{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.2999789714813232, "learning_rate": 1.8e-07, "loss": 1.5087, "step": 10 }, { "grad_norm": 1.0337468385696411, "learning_rate": 3.8e-07, "loss": 1.4805, "step": 20 }, { "grad_norm": 0.9054698944091797, "learning_rate": 5.8e-07, "loss": 1.5214, "step": 30 }, { "grad_norm": 0.91335129737854, "learning_rate": 7.8e-07, "loss": 1.5301, "step": 40 }, { "grad_norm": 0.9347710013389587, "learning_rate": 9.8e-07, "loss": 1.5449, "step": 50 }, { "grad_norm": 1.1382232904434204, "learning_rate": 1.18e-06, "loss": 1.497, "step": 60 }, { "grad_norm": 1.7571982145309448, "learning_rate": 1.3800000000000001e-06, "loss": 1.4378, "step": 70 }, { "grad_norm": 2.436497211456299, "learning_rate": 1.5800000000000003e-06, "loss": 1.3454, "step": 80 }, { "grad_norm": 2.6342053413391113, "learning_rate": 1.7800000000000001e-06, "loss": 1.3535, "step": 90 }, { "grad_norm": 2.019131660461426, "learning_rate": 1.98e-06, "loss": 1.318, "step": 100 }, { "grad_norm": 1.5716463327407837, "learning_rate": 2.1800000000000003e-06, "loss": 1.3347, "step": 110 }, { "grad_norm": 1.8884512186050415, "learning_rate": 2.38e-06, "loss": 1.3143, "step": 120 }, { "grad_norm": 2.249955177307129, "learning_rate": 2.5800000000000003e-06, "loss": 1.2474, "step": 130 }, { "grad_norm": 2.935065746307373, "learning_rate": 2.78e-06, "loss": 1.2413, "step": 140 }, { "grad_norm": 1.5890040397644043, "learning_rate": 2.9800000000000003e-06, "loss": 1.2671, "step": 150 }, { "grad_norm": 2.3993825912475586, "learning_rate": 3.1800000000000005e-06, "loss": 1.2319, "step": 160 }, { "grad_norm": 2.7814767360687256, "learning_rate": 3.38e-06, "loss": 1.2231, "step": 170 }, { "grad_norm": 1.7896043062210083, "learning_rate": 3.58e-06, "loss": 1.1996, "step": 180 }, { "grad_norm": 2.1708357334136963, "learning_rate": 3.7800000000000002e-06, "loss": 1.2079, "step": 190 }, { "grad_norm": 4.219785690307617, "learning_rate": 3.98e-06, "loss": 1.2038, "step": 200 }, { "grad_norm": 2.4833922386169434, "learning_rate": 4.18e-06, "loss": 1.1769, "step": 210 }, { "grad_norm": 3.116649627685547, "learning_rate": 4.38e-06, "loss": 1.1742, "step": 220 }, { "grad_norm": 1.797316312789917, "learning_rate": 4.58e-06, "loss": 1.1588, "step": 230 }, { "grad_norm": 1.8282727003097534, "learning_rate": 4.780000000000001e-06, "loss": 1.1553, "step": 240 }, { "grad_norm": 2.1893136501312256, "learning_rate": 4.98e-06, "loss": 1.1521, "step": 250 }, { "grad_norm": 2.0851407051086426, "learning_rate": 5.18e-06, "loss": 1.1563, "step": 260 }, { "grad_norm": 5.005786418914795, "learning_rate": 5.38e-06, "loss": 1.0875, "step": 270 }, { "grad_norm": 1.7040501832962036, "learning_rate": 5.580000000000001e-06, "loss": 1.1065, "step": 280 }, { "grad_norm": 2.5409388542175293, "learning_rate": 5.78e-06, "loss": 1.092, "step": 290 }, { "grad_norm": 2.1788413524627686, "learning_rate": 5.98e-06, "loss": 1.1419, "step": 300 }, { "grad_norm": 2.443389415740967, "learning_rate": 6.18e-06, "loss": 1.1088, "step": 310 }, { "grad_norm": 2.4896345138549805, "learning_rate": 6.38e-06, "loss": 1.1025, "step": 320 }, { "grad_norm": 2.3804075717926025, "learning_rate": 6.58e-06, "loss": 1.1126, "step": 330 }, { "grad_norm": 1.9432756900787354, "learning_rate": 6.78e-06, "loss": 1.1199, "step": 340 }, { "grad_norm": 1.7413837909698486, "learning_rate": 6.98e-06, "loss": 1.1131, "step": 350 }, { "grad_norm": 1.4792907238006592, "learning_rate": 7.180000000000001e-06, "loss": 1.1046, "step": 360 }, { "grad_norm": 1.8567938804626465, "learning_rate": 7.3800000000000005e-06, "loss": 1.0942, "step": 370 }, { "grad_norm": 2.247176170349121, "learning_rate": 7.580000000000001e-06, "loss": 1.1147, "step": 380 }, { "grad_norm": 1.9300649166107178, "learning_rate": 7.78e-06, "loss": 1.0863, "step": 390 }, { "grad_norm": 2.032555341720581, "learning_rate": 7.98e-06, "loss": 1.1085, "step": 400 }, { "grad_norm": 2.139118194580078, "learning_rate": 8.18e-06, "loss": 1.103, "step": 410 }, { "grad_norm": 1.4397512674331665, "learning_rate": 8.380000000000001e-06, "loss": 1.0742, "step": 420 }, { "grad_norm": 2.9834728240966797, "learning_rate": 8.580000000000001e-06, "loss": 1.0992, "step": 430 }, { "grad_norm": 2.333289384841919, "learning_rate": 8.78e-06, "loss": 1.1531, "step": 440 }, { "grad_norm": 1.6748785972595215, "learning_rate": 8.98e-06, "loss": 1.1056, "step": 450 }, { "grad_norm": 1.5313407182693481, "learning_rate": 9.180000000000002e-06, "loss": 1.0828, "step": 460 }, { "grad_norm": 1.957666277885437, "learning_rate": 9.38e-06, "loss": 1.0848, "step": 470 }, { "grad_norm": 2.08618426322937, "learning_rate": 9.58e-06, "loss": 1.0968, "step": 480 }, { "grad_norm": 2.2431857585906982, "learning_rate": 9.78e-06, "loss": 1.0948, "step": 490 }, { "grad_norm": 1.7985305786132812, "learning_rate": 9.980000000000001e-06, "loss": 1.0487, "step": 500 }, { "grad_norm": 1.7567179203033447, "learning_rate": 1.018e-05, "loss": 1.0746, "step": 510 }, { "grad_norm": 1.4796762466430664, "learning_rate": 1.038e-05, "loss": 1.0856, "step": 520 }, { "grad_norm": 2.31488299369812, "learning_rate": 1.058e-05, "loss": 1.0759, "step": 530 }, { "grad_norm": 1.4274725914001465, "learning_rate": 1.0780000000000002e-05, "loss": 1.0775, "step": 540 }, { "grad_norm": 2.05112361907959, "learning_rate": 1.098e-05, "loss": 1.0675, "step": 550 }, { "grad_norm": 1.8026702404022217, "learning_rate": 1.118e-05, "loss": 1.0616, "step": 560 }, { "grad_norm": 1.490317702293396, "learning_rate": 1.1380000000000001e-05, "loss": 1.0368, "step": 570 }, { "grad_norm": 1.3672175407409668, "learning_rate": 1.1580000000000001e-05, "loss": 1.0594, "step": 580 }, { "grad_norm": 1.5663177967071533, "learning_rate": 1.178e-05, "loss": 1.0764, "step": 590 }, { "grad_norm": 1.563858985900879, "learning_rate": 1.198e-05, "loss": 1.0556, "step": 600 }, { "grad_norm": 1.3983826637268066, "learning_rate": 1.2180000000000002e-05, "loss": 1.0787, "step": 610 }, { "grad_norm": 1.8076251745224, "learning_rate": 1.238e-05, "loss": 1.0639, "step": 620 }, { "grad_norm": 1.7688864469528198, "learning_rate": 1.258e-05, "loss": 1.0736, "step": 630 }, { "grad_norm": 1.8916254043579102, "learning_rate": 1.278e-05, "loss": 1.105, "step": 640 }, { "grad_norm": 1.3577286005020142, "learning_rate": 1.2980000000000001e-05, "loss": 1.0614, "step": 650 }, { "grad_norm": 1.709441900253296, "learning_rate": 1.3180000000000001e-05, "loss": 1.076, "step": 660 }, { "grad_norm": 1.3422187566757202, "learning_rate": 1.338e-05, "loss": 1.0562, "step": 670 }, { "grad_norm": 1.918203353881836, "learning_rate": 1.358e-05, "loss": 1.0812, "step": 680 }, { "grad_norm": 2.081998825073242, "learning_rate": 1.3780000000000002e-05, "loss": 1.0461, "step": 690 }, { "grad_norm": 4.681004047393799, "learning_rate": 1.3980000000000002e-05, "loss": 1.071, "step": 700 }, { "grad_norm": 2.0163745880126953, "learning_rate": 1.4180000000000001e-05, "loss": 1.0342, "step": 710 }, { "grad_norm": 1.6080355644226074, "learning_rate": 1.4380000000000001e-05, "loss": 0.9941, "step": 720 }, { "grad_norm": 1.7643494606018066, "learning_rate": 1.4580000000000003e-05, "loss": 1.0441, "step": 730 }, { "grad_norm": 1.7233377695083618, "learning_rate": 1.4779999999999999e-05, "loss": 1.0171, "step": 740 }, { "grad_norm": 2.4450912475585938, "learning_rate": 1.4979999999999999e-05, "loss": 1.0122, "step": 750 }, { "grad_norm": 4.247749328613281, "learning_rate": 1.518e-05, "loss": 0.9903, "step": 760 }, { "grad_norm": 2.223236083984375, "learning_rate": 1.538e-05, "loss": 0.9551, "step": 770 }, { "grad_norm": 1.9361231327056885, "learning_rate": 1.558e-05, "loss": 0.9451, "step": 780 }, { "grad_norm": 2.192382335662842, "learning_rate": 1.578e-05, "loss": 0.9257, "step": 790 }, { "grad_norm": 2.241964101791382, "learning_rate": 1.598e-05, "loss": 0.9297, "step": 800 }, { "grad_norm": 2.6011528968811035, "learning_rate": 1.618e-05, "loss": 0.9166, "step": 810 }, { "grad_norm": 2.1114304065704346, "learning_rate": 1.6380000000000002e-05, "loss": 0.9082, "step": 820 }, { "grad_norm": 1.9717155694961548, "learning_rate": 1.658e-05, "loss": 0.8371, "step": 830 }, { "grad_norm": 2.6632437705993652, "learning_rate": 1.6780000000000002e-05, "loss": 0.8341, "step": 840 }, { "grad_norm": 2.3324711322784424, "learning_rate": 1.698e-05, "loss": 0.8618, "step": 850 }, { "grad_norm": 3.3334715366363525, "learning_rate": 1.718e-05, "loss": 0.811, "step": 860 }, { "grad_norm": 2.9062857627868652, "learning_rate": 1.7380000000000003e-05, "loss": 0.7997, "step": 870 }, { "grad_norm": 3.9300436973571777, "learning_rate": 1.758e-05, "loss": 0.7681, "step": 880 }, { "grad_norm": 2.6380813121795654, "learning_rate": 1.7780000000000003e-05, "loss": 0.7446, "step": 890 }, { "grad_norm": 2.7459309101104736, "learning_rate": 1.798e-05, "loss": 0.7183, "step": 900 }, { "grad_norm": 2.4614686965942383, "learning_rate": 1.818e-05, "loss": 0.7139, "step": 910 }, { "grad_norm": 2.9515552520751953, "learning_rate": 1.838e-05, "loss": 0.7019, "step": 920 }, { "grad_norm": 2.851221799850464, "learning_rate": 1.858e-05, "loss": 0.7018, "step": 930 }, { "grad_norm": 3.3028831481933594, "learning_rate": 1.878e-05, "loss": 0.6946, "step": 940 }, { "grad_norm": 2.872344970703125, "learning_rate": 1.898e-05, "loss": 0.6506, "step": 950 }, { "grad_norm": 2.7236738204956055, "learning_rate": 1.918e-05, "loss": 0.6315, "step": 960 }, { "grad_norm": 2.1923019886016846, "learning_rate": 1.938e-05, "loss": 0.6157, "step": 970 }, { "grad_norm": 2.8887786865234375, "learning_rate": 1.9580000000000002e-05, "loss": 0.5832, "step": 980 }, { "grad_norm": 3.141160249710083, "learning_rate": 1.978e-05, "loss": 0.5789, "step": 990 }, { "grad_norm": 4.650730133056641, "learning_rate": 1.9980000000000002e-05, "loss": 0.5502, "step": 1000 }, { "grad_norm": 2.787749767303467, "learning_rate": 2.0180000000000003e-05, "loss": 0.5669, "step": 1010 }, { "grad_norm": 3.048600673675537, "learning_rate": 2.038e-05, "loss": 0.5695, "step": 1020 }, { "grad_norm": 3.3068745136260986, "learning_rate": 2.0580000000000003e-05, "loss": 0.5402, "step": 1030 }, { "grad_norm": 2.3895530700683594, "learning_rate": 2.078e-05, "loss": 0.4982, "step": 1040 }, { "grad_norm": 4.58560848236084, "learning_rate": 2.098e-05, "loss": 0.5097, "step": 1050 }, { "grad_norm": 3.593337297439575, "learning_rate": 2.118e-05, "loss": 0.4804, "step": 1060 }, { "grad_norm": 2.2792272567749023, "learning_rate": 2.138e-05, "loss": 0.4675, "step": 1070 }, { "grad_norm": 3.5376522541046143, "learning_rate": 2.158e-05, "loss": 0.3935, "step": 1080 }, { "grad_norm": 3.289968252182007, "learning_rate": 2.178e-05, "loss": 0.4815, "step": 1090 }, { "grad_norm": 2.9650022983551025, "learning_rate": 2.198e-05, "loss": 0.4262, "step": 1100 }, { "grad_norm": 2.671862840652466, "learning_rate": 2.218e-05, "loss": 0.4662, "step": 1110 }, { "grad_norm": 3.180450916290283, "learning_rate": 2.2380000000000003e-05, "loss": 0.4337, "step": 1120 }, { "grad_norm": 2.949124574661255, "learning_rate": 2.258e-05, "loss": 0.3941, "step": 1130 }, { "grad_norm": 3.691068649291992, "learning_rate": 2.2780000000000002e-05, "loss": 0.3617, "step": 1140 }, { "grad_norm": 2.5901496410369873, "learning_rate": 2.298e-05, "loss": 0.3498, "step": 1150 }, { "grad_norm": 2.9564218521118164, "learning_rate": 2.318e-05, "loss": 0.359, "step": 1160 }, { "grad_norm": 3.414628744125366, "learning_rate": 2.3380000000000003e-05, "loss": 0.3653, "step": 1170 }, { "grad_norm": 3.077230930328369, "learning_rate": 2.358e-05, "loss": 0.3594, "step": 1180 }, { "grad_norm": 2.534036874771118, "learning_rate": 2.3780000000000003e-05, "loss": 0.339, "step": 1190 }, { "grad_norm": 2.719104290008545, "learning_rate": 2.398e-05, "loss": 0.3304, "step": 1200 }, { "grad_norm": 3.117006540298462, "learning_rate": 2.418e-05, "loss": 0.3291, "step": 1210 }, { "grad_norm": 3.09615159034729, "learning_rate": 2.438e-05, "loss": 0.3109, "step": 1220 }, { "grad_norm": 2.7593789100646973, "learning_rate": 2.4580000000000002e-05, "loss": 0.3074, "step": 1230 }, { "grad_norm": 3.0724587440490723, "learning_rate": 2.478e-05, "loss": 0.2966, "step": 1240 }, { "grad_norm": 3.270805597305298, "learning_rate": 2.498e-05, "loss": 0.289, "step": 1250 }, { "grad_norm": 2.8866846561431885, "learning_rate": 2.5180000000000003e-05, "loss": 0.2937, "step": 1260 }, { "grad_norm": 2.891955852508545, "learning_rate": 2.5380000000000004e-05, "loss": 0.2684, "step": 1270 }, { "grad_norm": 3.281022787094116, "learning_rate": 2.5580000000000002e-05, "loss": 0.2679, "step": 1280 }, { "grad_norm": 3.0849478244781494, "learning_rate": 2.5779999999999997e-05, "loss": 0.2183, "step": 1290 }, { "grad_norm": 3.119767904281616, "learning_rate": 2.598e-05, "loss": 0.2461, "step": 1300 }, { "grad_norm": 3.1522510051727295, "learning_rate": 2.618e-05, "loss": 0.254, "step": 1310 }, { "grad_norm": 2.6708731651306152, "learning_rate": 2.6379999999999998e-05, "loss": 0.2786, "step": 1320 }, { "grad_norm": 2.3003101348876953, "learning_rate": 2.658e-05, "loss": 0.197, "step": 1330 }, { "grad_norm": 3.6105332374572754, "learning_rate": 2.678e-05, "loss": 0.2099, "step": 1340 }, { "grad_norm": 2.971222400665283, "learning_rate": 2.698e-05, "loss": 0.2103, "step": 1350 }, { "grad_norm": 2.6109871864318848, "learning_rate": 2.718e-05, "loss": 0.2627, "step": 1360 }, { "grad_norm": 2.7403345108032227, "learning_rate": 2.738e-05, "loss": 0.236, "step": 1370 }, { "grad_norm": 3.902864933013916, "learning_rate": 2.758e-05, "loss": 0.2476, "step": 1380 }, { "grad_norm": 2.866886854171753, "learning_rate": 2.778e-05, "loss": 0.2094, "step": 1390 }, { "grad_norm": 2.7701027393341064, "learning_rate": 2.798e-05, "loss": 0.2477, "step": 1400 }, { "grad_norm": 2.44081711769104, "learning_rate": 2.818e-05, "loss": 0.1811, "step": 1410 }, { "grad_norm": 3.196608066558838, "learning_rate": 2.8380000000000003e-05, "loss": 0.2455, "step": 1420 }, { "grad_norm": 3.1678378582000732, "learning_rate": 2.858e-05, "loss": 0.2358, "step": 1430 }, { "grad_norm": 2.8300628662109375, "learning_rate": 2.8780000000000002e-05, "loss": 0.2391, "step": 1440 }, { "grad_norm": 1.8216831684112549, "learning_rate": 2.898e-05, "loss": 0.2051, "step": 1450 }, { "grad_norm": 2.7266488075256348, "learning_rate": 2.9180000000000002e-05, "loss": 0.2297, "step": 1460 }, { "grad_norm": 2.6649117469787598, "learning_rate": 2.9380000000000003e-05, "loss": 0.1908, "step": 1470 }, { "grad_norm": 1.6623284816741943, "learning_rate": 2.958e-05, "loss": 0.1731, "step": 1480 }, { "grad_norm": 2.961717367172241, "learning_rate": 2.9780000000000003e-05, "loss": 0.2151, "step": 1490 }, { "grad_norm": 1.8108640909194946, "learning_rate": 2.998e-05, "loss": 0.1987, "step": 1500 }, { "grad_norm": 2.2156107425689697, "learning_rate": 3.0180000000000002e-05, "loss": 0.199, "step": 1510 }, { "grad_norm": 2.324927568435669, "learning_rate": 3.0380000000000004e-05, "loss": 0.2123, "step": 1520 }, { "grad_norm": 2.398017168045044, "learning_rate": 3.058e-05, "loss": 0.2089, "step": 1530 }, { "grad_norm": 2.2075862884521484, "learning_rate": 3.078e-05, "loss": 0.2244, "step": 1540 }, { "grad_norm": 1.6411654949188232, "learning_rate": 3.0980000000000005e-05, "loss": 0.265, "step": 1550 }, { "grad_norm": 1.8049588203430176, "learning_rate": 3.118e-05, "loss": 0.2316, "step": 1560 }, { "grad_norm": 1.6742442846298218, "learning_rate": 3.138e-05, "loss": 0.1717, "step": 1570 }, { "grad_norm": 3.1772332191467285, "learning_rate": 3.1580000000000006e-05, "loss": 0.2177, "step": 1580 }, { "grad_norm": 2.0937283039093018, "learning_rate": 3.1780000000000004e-05, "loss": 0.2073, "step": 1590 }, { "grad_norm": 2.272651195526123, "learning_rate": 3.198e-05, "loss": 0.1927, "step": 1600 }, { "grad_norm": 1.463700771331787, "learning_rate": 3.218e-05, "loss": 0.2058, "step": 1610 }, { "grad_norm": 2.8637938499450684, "learning_rate": 3.238e-05, "loss": 0.2231, "step": 1620 }, { "grad_norm": 2.1848671436309814, "learning_rate": 3.2579999999999996e-05, "loss": 0.2095, "step": 1630 }, { "grad_norm": 2.4535088539123535, "learning_rate": 3.278e-05, "loss": 0.2315, "step": 1640 }, { "grad_norm": 2.6135475635528564, "learning_rate": 3.298e-05, "loss": 0.2679, "step": 1650 }, { "grad_norm": 2.1864306926727295, "learning_rate": 3.318e-05, "loss": 0.1964, "step": 1660 }, { "grad_norm": 2.3879587650299072, "learning_rate": 3.338e-05, "loss": 0.1855, "step": 1670 }, { "grad_norm": 1.8848532438278198, "learning_rate": 3.358e-05, "loss": 0.234, "step": 1680 }, { "grad_norm": 2.0575644969940186, "learning_rate": 3.378e-05, "loss": 0.1926, "step": 1690 }, { "grad_norm": 2.3039746284484863, "learning_rate": 3.398e-05, "loss": 0.2163, "step": 1700 }, { "grad_norm": 2.213229179382324, "learning_rate": 3.418e-05, "loss": 0.19, "step": 1710 }, { "grad_norm": 2.1504247188568115, "learning_rate": 3.438e-05, "loss": 0.2082, "step": 1720 }, { "grad_norm": 1.8578407764434814, "learning_rate": 3.4580000000000004e-05, "loss": 0.2092, "step": 1730 }, { "grad_norm": 2.4798247814178467, "learning_rate": 3.478e-05, "loss": 0.2154, "step": 1740 }, { "grad_norm": 1.6996171474456787, "learning_rate": 3.498e-05, "loss": 0.2061, "step": 1750 }, { "grad_norm": 2.8861889839172363, "learning_rate": 3.518e-05, "loss": 0.2049, "step": 1760 }, { "grad_norm": 2.173691749572754, "learning_rate": 3.5380000000000003e-05, "loss": 0.2104, "step": 1770 }, { "grad_norm": 2.314204692840576, "learning_rate": 3.558e-05, "loss": 0.2164, "step": 1780 }, { "grad_norm": 1.9529485702514648, "learning_rate": 3.578e-05, "loss": 0.1963, "step": 1790 }, { "grad_norm": 1.7303597927093506, "learning_rate": 3.5980000000000004e-05, "loss": 0.2108, "step": 1800 }, { "grad_norm": 2.510887622833252, "learning_rate": 3.618e-05, "loss": 0.1837, "step": 1810 }, { "grad_norm": 1.6862355470657349, "learning_rate": 3.638e-05, "loss": 0.2011, "step": 1820 }, { "grad_norm": 2.2998788356781006, "learning_rate": 3.6580000000000006e-05, "loss": 0.1877, "step": 1830 }, { "grad_norm": 2.068141222000122, "learning_rate": 3.6780000000000004e-05, "loss": 0.1649, "step": 1840 }, { "grad_norm": 2.2298662662506104, "learning_rate": 3.698e-05, "loss": 0.1932, "step": 1850 }, { "grad_norm": 2.505981922149658, "learning_rate": 3.7180000000000007e-05, "loss": 0.2282, "step": 1860 }, { "grad_norm": 1.80564546585083, "learning_rate": 3.7380000000000005e-05, "loss": 0.2172, "step": 1870 }, { "grad_norm": 2.4618706703186035, "learning_rate": 3.758e-05, "loss": 0.1913, "step": 1880 }, { "grad_norm": 1.8275238275527954, "learning_rate": 3.778000000000001e-05, "loss": 0.184, "step": 1890 }, { "grad_norm": 1.7478289604187012, "learning_rate": 3.7980000000000006e-05, "loss": 0.1747, "step": 1900 }, { "grad_norm": 1.9337714910507202, "learning_rate": 3.818e-05, "loss": 0.2182, "step": 1910 }, { "grad_norm": 1.7661222219467163, "learning_rate": 3.838e-05, "loss": 0.1751, "step": 1920 }, { "grad_norm": 2.090662956237793, "learning_rate": 3.858e-05, "loss": 0.2299, "step": 1930 }, { "grad_norm": 1.8009949922561646, "learning_rate": 3.878e-05, "loss": 0.2029, "step": 1940 }, { "grad_norm": 1.6500612497329712, "learning_rate": 3.898e-05, "loss": 0.1585, "step": 1950 }, { "grad_norm": 1.700050711631775, "learning_rate": 3.918e-05, "loss": 0.182, "step": 1960 }, { "grad_norm": 1.75567626953125, "learning_rate": 3.938e-05, "loss": 0.1732, "step": 1970 }, { "grad_norm": 1.8731763362884521, "learning_rate": 3.958e-05, "loss": 0.1455, "step": 1980 }, { "grad_norm": 2.288756847381592, "learning_rate": 3.978e-05, "loss": 0.1783, "step": 1990 }, { "grad_norm": 1.9242637157440186, "learning_rate": 3.998e-05, "loss": 0.1636, "step": 2000 }, { "grad_norm": 1.3852201700210571, "learning_rate": 4.018e-05, "loss": 0.1464, "step": 2010 }, { "grad_norm": 2.021758794784546, "learning_rate": 4.038e-05, "loss": 0.1771, "step": 2020 }, { "grad_norm": 2.401139259338379, "learning_rate": 4.058e-05, "loss": 0.2116, "step": 2030 }, { "grad_norm": 2.54866886138916, "learning_rate": 4.078e-05, "loss": 0.1922, "step": 2040 }, { "grad_norm": 1.8465160131454468, "learning_rate": 4.0980000000000004e-05, "loss": 0.1845, "step": 2050 }, { "grad_norm": 1.711256980895996, "learning_rate": 4.118e-05, "loss": 0.1942, "step": 2060 }, { "grad_norm": 2.2238521575927734, "learning_rate": 4.138e-05, "loss": 0.1737, "step": 2070 }, { "grad_norm": 1.692217230796814, "learning_rate": 4.1580000000000005e-05, "loss": 0.1805, "step": 2080 }, { "grad_norm": 1.7062296867370605, "learning_rate": 4.178e-05, "loss": 0.1775, "step": 2090 }, { "grad_norm": 1.4412841796875, "learning_rate": 4.198e-05, "loss": 0.1819, "step": 2100 }, { "grad_norm": 2.0939319133758545, "learning_rate": 4.2180000000000006e-05, "loss": 0.1771, "step": 2110 }, { "grad_norm": 1.464255690574646, "learning_rate": 4.2380000000000004e-05, "loss": 0.1775, "step": 2120 }, { "grad_norm": 2.0018320083618164, "learning_rate": 4.258e-05, "loss": 0.2117, "step": 2130 }, { "grad_norm": 1.5361099243164062, "learning_rate": 4.278e-05, "loss": 0.1756, "step": 2140 }, { "grad_norm": 1.7851722240447998, "learning_rate": 4.2980000000000005e-05, "loss": 0.2117, "step": 2150 }, { "grad_norm": 2.024303913116455, "learning_rate": 4.318e-05, "loss": 0.1921, "step": 2160 }, { "grad_norm": 1.9656016826629639, "learning_rate": 4.338e-05, "loss": 0.161, "step": 2170 }, { "grad_norm": 1.263687252998352, "learning_rate": 4.3580000000000006e-05, "loss": 0.1743, "step": 2180 }, { "grad_norm": 2.342334270477295, "learning_rate": 4.3780000000000004e-05, "loss": 0.1434, "step": 2190 }, { "grad_norm": 1.799949049949646, "learning_rate": 4.398e-05, "loss": 0.1528, "step": 2200 }, { "grad_norm": 1.7522125244140625, "learning_rate": 4.418000000000001e-05, "loss": 0.1996, "step": 2210 }, { "grad_norm": 2.306530714035034, "learning_rate": 4.438e-05, "loss": 0.2332, "step": 2220 }, { "grad_norm": 1.6935690641403198, "learning_rate": 4.458e-05, "loss": 0.1644, "step": 2230 }, { "grad_norm": 1.6619549989700317, "learning_rate": 4.478e-05, "loss": 0.1582, "step": 2240 }, { "grad_norm": 1.8525400161743164, "learning_rate": 4.498e-05, "loss": 0.1631, "step": 2250 }, { "grad_norm": 1.9716410636901855, "learning_rate": 4.518e-05, "loss": 0.1948, "step": 2260 }, { "grad_norm": 1.7621270418167114, "learning_rate": 4.538e-05, "loss": 0.2147, "step": 2270 }, { "grad_norm": 2.2234556674957275, "learning_rate": 4.558e-05, "loss": 0.1999, "step": 2280 }, { "grad_norm": 1.9681966304779053, "learning_rate": 4.578e-05, "loss": 0.2014, "step": 2290 }, { "grad_norm": 1.701026439666748, "learning_rate": 4.5980000000000004e-05, "loss": 0.2283, "step": 2300 }, { "grad_norm": 1.5654799938201904, "learning_rate": 4.618e-05, "loss": 0.1778, "step": 2310 }, { "grad_norm": 1.1328321695327759, "learning_rate": 4.638e-05, "loss": 0.1933, "step": 2320 }, { "grad_norm": 1.9506208896636963, "learning_rate": 4.6580000000000005e-05, "loss": 0.175, "step": 2330 }, { "grad_norm": 1.3923678398132324, "learning_rate": 4.678e-05, "loss": 0.1987, "step": 2340 }, { "grad_norm": 1.4993773698806763, "learning_rate": 4.698e-05, "loss": 0.1699, "step": 2350 }, { "grad_norm": 1.9660732746124268, "learning_rate": 4.718e-05, "loss": 0.1673, "step": 2360 }, { "grad_norm": 1.729556679725647, "learning_rate": 4.7380000000000004e-05, "loss": 0.2053, "step": 2370 }, { "grad_norm": 1.7407022714614868, "learning_rate": 4.758e-05, "loss": 0.1612, "step": 2380 }, { "grad_norm": 1.7002776861190796, "learning_rate": 4.778e-05, "loss": 0.1724, "step": 2390 }, { "grad_norm": 2.012495756149292, "learning_rate": 4.7980000000000005e-05, "loss": 0.1999, "step": 2400 }, { "grad_norm": 1.1519384384155273, "learning_rate": 4.818e-05, "loss": 0.2354, "step": 2410 }, { "grad_norm": 3.005587577819824, "learning_rate": 4.838e-05, "loss": 0.1857, "step": 2420 }, { "grad_norm": 1.1600761413574219, "learning_rate": 4.8580000000000006e-05, "loss": 0.1582, "step": 2430 }, { "grad_norm": 2.3356668949127197, "learning_rate": 4.8780000000000004e-05, "loss": 0.1958, "step": 2440 }, { "grad_norm": 1.7926123142242432, "learning_rate": 4.898e-05, "loss": 0.2071, "step": 2450 }, { "grad_norm": 1.4073736667633057, "learning_rate": 4.918000000000001e-05, "loss": 0.15, "step": 2460 }, { "grad_norm": 1.2159595489501953, "learning_rate": 4.9380000000000005e-05, "loss": 0.2056, "step": 2470 }, { "grad_norm": 1.4592134952545166, "learning_rate": 4.958e-05, "loss": 0.1435, "step": 2480 }, { "grad_norm": 1.6708757877349854, "learning_rate": 4.978e-05, "loss": 0.172, "step": 2490 }, { "grad_norm": 1.5542018413543701, "learning_rate": 4.9980000000000006e-05, "loss": 0.1664, "step": 2500 }, { "grad_norm": 1.7101271152496338, "learning_rate": 5.0180000000000004e-05, "loss": 0.1903, "step": 2510 }, { "grad_norm": 1.3640968799591064, "learning_rate": 5.038e-05, "loss": 0.1887, "step": 2520 }, { "grad_norm": 1.1433956623077393, "learning_rate": 5.058000000000001e-05, "loss": 0.1705, "step": 2530 }, { "grad_norm": 2.064706325531006, "learning_rate": 5.0780000000000005e-05, "loss": 0.1424, "step": 2540 }, { "grad_norm": 1.268963098526001, "learning_rate": 5.098e-05, "loss": 0.1804, "step": 2550 }, { "grad_norm": 1.4193394184112549, "learning_rate": 5.118000000000001e-05, "loss": 0.1767, "step": 2560 }, { "grad_norm": 1.9921258687973022, "learning_rate": 5.1380000000000006e-05, "loss": 0.1852, "step": 2570 }, { "grad_norm": 1.477550983428955, "learning_rate": 5.1580000000000004e-05, "loss": 0.1724, "step": 2580 }, { "grad_norm": 1.841800332069397, "learning_rate": 5.178000000000001e-05, "loss": 0.1709, "step": 2590 }, { "grad_norm": 1.3974024057388306, "learning_rate": 5.198000000000001e-05, "loss": 0.1788, "step": 2600 }, { "grad_norm": 1.4675047397613525, "learning_rate": 5.2180000000000005e-05, "loss": 0.1817, "step": 2610 }, { "grad_norm": 1.2433561086654663, "learning_rate": 5.238000000000001e-05, "loss": 0.1674, "step": 2620 }, { "grad_norm": 1.522177815437317, "learning_rate": 5.258000000000001e-05, "loss": 0.156, "step": 2630 }, { "grad_norm": 1.1888940334320068, "learning_rate": 5.2780000000000006e-05, "loss": 0.1923, "step": 2640 }, { "grad_norm": 1.2990227937698364, "learning_rate": 5.2980000000000004e-05, "loss": 0.1428, "step": 2650 }, { "grad_norm": 1.2002696990966797, "learning_rate": 5.318000000000001e-05, "loss": 0.1681, "step": 2660 }, { "grad_norm": 1.4996758699417114, "learning_rate": 5.338000000000001e-05, "loss": 0.1496, "step": 2670 }, { "grad_norm": 1.1934601068496704, "learning_rate": 5.3580000000000005e-05, "loss": 0.1716, "step": 2680 }, { "grad_norm": 1.0915848016738892, "learning_rate": 5.378e-05, "loss": 0.1502, "step": 2690 }, { "grad_norm": 1.250162124633789, "learning_rate": 5.3979999999999995e-05, "loss": 0.2191, "step": 2700 }, { "grad_norm": 1.488775372505188, "learning_rate": 5.418e-05, "loss": 0.1656, "step": 2710 }, { "grad_norm": 1.6625032424926758, "learning_rate": 5.438e-05, "loss": 0.1785, "step": 2720 }, { "grad_norm": 1.3809292316436768, "learning_rate": 5.4579999999999996e-05, "loss": 0.1456, "step": 2730 }, { "grad_norm": 1.3022693395614624, "learning_rate": 5.478e-05, "loss": 0.1336, "step": 2740 }, { "grad_norm": 1.1500828266143799, "learning_rate": 5.498e-05, "loss": 0.129, "step": 2750 }, { "grad_norm": 1.620293378829956, "learning_rate": 5.518e-05, "loss": 0.1725, "step": 2760 }, { "grad_norm": 1.027129054069519, "learning_rate": 5.538e-05, "loss": 0.1577, "step": 2770 }, { "grad_norm": 1.686100959777832, "learning_rate": 5.558e-05, "loss": 0.1618, "step": 2780 }, { "grad_norm": 1.6256698369979858, "learning_rate": 5.578e-05, "loss": 0.1656, "step": 2790 }, { "grad_norm": 1.3270959854125977, "learning_rate": 5.5979999999999996e-05, "loss": 0.163, "step": 2800 }, { "grad_norm": 1.3683511018753052, "learning_rate": 5.618e-05, "loss": 0.1626, "step": 2810 }, { "grad_norm": 1.3252032995224, "learning_rate": 5.638e-05, "loss": 0.1503, "step": 2820 }, { "grad_norm": 1.4887455701828003, "learning_rate": 5.658e-05, "loss": 0.1545, "step": 2830 }, { "grad_norm": 1.5533909797668457, "learning_rate": 5.678e-05, "loss": 0.1798, "step": 2840 }, { "grad_norm": 1.7016644477844238, "learning_rate": 5.698e-05, "loss": 0.1204, "step": 2850 }, { "grad_norm": 1.5568411350250244, "learning_rate": 5.718e-05, "loss": 0.1507, "step": 2860 }, { "grad_norm": 1.772148609161377, "learning_rate": 5.738e-05, "loss": 0.1667, "step": 2870 }, { "grad_norm": 1.483229160308838, "learning_rate": 5.758e-05, "loss": 0.1747, "step": 2880 }, { "grad_norm": 1.3077517747879028, "learning_rate": 5.778e-05, "loss": 0.2125, "step": 2890 }, { "grad_norm": 1.2430123090744019, "learning_rate": 5.7980000000000004e-05, "loss": 0.16, "step": 2900 }, { "grad_norm": 1.2719041109085083, "learning_rate": 5.818e-05, "loss": 0.1964, "step": 2910 }, { "grad_norm": 1.1183756589889526, "learning_rate": 5.838e-05, "loss": 0.1457, "step": 2920 }, { "grad_norm": 1.3009506464004517, "learning_rate": 5.858e-05, "loss": 0.1716, "step": 2930 }, { "grad_norm": 1.1314444541931152, "learning_rate": 5.878e-05, "loss": 0.1592, "step": 2940 }, { "grad_norm": 1.1538525819778442, "learning_rate": 5.898e-05, "loss": 0.1533, "step": 2950 }, { "grad_norm": 1.9273884296417236, "learning_rate": 5.918e-05, "loss": 0.1509, "step": 2960 }, { "grad_norm": 1.1569786071777344, "learning_rate": 5.9380000000000004e-05, "loss": 0.137, "step": 2970 }, { "grad_norm": 0.9925075769424438, "learning_rate": 5.958e-05, "loss": 0.1554, "step": 2980 }, { "grad_norm": 1.1413403749465942, "learning_rate": 5.978e-05, "loss": 0.1566, "step": 2990 }, { "grad_norm": 1.1765592098236084, "learning_rate": 5.9980000000000005e-05, "loss": 0.1665, "step": 3000 }, { "grad_norm": 1.0591037273406982, "learning_rate": 6.018e-05, "loss": 0.1555, "step": 3010 }, { "grad_norm": 1.0888968706130981, "learning_rate": 6.038e-05, "loss": 0.149, "step": 3020 }, { "grad_norm": 1.4010485410690308, "learning_rate": 6.0580000000000006e-05, "loss": 0.1547, "step": 3030 }, { "grad_norm": 1.8744181394577026, "learning_rate": 6.0780000000000004e-05, "loss": 0.1735, "step": 3040 }, { "grad_norm": 0.8931981325149536, "learning_rate": 6.098e-05, "loss": 0.1359, "step": 3050 }, { "grad_norm": 1.1735831499099731, "learning_rate": 6.118000000000001e-05, "loss": 0.2062, "step": 3060 }, { "grad_norm": 1.36512291431427, "learning_rate": 6.138e-05, "loss": 0.1828, "step": 3070 }, { "grad_norm": 1.2349743843078613, "learning_rate": 6.158e-05, "loss": 0.1361, "step": 3080 }, { "grad_norm": 1.1712849140167236, "learning_rate": 6.178000000000001e-05, "loss": 0.144, "step": 3090 }, { "grad_norm": 1.3344475030899048, "learning_rate": 6.198e-05, "loss": 0.1601, "step": 3100 }, { "grad_norm": 1.9994498491287231, "learning_rate": 6.218e-05, "loss": 0.1853, "step": 3110 }, { "grad_norm": 1.1980791091918945, "learning_rate": 6.238000000000001e-05, "loss": 0.1729, "step": 3120 }, { "grad_norm": 1.800714373588562, "learning_rate": 6.258e-05, "loss": 0.1446, "step": 3130 }, { "grad_norm": 1.0978639125823975, "learning_rate": 6.278e-05, "loss": 0.1433, "step": 3140 }, { "grad_norm": 0.8783717751502991, "learning_rate": 6.298000000000001e-05, "loss": 0.1441, "step": 3150 }, { "grad_norm": 1.1003254652023315, "learning_rate": 6.318e-05, "loss": 0.1432, "step": 3160 }, { "grad_norm": 1.1823978424072266, "learning_rate": 6.338e-05, "loss": 0.1428, "step": 3170 }, { "grad_norm": 2.0096426010131836, "learning_rate": 6.358000000000001e-05, "loss": 0.1894, "step": 3180 }, { "grad_norm": 1.01858651638031, "learning_rate": 6.378e-05, "loss": 0.1339, "step": 3190 }, { "grad_norm": 1.2110193967819214, "learning_rate": 6.398000000000001e-05, "loss": 0.1631, "step": 3200 }, { "grad_norm": 1.3427802324295044, "learning_rate": 6.418000000000001e-05, "loss": 0.1395, "step": 3210 }, { "grad_norm": 0.7922284603118896, "learning_rate": 6.438e-05, "loss": 0.1444, "step": 3220 }, { "grad_norm": 1.4838407039642334, "learning_rate": 6.458000000000001e-05, "loss": 0.1429, "step": 3230 }, { "grad_norm": 1.2851542234420776, "learning_rate": 6.478000000000001e-05, "loss": 0.1594, "step": 3240 }, { "grad_norm": 1.3332663774490356, "learning_rate": 6.498e-05, "loss": 0.1626, "step": 3250 }, { "grad_norm": 1.2385560274124146, "learning_rate": 6.518000000000001e-05, "loss": 0.1957, "step": 3260 }, { "grad_norm": 1.2212804555892944, "learning_rate": 6.538000000000001e-05, "loss": 0.1589, "step": 3270 }, { "grad_norm": 0.9331560730934143, "learning_rate": 6.558e-05, "loss": 0.1602, "step": 3280 }, { "grad_norm": 1.3164340257644653, "learning_rate": 6.578000000000001e-05, "loss": 0.1649, "step": 3290 }, { "grad_norm": 1.129140853881836, "learning_rate": 6.598e-05, "loss": 0.154, "step": 3300 }, { "grad_norm": 1.4561598300933838, "learning_rate": 6.618e-05, "loss": 0.1383, "step": 3310 }, { "grad_norm": 1.2587332725524902, "learning_rate": 6.638e-05, "loss": 0.1782, "step": 3320 }, { "grad_norm": 1.3310102224349976, "learning_rate": 6.658e-05, "loss": 0.1293, "step": 3330 }, { "grad_norm": 1.0909537076950073, "learning_rate": 6.678e-05, "loss": 0.1263, "step": 3340 }, { "grad_norm": 1.2329695224761963, "learning_rate": 6.698e-05, "loss": 0.1586, "step": 3350 }, { "grad_norm": 1.3737616539001465, "learning_rate": 6.718e-05, "loss": 0.1399, "step": 3360 }, { "grad_norm": 1.2844771146774292, "learning_rate": 6.738e-05, "loss": 0.1398, "step": 3370 }, { "grad_norm": 0.8916271328926086, "learning_rate": 6.758e-05, "loss": 0.154, "step": 3380 }, { "grad_norm": 1.3158564567565918, "learning_rate": 6.778e-05, "loss": 0.1551, "step": 3390 }, { "grad_norm": 0.7972403168678284, "learning_rate": 6.798e-05, "loss": 0.1587, "step": 3400 }, { "grad_norm": 1.4735355377197266, "learning_rate": 6.818e-05, "loss": 0.1585, "step": 3410 }, { "grad_norm": 1.2610212564468384, "learning_rate": 6.838e-05, "loss": 0.1879, "step": 3420 }, { "grad_norm": 0.9174610376358032, "learning_rate": 6.858e-05, "loss": 0.1537, "step": 3430 }, { "grad_norm": 1.0147756338119507, "learning_rate": 6.878e-05, "loss": 0.1573, "step": 3440 }, { "grad_norm": 1.218732476234436, "learning_rate": 6.898e-05, "loss": 0.1601, "step": 3450 }, { "grad_norm": 1.5361930131912231, "learning_rate": 6.918e-05, "loss": 0.154, "step": 3460 }, { "grad_norm": 1.3966912031173706, "learning_rate": 6.938e-05, "loss": 0.1629, "step": 3470 }, { "grad_norm": 2.3338963985443115, "learning_rate": 6.958e-05, "loss": 0.1602, "step": 3480 }, { "grad_norm": 1.4295283555984497, "learning_rate": 6.978e-05, "loss": 0.1619, "step": 3490 }, { "grad_norm": 1.278865933418274, "learning_rate": 6.998e-05, "loss": 0.1582, "step": 3500 }, { "grad_norm": 1.4031400680541992, "learning_rate": 7.018e-05, "loss": 0.1575, "step": 3510 }, { "grad_norm": 1.4104089736938477, "learning_rate": 7.038e-05, "loss": 0.1678, "step": 3520 }, { "grad_norm": 1.6019668579101562, "learning_rate": 7.058e-05, "loss": 0.1525, "step": 3530 }, { "grad_norm": 1.1907323598861694, "learning_rate": 7.078e-05, "loss": 0.1395, "step": 3540 }, { "grad_norm": 1.3832788467407227, "learning_rate": 7.098e-05, "loss": 0.1734, "step": 3550 }, { "grad_norm": 1.2436381578445435, "learning_rate": 7.118e-05, "loss": 0.1741, "step": 3560 }, { "grad_norm": 1.016220211982727, "learning_rate": 7.138e-05, "loss": 0.1358, "step": 3570 }, { "grad_norm": 1.0402131080627441, "learning_rate": 7.158e-05, "loss": 0.1684, "step": 3580 }, { "grad_norm": 1.547102928161621, "learning_rate": 7.178000000000001e-05, "loss": 0.1636, "step": 3590 }, { "grad_norm": 0.9351200461387634, "learning_rate": 7.198e-05, "loss": 0.1587, "step": 3600 }, { "grad_norm": 1.1675076484680176, "learning_rate": 7.218e-05, "loss": 0.1586, "step": 3610 }, { "grad_norm": 0.870975136756897, "learning_rate": 7.238000000000001e-05, "loss": 0.1245, "step": 3620 }, { "grad_norm": 0.9535595774650574, "learning_rate": 7.258e-05, "loss": 0.1357, "step": 3630 }, { "grad_norm": 1.4394817352294922, "learning_rate": 7.278e-05, "loss": 0.1731, "step": 3640 }, { "grad_norm": 1.0299646854400635, "learning_rate": 7.298000000000001e-05, "loss": 0.1321, "step": 3650 }, { "grad_norm": 1.352728009223938, "learning_rate": 7.318e-05, "loss": 0.1714, "step": 3660 }, { "grad_norm": 1.1838628053665161, "learning_rate": 7.338e-05, "loss": 0.2168, "step": 3670 }, { "grad_norm": 1.299118161201477, "learning_rate": 7.358000000000001e-05, "loss": 0.1364, "step": 3680 }, { "grad_norm": 1.746370553970337, "learning_rate": 7.378e-05, "loss": 0.1865, "step": 3690 }, { "grad_norm": 1.1340748071670532, "learning_rate": 7.398e-05, "loss": 0.1511, "step": 3700 }, { "grad_norm": 1.4865782260894775, "learning_rate": 7.418000000000001e-05, "loss": 0.1367, "step": 3710 }, { "grad_norm": 1.038588285446167, "learning_rate": 7.438e-05, "loss": 0.2037, "step": 3720 }, { "grad_norm": 1.076218605041504, "learning_rate": 7.458000000000001e-05, "loss": 0.1247, "step": 3730 }, { "grad_norm": 1.0192058086395264, "learning_rate": 7.478e-05, "loss": 0.1605, "step": 3740 }, { "grad_norm": 1.1384590864181519, "learning_rate": 7.498e-05, "loss": 0.126, "step": 3750 }, { "grad_norm": 1.2891685962677002, "learning_rate": 7.518000000000001e-05, "loss": 0.1426, "step": 3760 }, { "grad_norm": 0.8637300729751587, "learning_rate": 7.538e-05, "loss": 0.1425, "step": 3770 }, { "grad_norm": 1.2955814599990845, "learning_rate": 7.558e-05, "loss": 0.122, "step": 3780 }, { "grad_norm": 0.867163360118866, "learning_rate": 7.578000000000001e-05, "loss": 0.1256, "step": 3790 }, { "grad_norm": 0.7795839905738831, "learning_rate": 7.598e-05, "loss": 0.1143, "step": 3800 }, { "grad_norm": 1.0306341648101807, "learning_rate": 7.618e-05, "loss": 0.1511, "step": 3810 }, { "grad_norm": 1.624916434288025, "learning_rate": 7.638000000000001e-05, "loss": 0.1763, "step": 3820 }, { "grad_norm": 1.5180503129959106, "learning_rate": 7.658e-05, "loss": 0.157, "step": 3830 }, { "grad_norm": 1.2227530479431152, "learning_rate": 7.678000000000001e-05, "loss": 0.139, "step": 3840 }, { "grad_norm": 1.131988286972046, "learning_rate": 7.698000000000001e-05, "loss": 0.151, "step": 3850 }, { "grad_norm": 1.3207285404205322, "learning_rate": 7.718e-05, "loss": 0.1293, "step": 3860 }, { "grad_norm": 0.967350959777832, "learning_rate": 7.738000000000001e-05, "loss": 0.1511, "step": 3870 }, { "grad_norm": 1.5703362226486206, "learning_rate": 7.758000000000001e-05, "loss": 0.1636, "step": 3880 }, { "grad_norm": 0.869243323802948, "learning_rate": 7.778e-05, "loss": 0.133, "step": 3890 }, { "grad_norm": 1.1074692010879517, "learning_rate": 7.798000000000001e-05, "loss": 0.1289, "step": 3900 }, { "grad_norm": 1.2070090770721436, "learning_rate": 7.818000000000001e-05, "loss": 0.1661, "step": 3910 }, { "grad_norm": 0.8063422441482544, "learning_rate": 7.838e-05, "loss": 0.1669, "step": 3920 }, { "grad_norm": 0.9206150770187378, "learning_rate": 7.858000000000001e-05, "loss": 0.1349, "step": 3930 }, { "grad_norm": 1.0258771181106567, "learning_rate": 7.878e-05, "loss": 0.1328, "step": 3940 }, { "grad_norm": 0.9937017560005188, "learning_rate": 7.897999999999999e-05, "loss": 0.1654, "step": 3950 }, { "grad_norm": 1.5101584196090698, "learning_rate": 7.918e-05, "loss": 0.1699, "step": 3960 }, { "grad_norm": 1.0575716495513916, "learning_rate": 7.938e-05, "loss": 0.1394, "step": 3970 }, { "grad_norm": 1.2165272235870361, "learning_rate": 7.958e-05, "loss": 0.1125, "step": 3980 }, { "grad_norm": 1.299094796180725, "learning_rate": 7.978e-05, "loss": 0.1884, "step": 3990 }, { "grad_norm": 0.9744162559509277, "learning_rate": 7.998e-05, "loss": 0.1329, "step": 4000 }, { "grad_norm": 1.033716082572937, "learning_rate": 8.018e-05, "loss": 0.1567, "step": 4010 }, { "grad_norm": 0.737687885761261, "learning_rate": 8.038e-05, "loss": 0.1884, "step": 4020 }, { "grad_norm": 1.1167442798614502, "learning_rate": 8.058e-05, "loss": 0.1363, "step": 4030 }, { "grad_norm": 0.7995310425758362, "learning_rate": 8.078e-05, "loss": 0.1783, "step": 4040 }, { "grad_norm": 1.0326181650161743, "learning_rate": 8.098e-05, "loss": 0.1175, "step": 4050 }, { "grad_norm": 1.0610944032669067, "learning_rate": 8.118e-05, "loss": 0.1433, "step": 4060 }, { "grad_norm": 0.9086894989013672, "learning_rate": 8.138e-05, "loss": 0.1254, "step": 4070 }, { "grad_norm": 0.8485583662986755, "learning_rate": 8.158e-05, "loss": 0.1632, "step": 4080 }, { "grad_norm": 1.1573952436447144, "learning_rate": 8.178e-05, "loss": 0.1578, "step": 4090 }, { "grad_norm": 1.4135584831237793, "learning_rate": 8.198e-05, "loss": 0.1597, "step": 4100 }, { "grad_norm": 1.0028961896896362, "learning_rate": 8.218e-05, "loss": 0.1341, "step": 4110 }, { "grad_norm": 0.9223678112030029, "learning_rate": 8.238000000000001e-05, "loss": 0.1418, "step": 4120 }, { "grad_norm": 1.0407304763793945, "learning_rate": 8.258e-05, "loss": 0.1157, "step": 4130 }, { "grad_norm": 1.2791284322738647, "learning_rate": 8.278e-05, "loss": 0.1406, "step": 4140 }, { "grad_norm": 0.8246432542800903, "learning_rate": 8.298000000000001e-05, "loss": 0.1316, "step": 4150 }, { "grad_norm": 0.8036928772926331, "learning_rate": 8.318e-05, "loss": 0.1411, "step": 4160 }, { "grad_norm": 1.0800267457962036, "learning_rate": 8.338e-05, "loss": 0.1115, "step": 4170 }, { "grad_norm": 0.9871774315834045, "learning_rate": 8.358e-05, "loss": 0.1753, "step": 4180 }, { "grad_norm": 0.9774514436721802, "learning_rate": 8.378e-05, "loss": 0.1223, "step": 4190 }, { "grad_norm": 1.1900697946548462, "learning_rate": 8.398e-05, "loss": 0.1739, "step": 4200 }, { "grad_norm": 1.5394784212112427, "learning_rate": 8.418e-05, "loss": 0.1704, "step": 4210 }, { "grad_norm": 1.0612784624099731, "learning_rate": 8.438e-05, "loss": 0.1658, "step": 4220 }, { "grad_norm": 1.1888530254364014, "learning_rate": 8.458e-05, "loss": 0.1329, "step": 4230 }, { "grad_norm": 1.1009858846664429, "learning_rate": 8.478e-05, "loss": 0.1685, "step": 4240 }, { "grad_norm": 0.861624002456665, "learning_rate": 8.498e-05, "loss": 0.1558, "step": 4250 }, { "grad_norm": 0.9536699056625366, "learning_rate": 8.518000000000001e-05, "loss": 0.1293, "step": 4260 }, { "grad_norm": 0.8749063611030579, "learning_rate": 8.538e-05, "loss": 0.1127, "step": 4270 }, { "grad_norm": 0.9660388827323914, "learning_rate": 8.558e-05, "loss": 0.1097, "step": 4280 }, { "grad_norm": 0.9676087498664856, "learning_rate": 8.578000000000001e-05, "loss": 0.171, "step": 4290 }, { "grad_norm": 1.2251026630401611, "learning_rate": 8.598e-05, "loss": 0.1521, "step": 4300 }, { "grad_norm": 0.92543625831604, "learning_rate": 8.618e-05, "loss": 0.1344, "step": 4310 }, { "grad_norm": 1.4106054306030273, "learning_rate": 8.638000000000001e-05, "loss": 0.1396, "step": 4320 }, { "grad_norm": 1.048874855041504, "learning_rate": 8.658e-05, "loss": 0.1396, "step": 4330 }, { "grad_norm": 0.8251035213470459, "learning_rate": 8.678e-05, "loss": 0.1182, "step": 4340 }, { "grad_norm": 1.229075312614441, "learning_rate": 8.698000000000001e-05, "loss": 0.1394, "step": 4350 }, { "grad_norm": 0.9016222357749939, "learning_rate": 8.718e-05, "loss": 0.1528, "step": 4360 }, { "grad_norm": 1.0396432876586914, "learning_rate": 8.738000000000001e-05, "loss": 0.1524, "step": 4370 }, { "grad_norm": 1.235166072845459, "learning_rate": 8.758000000000001e-05, "loss": 0.1374, "step": 4380 }, { "grad_norm": 0.8655564188957214, "learning_rate": 8.778e-05, "loss": 0.1537, "step": 4390 }, { "grad_norm": 1.4228664636611938, "learning_rate": 8.798000000000001e-05, "loss": 0.1226, "step": 4400 }, { "grad_norm": 0.8612837195396423, "learning_rate": 8.818000000000001e-05, "loss": 0.1551, "step": 4410 }, { "grad_norm": 0.854858934879303, "learning_rate": 8.838e-05, "loss": 0.1166, "step": 4420 }, { "grad_norm": 1.0681464672088623, "learning_rate": 8.858000000000001e-05, "loss": 0.15, "step": 4430 }, { "grad_norm": 0.7745836973190308, "learning_rate": 8.878000000000001e-05, "loss": 0.1117, "step": 4440 }, { "grad_norm": 1.0793548822402954, "learning_rate": 8.898e-05, "loss": 0.1463, "step": 4450 }, { "grad_norm": 1.0379091501235962, "learning_rate": 8.918000000000001e-05, "loss": 0.1391, "step": 4460 }, { "grad_norm": 1.0401362180709839, "learning_rate": 8.938e-05, "loss": 0.1387, "step": 4470 }, { "grad_norm": 0.9853309392929077, "learning_rate": 8.958e-05, "loss": 0.1437, "step": 4480 }, { "grad_norm": 1.1163932085037231, "learning_rate": 8.978000000000001e-05, "loss": 0.1394, "step": 4490 }, { "grad_norm": 1.0072810649871826, "learning_rate": 8.998e-05, "loss": 0.1637, "step": 4500 }, { "grad_norm": 1.0289998054504395, "learning_rate": 9.018000000000001e-05, "loss": 0.138, "step": 4510 }, { "grad_norm": 1.10483717918396, "learning_rate": 9.038000000000001e-05, "loss": 0.111, "step": 4520 }, { "grad_norm": 1.485826015472412, "learning_rate": 9.058e-05, "loss": 0.1458, "step": 4530 }, { "grad_norm": 0.8545111417770386, "learning_rate": 9.078000000000001e-05, "loss": 0.1335, "step": 4540 }, { "grad_norm": 0.7175036072731018, "learning_rate": 9.098000000000001e-05, "loss": 0.1285, "step": 4550 }, { "grad_norm": 0.9614787101745605, "learning_rate": 9.118e-05, "loss": 0.1355, "step": 4560 }, { "grad_norm": 0.9151699542999268, "learning_rate": 9.138e-05, "loss": 0.1632, "step": 4570 }, { "grad_norm": 1.7229788303375244, "learning_rate": 9.158e-05, "loss": 0.1688, "step": 4580 }, { "grad_norm": 0.9914770722389221, "learning_rate": 9.178e-05, "loss": 0.1322, "step": 4590 }, { "grad_norm": 0.960816502571106, "learning_rate": 9.198e-05, "loss": 0.145, "step": 4600 }, { "grad_norm": 0.8745008111000061, "learning_rate": 9.218e-05, "loss": 0.1374, "step": 4610 }, { "grad_norm": 0.9675936698913574, "learning_rate": 9.238e-05, "loss": 0.1413, "step": 4620 }, { "grad_norm": 0.8551729917526245, "learning_rate": 9.258e-05, "loss": 0.1207, "step": 4630 }, { "grad_norm": 0.8341443538665771, "learning_rate": 9.278e-05, "loss": 0.1127, "step": 4640 }, { "grad_norm": 1.3366824388504028, "learning_rate": 9.298e-05, "loss": 0.1425, "step": 4650 }, { "grad_norm": 0.992962658405304, "learning_rate": 9.318e-05, "loss": 0.1369, "step": 4660 }, { "grad_norm": 0.8486049175262451, "learning_rate": 9.338e-05, "loss": 0.1262, "step": 4670 }, { "grad_norm": 1.14008367061615, "learning_rate": 9.358e-05, "loss": 0.1172, "step": 4680 }, { "grad_norm": 0.6046943664550781, "learning_rate": 9.378e-05, "loss": 0.1326, "step": 4690 }, { "grad_norm": 0.9369427561759949, "learning_rate": 9.398e-05, "loss": 0.1437, "step": 4700 }, { "grad_norm": 0.7061327695846558, "learning_rate": 9.418e-05, "loss": 0.1305, "step": 4710 }, { "grad_norm": 0.9792536497116089, "learning_rate": 9.438e-05, "loss": 0.145, "step": 4720 }, { "grad_norm": 0.9217740893363953, "learning_rate": 9.458e-05, "loss": 0.1525, "step": 4730 }, { "grad_norm": 0.9774044156074524, "learning_rate": 9.478e-05, "loss": 0.1297, "step": 4740 }, { "grad_norm": 0.9653123021125793, "learning_rate": 9.498e-05, "loss": 0.1482, "step": 4750 }, { "grad_norm": 0.9273828864097595, "learning_rate": 9.518000000000001e-05, "loss": 0.1549, "step": 4760 }, { "grad_norm": 0.9491762518882751, "learning_rate": 9.538e-05, "loss": 0.1537, "step": 4770 }, { "grad_norm": 1.044560432434082, "learning_rate": 9.558e-05, "loss": 0.1484, "step": 4780 }, { "grad_norm": 1.0365092754364014, "learning_rate": 9.578000000000001e-05, "loss": 0.1477, "step": 4790 }, { "grad_norm": 0.944820761680603, "learning_rate": 9.598e-05, "loss": 0.1281, "step": 4800 }, { "grad_norm": 0.9743428230285645, "learning_rate": 9.618e-05, "loss": 0.1458, "step": 4810 }, { "grad_norm": 1.000309705734253, "learning_rate": 9.638000000000001e-05, "loss": 0.1335, "step": 4820 }, { "grad_norm": 2.228243112564087, "learning_rate": 9.658e-05, "loss": 0.1888, "step": 4830 }, { "grad_norm": 0.7775306105613708, "learning_rate": 9.678e-05, "loss": 0.158, "step": 4840 }, { "grad_norm": 1.3716679811477661, "learning_rate": 9.698000000000001e-05, "loss": 0.1658, "step": 4850 }, { "grad_norm": 1.0272443294525146, "learning_rate": 9.718e-05, "loss": 0.1406, "step": 4860 }, { "grad_norm": 1.029098391532898, "learning_rate": 9.738e-05, "loss": 0.1799, "step": 4870 }, { "grad_norm": 0.9217886924743652, "learning_rate": 9.758000000000001e-05, "loss": 0.1654, "step": 4880 }, { "grad_norm": 0.90045565366745, "learning_rate": 9.778e-05, "loss": 0.136, "step": 4890 }, { "grad_norm": 1.2500791549682617, "learning_rate": 9.798000000000001e-05, "loss": 0.1375, "step": 4900 }, { "grad_norm": 0.9312176704406738, "learning_rate": 9.818000000000001e-05, "loss": 0.1401, "step": 4910 }, { "grad_norm": 0.871899425983429, "learning_rate": 9.838e-05, "loss": 0.1407, "step": 4920 }, { "grad_norm": 1.2025597095489502, "learning_rate": 9.858000000000001e-05, "loss": 0.1531, "step": 4930 }, { "grad_norm": 1.0698890686035156, "learning_rate": 9.878e-05, "loss": 0.1131, "step": 4940 }, { "grad_norm": 0.9464985728263855, "learning_rate": 9.898e-05, "loss": 0.1634, "step": 4950 }, { "grad_norm": 0.9783420562744141, "learning_rate": 9.918000000000001e-05, "loss": 0.1551, "step": 4960 }, { "grad_norm": 1.062891960144043, "learning_rate": 9.938e-05, "loss": 0.1272, "step": 4970 }, { "grad_norm": 0.929410457611084, "learning_rate": 9.958e-05, "loss": 0.1448, "step": 4980 }, { "grad_norm": 1.0384961366653442, "learning_rate": 9.978000000000001e-05, "loss": 0.1522, "step": 4990 }, { "grad_norm": 0.7811340093612671, "learning_rate": 9.998e-05, "loss": 0.1263, "step": 5000 }, { "grad_norm": 0.7950204014778137, "learning_rate": 9.999999778549045e-05, "loss": 0.1211, "step": 5010 }, { "grad_norm": 1.1428141593933105, "learning_rate": 9.999999013039593e-05, "loss": 0.122, "step": 5020 }, { "grad_norm": 0.9436079859733582, "learning_rate": 9.999997700737766e-05, "loss": 0.1325, "step": 5030 }, { "grad_norm": 0.8943577408790588, "learning_rate": 9.999995841643709e-05, "loss": 0.1298, "step": 5040 }, { "grad_norm": 1.159459114074707, "learning_rate": 9.999993435757623e-05, "loss": 0.1216, "step": 5050 }, { "grad_norm": 0.8930107355117798, "learning_rate": 9.999990483079773e-05, "loss": 0.1045, "step": 5060 }, { "grad_norm": 1.0827443599700928, "learning_rate": 9.999986983610481e-05, "loss": 0.1281, "step": 5070 }, { "grad_norm": 1.0148978233337402, "learning_rate": 9.99998293735013e-05, "loss": 0.1318, "step": 5080 }, { "grad_norm": 1.0747365951538086, "learning_rate": 9.999978344299161e-05, "loss": 0.1493, "step": 5090 }, { "grad_norm": 1.1784765720367432, "learning_rate": 9.99997320445808e-05, "loss": 0.112, "step": 5100 }, { "grad_norm": 0.9150728583335876, "learning_rate": 9.999967517827444e-05, "loss": 0.1538, "step": 5110 }, { "grad_norm": 0.8517307639122009, "learning_rate": 9.999961284407879e-05, "loss": 0.1183, "step": 5120 }, { "grad_norm": 0.9224507212638855, "learning_rate": 9.999954504200067e-05, "loss": 0.1288, "step": 5130 }, { "grad_norm": 0.9588322639465332, "learning_rate": 9.999947177204744e-05, "loss": 0.1417, "step": 5140 }, { "grad_norm": 0.7951157093048096, "learning_rate": 9.999939303422718e-05, "loss": 0.1543, "step": 5150 }, { "grad_norm": 0.785211980342865, "learning_rate": 9.999930882854847e-05, "loss": 0.1283, "step": 5160 }, { "grad_norm": 0.9815780520439148, "learning_rate": 9.999921915502051e-05, "loss": 0.1348, "step": 5170 }, { "grad_norm": 0.9084077477455139, "learning_rate": 9.99991240136531e-05, "loss": 0.1211, "step": 5180 }, { "grad_norm": 0.7913064956665039, "learning_rate": 9.999902340445668e-05, "loss": 0.1637, "step": 5190 }, { "grad_norm": 0.7784050703048706, "learning_rate": 9.999891732744224e-05, "loss": 0.1134, "step": 5200 }, { "grad_norm": 0.8181003332138062, "learning_rate": 9.999880578262135e-05, "loss": 0.0984, "step": 5210 }, { "grad_norm": 1.1222312450408936, "learning_rate": 9.999868877000624e-05, "loss": 0.1524, "step": 5220 }, { "grad_norm": 1.0285342931747437, "learning_rate": 9.99985662896097e-05, "loss": 0.1216, "step": 5230 }, { "grad_norm": 0.9088329076766968, "learning_rate": 9.999843834144513e-05, "loss": 0.1184, "step": 5240 }, { "grad_norm": 0.7600256204605103, "learning_rate": 9.99983049255265e-05, "loss": 0.1052, "step": 5250 }, { "grad_norm": 0.753591775894165, "learning_rate": 9.999816604186843e-05, "loss": 0.1497, "step": 5260 }, { "grad_norm": 1.1479241847991943, "learning_rate": 9.999802169048609e-05, "loss": 0.1339, "step": 5270 }, { "grad_norm": 0.7621347904205322, "learning_rate": 9.999787187139527e-05, "loss": 0.1486, "step": 5280 }, { "grad_norm": 1.2159819602966309, "learning_rate": 9.999771658461234e-05, "loss": 0.1551, "step": 5290 }, { "grad_norm": 0.8471705317497253, "learning_rate": 9.999755583015431e-05, "loss": 0.1283, "step": 5300 }, { "grad_norm": 0.946432888507843, "learning_rate": 9.999738960803874e-05, "loss": 0.1387, "step": 5310 }, { "grad_norm": 0.9538792967796326, "learning_rate": 9.99972179182838e-05, "loss": 0.1446, "step": 5320 }, { "grad_norm": 1.0150171518325806, "learning_rate": 9.99970407609083e-05, "loss": 0.1704, "step": 5330 }, { "grad_norm": 0.9506930112838745, "learning_rate": 9.999685813593159e-05, "loss": 0.1589, "step": 5340 }, { "grad_norm": 0.7460045218467712, "learning_rate": 9.999667004337362e-05, "loss": 0.1528, "step": 5350 }, { "grad_norm": 0.8496687412261963, "learning_rate": 9.9996476483255e-05, "loss": 0.1701, "step": 5360 }, { "grad_norm": 0.8978176116943359, "learning_rate": 9.999627745559688e-05, "loss": 0.164, "step": 5370 }, { "grad_norm": 0.8294360041618347, "learning_rate": 9.999607296042101e-05, "loss": 0.124, "step": 5380 }, { "grad_norm": 1.0832182168960571, "learning_rate": 9.99958629977498e-05, "loss": 0.1158, "step": 5390 }, { "grad_norm": 1.1089184284210205, "learning_rate": 9.999564756760615e-05, "loss": 0.1152, "step": 5400 }, { "grad_norm": 0.9324980974197388, "learning_rate": 9.999542667001366e-05, "loss": 0.1168, "step": 5410 }, { "grad_norm": 0.7165359854698181, "learning_rate": 9.999520030499647e-05, "loss": 0.1168, "step": 5420 }, { "grad_norm": 0.8465978503227234, "learning_rate": 9.999496847257936e-05, "loss": 0.1081, "step": 5430 }, { "grad_norm": 0.9086726903915405, "learning_rate": 9.999473117278764e-05, "loss": 0.1492, "step": 5440 }, { "grad_norm": 0.9058041572570801, "learning_rate": 9.999448840564731e-05, "loss": 0.1422, "step": 5450 }, { "grad_norm": 0.8960558772087097, "learning_rate": 9.999424017118488e-05, "loss": 0.121, "step": 5460 }, { "grad_norm": 1.078031063079834, "learning_rate": 9.999398646942751e-05, "loss": 0.1129, "step": 5470 }, { "grad_norm": 0.869474470615387, "learning_rate": 9.999372730040296e-05, "loss": 0.1266, "step": 5480 }, { "grad_norm": 0.9434345364570618, "learning_rate": 9.999346266413953e-05, "loss": 0.1334, "step": 5490 }, { "grad_norm": 0.8159617781639099, "learning_rate": 9.99931925606662e-05, "loss": 0.1386, "step": 5500 }, { "grad_norm": 1.2069973945617676, "learning_rate": 9.99929169900125e-05, "loss": 0.107, "step": 5510 }, { "grad_norm": 1.0130586624145508, "learning_rate": 9.999263595220855e-05, "loss": 0.157, "step": 5520 }, { "grad_norm": 0.9699564576148987, "learning_rate": 9.99923494472851e-05, "loss": 0.1204, "step": 5530 }, { "grad_norm": 1.15045166015625, "learning_rate": 9.999205747527348e-05, "loss": 0.1335, "step": 5540 }, { "grad_norm": 0.8134603500366211, "learning_rate": 9.999176003620561e-05, "loss": 0.1193, "step": 5550 }, { "grad_norm": 1.0244261026382446, "learning_rate": 9.999145713011405e-05, "loss": 0.1288, "step": 5560 }, { "grad_norm": 1.0129117965698242, "learning_rate": 9.999114875703186e-05, "loss": 0.1482, "step": 5570 }, { "grad_norm": 0.8712672591209412, "learning_rate": 9.999083491699281e-05, "loss": 0.1229, "step": 5580 }, { "grad_norm": 1.601746678352356, "learning_rate": 9.999051561003123e-05, "loss": 0.1267, "step": 5590 }, { "grad_norm": 0.7339165806770325, "learning_rate": 9.999019083618202e-05, "loss": 0.118, "step": 5600 }, { "grad_norm": 1.106119990348816, "learning_rate": 9.99898605954807e-05, "loss": 0.1429, "step": 5610 }, { "grad_norm": 1.0619242191314697, "learning_rate": 9.998952488796338e-05, "loss": 0.1792, "step": 5620 }, { "grad_norm": 1.0516984462738037, "learning_rate": 9.998918371366676e-05, "loss": 0.1648, "step": 5630 }, { "grad_norm": 1.1541635990142822, "learning_rate": 9.99888370726282e-05, "loss": 0.1517, "step": 5640 }, { "grad_norm": 0.9341591000556946, "learning_rate": 9.998848496488556e-05, "loss": 0.1346, "step": 5650 }, { "grad_norm": 1.4681689739227295, "learning_rate": 9.998812739047736e-05, "loss": 0.1318, "step": 5660 }, { "grad_norm": 1.0708715915679932, "learning_rate": 9.99877643494427e-05, "loss": 0.1357, "step": 5670 }, { "grad_norm": 0.8539832234382629, "learning_rate": 9.998739584182128e-05, "loss": 0.1486, "step": 5680 }, { "grad_norm": 0.8722744584083557, "learning_rate": 9.998702186765342e-05, "loss": 0.1493, "step": 5690 }, { "grad_norm": 1.0466628074645996, "learning_rate": 9.998664242698e-05, "loss": 0.1335, "step": 5700 }, { "grad_norm": 1.0694522857666016, "learning_rate": 9.998625751984251e-05, "loss": 0.1081, "step": 5710 }, { "grad_norm": 1.109666109085083, "learning_rate": 9.998586714628307e-05, "loss": 0.1184, "step": 5720 }, { "grad_norm": 0.9099955558776855, "learning_rate": 9.998547130634432e-05, "loss": 0.1233, "step": 5730 }, { "grad_norm": 0.9360054731369019, "learning_rate": 9.99850700000696e-05, "loss": 0.1129, "step": 5740 }, { "grad_norm": 0.9608575701713562, "learning_rate": 9.998466322750278e-05, "loss": 0.1056, "step": 5750 }, { "grad_norm": 0.9387140274047852, "learning_rate": 9.998425098868834e-05, "loss": 0.1096, "step": 5760 }, { "grad_norm": 0.7619345188140869, "learning_rate": 9.998383328367136e-05, "loss": 0.1422, "step": 5770 }, { "grad_norm": 0.8362706899642944, "learning_rate": 9.99834101124975e-05, "loss": 0.1279, "step": 5780 }, { "grad_norm": 0.8881763219833374, "learning_rate": 9.998298147521309e-05, "loss": 0.115, "step": 5790 }, { "grad_norm": 1.0351622104644775, "learning_rate": 9.998254737186496e-05, "loss": 0.1551, "step": 5800 }, { "grad_norm": 1.0193599462509155, "learning_rate": 9.99821078025006e-05, "loss": 0.1348, "step": 5810 }, { "grad_norm": 1.1003669500350952, "learning_rate": 9.998166276716807e-05, "loss": 0.1429, "step": 5820 }, { "grad_norm": 1.4043207168579102, "learning_rate": 9.998121226591606e-05, "loss": 0.1268, "step": 5830 }, { "grad_norm": 0.978020191192627, "learning_rate": 9.998075629879382e-05, "loss": 0.1281, "step": 5840 }, { "grad_norm": 1.0763921737670898, "learning_rate": 9.99802948658512e-05, "loss": 0.1037, "step": 5850 }, { "grad_norm": 0.827333390712738, "learning_rate": 9.99798279671387e-05, "loss": 0.1152, "step": 5860 }, { "grad_norm": 0.683734655380249, "learning_rate": 9.997935560270734e-05, "loss": 0.1343, "step": 5870 }, { "grad_norm": 0.7756742835044861, "learning_rate": 9.997887777260879e-05, "loss": 0.0965, "step": 5880 }, { "grad_norm": 0.7500959038734436, "learning_rate": 9.997839447689532e-05, "loss": 0.1056, "step": 5890 }, { "grad_norm": 0.7433828115463257, "learning_rate": 9.997790571561978e-05, "loss": 0.1079, "step": 5900 }, { "grad_norm": 0.6339647769927979, "learning_rate": 9.99774114888356e-05, "loss": 0.1199, "step": 5910 }, { "grad_norm": 0.6103628873825073, "learning_rate": 9.997691179659684e-05, "loss": 0.1002, "step": 5920 }, { "grad_norm": 1.0061120986938477, "learning_rate": 9.997640663895815e-05, "loss": 0.109, "step": 5930 }, { "grad_norm": 1.0779772996902466, "learning_rate": 9.997589601597477e-05, "loss": 0.1191, "step": 5940 }, { "grad_norm": 0.9823411703109741, "learning_rate": 9.997537992770252e-05, "loss": 0.1288, "step": 5950 }, { "grad_norm": 1.0091222524642944, "learning_rate": 9.997485837419788e-05, "loss": 0.1294, "step": 5960 }, { "grad_norm": 0.9518134593963623, "learning_rate": 9.997433135551786e-05, "loss": 0.1127, "step": 5970 }, { "grad_norm": 0.9507105946540833, "learning_rate": 9.997379887172009e-05, "loss": 0.1512, "step": 5980 }, { "grad_norm": 0.8716434240341187, "learning_rate": 9.997326092286281e-05, "loss": 0.1694, "step": 5990 }, { "grad_norm": 0.9360793828964233, "learning_rate": 9.997271750900486e-05, "loss": 0.1365, "step": 6000 }, { "grad_norm": 1.3751652240753174, "learning_rate": 9.997216863020565e-05, "loss": 0.161, "step": 6010 }, { "grad_norm": 0.7708413600921631, "learning_rate": 9.99716142865252e-05, "loss": 0.1352, "step": 6020 }, { "grad_norm": 0.9521217942237854, "learning_rate": 9.997105447802415e-05, "loss": 0.1429, "step": 6030 }, { "grad_norm": 0.96779465675354, "learning_rate": 9.997048920476373e-05, "loss": 0.1561, "step": 6040 }, { "grad_norm": 1.2222915887832642, "learning_rate": 9.996991846680572e-05, "loss": 0.1442, "step": 6050 }, { "grad_norm": 1.6202744245529175, "learning_rate": 9.996934226421257e-05, "loss": 0.1309, "step": 6060 }, { "grad_norm": 0.7285489439964294, "learning_rate": 9.996876059704726e-05, "loss": 0.1055, "step": 6070 }, { "grad_norm": 0.8213813900947571, "learning_rate": 9.996817346537343e-05, "loss": 0.1268, "step": 6080 }, { "grad_norm": 0.9583041667938232, "learning_rate": 9.996758086925526e-05, "loss": 0.124, "step": 6090 }, { "grad_norm": 0.6623499393463135, "learning_rate": 9.996698280875759e-05, "loss": 0.1404, "step": 6100 }, { "grad_norm": 0.8927132487297058, "learning_rate": 9.99663792839458e-05, "loss": 0.126, "step": 6110 }, { "grad_norm": 0.7962413430213928, "learning_rate": 9.99657702948859e-05, "loss": 0.1099, "step": 6120 }, { "grad_norm": 0.7729344964027405, "learning_rate": 9.996515584164448e-05, "loss": 0.0941, "step": 6130 }, { "grad_norm": 0.9036170840263367, "learning_rate": 9.996453592428873e-05, "loss": 0.0989, "step": 6140 }, { "grad_norm": 1.3603142499923706, "learning_rate": 9.996391054288646e-05, "loss": 0.1311, "step": 6150 }, { "grad_norm": 0.9750224351882935, "learning_rate": 9.996327969750605e-05, "loss": 0.1357, "step": 6160 }, { "grad_norm": 1.152574062347412, "learning_rate": 9.996264338821649e-05, "loss": 0.1284, "step": 6170 }, { "grad_norm": 0.8891603350639343, "learning_rate": 9.996200161508735e-05, "loss": 0.1943, "step": 6180 }, { "grad_norm": 0.9453259706497192, "learning_rate": 9.996135437818885e-05, "loss": 0.1289, "step": 6190 }, { "grad_norm": 0.7228505611419678, "learning_rate": 9.996070167759175e-05, "loss": 0.1232, "step": 6200 }, { "grad_norm": 0.674874484539032, "learning_rate": 9.996004351336743e-05, "loss": 0.1074, "step": 6210 }, { "grad_norm": 0.6328564882278442, "learning_rate": 9.995937988558785e-05, "loss": 0.1195, "step": 6220 }, { "grad_norm": 0.7432196736335754, "learning_rate": 9.995871079432561e-05, "loss": 0.1134, "step": 6230 }, { "grad_norm": 0.7951350808143616, "learning_rate": 9.995803623965389e-05, "loss": 0.0894, "step": 6240 }, { "grad_norm": 1.043434500694275, "learning_rate": 9.995735622164641e-05, "loss": 0.1022, "step": 6250 }, { "grad_norm": 0.789287805557251, "learning_rate": 9.995667074037758e-05, "loss": 0.1224, "step": 6260 }, { "grad_norm": 0.8912470936775208, "learning_rate": 9.995597979592232e-05, "loss": 0.1184, "step": 6270 }, { "grad_norm": 0.9137488007545471, "learning_rate": 9.995528338835625e-05, "loss": 0.1246, "step": 6280 }, { "grad_norm": 0.8150200247764587, "learning_rate": 9.995458151775547e-05, "loss": 0.1044, "step": 6290 }, { "grad_norm": 0.5725212693214417, "learning_rate": 9.995387418419677e-05, "loss": 0.1053, "step": 6300 }, { "grad_norm": 0.9270899891853333, "learning_rate": 9.99531613877575e-05, "loss": 0.1055, "step": 6310 }, { "grad_norm": 0.6959233283996582, "learning_rate": 9.995244312851559e-05, "loss": 0.12, "step": 6320 }, { "grad_norm": 1.2582374811172485, "learning_rate": 9.995171940654961e-05, "loss": 0.1132, "step": 6330 }, { "grad_norm": 0.8996168971061707, "learning_rate": 9.995099022193871e-05, "loss": 0.108, "step": 6340 }, { "grad_norm": 0.763170063495636, "learning_rate": 9.995025557476261e-05, "loss": 0.1226, "step": 6350 }, { "grad_norm": 0.7138766050338745, "learning_rate": 9.994951546510165e-05, "loss": 0.1306, "step": 6360 }, { "grad_norm": 0.7848260402679443, "learning_rate": 9.994876989303679e-05, "loss": 0.1194, "step": 6370 }, { "grad_norm": 0.9666366577148438, "learning_rate": 9.994801885864955e-05, "loss": 0.1239, "step": 6380 }, { "grad_norm": 0.9111448526382446, "learning_rate": 9.994726236202205e-05, "loss": 0.1123, "step": 6390 }, { "grad_norm": 0.9341292977333069, "learning_rate": 9.994650040323704e-05, "loss": 0.1238, "step": 6400 }, { "grad_norm": 0.8286621570587158, "learning_rate": 9.994573298237784e-05, "loss": 0.1317, "step": 6410 }, { "grad_norm": 0.7254128456115723, "learning_rate": 9.994496009952837e-05, "loss": 0.1058, "step": 6420 }, { "grad_norm": 1.006181001663208, "learning_rate": 9.994418175477316e-05, "loss": 0.1308, "step": 6430 }, { "grad_norm": 0.6626794934272766, "learning_rate": 9.994339794819733e-05, "loss": 0.1616, "step": 6440 }, { "grad_norm": 0.805599570274353, "learning_rate": 9.994260867988658e-05, "loss": 0.1581, "step": 6450 }, { "grad_norm": 0.818274974822998, "learning_rate": 9.994181394992723e-05, "loss": 0.1248, "step": 6460 }, { "grad_norm": 0.8211098313331604, "learning_rate": 9.994101375840618e-05, "loss": 0.1292, "step": 6470 }, { "grad_norm": 0.5861597061157227, "learning_rate": 9.994020810541098e-05, "loss": 0.098, "step": 6480 }, { "grad_norm": 0.8806034326553345, "learning_rate": 9.99393969910297e-05, "loss": 0.0976, "step": 6490 }, { "grad_norm": 0.8553801774978638, "learning_rate": 9.993858041535104e-05, "loss": 0.1537, "step": 6500 }, { "grad_norm": 0.6457381248474121, "learning_rate": 9.99377583784643e-05, "loss": 0.1151, "step": 6510 }, { "grad_norm": 0.744419276714325, "learning_rate": 9.993693088045939e-05, "loss": 0.1196, "step": 6520 }, { "grad_norm": 0.8464398980140686, "learning_rate": 9.99360979214268e-05, "loss": 0.1284, "step": 6530 }, { "grad_norm": 0.8561275005340576, "learning_rate": 9.99352595014576e-05, "loss": 0.1205, "step": 6540 }, { "grad_norm": 1.0398883819580078, "learning_rate": 9.993441562064354e-05, "loss": 0.1283, "step": 6550 }, { "grad_norm": 1.1236913204193115, "learning_rate": 9.993356627907685e-05, "loss": 0.1128, "step": 6560 }, { "grad_norm": 1.3543939590454102, "learning_rate": 9.99327114768504e-05, "loss": 0.1215, "step": 6570 }, { "grad_norm": 0.8821060061454773, "learning_rate": 9.99318512140577e-05, "loss": 0.113, "step": 6580 }, { "grad_norm": 0.584140956401825, "learning_rate": 9.993098549079284e-05, "loss": 0.1092, "step": 6590 }, { "grad_norm": 0.6920478940010071, "learning_rate": 9.993011430715047e-05, "loss": 0.0974, "step": 6600 }, { "grad_norm": 0.5935183167457581, "learning_rate": 9.992923766322586e-05, "loss": 0.1371, "step": 6610 }, { "grad_norm": 0.6408575177192688, "learning_rate": 9.99283555591149e-05, "loss": 0.1115, "step": 6620 }, { "grad_norm": 0.6220712661743164, "learning_rate": 9.992746799491404e-05, "loss": 0.142, "step": 6630 }, { "grad_norm": 0.7680237889289856, "learning_rate": 9.992657497072033e-05, "loss": 0.1302, "step": 6640 }, { "grad_norm": 0.9695742130279541, "learning_rate": 9.992567648663147e-05, "loss": 0.1145, "step": 6650 }, { "grad_norm": 0.8711234331130981, "learning_rate": 9.992477254274568e-05, "loss": 0.1309, "step": 6660 }, { "grad_norm": 0.7893824577331543, "learning_rate": 9.992386313916183e-05, "loss": 0.1258, "step": 6670 }, { "grad_norm": 0.7778504490852356, "learning_rate": 9.992294827597934e-05, "loss": 0.093, "step": 6680 }, { "grad_norm": 0.6263057589530945, "learning_rate": 9.992202795329831e-05, "loss": 0.1164, "step": 6690 }, { "grad_norm": 0.8931609988212585, "learning_rate": 9.992110217121936e-05, "loss": 0.1066, "step": 6700 }, { "grad_norm": 0.8929260969161987, "learning_rate": 9.992017092984372e-05, "loss": 0.111, "step": 6710 }, { "grad_norm": 0.7664082646369934, "learning_rate": 9.991923422927326e-05, "loss": 0.1297, "step": 6720 }, { "grad_norm": 0.9767596125602722, "learning_rate": 9.991829206961037e-05, "loss": 0.1301, "step": 6730 }, { "grad_norm": 0.6199769377708435, "learning_rate": 9.991734445095813e-05, "loss": 0.1262, "step": 6740 }, { "grad_norm": 0.8687585592269897, "learning_rate": 9.991639137342015e-05, "loss": 0.1387, "step": 6750 }, { "grad_norm": 1.0184540748596191, "learning_rate": 9.991543283710064e-05, "loss": 0.1795, "step": 6760 }, { "grad_norm": 0.9459222555160522, "learning_rate": 9.991446884210445e-05, "loss": 0.1284, "step": 6770 }, { "grad_norm": 0.7307919859886169, "learning_rate": 9.9913499388537e-05, "loss": 0.1046, "step": 6780 }, { "grad_norm": 0.630869448184967, "learning_rate": 9.99125244765043e-05, "loss": 0.1183, "step": 6790 }, { "grad_norm": 0.9947811365127563, "learning_rate": 9.991154410611296e-05, "loss": 0.103, "step": 6800 }, { "grad_norm": 0.7738773226737976, "learning_rate": 9.99105582774702e-05, "loss": 0.1152, "step": 6810 }, { "grad_norm": 0.8058304786682129, "learning_rate": 9.990956699068384e-05, "loss": 0.0973, "step": 6820 }, { "grad_norm": 0.7895737290382385, "learning_rate": 9.990857024586224e-05, "loss": 0.1077, "step": 6830 }, { "grad_norm": 1.005699634552002, "learning_rate": 9.990756804311446e-05, "loss": 0.1012, "step": 6840 }, { "grad_norm": 0.8369748592376709, "learning_rate": 9.990656038255006e-05, "loss": 0.1124, "step": 6850 }, { "grad_norm": 0.8035370707511902, "learning_rate": 9.990554726427926e-05, "loss": 0.1411, "step": 6860 }, { "grad_norm": 0.9230621457099915, "learning_rate": 9.990452868841284e-05, "loss": 0.1308, "step": 6870 }, { "grad_norm": 0.8386424779891968, "learning_rate": 9.99035046550622e-05, "loss": 0.1626, "step": 6880 }, { "grad_norm": 0.8783845901489258, "learning_rate": 9.99024751643393e-05, "loss": 0.1251, "step": 6890 }, { "grad_norm": 0.8673522472381592, "learning_rate": 9.990144021635677e-05, "loss": 0.1223, "step": 6900 }, { "grad_norm": 0.7029398083686829, "learning_rate": 9.990039981122775e-05, "loss": 0.1088, "step": 6910 }, { "grad_norm": 0.9421749711036682, "learning_rate": 9.989935394906602e-05, "loss": 0.1141, "step": 6920 }, { "grad_norm": 0.8853627443313599, "learning_rate": 9.989830262998598e-05, "loss": 0.1515, "step": 6930 }, { "grad_norm": 0.8241662383079529, "learning_rate": 9.989724585410259e-05, "loss": 0.1224, "step": 6940 }, { "grad_norm": 0.7687853574752808, "learning_rate": 9.989618362153139e-05, "loss": 0.1206, "step": 6950 }, { "grad_norm": 0.7951071262359619, "learning_rate": 9.989511593238859e-05, "loss": 0.1266, "step": 6960 }, { "grad_norm": 0.7036670446395874, "learning_rate": 9.98940427867909e-05, "loss": 0.1021, "step": 6970 }, { "grad_norm": 1.4277465343475342, "learning_rate": 9.989296418485573e-05, "loss": 0.1283, "step": 6980 }, { "grad_norm": 0.8923998475074768, "learning_rate": 9.989188012670101e-05, "loss": 0.1007, "step": 6990 }, { "grad_norm": 0.9818207025527954, "learning_rate": 9.989079061244528e-05, "loss": 0.1, "step": 7000 }, { "grad_norm": 0.9210498929023743, "learning_rate": 9.988969564220769e-05, "loss": 0.1149, "step": 7010 }, { "grad_norm": 0.736630380153656, "learning_rate": 9.988859521610801e-05, "loss": 0.0982, "step": 7020 }, { "grad_norm": 0.579588770866394, "learning_rate": 9.988748933426656e-05, "loss": 0.0972, "step": 7030 }, { "grad_norm": 0.8542248010635376, "learning_rate": 9.988637799680428e-05, "loss": 0.1224, "step": 7040 }, { "grad_norm": 0.8072926998138428, "learning_rate": 9.98852612038427e-05, "loss": 0.1128, "step": 7050 }, { "grad_norm": 0.7244641184806824, "learning_rate": 9.988413895550397e-05, "loss": 0.1096, "step": 7060 }, { "grad_norm": 0.7837875485420227, "learning_rate": 9.98830112519108e-05, "loss": 0.1003, "step": 7070 }, { "grad_norm": 0.771587073802948, "learning_rate": 9.98818780931865e-05, "loss": 0.1245, "step": 7080 }, { "grad_norm": 0.7157162427902222, "learning_rate": 9.988073947945502e-05, "loss": 0.0987, "step": 7090 }, { "grad_norm": 0.734557569026947, "learning_rate": 9.987959541084087e-05, "loss": 0.1156, "step": 7100 }, { "grad_norm": 0.7000184059143066, "learning_rate": 9.987844588746915e-05, "loss": 0.086, "step": 7110 }, { "grad_norm": 0.583631694316864, "learning_rate": 9.987729090946558e-05, "loss": 0.0868, "step": 7120 }, { "grad_norm": 0.7219714522361755, "learning_rate": 9.987613047695647e-05, "loss": 0.117, "step": 7130 }, { "grad_norm": 0.7350841760635376, "learning_rate": 9.987496459006871e-05, "loss": 0.123, "step": 7140 }, { "grad_norm": 0.7094064354896545, "learning_rate": 9.987379324892982e-05, "loss": 0.1202, "step": 7150 }, { "grad_norm": 0.8231486082077026, "learning_rate": 9.987261645366788e-05, "loss": 0.1258, "step": 7160 }, { "grad_norm": 0.7094378471374512, "learning_rate": 9.987143420441158e-05, "loss": 0.1284, "step": 7170 }, { "grad_norm": 1.2684569358825684, "learning_rate": 9.987024650129022e-05, "loss": 0.1134, "step": 7180 }, { "grad_norm": 0.5695740580558777, "learning_rate": 9.986905334443368e-05, "loss": 0.1147, "step": 7190 }, { "grad_norm": 0.6490839123725891, "learning_rate": 9.986785473397245e-05, "loss": 0.1052, "step": 7200 }, { "grad_norm": 0.6231657266616821, "learning_rate": 9.98666506700376e-05, "loss": 0.096, "step": 7210 }, { "grad_norm": 0.7952033281326294, "learning_rate": 9.986544115276081e-05, "loss": 0.1068, "step": 7220 }, { "grad_norm": 0.9645623564720154, "learning_rate": 9.986422618227433e-05, "loss": 0.1095, "step": 7230 }, { "grad_norm": 0.7535861730575562, "learning_rate": 9.986300575871106e-05, "loss": 0.0944, "step": 7240 }, { "grad_norm": 0.7801653742790222, "learning_rate": 9.986177988220444e-05, "loss": 0.0845, "step": 7250 }, { "grad_norm": 1.1970727443695068, "learning_rate": 9.986054855288856e-05, "loss": 0.0774, "step": 7260 }, { "grad_norm": 0.6277251839637756, "learning_rate": 9.985931177089802e-05, "loss": 0.0981, "step": 7270 }, { "grad_norm": 0.9096134305000305, "learning_rate": 9.985806953636814e-05, "loss": 0.1089, "step": 7280 }, { "grad_norm": 0.7727604508399963, "learning_rate": 9.985682184943471e-05, "loss": 0.1088, "step": 7290 }, { "grad_norm": 0.6391924619674683, "learning_rate": 9.98555687102342e-05, "loss": 0.1304, "step": 7300 }, { "grad_norm": 0.5740845203399658, "learning_rate": 9.985431011890367e-05, "loss": 0.1163, "step": 7310 }, { "grad_norm": 0.8446229696273804, "learning_rate": 9.985304607558075e-05, "loss": 0.1011, "step": 7320 }, { "grad_norm": 0.7666597366333008, "learning_rate": 9.985177658040364e-05, "loss": 0.1121, "step": 7330 }, { "grad_norm": 0.9583081603050232, "learning_rate": 9.985050163351119e-05, "loss": 0.1066, "step": 7340 }, { "grad_norm": 0.8239176869392395, "learning_rate": 9.984922123504286e-05, "loss": 0.1124, "step": 7350 }, { "grad_norm": 0.6881431341171265, "learning_rate": 9.984793538513862e-05, "loss": 0.1039, "step": 7360 }, { "grad_norm": 0.7996229529380798, "learning_rate": 9.984664408393912e-05, "loss": 0.1052, "step": 7370 }, { "grad_norm": 0.7205153107643127, "learning_rate": 9.984534733158556e-05, "loss": 0.0937, "step": 7380 }, { "grad_norm": 0.6604582071304321, "learning_rate": 9.984404512821977e-05, "loss": 0.1018, "step": 7390 }, { "grad_norm": 0.653145968914032, "learning_rate": 9.984273747398411e-05, "loss": 0.0978, "step": 7400 }, { "grad_norm": 0.5815316438674927, "learning_rate": 9.984142436902165e-05, "loss": 0.1034, "step": 7410 }, { "grad_norm": 0.9117770791053772, "learning_rate": 9.984010581347596e-05, "loss": 0.1036, "step": 7420 }, { "grad_norm": 0.8603994846343994, "learning_rate": 9.983878180749121e-05, "loss": 0.11, "step": 7430 }, { "grad_norm": 0.9266615509986877, "learning_rate": 9.983745235121222e-05, "loss": 0.1173, "step": 7440 }, { "grad_norm": 0.879861056804657, "learning_rate": 9.983611744478438e-05, "loss": 0.0943, "step": 7450 }, { "grad_norm": 0.5892406702041626, "learning_rate": 9.983477708835365e-05, "loss": 0.0882, "step": 7460 }, { "grad_norm": 0.7312259674072266, "learning_rate": 9.983343128206664e-05, "loss": 0.128, "step": 7470 }, { "grad_norm": 0.9973168969154358, "learning_rate": 9.983208002607049e-05, "loss": 0.1106, "step": 7480 }, { "grad_norm": 0.7887935042381287, "learning_rate": 9.9830723320513e-05, "loss": 0.1063, "step": 7490 }, { "grad_norm": 0.5704343914985657, "learning_rate": 9.982936116554254e-05, "loss": 0.1116, "step": 7500 }, { "grad_norm": 0.790701150894165, "learning_rate": 9.982799356130803e-05, "loss": 0.1062, "step": 7510 }, { "grad_norm": 0.6250545978546143, "learning_rate": 9.982662050795908e-05, "loss": 0.0967, "step": 7520 }, { "grad_norm": 0.738427460193634, "learning_rate": 9.982524200564583e-05, "loss": 0.1173, "step": 7530 }, { "grad_norm": 1.0064451694488525, "learning_rate": 9.982385805451901e-05, "loss": 0.137, "step": 7540 }, { "grad_norm": 0.7911376357078552, "learning_rate": 9.982246865472998e-05, "loss": 0.1314, "step": 7550 }, { "grad_norm": 0.7766091227531433, "learning_rate": 9.982107380643069e-05, "loss": 0.1181, "step": 7560 }, { "grad_norm": 0.9999731183052063, "learning_rate": 9.981967350977368e-05, "loss": 0.1235, "step": 7570 }, { "grad_norm": 0.8162015676498413, "learning_rate": 9.981826776491208e-05, "loss": 0.1154, "step": 7580 }, { "grad_norm": 0.6944808959960938, "learning_rate": 9.98168565719996e-05, "loss": 0.1322, "step": 7590 }, { "grad_norm": 0.7759225964546204, "learning_rate": 9.98154399311906e-05, "loss": 0.1109, "step": 7600 }, { "grad_norm": 0.7320502400398254, "learning_rate": 9.981401784263997e-05, "loss": 0.1017, "step": 7610 }, { "grad_norm": 0.6754457354545593, "learning_rate": 9.981259030650326e-05, "loss": 0.0783, "step": 7620 }, { "grad_norm": 0.6483884453773499, "learning_rate": 9.981115732293655e-05, "loss": 0.1132, "step": 7630 }, { "grad_norm": 0.6610525250434875, "learning_rate": 9.980971889209659e-05, "loss": 0.1071, "step": 7640 }, { "grad_norm": 0.7511760592460632, "learning_rate": 9.980827501414064e-05, "loss": 0.0952, "step": 7650 }, { "grad_norm": 0.5372986793518066, "learning_rate": 9.980682568922663e-05, "loss": 0.0988, "step": 7660 }, { "grad_norm": 0.8947480916976929, "learning_rate": 9.980537091751304e-05, "loss": 0.0856, "step": 7670 }, { "grad_norm": 0.9834985136985779, "learning_rate": 9.980391069915897e-05, "loss": 0.1168, "step": 7680 }, { "grad_norm": 0.8126869201660156, "learning_rate": 9.98024450343241e-05, "loss": 0.1304, "step": 7690 }, { "grad_norm": 1.0007219314575195, "learning_rate": 9.980097392316872e-05, "loss": 0.1266, "step": 7700 }, { "grad_norm": 0.9393213987350464, "learning_rate": 9.97994973658537e-05, "loss": 0.1, "step": 7710 }, { "grad_norm": 0.7474406957626343, "learning_rate": 9.979801536254054e-05, "loss": 0.0996, "step": 7720 }, { "grad_norm": 0.8309262990951538, "learning_rate": 9.979652791339127e-05, "loss": 0.1066, "step": 7730 }, { "grad_norm": 0.8329525589942932, "learning_rate": 9.97950350185686e-05, "loss": 0.0925, "step": 7740 }, { "grad_norm": 0.6326771378517151, "learning_rate": 9.979353667823574e-05, "loss": 0.0985, "step": 7750 }, { "grad_norm": 0.9652905464172363, "learning_rate": 9.979203289255658e-05, "loss": 0.0876, "step": 7760 }, { "grad_norm": 0.8166513442993164, "learning_rate": 9.979052366169557e-05, "loss": 0.1145, "step": 7770 }, { "grad_norm": 0.952558696269989, "learning_rate": 9.978900898581775e-05, "loss": 0.0926, "step": 7780 }, { "grad_norm": 0.9395313858985901, "learning_rate": 9.978748886508875e-05, "loss": 0.0995, "step": 7790 }, { "grad_norm": 1.0647861957550049, "learning_rate": 9.978596329967484e-05, "loss": 0.1019, "step": 7800 }, { "grad_norm": 0.6119666695594788, "learning_rate": 9.978443228974284e-05, "loss": 0.0998, "step": 7810 }, { "grad_norm": 0.6195831298828125, "learning_rate": 9.978289583546015e-05, "loss": 0.1116, "step": 7820 }, { "grad_norm": 1.3978350162506104, "learning_rate": 9.978135393699484e-05, "loss": 0.1105, "step": 7830 }, { "grad_norm": 0.8016823530197144, "learning_rate": 9.977980659451548e-05, "loss": 0.1175, "step": 7840 }, { "grad_norm": 0.7295243144035339, "learning_rate": 9.977825380819135e-05, "loss": 0.0929, "step": 7850 }, { "grad_norm": 0.6421626210212708, "learning_rate": 9.97766955781922e-05, "loss": 0.091, "step": 7860 }, { "grad_norm": 1.0991190671920776, "learning_rate": 9.977513190468848e-05, "loss": 0.1275, "step": 7870 }, { "grad_norm": 1.0543222427368164, "learning_rate": 9.977356278785116e-05, "loss": 0.1261, "step": 7880 }, { "grad_norm": 0.7575540542602539, "learning_rate": 9.977198822785184e-05, "loss": 0.138, "step": 7890 }, { "grad_norm": 0.7555174827575684, "learning_rate": 9.977040822486273e-05, "loss": 0.0962, "step": 7900 }, { "grad_norm": 0.7185583114624023, "learning_rate": 9.97688227790566e-05, "loss": 0.1574, "step": 7910 }, { "grad_norm": 0.9485403299331665, "learning_rate": 9.976723189060684e-05, "loss": 0.1329, "step": 7920 }, { "grad_norm": 0.6946051716804504, "learning_rate": 9.976563555968742e-05, "loss": 0.0918, "step": 7930 }, { "grad_norm": 0.7765624523162842, "learning_rate": 9.976403378647292e-05, "loss": 0.1144, "step": 7940 }, { "grad_norm": 0.6201130747795105, "learning_rate": 9.97624265711385e-05, "loss": 0.1229, "step": 7950 }, { "grad_norm": 0.9678464531898499, "learning_rate": 9.976081391385993e-05, "loss": 0.1046, "step": 7960 }, { "grad_norm": 0.7910401225090027, "learning_rate": 9.975919581481356e-05, "loss": 0.1021, "step": 7970 }, { "grad_norm": 0.8938075304031372, "learning_rate": 9.975757227417634e-05, "loss": 0.1111, "step": 7980 }, { "grad_norm": 1.086105465888977, "learning_rate": 9.975594329212586e-05, "loss": 0.1033, "step": 7990 }, { "grad_norm": 0.8023087382316589, "learning_rate": 9.97543088688402e-05, "loss": 0.1056, "step": 8000 }, { "grad_norm": 0.7730830907821655, "learning_rate": 9.975266900449814e-05, "loss": 0.0817, "step": 8010 }, { "grad_norm": 0.8706198930740356, "learning_rate": 9.975102369927898e-05, "loss": 0.1332, "step": 8020 }, { "grad_norm": 1.5995371341705322, "learning_rate": 9.974937295336269e-05, "loss": 0.1138, "step": 8030 }, { "grad_norm": 0.6261574029922485, "learning_rate": 9.974771676692975e-05, "loss": 0.0888, "step": 8040 }, { "grad_norm": 0.9045760035514832, "learning_rate": 9.974605514016131e-05, "loss": 0.1082, "step": 8050 }, { "grad_norm": 0.9409982562065125, "learning_rate": 9.974438807323907e-05, "loss": 0.1344, "step": 8060 }, { "grad_norm": 0.8122219443321228, "learning_rate": 9.974271556634535e-05, "loss": 0.0937, "step": 8070 }, { "grad_norm": 0.846705436706543, "learning_rate": 9.974103761966302e-05, "loss": 0.104, "step": 8080 }, { "grad_norm": 0.7308586239814758, "learning_rate": 9.973935423337563e-05, "loss": 0.1193, "step": 8090 }, { "grad_norm": 0.6171557307243347, "learning_rate": 9.973766540766722e-05, "loss": 0.1094, "step": 8100 }, { "grad_norm": 0.6030229330062866, "learning_rate": 9.97359711427225e-05, "loss": 0.0988, "step": 8110 }, { "grad_norm": 0.7807687520980835, "learning_rate": 9.973427143872677e-05, "loss": 0.1145, "step": 8120 }, { "grad_norm": 0.6997115612030029, "learning_rate": 9.973256629586589e-05, "loss": 0.0963, "step": 8130 }, { "grad_norm": 0.65814608335495, "learning_rate": 9.973085571432632e-05, "loss": 0.0872, "step": 8140 }, { "grad_norm": 0.738318681716919, "learning_rate": 9.972913969429513e-05, "loss": 0.1269, "step": 8150 }, { "grad_norm": 0.5460048317909241, "learning_rate": 9.972741823596e-05, "loss": 0.1017, "step": 8160 }, { "grad_norm": 0.7045845985412598, "learning_rate": 9.972569133950917e-05, "loss": 0.0984, "step": 8170 }, { "grad_norm": 0.8220614194869995, "learning_rate": 9.972395900513151e-05, "loss": 0.0939, "step": 8180 }, { "grad_norm": 0.6868319511413574, "learning_rate": 9.972222123301645e-05, "loss": 0.0775, "step": 8190 }, { "grad_norm": 0.7684982419013977, "learning_rate": 9.972047802335403e-05, "loss": 0.1017, "step": 8200 }, { "grad_norm": 0.8984031081199646, "learning_rate": 9.971872937633488e-05, "loss": 0.1026, "step": 8210 }, { "grad_norm": 0.5528653264045715, "learning_rate": 9.971697529215024e-05, "loss": 0.1132, "step": 8220 }, { "grad_norm": 0.9615933895111084, "learning_rate": 9.971521577099192e-05, "loss": 0.1114, "step": 8230 }, { "grad_norm": 0.8777275085449219, "learning_rate": 9.971345081305236e-05, "loss": 0.1062, "step": 8240 }, { "grad_norm": 0.6593096256256104, "learning_rate": 9.971168041852456e-05, "loss": 0.0912, "step": 8250 }, { "grad_norm": 0.7927566170692444, "learning_rate": 9.970990458760215e-05, "loss": 0.1129, "step": 8260 }, { "grad_norm": 0.8494217991828918, "learning_rate": 9.970812332047929e-05, "loss": 0.1326, "step": 8270 }, { "grad_norm": 0.8364444375038147, "learning_rate": 9.97063366173508e-05, "loss": 0.0989, "step": 8280 }, { "grad_norm": 0.6668582558631897, "learning_rate": 9.970454447841207e-05, "loss": 0.093, "step": 8290 }, { "grad_norm": 0.7417623400688171, "learning_rate": 9.970274690385909e-05, "loss": 0.1168, "step": 8300 }, { "grad_norm": 0.8347209692001343, "learning_rate": 9.970094389388844e-05, "loss": 0.097, "step": 8310 }, { "grad_norm": 0.7736817002296448, "learning_rate": 9.969913544869728e-05, "loss": 0.0843, "step": 8320 }, { "grad_norm": 0.7605167031288147, "learning_rate": 9.96973215684834e-05, "loss": 0.0919, "step": 8330 }, { "grad_norm": 0.7727516889572144, "learning_rate": 9.969550225344513e-05, "loss": 0.1029, "step": 8340 }, { "grad_norm": 0.7069677114486694, "learning_rate": 9.969367750378147e-05, "loss": 0.0975, "step": 8350 }, { "grad_norm": 0.5139187574386597, "learning_rate": 9.969184731969194e-05, "loss": 0.1104, "step": 8360 }, { "grad_norm": 0.7152842283248901, "learning_rate": 9.96900117013767e-05, "loss": 0.102, "step": 8370 }, { "grad_norm": 0.7714475989341736, "learning_rate": 9.96881706490365e-05, "loss": 0.0922, "step": 8380 }, { "grad_norm": 0.7734650373458862, "learning_rate": 9.968632416287265e-05, "loss": 0.1015, "step": 8390 }, { "grad_norm": 0.7066207528114319, "learning_rate": 9.96844722430871e-05, "loss": 0.0922, "step": 8400 }, { "grad_norm": 1.0061308145523071, "learning_rate": 9.968261488988235e-05, "loss": 0.095, "step": 8410 }, { "grad_norm": 0.791141927242279, "learning_rate": 9.968075210346155e-05, "loss": 0.1105, "step": 8420 }, { "grad_norm": 0.6860697865486145, "learning_rate": 9.967888388402839e-05, "loss": 0.11, "step": 8430 }, { "grad_norm": 0.7238338589668274, "learning_rate": 9.967701023178717e-05, "loss": 0.116, "step": 8440 }, { "grad_norm": 0.6999697089195251, "learning_rate": 9.967513114694282e-05, "loss": 0.1105, "step": 8450 }, { "grad_norm": 0.6534947752952576, "learning_rate": 9.967324662970079e-05, "loss": 0.1295, "step": 8460 }, { "grad_norm": 0.8407314419746399, "learning_rate": 9.96713566802672e-05, "loss": 0.1155, "step": 8470 }, { "grad_norm": 0.5190292000770569, "learning_rate": 9.966946129884873e-05, "loss": 0.1078, "step": 8480 }, { "grad_norm": 0.837830126285553, "learning_rate": 9.966756048565265e-05, "loss": 0.1327, "step": 8490 }, { "grad_norm": 0.7224282026290894, "learning_rate": 9.966565424088681e-05, "loss": 0.1091, "step": 8500 }, { "grad_norm": 0.6102551817893982, "learning_rate": 9.96637425647597e-05, "loss": 0.1045, "step": 8510 }, { "grad_norm": 0.7822325229644775, "learning_rate": 9.966182545748038e-05, "loss": 0.1047, "step": 8520 }, { "grad_norm": 0.6670367121696472, "learning_rate": 9.96599029192585e-05, "loss": 0.1005, "step": 8530 }, { "grad_norm": 0.6291562914848328, "learning_rate": 9.965797495030428e-05, "loss": 0.0917, "step": 8540 }, { "grad_norm": 0.7983969449996948, "learning_rate": 9.96560415508286e-05, "loss": 0.0821, "step": 8550 }, { "grad_norm": 0.5035898685455322, "learning_rate": 9.965410272104286e-05, "loss": 0.0915, "step": 8560 }, { "grad_norm": 0.5219632387161255, "learning_rate": 9.96521584611591e-05, "loss": 0.1054, "step": 8570 }, { "grad_norm": 0.5778961777687073, "learning_rate": 9.965020877138994e-05, "loss": 0.0921, "step": 8580 }, { "grad_norm": 0.5173657536506653, "learning_rate": 9.964825365194861e-05, "loss": 0.0881, "step": 8590 }, { "grad_norm": 0.743675708770752, "learning_rate": 9.96462931030489e-05, "loss": 0.0888, "step": 8600 }, { "grad_norm": 0.9445204734802246, "learning_rate": 9.96443271249052e-05, "loss": 0.1063, "step": 8610 }, { "grad_norm": 1.163183331489563, "learning_rate": 9.964235571773255e-05, "loss": 0.1054, "step": 8620 }, { "grad_norm": 1.0260899066925049, "learning_rate": 9.96403788817465e-05, "loss": 0.1551, "step": 8630 }, { "grad_norm": 0.8038184642791748, "learning_rate": 9.963839661716325e-05, "loss": 0.0899, "step": 8640 }, { "grad_norm": 0.65468829870224, "learning_rate": 9.963640892419958e-05, "loss": 0.092, "step": 8650 }, { "grad_norm": 0.6752349138259888, "learning_rate": 9.963441580307286e-05, "loss": 0.1012, "step": 8660 }, { "grad_norm": 0.6656104922294617, "learning_rate": 9.963241725400104e-05, "loss": 0.1078, "step": 8670 }, { "grad_norm": 0.9625158309936523, "learning_rate": 9.963041327720271e-05, "loss": 0.095, "step": 8680 }, { "grad_norm": 0.612346887588501, "learning_rate": 9.962840387289697e-05, "loss": 0.0902, "step": 8690 }, { "grad_norm": 0.8205859065055847, "learning_rate": 9.962638904130363e-05, "loss": 0.0844, "step": 8700 }, { "grad_norm": 0.6791262626647949, "learning_rate": 9.962436878264298e-05, "loss": 0.0974, "step": 8710 }, { "grad_norm": 0.5628489255905151, "learning_rate": 9.962234309713598e-05, "loss": 0.0921, "step": 8720 }, { "grad_norm": 0.7624514698982239, "learning_rate": 9.962031198500414e-05, "loss": 0.0932, "step": 8730 }, { "grad_norm": 0.5268611311912537, "learning_rate": 9.961827544646958e-05, "loss": 0.1035, "step": 8740 }, { "grad_norm": 0.5704571604728699, "learning_rate": 9.961623348175501e-05, "loss": 0.1131, "step": 8750 }, { "grad_norm": 0.7937997579574585, "learning_rate": 9.961418609108377e-05, "loss": 0.0999, "step": 8760 }, { "grad_norm": 0.6081406474113464, "learning_rate": 9.961213327467971e-05, "loss": 0.089, "step": 8770 }, { "grad_norm": 0.8623697757720947, "learning_rate": 9.961007503276736e-05, "loss": 0.0772, "step": 8780 }, { "grad_norm": 0.7318239808082581, "learning_rate": 9.960801136557179e-05, "loss": 0.139, "step": 8790 }, { "grad_norm": 0.6521780490875244, "learning_rate": 9.960594227331866e-05, "loss": 0.0779, "step": 8800 }, { "grad_norm": 0.6897981762886047, "learning_rate": 9.960386775623429e-05, "loss": 0.1067, "step": 8810 }, { "grad_norm": 0.6679515242576599, "learning_rate": 9.96017878145455e-05, "loss": 0.1036, "step": 8820 }, { "grad_norm": 0.8250641226768494, "learning_rate": 9.959970244847977e-05, "loss": 0.1242, "step": 8830 }, { "grad_norm": 1.1327412128448486, "learning_rate": 9.959761165826518e-05, "loss": 0.1268, "step": 8840 }, { "grad_norm": 0.7180212140083313, "learning_rate": 9.959551544413033e-05, "loss": 0.1134, "step": 8850 }, { "grad_norm": 0.6872667670249939, "learning_rate": 9.959341380630448e-05, "loss": 0.0877, "step": 8860 }, { "grad_norm": 1.152207851409912, "learning_rate": 9.959130674501746e-05, "loss": 0.1267, "step": 8870 }, { "grad_norm": 0.9801048636436462, "learning_rate": 9.958919426049968e-05, "loss": 0.1186, "step": 8880 }, { "grad_norm": 0.7497374415397644, "learning_rate": 9.958707635298219e-05, "loss": 0.1134, "step": 8890 }, { "grad_norm": 0.8693662881851196, "learning_rate": 9.958495302269657e-05, "loss": 0.1217, "step": 8900 }, { "grad_norm": 0.7025782465934753, "learning_rate": 9.958282426987503e-05, "loss": 0.1221, "step": 8910 }, { "grad_norm": 0.5688144564628601, "learning_rate": 9.95806900947504e-05, "loss": 0.097, "step": 8920 }, { "grad_norm": 0.8843528032302856, "learning_rate": 9.957855049755604e-05, "loss": 0.1082, "step": 8930 }, { "grad_norm": 0.6852983236312866, "learning_rate": 9.957640547852593e-05, "loss": 0.0891, "step": 8940 }, { "grad_norm": 0.7827601432800293, "learning_rate": 9.957425503789466e-05, "loss": 0.096, "step": 8950 }, { "grad_norm": 0.4809911847114563, "learning_rate": 9.957209917589738e-05, "loss": 0.1143, "step": 8960 }, { "grad_norm": 0.6981006264686584, "learning_rate": 9.956993789276987e-05, "loss": 0.1138, "step": 8970 }, { "grad_norm": 0.7714274525642395, "learning_rate": 9.956777118874847e-05, "loss": 0.0977, "step": 8980 }, { "grad_norm": 0.8096242547035217, "learning_rate": 9.956559906407016e-05, "loss": 0.1077, "step": 8990 }, { "grad_norm": 0.7002239227294922, "learning_rate": 9.956342151897245e-05, "loss": 0.092, "step": 9000 }, { "grad_norm": 0.718513548374176, "learning_rate": 9.956123855369346e-05, "loss": 0.1243, "step": 9010 }, { "grad_norm": 0.6028826236724854, "learning_rate": 9.955905016847196e-05, "loss": 0.1093, "step": 9020 }, { "grad_norm": 0.622805655002594, "learning_rate": 9.955685636354723e-05, "loss": 0.1047, "step": 9030 }, { "grad_norm": 0.7161643505096436, "learning_rate": 9.95546571391592e-05, "loss": 0.1238, "step": 9040 }, { "grad_norm": 0.7353559136390686, "learning_rate": 9.955245249554837e-05, "loss": 0.0977, "step": 9050 }, { "grad_norm": 0.822711169719696, "learning_rate": 9.955024243295582e-05, "loss": 0.1098, "step": 9060 }, { "grad_norm": 0.6552707552909851, "learning_rate": 9.954802695162328e-05, "loss": 0.1305, "step": 9070 }, { "grad_norm": 0.8241406679153442, "learning_rate": 9.954580605179302e-05, "loss": 0.1057, "step": 9080 }, { "grad_norm": 0.579561173915863, "learning_rate": 9.954357973370788e-05, "loss": 0.1195, "step": 9090 }, { "grad_norm": 0.6531955599784851, "learning_rate": 9.954134799761135e-05, "loss": 0.0828, "step": 9100 }, { "grad_norm": 0.671127438545227, "learning_rate": 9.953911084374748e-05, "loss": 0.0793, "step": 9110 }, { "grad_norm": 0.6417557001113892, "learning_rate": 9.953686827236093e-05, "loss": 0.0997, "step": 9120 }, { "grad_norm": 0.636294960975647, "learning_rate": 9.953462028369695e-05, "loss": 0.1074, "step": 9130 }, { "grad_norm": 0.6858224272727966, "learning_rate": 9.953236687800136e-05, "loss": 0.1183, "step": 9140 }, { "grad_norm": 0.6135473847389221, "learning_rate": 9.95301080555206e-05, "loss": 0.0834, "step": 9150 }, { "grad_norm": 0.6411691308021545, "learning_rate": 9.952784381650171e-05, "loss": 0.1094, "step": 9160 }, { "grad_norm": 0.990227460861206, "learning_rate": 9.952557416119226e-05, "loss": 0.1, "step": 9170 }, { "grad_norm": 0.7935163974761963, "learning_rate": 9.95232990898405e-05, "loss": 0.1087, "step": 9180 }, { "grad_norm": 0.8039679527282715, "learning_rate": 9.95210186026952e-05, "loss": 0.1176, "step": 9190 }, { "grad_norm": 0.8036527633666992, "learning_rate": 9.951873270000576e-05, "loss": 0.0933, "step": 9200 }, { "grad_norm": 0.8191133737564087, "learning_rate": 9.951644138202216e-05, "loss": 0.0889, "step": 9210 }, { "grad_norm": 0.8390368223190308, "learning_rate": 9.951414464899498e-05, "loss": 0.0937, "step": 9220 }, { "grad_norm": 0.6015384197235107, "learning_rate": 9.951184250117538e-05, "loss": 0.0729, "step": 9230 }, { "grad_norm": 0.7270970344543457, "learning_rate": 9.950953493881513e-05, "loss": 0.0995, "step": 9240 }, { "grad_norm": 0.756011426448822, "learning_rate": 9.950722196216658e-05, "loss": 0.0854, "step": 9250 }, { "grad_norm": 0.9864425659179688, "learning_rate": 9.950490357148265e-05, "loss": 0.1149, "step": 9260 }, { "grad_norm": 0.6522082686424255, "learning_rate": 9.950257976701692e-05, "loss": 0.0957, "step": 9270 }, { "grad_norm": 0.6664878129959106, "learning_rate": 9.950025054902348e-05, "loss": 0.1043, "step": 9280 }, { "grad_norm": 0.5951300263404846, "learning_rate": 9.949791591775706e-05, "loss": 0.0932, "step": 9290 }, { "grad_norm": 0.9306336641311646, "learning_rate": 9.949557587347298e-05, "loss": 0.0858, "step": 9300 }, { "grad_norm": 0.6356043219566345, "learning_rate": 9.949323041642713e-05, "loss": 0.0992, "step": 9310 }, { "grad_norm": 0.8273126482963562, "learning_rate": 9.949087954687602e-05, "loss": 0.0878, "step": 9320 }, { "grad_norm": 1.3741284608840942, "learning_rate": 9.948852326507672e-05, "loss": 0.109, "step": 9330 }, { "grad_norm": 1.080744743347168, "learning_rate": 9.948616157128694e-05, "loss": 0.098, "step": 9340 }, { "grad_norm": 0.7364424467086792, "learning_rate": 9.948379446576493e-05, "loss": 0.108, "step": 9350 }, { "grad_norm": 0.5593907833099365, "learning_rate": 9.948142194876952e-05, "loss": 0.1024, "step": 9360 }, { "grad_norm": 0.6034523844718933, "learning_rate": 9.947904402056024e-05, "loss": 0.0928, "step": 9370 }, { "grad_norm": 0.6886141300201416, "learning_rate": 9.947666068139708e-05, "loss": 0.1086, "step": 9380 }, { "grad_norm": 0.7772342562675476, "learning_rate": 9.947427193154071e-05, "loss": 0.1052, "step": 9390 }, { "grad_norm": 0.7106131911277771, "learning_rate": 9.947187777125233e-05, "loss": 0.0975, "step": 9400 }, { "grad_norm": 0.9224360585212708, "learning_rate": 9.946947820079377e-05, "loss": 0.0882, "step": 9410 }, { "grad_norm": 0.8169370889663696, "learning_rate": 9.946707322042747e-05, "loss": 0.1126, "step": 9420 }, { "grad_norm": 0.6158862709999084, "learning_rate": 9.94646628304164e-05, "loss": 0.0933, "step": 9430 }, { "grad_norm": 0.6766100525856018, "learning_rate": 9.946224703102418e-05, "loss": 0.1032, "step": 9440 }, { "grad_norm": 0.8435202240943909, "learning_rate": 9.945982582251498e-05, "loss": 0.1027, "step": 9450 }, { "grad_norm": 0.5923638939857483, "learning_rate": 9.94573992051536e-05, "loss": 0.1367, "step": 9460 }, { "grad_norm": 0.9487195014953613, "learning_rate": 9.94549671792054e-05, "loss": 0.1247, "step": 9470 }, { "grad_norm": 0.5976546406745911, "learning_rate": 9.945252974493635e-05, "loss": 0.0811, "step": 9480 }, { "grad_norm": 0.5575807094573975, "learning_rate": 9.9450086902613e-05, "loss": 0.0866, "step": 9490 }, { "grad_norm": 0.7826381921768188, "learning_rate": 9.944763865250248e-05, "loss": 0.0927, "step": 9500 }, { "grad_norm": 1.4499382972717285, "learning_rate": 9.944518499487254e-05, "loss": 0.1055, "step": 9510 }, { "grad_norm": 0.7266030311584473, "learning_rate": 9.944272592999151e-05, "loss": 0.1195, "step": 9520 }, { "grad_norm": 0.6813144087791443, "learning_rate": 9.94402614581283e-05, "loss": 0.1084, "step": 9530 }, { "grad_norm": 0.7864287495613098, "learning_rate": 9.943779157955244e-05, "loss": 0.0933, "step": 9540 }, { "grad_norm": 0.8016054034233093, "learning_rate": 9.943531629453403e-05, "loss": 0.1276, "step": 9550 }, { "grad_norm": 1.0885285139083862, "learning_rate": 9.943283560334375e-05, "loss": 0.1059, "step": 9560 }, { "grad_norm": 1.0552014112472534, "learning_rate": 9.943034950625288e-05, "loss": 0.0927, "step": 9570 }, { "grad_norm": 0.4346337616443634, "learning_rate": 9.942785800353332e-05, "loss": 0.0897, "step": 9580 }, { "grad_norm": 0.7084578275680542, "learning_rate": 9.942536109545751e-05, "loss": 0.0985, "step": 9590 }, { "grad_norm": 0.6862210631370544, "learning_rate": 9.942285878229853e-05, "loss": 0.0992, "step": 9600 }, { "grad_norm": 0.8918427228927612, "learning_rate": 9.942035106433001e-05, "loss": 0.094, "step": 9610 }, { "grad_norm": 0.7727041244506836, "learning_rate": 9.94178379418262e-05, "loss": 0.0921, "step": 9620 }, { "grad_norm": 0.774603545665741, "learning_rate": 9.941531941506194e-05, "loss": 0.0931, "step": 9630 }, { "grad_norm": 0.6606017351150513, "learning_rate": 9.941279548431263e-05, "loss": 0.1214, "step": 9640 }, { "grad_norm": 0.944131076335907, "learning_rate": 9.941026614985431e-05, "loss": 0.0989, "step": 9650 }, { "grad_norm": 0.5593726634979248, "learning_rate": 9.940773141196357e-05, "loss": 0.1057, "step": 9660 }, { "grad_norm": 0.7688432335853577, "learning_rate": 9.94051912709176e-05, "loss": 0.0985, "step": 9670 }, { "grad_norm": 0.6819236278533936, "learning_rate": 9.940264572699421e-05, "loss": 0.0937, "step": 9680 }, { "grad_norm": 0.5919490456581116, "learning_rate": 9.940009478047174e-05, "loss": 0.1014, "step": 9690 }, { "grad_norm": 1.1063140630722046, "learning_rate": 9.939753843162918e-05, "loss": 0.1015, "step": 9700 }, { "grad_norm": 0.6042591333389282, "learning_rate": 9.939497668074609e-05, "loss": 0.1003, "step": 9710 }, { "grad_norm": 0.9724153876304626, "learning_rate": 9.93924095281026e-05, "loss": 0.0919, "step": 9720 }, { "grad_norm": 0.5387865900993347, "learning_rate": 9.938983697397948e-05, "loss": 0.1021, "step": 9730 }, { "grad_norm": 0.5631493926048279, "learning_rate": 9.938725901865805e-05, "loss": 0.0815, "step": 9740 }, { "grad_norm": 0.8214746713638306, "learning_rate": 9.93846756624202e-05, "loss": 0.0891, "step": 9750 }, { "grad_norm": 0.5336772799491882, "learning_rate": 9.938208690554849e-05, "loss": 0.0957, "step": 9760 }, { "grad_norm": 0.6339232325553894, "learning_rate": 9.9379492748326e-05, "loss": 0.1156, "step": 9770 }, { "grad_norm": 0.6766915321350098, "learning_rate": 9.937689319103641e-05, "loss": 0.1023, "step": 9780 }, { "grad_norm": 0.5530292987823486, "learning_rate": 9.937428823396404e-05, "loss": 0.0938, "step": 9790 }, { "grad_norm": 0.5837986469268799, "learning_rate": 9.937167787739372e-05, "loss": 0.0818, "step": 9800 }, { "grad_norm": 0.7374151945114136, "learning_rate": 9.936906212161095e-05, "loss": 0.0914, "step": 9810 }, { "grad_norm": 0.5743416547775269, "learning_rate": 9.936644096690176e-05, "loss": 0.0959, "step": 9820 }, { "grad_norm": 0.8877694010734558, "learning_rate": 9.936381441355282e-05, "loss": 0.1274, "step": 9830 }, { "grad_norm": 0.6050177812576294, "learning_rate": 9.936118246185136e-05, "loss": 0.0865, "step": 9840 }, { "grad_norm": 0.6698606610298157, "learning_rate": 9.935854511208518e-05, "loss": 0.0885, "step": 9850 }, { "grad_norm": 0.9347379207611084, "learning_rate": 9.935590236454272e-05, "loss": 0.1059, "step": 9860 }, { "grad_norm": 1.0032099485397339, "learning_rate": 9.935325421951298e-05, "loss": 0.0955, "step": 9870 }, { "grad_norm": 0.594611644744873, "learning_rate": 9.935060067728557e-05, "loss": 0.0814, "step": 9880 }, { "grad_norm": 0.791086733341217, "learning_rate": 9.934794173815067e-05, "loss": 0.1086, "step": 9890 }, { "grad_norm": 0.7268203496932983, "learning_rate": 9.934527740239906e-05, "loss": 0.0909, "step": 9900 }, { "grad_norm": 0.7075511813163757, "learning_rate": 9.934260767032209e-05, "loss": 0.0963, "step": 9910 }, { "grad_norm": 1.0830473899841309, "learning_rate": 9.933993254221172e-05, "loss": 0.1105, "step": 9920 }, { "grad_norm": 0.7769351005554199, "learning_rate": 9.933725201836053e-05, "loss": 0.1127, "step": 9930 }, { "grad_norm": 0.5491613745689392, "learning_rate": 9.933456609906162e-05, "loss": 0.1064, "step": 9940 }, { "grad_norm": 0.8459573984146118, "learning_rate": 9.933187478460875e-05, "loss": 0.1089, "step": 9950 }, { "grad_norm": 0.7941088080406189, "learning_rate": 9.93291780752962e-05, "loss": 0.0941, "step": 9960 }, { "grad_norm": 0.6158515214920044, "learning_rate": 9.932647597141893e-05, "loss": 0.1009, "step": 9970 }, { "grad_norm": 0.6712644100189209, "learning_rate": 9.932376847327239e-05, "loss": 0.1109, "step": 9980 }, { "grad_norm": 0.6559886932373047, "learning_rate": 9.932105558115268e-05, "loss": 0.077, "step": 9990 }, { "grad_norm": 0.7052610516548157, "learning_rate": 9.931833729535651e-05, "loss": 0.1128, "step": 10000 }, { "grad_norm": 0.8831080794334412, "learning_rate": 9.931561361618111e-05, "loss": 0.098, "step": 10010 }, { "grad_norm": 0.6475574970245361, "learning_rate": 9.931288454392435e-05, "loss": 0.0874, "step": 10020 }, { "grad_norm": 0.5243498086929321, "learning_rate": 9.931015007888467e-05, "loss": 0.0857, "step": 10030 }, { "grad_norm": 0.6990725994110107, "learning_rate": 9.930741022136112e-05, "loss": 0.1088, "step": 10040 }, { "grad_norm": 0.683612048625946, "learning_rate": 9.930466497165333e-05, "loss": 0.0944, "step": 10050 }, { "grad_norm": 0.7292007803916931, "learning_rate": 9.93019143300615e-05, "loss": 0.0897, "step": 10060 }, { "grad_norm": 0.8674601912498474, "learning_rate": 9.929915829688644e-05, "loss": 0.0736, "step": 10070 }, { "grad_norm": 0.6011753082275391, "learning_rate": 9.929639687242955e-05, "loss": 0.101, "step": 10080 }, { "grad_norm": 0.5559348464012146, "learning_rate": 9.929363005699281e-05, "loss": 0.116, "step": 10090 }, { "grad_norm": 0.7091989517211914, "learning_rate": 9.92908578508788e-05, "loss": 0.098, "step": 10100 }, { "grad_norm": 1.0187304019927979, "learning_rate": 9.928808025439069e-05, "loss": 0.0773, "step": 10110 }, { "grad_norm": 0.5846640467643738, "learning_rate": 9.928529726783223e-05, "loss": 0.0759, "step": 10120 }, { "grad_norm": 0.6122124791145325, "learning_rate": 9.928250889150774e-05, "loss": 0.1214, "step": 10130 }, { "grad_norm": 0.632436990737915, "learning_rate": 9.92797151257222e-05, "loss": 0.1163, "step": 10140 }, { "grad_norm": 0.6434361934661865, "learning_rate": 9.927691597078108e-05, "loss": 0.1204, "step": 10150 }, { "grad_norm": 0.9273130297660828, "learning_rate": 9.927411142699053e-05, "loss": 0.0948, "step": 10160 }, { "grad_norm": 0.7143503427505493, "learning_rate": 9.927130149465725e-05, "loss": 0.0838, "step": 10170 }, { "grad_norm": 0.7318591475486755, "learning_rate": 9.92684861740885e-05, "loss": 0.0827, "step": 10180 }, { "grad_norm": 0.6795605421066284, "learning_rate": 9.926566546559217e-05, "loss": 0.0799, "step": 10190 }, { "grad_norm": 0.6757745742797852, "learning_rate": 9.926283936947673e-05, "loss": 0.0849, "step": 10200 }, { "grad_norm": 0.5569467544555664, "learning_rate": 9.926000788605126e-05, "loss": 0.0815, "step": 10210 }, { "grad_norm": 0.6219619512557983, "learning_rate": 9.92571710156254e-05, "loss": 0.1157, "step": 10220 }, { "grad_norm": 0.6906128525733948, "learning_rate": 9.925432875850936e-05, "loss": 0.1184, "step": 10230 }, { "grad_norm": 1.0247493982315063, "learning_rate": 9.925148111501396e-05, "loss": 0.1222, "step": 10240 }, { "grad_norm": 1.1580851078033447, "learning_rate": 9.924862808545066e-05, "loss": 0.1044, "step": 10250 }, { "grad_norm": 0.8616909980773926, "learning_rate": 9.924576967013141e-05, "loss": 0.1267, "step": 10260 }, { "grad_norm": 0.7946112751960754, "learning_rate": 9.924290586936887e-05, "loss": 0.1066, "step": 10270 }, { "grad_norm": 0.598859429359436, "learning_rate": 9.924003668347614e-05, "loss": 0.0969, "step": 10280 }, { "grad_norm": 0.6558599472045898, "learning_rate": 9.923716211276704e-05, "loss": 0.0845, "step": 10290 }, { "grad_norm": 0.9024509787559509, "learning_rate": 9.923428215755594e-05, "loss": 0.0745, "step": 10300 }, { "grad_norm": 0.7386588454246521, "learning_rate": 9.923139681815775e-05, "loss": 0.1169, "step": 10310 }, { "grad_norm": 0.54670250415802, "learning_rate": 9.922850609488801e-05, "loss": 0.0961, "step": 10320 }, { "grad_norm": 0.6917989253997803, "learning_rate": 9.922560998806287e-05, "loss": 0.1276, "step": 10330 }, { "grad_norm": 0.6885901689529419, "learning_rate": 9.922270849799905e-05, "loss": 0.0895, "step": 10340 }, { "grad_norm": 0.7626888155937195, "learning_rate": 9.92198016250138e-05, "loss": 0.0847, "step": 10350 }, { "grad_norm": 0.6349205374717712, "learning_rate": 9.921688936942506e-05, "loss": 0.0894, "step": 10360 }, { "grad_norm": 1.0068429708480835, "learning_rate": 9.921397173155129e-05, "loss": 0.1238, "step": 10370 }, { "grad_norm": 1.0163800716400146, "learning_rate": 9.921104871171157e-05, "loss": 0.1427, "step": 10380 }, { "grad_norm": 0.9335481524467468, "learning_rate": 9.920812031022554e-05, "loss": 0.1026, "step": 10390 }, { "grad_norm": 0.5322441458702087, "learning_rate": 9.920518652741348e-05, "loss": 0.0909, "step": 10400 }, { "grad_norm": 0.8925983309745789, "learning_rate": 9.920224736359618e-05, "loss": 0.1003, "step": 10410 }, { "grad_norm": 0.7834806442260742, "learning_rate": 9.91993028190951e-05, "loss": 0.108, "step": 10420 }, { "grad_norm": 0.7792406678199768, "learning_rate": 9.919635289423222e-05, "loss": 0.107, "step": 10430 }, { "grad_norm": 0.6226374506950378, "learning_rate": 9.919339758933015e-05, "loss": 0.0877, "step": 10440 }, { "grad_norm": 0.7339866757392883, "learning_rate": 9.919043690471209e-05, "loss": 0.0811, "step": 10450 }, { "grad_norm": 0.6106528043746948, "learning_rate": 9.91874708407018e-05, "loss": 0.0947, "step": 10460 }, { "grad_norm": 0.6227996349334717, "learning_rate": 9.918449939762367e-05, "loss": 0.1104, "step": 10470 }, { "grad_norm": 0.47312209010124207, "learning_rate": 9.91815225758026e-05, "loss": 0.0814, "step": 10480 }, { "grad_norm": 0.5865525603294373, "learning_rate": 9.917854037556419e-05, "loss": 0.116, "step": 10490 }, { "grad_norm": 0.8323413133621216, "learning_rate": 9.917555279723454e-05, "loss": 0.1135, "step": 10500 }, { "grad_norm": 0.6536474823951721, "learning_rate": 9.917255984114036e-05, "loss": 0.0808, "step": 10510 }, { "grad_norm": 0.7488875985145569, "learning_rate": 9.916956150760896e-05, "loss": 0.0942, "step": 10520 }, { "grad_norm": 0.9507960081100464, "learning_rate": 9.916655779696826e-05, "loss": 0.1074, "step": 10530 }, { "grad_norm": 0.7967543601989746, "learning_rate": 9.916354870954671e-05, "loss": 0.0919, "step": 10540 }, { "grad_norm": 0.6412456631660461, "learning_rate": 9.91605342456734e-05, "loss": 0.1116, "step": 10550 }, { "grad_norm": 0.7641332149505615, "learning_rate": 9.915751440567795e-05, "loss": 0.0992, "step": 10560 }, { "grad_norm": 0.5693318247795105, "learning_rate": 9.915448918989066e-05, "loss": 0.1126, "step": 10570 }, { "grad_norm": 0.8135589361190796, "learning_rate": 9.915145859864232e-05, "loss": 0.0844, "step": 10580 }, { "grad_norm": 0.7450034618377686, "learning_rate": 9.914842263226437e-05, "loss": 0.0844, "step": 10590 }, { "grad_norm": 0.5833271741867065, "learning_rate": 9.914538129108882e-05, "loss": 0.0815, "step": 10600 }, { "grad_norm": 0.6300477981567383, "learning_rate": 9.914233457544825e-05, "loss": 0.0881, "step": 10610 }, { "grad_norm": 0.7246978282928467, "learning_rate": 9.913928248567586e-05, "loss": 0.0773, "step": 10620 }, { "grad_norm": 0.6122862100601196, "learning_rate": 9.913622502210542e-05, "loss": 0.0959, "step": 10630 }, { "grad_norm": 0.7009672522544861, "learning_rate": 9.913316218507128e-05, "loss": 0.0851, "step": 10640 }, { "grad_norm": 0.7735552191734314, "learning_rate": 9.91300939749084e-05, "loss": 0.1091, "step": 10650 }, { "grad_norm": 0.8672723174095154, "learning_rate": 9.91270203919523e-05, "loss": 0.0954, "step": 10660 }, { "grad_norm": 0.7132527828216553, "learning_rate": 9.912394143653912e-05, "loss": 0.0796, "step": 10670 }, { "grad_norm": 0.7497026920318604, "learning_rate": 9.912085710900555e-05, "loss": 0.0945, "step": 10680 }, { "grad_norm": 0.5969906449317932, "learning_rate": 9.911776740968892e-05, "loss": 0.0699, "step": 10690 }, { "grad_norm": 1.0556495189666748, "learning_rate": 9.911467233892709e-05, "loss": 0.0857, "step": 10700 }, { "grad_norm": 0.6905048489570618, "learning_rate": 9.911157189705853e-05, "loss": 0.0808, "step": 10710 }, { "grad_norm": 0.5863040089607239, "learning_rate": 9.910846608442229e-05, "loss": 0.1021, "step": 10720 }, { "grad_norm": 0.5892178416252136, "learning_rate": 9.910535490135805e-05, "loss": 0.0742, "step": 10730 }, { "grad_norm": 0.8359094262123108, "learning_rate": 9.910223834820603e-05, "loss": 0.0944, "step": 10740 }, { "grad_norm": 0.6058107614517212, "learning_rate": 9.909911642530703e-05, "loss": 0.0771, "step": 10750 }, { "grad_norm": 0.563774824142456, "learning_rate": 9.909598913300249e-05, "loss": 0.078, "step": 10760 }, { "grad_norm": 0.8015226125717163, "learning_rate": 9.909285647163438e-05, "loss": 0.0939, "step": 10770 }, { "grad_norm": 0.8552261590957642, "learning_rate": 9.908971844154531e-05, "loss": 0.0874, "step": 10780 }, { "grad_norm": 0.7978106141090393, "learning_rate": 9.908657504307843e-05, "loss": 0.1064, "step": 10790 }, { "grad_norm": 0.6496439576148987, "learning_rate": 9.908342627657751e-05, "loss": 0.0888, "step": 10800 }, { "grad_norm": 0.9398994445800781, "learning_rate": 9.908027214238689e-05, "loss": 0.0829, "step": 10810 }, { "grad_norm": 0.8669595718383789, "learning_rate": 9.90771126408515e-05, "loss": 0.0883, "step": 10820 }, { "grad_norm": 0.6220933794975281, "learning_rate": 9.907394777231685e-05, "loss": 0.102, "step": 10830 }, { "grad_norm": 0.9405134916305542, "learning_rate": 9.907077753712905e-05, "loss": 0.0886, "step": 10840 }, { "grad_norm": 0.7304231524467468, "learning_rate": 9.906760193563482e-05, "loss": 0.0859, "step": 10850 }, { "grad_norm": 0.6230496764183044, "learning_rate": 9.906442096818139e-05, "loss": 0.0979, "step": 10860 }, { "grad_norm": 0.6590986847877502, "learning_rate": 9.906123463511665e-05, "loss": 0.1261, "step": 10870 }, { "grad_norm": 0.7309950590133667, "learning_rate": 9.905804293678907e-05, "loss": 0.0912, "step": 10880 }, { "grad_norm": 0.9241798520088196, "learning_rate": 9.905484587354766e-05, "loss": 0.0824, "step": 10890 }, { "grad_norm": 0.5512362122535706, "learning_rate": 9.905164344574205e-05, "loss": 0.0776, "step": 10900 }, { "grad_norm": 0.6941649317741394, "learning_rate": 9.904843565372248e-05, "loss": 0.1011, "step": 10910 }, { "grad_norm": 0.7262336611747742, "learning_rate": 9.904522249783972e-05, "loss": 0.084, "step": 10920 }, { "grad_norm": 0.7727411985397339, "learning_rate": 9.904200397844517e-05, "loss": 0.1092, "step": 10930 }, { "grad_norm": 0.67280113697052, "learning_rate": 9.903878009589078e-05, "loss": 0.0906, "step": 10940 }, { "grad_norm": 0.5591420531272888, "learning_rate": 9.903555085052915e-05, "loss": 0.0966, "step": 10950 }, { "grad_norm": 0.7549257278442383, "learning_rate": 9.903231624271338e-05, "loss": 0.1033, "step": 10960 }, { "grad_norm": 0.7125890254974365, "learning_rate": 9.902907627279724e-05, "loss": 0.0911, "step": 10970 }, { "grad_norm": 0.805566668510437, "learning_rate": 9.902583094113504e-05, "loss": 0.0966, "step": 10980 }, { "grad_norm": 0.8194602727890015, "learning_rate": 9.902258024808168e-05, "loss": 0.0998, "step": 10990 }, { "grad_norm": 0.6936808824539185, "learning_rate": 9.901932419399264e-05, "loss": 0.1096, "step": 11000 }, { "grad_norm": 0.8833263516426086, "learning_rate": 9.9016062779224e-05, "loss": 0.1078, "step": 11010 }, { "grad_norm": 0.8219338655471802, "learning_rate": 9.901279600413242e-05, "loss": 0.0949, "step": 11020 }, { "grad_norm": 0.7093368172645569, "learning_rate": 9.900952386907518e-05, "loss": 0.0959, "step": 11030 }, { "grad_norm": 0.7903746366500854, "learning_rate": 9.90062463744101e-05, "loss": 0.0905, "step": 11040 }, { "grad_norm": 0.4396240711212158, "learning_rate": 9.900296352049558e-05, "loss": 0.0953, "step": 11050 }, { "grad_norm": 0.627068817615509, "learning_rate": 9.899967530769065e-05, "loss": 0.0913, "step": 11060 }, { "grad_norm": 0.5823429226875305, "learning_rate": 9.899638173635489e-05, "loss": 0.0749, "step": 11070 }, { "grad_norm": 0.7402346730232239, "learning_rate": 9.899308280684849e-05, "loss": 0.0828, "step": 11080 }, { "grad_norm": 0.5673876404762268, "learning_rate": 9.898977851953222e-05, "loss": 0.0963, "step": 11090 }, { "grad_norm": 0.6602680087089539, "learning_rate": 9.898646887476741e-05, "loss": 0.1009, "step": 11100 }, { "grad_norm": 0.9197094440460205, "learning_rate": 9.898315387291603e-05, "loss": 0.1048, "step": 11110 }, { "grad_norm": 0.7837830781936646, "learning_rate": 9.89798335143406e-05, "loss": 0.0827, "step": 11120 }, { "grad_norm": 0.658967912197113, "learning_rate": 9.897650779940419e-05, "loss": 0.0936, "step": 11130 }, { "grad_norm": 0.6112364530563354, "learning_rate": 9.897317672847054e-05, "loss": 0.0891, "step": 11140 }, { "grad_norm": 0.8724700212478638, "learning_rate": 9.89698403019039e-05, "loss": 0.1078, "step": 11150 }, { "grad_norm": 0.820989191532135, "learning_rate": 9.896649852006917e-05, "loss": 0.1163, "step": 11160 }, { "grad_norm": 0.4851023852825165, "learning_rate": 9.896315138333177e-05, "loss": 0.0738, "step": 11170 }, { "grad_norm": 0.8140140175819397, "learning_rate": 9.895979889205774e-05, "loss": 0.0828, "step": 11180 }, { "grad_norm": 0.7151710391044617, "learning_rate": 9.895644104661372e-05, "loss": 0.0952, "step": 11190 }, { "grad_norm": 0.5379443764686584, "learning_rate": 9.895307784736691e-05, "loss": 0.0875, "step": 11200 }, { "grad_norm": 0.9098193645477295, "learning_rate": 9.894970929468512e-05, "loss": 0.1249, "step": 11210 }, { "grad_norm": 0.5850266218185425, "learning_rate": 9.89463353889367e-05, "loss": 0.0806, "step": 11220 }, { "grad_norm": 0.9088364839553833, "learning_rate": 9.894295613049065e-05, "loss": 0.0815, "step": 11230 }, { "grad_norm": 0.696374773979187, "learning_rate": 9.893957151971649e-05, "loss": 0.0757, "step": 11240 }, { "grad_norm": 0.7791599631309509, "learning_rate": 9.893618155698436e-05, "loss": 0.0731, "step": 11250 }, { "grad_norm": 0.9247416257858276, "learning_rate": 9.8932786242665e-05, "loss": 0.087, "step": 11260 }, { "grad_norm": 0.8178786635398865, "learning_rate": 9.89293855771297e-05, "loss": 0.0921, "step": 11270 }, { "grad_norm": 0.8606494665145874, "learning_rate": 9.892597956075036e-05, "loss": 0.0882, "step": 11280 }, { "grad_norm": 0.8249112367630005, "learning_rate": 9.892256819389947e-05, "loss": 0.0833, "step": 11290 }, { "grad_norm": 0.537717878818512, "learning_rate": 9.891915147695006e-05, "loss": 0.1189, "step": 11300 }, { "grad_norm": 0.7175687551498413, "learning_rate": 9.891572941027577e-05, "loss": 0.0944, "step": 11310 }, { "grad_norm": 0.7245373129844666, "learning_rate": 9.89123019942509e-05, "loss": 0.0873, "step": 11320 }, { "grad_norm": 0.9567139744758606, "learning_rate": 9.89088692292502e-05, "loss": 0.1027, "step": 11330 }, { "grad_norm": 0.9378630518913269, "learning_rate": 9.89054311156491e-05, "loss": 0.0985, "step": 11340 }, { "grad_norm": 0.6338625550270081, "learning_rate": 9.890198765382357e-05, "loss": 0.0846, "step": 11350 }, { "grad_norm": 0.7017553448677063, "learning_rate": 9.889853884415021e-05, "loss": 0.1149, "step": 11360 }, { "grad_norm": 0.7856931686401367, "learning_rate": 9.889508468700614e-05, "loss": 0.0692, "step": 11370 }, { "grad_norm": 0.4726383090019226, "learning_rate": 9.889162518276915e-05, "loss": 0.0996, "step": 11380 }, { "grad_norm": 0.9675601124763489, "learning_rate": 9.888816033181752e-05, "loss": 0.0881, "step": 11390 }, { "grad_norm": 0.6002920269966125, "learning_rate": 9.888469013453018e-05, "loss": 0.0965, "step": 11400 }, { "grad_norm": 0.6597592234611511, "learning_rate": 9.888121459128663e-05, "loss": 0.0943, "step": 11410 }, { "grad_norm": 0.5260646343231201, "learning_rate": 9.887773370246693e-05, "loss": 0.1059, "step": 11420 }, { "grad_norm": 0.7883245348930359, "learning_rate": 9.887424746845177e-05, "loss": 0.0986, "step": 11430 }, { "grad_norm": 0.9252811670303345, "learning_rate": 9.887075588962239e-05, "loss": 0.0949, "step": 11440 }, { "grad_norm": 0.6916220784187317, "learning_rate": 9.88672589663606e-05, "loss": 0.0984, "step": 11450 }, { "grad_norm": 0.7035229802131653, "learning_rate": 9.886375669904886e-05, "loss": 0.0935, "step": 11460 }, { "grad_norm": 1.0996612310409546, "learning_rate": 9.886024908807014e-05, "loss": 0.0942, "step": 11470 }, { "grad_norm": 0.6782337427139282, "learning_rate": 9.885673613380806e-05, "loss": 0.0957, "step": 11480 }, { "grad_norm": 0.6553600430488586, "learning_rate": 9.885321783664676e-05, "loss": 0.0959, "step": 11490 }, { "grad_norm": 0.4976503551006317, "learning_rate": 9.884969419697101e-05, "loss": 0.0773, "step": 11500 }, { "grad_norm": 0.5589221715927124, "learning_rate": 9.884616521516614e-05, "loss": 0.0928, "step": 11510 }, { "grad_norm": 0.6925768256187439, "learning_rate": 9.88426308916181e-05, "loss": 0.0849, "step": 11520 }, { "grad_norm": 0.9149131178855896, "learning_rate": 9.883909122671335e-05, "loss": 0.0787, "step": 11530 }, { "grad_norm": 0.6961629986763, "learning_rate": 9.883554622083904e-05, "loss": 0.1079, "step": 11540 }, { "grad_norm": 1.0993218421936035, "learning_rate": 9.88319958743828e-05, "loss": 0.1201, "step": 11550 }, { "grad_norm": 0.5884571671485901, "learning_rate": 9.882844018773291e-05, "loss": 0.0961, "step": 11560 }, { "grad_norm": 0.8506823778152466, "learning_rate": 9.882487916127823e-05, "loss": 0.083, "step": 11570 }, { "grad_norm": 0.8243205547332764, "learning_rate": 9.882131279540815e-05, "loss": 0.0858, "step": 11580 }, { "grad_norm": 0.6623758673667908, "learning_rate": 9.881774109051271e-05, "loss": 0.1243, "step": 11590 }, { "grad_norm": 0.9279493093490601, "learning_rate": 9.881416404698252e-05, "loss": 0.092, "step": 11600 }, { "grad_norm": 0.47777611017227173, "learning_rate": 9.881058166520873e-05, "loss": 0.088, "step": 11610 }, { "grad_norm": 0.6199489235877991, "learning_rate": 9.880699394558311e-05, "loss": 0.0713, "step": 11620 }, { "grad_norm": 0.6886093616485596, "learning_rate": 9.880340088849801e-05, "loss": 0.0699, "step": 11630 }, { "grad_norm": 0.5512795448303223, "learning_rate": 9.879980249434637e-05, "loss": 0.0833, "step": 11640 }, { "grad_norm": 0.7624636292457581, "learning_rate": 9.879619876352168e-05, "loss": 0.0958, "step": 11650 }, { "grad_norm": 0.7758662104606628, "learning_rate": 9.879258969641809e-05, "loss": 0.0773, "step": 11660 }, { "grad_norm": 0.949623703956604, "learning_rate": 9.878897529343023e-05, "loss": 0.1035, "step": 11670 }, { "grad_norm": 0.6185071468353271, "learning_rate": 9.878535555495338e-05, "loss": 0.0829, "step": 11680 }, { "grad_norm": 0.6162616014480591, "learning_rate": 9.87817304813834e-05, "loss": 0.0811, "step": 11690 }, { "grad_norm": 0.7161450386047363, "learning_rate": 9.877810007311671e-05, "loss": 0.0716, "step": 11700 }, { "grad_norm": 0.752868115901947, "learning_rate": 9.877446433055035e-05, "loss": 0.1045, "step": 11710 }, { "grad_norm": 0.7068431377410889, "learning_rate": 9.877082325408191e-05, "loss": 0.0801, "step": 11720 }, { "grad_norm": 0.7809284925460815, "learning_rate": 9.876717684410954e-05, "loss": 0.0812, "step": 11730 }, { "grad_norm": 0.8699203729629517, "learning_rate": 9.876352510103204e-05, "loss": 0.092, "step": 11740 }, { "grad_norm": 0.8121247291564941, "learning_rate": 9.875986802524875e-05, "loss": 0.1074, "step": 11750 }, { "grad_norm": 0.704228937625885, "learning_rate": 9.87562056171596e-05, "loss": 0.097, "step": 11760 }, { "grad_norm": 0.9095374941825867, "learning_rate": 9.875253787716511e-05, "loss": 0.1084, "step": 11770 }, { "grad_norm": 0.5193262696266174, "learning_rate": 9.874886480566637e-05, "loss": 0.0779, "step": 11780 }, { "grad_norm": 0.5570003986358643, "learning_rate": 9.874518640306507e-05, "loss": 0.0959, "step": 11790 }, { "grad_norm": 0.5541623830795288, "learning_rate": 9.874150266976347e-05, "loss": 0.1326, "step": 11800 }, { "grad_norm": 0.8211928606033325, "learning_rate": 9.873781360616443e-05, "loss": 0.1126, "step": 11810 }, { "grad_norm": 1.0955638885498047, "learning_rate": 9.873411921267137e-05, "loss": 0.1003, "step": 11820 }, { "grad_norm": 0.5859614014625549, "learning_rate": 9.873041948968829e-05, "loss": 0.0839, "step": 11830 }, { "grad_norm": 0.5233978033065796, "learning_rate": 9.872671443761981e-05, "loss": 0.0844, "step": 11840 }, { "grad_norm": 0.6079379320144653, "learning_rate": 9.872300405687109e-05, "loss": 0.0648, "step": 11850 }, { "grad_norm": 0.5414599776268005, "learning_rate": 9.871928834784792e-05, "loss": 0.0723, "step": 11860 }, { "grad_norm": 0.5711613297462463, "learning_rate": 9.871556731095661e-05, "loss": 0.0707, "step": 11870 }, { "grad_norm": 0.6194420456886292, "learning_rate": 9.871184094660411e-05, "loss": 0.0945, "step": 11880 }, { "grad_norm": 0.6812218427658081, "learning_rate": 9.870810925519791e-05, "loss": 0.102, "step": 11890 }, { "grad_norm": 0.4926844537258148, "learning_rate": 9.870437223714612e-05, "loss": 0.0833, "step": 11900 }, { "grad_norm": 0.6003040075302124, "learning_rate": 9.87006298928574e-05, "loss": 0.0712, "step": 11910 }, { "grad_norm": 0.5842589139938354, "learning_rate": 9.869688222274103e-05, "loss": 0.0781, "step": 11920 }, { "grad_norm": 0.48863545060157776, "learning_rate": 9.869312922720681e-05, "loss": 0.0785, "step": 11930 }, { "grad_norm": 0.8738866448402405, "learning_rate": 9.868937090666521e-05, "loss": 0.0979, "step": 11940 }, { "grad_norm": 0.696704089641571, "learning_rate": 9.86856072615272e-05, "loss": 0.0851, "step": 11950 }, { "grad_norm": 0.7684020400047302, "learning_rate": 9.868183829220438e-05, "loss": 0.0964, "step": 11960 }, { "grad_norm": 0.5260816812515259, "learning_rate": 9.867806399910893e-05, "loss": 0.0919, "step": 11970 }, { "grad_norm": 1.2168495655059814, "learning_rate": 9.867428438265356e-05, "loss": 0.1441, "step": 11980 }, { "grad_norm": 1.3358420133590698, "learning_rate": 9.867049944325165e-05, "loss": 0.1004, "step": 11990 }, { "grad_norm": 0.7750048041343689, "learning_rate": 9.86667091813171e-05, "loss": 0.1281, "step": 12000 }, { "grad_norm": 0.6206136345863342, "learning_rate": 9.866291359726438e-05, "loss": 0.0908, "step": 12010 }, { "grad_norm": 0.6608884334564209, "learning_rate": 9.865911269150861e-05, "loss": 0.0858, "step": 12020 }, { "grad_norm": 0.6891574263572693, "learning_rate": 9.865530646446544e-05, "loss": 0.0943, "step": 12030 }, { "grad_norm": 0.7198166847229004, "learning_rate": 9.86514949165511e-05, "loss": 0.0987, "step": 12040 }, { "grad_norm": 0.8544383645057678, "learning_rate": 9.864767804818243e-05, "loss": 0.1032, "step": 12050 }, { "grad_norm": 0.5950340628623962, "learning_rate": 9.86438558597768e-05, "loss": 0.1039, "step": 12060 }, { "grad_norm": 0.65531325340271, "learning_rate": 9.864002835175225e-05, "loss": 0.0892, "step": 12070 }, { "grad_norm": 0.5511330962181091, "learning_rate": 9.863619552452734e-05, "loss": 0.0819, "step": 12080 }, { "grad_norm": 0.568496823310852, "learning_rate": 9.863235737852119e-05, "loss": 0.1035, "step": 12090 }, { "grad_norm": 0.714654803276062, "learning_rate": 9.862851391415356e-05, "loss": 0.0966, "step": 12100 }, { "grad_norm": 0.6548973321914673, "learning_rate": 9.862466513184477e-05, "loss": 0.088, "step": 12110 }, { "grad_norm": 0.8692603707313538, "learning_rate": 9.86208110320157e-05, "loss": 0.0771, "step": 12120 }, { "grad_norm": 1.2454696893692017, "learning_rate": 9.861695161508784e-05, "loss": 0.0917, "step": 12130 }, { "grad_norm": 0.7236482501029968, "learning_rate": 9.861308688148324e-05, "loss": 0.0889, "step": 12140 }, { "grad_norm": 0.8764500021934509, "learning_rate": 9.860921683162455e-05, "loss": 0.0917, "step": 12150 }, { "grad_norm": 0.7515614628791809, "learning_rate": 9.860534146593499e-05, "loss": 0.1036, "step": 12160 }, { "grad_norm": 0.9978235363960266, "learning_rate": 9.860146078483836e-05, "loss": 0.1075, "step": 12170 }, { "grad_norm": 0.72822505235672, "learning_rate": 9.859757478875905e-05, "loss": 0.0848, "step": 12180 }, { "grad_norm": 1.0301406383514404, "learning_rate": 9.859368347812204e-05, "loss": 0.1048, "step": 12190 }, { "grad_norm": 1.1146748065948486, "learning_rate": 9.858978685335285e-05, "loss": 0.0994, "step": 12200 }, { "grad_norm": 0.5994881391525269, "learning_rate": 9.858588491487763e-05, "loss": 0.0701, "step": 12210 }, { "grad_norm": 0.6739115715026855, "learning_rate": 9.858197766312308e-05, "loss": 0.085, "step": 12220 }, { "grad_norm": 0.5924992561340332, "learning_rate": 9.857806509851649e-05, "loss": 0.0669, "step": 12230 }, { "grad_norm": 0.6195202469825745, "learning_rate": 9.857414722148574e-05, "loss": 0.0998, "step": 12240 }, { "grad_norm": 0.5795271396636963, "learning_rate": 9.857022403245928e-05, "loss": 0.0741, "step": 12250 }, { "grad_norm": 0.8409627079963684, "learning_rate": 9.856629553186615e-05, "loss": 0.0769, "step": 12260 }, { "grad_norm": 0.6265002489089966, "learning_rate": 9.856236172013595e-05, "loss": 0.0739, "step": 12270 }, { "grad_norm": 0.7410828471183777, "learning_rate": 9.85584225976989e-05, "loss": 0.0943, "step": 12280 }, { "grad_norm": 0.6414968371391296, "learning_rate": 9.855447816498575e-05, "loss": 0.1167, "step": 12290 }, { "grad_norm": 0.8293763995170593, "learning_rate": 9.855052842242787e-05, "loss": 0.1112, "step": 12300 }, { "grad_norm": 1.034382700920105, "learning_rate": 9.85465733704572e-05, "loss": 0.1343, "step": 12310 }, { "grad_norm": 0.8414400219917297, "learning_rate": 9.854261300950624e-05, "loss": 0.1376, "step": 12320 }, { "grad_norm": 0.7409482598304749, "learning_rate": 9.853864734000813e-05, "loss": 0.126, "step": 12330 }, { "grad_norm": 0.6619705557823181, "learning_rate": 9.85346763623965e-05, "loss": 0.1046, "step": 12340 }, { "grad_norm": 0.6154488325119019, "learning_rate": 9.853070007710564e-05, "loss": 0.0853, "step": 12350 }, { "grad_norm": 0.8292258977890015, "learning_rate": 9.85267184845704e-05, "loss": 0.0969, "step": 12360 }, { "grad_norm": 0.7497810125350952, "learning_rate": 9.852273158522616e-05, "loss": 0.1054, "step": 12370 }, { "grad_norm": 0.5383074283599854, "learning_rate": 9.851873937950896e-05, "loss": 0.0883, "step": 12380 }, { "grad_norm": 0.6687007546424866, "learning_rate": 9.851474186785537e-05, "loss": 0.1058, "step": 12390 }, { "grad_norm": 0.5410100221633911, "learning_rate": 9.851073905070254e-05, "loss": 0.0673, "step": 12400 }, { "grad_norm": 0.60079425573349, "learning_rate": 9.850673092848824e-05, "loss": 0.0899, "step": 12410 }, { "grad_norm": 0.8840530514717102, "learning_rate": 9.850271750165077e-05, "loss": 0.099, "step": 12420 }, { "grad_norm": 0.6735556125640869, "learning_rate": 9.849869877062902e-05, "loss": 0.0886, "step": 12430 }, { "grad_norm": 0.6123524308204651, "learning_rate": 9.849467473586252e-05, "loss": 0.0922, "step": 12440 }, { "grad_norm": 0.6740984320640564, "learning_rate": 9.849064539779127e-05, "loss": 0.1106, "step": 12450 }, { "grad_norm": 0.692608118057251, "learning_rate": 9.848661075685594e-05, "loss": 0.0703, "step": 12460 }, { "grad_norm": 0.6875949501991272, "learning_rate": 9.848257081349778e-05, "loss": 0.1059, "step": 12470 }, { "grad_norm": 0.506828784942627, "learning_rate": 9.847852556815856e-05, "loss": 0.0754, "step": 12480 }, { "grad_norm": 1.0448943376541138, "learning_rate": 9.847447502128067e-05, "loss": 0.0896, "step": 12490 }, { "grad_norm": 0.5220162272453308, "learning_rate": 9.847041917330708e-05, "loss": 0.0975, "step": 12500 }, { "grad_norm": 0.6641718149185181, "learning_rate": 9.846635802468132e-05, "loss": 0.0679, "step": 12510 }, { "grad_norm": 0.8661326169967651, "learning_rate": 9.84622915758475e-05, "loss": 0.0757, "step": 12520 }, { "grad_norm": 0.6032021641731262, "learning_rate": 9.845821982725034e-05, "loss": 0.0665, "step": 12530 }, { "grad_norm": 0.5680547952651978, "learning_rate": 9.845414277933514e-05, "loss": 0.0967, "step": 12540 }, { "grad_norm": 0.7650204300880432, "learning_rate": 9.845006043254771e-05, "loss": 0.0899, "step": 12550 }, { "grad_norm": 0.6628085970878601, "learning_rate": 9.844597278733451e-05, "loss": 0.0943, "step": 12560 }, { "grad_norm": 0.6043098568916321, "learning_rate": 9.844187984414259e-05, "loss": 0.0817, "step": 12570 }, { "grad_norm": 0.811124861240387, "learning_rate": 9.84377816034195e-05, "loss": 0.0928, "step": 12580 }, { "grad_norm": 0.7413172125816345, "learning_rate": 9.843367806561345e-05, "loss": 0.1031, "step": 12590 }, { "grad_norm": 0.7879091501235962, "learning_rate": 9.842956923117317e-05, "loss": 0.0845, "step": 12600 }, { "grad_norm": 0.7739408016204834, "learning_rate": 9.842545510054802e-05, "loss": 0.08, "step": 12610 }, { "grad_norm": 1.075394868850708, "learning_rate": 9.842133567418792e-05, "loss": 0.0926, "step": 12620 }, { "grad_norm": 0.60719233751297, "learning_rate": 9.841721095254333e-05, "loss": 0.0869, "step": 12630 }, { "grad_norm": 0.8767765760421753, "learning_rate": 9.841308093606537e-05, "loss": 0.0888, "step": 12640 }, { "grad_norm": 1.0920801162719727, "learning_rate": 9.840894562520565e-05, "loss": 0.0839, "step": 12650 }, { "grad_norm": 0.7536035180091858, "learning_rate": 9.840480502041642e-05, "loss": 0.0982, "step": 12660 }, { "grad_norm": 0.6912155747413635, "learning_rate": 9.840065912215049e-05, "loss": 0.1032, "step": 12670 }, { "grad_norm": 0.8741875886917114, "learning_rate": 9.839650793086124e-05, "loss": 0.083, "step": 12680 }, { "grad_norm": 0.7823854684829712, "learning_rate": 9.839235144700265e-05, "loss": 0.0877, "step": 12690 }, { "grad_norm": 0.6214651465415955, "learning_rate": 9.838818967102926e-05, "loss": 0.0796, "step": 12700 }, { "grad_norm": 0.8013910055160522, "learning_rate": 9.83840226033962e-05, "loss": 0.1054, "step": 12710 }, { "grad_norm": 0.6989085078239441, "learning_rate": 9.837985024455918e-05, "loss": 0.0788, "step": 12720 }, { "grad_norm": 0.6571138501167297, "learning_rate": 9.837567259497447e-05, "loss": 0.079, "step": 12730 }, { "grad_norm": 0.6946267485618591, "learning_rate": 9.837148965509894e-05, "loss": 0.0897, "step": 12740 }, { "grad_norm": 0.6561670303344727, "learning_rate": 9.836730142539001e-05, "loss": 0.0746, "step": 12750 }, { "grad_norm": 0.7085579633712769, "learning_rate": 9.836310790630574e-05, "loss": 0.0793, "step": 12760 }, { "grad_norm": 0.6762197613716125, "learning_rate": 9.83589090983047e-05, "loss": 0.0908, "step": 12770 }, { "grad_norm": 1.0534002780914307, "learning_rate": 9.835470500184605e-05, "loss": 0.1032, "step": 12780 }, { "grad_norm": 0.8248004913330078, "learning_rate": 9.835049561738957e-05, "loss": 0.0751, "step": 12790 }, { "grad_norm": 0.707284152507782, "learning_rate": 9.834628094539558e-05, "loss": 0.1058, "step": 12800 }, { "grad_norm": 0.8876157402992249, "learning_rate": 9.834206098632499e-05, "loss": 0.0684, "step": 12810 }, { "grad_norm": 0.7433859705924988, "learning_rate": 9.833783574063931e-05, "loss": 0.0799, "step": 12820 }, { "grad_norm": 0.3804192841053009, "learning_rate": 9.833360520880058e-05, "loss": 0.0922, "step": 12830 }, { "grad_norm": 0.7510034441947937, "learning_rate": 9.832936939127144e-05, "loss": 0.0639, "step": 12840 }, { "grad_norm": 0.6886688470840454, "learning_rate": 9.832512828851515e-05, "loss": 0.0713, "step": 12850 }, { "grad_norm": 0.5817186832427979, "learning_rate": 9.832088190099546e-05, "loss": 0.061, "step": 12860 }, { "grad_norm": 0.694739580154419, "learning_rate": 9.831663022917679e-05, "loss": 0.0818, "step": 12870 }, { "grad_norm": 0.7291336059570312, "learning_rate": 9.831237327352407e-05, "loss": 0.0758, "step": 12880 }, { "grad_norm": 0.47045910358428955, "learning_rate": 9.830811103450286e-05, "loss": 0.0785, "step": 12890 }, { "grad_norm": 0.7950178384780884, "learning_rate": 9.830384351257924e-05, "loss": 0.0827, "step": 12900 }, { "grad_norm": 0.6520109176635742, "learning_rate": 9.829957070821993e-05, "loss": 0.0893, "step": 12910 }, { "grad_norm": 0.7435853481292725, "learning_rate": 9.829529262189218e-05, "loss": 0.0911, "step": 12920 }, { "grad_norm": 0.9777659773826599, "learning_rate": 9.829100925406385e-05, "loss": 0.0803, "step": 12930 }, { "grad_norm": 0.5732824802398682, "learning_rate": 9.828672060520333e-05, "loss": 0.0965, "step": 12940 }, { "grad_norm": 0.6263505220413208, "learning_rate": 9.828242667577966e-05, "loss": 0.0725, "step": 12950 }, { "grad_norm": 0.888172447681427, "learning_rate": 9.82781274662624e-05, "loss": 0.0843, "step": 12960 }, { "grad_norm": 0.6704139709472656, "learning_rate": 9.82738229771217e-05, "loss": 0.0732, "step": 12970 }, { "grad_norm": 0.7852035164833069, "learning_rate": 9.826951320882829e-05, "loss": 0.104, "step": 12980 }, { "grad_norm": 0.6321118474006653, "learning_rate": 9.826519816185351e-05, "loss": 0.0765, "step": 12990 }, { "grad_norm": 0.4818579852581024, "learning_rate": 9.826087783666921e-05, "loss": 0.0728, "step": 13000 }, { "grad_norm": 0.6210394501686096, "learning_rate": 9.825655223374787e-05, "loss": 0.0684, "step": 13010 }, { "grad_norm": 0.6256142258644104, "learning_rate": 9.825222135356253e-05, "loss": 0.0653, "step": 13020 }, { "grad_norm": 0.5330297350883484, "learning_rate": 9.82478851965868e-05, "loss": 0.0855, "step": 13030 }, { "grad_norm": 0.6794848442077637, "learning_rate": 9.82435437632949e-05, "loss": 0.0756, "step": 13040 }, { "grad_norm": 0.6833067536354065, "learning_rate": 9.823919705416158e-05, "loss": 0.0712, "step": 13050 }, { "grad_norm": 0.7612708210945129, "learning_rate": 9.82348450696622e-05, "loss": 0.0841, "step": 13060 }, { "grad_norm": 0.5227667093276978, "learning_rate": 9.823048781027268e-05, "loss": 0.0948, "step": 13070 }, { "grad_norm": 0.6041677594184875, "learning_rate": 9.822612527646953e-05, "loss": 0.0828, "step": 13080 }, { "grad_norm": 0.6319622993469238, "learning_rate": 9.822175746872984e-05, "loss": 0.0842, "step": 13090 }, { "grad_norm": 0.7433093786239624, "learning_rate": 9.821738438753123e-05, "loss": 0.0815, "step": 13100 }, { "grad_norm": 0.9737274050712585, "learning_rate": 9.821300603335196e-05, "loss": 0.0823, "step": 13110 }, { "grad_norm": 0.8855244517326355, "learning_rate": 9.820862240667085e-05, "loss": 0.0956, "step": 13120 }, { "grad_norm": 0.7372452020645142, "learning_rate": 9.820423350796726e-05, "loss": 0.0829, "step": 13130 }, { "grad_norm": 0.8542646765708923, "learning_rate": 9.819983933772118e-05, "loss": 0.1106, "step": 13140 }, { "grad_norm": 0.784464418888092, "learning_rate": 9.819543989641314e-05, "loss": 0.0759, "step": 13150 }, { "grad_norm": 0.7396730780601501, "learning_rate": 9.819103518452423e-05, "loss": 0.0774, "step": 13160 }, { "grad_norm": 0.7358533143997192, "learning_rate": 9.818662520253618e-05, "loss": 0.098, "step": 13170 }, { "grad_norm": 0.8343899250030518, "learning_rate": 9.818220995093126e-05, "loss": 0.0897, "step": 13180 }, { "grad_norm": 0.906176745891571, "learning_rate": 9.817778943019228e-05, "loss": 0.101, "step": 13190 }, { "grad_norm": 0.8721826076507568, "learning_rate": 9.81733636408027e-05, "loss": 0.0977, "step": 13200 }, { "grad_norm": 0.5094088912010193, "learning_rate": 9.816893258324649e-05, "loss": 0.0802, "step": 13210 }, { "grad_norm": 0.6993772983551025, "learning_rate": 9.816449625800823e-05, "loss": 0.09, "step": 13220 }, { "grad_norm": 0.507606029510498, "learning_rate": 9.816005466557308e-05, "loss": 0.0864, "step": 13230 }, { "grad_norm": 0.8506569266319275, "learning_rate": 9.815560780642674e-05, "loss": 0.0752, "step": 13240 }, { "grad_norm": 0.7813331484794617, "learning_rate": 9.815115568105555e-05, "loss": 0.0831, "step": 13250 }, { "grad_norm": 0.6797144412994385, "learning_rate": 9.814669828994638e-05, "loss": 0.0802, "step": 13260 }, { "grad_norm": 0.46400606632232666, "learning_rate": 9.814223563358665e-05, "loss": 0.0853, "step": 13270 }, { "grad_norm": 0.5177779793739319, "learning_rate": 9.813776771246443e-05, "loss": 0.0663, "step": 13280 }, { "grad_norm": 0.6448492407798767, "learning_rate": 9.813329452706829e-05, "loss": 0.0681, "step": 13290 }, { "grad_norm": 0.7333129048347473, "learning_rate": 9.812881607788744e-05, "loss": 0.1194, "step": 13300 }, { "grad_norm": 0.6062377691268921, "learning_rate": 9.812433236541163e-05, "loss": 0.0814, "step": 13310 }, { "grad_norm": 0.5237687826156616, "learning_rate": 9.811984339013116e-05, "loss": 0.0662, "step": 13320 }, { "grad_norm": 0.5150750875473022, "learning_rate": 9.811534915253698e-05, "loss": 0.0759, "step": 13330 }, { "grad_norm": 0.6421722769737244, "learning_rate": 9.811084965312056e-05, "loss": 0.0621, "step": 13340 }, { "grad_norm": 0.5892778038978577, "learning_rate": 9.810634489237396e-05, "loss": 0.0966, "step": 13350 }, { "grad_norm": 0.7754455208778381, "learning_rate": 9.81018348707898e-05, "loss": 0.0956, "step": 13360 }, { "grad_norm": 0.5538326501846313, "learning_rate": 9.809731958886131e-05, "loss": 0.0791, "step": 13370 }, { "grad_norm": 0.5911293625831604, "learning_rate": 9.809279904708224e-05, "loss": 0.096, "step": 13380 }, { "grad_norm": 0.4656468331813812, "learning_rate": 9.808827324594699e-05, "loss": 0.0872, "step": 13390 }, { "grad_norm": 0.6218768954277039, "learning_rate": 9.808374218595046e-05, "loss": 0.0822, "step": 13400 }, { "grad_norm": 0.629659116268158, "learning_rate": 9.80792058675882e-05, "loss": 0.0871, "step": 13410 }, { "grad_norm": 0.7592500448226929, "learning_rate": 9.807466429135627e-05, "loss": 0.0898, "step": 13420 }, { "grad_norm": 0.476072758436203, "learning_rate": 9.807011745775132e-05, "loss": 0.1006, "step": 13430 }, { "grad_norm": 0.3960818350315094, "learning_rate": 9.806556536727061e-05, "loss": 0.0939, "step": 13440 }, { "grad_norm": 0.49497923254966736, "learning_rate": 9.806100802041193e-05, "loss": 0.0805, "step": 13450 }, { "grad_norm": 0.6094144582748413, "learning_rate": 9.805644541767368e-05, "loss": 0.0811, "step": 13460 }, { "grad_norm": 1.1577450037002563, "learning_rate": 9.805187755955478e-05, "loss": 0.1199, "step": 13470 }, { "grad_norm": 0.6019378304481506, "learning_rate": 9.804730444655483e-05, "loss": 0.0849, "step": 13480 }, { "grad_norm": 0.502974808216095, "learning_rate": 9.804272607917388e-05, "loss": 0.0781, "step": 13490 }, { "grad_norm": 0.5834264159202576, "learning_rate": 9.803814245791265e-05, "loss": 0.0812, "step": 13500 }, { "grad_norm": 0.6349467635154724, "learning_rate": 9.803355358327239e-05, "loss": 0.0829, "step": 13510 }, { "grad_norm": 0.5462818741798401, "learning_rate": 9.802895945575492e-05, "loss": 0.0801, "step": 13520 }, { "grad_norm": 0.5331334471702576, "learning_rate": 9.802436007586266e-05, "loss": 0.0693, "step": 13530 }, { "grad_norm": 0.552194356918335, "learning_rate": 9.801975544409858e-05, "loss": 0.0817, "step": 13540 }, { "grad_norm": 0.7883672118186951, "learning_rate": 9.801514556096625e-05, "loss": 0.0921, "step": 13550 }, { "grad_norm": 0.7582927346229553, "learning_rate": 9.801053042696977e-05, "loss": 0.1061, "step": 13560 }, { "grad_norm": 0.5426456928253174, "learning_rate": 9.800591004261388e-05, "loss": 0.0863, "step": 13570 }, { "grad_norm": 0.7464055418968201, "learning_rate": 9.800128440840385e-05, "loss": 0.0714, "step": 13580 }, { "grad_norm": 0.6053320169448853, "learning_rate": 9.799665352484552e-05, "loss": 0.0696, "step": 13590 }, { "grad_norm": 0.6217365860939026, "learning_rate": 9.799201739244532e-05, "loss": 0.1221, "step": 13600 }, { "grad_norm": 0.6634798645973206, "learning_rate": 9.798737601171025e-05, "loss": 0.0721, "step": 13610 }, { "grad_norm": 0.7919149398803711, "learning_rate": 9.79827293831479e-05, "loss": 0.0758, "step": 13620 }, { "grad_norm": 0.4742473363876343, "learning_rate": 9.797807750726638e-05, "loss": 0.0791, "step": 13630 }, { "grad_norm": 0.6170284748077393, "learning_rate": 9.797342038457446e-05, "loss": 0.0903, "step": 13640 }, { "grad_norm": 0.4977993369102478, "learning_rate": 9.796875801558141e-05, "loss": 0.0796, "step": 13650 }, { "grad_norm": 1.102381706237793, "learning_rate": 9.79640904007971e-05, "loss": 0.0972, "step": 13660 }, { "grad_norm": 0.840755045413971, "learning_rate": 9.795941754073199e-05, "loss": 0.0966, "step": 13670 }, { "grad_norm": 0.584327757358551, "learning_rate": 9.795473943589705e-05, "loss": 0.0841, "step": 13680 }, { "grad_norm": 0.6246228218078613, "learning_rate": 9.795005608680394e-05, "loss": 0.0754, "step": 13690 }, { "grad_norm": 0.722172200679779, "learning_rate": 9.794536749396477e-05, "loss": 0.1108, "step": 13700 }, { "grad_norm": 0.7421950697898865, "learning_rate": 9.79406736578923e-05, "loss": 0.0907, "step": 13710 }, { "grad_norm": 0.6988069415092468, "learning_rate": 9.793597457909984e-05, "loss": 0.0713, "step": 13720 }, { "grad_norm": 0.7194722890853882, "learning_rate": 9.793127025810127e-05, "loss": 0.0904, "step": 13730 }, { "grad_norm": 0.9214552640914917, "learning_rate": 9.792656069541104e-05, "loss": 0.1082, "step": 13740 }, { "grad_norm": 0.668215274810791, "learning_rate": 9.79218458915442e-05, "loss": 0.0875, "step": 13750 }, { "grad_norm": 0.5377021431922913, "learning_rate": 9.791712584701634e-05, "loss": 0.1219, "step": 13760 }, { "grad_norm": 0.7711840867996216, "learning_rate": 9.791240056234364e-05, "loss": 0.0793, "step": 13770 }, { "grad_norm": 0.7456702589988708, "learning_rate": 9.790767003804283e-05, "loss": 0.0908, "step": 13780 }, { "grad_norm": 0.5550426840782166, "learning_rate": 9.790293427463126e-05, "loss": 0.0743, "step": 13790 }, { "grad_norm": 0.47637009620666504, "learning_rate": 9.789819327262684e-05, "loss": 0.0862, "step": 13800 }, { "grad_norm": 0.6394058465957642, "learning_rate": 9.7893447032548e-05, "loss": 0.0799, "step": 13810 }, { "grad_norm": 0.8183403015136719, "learning_rate": 9.78886955549138e-05, "loss": 0.0879, "step": 13820 }, { "grad_norm": 0.8722814917564392, "learning_rate": 9.788393884024387e-05, "loss": 0.0832, "step": 13830 }, { "grad_norm": 0.8881000876426697, "learning_rate": 9.787917688905836e-05, "loss": 0.1012, "step": 13840 }, { "grad_norm": 0.8393563628196716, "learning_rate": 9.787440970187807e-05, "loss": 0.096, "step": 13850 }, { "grad_norm": 0.7081876993179321, "learning_rate": 9.786963727922429e-05, "loss": 0.1108, "step": 13860 }, { "grad_norm": 0.6804845333099365, "learning_rate": 9.786485962161897e-05, "loss": 0.0793, "step": 13870 }, { "grad_norm": 0.6485002636909485, "learning_rate": 9.786007672958455e-05, "loss": 0.0697, "step": 13880 }, { "grad_norm": 0.5333536267280579, "learning_rate": 9.78552886036441e-05, "loss": 0.0681, "step": 13890 }, { "grad_norm": 0.6943023204803467, "learning_rate": 9.785049524432124e-05, "loss": 0.0731, "step": 13900 }, { "grad_norm": 0.8072029948234558, "learning_rate": 9.784569665214016e-05, "loss": 0.0889, "step": 13910 }, { "grad_norm": 1.0054343938827515, "learning_rate": 9.784089282762563e-05, "loss": 0.0881, "step": 13920 }, { "grad_norm": 0.5961007475852966, "learning_rate": 9.7836083771303e-05, "loss": 0.0878, "step": 13930 }, { "grad_norm": 0.9535620808601379, "learning_rate": 9.783126948369817e-05, "loss": 0.0964, "step": 13940 }, { "grad_norm": 0.6741158962249756, "learning_rate": 9.78264499653376e-05, "loss": 0.082, "step": 13950 }, { "grad_norm": 0.6312775611877441, "learning_rate": 9.782162521674838e-05, "loss": 0.1026, "step": 13960 }, { "grad_norm": 0.5933154821395874, "learning_rate": 9.781679523845812e-05, "loss": 0.0801, "step": 13970 }, { "grad_norm": 0.8021098375320435, "learning_rate": 9.781196003099502e-05, "loss": 0.0842, "step": 13980 }, { "grad_norm": 0.6406062245368958, "learning_rate": 9.780711959488786e-05, "loss": 0.0934, "step": 13990 }, { "grad_norm": 0.9914801120758057, "learning_rate": 9.780227393066599e-05, "loss": 0.0848, "step": 14000 }, { "grad_norm": 0.4760849177837372, "learning_rate": 9.77974230388593e-05, "loss": 0.089, "step": 14010 }, { "grad_norm": 0.6069236993789673, "learning_rate": 9.779256691999829e-05, "loss": 0.0997, "step": 14020 }, { "grad_norm": 1.0527772903442383, "learning_rate": 9.778770557461403e-05, "loss": 0.0854, "step": 14030 }, { "grad_norm": 0.8502780795097351, "learning_rate": 9.778283900323812e-05, "loss": 0.0995, "step": 14040 }, { "grad_norm": 0.5525216460227966, "learning_rate": 9.777796720640277e-05, "loss": 0.1064, "step": 14050 }, { "grad_norm": 0.7617731094360352, "learning_rate": 9.777309018464078e-05, "loss": 0.0896, "step": 14060 }, { "grad_norm": 0.6909792423248291, "learning_rate": 9.776820793848547e-05, "loss": 0.0701, "step": 14070 }, { "grad_norm": 0.6908486485481262, "learning_rate": 9.776332046847075e-05, "loss": 0.0746, "step": 14080 }, { "grad_norm": 0.8444229960441589, "learning_rate": 9.775842777513111e-05, "loss": 0.0762, "step": 14090 }, { "grad_norm": 0.6135054230690002, "learning_rate": 9.775352985900163e-05, "loss": 0.0692, "step": 14100 }, { "grad_norm": 0.7687132954597473, "learning_rate": 9.774862672061791e-05, "loss": 0.0854, "step": 14110 }, { "grad_norm": 0.6026694178581238, "learning_rate": 9.774371836051616e-05, "loss": 0.0751, "step": 14120 }, { "grad_norm": 0.5173074007034302, "learning_rate": 9.773880477923315e-05, "loss": 0.0833, "step": 14130 }, { "grad_norm": 0.6886129975318909, "learning_rate": 9.773388597730623e-05, "loss": 0.0731, "step": 14140 }, { "grad_norm": 0.632087767124176, "learning_rate": 9.77289619552733e-05, "loss": 0.0976, "step": 14150 }, { "grad_norm": 0.7503756880760193, "learning_rate": 9.772403271367285e-05, "loss": 0.1073, "step": 14160 }, { "grad_norm": 1.008804202079773, "learning_rate": 9.771909825304396e-05, "loss": 0.0775, "step": 14170 }, { "grad_norm": 0.6422093510627747, "learning_rate": 9.771415857392619e-05, "loss": 0.1069, "step": 14180 }, { "grad_norm": 0.6144077181816101, "learning_rate": 9.770921367685978e-05, "loss": 0.0661, "step": 14190 }, { "grad_norm": 0.8255826830863953, "learning_rate": 9.770426356238551e-05, "loss": 0.0875, "step": 14200 }, { "grad_norm": 0.5927662253379822, "learning_rate": 9.769930823104469e-05, "loss": 0.104, "step": 14210 }, { "grad_norm": 0.5555858612060547, "learning_rate": 9.769434768337926e-05, "loss": 0.0922, "step": 14220 }, { "grad_norm": 0.726523220539093, "learning_rate": 9.768938191993164e-05, "loss": 0.0797, "step": 14230 }, { "grad_norm": 0.6773883700370789, "learning_rate": 9.768441094124494e-05, "loss": 0.0767, "step": 14240 }, { "grad_norm": 0.5833565592765808, "learning_rate": 9.767943474786275e-05, "loss": 0.0674, "step": 14250 }, { "grad_norm": 0.6239273548126221, "learning_rate": 9.767445334032923e-05, "loss": 0.0823, "step": 14260 }, { "grad_norm": 0.6207653880119324, "learning_rate": 9.766946671918919e-05, "loss": 0.0897, "step": 14270 }, { "grad_norm": 0.5597454309463501, "learning_rate": 9.766447488498796e-05, "loss": 0.0958, "step": 14280 }, { "grad_norm": 0.820809006690979, "learning_rate": 9.765947783827139e-05, "loss": 0.1021, "step": 14290 }, { "grad_norm": 0.915454626083374, "learning_rate": 9.765447557958599e-05, "loss": 0.1101, "step": 14300 }, { "grad_norm": 0.6188080906867981, "learning_rate": 9.764946810947879e-05, "loss": 0.0872, "step": 14310 }, { "grad_norm": 1.069104552268982, "learning_rate": 9.764445542849738e-05, "loss": 0.1011, "step": 14320 }, { "grad_norm": 2.1067991256713867, "learning_rate": 9.763943753718998e-05, "loss": 0.0908, "step": 14330 }, { "grad_norm": 0.5894839763641357, "learning_rate": 9.76344144361053e-05, "loss": 0.0774, "step": 14340 }, { "grad_norm": 0.8407804369926453, "learning_rate": 9.762938612579269e-05, "loss": 0.0731, "step": 14350 }, { "grad_norm": 0.8094589114189148, "learning_rate": 9.762435260680202e-05, "loss": 0.0884, "step": 14360 }, { "grad_norm": 0.6848286390304565, "learning_rate": 9.761931387968373e-05, "loss": 0.0859, "step": 14370 }, { "grad_norm": 0.7664588689804077, "learning_rate": 9.76142699449889e-05, "loss": 0.0808, "step": 14380 }, { "grad_norm": 0.7966979146003723, "learning_rate": 9.760922080326908e-05, "loss": 0.0783, "step": 14390 }, { "grad_norm": 1.2201802730560303, "learning_rate": 9.760416645507644e-05, "loss": 0.0887, "step": 14400 }, { "grad_norm": 0.9594006538391113, "learning_rate": 9.759910690096375e-05, "loss": 0.0755, "step": 14410 }, { "grad_norm": 0.6531534194946289, "learning_rate": 9.759404214148429e-05, "loss": 0.0913, "step": 14420 }, { "grad_norm": 1.0657089948654175, "learning_rate": 9.758897217719191e-05, "loss": 0.1058, "step": 14430 }, { "grad_norm": 0.7536860108375549, "learning_rate": 9.758389700864113e-05, "loss": 0.0931, "step": 14440 }, { "grad_norm": 0.6678152680397034, "learning_rate": 9.757881663638688e-05, "loss": 0.0753, "step": 14450 }, { "grad_norm": 0.6126086711883545, "learning_rate": 9.757373106098478e-05, "loss": 0.0802, "step": 14460 }, { "grad_norm": 0.6436264514923096, "learning_rate": 9.756864028299097e-05, "loss": 0.084, "step": 14470 }, { "grad_norm": 0.6883033514022827, "learning_rate": 9.75635443029622e-05, "loss": 0.0875, "step": 14480 }, { "grad_norm": 0.6577674746513367, "learning_rate": 9.755844312145572e-05, "loss": 0.0832, "step": 14490 }, { "grad_norm": 0.6655340790748596, "learning_rate": 9.755333673902941e-05, "loss": 0.0747, "step": 14500 }, { "grad_norm": 0.8056566715240479, "learning_rate": 9.75482251562417e-05, "loss": 0.0768, "step": 14510 }, { "grad_norm": 0.6558332443237305, "learning_rate": 9.754310837365155e-05, "loss": 0.0779, "step": 14520 }, { "grad_norm": 0.6221230626106262, "learning_rate": 9.753798639181856e-05, "loss": 0.0962, "step": 14530 }, { "grad_norm": 0.6591148972511292, "learning_rate": 9.753285921130286e-05, "loss": 0.077, "step": 14540 }, { "grad_norm": 0.5259786248207092, "learning_rate": 9.752772683266512e-05, "loss": 0.1045, "step": 14550 }, { "grad_norm": 0.8511868119239807, "learning_rate": 9.752258925646665e-05, "loss": 0.093, "step": 14560 }, { "grad_norm": 0.761447548866272, "learning_rate": 9.751744648326926e-05, "loss": 0.0866, "step": 14570 }, { "grad_norm": 0.8746169805526733, "learning_rate": 9.751229851363536e-05, "loss": 0.082, "step": 14580 }, { "grad_norm": 0.7636524438858032, "learning_rate": 9.750714534812793e-05, "loss": 0.0828, "step": 14590 }, { "grad_norm": 0.4698547422885895, "learning_rate": 9.750198698731053e-05, "loss": 0.0917, "step": 14600 }, { "grad_norm": 0.6462781429290771, "learning_rate": 9.749682343174722e-05, "loss": 0.1044, "step": 14610 }, { "grad_norm": 0.6102008819580078, "learning_rate": 9.749165468200272e-05, "loss": 0.0862, "step": 14620 }, { "grad_norm": 0.7391397953033447, "learning_rate": 9.748648073864229e-05, "loss": 0.0779, "step": 14630 }, { "grad_norm": 0.6621421575546265, "learning_rate": 9.748130160223168e-05, "loss": 0.0856, "step": 14640 }, { "grad_norm": 0.7096847295761108, "learning_rate": 9.747611727333734e-05, "loss": 0.0917, "step": 14650 }, { "grad_norm": 0.8713374733924866, "learning_rate": 9.74709277525262e-05, "loss": 0.0805, "step": 14660 }, { "grad_norm": 0.6073492169380188, "learning_rate": 9.746573304036576e-05, "loss": 0.086, "step": 14670 }, { "grad_norm": 0.596779465675354, "learning_rate": 9.746053313742412e-05, "loss": 0.0799, "step": 14680 }, { "grad_norm": 0.49764567613601685, "learning_rate": 9.745532804426994e-05, "loss": 0.0631, "step": 14690 }, { "grad_norm": 0.638648271560669, "learning_rate": 9.745011776147242e-05, "loss": 0.0641, "step": 14700 }, { "grad_norm": 0.6272572875022888, "learning_rate": 9.744490228960138e-05, "loss": 0.0699, "step": 14710 }, { "grad_norm": 0.8457241058349609, "learning_rate": 9.743968162922713e-05, "loss": 0.0752, "step": 14720 }, { "grad_norm": 0.7167947888374329, "learning_rate": 9.743445578092064e-05, "loss": 0.0865, "step": 14730 }, { "grad_norm": 0.6098755598068237, "learning_rate": 9.742922474525338e-05, "loss": 0.0934, "step": 14740 }, { "grad_norm": 0.6866421103477478, "learning_rate": 9.742398852279741e-05, "loss": 0.0578, "step": 14750 }, { "grad_norm": 0.5027018189430237, "learning_rate": 9.741874711412535e-05, "loss": 0.0736, "step": 14760 }, { "grad_norm": 0.5664196014404297, "learning_rate": 9.741350051981042e-05, "loss": 0.0745, "step": 14770 }, { "grad_norm": 0.6953340172767639, "learning_rate": 9.740824874042633e-05, "loss": 0.0743, "step": 14780 }, { "grad_norm": 0.9705739617347717, "learning_rate": 9.740299177654746e-05, "loss": 0.1005, "step": 14790 }, { "grad_norm": 0.5676624178886414, "learning_rate": 9.739772962874867e-05, "loss": 0.0777, "step": 14800 }, { "grad_norm": 0.6932673454284668, "learning_rate": 9.739246229760541e-05, "loss": 0.0925, "step": 14810 }, { "grad_norm": 0.4453347623348236, "learning_rate": 9.738718978369376e-05, "loss": 0.0725, "step": 14820 }, { "grad_norm": 0.6093139052391052, "learning_rate": 9.738191208759025e-05, "loss": 0.0779, "step": 14830 }, { "grad_norm": 0.9325056672096252, "learning_rate": 9.73766292098721e-05, "loss": 0.1085, "step": 14840 }, { "grad_norm": 0.6621383428573608, "learning_rate": 9.737134115111699e-05, "loss": 0.0885, "step": 14850 }, { "grad_norm": 0.8618723154067993, "learning_rate": 9.736604791190323e-05, "loss": 0.0808, "step": 14860 }, { "grad_norm": 0.5649840235710144, "learning_rate": 9.73607494928097e-05, "loss": 0.0823, "step": 14870 }, { "grad_norm": 0.6426898837089539, "learning_rate": 9.735544589441581e-05, "loss": 0.0836, "step": 14880 }, { "grad_norm": 0.505989670753479, "learning_rate": 9.735013711730154e-05, "loss": 0.0921, "step": 14890 }, { "grad_norm": 0.6072019338607788, "learning_rate": 9.734482316204747e-05, "loss": 0.07, "step": 14900 }, { "grad_norm": 0.8920980095863342, "learning_rate": 9.733950402923473e-05, "loss": 0.1042, "step": 14910 }, { "grad_norm": 0.6890347599983215, "learning_rate": 9.7334179719445e-05, "loss": 0.0776, "step": 14920 }, { "grad_norm": 0.8142871856689453, "learning_rate": 9.732885023326053e-05, "loss": 0.0935, "step": 14930 }, { "grad_norm": 0.6911337971687317, "learning_rate": 9.732351557126418e-05, "loss": 0.0894, "step": 14940 }, { "grad_norm": 0.9495934247970581, "learning_rate": 9.731817573403929e-05, "loss": 0.0924, "step": 14950 }, { "grad_norm": 0.776615560054779, "learning_rate": 9.731283072216985e-05, "loss": 0.1053, "step": 14960 }, { "grad_norm": 0.5330576300621033, "learning_rate": 9.730748053624039e-05, "loss": 0.0981, "step": 14970 }, { "grad_norm": 0.7050519585609436, "learning_rate": 9.730212517683598e-05, "loss": 0.0906, "step": 14980 }, { "grad_norm": 0.6032552719116211, "learning_rate": 9.729676464454228e-05, "loss": 0.0786, "step": 14990 }, { "grad_norm": 0.5848975777626038, "learning_rate": 9.72913989399455e-05, "loss": 0.0839, "step": 15000 }, { "grad_norm": 0.3357287645339966, "learning_rate": 9.728602806363242e-05, "loss": 0.0822, "step": 15010 }, { "grad_norm": 0.8491060137748718, "learning_rate": 9.728065201619043e-05, "loss": 0.0771, "step": 15020 }, { "grad_norm": 1.1087535619735718, "learning_rate": 9.727527079820742e-05, "loss": 0.0711, "step": 15030 }, { "grad_norm": 0.703381359577179, "learning_rate": 9.726988441027186e-05, "loss": 0.0696, "step": 15040 }, { "grad_norm": 0.6301773190498352, "learning_rate": 9.726449285297281e-05, "loss": 0.0739, "step": 15050 }, { "grad_norm": 0.5217711329460144, "learning_rate": 9.72590961268999e-05, "loss": 0.0632, "step": 15060 }, { "grad_norm": 0.7165346145629883, "learning_rate": 9.725369423264328e-05, "loss": 0.0624, "step": 15070 }, { "grad_norm": 0.5274738073348999, "learning_rate": 9.72482871707937e-05, "loss": 0.0837, "step": 15080 }, { "grad_norm": 0.6059780716896057, "learning_rate": 9.724287494194247e-05, "loss": 0.0787, "step": 15090 }, { "grad_norm": 0.5821444392204285, "learning_rate": 9.723745754668147e-05, "loss": 0.0742, "step": 15100 }, { "grad_norm": 0.8383523225784302, "learning_rate": 9.723203498560313e-05, "loss": 0.0631, "step": 15110 }, { "grad_norm": 0.9296335577964783, "learning_rate": 9.722660725930046e-05, "loss": 0.0757, "step": 15120 }, { "grad_norm": 0.6347447037696838, "learning_rate": 9.722117436836702e-05, "loss": 0.0905, "step": 15130 }, { "grad_norm": 0.7784625887870789, "learning_rate": 9.721573631339696e-05, "loss": 0.0817, "step": 15140 }, { "grad_norm": 0.5654709339141846, "learning_rate": 9.721029309498494e-05, "loss": 0.074, "step": 15150 }, { "grad_norm": 0.883213460445404, "learning_rate": 9.720484471372627e-05, "loss": 0.1198, "step": 15160 }, { "grad_norm": 0.6585603356361389, "learning_rate": 9.719939117021673e-05, "loss": 0.1135, "step": 15170 }, { "grad_norm": 0.5112322568893433, "learning_rate": 9.719393246505275e-05, "loss": 0.0823, "step": 15180 }, { "grad_norm": 0.8021109104156494, "learning_rate": 9.718846859883128e-05, "loss": 0.1088, "step": 15190 }, { "grad_norm": 0.706506073474884, "learning_rate": 9.718299957214982e-05, "loss": 0.0816, "step": 15200 }, { "grad_norm": 0.741327702999115, "learning_rate": 9.717752538560646e-05, "loss": 0.0874, "step": 15210 }, { "grad_norm": 0.5576861500740051, "learning_rate": 9.717204603979986e-05, "loss": 0.0926, "step": 15220 }, { "grad_norm": 0.5629270076751709, "learning_rate": 9.716656153532922e-05, "loss": 0.0992, "step": 15230 }, { "grad_norm": 0.6289879679679871, "learning_rate": 9.716107187279434e-05, "loss": 0.0673, "step": 15240 }, { "grad_norm": 0.48265406489372253, "learning_rate": 9.715557705279555e-05, "loss": 0.0957, "step": 15250 }, { "grad_norm": 0.7709991335868835, "learning_rate": 9.715007707593372e-05, "loss": 0.0775, "step": 15260 }, { "grad_norm": 0.6922102570533752, "learning_rate": 9.714457194281036e-05, "loss": 0.0711, "step": 15270 }, { "grad_norm": 0.6314000487327576, "learning_rate": 9.713906165402751e-05, "loss": 0.0942, "step": 15280 }, { "grad_norm": 0.46801626682281494, "learning_rate": 9.713354621018774e-05, "loss": 0.069, "step": 15290 }, { "grad_norm": 0.6137934923171997, "learning_rate": 9.712802561189422e-05, "loss": 0.0876, "step": 15300 }, { "grad_norm": 1.3198145627975464, "learning_rate": 9.712249985975069e-05, "loss": 0.0887, "step": 15310 }, { "grad_norm": 0.49987345933914185, "learning_rate": 9.71169689543614e-05, "loss": 0.0741, "step": 15320 }, { "grad_norm": 0.48604142665863037, "learning_rate": 9.711143289633123e-05, "loss": 0.0919, "step": 15330 }, { "grad_norm": 0.6488223075866699, "learning_rate": 9.710589168626561e-05, "loss": 0.0892, "step": 15340 }, { "grad_norm": 0.5845522284507751, "learning_rate": 9.710034532477048e-05, "loss": 0.0881, "step": 15350 }, { "grad_norm": 0.814625084400177, "learning_rate": 9.709479381245239e-05, "loss": 0.1029, "step": 15360 }, { "grad_norm": 0.7161587476730347, "learning_rate": 9.708923714991847e-05, "loss": 0.0908, "step": 15370 }, { "grad_norm": 1.244803786277771, "learning_rate": 9.708367533777638e-05, "loss": 0.0879, "step": 15380 }, { "grad_norm": 0.6170608401298523, "learning_rate": 9.707810837663431e-05, "loss": 0.0954, "step": 15390 }, { "grad_norm": 0.5337232351303101, "learning_rate": 9.707253626710113e-05, "loss": 0.0768, "step": 15400 }, { "grad_norm": 0.7361584901809692, "learning_rate": 9.706695900978613e-05, "loss": 0.0804, "step": 15410 }, { "grad_norm": 0.5584025979042053, "learning_rate": 9.706137660529926e-05, "loss": 0.1102, "step": 15420 }, { "grad_norm": 0.7316085696220398, "learning_rate": 9.705578905425101e-05, "loss": 0.0881, "step": 15430 }, { "grad_norm": 0.6465297341346741, "learning_rate": 9.705019635725241e-05, "loss": 0.092, "step": 15440 }, { "grad_norm": 0.5163934826850891, "learning_rate": 9.704459851491508e-05, "loss": 0.0813, "step": 15450 }, { "grad_norm": 0.7685633897781372, "learning_rate": 9.703899552785118e-05, "loss": 0.0942, "step": 15460 }, { "grad_norm": 0.3983122408390045, "learning_rate": 9.703338739667346e-05, "loss": 0.0872, "step": 15470 }, { "grad_norm": 0.6800843477249146, "learning_rate": 9.70277741219952e-05, "loss": 0.089, "step": 15480 }, { "grad_norm": 0.7147240042686462, "learning_rate": 9.702215570443027e-05, "loss": 0.0738, "step": 15490 }, { "grad_norm": 0.6377452611923218, "learning_rate": 9.701653214459309e-05, "loss": 0.0823, "step": 15500 }, { "grad_norm": 0.7086158394813538, "learning_rate": 9.701090344309865e-05, "loss": 0.0783, "step": 15510 }, { "grad_norm": 0.6517870426177979, "learning_rate": 9.700526960056247e-05, "loss": 0.0829, "step": 15520 }, { "grad_norm": 0.6385043859481812, "learning_rate": 9.699963061760068e-05, "loss": 0.0935, "step": 15530 }, { "grad_norm": 0.5659664869308472, "learning_rate": 9.699398649482997e-05, "loss": 0.086, "step": 15540 }, { "grad_norm": 0.5123661756515503, "learning_rate": 9.698833723286753e-05, "loss": 0.0797, "step": 15550 }, { "grad_norm": 0.890531063079834, "learning_rate": 9.698268283233118e-05, "loss": 0.0914, "step": 15560 }, { "grad_norm": 0.6842980980873108, "learning_rate": 9.697702329383929e-05, "loss": 0.0883, "step": 15570 }, { "grad_norm": 0.6431528925895691, "learning_rate": 9.697135861801074e-05, "loss": 0.0773, "step": 15580 }, { "grad_norm": 0.6063172817230225, "learning_rate": 9.696568880546505e-05, "loss": 0.0812, "step": 15590 }, { "grad_norm": 0.6822283864021301, "learning_rate": 9.696001385682223e-05, "loss": 0.0887, "step": 15600 }, { "grad_norm": 0.7217862010002136, "learning_rate": 9.695433377270291e-05, "loss": 0.0929, "step": 15610 }, { "grad_norm": 0.7173516154289246, "learning_rate": 9.694864855372824e-05, "loss": 0.0727, "step": 15620 }, { "grad_norm": 0.7729542255401611, "learning_rate": 9.694295820051995e-05, "loss": 0.0762, "step": 15630 }, { "grad_norm": 0.7312710285186768, "learning_rate": 9.693726271370032e-05, "loss": 0.0689, "step": 15640 }, { "grad_norm": 0.5927479863166809, "learning_rate": 9.693156209389221e-05, "loss": 0.0735, "step": 15650 }, { "grad_norm": 0.5339921116828918, "learning_rate": 9.692585634171905e-05, "loss": 0.0782, "step": 15660 }, { "grad_norm": 0.6110394597053528, "learning_rate": 9.692014545780476e-05, "loss": 0.0722, "step": 15670 }, { "grad_norm": 0.5507495403289795, "learning_rate": 9.691442944277393e-05, "loss": 0.0834, "step": 15680 }, { "grad_norm": 1.0826598405838013, "learning_rate": 9.690870829725162e-05, "loss": 0.0886, "step": 15690 }, { "grad_norm": 0.7728284597396851, "learning_rate": 9.69029820218635e-05, "loss": 0.0714, "step": 15700 }, { "grad_norm": 0.9538499116897583, "learning_rate": 9.689725061723579e-05, "loss": 0.0904, "step": 15710 }, { "grad_norm": 0.5562549233436584, "learning_rate": 9.689151408399527e-05, "loss": 0.0801, "step": 15720 }, { "grad_norm": 0.6211766004562378, "learning_rate": 9.688577242276924e-05, "loss": 0.0788, "step": 15730 }, { "grad_norm": 0.5908359885215759, "learning_rate": 9.688002563418566e-05, "loss": 0.107, "step": 15740 }, { "grad_norm": 0.5952829718589783, "learning_rate": 9.687427371887293e-05, "loss": 0.0891, "step": 15750 }, { "grad_norm": 0.8385263681411743, "learning_rate": 9.686851667746012e-05, "loss": 0.0711, "step": 15760 }, { "grad_norm": 0.5549042224884033, "learning_rate": 9.686275451057677e-05, "loss": 0.0789, "step": 15770 }, { "grad_norm": 0.4018065929412842, "learning_rate": 9.685698721885308e-05, "loss": 0.0841, "step": 15780 }, { "grad_norm": 0.9975634813308716, "learning_rate": 9.68512148029197e-05, "loss": 0.1036, "step": 15790 }, { "grad_norm": 0.7940650582313538, "learning_rate": 9.684543726340791e-05, "loss": 0.1014, "step": 15800 }, { "grad_norm": 0.7561179399490356, "learning_rate": 9.683965460094952e-05, "loss": 0.1031, "step": 15810 }, { "grad_norm": 0.5333911180496216, "learning_rate": 9.683386681617694e-05, "loss": 0.0669, "step": 15820 }, { "grad_norm": 0.8507370948791504, "learning_rate": 9.68280739097231e-05, "loss": 0.1105, "step": 15830 }, { "grad_norm": 0.8557125926017761, "learning_rate": 9.682227588222148e-05, "loss": 0.1068, "step": 15840 }, { "grad_norm": 0.7904860377311707, "learning_rate": 9.681647273430618e-05, "loss": 0.0889, "step": 15850 }, { "grad_norm": 0.5272907614707947, "learning_rate": 9.681066446661182e-05, "loss": 0.0841, "step": 15860 }, { "grad_norm": 0.5635256171226501, "learning_rate": 9.680485107977357e-05, "loss": 0.0751, "step": 15870 }, { "grad_norm": 0.869258463382721, "learning_rate": 9.679903257442716e-05, "loss": 0.099, "step": 15880 }, { "grad_norm": 0.6761395931243896, "learning_rate": 9.679320895120891e-05, "loss": 0.0722, "step": 15890 }, { "grad_norm": 1.0229835510253906, "learning_rate": 9.67873802107557e-05, "loss": 0.0733, "step": 15900 }, { "grad_norm": 0.5183296203613281, "learning_rate": 9.67815463537049e-05, "loss": 0.0654, "step": 15910 }, { "grad_norm": 0.4419171214103699, "learning_rate": 9.677570738069457e-05, "loss": 0.0684, "step": 15920 }, { "grad_norm": 0.8234242796897888, "learning_rate": 9.676986329236318e-05, "loss": 0.067, "step": 15930 }, { "grad_norm": 0.7143159508705139, "learning_rate": 9.676401408934987e-05, "loss": 0.0723, "step": 15940 }, { "grad_norm": 0.5452805161476135, "learning_rate": 9.675815977229428e-05, "loss": 0.0807, "step": 15950 }, { "grad_norm": 0.7869433760643005, "learning_rate": 9.675230034183664e-05, "loss": 0.0816, "step": 15960 }, { "grad_norm": 0.5913022756576538, "learning_rate": 9.674643579861773e-05, "loss": 0.0614, "step": 15970 }, { "grad_norm": 0.9824488162994385, "learning_rate": 9.674056614327886e-05, "loss": 0.0902, "step": 15980 }, { "grad_norm": 0.7739496827125549, "learning_rate": 9.673469137646198e-05, "loss": 0.0768, "step": 15990 }, { "grad_norm": 0.5659932494163513, "learning_rate": 9.67288114988095e-05, "loss": 0.0715, "step": 16000 }, { "grad_norm": 0.5061504244804382, "learning_rate": 9.672292651096447e-05, "loss": 0.0699, "step": 16010 }, { "grad_norm": 0.6366164684295654, "learning_rate": 9.671703641357042e-05, "loss": 0.0671, "step": 16020 }, { "grad_norm": 0.793037474155426, "learning_rate": 9.67111412072715e-05, "loss": 0.0822, "step": 16030 }, { "grad_norm": 0.5163495540618896, "learning_rate": 9.670524089271242e-05, "loss": 0.0672, "step": 16040 }, { "grad_norm": 0.715600848197937, "learning_rate": 9.669933547053842e-05, "loss": 0.0748, "step": 16050 }, { "grad_norm": 0.3897915184497833, "learning_rate": 9.669342494139531e-05, "loss": 0.0707, "step": 16060 }, { "grad_norm": 0.4597914218902588, "learning_rate": 9.668750930592943e-05, "loss": 0.0668, "step": 16070 }, { "grad_norm": 0.593494176864624, "learning_rate": 9.668158856478775e-05, "loss": 0.0698, "step": 16080 }, { "grad_norm": 0.5753305554389954, "learning_rate": 9.66756627186177e-05, "loss": 0.0729, "step": 16090 }, { "grad_norm": 0.6978012323379517, "learning_rate": 9.666973176806737e-05, "loss": 0.0766, "step": 16100 }, { "grad_norm": 0.7409442067146301, "learning_rate": 9.666379571378534e-05, "loss": 0.0642, "step": 16110 }, { "grad_norm": 0.49745264649391174, "learning_rate": 9.665785455642076e-05, "loss": 0.0652, "step": 16120 }, { "grad_norm": 0.7617893815040588, "learning_rate": 9.665190829662337e-05, "loss": 0.0821, "step": 16130 }, { "grad_norm": 0.7273932099342346, "learning_rate": 9.664595693504342e-05, "loss": 0.0698, "step": 16140 }, { "grad_norm": 0.5531633496284485, "learning_rate": 9.664000047233175e-05, "loss": 0.0875, "step": 16150 }, { "grad_norm": 1.1315433979034424, "learning_rate": 9.663403890913976e-05, "loss": 0.0898, "step": 16160 }, { "grad_norm": 0.9285862445831299, "learning_rate": 9.662807224611938e-05, "loss": 0.0991, "step": 16170 }, { "grad_norm": 0.8018402457237244, "learning_rate": 9.662210048392311e-05, "loss": 0.0848, "step": 16180 }, { "grad_norm": 0.5602613687515259, "learning_rate": 9.661612362320405e-05, "loss": 0.1016, "step": 16190 }, { "grad_norm": 0.6549078822135925, "learning_rate": 9.661014166461579e-05, "loss": 0.0792, "step": 16200 }, { "grad_norm": 0.7366273403167725, "learning_rate": 9.66041546088125e-05, "loss": 0.1129, "step": 16210 }, { "grad_norm": 0.6056944131851196, "learning_rate": 9.659816245644895e-05, "loss": 0.0726, "step": 16220 }, { "grad_norm": 0.5891206860542297, "learning_rate": 9.65921652081804e-05, "loss": 0.085, "step": 16230 }, { "grad_norm": 0.5524492263793945, "learning_rate": 9.658616286466271e-05, "loss": 0.0604, "step": 16240 }, { "grad_norm": 0.663233757019043, "learning_rate": 9.65801554265523e-05, "loss": 0.0708, "step": 16250 }, { "grad_norm": 0.6206968426704407, "learning_rate": 9.657414289450612e-05, "loss": 0.0963, "step": 16260 }, { "grad_norm": 0.6545518636703491, "learning_rate": 9.656812526918171e-05, "loss": 0.0992, "step": 16270 }, { "grad_norm": 0.49591779708862305, "learning_rate": 9.656210255123712e-05, "loss": 0.0614, "step": 16280 }, { "grad_norm": 0.4582091271877289, "learning_rate": 9.6556074741331e-05, "loss": 0.0642, "step": 16290 }, { "grad_norm": 0.5696893334388733, "learning_rate": 9.655004184012256e-05, "loss": 0.0831, "step": 16300 }, { "grad_norm": 0.5788630843162537, "learning_rate": 9.654400384827152e-05, "loss": 0.0994, "step": 16310 }, { "grad_norm": 0.6371676921844482, "learning_rate": 9.653796076643818e-05, "loss": 0.0834, "step": 16320 }, { "grad_norm": 0.6287057995796204, "learning_rate": 9.653191259528344e-05, "loss": 0.0916, "step": 16330 }, { "grad_norm": 0.7303736805915833, "learning_rate": 9.65258593354687e-05, "loss": 0.0795, "step": 16340 }, { "grad_norm": 0.9429373145103455, "learning_rate": 9.651980098765591e-05, "loss": 0.0794, "step": 16350 }, { "grad_norm": 0.5321341753005981, "learning_rate": 9.651373755250765e-05, "loss": 0.0613, "step": 16360 }, { "grad_norm": 0.46619972586631775, "learning_rate": 9.650766903068697e-05, "loss": 0.0688, "step": 16370 }, { "grad_norm": 0.6730569005012512, "learning_rate": 9.650159542285753e-05, "loss": 0.0951, "step": 16380 }, { "grad_norm": 0.5576658844947815, "learning_rate": 9.649551672968353e-05, "loss": 0.0764, "step": 16390 }, { "grad_norm": 0.5414199233055115, "learning_rate": 9.648943295182973e-05, "loss": 0.0816, "step": 16400 }, { "grad_norm": 0.5868863463401794, "learning_rate": 9.648334408996144e-05, "loss": 0.0565, "step": 16410 }, { "grad_norm": 0.6434422731399536, "learning_rate": 9.647725014474452e-05, "loss": 0.067, "step": 16420 }, { "grad_norm": 0.640434205532074, "learning_rate": 9.64711511168454e-05, "loss": 0.0802, "step": 16430 }, { "grad_norm": 0.8084682822227478, "learning_rate": 9.646504700693108e-05, "loss": 0.0891, "step": 16440 }, { "grad_norm": 0.6004903316497803, "learning_rate": 9.645893781566907e-05, "loss": 0.0846, "step": 16450 }, { "grad_norm": 0.6926470994949341, "learning_rate": 9.645282354372744e-05, "loss": 0.0685, "step": 16460 }, { "grad_norm": 0.6717531085014343, "learning_rate": 9.644670419177491e-05, "loss": 0.0915, "step": 16470 }, { "grad_norm": 0.49992579221725464, "learning_rate": 9.644057976048062e-05, "loss": 0.0817, "step": 16480 }, { "grad_norm": 0.6759812235832214, "learning_rate": 9.643445025051435e-05, "loss": 0.0801, "step": 16490 }, { "grad_norm": 0.5600285530090332, "learning_rate": 9.642831566254641e-05, "loss": 0.0718, "step": 16500 }, { "grad_norm": 0.6116035580635071, "learning_rate": 9.642217599724769e-05, "loss": 0.0578, "step": 16510 }, { "grad_norm": 0.7194778323173523, "learning_rate": 9.64160312552896e-05, "loss": 0.0609, "step": 16520 }, { "grad_norm": 0.5255911350250244, "learning_rate": 9.64098814373441e-05, "loss": 0.0725, "step": 16530 }, { "grad_norm": 0.7392768263816833, "learning_rate": 9.640372654408374e-05, "loss": 0.0699, "step": 16540 }, { "grad_norm": 0.6428754329681396, "learning_rate": 9.639756657618162e-05, "loss": 0.0742, "step": 16550 }, { "grad_norm": 0.73419588804245, "learning_rate": 9.639140153431138e-05, "loss": 0.0899, "step": 16560 }, { "grad_norm": 0.5422863960266113, "learning_rate": 9.638523141914721e-05, "loss": 0.0661, "step": 16570 }, { "grad_norm": 0.7828894853591919, "learning_rate": 9.637905623136388e-05, "loss": 0.0686, "step": 16580 }, { "grad_norm": 0.9177446961402893, "learning_rate": 9.637287597163669e-05, "loss": 0.0821, "step": 16590 }, { "grad_norm": 0.6024459004402161, "learning_rate": 9.63666906406415e-05, "loss": 0.0777, "step": 16600 }, { "grad_norm": 0.5725513696670532, "learning_rate": 9.636050023905473e-05, "loss": 0.0621, "step": 16610 }, { "grad_norm": 0.5020573139190674, "learning_rate": 9.635430476755336e-05, "loss": 0.0685, "step": 16620 }, { "grad_norm": 0.5355796813964844, "learning_rate": 9.63481042268149e-05, "loss": 0.0636, "step": 16630 }, { "grad_norm": 0.5088977217674255, "learning_rate": 9.634189861751745e-05, "loss": 0.0522, "step": 16640 }, { "grad_norm": 0.49457719922065735, "learning_rate": 9.633568794033967e-05, "loss": 0.0645, "step": 16650 }, { "grad_norm": 0.6169316172599792, "learning_rate": 9.63294721959607e-05, "loss": 0.0793, "step": 16660 }, { "grad_norm": 0.4463939368724823, "learning_rate": 9.63232513850603e-05, "loss": 0.0783, "step": 16670 }, { "grad_norm": 1.1846108436584473, "learning_rate": 9.631702550831878e-05, "loss": 0.0729, "step": 16680 }, { "grad_norm": 0.6214651465415955, "learning_rate": 9.631079456641698e-05, "loss": 0.0853, "step": 16690 }, { "grad_norm": 0.8636924028396606, "learning_rate": 9.630455856003632e-05, "loss": 0.0695, "step": 16700 }, { "grad_norm": 0.5624796748161316, "learning_rate": 9.629831748985876e-05, "loss": 0.0696, "step": 16710 }, { "grad_norm": 0.672736644744873, "learning_rate": 9.629207135656679e-05, "loss": 0.0716, "step": 16720 }, { "grad_norm": 0.5735648274421692, "learning_rate": 9.628582016084353e-05, "loss": 0.0585, "step": 16730 }, { "grad_norm": 0.6815701127052307, "learning_rate": 9.627956390337254e-05, "loss": 0.0883, "step": 16740 }, { "grad_norm": 0.5247380137443542, "learning_rate": 9.627330258483802e-05, "loss": 0.0801, "step": 16750 }, { "grad_norm": 0.7898552417755127, "learning_rate": 9.62670362059247e-05, "loss": 0.0735, "step": 16760 }, { "grad_norm": 0.6837034821510315, "learning_rate": 9.626076476731786e-05, "loss": 0.0635, "step": 16770 }, { "grad_norm": 0.4752463698387146, "learning_rate": 9.625448826970336e-05, "loss": 0.0685, "step": 16780 }, { "grad_norm": 0.5913777351379395, "learning_rate": 9.624820671376755e-05, "loss": 0.0651, "step": 16790 }, { "grad_norm": 0.5499190092086792, "learning_rate": 9.62419201001974e-05, "loss": 0.0659, "step": 16800 }, { "grad_norm": 0.6721624135971069, "learning_rate": 9.623562842968037e-05, "loss": 0.0828, "step": 16810 }, { "grad_norm": 0.6084586381912231, "learning_rate": 9.622933170290454e-05, "loss": 0.067, "step": 16820 }, { "grad_norm": 0.611526370048523, "learning_rate": 9.622302992055849e-05, "loss": 0.0582, "step": 16830 }, { "grad_norm": 0.5338876843452454, "learning_rate": 9.62167230833314e-05, "loss": 0.0622, "step": 16840 }, { "grad_norm": 0.6360195875167847, "learning_rate": 9.621041119191295e-05, "loss": 0.0605, "step": 16850 }, { "grad_norm": 0.5421384572982788, "learning_rate": 9.620409424699342e-05, "loss": 0.0763, "step": 16860 }, { "grad_norm": 0.5184112191200256, "learning_rate": 9.619777224926359e-05, "loss": 0.0793, "step": 16870 }, { "grad_norm": 0.6283755898475647, "learning_rate": 9.619144519941485e-05, "loss": 0.0602, "step": 16880 }, { "grad_norm": 0.5814446210861206, "learning_rate": 9.618511309813912e-05, "loss": 0.0868, "step": 16890 }, { "grad_norm": 3.1230058670043945, "learning_rate": 9.617877594612886e-05, "loss": 0.0646, "step": 16900 }, { "grad_norm": 0.7935622930526733, "learning_rate": 9.617243374407707e-05, "loss": 0.0729, "step": 16910 }, { "grad_norm": 0.6425301432609558, "learning_rate": 9.616608649267736e-05, "loss": 0.0779, "step": 16920 }, { "grad_norm": 0.5498632192611694, "learning_rate": 9.615973419262385e-05, "loss": 0.0649, "step": 16930 }, { "grad_norm": 0.5360889434814453, "learning_rate": 9.615337684461119e-05, "loss": 0.0747, "step": 16940 }, { "grad_norm": 0.5657561421394348, "learning_rate": 9.614701444933465e-05, "loss": 0.0584, "step": 16950 }, { "grad_norm": 0.710166871547699, "learning_rate": 9.614064700748997e-05, "loss": 0.0807, "step": 16960 }, { "grad_norm": 0.7647745013237, "learning_rate": 9.613427451977352e-05, "loss": 0.0692, "step": 16970 }, { "grad_norm": 0.7070757746696472, "learning_rate": 9.612789698688216e-05, "loss": 0.0845, "step": 16980 }, { "grad_norm": 0.7258146405220032, "learning_rate": 9.612151440951334e-05, "loss": 0.0827, "step": 16990 }, { "grad_norm": 0.6832115054130554, "learning_rate": 9.611512678836506e-05, "loss": 0.0843, "step": 17000 }, { "grad_norm": 0.7902984619140625, "learning_rate": 9.610873412413584e-05, "loss": 0.0818, "step": 17010 }, { "grad_norm": 0.6463161110877991, "learning_rate": 9.610233641752476e-05, "loss": 0.0723, "step": 17020 }, { "grad_norm": 0.4439050257205963, "learning_rate": 9.609593366923151e-05, "loss": 0.0682, "step": 17030 }, { "grad_norm": 0.7155143618583679, "learning_rate": 9.608952587995625e-05, "loss": 0.0677, "step": 17040 }, { "grad_norm": 0.7251890301704407, "learning_rate": 9.608311305039972e-05, "loss": 0.0772, "step": 17050 }, { "grad_norm": 0.5344513654708862, "learning_rate": 9.607669518126326e-05, "loss": 0.0751, "step": 17060 }, { "grad_norm": 0.4397137761116028, "learning_rate": 9.607027227324866e-05, "loss": 0.0678, "step": 17070 }, { "grad_norm": 0.6215252876281738, "learning_rate": 9.606384432705837e-05, "loss": 0.0696, "step": 17080 }, { "grad_norm": 0.5227239727973938, "learning_rate": 9.60574113433953e-05, "loss": 0.0759, "step": 17090 }, { "grad_norm": 0.7509779334068298, "learning_rate": 9.6050973322963e-05, "loss": 0.0689, "step": 17100 }, { "grad_norm": 0.787949800491333, "learning_rate": 9.604453026646547e-05, "loss": 0.0853, "step": 17110 }, { "grad_norm": 0.42382171750068665, "learning_rate": 9.603808217460735e-05, "loss": 0.0715, "step": 17120 }, { "grad_norm": 0.7806867957115173, "learning_rate": 9.603162904809377e-05, "loss": 0.1222, "step": 17130 }, { "grad_norm": 0.4605365991592407, "learning_rate": 9.602517088763045e-05, "loss": 0.0839, "step": 17140 }, { "grad_norm": 0.5237055420875549, "learning_rate": 9.601870769392365e-05, "loss": 0.0726, "step": 17150 }, { "grad_norm": 0.71216881275177, "learning_rate": 9.601223946768017e-05, "loss": 0.0976, "step": 17160 }, { "grad_norm": 0.8589075207710266, "learning_rate": 9.600576620960734e-05, "loss": 0.1009, "step": 17170 }, { "grad_norm": 0.6839165091514587, "learning_rate": 9.599928792041308e-05, "loss": 0.0716, "step": 17180 }, { "grad_norm": 0.7751689553260803, "learning_rate": 9.599280460080587e-05, "loss": 0.0575, "step": 17190 }, { "grad_norm": 0.5959526896476746, "learning_rate": 9.59863162514947e-05, "loss": 0.062, "step": 17200 }, { "grad_norm": 0.5097834467887878, "learning_rate": 9.597982287318911e-05, "loss": 0.0708, "step": 17210 }, { "grad_norm": 0.8060693144798279, "learning_rate": 9.597332446659923e-05, "loss": 0.0706, "step": 17220 }, { "grad_norm": 0.5786627531051636, "learning_rate": 9.59668210324357e-05, "loss": 0.0812, "step": 17230 }, { "grad_norm": 0.8168007135391235, "learning_rate": 9.596031257140974e-05, "loss": 0.0915, "step": 17240 }, { "grad_norm": 0.4856245219707489, "learning_rate": 9.59537990842331e-05, "loss": 0.0627, "step": 17250 }, { "grad_norm": 0.8107638955116272, "learning_rate": 9.594728057161806e-05, "loss": 0.066, "step": 17260 }, { "grad_norm": 1.7075380086898804, "learning_rate": 9.594075703427752e-05, "loss": 0.0864, "step": 17270 }, { "grad_norm": 0.7147572636604309, "learning_rate": 9.593422847292486e-05, "loss": 0.0888, "step": 17280 }, { "grad_norm": 0.5097848176956177, "learning_rate": 9.592769488827402e-05, "loss": 0.0796, "step": 17290 }, { "grad_norm": 0.6661931872367859, "learning_rate": 9.592115628103952e-05, "loss": 0.0896, "step": 17300 }, { "grad_norm": 0.5314740538597107, "learning_rate": 9.591461265193643e-05, "loss": 0.0712, "step": 17310 }, { "grad_norm": 0.5604088306427002, "learning_rate": 9.590806400168032e-05, "loss": 0.0776, "step": 17320 }, { "grad_norm": 0.6044672131538391, "learning_rate": 9.590151033098735e-05, "loss": 0.0724, "step": 17330 }, { "grad_norm": 0.8310844898223877, "learning_rate": 9.589495164057423e-05, "loss": 0.0733, "step": 17340 }, { "grad_norm": 0.6187826991081238, "learning_rate": 9.58883879311582e-05, "loss": 0.0696, "step": 17350 }, { "grad_norm": 0.5211523175239563, "learning_rate": 9.588181920345705e-05, "loss": 0.0612, "step": 17360 }, { "grad_norm": 0.5786013603210449, "learning_rate": 9.587524545818913e-05, "loss": 0.0665, "step": 17370 }, { "grad_norm": 0.49403178691864014, "learning_rate": 9.586866669607335e-05, "loss": 0.074, "step": 17380 }, { "grad_norm": 0.6985087990760803, "learning_rate": 9.586208291782915e-05, "loss": 0.0712, "step": 17390 }, { "grad_norm": 0.6106632947921753, "learning_rate": 9.58554941241765e-05, "loss": 0.0573, "step": 17400 }, { "grad_norm": 0.6605665683746338, "learning_rate": 9.584890031583596e-05, "loss": 0.0784, "step": 17410 }, { "grad_norm": 0.5980848670005798, "learning_rate": 9.584230149352861e-05, "loss": 0.0939, "step": 17420 }, { "grad_norm": 0.8111052513122559, "learning_rate": 9.58356976579761e-05, "loss": 0.0804, "step": 17430 }, { "grad_norm": 0.9167343378067017, "learning_rate": 9.58290888099006e-05, "loss": 0.1148, "step": 17440 }, { "grad_norm": 1.2989205121994019, "learning_rate": 9.582247495002486e-05, "loss": 0.0834, "step": 17450 }, { "grad_norm": 0.5812659859657288, "learning_rate": 9.581585607907214e-05, "loss": 0.0649, "step": 17460 }, { "grad_norm": 0.6264286637306213, "learning_rate": 9.580923219776628e-05, "loss": 0.0761, "step": 17470 }, { "grad_norm": 0.7067257165908813, "learning_rate": 9.580260330683167e-05, "loss": 0.0646, "step": 17480 }, { "grad_norm": 0.5403013229370117, "learning_rate": 9.579596940699322e-05, "loss": 0.0688, "step": 17490 }, { "grad_norm": 0.5214893817901611, "learning_rate": 9.578933049897643e-05, "loss": 0.0988, "step": 17500 }, { "grad_norm": 0.5629190802574158, "learning_rate": 9.578268658350728e-05, "loss": 0.062, "step": 17510 }, { "grad_norm": 0.8172369003295898, "learning_rate": 9.577603766131235e-05, "loss": 0.0715, "step": 17520 }, { "grad_norm": 0.8247351050376892, "learning_rate": 9.576938373311878e-05, "loss": 0.0797, "step": 17530 }, { "grad_norm": 0.6454036235809326, "learning_rate": 9.576272479965421e-05, "loss": 0.0696, "step": 17540 }, { "grad_norm": 0.4982103705406189, "learning_rate": 9.575606086164687e-05, "loss": 0.0693, "step": 17550 }, { "grad_norm": 0.5480557084083557, "learning_rate": 9.57493919198255e-05, "loss": 0.0942, "step": 17560 }, { "grad_norm": 0.674802839756012, "learning_rate": 9.57427179749194e-05, "loss": 0.0769, "step": 17570 }, { "grad_norm": 1.0963047742843628, "learning_rate": 9.573603902765846e-05, "loss": 0.0696, "step": 17580 }, { "grad_norm": 0.5843009948730469, "learning_rate": 9.572935507877304e-05, "loss": 0.0674, "step": 17590 }, { "grad_norm": 0.6518981456756592, "learning_rate": 9.57226661289941e-05, "loss": 0.0631, "step": 17600 }, { "grad_norm": 0.5132079720497131, "learning_rate": 9.571597217905315e-05, "loss": 0.0647, "step": 17610 }, { "grad_norm": 1.1572028398513794, "learning_rate": 9.57092732296822e-05, "loss": 0.0782, "step": 17620 }, { "grad_norm": 0.7094407081604004, "learning_rate": 9.570256928161385e-05, "loss": 0.0763, "step": 17630 }, { "grad_norm": 0.7904301881790161, "learning_rate": 9.569586033558126e-05, "loss": 0.0805, "step": 17640 }, { "grad_norm": 0.5264331698417664, "learning_rate": 9.568914639231807e-05, "loss": 0.0753, "step": 17650 }, { "grad_norm": 0.5862482786178589, "learning_rate": 9.568242745255852e-05, "loss": 0.0754, "step": 17660 }, { "grad_norm": 0.5782593488693237, "learning_rate": 9.567570351703739e-05, "loss": 0.0661, "step": 17670 }, { "grad_norm": 0.710760772228241, "learning_rate": 9.566897458649001e-05, "loss": 0.0637, "step": 17680 }, { "grad_norm": 0.6060661673545837, "learning_rate": 9.566224066165221e-05, "loss": 0.0603, "step": 17690 }, { "grad_norm": 0.4786101281642914, "learning_rate": 9.565550174326043e-05, "loss": 0.0664, "step": 17700 }, { "grad_norm": 0.5959997773170471, "learning_rate": 9.564875783205162e-05, "loss": 0.0722, "step": 17710 }, { "grad_norm": 0.5369310975074768, "learning_rate": 9.564200892876328e-05, "loss": 0.0631, "step": 17720 }, { "grad_norm": 0.5946212410926819, "learning_rate": 9.563525503413348e-05, "loss": 0.0705, "step": 17730 }, { "grad_norm": 0.8047574758529663, "learning_rate": 9.562849614890079e-05, "loss": 0.0943, "step": 17740 }, { "grad_norm": 0.6561145186424255, "learning_rate": 9.562173227380436e-05, "loss": 0.1039, "step": 17750 }, { "grad_norm": 0.8872556090354919, "learning_rate": 9.561496340958389e-05, "loss": 0.0722, "step": 17760 }, { "grad_norm": 0.6944124698638916, "learning_rate": 9.560818955697959e-05, "loss": 0.0924, "step": 17770 }, { "grad_norm": 0.5380029082298279, "learning_rate": 9.560141071673228e-05, "loss": 0.0647, "step": 17780 }, { "grad_norm": 0.6882248520851135, "learning_rate": 9.559462688958323e-05, "loss": 0.0807, "step": 17790 }, { "grad_norm": 0.5105918645858765, "learning_rate": 9.558783807627434e-05, "loss": 0.075, "step": 17800 }, { "grad_norm": 0.7982970476150513, "learning_rate": 9.558104427754801e-05, "loss": 0.077, "step": 17810 }, { "grad_norm": 0.960090160369873, "learning_rate": 9.557424549414722e-05, "loss": 0.0898, "step": 17820 }, { "grad_norm": 1.0854756832122803, "learning_rate": 9.556744172681546e-05, "loss": 0.0875, "step": 17830 }, { "grad_norm": 0.8215665817260742, "learning_rate": 9.556063297629677e-05, "loss": 0.0895, "step": 17840 }, { "grad_norm": 0.5431113839149475, "learning_rate": 9.555381924333578e-05, "loss": 0.0689, "step": 17850 }, { "grad_norm": 0.656074583530426, "learning_rate": 9.554700052867758e-05, "loss": 0.0795, "step": 17860 }, { "grad_norm": 0.7654755115509033, "learning_rate": 9.554017683306789e-05, "loss": 0.0704, "step": 17870 }, { "grad_norm": 0.5764421820640564, "learning_rate": 9.553334815725294e-05, "loss": 0.0707, "step": 17880 }, { "grad_norm": 0.5727317929267883, "learning_rate": 9.552651450197949e-05, "loss": 0.0854, "step": 17890 }, { "grad_norm": 0.6447656154632568, "learning_rate": 9.551967586799486e-05, "loss": 0.0541, "step": 17900 }, { "grad_norm": 1.1069917678833008, "learning_rate": 9.551283225604692e-05, "loss": 0.0557, "step": 17910 }, { "grad_norm": 0.38275811076164246, "learning_rate": 9.550598366688406e-05, "loss": 0.0678, "step": 17920 }, { "grad_norm": 0.44320210814476013, "learning_rate": 9.549913010125526e-05, "loss": 0.0634, "step": 17930 }, { "grad_norm": 0.8160370588302612, "learning_rate": 9.549227155990999e-05, "loss": 0.056, "step": 17940 }, { "grad_norm": 0.5259928107261658, "learning_rate": 9.548540804359828e-05, "loss": 0.0771, "step": 17950 }, { "grad_norm": 0.717805027961731, "learning_rate": 9.547853955307077e-05, "loss": 0.0667, "step": 17960 }, { "grad_norm": 0.5925651788711548, "learning_rate": 9.547166608907853e-05, "loss": 0.0969, "step": 17970 }, { "grad_norm": 0.503223180770874, "learning_rate": 9.546478765237326e-05, "loss": 0.0971, "step": 17980 }, { "grad_norm": 0.5206246972084045, "learning_rate": 9.545790424370715e-05, "loss": 0.0755, "step": 17990 }, { "grad_norm": 0.7120431661605835, "learning_rate": 9.5451015863833e-05, "loss": 0.0593, "step": 18000 }, { "grad_norm": 0.8566437363624573, "learning_rate": 9.544412251350408e-05, "loss": 0.0609, "step": 18010 }, { "grad_norm": 0.5880116820335388, "learning_rate": 9.543722419347422e-05, "loss": 0.0704, "step": 18020 }, { "grad_norm": 0.7444019913673401, "learning_rate": 9.543032090449788e-05, "loss": 0.062, "step": 18030 }, { "grad_norm": 0.5640814304351807, "learning_rate": 9.542341264732992e-05, "loss": 0.0826, "step": 18040 }, { "grad_norm": 0.6664976477622986, "learning_rate": 9.541649942272585e-05, "loss": 0.0917, "step": 18050 }, { "grad_norm": 0.7035025954246521, "learning_rate": 9.54095812314417e-05, "loss": 0.0883, "step": 18060 }, { "grad_norm": 0.5042678117752075, "learning_rate": 9.540265807423401e-05, "loss": 0.0837, "step": 18070 }, { "grad_norm": 0.6279459595680237, "learning_rate": 9.53957299518599e-05, "loss": 0.0854, "step": 18080 }, { "grad_norm": 0.492063045501709, "learning_rate": 9.5388796865077e-05, "loss": 0.0681, "step": 18090 }, { "grad_norm": 0.5254278779029846, "learning_rate": 9.538185881464353e-05, "loss": 0.0707, "step": 18100 }, { "grad_norm": 0.6857428550720215, "learning_rate": 9.537491580131821e-05, "loss": 0.0815, "step": 18110 }, { "grad_norm": 0.4820834696292877, "learning_rate": 9.53679678258603e-05, "loss": 0.0721, "step": 18120 }, { "grad_norm": 0.5638211965560913, "learning_rate": 9.536101488902966e-05, "loss": 0.0639, "step": 18130 }, { "grad_norm": 0.6451060771942139, "learning_rate": 9.535405699158663e-05, "loss": 0.0624, "step": 18140 }, { "grad_norm": 0.8014817833900452, "learning_rate": 9.53470941342921e-05, "loss": 0.0718, "step": 18150 }, { "grad_norm": 0.5270816683769226, "learning_rate": 9.534012631790756e-05, "loss": 0.0744, "step": 18160 }, { "grad_norm": 0.6855354309082031, "learning_rate": 9.533315354319494e-05, "loss": 0.0838, "step": 18170 }, { "grad_norm": 0.9169464707374573, "learning_rate": 9.532617581091682e-05, "loss": 0.0683, "step": 18180 }, { "grad_norm": 0.6333739757537842, "learning_rate": 9.531919312183629e-05, "loss": 0.0908, "step": 18190 }, { "grad_norm": 0.5241528749465942, "learning_rate": 9.531220547671688e-05, "loss": 0.0778, "step": 18200 }, { "grad_norm": 0.5776856541633606, "learning_rate": 9.530521287632285e-05, "loss": 0.0682, "step": 18210 }, { "grad_norm": 0.7883760929107666, "learning_rate": 9.529821532141884e-05, "loss": 0.077, "step": 18220 }, { "grad_norm": 0.6620287299156189, "learning_rate": 9.52912128127701e-05, "loss": 0.0777, "step": 18230 }, { "grad_norm": 1.0540341138839722, "learning_rate": 9.528420535114244e-05, "loss": 0.0765, "step": 18240 }, { "grad_norm": 0.8293753266334534, "learning_rate": 9.527719293730215e-05, "loss": 0.0979, "step": 18250 }, { "grad_norm": 0.8406115770339966, "learning_rate": 9.527017557201611e-05, "loss": 0.0599, "step": 18260 }, { "grad_norm": 0.5504888296127319, "learning_rate": 9.526315325605176e-05, "loss": 0.0636, "step": 18270 }, { "grad_norm": 0.7040435075759888, "learning_rate": 9.525612599017699e-05, "loss": 0.0578, "step": 18280 }, { "grad_norm": 0.5555473566055298, "learning_rate": 9.524909377516033e-05, "loss": 0.0949, "step": 18290 }, { "grad_norm": 0.6648399233818054, "learning_rate": 9.524205661177081e-05, "loss": 0.0806, "step": 18300 }, { "grad_norm": 0.794650673866272, "learning_rate": 9.523501450077801e-05, "loss": 0.0775, "step": 18310 }, { "grad_norm": 0.6736153960227966, "learning_rate": 9.522796744295202e-05, "loss": 0.0621, "step": 18320 }, { "grad_norm": 0.8451866507530212, "learning_rate": 9.522091543906352e-05, "loss": 0.068, "step": 18330 }, { "grad_norm": 0.6818528771400452, "learning_rate": 9.521385848988369e-05, "loss": 0.0569, "step": 18340 }, { "grad_norm": 0.6062413454055786, "learning_rate": 9.520679659618428e-05, "loss": 0.0843, "step": 18350 }, { "grad_norm": 0.6784780621528625, "learning_rate": 9.519972975873754e-05, "loss": 0.0792, "step": 18360 }, { "grad_norm": 0.594805121421814, "learning_rate": 9.519265797831633e-05, "loss": 0.0706, "step": 18370 }, { "grad_norm": 0.34453198313713074, "learning_rate": 9.518558125569399e-05, "loss": 0.0624, "step": 18380 }, { "grad_norm": 0.5851852893829346, "learning_rate": 9.517849959164442e-05, "loss": 0.0751, "step": 18390 }, { "grad_norm": 0.3491629660129547, "learning_rate": 9.517141298694205e-05, "loss": 0.0641, "step": 18400 }, { "grad_norm": 0.7178106307983398, "learning_rate": 9.516432144236188e-05, "loss": 0.0709, "step": 18410 }, { "grad_norm": 0.6112589836120605, "learning_rate": 9.515722495867941e-05, "loss": 0.1059, "step": 18420 }, { "grad_norm": 0.7698385119438171, "learning_rate": 9.515012353667072e-05, "loss": 0.0825, "step": 18430 }, { "grad_norm": 0.8898606300354004, "learning_rate": 9.51430171771124e-05, "loss": 0.0795, "step": 18440 }, { "grad_norm": 0.659127414226532, "learning_rate": 9.513590588078159e-05, "loss": 0.0679, "step": 18450 }, { "grad_norm": 0.5144343972206116, "learning_rate": 9.512878964845597e-05, "loss": 0.0768, "step": 18460 }, { "grad_norm": 0.5907641649246216, "learning_rate": 9.512166848091377e-05, "loss": 0.0593, "step": 18470 }, { "grad_norm": 0.6018690466880798, "learning_rate": 9.511454237893376e-05, "loss": 0.1028, "step": 18480 }, { "grad_norm": 0.4118725657463074, "learning_rate": 9.51074113432952e-05, "loss": 0.0476, "step": 18490 }, { "grad_norm": 0.7113272547721863, "learning_rate": 9.510027537477797e-05, "loss": 0.0814, "step": 18500 }, { "grad_norm": 0.5890058279037476, "learning_rate": 9.509313447416242e-05, "loss": 0.0618, "step": 18510 }, { "grad_norm": 3.050041913986206, "learning_rate": 9.508598864222949e-05, "loss": 0.0658, "step": 18520 }, { "grad_norm": 0.5345273613929749, "learning_rate": 9.507883787976062e-05, "loss": 0.062, "step": 18530 }, { "grad_norm": 1.5350805521011353, "learning_rate": 9.507168218753781e-05, "loss": 0.0936, "step": 18540 }, { "grad_norm": 0.5060842037200928, "learning_rate": 9.506452156634362e-05, "loss": 0.0772, "step": 18550 }, { "grad_norm": 0.6271471381187439, "learning_rate": 9.505735601696109e-05, "loss": 0.1065, "step": 18560 }, { "grad_norm": 0.8765199184417725, "learning_rate": 9.505018554017385e-05, "loss": 0.083, "step": 18570 }, { "grad_norm": 0.9206939339637756, "learning_rate": 9.504301013676604e-05, "loss": 0.0995, "step": 18580 }, { "grad_norm": 0.8988001942634583, "learning_rate": 9.503582980752238e-05, "loss": 0.0776, "step": 18590 }, { "grad_norm": 0.5227871537208557, "learning_rate": 9.502864455322809e-05, "loss": 0.107, "step": 18600 }, { "grad_norm": 0.7100628614425659, "learning_rate": 9.502145437466891e-05, "loss": 0.0786, "step": 18610 }, { "grad_norm": 0.601809024810791, "learning_rate": 9.501425927263116e-05, "loss": 0.0679, "step": 18620 }, { "grad_norm": 0.43565618991851807, "learning_rate": 9.500705924790172e-05, "loss": 0.079, "step": 18630 }, { "grad_norm": 0.6174913644790649, "learning_rate": 9.499985430126794e-05, "loss": 0.1005, "step": 18640 }, { "grad_norm": 0.7806422114372253, "learning_rate": 9.499264443351775e-05, "loss": 0.1049, "step": 18650 }, { "grad_norm": 0.5551429390907288, "learning_rate": 9.498542964543961e-05, "loss": 0.078, "step": 18660 }, { "grad_norm": 0.7905206084251404, "learning_rate": 9.497820993782252e-05, "loss": 0.0909, "step": 18670 }, { "grad_norm": 0.750825047492981, "learning_rate": 9.497098531145601e-05, "loss": 0.0727, "step": 18680 }, { "grad_norm": 0.6804237365722656, "learning_rate": 9.496375576713017e-05, "loss": 0.0715, "step": 18690 }, { "grad_norm": 0.7806549072265625, "learning_rate": 9.49565213056356e-05, "loss": 0.0697, "step": 18700 }, { "grad_norm": 0.5271267294883728, "learning_rate": 9.494928192776342e-05, "loss": 0.0695, "step": 18710 }, { "grad_norm": 0.6635415554046631, "learning_rate": 9.494203763430538e-05, "loss": 0.0731, "step": 18720 }, { "grad_norm": 0.8021860718727112, "learning_rate": 9.493478842605366e-05, "loss": 0.0773, "step": 18730 }, { "grad_norm": 0.6146034598350525, "learning_rate": 9.492753430380105e-05, "loss": 0.0888, "step": 18740 }, { "grad_norm": 0.5044208765029907, "learning_rate": 9.492027526834083e-05, "loss": 0.0777, "step": 18750 }, { "grad_norm": 0.4603402614593506, "learning_rate": 9.491301132046684e-05, "loss": 0.0643, "step": 18760 }, { "grad_norm": 0.6083039045333862, "learning_rate": 9.490574246097345e-05, "loss": 0.0625, "step": 18770 }, { "grad_norm": 0.6631247997283936, "learning_rate": 9.48984686906556e-05, "loss": 0.0688, "step": 18780 }, { "grad_norm": 0.6765589118003845, "learning_rate": 9.489119001030871e-05, "loss": 0.0754, "step": 18790 }, { "grad_norm": 0.7144650220870972, "learning_rate": 9.488390642072878e-05, "loss": 0.0775, "step": 18800 }, { "grad_norm": 0.9713727831840515, "learning_rate": 9.48766179227123e-05, "loss": 0.0929, "step": 18810 }, { "grad_norm": 0.560266375541687, "learning_rate": 9.486932451705636e-05, "loss": 0.0648, "step": 18820 }, { "grad_norm": 0.5893486142158508, "learning_rate": 9.486202620455857e-05, "loss": 0.0607, "step": 18830 }, { "grad_norm": 0.9137300252914429, "learning_rate": 9.485472298601704e-05, "loss": 0.0844, "step": 18840 }, { "grad_norm": 0.5876896977424622, "learning_rate": 9.484741486223043e-05, "loss": 0.0716, "step": 18850 }, { "grad_norm": 0.5518268346786499, "learning_rate": 9.484010183399797e-05, "loss": 0.0661, "step": 18860 }, { "grad_norm": 0.5204907655715942, "learning_rate": 9.483278390211938e-05, "loss": 0.0704, "step": 18870 }, { "grad_norm": 0.4890603721141815, "learning_rate": 9.482546106739496e-05, "loss": 0.0823, "step": 18880 }, { "grad_norm": 0.6570711731910706, "learning_rate": 9.48181333306255e-05, "loss": 0.0681, "step": 18890 }, { "grad_norm": 0.6260334253311157, "learning_rate": 9.481080069261237e-05, "loss": 0.0624, "step": 18900 }, { "grad_norm": 1.1790457963943481, "learning_rate": 9.480346315415745e-05, "loss": 0.0964, "step": 18910 }, { "grad_norm": 0.7482521533966064, "learning_rate": 9.479612071606314e-05, "loss": 0.0655, "step": 18920 }, { "grad_norm": 0.6475679874420166, "learning_rate": 9.478877337913244e-05, "loss": 0.073, "step": 18930 }, { "grad_norm": 0.8785180449485779, "learning_rate": 9.478142114416881e-05, "loss": 0.1045, "step": 18940 }, { "grad_norm": 1.0740880966186523, "learning_rate": 9.47740640119763e-05, "loss": 0.0636, "step": 18950 }, { "grad_norm": 0.9021574258804321, "learning_rate": 9.476670198335947e-05, "loss": 0.0692, "step": 18960 }, { "grad_norm": 0.4946148693561554, "learning_rate": 9.47593350591234e-05, "loss": 0.0616, "step": 18970 }, { "grad_norm": 0.6248638033866882, "learning_rate": 9.475196324007376e-05, "loss": 0.0641, "step": 18980 }, { "grad_norm": 0.8269971013069153, "learning_rate": 9.474458652701669e-05, "loss": 0.0686, "step": 18990 }, { "grad_norm": 0.8658469319343567, "learning_rate": 9.473720492075892e-05, "loss": 0.0702, "step": 19000 }, { "grad_norm": 0.8126246333122253, "learning_rate": 9.472981842210768e-05, "loss": 0.0657, "step": 19010 }, { "grad_norm": 0.47394922375679016, "learning_rate": 9.472242703187074e-05, "loss": 0.0757, "step": 19020 }, { "grad_norm": 0.5412049889564514, "learning_rate": 9.471503075085643e-05, "loss": 0.0743, "step": 19030 }, { "grad_norm": 0.7414551377296448, "learning_rate": 9.470762957987359e-05, "loss": 0.0809, "step": 19040 }, { "grad_norm": 0.7187448740005493, "learning_rate": 9.470022351973158e-05, "loss": 0.0801, "step": 19050 }, { "grad_norm": 0.5149674415588379, "learning_rate": 9.469281257124034e-05, "loss": 0.0726, "step": 19060 }, { "grad_norm": 0.6836419701576233, "learning_rate": 9.46853967352103e-05, "loss": 0.0802, "step": 19070 }, { "grad_norm": 0.8691694736480713, "learning_rate": 9.467797601245246e-05, "loss": 0.0939, "step": 19080 }, { "grad_norm": 0.45936688780784607, "learning_rate": 9.467055040377834e-05, "loss": 0.0894, "step": 19090 }, { "grad_norm": 0.9495720863342285, "learning_rate": 9.466311990999999e-05, "loss": 0.0642, "step": 19100 }, { "grad_norm": 0.5822500586509705, "learning_rate": 9.465568453193e-05, "loss": 0.0706, "step": 19110 }, { "grad_norm": 0.48714128136634827, "learning_rate": 9.464824427038148e-05, "loss": 0.0711, "step": 19120 }, { "grad_norm": 1.0257477760314941, "learning_rate": 9.46407991261681e-05, "loss": 0.0905, "step": 19130 }, { "grad_norm": 0.6654970645904541, "learning_rate": 9.463334910010404e-05, "loss": 0.0901, "step": 19140 }, { "grad_norm": 0.7685636281967163, "learning_rate": 9.462589419300403e-05, "loss": 0.0843, "step": 19150 }, { "grad_norm": 0.7239675521850586, "learning_rate": 9.461843440568333e-05, "loss": 0.0791, "step": 19160 }, { "grad_norm": 0.6524497270584106, "learning_rate": 9.461096973895773e-05, "loss": 0.0721, "step": 19170 }, { "grad_norm": 0.5444430708885193, "learning_rate": 9.460350019364355e-05, "loss": 0.0851, "step": 19180 }, { "grad_norm": 0.7904216647148132, "learning_rate": 9.459602577055764e-05, "loss": 0.0745, "step": 19190 }, { "grad_norm": 0.8549227714538574, "learning_rate": 9.45885464705174e-05, "loss": 0.0769, "step": 19200 }, { "grad_norm": 0.4970335364341736, "learning_rate": 9.458106229434076e-05, "loss": 0.0692, "step": 19210 }, { "grad_norm": 0.7156646251678467, "learning_rate": 9.457357324284617e-05, "loss": 0.0691, "step": 19220 }, { "grad_norm": 0.6011540293693542, "learning_rate": 9.456607931685262e-05, "loss": 0.0639, "step": 19230 }, { "grad_norm": 0.482560396194458, "learning_rate": 9.455858051717965e-05, "loss": 0.0748, "step": 19240 }, { "grad_norm": 0.6174077987670898, "learning_rate": 9.45510768446473e-05, "loss": 0.0662, "step": 19250 }, { "grad_norm": 0.6870624423027039, "learning_rate": 9.454356830007618e-05, "loss": 0.0657, "step": 19260 }, { "grad_norm": 0.7406558394432068, "learning_rate": 9.45360548842874e-05, "loss": 0.0633, "step": 19270 }, { "grad_norm": 0.7809652090072632, "learning_rate": 9.452853659810261e-05, "loss": 0.0704, "step": 19280 }, { "grad_norm": 0.6403831839561462, "learning_rate": 9.452101344234401e-05, "loss": 0.0839, "step": 19290 }, { "grad_norm": 0.7329429984092712, "learning_rate": 9.451348541783431e-05, "loss": 0.0547, "step": 19300 }, { "grad_norm": 0.9271433353424072, "learning_rate": 9.450595252539678e-05, "loss": 0.0721, "step": 19310 }, { "grad_norm": 0.5841982960700989, "learning_rate": 9.449841476585518e-05, "loss": 0.0617, "step": 19320 }, { "grad_norm": 0.6060729622840881, "learning_rate": 9.449087214003384e-05, "loss": 0.0776, "step": 19330 }, { "grad_norm": 0.44056734442710876, "learning_rate": 9.448332464875765e-05, "loss": 0.0795, "step": 19340 }, { "grad_norm": 0.5234692096710205, "learning_rate": 9.447577229285192e-05, "loss": 0.0651, "step": 19350 }, { "grad_norm": 0.5647335648536682, "learning_rate": 9.446821507314261e-05, "loss": 0.0615, "step": 19360 }, { "grad_norm": 0.5665611624717712, "learning_rate": 9.446065299045617e-05, "loss": 0.055, "step": 19370 }, { "grad_norm": 0.7670699954032898, "learning_rate": 9.445308604561955e-05, "loss": 0.0568, "step": 19380 }, { "grad_norm": 0.637904703617096, "learning_rate": 9.444551423946028e-05, "loss": 0.0593, "step": 19390 }, { "grad_norm": 0.577528715133667, "learning_rate": 9.443793757280638e-05, "loss": 0.072, "step": 19400 }, { "grad_norm": 0.6799170970916748, "learning_rate": 9.443035604648646e-05, "loss": 0.058, "step": 19410 }, { "grad_norm": 0.717399537563324, "learning_rate": 9.44227696613296e-05, "loss": 0.0558, "step": 19420 }, { "grad_norm": 0.5136563181877136, "learning_rate": 9.441517841816542e-05, "loss": 0.0702, "step": 19430 }, { "grad_norm": 0.5926981568336487, "learning_rate": 9.440758231782413e-05, "loss": 0.0742, "step": 19440 }, { "grad_norm": 0.7299193143844604, "learning_rate": 9.439998136113639e-05, "loss": 0.1076, "step": 19450 }, { "grad_norm": 0.6911899447441101, "learning_rate": 9.439237554893344e-05, "loss": 0.0742, "step": 19460 }, { "grad_norm": 0.5311341881752014, "learning_rate": 9.438476488204705e-05, "loss": 0.0592, "step": 19470 }, { "grad_norm": 0.6009157299995422, "learning_rate": 9.43771493613095e-05, "loss": 0.0641, "step": 19480 }, { "grad_norm": 0.6631041765213013, "learning_rate": 9.436952898755362e-05, "loss": 0.0583, "step": 19490 }, { "grad_norm": 0.7152264714241028, "learning_rate": 9.436190376161276e-05, "loss": 0.0591, "step": 19500 }, { "grad_norm": 0.5988537669181824, "learning_rate": 9.43542736843208e-05, "loss": 0.0709, "step": 19510 }, { "grad_norm": 0.37228164076805115, "learning_rate": 9.434663875651216e-05, "loss": 0.0665, "step": 19520 }, { "grad_norm": 0.5641106963157654, "learning_rate": 9.433899897902177e-05, "loss": 0.0672, "step": 19530 }, { "grad_norm": 0.8936834335327148, "learning_rate": 9.433135435268511e-05, "loss": 0.0653, "step": 19540 }, { "grad_norm": 0.6004319787025452, "learning_rate": 9.432370487833819e-05, "loss": 0.0628, "step": 19550 }, { "grad_norm": 0.5263289213180542, "learning_rate": 9.431605055681756e-05, "loss": 0.0614, "step": 19560 }, { "grad_norm": 0.6684754490852356, "learning_rate": 9.430839138896026e-05, "loss": 0.076, "step": 19570 }, { "grad_norm": 0.7611179947853088, "learning_rate": 9.43007273756039e-05, "loss": 0.074, "step": 19580 }, { "grad_norm": 0.5625564455986023, "learning_rate": 9.429305851758658e-05, "loss": 0.0651, "step": 19590 }, { "grad_norm": 0.732709527015686, "learning_rate": 9.428538481574699e-05, "loss": 0.0854, "step": 19600 }, { "grad_norm": 0.721360981464386, "learning_rate": 9.42777062709243e-05, "loss": 0.0696, "step": 19610 }, { "grad_norm": 0.5432191491127014, "learning_rate": 9.427002288395821e-05, "loss": 0.0737, "step": 19620 }, { "grad_norm": 0.6333573460578918, "learning_rate": 9.426233465568898e-05, "loss": 0.0867, "step": 19630 }, { "grad_norm": 0.5273157358169556, "learning_rate": 9.42546415869574e-05, "loss": 0.0714, "step": 19640 }, { "grad_norm": 0.6105028390884399, "learning_rate": 9.424694367860473e-05, "loss": 0.0769, "step": 19650 }, { "grad_norm": 0.7265188694000244, "learning_rate": 9.423924093147284e-05, "loss": 0.0657, "step": 19660 }, { "grad_norm": 0.8470976948738098, "learning_rate": 9.423153334640407e-05, "loss": 0.0806, "step": 19670 }, { "grad_norm": 0.5588187575340271, "learning_rate": 9.42238209242413e-05, "loss": 0.082, "step": 19680 }, { "grad_norm": 0.6383838057518005, "learning_rate": 9.421610366582798e-05, "loss": 0.0801, "step": 19690 }, { "grad_norm": 0.7271637320518494, "learning_rate": 9.420838157200803e-05, "loss": 0.0781, "step": 19700 }, { "grad_norm": 0.5671402215957642, "learning_rate": 9.420065464362594e-05, "loss": 0.0615, "step": 19710 }, { "grad_norm": 0.3716026544570923, "learning_rate": 9.419292288152673e-05, "loss": 0.0639, "step": 19720 }, { "grad_norm": 0.5240839719772339, "learning_rate": 9.418518628655588e-05, "loss": 0.0663, "step": 19730 }, { "grad_norm": 0.5747740864753723, "learning_rate": 9.417744485955951e-05, "loss": 0.0598, "step": 19740 }, { "grad_norm": 0.787227213382721, "learning_rate": 9.41696986013842e-05, "loss": 0.0656, "step": 19750 }, { "grad_norm": 0.7691651582717896, "learning_rate": 9.416194751287705e-05, "loss": 0.0907, "step": 19760 }, { "grad_norm": 0.6036479473114014, "learning_rate": 9.415419159488572e-05, "loss": 0.0632, "step": 19770 }, { "grad_norm": 0.5055239200592041, "learning_rate": 9.414643084825837e-05, "loss": 0.0728, "step": 19780 }, { "grad_norm": 0.517854630947113, "learning_rate": 9.413866527384372e-05, "loss": 0.0697, "step": 19790 }, { "grad_norm": 0.9855899810791016, "learning_rate": 9.4130894872491e-05, "loss": 0.0745, "step": 19800 }, { "grad_norm": 0.644138514995575, "learning_rate": 9.412311964504998e-05, "loss": 0.0794, "step": 19810 }, { "grad_norm": 0.5729483366012573, "learning_rate": 9.411533959237091e-05, "loss": 0.0648, "step": 19820 }, { "grad_norm": 0.6830821633338928, "learning_rate": 9.410755471530464e-05, "loss": 0.0756, "step": 19830 }, { "grad_norm": 0.9470508694648743, "learning_rate": 9.40997650147025e-05, "loss": 0.0801, "step": 19840 }, { "grad_norm": 0.4756442606449127, "learning_rate": 9.409197049141637e-05, "loss": 0.0592, "step": 19850 }, { "grad_norm": 0.7060738205909729, "learning_rate": 9.408417114629863e-05, "loss": 0.0791, "step": 19860 }, { "grad_norm": 0.663348913192749, "learning_rate": 9.40763669802022e-05, "loss": 0.0616, "step": 19870 }, { "grad_norm": 0.6289646625518799, "learning_rate": 9.406855799398056e-05, "loss": 0.0665, "step": 19880 }, { "grad_norm": 0.4408925771713257, "learning_rate": 9.406074418848767e-05, "loss": 0.0546, "step": 19890 }, { "grad_norm": 0.6283482909202576, "learning_rate": 9.405292556457805e-05, "loss": 0.0655, "step": 19900 }, { "grad_norm": 0.4329817593097687, "learning_rate": 9.404510212310671e-05, "loss": 0.0583, "step": 19910 }, { "grad_norm": 0.49501192569732666, "learning_rate": 9.403727386492924e-05, "loss": 0.0683, "step": 19920 }, { "grad_norm": 0.6330239176750183, "learning_rate": 9.40294407909017e-05, "loss": 0.0612, "step": 19930 }, { "grad_norm": 0.6503991484642029, "learning_rate": 9.40216029018807e-05, "loss": 0.0888, "step": 19940 }, { "grad_norm": 0.8182486295700073, "learning_rate": 9.401376019872338e-05, "loss": 0.0768, "step": 19950 }, { "grad_norm": 0.6408423185348511, "learning_rate": 9.400591268228746e-05, "loss": 0.049, "step": 19960 }, { "grad_norm": 0.5057446360588074, "learning_rate": 9.399806035343106e-05, "loss": 0.0795, "step": 19970 }, { "grad_norm": 0.9246703386306763, "learning_rate": 9.399020321301294e-05, "loss": 0.0742, "step": 19980 }, { "grad_norm": 0.6623418927192688, "learning_rate": 9.398234126189234e-05, "loss": 0.0657, "step": 19990 }, { "grad_norm": 0.9346329569816589, "learning_rate": 9.397447450092902e-05, "loss": 0.0828, "step": 20000 }, { "grad_norm": 1.0361441373825073, "learning_rate": 9.39666029309833e-05, "loss": 0.0976, "step": 20010 }, { "grad_norm": 0.6668805480003357, "learning_rate": 9.395872655291596e-05, "loss": 0.0642, "step": 20020 }, { "grad_norm": 0.5014497637748718, "learning_rate": 9.395084536758838e-05, "loss": 0.0796, "step": 20030 }, { "grad_norm": 0.7664920091629028, "learning_rate": 9.394295937586243e-05, "loss": 0.0835, "step": 20040 }, { "grad_norm": 0.6493426561355591, "learning_rate": 9.393506857860052e-05, "loss": 0.0661, "step": 20050 }, { "grad_norm": 0.771082878112793, "learning_rate": 9.392717297666555e-05, "loss": 0.0631, "step": 20060 }, { "grad_norm": 0.6315302848815918, "learning_rate": 9.391927257092101e-05, "loss": 0.0818, "step": 20070 }, { "grad_norm": 0.7954522371292114, "learning_rate": 9.391136736223085e-05, "loss": 0.0921, "step": 20080 }, { "grad_norm": 0.7146672606468201, "learning_rate": 9.390345735145956e-05, "loss": 0.0837, "step": 20090 }, { "grad_norm": 0.9130012392997742, "learning_rate": 9.389554253947219e-05, "loss": 0.0813, "step": 20100 }, { "grad_norm": 0.5137489438056946, "learning_rate": 9.388762292713428e-05, "loss": 0.0875, "step": 20110 }, { "grad_norm": 0.5745056867599487, "learning_rate": 9.38796985153119e-05, "loss": 0.068, "step": 20120 }, { "grad_norm": 0.6060032248497009, "learning_rate": 9.387176930487169e-05, "loss": 0.055, "step": 20130 }, { "grad_norm": 0.5147561430931091, "learning_rate": 9.386383529668072e-05, "loss": 0.0648, "step": 20140 }, { "grad_norm": 3.523080587387085, "learning_rate": 9.385589649160669e-05, "loss": 0.0804, "step": 20150 }, { "grad_norm": 0.6152851581573486, "learning_rate": 9.384795289051775e-05, "loss": 0.0689, "step": 20160 }, { "grad_norm": 0.584775447845459, "learning_rate": 9.384000449428261e-05, "loss": 0.0774, "step": 20170 }, { "grad_norm": 0.37020090222358704, "learning_rate": 9.383205130377048e-05, "loss": 0.0833, "step": 20180 }, { "grad_norm": 0.6445564031600952, "learning_rate": 9.382409331985114e-05, "loss": 0.0697, "step": 20190 }, { "grad_norm": 0.6974520683288574, "learning_rate": 9.381613054339482e-05, "loss": 0.0741, "step": 20200 }, { "grad_norm": 0.8170385360717773, "learning_rate": 9.380816297527235e-05, "loss": 0.0662, "step": 20210 }, { "grad_norm": 0.5899627804756165, "learning_rate": 9.380019061635506e-05, "loss": 0.0743, "step": 20220 }, { "grad_norm": 0.6431815028190613, "learning_rate": 9.379221346751474e-05, "loss": 0.057, "step": 20230 }, { "grad_norm": 0.5273571610450745, "learning_rate": 9.378423152962382e-05, "loss": 0.0807, "step": 20240 }, { "grad_norm": 0.4919365644454956, "learning_rate": 9.377624480355517e-05, "loss": 0.0558, "step": 20250 }, { "grad_norm": 0.723105788230896, "learning_rate": 9.376825329018219e-05, "loss": 0.0904, "step": 20260 }, { "grad_norm": 0.4397668242454529, "learning_rate": 9.376025699037884e-05, "loss": 0.0746, "step": 20270 }, { "grad_norm": 0.6903219819068909, "learning_rate": 9.37522559050196e-05, "loss": 0.0698, "step": 20280 }, { "grad_norm": 0.6357010006904602, "learning_rate": 9.37442500349794e-05, "loss": 0.0802, "step": 20290 }, { "grad_norm": 0.5688610672950745, "learning_rate": 9.373623938113381e-05, "loss": 0.0745, "step": 20300 }, { "grad_norm": 0.8605440855026245, "learning_rate": 9.372822394435883e-05, "loss": 0.0652, "step": 20310 }, { "grad_norm": 0.9222672581672668, "learning_rate": 9.372020372553102e-05, "loss": 0.0751, "step": 20320 }, { "grad_norm": 0.7834315896034241, "learning_rate": 9.371217872552746e-05, "loss": 0.0667, "step": 20330 }, { "grad_norm": 0.7067381143569946, "learning_rate": 9.370414894522576e-05, "loss": 0.0702, "step": 20340 }, { "grad_norm": 0.5789878964424133, "learning_rate": 9.369611438550406e-05, "loss": 0.0793, "step": 20350 }, { "grad_norm": 0.3641308844089508, "learning_rate": 9.368807504724095e-05, "loss": 0.0858, "step": 20360 }, { "grad_norm": 0.8003368973731995, "learning_rate": 9.368003093131565e-05, "loss": 0.0678, "step": 20370 }, { "grad_norm": 0.5890169739723206, "learning_rate": 9.367198203860785e-05, "loss": 0.1001, "step": 20380 }, { "grad_norm": 0.6020205616950989, "learning_rate": 9.366392836999774e-05, "loss": 0.0806, "step": 20390 }, { "grad_norm": 0.7753300070762634, "learning_rate": 9.365586992636607e-05, "loss": 0.0793, "step": 20400 }, { "grad_norm": 0.729276716709137, "learning_rate": 9.364780670859412e-05, "loss": 0.0725, "step": 20410 }, { "grad_norm": 0.46620652079582214, "learning_rate": 9.363973871756364e-05, "loss": 0.0673, "step": 20420 }, { "grad_norm": 0.45251932740211487, "learning_rate": 9.363166595415696e-05, "loss": 0.0709, "step": 20430 }, { "grad_norm": 0.3554857671260834, "learning_rate": 9.362358841925686e-05, "loss": 0.0627, "step": 20440 }, { "grad_norm": 0.5459142923355103, "learning_rate": 9.361550611374674e-05, "loss": 0.0632, "step": 20450 }, { "grad_norm": 0.592802882194519, "learning_rate": 9.360741903851043e-05, "loss": 0.0765, "step": 20460 }, { "grad_norm": 0.5593705177307129, "learning_rate": 9.359932719443236e-05, "loss": 0.0796, "step": 20470 }, { "grad_norm": 0.6313785314559937, "learning_rate": 9.35912305823974e-05, "loss": 0.0503, "step": 20480 }, { "grad_norm": 0.47728437185287476, "learning_rate": 9.358312920329101e-05, "loss": 0.0589, "step": 20490 }, { "grad_norm": 0.43476343154907227, "learning_rate": 9.357502305799914e-05, "loss": 0.0655, "step": 20500 }, { "grad_norm": 0.5366076231002808, "learning_rate": 9.356691214740824e-05, "loss": 0.0508, "step": 20510 }, { "grad_norm": 0.7885556817054749, "learning_rate": 9.355879647240535e-05, "loss": 0.0622, "step": 20520 }, { "grad_norm": 0.945249617099762, "learning_rate": 9.355067603387798e-05, "loss": 0.0667, "step": 20530 }, { "grad_norm": 0.5886598825454712, "learning_rate": 9.354255083271412e-05, "loss": 0.0698, "step": 20540 }, { "grad_norm": 0.7147386074066162, "learning_rate": 9.353442086980239e-05, "loss": 0.0734, "step": 20550 }, { "grad_norm": 0.41213884949684143, "learning_rate": 9.352628614603185e-05, "loss": 0.0866, "step": 20560 }, { "grad_norm": 0.7829113006591797, "learning_rate": 9.351814666229209e-05, "loss": 0.0955, "step": 20570 }, { "grad_norm": 0.8448135256767273, "learning_rate": 9.351000241947324e-05, "loss": 0.0619, "step": 20580 }, { "grad_norm": 0.7528838515281677, "learning_rate": 9.350185341846594e-05, "loss": 0.0616, "step": 20590 }, { "grad_norm": 1.0267635583877563, "learning_rate": 9.349369966016134e-05, "loss": 0.082, "step": 20600 }, { "grad_norm": 0.7192566990852356, "learning_rate": 9.348554114545117e-05, "loss": 0.0612, "step": 20610 }, { "grad_norm": 0.6007952690124512, "learning_rate": 9.347737787522758e-05, "loss": 0.0614, "step": 20620 }, { "grad_norm": 0.8295806646347046, "learning_rate": 9.346920985038332e-05, "loss": 0.0763, "step": 20630 }, { "grad_norm": 0.7471820116043091, "learning_rate": 9.346103707181162e-05, "loss": 0.0765, "step": 20640 }, { "grad_norm": 0.5574135184288025, "learning_rate": 9.345285954040626e-05, "loss": 0.0709, "step": 20650 }, { "grad_norm": 0.8248409628868103, "learning_rate": 9.34446772570615e-05, "loss": 0.0724, "step": 20660 }, { "grad_norm": 0.42098188400268555, "learning_rate": 9.343649022267214e-05, "loss": 0.0579, "step": 20670 }, { "grad_norm": 0.7575330138206482, "learning_rate": 9.342829843813353e-05, "loss": 0.0882, "step": 20680 }, { "grad_norm": 0.486228883266449, "learning_rate": 9.342010190434149e-05, "loss": 0.0789, "step": 20690 }, { "grad_norm": 0.6580768823623657, "learning_rate": 9.34119006221924e-05, "loss": 0.0829, "step": 20700 }, { "grad_norm": 0.7637659311294556, "learning_rate": 9.340369459258313e-05, "loss": 0.0678, "step": 20710 }, { "grad_norm": 0.5977748036384583, "learning_rate": 9.339548381641106e-05, "loss": 0.0561, "step": 20720 }, { "grad_norm": 0.5350543260574341, "learning_rate": 9.338726829457413e-05, "loss": 0.0792, "step": 20730 }, { "grad_norm": 0.5389119386672974, "learning_rate": 9.337904802797078e-05, "loss": 0.0504, "step": 20740 }, { "grad_norm": 0.5892829895019531, "learning_rate": 9.337082301749993e-05, "loss": 0.0781, "step": 20750 }, { "grad_norm": 0.6770751476287842, "learning_rate": 9.336259326406109e-05, "loss": 0.0749, "step": 20760 }, { "grad_norm": 1.0130490064620972, "learning_rate": 9.335435876855427e-05, "loss": 0.0785, "step": 20770 }, { "grad_norm": 0.5954046845436096, "learning_rate": 9.334611953187994e-05, "loss": 0.0627, "step": 20780 }, { "grad_norm": 0.808966338634491, "learning_rate": 9.333787555493914e-05, "loss": 0.0828, "step": 20790 }, { "grad_norm": 0.4969041347503662, "learning_rate": 9.332962683863345e-05, "loss": 0.0695, "step": 20800 }, { "grad_norm": 0.7238913774490356, "learning_rate": 9.332137338386489e-05, "loss": 0.0851, "step": 20810 }, { "grad_norm": 0.669813871383667, "learning_rate": 9.33131151915361e-05, "loss": 0.0663, "step": 20820 }, { "grad_norm": 0.679715633392334, "learning_rate": 9.330485226255012e-05, "loss": 0.0733, "step": 20830 }, { "grad_norm": 0.5247777104377747, "learning_rate": 9.329658459781061e-05, "loss": 0.0828, "step": 20840 }, { "grad_norm": 0.6522963047027588, "learning_rate": 9.328831219822172e-05, "loss": 0.0625, "step": 20850 }, { "grad_norm": 0.6723488569259644, "learning_rate": 9.328003506468808e-05, "loss": 0.0713, "step": 20860 }, { "grad_norm": 0.7281212210655212, "learning_rate": 9.327175319811488e-05, "loss": 0.0577, "step": 20870 }, { "grad_norm": 0.6374409794807434, "learning_rate": 9.326346659940781e-05, "loss": 0.0656, "step": 20880 }, { "grad_norm": 0.8395130634307861, "learning_rate": 9.325517526947308e-05, "loss": 0.0877, "step": 20890 }, { "grad_norm": 0.6862437725067139, "learning_rate": 9.32468792092174e-05, "loss": 0.0779, "step": 20900 }, { "grad_norm": 0.8853892683982849, "learning_rate": 9.323857841954803e-05, "loss": 0.0914, "step": 20910 }, { "grad_norm": 0.8475831747055054, "learning_rate": 9.323027290137276e-05, "loss": 0.0768, "step": 20920 }, { "grad_norm": 0.5987755060195923, "learning_rate": 9.322196265559981e-05, "loss": 0.0726, "step": 20930 }, { "grad_norm": 0.5797744393348694, "learning_rate": 9.321364768313803e-05, "loss": 0.0678, "step": 20940 }, { "grad_norm": 0.6625754833221436, "learning_rate": 9.32053279848967e-05, "loss": 0.0707, "step": 20950 }, { "grad_norm": 0.7380238175392151, "learning_rate": 9.319700356178567e-05, "loss": 0.0674, "step": 20960 }, { "grad_norm": 0.7256823182106018, "learning_rate": 9.318867441471527e-05, "loss": 0.0701, "step": 20970 }, { "grad_norm": 0.5193080902099609, "learning_rate": 9.318034054459637e-05, "loss": 0.0607, "step": 20980 }, { "grad_norm": 0.6023196578025818, "learning_rate": 9.317200195234034e-05, "loss": 0.0606, "step": 20990 }, { "grad_norm": 0.47614362835884094, "learning_rate": 9.316365863885909e-05, "loss": 0.0755, "step": 21000 }, { "grad_norm": 0.4583386778831482, "learning_rate": 9.315531060506502e-05, "loss": 0.0599, "step": 21010 }, { "grad_norm": 0.6586698293685913, "learning_rate": 9.314695785187108e-05, "loss": 0.1006, "step": 21020 }, { "grad_norm": 0.4686852991580963, "learning_rate": 9.313860038019069e-05, "loss": 0.0645, "step": 21030 }, { "grad_norm": 0.5779497027397156, "learning_rate": 9.313023819093782e-05, "loss": 0.0567, "step": 21040 }, { "grad_norm": 0.7140873670578003, "learning_rate": 9.312187128502695e-05, "loss": 0.0699, "step": 21050 }, { "grad_norm": 0.47294124960899353, "learning_rate": 9.311349966337307e-05, "loss": 0.0555, "step": 21060 }, { "grad_norm": 0.6648185849189758, "learning_rate": 9.310512332689169e-05, "loss": 0.0591, "step": 21070 }, { "grad_norm": 0.5618964433670044, "learning_rate": 9.309674227649883e-05, "loss": 0.055, "step": 21080 }, { "grad_norm": 0.4981265962123871, "learning_rate": 9.308835651311103e-05, "loss": 0.0604, "step": 21090 }, { "grad_norm": 0.5366096496582031, "learning_rate": 9.307996603764533e-05, "loss": 0.051, "step": 21100 }, { "grad_norm": 0.6153198480606079, "learning_rate": 9.307157085101932e-05, "loss": 0.0614, "step": 21110 }, { "grad_norm": 0.5585233569145203, "learning_rate": 9.306317095415109e-05, "loss": 0.0529, "step": 21120 }, { "grad_norm": 0.6962798237800598, "learning_rate": 9.305476634795922e-05, "loss": 0.0571, "step": 21130 }, { "grad_norm": 0.49859142303466797, "learning_rate": 9.304635703336284e-05, "loss": 0.0445, "step": 21140 }, { "grad_norm": 0.6494801640510559, "learning_rate": 9.303794301128157e-05, "loss": 0.0728, "step": 21150 }, { "grad_norm": 0.6046818494796753, "learning_rate": 9.302952428263555e-05, "loss": 0.0724, "step": 21160 }, { "grad_norm": 0.5926865339279175, "learning_rate": 9.302110084834545e-05, "loss": 0.0758, "step": 21170 }, { "grad_norm": 0.5409799814224243, "learning_rate": 9.301267270933245e-05, "loss": 0.0671, "step": 21180 }, { "grad_norm": 0.8086435794830322, "learning_rate": 9.300423986651823e-05, "loss": 0.0829, "step": 21190 }, { "grad_norm": 0.6733336448669434, "learning_rate": 9.299580232082501e-05, "loss": 0.0873, "step": 21200 }, { "grad_norm": 1.1347447633743286, "learning_rate": 9.298736007317547e-05, "loss": 0.124, "step": 21210 }, { "grad_norm": 0.6141234636306763, "learning_rate": 9.297891312449288e-05, "loss": 0.0662, "step": 21220 }, { "grad_norm": 0.5777072906494141, "learning_rate": 9.297046147570094e-05, "loss": 0.0841, "step": 21230 }, { "grad_norm": 0.925952672958374, "learning_rate": 9.296200512772396e-05, "loss": 0.0874, "step": 21240 }, { "grad_norm": 0.6312722563743591, "learning_rate": 9.295354408148668e-05, "loss": 0.0641, "step": 21250 }, { "grad_norm": 0.5799328088760376, "learning_rate": 9.294507833791441e-05, "loss": 0.0577, "step": 21260 }, { "grad_norm": 0.8232540488243103, "learning_rate": 9.293660789793295e-05, "loss": 0.0774, "step": 21270 }, { "grad_norm": 0.40277934074401855, "learning_rate": 9.292813276246858e-05, "loss": 0.0741, "step": 21280 }, { "grad_norm": 0.7022687196731567, "learning_rate": 9.291965293244816e-05, "loss": 0.0782, "step": 21290 }, { "grad_norm": 0.6312326192855835, "learning_rate": 9.291116840879904e-05, "loss": 0.0594, "step": 21300 }, { "grad_norm": 0.797996997833252, "learning_rate": 9.290267919244904e-05, "loss": 0.0786, "step": 21310 }, { "grad_norm": 0.5394517183303833, "learning_rate": 9.289418528432655e-05, "loss": 0.0516, "step": 21320 }, { "grad_norm": 0.849769651889801, "learning_rate": 9.288568668536045e-05, "loss": 0.0699, "step": 21330 }, { "grad_norm": 0.7952023148536682, "learning_rate": 9.287718339648013e-05, "loss": 0.0607, "step": 21340 }, { "grad_norm": 0.3806271255016327, "learning_rate": 9.28686754186155e-05, "loss": 0.0562, "step": 21350 }, { "grad_norm": 0.4125669300556183, "learning_rate": 9.286016275269698e-05, "loss": 0.0643, "step": 21360 }, { "grad_norm": 0.5770158171653748, "learning_rate": 9.285164539965551e-05, "loss": 0.0661, "step": 21370 }, { "grad_norm": 0.6141345500946045, "learning_rate": 9.284312336042251e-05, "loss": 0.0527, "step": 21380 }, { "grad_norm": 0.47807925939559937, "learning_rate": 9.283459663592996e-05, "loss": 0.0572, "step": 21390 }, { "grad_norm": 0.830294132232666, "learning_rate": 9.282606522711033e-05, "loss": 0.1088, "step": 21400 }, { "grad_norm": 0.6186036467552185, "learning_rate": 9.281752913489657e-05, "loss": 0.0804, "step": 21410 }, { "grad_norm": 0.5436728000640869, "learning_rate": 9.280898836022222e-05, "loss": 0.0593, "step": 21420 }, { "grad_norm": 0.4453252851963043, "learning_rate": 9.280044290402126e-05, "loss": 0.0644, "step": 21430 }, { "grad_norm": 0.6773169636726379, "learning_rate": 9.279189276722821e-05, "loss": 0.0722, "step": 21440 }, { "grad_norm": 0.7817847728729248, "learning_rate": 9.278333795077812e-05, "loss": 0.0724, "step": 21450 }, { "grad_norm": 0.5802537798881531, "learning_rate": 9.27747784556065e-05, "loss": 0.0577, "step": 21460 }, { "grad_norm": 0.6153376698493958, "learning_rate": 9.276621428264942e-05, "loss": 0.0703, "step": 21470 }, { "grad_norm": 0.6807664036750793, "learning_rate": 9.275764543284345e-05, "loss": 0.0685, "step": 21480 }, { "grad_norm": 0.7024836540222168, "learning_rate": 9.274907190712566e-05, "loss": 0.0821, "step": 21490 }, { "grad_norm": 0.37128862738609314, "learning_rate": 9.274049370643363e-05, "loss": 0.0645, "step": 21500 }, { "grad_norm": 0.9258514046669006, "learning_rate": 9.273191083170547e-05, "loss": 0.0876, "step": 21510 }, { "grad_norm": 0.8183388113975525, "learning_rate": 9.27233232838798e-05, "loss": 0.0605, "step": 21520 }, { "grad_norm": 1.3546875715255737, "learning_rate": 9.27147310638957e-05, "loss": 0.0794, "step": 21530 }, { "grad_norm": 0.8429300785064697, "learning_rate": 9.270613417269286e-05, "loss": 0.0917, "step": 21540 }, { "grad_norm": 0.6437114477157593, "learning_rate": 9.269753261121138e-05, "loss": 0.0635, "step": 21550 }, { "grad_norm": 0.5753298997879028, "learning_rate": 9.268892638039194e-05, "loss": 0.0648, "step": 21560 }, { "grad_norm": 0.6502668261528015, "learning_rate": 9.268031548117569e-05, "loss": 0.0572, "step": 21570 }, { "grad_norm": 0.5480411052703857, "learning_rate": 9.26716999145043e-05, "loss": 0.0654, "step": 21580 }, { "grad_norm": 0.6284515857696533, "learning_rate": 9.266307968131998e-05, "loss": 0.0793, "step": 21590 }, { "grad_norm": 1.8373998403549194, "learning_rate": 9.26544547825654e-05, "loss": 0.0725, "step": 21600 }, { "grad_norm": 0.38064339756965637, "learning_rate": 9.264582521918376e-05, "loss": 0.0517, "step": 21610 }, { "grad_norm": 0.767288863658905, "learning_rate": 9.263719099211881e-05, "loss": 0.0711, "step": 21620 }, { "grad_norm": 0.7587454915046692, "learning_rate": 9.262855210231476e-05, "loss": 0.0727, "step": 21630 }, { "grad_norm": 0.8926063776016235, "learning_rate": 9.261990855071633e-05, "loss": 0.0711, "step": 21640 }, { "grad_norm": 0.6111559271812439, "learning_rate": 9.261126033826878e-05, "loss": 0.0576, "step": 21650 }, { "grad_norm": 0.533305823802948, "learning_rate": 9.260260746591786e-05, "loss": 0.0585, "step": 21660 }, { "grad_norm": 0.8500757217407227, "learning_rate": 9.259394993460985e-05, "loss": 0.1189, "step": 21670 }, { "grad_norm": 0.6525083780288696, "learning_rate": 9.258528774529151e-05, "loss": 0.0603, "step": 21680 }, { "grad_norm": 0.49223920702934265, "learning_rate": 9.257662089891013e-05, "loss": 0.0657, "step": 21690 }, { "grad_norm": 0.643341064453125, "learning_rate": 9.25679493964135e-05, "loss": 0.0615, "step": 21700 }, { "grad_norm": 0.7487446069717407, "learning_rate": 9.255927323874994e-05, "loss": 0.0557, "step": 21710 }, { "grad_norm": 0.6302087903022766, "learning_rate": 9.255059242686822e-05, "loss": 0.0622, "step": 21720 }, { "grad_norm": 0.8295841217041016, "learning_rate": 9.254190696171769e-05, "loss": 0.0534, "step": 21730 }, { "grad_norm": 0.3462629020214081, "learning_rate": 9.25332168442482e-05, "loss": 0.0598, "step": 21740 }, { "grad_norm": 0.8467715978622437, "learning_rate": 9.252452207541004e-05, "loss": 0.0867, "step": 21750 }, { "grad_norm": 0.6065700054168701, "learning_rate": 9.251582265615409e-05, "loss": 0.0642, "step": 21760 }, { "grad_norm": 0.7356733083724976, "learning_rate": 9.250711858743169e-05, "loss": 0.0682, "step": 21770 }, { "grad_norm": 0.549900233745575, "learning_rate": 9.24984098701947e-05, "loss": 0.0654, "step": 21780 }, { "grad_norm": 0.5070098638534546, "learning_rate": 9.248969650539552e-05, "loss": 0.0747, "step": 21790 }, { "grad_norm": 0.6859097480773926, "learning_rate": 9.2480978493987e-05, "loss": 0.06, "step": 21800 }, { "grad_norm": 0.6198469996452332, "learning_rate": 9.247225583692256e-05, "loss": 0.0606, "step": 21810 }, { "grad_norm": 0.4260503649711609, "learning_rate": 9.246352853515607e-05, "loss": 0.0486, "step": 21820 }, { "grad_norm": 0.48189008235931396, "learning_rate": 9.245479658964194e-05, "loss": 0.0476, "step": 21830 }, { "grad_norm": 0.6153523325920105, "learning_rate": 9.244606000133507e-05, "loss": 0.0819, "step": 21840 }, { "grad_norm": 1.3588508367538452, "learning_rate": 9.24373187711909e-05, "loss": 0.081, "step": 21850 }, { "grad_norm": 0.6653304100036621, "learning_rate": 9.242857290016537e-05, "loss": 0.0736, "step": 21860 }, { "grad_norm": 0.4157271385192871, "learning_rate": 9.241982238921488e-05, "loss": 0.0595, "step": 21870 }, { "grad_norm": 0.5851463079452515, "learning_rate": 9.24110672392964e-05, "loss": 0.0838, "step": 21880 }, { "grad_norm": 0.5434345006942749, "learning_rate": 9.240230745136737e-05, "loss": 0.07, "step": 21890 }, { "grad_norm": 0.7383832335472107, "learning_rate": 9.239354302638575e-05, "loss": 0.0666, "step": 21900 }, { "grad_norm": 0.6097307205200195, "learning_rate": 9.238477396531e-05, "loss": 0.0581, "step": 21910 }, { "grad_norm": 0.7797865867614746, "learning_rate": 9.23760002690991e-05, "loss": 0.0659, "step": 21920 }, { "grad_norm": 0.518217921257019, "learning_rate": 9.236722193871252e-05, "loss": 0.0642, "step": 21930 }, { "grad_norm": 0.47409671545028687, "learning_rate": 9.235843897511023e-05, "loss": 0.0574, "step": 21940 }, { "grad_norm": 0.44417300820350647, "learning_rate": 9.234965137925276e-05, "loss": 0.0592, "step": 21950 }, { "grad_norm": 0.7110503911972046, "learning_rate": 9.234085915210108e-05, "loss": 0.0635, "step": 21960 }, { "grad_norm": 0.37293529510498047, "learning_rate": 9.23320622946167e-05, "loss": 0.0455, "step": 21970 }, { "grad_norm": 0.8082855939865112, "learning_rate": 9.232326080776163e-05, "loss": 0.0755, "step": 21980 }, { "grad_norm": 0.7169431447982788, "learning_rate": 9.23144546924984e-05, "loss": 0.0682, "step": 21990 }, { "grad_norm": 0.5213342905044556, "learning_rate": 9.230564394979e-05, "loss": 0.0596, "step": 22000 }, { "grad_norm": 0.8078168034553528, "learning_rate": 9.22968285806e-05, "loss": 0.0686, "step": 22010 }, { "grad_norm": 0.3153551518917084, "learning_rate": 9.228800858589242e-05, "loss": 0.0525, "step": 22020 }, { "grad_norm": 0.5482958555221558, "learning_rate": 9.227918396663179e-05, "loss": 0.077, "step": 22030 }, { "grad_norm": 0.6067166328430176, "learning_rate": 9.227035472378319e-05, "loss": 0.0709, "step": 22040 }, { "grad_norm": 0.8769850134849548, "learning_rate": 9.226152085831213e-05, "loss": 0.0666, "step": 22050 }, { "grad_norm": 0.8901210427284241, "learning_rate": 9.22526823711847e-05, "loss": 0.0817, "step": 22060 }, { "grad_norm": 0.7248391509056091, "learning_rate": 9.224383926336745e-05, "loss": 0.0667, "step": 22070 }, { "grad_norm": 0.5270617008209229, "learning_rate": 9.223499153582744e-05, "loss": 0.0641, "step": 22080 }, { "grad_norm": 0.6083806753158569, "learning_rate": 9.222613918953226e-05, "loss": 0.0773, "step": 22090 }, { "grad_norm": 0.749331533908844, "learning_rate": 9.221728222544999e-05, "loss": 0.0581, "step": 22100 }, { "grad_norm": 0.806287944316864, "learning_rate": 9.22084206445492e-05, "loss": 0.0696, "step": 22110 }, { "grad_norm": 0.6502721905708313, "learning_rate": 9.2199554447799e-05, "loss": 0.0725, "step": 22120 }, { "grad_norm": 0.6289002299308777, "learning_rate": 9.219068363616897e-05, "loss": 0.0618, "step": 22130 }, { "grad_norm": 0.521557092666626, "learning_rate": 9.218180821062919e-05, "loss": 0.062, "step": 22140 }, { "grad_norm": 0.47051191329956055, "learning_rate": 9.21729281721503e-05, "loss": 0.0582, "step": 22150 }, { "grad_norm": 0.5287129282951355, "learning_rate": 9.216404352170339e-05, "loss": 0.0737, "step": 22160 }, { "grad_norm": 0.48780834674835205, "learning_rate": 9.215515426026007e-05, "loss": 0.0576, "step": 22170 }, { "grad_norm": 0.6389613747596741, "learning_rate": 9.214626038879246e-05, "loss": 0.075, "step": 22180 }, { "grad_norm": 0.6521708965301514, "learning_rate": 9.21373619082732e-05, "loss": 0.0503, "step": 22190 }, { "grad_norm": 0.6683319211006165, "learning_rate": 9.212845881967535e-05, "loss": 0.0954, "step": 22200 }, { "grad_norm": 0.5853782892227173, "learning_rate": 9.211955112397262e-05, "loss": 0.0527, "step": 22210 }, { "grad_norm": 0.6208635568618774, "learning_rate": 9.211063882213909e-05, "loss": 0.0839, "step": 22220 }, { "grad_norm": 0.5581889152526855, "learning_rate": 9.210172191514942e-05, "loss": 0.0624, "step": 22230 }, { "grad_norm": 0.4542239308357239, "learning_rate": 9.209280040397874e-05, "loss": 0.0682, "step": 22240 }, { "grad_norm": 0.5391443371772766, "learning_rate": 9.208387428960268e-05, "loss": 0.087, "step": 22250 }, { "grad_norm": 0.5228466391563416, "learning_rate": 9.20749435729974e-05, "loss": 0.0534, "step": 22260 }, { "grad_norm": 0.6987148523330688, "learning_rate": 9.206600825513957e-05, "loss": 0.0621, "step": 22270 }, { "grad_norm": 0.5180585384368896, "learning_rate": 9.20570683370063e-05, "loss": 0.0588, "step": 22280 }, { "grad_norm": 0.44511714577674866, "learning_rate": 9.204812381957528e-05, "loss": 0.0486, "step": 22290 }, { "grad_norm": 0.5286006927490234, "learning_rate": 9.203917470382465e-05, "loss": 0.0619, "step": 22300 }, { "grad_norm": 0.48091983795166016, "learning_rate": 9.203022099073309e-05, "loss": 0.0626, "step": 22310 }, { "grad_norm": 0.6217516660690308, "learning_rate": 9.202126268127976e-05, "loss": 0.0667, "step": 22320 }, { "grad_norm": 0.5424116849899292, "learning_rate": 9.20122997764443e-05, "loss": 0.0502, "step": 22330 }, { "grad_norm": 0.7511638402938843, "learning_rate": 9.200333227720692e-05, "loss": 0.0596, "step": 22340 }, { "grad_norm": 0.5139583945274353, "learning_rate": 9.199436018454826e-05, "loss": 0.0671, "step": 22350 }, { "grad_norm": 0.703448474407196, "learning_rate": 9.198538349944952e-05, "loss": 0.0536, "step": 22360 }, { "grad_norm": 0.6666685938835144, "learning_rate": 9.197640222289234e-05, "loss": 0.0941, "step": 22370 }, { "grad_norm": 0.48722583055496216, "learning_rate": 9.196741635585895e-05, "loss": 0.0514, "step": 22380 }, { "grad_norm": 0.7075916528701782, "learning_rate": 9.195842589933199e-05, "loss": 0.0723, "step": 22390 }, { "grad_norm": 0.5533217191696167, "learning_rate": 9.194943085429466e-05, "loss": 0.0635, "step": 22400 }, { "grad_norm": 0.5362560749053955, "learning_rate": 9.194043122173065e-05, "loss": 0.0675, "step": 22410 }, { "grad_norm": 0.843325674533844, "learning_rate": 9.193142700262413e-05, "loss": 0.073, "step": 22420 }, { "grad_norm": 0.6978311538696289, "learning_rate": 9.192241819795979e-05, "loss": 0.0611, "step": 22430 }, { "grad_norm": 0.639589250087738, "learning_rate": 9.191340480872284e-05, "loss": 0.0695, "step": 22440 }, { "grad_norm": 0.5531908869743347, "learning_rate": 9.190438683589895e-05, "loss": 0.0905, "step": 22450 }, { "grad_norm": 0.8311835527420044, "learning_rate": 9.189536428047432e-05, "loss": 0.0591, "step": 22460 }, { "grad_norm": 0.9996111989021301, "learning_rate": 9.188633714343564e-05, "loss": 0.0731, "step": 22470 }, { "grad_norm": 0.44762122631073, "learning_rate": 9.18773054257701e-05, "loss": 0.0665, "step": 22480 }, { "grad_norm": 0.7848110198974609, "learning_rate": 9.18682691284654e-05, "loss": 0.0571, "step": 22490 }, { "grad_norm": 0.5668478608131409, "learning_rate": 9.185922825250974e-05, "loss": 0.0555, "step": 22500 }, { "grad_norm": 0.8808603286743164, "learning_rate": 9.185018279889181e-05, "loss": 0.0474, "step": 22510 }, { "grad_norm": 0.5304011702537537, "learning_rate": 9.184113276860082e-05, "loss": 0.0632, "step": 22520 }, { "grad_norm": 0.465288907289505, "learning_rate": 9.183207816262645e-05, "loss": 0.071, "step": 22530 }, { "grad_norm": 0.6966820955276489, "learning_rate": 9.182301898195891e-05, "loss": 0.0578, "step": 22540 }, { "grad_norm": 0.5839657783508301, "learning_rate": 9.181395522758889e-05, "loss": 0.0621, "step": 22550 }, { "grad_norm": 1.1160576343536377, "learning_rate": 9.180488690050759e-05, "loss": 0.0995, "step": 22560 }, { "grad_norm": 0.48687490820884705, "learning_rate": 9.179581400170671e-05, "loss": 0.0597, "step": 22570 }, { "grad_norm": 0.5702658891677856, "learning_rate": 9.178673653217845e-05, "loss": 0.0614, "step": 22580 }, { "grad_norm": 0.5181484818458557, "learning_rate": 9.177765449291551e-05, "loss": 0.0645, "step": 22590 }, { "grad_norm": 0.8389971256256104, "learning_rate": 9.176856788491109e-05, "loss": 0.0697, "step": 22600 }, { "grad_norm": 0.8034440279006958, "learning_rate": 9.175947670915887e-05, "loss": 0.0637, "step": 22610 }, { "grad_norm": 0.59869384765625, "learning_rate": 9.175038096665309e-05, "loss": 0.0694, "step": 22620 }, { "grad_norm": 0.5762763023376465, "learning_rate": 9.17412806583884e-05, "loss": 0.0599, "step": 22630 }, { "grad_norm": 0.3425946533679962, "learning_rate": 9.173217578536002e-05, "loss": 0.057, "step": 22640 }, { "grad_norm": 0.6401988863945007, "learning_rate": 9.172306634856362e-05, "loss": 0.0607, "step": 22650 }, { "grad_norm": 0.675510823726654, "learning_rate": 9.171395234899545e-05, "loss": 0.0633, "step": 22660 }, { "grad_norm": 0.6376674175262451, "learning_rate": 9.170483378765214e-05, "loss": 0.0524, "step": 22670 }, { "grad_norm": 0.7800777554512024, "learning_rate": 9.169571066553091e-05, "loss": 0.0745, "step": 22680 }, { "grad_norm": 0.5921436548233032, "learning_rate": 9.168658298362946e-05, "loss": 0.0968, "step": 22690 }, { "grad_norm": 0.6155561208724976, "learning_rate": 9.167745074294598e-05, "loss": 0.0998, "step": 22700 }, { "grad_norm": 0.6751657128334045, "learning_rate": 9.166831394447913e-05, "loss": 0.0612, "step": 22710 }, { "grad_norm": 0.6507332921028137, "learning_rate": 9.165917258922812e-05, "loss": 0.0588, "step": 22720 }, { "grad_norm": 0.509975254535675, "learning_rate": 9.165002667819262e-05, "loss": 0.0572, "step": 22730 }, { "grad_norm": 0.5200465321540833, "learning_rate": 9.164087621237282e-05, "loss": 0.0523, "step": 22740 }, { "grad_norm": 0.7532728910446167, "learning_rate": 9.163172119276942e-05, "loss": 0.063, "step": 22750 }, { "grad_norm": 0.534299910068512, "learning_rate": 9.162256162038358e-05, "loss": 0.0629, "step": 22760 }, { "grad_norm": 0.5692771077156067, "learning_rate": 9.161339749621698e-05, "loss": 0.0846, "step": 22770 }, { "grad_norm": 0.6650581359863281, "learning_rate": 9.160422882127177e-05, "loss": 0.065, "step": 22780 }, { "grad_norm": 1.0179705619812012, "learning_rate": 9.159505559655069e-05, "loss": 0.0687, "step": 22790 }, { "grad_norm": 0.5103414058685303, "learning_rate": 9.158587782305684e-05, "loss": 0.0764, "step": 22800 }, { "grad_norm": 0.5204548835754395, "learning_rate": 9.157669550179391e-05, "loss": 0.0502, "step": 22810 }, { "grad_norm": 0.636674702167511, "learning_rate": 9.156750863376609e-05, "loss": 0.0647, "step": 22820 }, { "grad_norm": 0.4882219433784485, "learning_rate": 9.155831721997801e-05, "loss": 0.096, "step": 22830 }, { "grad_norm": 1.0270887613296509, "learning_rate": 9.154912126143484e-05, "loss": 0.0592, "step": 22840 }, { "grad_norm": 0.7421218156814575, "learning_rate": 9.153992075914224e-05, "loss": 0.0606, "step": 22850 }, { "grad_norm": 0.5360786318778992, "learning_rate": 9.153071571410635e-05, "loss": 0.0733, "step": 22860 }, { "grad_norm": 0.6558576226234436, "learning_rate": 9.152150612733384e-05, "loss": 0.0589, "step": 22870 }, { "grad_norm": 0.7165749073028564, "learning_rate": 9.151229199983184e-05, "loss": 0.0573, "step": 22880 }, { "grad_norm": 0.5732967257499695, "learning_rate": 9.150307333260802e-05, "loss": 0.0494, "step": 22890 }, { "grad_norm": 0.8142310976982117, "learning_rate": 9.149385012667048e-05, "loss": 0.0833, "step": 22900 }, { "grad_norm": 0.5306904315948486, "learning_rate": 9.148462238302788e-05, "loss": 0.0777, "step": 22910 }, { "grad_norm": 0.6414266228675842, "learning_rate": 9.147539010268936e-05, "loss": 0.0501, "step": 22920 }, { "grad_norm": 0.7906873226165771, "learning_rate": 9.14661532866645e-05, "loss": 0.0576, "step": 22930 }, { "grad_norm": 0.4819842576980591, "learning_rate": 9.145691193596348e-05, "loss": 0.0571, "step": 22940 }, { "grad_norm": 1.2076966762542725, "learning_rate": 9.144766605159691e-05, "loss": 0.061, "step": 22950 }, { "grad_norm": 0.6518391966819763, "learning_rate": 9.14384156345759e-05, "loss": 0.0564, "step": 22960 }, { "grad_norm": 0.7647814750671387, "learning_rate": 9.142916068591204e-05, "loss": 0.0802, "step": 22970 }, { "grad_norm": 0.7192773818969727, "learning_rate": 9.141990120661746e-05, "loss": 0.0726, "step": 22980 }, { "grad_norm": 0.7298362255096436, "learning_rate": 9.141063719770475e-05, "loss": 0.0665, "step": 22990 }, { "grad_norm": 0.6717017889022827, "learning_rate": 9.140136866018704e-05, "loss": 0.068, "step": 23000 }, { "grad_norm": 0.5893793106079102, "learning_rate": 9.139209559507788e-05, "loss": 0.0599, "step": 23010 }, { "grad_norm": 0.656234622001648, "learning_rate": 9.13828180033914e-05, "loss": 0.0602, "step": 23020 }, { "grad_norm": 0.55562424659729, "learning_rate": 9.137353588614212e-05, "loss": 0.0692, "step": 23030 }, { "grad_norm": 0.605360746383667, "learning_rate": 9.136424924434519e-05, "loss": 0.0581, "step": 23040 }, { "grad_norm": 0.755123496055603, "learning_rate": 9.135495807901615e-05, "loss": 0.0688, "step": 23050 }, { "grad_norm": 0.46289050579071045, "learning_rate": 9.134566239117108e-05, "loss": 0.0544, "step": 23060 }, { "grad_norm": 0.5391168594360352, "learning_rate": 9.13363621818265e-05, "loss": 0.0586, "step": 23070 }, { "grad_norm": 0.49454253911972046, "learning_rate": 9.132705745199953e-05, "loss": 0.0694, "step": 23080 }, { "grad_norm": 0.6979235410690308, "learning_rate": 9.131774820270768e-05, "loss": 0.0712, "step": 23090 }, { "grad_norm": 0.7874665856361389, "learning_rate": 9.130843443496901e-05, "loss": 0.0682, "step": 23100 }, { "grad_norm": 0.5657044053077698, "learning_rate": 9.129911614980206e-05, "loss": 0.0762, "step": 23110 }, { "grad_norm": 0.5772814154624939, "learning_rate": 9.128979334822584e-05, "loss": 0.0533, "step": 23120 }, { "grad_norm": 0.44908857345581055, "learning_rate": 9.128046603125992e-05, "loss": 0.0599, "step": 23130 }, { "grad_norm": 0.6933352947235107, "learning_rate": 9.12711341999243e-05, "loss": 0.0632, "step": 23140 }, { "grad_norm": 0.6843278408050537, "learning_rate": 9.12617978552395e-05, "loss": 0.0727, "step": 23150 }, { "grad_norm": 0.6531952619552612, "learning_rate": 9.12524569982265e-05, "loss": 0.0736, "step": 23160 }, { "grad_norm": 0.5552300214767456, "learning_rate": 9.124311162990684e-05, "loss": 0.0749, "step": 23170 }, { "grad_norm": 0.47173503041267395, "learning_rate": 9.12337617513025e-05, "loss": 0.0673, "step": 23180 }, { "grad_norm": 0.485993891954422, "learning_rate": 9.122440736343596e-05, "loss": 0.0598, "step": 23190 }, { "grad_norm": 0.5755847096443176, "learning_rate": 9.12150484673302e-05, "loss": 0.0524, "step": 23200 }, { "grad_norm": 0.4116760194301605, "learning_rate": 9.120568506400873e-05, "loss": 0.09, "step": 23210 }, { "grad_norm": 1.013307809829712, "learning_rate": 9.119631715449548e-05, "loss": 0.0724, "step": 23220 }, { "grad_norm": 0.46334144473075867, "learning_rate": 9.118694473981493e-05, "loss": 0.0713, "step": 23230 }, { "grad_norm": 0.7590882182121277, "learning_rate": 9.117756782099203e-05, "loss": 0.0712, "step": 23240 }, { "grad_norm": 0.503174364566803, "learning_rate": 9.11681863990522e-05, "loss": 0.0821, "step": 23250 }, { "grad_norm": 0.5532654523849487, "learning_rate": 9.115880047502142e-05, "loss": 0.0959, "step": 23260 }, { "grad_norm": 0.4137432873249054, "learning_rate": 9.114941004992609e-05, "loss": 0.0569, "step": 23270 }, { "grad_norm": 0.7407957315444946, "learning_rate": 9.114001512479317e-05, "loss": 0.0626, "step": 23280 }, { "grad_norm": 0.7147806882858276, "learning_rate": 9.113061570065003e-05, "loss": 0.0714, "step": 23290 }, { "grad_norm": 0.8099187612533569, "learning_rate": 9.112121177852459e-05, "loss": 0.0732, "step": 23300 }, { "grad_norm": 0.6858149170875549, "learning_rate": 9.111180335944527e-05, "loss": 0.0629, "step": 23310 }, { "grad_norm": 0.5956851243972778, "learning_rate": 9.110239044444093e-05, "loss": 0.0567, "step": 23320 }, { "grad_norm": 0.7120476961135864, "learning_rate": 9.109297303454099e-05, "loss": 0.0639, "step": 23330 }, { "grad_norm": 0.5340265035629272, "learning_rate": 9.108355113077526e-05, "loss": 0.0475, "step": 23340 }, { "grad_norm": 0.5687625408172607, "learning_rate": 9.107412473417419e-05, "loss": 0.0656, "step": 23350 }, { "grad_norm": 0.43560048937797546, "learning_rate": 9.106469384576858e-05, "loss": 0.0611, "step": 23360 }, { "grad_norm": 0.5287518501281738, "learning_rate": 9.105525846658978e-05, "loss": 0.0575, "step": 23370 }, { "grad_norm": 0.7333928942680359, "learning_rate": 9.104581859766965e-05, "loss": 0.0625, "step": 23380 }, { "grad_norm": 0.6420456171035767, "learning_rate": 9.10363742400405e-05, "loss": 0.0624, "step": 23390 }, { "grad_norm": 1.9934180974960327, "learning_rate": 9.102692539473518e-05, "loss": 0.064, "step": 23400 }, { "grad_norm": 0.5922147035598755, "learning_rate": 9.101747206278697e-05, "loss": 0.0509, "step": 23410 }, { "grad_norm": 0.5387743711471558, "learning_rate": 9.100801424522968e-05, "loss": 0.0604, "step": 23420 }, { "grad_norm": 0.7268848419189453, "learning_rate": 9.099855194309762e-05, "loss": 0.0686, "step": 23430 }, { "grad_norm": 0.5674452781677246, "learning_rate": 9.098908515742554e-05, "loss": 0.057, "step": 23440 }, { "grad_norm": 0.5196605920791626, "learning_rate": 9.097961388924873e-05, "loss": 0.0522, "step": 23450 }, { "grad_norm": 0.5196019411087036, "learning_rate": 9.097013813960298e-05, "loss": 0.0471, "step": 23460 }, { "grad_norm": 0.46149006485939026, "learning_rate": 9.09606579095245e-05, "loss": 0.0554, "step": 23470 }, { "grad_norm": 0.7960423231124878, "learning_rate": 9.095117320005008e-05, "loss": 0.0722, "step": 23480 }, { "grad_norm": 0.7193050980567932, "learning_rate": 9.094168401221691e-05, "loss": 0.056, "step": 23490 }, { "grad_norm": 0.5614984631538391, "learning_rate": 9.093219034706273e-05, "loss": 0.075, "step": 23500 }, { "grad_norm": 0.7614705562591553, "learning_rate": 9.092269220562577e-05, "loss": 0.0602, "step": 23510 }, { "grad_norm": 0.5399461984634399, "learning_rate": 9.09131895889447e-05, "loss": 0.045, "step": 23520 }, { "grad_norm": 0.5022040605545044, "learning_rate": 9.090368249805873e-05, "loss": 0.0611, "step": 23530 }, { "grad_norm": 0.854101836681366, "learning_rate": 9.089417093400754e-05, "loss": 0.0692, "step": 23540 }, { "grad_norm": 0.6588537693023682, "learning_rate": 9.088465489783131e-05, "loss": 0.0674, "step": 23550 }, { "grad_norm": 0.7216013669967651, "learning_rate": 9.087513439057068e-05, "loss": 0.0727, "step": 23560 }, { "grad_norm": 0.6452583074569702, "learning_rate": 9.08656094132668e-05, "loss": 0.0661, "step": 23570 }, { "grad_norm": 0.9707720875740051, "learning_rate": 9.085607996696134e-05, "loss": 0.0738, "step": 23580 }, { "grad_norm": 0.8386173248291016, "learning_rate": 9.084654605269639e-05, "loss": 0.0769, "step": 23590 }, { "grad_norm": 0.5353692770004272, "learning_rate": 9.083700767151457e-05, "loss": 0.0719, "step": 23600 }, { "grad_norm": 0.985127866268158, "learning_rate": 9.082746482445898e-05, "loss": 0.0804, "step": 23610 }, { "grad_norm": 0.6578841209411621, "learning_rate": 9.081791751257325e-05, "loss": 0.0643, "step": 23620 }, { "grad_norm": 0.5634751915931702, "learning_rate": 9.080836573690142e-05, "loss": 0.0657, "step": 23630 }, { "grad_norm": 0.5432795286178589, "learning_rate": 9.079880949848805e-05, "loss": 0.0575, "step": 23640 }, { "grad_norm": 0.6784237623214722, "learning_rate": 9.078924879837822e-05, "loss": 0.0587, "step": 23650 }, { "grad_norm": 0.9184025526046753, "learning_rate": 9.077968363761747e-05, "loss": 0.0917, "step": 23660 }, { "grad_norm": 1.197340726852417, "learning_rate": 9.077011401725182e-05, "loss": 0.0547, "step": 23670 }, { "grad_norm": 0.6155344843864441, "learning_rate": 9.07605399383278e-05, "loss": 0.058, "step": 23680 }, { "grad_norm": 0.5075444579124451, "learning_rate": 9.075096140189243e-05, "loss": 0.064, "step": 23690 }, { "grad_norm": 0.8639175891876221, "learning_rate": 9.074137840899318e-05, "loss": 0.0664, "step": 23700 }, { "grad_norm": 0.6487533450126648, "learning_rate": 9.073179096067804e-05, "loss": 0.0713, "step": 23710 }, { "grad_norm": 0.6027271747589111, "learning_rate": 9.072219905799549e-05, "loss": 0.0584, "step": 23720 }, { "grad_norm": 0.6247918009757996, "learning_rate": 9.071260270199447e-05, "loss": 0.0645, "step": 23730 }, { "grad_norm": 0.4511493444442749, "learning_rate": 9.070300189372441e-05, "loss": 0.0585, "step": 23740 }, { "grad_norm": 0.7477952241897583, "learning_rate": 9.069339663423528e-05, "loss": 0.0913, "step": 23750 }, { "grad_norm": 0.6259909868240356, "learning_rate": 9.068378692457747e-05, "loss": 0.0661, "step": 23760 }, { "grad_norm": 0.5415745973587036, "learning_rate": 9.067417276580189e-05, "loss": 0.0575, "step": 23770 }, { "grad_norm": 0.4444354474544525, "learning_rate": 9.066455415895993e-05, "loss": 0.0559, "step": 23780 }, { "grad_norm": 0.6461361050605774, "learning_rate": 9.065493110510346e-05, "loss": 0.0559, "step": 23790 }, { "grad_norm": 0.5823233127593994, "learning_rate": 9.064530360528484e-05, "loss": 0.0627, "step": 23800 }, { "grad_norm": 0.9023003578186035, "learning_rate": 9.063567166055695e-05, "loss": 0.0701, "step": 23810 }, { "grad_norm": 0.8120307326316833, "learning_rate": 9.062603527197308e-05, "loss": 0.0588, "step": 23820 }, { "grad_norm": 0.6789265871047974, "learning_rate": 9.06163944405871e-05, "loss": 0.0473, "step": 23830 }, { "grad_norm": 0.597449779510498, "learning_rate": 9.060674916745327e-05, "loss": 0.0536, "step": 23840 }, { "grad_norm": 0.7102142572402954, "learning_rate": 9.05970994536264e-05, "loss": 0.0743, "step": 23850 }, { "grad_norm": 0.6369552612304688, "learning_rate": 9.05874453001618e-05, "loss": 0.0713, "step": 23860 }, { "grad_norm": 0.6240330338478088, "learning_rate": 9.057778670811517e-05, "loss": 0.0553, "step": 23870 }, { "grad_norm": 0.6095353364944458, "learning_rate": 9.056812367854281e-05, "loss": 0.0641, "step": 23880 }, { "grad_norm": 0.5308515429496765, "learning_rate": 9.055845621250143e-05, "loss": 0.0576, "step": 23890 }, { "grad_norm": 0.6675844192504883, "learning_rate": 9.054878431104825e-05, "loss": 0.0571, "step": 23900 }, { "grad_norm": 0.7340705990791321, "learning_rate": 9.0539107975241e-05, "loss": 0.0575, "step": 23910 }, { "grad_norm": 0.616741418838501, "learning_rate": 9.052942720613784e-05, "loss": 0.0548, "step": 23920 }, { "grad_norm": 0.6079967617988586, "learning_rate": 9.051974200479745e-05, "loss": 0.0519, "step": 23930 }, { "grad_norm": 0.5797609686851501, "learning_rate": 9.051005237227901e-05, "loss": 0.0654, "step": 23940 }, { "grad_norm": 0.644737958908081, "learning_rate": 9.050035830964215e-05, "loss": 0.0647, "step": 23950 }, { "grad_norm": 0.6237507462501526, "learning_rate": 9.049065981794698e-05, "loss": 0.0663, "step": 23960 }, { "grad_norm": 0.5485249161720276, "learning_rate": 9.048095689825414e-05, "loss": 0.0697, "step": 23970 }, { "grad_norm": 0.4986760914325714, "learning_rate": 9.047124955162472e-05, "loss": 0.0612, "step": 23980 }, { "grad_norm": 1.2353469133377075, "learning_rate": 9.046153777912028e-05, "loss": 0.0843, "step": 23990 }, { "grad_norm": 1.0663981437683105, "learning_rate": 9.045182158180292e-05, "loss": 0.0853, "step": 24000 }, { "grad_norm": 0.6710480451583862, "learning_rate": 9.044210096073516e-05, "loss": 0.0782, "step": 24010 }, { "grad_norm": 0.9554337859153748, "learning_rate": 9.043237591698004e-05, "loss": 0.0657, "step": 24020 }, { "grad_norm": 0.7013730406761169, "learning_rate": 9.04226464516011e-05, "loss": 0.0622, "step": 24030 }, { "grad_norm": 0.610946536064148, "learning_rate": 9.041291256566229e-05, "loss": 0.059, "step": 24040 }, { "grad_norm": 0.6270118951797485, "learning_rate": 9.040317426022814e-05, "loss": 0.059, "step": 24050 }, { "grad_norm": 0.5724241733551025, "learning_rate": 9.03934315363636e-05, "loss": 0.0494, "step": 24060 }, { "grad_norm": 0.7155265808105469, "learning_rate": 9.038368439513409e-05, "loss": 0.0591, "step": 24070 }, { "grad_norm": 0.8880976438522339, "learning_rate": 9.03739328376056e-05, "loss": 0.0833, "step": 24080 }, { "grad_norm": 0.6099636554718018, "learning_rate": 9.036417686484451e-05, "loss": 0.0568, "step": 24090 }, { "grad_norm": 0.5730770230293274, "learning_rate": 9.035441647791773e-05, "loss": 0.066, "step": 24100 }, { "grad_norm": 0.49644267559051514, "learning_rate": 9.034465167789263e-05, "loss": 0.058, "step": 24110 }, { "grad_norm": 0.6206948161125183, "learning_rate": 9.033488246583706e-05, "loss": 0.0649, "step": 24120 }, { "grad_norm": 0.5309808254241943, "learning_rate": 9.032510884281941e-05, "loss": 0.0653, "step": 24130 }, { "grad_norm": 0.545952320098877, "learning_rate": 9.031533080990848e-05, "loss": 0.0788, "step": 24140 }, { "grad_norm": 0.7698030471801758, "learning_rate": 9.030554836817358e-05, "loss": 0.0602, "step": 24150 }, { "grad_norm": 0.6144784092903137, "learning_rate": 9.029576151868451e-05, "loss": 0.0656, "step": 24160 }, { "grad_norm": 0.6520487666130066, "learning_rate": 9.028597026251155e-05, "loss": 0.0538, "step": 24170 }, { "grad_norm": 0.9589599967002869, "learning_rate": 9.027617460072547e-05, "loss": 0.0586, "step": 24180 }, { "grad_norm": 0.7894977927207947, "learning_rate": 9.026637453439745e-05, "loss": 0.0683, "step": 24190 }, { "grad_norm": 0.42778879404067993, "learning_rate": 9.025657006459927e-05, "loss": 0.069, "step": 24200 }, { "grad_norm": 0.8622857332229614, "learning_rate": 9.024676119240311e-05, "loss": 0.0663, "step": 24210 }, { "grad_norm": 0.43775931000709534, "learning_rate": 9.023694791888166e-05, "loss": 0.062, "step": 24220 }, { "grad_norm": 0.4888722896575928, "learning_rate": 9.022713024510808e-05, "loss": 0.0415, "step": 24230 }, { "grad_norm": 0.7335543036460876, "learning_rate": 9.021730817215601e-05, "loss": 0.0622, "step": 24240 }, { "grad_norm": 0.662582278251648, "learning_rate": 9.02074817010996e-05, "loss": 0.0542, "step": 24250 }, { "grad_norm": 0.4036911725997925, "learning_rate": 9.019765083301342e-05, "loss": 0.0638, "step": 24260 }, { "grad_norm": 0.6039438843727112, "learning_rate": 9.01878155689726e-05, "loss": 0.0623, "step": 24270 }, { "grad_norm": 0.4356478154659271, "learning_rate": 9.017797591005268e-05, "loss": 0.0616, "step": 24280 }, { "grad_norm": 0.547177255153656, "learning_rate": 9.016813185732972e-05, "loss": 0.0491, "step": 24290 }, { "grad_norm": 1.0306501388549805, "learning_rate": 9.015828341188027e-05, "loss": 0.0563, "step": 24300 }, { "grad_norm": 0.6381553411483765, "learning_rate": 9.01484305747813e-05, "loss": 0.0601, "step": 24310 }, { "grad_norm": 1.1495602130889893, "learning_rate": 9.013857334711033e-05, "loss": 0.0505, "step": 24320 }, { "grad_norm": 0.632123589515686, "learning_rate": 9.012871172994534e-05, "loss": 0.0447, "step": 24330 }, { "grad_norm": 0.5180799961090088, "learning_rate": 9.011884572436476e-05, "loss": 0.0561, "step": 24340 }, { "grad_norm": 0.6772035360336304, "learning_rate": 9.010897533144754e-05, "loss": 0.0704, "step": 24350 }, { "grad_norm": 1.0022151470184326, "learning_rate": 9.009910055227306e-05, "loss": 0.0798, "step": 24360 }, { "grad_norm": 0.8269349336624146, "learning_rate": 9.008922138792124e-05, "loss": 0.0598, "step": 24370 }, { "grad_norm": 0.7772126793861389, "learning_rate": 9.007933783947244e-05, "loss": 0.0642, "step": 24380 }, { "grad_norm": 0.6046224236488342, "learning_rate": 9.006944990800752e-05, "loss": 0.0895, "step": 24390 }, { "grad_norm": 1.222133755683899, "learning_rate": 9.005955759460779e-05, "loss": 0.0637, "step": 24400 }, { "grad_norm": 0.4686540961265564, "learning_rate": 9.004966090035508e-05, "loss": 0.098, "step": 24410 }, { "grad_norm": 0.5779590606689453, "learning_rate": 9.003975982633166e-05, "loss": 0.0631, "step": 24420 }, { "grad_norm": 0.45769834518432617, "learning_rate": 9.00298543736203e-05, "loss": 0.0577, "step": 24430 }, { "grad_norm": 0.509373664855957, "learning_rate": 9.001994454330427e-05, "loss": 0.0611, "step": 24440 }, { "grad_norm": 0.7192037105560303, "learning_rate": 9.001003033646727e-05, "loss": 0.0629, "step": 24450 }, { "grad_norm": 0.5596395134925842, "learning_rate": 9.00001117541935e-05, "loss": 0.0735, "step": 24460 }, { "grad_norm": 0.4470633864402771, "learning_rate": 8.999018879756764e-05, "loss": 0.0505, "step": 24470 }, { "grad_norm": 0.46312248706817627, "learning_rate": 8.998026146767487e-05, "loss": 0.0525, "step": 24480 }, { "grad_norm": 0.7938422560691833, "learning_rate": 8.99703297656008e-05, "loss": 0.0985, "step": 24490 }, { "grad_norm": 0.6567539572715759, "learning_rate": 8.996039369243156e-05, "loss": 0.0627, "step": 24500 }, { "grad_norm": 0.5918510556221008, "learning_rate": 8.995045324925378e-05, "loss": 0.0848, "step": 24510 }, { "grad_norm": 0.65004962682724, "learning_rate": 8.994050843715448e-05, "loss": 0.0832, "step": 24520 }, { "grad_norm": 1.0291633605957031, "learning_rate": 8.993055925722121e-05, "loss": 0.0716, "step": 24530 }, { "grad_norm": 0.6142466068267822, "learning_rate": 8.992060571054202e-05, "loss": 0.0723, "step": 24540 }, { "grad_norm": 0.7506587505340576, "learning_rate": 8.991064779820542e-05, "loss": 0.0622, "step": 24550 }, { "grad_norm": 0.5549033284187317, "learning_rate": 8.990068552130036e-05, "loss": 0.0601, "step": 24560 }, { "grad_norm": 0.7754573225975037, "learning_rate": 8.989071888091634e-05, "loss": 0.0779, "step": 24570 }, { "grad_norm": 0.5821207761764526, "learning_rate": 8.988074787814329e-05, "loss": 0.0652, "step": 24580 }, { "grad_norm": 0.4542984366416931, "learning_rate": 8.987077251407158e-05, "loss": 0.0522, "step": 24590 }, { "grad_norm": 0.5828810334205627, "learning_rate": 8.986079278979216e-05, "loss": 0.0609, "step": 24600 }, { "grad_norm": 0.8342454433441162, "learning_rate": 8.985080870639635e-05, "loss": 0.0703, "step": 24610 }, { "grad_norm": 0.7165471911430359, "learning_rate": 8.984082026497603e-05, "loss": 0.0561, "step": 24620 }, { "grad_norm": 0.4694603383541107, "learning_rate": 8.98308274666235e-05, "loss": 0.0744, "step": 24630 }, { "grad_norm": 0.6157486438751221, "learning_rate": 8.982083031243155e-05, "loss": 0.0572, "step": 24640 }, { "grad_norm": 0.6622006893157959, "learning_rate": 8.98108288034935e-05, "loss": 0.0602, "step": 24650 }, { "grad_norm": 0.5775927901268005, "learning_rate": 8.980082294090305e-05, "loss": 0.0556, "step": 24660 }, { "grad_norm": 0.47657859325408936, "learning_rate": 8.979081272575443e-05, "loss": 0.0643, "step": 24670 }, { "grad_norm": 0.5854668021202087, "learning_rate": 8.978079815914236e-05, "loss": 0.0729, "step": 24680 }, { "grad_norm": 0.5096737742424011, "learning_rate": 8.977077924216202e-05, "loss": 0.0569, "step": 24690 }, { "grad_norm": 0.8622095584869385, "learning_rate": 8.976075597590905e-05, "loss": 0.0759, "step": 24700 }, { "grad_norm": 0.8851601481437683, "learning_rate": 8.975072836147958e-05, "loss": 0.0876, "step": 24710 }, { "grad_norm": 0.4805409610271454, "learning_rate": 8.974069639997025e-05, "loss": 0.0629, "step": 24720 }, { "grad_norm": 0.44932690262794495, "learning_rate": 8.973066009247808e-05, "loss": 0.079, "step": 24730 }, { "grad_norm": 0.4485713839530945, "learning_rate": 8.972061944010066e-05, "loss": 0.0644, "step": 24740 }, { "grad_norm": 0.5730527639389038, "learning_rate": 8.971057444393603e-05, "loss": 0.0686, "step": 24750 }, { "grad_norm": 0.8202264904975891, "learning_rate": 8.970052510508268e-05, "loss": 0.0699, "step": 24760 }, { "grad_norm": 0.6030222177505493, "learning_rate": 8.969047142463959e-05, "loss": 0.0651, "step": 24770 }, { "grad_norm": 0.8423622846603394, "learning_rate": 8.968041340370621e-05, "loss": 0.0706, "step": 24780 }, { "grad_norm": 0.7619331479072571, "learning_rate": 8.96703510433825e-05, "loss": 0.0741, "step": 24790 }, { "grad_norm": 0.48075371980667114, "learning_rate": 8.966028434476883e-05, "loss": 0.0677, "step": 24800 }, { "grad_norm": 0.5112065076828003, "learning_rate": 8.96502133089661e-05, "loss": 0.0731, "step": 24810 }, { "grad_norm": 0.514659583568573, "learning_rate": 8.964013793707564e-05, "loss": 0.0699, "step": 24820 }, { "grad_norm": 0.7293438911437988, "learning_rate": 8.963005823019932e-05, "loss": 0.062, "step": 24830 }, { "grad_norm": 0.5212116837501526, "learning_rate": 8.961997418943939e-05, "loss": 0.0684, "step": 24840 }, { "grad_norm": 0.8444229960441589, "learning_rate": 8.960988581589865e-05, "loss": 0.0627, "step": 24850 }, { "grad_norm": 1.0088636875152588, "learning_rate": 8.959979311068037e-05, "loss": 0.0866, "step": 24860 }, { "grad_norm": 0.7978466749191284, "learning_rate": 8.958969607488823e-05, "loss": 0.0869, "step": 24870 }, { "grad_norm": 0.6540276408195496, "learning_rate": 8.957959470962647e-05, "loss": 0.0784, "step": 24880 }, { "grad_norm": 0.7187244296073914, "learning_rate": 8.956948901599971e-05, "loss": 0.061, "step": 24890 }, { "grad_norm": 0.5153450965881348, "learning_rate": 8.955937899511315e-05, "loss": 0.0621, "step": 24900 }, { "grad_norm": 0.8027661442756653, "learning_rate": 8.954926464807238e-05, "loss": 0.0654, "step": 24910 }, { "grad_norm": 0.7076085209846497, "learning_rate": 8.953914597598347e-05, "loss": 0.0788, "step": 24920 }, { "grad_norm": 0.5480613112449646, "learning_rate": 8.952902297995303e-05, "loss": 0.061, "step": 24930 }, { "grad_norm": 0.6596411466598511, "learning_rate": 8.951889566108804e-05, "loss": 0.0664, "step": 24940 }, { "grad_norm": 0.5503808856010437, "learning_rate": 8.950876402049606e-05, "loss": 0.06, "step": 24950 }, { "grad_norm": 1.2386629581451416, "learning_rate": 8.949862805928504e-05, "loss": 0.0705, "step": 24960 }, { "grad_norm": 0.6561667919158936, "learning_rate": 8.948848777856343e-05, "loss": 0.0558, "step": 24970 }, { "grad_norm": 0.9050532579421997, "learning_rate": 8.947834317944017e-05, "loss": 0.0542, "step": 24980 }, { "grad_norm": 0.48528799414634705, "learning_rate": 8.946819426302466e-05, "loss": 0.0689, "step": 24990 }, { "grad_norm": 0.7856316566467285, "learning_rate": 8.945804103042676e-05, "loss": 0.0486, "step": 25000 }, { "grad_norm": 0.9442822337150574, "learning_rate": 8.944788348275681e-05, "loss": 0.0634, "step": 25010 }, { "grad_norm": 0.6154482960700989, "learning_rate": 8.943772162112565e-05, "loss": 0.0844, "step": 25020 }, { "grad_norm": 0.5514326691627502, "learning_rate": 8.942755544664454e-05, "loss": 0.0739, "step": 25030 }, { "grad_norm": 0.8631011843681335, "learning_rate": 8.941738496042525e-05, "loss": 0.0595, "step": 25040 }, { "grad_norm": 0.7732834219932556, "learning_rate": 8.940721016357999e-05, "loss": 0.0723, "step": 25050 }, { "grad_norm": 0.6733863353729248, "learning_rate": 8.939703105722148e-05, "loss": 0.059, "step": 25060 }, { "grad_norm": 0.38918620347976685, "learning_rate": 8.93868476424629e-05, "loss": 0.0634, "step": 25070 }, { "grad_norm": 1.040685772895813, "learning_rate": 8.937665992041786e-05, "loss": 0.0785, "step": 25080 }, { "grad_norm": 1.2819304466247559, "learning_rate": 8.93664678922005e-05, "loss": 0.0769, "step": 25090 }, { "grad_norm": 0.7075808048248291, "learning_rate": 8.93562715589254e-05, "loss": 0.065, "step": 25100 }, { "grad_norm": 0.5808641314506531, "learning_rate": 8.934607092170762e-05, "loss": 0.0667, "step": 25110 }, { "grad_norm": 0.6855165958404541, "learning_rate": 8.933586598166266e-05, "loss": 0.0605, "step": 25120 }, { "grad_norm": 0.4893777370452881, "learning_rate": 8.932565673990655e-05, "loss": 0.0703, "step": 25130 }, { "grad_norm": 0.5087274312973022, "learning_rate": 8.931544319755574e-05, "loss": 0.0697, "step": 25140 }, { "grad_norm": 0.5444875955581665, "learning_rate": 8.930522535572718e-05, "loss": 0.0563, "step": 25150 }, { "grad_norm": 0.45097610354423523, "learning_rate": 8.929500321553826e-05, "loss": 0.0591, "step": 25160 }, { "grad_norm": 0.6113709807395935, "learning_rate": 8.928477677810686e-05, "loss": 0.0904, "step": 25170 }, { "grad_norm": 0.43256476521492004, "learning_rate": 8.927454604455137e-05, "loss": 0.0433, "step": 25180 }, { "grad_norm": 0.4912196397781372, "learning_rate": 8.926431101599053e-05, "loss": 0.0465, "step": 25190 }, { "grad_norm": 0.6865953803062439, "learning_rate": 8.925407169354369e-05, "loss": 0.0697, "step": 25200 }, { "grad_norm": 0.4513121545314789, "learning_rate": 8.92438280783306e-05, "loss": 0.0487, "step": 25210 }, { "grad_norm": 0.4104909300804138, "learning_rate": 8.923358017147146e-05, "loss": 0.0484, "step": 25220 }, { "grad_norm": 0.6918448209762573, "learning_rate": 8.922332797408697e-05, "loss": 0.054, "step": 25230 }, { "grad_norm": 0.524768590927124, "learning_rate": 8.921307148729831e-05, "loss": 0.0521, "step": 25240 }, { "grad_norm": 0.42721429467201233, "learning_rate": 8.920281071222712e-05, "loss": 0.0466, "step": 25250 }, { "grad_norm": 0.4675832986831665, "learning_rate": 8.919254564999548e-05, "loss": 0.0594, "step": 25260 }, { "grad_norm": 0.5682530999183655, "learning_rate": 8.918227630172598e-05, "loss": 0.0628, "step": 25270 }, { "grad_norm": 0.7192248702049255, "learning_rate": 8.917200266854165e-05, "loss": 0.0682, "step": 25280 }, { "grad_norm": 0.418334424495697, "learning_rate": 8.9161724751566e-05, "loss": 0.052, "step": 25290 }, { "grad_norm": 0.6457694172859192, "learning_rate": 8.915144255192302e-05, "loss": 0.0531, "step": 25300 }, { "grad_norm": 0.4309857189655304, "learning_rate": 8.914115607073714e-05, "loss": 0.0425, "step": 25310 }, { "grad_norm": 0.8977031707763672, "learning_rate": 8.913086530913327e-05, "loss": 0.0822, "step": 25320 }, { "grad_norm": 0.5137282013893127, "learning_rate": 8.912057026823681e-05, "loss": 0.0569, "step": 25330 }, { "grad_norm": 0.4598686695098877, "learning_rate": 8.91102709491736e-05, "loss": 0.0574, "step": 25340 }, { "grad_norm": 1.0068782567977905, "learning_rate": 8.909996735306996e-05, "loss": 0.0612, "step": 25350 }, { "grad_norm": 0.35684365034103394, "learning_rate": 8.908965948105268e-05, "loss": 0.074, "step": 25360 }, { "grad_norm": 0.5720511674880981, "learning_rate": 8.907934733424901e-05, "loss": 0.0616, "step": 25370 }, { "grad_norm": 0.9088981747627258, "learning_rate": 8.906903091378666e-05, "loss": 0.0732, "step": 25380 }, { "grad_norm": 0.5194448828697205, "learning_rate": 8.905871022079384e-05, "loss": 0.0503, "step": 25390 }, { "grad_norm": 0.8052963018417358, "learning_rate": 8.90483852563992e-05, "loss": 0.0691, "step": 25400 }, { "grad_norm": 0.5402407050132751, "learning_rate": 8.903805602173185e-05, "loss": 0.0483, "step": 25410 }, { "grad_norm": 0.6598817110061646, "learning_rate": 8.902772251792137e-05, "loss": 0.0576, "step": 25420 }, { "grad_norm": 0.5453411340713501, "learning_rate": 8.901738474609786e-05, "loss": 0.0664, "step": 25430 }, { "grad_norm": 0.7799461483955383, "learning_rate": 8.900704270739179e-05, "loss": 0.0587, "step": 25440 }, { "grad_norm": 0.6327877044677734, "learning_rate": 8.89966964029342e-05, "loss": 0.0541, "step": 25450 }, { "grad_norm": 0.531284511089325, "learning_rate": 8.898634583385652e-05, "loss": 0.0619, "step": 25460 }, { "grad_norm": 0.6107194423675537, "learning_rate": 8.897599100129065e-05, "loss": 0.0584, "step": 25470 }, { "grad_norm": 0.47323888540267944, "learning_rate": 8.896563190636903e-05, "loss": 0.0587, "step": 25480 }, { "grad_norm": 0.43650394678115845, "learning_rate": 8.895526855022448e-05, "loss": 0.042, "step": 25490 }, { "grad_norm": 0.5693970918655396, "learning_rate": 8.894490093399033e-05, "loss": 0.057, "step": 25500 }, { "grad_norm": 0.6039342284202576, "learning_rate": 8.893452905880035e-05, "loss": 0.0715, "step": 25510 }, { "grad_norm": 0.9000293612480164, "learning_rate": 8.892415292578883e-05, "loss": 0.0736, "step": 25520 }, { "grad_norm": 0.6959168314933777, "learning_rate": 8.891377253609046e-05, "loss": 0.0813, "step": 25530 }, { "grad_norm": 0.5444051027297974, "learning_rate": 8.890338789084043e-05, "loss": 0.0572, "step": 25540 }, { "grad_norm": 0.5764027833938599, "learning_rate": 8.88929989911744e-05, "loss": 0.0631, "step": 25550 }, { "grad_norm": 0.5739176869392395, "learning_rate": 8.888260583822847e-05, "loss": 0.0639, "step": 25560 }, { "grad_norm": 0.7358345985412598, "learning_rate": 8.887220843313921e-05, "loss": 0.0741, "step": 25570 }, { "grad_norm": 0.5876644849777222, "learning_rate": 8.88618067770437e-05, "loss": 0.0508, "step": 25580 }, { "grad_norm": 0.6245595812797546, "learning_rate": 8.885140087107942e-05, "loss": 0.056, "step": 25590 }, { "grad_norm": 0.6732368469238281, "learning_rate": 8.884099071638436e-05, "loss": 0.071, "step": 25600 }, { "grad_norm": 0.6022033095359802, "learning_rate": 8.883057631409695e-05, "loss": 0.0685, "step": 25610 }, { "grad_norm": 0.6269504427909851, "learning_rate": 8.882015766535608e-05, "loss": 0.0555, "step": 25620 }, { "grad_norm": 0.4246881902217865, "learning_rate": 8.880973477130115e-05, "loss": 0.0557, "step": 25630 }, { "grad_norm": 0.6342000961303711, "learning_rate": 8.879930763307197e-05, "loss": 0.0546, "step": 25640 }, { "grad_norm": 0.6798269748687744, "learning_rate": 8.878887625180884e-05, "loss": 0.0552, "step": 25650 }, { "grad_norm": 0.844990611076355, "learning_rate": 8.877844062865253e-05, "loss": 0.0691, "step": 25660 }, { "grad_norm": 0.5903769135475159, "learning_rate": 8.876800076474424e-05, "loss": 0.0436, "step": 25670 }, { "grad_norm": 0.4599197506904602, "learning_rate": 8.875755666122568e-05, "loss": 0.0533, "step": 25680 }, { "grad_norm": 0.5671740174293518, "learning_rate": 8.8747108319239e-05, "loss": 0.0546, "step": 25690 }, { "grad_norm": 0.6423686742782593, "learning_rate": 8.87366557399268e-05, "loss": 0.0588, "step": 25700 }, { "grad_norm": 0.6758549809455872, "learning_rate": 8.872619892443217e-05, "loss": 0.069, "step": 25710 }, { "grad_norm": 0.456731379032135, "learning_rate": 8.871573787389865e-05, "loss": 0.0601, "step": 25720 }, { "grad_norm": 0.46491867303848267, "learning_rate": 8.870527258947024e-05, "loss": 0.0599, "step": 25730 }, { "grad_norm": 0.7233842015266418, "learning_rate": 8.869480307229143e-05, "loss": 0.0563, "step": 25740 }, { "grad_norm": 0.760837197303772, "learning_rate": 8.868432932350712e-05, "loss": 0.0573, "step": 25750 }, { "grad_norm": 0.9364517331123352, "learning_rate": 8.867385134426272e-05, "loss": 0.0698, "step": 25760 }, { "grad_norm": 0.4391074478626251, "learning_rate": 8.866336913570407e-05, "loss": 0.0577, "step": 25770 }, { "grad_norm": 0.7591128349304199, "learning_rate": 8.865288269897751e-05, "loss": 0.0714, "step": 25780 }, { "grad_norm": 0.5375124216079712, "learning_rate": 8.864239203522981e-05, "loss": 0.0748, "step": 25790 }, { "grad_norm": 0.44449856877326965, "learning_rate": 8.863189714560822e-05, "loss": 0.0649, "step": 25800 }, { "grad_norm": 0.8169793486595154, "learning_rate": 8.862139803126043e-05, "loss": 0.0659, "step": 25810 }, { "grad_norm": 0.5189172625541687, "learning_rate": 8.861089469333463e-05, "loss": 0.0519, "step": 25820 }, { "grad_norm": 0.4896983504295349, "learning_rate": 8.860038713297944e-05, "loss": 0.0477, "step": 25830 }, { "grad_norm": 0.5876211524009705, "learning_rate": 8.858987535134394e-05, "loss": 0.0573, "step": 25840 }, { "grad_norm": 0.616701066493988, "learning_rate": 8.857935934957769e-05, "loss": 0.0455, "step": 25850 }, { "grad_norm": 0.473259299993515, "learning_rate": 8.856883912883071e-05, "loss": 0.0694, "step": 25860 }, { "grad_norm": 0.797680675983429, "learning_rate": 8.855831469025346e-05, "loss": 0.0567, "step": 25870 }, { "grad_norm": 0.6511815786361694, "learning_rate": 8.854778603499689e-05, "loss": 0.063, "step": 25880 }, { "grad_norm": 0.8328746557235718, "learning_rate": 8.85372531642124e-05, "loss": 0.0569, "step": 25890 }, { "grad_norm": 0.6366116404533386, "learning_rate": 8.852671607905185e-05, "loss": 0.0493, "step": 25900 }, { "grad_norm": 0.6113909482955933, "learning_rate": 8.851617478066754e-05, "loss": 0.0475, "step": 25910 }, { "grad_norm": 0.944637656211853, "learning_rate": 8.850562927021227e-05, "loss": 0.0618, "step": 25920 }, { "grad_norm": 1.1985292434692383, "learning_rate": 8.849507954883928e-05, "loss": 0.0627, "step": 25930 }, { "grad_norm": 0.4354534447193146, "learning_rate": 8.848452561770226e-05, "loss": 0.0525, "step": 25940 }, { "grad_norm": 0.518623411655426, "learning_rate": 8.847396747795538e-05, "loss": 0.0457, "step": 25950 }, { "grad_norm": 0.6880924105644226, "learning_rate": 8.846340513075327e-05, "loss": 0.0482, "step": 25960 }, { "grad_norm": 0.3822615146636963, "learning_rate": 8.845283857725099e-05, "loss": 0.0537, "step": 25970 }, { "grad_norm": 0.389472097158432, "learning_rate": 8.844226781860409e-05, "loss": 0.0487, "step": 25980 }, { "grad_norm": 0.44964101910591125, "learning_rate": 8.84316928559686e-05, "loss": 0.0596, "step": 25990 }, { "grad_norm": 0.8744215965270996, "learning_rate": 8.842111369050094e-05, "loss": 0.0529, "step": 26000 }, { "grad_norm": 0.9028172492980957, "learning_rate": 8.841053032335808e-05, "loss": 0.0744, "step": 26010 }, { "grad_norm": 0.7772029042243958, "learning_rate": 8.839994275569735e-05, "loss": 0.0795, "step": 26020 }, { "grad_norm": 1.0182366371154785, "learning_rate": 8.838935098867662e-05, "loss": 0.0695, "step": 26030 }, { "grad_norm": 0.6821036338806152, "learning_rate": 8.837875502345418e-05, "loss": 0.0726, "step": 26040 }, { "grad_norm": 0.576251745223999, "learning_rate": 8.83681548611888e-05, "loss": 0.071, "step": 26050 }, { "grad_norm": 0.4824022054672241, "learning_rate": 8.835755050303969e-05, "loss": 0.0726, "step": 26060 }, { "grad_norm": 0.6485371589660645, "learning_rate": 8.834694195016653e-05, "loss": 0.0573, "step": 26070 }, { "grad_norm": 0.6002235412597656, "learning_rate": 8.833632920372942e-05, "loss": 0.0645, "step": 26080 }, { "grad_norm": 0.6001245379447937, "learning_rate": 8.832571226488903e-05, "loss": 0.0602, "step": 26090 }, { "grad_norm": 0.5402376055717468, "learning_rate": 8.831509113480634e-05, "loss": 0.0584, "step": 26100 }, { "grad_norm": 0.6885974407196045, "learning_rate": 8.83044658146429e-05, "loss": 0.0581, "step": 26110 }, { "grad_norm": 0.892504096031189, "learning_rate": 8.829383630556067e-05, "loss": 0.0695, "step": 26120 }, { "grad_norm": 0.5700886845588684, "learning_rate": 8.828320260872207e-05, "loss": 0.067, "step": 26130 }, { "grad_norm": 0.8246976137161255, "learning_rate": 8.827256472529e-05, "loss": 0.0938, "step": 26140 }, { "grad_norm": 0.567164421081543, "learning_rate": 8.826192265642778e-05, "loss": 0.0617, "step": 26150 }, { "grad_norm": 0.7868924140930176, "learning_rate": 8.825127640329923e-05, "loss": 0.059, "step": 26160 }, { "grad_norm": 0.6846358776092529, "learning_rate": 8.824062596706861e-05, "loss": 0.0464, "step": 26170 }, { "grad_norm": 0.6757090091705322, "learning_rate": 8.822997134890062e-05, "loss": 0.0658, "step": 26180 }, { "grad_norm": 0.574839174747467, "learning_rate": 8.821931254996044e-05, "loss": 0.0583, "step": 26190 }, { "grad_norm": 0.7182226181030273, "learning_rate": 8.82086495714137e-05, "loss": 0.061, "step": 26200 }, { "grad_norm": 1.3330241441726685, "learning_rate": 8.81979824144265e-05, "loss": 0.0613, "step": 26210 }, { "grad_norm": 0.5286248922348022, "learning_rate": 8.818731108016536e-05, "loss": 0.0546, "step": 26220 }, { "grad_norm": 0.4265792667865753, "learning_rate": 8.81766355697973e-05, "loss": 0.0591, "step": 26230 }, { "grad_norm": 0.5471775531768799, "learning_rate": 8.816595588448977e-05, "loss": 0.0549, "step": 26240 }, { "grad_norm": 0.7905595302581787, "learning_rate": 8.81552720254107e-05, "loss": 0.0705, "step": 26250 }, { "grad_norm": 0.3923071622848511, "learning_rate": 8.814458399372842e-05, "loss": 0.0728, "step": 26260 }, { "grad_norm": 0.5520627498626709, "learning_rate": 8.813389179061181e-05, "loss": 0.0675, "step": 26270 }, { "grad_norm": 0.8669334650039673, "learning_rate": 8.812319541723012e-05, "loss": 0.0723, "step": 26280 }, { "grad_norm": 0.4798571765422821, "learning_rate": 8.811249487475309e-05, "loss": 0.0551, "step": 26290 }, { "grad_norm": 0.3845236897468567, "learning_rate": 8.810179016435092e-05, "loss": 0.0579, "step": 26300 }, { "grad_norm": 0.5264618992805481, "learning_rate": 8.809108128719428e-05, "loss": 0.0411, "step": 26310 }, { "grad_norm": 0.5430472493171692, "learning_rate": 8.808036824445424e-05, "loss": 0.0545, "step": 26320 }, { "grad_norm": 0.6157332062721252, "learning_rate": 8.806965103730238e-05, "loss": 0.0563, "step": 26330 }, { "grad_norm": 0.7609618306159973, "learning_rate": 8.805892966691074e-05, "loss": 0.043, "step": 26340 }, { "grad_norm": 0.44957056641578674, "learning_rate": 8.804820413445175e-05, "loss": 0.0636, "step": 26350 }, { "grad_norm": 0.4398247003555298, "learning_rate": 8.803747444109837e-05, "loss": 0.0481, "step": 26360 }, { "grad_norm": 1.2254054546356201, "learning_rate": 8.802674058802399e-05, "loss": 0.0655, "step": 26370 }, { "grad_norm": 0.40463054180145264, "learning_rate": 8.801600257640241e-05, "loss": 0.0614, "step": 26380 }, { "grad_norm": 0.3798496425151825, "learning_rate": 8.800526040740795e-05, "loss": 0.0542, "step": 26390 }, { "grad_norm": 0.9082995057106018, "learning_rate": 8.799451408221535e-05, "loss": 0.0784, "step": 26400 }, { "grad_norm": 0.6670491099357605, "learning_rate": 8.798376360199982e-05, "loss": 0.0536, "step": 26410 }, { "grad_norm": 0.5781992077827454, "learning_rate": 8.797300896793701e-05, "loss": 0.057, "step": 26420 }, { "grad_norm": 0.4892534613609314, "learning_rate": 8.796225018120302e-05, "loss": 0.0553, "step": 26430 }, { "grad_norm": 0.6434653401374817, "learning_rate": 8.795148724297444e-05, "loss": 0.0534, "step": 26440 }, { "grad_norm": 0.7671406865119934, "learning_rate": 8.794072015442825e-05, "loss": 0.0698, "step": 26450 }, { "grad_norm": 0.5509716272354126, "learning_rate": 8.792994891674198e-05, "loss": 0.0631, "step": 26460 }, { "grad_norm": 0.855464518070221, "learning_rate": 8.79191735310935e-05, "loss": 0.0772, "step": 26470 }, { "grad_norm": 0.6462153792381287, "learning_rate": 8.790839399866122e-05, "loss": 0.0516, "step": 26480 }, { "grad_norm": 0.6959033012390137, "learning_rate": 8.789761032062397e-05, "loss": 0.0527, "step": 26490 }, { "grad_norm": 0.5652463436126709, "learning_rate": 8.788682249816103e-05, "loss": 0.0585, "step": 26500 }, { "grad_norm": 0.6763012409210205, "learning_rate": 8.787603053245215e-05, "loss": 0.0539, "step": 26510 }, { "grad_norm": 0.7252109050750732, "learning_rate": 8.78652344246775e-05, "loss": 0.0734, "step": 26520 }, { "grad_norm": 0.7253211140632629, "learning_rate": 8.785443417601776e-05, "loss": 0.0505, "step": 26530 }, { "grad_norm": 0.4010055661201477, "learning_rate": 8.784362978765401e-05, "loss": 0.0452, "step": 26540 }, { "grad_norm": 0.9225078821182251, "learning_rate": 8.783282126076779e-05, "loss": 0.0581, "step": 26550 }, { "grad_norm": 0.6259317398071289, "learning_rate": 8.782200859654112e-05, "loss": 0.0546, "step": 26560 }, { "grad_norm": 0.36468008160591125, "learning_rate": 8.781119179615646e-05, "loss": 0.043, "step": 26570 }, { "grad_norm": 0.7469316124916077, "learning_rate": 8.780037086079674e-05, "loss": 0.0615, "step": 26580 }, { "grad_norm": 0.9403184056282043, "learning_rate": 8.778954579164527e-05, "loss": 0.0556, "step": 26590 }, { "grad_norm": 1.4698870182037354, "learning_rate": 8.777871658988588e-05, "loss": 0.0575, "step": 26600 }, { "grad_norm": 0.6140490770339966, "learning_rate": 8.776788325670285e-05, "loss": 0.058, "step": 26610 }, { "grad_norm": 0.7283746004104614, "learning_rate": 8.775704579328089e-05, "loss": 0.0607, "step": 26620 }, { "grad_norm": 0.5904834866523743, "learning_rate": 8.774620420080517e-05, "loss": 0.0708, "step": 26630 }, { "grad_norm": 0.7860304117202759, "learning_rate": 8.773535848046131e-05, "loss": 0.0749, "step": 26640 }, { "grad_norm": 0.8626939058303833, "learning_rate": 8.772450863343538e-05, "loss": 0.0681, "step": 26650 }, { "grad_norm": 0.5159951448440552, "learning_rate": 8.77136546609139e-05, "loss": 0.0561, "step": 26660 }, { "grad_norm": 0.5669646859169006, "learning_rate": 8.770279656408385e-05, "loss": 0.0695, "step": 26670 }, { "grad_norm": 0.8891618251800537, "learning_rate": 8.769193434413265e-05, "loss": 0.0569, "step": 26680 }, { "grad_norm": 0.605242133140564, "learning_rate": 8.76810680022482e-05, "loss": 0.049, "step": 26690 }, { "grad_norm": 0.7151862978935242, "learning_rate": 8.767019753961878e-05, "loss": 0.0617, "step": 26700 }, { "grad_norm": 0.4990446865558624, "learning_rate": 8.765932295743321e-05, "loss": 0.0431, "step": 26710 }, { "grad_norm": 0.6810982823371887, "learning_rate": 8.764844425688068e-05, "loss": 0.0462, "step": 26720 }, { "grad_norm": 0.6240711212158203, "learning_rate": 8.763756143915092e-05, "loss": 0.0597, "step": 26730 }, { "grad_norm": 0.3118511438369751, "learning_rate": 8.7626674505434e-05, "loss": 0.0611, "step": 26740 }, { "grad_norm": 0.8332448601722717, "learning_rate": 8.761578345692053e-05, "loss": 0.0527, "step": 26750 }, { "grad_norm": 0.9550560116767883, "learning_rate": 8.760488829480156e-05, "loss": 0.0551, "step": 26760 }, { "grad_norm": 0.8246656060218811, "learning_rate": 8.759398902026854e-05, "loss": 0.0692, "step": 26770 }, { "grad_norm": 0.7638725638389587, "learning_rate": 8.758308563451339e-05, "loss": 0.0739, "step": 26780 }, { "grad_norm": 0.6771754026412964, "learning_rate": 8.75721781387285e-05, "loss": 0.067, "step": 26790 }, { "grad_norm": 0.40386566519737244, "learning_rate": 8.75612665341067e-05, "loss": 0.0664, "step": 26800 }, { "grad_norm": 0.7850888967514038, "learning_rate": 8.755035082184126e-05, "loss": 0.087, "step": 26810 }, { "grad_norm": 0.6851113438606262, "learning_rate": 8.753943100312592e-05, "loss": 0.0572, "step": 26820 }, { "grad_norm": 0.5028424263000488, "learning_rate": 8.752850707915484e-05, "loss": 0.0718, "step": 26830 }, { "grad_norm": 0.31102254986763, "learning_rate": 8.751757905112264e-05, "loss": 0.0608, "step": 26840 }, { "grad_norm": 0.5613064169883728, "learning_rate": 8.75066469202244e-05, "loss": 0.0556, "step": 26850 }, { "grad_norm": 0.6302586793899536, "learning_rate": 8.749571068765567e-05, "loss": 0.0535, "step": 26860 }, { "grad_norm": 0.48163118958473206, "learning_rate": 8.748477035461238e-05, "loss": 0.0518, "step": 26870 }, { "grad_norm": 0.6079374551773071, "learning_rate": 8.747382592229095e-05, "loss": 0.0536, "step": 26880 }, { "grad_norm": 0.5389466285705566, "learning_rate": 8.746287739188828e-05, "loss": 0.0602, "step": 26890 }, { "grad_norm": 0.7481048703193665, "learning_rate": 8.745192476460165e-05, "loss": 0.0718, "step": 26900 }, { "grad_norm": 0.6326317191123962, "learning_rate": 8.744096804162882e-05, "loss": 0.0777, "step": 26910 }, { "grad_norm": 0.466959685087204, "learning_rate": 8.743000722416804e-05, "loss": 0.053, "step": 26920 }, { "grad_norm": 0.7048660516738892, "learning_rate": 8.741904231341793e-05, "loss": 0.0756, "step": 26930 }, { "grad_norm": 1.1052809953689575, "learning_rate": 8.740807331057762e-05, "loss": 0.0706, "step": 26940 }, { "grad_norm": 0.5730422139167786, "learning_rate": 8.739710021684667e-05, "loss": 0.0601, "step": 26950 }, { "grad_norm": 0.42967063188552856, "learning_rate": 8.738612303342503e-05, "loss": 0.0772, "step": 26960 }, { "grad_norm": 0.5845295190811157, "learning_rate": 8.73751417615132e-05, "loss": 0.0793, "step": 26970 }, { "grad_norm": 0.5730488300323486, "learning_rate": 8.736415640231208e-05, "loss": 0.0599, "step": 26980 }, { "grad_norm": 0.39244335889816284, "learning_rate": 8.735316695702297e-05, "loss": 0.0401, "step": 26990 }, { "grad_norm": 0.5521454215049744, "learning_rate": 8.734217342684769e-05, "loss": 0.0446, "step": 27000 }, { "grad_norm": 0.8684660196304321, "learning_rate": 8.733117581298847e-05, "loss": 0.0633, "step": 27010 }, { "grad_norm": 0.7772694230079651, "learning_rate": 8.732017411664796e-05, "loss": 0.0551, "step": 27020 }, { "grad_norm": 0.6003633141517639, "learning_rate": 8.730916833902936e-05, "loss": 0.0814, "step": 27030 }, { "grad_norm": 0.351854532957077, "learning_rate": 8.729815848133618e-05, "loss": 0.0532, "step": 27040 }, { "grad_norm": 0.5272319912910461, "learning_rate": 8.728714454477247e-05, "loss": 0.059, "step": 27050 }, { "grad_norm": 0.7563838958740234, "learning_rate": 8.727612653054269e-05, "loss": 0.0528, "step": 27060 }, { "grad_norm": 0.6300396919250488, "learning_rate": 8.726510443985176e-05, "loss": 0.0661, "step": 27070 }, { "grad_norm": 0.495805948972702, "learning_rate": 8.725407827390503e-05, "loss": 0.0441, "step": 27080 }, { "grad_norm": 0.5963503122329712, "learning_rate": 8.724304803390833e-05, "loss": 0.0478, "step": 27090 }, { "grad_norm": 0.5231837630271912, "learning_rate": 8.723201372106788e-05, "loss": 0.0491, "step": 27100 }, { "grad_norm": 0.6507219672203064, "learning_rate": 8.722097533659038e-05, "loss": 0.0731, "step": 27110 }, { "grad_norm": 0.48716554045677185, "learning_rate": 8.720993288168299e-05, "loss": 0.0647, "step": 27120 }, { "grad_norm": 0.3885786533355713, "learning_rate": 8.719888635755327e-05, "loss": 0.0553, "step": 27130 }, { "grad_norm": 0.6794624924659729, "learning_rate": 8.718783576540928e-05, "loss": 0.0556, "step": 27140 }, { "grad_norm": 0.5096696615219116, "learning_rate": 8.717678110645948e-05, "loss": 0.0472, "step": 27150 }, { "grad_norm": 0.5858005285263062, "learning_rate": 8.716572238191279e-05, "loss": 0.0535, "step": 27160 }, { "grad_norm": 0.6732805371284485, "learning_rate": 8.715465959297857e-05, "loss": 0.0806, "step": 27170 }, { "grad_norm": 0.6370263695716858, "learning_rate": 8.714359274086665e-05, "loss": 0.0541, "step": 27180 }, { "grad_norm": 0.5658743977546692, "learning_rate": 8.713252182678726e-05, "loss": 0.0928, "step": 27190 }, { "grad_norm": 0.4948467016220093, "learning_rate": 8.712144685195112e-05, "loss": 0.0616, "step": 27200 }, { "grad_norm": 0.49536481499671936, "learning_rate": 8.711036781756936e-05, "loss": 0.0594, "step": 27210 }, { "grad_norm": 0.5411480069160461, "learning_rate": 8.709928472485357e-05, "loss": 0.0537, "step": 27220 }, { "grad_norm": 0.5927401185035706, "learning_rate": 8.708819757501579e-05, "loss": 0.0544, "step": 27230 }, { "grad_norm": 0.5603664517402649, "learning_rate": 8.707710636926846e-05, "loss": 0.0853, "step": 27240 }, { "grad_norm": 0.6723938584327698, "learning_rate": 8.706601110882455e-05, "loss": 0.0606, "step": 27250 }, { "grad_norm": 0.6118679642677307, "learning_rate": 8.705491179489738e-05, "loss": 0.0625, "step": 27260 }, { "grad_norm": 0.4806278347969055, "learning_rate": 8.704380842870077e-05, "loss": 0.0569, "step": 27270 }, { "grad_norm": 0.5778734087944031, "learning_rate": 8.703270101144895e-05, "loss": 0.065, "step": 27280 }, { "grad_norm": 0.4961611330509186, "learning_rate": 8.702158954435664e-05, "loss": 0.0508, "step": 27290 }, { "grad_norm": 0.40023714303970337, "learning_rate": 8.701047402863896e-05, "loss": 0.0711, "step": 27300 }, { "grad_norm": 0.5135348439216614, "learning_rate": 8.699935446551148e-05, "loss": 0.0574, "step": 27310 }, { "grad_norm": 0.5986130237579346, "learning_rate": 8.698823085619022e-05, "loss": 0.0618, "step": 27320 }, { "grad_norm": 0.559266209602356, "learning_rate": 8.697710320189166e-05, "loss": 0.0623, "step": 27330 }, { "grad_norm": 0.5865124464035034, "learning_rate": 8.696597150383268e-05, "loss": 0.0572, "step": 27340 }, { "grad_norm": 0.5268883109092712, "learning_rate": 8.695483576323063e-05, "loss": 0.0562, "step": 27350 }, { "grad_norm": 0.46771240234375, "learning_rate": 8.69436959813033e-05, "loss": 0.0745, "step": 27360 }, { "grad_norm": 0.5321170687675476, "learning_rate": 8.693255215926892e-05, "loss": 0.0523, "step": 27370 }, { "grad_norm": 0.8281797170639038, "learning_rate": 8.692140429834617e-05, "loss": 0.064, "step": 27380 }, { "grad_norm": 0.7117178440093994, "learning_rate": 8.691025239975415e-05, "loss": 0.068, "step": 27390 }, { "grad_norm": 0.7056095600128174, "learning_rate": 8.689909646471243e-05, "loss": 0.0563, "step": 27400 }, { "grad_norm": 0.5577089190483093, "learning_rate": 8.688793649444099e-05, "loss": 0.0679, "step": 27410 }, { "grad_norm": 0.5780560970306396, "learning_rate": 8.687677249016029e-05, "loss": 0.0612, "step": 27420 }, { "grad_norm": 0.5524622797966003, "learning_rate": 8.686560445309118e-05, "loss": 0.0499, "step": 27430 }, { "grad_norm": 0.5701315999031067, "learning_rate": 8.685443238445499e-05, "loss": 0.0509, "step": 27440 }, { "grad_norm": 0.4910880923271179, "learning_rate": 8.68432562854735e-05, "loss": 0.051, "step": 27450 }, { "grad_norm": 0.5946257710456848, "learning_rate": 8.683207615736887e-05, "loss": 0.0459, "step": 27460 }, { "grad_norm": 0.5899688601493835, "learning_rate": 8.682089200136379e-05, "loss": 0.0417, "step": 27470 }, { "grad_norm": 0.49219557642936707, "learning_rate": 8.680970381868132e-05, "loss": 0.0591, "step": 27480 }, { "grad_norm": 0.6193950176239014, "learning_rate": 8.679851161054498e-05, "loss": 0.0601, "step": 27490 }, { "grad_norm": 0.7443172335624695, "learning_rate": 8.678731537817873e-05, "loss": 0.0523, "step": 27500 }, { "grad_norm": 0.4939221143722534, "learning_rate": 8.677611512280697e-05, "loss": 0.0623, "step": 27510 }, { "grad_norm": 0.8887459635734558, "learning_rate": 8.676491084565457e-05, "loss": 0.1009, "step": 27520 }, { "grad_norm": 0.4355841279029846, "learning_rate": 8.675370254794678e-05, "loss": 0.0558, "step": 27530 }, { "grad_norm": 0.5479815006256104, "learning_rate": 8.674249023090935e-05, "loss": 0.0584, "step": 27540 }, { "grad_norm": 1.1857203245162964, "learning_rate": 8.673127389576843e-05, "loss": 0.0633, "step": 27550 }, { "grad_norm": 0.6348561644554138, "learning_rate": 8.67200535437506e-05, "loss": 0.0555, "step": 27560 }, { "grad_norm": 0.5068132877349854, "learning_rate": 8.670882917608296e-05, "loss": 0.0615, "step": 27570 }, { "grad_norm": 0.5589273571968079, "learning_rate": 8.669760079399292e-05, "loss": 0.0682, "step": 27580 }, { "grad_norm": 0.464189738035202, "learning_rate": 8.668636839870845e-05, "loss": 0.0489, "step": 27590 }, { "grad_norm": 0.5929135680198669, "learning_rate": 8.667513199145789e-05, "loss": 0.0554, "step": 27600 }, { "grad_norm": 0.5485330820083618, "learning_rate": 8.666389157347002e-05, "loss": 0.069, "step": 27610 }, { "grad_norm": 0.5560389161109924, "learning_rate": 8.66526471459741e-05, "loss": 0.0523, "step": 27620 }, { "grad_norm": 0.5867146253585815, "learning_rate": 8.66413987101998e-05, "loss": 0.0609, "step": 27630 }, { "grad_norm": 0.6239184737205505, "learning_rate": 8.663014626737723e-05, "loss": 0.0666, "step": 27640 }, { "grad_norm": 0.5159248113632202, "learning_rate": 8.661888981873691e-05, "loss": 0.0569, "step": 27650 }, { "grad_norm": 0.8079900741577148, "learning_rate": 8.660762936550988e-05, "loss": 0.0696, "step": 27660 }, { "grad_norm": 0.6634500622749329, "learning_rate": 8.659636490892753e-05, "loss": 0.0579, "step": 27670 }, { "grad_norm": 0.6482571959495544, "learning_rate": 8.658509645022174e-05, "loss": 0.0611, "step": 27680 }, { "grad_norm": 0.6843475103378296, "learning_rate": 8.657382399062481e-05, "loss": 0.0501, "step": 27690 }, { "grad_norm": 0.5441847443580627, "learning_rate": 8.656254753136946e-05, "loss": 0.0621, "step": 27700 }, { "grad_norm": 0.6639117002487183, "learning_rate": 8.655126707368891e-05, "loss": 0.0585, "step": 27710 }, { "grad_norm": 0.3018341660499573, "learning_rate": 8.653998261881672e-05, "loss": 0.0438, "step": 27720 }, { "grad_norm": 0.6039237976074219, "learning_rate": 8.652869416798699e-05, "loss": 0.0647, "step": 27730 }, { "grad_norm": 0.9830156564712524, "learning_rate": 8.651740172243417e-05, "loss": 0.0544, "step": 27740 }, { "grad_norm": 0.7637002468109131, "learning_rate": 8.65061052833932e-05, "loss": 0.0454, "step": 27750 }, { "grad_norm": 0.41839873790740967, "learning_rate": 8.649480485209945e-05, "loss": 0.0514, "step": 27760 }, { "grad_norm": 0.5580145716667175, "learning_rate": 8.64835004297887e-05, "loss": 0.0636, "step": 27770 }, { "grad_norm": 0.8294304013252258, "learning_rate": 8.64721920176972e-05, "loss": 0.0824, "step": 27780 }, { "grad_norm": 0.5831010341644287, "learning_rate": 8.646087961706164e-05, "loss": 0.0446, "step": 27790 }, { "grad_norm": 0.4415002167224884, "learning_rate": 8.644956322911908e-05, "loss": 0.0569, "step": 27800 }, { "grad_norm": 0.504245400428772, "learning_rate": 8.643824285510709e-05, "loss": 0.0676, "step": 27810 }, { "grad_norm": 0.42963695526123047, "learning_rate": 8.642691849626364e-05, "loss": 0.038, "step": 27820 }, { "grad_norm": 0.6025401949882507, "learning_rate": 8.641559015382717e-05, "loss": 0.0743, "step": 27830 }, { "grad_norm": 0.6004102230072021, "learning_rate": 8.640425782903649e-05, "loss": 0.0567, "step": 27840 }, { "grad_norm": 0.5849819779396057, "learning_rate": 8.639292152313091e-05, "loss": 0.0477, "step": 27850 }, { "grad_norm": 0.5308045744895935, "learning_rate": 8.638158123735015e-05, "loss": 0.0664, "step": 27860 }, { "grad_norm": 0.7524694204330444, "learning_rate": 8.637023697293436e-05, "loss": 0.0642, "step": 27870 }, { "grad_norm": 0.6085582971572876, "learning_rate": 8.635888873112414e-05, "loss": 0.0536, "step": 27880 }, { "grad_norm": 0.4988381564617157, "learning_rate": 8.634753651316052e-05, "loss": 0.0556, "step": 27890 }, { "grad_norm": 0.543735146522522, "learning_rate": 8.633618032028496e-05, "loss": 0.0725, "step": 27900 }, { "grad_norm": 0.5141010880470276, "learning_rate": 8.632482015373934e-05, "loss": 0.0631, "step": 27910 }, { "grad_norm": 0.5190088748931885, "learning_rate": 8.6313456014766e-05, "loss": 0.055, "step": 27920 }, { "grad_norm": 0.5167139768600464, "learning_rate": 8.630208790460771e-05, "loss": 0.0829, "step": 27930 }, { "grad_norm": 0.43280133605003357, "learning_rate": 8.629071582450768e-05, "loss": 0.0816, "step": 27940 }, { "grad_norm": 0.9736926555633545, "learning_rate": 8.62793397757095e-05, "loss": 0.0582, "step": 27950 }, { "grad_norm": 0.45159876346588135, "learning_rate": 8.626795975945729e-05, "loss": 0.0841, "step": 27960 }, { "grad_norm": 0.706610381603241, "learning_rate": 8.625657577699551e-05, "loss": 0.0715, "step": 27970 }, { "grad_norm": 0.5385169982910156, "learning_rate": 8.624518782956914e-05, "loss": 0.0555, "step": 27980 }, { "grad_norm": 0.46359783411026, "learning_rate": 8.62337959184235e-05, "loss": 0.057, "step": 27990 }, { "grad_norm": 0.5080866813659668, "learning_rate": 8.622240004480441e-05, "loss": 0.0543, "step": 28000 }, { "grad_norm": 0.39944902062416077, "learning_rate": 8.621100020995814e-05, "loss": 0.0727, "step": 28010 }, { "grad_norm": 0.46233993768692017, "learning_rate": 8.619959641513132e-05, "loss": 0.0572, "step": 28020 }, { "grad_norm": 0.6117286086082458, "learning_rate": 8.618818866157105e-05, "loss": 0.0815, "step": 28030 }, { "grad_norm": 0.7534019351005554, "learning_rate": 8.617677695052487e-05, "loss": 0.0686, "step": 28040 }, { "grad_norm": 0.586935818195343, "learning_rate": 8.616536128324078e-05, "loss": 0.0572, "step": 28050 }, { "grad_norm": 0.529074490070343, "learning_rate": 8.615394166096712e-05, "loss": 0.0743, "step": 28060 }, { "grad_norm": 0.8372851610183716, "learning_rate": 8.614251808495279e-05, "loss": 0.0558, "step": 28070 }, { "grad_norm": 0.7758941650390625, "learning_rate": 8.6131090556447e-05, "loss": 0.0533, "step": 28080 }, { "grad_norm": 0.8557515740394592, "learning_rate": 8.611965907669947e-05, "loss": 0.0616, "step": 28090 }, { "grad_norm": 0.6541391611099243, "learning_rate": 8.610822364696034e-05, "loss": 0.0524, "step": 28100 }, { "grad_norm": 0.7302231192588806, "learning_rate": 8.609678426848015e-05, "loss": 0.0589, "step": 28110 }, { "grad_norm": 0.5607210993766785, "learning_rate": 8.60853409425099e-05, "loss": 0.0483, "step": 28120 }, { "grad_norm": 0.5223907232284546, "learning_rate": 8.607389367030104e-05, "loss": 0.0526, "step": 28130 }, { "grad_norm": 0.7000944018363953, "learning_rate": 8.606244245310538e-05, "loss": 0.0535, "step": 28140 }, { "grad_norm": 0.9036111831665039, "learning_rate": 8.605098729217525e-05, "loss": 0.0682, "step": 28150 }, { "grad_norm": 0.36927857995033264, "learning_rate": 8.603952818876335e-05, "loss": 0.0502, "step": 28160 }, { "grad_norm": 0.8298384547233582, "learning_rate": 8.602806514412281e-05, "loss": 0.0627, "step": 28170 }, { "grad_norm": 0.4664284884929657, "learning_rate": 8.601659815950726e-05, "loss": 0.0433, "step": 28180 }, { "grad_norm": 0.8475841283798218, "learning_rate": 8.600512723617067e-05, "loss": 0.071, "step": 28190 }, { "grad_norm": 0.39555084705352783, "learning_rate": 8.59936523753675e-05, "loss": 0.0526, "step": 28200 }, { "grad_norm": 0.5015844106674194, "learning_rate": 8.598217357835264e-05, "loss": 0.0506, "step": 28210 }, { "grad_norm": 0.606675922870636, "learning_rate": 8.597069084638135e-05, "loss": 0.0682, "step": 28220 }, { "grad_norm": 0.3202034533023834, "learning_rate": 8.595920418070939e-05, "loss": 0.0454, "step": 28230 }, { "grad_norm": 0.5324503779411316, "learning_rate": 8.594771358259295e-05, "loss": 0.0679, "step": 28240 }, { "grad_norm": 0.49151167273521423, "learning_rate": 8.593621905328858e-05, "loss": 0.0418, "step": 28250 }, { "grad_norm": 0.6395747661590576, "learning_rate": 8.592472059405333e-05, "loss": 0.0807, "step": 28260 }, { "grad_norm": 0.697195827960968, "learning_rate": 8.591321820614464e-05, "loss": 0.0735, "step": 28270 }, { "grad_norm": 0.6800078749656677, "learning_rate": 8.590171189082041e-05, "loss": 0.0496, "step": 28280 }, { "grad_norm": 0.2748216986656189, "learning_rate": 8.589020164933894e-05, "loss": 0.0476, "step": 28290 }, { "grad_norm": 0.7202867269515991, "learning_rate": 8.587868748295898e-05, "loss": 0.0655, "step": 28300 }, { "grad_norm": 1.0810827016830444, "learning_rate": 8.586716939293971e-05, "loss": 0.0973, "step": 28310 }, { "grad_norm": 0.911271333694458, "learning_rate": 8.58556473805407e-05, "loss": 0.0825, "step": 28320 }, { "grad_norm": 0.5448923707008362, "learning_rate": 8.584412144702202e-05, "loss": 0.0612, "step": 28330 }, { "grad_norm": 0.8301641941070557, "learning_rate": 8.58325915936441e-05, "loss": 0.0768, "step": 28340 }, { "grad_norm": 0.6246988773345947, "learning_rate": 8.582105782166783e-05, "loss": 0.0601, "step": 28350 }, { "grad_norm": 0.48858577013015747, "learning_rate": 8.580952013235455e-05, "loss": 0.0614, "step": 28360 }, { "grad_norm": 0.5982487201690674, "learning_rate": 8.579797852696596e-05, "loss": 0.0655, "step": 28370 }, { "grad_norm": 0.5854908227920532, "learning_rate": 8.578643300676428e-05, "loss": 0.0548, "step": 28380 }, { "grad_norm": 0.6464110016822815, "learning_rate": 8.577488357301209e-05, "loss": 0.062, "step": 28390 }, { "grad_norm": 0.8202504515647888, "learning_rate": 8.576333022697242e-05, "loss": 0.0508, "step": 28400 }, { "grad_norm": 0.5193952322006226, "learning_rate": 8.575177296990873e-05, "loss": 0.0511, "step": 28410 }, { "grad_norm": 0.515953540802002, "learning_rate": 8.574021180308489e-05, "loss": 0.053, "step": 28420 }, { "grad_norm": 0.3863006830215454, "learning_rate": 8.572864672776523e-05, "loss": 0.0765, "step": 28430 }, { "grad_norm": 0.4164190888404846, "learning_rate": 8.571707774521447e-05, "loss": 0.0475, "step": 28440 }, { "grad_norm": 0.5355556607246399, "learning_rate": 8.57055048566978e-05, "loss": 0.0562, "step": 28450 }, { "grad_norm": 0.6347382068634033, "learning_rate": 8.569392806348078e-05, "loss": 0.0568, "step": 28460 }, { "grad_norm": 0.8519576191902161, "learning_rate": 8.568234736682947e-05, "loss": 0.0603, "step": 28470 }, { "grad_norm": 0.6933019161224365, "learning_rate": 8.567076276801029e-05, "loss": 0.0592, "step": 28480 }, { "grad_norm": 0.42765167355537415, "learning_rate": 8.565917426829013e-05, "loss": 0.0811, "step": 28490 }, { "grad_norm": 0.5001162886619568, "learning_rate": 8.564758186893628e-05, "loss": 0.0559, "step": 28500 }, { "grad_norm": 0.57770174741745, "learning_rate": 8.563598557121649e-05, "loss": 0.0605, "step": 28510 }, { "grad_norm": 0.5894272327423096, "learning_rate": 8.562438537639888e-05, "loss": 0.0561, "step": 28520 }, { "grad_norm": 0.5082302689552307, "learning_rate": 8.561278128575206e-05, "loss": 0.0805, "step": 28530 }, { "grad_norm": 0.6702529191970825, "learning_rate": 8.5601173300545e-05, "loss": 0.0706, "step": 28540 }, { "grad_norm": 0.5319708585739136, "learning_rate": 8.558956142204717e-05, "loss": 0.0501, "step": 28550 }, { "grad_norm": 0.9045563340187073, "learning_rate": 8.55779456515284e-05, "loss": 0.0585, "step": 28560 }, { "grad_norm": 0.6504054665565491, "learning_rate": 8.556632599025898e-05, "loss": 0.0499, "step": 28570 }, { "grad_norm": 0.40171942114830017, "learning_rate": 8.555470243950964e-05, "loss": 0.0492, "step": 28580 }, { "grad_norm": 0.5375142097473145, "learning_rate": 8.554307500055148e-05, "loss": 0.055, "step": 28590 }, { "grad_norm": 0.5401287078857422, "learning_rate": 8.553144367465609e-05, "loss": 0.0842, "step": 28600 }, { "grad_norm": 0.5273222327232361, "learning_rate": 8.551980846309544e-05, "loss": 0.0523, "step": 28610 }, { "grad_norm": 0.6392686367034912, "learning_rate": 8.550816936714193e-05, "loss": 0.0464, "step": 28620 }, { "grad_norm": 0.43525782227516174, "learning_rate": 8.549652638806841e-05, "loss": 0.0766, "step": 28630 }, { "grad_norm": 0.34824511408805847, "learning_rate": 8.548487952714812e-05, "loss": 0.0538, "step": 28640 }, { "grad_norm": 0.4614686667919159, "learning_rate": 8.547322878565478e-05, "loss": 0.0604, "step": 28650 }, { "grad_norm": 0.6155575513839722, "learning_rate": 8.546157416486245e-05, "loss": 0.0454, "step": 28660 }, { "grad_norm": 0.6488193869590759, "learning_rate": 8.54499156660457e-05, "loss": 0.0523, "step": 28670 }, { "grad_norm": 0.6008784770965576, "learning_rate": 8.543825329047947e-05, "loss": 0.0506, "step": 28680 }, { "grad_norm": 0.7475461363792419, "learning_rate": 8.542658703943913e-05, "loss": 0.0565, "step": 28690 }, { "grad_norm": 0.7801908254623413, "learning_rate": 8.541491691420051e-05, "loss": 0.0518, "step": 28700 }, { "grad_norm": 0.5711467266082764, "learning_rate": 8.54032429160398e-05, "loss": 0.0514, "step": 28710 }, { "grad_norm": 0.6394297480583191, "learning_rate": 8.539156504623369e-05, "loss": 0.0527, "step": 28720 }, { "grad_norm": 0.5958387851715088, "learning_rate": 8.537988330605923e-05, "loss": 0.0579, "step": 28730 }, { "grad_norm": 0.529841423034668, "learning_rate": 8.536819769679393e-05, "loss": 0.0587, "step": 28740 }, { "grad_norm": 0.3970388174057007, "learning_rate": 8.53565082197157e-05, "loss": 0.0541, "step": 28750 }, { "grad_norm": 0.36893072724342346, "learning_rate": 8.534481487610289e-05, "loss": 0.0625, "step": 28760 }, { "grad_norm": 0.60494464635849, "learning_rate": 8.533311766723428e-05, "loss": 0.0539, "step": 28770 }, { "grad_norm": 0.7734112739562988, "learning_rate": 8.532141659438901e-05, "loss": 0.0901, "step": 28780 }, { "grad_norm": 1.2206306457519531, "learning_rate": 8.530971165884675e-05, "loss": 0.0643, "step": 28790 }, { "grad_norm": 0.5458928346633911, "learning_rate": 8.529800286188752e-05, "loss": 0.0591, "step": 28800 }, { "grad_norm": 0.713875949382782, "learning_rate": 8.528629020479175e-05, "loss": 0.0563, "step": 28810 }, { "grad_norm": 0.5120625495910645, "learning_rate": 8.527457368884033e-05, "loss": 0.0617, "step": 28820 }, { "grad_norm": 0.4698549509048462, "learning_rate": 8.526285331531458e-05, "loss": 0.0534, "step": 28830 }, { "grad_norm": 0.6671509146690369, "learning_rate": 8.525112908549621e-05, "loss": 0.0495, "step": 28840 }, { "grad_norm": 0.4837821125984192, "learning_rate": 8.523940100066735e-05, "loss": 0.0499, "step": 28850 }, { "grad_norm": 0.6541001796722412, "learning_rate": 8.52276690621106e-05, "loss": 0.0522, "step": 28860 }, { "grad_norm": 0.6600313186645508, "learning_rate": 8.521593327110889e-05, "loss": 0.0594, "step": 28870 }, { "grad_norm": 0.5282735824584961, "learning_rate": 8.520419362894569e-05, "loss": 0.0693, "step": 28880 }, { "grad_norm": 0.5456072092056274, "learning_rate": 8.51924501369048e-05, "loss": 0.0621, "step": 28890 }, { "grad_norm": 0.5123204588890076, "learning_rate": 8.518070279627047e-05, "loss": 0.052, "step": 28900 }, { "grad_norm": 0.5668473839759827, "learning_rate": 8.516895160832737e-05, "loss": 0.069, "step": 28910 }, { "grad_norm": 0.629263162612915, "learning_rate": 8.515719657436061e-05, "loss": 0.056, "step": 28920 }, { "grad_norm": 0.7928341031074524, "learning_rate": 8.514543769565568e-05, "loss": 0.055, "step": 28930 }, { "grad_norm": 0.6981499195098877, "learning_rate": 8.513367497349853e-05, "loss": 0.0852, "step": 28940 }, { "grad_norm": 0.5184730887413025, "learning_rate": 8.51219084091755e-05, "loss": 0.0619, "step": 28950 }, { "grad_norm": 0.5684282779693604, "learning_rate": 8.511013800397338e-05, "loss": 0.0586, "step": 28960 }, { "grad_norm": 0.5579249262809753, "learning_rate": 8.509836375917937e-05, "loss": 0.0533, "step": 28970 }, { "grad_norm": 0.5960132479667664, "learning_rate": 8.508658567608104e-05, "loss": 0.0641, "step": 28980 }, { "grad_norm": 0.7255255579948425, "learning_rate": 8.507480375596647e-05, "loss": 0.065, "step": 28990 }, { "grad_norm": 0.43645110726356506, "learning_rate": 8.506301800012408e-05, "loss": 0.0532, "step": 29000 }, { "grad_norm": 0.8552150130271912, "learning_rate": 8.505122840984278e-05, "loss": 0.0556, "step": 29010 }, { "grad_norm": 0.5060554146766663, "learning_rate": 8.503943498641182e-05, "loss": 0.0667, "step": 29020 }, { "grad_norm": 0.9336883425712585, "learning_rate": 8.502763773112095e-05, "loss": 0.055, "step": 29030 }, { "grad_norm": 0.5500101447105408, "learning_rate": 8.501583664526026e-05, "loss": 0.0506, "step": 29040 }, { "grad_norm": 0.5825344920158386, "learning_rate": 8.500403173012032e-05, "loss": 0.1077, "step": 29050 }, { "grad_norm": 0.5641682744026184, "learning_rate": 8.499222298699211e-05, "loss": 0.0531, "step": 29060 }, { "grad_norm": 0.43981796503067017, "learning_rate": 8.498041041716701e-05, "loss": 0.0589, "step": 29070 }, { "grad_norm": 0.48960310220718384, "learning_rate": 8.496859402193681e-05, "loss": 0.0465, "step": 29080 }, { "grad_norm": 0.4834069609642029, "learning_rate": 8.495677380259374e-05, "loss": 0.0558, "step": 29090 }, { "grad_norm": 0.575516402721405, "learning_rate": 8.494494976043045e-05, "loss": 0.0465, "step": 29100 }, { "grad_norm": 0.6245707869529724, "learning_rate": 8.493312189673998e-05, "loss": 0.0683, "step": 29110 }, { "grad_norm": 0.5037709474563599, "learning_rate": 8.492129021281584e-05, "loss": 0.0563, "step": 29120 }, { "grad_norm": 0.5604478716850281, "learning_rate": 8.490945470995188e-05, "loss": 0.0443, "step": 29130 }, { "grad_norm": 0.44342857599258423, "learning_rate": 8.489761538944247e-05, "loss": 0.0581, "step": 29140 }, { "grad_norm": 0.5047421455383301, "learning_rate": 8.48857722525823e-05, "loss": 0.0717, "step": 29150 }, { "grad_norm": 0.8477599620819092, "learning_rate": 8.487392530066652e-05, "loss": 0.068, "step": 29160 }, { "grad_norm": 0.4327845275402069, "learning_rate": 8.486207453499069e-05, "loss": 0.067, "step": 29170 }, { "grad_norm": 0.7533409595489502, "learning_rate": 8.485021995685082e-05, "loss": 0.0678, "step": 29180 }, { "grad_norm": 0.7082216739654541, "learning_rate": 8.483836156754328e-05, "loss": 0.0527, "step": 29190 }, { "grad_norm": 0.4540119171142578, "learning_rate": 8.482649936836491e-05, "loss": 0.0587, "step": 29200 }, { "grad_norm": 0.5525249242782593, "learning_rate": 8.481463336061293e-05, "loss": 0.058, "step": 29210 }, { "grad_norm": 0.6985135078430176, "learning_rate": 8.480276354558496e-05, "loss": 0.0599, "step": 29220 }, { "grad_norm": 0.7195479869842529, "learning_rate": 8.479088992457913e-05, "loss": 0.0486, "step": 29230 }, { "grad_norm": 0.5738096833229065, "learning_rate": 8.477901249889387e-05, "loss": 0.049, "step": 29240 }, { "grad_norm": 0.6562507152557373, "learning_rate": 8.47671312698281e-05, "loss": 0.0499, "step": 29250 }, { "grad_norm": 0.762610673904419, "learning_rate": 8.475524623868112e-05, "loss": 0.07, "step": 29260 }, { "grad_norm": 0.5055481195449829, "learning_rate": 8.474335740675266e-05, "loss": 0.0506, "step": 29270 }, { "grad_norm": 0.7030470371246338, "learning_rate": 8.473146477534289e-05, "loss": 0.0559, "step": 29280 }, { "grad_norm": 0.9302915930747986, "learning_rate": 8.471956834575232e-05, "loss": 0.0483, "step": 29290 }, { "grad_norm": 0.6413247585296631, "learning_rate": 8.470766811928197e-05, "loss": 0.0639, "step": 29300 }, { "grad_norm": 0.5007402300834656, "learning_rate": 8.469576409723323e-05, "loss": 0.0547, "step": 29310 }, { "grad_norm": 0.7995347380638123, "learning_rate": 8.468385628090788e-05, "loss": 0.061, "step": 29320 }, { "grad_norm": 0.5318570733070374, "learning_rate": 8.467194467160815e-05, "loss": 0.0713, "step": 29330 }, { "grad_norm": 0.4951888620853424, "learning_rate": 8.466002927063667e-05, "loss": 0.0662, "step": 29340 }, { "grad_norm": 0.4771537184715271, "learning_rate": 8.464811007929651e-05, "loss": 0.0466, "step": 29350 }, { "grad_norm": 0.5165125131607056, "learning_rate": 8.463618709889114e-05, "loss": 0.0596, "step": 29360 }, { "grad_norm": 0.6152545213699341, "learning_rate": 8.462426033072442e-05, "loss": 0.056, "step": 29370 }, { "grad_norm": 0.5850260853767395, "learning_rate": 8.461232977610061e-05, "loss": 0.0684, "step": 29380 }, { "grad_norm": 0.45163124799728394, "learning_rate": 8.46003954363245e-05, "loss": 0.0484, "step": 29390 }, { "grad_norm": 0.33560171723365784, "learning_rate": 8.458845731270115e-05, "loss": 0.0527, "step": 29400 }, { "grad_norm": 0.5364236831665039, "learning_rate": 8.45765154065361e-05, "loss": 0.0463, "step": 29410 }, { "grad_norm": 0.7258803844451904, "learning_rate": 8.456456971913532e-05, "loss": 0.0848, "step": 29420 }, { "grad_norm": 1.5809885263442993, "learning_rate": 8.455262025180517e-05, "loss": 0.0601, "step": 29430 }, { "grad_norm": 0.368939608335495, "learning_rate": 8.454066700585242e-05, "loss": 0.0533, "step": 29440 }, { "grad_norm": 0.4706217348575592, "learning_rate": 8.452870998258423e-05, "loss": 0.0605, "step": 29450 }, { "grad_norm": 0.6209795475006104, "learning_rate": 8.451674918330825e-05, "loss": 0.0667, "step": 29460 }, { "grad_norm": 0.6341423988342285, "learning_rate": 8.450478460933246e-05, "loss": 0.0736, "step": 29470 }, { "grad_norm": 0.5205168128013611, "learning_rate": 8.449281626196532e-05, "loss": 0.0568, "step": 29480 }, { "grad_norm": 0.6524850726127625, "learning_rate": 8.448084414251564e-05, "loss": 0.0478, "step": 29490 }, { "grad_norm": 0.6124391555786133, "learning_rate": 8.446886825229271e-05, "loss": 0.0618, "step": 29500 }, { "grad_norm": 0.41590917110443115, "learning_rate": 8.445688859260615e-05, "loss": 0.0564, "step": 29510 }, { "grad_norm": 0.47874918580055237, "learning_rate": 8.444490516476606e-05, "loss": 0.0645, "step": 29520 }, { "grad_norm": 0.7149626016616821, "learning_rate": 8.443291797008293e-05, "loss": 0.0625, "step": 29530 }, { "grad_norm": 0.5259549617767334, "learning_rate": 8.442092700986765e-05, "loss": 0.0486, "step": 29540 }, { "grad_norm": 0.5096343755722046, "learning_rate": 8.440893228543156e-05, "loss": 0.0516, "step": 29550 }, { "grad_norm": 0.594550371170044, "learning_rate": 8.439693379808638e-05, "loss": 0.0664, "step": 29560 }, { "grad_norm": 0.45543360710144043, "learning_rate": 8.43849315491442e-05, "loss": 0.0424, "step": 29570 }, { "grad_norm": 0.7031232118606567, "learning_rate": 8.437292553991763e-05, "loss": 0.0466, "step": 29580 }, { "grad_norm": 0.5608804225921631, "learning_rate": 8.436091577171959e-05, "loss": 0.0481, "step": 29590 }, { "grad_norm": 0.8568444848060608, "learning_rate": 8.434890224586347e-05, "loss": 0.0649, "step": 29600 }, { "grad_norm": 0.6611362099647522, "learning_rate": 8.433688496366303e-05, "loss": 0.0594, "step": 29610 }, { "grad_norm": 0.48393023014068604, "learning_rate": 8.432486392643248e-05, "loss": 0.0526, "step": 29620 }, { "grad_norm": 0.447365403175354, "learning_rate": 8.431283913548643e-05, "loss": 0.0753, "step": 29630 }, { "grad_norm": 0.46912673115730286, "learning_rate": 8.430081059213985e-05, "loss": 0.0356, "step": 29640 }, { "grad_norm": 0.45662349462509155, "learning_rate": 8.428877829770823e-05, "loss": 0.085, "step": 29650 }, { "grad_norm": 0.6590185165405273, "learning_rate": 8.427674225350735e-05, "loss": 0.0392, "step": 29660 }, { "grad_norm": 0.5507242679595947, "learning_rate": 8.426470246085347e-05, "loss": 0.0972, "step": 29670 }, { "grad_norm": 0.38552698493003845, "learning_rate": 8.425265892106324e-05, "loss": 0.0445, "step": 29680 }, { "grad_norm": 0.5400587916374207, "learning_rate": 8.424061163545374e-05, "loss": 0.0465, "step": 29690 }, { "grad_norm": 0.4532459080219269, "learning_rate": 8.422856060534243e-05, "loss": 0.0949, "step": 29700 }, { "grad_norm": 0.6098496913909912, "learning_rate": 8.421650583204718e-05, "loss": 0.048, "step": 29710 }, { "grad_norm": 8.299079895019531, "learning_rate": 8.420444731688633e-05, "loss": 0.0711, "step": 29720 }, { "grad_norm": 0.6118634939193726, "learning_rate": 8.419238506117852e-05, "loss": 0.047, "step": 29730 }, { "grad_norm": 0.46592801809310913, "learning_rate": 8.418031906624289e-05, "loss": 0.0573, "step": 29740 }, { "grad_norm": 0.46653491258621216, "learning_rate": 8.416824933339898e-05, "loss": 0.0543, "step": 29750 }, { "grad_norm": 0.4883164167404175, "learning_rate": 8.415617586396667e-05, "loss": 0.0518, "step": 29760 }, { "grad_norm": 0.48428744077682495, "learning_rate": 8.414409865926632e-05, "loss": 0.0495, "step": 29770 }, { "grad_norm": 0.6170511841773987, "learning_rate": 8.413201772061867e-05, "loss": 0.0512, "step": 29780 }, { "grad_norm": 2.5517842769622803, "learning_rate": 8.411993304934488e-05, "loss": 0.0613, "step": 29790 }, { "grad_norm": 0.5958227515220642, "learning_rate": 8.410784464676654e-05, "loss": 0.04, "step": 29800 }, { "grad_norm": 0.46351611614227295, "learning_rate": 8.409575251420556e-05, "loss": 0.0659, "step": 29810 }, { "grad_norm": 1.1603444814682007, "learning_rate": 8.408365665298435e-05, "loss": 0.0619, "step": 29820 }, { "grad_norm": 0.9262927770614624, "learning_rate": 8.40715570644257e-05, "loss": 0.0548, "step": 29830 }, { "grad_norm": 0.8301175832748413, "learning_rate": 8.40594537498528e-05, "loss": 0.0591, "step": 29840 }, { "grad_norm": 0.3867257535457611, "learning_rate": 8.404734671058924e-05, "loss": 0.0548, "step": 29850 }, { "grad_norm": 0.7504797577857971, "learning_rate": 8.403523594795902e-05, "loss": 0.0415, "step": 29860 }, { "grad_norm": 0.9997333884239197, "learning_rate": 8.402312146328659e-05, "loss": 0.0616, "step": 29870 }, { "grad_norm": 0.576662540435791, "learning_rate": 8.401100325789675e-05, "loss": 0.0576, "step": 29880 }, { "grad_norm": 0.6676989197731018, "learning_rate": 8.399888133311472e-05, "loss": 0.0424, "step": 29890 }, { "grad_norm": 0.7006317973136902, "learning_rate": 8.398675569026613e-05, "loss": 0.0935, "step": 29900 }, { "grad_norm": 0.5106142163276672, "learning_rate": 8.397462633067705e-05, "loss": 0.0479, "step": 29910 }, { "grad_norm": 0.5501518845558167, "learning_rate": 8.396249325567392e-05, "loss": 0.0753, "step": 29920 }, { "grad_norm": 0.7981583476066589, "learning_rate": 8.395035646658357e-05, "loss": 0.0391, "step": 29930 }, { "grad_norm": 0.4239071309566498, "learning_rate": 8.39382159647333e-05, "loss": 0.0606, "step": 29940 }, { "grad_norm": 1.2354085445404053, "learning_rate": 8.392607175145075e-05, "loss": 0.066, "step": 29950 }, { "grad_norm": 0.3798169791698456, "learning_rate": 8.3913923828064e-05, "loss": 0.0521, "step": 29960 }, { "grad_norm": 0.6724990606307983, "learning_rate": 8.390177219590152e-05, "loss": 0.0589, "step": 29970 }, { "grad_norm": 0.7028193473815918, "learning_rate": 8.388961685629222e-05, "loss": 0.0575, "step": 29980 }, { "grad_norm": 0.517053484916687, "learning_rate": 8.387745781056536e-05, "loss": 0.0526, "step": 29990 }, { "grad_norm": 0.779911994934082, "learning_rate": 8.386529506005065e-05, "loss": 0.0685, "step": 30000 }, { "grad_norm": 0.6749436259269714, "learning_rate": 8.38531286060782e-05, "loss": 0.0441, "step": 30010 }, { "grad_norm": 0.4399740397930145, "learning_rate": 8.384095844997849e-05, "loss": 0.0413, "step": 30020 }, { "grad_norm": 0.6427057981491089, "learning_rate": 8.382878459308245e-05, "loss": 0.044, "step": 30030 }, { "grad_norm": 0.6834813356399536, "learning_rate": 8.381660703672138e-05, "loss": 0.062, "step": 30040 }, { "grad_norm": 0.46906813979148865, "learning_rate": 8.380442578222702e-05, "loss": 0.0463, "step": 30050 }, { "grad_norm": 0.6849527359008789, "learning_rate": 8.379224083093146e-05, "loss": 0.0531, "step": 30060 }, { "grad_norm": 0.7611480355262756, "learning_rate": 8.378005218416727e-05, "loss": 0.0853, "step": 30070 }, { "grad_norm": 0.86512291431427, "learning_rate": 8.376785984326735e-05, "loss": 0.0422, "step": 30080 }, { "grad_norm": 0.5106607675552368, "learning_rate": 8.375566380956506e-05, "loss": 0.0592, "step": 30090 }, { "grad_norm": 0.6429929733276367, "learning_rate": 8.374346408439411e-05, "loss": 0.0816, "step": 30100 }, { "grad_norm": 0.5754761099815369, "learning_rate": 8.373126066908868e-05, "loss": 0.0669, "step": 30110 }, { "grad_norm": 0.7356120347976685, "learning_rate": 8.371905356498326e-05, "loss": 0.064, "step": 30120 }, { "grad_norm": 0.5176293253898621, "learning_rate": 8.370684277341288e-05, "loss": 0.046, "step": 30130 }, { "grad_norm": 0.5260486602783203, "learning_rate": 8.369462829571282e-05, "loss": 0.0757, "step": 30140 }, { "grad_norm": 0.47067397832870483, "learning_rate": 8.36824101332189e-05, "loss": 0.0592, "step": 30150 }, { "grad_norm": 0.553410530090332, "learning_rate": 8.367018828726721e-05, "loss": 0.0526, "step": 30160 }, { "grad_norm": 0.577317476272583, "learning_rate": 8.365796275919438e-05, "loss": 0.0595, "step": 30170 }, { "grad_norm": 0.5562425851821899, "learning_rate": 8.364573355033734e-05, "loss": 0.0503, "step": 30180 }, { "grad_norm": 0.7059834003448486, "learning_rate": 8.363350066203346e-05, "loss": 0.0613, "step": 30190 }, { "grad_norm": 0.44053399562835693, "learning_rate": 8.362126409562053e-05, "loss": 0.0434, "step": 30200 }, { "grad_norm": 0.41345325112342834, "learning_rate": 8.360902385243667e-05, "loss": 0.0452, "step": 30210 }, { "grad_norm": 0.4138906002044678, "learning_rate": 8.359677993382052e-05, "loss": 0.0593, "step": 30220 }, { "grad_norm": 0.8812946081161499, "learning_rate": 8.358453234111103e-05, "loss": 0.0648, "step": 30230 }, { "grad_norm": 0.356072336435318, "learning_rate": 8.357228107564756e-05, "loss": 0.038, "step": 30240 }, { "grad_norm": 0.4170108437538147, "learning_rate": 8.356002613876993e-05, "loss": 0.0479, "step": 30250 }, { "grad_norm": 0.47257354855537415, "learning_rate": 8.35477675318183e-05, "loss": 0.0482, "step": 30260 }, { "grad_norm": 0.7944028973579407, "learning_rate": 8.353550525613323e-05, "loss": 0.0498, "step": 30270 }, { "grad_norm": 0.5925614237785339, "learning_rate": 8.352323931305572e-05, "loss": 0.0375, "step": 30280 }, { "grad_norm": 0.9693601727485657, "learning_rate": 8.351096970392717e-05, "loss": 0.0662, "step": 30290 }, { "grad_norm": 0.6806278228759766, "learning_rate": 8.349869643008937e-05, "loss": 0.0514, "step": 30300 }, { "grad_norm": 0.6723130345344543, "learning_rate": 8.348641949288449e-05, "loss": 0.0611, "step": 30310 }, { "grad_norm": 0.7177345752716064, "learning_rate": 8.34741388936551e-05, "loss": 0.0547, "step": 30320 }, { "grad_norm": 0.5758967995643616, "learning_rate": 8.346185463374423e-05, "loss": 0.0481, "step": 30330 }, { "grad_norm": 0.7276887893676758, "learning_rate": 8.344956671449524e-05, "loss": 0.0781, "step": 30340 }, { "grad_norm": 0.8620285987854004, "learning_rate": 8.343727513725192e-05, "loss": 0.0584, "step": 30350 }, { "grad_norm": 0.6229957938194275, "learning_rate": 8.342497990335847e-05, "loss": 0.0593, "step": 30360 }, { "grad_norm": 0.7466050386428833, "learning_rate": 8.341268101415946e-05, "loss": 0.0588, "step": 30370 }, { "grad_norm": 0.5242812633514404, "learning_rate": 8.34003784709999e-05, "loss": 0.0458, "step": 30380 }, { "grad_norm": 0.732214629650116, "learning_rate": 8.338807227522516e-05, "loss": 0.0763, "step": 30390 }, { "grad_norm": 0.6274590492248535, "learning_rate": 8.337576242818103e-05, "loss": 0.051, "step": 30400 }, { "grad_norm": 0.7453847527503967, "learning_rate": 8.336344893121372e-05, "loss": 0.0542, "step": 30410 }, { "grad_norm": 0.6313483119010925, "learning_rate": 8.335113178566977e-05, "loss": 0.0361, "step": 30420 }, { "grad_norm": 0.9709190130233765, "learning_rate": 8.33388109928962e-05, "loss": 0.0574, "step": 30430 }, { "grad_norm": 0.8076876401901245, "learning_rate": 8.33264865542404e-05, "loss": 0.0505, "step": 30440 }, { "grad_norm": 0.36295968294143677, "learning_rate": 8.331415847105013e-05, "loss": 0.04, "step": 30450 }, { "grad_norm": 0.6305891275405884, "learning_rate": 8.330182674467357e-05, "loss": 0.0459, "step": 30460 }, { "grad_norm": 0.7581842541694641, "learning_rate": 8.32894913764593e-05, "loss": 0.0423, "step": 30470 }, { "grad_norm": 0.6527085900306702, "learning_rate": 8.327715236775633e-05, "loss": 0.0561, "step": 30480 }, { "grad_norm": 0.8457051515579224, "learning_rate": 8.326480971991398e-05, "loss": 0.0498, "step": 30490 }, { "grad_norm": 0.7708380818367004, "learning_rate": 8.325246343428206e-05, "loss": 0.0764, "step": 30500 }, { "grad_norm": 0.6664738059043884, "learning_rate": 8.324011351221072e-05, "loss": 0.0593, "step": 30510 }, { "grad_norm": 0.4804321825504303, "learning_rate": 8.322775995505057e-05, "loss": 0.0563, "step": 30520 }, { "grad_norm": 0.7990466952323914, "learning_rate": 8.321540276415254e-05, "loss": 0.0478, "step": 30530 }, { "grad_norm": 0.5122813582420349, "learning_rate": 8.320304194086798e-05, "loss": 0.0445, "step": 30540 }, { "grad_norm": 0.5968475937843323, "learning_rate": 8.31906774865487e-05, "loss": 0.0563, "step": 30550 }, { "grad_norm": 0.4001173675060272, "learning_rate": 8.317830940254682e-05, "loss": 0.0384, "step": 30560 }, { "grad_norm": 0.728603184223175, "learning_rate": 8.316593769021491e-05, "loss": 0.0595, "step": 30570 }, { "grad_norm": 0.8202416896820068, "learning_rate": 8.315356235090592e-05, "loss": 0.0687, "step": 30580 }, { "grad_norm": 1.369171380996704, "learning_rate": 8.314118338597319e-05, "loss": 0.0543, "step": 30590 }, { "grad_norm": 0.6389705538749695, "learning_rate": 8.312880079677048e-05, "loss": 0.0621, "step": 30600 }, { "grad_norm": 0.5960790514945984, "learning_rate": 8.311641458465191e-05, "loss": 0.085, "step": 30610 }, { "grad_norm": 0.76337730884552, "learning_rate": 8.310402475097205e-05, "loss": 0.0633, "step": 30620 }, { "grad_norm": 0.26117298007011414, "learning_rate": 8.309163129708581e-05, "loss": 0.0566, "step": 30630 }, { "grad_norm": 0.6590617895126343, "learning_rate": 8.307923422434852e-05, "loss": 0.0493, "step": 30640 }, { "grad_norm": 0.5911774039268494, "learning_rate": 8.30668335341159e-05, "loss": 0.0504, "step": 30650 }, { "grad_norm": 0.8272029757499695, "learning_rate": 8.305442922774408e-05, "loss": 0.0621, "step": 30660 }, { "grad_norm": 1.403633952140808, "learning_rate": 8.304202130658959e-05, "loss": 0.0723, "step": 30670 }, { "grad_norm": 0.435038685798645, "learning_rate": 8.302960977200931e-05, "loss": 0.0656, "step": 30680 }, { "grad_norm": 0.5426539778709412, "learning_rate": 8.301719462536058e-05, "loss": 0.0549, "step": 30690 }, { "grad_norm": 0.42151641845703125, "learning_rate": 8.300477586800108e-05, "loss": 0.063, "step": 30700 }, { "grad_norm": 0.5289168357849121, "learning_rate": 8.299235350128892e-05, "loss": 0.0513, "step": 30710 }, { "grad_norm": 0.4551810920238495, "learning_rate": 8.297992752658258e-05, "loss": 0.0484, "step": 30720 }, { "grad_norm": 0.5356563925743103, "learning_rate": 8.296749794524094e-05, "loss": 0.0662, "step": 30730 }, { "grad_norm": 0.5363472700119019, "learning_rate": 8.295506475862332e-05, "loss": 0.0444, "step": 30740 }, { "grad_norm": 1.0621157884597778, "learning_rate": 8.294262796808933e-05, "loss": 0.0623, "step": 30750 }, { "grad_norm": 0.4786853790283203, "learning_rate": 8.293018757499909e-05, "loss": 0.045, "step": 30760 }, { "grad_norm": 0.5430902242660522, "learning_rate": 8.291774358071305e-05, "loss": 0.0502, "step": 30770 }, { "grad_norm": 0.6962375044822693, "learning_rate": 8.290529598659205e-05, "loss": 0.0516, "step": 30780 }, { "grad_norm": 1.7019578218460083, "learning_rate": 8.289284479399738e-05, "loss": 0.0768, "step": 30790 }, { "grad_norm": 0.5186309218406677, "learning_rate": 8.288039000429064e-05, "loss": 0.0579, "step": 30800 }, { "grad_norm": 0.5947262644767761, "learning_rate": 8.286793161883388e-05, "loss": 0.0419, "step": 30810 }, { "grad_norm": 0.7808212637901306, "learning_rate": 8.285546963898954e-05, "loss": 0.0503, "step": 30820 }, { "grad_norm": 0.5906331539154053, "learning_rate": 8.284300406612044e-05, "loss": 0.0601, "step": 30830 }, { "grad_norm": 0.5522522330284119, "learning_rate": 8.283053490158978e-05, "loss": 0.0538, "step": 30840 }, { "grad_norm": 0.601551353931427, "learning_rate": 8.28180621467612e-05, "loss": 0.0597, "step": 30850 }, { "grad_norm": 0.6994083523750305, "learning_rate": 8.280558580299867e-05, "loss": 0.0876, "step": 30860 }, { "grad_norm": 0.5606354475021362, "learning_rate": 8.279310587166661e-05, "loss": 0.0593, "step": 30870 }, { "grad_norm": 0.5768272280693054, "learning_rate": 8.278062235412978e-05, "loss": 0.049, "step": 30880 }, { "grad_norm": 0.6391192674636841, "learning_rate": 8.276813525175339e-05, "loss": 0.0781, "step": 30890 }, { "grad_norm": 0.7017138004302979, "learning_rate": 8.2755644565903e-05, "loss": 0.0562, "step": 30900 }, { "grad_norm": 1.156541347503662, "learning_rate": 8.274315029794454e-05, "loss": 0.0595, "step": 30910 }, { "grad_norm": 0.6314411163330078, "learning_rate": 8.273065244924443e-05, "loss": 0.0498, "step": 30920 }, { "grad_norm": 0.8353999257087708, "learning_rate": 8.271815102116936e-05, "loss": 0.0591, "step": 30930 }, { "grad_norm": 0.7308289408683777, "learning_rate": 8.270564601508648e-05, "loss": 0.0546, "step": 30940 }, { "grad_norm": 0.5949510335922241, "learning_rate": 8.269313743236333e-05, "loss": 0.0481, "step": 30950 }, { "grad_norm": 0.6684545278549194, "learning_rate": 8.268062527436783e-05, "loss": 0.0604, "step": 30960 }, { "grad_norm": 0.5969019532203674, "learning_rate": 8.266810954246828e-05, "loss": 0.0597, "step": 30970 }, { "grad_norm": 0.896986186504364, "learning_rate": 8.265559023803338e-05, "loss": 0.064, "step": 30980 }, { "grad_norm": 0.5051249265670776, "learning_rate": 8.264306736243223e-05, "loss": 0.0539, "step": 30990 }, { "grad_norm": 0.6398352384567261, "learning_rate": 8.263054091703432e-05, "loss": 0.0693, "step": 31000 }, { "grad_norm": 0.7672916054725647, "learning_rate": 8.26180109032095e-05, "loss": 0.0792, "step": 31010 }, { "grad_norm": 0.578074038028717, "learning_rate": 8.260547732232807e-05, "loss": 0.0536, "step": 31020 }, { "grad_norm": 0.5123610496520996, "learning_rate": 8.259294017576063e-05, "loss": 0.0529, "step": 31030 }, { "grad_norm": 0.490816593170166, "learning_rate": 8.258039946487828e-05, "loss": 0.0543, "step": 31040 }, { "grad_norm": 0.9279174208641052, "learning_rate": 8.256785519105241e-05, "loss": 0.0545, "step": 31050 }, { "grad_norm": 0.5031466484069824, "learning_rate": 8.255530735565488e-05, "loss": 0.0526, "step": 31060 }, { "grad_norm": 0.6267079710960388, "learning_rate": 8.254275596005787e-05, "loss": 0.0612, "step": 31070 }, { "grad_norm": 0.5941689014434814, "learning_rate": 8.2530201005634e-05, "loss": 0.0473, "step": 31080 }, { "grad_norm": 0.669394314289093, "learning_rate": 8.251764249375626e-05, "loss": 0.052, "step": 31090 }, { "grad_norm": 0.5566632747650146, "learning_rate": 8.250508042579803e-05, "loss": 0.0396, "step": 31100 }, { "grad_norm": 0.5700597167015076, "learning_rate": 8.249251480313307e-05, "loss": 0.061, "step": 31110 }, { "grad_norm": 0.8568575382232666, "learning_rate": 8.247994562713555e-05, "loss": 0.0519, "step": 31120 }, { "grad_norm": 0.4872756898403168, "learning_rate": 8.246737289917998e-05, "loss": 0.0562, "step": 31130 }, { "grad_norm": 0.5570693016052246, "learning_rate": 8.245479662064135e-05, "loss": 0.0504, "step": 31140 }, { "grad_norm": 0.5785778164863586, "learning_rate": 8.244221679289496e-05, "loss": 0.0939, "step": 31150 }, { "grad_norm": 0.6356174945831299, "learning_rate": 8.242963341731652e-05, "loss": 0.0666, "step": 31160 }, { "grad_norm": 0.5641206502914429, "learning_rate": 8.24170464952821e-05, "loss": 0.0411, "step": 31170 }, { "grad_norm": 0.3928571939468384, "learning_rate": 8.240445602816824e-05, "loss": 0.0422, "step": 31180 }, { "grad_norm": 0.5876401662826538, "learning_rate": 8.239186201735179e-05, "loss": 0.0518, "step": 31190 }, { "grad_norm": 0.8788478374481201, "learning_rate": 8.237926446420998e-05, "loss": 0.0506, "step": 31200 }, { "grad_norm": 0.38259658217430115, "learning_rate": 8.236666337012052e-05, "loss": 0.0519, "step": 31210 }, { "grad_norm": 0.7560259699821472, "learning_rate": 8.23540587364614e-05, "loss": 0.0538, "step": 31220 }, { "grad_norm": 0.7879192233085632, "learning_rate": 8.234145056461107e-05, "loss": 0.0487, "step": 31230 }, { "grad_norm": 0.3813084661960602, "learning_rate": 8.232883885594832e-05, "loss": 0.0478, "step": 31240 }, { "grad_norm": 0.6910598874092102, "learning_rate": 8.231622361185236e-05, "loss": 0.0509, "step": 31250 }, { "grad_norm": 0.6722486615180969, "learning_rate": 8.230360483370278e-05, "loss": 0.0518, "step": 31260 }, { "grad_norm": 0.5263000130653381, "learning_rate": 8.229098252287953e-05, "loss": 0.0792, "step": 31270 }, { "grad_norm": 0.634817361831665, "learning_rate": 8.2278356680763e-05, "loss": 0.052, "step": 31280 }, { "grad_norm": 0.6145200729370117, "learning_rate": 8.22657273087339e-05, "loss": 0.0724, "step": 31290 }, { "grad_norm": 0.5443637371063232, "learning_rate": 8.225309440817336e-05, "loss": 0.053, "step": 31300 }, { "grad_norm": 0.6732315421104431, "learning_rate": 8.224045798046293e-05, "loss": 0.0512, "step": 31310 }, { "grad_norm": 0.704111635684967, "learning_rate": 8.22278180269845e-05, "loss": 0.0587, "step": 31320 }, { "grad_norm": 0.5324612259864807, "learning_rate": 8.221517454912031e-05, "loss": 0.0532, "step": 31330 }, { "grad_norm": 0.5609002113342285, "learning_rate": 8.220252754825308e-05, "loss": 0.0645, "step": 31340 }, { "grad_norm": 0.3669925034046173, "learning_rate": 8.218987702576586e-05, "loss": 0.0436, "step": 31350 }, { "grad_norm": 0.5448737740516663, "learning_rate": 8.217722298304207e-05, "loss": 0.0472, "step": 31360 }, { "grad_norm": 0.8444020748138428, "learning_rate": 8.216456542146557e-05, "loss": 0.0502, "step": 31370 }, { "grad_norm": 0.5080223083496094, "learning_rate": 8.215190434242055e-05, "loss": 0.0466, "step": 31380 }, { "grad_norm": 0.5761906504631042, "learning_rate": 8.213923974729161e-05, "loss": 0.0527, "step": 31390 }, { "grad_norm": 1.0112996101379395, "learning_rate": 8.212657163746373e-05, "loss": 0.0586, "step": 31400 }, { "grad_norm": 0.5937061905860901, "learning_rate": 8.211390001432227e-05, "loss": 0.0509, "step": 31410 }, { "grad_norm": 0.5272349715232849, "learning_rate": 8.210122487925297e-05, "loss": 0.0746, "step": 31420 }, { "grad_norm": 0.5299913287162781, "learning_rate": 8.208854623364202e-05, "loss": 0.0536, "step": 31430 }, { "grad_norm": 0.46798175573349, "learning_rate": 8.207586407887589e-05, "loss": 0.0568, "step": 31440 }, { "grad_norm": 0.7504488229751587, "learning_rate": 8.206317841634148e-05, "loss": 0.0597, "step": 31450 }, { "grad_norm": 0.7285634279251099, "learning_rate": 8.205048924742609e-05, "loss": 0.0549, "step": 31460 }, { "grad_norm": 0.6056930422782898, "learning_rate": 8.203779657351738e-05, "loss": 0.049, "step": 31470 }, { "grad_norm": 0.47627297043800354, "learning_rate": 8.20251003960034e-05, "loss": 0.0468, "step": 31480 }, { "grad_norm": 0.6253790855407715, "learning_rate": 8.201240071627258e-05, "loss": 0.0483, "step": 31490 }, { "grad_norm": 0.6653507351875305, "learning_rate": 8.199969753571377e-05, "loss": 0.0552, "step": 31500 }, { "grad_norm": 0.3982885777950287, "learning_rate": 8.198699085571615e-05, "loss": 0.041, "step": 31510 }, { "grad_norm": 0.6066991686820984, "learning_rate": 8.197428067766928e-05, "loss": 0.0476, "step": 31520 }, { "grad_norm": 0.40723809599876404, "learning_rate": 8.196156700296316e-05, "loss": 0.0376, "step": 31530 }, { "grad_norm": 0.46467354893684387, "learning_rate": 8.194884983298814e-05, "loss": 0.0634, "step": 31540 }, { "grad_norm": 0.39901894330978394, "learning_rate": 8.193612916913491e-05, "loss": 0.0642, "step": 31550 }, { "grad_norm": 0.6364478468894958, "learning_rate": 8.192340501279463e-05, "loss": 0.046, "step": 31560 }, { "grad_norm": 0.6673020124435425, "learning_rate": 8.191067736535876e-05, "loss": 0.0622, "step": 31570 }, { "grad_norm": 1.011710286140442, "learning_rate": 8.18979462282192e-05, "loss": 0.0559, "step": 31580 }, { "grad_norm": 0.6421170234680176, "learning_rate": 8.188521160276819e-05, "loss": 0.0532, "step": 31590 }, { "grad_norm": 0.5509462356567383, "learning_rate": 8.187247349039837e-05, "loss": 0.0466, "step": 31600 }, { "grad_norm": 0.4774319529533386, "learning_rate": 8.185973189250278e-05, "loss": 0.0503, "step": 31610 }, { "grad_norm": 0.5968803763389587, "learning_rate": 8.184698681047482e-05, "loss": 0.0514, "step": 31620 }, { "grad_norm": 0.6321961879730225, "learning_rate": 8.183423824570827e-05, "loss": 0.0504, "step": 31630 }, { "grad_norm": 0.5054248571395874, "learning_rate": 8.182148619959725e-05, "loss": 0.0494, "step": 31640 }, { "grad_norm": 0.5265031456947327, "learning_rate": 8.180873067353636e-05, "loss": 0.0566, "step": 31650 }, { "grad_norm": 0.5074480175971985, "learning_rate": 8.179597166892052e-05, "loss": 0.0537, "step": 31660 }, { "grad_norm": 0.7040980458259583, "learning_rate": 8.178320918714501e-05, "loss": 0.0485, "step": 31670 }, { "grad_norm": 0.5555534362792969, "learning_rate": 8.177044322960554e-05, "loss": 0.0577, "step": 31680 }, { "grad_norm": 0.4920259118080139, "learning_rate": 8.175767379769816e-05, "loss": 0.0449, "step": 31690 }, { "grad_norm": 0.897796094417572, "learning_rate": 8.174490089281932e-05, "loss": 0.0605, "step": 31700 }, { "grad_norm": 0.6817065477371216, "learning_rate": 8.173212451636584e-05, "loss": 0.0592, "step": 31710 }, { "grad_norm": 0.4585786461830139, "learning_rate": 8.171934466973493e-05, "loss": 0.0508, "step": 31720 }, { "grad_norm": 0.5071755647659302, "learning_rate": 8.170656135432418e-05, "loss": 0.06, "step": 31730 }, { "grad_norm": 0.4139961302280426, "learning_rate": 8.169377457153155e-05, "loss": 0.052, "step": 31740 }, { "grad_norm": 0.6883107423782349, "learning_rate": 8.168098432275539e-05, "loss": 0.0801, "step": 31750 }, { "grad_norm": 0.5099492073059082, "learning_rate": 8.166819060939442e-05, "loss": 0.047, "step": 31760 }, { "grad_norm": 0.8338611125946045, "learning_rate": 8.165539343284772e-05, "loss": 0.0654, "step": 31770 }, { "grad_norm": 0.6045947074890137, "learning_rate": 8.16425927945148e-05, "loss": 0.0515, "step": 31780 }, { "grad_norm": 0.7129067778587341, "learning_rate": 8.162978869579551e-05, "loss": 0.0539, "step": 31790 }, { "grad_norm": 0.5324943661689758, "learning_rate": 8.161698113809007e-05, "loss": 0.0568, "step": 31800 }, { "grad_norm": 0.517873227596283, "learning_rate": 8.160417012279911e-05, "loss": 0.0575, "step": 31810 }, { "grad_norm": 0.5369989275932312, "learning_rate": 8.159135565132363e-05, "loss": 0.044, "step": 31820 }, { "grad_norm": 0.4751207232475281, "learning_rate": 8.157853772506498e-05, "loss": 0.0533, "step": 31830 }, { "grad_norm": 0.6486778259277344, "learning_rate": 8.156571634542494e-05, "loss": 0.0617, "step": 31840 }, { "grad_norm": 0.9163649678230286, "learning_rate": 8.15528915138056e-05, "loss": 0.0561, "step": 31850 }, { "grad_norm": 0.7181866765022278, "learning_rate": 8.154006323160949e-05, "loss": 0.0641, "step": 31860 }, { "grad_norm": 0.5021969079971313, "learning_rate": 8.152723150023949e-05, "loss": 0.056, "step": 31870 }, { "grad_norm": 0.6817652583122253, "learning_rate": 8.151439632109886e-05, "loss": 0.055, "step": 31880 }, { "grad_norm": 0.6153611540794373, "learning_rate": 8.150155769559122e-05, "loss": 0.0618, "step": 31890 }, { "grad_norm": 0.4537021815776825, "learning_rate": 8.148871562512058e-05, "loss": 0.0445, "step": 31900 }, { "grad_norm": 0.4473828971385956, "learning_rate": 8.147587011109136e-05, "loss": 0.0581, "step": 31910 }, { "grad_norm": 0.6230180859565735, "learning_rate": 8.14630211549083e-05, "loss": 0.0431, "step": 31920 }, { "grad_norm": 0.5693441033363342, "learning_rate": 8.145016875797655e-05, "loss": 0.0584, "step": 31930 }, { "grad_norm": 0.39715778827667236, "learning_rate": 8.143731292170164e-05, "loss": 0.0631, "step": 31940 }, { "grad_norm": 0.5977907180786133, "learning_rate": 8.142445364748944e-05, "loss": 0.0594, "step": 31950 }, { "grad_norm": 0.7245537042617798, "learning_rate": 8.141159093674624e-05, "loss": 0.047, "step": 31960 }, { "grad_norm": 0.6130446195602417, "learning_rate": 8.139872479087869e-05, "loss": 0.0646, "step": 31970 }, { "grad_norm": 0.6301936507225037, "learning_rate": 8.13858552112938e-05, "loss": 0.0608, "step": 31980 }, { "grad_norm": 0.7350335717201233, "learning_rate": 8.137298219939895e-05, "loss": 0.0493, "step": 31990 }, { "grad_norm": 0.6153267025947571, "learning_rate": 8.136010575660196e-05, "loss": 0.0693, "step": 32000 }, { "grad_norm": 0.5947290658950806, "learning_rate": 8.134722588431095e-05, "loss": 0.06, "step": 32010 }, { "grad_norm": 0.6807273626327515, "learning_rate": 8.133434258393444e-05, "loss": 0.0605, "step": 32020 }, { "grad_norm": 0.5284030437469482, "learning_rate": 8.132145585688134e-05, "loss": 0.0523, "step": 32030 }, { "grad_norm": 0.5053837299346924, "learning_rate": 8.130856570456093e-05, "loss": 0.0425, "step": 32040 }, { "grad_norm": 1.0955045223236084, "learning_rate": 8.129567212838283e-05, "loss": 0.0721, "step": 32050 }, { "grad_norm": 0.5912254452705383, "learning_rate": 8.128277512975708e-05, "loss": 0.0497, "step": 32060 }, { "grad_norm": 0.641704797744751, "learning_rate": 8.126987471009408e-05, "loss": 0.0592, "step": 32070 }, { "grad_norm": 0.5449431538581848, "learning_rate": 8.125697087080459e-05, "loss": 0.0499, "step": 32080 }, { "grad_norm": 0.32773613929748535, "learning_rate": 8.124406361329976e-05, "loss": 0.0451, "step": 32090 }, { "grad_norm": 0.8254051804542542, "learning_rate": 8.123115293899111e-05, "loss": 0.0532, "step": 32100 }, { "grad_norm": 0.5405400991439819, "learning_rate": 8.121823884929055e-05, "loss": 0.0534, "step": 32110 }, { "grad_norm": 0.4606790244579315, "learning_rate": 8.12053213456103e-05, "loss": 0.0516, "step": 32120 }, { "grad_norm": 0.5091182589530945, "learning_rate": 8.119240042936303e-05, "loss": 0.0473, "step": 32130 }, { "grad_norm": 0.8549163937568665, "learning_rate": 8.117947610196175e-05, "loss": 0.0492, "step": 32140 }, { "grad_norm": 0.5928491950035095, "learning_rate": 8.116654836481982e-05, "loss": 0.0521, "step": 32150 }, { "grad_norm": 0.4165918231010437, "learning_rate": 8.115361721935106e-05, "loss": 0.0578, "step": 32160 }, { "grad_norm": 0.5074857473373413, "learning_rate": 8.114068266696953e-05, "loss": 0.0446, "step": 32170 }, { "grad_norm": 0.6454291939735413, "learning_rate": 8.112774470908978e-05, "loss": 0.0512, "step": 32180 }, { "grad_norm": 0.7047528028488159, "learning_rate": 8.111480334712665e-05, "loss": 0.05, "step": 32190 }, { "grad_norm": 0.7095699310302734, "learning_rate": 8.110185858249542e-05, "loss": 0.0462, "step": 32200 }, { "grad_norm": 0.793553352355957, "learning_rate": 8.108891041661168e-05, "loss": 0.0499, "step": 32210 }, { "grad_norm": 0.9311846494674683, "learning_rate": 8.107595885089146e-05, "loss": 0.0511, "step": 32220 }, { "grad_norm": 0.5004375576972961, "learning_rate": 8.106300388675108e-05, "loss": 0.0656, "step": 32230 }, { "grad_norm": 0.6049401760101318, "learning_rate": 8.105004552560731e-05, "loss": 0.0563, "step": 32240 }, { "grad_norm": 0.5281969904899597, "learning_rate": 8.103708376887724e-05, "loss": 0.0417, "step": 32250 }, { "grad_norm": 0.8970789313316345, "learning_rate": 8.102411861797836e-05, "loss": 0.062, "step": 32260 }, { "grad_norm": 0.659622311592102, "learning_rate": 8.101115007432851e-05, "loss": 0.0452, "step": 32270 }, { "grad_norm": 0.6835274696350098, "learning_rate": 8.09981781393459e-05, "loss": 0.0624, "step": 32280 }, { "grad_norm": 0.7955119013786316, "learning_rate": 8.098520281444915e-05, "loss": 0.0723, "step": 32290 }, { "grad_norm": 0.40713441371917725, "learning_rate": 8.09722241010572e-05, "loss": 0.0584, "step": 32300 }, { "grad_norm": 0.6622258424758911, "learning_rate": 8.095924200058939e-05, "loss": 0.0688, "step": 32310 }, { "grad_norm": 0.5303836464881897, "learning_rate": 8.094625651446541e-05, "loss": 0.0396, "step": 32320 }, { "grad_norm": 0.5996628999710083, "learning_rate": 8.093326764410536e-05, "loss": 0.0467, "step": 32330 }, { "grad_norm": 0.8029459118843079, "learning_rate": 8.092027539092966e-05, "loss": 0.044, "step": 32340 }, { "grad_norm": 0.6611806154251099, "learning_rate": 8.090727975635913e-05, "loss": 0.0429, "step": 32350 }, { "grad_norm": 0.4777602553367615, "learning_rate": 8.089428074181497e-05, "loss": 0.0454, "step": 32360 }, { "grad_norm": 0.4356628358364105, "learning_rate": 8.088127834871871e-05, "loss": 0.0473, "step": 32370 }, { "grad_norm": 0.7179421782493591, "learning_rate": 8.086827257849226e-05, "loss": 0.0493, "step": 32380 }, { "grad_norm": 0.4433465600013733, "learning_rate": 8.085526343255795e-05, "loss": 0.0528, "step": 32390 }, { "grad_norm": 0.6582457423210144, "learning_rate": 8.084225091233842e-05, "loss": 0.048, "step": 32400 }, { "grad_norm": 0.5331718325614929, "learning_rate": 8.082923501925668e-05, "loss": 0.0763, "step": 32410 }, { "grad_norm": 0.6036883592605591, "learning_rate": 8.081621575473617e-05, "loss": 0.0537, "step": 32420 }, { "grad_norm": 0.37275075912475586, "learning_rate": 8.080319312020064e-05, "loss": 0.0516, "step": 32430 }, { "grad_norm": 0.385597288608551, "learning_rate": 8.079016711707421e-05, "loss": 0.0414, "step": 32440 }, { "grad_norm": 0.7590147256851196, "learning_rate": 8.077713774678139e-05, "loss": 0.0576, "step": 32450 }, { "grad_norm": 0.6790410280227661, "learning_rate": 8.076410501074707e-05, "loss": 0.0717, "step": 32460 }, { "grad_norm": 0.7098140716552734, "learning_rate": 8.075106891039647e-05, "loss": 0.0588, "step": 32470 }, { "grad_norm": 0.6283175349235535, "learning_rate": 8.073802944715523e-05, "loss": 0.0662, "step": 32480 }, { "grad_norm": 0.5381463170051575, "learning_rate": 8.072498662244929e-05, "loss": 0.0508, "step": 32490 }, { "grad_norm": 0.609125554561615, "learning_rate": 8.0711940437705e-05, "loss": 0.0578, "step": 32500 }, { "grad_norm": 0.622779905796051, "learning_rate": 8.06988908943491e-05, "loss": 0.0551, "step": 32510 }, { "grad_norm": 0.7575052380561829, "learning_rate": 8.068583799380863e-05, "loss": 0.052, "step": 32520 }, { "grad_norm": 0.5542747378349304, "learning_rate": 8.067278173751104e-05, "loss": 0.0445, "step": 32530 }, { "grad_norm": 0.7524160146713257, "learning_rate": 8.065972212688417e-05, "loss": 0.0587, "step": 32540 }, { "grad_norm": 0.49961182475090027, "learning_rate": 8.064665916335618e-05, "loss": 0.0682, "step": 32550 }, { "grad_norm": 0.49039551615715027, "learning_rate": 8.063359284835564e-05, "loss": 0.0358, "step": 32560 }, { "grad_norm": 1.0708342790603638, "learning_rate": 8.062052318331142e-05, "loss": 0.0544, "step": 32570 }, { "grad_norm": 0.4576076865196228, "learning_rate": 8.060745016965283e-05, "loss": 0.0572, "step": 32580 }, { "grad_norm": 0.4821472465991974, "learning_rate": 8.059437380880952e-05, "loss": 0.042, "step": 32590 }, { "grad_norm": 0.5226573944091797, "learning_rate": 8.058129410221146e-05, "loss": 0.0539, "step": 32600 }, { "grad_norm": 0.5808736085891724, "learning_rate": 8.056821105128908e-05, "loss": 0.0524, "step": 32610 }, { "grad_norm": 0.4145182967185974, "learning_rate": 8.05551246574731e-05, "loss": 0.0567, "step": 32620 }, { "grad_norm": 0.5080671310424805, "learning_rate": 8.05420349221946e-05, "loss": 0.0429, "step": 32630 }, { "grad_norm": 0.8884959816932678, "learning_rate": 8.05289418468851e-05, "loss": 0.051, "step": 32640 }, { "grad_norm": 0.9098924994468689, "learning_rate": 8.051584543297642e-05, "loss": 0.0681, "step": 32650 }, { "grad_norm": 0.401007741689682, "learning_rate": 8.050274568190074e-05, "loss": 0.0388, "step": 32660 }, { "grad_norm": 0.6648406982421875, "learning_rate": 8.048964259509067e-05, "loss": 0.0502, "step": 32670 }, { "grad_norm": 0.3838648498058319, "learning_rate": 8.047653617397914e-05, "loss": 0.0456, "step": 32680 }, { "grad_norm": 0.4706100821495056, "learning_rate": 8.046342641999941e-05, "loss": 0.0516, "step": 32690 }, { "grad_norm": 0.6138302087783813, "learning_rate": 8.045031333458517e-05, "loss": 0.053, "step": 32700 }, { "grad_norm": 0.5693729519844055, "learning_rate": 8.043719691917047e-05, "loss": 0.0591, "step": 32710 }, { "grad_norm": 0.5874732136726379, "learning_rate": 8.042407717518966e-05, "loss": 0.0454, "step": 32720 }, { "grad_norm": 0.5838337540626526, "learning_rate": 8.041095410407751e-05, "loss": 0.051, "step": 32730 }, { "grad_norm": 0.440300315618515, "learning_rate": 8.039782770726913e-05, "loss": 0.0607, "step": 32740 }, { "grad_norm": 0.636647641658783, "learning_rate": 8.038469798620004e-05, "loss": 0.0574, "step": 32750 }, { "grad_norm": 0.4636276066303253, "learning_rate": 8.037156494230604e-05, "loss": 0.0426, "step": 32760 }, { "grad_norm": 0.649649441242218, "learning_rate": 8.035842857702338e-05, "loss": 0.0638, "step": 32770 }, { "grad_norm": 0.4673628807067871, "learning_rate": 8.03452888917886e-05, "loss": 0.0644, "step": 32780 }, { "grad_norm": 0.47702619433403015, "learning_rate": 8.033214588803866e-05, "loss": 0.0494, "step": 32790 }, { "grad_norm": 0.702059268951416, "learning_rate": 8.031899956721083e-05, "loss": 0.0518, "step": 32800 }, { "grad_norm": 0.7588040828704834, "learning_rate": 8.030584993074282e-05, "loss": 0.0487, "step": 32810 }, { "grad_norm": 0.724223792552948, "learning_rate": 8.02926969800726e-05, "loss": 0.0545, "step": 32820 }, { "grad_norm": 0.6231944561004639, "learning_rate": 8.027954071663859e-05, "loss": 0.0565, "step": 32830 }, { "grad_norm": 0.6719257831573486, "learning_rate": 8.026638114187954e-05, "loss": 0.0669, "step": 32840 }, { "grad_norm": 0.7316499948501587, "learning_rate": 8.025321825723456e-05, "loss": 0.0631, "step": 32850 }, { "grad_norm": 0.7952733039855957, "learning_rate": 8.02400520641431e-05, "loss": 0.0553, "step": 32860 }, { "grad_norm": 0.4196450412273407, "learning_rate": 8.022688256404501e-05, "loss": 0.043, "step": 32870 }, { "grad_norm": 0.3525383472442627, "learning_rate": 8.02137097583805e-05, "loss": 0.0489, "step": 32880 }, { "grad_norm": 0.7068825960159302, "learning_rate": 8.02005336485901e-05, "loss": 0.0405, "step": 32890 }, { "grad_norm": 0.4125959575176239, "learning_rate": 8.018735423611476e-05, "loss": 0.0482, "step": 32900 }, { "grad_norm": 0.31077075004577637, "learning_rate": 8.017417152239574e-05, "loss": 0.0314, "step": 32910 }, { "grad_norm": 0.43971607089042664, "learning_rate": 8.01609855088747e-05, "loss": 0.0659, "step": 32920 }, { "grad_norm": 0.5390576720237732, "learning_rate": 8.014779619699362e-05, "loss": 0.0501, "step": 32930 }, { "grad_norm": 0.5866856575012207, "learning_rate": 8.013460358819489e-05, "loss": 0.05, "step": 32940 }, { "grad_norm": 0.9153217673301697, "learning_rate": 8.01214076839212e-05, "loss": 0.0459, "step": 32950 }, { "grad_norm": 0.611281156539917, "learning_rate": 8.010820848561565e-05, "loss": 0.0628, "step": 32960 }, { "grad_norm": 0.546671450138092, "learning_rate": 8.009500599472171e-05, "loss": 0.0387, "step": 32970 }, { "grad_norm": 0.4061844050884247, "learning_rate": 8.008180021268314e-05, "loss": 0.0567, "step": 32980 }, { "grad_norm": 0.6055881977081299, "learning_rate": 8.006859114094414e-05, "loss": 0.0397, "step": 32990 }, { "grad_norm": 0.8733118772506714, "learning_rate": 8.005537878094921e-05, "loss": 0.0483, "step": 33000 }, { "grad_norm": 0.5876281261444092, "learning_rate": 8.004216313414323e-05, "loss": 0.0442, "step": 33010 }, { "grad_norm": 0.7235252261161804, "learning_rate": 8.002894420197149e-05, "loss": 0.0397, "step": 33020 }, { "grad_norm": 0.5208247900009155, "learning_rate": 8.001572198587954e-05, "loss": 0.0421, "step": 33030 }, { "grad_norm": 0.6476679444313049, "learning_rate": 8.000249648731338e-05, "loss": 0.0484, "step": 33040 }, { "grad_norm": 0.4655742943286896, "learning_rate": 7.998926770771928e-05, "loss": 0.0532, "step": 33050 }, { "grad_norm": 0.5073981881141663, "learning_rate": 7.997603564854397e-05, "loss": 0.0536, "step": 33060 }, { "grad_norm": 0.8659059405326843, "learning_rate": 7.996280031123448e-05, "loss": 0.0516, "step": 33070 }, { "grad_norm": 0.7470903992652893, "learning_rate": 7.994956169723818e-05, "loss": 0.0507, "step": 33080 }, { "grad_norm": 0.5483083724975586, "learning_rate": 7.993631980800285e-05, "loss": 0.0433, "step": 33090 }, { "grad_norm": 0.6262773275375366, "learning_rate": 7.992307464497659e-05, "loss": 0.0575, "step": 33100 }, { "grad_norm": 0.4425686299800873, "learning_rate": 7.990982620960787e-05, "loss": 0.0603, "step": 33110 }, { "grad_norm": 0.49310973286628723, "learning_rate": 7.989657450334554e-05, "loss": 0.0607, "step": 33120 }, { "grad_norm": 0.4535355567932129, "learning_rate": 7.988331952763877e-05, "loss": 0.0548, "step": 33130 }, { "grad_norm": 0.7136440873146057, "learning_rate": 7.987006128393709e-05, "loss": 0.048, "step": 33140 }, { "grad_norm": 0.6988893151283264, "learning_rate": 7.985679977369043e-05, "loss": 0.0611, "step": 33150 }, { "grad_norm": 0.48911744356155396, "learning_rate": 7.984353499834902e-05, "loss": 0.0552, "step": 33160 }, { "grad_norm": 0.5607160329818726, "learning_rate": 7.983026695936351e-05, "loss": 0.0504, "step": 33170 }, { "grad_norm": 0.3788264989852905, "learning_rate": 7.981699565818486e-05, "loss": 0.0509, "step": 33180 }, { "grad_norm": 0.4891946017742157, "learning_rate": 7.980372109626437e-05, "loss": 0.0412, "step": 33190 }, { "grad_norm": 1.2967795133590698, "learning_rate": 7.979044327505375e-05, "loss": 0.0448, "step": 33200 }, { "grad_norm": 0.26123860478401184, "learning_rate": 7.977716219600506e-05, "loss": 0.0385, "step": 33210 }, { "grad_norm": 0.7598969340324402, "learning_rate": 7.97638778605707e-05, "loss": 0.0372, "step": 33220 }, { "grad_norm": 0.4015129804611206, "learning_rate": 7.975059027020338e-05, "loss": 0.0491, "step": 33230 }, { "grad_norm": 0.4871114492416382, "learning_rate": 7.973729942635623e-05, "loss": 0.0322, "step": 33240 }, { "grad_norm": 0.4347975552082062, "learning_rate": 7.972400533048273e-05, "loss": 0.038, "step": 33250 }, { "grad_norm": 0.37724483013153076, "learning_rate": 7.97107079840367e-05, "loss": 0.0498, "step": 33260 }, { "grad_norm": 0.36784833669662476, "learning_rate": 7.969740738847231e-05, "loss": 0.0479, "step": 33270 }, { "grad_norm": 0.37558797001838684, "learning_rate": 7.968410354524411e-05, "loss": 0.0604, "step": 33280 }, { "grad_norm": 0.4950485825538635, "learning_rate": 7.967079645580697e-05, "loss": 0.0588, "step": 33290 }, { "grad_norm": 0.45179128646850586, "learning_rate": 7.965748612161612e-05, "loss": 0.0488, "step": 33300 }, { "grad_norm": 0.5452841520309448, "learning_rate": 7.96441725441272e-05, "loss": 0.0565, "step": 33310 }, { "grad_norm": 0.5426152944564819, "learning_rate": 7.963085572479614e-05, "loss": 0.0517, "step": 33320 }, { "grad_norm": 0.5464736223220825, "learning_rate": 7.961753566507924e-05, "loss": 0.0445, "step": 33330 }, { "grad_norm": 0.5418044924736023, "learning_rate": 7.960421236643316e-05, "loss": 0.0803, "step": 33340 }, { "grad_norm": 0.4443526864051819, "learning_rate": 7.959088583031496e-05, "loss": 0.0491, "step": 33350 }, { "grad_norm": 0.42071962356567383, "learning_rate": 7.957755605818194e-05, "loss": 0.0685, "step": 33360 }, { "grad_norm": 0.4718034267425537, "learning_rate": 7.956422305149185e-05, "loss": 0.0433, "step": 33370 }, { "grad_norm": 0.5396480560302734, "learning_rate": 7.95508868117028e-05, "loss": 0.0504, "step": 33380 }, { "grad_norm": 0.5048331022262573, "learning_rate": 7.953754734027318e-05, "loss": 0.0518, "step": 33390 }, { "grad_norm": 0.5975258946418762, "learning_rate": 7.952420463866182e-05, "loss": 0.0506, "step": 33400 }, { "grad_norm": 0.7550088763237, "learning_rate": 7.951085870832782e-05, "loss": 0.0698, "step": 33410 }, { "grad_norm": 0.6353141665458679, "learning_rate": 7.949750955073067e-05, "loss": 0.0536, "step": 33420 }, { "grad_norm": 0.9085369110107422, "learning_rate": 7.948415716733022e-05, "loss": 0.0698, "step": 33430 }, { "grad_norm": 0.8921648263931274, "learning_rate": 7.947080155958669e-05, "loss": 0.0636, "step": 33440 }, { "grad_norm": 0.8895484805107117, "learning_rate": 7.94574427289606e-05, "loss": 0.0421, "step": 33450 }, { "grad_norm": 0.7928355932235718, "learning_rate": 7.944408067691284e-05, "loss": 0.0882, "step": 33460 }, { "grad_norm": 0.6266704797744751, "learning_rate": 7.943071540490473e-05, "loss": 0.0403, "step": 33470 }, { "grad_norm": 0.4508627653121948, "learning_rate": 7.94173469143978e-05, "loss": 0.0519, "step": 33480 }, { "grad_norm": 0.4395706057548523, "learning_rate": 7.940397520685406e-05, "loss": 0.044, "step": 33490 }, { "grad_norm": 0.4709222912788391, "learning_rate": 7.939060028373577e-05, "loss": 0.0744, "step": 33500 }, { "grad_norm": 0.9113515615463257, "learning_rate": 7.937722214650565e-05, "loss": 0.0547, "step": 33510 }, { "grad_norm": 0.5411149859428406, "learning_rate": 7.936384079662666e-05, "loss": 0.0555, "step": 33520 }, { "grad_norm": 0.5938262343406677, "learning_rate": 7.93504562355622e-05, "loss": 0.0552, "step": 33530 }, { "grad_norm": 0.6799737215042114, "learning_rate": 7.933706846477599e-05, "loss": 0.0438, "step": 33540 }, { "grad_norm": 0.5965458154678345, "learning_rate": 7.932367748573206e-05, "loss": 0.0478, "step": 33550 }, { "grad_norm": 0.5638999342918396, "learning_rate": 7.931028329989485e-05, "loss": 0.0519, "step": 33560 }, { "grad_norm": 0.5988129377365112, "learning_rate": 7.929688590872913e-05, "loss": 0.0512, "step": 33570 }, { "grad_norm": 0.8142867088317871, "learning_rate": 7.928348531370003e-05, "loss": 0.0454, "step": 33580 }, { "grad_norm": 0.5871816873550415, "learning_rate": 7.927008151627297e-05, "loss": 0.047, "step": 33590 }, { "grad_norm": 0.4826081097126007, "learning_rate": 7.925667451791383e-05, "loss": 0.0516, "step": 33600 }, { "grad_norm": 0.4254705607891083, "learning_rate": 7.924326432008874e-05, "loss": 0.0413, "step": 33610 }, { "grad_norm": 0.6206925511360168, "learning_rate": 7.922985092426422e-05, "loss": 0.0566, "step": 33620 }, { "grad_norm": 0.5993123650550842, "learning_rate": 7.921643433190717e-05, "loss": 0.0582, "step": 33630 }, { "grad_norm": 0.4087766110897064, "learning_rate": 7.920301454448478e-05, "loss": 0.0434, "step": 33640 }, { "grad_norm": 0.8726099133491516, "learning_rate": 7.918959156346461e-05, "loss": 0.0505, "step": 33650 }, { "grad_norm": 0.4633631408214569, "learning_rate": 7.91761653903146e-05, "loss": 0.0544, "step": 33660 }, { "grad_norm": 0.5651317834854126, "learning_rate": 7.916273602650302e-05, "loss": 0.0578, "step": 33670 }, { "grad_norm": 0.7356703281402588, "learning_rate": 7.914930347349847e-05, "loss": 0.0529, "step": 33680 }, { "grad_norm": 0.5283923745155334, "learning_rate": 7.913586773276992e-05, "loss": 0.0522, "step": 33690 }, { "grad_norm": 0.5190812349319458, "learning_rate": 7.912242880578667e-05, "loss": 0.0454, "step": 33700 }, { "grad_norm": 0.42858025431632996, "learning_rate": 7.910898669401839e-05, "loss": 0.0471, "step": 33710 }, { "grad_norm": 0.7722651362419128, "learning_rate": 7.909554139893511e-05, "loss": 0.0708, "step": 33720 }, { "grad_norm": 0.4566277861595154, "learning_rate": 7.908209292200715e-05, "loss": 0.0457, "step": 33730 }, { "grad_norm": 0.5674446821212769, "learning_rate": 7.906864126470523e-05, "loss": 0.0393, "step": 33740 }, { "grad_norm": 0.7794837355613708, "learning_rate": 7.905518642850041e-05, "loss": 0.0449, "step": 33750 }, { "grad_norm": 0.44577500224113464, "learning_rate": 7.904172841486409e-05, "loss": 0.0439, "step": 33760 }, { "grad_norm": 0.530280590057373, "learning_rate": 7.902826722526801e-05, "loss": 0.0602, "step": 33770 }, { "grad_norm": 0.5935344696044922, "learning_rate": 7.901480286118427e-05, "loss": 0.0598, "step": 33780 }, { "grad_norm": 0.7040578722953796, "learning_rate": 7.900133532408531e-05, "loss": 0.0477, "step": 33790 }, { "grad_norm": 0.6248672604560852, "learning_rate": 7.898786461544395e-05, "loss": 0.0408, "step": 33800 }, { "grad_norm": 0.40181633830070496, "learning_rate": 7.897439073673325e-05, "loss": 0.0362, "step": 33810 }, { "grad_norm": 0.5367425680160522, "learning_rate": 7.896091368942677e-05, "loss": 0.0504, "step": 33820 }, { "grad_norm": 0.8294945359230042, "learning_rate": 7.894743347499832e-05, "loss": 0.0829, "step": 33830 }, { "grad_norm": 0.8300879001617432, "learning_rate": 7.893395009492203e-05, "loss": 0.0468, "step": 33840 }, { "grad_norm": 0.538795530796051, "learning_rate": 7.892046355067248e-05, "loss": 0.0605, "step": 33850 }, { "grad_norm": 0.9739481210708618, "learning_rate": 7.890697384372451e-05, "loss": 0.0734, "step": 33860 }, { "grad_norm": 0.6928396224975586, "learning_rate": 7.889348097555336e-05, "loss": 0.0508, "step": 33870 }, { "grad_norm": 0.6469647288322449, "learning_rate": 7.887998494763455e-05, "loss": 0.0503, "step": 33880 }, { "grad_norm": 0.5522807240486145, "learning_rate": 7.886648576144404e-05, "loss": 0.0551, "step": 33890 }, { "grad_norm": 0.6458017826080322, "learning_rate": 7.885298341845802e-05, "loss": 0.0663, "step": 33900 }, { "grad_norm": 0.32540345191955566, "learning_rate": 7.883947792015311e-05, "loss": 0.0469, "step": 33910 }, { "grad_norm": 0.5669247508049011, "learning_rate": 7.882596926800628e-05, "loss": 0.0558, "step": 33920 }, { "grad_norm": 0.6093870997428894, "learning_rate": 7.881245746349477e-05, "loss": 0.0589, "step": 33930 }, { "grad_norm": 0.6137499213218689, "learning_rate": 7.879894250809623e-05, "loss": 0.0479, "step": 33940 }, { "grad_norm": 0.6407944560050964, "learning_rate": 7.878542440328865e-05, "loss": 0.0493, "step": 33950 }, { "grad_norm": 0.5967684984207153, "learning_rate": 7.877190315055031e-05, "loss": 0.0501, "step": 33960 }, { "grad_norm": 1.0850931406021118, "learning_rate": 7.875837875135991e-05, "loss": 0.0678, "step": 33970 }, { "grad_norm": 0.5909207463264465, "learning_rate": 7.874485120719646e-05, "loss": 0.0646, "step": 33980 }, { "grad_norm": 0.4573403298854828, "learning_rate": 7.873132051953928e-05, "loss": 0.0458, "step": 33990 }, { "grad_norm": 0.718601644039154, "learning_rate": 7.87177866898681e-05, "loss": 0.0505, "step": 34000 }, { "grad_norm": 0.5881073474884033, "learning_rate": 7.870424971966294e-05, "loss": 0.0562, "step": 34010 }, { "grad_norm": 0.7463275194168091, "learning_rate": 7.869070961040419e-05, "loss": 0.073, "step": 34020 }, { "grad_norm": 0.5896986126899719, "learning_rate": 7.867716636357257e-05, "loss": 0.0522, "step": 34030 }, { "grad_norm": 0.6327846646308899, "learning_rate": 7.866361998064915e-05, "loss": 0.0602, "step": 34040 }, { "grad_norm": 0.8246455788612366, "learning_rate": 7.865007046311534e-05, "loss": 0.0647, "step": 34050 }, { "grad_norm": 0.4672936201095581, "learning_rate": 7.86365178124529e-05, "loss": 0.0424, "step": 34060 }, { "grad_norm": 0.5719612836837769, "learning_rate": 7.862296203014394e-05, "loss": 0.0541, "step": 34070 }, { "grad_norm": 1.2524226903915405, "learning_rate": 7.860940311767088e-05, "loss": 0.0513, "step": 34080 }, { "grad_norm": 0.6656429171562195, "learning_rate": 7.85958410765165e-05, "loss": 0.0461, "step": 34090 }, { "grad_norm": 0.7866065502166748, "learning_rate": 7.858227590816394e-05, "loss": 0.0543, "step": 34100 }, { "grad_norm": 0.3825374245643616, "learning_rate": 7.856870761409664e-05, "loss": 0.0427, "step": 34110 }, { "grad_norm": 0.7608758211135864, "learning_rate": 7.855513619579846e-05, "loss": 0.0757, "step": 34120 }, { "grad_norm": 0.30133822560310364, "learning_rate": 7.85415616547535e-05, "loss": 0.0514, "step": 34130 }, { "grad_norm": 0.8412281274795532, "learning_rate": 7.852798399244627e-05, "loss": 0.0551, "step": 34140 }, { "grad_norm": 0.44482889771461487, "learning_rate": 7.851440321036161e-05, "loss": 0.0687, "step": 34150 }, { "grad_norm": 0.6351081728935242, "learning_rate": 7.850081930998468e-05, "loss": 0.0447, "step": 34160 }, { "grad_norm": 0.5813955664634705, "learning_rate": 7.8487232292801e-05, "loss": 0.0506, "step": 34170 }, { "grad_norm": 0.3917408883571625, "learning_rate": 7.847364216029642e-05, "loss": 0.057, "step": 34180 }, { "grad_norm": 0.550289511680603, "learning_rate": 7.846004891395716e-05, "loss": 0.0467, "step": 34190 }, { "grad_norm": 0.5745021104812622, "learning_rate": 7.844645255526972e-05, "loss": 0.0459, "step": 34200 }, { "grad_norm": 0.6945371031761169, "learning_rate": 7.843285308572101e-05, "loss": 0.0407, "step": 34210 }, { "grad_norm": 0.5446130037307739, "learning_rate": 7.841925050679823e-05, "loss": 0.0642, "step": 34220 }, { "grad_norm": 0.5925203561782837, "learning_rate": 7.840564481998895e-05, "loss": 0.0489, "step": 34230 }, { "grad_norm": 0.7772892117500305, "learning_rate": 7.839203602678105e-05, "loss": 0.0546, "step": 34240 }, { "grad_norm": 0.6108940839767456, "learning_rate": 7.837842412866279e-05, "loss": 0.0499, "step": 34250 }, { "grad_norm": 0.7964046597480774, "learning_rate": 7.836480912712272e-05, "loss": 0.0515, "step": 34260 }, { "grad_norm": 0.4998132884502411, "learning_rate": 7.835119102364976e-05, "loss": 0.0512, "step": 34270 }, { "grad_norm": 0.7608523964881897, "learning_rate": 7.83375698197332e-05, "loss": 0.0465, "step": 34280 }, { "grad_norm": 0.7598235011100769, "learning_rate": 7.83239455168626e-05, "loss": 0.0597, "step": 34290 }, { "grad_norm": 0.6434313654899597, "learning_rate": 7.83103181165279e-05, "loss": 0.0574, "step": 34300 }, { "grad_norm": 0.5536467432975769, "learning_rate": 7.829668762021937e-05, "loss": 0.0447, "step": 34310 }, { "grad_norm": 0.4387056827545166, "learning_rate": 7.828305402942764e-05, "loss": 0.0576, "step": 34320 }, { "grad_norm": 0.4804636538028717, "learning_rate": 7.826941734564363e-05, "loss": 0.071, "step": 34330 }, { "grad_norm": 0.47118768095970154, "learning_rate": 7.825577757035865e-05, "loss": 0.0534, "step": 34340 }, { "grad_norm": 0.35253849625587463, "learning_rate": 7.824213470506431e-05, "loss": 0.0562, "step": 34350 }, { "grad_norm": 0.6222870945930481, "learning_rate": 7.822848875125257e-05, "loss": 0.0548, "step": 34360 }, { "grad_norm": 0.6383029818534851, "learning_rate": 7.821483971041576e-05, "loss": 0.0485, "step": 34370 }, { "grad_norm": 0.6916298270225525, "learning_rate": 7.820118758404649e-05, "loss": 0.0418, "step": 34380 }, { "grad_norm": 0.9351388812065125, "learning_rate": 7.818753237363776e-05, "loss": 0.0523, "step": 34390 }, { "grad_norm": 0.5378203988075256, "learning_rate": 7.817387408068286e-05, "loss": 0.0563, "step": 34400 }, { "grad_norm": 0.436878502368927, "learning_rate": 7.816021270667544e-05, "loss": 0.0531, "step": 34410 }, { "grad_norm": 0.7944342494010925, "learning_rate": 7.81465482531095e-05, "loss": 0.0497, "step": 34420 }, { "grad_norm": 0.41429027915000916, "learning_rate": 7.813288072147938e-05, "loss": 0.0506, "step": 34430 }, { "grad_norm": 0.494333952665329, "learning_rate": 7.811921011327972e-05, "loss": 0.0475, "step": 34440 }, { "grad_norm": 0.8018460273742676, "learning_rate": 7.810553643000549e-05, "loss": 0.0454, "step": 34450 }, { "grad_norm": 0.5311409831047058, "learning_rate": 7.809185967315206e-05, "loss": 0.0436, "step": 34460 }, { "grad_norm": 0.6762900352478027, "learning_rate": 7.80781798442151e-05, "loss": 0.0416, "step": 34470 }, { "grad_norm": 0.7432007789611816, "learning_rate": 7.806449694469061e-05, "loss": 0.0607, "step": 34480 }, { "grad_norm": 0.7603092789649963, "learning_rate": 7.805081097607492e-05, "loss": 0.0681, "step": 34490 }, { "grad_norm": 0.5166007876396179, "learning_rate": 7.803712193986474e-05, "loss": 0.047, "step": 34500 }, { "grad_norm": 0.47539061307907104, "learning_rate": 7.802342983755702e-05, "loss": 0.0447, "step": 34510 }, { "grad_norm": 0.593461811542511, "learning_rate": 7.800973467064918e-05, "loss": 0.0394, "step": 34520 }, { "grad_norm": 0.8103374242782593, "learning_rate": 7.799603644063884e-05, "loss": 0.0532, "step": 34530 }, { "grad_norm": 0.44475388526916504, "learning_rate": 7.798233514902405e-05, "loss": 0.0467, "step": 34540 }, { "grad_norm": 0.7544155120849609, "learning_rate": 7.796863079730318e-05, "loss": 0.0474, "step": 34550 }, { "grad_norm": 0.3735411763191223, "learning_rate": 7.795492338697488e-05, "loss": 0.0365, "step": 34560 }, { "grad_norm": 0.4050881862640381, "learning_rate": 7.794121291953819e-05, "loss": 0.0458, "step": 34570 }, { "grad_norm": 0.4344370663166046, "learning_rate": 7.792749939649246e-05, "loss": 0.065, "step": 34580 }, { "grad_norm": 0.6179997324943542, "learning_rate": 7.79137828193374e-05, "loss": 0.0571, "step": 34590 }, { "grad_norm": 0.3618978261947632, "learning_rate": 7.790006318957301e-05, "loss": 0.0439, "step": 34600 }, { "grad_norm": 1.1461695432662964, "learning_rate": 7.788634050869965e-05, "loss": 0.0674, "step": 34610 }, { "grad_norm": 0.5276060104370117, "learning_rate": 7.787261477821803e-05, "loss": 0.0417, "step": 34620 }, { "grad_norm": 0.3541554808616638, "learning_rate": 7.785888599962916e-05, "loss": 0.0312, "step": 34630 }, { "grad_norm": 0.4313119351863861, "learning_rate": 7.784515417443439e-05, "loss": 0.0437, "step": 34640 }, { "grad_norm": 0.6446402668952942, "learning_rate": 7.783141930413545e-05, "loss": 0.048, "step": 34650 }, { "grad_norm": 0.381531685590744, "learning_rate": 7.78176813902343e-05, "loss": 0.0369, "step": 34660 }, { "grad_norm": 0.666385293006897, "learning_rate": 7.780394043423336e-05, "loss": 0.0487, "step": 34670 }, { "grad_norm": 0.5840229988098145, "learning_rate": 7.77901964376353e-05, "loss": 0.0437, "step": 34680 }, { "grad_norm": 0.5769203901290894, "learning_rate": 7.777644940194316e-05, "loss": 0.0482, "step": 34690 }, { "grad_norm": 0.7753841280937195, "learning_rate": 7.776269932866023e-05, "loss": 0.0401, "step": 34700 }, { "grad_norm": 0.7265809774398804, "learning_rate": 7.774894621929026e-05, "loss": 0.0511, "step": 34710 }, { "grad_norm": 0.630693256855011, "learning_rate": 7.773519007533725e-05, "loss": 0.0538, "step": 34720 }, { "grad_norm": 0.9908807873725891, "learning_rate": 7.772143089830556e-05, "loss": 0.0473, "step": 34730 }, { "grad_norm": 0.8212049603462219, "learning_rate": 7.770766868969985e-05, "loss": 0.0575, "step": 34740 }, { "grad_norm": 0.5259498953819275, "learning_rate": 7.769390345102518e-05, "loss": 0.0581, "step": 34750 }, { "grad_norm": 0.540101945400238, "learning_rate": 7.768013518378683e-05, "loss": 0.0422, "step": 34760 }, { "grad_norm": 0.7648714184761047, "learning_rate": 7.766636388949053e-05, "loss": 0.0581, "step": 34770 }, { "grad_norm": 0.4760996997356415, "learning_rate": 7.765258956964229e-05, "loss": 0.045, "step": 34780 }, { "grad_norm": 0.4050212800502777, "learning_rate": 7.76388122257484e-05, "loss": 0.0475, "step": 34790 }, { "grad_norm": 0.5135324001312256, "learning_rate": 7.762503185931558e-05, "loss": 0.0525, "step": 34800 }, { "grad_norm": 0.5394927859306335, "learning_rate": 7.76112484718508e-05, "loss": 0.0421, "step": 34810 }, { "grad_norm": 0.5418906211853027, "learning_rate": 7.75974620648614e-05, "loss": 0.0326, "step": 34820 }, { "grad_norm": 0.6413031220436096, "learning_rate": 7.758367263985503e-05, "loss": 0.038, "step": 34830 }, { "grad_norm": 0.40761512517929077, "learning_rate": 7.75698801983397e-05, "loss": 0.0471, "step": 34840 }, { "grad_norm": 0.5289561748504639, "learning_rate": 7.755608474182372e-05, "loss": 0.0439, "step": 34850 }, { "grad_norm": 0.5418462753295898, "learning_rate": 7.754228627181574e-05, "loss": 0.0528, "step": 34860 }, { "grad_norm": 0.32330387830734253, "learning_rate": 7.752848478982476e-05, "loss": 0.0435, "step": 34870 }, { "grad_norm": 0.5767136812210083, "learning_rate": 7.751468029736006e-05, "loss": 0.067, "step": 34880 }, { "grad_norm": 0.5709933638572693, "learning_rate": 7.750087279593129e-05, "loss": 0.0471, "step": 34890 }, { "grad_norm": 0.809452474117279, "learning_rate": 7.748706228704843e-05, "loss": 0.0546, "step": 34900 }, { "grad_norm": 0.4409797489643097, "learning_rate": 7.747324877222176e-05, "loss": 0.0305, "step": 34910 }, { "grad_norm": 0.6651010513305664, "learning_rate": 7.745943225296188e-05, "loss": 0.0555, "step": 34920 }, { "grad_norm": 0.5241127610206604, "learning_rate": 7.744561273077981e-05, "loss": 0.0459, "step": 34930 }, { "grad_norm": 2.5079352855682373, "learning_rate": 7.743179020718678e-05, "loss": 0.0425, "step": 34940 }, { "grad_norm": 0.8686830401420593, "learning_rate": 7.741796468369443e-05, "loss": 0.0564, "step": 34950 }, { "grad_norm": 0.40363165736198425, "learning_rate": 7.740413616181466e-05, "loss": 0.0468, "step": 34960 }, { "grad_norm": 0.6986941695213318, "learning_rate": 7.739030464305978e-05, "loss": 0.0491, "step": 34970 }, { "grad_norm": 0.43169862031936646, "learning_rate": 7.737647012894235e-05, "loss": 0.0428, "step": 34980 }, { "grad_norm": 0.5538209080696106, "learning_rate": 7.736263262097532e-05, "loss": 0.0496, "step": 34990 }, { "grad_norm": 0.41967660188674927, "learning_rate": 7.734879212067192e-05, "loss": 0.0424, "step": 35000 }, { "grad_norm": 0.4856729507446289, "learning_rate": 7.733494862954573e-05, "loss": 0.0487, "step": 35010 }, { "grad_norm": 0.3719133138656616, "learning_rate": 7.732110214911066e-05, "loss": 0.0492, "step": 35020 }, { "grad_norm": 0.4222795069217682, "learning_rate": 7.730725268088094e-05, "loss": 0.0482, "step": 35030 }, { "grad_norm": 0.32974371314048767, "learning_rate": 7.729340022637112e-05, "loss": 0.0452, "step": 35040 }, { "grad_norm": 0.7355738878250122, "learning_rate": 7.727954478709607e-05, "loss": 0.0488, "step": 35050 }, { "grad_norm": 0.47854214906692505, "learning_rate": 7.726568636457103e-05, "loss": 0.0491, "step": 35060 }, { "grad_norm": 0.5146725177764893, "learning_rate": 7.725182496031153e-05, "loss": 0.0691, "step": 35070 }, { "grad_norm": 0.40901899337768555, "learning_rate": 7.723796057583341e-05, "loss": 0.0407, "step": 35080 }, { "grad_norm": 0.5670910477638245, "learning_rate": 7.722409321265287e-05, "loss": 0.0396, "step": 35090 }, { "grad_norm": 0.938880443572998, "learning_rate": 7.721022287228645e-05, "loss": 0.0495, "step": 35100 }, { "grad_norm": 0.4506698250770569, "learning_rate": 7.719634955625094e-05, "loss": 0.0635, "step": 35110 }, { "grad_norm": 0.6686111688613892, "learning_rate": 7.718247326606355e-05, "loss": 0.0569, "step": 35120 }, { "grad_norm": 0.6487115621566772, "learning_rate": 7.716859400324175e-05, "loss": 0.0493, "step": 35130 }, { "grad_norm": 0.7855050563812256, "learning_rate": 7.715471176930335e-05, "loss": 0.0656, "step": 35140 }, { "grad_norm": 0.4313220679759979, "learning_rate": 7.714082656576651e-05, "loss": 0.0567, "step": 35150 }, { "grad_norm": 0.6244755983352661, "learning_rate": 7.712693839414968e-05, "loss": 0.0438, "step": 35160 }, { "grad_norm": 0.6662998199462891, "learning_rate": 7.711304725597164e-05, "loss": 0.0563, "step": 35170 }, { "grad_norm": 0.6250693202018738, "learning_rate": 7.709915315275151e-05, "loss": 0.0543, "step": 35180 }, { "grad_norm": 0.7831920981407166, "learning_rate": 7.708525608600876e-05, "loss": 0.0471, "step": 35190 }, { "grad_norm": 0.5358557105064392, "learning_rate": 7.707135605726311e-05, "loss": 0.0887, "step": 35200 }, { "grad_norm": 0.6756272315979004, "learning_rate": 7.705745306803466e-05, "loss": 0.0552, "step": 35210 }, { "grad_norm": 0.5002084374427795, "learning_rate": 7.704354711984383e-05, "loss": 0.0403, "step": 35220 }, { "grad_norm": 0.30878886580467224, "learning_rate": 7.702963821421133e-05, "loss": 0.0448, "step": 35230 }, { "grad_norm": 0.5891147255897522, "learning_rate": 7.701572635265826e-05, "loss": 0.0491, "step": 35240 }, { "grad_norm": 0.4729122519493103, "learning_rate": 7.700181153670596e-05, "loss": 0.0488, "step": 35250 }, { "grad_norm": 0.7050588130950928, "learning_rate": 7.698789376787616e-05, "loss": 0.0465, "step": 35260 }, { "grad_norm": 0.6832442283630371, "learning_rate": 7.697397304769085e-05, "loss": 0.058, "step": 35270 }, { "grad_norm": 0.3942515552043915, "learning_rate": 7.696004937767241e-05, "loss": 0.0466, "step": 35280 }, { "grad_norm": 0.7535952925682068, "learning_rate": 7.694612275934352e-05, "loss": 0.0584, "step": 35290 }, { "grad_norm": 0.6632048487663269, "learning_rate": 7.693219319422714e-05, "loss": 0.0384, "step": 35300 }, { "grad_norm": 0.5400245189666748, "learning_rate": 7.69182606838466e-05, "loss": 0.0347, "step": 35310 }, { "grad_norm": 0.5650948286056519, "learning_rate": 7.690432522972558e-05, "loss": 0.0871, "step": 35320 }, { "grad_norm": 0.44987356662750244, "learning_rate": 7.689038683338796e-05, "loss": 0.0375, "step": 35330 }, { "grad_norm": 0.44046807289123535, "learning_rate": 7.687644549635808e-05, "loss": 0.0506, "step": 35340 }, { "grad_norm": 0.45744532346725464, "learning_rate": 7.686250122016053e-05, "loss": 0.0562, "step": 35350 }, { "grad_norm": 1.1545283794403076, "learning_rate": 7.684855400632022e-05, "loss": 0.0496, "step": 35360 }, { "grad_norm": 0.34720832109451294, "learning_rate": 7.683460385636243e-05, "loss": 0.0554, "step": 35370 }, { "grad_norm": 0.8829121589660645, "learning_rate": 7.68206507718127e-05, "loss": 0.0431, "step": 35380 }, { "grad_norm": 0.4780283272266388, "learning_rate": 7.680669475419693e-05, "loss": 0.0538, "step": 35390 }, { "grad_norm": 0.6282024383544922, "learning_rate": 7.679273580504132e-05, "loss": 0.0361, "step": 35400 }, { "grad_norm": 0.5410629510879517, "learning_rate": 7.67787739258724e-05, "loss": 0.0415, "step": 35410 }, { "grad_norm": 1.0911850929260254, "learning_rate": 7.676480911821705e-05, "loss": 0.0607, "step": 35420 }, { "grad_norm": 0.9909916520118713, "learning_rate": 7.675084138360238e-05, "loss": 0.0469, "step": 35430 }, { "grad_norm": 0.34993019700050354, "learning_rate": 7.673687072355592e-05, "loss": 0.0536, "step": 35440 }, { "grad_norm": 0.4970482289791107, "learning_rate": 7.67228971396055e-05, "loss": 0.0436, "step": 35450 }, { "grad_norm": 0.785029411315918, "learning_rate": 7.670892063327922e-05, "loss": 0.0449, "step": 35460 }, { "grad_norm": 0.5112438201904297, "learning_rate": 7.669494120610552e-05, "loss": 0.0397, "step": 35470 }, { "grad_norm": 0.7446653842926025, "learning_rate": 7.66809588596132e-05, "loss": 0.0751, "step": 35480 }, { "grad_norm": 0.9342462420463562, "learning_rate": 7.666697359533132e-05, "loss": 0.0468, "step": 35490 }, { "grad_norm": 0.5291959047317505, "learning_rate": 7.665298541478932e-05, "loss": 0.0718, "step": 35500 }, { "grad_norm": 0.6815918684005737, "learning_rate": 7.66389943195169e-05, "loss": 0.0439, "step": 35510 }, { "grad_norm": 0.7470179200172424, "learning_rate": 7.66250003110441e-05, "loss": 0.0706, "step": 35520 }, { "grad_norm": 0.46516478061676025, "learning_rate": 7.66110033909013e-05, "loss": 0.0414, "step": 35530 }, { "grad_norm": 0.5297002196311951, "learning_rate": 7.659700356061918e-05, "loss": 0.0445, "step": 35540 }, { "grad_norm": 0.8187874555587769, "learning_rate": 7.658300082172875e-05, "loss": 0.0426, "step": 35550 }, { "grad_norm": 0.461933434009552, "learning_rate": 7.65689951757613e-05, "loss": 0.0679, "step": 35560 }, { "grad_norm": 1.2603672742843628, "learning_rate": 7.655498662424849e-05, "loss": 0.0476, "step": 35570 }, { "grad_norm": 0.3819424510002136, "learning_rate": 7.654097516872227e-05, "loss": 0.0388, "step": 35580 }, { "grad_norm": 0.4526887536048889, "learning_rate": 7.65269608107149e-05, "loss": 0.0541, "step": 35590 }, { "grad_norm": 0.6562023758888245, "learning_rate": 7.6512943551759e-05, "loss": 0.0473, "step": 35600 }, { "grad_norm": 1.721889853477478, "learning_rate": 7.649892339338743e-05, "loss": 0.0558, "step": 35610 }, { "grad_norm": 0.5364156365394592, "learning_rate": 7.648490033713344e-05, "loss": 0.055, "step": 35620 }, { "grad_norm": 0.49886244535446167, "learning_rate": 7.647087438453058e-05, "loss": 0.041, "step": 35630 }, { "grad_norm": 0.8022003769874573, "learning_rate": 7.64568455371127e-05, "loss": 0.0514, "step": 35640 }, { "grad_norm": 0.42282000184059143, "learning_rate": 7.644281379641396e-05, "loss": 0.0527, "step": 35650 }, { "grad_norm": 0.6209678649902344, "learning_rate": 7.642877916396887e-05, "loss": 0.0449, "step": 35660 }, { "grad_norm": 0.5557528138160706, "learning_rate": 7.641474164131221e-05, "loss": 0.0458, "step": 35670 }, { "grad_norm": 0.6175469756126404, "learning_rate": 7.640070122997913e-05, "loss": 0.0488, "step": 35680 }, { "grad_norm": 0.7306159734725952, "learning_rate": 7.638665793150505e-05, "loss": 0.0495, "step": 35690 }, { "grad_norm": 0.8010202050209045, "learning_rate": 7.637261174742574e-05, "loss": 0.0539, "step": 35700 }, { "grad_norm": 0.6574764847755432, "learning_rate": 7.635856267927725e-05, "loss": 0.0507, "step": 35710 }, { "grad_norm": 0.43944305181503296, "learning_rate": 7.634451072859597e-05, "loss": 0.051, "step": 35720 }, { "grad_norm": 0.6159563660621643, "learning_rate": 7.633045589691863e-05, "loss": 0.0389, "step": 35730 }, { "grad_norm": 0.8099306225776672, "learning_rate": 7.63163981857822e-05, "loss": 0.0507, "step": 35740 }, { "grad_norm": 0.6352263689041138, "learning_rate": 7.630233759672403e-05, "loss": 0.0485, "step": 35750 }, { "grad_norm": 0.5761810541152954, "learning_rate": 7.628827413128177e-05, "loss": 0.0503, "step": 35760 }, { "grad_norm": 0.7697047591209412, "learning_rate": 7.627420779099338e-05, "loss": 0.0445, "step": 35770 }, { "grad_norm": 0.9567749500274658, "learning_rate": 7.626013857739711e-05, "loss": 0.0497, "step": 35780 }, { "grad_norm": 0.7990264892578125, "learning_rate": 7.624606649203158e-05, "loss": 0.0453, "step": 35790 }, { "grad_norm": 0.9529138207435608, "learning_rate": 7.623199153643569e-05, "loss": 0.0573, "step": 35800 }, { "grad_norm": 0.7328391671180725, "learning_rate": 7.621791371214863e-05, "loss": 0.0438, "step": 35810 }, { "grad_norm": 0.62629234790802, "learning_rate": 7.620383302070995e-05, "loss": 0.0586, "step": 35820 }, { "grad_norm": 0.8249144554138184, "learning_rate": 7.61897494636595e-05, "loss": 0.0493, "step": 35830 }, { "grad_norm": 0.5661147832870483, "learning_rate": 7.617566304253739e-05, "loss": 0.0475, "step": 35840 }, { "grad_norm": 0.53675377368927, "learning_rate": 7.616157375888416e-05, "loss": 0.0411, "step": 35850 }, { "grad_norm": 0.5011516213417053, "learning_rate": 7.614748161424053e-05, "loss": 0.0476, "step": 35860 }, { "grad_norm": 0.46120402216911316, "learning_rate": 7.613338661014763e-05, "loss": 0.0571, "step": 35870 }, { "grad_norm": 0.4354272782802582, "learning_rate": 7.611928874814686e-05, "loss": 0.0529, "step": 35880 }, { "grad_norm": 0.4098428785800934, "learning_rate": 7.610518802977996e-05, "loss": 0.0692, "step": 35890 }, { "grad_norm": 0.526465654373169, "learning_rate": 7.609108445658893e-05, "loss": 0.0428, "step": 35900 }, { "grad_norm": 0.7878656387329102, "learning_rate": 7.607697803011612e-05, "loss": 0.0493, "step": 35910 }, { "grad_norm": 0.7664912343025208, "learning_rate": 7.606286875190421e-05, "loss": 0.0577, "step": 35920 }, { "grad_norm": 0.7248660922050476, "learning_rate": 7.604875662349617e-05, "loss": 0.0458, "step": 35930 }, { "grad_norm": 0.44310227036476135, "learning_rate": 7.603464164643525e-05, "loss": 0.0441, "step": 35940 }, { "grad_norm": 0.6491826772689819, "learning_rate": 7.602052382226507e-05, "loss": 0.0637, "step": 35950 }, { "grad_norm": 0.6958886384963989, "learning_rate": 7.600640315252954e-05, "loss": 0.0602, "step": 35960 }, { "grad_norm": 0.9661158323287964, "learning_rate": 7.599227963877284e-05, "loss": 0.0533, "step": 35970 }, { "grad_norm": 0.43400707840919495, "learning_rate": 7.597815328253954e-05, "loss": 0.0474, "step": 35980 }, { "grad_norm": 0.443435937166214, "learning_rate": 7.596402408537443e-05, "loss": 0.0629, "step": 35990 }, { "grad_norm": 0.45850691199302673, "learning_rate": 7.59498920488227e-05, "loss": 0.0431, "step": 36000 }, { "grad_norm": 0.5397999286651611, "learning_rate": 7.593575717442979e-05, "loss": 0.0532, "step": 36010 }, { "grad_norm": 0.4742410182952881, "learning_rate": 7.592161946374147e-05, "loss": 0.0642, "step": 36020 }, { "grad_norm": 0.7051717042922974, "learning_rate": 7.590747891830381e-05, "loss": 0.0628, "step": 36030 }, { "grad_norm": 0.3869585692882538, "learning_rate": 7.58933355396632e-05, "loss": 0.0492, "step": 36040 }, { "grad_norm": 0.7109806537628174, "learning_rate": 7.587918932936636e-05, "loss": 0.0694, "step": 36050 }, { "grad_norm": 0.6120812296867371, "learning_rate": 7.586504028896028e-05, "loss": 0.0437, "step": 36060 }, { "grad_norm": 0.4119345247745514, "learning_rate": 7.585088841999228e-05, "loss": 0.0492, "step": 36070 }, { "grad_norm": 0.4882510304450989, "learning_rate": 7.583673372400999e-05, "loss": 0.0531, "step": 36080 }, { "grad_norm": 0.6598861217498779, "learning_rate": 7.582257620256134e-05, "loss": 0.054, "step": 36090 }, { "grad_norm": 0.6321027278900146, "learning_rate": 7.580841585719458e-05, "loss": 0.0491, "step": 36100 }, { "grad_norm": 0.8340703248977661, "learning_rate": 7.579425268945825e-05, "loss": 0.0433, "step": 36110 }, { "grad_norm": 0.6037326455116272, "learning_rate": 7.578008670090127e-05, "loss": 0.0514, "step": 36120 }, { "grad_norm": 0.7766664028167725, "learning_rate": 7.576591789307272e-05, "loss": 0.0614, "step": 36130 }, { "grad_norm": 0.6376461982727051, "learning_rate": 7.575174626752216e-05, "loss": 0.0408, "step": 36140 }, { "grad_norm": 0.5394293069839478, "learning_rate": 7.573757182579934e-05, "loss": 0.0451, "step": 36150 }, { "grad_norm": 0.4737640619277954, "learning_rate": 7.572339456945435e-05, "loss": 0.036, "step": 36160 }, { "grad_norm": 0.8038409948348999, "learning_rate": 7.570921450003762e-05, "loss": 0.0503, "step": 36170 }, { "grad_norm": 0.6120113134384155, "learning_rate": 7.569503161909983e-05, "loss": 0.0416, "step": 36180 }, { "grad_norm": 0.4639597237110138, "learning_rate": 7.568084592819202e-05, "loss": 0.0516, "step": 36190 }, { "grad_norm": 0.5368704795837402, "learning_rate": 7.566665742886551e-05, "loss": 0.0631, "step": 36200 }, { "grad_norm": 0.40845340490341187, "learning_rate": 7.565246612267194e-05, "loss": 0.0549, "step": 36210 }, { "grad_norm": 0.9820067286491394, "learning_rate": 7.563827201116325e-05, "loss": 0.0583, "step": 36220 }, { "grad_norm": 0.5227813720703125, "learning_rate": 7.562407509589164e-05, "loss": 0.0494, "step": 36230 }, { "grad_norm": 0.4851265549659729, "learning_rate": 7.560987537840973e-05, "loss": 0.0448, "step": 36240 }, { "grad_norm": 0.36458730697631836, "learning_rate": 7.559567286027036e-05, "loss": 0.0422, "step": 36250 }, { "grad_norm": 0.809861421585083, "learning_rate": 7.558146754302668e-05, "loss": 0.0493, "step": 36260 }, { "grad_norm": 0.5310388803482056, "learning_rate": 7.556725942823217e-05, "loss": 0.0446, "step": 36270 }, { "grad_norm": 0.4515098035335541, "learning_rate": 7.555304851744061e-05, "loss": 0.0446, "step": 36280 }, { "grad_norm": 0.5198311805725098, "learning_rate": 7.553883481220608e-05, "loss": 0.041, "step": 36290 }, { "grad_norm": 0.5547401905059814, "learning_rate": 7.552461831408298e-05, "loss": 0.0526, "step": 36300 }, { "grad_norm": 0.6559155583381653, "learning_rate": 7.551039902462599e-05, "loss": 0.0487, "step": 36310 }, { "grad_norm": 0.4773240089416504, "learning_rate": 7.549617694539011e-05, "loss": 0.0412, "step": 36320 }, { "grad_norm": 0.3483622372150421, "learning_rate": 7.548195207793066e-05, "loss": 0.0507, "step": 36330 }, { "grad_norm": 0.49280649423599243, "learning_rate": 7.546772442380323e-05, "loss": 0.0376, "step": 36340 }, { "grad_norm": 0.5440158843994141, "learning_rate": 7.545349398456376e-05, "loss": 0.0311, "step": 36350 }, { "grad_norm": 0.3818818926811218, "learning_rate": 7.543926076176845e-05, "loss": 0.0394, "step": 36360 }, { "grad_norm": 0.9357197880744934, "learning_rate": 7.542502475697385e-05, "loss": 0.0362, "step": 36370 }, { "grad_norm": 0.32892724871635437, "learning_rate": 7.541078597173675e-05, "loss": 0.0409, "step": 36380 }, { "grad_norm": 0.544319748878479, "learning_rate": 7.539654440761431e-05, "loss": 0.0501, "step": 36390 }, { "grad_norm": 0.41969290375709534, "learning_rate": 7.538230006616395e-05, "loss": 0.0468, "step": 36400 }, { "grad_norm": 1.0718376636505127, "learning_rate": 7.536805294894342e-05, "loss": 0.051, "step": 36410 }, { "grad_norm": 0.7706447839736938, "learning_rate": 7.535380305751076e-05, "loss": 0.0525, "step": 36420 }, { "grad_norm": 0.8216004371643066, "learning_rate": 7.533955039342431e-05, "loss": 0.0382, "step": 36430 }, { "grad_norm": 0.738533079624176, "learning_rate": 7.532529495824274e-05, "loss": 0.0388, "step": 36440 }, { "grad_norm": 1.1539075374603271, "learning_rate": 7.531103675352497e-05, "loss": 0.0489, "step": 36450 }, { "grad_norm": 0.6111423373222351, "learning_rate": 7.529677578083029e-05, "loss": 0.0483, "step": 36460 }, { "grad_norm": 0.5958170294761658, "learning_rate": 7.528251204171823e-05, "loss": 0.0485, "step": 36470 }, { "grad_norm": 0.6615422368049622, "learning_rate": 7.526824553774867e-05, "loss": 0.042, "step": 36480 }, { "grad_norm": 0.447603315114975, "learning_rate": 7.525397627048178e-05, "loss": 0.0685, "step": 36490 }, { "grad_norm": 0.8933366537094116, "learning_rate": 7.523970424147802e-05, "loss": 0.0549, "step": 36500 }, { "grad_norm": 0.5262845754623413, "learning_rate": 7.522542945229813e-05, "loss": 0.0483, "step": 36510 }, { "grad_norm": 0.5158271789550781, "learning_rate": 7.521115190450321e-05, "loss": 0.05, "step": 36520 }, { "grad_norm": 0.4380260407924652, "learning_rate": 7.519687159965462e-05, "loss": 0.057, "step": 36530 }, { "grad_norm": 0.836806058883667, "learning_rate": 7.518258853931403e-05, "loss": 0.058, "step": 36540 }, { "grad_norm": 0.43774324655532837, "learning_rate": 7.516830272504342e-05, "loss": 0.0415, "step": 36550 }, { "grad_norm": 0.5467644929885864, "learning_rate": 7.515401415840509e-05, "loss": 0.0435, "step": 36560 }, { "grad_norm": 0.6774062514305115, "learning_rate": 7.513972284096155e-05, "loss": 0.0569, "step": 36570 }, { "grad_norm": 0.7470629215240479, "learning_rate": 7.512542877427574e-05, "loss": 0.0438, "step": 36580 }, { "grad_norm": 0.4838353991508484, "learning_rate": 7.51111319599108e-05, "loss": 0.0442, "step": 36590 }, { "grad_norm": 0.5635437965393066, "learning_rate": 7.509683239943024e-05, "loss": 0.0458, "step": 36600 }, { "grad_norm": 0.6100821495056152, "learning_rate": 7.50825300943978e-05, "loss": 0.0498, "step": 36610 }, { "grad_norm": 0.419900506734848, "learning_rate": 7.50682250463776e-05, "loss": 0.0567, "step": 36620 }, { "grad_norm": 0.42976897954940796, "learning_rate": 7.5053917256934e-05, "loss": 0.0466, "step": 36630 }, { "grad_norm": 0.8417443037033081, "learning_rate": 7.503960672763166e-05, "loss": 0.055, "step": 36640 }, { "grad_norm": 0.42285585403442383, "learning_rate": 7.502529346003559e-05, "loss": 0.0489, "step": 36650 }, { "grad_norm": 0.4942253828048706, "learning_rate": 7.501097745571106e-05, "loss": 0.0364, "step": 36660 }, { "grad_norm": 0.4978860914707184, "learning_rate": 7.499665871622361e-05, "loss": 0.0529, "step": 36670 }, { "grad_norm": 0.6088745594024658, "learning_rate": 7.498233724313917e-05, "loss": 0.0461, "step": 36680 }, { "grad_norm": 0.5017567873001099, "learning_rate": 7.496801303802387e-05, "loss": 0.0392, "step": 36690 }, { "grad_norm": 0.8718788623809814, "learning_rate": 7.495368610244423e-05, "loss": 0.0427, "step": 36700 }, { "grad_norm": 0.2972511351108551, "learning_rate": 7.493935643796697e-05, "loss": 0.0349, "step": 36710 }, { "grad_norm": 0.4978031814098358, "learning_rate": 7.49250240461592e-05, "loss": 0.038, "step": 36720 }, { "grad_norm": 0.4665164351463318, "learning_rate": 7.491068892858826e-05, "loss": 0.0416, "step": 36730 }, { "grad_norm": 0.5516916513442993, "learning_rate": 7.489635108682184e-05, "loss": 0.0466, "step": 36740 }, { "grad_norm": 0.6870456337928772, "learning_rate": 7.48820105224279e-05, "loss": 0.0513, "step": 36750 }, { "grad_norm": 0.8327728509902954, "learning_rate": 7.48676672369747e-05, "loss": 0.0388, "step": 36760 }, { "grad_norm": 0.5777459740638733, "learning_rate": 7.485332123203079e-05, "loss": 0.0438, "step": 36770 }, { "grad_norm": 0.463329941034317, "learning_rate": 7.483897250916504e-05, "loss": 0.0379, "step": 36780 }, { "grad_norm": 0.5961059331893921, "learning_rate": 7.48246210699466e-05, "loss": 0.043, "step": 36790 }, { "grad_norm": 0.6239539384841919, "learning_rate": 7.481026691594492e-05, "loss": 0.0835, "step": 36800 }, { "grad_norm": 0.5314226150512695, "learning_rate": 7.479591004872975e-05, "loss": 0.0413, "step": 36810 }, { "grad_norm": 1.280177354812622, "learning_rate": 7.478155046987114e-05, "loss": 0.0622, "step": 36820 }, { "grad_norm": 0.3619651198387146, "learning_rate": 7.476718818093944e-05, "loss": 0.049, "step": 36830 }, { "grad_norm": 0.6678715944290161, "learning_rate": 7.475282318350527e-05, "loss": 0.0549, "step": 36840 }, { "grad_norm": 0.5825868844985962, "learning_rate": 7.473845547913957e-05, "loss": 0.0508, "step": 36850 }, { "grad_norm": 0.37361302971839905, "learning_rate": 7.472408506941357e-05, "loss": 0.0485, "step": 36860 }, { "grad_norm": 0.591564953327179, "learning_rate": 7.470971195589881e-05, "loss": 0.0457, "step": 36870 }, { "grad_norm": 0.7710236310958862, "learning_rate": 7.46953361401671e-05, "loss": 0.0456, "step": 36880 }, { "grad_norm": 0.4775368273258209, "learning_rate": 7.468095762379055e-05, "loss": 0.0491, "step": 36890 }, { "grad_norm": 0.692466139793396, "learning_rate": 7.466657640834158e-05, "loss": 0.0717, "step": 36900 }, { "grad_norm": 0.520477831363678, "learning_rate": 7.46521924953929e-05, "loss": 0.0401, "step": 36910 }, { "grad_norm": 0.5586643815040588, "learning_rate": 7.463780588651752e-05, "loss": 0.0455, "step": 36920 }, { "grad_norm": 0.4850374758243561, "learning_rate": 7.462341658328873e-05, "loss": 0.0502, "step": 36930 }, { "grad_norm": 0.7914818525314331, "learning_rate": 7.460902458728012e-05, "loss": 0.0661, "step": 36940 }, { "grad_norm": 0.5099671483039856, "learning_rate": 7.459462990006558e-05, "loss": 0.0566, "step": 36950 }, { "grad_norm": 0.4548477530479431, "learning_rate": 7.458023252321929e-05, "loss": 0.0462, "step": 36960 }, { "grad_norm": 0.7166053652763367, "learning_rate": 7.456583245831573e-05, "loss": 0.0505, "step": 36970 }, { "grad_norm": 0.6676123738288879, "learning_rate": 7.455142970692968e-05, "loss": 0.0497, "step": 36980 }, { "grad_norm": 0.49561429023742676, "learning_rate": 7.453702427063615e-05, "loss": 0.059, "step": 36990 }, { "grad_norm": 0.5325804352760315, "learning_rate": 7.452261615101057e-05, "loss": 0.0559, "step": 37000 }, { "grad_norm": 0.7258409857749939, "learning_rate": 7.450820534962856e-05, "loss": 0.0578, "step": 37010 }, { "grad_norm": 0.3735073208808899, "learning_rate": 7.449379186806604e-05, "loss": 0.041, "step": 37020 }, { "grad_norm": 0.5195798873901367, "learning_rate": 7.44793757078993e-05, "loss": 0.0438, "step": 37030 }, { "grad_norm": 0.7822752594947815, "learning_rate": 7.44649568707048e-05, "loss": 0.048, "step": 37040 }, { "grad_norm": 0.24939318001270294, "learning_rate": 7.445053535805942e-05, "loss": 0.0391, "step": 37050 }, { "grad_norm": 0.5470792651176453, "learning_rate": 7.443611117154024e-05, "loss": 0.0421, "step": 37060 }, { "grad_norm": 0.3478478491306305, "learning_rate": 7.442168431272469e-05, "loss": 0.0436, "step": 37070 }, { "grad_norm": 0.4224545955657959, "learning_rate": 7.440725478319044e-05, "loss": 0.0375, "step": 37080 }, { "grad_norm": 0.5149595141410828, "learning_rate": 7.439282258451553e-05, "loss": 0.0437, "step": 37090 }, { "grad_norm": 0.571661651134491, "learning_rate": 7.43783877182782e-05, "loss": 0.0407, "step": 37100 }, { "grad_norm": 0.43452972173690796, "learning_rate": 7.436395018605707e-05, "loss": 0.0525, "step": 37110 }, { "grad_norm": 0.6391600370407104, "learning_rate": 7.434950998943094e-05, "loss": 0.0609, "step": 37120 }, { "grad_norm": 0.5740234851837158, "learning_rate": 7.433506712997902e-05, "loss": 0.0457, "step": 37130 }, { "grad_norm": 0.7776255011558533, "learning_rate": 7.432062160928077e-05, "loss": 0.0446, "step": 37140 }, { "grad_norm": 0.4311409294605255, "learning_rate": 7.430617342891588e-05, "loss": 0.0312, "step": 37150 }, { "grad_norm": 0.49627214670181274, "learning_rate": 7.429172259046442e-05, "loss": 0.0478, "step": 37160 }, { "grad_norm": 0.3506421744823456, "learning_rate": 7.42772690955067e-05, "loss": 0.0381, "step": 37170 }, { "grad_norm": 0.49575987458229065, "learning_rate": 7.426281294562332e-05, "loss": 0.0359, "step": 37180 }, { "grad_norm": 0.7124703526496887, "learning_rate": 7.424835414239519e-05, "loss": 0.0585, "step": 37190 }, { "grad_norm": 0.37932777404785156, "learning_rate": 7.423389268740352e-05, "loss": 0.0578, "step": 37200 }, { "grad_norm": 0.5810487270355225, "learning_rate": 7.42194285822298e-05, "loss": 0.0453, "step": 37210 }, { "grad_norm": 0.5394600629806519, "learning_rate": 7.420496182845576e-05, "loss": 0.0418, "step": 37220 }, { "grad_norm": 0.6420701146125793, "learning_rate": 7.419049242766348e-05, "loss": 0.043, "step": 37230 }, { "grad_norm": 0.420746773481369, "learning_rate": 7.417602038143533e-05, "loss": 0.0737, "step": 37240 }, { "grad_norm": 0.7382028102874756, "learning_rate": 7.416154569135393e-05, "loss": 0.0844, "step": 37250 }, { "grad_norm": 0.8399494290351868, "learning_rate": 7.414706835900222e-05, "loss": 0.084, "step": 37260 }, { "grad_norm": 0.5306681990623474, "learning_rate": 7.413258838596345e-05, "loss": 0.046, "step": 37270 }, { "grad_norm": 0.44950324296951294, "learning_rate": 7.411810577382106e-05, "loss": 0.0506, "step": 37280 }, { "grad_norm": 0.8484233021736145, "learning_rate": 7.410362052415889e-05, "loss": 0.0545, "step": 37290 }, { "grad_norm": 0.4943585693836212, "learning_rate": 7.408913263856102e-05, "loss": 0.0374, "step": 37300 }, { "grad_norm": 0.38751816749572754, "learning_rate": 7.407464211861183e-05, "loss": 0.0482, "step": 37310 }, { "grad_norm": 0.5573400855064392, "learning_rate": 7.406014896589597e-05, "loss": 0.0502, "step": 37320 }, { "grad_norm": 0.3040142357349396, "learning_rate": 7.404565318199838e-05, "loss": 0.0456, "step": 37330 }, { "grad_norm": 0.48362234234809875, "learning_rate": 7.403115476850432e-05, "loss": 0.0438, "step": 37340 }, { "grad_norm": 0.5370354652404785, "learning_rate": 7.401665372699932e-05, "loss": 0.0435, "step": 37350 }, { "grad_norm": 0.5391585230827332, "learning_rate": 7.400215005906916e-05, "loss": 0.0492, "step": 37360 }, { "grad_norm": 0.5764718055725098, "learning_rate": 7.398764376629996e-05, "loss": 0.0453, "step": 37370 }, { "grad_norm": 0.6777543425559998, "learning_rate": 7.39731348502781e-05, "loss": 0.043, "step": 37380 }, { "grad_norm": 0.6832917928695679, "learning_rate": 7.395862331259028e-05, "loss": 0.0385, "step": 37390 }, { "grad_norm": 0.5661066174507141, "learning_rate": 7.394410915482341e-05, "loss": 0.0512, "step": 37400 }, { "grad_norm": 0.3911260962486267, "learning_rate": 7.392959237856476e-05, "loss": 0.0394, "step": 37410 }, { "grad_norm": 0.4938988983631134, "learning_rate": 7.391507298540187e-05, "loss": 0.0485, "step": 37420 }, { "grad_norm": 0.5136246085166931, "learning_rate": 7.390055097692257e-05, "loss": 0.0457, "step": 37430 }, { "grad_norm": 0.5597466826438904, "learning_rate": 7.388602635471494e-05, "loss": 0.0421, "step": 37440 }, { "grad_norm": 0.7054430842399597, "learning_rate": 7.387149912036738e-05, "loss": 0.0792, "step": 37450 }, { "grad_norm": 0.721835732460022, "learning_rate": 7.385696927546854e-05, "loss": 0.0596, "step": 37460 }, { "grad_norm": 0.570965051651001, "learning_rate": 7.384243682160743e-05, "loss": 0.0411, "step": 37470 }, { "grad_norm": 0.44021111726760864, "learning_rate": 7.382790176037329e-05, "loss": 0.0432, "step": 37480 }, { "grad_norm": 0.49982213973999023, "learning_rate": 7.381336409335562e-05, "loss": 0.0416, "step": 37490 }, { "grad_norm": 0.8467102646827698, "learning_rate": 7.379882382214426e-05, "loss": 0.0456, "step": 37500 }, { "grad_norm": 0.3856985569000244, "learning_rate": 7.378428094832931e-05, "loss": 0.0406, "step": 37510 }, { "grad_norm": 0.6999778747558594, "learning_rate": 7.376973547350114e-05, "loss": 0.0511, "step": 37520 }, { "grad_norm": 0.7223080396652222, "learning_rate": 7.375518739925046e-05, "loss": 0.0641, "step": 37530 }, { "grad_norm": 0.38546687364578247, "learning_rate": 7.374063672716818e-05, "loss": 0.0663, "step": 37540 }, { "grad_norm": 0.6033527255058289, "learning_rate": 7.372608345884558e-05, "loss": 0.043, "step": 37550 }, { "grad_norm": 0.41177898645401, "learning_rate": 7.371152759587415e-05, "loss": 0.0668, "step": 37560 }, { "grad_norm": 0.48843032121658325, "learning_rate": 7.369696913984572e-05, "loss": 0.0436, "step": 37570 }, { "grad_norm": 0.7193530797958374, "learning_rate": 7.368240809235237e-05, "loss": 0.0474, "step": 37580 }, { "grad_norm": 0.7511984705924988, "learning_rate": 7.366784445498649e-05, "loss": 0.052, "step": 37590 }, { "grad_norm": 0.6503974199295044, "learning_rate": 7.36532782293407e-05, "loss": 0.0448, "step": 37600 }, { "grad_norm": 0.42944830656051636, "learning_rate": 7.363870941700797e-05, "loss": 0.0505, "step": 37610 }, { "grad_norm": 0.7843252420425415, "learning_rate": 7.362413801958152e-05, "loss": 0.0488, "step": 37620 }, { "grad_norm": 0.5698833465576172, "learning_rate": 7.360956403865486e-05, "loss": 0.0523, "step": 37630 }, { "grad_norm": 0.5462479591369629, "learning_rate": 7.359498747582177e-05, "loss": 0.0449, "step": 37640 }, { "grad_norm": 0.47928082942962646, "learning_rate": 7.358040833267634e-05, "loss": 0.0457, "step": 37650 }, { "grad_norm": 0.4923282563686371, "learning_rate": 7.356582661081288e-05, "loss": 0.0484, "step": 37660 }, { "grad_norm": 0.3323519229888916, "learning_rate": 7.355124231182607e-05, "loss": 0.0415, "step": 37670 }, { "grad_norm": 0.5327914953231812, "learning_rate": 7.353665543731083e-05, "loss": 0.0413, "step": 37680 }, { "grad_norm": 0.855122983455658, "learning_rate": 7.352206598886229e-05, "loss": 0.0492, "step": 37690 }, { "grad_norm": 0.5175674557685852, "learning_rate": 7.350747396807601e-05, "loss": 0.0441, "step": 37700 }, { "grad_norm": 0.9249626994132996, "learning_rate": 7.349287937654772e-05, "loss": 0.054, "step": 37710 }, { "grad_norm": 0.49577847123146057, "learning_rate": 7.347828221587345e-05, "loss": 0.0478, "step": 37720 }, { "grad_norm": 1.0142900943756104, "learning_rate": 7.346368248764955e-05, "loss": 0.0472, "step": 37730 }, { "grad_norm": 0.6254013776779175, "learning_rate": 7.34490801934726e-05, "loss": 0.051, "step": 37740 }, { "grad_norm": 0.5673099756240845, "learning_rate": 7.343447533493947e-05, "loss": 0.0447, "step": 37750 }, { "grad_norm": 0.3704298734664917, "learning_rate": 7.341986791364739e-05, "loss": 0.0443, "step": 37760 }, { "grad_norm": 0.399148553609848, "learning_rate": 7.340525793119374e-05, "loss": 0.042, "step": 37770 }, { "grad_norm": 0.38722777366638184, "learning_rate": 7.33906453891763e-05, "loss": 0.0412, "step": 37780 }, { "grad_norm": 0.5071077942848206, "learning_rate": 7.337603028919303e-05, "loss": 0.0489, "step": 37790 }, { "grad_norm": 0.3932492136955261, "learning_rate": 7.336141263284225e-05, "loss": 0.0671, "step": 37800 }, { "grad_norm": 0.8127312660217285, "learning_rate": 7.334679242172249e-05, "loss": 0.0528, "step": 37810 }, { "grad_norm": 0.40669649839401245, "learning_rate": 7.33321696574326e-05, "loss": 0.0429, "step": 37820 }, { "grad_norm": 0.49160751700401306, "learning_rate": 7.331754434157174e-05, "loss": 0.0461, "step": 37830 }, { "grad_norm": 1.0743194818496704, "learning_rate": 7.330291647573928e-05, "loss": 0.049, "step": 37840 }, { "grad_norm": 0.4336630403995514, "learning_rate": 7.32882860615349e-05, "loss": 0.0505, "step": 37850 }, { "grad_norm": 0.7185769081115723, "learning_rate": 7.327365310055858e-05, "loss": 0.0476, "step": 37860 }, { "grad_norm": 0.4979609549045563, "learning_rate": 7.325901759441055e-05, "loss": 0.0431, "step": 37870 }, { "grad_norm": 0.5426579713821411, "learning_rate": 7.324437954469132e-05, "loss": 0.0404, "step": 37880 }, { "grad_norm": 1.0885443687438965, "learning_rate": 7.32297389530017e-05, "loss": 0.0391, "step": 37890 }, { "grad_norm": 0.9076901078224182, "learning_rate": 7.321509582094274e-05, "loss": 0.0463, "step": 37900 }, { "grad_norm": 0.4595930576324463, "learning_rate": 7.320045015011582e-05, "loss": 0.0371, "step": 37910 }, { "grad_norm": 0.6600121259689331, "learning_rate": 7.318580194212253e-05, "loss": 0.063, "step": 37920 }, { "grad_norm": 0.5458217263221741, "learning_rate": 7.317115119856483e-05, "loss": 0.0449, "step": 37930 }, { "grad_norm": 0.4924132525920868, "learning_rate": 7.315649792104485e-05, "loss": 0.0507, "step": 37940 }, { "grad_norm": 0.6033317446708679, "learning_rate": 7.314184211116507e-05, "loss": 0.0472, "step": 37950 }, { "grad_norm": 0.5686862468719482, "learning_rate": 7.312718377052823e-05, "loss": 0.0445, "step": 37960 }, { "grad_norm": 0.5774927735328674, "learning_rate": 7.311252290073737e-05, "loss": 0.0389, "step": 37970 }, { "grad_norm": 0.6024919748306274, "learning_rate": 7.309785950339572e-05, "loss": 0.045, "step": 37980 }, { "grad_norm": 0.786126971244812, "learning_rate": 7.308319358010692e-05, "loss": 0.0474, "step": 37990 }, { "grad_norm": 0.5731559991836548, "learning_rate": 7.306852513247474e-05, "loss": 0.0689, "step": 38000 }, { "grad_norm": 0.6418901681900024, "learning_rate": 7.305385416210336e-05, "loss": 0.0572, "step": 38010 }, { "grad_norm": 0.9440637230873108, "learning_rate": 7.303918067059711e-05, "loss": 0.0385, "step": 38020 }, { "grad_norm": 1.383899211883545, "learning_rate": 7.302450465956075e-05, "loss": 0.0447, "step": 38030 }, { "grad_norm": 0.6057636737823486, "learning_rate": 7.300982613059914e-05, "loss": 0.0399, "step": 38040 }, { "grad_norm": 0.6825888752937317, "learning_rate": 7.299514508531757e-05, "loss": 0.0358, "step": 38050 }, { "grad_norm": 0.5339831113815308, "learning_rate": 7.298046152532148e-05, "loss": 0.0392, "step": 38060 }, { "grad_norm": 0.7462372183799744, "learning_rate": 7.296577545221668e-05, "loss": 0.0589, "step": 38070 }, { "grad_norm": 0.46715518832206726, "learning_rate": 7.29510868676092e-05, "loss": 0.0431, "step": 38080 }, { "grad_norm": 0.691275954246521, "learning_rate": 7.293639577310538e-05, "loss": 0.0435, "step": 38090 }, { "grad_norm": 0.295631468296051, "learning_rate": 7.292170217031179e-05, "loss": 0.0497, "step": 38100 }, { "grad_norm": 0.33774876594543457, "learning_rate": 7.290700606083532e-05, "loss": 0.0464, "step": 38110 }, { "grad_norm": 0.6689620614051819, "learning_rate": 7.28923074462831e-05, "loss": 0.0525, "step": 38120 }, { "grad_norm": 0.5255529880523682, "learning_rate": 7.287760632826255e-05, "loss": 0.0506, "step": 38130 }, { "grad_norm": 0.8234825134277344, "learning_rate": 7.286290270838139e-05, "loss": 0.0477, "step": 38140 }, { "grad_norm": 0.7323631048202515, "learning_rate": 7.284819658824756e-05, "loss": 0.036, "step": 38150 }, { "grad_norm": 0.5687479376792908, "learning_rate": 7.283348796946929e-05, "loss": 0.0513, "step": 38160 }, { "grad_norm": 0.5487955808639526, "learning_rate": 7.28187768536551e-05, "loss": 0.0564, "step": 38170 }, { "grad_norm": 0.49299347400665283, "learning_rate": 7.280406324241379e-05, "loss": 0.0421, "step": 38180 }, { "grad_norm": 0.48399221897125244, "learning_rate": 7.278934713735441e-05, "loss": 0.0734, "step": 38190 }, { "grad_norm": 0.6726310849189758, "learning_rate": 7.277462854008629e-05, "loss": 0.0424, "step": 38200 }, { "grad_norm": 0.7080036997795105, "learning_rate": 7.275990745221904e-05, "loss": 0.0346, "step": 38210 }, { "grad_norm": 0.47733110189437866, "learning_rate": 7.274518387536252e-05, "loss": 0.0478, "step": 38220 }, { "grad_norm": 0.6191864609718323, "learning_rate": 7.273045781112689e-05, "loss": 0.0641, "step": 38230 }, { "grad_norm": 0.5905336141586304, "learning_rate": 7.271572926112257e-05, "loss": 0.0444, "step": 38240 }, { "grad_norm": 0.5177146792411804, "learning_rate": 7.270099822696024e-05, "loss": 0.0457, "step": 38250 }, { "grad_norm": 0.489498496055603, "learning_rate": 7.26862647102509e-05, "loss": 0.0668, "step": 38260 }, { "grad_norm": 0.594322681427002, "learning_rate": 7.267152871260573e-05, "loss": 0.0462, "step": 38270 }, { "grad_norm": 0.4830127954483032, "learning_rate": 7.265679023563629e-05, "loss": 0.0393, "step": 38280 }, { "grad_norm": 0.5042250752449036, "learning_rate": 7.264204928095433e-05, "loss": 0.0372, "step": 38290 }, { "grad_norm": 0.5048248767852783, "learning_rate": 7.262730585017188e-05, "loss": 0.0394, "step": 38300 }, { "grad_norm": 0.3974741995334625, "learning_rate": 7.261255994490131e-05, "loss": 0.0421, "step": 38310 }, { "grad_norm": 0.7011465430259705, "learning_rate": 7.259781156675517e-05, "loss": 0.0431, "step": 38320 }, { "grad_norm": 0.40396466851234436, "learning_rate": 7.258306071734633e-05, "loss": 0.0386, "step": 38330 }, { "grad_norm": 0.6361738443374634, "learning_rate": 7.256830739828792e-05, "loss": 0.053, "step": 38340 }, { "grad_norm": 0.8591145277023315, "learning_rate": 7.255355161119336e-05, "loss": 0.0587, "step": 38350 }, { "grad_norm": 0.5891491174697876, "learning_rate": 7.253879335767628e-05, "loss": 0.0533, "step": 38360 }, { "grad_norm": 0.7207721471786499, "learning_rate": 7.252403263935067e-05, "loss": 0.0834, "step": 38370 }, { "grad_norm": 0.5810479521751404, "learning_rate": 7.25092694578307e-05, "loss": 0.047, "step": 38380 }, { "grad_norm": 0.4705394506454468, "learning_rate": 7.249450381473087e-05, "loss": 0.0428, "step": 38390 }, { "grad_norm": 0.5035507082939148, "learning_rate": 7.247973571166593e-05, "loss": 0.0647, "step": 38400 }, { "grad_norm": 1.0852159261703491, "learning_rate": 7.246496515025089e-05, "loss": 0.0502, "step": 38410 }, { "grad_norm": 0.7128679752349854, "learning_rate": 7.245019213210101e-05, "loss": 0.0397, "step": 38420 }, { "grad_norm": 1.7290593385696411, "learning_rate": 7.24354166588319e-05, "loss": 0.0486, "step": 38430 }, { "grad_norm": 0.48496586084365845, "learning_rate": 7.242063873205935e-05, "loss": 0.0493, "step": 38440 }, { "grad_norm": 0.6177050471305847, "learning_rate": 7.240585835339946e-05, "loss": 0.0482, "step": 38450 }, { "grad_norm": 0.53780597448349, "learning_rate": 7.239107552446857e-05, "loss": 0.0622, "step": 38460 }, { "grad_norm": 0.571123480796814, "learning_rate": 7.237629024688334e-05, "loss": 0.0528, "step": 38470 }, { "grad_norm": 0.5448997616767883, "learning_rate": 7.236150252226064e-05, "loss": 0.0616, "step": 38480 }, { "grad_norm": 0.5214797854423523, "learning_rate": 7.234671235221765e-05, "loss": 0.051, "step": 38490 }, { "grad_norm": 0.9420903325080872, "learning_rate": 7.233191973837179e-05, "loss": 0.0364, "step": 38500 }, { "grad_norm": 0.5452425479888916, "learning_rate": 7.231712468234078e-05, "loss": 0.0447, "step": 38510 }, { "grad_norm": 0.555502712726593, "learning_rate": 7.230232718574254e-05, "loss": 0.0558, "step": 38520 }, { "grad_norm": 0.42422062158584595, "learning_rate": 7.228752725019535e-05, "loss": 0.0356, "step": 38530 }, { "grad_norm": 0.5794813632965088, "learning_rate": 7.227272487731769e-05, "loss": 0.0383, "step": 38540 }, { "grad_norm": 1.4423596858978271, "learning_rate": 7.225792006872831e-05, "loss": 0.0557, "step": 38550 }, { "grad_norm": 0.6624144315719604, "learning_rate": 7.224311282604628e-05, "loss": 0.0377, "step": 38560 }, { "grad_norm": 0.4705948531627655, "learning_rate": 7.222830315089085e-05, "loss": 0.0592, "step": 38570 }, { "grad_norm": 0.6501269340515137, "learning_rate": 7.22134910448816e-05, "loss": 0.0608, "step": 38580 }, { "grad_norm": 0.5939269065856934, "learning_rate": 7.219867650963839e-05, "loss": 0.0403, "step": 38590 }, { "grad_norm": 0.5868947505950928, "learning_rate": 7.21838595467813e-05, "loss": 0.0545, "step": 38600 }, { "grad_norm": 0.5065101385116577, "learning_rate": 7.216904015793067e-05, "loss": 0.0435, "step": 38610 }, { "grad_norm": 0.5563296675682068, "learning_rate": 7.215421834470713e-05, "loss": 0.0505, "step": 38620 }, { "grad_norm": 0.5763861536979675, "learning_rate": 7.213939410873158e-05, "loss": 0.0538, "step": 38630 }, { "grad_norm": 0.6494434475898743, "learning_rate": 7.212456745162519e-05, "loss": 0.0553, "step": 38640 }, { "grad_norm": 0.4540240168571472, "learning_rate": 7.210973837500937e-05, "loss": 0.0519, "step": 38650 }, { "grad_norm": 0.6302247047424316, "learning_rate": 7.209490688050578e-05, "loss": 0.0765, "step": 38660 }, { "grad_norm": 0.5527412295341492, "learning_rate": 7.208007296973641e-05, "loss": 0.0486, "step": 38670 }, { "grad_norm": 0.8432797193527222, "learning_rate": 7.206523664432345e-05, "loss": 0.0916, "step": 38680 }, { "grad_norm": 0.4887544512748718, "learning_rate": 7.205039790588939e-05, "loss": 0.0665, "step": 38690 }, { "grad_norm": 0.4333454966545105, "learning_rate": 7.203555675605697e-05, "loss": 0.0718, "step": 38700 }, { "grad_norm": 0.5639246106147766, "learning_rate": 7.202071319644917e-05, "loss": 0.0421, "step": 38710 }, { "grad_norm": 0.6270932555198669, "learning_rate": 7.200586722868932e-05, "loss": 0.0354, "step": 38720 }, { "grad_norm": 0.6343622207641602, "learning_rate": 7.199101885440088e-05, "loss": 0.0506, "step": 38730 }, { "grad_norm": 0.5097047090530396, "learning_rate": 7.197616807520771e-05, "loss": 0.0457, "step": 38740 }, { "grad_norm": 0.492422491312027, "learning_rate": 7.196131489273381e-05, "loss": 0.035, "step": 38750 }, { "grad_norm": 0.3513105511665344, "learning_rate": 7.194645930860355e-05, "loss": 0.0613, "step": 38760 }, { "grad_norm": 0.6531484127044678, "learning_rate": 7.19316013244415e-05, "loss": 0.0414, "step": 38770 }, { "grad_norm": 0.516596794128418, "learning_rate": 7.191674094187248e-05, "loss": 0.0503, "step": 38780 }, { "grad_norm": 0.5788313150405884, "learning_rate": 7.190187816252165e-05, "loss": 0.034, "step": 38790 }, { "grad_norm": 0.5773788094520569, "learning_rate": 7.188701298801435e-05, "loss": 0.0495, "step": 38800 }, { "grad_norm": 0.6485369205474854, "learning_rate": 7.18721454199762e-05, "loss": 0.0381, "step": 38810 }, { "grad_norm": 0.8252289891242981, "learning_rate": 7.185727546003312e-05, "loss": 0.0447, "step": 38820 }, { "grad_norm": 2.58488130569458, "learning_rate": 7.184240310981126e-05, "loss": 0.0506, "step": 38830 }, { "grad_norm": 0.7923653721809387, "learning_rate": 7.182752837093704e-05, "loss": 0.0392, "step": 38840 }, { "grad_norm": 0.46430352330207825, "learning_rate": 7.181265124503711e-05, "loss": 0.0422, "step": 38850 }, { "grad_norm": 1.2343676090240479, "learning_rate": 7.179777173373847e-05, "loss": 0.0822, "step": 38860 }, { "grad_norm": 0.4656412899494171, "learning_rate": 7.178288983866826e-05, "loss": 0.0446, "step": 38870 }, { "grad_norm": 0.7966891527175903, "learning_rate": 7.176800556145397e-05, "loss": 0.0507, "step": 38880 }, { "grad_norm": 0.5794382691383362, "learning_rate": 7.175311890372334e-05, "loss": 0.0479, "step": 38890 }, { "grad_norm": 0.5338435769081116, "learning_rate": 7.17382298671043e-05, "loss": 0.0468, "step": 38900 }, { "grad_norm": 0.5502517223358154, "learning_rate": 7.172333845322515e-05, "loss": 0.0487, "step": 38910 }, { "grad_norm": 0.3867968022823334, "learning_rate": 7.170844466371436e-05, "loss": 0.0372, "step": 38920 }, { "grad_norm": 0.6406056880950928, "learning_rate": 7.16935485002007e-05, "loss": 0.0553, "step": 38930 }, { "grad_norm": 0.6046006679534912, "learning_rate": 7.167864996431319e-05, "loss": 0.0509, "step": 38940 }, { "grad_norm": 0.6533076763153076, "learning_rate": 7.166374905768111e-05, "loss": 0.0402, "step": 38950 }, { "grad_norm": 0.48951947689056396, "learning_rate": 7.164884578193401e-05, "loss": 0.0358, "step": 38960 }, { "grad_norm": 1.0425186157226562, "learning_rate": 7.163394013870168e-05, "loss": 0.0508, "step": 38970 }, { "grad_norm": 0.44805917143821716, "learning_rate": 7.161903212961419e-05, "loss": 0.0604, "step": 38980 }, { "grad_norm": 0.5754992961883545, "learning_rate": 7.160412175630183e-05, "loss": 0.0431, "step": 38990 }, { "grad_norm": 0.5665168762207031, "learning_rate": 7.158920902039521e-05, "loss": 0.0406, "step": 39000 }, { "grad_norm": 0.4708809554576874, "learning_rate": 7.157429392352514e-05, "loss": 0.061, "step": 39010 }, { "grad_norm": 0.5567655563354492, "learning_rate": 7.155937646732275e-05, "loss": 0.055, "step": 39020 }, { "grad_norm": 0.45248207449913025, "learning_rate": 7.154445665341933e-05, "loss": 0.0365, "step": 39030 }, { "grad_norm": 0.5187005996704102, "learning_rate": 7.152953448344654e-05, "loss": 0.0444, "step": 39040 }, { "grad_norm": 0.8570083975791931, "learning_rate": 7.151460995903624e-05, "loss": 0.0454, "step": 39050 }, { "grad_norm": 1.0851387977600098, "learning_rate": 7.149968308182052e-05, "loss": 0.0424, "step": 39060 }, { "grad_norm": 0.4541621208190918, "learning_rate": 7.14847538534318e-05, "loss": 0.0382, "step": 39070 }, { "grad_norm": 0.5894619226455688, "learning_rate": 7.14698222755027e-05, "loss": 0.0479, "step": 39080 }, { "grad_norm": 1.0807710886001587, "learning_rate": 7.14548883496661e-05, "loss": 0.0507, "step": 39090 }, { "grad_norm": 0.506633996963501, "learning_rate": 7.143995207755517e-05, "loss": 0.0514, "step": 39100 }, { "grad_norm": 0.5656330585479736, "learning_rate": 7.142501346080333e-05, "loss": 0.0696, "step": 39110 }, { "grad_norm": 0.5790734887123108, "learning_rate": 7.141007250104421e-05, "loss": 0.0571, "step": 39120 }, { "grad_norm": 0.7883449196815491, "learning_rate": 7.139512919991176e-05, "loss": 0.0528, "step": 39130 }, { "grad_norm": 0.44900164008140564, "learning_rate": 7.138018355904014e-05, "loss": 0.0419, "step": 39140 }, { "grad_norm": 0.6532371640205383, "learning_rate": 7.13652355800638e-05, "loss": 0.0558, "step": 39150 }, { "grad_norm": 0.4168407618999481, "learning_rate": 7.135028526461743e-05, "loss": 0.0411, "step": 39160 }, { "grad_norm": 0.5015094876289368, "learning_rate": 7.133533261433594e-05, "loss": 0.0439, "step": 39170 }, { "grad_norm": 1.1165250539779663, "learning_rate": 7.132037763085457e-05, "loss": 0.0439, "step": 39180 }, { "grad_norm": 0.4100339710712433, "learning_rate": 7.130542031580875e-05, "loss": 0.0506, "step": 39190 }, { "grad_norm": 0.6939288973808289, "learning_rate": 7.12904606708342e-05, "loss": 0.0423, "step": 39200 }, { "grad_norm": 0.6767664551734924, "learning_rate": 7.127549869756687e-05, "loss": 0.0485, "step": 39210 }, { "grad_norm": 0.7650458812713623, "learning_rate": 7.126053439764299e-05, "loss": 0.0495, "step": 39220 }, { "grad_norm": 0.7488805651664734, "learning_rate": 7.124556777269904e-05, "loss": 0.0529, "step": 39230 }, { "grad_norm": 0.8304815292358398, "learning_rate": 7.123059882437174e-05, "loss": 0.0426, "step": 39240 }, { "grad_norm": 0.6170242428779602, "learning_rate": 7.121562755429807e-05, "loss": 0.0515, "step": 39250 }, { "grad_norm": 1.1303527355194092, "learning_rate": 7.120065396411528e-05, "loss": 0.0404, "step": 39260 }, { "grad_norm": 0.3411087393760681, "learning_rate": 7.118567805546084e-05, "loss": 0.0376, "step": 39270 }, { "grad_norm": 0.527559220790863, "learning_rate": 7.117069982997248e-05, "loss": 0.0529, "step": 39280 }, { "grad_norm": 0.48049312829971313, "learning_rate": 7.115571928928825e-05, "loss": 0.042, "step": 39290 }, { "grad_norm": 0.9726839661598206, "learning_rate": 7.114073643504635e-05, "loss": 0.0434, "step": 39300 }, { "grad_norm": 0.6862877607345581, "learning_rate": 7.11257512688853e-05, "loss": 0.0532, "step": 39310 }, { "grad_norm": 0.6706361174583435, "learning_rate": 7.111076379244384e-05, "loss": 0.0429, "step": 39320 }, { "grad_norm": 0.6200565695762634, "learning_rate": 7.109577400736101e-05, "loss": 0.0393, "step": 39330 }, { "grad_norm": 0.6307014226913452, "learning_rate": 7.108078191527605e-05, "loss": 0.04, "step": 39340 }, { "grad_norm": 0.9073326587677002, "learning_rate": 7.106578751782847e-05, "loss": 0.0435, "step": 39350 }, { "grad_norm": 0.5177494287490845, "learning_rate": 7.105079081665803e-05, "loss": 0.0467, "step": 39360 }, { "grad_norm": 0.4154543876647949, "learning_rate": 7.103579181340476e-05, "loss": 0.0571, "step": 39370 }, { "grad_norm": 0.46970686316490173, "learning_rate": 7.102079050970893e-05, "loss": 0.0413, "step": 39380 }, { "grad_norm": 0.39937904477119446, "learning_rate": 7.100578690721105e-05, "loss": 0.0537, "step": 39390 }, { "grad_norm": 0.5592450499534607, "learning_rate": 7.09907810075519e-05, "loss": 0.0425, "step": 39400 }, { "grad_norm": 0.49360474944114685, "learning_rate": 7.097577281237249e-05, "loss": 0.0507, "step": 39410 }, { "grad_norm": 0.4716110825538635, "learning_rate": 7.09607623233141e-05, "loss": 0.0468, "step": 39420 }, { "grad_norm": 0.7592474818229675, "learning_rate": 7.094574954201827e-05, "loss": 0.0465, "step": 39430 }, { "grad_norm": 0.9429792761802673, "learning_rate": 7.093073447012675e-05, "loss": 0.054, "step": 39440 }, { "grad_norm": 0.40820804238319397, "learning_rate": 7.09157171092816e-05, "loss": 0.0385, "step": 39450 }, { "grad_norm": 0.510796308517456, "learning_rate": 7.090069746112504e-05, "loss": 0.048, "step": 39460 }, { "grad_norm": 0.6286216974258423, "learning_rate": 7.088567552729965e-05, "loss": 0.045, "step": 39470 }, { "grad_norm": 0.5194845199584961, "learning_rate": 7.087065130944818e-05, "loss": 0.0392, "step": 39480 }, { "grad_norm": 0.464595228433609, "learning_rate": 7.085562480921366e-05, "loss": 0.0403, "step": 39490 }, { "grad_norm": 0.48753589391708374, "learning_rate": 7.084059602823937e-05, "loss": 0.0433, "step": 39500 }, { "grad_norm": 0.5999153256416321, "learning_rate": 7.082556496816882e-05, "loss": 0.046, "step": 39510 }, { "grad_norm": 0.5841097831726074, "learning_rate": 7.081053163064582e-05, "loss": 0.0513, "step": 39520 }, { "grad_norm": 0.6871664524078369, "learning_rate": 7.079549601731434e-05, "loss": 0.0561, "step": 39530 }, { "grad_norm": 0.7213456034660339, "learning_rate": 7.07804581298187e-05, "loss": 0.0406, "step": 39540 }, { "grad_norm": 0.46218207478523254, "learning_rate": 7.07654179698034e-05, "loss": 0.0605, "step": 39550 }, { "grad_norm": 0.4602068066596985, "learning_rate": 7.075037553891321e-05, "loss": 0.0487, "step": 39560 }, { "grad_norm": 0.47536858916282654, "learning_rate": 7.073533083879315e-05, "loss": 0.0501, "step": 39570 }, { "grad_norm": 0.6546916365623474, "learning_rate": 7.072028387108849e-05, "loss": 0.0562, "step": 39580 }, { "grad_norm": 0.3026888370513916, "learning_rate": 7.070523463744472e-05, "loss": 0.04, "step": 39590 }, { "grad_norm": 0.35924771428108215, "learning_rate": 7.069018313950763e-05, "loss": 0.0571, "step": 39600 }, { "grad_norm": 0.485946923494339, "learning_rate": 7.067512937892322e-05, "loss": 0.0445, "step": 39610 }, { "grad_norm": 0.4097702205181122, "learning_rate": 7.066007335733775e-05, "loss": 0.0405, "step": 39620 }, { "grad_norm": 0.43453049659729004, "learning_rate": 7.064501507639772e-05, "loss": 0.063, "step": 39630 }, { "grad_norm": 0.4854421019554138, "learning_rate": 7.062995453774987e-05, "loss": 0.04, "step": 39640 }, { "grad_norm": 0.46419045329093933, "learning_rate": 7.061489174304121e-05, "loss": 0.0364, "step": 39650 }, { "grad_norm": 0.6746165752410889, "learning_rate": 7.0599826693919e-05, "loss": 0.0438, "step": 39660 }, { "grad_norm": 0.4272690415382385, "learning_rate": 7.05847593920307e-05, "loss": 0.0375, "step": 39670 }, { "grad_norm": 0.5221607685089111, "learning_rate": 7.056968983902406e-05, "loss": 0.0432, "step": 39680 }, { "grad_norm": 0.6951592564582825, "learning_rate": 7.055461803654706e-05, "loss": 0.0517, "step": 39690 }, { "grad_norm": 0.9265437126159668, "learning_rate": 7.053954398624794e-05, "loss": 0.0579, "step": 39700 }, { "grad_norm": 0.48147261142730713, "learning_rate": 7.052446768977518e-05, "loss": 0.0357, "step": 39710 }, { "grad_norm": 0.7328177690505981, "learning_rate": 7.050938914877748e-05, "loss": 0.081, "step": 39720 }, { "grad_norm": 0.5476903319358826, "learning_rate": 7.049430836490382e-05, "loss": 0.0409, "step": 39730 }, { "grad_norm": 0.500540018081665, "learning_rate": 7.047922533980342e-05, "loss": 0.054, "step": 39740 }, { "grad_norm": 1.1767349243164062, "learning_rate": 7.046414007512571e-05, "loss": 0.0625, "step": 39750 }, { "grad_norm": 0.6535723209381104, "learning_rate": 7.044905257252042e-05, "loss": 0.0487, "step": 39760 }, { "grad_norm": 0.9151268005371094, "learning_rate": 7.04339628336375e-05, "loss": 0.0549, "step": 39770 }, { "grad_norm": 0.6908924579620361, "learning_rate": 7.041887086012711e-05, "loss": 0.0433, "step": 39780 }, { "grad_norm": 0.35101091861724854, "learning_rate": 7.040377665363969e-05, "loss": 0.0413, "step": 39790 }, { "grad_norm": 0.5185426473617554, "learning_rate": 7.038868021582594e-05, "loss": 0.0787, "step": 39800 }, { "grad_norm": 0.4865982234477997, "learning_rate": 7.037358154833679e-05, "loss": 0.0391, "step": 39810 }, { "grad_norm": 0.5597221255302429, "learning_rate": 7.035848065282339e-05, "loss": 0.0482, "step": 39820 }, { "grad_norm": 0.6176406741142273, "learning_rate": 7.034337753093714e-05, "loss": 0.0395, "step": 39830 }, { "grad_norm": 0.45254746079444885, "learning_rate": 7.032827218432972e-05, "loss": 0.0391, "step": 39840 }, { "grad_norm": 0.8398632407188416, "learning_rate": 7.031316461465302e-05, "loss": 0.0482, "step": 39850 }, { "grad_norm": 1.4439772367477417, "learning_rate": 7.029805482355915e-05, "loss": 0.0455, "step": 39860 }, { "grad_norm": 0.4110097587108612, "learning_rate": 7.028294281270055e-05, "loss": 0.0318, "step": 39870 }, { "grad_norm": 0.7107399702072144, "learning_rate": 7.026782858372982e-05, "loss": 0.0444, "step": 39880 }, { "grad_norm": 0.9574864506721497, "learning_rate": 7.025271213829982e-05, "loss": 0.0405, "step": 39890 }, { "grad_norm": 0.5539586544036865, "learning_rate": 7.023759347806366e-05, "loss": 0.0492, "step": 39900 }, { "grad_norm": 0.4858414828777313, "learning_rate": 7.022247260467472e-05, "loss": 0.0476, "step": 39910 }, { "grad_norm": 0.44449546933174133, "learning_rate": 7.020734951978658e-05, "loss": 0.0444, "step": 39920 }, { "grad_norm": 0.37348541617393494, "learning_rate": 7.019222422505307e-05, "loss": 0.0512, "step": 39930 }, { "grad_norm": 1.2348772287368774, "learning_rate": 7.017709672212827e-05, "loss": 0.0418, "step": 39940 }, { "grad_norm": 0.31962841749191284, "learning_rate": 7.016196701266652e-05, "loss": 0.0389, "step": 39950 }, { "grad_norm": 0.5165780782699585, "learning_rate": 7.014683509832238e-05, "loss": 0.0405, "step": 39960 }, { "grad_norm": 0.5680296421051025, "learning_rate": 7.013170098075063e-05, "loss": 0.0383, "step": 39970 }, { "grad_norm": 0.4553174674510956, "learning_rate": 7.011656466160632e-05, "loss": 0.0398, "step": 39980 }, { "grad_norm": 0.5753946304321289, "learning_rate": 7.010142614254475e-05, "loss": 0.0411, "step": 39990 }, { "grad_norm": 0.527683675289154, "learning_rate": 7.008628542522147e-05, "loss": 0.0553, "step": 40000 }, { "grad_norm": 0.45682913064956665, "learning_rate": 7.007114251129218e-05, "loss": 0.0407, "step": 40010 }, { "grad_norm": 0.6027337908744812, "learning_rate": 7.005599740241293e-05, "loss": 0.0323, "step": 40020 }, { "grad_norm": 1.101852536201477, "learning_rate": 7.004085010023996e-05, "loss": 0.0478, "step": 40030 }, { "grad_norm": 0.5889971852302551, "learning_rate": 7.002570060642976e-05, "loss": 0.0495, "step": 40040 }, { "grad_norm": 0.5987557172775269, "learning_rate": 7.001054892263903e-05, "loss": 0.0456, "step": 40050 }, { "grad_norm": 0.7727826237678528, "learning_rate": 6.999539505052477e-05, "loss": 0.0334, "step": 40060 }, { "grad_norm": 0.5336470603942871, "learning_rate": 6.998023899174418e-05, "loss": 0.0787, "step": 40070 }, { "grad_norm": 0.7072558999061584, "learning_rate": 6.996508074795467e-05, "loss": 0.0496, "step": 40080 }, { "grad_norm": 0.7561469674110413, "learning_rate": 6.994992032081396e-05, "loss": 0.0453, "step": 40090 }, { "grad_norm": 0.3712131381034851, "learning_rate": 6.993475771197995e-05, "loss": 0.0503, "step": 40100 }, { "grad_norm": 0.5914998650550842, "learning_rate": 6.991959292311082e-05, "loss": 0.0434, "step": 40110 }, { "grad_norm": 0.42949461936950684, "learning_rate": 6.990442595586495e-05, "loss": 0.0414, "step": 40120 }, { "grad_norm": 0.38167905807495117, "learning_rate": 6.988925681190098e-05, "loss": 0.0376, "step": 40130 }, { "grad_norm": 0.84131920337677, "learning_rate": 6.987408549287778e-05, "loss": 0.0782, "step": 40140 }, { "grad_norm": 0.6176363229751587, "learning_rate": 6.985891200045449e-05, "loss": 0.0408, "step": 40150 }, { "grad_norm": 0.38016223907470703, "learning_rate": 6.984373633629045e-05, "loss": 0.04, "step": 40160 }, { "grad_norm": 0.5696948766708374, "learning_rate": 6.98285585020452e-05, "loss": 0.0536, "step": 40170 }, { "grad_norm": 0.5123939514160156, "learning_rate": 6.981337849937864e-05, "loss": 0.0408, "step": 40180 }, { "grad_norm": 0.571302056312561, "learning_rate": 6.979819632995078e-05, "loss": 0.0378, "step": 40190 }, { "grad_norm": 0.4403376877307892, "learning_rate": 6.978301199542193e-05, "loss": 0.0461, "step": 40200 }, { "grad_norm": 0.4118785858154297, "learning_rate": 6.976782549745263e-05, "loss": 0.05, "step": 40210 }, { "grad_norm": 0.46436166763305664, "learning_rate": 6.975263683770367e-05, "loss": 0.0388, "step": 40220 }, { "grad_norm": 0.3583533763885498, "learning_rate": 6.973744601783604e-05, "loss": 0.0487, "step": 40230 }, { "grad_norm": 0.5591039657592773, "learning_rate": 6.972225303951097e-05, "loss": 0.0514, "step": 40240 }, { "grad_norm": 0.5262925624847412, "learning_rate": 6.970705790438998e-05, "loss": 0.0504, "step": 40250 }, { "grad_norm": 0.7205265760421753, "learning_rate": 6.969186061413477e-05, "loss": 0.0578, "step": 40260 }, { "grad_norm": 0.5715988874435425, "learning_rate": 6.967666117040727e-05, "loss": 0.0416, "step": 40270 }, { "grad_norm": 0.5945531725883484, "learning_rate": 6.966145957486972e-05, "loss": 0.0453, "step": 40280 }, { "grad_norm": 0.4642985761165619, "learning_rate": 6.964625582918449e-05, "loss": 0.0521, "step": 40290 }, { "grad_norm": 0.39651918411254883, "learning_rate": 6.963104993501425e-05, "loss": 0.0286, "step": 40300 }, { "grad_norm": 0.6178165674209595, "learning_rate": 6.961584189402192e-05, "loss": 0.0539, "step": 40310 }, { "grad_norm": 0.40252116322517395, "learning_rate": 6.96006317078706e-05, "loss": 0.044, "step": 40320 }, { "grad_norm": 0.4192394018173218, "learning_rate": 6.958541937822367e-05, "loss": 0.0611, "step": 40330 }, { "grad_norm": 0.5907796621322632, "learning_rate": 6.957020490674473e-05, "loss": 0.0513, "step": 40340 }, { "grad_norm": 0.5419631004333496, "learning_rate": 6.95549882950976e-05, "loss": 0.0481, "step": 40350 }, { "grad_norm": 0.43746718764305115, "learning_rate": 6.953976954494635e-05, "loss": 0.0485, "step": 40360 }, { "grad_norm": 0.6419054865837097, "learning_rate": 6.952454865795528e-05, "loss": 0.0422, "step": 40370 }, { "grad_norm": 0.6176126003265381, "learning_rate": 6.950932563578892e-05, "loss": 0.0429, "step": 40380 }, { "grad_norm": 0.4712664783000946, "learning_rate": 6.949410048011206e-05, "loss": 0.0381, "step": 40390 }, { "grad_norm": 0.8473241925239563, "learning_rate": 6.947887319258966e-05, "loss": 0.0382, "step": 40400 }, { "grad_norm": 0.23681877553462982, "learning_rate": 6.946364377488696e-05, "loss": 0.0349, "step": 40410 }, { "grad_norm": 0.6248292326927185, "learning_rate": 6.944841222866947e-05, "loss": 0.0402, "step": 40420 }, { "grad_norm": 0.320196270942688, "learning_rate": 6.943317855560284e-05, "loss": 0.0398, "step": 40430 }, { "grad_norm": 0.41711974143981934, "learning_rate": 6.941794275735302e-05, "loss": 0.036, "step": 40440 }, { "grad_norm": 0.5209223031997681, "learning_rate": 6.94027048355862e-05, "loss": 0.0311, "step": 40450 }, { "grad_norm": 0.8223759531974792, "learning_rate": 6.938746479196871e-05, "loss": 0.0621, "step": 40460 }, { "grad_norm": 0.5254886746406555, "learning_rate": 6.937222262816724e-05, "loss": 0.032, "step": 40470 }, { "grad_norm": 0.5257689952850342, "learning_rate": 6.935697834584865e-05, "loss": 0.0465, "step": 40480 }, { "grad_norm": 0.4229801297187805, "learning_rate": 6.934173194667998e-05, "loss": 0.0389, "step": 40490 }, { "grad_norm": 0.4018251299858093, "learning_rate": 6.93264834323286e-05, "loss": 0.0419, "step": 40500 }, { "grad_norm": 0.5766407251358032, "learning_rate": 6.931123280446206e-05, "loss": 0.0378, "step": 40510 }, { "grad_norm": 0.5663790702819824, "learning_rate": 6.929598006474811e-05, "loss": 0.0589, "step": 40520 }, { "grad_norm": 0.5319030284881592, "learning_rate": 6.92807252148548e-05, "loss": 0.0396, "step": 40530 }, { "grad_norm": 0.4867230951786041, "learning_rate": 6.926546825645037e-05, "loss": 0.0499, "step": 40540 }, { "grad_norm": 0.4609411954879761, "learning_rate": 6.92502091912033e-05, "loss": 0.0385, "step": 40550 }, { "grad_norm": 1.1442420482635498, "learning_rate": 6.923494802078229e-05, "loss": 0.0484, "step": 40560 }, { "grad_norm": 0.489590048789978, "learning_rate": 6.921968474685628e-05, "loss": 0.0554, "step": 40570 }, { "grad_norm": 0.7033159136772156, "learning_rate": 6.920441937109445e-05, "loss": 0.0468, "step": 40580 }, { "grad_norm": 0.6389578580856323, "learning_rate": 6.91891518951662e-05, "loss": 0.0378, "step": 40590 }, { "grad_norm": 0.6856451034545898, "learning_rate": 6.917388232074114e-05, "loss": 0.0404, "step": 40600 }, { "grad_norm": 0.5801604986190796, "learning_rate": 6.915861064948914e-05, "loss": 0.0598, "step": 40610 }, { "grad_norm": 0.5038446187973022, "learning_rate": 6.914333688308025e-05, "loss": 0.0418, "step": 40620 }, { "grad_norm": 0.6245686411857605, "learning_rate": 6.912806102318487e-05, "loss": 0.0381, "step": 40630 }, { "grad_norm": 0.5920082926750183, "learning_rate": 6.911278307147347e-05, "loss": 0.0472, "step": 40640 }, { "grad_norm": 0.5502541065216064, "learning_rate": 6.909750302961684e-05, "loss": 0.0388, "step": 40650 }, { "grad_norm": 0.6678645014762878, "learning_rate": 6.9082220899286e-05, "loss": 0.0443, "step": 40660 }, { "grad_norm": 0.6900463700294495, "learning_rate": 6.906693668215219e-05, "loss": 0.0513, "step": 40670 }, { "grad_norm": 0.5128227472305298, "learning_rate": 6.905165037988683e-05, "loss": 0.0487, "step": 40680 }, { "grad_norm": 1.8829165697097778, "learning_rate": 6.903636199416164e-05, "loss": 0.0748, "step": 40690 }, { "grad_norm": 0.6064891219139099, "learning_rate": 6.902107152664851e-05, "loss": 0.0473, "step": 40700 }, { "grad_norm": 0.5450998544692993, "learning_rate": 6.90057789790196e-05, "loss": 0.0464, "step": 40710 }, { "grad_norm": 0.3568786084651947, "learning_rate": 6.899048435294728e-05, "loss": 0.0408, "step": 40720 }, { "grad_norm": 0.5553327202796936, "learning_rate": 6.897518765010415e-05, "loss": 0.0493, "step": 40730 }, { "grad_norm": 1.2444125413894653, "learning_rate": 6.895988887216302e-05, "loss": 0.0519, "step": 40740 }, { "grad_norm": 0.4092993140220642, "learning_rate": 6.894458802079694e-05, "loss": 0.0439, "step": 40750 }, { "grad_norm": 0.47884678840637207, "learning_rate": 6.892928509767922e-05, "loss": 0.0707, "step": 40760 }, { "grad_norm": 0.43306228518486023, "learning_rate": 6.891398010448333e-05, "loss": 0.0432, "step": 40770 }, { "grad_norm": 0.5939105749130249, "learning_rate": 6.8898673042883e-05, "loss": 0.0392, "step": 40780 }, { "grad_norm": 0.5333742499351501, "learning_rate": 6.888336391455222e-05, "loss": 0.0389, "step": 40790 }, { "grad_norm": 0.43435609340667725, "learning_rate": 6.886805272116513e-05, "loss": 0.0388, "step": 40800 }, { "grad_norm": 0.8269054293632507, "learning_rate": 6.885273946439617e-05, "loss": 0.0485, "step": 40810 }, { "grad_norm": 0.7525169849395752, "learning_rate": 6.883742414591998e-05, "loss": 0.0519, "step": 40820 }, { "grad_norm": 0.4070303738117218, "learning_rate": 6.88221067674114e-05, "loss": 0.0374, "step": 40830 }, { "grad_norm": 0.6305745244026184, "learning_rate": 6.88067873305455e-05, "loss": 0.0364, "step": 40840 }, { "grad_norm": 0.6800611019134521, "learning_rate": 6.879146583699765e-05, "loss": 0.0414, "step": 40850 }, { "grad_norm": 0.5766798853874207, "learning_rate": 6.877614228844334e-05, "loss": 0.0364, "step": 40860 }, { "grad_norm": 0.6637930870056152, "learning_rate": 6.876081668655832e-05, "loss": 0.0483, "step": 40870 }, { "grad_norm": 0.3587038218975067, "learning_rate": 6.874548903301861e-05, "loss": 0.0393, "step": 40880 }, { "grad_norm": 0.4294865131378174, "learning_rate": 6.873015932950039e-05, "loss": 0.0466, "step": 40890 }, { "grad_norm": 0.3498381972312927, "learning_rate": 6.871482757768012e-05, "loss": 0.0371, "step": 40900 }, { "grad_norm": 0.6151151061058044, "learning_rate": 6.869949377923442e-05, "loss": 0.0769, "step": 40910 }, { "grad_norm": 0.6070870757102966, "learning_rate": 6.868415793584022e-05, "loss": 0.0632, "step": 40920 }, { "grad_norm": 1.0249640941619873, "learning_rate": 6.86688200491746e-05, "loss": 0.0445, "step": 40930 }, { "grad_norm": 0.4362978935241699, "learning_rate": 6.865348012091488e-05, "loss": 0.0576, "step": 40940 }, { "grad_norm": 0.6560449004173279, "learning_rate": 6.86381381527386e-05, "loss": 0.0375, "step": 40950 }, { "grad_norm": 0.7590524554252625, "learning_rate": 6.862279414632359e-05, "loss": 0.0386, "step": 40960 }, { "grad_norm": 0.39509597420692444, "learning_rate": 6.860744810334778e-05, "loss": 0.0347, "step": 40970 }, { "grad_norm": 0.45228978991508484, "learning_rate": 6.859210002548943e-05, "loss": 0.0387, "step": 40980 }, { "grad_norm": 0.6637909412384033, "learning_rate": 6.8576749914427e-05, "loss": 0.0395, "step": 40990 }, { "grad_norm": 0.586742103099823, "learning_rate": 6.85613977718391e-05, "loss": 0.0518, "step": 41000 }, { "grad_norm": 0.6390001177787781, "learning_rate": 6.854604359940464e-05, "loss": 0.0495, "step": 41010 }, { "grad_norm": 0.9801166653633118, "learning_rate": 6.853068739880276e-05, "loss": 0.0428, "step": 41020 }, { "grad_norm": 0.567476212978363, "learning_rate": 6.851532917171276e-05, "loss": 0.0526, "step": 41030 }, { "grad_norm": 0.9031907916069031, "learning_rate": 6.849996891981418e-05, "loss": 0.0525, "step": 41040 }, { "grad_norm": 0.36689189076423645, "learning_rate": 6.848460664478684e-05, "loss": 0.0391, "step": 41050 }, { "grad_norm": 0.45014429092407227, "learning_rate": 6.846924234831069e-05, "loss": 0.0562, "step": 41060 }, { "grad_norm": 0.3767226040363312, "learning_rate": 6.845387603206597e-05, "loss": 0.0308, "step": 41070 }, { "grad_norm": 0.7338601350784302, "learning_rate": 6.843850769773312e-05, "loss": 0.0295, "step": 41080 }, { "grad_norm": 0.6242253184318542, "learning_rate": 6.842313734699278e-05, "loss": 0.0443, "step": 41090 }, { "grad_norm": 0.5136945843696594, "learning_rate": 6.840776498152584e-05, "loss": 0.0428, "step": 41100 }, { "grad_norm": 0.4054807126522064, "learning_rate": 6.83923906030134e-05, "loss": 0.0371, "step": 41110 }, { "grad_norm": 0.8072311878204346, "learning_rate": 6.837701421313677e-05, "loss": 0.0507, "step": 41120 }, { "grad_norm": 0.5413999557495117, "learning_rate": 6.83616358135775e-05, "loss": 0.0408, "step": 41130 }, { "grad_norm": 0.4976857304573059, "learning_rate": 6.834625540601734e-05, "loss": 0.0317, "step": 41140 }, { "grad_norm": 0.6035829782485962, "learning_rate": 6.833087299213829e-05, "loss": 0.0534, "step": 41150 }, { "grad_norm": 0.4787043631076813, "learning_rate": 6.83154885736225e-05, "loss": 0.0422, "step": 41160 }, { "grad_norm": 0.5900180339813232, "learning_rate": 6.830010215215244e-05, "loss": 0.0567, "step": 41170 }, { "grad_norm": 0.5402771830558777, "learning_rate": 6.828471372941072e-05, "loss": 0.0365, "step": 41180 }, { "grad_norm": 0.8311694264411926, "learning_rate": 6.826932330708018e-05, "loss": 0.048, "step": 41190 }, { "grad_norm": 0.4556559920310974, "learning_rate": 6.825393088684393e-05, "loss": 0.041, "step": 41200 }, { "grad_norm": 0.45824965834617615, "learning_rate": 6.823853647038524e-05, "loss": 0.0323, "step": 41210 }, { "grad_norm": 0.3130991458892822, "learning_rate": 6.822314005938762e-05, "loss": 0.0513, "step": 41220 }, { "grad_norm": 0.2960008382797241, "learning_rate": 6.820774165553479e-05, "loss": 0.0291, "step": 41230 }, { "grad_norm": 0.5045982003211975, "learning_rate": 6.819234126051072e-05, "loss": 0.04, "step": 41240 }, { "grad_norm": 0.5964311361312866, "learning_rate": 6.817693887599956e-05, "loss": 0.0418, "step": 41250 }, { "grad_norm": 0.5947298407554626, "learning_rate": 6.816153450368568e-05, "loss": 0.0494, "step": 41260 }, { "grad_norm": 0.3642325699329376, "learning_rate": 6.814612814525372e-05, "loss": 0.0411, "step": 41270 }, { "grad_norm": 0.3875468373298645, "learning_rate": 6.813071980238846e-05, "loss": 0.0426, "step": 41280 }, { "grad_norm": 0.7499755024909973, "learning_rate": 6.811530947677492e-05, "loss": 0.0387, "step": 41290 }, { "grad_norm": 0.5496392846107483, "learning_rate": 6.809989717009839e-05, "loss": 0.0363, "step": 41300 }, { "grad_norm": 0.4952731430530548, "learning_rate": 6.808448288404431e-05, "loss": 0.0391, "step": 41310 }, { "grad_norm": 0.27739742398262024, "learning_rate": 6.806906662029838e-05, "loss": 0.0429, "step": 41320 }, { "grad_norm": 0.6219310164451599, "learning_rate": 6.805364838054647e-05, "loss": 0.0366, "step": 41330 }, { "grad_norm": 0.5652605891227722, "learning_rate": 6.803822816647474e-05, "loss": 0.0409, "step": 41340 }, { "grad_norm": 0.6853804588317871, "learning_rate": 6.802280597976949e-05, "loss": 0.0496, "step": 41350 }, { "grad_norm": 0.5666864514350891, "learning_rate": 6.800738182211727e-05, "loss": 0.0333, "step": 41360 }, { "grad_norm": 0.6694772839546204, "learning_rate": 6.799195569520485e-05, "loss": 0.0317, "step": 41370 }, { "grad_norm": 0.5874166488647461, "learning_rate": 6.79765276007192e-05, "loss": 0.0477, "step": 41380 }, { "grad_norm": 0.3723728656768799, "learning_rate": 6.796109754034753e-05, "loss": 0.0352, "step": 41390 }, { "grad_norm": 0.3499959111213684, "learning_rate": 6.794566551577724e-05, "loss": 0.0426, "step": 41400 }, { "grad_norm": 0.6782247424125671, "learning_rate": 6.793023152869594e-05, "loss": 0.0364, "step": 41410 }, { "grad_norm": 0.5687814354896545, "learning_rate": 6.791479558079147e-05, "loss": 0.0379, "step": 41420 }, { "grad_norm": 0.4419838786125183, "learning_rate": 6.789935767375191e-05, "loss": 0.0486, "step": 41430 }, { "grad_norm": 0.5305423736572266, "learning_rate": 6.78839178092655e-05, "loss": 0.0417, "step": 41440 }, { "grad_norm": 0.7309492230415344, "learning_rate": 6.786847598902072e-05, "loss": 0.0381, "step": 41450 }, { "grad_norm": 0.6213575005531311, "learning_rate": 6.785303221470627e-05, "loss": 0.0693, "step": 41460 }, { "grad_norm": 0.5194948315620422, "learning_rate": 6.783758648801108e-05, "loss": 0.0403, "step": 41470 }, { "grad_norm": 0.7000337839126587, "learning_rate": 6.782213881062422e-05, "loss": 0.0357, "step": 41480 }, { "grad_norm": 0.4717308282852173, "learning_rate": 6.780668918423508e-05, "loss": 0.0456, "step": 41490 }, { "grad_norm": 0.6473380923271179, "learning_rate": 6.779123761053317e-05, "loss": 0.0445, "step": 41500 }, { "grad_norm": 0.38351336121559143, "learning_rate": 6.777578409120826e-05, "loss": 0.0383, "step": 41510 }, { "grad_norm": 0.5914798378944397, "learning_rate": 6.776032862795033e-05, "loss": 0.0654, "step": 41520 }, { "grad_norm": 0.549963116645813, "learning_rate": 6.774487122244956e-05, "loss": 0.0422, "step": 41530 }, { "grad_norm": 0.5069546103477478, "learning_rate": 6.772941187639637e-05, "loss": 0.0602, "step": 41540 }, { "grad_norm": 0.49091267585754395, "learning_rate": 6.771395059148134e-05, "loss": 0.042, "step": 41550 }, { "grad_norm": 1.0070749521255493, "learning_rate": 6.76984873693953e-05, "loss": 0.0591, "step": 41560 }, { "grad_norm": 0.563797652721405, "learning_rate": 6.768302221182931e-05, "loss": 0.0435, "step": 41570 }, { "grad_norm": 0.5033173561096191, "learning_rate": 6.766755512047457e-05, "loss": 0.0338, "step": 41580 }, { "grad_norm": 0.5109729170799255, "learning_rate": 6.765208609702259e-05, "loss": 0.0526, "step": 41590 }, { "grad_norm": 0.31700512766838074, "learning_rate": 6.763661514316499e-05, "loss": 0.0496, "step": 41600 }, { "grad_norm": 0.7454653382301331, "learning_rate": 6.76211422605937e-05, "loss": 0.0382, "step": 41610 }, { "grad_norm": 0.37637022137641907, "learning_rate": 6.760566745100076e-05, "loss": 0.036, "step": 41620 }, { "grad_norm": 0.6041867136955261, "learning_rate": 6.75901907160785e-05, "loss": 0.0472, "step": 41630 }, { "grad_norm": 0.723632276058197, "learning_rate": 6.757471205751943e-05, "loss": 0.0599, "step": 41640 }, { "grad_norm": 0.7400722503662109, "learning_rate": 6.75592314770163e-05, "loss": 0.0413, "step": 41650 }, { "grad_norm": 0.7064082026481628, "learning_rate": 6.7543748976262e-05, "loss": 0.045, "step": 41660 }, { "grad_norm": 0.43615588545799255, "learning_rate": 6.752826455694968e-05, "loss": 0.0496, "step": 41670 }, { "grad_norm": 0.4688149392604828, "learning_rate": 6.751277822077271e-05, "loss": 0.0466, "step": 41680 }, { "grad_norm": 0.4899342656135559, "learning_rate": 6.749728996942465e-05, "loss": 0.0402, "step": 41690 }, { "grad_norm": 0.5326799750328064, "learning_rate": 6.748179980459924e-05, "loss": 0.0428, "step": 41700 }, { "grad_norm": 0.4633171260356903, "learning_rate": 6.746630772799052e-05, "loss": 0.049, "step": 41710 }, { "grad_norm": 0.35187941789627075, "learning_rate": 6.745081374129262e-05, "loss": 0.0447, "step": 41720 }, { "grad_norm": 0.49085161089897156, "learning_rate": 6.743531784619998e-05, "loss": 0.0445, "step": 41730 }, { "grad_norm": 0.4827888607978821, "learning_rate": 6.741982004440719e-05, "loss": 0.0351, "step": 41740 }, { "grad_norm": 0.37287795543670654, "learning_rate": 6.740432033760907e-05, "loss": 0.0677, "step": 41750 }, { "grad_norm": 0.43339550495147705, "learning_rate": 6.738881872750066e-05, "loss": 0.0353, "step": 41760 }, { "grad_norm": 0.6103796362876892, "learning_rate": 6.737331521577715e-05, "loss": 0.0537, "step": 41770 }, { "grad_norm": 0.4372543394565582, "learning_rate": 6.735780980413403e-05, "loss": 0.042, "step": 41780 }, { "grad_norm": 0.43163514137268066, "learning_rate": 6.734230249426692e-05, "loss": 0.053, "step": 41790 }, { "grad_norm": 0.9581590890884399, "learning_rate": 6.732679328787168e-05, "loss": 0.0504, "step": 41800 }, { "grad_norm": 0.8226542472839355, "learning_rate": 6.731128218664438e-05, "loss": 0.0385, "step": 41810 }, { "grad_norm": 0.7388395071029663, "learning_rate": 6.729576919228129e-05, "loss": 0.0629, "step": 41820 }, { "grad_norm": 0.8921256065368652, "learning_rate": 6.728025430647888e-05, "loss": 0.046, "step": 41830 }, { "grad_norm": 0.6491287350654602, "learning_rate": 6.726473753093383e-05, "loss": 0.042, "step": 41840 }, { "grad_norm": 0.44086983799934387, "learning_rate": 6.724921886734305e-05, "loss": 0.0359, "step": 41850 }, { "grad_norm": 0.4379916489124298, "learning_rate": 6.723369831740363e-05, "loss": 0.0345, "step": 41860 }, { "grad_norm": 0.5606244206428528, "learning_rate": 6.721817588281288e-05, "loss": 0.039, "step": 41870 }, { "grad_norm": 0.39605510234832764, "learning_rate": 6.720265156526828e-05, "loss": 0.0441, "step": 41880 }, { "grad_norm": 0.45130279660224915, "learning_rate": 6.718712536646758e-05, "loss": 0.0358, "step": 41890 }, { "grad_norm": 0.4306517243385315, "learning_rate": 6.71715972881087e-05, "loss": 0.045, "step": 41900 }, { "grad_norm": 0.840776264667511, "learning_rate": 6.715606733188974e-05, "loss": 0.0448, "step": 41910 }, { "grad_norm": 0.5084173679351807, "learning_rate": 6.714053549950909e-05, "loss": 0.0487, "step": 41920 }, { "grad_norm": 0.5319651961326599, "learning_rate": 6.71250017926652e-05, "loss": 0.0418, "step": 41930 }, { "grad_norm": 0.3927677869796753, "learning_rate": 6.71094662130569e-05, "loss": 0.0397, "step": 41940 }, { "grad_norm": 0.5113137364387512, "learning_rate": 6.709392876238307e-05, "loss": 0.0442, "step": 41950 }, { "grad_norm": 2.169240713119507, "learning_rate": 6.707838944234289e-05, "loss": 0.0464, "step": 41960 }, { "grad_norm": 0.8539868593215942, "learning_rate": 6.706284825463574e-05, "loss": 0.0464, "step": 41970 }, { "grad_norm": 0.4061463475227356, "learning_rate": 6.704730520096114e-05, "loss": 0.0393, "step": 41980 }, { "grad_norm": 0.6624677777290344, "learning_rate": 6.703176028301888e-05, "loss": 0.0452, "step": 41990 }, { "grad_norm": 0.4047132134437561, "learning_rate": 6.701621350250892e-05, "loss": 0.0414, "step": 42000 }, { "grad_norm": 0.43298736214637756, "learning_rate": 6.700066486113144e-05, "loss": 0.0546, "step": 42010 }, { "grad_norm": 0.5238221287727356, "learning_rate": 6.69851143605868e-05, "loss": 0.0493, "step": 42020 }, { "grad_norm": 1.2080376148223877, "learning_rate": 6.696956200257561e-05, "loss": 0.0403, "step": 42030 }, { "grad_norm": 0.6366239190101624, "learning_rate": 6.695400778879863e-05, "loss": 0.037, "step": 42040 }, { "grad_norm": 0.5186730623245239, "learning_rate": 6.693845172095683e-05, "loss": 0.0369, "step": 42050 }, { "grad_norm": 0.5088835954666138, "learning_rate": 6.692289380075142e-05, "loss": 0.0598, "step": 42060 }, { "grad_norm": 0.45400261878967285, "learning_rate": 6.690733402988379e-05, "loss": 0.0609, "step": 42070 }, { "grad_norm": 0.6033270359039307, "learning_rate": 6.689177241005553e-05, "loss": 0.0351, "step": 42080 }, { "grad_norm": 0.5765570402145386, "learning_rate": 6.687620894296844e-05, "loss": 0.0507, "step": 42090 }, { "grad_norm": 0.9781789183616638, "learning_rate": 6.686064363032451e-05, "loss": 0.0494, "step": 42100 }, { "grad_norm": 0.4127419888973236, "learning_rate": 6.684507647382596e-05, "loss": 0.0497, "step": 42110 }, { "grad_norm": 0.5516434907913208, "learning_rate": 6.682950747517513e-05, "loss": 0.0628, "step": 42120 }, { "grad_norm": 0.2722074091434479, "learning_rate": 6.68139366360747e-05, "loss": 0.0428, "step": 42130 }, { "grad_norm": 0.5555918216705322, "learning_rate": 6.679836395822744e-05, "loss": 0.0353, "step": 42140 }, { "grad_norm": 0.6270138025283813, "learning_rate": 6.678278944333633e-05, "loss": 0.0403, "step": 42150 }, { "grad_norm": 0.42538172006607056, "learning_rate": 6.676721309310462e-05, "loss": 0.0441, "step": 42160 }, { "grad_norm": 0.4440534710884094, "learning_rate": 6.67516349092357e-05, "loss": 0.0524, "step": 42170 }, { "grad_norm": 0.5950422286987305, "learning_rate": 6.673605489343315e-05, "loss": 0.0557, "step": 42180 }, { "grad_norm": 0.5478598475456238, "learning_rate": 6.672047304740081e-05, "loss": 0.0509, "step": 42190 }, { "grad_norm": 0.47505053877830505, "learning_rate": 6.670488937284268e-05, "loss": 0.0503, "step": 42200 }, { "grad_norm": 0.5131835341453552, "learning_rate": 6.668930387146294e-05, "loss": 0.0417, "step": 42210 }, { "grad_norm": 0.5989542007446289, "learning_rate": 6.667371654496605e-05, "loss": 0.0393, "step": 42220 }, { "grad_norm": 0.5776965022087097, "learning_rate": 6.665812739505659e-05, "loss": 0.0398, "step": 42230 }, { "grad_norm": 0.5322036147117615, "learning_rate": 6.664253642343935e-05, "loss": 0.032, "step": 42240 }, { "grad_norm": 0.511843204498291, "learning_rate": 6.662694363181935e-05, "loss": 0.0513, "step": 42250 }, { "grad_norm": 0.40350809693336487, "learning_rate": 6.66113490219018e-05, "loss": 0.0486, "step": 42260 }, { "grad_norm": 0.4757066071033478, "learning_rate": 6.659575259539207e-05, "loss": 0.0514, "step": 42270 }, { "grad_norm": 0.4412771761417389, "learning_rate": 6.65801543539958e-05, "loss": 0.0403, "step": 42280 }, { "grad_norm": 0.46555277705192566, "learning_rate": 6.65645542994188e-05, "loss": 0.0475, "step": 42290 }, { "grad_norm": 0.36259761452674866, "learning_rate": 6.654895243336702e-05, "loss": 0.0375, "step": 42300 }, { "grad_norm": 0.782783567905426, "learning_rate": 6.65333487575467e-05, "loss": 0.0447, "step": 42310 }, { "grad_norm": 0.5011155009269714, "learning_rate": 6.651774327366422e-05, "loss": 0.0338, "step": 42320 }, { "grad_norm": 0.553511381149292, "learning_rate": 6.650213598342613e-05, "loss": 0.0405, "step": 42330 }, { "grad_norm": 0.44466307759284973, "learning_rate": 6.648652688853931e-05, "loss": 0.0415, "step": 42340 }, { "grad_norm": 0.7214577198028564, "learning_rate": 6.647091599071066e-05, "loss": 0.0428, "step": 42350 }, { "grad_norm": 0.47072309255599976, "learning_rate": 6.645530329164743e-05, "loss": 0.0351, "step": 42360 }, { "grad_norm": 0.9330509901046753, "learning_rate": 6.643968879305694e-05, "loss": 0.0463, "step": 42370 }, { "grad_norm": 0.6510629653930664, "learning_rate": 6.642407249664683e-05, "loss": 0.0425, "step": 42380 }, { "grad_norm": 0.471566766500473, "learning_rate": 6.640845440412483e-05, "loss": 0.0366, "step": 42390 }, { "grad_norm": 0.37082234025001526, "learning_rate": 6.639283451719893e-05, "loss": 0.0463, "step": 42400 }, { "grad_norm": 0.48445284366607666, "learning_rate": 6.637721283757729e-05, "loss": 0.0482, "step": 42410 }, { "grad_norm": 0.5870599150657654, "learning_rate": 6.636158936696828e-05, "loss": 0.031, "step": 42420 }, { "grad_norm": 0.46950021386146545, "learning_rate": 6.634596410708047e-05, "loss": 0.0408, "step": 42430 }, { "grad_norm": 0.3707232177257538, "learning_rate": 6.633033705962257e-05, "loss": 0.0361, "step": 42440 }, { "grad_norm": 0.45931223034858704, "learning_rate": 6.631470822630359e-05, "loss": 0.0426, "step": 42450 }, { "grad_norm": 0.42160773277282715, "learning_rate": 6.629907760883263e-05, "loss": 0.034, "step": 42460 }, { "grad_norm": 1.0264123678207397, "learning_rate": 6.628344520891905e-05, "loss": 0.0449, "step": 42470 }, { "grad_norm": 0.40700510144233704, "learning_rate": 6.62678110282724e-05, "loss": 0.0313, "step": 42480 }, { "grad_norm": 0.5475713014602661, "learning_rate": 6.625217506860237e-05, "loss": 0.0468, "step": 42490 }, { "grad_norm": 0.7526366114616394, "learning_rate": 6.623653733161892e-05, "loss": 0.0308, "step": 42500 }, { "grad_norm": 0.5052512884140015, "learning_rate": 6.622089781903217e-05, "loss": 0.048, "step": 42510 }, { "grad_norm": 0.4333740174770355, "learning_rate": 6.620525653255241e-05, "loss": 0.046, "step": 42520 }, { "grad_norm": 0.7012739181518555, "learning_rate": 6.618961347389015e-05, "loss": 0.0504, "step": 42530 }, { "grad_norm": 0.47392162680625916, "learning_rate": 6.617396864475613e-05, "loss": 0.0388, "step": 42540 }, { "grad_norm": 0.42330893874168396, "learning_rate": 6.61583220468612e-05, "loss": 0.0352, "step": 42550 }, { "grad_norm": 0.5017439126968384, "learning_rate": 6.614267368191645e-05, "loss": 0.0508, "step": 42560 }, { "grad_norm": 0.6065912842750549, "learning_rate": 6.61270235516332e-05, "loss": 0.0431, "step": 42570 }, { "grad_norm": 0.6546527147293091, "learning_rate": 6.611137165772287e-05, "loss": 0.0349, "step": 42580 }, { "grad_norm": 0.32935282588005066, "learning_rate": 6.60957180018972e-05, "loss": 0.0277, "step": 42590 }, { "grad_norm": 0.38485851883888245, "learning_rate": 6.608006258586797e-05, "loss": 0.0376, "step": 42600 }, { "grad_norm": 0.4647946059703827, "learning_rate": 6.606440541134731e-05, "loss": 0.0408, "step": 42610 }, { "grad_norm": 0.6706732511520386, "learning_rate": 6.60487464800474e-05, "loss": 0.0444, "step": 42620 }, { "grad_norm": 0.5304138660430908, "learning_rate": 6.603308579368071e-05, "loss": 0.0407, "step": 42630 }, { "grad_norm": 0.3954164683818817, "learning_rate": 6.601742335395987e-05, "loss": 0.0448, "step": 42640 }, { "grad_norm": 0.546974778175354, "learning_rate": 6.600175916259769e-05, "loss": 0.0387, "step": 42650 }, { "grad_norm": 0.7170104384422302, "learning_rate": 6.598609322130718e-05, "loss": 0.0542, "step": 42660 }, { "grad_norm": 0.3928201198577881, "learning_rate": 6.597042553180155e-05, "loss": 0.0424, "step": 42670 }, { "grad_norm": 0.5701808929443359, "learning_rate": 6.59547560957942e-05, "loss": 0.0263, "step": 42680 }, { "grad_norm": 0.5330915451049805, "learning_rate": 6.59390849149987e-05, "loss": 0.0448, "step": 42690 }, { "grad_norm": 0.4369414150714874, "learning_rate": 6.592341199112886e-05, "loss": 0.0344, "step": 42700 }, { "grad_norm": 0.5071912407875061, "learning_rate": 6.590773732589861e-05, "loss": 0.0307, "step": 42710 }, { "grad_norm": 0.7295581102371216, "learning_rate": 6.589206092102213e-05, "loss": 0.0457, "step": 42720 }, { "grad_norm": 0.5733314752578735, "learning_rate": 6.587638277821375e-05, "loss": 0.0435, "step": 42730 }, { "grad_norm": 0.7225358486175537, "learning_rate": 6.586070289918803e-05, "loss": 0.043, "step": 42740 }, { "grad_norm": 0.40752387046813965, "learning_rate": 6.584502128565968e-05, "loss": 0.038, "step": 42750 }, { "grad_norm": 0.526567816734314, "learning_rate": 6.582933793934363e-05, "loss": 0.0394, "step": 42760 }, { "grad_norm": 0.38994863629341125, "learning_rate": 6.581365286195498e-05, "loss": 0.0424, "step": 42770 }, { "grad_norm": 0.4746730625629425, "learning_rate": 6.579796605520903e-05, "loss": 0.0467, "step": 42780 }, { "grad_norm": 0.7575257420539856, "learning_rate": 6.578227752082127e-05, "loss": 0.0506, "step": 42790 }, { "grad_norm": 0.6365805864334106, "learning_rate": 6.576658726050735e-05, "loss": 0.0398, "step": 42800 }, { "grad_norm": 0.5028039813041687, "learning_rate": 6.575089527598316e-05, "loss": 0.0381, "step": 42810 }, { "grad_norm": 0.3802008628845215, "learning_rate": 6.573520156896476e-05, "loss": 0.0544, "step": 42820 }, { "grad_norm": 0.37014538049697876, "learning_rate": 6.571950614116835e-05, "loss": 0.0357, "step": 42830 }, { "grad_norm": 0.4437294900417328, "learning_rate": 6.570380899431039e-05, "loss": 0.0527, "step": 42840 }, { "grad_norm": 0.7167976498603821, "learning_rate": 6.568811013010749e-05, "loss": 0.0425, "step": 42850 }, { "grad_norm": 0.4360518753528595, "learning_rate": 6.567240955027645e-05, "loss": 0.0367, "step": 42860 }, { "grad_norm": 0.727110743522644, "learning_rate": 6.565670725653427e-05, "loss": 0.0472, "step": 42870 }, { "grad_norm": 0.3699495494365692, "learning_rate": 6.564100325059812e-05, "loss": 0.0347, "step": 42880 }, { "grad_norm": 0.6229885816574097, "learning_rate": 6.562529753418538e-05, "loss": 0.0493, "step": 42890 }, { "grad_norm": 0.7199077606201172, "learning_rate": 6.56095901090136e-05, "loss": 0.0571, "step": 42900 }, { "grad_norm": 0.54872065782547, "learning_rate": 6.55938809768005e-05, "loss": 0.0555, "step": 42910 }, { "grad_norm": 0.6411569118499756, "learning_rate": 6.557817013926402e-05, "loss": 0.0431, "step": 42920 }, { "grad_norm": 0.5757753849029541, "learning_rate": 6.556245759812227e-05, "loss": 0.0614, "step": 42930 }, { "grad_norm": 0.6177501082420349, "learning_rate": 6.554674335509357e-05, "loss": 0.0528, "step": 42940 }, { "grad_norm": 0.6599146723747253, "learning_rate": 6.553102741189638e-05, "loss": 0.0395, "step": 42950 }, { "grad_norm": 0.6935361623764038, "learning_rate": 6.55153097702494e-05, "loss": 0.0442, "step": 42960 }, { "grad_norm": 0.4286591112613678, "learning_rate": 6.549959043187146e-05, "loss": 0.0354, "step": 42970 }, { "grad_norm": 0.5333662033081055, "learning_rate": 6.54838693984816e-05, "loss": 0.0475, "step": 42980 }, { "grad_norm": 0.5802068114280701, "learning_rate": 6.546814667179909e-05, "loss": 0.0507, "step": 42990 }, { "grad_norm": 0.5337920188903809, "learning_rate": 6.545242225354328e-05, "loss": 0.048, "step": 43000 }, { "grad_norm": 0.36583980917930603, "learning_rate": 6.543669614543383e-05, "loss": 0.0536, "step": 43010 }, { "grad_norm": 0.35955217480659485, "learning_rate": 6.542096834919049e-05, "loss": 0.0312, "step": 43020 }, { "grad_norm": 0.39156362414360046, "learning_rate": 6.540523886653324e-05, "loss": 0.0359, "step": 43030 }, { "grad_norm": 0.6336883902549744, "learning_rate": 6.538950769918222e-05, "loss": 0.0445, "step": 43040 }, { "grad_norm": 0.6080146431922913, "learning_rate": 6.537377484885779e-05, "loss": 0.0375, "step": 43050 }, { "grad_norm": 0.7677339315414429, "learning_rate": 6.535804031728045e-05, "loss": 0.036, "step": 43060 }, { "grad_norm": 0.9474451541900635, "learning_rate": 6.534230410617092e-05, "loss": 0.0435, "step": 43070 }, { "grad_norm": 0.5684894323348999, "learning_rate": 6.532656621725007e-05, "loss": 0.0484, "step": 43080 }, { "grad_norm": 0.5860534310340881, "learning_rate": 6.531082665223898e-05, "loss": 0.0344, "step": 43090 }, { "grad_norm": 0.6707851886749268, "learning_rate": 6.529508541285889e-05, "loss": 0.0556, "step": 43100 }, { "grad_norm": 0.5345644950866699, "learning_rate": 6.527934250083128e-05, "loss": 0.043, "step": 43110 }, { "grad_norm": 0.4097723066806793, "learning_rate": 6.526359791787775e-05, "loss": 0.0419, "step": 43120 }, { "grad_norm": 0.6364563703536987, "learning_rate": 6.524785166572009e-05, "loss": 0.0365, "step": 43130 }, { "grad_norm": 0.41711124777793884, "learning_rate": 6.523210374608029e-05, "loss": 0.045, "step": 43140 }, { "grad_norm": 0.33380088210105896, "learning_rate": 6.521635416068054e-05, "loss": 0.0399, "step": 43150 }, { "grad_norm": 0.5419729948043823, "learning_rate": 6.520060291124317e-05, "loss": 0.0293, "step": 43160 }, { "grad_norm": 0.3573084771633148, "learning_rate": 6.518484999949072e-05, "loss": 0.0473, "step": 43170 }, { "grad_norm": 0.4138779044151306, "learning_rate": 6.516909542714591e-05, "loss": 0.0498, "step": 43180 }, { "grad_norm": 0.6072503328323364, "learning_rate": 6.515333919593161e-05, "loss": 0.0381, "step": 43190 }, { "grad_norm": 0.3897024393081665, "learning_rate": 6.513758130757094e-05, "loss": 0.0311, "step": 43200 }, { "grad_norm": 0.48336756229400635, "learning_rate": 6.512182176378713e-05, "loss": 0.0373, "step": 43210 }, { "grad_norm": 0.5636741518974304, "learning_rate": 6.510606056630362e-05, "loss": 0.0564, "step": 43220 }, { "grad_norm": 0.3437434434890747, "learning_rate": 6.509029771684405e-05, "loss": 0.0367, "step": 43230 }, { "grad_norm": 0.6932621002197266, "learning_rate": 6.50745332171322e-05, "loss": 0.0383, "step": 43240 }, { "grad_norm": 0.5135777592658997, "learning_rate": 6.505876706889207e-05, "loss": 0.0496, "step": 43250 }, { "grad_norm": 1.0944867134094238, "learning_rate": 6.504299927384781e-05, "loss": 0.0454, "step": 43260 }, { "grad_norm": 0.6586645841598511, "learning_rate": 6.502722983372378e-05, "loss": 0.0387, "step": 43270 }, { "grad_norm": 0.4343865215778351, "learning_rate": 6.501145875024446e-05, "loss": 0.0311, "step": 43280 }, { "grad_norm": 0.3587062954902649, "learning_rate": 6.499568602513459e-05, "loss": 0.04, "step": 43290 }, { "grad_norm": 0.418287456035614, "learning_rate": 6.497991166011903e-05, "loss": 0.0317, "step": 43300 }, { "grad_norm": 0.4621419310569763, "learning_rate": 6.49641356569229e-05, "loss": 0.0603, "step": 43310 }, { "grad_norm": 0.7677015662193298, "learning_rate": 6.494835801727135e-05, "loss": 0.0392, "step": 43320 }, { "grad_norm": 0.6712169647216797, "learning_rate": 6.493257874288986e-05, "loss": 0.0415, "step": 43330 }, { "grad_norm": 0.2643676996231079, "learning_rate": 6.491679783550399e-05, "loss": 0.0388, "step": 43340 }, { "grad_norm": 0.5388785004615784, "learning_rate": 6.490101529683957e-05, "loss": 0.0385, "step": 43350 }, { "grad_norm": 0.4019958972930908, "learning_rate": 6.488523112862248e-05, "loss": 0.0458, "step": 43360 }, { "grad_norm": 0.40112635493278503, "learning_rate": 6.486944533257891e-05, "loss": 0.0391, "step": 43370 }, { "grad_norm": 0.6758356690406799, "learning_rate": 6.485365791043517e-05, "loss": 0.0424, "step": 43380 }, { "grad_norm": 0.4703682065010071, "learning_rate": 6.483786886391772e-05, "loss": 0.0429, "step": 43390 }, { "grad_norm": 0.8124924302101135, "learning_rate": 6.482207819475323e-05, "loss": 0.0513, "step": 43400 }, { "grad_norm": 0.537169873714447, "learning_rate": 6.480628590466857e-05, "loss": 0.0362, "step": 43410 }, { "grad_norm": 0.6076358556747437, "learning_rate": 6.479049199539074e-05, "loss": 0.0404, "step": 43420 }, { "grad_norm": 0.5873460173606873, "learning_rate": 6.477469646864693e-05, "loss": 0.0415, "step": 43430 }, { "grad_norm": 0.4936930537223816, "learning_rate": 6.475889932616454e-05, "loss": 0.0431, "step": 43440 }, { "grad_norm": 0.4926062524318695, "learning_rate": 6.474310056967111e-05, "loss": 0.0325, "step": 43450 }, { "grad_norm": 0.5986281037330627, "learning_rate": 6.472730020089437e-05, "loss": 0.0338, "step": 43460 }, { "grad_norm": 0.7162584066390991, "learning_rate": 6.471149822156222e-05, "loss": 0.0518, "step": 43470 }, { "grad_norm": 0.4980444610118866, "learning_rate": 6.469569463340275e-05, "loss": 0.0426, "step": 43480 }, { "grad_norm": 0.3679739236831665, "learning_rate": 6.467988943814418e-05, "loss": 0.0345, "step": 43490 }, { "grad_norm": 0.405997097492218, "learning_rate": 6.4664082637515e-05, "loss": 0.0382, "step": 43500 }, { "grad_norm": 0.35041260719299316, "learning_rate": 6.464827423324377e-05, "loss": 0.0405, "step": 43510 }, { "grad_norm": 0.49038323760032654, "learning_rate": 6.463246422705931e-05, "loss": 0.0639, "step": 43520 }, { "grad_norm": 0.5826379060745239, "learning_rate": 6.461665262069055e-05, "loss": 0.0443, "step": 43530 }, { "grad_norm": 0.33756986260414124, "learning_rate": 6.460083941586665e-05, "loss": 0.0462, "step": 43540 }, { "grad_norm": 0.5897189974784851, "learning_rate": 6.45850246143169e-05, "loss": 0.049, "step": 43550 }, { "grad_norm": 0.6719038486480713, "learning_rate": 6.456920821777077e-05, "loss": 0.0783, "step": 43560 }, { "grad_norm": 0.3545362651348114, "learning_rate": 6.455339022795796e-05, "loss": 0.0375, "step": 43570 }, { "grad_norm": 0.30928120017051697, "learning_rate": 6.453757064660825e-05, "loss": 0.0377, "step": 43580 }, { "grad_norm": 0.43741104006767273, "learning_rate": 6.452174947545169e-05, "loss": 0.0344, "step": 43590 }, { "grad_norm": 0.5096715688705444, "learning_rate": 6.450592671621842e-05, "loss": 0.0566, "step": 43600 }, { "grad_norm": 0.4339558780193329, "learning_rate": 6.449010237063882e-05, "loss": 0.0363, "step": 43610 }, { "grad_norm": 0.3310670256614685, "learning_rate": 6.447427644044343e-05, "loss": 0.0309, "step": 43620 }, { "grad_norm": 0.36194631457328796, "learning_rate": 6.44584489273629e-05, "loss": 0.0425, "step": 43630 }, { "grad_norm": 0.5759273171424866, "learning_rate": 6.444261983312817e-05, "loss": 0.0453, "step": 43640 }, { "grad_norm": 0.5898918509483337, "learning_rate": 6.442678915947023e-05, "loss": 0.0445, "step": 43650 }, { "grad_norm": 0.8052988648414612, "learning_rate": 6.441095690812032e-05, "loss": 0.0563, "step": 43660 }, { "grad_norm": 0.525023877620697, "learning_rate": 6.439512308080982e-05, "loss": 0.0444, "step": 43670 }, { "grad_norm": 0.5007182359695435, "learning_rate": 6.43792876792703e-05, "loss": 0.0376, "step": 43680 }, { "grad_norm": 0.7774642705917358, "learning_rate": 6.43634507052335e-05, "loss": 0.0477, "step": 43690 }, { "grad_norm": 0.4640820622444153, "learning_rate": 6.434761216043135e-05, "loss": 0.0368, "step": 43700 }, { "grad_norm": 0.549830436706543, "learning_rate": 6.433177204659589e-05, "loss": 0.0602, "step": 43710 }, { "grad_norm": 1.4744583368301392, "learning_rate": 6.431593036545938e-05, "loss": 0.0402, "step": 43720 }, { "grad_norm": 0.43271175026893616, "learning_rate": 6.430008711875426e-05, "loss": 0.0495, "step": 43730 }, { "grad_norm": 1.2244654893875122, "learning_rate": 6.42842423082131e-05, "loss": 0.0379, "step": 43740 }, { "grad_norm": 0.8538122773170471, "learning_rate": 6.426839593556868e-05, "loss": 0.0407, "step": 43750 }, { "grad_norm": 0.3697182536125183, "learning_rate": 6.425254800255392e-05, "loss": 0.0322, "step": 43760 }, { "grad_norm": 0.6244137287139893, "learning_rate": 6.423669851090196e-05, "loss": 0.0429, "step": 43770 }, { "grad_norm": 0.6066017746925354, "learning_rate": 6.422084746234604e-05, "loss": 0.0451, "step": 43780 }, { "grad_norm": 0.44656437635421753, "learning_rate": 6.420499485861965e-05, "loss": 0.0399, "step": 43790 }, { "grad_norm": 0.5443257093429565, "learning_rate": 6.418914070145636e-05, "loss": 0.0458, "step": 43800 }, { "grad_norm": 0.5487172603607178, "learning_rate": 6.417328499258996e-05, "loss": 0.0401, "step": 43810 }, { "grad_norm": 0.48440495133399963, "learning_rate": 6.415742773375446e-05, "loss": 0.05, "step": 43820 }, { "grad_norm": 0.453799307346344, "learning_rate": 6.414156892668393e-05, "loss": 0.0354, "step": 43830 }, { "grad_norm": 0.5162604451179504, "learning_rate": 6.412570857311267e-05, "loss": 0.0489, "step": 43840 }, { "grad_norm": 0.48847317695617676, "learning_rate": 6.410984667477518e-05, "loss": 0.04, "step": 43850 }, { "grad_norm": 0.3901398181915283, "learning_rate": 6.409398323340607e-05, "loss": 0.0445, "step": 43860 }, { "grad_norm": 0.613660991191864, "learning_rate": 6.407811825074012e-05, "loss": 0.0368, "step": 43870 }, { "grad_norm": 0.5899502635002136, "learning_rate": 6.406225172851234e-05, "loss": 0.0464, "step": 43880 }, { "grad_norm": 0.8873897194862366, "learning_rate": 6.404638366845786e-05, "loss": 0.049, "step": 43890 }, { "grad_norm": 0.43410053849220276, "learning_rate": 6.403051407231196e-05, "loss": 0.037, "step": 43900 }, { "grad_norm": 0.734235405921936, "learning_rate": 6.401464294181016e-05, "loss": 0.0435, "step": 43910 }, { "grad_norm": 0.7782819271087646, "learning_rate": 6.399877027868808e-05, "loss": 0.0446, "step": 43920 }, { "grad_norm": 0.41650423407554626, "learning_rate": 6.39828960846815e-05, "loss": 0.0418, "step": 43930 }, { "grad_norm": 0.5778665542602539, "learning_rate": 6.396702036152644e-05, "loss": 0.0435, "step": 43940 }, { "grad_norm": 0.2542087733745575, "learning_rate": 6.395114311095905e-05, "loss": 0.0283, "step": 43950 }, { "grad_norm": 0.4350090026855469, "learning_rate": 6.393526433471559e-05, "loss": 0.0371, "step": 43960 }, { "grad_norm": 0.6901717782020569, "learning_rate": 6.391938403453262e-05, "loss": 0.0268, "step": 43970 }, { "grad_norm": 0.3377622365951538, "learning_rate": 6.390350221214671e-05, "loss": 0.0434, "step": 43980 }, { "grad_norm": 0.5427712798118591, "learning_rate": 6.38876188692947e-05, "loss": 0.0404, "step": 43990 }, { "grad_norm": 0.5750882029533386, "learning_rate": 6.387173400771358e-05, "loss": 0.0441, "step": 44000 }, { "grad_norm": 0.6890609264373779, "learning_rate": 6.385584762914047e-05, "loss": 0.0343, "step": 44010 }, { "grad_norm": 0.5002971887588501, "learning_rate": 6.38399597353127e-05, "loss": 0.0411, "step": 44020 }, { "grad_norm": 0.33075037598609924, "learning_rate": 6.382407032796775e-05, "loss": 0.047, "step": 44030 }, { "grad_norm": 0.49206972122192383, "learning_rate": 6.380817940884325e-05, "loss": 0.041, "step": 44040 }, { "grad_norm": 0.6789848804473877, "learning_rate": 6.379228697967702e-05, "loss": 0.0635, "step": 44050 }, { "grad_norm": 0.39309608936309814, "learning_rate": 6.3776393042207e-05, "loss": 0.0326, "step": 44060 }, { "grad_norm": 1.0071054697036743, "learning_rate": 6.376049759817137e-05, "loss": 0.057, "step": 44070 }, { "grad_norm": 0.5168696641921997, "learning_rate": 6.37446006493084e-05, "loss": 0.0406, "step": 44080 }, { "grad_norm": 0.37102237343788147, "learning_rate": 6.372870219735656e-05, "loss": 0.0379, "step": 44090 }, { "grad_norm": 0.6251283884048462, "learning_rate": 6.371280224405452e-05, "loss": 0.0417, "step": 44100 }, { "grad_norm": 0.507343590259552, "learning_rate": 6.369690079114101e-05, "loss": 0.0308, "step": 44110 }, { "grad_norm": 0.554843544960022, "learning_rate": 6.368099784035504e-05, "loss": 0.0472, "step": 44120 }, { "grad_norm": 0.6292346119880676, "learning_rate": 6.366509339343572e-05, "loss": 0.0485, "step": 44130 }, { "grad_norm": 0.49689382314682007, "learning_rate": 6.364918745212232e-05, "loss": 0.042, "step": 44140 }, { "grad_norm": 0.5309044718742371, "learning_rate": 6.363328001815431e-05, "loss": 0.055, "step": 44150 }, { "grad_norm": 0.5720087885856628, "learning_rate": 6.361737109327128e-05, "loss": 0.0322, "step": 44160 }, { "grad_norm": 0.5511709451675415, "learning_rate": 6.360146067921301e-05, "loss": 0.0487, "step": 44170 }, { "grad_norm": 0.8359546661376953, "learning_rate": 6.35855487777195e-05, "loss": 0.0554, "step": 44180 }, { "grad_norm": 0.5906589031219482, "learning_rate": 6.356963539053076e-05, "loss": 0.0457, "step": 44190 }, { "grad_norm": 0.4364548325538635, "learning_rate": 6.355372051938711e-05, "loss": 0.0362, "step": 44200 }, { "grad_norm": 0.8833779692649841, "learning_rate": 6.353780416602894e-05, "loss": 0.0448, "step": 44210 }, { "grad_norm": 0.5581071972846985, "learning_rate": 6.352188633219689e-05, "loss": 0.0564, "step": 44220 }, { "grad_norm": 0.5598196983337402, "learning_rate": 6.350596701963166e-05, "loss": 0.0592, "step": 44230 }, { "grad_norm": 0.40972062945365906, "learning_rate": 6.349004623007419e-05, "loss": 0.0443, "step": 44240 }, { "grad_norm": 0.7565534114837646, "learning_rate": 6.347412396526555e-05, "loss": 0.0494, "step": 44250 }, { "grad_norm": 0.3963280916213989, "learning_rate": 6.345820022694696e-05, "loss": 0.0385, "step": 44260 }, { "grad_norm": 0.7846320271492004, "learning_rate": 6.344227501685984e-05, "loss": 0.0381, "step": 44270 }, { "grad_norm": 0.48491138219833374, "learning_rate": 6.342634833674572e-05, "loss": 0.0405, "step": 44280 }, { "grad_norm": 0.3853797912597656, "learning_rate": 6.341042018834635e-05, "loss": 0.034, "step": 44290 }, { "grad_norm": 0.4836513102054596, "learning_rate": 6.339449057340359e-05, "loss": 0.0358, "step": 44300 }, { "grad_norm": 0.5983522534370422, "learning_rate": 6.337855949365945e-05, "loss": 0.0349, "step": 44310 }, { "grad_norm": 0.48196813464164734, "learning_rate": 6.336262695085619e-05, "loss": 0.0315, "step": 44320 }, { "grad_norm": 0.6400942802429199, "learning_rate": 6.334669294673612e-05, "loss": 0.0353, "step": 44330 }, { "grad_norm": 0.5352827310562134, "learning_rate": 6.333075748304179e-05, "loss": 0.0324, "step": 44340 }, { "grad_norm": 0.6317440867424011, "learning_rate": 6.331482056151585e-05, "loss": 0.0384, "step": 44350 }, { "grad_norm": 0.7721714973449707, "learning_rate": 6.329888218390117e-05, "loss": 0.0461, "step": 44360 }, { "grad_norm": 0.834617555141449, "learning_rate": 6.328294235194072e-05, "loss": 0.0528, "step": 44370 }, { "grad_norm": 0.3761974275112152, "learning_rate": 6.326700106737765e-05, "loss": 0.0404, "step": 44380 }, { "grad_norm": 0.7395943403244019, "learning_rate": 6.325105833195531e-05, "loss": 0.0414, "step": 44390 }, { "grad_norm": 0.5186313390731812, "learning_rate": 6.323511414741715e-05, "loss": 0.0532, "step": 44400 }, { "grad_norm": 0.6309784650802612, "learning_rate": 6.321916851550678e-05, "loss": 0.0409, "step": 44410 }, { "grad_norm": 0.6746063232421875, "learning_rate": 6.320322143796806e-05, "loss": 0.0417, "step": 44420 }, { "grad_norm": 1.149919033050537, "learning_rate": 6.318727291654488e-05, "loss": 0.0476, "step": 44430 }, { "grad_norm": 0.6290131211280823, "learning_rate": 6.317132295298134e-05, "loss": 0.0412, "step": 44440 }, { "grad_norm": 0.49697622656822205, "learning_rate": 6.315537154902173e-05, "loss": 0.0285, "step": 44450 }, { "grad_norm": 0.49185845255851746, "learning_rate": 6.313941870641048e-05, "loss": 0.0397, "step": 44460 }, { "grad_norm": 0.6687734127044678, "learning_rate": 6.312346442689214e-05, "loss": 0.0518, "step": 44470 }, { "grad_norm": 0.6971397995948792, "learning_rate": 6.310750871221147e-05, "loss": 0.052, "step": 44480 }, { "grad_norm": 0.4333742558956146, "learning_rate": 6.309155156411335e-05, "loss": 0.0434, "step": 44490 }, { "grad_norm": 0.6574097275733948, "learning_rate": 6.307559298434284e-05, "loss": 0.0273, "step": 44500 }, { "grad_norm": 0.3980412185192108, "learning_rate": 6.305963297464512e-05, "loss": 0.0448, "step": 44510 }, { "grad_norm": 0.5608577728271484, "learning_rate": 6.304367153676561e-05, "loss": 0.0523, "step": 44520 }, { "grad_norm": 0.8487575650215149, "learning_rate": 6.302770867244978e-05, "loss": 0.0561, "step": 44530 }, { "grad_norm": 0.737116813659668, "learning_rate": 6.301174438344328e-05, "loss": 0.0316, "step": 44540 }, { "grad_norm": 0.6678928732872009, "learning_rate": 6.299577867149202e-05, "loss": 0.0423, "step": 44550 }, { "grad_norm": 0.36538103222846985, "learning_rate": 6.297981153834192e-05, "loss": 0.0505, "step": 44560 }, { "grad_norm": 0.8203524947166443, "learning_rate": 6.296384298573916e-05, "loss": 0.0549, "step": 44570 }, { "grad_norm": 1.076899766921997, "learning_rate": 6.294787301543001e-05, "loss": 0.0406, "step": 44580 }, { "grad_norm": 0.5782814621925354, "learning_rate": 6.293190162916095e-05, "loss": 0.0369, "step": 44590 }, { "grad_norm": 0.41301098465919495, "learning_rate": 6.291592882867855e-05, "loss": 0.0428, "step": 44600 }, { "grad_norm": 0.537738561630249, "learning_rate": 6.28999546157296e-05, "loss": 0.0492, "step": 44610 }, { "grad_norm": 1.1340872049331665, "learning_rate": 6.288397899206102e-05, "loss": 0.0342, "step": 44620 }, { "grad_norm": 0.4770452380180359, "learning_rate": 6.286800195941984e-05, "loss": 0.0495, "step": 44630 }, { "grad_norm": 0.49273157119750977, "learning_rate": 6.285202351955334e-05, "loss": 0.0433, "step": 44640 }, { "grad_norm": 0.5624255537986755, "learning_rate": 6.283604367420887e-05, "loss": 0.0359, "step": 44650 }, { "grad_norm": 0.5346558690071106, "learning_rate": 6.282006242513394e-05, "loss": 0.0479, "step": 44660 }, { "grad_norm": 0.6310287117958069, "learning_rate": 6.280407977407628e-05, "loss": 0.0593, "step": 44670 }, { "grad_norm": 0.3088177740573883, "learning_rate": 6.27880957227837e-05, "loss": 0.0452, "step": 44680 }, { "grad_norm": 0.6952694058418274, "learning_rate": 6.27721102730042e-05, "loss": 0.0529, "step": 44690 }, { "grad_norm": 0.5793741345405579, "learning_rate": 6.275612342648592e-05, "loss": 0.0346, "step": 44700 }, { "grad_norm": 1.0100376605987549, "learning_rate": 6.274013518497716e-05, "loss": 0.0443, "step": 44710 }, { "grad_norm": 0.47643014788627625, "learning_rate": 6.272414555022636e-05, "loss": 0.0369, "step": 44720 }, { "grad_norm": 0.5566343665122986, "learning_rate": 6.270815452398215e-05, "loss": 0.045, "step": 44730 }, { "grad_norm": 0.4757605493068695, "learning_rate": 6.269216210799326e-05, "loss": 0.0645, "step": 44740 }, { "grad_norm": 0.4694156050682068, "learning_rate": 6.26761683040086e-05, "loss": 0.0407, "step": 44750 }, { "grad_norm": 0.44151920080184937, "learning_rate": 6.266017311377723e-05, "loss": 0.0405, "step": 44760 }, { "grad_norm": 0.35809600353240967, "learning_rate": 6.264417653904839e-05, "loss": 0.0357, "step": 44770 }, { "grad_norm": 0.44511863589286804, "learning_rate": 6.262817858157139e-05, "loss": 0.041, "step": 44780 }, { "grad_norm": 0.6944315433502197, "learning_rate": 6.261217924309576e-05, "loss": 0.0453, "step": 44790 }, { "grad_norm": 0.5568474531173706, "learning_rate": 6.259617852537118e-05, "loss": 0.0473, "step": 44800 }, { "grad_norm": 0.809406578540802, "learning_rate": 6.258017643014747e-05, "loss": 0.0578, "step": 44810 }, { "grad_norm": 0.5917201638221741, "learning_rate": 6.256417295917456e-05, "loss": 0.0543, "step": 44820 }, { "grad_norm": 0.562535285949707, "learning_rate": 6.254816811420258e-05, "loss": 0.0353, "step": 44830 }, { "grad_norm": 0.6091095805168152, "learning_rate": 6.253216189698183e-05, "loss": 0.0544, "step": 44840 }, { "grad_norm": 0.47025319933891296, "learning_rate": 6.251615430926267e-05, "loss": 0.0374, "step": 44850 }, { "grad_norm": 0.6573266983032227, "learning_rate": 6.25001453527957e-05, "loss": 0.0444, "step": 44860 }, { "grad_norm": 0.5730999112129211, "learning_rate": 6.248413502933164e-05, "loss": 0.052, "step": 44870 }, { "grad_norm": 0.6611257195472717, "learning_rate": 6.246812334062133e-05, "loss": 0.0405, "step": 44880 }, { "grad_norm": 1.8914384841918945, "learning_rate": 6.245211028841579e-05, "loss": 0.0408, "step": 44890 }, { "grad_norm": 0.5172659158706665, "learning_rate": 6.24360958744662e-05, "loss": 0.0468, "step": 44900 }, { "grad_norm": 0.691758930683136, "learning_rate": 6.242008010052387e-05, "loss": 0.0386, "step": 44910 }, { "grad_norm": 0.3479919135570526, "learning_rate": 6.240406296834024e-05, "loss": 0.0447, "step": 44920 }, { "grad_norm": 1.6070789098739624, "learning_rate": 6.238804447966694e-05, "loss": 0.0445, "step": 44930 }, { "grad_norm": 0.4224817454814911, "learning_rate": 6.237202463625573e-05, "loss": 0.0484, "step": 44940 }, { "grad_norm": 0.4049340486526489, "learning_rate": 6.235600343985848e-05, "loss": 0.0406, "step": 44950 }, { "grad_norm": 0.7198245525360107, "learning_rate": 6.233998089222729e-05, "loss": 0.047, "step": 44960 }, { "grad_norm": 0.8944703936576843, "learning_rate": 6.232395699511433e-05, "loss": 0.0386, "step": 44970 }, { "grad_norm": 0.7744218707084656, "learning_rate": 6.230793175027199e-05, "loss": 0.0412, "step": 44980 }, { "grad_norm": 1.2381671667099, "learning_rate": 6.22919051594527e-05, "loss": 0.0401, "step": 44990 }, { "grad_norm": 0.5182633399963379, "learning_rate": 6.227587722440917e-05, "loss": 0.0355, "step": 45000 }, { "grad_norm": 0.2853468954563141, "learning_rate": 6.225984794689414e-05, "loss": 0.034, "step": 45010 }, { "grad_norm": 0.3918360769748688, "learning_rate": 6.22438173286606e-05, "loss": 0.0334, "step": 45020 }, { "grad_norm": 0.6074173450469971, "learning_rate": 6.222778537146159e-05, "loss": 0.0327, "step": 45030 }, { "grad_norm": 0.5218637585639954, "learning_rate": 6.221175207705037e-05, "loss": 0.0535, "step": 45040 }, { "grad_norm": 0.5674593448638916, "learning_rate": 6.219571744718027e-05, "loss": 0.0517, "step": 45050 }, { "grad_norm": 0.5996374487876892, "learning_rate": 6.217968148360487e-05, "loss": 0.0388, "step": 45060 }, { "grad_norm": 0.4177427589893341, "learning_rate": 6.216364418807782e-05, "loss": 0.0366, "step": 45070 }, { "grad_norm": 0.9529910087585449, "learning_rate": 6.214760556235292e-05, "loss": 0.0405, "step": 45080 }, { "grad_norm": 0.3724326193332672, "learning_rate": 6.213156560818414e-05, "loss": 0.047, "step": 45090 }, { "grad_norm": 0.7693004608154297, "learning_rate": 6.211552432732559e-05, "loss": 0.0508, "step": 45100 }, { "grad_norm": 0.3848029375076294, "learning_rate": 6.20994817215315e-05, "loss": 0.0434, "step": 45110 }, { "grad_norm": 0.7058331966400146, "learning_rate": 6.208343779255628e-05, "loss": 0.0588, "step": 45120 }, { "grad_norm": 0.4543534517288208, "learning_rate": 6.206739254215449e-05, "loss": 0.0408, "step": 45130 }, { "grad_norm": 0.5346185564994812, "learning_rate": 6.205134597208077e-05, "loss": 0.0423, "step": 45140 }, { "grad_norm": 0.5986899733543396, "learning_rate": 6.203529808409e-05, "loss": 0.0367, "step": 45150 }, { "grad_norm": 0.32319167256355286, "learning_rate": 6.201924887993712e-05, "loss": 0.0304, "step": 45160 }, { "grad_norm": 0.6218602061271667, "learning_rate": 6.200319836137725e-05, "loss": 0.0447, "step": 45170 }, { "grad_norm": 0.6922245025634766, "learning_rate": 6.198714653016565e-05, "loss": 0.0317, "step": 45180 }, { "grad_norm": 0.451594740152359, "learning_rate": 6.197109338805774e-05, "loss": 0.0521, "step": 45190 }, { "grad_norm": 0.3892284631729126, "learning_rate": 6.195503893680903e-05, "loss": 0.0314, "step": 45200 }, { "grad_norm": 0.38813358545303345, "learning_rate": 6.193898317817524e-05, "loss": 0.0342, "step": 45210 }, { "grad_norm": 0.9719483852386475, "learning_rate": 6.192292611391221e-05, "loss": 0.0378, "step": 45220 }, { "grad_norm": 0.5445789694786072, "learning_rate": 6.190686774577591e-05, "loss": 0.0345, "step": 45230 }, { "grad_norm": 0.4747101068496704, "learning_rate": 6.189080807552245e-05, "loss": 0.0372, "step": 45240 }, { "grad_norm": 0.5703929662704468, "learning_rate": 6.187474710490809e-05, "loss": 0.0411, "step": 45250 }, { "grad_norm": 0.45766139030456543, "learning_rate": 6.185868483568926e-05, "loss": 0.0483, "step": 45260 }, { "grad_norm": 0.7678465843200684, "learning_rate": 6.184262126962245e-05, "loss": 0.0668, "step": 45270 }, { "grad_norm": 0.45947960019111633, "learning_rate": 6.182655640846442e-05, "loss": 0.0461, "step": 45280 }, { "grad_norm": 0.953541100025177, "learning_rate": 6.181049025397196e-05, "loss": 0.0464, "step": 45290 }, { "grad_norm": 0.514866054058075, "learning_rate": 6.179442280790202e-05, "loss": 0.057, "step": 45300 }, { "grad_norm": 0.4154534935951233, "learning_rate": 6.177835407201174e-05, "loss": 0.0399, "step": 45310 }, { "grad_norm": 0.7734084129333496, "learning_rate": 6.176228404805839e-05, "loss": 0.0429, "step": 45320 }, { "grad_norm": 0.4702768325805664, "learning_rate": 6.174621273779932e-05, "loss": 0.0318, "step": 45330 }, { "grad_norm": 0.44692015647888184, "learning_rate": 6.17301401429921e-05, "loss": 0.0322, "step": 45340 }, { "grad_norm": 0.6161184310913086, "learning_rate": 6.171406626539436e-05, "loss": 0.0443, "step": 45350 }, { "grad_norm": 0.5915704369544983, "learning_rate": 6.169799110676398e-05, "loss": 0.0478, "step": 45360 }, { "grad_norm": 0.6058269143104553, "learning_rate": 6.168191466885885e-05, "loss": 0.0365, "step": 45370 }, { "grad_norm": 2.469165325164795, "learning_rate": 6.166583695343711e-05, "loss": 0.0308, "step": 45380 }, { "grad_norm": 0.5743553042411804, "learning_rate": 6.164975796225698e-05, "loss": 0.0346, "step": 45390 }, { "grad_norm": 0.48040643334388733, "learning_rate": 6.163367769707683e-05, "loss": 0.0358, "step": 45400 }, { "grad_norm": 0.3630375266075134, "learning_rate": 6.161759615965519e-05, "loss": 0.0367, "step": 45410 }, { "grad_norm": 0.37732791900634766, "learning_rate": 6.16015133517507e-05, "loss": 0.0542, "step": 45420 }, { "grad_norm": 0.4020897448062897, "learning_rate": 6.158542927512214e-05, "loss": 0.0373, "step": 45430 }, { "grad_norm": 0.641842782497406, "learning_rate": 6.156934393152846e-05, "loss": 0.0346, "step": 45440 }, { "grad_norm": 0.696502685546875, "learning_rate": 6.15532573227287e-05, "loss": 0.0289, "step": 45450 }, { "grad_norm": 1.0110794305801392, "learning_rate": 6.153716945048212e-05, "loss": 0.0449, "step": 45460 }, { "grad_norm": 0.5839661955833435, "learning_rate": 6.152108031654802e-05, "loss": 0.0428, "step": 45470 }, { "grad_norm": 0.41704457998275757, "learning_rate": 6.15049899226859e-05, "loss": 0.0414, "step": 45480 }, { "grad_norm": 0.6608023643493652, "learning_rate": 6.148889827065537e-05, "loss": 0.0662, "step": 45490 }, { "grad_norm": 0.6338849663734436, "learning_rate": 6.147280536221622e-05, "loss": 0.0574, "step": 45500 }, { "grad_norm": 0.5172257423400879, "learning_rate": 6.145671119912832e-05, "loss": 0.0482, "step": 45510 }, { "grad_norm": 0.42783159017562866, "learning_rate": 6.144061578315169e-05, "loss": 0.0493, "step": 45520 }, { "grad_norm": 0.6203200221061707, "learning_rate": 6.142451911604654e-05, "loss": 0.0391, "step": 45530 }, { "grad_norm": 0.35779350996017456, "learning_rate": 6.140842119957315e-05, "loss": 0.0373, "step": 45540 }, { "grad_norm": 0.43417930603027344, "learning_rate": 6.139232203549197e-05, "loss": 0.0373, "step": 45550 }, { "grad_norm": 0.37358883023262024, "learning_rate": 6.137622162556357e-05, "loss": 0.0327, "step": 45560 }, { "grad_norm": 0.48539265990257263, "learning_rate": 6.13601199715487e-05, "loss": 0.0377, "step": 45570 }, { "grad_norm": 0.40790796279907227, "learning_rate": 6.134401707520816e-05, "loss": 0.0369, "step": 45580 }, { "grad_norm": 0.7029542326927185, "learning_rate": 6.132791293830298e-05, "loss": 0.0403, "step": 45590 }, { "grad_norm": 0.5918486714363098, "learning_rate": 6.131180756259428e-05, "loss": 0.0303, "step": 45600 }, { "grad_norm": 0.3946326673030853, "learning_rate": 6.129570094984331e-05, "loss": 0.0451, "step": 45610 }, { "grad_norm": 0.6956355571746826, "learning_rate": 6.127959310181145e-05, "loss": 0.0458, "step": 45620 }, { "grad_norm": 0.5737610459327698, "learning_rate": 6.126348402026026e-05, "loss": 0.0545, "step": 45630 }, { "grad_norm": 0.4066620171070099, "learning_rate": 6.12473737069514e-05, "loss": 0.053, "step": 45640 }, { "grad_norm": 0.7387921810150146, "learning_rate": 6.123126216364665e-05, "loss": 0.0407, "step": 45650 }, { "grad_norm": 0.7911610007286072, "learning_rate": 6.121514939210797e-05, "loss": 0.0456, "step": 45660 }, { "grad_norm": 0.6414231061935425, "learning_rate": 6.119903539409741e-05, "loss": 0.0391, "step": 45670 }, { "grad_norm": 0.41641902923583984, "learning_rate": 6.118292017137716e-05, "loss": 0.0499, "step": 45680 }, { "grad_norm": 0.4647996127605438, "learning_rate": 6.116680372570959e-05, "loss": 0.0757, "step": 45690 }, { "grad_norm": 0.5878711342811584, "learning_rate": 6.115068605885713e-05, "loss": 0.0482, "step": 45700 }, { "grad_norm": 0.48275601863861084, "learning_rate": 6.113456717258243e-05, "loss": 0.0456, "step": 45710 }, { "grad_norm": 0.4536961019039154, "learning_rate": 6.11184470686482e-05, "loss": 0.0367, "step": 45720 }, { "grad_norm": 0.5151795744895935, "learning_rate": 6.110232574881731e-05, "loss": 0.0368, "step": 45730 }, { "grad_norm": 0.34834253787994385, "learning_rate": 6.108620321485277e-05, "loss": 0.0404, "step": 45740 }, { "grad_norm": 0.4572172164916992, "learning_rate": 6.107007946851773e-05, "loss": 0.0376, "step": 45750 }, { "grad_norm": 0.5823362469673157, "learning_rate": 6.105395451157542e-05, "loss": 0.0548, "step": 45760 }, { "grad_norm": 0.5699272155761719, "learning_rate": 6.103782834578928e-05, "loss": 0.038, "step": 45770 }, { "grad_norm": 0.9772704243659973, "learning_rate": 6.102170097292281e-05, "loss": 0.0386, "step": 45780 }, { "grad_norm": 0.4179628789424896, "learning_rate": 6.1005572394739716e-05, "loss": 0.053, "step": 45790 }, { "grad_norm": 0.7688419818878174, "learning_rate": 6.0989442613003765e-05, "loss": 0.0467, "step": 45800 }, { "grad_norm": 0.46336179971694946, "learning_rate": 6.097331162947888e-05, "loss": 0.0378, "step": 45810 }, { "grad_norm": 0.6838908791542053, "learning_rate": 6.095717944592914e-05, "loss": 0.0397, "step": 45820 }, { "grad_norm": 0.4060513973236084, "learning_rate": 6.094104606411873e-05, "loss": 0.062, "step": 45830 }, { "grad_norm": 0.5236443877220154, "learning_rate": 6.0924911485811966e-05, "loss": 0.0382, "step": 45840 }, { "grad_norm": 0.6925460696220398, "learning_rate": 6.090877571277331e-05, "loss": 0.0379, "step": 45850 }, { "grad_norm": 0.4258348345756531, "learning_rate": 6.089263874676736e-05, "loss": 0.0391, "step": 45860 }, { "grad_norm": 0.4086194932460785, "learning_rate": 6.0876500589558796e-05, "loss": 0.033, "step": 45870 }, { "grad_norm": 0.3936164677143097, "learning_rate": 6.086036124291248e-05, "loss": 0.0328, "step": 45880 }, { "grad_norm": 0.491117000579834, "learning_rate": 6.084422070859339e-05, "loss": 0.0437, "step": 45890 }, { "grad_norm": 0.41413015127182007, "learning_rate": 6.082807898836663e-05, "loss": 0.0568, "step": 45900 }, { "grad_norm": 0.5163335800170898, "learning_rate": 6.081193608399742e-05, "loss": 0.0356, "step": 45910 }, { "grad_norm": 0.842714250087738, "learning_rate": 6.0795791997251164e-05, "loss": 0.0413, "step": 45920 }, { "grad_norm": 0.5577239394187927, "learning_rate": 6.0779646729893294e-05, "loss": 0.0447, "step": 45930 }, { "grad_norm": 0.4090389013290405, "learning_rate": 6.0763500283689476e-05, "loss": 0.0318, "step": 45940 }, { "grad_norm": 1.7371827363967896, "learning_rate": 6.0747352660405455e-05, "loss": 0.0398, "step": 45950 }, { "grad_norm": 0.39944061636924744, "learning_rate": 6.073120386180709e-05, "loss": 0.0399, "step": 45960 }, { "grad_norm": 0.4045126140117645, "learning_rate": 6.0715053889660425e-05, "loss": 0.0484, "step": 45970 }, { "grad_norm": 0.5078149437904358, "learning_rate": 6.069890274573157e-05, "loss": 0.0416, "step": 45980 }, { "grad_norm": 0.4134160578250885, "learning_rate": 6.068275043178679e-05, "loss": 0.0387, "step": 45990 }, { "grad_norm": 0.4969329833984375, "learning_rate": 6.066659694959248e-05, "loss": 0.0355, "step": 46000 }, { "grad_norm": 0.512391209602356, "learning_rate": 6.0650442300915176e-05, "loss": 0.0333, "step": 46010 }, { "grad_norm": 0.42430564761161804, "learning_rate": 6.063428648752152e-05, "loss": 0.0317, "step": 46020 }, { "grad_norm": 0.7442826628684998, "learning_rate": 6.0618129511178266e-05, "loss": 0.0411, "step": 46030 }, { "grad_norm": 0.42601656913757324, "learning_rate": 6.060197137365234e-05, "loss": 0.0293, "step": 46040 }, { "grad_norm": 0.5687893629074097, "learning_rate": 6.058581207671077e-05, "loss": 0.044, "step": 46050 }, { "grad_norm": 0.45733293890953064, "learning_rate": 6.056965162212072e-05, "loss": 0.0377, "step": 46060 }, { "grad_norm": 0.5185626745223999, "learning_rate": 6.055349001164943e-05, "loss": 0.0439, "step": 46070 }, { "grad_norm": 0.6992849111557007, "learning_rate": 6.0537327247064347e-05, "loss": 0.0309, "step": 46080 }, { "grad_norm": 0.3421756327152252, "learning_rate": 6.0521163330133014e-05, "loss": 0.0522, "step": 46090 }, { "grad_norm": 0.709303617477417, "learning_rate": 6.050499826262306e-05, "loss": 0.0468, "step": 46100 }, { "grad_norm": 0.5964294075965881, "learning_rate": 6.0488832046302294e-05, "loss": 0.0444, "step": 46110 }, { "grad_norm": 0.4368632435798645, "learning_rate": 6.0472664682938626e-05, "loss": 0.0429, "step": 46120 }, { "grad_norm": 0.48397359251976013, "learning_rate": 6.045649617430009e-05, "loss": 0.0329, "step": 46130 }, { "grad_norm": 0.5453622341156006, "learning_rate": 6.0440326522154866e-05, "loss": 0.0437, "step": 46140 }, { "grad_norm": 0.5536027550697327, "learning_rate": 6.0424155728271224e-05, "loss": 0.0438, "step": 46150 }, { "grad_norm": 0.6921869516372681, "learning_rate": 6.040798379441758e-05, "loss": 0.0417, "step": 46160 }, { "grad_norm": 0.6430570483207703, "learning_rate": 6.0391810722362485e-05, "loss": 0.0367, "step": 46170 }, { "grad_norm": 0.49166667461395264, "learning_rate": 6.037563651387458e-05, "loss": 0.0374, "step": 46180 }, { "grad_norm": 0.5464339852333069, "learning_rate": 6.0359461170722666e-05, "loss": 0.0392, "step": 46190 }, { "grad_norm": 0.48282763361930847, "learning_rate": 6.034328469467566e-05, "loss": 0.046, "step": 46200 }, { "grad_norm": 0.5072340369224548, "learning_rate": 6.0327107087502596e-05, "loss": 0.0608, "step": 46210 }, { "grad_norm": 0.5271778702735901, "learning_rate": 6.031092835097262e-05, "loss": 0.0397, "step": 46220 }, { "grad_norm": 0.4291658103466034, "learning_rate": 6.0294748486855024e-05, "loss": 0.034, "step": 46230 }, { "grad_norm": 0.7136889696121216, "learning_rate": 6.0278567496919216e-05, "loss": 0.0513, "step": 46240 }, { "grad_norm": 0.6780015826225281, "learning_rate": 6.026238538293472e-05, "loss": 0.0455, "step": 46250 }, { "grad_norm": 0.2978726923465729, "learning_rate": 6.024620214667118e-05, "loss": 0.0288, "step": 46260 }, { "grad_norm": 0.422055184841156, "learning_rate": 6.0230017789898384e-05, "loss": 0.0288, "step": 46270 }, { "grad_norm": 0.5951910614967346, "learning_rate": 6.021383231438622e-05, "loss": 0.0315, "step": 46280 }, { "grad_norm": 0.7760651111602783, "learning_rate": 6.0197645721904704e-05, "loss": 0.0736, "step": 46290 }, { "grad_norm": 0.4167468249797821, "learning_rate": 6.0181458014224e-05, "loss": 0.0335, "step": 46300 }, { "grad_norm": 0.3626370131969452, "learning_rate": 6.016526919311435e-05, "loss": 0.0402, "step": 46310 }, { "grad_norm": 0.6222135424613953, "learning_rate": 6.014907926034613e-05, "loss": 0.0348, "step": 46320 }, { "grad_norm": 0.7310493588447571, "learning_rate": 6.0132888217689875e-05, "loss": 0.0363, "step": 46330 }, { "grad_norm": 0.5092610716819763, "learning_rate": 6.01166960669162e-05, "loss": 0.0458, "step": 46340 }, { "grad_norm": 0.5146347880363464, "learning_rate": 6.0100502809795845e-05, "loss": 0.034, "step": 46350 }, { "grad_norm": 0.3970588743686676, "learning_rate": 6.00843084480997e-05, "loss": 0.0321, "step": 46360 }, { "grad_norm": 0.5574559569358826, "learning_rate": 6.0068112983598736e-05, "loss": 0.0491, "step": 46370 }, { "grad_norm": 0.6464364528656006, "learning_rate": 6.005191641806407e-05, "loss": 0.0302, "step": 46380 }, { "grad_norm": 0.3826003968715668, "learning_rate": 6.003571875326694e-05, "loss": 0.0314, "step": 46390 }, { "grad_norm": 0.8458762168884277, "learning_rate": 6.00195199909787e-05, "loss": 0.0364, "step": 46400 }, { "grad_norm": 0.5238388180732727, "learning_rate": 6.000332013297082e-05, "loss": 0.039, "step": 46410 }, { "grad_norm": 0.3330869674682617, "learning_rate": 5.998711918101487e-05, "loss": 0.0336, "step": 46420 }, { "grad_norm": 0.6391535401344299, "learning_rate": 5.997091713688261e-05, "loss": 0.0431, "step": 46430 }, { "grad_norm": 0.3917573392391205, "learning_rate": 5.9954714002345836e-05, "loss": 0.0358, "step": 46440 }, { "grad_norm": 0.48452624678611755, "learning_rate": 5.993850977917649e-05, "loss": 0.0409, "step": 46450 }, { "grad_norm": 0.35573723912239075, "learning_rate": 5.992230446914667e-05, "loss": 0.0495, "step": 46460 }, { "grad_norm": 0.4620383083820343, "learning_rate": 5.990609807402855e-05, "loss": 0.0436, "step": 46470 }, { "grad_norm": 0.5387977361679077, "learning_rate": 5.988989059559443e-05, "loss": 0.0357, "step": 46480 }, { "grad_norm": 0.32512515783309937, "learning_rate": 5.987368203561675e-05, "loss": 0.0414, "step": 46490 }, { "grad_norm": 0.5752952694892883, "learning_rate": 5.9857472395868055e-05, "loss": 0.0366, "step": 46500 }, { "grad_norm": 0.38918739557266235, "learning_rate": 5.9841261678120983e-05, "loss": 0.0337, "step": 46510 }, { "grad_norm": 2.110589027404785, "learning_rate": 5.982504988414834e-05, "loss": 0.0387, "step": 46520 }, { "grad_norm": 0.6426323056221008, "learning_rate": 5.9808837015723015e-05, "loss": 0.0287, "step": 46530 }, { "grad_norm": 1.1169507503509521, "learning_rate": 5.9792623074618016e-05, "loss": 0.0521, "step": 46540 }, { "grad_norm": 0.5741776823997498, "learning_rate": 5.977640806260648e-05, "loss": 0.0408, "step": 46550 }, { "grad_norm": 0.4246945083141327, "learning_rate": 5.976019198146166e-05, "loss": 0.0458, "step": 46560 }, { "grad_norm": 0.7887396812438965, "learning_rate": 5.974397483295692e-05, "loss": 0.0515, "step": 46570 }, { "grad_norm": 0.4137740731239319, "learning_rate": 5.972775661886575e-05, "loss": 0.0297, "step": 46580 }, { "grad_norm": 0.48000824451446533, "learning_rate": 5.971153734096173e-05, "loss": 0.0384, "step": 46590 }, { "grad_norm": 0.68959641456604, "learning_rate": 5.969531700101859e-05, "loss": 0.0552, "step": 46600 }, { "grad_norm": 0.36638543009757996, "learning_rate": 5.9679095600810155e-05, "loss": 0.0388, "step": 46610 }, { "grad_norm": 0.362493634223938, "learning_rate": 5.9662873142110384e-05, "loss": 0.0496, "step": 46620 }, { "grad_norm": 0.4587024748325348, "learning_rate": 5.964664962669333e-05, "loss": 0.0355, "step": 46630 }, { "grad_norm": 0.5570513606071472, "learning_rate": 5.9630425056333186e-05, "loss": 0.0409, "step": 46640 }, { "grad_norm": 0.5475870966911316, "learning_rate": 5.961419943280422e-05, "loss": 0.0319, "step": 46650 }, { "grad_norm": 0.2564779222011566, "learning_rate": 5.959797275788087e-05, "loss": 0.0439, "step": 46660 }, { "grad_norm": 0.42731454968452454, "learning_rate": 5.958174503333765e-05, "loss": 0.0384, "step": 46670 }, { "grad_norm": 0.5424349308013916, "learning_rate": 5.9565516260949195e-05, "loss": 0.0445, "step": 46680 }, { "grad_norm": 0.3533601462841034, "learning_rate": 5.9549286442490273e-05, "loss": 0.0308, "step": 46690 }, { "grad_norm": 0.3725772798061371, "learning_rate": 5.953305557973572e-05, "loss": 0.0513, "step": 46700 }, { "grad_norm": 0.3937251567840576, "learning_rate": 5.9516823674460564e-05, "loss": 0.0307, "step": 46710 }, { "grad_norm": 1.5597376823425293, "learning_rate": 5.9500590728439875e-05, "loss": 0.049, "step": 46720 }, { "grad_norm": 0.5936490893363953, "learning_rate": 5.948435674344886e-05, "loss": 0.0373, "step": 46730 }, { "grad_norm": 0.49974721670150757, "learning_rate": 5.946812172126285e-05, "loss": 0.0543, "step": 46740 }, { "grad_norm": 0.4534912407398224, "learning_rate": 5.945188566365729e-05, "loss": 0.0485, "step": 46750 }, { "grad_norm": 0.8322950601577759, "learning_rate": 5.943564857240773e-05, "loss": 0.0596, "step": 46760 }, { "grad_norm": 0.49303045868873596, "learning_rate": 5.941941044928981e-05, "loss": 0.0423, "step": 46770 }, { "grad_norm": 0.4963110387325287, "learning_rate": 5.940317129607935e-05, "loss": 0.0758, "step": 46780 }, { "grad_norm": 0.25526538491249084, "learning_rate": 5.9386931114552204e-05, "loss": 0.0346, "step": 46790 }, { "grad_norm": 0.47202950716018677, "learning_rate": 5.937068990648438e-05, "loss": 0.0671, "step": 46800 }, { "grad_norm": 0.4231981933116913, "learning_rate": 5.935444767365199e-05, "loss": 0.0544, "step": 46810 }, { "grad_norm": 0.7897409796714783, "learning_rate": 5.933820441783129e-05, "loss": 0.0388, "step": 46820 }, { "grad_norm": 0.4406619369983673, "learning_rate": 5.932196014079857e-05, "loss": 0.0428, "step": 46830 }, { "grad_norm": 0.397754430770874, "learning_rate": 5.930571484433032e-05, "loss": 0.0299, "step": 46840 }, { "grad_norm": 0.6866399645805359, "learning_rate": 5.928946853020309e-05, "loss": 0.0421, "step": 46850 }, { "grad_norm": 0.5879068970680237, "learning_rate": 5.927322120019352e-05, "loss": 0.0467, "step": 46860 }, { "grad_norm": 1.0448890924453735, "learning_rate": 5.925697285607844e-05, "loss": 0.0326, "step": 46870 }, { "grad_norm": 0.8869717717170715, "learning_rate": 5.924072349963472e-05, "loss": 0.0392, "step": 46880 }, { "grad_norm": 0.49579450488090515, "learning_rate": 5.922447313263938e-05, "loss": 0.0426, "step": 46890 }, { "grad_norm": 0.6034950017929077, "learning_rate": 5.920822175686952e-05, "loss": 0.0526, "step": 46900 }, { "grad_norm": 0.3562045693397522, "learning_rate": 5.919196937410239e-05, "loss": 0.043, "step": 46910 }, { "grad_norm": 0.6988388895988464, "learning_rate": 5.917571598611529e-05, "loss": 0.0364, "step": 46920 }, { "grad_norm": 0.6728503704071045, "learning_rate": 5.9159461594685686e-05, "loss": 0.0349, "step": 46930 }, { "grad_norm": 0.4933256506919861, "learning_rate": 5.914320620159114e-05, "loss": 0.0466, "step": 46940 }, { "grad_norm": 0.4733797609806061, "learning_rate": 5.912694980860931e-05, "loss": 0.0428, "step": 46950 }, { "grad_norm": 0.3030823767185211, "learning_rate": 5.9110692417517964e-05, "loss": 0.0247, "step": 46960 }, { "grad_norm": 0.3514469265937805, "learning_rate": 5.9094434030095e-05, "loss": 0.0454, "step": 46970 }, { "grad_norm": 0.3567997217178345, "learning_rate": 5.907817464811841e-05, "loss": 0.046, "step": 46980 }, { "grad_norm": 0.5863044261932373, "learning_rate": 5.906191427336627e-05, "loss": 0.0484, "step": 46990 }, { "grad_norm": 0.36462023854255676, "learning_rate": 5.9045652907616835e-05, "loss": 0.0773, "step": 47000 }, { "grad_norm": 0.39290034770965576, "learning_rate": 5.902939055264838e-05, "loss": 0.0479, "step": 47010 }, { "grad_norm": 0.7403133511543274, "learning_rate": 5.901312721023935e-05, "loss": 0.0352, "step": 47020 }, { "grad_norm": 0.7975200414657593, "learning_rate": 5.8996862882168294e-05, "loss": 0.0749, "step": 47030 }, { "grad_norm": 0.6550228595733643, "learning_rate": 5.8980597570213826e-05, "loss": 0.0528, "step": 47040 }, { "grad_norm": 0.6139930486679077, "learning_rate": 5.896433127615471e-05, "loss": 0.035, "step": 47050 }, { "grad_norm": 0.4444611966609955, "learning_rate": 5.894806400176981e-05, "loss": 0.0377, "step": 47060 }, { "grad_norm": 0.4563363492488861, "learning_rate": 5.893179574883808e-05, "loss": 0.0531, "step": 47070 }, { "grad_norm": 0.4357072114944458, "learning_rate": 5.8915526519138585e-05, "loss": 0.0421, "step": 47080 }, { "grad_norm": 0.49653586745262146, "learning_rate": 5.889925631445053e-05, "loss": 0.0408, "step": 47090 }, { "grad_norm": 0.4943198561668396, "learning_rate": 5.888298513655318e-05, "loss": 0.0313, "step": 47100 }, { "grad_norm": 0.6325770616531372, "learning_rate": 5.886671298722592e-05, "loss": 0.0385, "step": 47110 }, { "grad_norm": 1.1215075254440308, "learning_rate": 5.885043986824828e-05, "loss": 0.0357, "step": 47120 }, { "grad_norm": 0.23863890767097473, "learning_rate": 5.883416578139982e-05, "loss": 0.0365, "step": 47130 }, { "grad_norm": 0.5605230927467346, "learning_rate": 5.881789072846029e-05, "loss": 0.0353, "step": 47140 }, { "grad_norm": 0.7920058965682983, "learning_rate": 5.8801614711209474e-05, "loss": 0.0345, "step": 47150 }, { "grad_norm": 0.3855210244655609, "learning_rate": 5.878533773142732e-05, "loss": 0.0407, "step": 47160 }, { "grad_norm": 0.3933066129684448, "learning_rate": 5.8769059790893834e-05, "loss": 0.0381, "step": 47170 }, { "grad_norm": 0.4729079008102417, "learning_rate": 5.8752780891389145e-05, "loss": 0.0613, "step": 47180 }, { "grad_norm": 0.3636566698551178, "learning_rate": 5.873650103469351e-05, "loss": 0.0481, "step": 47190 }, { "grad_norm": 0.5492238402366638, "learning_rate": 5.872022022258725e-05, "loss": 0.0446, "step": 47200 }, { "grad_norm": 0.46826836466789246, "learning_rate": 5.8703938456850814e-05, "loss": 0.0543, "step": 47210 }, { "grad_norm": 1.0038912296295166, "learning_rate": 5.868765573926477e-05, "loss": 0.0327, "step": 47220 }, { "grad_norm": 0.45307278633117676, "learning_rate": 5.867137207160974e-05, "loss": 0.0342, "step": 47230 }, { "grad_norm": 0.471038818359375, "learning_rate": 5.86550874556665e-05, "loss": 0.0361, "step": 47240 }, { "grad_norm": 0.3293292820453644, "learning_rate": 5.863880189321592e-05, "loss": 0.0376, "step": 47250 }, { "grad_norm": 0.41267839074134827, "learning_rate": 5.862251538603894e-05, "loss": 0.0483, "step": 47260 }, { "grad_norm": 0.37480616569519043, "learning_rate": 5.8606227935916656e-05, "loss": 0.0447, "step": 47270 }, { "grad_norm": 0.5969026684761047, "learning_rate": 5.858993954463021e-05, "loss": 0.0524, "step": 47280 }, { "grad_norm": 0.28706398606300354, "learning_rate": 5.8573650213960896e-05, "loss": 0.0316, "step": 47290 }, { "grad_norm": 1.710071086883545, "learning_rate": 5.8557359945690084e-05, "loss": 0.0503, "step": 47300 }, { "grad_norm": 0.3224284052848816, "learning_rate": 5.8541068741599246e-05, "loss": 0.0334, "step": 47310 }, { "grad_norm": 0.30049481987953186, "learning_rate": 5.8524776603469985e-05, "loss": 0.0297, "step": 47320 }, { "grad_norm": 0.4913868010044098, "learning_rate": 5.850848353308397e-05, "loss": 0.0391, "step": 47330 }, { "grad_norm": 0.5125539898872375, "learning_rate": 5.849218953222297e-05, "loss": 0.0301, "step": 47340 }, { "grad_norm": 0.8712201118469238, "learning_rate": 5.847589460266891e-05, "loss": 0.0532, "step": 47350 }, { "grad_norm": 0.541245698928833, "learning_rate": 5.845959874620375e-05, "loss": 0.0546, "step": 47360 }, { "grad_norm": 0.43786123394966125, "learning_rate": 5.84433019646096e-05, "loss": 0.0356, "step": 47370 }, { "grad_norm": 0.3137793242931366, "learning_rate": 5.842700425966863e-05, "loss": 0.0297, "step": 47380 }, { "grad_norm": 0.3271837830543518, "learning_rate": 5.841070563316315e-05, "loss": 0.0412, "step": 47390 }, { "grad_norm": 0.4429480731487274, "learning_rate": 5.839440608687554e-05, "loss": 0.038, "step": 47400 }, { "grad_norm": 0.4595009386539459, "learning_rate": 5.837810562258831e-05, "loss": 0.0402, "step": 47410 }, { "grad_norm": 0.4882444739341736, "learning_rate": 5.836180424208405e-05, "loss": 0.036, "step": 47420 }, { "grad_norm": 0.5506543517112732, "learning_rate": 5.834550194714543e-05, "loss": 0.0401, "step": 47430 }, { "grad_norm": 0.3063279390335083, "learning_rate": 5.832919873955528e-05, "loss": 0.0338, "step": 47440 }, { "grad_norm": 0.4381895661354065, "learning_rate": 5.831289462109648e-05, "loss": 0.0424, "step": 47450 }, { "grad_norm": 0.6903694272041321, "learning_rate": 5.829658959355202e-05, "loss": 0.0423, "step": 47460 }, { "grad_norm": 0.5775061845779419, "learning_rate": 5.8280283658704995e-05, "loss": 0.0368, "step": 47470 }, { "grad_norm": 0.47947195172309875, "learning_rate": 5.826397681833859e-05, "loss": 0.0415, "step": 47480 }, { "grad_norm": 0.4437359571456909, "learning_rate": 5.824766907423612e-05, "loss": 0.0342, "step": 47490 }, { "grad_norm": 0.5792703628540039, "learning_rate": 5.8231360428180945e-05, "loss": 0.035, "step": 47500 }, { "grad_norm": 0.4532638490200043, "learning_rate": 5.821505088195658e-05, "loss": 0.0399, "step": 47510 }, { "grad_norm": 0.3504635989665985, "learning_rate": 5.819874043734661e-05, "loss": 0.0364, "step": 47520 }, { "grad_norm": 2.000516414642334, "learning_rate": 5.8182429096134695e-05, "loss": 0.0323, "step": 47530 }, { "grad_norm": 0.4500810503959656, "learning_rate": 5.816611686010465e-05, "loss": 0.0424, "step": 47540 }, { "grad_norm": 0.6095982193946838, "learning_rate": 5.814980373104033e-05, "loss": 0.0401, "step": 47550 }, { "grad_norm": 0.45677822828292847, "learning_rate": 5.813348971072572e-05, "loss": 0.0408, "step": 47560 }, { "grad_norm": 0.6718471050262451, "learning_rate": 5.811717480094492e-05, "loss": 0.036, "step": 47570 }, { "grad_norm": 0.5439592599868774, "learning_rate": 5.810085900348209e-05, "loss": 0.0393, "step": 47580 }, { "grad_norm": 0.8159199357032776, "learning_rate": 5.8084542320121483e-05, "loss": 0.0406, "step": 47590 }, { "grad_norm": 0.9847373962402344, "learning_rate": 5.8068224752647497e-05, "loss": 0.0366, "step": 47600 }, { "grad_norm": 0.48955237865448, "learning_rate": 5.805190630284457e-05, "loss": 0.0314, "step": 47610 }, { "grad_norm": 0.3159424662590027, "learning_rate": 5.803558697249729e-05, "loss": 0.0436, "step": 47620 }, { "grad_norm": 1.0164622068405151, "learning_rate": 5.8019266763390276e-05, "loss": 0.0396, "step": 47630 }, { "grad_norm": 0.5813717842102051, "learning_rate": 5.800294567730833e-05, "loss": 0.0327, "step": 47640 }, { "grad_norm": 0.44126611948013306, "learning_rate": 5.798662371603628e-05, "loss": 0.0335, "step": 47650 }, { "grad_norm": 0.4284457266330719, "learning_rate": 5.797030088135904e-05, "loss": 0.0408, "step": 47660 }, { "grad_norm": 0.562430739402771, "learning_rate": 5.7953977175061705e-05, "loss": 0.0315, "step": 47670 }, { "grad_norm": 0.6695346236228943, "learning_rate": 5.793765259892938e-05, "loss": 0.0419, "step": 47680 }, { "grad_norm": 0.5891830921173096, "learning_rate": 5.792132715474729e-05, "loss": 0.0329, "step": 47690 }, { "grad_norm": 0.6158193349838257, "learning_rate": 5.790500084430078e-05, "loss": 0.036, "step": 47700 }, { "grad_norm": 0.38724955916404724, "learning_rate": 5.7888673669375274e-05, "loss": 0.0353, "step": 47710 }, { "grad_norm": 0.7279627919197083, "learning_rate": 5.787234563175625e-05, "loss": 0.0395, "step": 47720 }, { "grad_norm": 0.5389918684959412, "learning_rate": 5.7856016733229355e-05, "loss": 0.0539, "step": 47730 }, { "grad_norm": 0.4777320921421051, "learning_rate": 5.7839686975580297e-05, "loss": 0.0384, "step": 47740 }, { "grad_norm": 0.41714543104171753, "learning_rate": 5.782335636059484e-05, "loss": 0.0335, "step": 47750 }, { "grad_norm": 0.519906759262085, "learning_rate": 5.780702489005889e-05, "loss": 0.0375, "step": 47760 }, { "grad_norm": 0.36957451701164246, "learning_rate": 5.779069256575845e-05, "loss": 0.0338, "step": 47770 }, { "grad_norm": 2.9784088134765625, "learning_rate": 5.7774359389479574e-05, "loss": 0.0645, "step": 47780 }, { "grad_norm": 0.47374433279037476, "learning_rate": 5.775802536300845e-05, "loss": 0.0362, "step": 47790 }, { "grad_norm": 0.8532683253288269, "learning_rate": 5.774169048813134e-05, "loss": 0.0346, "step": 47800 }, { "grad_norm": 0.4654887914657593, "learning_rate": 5.77253547666346e-05, "loss": 0.038, "step": 47810 }, { "grad_norm": 0.5415883660316467, "learning_rate": 5.770901820030465e-05, "loss": 0.0264, "step": 47820 }, { "grad_norm": 0.7120095491409302, "learning_rate": 5.769268079092809e-05, "loss": 0.0509, "step": 47830 }, { "grad_norm": 0.48915231227874756, "learning_rate": 5.767634254029151e-05, "loss": 0.0439, "step": 47840 }, { "grad_norm": 0.4115801751613617, "learning_rate": 5.7660003450181655e-05, "loss": 0.0308, "step": 47850 }, { "grad_norm": 0.2613096535205841, "learning_rate": 5.764366352238534e-05, "loss": 0.0309, "step": 47860 }, { "grad_norm": 0.4989013969898224, "learning_rate": 5.7627322758689474e-05, "loss": 0.032, "step": 47870 }, { "grad_norm": 0.4276750683784485, "learning_rate": 5.761098116088105e-05, "loss": 0.0313, "step": 47880 }, { "grad_norm": 0.4359484016895294, "learning_rate": 5.759463873074717e-05, "loss": 0.0312, "step": 47890 }, { "grad_norm": 0.4497110843658447, "learning_rate": 5.757829547007504e-05, "loss": 0.0294, "step": 47900 }, { "grad_norm": 0.4523111581802368, "learning_rate": 5.756195138065189e-05, "loss": 0.0494, "step": 47910 }, { "grad_norm": 0.4125995337963104, "learning_rate": 5.754560646426511e-05, "loss": 0.0489, "step": 47920 }, { "grad_norm": 0.5186487436294556, "learning_rate": 5.752926072270216e-05, "loss": 0.0402, "step": 47930 }, { "grad_norm": 0.3748646378517151, "learning_rate": 5.7512914157750563e-05, "loss": 0.0418, "step": 47940 }, { "grad_norm": 0.5864193439483643, "learning_rate": 5.749656677119798e-05, "loss": 0.033, "step": 47950 }, { "grad_norm": 0.5562973022460938, "learning_rate": 5.7480218564832125e-05, "loss": 0.0411, "step": 47960 }, { "grad_norm": 0.5695817470550537, "learning_rate": 5.746386954044082e-05, "loss": 0.0436, "step": 47970 }, { "grad_norm": 0.5415360927581787, "learning_rate": 5.744751969981195e-05, "loss": 0.0412, "step": 47980 }, { "grad_norm": 0.5599923133850098, "learning_rate": 5.7431169044733526e-05, "loss": 0.0364, "step": 47990 }, { "grad_norm": 0.9092739820480347, "learning_rate": 5.741481757699364e-05, "loss": 0.0428, "step": 48000 }, { "grad_norm": 0.39030399918556213, "learning_rate": 5.7398465298380434e-05, "loss": 0.0397, "step": 48010 }, { "grad_norm": 0.5111032128334045, "learning_rate": 5.7382112210682193e-05, "loss": 0.0284, "step": 48020 }, { "grad_norm": 0.3659689128398895, "learning_rate": 5.7365758315687266e-05, "loss": 0.0342, "step": 48030 }, { "grad_norm": 0.4372515082359314, "learning_rate": 5.734940361518407e-05, "loss": 0.0281, "step": 48040 }, { "grad_norm": 0.37658634781837463, "learning_rate": 5.733304811096116e-05, "loss": 0.0306, "step": 48050 }, { "grad_norm": 0.6127350330352783, "learning_rate": 5.731669180480713e-05, "loss": 0.0409, "step": 48060 }, { "grad_norm": 0.5230758786201477, "learning_rate": 5.730033469851067e-05, "loss": 0.0325, "step": 48070 }, { "grad_norm": 0.5683586597442627, "learning_rate": 5.7283976793860607e-05, "loss": 0.0448, "step": 48080 }, { "grad_norm": 0.516910970211029, "learning_rate": 5.726761809264577e-05, "loss": 0.0449, "step": 48090 }, { "grad_norm": 0.8700975775718689, "learning_rate": 5.7251258596655155e-05, "loss": 0.0352, "step": 48100 }, { "grad_norm": 0.4234963059425354, "learning_rate": 5.72348983076778e-05, "loss": 0.0236, "step": 48110 }, { "grad_norm": 0.3810742497444153, "learning_rate": 5.7218537227502854e-05, "loss": 0.0696, "step": 48120 }, { "grad_norm": 0.7972449660301208, "learning_rate": 5.720217535791951e-05, "loss": 0.0441, "step": 48130 }, { "grad_norm": 0.45659056305885315, "learning_rate": 5.718581270071711e-05, "loss": 0.0321, "step": 48140 }, { "grad_norm": 0.40444424748420715, "learning_rate": 5.716944925768505e-05, "loss": 0.0461, "step": 48150 }, { "grad_norm": 0.4505762755870819, "learning_rate": 5.7153085030612786e-05, "loss": 0.0348, "step": 48160 }, { "grad_norm": 0.5144041776657104, "learning_rate": 5.713672002128989e-05, "loss": 0.0455, "step": 48170 }, { "grad_norm": 0.3935360312461853, "learning_rate": 5.712035423150602e-05, "loss": 0.0284, "step": 48180 }, { "grad_norm": 0.42364516854286194, "learning_rate": 5.710398766305094e-05, "loss": 0.0512, "step": 48190 }, { "grad_norm": 0.559459388256073, "learning_rate": 5.708762031771442e-05, "loss": 0.0301, "step": 48200 }, { "grad_norm": 0.5136794447898865, "learning_rate": 5.707125219728642e-05, "loss": 0.0425, "step": 48210 }, { "grad_norm": 0.28613683581352234, "learning_rate": 5.7054883303556905e-05, "loss": 0.0288, "step": 48220 }, { "grad_norm": 0.4951804280281067, "learning_rate": 5.703851363831595e-05, "loss": 0.0242, "step": 48230 }, { "grad_norm": 0.444844126701355, "learning_rate": 5.702214320335374e-05, "loss": 0.0373, "step": 48240 }, { "grad_norm": 0.3169440031051636, "learning_rate": 5.7005772000460514e-05, "loss": 0.0453, "step": 48250 }, { "grad_norm": 0.5095341801643372, "learning_rate": 5.698940003142656e-05, "loss": 0.0445, "step": 48260 }, { "grad_norm": 0.9441083669662476, "learning_rate": 5.697302729804236e-05, "loss": 0.0349, "step": 48270 }, { "grad_norm": 0.43760690093040466, "learning_rate": 5.695665380209837e-05, "loss": 0.0425, "step": 48280 }, { "grad_norm": 0.250860333442688, "learning_rate": 5.6940279545385165e-05, "loss": 0.0299, "step": 48290 }, { "grad_norm": 0.5779277682304382, "learning_rate": 5.692390452969344e-05, "loss": 0.034, "step": 48300 }, { "grad_norm": 1.1514688730239868, "learning_rate": 5.690752875681392e-05, "loss": 0.041, "step": 48310 }, { "grad_norm": 0.45400792360305786, "learning_rate": 5.6891152228537435e-05, "loss": 0.045, "step": 48320 }, { "grad_norm": 0.4857809245586395, "learning_rate": 5.687477494665492e-05, "loss": 0.0311, "step": 48330 }, { "grad_norm": 0.49224403500556946, "learning_rate": 5.685839691295733e-05, "loss": 0.034, "step": 48340 }, { "grad_norm": 0.554252564907074, "learning_rate": 5.6842018129235786e-05, "loss": 0.0391, "step": 48350 }, { "grad_norm": 0.5727197527885437, "learning_rate": 5.6825638597281404e-05, "loss": 0.0343, "step": 48360 }, { "grad_norm": 0.6183748245239258, "learning_rate": 5.680925831888546e-05, "loss": 0.0348, "step": 48370 }, { "grad_norm": 0.5433884263038635, "learning_rate": 5.6792877295839274e-05, "loss": 0.0315, "step": 48380 }, { "grad_norm": 0.29810595512390137, "learning_rate": 5.6776495529934224e-05, "loss": 0.0259, "step": 48390 }, { "grad_norm": 0.4995187819004059, "learning_rate": 5.6760113022961824e-05, "loss": 0.0312, "step": 48400 }, { "grad_norm": 0.44556859135627747, "learning_rate": 5.6743729776713617e-05, "loss": 0.028, "step": 48410 }, { "grad_norm": 0.5476740598678589, "learning_rate": 5.672734579298126e-05, "loss": 0.0342, "step": 48420 }, { "grad_norm": 0.40944260358810425, "learning_rate": 5.671096107355649e-05, "loss": 0.0406, "step": 48430 }, { "grad_norm": 0.6138590574264526, "learning_rate": 5.66945756202311e-05, "loss": 0.0526, "step": 48440 }, { "grad_norm": 0.39492183923721313, "learning_rate": 5.667818943479699e-05, "loss": 0.037, "step": 48450 }, { "grad_norm": 0.4991172254085541, "learning_rate": 5.666180251904612e-05, "loss": 0.0374, "step": 48460 }, { "grad_norm": 0.5685911178588867, "learning_rate": 5.6645414874770555e-05, "loss": 0.032, "step": 48470 }, { "grad_norm": 0.5011997818946838, "learning_rate": 5.66290265037624e-05, "loss": 0.0408, "step": 48480 }, { "grad_norm": 0.4303417503833771, "learning_rate": 5.661263740781386e-05, "loss": 0.0332, "step": 48490 }, { "grad_norm": 0.4993404150009155, "learning_rate": 5.6596247588717254e-05, "loss": 0.0411, "step": 48500 }, { "grad_norm": 0.4936252534389496, "learning_rate": 5.6579857048264926e-05, "loss": 0.0342, "step": 48510 }, { "grad_norm": 0.5233039259910583, "learning_rate": 5.6563465788249314e-05, "loss": 0.036, "step": 48520 }, { "grad_norm": 0.5633476972579956, "learning_rate": 5.6547073810462956e-05, "loss": 0.0381, "step": 48530 }, { "grad_norm": 0.8464091420173645, "learning_rate": 5.653068111669846e-05, "loss": 0.0385, "step": 48540 }, { "grad_norm": 0.7335649132728577, "learning_rate": 5.651428770874848e-05, "loss": 0.0326, "step": 48550 }, { "grad_norm": 0.5587158799171448, "learning_rate": 5.64978935884058e-05, "loss": 0.0411, "step": 48560 }, { "grad_norm": 0.9002370238304138, "learning_rate": 5.6481498757463244e-05, "loss": 0.0441, "step": 48570 }, { "grad_norm": 0.6654734015464783, "learning_rate": 5.646510321771373e-05, "loss": 0.0544, "step": 48580 }, { "grad_norm": 0.7066807150840759, "learning_rate": 5.644870697095024e-05, "loss": 0.055, "step": 48590 }, { "grad_norm": 0.5506753921508789, "learning_rate": 5.643231001896586e-05, "loss": 0.0333, "step": 48600 }, { "grad_norm": 0.7506044507026672, "learning_rate": 5.6415912363553726e-05, "loss": 0.0303, "step": 48610 }, { "grad_norm": 0.4439100921154022, "learning_rate": 5.639951400650706e-05, "loss": 0.0526, "step": 48620 }, { "grad_norm": 0.4903976023197174, "learning_rate": 5.6383114949619165e-05, "loss": 0.0337, "step": 48630 }, { "grad_norm": 0.7725608348846436, "learning_rate": 5.636671519468342e-05, "loss": 0.0381, "step": 48640 }, { "grad_norm": 0.5154516100883484, "learning_rate": 5.635031474349327e-05, "loss": 0.0377, "step": 48650 }, { "grad_norm": 0.33049699664115906, "learning_rate": 5.6333913597842246e-05, "loss": 0.034, "step": 48660 }, { "grad_norm": 0.33392810821533203, "learning_rate": 5.6317511759523955e-05, "loss": 0.0329, "step": 48670 }, { "grad_norm": 0.5927185416221619, "learning_rate": 5.630110923033207e-05, "loss": 0.0359, "step": 48680 }, { "grad_norm": 0.36896035075187683, "learning_rate": 5.628470601206036e-05, "loss": 0.0459, "step": 48690 }, { "grad_norm": 0.3958703279495239, "learning_rate": 5.626830210650266e-05, "loss": 0.0405, "step": 48700 }, { "grad_norm": 0.3583175539970398, "learning_rate": 5.625189751545285e-05, "loss": 0.0406, "step": 48710 }, { "grad_norm": 0.38857612013816833, "learning_rate": 5.6235492240704936e-05, "loss": 0.0299, "step": 48720 }, { "grad_norm": 0.3896867036819458, "learning_rate": 5.621908628405296e-05, "loss": 0.0377, "step": 48730 }, { "grad_norm": 0.44088441133499146, "learning_rate": 5.620267964729106e-05, "loss": 0.0272, "step": 48740 }, { "grad_norm": 0.6173848509788513, "learning_rate": 5.618627233221344e-05, "loss": 0.0304, "step": 48750 }, { "grad_norm": 0.43577006459236145, "learning_rate": 5.616986434061438e-05, "loss": 0.0278, "step": 48760 }, { "grad_norm": 0.6405341029167175, "learning_rate": 5.615345567428822e-05, "loss": 0.0325, "step": 48770 }, { "grad_norm": 0.3999653160572052, "learning_rate": 5.613704633502941e-05, "loss": 0.0365, "step": 48780 }, { "grad_norm": 0.27459391951560974, "learning_rate": 5.612063632463245e-05, "loss": 0.0285, "step": 48790 }, { "grad_norm": 0.8107383251190186, "learning_rate": 5.610422564489188e-05, "loss": 0.0515, "step": 48800 }, { "grad_norm": 0.5076420903205872, "learning_rate": 5.608781429760239e-05, "loss": 0.0394, "step": 48810 }, { "grad_norm": 1.1868773698806763, "learning_rate": 5.607140228455866e-05, "loss": 0.0327, "step": 48820 }, { "grad_norm": 0.4985607862472534, "learning_rate": 5.605498960755553e-05, "loss": 0.0415, "step": 48830 }, { "grad_norm": 0.7974625825881958, "learning_rate": 5.603857626838782e-05, "loss": 0.0408, "step": 48840 }, { "grad_norm": 0.6240688562393188, "learning_rate": 5.60221622688505e-05, "loss": 0.0373, "step": 48850 }, { "grad_norm": 0.6267910599708557, "learning_rate": 5.6005747610738565e-05, "loss": 0.0425, "step": 48860 }, { "grad_norm": 0.3552992343902588, "learning_rate": 5.59893322958471e-05, "loss": 0.0352, "step": 48870 }, { "grad_norm": 0.3619730770587921, "learning_rate": 5.5972916325971256e-05, "loss": 0.0277, "step": 48880 }, { "grad_norm": 0.3501209318637848, "learning_rate": 5.595649970290628e-05, "loss": 0.03, "step": 48890 }, { "grad_norm": 0.8668941855430603, "learning_rate": 5.5940082428447426e-05, "loss": 0.054, "step": 48900 }, { "grad_norm": 0.44083812832832336, "learning_rate": 5.592366450439012e-05, "loss": 0.0312, "step": 48910 }, { "grad_norm": 0.34787046909332275, "learning_rate": 5.590724593252975e-05, "loss": 0.0341, "step": 48920 }, { "grad_norm": 0.4094350337982178, "learning_rate": 5.589082671466184e-05, "loss": 0.0376, "step": 48930 }, { "grad_norm": 0.4093480110168457, "learning_rate": 5.587440685258199e-05, "loss": 0.0391, "step": 48940 }, { "grad_norm": 0.2603226602077484, "learning_rate": 5.585798634808583e-05, "loss": 0.0331, "step": 48950 }, { "grad_norm": 0.4044131338596344, "learning_rate": 5.584156520296909e-05, "loss": 0.0377, "step": 48960 }, { "grad_norm": 0.4902399182319641, "learning_rate": 5.582514341902757e-05, "loss": 0.0328, "step": 48970 }, { "grad_norm": 0.4806082546710968, "learning_rate": 5.580872099805713e-05, "loss": 0.0375, "step": 48980 }, { "grad_norm": 0.5215768814086914, "learning_rate": 5.5792297941853674e-05, "loss": 0.0377, "step": 48990 }, { "grad_norm": 0.7420658469200134, "learning_rate": 5.5775874252213247e-05, "loss": 0.0405, "step": 49000 }, { "grad_norm": 0.5399541854858398, "learning_rate": 5.575944993093189e-05, "loss": 0.0402, "step": 49010 }, { "grad_norm": 0.590316653251648, "learning_rate": 5.574302497980574e-05, "loss": 0.0317, "step": 49020 }, { "grad_norm": 0.6673380136489868, "learning_rate": 5.5726599400631005e-05, "loss": 0.0566, "step": 49030 }, { "grad_norm": 0.47115567326545715, "learning_rate": 5.571017319520399e-05, "loss": 0.028, "step": 49040 }, { "grad_norm": 0.5562276840209961, "learning_rate": 5.5693746365320984e-05, "loss": 0.0462, "step": 49050 }, { "grad_norm": 0.41832536458969116, "learning_rate": 5.567731891277848e-05, "loss": 0.0328, "step": 49060 }, { "grad_norm": 0.42936787009239197, "learning_rate": 5.566089083937288e-05, "loss": 0.037, "step": 49070 }, { "grad_norm": 0.5716803669929504, "learning_rate": 5.564446214690079e-05, "loss": 0.0345, "step": 49080 }, { "grad_norm": 0.821274995803833, "learning_rate": 5.56280328371588e-05, "loss": 0.0442, "step": 49090 }, { "grad_norm": 0.6923881769180298, "learning_rate": 5.5611602911943596e-05, "loss": 0.0326, "step": 49100 }, { "grad_norm": 0.47372645139694214, "learning_rate": 5.559517237305194e-05, "loss": 0.0313, "step": 49110 }, { "grad_norm": 0.6273805499076843, "learning_rate": 5.557874122228064e-05, "loss": 0.0299, "step": 49120 }, { "grad_norm": 0.4646921455860138, "learning_rate": 5.55623094614266e-05, "loss": 0.0277, "step": 49130 }, { "grad_norm": 0.46699222922325134, "learning_rate": 5.5545877092286743e-05, "loss": 0.0281, "step": 49140 }, { "grad_norm": 0.5067638754844666, "learning_rate": 5.552944411665811e-05, "loss": 0.0283, "step": 49150 }, { "grad_norm": 0.3698894679546356, "learning_rate": 5.5513010536337795e-05, "loss": 0.0391, "step": 49160 }, { "grad_norm": 0.4574033319950104, "learning_rate": 5.5496576353122934e-05, "loss": 0.0305, "step": 49170 }, { "grad_norm": 0.6415001749992371, "learning_rate": 5.548014156881074e-05, "loss": 0.029, "step": 49180 }, { "grad_norm": 0.46556130051612854, "learning_rate": 5.5463706185198494e-05, "loss": 0.0322, "step": 49190 }, { "grad_norm": 0.3084358870983124, "learning_rate": 5.544727020408358e-05, "loss": 0.0327, "step": 49200 }, { "grad_norm": 0.3073261082172394, "learning_rate": 5.5430833627263366e-05, "loss": 0.0454, "step": 49210 }, { "grad_norm": 0.4910079836845398, "learning_rate": 5.5414396456535354e-05, "loss": 0.0403, "step": 49220 }, { "grad_norm": 0.6554314494132996, "learning_rate": 5.539795869369709e-05, "loss": 0.0309, "step": 49230 }, { "grad_norm": 0.8189933896064758, "learning_rate": 5.538152034054618e-05, "loss": 0.0321, "step": 49240 }, { "grad_norm": 0.6348269581794739, "learning_rate": 5.536508139888028e-05, "loss": 0.0475, "step": 49250 }, { "grad_norm": 0.37957993149757385, "learning_rate": 5.534864187049716e-05, "loss": 0.0322, "step": 49260 }, { "grad_norm": 0.4670672118663788, "learning_rate": 5.5332201757194615e-05, "loss": 0.0438, "step": 49270 }, { "grad_norm": 0.5221475958824158, "learning_rate": 5.531576106077049e-05, "loss": 0.0439, "step": 49280 }, { "grad_norm": 1.2807247638702393, "learning_rate": 5.529931978302272e-05, "loss": 0.0434, "step": 49290 }, { "grad_norm": 0.4394684433937073, "learning_rate": 5.528287792574932e-05, "loss": 0.0362, "step": 49300 }, { "grad_norm": 0.5757120847702026, "learning_rate": 5.5266435490748324e-05, "loss": 0.037, "step": 49310 }, { "grad_norm": 0.543403685092926, "learning_rate": 5.524999247981787e-05, "loss": 0.0387, "step": 49320 }, { "grad_norm": 0.5090606808662415, "learning_rate": 5.523354889475613e-05, "loss": 0.0415, "step": 49330 }, { "grad_norm": 0.4354216754436493, "learning_rate": 5.521710473736134e-05, "loss": 0.0426, "step": 49340 }, { "grad_norm": 0.4370010793209076, "learning_rate": 5.5200660009431826e-05, "loss": 0.039, "step": 49350 }, { "grad_norm": 0.4438594877719879, "learning_rate": 5.518421471276596e-05, "loss": 0.0316, "step": 49360 }, { "grad_norm": 0.4692568778991699, "learning_rate": 5.5167768849162174e-05, "loss": 0.0352, "step": 49370 }, { "grad_norm": 0.5479319095611572, "learning_rate": 5.515132242041893e-05, "loss": 0.0275, "step": 49380 }, { "grad_norm": 1.1332085132598877, "learning_rate": 5.513487542833483e-05, "loss": 0.0399, "step": 49390 }, { "grad_norm": 0.6412266492843628, "learning_rate": 5.5118427874708466e-05, "loss": 0.0438, "step": 49400 }, { "grad_norm": 0.8235276937484741, "learning_rate": 5.510197976133853e-05, "loss": 0.0439, "step": 49410 }, { "grad_norm": 0.8317351341247559, "learning_rate": 5.508553109002376e-05, "loss": 0.0374, "step": 49420 }, { "grad_norm": 0.5678921341896057, "learning_rate": 5.5069081862562957e-05, "loss": 0.0338, "step": 49430 }, { "grad_norm": 0.52516770362854, "learning_rate": 5.5052632080754965e-05, "loss": 0.0448, "step": 49440 }, { "grad_norm": 0.604753851890564, "learning_rate": 5.503618174639874e-05, "loss": 0.0332, "step": 49450 }, { "grad_norm": 0.6531098484992981, "learning_rate": 5.501973086129325e-05, "loss": 0.0365, "step": 49460 }, { "grad_norm": 0.7864130139350891, "learning_rate": 5.500327942723753e-05, "loss": 0.0421, "step": 49470 }, { "grad_norm": 0.48945364356040955, "learning_rate": 5.498682744603071e-05, "loss": 0.0308, "step": 49480 }, { "grad_norm": 0.3098685145378113, "learning_rate": 5.497037491947194e-05, "loss": 0.04, "step": 49490 }, { "grad_norm": 0.4311908483505249, "learning_rate": 5.4953921849360424e-05, "loss": 0.0353, "step": 49500 }, { "grad_norm": 0.46951842308044434, "learning_rate": 5.493746823749547e-05, "loss": 0.0331, "step": 49510 }, { "grad_norm": 0.6004273295402527, "learning_rate": 5.49210140856764e-05, "loss": 0.0518, "step": 49520 }, { "grad_norm": 0.5140829086303711, "learning_rate": 5.4904559395702647e-05, "loss": 0.0442, "step": 49530 }, { "grad_norm": 1.3726977109909058, "learning_rate": 5.488810416937364e-05, "loss": 0.0475, "step": 49540 }, { "grad_norm": 0.5161929726600647, "learning_rate": 5.4871648408488926e-05, "loss": 0.0354, "step": 49550 }, { "grad_norm": 0.4320122301578522, "learning_rate": 5.485519211484807e-05, "loss": 0.0425, "step": 49560 }, { "grad_norm": 0.35320237278938293, "learning_rate": 5.4838735290250676e-05, "loss": 0.0289, "step": 49570 }, { "grad_norm": 0.7228052616119385, "learning_rate": 5.482227793649649e-05, "loss": 0.0396, "step": 49580 }, { "grad_norm": 0.44102418422698975, "learning_rate": 5.480582005538524e-05, "loss": 0.0214, "step": 49590 }, { "grad_norm": 0.4301331639289856, "learning_rate": 5.478936164871671e-05, "loss": 0.0333, "step": 49600 }, { "grad_norm": 0.3390357196331024, "learning_rate": 5.477290271829081e-05, "loss": 0.0389, "step": 49610 }, { "grad_norm": 0.3929409086704254, "learning_rate": 5.475644326590744e-05, "loss": 0.0353, "step": 49620 }, { "grad_norm": 0.6068605780601501, "learning_rate": 5.473998329336658e-05, "loss": 0.0372, "step": 49630 }, { "grad_norm": 0.5941483378410339, "learning_rate": 5.4723522802468286e-05, "loss": 0.0307, "step": 49640 }, { "grad_norm": 0.7112465500831604, "learning_rate": 5.470706179501264e-05, "loss": 0.0452, "step": 49650 }, { "grad_norm": 0.4028644859790802, "learning_rate": 5.4690600272799776e-05, "loss": 0.0289, "step": 49660 }, { "grad_norm": 0.42064112424850464, "learning_rate": 5.467413823762993e-05, "loss": 0.0332, "step": 49670 }, { "grad_norm": 0.5641224980354309, "learning_rate": 5.465767569130335e-05, "loss": 0.0331, "step": 49680 }, { "grad_norm": 0.604812502861023, "learning_rate": 5.464121263562036e-05, "loss": 0.0351, "step": 49690 }, { "grad_norm": 0.4191550016403198, "learning_rate": 5.4624749072381343e-05, "loss": 0.0324, "step": 49700 }, { "grad_norm": 0.8594871163368225, "learning_rate": 5.460828500338672e-05, "loss": 0.0393, "step": 49710 }, { "grad_norm": 0.7554093599319458, "learning_rate": 5.459182043043698e-05, "loss": 0.0253, "step": 49720 }, { "grad_norm": 0.5676077604293823, "learning_rate": 5.457535535533265e-05, "loss": 0.0352, "step": 49730 }, { "grad_norm": 0.4273280203342438, "learning_rate": 5.4558889779874334e-05, "loss": 0.0511, "step": 49740 }, { "grad_norm": 0.5625126361846924, "learning_rate": 5.454242370586269e-05, "loss": 0.0344, "step": 49750 }, { "grad_norm": 0.7117862701416016, "learning_rate": 5.452595713509843e-05, "loss": 0.0334, "step": 49760 }, { "grad_norm": 0.3893553614616394, "learning_rate": 5.4509490069382274e-05, "loss": 0.0458, "step": 49770 }, { "grad_norm": 0.9143555760383606, "learning_rate": 5.449302251051509e-05, "loss": 0.0433, "step": 49780 }, { "grad_norm": 0.47438666224479675, "learning_rate": 5.44765544602977e-05, "loss": 0.0394, "step": 49790 }, { "grad_norm": 0.6840224862098694, "learning_rate": 5.446008592053107e-05, "loss": 0.0504, "step": 49800 }, { "grad_norm": 0.5380843877792358, "learning_rate": 5.4443616893016135e-05, "loss": 0.0376, "step": 49810 }, { "grad_norm": 0.5044940710067749, "learning_rate": 5.4427147379553934e-05, "loss": 0.0347, "step": 49820 }, { "grad_norm": 0.5206811428070068, "learning_rate": 5.441067738194556e-05, "loss": 0.0359, "step": 49830 }, { "grad_norm": 0.2554585635662079, "learning_rate": 5.439420690199214e-05, "loss": 0.0329, "step": 49840 }, { "grad_norm": 0.6646296381950378, "learning_rate": 5.437773594149484e-05, "loss": 0.0331, "step": 49850 }, { "grad_norm": 0.5625228881835938, "learning_rate": 5.436126450225495e-05, "loss": 0.0349, "step": 49860 }, { "grad_norm": 0.6113070845603943, "learning_rate": 5.434479258607371e-05, "loss": 0.0369, "step": 49870 }, { "grad_norm": 0.47774481773376465, "learning_rate": 5.43283201947525e-05, "loss": 0.0479, "step": 49880 }, { "grad_norm": 0.9460112452507019, "learning_rate": 5.431184733009268e-05, "loss": 0.0351, "step": 49890 }, { "grad_norm": 1.1644179821014404, "learning_rate": 5.4295373993895736e-05, "loss": 0.0488, "step": 49900 }, { "grad_norm": 1.0999773740768433, "learning_rate": 5.4278900187963157e-05, "loss": 0.0316, "step": 49910 }, { "grad_norm": 0.9628614783287048, "learning_rate": 5.426242591409646e-05, "loss": 0.0465, "step": 49920 }, { "grad_norm": 0.3426874577999115, "learning_rate": 5.42459511740973e-05, "loss": 0.0343, "step": 49930 }, { "grad_norm": 0.4273504614830017, "learning_rate": 5.422947596976729e-05, "loss": 0.0532, "step": 49940 }, { "grad_norm": 0.5397392511367798, "learning_rate": 5.4213000302908134e-05, "loss": 0.0356, "step": 49950 }, { "grad_norm": 0.5177704095840454, "learning_rate": 5.419652417532162e-05, "loss": 0.0358, "step": 49960 }, { "grad_norm": 0.3996940851211548, "learning_rate": 5.4180047588809534e-05, "loss": 0.0376, "step": 49970 }, { "grad_norm": 0.4403967559337616, "learning_rate": 5.4163570545173704e-05, "loss": 0.0649, "step": 49980 }, { "grad_norm": 0.6126241087913513, "learning_rate": 5.414709304621608e-05, "loss": 0.0349, "step": 49990 }, { "grad_norm": 0.8049873113632202, "learning_rate": 5.413061509373858e-05, "loss": 0.0349, "step": 50000 }, { "grad_norm": 0.6413583755493164, "learning_rate": 5.411413668954324e-05, "loss": 0.0457, "step": 50010 }, { "grad_norm": 0.3690757751464844, "learning_rate": 5.4097657835432083e-05, "loss": 0.0368, "step": 50020 }, { "grad_norm": 0.6286015510559082, "learning_rate": 5.408117853320723e-05, "loss": 0.0315, "step": 50030 }, { "grad_norm": 0.6294805407524109, "learning_rate": 5.406469878467084e-05, "loss": 0.0437, "step": 50040 }, { "grad_norm": 0.4422198235988617, "learning_rate": 5.404821859162509e-05, "loss": 0.0294, "step": 50050 }, { "grad_norm": 0.7431378960609436, "learning_rate": 5.403173795587225e-05, "loss": 0.0516, "step": 50060 }, { "grad_norm": 0.5036113858222961, "learning_rate": 5.4015256879214606e-05, "loss": 0.0318, "step": 50070 }, { "grad_norm": 0.8181453347206116, "learning_rate": 5.39987753634545e-05, "loss": 0.0402, "step": 50080 }, { "grad_norm": 0.5935421586036682, "learning_rate": 5.3982293410394335e-05, "loss": 0.044, "step": 50090 }, { "grad_norm": 0.389945387840271, "learning_rate": 5.396581102183655e-05, "loss": 0.0312, "step": 50100 }, { "grad_norm": 0.5626307129859924, "learning_rate": 5.3949328199583634e-05, "loss": 0.0452, "step": 50110 }, { "grad_norm": 0.4577651917934418, "learning_rate": 5.393284494543813e-05, "loss": 0.0324, "step": 50120 }, { "grad_norm": 0.6305383443832397, "learning_rate": 5.391636126120262e-05, "loss": 0.0364, "step": 50130 }, { "grad_norm": 0.6012268662452698, "learning_rate": 5.389987714867971e-05, "loss": 0.0328, "step": 50140 }, { "grad_norm": 0.39246538281440735, "learning_rate": 5.3883392609672114e-05, "loss": 0.0293, "step": 50150 }, { "grad_norm": 0.2774273455142975, "learning_rate": 5.3866907645982545e-05, "loss": 0.0437, "step": 50160 }, { "grad_norm": 0.4979347586631775, "learning_rate": 5.385042225941375e-05, "loss": 0.0397, "step": 50170 }, { "grad_norm": 0.8055859208106995, "learning_rate": 5.3833936451768585e-05, "loss": 0.0423, "step": 50180 }, { "grad_norm": 0.5104636549949646, "learning_rate": 5.3817450224849884e-05, "loss": 0.0331, "step": 50190 }, { "grad_norm": 1.1469179391860962, "learning_rate": 5.380096358046056e-05, "loss": 0.0386, "step": 50200 }, { "grad_norm": 0.5093704462051392, "learning_rate": 5.378447652040359e-05, "loss": 0.0328, "step": 50210 }, { "grad_norm": 0.48083752393722534, "learning_rate": 5.376798904648195e-05, "loss": 0.0329, "step": 50220 }, { "grad_norm": 0.6093577146530151, "learning_rate": 5.375150116049869e-05, "loss": 0.0293, "step": 50230 }, { "grad_norm": 0.5481991767883301, "learning_rate": 5.373501286425691e-05, "loss": 0.0346, "step": 50240 }, { "grad_norm": 0.5178562998771667, "learning_rate": 5.3718524159559726e-05, "loss": 0.0503, "step": 50250 }, { "grad_norm": 0.38251379132270813, "learning_rate": 5.370203504821034e-05, "loss": 0.0381, "step": 50260 }, { "grad_norm": 0.533612072467804, "learning_rate": 5.368554553201196e-05, "loss": 0.0449, "step": 50270 }, { "grad_norm": 0.5346506237983704, "learning_rate": 5.3669055612767874e-05, "loss": 0.0373, "step": 50280 }, { "grad_norm": 0.40015512704849243, "learning_rate": 5.3652565292281376e-05, "loss": 0.0374, "step": 50290 }, { "grad_norm": 1.003159523010254, "learning_rate": 5.3636074572355824e-05, "loss": 0.033, "step": 50300 }, { "grad_norm": 0.5003423094749451, "learning_rate": 5.361958345479463e-05, "loss": 0.0507, "step": 50310 }, { "grad_norm": 0.481934130191803, "learning_rate": 5.360309194140123e-05, "loss": 0.0369, "step": 50320 }, { "grad_norm": 0.4630373418331146, "learning_rate": 5.3586600033979106e-05, "loss": 0.0338, "step": 50330 }, { "grad_norm": 0.4770975708961487, "learning_rate": 5.357010773433181e-05, "loss": 0.0273, "step": 50340 }, { "grad_norm": 0.6067084670066833, "learning_rate": 5.3553615044262895e-05, "loss": 0.0333, "step": 50350 }, { "grad_norm": 0.5349823236465454, "learning_rate": 5.3537121965575964e-05, "loss": 0.0335, "step": 50360 }, { "grad_norm": 0.6858391761779785, "learning_rate": 5.352062850007471e-05, "loss": 0.059, "step": 50370 }, { "grad_norm": 0.6872913837432861, "learning_rate": 5.350413464956282e-05, "loss": 0.0362, "step": 50380 }, { "grad_norm": 0.36305204033851624, "learning_rate": 5.348764041584403e-05, "loss": 0.0418, "step": 50390 }, { "grad_norm": 0.684984564781189, "learning_rate": 5.347114580072212e-05, "loss": 0.0335, "step": 50400 }, { "grad_norm": 0.48757851123809814, "learning_rate": 5.345465080600093e-05, "loss": 0.0351, "step": 50410 }, { "grad_norm": 0.3727697730064392, "learning_rate": 5.3438155433484326e-05, "loss": 0.039, "step": 50420 }, { "grad_norm": 0.9415730237960815, "learning_rate": 5.3421659684976197e-05, "loss": 0.041, "step": 50430 }, { "grad_norm": 0.40164390206336975, "learning_rate": 5.340516356228052e-05, "loss": 0.0327, "step": 50440 }, { "grad_norm": 0.6830030083656311, "learning_rate": 5.338866706720128e-05, "loss": 0.0445, "step": 50450 }, { "grad_norm": 0.6323651671409607, "learning_rate": 5.337217020154249e-05, "loss": 0.0434, "step": 50460 }, { "grad_norm": 0.4911954998970032, "learning_rate": 5.335567296710825e-05, "loss": 0.0315, "step": 50470 }, { "grad_norm": 0.5641705989837646, "learning_rate": 5.333917536570265e-05, "loss": 0.0369, "step": 50480 }, { "grad_norm": 0.6028375625610352, "learning_rate": 5.332267739912986e-05, "loss": 0.0382, "step": 50490 }, { "grad_norm": 1.5713081359863281, "learning_rate": 5.330617906919405e-05, "loss": 0.0339, "step": 50500 }, { "grad_norm": 0.3356788754463196, "learning_rate": 5.328968037769949e-05, "loss": 0.0354, "step": 50510 }, { "grad_norm": 0.3442026972770691, "learning_rate": 5.327318132645042e-05, "loss": 0.0258, "step": 50520 }, { "grad_norm": 0.5568965673446655, "learning_rate": 5.325668191725116e-05, "loss": 0.0345, "step": 50530 }, { "grad_norm": 0.5943773984909058, "learning_rate": 5.324018215190606e-05, "loss": 0.0399, "step": 50540 }, { "grad_norm": 0.3254600167274475, "learning_rate": 5.3223682032219515e-05, "loss": 0.0343, "step": 50550 }, { "grad_norm": 0.4056698679924011, "learning_rate": 5.320718155999595e-05, "loss": 0.0339, "step": 50560 }, { "grad_norm": 0.3804103136062622, "learning_rate": 5.3190680737039835e-05, "loss": 0.0301, "step": 50570 }, { "grad_norm": 0.6275412440299988, "learning_rate": 5.317417956515567e-05, "loss": 0.0427, "step": 50580 }, { "grad_norm": 0.3027311861515045, "learning_rate": 5.315767804614798e-05, "loss": 0.0324, "step": 50590 }, { "grad_norm": 0.446185439825058, "learning_rate": 5.3141176181821395e-05, "loss": 0.0362, "step": 50600 }, { "grad_norm": 0.41501474380493164, "learning_rate": 5.31246739739805e-05, "loss": 0.0329, "step": 50610 }, { "grad_norm": 0.6002718806266785, "learning_rate": 5.310817142442994e-05, "loss": 0.0323, "step": 50620 }, { "grad_norm": 0.7564369440078735, "learning_rate": 5.309166853497445e-05, "loss": 0.0303, "step": 50630 }, { "grad_norm": 0.8587841987609863, "learning_rate": 5.307516530741873e-05, "loss": 0.0452, "step": 50640 }, { "grad_norm": 0.6065633893013, "learning_rate": 5.305866174356754e-05, "loss": 0.057, "step": 50650 }, { "grad_norm": 0.6465681195259094, "learning_rate": 5.304215784522571e-05, "loss": 0.0344, "step": 50660 }, { "grad_norm": 0.42464080452919006, "learning_rate": 5.302565361419808e-05, "loss": 0.0324, "step": 50670 }, { "grad_norm": 0.54108065366745, "learning_rate": 5.3009149052289507e-05, "loss": 0.0304, "step": 50680 }, { "grad_norm": 0.2620874047279358, "learning_rate": 5.299264416130493e-05, "loss": 0.0315, "step": 50690 }, { "grad_norm": 0.3445957601070404, "learning_rate": 5.297613894304928e-05, "loss": 0.0481, "step": 50700 }, { "grad_norm": 0.45900607109069824, "learning_rate": 5.2959633399327534e-05, "loss": 0.0375, "step": 50710 }, { "grad_norm": 0.5127825736999512, "learning_rate": 5.294312753194476e-05, "loss": 0.0298, "step": 50720 }, { "grad_norm": 0.6048670411109924, "learning_rate": 5.292662134270596e-05, "loss": 0.0416, "step": 50730 }, { "grad_norm": 0.7722629308700562, "learning_rate": 5.291011483341626e-05, "loss": 0.0387, "step": 50740 }, { "grad_norm": 0.4580465853214264, "learning_rate": 5.2893608005880767e-05, "loss": 0.0342, "step": 50750 }, { "grad_norm": 0.515717625617981, "learning_rate": 5.287710086190467e-05, "loss": 0.0351, "step": 50760 }, { "grad_norm": 0.430679053068161, "learning_rate": 5.286059340329314e-05, "loss": 0.0365, "step": 50770 }, { "grad_norm": 0.28365641832351685, "learning_rate": 5.284408563185141e-05, "loss": 0.0313, "step": 50780 }, { "grad_norm": 0.45067688822746277, "learning_rate": 5.282757754938476e-05, "loss": 0.0268, "step": 50790 }, { "grad_norm": 0.31454023718833923, "learning_rate": 5.281106915769849e-05, "loss": 0.0287, "step": 50800 }, { "grad_norm": 0.5366957187652588, "learning_rate": 5.2794560458597897e-05, "loss": 0.0342, "step": 50810 }, { "grad_norm": 0.5985012054443359, "learning_rate": 5.27780514538884e-05, "loss": 0.0323, "step": 50820 }, { "grad_norm": 0.3654036521911621, "learning_rate": 5.2761542145375365e-05, "loss": 0.0198, "step": 50830 }, { "grad_norm": 0.6232225298881531, "learning_rate": 5.274503253486421e-05, "loss": 0.0472, "step": 50840 }, { "grad_norm": 0.5917954444885254, "learning_rate": 5.272852262416046e-05, "loss": 0.035, "step": 50850 }, { "grad_norm": 0.7303821444511414, "learning_rate": 5.2712012415069555e-05, "loss": 0.0583, "step": 50860 }, { "grad_norm": 0.6187044382095337, "learning_rate": 5.269550190939705e-05, "loss": 0.0412, "step": 50870 }, { "grad_norm": 0.6679890155792236, "learning_rate": 5.267899110894852e-05, "loss": 0.0311, "step": 50880 }, { "grad_norm": 0.42154204845428467, "learning_rate": 5.266248001552955e-05, "loss": 0.0398, "step": 50890 }, { "grad_norm": 0.5885964632034302, "learning_rate": 5.264596863094575e-05, "loss": 0.0392, "step": 50900 }, { "grad_norm": 0.472382515668869, "learning_rate": 5.262945695700282e-05, "loss": 0.0407, "step": 50910 }, { "grad_norm": 0.47242090106010437, "learning_rate": 5.261294499550643e-05, "loss": 0.032, "step": 50920 }, { "grad_norm": 0.7386075258255005, "learning_rate": 5.25964327482623e-05, "loss": 0.046, "step": 50930 }, { "grad_norm": 0.39989879727363586, "learning_rate": 5.257992021707617e-05, "loss": 0.0317, "step": 50940 }, { "grad_norm": 0.2656809985637665, "learning_rate": 5.256340740375387e-05, "loss": 0.0351, "step": 50950 }, { "grad_norm": 0.32636335492134094, "learning_rate": 5.254689431010117e-05, "loss": 0.036, "step": 50960 }, { "grad_norm": 0.2722826600074768, "learning_rate": 5.253038093792395e-05, "loss": 0.0309, "step": 50970 }, { "grad_norm": 0.568524181842804, "learning_rate": 5.251386728902806e-05, "loss": 0.0362, "step": 50980 }, { "grad_norm": 0.610887348651886, "learning_rate": 5.2497353365219446e-05, "loss": 0.0317, "step": 50990 }, { "grad_norm": 0.6313692927360535, "learning_rate": 5.2480839168304e-05, "loss": 0.0561, "step": 51000 }, { "grad_norm": 0.5306339859962463, "learning_rate": 5.246432470008773e-05, "loss": 0.0324, "step": 51010 }, { "grad_norm": 0.7675139307975769, "learning_rate": 5.2447809962376616e-05, "loss": 0.0317, "step": 51020 }, { "grad_norm": 0.39529046416282654, "learning_rate": 5.243129495697666e-05, "loss": 0.0341, "step": 51030 }, { "grad_norm": 0.5533754229545593, "learning_rate": 5.241477968569397e-05, "loss": 0.0373, "step": 51040 }, { "grad_norm": 0.6223902702331543, "learning_rate": 5.23982641503346e-05, "loss": 0.0371, "step": 51050 }, { "grad_norm": 0.44710952043533325, "learning_rate": 5.238174835270465e-05, "loss": 0.0437, "step": 51060 }, { "grad_norm": 0.3616122901439667, "learning_rate": 5.23652322946103e-05, "loss": 0.0305, "step": 51070 }, { "grad_norm": 0.6686208844184875, "learning_rate": 5.234871597785771e-05, "loss": 0.0321, "step": 51080 }, { "grad_norm": 0.351514607667923, "learning_rate": 5.233219940425306e-05, "loss": 0.0285, "step": 51090 }, { "grad_norm": 0.5439487099647522, "learning_rate": 5.2315682575602576e-05, "loss": 0.0318, "step": 51100 }, { "grad_norm": 0.9054826498031616, "learning_rate": 5.229916549371254e-05, "loss": 0.0267, "step": 51110 }, { "grad_norm": 0.3737834393978119, "learning_rate": 5.228264816038924e-05, "loss": 0.0361, "step": 51120 }, { "grad_norm": 0.4694058299064636, "learning_rate": 5.2266130577438935e-05, "loss": 0.0405, "step": 51130 }, { "grad_norm": 0.41614773869514465, "learning_rate": 5.224961274666802e-05, "loss": 0.0394, "step": 51140 }, { "grad_norm": 0.4041460454463959, "learning_rate": 5.223309466988283e-05, "loss": 0.035, "step": 51150 }, { "grad_norm": 0.3634231388568878, "learning_rate": 5.221657634888976e-05, "loss": 0.0321, "step": 51160 }, { "grad_norm": 0.5810183882713318, "learning_rate": 5.2200057785495235e-05, "loss": 0.0287, "step": 51170 }, { "grad_norm": 0.452157199382782, "learning_rate": 5.2183538981505706e-05, "loss": 0.033, "step": 51180 }, { "grad_norm": 0.46379151940345764, "learning_rate": 5.216701993872762e-05, "loss": 0.033, "step": 51190 }, { "grad_norm": 2.107563018798828, "learning_rate": 5.215050065896752e-05, "loss": 0.0466, "step": 51200 }, { "grad_norm": 0.5696273446083069, "learning_rate": 5.2133981144031875e-05, "loss": 0.0491, "step": 51210 }, { "grad_norm": 0.29391759634017944, "learning_rate": 5.211746139572727e-05, "loss": 0.0289, "step": 51220 }, { "grad_norm": 0.9068784713745117, "learning_rate": 5.210094141586026e-05, "loss": 0.0443, "step": 51230 }, { "grad_norm": 0.435458779335022, "learning_rate": 5.2084421206237476e-05, "loss": 0.0421, "step": 51240 }, { "grad_norm": 0.43437322974205017, "learning_rate": 5.20679007686655e-05, "loss": 0.0349, "step": 51250 }, { "grad_norm": 0.6195080280303955, "learning_rate": 5.205138010495102e-05, "loss": 0.0424, "step": 51260 }, { "grad_norm": 0.440188467502594, "learning_rate": 5.20348592169007e-05, "loss": 0.0349, "step": 51270 }, { "grad_norm": 0.3354952931404114, "learning_rate": 5.201833810632123e-05, "loss": 0.0277, "step": 51280 }, { "grad_norm": 0.5333068370819092, "learning_rate": 5.200181677501933e-05, "loss": 0.0325, "step": 51290 }, { "grad_norm": 0.7042534351348877, "learning_rate": 5.1985295224801777e-05, "loss": 0.0498, "step": 51300 }, { "grad_norm": 0.7252212166786194, "learning_rate": 5.196877345747531e-05, "loss": 0.0329, "step": 51310 }, { "grad_norm": 1.4311764240264893, "learning_rate": 5.195225147484674e-05, "loss": 0.0356, "step": 51320 }, { "grad_norm": 0.4808245301246643, "learning_rate": 5.193572927872288e-05, "loss": 0.0455, "step": 51330 }, { "grad_norm": 0.44582170248031616, "learning_rate": 5.191920687091059e-05, "loss": 0.0471, "step": 51340 }, { "grad_norm": 0.41684114933013916, "learning_rate": 5.1902684253216695e-05, "loss": 0.0322, "step": 51350 }, { "grad_norm": 0.6920727491378784, "learning_rate": 5.1886161427448134e-05, "loss": 0.0236, "step": 51360 }, { "grad_norm": 0.27586477994918823, "learning_rate": 5.186963839541179e-05, "loss": 0.0291, "step": 51370 }, { "grad_norm": 0.5589306354522705, "learning_rate": 5.185311515891459e-05, "loss": 0.027, "step": 51380 }, { "grad_norm": 0.5036526322364807, "learning_rate": 5.1836591719763516e-05, "loss": 0.0369, "step": 51390 }, { "grad_norm": 0.5310055017471313, "learning_rate": 5.182006807976554e-05, "loss": 0.031, "step": 51400 }, { "grad_norm": 0.27660828828811646, "learning_rate": 5.180354424072763e-05, "loss": 0.0334, "step": 51410 }, { "grad_norm": 0.4512002170085907, "learning_rate": 5.178702020445686e-05, "loss": 0.0405, "step": 51420 }, { "grad_norm": 0.33840200304985046, "learning_rate": 5.177049597276024e-05, "loss": 0.0303, "step": 51430 }, { "grad_norm": 0.42594826221466064, "learning_rate": 5.175397154744482e-05, "loss": 0.0372, "step": 51440 }, { "grad_norm": 0.6073659658432007, "learning_rate": 5.1737446930317736e-05, "loss": 0.0448, "step": 51450 }, { "grad_norm": 0.3077083230018616, "learning_rate": 5.172092212318604e-05, "loss": 0.037, "step": 51460 }, { "grad_norm": 0.5075892806053162, "learning_rate": 5.170439712785692e-05, "loss": 0.0386, "step": 51470 }, { "grad_norm": 0.4655303359031677, "learning_rate": 5.1687871946137467e-05, "loss": 0.0258, "step": 51480 }, { "grad_norm": 0.4304044544696808, "learning_rate": 5.16713465798349e-05, "loss": 0.0272, "step": 51490 }, { "grad_norm": 0.4939371943473816, "learning_rate": 5.165482103075637e-05, "loss": 0.0345, "step": 51500 }, { "grad_norm": 0.45185914635658264, "learning_rate": 5.163829530070909e-05, "loss": 0.0316, "step": 51510 }, { "grad_norm": 0.7706164717674255, "learning_rate": 5.162176939150032e-05, "loss": 0.0316, "step": 51520 }, { "grad_norm": 0.3914806544780731, "learning_rate": 5.160524330493728e-05, "loss": 0.032, "step": 51530 }, { "grad_norm": 0.4799436628818512, "learning_rate": 5.158871704282724e-05, "loss": 0.0277, "step": 51540 }, { "grad_norm": 0.4329206943511963, "learning_rate": 5.15721906069775e-05, "loss": 0.0273, "step": 51550 }, { "grad_norm": 0.41145867109298706, "learning_rate": 5.155566399919536e-05, "loss": 0.0318, "step": 51560 }, { "grad_norm": 0.4313386380672455, "learning_rate": 5.153913722128813e-05, "loss": 0.0508, "step": 51570 }, { "grad_norm": 0.7304286956787109, "learning_rate": 5.1522610275063176e-05, "loss": 0.0296, "step": 51580 }, { "grad_norm": 0.5431072115898132, "learning_rate": 5.150608316232786e-05, "loss": 0.0357, "step": 51590 }, { "grad_norm": 0.6868614554405212, "learning_rate": 5.148955588488955e-05, "loss": 0.0367, "step": 51600 }, { "grad_norm": 0.43191447854042053, "learning_rate": 5.147302844455564e-05, "loss": 0.0311, "step": 51610 }, { "grad_norm": 0.531554102897644, "learning_rate": 5.145650084313356e-05, "loss": 0.0367, "step": 51620 }, { "grad_norm": 2.859622001647949, "learning_rate": 5.1439973082430734e-05, "loss": 0.0365, "step": 51630 }, { "grad_norm": 1.8431148529052734, "learning_rate": 5.142344516425461e-05, "loss": 0.0249, "step": 51640 }, { "grad_norm": 0.6356233954429626, "learning_rate": 5.1406917090412673e-05, "loss": 0.0345, "step": 51650 }, { "grad_norm": 0.37837910652160645, "learning_rate": 5.1390388862712404e-05, "loss": 0.0249, "step": 51660 }, { "grad_norm": 0.4770631194114685, "learning_rate": 5.137386048296129e-05, "loss": 0.0384, "step": 51670 }, { "grad_norm": 0.32959508895874023, "learning_rate": 5.135733195296686e-05, "loss": 0.0349, "step": 51680 }, { "grad_norm": 0.5247185826301575, "learning_rate": 5.134080327453664e-05, "loss": 0.0364, "step": 51690 }, { "grad_norm": 0.36251288652420044, "learning_rate": 5.1324274449478204e-05, "loss": 0.0323, "step": 51700 }, { "grad_norm": 0.4681988060474396, "learning_rate": 5.130774547959909e-05, "loss": 0.0316, "step": 51710 }, { "grad_norm": 0.38344717025756836, "learning_rate": 5.129121636670692e-05, "loss": 0.0252, "step": 51720 }, { "grad_norm": 0.5805899500846863, "learning_rate": 5.1274687112609246e-05, "loss": 0.0329, "step": 51730 }, { "grad_norm": 0.4115343689918518, "learning_rate": 5.125815771911372e-05, "loss": 0.0331, "step": 51740 }, { "grad_norm": 0.5407361388206482, "learning_rate": 5.1241628188027956e-05, "loss": 0.0582, "step": 51750 }, { "grad_norm": 0.6200650334358215, "learning_rate": 5.1225098521159585e-05, "loss": 0.0352, "step": 51760 }, { "grad_norm": 0.47344112396240234, "learning_rate": 5.12085687203163e-05, "loss": 0.037, "step": 51770 }, { "grad_norm": 0.32590770721435547, "learning_rate": 5.119203878730575e-05, "loss": 0.0248, "step": 51780 }, { "grad_norm": 0.48674923181533813, "learning_rate": 5.117550872393564e-05, "loss": 0.0253, "step": 51790 }, { "grad_norm": 0.4404297173023224, "learning_rate": 5.1158978532013645e-05, "loss": 0.0318, "step": 51800 }, { "grad_norm": 0.5702709555625916, "learning_rate": 5.114244821334751e-05, "loss": 0.0365, "step": 51810 }, { "grad_norm": 0.3664301633834839, "learning_rate": 5.112591776974497e-05, "loss": 0.0424, "step": 51820 }, { "grad_norm": 0.3556475043296814, "learning_rate": 5.110938720301374e-05, "loss": 0.0311, "step": 51830 }, { "grad_norm": 0.47599703073501587, "learning_rate": 5.1092856514961595e-05, "loss": 0.0311, "step": 51840 }, { "grad_norm": 0.3738378882408142, "learning_rate": 5.107632570739631e-05, "loss": 0.0238, "step": 51850 }, { "grad_norm": 0.5883981585502625, "learning_rate": 5.1059794782125645e-05, "loss": 0.0422, "step": 51860 }, { "grad_norm": 0.748015284538269, "learning_rate": 5.104326374095744e-05, "loss": 0.0297, "step": 51870 }, { "grad_norm": 0.6318260431289673, "learning_rate": 5.102673258569947e-05, "loss": 0.0236, "step": 51880 }, { "grad_norm": 0.7138366103172302, "learning_rate": 5.101020131815956e-05, "loss": 0.043, "step": 51890 }, { "grad_norm": 0.5087956190109253, "learning_rate": 5.099366994014556e-05, "loss": 0.0321, "step": 51900 }, { "grad_norm": 0.34528297185897827, "learning_rate": 5.097713845346532e-05, "loss": 0.0431, "step": 51910 }, { "grad_norm": 0.3413616120815277, "learning_rate": 5.096060685992668e-05, "loss": 0.0268, "step": 51920 }, { "grad_norm": 0.4499937891960144, "learning_rate": 5.0944075161337514e-05, "loss": 0.0289, "step": 51930 }, { "grad_norm": 0.3012661039829254, "learning_rate": 5.09275433595057e-05, "loss": 0.0361, "step": 51940 }, { "grad_norm": 0.4573654234409332, "learning_rate": 5.0911011456239157e-05, "loss": 0.039, "step": 51950 }, { "grad_norm": 0.6312633156776428, "learning_rate": 5.089447945334574e-05, "loss": 0.0351, "step": 51960 }, { "grad_norm": 0.44966304302215576, "learning_rate": 5.087794735263343e-05, "loss": 0.0317, "step": 51970 }, { "grad_norm": 0.2828112542629242, "learning_rate": 5.086141515591011e-05, "loss": 0.027, "step": 51980 }, { "grad_norm": 0.5037036538124084, "learning_rate": 5.084488286498371e-05, "loss": 0.0306, "step": 51990 }, { "grad_norm": 0.3367908298969269, "learning_rate": 5.08283504816622e-05, "loss": 0.0341, "step": 52000 }, { "grad_norm": 0.4509970247745514, "learning_rate": 5.081181800775353e-05, "loss": 0.0321, "step": 52010 }, { "grad_norm": 0.41579946875572205, "learning_rate": 5.079528544506565e-05, "loss": 0.0518, "step": 52020 }, { "grad_norm": 0.7197730541229248, "learning_rate": 5.077875279540657e-05, "loss": 0.0295, "step": 52030 }, { "grad_norm": 0.4896433651447296, "learning_rate": 5.076222006058424e-05, "loss": 0.039, "step": 52040 }, { "grad_norm": 0.3988996148109436, "learning_rate": 5.074568724240667e-05, "loss": 0.0282, "step": 52050 }, { "grad_norm": 0.43555065989494324, "learning_rate": 5.072915434268187e-05, "loss": 0.0372, "step": 52060 }, { "grad_norm": 0.2919551134109497, "learning_rate": 5.071262136321785e-05, "loss": 0.0471, "step": 52070 }, { "grad_norm": 0.6667822599411011, "learning_rate": 5.069608830582262e-05, "loss": 0.0364, "step": 52080 }, { "grad_norm": 0.3267575204372406, "learning_rate": 5.067955517230423e-05, "loss": 0.0336, "step": 52090 }, { "grad_norm": 0.3654749095439911, "learning_rate": 5.06630219644707e-05, "loss": 0.0395, "step": 52100 }, { "grad_norm": 0.5883592367172241, "learning_rate": 5.064648868413008e-05, "loss": 0.0232, "step": 52110 }, { "grad_norm": 0.3170010447502136, "learning_rate": 5.062995533309044e-05, "loss": 0.0275, "step": 52120 }, { "grad_norm": 0.7517979741096497, "learning_rate": 5.061342191315983e-05, "loss": 0.0482, "step": 52130 }, { "grad_norm": 0.4089558720588684, "learning_rate": 5.0596888426146325e-05, "loss": 0.0299, "step": 52140 }, { "grad_norm": 0.41114726662635803, "learning_rate": 5.058035487385798e-05, "loss": 0.0297, "step": 52150 }, { "grad_norm": 0.4637659788131714, "learning_rate": 5.0563821258102915e-05, "loss": 0.0648, "step": 52160 }, { "grad_norm": 0.7785091996192932, "learning_rate": 5.05472875806892e-05, "loss": 0.0372, "step": 52170 }, { "grad_norm": 0.5343731641769409, "learning_rate": 5.053075384342494e-05, "loss": 0.0432, "step": 52180 }, { "grad_norm": 0.5768067240715027, "learning_rate": 5.0514220048118234e-05, "loss": 0.0326, "step": 52190 }, { "grad_norm": 0.43496131896972656, "learning_rate": 5.049768619657721e-05, "loss": 0.047, "step": 52200 }, { "grad_norm": 0.3306116759777069, "learning_rate": 5.048115229060996e-05, "loss": 0.0261, "step": 52210 }, { "grad_norm": 0.43891361355781555, "learning_rate": 5.046461833202462e-05, "loss": 0.0313, "step": 52220 }, { "grad_norm": 0.3918397128582001, "learning_rate": 5.044808432262932e-05, "loss": 0.0298, "step": 52230 }, { "grad_norm": 0.5291480422019958, "learning_rate": 5.043155026423219e-05, "loss": 0.031, "step": 52240 }, { "grad_norm": 0.39845603704452515, "learning_rate": 5.041501615864137e-05, "loss": 0.0406, "step": 52250 }, { "grad_norm": 0.7438106536865234, "learning_rate": 5.039848200766502e-05, "loss": 0.0306, "step": 52260 }, { "grad_norm": 0.7452266216278076, "learning_rate": 5.038194781311125e-05, "loss": 0.0413, "step": 52270 }, { "grad_norm": 0.29586824774742126, "learning_rate": 5.0365413576788264e-05, "loss": 0.0398, "step": 52280 }, { "grad_norm": 0.3089830279350281, "learning_rate": 5.0348879300504194e-05, "loss": 0.0521, "step": 52290 }, { "grad_norm": 0.5669301152229309, "learning_rate": 5.033234498606721e-05, "loss": 0.0338, "step": 52300 }, { "grad_norm": 0.4200313687324524, "learning_rate": 5.0315810635285476e-05, "loss": 0.0355, "step": 52310 }, { "grad_norm": 0.534332275390625, "learning_rate": 5.0299276249967164e-05, "loss": 0.0223, "step": 52320 }, { "grad_norm": 0.44266805052757263, "learning_rate": 5.0282741831920454e-05, "loss": 0.0311, "step": 52330 }, { "grad_norm": 0.5614042282104492, "learning_rate": 5.026620738295351e-05, "loss": 0.033, "step": 52340 }, { "grad_norm": 0.5927354097366333, "learning_rate": 5.024967290487455e-05, "loss": 0.0323, "step": 52350 }, { "grad_norm": 0.3066800832748413, "learning_rate": 5.023313839949172e-05, "loss": 0.0443, "step": 52360 }, { "grad_norm": 0.614177942276001, "learning_rate": 5.0216603868613224e-05, "loss": 0.0332, "step": 52370 }, { "grad_norm": 0.3981625735759735, "learning_rate": 5.020006931404726e-05, "loss": 0.0306, "step": 52380 }, { "grad_norm": 0.8123356699943542, "learning_rate": 5.0183534737602014e-05, "loss": 0.0438, "step": 52390 }, { "grad_norm": 0.3507777452468872, "learning_rate": 5.0167000141085686e-05, "loss": 0.0373, "step": 52400 }, { "grad_norm": 0.7227878570556641, "learning_rate": 5.015046552630647e-05, "loss": 0.0354, "step": 52410 }, { "grad_norm": 0.5005607008934021, "learning_rate": 5.013393089507258e-05, "loss": 0.0504, "step": 52420 }, { "grad_norm": 0.8666332960128784, "learning_rate": 5.011739624919219e-05, "loss": 0.0299, "step": 52430 }, { "grad_norm": 0.45434218645095825, "learning_rate": 5.0100861590473545e-05, "loss": 0.0395, "step": 52440 }, { "grad_norm": 0.7667002081871033, "learning_rate": 5.008432692072482e-05, "loss": 0.0411, "step": 52450 }, { "grad_norm": 0.5011823177337646, "learning_rate": 5.006779224175423e-05, "loss": 0.0324, "step": 52460 }, { "grad_norm": 0.3952452838420868, "learning_rate": 5.005125755536998e-05, "loss": 0.0427, "step": 52470 }, { "grad_norm": 0.4458942413330078, "learning_rate": 5.003472286338029e-05, "loss": 0.0242, "step": 52480 }, { "grad_norm": 0.5879467725753784, "learning_rate": 5.001818816759335e-05, "loss": 0.0271, "step": 52490 }, { "grad_norm": 0.406063973903656, "learning_rate": 5.000165346981738e-05, "loss": 0.0386, "step": 52500 }, { "grad_norm": 0.8427304625511169, "learning_rate": 4.9985118771860595e-05, "loss": 0.0349, "step": 52510 }, { "grad_norm": 0.5285137891769409, "learning_rate": 4.996858407553119e-05, "loss": 0.0432, "step": 52520 }, { "grad_norm": 0.692129373550415, "learning_rate": 4.99520493826374e-05, "loss": 0.0302, "step": 52530 }, { "grad_norm": 0.2546510696411133, "learning_rate": 4.9935514694987384e-05, "loss": 0.028, "step": 52540 }, { "grad_norm": 0.8881853222846985, "learning_rate": 4.991898001438941e-05, "loss": 0.0743, "step": 52550 }, { "grad_norm": 0.5834561586380005, "learning_rate": 4.990244534265166e-05, "loss": 0.0303, "step": 52560 }, { "grad_norm": 0.38770410418510437, "learning_rate": 4.988591068158234e-05, "loss": 0.0361, "step": 52570 }, { "grad_norm": 0.23749427497386932, "learning_rate": 4.986937603298965e-05, "loss": 0.0466, "step": 52580 }, { "grad_norm": 0.47968101501464844, "learning_rate": 4.985284139868181e-05, "loss": 0.0371, "step": 52590 }, { "grad_norm": 0.42248740792274475, "learning_rate": 4.9836306780467e-05, "loss": 0.0447, "step": 52600 }, { "grad_norm": 0.4425451159477234, "learning_rate": 4.981977218015343e-05, "loss": 0.0462, "step": 52610 }, { "grad_norm": 0.5728512406349182, "learning_rate": 4.980323759954932e-05, "loss": 0.0589, "step": 52620 }, { "grad_norm": 0.5048611760139465, "learning_rate": 4.978670304046284e-05, "loss": 0.027, "step": 52630 }, { "grad_norm": 1.3704220056533813, "learning_rate": 4.9770168504702195e-05, "loss": 0.0324, "step": 52640 }, { "grad_norm": 0.38172754645347595, "learning_rate": 4.975363399407559e-05, "loss": 0.0415, "step": 52650 }, { "grad_norm": 0.45604580640792847, "learning_rate": 4.9737099510391194e-05, "loss": 0.041, "step": 52660 }, { "grad_norm": 0.48802223801612854, "learning_rate": 4.972056505545719e-05, "loss": 0.0262, "step": 52670 }, { "grad_norm": 1.2898446321487427, "learning_rate": 4.9704030631081775e-05, "loss": 0.044, "step": 52680 }, { "grad_norm": 0.44575273990631104, "learning_rate": 4.968749623907313e-05, "loss": 0.0392, "step": 52690 }, { "grad_norm": 0.39163193106651306, "learning_rate": 4.967096188123943e-05, "loss": 0.0301, "step": 52700 }, { "grad_norm": 0.3021494150161743, "learning_rate": 4.965442755938884e-05, "loss": 0.0391, "step": 52710 }, { "grad_norm": 0.6403135657310486, "learning_rate": 4.963789327532954e-05, "loss": 0.0457, "step": 52720 }, { "grad_norm": 0.6180774569511414, "learning_rate": 4.962135903086968e-05, "loss": 0.0279, "step": 52730 }, { "grad_norm": 0.5008919835090637, "learning_rate": 4.960482482781741e-05, "loss": 0.033, "step": 52740 }, { "grad_norm": 0.7279577255249023, "learning_rate": 4.958829066798092e-05, "loss": 0.0358, "step": 52750 }, { "grad_norm": 1.4558650255203247, "learning_rate": 4.957175655316835e-05, "loss": 0.0399, "step": 52760 }, { "grad_norm": 0.7625278234481812, "learning_rate": 4.9555222485187834e-05, "loss": 0.0378, "step": 52770 }, { "grad_norm": 0.4385685324668884, "learning_rate": 4.953868846584752e-05, "loss": 0.0287, "step": 52780 }, { "grad_norm": 0.6686869263648987, "learning_rate": 4.9522154496955544e-05, "loss": 0.039, "step": 52790 }, { "grad_norm": 0.4020437002182007, "learning_rate": 4.950562058032001e-05, "loss": 0.0381, "step": 52800 }, { "grad_norm": 0.46255943179130554, "learning_rate": 4.948908671774909e-05, "loss": 0.0316, "step": 52810 }, { "grad_norm": 0.5948241949081421, "learning_rate": 4.947255291105087e-05, "loss": 0.0333, "step": 52820 }, { "grad_norm": 0.36425548791885376, "learning_rate": 4.945601916203348e-05, "loss": 0.024, "step": 52830 }, { "grad_norm": 0.4816477298736572, "learning_rate": 4.9439485472505016e-05, "loss": 0.0316, "step": 52840 }, { "grad_norm": 0.4859177768230438, "learning_rate": 4.942295184427356e-05, "loss": 0.0334, "step": 52850 }, { "grad_norm": 0.4590630531311035, "learning_rate": 4.9406418279147234e-05, "loss": 0.0358, "step": 52860 }, { "grad_norm": 0.508765697479248, "learning_rate": 4.9389884778934116e-05, "loss": 0.0441, "step": 52870 }, { "grad_norm": 0.706359326839447, "learning_rate": 4.937335134544228e-05, "loss": 0.0323, "step": 52880 }, { "grad_norm": 0.433352530002594, "learning_rate": 4.935681798047979e-05, "loss": 0.0417, "step": 52890 }, { "grad_norm": 0.3399324417114258, "learning_rate": 4.934028468585473e-05, "loss": 0.0228, "step": 52900 }, { "grad_norm": 0.5502448081970215, "learning_rate": 4.932375146337514e-05, "loss": 0.0314, "step": 52910 }, { "grad_norm": 0.556752622127533, "learning_rate": 4.930721831484908e-05, "loss": 0.053, "step": 52920 }, { "grad_norm": 0.28492558002471924, "learning_rate": 4.929068524208456e-05, "loss": 0.0236, "step": 52930 }, { "grad_norm": 0.5235396027565002, "learning_rate": 4.927415224688965e-05, "loss": 0.0325, "step": 52940 }, { "grad_norm": 0.3064528703689575, "learning_rate": 4.9257619331072375e-05, "loss": 0.0303, "step": 52950 }, { "grad_norm": 0.3210107982158661, "learning_rate": 4.924108649644072e-05, "loss": 0.0266, "step": 52960 }, { "grad_norm": 0.49757471680641174, "learning_rate": 4.922455374480271e-05, "loss": 0.0449, "step": 52970 }, { "grad_norm": 1.45937180519104, "learning_rate": 4.920802107796634e-05, "loss": 0.0305, "step": 52980 }, { "grad_norm": 0.46599051356315613, "learning_rate": 4.9191488497739564e-05, "loss": 0.0459, "step": 52990 }, { "grad_norm": 0.9295778870582581, "learning_rate": 4.917495600593042e-05, "loss": 0.0376, "step": 53000 }, { "grad_norm": 0.34328609704971313, "learning_rate": 4.915842360434684e-05, "loss": 0.0556, "step": 53010 }, { "grad_norm": 0.749718189239502, "learning_rate": 4.914189129479679e-05, "loss": 0.0348, "step": 53020 }, { "grad_norm": 0.4329160451889038, "learning_rate": 4.912535907908822e-05, "loss": 0.0377, "step": 53030 }, { "grad_norm": 0.4990832805633545, "learning_rate": 4.910882695902907e-05, "loss": 0.0398, "step": 53040 }, { "grad_norm": 1.8649849891662598, "learning_rate": 4.9092294936427234e-05, "loss": 0.0301, "step": 53050 }, { "grad_norm": 0.2934822142124176, "learning_rate": 4.9075763013090685e-05, "loss": 0.0409, "step": 53060 }, { "grad_norm": 0.3315041661262512, "learning_rate": 4.9059231190827305e-05, "loss": 0.0329, "step": 53070 }, { "grad_norm": 0.5230976939201355, "learning_rate": 4.904269947144499e-05, "loss": 0.0324, "step": 53080 }, { "grad_norm": 0.46520888805389404, "learning_rate": 4.902616785675161e-05, "loss": 0.0263, "step": 53090 }, { "grad_norm": 0.37640702724456787, "learning_rate": 4.900963634855505e-05, "loss": 0.0292, "step": 53100 }, { "grad_norm": 0.6257944107055664, "learning_rate": 4.8993104948663173e-05, "loss": 0.037, "step": 53110 }, { "grad_norm": 0.33816349506378174, "learning_rate": 4.8976573658883823e-05, "loss": 0.0335, "step": 53120 }, { "grad_norm": 1.2216203212738037, "learning_rate": 4.896004248102484e-05, "loss": 0.0368, "step": 53130 }, { "grad_norm": 0.31669482588768005, "learning_rate": 4.8943511416894045e-05, "loss": 0.0275, "step": 53140 }, { "grad_norm": 0.35258904099464417, "learning_rate": 4.892698046829926e-05, "loss": 0.0272, "step": 53150 }, { "grad_norm": 0.7553477883338928, "learning_rate": 4.8910449637048275e-05, "loss": 0.0345, "step": 53160 }, { "grad_norm": 0.7009041905403137, "learning_rate": 4.8893918924948884e-05, "loss": 0.037, "step": 53170 }, { "grad_norm": 1.0470225811004639, "learning_rate": 4.8877388333808835e-05, "loss": 0.0373, "step": 53180 }, { "grad_norm": 0.4334055185317993, "learning_rate": 4.8860857865435925e-05, "loss": 0.0341, "step": 53190 }, { "grad_norm": 0.49840250611305237, "learning_rate": 4.884432752163789e-05, "loss": 0.0329, "step": 53200 }, { "grad_norm": 0.45876193046569824, "learning_rate": 4.8827797304222466e-05, "loss": 0.0323, "step": 53210 }, { "grad_norm": 0.356576144695282, "learning_rate": 4.8811267214997366e-05, "loss": 0.0345, "step": 53220 }, { "grad_norm": 0.3005582094192505, "learning_rate": 4.879473725577029e-05, "loss": 0.0283, "step": 53230 }, { "grad_norm": 0.33021920919418335, "learning_rate": 4.877820742834891e-05, "loss": 0.0486, "step": 53240 }, { "grad_norm": 0.39868661761283875, "learning_rate": 4.8761677734540956e-05, "loss": 0.039, "step": 53250 }, { "grad_norm": 0.6749054789543152, "learning_rate": 4.874514817615407e-05, "loss": 0.0295, "step": 53260 }, { "grad_norm": 0.4590545892715454, "learning_rate": 4.872861875499588e-05, "loss": 0.0357, "step": 53270 }, { "grad_norm": 0.520270586013794, "learning_rate": 4.871208947287404e-05, "loss": 0.0306, "step": 53280 }, { "grad_norm": 0.568151593208313, "learning_rate": 4.869556033159615e-05, "loss": 0.0374, "step": 53290 }, { "grad_norm": 0.5273966193199158, "learning_rate": 4.8679031332969835e-05, "loss": 0.03, "step": 53300 }, { "grad_norm": 0.5013478398323059, "learning_rate": 4.8662502478802635e-05, "loss": 0.0453, "step": 53310 }, { "grad_norm": 0.38333579897880554, "learning_rate": 4.8645973770902176e-05, "loss": 0.0421, "step": 53320 }, { "grad_norm": 0.48671755194664, "learning_rate": 4.862944521107599e-05, "loss": 0.0333, "step": 53330 }, { "grad_norm": 0.4081857204437256, "learning_rate": 4.861291680113162e-05, "loss": 0.0443, "step": 53340 }, { "grad_norm": 0.45790135860443115, "learning_rate": 4.8596388542876563e-05, "loss": 0.0425, "step": 53350 }, { "grad_norm": 0.5644410848617554, "learning_rate": 4.857986043811836e-05, "loss": 0.0352, "step": 53360 }, { "grad_norm": 0.5118959546089172, "learning_rate": 4.856333248866448e-05, "loss": 0.0297, "step": 53370 }, { "grad_norm": 0.47911593317985535, "learning_rate": 4.85468046963224e-05, "loss": 0.038, "step": 53380 }, { "grad_norm": 0.25927823781967163, "learning_rate": 4.853027706289957e-05, "loss": 0.0383, "step": 53390 }, { "grad_norm": 0.6137120723724365, "learning_rate": 4.851374959020344e-05, "loss": 0.0464, "step": 53400 }, { "grad_norm": 0.3062782287597656, "learning_rate": 4.849722228004141e-05, "loss": 0.0283, "step": 53410 }, { "grad_norm": 0.48176953196525574, "learning_rate": 4.848069513422091e-05, "loss": 0.0404, "step": 53420 }, { "grad_norm": 0.3440808653831482, "learning_rate": 4.8464168154549294e-05, "loss": 0.0272, "step": 53430 }, { "grad_norm": 0.49497687816619873, "learning_rate": 4.844764134283392e-05, "loss": 0.0247, "step": 53440 }, { "grad_norm": 0.29547566175460815, "learning_rate": 4.843111470088218e-05, "loss": 0.0248, "step": 53450 }, { "grad_norm": 0.5831066966056824, "learning_rate": 4.841458823050137e-05, "loss": 0.0285, "step": 53460 }, { "grad_norm": 0.3945498764514923, "learning_rate": 4.8398061933498816e-05, "loss": 0.0257, "step": 53470 }, { "grad_norm": 0.44741979241371155, "learning_rate": 4.83815358116818e-05, "loss": 0.0303, "step": 53480 }, { "grad_norm": 0.5350257158279419, "learning_rate": 4.8365009866857584e-05, "loss": 0.0311, "step": 53490 }, { "grad_norm": 0.6570818424224854, "learning_rate": 4.834848410083342e-05, "loss": 0.0445, "step": 53500 }, { "grad_norm": 0.4447086453437805, "learning_rate": 4.833195851541657e-05, "loss": 0.0342, "step": 53510 }, { "grad_norm": 0.5015693306922913, "learning_rate": 4.831543311241423e-05, "loss": 0.0365, "step": 53520 }, { "grad_norm": 0.8470878005027771, "learning_rate": 4.829890789363359e-05, "loss": 0.0346, "step": 53530 }, { "grad_norm": 0.4472026526927948, "learning_rate": 4.828238286088182e-05, "loss": 0.0337, "step": 53540 }, { "grad_norm": 0.6416880488395691, "learning_rate": 4.826585801596609e-05, "loss": 0.0403, "step": 53550 }, { "grad_norm": 0.4226163923740387, "learning_rate": 4.8249333360693486e-05, "loss": 0.0313, "step": 53560 }, { "grad_norm": 0.25284823775291443, "learning_rate": 4.8232808896871185e-05, "loss": 0.0407, "step": 53570 }, { "grad_norm": 0.39780643582344055, "learning_rate": 4.821628462630624e-05, "loss": 0.0231, "step": 53580 }, { "grad_norm": 0.3594866394996643, "learning_rate": 4.8199760550805735e-05, "loss": 0.0361, "step": 53590 }, { "grad_norm": 0.43605268001556396, "learning_rate": 4.818323667217669e-05, "loss": 0.033, "step": 53600 }, { "grad_norm": 0.5919357538223267, "learning_rate": 4.8166712992226165e-05, "loss": 0.0338, "step": 53610 }, { "grad_norm": 0.8303390741348267, "learning_rate": 4.8150189512761154e-05, "loss": 0.027, "step": 53620 }, { "grad_norm": 0.29911699891090393, "learning_rate": 4.813366623558862e-05, "loss": 0.031, "step": 53630 }, { "grad_norm": 0.44811511039733887, "learning_rate": 4.811714316251554e-05, "loss": 0.0326, "step": 53640 }, { "grad_norm": 0.40874338150024414, "learning_rate": 4.810062029534886e-05, "loss": 0.0279, "step": 53650 }, { "grad_norm": 0.41754743456840515, "learning_rate": 4.8084097635895505e-05, "loss": 0.0352, "step": 53660 }, { "grad_norm": 0.7416878342628479, "learning_rate": 4.8067575185962335e-05, "loss": 0.0286, "step": 53670 }, { "grad_norm": 0.468153178691864, "learning_rate": 4.805105294735625e-05, "loss": 0.0372, "step": 53680 }, { "grad_norm": 0.5171918272972107, "learning_rate": 4.803453092188406e-05, "loss": 0.0307, "step": 53690 }, { "grad_norm": 0.5479240417480469, "learning_rate": 4.801800911135263e-05, "loss": 0.0332, "step": 53700 }, { "grad_norm": 0.5914439558982849, "learning_rate": 4.8001487517568744e-05, "loss": 0.027, "step": 53710 }, { "grad_norm": 0.35606124997138977, "learning_rate": 4.798496614233918e-05, "loss": 0.0236, "step": 53720 }, { "grad_norm": 0.4048880636692047, "learning_rate": 4.796844498747069e-05, "loss": 0.0237, "step": 53730 }, { "grad_norm": 1.1653815507888794, "learning_rate": 4.795192405477e-05, "loss": 0.0283, "step": 53740 }, { "grad_norm": 0.4454186260700226, "learning_rate": 4.79354033460438e-05, "loss": 0.0339, "step": 53750 }, { "grad_norm": 0.34872499108314514, "learning_rate": 4.79188828630988e-05, "loss": 0.0297, "step": 53760 }, { "grad_norm": 0.34466493129730225, "learning_rate": 4.790236260774165e-05, "loss": 0.0329, "step": 53770 }, { "grad_norm": 0.3894135653972626, "learning_rate": 4.788584258177896e-05, "loss": 0.0263, "step": 53780 }, { "grad_norm": 0.35901838541030884, "learning_rate": 4.7869322787017356e-05, "loss": 0.0292, "step": 53790 }, { "grad_norm": 0.38139334321022034, "learning_rate": 4.78528032252634e-05, "loss": 0.0324, "step": 53800 }, { "grad_norm": 0.30550748109817505, "learning_rate": 4.783628389832366e-05, "loss": 0.0427, "step": 53810 }, { "grad_norm": 0.42684710025787354, "learning_rate": 4.7819764808004644e-05, "loss": 0.0318, "step": 53820 }, { "grad_norm": 1.002841591835022, "learning_rate": 4.780324595611289e-05, "loss": 0.0402, "step": 53830 }, { "grad_norm": 0.6878487467765808, "learning_rate": 4.7786727344454854e-05, "loss": 0.0415, "step": 53840 }, { "grad_norm": 0.46168380975723267, "learning_rate": 4.777020897483697e-05, "loss": 0.0292, "step": 53850 }, { "grad_norm": 0.5669446587562561, "learning_rate": 4.775369084906569e-05, "loss": 0.032, "step": 53860 }, { "grad_norm": 0.591606080532074, "learning_rate": 4.773717296894742e-05, "loss": 0.0413, "step": 53870 }, { "grad_norm": 0.7782503366470337, "learning_rate": 4.772065533628848e-05, "loss": 0.043, "step": 53880 }, { "grad_norm": 0.5232882499694824, "learning_rate": 4.7704137952895265e-05, "loss": 0.0391, "step": 53890 }, { "grad_norm": 0.5564358830451965, "learning_rate": 4.768762082057407e-05, "loss": 0.0391, "step": 53900 }, { "grad_norm": 0.42609092593193054, "learning_rate": 4.7671103941131186e-05, "loss": 0.0245, "step": 53910 }, { "grad_norm": 0.407428503036499, "learning_rate": 4.765458731637287e-05, "loss": 0.0352, "step": 53920 }, { "grad_norm": 0.4165244400501251, "learning_rate": 4.763807094810537e-05, "loss": 0.0428, "step": 53930 }, { "grad_norm": 0.5716474652290344, "learning_rate": 4.762155483813487e-05, "loss": 0.0336, "step": 53940 }, { "grad_norm": 0.8690997362136841, "learning_rate": 4.760503898826754e-05, "loss": 0.0304, "step": 53950 }, { "grad_norm": 0.3105238080024719, "learning_rate": 4.758852340030957e-05, "loss": 0.0253, "step": 53960 }, { "grad_norm": 2.0775020122528076, "learning_rate": 4.757200807606705e-05, "loss": 0.0337, "step": 53970 }, { "grad_norm": 0.4465968608856201, "learning_rate": 4.755549301734607e-05, "loss": 0.0323, "step": 53980 }, { "grad_norm": 0.337924063205719, "learning_rate": 4.753897822595271e-05, "loss": 0.0313, "step": 53990 }, { "grad_norm": 0.4936814606189728, "learning_rate": 4.7522463703692994e-05, "loss": 0.0298, "step": 54000 }, { "grad_norm": 0.471444308757782, "learning_rate": 4.75059494523729e-05, "loss": 0.0284, "step": 54010 }, { "grad_norm": 0.4453575611114502, "learning_rate": 4.748943547379843e-05, "loss": 0.0259, "step": 54020 }, { "grad_norm": 0.25824230909347534, "learning_rate": 4.7472921769775535e-05, "loss": 0.0263, "step": 54030 }, { "grad_norm": 0.43844813108444214, "learning_rate": 4.745640834211011e-05, "loss": 0.0275, "step": 54040 }, { "grad_norm": 0.49120888113975525, "learning_rate": 4.7439895192608054e-05, "loss": 0.033, "step": 54050 }, { "grad_norm": 0.39640679955482483, "learning_rate": 4.7423382323075186e-05, "loss": 0.0269, "step": 54060 }, { "grad_norm": 0.502277135848999, "learning_rate": 4.740686973531737e-05, "loss": 0.0325, "step": 54070 }, { "grad_norm": 0.8915520310401917, "learning_rate": 4.7390357431140375e-05, "loss": 0.0453, "step": 54080 }, { "grad_norm": 0.43855372071266174, "learning_rate": 4.737384541234998e-05, "loss": 0.0258, "step": 54090 }, { "grad_norm": 0.5388268232345581, "learning_rate": 4.735733368075189e-05, "loss": 0.0419, "step": 54100 }, { "grad_norm": 0.5603451728820801, "learning_rate": 4.734082223815182e-05, "loss": 0.0343, "step": 54110 }, { "grad_norm": 0.2972150146961212, "learning_rate": 4.7324311086355436e-05, "loss": 0.0184, "step": 54120 }, { "grad_norm": 0.38688603043556213, "learning_rate": 4.730780022716836e-05, "loss": 0.0326, "step": 54130 }, { "grad_norm": 0.4538649022579193, "learning_rate": 4.729128966239619e-05, "loss": 0.0257, "step": 54140 }, { "grad_norm": 0.47157353162765503, "learning_rate": 4.7274779393844536e-05, "loss": 0.0264, "step": 54150 }, { "grad_norm": 0.5261023640632629, "learning_rate": 4.725826942331891e-05, "loss": 0.0284, "step": 54160 }, { "grad_norm": 0.46263447403907776, "learning_rate": 4.724175975262481e-05, "loss": 0.0242, "step": 54170 }, { "grad_norm": 0.2679344713687897, "learning_rate": 4.722525038356772e-05, "loss": 0.0243, "step": 54180 }, { "grad_norm": 0.4022999703884125, "learning_rate": 4.7208741317953075e-05, "loss": 0.0238, "step": 54190 }, { "grad_norm": 0.5915663242340088, "learning_rate": 4.7192232557586264e-05, "loss": 0.0223, "step": 54200 }, { "grad_norm": 0.2858779728412628, "learning_rate": 4.717572410427269e-05, "loss": 0.0297, "step": 54210 }, { "grad_norm": 0.5574103593826294, "learning_rate": 4.71592159598177e-05, "loss": 0.033, "step": 54220 }, { "grad_norm": 0.4375740885734558, "learning_rate": 4.714270812602657e-05, "loss": 0.0263, "step": 54230 }, { "grad_norm": 0.5255787372589111, "learning_rate": 4.712620060470458e-05, "loss": 0.0396, "step": 54240 }, { "grad_norm": 0.43417349457740784, "learning_rate": 4.7109693397656986e-05, "loss": 0.0326, "step": 54250 }, { "grad_norm": 0.6627196073532104, "learning_rate": 4.709318650668894e-05, "loss": 0.022, "step": 54260 }, { "grad_norm": 0.6293396949768066, "learning_rate": 4.7076679933605676e-05, "loss": 0.049, "step": 54270 }, { "grad_norm": 0.4050044119358063, "learning_rate": 4.70601736802123e-05, "loss": 0.0304, "step": 54280 }, { "grad_norm": 0.45381858944892883, "learning_rate": 4.704366774831391e-05, "loss": 0.0331, "step": 54290 }, { "grad_norm": 0.5722429156303406, "learning_rate": 4.702716213971557e-05, "loss": 0.0297, "step": 54300 }, { "grad_norm": 0.4803454279899597, "learning_rate": 4.70106568562223e-05, "loss": 0.0279, "step": 54310 }, { "grad_norm": 0.3847081959247589, "learning_rate": 4.6994151899639105e-05, "loss": 0.0289, "step": 54320 }, { "grad_norm": 0.49943745136260986, "learning_rate": 4.697764727177093e-05, "loss": 0.0314, "step": 54330 }, { "grad_norm": 0.5707786679267883, "learning_rate": 4.696114297442272e-05, "loss": 0.0392, "step": 54340 }, { "grad_norm": 0.6952166557312012, "learning_rate": 4.694463900939933e-05, "loss": 0.031, "step": 54350 }, { "grad_norm": 0.674888014793396, "learning_rate": 4.692813537850564e-05, "loss": 0.0479, "step": 54360 }, { "grad_norm": 0.4849327504634857, "learning_rate": 4.6911632083546445e-05, "loss": 0.0404, "step": 54370 }, { "grad_norm": 0.382771760225296, "learning_rate": 4.6895129126326525e-05, "loss": 0.0325, "step": 54380 }, { "grad_norm": 0.6090958714485168, "learning_rate": 4.687862650865059e-05, "loss": 0.0385, "step": 54390 }, { "grad_norm": 0.7443493008613586, "learning_rate": 4.686212423232339e-05, "loss": 0.0396, "step": 54400 }, { "grad_norm": 0.5225640535354614, "learning_rate": 4.6845622299149565e-05, "loss": 0.0346, "step": 54410 }, { "grad_norm": 0.38558927178382874, "learning_rate": 4.682912071093374e-05, "loss": 0.0296, "step": 54420 }, { "grad_norm": 0.5448931455612183, "learning_rate": 4.681261946948052e-05, "loss": 0.0248, "step": 54430 }, { "grad_norm": 0.4910668134689331, "learning_rate": 4.679611857659443e-05, "loss": 0.034, "step": 54440 }, { "grad_norm": 0.5231115818023682, "learning_rate": 4.6779618034079975e-05, "loss": 0.0282, "step": 54450 }, { "grad_norm": 0.3602529764175415, "learning_rate": 4.6763117843741675e-05, "loss": 0.0366, "step": 54460 }, { "grad_norm": 0.3378746211528778, "learning_rate": 4.674661800738394e-05, "loss": 0.0421, "step": 54470 }, { "grad_norm": 0.38968175649642944, "learning_rate": 4.673011852681117e-05, "loss": 0.0356, "step": 54480 }, { "grad_norm": 1.3396779298782349, "learning_rate": 4.671361940382772e-05, "loss": 0.0438, "step": 54490 }, { "grad_norm": 0.7799427509307861, "learning_rate": 4.669712064023791e-05, "loss": 0.0329, "step": 54500 }, { "grad_norm": 0.4750667214393616, "learning_rate": 4.668062223784602e-05, "loss": 0.0266, "step": 54510 }, { "grad_norm": 0.29357168078422546, "learning_rate": 4.6664124198456274e-05, "loss": 0.0353, "step": 54520 }, { "grad_norm": 0.47818076610565186, "learning_rate": 4.6647626523872905e-05, "loss": 0.0253, "step": 54530 }, { "grad_norm": 0.5805109739303589, "learning_rate": 4.6631129215900066e-05, "loss": 0.0284, "step": 54540 }, { "grad_norm": 0.4551585018634796, "learning_rate": 4.661463227634187e-05, "loss": 0.0347, "step": 54550 }, { "grad_norm": 0.485027015209198, "learning_rate": 4.659813570700238e-05, "loss": 0.0334, "step": 54560 }, { "grad_norm": 0.5250878930091858, "learning_rate": 4.658163950968568e-05, "loss": 0.0299, "step": 54570 }, { "grad_norm": 0.6295104622840881, "learning_rate": 4.6565143686195726e-05, "loss": 0.0299, "step": 54580 }, { "grad_norm": 0.23646630346775055, "learning_rate": 4.654864823833649e-05, "loss": 0.0294, "step": 54590 }, { "grad_norm": 1.0920004844665527, "learning_rate": 4.65321531679119e-05, "loss": 0.0384, "step": 54600 }, { "grad_norm": 0.4234474301338196, "learning_rate": 4.6515658476725834e-05, "loss": 0.0372, "step": 54610 }, { "grad_norm": 0.5948024392127991, "learning_rate": 4.649916416658212e-05, "loss": 0.0388, "step": 54620 }, { "grad_norm": 0.4203566312789917, "learning_rate": 4.648267023928455e-05, "loss": 0.0346, "step": 54630 }, { "grad_norm": 0.8050843477249146, "learning_rate": 4.646617669663687e-05, "loss": 0.0365, "step": 54640 }, { "grad_norm": 0.35207098722457886, "learning_rate": 4.644968354044277e-05, "loss": 0.0231, "step": 54650 }, { "grad_norm": 0.3152943253517151, "learning_rate": 4.643319077250596e-05, "loss": 0.0224, "step": 54660 }, { "grad_norm": 0.4751321077346802, "learning_rate": 4.6416698394630055e-05, "loss": 0.0252, "step": 54670 }, { "grad_norm": 0.8342933058738708, "learning_rate": 4.640020640861862e-05, "loss": 0.0374, "step": 54680 }, { "grad_norm": 0.5819252133369446, "learning_rate": 4.6383714816275195e-05, "loss": 0.0518, "step": 54690 }, { "grad_norm": 0.2915806472301483, "learning_rate": 4.636722361940329e-05, "loss": 0.0293, "step": 54700 }, { "grad_norm": 0.3259567618370056, "learning_rate": 4.635073281980632e-05, "loss": 0.0332, "step": 54710 }, { "grad_norm": 0.42482730746269226, "learning_rate": 4.633424241928775e-05, "loss": 0.0307, "step": 54720 }, { "grad_norm": 0.6765640377998352, "learning_rate": 4.6317752419650914e-05, "loss": 0.0337, "step": 54730 }, { "grad_norm": 0.37840011715888977, "learning_rate": 4.630126282269914e-05, "loss": 0.034, "step": 54740 }, { "grad_norm": 0.4337882697582245, "learning_rate": 4.628477363023569e-05, "loss": 0.0242, "step": 54750 }, { "grad_norm": 0.37210169434547424, "learning_rate": 4.626828484406383e-05, "loss": 0.0356, "step": 54760 }, { "grad_norm": 0.4766532778739929, "learning_rate": 4.625179646598669e-05, "loss": 0.034, "step": 54770 }, { "grad_norm": 0.5008560419082642, "learning_rate": 4.6235308497807487e-05, "loss": 0.023, "step": 54780 }, { "grad_norm": 0.9525343179702759, "learning_rate": 4.6218820941329276e-05, "loss": 0.0339, "step": 54790 }, { "grad_norm": 0.4470615088939667, "learning_rate": 4.620233379835513e-05, "loss": 0.0287, "step": 54800 }, { "grad_norm": 0.5310224890708923, "learning_rate": 4.618584707068804e-05, "loss": 0.0451, "step": 54810 }, { "grad_norm": 0.3215891122817993, "learning_rate": 4.616936076013099e-05, "loss": 0.0375, "step": 54820 }, { "grad_norm": 0.40979447960853577, "learning_rate": 4.615287486848688e-05, "loss": 0.0354, "step": 54830 }, { "grad_norm": 0.43390989303588867, "learning_rate": 4.6136389397558584e-05, "loss": 0.0258, "step": 54840 }, { "grad_norm": 0.49904245138168335, "learning_rate": 4.6119904349148935e-05, "loss": 0.0259, "step": 54850 }, { "grad_norm": 0.544006884098053, "learning_rate": 4.6103419725060726e-05, "loss": 0.0372, "step": 54860 }, { "grad_norm": 0.4986349940299988, "learning_rate": 4.608693552709667e-05, "loss": 0.0315, "step": 54870 }, { "grad_norm": 0.5007418990135193, "learning_rate": 4.607045175705947e-05, "loss": 0.0405, "step": 54880 }, { "grad_norm": 0.7889307737350464, "learning_rate": 4.605396841675176e-05, "loss": 0.0315, "step": 54890 }, { "grad_norm": 0.35121092200279236, "learning_rate": 4.6037485507976094e-05, "loss": 0.0344, "step": 54900 }, { "grad_norm": 0.5250710844993591, "learning_rate": 4.602100303253509e-05, "loss": 0.042, "step": 54910 }, { "grad_norm": 1.091752290725708, "learning_rate": 4.600452099223122e-05, "loss": 0.0293, "step": 54920 }, { "grad_norm": 0.5448305606842041, "learning_rate": 4.598803938886691e-05, "loss": 0.029, "step": 54930 }, { "grad_norm": 0.5833532214164734, "learning_rate": 4.597155822424459e-05, "loss": 0.0489, "step": 54940 }, { "grad_norm": 0.33368349075317383, "learning_rate": 4.59550775001666e-05, "loss": 0.0323, "step": 54950 }, { "grad_norm": 0.5970183610916138, "learning_rate": 4.5938597218435234e-05, "loss": 0.0333, "step": 54960 }, { "grad_norm": 1.0201539993286133, "learning_rate": 4.592211738085279e-05, "loss": 0.0327, "step": 54970 }, { "grad_norm": 0.4173247516155243, "learning_rate": 4.590563798922146e-05, "loss": 0.0369, "step": 54980 }, { "grad_norm": 0.5228488445281982, "learning_rate": 4.5889159045343404e-05, "loss": 0.0267, "step": 54990 }, { "grad_norm": 0.5762671828269958, "learning_rate": 4.5872680551020734e-05, "loss": 0.0542, "step": 55000 }, { "grad_norm": 0.33737465739250183, "learning_rate": 4.585620250805552e-05, "loss": 0.0294, "step": 55010 }, { "grad_norm": 0.6602696776390076, "learning_rate": 4.583972491824975e-05, "loss": 0.0342, "step": 55020 }, { "grad_norm": 0.2508729100227356, "learning_rate": 4.582324778340541e-05, "loss": 0.0236, "step": 55030 }, { "grad_norm": 7.009443759918213, "learning_rate": 4.5806771105324424e-05, "loss": 0.0507, "step": 55040 }, { "grad_norm": 0.6951577067375183, "learning_rate": 4.579029488580865e-05, "loss": 0.0355, "step": 55050 }, { "grad_norm": 0.6418396830558777, "learning_rate": 4.5773819126659876e-05, "loss": 0.0331, "step": 55060 }, { "grad_norm": 0.34423694014549255, "learning_rate": 4.575734382967991e-05, "loss": 0.0371, "step": 55070 }, { "grad_norm": 1.2561765909194946, "learning_rate": 4.574086899667044e-05, "loss": 0.03, "step": 55080 }, { "grad_norm": 0.478003591299057, "learning_rate": 4.5724394629433106e-05, "loss": 0.0353, "step": 55090 }, { "grad_norm": 0.5288689732551575, "learning_rate": 4.570792072976958e-05, "loss": 0.0424, "step": 55100 }, { "grad_norm": 0.6653320789337158, "learning_rate": 4.569144729948138e-05, "loss": 0.0298, "step": 55110 }, { "grad_norm": 0.4108215868473053, "learning_rate": 4.5674974340370033e-05, "loss": 0.0259, "step": 55120 }, { "grad_norm": 1.4875612258911133, "learning_rate": 4.565850185423699e-05, "loss": 0.0399, "step": 55130 }, { "grad_norm": 0.5882405638694763, "learning_rate": 4.564202984288365e-05, "loss": 0.0382, "step": 55140 }, { "grad_norm": 0.5301520824432373, "learning_rate": 4.5625558308111356e-05, "loss": 0.0291, "step": 55150 }, { "grad_norm": 0.46100589632987976, "learning_rate": 4.560908725172145e-05, "loss": 0.0266, "step": 55160 }, { "grad_norm": 0.8683850765228271, "learning_rate": 4.559261667551516e-05, "loss": 0.0346, "step": 55170 }, { "grad_norm": 0.4877939522266388, "learning_rate": 4.5576146581293685e-05, "loss": 0.0239, "step": 55180 }, { "grad_norm": 0.262178897857666, "learning_rate": 4.5559676970858164e-05, "loss": 0.0403, "step": 55190 }, { "grad_norm": 0.6409838795661926, "learning_rate": 4.5543207846009705e-05, "loss": 0.0402, "step": 55200 }, { "grad_norm": 0.7198663949966431, "learning_rate": 4.552673920854933e-05, "loss": 0.0455, "step": 55210 }, { "grad_norm": 0.4129725694656372, "learning_rate": 4.5510271060278006e-05, "loss": 0.0319, "step": 55220 }, { "grad_norm": 0.31348150968551636, "learning_rate": 4.5493803402996704e-05, "loss": 0.0337, "step": 55230 }, { "grad_norm": 0.2871032655239105, "learning_rate": 4.5477336238506295e-05, "loss": 0.0292, "step": 55240 }, { "grad_norm": 0.5050691962242126, "learning_rate": 4.546086956860759e-05, "loss": 0.0211, "step": 55250 }, { "grad_norm": 0.3277261555194855, "learning_rate": 4.5444403395101374e-05, "loss": 0.0375, "step": 55260 }, { "grad_norm": 0.3849794566631317, "learning_rate": 4.542793771978834e-05, "loss": 0.0268, "step": 55270 }, { "grad_norm": 0.44776278734207153, "learning_rate": 4.541147254446916e-05, "loss": 0.0347, "step": 55280 }, { "grad_norm": 0.6271728277206421, "learning_rate": 4.539500787094445e-05, "loss": 0.0327, "step": 55290 }, { "grad_norm": 0.4206051826477051, "learning_rate": 4.537854370101475e-05, "loss": 0.0479, "step": 55300 }, { "grad_norm": 0.21932117640972137, "learning_rate": 4.5362080036480566e-05, "loss": 0.0206, "step": 55310 }, { "grad_norm": 0.36424732208251953, "learning_rate": 4.534561687914235e-05, "loss": 0.049, "step": 55320 }, { "grad_norm": 0.5207931995391846, "learning_rate": 4.5329154230800466e-05, "loss": 0.054, "step": 55330 }, { "grad_norm": 0.8081684112548828, "learning_rate": 4.531269209325525e-05, "loss": 0.0351, "step": 55340 }, { "grad_norm": 0.32511231303215027, "learning_rate": 4.529623046830696e-05, "loss": 0.0331, "step": 55350 }, { "grad_norm": 0.5600776672363281, "learning_rate": 4.5279769357755866e-05, "loss": 0.0371, "step": 55360 }, { "grad_norm": 0.9983066320419312, "learning_rate": 4.5263308763402084e-05, "loss": 0.0276, "step": 55370 }, { "grad_norm": 0.7586212754249573, "learning_rate": 4.524684868704574e-05, "loss": 0.0342, "step": 55380 }, { "grad_norm": 0.5605604648590088, "learning_rate": 4.523038913048688e-05, "loss": 0.0325, "step": 55390 }, { "grad_norm": 0.3945106565952301, "learning_rate": 4.521393009552549e-05, "loss": 0.0315, "step": 55400 }, { "grad_norm": 0.39432576298713684, "learning_rate": 4.5197471583961484e-05, "loss": 0.0223, "step": 55410 }, { "grad_norm": 0.4216057062149048, "learning_rate": 4.51810135975948e-05, "loss": 0.0268, "step": 55420 }, { "grad_norm": 0.3694663643836975, "learning_rate": 4.516455613822521e-05, "loss": 0.032, "step": 55430 }, { "grad_norm": 0.3055562973022461, "learning_rate": 4.514809920765249e-05, "loss": 0.0368, "step": 55440 }, { "grad_norm": 0.4836726784706116, "learning_rate": 4.5131642807676356e-05, "loss": 0.0288, "step": 55450 }, { "grad_norm": 0.5555112957954407, "learning_rate": 4.511518694009644e-05, "loss": 0.0365, "step": 55460 }, { "grad_norm": 0.32276231050491333, "learning_rate": 4.509873160671231e-05, "loss": 0.0235, "step": 55470 }, { "grad_norm": 1.176817536354065, "learning_rate": 4.508227680932356e-05, "loss": 0.0331, "step": 55480 }, { "grad_norm": 0.40703052282333374, "learning_rate": 4.5065822549729605e-05, "loss": 0.0267, "step": 55490 }, { "grad_norm": 0.6075600981712341, "learning_rate": 4.504936882972989e-05, "loss": 0.0328, "step": 55500 }, { "grad_norm": 0.5671252608299255, "learning_rate": 4.503291565112374e-05, "loss": 0.0292, "step": 55510 }, { "grad_norm": 0.4472901225090027, "learning_rate": 4.501646301571048e-05, "loss": 0.0283, "step": 55520 }, { "grad_norm": 0.6749748587608337, "learning_rate": 4.500001092528933e-05, "loss": 0.0352, "step": 55530 }, { "grad_norm": 0.3180055618286133, "learning_rate": 4.4983559381659455e-05, "loss": 0.0361, "step": 55540 }, { "grad_norm": 0.7884342670440674, "learning_rate": 4.496710838661999e-05, "loss": 0.0365, "step": 55550 }, { "grad_norm": 4.810285568237305, "learning_rate": 4.495065794196999e-05, "loss": 0.0319, "step": 55560 }, { "grad_norm": 1.753098726272583, "learning_rate": 4.4934208049508455e-05, "loss": 0.0401, "step": 55570 }, { "grad_norm": 0.5153861045837402, "learning_rate": 4.4917758711034303e-05, "loss": 0.0431, "step": 55580 }, { "grad_norm": 0.6177926063537598, "learning_rate": 4.490130992834642e-05, "loss": 0.0326, "step": 55590 }, { "grad_norm": 0.5583807826042175, "learning_rate": 4.488486170324359e-05, "loss": 0.0319, "step": 55600 }, { "grad_norm": 0.6352322101593018, "learning_rate": 4.486841403752463e-05, "loss": 0.0324, "step": 55610 }, { "grad_norm": 0.4307059347629547, "learning_rate": 4.4851966932988174e-05, "loss": 0.0507, "step": 55620 }, { "grad_norm": 0.5078892111778259, "learning_rate": 4.4835520391432894e-05, "loss": 0.0395, "step": 55630 }, { "grad_norm": 0.37493255734443665, "learning_rate": 4.481907441465733e-05, "loss": 0.0249, "step": 55640 }, { "grad_norm": 0.4824991822242737, "learning_rate": 4.480262900446e-05, "loss": 0.0258, "step": 55650 }, { "grad_norm": 0.36188405752182007, "learning_rate": 4.478618416263933e-05, "loss": 0.0232, "step": 55660 }, { "grad_norm": 0.3844187557697296, "learning_rate": 4.476973989099373e-05, "loss": 0.0344, "step": 55670 }, { "grad_norm": 0.5509741902351379, "learning_rate": 4.475329619132153e-05, "loss": 0.0346, "step": 55680 }, { "grad_norm": 0.7126777768135071, "learning_rate": 4.4736853065420956e-05, "loss": 0.024, "step": 55690 }, { "grad_norm": 0.46763524413108826, "learning_rate": 4.472041051509023e-05, "loss": 0.0257, "step": 55700 }, { "grad_norm": 0.33638808131217957, "learning_rate": 4.470396854212746e-05, "loss": 0.0227, "step": 55710 }, { "grad_norm": 0.30464622378349304, "learning_rate": 4.468752714833072e-05, "loss": 0.0259, "step": 55720 }, { "grad_norm": 0.4168369472026825, "learning_rate": 4.467108633549803e-05, "loss": 0.0296, "step": 55730 }, { "grad_norm": 0.9276061654090881, "learning_rate": 4.4654646105427335e-05, "loss": 0.028, "step": 55740 }, { "grad_norm": 0.6909464001655579, "learning_rate": 4.463820645991651e-05, "loss": 0.0312, "step": 55750 }, { "grad_norm": 0.3257536292076111, "learning_rate": 4.462176740076334e-05, "loss": 0.0519, "step": 55760 }, { "grad_norm": 0.5314671993255615, "learning_rate": 4.4605328929765624e-05, "loss": 0.0371, "step": 55770 }, { "grad_norm": 0.34192952513694763, "learning_rate": 4.458889104872102e-05, "loss": 0.0263, "step": 55780 }, { "grad_norm": 0.3909381330013275, "learning_rate": 4.457245375942715e-05, "loss": 0.0367, "step": 55790 }, { "grad_norm": 0.3541682958602905, "learning_rate": 4.455601706368157e-05, "loss": 0.0284, "step": 55800 }, { "grad_norm": 0.31134670972824097, "learning_rate": 4.453958096328179e-05, "loss": 0.0252, "step": 55810 }, { "grad_norm": 0.5170889496803284, "learning_rate": 4.452314546002523e-05, "loss": 0.0326, "step": 55820 }, { "grad_norm": 0.509580671787262, "learning_rate": 4.4506710555709246e-05, "loss": 0.0292, "step": 55830 }, { "grad_norm": 0.2581210136413574, "learning_rate": 4.449027625213114e-05, "loss": 0.0269, "step": 55840 }, { "grad_norm": 0.4810841679573059, "learning_rate": 4.4473842551088135e-05, "loss": 0.0311, "step": 55850 }, { "grad_norm": 0.3645956516265869, "learning_rate": 4.445740945437738e-05, "loss": 0.0361, "step": 55860 }, { "grad_norm": 0.6146658062934875, "learning_rate": 4.444097696379602e-05, "loss": 0.0315, "step": 55870 }, { "grad_norm": 0.3010907471179962, "learning_rate": 4.442454508114106e-05, "loss": 0.0334, "step": 55880 }, { "grad_norm": 0.3264015316963196, "learning_rate": 4.4408113808209465e-05, "loss": 0.0279, "step": 55890 }, { "grad_norm": 0.3869449198246002, "learning_rate": 4.439168314679813e-05, "loss": 0.0421, "step": 55900 }, { "grad_norm": 0.5111698508262634, "learning_rate": 4.43752530987039e-05, "loss": 0.0266, "step": 55910 }, { "grad_norm": 0.5479125380516052, "learning_rate": 4.4358823665723515e-05, "loss": 0.0465, "step": 55920 }, { "grad_norm": 0.7108932137489319, "learning_rate": 4.434239484965371e-05, "loss": 0.0351, "step": 55930 }, { "grad_norm": 0.3001307249069214, "learning_rate": 4.4325966652291103e-05, "loss": 0.0233, "step": 55940 }, { "grad_norm": 0.37051233649253845, "learning_rate": 4.430953907543225e-05, "loss": 0.0256, "step": 55950 }, { "grad_norm": 0.2587210237979889, "learning_rate": 4.4293112120873645e-05, "loss": 0.0224, "step": 55960 }, { "grad_norm": 0.8669247031211853, "learning_rate": 4.427668579041172e-05, "loss": 0.0309, "step": 55970 }, { "grad_norm": 0.3420785367488861, "learning_rate": 4.4260260085842816e-05, "loss": 0.0288, "step": 55980 }, { "grad_norm": 0.7178967595100403, "learning_rate": 4.4243835008963256e-05, "loss": 0.0442, "step": 55990 }, { "grad_norm": 0.2685697674751282, "learning_rate": 4.422741056156925e-05, "loss": 0.0338, "step": 56000 }, { "grad_norm": 0.6757442951202393, "learning_rate": 4.421098674545693e-05, "loss": 0.0281, "step": 56010 }, { "grad_norm": 0.3593725562095642, "learning_rate": 4.419456356242241e-05, "loss": 0.0229, "step": 56020 }, { "grad_norm": 0.5727565288543701, "learning_rate": 4.4178141014261685e-05, "loss": 0.0384, "step": 56030 }, { "grad_norm": 0.5952348709106445, "learning_rate": 4.416171910277071e-05, "loss": 0.0322, "step": 56040 }, { "grad_norm": 0.34897327423095703, "learning_rate": 4.414529782974533e-05, "loss": 0.0284, "step": 56050 }, { "grad_norm": 0.7916595339775085, "learning_rate": 4.41288771969814e-05, "loss": 0.0325, "step": 56060 }, { "grad_norm": 0.4708816409111023, "learning_rate": 4.411245720627464e-05, "loss": 0.0386, "step": 56070 }, { "grad_norm": 0.27650532126426697, "learning_rate": 4.40960378594207e-05, "loss": 0.0323, "step": 56080 }, { "grad_norm": 0.4070393443107605, "learning_rate": 4.407961915821519e-05, "loss": 0.0243, "step": 56090 }, { "grad_norm": 0.517362117767334, "learning_rate": 4.406320110445362e-05, "loss": 0.0339, "step": 56100 }, { "grad_norm": 0.455394446849823, "learning_rate": 4.404678369993144e-05, "loss": 0.0275, "step": 56110 }, { "grad_norm": 0.4854705035686493, "learning_rate": 4.403036694644406e-05, "loss": 0.0347, "step": 56120 }, { "grad_norm": 0.4317023456096649, "learning_rate": 4.4013950845786764e-05, "loss": 0.0253, "step": 56130 }, { "grad_norm": 1.0721795558929443, "learning_rate": 4.399753539975482e-05, "loss": 0.0334, "step": 56140 }, { "grad_norm": 0.6817452311515808, "learning_rate": 4.398112061014337e-05, "loss": 0.033, "step": 56150 }, { "grad_norm": 0.324517160654068, "learning_rate": 4.396470647874753e-05, "loss": 0.0293, "step": 56160 }, { "grad_norm": 0.5252076983451843, "learning_rate": 4.394829300736229e-05, "loss": 0.037, "step": 56170 }, { "grad_norm": 0.5006980299949646, "learning_rate": 4.393188019778265e-05, "loss": 0.0367, "step": 56180 }, { "grad_norm": 0.39875197410583496, "learning_rate": 4.391546805180347e-05, "loss": 0.0318, "step": 56190 }, { "grad_norm": 0.3292177617549896, "learning_rate": 4.389905657121955e-05, "loss": 0.029, "step": 56200 }, { "grad_norm": 0.3833238184452057, "learning_rate": 4.3882645757825635e-05, "loss": 0.0315, "step": 56210 }, { "grad_norm": 0.645062267780304, "learning_rate": 4.386623561341637e-05, "loss": 0.0306, "step": 56220 }, { "grad_norm": 0.9738255739212036, "learning_rate": 4.384982613978637e-05, "loss": 0.036, "step": 56230 }, { "grad_norm": 0.6781898736953735, "learning_rate": 4.383341733873012e-05, "loss": 0.0339, "step": 56240 }, { "grad_norm": 0.5865494012832642, "learning_rate": 4.381700921204209e-05, "loss": 0.0348, "step": 56250 }, { "grad_norm": 0.3511054813861847, "learning_rate": 4.380060176151663e-05, "loss": 0.0314, "step": 56260 }, { "grad_norm": 0.43774986267089844, "learning_rate": 4.378419498894805e-05, "loss": 0.0291, "step": 56270 }, { "grad_norm": 0.7705450057983398, "learning_rate": 4.376778889613056e-05, "loss": 0.0329, "step": 56280 }, { "grad_norm": 0.5468838810920715, "learning_rate": 4.375138348485831e-05, "loss": 0.0296, "step": 56290 }, { "grad_norm": 0.6116770505905151, "learning_rate": 4.3734978756925346e-05, "loss": 0.066, "step": 56300 }, { "grad_norm": 0.420110821723938, "learning_rate": 4.3718574714125706e-05, "loss": 0.0321, "step": 56310 }, { "grad_norm": 0.5167644619941711, "learning_rate": 4.370217135825329e-05, "loss": 0.0339, "step": 56320 }, { "grad_norm": 0.758968710899353, "learning_rate": 4.368576869110194e-05, "loss": 0.0317, "step": 56330 }, { "grad_norm": 0.40259167551994324, "learning_rate": 4.366936671446544e-05, "loss": 0.0368, "step": 56340 }, { "grad_norm": 0.4588698446750641, "learning_rate": 4.3652965430137474e-05, "loss": 0.0464, "step": 56350 }, { "grad_norm": 1.4715826511383057, "learning_rate": 4.3636564839911646e-05, "loss": 0.0301, "step": 56360 }, { "grad_norm": 0.8461122512817383, "learning_rate": 4.3620164945581545e-05, "loss": 0.0472, "step": 56370 }, { "grad_norm": 0.39124763011932373, "learning_rate": 4.360376574894061e-05, "loss": 0.0538, "step": 56380 }, { "grad_norm": 0.4375603199005127, "learning_rate": 4.358736725178224e-05, "loss": 0.0382, "step": 56390 }, { "grad_norm": 0.5647441744804382, "learning_rate": 4.357096945589974e-05, "loss": 0.0387, "step": 56400 }, { "grad_norm": 0.4571570158004761, "learning_rate": 4.355457236308636e-05, "loss": 0.0327, "step": 56410 }, { "grad_norm": 0.36119964718818665, "learning_rate": 4.353817597513526e-05, "loss": 0.0325, "step": 56420 }, { "grad_norm": 0.6286057233810425, "learning_rate": 4.352178029383948e-05, "loss": 0.0325, "step": 56430 }, { "grad_norm": 0.3531110882759094, "learning_rate": 4.3505385320992105e-05, "loss": 0.031, "step": 56440 }, { "grad_norm": 0.42065608501434326, "learning_rate": 4.348899105838602e-05, "loss": 0.0318, "step": 56450 }, { "grad_norm": 0.2610779106616974, "learning_rate": 4.3472597507814087e-05, "loss": 0.0334, "step": 56460 }, { "grad_norm": 0.29916608333587646, "learning_rate": 4.3456204671069066e-05, "loss": 0.0291, "step": 56470 }, { "grad_norm": 0.4074668288230896, "learning_rate": 4.343981254994367e-05, "loss": 0.024, "step": 56480 }, { "grad_norm": 0.43044695258140564, "learning_rate": 4.34234211462305e-05, "loss": 0.038, "step": 56490 }, { "grad_norm": 1.009812593460083, "learning_rate": 4.340703046172213e-05, "loss": 0.0497, "step": 56500 }, { "grad_norm": 0.32046937942504883, "learning_rate": 4.339064049821097e-05, "loss": 0.0382, "step": 56510 }, { "grad_norm": 0.49769800901412964, "learning_rate": 4.3374251257489446e-05, "loss": 0.0355, "step": 56520 }, { "grad_norm": 0.322119802236557, "learning_rate": 4.335786274134984e-05, "loss": 0.0284, "step": 56530 }, { "grad_norm": 0.17914137244224548, "learning_rate": 4.334147495158438e-05, "loss": 0.03, "step": 56540 }, { "grad_norm": 0.42114338278770447, "learning_rate": 4.33250878899852e-05, "loss": 0.033, "step": 56550 }, { "grad_norm": 0.44115689396858215, "learning_rate": 4.330870155834437e-05, "loss": 0.0293, "step": 56560 }, { "grad_norm": 0.3835998773574829, "learning_rate": 4.3292315958453884e-05, "loss": 0.0288, "step": 56570 }, { "grad_norm": 0.24839462339878082, "learning_rate": 4.3275931092105646e-05, "loss": 0.0266, "step": 56580 }, { "grad_norm": 0.39805617928504944, "learning_rate": 4.325954696109147e-05, "loss": 0.0261, "step": 56590 }, { "grad_norm": 0.6157453060150146, "learning_rate": 4.32431635672031e-05, "loss": 0.0236, "step": 56600 }, { "grad_norm": 0.3273386061191559, "learning_rate": 4.3226780912232214e-05, "loss": 0.0283, "step": 56610 }, { "grad_norm": 0.4352662265300751, "learning_rate": 4.321039899797036e-05, "loss": 0.0315, "step": 56620 }, { "grad_norm": 1.0407651662826538, "learning_rate": 4.319401782620908e-05, "loss": 0.0327, "step": 56630 }, { "grad_norm": 0.4296764135360718, "learning_rate": 4.317763739873978e-05, "loss": 0.0352, "step": 56640 }, { "grad_norm": 0.3398081660270691, "learning_rate": 4.31612577173538e-05, "loss": 0.0341, "step": 56650 }, { "grad_norm": 0.25446027517318726, "learning_rate": 4.314487878384239e-05, "loss": 0.032, "step": 56660 }, { "grad_norm": 0.4605749249458313, "learning_rate": 4.3128500599996736e-05, "loss": 0.0302, "step": 56670 }, { "grad_norm": 0.3291815221309662, "learning_rate": 4.31121231676079e-05, "loss": 0.0321, "step": 56680 }, { "grad_norm": 0.5264405608177185, "learning_rate": 4.309574648846694e-05, "loss": 0.0261, "step": 56690 }, { "grad_norm": 0.4837823212146759, "learning_rate": 4.3079370564364755e-05, "loss": 0.023, "step": 56700 }, { "grad_norm": 0.2823816239833832, "learning_rate": 4.306299539709221e-05, "loss": 0.0222, "step": 56710 }, { "grad_norm": 0.3747287094593048, "learning_rate": 4.3046620988440046e-05, "loss": 0.0278, "step": 56720 }, { "grad_norm": 0.464180588722229, "learning_rate": 4.303024734019897e-05, "loss": 0.0351, "step": 56730 }, { "grad_norm": 0.34486010670661926, "learning_rate": 4.3013874454159565e-05, "loss": 0.0333, "step": 56740 }, { "grad_norm": 0.6022781133651733, "learning_rate": 4.299750233211233e-05, "loss": 0.0309, "step": 56750 }, { "grad_norm": 0.40956729650497437, "learning_rate": 4.2981130975847715e-05, "loss": 0.036, "step": 56760 }, { "grad_norm": 0.4054853916168213, "learning_rate": 4.2964760387156076e-05, "loss": 0.0302, "step": 56770 }, { "grad_norm": 0.30040499567985535, "learning_rate": 4.294839056782767e-05, "loss": 0.0366, "step": 56780 }, { "grad_norm": 0.7271663546562195, "learning_rate": 4.293202151965266e-05, "loss": 0.0313, "step": 56790 }, { "grad_norm": 0.6732029318809509, "learning_rate": 4.291565324442115e-05, "loss": 0.0261, "step": 56800 }, { "grad_norm": 0.515300989151001, "learning_rate": 4.2899285743923135e-05, "loss": 0.0244, "step": 56810 }, { "grad_norm": 0.700061023235321, "learning_rate": 4.288291901994857e-05, "loss": 0.0263, "step": 56820 }, { "grad_norm": 0.4074278771877289, "learning_rate": 4.286655307428729e-05, "loss": 0.0367, "step": 56830 }, { "grad_norm": 0.47207877039909363, "learning_rate": 4.285018790872903e-05, "loss": 0.0296, "step": 56840 }, { "grad_norm": 0.6320184469223022, "learning_rate": 4.283382352506347e-05, "loss": 0.0291, "step": 56850 }, { "grad_norm": 0.5238053202629089, "learning_rate": 4.2817459925080205e-05, "loss": 0.0301, "step": 56860 }, { "grad_norm": 0.4234921634197235, "learning_rate": 4.28010971105687e-05, "loss": 0.0242, "step": 56870 }, { "grad_norm": 0.30805879831314087, "learning_rate": 4.278473508331841e-05, "loss": 0.026, "step": 56880 }, { "grad_norm": 0.32887428998947144, "learning_rate": 4.276837384511864e-05, "loss": 0.0435, "step": 56890 }, { "grad_norm": 0.33355027437210083, "learning_rate": 4.275201339775864e-05, "loss": 0.0328, "step": 56900 }, { "grad_norm": 0.2129196673631668, "learning_rate": 4.273565374302756e-05, "loss": 0.0299, "step": 56910 }, { "grad_norm": 0.43309903144836426, "learning_rate": 4.2719294882714454e-05, "loss": 0.0314, "step": 56920 }, { "grad_norm": 0.394227534532547, "learning_rate": 4.270293681860831e-05, "loss": 0.0264, "step": 56930 }, { "grad_norm": 0.40957409143447876, "learning_rate": 4.268657955249802e-05, "loss": 0.0322, "step": 56940 }, { "grad_norm": 0.3499370515346527, "learning_rate": 4.267022308617241e-05, "loss": 0.0433, "step": 56950 }, { "grad_norm": 0.45608678460121155, "learning_rate": 4.265386742142018e-05, "loss": 0.0415, "step": 56960 }, { "grad_norm": 0.2838801145553589, "learning_rate": 4.263751256002996e-05, "loss": 0.0191, "step": 56970 }, { "grad_norm": 0.4914033114910126, "learning_rate": 4.26211585037903e-05, "loss": 0.031, "step": 56980 }, { "grad_norm": 0.5062592625617981, "learning_rate": 4.2604805254489656e-05, "loss": 0.0332, "step": 56990 }, { "grad_norm": 0.34647366404533386, "learning_rate": 4.258845281391638e-05, "loss": 0.0246, "step": 57000 }, { "grad_norm": 0.4502786099910736, "learning_rate": 4.257210118385876e-05, "loss": 0.0236, "step": 57010 }, { "grad_norm": 0.7312574982643127, "learning_rate": 4.2555750366104994e-05, "loss": 0.0382, "step": 57020 }, { "grad_norm": 0.43725332617759705, "learning_rate": 4.2539400362443184e-05, "loss": 0.0328, "step": 57030 }, { "grad_norm": 0.37640854716300964, "learning_rate": 4.252305117466133e-05, "loss": 0.0292, "step": 57040 }, { "grad_norm": 0.6387448310852051, "learning_rate": 4.2506702804547356e-05, "loss": 0.025, "step": 57050 }, { "grad_norm": 0.48950257897377014, "learning_rate": 4.249035525388907e-05, "loss": 0.0367, "step": 57060 }, { "grad_norm": 0.546932578086853, "learning_rate": 4.247400852447427e-05, "loss": 0.0236, "step": 57070 }, { "grad_norm": 0.2916882336139679, "learning_rate": 4.245766261809059e-05, "loss": 0.025, "step": 57080 }, { "grad_norm": 1.8284951448440552, "learning_rate": 4.244131753652558e-05, "loss": 0.0422, "step": 57090 }, { "grad_norm": 0.366610586643219, "learning_rate": 4.242497328156672e-05, "loss": 0.0326, "step": 57100 }, { "grad_norm": 0.4586455821990967, "learning_rate": 4.240862985500139e-05, "loss": 0.0341, "step": 57110 }, { "grad_norm": 0.37814098596572876, "learning_rate": 4.2392287258616894e-05, "loss": 0.029, "step": 57120 }, { "grad_norm": 0.39847004413604736, "learning_rate": 4.2375945494200396e-05, "loss": 0.0553, "step": 57130 }, { "grad_norm": 0.4351543188095093, "learning_rate": 4.2359604563539054e-05, "loss": 0.0353, "step": 57140 }, { "grad_norm": 0.34620070457458496, "learning_rate": 4.2343264468419885e-05, "loss": 0.0474, "step": 57150 }, { "grad_norm": 0.4106017053127289, "learning_rate": 4.232692521062979e-05, "loss": 0.0303, "step": 57160 }, { "grad_norm": 0.4351043105125427, "learning_rate": 4.231058679195562e-05, "loss": 0.0335, "step": 57170 }, { "grad_norm": 0.446392685174942, "learning_rate": 4.229424921418411e-05, "loss": 0.0363, "step": 57180 }, { "grad_norm": 0.4449487626552582, "learning_rate": 4.2277912479101916e-05, "loss": 0.0307, "step": 57190 }, { "grad_norm": 0.45945337414741516, "learning_rate": 4.226157658849562e-05, "loss": 0.0305, "step": 57200 }, { "grad_norm": 0.4263545274734497, "learning_rate": 4.2245241544151664e-05, "loss": 0.0409, "step": 57210 }, { "grad_norm": 0.5662491917610168, "learning_rate": 4.2228907347856425e-05, "loss": 0.0396, "step": 57220 }, { "grad_norm": 0.4924752712249756, "learning_rate": 4.2212574001396214e-05, "loss": 0.0341, "step": 57230 }, { "grad_norm": 0.4691201150417328, "learning_rate": 4.21962415065572e-05, "loss": 0.0326, "step": 57240 }, { "grad_norm": 0.7280324101448059, "learning_rate": 4.217990986512548e-05, "loss": 0.0319, "step": 57250 }, { "grad_norm": 0.31368085741996765, "learning_rate": 4.216357907888704e-05, "loss": 0.0248, "step": 57260 }, { "grad_norm": 0.5678796768188477, "learning_rate": 4.2147249149627824e-05, "loss": 0.0355, "step": 57270 }, { "grad_norm": 8.210208892822266, "learning_rate": 4.213092007913364e-05, "loss": 0.0351, "step": 57280 }, { "grad_norm": 0.4268556535243988, "learning_rate": 4.2114591869190205e-05, "loss": 0.0225, "step": 57290 }, { "grad_norm": 0.27300402522087097, "learning_rate": 4.209826452158315e-05, "loss": 0.0294, "step": 57300 }, { "grad_norm": 0.5542567372322083, "learning_rate": 4.2081938038098e-05, "loss": 0.032, "step": 57310 }, { "grad_norm": 0.7557511329650879, "learning_rate": 4.2065612420520185e-05, "loss": 0.0394, "step": 57320 }, { "grad_norm": 2.416050672531128, "learning_rate": 4.204928767063509e-05, "loss": 0.0339, "step": 57330 }, { "grad_norm": 0.6643273234367371, "learning_rate": 4.2032963790227944e-05, "loss": 0.0329, "step": 57340 }, { "grad_norm": 0.4660249650478363, "learning_rate": 4.2016640781083887e-05, "loss": 0.0342, "step": 57350 }, { "grad_norm": 0.44584840536117554, "learning_rate": 4.2000318644988004e-05, "loss": 0.037, "step": 57360 }, { "grad_norm": 0.3177841007709503, "learning_rate": 4.198399738372524e-05, "loss": 0.0322, "step": 57370 }, { "grad_norm": 0.3660814166069031, "learning_rate": 4.196767699908045e-05, "loss": 0.0302, "step": 57380 }, { "grad_norm": 0.33114391565322876, "learning_rate": 4.195135749283845e-05, "loss": 0.0372, "step": 57390 }, { "grad_norm": 0.5110266208648682, "learning_rate": 4.1935038866783894e-05, "loss": 0.0274, "step": 57400 }, { "grad_norm": 0.36202770471572876, "learning_rate": 4.1918721122701355e-05, "loss": 0.0294, "step": 57410 }, { "grad_norm": 0.33830001950263977, "learning_rate": 4.190240426237533e-05, "loss": 0.0288, "step": 57420 }, { "grad_norm": 0.49464184045791626, "learning_rate": 4.188608828759019e-05, "loss": 0.0263, "step": 57430 }, { "grad_norm": 0.9874659180641174, "learning_rate": 4.1869773200130246e-05, "loss": 0.0472, "step": 57440 }, { "grad_norm": 0.502974808216095, "learning_rate": 4.185345900177967e-05, "loss": 0.0297, "step": 57450 }, { "grad_norm": 1.4356189966201782, "learning_rate": 4.183714569432258e-05, "loss": 0.0629, "step": 57460 }, { "grad_norm": 0.6896640062332153, "learning_rate": 4.182083327954295e-05, "loss": 0.031, "step": 57470 }, { "grad_norm": 0.3672140836715698, "learning_rate": 4.1804521759224714e-05, "loss": 0.0317, "step": 57480 }, { "grad_norm": 0.48039379715919495, "learning_rate": 4.178821113515165e-05, "loss": 0.0307, "step": 57490 }, { "grad_norm": 0.4784248173236847, "learning_rate": 4.1771901409107476e-05, "loss": 0.029, "step": 57500 }, { "grad_norm": 0.40446802973747253, "learning_rate": 4.175559258287577e-05, "loss": 0.0602, "step": 57510 }, { "grad_norm": 0.45742812752723694, "learning_rate": 4.173928465824008e-05, "loss": 0.0353, "step": 57520 }, { "grad_norm": 0.5692225694656372, "learning_rate": 4.17229776369838e-05, "loss": 0.0324, "step": 57530 }, { "grad_norm": 0.5691127181053162, "learning_rate": 4.170667152089025e-05, "loss": 0.03, "step": 57540 }, { "grad_norm": 0.4513644576072693, "learning_rate": 4.1690366311742626e-05, "loss": 0.0278, "step": 57550 }, { "grad_norm": 0.32831794023513794, "learning_rate": 4.1674062011324055e-05, "loss": 0.0266, "step": 57560 }, { "grad_norm": 0.6369884610176086, "learning_rate": 4.165775862141752e-05, "loss": 0.0408, "step": 57570 }, { "grad_norm": 0.5282418727874756, "learning_rate": 4.164145614380598e-05, "loss": 0.0307, "step": 57580 }, { "grad_norm": 0.9514092803001404, "learning_rate": 4.162515458027223e-05, "loss": 0.0348, "step": 57590 }, { "grad_norm": 0.8998900651931763, "learning_rate": 4.160885393259899e-05, "loss": 0.023, "step": 57600 }, { "grad_norm": 0.49419689178466797, "learning_rate": 4.159255420256886e-05, "loss": 0.028, "step": 57610 }, { "grad_norm": 0.4013914465904236, "learning_rate": 4.157625539196436e-05, "loss": 0.024, "step": 57620 }, { "grad_norm": 0.47240716218948364, "learning_rate": 4.155995750256791e-05, "loss": 0.0388, "step": 57630 }, { "grad_norm": 0.22832991182804108, "learning_rate": 4.15436605361618e-05, "loss": 0.0409, "step": 57640 }, { "grad_norm": 0.3671833872795105, "learning_rate": 4.152736449452827e-05, "loss": 0.0285, "step": 57650 }, { "grad_norm": 0.40035951137542725, "learning_rate": 4.151106937944943e-05, "loss": 0.0261, "step": 57660 }, { "grad_norm": 0.8436069488525391, "learning_rate": 4.149477519270727e-05, "loss": 0.0273, "step": 57670 }, { "grad_norm": 0.5694953799247742, "learning_rate": 4.147848193608371e-05, "loss": 0.0317, "step": 57680 }, { "grad_norm": 0.6013928651809692, "learning_rate": 4.146218961136056e-05, "loss": 0.0305, "step": 57690 }, { "grad_norm": 0.7920355200767517, "learning_rate": 4.14458982203195e-05, "loss": 0.0391, "step": 57700 }, { "grad_norm": 0.32651573419570923, "learning_rate": 4.1429607764742165e-05, "loss": 0.0401, "step": 57710 }, { "grad_norm": 0.3598765432834625, "learning_rate": 4.141331824641003e-05, "loss": 0.0354, "step": 57720 }, { "grad_norm": 0.43789416551589966, "learning_rate": 4.139702966710452e-05, "loss": 0.0272, "step": 57730 }, { "grad_norm": 0.7364853620529175, "learning_rate": 4.13807420286069e-05, "loss": 0.0306, "step": 57740 }, { "grad_norm": 0.35938069224357605, "learning_rate": 4.136445533269837e-05, "loss": 0.0539, "step": 57750 }, { "grad_norm": 0.4105094373226166, "learning_rate": 4.134816958116003e-05, "loss": 0.0367, "step": 57760 }, { "grad_norm": 0.37423527240753174, "learning_rate": 4.133188477577283e-05, "loss": 0.0295, "step": 57770 }, { "grad_norm": 0.38107815384864807, "learning_rate": 4.131560091831771e-05, "loss": 0.0265, "step": 57780 }, { "grad_norm": 0.5532119870185852, "learning_rate": 4.1299318010575405e-05, "loss": 0.0366, "step": 57790 }, { "grad_norm": 0.4639858901500702, "learning_rate": 4.128303605432661e-05, "loss": 0.0228, "step": 57800 }, { "grad_norm": 0.551135778427124, "learning_rate": 4.126675505135188e-05, "loss": 0.0303, "step": 57810 }, { "grad_norm": 0.32495132088661194, "learning_rate": 4.125047500343169e-05, "loss": 0.0264, "step": 57820 }, { "grad_norm": 0.40327921509742737, "learning_rate": 4.1234195912346376e-05, "loss": 0.0237, "step": 57830 }, { "grad_norm": 0.6458893418312073, "learning_rate": 4.1217917779876235e-05, "loss": 0.0337, "step": 57840 }, { "grad_norm": 1.0305622816085815, "learning_rate": 4.120164060780141e-05, "loss": 0.0328, "step": 57850 }, { "grad_norm": 0.43563395738601685, "learning_rate": 4.1185364397901935e-05, "loss": 0.0243, "step": 57860 }, { "grad_norm": 0.5440017580986023, "learning_rate": 4.116908915195776e-05, "loss": 0.0252, "step": 57870 }, { "grad_norm": 0.5756111741065979, "learning_rate": 4.115281487174872e-05, "loss": 0.0268, "step": 57880 }, { "grad_norm": 0.38768649101257324, "learning_rate": 4.113654155905452e-05, "loss": 0.0291, "step": 57890 }, { "grad_norm": 0.5491419434547424, "learning_rate": 4.112026921565483e-05, "loss": 0.0308, "step": 57900 }, { "grad_norm": 0.5122949481010437, "learning_rate": 4.1103997843329147e-05, "loss": 0.0342, "step": 57910 }, { "grad_norm": 0.3636661171913147, "learning_rate": 4.108772744385689e-05, "loss": 0.029, "step": 57920 }, { "grad_norm": 0.5998072624206543, "learning_rate": 4.1071458019017346e-05, "loss": 0.0341, "step": 57930 }, { "grad_norm": 0.4416843354701996, "learning_rate": 4.1055189570589734e-05, "loss": 0.0293, "step": 57940 }, { "grad_norm": 0.3706431984901428, "learning_rate": 4.103892210035315e-05, "loss": 0.0279, "step": 57950 }, { "grad_norm": 0.39188697934150696, "learning_rate": 4.102265561008656e-05, "loss": 0.0271, "step": 57960 }, { "grad_norm": 0.4625032842159271, "learning_rate": 4.100639010156885e-05, "loss": 0.0259, "step": 57970 }, { "grad_norm": 0.605949878692627, "learning_rate": 4.099012557657882e-05, "loss": 0.0322, "step": 57980 }, { "grad_norm": 0.687527596950531, "learning_rate": 4.09738620368951e-05, "loss": 0.0305, "step": 57990 }, { "grad_norm": 0.5571247935295105, "learning_rate": 4.095759948429626e-05, "loss": 0.0258, "step": 58000 }, { "grad_norm": 0.5919150710105896, "learning_rate": 4.094133792056075e-05, "loss": 0.0346, "step": 58010 }, { "grad_norm": 0.3365679979324341, "learning_rate": 4.092507734746688e-05, "loss": 0.0246, "step": 58020 }, { "grad_norm": 0.31404632329940796, "learning_rate": 4.090881776679293e-05, "loss": 0.0254, "step": 58030 }, { "grad_norm": 1.4144538640975952, "learning_rate": 4.0892559180316994e-05, "loss": 0.0326, "step": 58040 }, { "grad_norm": 0.4670717120170593, "learning_rate": 4.08763015898171e-05, "loss": 0.0283, "step": 58050 }, { "grad_norm": 0.5308701992034912, "learning_rate": 4.0860044997071145e-05, "loss": 0.0222, "step": 58060 }, { "grad_norm": 0.4532700479030609, "learning_rate": 4.084378940385693e-05, "loss": 0.0292, "step": 58070 }, { "grad_norm": 0.9132925271987915, "learning_rate": 4.082753481195211e-05, "loss": 0.0387, "step": 58080 }, { "grad_norm": 0.507443904876709, "learning_rate": 4.0811281223134324e-05, "loss": 0.0336, "step": 58090 }, { "grad_norm": 0.3680807054042816, "learning_rate": 4.0795028639181e-05, "loss": 0.0289, "step": 58100 }, { "grad_norm": 0.43899938464164734, "learning_rate": 4.077877706186951e-05, "loss": 0.0287, "step": 58110 }, { "grad_norm": 0.5631065964698792, "learning_rate": 4.076252649297709e-05, "loss": 0.0341, "step": 58120 }, { "grad_norm": 0.6206502318382263, "learning_rate": 4.07462769342809e-05, "loss": 0.0307, "step": 58130 }, { "grad_norm": 0.6389527916908264, "learning_rate": 4.073002838755793e-05, "loss": 0.0313, "step": 58140 }, { "grad_norm": 0.49845001101493835, "learning_rate": 4.0713780854585115e-05, "loss": 0.0245, "step": 58150 }, { "grad_norm": 0.4254895746707916, "learning_rate": 4.069753433713927e-05, "loss": 0.0372, "step": 58160 }, { "grad_norm": 0.614290714263916, "learning_rate": 4.068128883699709e-05, "loss": 0.0347, "step": 58170 }, { "grad_norm": 0.4686138331890106, "learning_rate": 4.0665044355935155e-05, "loss": 0.0269, "step": 58180 }, { "grad_norm": 0.8050917983055115, "learning_rate": 4.064880089572993e-05, "loss": 0.0292, "step": 58190 }, { "grad_norm": 3.1370034217834473, "learning_rate": 4.063255845815779e-05, "loss": 0.0328, "step": 58200 }, { "grad_norm": 0.5412978529930115, "learning_rate": 4.0616317044994945e-05, "loss": 0.0301, "step": 58210 }, { "grad_norm": 0.3329732418060303, "learning_rate": 4.0600076658017585e-05, "loss": 0.0277, "step": 58220 }, { "grad_norm": 0.7578065395355225, "learning_rate": 4.0583837299001706e-05, "loss": 0.0258, "step": 58230 }, { "grad_norm": 0.36733880639076233, "learning_rate": 4.056759896972322e-05, "loss": 0.0296, "step": 58240 }, { "grad_norm": 0.5372344255447388, "learning_rate": 4.055136167195793e-05, "loss": 0.0216, "step": 58250 }, { "grad_norm": 0.36636921763420105, "learning_rate": 4.053512540748152e-05, "loss": 0.0242, "step": 58260 }, { "grad_norm": 0.4373590350151062, "learning_rate": 4.051889017806955e-05, "loss": 0.0337, "step": 58270 }, { "grad_norm": 0.8011927604675293, "learning_rate": 4.0502655985497516e-05, "loss": 0.0332, "step": 58280 }, { "grad_norm": 0.7541471719741821, "learning_rate": 4.048642283154074e-05, "loss": 0.0303, "step": 58290 }, { "grad_norm": 0.4648355543613434, "learning_rate": 4.0470190717974464e-05, "loss": 0.0273, "step": 58300 }, { "grad_norm": 0.5598299503326416, "learning_rate": 4.045395964657381e-05, "loss": 0.0412, "step": 58310 }, { "grad_norm": 0.3987797796726227, "learning_rate": 4.043772961911376e-05, "loss": 0.0352, "step": 58320 }, { "grad_norm": 0.39154553413391113, "learning_rate": 4.042150063736924e-05, "loss": 0.0354, "step": 58330 }, { "grad_norm": 0.4514407813549042, "learning_rate": 4.0405272703114976e-05, "loss": 0.045, "step": 58340 }, { "grad_norm": 0.6605947017669678, "learning_rate": 4.038904581812569e-05, "loss": 0.0352, "step": 58350 }, { "grad_norm": 0.9587211012840271, "learning_rate": 4.03728199841759e-05, "loss": 0.0317, "step": 58360 }, { "grad_norm": 0.4480747878551483, "learning_rate": 4.0356595203040044e-05, "loss": 0.0247, "step": 58370 }, { "grad_norm": 0.302949458360672, "learning_rate": 4.0340371476492426e-05, "loss": 0.03, "step": 58380 }, { "grad_norm": 0.4715108871459961, "learning_rate": 4.032414880630727e-05, "loss": 0.0241, "step": 58390 }, { "grad_norm": 0.28471049666404724, "learning_rate": 4.0307927194258646e-05, "loss": 0.0263, "step": 58400 }, { "grad_norm": 0.6315518021583557, "learning_rate": 4.029170664212054e-05, "loss": 0.0284, "step": 58410 }, { "grad_norm": 0.6439913511276245, "learning_rate": 4.0275487151666794e-05, "loss": 0.0264, "step": 58420 }, { "grad_norm": 0.31446799635887146, "learning_rate": 4.025926872467116e-05, "loss": 0.0416, "step": 58430 }, { "grad_norm": 0.8247591853141785, "learning_rate": 4.024305136290725e-05, "loss": 0.035, "step": 58440 }, { "grad_norm": 0.3131505250930786, "learning_rate": 4.022683506814858e-05, "loss": 0.0268, "step": 58450 }, { "grad_norm": 0.2861337661743164, "learning_rate": 4.021061984216853e-05, "loss": 0.0288, "step": 58460 }, { "grad_norm": 0.3219083249568939, "learning_rate": 4.019440568674035e-05, "loss": 0.0552, "step": 58470 }, { "grad_norm": 0.3150891661643982, "learning_rate": 4.017819260363724e-05, "loss": 0.0276, "step": 58480 }, { "grad_norm": 0.2813694477081299, "learning_rate": 4.016198059463223e-05, "loss": 0.0229, "step": 58490 }, { "grad_norm": 0.7676397562026978, "learning_rate": 4.0145769661498233e-05, "loss": 0.0258, "step": 58500 }, { "grad_norm": 0.4657648205757141, "learning_rate": 4.012955980600804e-05, "loss": 0.0322, "step": 58510 }, { "grad_norm": 1.0972257852554321, "learning_rate": 4.0113351029934355e-05, "loss": 0.0333, "step": 58520 }, { "grad_norm": 0.5117760896682739, "learning_rate": 4.0097143335049714e-05, "loss": 0.0349, "step": 58530 }, { "grad_norm": 0.2894780933856964, "learning_rate": 4.0080936723126606e-05, "loss": 0.0288, "step": 58540 }, { "grad_norm": 0.3889596462249756, "learning_rate": 4.006473119593735e-05, "loss": 0.0212, "step": 58550 }, { "grad_norm": 0.4849584102630615, "learning_rate": 4.004852675525415e-05, "loss": 0.0253, "step": 58560 }, { "grad_norm": 0.46187150478363037, "learning_rate": 4.00323234028491e-05, "loss": 0.0331, "step": 58570 }, { "grad_norm": 0.5726417899131775, "learning_rate": 4.0016121140494186e-05, "loss": 0.0335, "step": 58580 }, { "grad_norm": 0.4335012137889862, "learning_rate": 3.999991996996122e-05, "loss": 0.0367, "step": 58590 }, { "grad_norm": 0.3294748067855835, "learning_rate": 3.9983719893022e-05, "loss": 0.0231, "step": 58600 }, { "grad_norm": 0.392392635345459, "learning_rate": 3.996752091144812e-05, "loss": 0.0368, "step": 58610 }, { "grad_norm": 0.3483487665653229, "learning_rate": 3.9951323027011055e-05, "loss": 0.041, "step": 58620 }, { "grad_norm": 0.368937611579895, "learning_rate": 3.993512624148219e-05, "loss": 0.0319, "step": 58630 }, { "grad_norm": 0.2679797410964966, "learning_rate": 3.99189305566328e-05, "loss": 0.0306, "step": 58640 }, { "grad_norm": 0.73670494556427, "learning_rate": 3.990273597423401e-05, "loss": 0.0263, "step": 58650 }, { "grad_norm": 0.5373960733413696, "learning_rate": 3.988654249605681e-05, "loss": 0.0332, "step": 58660 }, { "grad_norm": 0.28850144147872925, "learning_rate": 3.987035012387213e-05, "loss": 0.0313, "step": 58670 }, { "grad_norm": 0.4609701633453369, "learning_rate": 3.9854158859450735e-05, "loss": 0.0222, "step": 58680 }, { "grad_norm": 0.40640565752983093, "learning_rate": 3.983796870456328e-05, "loss": 0.0283, "step": 58690 }, { "grad_norm": 0.4598390460014343, "learning_rate": 3.982177966098028e-05, "loss": 0.0303, "step": 58700 }, { "grad_norm": 0.4736950099468231, "learning_rate": 3.980559173047216e-05, "loss": 0.0304, "step": 58710 }, { "grad_norm": 0.27287766337394714, "learning_rate": 3.9789404914809176e-05, "loss": 0.0315, "step": 58720 }, { "grad_norm": 0.4867401123046875, "learning_rate": 3.9773219215761544e-05, "loss": 0.0326, "step": 58730 }, { "grad_norm": 0.27106261253356934, "learning_rate": 3.975703463509928e-05, "loss": 0.0317, "step": 58740 }, { "grad_norm": 0.537505030632019, "learning_rate": 3.974085117459231e-05, "loss": 0.02, "step": 58750 }, { "grad_norm": 0.40335723757743835, "learning_rate": 3.972466883601043e-05, "loss": 0.0309, "step": 58760 }, { "grad_norm": 0.32499635219573975, "learning_rate": 3.9708487621123316e-05, "loss": 0.0236, "step": 58770 }, { "grad_norm": 0.44005823135375977, "learning_rate": 3.969230753170051e-05, "loss": 0.0208, "step": 58780 }, { "grad_norm": 0.4775083363056183, "learning_rate": 3.967612856951146e-05, "loss": 0.046, "step": 58790 }, { "grad_norm": 0.44466450810432434, "learning_rate": 3.9659950736325476e-05, "loss": 0.0253, "step": 58800 }, { "grad_norm": 0.504437267780304, "learning_rate": 3.964377403391174e-05, "loss": 0.0307, "step": 58810 }, { "grad_norm": 0.6843222975730896, "learning_rate": 3.9627598464039303e-05, "loss": 0.0335, "step": 58820 }, { "grad_norm": 0.529901921749115, "learning_rate": 3.9611424028477096e-05, "loss": 0.0253, "step": 58830 }, { "grad_norm": 0.4853459894657135, "learning_rate": 3.959525072899394e-05, "loss": 0.0271, "step": 58840 }, { "grad_norm": 0.3799201548099518, "learning_rate": 3.9579078567358506e-05, "loss": 0.0224, "step": 58850 }, { "grad_norm": 0.4142570197582245, "learning_rate": 3.956290754533939e-05, "loss": 0.0327, "step": 58860 }, { "grad_norm": 0.44886699318885803, "learning_rate": 3.954673766470501e-05, "loss": 0.0247, "step": 58870 }, { "grad_norm": 0.5459022521972656, "learning_rate": 3.9530568927223674e-05, "loss": 0.0324, "step": 58880 }, { "grad_norm": 0.42415204644203186, "learning_rate": 3.95144013346636e-05, "loss": 0.0407, "step": 58890 }, { "grad_norm": 0.2498513013124466, "learning_rate": 3.949823488879282e-05, "loss": 0.0255, "step": 58900 }, { "grad_norm": 0.7747823596000671, "learning_rate": 3.9482069591379276e-05, "loss": 0.0283, "step": 58910 }, { "grad_norm": 0.30269092321395874, "learning_rate": 3.9465905444190784e-05, "loss": 0.0214, "step": 58920 }, { "grad_norm": 0.3608746826648712, "learning_rate": 3.944974244899505e-05, "loss": 0.023, "step": 58930 }, { "grad_norm": 0.3171287477016449, "learning_rate": 3.9433580607559616e-05, "loss": 0.0303, "step": 58940 }, { "grad_norm": 0.3578379154205322, "learning_rate": 3.941741992165193e-05, "loss": 0.0515, "step": 58950 }, { "grad_norm": 0.4869542121887207, "learning_rate": 3.940126039303928e-05, "loss": 0.0214, "step": 58960 }, { "grad_norm": 0.4997763931751251, "learning_rate": 3.938510202348886e-05, "loss": 0.0316, "step": 58970 }, { "grad_norm": 0.5390943288803101, "learning_rate": 3.93689448147677e-05, "loss": 0.0287, "step": 58980 }, { "grad_norm": 0.6590879559516907, "learning_rate": 3.935278876864278e-05, "loss": 0.037, "step": 58990 }, { "grad_norm": 0.76139235496521, "learning_rate": 3.933663388688087e-05, "loss": 0.03, "step": 59000 }, { "grad_norm": 0.36661699414253235, "learning_rate": 3.9320480171248645e-05, "loss": 0.0338, "step": 59010 }, { "grad_norm": 0.3982982635498047, "learning_rate": 3.930432762351265e-05, "loss": 0.0219, "step": 59020 }, { "grad_norm": 1.1834790706634521, "learning_rate": 3.928817624543931e-05, "loss": 0.032, "step": 59030 }, { "grad_norm": 0.3194866478443146, "learning_rate": 3.9272026038794885e-05, "loss": 0.0254, "step": 59040 }, { "grad_norm": 0.5134322643280029, "learning_rate": 3.9255877005345586e-05, "loss": 0.0338, "step": 59050 }, { "grad_norm": 0.426866739988327, "learning_rate": 3.923972914685743e-05, "loss": 0.0318, "step": 59060 }, { "grad_norm": 0.48156246542930603, "learning_rate": 3.922358246509631e-05, "loss": 0.0227, "step": 59070 }, { "grad_norm": 0.3090980648994446, "learning_rate": 3.920743696182801e-05, "loss": 0.034, "step": 59080 }, { "grad_norm": 0.3349304497241974, "learning_rate": 3.9191292638818164e-05, "loss": 0.0255, "step": 59090 }, { "grad_norm": 2.776897668838501, "learning_rate": 3.9175149497832316e-05, "loss": 0.0265, "step": 59100 }, { "grad_norm": 0.32743266224861145, "learning_rate": 3.915900754063584e-05, "loss": 0.0263, "step": 59110 }, { "grad_norm": 0.3262733221054077, "learning_rate": 3.9142866768994014e-05, "loss": 0.0278, "step": 59120 }, { "grad_norm": 0.5474258065223694, "learning_rate": 3.912672718467193e-05, "loss": 0.0266, "step": 59130 }, { "grad_norm": 0.4629231095314026, "learning_rate": 3.911058878943463e-05, "loss": 0.0328, "step": 59140 }, { "grad_norm": 1.99441659450531, "learning_rate": 3.9094451585046974e-05, "loss": 0.0386, "step": 59150 }, { "grad_norm": 0.325487345457077, "learning_rate": 3.90783155732737e-05, "loss": 0.0326, "step": 59160 }, { "grad_norm": 0.37532058358192444, "learning_rate": 3.9062180755879373e-05, "loss": 0.0249, "step": 59170 }, { "grad_norm": 0.4202859699726105, "learning_rate": 3.904604713462855e-05, "loss": 0.038, "step": 59180 }, { "grad_norm": 0.8206967115402222, "learning_rate": 3.902991471128554e-05, "loss": 0.0301, "step": 59190 }, { "grad_norm": 0.28966182470321655, "learning_rate": 3.901378348761456e-05, "loss": 0.0228, "step": 59200 }, { "grad_norm": 0.42539703845977783, "learning_rate": 3.899765346537971e-05, "loss": 0.0285, "step": 59210 }, { "grad_norm": 0.3280603885650635, "learning_rate": 3.898152464634493e-05, "loss": 0.0214, "step": 59220 }, { "grad_norm": 0.3943968713283539, "learning_rate": 3.896539703227402e-05, "loss": 0.0229, "step": 59230 }, { "grad_norm": 0.3319185972213745, "learning_rate": 3.894927062493073e-05, "loss": 0.0332, "step": 59240 }, { "grad_norm": 0.9774654507637024, "learning_rate": 3.893314542607859e-05, "loss": 0.0376, "step": 59250 }, { "grad_norm": 0.38259023427963257, "learning_rate": 3.891702143748103e-05, "loss": 0.0279, "step": 59260 }, { "grad_norm": 0.6280010938644409, "learning_rate": 3.890089866090133e-05, "loss": 0.0231, "step": 59270 }, { "grad_norm": 0.6632233262062073, "learning_rate": 3.888477709810267e-05, "loss": 0.0243, "step": 59280 }, { "grad_norm": 0.6730365753173828, "learning_rate": 3.886865675084806e-05, "loss": 0.031, "step": 59290 }, { "grad_norm": 0.42536449432373047, "learning_rate": 3.885253762090043e-05, "loss": 0.0283, "step": 59300 }, { "grad_norm": 1.6083348989486694, "learning_rate": 3.8836419710022534e-05, "loss": 0.0248, "step": 59310 }, { "grad_norm": 0.4965762794017792, "learning_rate": 3.882030301997698e-05, "loss": 0.036, "step": 59320 }, { "grad_norm": 0.6163048148155212, "learning_rate": 3.880418755252629e-05, "loss": 0.0402, "step": 59330 }, { "grad_norm": 0.5649992227554321, "learning_rate": 3.87880733094328e-05, "loss": 0.0419, "step": 59340 }, { "grad_norm": 0.30007871985435486, "learning_rate": 3.877196029245877e-05, "loss": 0.0382, "step": 59350 }, { "grad_norm": 0.3784453570842743, "learning_rate": 3.875584850336627e-05, "loss": 0.0313, "step": 59360 }, { "grad_norm": 0.5303940176963806, "learning_rate": 3.873973794391728e-05, "loss": 0.0257, "step": 59370 }, { "grad_norm": 0.48657214641571045, "learning_rate": 3.872362861587361e-05, "loss": 0.0368, "step": 59380 }, { "grad_norm": 0.6678881049156189, "learning_rate": 3.870752052099698e-05, "loss": 0.0233, "step": 59390 }, { "grad_norm": 0.39845871925354004, "learning_rate": 3.8691413661048916e-05, "loss": 0.032, "step": 59400 }, { "grad_norm": 0.42992499470710754, "learning_rate": 3.867530803779085e-05, "loss": 0.0263, "step": 59410 }, { "grad_norm": 0.4373185634613037, "learning_rate": 3.8659203652984054e-05, "loss": 0.0395, "step": 59420 }, { "grad_norm": 0.5941739082336426, "learning_rate": 3.864310050838972e-05, "loss": 0.0302, "step": 59430 }, { "grad_norm": 0.3626922369003296, "learning_rate": 3.8626998605768835e-05, "loss": 0.037, "step": 59440 }, { "grad_norm": 0.44218623638153076, "learning_rate": 3.8610897946882294e-05, "loss": 0.0284, "step": 59450 }, { "grad_norm": 0.2804527282714844, "learning_rate": 3.859479853349082e-05, "loss": 0.0458, "step": 59460 }, { "grad_norm": 0.8189603686332703, "learning_rate": 3.857870036735503e-05, "loss": 0.0162, "step": 59470 }, { "grad_norm": 0.5541741251945496, "learning_rate": 3.856260345023538e-05, "loss": 0.0318, "step": 59480 }, { "grad_norm": 0.7741492986679077, "learning_rate": 3.854650778389224e-05, "loss": 0.0322, "step": 59490 }, { "grad_norm": 0.40665289759635925, "learning_rate": 3.8530413370085786e-05, "loss": 0.0479, "step": 59500 }, { "grad_norm": 0.41275566816329956, "learning_rate": 3.851432021057608e-05, "loss": 0.0266, "step": 59510 }, { "grad_norm": 0.5070011019706726, "learning_rate": 3.849822830712306e-05, "loss": 0.0298, "step": 59520 }, { "grad_norm": 0.44299450516700745, "learning_rate": 3.848213766148649e-05, "loss": 0.0313, "step": 59530 }, { "grad_norm": 0.771239161491394, "learning_rate": 3.846604827542602e-05, "loss": 0.0352, "step": 59540 }, { "grad_norm": 0.6699628233909607, "learning_rate": 3.844996015070117e-05, "loss": 0.0345, "step": 59550 }, { "grad_norm": 0.3070719242095947, "learning_rate": 3.843387328907132e-05, "loss": 0.036, "step": 59560 }, { "grad_norm": 0.3568853437900543, "learning_rate": 3.8417787692295696e-05, "loss": 0.025, "step": 59570 }, { "grad_norm": 0.3706744909286499, "learning_rate": 3.84017033621334e-05, "loss": 0.0172, "step": 59580 }, { "grad_norm": 0.45525652170181274, "learning_rate": 3.838562030034338e-05, "loss": 0.0244, "step": 59590 }, { "grad_norm": 0.35901734232902527, "learning_rate": 3.836953850868447e-05, "loss": 0.0243, "step": 59600 }, { "grad_norm": 0.2852036952972412, "learning_rate": 3.835345798891533e-05, "loss": 0.0261, "step": 59610 }, { "grad_norm": 0.5596430897712708, "learning_rate": 3.833737874279452e-05, "loss": 0.0307, "step": 59620 }, { "grad_norm": 0.7991923689842224, "learning_rate": 3.8321300772080433e-05, "loss": 0.0255, "step": 59630 }, { "grad_norm": 0.8829084634780884, "learning_rate": 3.8305224078531345e-05, "loss": 0.0283, "step": 59640 }, { "grad_norm": 0.5055646896362305, "learning_rate": 3.828914866390537e-05, "loss": 0.0415, "step": 59650 }, { "grad_norm": 0.75423264503479, "learning_rate": 3.827307452996048e-05, "loss": 0.0332, "step": 59660 }, { "grad_norm": 0.37501636147499084, "learning_rate": 3.825700167845454e-05, "loss": 0.0473, "step": 59670 }, { "grad_norm": 0.6978443264961243, "learning_rate": 3.8240930111145214e-05, "loss": 0.0276, "step": 59680 }, { "grad_norm": 0.3371143639087677, "learning_rate": 3.8224859829790116e-05, "loss": 0.0445, "step": 59690 }, { "grad_norm": 0.5714054107666016, "learning_rate": 3.820879083614664e-05, "loss": 0.0303, "step": 59700 }, { "grad_norm": 0.59577876329422, "learning_rate": 3.819272313197208e-05, "loss": 0.0306, "step": 59710 }, { "grad_norm": 0.5762626528739929, "learning_rate": 3.817665671902355e-05, "loss": 0.0312, "step": 59720 }, { "grad_norm": 0.34104153513908386, "learning_rate": 3.816059159905808e-05, "loss": 0.03, "step": 59730 }, { "grad_norm": 0.39488187432289124, "learning_rate": 3.8144527773832476e-05, "loss": 0.0277, "step": 59740 }, { "grad_norm": 0.2863599956035614, "learning_rate": 3.812846524510352e-05, "loss": 0.0344, "step": 59750 }, { "grad_norm": 0.3446182608604431, "learning_rate": 3.811240401462775e-05, "loss": 0.0315, "step": 59760 }, { "grad_norm": 0.3927375376224518, "learning_rate": 3.809634408416162e-05, "loss": 0.0285, "step": 59770 }, { "grad_norm": 0.4885459840297699, "learning_rate": 3.8080285455461394e-05, "loss": 0.0212, "step": 59780 }, { "grad_norm": 0.401644229888916, "learning_rate": 3.806422813028323e-05, "loss": 0.0573, "step": 59790 }, { "grad_norm": 0.6875494718551636, "learning_rate": 3.8048172110383114e-05, "loss": 0.0337, "step": 59800 }, { "grad_norm": 0.24469026923179626, "learning_rate": 3.8032117397516944e-05, "loss": 0.0239, "step": 59810 }, { "grad_norm": 0.5768913626670837, "learning_rate": 3.801606399344042e-05, "loss": 0.0327, "step": 59820 }, { "grad_norm": 0.4294346272945404, "learning_rate": 3.800001189990913e-05, "loss": 0.0335, "step": 59830 }, { "grad_norm": 0.3973678648471832, "learning_rate": 3.798396111867847e-05, "loss": 0.0232, "step": 59840 }, { "grad_norm": 0.31387922167778015, "learning_rate": 3.796791165150378e-05, "loss": 0.0266, "step": 59850 }, { "grad_norm": 0.2732102572917938, "learning_rate": 3.795186350014018e-05, "loss": 0.0397, "step": 59860 }, { "grad_norm": 0.4043053388595581, "learning_rate": 3.793581666634266e-05, "loss": 0.031, "step": 59870 }, { "grad_norm": 0.47521209716796875, "learning_rate": 3.791977115186609e-05, "loss": 0.0354, "step": 59880 }, { "grad_norm": 0.47201868891716003, "learning_rate": 3.79037269584652e-05, "loss": 0.0299, "step": 59890 }, { "grad_norm": 0.7672237157821655, "learning_rate": 3.7887684087894545e-05, "loss": 0.0247, "step": 59900 }, { "grad_norm": 0.35539644956588745, "learning_rate": 3.787164254190854e-05, "loss": 0.0378, "step": 59910 }, { "grad_norm": 0.3540664613246918, "learning_rate": 3.7855602322261473e-05, "loss": 0.0251, "step": 59920 }, { "grad_norm": 0.28359121084213257, "learning_rate": 3.783956343070746e-05, "loss": 0.0291, "step": 59930 }, { "grad_norm": 0.467313677072525, "learning_rate": 3.7823525869000526e-05, "loss": 0.0274, "step": 59940 }, { "grad_norm": 0.5705638527870178, "learning_rate": 3.78074896388945e-05, "loss": 0.0341, "step": 59950 }, { "grad_norm": 0.608953595161438, "learning_rate": 3.7791454742143075e-05, "loss": 0.0353, "step": 59960 }, { "grad_norm": 0.5342438817024231, "learning_rate": 3.7775421180499795e-05, "loss": 0.0258, "step": 59970 }, { "grad_norm": 0.656065821647644, "learning_rate": 3.7759388955718086e-05, "loss": 0.0323, "step": 59980 }, { "grad_norm": 0.443318247795105, "learning_rate": 3.774335806955117e-05, "loss": 0.0268, "step": 59990 }, { "grad_norm": 0.35304537415504456, "learning_rate": 3.7727328523752215e-05, "loss": 0.0254, "step": 60000 }, { "grad_norm": 0.42252278327941895, "learning_rate": 3.771130032007415e-05, "loss": 0.0245, "step": 60010 }, { "grad_norm": 0.49290260672569275, "learning_rate": 3.769527346026982e-05, "loss": 0.0396, "step": 60020 }, { "grad_norm": 0.3219473361968994, "learning_rate": 3.767924794609188e-05, "loss": 0.0199, "step": 60030 }, { "grad_norm": 0.7266058921813965, "learning_rate": 3.7663223779292854e-05, "loss": 0.0298, "step": 60040 }, { "grad_norm": 0.6468527913093567, "learning_rate": 3.764720096162512e-05, "loss": 0.034, "step": 60050 }, { "grad_norm": 0.4109663963317871, "learning_rate": 3.763117949484092e-05, "loss": 0.0189, "step": 60060 }, { "grad_norm": 0.4169881045818329, "learning_rate": 3.7615159380692344e-05, "loss": 0.0259, "step": 60070 }, { "grad_norm": 0.569542646408081, "learning_rate": 3.759914062093132e-05, "loss": 0.0305, "step": 60080 }, { "grad_norm": 0.9161638021469116, "learning_rate": 3.7583123217309615e-05, "loss": 0.0532, "step": 60090 }, { "grad_norm": 0.3082251250743866, "learning_rate": 3.7567107171578904e-05, "loss": 0.0276, "step": 60100 }, { "grad_norm": 0.3309977650642395, "learning_rate": 3.755109248549066e-05, "loss": 0.0457, "step": 60110 }, { "grad_norm": 0.3937424123287201, "learning_rate": 3.75350791607962e-05, "loss": 0.0218, "step": 60120 }, { "grad_norm": 0.3137868046760559, "learning_rate": 3.751906719924676e-05, "loss": 0.0318, "step": 60130 }, { "grad_norm": 0.4123350977897644, "learning_rate": 3.750305660259337e-05, "loss": 0.026, "step": 60140 }, { "grad_norm": 0.7060155272483826, "learning_rate": 3.748704737258693e-05, "loss": 0.0361, "step": 60150 }, { "grad_norm": 0.19530853629112244, "learning_rate": 3.747103951097816e-05, "loss": 0.0308, "step": 60160 }, { "grad_norm": 0.42051807045936584, "learning_rate": 3.745503301951768e-05, "loss": 0.0199, "step": 60170 }, { "grad_norm": 0.33203890919685364, "learning_rate": 3.743902789995592e-05, "loss": 0.029, "step": 60180 }, { "grad_norm": 0.5462640523910522, "learning_rate": 3.7423024154043183e-05, "loss": 0.0239, "step": 60190 }, { "grad_norm": 0.41501158475875854, "learning_rate": 3.740702178352963e-05, "loss": 0.0302, "step": 60200 }, { "grad_norm": 0.4875243902206421, "learning_rate": 3.739102079016523e-05, "loss": 0.0359, "step": 60210 }, { "grad_norm": 0.8108805418014526, "learning_rate": 3.7375021175699846e-05, "loss": 0.0354, "step": 60220 }, { "grad_norm": 0.4001465439796448, "learning_rate": 3.735902294188317e-05, "loss": 0.0379, "step": 60230 }, { "grad_norm": 0.45787835121154785, "learning_rate": 3.734302609046473e-05, "loss": 0.0263, "step": 60240 }, { "grad_norm": 0.2313113510608673, "learning_rate": 3.732703062319391e-05, "loss": 0.0274, "step": 60250 }, { "grad_norm": 0.8778553605079651, "learning_rate": 3.731103654181999e-05, "loss": 0.0316, "step": 60260 }, { "grad_norm": 0.4288666844367981, "learning_rate": 3.729504384809203e-05, "loss": 0.0504, "step": 60270 }, { "grad_norm": 0.5842083096504211, "learning_rate": 3.727905254375898e-05, "loss": 0.0336, "step": 60280 }, { "grad_norm": 0.4654477834701538, "learning_rate": 3.7263062630569614e-05, "loss": 0.0362, "step": 60290 }, { "grad_norm": 0.5964536070823669, "learning_rate": 3.724707411027256e-05, "loss": 0.0256, "step": 60300 }, { "grad_norm": 0.41780421137809753, "learning_rate": 3.723108698461631e-05, "loss": 0.0283, "step": 60310 }, { "grad_norm": 0.622950553894043, "learning_rate": 3.72151012553492e-05, "loss": 0.0252, "step": 60320 }, { "grad_norm": 0.4227463901042938, "learning_rate": 3.719911692421939e-05, "loss": 0.0295, "step": 60330 }, { "grad_norm": 0.4423370063304901, "learning_rate": 3.71831339929749e-05, "loss": 0.0321, "step": 60340 }, { "grad_norm": 0.38398635387420654, "learning_rate": 3.716715246336362e-05, "loss": 0.0296, "step": 60350 }, { "grad_norm": 0.3881632089614868, "learning_rate": 3.715117233713324e-05, "loss": 0.0259, "step": 60360 }, { "grad_norm": 0.6212233304977417, "learning_rate": 3.713519361603135e-05, "loss": 0.0262, "step": 60370 }, { "grad_norm": 0.5346513390541077, "learning_rate": 3.711921630180532e-05, "loss": 0.0309, "step": 60380 }, { "grad_norm": 0.41349726915359497, "learning_rate": 3.710324039620245e-05, "loss": 0.0277, "step": 60390 }, { "grad_norm": 0.33510687947273254, "learning_rate": 3.708726590096982e-05, "loss": 0.0316, "step": 60400 }, { "grad_norm": 0.5343237519264221, "learning_rate": 3.707129281785437e-05, "loss": 0.0292, "step": 60410 }, { "grad_norm": 0.5555601716041565, "learning_rate": 3.705532114860291e-05, "loss": 0.0344, "step": 60420 }, { "grad_norm": 0.6355599761009216, "learning_rate": 3.703935089496207e-05, "loss": 0.0296, "step": 60430 }, { "grad_norm": 0.34164857864379883, "learning_rate": 3.702338205867829e-05, "loss": 0.0205, "step": 60440 }, { "grad_norm": 0.5579012632369995, "learning_rate": 3.7007414641497976e-05, "loss": 0.0221, "step": 60450 }, { "grad_norm": 0.417224645614624, "learning_rate": 3.699144864516726e-05, "loss": 0.0279, "step": 60460 }, { "grad_norm": 0.552141547203064, "learning_rate": 3.6975484071432153e-05, "loss": 0.0273, "step": 60470 }, { "grad_norm": 0.25541865825653076, "learning_rate": 3.695952092203854e-05, "loss": 0.0308, "step": 60480 }, { "grad_norm": 0.37452101707458496, "learning_rate": 3.69435591987321e-05, "loss": 0.0206, "step": 60490 }, { "grad_norm": 0.3278065025806427, "learning_rate": 3.6927598903258374e-05, "loss": 0.0265, "step": 60500 }, { "grad_norm": 0.30043143033981323, "learning_rate": 3.691164003736281e-05, "loss": 0.0235, "step": 60510 }, { "grad_norm": 0.4282773435115814, "learning_rate": 3.689568260279061e-05, "loss": 0.0215, "step": 60520 }, { "grad_norm": 0.2562209665775299, "learning_rate": 3.687972660128686e-05, "loss": 0.041, "step": 60530 }, { "grad_norm": 0.6409203410148621, "learning_rate": 3.686377203459648e-05, "loss": 0.0377, "step": 60540 }, { "grad_norm": 0.3965228497982025, "learning_rate": 3.6847818904464226e-05, "loss": 0.0268, "step": 60550 }, { "grad_norm": 0.4586258828639984, "learning_rate": 3.683186721263474e-05, "loss": 0.0476, "step": 60560 }, { "grad_norm": 0.5218170881271362, "learning_rate": 3.6815916960852435e-05, "loss": 0.0251, "step": 60570 }, { "grad_norm": 0.7154523134231567, "learning_rate": 3.679996815086165e-05, "loss": 0.0333, "step": 60580 }, { "grad_norm": 0.39763343334198, "learning_rate": 3.678402078440649e-05, "loss": 0.0188, "step": 60590 }, { "grad_norm": 0.5924580097198486, "learning_rate": 3.676807486323096e-05, "loss": 0.0212, "step": 60600 }, { "grad_norm": 2.099902868270874, "learning_rate": 3.6752130389078864e-05, "loss": 0.0362, "step": 60610 }, { "grad_norm": 0.3651018440723419, "learning_rate": 3.673618736369388e-05, "loss": 0.0254, "step": 60620 }, { "grad_norm": 0.30539670586586, "learning_rate": 3.6720245788819474e-05, "loss": 0.0268, "step": 60630 }, { "grad_norm": 0.45568013191223145, "learning_rate": 3.6704305666199044e-05, "loss": 0.0359, "step": 60640 }, { "grad_norm": 0.49621981382369995, "learning_rate": 3.668836699757576e-05, "loss": 0.02, "step": 60650 }, { "grad_norm": 0.3204532265663147, "learning_rate": 3.667242978469265e-05, "loss": 0.0378, "step": 60660 }, { "grad_norm": 0.5126598477363586, "learning_rate": 3.6656494029292575e-05, "loss": 0.0327, "step": 60670 }, { "grad_norm": 1.504430890083313, "learning_rate": 3.664055973311825e-05, "loss": 0.0332, "step": 60680 }, { "grad_norm": 0.292299747467041, "learning_rate": 3.662462689791221e-05, "loss": 0.019, "step": 60690 }, { "grad_norm": 0.6091998815536499, "learning_rate": 3.660869552541689e-05, "loss": 0.0313, "step": 60700 }, { "grad_norm": 0.6179893612861633, "learning_rate": 3.659276561737448e-05, "loss": 0.031, "step": 60710 }, { "grad_norm": 0.4427632987499237, "learning_rate": 3.657683717552708e-05, "loss": 0.0328, "step": 60720 }, { "grad_norm": 0.38404062390327454, "learning_rate": 3.6560910201616574e-05, "loss": 0.0213, "step": 60730 }, { "grad_norm": 1.192830204963684, "learning_rate": 3.654498469738472e-05, "loss": 0.0339, "step": 60740 }, { "grad_norm": 0.26612499356269836, "learning_rate": 3.6529060664573126e-05, "loss": 0.027, "step": 60750 }, { "grad_norm": 0.3307824432849884, "learning_rate": 3.6513138104923176e-05, "loss": 0.0275, "step": 60760 }, { "grad_norm": 0.3970581591129303, "learning_rate": 3.6497217020176176e-05, "loss": 0.0302, "step": 60770 }, { "grad_norm": 0.3469735085964203, "learning_rate": 3.648129741207323e-05, "loss": 0.0353, "step": 60780 }, { "grad_norm": 0.4612467288970947, "learning_rate": 3.646537928235527e-05, "loss": 0.0268, "step": 60790 }, { "grad_norm": 0.5980446934700012, "learning_rate": 3.6449462632763075e-05, "loss": 0.0242, "step": 60800 }, { "grad_norm": 0.3741113245487213, "learning_rate": 3.643354746503729e-05, "loss": 0.0208, "step": 60810 }, { "grad_norm": 0.5086674094200134, "learning_rate": 3.6417633780918345e-05, "loss": 0.0293, "step": 60820 }, { "grad_norm": 0.8019403219223022, "learning_rate": 3.640172158214655e-05, "loss": 0.0267, "step": 60830 }, { "grad_norm": 1.9110223054885864, "learning_rate": 3.6385810870462035e-05, "loss": 0.035, "step": 60840 }, { "grad_norm": 0.851941704750061, "learning_rate": 3.636990164760479e-05, "loss": 0.0289, "step": 60850 }, { "grad_norm": 0.670619547367096, "learning_rate": 3.6353993915314596e-05, "loss": 0.0347, "step": 60860 }, { "grad_norm": 0.523838460445404, "learning_rate": 3.6338087675331124e-05, "loss": 0.0215, "step": 60870 }, { "grad_norm": 0.6869264245033264, "learning_rate": 3.632218292939383e-05, "loss": 0.0363, "step": 60880 }, { "grad_norm": 0.35292699933052063, "learning_rate": 3.6306279679242033e-05, "loss": 0.0301, "step": 60890 }, { "grad_norm": 0.31609660387039185, "learning_rate": 3.629037792661492e-05, "loss": 0.0284, "step": 60900 }, { "grad_norm": 0.40649598836898804, "learning_rate": 3.627447767325147e-05, "loss": 0.0276, "step": 60910 }, { "grad_norm": 1.3275136947631836, "learning_rate": 3.62585789208905e-05, "loss": 0.0303, "step": 60920 }, { "grad_norm": 0.315168172121048, "learning_rate": 3.624268167127068e-05, "loss": 0.0257, "step": 60930 }, { "grad_norm": 0.40465331077575684, "learning_rate": 3.622678592613051e-05, "loss": 0.036, "step": 60940 }, { "grad_norm": 1.864091157913208, "learning_rate": 3.6210891687208304e-05, "loss": 0.0301, "step": 60950 }, { "grad_norm": 0.48558834195137024, "learning_rate": 3.619499895624228e-05, "loss": 0.0357, "step": 60960 }, { "grad_norm": 0.3829449713230133, "learning_rate": 3.61791077349704e-05, "loss": 0.025, "step": 60970 }, { "grad_norm": 0.6280152797698975, "learning_rate": 3.616321802513053e-05, "loss": 0.024, "step": 60980 }, { "grad_norm": 0.377719521522522, "learning_rate": 3.6147329828460325e-05, "loss": 0.0338, "step": 60990 }, { "grad_norm": 0.49339303374290466, "learning_rate": 3.6131443146697306e-05, "loss": 0.0309, "step": 61000 }, { "grad_norm": 0.5376359820365906, "learning_rate": 3.611555798157879e-05, "loss": 0.0236, "step": 61010 }, { "grad_norm": 0.1887870579957962, "learning_rate": 3.6099674334842e-05, "loss": 0.0305, "step": 61020 }, { "grad_norm": 0.4548444449901581, "learning_rate": 3.6083792208223934e-05, "loss": 0.0237, "step": 61030 }, { "grad_norm": 0.5048812031745911, "learning_rate": 3.606791160346142e-05, "loss": 0.0247, "step": 61040 }, { "grad_norm": 0.4048689901828766, "learning_rate": 3.6052032522291134e-05, "loss": 0.0312, "step": 61050 }, { "grad_norm": 0.4580308496952057, "learning_rate": 3.603615496644962e-05, "loss": 0.031, "step": 61060 }, { "grad_norm": 0.4003125727176666, "learning_rate": 3.60202789376732e-05, "loss": 0.0248, "step": 61070 }, { "grad_norm": 0.6949440240859985, "learning_rate": 3.6004404437698036e-05, "loss": 0.0299, "step": 61080 }, { "grad_norm": 0.4311908781528473, "learning_rate": 3.5988531468260184e-05, "loss": 0.0254, "step": 61090 }, { "grad_norm": 0.37133240699768066, "learning_rate": 3.597266003109546e-05, "loss": 0.028, "step": 61100 }, { "grad_norm": 0.8633913397789001, "learning_rate": 3.5956790127939557e-05, "loss": 0.0327, "step": 61110 }, { "grad_norm": 0.6743133068084717, "learning_rate": 3.594092176052796e-05, "loss": 0.0325, "step": 61120 }, { "grad_norm": 0.9828663468360901, "learning_rate": 3.592505493059603e-05, "loss": 0.0464, "step": 61130 }, { "grad_norm": 0.4675065577030182, "learning_rate": 3.5909189639878905e-05, "loss": 0.0366, "step": 61140 }, { "grad_norm": 0.675516664981842, "learning_rate": 3.5893325890111626e-05, "loss": 0.0414, "step": 61150 }, { "grad_norm": 0.530068039894104, "learning_rate": 3.5877463683029034e-05, "loss": 0.0293, "step": 61160 }, { "grad_norm": 0.35917365550994873, "learning_rate": 3.586160302036578e-05, "loss": 0.0271, "step": 61170 }, { "grad_norm": 0.23870410025119781, "learning_rate": 3.584574390385636e-05, "loss": 0.0312, "step": 61180 }, { "grad_norm": 0.7204465866088867, "learning_rate": 3.582988633523511e-05, "loss": 0.0281, "step": 61190 }, { "grad_norm": 0.4705207943916321, "learning_rate": 3.581403031623616e-05, "loss": 0.033, "step": 61200 }, { "grad_norm": 0.3685780465602875, "learning_rate": 3.579817584859354e-05, "loss": 0.035, "step": 61210 }, { "grad_norm": 0.8552918434143066, "learning_rate": 3.578232293404107e-05, "loss": 0.0323, "step": 61220 }, { "grad_norm": 0.7068869471549988, "learning_rate": 3.5766471574312385e-05, "loss": 0.0392, "step": 61230 }, { "grad_norm": 0.44704627990722656, "learning_rate": 3.5750621771140964e-05, "loss": 0.0317, "step": 61240 }, { "grad_norm": 0.5858715772628784, "learning_rate": 3.5734773526260115e-05, "loss": 0.0277, "step": 61250 }, { "grad_norm": 0.3095904290676117, "learning_rate": 3.5718926841402996e-05, "loss": 0.0292, "step": 61260 }, { "grad_norm": 0.5398325324058533, "learning_rate": 3.5703081718302555e-05, "loss": 0.0243, "step": 61270 }, { "grad_norm": 0.37090596556663513, "learning_rate": 3.568723815869162e-05, "loss": 0.0245, "step": 61280 }, { "grad_norm": 0.3079223930835724, "learning_rate": 3.5671396164302777e-05, "loss": 0.0368, "step": 61290 }, { "grad_norm": 0.42753273248672485, "learning_rate": 3.565555573686853e-05, "loss": 0.0311, "step": 61300 }, { "grad_norm": 0.5753559470176697, "learning_rate": 3.563971687812113e-05, "loss": 0.0389, "step": 61310 }, { "grad_norm": 0.4418170154094696, "learning_rate": 3.5623879589792706e-05, "loss": 0.0252, "step": 61320 }, { "grad_norm": 0.3033251166343689, "learning_rate": 3.5608043873615185e-05, "loss": 0.0278, "step": 61330 }, { "grad_norm": 0.41869208216667175, "learning_rate": 3.559220973132035e-05, "loss": 0.0291, "step": 61340 }, { "grad_norm": 0.34885647892951965, "learning_rate": 3.557637716463981e-05, "loss": 0.0249, "step": 61350 }, { "grad_norm": 0.37020936608314514, "learning_rate": 3.556054617530497e-05, "loss": 0.0283, "step": 61360 }, { "grad_norm": 0.3142601251602173, "learning_rate": 3.5544716765047106e-05, "loss": 0.0223, "step": 61370 }, { "grad_norm": 0.34656327962875366, "learning_rate": 3.5528888935597274e-05, "loss": 0.0252, "step": 61380 }, { "grad_norm": 0.29398009181022644, "learning_rate": 3.551306268868636e-05, "loss": 0.026, "step": 61390 }, { "grad_norm": 0.49379026889801025, "learning_rate": 3.5497238026045166e-05, "loss": 0.0282, "step": 61400 }, { "grad_norm": 0.3393397033214569, "learning_rate": 3.548141494940421e-05, "loss": 0.0267, "step": 61410 }, { "grad_norm": 0.5641646981239319, "learning_rate": 3.546559346049388e-05, "loss": 0.0199, "step": 61420 }, { "grad_norm": 0.5837070345878601, "learning_rate": 3.544977356104441e-05, "loss": 0.0267, "step": 61430 }, { "grad_norm": 0.6585462689399719, "learning_rate": 3.543395525278582e-05, "loss": 0.0278, "step": 61440 }, { "grad_norm": 0.3648114800453186, "learning_rate": 3.5418138537447985e-05, "loss": 0.0286, "step": 61450 }, { "grad_norm": 0.3962128758430481, "learning_rate": 3.5402323416760574e-05, "loss": 0.0453, "step": 61460 }, { "grad_norm": 0.5719403028488159, "learning_rate": 3.538650989245316e-05, "loss": 0.0345, "step": 61470 }, { "grad_norm": 0.3869677782058716, "learning_rate": 3.537069796625504e-05, "loss": 0.0208, "step": 61480 }, { "grad_norm": 0.3119337856769562, "learning_rate": 3.53548876398954e-05, "loss": 0.0255, "step": 61490 }, { "grad_norm": 0.45262059569358826, "learning_rate": 3.5339078915103214e-05, "loss": 0.0247, "step": 61500 }, { "grad_norm": 0.4347788393497467, "learning_rate": 3.532327179360733e-05, "loss": 0.032, "step": 61510 }, { "grad_norm": 0.3964073657989502, "learning_rate": 3.530746627713636e-05, "loss": 0.0258, "step": 61520 }, { "grad_norm": 0.3359168767929077, "learning_rate": 3.5291662367418805e-05, "loss": 0.0243, "step": 61530 }, { "grad_norm": 0.8357008695602417, "learning_rate": 3.527586006618293e-05, "loss": 0.0281, "step": 61540 }, { "grad_norm": 0.5313315391540527, "learning_rate": 3.526005937515687e-05, "loss": 0.0306, "step": 61550 }, { "grad_norm": 0.24764375388622284, "learning_rate": 3.524426029606856e-05, "loss": 0.0242, "step": 61560 }, { "grad_norm": 0.5180232524871826, "learning_rate": 3.5228462830645746e-05, "loss": 0.032, "step": 61570 }, { "grad_norm": 1.2640199661254883, "learning_rate": 3.5212666980616036e-05, "loss": 0.0295, "step": 61580 }, { "grad_norm": 0.315004825592041, "learning_rate": 3.519687274770681e-05, "loss": 0.0273, "step": 61590 }, { "grad_norm": 0.5499024391174316, "learning_rate": 3.518108013364535e-05, "loss": 0.0277, "step": 61600 }, { "grad_norm": 0.44399869441986084, "learning_rate": 3.516528914015868e-05, "loss": 0.0261, "step": 61610 }, { "grad_norm": 0.5884255170822144, "learning_rate": 3.5149499768973694e-05, "loss": 0.0417, "step": 61620 }, { "grad_norm": 0.46798551082611084, "learning_rate": 3.513371202181708e-05, "loss": 0.0336, "step": 61630 }, { "grad_norm": 0.26201578974723816, "learning_rate": 3.511792590041537e-05, "loss": 0.0299, "step": 61640 }, { "grad_norm": 0.8146328926086426, "learning_rate": 3.510214140649488e-05, "loss": 0.0295, "step": 61650 }, { "grad_norm": 0.33058393001556396, "learning_rate": 3.508635854178184e-05, "loss": 0.0396, "step": 61660 }, { "grad_norm": 0.5726099014282227, "learning_rate": 3.50705773080022e-05, "loss": 0.034, "step": 61670 }, { "grad_norm": 0.756354570388794, "learning_rate": 3.505479770688178e-05, "loss": 0.0381, "step": 61680 }, { "grad_norm": 0.3221357464790344, "learning_rate": 3.503901974014622e-05, "loss": 0.0238, "step": 61690 }, { "grad_norm": 0.3158215284347534, "learning_rate": 3.502324340952095e-05, "loss": 0.0232, "step": 61700 }, { "grad_norm": 0.8204108476638794, "learning_rate": 3.500746871673126e-05, "loss": 0.0321, "step": 61710 }, { "grad_norm": 0.23793968558311462, "learning_rate": 3.499169566350227e-05, "loss": 0.0274, "step": 61720 }, { "grad_norm": 0.3645409643650055, "learning_rate": 3.4975924251558876e-05, "loss": 0.0219, "step": 61730 }, { "grad_norm": 0.5083181858062744, "learning_rate": 3.496015448262582e-05, "loss": 0.0269, "step": 61740 }, { "grad_norm": 0.31734299659729004, "learning_rate": 3.494438635842764e-05, "loss": 0.0217, "step": 61750 }, { "grad_norm": 0.40992531180381775, "learning_rate": 3.492861988068874e-05, "loss": 0.0287, "step": 61760 }, { "grad_norm": 1.2175222635269165, "learning_rate": 3.491285505113332e-05, "loss": 0.0277, "step": 61770 }, { "grad_norm": 0.4227878153324127, "learning_rate": 3.4897091871485376e-05, "loss": 0.0271, "step": 61780 }, { "grad_norm": 0.24444697797298431, "learning_rate": 3.488133034346875e-05, "loss": 0.0352, "step": 61790 }, { "grad_norm": 0.38983458280563354, "learning_rate": 3.486557046880713e-05, "loss": 0.0303, "step": 61800 }, { "grad_norm": 0.8480038046836853, "learning_rate": 3.484981224922395e-05, "loss": 0.0398, "step": 61810 }, { "grad_norm": 0.5332070589065552, "learning_rate": 3.4834055686442544e-05, "loss": 0.0303, "step": 61820 }, { "grad_norm": 0.4578450620174408, "learning_rate": 3.4818300782186e-05, "loss": 0.0376, "step": 61830 }, { "grad_norm": 0.42662400007247925, "learning_rate": 3.4802547538177236e-05, "loss": 0.0313, "step": 61840 }, { "grad_norm": 0.33490967750549316, "learning_rate": 3.478679595613905e-05, "loss": 0.0265, "step": 61850 }, { "grad_norm": 0.34592509269714355, "learning_rate": 3.4771046037793975e-05, "loss": 0.0221, "step": 61860 }, { "grad_norm": 0.5207197666168213, "learning_rate": 3.4755297784864416e-05, "loss": 0.0216, "step": 61870 }, { "grad_norm": 0.47591105103492737, "learning_rate": 3.473955119907257e-05, "loss": 0.0298, "step": 61880 }, { "grad_norm": 0.5869019627571106, "learning_rate": 3.472380628214046e-05, "loss": 0.0405, "step": 61890 }, { "grad_norm": 0.42087578773498535, "learning_rate": 3.4708063035789906e-05, "loss": 0.0239, "step": 61900 }, { "grad_norm": 0.24170444905757904, "learning_rate": 3.4692321461742614e-05, "loss": 0.029, "step": 61910 }, { "grad_norm": 0.3345700204372406, "learning_rate": 3.4676581561720023e-05, "loss": 0.0241, "step": 61920 }, { "grad_norm": 0.4584089517593384, "learning_rate": 3.466084333744344e-05, "loss": 0.0216, "step": 61930 }, { "grad_norm": 0.3273712396621704, "learning_rate": 3.4645106790633964e-05, "loss": 0.0332, "step": 61940 }, { "grad_norm": 0.4773092269897461, "learning_rate": 3.4629371923012526e-05, "loss": 0.0338, "step": 61950 }, { "grad_norm": 0.49678003787994385, "learning_rate": 3.4613638736299856e-05, "loss": 0.0317, "step": 61960 }, { "grad_norm": 0.4945352375507355, "learning_rate": 3.459790723221652e-05, "loss": 0.0251, "step": 61970 }, { "grad_norm": 0.5340889692306519, "learning_rate": 3.45821774124829e-05, "loss": 0.0298, "step": 61980 }, { "grad_norm": 0.4901738166809082, "learning_rate": 3.456644927881918e-05, "loss": 0.0324, "step": 61990 }, { "grad_norm": 0.3011804223060608, "learning_rate": 3.4550722832945354e-05, "loss": 0.0274, "step": 62000 }, { "grad_norm": 0.2711542844772339, "learning_rate": 3.453499807658126e-05, "loss": 0.0232, "step": 62010 }, { "grad_norm": 0.31722861528396606, "learning_rate": 3.451927501144653e-05, "loss": 0.0279, "step": 62020 }, { "grad_norm": 0.6763314604759216, "learning_rate": 3.450355363926059e-05, "loss": 0.0296, "step": 62030 }, { "grad_norm": 0.413644939661026, "learning_rate": 3.448783396174273e-05, "loss": 0.0247, "step": 62040 }, { "grad_norm": 0.3645049035549164, "learning_rate": 3.4472115980612034e-05, "loss": 0.0347, "step": 62050 }, { "grad_norm": 0.5891615152359009, "learning_rate": 3.4456399697587396e-05, "loss": 0.0283, "step": 62060 }, { "grad_norm": 0.4586580991744995, "learning_rate": 3.4440685114387506e-05, "loss": 0.0303, "step": 62070 }, { "grad_norm": 0.6980863809585571, "learning_rate": 3.442497223273091e-05, "loss": 0.0326, "step": 62080 }, { "grad_norm": 0.36061209440231323, "learning_rate": 3.4409261054335926e-05, "loss": 0.0277, "step": 62090 }, { "grad_norm": 0.46250325441360474, "learning_rate": 3.439355158092069e-05, "loss": 0.0245, "step": 62100 }, { "grad_norm": 0.35358738899230957, "learning_rate": 3.437784381420322e-05, "loss": 0.0228, "step": 62110 }, { "grad_norm": 0.4086476266384125, "learning_rate": 3.436213775590126e-05, "loss": 0.0226, "step": 62120 }, { "grad_norm": 0.4872191250324249, "learning_rate": 3.4346433407732405e-05, "loss": 0.0268, "step": 62130 }, { "grad_norm": 0.5209673643112183, "learning_rate": 3.433073077141405e-05, "loss": 0.0241, "step": 62140 }, { "grad_norm": 0.6512956023216248, "learning_rate": 3.431502984866343e-05, "loss": 0.0244, "step": 62150 }, { "grad_norm": 0.4368506073951721, "learning_rate": 3.429933064119753e-05, "loss": 0.0315, "step": 62160 }, { "grad_norm": 0.6994372606277466, "learning_rate": 3.4283633150733255e-05, "loss": 0.0395, "step": 62170 }, { "grad_norm": 0.24282078444957733, "learning_rate": 3.4267937378987225e-05, "loss": 0.0193, "step": 62180 }, { "grad_norm": 0.38384899497032166, "learning_rate": 3.425224332767592e-05, "loss": 0.0238, "step": 62190 }, { "grad_norm": 0.5563178062438965, "learning_rate": 3.4236550998515595e-05, "loss": 0.0266, "step": 62200 }, { "grad_norm": 0.29210448265075684, "learning_rate": 3.422086039322235e-05, "loss": 0.0303, "step": 62210 }, { "grad_norm": 0.5940820574760437, "learning_rate": 3.420517151351209e-05, "loss": 0.0321, "step": 62220 }, { "grad_norm": 0.7116878628730774, "learning_rate": 3.418948436110053e-05, "loss": 0.0241, "step": 62230 }, { "grad_norm": 0.4177325367927551, "learning_rate": 3.417379893770319e-05, "loss": 0.0215, "step": 62240 }, { "grad_norm": 0.9203541874885559, "learning_rate": 3.41581152450354e-05, "loss": 0.0263, "step": 62250 }, { "grad_norm": 0.3730471134185791, "learning_rate": 3.414243328481231e-05, "loss": 0.0381, "step": 62260 }, { "grad_norm": 0.35548558831214905, "learning_rate": 3.412675305874888e-05, "loss": 0.0353, "step": 62270 }, { "grad_norm": 0.6750905513763428, "learning_rate": 3.4111074568559876e-05, "loss": 0.0281, "step": 62280 }, { "grad_norm": 0.2903002202510834, "learning_rate": 3.4095397815959846e-05, "loss": 0.0223, "step": 62290 }, { "grad_norm": 0.48154884576797485, "learning_rate": 3.407972280266322e-05, "loss": 0.031, "step": 62300 }, { "grad_norm": 0.43187567591667175, "learning_rate": 3.4064049530384166e-05, "loss": 0.0282, "step": 62310 }, { "grad_norm": 0.4605366289615631, "learning_rate": 3.40483780008367e-05, "loss": 0.0225, "step": 62320 }, { "grad_norm": 0.8331908583641052, "learning_rate": 3.4032708215734635e-05, "loss": 0.0269, "step": 62330 }, { "grad_norm": 0.6382821202278137, "learning_rate": 3.4017040176791584e-05, "loss": 0.0313, "step": 62340 }, { "grad_norm": 0.2913292944431305, "learning_rate": 3.4001373885720976e-05, "loss": 0.0467, "step": 62350 }, { "grad_norm": 0.7065551280975342, "learning_rate": 3.398570934423608e-05, "loss": 0.033, "step": 62360 }, { "grad_norm": 0.3529423475265503, "learning_rate": 3.397004655404994e-05, "loss": 0.0192, "step": 62370 }, { "grad_norm": 0.3893466889858246, "learning_rate": 3.395438551687541e-05, "loss": 0.0327, "step": 62380 }, { "grad_norm": 0.45087170600891113, "learning_rate": 3.393872623442515e-05, "loss": 0.022, "step": 62390 }, { "grad_norm": 2.3346219062805176, "learning_rate": 3.392306870841164e-05, "loss": 0.0228, "step": 62400 }, { "grad_norm": 1.1428959369659424, "learning_rate": 3.390741294054715e-05, "loss": 0.0286, "step": 62410 }, { "grad_norm": 0.44742152094841003, "learning_rate": 3.38917589325438e-05, "loss": 0.0275, "step": 62420 }, { "grad_norm": 0.24201752245426178, "learning_rate": 3.387610668611349e-05, "loss": 0.0247, "step": 62430 }, { "grad_norm": 0.24777892231941223, "learning_rate": 3.38604562029679e-05, "loss": 0.0271, "step": 62440 }, { "grad_norm": 0.7192589640617371, "learning_rate": 3.384480748481855e-05, "loss": 0.0278, "step": 62450 }, { "grad_norm": 0.4590148329734802, "learning_rate": 3.382916053337676e-05, "loss": 0.0221, "step": 62460 }, { "grad_norm": 1.4237908124923706, "learning_rate": 3.381351535035366e-05, "loss": 0.0474, "step": 62470 }, { "grad_norm": 0.5231197476387024, "learning_rate": 3.379787193746019e-05, "loss": 0.0258, "step": 62480 }, { "grad_norm": 0.3473513722419739, "learning_rate": 3.378223029640709e-05, "loss": 0.0233, "step": 62490 }, { "grad_norm": 0.6989523768424988, "learning_rate": 3.376659042890489e-05, "loss": 0.0246, "step": 62500 }, { "grad_norm": 0.4655539393424988, "learning_rate": 3.375095233666397e-05, "loss": 0.0225, "step": 62510 }, { "grad_norm": 0.6489865779876709, "learning_rate": 3.373531602139447e-05, "loss": 0.0547, "step": 62520 }, { "grad_norm": 0.46684250235557556, "learning_rate": 3.371968148480635e-05, "loss": 0.0242, "step": 62530 }, { "grad_norm": 0.4936048984527588, "learning_rate": 3.370404872860937e-05, "loss": 0.016, "step": 62540 }, { "grad_norm": 0.7131245136260986, "learning_rate": 3.3688417754513145e-05, "loss": 0.0287, "step": 62550 }, { "grad_norm": 0.5056888461112976, "learning_rate": 3.367278856422703e-05, "loss": 0.0385, "step": 62560 }, { "grad_norm": 0.5147179961204529, "learning_rate": 3.365716115946021e-05, "loss": 0.0231, "step": 62570 }, { "grad_norm": 0.23161238431930542, "learning_rate": 3.3641535541921666e-05, "loss": 0.032, "step": 62580 }, { "grad_norm": 0.3680500388145447, "learning_rate": 3.36259117133202e-05, "loss": 0.0302, "step": 62590 }, { "grad_norm": 0.49437040090560913, "learning_rate": 3.36102896753644e-05, "loss": 0.0456, "step": 62600 }, { "grad_norm": 0.46317189931869507, "learning_rate": 3.35946694297627e-05, "loss": 0.0417, "step": 62610 }, { "grad_norm": 0.49526479840278625, "learning_rate": 3.357905097822328e-05, "loss": 0.0384, "step": 62620 }, { "grad_norm": 0.4025092124938965, "learning_rate": 3.356343432245416e-05, "loss": 0.022, "step": 62630 }, { "grad_norm": 0.37813690304756165, "learning_rate": 3.354781946416315e-05, "loss": 0.0254, "step": 62640 }, { "grad_norm": 0.38379475474357605, "learning_rate": 3.3532206405057864e-05, "loss": 0.0242, "step": 62650 }, { "grad_norm": 0.5277918577194214, "learning_rate": 3.351659514684573e-05, "loss": 0.0341, "step": 62660 }, { "grad_norm": 0.3822339177131653, "learning_rate": 3.350098569123395e-05, "loss": 0.025, "step": 62670 }, { "grad_norm": 0.42213544249534607, "learning_rate": 3.348537803992958e-05, "loss": 0.0309, "step": 62680 }, { "grad_norm": 0.5187427997589111, "learning_rate": 3.346977219463945e-05, "loss": 0.0241, "step": 62690 }, { "grad_norm": 0.5098176598548889, "learning_rate": 3.345416815707017e-05, "loss": 0.0271, "step": 62700 }, { "grad_norm": 0.4577886164188385, "learning_rate": 3.3438565928928185e-05, "loss": 0.0219, "step": 62710 }, { "grad_norm": 0.2951989769935608, "learning_rate": 3.3422965511919736e-05, "loss": 0.0445, "step": 62720 }, { "grad_norm": 0.22996929287910461, "learning_rate": 3.3407366907750846e-05, "loss": 0.0274, "step": 62730 }, { "grad_norm": 0.48836439847946167, "learning_rate": 3.339177011812737e-05, "loss": 0.0317, "step": 62740 }, { "grad_norm": 0.346241295337677, "learning_rate": 3.3376175144754935e-05, "loss": 0.0249, "step": 62750 }, { "grad_norm": 0.37592941522598267, "learning_rate": 3.3360581989339004e-05, "loss": 0.0229, "step": 62760 }, { "grad_norm": 0.5412411093711853, "learning_rate": 3.3344990653584804e-05, "loss": 0.0245, "step": 62770 }, { "grad_norm": 0.451134592294693, "learning_rate": 3.332940113919739e-05, "loss": 0.0233, "step": 62780 }, { "grad_norm": 0.558768093585968, "learning_rate": 3.331381344788161e-05, "loss": 0.0215, "step": 62790 }, { "grad_norm": 0.3473195731639862, "learning_rate": 3.3298227581342076e-05, "loss": 0.0264, "step": 62800 }, { "grad_norm": 0.5598569512367249, "learning_rate": 3.328264354128329e-05, "loss": 0.0292, "step": 62810 }, { "grad_norm": 0.5702947974205017, "learning_rate": 3.3267061329409465e-05, "loss": 0.0234, "step": 62820 }, { "grad_norm": 0.3631569743156433, "learning_rate": 3.325148094742466e-05, "loss": 0.0162, "step": 62830 }, { "grad_norm": 0.5320431590080261, "learning_rate": 3.323590239703271e-05, "loss": 0.0215, "step": 62840 }, { "grad_norm": 0.27354076504707336, "learning_rate": 3.322032567993728e-05, "loss": 0.0276, "step": 62850 }, { "grad_norm": 0.4779166281223297, "learning_rate": 3.3204750797841774e-05, "loss": 0.0214, "step": 62860 }, { "grad_norm": 0.43574783205986023, "learning_rate": 3.31891777524495e-05, "loss": 0.0265, "step": 62870 }, { "grad_norm": 0.33085715770721436, "learning_rate": 3.317360654546347e-05, "loss": 0.0225, "step": 62880 }, { "grad_norm": 0.4530380666255951, "learning_rate": 3.3158037178586534e-05, "loss": 0.0279, "step": 62890 }, { "grad_norm": 0.21727867424488068, "learning_rate": 3.3142469653521316e-05, "loss": 0.0237, "step": 62900 }, { "grad_norm": 1.1940089464187622, "learning_rate": 3.3126903971970284e-05, "loss": 0.0257, "step": 62910 }, { "grad_norm": 0.30825671553611755, "learning_rate": 3.311134013563564e-05, "loss": 0.0317, "step": 62920 }, { "grad_norm": 0.8775857090950012, "learning_rate": 3.309577814621947e-05, "loss": 0.0267, "step": 62930 }, { "grad_norm": 0.3580402433872223, "learning_rate": 3.308021800542359e-05, "loss": 0.0197, "step": 62940 }, { "grad_norm": 0.19542476534843445, "learning_rate": 3.306465971494963e-05, "loss": 0.0198, "step": 62950 }, { "grad_norm": 0.46781447529792786, "learning_rate": 3.304910327649901e-05, "loss": 0.0275, "step": 62960 }, { "grad_norm": 0.38359642028808594, "learning_rate": 3.303354869177297e-05, "loss": 0.0239, "step": 62970 }, { "grad_norm": 0.5287337303161621, "learning_rate": 3.301799596247255e-05, "loss": 0.0265, "step": 62980 }, { "grad_norm": 0.5903441905975342, "learning_rate": 3.300244509029854e-05, "loss": 0.0404, "step": 62990 }, { "grad_norm": 0.4915238320827484, "learning_rate": 3.2986896076951585e-05, "loss": 0.0198, "step": 63000 }, { "grad_norm": 1.129006266593933, "learning_rate": 3.297134892413211e-05, "loss": 0.0212, "step": 63010 }, { "grad_norm": 0.3637818694114685, "learning_rate": 3.295580363354031e-05, "loss": 0.0291, "step": 63020 }, { "grad_norm": 0.3485148549079895, "learning_rate": 3.2940260206876194e-05, "loss": 0.0268, "step": 63030 }, { "grad_norm": 0.6127119660377502, "learning_rate": 3.292471864583958e-05, "loss": 0.0237, "step": 63040 }, { "grad_norm": 0.4378986358642578, "learning_rate": 3.290917895213004e-05, "loss": 0.0283, "step": 63050 }, { "grad_norm": 0.3595097064971924, "learning_rate": 3.289364112744701e-05, "loss": 0.0312, "step": 63060 }, { "grad_norm": 0.3858603537082672, "learning_rate": 3.287810517348967e-05, "loss": 0.026, "step": 63070 }, { "grad_norm": 0.35169678926467896, "learning_rate": 3.2862571091957014e-05, "loss": 0.0177, "step": 63080 }, { "grad_norm": 0.2516200840473175, "learning_rate": 3.284703888454781e-05, "loss": 0.0223, "step": 63090 }, { "grad_norm": 0.3815617263317108, "learning_rate": 3.2831508552960646e-05, "loss": 0.0237, "step": 63100 }, { "grad_norm": 1.1271636486053467, "learning_rate": 3.281598009889387e-05, "loss": 0.0279, "step": 63110 }, { "grad_norm": 0.4415333569049835, "learning_rate": 3.28004535240457e-05, "loss": 0.0243, "step": 63120 }, { "grad_norm": 0.4067465364933014, "learning_rate": 3.278492883011407e-05, "loss": 0.0284, "step": 63130 }, { "grad_norm": 0.28413355350494385, "learning_rate": 3.276940601879675e-05, "loss": 0.022, "step": 63140 }, { "grad_norm": 0.4908362627029419, "learning_rate": 3.275388509179127e-05, "loss": 0.0212, "step": 63150 }, { "grad_norm": 0.5249181985855103, "learning_rate": 3.2738366050794985e-05, "loss": 0.0261, "step": 63160 }, { "grad_norm": 0.41462278366088867, "learning_rate": 3.2722848897505034e-05, "loss": 0.0255, "step": 63170 }, { "grad_norm": 0.39273178577423096, "learning_rate": 3.2707333633618356e-05, "loss": 0.0238, "step": 63180 }, { "grad_norm": 0.2087741494178772, "learning_rate": 3.269182026083167e-05, "loss": 0.024, "step": 63190 }, { "grad_norm": 0.9545254111289978, "learning_rate": 3.2676308780841516e-05, "loss": 0.0244, "step": 63200 }, { "grad_norm": 0.5747256875038147, "learning_rate": 3.266079919534416e-05, "loss": 0.0196, "step": 63210 }, { "grad_norm": 0.644241213798523, "learning_rate": 3.264529150603575e-05, "loss": 0.0381, "step": 63220 }, { "grad_norm": 0.6270655393600464, "learning_rate": 3.262978571461217e-05, "loss": 0.0398, "step": 63230 }, { "grad_norm": 0.6345661878585815, "learning_rate": 3.26142818227691e-05, "loss": 0.0328, "step": 63240 }, { "grad_norm": 1.0237935781478882, "learning_rate": 3.259877983220203e-05, "loss": 0.0358, "step": 63250 }, { "grad_norm": 0.4858332574367523, "learning_rate": 3.2583279744606246e-05, "loss": 0.0312, "step": 63260 }, { "grad_norm": 4.246783256530762, "learning_rate": 3.25677815616768e-05, "loss": 0.0299, "step": 63270 }, { "grad_norm": 0.5645500421524048, "learning_rate": 3.255228528510856e-05, "loss": 0.026, "step": 63280 }, { "grad_norm": 0.24266992509365082, "learning_rate": 3.253679091659617e-05, "loss": 0.028, "step": 63290 }, { "grad_norm": 0.9289735555648804, "learning_rate": 3.252129845783404e-05, "loss": 0.0352, "step": 63300 }, { "grad_norm": 0.303756982088089, "learning_rate": 3.2505807910516456e-05, "loss": 0.0293, "step": 63310 }, { "grad_norm": 0.45070523023605347, "learning_rate": 3.249031927633742e-05, "loss": 0.0239, "step": 63320 }, { "grad_norm": 0.4547960162162781, "learning_rate": 3.2474832556990737e-05, "loss": 0.031, "step": 63330 }, { "grad_norm": 0.4782088100910187, "learning_rate": 3.245934775417002e-05, "loss": 0.0286, "step": 63340 }, { "grad_norm": 0.6854742169380188, "learning_rate": 3.244386486956866e-05, "loss": 0.0255, "step": 63350 }, { "grad_norm": 0.6999183893203735, "learning_rate": 3.2428383904879846e-05, "loss": 0.0318, "step": 63360 }, { "grad_norm": 0.5984449982643127, "learning_rate": 3.241290486179652e-05, "loss": 0.026, "step": 63370 }, { "grad_norm": 0.28886356949806213, "learning_rate": 3.239742774201151e-05, "loss": 0.0264, "step": 63380 }, { "grad_norm": 0.4082147479057312, "learning_rate": 3.238195254721733e-05, "loss": 0.0299, "step": 63390 }, { "grad_norm": 0.26022523641586304, "learning_rate": 3.236647927910633e-05, "loss": 0.0288, "step": 63400 }, { "grad_norm": 0.5710222125053406, "learning_rate": 3.2351007939370646e-05, "loss": 0.0326, "step": 63410 }, { "grad_norm": 0.22310341894626617, "learning_rate": 3.2335538529702184e-05, "loss": 0.0236, "step": 63420 }, { "grad_norm": 0.3882097899913788, "learning_rate": 3.232007105179268e-05, "loss": 0.0309, "step": 63430 }, { "grad_norm": 1.608930230140686, "learning_rate": 3.230460550733363e-05, "loss": 0.0439, "step": 63440 }, { "grad_norm": 0.6969218254089355, "learning_rate": 3.2289141898016315e-05, "loss": 0.0303, "step": 63450 }, { "grad_norm": 0.5669470429420471, "learning_rate": 3.227368022553181e-05, "loss": 0.0309, "step": 63460 }, { "grad_norm": 0.35936084389686584, "learning_rate": 3.225822049157099e-05, "loss": 0.0352, "step": 63470 }, { "grad_norm": 0.4528197646141052, "learning_rate": 3.2242762697824494e-05, "loss": 0.0253, "step": 63480 }, { "grad_norm": 0.5701975226402283, "learning_rate": 3.222730684598279e-05, "loss": 0.027, "step": 63490 }, { "grad_norm": 0.5542851090431213, "learning_rate": 3.221185293773605e-05, "loss": 0.0258, "step": 63500 }, { "grad_norm": 0.5948411822319031, "learning_rate": 3.2196400974774354e-05, "loss": 0.0333, "step": 63510 }, { "grad_norm": 0.39349350333213806, "learning_rate": 3.2180950958787485e-05, "loss": 0.0244, "step": 63520 }, { "grad_norm": 0.42005643248558044, "learning_rate": 3.2165502891465024e-05, "loss": 0.0455, "step": 63530 }, { "grad_norm": 0.3534899353981018, "learning_rate": 3.215005677449636e-05, "loss": 0.035, "step": 63540 }, { "grad_norm": 0.6287833452224731, "learning_rate": 3.2134612609570644e-05, "loss": 0.0363, "step": 63550 }, { "grad_norm": 0.5403743386268616, "learning_rate": 3.211917039837682e-05, "loss": 0.0312, "step": 63560 }, { "grad_norm": 0.580374538898468, "learning_rate": 3.210373014260365e-05, "loss": 0.0304, "step": 63570 }, { "grad_norm": 0.33007732033729553, "learning_rate": 3.208829184393965e-05, "loss": 0.0536, "step": 63580 }, { "grad_norm": 0.3675975799560547, "learning_rate": 3.207285550407312e-05, "loss": 0.0339, "step": 63590 }, { "grad_norm": 0.40473130345344543, "learning_rate": 3.205742112469217e-05, "loss": 0.0216, "step": 63600 }, { "grad_norm": 0.8318395614624023, "learning_rate": 3.204198870748465e-05, "loss": 0.0547, "step": 63610 }, { "grad_norm": 0.18979793787002563, "learning_rate": 3.202655825413824e-05, "loss": 0.0221, "step": 63620 }, { "grad_norm": 0.5344874262809753, "learning_rate": 3.2011129766340416e-05, "loss": 0.0313, "step": 63630 }, { "grad_norm": 0.3686073422431946, "learning_rate": 3.1995703245778384e-05, "loss": 0.0237, "step": 63640 }, { "grad_norm": 0.4250020682811737, "learning_rate": 3.1980278694139186e-05, "loss": 0.0291, "step": 63650 }, { "grad_norm": 0.2654089331626892, "learning_rate": 3.196485611310962e-05, "loss": 0.0372, "step": 63660 }, { "grad_norm": 0.287929505109787, "learning_rate": 3.1949435504376257e-05, "loss": 0.0355, "step": 63670 }, { "grad_norm": 0.48657774925231934, "learning_rate": 3.1934016869625505e-05, "loss": 0.0235, "step": 63680 }, { "grad_norm": 0.6288666129112244, "learning_rate": 3.191860021054349e-05, "loss": 0.0254, "step": 63690 }, { "grad_norm": 0.5807739496231079, "learning_rate": 3.190318552881618e-05, "loss": 0.0355, "step": 63700 }, { "grad_norm": 0.3019794225692749, "learning_rate": 3.1887772826129285e-05, "loss": 0.0219, "step": 63710 }, { "grad_norm": 0.8626832365989685, "learning_rate": 3.1872362104168335e-05, "loss": 0.0261, "step": 63720 }, { "grad_norm": 6.172033309936523, "learning_rate": 3.185695336461861e-05, "loss": 0.0358, "step": 63730 }, { "grad_norm": 0.47055181860923767, "learning_rate": 3.184154660916519e-05, "loss": 0.0387, "step": 63740 }, { "grad_norm": 0.3635861277580261, "learning_rate": 3.1826141839492906e-05, "loss": 0.0294, "step": 63750 }, { "grad_norm": 0.5111253261566162, "learning_rate": 3.181073905728644e-05, "loss": 0.0247, "step": 63760 }, { "grad_norm": 0.3636792302131653, "learning_rate": 3.1795338264230214e-05, "loss": 0.0238, "step": 63770 }, { "grad_norm": 0.49417123198509216, "learning_rate": 3.1779939462008426e-05, "loss": 0.0476, "step": 63780 }, { "grad_norm": 0.3970247209072113, "learning_rate": 3.1764542652305066e-05, "loss": 0.0208, "step": 63790 }, { "grad_norm": 0.35200241208076477, "learning_rate": 3.17491478368039e-05, "loss": 0.0283, "step": 63800 }, { "grad_norm": 0.4800252318382263, "learning_rate": 3.173375501718847e-05, "loss": 0.0265, "step": 63810 }, { "grad_norm": 0.5126370191574097, "learning_rate": 3.171836419514215e-05, "loss": 0.0422, "step": 63820 }, { "grad_norm": 0.37559080123901367, "learning_rate": 3.170297537234803e-05, "loss": 0.0271, "step": 63830 }, { "grad_norm": 0.688334047794342, "learning_rate": 3.168758855048901e-05, "loss": 0.0605, "step": 63840 }, { "grad_norm": 0.4848240315914154, "learning_rate": 3.1672203731247785e-05, "loss": 0.0211, "step": 63850 }, { "grad_norm": 0.4340948462486267, "learning_rate": 3.16568209163068e-05, "loss": 0.0225, "step": 63860 }, { "grad_norm": 0.5053966045379639, "learning_rate": 3.1641440107348295e-05, "loss": 0.0222, "step": 63870 }, { "grad_norm": 0.5031121969223022, "learning_rate": 3.162606130605429e-05, "loss": 0.029, "step": 63880 }, { "grad_norm": 0.5113721489906311, "learning_rate": 3.16106845141066e-05, "loss": 0.0421, "step": 63890 }, { "grad_norm": 0.3756970465183258, "learning_rate": 3.159530973318682e-05, "loss": 0.0257, "step": 63900 }, { "grad_norm": 0.42298468947410583, "learning_rate": 3.1579936964976286e-05, "loss": 0.0257, "step": 63910 }, { "grad_norm": 0.333699494600296, "learning_rate": 3.156456621115615e-05, "loss": 0.0426, "step": 63920 }, { "grad_norm": 0.3625761866569519, "learning_rate": 3.154919747340734e-05, "loss": 0.0277, "step": 63930 }, { "grad_norm": 0.9022161960601807, "learning_rate": 3.153383075341054e-05, "loss": 0.0282, "step": 63940 }, { "grad_norm": 0.33338841795921326, "learning_rate": 3.1518466052846264e-05, "loss": 0.0176, "step": 63950 }, { "grad_norm": 0.32215991616249084, "learning_rate": 3.150310337339474e-05, "loss": 0.0201, "step": 63960 }, { "grad_norm": 0.5880865454673767, "learning_rate": 3.148774271673603e-05, "loss": 0.0426, "step": 63970 }, { "grad_norm": 0.26007911562919617, "learning_rate": 3.147238408454995e-05, "loss": 0.0315, "step": 63980 }, { "grad_norm": 0.31930461525917053, "learning_rate": 3.1457027478516086e-05, "loss": 0.0196, "step": 63990 }, { "grad_norm": 0.5601866841316223, "learning_rate": 3.144167290031382e-05, "loss": 0.0283, "step": 64000 }, { "grad_norm": 0.6097289323806763, "learning_rate": 3.142632035162227e-05, "loss": 0.0371, "step": 64010 }, { "grad_norm": 0.3872638940811157, "learning_rate": 3.141096983412044e-05, "loss": 0.0249, "step": 64020 }, { "grad_norm": 0.3461499810218811, "learning_rate": 3.139562134948699e-05, "loss": 0.0216, "step": 64030 }, { "grad_norm": 0.29400426149368286, "learning_rate": 3.1380274899400424e-05, "loss": 0.0283, "step": 64040 }, { "grad_norm": 0.2508418560028076, "learning_rate": 3.1364930485539e-05, "loss": 0.0266, "step": 64050 }, { "grad_norm": 0.46694114804267883, "learning_rate": 3.134958810958076e-05, "loss": 0.0257, "step": 64060 }, { "grad_norm": 0.3830202519893646, "learning_rate": 3.13342477732035e-05, "loss": 0.0256, "step": 64070 }, { "grad_norm": 0.46030908823013306, "learning_rate": 3.131890947808487e-05, "loss": 0.0277, "step": 64080 }, { "grad_norm": 0.36908450722694397, "learning_rate": 3.13035732259022e-05, "loss": 0.0265, "step": 64090 }, { "grad_norm": 0.6453779339790344, "learning_rate": 3.1288239018332656e-05, "loss": 0.0238, "step": 64100 }, { "grad_norm": 0.2491956502199173, "learning_rate": 3.1272906857053164e-05, "loss": 0.0292, "step": 64110 }, { "grad_norm": 0.46998947858810425, "learning_rate": 3.125757674374042e-05, "loss": 0.0314, "step": 64120 }, { "grad_norm": 0.805273175239563, "learning_rate": 3.124224868007088e-05, "loss": 0.0297, "step": 64130 }, { "grad_norm": 0.40920281410217285, "learning_rate": 3.1226922667720854e-05, "loss": 0.0327, "step": 64140 }, { "grad_norm": 0.34631848335266113, "learning_rate": 3.121159870836633e-05, "loss": 0.0233, "step": 64150 }, { "grad_norm": 0.24763235449790955, "learning_rate": 3.119627680368312e-05, "loss": 0.0209, "step": 64160 }, { "grad_norm": 1.3667373657226562, "learning_rate": 3.1180956955346816e-05, "loss": 0.0486, "step": 64170 }, { "grad_norm": 0.6212294697761536, "learning_rate": 3.116563916503276e-05, "loss": 0.0348, "step": 64180 }, { "grad_norm": 0.39105623960494995, "learning_rate": 3.1150323434416095e-05, "loss": 0.0277, "step": 64190 }, { "grad_norm": 0.4715050160884857, "learning_rate": 3.1135009765171696e-05, "loss": 0.0233, "step": 64200 }, { "grad_norm": 0.48524022102355957, "learning_rate": 3.111969815897429e-05, "loss": 0.0258, "step": 64210 }, { "grad_norm": 0.6812503933906555, "learning_rate": 3.1104388617498295e-05, "loss": 0.0247, "step": 64220 }, { "grad_norm": 3.750030755996704, "learning_rate": 3.1089081142417966e-05, "loss": 0.0299, "step": 64230 }, { "grad_norm": 0.3681304454803467, "learning_rate": 3.1073775735407285e-05, "loss": 0.0231, "step": 64240 }, { "grad_norm": 0.2690269351005554, "learning_rate": 3.1058472398140035e-05, "loss": 0.0179, "step": 64250 }, { "grad_norm": 0.6603783369064331, "learning_rate": 3.1043171132289736e-05, "loss": 0.0324, "step": 64260 }, { "grad_norm": 0.39309078454971313, "learning_rate": 3.1027871939529774e-05, "loss": 0.0247, "step": 64270 }, { "grad_norm": 0.4797961711883545, "learning_rate": 3.1012574821533214e-05, "loss": 0.0273, "step": 64280 }, { "grad_norm": 0.2473805844783783, "learning_rate": 3.099727977997291e-05, "loss": 0.0286, "step": 64290 }, { "grad_norm": 0.4803331196308136, "learning_rate": 3.098198681652154e-05, "loss": 0.0185, "step": 64300 }, { "grad_norm": 0.3576025068759918, "learning_rate": 3.0966695932851475e-05, "loss": 0.0182, "step": 64310 }, { "grad_norm": 0.5229568481445312, "learning_rate": 3.0951407130634926e-05, "loss": 0.0273, "step": 64320 }, { "grad_norm": 0.6650587916374207, "learning_rate": 3.0936120411543864e-05, "loss": 0.0347, "step": 64330 }, { "grad_norm": 0.3021101653575897, "learning_rate": 3.0920835777250015e-05, "loss": 0.0264, "step": 64340 }, { "grad_norm": 1.1193768978118896, "learning_rate": 3.090555322942488e-05, "loss": 0.0281, "step": 64350 }, { "grad_norm": 0.37115952372550964, "learning_rate": 3.089027276973973e-05, "loss": 0.0241, "step": 64360 }, { "grad_norm": 0.3637750744819641, "learning_rate": 3.0874994399865615e-05, "loss": 0.0195, "step": 64370 }, { "grad_norm": 0.47404852509498596, "learning_rate": 3.085971812147337e-05, "loss": 0.0195, "step": 64380 }, { "grad_norm": 0.3804216682910919, "learning_rate": 3.0844443936233557e-05, "loss": 0.0242, "step": 64390 }, { "grad_norm": 0.28371089696884155, "learning_rate": 3.082917184581657e-05, "loss": 0.0257, "step": 64400 }, { "grad_norm": 0.3078831434249878, "learning_rate": 3.0813901851892513e-05, "loss": 0.0318, "step": 64410 }, { "grad_norm": 0.460525780916214, "learning_rate": 3.0798633956131316e-05, "loss": 0.0265, "step": 64420 }, { "grad_norm": 0.3783182203769684, "learning_rate": 3.078336816020264e-05, "loss": 0.0272, "step": 64430 }, { "grad_norm": 0.9864034056663513, "learning_rate": 3.0768104465775935e-05, "loss": 0.0362, "step": 64440 }, { "grad_norm": 0.14381837844848633, "learning_rate": 3.0752842874520396e-05, "loss": 0.0234, "step": 64450 }, { "grad_norm": 0.2830413281917572, "learning_rate": 3.073758338810503e-05, "loss": 0.0247, "step": 64460 }, { "grad_norm": 0.45401978492736816, "learning_rate": 3.072232600819859e-05, "loss": 0.0433, "step": 64470 }, { "grad_norm": 0.9160100221633911, "learning_rate": 3.070707073646959e-05, "loss": 0.0267, "step": 64480 }, { "grad_norm": 0.45128288865089417, "learning_rate": 3.069181757458633e-05, "loss": 0.0298, "step": 64490 }, { "grad_norm": 0.6540843844413757, "learning_rate": 3.067656652421687e-05, "loss": 0.0326, "step": 64500 }, { "grad_norm": 0.2897379398345947, "learning_rate": 3.0661317587029025e-05, "loss": 0.0257, "step": 64510 }, { "grad_norm": 0.6199365854263306, "learning_rate": 3.064607076469043e-05, "loss": 0.026, "step": 64520 }, { "grad_norm": 0.7501678466796875, "learning_rate": 3.063082605886845e-05, "loss": 0.027, "step": 64530 }, { "grad_norm": 0.3456689119338989, "learning_rate": 3.061558347123019e-05, "loss": 0.0215, "step": 64540 }, { "grad_norm": 0.24904625117778778, "learning_rate": 3.060034300344259e-05, "loss": 0.016, "step": 64550 }, { "grad_norm": 0.2588728070259094, "learning_rate": 3.058510465717231e-05, "loss": 0.0202, "step": 64560 }, { "grad_norm": 0.416856586933136, "learning_rate": 3.0569868434085804e-05, "loss": 0.0258, "step": 64570 }, { "grad_norm": 0.3650359511375427, "learning_rate": 3.055463433584924e-05, "loss": 0.0217, "step": 64580 }, { "grad_norm": 0.28723466396331787, "learning_rate": 3.0539402364128656e-05, "loss": 0.0197, "step": 64590 }, { "grad_norm": 0.4515955448150635, "learning_rate": 3.052417252058977e-05, "loss": 0.0251, "step": 64600 }, { "grad_norm": 0.29868823289871216, "learning_rate": 3.0508944806898092e-05, "loss": 0.0285, "step": 64610 }, { "grad_norm": 0.39467084407806396, "learning_rate": 3.04937192247189e-05, "loss": 0.0337, "step": 64620 }, { "grad_norm": 0.5837359428405762, "learning_rate": 3.047849577571725e-05, "loss": 0.0218, "step": 64630 }, { "grad_norm": 0.651253342628479, "learning_rate": 3.046327446155795e-05, "loss": 0.0246, "step": 64640 }, { "grad_norm": 0.4663023352622986, "learning_rate": 3.0448055283905587e-05, "loss": 0.0327, "step": 64650 }, { "grad_norm": 0.2459907978773117, "learning_rate": 3.0432838244424494e-05, "loss": 0.0186, "step": 64660 }, { "grad_norm": 0.43122050166130066, "learning_rate": 3.04176233447788e-05, "loss": 0.0307, "step": 64670 }, { "grad_norm": 0.36864781379699707, "learning_rate": 3.040241058663238e-05, "loss": 0.0263, "step": 64680 }, { "grad_norm": 0.45563170313835144, "learning_rate": 3.038719997164887e-05, "loss": 0.0225, "step": 64690 }, { "grad_norm": 0.4637534022331238, "learning_rate": 3.0371991501491683e-05, "loss": 0.0265, "step": 64700 }, { "grad_norm": 0.38594338297843933, "learning_rate": 3.0356785177823975e-05, "loss": 0.0364, "step": 64710 }, { "grad_norm": 0.5408385992050171, "learning_rate": 3.0341581002308726e-05, "loss": 0.0245, "step": 64720 }, { "grad_norm": 0.5511568784713745, "learning_rate": 3.0326378976608627e-05, "loss": 0.0339, "step": 64730 }, { "grad_norm": 1.1132707595825195, "learning_rate": 3.0311179102386146e-05, "loss": 0.0286, "step": 64740 }, { "grad_norm": 0.4057772755622864, "learning_rate": 3.0295981381303513e-05, "loss": 0.0291, "step": 64750 }, { "grad_norm": 0.28483644127845764, "learning_rate": 3.0280785815022727e-05, "loss": 0.0223, "step": 64760 }, { "grad_norm": 0.49949225783348083, "learning_rate": 3.0265592405205544e-05, "loss": 0.0182, "step": 64770 }, { "grad_norm": 0.6452601552009583, "learning_rate": 3.0250401153513517e-05, "loss": 0.0253, "step": 64780 }, { "grad_norm": 0.530592143535614, "learning_rate": 3.0235212061607936e-05, "loss": 0.0224, "step": 64790 }, { "grad_norm": 0.845431387424469, "learning_rate": 3.0220025131149842e-05, "loss": 0.0338, "step": 64800 }, { "grad_norm": 0.44454509019851685, "learning_rate": 3.020484036380007e-05, "loss": 0.0257, "step": 64810 }, { "grad_norm": 0.500033974647522, "learning_rate": 3.0189657761219188e-05, "loss": 0.025, "step": 64820 }, { "grad_norm": 0.2877444624900818, "learning_rate": 3.017447732506753e-05, "loss": 0.02, "step": 64830 }, { "grad_norm": 0.3280223309993744, "learning_rate": 3.015929905700524e-05, "loss": 0.0316, "step": 64840 }, { "grad_norm": 0.5566269159317017, "learning_rate": 3.014412295869218e-05, "loss": 0.0296, "step": 64850 }, { "grad_norm": 0.6781213879585266, "learning_rate": 3.0128949031787977e-05, "loss": 0.0243, "step": 64860 }, { "grad_norm": 0.2667594254016876, "learning_rate": 3.0113777277952022e-05, "loss": 0.0245, "step": 64870 }, { "grad_norm": 0.32510778307914734, "learning_rate": 3.0098607698843496e-05, "loss": 0.0187, "step": 64880 }, { "grad_norm": 0.7516671419143677, "learning_rate": 3.0083440296121312e-05, "loss": 0.0283, "step": 64890 }, { "grad_norm": 0.3713901937007904, "learning_rate": 3.0068275071444142e-05, "loss": 0.0285, "step": 64900 }, { "grad_norm": 0.2769542932510376, "learning_rate": 3.0053112026470442e-05, "loss": 0.0225, "step": 64910 }, { "grad_norm": 1.0067713260650635, "learning_rate": 3.0037951162858434e-05, "loss": 0.0299, "step": 64920 }, { "grad_norm": 0.26491278409957886, "learning_rate": 3.002279248226607e-05, "loss": 0.0251, "step": 64930 }, { "grad_norm": 0.5085199475288391, "learning_rate": 3.000763598635109e-05, "loss": 0.0411, "step": 64940 }, { "grad_norm": 0.5113978981971741, "learning_rate": 2.999248167677098e-05, "loss": 0.0318, "step": 64950 }, { "grad_norm": 0.38232263922691345, "learning_rate": 2.9977329555182976e-05, "loss": 0.0242, "step": 64960 }, { "grad_norm": 0.34160515666007996, "learning_rate": 2.9962179623244124e-05, "loss": 0.0419, "step": 64970 }, { "grad_norm": 0.5538060069084167, "learning_rate": 2.9947031882611175e-05, "loss": 0.0228, "step": 64980 }, { "grad_norm": 0.26884177327156067, "learning_rate": 2.9931886334940685e-05, "loss": 0.0293, "step": 64990 }, { "grad_norm": 0.4211253225803375, "learning_rate": 2.9916742981888923e-05, "loss": 0.0228, "step": 65000 }, { "grad_norm": 0.3924848437309265, "learning_rate": 2.990160182511196e-05, "loss": 0.0171, "step": 65010 }, { "grad_norm": 0.33388859033584595, "learning_rate": 2.9886462866265574e-05, "loss": 0.0213, "step": 65020 }, { "grad_norm": 0.49933311343193054, "learning_rate": 2.9871326107005392e-05, "loss": 0.0275, "step": 65030 }, { "grad_norm": 0.4213320314884186, "learning_rate": 2.9856191548986735e-05, "loss": 0.0249, "step": 65040 }, { "grad_norm": 0.5296128392219543, "learning_rate": 2.9841059193864672e-05, "loss": 0.0213, "step": 65050 }, { "grad_norm": 0.3549690842628479, "learning_rate": 2.982592904329407e-05, "loss": 0.0227, "step": 65060 }, { "grad_norm": 0.3944668769836426, "learning_rate": 2.981080109892954e-05, "loss": 0.0232, "step": 65070 }, { "grad_norm": 0.43349042534828186, "learning_rate": 2.9795675362425434e-05, "loss": 0.0328, "step": 65080 }, { "grad_norm": 0.4919166564941406, "learning_rate": 2.97805518354359e-05, "loss": 0.0458, "step": 65090 }, { "grad_norm": 0.4149213433265686, "learning_rate": 2.9765430519614825e-05, "loss": 0.0326, "step": 65100 }, { "grad_norm": 0.4136279225349426, "learning_rate": 2.975031141661584e-05, "loss": 0.0291, "step": 65110 }, { "grad_norm": 0.33553165197372437, "learning_rate": 2.9735194528092347e-05, "loss": 0.0431, "step": 65120 }, { "grad_norm": 0.2966233789920807, "learning_rate": 2.9720079855697525e-05, "loss": 0.0258, "step": 65130 }, { "grad_norm": 0.37912696599960327, "learning_rate": 2.970496740108427e-05, "loss": 0.0348, "step": 65140 }, { "grad_norm": 0.4393080770969391, "learning_rate": 2.968985716590527e-05, "loss": 0.02, "step": 65150 }, { "grad_norm": 0.685769259929657, "learning_rate": 2.9674749151812953e-05, "loss": 0.0264, "step": 65160 }, { "grad_norm": 0.6876621246337891, "learning_rate": 2.965964336045951e-05, "loss": 0.0281, "step": 65170 }, { "grad_norm": 0.36428308486938477, "learning_rate": 2.9644539793496894e-05, "loss": 0.0226, "step": 65180 }, { "grad_norm": 0.3394888639450073, "learning_rate": 2.9629438452576807e-05, "loss": 0.0231, "step": 65190 }, { "grad_norm": 0.7810846567153931, "learning_rate": 2.9614339339350698e-05, "loss": 0.0293, "step": 65200 }, { "grad_norm": 0.6046932339668274, "learning_rate": 2.959924245546979e-05, "loss": 0.0409, "step": 65210 }, { "grad_norm": 0.4385223686695099, "learning_rate": 2.9584147802585038e-05, "loss": 0.0216, "step": 65220 }, { "grad_norm": 0.26206690073013306, "learning_rate": 2.956905538234721e-05, "loss": 0.0284, "step": 65230 }, { "grad_norm": 0.2200259268283844, "learning_rate": 2.9553965196406764e-05, "loss": 0.0243, "step": 65240 }, { "grad_norm": 0.25283586978912354, "learning_rate": 2.9538877246413943e-05, "loss": 0.0231, "step": 65250 }, { "grad_norm": 0.555443525314331, "learning_rate": 2.952379153401874e-05, "loss": 0.0238, "step": 65260 }, { "grad_norm": 0.2819216847419739, "learning_rate": 2.9508708060870914e-05, "loss": 0.0227, "step": 65270 }, { "grad_norm": 0.4063030481338501, "learning_rate": 2.9493626828619948e-05, "loss": 0.019, "step": 65280 }, { "grad_norm": 0.4473826587200165, "learning_rate": 2.9478547838915137e-05, "loss": 0.0255, "step": 65290 }, { "grad_norm": 0.37141507863998413, "learning_rate": 2.946347109340548e-05, "loss": 0.0242, "step": 65300 }, { "grad_norm": 0.3945533335208893, "learning_rate": 2.944839659373976e-05, "loss": 0.02, "step": 65310 }, { "grad_norm": 0.3812294900417328, "learning_rate": 2.943332434156647e-05, "loss": 0.0281, "step": 65320 }, { "grad_norm": 0.2927621901035309, "learning_rate": 2.9418254338533912e-05, "loss": 0.0373, "step": 65330 }, { "grad_norm": 0.2232895791530609, "learning_rate": 2.940318658629011e-05, "loss": 0.0248, "step": 65340 }, { "grad_norm": 0.5333719253540039, "learning_rate": 2.938812108648287e-05, "loss": 0.0234, "step": 65350 }, { "grad_norm": 0.37465694546699524, "learning_rate": 2.937305784075972e-05, "loss": 0.0238, "step": 65360 }, { "grad_norm": 0.30780065059661865, "learning_rate": 2.9357996850767937e-05, "loss": 0.0242, "step": 65370 }, { "grad_norm": 0.5592530965805054, "learning_rate": 2.9342938118154595e-05, "loss": 0.0355, "step": 65380 }, { "grad_norm": 0.6654554009437561, "learning_rate": 2.9327881644566485e-05, "loss": 0.0257, "step": 65390 }, { "grad_norm": 0.36830562353134155, "learning_rate": 2.9312827431650157e-05, "loss": 0.027, "step": 65400 }, { "grad_norm": 0.5075503587722778, "learning_rate": 2.92977754810519e-05, "loss": 0.0232, "step": 65410 }, { "grad_norm": 1.1535110473632812, "learning_rate": 2.928272579441781e-05, "loss": 0.0429, "step": 65420 }, { "grad_norm": 0.46589696407318115, "learning_rate": 2.9267678373393676e-05, "loss": 0.0269, "step": 65430 }, { "grad_norm": 0.3469230532646179, "learning_rate": 2.925263321962507e-05, "loss": 0.0303, "step": 65440 }, { "grad_norm": 0.5506428480148315, "learning_rate": 2.9237590334757292e-05, "loss": 0.0294, "step": 65450 }, { "grad_norm": 0.35359886288642883, "learning_rate": 2.922254972043543e-05, "loss": 0.0195, "step": 65460 }, { "grad_norm": 0.7521533966064453, "learning_rate": 2.9207511378304265e-05, "loss": 0.0293, "step": 65470 }, { "grad_norm": 0.22855360805988312, "learning_rate": 2.9192475310008428e-05, "loss": 0.0314, "step": 65480 }, { "grad_norm": 0.540844202041626, "learning_rate": 2.917744151719219e-05, "loss": 0.0221, "step": 65490 }, { "grad_norm": 0.2935408651828766, "learning_rate": 2.916241000149964e-05, "loss": 0.0204, "step": 65500 }, { "grad_norm": 0.4475685954093933, "learning_rate": 2.9147380764574622e-05, "loss": 0.019, "step": 65510 }, { "grad_norm": 0.39133962988853455, "learning_rate": 2.9132353808060674e-05, "loss": 0.0239, "step": 65520 }, { "grad_norm": 0.2092990130186081, "learning_rate": 2.9117329133601134e-05, "loss": 0.0271, "step": 65530 }, { "grad_norm": 0.475359171628952, "learning_rate": 2.9102306742839093e-05, "loss": 0.0232, "step": 65540 }, { "grad_norm": 0.32503920793533325, "learning_rate": 2.9087286637417375e-05, "loss": 0.034, "step": 65550 }, { "grad_norm": 0.33349186182022095, "learning_rate": 2.9072268818978537e-05, "loss": 0.0252, "step": 65560 }, { "grad_norm": 0.3284253180027008, "learning_rate": 2.905725328916493e-05, "loss": 0.0192, "step": 65570 }, { "grad_norm": 0.33465319871902466, "learning_rate": 2.9042240049618608e-05, "loss": 0.0192, "step": 65580 }, { "grad_norm": 0.541718065738678, "learning_rate": 2.9027229101981413e-05, "loss": 0.0262, "step": 65590 }, { "grad_norm": 0.4756763279438019, "learning_rate": 2.901222044789488e-05, "loss": 0.0283, "step": 65600 }, { "grad_norm": 0.31294727325439453, "learning_rate": 2.89972140890004e-05, "loss": 0.0258, "step": 65610 }, { "grad_norm": 0.4328734576702118, "learning_rate": 2.8982210026938988e-05, "loss": 0.0237, "step": 65620 }, { "grad_norm": 0.5151239633560181, "learning_rate": 2.896720826335151e-05, "loss": 0.0234, "step": 65630 }, { "grad_norm": 0.3976406157016754, "learning_rate": 2.8952208799878493e-05, "loss": 0.025, "step": 65640 }, { "grad_norm": 0.2918260991573334, "learning_rate": 2.89372116381603e-05, "loss": 0.022, "step": 65650 }, { "grad_norm": 0.3055170178413391, "learning_rate": 2.8922216779836932e-05, "loss": 0.0414, "step": 65660 }, { "grad_norm": 0.3855399191379547, "learning_rate": 2.8907224226548285e-05, "loss": 0.0226, "step": 65670 }, { "grad_norm": 0.6057531833648682, "learning_rate": 2.8892233979933868e-05, "loss": 0.0265, "step": 65680 }, { "grad_norm": 0.4475123882293701, "learning_rate": 2.8877246041633026e-05, "loss": 0.0281, "step": 65690 }, { "grad_norm": 0.34638771414756775, "learning_rate": 2.886226041328478e-05, "loss": 0.0276, "step": 65700 }, { "grad_norm": 0.4912141263484955, "learning_rate": 2.8847277096527957e-05, "loss": 0.0244, "step": 65710 }, { "grad_norm": 0.32661527395248413, "learning_rate": 2.8832296093001105e-05, "loss": 0.0282, "step": 65720 }, { "grad_norm": 0.666584312915802, "learning_rate": 2.8817317404342547e-05, "loss": 0.0263, "step": 65730 }, { "grad_norm": 0.5300440788269043, "learning_rate": 2.880234103219029e-05, "loss": 0.0267, "step": 65740 }, { "grad_norm": 0.49483901262283325, "learning_rate": 2.878736697818215e-05, "loss": 0.0229, "step": 65750 }, { "grad_norm": 0.41088351607322693, "learning_rate": 2.877239524395568e-05, "loss": 0.0235, "step": 65760 }, { "grad_norm": 0.42805787920951843, "learning_rate": 2.8757425831148134e-05, "loss": 0.0184, "step": 65770 }, { "grad_norm": 0.3562764525413513, "learning_rate": 2.874245874139658e-05, "loss": 0.0175, "step": 65780 }, { "grad_norm": 0.6563865542411804, "learning_rate": 2.8727493976337737e-05, "loss": 0.0286, "step": 65790 }, { "grad_norm": 0.3896459639072418, "learning_rate": 2.8712531537608218e-05, "loss": 0.0226, "step": 65800 }, { "grad_norm": 0.4491249620914459, "learning_rate": 2.8697571426844217e-05, "loss": 0.0333, "step": 65810 }, { "grad_norm": 0.4206230938434601, "learning_rate": 2.86826136456818e-05, "loss": 0.0257, "step": 65820 }, { "grad_norm": 0.6771740913391113, "learning_rate": 2.8667658195756685e-05, "loss": 0.0221, "step": 65830 }, { "grad_norm": 0.5155565142631531, "learning_rate": 2.8652705078704412e-05, "loss": 0.0366, "step": 65840 }, { "grad_norm": 0.3530409038066864, "learning_rate": 2.863775429616018e-05, "loss": 0.0268, "step": 65850 }, { "grad_norm": 0.27949368953704834, "learning_rate": 2.862280584975906e-05, "loss": 0.0227, "step": 65860 }, { "grad_norm": 0.4340110421180725, "learning_rate": 2.8607859741135735e-05, "loss": 0.0242, "step": 65870 }, { "grad_norm": 0.28572338819503784, "learning_rate": 2.8592915971924723e-05, "loss": 0.0226, "step": 65880 }, { "grad_norm": 0.7439230680465698, "learning_rate": 2.8577974543760212e-05, "loss": 0.0289, "step": 65890 }, { "grad_norm": 0.39529529213905334, "learning_rate": 2.8563035458276222e-05, "loss": 0.0214, "step": 65900 }, { "grad_norm": 0.48259037733078003, "learning_rate": 2.8548098717106424e-05, "loss": 0.0177, "step": 65910 }, { "grad_norm": 0.3674372732639313, "learning_rate": 2.85331643218843e-05, "loss": 0.0169, "step": 65920 }, { "grad_norm": 0.43455371260643005, "learning_rate": 2.8518232274243046e-05, "loss": 0.0249, "step": 65930 }, { "grad_norm": 1.399378776550293, "learning_rate": 2.8503302575815638e-05, "loss": 0.0228, "step": 65940 }, { "grad_norm": 0.5164905786514282, "learning_rate": 2.8488375228234716e-05, "loss": 0.0203, "step": 65950 }, { "grad_norm": 0.34993401169776917, "learning_rate": 2.8473450233132738e-05, "loss": 0.0256, "step": 65960 }, { "grad_norm": 0.24405786395072937, "learning_rate": 2.8458527592141898e-05, "loss": 0.0263, "step": 65970 }, { "grad_norm": 0.6820975542068481, "learning_rate": 2.8443607306894073e-05, "loss": 0.0191, "step": 65980 }, { "grad_norm": 0.44648462533950806, "learning_rate": 2.8428689379020946e-05, "loss": 0.0241, "step": 65990 }, { "grad_norm": 0.6061809062957764, "learning_rate": 2.841377381015391e-05, "loss": 0.0247, "step": 66000 }, { "grad_norm": 0.3255670666694641, "learning_rate": 2.8398860601924144e-05, "loss": 0.0206, "step": 66010 }, { "grad_norm": 0.28087323904037476, "learning_rate": 2.838394975596248e-05, "loss": 0.0316, "step": 66020 }, { "grad_norm": 0.2632646858692169, "learning_rate": 2.83690412738996e-05, "loss": 0.0276, "step": 66030 }, { "grad_norm": 0.44974493980407715, "learning_rate": 2.835413515736579e-05, "loss": 0.0295, "step": 66040 }, { "grad_norm": 0.40213778614997864, "learning_rate": 2.8339231407991267e-05, "loss": 0.0358, "step": 66050 }, { "grad_norm": 0.40450865030288696, "learning_rate": 2.832433002740581e-05, "loss": 0.0212, "step": 66060 }, { "grad_norm": 0.518873393535614, "learning_rate": 2.8309431017239047e-05, "loss": 0.0226, "step": 66070 }, { "grad_norm": 0.43748655915260315, "learning_rate": 2.829453437912028e-05, "loss": 0.0207, "step": 66080 }, { "grad_norm": 0.3845464289188385, "learning_rate": 2.8279640114678617e-05, "loss": 0.0321, "step": 66090 }, { "grad_norm": 1.2134504318237305, "learning_rate": 2.8264748225542836e-05, "loss": 0.035, "step": 66100 }, { "grad_norm": 0.1968330591917038, "learning_rate": 2.8249858713341516e-05, "loss": 0.0281, "step": 66110 }, { "grad_norm": 0.3841915428638458, "learning_rate": 2.823497157970294e-05, "loss": 0.0296, "step": 66120 }, { "grad_norm": 0.47020748257637024, "learning_rate": 2.8220086826255164e-05, "loss": 0.029, "step": 66130 }, { "grad_norm": 0.34335359930992126, "learning_rate": 2.8205204454625934e-05, "loss": 0.026, "step": 66140 }, { "grad_norm": 0.4135655164718628, "learning_rate": 2.8190324466442798e-05, "loss": 0.0265, "step": 66150 }, { "grad_norm": 0.3827632665634155, "learning_rate": 2.8175446863332965e-05, "loss": 0.0475, "step": 66160 }, { "grad_norm": 0.26395168900489807, "learning_rate": 2.816057164692345e-05, "loss": 0.0241, "step": 66170 }, { "grad_norm": 0.31834521889686584, "learning_rate": 2.8145698818840992e-05, "loss": 0.0268, "step": 66180 }, { "grad_norm": 0.33816957473754883, "learning_rate": 2.8130828380712065e-05, "loss": 0.0487, "step": 66190 }, { "grad_norm": 0.5126754641532898, "learning_rate": 2.811596033416285e-05, "loss": 0.0255, "step": 66200 }, { "grad_norm": 0.685847282409668, "learning_rate": 2.8101094680819307e-05, "loss": 0.027, "step": 66210 }, { "grad_norm": 0.3959180414676666, "learning_rate": 2.808623142230715e-05, "loss": 0.0279, "step": 66220 }, { "grad_norm": 0.2769574522972107, "learning_rate": 2.8071370560251757e-05, "loss": 0.0302, "step": 66230 }, { "grad_norm": 0.2742308974266052, "learning_rate": 2.805651209627831e-05, "loss": 0.0239, "step": 66240 }, { "grad_norm": 0.28090792894363403, "learning_rate": 2.804165603201171e-05, "loss": 0.0304, "step": 66250 }, { "grad_norm": 0.3821337819099426, "learning_rate": 2.802680236907661e-05, "loss": 0.023, "step": 66260 }, { "grad_norm": 0.32320818305015564, "learning_rate": 2.8011951109097344e-05, "loss": 0.0365, "step": 66270 }, { "grad_norm": 0.4646936357021332, "learning_rate": 2.7997102253698072e-05, "loss": 0.027, "step": 66280 }, { "grad_norm": 0.4256584048271179, "learning_rate": 2.7982255804502588e-05, "loss": 0.0208, "step": 66290 }, { "grad_norm": 0.6243117451667786, "learning_rate": 2.7967411763134516e-05, "loss": 0.0209, "step": 66300 }, { "grad_norm": 0.484075129032135, "learning_rate": 2.7952570131217166e-05, "loss": 0.0258, "step": 66310 }, { "grad_norm": 0.3552759289741516, "learning_rate": 2.7937730910373618e-05, "loss": 0.025, "step": 66320 }, { "grad_norm": 0.43171778321266174, "learning_rate": 2.7922894102226628e-05, "loss": 0.0184, "step": 66330 }, { "grad_norm": 0.33439671993255615, "learning_rate": 2.790805970839876e-05, "loss": 0.0293, "step": 66340 }, { "grad_norm": 0.3996087610721588, "learning_rate": 2.789322773051225e-05, "loss": 0.0262, "step": 66350 }, { "grad_norm": 0.6477504968643188, "learning_rate": 2.7878398170189124e-05, "loss": 0.0271, "step": 66360 }, { "grad_norm": 0.6785930395126343, "learning_rate": 2.7863571029051112e-05, "loss": 0.0324, "step": 66370 }, { "grad_norm": 0.374799907207489, "learning_rate": 2.784874630871971e-05, "loss": 0.0162, "step": 66380 }, { "grad_norm": 0.402656614780426, "learning_rate": 2.7833924010816086e-05, "loss": 0.0214, "step": 66390 }, { "grad_norm": 0.33364930748939514, "learning_rate": 2.7819104136961226e-05, "loss": 0.0169, "step": 66400 }, { "grad_norm": 0.45369893312454224, "learning_rate": 2.780428668877577e-05, "loss": 0.0262, "step": 66410 }, { "grad_norm": 0.37213560938835144, "learning_rate": 2.7789471667880152e-05, "loss": 0.0255, "step": 66420 }, { "grad_norm": 0.9657806754112244, "learning_rate": 2.7774659075894523e-05, "loss": 0.0362, "step": 66430 }, { "grad_norm": 0.4158603847026825, "learning_rate": 2.7759848914438778e-05, "loss": 0.0191, "step": 66440 }, { "grad_norm": 0.8535493612289429, "learning_rate": 2.7745041185132496e-05, "loss": 0.0591, "step": 66450 }, { "grad_norm": 0.591077446937561, "learning_rate": 2.773023588959506e-05, "loss": 0.0279, "step": 66460 }, { "grad_norm": 0.4723769426345825, "learning_rate": 2.7715433029445555e-05, "loss": 0.0221, "step": 66470 }, { "grad_norm": 0.22894124686717987, "learning_rate": 2.7700632606302766e-05, "loss": 0.0311, "step": 66480 }, { "grad_norm": 0.42811504006385803, "learning_rate": 2.768583462178528e-05, "loss": 0.0254, "step": 66490 }, { "grad_norm": 0.5917727947235107, "learning_rate": 2.767103907751136e-05, "loss": 0.0337, "step": 66500 }, { "grad_norm": 0.2981339395046234, "learning_rate": 2.7656245975099065e-05, "loss": 0.0373, "step": 66510 }, { "grad_norm": 0.39831072092056274, "learning_rate": 2.7641455316166093e-05, "loss": 0.0266, "step": 66520 }, { "grad_norm": 0.48283469676971436, "learning_rate": 2.7626667102329973e-05, "loss": 0.0402, "step": 66530 }, { "grad_norm": 0.4972079396247864, "learning_rate": 2.761188133520788e-05, "loss": 0.0282, "step": 66540 }, { "grad_norm": 0.4034316837787628, "learning_rate": 2.7597098016416784e-05, "loss": 0.0235, "step": 66550 }, { "grad_norm": 0.45366960763931274, "learning_rate": 2.7582317147573368e-05, "loss": 0.032, "step": 66560 }, { "grad_norm": 0.4837288558483124, "learning_rate": 2.7567538730294064e-05, "loss": 0.0251, "step": 66570 }, { "grad_norm": 0.3260473608970642, "learning_rate": 2.7552762766194973e-05, "loss": 0.0235, "step": 66580 }, { "grad_norm": 0.237508162856102, "learning_rate": 2.753798925689202e-05, "loss": 0.0182, "step": 66590 }, { "grad_norm": 0.4322165548801422, "learning_rate": 2.7523218204000774e-05, "loss": 0.0293, "step": 66600 }, { "grad_norm": 0.48230642080307007, "learning_rate": 2.7508449609136612e-05, "loss": 0.0295, "step": 66610 }, { "grad_norm": 0.47780442237854004, "learning_rate": 2.749368347391454e-05, "loss": 0.0259, "step": 66620 }, { "grad_norm": 0.4174886643886566, "learning_rate": 2.7478919799949443e-05, "loss": 0.0188, "step": 66630 }, { "grad_norm": 0.4154300093650818, "learning_rate": 2.7464158588855803e-05, "loss": 0.0256, "step": 66640 }, { "grad_norm": 0.7834829092025757, "learning_rate": 2.7449399842247918e-05, "loss": 0.0225, "step": 66650 }, { "grad_norm": 0.3862415552139282, "learning_rate": 2.7434643561739736e-05, "loss": 0.028, "step": 66660 }, { "grad_norm": 0.5620108842849731, "learning_rate": 2.7419889748945016e-05, "loss": 0.0293, "step": 66670 }, { "grad_norm": 0.3075421154499054, "learning_rate": 2.7405138405477203e-05, "loss": 0.0385, "step": 66680 }, { "grad_norm": 0.44637125730514526, "learning_rate": 2.73903895329495e-05, "loss": 0.0241, "step": 66690 }, { "grad_norm": 0.5069699287414551, "learning_rate": 2.737564313297478e-05, "loss": 0.0276, "step": 66700 }, { "grad_norm": 0.42736130952835083, "learning_rate": 2.736089920716572e-05, "loss": 0.021, "step": 66710 }, { "grad_norm": 0.4050745964050293, "learning_rate": 2.7346157757134706e-05, "loss": 0.0224, "step": 66720 }, { "grad_norm": 0.3796590268611908, "learning_rate": 2.733141878449379e-05, "loss": 0.0302, "step": 66730 }, { "grad_norm": 0.596798837184906, "learning_rate": 2.7316682290854844e-05, "loss": 0.0247, "step": 66740 }, { "grad_norm": 0.20111769437789917, "learning_rate": 2.7301948277829414e-05, "loss": 0.0303, "step": 66750 }, { "grad_norm": 0.676153302192688, "learning_rate": 2.7287216747028814e-05, "loss": 0.0262, "step": 66760 }, { "grad_norm": 0.44681790471076965, "learning_rate": 2.7272487700064025e-05, "loss": 0.0374, "step": 66770 }, { "grad_norm": 0.4836429953575134, "learning_rate": 2.725776113854583e-05, "loss": 0.0264, "step": 66780 }, { "grad_norm": 0.37187936902046204, "learning_rate": 2.7243037064084664e-05, "loss": 0.0231, "step": 66790 }, { "grad_norm": 0.3065125048160553, "learning_rate": 2.7228315478290766e-05, "loss": 0.0211, "step": 66800 }, { "grad_norm": 0.5113776326179504, "learning_rate": 2.721359638277402e-05, "loss": 0.0205, "step": 66810 }, { "grad_norm": 0.42574265599250793, "learning_rate": 2.7198879779144148e-05, "loss": 0.0358, "step": 66820 }, { "grad_norm": 0.4575793147087097, "learning_rate": 2.7184165669010485e-05, "loss": 0.0256, "step": 66830 }, { "grad_norm": 0.6485100388526917, "learning_rate": 2.716945405398218e-05, "loss": 0.0217, "step": 66840 }, { "grad_norm": 0.3989992141723633, "learning_rate": 2.715474493566803e-05, "loss": 0.0254, "step": 66850 }, { "grad_norm": 0.408457487821579, "learning_rate": 2.7140038315676652e-05, "loss": 0.0277, "step": 66860 }, { "grad_norm": 0.4449586868286133, "learning_rate": 2.7125334195616275e-05, "loss": 0.0248, "step": 66870 }, { "grad_norm": 0.3069518506526947, "learning_rate": 2.7110632577094997e-05, "loss": 0.023, "step": 66880 }, { "grad_norm": 0.3391364812850952, "learning_rate": 2.709593346172051e-05, "loss": 0.0326, "step": 66890 }, { "grad_norm": 0.31183841824531555, "learning_rate": 2.708123685110032e-05, "loss": 0.0212, "step": 66900 }, { "grad_norm": 0.6499873995780945, "learning_rate": 2.7066542746841588e-05, "loss": 0.0241, "step": 66910 }, { "grad_norm": 0.4109303057193756, "learning_rate": 2.7051851150551267e-05, "loss": 0.0207, "step": 66920 }, { "grad_norm": 0.3518022298812866, "learning_rate": 2.7037162063835996e-05, "loss": 0.0231, "step": 66930 }, { "grad_norm": 0.5281900763511658, "learning_rate": 2.702247548830218e-05, "loss": 0.0172, "step": 66940 }, { "grad_norm": 0.15780536830425262, "learning_rate": 2.7007791425555883e-05, "loss": 0.0163, "step": 66950 }, { "grad_norm": 0.4357612431049347, "learning_rate": 2.6993109877202945e-05, "loss": 0.0252, "step": 66960 }, { "grad_norm": 0.5022703409194946, "learning_rate": 2.6978430844848935e-05, "loss": 0.0258, "step": 66970 }, { "grad_norm": 0.3403777480125427, "learning_rate": 2.69637543300991e-05, "loss": 0.0245, "step": 66980 }, { "grad_norm": 0.4188486337661743, "learning_rate": 2.694908033455848e-05, "loss": 0.0298, "step": 66990 }, { "grad_norm": 0.5251174569129944, "learning_rate": 2.6934408859831733e-05, "loss": 0.0323, "step": 67000 }, { "grad_norm": 0.7426025867462158, "learning_rate": 2.69197399075234e-05, "loss": 0.0417, "step": 67010 }, { "grad_norm": 0.3432447016239166, "learning_rate": 2.6905073479237584e-05, "loss": 0.0261, "step": 67020 }, { "grad_norm": 0.5348137617111206, "learning_rate": 2.6890409576578235e-05, "loss": 0.0348, "step": 67030 }, { "grad_norm": 0.32771191000938416, "learning_rate": 2.687574820114892e-05, "loss": 0.0291, "step": 67040 }, { "grad_norm": 0.627467930316925, "learning_rate": 2.686108935455305e-05, "loss": 0.0269, "step": 67050 }, { "grad_norm": 0.33939388394355774, "learning_rate": 2.6846433038393603e-05, "loss": 0.033, "step": 67060 }, { "grad_norm": 0.3970128893852234, "learning_rate": 2.6831779254273482e-05, "loss": 0.0295, "step": 67070 }, { "grad_norm": 0.48303502798080444, "learning_rate": 2.6817128003795122e-05, "loss": 0.0216, "step": 67080 }, { "grad_norm": 0.6615509390830994, "learning_rate": 2.680247928856081e-05, "loss": 0.0338, "step": 67090 }, { "grad_norm": 0.4437271058559418, "learning_rate": 2.6787833110172466e-05, "loss": 0.0248, "step": 67100 }, { "grad_norm": 1.3843239545822144, "learning_rate": 2.6773189470231807e-05, "loss": 0.0284, "step": 67110 }, { "grad_norm": 0.35172513127326965, "learning_rate": 2.675854837034022e-05, "loss": 0.0278, "step": 67120 }, { "grad_norm": 0.5436289310455322, "learning_rate": 2.6743909812098833e-05, "loss": 0.0215, "step": 67130 }, { "grad_norm": 0.4184907376766205, "learning_rate": 2.67292737971085e-05, "loss": 0.0213, "step": 67140 }, { "grad_norm": 0.22676603496074677, "learning_rate": 2.671464032696982e-05, "loss": 0.0177, "step": 67150 }, { "grad_norm": 0.15222907066345215, "learning_rate": 2.6700009403283055e-05, "loss": 0.0182, "step": 67160 }, { "grad_norm": 0.8840603828430176, "learning_rate": 2.6685381027648214e-05, "loss": 0.0278, "step": 67170 }, { "grad_norm": 1.2067688703536987, "learning_rate": 2.6670755201665086e-05, "loss": 0.0283, "step": 67180 }, { "grad_norm": 0.4895241856575012, "learning_rate": 2.6656131926933066e-05, "loss": 0.0335, "step": 67190 }, { "grad_norm": 0.424333781003952, "learning_rate": 2.664151120505136e-05, "loss": 0.0247, "step": 67200 }, { "grad_norm": 0.3507438600063324, "learning_rate": 2.6626893037618873e-05, "loss": 0.0376, "step": 67210 }, { "grad_norm": 0.4722951650619507, "learning_rate": 2.6612277426234243e-05, "loss": 0.0265, "step": 67220 }, { "grad_norm": 0.5142638087272644, "learning_rate": 2.659766437249577e-05, "loss": 0.0257, "step": 67230 }, { "grad_norm": 0.4877399504184723, "learning_rate": 2.6583053878001557e-05, "loss": 0.0202, "step": 67240 }, { "grad_norm": 0.31045109033584595, "learning_rate": 2.6568445944349323e-05, "loss": 0.026, "step": 67250 }, { "grad_norm": 0.35463306307792664, "learning_rate": 2.6553840573136657e-05, "loss": 0.0227, "step": 67260 }, { "grad_norm": 0.46052032709121704, "learning_rate": 2.6539237765960706e-05, "loss": 0.0202, "step": 67270 }, { "grad_norm": 0.730862021446228, "learning_rate": 2.6524637524418465e-05, "loss": 0.0276, "step": 67280 }, { "grad_norm": 0.5021882057189941, "learning_rate": 2.651003985010655e-05, "loss": 0.0266, "step": 67290 }, { "grad_norm": 0.2372298687696457, "learning_rate": 2.649544474462138e-05, "loss": 0.0217, "step": 67300 }, { "grad_norm": 0.4615952968597412, "learning_rate": 2.6480852209559015e-05, "loss": 0.0254, "step": 67310 }, { "grad_norm": 0.2641162574291229, "learning_rate": 2.6466262246515296e-05, "loss": 0.0181, "step": 67320 }, { "grad_norm": 0.5035140514373779, "learning_rate": 2.6451674857085746e-05, "loss": 0.0344, "step": 67330 }, { "grad_norm": 0.41325390338897705, "learning_rate": 2.6437090042865655e-05, "loss": 0.03, "step": 67340 }, { "grad_norm": 0.2953380346298218, "learning_rate": 2.6422507805449947e-05, "loss": 0.0239, "step": 67350 }, { "grad_norm": 0.49827003479003906, "learning_rate": 2.6407928146433357e-05, "loss": 0.0191, "step": 67360 }, { "grad_norm": 0.3452047109603882, "learning_rate": 2.639335106741026e-05, "loss": 0.0243, "step": 67370 }, { "grad_norm": 0.24726513028144836, "learning_rate": 2.6378776569974796e-05, "loss": 0.0253, "step": 67380 }, { "grad_norm": 0.725450873374939, "learning_rate": 2.6364204655720814e-05, "loss": 0.0213, "step": 67390 }, { "grad_norm": 0.25394394993782043, "learning_rate": 2.6349635326241894e-05, "loss": 0.0338, "step": 67400 }, { "grad_norm": 0.3398728370666504, "learning_rate": 2.633506858313127e-05, "loss": 0.0321, "step": 67410 }, { "grad_norm": 0.5077832341194153, "learning_rate": 2.6320504427981975e-05, "loss": 0.0183, "step": 67420 }, { "grad_norm": 0.4317280352115631, "learning_rate": 2.6305942862386734e-05, "loss": 0.033, "step": 67430 }, { "grad_norm": 0.369964599609375, "learning_rate": 2.629138388793793e-05, "loss": 0.021, "step": 67440 }, { "grad_norm": 0.41788628697395325, "learning_rate": 2.6276827506227743e-05, "loss": 0.0242, "step": 67450 }, { "grad_norm": 0.25304940342903137, "learning_rate": 2.6262273718848037e-05, "loss": 0.0252, "step": 67460 }, { "grad_norm": 0.3651926517486572, "learning_rate": 2.6247722527390407e-05, "loss": 0.0213, "step": 67470 }, { "grad_norm": 0.25029584765434265, "learning_rate": 2.6233173933446104e-05, "loss": 0.031, "step": 67480 }, { "grad_norm": 0.45847731828689575, "learning_rate": 2.6218627938606188e-05, "loss": 0.0319, "step": 67490 }, { "grad_norm": 0.18632522225379944, "learning_rate": 2.620408454446135e-05, "loss": 0.0124, "step": 67500 }, { "grad_norm": 0.3092445433139801, "learning_rate": 2.6189543752602046e-05, "loss": 0.0229, "step": 67510 }, { "grad_norm": 0.52432781457901, "learning_rate": 2.6175005564618444e-05, "loss": 0.0351, "step": 67520 }, { "grad_norm": 0.1916709691286087, "learning_rate": 2.6160469982100428e-05, "loss": 0.0264, "step": 67530 }, { "grad_norm": 0.34782660007476807, "learning_rate": 2.6145937006637555e-05, "loss": 0.0278, "step": 67540 }, { "grad_norm": 0.5828356742858887, "learning_rate": 2.6131406639819166e-05, "loss": 0.0263, "step": 67550 }, { "grad_norm": 0.5024917125701904, "learning_rate": 2.611687888323424e-05, "loss": 0.0235, "step": 67560 }, { "grad_norm": 0.2950228154659271, "learning_rate": 2.610235373847153e-05, "loss": 0.0214, "step": 67570 }, { "grad_norm": 0.5397739410400391, "learning_rate": 2.6087831207119494e-05, "loss": 0.0269, "step": 67580 }, { "grad_norm": 0.27428650856018066, "learning_rate": 2.60733112907663e-05, "loss": 0.0286, "step": 67590 }, { "grad_norm": 0.5717750787734985, "learning_rate": 2.6058793990999786e-05, "loss": 0.0249, "step": 67600 }, { "grad_norm": 0.23670953512191772, "learning_rate": 2.6044279309407593e-05, "loss": 0.0213, "step": 67610 }, { "grad_norm": 0.5316188335418701, "learning_rate": 2.6029767247576976e-05, "loss": 0.0328, "step": 67620 }, { "grad_norm": 0.23837974667549133, "learning_rate": 2.601525780709497e-05, "loss": 0.0208, "step": 67630 }, { "grad_norm": 0.27643248438835144, "learning_rate": 2.6000750989548318e-05, "loss": 0.0244, "step": 67640 }, { "grad_norm": 0.3227013945579529, "learning_rate": 2.5986246796523462e-05, "loss": 0.023, "step": 67650 }, { "grad_norm": 0.3686589002609253, "learning_rate": 2.5971745229606546e-05, "loss": 0.0226, "step": 67660 }, { "grad_norm": 0.46998873353004456, "learning_rate": 2.5957246290383442e-05, "loss": 0.0328, "step": 67670 }, { "grad_norm": 0.36110401153564453, "learning_rate": 2.594274998043975e-05, "loss": 0.0255, "step": 67680 }, { "grad_norm": 0.4816683828830719, "learning_rate": 2.5928256301360733e-05, "loss": 0.0233, "step": 67690 }, { "grad_norm": 0.9204339981079102, "learning_rate": 2.591376525473142e-05, "loss": 0.0281, "step": 67700 }, { "grad_norm": 0.27840033173561096, "learning_rate": 2.5899276842136523e-05, "loss": 0.0276, "step": 67710 }, { "grad_norm": 0.6652148365974426, "learning_rate": 2.5884791065160495e-05, "loss": 0.023, "step": 67720 }, { "grad_norm": 0.5789492130279541, "learning_rate": 2.5870307925387436e-05, "loss": 0.0199, "step": 67730 }, { "grad_norm": 0.37155893445014954, "learning_rate": 2.585582742440125e-05, "loss": 0.0242, "step": 67740 }, { "grad_norm": 0.5314699411392212, "learning_rate": 2.5841349563785455e-05, "loss": 0.024, "step": 67750 }, { "grad_norm": 0.47963428497314453, "learning_rate": 2.5826874345123355e-05, "loss": 0.0275, "step": 67760 }, { "grad_norm": 0.35520368814468384, "learning_rate": 2.581240176999792e-05, "loss": 0.0197, "step": 67770 }, { "grad_norm": 0.41593101620674133, "learning_rate": 2.579793183999189e-05, "loss": 0.0302, "step": 67780 }, { "grad_norm": 0.4851226508617401, "learning_rate": 2.5783464556687627e-05, "loss": 0.023, "step": 67790 }, { "grad_norm": 0.3390204608440399, "learning_rate": 2.576899992166729e-05, "loss": 0.0216, "step": 67800 }, { "grad_norm": 0.6257338523864746, "learning_rate": 2.5754537936512668e-05, "loss": 0.019, "step": 67810 }, { "grad_norm": 0.5652239918708801, "learning_rate": 2.574007860280535e-05, "loss": 0.0325, "step": 67820 }, { "grad_norm": 0.3386628031730652, "learning_rate": 2.5725621922126518e-05, "loss": 0.0551, "step": 67830 }, { "grad_norm": 0.30987703800201416, "learning_rate": 2.571116789605722e-05, "loss": 0.0323, "step": 67840 }, { "grad_norm": 0.2975413501262665, "learning_rate": 2.5696716526178054e-05, "loss": 0.0241, "step": 67850 }, { "grad_norm": 0.36012527346611023, "learning_rate": 2.568226781406946e-05, "loss": 0.0197, "step": 67860 }, { "grad_norm": 0.40430670976638794, "learning_rate": 2.5667821761311474e-05, "loss": 0.0187, "step": 67870 }, { "grad_norm": 0.2453158050775528, "learning_rate": 2.5653378369483917e-05, "loss": 0.018, "step": 67880 }, { "grad_norm": 0.38402441143989563, "learning_rate": 2.5638937640166292e-05, "loss": 0.0485, "step": 67890 }, { "grad_norm": 0.4153466522693634, "learning_rate": 2.5624499574937842e-05, "loss": 0.0323, "step": 67900 }, { "grad_norm": 0.48405876755714417, "learning_rate": 2.5610064175377456e-05, "loss": 0.0233, "step": 67910 }, { "grad_norm": 0.4816855192184448, "learning_rate": 2.5595631443063782e-05, "loss": 0.0207, "step": 67920 }, { "grad_norm": 0.3962787985801697, "learning_rate": 2.5581201379575177e-05, "loss": 0.0258, "step": 67930 }, { "grad_norm": 3.504836082458496, "learning_rate": 2.5566773986489655e-05, "loss": 0.0179, "step": 67940 }, { "grad_norm": 0.2908460795879364, "learning_rate": 2.5552349265385e-05, "loss": 0.0205, "step": 67950 }, { "grad_norm": 0.28280842304229736, "learning_rate": 2.5537927217838675e-05, "loss": 0.0287, "step": 67960 }, { "grad_norm": 3.9707071781158447, "learning_rate": 2.5523507845427864e-05, "loss": 0.0332, "step": 67970 }, { "grad_norm": 0.49902909994125366, "learning_rate": 2.5509091149729413e-05, "loss": 0.0232, "step": 67980 }, { "grad_norm": 0.20311477780342102, "learning_rate": 2.5494677132319962e-05, "loss": 0.0244, "step": 67990 }, { "grad_norm": 0.4743984043598175, "learning_rate": 2.548026579477575e-05, "loss": 0.0212, "step": 68000 }, { "grad_norm": 0.32136011123657227, "learning_rate": 2.5465857138672834e-05, "loss": 0.0248, "step": 68010 }, { "grad_norm": 0.3773338794708252, "learning_rate": 2.5451451165586848e-05, "loss": 0.0368, "step": 68020 }, { "grad_norm": 0.5284589529037476, "learning_rate": 2.543704787709329e-05, "loss": 0.0229, "step": 68030 }, { "grad_norm": 0.33091557025909424, "learning_rate": 2.542264727476723e-05, "loss": 0.0247, "step": 68040 }, { "grad_norm": 0.4989772439002991, "learning_rate": 2.5408249360183535e-05, "loss": 0.0342, "step": 68050 }, { "grad_norm": 0.44751688838005066, "learning_rate": 2.5393854134916694e-05, "loss": 0.0263, "step": 68060 }, { "grad_norm": 0.4648325443267822, "learning_rate": 2.5379461600540993e-05, "loss": 0.0305, "step": 68070 }, { "grad_norm": 0.3590047359466553, "learning_rate": 2.5365071758630317e-05, "loss": 0.0302, "step": 68080 }, { "grad_norm": 0.48028168082237244, "learning_rate": 2.535068461075839e-05, "loss": 0.0193, "step": 68090 }, { "grad_norm": 0.3190622329711914, "learning_rate": 2.5336300158498516e-05, "loss": 0.0179, "step": 68100 }, { "grad_norm": 0.275606632232666, "learning_rate": 2.5321918403423793e-05, "loss": 0.0191, "step": 68110 }, { "grad_norm": 0.7763818502426147, "learning_rate": 2.5307539347106957e-05, "loss": 0.0336, "step": 68120 }, { "grad_norm": 0.4731540381908417, "learning_rate": 2.529316299112049e-05, "loss": 0.028, "step": 68130 }, { "grad_norm": 0.24183174967765808, "learning_rate": 2.5278789337036567e-05, "loss": 0.0232, "step": 68140 }, { "grad_norm": 0.398458331823349, "learning_rate": 2.52644183864271e-05, "loss": 0.0208, "step": 68150 }, { "grad_norm": 0.24068784713745117, "learning_rate": 2.525005014086363e-05, "loss": 0.023, "step": 68160 }, { "grad_norm": 0.3579997718334198, "learning_rate": 2.5235684601917465e-05, "loss": 0.0194, "step": 68170 }, { "grad_norm": 0.7688282132148743, "learning_rate": 2.5221321771159618e-05, "loss": 0.0193, "step": 68180 }, { "grad_norm": 0.19767066836357117, "learning_rate": 2.5206961650160743e-05, "loss": 0.0311, "step": 68190 }, { "grad_norm": 0.2348909080028534, "learning_rate": 2.519260424049129e-05, "loss": 0.0285, "step": 68200 }, { "grad_norm": 0.4424133002758026, "learning_rate": 2.5178249543721304e-05, "loss": 0.021, "step": 68210 }, { "grad_norm": 0.2620998024940491, "learning_rate": 2.5163897561420668e-05, "loss": 0.0231, "step": 68220 }, { "grad_norm": 0.4665391445159912, "learning_rate": 2.514954829515883e-05, "loss": 0.0224, "step": 68230 }, { "grad_norm": 0.46823957562446594, "learning_rate": 2.513520174650505e-05, "loss": 0.0246, "step": 68240 }, { "grad_norm": 0.4010177254676819, "learning_rate": 2.5120857917028207e-05, "loss": 0.0188, "step": 68250 }, { "grad_norm": 0.39979979395866394, "learning_rate": 2.5106516808296953e-05, "loss": 0.0292, "step": 68260 }, { "grad_norm": 0.21884797513484955, "learning_rate": 2.5092178421879552e-05, "loss": 0.0157, "step": 68270 }, { "grad_norm": 0.43964776396751404, "learning_rate": 2.5077842759344116e-05, "loss": 0.0155, "step": 68280 }, { "grad_norm": 0.43612533807754517, "learning_rate": 2.5063509822258304e-05, "loss": 0.0189, "step": 68290 }, { "grad_norm": 0.49694696068763733, "learning_rate": 2.5049179612189576e-05, "loss": 0.029, "step": 68300 }, { "grad_norm": 0.29450953006744385, "learning_rate": 2.5034852130705043e-05, "loss": 0.0358, "step": 68310 }, { "grad_norm": 0.3778190612792969, "learning_rate": 2.5020527379371572e-05, "loss": 0.0202, "step": 68320 }, { "grad_norm": 0.6144751906394958, "learning_rate": 2.500620535975564e-05, "loss": 0.0339, "step": 68330 }, { "grad_norm": 0.5169597268104553, "learning_rate": 2.4991886073423515e-05, "loss": 0.0254, "step": 68340 }, { "grad_norm": 0.3654748797416687, "learning_rate": 2.4977569521941134e-05, "loss": 0.0239, "step": 68350 }, { "grad_norm": 0.5070598721504211, "learning_rate": 2.4963255706874143e-05, "loss": 0.0293, "step": 68360 }, { "grad_norm": 0.3946880102157593, "learning_rate": 2.4948944629787853e-05, "loss": 0.0314, "step": 68370 }, { "grad_norm": 0.5543596148490906, "learning_rate": 2.493463629224731e-05, "loss": 0.0258, "step": 68380 }, { "grad_norm": 0.6150648593902588, "learning_rate": 2.492033069581728e-05, "loss": 0.0271, "step": 68390 }, { "grad_norm": 0.24239513278007507, "learning_rate": 2.4906027842062137e-05, "loss": 0.0246, "step": 68400 }, { "grad_norm": 0.474803626537323, "learning_rate": 2.48917277325461e-05, "loss": 0.0236, "step": 68410 }, { "grad_norm": 0.44591814279556274, "learning_rate": 2.487743036883295e-05, "loss": 0.0266, "step": 68420 }, { "grad_norm": 0.36412155628204346, "learning_rate": 2.486313575248626e-05, "loss": 0.0229, "step": 68430 }, { "grad_norm": 1.067884087562561, "learning_rate": 2.4848843885069235e-05, "loss": 0.0267, "step": 68440 }, { "grad_norm": 0.2707211971282959, "learning_rate": 2.483455476814484e-05, "loss": 0.0259, "step": 68450 }, { "grad_norm": 0.27981996536254883, "learning_rate": 2.4820268403275672e-05, "loss": 0.0213, "step": 68460 }, { "grad_norm": 0.29172801971435547, "learning_rate": 2.480598479202413e-05, "loss": 0.023, "step": 68470 }, { "grad_norm": 0.3333549499511719, "learning_rate": 2.4791703935952193e-05, "loss": 0.018, "step": 68480 }, { "grad_norm": 0.23251241445541382, "learning_rate": 2.477742583662163e-05, "loss": 0.0177, "step": 68490 }, { "grad_norm": 0.24028927087783813, "learning_rate": 2.476315049559384e-05, "loss": 0.0212, "step": 68500 }, { "grad_norm": 0.36881375312805176, "learning_rate": 2.4748877914429987e-05, "loss": 0.0287, "step": 68510 }, { "grad_norm": 0.4349064826965332, "learning_rate": 2.4734608094690857e-05, "loss": 0.0215, "step": 68520 }, { "grad_norm": 0.4615863859653473, "learning_rate": 2.4720341037937006e-05, "loss": 0.0269, "step": 68530 }, { "grad_norm": 0.8958690762519836, "learning_rate": 2.4706076745728645e-05, "loss": 0.0346, "step": 68540 }, { "grad_norm": 0.2844848036766052, "learning_rate": 2.469181521962573e-05, "loss": 0.0293, "step": 68550 }, { "grad_norm": 0.4025944769382477, "learning_rate": 2.467755646118783e-05, "loss": 0.0229, "step": 68560 }, { "grad_norm": 0.4214291572570801, "learning_rate": 2.46633004719743e-05, "loss": 0.0235, "step": 68570 }, { "grad_norm": 0.5640984773635864, "learning_rate": 2.4649047253544123e-05, "loss": 0.0171, "step": 68580 }, { "grad_norm": 0.5505059957504272, "learning_rate": 2.4634796807456016e-05, "loss": 0.0287, "step": 68590 }, { "grad_norm": 0.18022289872169495, "learning_rate": 2.46205491352684e-05, "loss": 0.0181, "step": 68600 }, { "grad_norm": 0.25355732440948486, "learning_rate": 2.4606304238539375e-05, "loss": 0.0224, "step": 68610 }, { "grad_norm": 0.6905607581138611, "learning_rate": 2.4592062118826753e-05, "loss": 0.0265, "step": 68620 }, { "grad_norm": 0.5029525756835938, "learning_rate": 2.4577822777688004e-05, "loss": 0.0317, "step": 68630 }, { "grad_norm": 0.5920765399932861, "learning_rate": 2.4563586216680347e-05, "loss": 0.0193, "step": 68640 }, { "grad_norm": 0.6564638614654541, "learning_rate": 2.4549352437360622e-05, "loss": 0.0254, "step": 68650 }, { "grad_norm": 0.2926018238067627, "learning_rate": 2.4535121441285493e-05, "loss": 0.0273, "step": 68660 }, { "grad_norm": 0.1595536470413208, "learning_rate": 2.4520893230011172e-05, "loss": 0.0204, "step": 68670 }, { "grad_norm": 0.7658834457397461, "learning_rate": 2.450666780509368e-05, "loss": 0.0243, "step": 68680 }, { "grad_norm": 0.3949286937713623, "learning_rate": 2.4492445168088656e-05, "loss": 0.0197, "step": 68690 }, { "grad_norm": 0.29279112815856934, "learning_rate": 2.447822532055149e-05, "loss": 0.024, "step": 68700 }, { "grad_norm": 1.284899115562439, "learning_rate": 2.4464008264037212e-05, "loss": 0.0263, "step": 68710 }, { "grad_norm": 0.5297040343284607, "learning_rate": 2.4449794000100605e-05, "loss": 0.0227, "step": 68720 }, { "grad_norm": 0.5422325134277344, "learning_rate": 2.4435582530296108e-05, "loss": 0.0268, "step": 68730 }, { "grad_norm": 0.5080130696296692, "learning_rate": 2.4421373856177887e-05, "loss": 0.0286, "step": 68740 }, { "grad_norm": 0.25103959441185, "learning_rate": 2.440716797929975e-05, "loss": 0.0208, "step": 68750 }, { "grad_norm": 0.49960657954216003, "learning_rate": 2.439296490121526e-05, "loss": 0.0307, "step": 68760 }, { "grad_norm": 0.36692339181900024, "learning_rate": 2.4378764623477608e-05, "loss": 0.0223, "step": 68770 }, { "grad_norm": 0.33917951583862305, "learning_rate": 2.436456714763974e-05, "loss": 0.0232, "step": 68780 }, { "grad_norm": 0.33778002858161926, "learning_rate": 2.4350372475254262e-05, "loss": 0.022, "step": 68790 }, { "grad_norm": 0.3060073256492615, "learning_rate": 2.433618060787351e-05, "loss": 0.0239, "step": 68800 }, { "grad_norm": 0.4415440559387207, "learning_rate": 2.432199154704944e-05, "loss": 0.0181, "step": 68810 }, { "grad_norm": 0.376618891954422, "learning_rate": 2.430780529433377e-05, "loss": 0.0236, "step": 68820 }, { "grad_norm": 0.3849411904811859, "learning_rate": 2.42936218512779e-05, "loss": 0.0232, "step": 68830 }, { "grad_norm": 0.41877374053001404, "learning_rate": 2.427944121943288e-05, "loss": 0.0196, "step": 68840 }, { "grad_norm": 0.3749803900718689, "learning_rate": 2.4265263400349497e-05, "loss": 0.0241, "step": 68850 }, { "grad_norm": 0.4273979067802429, "learning_rate": 2.4251088395578213e-05, "loss": 0.0234, "step": 68860 }, { "grad_norm": 0.3917028307914734, "learning_rate": 2.4236916206669214e-05, "loss": 0.0327, "step": 68870 }, { "grad_norm": 0.42102158069610596, "learning_rate": 2.4222746835172304e-05, "loss": 0.0269, "step": 68880 }, { "grad_norm": 0.9677636623382568, "learning_rate": 2.4208580282637065e-05, "loss": 0.0461, "step": 68890 }, { "grad_norm": 0.2875545024871826, "learning_rate": 2.4194416550612687e-05, "loss": 0.0213, "step": 68900 }, { "grad_norm": 0.670371413230896, "learning_rate": 2.418025564064812e-05, "loss": 0.0285, "step": 68910 }, { "grad_norm": 0.3043742775917053, "learning_rate": 2.4166097554291978e-05, "loss": 0.0259, "step": 68920 }, { "grad_norm": 0.5389046669006348, "learning_rate": 2.4151942293092583e-05, "loss": 0.0323, "step": 68930 }, { "grad_norm": 0.5237337350845337, "learning_rate": 2.41377898585979e-05, "loss": 0.0298, "step": 68940 }, { "grad_norm": 0.244812473654747, "learning_rate": 2.4123640252355652e-05, "loss": 0.0256, "step": 68950 }, { "grad_norm": 0.2707364559173584, "learning_rate": 2.410949347591318e-05, "loss": 0.0267, "step": 68960 }, { "grad_norm": 1.8069958686828613, "learning_rate": 2.4095349530817578e-05, "loss": 0.0222, "step": 68970 }, { "grad_norm": 0.24032972753047943, "learning_rate": 2.4081208418615603e-05, "loss": 0.0184, "step": 68980 }, { "grad_norm": 1.4792072772979736, "learning_rate": 2.4067070140853725e-05, "loss": 0.0295, "step": 68990 }, { "grad_norm": 0.3165592551231384, "learning_rate": 2.405293469907805e-05, "loss": 0.0225, "step": 69000 }, { "grad_norm": 0.31191200017929077, "learning_rate": 2.4038802094834446e-05, "loss": 0.0267, "step": 69010 }, { "grad_norm": 0.4104122817516327, "learning_rate": 2.4024672329668395e-05, "loss": 0.0224, "step": 69020 }, { "grad_norm": 0.3622852861881256, "learning_rate": 2.4010545405125125e-05, "loss": 0.0222, "step": 69030 }, { "grad_norm": 0.23065313696861267, "learning_rate": 2.3996421322749528e-05, "loss": 0.0264, "step": 69040 }, { "grad_norm": 0.5639175176620483, "learning_rate": 2.3982300084086224e-05, "loss": 0.0293, "step": 69050 }, { "grad_norm": 0.41372939944267273, "learning_rate": 2.396818169067944e-05, "loss": 0.0293, "step": 69060 }, { "grad_norm": 0.4239443242549896, "learning_rate": 2.3954066144073178e-05, "loss": 0.0193, "step": 69070 }, { "grad_norm": 0.7219495177268982, "learning_rate": 2.39399534458111e-05, "loss": 0.0203, "step": 69080 }, { "grad_norm": 0.39663076400756836, "learning_rate": 2.392584359743651e-05, "loss": 0.0174, "step": 69090 }, { "grad_norm": 0.4662262499332428, "learning_rate": 2.3911736600492463e-05, "loss": 0.0229, "step": 69100 }, { "grad_norm": 0.2895304560661316, "learning_rate": 2.3897632456521672e-05, "loss": 0.02, "step": 69110 }, { "grad_norm": 0.2593982517719269, "learning_rate": 2.388353116706658e-05, "loss": 0.0155, "step": 69120 }, { "grad_norm": 0.32487308979034424, "learning_rate": 2.3869432733669232e-05, "loss": 0.0293, "step": 69130 }, { "grad_norm": 0.2831747233867645, "learning_rate": 2.385533715787145e-05, "loss": 0.0267, "step": 69140 }, { "grad_norm": 0.5613048076629639, "learning_rate": 2.3841244441214673e-05, "loss": 0.0245, "step": 69150 }, { "grad_norm": 0.4802631139755249, "learning_rate": 2.3827154585240073e-05, "loss": 0.0243, "step": 69160 }, { "grad_norm": 0.2401294708251953, "learning_rate": 2.3813067591488498e-05, "loss": 0.0253, "step": 69170 }, { "grad_norm": 0.3536181151866913, "learning_rate": 2.3798983461500494e-05, "loss": 0.0142, "step": 69180 }, { "grad_norm": 0.26678186655044556, "learning_rate": 2.3784902196816256e-05, "loss": 0.0404, "step": 69190 }, { "grad_norm": 0.7847511172294617, "learning_rate": 2.3770823798975716e-05, "loss": 0.0263, "step": 69200 }, { "grad_norm": 0.7121303677558899, "learning_rate": 2.3756748269518424e-05, "loss": 0.0242, "step": 69210 }, { "grad_norm": 0.30576595664024353, "learning_rate": 2.374267560998371e-05, "loss": 0.0301, "step": 69220 }, { "grad_norm": 0.4135911166667938, "learning_rate": 2.3728605821910477e-05, "loss": 0.024, "step": 69230 }, { "grad_norm": 0.4364771842956543, "learning_rate": 2.3714538906837453e-05, "loss": 0.0232, "step": 69240 }, { "grad_norm": 0.5084584951400757, "learning_rate": 2.3700474866302913e-05, "loss": 0.0237, "step": 69250 }, { "grad_norm": 0.3294185400009155, "learning_rate": 2.3686413701844917e-05, "loss": 0.0265, "step": 69260 }, { "grad_norm": 0.36997315287590027, "learning_rate": 2.367235541500114e-05, "loss": 0.0219, "step": 69270 }, { "grad_norm": 0.5583509802818298, "learning_rate": 2.3658300007308993e-05, "loss": 0.0283, "step": 69280 }, { "grad_norm": 1.0447500944137573, "learning_rate": 2.364424748030555e-05, "loss": 0.0308, "step": 69290 }, { "grad_norm": 0.326799601316452, "learning_rate": 2.3630197835527595e-05, "loss": 0.0209, "step": 69300 }, { "grad_norm": 0.7437719106674194, "learning_rate": 2.3616151074511546e-05, "loss": 0.0191, "step": 69310 }, { "grad_norm": 0.34660258889198303, "learning_rate": 2.3602107198793545e-05, "loss": 0.0199, "step": 69320 }, { "grad_norm": 0.4756001830101013, "learning_rate": 2.3588066209909427e-05, "loss": 0.0218, "step": 69330 }, { "grad_norm": 0.41319555044174194, "learning_rate": 2.3574028109394662e-05, "loss": 0.0277, "step": 69340 }, { "grad_norm": 0.1913154125213623, "learning_rate": 2.3559992898784454e-05, "loss": 0.0224, "step": 69350 }, { "grad_norm": 0.4912567734718323, "learning_rate": 2.354596057961367e-05, "loss": 0.0305, "step": 69360 }, { "grad_norm": 0.204936221241951, "learning_rate": 2.353193115341688e-05, "loss": 0.02, "step": 69370 }, { "grad_norm": 0.15427252650260925, "learning_rate": 2.3517904621728294e-05, "loss": 0.0274, "step": 69380 }, { "grad_norm": 0.6746392846107483, "learning_rate": 2.3503880986081855e-05, "loss": 0.0256, "step": 69390 }, { "grad_norm": 0.32100576162338257, "learning_rate": 2.3489860248011136e-05, "loss": 0.0248, "step": 69400 }, { "grad_norm": 0.4132634401321411, "learning_rate": 2.3475842409049464e-05, "loss": 0.0221, "step": 69410 }, { "grad_norm": 0.30447396636009216, "learning_rate": 2.346182747072975e-05, "loss": 0.023, "step": 69420 }, { "grad_norm": 0.5884842276573181, "learning_rate": 2.3447815434584718e-05, "loss": 0.0393, "step": 69430 }, { "grad_norm": 0.3042258620262146, "learning_rate": 2.3433806302146656e-05, "loss": 0.0286, "step": 69440 }, { "grad_norm": 0.27471962571144104, "learning_rate": 2.3419800074947612e-05, "loss": 0.0207, "step": 69450 }, { "grad_norm": 0.3865518569946289, "learning_rate": 2.3405796754519248e-05, "loss": 0.0194, "step": 69460 }, { "grad_norm": 0.2834334969520569, "learning_rate": 2.3391796342392992e-05, "loss": 0.0183, "step": 69470 }, { "grad_norm": 0.30985531210899353, "learning_rate": 2.337779884009984e-05, "loss": 0.0238, "step": 69480 }, { "grad_norm": 0.4151986539363861, "learning_rate": 2.3363804249170624e-05, "loss": 0.025, "step": 69490 }, { "grad_norm": 0.5487521290779114, "learning_rate": 2.3349812571135703e-05, "loss": 0.0273, "step": 69500 }, { "grad_norm": 0.40987682342529297, "learning_rate": 2.333582380752523e-05, "loss": 0.0318, "step": 69510 }, { "grad_norm": 0.3753816783428192, "learning_rate": 2.3321837959868963e-05, "loss": 0.0281, "step": 69520 }, { "grad_norm": 0.49567562341690063, "learning_rate": 2.3307855029696383e-05, "loss": 0.0271, "step": 69530 }, { "grad_norm": 0.5255765914916992, "learning_rate": 2.3293875018536648e-05, "loss": 0.0286, "step": 69540 }, { "grad_norm": 0.33068621158599854, "learning_rate": 2.327989792791861e-05, "loss": 0.0385, "step": 69550 }, { "grad_norm": 0.14268194139003754, "learning_rate": 2.326592375937074e-05, "loss": 0.0301, "step": 69560 }, { "grad_norm": 0.4564366042613983, "learning_rate": 2.3251952514421248e-05, "loss": 0.0213, "step": 69570 }, { "grad_norm": 0.36404892802238464, "learning_rate": 2.323798419459804e-05, "loss": 0.0267, "step": 69580 }, { "grad_norm": 0.8315600752830505, "learning_rate": 2.3224018801428625e-05, "loss": 0.0245, "step": 69590 }, { "grad_norm": 0.38101714849472046, "learning_rate": 2.3210056336440273e-05, "loss": 0.0158, "step": 69600 }, { "grad_norm": 0.24494297802448273, "learning_rate": 2.3196096801159844e-05, "loss": 0.0227, "step": 69610 }, { "grad_norm": 0.4223444163799286, "learning_rate": 2.3182140197114012e-05, "loss": 0.0223, "step": 69620 }, { "grad_norm": 0.2653914988040924, "learning_rate": 2.3168186525828985e-05, "loss": 0.0264, "step": 69630 }, { "grad_norm": 0.40244635939598083, "learning_rate": 2.3154235788830757e-05, "loss": 0.021, "step": 69640 }, { "grad_norm": 0.3035300374031067, "learning_rate": 2.3140287987644927e-05, "loss": 0.0217, "step": 69650 }, { "grad_norm": 0.8822260499000549, "learning_rate": 2.3126343123796833e-05, "loss": 0.0193, "step": 69660 }, { "grad_norm": 0.42602482438087463, "learning_rate": 2.3112401198811418e-05, "loss": 0.02, "step": 69670 }, { "grad_norm": 0.4217792749404907, "learning_rate": 2.3098462214213418e-05, "loss": 0.0244, "step": 69680 }, { "grad_norm": 0.3448096513748169, "learning_rate": 2.308452617152712e-05, "loss": 0.0214, "step": 69690 }, { "grad_norm": 0.38045284152030945, "learning_rate": 2.3070593072276603e-05, "loss": 0.0269, "step": 69700 }, { "grad_norm": 0.31298673152923584, "learning_rate": 2.3056662917985518e-05, "loss": 0.0252, "step": 69710 }, { "grad_norm": 0.4215264916419983, "learning_rate": 2.3042735710177283e-05, "loss": 0.0241, "step": 69720 }, { "grad_norm": 1.290117621421814, "learning_rate": 2.302881145037492e-05, "loss": 0.0333, "step": 69730 }, { "grad_norm": 0.24625742435455322, "learning_rate": 2.3014890140101192e-05, "loss": 0.0278, "step": 69740 }, { "grad_norm": 0.41923126578330994, "learning_rate": 2.30009717808785e-05, "loss": 0.0317, "step": 69750 }, { "grad_norm": 0.20881287753582, "learning_rate": 2.2987056374228967e-05, "loss": 0.0253, "step": 69760 }, { "grad_norm": 0.5532712936401367, "learning_rate": 2.297314392167431e-05, "loss": 0.0333, "step": 69770 }, { "grad_norm": 0.6562377214431763, "learning_rate": 2.295923442473601e-05, "loss": 0.0202, "step": 69780 }, { "grad_norm": 0.4094034731388092, "learning_rate": 2.294532788493519e-05, "loss": 0.018, "step": 69790 }, { "grad_norm": 0.30672791600227356, "learning_rate": 2.293142430379262e-05, "loss": 0.0233, "step": 69800 }, { "grad_norm": 0.3127037584781647, "learning_rate": 2.291752368282879e-05, "loss": 0.0224, "step": 69810 }, { "grad_norm": 0.8124366998672485, "learning_rate": 2.290362602356385e-05, "loss": 0.0208, "step": 69820 }, { "grad_norm": 0.4616157114505768, "learning_rate": 2.288973132751765e-05, "loss": 0.0314, "step": 69830 }, { "grad_norm": 0.2695910632610321, "learning_rate": 2.287583959620965e-05, "loss": 0.0186, "step": 69840 }, { "grad_norm": 0.3958037793636322, "learning_rate": 2.2861950831159074e-05, "loss": 0.0216, "step": 69850 }, { "grad_norm": 0.43606308102607727, "learning_rate": 2.284806503388472e-05, "loss": 0.0216, "step": 69860 }, { "grad_norm": 0.2038428783416748, "learning_rate": 2.283418220590518e-05, "loss": 0.0214, "step": 69870 }, { "grad_norm": 0.5370387434959412, "learning_rate": 2.282030234873862e-05, "loss": 0.0245, "step": 69880 }, { "grad_norm": 0.26508864760398865, "learning_rate": 2.2806425463902948e-05, "loss": 0.0345, "step": 69890 }, { "grad_norm": 0.47697556018829346, "learning_rate": 2.279255155291568e-05, "loss": 0.024, "step": 69900 }, { "grad_norm": 0.512816846370697, "learning_rate": 2.277868061729409e-05, "loss": 0.0249, "step": 69910 }, { "grad_norm": 0.4719209671020508, "learning_rate": 2.276481265855504e-05, "loss": 0.0233, "step": 69920 }, { "grad_norm": 0.20552611351013184, "learning_rate": 2.2750947678215128e-05, "loss": 0.0239, "step": 69930 }, { "grad_norm": 0.3114805221557617, "learning_rate": 2.2737085677790616e-05, "loss": 0.0313, "step": 69940 }, { "grad_norm": 0.6256904006004333, "learning_rate": 2.2723226658797443e-05, "loss": 0.0344, "step": 69950 }, { "grad_norm": 0.20385809242725372, "learning_rate": 2.270937062275117e-05, "loss": 0.0279, "step": 69960 }, { "grad_norm": 0.36610737442970276, "learning_rate": 2.2695517571167113e-05, "loss": 0.0245, "step": 69970 }, { "grad_norm": 0.39833348989486694, "learning_rate": 2.268166750556019e-05, "loss": 0.0181, "step": 69980 }, { "grad_norm": 0.4226464331150055, "learning_rate": 2.266782042744503e-05, "loss": 0.0167, "step": 69990 }, { "grad_norm": 0.377087265253067, "learning_rate": 2.2653976338335935e-05, "loss": 0.0297, "step": 70000 }, { "grad_norm": 0.5592187643051147, "learning_rate": 2.2640135239746895e-05, "loss": 0.026, "step": 70010 }, { "grad_norm": 0.43477144837379456, "learning_rate": 2.2626297133191515e-05, "loss": 0.021, "step": 70020 }, { "grad_norm": 0.5622238516807556, "learning_rate": 2.261246202018312e-05, "loss": 0.0241, "step": 70030 }, { "grad_norm": 0.27586066722869873, "learning_rate": 2.2598629902234718e-05, "loss": 0.0142, "step": 70040 }, { "grad_norm": 0.33736738562583923, "learning_rate": 2.258480078085894e-05, "loss": 0.0288, "step": 70050 }, { "grad_norm": 0.48532506823539734, "learning_rate": 2.2570974657568126e-05, "loss": 0.0201, "step": 70060 }, { "grad_norm": 0.258364200592041, "learning_rate": 2.2557151533874286e-05, "loss": 0.0256, "step": 70070 }, { "grad_norm": 1.656656265258789, "learning_rate": 2.2543331411289116e-05, "loss": 0.0317, "step": 70080 }, { "grad_norm": 0.2524386942386627, "learning_rate": 2.252951429132392e-05, "loss": 0.0196, "step": 70090 }, { "grad_norm": 1.4416227340698242, "learning_rate": 2.2515700175489762e-05, "loss": 0.026, "step": 70100 }, { "grad_norm": 0.4996911883354187, "learning_rate": 2.2501889065297288e-05, "loss": 0.0258, "step": 70110 }, { "grad_norm": 0.3336188495159149, "learning_rate": 2.2488080962256887e-05, "loss": 0.0224, "step": 70120 }, { "grad_norm": 0.47483164072036743, "learning_rate": 2.2474275867878585e-05, "loss": 0.0242, "step": 70130 }, { "grad_norm": 0.39121365547180176, "learning_rate": 2.246047378367211e-05, "loss": 0.0237, "step": 70140 }, { "grad_norm": 0.18906021118164062, "learning_rate": 2.2446674711146798e-05, "loss": 0.0256, "step": 70150 }, { "grad_norm": 0.36160486936569214, "learning_rate": 2.2432878651811734e-05, "loss": 0.0205, "step": 70160 }, { "grad_norm": 0.7261502146720886, "learning_rate": 2.2419085607175594e-05, "loss": 0.0236, "step": 70170 }, { "grad_norm": 0.20722557604312897, "learning_rate": 2.240529557874678e-05, "loss": 0.0192, "step": 70180 }, { "grad_norm": 1.2954398393630981, "learning_rate": 2.2391508568033358e-05, "loss": 0.03, "step": 70190 }, { "grad_norm": 0.3207668364048004, "learning_rate": 2.237772457654307e-05, "loss": 0.0328, "step": 70200 }, { "grad_norm": 0.24456006288528442, "learning_rate": 2.236394360578327e-05, "loss": 0.0198, "step": 70210 }, { "grad_norm": 0.4484371542930603, "learning_rate": 2.2350165657261073e-05, "loss": 0.0289, "step": 70220 }, { "grad_norm": 0.3601720631122589, "learning_rate": 2.2336390732483164e-05, "loss": 0.0192, "step": 70230 }, { "grad_norm": 0.4662289619445801, "learning_rate": 2.2322618832955976e-05, "loss": 0.0276, "step": 70240 }, { "grad_norm": 0.41638079285621643, "learning_rate": 2.2308849960185578e-05, "loss": 0.0294, "step": 70250 }, { "grad_norm": 0.5294591784477234, "learning_rate": 2.229508411567774e-05, "loss": 0.0258, "step": 70260 }, { "grad_norm": 0.4104349613189697, "learning_rate": 2.2281321300937823e-05, "loss": 0.0298, "step": 70270 }, { "grad_norm": 0.45585694909095764, "learning_rate": 2.2267561517470937e-05, "loss": 0.0353, "step": 70280 }, { "grad_norm": 0.8073983192443848, "learning_rate": 2.2253804766781845e-05, "loss": 0.0281, "step": 70290 }, { "grad_norm": 0.5759263634681702, "learning_rate": 2.2240051050374933e-05, "loss": 0.0241, "step": 70300 }, { "grad_norm": 0.4060220718383789, "learning_rate": 2.22263003697543e-05, "loss": 0.0215, "step": 70310 }, { "grad_norm": 0.5926268100738525, "learning_rate": 2.2212552726423702e-05, "loss": 0.0257, "step": 70320 }, { "grad_norm": 0.602275550365448, "learning_rate": 2.2198808121886578e-05, "loss": 0.035, "step": 70330 }, { "grad_norm": 0.24294525384902954, "learning_rate": 2.218506655764598e-05, "loss": 0.0235, "step": 70340 }, { "grad_norm": 0.32766157388687134, "learning_rate": 2.2171328035204704e-05, "loss": 0.0233, "step": 70350 }, { "grad_norm": 0.1785789281129837, "learning_rate": 2.2157592556065133e-05, "loss": 0.0219, "step": 70360 }, { "grad_norm": 0.49671292304992676, "learning_rate": 2.2143860121729382e-05, "loss": 0.0303, "step": 70370 }, { "grad_norm": 0.5048261880874634, "learning_rate": 2.2130130733699206e-05, "loss": 0.024, "step": 70380 }, { "grad_norm": 0.3377968668937683, "learning_rate": 2.2116404393476052e-05, "loss": 0.0224, "step": 70390 }, { "grad_norm": 0.39835435152053833, "learning_rate": 2.2102681102560967e-05, "loss": 0.0407, "step": 70400 }, { "grad_norm": 0.509598433971405, "learning_rate": 2.208896086245476e-05, "loss": 0.0402, "step": 70410 }, { "grad_norm": 0.3333780765533447, "learning_rate": 2.2075243674657808e-05, "loss": 0.0319, "step": 70420 }, { "grad_norm": 0.5463317632675171, "learning_rate": 2.206152954067024e-05, "loss": 0.0294, "step": 70430 }, { "grad_norm": 0.3994809687137604, "learning_rate": 2.2047818461991758e-05, "loss": 0.0231, "step": 70440 }, { "grad_norm": 0.320571631193161, "learning_rate": 2.2034110440121864e-05, "loss": 0.0235, "step": 70450 }, { "grad_norm": 0.34897446632385254, "learning_rate": 2.2020405476559585e-05, "loss": 0.0206, "step": 70460 }, { "grad_norm": 0.7459071278572083, "learning_rate": 2.2006703572803722e-05, "loss": 0.027, "step": 70470 }, { "grad_norm": 0.2407209873199463, "learning_rate": 2.1993004730352646e-05, "loss": 0.0194, "step": 70480 }, { "grad_norm": 0.3895094692707062, "learning_rate": 2.1979308950704464e-05, "loss": 0.0195, "step": 70490 }, { "grad_norm": 0.4735594093799591, "learning_rate": 2.1965616235356923e-05, "loss": 0.0276, "step": 70500 }, { "grad_norm": 0.30089715123176575, "learning_rate": 2.1951926585807465e-05, "loss": 0.0178, "step": 70510 }, { "grad_norm": 0.4352789521217346, "learning_rate": 2.193824000355313e-05, "loss": 0.019, "step": 70520 }, { "grad_norm": 0.36887967586517334, "learning_rate": 2.192455649009067e-05, "loss": 0.0249, "step": 70530 }, { "grad_norm": 0.4429861307144165, "learning_rate": 2.1910876046916523e-05, "loss": 0.0236, "step": 70540 }, { "grad_norm": 0.5151901841163635, "learning_rate": 2.1897198675526725e-05, "loss": 0.0352, "step": 70550 }, { "grad_norm": 0.34755054116249084, "learning_rate": 2.1883524377417024e-05, "loss": 0.0418, "step": 70560 }, { "grad_norm": 0.40521004796028137, "learning_rate": 2.1869853154082825e-05, "loss": 0.0288, "step": 70570 }, { "grad_norm": 0.3744193911552429, "learning_rate": 2.1856185007019215e-05, "loss": 0.0297, "step": 70580 }, { "grad_norm": 0.2841062545776367, "learning_rate": 2.1842519937720874e-05, "loss": 0.0299, "step": 70590 }, { "grad_norm": 0.4161204397678375, "learning_rate": 2.182885794768224e-05, "loss": 0.0257, "step": 70600 }, { "grad_norm": 0.25587084889411926, "learning_rate": 2.1815199038397326e-05, "loss": 0.0256, "step": 70610 }, { "grad_norm": 0.4252419173717499, "learning_rate": 2.1801543211359893e-05, "loss": 0.0281, "step": 70620 }, { "grad_norm": 0.3274235129356384, "learning_rate": 2.178789046806326e-05, "loss": 0.025, "step": 70630 }, { "grad_norm": 0.9173341393470764, "learning_rate": 2.1774240810000547e-05, "loss": 0.0319, "step": 70640 }, { "grad_norm": 0.4570322632789612, "learning_rate": 2.17605942386644e-05, "loss": 0.0378, "step": 70650 }, { "grad_norm": 0.36516034603118896, "learning_rate": 2.174695075554723e-05, "loss": 0.0194, "step": 70660 }, { "grad_norm": 0.4257771968841553, "learning_rate": 2.173331036214103e-05, "loss": 0.022, "step": 70670 }, { "grad_norm": 0.3295445442199707, "learning_rate": 2.171967305993753e-05, "loss": 0.0161, "step": 70680 }, { "grad_norm": 0.3054327070713043, "learning_rate": 2.1706038850428024e-05, "loss": 0.0247, "step": 70690 }, { "grad_norm": 0.5561276078224182, "learning_rate": 2.169240773510361e-05, "loss": 0.0265, "step": 70700 }, { "grad_norm": 0.42798739671707153, "learning_rate": 2.1678779715454906e-05, "loss": 0.025, "step": 70710 }, { "grad_norm": 0.3753058910369873, "learning_rate": 2.166515479297229e-05, "loss": 0.0352, "step": 70720 }, { "grad_norm": 0.20497757196426392, "learning_rate": 2.1651532969145732e-05, "loss": 0.0199, "step": 70730 }, { "grad_norm": 0.49264758825302124, "learning_rate": 2.1637914245464903e-05, "loss": 0.0298, "step": 70740 }, { "grad_norm": 0.5227330923080444, "learning_rate": 2.1624298623419133e-05, "loss": 0.0224, "step": 70750 }, { "grad_norm": 0.47712698578834534, "learning_rate": 2.1610686104497413e-05, "loss": 0.0225, "step": 70760 }, { "grad_norm": 0.4190908968448639, "learning_rate": 2.1597076690188366e-05, "loss": 0.0197, "step": 70770 }, { "grad_norm": 0.524860143661499, "learning_rate": 2.1583470381980315e-05, "loss": 0.0243, "step": 70780 }, { "grad_norm": 0.9051259160041809, "learning_rate": 2.156986718136123e-05, "loss": 0.0283, "step": 70790 }, { "grad_norm": 0.6196923851966858, "learning_rate": 2.155626708981871e-05, "loss": 0.0284, "step": 70800 }, { "grad_norm": 0.4180694818496704, "learning_rate": 2.1542670108840075e-05, "loss": 0.0179, "step": 70810 }, { "grad_norm": 0.31907328963279724, "learning_rate": 2.152907623991221e-05, "loss": 0.0205, "step": 70820 }, { "grad_norm": 0.2269396185874939, "learning_rate": 2.1515485484521802e-05, "loss": 0.0339, "step": 70830 }, { "grad_norm": 0.7792335748672485, "learning_rate": 2.150189784415506e-05, "loss": 0.0283, "step": 70840 }, { "grad_norm": 0.6577445268630981, "learning_rate": 2.148831332029794e-05, "loss": 0.0301, "step": 70850 }, { "grad_norm": 0.28549206256866455, "learning_rate": 2.147473191443599e-05, "loss": 0.0211, "step": 70860 }, { "grad_norm": 0.39828017354011536, "learning_rate": 2.1461153628054492e-05, "loss": 0.02, "step": 70870 }, { "grad_norm": 0.43896207213401794, "learning_rate": 2.1447578462638284e-05, "loss": 0.0262, "step": 70880 }, { "grad_norm": 0.37370845675468445, "learning_rate": 2.143400641967201e-05, "loss": 0.0256, "step": 70890 }, { "grad_norm": 0.8690947890281677, "learning_rate": 2.1420437500639823e-05, "loss": 0.0291, "step": 70900 }, { "grad_norm": 0.37874144315719604, "learning_rate": 2.1406871707025646e-05, "loss": 0.0333, "step": 70910 }, { "grad_norm": 0.3449559807777405, "learning_rate": 2.1393309040312963e-05, "loss": 0.0242, "step": 70920 }, { "grad_norm": 0.3891867995262146, "learning_rate": 2.137974950198502e-05, "loss": 0.0222, "step": 70930 }, { "grad_norm": 0.4156399369239807, "learning_rate": 2.1366193093524617e-05, "loss": 0.0281, "step": 70940 }, { "grad_norm": 0.1740228235721588, "learning_rate": 2.1352639816414288e-05, "loss": 0.0183, "step": 70950 }, { "grad_norm": 0.34180471301078796, "learning_rate": 2.1339089672136202e-05, "loss": 0.0222, "step": 70960 }, { "grad_norm": 0.42414435744285583, "learning_rate": 2.1325542662172193e-05, "loss": 0.0223, "step": 70970 }, { "grad_norm": 0.3799152970314026, "learning_rate": 2.1311998788003712e-05, "loss": 0.023, "step": 70980 }, { "grad_norm": 0.3386235535144806, "learning_rate": 2.1298458051111907e-05, "loss": 0.0224, "step": 70990 }, { "grad_norm": 0.5095663666725159, "learning_rate": 2.1284920452977602e-05, "loss": 0.0228, "step": 71000 }, { "grad_norm": 0.4353993535041809, "learning_rate": 2.12713859950812e-05, "loss": 0.0181, "step": 71010 }, { "grad_norm": 0.4422358572483063, "learning_rate": 2.1257854678902832e-05, "loss": 0.0269, "step": 71020 }, { "grad_norm": 0.30993354320526123, "learning_rate": 2.124432650592227e-05, "loss": 0.0241, "step": 71030 }, { "grad_norm": 0.4082016944885254, "learning_rate": 2.1230801477618937e-05, "loss": 0.0288, "step": 71040 }, { "grad_norm": 0.3246501088142395, "learning_rate": 2.121727959547189e-05, "loss": 0.0207, "step": 71050 }, { "grad_norm": 0.6152711510658264, "learning_rate": 2.1203760860959886e-05, "loss": 0.0285, "step": 71060 }, { "grad_norm": 0.2618725597858429, "learning_rate": 2.119024527556127e-05, "loss": 0.0196, "step": 71070 }, { "grad_norm": 0.20697928965091705, "learning_rate": 2.117673284075415e-05, "loss": 0.0273, "step": 71080 }, { "grad_norm": 0.3907408118247986, "learning_rate": 2.1163223558016165e-05, "loss": 0.0208, "step": 71090 }, { "grad_norm": 0.5409773588180542, "learning_rate": 2.1149717428824722e-05, "loss": 0.0258, "step": 71100 }, { "grad_norm": 0.3823133707046509, "learning_rate": 2.1136214454656784e-05, "loss": 0.0236, "step": 71110 }, { "grad_norm": 0.48253533244132996, "learning_rate": 2.1122714636989056e-05, "loss": 0.0231, "step": 71120 }, { "grad_norm": 0.35523056983947754, "learning_rate": 2.1109217977297823e-05, "loss": 0.0212, "step": 71130 }, { "grad_norm": 0.6047936677932739, "learning_rate": 2.1095724477059077e-05, "loss": 0.0215, "step": 71140 }, { "grad_norm": 0.1509474813938141, "learning_rate": 2.1082234137748434e-05, "loss": 0.0155, "step": 71150 }, { "grad_norm": 0.30131998658180237, "learning_rate": 2.106874696084122e-05, "loss": 0.022, "step": 71160 }, { "grad_norm": 0.4338862895965576, "learning_rate": 2.1055262947812315e-05, "loss": 0.0274, "step": 71170 }, { "grad_norm": 0.705929696559906, "learning_rate": 2.104178210013636e-05, "loss": 0.0216, "step": 71180 }, { "grad_norm": 1.0070358514785767, "learning_rate": 2.1028304419287554e-05, "loss": 0.0297, "step": 71190 }, { "grad_norm": 0.26430901885032654, "learning_rate": 2.101482990673982e-05, "loss": 0.0392, "step": 71200 }, { "grad_norm": 0.28872448205947876, "learning_rate": 2.1001358563966707e-05, "loss": 0.0223, "step": 71210 }, { "grad_norm": 0.24618999660015106, "learning_rate": 2.098789039244144e-05, "loss": 0.0202, "step": 71220 }, { "grad_norm": 0.2653791308403015, "learning_rate": 2.097442539363685e-05, "loss": 0.0191, "step": 71230 }, { "grad_norm": 0.2736259996891022, "learning_rate": 2.0960963569025448e-05, "loss": 0.0162, "step": 71240 }, { "grad_norm": 0.7528740167617798, "learning_rate": 2.0947504920079435e-05, "loss": 0.0209, "step": 71250 }, { "grad_norm": 0.35285553336143494, "learning_rate": 2.093404944827059e-05, "loss": 0.02, "step": 71260 }, { "grad_norm": 0.3324965238571167, "learning_rate": 2.0920597155070394e-05, "loss": 0.0186, "step": 71270 }, { "grad_norm": 0.510890781879425, "learning_rate": 2.090714804194997e-05, "loss": 0.0301, "step": 71280 }, { "grad_norm": 0.40826550126075745, "learning_rate": 2.0893702110380114e-05, "loss": 0.0223, "step": 71290 }, { "grad_norm": 0.5139980912208557, "learning_rate": 2.0880259361831216e-05, "loss": 0.022, "step": 71300 }, { "grad_norm": 0.3741019070148468, "learning_rate": 2.0866819797773397e-05, "loss": 0.0192, "step": 71310 }, { "grad_norm": 0.1471608579158783, "learning_rate": 2.085338341967634e-05, "loss": 0.0206, "step": 71320 }, { "grad_norm": 0.21213959157466888, "learning_rate": 2.083995022900946e-05, "loss": 0.0214, "step": 71330 }, { "grad_norm": 0.360958456993103, "learning_rate": 2.0826520227241768e-05, "loss": 0.0292, "step": 71340 }, { "grad_norm": 0.14519955217838287, "learning_rate": 2.081309341584199e-05, "loss": 0.0187, "step": 71350 }, { "grad_norm": 0.2514846920967102, "learning_rate": 2.0799669796278413e-05, "loss": 0.0208, "step": 71360 }, { "grad_norm": 0.3556397557258606, "learning_rate": 2.078624937001906e-05, "loss": 0.0159, "step": 71370 }, { "grad_norm": 0.27760976552963257, "learning_rate": 2.0772832138531544e-05, "loss": 0.0209, "step": 71380 }, { "grad_norm": 2.215993642807007, "learning_rate": 2.0759418103283158e-05, "loss": 0.0321, "step": 71390 }, { "grad_norm": 0.37349221110343933, "learning_rate": 2.0746007265740846e-05, "loss": 0.0232, "step": 71400 }, { "grad_norm": 0.6631950736045837, "learning_rate": 2.0732599627371215e-05, "loss": 0.0238, "step": 71410 }, { "grad_norm": 0.3414294719696045, "learning_rate": 2.0719195189640468e-05, "loss": 0.0239, "step": 71420 }, { "grad_norm": 0.2553479075431824, "learning_rate": 2.0705793954014528e-05, "loss": 0.0186, "step": 71430 }, { "grad_norm": 0.21485574543476105, "learning_rate": 2.06923959219589e-05, "loss": 0.0149, "step": 71440 }, { "grad_norm": 0.22734591364860535, "learning_rate": 2.0679001094938793e-05, "loss": 0.0266, "step": 71450 }, { "grad_norm": 0.5098473429679871, "learning_rate": 2.066560947441904e-05, "loss": 0.036, "step": 71460 }, { "grad_norm": 0.49321216344833374, "learning_rate": 2.0652221061864152e-05, "loss": 0.0237, "step": 71470 }, { "grad_norm": 0.20103338360786438, "learning_rate": 2.0638835858738226e-05, "loss": 0.0187, "step": 71480 }, { "grad_norm": 0.5156283378601074, "learning_rate": 2.0625453866505068e-05, "loss": 0.0327, "step": 71490 }, { "grad_norm": 0.38726118206977844, "learning_rate": 2.0612075086628123e-05, "loss": 0.0238, "step": 71500 }, { "grad_norm": 0.24915556609630585, "learning_rate": 2.059869952057045e-05, "loss": 0.0188, "step": 71510 }, { "grad_norm": 0.5549647212028503, "learning_rate": 2.0585327169794793e-05, "loss": 0.0258, "step": 71520 }, { "grad_norm": 0.15158987045288086, "learning_rate": 2.0571958035763534e-05, "loss": 0.0166, "step": 71530 }, { "grad_norm": 0.1496596485376358, "learning_rate": 2.0558592119938715e-05, "loss": 0.0272, "step": 71540 }, { "grad_norm": 0.38773903250694275, "learning_rate": 2.054522942378198e-05, "loss": 0.031, "step": 71550 }, { "grad_norm": 0.7703311443328857, "learning_rate": 2.053186994875469e-05, "loss": 0.0211, "step": 71560 }, { "grad_norm": 0.3301883339881897, "learning_rate": 2.0518513696317775e-05, "loss": 0.0234, "step": 71570 }, { "grad_norm": 0.33332857489585876, "learning_rate": 2.050516066793188e-05, "loss": 0.013, "step": 71580 }, { "grad_norm": 0.4538692533969879, "learning_rate": 2.0491810865057276e-05, "loss": 0.0369, "step": 71590 }, { "grad_norm": 0.532942533493042, "learning_rate": 2.0478464289153882e-05, "loss": 0.0221, "step": 71600 }, { "grad_norm": 0.5535504817962646, "learning_rate": 2.0465120941681232e-05, "loss": 0.0256, "step": 71610 }, { "grad_norm": 0.40073567628860474, "learning_rate": 2.0451780824098566e-05, "loss": 0.0193, "step": 71620 }, { "grad_norm": 0.3003835380077362, "learning_rate": 2.043844393786471e-05, "loss": 0.0195, "step": 71630 }, { "grad_norm": 0.3160027265548706, "learning_rate": 2.0425110284438193e-05, "loss": 0.0287, "step": 71640 }, { "grad_norm": 0.2541641294956207, "learning_rate": 2.041177986527712e-05, "loss": 0.0167, "step": 71650 }, { "grad_norm": 0.36254552006721497, "learning_rate": 2.0398452681839347e-05, "loss": 0.0343, "step": 71660 }, { "grad_norm": 0.7611741423606873, "learning_rate": 2.0385128735582272e-05, "loss": 0.0348, "step": 71670 }, { "grad_norm": 0.530092716217041, "learning_rate": 2.0371808027963003e-05, "loss": 0.0266, "step": 71680 }, { "grad_norm": 0.28476861119270325, "learning_rate": 2.035849056043825e-05, "loss": 0.0253, "step": 71690 }, { "grad_norm": 0.43987610936164856, "learning_rate": 2.03451763344644e-05, "loss": 0.0225, "step": 71700 }, { "grad_norm": 0.8720492720603943, "learning_rate": 2.0331865351497482e-05, "loss": 0.0289, "step": 71710 }, { "grad_norm": 0.3535493314266205, "learning_rate": 2.0318557612993185e-05, "loss": 0.018, "step": 71720 }, { "grad_norm": 0.22004249691963196, "learning_rate": 2.0305253120406782e-05, "loss": 0.0138, "step": 71730 }, { "grad_norm": 0.409342885017395, "learning_rate": 2.029195187519325e-05, "loss": 0.0197, "step": 71740 }, { "grad_norm": 0.4022339880466461, "learning_rate": 2.0278653878807217e-05, "loss": 0.0216, "step": 71750 }, { "grad_norm": 0.9490525126457214, "learning_rate": 2.0265359132702892e-05, "loss": 0.0266, "step": 71760 }, { "grad_norm": 0.7578752040863037, "learning_rate": 2.0252067638334182e-05, "loss": 0.0218, "step": 71770 }, { "grad_norm": 0.37122875452041626, "learning_rate": 2.023877939715464e-05, "loss": 0.0277, "step": 71780 }, { "grad_norm": 0.37210896611213684, "learning_rate": 2.0225494410617448e-05, "loss": 0.0253, "step": 71790 }, { "grad_norm": 0.2986565828323364, "learning_rate": 2.02122126801754e-05, "loss": 0.02, "step": 71800 }, { "grad_norm": 0.4313961863517761, "learning_rate": 2.0198934207281018e-05, "loss": 0.0478, "step": 71810 }, { "grad_norm": 0.28621891140937805, "learning_rate": 2.0185658993386357e-05, "loss": 0.0194, "step": 71820 }, { "grad_norm": 0.4248630404472351, "learning_rate": 2.0172387039943226e-05, "loss": 0.0287, "step": 71830 }, { "grad_norm": 0.23822571337223053, "learning_rate": 2.0159118348402968e-05, "loss": 0.0198, "step": 71840 }, { "grad_norm": 0.4272685945034027, "learning_rate": 2.0145852920216697e-05, "loss": 0.0227, "step": 71850 }, { "grad_norm": 0.281990647315979, "learning_rate": 2.0132590756835045e-05, "loss": 0.0252, "step": 71860 }, { "grad_norm": 0.7267889976501465, "learning_rate": 2.0119331859708386e-05, "loss": 0.0249, "step": 71870 }, { "grad_norm": 0.8306824564933777, "learning_rate": 2.0106076230286654e-05, "loss": 0.0314, "step": 71880 }, { "grad_norm": 0.2830713987350464, "learning_rate": 2.0092823870019488e-05, "loss": 0.0244, "step": 71890 }, { "grad_norm": 0.3803452253341675, "learning_rate": 2.0079574780356115e-05, "loss": 0.0251, "step": 71900 }, { "grad_norm": 0.5295437574386597, "learning_rate": 2.0066328962745486e-05, "loss": 0.0201, "step": 71910 }, { "grad_norm": 0.3987143039703369, "learning_rate": 2.00530864186361e-05, "loss": 0.0327, "step": 71920 }, { "grad_norm": 0.35690855979919434, "learning_rate": 2.003984714947618e-05, "loss": 0.0321, "step": 71930 }, { "grad_norm": 0.36436423659324646, "learning_rate": 2.0026611156713515e-05, "loss": 0.0197, "step": 71940 }, { "grad_norm": 0.42101141810417175, "learning_rate": 2.0013378441795584e-05, "loss": 0.0206, "step": 71950 }, { "grad_norm": 0.3855060636997223, "learning_rate": 2.0000149006169506e-05, "loss": 0.0344, "step": 71960 }, { "grad_norm": 0.39573413133621216, "learning_rate": 1.9986922851282048e-05, "loss": 0.0272, "step": 71970 }, { "grad_norm": 0.4833182394504547, "learning_rate": 1.997369997857956e-05, "loss": 0.0179, "step": 71980 }, { "grad_norm": 0.2323649376630783, "learning_rate": 1.9960480389508103e-05, "loss": 0.0143, "step": 71990 }, { "grad_norm": 0.5027960538864136, "learning_rate": 1.9947264085513363e-05, "loss": 0.0354, "step": 72000 }, { "grad_norm": 0.36823874711990356, "learning_rate": 1.993405106804062e-05, "loss": 0.0167, "step": 72010 }, { "grad_norm": 0.3163206875324249, "learning_rate": 1.9920841338534867e-05, "loss": 0.0277, "step": 72020 }, { "grad_norm": 0.38604167103767395, "learning_rate": 1.9907634898440646e-05, "loss": 0.0205, "step": 72030 }, { "grad_norm": 0.8038565516471863, "learning_rate": 1.989443174920227e-05, "loss": 0.0304, "step": 72040 }, { "grad_norm": 0.5019468069076538, "learning_rate": 1.988123189226355e-05, "loss": 0.0219, "step": 72050 }, { "grad_norm": 0.5774677395820618, "learning_rate": 1.986803532906805e-05, "loss": 0.0235, "step": 72060 }, { "grad_norm": 0.42838355898857117, "learning_rate": 1.985484206105889e-05, "loss": 0.0356, "step": 72070 }, { "grad_norm": 0.33247512578964233, "learning_rate": 1.9841652089678887e-05, "loss": 0.0123, "step": 72080 }, { "grad_norm": 0.47091174125671387, "learning_rate": 1.9828465416370435e-05, "loss": 0.0186, "step": 72090 }, { "grad_norm": 0.2714831829071045, "learning_rate": 1.9815282042575685e-05, "loss": 0.0151, "step": 72100 }, { "grad_norm": 0.28378838300704956, "learning_rate": 1.9802101969736285e-05, "loss": 0.0247, "step": 72110 }, { "grad_norm": 0.4320545494556427, "learning_rate": 1.9788925199293633e-05, "loss": 0.0203, "step": 72120 }, { "grad_norm": 0.7499855756759644, "learning_rate": 1.977575173268868e-05, "loss": 0.018, "step": 72130 }, { "grad_norm": 0.758172869682312, "learning_rate": 1.9762581571362094e-05, "loss": 0.0232, "step": 72140 }, { "grad_norm": 0.49468284845352173, "learning_rate": 1.9749414716754112e-05, "loss": 0.0192, "step": 72150 }, { "grad_norm": 0.35135746002197266, "learning_rate": 1.973625117030465e-05, "loss": 0.03, "step": 72160 }, { "grad_norm": 0.448396235704422, "learning_rate": 1.972309093345326e-05, "loss": 0.0195, "step": 72170 }, { "grad_norm": 0.2470799833536148, "learning_rate": 1.970993400763915e-05, "loss": 0.0187, "step": 72180 }, { "grad_norm": 0.42939460277557373, "learning_rate": 1.969678039430109e-05, "loss": 0.0182, "step": 72190 }, { "grad_norm": 0.5625166893005371, "learning_rate": 1.9683630094877564e-05, "loss": 0.0336, "step": 72200 }, { "grad_norm": 0.34292900562286377, "learning_rate": 1.9670483110806696e-05, "loss": 0.0195, "step": 72210 }, { "grad_norm": 0.6584309339523315, "learning_rate": 1.9657339443526172e-05, "loss": 0.0266, "step": 72220 }, { "grad_norm": 0.28779152035713196, "learning_rate": 1.964419909447339e-05, "loss": 0.0225, "step": 72230 }, { "grad_norm": 0.21147339046001434, "learning_rate": 1.963106206508536e-05, "loss": 0.0181, "step": 72240 }, { "grad_norm": 0.5289291143417358, "learning_rate": 1.9617928356798738e-05, "loss": 0.0158, "step": 72250 }, { "grad_norm": 0.44217148423194885, "learning_rate": 1.960479797104977e-05, "loss": 0.0262, "step": 72260 }, { "grad_norm": 0.2581183612346649, "learning_rate": 1.9591670909274418e-05, "loss": 0.0194, "step": 72270 }, { "grad_norm": 0.24589429795742035, "learning_rate": 1.9578547172908184e-05, "loss": 0.0143, "step": 72280 }, { "grad_norm": 0.25687527656555176, "learning_rate": 1.9565426763386325e-05, "loss": 0.0181, "step": 72290 }, { "grad_norm": 0.3717920184135437, "learning_rate": 1.9552309682143622e-05, "loss": 0.0145, "step": 72300 }, { "grad_norm": 0.2678651213645935, "learning_rate": 1.9539195930614574e-05, "loss": 0.0214, "step": 72310 }, { "grad_norm": 0.6725535988807678, "learning_rate": 1.952608551023324e-05, "loss": 0.0219, "step": 72320 }, { "grad_norm": 0.6041199564933777, "learning_rate": 1.9512978422433397e-05, "loss": 0.0254, "step": 72330 }, { "grad_norm": 0.36046263575553894, "learning_rate": 1.9499874668648376e-05, "loss": 0.0242, "step": 72340 }, { "grad_norm": 0.22507360577583313, "learning_rate": 1.9486774250311213e-05, "loss": 0.0208, "step": 72350 }, { "grad_norm": 0.5430155992507935, "learning_rate": 1.9473677168854537e-05, "loss": 0.0265, "step": 72360 }, { "grad_norm": 0.34351643919944763, "learning_rate": 1.946058342571065e-05, "loss": 0.0328, "step": 72370 }, { "grad_norm": 0.3146418035030365, "learning_rate": 1.9447493022311424e-05, "loss": 0.0181, "step": 72380 }, { "grad_norm": 0.42650723457336426, "learning_rate": 1.9434405960088442e-05, "loss": 0.0226, "step": 72390 }, { "grad_norm": 0.6176966428756714, "learning_rate": 1.9421322240472857e-05, "loss": 0.0227, "step": 72400 }, { "grad_norm": 0.575050413608551, "learning_rate": 1.9408241864895494e-05, "loss": 0.022, "step": 72410 }, { "grad_norm": 0.6451215744018555, "learning_rate": 1.939516483478681e-05, "loss": 0.0285, "step": 72420 }, { "grad_norm": 0.400526225566864, "learning_rate": 1.9382091151576902e-05, "loss": 0.032, "step": 72430 }, { "grad_norm": 0.33794090151786804, "learning_rate": 1.9369020816695454e-05, "loss": 0.0213, "step": 72440 }, { "grad_norm": 0.23727315664291382, "learning_rate": 1.935595383157184e-05, "loss": 0.0213, "step": 72450 }, { "grad_norm": 0.5290024876594543, "learning_rate": 1.9342890197635057e-05, "loss": 0.0162, "step": 72460 }, { "grad_norm": 0.2607986330986023, "learning_rate": 1.9329829916313685e-05, "loss": 0.0197, "step": 72470 }, { "grad_norm": 0.3571247458457947, "learning_rate": 1.9316772989036013e-05, "loss": 0.0156, "step": 72480 }, { "grad_norm": 0.5386734008789062, "learning_rate": 1.930371941722991e-05, "loss": 0.0236, "step": 72490 }, { "grad_norm": 0.32442644238471985, "learning_rate": 1.929066920232292e-05, "loss": 0.0165, "step": 72500 }, { "grad_norm": 0.4611949920654297, "learning_rate": 1.9277622345742152e-05, "loss": 0.0333, "step": 72510 }, { "grad_norm": 0.22815939784049988, "learning_rate": 1.926457884891444e-05, "loss": 0.0148, "step": 72520 }, { "grad_norm": 0.45551109313964844, "learning_rate": 1.9251538713266147e-05, "loss": 0.0226, "step": 72530 }, { "grad_norm": 0.7738412618637085, "learning_rate": 1.9238501940223358e-05, "loss": 0.0277, "step": 72540 }, { "grad_norm": 0.13952121138572693, "learning_rate": 1.922546853121175e-05, "loss": 0.0224, "step": 72550 }, { "grad_norm": 0.43935924768447876, "learning_rate": 1.9212438487656647e-05, "loss": 0.0255, "step": 72560 }, { "grad_norm": 0.5477132797241211, "learning_rate": 1.9199411810982965e-05, "loss": 0.0465, "step": 72570 }, { "grad_norm": 0.6520833969116211, "learning_rate": 1.918638850261532e-05, "loss": 0.0213, "step": 72580 }, { "grad_norm": 0.8389623761177063, "learning_rate": 1.9173368563977885e-05, "loss": 0.0166, "step": 72590 }, { "grad_norm": 0.28327086567878723, "learning_rate": 1.9160351996494513e-05, "loss": 0.0153, "step": 72600 }, { "grad_norm": 0.34019410610198975, "learning_rate": 1.9147338801588677e-05, "loss": 0.0167, "step": 72610 }, { "grad_norm": 0.36258652806282043, "learning_rate": 1.913432898068351e-05, "loss": 0.0214, "step": 72620 }, { "grad_norm": 0.22359497845172882, "learning_rate": 1.912132253520169e-05, "loss": 0.0128, "step": 72630 }, { "grad_norm": 0.2842255234718323, "learning_rate": 1.910831946656564e-05, "loss": 0.0217, "step": 72640 }, { "grad_norm": 0.6380869746208191, "learning_rate": 1.909531977619731e-05, "loss": 0.0337, "step": 72650 }, { "grad_norm": 0.3643248677253723, "learning_rate": 1.908232346551834e-05, "loss": 0.0227, "step": 72660 }, { "grad_norm": 0.38755518198013306, "learning_rate": 1.9069330535949993e-05, "loss": 0.0183, "step": 72670 }, { "grad_norm": 0.2804047465324402, "learning_rate": 1.9056340988913167e-05, "loss": 0.0223, "step": 72680 }, { "grad_norm": 0.2551400661468506, "learning_rate": 1.9043354825828357e-05, "loss": 0.023, "step": 72690 }, { "grad_norm": 0.3118516802787781, "learning_rate": 1.9030372048115724e-05, "loss": 0.0272, "step": 72700 }, { "grad_norm": 0.344602108001709, "learning_rate": 1.9017392657195056e-05, "loss": 0.0236, "step": 72710 }, { "grad_norm": 0.23134201765060425, "learning_rate": 1.9004416654485724e-05, "loss": 0.0138, "step": 72720 }, { "grad_norm": 0.39004552364349365, "learning_rate": 1.8991444041406785e-05, "loss": 0.0185, "step": 72730 }, { "grad_norm": 1.5983872413635254, "learning_rate": 1.8978474819376912e-05, "loss": 0.0194, "step": 72740 }, { "grad_norm": 0.3059048652648926, "learning_rate": 1.8965508989814406e-05, "loss": 0.0136, "step": 72750 }, { "grad_norm": 0.394956111907959, "learning_rate": 1.8952546554137158e-05, "loss": 0.0224, "step": 72760 }, { "grad_norm": 0.3476746082305908, "learning_rate": 1.8939587513762757e-05, "loss": 0.0194, "step": 72770 }, { "grad_norm": 0.27217164635658264, "learning_rate": 1.892663187010835e-05, "loss": 0.0281, "step": 72780 }, { "grad_norm": 0.3754786252975464, "learning_rate": 1.891367962459077e-05, "loss": 0.0271, "step": 72790 }, { "grad_norm": 0.2845408022403717, "learning_rate": 1.890073077862644e-05, "loss": 0.0217, "step": 72800 }, { "grad_norm": 0.8614521622657776, "learning_rate": 1.888778533363146e-05, "loss": 0.0424, "step": 72810 }, { "grad_norm": 0.4489211142063141, "learning_rate": 1.8874843291021478e-05, "loss": 0.0211, "step": 72820 }, { "grad_norm": 0.35515084862709045, "learning_rate": 1.8861904652211853e-05, "loss": 0.029, "step": 72830 }, { "grad_norm": 0.2905876934528351, "learning_rate": 1.8848969418617507e-05, "loss": 0.0176, "step": 72840 }, { "grad_norm": 0.525184154510498, "learning_rate": 1.8836037591653044e-05, "loss": 0.0189, "step": 72850 }, { "grad_norm": 0.4680921137332916, "learning_rate": 1.882310917273262e-05, "loss": 0.0278, "step": 72860 }, { "grad_norm": 0.5845322012901306, "learning_rate": 1.881018416327014e-05, "loss": 0.0365, "step": 72870 }, { "grad_norm": 0.6032387614250183, "learning_rate": 1.8797262564679014e-05, "loss": 0.0199, "step": 72880 }, { "grad_norm": 0.23099283874034882, "learning_rate": 1.8784344378372355e-05, "loss": 0.0322, "step": 72890 }, { "grad_norm": 0.3005492687225342, "learning_rate": 1.8771429605762843e-05, "loss": 0.0215, "step": 72900 }, { "grad_norm": 0.479621022939682, "learning_rate": 1.8758518248262835e-05, "loss": 0.0332, "step": 72910 }, { "grad_norm": 0.3957364857196808, "learning_rate": 1.8745610307284305e-05, "loss": 0.0226, "step": 72920 }, { "grad_norm": 0.4647309184074402, "learning_rate": 1.8732705784238862e-05, "loss": 0.021, "step": 72930 }, { "grad_norm": 0.28601351380348206, "learning_rate": 1.8719804680537682e-05, "loss": 0.0218, "step": 72940 }, { "grad_norm": 0.3420636057853699, "learning_rate": 1.870690699759164e-05, "loss": 0.0217, "step": 72950 }, { "grad_norm": 0.4558239281177521, "learning_rate": 1.8694012736811213e-05, "loss": 0.0233, "step": 72960 }, { "grad_norm": 1.1445130109786987, "learning_rate": 1.8681121899606463e-05, "loss": 0.0198, "step": 72970 }, { "grad_norm": 0.20728230476379395, "learning_rate": 1.8668234487387144e-05, "loss": 0.0175, "step": 72980 }, { "grad_norm": 0.2169620841741562, "learning_rate": 1.8655350501562592e-05, "loss": 0.0192, "step": 72990 }, { "grad_norm": 0.29744482040405273, "learning_rate": 1.8642469943541796e-05, "loss": 0.0209, "step": 73000 }, { "grad_norm": 0.4894568622112274, "learning_rate": 1.8629592814733328e-05, "loss": 0.0221, "step": 73010 }, { "grad_norm": 0.29531580209732056, "learning_rate": 1.8616719116545438e-05, "loss": 0.0208, "step": 73020 }, { "grad_norm": 0.28612300753593445, "learning_rate": 1.8603848850385952e-05, "loss": 0.0171, "step": 73030 }, { "grad_norm": 0.3821662962436676, "learning_rate": 1.859098201766236e-05, "loss": 0.0162, "step": 73040 }, { "grad_norm": 0.8185142874717712, "learning_rate": 1.8578118619781726e-05, "loss": 0.0193, "step": 73050 }, { "grad_norm": 0.47713398933410645, "learning_rate": 1.856525865815083e-05, "loss": 0.0229, "step": 73060 }, { "grad_norm": 0.4459001421928406, "learning_rate": 1.8552402134175966e-05, "loss": 0.0225, "step": 73070 }, { "grad_norm": 0.25339019298553467, "learning_rate": 1.8539549049263143e-05, "loss": 0.0234, "step": 73080 }, { "grad_norm": 0.3351482152938843, "learning_rate": 1.8526699404817917e-05, "loss": 0.0223, "step": 73090 }, { "grad_norm": 0.38018953800201416, "learning_rate": 1.8513853202245546e-05, "loss": 0.0223, "step": 73100 }, { "grad_norm": 0.5509732961654663, "learning_rate": 1.850101044295081e-05, "loss": 0.023, "step": 73110 }, { "grad_norm": 0.30356669425964355, "learning_rate": 1.848817112833825e-05, "loss": 0.0267, "step": 73120 }, { "grad_norm": 0.5736671686172485, "learning_rate": 1.8475335259811905e-05, "loss": 0.018, "step": 73130 }, { "grad_norm": 0.42663484811782837, "learning_rate": 1.8462502838775508e-05, "loss": 0.0187, "step": 73140 }, { "grad_norm": 0.15868975222110748, "learning_rate": 1.8449673866632378e-05, "loss": 0.0172, "step": 73150 }, { "grad_norm": 0.19473209977149963, "learning_rate": 1.8436848344785473e-05, "loss": 0.0421, "step": 73160 }, { "grad_norm": 0.44217604398727417, "learning_rate": 1.8424026274637378e-05, "loss": 0.0233, "step": 73170 }, { "grad_norm": 0.5980053544044495, "learning_rate": 1.8411207657590312e-05, "loss": 0.0333, "step": 73180 }, { "grad_norm": 0.2890612483024597, "learning_rate": 1.839839249504607e-05, "loss": 0.0235, "step": 73190 }, { "grad_norm": 0.7234116792678833, "learning_rate": 1.8385580788406108e-05, "loss": 0.0299, "step": 73200 }, { "grad_norm": 0.45973441004753113, "learning_rate": 1.837277253907152e-05, "loss": 0.026, "step": 73210 }, { "grad_norm": 0.2887592911720276, "learning_rate": 1.8359967748442957e-05, "loss": 0.0204, "step": 73220 }, { "grad_norm": 0.3389599919319153, "learning_rate": 1.8347166417920774e-05, "loss": 0.0191, "step": 73230 }, { "grad_norm": 0.2134077399969101, "learning_rate": 1.833436854890484e-05, "loss": 0.0211, "step": 73240 }, { "grad_norm": 0.5004751682281494, "learning_rate": 1.8321574142794796e-05, "loss": 0.0219, "step": 73250 }, { "grad_norm": 0.39236265420913696, "learning_rate": 1.8308783200989748e-05, "loss": 0.0289, "step": 73260 }, { "grad_norm": 0.2158758044242859, "learning_rate": 1.829599572488855e-05, "loss": 0.0187, "step": 73270 }, { "grad_norm": 0.15024779736995697, "learning_rate": 1.828321171588957e-05, "loss": 0.0289, "step": 73280 }, { "grad_norm": 0.4951125383377075, "learning_rate": 1.827043117539089e-05, "loss": 0.0291, "step": 73290 }, { "grad_norm": 0.4587057828903198, "learning_rate": 1.8257654104790116e-05, "loss": 0.0202, "step": 73300 }, { "grad_norm": 0.37393590807914734, "learning_rate": 1.82448805054846e-05, "loss": 0.0226, "step": 73310 }, { "grad_norm": 0.3820720314979553, "learning_rate": 1.823211037887119e-05, "loss": 0.0265, "step": 73320 }, { "grad_norm": 0.2688165307044983, "learning_rate": 1.821934372634645e-05, "loss": 0.0298, "step": 73330 }, { "grad_norm": 0.1879294216632843, "learning_rate": 1.8206580549306472e-05, "loss": 0.0228, "step": 73340 }, { "grad_norm": 0.35631048679351807, "learning_rate": 1.819382084914707e-05, "loss": 0.0321, "step": 73350 }, { "grad_norm": 0.40235432982444763, "learning_rate": 1.8181064627263577e-05, "loss": 0.0287, "step": 73360 }, { "grad_norm": 0.5975311994552612, "learning_rate": 1.8168311885051022e-05, "loss": 0.03, "step": 73370 }, { "grad_norm": 0.48379671573638916, "learning_rate": 1.815556262390402e-05, "loss": 0.0272, "step": 73380 }, { "grad_norm": 0.5406438708305359, "learning_rate": 1.8142816845216826e-05, "loss": 0.0193, "step": 73390 }, { "grad_norm": 0.367220014333725, "learning_rate": 1.8130074550383276e-05, "loss": 0.0177, "step": 73400 }, { "grad_norm": 0.3315453827381134, "learning_rate": 1.8117335740796848e-05, "loss": 0.0248, "step": 73410 }, { "grad_norm": 0.46480056643486023, "learning_rate": 1.810460041785067e-05, "loss": 0.0302, "step": 73420 }, { "grad_norm": 0.21837331354618073, "learning_rate": 1.809186858293742e-05, "loss": 0.0218, "step": 73430 }, { "grad_norm": 0.5349356532096863, "learning_rate": 1.8079140237449448e-05, "loss": 0.0278, "step": 73440 }, { "grad_norm": 0.3405110836029053, "learning_rate": 1.8066415382778707e-05, "loss": 0.0167, "step": 73450 }, { "grad_norm": 0.5501185059547424, "learning_rate": 1.8053694020316786e-05, "loss": 0.0317, "step": 73460 }, { "grad_norm": 0.513748049736023, "learning_rate": 1.8040976151454837e-05, "loss": 0.019, "step": 73470 }, { "grad_norm": 0.28435590863227844, "learning_rate": 1.8028261777583704e-05, "loss": 0.0211, "step": 73480 }, { "grad_norm": 0.4768966734409332, "learning_rate": 1.8015550900093758e-05, "loss": 0.023, "step": 73490 }, { "grad_norm": 0.6667037010192871, "learning_rate": 1.800284352037512e-05, "loss": 0.0256, "step": 73500 }, { "grad_norm": 0.5720049142837524, "learning_rate": 1.799013963981738e-05, "loss": 0.0252, "step": 73510 }, { "grad_norm": 0.36288121342658997, "learning_rate": 1.797743925980987e-05, "loss": 0.0206, "step": 73520 }, { "grad_norm": 0.5699130892753601, "learning_rate": 1.7964742381741435e-05, "loss": 0.0216, "step": 73530 }, { "grad_norm": 0.381778359413147, "learning_rate": 1.7952049007000627e-05, "loss": 0.0228, "step": 73540 }, { "grad_norm": 0.33136945962905884, "learning_rate": 1.7939359136975535e-05, "loss": 0.0198, "step": 73550 }, { "grad_norm": 0.2814822196960449, "learning_rate": 1.7926672773053932e-05, "loss": 0.0326, "step": 73560 }, { "grad_norm": 0.3281468152999878, "learning_rate": 1.791398991662317e-05, "loss": 0.0284, "step": 73570 }, { "grad_norm": 0.2180354744195938, "learning_rate": 1.7901310569070244e-05, "loss": 0.019, "step": 73580 }, { "grad_norm": 0.6045945882797241, "learning_rate": 1.7888634731781724e-05, "loss": 0.029, "step": 73590 }, { "grad_norm": 0.32430097460746765, "learning_rate": 1.7875962406143844e-05, "loss": 0.0226, "step": 73600 }, { "grad_norm": 0.6380521655082703, "learning_rate": 1.7863293593542403e-05, "loss": 0.0186, "step": 73610 }, { "grad_norm": 0.16917836666107178, "learning_rate": 1.7850628295362853e-05, "loss": 0.018, "step": 73620 }, { "grad_norm": 0.36521580815315247, "learning_rate": 1.7837966512990263e-05, "loss": 0.0189, "step": 73630 }, { "grad_norm": 0.3242306411266327, "learning_rate": 1.782530824780931e-05, "loss": 0.0263, "step": 73640 }, { "grad_norm": 0.33447957038879395, "learning_rate": 1.7812653501204263e-05, "loss": 0.0212, "step": 73650 }, { "grad_norm": 0.23050104081630707, "learning_rate": 1.780000227455903e-05, "loss": 0.0212, "step": 73660 }, { "grad_norm": 0.24049465358257294, "learning_rate": 1.778735456925716e-05, "loss": 0.0126, "step": 73670 }, { "grad_norm": 0.47227200865745544, "learning_rate": 1.7774710386681737e-05, "loss": 0.0188, "step": 73680 }, { "grad_norm": 0.27229198813438416, "learning_rate": 1.7762069728215545e-05, "loss": 0.0281, "step": 73690 }, { "grad_norm": 0.3831939697265625, "learning_rate": 1.7749432595240932e-05, "loss": 0.0189, "step": 73700 }, { "grad_norm": 1.8670294284820557, "learning_rate": 1.773679898913991e-05, "loss": 0.0282, "step": 73710 }, { "grad_norm": 0.5719268918037415, "learning_rate": 1.7724168911294015e-05, "loss": 0.0178, "step": 73720 }, { "grad_norm": 0.24504990875720978, "learning_rate": 1.7711542363084504e-05, "loss": 0.0156, "step": 73730 }, { "grad_norm": 0.22922632098197937, "learning_rate": 1.7698919345892158e-05, "loss": 0.0276, "step": 73740 }, { "grad_norm": 0.486294150352478, "learning_rate": 1.768629986109743e-05, "loss": 0.0274, "step": 73750 }, { "grad_norm": 0.18670669198036194, "learning_rate": 1.7673683910080357e-05, "loss": 0.0345, "step": 73760 }, { "grad_norm": 0.32496827840805054, "learning_rate": 1.7661071494220638e-05, "loss": 0.0287, "step": 73770 }, { "grad_norm": 0.37605226039886475, "learning_rate": 1.7648462614897498e-05, "loss": 0.0199, "step": 73780 }, { "grad_norm": 0.35372206568717957, "learning_rate": 1.7635857273489863e-05, "loss": 0.0203, "step": 73790 }, { "grad_norm": 0.1856709122657776, "learning_rate": 1.7623255471376198e-05, "loss": 0.0272, "step": 73800 }, { "grad_norm": 0.28757670521736145, "learning_rate": 1.7610657209934646e-05, "loss": 0.0278, "step": 73810 }, { "grad_norm": 0.29191330075263977, "learning_rate": 1.7598062490542915e-05, "loss": 0.0186, "step": 73820 }, { "grad_norm": 0.2836031913757324, "learning_rate": 1.7585471314578377e-05, "loss": 0.0237, "step": 73830 }, { "grad_norm": 0.38253122568130493, "learning_rate": 1.7572883683417945e-05, "loss": 0.0269, "step": 73840 }, { "grad_norm": 0.2148067057132721, "learning_rate": 1.7560299598438212e-05, "loss": 0.0246, "step": 73850 }, { "grad_norm": 0.39056292176246643, "learning_rate": 1.7547719061015334e-05, "loss": 0.0373, "step": 73860 }, { "grad_norm": 0.5244193077087402, "learning_rate": 1.75351420725251e-05, "loss": 0.0265, "step": 73870 }, { "grad_norm": 0.33789774775505066, "learning_rate": 1.7522568634342923e-05, "loss": 0.0228, "step": 73880 }, { "grad_norm": 0.26138556003570557, "learning_rate": 1.7509998747843825e-05, "loss": 0.0146, "step": 73890 }, { "grad_norm": 0.48919016122817993, "learning_rate": 1.7497432414402403e-05, "loss": 0.0175, "step": 73900 }, { "grad_norm": 0.45898300409317017, "learning_rate": 1.7484869635392903e-05, "loss": 0.0215, "step": 73910 }, { "grad_norm": 0.3856978118419647, "learning_rate": 1.747231041218919e-05, "loss": 0.0216, "step": 73920 }, { "grad_norm": 0.28657591342926025, "learning_rate": 1.7459754746164692e-05, "loss": 0.0203, "step": 73930 }, { "grad_norm": 0.32660242915153503, "learning_rate": 1.744720263869249e-05, "loss": 0.0202, "step": 73940 }, { "grad_norm": 0.5833056569099426, "learning_rate": 1.7434654091145263e-05, "loss": 0.0425, "step": 73950 }, { "grad_norm": 0.34434419870376587, "learning_rate": 1.742210910489532e-05, "loss": 0.0224, "step": 73960 }, { "grad_norm": 0.1998898833990097, "learning_rate": 1.7409567681314532e-05, "loss": 0.0189, "step": 73970 }, { "grad_norm": 0.4765065312385559, "learning_rate": 1.7397029821774434e-05, "loss": 0.0213, "step": 73980 }, { "grad_norm": 0.453700989484787, "learning_rate": 1.7384495527646126e-05, "loss": 0.0177, "step": 73990 }, { "grad_norm": 0.26679250597953796, "learning_rate": 1.737196480030035e-05, "loss": 0.0192, "step": 74000 }, { "grad_norm": 0.24295029044151306, "learning_rate": 1.7359437641107447e-05, "loss": 0.0313, "step": 74010 }, { "grad_norm": 0.33146947622299194, "learning_rate": 1.7346914051437387e-05, "loss": 0.0246, "step": 74020 }, { "grad_norm": 0.8460005521774292, "learning_rate": 1.733439403265969e-05, "loss": 0.0206, "step": 74030 }, { "grad_norm": 0.28238075971603394, "learning_rate": 1.732187758614357e-05, "loss": 0.0231, "step": 74040 }, { "grad_norm": 0.3074817359447479, "learning_rate": 1.730936471325777e-05, "loss": 0.018, "step": 74050 }, { "grad_norm": 0.21693874895572662, "learning_rate": 1.7296855415370718e-05, "loss": 0.0184, "step": 74060 }, { "grad_norm": 0.46360278129577637, "learning_rate": 1.728434969385035e-05, "loss": 0.0192, "step": 74070 }, { "grad_norm": 0.4856940805912018, "learning_rate": 1.7271847550064346e-05, "loss": 0.0219, "step": 74080 }, { "grad_norm": 0.3281841576099396, "learning_rate": 1.725934898537987e-05, "loss": 0.0309, "step": 74090 }, { "grad_norm": 0.2121138572692871, "learning_rate": 1.7246854001163783e-05, "loss": 0.0215, "step": 74100 }, { "grad_norm": 0.24399515986442566, "learning_rate": 1.7234362598782476e-05, "loss": 0.0118, "step": 74110 }, { "grad_norm": 0.33155837655067444, "learning_rate": 1.722187477960202e-05, "loss": 0.0243, "step": 74120 }, { "grad_norm": 0.20221374928951263, "learning_rate": 1.7209390544988057e-05, "loss": 0.0174, "step": 74130 }, { "grad_norm": 0.5479642748832703, "learning_rate": 1.7196909896305856e-05, "loss": 0.0279, "step": 74140 }, { "grad_norm": 0.5476112961769104, "learning_rate": 1.718443283492026e-05, "loss": 0.0237, "step": 74150 }, { "grad_norm": 0.5651607513427734, "learning_rate": 1.7171959362195746e-05, "loss": 0.0227, "step": 74160 }, { "grad_norm": 0.4415999948978424, "learning_rate": 1.715948947949642e-05, "loss": 0.018, "step": 74170 }, { "grad_norm": 0.5692526698112488, "learning_rate": 1.714702318818593e-05, "loss": 0.0211, "step": 74180 }, { "grad_norm": 0.36014166474342346, "learning_rate": 1.7134560489627592e-05, "loss": 0.0185, "step": 74190 }, { "grad_norm": 0.38848185539245605, "learning_rate": 1.712210138518431e-05, "loss": 0.0222, "step": 74200 }, { "grad_norm": 0.2576299011707306, "learning_rate": 1.71096458762186e-05, "loss": 0.0231, "step": 74210 }, { "grad_norm": 0.27071577310562134, "learning_rate": 1.7097193964092552e-05, "loss": 0.0184, "step": 74220 }, { "grad_norm": 0.4176015257835388, "learning_rate": 1.708474565016792e-05, "loss": 0.0212, "step": 74230 }, { "grad_norm": 0.48518988490104675, "learning_rate": 1.7072300935806e-05, "loss": 0.0233, "step": 74240 }, { "grad_norm": 0.2661946713924408, "learning_rate": 1.705985982236776e-05, "loss": 0.0145, "step": 74250 }, { "grad_norm": 0.2180866152048111, "learning_rate": 1.7047422311213697e-05, "loss": 0.0322, "step": 74260 }, { "grad_norm": 0.4061562120914459, "learning_rate": 1.703498840370401e-05, "loss": 0.0232, "step": 74270 }, { "grad_norm": 0.18711599707603455, "learning_rate": 1.702255810119842e-05, "loss": 0.0188, "step": 74280 }, { "grad_norm": 0.2772097587585449, "learning_rate": 1.7010131405056305e-05, "loss": 0.0158, "step": 74290 }, { "grad_norm": 0.2441437989473343, "learning_rate": 1.6997708316636613e-05, "loss": 0.0203, "step": 74300 }, { "grad_norm": 0.3484398126602173, "learning_rate": 1.6985288837297932e-05, "loss": 0.0187, "step": 74310 }, { "grad_norm": 0.5768960118293762, "learning_rate": 1.6972872968398394e-05, "loss": 0.0174, "step": 74320 }, { "grad_norm": 0.41078805923461914, "learning_rate": 1.6960460711295845e-05, "loss": 0.0226, "step": 74330 }, { "grad_norm": 0.22563384473323822, "learning_rate": 1.6948052067347623e-05, "loss": 0.0161, "step": 74340 }, { "grad_norm": 0.4731139540672302, "learning_rate": 1.693564703791075e-05, "loss": 0.0252, "step": 74350 }, { "grad_norm": 0.5055983662605286, "learning_rate": 1.6923245624341793e-05, "loss": 0.048, "step": 74360 }, { "grad_norm": 0.6761993169784546, "learning_rate": 1.6910847827996962e-05, "loss": 0.0374, "step": 74370 }, { "grad_norm": 0.47179391980171204, "learning_rate": 1.689845365023206e-05, "loss": 0.0219, "step": 74380 }, { "grad_norm": 0.525623083114624, "learning_rate": 1.6886063092402505e-05, "loss": 0.0197, "step": 74390 }, { "grad_norm": 0.25100407004356384, "learning_rate": 1.6873676155863317e-05, "loss": 0.0222, "step": 74400 }, { "grad_norm": 0.2340967357158661, "learning_rate": 1.6861292841969085e-05, "loss": 0.0152, "step": 74410 }, { "grad_norm": 0.2576577663421631, "learning_rate": 1.684891315207406e-05, "loss": 0.0176, "step": 74420 }, { "grad_norm": 0.6290748715400696, "learning_rate": 1.6836537087532034e-05, "loss": 0.0361, "step": 74430 }, { "grad_norm": 0.21053168177604675, "learning_rate": 1.6824164649696465e-05, "loss": 0.0239, "step": 74440 }, { "grad_norm": 1.1213922500610352, "learning_rate": 1.681179583992034e-05, "loss": 0.0387, "step": 74450 }, { "grad_norm": 0.3834111988544464, "learning_rate": 1.679943065955636e-05, "loss": 0.0195, "step": 74460 }, { "grad_norm": 0.32772520184516907, "learning_rate": 1.678706910995671e-05, "loss": 0.0163, "step": 74470 }, { "grad_norm": 0.5639414191246033, "learning_rate": 1.6774711192473262e-05, "loss": 0.0232, "step": 74480 }, { "grad_norm": 1.0529625415802002, "learning_rate": 1.676235690845743e-05, "loss": 0.021, "step": 74490 }, { "grad_norm": 0.31872138381004333, "learning_rate": 1.6750006259260285e-05, "loss": 0.0254, "step": 74500 }, { "grad_norm": 0.27370381355285645, "learning_rate": 1.673765924623244e-05, "loss": 0.0205, "step": 74510 }, { "grad_norm": 0.4688716232776642, "learning_rate": 1.67253158707242e-05, "loss": 0.0174, "step": 74520 }, { "grad_norm": 0.31648939847946167, "learning_rate": 1.671297613408537e-05, "loss": 0.0288, "step": 74530 }, { "grad_norm": 0.4347221255302429, "learning_rate": 1.6700640037665444e-05, "loss": 0.0178, "step": 74540 }, { "grad_norm": 0.39189568161964417, "learning_rate": 1.668830758281344e-05, "loss": 0.0252, "step": 74550 }, { "grad_norm": 0.655859112739563, "learning_rate": 1.6675978770878052e-05, "loss": 0.0297, "step": 74560 }, { "grad_norm": 0.35609501600265503, "learning_rate": 1.6663653603207506e-05, "loss": 0.0189, "step": 74570 }, { "grad_norm": 0.6960703134536743, "learning_rate": 1.6651332081149685e-05, "loss": 0.0238, "step": 74580 }, { "grad_norm": 0.522392213344574, "learning_rate": 1.6639014206052046e-05, "loss": 0.0244, "step": 74590 }, { "grad_norm": 0.30701494216918945, "learning_rate": 1.6626699979261657e-05, "loss": 0.0165, "step": 74600 }, { "grad_norm": 0.2746400833129883, "learning_rate": 1.6614389402125197e-05, "loss": 0.0207, "step": 74610 }, { "grad_norm": 0.2700764238834381, "learning_rate": 1.66020824759889e-05, "loss": 0.0299, "step": 74620 }, { "grad_norm": 0.32225263118743896, "learning_rate": 1.6589779202198675e-05, "loss": 0.0193, "step": 74630 }, { "grad_norm": 0.5376808047294617, "learning_rate": 1.6577479582099924e-05, "loss": 0.0327, "step": 74640 }, { "grad_norm": 0.27560240030288696, "learning_rate": 1.6565183617037788e-05, "loss": 0.0195, "step": 74650 }, { "grad_norm": 0.4302312433719635, "learning_rate": 1.6552891308356894e-05, "loss": 0.0193, "step": 74660 }, { "grad_norm": 0.3395857810974121, "learning_rate": 1.6540602657401526e-05, "loss": 0.0237, "step": 74670 }, { "grad_norm": 0.6337274312973022, "learning_rate": 1.6528317665515537e-05, "loss": 0.0295, "step": 74680 }, { "grad_norm": 0.570875346660614, "learning_rate": 1.6516036334042416e-05, "loss": 0.0181, "step": 74690 }, { "grad_norm": 0.37789613008499146, "learning_rate": 1.650375866432519e-05, "loss": 0.0245, "step": 74700 }, { "grad_norm": 1.0555046796798706, "learning_rate": 1.649148465770659e-05, "loss": 0.0304, "step": 74710 }, { "grad_norm": 0.4750123620033264, "learning_rate": 1.647921431552883e-05, "loss": 0.0249, "step": 74720 }, { "grad_norm": 0.39367592334747314, "learning_rate": 1.646694763913381e-05, "loss": 0.0233, "step": 74730 }, { "grad_norm": 0.3592546880245209, "learning_rate": 1.6454684629862966e-05, "loss": 0.0189, "step": 74740 }, { "grad_norm": 0.38708481192588806, "learning_rate": 1.6442425289057388e-05, "loss": 0.0231, "step": 74750 }, { "grad_norm": 0.2944202423095703, "learning_rate": 1.6430169618057718e-05, "loss": 0.0218, "step": 74760 }, { "grad_norm": 0.33575356006622314, "learning_rate": 1.641791761820422e-05, "loss": 0.0193, "step": 74770 }, { "grad_norm": 0.4286958873271942, "learning_rate": 1.640566929083676e-05, "loss": 0.0218, "step": 74780 }, { "grad_norm": 0.3673352897167206, "learning_rate": 1.6393424637294823e-05, "loss": 0.0202, "step": 74790 }, { "grad_norm": 0.19363225996494293, "learning_rate": 1.638118365891742e-05, "loss": 0.0153, "step": 74800 }, { "grad_norm": 0.31234872341156006, "learning_rate": 1.636894635704323e-05, "loss": 0.0195, "step": 74810 }, { "grad_norm": 0.49092379212379456, "learning_rate": 1.6356712733010517e-05, "loss": 0.0299, "step": 74820 }, { "grad_norm": 0.45049333572387695, "learning_rate": 1.6344482788157096e-05, "loss": 0.024, "step": 74830 }, { "grad_norm": 0.3743048906326294, "learning_rate": 1.6332256523820444e-05, "loss": 0.017, "step": 74840 }, { "grad_norm": 0.24966798722743988, "learning_rate": 1.6320033941337597e-05, "loss": 0.0146, "step": 74850 }, { "grad_norm": 0.4717600643634796, "learning_rate": 1.6307815042045216e-05, "loss": 0.0172, "step": 74860 }, { "grad_norm": 0.20566216111183167, "learning_rate": 1.6295599827279513e-05, "loss": 0.0425, "step": 74870 }, { "grad_norm": 0.4203735888004303, "learning_rate": 1.6283388298376356e-05, "loss": 0.0219, "step": 74880 }, { "grad_norm": 0.3971867859363556, "learning_rate": 1.6271180456671125e-05, "loss": 0.025, "step": 74890 }, { "grad_norm": 1.5520806312561035, "learning_rate": 1.625897630349893e-05, "loss": 0.036, "step": 74900 }, { "grad_norm": 0.319839745759964, "learning_rate": 1.6246775840194335e-05, "loss": 0.0152, "step": 74910 }, { "grad_norm": 0.5240956544876099, "learning_rate": 1.6234579068091604e-05, "loss": 0.02, "step": 74920 }, { "grad_norm": 0.4950752556324005, "learning_rate": 1.6222385988524524e-05, "loss": 0.0362, "step": 74930 }, { "grad_norm": 0.30348795652389526, "learning_rate": 1.6210196602826544e-05, "loss": 0.0257, "step": 74940 }, { "grad_norm": 0.424442857503891, "learning_rate": 1.6198010912330646e-05, "loss": 0.0204, "step": 74950 }, { "grad_norm": 0.3743257224559784, "learning_rate": 1.6185828918369456e-05, "loss": 0.0163, "step": 74960 }, { "grad_norm": 0.4771255552768707, "learning_rate": 1.617365062227517e-05, "loss": 0.0267, "step": 74970 }, { "grad_norm": 0.6863628625869751, "learning_rate": 1.616147602537962e-05, "loss": 0.0272, "step": 74980 }, { "grad_norm": 0.2981652319431305, "learning_rate": 1.614930512901415e-05, "loss": 0.0217, "step": 74990 }, { "grad_norm": 0.3088322579860687, "learning_rate": 1.6137137934509804e-05, "loss": 0.0206, "step": 75000 }, { "grad_norm": 0.21002735197544098, "learning_rate": 1.6124974443197123e-05, "loss": 0.0232, "step": 75010 }, { "grad_norm": 0.4607020616531372, "learning_rate": 1.6112814656406306e-05, "loss": 0.0293, "step": 75020 }, { "grad_norm": 0.22660508751869202, "learning_rate": 1.6100658575467125e-05, "loss": 0.0287, "step": 75030 }, { "grad_norm": 0.5410041213035583, "learning_rate": 1.6088506201708974e-05, "loss": 0.0264, "step": 75040 }, { "grad_norm": 0.2476169914007187, "learning_rate": 1.6076357536460785e-05, "loss": 0.0171, "step": 75050 }, { "grad_norm": 0.4622561037540436, "learning_rate": 1.606421258105113e-05, "loss": 0.0274, "step": 75060 }, { "grad_norm": 0.5574555993080139, "learning_rate": 1.6052071336808183e-05, "loss": 0.0266, "step": 75070 }, { "grad_norm": 0.630345344543457, "learning_rate": 1.6039933805059658e-05, "loss": 0.0205, "step": 75080 }, { "grad_norm": 0.27298370003700256, "learning_rate": 1.6027799987132915e-05, "loss": 0.0245, "step": 75090 }, { "grad_norm": 0.3490273952484131, "learning_rate": 1.6015669884354888e-05, "loss": 0.0168, "step": 75100 }, { "grad_norm": 0.28352344036102295, "learning_rate": 1.6003543498052125e-05, "loss": 0.0172, "step": 75110 }, { "grad_norm": 0.6598345041275024, "learning_rate": 1.5991420829550715e-05, "loss": 0.0292, "step": 75120 }, { "grad_norm": 0.5839701294898987, "learning_rate": 1.5979301880176405e-05, "loss": 0.0223, "step": 75130 }, { "grad_norm": 0.12792183458805084, "learning_rate": 1.5967186651254485e-05, "loss": 0.0202, "step": 75140 }, { "grad_norm": 0.32526761293411255, "learning_rate": 1.5955075144109853e-05, "loss": 0.0272, "step": 75150 }, { "grad_norm": 1.0412777662277222, "learning_rate": 1.594296736006703e-05, "loss": 0.0219, "step": 75160 }, { "grad_norm": 0.6234603524208069, "learning_rate": 1.59308633004501e-05, "loss": 0.0205, "step": 75170 }, { "grad_norm": 0.254707932472229, "learning_rate": 1.5918762966582724e-05, "loss": 0.0251, "step": 75180 }, { "grad_norm": 0.19469216465950012, "learning_rate": 1.5906666359788203e-05, "loss": 0.0206, "step": 75190 }, { "grad_norm": 0.63755202293396, "learning_rate": 1.589457348138938e-05, "loss": 0.0221, "step": 75200 }, { "grad_norm": 0.29373493790626526, "learning_rate": 1.588248433270872e-05, "loss": 0.0165, "step": 75210 }, { "grad_norm": 0.24278071522712708, "learning_rate": 1.587039891506828e-05, "loss": 0.0209, "step": 75220 }, { "grad_norm": 0.2707423269748688, "learning_rate": 1.5858317229789715e-05, "loss": 0.0225, "step": 75230 }, { "grad_norm": 0.25412681698799133, "learning_rate": 1.584623927819423e-05, "loss": 0.0164, "step": 75240 }, { "grad_norm": 0.2624457776546478, "learning_rate": 1.583416506160269e-05, "loss": 0.0172, "step": 75250 }, { "grad_norm": 0.5157018899917603, "learning_rate": 1.5822094581335472e-05, "loss": 0.0222, "step": 75260 }, { "grad_norm": 0.35225555300712585, "learning_rate": 1.5810027838712604e-05, "loss": 0.0244, "step": 75270 }, { "grad_norm": 0.2549506425857544, "learning_rate": 1.5797964835053687e-05, "loss": 0.0207, "step": 75280 }, { "grad_norm": 0.30573102831840515, "learning_rate": 1.5785905571677933e-05, "loss": 0.0266, "step": 75290 }, { "grad_norm": 0.935799777507782, "learning_rate": 1.5773850049904086e-05, "loss": 0.0199, "step": 75300 }, { "grad_norm": 0.4225609302520752, "learning_rate": 1.5761798271050543e-05, "loss": 0.0353, "step": 75310 }, { "grad_norm": 0.5646362900733948, "learning_rate": 1.5749750236435277e-05, "loss": 0.0197, "step": 75320 }, { "grad_norm": 0.3832706809043884, "learning_rate": 1.5737705947375815e-05, "loss": 0.0232, "step": 75330 }, { "grad_norm": 0.27379634976387024, "learning_rate": 1.5725665405189322e-05, "loss": 0.0212, "step": 75340 }, { "grad_norm": 0.48058703541755676, "learning_rate": 1.571362861119253e-05, "loss": 0.0198, "step": 75350 }, { "grad_norm": 0.49478965997695923, "learning_rate": 1.570159556670177e-05, "loss": 0.0219, "step": 75360 }, { "grad_norm": 0.33148089051246643, "learning_rate": 1.5689566273032946e-05, "loss": 0.0214, "step": 75370 }, { "grad_norm": 0.3222494125366211, "learning_rate": 1.5677540731501583e-05, "loss": 0.0172, "step": 75380 }, { "grad_norm": 0.58527010679245, "learning_rate": 1.5665518943422746e-05, "loss": 0.0369, "step": 75390 }, { "grad_norm": 0.31645864248275757, "learning_rate": 1.5653500910111134e-05, "loss": 0.0181, "step": 75400 }, { "grad_norm": 0.31200525164604187, "learning_rate": 1.5641486632881023e-05, "loss": 0.0226, "step": 75410 }, { "grad_norm": 0.37405309081077576, "learning_rate": 1.56294761130463e-05, "loss": 0.0281, "step": 75420 }, { "grad_norm": 0.3522469103336334, "learning_rate": 1.5617469351920373e-05, "loss": 0.0276, "step": 75430 }, { "grad_norm": 0.3278689384460449, "learning_rate": 1.5605466350816315e-05, "loss": 0.0242, "step": 75440 }, { "grad_norm": 0.3295767903327942, "learning_rate": 1.559346711104674e-05, "loss": 0.0159, "step": 75450 }, { "grad_norm": 0.669693112373352, "learning_rate": 1.5581471633923883e-05, "loss": 0.019, "step": 75460 }, { "grad_norm": 0.5274083614349365, "learning_rate": 1.5569479920759505e-05, "loss": 0.0174, "step": 75470 }, { "grad_norm": 0.3513161242008209, "learning_rate": 1.5557491972865072e-05, "loss": 0.0203, "step": 75480 }, { "grad_norm": 0.37782955169677734, "learning_rate": 1.554550779155152e-05, "loss": 0.0243, "step": 75490 }, { "grad_norm": 0.3550550043582916, "learning_rate": 1.5533527378129453e-05, "loss": 0.0324, "step": 75500 }, { "grad_norm": 0.3682153522968292, "learning_rate": 1.552155073390899e-05, "loss": 0.0366, "step": 75510 }, { "grad_norm": 0.2766743302345276, "learning_rate": 1.5509577860199903e-05, "loss": 0.0255, "step": 75520 }, { "grad_norm": 0.38980165123939514, "learning_rate": 1.5497608758311525e-05, "loss": 0.0176, "step": 75530 }, { "grad_norm": 0.1797700971364975, "learning_rate": 1.5485643429552797e-05, "loss": 0.0196, "step": 75540 }, { "grad_norm": 0.3837224841117859, "learning_rate": 1.5473681875232197e-05, "loss": 0.0199, "step": 75550 }, { "grad_norm": 0.7053448557853699, "learning_rate": 1.5461724096657837e-05, "loss": 0.0252, "step": 75560 }, { "grad_norm": 0.2610953152179718, "learning_rate": 1.5449770095137416e-05, "loss": 0.0154, "step": 75570 }, { "grad_norm": 0.2115064263343811, "learning_rate": 1.5437819871978178e-05, "loss": 0.015, "step": 75580 }, { "grad_norm": 0.30472588539123535, "learning_rate": 1.5425873428486992e-05, "loss": 0.018, "step": 75590 }, { "grad_norm": 0.7013052105903625, "learning_rate": 1.5413930765970312e-05, "loss": 0.0196, "step": 75600 }, { "grad_norm": 0.1958901733160019, "learning_rate": 1.5401991885734175e-05, "loss": 0.0385, "step": 75610 }, { "grad_norm": 0.47917401790618896, "learning_rate": 1.5390056789084174e-05, "loss": 0.0196, "step": 75620 }, { "grad_norm": 0.40431976318359375, "learning_rate": 1.5378125477325545e-05, "loss": 0.028, "step": 75630 }, { "grad_norm": 0.23497700691223145, "learning_rate": 1.5366197951763046e-05, "loss": 0.0192, "step": 75640 }, { "grad_norm": 0.23893749713897705, "learning_rate": 1.5354274213701075e-05, "loss": 0.0231, "step": 75650 }, { "grad_norm": 0.4131861925125122, "learning_rate": 1.5342354264443566e-05, "loss": 0.0254, "step": 75660 }, { "grad_norm": 0.42535221576690674, "learning_rate": 1.533043810529411e-05, "loss": 0.0171, "step": 75670 }, { "grad_norm": 0.4398181736469269, "learning_rate": 1.5318525737555807e-05, "loss": 0.0268, "step": 75680 }, { "grad_norm": 0.3710341453552246, "learning_rate": 1.53066171625314e-05, "loss": 0.0292, "step": 75690 }, { "grad_norm": 0.3266353905200958, "learning_rate": 1.529471238152317e-05, "loss": 0.0149, "step": 75700 }, { "grad_norm": 0.7363366484642029, "learning_rate": 1.5282811395833023e-05, "loss": 0.0332, "step": 75710 }, { "grad_norm": 0.3435134291648865, "learning_rate": 1.5270914206762397e-05, "loss": 0.0162, "step": 75720 }, { "grad_norm": 0.2686227560043335, "learning_rate": 1.525902081561241e-05, "loss": 0.0165, "step": 75730 }, { "grad_norm": 0.552375316619873, "learning_rate": 1.5247131223683664e-05, "loss": 0.026, "step": 75740 }, { "grad_norm": 0.3167654275894165, "learning_rate": 1.5235245432276408e-05, "loss": 0.0273, "step": 75750 }, { "grad_norm": 0.19817225635051727, "learning_rate": 1.5223363442690424e-05, "loss": 0.0163, "step": 75760 }, { "grad_norm": 0.222063347697258, "learning_rate": 1.5211485256225133e-05, "loss": 0.014, "step": 75770 }, { "grad_norm": 0.4694114029407501, "learning_rate": 1.519961087417951e-05, "loss": 0.0205, "step": 75780 }, { "grad_norm": 0.1625031679868698, "learning_rate": 1.5187740297852137e-05, "loss": 0.0143, "step": 75790 }, { "grad_norm": 0.371656209230423, "learning_rate": 1.517587352854113e-05, "loss": 0.0286, "step": 75800 }, { "grad_norm": 0.1942983865737915, "learning_rate": 1.516401056754423e-05, "loss": 0.0267, "step": 75810 }, { "grad_norm": 0.5908242464065552, "learning_rate": 1.5152151416158783e-05, "loss": 0.0193, "step": 75820 }, { "grad_norm": 0.2798382341861725, "learning_rate": 1.5140296075681642e-05, "loss": 0.0187, "step": 75830 }, { "grad_norm": 0.22631317377090454, "learning_rate": 1.5128444547409331e-05, "loss": 0.0237, "step": 75840 }, { "grad_norm": 0.32005906105041504, "learning_rate": 1.5116596832637864e-05, "loss": 0.0164, "step": 75850 }, { "grad_norm": 0.2723402976989746, "learning_rate": 1.5104752932662952e-05, "loss": 0.0213, "step": 75860 }, { "grad_norm": 0.4616855978965759, "learning_rate": 1.5092912848779778e-05, "loss": 0.0163, "step": 75870 }, { "grad_norm": 0.584550678730011, "learning_rate": 1.5081076582283194e-05, "loss": 0.0227, "step": 75880 }, { "grad_norm": 0.736713707447052, "learning_rate": 1.5069244134467553e-05, "loss": 0.0301, "step": 75890 }, { "grad_norm": 0.8151888847351074, "learning_rate": 1.5057415506626876e-05, "loss": 0.0323, "step": 75900 }, { "grad_norm": 0.25426509976387024, "learning_rate": 1.5045590700054668e-05, "loss": 0.0176, "step": 75910 }, { "grad_norm": 0.40820956230163574, "learning_rate": 1.5033769716044144e-05, "loss": 0.0262, "step": 75920 }, { "grad_norm": 0.5339818596839905, "learning_rate": 1.5021952555887975e-05, "loss": 0.0197, "step": 75930 }, { "grad_norm": 0.1986764371395111, "learning_rate": 1.50101392208785e-05, "loss": 0.0179, "step": 75940 }, { "grad_norm": 0.3248695433139801, "learning_rate": 1.4998329712307579e-05, "loss": 0.0154, "step": 75950 }, { "grad_norm": 0.35401591658592224, "learning_rate": 1.4986524031466708e-05, "loss": 0.0206, "step": 75960 }, { "grad_norm": 0.3632813096046448, "learning_rate": 1.4974722179646916e-05, "loss": 0.0215, "step": 75970 }, { "grad_norm": 0.5955327749252319, "learning_rate": 1.4962924158138847e-05, "loss": 0.035, "step": 75980 }, { "grad_norm": 0.23430036008358002, "learning_rate": 1.4951129968232714e-05, "loss": 0.0341, "step": 75990 }, { "grad_norm": 1.5945987701416016, "learning_rate": 1.4939339611218329e-05, "loss": 0.0255, "step": 76000 }, { "grad_norm": 0.9517929553985596, "learning_rate": 1.4927553088385032e-05, "loss": 0.0263, "step": 76010 }, { "grad_norm": 0.23980502784252167, "learning_rate": 1.4915770401021801e-05, "loss": 0.0291, "step": 76020 }, { "grad_norm": 0.2648184895515442, "learning_rate": 1.4903991550417185e-05, "loss": 0.0162, "step": 76030 }, { "grad_norm": 0.8533961176872253, "learning_rate": 1.4892216537859277e-05, "loss": 0.0208, "step": 76040 }, { "grad_norm": 0.22694025933742523, "learning_rate": 1.4880445364635775e-05, "loss": 0.0224, "step": 76050 }, { "grad_norm": 0.2888917922973633, "learning_rate": 1.4868678032033973e-05, "loss": 0.0174, "step": 76060 }, { "grad_norm": 0.37874871492385864, "learning_rate": 1.4856914541340739e-05, "loss": 0.0152, "step": 76070 }, { "grad_norm": 0.21216754615306854, "learning_rate": 1.4845154893842472e-05, "loss": 0.026, "step": 76080 }, { "grad_norm": 0.27391213178634644, "learning_rate": 1.4833399090825234e-05, "loss": 0.0282, "step": 76090 }, { "grad_norm": 0.5519241094589233, "learning_rate": 1.4821647133574562e-05, "loss": 0.0181, "step": 76100 }, { "grad_norm": 1.3452990055084229, "learning_rate": 1.4809899023375706e-05, "loss": 0.0283, "step": 76110 }, { "grad_norm": 0.23647728562355042, "learning_rate": 1.4798154761513366e-05, "loss": 0.0211, "step": 76120 }, { "grad_norm": 0.2938007116317749, "learning_rate": 1.4786414349271916e-05, "loss": 0.0168, "step": 76130 }, { "grad_norm": 0.25130975246429443, "learning_rate": 1.4774677787935226e-05, "loss": 0.016, "step": 76140 }, { "grad_norm": 0.2671812176704407, "learning_rate": 1.4762945078786838e-05, "loss": 0.0258, "step": 76150 }, { "grad_norm": 0.6141445636749268, "learning_rate": 1.4751216223109777e-05, "loss": 0.034, "step": 76160 }, { "grad_norm": 0.25654149055480957, "learning_rate": 1.4739491222186714e-05, "loss": 0.0189, "step": 76170 }, { "grad_norm": 0.4715595841407776, "learning_rate": 1.4727770077299879e-05, "loss": 0.0254, "step": 76180 }, { "grad_norm": 0.34106847643852234, "learning_rate": 1.4716052789731094e-05, "loss": 0.0211, "step": 76190 }, { "grad_norm": 0.3864138126373291, "learning_rate": 1.470433936076171e-05, "loss": 0.0217, "step": 76200 }, { "grad_norm": 0.3602220416069031, "learning_rate": 1.4692629791672719e-05, "loss": 0.0151, "step": 76210 }, { "grad_norm": 0.5210683345794678, "learning_rate": 1.468092408374464e-05, "loss": 0.0278, "step": 76220 }, { "grad_norm": 0.2763836085796356, "learning_rate": 1.4669222238257602e-05, "loss": 0.0146, "step": 76230 }, { "grad_norm": 0.3490007221698761, "learning_rate": 1.4657524256491301e-05, "loss": 0.018, "step": 76240 }, { "grad_norm": 0.30417948961257935, "learning_rate": 1.464583013972502e-05, "loss": 0.0165, "step": 76250 }, { "grad_norm": 0.623517632484436, "learning_rate": 1.4634139889237591e-05, "loss": 0.0353, "step": 76260 }, { "grad_norm": 0.6040552258491516, "learning_rate": 1.4622453506307448e-05, "loss": 0.0214, "step": 76270 }, { "grad_norm": 0.6246675252914429, "learning_rate": 1.461077099221262e-05, "loss": 0.0315, "step": 76280 }, { "grad_norm": 0.15923945605754852, "learning_rate": 1.4599092348230652e-05, "loss": 0.0176, "step": 76290 }, { "grad_norm": 0.6918943524360657, "learning_rate": 1.4587417575638724e-05, "loss": 0.0176, "step": 76300 }, { "grad_norm": 0.4312530755996704, "learning_rate": 1.4575746675713563e-05, "loss": 0.0239, "step": 76310 }, { "grad_norm": 0.47351956367492676, "learning_rate": 1.4564079649731504e-05, "loss": 0.022, "step": 76320 }, { "grad_norm": 0.3994336426258087, "learning_rate": 1.4552416498968402e-05, "loss": 0.0276, "step": 76330 }, { "grad_norm": 0.39652153849601746, "learning_rate": 1.4540757224699747e-05, "loss": 0.025, "step": 76340 }, { "grad_norm": 0.46205535531044006, "learning_rate": 1.452910182820056e-05, "loss": 0.0285, "step": 76350 }, { "grad_norm": 0.2091328501701355, "learning_rate": 1.4517450310745468e-05, "loss": 0.0151, "step": 76360 }, { "grad_norm": 0.3962766230106354, "learning_rate": 1.450580267360866e-05, "loss": 0.0305, "step": 76370 }, { "grad_norm": 0.4103747308254242, "learning_rate": 1.4494158918063922e-05, "loss": 0.021, "step": 76380 }, { "grad_norm": 0.2585514485836029, "learning_rate": 1.4482519045384568e-05, "loss": 0.0165, "step": 76390 }, { "grad_norm": 2.8437583446502686, "learning_rate": 1.4470883056843548e-05, "loss": 0.0265, "step": 76400 }, { "grad_norm": 0.454918771982193, "learning_rate": 1.4459250953713322e-05, "loss": 0.0188, "step": 76410 }, { "grad_norm": 0.5680842995643616, "learning_rate": 1.444762273726598e-05, "loss": 0.0195, "step": 76420 }, { "grad_norm": 0.23482267558574677, "learning_rate": 1.443599840877316e-05, "loss": 0.0151, "step": 76430 }, { "grad_norm": 0.30072587728500366, "learning_rate": 1.4424377969506103e-05, "loss": 0.0195, "step": 76440 }, { "grad_norm": 0.2793407440185547, "learning_rate": 1.4412761420735565e-05, "loss": 0.0217, "step": 76450 }, { "grad_norm": 0.7333107590675354, "learning_rate": 1.4401148763731953e-05, "loss": 0.0336, "step": 76460 }, { "grad_norm": 0.7213059067726135, "learning_rate": 1.4389539999765173e-05, "loss": 0.0243, "step": 76470 }, { "grad_norm": 0.29020923376083374, "learning_rate": 1.4377935130104758e-05, "loss": 0.0206, "step": 76480 }, { "grad_norm": 0.28964799642562866, "learning_rate": 1.4366334156019801e-05, "loss": 0.0183, "step": 76490 }, { "grad_norm": 0.46026837825775146, "learning_rate": 1.435473707877898e-05, "loss": 0.0183, "step": 76500 }, { "grad_norm": 0.22117885947227478, "learning_rate": 1.4343143899650508e-05, "loss": 0.0233, "step": 76510 }, { "grad_norm": 0.24238590896129608, "learning_rate": 1.4331554619902204e-05, "loss": 0.0218, "step": 76520 }, { "grad_norm": 0.19517838954925537, "learning_rate": 1.4319969240801484e-05, "loss": 0.0252, "step": 76530 }, { "grad_norm": 0.42305266857147217, "learning_rate": 1.430838776361526e-05, "loss": 0.0251, "step": 76540 }, { "grad_norm": 0.13605976104736328, "learning_rate": 1.4296810189610093e-05, "loss": 0.0167, "step": 76550 }, { "grad_norm": 0.3565489649772644, "learning_rate": 1.4285236520052087e-05, "loss": 0.0228, "step": 76560 }, { "grad_norm": 0.5428263545036316, "learning_rate": 1.4273666756206933e-05, "loss": 0.0224, "step": 76570 }, { "grad_norm": 0.29956838488578796, "learning_rate": 1.4262100899339848e-05, "loss": 0.0262, "step": 76580 }, { "grad_norm": 0.8076500296592712, "learning_rate": 1.4250538950715696e-05, "loss": 0.035, "step": 76590 }, { "grad_norm": 0.1483471393585205, "learning_rate": 1.4238980911598842e-05, "loss": 0.015, "step": 76600 }, { "grad_norm": 0.4440970718860626, "learning_rate": 1.4227426783253267e-05, "loss": 0.0155, "step": 76610 }, { "grad_norm": 0.3040172755718231, "learning_rate": 1.4215876566942521e-05, "loss": 0.0318, "step": 76620 }, { "grad_norm": 0.3683445155620575, "learning_rate": 1.4204330263929721e-05, "loss": 0.0322, "step": 76630 }, { "grad_norm": 0.24884898960590363, "learning_rate": 1.4192787875477537e-05, "loss": 0.0211, "step": 76640 }, { "grad_norm": 0.9434090852737427, "learning_rate": 1.4181249402848246e-05, "loss": 0.0213, "step": 76650 }, { "grad_norm": 0.3620006740093231, "learning_rate": 1.4169714847303656e-05, "loss": 0.024, "step": 76660 }, { "grad_norm": 0.4035620391368866, "learning_rate": 1.4158184210105197e-05, "loss": 0.0269, "step": 76670 }, { "grad_norm": 0.4072897434234619, "learning_rate": 1.4146657492513787e-05, "loss": 0.0293, "step": 76680 }, { "grad_norm": 0.29654210805892944, "learning_rate": 1.4135134695790047e-05, "loss": 0.0182, "step": 76690 }, { "grad_norm": 0.4352398216724396, "learning_rate": 1.4123615821194037e-05, "loss": 0.0234, "step": 76700 }, { "grad_norm": 0.22895930707454681, "learning_rate": 1.4112100869985473e-05, "loss": 0.0162, "step": 76710 }, { "grad_norm": 0.19436219334602356, "learning_rate": 1.4100589843423589e-05, "loss": 0.0148, "step": 76720 }, { "grad_norm": 0.2840072810649872, "learning_rate": 1.408908274276723e-05, "loss": 0.0255, "step": 76730 }, { "grad_norm": 0.2390078902244568, "learning_rate": 1.4077579569274786e-05, "loss": 0.0146, "step": 76740 }, { "grad_norm": 0.3220824599266052, "learning_rate": 1.4066080324204245e-05, "loss": 0.0383, "step": 76750 }, { "grad_norm": 0.21742190420627594, "learning_rate": 1.405458500881312e-05, "loss": 0.0132, "step": 76760 }, { "grad_norm": 0.3081470727920532, "learning_rate": 1.4043093624358538e-05, "loss": 0.0147, "step": 76770 }, { "grad_norm": 0.17340010404586792, "learning_rate": 1.4031606172097189e-05, "loss": 0.0204, "step": 76780 }, { "grad_norm": 0.37698525190353394, "learning_rate": 1.40201226532853e-05, "loss": 0.0279, "step": 76790 }, { "grad_norm": 0.2915542423725128, "learning_rate": 1.4008643069178701e-05, "loss": 0.0222, "step": 76800 }, { "grad_norm": 0.5051752924919128, "learning_rate": 1.399716742103278e-05, "loss": 0.0176, "step": 76810 }, { "grad_norm": 0.34732189774513245, "learning_rate": 1.398569571010252e-05, "loss": 0.0169, "step": 76820 }, { "grad_norm": 0.31498053669929504, "learning_rate": 1.3974227937642414e-05, "loss": 0.0254, "step": 76830 }, { "grad_norm": 1.0798429250717163, "learning_rate": 1.3962764104906595e-05, "loss": 0.0234, "step": 76840 }, { "grad_norm": 0.5199583768844604, "learning_rate": 1.3951304213148696e-05, "loss": 0.0233, "step": 76850 }, { "grad_norm": 0.4199957549571991, "learning_rate": 1.3939848263621979e-05, "loss": 0.0246, "step": 76860 }, { "grad_norm": 0.5116413235664368, "learning_rate": 1.3928396257579219e-05, "loss": 0.0178, "step": 76870 }, { "grad_norm": 0.5590531229972839, "learning_rate": 1.3916948196272833e-05, "loss": 0.0317, "step": 76880 }, { "grad_norm": 0.6331314444541931, "learning_rate": 1.3905504080954734e-05, "loss": 0.0207, "step": 76890 }, { "grad_norm": 0.7094355821609497, "learning_rate": 1.3894063912876454e-05, "loss": 0.0285, "step": 76900 }, { "grad_norm": 0.4970118999481201, "learning_rate": 1.3882627693289046e-05, "loss": 0.0205, "step": 76910 }, { "grad_norm": 0.26811501383781433, "learning_rate": 1.3871195423443183e-05, "loss": 0.0274, "step": 76920 }, { "grad_norm": 0.15279625356197357, "learning_rate": 1.3859767104589044e-05, "loss": 0.0191, "step": 76930 }, { "grad_norm": 0.6903086304664612, "learning_rate": 1.3848342737976467e-05, "loss": 0.0234, "step": 76940 }, { "grad_norm": 0.4197452664375305, "learning_rate": 1.3836922324854757e-05, "loss": 0.0272, "step": 76950 }, { "grad_norm": 0.289502888917923, "learning_rate": 1.382550586647287e-05, "loss": 0.026, "step": 76960 }, { "grad_norm": 0.31199562549591064, "learning_rate": 1.3814093364079257e-05, "loss": 0.0437, "step": 76970 }, { "grad_norm": 0.42360833287239075, "learning_rate": 1.3802684818921991e-05, "loss": 0.0207, "step": 76980 }, { "grad_norm": 0.2946116626262665, "learning_rate": 1.3791280232248687e-05, "loss": 0.0236, "step": 76990 }, { "grad_norm": 0.47148001194000244, "learning_rate": 1.3779879605306555e-05, "loss": 0.0247, "step": 77000 }, { "grad_norm": 0.24659278988838196, "learning_rate": 1.376848293934232e-05, "loss": 0.0133, "step": 77010 }, { "grad_norm": 0.3068148195743561, "learning_rate": 1.3757090235602316e-05, "loss": 0.0229, "step": 77020 }, { "grad_norm": 0.404348224401474, "learning_rate": 1.3745701495332447e-05, "loss": 0.0255, "step": 77030 }, { "grad_norm": 0.36960121989250183, "learning_rate": 1.3734316719778145e-05, "loss": 0.0236, "step": 77040 }, { "grad_norm": 0.1219753548502922, "learning_rate": 1.3722935910184454e-05, "loss": 0.0225, "step": 77050 }, { "grad_norm": 0.30405646562576294, "learning_rate": 1.3711559067795921e-05, "loss": 0.0193, "step": 77060 }, { "grad_norm": 0.48917868733406067, "learning_rate": 1.3700186193856756e-05, "loss": 0.0249, "step": 77070 }, { "grad_norm": 0.318799763917923, "learning_rate": 1.368881728961064e-05, "loss": 0.0148, "step": 77080 }, { "grad_norm": 0.1434241235256195, "learning_rate": 1.3677452356300885e-05, "loss": 0.0168, "step": 77090 }, { "grad_norm": 0.34035050868988037, "learning_rate": 1.3666091395170321e-05, "loss": 0.0177, "step": 77100 }, { "grad_norm": 0.17621371150016785, "learning_rate": 1.3654734407461384e-05, "loss": 0.02, "step": 77110 }, { "grad_norm": 0.27278387546539307, "learning_rate": 1.3643381394416022e-05, "loss": 0.0218, "step": 77120 }, { "grad_norm": 0.33374887704849243, "learning_rate": 1.3632032357275838e-05, "loss": 0.0228, "step": 77130 }, { "grad_norm": 0.4099762737751007, "learning_rate": 1.3620687297281898e-05, "loss": 0.0242, "step": 77140 }, { "grad_norm": 0.4469817578792572, "learning_rate": 1.3609346215674917e-05, "loss": 0.0228, "step": 77150 }, { "grad_norm": 0.27466002106666565, "learning_rate": 1.3598009113695109e-05, "loss": 0.0251, "step": 77160 }, { "grad_norm": 0.26160696148872375, "learning_rate": 1.358667599258231e-05, "loss": 0.0253, "step": 77170 }, { "grad_norm": 0.6073052883148193, "learning_rate": 1.357534685357586e-05, "loss": 0.0208, "step": 77180 }, { "grad_norm": 0.20681974291801453, "learning_rate": 1.3564021697914715e-05, "loss": 0.012, "step": 77190 }, { "grad_norm": 0.8831804394721985, "learning_rate": 1.3552700526837381e-05, "loss": 0.0367, "step": 77200 }, { "grad_norm": 0.4258406460285187, "learning_rate": 1.3541383341581925e-05, "loss": 0.0218, "step": 77210 }, { "grad_norm": 0.38046690821647644, "learning_rate": 1.3530070143385965e-05, "loss": 0.0241, "step": 77220 }, { "grad_norm": 0.2840343415737152, "learning_rate": 1.3518760933486702e-05, "loss": 0.0152, "step": 77230 }, { "grad_norm": 0.48012077808380127, "learning_rate": 1.3507455713120908e-05, "loss": 0.0143, "step": 77240 }, { "grad_norm": 0.2414473593235016, "learning_rate": 1.3496154483524876e-05, "loss": 0.0237, "step": 77250 }, { "grad_norm": 0.5393559336662292, "learning_rate": 1.3484857245934506e-05, "loss": 0.0205, "step": 77260 }, { "grad_norm": 0.3839149475097656, "learning_rate": 1.3473564001585248e-05, "loss": 0.0192, "step": 77270 }, { "grad_norm": 0.328204482793808, "learning_rate": 1.3462274751712129e-05, "loss": 0.0159, "step": 77280 }, { "grad_norm": 0.39812716841697693, "learning_rate": 1.3450989497549698e-05, "loss": 0.0226, "step": 77290 }, { "grad_norm": 0.8038043975830078, "learning_rate": 1.3439708240332115e-05, "loss": 0.0273, "step": 77300 }, { "grad_norm": 0.23272262513637543, "learning_rate": 1.3428430981293043e-05, "loss": 0.0173, "step": 77310 }, { "grad_norm": 0.15053749084472656, "learning_rate": 1.3417157721665808e-05, "loss": 0.0159, "step": 77320 }, { "grad_norm": 0.29152122139930725, "learning_rate": 1.3405888462683186e-05, "loss": 0.0298, "step": 77330 }, { "grad_norm": 0.36473363637924194, "learning_rate": 1.3394623205577595e-05, "loss": 0.0154, "step": 77340 }, { "grad_norm": 0.3155360221862793, "learning_rate": 1.3383361951580963e-05, "loss": 0.0202, "step": 77350 }, { "grad_norm": 0.5014823079109192, "learning_rate": 1.337210470192483e-05, "loss": 0.0255, "step": 77360 }, { "grad_norm": 0.5118979811668396, "learning_rate": 1.3360851457840245e-05, "loss": 0.0243, "step": 77370 }, { "grad_norm": 0.23881855607032776, "learning_rate": 1.3349602220557856e-05, "loss": 0.0168, "step": 77380 }, { "grad_norm": 0.19958360493183136, "learning_rate": 1.3338356991307865e-05, "loss": 0.0249, "step": 77390 }, { "grad_norm": 0.6079471111297607, "learning_rate": 1.3327115771320048e-05, "loss": 0.0236, "step": 77400 }, { "grad_norm": 0.33620885014533997, "learning_rate": 1.3315878561823697e-05, "loss": 0.0302, "step": 77410 }, { "grad_norm": 0.5709415674209595, "learning_rate": 1.330464536404773e-05, "loss": 0.0185, "step": 77420 }, { "grad_norm": 0.4128738045692444, "learning_rate": 1.3293416179220563e-05, "loss": 0.0216, "step": 77430 }, { "grad_norm": 0.46278953552246094, "learning_rate": 1.328219100857021e-05, "loss": 0.0175, "step": 77440 }, { "grad_norm": 0.23741157352924347, "learning_rate": 1.327096985332425e-05, "loss": 0.0171, "step": 77450 }, { "grad_norm": 0.41719773411750793, "learning_rate": 1.3259752714709816e-05, "loss": 0.0234, "step": 77460 }, { "grad_norm": 0.3513175845146179, "learning_rate": 1.3248539593953569e-05, "loss": 0.0198, "step": 77470 }, { "grad_norm": 0.42635270953178406, "learning_rate": 1.323733049228178e-05, "loss": 0.0154, "step": 77480 }, { "grad_norm": 0.5503911375999451, "learning_rate": 1.3226125410920276e-05, "loss": 0.0231, "step": 77490 }, { "grad_norm": 0.28467002511024475, "learning_rate": 1.3214924351094388e-05, "loss": 0.0224, "step": 77500 }, { "grad_norm": 0.3879564106464386, "learning_rate": 1.3203727314029069e-05, "loss": 0.0192, "step": 77510 }, { "grad_norm": 0.3194009065628052, "learning_rate": 1.3192534300948811e-05, "loss": 0.0249, "step": 77520 }, { "grad_norm": 0.2398267686367035, "learning_rate": 1.3181345313077675e-05, "loss": 0.0158, "step": 77530 }, { "grad_norm": 0.47126269340515137, "learning_rate": 1.3170160351639243e-05, "loss": 0.0221, "step": 77540 }, { "grad_norm": 0.5667023062705994, "learning_rate": 1.3158979417856721e-05, "loss": 0.0186, "step": 77550 }, { "grad_norm": 0.972172737121582, "learning_rate": 1.31478025129528e-05, "loss": 0.025, "step": 77560 }, { "grad_norm": 0.4230310618877411, "learning_rate": 1.3136629638149794e-05, "loss": 0.0182, "step": 77570 }, { "grad_norm": 0.6323590874671936, "learning_rate": 1.3125460794669547e-05, "loss": 0.02, "step": 77580 }, { "grad_norm": 0.1721193790435791, "learning_rate": 1.3114295983733483e-05, "loss": 0.0214, "step": 77590 }, { "grad_norm": 0.26312562823295593, "learning_rate": 1.3103135206562534e-05, "loss": 0.014, "step": 77600 }, { "grad_norm": 0.7930710315704346, "learning_rate": 1.3091978464377264e-05, "loss": 0.0293, "step": 77610 }, { "grad_norm": 0.40494945645332336, "learning_rate": 1.3080825758397724e-05, "loss": 0.0274, "step": 77620 }, { "grad_norm": 0.7599021792411804, "learning_rate": 1.3069677089843568e-05, "loss": 0.0206, "step": 77630 }, { "grad_norm": 0.4759265184402466, "learning_rate": 1.305853245993401e-05, "loss": 0.0224, "step": 77640 }, { "grad_norm": 0.4209524095058441, "learning_rate": 1.3047391869887809e-05, "loss": 0.0194, "step": 77650 }, { "grad_norm": 0.35877227783203125, "learning_rate": 1.3036255320923263e-05, "loss": 0.0287, "step": 77660 }, { "grad_norm": 0.6735733151435852, "learning_rate": 1.3025122814258273e-05, "loss": 0.017, "step": 77670 }, { "grad_norm": 0.5159377455711365, "learning_rate": 1.3013994351110249e-05, "loss": 0.0211, "step": 77680 }, { "grad_norm": 0.32851889729499817, "learning_rate": 1.3002869932696194e-05, "loss": 0.0319, "step": 77690 }, { "grad_norm": 0.33890119194984436, "learning_rate": 1.2991749560232664e-05, "loss": 0.0148, "step": 77700 }, { "grad_norm": 0.3478689193725586, "learning_rate": 1.2980633234935768e-05, "loss": 0.0167, "step": 77710 }, { "grad_norm": 0.2500384449958801, "learning_rate": 1.2969520958021153e-05, "loss": 0.0163, "step": 77720 }, { "grad_norm": 0.24704934656620026, "learning_rate": 1.295841273070405e-05, "loss": 0.017, "step": 77730 }, { "grad_norm": 0.32684579491615295, "learning_rate": 1.2947308554199255e-05, "loss": 0.0167, "step": 77740 }, { "grad_norm": 0.36636480689048767, "learning_rate": 1.2936208429721075e-05, "loss": 0.0273, "step": 77750 }, { "grad_norm": 0.3771473467350006, "learning_rate": 1.2925112358483416e-05, "loss": 0.019, "step": 77760 }, { "grad_norm": 0.19341926276683807, "learning_rate": 1.2914020341699728e-05, "loss": 0.0181, "step": 77770 }, { "grad_norm": 0.34001031517982483, "learning_rate": 1.2902932380583032e-05, "loss": 0.0232, "step": 77780 }, { "grad_norm": 0.271290123462677, "learning_rate": 1.2891848476345863e-05, "loss": 0.0181, "step": 77790 }, { "grad_norm": 0.44449299573898315, "learning_rate": 1.2880768630200368e-05, "loss": 0.0222, "step": 77800 }, { "grad_norm": 0.2465953528881073, "learning_rate": 1.286969284335819e-05, "loss": 0.0173, "step": 77810 }, { "grad_norm": 0.743943989276886, "learning_rate": 1.2858621117030578e-05, "loss": 0.0291, "step": 77820 }, { "grad_norm": 0.24276910722255707, "learning_rate": 1.2847553452428324e-05, "loss": 0.0261, "step": 77830 }, { "grad_norm": 0.31187283992767334, "learning_rate": 1.283648985076178e-05, "loss": 0.0222, "step": 77840 }, { "grad_norm": 0.4628283381462097, "learning_rate": 1.2825430313240817e-05, "loss": 0.0219, "step": 77850 }, { "grad_norm": 0.38252556324005127, "learning_rate": 1.281437484107491e-05, "loss": 0.0209, "step": 77860 }, { "grad_norm": 0.39042556285858154, "learning_rate": 1.2803323435473058e-05, "loss": 0.0302, "step": 77870 }, { "grad_norm": 0.4374462366104126, "learning_rate": 1.2792276097643841e-05, "loss": 0.0179, "step": 77880 }, { "grad_norm": 0.2120361626148224, "learning_rate": 1.2781232828795342e-05, "loss": 0.0193, "step": 77890 }, { "grad_norm": 0.2339409738779068, "learning_rate": 1.2770193630135292e-05, "loss": 0.0177, "step": 77900 }, { "grad_norm": 0.4671952724456787, "learning_rate": 1.2759158502870877e-05, "loss": 0.0229, "step": 77910 }, { "grad_norm": 0.2360171377658844, "learning_rate": 1.2748127448208913e-05, "loss": 0.0183, "step": 77920 }, { "grad_norm": 0.27385637164115906, "learning_rate": 1.2737100467355706e-05, "loss": 0.0157, "step": 77930 }, { "grad_norm": 0.5584076046943665, "learning_rate": 1.2726077561517169e-05, "loss": 0.0232, "step": 77940 }, { "grad_norm": 0.17930899560451508, "learning_rate": 1.2715058731898755e-05, "loss": 0.0217, "step": 77950 }, { "grad_norm": 0.26173385977745056, "learning_rate": 1.2704043979705471e-05, "loss": 0.0248, "step": 77960 }, { "grad_norm": 0.6021007895469666, "learning_rate": 1.269303330614185e-05, "loss": 0.0167, "step": 77970 }, { "grad_norm": 0.19659841060638428, "learning_rate": 1.2682026712412015e-05, "loss": 0.0239, "step": 77980 }, { "grad_norm": 0.3059055507183075, "learning_rate": 1.2671024199719644e-05, "loss": 0.0193, "step": 77990 }, { "grad_norm": 0.24411463737487793, "learning_rate": 1.2660025769267936e-05, "loss": 0.0168, "step": 78000 }, { "grad_norm": 0.25252681970596313, "learning_rate": 1.2649031422259667e-05, "loss": 0.018, "step": 78010 }, { "grad_norm": 0.4374499022960663, "learning_rate": 1.2638041159897163e-05, "loss": 0.0184, "step": 78020 }, { "grad_norm": 0.21184630692005157, "learning_rate": 1.262705498338232e-05, "loss": 0.0202, "step": 78030 }, { "grad_norm": 0.5778383612632751, "learning_rate": 1.2616072893916542e-05, "loss": 0.0247, "step": 78040 }, { "grad_norm": 0.2807106673717499, "learning_rate": 1.2605094892700836e-05, "loss": 0.0217, "step": 78050 }, { "grad_norm": 0.15628887712955475, "learning_rate": 1.259412098093572e-05, "loss": 0.0159, "step": 78060 }, { "grad_norm": 0.24170778691768646, "learning_rate": 1.258315115982131e-05, "loss": 0.016, "step": 78070 }, { "grad_norm": 0.6076437830924988, "learning_rate": 1.2572185430557203e-05, "loss": 0.0204, "step": 78080 }, { "grad_norm": 0.4331291615962982, "learning_rate": 1.2561223794342659e-05, "loss": 0.0153, "step": 78090 }, { "grad_norm": 0.3037911653518677, "learning_rate": 1.255026625237637e-05, "loss": 0.0292, "step": 78100 }, { "grad_norm": 0.3919336795806885, "learning_rate": 1.2539312805856678e-05, "loss": 0.0163, "step": 78110 }, { "grad_norm": 0.4542250335216522, "learning_rate": 1.25283634559814e-05, "loss": 0.0186, "step": 78120 }, { "grad_norm": 0.22356802225112915, "learning_rate": 1.2517418203947972e-05, "loss": 0.0265, "step": 78130 }, { "grad_norm": 0.7667443752288818, "learning_rate": 1.2506477050953308e-05, "loss": 0.0212, "step": 78140 }, { "grad_norm": 0.26439714431762695, "learning_rate": 1.2495539998193972e-05, "loss": 0.0214, "step": 78150 }, { "grad_norm": 0.20085030794143677, "learning_rate": 1.2484607046865975e-05, "loss": 0.0164, "step": 78160 }, { "grad_norm": 0.4415642321109772, "learning_rate": 1.2473678198164968e-05, "loss": 0.0212, "step": 78170 }, { "grad_norm": 0.4393148422241211, "learning_rate": 1.2462753453286075e-05, "loss": 0.0219, "step": 78180 }, { "grad_norm": 0.3537779450416565, "learning_rate": 1.2451832813424031e-05, "loss": 0.0241, "step": 78190 }, { "grad_norm": 0.21512624621391296, "learning_rate": 1.2440916279773095e-05, "loss": 0.0167, "step": 78200 }, { "grad_norm": 0.40235668420791626, "learning_rate": 1.2430003853527101e-05, "loss": 0.0158, "step": 78210 }, { "grad_norm": 0.5350087285041809, "learning_rate": 1.2419095535879383e-05, "loss": 0.025, "step": 78220 }, { "grad_norm": 0.20707647502422333, "learning_rate": 1.2408191328022878e-05, "loss": 0.0153, "step": 78230 }, { "grad_norm": 0.31534555554389954, "learning_rate": 1.2397291231150066e-05, "loss": 0.0144, "step": 78240 }, { "grad_norm": 0.5528734922409058, "learning_rate": 1.2386395246452937e-05, "loss": 0.0202, "step": 78250 }, { "grad_norm": 0.3820365369319916, "learning_rate": 1.2375503375123083e-05, "loss": 0.0189, "step": 78260 }, { "grad_norm": 0.8245673775672913, "learning_rate": 1.2364615618351583e-05, "loss": 0.0204, "step": 78270 }, { "grad_norm": 0.3525571823120117, "learning_rate": 1.2353731977329169e-05, "loss": 0.0178, "step": 78280 }, { "grad_norm": 0.35668647289276123, "learning_rate": 1.2342852453246007e-05, "loss": 0.0154, "step": 78290 }, { "grad_norm": 0.344804972410202, "learning_rate": 1.2331977047291898e-05, "loss": 0.0445, "step": 78300 }, { "grad_norm": 0.27001628279685974, "learning_rate": 1.2321105760656137e-05, "loss": 0.0169, "step": 78310 }, { "grad_norm": 0.40054020285606384, "learning_rate": 1.2310238594527613e-05, "loss": 0.0179, "step": 78320 }, { "grad_norm": 0.3646017909049988, "learning_rate": 1.2299375550094693e-05, "loss": 0.0198, "step": 78330 }, { "grad_norm": 0.30739957094192505, "learning_rate": 1.2288516628545421e-05, "loss": 0.0187, "step": 78340 }, { "grad_norm": 0.49028274416923523, "learning_rate": 1.2277661831067255e-05, "loss": 0.0261, "step": 78350 }, { "grad_norm": 0.2162180095911026, "learning_rate": 1.2266811158847286e-05, "loss": 0.0128, "step": 78360 }, { "grad_norm": 0.23208047449588776, "learning_rate": 1.2255964613072107e-05, "loss": 0.0194, "step": 78370 }, { "grad_norm": 0.6828669309616089, "learning_rate": 1.2245122194927905e-05, "loss": 0.0226, "step": 78380 }, { "grad_norm": 0.3871423304080963, "learning_rate": 1.2234283905600364e-05, "loss": 0.018, "step": 78390 }, { "grad_norm": 0.3711598515510559, "learning_rate": 1.222344974627475e-05, "loss": 0.0262, "step": 78400 }, { "grad_norm": 0.3446462154388428, "learning_rate": 1.2212619718135875e-05, "loss": 0.023, "step": 78410 }, { "grad_norm": 0.33538559079170227, "learning_rate": 1.2201793822368118e-05, "loss": 0.0148, "step": 78420 }, { "grad_norm": 0.27017447352409363, "learning_rate": 1.2190972060155337e-05, "loss": 0.0206, "step": 78430 }, { "grad_norm": 0.267677366733551, "learning_rate": 1.2180154432681013e-05, "loss": 0.015, "step": 78440 }, { "grad_norm": 0.332229346036911, "learning_rate": 1.2169340941128148e-05, "loss": 0.014, "step": 78450 }, { "grad_norm": 0.28279197216033936, "learning_rate": 1.2158531586679267e-05, "loss": 0.029, "step": 78460 }, { "grad_norm": 0.33906838297843933, "learning_rate": 1.2147726370516477e-05, "loss": 0.0319, "step": 78470 }, { "grad_norm": 0.29779788851737976, "learning_rate": 1.2136925293821422e-05, "loss": 0.0206, "step": 78480 }, { "grad_norm": 0.4047781527042389, "learning_rate": 1.2126128357775306e-05, "loss": 0.0178, "step": 78490 }, { "grad_norm": 0.3743848204612732, "learning_rate": 1.2115335563558838e-05, "loss": 0.0191, "step": 78500 }, { "grad_norm": 0.24668051302433014, "learning_rate": 1.2104546912352327e-05, "loss": 0.0209, "step": 78510 }, { "grad_norm": 0.20962098240852356, "learning_rate": 1.2093762405335557e-05, "loss": 0.014, "step": 78520 }, { "grad_norm": 0.6302686333656311, "learning_rate": 1.2082982043687974e-05, "loss": 0.0237, "step": 78530 }, { "grad_norm": 0.1944671869277954, "learning_rate": 1.207220582858845e-05, "loss": 0.0197, "step": 78540 }, { "grad_norm": 0.41068175435066223, "learning_rate": 1.206143376121549e-05, "loss": 0.0174, "step": 78550 }, { "grad_norm": 0.4339084029197693, "learning_rate": 1.2050665842747078e-05, "loss": 0.0205, "step": 78560 }, { "grad_norm": 0.4585222601890564, "learning_rate": 1.2039902074360804e-05, "loss": 0.049, "step": 78570 }, { "grad_norm": 0.39377808570861816, "learning_rate": 1.2029142457233755e-05, "loss": 0.0319, "step": 78580 }, { "grad_norm": 0.23543862998485565, "learning_rate": 1.2018386992542601e-05, "loss": 0.0149, "step": 78590 }, { "grad_norm": 0.40781134366989136, "learning_rate": 1.2007635681463542e-05, "loss": 0.0172, "step": 78600 }, { "grad_norm": 0.2785528302192688, "learning_rate": 1.1996888525172334e-05, "loss": 0.0199, "step": 78610 }, { "grad_norm": 0.4395654797554016, "learning_rate": 1.198614552484425e-05, "loss": 0.0188, "step": 78620 }, { "grad_norm": 0.4600140452384949, "learning_rate": 1.197540668165415e-05, "loss": 0.0237, "step": 78630 }, { "grad_norm": 0.15034575760364532, "learning_rate": 1.1964671996776395e-05, "loss": 0.0178, "step": 78640 }, { "grad_norm": 0.1808723658323288, "learning_rate": 1.1953941471384922e-05, "loss": 0.0193, "step": 78650 }, { "grad_norm": 0.28184181451797485, "learning_rate": 1.1943215106653205e-05, "loss": 0.0188, "step": 78660 }, { "grad_norm": 0.45271310210227966, "learning_rate": 1.193249290375429e-05, "loss": 0.02, "step": 78670 }, { "grad_norm": 0.5642059445381165, "learning_rate": 1.1921774863860696e-05, "loss": 0.0219, "step": 78680 }, { "grad_norm": 0.20008495450019836, "learning_rate": 1.1911060988144556e-05, "loss": 0.0237, "step": 78690 }, { "grad_norm": 0.3110145330429077, "learning_rate": 1.1900351277777533e-05, "loss": 0.0284, "step": 78700 }, { "grad_norm": 0.48113977909088135, "learning_rate": 1.1889645733930798e-05, "loss": 0.022, "step": 78710 }, { "grad_norm": 0.3867238759994507, "learning_rate": 1.1878944357775107e-05, "loss": 0.0206, "step": 78720 }, { "grad_norm": 0.2887251079082489, "learning_rate": 1.1868247150480749e-05, "loss": 0.0258, "step": 78730 }, { "grad_norm": 0.28423866629600525, "learning_rate": 1.1857554113217567e-05, "loss": 0.0257, "step": 78740 }, { "grad_norm": 0.44016093015670776, "learning_rate": 1.1846865247154903e-05, "loss": 0.0212, "step": 78750 }, { "grad_norm": 0.3543722331523895, "learning_rate": 1.1836180553461711e-05, "loss": 0.019, "step": 78760 }, { "grad_norm": 0.6665499210357666, "learning_rate": 1.1825500033306419e-05, "loss": 0.0279, "step": 78770 }, { "grad_norm": 0.18407902121543884, "learning_rate": 1.1814823687857052e-05, "loss": 0.0197, "step": 78780 }, { "grad_norm": 0.4385817348957062, "learning_rate": 1.1804151518281153e-05, "loss": 0.0315, "step": 78790 }, { "grad_norm": 0.3694000244140625, "learning_rate": 1.179348352574583e-05, "loss": 0.0295, "step": 78800 }, { "grad_norm": 0.38739508390426636, "learning_rate": 1.1782819711417697e-05, "loss": 0.0204, "step": 78810 }, { "grad_norm": 0.7058584094047546, "learning_rate": 1.1772160076462951e-05, "loss": 0.0389, "step": 78820 }, { "grad_norm": 0.3958493769168854, "learning_rate": 1.1761504622047293e-05, "loss": 0.0189, "step": 78830 }, { "grad_norm": 0.21258248388767242, "learning_rate": 1.1750853349335999e-05, "loss": 0.0155, "step": 78840 }, { "grad_norm": 0.25963741540908813, "learning_rate": 1.1740206259493874e-05, "loss": 0.0222, "step": 78850 }, { "grad_norm": 0.3418426215648651, "learning_rate": 1.1729563353685286e-05, "loss": 0.0201, "step": 78860 }, { "grad_norm": 0.38657933473587036, "learning_rate": 1.1718924633074091e-05, "loss": 0.0217, "step": 78870 }, { "grad_norm": 0.3165491223335266, "learning_rate": 1.1708290098823765e-05, "loss": 0.0252, "step": 78880 }, { "grad_norm": 1.1410939693450928, "learning_rate": 1.1697659752097245e-05, "loss": 0.0171, "step": 78890 }, { "grad_norm": 0.3690069913864136, "learning_rate": 1.1687033594057068e-05, "loss": 0.0221, "step": 78900 }, { "grad_norm": 0.2116233855485916, "learning_rate": 1.1676411625865297e-05, "loss": 0.0252, "step": 78910 }, { "grad_norm": 0.32994821667671204, "learning_rate": 1.1665793848683542e-05, "loss": 0.0213, "step": 78920 }, { "grad_norm": 0.40082797408103943, "learning_rate": 1.1655180263672927e-05, "loss": 0.0134, "step": 78930 }, { "grad_norm": 0.6729310154914856, "learning_rate": 1.1644570871994142e-05, "loss": 0.014, "step": 78940 }, { "grad_norm": 0.1656494140625, "learning_rate": 1.163396567480744e-05, "loss": 0.0215, "step": 78950 }, { "grad_norm": 0.32373955845832825, "learning_rate": 1.1623364673272552e-05, "loss": 0.0369, "step": 78960 }, { "grad_norm": 0.5850350856781006, "learning_rate": 1.1612767868548807e-05, "loss": 0.0182, "step": 78970 }, { "grad_norm": 0.4229428470134735, "learning_rate": 1.1602175261795057e-05, "loss": 0.0436, "step": 78980 }, { "grad_norm": 0.45730602741241455, "learning_rate": 1.1591586854169696e-05, "loss": 0.0209, "step": 78990 }, { "grad_norm": 0.2726195454597473, "learning_rate": 1.158100264683064e-05, "loss": 0.0196, "step": 79000 }, { "grad_norm": 0.2545676529407501, "learning_rate": 1.1570422640935386e-05, "loss": 0.0178, "step": 79010 }, { "grad_norm": 0.3149319887161255, "learning_rate": 1.155984683764092e-05, "loss": 0.022, "step": 79020 }, { "grad_norm": 0.3256545662879944, "learning_rate": 1.1549275238103812e-05, "loss": 0.0192, "step": 79030 }, { "grad_norm": 0.4206446707248688, "learning_rate": 1.1538707843480152e-05, "loss": 0.0201, "step": 79040 }, { "grad_norm": 0.5806085467338562, "learning_rate": 1.1528144654925593e-05, "loss": 0.0276, "step": 79050 }, { "grad_norm": 0.49031707644462585, "learning_rate": 1.1517585673595277e-05, "loss": 0.0189, "step": 79060 }, { "grad_norm": 0.2972413897514343, "learning_rate": 1.150703090064395e-05, "loss": 0.0192, "step": 79070 }, { "grad_norm": 0.3130649924278259, "learning_rate": 1.1496480337225835e-05, "loss": 0.0221, "step": 79080 }, { "grad_norm": 0.5165964961051941, "learning_rate": 1.1485933984494751e-05, "loss": 0.025, "step": 79090 }, { "grad_norm": 0.3498785197734833, "learning_rate": 1.1475391843604e-05, "loss": 0.031, "step": 79100 }, { "grad_norm": 0.2901175022125244, "learning_rate": 1.1464853915706503e-05, "loss": 0.0197, "step": 79110 }, { "grad_norm": 0.3261171281337738, "learning_rate": 1.1454320201954626e-05, "loss": 0.024, "step": 79120 }, { "grad_norm": 0.7593121528625488, "learning_rate": 1.1443790703500356e-05, "loss": 0.0226, "step": 79130 }, { "grad_norm": 0.5205875039100647, "learning_rate": 1.1433265421495154e-05, "loss": 0.0276, "step": 79140 }, { "grad_norm": 0.376662015914917, "learning_rate": 1.1422744357090065e-05, "loss": 0.016, "step": 79150 }, { "grad_norm": 0.49156132340431213, "learning_rate": 1.1412227511435647e-05, "loss": 0.0264, "step": 79160 }, { "grad_norm": 0.17723794281482697, "learning_rate": 1.1401714885682025e-05, "loss": 0.0246, "step": 79170 }, { "grad_norm": 0.5150497555732727, "learning_rate": 1.1391206480978823e-05, "loss": 0.0247, "step": 79180 }, { "grad_norm": 0.5134043097496033, "learning_rate": 1.1380702298475227e-05, "loss": 0.0163, "step": 79190 }, { "grad_norm": 0.23085063695907593, "learning_rate": 1.1370202339319985e-05, "loss": 0.0165, "step": 79200 }, { "grad_norm": 0.6332409381866455, "learning_rate": 1.1359706604661319e-05, "loss": 0.0186, "step": 79210 }, { "grad_norm": 0.3505910336971283, "learning_rate": 1.1349215095647042e-05, "loss": 0.0169, "step": 79220 }, { "grad_norm": 0.3335820734500885, "learning_rate": 1.1338727813424493e-05, "loss": 0.0151, "step": 79230 }, { "grad_norm": 0.36832523345947266, "learning_rate": 1.132824475914056e-05, "loss": 0.0174, "step": 79240 }, { "grad_norm": 0.3467845618724823, "learning_rate": 1.131776593394162e-05, "loss": 0.0267, "step": 79250 }, { "grad_norm": 0.25328677892684937, "learning_rate": 1.1307291338973657e-05, "loss": 0.0207, "step": 79260 }, { "grad_norm": 0.40004026889801025, "learning_rate": 1.1296820975382121e-05, "loss": 0.0221, "step": 79270 }, { "grad_norm": 0.1397101730108261, "learning_rate": 1.1286354844312074e-05, "loss": 0.031, "step": 79280 }, { "grad_norm": 0.8442395329475403, "learning_rate": 1.1275892946908023e-05, "loss": 0.0527, "step": 79290 }, { "grad_norm": 0.4135720431804657, "learning_rate": 1.126543528431413e-05, "loss": 0.0297, "step": 79300 }, { "grad_norm": 0.16150234639644623, "learning_rate": 1.125498185767398e-05, "loss": 0.0174, "step": 79310 }, { "grad_norm": 0.27316364645957947, "learning_rate": 1.1244532668130781e-05, "loss": 0.0179, "step": 79320 }, { "grad_norm": 0.37466299533843994, "learning_rate": 1.1234087716827203e-05, "loss": 0.0298, "step": 79330 }, { "grad_norm": 0.25439655780792236, "learning_rate": 1.1223647004905524e-05, "loss": 0.0154, "step": 79340 }, { "grad_norm": 0.3470919132232666, "learning_rate": 1.1213210533507485e-05, "loss": 0.0188, "step": 79350 }, { "grad_norm": 0.7717300653457642, "learning_rate": 1.1202778303774447e-05, "loss": 0.0229, "step": 79360 }, { "grad_norm": 0.16120824217796326, "learning_rate": 1.1192350316847234e-05, "loss": 0.0183, "step": 79370 }, { "grad_norm": 0.16666513681411743, "learning_rate": 1.1181926573866258e-05, "loss": 0.0228, "step": 79380 }, { "grad_norm": 0.16224436461925507, "learning_rate": 1.1171507075971416e-05, "loss": 0.0149, "step": 79390 }, { "grad_norm": 0.18527205288410187, "learning_rate": 1.1161091824302177e-05, "loss": 0.0174, "step": 79400 }, { "grad_norm": 0.3390168249607086, "learning_rate": 1.115068081999755e-05, "loss": 0.0181, "step": 79410 }, { "grad_norm": 0.22799809277057648, "learning_rate": 1.1140274064196071e-05, "loss": 0.0124, "step": 79420 }, { "grad_norm": 0.31288230419158936, "learning_rate": 1.112987155803578e-05, "loss": 0.0146, "step": 79430 }, { "grad_norm": 0.27088817954063416, "learning_rate": 1.1119473302654298e-05, "loss": 0.0165, "step": 79440 }, { "grad_norm": 0.5948955416679382, "learning_rate": 1.1109079299188769e-05, "loss": 0.03, "step": 79450 }, { "grad_norm": 0.41884446144104004, "learning_rate": 1.1098689548775847e-05, "loss": 0.0226, "step": 79460 }, { "grad_norm": 0.37912023067474365, "learning_rate": 1.1088304052551762e-05, "loss": 0.0183, "step": 79470 }, { "grad_norm": 0.1025971919298172, "learning_rate": 1.1077922811652215e-05, "loss": 0.0152, "step": 79480 }, { "grad_norm": 0.34184181690216064, "learning_rate": 1.1067545827212534e-05, "loss": 0.0219, "step": 79490 }, { "grad_norm": 0.48448747396469116, "learning_rate": 1.1057173100367496e-05, "loss": 0.0205, "step": 79500 }, { "grad_norm": 0.277834951877594, "learning_rate": 1.1046804632251474e-05, "loss": 0.0247, "step": 79510 }, { "grad_norm": 0.1769312024116516, "learning_rate": 1.1036440423998318e-05, "loss": 0.0156, "step": 79520 }, { "grad_norm": 0.5034497380256653, "learning_rate": 1.1026080476741469e-05, "loss": 0.0232, "step": 79530 }, { "grad_norm": 0.2818979322910309, "learning_rate": 1.1015724791613836e-05, "loss": 0.0208, "step": 79540 }, { "grad_norm": 0.9407973885536194, "learning_rate": 1.1005373369747951e-05, "loss": 0.0218, "step": 79550 }, { "grad_norm": 0.2969914972782135, "learning_rate": 1.09950262122758e-05, "loss": 0.0197, "step": 79560 }, { "grad_norm": 0.48372840881347656, "learning_rate": 1.0984683320328948e-05, "loss": 0.0207, "step": 79570 }, { "grad_norm": 0.2385089248418808, "learning_rate": 1.0974344695038458e-05, "loss": 0.0137, "step": 79580 }, { "grad_norm": 0.32637715339660645, "learning_rate": 1.0964010337534968e-05, "loss": 0.0218, "step": 79590 }, { "grad_norm": 0.2682972252368927, "learning_rate": 1.0953680248948611e-05, "loss": 0.0307, "step": 79600 }, { "grad_norm": 0.2397710382938385, "learning_rate": 1.0943354430409069e-05, "loss": 0.0212, "step": 79610 }, { "grad_norm": 0.3704683482646942, "learning_rate": 1.0933032883045574e-05, "loss": 0.0188, "step": 79620 }, { "grad_norm": 0.4173579812049866, "learning_rate": 1.0922715607986872e-05, "loss": 0.0188, "step": 79630 }, { "grad_norm": 0.4851056635379791, "learning_rate": 1.0912402606361234e-05, "loss": 0.0289, "step": 79640 }, { "grad_norm": 0.16596458852291107, "learning_rate": 1.090209387929647e-05, "loss": 0.022, "step": 79650 }, { "grad_norm": 0.2603469491004944, "learning_rate": 1.0891789427919957e-05, "loss": 0.0159, "step": 79660 }, { "grad_norm": 0.9590813517570496, "learning_rate": 1.088148925335854e-05, "loss": 0.0228, "step": 79670 }, { "grad_norm": 0.26502445340156555, "learning_rate": 1.087119335673864e-05, "loss": 0.0213, "step": 79680 }, { "grad_norm": 0.6817592978477478, "learning_rate": 1.0860901739186208e-05, "loss": 0.0215, "step": 79690 }, { "grad_norm": 0.4847218692302704, "learning_rate": 1.085061440182673e-05, "loss": 0.0276, "step": 79700 }, { "grad_norm": 0.2732233703136444, "learning_rate": 1.0840331345785193e-05, "loss": 0.0147, "step": 79710 }, { "grad_norm": 1.4405275583267212, "learning_rate": 1.083005257218615e-05, "loss": 0.0216, "step": 79720 }, { "grad_norm": 0.284073144197464, "learning_rate": 1.0819778082153642e-05, "loss": 0.0141, "step": 79730 }, { "grad_norm": 0.3297590911388397, "learning_rate": 1.0809507876811325e-05, "loss": 0.0209, "step": 79740 }, { "grad_norm": 0.20658843219280243, "learning_rate": 1.0799241957282292e-05, "loss": 0.021, "step": 79750 }, { "grad_norm": 0.22734159231185913, "learning_rate": 1.078898032468923e-05, "loss": 0.0252, "step": 79760 }, { "grad_norm": 0.2920317053794861, "learning_rate": 1.0778722980154315e-05, "loss": 0.0227, "step": 79770 }, { "grad_norm": 0.31361040472984314, "learning_rate": 1.0768469924799302e-05, "loss": 0.0159, "step": 79780 }, { "grad_norm": 0.3011448383331299, "learning_rate": 1.0758221159745419e-05, "loss": 0.0169, "step": 79790 }, { "grad_norm": 0.49768757820129395, "learning_rate": 1.074797668611347e-05, "loss": 0.0139, "step": 79800 }, { "grad_norm": 0.378734827041626, "learning_rate": 1.0737736505023777e-05, "loss": 0.0221, "step": 79810 }, { "grad_norm": 0.21321897208690643, "learning_rate": 1.0727500617596203e-05, "loss": 0.0228, "step": 79820 }, { "grad_norm": 0.3952007591724396, "learning_rate": 1.0717269024950104e-05, "loss": 0.0348, "step": 79830 }, { "grad_norm": 0.38685140013694763, "learning_rate": 1.070704172820442e-05, "loss": 0.0169, "step": 79840 }, { "grad_norm": 0.9465905427932739, "learning_rate": 1.0696818728477559e-05, "loss": 0.0323, "step": 79850 }, { "grad_norm": 0.23305323719978333, "learning_rate": 1.0686600026887511e-05, "loss": 0.0155, "step": 79860 }, { "grad_norm": 0.4418228566646576, "learning_rate": 1.0676385624551777e-05, "loss": 0.0186, "step": 79870 }, { "grad_norm": 0.478742778301239, "learning_rate": 1.0666175522587402e-05, "loss": 0.0198, "step": 79880 }, { "grad_norm": 0.26843640208244324, "learning_rate": 1.0655969722110925e-05, "loss": 0.017, "step": 79890 }, { "grad_norm": 0.4049604833126068, "learning_rate": 1.0645768224238439e-05, "loss": 0.0197, "step": 79900 }, { "grad_norm": 0.4226842224597931, "learning_rate": 1.063557103008559e-05, "loss": 0.0248, "step": 79910 }, { "grad_norm": 0.26169466972351074, "learning_rate": 1.0625378140767494e-05, "loss": 0.0183, "step": 79920 }, { "grad_norm": 0.7277010083198547, "learning_rate": 1.0615189557398846e-05, "loss": 0.0293, "step": 79930 }, { "grad_norm": 0.2318899780511856, "learning_rate": 1.060500528109386e-05, "loss": 0.0182, "step": 79940 }, { "grad_norm": 0.26536238193511963, "learning_rate": 1.0594825312966272e-05, "loss": 0.0224, "step": 79950 }, { "grad_norm": 0.8159878849983215, "learning_rate": 1.0584649654129336e-05, "loss": 0.0266, "step": 79960 }, { "grad_norm": 0.28510499000549316, "learning_rate": 1.0574478305695863e-05, "loss": 0.0235, "step": 79970 }, { "grad_norm": 0.5491560697555542, "learning_rate": 1.0564311268778154e-05, "loss": 0.0224, "step": 79980 }, { "grad_norm": 0.24158860743045807, "learning_rate": 1.0554148544488073e-05, "loss": 0.0257, "step": 79990 }, { "grad_norm": 0.41400447487831116, "learning_rate": 1.0543990133937004e-05, "loss": 0.0215, "step": 80000 }, { "grad_norm": 0.3780239224433899, "learning_rate": 1.0533836038235868e-05, "loss": 0.0214, "step": 80010 }, { "grad_norm": 0.453387051820755, "learning_rate": 1.052368625849507e-05, "loss": 0.0222, "step": 80020 }, { "grad_norm": 0.6042959094047546, "learning_rate": 1.05135407958246e-05, "loss": 0.0191, "step": 80030 }, { "grad_norm": 0.21049462258815765, "learning_rate": 1.0503399651333934e-05, "loss": 0.0157, "step": 80040 }, { "grad_norm": 0.36565518379211426, "learning_rate": 1.04932628261321e-05, "loss": 0.019, "step": 80050 }, { "grad_norm": 0.42532873153686523, "learning_rate": 1.0483130321327644e-05, "loss": 0.0307, "step": 80060 }, { "grad_norm": 0.24407429993152618, "learning_rate": 1.0473002138028654e-05, "loss": 0.015, "step": 80070 }, { "grad_norm": 0.18199557065963745, "learning_rate": 1.0462878277342713e-05, "loss": 0.0237, "step": 80080 }, { "grad_norm": 0.2498050481081009, "learning_rate": 1.0452758740376972e-05, "loss": 0.0223, "step": 80090 }, { "grad_norm": 0.2606622576713562, "learning_rate": 1.0442643528238067e-05, "loss": 0.0239, "step": 80100 }, { "grad_norm": 0.4912104606628418, "learning_rate": 1.0432532642032188e-05, "loss": 0.0194, "step": 80110 }, { "grad_norm": 0.3937843441963196, "learning_rate": 1.042242608286505e-05, "loss": 0.017, "step": 80120 }, { "grad_norm": 0.3604692220687866, "learning_rate": 1.041232385184191e-05, "loss": 0.0203, "step": 80130 }, { "grad_norm": 0.3245421350002289, "learning_rate": 1.0402225950067495e-05, "loss": 0.0233, "step": 80140 }, { "grad_norm": 1.0846315622329712, "learning_rate": 1.0392132378646119e-05, "loss": 0.0188, "step": 80150 }, { "grad_norm": 0.45622357726097107, "learning_rate": 1.0382043138681607e-05, "loss": 0.0219, "step": 80160 }, { "grad_norm": 0.22035159170627594, "learning_rate": 1.0371958231277284e-05, "loss": 0.0183, "step": 80170 }, { "grad_norm": 0.5446385145187378, "learning_rate": 1.0361877657536024e-05, "loss": 0.0335, "step": 80180 }, { "grad_norm": 0.3136674165725708, "learning_rate": 1.035180141856023e-05, "loss": 0.0247, "step": 80190 }, { "grad_norm": 0.36922305822372437, "learning_rate": 1.0341729515451836e-05, "loss": 0.018, "step": 80200 }, { "grad_norm": 0.4320017397403717, "learning_rate": 1.0331661949312265e-05, "loss": 0.0207, "step": 80210 }, { "grad_norm": 0.23710931837558746, "learning_rate": 1.0321598721242514e-05, "loss": 0.0155, "step": 80220 }, { "grad_norm": 0.5045467615127563, "learning_rate": 1.0311539832343064e-05, "loss": 0.0158, "step": 80230 }, { "grad_norm": 0.5448340177536011, "learning_rate": 1.0301485283713942e-05, "loss": 0.0269, "step": 80240 }, { "grad_norm": 0.4287635087966919, "learning_rate": 1.0291435076454703e-05, "loss": 0.0157, "step": 80250 }, { "grad_norm": 0.21630065143108368, "learning_rate": 1.028138921166444e-05, "loss": 0.0187, "step": 80260 }, { "grad_norm": 0.41400864720344543, "learning_rate": 1.0271347690441719e-05, "loss": 0.0189, "step": 80270 }, { "grad_norm": 0.351502925157547, "learning_rate": 1.0261310513884698e-05, "loss": 0.0215, "step": 80280 }, { "grad_norm": 0.4712991416454315, "learning_rate": 1.0251277683090998e-05, "loss": 0.0247, "step": 80290 }, { "grad_norm": 0.69432133436203, "learning_rate": 1.0241249199157826e-05, "loss": 0.0307, "step": 80300 }, { "grad_norm": 0.40338510274887085, "learning_rate": 1.0231225063181837e-05, "loss": 0.0222, "step": 80310 }, { "grad_norm": 0.3110676407814026, "learning_rate": 1.0221205276259316e-05, "loss": 0.0264, "step": 80320 }, { "grad_norm": 0.4364761710166931, "learning_rate": 1.0211189839485958e-05, "loss": 0.0254, "step": 80330 }, { "grad_norm": 0.23902279138565063, "learning_rate": 1.0201178753957075e-05, "loss": 0.014, "step": 80340 }, { "grad_norm": 0.5265188813209534, "learning_rate": 1.0191172020767432e-05, "loss": 0.0208, "step": 80350 }, { "grad_norm": 0.5870895981788635, "learning_rate": 1.0181169641011362e-05, "loss": 0.021, "step": 80360 }, { "grad_norm": 0.6034277677536011, "learning_rate": 1.0171171615782721e-05, "loss": 0.0264, "step": 80370 }, { "grad_norm": 0.2980557084083557, "learning_rate": 1.0161177946174866e-05, "loss": 0.018, "step": 80380 }, { "grad_norm": 0.30928999185562134, "learning_rate": 1.0151188633280706e-05, "loss": 0.0116, "step": 80390 }, { "grad_norm": 0.2459348738193512, "learning_rate": 1.0141203678192635e-05, "loss": 0.0145, "step": 80400 }, { "grad_norm": 0.2367127686738968, "learning_rate": 1.0131223082002612e-05, "loss": 0.0232, "step": 80410 }, { "grad_norm": 0.33060917258262634, "learning_rate": 1.012124684580208e-05, "loss": 0.0197, "step": 80420 }, { "grad_norm": 0.2575685977935791, "learning_rate": 1.0111274970682039e-05, "loss": 0.021, "step": 80430 }, { "grad_norm": 0.3473167419433594, "learning_rate": 1.010130745773299e-05, "loss": 0.0216, "step": 80440 }, { "grad_norm": 0.22445610165596008, "learning_rate": 1.0091344308044986e-05, "loss": 0.0232, "step": 80450 }, { "grad_norm": 0.3056889772415161, "learning_rate": 1.0081385522707554e-05, "loss": 0.0146, "step": 80460 }, { "grad_norm": 0.3902122676372528, "learning_rate": 1.0071431102809803e-05, "loss": 0.0261, "step": 80470 }, { "grad_norm": 0.2811206877231598, "learning_rate": 1.0061481049440297e-05, "loss": 0.0129, "step": 80480 }, { "grad_norm": 0.44785335659980774, "learning_rate": 1.0051535363687187e-05, "loss": 0.0164, "step": 80490 }, { "grad_norm": 0.3399108052253723, "learning_rate": 1.0041594046638087e-05, "loss": 0.0126, "step": 80500 }, { "grad_norm": 0.12964126467704773, "learning_rate": 1.0031657099380204e-05, "loss": 0.0148, "step": 80510 }, { "grad_norm": 0.3412955403327942, "learning_rate": 1.0021724523000204e-05, "loss": 0.0278, "step": 80520 }, { "grad_norm": 0.4198838174343109, "learning_rate": 1.001179631858431e-05, "loss": 0.0321, "step": 80530 }, { "grad_norm": 0.4434244632720947, "learning_rate": 1.0001872487218245e-05, "loss": 0.0145, "step": 80540 }, { "grad_norm": 0.4888918399810791, "learning_rate": 9.991953029987272e-06, "loss": 0.0207, "step": 80550 }, { "grad_norm": 0.3217849135398865, "learning_rate": 9.982037947976147e-06, "loss": 0.0171, "step": 80560 }, { "grad_norm": 0.1919427365064621, "learning_rate": 9.972127242269208e-06, "loss": 0.0235, "step": 80570 }, { "grad_norm": 1.333297848701477, "learning_rate": 9.962220913950238e-06, "loss": 0.0243, "step": 80580 }, { "grad_norm": 0.645839273929596, "learning_rate": 9.952318964102591e-06, "loss": 0.0293, "step": 80590 }, { "grad_norm": 0.45676958560943604, "learning_rate": 9.942421393809148e-06, "loss": 0.0157, "step": 80600 }, { "grad_norm": 0.4530580937862396, "learning_rate": 9.932528204152264e-06, "loss": 0.0209, "step": 80610 }, { "grad_norm": 0.48864296078681946, "learning_rate": 9.922639396213856e-06, "loss": 0.0235, "step": 80620 }, { "grad_norm": 0.35204818844795227, "learning_rate": 9.912754971075344e-06, "loss": 0.0216, "step": 80630 }, { "grad_norm": 0.27941420674324036, "learning_rate": 9.902874929817696e-06, "loss": 0.0152, "step": 80640 }, { "grad_norm": 0.5247713923454285, "learning_rate": 9.892999273521342e-06, "loss": 0.0253, "step": 80650 }, { "grad_norm": 0.22181634604930878, "learning_rate": 9.883128003266302e-06, "loss": 0.0214, "step": 80660 }, { "grad_norm": 0.3404107987880707, "learning_rate": 9.87326112013206e-06, "loss": 0.0192, "step": 80670 }, { "grad_norm": 1.0529173612594604, "learning_rate": 9.863398625197662e-06, "loss": 0.0193, "step": 80680 }, { "grad_norm": 0.22144092619419098, "learning_rate": 9.853540519541615e-06, "loss": 0.0226, "step": 80690 }, { "grad_norm": 0.4615080952644348, "learning_rate": 9.843686804242053e-06, "loss": 0.0202, "step": 80700 }, { "grad_norm": 0.33528539538383484, "learning_rate": 9.833837480376506e-06, "loss": 0.0219, "step": 80710 }, { "grad_norm": 0.3123117685317993, "learning_rate": 9.82399254902212e-06, "loss": 0.0197, "step": 80720 }, { "grad_norm": 0.6151888966560364, "learning_rate": 9.814152011255494e-06, "loss": 0.0214, "step": 80730 }, { "grad_norm": 0.25240465998649597, "learning_rate": 9.804315868152797e-06, "loss": 0.0182, "step": 80740 }, { "grad_norm": 0.44978299736976624, "learning_rate": 9.79448412078966e-06, "loss": 0.0177, "step": 80750 }, { "grad_norm": 0.32729777693748474, "learning_rate": 9.784656770241312e-06, "loss": 0.0151, "step": 80760 }, { "grad_norm": 0.4180264472961426, "learning_rate": 9.77483381758243e-06, "loss": 0.0176, "step": 80770 }, { "grad_norm": 0.5657193064689636, "learning_rate": 9.76501526388725e-06, "loss": 0.021, "step": 80780 }, { "grad_norm": 0.34059110283851624, "learning_rate": 9.7552011102295e-06, "loss": 0.0183, "step": 80790 }, { "grad_norm": 0.3611672818660736, "learning_rate": 9.745391357682449e-06, "loss": 0.0275, "step": 80800 }, { "grad_norm": 0.29202067852020264, "learning_rate": 9.735586007318875e-06, "loss": 0.0177, "step": 80810 }, { "grad_norm": 0.30951815843582153, "learning_rate": 9.725785060211096e-06, "loss": 0.0144, "step": 80820 }, { "grad_norm": 0.4473320543766022, "learning_rate": 9.715988517430896e-06, "loss": 0.0266, "step": 80830 }, { "grad_norm": 0.1725650131702423, "learning_rate": 9.706196380049625e-06, "loss": 0.0145, "step": 80840 }, { "grad_norm": 0.5731205940246582, "learning_rate": 9.69640864913815e-06, "loss": 0.029, "step": 80850 }, { "grad_norm": 0.4462359845638275, "learning_rate": 9.686625325766818e-06, "loss": 0.0182, "step": 80860 }, { "grad_norm": 0.31715530157089233, "learning_rate": 9.67684641100554e-06, "loss": 0.0128, "step": 80870 }, { "grad_norm": 0.461642861366272, "learning_rate": 9.667071905923692e-06, "loss": 0.016, "step": 80880 }, { "grad_norm": 1.1253267526626587, "learning_rate": 9.657301811590242e-06, "loss": 0.026, "step": 80890 }, { "grad_norm": 0.38738197088241577, "learning_rate": 9.647536129073603e-06, "loss": 0.02, "step": 80900 }, { "grad_norm": 1.0738290548324585, "learning_rate": 9.637774859441751e-06, "loss": 0.0199, "step": 80910 }, { "grad_norm": 0.5885433554649353, "learning_rate": 9.628018003762152e-06, "loss": 0.025, "step": 80920 }, { "grad_norm": 0.4531804323196411, "learning_rate": 9.618265563101813e-06, "loss": 0.0197, "step": 80930 }, { "grad_norm": 0.29370108246803284, "learning_rate": 9.608517538527211e-06, "loss": 0.0139, "step": 80940 }, { "grad_norm": 0.2714582085609436, "learning_rate": 9.598773931104431e-06, "loss": 0.0214, "step": 80950 }, { "grad_norm": 0.3132172226905823, "learning_rate": 9.589034741898983e-06, "loss": 0.0231, "step": 80960 }, { "grad_norm": 0.4016302824020386, "learning_rate": 9.579299971975947e-06, "loss": 0.0245, "step": 80970 }, { "grad_norm": 0.39215871691703796, "learning_rate": 9.569569622399882e-06, "loss": 0.0177, "step": 80980 }, { "grad_norm": 0.4429563879966736, "learning_rate": 9.559843694234916e-06, "loss": 0.0135, "step": 80990 }, { "grad_norm": 0.30330973863601685, "learning_rate": 9.550122188544625e-06, "loss": 0.0224, "step": 81000 }, { "grad_norm": 0.616601288318634, "learning_rate": 9.540405106392152e-06, "loss": 0.0228, "step": 81010 }, { "grad_norm": 0.26626813411712646, "learning_rate": 9.53069244884015e-06, "loss": 0.0172, "step": 81020 }, { "grad_norm": 0.45312049984931946, "learning_rate": 9.52098421695079e-06, "loss": 0.0223, "step": 81030 }, { "grad_norm": 0.2380315214395523, "learning_rate": 9.511280411785728e-06, "loss": 0.0189, "step": 81040 }, { "grad_norm": 0.2444254457950592, "learning_rate": 9.50158103440616e-06, "loss": 0.0201, "step": 81050 }, { "grad_norm": 0.40040111541748047, "learning_rate": 9.491886085872815e-06, "loss": 0.0136, "step": 81060 }, { "grad_norm": 0.41619232296943665, "learning_rate": 9.482195567245895e-06, "loss": 0.0204, "step": 81070 }, { "grad_norm": 0.2543231248855591, "learning_rate": 9.472509479585151e-06, "loss": 0.021, "step": 81080 }, { "grad_norm": 0.4402531683444977, "learning_rate": 9.462827823949833e-06, "loss": 0.0206, "step": 81090 }, { "grad_norm": 0.3599512279033661, "learning_rate": 9.453150601398735e-06, "loss": 0.0241, "step": 81100 }, { "grad_norm": 0.3656105399131775, "learning_rate": 9.44347781299011e-06, "loss": 0.0173, "step": 81110 }, { "grad_norm": 0.3159905672073364, "learning_rate": 9.43380945978179e-06, "loss": 0.0183, "step": 81120 }, { "grad_norm": 0.35267215967178345, "learning_rate": 9.424145542831047e-06, "loss": 0.0134, "step": 81130 }, { "grad_norm": 0.29320502281188965, "learning_rate": 9.414486063194778e-06, "loss": 0.0174, "step": 81140 }, { "grad_norm": 0.21315090358257294, "learning_rate": 9.404831021929273e-06, "loss": 0.0177, "step": 81150 }, { "grad_norm": 0.31718909740448, "learning_rate": 9.395180420090427e-06, "loss": 0.0168, "step": 81160 }, { "grad_norm": 0.26815515756607056, "learning_rate": 9.385534258733591e-06, "loss": 0.025, "step": 81170 }, { "grad_norm": 0.3782734274864197, "learning_rate": 9.375892538913677e-06, "loss": 0.0151, "step": 81180 }, { "grad_norm": 0.26036685705184937, "learning_rate": 9.36625526168507e-06, "loss": 0.0207, "step": 81190 }, { "grad_norm": 0.20278756320476532, "learning_rate": 9.356622428101686e-06, "loss": 0.0168, "step": 81200 }, { "grad_norm": 0.27570641040802, "learning_rate": 9.346994039216972e-06, "loss": 0.0197, "step": 81210 }, { "grad_norm": 0.3820480704307556, "learning_rate": 9.337370096083886e-06, "loss": 0.0177, "step": 81220 }, { "grad_norm": 0.3102254271507263, "learning_rate": 9.327750599754854e-06, "loss": 0.0168, "step": 81230 }, { "grad_norm": 0.5929415822029114, "learning_rate": 9.318135551281875e-06, "loss": 0.0149, "step": 81240 }, { "grad_norm": 0.3628952205181122, "learning_rate": 9.308524951716424e-06, "loss": 0.0164, "step": 81250 }, { "grad_norm": 0.9576228260993958, "learning_rate": 9.298918802109497e-06, "loss": 0.0212, "step": 81260 }, { "grad_norm": 0.64048171043396, "learning_rate": 9.289317103511625e-06, "loss": 0.0276, "step": 81270 }, { "grad_norm": 0.2728816270828247, "learning_rate": 9.279719856972835e-06, "loss": 0.0258, "step": 81280 }, { "grad_norm": 0.3512522876262665, "learning_rate": 9.270127063542645e-06, "loss": 0.0214, "step": 81290 }, { "grad_norm": 0.273210346698761, "learning_rate": 9.260538724270124e-06, "loss": 0.0217, "step": 81300 }, { "grad_norm": 0.7406069040298462, "learning_rate": 9.250954840203852e-06, "loss": 0.0367, "step": 81310 }, { "grad_norm": 0.22833351790905, "learning_rate": 9.241375412391878e-06, "loss": 0.0229, "step": 81320 }, { "grad_norm": 0.39796486496925354, "learning_rate": 9.23180044188181e-06, "loss": 0.0168, "step": 81330 }, { "grad_norm": 0.24828657507896423, "learning_rate": 9.222229929720755e-06, "loss": 0.0323, "step": 81340 }, { "grad_norm": 0.23467634618282318, "learning_rate": 9.212663876955336e-06, "loss": 0.0134, "step": 81350 }, { "grad_norm": 0.2924162745475769, "learning_rate": 9.20310228463166e-06, "loss": 0.0197, "step": 81360 }, { "grad_norm": 0.25766071677207947, "learning_rate": 9.19354515379539e-06, "loss": 0.0189, "step": 81370 }, { "grad_norm": 0.4310612380504608, "learning_rate": 9.183992485491666e-06, "loss": 0.0216, "step": 81380 }, { "grad_norm": 0.2765111029148102, "learning_rate": 9.174444280765154e-06, "loss": 0.0157, "step": 81390 }, { "grad_norm": 0.27035337686538696, "learning_rate": 9.164900540660032e-06, "loss": 0.0196, "step": 81400 }, { "grad_norm": 0.35802873969078064, "learning_rate": 9.155361266220008e-06, "loss": 0.0131, "step": 81410 }, { "grad_norm": 0.3349323272705078, "learning_rate": 9.145826458488254e-06, "loss": 0.0213, "step": 81420 }, { "grad_norm": 0.5166146159172058, "learning_rate": 9.136296118507509e-06, "loss": 0.0237, "step": 81430 }, { "grad_norm": 0.3527894616127014, "learning_rate": 9.126770247319966e-06, "loss": 0.0174, "step": 81440 }, { "grad_norm": 0.2522493600845337, "learning_rate": 9.11724884596738e-06, "loss": 0.0243, "step": 81450 }, { "grad_norm": 0.4600198268890381, "learning_rate": 9.107731915490996e-06, "loss": 0.0209, "step": 81460 }, { "grad_norm": 0.4320026934146881, "learning_rate": 9.098219456931579e-06, "loss": 0.0183, "step": 81470 }, { "grad_norm": 0.3505656123161316, "learning_rate": 9.088711471329376e-06, "loss": 0.0209, "step": 81480 }, { "grad_norm": 0.4713592827320099, "learning_rate": 9.079207959724184e-06, "loss": 0.0209, "step": 81490 }, { "grad_norm": 0.30872809886932373, "learning_rate": 9.069708923155279e-06, "loss": 0.016, "step": 81500 }, { "grad_norm": 0.29044222831726074, "learning_rate": 9.060214362661468e-06, "loss": 0.0197, "step": 81510 }, { "grad_norm": 0.325469434261322, "learning_rate": 9.050724279281064e-06, "loss": 0.0166, "step": 81520 }, { "grad_norm": 0.3536531925201416, "learning_rate": 9.041238674051894e-06, "loss": 0.0196, "step": 81530 }, { "grad_norm": 1.1605643033981323, "learning_rate": 9.031757548011277e-06, "loss": 0.0178, "step": 81540 }, { "grad_norm": 0.3776548206806183, "learning_rate": 9.022280902196051e-06, "loss": 0.02, "step": 81550 }, { "grad_norm": 0.31668564677238464, "learning_rate": 9.012808737642592e-06, "loss": 0.0286, "step": 81560 }, { "grad_norm": 0.5329874157905579, "learning_rate": 9.003341055386739e-06, "loss": 0.017, "step": 81570 }, { "grad_norm": 0.4769774377346039, "learning_rate": 8.993877856463862e-06, "loss": 0.0164, "step": 81580 }, { "grad_norm": 0.4685671627521515, "learning_rate": 8.984419141908856e-06, "loss": 0.022, "step": 81590 }, { "grad_norm": 0.3694010078907013, "learning_rate": 8.974964912756117e-06, "loss": 0.0197, "step": 81600 }, { "grad_norm": 0.524249792098999, "learning_rate": 8.965515170039523e-06, "loss": 0.0329, "step": 81610 }, { "grad_norm": 0.46644046902656555, "learning_rate": 8.956069914792514e-06, "loss": 0.017, "step": 81620 }, { "grad_norm": 0.1503288894891739, "learning_rate": 8.946629148047969e-06, "loss": 0.0199, "step": 81630 }, { "grad_norm": 0.5491594076156616, "learning_rate": 8.93719287083834e-06, "loss": 0.0201, "step": 81640 }, { "grad_norm": 0.4616704285144806, "learning_rate": 8.927761084195569e-06, "loss": 0.0215, "step": 81650 }, { "grad_norm": 0.36159563064575195, "learning_rate": 8.918333789151096e-06, "loss": 0.0181, "step": 81660 }, { "grad_norm": 0.5661479234695435, "learning_rate": 8.90891098673587e-06, "loss": 0.0194, "step": 81670 }, { "grad_norm": 0.25963667035102844, "learning_rate": 8.89949267798037e-06, "loss": 0.0205, "step": 81680 }, { "grad_norm": 0.31214088201522827, "learning_rate": 8.890078863914542e-06, "loss": 0.0168, "step": 81690 }, { "grad_norm": 0.24092482030391693, "learning_rate": 8.880669545567894e-06, "loss": 0.0152, "step": 81700 }, { "grad_norm": 0.529495894908905, "learning_rate": 8.871264723969381e-06, "loss": 0.0199, "step": 81710 }, { "grad_norm": 1.0622155666351318, "learning_rate": 8.86186440014754e-06, "loss": 0.0237, "step": 81720 }, { "grad_norm": 0.4198300242424011, "learning_rate": 8.852468575130351e-06, "loss": 0.0268, "step": 81730 }, { "grad_norm": 0.4124472141265869, "learning_rate": 8.843077249945347e-06, "loss": 0.0162, "step": 81740 }, { "grad_norm": 0.45750144124031067, "learning_rate": 8.833690425619523e-06, "loss": 0.0166, "step": 81750 }, { "grad_norm": 0.42611873149871826, "learning_rate": 8.82430810317942e-06, "loss": 0.0206, "step": 81760 }, { "grad_norm": 0.3362917900085449, "learning_rate": 8.814930283651074e-06, "loss": 0.0199, "step": 81770 }, { "grad_norm": 0.4984387457370758, "learning_rate": 8.805556968060048e-06, "loss": 0.022, "step": 81780 }, { "grad_norm": 0.3782373368740082, "learning_rate": 8.796188157431362e-06, "loss": 0.0193, "step": 81790 }, { "grad_norm": 0.2670893967151642, "learning_rate": 8.786823852789595e-06, "loss": 0.0282, "step": 81800 }, { "grad_norm": 0.3695670962333679, "learning_rate": 8.777464055158813e-06, "loss": 0.0232, "step": 81810 }, { "grad_norm": 0.4540467858314514, "learning_rate": 8.768108765562578e-06, "loss": 0.0207, "step": 81820 }, { "grad_norm": 0.2627059519290924, "learning_rate": 8.758757985023975e-06, "loss": 0.0145, "step": 81830 }, { "grad_norm": 0.4062761068344116, "learning_rate": 8.749411714565598e-06, "loss": 0.0263, "step": 81840 }, { "grad_norm": 0.332442969083786, "learning_rate": 8.74006995520954e-06, "loss": 0.0179, "step": 81850 }, { "grad_norm": 0.2806982100009918, "learning_rate": 8.730732707977397e-06, "loss": 0.0302, "step": 81860 }, { "grad_norm": 0.31748494505882263, "learning_rate": 8.721399973890281e-06, "loss": 0.0182, "step": 81870 }, { "grad_norm": 0.27791929244995117, "learning_rate": 8.712071753968792e-06, "loss": 0.0209, "step": 81880 }, { "grad_norm": 0.4158632457256317, "learning_rate": 8.702748049233072e-06, "loss": 0.0177, "step": 81890 }, { "grad_norm": 0.3785021901130676, "learning_rate": 8.693428860702718e-06, "loss": 0.0205, "step": 81900 }, { "grad_norm": 0.5149484872817993, "learning_rate": 8.684114189396898e-06, "loss": 0.0258, "step": 81910 }, { "grad_norm": 0.21628771722316742, "learning_rate": 8.674804036334222e-06, "loss": 0.0172, "step": 81920 }, { "grad_norm": 0.4502840042114258, "learning_rate": 8.665498402532857e-06, "loss": 0.0277, "step": 81930 }, { "grad_norm": 0.5317888855934143, "learning_rate": 8.656197289010432e-06, "loss": 0.0229, "step": 81940 }, { "grad_norm": 0.32788583636283875, "learning_rate": 8.646900696784122e-06, "loss": 0.0167, "step": 81950 }, { "grad_norm": 0.4827021062374115, "learning_rate": 8.637608626870553e-06, "loss": 0.018, "step": 81960 }, { "grad_norm": 0.3732786774635315, "learning_rate": 8.628321080285944e-06, "loss": 0.017, "step": 81970 }, { "grad_norm": 0.28411123156547546, "learning_rate": 8.619038058045925e-06, "loss": 0.0173, "step": 81980 }, { "grad_norm": 0.5403273105621338, "learning_rate": 8.609759561165704e-06, "loss": 0.0224, "step": 81990 }, { "grad_norm": 0.29792818427085876, "learning_rate": 8.600485590659934e-06, "loss": 0.0233, "step": 82000 }, { "grad_norm": 0.47041186690330505, "learning_rate": 8.591216147542813e-06, "loss": 0.0185, "step": 82010 }, { "grad_norm": 0.3437325358390808, "learning_rate": 8.581951232828045e-06, "loss": 0.0126, "step": 82020 }, { "grad_norm": 0.32763364911079407, "learning_rate": 8.572690847528825e-06, "loss": 0.0195, "step": 82030 }, { "grad_norm": 0.42062172293663025, "learning_rate": 8.563434992657838e-06, "loss": 0.0155, "step": 82040 }, { "grad_norm": 0.289908230304718, "learning_rate": 8.554183669227307e-06, "loss": 0.0156, "step": 82050 }, { "grad_norm": 0.9629996418952942, "learning_rate": 8.544936878248937e-06, "loss": 0.0357, "step": 82060 }, { "grad_norm": 0.16682109236717224, "learning_rate": 8.535694620733942e-06, "loss": 0.0278, "step": 82070 }, { "grad_norm": 0.317319393157959, "learning_rate": 8.526456897693053e-06, "loss": 0.0191, "step": 82080 }, { "grad_norm": 0.41497302055358887, "learning_rate": 8.517223710136456e-06, "loss": 0.0274, "step": 82090 }, { "grad_norm": 0.23964858055114746, "learning_rate": 8.507995059073931e-06, "loss": 0.0194, "step": 82100 }, { "grad_norm": 0.25524404644966125, "learning_rate": 8.498770945514673e-06, "loss": 0.0168, "step": 82110 }, { "grad_norm": 0.2878981828689575, "learning_rate": 8.489551370467436e-06, "loss": 0.0187, "step": 82120 }, { "grad_norm": 0.1924455761909485, "learning_rate": 8.480336334940442e-06, "loss": 0.0172, "step": 82130 }, { "grad_norm": 0.21902768313884735, "learning_rate": 8.471125839941458e-06, "loss": 0.0266, "step": 82140 }, { "grad_norm": 0.3438771665096283, "learning_rate": 8.461919886477682e-06, "loss": 0.0165, "step": 82150 }, { "grad_norm": 0.41996920108795166, "learning_rate": 8.452718475555927e-06, "loss": 0.017, "step": 82160 }, { "grad_norm": 0.296815425157547, "learning_rate": 8.443521608182398e-06, "loss": 0.0207, "step": 82170 }, { "grad_norm": 0.5649834871292114, "learning_rate": 8.43432928536288e-06, "loss": 0.0344, "step": 82180 }, { "grad_norm": 0.3513789772987366, "learning_rate": 8.425141508102607e-06, "loss": 0.0223, "step": 82190 }, { "grad_norm": 0.1313147097826004, "learning_rate": 8.415958277406365e-06, "loss": 0.0278, "step": 82200 }, { "grad_norm": 0.3658313751220703, "learning_rate": 8.406779594278392e-06, "loss": 0.0177, "step": 82210 }, { "grad_norm": 0.16363558173179626, "learning_rate": 8.397605459722468e-06, "loss": 0.0246, "step": 82220 }, { "grad_norm": 0.24519997835159302, "learning_rate": 8.388435874741863e-06, "loss": 0.0184, "step": 82230 }, { "grad_norm": 0.24198904633522034, "learning_rate": 8.37927084033936e-06, "loss": 0.0274, "step": 82240 }, { "grad_norm": 0.22516703605651855, "learning_rate": 8.370110357517208e-06, "loss": 0.0192, "step": 82250 }, { "grad_norm": 0.43013840913772583, "learning_rate": 8.360954427277201e-06, "loss": 0.016, "step": 82260 }, { "grad_norm": 0.3943328559398651, "learning_rate": 8.351803050620627e-06, "loss": 0.0184, "step": 82270 }, { "grad_norm": 0.7371727228164673, "learning_rate": 8.34265622854824e-06, "loss": 0.0302, "step": 82280 }, { "grad_norm": 0.36558565497398376, "learning_rate": 8.333513962060335e-06, "loss": 0.0177, "step": 82290 }, { "grad_norm": 0.9359576106071472, "learning_rate": 8.324376252156701e-06, "loss": 0.0213, "step": 82300 }, { "grad_norm": 0.4267309606075287, "learning_rate": 8.315243099836633e-06, "loss": 0.0289, "step": 82310 }, { "grad_norm": 0.3684951961040497, "learning_rate": 8.3061145060989e-06, "loss": 0.0237, "step": 82320 }, { "grad_norm": 0.26021239161491394, "learning_rate": 8.296990471941806e-06, "loss": 0.0183, "step": 82330 }, { "grad_norm": 0.43000122904777527, "learning_rate": 8.287870998363112e-06, "loss": 0.0272, "step": 82340 }, { "grad_norm": 0.3085313141345978, "learning_rate": 8.278756086360156e-06, "loss": 0.0166, "step": 82350 }, { "grad_norm": 0.3934686779975891, "learning_rate": 8.269645736929693e-06, "loss": 0.0185, "step": 82360 }, { "grad_norm": 0.4057989716529846, "learning_rate": 8.260539951068048e-06, "loss": 0.0198, "step": 82370 }, { "grad_norm": 0.44668248295783997, "learning_rate": 8.251438729770983e-06, "loss": 0.0151, "step": 82380 }, { "grad_norm": 0.3891921043395996, "learning_rate": 8.242342074033826e-06, "loss": 0.015, "step": 82390 }, { "grad_norm": 0.4086272418498993, "learning_rate": 8.233249984851344e-06, "loss": 0.0136, "step": 82400 }, { "grad_norm": 0.16633404791355133, "learning_rate": 8.224162463217849e-06, "loss": 0.022, "step": 82410 }, { "grad_norm": 0.4847957491874695, "learning_rate": 8.215079510127139e-06, "loss": 0.0167, "step": 82420 }, { "grad_norm": 0.2971716821193695, "learning_rate": 8.206001126572521e-06, "loss": 0.0137, "step": 82430 }, { "grad_norm": 0.681829571723938, "learning_rate": 8.196927313546776e-06, "loss": 0.0136, "step": 82440 }, { "grad_norm": 0.4887486696243286, "learning_rate": 8.18785807204222e-06, "loss": 0.0252, "step": 82450 }, { "grad_norm": 0.36561834812164307, "learning_rate": 8.178793403050627e-06, "loss": 0.0261, "step": 82460 }, { "grad_norm": 0.5756561160087585, "learning_rate": 8.16973330756331e-06, "loss": 0.0174, "step": 82470 }, { "grad_norm": 0.43584373593330383, "learning_rate": 8.16067778657107e-06, "loss": 0.0124, "step": 82480 }, { "grad_norm": 0.31859731674194336, "learning_rate": 8.151626841064214e-06, "loss": 0.016, "step": 82490 }, { "grad_norm": 0.2195834517478943, "learning_rate": 8.142580472032518e-06, "loss": 0.0274, "step": 82500 }, { "grad_norm": 0.2702949345111847, "learning_rate": 8.133538680465285e-06, "loss": 0.0293, "step": 82510 }, { "grad_norm": 0.33222585916519165, "learning_rate": 8.124501467351336e-06, "loss": 0.0191, "step": 82520 }, { "grad_norm": 0.26212698221206665, "learning_rate": 8.11546883367893e-06, "loss": 0.0169, "step": 82530 }, { "grad_norm": 0.3984968364238739, "learning_rate": 8.106440780435882e-06, "loss": 0.0291, "step": 82540 }, { "grad_norm": 0.8763499855995178, "learning_rate": 8.097417308609485e-06, "loss": 0.0262, "step": 82550 }, { "grad_norm": 0.4114561378955841, "learning_rate": 8.088398419186544e-06, "loss": 0.0176, "step": 82560 }, { "grad_norm": 0.3386181890964508, "learning_rate": 8.079384113153321e-06, "loss": 0.0112, "step": 82570 }, { "grad_norm": 0.3235434591770172, "learning_rate": 8.070374391495644e-06, "loss": 0.0216, "step": 82580 }, { "grad_norm": 0.2914975583553314, "learning_rate": 8.061369255198764e-06, "loss": 0.0195, "step": 82590 }, { "grad_norm": 0.23988251388072968, "learning_rate": 8.052368705247493e-06, "loss": 0.0184, "step": 82600 }, { "grad_norm": 0.43477460741996765, "learning_rate": 8.043372742626114e-06, "loss": 0.0215, "step": 82610 }, { "grad_norm": 0.49528244137763977, "learning_rate": 8.034381368318422e-06, "loss": 0.0196, "step": 82620 }, { "grad_norm": 0.31701043248176575, "learning_rate": 8.025394583307677e-06, "loss": 0.014, "step": 82630 }, { "grad_norm": 0.165786013007164, "learning_rate": 8.016412388576683e-06, "loss": 0.0184, "step": 82640 }, { "grad_norm": 0.43381550908088684, "learning_rate": 8.007434785107699e-06, "loss": 0.0208, "step": 82650 }, { "grad_norm": 0.46347227692604065, "learning_rate": 7.998461773882509e-06, "loss": 0.0191, "step": 82660 }, { "grad_norm": 0.24407735466957092, "learning_rate": 7.989493355882393e-06, "loss": 0.0126, "step": 82670 }, { "grad_norm": 0.4049835205078125, "learning_rate": 7.980529532088137e-06, "loss": 0.0178, "step": 82680 }, { "grad_norm": 0.2726251780986786, "learning_rate": 7.97157030347998e-06, "loss": 0.0231, "step": 82690 }, { "grad_norm": 0.48565673828125, "learning_rate": 7.96261567103772e-06, "loss": 0.0164, "step": 82700 }, { "grad_norm": 0.3231956660747528, "learning_rate": 7.953665635740593e-06, "loss": 0.0135, "step": 82710 }, { "grad_norm": 0.3311989903450012, "learning_rate": 7.94472019856738e-06, "loss": 0.0176, "step": 82720 }, { "grad_norm": 0.1344875991344452, "learning_rate": 7.935779360496337e-06, "loss": 0.0155, "step": 82730 }, { "grad_norm": 0.281660795211792, "learning_rate": 7.92684312250523e-06, "loss": 0.0183, "step": 82740 }, { "grad_norm": 0.41104796528816223, "learning_rate": 7.917911485571289e-06, "loss": 0.0264, "step": 82750 }, { "grad_norm": 0.33731159567832947, "learning_rate": 7.90898445067128e-06, "loss": 0.0159, "step": 82760 }, { "grad_norm": 0.41117677092552185, "learning_rate": 7.900062018781462e-06, "loss": 0.0221, "step": 82770 }, { "grad_norm": 0.4281074106693268, "learning_rate": 7.891144190877553e-06, "loss": 0.0154, "step": 82780 }, { "grad_norm": 0.2542343735694885, "learning_rate": 7.882230967934806e-06, "loss": 0.0128, "step": 82790 }, { "grad_norm": 0.3904515206813812, "learning_rate": 7.873322350927959e-06, "loss": 0.0145, "step": 82800 }, { "grad_norm": 0.7103496789932251, "learning_rate": 7.864418340831253e-06, "loss": 0.0212, "step": 82810 }, { "grad_norm": 0.2719082534313202, "learning_rate": 7.855518938618395e-06, "loss": 0.013, "step": 82820 }, { "grad_norm": 0.3391469717025757, "learning_rate": 7.846624145262633e-06, "loss": 0.0176, "step": 82830 }, { "grad_norm": 0.43213146924972534, "learning_rate": 7.837733961736672e-06, "loss": 0.0148, "step": 82840 }, { "grad_norm": 0.6398100256919861, "learning_rate": 7.828848389012727e-06, "loss": 0.0155, "step": 82850 }, { "grad_norm": 0.4184664189815521, "learning_rate": 7.819967428062524e-06, "loss": 0.0128, "step": 82860 }, { "grad_norm": 0.3476303815841675, "learning_rate": 7.811091079857274e-06, "loss": 0.0196, "step": 82870 }, { "grad_norm": 0.3236633539199829, "learning_rate": 7.802219345367667e-06, "loss": 0.0159, "step": 82880 }, { "grad_norm": 0.30223312973976135, "learning_rate": 7.793352225563916e-06, "loss": 0.0231, "step": 82890 }, { "grad_norm": 0.26600581407546997, "learning_rate": 7.7844897214157e-06, "loss": 0.0181, "step": 82900 }, { "grad_norm": 0.364523321390152, "learning_rate": 7.775631833892228e-06, "loss": 0.0246, "step": 82910 }, { "grad_norm": 0.3228280246257782, "learning_rate": 7.76677856396215e-06, "loss": 0.0222, "step": 82920 }, { "grad_norm": 0.39946091175079346, "learning_rate": 7.757929912593697e-06, "loss": 0.0184, "step": 82930 }, { "grad_norm": 0.4509463906288147, "learning_rate": 7.749085880754509e-06, "loss": 0.0146, "step": 82940 }, { "grad_norm": 0.30306166410446167, "learning_rate": 7.740246469411777e-06, "loss": 0.0165, "step": 82950 }, { "grad_norm": 0.27699610590934753, "learning_rate": 7.731411679532142e-06, "loss": 0.0283, "step": 82960 }, { "grad_norm": 0.2756240963935852, "learning_rate": 7.722581512081779e-06, "loss": 0.0193, "step": 82970 }, { "grad_norm": 0.24541379511356354, "learning_rate": 7.713755968026331e-06, "loss": 0.0211, "step": 82980 }, { "grad_norm": 0.12423725426197052, "learning_rate": 7.704935048330969e-06, "loss": 0.0121, "step": 82990 }, { "grad_norm": 0.3035317659378052, "learning_rate": 7.696118753960313e-06, "loss": 0.016, "step": 83000 }, { "grad_norm": 0.1921539306640625, "learning_rate": 7.687307085878503e-06, "loss": 0.0173, "step": 83010 }, { "grad_norm": 0.2660820484161377, "learning_rate": 7.678500045049186e-06, "loss": 0.0262, "step": 83020 }, { "grad_norm": 0.3811953663825989, "learning_rate": 7.669697632435463e-06, "loss": 0.0275, "step": 83030 }, { "grad_norm": 0.37070798873901367, "learning_rate": 7.660899848999963e-06, "loss": 0.0231, "step": 83040 }, { "grad_norm": 0.2807198762893677, "learning_rate": 7.652106695704803e-06, "loss": 0.0254, "step": 83050 }, { "grad_norm": 0.33659517765045166, "learning_rate": 7.643318173511598e-06, "loss": 0.0159, "step": 83060 }, { "grad_norm": 0.49478673934936523, "learning_rate": 7.634534283381423e-06, "loss": 0.0137, "step": 83070 }, { "grad_norm": 0.3861931562423706, "learning_rate": 7.625755026274889e-06, "loss": 0.014, "step": 83080 }, { "grad_norm": 0.7021096348762512, "learning_rate": 7.616980403152069e-06, "loss": 0.0185, "step": 83090 }, { "grad_norm": 0.3116879165172577, "learning_rate": 7.608210414972561e-06, "loss": 0.0178, "step": 83100 }, { "grad_norm": 0.36114856600761414, "learning_rate": 7.599445062695404e-06, "loss": 0.013, "step": 83110 }, { "grad_norm": 0.33896058797836304, "learning_rate": 7.590684347279214e-06, "loss": 0.0212, "step": 83120 }, { "grad_norm": 1.8142708539962769, "learning_rate": 7.581928269682004e-06, "loss": 0.0175, "step": 83130 }, { "grad_norm": 0.7611607909202576, "learning_rate": 7.573176830861361e-06, "loss": 0.0309, "step": 83140 }, { "grad_norm": 0.383319228887558, "learning_rate": 7.5644300317743e-06, "loss": 0.02, "step": 83150 }, { "grad_norm": 0.22999761998653412, "learning_rate": 7.555687873377376e-06, "loss": 0.0333, "step": 83160 }, { "grad_norm": 0.39486798644065857, "learning_rate": 7.546950356626592e-06, "loss": 0.011, "step": 83170 }, { "grad_norm": 0.3492144048213959, "learning_rate": 7.538217482477516e-06, "loss": 0.0226, "step": 83180 }, { "grad_norm": 0.6305081844329834, "learning_rate": 7.529489251885119e-06, "loss": 0.0188, "step": 83190 }, { "grad_norm": 0.18135562539100647, "learning_rate": 7.520765665803941e-06, "loss": 0.0142, "step": 83200 }, { "grad_norm": 0.513916015625, "learning_rate": 7.512046725187949e-06, "loss": 0.0217, "step": 83210 }, { "grad_norm": 0.21657219529151917, "learning_rate": 7.503332430990645e-06, "loss": 0.0169, "step": 83220 }, { "grad_norm": 0.1755487620830536, "learning_rate": 7.494622784165017e-06, "loss": 0.0179, "step": 83230 }, { "grad_norm": 0.39091062545776367, "learning_rate": 7.485917785663543e-06, "loss": 0.0243, "step": 83240 }, { "grad_norm": 0.4450530707836151, "learning_rate": 7.477217436438172e-06, "loss": 0.0292, "step": 83250 }, { "grad_norm": 0.367420494556427, "learning_rate": 7.46852173744037e-06, "loss": 0.0143, "step": 83260 }, { "grad_norm": 0.19958896934986115, "learning_rate": 7.4598306896211004e-06, "loss": 0.0135, "step": 83270 }, { "grad_norm": 0.42953017354011536, "learning_rate": 7.451144293930773e-06, "loss": 0.021, "step": 83280 }, { "grad_norm": 0.2910695970058441, "learning_rate": 7.4424625513193455e-06, "loss": 0.0162, "step": 83290 }, { "grad_norm": 0.3772861361503601, "learning_rate": 7.433785462736209e-06, "loss": 0.0165, "step": 83300 }, { "grad_norm": 0.14227233827114105, "learning_rate": 7.425113029130315e-06, "loss": 0.0327, "step": 83310 }, { "grad_norm": 0.3383043706417084, "learning_rate": 7.416445251450044e-06, "loss": 0.0196, "step": 83320 }, { "grad_norm": 0.23892779648303986, "learning_rate": 7.4077821306433035e-06, "loss": 0.0154, "step": 83330 }, { "grad_norm": 0.4464219808578491, "learning_rate": 7.399123667657465e-06, "loss": 0.0182, "step": 83340 }, { "grad_norm": 0.6078747510910034, "learning_rate": 7.390469863439426e-06, "loss": 0.0161, "step": 83350 }, { "grad_norm": 0.3930400609970093, "learning_rate": 7.381820718935512e-06, "loss": 0.0218, "step": 83360 }, { "grad_norm": 0.232501819729805, "learning_rate": 7.373176235091633e-06, "loss": 0.0204, "step": 83370 }, { "grad_norm": 0.3574502170085907, "learning_rate": 7.364536412853101e-06, "loss": 0.0119, "step": 83380 }, { "grad_norm": 0.3275676369667053, "learning_rate": 7.355901253164776e-06, "loss": 0.0203, "step": 83390 }, { "grad_norm": 0.3502048850059509, "learning_rate": 7.34727075697097e-06, "loss": 0.0171, "step": 83400 }, { "grad_norm": 0.5232555866241455, "learning_rate": 7.3386449252155095e-06, "loss": 0.0162, "step": 83410 }, { "grad_norm": 0.4651315212249756, "learning_rate": 7.330023758841692e-06, "loss": 0.0184, "step": 83420 }, { "grad_norm": 0.23021528124809265, "learning_rate": 7.321407258792323e-06, "loss": 0.0136, "step": 83430 }, { "grad_norm": 0.3771357834339142, "learning_rate": 7.312795426009694e-06, "loss": 0.0213, "step": 83440 }, { "grad_norm": 0.48475390672683716, "learning_rate": 7.304188261435585e-06, "loss": 0.0297, "step": 83450 }, { "grad_norm": 0.44057509303092957, "learning_rate": 7.295585766011242e-06, "loss": 0.0282, "step": 83460 }, { "grad_norm": 0.3825509548187256, "learning_rate": 7.286987940677436e-06, "loss": 0.0171, "step": 83470 }, { "grad_norm": 0.34443119168281555, "learning_rate": 7.278394786374426e-06, "loss": 0.0202, "step": 83480 }, { "grad_norm": 0.39985817670822144, "learning_rate": 7.269806304041915e-06, "loss": 0.0157, "step": 83490 }, { "grad_norm": 0.3619075417518616, "learning_rate": 7.261222494619147e-06, "loss": 0.0194, "step": 83500 }, { "grad_norm": 0.3843230903148651, "learning_rate": 7.252643359044825e-06, "loss": 0.0212, "step": 83510 }, { "grad_norm": 0.2846147418022156, "learning_rate": 7.244068898257172e-06, "loss": 0.0224, "step": 83520 }, { "grad_norm": 1.3010402917861938, "learning_rate": 7.2354991131938495e-06, "loss": 0.0201, "step": 83530 }, { "grad_norm": 0.3604026138782501, "learning_rate": 7.2269340047920574e-06, "loss": 0.016, "step": 83540 }, { "grad_norm": 0.5152834057807922, "learning_rate": 7.218373573988435e-06, "loss": 0.0199, "step": 83550 }, { "grad_norm": 0.26786351203918457, "learning_rate": 7.209817821719178e-06, "loss": 0.0128, "step": 83560 }, { "grad_norm": 0.3169253170490265, "learning_rate": 7.201266748919894e-06, "loss": 0.0205, "step": 83570 }, { "grad_norm": 0.5624637007713318, "learning_rate": 7.19272035652575e-06, "loss": 0.015, "step": 83580 }, { "grad_norm": 0.7607636451721191, "learning_rate": 7.184178645471334e-06, "loss": 0.0216, "step": 83590 }, { "grad_norm": 0.3137294054031372, "learning_rate": 7.175641616690775e-06, "loss": 0.0116, "step": 83600 }, { "grad_norm": 0.2628992795944214, "learning_rate": 7.167109271117661e-06, "loss": 0.0143, "step": 83610 }, { "grad_norm": 0.24275867640972137, "learning_rate": 7.158581609685072e-06, "loss": 0.0193, "step": 83620 }, { "grad_norm": 0.46131354570388794, "learning_rate": 7.150058633325591e-06, "loss": 0.0175, "step": 83630 }, { "grad_norm": 0.4566955268383026, "learning_rate": 7.141540342971287e-06, "loss": 0.016, "step": 83640 }, { "grad_norm": 0.31791627407073975, "learning_rate": 7.133026739553677e-06, "loss": 0.021, "step": 83650 }, { "grad_norm": 0.4804497957229614, "learning_rate": 7.124517824003829e-06, "loss": 0.0211, "step": 83660 }, { "grad_norm": 0.09806106984615326, "learning_rate": 7.116013597252236e-06, "loss": 0.0213, "step": 83670 }, { "grad_norm": 0.45691636204719543, "learning_rate": 7.107514060228921e-06, "loss": 0.0134, "step": 83680 }, { "grad_norm": 0.40225279331207275, "learning_rate": 7.099019213863378e-06, "loss": 0.02, "step": 83690 }, { "grad_norm": 0.37579721212387085, "learning_rate": 7.090529059084611e-06, "loss": 0.0163, "step": 83700 }, { "grad_norm": 0.36441048979759216, "learning_rate": 7.082043596821058e-06, "loss": 0.0165, "step": 83710 }, { "grad_norm": 0.3725510835647583, "learning_rate": 7.07356282800069e-06, "loss": 0.044, "step": 83720 }, { "grad_norm": 0.1644771844148636, "learning_rate": 7.0650867535509645e-06, "loss": 0.0241, "step": 83730 }, { "grad_norm": 0.2610180974006653, "learning_rate": 7.0566153743987895e-06, "loss": 0.0192, "step": 83740 }, { "grad_norm": 0.29747629165649414, "learning_rate": 7.04814869147059e-06, "loss": 0.0138, "step": 83750 }, { "grad_norm": 0.18327273428440094, "learning_rate": 7.0396867056922725e-06, "loss": 0.016, "step": 83760 }, { "grad_norm": 0.2689412236213684, "learning_rate": 7.031229417989244e-06, "loss": 0.0182, "step": 83770 }, { "grad_norm": 0.27595940232276917, "learning_rate": 7.022776829286348e-06, "loss": 0.0167, "step": 83780 }, { "grad_norm": 0.32217660546302795, "learning_rate": 7.014328940507975e-06, "loss": 0.029, "step": 83790 }, { "grad_norm": 0.22727717459201813, "learning_rate": 7.005885752577945e-06, "loss": 0.0149, "step": 83800 }, { "grad_norm": 0.3341604173183441, "learning_rate": 6.997447266419615e-06, "loss": 0.024, "step": 83810 }, { "grad_norm": 0.45143646001815796, "learning_rate": 6.989013482955797e-06, "loss": 0.021, "step": 83820 }, { "grad_norm": 0.29248422384262085, "learning_rate": 6.980584403108808e-06, "loss": 0.02, "step": 83830 }, { "grad_norm": 0.1796644777059555, "learning_rate": 6.972160027800417e-06, "loss": 0.017, "step": 83840 }, { "grad_norm": 0.3879447281360626, "learning_rate": 6.963740357951931e-06, "loss": 0.0201, "step": 83850 }, { "grad_norm": 0.2948426306247711, "learning_rate": 6.95532539448408e-06, "loss": 0.0159, "step": 83860 }, { "grad_norm": 0.29669007658958435, "learning_rate": 6.946915138317128e-06, "loss": 0.0124, "step": 83870 }, { "grad_norm": 0.5202192664146423, "learning_rate": 6.9385095903708065e-06, "loss": 0.0202, "step": 83880 }, { "grad_norm": 0.2741960883140564, "learning_rate": 6.930108751564346e-06, "loss": 0.03, "step": 83890 }, { "grad_norm": 0.23378804326057434, "learning_rate": 6.921712622816428e-06, "loss": 0.0218, "step": 83900 }, { "grad_norm": 0.3297278881072998, "learning_rate": 6.913321205045259e-06, "loss": 0.0166, "step": 83910 }, { "grad_norm": 0.4433566629886627, "learning_rate": 6.904934499168491e-06, "loss": 0.0175, "step": 83920 }, { "grad_norm": 0.41037777066230774, "learning_rate": 6.896552506103299e-06, "loss": 0.0202, "step": 83930 }, { "grad_norm": 0.25350868701934814, "learning_rate": 6.888175226766313e-06, "loss": 0.016, "step": 83940 }, { "grad_norm": 0.3841308355331421, "learning_rate": 6.879802662073686e-06, "loss": 0.0093, "step": 83950 }, { "grad_norm": 0.2150317281484604, "learning_rate": 6.871434812940991e-06, "loss": 0.0141, "step": 83960 }, { "grad_norm": 0.272213876247406, "learning_rate": 6.863071680283345e-06, "loss": 0.0177, "step": 83970 }, { "grad_norm": 0.22703389823436737, "learning_rate": 6.854713265015339e-06, "loss": 0.0252, "step": 83980 }, { "grad_norm": 0.3387727737426758, "learning_rate": 6.8463595680510025e-06, "loss": 0.0218, "step": 83990 }, { "grad_norm": 0.1809462308883667, "learning_rate": 6.838010590303911e-06, "loss": 0.0109, "step": 84000 }, { "grad_norm": 0.19672849774360657, "learning_rate": 6.829666332687085e-06, "loss": 0.016, "step": 84010 }, { "grad_norm": 0.4299977719783783, "learning_rate": 6.821326796113059e-06, "loss": 0.0163, "step": 84020 }, { "grad_norm": 0.4093586802482605, "learning_rate": 6.812991981493799e-06, "loss": 0.0154, "step": 84030 }, { "grad_norm": 0.2043880820274353, "learning_rate": 6.804661889740821e-06, "loss": 0.0237, "step": 84040 }, { "grad_norm": 0.2570289969444275, "learning_rate": 6.7963365217650625e-06, "loss": 0.0258, "step": 84050 }, { "grad_norm": 0.2827862501144409, "learning_rate": 6.788015878476983e-06, "loss": 0.0191, "step": 84060 }, { "grad_norm": 0.3043424189090729, "learning_rate": 6.779699960786523e-06, "loss": 0.0202, "step": 84070 }, { "grad_norm": 0.22405262291431427, "learning_rate": 6.771388769603104e-06, "loss": 0.0445, "step": 84080 }, { "grad_norm": 0.5879831910133362, "learning_rate": 6.763082305835605e-06, "loss": 0.0189, "step": 84090 }, { "grad_norm": 0.322864830493927, "learning_rate": 6.754780570392433e-06, "loss": 0.0245, "step": 84100 }, { "grad_norm": 0.285577654838562, "learning_rate": 6.7464835641814215e-06, "loss": 0.0152, "step": 84110 }, { "grad_norm": 0.22100839018821716, "learning_rate": 6.7381912881099515e-06, "loss": 0.0195, "step": 84120 }, { "grad_norm": 0.45981350541114807, "learning_rate": 6.72990374308482e-06, "loss": 0.0194, "step": 84130 }, { "grad_norm": 0.4358547329902649, "learning_rate": 6.721620930012379e-06, "loss": 0.0164, "step": 84140 }, { "grad_norm": 0.19239988923072815, "learning_rate": 6.713342849798393e-06, "loss": 0.0163, "step": 84150 }, { "grad_norm": 0.5098431706428528, "learning_rate": 6.705069503348166e-06, "loss": 0.023, "step": 84160 }, { "grad_norm": 0.3069351315498352, "learning_rate": 6.696800891566435e-06, "loss": 0.0142, "step": 84170 }, { "grad_norm": 0.2613096833229065, "learning_rate": 6.688537015357449e-06, "loss": 0.0345, "step": 84180 }, { "grad_norm": 0.25789257884025574, "learning_rate": 6.680277875624941e-06, "loss": 0.0236, "step": 84190 }, { "grad_norm": 0.3646676242351532, "learning_rate": 6.672023473272121e-06, "loss": 0.0217, "step": 84200 }, { "grad_norm": 0.639302670955658, "learning_rate": 6.663773809201667e-06, "loss": 0.0197, "step": 84210 }, { "grad_norm": 0.3565708100795746, "learning_rate": 6.655528884315748e-06, "loss": 0.0249, "step": 84220 }, { "grad_norm": 0.4244288504123688, "learning_rate": 6.647288699516036e-06, "loss": 0.0277, "step": 84230 }, { "grad_norm": 0.3917115032672882, "learning_rate": 6.639053255703642e-06, "loss": 0.0165, "step": 84240 }, { "grad_norm": 0.485258549451828, "learning_rate": 6.6308225537791925e-06, "loss": 0.019, "step": 84250 }, { "grad_norm": 0.3088136613368988, "learning_rate": 6.622596594642777e-06, "loss": 0.0193, "step": 84260 }, { "grad_norm": 0.30200549960136414, "learning_rate": 6.614375379193999e-06, "loss": 0.0182, "step": 84270 }, { "grad_norm": 0.09898293018341064, "learning_rate": 6.606158908331889e-06, "loss": 0.0159, "step": 84280 }, { "grad_norm": 0.31681951880455017, "learning_rate": 6.597947182955005e-06, "loss": 0.02, "step": 84290 }, { "grad_norm": 0.45551419258117676, "learning_rate": 6.589740203961359e-06, "loss": 0.0178, "step": 84300 }, { "grad_norm": 0.2296694666147232, "learning_rate": 6.581537972248464e-06, "loss": 0.0168, "step": 84310 }, { "grad_norm": 0.4692140221595764, "learning_rate": 6.573340488713275e-06, "loss": 0.0192, "step": 84320 }, { "grad_norm": 0.6102741360664368, "learning_rate": 6.565147754252304e-06, "loss": 0.0162, "step": 84330 }, { "grad_norm": 0.40428075194358826, "learning_rate": 6.556959769761462e-06, "loss": 0.016, "step": 84340 }, { "grad_norm": 0.3311254382133484, "learning_rate": 6.548776536136192e-06, "loss": 0.0165, "step": 84350 }, { "grad_norm": 0.3041536509990692, "learning_rate": 6.54059805427138e-06, "loss": 0.0181, "step": 84360 }, { "grad_norm": 0.1336905062198639, "learning_rate": 6.532424325061437e-06, "loss": 0.0159, "step": 84370 }, { "grad_norm": 0.33962517976760864, "learning_rate": 6.5242553494001935e-06, "loss": 0.0175, "step": 84380 }, { "grad_norm": 0.5034884810447693, "learning_rate": 6.516091128181046e-06, "loss": 0.0145, "step": 84390 }, { "grad_norm": 0.5524750351905823, "learning_rate": 6.507931662296785e-06, "loss": 0.0161, "step": 84400 }, { "grad_norm": 0.5432736277580261, "learning_rate": 6.4997769526397354e-06, "loss": 0.0324, "step": 84410 }, { "grad_norm": 0.2718958854675293, "learning_rate": 6.491627000101663e-06, "loss": 0.0149, "step": 84420 }, { "grad_norm": 0.2637344300746918, "learning_rate": 6.4834818055738545e-06, "loss": 0.0097, "step": 84430 }, { "grad_norm": 0.40841367840766907, "learning_rate": 6.4753413699470465e-06, "loss": 0.0192, "step": 84440 }, { "grad_norm": 0.6006030440330505, "learning_rate": 6.467205694111478e-06, "loss": 0.0143, "step": 84450 }, { "grad_norm": 0.22414936125278473, "learning_rate": 6.459074778956836e-06, "loss": 0.0195, "step": 84460 }, { "grad_norm": 0.26072484254837036, "learning_rate": 6.450948625372311e-06, "loss": 0.0191, "step": 84470 }, { "grad_norm": 1.165408968925476, "learning_rate": 6.442827234246585e-06, "loss": 0.0234, "step": 84480 }, { "grad_norm": 0.5462142825126648, "learning_rate": 6.43471060646777e-06, "loss": 0.018, "step": 84490 }, { "grad_norm": 0.6462931036949158, "learning_rate": 6.426598742923512e-06, "loss": 0.0192, "step": 84500 }, { "grad_norm": 0.34171605110168457, "learning_rate": 6.418491644500885e-06, "loss": 0.0164, "step": 84510 }, { "grad_norm": 0.7052497267723083, "learning_rate": 6.410389312086512e-06, "loss": 0.0181, "step": 84520 }, { "grad_norm": 0.24168963730335236, "learning_rate": 6.4022917465664135e-06, "loss": 0.0162, "step": 84530 }, { "grad_norm": 0.23763316869735718, "learning_rate": 6.3941989488261575e-06, "loss": 0.0171, "step": 84540 }, { "grad_norm": 0.4465181827545166, "learning_rate": 6.386110919750726e-06, "loss": 0.0166, "step": 84550 }, { "grad_norm": 0.14601926505565643, "learning_rate": 6.378027660224651e-06, "loss": 0.0177, "step": 84560 }, { "grad_norm": 0.35340043902397156, "learning_rate": 6.369949171131856e-06, "loss": 0.0154, "step": 84570 }, { "grad_norm": 0.5136469006538391, "learning_rate": 6.361875453355848e-06, "loss": 0.0292, "step": 84580 }, { "grad_norm": 0.21926358342170715, "learning_rate": 6.353806507779525e-06, "loss": 0.0148, "step": 84590 }, { "grad_norm": 0.40964093804359436, "learning_rate": 6.3457423352853044e-06, "loss": 0.014, "step": 84600 }, { "grad_norm": 0.2815467417240143, "learning_rate": 6.337682936755062e-06, "loss": 0.0202, "step": 84610 }, { "grad_norm": 0.6403186917304993, "learning_rate": 6.329628313070179e-06, "loss": 0.0152, "step": 84620 }, { "grad_norm": 0.3598829209804535, "learning_rate": 6.321578465111477e-06, "loss": 0.0166, "step": 84630 }, { "grad_norm": 0.47242122888565063, "learning_rate": 6.31353339375928e-06, "loss": 0.0215, "step": 84640 }, { "grad_norm": 0.21263247728347778, "learning_rate": 6.305493099893395e-06, "loss": 0.017, "step": 84650 }, { "grad_norm": 0.3575522005558014, "learning_rate": 6.297457584393096e-06, "loss": 0.0172, "step": 84660 }, { "grad_norm": 0.49420785903930664, "learning_rate": 6.289426848137126e-06, "loss": 0.023, "step": 84670 }, { "grad_norm": 0.6698322892189026, "learning_rate": 6.281400892003713e-06, "loss": 0.0278, "step": 84680 }, { "grad_norm": 0.6725016236305237, "learning_rate": 6.273379716870581e-06, "loss": 0.0147, "step": 84690 }, { "grad_norm": 0.338456392288208, "learning_rate": 6.265363323614887e-06, "loss": 0.027, "step": 84700 }, { "grad_norm": 0.3110547959804535, "learning_rate": 6.257351713113308e-06, "loss": 0.0156, "step": 84710 }, { "grad_norm": 0.4048813581466675, "learning_rate": 6.249344886241981e-06, "loss": 0.0199, "step": 84720 }, { "grad_norm": 0.20712675154209137, "learning_rate": 6.2413428438765295e-06, "loss": 0.0162, "step": 84730 }, { "grad_norm": 0.4925706088542938, "learning_rate": 6.233345586892026e-06, "loss": 0.0214, "step": 84740 }, { "grad_norm": 0.32487666606903076, "learning_rate": 6.225353116163052e-06, "loss": 0.0143, "step": 84750 }, { "grad_norm": 0.22663025557994843, "learning_rate": 6.217365432563627e-06, "loss": 0.0139, "step": 84760 }, { "grad_norm": 0.37894049286842346, "learning_rate": 6.209382536967318e-06, "loss": 0.0177, "step": 84770 }, { "grad_norm": 0.48548179864883423, "learning_rate": 6.201404430247082e-06, "loss": 0.0175, "step": 84780 }, { "grad_norm": 1.694295048713684, "learning_rate": 6.1934311132754165e-06, "loss": 0.0316, "step": 84790 }, { "grad_norm": 0.4830690324306488, "learning_rate": 6.185462586924257e-06, "loss": 0.0207, "step": 84800 }, { "grad_norm": 0.46956396102905273, "learning_rate": 6.177498852065039e-06, "loss": 0.017, "step": 84810 }, { "grad_norm": 1.192205548286438, "learning_rate": 6.169539909568655e-06, "loss": 0.0185, "step": 84820 }, { "grad_norm": 0.46896886825561523, "learning_rate": 6.161585760305488e-06, "loss": 0.0175, "step": 84830 }, { "grad_norm": 0.2461612969636917, "learning_rate": 6.153636405145391e-06, "loss": 0.0132, "step": 84840 }, { "grad_norm": 0.653302013874054, "learning_rate": 6.145691844957702e-06, "loss": 0.0263, "step": 84850 }, { "grad_norm": 0.26268503069877625, "learning_rate": 6.1377520806112145e-06, "loss": 0.0178, "step": 84860 }, { "grad_norm": 0.3846219480037689, "learning_rate": 6.129817112974218e-06, "loss": 0.02, "step": 84870 }, { "grad_norm": 0.3693431615829468, "learning_rate": 6.121886942914457e-06, "loss": 0.0204, "step": 84880 }, { "grad_norm": 0.4180566966533661, "learning_rate": 6.11396157129917e-06, "loss": 0.0183, "step": 84890 }, { "grad_norm": 0.24747976660728455, "learning_rate": 6.106040998995066e-06, "loss": 0.0181, "step": 84900 }, { "grad_norm": 0.30835282802581787, "learning_rate": 6.098125226868329e-06, "loss": 0.0227, "step": 84910 }, { "grad_norm": 0.16841524839401245, "learning_rate": 6.090214255784604e-06, "loss": 0.0132, "step": 84920 }, { "grad_norm": 0.5795215964317322, "learning_rate": 6.0823080866090346e-06, "loss": 0.0228, "step": 84930 }, { "grad_norm": 0.3783376216888428, "learning_rate": 6.074406720206233e-06, "loss": 0.0171, "step": 84940 }, { "grad_norm": 0.4315229654312134, "learning_rate": 6.066510157440259e-06, "loss": 0.0176, "step": 84950 }, { "grad_norm": 0.44364550709724426, "learning_rate": 6.058618399174676e-06, "loss": 0.0135, "step": 84960 }, { "grad_norm": 0.37565210461616516, "learning_rate": 6.050731446272528e-06, "loss": 0.0273, "step": 84970 }, { "grad_norm": 0.36214470863342285, "learning_rate": 6.042849299596321e-06, "loss": 0.0156, "step": 84980 }, { "grad_norm": 1.347301959991455, "learning_rate": 6.034971960008013e-06, "loss": 0.0231, "step": 84990 }, { "grad_norm": 0.26660165190696716, "learning_rate": 6.027099428369082e-06, "loss": 0.015, "step": 85000 }, { "grad_norm": 0.33176785707473755, "learning_rate": 6.019231705540435e-06, "loss": 0.0209, "step": 85010 }, { "grad_norm": 0.43906691670417786, "learning_rate": 6.011368792382488e-06, "loss": 0.0153, "step": 85020 }, { "grad_norm": 0.35091811418533325, "learning_rate": 6.00351068975511e-06, "loss": 0.0215, "step": 85030 }, { "grad_norm": 0.27279922366142273, "learning_rate": 5.9956573985176674e-06, "loss": 0.0176, "step": 85040 }, { "grad_norm": 0.5152726173400879, "learning_rate": 5.987808919528959e-06, "loss": 0.0178, "step": 85050 }, { "grad_norm": 0.38575541973114014, "learning_rate": 5.979965253647307e-06, "loss": 0.0189, "step": 85060 }, { "grad_norm": 0.4016040861606598, "learning_rate": 5.9721264017304655e-06, "loss": 0.0254, "step": 85070 }, { "grad_norm": 0.24336737394332886, "learning_rate": 5.964292364635682e-06, "loss": 0.0131, "step": 85080 }, { "grad_norm": 0.3686710000038147, "learning_rate": 5.956463143219682e-06, "loss": 0.0157, "step": 85090 }, { "grad_norm": 0.3157000243663788, "learning_rate": 5.948638738338663e-06, "loss": 0.0173, "step": 85100 }, { "grad_norm": 0.3465903401374817, "learning_rate": 5.940819150848271e-06, "loss": 0.017, "step": 85110 }, { "grad_norm": 0.1581512838602066, "learning_rate": 5.933004381603663e-06, "loss": 0.0166, "step": 85120 }, { "grad_norm": 0.35579344630241394, "learning_rate": 5.92519443145943e-06, "loss": 0.0132, "step": 85130 }, { "grad_norm": 0.3390752077102661, "learning_rate": 5.917389301269671e-06, "loss": 0.0155, "step": 85140 }, { "grad_norm": 0.19238321483135223, "learning_rate": 5.9095889918879435e-06, "loss": 0.0178, "step": 85150 }, { "grad_norm": 0.19690628349781036, "learning_rate": 5.901793504167286e-06, "loss": 0.021, "step": 85160 }, { "grad_norm": 0.494832307100296, "learning_rate": 5.894002838960172e-06, "loss": 0.0205, "step": 85170 }, { "grad_norm": 0.485470175743103, "learning_rate": 5.886216997118604e-06, "loss": 0.0309, "step": 85180 }, { "grad_norm": 0.17522716522216797, "learning_rate": 5.878435979494023e-06, "loss": 0.0277, "step": 85190 }, { "grad_norm": 0.37991827726364136, "learning_rate": 5.870659786937344e-06, "loss": 0.0158, "step": 85200 }, { "grad_norm": 0.2798062860965729, "learning_rate": 5.862888420298962e-06, "loss": 0.0154, "step": 85210 }, { "grad_norm": 0.10251456499099731, "learning_rate": 5.855121880428743e-06, "loss": 0.0166, "step": 85220 }, { "grad_norm": 0.3010888993740082, "learning_rate": 5.84736016817603e-06, "loss": 0.0333, "step": 85230 }, { "grad_norm": 0.27797114849090576, "learning_rate": 5.839603284389622e-06, "loss": 0.0193, "step": 85240 }, { "grad_norm": 0.22705501317977905, "learning_rate": 5.831851229917812e-06, "loss": 0.0153, "step": 85250 }, { "grad_norm": 0.3960966169834137, "learning_rate": 5.824104005608338e-06, "loss": 0.0249, "step": 85260 }, { "grad_norm": 0.645893394947052, "learning_rate": 5.81636161230843e-06, "loss": 0.0261, "step": 85270 }, { "grad_norm": 3.5279319286346436, "learning_rate": 5.808624050864791e-06, "loss": 0.0209, "step": 85280 }, { "grad_norm": 0.22936134040355682, "learning_rate": 5.800891322123592e-06, "loss": 0.0235, "step": 85290 }, { "grad_norm": 0.30406633019447327, "learning_rate": 5.793163426930454e-06, "loss": 0.0157, "step": 85300 }, { "grad_norm": 0.3162687122821808, "learning_rate": 5.785440366130512e-06, "loss": 0.0126, "step": 85310 }, { "grad_norm": 0.4146519601345062, "learning_rate": 5.777722140568326e-06, "loss": 0.014, "step": 85320 }, { "grad_norm": 0.30830949544906616, "learning_rate": 5.7700087510879685e-06, "loss": 0.0158, "step": 85330 }, { "grad_norm": 0.3812693953514099, "learning_rate": 5.762300198532933e-06, "loss": 0.0161, "step": 85340 }, { "grad_norm": 0.40162986516952515, "learning_rate": 5.7545964837462606e-06, "loss": 0.0173, "step": 85350 }, { "grad_norm": 0.46398812532424927, "learning_rate": 5.746897607570384e-06, "loss": 0.0129, "step": 85360 }, { "grad_norm": 0.34190425276756287, "learning_rate": 5.7392035708472626e-06, "loss": 0.031, "step": 85370 }, { "grad_norm": 0.4168718755245209, "learning_rate": 5.731514374418278e-06, "loss": 0.0167, "step": 85380 }, { "grad_norm": 0.3098355829715729, "learning_rate": 5.7238300191243325e-06, "loss": 0.0135, "step": 85390 }, { "grad_norm": 0.28849828243255615, "learning_rate": 5.716150505805762e-06, "loss": 0.0203, "step": 85400 }, { "grad_norm": 0.4117805063724518, "learning_rate": 5.708475835302407e-06, "loss": 0.0227, "step": 85410 }, { "grad_norm": 0.24023835361003876, "learning_rate": 5.70080600845353e-06, "loss": 0.0152, "step": 85420 }, { "grad_norm": 0.3258214294910431, "learning_rate": 5.693141026097909e-06, "loss": 0.0443, "step": 85430 }, { "grad_norm": 0.3990597724914551, "learning_rate": 5.6854808890737745e-06, "loss": 0.0217, "step": 85440 }, { "grad_norm": 0.8059405088424683, "learning_rate": 5.6778255982188225e-06, "loss": 0.018, "step": 85450 }, { "grad_norm": 0.3393852710723877, "learning_rate": 5.670175154370217e-06, "loss": 0.012, "step": 85460 }, { "grad_norm": 0.3604010343551636, "learning_rate": 5.662529558364616e-06, "loss": 0.0187, "step": 85470 }, { "grad_norm": 0.19868755340576172, "learning_rate": 5.654888811038128e-06, "loss": 0.0118, "step": 85480 }, { "grad_norm": 0.4901329278945923, "learning_rate": 5.647252913226314e-06, "loss": 0.0262, "step": 85490 }, { "grad_norm": 0.4053319990634918, "learning_rate": 5.639621865764255e-06, "loss": 0.0223, "step": 85500 }, { "grad_norm": 0.3915656805038452, "learning_rate": 5.63199566948644e-06, "loss": 0.023, "step": 85510 }, { "grad_norm": 0.3006805181503296, "learning_rate": 5.6243743252268835e-06, "loss": 0.0179, "step": 85520 }, { "grad_norm": 0.09665346890687943, "learning_rate": 5.616757833819009e-06, "loss": 0.0139, "step": 85530 }, { "grad_norm": 0.37478217482566833, "learning_rate": 5.609146196095794e-06, "loss": 0.0187, "step": 85540 }, { "grad_norm": 0.41250017285346985, "learning_rate": 5.6015394128896005e-06, "loss": 0.0191, "step": 85550 }, { "grad_norm": 0.1271449774503708, "learning_rate": 5.593937485032308e-06, "loss": 0.0167, "step": 85560 }, { "grad_norm": 0.5905020236968994, "learning_rate": 5.586340413355246e-06, "loss": 0.0214, "step": 85570 }, { "grad_norm": 0.4735807776451111, "learning_rate": 5.578748198689227e-06, "loss": 0.0171, "step": 85580 }, { "grad_norm": 0.42957305908203125, "learning_rate": 5.571160841864504e-06, "loss": 0.0143, "step": 85590 }, { "grad_norm": 0.5245488882064819, "learning_rate": 5.563578343710846e-06, "loss": 0.0165, "step": 85600 }, { "grad_norm": 0.35028716921806335, "learning_rate": 5.556000705057446e-06, "loss": 0.0154, "step": 85610 }, { "grad_norm": 0.1623571813106537, "learning_rate": 5.548427926732996e-06, "loss": 0.014, "step": 85620 }, { "grad_norm": 0.25520461797714233, "learning_rate": 5.540860009565629e-06, "loss": 0.0187, "step": 85630 }, { "grad_norm": 0.3924519717693329, "learning_rate": 5.533296954382966e-06, "loss": 0.0207, "step": 85640 }, { "grad_norm": 0.20459280908107758, "learning_rate": 5.525738762012095e-06, "loss": 0.0155, "step": 85650 }, { "grad_norm": 0.23464466631412506, "learning_rate": 5.518185433279577e-06, "loss": 0.0222, "step": 85660 }, { "grad_norm": 0.33501431345939636, "learning_rate": 5.510636969011407e-06, "loss": 0.0136, "step": 85670 }, { "grad_norm": 0.5720400214195251, "learning_rate": 5.503093370033096e-06, "loss": 0.0446, "step": 85680 }, { "grad_norm": 0.38134893774986267, "learning_rate": 5.495554637169603e-06, "loss": 0.0107, "step": 85690 }, { "grad_norm": 0.32319989800453186, "learning_rate": 5.488020771245328e-06, "loss": 0.0178, "step": 85700 }, { "grad_norm": 0.5663841366767883, "learning_rate": 5.480491773084195e-06, "loss": 0.0183, "step": 85710 }, { "grad_norm": 0.3424653112888336, "learning_rate": 5.472967643509525e-06, "loss": 0.0139, "step": 85720 }, { "grad_norm": 0.24847817420959473, "learning_rate": 5.46544838334419e-06, "loss": 0.0125, "step": 85730 }, { "grad_norm": 0.3917136788368225, "learning_rate": 5.4579339934104525e-06, "loss": 0.0146, "step": 85740 }, { "grad_norm": 0.27746447920799255, "learning_rate": 5.450424474530091e-06, "loss": 0.0188, "step": 85750 }, { "grad_norm": 0.21282902359962463, "learning_rate": 5.442919827524323e-06, "loss": 0.0127, "step": 85760 }, { "grad_norm": 0.18405769765377045, "learning_rate": 5.435420053213863e-06, "loss": 0.0112, "step": 85770 }, { "grad_norm": 0.5969107747077942, "learning_rate": 5.427925152418845e-06, "loss": 0.0237, "step": 85780 }, { "grad_norm": 0.3014260530471802, "learning_rate": 5.420435125958934e-06, "loss": 0.0165, "step": 85790 }, { "grad_norm": 0.5567922592163086, "learning_rate": 5.412949974653209e-06, "loss": 0.0191, "step": 85800 }, { "grad_norm": 0.3681000769138336, "learning_rate": 5.405469699320248e-06, "loss": 0.0158, "step": 85810 }, { "grad_norm": 0.11685485392808914, "learning_rate": 5.397994300778064e-06, "loss": 0.013, "step": 85820 }, { "grad_norm": 0.42477622628211975, "learning_rate": 5.39052377984417e-06, "loss": 0.0176, "step": 85830 }, { "grad_norm": 1.2432724237442017, "learning_rate": 5.383058137335523e-06, "loss": 0.021, "step": 85840 }, { "grad_norm": 0.28490734100341797, "learning_rate": 5.375597374068553e-06, "loss": 0.0374, "step": 85850 }, { "grad_norm": 0.36089780926704407, "learning_rate": 5.368141490859163e-06, "loss": 0.014, "step": 85860 }, { "grad_norm": 0.6311886310577393, "learning_rate": 5.360690488522729e-06, "loss": 0.0216, "step": 85870 }, { "grad_norm": 0.3184993863105774, "learning_rate": 5.353244367874055e-06, "loss": 0.0218, "step": 85880 }, { "grad_norm": 0.22511273622512817, "learning_rate": 5.345803129727456e-06, "loss": 0.0134, "step": 85890 }, { "grad_norm": 0.251819372177124, "learning_rate": 5.338366774896697e-06, "loss": 0.0201, "step": 85900 }, { "grad_norm": 0.3578317165374756, "learning_rate": 5.330935304194995e-06, "loss": 0.0217, "step": 85910 }, { "grad_norm": 0.3422183692455292, "learning_rate": 5.323508718435044e-06, "loss": 0.0218, "step": 85920 }, { "grad_norm": 0.13589410483837128, "learning_rate": 5.316087018429011e-06, "loss": 0.0138, "step": 85930 }, { "grad_norm": 0.48049068450927734, "learning_rate": 5.30867020498853e-06, "loss": 0.0196, "step": 85940 }, { "grad_norm": 0.4178045094013214, "learning_rate": 5.30125827892467e-06, "loss": 0.0121, "step": 85950 }, { "grad_norm": 0.19691261649131775, "learning_rate": 5.293851241048015e-06, "loss": 0.0156, "step": 85960 }, { "grad_norm": 0.8582878112792969, "learning_rate": 5.286449092168555e-06, "loss": 0.0172, "step": 85970 }, { "grad_norm": 0.16508345305919647, "learning_rate": 5.2790518330958105e-06, "loss": 0.0186, "step": 85980 }, { "grad_norm": 0.2124498337507248, "learning_rate": 5.271659464638717e-06, "loss": 0.014, "step": 85990 }, { "grad_norm": 0.41654258966445923, "learning_rate": 5.264271987605701e-06, "loss": 0.012, "step": 86000 }, { "grad_norm": 0.3836156725883484, "learning_rate": 5.256889402804632e-06, "loss": 0.0169, "step": 86010 }, { "grad_norm": 0.3321636915206909, "learning_rate": 5.249511711042881e-06, "loss": 0.016, "step": 86020 }, { "grad_norm": 0.2492244690656662, "learning_rate": 5.242138913127237e-06, "loss": 0.0212, "step": 86030 }, { "grad_norm": 0.2881062626838684, "learning_rate": 5.234771009863987e-06, "loss": 0.0173, "step": 86040 }, { "grad_norm": 0.5331088900566101, "learning_rate": 5.227408002058881e-06, "loss": 0.0181, "step": 86050 }, { "grad_norm": 0.38517680764198303, "learning_rate": 5.220049890517126e-06, "loss": 0.033, "step": 86060 }, { "grad_norm": 0.5755009651184082, "learning_rate": 5.212696676043377e-06, "loss": 0.0224, "step": 86070 }, { "grad_norm": 0.39078572392463684, "learning_rate": 5.205348359441798e-06, "loss": 0.017, "step": 86080 }, { "grad_norm": 0.17152278125286102, "learning_rate": 5.198004941515966e-06, "loss": 0.0156, "step": 86090 }, { "grad_norm": 0.4881516098976135, "learning_rate": 5.190666423068952e-06, "loss": 0.018, "step": 86100 }, { "grad_norm": 0.4601963758468628, "learning_rate": 5.183332804903284e-06, "loss": 0.0128, "step": 86110 }, { "grad_norm": 0.49918437004089355, "learning_rate": 5.176004087820979e-06, "loss": 0.0203, "step": 86120 }, { "grad_norm": 0.512901782989502, "learning_rate": 5.168680272623455e-06, "loss": 0.0194, "step": 86130 }, { "grad_norm": 0.39015403389930725, "learning_rate": 5.161361360111661e-06, "loss": 0.0214, "step": 86140 }, { "grad_norm": 0.25199365615844727, "learning_rate": 5.154047351085984e-06, "loss": 0.0161, "step": 86150 }, { "grad_norm": 0.20454321801662445, "learning_rate": 5.1467382463462385e-06, "loss": 0.0167, "step": 86160 }, { "grad_norm": 0.45506784319877625, "learning_rate": 5.139434046691788e-06, "loss": 0.0253, "step": 86170 }, { "grad_norm": 0.34684962034225464, "learning_rate": 5.13213475292137e-06, "loss": 0.0158, "step": 86180 }, { "grad_norm": 0.37378236651420593, "learning_rate": 5.12484036583325e-06, "loss": 0.0258, "step": 86190 }, { "grad_norm": 0.12265775352716446, "learning_rate": 5.117550886225109e-06, "loss": 0.0133, "step": 86200 }, { "grad_norm": 0.1205064058303833, "learning_rate": 5.11026631489413e-06, "loss": 0.0134, "step": 86210 }, { "grad_norm": 0.36507654190063477, "learning_rate": 5.102986652636926e-06, "loss": 0.0207, "step": 86220 }, { "grad_norm": 0.29003292322158813, "learning_rate": 5.095711900249606e-06, "loss": 0.0172, "step": 86230 }, { "grad_norm": 0.3533005118370056, "learning_rate": 5.088442058527715e-06, "loss": 0.016, "step": 86240 }, { "grad_norm": 0.4152103364467621, "learning_rate": 5.08117712826629e-06, "loss": 0.0189, "step": 86250 }, { "grad_norm": 0.7368568778038025, "learning_rate": 5.073917110259785e-06, "loss": 0.0169, "step": 86260 }, { "grad_norm": 0.4057728946208954, "learning_rate": 5.066662005302175e-06, "loss": 0.0249, "step": 86270 }, { "grad_norm": 0.2616691291332245, "learning_rate": 5.059411814186838e-06, "loss": 0.022, "step": 86280 }, { "grad_norm": 0.34615358710289, "learning_rate": 5.05216653770666e-06, "loss": 0.0164, "step": 86290 }, { "grad_norm": 0.4030688405036926, "learning_rate": 5.044926176653969e-06, "loss": 0.0191, "step": 86300 }, { "grad_norm": 0.2867886424064636, "learning_rate": 5.037690731820571e-06, "loss": 0.042, "step": 86310 }, { "grad_norm": 0.2792796492576599, "learning_rate": 5.030460203997711e-06, "loss": 0.0151, "step": 86320 }, { "grad_norm": 0.22194750607013702, "learning_rate": 5.0232345939761174e-06, "loss": 0.0238, "step": 86330 }, { "grad_norm": 0.23600764572620392, "learning_rate": 5.016013902545957e-06, "loss": 0.0205, "step": 86340 }, { "grad_norm": 0.25517070293426514, "learning_rate": 5.008798130496889e-06, "loss": 0.02, "step": 86350 }, { "grad_norm": 0.15475806593894958, "learning_rate": 5.001587278618008e-06, "loss": 0.0182, "step": 86360 }, { "grad_norm": 0.3129792809486389, "learning_rate": 4.994381347697891e-06, "loss": 0.0175, "step": 86370 }, { "grad_norm": 0.5165646076202393, "learning_rate": 4.987180338524572e-06, "loss": 0.0169, "step": 86380 }, { "grad_norm": 0.5402804017066956, "learning_rate": 4.9799842518855275e-06, "loss": 0.0188, "step": 86390 }, { "grad_norm": 0.25121793150901794, "learning_rate": 4.9727930885677285e-06, "loss": 0.0236, "step": 86400 }, { "grad_norm": 0.25663772225379944, "learning_rate": 4.965606849357573e-06, "loss": 0.0192, "step": 86410 }, { "grad_norm": 0.3320152163505554, "learning_rate": 4.9584255350409445e-06, "loss": 0.0179, "step": 86420 }, { "grad_norm": 0.2892659604549408, "learning_rate": 4.951249146403175e-06, "loss": 0.0151, "step": 86430 }, { "grad_norm": 0.3239547908306122, "learning_rate": 4.9440776842290815e-06, "loss": 0.0114, "step": 86440 }, { "grad_norm": 0.3003155589103699, "learning_rate": 4.936911149302903e-06, "loss": 0.0286, "step": 86450 }, { "grad_norm": 0.3096328377723694, "learning_rate": 4.929749542408379e-06, "loss": 0.0134, "step": 86460 }, { "grad_norm": 0.6659452319145203, "learning_rate": 4.922592864328673e-06, "loss": 0.0236, "step": 86470 }, { "grad_norm": 0.25048819184303284, "learning_rate": 4.915441115846436e-06, "loss": 0.0148, "step": 86480 }, { "grad_norm": 0.17986132204532623, "learning_rate": 4.908294297743782e-06, "loss": 0.0155, "step": 86490 }, { "grad_norm": 0.5361748933792114, "learning_rate": 4.901152410802273e-06, "loss": 0.0268, "step": 86500 }, { "grad_norm": 0.5528593063354492, "learning_rate": 4.894015455802919e-06, "loss": 0.0265, "step": 86510 }, { "grad_norm": 0.3305463194847107, "learning_rate": 4.8868834335262346e-06, "loss": 0.0233, "step": 86520 }, { "grad_norm": 0.319976806640625, "learning_rate": 4.87975634475214e-06, "loss": 0.0213, "step": 86530 }, { "grad_norm": 0.2772371172904968, "learning_rate": 4.8726341902600614e-06, "loss": 0.0179, "step": 86540 }, { "grad_norm": 0.5903798341751099, "learning_rate": 4.865516970828843e-06, "loss": 0.0191, "step": 86550 }, { "grad_norm": 0.29320842027664185, "learning_rate": 4.85840468723685e-06, "loss": 0.0172, "step": 86560 }, { "grad_norm": 0.3191828727722168, "learning_rate": 4.85129734026184e-06, "loss": 0.0175, "step": 86570 }, { "grad_norm": 0.25057440996170044, "learning_rate": 4.844194930681073e-06, "loss": 0.0206, "step": 86580 }, { "grad_norm": 0.33520880341529846, "learning_rate": 4.837097459271267e-06, "loss": 0.0233, "step": 86590 }, { "grad_norm": 0.19067078828811646, "learning_rate": 4.830004926808573e-06, "loss": 0.0164, "step": 86600 }, { "grad_norm": 0.47251951694488525, "learning_rate": 4.822917334068627e-06, "loss": 0.0219, "step": 86610 }, { "grad_norm": 0.39311477541923523, "learning_rate": 4.81583468182652e-06, "loss": 0.0181, "step": 86620 }, { "grad_norm": 0.4209930896759033, "learning_rate": 4.808756970856809e-06, "loss": 0.02, "step": 86630 }, { "grad_norm": 0.12716540694236755, "learning_rate": 4.801684201933476e-06, "loss": 0.0233, "step": 86640 }, { "grad_norm": 0.2539947032928467, "learning_rate": 4.794616375830019e-06, "loss": 0.022, "step": 86650 }, { "grad_norm": 0.4393406808376312, "learning_rate": 4.787553493319341e-06, "loss": 0.014, "step": 86660 }, { "grad_norm": 0.4825105667114258, "learning_rate": 4.780495555173836e-06, "loss": 0.0279, "step": 86670 }, { "grad_norm": 0.23523588478565216, "learning_rate": 4.7734425621653475e-06, "loss": 0.0225, "step": 86680 }, { "grad_norm": 0.2537190318107605, "learning_rate": 4.766394515065198e-06, "loss": 0.0138, "step": 86690 }, { "grad_norm": 0.31077709794044495, "learning_rate": 4.75935141464412e-06, "loss": 0.013, "step": 86700 }, { "grad_norm": 0.4219505190849304, "learning_rate": 4.75231326167237e-06, "loss": 0.0155, "step": 86710 }, { "grad_norm": 0.3222895562648773, "learning_rate": 4.745280056919599e-06, "loss": 0.0163, "step": 86720 }, { "grad_norm": 0.22972361743450165, "learning_rate": 4.738251801154975e-06, "loss": 0.0257, "step": 86730 }, { "grad_norm": 0.15176627039909363, "learning_rate": 4.731228495147061e-06, "loss": 0.013, "step": 86740 }, { "grad_norm": 0.15680310130119324, "learning_rate": 4.7242101396639595e-06, "loss": 0.0203, "step": 86750 }, { "grad_norm": 0.3515164256095886, "learning_rate": 4.717196735473151e-06, "loss": 0.0194, "step": 86760 }, { "grad_norm": 0.32259419560432434, "learning_rate": 4.710188283341643e-06, "loss": 0.0211, "step": 86770 }, { "grad_norm": 0.3498087525367737, "learning_rate": 4.703184784035835e-06, "loss": 0.0154, "step": 86780 }, { "grad_norm": 0.40030258893966675, "learning_rate": 4.696186238321642e-06, "loss": 0.0142, "step": 86790 }, { "grad_norm": 0.27644166350364685, "learning_rate": 4.689192646964408e-06, "loss": 0.014, "step": 86800 }, { "grad_norm": 0.36349931359291077, "learning_rate": 4.682204010728952e-06, "loss": 0.0117, "step": 86810 }, { "grad_norm": 0.35483118891716003, "learning_rate": 4.67522033037952e-06, "loss": 0.0165, "step": 86820 }, { "grad_norm": 0.35942649841308594, "learning_rate": 4.668241606679846e-06, "loss": 0.0214, "step": 86830 }, { "grad_norm": 0.3115552067756653, "learning_rate": 4.66126784039313e-06, "loss": 0.019, "step": 86840 }, { "grad_norm": 0.21609415113925934, "learning_rate": 4.6542990322819836e-06, "loss": 0.0172, "step": 86850 }, { "grad_norm": 0.25966888666152954, "learning_rate": 4.64733518310852e-06, "loss": 0.0231, "step": 86860 }, { "grad_norm": 0.4218379557132721, "learning_rate": 4.640376293634291e-06, "loss": 0.0273, "step": 86870 }, { "grad_norm": 0.3653388023376465, "learning_rate": 4.633422364620327e-06, "loss": 0.0141, "step": 86880 }, { "grad_norm": 0.3169845640659332, "learning_rate": 4.6264733968270745e-06, "loss": 0.0115, "step": 86890 }, { "grad_norm": 0.4000776708126068, "learning_rate": 4.619529391014482e-06, "loss": 0.0175, "step": 86900 }, { "grad_norm": 0.30557239055633545, "learning_rate": 4.61259034794192e-06, "loss": 0.0116, "step": 86910 }, { "grad_norm": 0.28612950444221497, "learning_rate": 4.605656268368247e-06, "loss": 0.0212, "step": 86920 }, { "grad_norm": 0.20893003046512604, "learning_rate": 4.598727153051735e-06, "loss": 0.0146, "step": 86930 }, { "grad_norm": 0.5528314709663391, "learning_rate": 4.591803002750178e-06, "loss": 0.0171, "step": 86940 }, { "grad_norm": 0.3378910422325134, "learning_rate": 4.58488381822077e-06, "loss": 0.0158, "step": 86950 }, { "grad_norm": 0.4166751503944397, "learning_rate": 4.5779696002201946e-06, "loss": 0.0146, "step": 86960 }, { "grad_norm": 0.2749593257904053, "learning_rate": 4.571060349504563e-06, "loss": 0.0213, "step": 86970 }, { "grad_norm": 0.2673676908016205, "learning_rate": 4.564156066829478e-06, "loss": 0.0099, "step": 86980 }, { "grad_norm": 0.618017852306366, "learning_rate": 4.557256752949957e-06, "loss": 0.0139, "step": 86990 }, { "grad_norm": 0.9600791335105896, "learning_rate": 4.550362408620534e-06, "loss": 0.018, "step": 87000 }, { "grad_norm": 0.1968054175376892, "learning_rate": 4.543473034595136e-06, "loss": 0.02, "step": 87010 }, { "grad_norm": 0.5394284725189209, "learning_rate": 4.536588631627192e-06, "loss": 0.0163, "step": 87020 }, { "grad_norm": 0.29717662930488586, "learning_rate": 4.5297092004695565e-06, "loss": 0.0224, "step": 87030 }, { "grad_norm": 0.3530254662036896, "learning_rate": 4.522834741874555e-06, "loss": 0.0175, "step": 87040 }, { "grad_norm": 0.5696738362312317, "learning_rate": 4.515965256593974e-06, "loss": 0.0143, "step": 87050 }, { "grad_norm": 0.4548576772212982, "learning_rate": 4.509100745379058e-06, "loss": 0.0224, "step": 87060 }, { "grad_norm": 0.4158739745616913, "learning_rate": 4.502241208980479e-06, "loss": 0.0241, "step": 87070 }, { "grad_norm": 0.19177410006523132, "learning_rate": 4.495386648148398e-06, "loss": 0.0122, "step": 87080 }, { "grad_norm": 0.3204243779182434, "learning_rate": 4.488537063632431e-06, "loss": 0.0178, "step": 87090 }, { "grad_norm": 0.30388227105140686, "learning_rate": 4.481692456181608e-06, "loss": 0.0235, "step": 87100 }, { "grad_norm": 0.1996486485004425, "learning_rate": 4.474852826544479e-06, "loss": 0.0188, "step": 87110 }, { "grad_norm": 0.12930229306221008, "learning_rate": 4.468018175468974e-06, "loss": 0.0187, "step": 87120 }, { "grad_norm": 0.3633947968482971, "learning_rate": 4.461188503702568e-06, "loss": 0.0183, "step": 87130 }, { "grad_norm": 0.47516414523124695, "learning_rate": 4.454363811992107e-06, "loss": 0.0286, "step": 87140 }, { "grad_norm": 0.6007207036018372, "learning_rate": 4.447544101083956e-06, "loss": 0.0202, "step": 87150 }, { "grad_norm": 0.282694548368454, "learning_rate": 4.440729371723884e-06, "loss": 0.0146, "step": 87160 }, { "grad_norm": 0.8882516622543335, "learning_rate": 4.433919624657162e-06, "loss": 0.0154, "step": 87170 }, { "grad_norm": 0.22909271717071533, "learning_rate": 4.427114860628462e-06, "loss": 0.0138, "step": 87180 }, { "grad_norm": 0.2787449359893799, "learning_rate": 4.420315080381982e-06, "loss": 0.0131, "step": 87190 }, { "grad_norm": 0.36933669447898865, "learning_rate": 4.4135202846613045e-06, "loss": 0.0123, "step": 87200 }, { "grad_norm": 0.20687562227249146, "learning_rate": 4.406730474209525e-06, "loss": 0.0226, "step": 87210 }, { "grad_norm": 0.3516415059566498, "learning_rate": 4.399945649769138e-06, "loss": 0.0144, "step": 87220 }, { "grad_norm": 0.4723491072654724, "learning_rate": 4.393165812082151e-06, "loss": 0.0146, "step": 87230 }, { "grad_norm": 0.3630905747413635, "learning_rate": 4.386390961889969e-06, "loss": 0.0294, "step": 87240 }, { "grad_norm": 0.14894706010818481, "learning_rate": 4.3796210999334895e-06, "loss": 0.0158, "step": 87250 }, { "grad_norm": 0.18984544277191162, "learning_rate": 4.372856226953065e-06, "loss": 0.0095, "step": 87260 }, { "grad_norm": 0.2931582033634186, "learning_rate": 4.366096343688486e-06, "loss": 0.0138, "step": 87270 }, { "grad_norm": 0.48527565598487854, "learning_rate": 4.359341450878995e-06, "loss": 0.0159, "step": 87280 }, { "grad_norm": 0.23401078581809998, "learning_rate": 4.352591549263302e-06, "loss": 0.0174, "step": 87290 }, { "grad_norm": 0.2728508412837982, "learning_rate": 4.345846639579582e-06, "loss": 0.0221, "step": 87300 }, { "grad_norm": 0.5554161071777344, "learning_rate": 4.339106722565417e-06, "loss": 0.016, "step": 87310 }, { "grad_norm": 0.5624527931213379, "learning_rate": 4.332371798957896e-06, "loss": 0.019, "step": 87320 }, { "grad_norm": 0.332448810338974, "learning_rate": 4.3256418694935295e-06, "loss": 0.0214, "step": 87330 }, { "grad_norm": 0.21166005730628967, "learning_rate": 4.3189169349083125e-06, "loss": 0.0174, "step": 87340 }, { "grad_norm": 0.26651090383529663, "learning_rate": 4.312196995937645e-06, "loss": 0.0118, "step": 87350 }, { "grad_norm": 0.6702367663383484, "learning_rate": 4.305482053316434e-06, "loss": 0.0136, "step": 87360 }, { "grad_norm": 0.3446512520313263, "learning_rate": 4.2987721077789925e-06, "loss": 0.0275, "step": 87370 }, { "grad_norm": 0.21699005365371704, "learning_rate": 4.2920671600591325e-06, "loss": 0.0112, "step": 87380 }, { "grad_norm": 0.26393336057662964, "learning_rate": 4.285367210890084e-06, "loss": 0.0161, "step": 87390 }, { "grad_norm": 0.5605329275131226, "learning_rate": 4.278672261004551e-06, "loss": 0.0167, "step": 87400 }, { "grad_norm": 0.25032034516334534, "learning_rate": 4.2719823111346704e-06, "loss": 0.0152, "step": 87410 }, { "grad_norm": 0.5120654106140137, "learning_rate": 4.265297362012061e-06, "loss": 0.0259, "step": 87420 }, { "grad_norm": 0.3664799928665161, "learning_rate": 4.258617414367766e-06, "loss": 0.02, "step": 87430 }, { "grad_norm": 0.24095609784126282, "learning_rate": 4.251942468932296e-06, "loss": 0.0193, "step": 87440 }, { "grad_norm": 0.32143282890319824, "learning_rate": 4.2452725264356175e-06, "loss": 0.0168, "step": 87450 }, { "grad_norm": 0.295401006937027, "learning_rate": 4.2386075876071575e-06, "loss": 0.0134, "step": 87460 }, { "grad_norm": 0.1513039469718933, "learning_rate": 4.231947653175755e-06, "loss": 0.0261, "step": 87470 }, { "grad_norm": 0.344079852104187, "learning_rate": 4.225292723869761e-06, "loss": 0.0156, "step": 87480 }, { "grad_norm": 0.5488795042037964, "learning_rate": 4.218642800416916e-06, "loss": 0.0213, "step": 87490 }, { "grad_norm": 0.2174888402223587, "learning_rate": 4.211997883544472e-06, "loss": 0.0142, "step": 87500 }, { "grad_norm": 0.260865718126297, "learning_rate": 4.205357973979096e-06, "loss": 0.0151, "step": 87510 }, { "grad_norm": 0.24950186908245087, "learning_rate": 4.198723072446925e-06, "loss": 0.015, "step": 87520 }, { "grad_norm": 0.29471415281295776, "learning_rate": 4.192093179673534e-06, "loss": 0.0173, "step": 87530 }, { "grad_norm": 0.46857303380966187, "learning_rate": 4.185468296383959e-06, "loss": 0.0245, "step": 87540 }, { "grad_norm": 0.4616527855396271, "learning_rate": 4.1788484233027005e-06, "loss": 0.0263, "step": 87550 }, { "grad_norm": 0.38101667165756226, "learning_rate": 4.172233561153677e-06, "loss": 0.018, "step": 87560 }, { "grad_norm": 0.2827847898006439, "learning_rate": 4.165623710660288e-06, "loss": 0.0129, "step": 87570 }, { "grad_norm": 0.4656788408756256, "learning_rate": 4.1590188725453835e-06, "loss": 0.0195, "step": 87580 }, { "grad_norm": 0.32297199964523315, "learning_rate": 4.152419047531259e-06, "loss": 0.0181, "step": 87590 }, { "grad_norm": 0.2748233377933502, "learning_rate": 4.145824236339651e-06, "loss": 0.0154, "step": 87600 }, { "grad_norm": 0.3137209117412567, "learning_rate": 4.1392344396917695e-06, "loss": 0.0146, "step": 87610 }, { "grad_norm": 0.36660927534103394, "learning_rate": 4.132649658308252e-06, "loss": 0.0169, "step": 87620 }, { "grad_norm": 0.25194934010505676, "learning_rate": 4.126069892909213e-06, "loss": 0.015, "step": 87630 }, { "grad_norm": 0.4766641855239868, "learning_rate": 4.119495144214191e-06, "loss": 0.0163, "step": 87640 }, { "grad_norm": 0.29082900285720825, "learning_rate": 4.112925412942214e-06, "loss": 0.0143, "step": 87650 }, { "grad_norm": 0.3423951268196106, "learning_rate": 4.106360699811712e-06, "loss": 0.0167, "step": 87660 }, { "grad_norm": 0.4003971517086029, "learning_rate": 4.099801005540616e-06, "loss": 0.0131, "step": 87670 }, { "grad_norm": 0.8859373331069946, "learning_rate": 4.093246330846256e-06, "loss": 0.0188, "step": 87680 }, { "grad_norm": 0.3776008188724518, "learning_rate": 4.086696676445467e-06, "loss": 0.0187, "step": 87690 }, { "grad_norm": 0.49146199226379395, "learning_rate": 4.0801520430544914e-06, "loss": 0.0221, "step": 87700 }, { "grad_norm": 0.2779536843299866, "learning_rate": 4.073612431389062e-06, "loss": 0.0154, "step": 87710 }, { "grad_norm": 0.4709286093711853, "learning_rate": 4.067077842164319e-06, "loss": 0.0152, "step": 87720 }, { "grad_norm": 0.2547358274459839, "learning_rate": 4.0605482760948945e-06, "loss": 0.0223, "step": 87730 }, { "grad_norm": 0.1579495668411255, "learning_rate": 4.054023733894829e-06, "loss": 0.0158, "step": 87740 }, { "grad_norm": 0.1755279302597046, "learning_rate": 4.047504216277653e-06, "loss": 0.0165, "step": 87750 }, { "grad_norm": 0.4120914340019226, "learning_rate": 4.040989723956323e-06, "loss": 0.0264, "step": 87760 }, { "grad_norm": 0.4291023910045624, "learning_rate": 4.03448025764327e-06, "loss": 0.0212, "step": 87770 }, { "grad_norm": 0.3951558470726013, "learning_rate": 4.027975818050339e-06, "loss": 0.015, "step": 87780 }, { "grad_norm": 0.9170981645584106, "learning_rate": 4.021476405888858e-06, "loss": 0.0217, "step": 87790 }, { "grad_norm": 0.296537846326828, "learning_rate": 4.01498202186959e-06, "loss": 0.0227, "step": 87800 }, { "grad_norm": 0.40555551648139954, "learning_rate": 4.008492666702751e-06, "loss": 0.0175, "step": 87810 }, { "grad_norm": 0.7667532563209534, "learning_rate": 4.002008341098001e-06, "loss": 0.0196, "step": 87820 }, { "grad_norm": 0.3305886387825012, "learning_rate": 3.995529045764462e-06, "loss": 0.013, "step": 87830 }, { "grad_norm": 0.11268824338912964, "learning_rate": 3.989054781410712e-06, "loss": 0.0201, "step": 87840 }, { "grad_norm": 0.2598702311515808, "learning_rate": 3.982585548744744e-06, "loss": 0.0172, "step": 87850 }, { "grad_norm": 0.48945164680480957, "learning_rate": 3.9761213484740435e-06, "loss": 0.022, "step": 87860 }, { "grad_norm": 0.2683058977127075, "learning_rate": 3.9696621813055055e-06, "loss": 0.0154, "step": 87870 }, { "grad_norm": 0.3676549792289734, "learning_rate": 3.9632080479455044e-06, "loss": 0.0146, "step": 87880 }, { "grad_norm": 0.19278451800346375, "learning_rate": 3.956758949099859e-06, "loss": 0.0268, "step": 87890 }, { "grad_norm": 0.2864964008331299, "learning_rate": 3.9503148854738385e-06, "loss": 0.0251, "step": 87900 }, { "grad_norm": 0.20765528082847595, "learning_rate": 3.943875857772139e-06, "loss": 0.0206, "step": 87910 }, { "grad_norm": 0.2719510793685913, "learning_rate": 3.9374418666989375e-06, "loss": 0.0189, "step": 87920 }, { "grad_norm": 0.33467841148376465, "learning_rate": 3.931012912957838e-06, "loss": 0.0208, "step": 87930 }, { "grad_norm": 0.42249608039855957, "learning_rate": 3.9245889972519045e-06, "loss": 0.0319, "step": 87940 }, { "grad_norm": 0.46752792596817017, "learning_rate": 3.9181701202836265e-06, "loss": 0.0186, "step": 87950 }, { "grad_norm": 0.7610180974006653, "learning_rate": 3.911756282755003e-06, "loss": 0.0225, "step": 87960 }, { "grad_norm": 3.991708993911743, "learning_rate": 3.905347485367411e-06, "loss": 0.0179, "step": 87970 }, { "grad_norm": 0.37611123919487, "learning_rate": 3.898943728821725e-06, "loss": 0.0367, "step": 87980 }, { "grad_norm": 0.32872337102890015, "learning_rate": 3.892545013818227e-06, "loss": 0.0177, "step": 87990 }, { "grad_norm": 0.407187283039093, "learning_rate": 3.886151341056687e-06, "loss": 0.0138, "step": 88000 }, { "grad_norm": 0.20248371362686157, "learning_rate": 3.8797627112363066e-06, "loss": 0.0119, "step": 88010 }, { "grad_norm": 0.22402435541152954, "learning_rate": 3.873379125055743e-06, "loss": 0.0152, "step": 88020 }, { "grad_norm": 0.6066024899482727, "learning_rate": 3.867000583213082e-06, "loss": 0.0171, "step": 88030 }, { "grad_norm": 0.3140544891357422, "learning_rate": 3.860627086405871e-06, "loss": 0.0163, "step": 88040 }, { "grad_norm": 0.46924611926078796, "learning_rate": 3.854258635331126e-06, "loss": 0.0142, "step": 88050 }, { "grad_norm": 0.5301024317741394, "learning_rate": 3.8478952306852665e-06, "loss": 0.0174, "step": 88060 }, { "grad_norm": 0.41346657276153564, "learning_rate": 3.841536873164198e-06, "loss": 0.0202, "step": 88070 }, { "grad_norm": 0.24871112406253815, "learning_rate": 3.8351835634632626e-06, "loss": 0.0203, "step": 88080 }, { "grad_norm": 0.46444299817085266, "learning_rate": 3.828835302277245e-06, "loss": 0.0186, "step": 88090 }, { "grad_norm": 0.337872177362442, "learning_rate": 3.822492090300378e-06, "loss": 0.0134, "step": 88100 }, { "grad_norm": 0.3696204125881195, "learning_rate": 3.8161539282263566e-06, "loss": 0.0178, "step": 88110 }, { "grad_norm": 0.2913484573364258, "learning_rate": 3.8098208167482995e-06, "loss": 0.0162, "step": 88120 }, { "grad_norm": 0.27326828241348267, "learning_rate": 3.803492756558796e-06, "loss": 0.0157, "step": 88130 }, { "grad_norm": 0.440631628036499, "learning_rate": 3.797169748349855e-06, "loss": 0.0252, "step": 88140 }, { "grad_norm": 0.42424312233924866, "learning_rate": 3.790851792812977e-06, "loss": 0.0142, "step": 88150 }, { "grad_norm": 0.31156668066978455, "learning_rate": 3.784538890639072e-06, "loss": 0.0214, "step": 88160 }, { "grad_norm": 0.3439320921897888, "learning_rate": 3.7782310425185153e-06, "loss": 0.0197, "step": 88170 }, { "grad_norm": 1.0746324062347412, "learning_rate": 3.77192824914111e-06, "loss": 0.0318, "step": 88180 }, { "grad_norm": 0.5345616340637207, "learning_rate": 3.7656305111961333e-06, "loss": 0.02, "step": 88190 }, { "grad_norm": 0.23425915837287903, "learning_rate": 3.7593378293722782e-06, "loss": 0.0235, "step": 88200 }, { "grad_norm": 0.28493744134902954, "learning_rate": 3.7530502043577333e-06, "loss": 0.0169, "step": 88210 }, { "grad_norm": 0.3370879590511322, "learning_rate": 3.746767636840076e-06, "loss": 0.0124, "step": 88220 }, { "grad_norm": 0.2498738318681717, "learning_rate": 3.74049012750638e-06, "loss": 0.0175, "step": 88230 }, { "grad_norm": 0.5196683406829834, "learning_rate": 3.734217677043128e-06, "loss": 0.018, "step": 88240 }, { "grad_norm": 0.3133505880832672, "learning_rate": 3.727950286136267e-06, "loss": 0.0191, "step": 88250 }, { "grad_norm": 0.31360140442848206, "learning_rate": 3.7216879554711925e-06, "loss": 0.0156, "step": 88260 }, { "grad_norm": 0.20318707823753357, "learning_rate": 3.715430685732757e-06, "loss": 0.0139, "step": 88270 }, { "grad_norm": 0.3924362063407898, "learning_rate": 3.7091784776052243e-06, "loss": 0.0169, "step": 88280 }, { "grad_norm": 0.15383268892765045, "learning_rate": 3.7029313317723356e-06, "loss": 0.017, "step": 88290 }, { "grad_norm": 0.2165347933769226, "learning_rate": 3.6966892489172787e-06, "loss": 0.0235, "step": 88300 }, { "grad_norm": 0.4944767653942108, "learning_rate": 3.6904522297226575e-06, "loss": 0.0234, "step": 88310 }, { "grad_norm": 0.21423646807670593, "learning_rate": 3.6842202748705647e-06, "loss": 0.0133, "step": 88320 }, { "grad_norm": 0.4359492361545563, "learning_rate": 3.677993385042494e-06, "loss": 0.0114, "step": 88330 }, { "grad_norm": 0.33306682109832764, "learning_rate": 3.67177156091944e-06, "loss": 0.0137, "step": 88340 }, { "grad_norm": 0.333678662776947, "learning_rate": 3.6655548031817866e-06, "loss": 0.0174, "step": 88350 }, { "grad_norm": 0.5557249784469604, "learning_rate": 3.6593431125094e-06, "loss": 0.0221, "step": 88360 }, { "grad_norm": 0.11798593401908875, "learning_rate": 3.6531364895815766e-06, "loss": 0.0243, "step": 88370 }, { "grad_norm": 0.2870202362537384, "learning_rate": 3.6469349350770677e-06, "loss": 0.028, "step": 88380 }, { "grad_norm": 0.27494680881500244, "learning_rate": 3.6407384496740525e-06, "loss": 0.017, "step": 88390 }, { "grad_norm": 1.1927742958068848, "learning_rate": 3.6345470340501954e-06, "loss": 0.0242, "step": 88400 }, { "grad_norm": 0.2941126525402069, "learning_rate": 3.6283606888825537e-06, "loss": 0.0159, "step": 88410 }, { "grad_norm": 0.34239983558654785, "learning_rate": 3.622179414847676e-06, "loss": 0.0222, "step": 88420 }, { "grad_norm": 0.1957802176475525, "learning_rate": 3.616003212621527e-06, "loss": 0.0135, "step": 88430 }, { "grad_norm": 1.1080235242843628, "learning_rate": 3.609832082879533e-06, "loss": 0.012, "step": 88440 }, { "grad_norm": 0.6819143891334534, "learning_rate": 3.6036660262965485e-06, "loss": 0.0143, "step": 88450 }, { "grad_norm": 0.3012922406196594, "learning_rate": 3.597505043546895e-06, "loss": 0.0131, "step": 88460 }, { "grad_norm": 0.3957688510417938, "learning_rate": 3.5913491353043227e-06, "loss": 0.0153, "step": 88470 }, { "grad_norm": 0.2965587377548218, "learning_rate": 3.5851983022420375e-06, "loss": 0.0139, "step": 88480 }, { "grad_norm": 0.3901008665561676, "learning_rate": 3.5790525450326797e-06, "loss": 0.0179, "step": 88490 }, { "grad_norm": 0.5113677978515625, "learning_rate": 3.572911864348344e-06, "loss": 0.0202, "step": 88500 }, { "grad_norm": 1.6937766075134277, "learning_rate": 3.5667762608605717e-06, "loss": 0.0274, "step": 88510 }, { "grad_norm": 0.2868695557117462, "learning_rate": 3.5606457352403255e-06, "loss": 0.0149, "step": 88520 }, { "grad_norm": 0.3227711617946625, "learning_rate": 3.554520288158042e-06, "loss": 0.0233, "step": 88530 }, { "grad_norm": 0.260126531124115, "learning_rate": 3.5483999202835903e-06, "loss": 0.0107, "step": 88540 }, { "grad_norm": 0.7553085684776306, "learning_rate": 3.5422846322862913e-06, "loss": 0.0288, "step": 88550 }, { "grad_norm": 0.2904132902622223, "learning_rate": 3.5361744248348925e-06, "loss": 0.0154, "step": 88560 }, { "grad_norm": 0.352894127368927, "learning_rate": 3.5300692985976047e-06, "loss": 0.0285, "step": 88570 }, { "grad_norm": 0.3535210192203522, "learning_rate": 3.52396925424206e-06, "loss": 0.0185, "step": 88580 }, { "grad_norm": 0.4931417405605316, "learning_rate": 3.5178742924353747e-06, "loss": 0.0179, "step": 88590 }, { "grad_norm": 0.31685659289360046, "learning_rate": 3.511784413844066e-06, "loss": 0.0137, "step": 88600 }, { "grad_norm": 0.4430544376373291, "learning_rate": 3.5056996191341287e-06, "loss": 0.0181, "step": 88610 }, { "grad_norm": 0.22314201295375824, "learning_rate": 3.4996199089709692e-06, "loss": 0.0153, "step": 88620 }, { "grad_norm": 0.19438552856445312, "learning_rate": 3.4935452840194725e-06, "loss": 0.0196, "step": 88630 }, { "grad_norm": 0.3007870614528656, "learning_rate": 3.48747574494393e-06, "loss": 0.0179, "step": 88640 }, { "grad_norm": 0.8055797219276428, "learning_rate": 3.4814112924081154e-06, "loss": 0.0182, "step": 88650 }, { "grad_norm": 0.20015151798725128, "learning_rate": 3.475351927075221e-06, "loss": 0.0109, "step": 88660 }, { "grad_norm": 0.31936684250831604, "learning_rate": 3.469297649607894e-06, "loss": 0.0279, "step": 88670 }, { "grad_norm": 0.7630925178527832, "learning_rate": 3.463248460668217e-06, "loss": 0.0195, "step": 88680 }, { "grad_norm": 0.28411924839019775, "learning_rate": 3.457204360917726e-06, "loss": 0.0147, "step": 88690 }, { "grad_norm": 0.3879885971546173, "learning_rate": 3.4511653510173827e-06, "loss": 0.0217, "step": 88700 }, { "grad_norm": 0.49619999527931213, "learning_rate": 3.4451314316276074e-06, "loss": 0.0194, "step": 88710 }, { "grad_norm": 0.38315069675445557, "learning_rate": 3.4391026034082674e-06, "loss": 0.0172, "step": 88720 }, { "grad_norm": 0.3801201581954956, "learning_rate": 3.433078867018674e-06, "loss": 0.0149, "step": 88730 }, { "grad_norm": 0.3249652087688446, "learning_rate": 3.427060223117551e-06, "loss": 0.0152, "step": 88740 }, { "grad_norm": 0.37010928988456726, "learning_rate": 3.4210466723631042e-06, "loss": 0.0282, "step": 88750 }, { "grad_norm": 0.21833491325378418, "learning_rate": 3.415038215412969e-06, "loss": 0.0145, "step": 88760 }, { "grad_norm": 0.4854271113872528, "learning_rate": 3.4090348529242088e-06, "loss": 0.0181, "step": 88770 }, { "grad_norm": 0.3893219828605652, "learning_rate": 3.4030365855533487e-06, "loss": 0.0265, "step": 88780 }, { "grad_norm": 0.3009646236896515, "learning_rate": 3.397043413956347e-06, "loss": 0.0204, "step": 88790 }, { "grad_norm": 0.20650525391101837, "learning_rate": 3.3910553387886237e-06, "loss": 0.0168, "step": 88800 }, { "grad_norm": 0.27040421962738037, "learning_rate": 3.3850723607049996e-06, "loss": 0.0294, "step": 88810 }, { "grad_norm": 0.596589982509613, "learning_rate": 3.3790944803597902e-06, "loss": 0.0193, "step": 88820 }, { "grad_norm": 0.3822047710418701, "learning_rate": 3.373121698406706e-06, "loss": 0.0151, "step": 88830 }, { "grad_norm": 0.21630533039569855, "learning_rate": 3.3671540154989302e-06, "loss": 0.0204, "step": 88840 }, { "grad_norm": 0.34171387553215027, "learning_rate": 3.361191432289079e-06, "loss": 0.0212, "step": 88850 }, { "grad_norm": 0.7549020648002625, "learning_rate": 3.3552339494292207e-06, "loss": 0.018, "step": 88860 }, { "grad_norm": 0.4274680018424988, "learning_rate": 3.349281567570839e-06, "loss": 0.0151, "step": 88870 }, { "grad_norm": 0.374386191368103, "learning_rate": 3.3433342873648965e-06, "loss": 0.0146, "step": 88880 }, { "grad_norm": 0.3452114164829254, "learning_rate": 3.337392109461762e-06, "loss": 0.0195, "step": 88890 }, { "grad_norm": 0.38955777883529663, "learning_rate": 3.331455034511266e-06, "loss": 0.0216, "step": 88900 }, { "grad_norm": 0.226644828915596, "learning_rate": 3.325523063162689e-06, "loss": 0.0164, "step": 88910 }, { "grad_norm": 0.6383726000785828, "learning_rate": 3.3195961960647393e-06, "loss": 0.0218, "step": 88920 }, { "grad_norm": 0.4639894664287567, "learning_rate": 3.31367443386556e-06, "loss": 0.016, "step": 88930 }, { "grad_norm": 0.3658353388309479, "learning_rate": 3.3077577772127556e-06, "loss": 0.0233, "step": 88940 }, { "grad_norm": 0.2812868654727936, "learning_rate": 3.301846226753358e-06, "loss": 0.0155, "step": 88950 }, { "grad_norm": 0.4029707908630371, "learning_rate": 3.2959397831338444e-06, "loss": 0.0183, "step": 88960 }, { "grad_norm": 0.6521267890930176, "learning_rate": 3.2900384470001432e-06, "loss": 0.0149, "step": 88970 }, { "grad_norm": 0.3209143280982971, "learning_rate": 3.28414221899761e-06, "loss": 0.0127, "step": 88980 }, { "grad_norm": 0.4073299467563629, "learning_rate": 3.278251099771046e-06, "loss": 0.0146, "step": 88990 }, { "grad_norm": 0.23314689099788666, "learning_rate": 3.2723650899646906e-06, "loss": 0.0179, "step": 89000 }, { "grad_norm": 0.2823001742362976, "learning_rate": 3.2664841902222466e-06, "loss": 0.0179, "step": 89010 }, { "grad_norm": 0.30641961097717285, "learning_rate": 3.2606084011868208e-06, "loss": 0.0159, "step": 89020 }, { "grad_norm": 0.22426338493824005, "learning_rate": 3.254737723500989e-06, "loss": 0.0202, "step": 89030 }, { "grad_norm": 0.13267143070697784, "learning_rate": 3.2488721578067595e-06, "loss": 0.0141, "step": 89040 }, { "grad_norm": 0.42486634850502014, "learning_rate": 3.243011704745591e-06, "loss": 0.0171, "step": 89050 }, { "grad_norm": 0.4064878821372986, "learning_rate": 3.237156364958355e-06, "loss": 0.0258, "step": 89060 }, { "grad_norm": 0.31349489092826843, "learning_rate": 3.231306139085394e-06, "loss": 0.0282, "step": 89070 }, { "grad_norm": 0.3098088204860687, "learning_rate": 3.2254610277664797e-06, "loss": 0.0196, "step": 89080 }, { "grad_norm": 0.34217241406440735, "learning_rate": 3.2196210316408183e-06, "loss": 0.0347, "step": 89090 }, { "grad_norm": 0.4452197253704071, "learning_rate": 3.2137861513470644e-06, "loss": 0.0165, "step": 89100 }, { "grad_norm": 0.15489856898784637, "learning_rate": 3.2079563875233252e-06, "loss": 0.0156, "step": 89110 }, { "grad_norm": 0.16014230251312256, "learning_rate": 3.2021317408071184e-06, "loss": 0.0142, "step": 89120 }, { "grad_norm": 0.3539826273918152, "learning_rate": 3.196312211835434e-06, "loss": 0.0177, "step": 89130 }, { "grad_norm": 0.29055455327033997, "learning_rate": 3.1904978012446686e-06, "loss": 0.0339, "step": 89140 }, { "grad_norm": 0.3128225803375244, "learning_rate": 3.184688509670691e-06, "loss": 0.0557, "step": 89150 }, { "grad_norm": 0.4089796543121338, "learning_rate": 3.178884337748783e-06, "loss": 0.0118, "step": 89160 }, { "grad_norm": 0.16513651609420776, "learning_rate": 3.173085286113697e-06, "loss": 0.011, "step": 89170 }, { "grad_norm": 0.2308167815208435, "learning_rate": 3.1672913553995986e-06, "loss": 0.0141, "step": 89180 }, { "grad_norm": 0.3855260908603668, "learning_rate": 3.1615025462401138e-06, "loss": 0.0165, "step": 89190 }, { "grad_norm": 0.3944316506385803, "learning_rate": 3.1557188592682808e-06, "loss": 0.0164, "step": 89200 }, { "grad_norm": 0.25708580017089844, "learning_rate": 3.1499402951166045e-06, "loss": 0.0152, "step": 89210 }, { "grad_norm": 0.5514475703239441, "learning_rate": 3.144166854417019e-06, "loss": 0.0213, "step": 89220 }, { "grad_norm": 0.43556180596351624, "learning_rate": 3.1383985378009073e-06, "loss": 0.0225, "step": 89230 }, { "grad_norm": 1.2519831657409668, "learning_rate": 3.1326353458990654e-06, "loss": 0.0174, "step": 89240 }, { "grad_norm": 0.5839671492576599, "learning_rate": 3.126877279341761e-06, "loss": 0.0307, "step": 89250 }, { "grad_norm": 0.5216960310935974, "learning_rate": 3.1211243387586964e-06, "loss": 0.0243, "step": 89260 }, { "grad_norm": 0.2074037343263626, "learning_rate": 3.115376524778979e-06, "loss": 0.0116, "step": 89270 }, { "grad_norm": 0.4467509984970093, "learning_rate": 3.109633838031195e-06, "loss": 0.0348, "step": 89280 }, { "grad_norm": 0.525485098361969, "learning_rate": 3.103896279143359e-06, "loss": 0.025, "step": 89290 }, { "grad_norm": 0.2015920877456665, "learning_rate": 3.098163848742924e-06, "loss": 0.0167, "step": 89300 }, { "grad_norm": 0.3349952697753906, "learning_rate": 3.0924365474567675e-06, "loss": 0.0113, "step": 89310 }, { "grad_norm": 0.34892508387565613, "learning_rate": 3.086714375911237e-06, "loss": 0.0172, "step": 89320 }, { "grad_norm": 0.1682289093732834, "learning_rate": 3.0809973347320775e-06, "loss": 0.0198, "step": 89330 }, { "grad_norm": 0.29076284170150757, "learning_rate": 3.075285424544516e-06, "loss": 0.0179, "step": 89340 }, { "grad_norm": 0.20149526000022888, "learning_rate": 3.069578645973176e-06, "loss": 0.0202, "step": 89350 }, { "grad_norm": 0.23324017226696014, "learning_rate": 3.0638769996421735e-06, "loss": 0.0182, "step": 89360 }, { "grad_norm": 0.3188660442829132, "learning_rate": 3.058180486175005e-06, "loss": 0.0162, "step": 89370 }, { "grad_norm": 0.5461264252662659, "learning_rate": 3.052489106194645e-06, "loss": 0.0124, "step": 89380 }, { "grad_norm": 0.47754207253456116, "learning_rate": 3.0468028603234946e-06, "loss": 0.0135, "step": 89390 }, { "grad_norm": 0.8267606496810913, "learning_rate": 3.04112174918339e-06, "loss": 0.0283, "step": 89400 }, { "grad_norm": 0.2500265836715698, "learning_rate": 3.0354457733955954e-06, "loss": 0.0109, "step": 89410 }, { "grad_norm": 0.5275958180427551, "learning_rate": 3.029774933580859e-06, "loss": 0.0226, "step": 89420 }, { "grad_norm": 0.13877297937870026, "learning_rate": 3.0241092303593065e-06, "loss": 0.0121, "step": 89430 }, { "grad_norm": 0.27054712176322937, "learning_rate": 3.018448664350554e-06, "loss": 0.0131, "step": 89440 }, { "grad_norm": 0.4636589586734772, "learning_rate": 3.0127932361736055e-06, "loss": 0.0127, "step": 89450 }, { "grad_norm": 0.3064288794994354, "learning_rate": 3.00714294644695e-06, "loss": 0.021, "step": 89460 }, { "grad_norm": 0.30531275272369385, "learning_rate": 3.0014977957884828e-06, "loss": 0.0332, "step": 89470 }, { "grad_norm": 0.2205638289451599, "learning_rate": 2.99585778481557e-06, "loss": 0.0164, "step": 89480 }, { "grad_norm": 0.41232016682624817, "learning_rate": 2.990222914144969e-06, "loss": 0.0158, "step": 89490 }, { "grad_norm": 0.25362494587898254, "learning_rate": 2.984593184392914e-06, "loss": 0.0337, "step": 89500 }, { "grad_norm": 0.5143895745277405, "learning_rate": 2.978968596175069e-06, "loss": 0.0213, "step": 89510 }, { "grad_norm": 0.4028792679309845, "learning_rate": 2.9733491501065137e-06, "loss": 0.019, "step": 89520 }, { "grad_norm": 0.2669559121131897, "learning_rate": 2.9677348468018005e-06, "loss": 0.0267, "step": 89530 }, { "grad_norm": 0.3459344804286957, "learning_rate": 2.9621256868748726e-06, "loss": 0.0175, "step": 89540 }, { "grad_norm": 0.30864396691322327, "learning_rate": 2.956521670939183e-06, "loss": 0.0214, "step": 89550 }, { "grad_norm": 0.40456971526145935, "learning_rate": 2.950922799607542e-06, "loss": 0.0181, "step": 89560 }, { "grad_norm": 0.19756706058979034, "learning_rate": 2.9453290734922535e-06, "loss": 0.0382, "step": 89570 }, { "grad_norm": 0.3841147720813751, "learning_rate": 2.939740493205029e-06, "loss": 0.0134, "step": 89580 }, { "grad_norm": 0.4544006884098053, "learning_rate": 2.9341570593570345e-06, "loss": 0.0155, "step": 89590 }, { "grad_norm": 0.23144195973873138, "learning_rate": 2.928578772558843e-06, "loss": 0.0129, "step": 89600 }, { "grad_norm": 0.14552152156829834, "learning_rate": 2.923005633420528e-06, "loss": 0.019, "step": 89610 }, { "grad_norm": 0.7538483142852783, "learning_rate": 2.9174376425515237e-06, "loss": 0.0287, "step": 89620 }, { "grad_norm": 0.14715763926506042, "learning_rate": 2.91187480056076e-06, "loss": 0.0161, "step": 89630 }, { "grad_norm": 0.29876643419265747, "learning_rate": 2.9063171080565664e-06, "loss": 0.0336, "step": 89640 }, { "grad_norm": 0.21813493967056274, "learning_rate": 2.900764565646741e-06, "loss": 0.0171, "step": 89650 }, { "grad_norm": 0.19176748394966125, "learning_rate": 2.895217173938475e-06, "loss": 0.0182, "step": 89660 }, { "grad_norm": 0.19680260121822357, "learning_rate": 2.8896749335384445e-06, "loss": 0.0149, "step": 89670 }, { "grad_norm": 0.22678819298744202, "learning_rate": 2.884137845052731e-06, "loss": 0.0171, "step": 89680 }, { "grad_norm": 0.2968055307865143, "learning_rate": 2.8786059090868734e-06, "loss": 0.0116, "step": 89690 }, { "grad_norm": 0.16150543093681335, "learning_rate": 2.87307912624582e-06, "loss": 0.0183, "step": 89700 }, { "grad_norm": 0.284757137298584, "learning_rate": 2.867557497133977e-06, "loss": 0.0121, "step": 89710 }, { "grad_norm": 0.4330591857433319, "learning_rate": 2.862041022355194e-06, "loss": 0.0191, "step": 89720 }, { "grad_norm": 0.5408307313919067, "learning_rate": 2.8565297025127226e-06, "loss": 0.0216, "step": 89730 }, { "grad_norm": 0.3283531069755554, "learning_rate": 2.8510235382092853e-06, "loss": 0.0197, "step": 89740 }, { "grad_norm": 0.25319042801856995, "learning_rate": 2.845522530047029e-06, "loss": 0.023, "step": 89750 }, { "grad_norm": 0.45791998505592346, "learning_rate": 2.8400266786275387e-06, "loss": 0.0153, "step": 89760 }, { "grad_norm": 0.5676982998847961, "learning_rate": 2.834535984551817e-06, "loss": 0.0205, "step": 89770 }, { "grad_norm": 0.35709771513938904, "learning_rate": 2.8290504484203397e-06, "loss": 0.0176, "step": 89780 }, { "grad_norm": 0.6144217848777771, "learning_rate": 2.8235700708329705e-06, "loss": 0.0186, "step": 89790 }, { "grad_norm": 0.27099812030792236, "learning_rate": 2.818094852389058e-06, "loss": 0.0226, "step": 89800 }, { "grad_norm": 0.30442988872528076, "learning_rate": 2.8126247936873516e-06, "loss": 0.0159, "step": 89810 }, { "grad_norm": 0.2133687138557434, "learning_rate": 2.8071598953260614e-06, "loss": 0.0246, "step": 89820 }, { "grad_norm": 0.3067578077316284, "learning_rate": 2.8017001579028035e-06, "loss": 0.0218, "step": 89830 }, { "grad_norm": 0.5535686612129211, "learning_rate": 2.7962455820146617e-06, "loss": 0.0306, "step": 89840 }, { "grad_norm": 0.16758239269256592, "learning_rate": 2.7907961682581253e-06, "loss": 0.018, "step": 89850 }, { "grad_norm": 0.3218153119087219, "learning_rate": 2.7853519172291453e-06, "loss": 0.0116, "step": 89860 }, { "grad_norm": 0.09574195742607117, "learning_rate": 2.7799128295230904e-06, "loss": 0.0202, "step": 89870 }, { "grad_norm": 0.3361106216907501, "learning_rate": 2.774478905734784e-06, "loss": 0.0194, "step": 89880 }, { "grad_norm": 0.5686173439025879, "learning_rate": 2.7690501464584516e-06, "loss": 0.023, "step": 89890 }, { "grad_norm": 0.21522822976112366, "learning_rate": 2.763626552287796e-06, "loss": 0.0126, "step": 89900 }, { "grad_norm": 0.15932489931583405, "learning_rate": 2.7582081238159083e-06, "loss": 0.0107, "step": 89910 }, { "grad_norm": 0.5153849124908447, "learning_rate": 2.7527948616353548e-06, "loss": 0.019, "step": 89920 }, { "grad_norm": 0.3031894266605377, "learning_rate": 2.7473867663381226e-06, "loss": 0.0141, "step": 89930 }, { "grad_norm": 0.2690145671367645, "learning_rate": 2.7419838385156383e-06, "loss": 0.0342, "step": 89940 }, { "grad_norm": 0.766780436038971, "learning_rate": 2.7365860787587407e-06, "loss": 0.0235, "step": 89950 }, { "grad_norm": 0.2589586079120636, "learning_rate": 2.7311934876577295e-06, "loss": 0.0174, "step": 89960 }, { "grad_norm": 0.38340508937835693, "learning_rate": 2.725806065802339e-06, "loss": 0.018, "step": 89970 }, { "grad_norm": 0.48031553626060486, "learning_rate": 2.720423813781714e-06, "loss": 0.0185, "step": 89980 }, { "grad_norm": 0.30522245168685913, "learning_rate": 2.715046732184462e-06, "loss": 0.0138, "step": 89990 }, { "grad_norm": 0.2780553102493286, "learning_rate": 2.7096748215986013e-06, "loss": 0.0275, "step": 90000 }, { "grad_norm": 0.46240171790122986, "learning_rate": 2.704308082611612e-06, "loss": 0.0205, "step": 90010 }, { "grad_norm": 0.3486839234828949, "learning_rate": 2.698946515810374e-06, "loss": 0.02, "step": 90020 }, { "grad_norm": 0.15045732259750366, "learning_rate": 2.6935901217812367e-06, "loss": 0.0237, "step": 90030 }, { "grad_norm": 0.4976900815963745, "learning_rate": 2.688238901109952e-06, "loss": 0.0225, "step": 90040 }, { "grad_norm": 0.5345525145530701, "learning_rate": 2.68289285438173e-06, "loss": 0.0182, "step": 90050 }, { "grad_norm": 0.26396316289901733, "learning_rate": 2.6775519821812033e-06, "loss": 0.0167, "step": 90060 }, { "grad_norm": 0.25266385078430176, "learning_rate": 2.672216285092449e-06, "loss": 0.0193, "step": 90070 }, { "grad_norm": 0.33810946345329285, "learning_rate": 2.666885763698962e-06, "loss": 0.019, "step": 90080 }, { "grad_norm": 0.3127090632915497, "learning_rate": 2.661560418583692e-06, "loss": 0.0137, "step": 90090 }, { "grad_norm": 0.2738665044307709, "learning_rate": 2.656240250328995e-06, "loss": 0.0143, "step": 90100 }, { "grad_norm": 0.3780265152454376, "learning_rate": 2.650925259516679e-06, "loss": 0.0214, "step": 90110 }, { "grad_norm": 0.38662394881248474, "learning_rate": 2.645615446727995e-06, "loss": 0.0214, "step": 90120 }, { "grad_norm": 0.21818657219409943, "learning_rate": 2.6403108125436115e-06, "loss": 0.0155, "step": 90130 }, { "grad_norm": 0.28957441449165344, "learning_rate": 2.6350113575436264e-06, "loss": 0.026, "step": 90140 }, { "grad_norm": 0.49269959330558777, "learning_rate": 2.629717082307598e-06, "loss": 0.0198, "step": 90150 }, { "grad_norm": 0.5573558807373047, "learning_rate": 2.6244279874144793e-06, "loss": 0.0118, "step": 90160 }, { "grad_norm": 0.21805617213249207, "learning_rate": 2.6191440734426862e-06, "loss": 0.0198, "step": 90170 }, { "grad_norm": 0.5140522718429565, "learning_rate": 2.613865340970062e-06, "loss": 0.0166, "step": 90180 }, { "grad_norm": 0.2975896894931793, "learning_rate": 2.608591790573889e-06, "loss": 0.0129, "step": 90190 }, { "grad_norm": 0.3795640170574188, "learning_rate": 2.603323422830861e-06, "loss": 0.0163, "step": 90200 }, { "grad_norm": 0.6074603199958801, "learning_rate": 2.5980602383171183e-06, "loss": 0.0165, "step": 90210 }, { "grad_norm": 0.18038158118724823, "learning_rate": 2.5928022376082496e-06, "loss": 0.0156, "step": 90220 }, { "grad_norm": 0.23304040729999542, "learning_rate": 2.5875494212792505e-06, "loss": 0.0197, "step": 90230 }, { "grad_norm": 0.17384320497512817, "learning_rate": 2.5823017899045565e-06, "loss": 0.0175, "step": 90240 }, { "grad_norm": 0.3583456873893738, "learning_rate": 2.577059344058047e-06, "loss": 0.0208, "step": 90250 }, { "grad_norm": 0.4326329231262207, "learning_rate": 2.5718220843130415e-06, "loss": 0.0231, "step": 90260 }, { "grad_norm": 0.3662385940551758, "learning_rate": 2.5665900112422535e-06, "loss": 0.0191, "step": 90270 }, { "grad_norm": 0.19126754999160767, "learning_rate": 2.561363125417876e-06, "loss": 0.0138, "step": 90280 }, { "grad_norm": 0.2608388066291809, "learning_rate": 2.556141427411496e-06, "loss": 0.0134, "step": 90290 }, { "grad_norm": 0.45722252130508423, "learning_rate": 2.550924917794156e-06, "loss": 0.0191, "step": 90300 }, { "grad_norm": 0.3189883232116699, "learning_rate": 2.545713597136329e-06, "loss": 0.0226, "step": 90310 }, { "grad_norm": 0.2708539366722107, "learning_rate": 2.5405074660079298e-06, "loss": 0.0302, "step": 90320 }, { "grad_norm": 0.1911844164133072, "learning_rate": 2.5353065249782647e-06, "loss": 0.0202, "step": 90330 }, { "grad_norm": 0.8838605284690857, "learning_rate": 2.530110774616129e-06, "loss": 0.0253, "step": 90340 }, { "grad_norm": 0.4688544273376465, "learning_rate": 2.524920215489701e-06, "loss": 0.0224, "step": 90350 }, { "grad_norm": 0.17700187861919403, "learning_rate": 2.519734848166627e-06, "loss": 0.013, "step": 90360 }, { "grad_norm": 0.705466628074646, "learning_rate": 2.514554673213948e-06, "loss": 0.0211, "step": 90370 }, { "grad_norm": 0.3202717900276184, "learning_rate": 2.5093796911981936e-06, "loss": 0.0153, "step": 90380 }, { "grad_norm": 0.3488284647464752, "learning_rate": 2.5042099026852672e-06, "loss": 0.0193, "step": 90390 }, { "grad_norm": 0.1953860968351364, "learning_rate": 2.499045308240544e-06, "loss": 0.0116, "step": 90400 }, { "grad_norm": 0.1353207528591156, "learning_rate": 2.4938859084288003e-06, "loss": 0.0115, "step": 90410 }, { "grad_norm": 0.22763583064079285, "learning_rate": 2.4887317038142734e-06, "loss": 0.0158, "step": 90420 }, { "grad_norm": 0.23421059548854828, "learning_rate": 2.4835826949606124e-06, "loss": 0.0182, "step": 90430 }, { "grad_norm": 0.20836499333381653, "learning_rate": 2.4784388824309167e-06, "loss": 0.0119, "step": 90440 }, { "grad_norm": 0.14078564941883087, "learning_rate": 2.473300266787687e-06, "loss": 0.0232, "step": 90450 }, { "grad_norm": 0.30534306168556213, "learning_rate": 2.46816684859289e-06, "loss": 0.021, "step": 90460 }, { "grad_norm": 0.23000292479991913, "learning_rate": 2.4630386284079043e-06, "loss": 0.0203, "step": 90470 }, { "grad_norm": 0.777263343334198, "learning_rate": 2.4579156067935427e-06, "loss": 0.0314, "step": 90480 }, { "grad_norm": 0.3448697626590729, "learning_rate": 2.452797784310046e-06, "loss": 0.0173, "step": 90490 }, { "grad_norm": 0.7124828696250916, "learning_rate": 2.447685161517105e-06, "loss": 0.022, "step": 90500 }, { "grad_norm": 0.29776546359062195, "learning_rate": 2.4425777389738224e-06, "loss": 0.0126, "step": 90510 }, { "grad_norm": 0.26607298851013184, "learning_rate": 2.43747551723873e-06, "loss": 0.0154, "step": 90520 }, { "grad_norm": 0.39173728227615356, "learning_rate": 2.4323784968698136e-06, "loss": 0.019, "step": 90530 }, { "grad_norm": 0.26493600010871887, "learning_rate": 2.4272866784244565e-06, "loss": 0.0185, "step": 90540 }, { "grad_norm": 0.26062631607055664, "learning_rate": 2.422200062459518e-06, "loss": 0.0186, "step": 90550 }, { "grad_norm": 0.24663816392421722, "learning_rate": 2.41711864953123e-06, "loss": 0.0188, "step": 90560 }, { "grad_norm": 0.3326071798801422, "learning_rate": 2.412042440195322e-06, "loss": 0.0222, "step": 90570 }, { "grad_norm": 0.12748809158802032, "learning_rate": 2.4069714350068993e-06, "loss": 0.0175, "step": 90580 }, { "grad_norm": 0.3371719717979431, "learning_rate": 2.4019056345205294e-06, "loss": 0.0295, "step": 90590 }, { "grad_norm": 0.526311457157135, "learning_rate": 2.396845039290191e-06, "loss": 0.0225, "step": 90600 }, { "grad_norm": 0.2007128745317459, "learning_rate": 2.3917896498693194e-06, "loss": 0.0153, "step": 90610 }, { "grad_norm": 0.3241325318813324, "learning_rate": 2.3867394668107335e-06, "loss": 0.0247, "step": 90620 }, { "grad_norm": 0.6240085363388062, "learning_rate": 2.3816944906667527e-06, "loss": 0.0177, "step": 90630 }, { "grad_norm": 0.1235136166214943, "learning_rate": 2.3766547219890634e-06, "loss": 0.0113, "step": 90640 }, { "grad_norm": 0.5452616214752197, "learning_rate": 2.371620161328819e-06, "loss": 0.0213, "step": 90650 }, { "grad_norm": 0.3766835927963257, "learning_rate": 2.366590809236574e-06, "loss": 0.0119, "step": 90660 }, { "grad_norm": 0.6965590715408325, "learning_rate": 2.3615666662623494e-06, "loss": 0.0214, "step": 90670 }, { "grad_norm": 0.3050747513771057, "learning_rate": 2.3565477329555664e-06, "loss": 0.0161, "step": 90680 }, { "grad_norm": 0.22423169016838074, "learning_rate": 2.3515340098650975e-06, "loss": 0.016, "step": 90690 }, { "grad_norm": 0.22260497510433197, "learning_rate": 2.3465254975392258e-06, "loss": 0.0173, "step": 90700 }, { "grad_norm": 0.6604501605033875, "learning_rate": 2.3415221965256805e-06, "loss": 0.0275, "step": 90710 }, { "grad_norm": 0.7603784203529358, "learning_rate": 2.3365241073716237e-06, "loss": 0.0225, "step": 90720 }, { "grad_norm": 0.4244334101676941, "learning_rate": 2.3315312306236183e-06, "loss": 0.0157, "step": 90730 }, { "grad_norm": 0.41485393047332764, "learning_rate": 2.326543566827699e-06, "loss": 0.0194, "step": 90740 }, { "grad_norm": 0.5801607370376587, "learning_rate": 2.3215611165292862e-06, "loss": 0.0167, "step": 90750 }, { "grad_norm": 0.21950940787792206, "learning_rate": 2.3165838802732765e-06, "loss": 0.0105, "step": 90760 }, { "grad_norm": 0.38747021555900574, "learning_rate": 2.311611858603957e-06, "loss": 0.0163, "step": 90770 }, { "grad_norm": 0.3413616120815277, "learning_rate": 2.3066450520650763e-06, "loss": 0.0175, "step": 90780 }, { "grad_norm": 0.22095395624637604, "learning_rate": 2.3016834611997773e-06, "loss": 0.0213, "step": 90790 }, { "grad_norm": 0.3717293441295624, "learning_rate": 2.296727086550665e-06, "loss": 0.0087, "step": 90800 }, { "grad_norm": 0.31752440333366394, "learning_rate": 2.2917759286597507e-06, "loss": 0.0172, "step": 90810 }, { "grad_norm": 0.21458832919597626, "learning_rate": 2.2868299880685006e-06, "loss": 0.0196, "step": 90820 }, { "grad_norm": 0.39799389243125916, "learning_rate": 2.2818892653177827e-06, "loss": 0.012, "step": 90830 }, { "grad_norm": 0.3028140366077423, "learning_rate": 2.2769537609479196e-06, "loss": 0.0159, "step": 90840 }, { "grad_norm": 0.20283864438533783, "learning_rate": 2.2720234754986357e-06, "loss": 0.0098, "step": 90850 }, { "grad_norm": 0.23877161741256714, "learning_rate": 2.2670984095091108e-06, "loss": 0.0162, "step": 90860 }, { "grad_norm": 0.3353394567966461, "learning_rate": 2.262178563517936e-06, "loss": 0.0172, "step": 90870 }, { "grad_norm": 0.3718346953392029, "learning_rate": 2.257263938063137e-06, "loss": 0.0165, "step": 90880 }, { "grad_norm": 0.2070276290178299, "learning_rate": 2.2523545336821715e-06, "loss": 0.0203, "step": 90890 }, { "grad_norm": 0.37867677211761475, "learning_rate": 2.247450350911939e-06, "loss": 0.0217, "step": 90900 }, { "grad_norm": 0.20223328471183777, "learning_rate": 2.2425513902887328e-06, "loss": 0.026, "step": 90910 }, { "grad_norm": 0.3158397674560547, "learning_rate": 2.2376576523483015e-06, "loss": 0.0206, "step": 90920 }, { "grad_norm": 0.3914634883403778, "learning_rate": 2.232769137625829e-06, "loss": 0.0176, "step": 90930 }, { "grad_norm": 0.2109338492155075, "learning_rate": 2.227885846655903e-06, "loss": 0.0204, "step": 90940 }, { "grad_norm": 0.6063592433929443, "learning_rate": 2.2230077799725523e-06, "loss": 0.0209, "step": 90950 }, { "grad_norm": 0.4301925003528595, "learning_rate": 2.2181349381092396e-06, "loss": 0.0167, "step": 90960 }, { "grad_norm": 1.177661418914795, "learning_rate": 2.2132673215988552e-06, "loss": 0.0236, "step": 90970 }, { "grad_norm": 0.3649384379386902, "learning_rate": 2.2084049309737065e-06, "loss": 0.0126, "step": 90980 }, { "grad_norm": 0.1460060328245163, "learning_rate": 2.2035477667655513e-06, "loss": 0.0117, "step": 90990 }, { "grad_norm": 0.31056976318359375, "learning_rate": 2.1986958295055315e-06, "loss": 0.019, "step": 91000 }, { "grad_norm": 0.1868407428264618, "learning_rate": 2.193849119724284e-06, "loss": 0.0196, "step": 91010 }, { "grad_norm": 0.4605613350868225, "learning_rate": 2.1890076379518175e-06, "loss": 0.0184, "step": 91020 }, { "grad_norm": 0.11772648245096207, "learning_rate": 2.1841713847175925e-06, "loss": 0.0194, "step": 91030 }, { "grad_norm": 0.35692098736763, "learning_rate": 2.1793403605504968e-06, "loss": 0.0146, "step": 91040 }, { "grad_norm": 0.16064400970935822, "learning_rate": 2.1745145659788413e-06, "loss": 0.0145, "step": 91050 }, { "grad_norm": 0.4491230845451355, "learning_rate": 2.1696940015303645e-06, "loss": 0.0157, "step": 91060 }, { "grad_norm": 0.2578727900981903, "learning_rate": 2.1648786677322397e-06, "loss": 0.0141, "step": 91070 }, { "grad_norm": 0.2605724334716797, "learning_rate": 2.1600685651110673e-06, "loss": 0.0124, "step": 91080 }, { "grad_norm": 0.25509607791900635, "learning_rate": 2.1552636941928717e-06, "loss": 0.0264, "step": 91090 }, { "grad_norm": 0.5429394245147705, "learning_rate": 2.1504640555031043e-06, "loss": 0.0169, "step": 91100 }, { "grad_norm": 0.19729101657867432, "learning_rate": 2.1456696495666506e-06, "loss": 0.0165, "step": 91110 }, { "grad_norm": 0.22583115100860596, "learning_rate": 2.1408804769078084e-06, "loss": 0.0155, "step": 91120 }, { "grad_norm": 0.1965501606464386, "learning_rate": 2.136096538050325e-06, "loss": 0.021, "step": 91130 }, { "grad_norm": 0.18351435661315918, "learning_rate": 2.13131783351736e-06, "loss": 0.016, "step": 91140 }, { "grad_norm": 0.48556575179100037, "learning_rate": 2.126544363831512e-06, "loss": 0.0138, "step": 91150 }, { "grad_norm": 0.26487910747528076, "learning_rate": 2.121776129514791e-06, "loss": 0.0205, "step": 91160 }, { "grad_norm": 0.14912083745002747, "learning_rate": 2.1170131310886523e-06, "loss": 0.0167, "step": 91170 }, { "grad_norm": 0.4410768449306488, "learning_rate": 2.112255369073968e-06, "loss": 0.0137, "step": 91180 }, { "grad_norm": 0.39901402592658997, "learning_rate": 2.1075028439910328e-06, "loss": 0.0114, "step": 91190 }, { "grad_norm": 0.1927812248468399, "learning_rate": 2.1027555563595813e-06, "loss": 0.0156, "step": 91200 }, { "grad_norm": 0.4082101881504059, "learning_rate": 2.0980135066987706e-06, "loss": 0.016, "step": 91210 }, { "grad_norm": 0.2628137767314911, "learning_rate": 2.093276695527191e-06, "loss": 0.0176, "step": 91220 }, { "grad_norm": 0.5885475277900696, "learning_rate": 2.0885451233628396e-06, "loss": 0.0254, "step": 91230 }, { "grad_norm": 0.6173306107521057, "learning_rate": 2.0838187907231633e-06, "loss": 0.0216, "step": 91240 }, { "grad_norm": 0.1216234564781189, "learning_rate": 2.0790976981250153e-06, "loss": 0.0169, "step": 91250 }, { "grad_norm": 0.2582767605781555, "learning_rate": 2.0743818460847e-06, "loss": 0.0224, "step": 91260 }, { "grad_norm": 0.29046568274497986, "learning_rate": 2.069671235117926e-06, "loss": 0.0177, "step": 91270 }, { "grad_norm": 0.32197844982147217, "learning_rate": 2.0649658657398484e-06, "loss": 0.0145, "step": 91280 }, { "grad_norm": 0.3323788344860077, "learning_rate": 2.060265738465034e-06, "loss": 0.0192, "step": 91290 }, { "grad_norm": 0.13240252435207367, "learning_rate": 2.055570853807487e-06, "loss": 0.0167, "step": 91300 }, { "grad_norm": 0.4150942862033844, "learning_rate": 2.0508812122806196e-06, "loss": 0.0185, "step": 91310 }, { "grad_norm": 0.46098655462265015, "learning_rate": 2.046196814397294e-06, "loss": 0.0209, "step": 91320 }, { "grad_norm": 0.3000038266181946, "learning_rate": 2.0415176606697896e-06, "loss": 0.0138, "step": 91330 }, { "grad_norm": 0.45291316509246826, "learning_rate": 2.0368437516098126e-06, "loss": 0.0183, "step": 91340 }, { "grad_norm": 0.32013019919395447, "learning_rate": 2.032175087728483e-06, "loss": 0.0218, "step": 91350 }, { "grad_norm": 0.3000510334968567, "learning_rate": 2.027511669536375e-06, "loss": 0.0284, "step": 91360 }, { "grad_norm": 0.33586302399635315, "learning_rate": 2.0228534975434643e-06, "loss": 0.012, "step": 91370 }, { "grad_norm": 0.19668391346931458, "learning_rate": 2.0182005722591548e-06, "loss": 0.0203, "step": 91380 }, { "grad_norm": 0.34650516510009766, "learning_rate": 2.0135528941922944e-06, "loss": 0.0188, "step": 91390 }, { "grad_norm": 0.20340876281261444, "learning_rate": 2.008910463851149e-06, "loss": 0.0184, "step": 91400 }, { "grad_norm": 0.36442914605140686, "learning_rate": 2.0042732817433963e-06, "loss": 0.017, "step": 91410 }, { "grad_norm": 0.8436615467071533, "learning_rate": 1.999641348376158e-06, "loss": 0.0188, "step": 91420 }, { "grad_norm": 0.26580485701560974, "learning_rate": 1.9950146642559787e-06, "loss": 0.0112, "step": 91430 }, { "grad_norm": 0.453117311000824, "learning_rate": 1.990393229888815e-06, "loss": 0.0236, "step": 91440 }, { "grad_norm": 0.28269001841545105, "learning_rate": 1.9857770457800684e-06, "loss": 0.0154, "step": 91450 }, { "grad_norm": 0.3028903305530548, "learning_rate": 1.9811661124345517e-06, "loss": 0.0175, "step": 91460 }, { "grad_norm": 0.44187915325164795, "learning_rate": 1.976560430356522e-06, "loss": 0.0164, "step": 91470 }, { "grad_norm": 0.1993170976638794, "learning_rate": 1.9719600000496383e-06, "loss": 0.0149, "step": 91480 }, { "grad_norm": 0.32965460419654846, "learning_rate": 1.967364822017004e-06, "loss": 0.0227, "step": 91490 }, { "grad_norm": 0.439411997795105, "learning_rate": 1.962774896761127e-06, "loss": 0.0217, "step": 91500 }, { "grad_norm": 0.7320567965507507, "learning_rate": 1.958190224783968e-06, "loss": 0.026, "step": 91510 }, { "grad_norm": 0.10618738830089569, "learning_rate": 1.9536108065868975e-06, "loss": 0.014, "step": 91520 }, { "grad_norm": 0.3235706686973572, "learning_rate": 1.949036642670715e-06, "loss": 0.0154, "step": 91530 }, { "grad_norm": 0.3170541226863861, "learning_rate": 1.9444677335356377e-06, "loss": 0.0127, "step": 91540 }, { "grad_norm": 0.4956020414829254, "learning_rate": 1.939904079681326e-06, "loss": 0.0188, "step": 91550 }, { "grad_norm": 0.35200420022010803, "learning_rate": 1.935345681606837e-06, "loss": 0.0157, "step": 91560 }, { "grad_norm": 0.3372592628002167, "learning_rate": 1.930792539810694e-06, "loss": 0.0225, "step": 91570 }, { "grad_norm": 0.5351267457008362, "learning_rate": 1.9262446547907877e-06, "loss": 0.0271, "step": 91580 }, { "grad_norm": 0.368552565574646, "learning_rate": 1.921702027044503e-06, "loss": 0.0158, "step": 91590 }, { "grad_norm": 0.3198544681072235, "learning_rate": 1.917164657068593e-06, "loss": 0.0162, "step": 91600 }, { "grad_norm": 0.28425005078315735, "learning_rate": 1.912632545359272e-06, "loss": 0.0178, "step": 91610 }, { "grad_norm": 0.6281026601791382, "learning_rate": 1.9081056924121488e-06, "loss": 0.0212, "step": 91620 }, { "grad_norm": 0.2047240436077118, "learning_rate": 1.9035840987222887e-06, "loss": 0.012, "step": 91630 }, { "grad_norm": 0.5546002388000488, "learning_rate": 1.8990677647841515e-06, "loss": 0.011, "step": 91640 }, { "grad_norm": 0.31646254658699036, "learning_rate": 1.8945566910916534e-06, "loss": 0.0272, "step": 91650 }, { "grad_norm": 0.20550385117530823, "learning_rate": 1.8900508781381054e-06, "loss": 0.018, "step": 91660 }, { "grad_norm": 0.39595526456832886, "learning_rate": 1.8855503264162633e-06, "loss": 0.0161, "step": 91670 }, { "grad_norm": 0.356353759765625, "learning_rate": 1.8810550364182998e-06, "loss": 0.0337, "step": 91680 }, { "grad_norm": 1.9841573238372803, "learning_rate": 1.8765650086358112e-06, "loss": 0.0179, "step": 91690 }, { "grad_norm": 0.6211792230606079, "learning_rate": 1.8720802435598151e-06, "loss": 0.0292, "step": 91700 }, { "grad_norm": 0.3484437167644501, "learning_rate": 1.8676007416807694e-06, "loss": 0.0181, "step": 91710 }, { "grad_norm": 0.515785276889801, "learning_rate": 1.8631265034885437e-06, "loss": 0.0212, "step": 91720 }, { "grad_norm": 0.4233548939228058, "learning_rate": 1.858657529472424e-06, "loss": 0.019, "step": 91730 }, { "grad_norm": 0.5128350853919983, "learning_rate": 1.8541938201211473e-06, "loss": 0.0138, "step": 91740 }, { "grad_norm": 0.2057700753211975, "learning_rate": 1.8497353759228398e-06, "loss": 0.0227, "step": 91750 }, { "grad_norm": 0.3830161690711975, "learning_rate": 1.8452821973650835e-06, "loss": 0.0136, "step": 91760 }, { "grad_norm": 0.45693421363830566, "learning_rate": 1.840834284934856e-06, "loss": 0.0371, "step": 91770 }, { "grad_norm": 0.5102992057800293, "learning_rate": 1.8363916391185954e-06, "loss": 0.0166, "step": 91780 }, { "grad_norm": 0.2794318199157715, "learning_rate": 1.8319542604021245e-06, "loss": 0.0218, "step": 91790 }, { "grad_norm": 0.2213670164346695, "learning_rate": 1.8275221492707273e-06, "loss": 0.0112, "step": 91800 }, { "grad_norm": 0.22657576203346252, "learning_rate": 1.8230953062090717e-06, "loss": 0.0175, "step": 91810 }, { "grad_norm": 0.5216246247291565, "learning_rate": 1.8186737317012926e-06, "loss": 0.0138, "step": 91820 }, { "grad_norm": 0.4152849316596985, "learning_rate": 1.8142574262308977e-06, "loss": 0.0157, "step": 91830 }, { "grad_norm": 0.2794351577758789, "learning_rate": 1.8098463902808727e-06, "loss": 0.0146, "step": 91840 }, { "grad_norm": 0.2679443955421448, "learning_rate": 1.805440624333593e-06, "loss": 0.0133, "step": 91850 }, { "grad_norm": 0.355825275182724, "learning_rate": 1.8010401288708734e-06, "loss": 0.0148, "step": 91860 }, { "grad_norm": 0.3068760931491852, "learning_rate": 1.7966449043739343e-06, "loss": 0.0231, "step": 91870 }, { "grad_norm": 0.6644296646118164, "learning_rate": 1.79225495132343e-06, "loss": 0.011, "step": 91880 }, { "grad_norm": 0.3680611848831177, "learning_rate": 1.7878702701994487e-06, "loss": 0.0184, "step": 91890 }, { "grad_norm": 0.4674619734287262, "learning_rate": 1.783490861481496e-06, "loss": 0.0317, "step": 91900 }, { "grad_norm": 0.2635815739631653, "learning_rate": 1.7791167256484832e-06, "loss": 0.0262, "step": 91910 }, { "grad_norm": 0.2984044551849365, "learning_rate": 1.774747863178766e-06, "loss": 0.0164, "step": 91920 }, { "grad_norm": 0.6541114449501038, "learning_rate": 1.7703842745501186e-06, "loss": 0.0193, "step": 91930 }, { "grad_norm": 0.5148703455924988, "learning_rate": 1.7660259602397366e-06, "loss": 0.0183, "step": 91940 }, { "grad_norm": 0.24054282903671265, "learning_rate": 1.7616729207242388e-06, "loss": 0.0158, "step": 91950 }, { "grad_norm": 0.29095134139060974, "learning_rate": 1.7573251564796556e-06, "loss": 0.0157, "step": 91960 }, { "grad_norm": 0.3052150011062622, "learning_rate": 1.7529826679814676e-06, "loss": 0.0286, "step": 91970 }, { "grad_norm": 0.27486786246299744, "learning_rate": 1.7486454557045562e-06, "loss": 0.0123, "step": 91980 }, { "grad_norm": 0.6752115488052368, "learning_rate": 1.7443135201232364e-06, "loss": 0.0167, "step": 91990 }, { "grad_norm": 0.19659960269927979, "learning_rate": 1.7399868617112347e-06, "loss": 0.0315, "step": 92000 }, { "grad_norm": 0.4746767282485962, "learning_rate": 1.7356654809417171e-06, "loss": 0.0148, "step": 92010 }, { "grad_norm": 0.24673108756542206, "learning_rate": 1.7313493782872448e-06, "loss": 0.0148, "step": 92020 }, { "grad_norm": 0.3244743347167969, "learning_rate": 1.7270385542198453e-06, "loss": 0.0106, "step": 92030 }, { "grad_norm": 0.2821040153503418, "learning_rate": 1.7227330092109307e-06, "loss": 0.016, "step": 92040 }, { "grad_norm": 0.2939043939113617, "learning_rate": 1.718432743731352e-06, "loss": 0.0154, "step": 92050 }, { "grad_norm": 0.4293147325515747, "learning_rate": 1.7141377582513718e-06, "loss": 0.0217, "step": 92060 }, { "grad_norm": 0.6767988204956055, "learning_rate": 1.709848053240698e-06, "loss": 0.0189, "step": 92070 }, { "grad_norm": 0.6245936751365662, "learning_rate": 1.7055636291684328e-06, "loss": 0.018, "step": 92080 }, { "grad_norm": 0.8396782875061035, "learning_rate": 1.7012844865031185e-06, "loss": 0.0145, "step": 92090 }, { "grad_norm": 0.3229235112667084, "learning_rate": 1.6970106257127138e-06, "loss": 0.0138, "step": 92100 }, { "grad_norm": 0.6615424156188965, "learning_rate": 1.6927420472646117e-06, "loss": 0.0192, "step": 92110 }, { "grad_norm": 0.271554559469223, "learning_rate": 1.6884787516256106e-06, "loss": 0.015, "step": 92120 }, { "grad_norm": 0.2125747799873352, "learning_rate": 1.6842207392619325e-06, "loss": 0.0143, "step": 92130 }, { "grad_norm": 0.1888517290353775, "learning_rate": 1.679968010639238e-06, "loss": 0.0321, "step": 92140 }, { "grad_norm": 0.19077926874160767, "learning_rate": 1.6757205662225829e-06, "loss": 0.0195, "step": 92150 }, { "grad_norm": 0.2233530431985855, "learning_rate": 1.6714784064764844e-06, "loss": 0.0121, "step": 92160 }, { "grad_norm": 0.45024538040161133, "learning_rate": 1.6672415318648383e-06, "loss": 0.019, "step": 92170 }, { "grad_norm": 0.3677493631839752, "learning_rate": 1.6630099428510016e-06, "loss": 0.0279, "step": 92180 }, { "grad_norm": 0.3394213616847992, "learning_rate": 1.6587836398977153e-06, "loss": 0.0148, "step": 92190 }, { "grad_norm": 0.273762047290802, "learning_rate": 1.6545626234671762e-06, "loss": 0.0225, "step": 92200 }, { "grad_norm": 0.11722040176391602, "learning_rate": 1.650346894020971e-06, "loss": 0.0144, "step": 92210 }, { "grad_norm": 0.2906894385814667, "learning_rate": 1.6461364520201472e-06, "loss": 0.0245, "step": 92220 }, { "grad_norm": 0.3780812919139862, "learning_rate": 1.6419312979251366e-06, "loss": 0.0165, "step": 92230 }, { "grad_norm": 0.25555646419525146, "learning_rate": 1.6377314321958214e-06, "loss": 0.0164, "step": 92240 }, { "grad_norm": 0.322925865650177, "learning_rate": 1.6335368552914787e-06, "loss": 0.0162, "step": 92250 }, { "grad_norm": 0.28303706645965576, "learning_rate": 1.6293475676708358e-06, "loss": 0.0169, "step": 92260 }, { "grad_norm": 0.2813328206539154, "learning_rate": 1.6251635697920097e-06, "loss": 0.0168, "step": 92270 }, { "grad_norm": 0.3846149742603302, "learning_rate": 1.6209848621125624e-06, "loss": 0.0153, "step": 92280 }, { "grad_norm": 0.28477632999420166, "learning_rate": 1.6168114450894778e-06, "loss": 0.0174, "step": 92290 }, { "grad_norm": 0.2137071043252945, "learning_rate": 1.612643319179158e-06, "loss": 0.0161, "step": 92300 }, { "grad_norm": 0.31620270013809204, "learning_rate": 1.6084804848374046e-06, "loss": 0.0133, "step": 92310 }, { "grad_norm": 0.264532208442688, "learning_rate": 1.6043229425194817e-06, "loss": 0.0152, "step": 92320 }, { "grad_norm": 0.3046804666519165, "learning_rate": 1.6001706926800308e-06, "loss": 0.0154, "step": 92330 }, { "grad_norm": 0.19318321347236633, "learning_rate": 1.5960237357731444e-06, "loss": 0.0206, "step": 92340 }, { "grad_norm": 0.2930539846420288, "learning_rate": 1.5918820722523264e-06, "loss": 0.0136, "step": 92350 }, { "grad_norm": 0.4020063579082489, "learning_rate": 1.587745702570509e-06, "loss": 0.0175, "step": 92360 }, { "grad_norm": 0.3749752342700958, "learning_rate": 1.5836146271800356e-06, "loss": 0.0165, "step": 92370 }, { "grad_norm": 0.47341758012771606, "learning_rate": 1.5794888465326728e-06, "loss": 0.0205, "step": 92380 }, { "grad_norm": 0.29917213320732117, "learning_rate": 1.5753683610796155e-06, "loss": 0.0186, "step": 92390 }, { "grad_norm": 0.20234982669353485, "learning_rate": 1.5712531712714528e-06, "loss": 0.0191, "step": 92400 }, { "grad_norm": 0.30628010630607605, "learning_rate": 1.5671432775582474e-06, "loss": 0.0166, "step": 92410 }, { "grad_norm": 0.34957754611968994, "learning_rate": 1.563038680389428e-06, "loss": 0.0162, "step": 92420 }, { "grad_norm": 0.35034406185150146, "learning_rate": 1.5589393802138808e-06, "loss": 0.0154, "step": 92430 }, { "grad_norm": 0.2811393439769745, "learning_rate": 1.554845377479891e-06, "loss": 0.0175, "step": 92440 }, { "grad_norm": 0.20250028371810913, "learning_rate": 1.550756672635173e-06, "loss": 0.0159, "step": 92450 }, { "grad_norm": 0.25362494587898254, "learning_rate": 1.5466732661268635e-06, "loss": 0.02, "step": 92460 }, { "grad_norm": 0.25522515177726746, "learning_rate": 1.5425951584015163e-06, "loss": 0.0183, "step": 92470 }, { "grad_norm": 0.11645526438951492, "learning_rate": 1.5385223499051026e-06, "loss": 0.0185, "step": 92480 }, { "grad_norm": 0.6565138101577759, "learning_rate": 1.5344548410830328e-06, "loss": 0.0185, "step": 92490 }, { "grad_norm": 0.4389561712741852, "learning_rate": 1.5303926323801121e-06, "loss": 0.0205, "step": 92500 }, { "grad_norm": 0.5072879195213318, "learning_rate": 1.5263357242405795e-06, "loss": 0.0254, "step": 92510 }, { "grad_norm": 0.1739325076341629, "learning_rate": 1.5222841171080915e-06, "loss": 0.0208, "step": 92520 }, { "grad_norm": 0.23863673210144043, "learning_rate": 1.5182378114257268e-06, "loss": 0.0133, "step": 92530 }, { "grad_norm": 0.22887492179870605, "learning_rate": 1.5141968076359869e-06, "loss": 0.0104, "step": 92540 }, { "grad_norm": 0.2687351703643799, "learning_rate": 1.5101611061807908e-06, "loss": 0.0263, "step": 92550 }, { "grad_norm": 0.17634250223636627, "learning_rate": 1.506130707501463e-06, "loss": 0.0173, "step": 92560 }, { "grad_norm": 0.3241625130176544, "learning_rate": 1.5021056120387734e-06, "loss": 0.0142, "step": 92570 }, { "grad_norm": 0.32096439599990845, "learning_rate": 1.4980858202329084e-06, "loss": 0.0226, "step": 92580 }, { "grad_norm": 0.30377328395843506, "learning_rate": 1.4940713325234445e-06, "loss": 0.0171, "step": 92590 }, { "grad_norm": 0.3290669918060303, "learning_rate": 1.4900621493494137e-06, "loss": 0.0214, "step": 92600 }, { "grad_norm": 0.4598779082298279, "learning_rate": 1.4860582711492544e-06, "loss": 0.014, "step": 92610 }, { "grad_norm": 0.3348885476589203, "learning_rate": 1.4820596983608325e-06, "loss": 0.0177, "step": 92620 }, { "grad_norm": 0.3226481080055237, "learning_rate": 1.4780664314214044e-06, "loss": 0.0169, "step": 92630 }, { "grad_norm": 0.2648168206214905, "learning_rate": 1.474078470767687e-06, "loss": 0.0117, "step": 92640 }, { "grad_norm": 0.33333906531333923, "learning_rate": 1.470095816835787e-06, "loss": 0.0132, "step": 92650 }, { "grad_norm": 0.6758955121040344, "learning_rate": 1.4661184700612395e-06, "loss": 0.0269, "step": 92660 }, { "grad_norm": 0.7269556522369385, "learning_rate": 1.4621464308790133e-06, "loss": 0.0174, "step": 92670 }, { "grad_norm": 0.9480869174003601, "learning_rate": 1.4581796997234775e-06, "loss": 0.0332, "step": 92680 }, { "grad_norm": 0.3212760090827942, "learning_rate": 1.4542182770284297e-06, "loss": 0.0133, "step": 92690 }, { "grad_norm": 0.6417223215103149, "learning_rate": 1.450262163227084e-06, "loss": 0.0176, "step": 92700 }, { "grad_norm": 0.27899235486984253, "learning_rate": 1.4463113587520671e-06, "loss": 0.02, "step": 92710 }, { "grad_norm": 0.23341703414916992, "learning_rate": 1.4423658640354443e-06, "loss": 0.0154, "step": 92720 }, { "grad_norm": 0.3627449572086334, "learning_rate": 1.438425679508687e-06, "loss": 0.0119, "step": 92730 }, { "grad_norm": 0.3933134078979492, "learning_rate": 1.4344908056026895e-06, "loss": 0.0136, "step": 92740 }, { "grad_norm": 0.1927458494901657, "learning_rate": 1.4305612427477521e-06, "loss": 0.0167, "step": 92750 }, { "grad_norm": 0.22173263132572174, "learning_rate": 1.42663699137362e-06, "loss": 0.0185, "step": 92760 }, { "grad_norm": 0.17515552043914795, "learning_rate": 1.4227180519094386e-06, "loss": 0.0173, "step": 92770 }, { "grad_norm": 0.41509950160980225, "learning_rate": 1.4188044247837706e-06, "loss": 0.0264, "step": 92780 }, { "grad_norm": 0.23646938800811768, "learning_rate": 1.4148961104246128e-06, "loss": 0.0154, "step": 92790 }, { "grad_norm": 0.24257633090019226, "learning_rate": 1.4109931092593731e-06, "loss": 0.012, "step": 92800 }, { "grad_norm": 0.15236635506153107, "learning_rate": 1.4070954217148658e-06, "loss": 0.0185, "step": 92810 }, { "grad_norm": 0.4367193281650543, "learning_rate": 1.4032030482173497e-06, "loss": 0.0142, "step": 92820 }, { "grad_norm": 0.27329689264297485, "learning_rate": 1.3993159891924901e-06, "loss": 0.0201, "step": 92830 }, { "grad_norm": 0.4615064859390259, "learning_rate": 1.3954342450653524e-06, "loss": 0.0169, "step": 92840 }, { "grad_norm": 0.16888104379177094, "learning_rate": 1.3915578162604582e-06, "loss": 0.0177, "step": 92850 }, { "grad_norm": 0.31526875495910645, "learning_rate": 1.3876867032017126e-06, "loss": 0.0125, "step": 92860 }, { "grad_norm": 1.0724655389785767, "learning_rate": 1.383820906312472e-06, "loss": 0.0251, "step": 92870 }, { "grad_norm": 0.23554709553718567, "learning_rate": 1.3799604260154752e-06, "loss": 0.0109, "step": 92880 }, { "grad_norm": 0.3960667550563812, "learning_rate": 1.376105262732913e-06, "loss": 0.0166, "step": 92890 }, { "grad_norm": 0.9065539240837097, "learning_rate": 1.37225541688637e-06, "loss": 0.0201, "step": 92900 }, { "grad_norm": 1.1636548042297363, "learning_rate": 1.3684108888968594e-06, "loss": 0.0295, "step": 92910 }, { "grad_norm": 0.6865300536155701, "learning_rate": 1.364571679184823e-06, "loss": 0.016, "step": 92920 }, { "grad_norm": 0.25763440132141113, "learning_rate": 1.3607377881701077e-06, "loss": 0.0199, "step": 92930 }, { "grad_norm": 0.1341230422258377, "learning_rate": 1.3569092162719787e-06, "loss": 0.02, "step": 92940 }, { "grad_norm": 0.4528488218784332, "learning_rate": 1.353085963909123e-06, "loss": 0.0224, "step": 92950 }, { "grad_norm": 0.23779112100601196, "learning_rate": 1.3492680314996454e-06, "loss": 0.0229, "step": 92960 }, { "grad_norm": 0.5324654579162598, "learning_rate": 1.345455419461078e-06, "loss": 0.0204, "step": 92970 }, { "grad_norm": 0.4233071506023407, "learning_rate": 1.3416481282103431e-06, "loss": 0.0112, "step": 92980 }, { "grad_norm": 0.17934969067573547, "learning_rate": 1.337846158163819e-06, "loss": 0.0136, "step": 92990 }, { "grad_norm": 0.20372314751148224, "learning_rate": 1.334049509737273e-06, "loss": 0.0165, "step": 93000 }, { "grad_norm": 0.36202695965766907, "learning_rate": 1.3302581833459116e-06, "loss": 0.0228, "step": 93010 }, { "grad_norm": 0.3767354488372803, "learning_rate": 1.3264721794043366e-06, "loss": 0.0185, "step": 93020 }, { "grad_norm": 0.3437046706676483, "learning_rate": 1.3226914983265836e-06, "loss": 0.0118, "step": 93030 }, { "grad_norm": 0.32087182998657227, "learning_rate": 1.3189161405260997e-06, "loss": 0.0152, "step": 93040 }, { "grad_norm": 0.3876629173755646, "learning_rate": 1.3151461064157656e-06, "loss": 0.0314, "step": 93050 }, { "grad_norm": 0.5325431227684021, "learning_rate": 1.3113813964078459e-06, "loss": 0.0332, "step": 93060 }, { "grad_norm": 0.3291172981262207, "learning_rate": 1.3076220109140614e-06, "loss": 0.0197, "step": 93070 }, { "grad_norm": 0.25843873620033264, "learning_rate": 1.3038679503455275e-06, "loss": 0.0195, "step": 93080 }, { "grad_norm": 0.6057460308074951, "learning_rate": 1.3001192151127773e-06, "loss": 0.0127, "step": 93090 }, { "grad_norm": 0.13632699847221375, "learning_rate": 1.2963758056257657e-06, "loss": 0.0153, "step": 93100 }, { "grad_norm": 0.605549693107605, "learning_rate": 1.2926377222938768e-06, "loss": 0.0113, "step": 93110 }, { "grad_norm": 0.9261333346366882, "learning_rate": 1.2889049655259e-06, "loss": 0.0199, "step": 93120 }, { "grad_norm": 0.31097307801246643, "learning_rate": 1.2851775357300312e-06, "loss": 0.0132, "step": 93130 }, { "grad_norm": 1.1778709888458252, "learning_rate": 1.281455433313916e-06, "loss": 0.0289, "step": 93140 }, { "grad_norm": 0.19593064486980438, "learning_rate": 1.2777386586845796e-06, "loss": 0.019, "step": 93150 }, { "grad_norm": 0.4913526475429535, "learning_rate": 1.2740272122484964e-06, "loss": 0.027, "step": 93160 }, { "grad_norm": 0.43142595887184143, "learning_rate": 1.2703210944115308e-06, "loss": 0.0123, "step": 93170 }, { "grad_norm": 0.2907671332359314, "learning_rate": 1.2666203055789916e-06, "loss": 0.0134, "step": 93180 }, { "grad_norm": 0.1960253119468689, "learning_rate": 1.2629248461555832e-06, "loss": 0.0171, "step": 93190 }, { "grad_norm": 0.2521071135997772, "learning_rate": 1.2592347165454431e-06, "loss": 0.0169, "step": 93200 }, { "grad_norm": 0.19835041463375092, "learning_rate": 1.2555499171521102e-06, "loss": 0.0146, "step": 93210 }, { "grad_norm": 0.2973982095718384, "learning_rate": 1.251870448378556e-06, "loss": 0.0208, "step": 93220 }, { "grad_norm": 0.12393404543399811, "learning_rate": 1.2481963106271478e-06, "loss": 0.0137, "step": 93230 }, { "grad_norm": 0.2242218255996704, "learning_rate": 1.2445275042997084e-06, "loss": 0.0154, "step": 93240 }, { "grad_norm": 0.2646588981151581, "learning_rate": 1.2408640297974283e-06, "loss": 0.0153, "step": 93250 }, { "grad_norm": 0.4327922463417053, "learning_rate": 1.2372058875209535e-06, "loss": 0.0104, "step": 93260 }, { "grad_norm": 0.642708420753479, "learning_rate": 1.2335530778703252e-06, "loss": 0.0152, "step": 93270 }, { "grad_norm": 0.3144482672214508, "learning_rate": 1.2299056012450182e-06, "loss": 0.0166, "step": 93280 }, { "grad_norm": 0.2084093540906906, "learning_rate": 1.2262634580439026e-06, "loss": 0.0176, "step": 93290 }, { "grad_norm": 0.47659674286842346, "learning_rate": 1.2226266486652926e-06, "loss": 0.0201, "step": 93300 }, { "grad_norm": 0.19025371968746185, "learning_rate": 1.2189951735068927e-06, "loss": 0.0137, "step": 93310 }, { "grad_norm": 0.2624365985393524, "learning_rate": 1.2153690329658407e-06, "loss": 0.0211, "step": 93320 }, { "grad_norm": 0.3407175838947296, "learning_rate": 1.2117482274386859e-06, "loss": 0.0166, "step": 93330 }, { "grad_norm": 0.22946159541606903, "learning_rate": 1.2081327573213953e-06, "loss": 0.0167, "step": 93340 }, { "grad_norm": 0.3278481662273407, "learning_rate": 1.2045226230093465e-06, "loss": 0.0171, "step": 93350 }, { "grad_norm": 0.3709166944026947, "learning_rate": 1.2009178248973296e-06, "loss": 0.0277, "step": 93360 }, { "grad_norm": 0.27996915578842163, "learning_rate": 1.1973183633795849e-06, "loss": 0.0145, "step": 93370 }, { "grad_norm": 0.2801191508769989, "learning_rate": 1.193724238849725e-06, "loss": 0.0145, "step": 93380 }, { "grad_norm": 0.25477567315101624, "learning_rate": 1.190135451700808e-06, "loss": 0.0159, "step": 93390 }, { "grad_norm": 0.41318845748901367, "learning_rate": 1.186552002325292e-06, "loss": 0.0118, "step": 93400 }, { "grad_norm": 0.24937684834003448, "learning_rate": 1.1829738911150578e-06, "loss": 0.0146, "step": 93410 }, { "grad_norm": 0.6266409158706665, "learning_rate": 1.1794011184613985e-06, "loss": 0.0168, "step": 93420 }, { "grad_norm": 0.4003139138221741, "learning_rate": 1.17583368475504e-06, "loss": 0.0194, "step": 93430 }, { "grad_norm": 0.3555789291858673, "learning_rate": 1.1722715903860982e-06, "loss": 0.0127, "step": 93440 }, { "grad_norm": 0.30981889367103577, "learning_rate": 1.168714835744128e-06, "loss": 0.016, "step": 93450 }, { "grad_norm": 0.22867916524410248, "learning_rate": 1.165163421218085e-06, "loss": 0.0283, "step": 93460 }, { "grad_norm": 0.2335449904203415, "learning_rate": 1.1616173471963477e-06, "loss": 0.0247, "step": 93470 }, { "grad_norm": 0.3700132369995117, "learning_rate": 1.1580766140667109e-06, "loss": 0.0106, "step": 93480 }, { "grad_norm": 0.15416482090950012, "learning_rate": 1.1545412222163764e-06, "loss": 0.0212, "step": 93490 }, { "grad_norm": 0.35220983624458313, "learning_rate": 1.1510111720319795e-06, "loss": 0.0143, "step": 93500 }, { "grad_norm": 0.2616611123085022, "learning_rate": 1.1474864638995663e-06, "loss": 0.0143, "step": 93510 }, { "grad_norm": 0.3649144768714905, "learning_rate": 1.1439670982045735e-06, "loss": 0.0231, "step": 93520 }, { "grad_norm": 0.7663136124610901, "learning_rate": 1.1404530753318877e-06, "loss": 0.021, "step": 93530 }, { "grad_norm": 0.3228204846382141, "learning_rate": 1.136944395665801e-06, "loss": 0.0194, "step": 93540 }, { "grad_norm": 0.8609336614608765, "learning_rate": 1.1334410595900014e-06, "loss": 0.0268, "step": 93550 }, { "grad_norm": 0.14048448204994202, "learning_rate": 1.129943067487621e-06, "loss": 0.0183, "step": 93560 }, { "grad_norm": 0.22688719630241394, "learning_rate": 1.126450419741193e-06, "loss": 0.0157, "step": 93570 }, { "grad_norm": 0.2946135699748993, "learning_rate": 1.1229631167326671e-06, "loss": 0.0219, "step": 93580 }, { "grad_norm": 0.3106890320777893, "learning_rate": 1.1194811588434052e-06, "loss": 0.015, "step": 93590 }, { "grad_norm": 0.32217419147491455, "learning_rate": 1.1160045464542023e-06, "loss": 0.0228, "step": 93600 }, { "grad_norm": 0.30260124802589417, "learning_rate": 1.1125332799452325e-06, "loss": 0.0402, "step": 93610 }, { "grad_norm": 0.541602611541748, "learning_rate": 1.109067359696131e-06, "loss": 0.021, "step": 93620 }, { "grad_norm": 0.4390869438648224, "learning_rate": 1.1056067860859166e-06, "loss": 0.0175, "step": 93630 }, { "grad_norm": 0.3426315486431122, "learning_rate": 1.102151559493031e-06, "loss": 0.017, "step": 93640 }, { "grad_norm": 0.2721237242221832, "learning_rate": 1.0987016802953332e-06, "loss": 0.0157, "step": 93650 }, { "grad_norm": 0.41368380188941956, "learning_rate": 1.0952571488701048e-06, "loss": 0.0136, "step": 93660 }, { "grad_norm": 0.8992475271224976, "learning_rate": 1.0918179655940163e-06, "loss": 0.0149, "step": 93670 }, { "grad_norm": 0.48131707310676575, "learning_rate": 1.088384130843184e-06, "loss": 0.0139, "step": 93680 }, { "grad_norm": 0.2814919352531433, "learning_rate": 1.0849556449931296e-06, "loss": 0.0144, "step": 93690 }, { "grad_norm": 0.39110708236694336, "learning_rate": 1.0815325084187866e-06, "loss": 0.0126, "step": 93700 }, { "grad_norm": 0.2350820153951645, "learning_rate": 1.0781147214944943e-06, "loss": 0.0089, "step": 93710 }, { "grad_norm": 0.4018249213695526, "learning_rate": 1.0747022845940313e-06, "loss": 0.0199, "step": 93720 }, { "grad_norm": 0.605703592300415, "learning_rate": 1.0712951980905605e-06, "loss": 0.0193, "step": 93730 }, { "grad_norm": 0.19011814892292023, "learning_rate": 1.0678934623566839e-06, "loss": 0.0166, "step": 93740 }, { "grad_norm": 0.5961885452270508, "learning_rate": 1.0644970777644091e-06, "loss": 0.0174, "step": 93750 }, { "grad_norm": 0.11276599019765854, "learning_rate": 1.0611060446851673e-06, "loss": 0.0099, "step": 93760 }, { "grad_norm": 0.32266926765441895, "learning_rate": 1.0577203634897836e-06, "loss": 0.0145, "step": 93770 }, { "grad_norm": 0.4315001368522644, "learning_rate": 1.0543400345485178e-06, "loss": 0.016, "step": 93780 }, { "grad_norm": 0.3418225049972534, "learning_rate": 1.0509650582310404e-06, "loss": 0.0133, "step": 93790 }, { "grad_norm": 0.3056154251098633, "learning_rate": 1.0475954349064287e-06, "loss": 0.0245, "step": 93800 }, { "grad_norm": 0.33758339285850525, "learning_rate": 1.0442311649431825e-06, "loss": 0.0215, "step": 93810 }, { "grad_norm": 0.20374177396297455, "learning_rate": 1.0408722487092071e-06, "loss": 0.0187, "step": 93820 }, { "grad_norm": 0.20336197316646576, "learning_rate": 1.0375186865718424e-06, "loss": 0.0212, "step": 93830 }, { "grad_norm": 0.11448408663272858, "learning_rate": 1.0341704788978113e-06, "loss": 0.0165, "step": 93840 }, { "grad_norm": 0.5973443984985352, "learning_rate": 1.0308276260532879e-06, "loss": 0.0201, "step": 93850 }, { "grad_norm": 0.3450334668159485, "learning_rate": 1.0274901284038241e-06, "loss": 0.0185, "step": 93860 }, { "grad_norm": 0.25988784432411194, "learning_rate": 1.0241579863144114e-06, "loss": 0.0155, "step": 93870 }, { "grad_norm": 0.3401760756969452, "learning_rate": 1.0208312001494413e-06, "loss": 0.0229, "step": 93880 }, { "grad_norm": 0.3327312469482422, "learning_rate": 1.0175097702727453e-06, "loss": 0.0221, "step": 93890 }, { "grad_norm": 0.19206462800502777, "learning_rate": 1.0141936970475275e-06, "loss": 0.0161, "step": 93900 }, { "grad_norm": 0.42252880334854126, "learning_rate": 1.0108829808364417e-06, "loss": 0.0118, "step": 93910 }, { "grad_norm": 0.8235872387886047, "learning_rate": 1.0075776220015376e-06, "loss": 0.0155, "step": 93920 }, { "grad_norm": 0.4167364835739136, "learning_rate": 1.0042776209042816e-06, "loss": 0.0183, "step": 93930 }, { "grad_norm": 0.31389811635017395, "learning_rate": 1.000982977905568e-06, "loss": 0.017, "step": 93940 }, { "grad_norm": 0.47404277324676514, "learning_rate": 9.976936933656867e-07, "loss": 0.0162, "step": 93950 }, { "grad_norm": 0.31900671124458313, "learning_rate": 9.944097676443442e-07, "loss": 0.0203, "step": 93960 }, { "grad_norm": 0.20005688071250916, "learning_rate": 9.911312011006757e-07, "loss": 0.0114, "step": 93970 }, { "grad_norm": 0.3850424885749817, "learning_rate": 9.878579940932109e-07, "loss": 0.0166, "step": 93980 }, { "grad_norm": 0.10402362793684006, "learning_rate": 9.84590146979908e-07, "loss": 0.0195, "step": 93990 }, { "grad_norm": 0.7004086971282959, "learning_rate": 9.813276601181365e-07, "loss": 0.0171, "step": 94000 }, { "grad_norm": 0.3295935094356537, "learning_rate": 9.780705338646724e-07, "loss": 0.0145, "step": 94010 }, { "grad_norm": 0.7222436666488647, "learning_rate": 9.748187685757083e-07, "loss": 0.0217, "step": 94020 }, { "grad_norm": 0.28762874007225037, "learning_rate": 9.715723646068543e-07, "loss": 0.0107, "step": 94030 }, { "grad_norm": 0.1347900927066803, "learning_rate": 9.68331322313143e-07, "loss": 0.024, "step": 94040 }, { "grad_norm": 0.266117125749588, "learning_rate": 9.650956420489909e-07, "loss": 0.0121, "step": 94050 }, { "grad_norm": 0.29257693886756897, "learning_rate": 9.618653241682595e-07, "loss": 0.0143, "step": 94060 }, { "grad_norm": 0.23893573880195618, "learning_rate": 9.586403690242053e-07, "loss": 0.0191, "step": 94070 }, { "grad_norm": 0.30285948514938354, "learning_rate": 9.554207769695122e-07, "loss": 0.0282, "step": 94080 }, { "grad_norm": 0.3327295482158661, "learning_rate": 9.522065483562659e-07, "loss": 0.0142, "step": 94090 }, { "grad_norm": 0.2726929783821106, "learning_rate": 9.489976835359681e-07, "loss": 0.0108, "step": 94100 }, { "grad_norm": 0.08255068957805634, "learning_rate": 9.457941828595329e-07, "loss": 0.0175, "step": 94110 }, { "grad_norm": 0.3484618067741394, "learning_rate": 9.425960466772965e-07, "loss": 0.0136, "step": 94120 }, { "grad_norm": 0.11367426812648773, "learning_rate": 9.394032753390014e-07, "loss": 0.0179, "step": 94130 }, { "grad_norm": 0.3005247116088867, "learning_rate": 9.362158691938017e-07, "loss": 0.0175, "step": 94140 }, { "grad_norm": 0.36076492071151733, "learning_rate": 9.330338285902685e-07, "loss": 0.0149, "step": 94150 }, { "grad_norm": 0.10403129458427429, "learning_rate": 9.298571538763845e-07, "loss": 0.0157, "step": 94160 }, { "grad_norm": 0.2694090008735657, "learning_rate": 9.266858453995441e-07, "loss": 0.0152, "step": 94170 }, { "grad_norm": 0.7920754551887512, "learning_rate": 9.235199035065645e-07, "loss": 0.0202, "step": 94180 }, { "grad_norm": 0.30064815282821655, "learning_rate": 9.203593285436574e-07, "loss": 0.0188, "step": 94190 }, { "grad_norm": 0.40561315417289734, "learning_rate": 9.172041208564686e-07, "loss": 0.0174, "step": 94200 }, { "grad_norm": 0.28851404786109924, "learning_rate": 9.140542807900332e-07, "loss": 0.016, "step": 94210 }, { "grad_norm": 0.25353458523750305, "learning_rate": 9.109098086888313e-07, "loss": 0.012, "step": 94220 }, { "grad_norm": 0.4126470983028412, "learning_rate": 9.077707048967265e-07, "loss": 0.0124, "step": 94230 }, { "grad_norm": 0.16671410202980042, "learning_rate": 9.046369697570057e-07, "loss": 0.0102, "step": 94240 }, { "grad_norm": 0.8332404494285583, "learning_rate": 9.015086036123721e-07, "loss": 0.019, "step": 94250 }, { "grad_norm": 0.2565310001373291, "learning_rate": 8.983856068049413e-07, "loss": 0.023, "step": 94260 }, { "grad_norm": 0.35040438175201416, "learning_rate": 8.952679796762398e-07, "loss": 0.0252, "step": 94270 }, { "grad_norm": 0.2790135145187378, "learning_rate": 8.921557225671951e-07, "loss": 0.0248, "step": 94280 }, { "grad_norm": 0.2609516680240631, "learning_rate": 8.890488358181792e-07, "loss": 0.0163, "step": 94290 }, { "grad_norm": 0.47742193937301636, "learning_rate": 8.859473197689372e-07, "loss": 0.0181, "step": 94300 }, { "grad_norm": 0.24962471425533295, "learning_rate": 8.828511747586588e-07, "loss": 0.0257, "step": 94310 }, { "grad_norm": 0.2681125998497009, "learning_rate": 8.797604011259287e-07, "loss": 0.0184, "step": 94320 }, { "grad_norm": 0.5836890339851379, "learning_rate": 8.766749992087487e-07, "loss": 0.0211, "step": 94330 }, { "grad_norm": 0.14489296078681946, "learning_rate": 8.735949693445378e-07, "loss": 0.0167, "step": 94340 }, { "grad_norm": 0.3056556284427643, "learning_rate": 8.705203118701211e-07, "loss": 0.0237, "step": 94350 }, { "grad_norm": 0.13792715966701508, "learning_rate": 8.674510271217295e-07, "loss": 0.0168, "step": 94360 }, { "grad_norm": 0.2511392831802368, "learning_rate": 8.643871154350336e-07, "loss": 0.0127, "step": 94370 }, { "grad_norm": 0.2236543595790863, "learning_rate": 8.613285771450818e-07, "loss": 0.0153, "step": 94380 }, { "grad_norm": 0.21912583708763123, "learning_rate": 8.582754125863623e-07, "loss": 0.0134, "step": 94390 }, { "grad_norm": 0.7526824474334717, "learning_rate": 8.55227622092758e-07, "loss": 0.0165, "step": 94400 }, { "grad_norm": 0.19087904691696167, "learning_rate": 8.521852059975743e-07, "loss": 0.0103, "step": 94410 }, { "grad_norm": 0.21562963724136353, "learning_rate": 8.491481646335231e-07, "loss": 0.014, "step": 94420 }, { "grad_norm": 0.33446815609931946, "learning_rate": 8.46116498332733e-07, "loss": 0.0182, "step": 94430 }, { "grad_norm": 0.3590986728668213, "learning_rate": 8.430902074267333e-07, "loss": 0.0145, "step": 94440 }, { "grad_norm": 0.9584903120994568, "learning_rate": 8.400692922464925e-07, "loss": 0.0162, "step": 94450 }, { "grad_norm": 0.4634028673171997, "learning_rate": 8.370537531223577e-07, "loss": 0.0224, "step": 94460 }, { "grad_norm": 0.46313977241516113, "learning_rate": 8.340435903841092e-07, "loss": 0.0129, "step": 94470 }, { "grad_norm": 0.3267430365085602, "learning_rate": 8.310388043609341e-07, "loss": 0.0107, "step": 94480 }, { "grad_norm": 0.2256767600774765, "learning_rate": 8.280393953814303e-07, "loss": 0.0103, "step": 94490 }, { "grad_norm": 0.2782229781150818, "learning_rate": 8.250453637736077e-07, "loss": 0.0217, "step": 94500 }, { "grad_norm": 0.35369396209716797, "learning_rate": 8.22056709864899e-07, "loss": 0.0204, "step": 94510 }, { "grad_norm": 0.9355019330978394, "learning_rate": 8.190734339821204e-07, "loss": 0.0314, "step": 94520 }, { "grad_norm": 0.8384156227111816, "learning_rate": 8.160955364515333e-07, "loss": 0.0149, "step": 94530 }, { "grad_norm": 0.2407824546098709, "learning_rate": 8.131230175987936e-07, "loss": 0.0193, "step": 94540 }, { "grad_norm": 0.6621560454368591, "learning_rate": 8.101558777489693e-07, "loss": 0.0191, "step": 94550 }, { "grad_norm": 0.35674846172332764, "learning_rate": 8.071941172265396e-07, "loss": 0.0107, "step": 94560 }, { "grad_norm": 0.38537442684173584, "learning_rate": 8.042377363553955e-07, "loss": 0.0171, "step": 94570 }, { "grad_norm": 0.3425426483154297, "learning_rate": 8.01286735458856e-07, "loss": 0.0222, "step": 94580 }, { "grad_norm": 0.25047793984413147, "learning_rate": 7.983411148596298e-07, "loss": 0.0182, "step": 94590 }, { "grad_norm": 0.29308056831359863, "learning_rate": 7.954008748798425e-07, "loss": 0.0131, "step": 94600 }, { "grad_norm": 0.684070348739624, "learning_rate": 7.924660158410369e-07, "loss": 0.0192, "step": 94610 }, { "grad_norm": 0.3149609863758087, "learning_rate": 7.895365380641728e-07, "loss": 0.0149, "step": 94620 }, { "grad_norm": 0.24151091277599335, "learning_rate": 7.866124418695941e-07, "loss": 0.0148, "step": 94630 }, { "grad_norm": 0.4291086494922638, "learning_rate": 7.83693727577095e-07, "loss": 0.0146, "step": 94640 }, { "grad_norm": 0.1992398202419281, "learning_rate": 7.807803955058535e-07, "loss": 0.0129, "step": 94650 }, { "grad_norm": 0.1530485302209854, "learning_rate": 7.778724459744702e-07, "loss": 0.0099, "step": 94660 }, { "grad_norm": 0.45399945974349976, "learning_rate": 7.749698793009464e-07, "loss": 0.0167, "step": 94670 }, { "grad_norm": 0.3538675308227539, "learning_rate": 7.720726958027169e-07, "loss": 0.0255, "step": 94680 }, { "grad_norm": 0.19199420511722565, "learning_rate": 7.691808957965951e-07, "loss": 0.0109, "step": 94690 }, { "grad_norm": 0.27204978466033936, "learning_rate": 7.662944795988336e-07, "loss": 0.0242, "step": 94700 }, { "grad_norm": 0.143480122089386, "learning_rate": 7.634134475250854e-07, "loss": 0.016, "step": 94710 }, { "grad_norm": 0.5976507067680359, "learning_rate": 7.605377998904262e-07, "loss": 0.0227, "step": 94720 }, { "grad_norm": 0.299196720123291, "learning_rate": 7.5766753700931e-07, "loss": 0.0159, "step": 94730 }, { "grad_norm": 0.2582564651966095, "learning_rate": 7.54802659195647e-07, "loss": 0.0378, "step": 94740 }, { "grad_norm": 0.2426293045282364, "learning_rate": 7.519431667627253e-07, "loss": 0.0153, "step": 94750 }, { "grad_norm": 0.2503434121608734, "learning_rate": 7.490890600232503e-07, "loss": 0.0199, "step": 94760 }, { "grad_norm": 0.2519861161708832, "learning_rate": 7.462403392893502e-07, "loss": 0.0188, "step": 94770 }, { "grad_norm": 0.2545951306819916, "learning_rate": 7.433970048725536e-07, "loss": 0.0145, "step": 94780 }, { "grad_norm": 0.4705195128917694, "learning_rate": 7.405590570838062e-07, "loss": 0.0244, "step": 94790 }, { "grad_norm": 0.19960229098796844, "learning_rate": 7.377264962334596e-07, "loss": 0.019, "step": 94800 }, { "grad_norm": 0.4839728772640228, "learning_rate": 7.348993226312773e-07, "loss": 0.0154, "step": 94810 }, { "grad_norm": 0.23326534032821655, "learning_rate": 7.32077536586434e-07, "loss": 0.0177, "step": 94820 }, { "grad_norm": 0.5241510272026062, "learning_rate": 7.292611384075221e-07, "loss": 0.0191, "step": 94830 }, { "grad_norm": 0.31481924653053284, "learning_rate": 7.264501284025338e-07, "loss": 0.0239, "step": 94840 }, { "grad_norm": 0.2672324776649475, "learning_rate": 7.236445068788789e-07, "loss": 0.016, "step": 94850 }, { "grad_norm": 0.11009605973958969, "learning_rate": 7.20844274143373e-07, "loss": 0.0151, "step": 94860 }, { "grad_norm": 0.3630830943584442, "learning_rate": 7.180494305022489e-07, "loss": 0.0231, "step": 94870 }, { "grad_norm": 0.27948999404907227, "learning_rate": 7.152599762611456e-07, "loss": 0.0186, "step": 94880 }, { "grad_norm": 0.495975136756897, "learning_rate": 7.124759117251079e-07, "loss": 0.0158, "step": 94890 }, { "grad_norm": 0.22870458662509918, "learning_rate": 7.096972371986033e-07, "loss": 0.0177, "step": 94900 }, { "grad_norm": 0.6133654117584229, "learning_rate": 7.069239529855054e-07, "loss": 0.0381, "step": 94910 }, { "grad_norm": 0.4054754972457886, "learning_rate": 7.041560593890939e-07, "loss": 0.0232, "step": 94920 }, { "grad_norm": 0.58090740442276, "learning_rate": 7.013935567120599e-07, "loss": 0.0186, "step": 94930 }, { "grad_norm": 0.13793611526489258, "learning_rate": 6.986364452565009e-07, "loss": 0.0191, "step": 94940 }, { "grad_norm": 0.43911829590797424, "learning_rate": 6.95884725323942e-07, "loss": 0.0118, "step": 94950 }, { "grad_norm": 0.26945480704307556, "learning_rate": 6.931383972152982e-07, "loss": 0.0129, "step": 94960 }, { "grad_norm": 0.42504677176475525, "learning_rate": 6.903974612309127e-07, "loss": 0.0106, "step": 94970 }, { "grad_norm": 1.73726487159729, "learning_rate": 6.876619176705179e-07, "loss": 0.0167, "step": 94980 }, { "grad_norm": 0.42049068212509155, "learning_rate": 6.8493176683328e-07, "loss": 0.0152, "step": 94990 }, { "grad_norm": 0.9301065802574158, "learning_rate": 6.822070090177657e-07, "loss": 0.0237, "step": 95000 }, { "grad_norm": 0.19941242039203644, "learning_rate": 6.794876445219367e-07, "loss": 0.0129, "step": 95010 }, { "grad_norm": 0.2712427079677582, "learning_rate": 6.767736736431885e-07, "loss": 0.0153, "step": 95020 }, { "grad_norm": 0.2714499831199646, "learning_rate": 6.740650966783113e-07, "loss": 0.0139, "step": 95030 }, { "grad_norm": 0.29479289054870605, "learning_rate": 6.713619139235239e-07, "loss": 0.0198, "step": 95040 }, { "grad_norm": 0.4529435932636261, "learning_rate": 6.686641256744286e-07, "loss": 0.0153, "step": 95050 }, { "grad_norm": 0.36034026741981506, "learning_rate": 6.659717322260562e-07, "loss": 0.0202, "step": 95060 }, { "grad_norm": 0.2618955969810486, "learning_rate": 6.632847338728377e-07, "loss": 0.0189, "step": 95070 }, { "grad_norm": 0.37598276138305664, "learning_rate": 6.60603130908627e-07, "loss": 0.0188, "step": 95080 }, { "grad_norm": 0.13705046474933624, "learning_rate": 6.579269236266783e-07, "loss": 0.0155, "step": 95090 }, { "grad_norm": 0.3283850848674774, "learning_rate": 6.552561123196632e-07, "loss": 0.017, "step": 95100 }, { "grad_norm": 0.41569042205810547, "learning_rate": 6.525906972796425e-07, "loss": 0.02, "step": 95110 }, { "grad_norm": 0.48694515228271484, "learning_rate": 6.499306787981108e-07, "loss": 0.0202, "step": 95120 }, { "grad_norm": 0.30243703722953796, "learning_rate": 6.472760571659631e-07, "loss": 0.0136, "step": 95130 }, { "grad_norm": 0.23369258642196655, "learning_rate": 6.446268326735061e-07, "loss": 0.0119, "step": 95140 }, { "grad_norm": 0.3319927752017975, "learning_rate": 6.41983005610458e-07, "loss": 0.0198, "step": 95150 }, { "grad_norm": 0.4099826514720917, "learning_rate": 6.393445762659378e-07, "loss": 0.018, "step": 95160 }, { "grad_norm": 0.5360973477363586, "learning_rate": 6.36711544928481e-07, "loss": 0.0143, "step": 95170 }, { "grad_norm": 0.6264635324478149, "learning_rate": 6.340839118860298e-07, "loss": 0.0156, "step": 95180 }, { "grad_norm": 0.3477305471897125, "learning_rate": 6.314616774259485e-07, "loss": 0.0165, "step": 95190 }, { "grad_norm": 0.3671077787876129, "learning_rate": 6.288448418349858e-07, "loss": 0.04, "step": 95200 }, { "grad_norm": 0.7534148097038269, "learning_rate": 6.262334053993235e-07, "loss": 0.0124, "step": 95210 }, { "grad_norm": 0.1910049319267273, "learning_rate": 6.236273684045501e-07, "loss": 0.0174, "step": 95220 }, { "grad_norm": 0.23774075508117676, "learning_rate": 6.210267311356488e-07, "loss": 0.0114, "step": 95230 }, { "grad_norm": 0.5930437445640564, "learning_rate": 6.184314938770197e-07, "loss": 0.0254, "step": 95240 }, { "grad_norm": 0.3842957317829132, "learning_rate": 6.158416569124858e-07, "loss": 0.0281, "step": 95250 }, { "grad_norm": 0.43970581889152527, "learning_rate": 6.132572205252541e-07, "loss": 0.0169, "step": 95260 }, { "grad_norm": 0.5334346890449524, "learning_rate": 6.106781849979648e-07, "loss": 0.0206, "step": 95270 }, { "grad_norm": 0.08384718745946884, "learning_rate": 6.081045506126482e-07, "loss": 0.015, "step": 95280 }, { "grad_norm": 0.3141170144081116, "learning_rate": 6.055363176507678e-07, "loss": 0.0126, "step": 95290 }, { "grad_norm": 0.29569223523139954, "learning_rate": 6.029734863931602e-07, "loss": 0.0152, "step": 95300 }, { "grad_norm": 0.593087375164032, "learning_rate": 6.004160571201179e-07, "loss": 0.0124, "step": 95310 }, { "grad_norm": 0.2958417534828186, "learning_rate": 5.978640301112947e-07, "loss": 0.0129, "step": 95320 }, { "grad_norm": 0.378817617893219, "learning_rate": 5.9531740564579e-07, "loss": 0.0195, "step": 95330 }, { "grad_norm": 0.2873236835002899, "learning_rate": 5.927761840020918e-07, "loss": 0.0127, "step": 95340 }, { "grad_norm": 0.1884264349937439, "learning_rate": 5.902403654581113e-07, "loss": 0.0239, "step": 95350 }, { "grad_norm": 0.30747437477111816, "learning_rate": 5.877099502911542e-07, "loss": 0.0187, "step": 95360 }, { "grad_norm": 0.42259082198143005, "learning_rate": 5.851849387779496e-07, "loss": 0.013, "step": 95370 }, { "grad_norm": 0.22732406854629517, "learning_rate": 5.826653311946262e-07, "loss": 0.0144, "step": 95380 }, { "grad_norm": 0.3047293722629547, "learning_rate": 5.801511278167249e-07, "loss": 0.0189, "step": 95390 }, { "grad_norm": 0.12885834276676178, "learning_rate": 5.776423289191868e-07, "loss": 0.0178, "step": 95400 }, { "grad_norm": 0.26211732625961304, "learning_rate": 5.751389347763869e-07, "loss": 0.0154, "step": 95410 }, { "grad_norm": 0.40673893690109253, "learning_rate": 5.726409456620841e-07, "loss": 0.0252, "step": 95420 }, { "grad_norm": 0.7356039881706238, "learning_rate": 5.701483618494541e-07, "loss": 0.0252, "step": 95430 }, { "grad_norm": 0.3675595223903656, "learning_rate": 5.67661183611079e-07, "loss": 0.0137, "step": 95440 }, { "grad_norm": 0.24513676762580872, "learning_rate": 5.651794112189579e-07, "loss": 0.018, "step": 95450 }, { "grad_norm": 0.3430768549442291, "learning_rate": 5.627030449444903e-07, "loss": 0.0154, "step": 95460 }, { "grad_norm": 0.4430689811706543, "learning_rate": 5.60232085058493e-07, "loss": 0.0315, "step": 95470 }, { "grad_norm": 0.18138441443443298, "learning_rate": 5.577665318311776e-07, "loss": 0.0126, "step": 95480 }, { "grad_norm": 0.2017395943403244, "learning_rate": 5.553063855321838e-07, "loss": 0.0112, "step": 95490 }, { "grad_norm": 0.6517603397369385, "learning_rate": 5.528516464305467e-07, "loss": 0.0155, "step": 95500 }, { "grad_norm": 0.40944740176200867, "learning_rate": 5.504023147947068e-07, "loss": 0.0182, "step": 95510 }, { "grad_norm": 0.3111271262168884, "learning_rate": 5.479583908925279e-07, "loss": 0.0142, "step": 95520 }, { "grad_norm": 0.33027318120002747, "learning_rate": 5.455198749912626e-07, "loss": 0.017, "step": 95530 }, { "grad_norm": 0.1923069804906845, "learning_rate": 5.430867673575979e-07, "loss": 0.0116, "step": 95540 }, { "grad_norm": 0.2664858400821686, "learning_rate": 5.406590682575985e-07, "loss": 0.0166, "step": 95550 }, { "grad_norm": 0.23245912790298462, "learning_rate": 5.382367779567687e-07, "loss": 0.0307, "step": 95560 }, { "grad_norm": 0.14778846502304077, "learning_rate": 5.358198967200023e-07, "loss": 0.0128, "step": 95570 }, { "grad_norm": 0.3479951322078705, "learning_rate": 5.334084248115989e-07, "loss": 0.016, "step": 95580 }, { "grad_norm": 0.12130923569202423, "learning_rate": 5.310023624952754e-07, "loss": 0.0121, "step": 95590 }, { "grad_norm": 0.32414886355400085, "learning_rate": 5.286017100341656e-07, "loss": 0.0157, "step": 95600 }, { "grad_norm": 0.5597681403160095, "learning_rate": 5.262064676907874e-07, "loss": 0.018, "step": 95610 }, { "grad_norm": 0.31903019547462463, "learning_rate": 5.238166357270924e-07, "loss": 0.014, "step": 95620 }, { "grad_norm": 0.2058018445968628, "learning_rate": 5.214322144044215e-07, "loss": 0.0116, "step": 95630 }, { "grad_norm": 0.21462537348270416, "learning_rate": 5.190532039835327e-07, "loss": 0.026, "step": 95640 }, { "grad_norm": 0.7461627125740051, "learning_rate": 5.166796047245903e-07, "loss": 0.0243, "step": 95650 }, { "grad_norm": 0.43489113450050354, "learning_rate": 5.143114168871699e-07, "loss": 0.0213, "step": 95660 }, { "grad_norm": 0.794730007648468, "learning_rate": 5.119486407302532e-07, "loss": 0.0152, "step": 95670 }, { "grad_norm": 0.12330164015293121, "learning_rate": 5.095912765122335e-07, "loss": 0.0146, "step": 95680 }, { "grad_norm": 1.1727707386016846, "learning_rate": 5.07239324490899e-07, "loss": 0.0406, "step": 95690 }, { "grad_norm": 0.4059120714664459, "learning_rate": 5.048927849234608e-07, "loss": 0.0244, "step": 95700 }, { "grad_norm": 0.2374630570411682, "learning_rate": 5.025516580665358e-07, "loss": 0.012, "step": 95710 }, { "grad_norm": 0.3719983994960785, "learning_rate": 5.002159441761412e-07, "loss": 0.014, "step": 95720 }, { "grad_norm": 0.43023681640625, "learning_rate": 4.978856435077062e-07, "loss": 0.0136, "step": 95730 }, { "grad_norm": 0.8011865019798279, "learning_rate": 4.955607563160769e-07, "loss": 0.0138, "step": 95740 }, { "grad_norm": 0.20367012917995453, "learning_rate": 4.932412828554944e-07, "loss": 0.0176, "step": 95750 }, { "grad_norm": 0.3450496196746826, "learning_rate": 4.909272233796114e-07, "loss": 0.0154, "step": 95760 }, { "grad_norm": 0.16385766863822937, "learning_rate": 4.88618578141492e-07, "loss": 0.0178, "step": 95770 }, { "grad_norm": 0.4330328106880188, "learning_rate": 4.863153473936011e-07, "loss": 0.0256, "step": 95780 }, { "grad_norm": 0.1859089881181717, "learning_rate": 4.840175313878259e-07, "loss": 0.0241, "step": 95790 }, { "grad_norm": 0.19478029012680054, "learning_rate": 4.817251303754433e-07, "loss": 0.0165, "step": 95800 }, { "grad_norm": 0.746207058429718, "learning_rate": 4.794381446071583e-07, "loss": 0.0141, "step": 95810 }, { "grad_norm": 0.42224010825157166, "learning_rate": 4.771565743330597e-07, "loss": 0.0184, "step": 95820 }, { "grad_norm": 0.4452589452266693, "learning_rate": 4.7488041980266463e-07, "loss": 0.0154, "step": 95830 }, { "grad_norm": 0.3862505555152893, "learning_rate": 4.726096812648795e-07, "loss": 0.0229, "step": 95840 }, { "grad_norm": 0.8196892142295837, "learning_rate": 4.703443589680445e-07, "loss": 0.0221, "step": 95850 }, { "grad_norm": 0.5131235718727112, "learning_rate": 4.6808445315987826e-07, "loss": 0.0158, "step": 95860 }, { "grad_norm": 0.34339696168899536, "learning_rate": 4.658299640875274e-07, "loss": 0.0214, "step": 95870 }, { "grad_norm": 0.47135892510414124, "learning_rate": 4.635808919975393e-07, "loss": 0.0153, "step": 95880 }, { "grad_norm": 0.5760918259620667, "learning_rate": 4.6133723713586706e-07, "loss": 0.024, "step": 95890 }, { "grad_norm": 0.27809417247772217, "learning_rate": 4.590989997478756e-07, "loss": 0.0221, "step": 95900 }, { "grad_norm": 0.22504882514476776, "learning_rate": 4.5686618007833027e-07, "loss": 0.0202, "step": 95910 }, { "grad_norm": 0.2835237383842468, "learning_rate": 4.54638778371419e-07, "loss": 0.0111, "step": 95920 }, { "grad_norm": 0.5406709909439087, "learning_rate": 4.524167948707192e-07, "loss": 0.0167, "step": 95930 }, { "grad_norm": 0.4362676441669464, "learning_rate": 4.5020022981921986e-07, "loss": 0.0169, "step": 95940 }, { "grad_norm": 0.3249443471431732, "learning_rate": 4.47989083459327e-07, "loss": 0.0114, "step": 95950 }, { "grad_norm": 1.11478853225708, "learning_rate": 4.457833560328528e-07, "loss": 0.0251, "step": 95960 }, { "grad_norm": 0.3095875680446625, "learning_rate": 4.4358304778100434e-07, "loss": 0.0245, "step": 95970 }, { "grad_norm": 0.3771102726459503, "learning_rate": 4.413881589444058e-07, "loss": 0.0165, "step": 95980 }, { "grad_norm": 0.3212653398513794, "learning_rate": 4.391986897630873e-07, "loss": 0.0128, "step": 95990 }, { "grad_norm": 0.3837381899356842, "learning_rate": 4.370146404764908e-07, "loss": 0.0148, "step": 96000 }, { "grad_norm": 0.3484165668487549, "learning_rate": 4.3483601132345285e-07, "loss": 0.0138, "step": 96010 }, { "grad_norm": 0.9353174567222595, "learning_rate": 4.32662802542233e-07, "loss": 0.0169, "step": 96020 }, { "grad_norm": 0.20895753800868988, "learning_rate": 4.3049501437047444e-07, "loss": 0.0389, "step": 96030 }, { "grad_norm": 0.6090052723884583, "learning_rate": 4.283326470452653e-07, "loss": 0.0203, "step": 96040 }, { "grad_norm": 0.35256198048591614, "learning_rate": 4.2617570080306647e-07, "loss": 0.0215, "step": 96050 }, { "grad_norm": 0.23904873430728912, "learning_rate": 4.2402417587975585e-07, "loss": 0.0157, "step": 96060 }, { "grad_norm": 0.350864440202713, "learning_rate": 4.218780725106286e-07, "loss": 0.0182, "step": 96070 }, { "grad_norm": 0.39016270637512207, "learning_rate": 4.197373909303748e-07, "loss": 0.0199, "step": 96080 }, { "grad_norm": 0.40908920764923096, "learning_rate": 4.17602131373096e-07, "loss": 0.0174, "step": 96090 }, { "grad_norm": 0.4985893666744232, "learning_rate": 4.154722940723055e-07, "loss": 0.027, "step": 96100 }, { "grad_norm": 0.4982309639453888, "learning_rate": 4.1334787926091136e-07, "loss": 0.0223, "step": 96110 }, { "grad_norm": 0.5506921410560608, "learning_rate": 4.1122888717124444e-07, "loss": 0.0161, "step": 96120 }, { "grad_norm": 0.1983920931816101, "learning_rate": 4.091153180350249e-07, "loss": 0.02, "step": 96130 }, { "grad_norm": 0.6179583668708801, "learning_rate": 4.0700717208340125e-07, "loss": 0.0285, "step": 96140 }, { "grad_norm": 0.2980634570121765, "learning_rate": 4.049044495469112e-07, "loss": 0.0166, "step": 96150 }, { "grad_norm": 0.5140543580055237, "learning_rate": 4.0280715065549866e-07, "loss": 0.0249, "step": 96160 }, { "grad_norm": 0.26604560017585754, "learning_rate": 4.0071527563853575e-07, "loss": 0.0141, "step": 96170 }, { "grad_norm": 0.26123884320259094, "learning_rate": 3.9862882472477823e-07, "loss": 0.0268, "step": 96180 }, { "grad_norm": 0.34459570050239563, "learning_rate": 3.9654779814239375e-07, "loss": 0.0183, "step": 96190 }, { "grad_norm": 0.3944358229637146, "learning_rate": 3.944721961189668e-07, "loss": 0.0168, "step": 96200 }, { "grad_norm": 0.30859366059303284, "learning_rate": 3.9240201888147697e-07, "loss": 0.0245, "step": 96210 }, { "grad_norm": 0.7586358785629272, "learning_rate": 3.9033726665632096e-07, "loss": 0.0188, "step": 96220 }, { "grad_norm": 0.30546125769615173, "learning_rate": 3.882779396692959e-07, "loss": 0.0154, "step": 96230 }, { "grad_norm": 0.35679155588150024, "learning_rate": 3.8622403814559947e-07, "loss": 0.0193, "step": 96240 }, { "grad_norm": 0.13471801578998566, "learning_rate": 3.841755623098575e-07, "loss": 0.0163, "step": 96250 }, { "grad_norm": 0.4214879870414734, "learning_rate": 3.821325123860742e-07, "loss": 0.0154, "step": 96260 }, { "grad_norm": 0.3258025646209717, "learning_rate": 3.800948885976818e-07, "loss": 0.0149, "step": 96270 }, { "grad_norm": 0.48774421215057373, "learning_rate": 3.780626911675078e-07, "loss": 0.0149, "step": 96280 }, { "grad_norm": 0.7213416695594788, "learning_rate": 3.76035920317791e-07, "loss": 0.0183, "step": 96290 }, { "grad_norm": 0.2521781623363495, "learning_rate": 3.740145762701819e-07, "loss": 0.0163, "step": 96300 }, { "grad_norm": 0.40674281120300293, "learning_rate": 3.7199865924572585e-07, "loss": 0.0219, "step": 96310 }, { "grad_norm": 0.3753105103969574, "learning_rate": 3.699881694648799e-07, "loss": 0.0116, "step": 96320 }, { "grad_norm": 0.4028529226779938, "learning_rate": 3.679831071475126e-07, "loss": 0.0148, "step": 96330 }, { "grad_norm": 0.28558993339538574, "learning_rate": 3.65983472512893e-07, "loss": 0.0141, "step": 96340 }, { "grad_norm": 0.4425192177295685, "learning_rate": 3.6398926577969616e-07, "loss": 0.0174, "step": 96350 }, { "grad_norm": 0.5682154297828674, "learning_rate": 3.620004871660032e-07, "loss": 0.0272, "step": 96360 }, { "grad_norm": 0.35902178287506104, "learning_rate": 3.600171368893124e-07, "loss": 0.0161, "step": 96370 }, { "grad_norm": 0.1921919733285904, "learning_rate": 3.580392151665113e-07, "loss": 0.016, "step": 96380 }, { "grad_norm": 0.1470402628183365, "learning_rate": 3.5606672221391024e-07, "loss": 0.0135, "step": 96390 }, { "grad_norm": 0.524969220161438, "learning_rate": 3.540996582472089e-07, "loss": 0.019, "step": 96400 }, { "grad_norm": 0.6029605865478516, "learning_rate": 3.521380234815297e-07, "loss": 0.0192, "step": 96410 }, { "grad_norm": 0.6217435002326965, "learning_rate": 3.5018181813138985e-07, "loss": 0.0196, "step": 96420 }, { "grad_norm": 0.2444109469652176, "learning_rate": 3.482310424107238e-07, "loss": 0.0148, "step": 96430 }, { "grad_norm": 0.30377084016799927, "learning_rate": 3.4628569653285537e-07, "loss": 0.0116, "step": 96440 }, { "grad_norm": 0.7896736264228821, "learning_rate": 3.44345780710531e-07, "loss": 0.0171, "step": 96450 }, { "grad_norm": 0.18176570534706116, "learning_rate": 3.424112951558978e-07, "loss": 0.0229, "step": 96460 }, { "grad_norm": 0.2558166980743408, "learning_rate": 3.40482240080503e-07, "loss": 0.0185, "step": 96470 }, { "grad_norm": 1.0787421464920044, "learning_rate": 3.3855861569531133e-07, "loss": 0.0416, "step": 96480 }, { "grad_norm": 0.3970029652118683, "learning_rate": 3.3664042221068226e-07, "loss": 0.0179, "step": 96490 }, { "grad_norm": 0.5047012567520142, "learning_rate": 3.347276598363924e-07, "loss": 0.0244, "step": 96500 }, { "grad_norm": 0.4089484214782715, "learning_rate": 3.328203287816134e-07, "loss": 0.0208, "step": 96510 }, { "grad_norm": 0.2964397072792053, "learning_rate": 3.3091842925493386e-07, "loss": 0.0095, "step": 96520 }, { "grad_norm": 1.3226310014724731, "learning_rate": 3.2902196146433196e-07, "loss": 0.0215, "step": 96530 }, { "grad_norm": 0.3242294192314148, "learning_rate": 3.2713092561720836e-07, "loss": 0.0142, "step": 96540 }, { "grad_norm": 0.20140056312084198, "learning_rate": 3.2524532192036996e-07, "loss": 0.0173, "step": 96550 }, { "grad_norm": 0.45497268438339233, "learning_rate": 3.2336515058001836e-07, "loss": 0.019, "step": 96560 }, { "grad_norm": 0.46837300062179565, "learning_rate": 3.2149041180176143e-07, "loss": 0.0203, "step": 96570 }, { "grad_norm": 0.5841184258460999, "learning_rate": 3.1962110579062954e-07, "loss": 0.0209, "step": 96580 }, { "grad_norm": 0.2805488407611847, "learning_rate": 3.177572327510314e-07, "loss": 0.0117, "step": 96590 }, { "grad_norm": 0.3711721897125244, "learning_rate": 3.158987928868151e-07, "loss": 0.0168, "step": 96600 }, { "grad_norm": 0.21689535677433014, "learning_rate": 3.1404578640120144e-07, "loss": 0.0176, "step": 96610 }, { "grad_norm": 0.8168206214904785, "learning_rate": 3.1219821349683933e-07, "loss": 0.0127, "step": 96620 }, { "grad_norm": 0.3268585205078125, "learning_rate": 3.103560743757783e-07, "loss": 0.0175, "step": 96630 }, { "grad_norm": 0.6288732290267944, "learning_rate": 3.0851936923946834e-07, "loss": 0.0263, "step": 96640 }, { "grad_norm": 0.2923785448074341, "learning_rate": 3.06688098288771e-07, "loss": 0.0133, "step": 96650 }, { "grad_norm": 0.2610327899456024, "learning_rate": 3.0486226172394825e-07, "loss": 0.0117, "step": 96660 }, { "grad_norm": 0.41806840896606445, "learning_rate": 3.030418597446738e-07, "loss": 0.0172, "step": 96670 }, { "grad_norm": 0.27400168776512146, "learning_rate": 3.012268925500272e-07, "loss": 0.0146, "step": 96680 }, { "grad_norm": 0.9123730063438416, "learning_rate": 2.994173603384831e-07, "loss": 0.0247, "step": 96690 }, { "grad_norm": 0.6798012256622314, "learning_rate": 2.9761326330793317e-07, "loss": 0.0199, "step": 96700 }, { "grad_norm": 0.2956596314907074, "learning_rate": 2.958146016556751e-07, "loss": 0.0147, "step": 96710 }, { "grad_norm": 0.232313871383667, "learning_rate": 2.94021375578396e-07, "loss": 0.0192, "step": 96720 }, { "grad_norm": 0.25271981954574585, "learning_rate": 2.9223358527221135e-07, "loss": 0.0168, "step": 96730 }, { "grad_norm": 0.2549397051334381, "learning_rate": 2.904512309326257e-07, "loss": 0.0181, "step": 96740 }, { "grad_norm": 0.44109776616096497, "learning_rate": 2.8867431275455545e-07, "loss": 0.0171, "step": 96750 }, { "grad_norm": 0.12249186635017395, "learning_rate": 2.86902830932323e-07, "loss": 0.0193, "step": 96760 }, { "grad_norm": 0.35100361704826355, "learning_rate": 2.851367856596565e-07, "loss": 0.0176, "step": 96770 }, { "grad_norm": 0.17935816943645477, "learning_rate": 2.83376177129685e-07, "loss": 0.0175, "step": 96780 }, { "grad_norm": 0.4266473948955536, "learning_rate": 2.8162100553494884e-07, "loss": 0.0208, "step": 96790 }, { "grad_norm": 0.45880067348480225, "learning_rate": 2.798712710673834e-07, "loss": 0.0147, "step": 96800 }, { "grad_norm": 0.1543002426624298, "learning_rate": 2.7812697391834676e-07, "loss": 0.0144, "step": 96810 }, { "grad_norm": 0.3613777756690979, "learning_rate": 2.7638811427858627e-07, "loss": 0.0145, "step": 96820 }, { "grad_norm": 0.20985855162143707, "learning_rate": 2.746546923382665e-07, "loss": 0.0152, "step": 96830 }, { "grad_norm": 0.2066601663827896, "learning_rate": 2.729267082869469e-07, "loss": 0.0148, "step": 96840 }, { "grad_norm": 0.23763564229011536, "learning_rate": 2.712041623135986e-07, "loss": 0.0115, "step": 96850 }, { "grad_norm": 0.30328160524368286, "learning_rate": 2.694870546065931e-07, "loss": 0.0275, "step": 96860 }, { "grad_norm": 0.22685903310775757, "learning_rate": 2.677753853537246e-07, "loss": 0.0237, "step": 96870 }, { "grad_norm": 0.3072012662887573, "learning_rate": 2.660691547421601e-07, "loss": 0.0137, "step": 96880 }, { "grad_norm": 0.37471306324005127, "learning_rate": 2.6436836295850033e-07, "loss": 0.0143, "step": 96890 }, { "grad_norm": 0.47862866520881653, "learning_rate": 2.6267301018874093e-07, "loss": 0.0146, "step": 96900 }, { "grad_norm": 0.3374823033809662, "learning_rate": 2.6098309661828355e-07, "loss": 0.0167, "step": 96910 }, { "grad_norm": 0.08320637047290802, "learning_rate": 2.592986224319305e-07, "loss": 0.0208, "step": 96920 }, { "grad_norm": 0.16789953410625458, "learning_rate": 2.576195878139009e-07, "loss": 0.0122, "step": 96930 }, { "grad_norm": 0.4483886659145355, "learning_rate": 2.5594599294780364e-07, "loss": 0.0214, "step": 96940 }, { "grad_norm": 0.19414031505584717, "learning_rate": 2.542778380166644e-07, "loss": 0.0154, "step": 96950 }, { "grad_norm": 0.21095797419548035, "learning_rate": 2.526151232029095e-07, "loss": 0.023, "step": 96960 }, { "grad_norm": 0.3408872187137604, "learning_rate": 2.5095784868836567e-07, "loss": 0.0157, "step": 96970 }, { "grad_norm": 0.31737157702445984, "learning_rate": 2.493060146542825e-07, "loss": 0.0146, "step": 96980 }, { "grad_norm": 0.18480423092842102, "learning_rate": 2.476596212812876e-07, "loss": 0.0207, "step": 96990 }, { "grad_norm": 0.22607387602329254, "learning_rate": 2.4601866874943703e-07, "loss": 0.0131, "step": 97000 }, { "grad_norm": 0.3634187579154968, "learning_rate": 2.4438315723818163e-07, "loss": 0.0149, "step": 97010 }, { "grad_norm": 0.26146242022514343, "learning_rate": 2.4275308692637833e-07, "loss": 0.0197, "step": 97020 }, { "grad_norm": 0.30626237392425537, "learning_rate": 2.411284579922901e-07, "loss": 0.0209, "step": 97030 }, { "grad_norm": 0.48563462495803833, "learning_rate": 2.395092706135804e-07, "loss": 0.0185, "step": 97040 }, { "grad_norm": 0.411226361989975, "learning_rate": 2.3789552496731872e-07, "loss": 0.015, "step": 97050 }, { "grad_norm": 0.2185826599597931, "learning_rate": 2.3628722122999158e-07, "loss": 0.0131, "step": 97060 }, { "grad_norm": 0.362753689289093, "learning_rate": 2.3468435957747503e-07, "loss": 0.0136, "step": 97070 }, { "grad_norm": 0.3392583727836609, "learning_rate": 2.330869401850566e-07, "loss": 0.0162, "step": 97080 }, { "grad_norm": 0.2790755033493042, "learning_rate": 2.3149496322742437e-07, "loss": 0.0189, "step": 97090 }, { "grad_norm": 0.20931066572666168, "learning_rate": 2.2990842887868346e-07, "loss": 0.0146, "step": 97100 }, { "grad_norm": 0.7314742207527161, "learning_rate": 2.2832733731232292e-07, "loss": 0.018, "step": 97110 }, { "grad_norm": 0.8330298662185669, "learning_rate": 2.2675168870124886e-07, "loss": 0.014, "step": 97120 }, { "grad_norm": 0.6754292249679565, "learning_rate": 2.2518148321778455e-07, "loss": 0.0195, "step": 97130 }, { "grad_norm": 0.45070669054985046, "learning_rate": 2.2361672103363706e-07, "loss": 0.0211, "step": 97140 }, { "grad_norm": 0.39765220880508423, "learning_rate": 2.220574023199251e-07, "loss": 0.0178, "step": 97150 }, { "grad_norm": 0.2900388836860657, "learning_rate": 2.2050352724717894e-07, "loss": 0.0172, "step": 97160 }, { "grad_norm": 0.07511425763368607, "learning_rate": 2.1895509598532372e-07, "loss": 0.0131, "step": 97170 }, { "grad_norm": 0.5149808526039124, "learning_rate": 2.1741210870369066e-07, "loss": 0.0126, "step": 97180 }, { "grad_norm": 0.24324609339237213, "learning_rate": 2.1587456557102814e-07, "loss": 0.0111, "step": 97190 }, { "grad_norm": 0.2951160669326782, "learning_rate": 2.143424667554683e-07, "loss": 0.0182, "step": 97200 }, { "grad_norm": 0.46895831823349, "learning_rate": 2.128158124245716e-07, "loss": 0.0266, "step": 97210 }, { "grad_norm": 0.2936376631259918, "learning_rate": 2.1129460274527668e-07, "loss": 0.06, "step": 97220 }, { "grad_norm": 0.5252715349197388, "learning_rate": 2.0977883788395046e-07, "loss": 0.0173, "step": 97230 }, { "grad_norm": 0.3376510739326477, "learning_rate": 2.0826851800634928e-07, "loss": 0.0141, "step": 97240 }, { "grad_norm": 0.4287285804748535, "learning_rate": 2.0676364327764653e-07, "loss": 0.0155, "step": 97250 }, { "grad_norm": 0.24039793014526367, "learning_rate": 2.0526421386240502e-07, "loss": 0.0149, "step": 97260 }, { "grad_norm": 0.3286582827568054, "learning_rate": 2.0377022992461025e-07, "loss": 0.011, "step": 97270 }, { "grad_norm": 0.44125115871429443, "learning_rate": 2.0228169162763156e-07, "loss": 0.019, "step": 97280 }, { "grad_norm": 0.36587244272232056, "learning_rate": 2.0079859913425536e-07, "loss": 0.0123, "step": 97290 }, { "grad_norm": 0.23946717381477356, "learning_rate": 1.9932095260667417e-07, "loss": 0.0223, "step": 97300 }, { "grad_norm": 0.970481276512146, "learning_rate": 1.9784875220648092e-07, "loss": 0.0165, "step": 97310 }, { "grad_norm": 0.3332318961620331, "learning_rate": 1.9638199809466907e-07, "loss": 0.013, "step": 97320 }, { "grad_norm": 0.44746363162994385, "learning_rate": 1.9492069043164364e-07, "loss": 0.0193, "step": 97330 }, { "grad_norm": 0.3598456382751465, "learning_rate": 1.9346482937721567e-07, "loss": 0.0222, "step": 97340 }, { "grad_norm": 0.3630090355873108, "learning_rate": 1.920144150905856e-07, "loss": 0.0134, "step": 97350 }, { "grad_norm": 0.3774467706680298, "learning_rate": 1.9056944773037655e-07, "loss": 0.018, "step": 97360 }, { "grad_norm": 0.3372405469417572, "learning_rate": 1.8912992745460102e-07, "loss": 0.0213, "step": 97370 }, { "grad_norm": 0.22256198525428772, "learning_rate": 1.876958544206886e-07, "loss": 0.0274, "step": 97380 }, { "grad_norm": 0.24069185554981232, "learning_rate": 1.8626722878546942e-07, "loss": 0.0302, "step": 97390 }, { "grad_norm": 0.28813236951828003, "learning_rate": 1.8484405070516851e-07, "loss": 0.0169, "step": 97400 }, { "grad_norm": 0.20968784391880035, "learning_rate": 1.8342632033542805e-07, "loss": 0.0146, "step": 97410 }, { "grad_norm": 0.19306586682796478, "learning_rate": 1.8201403783129066e-07, "loss": 0.022, "step": 97420 }, { "grad_norm": 0.2850344181060791, "learning_rate": 1.806072033471884e-07, "loss": 0.0156, "step": 97430 }, { "grad_norm": 0.18725457787513733, "learning_rate": 1.79205817036987e-07, "loss": 0.0149, "step": 97440 }, { "grad_norm": 0.3452484905719757, "learning_rate": 1.7780987905393065e-07, "loss": 0.023, "step": 97450 }, { "grad_norm": 0.37068670988082886, "learning_rate": 1.764193895506805e-07, "loss": 0.0187, "step": 97460 }, { "grad_norm": 0.3944615125656128, "learning_rate": 1.7503434867929824e-07, "loss": 0.0167, "step": 97470 }, { "grad_norm": 0.26344001293182373, "learning_rate": 1.7365475659125165e-07, "loss": 0.018, "step": 97480 }, { "grad_norm": 0.35025331377983093, "learning_rate": 1.7228061343740332e-07, "loss": 0.0145, "step": 97490 }, { "grad_norm": 0.19522696733474731, "learning_rate": 1.7091191936803308e-07, "loss": 0.0089, "step": 97500 }, { "grad_norm": 0.6197023987770081, "learning_rate": 1.6954867453281563e-07, "loss": 0.0162, "step": 97510 }, { "grad_norm": 0.24076206982135773, "learning_rate": 1.6819087908084285e-07, "loss": 0.017, "step": 97520 }, { "grad_norm": 0.6987725496292114, "learning_rate": 1.6683853316058485e-07, "loss": 0.0144, "step": 97530 }, { "grad_norm": 0.274036705493927, "learning_rate": 1.6549163691995106e-07, "loss": 0.012, "step": 97540 }, { "grad_norm": 0.27260807156562805, "learning_rate": 1.641501905062237e-07, "loss": 0.0162, "step": 97550 }, { "grad_norm": 0.24325616657733917, "learning_rate": 1.628141940661021e-07, "loss": 0.0214, "step": 97560 }, { "grad_norm": 0.6069096922874451, "learning_rate": 1.6148364774569158e-07, "loss": 0.0197, "step": 97570 }, { "grad_norm": 0.2981836199760437, "learning_rate": 1.6015855169050355e-07, "loss": 0.0138, "step": 97580 }, { "grad_norm": 0.1865309625864029, "learning_rate": 1.5883890604543873e-07, "loss": 0.0222, "step": 97590 }, { "grad_norm": 0.2487974762916565, "learning_rate": 1.5752471095482057e-07, "loss": 0.0124, "step": 97600 }, { "grad_norm": 0.7148568034172058, "learning_rate": 1.5621596656235638e-07, "loss": 0.0214, "step": 97610 }, { "grad_norm": 0.6214354634284973, "learning_rate": 1.5491267301117608e-07, "loss": 0.0212, "step": 97620 }, { "grad_norm": 0.15719099342823029, "learning_rate": 1.536148304438101e-07, "loss": 0.0158, "step": 97630 }, { "grad_norm": 0.4402359426021576, "learning_rate": 1.5232243900217823e-07, "loss": 0.0212, "step": 97640 }, { "grad_norm": 0.44539496302604675, "learning_rate": 1.51035498827623e-07, "loss": 0.0193, "step": 97650 }, { "grad_norm": 0.3470476269721985, "learning_rate": 1.497540100608763e-07, "loss": 0.0209, "step": 97660 }, { "grad_norm": 0.26694703102111816, "learning_rate": 1.4847797284208153e-07, "loss": 0.0157, "step": 97670 }, { "grad_norm": 0.6579695343971252, "learning_rate": 1.472073873107882e-07, "loss": 0.0235, "step": 97680 }, { "grad_norm": 0.2976507842540741, "learning_rate": 1.459422536059407e-07, "loss": 0.0132, "step": 97690 }, { "grad_norm": 0.24480624496936798, "learning_rate": 1.4468257186589508e-07, "loss": 0.0204, "step": 97700 }, { "grad_norm": 0.18939079344272614, "learning_rate": 1.4342834222840217e-07, "loss": 0.0191, "step": 97710 }, { "grad_norm": 0.2301453948020935, "learning_rate": 1.4217956483063566e-07, "loss": 0.0143, "step": 97720 }, { "grad_norm": 0.17190514504909515, "learning_rate": 1.409362398091474e-07, "loss": 0.0107, "step": 97730 }, { "grad_norm": 0.2599819302558899, "learning_rate": 1.3969836729990638e-07, "loss": 0.0162, "step": 97740 }, { "grad_norm": 0.15896672010421753, "learning_rate": 1.384659474382932e-07, "loss": 0.0217, "step": 97750 }, { "grad_norm": 0.566092848777771, "learning_rate": 1.372389803590779e-07, "loss": 0.0195, "step": 97760 }, { "grad_norm": 0.24835911393165588, "learning_rate": 1.360174661964364e-07, "loss": 0.0138, "step": 97770 }, { "grad_norm": 0.26111558079719543, "learning_rate": 1.3480140508396188e-07, "loss": 0.0146, "step": 97780 }, { "grad_norm": 0.5124934315681458, "learning_rate": 1.335907971546313e-07, "loss": 0.0257, "step": 97790 }, { "grad_norm": 0.41934332251548767, "learning_rate": 1.3238564254083875e-07, "loss": 0.0129, "step": 97800 }, { "grad_norm": 0.3786242604255676, "learning_rate": 1.3118594137437323e-07, "loss": 0.0167, "step": 97810 }, { "grad_norm": 0.268396258354187, "learning_rate": 1.2999169378644094e-07, "loss": 0.0327, "step": 97820 }, { "grad_norm": 0.44269099831581116, "learning_rate": 1.2880289990763738e-07, "loss": 0.0158, "step": 97830 }, { "grad_norm": 0.7477367520332336, "learning_rate": 1.2761955986796968e-07, "loss": 0.0247, "step": 97840 }, { "grad_norm": 0.3882962763309479, "learning_rate": 1.264416737968399e-07, "loss": 0.02, "step": 97850 }, { "grad_norm": 1.1944197416305542, "learning_rate": 1.2526924182307275e-07, "loss": 0.0176, "step": 97860 }, { "grad_norm": 0.32057228684425354, "learning_rate": 1.2410226407487124e-07, "loss": 0.0158, "step": 97870 }, { "grad_norm": 0.3434663712978363, "learning_rate": 1.229407406798555e-07, "loss": 0.0149, "step": 97880 }, { "grad_norm": 0.30147960782051086, "learning_rate": 1.2178467176505726e-07, "loss": 0.0138, "step": 97890 }, { "grad_norm": 0.2311314195394516, "learning_rate": 1.2063405745689205e-07, "loss": 0.0223, "step": 97900 }, { "grad_norm": 0.22939130663871765, "learning_rate": 1.1948889788119812e-07, "loss": 0.0181, "step": 97910 }, { "grad_norm": 0.3304961025714874, "learning_rate": 1.1834919316320303e-07, "loss": 0.0192, "step": 97920 }, { "grad_norm": 0.11599093675613403, "learning_rate": 1.1721494342754048e-07, "loss": 0.0172, "step": 97930 }, { "grad_norm": 0.2064911127090454, "learning_rate": 1.1608614879825563e-07, "loss": 0.0129, "step": 97940 }, { "grad_norm": 0.2557316720485687, "learning_rate": 1.1496280939879422e-07, "loss": 0.0197, "step": 97950 }, { "grad_norm": 0.41257843375205994, "learning_rate": 1.1384492535199687e-07, "loss": 0.0175, "step": 97960 }, { "grad_norm": 0.4083136022090912, "learning_rate": 1.1273249678011578e-07, "loss": 0.0153, "step": 97970 }, { "grad_norm": 0.3309977650642395, "learning_rate": 1.1162552380480362e-07, "loss": 0.0186, "step": 97980 }, { "grad_norm": 0.27132293581962585, "learning_rate": 1.1052400654711359e-07, "loss": 0.0107, "step": 97990 }, { "grad_norm": 0.38627326488494873, "learning_rate": 1.0942794512751597e-07, "loss": 0.0223, "step": 98000 }, { "grad_norm": 0.25221747159957886, "learning_rate": 1.0833733966587045e-07, "loss": 0.0149, "step": 98010 }, { "grad_norm": 0.5189645290374756, "learning_rate": 1.0725219028143718e-07, "loss": 0.0359, "step": 98020 }, { "grad_norm": 0.33502835035324097, "learning_rate": 1.0617249709289345e-07, "loss": 0.0162, "step": 98030 }, { "grad_norm": 0.3806988596916199, "learning_rate": 1.0509826021831703e-07, "loss": 0.0163, "step": 98040 }, { "grad_norm": 0.32194146513938904, "learning_rate": 1.0402947977517508e-07, "loss": 0.0107, "step": 98050 }, { "grad_norm": 0.4693092405796051, "learning_rate": 1.0296615588035185e-07, "loss": 0.0228, "step": 98060 }, { "grad_norm": 0.306338906288147, "learning_rate": 1.0190828865012659e-07, "loss": 0.0169, "step": 98070 }, { "grad_norm": 0.2693197727203369, "learning_rate": 1.008558782001956e-07, "loss": 0.0266, "step": 98080 }, { "grad_norm": 0.616392195224762, "learning_rate": 9.980892464563906e-08, "loss": 0.0188, "step": 98090 }, { "grad_norm": 0.29584500193595886, "learning_rate": 9.876742810095985e-08, "loss": 0.0176, "step": 98100 }, { "grad_norm": 0.27279916405677795, "learning_rate": 9.77313886800446e-08, "loss": 0.0181, "step": 98110 }, { "grad_norm": 0.16330629587173462, "learning_rate": 9.670080649619717e-08, "loss": 0.0158, "step": 98120 }, { "grad_norm": 0.27497315406799316, "learning_rate": 9.567568166212737e-08, "loss": 0.017, "step": 98130 }, { "grad_norm": 0.15615715086460114, "learning_rate": 9.465601428992888e-08, "loss": 0.0126, "step": 98140 }, { "grad_norm": 0.25383254885673523, "learning_rate": 9.364180449111803e-08, "loss": 0.013, "step": 98150 }, { "grad_norm": 0.22530479729175568, "learning_rate": 9.263305237661169e-08, "loss": 0.0252, "step": 98160 }, { "grad_norm": 0.26989150047302246, "learning_rate": 9.162975805671603e-08, "loss": 0.013, "step": 98170 }, { "grad_norm": 0.19822609424591064, "learning_rate": 9.063192164115442e-08, "loss": 0.0241, "step": 98180 }, { "grad_norm": 0.2599451541900635, "learning_rate": 8.963954323904511e-08, "loss": 0.0187, "step": 98190 }, { "grad_norm": 0.2954573631286621, "learning_rate": 8.865262295891796e-08, "loss": 0.0118, "step": 98200 }, { "grad_norm": 0.45029690861701965, "learning_rate": 8.767116090870331e-08, "loss": 0.0156, "step": 98210 }, { "grad_norm": 0.24743591248989105, "learning_rate": 8.669515719572086e-08, "loss": 0.0151, "step": 98220 }, { "grad_norm": 0.17834006249904633, "learning_rate": 8.572461192671855e-08, "loss": 0.0226, "step": 98230 }, { "grad_norm": 0.48644566535949707, "learning_rate": 8.475952520782815e-08, "loss": 0.0236, "step": 98240 }, { "grad_norm": 0.47168081998825073, "learning_rate": 8.379989714458747e-08, "loss": 0.0176, "step": 98250 }, { "grad_norm": 0.44159388542175293, "learning_rate": 8.284572784194034e-08, "loss": 0.0185, "step": 98260 }, { "grad_norm": 0.5607385635375977, "learning_rate": 8.189701740424216e-08, "loss": 0.019, "step": 98270 }, { "grad_norm": 0.4788631498813629, "learning_rate": 8.095376593522663e-08, "loss": 0.0229, "step": 98280 }, { "grad_norm": 0.24378849565982819, "learning_rate": 8.001597353806123e-08, "loss": 0.0147, "step": 98290 }, { "grad_norm": 0.38571906089782715, "learning_rate": 7.90836403152917e-08, "loss": 0.0176, "step": 98300 }, { "grad_norm": 0.22336353361606598, "learning_rate": 7.815676636888092e-08, "loss": 0.0172, "step": 98310 }, { "grad_norm": 0.2737765610218048, "learning_rate": 7.723535180019226e-08, "loss": 0.0203, "step": 98320 }, { "grad_norm": 0.503097653388977, "learning_rate": 7.6319396709984e-08, "loss": 0.0159, "step": 98330 }, { "grad_norm": 0.5470160841941833, "learning_rate": 7.540890119843158e-08, "loss": 0.0187, "step": 98340 }, { "grad_norm": 0.6323568820953369, "learning_rate": 7.450386536509979e-08, "loss": 0.0178, "step": 98350 }, { "grad_norm": 0.34282493591308594, "learning_rate": 7.360428930895947e-08, "loss": 0.0132, "step": 98360 }, { "grad_norm": 0.24406303465366364, "learning_rate": 7.271017312839302e-08, "loss": 0.0164, "step": 98370 }, { "grad_norm": 0.6340881586074829, "learning_rate": 7.182151692118333e-08, "loss": 0.0173, "step": 98380 }, { "grad_norm": 0.2514425814151764, "learning_rate": 7.093832078449713e-08, "loss": 0.0178, "step": 98390 }, { "grad_norm": 0.40508896112442017, "learning_rate": 7.006058481493493e-08, "loss": 0.0215, "step": 98400 }, { "grad_norm": 0.42869749665260315, "learning_rate": 6.918830910847552e-08, "loss": 0.0135, "step": 98410 }, { "grad_norm": 0.48829716444015503, "learning_rate": 6.832149376052033e-08, "loss": 0.017, "step": 98420 }, { "grad_norm": 0.4599863886833191, "learning_rate": 6.746013886584912e-08, "loss": 0.0164, "step": 98430 }, { "grad_norm": 0.6541864275932312, "learning_rate": 6.660424451866431e-08, "loss": 0.0202, "step": 98440 }, { "grad_norm": 0.2872515618801117, "learning_rate": 6.575381081256327e-08, "loss": 0.0109, "step": 98450 }, { "grad_norm": 0.24869872629642487, "learning_rate": 6.49088378405549e-08, "loss": 0.0138, "step": 98460 }, { "grad_norm": 0.4577551484107971, "learning_rate": 6.406932569503755e-08, "loss": 0.0235, "step": 98470 }, { "grad_norm": 0.3239578604698181, "learning_rate": 6.323527446782662e-08, "loss": 0.0194, "step": 98480 }, { "grad_norm": 0.24431400001049042, "learning_rate": 6.240668425012142e-08, "loss": 0.0175, "step": 98490 }, { "grad_norm": 0.3666805326938629, "learning_rate": 6.158355513254388e-08, "loss": 0.0165, "step": 98500 }, { "grad_norm": 0.21656498312950134, "learning_rate": 6.076588720510534e-08, "loss": 0.0166, "step": 98510 }, { "grad_norm": 0.14930947124958038, "learning_rate": 5.995368055722316e-08, "loss": 0.02, "step": 98520 }, { "grad_norm": 0.31461095809936523, "learning_rate": 5.914693527773185e-08, "loss": 0.0161, "step": 98530 }, { "grad_norm": 0.3993350863456726, "learning_rate": 5.834565145483861e-08, "loss": 0.0178, "step": 98540 }, { "grad_norm": 0.24540947377681732, "learning_rate": 5.754982917618446e-08, "loss": 0.0161, "step": 98550 }, { "grad_norm": 0.3878777027130127, "learning_rate": 5.675946852878866e-08, "loss": 0.0154, "step": 98560 }, { "grad_norm": 0.420257568359375, "learning_rate": 5.59745695990932e-08, "loss": 0.0158, "step": 98570 }, { "grad_norm": 0.293484091758728, "learning_rate": 5.51951324729294e-08, "loss": 0.014, "step": 98580 }, { "grad_norm": 0.3600734770298004, "learning_rate": 5.4421157235529093e-08, "loss": 0.0176, "step": 98590 }, { "grad_norm": 0.4565820097923279, "learning_rate": 5.365264397154679e-08, "loss": 0.0188, "step": 98600 }, { "grad_norm": 0.26940998435020447, "learning_rate": 5.288959276501526e-08, "loss": 0.0109, "step": 98610 }, { "grad_norm": 0.1904665231704712, "learning_rate": 5.2132003699378876e-08, "loss": 0.0185, "step": 98620 }, { "grad_norm": 0.22109638154506683, "learning_rate": 5.137987685749912e-08, "loss": 0.0174, "step": 98630 }, { "grad_norm": 0.17503035068511963, "learning_rate": 5.0633212321610227e-08, "loss": 0.0171, "step": 98640 }, { "grad_norm": 0.5191037654876709, "learning_rate": 4.989201017338019e-08, "loss": 0.0121, "step": 98650 }, { "grad_norm": 0.2739184498786926, "learning_rate": 4.9156270493860846e-08, "loss": 0.0126, "step": 98660 }, { "grad_norm": 0.3968559801578522, "learning_rate": 4.842599336351561e-08, "loss": 0.0228, "step": 98670 }, { "grad_norm": 0.41058921813964844, "learning_rate": 4.7701178862197274e-08, "loss": 0.0209, "step": 98680 }, { "grad_norm": 0.22236891090869904, "learning_rate": 4.698182706918131e-08, "loss": 0.0164, "step": 98690 }, { "grad_norm": 0.2694559395313263, "learning_rate": 4.6267938063121466e-08, "loss": 0.02, "step": 98700 }, { "grad_norm": 0.20240920782089233, "learning_rate": 4.555951192209973e-08, "loss": 0.0228, "step": 98710 }, { "grad_norm": 0.2414870709180832, "learning_rate": 4.485654872358747e-08, "loss": 0.0174, "step": 98720 }, { "grad_norm": 0.3010258376598358, "learning_rate": 4.415904854445097e-08, "loss": 0.013, "step": 98730 }, { "grad_norm": 0.4581926763057709, "learning_rate": 4.34670114609792e-08, "loss": 0.0164, "step": 98740 }, { "grad_norm": 0.5873720049858093, "learning_rate": 4.278043754884498e-08, "loss": 0.0289, "step": 98750 }, { "grad_norm": 0.34081584215164185, "learning_rate": 4.209932688312712e-08, "loss": 0.014, "step": 98760 }, { "grad_norm": 0.1681206226348877, "learning_rate": 4.14236795383216e-08, "loss": 0.0158, "step": 98770 }, { "grad_norm": 0.3149348795413971, "learning_rate": 4.075349558830821e-08, "loss": 0.0265, "step": 98780 }, { "grad_norm": 0.22719043493270874, "learning_rate": 4.008877510638387e-08, "loss": 0.0172, "step": 98790 }, { "grad_norm": 0.6112929582595825, "learning_rate": 3.942951816523488e-08, "loss": 0.0194, "step": 98800 }, { "grad_norm": 0.3908992111682892, "learning_rate": 3.8775724836959125e-08, "loss": 0.0142, "step": 98810 }, { "grad_norm": 0.5002874135971069, "learning_rate": 3.812739519305497e-08, "loss": 0.0153, "step": 98820 }, { "grad_norm": 0.3380928635597229, "learning_rate": 3.748452930442126e-08, "loss": 0.0182, "step": 98830 }, { "grad_norm": 0.2469554841518402, "learning_rate": 3.6847127241362855e-08, "loss": 0.0159, "step": 98840 }, { "grad_norm": 0.3533075749874115, "learning_rate": 3.621518907358512e-08, "loss": 0.0178, "step": 98850 }, { "grad_norm": 0.5575026273727417, "learning_rate": 3.558871487019388e-08, "loss": 0.0178, "step": 98860 }, { "grad_norm": 0.4088343381881714, "learning_rate": 3.4967704699701007e-08, "loss": 0.0142, "step": 98870 }, { "grad_norm": 0.3912276029586792, "learning_rate": 3.435215863001884e-08, "loss": 0.0149, "step": 98880 }, { "grad_norm": 0.4314599931240082, "learning_rate": 3.374207672846019e-08, "loss": 0.0139, "step": 98890 }, { "grad_norm": 0.3873180150985718, "learning_rate": 3.313745906174392e-08, "loss": 0.0156, "step": 98900 }, { "grad_norm": 0.4216981530189514, "learning_rate": 3.253830569599492e-08, "loss": 0.0105, "step": 98910 }, { "grad_norm": 0.5529046654701233, "learning_rate": 3.1944616696727436e-08, "loss": 0.0209, "step": 98920 }, { "grad_norm": 0.6051570773124695, "learning_rate": 3.135639212887287e-08, "loss": 0.0178, "step": 98930 }, { "grad_norm": 0.3268304765224457, "learning_rate": 3.077363205675754e-08, "loss": 0.0164, "step": 98940 }, { "grad_norm": 0.24911309778690338, "learning_rate": 3.0196336544113804e-08, "loss": 0.02, "step": 98950 }, { "grad_norm": 0.45831775665283203, "learning_rate": 2.9624505654063383e-08, "loss": 0.0161, "step": 98960 }, { "grad_norm": 0.4096488058567047, "learning_rate": 2.90581394491507e-08, "loss": 0.0196, "step": 98970 }, { "grad_norm": 0.2940226197242737, "learning_rate": 2.8497237991309545e-08, "loss": 0.0127, "step": 98980 }, { "grad_norm": 0.42293354868888855, "learning_rate": 2.7941801341879735e-08, "loss": 0.0178, "step": 98990 }, { "grad_norm": 0.360921174287796, "learning_rate": 2.7391829561601578e-08, "loss": 0.0181, "step": 99000 }, { "grad_norm": 0.18161234259605408, "learning_rate": 2.6847322710621403e-08, "loss": 0.0179, "step": 99010 }, { "grad_norm": 0.39731380343437195, "learning_rate": 2.6308280848486022e-08, "loss": 0.0296, "step": 99020 }, { "grad_norm": 0.5036655068397522, "learning_rate": 2.5774704034137177e-08, "loss": 0.015, "step": 99030 }, { "grad_norm": 0.3800545334815979, "learning_rate": 2.524659232593929e-08, "loss": 0.0177, "step": 99040 }, { "grad_norm": 0.15652580559253693, "learning_rate": 2.472394578163506e-08, "loss": 0.0153, "step": 99050 }, { "grad_norm": 0.4068228602409363, "learning_rate": 2.4206764458378772e-08, "loss": 0.0129, "step": 99060 }, { "grad_norm": 0.3100242614746094, "learning_rate": 2.3695048412736285e-08, "loss": 0.0173, "step": 99070 }, { "grad_norm": 0.27449920773506165, "learning_rate": 2.3188797700668395e-08, "loss": 0.017, "step": 99080 }, { "grad_norm": 0.5258236527442932, "learning_rate": 2.268801237753082e-08, "loss": 0.0154, "step": 99090 }, { "grad_norm": 0.1334501951932907, "learning_rate": 2.2192692498090862e-08, "loss": 0.0232, "step": 99100 }, { "grad_norm": 0.4134562313556671, "learning_rate": 2.170283811652185e-08, "loss": 0.0211, "step": 99110 }, { "grad_norm": 0.23034562170505524, "learning_rate": 2.12184492863865e-08, "loss": 0.012, "step": 99120 }, { "grad_norm": 0.359526664018631, "learning_rate": 2.073952606066465e-08, "loss": 0.0176, "step": 99130 }, { "grad_norm": 0.3422880172729492, "learning_rate": 2.026606849171997e-08, "loss": 0.0248, "step": 99140 }, { "grad_norm": 0.2767057418823242, "learning_rate": 1.9798076631333262e-08, "loss": 0.0153, "step": 99150 }, { "grad_norm": 0.4243469834327698, "learning_rate": 1.933555053069136e-08, "loss": 0.0134, "step": 99160 }, { "grad_norm": 0.48278287053108215, "learning_rate": 1.887849024036492e-08, "loss": 0.0182, "step": 99170 }, { "grad_norm": 0.4060365855693817, "learning_rate": 1.8426895810341736e-08, "loss": 0.0136, "step": 99180 }, { "grad_norm": 0.3138636648654938, "learning_rate": 1.798076729000453e-08, "loss": 0.0196, "step": 99190 }, { "grad_norm": 0.29829689860343933, "learning_rate": 1.75401047281476e-08, "loss": 0.0174, "step": 99200 }, { "grad_norm": 0.5128330588340759, "learning_rate": 1.7104908172954626e-08, "loss": 0.0181, "step": 99210 }, { "grad_norm": 0.5566613078117371, "learning_rate": 1.6675177672020868e-08, "loss": 0.0269, "step": 99220 }, { "grad_norm": 0.22890889644622803, "learning_rate": 1.6250913272342073e-08, "loss": 0.0163, "step": 99230 }, { "grad_norm": 0.2758342921733856, "learning_rate": 1.5832115020314453e-08, "loss": 0.0147, "step": 99240 }, { "grad_norm": 0.27227675914764404, "learning_rate": 1.5418782961734712e-08, "loss": 0.0214, "step": 99250 }, { "grad_norm": 0.19872169196605682, "learning_rate": 1.501091714181113e-08, "loss": 0.0179, "step": 99260 }, { "grad_norm": 0.1821938008069992, "learning_rate": 1.460851760513582e-08, "loss": 0.0093, "step": 99270 }, { "grad_norm": 0.47147753834724426, "learning_rate": 1.4211584395723566e-08, "loss": 0.0238, "step": 99280 }, { "grad_norm": 0.35108956694602966, "learning_rate": 1.382011755697854e-08, "loss": 0.0212, "step": 99290 }, { "grad_norm": 0.3390578329563141, "learning_rate": 1.3434117131716494e-08, "loss": 0.0116, "step": 99300 }, { "grad_norm": 0.321364164352417, "learning_rate": 1.3053583162142557e-08, "loss": 0.0254, "step": 99310 }, { "grad_norm": 0.20890848338603973, "learning_rate": 1.2678515689873438e-08, "loss": 0.0162, "step": 99320 }, { "grad_norm": 0.24227173626422882, "learning_rate": 1.2308914755931877e-08, "loss": 0.0173, "step": 99330 }, { "grad_norm": 0.76868736743927, "learning_rate": 1.1944780400729993e-08, "loss": 0.0278, "step": 99340 }, { "grad_norm": 0.28026989102363586, "learning_rate": 1.1586112664085935e-08, "loss": 0.014, "step": 99350 }, { "grad_norm": 0.3032866418361664, "learning_rate": 1.1232911585234984e-08, "loss": 0.0152, "step": 99360 }, { "grad_norm": 0.38287368416786194, "learning_rate": 1.0885177202790697e-08, "loss": 0.0206, "step": 99370 }, { "grad_norm": 0.25745993852615356, "learning_rate": 1.0542909554789315e-08, "loss": 0.0158, "step": 99380 }, { "grad_norm": 0.10130961984395981, "learning_rate": 1.0206108678656456e-08, "loss": 0.0156, "step": 99390 }, { "grad_norm": 0.24827787280082703, "learning_rate": 9.874774611223769e-09, "loss": 0.0094, "step": 99400 }, { "grad_norm": 0.19006076455116272, "learning_rate": 9.548907388728933e-09, "loss": 0.0201, "step": 99410 }, { "grad_norm": 0.21235835552215576, "learning_rate": 9.228507046804557e-09, "loss": 0.0137, "step": 99420 }, { "grad_norm": 0.24781711399555206, "learning_rate": 8.91357362048928e-09, "loss": 0.0136, "step": 99430 }, { "grad_norm": 0.21391160786151886, "learning_rate": 8.604107144227769e-09, "loss": 0.0258, "step": 99440 }, { "grad_norm": 0.23127904534339905, "learning_rate": 8.300107651859623e-09, "loss": 0.0155, "step": 99450 }, { "grad_norm": 0.17449264228343964, "learning_rate": 8.001575176630472e-09, "loss": 0.0162, "step": 99460 }, { "grad_norm": 0.14762595295906067, "learning_rate": 7.708509751186422e-09, "loss": 0.0137, "step": 99470 }, { "grad_norm": 0.2438325732946396, "learning_rate": 7.420911407579611e-09, "loss": 0.015, "step": 99480 }, { "grad_norm": 0.47667428851127625, "learning_rate": 7.13878017725711e-09, "loss": 0.0154, "step": 99490 }, { "grad_norm": 0.3916815519332886, "learning_rate": 6.8621160910720125e-09, "loss": 0.0159, "step": 99500 }, { "grad_norm": 0.16766957938671112, "learning_rate": 6.59091917928345e-09, "loss": 0.0161, "step": 99510 }, { "grad_norm": 0.7855103015899658, "learning_rate": 6.32518947155103e-09, "loss": 0.025, "step": 99520 }, { "grad_norm": 0.2659263610839844, "learning_rate": 6.064926996929288e-09, "loss": 0.0108, "step": 99530 }, { "grad_norm": 0.3464621305465698, "learning_rate": 5.810131783884343e-09, "loss": 0.0149, "step": 99540 }, { "grad_norm": 0.39257746934890747, "learning_rate": 5.5608038602772415e-09, "loss": 0.0145, "step": 99550 }, { "grad_norm": 0.5313186645507812, "learning_rate": 5.316943253375062e-09, "loss": 0.0156, "step": 99560 }, { "grad_norm": 0.15652352571487427, "learning_rate": 5.07854998984536e-09, "loss": 0.0179, "step": 99570 }, { "grad_norm": 0.4857019782066345, "learning_rate": 4.845624095756174e-09, "loss": 0.0116, "step": 99580 }, { "grad_norm": 0.4014892578125, "learning_rate": 4.618165596587121e-09, "loss": 0.0254, "step": 99590 }, { "grad_norm": 0.22347119450569153, "learning_rate": 4.396174517207197e-09, "loss": 0.0132, "step": 99600 }, { "grad_norm": 0.34794947504997253, "learning_rate": 4.179650881896979e-09, "loss": 0.026, "step": 99610 }, { "grad_norm": 0.586854100227356, "learning_rate": 3.968594714331975e-09, "loss": 0.015, "step": 99620 }, { "grad_norm": 0.3712857663631439, "learning_rate": 3.763006037593719e-09, "loss": 0.0194, "step": 99630 }, { "grad_norm": 0.24625350534915924, "learning_rate": 3.5628848741586786e-09, "loss": 0.0265, "step": 99640 }, { "grad_norm": 0.27556777000427246, "learning_rate": 3.368231245926001e-09, "loss": 0.0125, "step": 99650 }, { "grad_norm": 0.19717739522457123, "learning_rate": 3.179045174167561e-09, "loss": 0.0135, "step": 99660 }, { "grad_norm": 0.12376169115304947, "learning_rate": 2.9953266795834657e-09, "loss": 0.0196, "step": 99670 }, { "grad_norm": 0.19399100542068481, "learning_rate": 2.817075782263201e-09, "loss": 0.0231, "step": 99680 }, { "grad_norm": 0.29386940598487854, "learning_rate": 2.6442925016967325e-09, "loss": 0.0126, "step": 99690 }, { "grad_norm": 0.26361775398254395, "learning_rate": 2.4769768567800555e-09, "loss": 0.0136, "step": 99700 }, { "grad_norm": 0.3485375642776489, "learning_rate": 2.315128865809646e-09, "loss": 0.0138, "step": 99710 }, { "grad_norm": 0.12933756411075592, "learning_rate": 2.1587485464880097e-09, "loss": 0.0139, "step": 99720 }, { "grad_norm": 0.30623164772987366, "learning_rate": 2.0078359159125813e-09, "loss": 0.0187, "step": 99730 }, { "grad_norm": 0.44577276706695557, "learning_rate": 1.8623909905923776e-09, "loss": 0.013, "step": 99740 }, { "grad_norm": 0.3404328525066376, "learning_rate": 1.7224137864257916e-09, "loss": 0.0145, "step": 99750 }, { "grad_norm": 0.16817574203014374, "learning_rate": 1.5879043187283505e-09, "loss": 0.015, "step": 99760 }, { "grad_norm": 0.09258098155260086, "learning_rate": 1.458862602204958e-09, "loss": 0.03, "step": 99770 }, { "grad_norm": 0.37239307165145874, "learning_rate": 1.3352886509720996e-09, "loss": 0.0156, "step": 99780 }, { "grad_norm": 0.23258346319198608, "learning_rate": 1.2171824785356389e-09, "loss": 0.013, "step": 99790 }, { "grad_norm": 0.7953724265098572, "learning_rate": 1.1045440978185718e-09, "loss": 0.0173, "step": 99800 }, { "grad_norm": 0.21959452331066132, "learning_rate": 9.973735211332714e-10, "loss": 0.0145, "step": 99810 }, { "grad_norm": 0.432484894990921, "learning_rate": 8.956707602036929e-10, "loss": 0.0133, "step": 99820 }, { "grad_norm": 0.4297583997249603, "learning_rate": 7.994358261542712e-10, "loss": 0.0158, "step": 99830 }, { "grad_norm": 0.2361152470111847, "learning_rate": 7.086687295043692e-10, "loss": 0.0183, "step": 99840 }, { "grad_norm": 0.4212273359298706, "learning_rate": 6.23369480179381e-10, "loss": 0.0149, "step": 99850 }, { "grad_norm": 0.36861541867256165, "learning_rate": 5.43538087510731e-10, "loss": 0.0139, "step": 99860 }, { "grad_norm": 0.28918689489364624, "learning_rate": 4.691745602303233e-10, "loss": 0.0179, "step": 99870 }, { "grad_norm": 1.6965487003326416, "learning_rate": 4.0027890646499034e-10, "loss": 0.019, "step": 99880 }, { "grad_norm": 0.42132368683815, "learning_rate": 3.368511337531466e-10, "loss": 0.0195, "step": 99890 }, { "grad_norm": 0.6125609874725342, "learning_rate": 2.788912490281348e-10, "loss": 0.0217, "step": 99900 }, { "grad_norm": 0.4062001407146454, "learning_rate": 2.2639925863487954e-10, "loss": 0.0229, "step": 99910 }, { "grad_norm": 1.4377720355987549, "learning_rate": 1.7937516830213163e-10, "loss": 0.0204, "step": 99920 }, { "grad_norm": 0.1736665666103363, "learning_rate": 1.3781898318687702e-10, "loss": 0.012, "step": 99930 }, { "grad_norm": 0.3953746557235718, "learning_rate": 1.0173070781882566e-10, "loss": 0.0123, "step": 99940 }, { "grad_norm": 0.3171420991420746, "learning_rate": 7.111034615592261e-11, "loss": 0.0134, "step": 99950 }, { "grad_norm": 0.13814836740493774, "learning_rate": 4.595790153993918e-11, "loss": 0.0191, "step": 99960 }, { "grad_norm": 0.3042231500148773, "learning_rate": 2.6273376729779585e-11, "loss": 0.0114, "step": 99970 }, { "grad_norm": 0.30573901534080505, "learning_rate": 1.2056773868174276e-11, "loss": 0.0189, "step": 99980 }, { "grad_norm": 0.5084930062294006, "learning_rate": 3.3080945149865926e-12, "loss": 0.0155, "step": 99990 }, { "grad_norm": 0.48580417037010193, "learning_rate": 2.7339625008337977e-14, "loss": 0.0263, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }