 0.106974296271801, "learning_rate": 3.933663388688087e-05, "loss": 0.0061, "step": 59000 }, { "grad_norm": 0.11333934217691422, "learning_rate": 3.9320480171248645e-05, "loss": 0.0077, "step": 59010 }, { "grad_norm": 0.13485600054264069, "learning_rate": 3.930432762351265e-05, "loss": 0.0074, "step": 59020 }, { "grad_norm": 0.31576791405677795, "learning_rate": 3.928817624543931e-05, "loss": 0.0072, "step": 59030 }, { "grad_norm": 0.09422265738248825, "learning_rate": 3.9272026038794885e-05, "loss": 0.0087, "step": 59040 }, { "grad_norm": 0.15917891263961792, "learning_rate": 3.9255877005345586e-05, "loss": 0.007, "step": 59050 }, { "grad_norm": 0.13904766738414764, "learning_rate": 3.923972914685743e-05, "loss": 0.0087, "step": 59060 }, { "grad_norm": 0.16448241472244263, "learning_rate": 3.922358246509631e-05, "loss": 0.0071, "step": 59070 }, { "grad_norm": 0.12280980497598648, "learning_rate": 3.920743696182801e-05, "loss": 0.0095, "step": 59080 }, { "grad_norm": 0.125627338886261, "learning_rate": 3.9191292638818164e-05, "loss": 0.0075, "step": 59090 }, { "grad_norm": 0.12071317434310913, "learning_rate": 3.9175149497832316e-05, "loss": 0.0064, "step": 59100 }, { "grad_norm": 0.16719022393226624, "learning_rate": 3.915900754063584e-05, "loss": 0.0072, "step": 59110 }, { "grad_norm": 0.11379022151231766, "learning_rate": 3.9142866768994014e-05, "loss": 0.0059, "step": 59120 }, { "grad_norm": 0.13841095566749573, "learning_rate": 3.912672718467193e-05, "loss": 0.0076, "step": 59130 }, { "grad_norm": 0.11315363645553589, "learning_rate": 3.911058878943463e-05, "loss": 0.0081, "step": 59140 }, { "grad_norm": 0.21572360396385193, "learning_rate": 3.9094451585046974e-05, "loss": 0.0062, "step": 59150 }, { "grad_norm": 0.12332277745008469, "learning_rate": 3.90783155732737e-05, "loss": 0.0075, "step": 59160 }, { "grad_norm": 0.18732966482639313, "learning_rate": 3.9062180755879373e-05, "loss": 0.0082, "step": 59170 }, { "grad_norm": 0.147126242518425, "learning_rate": 3.904604713462855e-05, "loss": 0.0075, "step": 59180 }, { "grad_norm": 0.09784648567438126, "learning_rate": 3.902991471128554e-05, "loss": 0.0074, "step": 59190 }, { "grad_norm": 0.14688339829444885, "learning_rate": 3.901378348761456e-05, "loss": 0.0109, "step": 59200 }, { "grad_norm": 0.13375407457351685, "learning_rate": 3.899765346537971e-05, "loss": 0.0087, "step": 59210 }, { "grad_norm": 0.11987903714179993, "learning_rate": 3.898152464634493e-05, "loss": 0.0072, "step": 59220 }, { "grad_norm": 0.21464137732982635, "learning_rate": 3.896539703227402e-05, "loss": 0.0074, "step": 59230 }, { "grad_norm": 0.27281442284584045, "learning_rate": 3.894927062493073e-05, "loss": 0.0064, "step": 59240 }, { "grad_norm": 0.07539458572864532, "learning_rate": 3.893314542607859e-05, "loss": 0.007, "step": 59250 }, { "grad_norm": 0.08581376820802689, "learning_rate": 3.891702143748103e-05, "loss": 0.0066, "step": 59260 }, { "grad_norm": 0.13809429109096527, "learning_rate": 3.890089866090133e-05, "loss": 0.0092, "step": 59270 }, { "grad_norm": 0.11037803441286087, "learning_rate": 3.888477709810267e-05, "loss": 0.0072, "step": 59280 }, { "grad_norm": 0.14086446166038513, "learning_rate": 3.886865675084806e-05, "loss": 0.009, "step": 59290 }, { "grad_norm": 0.14562839269638062, "learning_rate": 3.885253762090043e-05, "loss": 0.0058, "step": 59300 }, { "grad_norm": 0.10730338841676712, "learning_rate": 3.8836419710022534e-05, "loss": 0.0074, "step": 59310 }, { "grad_norm": 0.1479569673538208, "learning_rate": 3.882030301997698e-05, "loss": 0.0076, "step": 59320 }, { "grad_norm": 0.1260756254196167, "learning_rate": 3.880418755252629e-05, "loss": 0.0076, "step": 59330 }, { "grad_norm": 0.14721280336380005, "learning_rate": 3.87880733094328e-05, "loss": 0.0081, "step": 59340 }, { "grad_norm": 0.1839635819196701, "learning_rate": 3.877196029245877e-05, "loss": 0.0079, "step": 59350 }, { "grad_norm": 0.1639978140592575, "learning_rate": 3.875584850336627e-05, "loss": 0.0084, "step": 59360 }, { "grad_norm": 0.15877355635166168, "learning_rate": 3.873973794391728e-05, "loss": 0.0072, "step": 59370 }, { "grad_norm": 0.22367604076862335, "learning_rate": 3.872362861587361e-05, "loss": 0.0067, "step": 59380 }, { "grad_norm": 0.15426823496818542, "learning_rate": 3.870752052099698e-05, "loss": 0.0068, "step": 59390 }, { "grad_norm": 0.14716960489749908, "learning_rate": 3.8691413661048916e-05, "loss": 0.0088, "step": 59400 }, { "grad_norm": 0.1685185730457306, "learning_rate": 3.867530803779085e-05, "loss": 0.009, "step": 59410 }, { "grad_norm": 0.15693753957748413, "learning_rate": 3.8659203652984054e-05, "loss": 0.0086, "step": 59420 }, { "grad_norm": 0.09625164419412613, "learning_rate": 3.864310050838972e-05, "loss": 0.0059, "step": 59430 }, { "grad_norm": 0.24108554422855377, "learning_rate": 3.8626998605768835e-05, "loss": 0.0075, "step": 59440 }, { "grad_norm": 0.15785227715969086, "learning_rate": 3.8610897946882294e-05, "loss": 0.0072, "step": 59450 }, { "grad_norm": 0.12706241011619568, "learning_rate": 3.859479853349082e-05, "loss": 0.0068, "step": 59460 }, { "grad_norm": 0.13774971663951874, "learning_rate": 3.857870036735503e-05, "loss": 0.0087, "step": 59470 }, { "grad_norm": 0.17539069056510925, "learning_rate": 3.856260345023538e-05, "loss": 0.0096, "step": 59480 }, { "grad_norm": 0.11878449469804764, "learning_rate": 3.854650778389224e-05, "loss": 0.0076, "step": 59490 }, { "grad_norm": 0.12319281697273254, "learning_rate": 3.8530413370085786e-05, "loss": 0.0081, "step": 59500 }, { "grad_norm": 0.10572129487991333, "learning_rate": 3.851432021057608e-05, "loss": 0.0072, "step": 59510 }, { "grad_norm": 0.12595266103744507, "learning_rate": 3.849822830712306e-05, "loss": 0.0085, "step": 59520 }, { "grad_norm": 0.1398131400346756, "learning_rate": 3.848213766148649e-05, "loss": 0.0085, "step": 59530 }, { "grad_norm": 0.10084165632724762, "learning_rate": 3.846604827542602e-05, "loss": 0.0084, "step": 59540 }, { "grad_norm": 0.11838921159505844, "learning_rate": 3.844996015070117e-05, "loss": 0.0079, "step": 59550 }, { "grad_norm": 0.1752660721540451, "learning_rate": 3.843387328907132e-05, "loss": 0.0095, "step": 59560 }, { "grad_norm": 0.19758424162864685, "learning_rate": 3.8417787692295696e-05, "loss": 0.0099, "step": 59570 }, { "grad_norm": 0.16919495165348053, "learning_rate": 3.84017033621334e-05, "loss": 0.0079, "step": 59580 }, { "grad_norm": 0.13042083382606506, "learning_rate": 3.838562030034338e-05, "loss": 0.0076, "step": 59590 }, { "grad_norm": 0.14359118044376373, "learning_rate": 3.836953850868447e-05, "loss": 0.008, "step": 59600 }, { "grad_norm": 0.07453650236129761, "learning_rate": 3.835345798891533e-05, "loss": 0.008, "step": 59610 }, { "grad_norm": 0.13128520548343658, "learning_rate": 3.833737874279452e-05, "loss": 0.009, "step": 59620 }, { "grad_norm": 0.15226559340953827, "learning_rate": 3.8321300772080433e-05, "loss": 0.0075, "step": 59630 }, { "grad_norm": 0.1602175235748291, "learning_rate": 3.8305224078531345e-05, "loss": 0.0065, "step": 59640 }, { "grad_norm": 0.13043971359729767, "learning_rate": 3.828914866390537e-05, "loss": 0.0074, "step": 59650 }, { "grad_norm": 0.1019454151391983, "learning_rate": 3.827307452996048e-05, "loss": 0.0075, "step": 59660 }, { "grad_norm": 0.10932717472314835, "learning_rate": 3.825700167845454e-05, "loss": 0.0077, "step": 59670 }, { "grad_norm": 0.1120416596531868, "learning_rate": 3.8240930111145214e-05, "loss": 0.0074, "step": 59680 }, { "grad_norm": 0.09663701057434082, "learning_rate": 3.8224859829790116e-05, "loss": 0.0059, "step": 59690 }, { "grad_norm": 0.137888103723526, "learning_rate": 3.820879083614664e-05, "loss": 0.0074, "step": 59700 }, { "grad_norm": 0.12367034703493118, "learning_rate": 3.819272313197208e-05, "loss": 0.0073, "step": 59710 }, { "grad_norm": 0.11811880022287369, "learning_rate": 3.817665671902355e-05, "loss": 0.0068, "step": 59720 }, { "grad_norm": 0.10866203159093857, "learning_rate": 3.816059159905808e-05, "loss": 0.0064, "step": 59730 }, { "grad_norm": 0.10660247504711151, "learning_rate": 3.8144527773832476e-05, "loss": 0.0086, "step": 59740 }, { "grad_norm": 0.2143060863018036, "learning_rate": 3.812846524510352e-05, "loss": 0.0064, "step": 59750 }, { "grad_norm": 0.12600959837436676, "learning_rate": 3.811240401462775e-05, "loss": 0.0079, "step": 59760 }, { "grad_norm": 0.27946341037750244, "learning_rate": 3.809634408416162e-05, "loss": 0.0086, "step": 59770 }, { "grad_norm": 0.19659028947353363, "learning_rate": 3.8080285455461394e-05, "loss": 0.0075, "step": 59780 }, { "grad_norm": 0.0820886492729187, "learning_rate": 3.806422813028323e-05, "loss": 0.0073, "step": 59790 }, { "grad_norm": 0.11920097470283508, "learning_rate": 3.8048172110383114e-05, "loss": 0.0069, "step": 59800 }, { "grad_norm": 0.16759735345840454, "learning_rate": 3.8032117397516944e-05, "loss": 0.0085, "step": 59810 }, { "grad_norm": 0.16418133676052094, "learning_rate": 3.801606399344042e-05, "loss": 0.0071, "step": 59820 }, { "grad_norm": 0.16613121330738068, "learning_rate": 3.800001189990913e-05, "loss": 0.0085, "step": 59830 }, { "grad_norm": 0.5240309834480286, "learning_rate": 3.798396111867847e-05, "loss": 0.008, "step": 59840 }, { "grad_norm": 0.2137855738401413, "learning_rate": 3.796791165150378e-05, "loss": 0.0065, "step": 59850 }, { "grad_norm": 0.17878089845180511, "learning_rate": 3.795186350014018e-05, "loss": 0.008, "step": 59860 }, { "grad_norm": 0.13296818733215332, "learning_rate": 3.793581666634266e-05, "loss": 0.0096, "step": 59870 }, { "grad_norm": 0.12114079296588898, "learning_rate": 3.791977115186609e-05, "loss": 0.0065, "step": 59880 }, { "grad_norm": 0.1672225296497345, "learning_rate": 3.79037269584652e-05, "loss": 0.0072, "step": 59890 }, { "grad_norm": 0.11391182988882065, "learning_rate": 3.7887684087894545e-05, "loss": 0.0074, "step": 59900 }, { "grad_norm": 0.09894261509180069, "learning_rate": 3.787164254190854e-05, "loss": 0.0081, "step": 59910 }, { "grad_norm": 0.14650431275367737, "learning_rate": 3.7855602322261473e-05, "loss": 0.0075, "step": 59920 }, { "grad_norm": 0.12721237540245056, "learning_rate": 3.783956343070746e-05, "loss": 0.0071, "step": 59930 }, { "grad_norm": 0.1162337064743042, "learning_rate": 3.7823525869000526e-05, "loss": 0.007, "step": 59940 }, { "grad_norm": 0.1682235151529312, "learning_rate": 3.78074896388945e-05, "loss": 0.0079, "step": 59950 }, { "grad_norm": 0.10019835084676743, "learning_rate": 3.7791454742143075e-05, "loss": 0.0071, "step": 59960 }, { "grad_norm": 0.11936400085687637, "learning_rate": 3.7775421180499795e-05, "loss": 0.0084, "step": 59970 }, { "grad_norm": 0.15611980855464935, "learning_rate": 3.7759388955718086e-05, "loss": 0.0068, "step": 59980 }, { "grad_norm": 0.12157132476568222, "learning_rate": 3.774335806955117e-05, "loss": 0.0072, "step": 59990 }, { "grad_norm": 0.11545956134796143, "learning_rate": 3.7727328523752215e-05, "loss": 0.0076, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }