{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.8319921493530273, "learning_rate": 1.8e-07, "loss": 1.4988, "step": 10 }, { "grad_norm": 0.979226291179657, "learning_rate": 3.8e-07, "loss": 1.5451, "step": 20 }, { "grad_norm": 1.0143461227416992, "learning_rate": 5.8e-07, "loss": 1.4929, "step": 30 }, { "grad_norm": 0.7791479825973511, "learning_rate": 7.8e-07, "loss": 1.4705, "step": 40 }, { "grad_norm": 0.8534229397773743, "learning_rate": 9.8e-07, "loss": 1.4965, "step": 50 }, { "grad_norm": 1.0628036260604858, "learning_rate": 1.18e-06, "loss": 1.5132, "step": 60 }, { "grad_norm": 1.9677515029907227, "learning_rate": 1.3800000000000001e-06, "loss": 1.4233, "step": 70 }, { "grad_norm": 2.3712735176086426, "learning_rate": 1.5800000000000003e-06, "loss": 1.3731, "step": 80 }, { "grad_norm": 1.5589920282363892, "learning_rate": 1.7800000000000001e-06, "loss": 1.3046, "step": 90 }, { "grad_norm": 1.4982181787490845, "learning_rate": 1.98e-06, "loss": 1.289, "step": 100 }, { "grad_norm": 1.228908896446228, "learning_rate": 2.1800000000000003e-06, "loss": 1.3073, "step": 110 }, { "grad_norm": 2.2068281173706055, "learning_rate": 2.38e-06, "loss": 1.2803, "step": 120 }, { "grad_norm": 1.885425329208374, "learning_rate": 2.5800000000000003e-06, "loss": 1.2191, "step": 130 }, { "grad_norm": 1.5699480772018433, "learning_rate": 2.78e-06, "loss": 1.2248, "step": 140 }, { "grad_norm": 1.6360689401626587, "learning_rate": 2.9800000000000003e-06, "loss": 1.212, "step": 150 }, { "grad_norm": 2.28049898147583, "learning_rate": 3.1800000000000005e-06, "loss": 1.1666, "step": 160 }, { "grad_norm": 1.643598198890686, "learning_rate": 3.38e-06, "loss": 1.1782, "step": 170 }, { "grad_norm": 2.4969725608825684, "learning_rate": 3.58e-06, "loss": 1.2278, "step": 180 }, { "grad_norm": 1.762095332145691, "learning_rate": 3.7800000000000002e-06, "loss": 1.2232, "step": 190 }, { "grad_norm": 2.209320306777954, "learning_rate": 3.98e-06, "loss": 1.1946, "step": 200 }, { "grad_norm": 1.461064100265503, "learning_rate": 4.18e-06, "loss": 1.1935, "step": 210 }, { "grad_norm": 1.8928930759429932, "learning_rate": 4.38e-06, "loss": 1.1836, "step": 220 }, { "grad_norm": 1.4957484006881714, "learning_rate": 4.58e-06, "loss": 1.1783, "step": 230 }, { "grad_norm": 1.4120131731033325, "learning_rate": 4.780000000000001e-06, "loss": 1.1709, "step": 240 }, { "grad_norm": 2.549839973449707, "learning_rate": 4.98e-06, "loss": 1.1383, "step": 250 }, { "grad_norm": 1.8038805723190308, "learning_rate": 5.18e-06, "loss": 1.1193, "step": 260 }, { "grad_norm": 2.2915360927581787, "learning_rate": 5.38e-06, "loss": 1.1292, "step": 270 }, { "grad_norm": 1.9512310028076172, "learning_rate": 5.580000000000001e-06, "loss": 1.1068, "step": 280 }, { "grad_norm": 2.062634229660034, "learning_rate": 5.78e-06, "loss": 1.1139, "step": 290 }, { "grad_norm": 2.1579437255859375, "learning_rate": 5.98e-06, "loss": 1.1303, "step": 300 }, { "grad_norm": 1.6182597875595093, "learning_rate": 6.18e-06, "loss": 1.1106, "step": 310 }, { "grad_norm": 2.1663429737091064, "learning_rate": 6.38e-06, "loss": 1.0966, "step": 320 }, { "grad_norm": 1.8079423904418945, "learning_rate": 6.58e-06, "loss": 1.0941, "step": 330 }, { "grad_norm": 1.9228447675704956, "learning_rate": 6.78e-06, "loss": 1.0864, "step": 340 }, { "grad_norm": 1.6065611839294434, "learning_rate": 6.98e-06, "loss": 1.0896, "step": 350 }, { "grad_norm": 2.8922693729400635, "learning_rate": 7.180000000000001e-06, "loss": 1.0924, "step": 360 }, { "grad_norm": 1.3320786952972412, "learning_rate": 7.3800000000000005e-06, "loss": 1.0816, "step": 370 }, { "grad_norm": 1.389448881149292, "learning_rate": 7.580000000000001e-06, "loss": 1.0769, "step": 380 }, { "grad_norm": 1.331244707107544, "learning_rate": 7.78e-06, "loss": 1.0346, "step": 390 }, { "grad_norm": 1.7130367755889893, "learning_rate": 7.98e-06, "loss": 1.0936, "step": 400 }, { "grad_norm": 1.6302123069763184, "learning_rate": 8.18e-06, "loss": 1.0886, "step": 410 }, { "grad_norm": 1.3476558923721313, "learning_rate": 8.380000000000001e-06, "loss": 1.0738, "step": 420 }, { "grad_norm": 1.956231713294983, "learning_rate": 8.580000000000001e-06, "loss": 1.069, "step": 430 }, { "grad_norm": 1.4689406156539917, "learning_rate": 8.78e-06, "loss": 1.0628, "step": 440 }, { "grad_norm": 1.6344307661056519, "learning_rate": 8.98e-06, "loss": 1.0632, "step": 450 }, { "grad_norm": 1.526954174041748, "learning_rate": 9.180000000000002e-06, "loss": 1.0682, "step": 460 }, { "grad_norm": 2.6690754890441895, "learning_rate": 9.38e-06, "loss": 1.0722, "step": 470 }, { "grad_norm": 1.5233640670776367, "learning_rate": 9.58e-06, "loss": 1.0951, "step": 480 }, { "grad_norm": 1.4242873191833496, "learning_rate": 9.78e-06, "loss": 1.0712, "step": 490 }, { "grad_norm": 1.3727644681930542, "learning_rate": 9.980000000000001e-06, "loss": 1.0819, "step": 500 }, { "grad_norm": 2.0561633110046387, "learning_rate": 1.018e-05, "loss": 1.0654, "step": 510 }, { "grad_norm": 2.2519619464874268, "learning_rate": 1.038e-05, "loss": 1.0609, "step": 520 }, { "grad_norm": 1.4152653217315674, "learning_rate": 1.058e-05, "loss": 1.0935, "step": 530 }, { "grad_norm": 1.8126860857009888, "learning_rate": 1.0780000000000002e-05, "loss": 1.0593, "step": 540 }, { "grad_norm": 1.5609835386276245, "learning_rate": 1.098e-05, "loss": 1.0762, "step": 550 }, { "grad_norm": 1.3712763786315918, "learning_rate": 1.118e-05, "loss": 1.0478, "step": 560 }, { "grad_norm": 1.5414834022521973, "learning_rate": 1.1380000000000001e-05, "loss": 0.9941, "step": 570 }, { "grad_norm": 1.8785865306854248, "learning_rate": 1.1580000000000001e-05, "loss": 0.9585, "step": 580 }, { "grad_norm": 2.1582157611846924, "learning_rate": 1.178e-05, "loss": 0.9465, "step": 590 }, { "grad_norm": 2.3648576736450195, "learning_rate": 1.198e-05, "loss": 0.9058, "step": 600 }, { "grad_norm": 2.437617778778076, "learning_rate": 1.2180000000000002e-05, "loss": 0.8774, "step": 610 }, { "grad_norm": 2.9582254886627197, "learning_rate": 1.238e-05, "loss": 0.8539, "step": 620 }, { "grad_norm": 1.9871716499328613, "learning_rate": 1.258e-05, "loss": 0.8419, "step": 630 }, { "grad_norm": 2.328138589859009, "learning_rate": 1.278e-05, "loss": 0.8126, "step": 640 }, { "grad_norm": 2.481762409210205, "learning_rate": 1.2980000000000001e-05, "loss": 0.7653, "step": 650 }, { "grad_norm": 2.761993169784546, "learning_rate": 1.3180000000000001e-05, "loss": 0.7325, "step": 660 }, { "grad_norm": 1.872991681098938, "learning_rate": 1.338e-05, "loss": 0.6793, "step": 670 }, { "grad_norm": 2.8386611938476562, "learning_rate": 1.358e-05, "loss": 0.6593, "step": 680 }, { "grad_norm": 2.2590830326080322, "learning_rate": 1.3780000000000002e-05, "loss": 0.6454, "step": 690 }, { "grad_norm": 3.323244094848633, "learning_rate": 1.3980000000000002e-05, "loss": 0.638, "step": 700 }, { "grad_norm": 2.5417678356170654, "learning_rate": 1.4180000000000001e-05, "loss": 0.5546, "step": 710 }, { "grad_norm": 3.292449951171875, "learning_rate": 1.4380000000000001e-05, "loss": 0.5752, "step": 720 }, { "grad_norm": 2.933755397796631, "learning_rate": 1.4580000000000003e-05, "loss": 0.515, "step": 730 }, { "grad_norm": 2.1128907203674316, "learning_rate": 1.4779999999999999e-05, "loss": 0.4905, "step": 740 }, { "grad_norm": 3.4277617931365967, "learning_rate": 1.4979999999999999e-05, "loss": 0.4811, "step": 750 }, { "grad_norm": 2.1755480766296387, "learning_rate": 1.518e-05, "loss": 0.4444, "step": 760 }, { "grad_norm": 2.485243797302246, "learning_rate": 1.538e-05, "loss": 0.4289, "step": 770 }, { "grad_norm": 2.7109596729278564, "learning_rate": 1.558e-05, "loss": 0.4366, "step": 780 }, { "grad_norm": 2.4366066455841064, "learning_rate": 1.578e-05, "loss": 0.3658, "step": 790 }, { "grad_norm": 2.5367727279663086, "learning_rate": 1.598e-05, "loss": 0.3598, "step": 800 }, { "grad_norm": 2.9604592323303223, "learning_rate": 1.618e-05, "loss": 0.3748, "step": 810 }, { "grad_norm": 2.3376221656799316, "learning_rate": 1.6380000000000002e-05, "loss": 0.3589, "step": 820 }, { "grad_norm": 2.6434285640716553, "learning_rate": 1.658e-05, "loss": 0.3572, "step": 830 }, { "grad_norm": 2.9546821117401123, "learning_rate": 1.6780000000000002e-05, "loss": 0.339, "step": 840 }, { "grad_norm": 2.281217575073242, "learning_rate": 1.698e-05, "loss": 0.3599, "step": 850 }, { "grad_norm": 2.1436195373535156, "learning_rate": 1.718e-05, "loss": 0.3187, "step": 860 }, { "grad_norm": 3.056840658187866, "learning_rate": 1.7380000000000003e-05, "loss": 0.3296, "step": 870 }, { "grad_norm": 3.054232120513916, "learning_rate": 1.758e-05, "loss": 0.2827, "step": 880 }, { "grad_norm": 2.473891019821167, "learning_rate": 1.7780000000000003e-05, "loss": 0.2887, "step": 890 }, { "grad_norm": 2.5847301483154297, "learning_rate": 1.798e-05, "loss": 0.2814, "step": 900 }, { "grad_norm": 2.000833034515381, "learning_rate": 1.818e-05, "loss": 0.2655, "step": 910 }, { "grad_norm": 2.5139248371124268, "learning_rate": 1.838e-05, "loss": 0.2256, "step": 920 }, { "grad_norm": 2.6325159072875977, "learning_rate": 1.858e-05, "loss": 0.2527, "step": 930 }, { "grad_norm": 2.6237545013427734, "learning_rate": 1.878e-05, "loss": 0.2314, "step": 940 }, { "grad_norm": 2.789783239364624, "learning_rate": 1.898e-05, "loss": 0.2164, "step": 950 }, { "grad_norm": 2.3315327167510986, "learning_rate": 1.918e-05, "loss": 0.2114, "step": 960 }, { "grad_norm": 1.858264446258545, "learning_rate": 1.938e-05, "loss": 0.2509, "step": 970 }, { "grad_norm": 2.451533555984497, "learning_rate": 1.9580000000000002e-05, "loss": 0.2126, "step": 980 }, { "grad_norm": 2.5016496181488037, "learning_rate": 1.978e-05, "loss": 0.1881, "step": 990 }, { "grad_norm": 2.847114086151123, "learning_rate": 1.9980000000000002e-05, "loss": 0.2037, "step": 1000 }, { "grad_norm": 2.188105344772339, "learning_rate": 2.0180000000000003e-05, "loss": 0.1964, "step": 1010 }, { "grad_norm": 2.5337986946105957, "learning_rate": 2.038e-05, "loss": 0.2372, "step": 1020 }, { "grad_norm": 2.505978584289551, "learning_rate": 2.0580000000000003e-05, "loss": 0.1841, "step": 1030 }, { "grad_norm": 2.1607985496520996, "learning_rate": 2.078e-05, "loss": 0.1967, "step": 1040 }, { "grad_norm": 3.6994681358337402, "learning_rate": 2.098e-05, "loss": 0.1981, "step": 1050 }, { "grad_norm": 1.9215929508209229, "learning_rate": 2.118e-05, "loss": 0.2155, "step": 1060 }, { "grad_norm": 2.6162185668945312, "learning_rate": 2.138e-05, "loss": 0.231, "step": 1070 }, { "grad_norm": 3.580110788345337, "learning_rate": 2.158e-05, "loss": 0.2179, "step": 1080 }, { "grad_norm": 2.1047720909118652, "learning_rate": 2.178e-05, "loss": 0.1998, "step": 1090 }, { "grad_norm": 2.305633544921875, "learning_rate": 2.198e-05, "loss": 0.156, "step": 1100 }, { "grad_norm": 2.3130409717559814, "learning_rate": 2.218e-05, "loss": 0.2292, "step": 1110 }, { "grad_norm": 1.8870161771774292, "learning_rate": 2.2380000000000003e-05, "loss": 0.1541, "step": 1120 }, { "grad_norm": 1.9240559339523315, "learning_rate": 2.258e-05, "loss": 0.1824, "step": 1130 }, { "grad_norm": 2.9124677181243896, "learning_rate": 2.2780000000000002e-05, "loss": 0.241, "step": 1140 }, { "grad_norm": 2.731112480163574, "learning_rate": 2.298e-05, "loss": 0.2183, "step": 1150 }, { "grad_norm": 1.986678123474121, "learning_rate": 2.318e-05, "loss": 0.1891, "step": 1160 }, { "grad_norm": 2.21352481842041, "learning_rate": 2.3380000000000003e-05, "loss": 0.1823, "step": 1170 }, { "grad_norm": 2.116909980773926, "learning_rate": 2.358e-05, "loss": 0.1658, "step": 1180 }, { "grad_norm": 1.9297009706497192, "learning_rate": 2.3780000000000003e-05, "loss": 0.1798, "step": 1190 }, { "grad_norm": 9.641035079956055, "learning_rate": 2.398e-05, "loss": 0.1965, "step": 1200 }, { "grad_norm": 3.3854033946990967, "learning_rate": 2.418e-05, "loss": 0.203, "step": 1210 }, { "grad_norm": 2.419337272644043, "learning_rate": 2.438e-05, "loss": 0.1891, "step": 1220 }, { "grad_norm": 2.5479044914245605, "learning_rate": 2.4580000000000002e-05, "loss": 0.1939, "step": 1230 }, { "grad_norm": 2.054959297180176, "learning_rate": 2.478e-05, "loss": 0.1988, "step": 1240 }, { "grad_norm": 2.409252405166626, "learning_rate": 2.498e-05, "loss": 0.1649, "step": 1250 }, { "grad_norm": 2.1842033863067627, "learning_rate": 2.5180000000000003e-05, "loss": 0.1677, "step": 1260 }, { "grad_norm": 1.9974092245101929, "learning_rate": 2.5380000000000004e-05, "loss": 0.1459, "step": 1270 }, { "grad_norm": 1.9214391708374023, "learning_rate": 2.5580000000000002e-05, "loss": 0.1763, "step": 1280 }, { "grad_norm": 1.9258081912994385, "learning_rate": 2.5779999999999997e-05, "loss": 0.1838, "step": 1290 }, { "grad_norm": 1.6528453826904297, "learning_rate": 2.598e-05, "loss": 0.1538, "step": 1300 }, { "grad_norm": 1.8944857120513916, "learning_rate": 2.618e-05, "loss": 0.1509, "step": 1310 }, { "grad_norm": 1.7308353185653687, "learning_rate": 2.6379999999999998e-05, "loss": 0.1621, "step": 1320 }, { "grad_norm": 1.8609681129455566, "learning_rate": 2.658e-05, "loss": 0.1479, "step": 1330 }, { "grad_norm": 2.0977182388305664, "learning_rate": 2.678e-05, "loss": 0.1446, "step": 1340 }, { "grad_norm": 1.779240369796753, "learning_rate": 2.698e-05, "loss": 0.1814, "step": 1350 }, { "grad_norm": 2.356377124786377, "learning_rate": 2.718e-05, "loss": 0.1806, "step": 1360 }, { "grad_norm": 2.14871883392334, "learning_rate": 2.738e-05, "loss": 0.1393, "step": 1370 }, { "grad_norm": 2.0160934925079346, "learning_rate": 2.758e-05, "loss": 0.1401, "step": 1380 }, { "grad_norm": 2.163691759109497, "learning_rate": 2.778e-05, "loss": 0.1608, "step": 1390 }, { "grad_norm": 1.9248416423797607, "learning_rate": 2.798e-05, "loss": 0.1485, "step": 1400 }, { "grad_norm": 1.788889765739441, "learning_rate": 2.818e-05, "loss": 0.1508, "step": 1410 }, { "grad_norm": 2.119739294052124, "learning_rate": 2.8380000000000003e-05, "loss": 0.1578, "step": 1420 }, { "grad_norm": 2.4013402462005615, "learning_rate": 2.858e-05, "loss": 0.1424, "step": 1430 }, { "grad_norm": 1.543299674987793, "learning_rate": 2.8780000000000002e-05, "loss": 0.1614, "step": 1440 }, { "grad_norm": 1.5483407974243164, "learning_rate": 2.898e-05, "loss": 0.1757, "step": 1450 }, { "grad_norm": 1.533655047416687, "learning_rate": 2.9180000000000002e-05, "loss": 0.1667, "step": 1460 }, { "grad_norm": 1.4807697534561157, "learning_rate": 2.9380000000000003e-05, "loss": 0.1727, "step": 1470 }, { "grad_norm": 1.4966527223587036, "learning_rate": 2.958e-05, "loss": 0.1435, "step": 1480 }, { "grad_norm": 1.6605730056762695, "learning_rate": 2.9780000000000003e-05, "loss": 0.1396, "step": 1490 }, { "grad_norm": 1.4634219408035278, "learning_rate": 2.998e-05, "loss": 0.1486, "step": 1500 }, { "grad_norm": 2.4116110801696777, "learning_rate": 3.0180000000000002e-05, "loss": 0.1419, "step": 1510 }, { "grad_norm": 2.2858731746673584, "learning_rate": 3.0380000000000004e-05, "loss": 0.1501, "step": 1520 }, { "grad_norm": 1.559465765953064, "learning_rate": 3.058e-05, "loss": 0.1484, "step": 1530 }, { "grad_norm": 1.5473023653030396, "learning_rate": 3.078e-05, "loss": 0.1532, "step": 1540 }, { "grad_norm": 1.3645858764648438, "learning_rate": 3.0980000000000005e-05, "loss": 0.1432, "step": 1550 }, { "grad_norm": 1.457279920578003, "learning_rate": 3.118e-05, "loss": 0.1656, "step": 1560 }, { "grad_norm": 1.6988894939422607, "learning_rate": 3.138e-05, "loss": 0.1297, "step": 1570 }, { "grad_norm": 1.9253556728363037, "learning_rate": 3.1580000000000006e-05, "loss": 0.149, "step": 1580 }, { "grad_norm": 1.7245028018951416, "learning_rate": 3.1780000000000004e-05, "loss": 0.1624, "step": 1590 }, { "grad_norm": 1.219441294670105, "learning_rate": 3.198e-05, "loss": 0.1312, "step": 1600 }, { "grad_norm": 1.5258797407150269, "learning_rate": 3.218e-05, "loss": 0.1234, "step": 1610 }, { "grad_norm": 1.477684497833252, "learning_rate": 3.238e-05, "loss": 0.1478, "step": 1620 }, { "grad_norm": 1.5087471008300781, "learning_rate": 3.2579999999999996e-05, "loss": 0.1776, "step": 1630 }, { "grad_norm": 1.7488338947296143, "learning_rate": 3.278e-05, "loss": 0.1512, "step": 1640 }, { "grad_norm": 1.7127940654754639, "learning_rate": 3.298e-05, "loss": 0.1545, "step": 1650 }, { "grad_norm": 1.569908618927002, "learning_rate": 3.318e-05, "loss": 0.149, "step": 1660 }, { "grad_norm": 1.5234943628311157, "learning_rate": 3.338e-05, "loss": 0.1215, "step": 1670 }, { "grad_norm": 1.9825690984725952, "learning_rate": 3.358e-05, "loss": 0.1324, "step": 1680 }, { "grad_norm": 1.1418622732162476, "learning_rate": 3.378e-05, "loss": 0.1388, "step": 1690 }, { "grad_norm": 0.9220991730690002, "learning_rate": 3.398e-05, "loss": 0.139, "step": 1700 }, { "grad_norm": 2.008700370788574, "learning_rate": 3.418e-05, "loss": 0.1624, "step": 1710 }, { "grad_norm": 1.557051420211792, "learning_rate": 3.438e-05, "loss": 0.1477, "step": 1720 }, { "grad_norm": 1.3777308464050293, "learning_rate": 3.4580000000000004e-05, "loss": 0.1411, "step": 1730 }, { "grad_norm": 1.7419958114624023, "learning_rate": 3.478e-05, "loss": 0.1266, "step": 1740 }, { "grad_norm": 2.0528554916381836, "learning_rate": 3.498e-05, "loss": 0.1464, "step": 1750 }, { "grad_norm": 1.9027354717254639, "learning_rate": 3.518e-05, "loss": 0.1477, "step": 1760 }, { "grad_norm": 1.2961217164993286, "learning_rate": 3.5380000000000003e-05, "loss": 0.1328, "step": 1770 }, { "grad_norm": 1.3667761087417603, "learning_rate": 3.558e-05, "loss": 0.1518, "step": 1780 }, { "grad_norm": 1.8511933088302612, "learning_rate": 3.578e-05, "loss": 0.1291, "step": 1790 }, { "grad_norm": 1.6485768556594849, "learning_rate": 3.5980000000000004e-05, "loss": 0.1327, "step": 1800 }, { "grad_norm": 1.4093079566955566, "learning_rate": 3.618e-05, "loss": 0.1357, "step": 1810 }, { "grad_norm": 1.7371770143508911, "learning_rate": 3.638e-05, "loss": 0.1467, "step": 1820 }, { "grad_norm": 1.733585238456726, "learning_rate": 3.6580000000000006e-05, "loss": 0.13, "step": 1830 }, { "grad_norm": 1.7813910245895386, "learning_rate": 3.6780000000000004e-05, "loss": 0.1532, "step": 1840 }, { "grad_norm": 1.3753647804260254, "learning_rate": 3.698e-05, "loss": 0.1653, "step": 1850 }, { "grad_norm": 1.3848049640655518, "learning_rate": 3.7180000000000007e-05, "loss": 0.1573, "step": 1860 }, { "grad_norm": 1.002572774887085, "learning_rate": 3.7380000000000005e-05, "loss": 0.1347, "step": 1870 }, { "grad_norm": 1.6885453462600708, "learning_rate": 3.758e-05, "loss": 0.1173, "step": 1880 }, { "grad_norm": 1.3248306512832642, "learning_rate": 3.778000000000001e-05, "loss": 0.1466, "step": 1890 }, { "grad_norm": 1.5058149099349976, "learning_rate": 3.7980000000000006e-05, "loss": 0.1274, "step": 1900 }, { "grad_norm": 2.1105051040649414, "learning_rate": 3.818e-05, "loss": 0.1711, "step": 1910 }, { "grad_norm": 1.6723319292068481, "learning_rate": 3.838e-05, "loss": 0.1542, "step": 1920 }, { "grad_norm": 1.3362632989883423, "learning_rate": 3.858e-05, "loss": 0.1212, "step": 1930 }, { "grad_norm": 1.8298591375350952, "learning_rate": 3.878e-05, "loss": 0.1437, "step": 1940 }, { "grad_norm": 1.2433574199676514, "learning_rate": 3.898e-05, "loss": 0.1075, "step": 1950 }, { "grad_norm": 1.742508053779602, "learning_rate": 3.918e-05, "loss": 0.1564, "step": 1960 }, { "grad_norm": 1.3012913465499878, "learning_rate": 3.938e-05, "loss": 0.1042, "step": 1970 }, { "grad_norm": 1.3433148860931396, "learning_rate": 3.958e-05, "loss": 0.1512, "step": 1980 }, { "grad_norm": 1.274713158607483, "learning_rate": 3.978e-05, "loss": 0.1146, "step": 1990 }, { "grad_norm": 1.4132325649261475, "learning_rate": 3.998e-05, "loss": 0.142, "step": 2000 }, { "grad_norm": 1.4561810493469238, "learning_rate": 4.018e-05, "loss": 0.1701, "step": 2010 }, { "grad_norm": 1.6718355417251587, "learning_rate": 4.038e-05, "loss": 0.1335, "step": 2020 }, { "grad_norm": 1.5752105712890625, "learning_rate": 4.058e-05, "loss": 0.1276, "step": 2030 }, { "grad_norm": 1.7042597532272339, "learning_rate": 4.078e-05, "loss": 0.1436, "step": 2040 }, { "grad_norm": 1.9708384275436401, "learning_rate": 4.0980000000000004e-05, "loss": 0.1448, "step": 2050 }, { "grad_norm": 1.3334574699401855, "learning_rate": 4.118e-05, "loss": 0.1326, "step": 2060 }, { "grad_norm": 1.401942253112793, "learning_rate": 4.138e-05, "loss": 0.1254, "step": 2070 }, { "grad_norm": 1.5671769380569458, "learning_rate": 4.1580000000000005e-05, "loss": 0.1315, "step": 2080 }, { "grad_norm": 1.1444892883300781, "learning_rate": 4.178e-05, "loss": 0.1405, "step": 2090 }, { "grad_norm": 1.0913183689117432, "learning_rate": 4.198e-05, "loss": 0.1274, "step": 2100 }, { "grad_norm": 1.5260156393051147, "learning_rate": 4.2180000000000006e-05, "loss": 0.1606, "step": 2110 }, { "grad_norm": 1.0564533472061157, "learning_rate": 4.2380000000000004e-05, "loss": 0.1218, "step": 2120 }, { "grad_norm": 1.236493706703186, "learning_rate": 4.258e-05, "loss": 0.123, "step": 2130 }, { "grad_norm": 1.0953059196472168, "learning_rate": 4.278e-05, "loss": 0.1233, "step": 2140 }, { "grad_norm": 1.3317339420318604, "learning_rate": 4.2980000000000005e-05, "loss": 0.1491, "step": 2150 }, { "grad_norm": 1.3660690784454346, "learning_rate": 4.318e-05, "loss": 0.1344, "step": 2160 }, { "grad_norm": 1.9389044046401978, "learning_rate": 4.338e-05, "loss": 0.1313, "step": 2170 }, { "grad_norm": 1.3334678411483765, "learning_rate": 4.3580000000000006e-05, "loss": 0.1514, "step": 2180 }, { "grad_norm": 1.261315941810608, "learning_rate": 4.3780000000000004e-05, "loss": 0.13, "step": 2190 }, { "grad_norm": 1.2376765012741089, "learning_rate": 4.398e-05, "loss": 0.1194, "step": 2200 }, { "grad_norm": 1.324395775794983, "learning_rate": 4.418000000000001e-05, "loss": 0.1258, "step": 2210 }, { "grad_norm": 1.4221713542938232, "learning_rate": 4.438e-05, "loss": 0.1248, "step": 2220 }, { "grad_norm": 1.2067327499389648, "learning_rate": 4.458e-05, "loss": 0.1249, "step": 2230 }, { "grad_norm": 0.933070182800293, "learning_rate": 4.478e-05, "loss": 0.1271, "step": 2240 }, { "grad_norm": 1.217827320098877, "learning_rate": 4.498e-05, "loss": 0.1277, "step": 2250 }, { "grad_norm": 0.9654669165611267, "learning_rate": 4.518e-05, "loss": 0.1352, "step": 2260 }, { "grad_norm": 1.2021384239196777, "learning_rate": 4.538e-05, "loss": 0.1185, "step": 2270 }, { "grad_norm": 1.1509746313095093, "learning_rate": 4.558e-05, "loss": 0.1179, "step": 2280 }, { "grad_norm": 1.8655121326446533, "learning_rate": 4.578e-05, "loss": 0.1662, "step": 2290 }, { "grad_norm": 1.1867153644561768, "learning_rate": 4.5980000000000004e-05, "loss": 0.1209, "step": 2300 }, { "grad_norm": 1.2002379894256592, "learning_rate": 4.618e-05, "loss": 0.1267, "step": 2310 }, { "grad_norm": 1.504643440246582, "learning_rate": 4.638e-05, "loss": 0.1174, "step": 2320 }, { "grad_norm": 1.1360352039337158, "learning_rate": 4.6580000000000005e-05, "loss": 0.1518, "step": 2330 }, { "grad_norm": 1.3510950803756714, "learning_rate": 4.678e-05, "loss": 0.1298, "step": 2340 }, { "grad_norm": 1.2953625917434692, "learning_rate": 4.698e-05, "loss": 0.1282, "step": 2350 }, { "grad_norm": 1.215833306312561, "learning_rate": 4.718e-05, "loss": 0.1366, "step": 2360 }, { "grad_norm": 1.0337731838226318, "learning_rate": 4.7380000000000004e-05, "loss": 0.1244, "step": 2370 }, { "grad_norm": 1.8429639339447021, "learning_rate": 4.758e-05, "loss": 0.1412, "step": 2380 }, { "grad_norm": 1.1169270277023315, "learning_rate": 4.778e-05, "loss": 0.1422, "step": 2390 }, { "grad_norm": 1.2038779258728027, "learning_rate": 4.7980000000000005e-05, "loss": 0.1261, "step": 2400 }, { "grad_norm": 1.3472580909729004, "learning_rate": 4.818e-05, "loss": 0.1252, "step": 2410 }, { "grad_norm": 1.07061767578125, "learning_rate": 4.838e-05, "loss": 0.1719, "step": 2420 }, { "grad_norm": 1.51702082157135, "learning_rate": 4.8580000000000006e-05, "loss": 0.1339, "step": 2430 }, { "grad_norm": 1.3928077220916748, "learning_rate": 4.8780000000000004e-05, "loss": 0.1412, "step": 2440 }, { "grad_norm": 1.3258342742919922, "learning_rate": 4.898e-05, "loss": 0.117, "step": 2450 }, { "grad_norm": 0.9601233601570129, "learning_rate": 4.918000000000001e-05, "loss": 0.1175, "step": 2460 }, { "grad_norm": 1.325241208076477, "learning_rate": 4.9380000000000005e-05, "loss": 0.1152, "step": 2470 }, { "grad_norm": 0.88174968957901, "learning_rate": 4.958e-05, "loss": 0.1398, "step": 2480 }, { "grad_norm": 1.4697591066360474, "learning_rate": 4.978e-05, "loss": 0.1247, "step": 2490 }, { "grad_norm": 0.7889934182167053, "learning_rate": 4.9980000000000006e-05, "loss": 0.1108, "step": 2500 }, { "grad_norm": 0.9123203158378601, "learning_rate": 5.0180000000000004e-05, "loss": 0.1137, "step": 2510 }, { "grad_norm": 0.8855170011520386, "learning_rate": 5.038e-05, "loss": 0.122, "step": 2520 }, { "grad_norm": 1.3609923124313354, "learning_rate": 5.058000000000001e-05, "loss": 0.1153, "step": 2530 }, { "grad_norm": 1.277131199836731, "learning_rate": 5.0780000000000005e-05, "loss": 0.1255, "step": 2540 }, { "grad_norm": 1.0363249778747559, "learning_rate": 5.098e-05, "loss": 0.1456, "step": 2550 }, { "grad_norm": 1.0871646404266357, "learning_rate": 5.118000000000001e-05, "loss": 0.1277, "step": 2560 }, { "grad_norm": 1.1360929012298584, "learning_rate": 5.1380000000000006e-05, "loss": 0.1178, "step": 2570 }, { "grad_norm": 1.7696651220321655, "learning_rate": 5.1580000000000004e-05, "loss": 0.1379, "step": 2580 }, { "grad_norm": 1.0277773141860962, "learning_rate": 5.178000000000001e-05, "loss": 0.1182, "step": 2590 }, { "grad_norm": 1.0740998983383179, "learning_rate": 5.198000000000001e-05, "loss": 0.1399, "step": 2600 }, { "grad_norm": 1.0056040287017822, "learning_rate": 5.2180000000000005e-05, "loss": 0.1246, "step": 2610 }, { "grad_norm": 1.1597778797149658, "learning_rate": 5.238000000000001e-05, "loss": 0.1142, "step": 2620 }, { "grad_norm": 1.0593783855438232, "learning_rate": 5.258000000000001e-05, "loss": 0.1142, "step": 2630 }, { "grad_norm": 1.150016188621521, "learning_rate": 5.2780000000000006e-05, "loss": 0.1438, "step": 2640 }, { "grad_norm": 1.106428861618042, "learning_rate": 5.2980000000000004e-05, "loss": 0.1319, "step": 2650 }, { "grad_norm": 0.9560917019844055, "learning_rate": 5.318000000000001e-05, "loss": 0.1279, "step": 2660 }, { "grad_norm": 1.0661330223083496, "learning_rate": 5.338000000000001e-05, "loss": 0.1494, "step": 2670 }, { "grad_norm": 1.4195537567138672, "learning_rate": 5.3580000000000005e-05, "loss": 0.1625, "step": 2680 }, { "grad_norm": 1.0601781606674194, "learning_rate": 5.378e-05, "loss": 0.1316, "step": 2690 }, { "grad_norm": 1.0715622901916504, "learning_rate": 5.3979999999999995e-05, "loss": 0.1131, "step": 2700 }, { "grad_norm": 0.7865069508552551, "learning_rate": 5.418e-05, "loss": 0.121, "step": 2710 }, { "grad_norm": 0.9128389358520508, "learning_rate": 5.438e-05, "loss": 0.1436, "step": 2720 }, { "grad_norm": 1.0770716667175293, "learning_rate": 5.4579999999999996e-05, "loss": 0.118, "step": 2730 }, { "grad_norm": 0.9192811250686646, "learning_rate": 5.478e-05, "loss": 0.1123, "step": 2740 }, { "grad_norm": 1.1102886199951172, "learning_rate": 5.498e-05, "loss": 0.1384, "step": 2750 }, { "grad_norm": 1.0491629838943481, "learning_rate": 5.518e-05, "loss": 0.1239, "step": 2760 }, { "grad_norm": 0.9397494196891785, "learning_rate": 5.538e-05, "loss": 0.1202, "step": 2770 }, { "grad_norm": 0.8657879829406738, "learning_rate": 5.558e-05, "loss": 0.1274, "step": 2780 }, { "grad_norm": 0.7999303936958313, "learning_rate": 5.578e-05, "loss": 0.1208, "step": 2790 }, { "grad_norm": 0.9641468524932861, "learning_rate": 5.5979999999999996e-05, "loss": 0.1113, "step": 2800 }, { "grad_norm": 1.0564862489700317, "learning_rate": 5.618e-05, "loss": 0.1568, "step": 2810 }, { "grad_norm": 1.1044371128082275, "learning_rate": 5.638e-05, "loss": 0.1172, "step": 2820 }, { "grad_norm": 1.305515170097351, "learning_rate": 5.658e-05, "loss": 0.1164, "step": 2830 }, { "grad_norm": 0.8982599377632141, "learning_rate": 5.678e-05, "loss": 0.1508, "step": 2840 }, { "grad_norm": 0.8992084860801697, "learning_rate": 5.698e-05, "loss": 0.1266, "step": 2850 }, { "grad_norm": 1.1611170768737793, "learning_rate": 5.718e-05, "loss": 0.1242, "step": 2860 }, { "grad_norm": 0.9200268983840942, "learning_rate": 5.738e-05, "loss": 0.1136, "step": 2870 }, { "grad_norm": 0.9439517259597778, "learning_rate": 5.758e-05, "loss": 0.1105, "step": 2880 }, { "grad_norm": 0.7876362800598145, "learning_rate": 5.778e-05, "loss": 0.1197, "step": 2890 }, { "grad_norm": 0.9426207542419434, "learning_rate": 5.7980000000000004e-05, "loss": 0.1249, "step": 2900 }, { "grad_norm": 1.2086584568023682, "learning_rate": 5.818e-05, "loss": 0.1115, "step": 2910 }, { "grad_norm": 0.8224611878395081, "learning_rate": 5.838e-05, "loss": 0.1229, "step": 2920 }, { "grad_norm": 0.9105862975120544, "learning_rate": 5.858e-05, "loss": 0.1013, "step": 2930 }, { "grad_norm": 1.0447237491607666, "learning_rate": 5.878e-05, "loss": 0.1059, "step": 2940 }, { "grad_norm": 0.7882480621337891, "learning_rate": 5.898e-05, "loss": 0.0985, "step": 2950 }, { "grad_norm": 0.783225417137146, "learning_rate": 5.918e-05, "loss": 0.1114, "step": 2960 }, { "grad_norm": 0.9121639132499695, "learning_rate": 5.9380000000000004e-05, "loss": 0.101, "step": 2970 }, { "grad_norm": 0.9084203839302063, "learning_rate": 5.958e-05, "loss": 0.1201, "step": 2980 }, { "grad_norm": 1.312212347984314, "learning_rate": 5.978e-05, "loss": 0.1426, "step": 2990 }, { "grad_norm": 0.7309980988502502, "learning_rate": 5.9980000000000005e-05, "loss": 0.1205, "step": 3000 }, { "grad_norm": 0.6957531571388245, "learning_rate": 6.018e-05, "loss": 0.0973, "step": 3010 }, { "grad_norm": 0.7199123501777649, "learning_rate": 6.038e-05, "loss": 0.107, "step": 3020 }, { "grad_norm": 1.1250040531158447, "learning_rate": 6.0580000000000006e-05, "loss": 0.1171, "step": 3030 }, { "grad_norm": 1.2636682987213135, "learning_rate": 6.0780000000000004e-05, "loss": 0.1155, "step": 3040 }, { "grad_norm": 0.9163708686828613, "learning_rate": 6.098e-05, "loss": 0.1239, "step": 3050 }, { "grad_norm": 1.1699362993240356, "learning_rate": 6.118000000000001e-05, "loss": 0.1387, "step": 3060 }, { "grad_norm": 0.9515336751937866, "learning_rate": 6.138e-05, "loss": 0.1059, "step": 3070 }, { "grad_norm": 0.7265612483024597, "learning_rate": 6.158e-05, "loss": 0.1232, "step": 3080 }, { "grad_norm": 1.3251049518585205, "learning_rate": 6.178000000000001e-05, "loss": 0.1337, "step": 3090 }, { "grad_norm": 1.109589695930481, "learning_rate": 6.198e-05, "loss": 0.108, "step": 3100 }, { "grad_norm": 0.6747388243675232, "learning_rate": 6.218e-05, "loss": 0.1026, "step": 3110 }, { "grad_norm": 0.8076696395874023, "learning_rate": 6.238000000000001e-05, "loss": 0.1006, "step": 3120 }, { "grad_norm": 1.2380974292755127, "learning_rate": 6.258e-05, "loss": 0.1072, "step": 3130 }, { "grad_norm": 0.8874322175979614, "learning_rate": 6.278e-05, "loss": 0.1143, "step": 3140 }, { "grad_norm": 1.181980848312378, "learning_rate": 6.298000000000001e-05, "loss": 0.1191, "step": 3150 }, { "grad_norm": 0.8712249994277954, "learning_rate": 6.318e-05, "loss": 0.1215, "step": 3160 }, { "grad_norm": 0.9881347417831421, "learning_rate": 6.338e-05, "loss": 0.1035, "step": 3170 }, { "grad_norm": 0.8854615688323975, "learning_rate": 6.358000000000001e-05, "loss": 0.1252, "step": 3180 }, { "grad_norm": 0.8469372987747192, "learning_rate": 6.378e-05, "loss": 0.1232, "step": 3190 }, { "grad_norm": 0.862534761428833, "learning_rate": 6.398000000000001e-05, "loss": 0.1156, "step": 3200 }, { "grad_norm": 0.8336877226829529, "learning_rate": 6.418000000000001e-05, "loss": 0.0978, "step": 3210 }, { "grad_norm": 0.9746348261833191, "learning_rate": 6.438e-05, "loss": 0.1366, "step": 3220 }, { "grad_norm": 0.8100742101669312, "learning_rate": 6.458000000000001e-05, "loss": 0.1139, "step": 3230 }, { "grad_norm": 0.7509434819221497, "learning_rate": 6.478000000000001e-05, "loss": 0.0992, "step": 3240 }, { "grad_norm": 0.6377643346786499, "learning_rate": 6.498e-05, "loss": 0.1085, "step": 3250 }, { "grad_norm": 1.2753113508224487, "learning_rate": 6.518000000000001e-05, "loss": 0.109, "step": 3260 }, { "grad_norm": 0.9938151836395264, "learning_rate": 6.538000000000001e-05, "loss": 0.1195, "step": 3270 }, { "grad_norm": 0.8599580526351929, "learning_rate": 6.558e-05, "loss": 0.0995, "step": 3280 }, { "grad_norm": 0.8528472781181335, "learning_rate": 6.578000000000001e-05, "loss": 0.1039, "step": 3290 }, { "grad_norm": 0.7089157104492188, "learning_rate": 6.598e-05, "loss": 0.1134, "step": 3300 }, { "grad_norm": 0.9560403227806091, "learning_rate": 6.618e-05, "loss": 0.1131, "step": 3310 }, { "grad_norm": 0.6209797263145447, "learning_rate": 6.638e-05, "loss": 0.1276, "step": 3320 }, { "grad_norm": 0.8128027319908142, "learning_rate": 6.658e-05, "loss": 0.1185, "step": 3330 }, { "grad_norm": 1.1302595138549805, "learning_rate": 6.678e-05, "loss": 0.1165, "step": 3340 }, { "grad_norm": 0.7956356406211853, "learning_rate": 6.698e-05, "loss": 0.1156, "step": 3350 }, { "grad_norm": 0.9217714071273804, "learning_rate": 6.718e-05, "loss": 0.1004, "step": 3360 }, { "grad_norm": 1.088843822479248, "learning_rate": 6.738e-05, "loss": 0.1107, "step": 3370 }, { "grad_norm": 0.9883176684379578, "learning_rate": 6.758e-05, "loss": 0.1634, "step": 3380 }, { "grad_norm": 1.0490084886550903, "learning_rate": 6.778e-05, "loss": 0.1088, "step": 3390 }, { "grad_norm": 0.7950611710548401, "learning_rate": 6.798e-05, "loss": 0.1131, "step": 3400 }, { "grad_norm": 1.1027005910873413, "learning_rate": 6.818e-05, "loss": 0.1125, "step": 3410 }, { "grad_norm": 0.9205275177955627, "learning_rate": 6.838e-05, "loss": 0.1032, "step": 3420 }, { "grad_norm": 0.6042933464050293, "learning_rate": 6.858e-05, "loss": 0.1093, "step": 3430 }, { "grad_norm": 0.7574987411499023, "learning_rate": 6.878e-05, "loss": 0.1288, "step": 3440 }, { "grad_norm": 0.9758443236351013, "learning_rate": 6.898e-05, "loss": 0.1089, "step": 3450 }, { "grad_norm": 0.6476743221282959, "learning_rate": 6.918e-05, "loss": 0.1035, "step": 3460 }, { "grad_norm": 0.8403541445732117, "learning_rate": 6.938e-05, "loss": 0.1227, "step": 3470 }, { "grad_norm": 0.9343002438545227, "learning_rate": 6.958e-05, "loss": 0.1139, "step": 3480 }, { "grad_norm": 0.8399502038955688, "learning_rate": 6.978e-05, "loss": 0.1145, "step": 3490 }, { "grad_norm": 0.8118539452552795, "learning_rate": 6.998e-05, "loss": 0.0923, "step": 3500 }, { "grad_norm": 0.8658353090286255, "learning_rate": 7.018e-05, "loss": 0.114, "step": 3510 }, { "grad_norm": 1.122899055480957, "learning_rate": 7.038e-05, "loss": 0.1036, "step": 3520 }, { "grad_norm": 0.9801620841026306, "learning_rate": 7.058e-05, "loss": 0.1099, "step": 3530 }, { "grad_norm": 0.8843816518783569, "learning_rate": 7.078e-05, "loss": 0.1071, "step": 3540 }, { "grad_norm": 0.7956583499908447, "learning_rate": 7.098e-05, "loss": 0.1157, "step": 3550 }, { "grad_norm": 0.6460185050964355, "learning_rate": 7.118e-05, "loss": 0.0966, "step": 3560 }, { "grad_norm": 0.8257346749305725, "learning_rate": 7.138e-05, "loss": 0.1216, "step": 3570 }, { "grad_norm": 0.8112382888793945, "learning_rate": 7.158e-05, "loss": 0.1309, "step": 3580 }, { "grad_norm": 0.9386283159255981, "learning_rate": 7.178000000000001e-05, "loss": 0.1202, "step": 3590 }, { "grad_norm": 0.7964998483657837, "learning_rate": 7.198e-05, "loss": 0.1003, "step": 3600 }, { "grad_norm": 0.8369283676147461, "learning_rate": 7.218e-05, "loss": 0.1075, "step": 3610 }, { "grad_norm": 0.8122748136520386, "learning_rate": 7.238000000000001e-05, "loss": 0.1031, "step": 3620 }, { "grad_norm": 0.6992153525352478, "learning_rate": 7.258e-05, "loss": 0.0933, "step": 3630 }, { "grad_norm": 0.6782217025756836, "learning_rate": 7.278e-05, "loss": 0.0979, "step": 3640 }, { "grad_norm": 0.7325298190116882, "learning_rate": 7.298000000000001e-05, "loss": 0.1047, "step": 3650 }, { "grad_norm": 0.6443296670913696, "learning_rate": 7.318e-05, "loss": 0.1082, "step": 3660 }, { "grad_norm": 0.8393664956092834, "learning_rate": 7.338e-05, "loss": 0.1247, "step": 3670 }, { "grad_norm": 0.7159592509269714, "learning_rate": 7.358000000000001e-05, "loss": 0.116, "step": 3680 }, { "grad_norm": 0.8066863417625427, "learning_rate": 7.378e-05, "loss": 0.1053, "step": 3690 }, { "grad_norm": 0.7961257100105286, "learning_rate": 7.398e-05, "loss": 0.1047, "step": 3700 }, { "grad_norm": 0.6610666513442993, "learning_rate": 7.418000000000001e-05, "loss": 0.0958, "step": 3710 }, { "grad_norm": 0.6787974834442139, "learning_rate": 7.438e-05, "loss": 0.1082, "step": 3720 }, { "grad_norm": 0.7137625813484192, "learning_rate": 7.458000000000001e-05, "loss": 0.0898, "step": 3730 }, { "grad_norm": 0.6573916077613831, "learning_rate": 7.478e-05, "loss": 0.1078, "step": 3740 }, { "grad_norm": 0.6974523663520813, "learning_rate": 7.498e-05, "loss": 0.1223, "step": 3750 }, { "grad_norm": 0.6490142345428467, "learning_rate": 7.518000000000001e-05, "loss": 0.0966, "step": 3760 }, { "grad_norm": 0.8600062131881714, "learning_rate": 7.538e-05, "loss": 0.1084, "step": 3770 }, { "grad_norm": 1.0486022233963013, "learning_rate": 7.558e-05, "loss": 0.115, "step": 3780 }, { "grad_norm": 0.7201052308082581, "learning_rate": 7.578000000000001e-05, "loss": 0.1209, "step": 3790 }, { "grad_norm": 0.8976091146469116, "learning_rate": 7.598e-05, "loss": 0.1592, "step": 3800 }, { "grad_norm": 0.556566596031189, "learning_rate": 7.618e-05, "loss": 0.0991, "step": 3810 }, { "grad_norm": 0.9882681369781494, "learning_rate": 7.638000000000001e-05, "loss": 0.1184, "step": 3820 }, { "grad_norm": 0.8187375068664551, "learning_rate": 7.658e-05, "loss": 0.0915, "step": 3830 }, { "grad_norm": 0.6023256778717041, "learning_rate": 7.678000000000001e-05, "loss": 0.0947, "step": 3840 }, { "grad_norm": 1.027752161026001, "learning_rate": 7.698000000000001e-05, "loss": 0.1144, "step": 3850 }, { "grad_norm": 0.7962757349014282, "learning_rate": 7.718e-05, "loss": 0.1102, "step": 3860 }, { "grad_norm": 0.885013222694397, "learning_rate": 7.738000000000001e-05, "loss": 0.0933, "step": 3870 }, { "grad_norm": 0.5872977375984192, "learning_rate": 7.758000000000001e-05, "loss": 0.0942, "step": 3880 }, { "grad_norm": 0.8423817157745361, "learning_rate": 7.778e-05, "loss": 0.1121, "step": 3890 }, { "grad_norm": 0.848065197467804, "learning_rate": 7.798000000000001e-05, "loss": 0.1154, "step": 3900 }, { "grad_norm": 0.7775471210479736, "learning_rate": 7.818000000000001e-05, "loss": 0.0924, "step": 3910 }, { "grad_norm": 0.9784713387489319, "learning_rate": 7.838e-05, "loss": 0.1158, "step": 3920 }, { "grad_norm": 0.6726303100585938, "learning_rate": 7.858000000000001e-05, "loss": 0.0945, "step": 3930 }, { "grad_norm": 0.9509754180908203, "learning_rate": 7.878e-05, "loss": 0.119, "step": 3940 }, { "grad_norm": 0.6215437054634094, "learning_rate": 7.897999999999999e-05, "loss": 0.1106, "step": 3950 }, { "grad_norm": 0.8098585605621338, "learning_rate": 7.918e-05, "loss": 0.1239, "step": 3960 }, { "grad_norm": 0.7471200823783875, "learning_rate": 7.938e-05, "loss": 0.0961, "step": 3970 }, { "grad_norm": 0.7879511713981628, "learning_rate": 7.958e-05, "loss": 0.1095, "step": 3980 }, { "grad_norm": 0.898028552532196, "learning_rate": 7.978e-05, "loss": 0.1092, "step": 3990 }, { "grad_norm": 0.8506356477737427, "learning_rate": 7.998e-05, "loss": 0.1036, "step": 4000 }, { "grad_norm": 0.729374349117279, "learning_rate": 8.018e-05, "loss": 0.0958, "step": 4010 }, { "grad_norm": 0.6232615113258362, "learning_rate": 8.038e-05, "loss": 0.1062, "step": 4020 }, { "grad_norm": 0.7349666357040405, "learning_rate": 8.058e-05, "loss": 0.1138, "step": 4030 }, { "grad_norm": 0.6654584407806396, "learning_rate": 8.078e-05, "loss": 0.0807, "step": 4040 }, { "grad_norm": 0.6293662786483765, "learning_rate": 8.098e-05, "loss": 0.1004, "step": 4050 }, { "grad_norm": 0.7680580019950867, "learning_rate": 8.118e-05, "loss": 0.1002, "step": 4060 }, { "grad_norm": 0.7482550740242004, "learning_rate": 8.138e-05, "loss": 0.0911, "step": 4070 }, { "grad_norm": 0.6892260909080505, "learning_rate": 8.158e-05, "loss": 0.107, "step": 4080 }, { "grad_norm": 0.8873735666275024, "learning_rate": 8.178e-05, "loss": 0.0993, "step": 4090 }, { "grad_norm": 0.9470645785331726, "learning_rate": 8.198e-05, "loss": 0.0956, "step": 4100 }, { "grad_norm": 0.7333166599273682, "learning_rate": 8.218e-05, "loss": 0.111, "step": 4110 }, { "grad_norm": 0.9249190092086792, "learning_rate": 8.238000000000001e-05, "loss": 0.1194, "step": 4120 }, { "grad_norm": 0.6392907500267029, "learning_rate": 8.258e-05, "loss": 0.1008, "step": 4130 }, { "grad_norm": 0.6714558005332947, "learning_rate": 8.278e-05, "loss": 0.1192, "step": 4140 }, { "grad_norm": 0.7249051928520203, "learning_rate": 8.298000000000001e-05, "loss": 0.1165, "step": 4150 }, { "grad_norm": 0.9072371125221252, "learning_rate": 8.318e-05, "loss": 0.1006, "step": 4160 }, { "grad_norm": 0.6944941282272339, "learning_rate": 8.338e-05, "loss": 0.0898, "step": 4170 }, { "grad_norm": 0.7280938029289246, "learning_rate": 8.358e-05, "loss": 0.0894, "step": 4180 }, { "grad_norm": 0.7174059152603149, "learning_rate": 8.378e-05, "loss": 0.0931, "step": 4190 }, { "grad_norm": 0.6008034348487854, "learning_rate": 8.398e-05, "loss": 0.0976, "step": 4200 }, { "grad_norm": 0.6430165767669678, "learning_rate": 8.418e-05, "loss": 0.0932, "step": 4210 }, { "grad_norm": 0.762456476688385, "learning_rate": 8.438e-05, "loss": 0.1097, "step": 4220 }, { "grad_norm": 0.8059826493263245, "learning_rate": 8.458e-05, "loss": 0.1226, "step": 4230 }, { "grad_norm": 0.7096475958824158, "learning_rate": 8.478e-05, "loss": 0.1235, "step": 4240 }, { "grad_norm": 0.708865225315094, "learning_rate": 8.498e-05, "loss": 0.1238, "step": 4250 }, { "grad_norm": 0.8385812044143677, "learning_rate": 8.518000000000001e-05, "loss": 0.1127, "step": 4260 }, { "grad_norm": 0.6186818480491638, "learning_rate": 8.538e-05, "loss": 0.1157, "step": 4270 }, { "grad_norm": 0.6110303401947021, "learning_rate": 8.558e-05, "loss": 0.0875, "step": 4280 }, { "grad_norm": 0.7852970361709595, "learning_rate": 8.578000000000001e-05, "loss": 0.1032, "step": 4290 }, { "grad_norm": 0.818342924118042, "learning_rate": 8.598e-05, "loss": 0.0859, "step": 4300 }, { "grad_norm": 0.7893379330635071, "learning_rate": 8.618e-05, "loss": 0.1019, "step": 4310 }, { "grad_norm": 0.7269447445869446, "learning_rate": 8.638000000000001e-05, "loss": 0.1292, "step": 4320 }, { "grad_norm": 0.6212019324302673, "learning_rate": 8.658e-05, "loss": 0.0971, "step": 4330 }, { "grad_norm": 0.5561820864677429, "learning_rate": 8.678e-05, "loss": 0.0759, "step": 4340 }, { "grad_norm": 0.6992970705032349, "learning_rate": 8.698000000000001e-05, "loss": 0.0903, "step": 4350 }, { "grad_norm": 0.8689740300178528, "learning_rate": 8.718e-05, "loss": 0.0956, "step": 4360 }, { "grad_norm": 0.8502849340438843, "learning_rate": 8.738000000000001e-05, "loss": 0.1035, "step": 4370 }, { "grad_norm": 0.6567395925521851, "learning_rate": 8.758000000000001e-05, "loss": 0.1066, "step": 4380 }, { "grad_norm": 0.8285807371139526, "learning_rate": 8.778e-05, "loss": 0.0899, "step": 4390 }, { "grad_norm": 0.6621524095535278, "learning_rate": 8.798000000000001e-05, "loss": 0.1132, "step": 4400 }, { "grad_norm": 0.8109601736068726, "learning_rate": 8.818000000000001e-05, "loss": 0.1206, "step": 4410 }, { "grad_norm": 0.6152639389038086, "learning_rate": 8.838e-05, "loss": 0.1255, "step": 4420 }, { "grad_norm": 0.7024684548377991, "learning_rate": 8.858000000000001e-05, "loss": 0.1044, "step": 4430 }, { "grad_norm": 0.5558955073356628, "learning_rate": 8.878000000000001e-05, "loss": 0.0995, "step": 4440 }, { "grad_norm": 0.6094262003898621, "learning_rate": 8.898e-05, "loss": 0.1138, "step": 4450 }, { "grad_norm": 0.6093554496765137, "learning_rate": 8.918000000000001e-05, "loss": 0.1008, "step": 4460 }, { "grad_norm": 0.6403636932373047, "learning_rate": 8.938e-05, "loss": 0.0968, "step": 4470 }, { "grad_norm": 0.8425240516662598, "learning_rate": 8.958e-05, "loss": 0.1125, "step": 4480 }, { "grad_norm": 0.6871523261070251, "learning_rate": 8.978000000000001e-05, "loss": 0.0992, "step": 4490 }, { "grad_norm": 0.646405816078186, "learning_rate": 8.998e-05, "loss": 0.1094, "step": 4500 }, { "grad_norm": 0.6913594603538513, "learning_rate": 9.018000000000001e-05, "loss": 0.1211, "step": 4510 }, { "grad_norm": 0.8319408297538757, "learning_rate": 9.038000000000001e-05, "loss": 0.1182, "step": 4520 }, { "grad_norm": 0.6286144852638245, "learning_rate": 9.058e-05, "loss": 0.1014, "step": 4530 }, { "grad_norm": 0.8677932620048523, "learning_rate": 9.078000000000001e-05, "loss": 0.1157, "step": 4540 }, { "grad_norm": 0.6409794092178345, "learning_rate": 9.098000000000001e-05, "loss": 0.1101, "step": 4550 }, { "grad_norm": 0.6159761548042297, "learning_rate": 9.118e-05, "loss": 0.0871, "step": 4560 }, { "grad_norm": 0.5888301730155945, "learning_rate": 9.138e-05, "loss": 0.111, "step": 4570 }, { "grad_norm": 0.9701294898986816, "learning_rate": 9.158e-05, "loss": 0.1017, "step": 4580 }, { "grad_norm": 0.7107503414154053, "learning_rate": 9.178e-05, "loss": 0.0923, "step": 4590 }, { "grad_norm": 0.7877063751220703, "learning_rate": 9.198e-05, "loss": 0.0991, "step": 4600 }, { "grad_norm": 0.6454213857650757, "learning_rate": 9.218e-05, "loss": 0.0934, "step": 4610 }, { "grad_norm": 0.6606343388557434, "learning_rate": 9.238e-05, "loss": 0.0914, "step": 4620 }, { "grad_norm": 0.8301350474357605, "learning_rate": 9.258e-05, "loss": 0.1033, "step": 4630 }, { "grad_norm": 0.6277066469192505, "learning_rate": 9.278e-05, "loss": 0.1079, "step": 4640 }, { "grad_norm": 0.781343936920166, "learning_rate": 9.298e-05, "loss": 0.0781, "step": 4650 }, { "grad_norm": 0.5855537056922913, "learning_rate": 9.318e-05, "loss": 0.103, "step": 4660 }, { "grad_norm": 0.8472692966461182, "learning_rate": 9.338e-05, "loss": 0.095, "step": 4670 }, { "grad_norm": 0.6760597825050354, "learning_rate": 9.358e-05, "loss": 0.0941, "step": 4680 }, { "grad_norm": 0.6736287474632263, "learning_rate": 9.378e-05, "loss": 0.1025, "step": 4690 }, { "grad_norm": 0.7855861783027649, "learning_rate": 9.398e-05, "loss": 0.1013, "step": 4700 }, { "grad_norm": 0.7316515445709229, "learning_rate": 9.418e-05, "loss": 0.1047, "step": 4710 }, { "grad_norm": 0.5546889305114746, "learning_rate": 9.438e-05, "loss": 0.0941, "step": 4720 }, { "grad_norm": 0.7251315116882324, "learning_rate": 9.458e-05, "loss": 0.1129, "step": 4730 }, { "grad_norm": 0.7052092552185059, "learning_rate": 9.478e-05, "loss": 0.1037, "step": 4740 }, { "grad_norm": 0.9338893294334412, "learning_rate": 9.498e-05, "loss": 0.0909, "step": 4750 }, { "grad_norm": 0.9852426648139954, "learning_rate": 9.518000000000001e-05, "loss": 0.1052, "step": 4760 }, { "grad_norm": 1.0033156871795654, "learning_rate": 9.538e-05, "loss": 0.1176, "step": 4770 }, { "grad_norm": 0.7370409965515137, "learning_rate": 9.558e-05, "loss": 0.1137, "step": 4780 }, { "grad_norm": 0.6542633771896362, "learning_rate": 9.578000000000001e-05, "loss": 0.0866, "step": 4790 }, { "grad_norm": 0.6617001295089722, "learning_rate": 9.598e-05, "loss": 0.0995, "step": 4800 }, { "grad_norm": 0.5866040587425232, "learning_rate": 9.618e-05, "loss": 0.1035, "step": 4810 }, { "grad_norm": 0.5972495675086975, "learning_rate": 9.638000000000001e-05, "loss": 0.1123, "step": 4820 }, { "grad_norm": 0.6268007159233093, "learning_rate": 9.658e-05, "loss": 0.1112, "step": 4830 }, { "grad_norm": 0.7164517045021057, "learning_rate": 9.678e-05, "loss": 0.1042, "step": 4840 }, { "grad_norm": 0.8248564600944519, "learning_rate": 9.698000000000001e-05, "loss": 0.1038, "step": 4850 }, { "grad_norm": 0.7356888055801392, "learning_rate": 9.718e-05, "loss": 0.0851, "step": 4860 }, { "grad_norm": 0.523141086101532, "learning_rate": 9.738e-05, "loss": 0.0898, "step": 4870 }, { "grad_norm": 0.6973906755447388, "learning_rate": 9.758000000000001e-05, "loss": 0.0822, "step": 4880 }, { "grad_norm": 0.803859531879425, "learning_rate": 9.778e-05, "loss": 0.1111, "step": 4890 }, { "grad_norm": 0.6532315611839294, "learning_rate": 9.798000000000001e-05, "loss": 0.0882, "step": 4900 }, { "grad_norm": 0.7049265503883362, "learning_rate": 9.818000000000001e-05, "loss": 0.0967, "step": 4910 }, { "grad_norm": 0.6357542872428894, "learning_rate": 9.838e-05, "loss": 0.0973, "step": 4920 }, { "grad_norm": 0.7446737885475159, "learning_rate": 9.858000000000001e-05, "loss": 0.1032, "step": 4930 }, { "grad_norm": 0.5829612612724304, "learning_rate": 9.878e-05, "loss": 0.0865, "step": 4940 }, { "grad_norm": 0.6021092534065247, "learning_rate": 9.898e-05, "loss": 0.1135, "step": 4950 }, { "grad_norm": 0.5933926701545715, "learning_rate": 9.918000000000001e-05, "loss": 0.0905, "step": 4960 }, { "grad_norm": 0.549641489982605, "learning_rate": 9.938e-05, "loss": 0.1109, "step": 4970 }, { "grad_norm": 0.5349535346031189, "learning_rate": 9.958e-05, "loss": 0.0969, "step": 4980 }, { "grad_norm": 0.6998079419136047, "learning_rate": 9.978000000000001e-05, "loss": 0.0944, "step": 4990 }, { "grad_norm": 0.6663141250610352, "learning_rate": 9.998e-05, "loss": 0.1064, "step": 5000 }, { "grad_norm": 0.5780472755432129, "learning_rate": 9.999999778549045e-05, "loss": 0.0972, "step": 5010 }, { "grad_norm": 0.9515729546546936, "learning_rate": 9.999999013039593e-05, "loss": 0.0991, "step": 5020 }, { "grad_norm": 0.6527934670448303, "learning_rate": 9.999997700737766e-05, "loss": 0.0795, "step": 5030 }, { "grad_norm": 0.8080113530158997, "learning_rate": 9.999995841643709e-05, "loss": 0.1001, "step": 5040 }, { "grad_norm": 0.708366334438324, "learning_rate": 9.999993435757623e-05, "loss": 0.1013, "step": 5050 }, { "grad_norm": 0.5948297381401062, "learning_rate": 9.999990483079773e-05, "loss": 0.082, "step": 5060 }, { "grad_norm": 0.5923698544502258, "learning_rate": 9.999986983610481e-05, "loss": 0.107, "step": 5070 }, { "grad_norm": 0.7575445771217346, "learning_rate": 9.99998293735013e-05, "loss": 0.1068, "step": 5080 }, { "grad_norm": 0.8642271161079407, "learning_rate": 9.999978344299161e-05, "loss": 0.0963, "step": 5090 }, { "grad_norm": 0.5847957134246826, "learning_rate": 9.99997320445808e-05, "loss": 0.1035, "step": 5100 }, { "grad_norm": 0.6888630986213684, "learning_rate": 9.999967517827444e-05, "loss": 0.1099, "step": 5110 }, { "grad_norm": 0.8324796557426453, "learning_rate": 9.999961284407879e-05, "loss": 0.1215, "step": 5120 }, { "grad_norm": 0.7159454822540283, "learning_rate": 9.999954504200067e-05, "loss": 0.1056, "step": 5130 }, { "grad_norm": 0.6430867910385132, "learning_rate": 9.999947177204744e-05, "loss": 0.1061, "step": 5140 }, { "grad_norm": 0.5232489705085754, "learning_rate": 9.999939303422718e-05, "loss": 0.0884, "step": 5150 }, { "grad_norm": 0.6383417248725891, "learning_rate": 9.999930882854847e-05, "loss": 0.1169, "step": 5160 }, { "grad_norm": 0.8307160139083862, "learning_rate": 9.999921915502051e-05, "loss": 0.101, "step": 5170 }, { "grad_norm": 0.7485915422439575, "learning_rate": 9.99991240136531e-05, "loss": 0.088, "step": 5180 }, { "grad_norm": 0.7129383087158203, "learning_rate": 9.999902340445668e-05, "loss": 0.1255, "step": 5190 }, { "grad_norm": 0.6299911141395569, "learning_rate": 9.999891732744224e-05, "loss": 0.0835, "step": 5200 }, { "grad_norm": 0.5997602939605713, "learning_rate": 9.999880578262135e-05, "loss": 0.0909, "step": 5210 }, { "grad_norm": 0.7947958707809448, "learning_rate": 9.999868877000624e-05, "loss": 0.0994, "step": 5220 }, { "grad_norm": 0.7014700770378113, "learning_rate": 9.99985662896097e-05, "loss": 0.0886, "step": 5230 }, { "grad_norm": 0.699691653251648, "learning_rate": 9.999843834144513e-05, "loss": 0.0993, "step": 5240 }, { "grad_norm": 0.7539181709289551, "learning_rate": 9.99983049255265e-05, "loss": 0.1053, "step": 5250 }, { "grad_norm": 0.7386245727539062, "learning_rate": 9.999816604186843e-05, "loss": 0.09, "step": 5260 }, { "grad_norm": 0.6962882280349731, "learning_rate": 9.999802169048609e-05, "loss": 0.0949, "step": 5270 }, { "grad_norm": 0.6678304672241211, "learning_rate": 9.999787187139527e-05, "loss": 0.1086, "step": 5280 }, { "grad_norm": 0.6598872542381287, "learning_rate": 9.999771658461234e-05, "loss": 0.1147, "step": 5290 }, { "grad_norm": 0.6872413754463196, "learning_rate": 9.999755583015431e-05, "loss": 0.0868, "step": 5300 }, { "grad_norm": 0.787338137626648, "learning_rate": 9.999738960803874e-05, "loss": 0.1025, "step": 5310 }, { "grad_norm": 0.6820301413536072, "learning_rate": 9.99972179182838e-05, "loss": 0.1, "step": 5320 }, { "grad_norm": 0.8151903748512268, "learning_rate": 9.99970407609083e-05, "loss": 0.1077, "step": 5330 }, { "grad_norm": 0.7049399018287659, "learning_rate": 9.999685813593159e-05, "loss": 0.0914, "step": 5340 }, { "grad_norm": 0.7578346133232117, "learning_rate": 9.999667004337362e-05, "loss": 0.0828, "step": 5350 }, { "grad_norm": 0.590035080909729, "learning_rate": 9.9996476483255e-05, "loss": 0.1073, "step": 5360 }, { "grad_norm": 0.7090608477592468, "learning_rate": 9.999627745559688e-05, "loss": 0.1035, "step": 5370 }, { "grad_norm": 0.6634621620178223, "learning_rate": 9.999607296042101e-05, "loss": 0.0903, "step": 5380 }, { "grad_norm": 0.7024953365325928, "learning_rate": 9.99958629977498e-05, "loss": 0.0881, "step": 5390 }, { "grad_norm": 0.5423393249511719, "learning_rate": 9.999564756760615e-05, "loss": 0.0967, "step": 5400 }, { "grad_norm": 0.587245762348175, "learning_rate": 9.999542667001366e-05, "loss": 0.0839, "step": 5410 }, { "grad_norm": 0.7456430196762085, "learning_rate": 9.999520030499647e-05, "loss": 0.0771, "step": 5420 }, { "grad_norm": 0.7928114533424377, "learning_rate": 9.999496847257936e-05, "loss": 0.0923, "step": 5430 }, { "grad_norm": 0.8389496803283691, "learning_rate": 9.999473117278764e-05, "loss": 0.1119, "step": 5440 }, { "grad_norm": 0.5853843092918396, "learning_rate": 9.999448840564731e-05, "loss": 0.0789, "step": 5450 }, { "grad_norm": 0.7703052759170532, "learning_rate": 9.999424017118488e-05, "loss": 0.0897, "step": 5460 }, { "grad_norm": 0.6673101782798767, "learning_rate": 9.999398646942751e-05, "loss": 0.0812, "step": 5470 }, { "grad_norm": 0.5670938491821289, "learning_rate": 9.999372730040296e-05, "loss": 0.0829, "step": 5480 }, { "grad_norm": 0.7222223281860352, "learning_rate": 9.999346266413953e-05, "loss": 0.0936, "step": 5490 }, { "grad_norm": 0.722626805305481, "learning_rate": 9.99931925606662e-05, "loss": 0.1065, "step": 5500 }, { "grad_norm": 0.6236171722412109, "learning_rate": 9.99929169900125e-05, "loss": 0.0944, "step": 5510 }, { "grad_norm": 0.6258969306945801, "learning_rate": 9.999263595220855e-05, "loss": 0.0828, "step": 5520 }, { "grad_norm": 0.6968818306922913, "learning_rate": 9.99923494472851e-05, "loss": 0.0824, "step": 5530 }, { "grad_norm": 0.6694757342338562, "learning_rate": 9.999205747527348e-05, "loss": 0.081, "step": 5540 }, { "grad_norm": 0.6815804839134216, "learning_rate": 9.999176003620561e-05, "loss": 0.0928, "step": 5550 }, { "grad_norm": 0.6448670029640198, "learning_rate": 9.999145713011405e-05, "loss": 0.0826, "step": 5560 }, { "grad_norm": 0.5240583419799805, "learning_rate": 9.999114875703186e-05, "loss": 0.0854, "step": 5570 }, { "grad_norm": 0.5644882917404175, "learning_rate": 9.999083491699281e-05, "loss": 0.1028, "step": 5580 }, { "grad_norm": 0.7380779385566711, "learning_rate": 9.999051561003123e-05, "loss": 0.092, "step": 5590 }, { "grad_norm": 0.7757815718650818, "learning_rate": 9.999019083618202e-05, "loss": 0.1004, "step": 5600 }, { "grad_norm": 1.0558971166610718, "learning_rate": 9.99898605954807e-05, "loss": 0.0929, "step": 5610 }, { "grad_norm": 0.6253312826156616, "learning_rate": 9.998952488796338e-05, "loss": 0.0803, "step": 5620 }, { "grad_norm": 0.8451449275016785, "learning_rate": 9.998918371366676e-05, "loss": 0.084, "step": 5630 }, { "grad_norm": 0.6669623851776123, "learning_rate": 9.99888370726282e-05, "loss": 0.0948, "step": 5640 }, { "grad_norm": 0.7315524816513062, "learning_rate": 9.998848496488556e-05, "loss": 0.1137, "step": 5650 }, { "grad_norm": 0.6763821244239807, "learning_rate": 9.998812739047736e-05, "loss": 0.0874, "step": 5660 }, { "grad_norm": 0.7099320292472839, "learning_rate": 9.99877643494427e-05, "loss": 0.0881, "step": 5670 }, { "grad_norm": 0.5367976427078247, "learning_rate": 9.998739584182128e-05, "loss": 0.1038, "step": 5680 }, { "grad_norm": 0.9082674384117126, "learning_rate": 9.998702186765342e-05, "loss": 0.0907, "step": 5690 }, { "grad_norm": 0.7115488648414612, "learning_rate": 9.998664242698e-05, "loss": 0.0956, "step": 5700 }, { "grad_norm": 0.7140019536018372, "learning_rate": 9.998625751984251e-05, "loss": 0.0911, "step": 5710 }, { "grad_norm": 0.6991792321205139, "learning_rate": 9.998586714628307e-05, "loss": 0.0836, "step": 5720 }, { "grad_norm": 0.6333011984825134, "learning_rate": 9.998547130634432e-05, "loss": 0.0863, "step": 5730 }, { "grad_norm": 0.4729132056236267, "learning_rate": 9.99850700000696e-05, "loss": 0.0906, "step": 5740 }, { "grad_norm": 0.5187201499938965, "learning_rate": 9.998466322750278e-05, "loss": 0.0724, "step": 5750 }, { "grad_norm": 0.5133963227272034, "learning_rate": 9.998425098868834e-05, "loss": 0.0825, "step": 5760 }, { "grad_norm": 0.6964494585990906, "learning_rate": 9.998383328367136e-05, "loss": 0.0985, "step": 5770 }, { "grad_norm": 0.7275825142860413, "learning_rate": 9.99834101124975e-05, "loss": 0.0995, "step": 5780 }, { "grad_norm": 0.6364948153495789, "learning_rate": 9.998298147521309e-05, "loss": 0.0772, "step": 5790 }, { "grad_norm": 0.6145317554473877, "learning_rate": 9.998254737186496e-05, "loss": 0.0972, "step": 5800 }, { "grad_norm": 0.6101065874099731, "learning_rate": 9.99821078025006e-05, "loss": 0.0832, "step": 5810 }, { "grad_norm": 0.6371209621429443, "learning_rate": 9.998166276716807e-05, "loss": 0.0838, "step": 5820 }, { "grad_norm": 0.6116976141929626, "learning_rate": 9.998121226591606e-05, "loss": 0.0897, "step": 5830 }, { "grad_norm": 0.6270226240158081, "learning_rate": 9.998075629879382e-05, "loss": 0.0802, "step": 5840 }, { "grad_norm": 0.6248165965080261, "learning_rate": 9.99802948658512e-05, "loss": 0.0727, "step": 5850 }, { "grad_norm": 0.5476576089859009, "learning_rate": 9.99798279671387e-05, "loss": 0.0944, "step": 5860 }, { "grad_norm": 0.6771354675292969, "learning_rate": 9.997935560270734e-05, "loss": 0.0805, "step": 5870 }, { "grad_norm": 0.6038175225257874, "learning_rate": 9.997887777260879e-05, "loss": 0.095, "step": 5880 }, { "grad_norm": 0.5173991918563843, "learning_rate": 9.997839447689532e-05, "loss": 0.0925, "step": 5890 }, { "grad_norm": 0.7767724990844727, "learning_rate": 9.997790571561978e-05, "loss": 0.1079, "step": 5900 }, { "grad_norm": 0.785749077796936, "learning_rate": 9.99774114888356e-05, "loss": 0.0923, "step": 5910 }, { "grad_norm": 0.8455897569656372, "learning_rate": 9.997691179659684e-05, "loss": 0.09, "step": 5920 }, { "grad_norm": 0.5847190618515015, "learning_rate": 9.997640663895815e-05, "loss": 0.0784, "step": 5930 }, { "grad_norm": 0.5841396450996399, "learning_rate": 9.997589601597477e-05, "loss": 0.0839, "step": 5940 }, { "grad_norm": 0.542029857635498, "learning_rate": 9.997537992770252e-05, "loss": 0.0724, "step": 5950 }, { "grad_norm": 0.6574005484580994, "learning_rate": 9.997485837419788e-05, "loss": 0.0699, "step": 5960 }, { "grad_norm": 0.7650313973426819, "learning_rate": 9.997433135551786e-05, "loss": 0.0846, "step": 5970 }, { "grad_norm": 0.7232289910316467, "learning_rate": 9.997379887172009e-05, "loss": 0.0872, "step": 5980 }, { "grad_norm": 0.5114104151725769, "learning_rate": 9.997326092286281e-05, "loss": 0.1025, "step": 5990 }, { "grad_norm": 0.596351146697998, "learning_rate": 9.997271750900486e-05, "loss": 0.1194, "step": 6000 }, { "grad_norm": 0.5507991313934326, "learning_rate": 9.997216863020565e-05, "loss": 0.0978, "step": 6010 }, { "grad_norm": 0.7642802596092224, "learning_rate": 9.99716142865252e-05, "loss": 0.0991, "step": 6020 }, { "grad_norm": 0.6161865592002869, "learning_rate": 9.997105447802415e-05, "loss": 0.0996, "step": 6030 }, { "grad_norm": 0.551948606967926, "learning_rate": 9.997048920476373e-05, "loss": 0.0878, "step": 6040 }, { "grad_norm": 0.6465252637863159, "learning_rate": 9.996991846680572e-05, "loss": 0.1065, "step": 6050 }, { "grad_norm": 0.49413183331489563, "learning_rate": 9.996934226421257e-05, "loss": 0.0802, "step": 6060 }, { "grad_norm": 0.6287209987640381, "learning_rate": 9.996876059704726e-05, "loss": 0.0806, "step": 6070 }, { "grad_norm": 0.5977752804756165, "learning_rate": 9.996817346537343e-05, "loss": 0.0954, "step": 6080 }, { "grad_norm": 0.666776716709137, "learning_rate": 9.996758086925526e-05, "loss": 0.1027, "step": 6090 }, { "grad_norm": 0.5652996897697449, "learning_rate": 9.996698280875759e-05, "loss": 0.1083, "step": 6100 }, { "grad_norm": 0.5807763338088989, "learning_rate": 9.99663792839458e-05, "loss": 0.0954, "step": 6110 }, { "grad_norm": 0.5508661270141602, "learning_rate": 9.99657702948859e-05, "loss": 0.0899, "step": 6120 }, { "grad_norm": 0.535129189491272, "learning_rate": 9.996515584164448e-05, "loss": 0.0668, "step": 6130 }, { "grad_norm": 0.6961092352867126, "learning_rate": 9.996453592428873e-05, "loss": 0.0898, "step": 6140 }, { "grad_norm": 0.5697332620620728, "learning_rate": 9.996391054288646e-05, "loss": 0.1141, "step": 6150 }, { "grad_norm": 0.9889100193977356, "learning_rate": 9.996327969750605e-05, "loss": 0.1201, "step": 6160 }, { "grad_norm": 0.5739118456840515, "learning_rate": 9.996264338821649e-05, "loss": 0.1043, "step": 6170 }, { "grad_norm": 0.5627540349960327, "learning_rate": 9.996200161508735e-05, "loss": 0.0863, "step": 6180 }, { "grad_norm": 0.7349043488502502, "learning_rate": 9.996135437818885e-05, "loss": 0.0982, "step": 6190 }, { "grad_norm": 0.6012932062149048, "learning_rate": 9.996070167759175e-05, "loss": 0.0782, "step": 6200 }, { "grad_norm": 0.8591396808624268, "learning_rate": 9.996004351336743e-05, "loss": 0.0936, "step": 6210 }, { "grad_norm": 0.670255184173584, "learning_rate": 9.995937988558785e-05, "loss": 0.0784, "step": 6220 }, { "grad_norm": 0.7117581963539124, "learning_rate": 9.995871079432561e-05, "loss": 0.0862, "step": 6230 }, { "grad_norm": 0.5782531499862671, "learning_rate": 9.995803623965389e-05, "loss": 0.0867, "step": 6240 }, { "grad_norm": 0.4739722013473511, "learning_rate": 9.995735622164641e-05, "loss": 0.0869, "step": 6250 }, { "grad_norm": 0.7375711798667908, "learning_rate": 9.995667074037758e-05, "loss": 0.0652, "step": 6260 }, { "grad_norm": 0.6019680500030518, "learning_rate": 9.995597979592232e-05, "loss": 0.0771, "step": 6270 }, { "grad_norm": 0.5606500506401062, "learning_rate": 9.995528338835625e-05, "loss": 0.0784, "step": 6280 }, { "grad_norm": 1.2010334730148315, "learning_rate": 9.995458151775547e-05, "loss": 0.0905, "step": 6290 }, { "grad_norm": 0.4641077518463135, "learning_rate": 9.995387418419677e-05, "loss": 0.0912, "step": 6300 }, { "grad_norm": 0.6613391637802124, "learning_rate": 9.99531613877575e-05, "loss": 0.088, "step": 6310 }, { "grad_norm": 0.6004403233528137, "learning_rate": 9.995244312851559e-05, "loss": 0.0844, "step": 6320 }, { "grad_norm": 0.7641587853431702, "learning_rate": 9.995171940654961e-05, "loss": 0.079, "step": 6330 }, { "grad_norm": 0.6576120853424072, "learning_rate": 9.995099022193871e-05, "loss": 0.0975, "step": 6340 }, { "grad_norm": 0.45216435194015503, "learning_rate": 9.995025557476261e-05, "loss": 0.0781, "step": 6350 }, { "grad_norm": 0.5096346735954285, "learning_rate": 9.994951546510165e-05, "loss": 0.0863, "step": 6360 }, { "grad_norm": 0.5797306299209595, "learning_rate": 9.994876989303679e-05, "loss": 0.0756, "step": 6370 }, { "grad_norm": 0.6938868761062622, "learning_rate": 9.994801885864955e-05, "loss": 0.0753, "step": 6380 }, { "grad_norm": 0.741682767868042, "learning_rate": 9.994726236202205e-05, "loss": 0.0904, "step": 6390 }, { "grad_norm": 0.7084558606147766, "learning_rate": 9.994650040323704e-05, "loss": 0.0891, "step": 6400 }, { "grad_norm": 0.5232657790184021, "learning_rate": 9.994573298237784e-05, "loss": 0.0718, "step": 6410 }, { "grad_norm": 0.43254154920578003, "learning_rate": 9.994496009952837e-05, "loss": 0.0764, "step": 6420 }, { "grad_norm": 0.5719223022460938, "learning_rate": 9.994418175477316e-05, "loss": 0.0811, "step": 6430 }, { "grad_norm": 0.4670836627483368, "learning_rate": 9.994339794819733e-05, "loss": 0.0767, "step": 6440 }, { "grad_norm": 0.6058287024497986, "learning_rate": 9.994260867988658e-05, "loss": 0.0804, "step": 6450 }, { "grad_norm": 0.5484968423843384, "learning_rate": 9.994181394992723e-05, "loss": 0.0738, "step": 6460 }, { "grad_norm": 0.5039815306663513, "learning_rate": 9.994101375840618e-05, "loss": 0.0754, "step": 6470 }, { "grad_norm": 0.4904034435749054, "learning_rate": 9.994020810541098e-05, "loss": 0.0656, "step": 6480 }, { "grad_norm": 0.6847476363182068, "learning_rate": 9.99393969910297e-05, "loss": 0.0958, "step": 6490 }, { "grad_norm": 0.7273229360580444, "learning_rate": 9.993858041535104e-05, "loss": 0.0778, "step": 6500 }, { "grad_norm": 0.6613147258758545, "learning_rate": 9.99377583784643e-05, "loss": 0.0728, "step": 6510 }, { "grad_norm": 0.6598995923995972, "learning_rate": 9.993693088045939e-05, "loss": 0.0759, "step": 6520 }, { "grad_norm": 0.6502384543418884, "learning_rate": 9.99360979214268e-05, "loss": 0.075, "step": 6530 }, { "grad_norm": 0.5533660054206848, "learning_rate": 9.99352595014576e-05, "loss": 0.0888, "step": 6540 }, { "grad_norm": 0.6724204421043396, "learning_rate": 9.993441562064354e-05, "loss": 0.0759, "step": 6550 }, { "grad_norm": 0.526387095451355, "learning_rate": 9.993356627907685e-05, "loss": 0.0794, "step": 6560 }, { "grad_norm": 0.7170687317848206, "learning_rate": 9.99327114768504e-05, "loss": 0.0933, "step": 6570 }, { "grad_norm": 0.43872880935668945, "learning_rate": 9.99318512140577e-05, "loss": 0.0756, "step": 6580 }, { "grad_norm": 0.511579155921936, "learning_rate": 9.993098549079284e-05, "loss": 0.0861, "step": 6590 }, { "grad_norm": 0.6334900856018066, "learning_rate": 9.993011430715047e-05, "loss": 0.0915, "step": 6600 }, { "grad_norm": 0.4452422261238098, "learning_rate": 9.992923766322586e-05, "loss": 0.066, "step": 6610 }, { "grad_norm": 0.3904229998588562, "learning_rate": 9.99283555591149e-05, "loss": 0.0655, "step": 6620 }, { "grad_norm": 0.7511595487594604, "learning_rate": 9.992746799491404e-05, "loss": 0.0756, "step": 6630 }, { "grad_norm": 0.4748516380786896, "learning_rate": 9.992657497072033e-05, "loss": 0.065, "step": 6640 }, { "grad_norm": 0.5261983275413513, "learning_rate": 9.992567648663147e-05, "loss": 0.0802, "step": 6650 }, { "grad_norm": 0.5062252879142761, "learning_rate": 9.992477254274568e-05, "loss": 0.0696, "step": 6660 }, { "grad_norm": 0.551323652267456, "learning_rate": 9.992386313916183e-05, "loss": 0.0626, "step": 6670 }, { "grad_norm": 0.5742862224578857, "learning_rate": 9.992294827597934e-05, "loss": 0.0789, "step": 6680 }, { "grad_norm": 0.48951202630996704, "learning_rate": 9.992202795329831e-05, "loss": 0.0841, "step": 6690 }, { "grad_norm": 0.7179400324821472, "learning_rate": 9.992110217121936e-05, "loss": 0.0879, "step": 6700 }, { "grad_norm": 0.6057223081588745, "learning_rate": 9.992017092984372e-05, "loss": 0.0845, "step": 6710 }, { "grad_norm": 0.5860945582389832, "learning_rate": 9.991923422927326e-05, "loss": 0.0952, "step": 6720 }, { "grad_norm": 0.4685378670692444, "learning_rate": 9.991829206961037e-05, "loss": 0.081, "step": 6730 }, { "grad_norm": 0.47239434719085693, "learning_rate": 9.991734445095813e-05, "loss": 0.1014, "step": 6740 }, { "grad_norm": 0.4441092610359192, "learning_rate": 9.991639137342015e-05, "loss": 0.078, "step": 6750 }, { "grad_norm": 0.6186733245849609, "learning_rate": 9.991543283710064e-05, "loss": 0.0814, "step": 6760 }, { "grad_norm": 0.7491062879562378, "learning_rate": 9.991446884210445e-05, "loss": 0.071, "step": 6770 }, { "grad_norm": 0.5468415021896362, "learning_rate": 9.9913499388537e-05, "loss": 0.0799, "step": 6780 }, { "grad_norm": 0.6468756198883057, "learning_rate": 9.99125244765043e-05, "loss": 0.0715, "step": 6790 }, { "grad_norm": 0.541450560092926, "learning_rate": 9.991154410611296e-05, "loss": 0.0822, "step": 6800 }, { "grad_norm": 0.4971008598804474, "learning_rate": 9.99105582774702e-05, "loss": 0.066, "step": 6810 }, { "grad_norm": 0.5939796566963196, "learning_rate": 9.990956699068384e-05, "loss": 0.0828, "step": 6820 }, { "grad_norm": 0.5490896105766296, "learning_rate": 9.990857024586224e-05, "loss": 0.0829, "step": 6830 }, { "grad_norm": 0.5159897208213806, "learning_rate": 9.990756804311446e-05, "loss": 0.0885, "step": 6840 }, { "grad_norm": 0.524047315120697, "learning_rate": 9.990656038255006e-05, "loss": 0.0851, "step": 6850 }, { "grad_norm": 0.4735812246799469, "learning_rate": 9.990554726427926e-05, "loss": 0.0755, "step": 6860 }, { "grad_norm": 0.6673099994659424, "learning_rate": 9.990452868841284e-05, "loss": 0.0906, "step": 6870 }, { "grad_norm": 0.6624678373336792, "learning_rate": 9.99035046550622e-05, "loss": 0.0857, "step": 6880 }, { "grad_norm": 0.7810471653938293, "learning_rate": 9.99024751643393e-05, "loss": 0.0835, "step": 6890 }, { "grad_norm": 0.8276792168617249, "learning_rate": 9.990144021635677e-05, "loss": 0.0927, "step": 6900 }, { "grad_norm": 0.615775465965271, "learning_rate": 9.990039981122775e-05, "loss": 0.0719, "step": 6910 }, { "grad_norm": 0.4475822448730469, "learning_rate": 9.989935394906602e-05, "loss": 0.0842, "step": 6920 }, { "grad_norm": 0.566516101360321, "learning_rate": 9.989830262998598e-05, "loss": 0.0837, "step": 6930 }, { "grad_norm": 0.9099370837211609, "learning_rate": 9.989724585410259e-05, "loss": 0.0711, "step": 6940 }, { "grad_norm": 0.6267075538635254, "learning_rate": 9.989618362153139e-05, "loss": 0.0716, "step": 6950 }, { "grad_norm": 0.5676437616348267, "learning_rate": 9.989511593238859e-05, "loss": 0.0836, "step": 6960 }, { "grad_norm": 0.566550612449646, "learning_rate": 9.98940427867909e-05, "loss": 0.0964, "step": 6970 }, { "grad_norm": 0.6707751154899597, "learning_rate": 9.989296418485573e-05, "loss": 0.0936, "step": 6980 }, { "grad_norm": 0.5205867290496826, "learning_rate": 9.989188012670101e-05, "loss": 0.0717, "step": 6990 }, { "grad_norm": 0.7330201864242554, "learning_rate": 9.989079061244528e-05, "loss": 0.0759, "step": 7000 }, { "grad_norm": 0.580059289932251, "learning_rate": 9.988969564220769e-05, "loss": 0.0837, "step": 7010 }, { "grad_norm": 0.6053109765052795, "learning_rate": 9.988859521610801e-05, "loss": 0.0698, "step": 7020 }, { "grad_norm": 0.6753238439559937, "learning_rate": 9.988748933426656e-05, "loss": 0.0942, "step": 7030 }, { "grad_norm": 0.5477245450019836, "learning_rate": 9.988637799680428e-05, "loss": 0.073, "step": 7040 }, { "grad_norm": 0.6070713400840759, "learning_rate": 9.98852612038427e-05, "loss": 0.07, "step": 7050 }, { "grad_norm": 0.5249080061912537, "learning_rate": 9.988413895550397e-05, "loss": 0.0817, "step": 7060 }, { "grad_norm": 0.5686826109886169, "learning_rate": 9.98830112519108e-05, "loss": 0.0791, "step": 7070 }, { "grad_norm": 0.6298580765724182, "learning_rate": 9.98818780931865e-05, "loss": 0.0909, "step": 7080 }, { "grad_norm": 0.803611159324646, "learning_rate": 9.988073947945502e-05, "loss": 0.0833, "step": 7090 }, { "grad_norm": 0.7247141003608704, "learning_rate": 9.987959541084087e-05, "loss": 0.0841, "step": 7100 }, { "grad_norm": 0.41248348355293274, "learning_rate": 9.987844588746915e-05, "loss": 0.0842, "step": 7110 }, { "grad_norm": 0.6060003638267517, "learning_rate": 9.987729090946558e-05, "loss": 0.0648, "step": 7120 }, { "grad_norm": 0.42526450753211975, "learning_rate": 9.987613047695647e-05, "loss": 0.0748, "step": 7130 }, { "grad_norm": 0.5188283324241638, "learning_rate": 9.987496459006871e-05, "loss": 0.0958, "step": 7140 }, { "grad_norm": 0.6701939702033997, "learning_rate": 9.987379324892982e-05, "loss": 0.0891, "step": 7150 }, { "grad_norm": 0.8528227806091309, "learning_rate": 9.987261645366788e-05, "loss": 0.0868, "step": 7160 }, { "grad_norm": 0.6214022636413574, "learning_rate": 9.987143420441158e-05, "loss": 0.0816, "step": 7170 }, { "grad_norm": 0.48640474677085876, "learning_rate": 9.987024650129022e-05, "loss": 0.0689, "step": 7180 }, { "grad_norm": 0.45794597268104553, "learning_rate": 9.986905334443368e-05, "loss": 0.0825, "step": 7190 }, { "grad_norm": 0.6395137310028076, "learning_rate": 9.986785473397245e-05, "loss": 0.0836, "step": 7200 }, { "grad_norm": 0.6933766603469849, "learning_rate": 9.98666506700376e-05, "loss": 0.0852, "step": 7210 }, { "grad_norm": 0.6115215420722961, "learning_rate": 9.986544115276081e-05, "loss": 0.0721, "step": 7220 }, { "grad_norm": 0.5975186824798584, "learning_rate": 9.986422618227433e-05, "loss": 0.0724, "step": 7230 }, { "grad_norm": 0.6195501685142517, "learning_rate": 9.986300575871106e-05, "loss": 0.0826, "step": 7240 }, { "grad_norm": 0.5619465708732605, "learning_rate": 9.986177988220444e-05, "loss": 0.0747, "step": 7250 }, { "grad_norm": 0.7321537137031555, "learning_rate": 9.986054855288856e-05, "loss": 0.0914, "step": 7260 }, { "grad_norm": 0.6640805602073669, "learning_rate": 9.985931177089802e-05, "loss": 0.0752, "step": 7270 }, { "grad_norm": 0.5150489211082458, "learning_rate": 9.985806953636814e-05, "loss": 0.0689, "step": 7280 }, { "grad_norm": 0.6594642996788025, "learning_rate": 9.985682184943471e-05, "loss": 0.0712, "step": 7290 }, { "grad_norm": 0.6049561500549316, "learning_rate": 9.98555687102342e-05, "loss": 0.0719, "step": 7300 }, { "grad_norm": 0.669260561466217, "learning_rate": 9.985431011890367e-05, "loss": 0.076, "step": 7310 }, { "grad_norm": 0.5750633478164673, "learning_rate": 9.985304607558075e-05, "loss": 0.0722, "step": 7320 }, { "grad_norm": 0.6486774682998657, "learning_rate": 9.985177658040364e-05, "loss": 0.0723, "step": 7330 }, { "grad_norm": 0.602841854095459, "learning_rate": 9.985050163351119e-05, "loss": 0.0863, "step": 7340 }, { "grad_norm": 0.6712275743484497, "learning_rate": 9.984922123504286e-05, "loss": 0.0705, "step": 7350 }, { "grad_norm": 0.41594651341438293, "learning_rate": 9.984793538513862e-05, "loss": 0.0706, "step": 7360 }, { "grad_norm": 0.526322066783905, "learning_rate": 9.984664408393912e-05, "loss": 0.0823, "step": 7370 }, { "grad_norm": 0.4976455271244049, "learning_rate": 9.984534733158556e-05, "loss": 0.0707, "step": 7380 }, { "grad_norm": 0.49437373876571655, "learning_rate": 9.984404512821977e-05, "loss": 0.0723, "step": 7390 }, { "grad_norm": 0.5633967518806458, "learning_rate": 9.984273747398411e-05, "loss": 0.0657, "step": 7400 }, { "grad_norm": 0.5874425768852234, "learning_rate": 9.984142436902165e-05, "loss": 0.0764, "step": 7410 }, { "grad_norm": 0.477426677942276, "learning_rate": 9.984010581347596e-05, "loss": 0.0796, "step": 7420 }, { "grad_norm": 0.7504816651344299, "learning_rate": 9.983878180749121e-05, "loss": 0.083, "step": 7430 }, { "grad_norm": 0.6473439931869507, "learning_rate": 9.983745235121222e-05, "loss": 0.0865, "step": 7440 }, { "grad_norm": 0.6186310648918152, "learning_rate": 9.983611744478438e-05, "loss": 0.0759, "step": 7450 }, { "grad_norm": 0.48286908864974976, "learning_rate": 9.983477708835365e-05, "loss": 0.0769, "step": 7460 }, { "grad_norm": 0.6252347230911255, "learning_rate": 9.983343128206664e-05, "loss": 0.0789, "step": 7470 }, { "grad_norm": 0.7324609756469727, "learning_rate": 9.983208002607049e-05, "loss": 0.0907, "step": 7480 }, { "grad_norm": 0.6513383984565735, "learning_rate": 9.9830723320513e-05, "loss": 0.0795, "step": 7490 }, { "grad_norm": 0.5544431209564209, "learning_rate": 9.982936116554254e-05, "loss": 0.0781, "step": 7500 }, { "grad_norm": 0.6083570718765259, "learning_rate": 9.982799356130803e-05, "loss": 0.0661, "step": 7510 }, { "grad_norm": 0.4932217597961426, "learning_rate": 9.982662050795908e-05, "loss": 0.076, "step": 7520 }, { "grad_norm": 0.6870560050010681, "learning_rate": 9.982524200564583e-05, "loss": 0.075, "step": 7530 }, { "grad_norm": 0.47274845838546753, "learning_rate": 9.982385805451901e-05, "loss": 0.0759, "step": 7540 }, { "grad_norm": 0.5252658128738403, "learning_rate": 9.982246865472998e-05, "loss": 0.0718, "step": 7550 }, { "grad_norm": 0.5671672821044922, "learning_rate": 9.982107380643069e-05, "loss": 0.0736, "step": 7560 }, { "grad_norm": 0.6073296666145325, "learning_rate": 9.981967350977368e-05, "loss": 0.0972, "step": 7570 }, { "grad_norm": 0.7944881916046143, "learning_rate": 9.981826776491208e-05, "loss": 0.0938, "step": 7580 }, { "grad_norm": 0.584837794303894, "learning_rate": 9.98168565719996e-05, "loss": 0.0852, "step": 7590 }, { "grad_norm": 0.6002262234687805, "learning_rate": 9.98154399311906e-05, "loss": 0.0972, "step": 7600 }, { "grad_norm": 0.5188521146774292, "learning_rate": 9.981401784263997e-05, "loss": 0.087, "step": 7610 }, { "grad_norm": 0.46076253056526184, "learning_rate": 9.981259030650326e-05, "loss": 0.0814, "step": 7620 }, { "grad_norm": 0.4761243164539337, "learning_rate": 9.981115732293655e-05, "loss": 0.0766, "step": 7630 }, { "grad_norm": 0.6862100958824158, "learning_rate": 9.980971889209659e-05, "loss": 0.0766, "step": 7640 }, { "grad_norm": 0.40463006496429443, "learning_rate": 9.980827501414064e-05, "loss": 0.0758, "step": 7650 }, { "grad_norm": 0.5173660516738892, "learning_rate": 9.980682568922663e-05, "loss": 0.0652, "step": 7660 }, { "grad_norm": 0.6446139216423035, "learning_rate": 9.980537091751304e-05, "loss": 0.088, "step": 7670 }, { "grad_norm": 0.566994309425354, "learning_rate": 9.980391069915897e-05, "loss": 0.0802, "step": 7680 }, { "grad_norm": 0.532892107963562, "learning_rate": 9.98024450343241e-05, "loss": 0.0786, "step": 7690 }, { "grad_norm": 0.4542059600353241, "learning_rate": 9.980097392316872e-05, "loss": 0.0667, "step": 7700 }, { "grad_norm": 0.4628738760948181, "learning_rate": 9.97994973658537e-05, "loss": 0.0939, "step": 7710 }, { "grad_norm": 0.6938408017158508, "learning_rate": 9.979801536254054e-05, "loss": 0.0844, "step": 7720 }, { "grad_norm": 0.49879521131515503, "learning_rate": 9.979652791339127e-05, "loss": 0.079, "step": 7730 }, { "grad_norm": 0.5465453863143921, "learning_rate": 9.97950350185686e-05, "loss": 0.0788, "step": 7740 }, { "grad_norm": 0.4239027500152588, "learning_rate": 9.979353667823574e-05, "loss": 0.0686, "step": 7750 }, { "grad_norm": 0.45924416184425354, "learning_rate": 9.979203289255658e-05, "loss": 0.0755, "step": 7760 }, { "grad_norm": 0.6220340728759766, "learning_rate": 9.979052366169557e-05, "loss": 0.0868, "step": 7770 }, { "grad_norm": 0.5418806076049805, "learning_rate": 9.978900898581775e-05, "loss": 0.0894, "step": 7780 }, { "grad_norm": 0.5265543460845947, "learning_rate": 9.978748886508875e-05, "loss": 0.079, "step": 7790 }, { "grad_norm": 0.6076542735099792, "learning_rate": 9.978596329967484e-05, "loss": 0.0839, "step": 7800 }, { "grad_norm": 0.5506042242050171, "learning_rate": 9.978443228974284e-05, "loss": 0.0861, "step": 7810 }, { "grad_norm": 0.5266274213790894, "learning_rate": 9.978289583546015e-05, "loss": 0.0698, "step": 7820 }, { "grad_norm": 0.6848101615905762, "learning_rate": 9.978135393699484e-05, "loss": 0.0803, "step": 7830 }, { "grad_norm": 0.5885761976242065, "learning_rate": 9.977980659451548e-05, "loss": 0.0767, "step": 7840 }, { "grad_norm": 0.5867879390716553, "learning_rate": 9.977825380819135e-05, "loss": 0.0871, "step": 7850 }, { "grad_norm": 0.4875785708427429, "learning_rate": 9.97766955781922e-05, "loss": 0.0828, "step": 7860 }, { "grad_norm": 0.6715461015701294, "learning_rate": 9.977513190468848e-05, "loss": 0.097, "step": 7870 }, { "grad_norm": 0.4613666534423828, "learning_rate": 9.977356278785116e-05, "loss": 0.0857, "step": 7880 }, { "grad_norm": 0.7143324613571167, "learning_rate": 9.977198822785184e-05, "loss": 0.0765, "step": 7890 }, { "grad_norm": 0.4525129199028015, "learning_rate": 9.977040822486273e-05, "loss": 0.0963, "step": 7900 }, { "grad_norm": 0.47830039262771606, "learning_rate": 9.97688227790566e-05, "loss": 0.0985, "step": 7910 }, { "grad_norm": 0.5309436321258545, "learning_rate": 9.976723189060684e-05, "loss": 0.1091, "step": 7920 }, { "grad_norm": 0.6492742896080017, "learning_rate": 9.976563555968742e-05, "loss": 0.0787, "step": 7930 }, { "grad_norm": 0.5649189352989197, "learning_rate": 9.976403378647292e-05, "loss": 0.0764, "step": 7940 }, { "grad_norm": 0.6107155084609985, "learning_rate": 9.97624265711385e-05, "loss": 0.0761, "step": 7950 }, { "grad_norm": 0.4522315561771393, "learning_rate": 9.976081391385993e-05, "loss": 0.0743, "step": 7960 }, { "grad_norm": 0.4722643196582794, "learning_rate": 9.975919581481356e-05, "loss": 0.08, "step": 7970 }, { "grad_norm": 0.46496817469596863, "learning_rate": 9.975757227417634e-05, "loss": 0.0656, "step": 7980 }, { "grad_norm": 0.570717990398407, "learning_rate": 9.975594329212586e-05, "loss": 0.0758, "step": 7990 }, { "grad_norm": 0.4886288642883301, "learning_rate": 9.97543088688402e-05, "loss": 0.0659, "step": 8000 }, { "grad_norm": 0.5527797341346741, "learning_rate": 9.975266900449814e-05, "loss": 0.0661, "step": 8010 }, { "grad_norm": 0.5069243907928467, "learning_rate": 9.975102369927898e-05, "loss": 0.0645, "step": 8020 }, { "grad_norm": 0.385810524225235, "learning_rate": 9.974937295336269e-05, "loss": 0.0638, "step": 8030 }, { "grad_norm": 0.6145714521408081, "learning_rate": 9.974771676692975e-05, "loss": 0.063, "step": 8040 }, { "grad_norm": 0.4625590443611145, "learning_rate": 9.974605514016131e-05, "loss": 0.0736, "step": 8050 }, { "grad_norm": 0.4838498830795288, "learning_rate": 9.974438807323907e-05, "loss": 0.0676, "step": 8060 }, { "grad_norm": 0.49213382601737976, "learning_rate": 9.974271556634535e-05, "loss": 0.0668, "step": 8070 }, { "grad_norm": 0.7374060750007629, "learning_rate": 9.974103761966302e-05, "loss": 0.0914, "step": 8080 }, { "grad_norm": 0.4567733407020569, "learning_rate": 9.973935423337563e-05, "loss": 0.0712, "step": 8090 }, { "grad_norm": 0.5095852017402649, "learning_rate": 9.973766540766722e-05, "loss": 0.0755, "step": 8100 }, { "grad_norm": 0.5341172814369202, "learning_rate": 9.97359711427225e-05, "loss": 0.0763, "step": 8110 }, { "grad_norm": 0.3869343101978302, "learning_rate": 9.973427143872677e-05, "loss": 0.0747, "step": 8120 }, { "grad_norm": 0.7308095693588257, "learning_rate": 9.973256629586589e-05, "loss": 0.0877, "step": 8130 }, { "grad_norm": 0.5212194323539734, "learning_rate": 9.973085571432632e-05, "loss": 0.0811, "step": 8140 }, { "grad_norm": 0.5269376039505005, "learning_rate": 9.972913969429513e-05, "loss": 0.0688, "step": 8150 }, { "grad_norm": 0.9371418952941895, "learning_rate": 9.972741823596e-05, "loss": 0.0868, "step": 8160 }, { "grad_norm": 0.5341587066650391, "learning_rate": 9.972569133950917e-05, "loss": 0.0866, "step": 8170 }, { "grad_norm": 0.5898665189743042, "learning_rate": 9.972395900513151e-05, "loss": 0.0753, "step": 8180 }, { "grad_norm": 0.5001934170722961, "learning_rate": 9.972222123301645e-05, "loss": 0.0827, "step": 8190 }, { "grad_norm": 0.6120069622993469, "learning_rate": 9.972047802335403e-05, "loss": 0.0687, "step": 8200 }, { "grad_norm": 0.37794795632362366, "learning_rate": 9.971872937633488e-05, "loss": 0.0625, "step": 8210 }, { "grad_norm": 0.4569222033023834, "learning_rate": 9.971697529215024e-05, "loss": 0.073, "step": 8220 }, { "grad_norm": 0.6726720929145813, "learning_rate": 9.971521577099192e-05, "loss": 0.0742, "step": 8230 }, { "grad_norm": 0.4542614817619324, "learning_rate": 9.971345081305236e-05, "loss": 0.0757, "step": 8240 }, { "grad_norm": 0.9851104617118835, "learning_rate": 9.971168041852456e-05, "loss": 0.0958, "step": 8250 }, { "grad_norm": 0.5607678294181824, "learning_rate": 9.970990458760215e-05, "loss": 0.0788, "step": 8260 }, { "grad_norm": 0.437242329120636, "learning_rate": 9.970812332047929e-05, "loss": 0.0671, "step": 8270 }, { "grad_norm": 0.5152568817138672, "learning_rate": 9.97063366173508e-05, "loss": 0.0819, "step": 8280 }, { "grad_norm": 0.49926507472991943, "learning_rate": 9.970454447841207e-05, "loss": 0.0717, "step": 8290 }, { "grad_norm": 0.42014724016189575, "learning_rate": 9.970274690385909e-05, "loss": 0.0762, "step": 8300 }, { "grad_norm": 0.5476452708244324, "learning_rate": 9.970094389388844e-05, "loss": 0.0813, "step": 8310 }, { "grad_norm": 0.6533055305480957, "learning_rate": 9.969913544869728e-05, "loss": 0.0726, "step": 8320 }, { "grad_norm": 0.47668692469596863, "learning_rate": 9.96973215684834e-05, "loss": 0.0683, "step": 8330 }, { "grad_norm": 0.5635125041007996, "learning_rate": 9.969550225344513e-05, "loss": 0.0653, "step": 8340 }, { "grad_norm": 0.6407223343849182, "learning_rate": 9.969367750378147e-05, "loss": 0.0874, "step": 8350 }, { "grad_norm": 0.6044015884399414, "learning_rate": 9.969184731969194e-05, "loss": 0.0676, "step": 8360 }, { "grad_norm": 0.6031338572502136, "learning_rate": 9.96900117013767e-05, "loss": 0.0716, "step": 8370 }, { "grad_norm": 0.5230648517608643, "learning_rate": 9.96881706490365e-05, "loss": 0.0734, "step": 8380 }, { "grad_norm": 0.44588935375213623, "learning_rate": 9.968632416287265e-05, "loss": 0.0736, "step": 8390 }, { "grad_norm": 0.5067813992500305, "learning_rate": 9.96844722430871e-05, "loss": 0.0717, "step": 8400 }, { "grad_norm": 0.46525996923446655, "learning_rate": 9.968261488988235e-05, "loss": 0.0828, "step": 8410 }, { "grad_norm": 0.48338374495506287, "learning_rate": 9.968075210346155e-05, "loss": 0.0748, "step": 8420 }, { "grad_norm": 0.7847834825515747, "learning_rate": 9.967888388402839e-05, "loss": 0.0792, "step": 8430 }, { "grad_norm": 0.5393670797348022, "learning_rate": 9.967701023178717e-05, "loss": 0.0707, "step": 8440 }, { "grad_norm": 0.5936521887779236, "learning_rate": 9.967513114694282e-05, "loss": 0.0709, "step": 8450 }, { "grad_norm": 0.45119085907936096, "learning_rate": 9.967324662970079e-05, "loss": 0.0683, "step": 8460 }, { "grad_norm": 0.49417805671691895, "learning_rate": 9.96713566802672e-05, "loss": 0.0655, "step": 8470 }, { "grad_norm": 0.4346674084663391, "learning_rate": 9.966946129884873e-05, "loss": 0.0709, "step": 8480 }, { "grad_norm": 0.673485517501831, "learning_rate": 9.966756048565265e-05, "loss": 0.0761, "step": 8490 }, { "grad_norm": 0.5993248820304871, "learning_rate": 9.966565424088681e-05, "loss": 0.0773, "step": 8500 }, { "grad_norm": 0.5165562629699707, "learning_rate": 9.96637425647597e-05, "loss": 0.0707, "step": 8510 }, { "grad_norm": 0.64022296667099, "learning_rate": 9.966182545748038e-05, "loss": 0.0655, "step": 8520 }, { "grad_norm": 0.508798360824585, "learning_rate": 9.96599029192585e-05, "loss": 0.0759, "step": 8530 }, { "grad_norm": 0.4519994258880615, "learning_rate": 9.965797495030428e-05, "loss": 0.0651, "step": 8540 }, { "grad_norm": 0.6334484815597534, "learning_rate": 9.96560415508286e-05, "loss": 0.0769, "step": 8550 }, { "grad_norm": 0.4821540415287018, "learning_rate": 9.965410272104286e-05, "loss": 0.0706, "step": 8560 }, { "grad_norm": 0.577206015586853, "learning_rate": 9.96521584611591e-05, "loss": 0.1033, "step": 8570 }, { "grad_norm": 0.5250496864318848, "learning_rate": 9.965020877138994e-05, "loss": 0.0847, "step": 8580 }, { "grad_norm": 0.5466002225875854, "learning_rate": 9.964825365194861e-05, "loss": 0.0812, "step": 8590 }, { "grad_norm": 0.4655781090259552, "learning_rate": 9.96462931030489e-05, "loss": 0.0651, "step": 8600 }, { "grad_norm": 0.5801318287849426, "learning_rate": 9.96443271249052e-05, "loss": 0.0664, "step": 8610 }, { "grad_norm": 0.6525282859802246, "learning_rate": 9.964235571773255e-05, "loss": 0.091, "step": 8620 }, { "grad_norm": 0.4144763946533203, "learning_rate": 9.96403788817465e-05, "loss": 0.0673, "step": 8630 }, { "grad_norm": 0.6374234557151794, "learning_rate": 9.963839661716325e-05, "loss": 0.0782, "step": 8640 }, { "grad_norm": 0.5224519968032837, "learning_rate": 9.963640892419958e-05, "loss": 0.0765, "step": 8650 }, { "grad_norm": 0.6026553511619568, "learning_rate": 9.963441580307286e-05, "loss": 0.0721, "step": 8660 }, { "grad_norm": 0.5134572386741638, "learning_rate": 9.963241725400104e-05, "loss": 0.0618, "step": 8670 }, { "grad_norm": 0.5257141590118408, "learning_rate": 9.963041327720271e-05, "loss": 0.0608, "step": 8680 }, { "grad_norm": 0.5906543135643005, "learning_rate": 9.962840387289697e-05, "loss": 0.0684, "step": 8690 }, { "grad_norm": 0.4562460780143738, "learning_rate": 9.962638904130363e-05, "loss": 0.0645, "step": 8700 }, { "grad_norm": 0.45447662472724915, "learning_rate": 9.962436878264298e-05, "loss": 0.0694, "step": 8710 }, { "grad_norm": 0.6880111694335938, "learning_rate": 9.962234309713598e-05, "loss": 0.0921, "step": 8720 }, { "grad_norm": 0.40074530243873596, "learning_rate": 9.962031198500414e-05, "loss": 0.0691, "step": 8730 }, { "grad_norm": 0.5927583575248718, "learning_rate": 9.961827544646958e-05, "loss": 0.0739, "step": 8740 }, { "grad_norm": 0.5247321724891663, "learning_rate": 9.961623348175501e-05, "loss": 0.0766, "step": 8750 }, { "grad_norm": 0.5753244161605835, "learning_rate": 9.961418609108377e-05, "loss": 0.0718, "step": 8760 }, { "grad_norm": 0.6366556882858276, "learning_rate": 9.961213327467971e-05, "loss": 0.0707, "step": 8770 }, { "grad_norm": 0.5722011923789978, "learning_rate": 9.961007503276736e-05, "loss": 0.0785, "step": 8780 }, { "grad_norm": 0.5187246203422546, "learning_rate": 9.960801136557179e-05, "loss": 0.0685, "step": 8790 }, { "grad_norm": 0.5229710340499878, "learning_rate": 9.960594227331866e-05, "loss": 0.0627, "step": 8800 }, { "grad_norm": 0.45116695761680603, "learning_rate": 9.960386775623429e-05, "loss": 0.0613, "step": 8810 }, { "grad_norm": 0.44209977984428406, "learning_rate": 9.96017878145455e-05, "loss": 0.0788, "step": 8820 }, { "grad_norm": 0.598173975944519, "learning_rate": 9.959970244847977e-05, "loss": 0.0756, "step": 8830 }, { "grad_norm": 0.6599321365356445, "learning_rate": 9.959761165826518e-05, "loss": 0.0813, "step": 8840 }, { "grad_norm": 0.5646312236785889, "learning_rate": 9.959551544413033e-05, "loss": 0.0672, "step": 8850 }, { "grad_norm": 0.6010459065437317, "learning_rate": 9.959341380630448e-05, "loss": 0.0773, "step": 8860 }, { "grad_norm": 1.854341983795166, "learning_rate": 9.959130674501746e-05, "loss": 0.0821, "step": 8870 }, { "grad_norm": 0.5280343294143677, "learning_rate": 9.958919426049968e-05, "loss": 0.0778, "step": 8880 }, { "grad_norm": 0.5351850986480713, "learning_rate": 9.958707635298219e-05, "loss": 0.0735, "step": 8890 }, { "grad_norm": 0.5301709175109863, "learning_rate": 9.958495302269657e-05, "loss": 0.0741, "step": 8900 }, { "grad_norm": 0.4962819218635559, "learning_rate": 9.958282426987503e-05, "loss": 0.0668, "step": 8910 }, { "grad_norm": 0.5532568097114563, "learning_rate": 9.95806900947504e-05, "loss": 0.078, "step": 8920 }, { "grad_norm": 0.5747710466384888, "learning_rate": 9.957855049755604e-05, "loss": 0.0607, "step": 8930 }, { "grad_norm": 0.4259873330593109, "learning_rate": 9.957640547852593e-05, "loss": 0.0727, "step": 8940 }, { "grad_norm": 0.5030853748321533, "learning_rate": 9.957425503789466e-05, "loss": 0.0713, "step": 8950 }, { "grad_norm": 0.6353023052215576, "learning_rate": 9.957209917589738e-05, "loss": 0.0731, "step": 8960 }, { "grad_norm": 0.5828598141670227, "learning_rate": 9.956993789276987e-05, "loss": 0.0693, "step": 8970 }, { "grad_norm": 0.617170512676239, "learning_rate": 9.956777118874847e-05, "loss": 0.0668, "step": 8980 }, { "grad_norm": 0.6534974575042725, "learning_rate": 9.956559906407016e-05, "loss": 0.093, "step": 8990 }, { "grad_norm": 0.5831897258758545, "learning_rate": 9.956342151897245e-05, "loss": 0.0771, "step": 9000 }, { "grad_norm": 0.38898777961730957, "learning_rate": 9.956123855369346e-05, "loss": 0.0604, "step": 9010 }, { "grad_norm": 0.5515241026878357, "learning_rate": 9.955905016847196e-05, "loss": 0.0635, "step": 9020 }, { "grad_norm": 0.52506422996521, "learning_rate": 9.955685636354723e-05, "loss": 0.0861, "step": 9030 }, { "grad_norm": 0.4907192587852478, "learning_rate": 9.95546571391592e-05, "loss": 0.0783, "step": 9040 }, { "grad_norm": 0.44356846809387207, "learning_rate": 9.955245249554837e-05, "loss": 0.0622, "step": 9050 }, { "grad_norm": 0.5010348558425903, "learning_rate": 9.955024243295582e-05, "loss": 0.0821, "step": 9060 }, { "grad_norm": 0.42698314785957336, "learning_rate": 9.954802695162328e-05, "loss": 0.065, "step": 9070 }, { "grad_norm": 0.600908637046814, "learning_rate": 9.954580605179302e-05, "loss": 0.0666, "step": 9080 }, { "grad_norm": 0.444741815328598, "learning_rate": 9.954357973370788e-05, "loss": 0.0694, "step": 9090 }, { "grad_norm": 0.5683979988098145, "learning_rate": 9.954134799761135e-05, "loss": 0.085, "step": 9100 }, { "grad_norm": 0.577621579170227, "learning_rate": 9.953911084374748e-05, "loss": 0.0821, "step": 9110 }, { "grad_norm": 0.5663365721702576, "learning_rate": 9.953686827236093e-05, "loss": 0.0701, "step": 9120 }, { "grad_norm": 0.4246208369731903, "learning_rate": 9.953462028369695e-05, "loss": 0.0747, "step": 9130 }, { "grad_norm": 0.5871152877807617, "learning_rate": 9.953236687800136e-05, "loss": 0.0708, "step": 9140 }, { "grad_norm": 0.5557388067245483, "learning_rate": 9.95301080555206e-05, "loss": 0.0801, "step": 9150 }, { "grad_norm": 0.3487769663333893, "learning_rate": 9.952784381650171e-05, "loss": 0.0569, "step": 9160 }, { "grad_norm": 0.5955565571784973, "learning_rate": 9.952557416119226e-05, "loss": 0.067, "step": 9170 }, { "grad_norm": 0.5235143899917603, "learning_rate": 9.95232990898405e-05, "loss": 0.061, "step": 9180 }, { "grad_norm": 0.39055392146110535, "learning_rate": 9.95210186026952e-05, "loss": 0.0561, "step": 9190 }, { "grad_norm": 0.5344921946525574, "learning_rate": 9.951873270000576e-05, "loss": 0.0643, "step": 9200 }, { "grad_norm": 0.40415188670158386, "learning_rate": 9.951644138202216e-05, "loss": 0.0806, "step": 9210 }, { "grad_norm": 0.7637374997138977, "learning_rate": 9.951414464899498e-05, "loss": 0.0801, "step": 9220 }, { "grad_norm": 0.7355364561080933, "learning_rate": 9.951184250117538e-05, "loss": 0.0678, "step": 9230 }, { "grad_norm": 0.48314207792282104, "learning_rate": 9.950953493881513e-05, "loss": 0.0783, "step": 9240 }, { "grad_norm": 0.5201412439346313, "learning_rate": 9.950722196216658e-05, "loss": 0.0632, "step": 9250 }, { "grad_norm": 0.4266493022441864, "learning_rate": 9.950490357148265e-05, "loss": 0.0637, "step": 9260 }, { "grad_norm": 0.6284483075141907, "learning_rate": 9.950257976701692e-05, "loss": 0.0742, "step": 9270 }, { "grad_norm": 0.5629191398620605, "learning_rate": 9.950025054902348e-05, "loss": 0.0748, "step": 9280 }, { "grad_norm": 0.6160375475883484, "learning_rate": 9.949791591775706e-05, "loss": 0.0783, "step": 9290 }, { "grad_norm": 0.5182037353515625, "learning_rate": 9.949557587347298e-05, "loss": 0.0866, "step": 9300 }, { "grad_norm": 0.5593995451927185, "learning_rate": 9.949323041642713e-05, "loss": 0.0624, "step": 9310 }, { "grad_norm": 0.41408824920654297, "learning_rate": 9.949087954687602e-05, "loss": 0.0894, "step": 9320 }, { "grad_norm": 0.4692794680595398, "learning_rate": 9.948852326507672e-05, "loss": 0.0737, "step": 9330 }, { "grad_norm": 0.5953066349029541, "learning_rate": 9.948616157128694e-05, "loss": 0.0709, "step": 9340 }, { "grad_norm": 0.4476054608821869, "learning_rate": 9.948379446576493e-05, "loss": 0.0627, "step": 9350 }, { "grad_norm": 0.3977835178375244, "learning_rate": 9.948142194876952e-05, "loss": 0.0686, "step": 9360 }, { "grad_norm": 0.484540194272995, "learning_rate": 9.947904402056024e-05, "loss": 0.0853, "step": 9370 }, { "grad_norm": 0.5279958248138428, "learning_rate": 9.947666068139708e-05, "loss": 0.0764, "step": 9380 }, { "grad_norm": 0.9781321287155151, "learning_rate": 9.947427193154071e-05, "loss": 0.0843, "step": 9390 }, { "grad_norm": 0.5487475991249084, "learning_rate": 9.947187777125233e-05, "loss": 0.0793, "step": 9400 }, { "grad_norm": 0.5921435952186584, "learning_rate": 9.946947820079377e-05, "loss": 0.0814, "step": 9410 }, { "grad_norm": 0.7337366342544556, "learning_rate": 9.946707322042747e-05, "loss": 0.0803, "step": 9420 }, { "grad_norm": 0.42888063192367554, "learning_rate": 9.94646628304164e-05, "loss": 0.0611, "step": 9430 }, { "grad_norm": 0.5089788436889648, "learning_rate": 9.946224703102418e-05, "loss": 0.0644, "step": 9440 }, { "grad_norm": 0.5632354021072388, "learning_rate": 9.945982582251498e-05, "loss": 0.0722, "step": 9450 }, { "grad_norm": 0.8915722966194153, "learning_rate": 9.94573992051536e-05, "loss": 0.0763, "step": 9460 }, { "grad_norm": 0.649315595626831, "learning_rate": 9.94549671792054e-05, "loss": 0.0915, "step": 9470 }, { "grad_norm": 0.6011292338371277, "learning_rate": 9.945252974493635e-05, "loss": 0.0646, "step": 9480 }, { "grad_norm": 0.5423955917358398, "learning_rate": 9.9450086902613e-05, "loss": 0.0756, "step": 9490 }, { "grad_norm": 0.5466525554656982, "learning_rate": 9.944763865250248e-05, "loss": 0.0643, "step": 9500 }, { "grad_norm": 0.4715823829174042, "learning_rate": 9.944518499487254e-05, "loss": 0.08, "step": 9510 }, { "grad_norm": 0.561896800994873, "learning_rate": 9.944272592999151e-05, "loss": 0.0715, "step": 9520 }, { "grad_norm": 0.5288023948669434, "learning_rate": 9.94402614581283e-05, "loss": 0.0641, "step": 9530 }, { "grad_norm": 0.7662191390991211, "learning_rate": 9.943779157955244e-05, "loss": 0.0794, "step": 9540 }, { "grad_norm": 0.3937842547893524, "learning_rate": 9.943531629453403e-05, "loss": 0.0668, "step": 9550 }, { "grad_norm": 0.5532125234603882, "learning_rate": 9.943283560334375e-05, "loss": 0.0694, "step": 9560 }, { "grad_norm": 0.49250537157058716, "learning_rate": 9.943034950625288e-05, "loss": 0.0785, "step": 9570 }, { "grad_norm": 0.4778591990470886, "learning_rate": 9.942785800353332e-05, "loss": 0.0668, "step": 9580 }, { "grad_norm": 0.5246980786323547, "learning_rate": 9.942536109545751e-05, "loss": 0.0705, "step": 9590 }, { "grad_norm": 0.5475078225135803, "learning_rate": 9.942285878229853e-05, "loss": 0.0622, "step": 9600 }, { "grad_norm": 0.4640369713306427, "learning_rate": 9.942035106433001e-05, "loss": 0.0649, "step": 9610 }, { "grad_norm": 0.6089525818824768, "learning_rate": 9.94178379418262e-05, "loss": 0.0845, "step": 9620 }, { "grad_norm": 0.4830666482448578, "learning_rate": 9.941531941506194e-05, "loss": 0.0742, "step": 9630 }, { "grad_norm": 0.488924115896225, "learning_rate": 9.941279548431263e-05, "loss": 0.0584, "step": 9640 }, { "grad_norm": 0.4627329707145691, "learning_rate": 9.941026614985431e-05, "loss": 0.0618, "step": 9650 }, { "grad_norm": 0.5855458378791809, "learning_rate": 9.940773141196357e-05, "loss": 0.0639, "step": 9660 }, { "grad_norm": 0.5420619249343872, "learning_rate": 9.94051912709176e-05, "loss": 0.0797, "step": 9670 }, { "grad_norm": 0.5095888376235962, "learning_rate": 9.940264572699421e-05, "loss": 0.0604, "step": 9680 }, { "grad_norm": 0.44878289103507996, "learning_rate": 9.940009478047174e-05, "loss": 0.0708, "step": 9690 }, { "grad_norm": 0.6318920850753784, "learning_rate": 9.939753843162918e-05, "loss": 0.0623, "step": 9700 }, { "grad_norm": 0.7163341641426086, "learning_rate": 9.939497668074609e-05, "loss": 0.0624, "step": 9710 }, { "grad_norm": 0.5771652460098267, "learning_rate": 9.93924095281026e-05, "loss": 0.0674, "step": 9720 }, { "grad_norm": 0.546261727809906, "learning_rate": 9.938983697397948e-05, "loss": 0.0806, "step": 9730 }, { "grad_norm": 0.4425870478153229, "learning_rate": 9.938725901865805e-05, "loss": 0.0707, "step": 9740 }, { "grad_norm": 0.45365574955940247, "learning_rate": 9.93846756624202e-05, "loss": 0.0714, "step": 9750 }, { "grad_norm": 0.5796294808387756, "learning_rate": 9.938208690554849e-05, "loss": 0.0662, "step": 9760 }, { "grad_norm": 0.6010634899139404, "learning_rate": 9.9379492748326e-05, "loss": 0.0685, "step": 9770 }, { "grad_norm": 0.5293237566947937, "learning_rate": 9.937689319103641e-05, "loss": 0.0685, "step": 9780 }, { "grad_norm": 0.5989262461662292, "learning_rate": 9.937428823396404e-05, "loss": 0.0696, "step": 9790 }, { "grad_norm": 0.4962805509567261, "learning_rate": 9.937167787739372e-05, "loss": 0.0611, "step": 9800 }, { "grad_norm": 0.5710523128509521, "learning_rate": 9.936906212161095e-05, "loss": 0.0711, "step": 9810 }, { "grad_norm": 0.4233280420303345, "learning_rate": 9.936644096690176e-05, "loss": 0.0773, "step": 9820 }, { "grad_norm": 0.5138810276985168, "learning_rate": 9.936381441355282e-05, "loss": 0.0742, "step": 9830 }, { "grad_norm": 0.46146121621131897, "learning_rate": 9.936118246185136e-05, "loss": 0.0731, "step": 9840 }, { "grad_norm": 0.5068540573120117, "learning_rate": 9.935854511208518e-05, "loss": 0.074, "step": 9850 }, { "grad_norm": 0.4964486360549927, "learning_rate": 9.935590236454272e-05, "loss": 0.0579, "step": 9860 }, { "grad_norm": 0.5363719463348389, "learning_rate": 9.935325421951298e-05, "loss": 0.062, "step": 9870 }, { "grad_norm": 0.6598889231681824, "learning_rate": 9.935060067728557e-05, "loss": 0.0791, "step": 9880 }, { "grad_norm": 0.3544688820838928, "learning_rate": 9.934794173815067e-05, "loss": 0.0639, "step": 9890 }, { "grad_norm": 0.6858711242675781, "learning_rate": 9.934527740239906e-05, "loss": 0.0767, "step": 9900 }, { "grad_norm": 0.46804267168045044, "learning_rate": 9.934260767032209e-05, "loss": 0.0709, "step": 9910 }, { "grad_norm": 0.4878183901309967, "learning_rate": 9.933993254221172e-05, "loss": 0.0728, "step": 9920 }, { "grad_norm": 0.5147188901901245, "learning_rate": 9.933725201836053e-05, "loss": 0.0595, "step": 9930 }, { "grad_norm": 0.49823930859565735, "learning_rate": 9.933456609906162e-05, "loss": 0.0631, "step": 9940 }, { "grad_norm": 0.5276349186897278, "learning_rate": 9.933187478460875e-05, "loss": 0.0614, "step": 9950 }, { "grad_norm": 0.4868726134300232, "learning_rate": 9.93291780752962e-05, "loss": 0.056, "step": 9960 }, { "grad_norm": 0.418486088514328, "learning_rate": 9.932647597141893e-05, "loss": 0.0668, "step": 9970 }, { "grad_norm": 0.4464324116706848, "learning_rate": 9.932376847327239e-05, "loss": 0.0628, "step": 9980 }, { "grad_norm": 0.5591195225715637, "learning_rate": 9.932105558115268e-05, "loss": 0.0652, "step": 9990 }, { "grad_norm": 0.6557487845420837, "learning_rate": 9.931833729535651e-05, "loss": 0.0786, "step": 10000 }, { "grad_norm": 0.5654444098472595, "learning_rate": 9.931561361618111e-05, "loss": 0.0691, "step": 10010 }, { "grad_norm": 0.40052512288093567, "learning_rate": 9.931288454392435e-05, "loss": 0.0696, "step": 10020 }, { "grad_norm": 1.0095646381378174, "learning_rate": 9.931015007888467e-05, "loss": 0.0658, "step": 10030 }, { "grad_norm": 0.708981454372406, "learning_rate": 9.930741022136112e-05, "loss": 0.0729, "step": 10040 }, { "grad_norm": 0.5627617239952087, "learning_rate": 9.930466497165333e-05, "loss": 0.0784, "step": 10050 }, { "grad_norm": 0.5586085319519043, "learning_rate": 9.93019143300615e-05, "loss": 0.0689, "step": 10060 }, { "grad_norm": 0.4309070408344269, "learning_rate": 9.929915829688644e-05, "loss": 0.0554, "step": 10070 }, { "grad_norm": 0.4841453731060028, "learning_rate": 9.929639687242955e-05, "loss": 0.0702, "step": 10080 }, { "grad_norm": 0.42601558566093445, "learning_rate": 9.929363005699281e-05, "loss": 0.0605, "step": 10090 }, { "grad_norm": 0.5236645340919495, "learning_rate": 9.92908578508788e-05, "loss": 0.0756, "step": 10100 }, { "grad_norm": 0.5388858318328857, "learning_rate": 9.928808025439069e-05, "loss": 0.0694, "step": 10110 }, { "grad_norm": 0.40773242712020874, "learning_rate": 9.928529726783223e-05, "loss": 0.0586, "step": 10120 }, { "grad_norm": 0.586212694644928, "learning_rate": 9.928250889150774e-05, "loss": 0.0773, "step": 10130 }, { "grad_norm": 0.519571840763092, "learning_rate": 9.92797151257222e-05, "loss": 0.0575, "step": 10140 }, { "grad_norm": 0.7756749987602234, "learning_rate": 9.927691597078108e-05, "loss": 0.0737, "step": 10150 }, { "grad_norm": 0.6925562620162964, "learning_rate": 9.927411142699053e-05, "loss": 0.0795, "step": 10160 }, { "grad_norm": 0.5676160454750061, "learning_rate": 9.927130149465725e-05, "loss": 0.0648, "step": 10170 }, { "grad_norm": 0.49927958846092224, "learning_rate": 9.92684861740885e-05, "loss": 0.0808, "step": 10180 }, { "grad_norm": 0.6014199256896973, "learning_rate": 9.926566546559217e-05, "loss": 0.084, "step": 10190 }, { "grad_norm": 0.5000758767127991, "learning_rate": 9.926283936947673e-05, "loss": 0.0721, "step": 10200 }, { "grad_norm": 0.4765237867832184, "learning_rate": 9.926000788605126e-05, "loss": 0.0658, "step": 10210 }, { "grad_norm": 0.46123260259628296, "learning_rate": 9.92571710156254e-05, "loss": 0.0617, "step": 10220 }, { "grad_norm": 0.48660382628440857, "learning_rate": 9.925432875850936e-05, "loss": 0.0656, "step": 10230 }, { "grad_norm": 0.4532853662967682, "learning_rate": 9.925148111501396e-05, "loss": 0.0662, "step": 10240 }, { "grad_norm": 0.6842005848884583, "learning_rate": 9.924862808545066e-05, "loss": 0.0716, "step": 10250 }, { "grad_norm": 0.5476815700531006, "learning_rate": 9.924576967013141e-05, "loss": 0.0666, "step": 10260 }, { "grad_norm": 0.7170971632003784, "learning_rate": 9.924290586936887e-05, "loss": 0.0752, "step": 10270 }, { "grad_norm": 0.6494933366775513, "learning_rate": 9.924003668347614e-05, "loss": 0.0678, "step": 10280 }, { "grad_norm": 0.4355655610561371, "learning_rate": 9.923716211276704e-05, "loss": 0.0797, "step": 10290 }, { "grad_norm": 0.6123600602149963, "learning_rate": 9.923428215755594e-05, "loss": 0.0674, "step": 10300 }, { "grad_norm": 0.42120087146759033, "learning_rate": 9.923139681815775e-05, "loss": 0.0583, "step": 10310 }, { "grad_norm": 0.4543569087982178, "learning_rate": 9.922850609488801e-05, "loss": 0.0636, "step": 10320 }, { "grad_norm": 0.712999165058136, "learning_rate": 9.922560998806287e-05, "loss": 0.0692, "step": 10330 }, { "grad_norm": 0.4371901750564575, "learning_rate": 9.922270849799905e-05, "loss": 0.0625, "step": 10340 }, { "grad_norm": 0.44084587693214417, "learning_rate": 9.92198016250138e-05, "loss": 0.0677, "step": 10350 }, { "grad_norm": 0.38911595940589905, "learning_rate": 9.921688936942506e-05, "loss": 0.0601, "step": 10360 }, { "grad_norm": 0.5000162124633789, "learning_rate": 9.921397173155129e-05, "loss": 0.0667, "step": 10370 }, { "grad_norm": 0.6227226257324219, "learning_rate": 9.921104871171157e-05, "loss": 0.0675, "step": 10380 }, { "grad_norm": 0.3981127440929413, "learning_rate": 9.920812031022554e-05, "loss": 0.0625, "step": 10390 }, { "grad_norm": 0.3930802643299103, "learning_rate": 9.920518652741348e-05, "loss": 0.0651, "step": 10400 }, { "grad_norm": 0.34237203001976013, "learning_rate": 9.920224736359618e-05, "loss": 0.0652, "step": 10410 }, { "grad_norm": 0.5428934097290039, "learning_rate": 9.91993028190951e-05, "loss": 0.0636, "step": 10420 }, { "grad_norm": 0.4820058047771454, "learning_rate": 9.919635289423222e-05, "loss": 0.0619, "step": 10430 }, { "grad_norm": 0.4861767888069153, "learning_rate": 9.919339758933015e-05, "loss": 0.052, "step": 10440 }, { "grad_norm": 0.6974349021911621, "learning_rate": 9.919043690471209e-05, "loss": 0.0631, "step": 10450 }, { "grad_norm": 0.529112696647644, "learning_rate": 9.91874708407018e-05, "loss": 0.0581, "step": 10460 }, { "grad_norm": 0.459317684173584, "learning_rate": 9.918449939762367e-05, "loss": 0.0609, "step": 10470 }, { "grad_norm": 0.45899349451065063, "learning_rate": 9.91815225758026e-05, "loss": 0.0499, "step": 10480 }, { "grad_norm": 0.4380313456058502, "learning_rate": 9.917854037556419e-05, "loss": 0.0638, "step": 10490 }, { "grad_norm": 0.4881557524204254, "learning_rate": 9.917555279723454e-05, "loss": 0.0574, "step": 10500 }, { "grad_norm": 0.48746880888938904, "learning_rate": 9.917255984114036e-05, "loss": 0.0691, "step": 10510 }, { "grad_norm": 0.49367788434028625, "learning_rate": 9.916956150760896e-05, "loss": 0.0526, "step": 10520 }, { "grad_norm": 0.44259563088417053, "learning_rate": 9.916655779696826e-05, "loss": 0.0767, "step": 10530 }, { "grad_norm": 0.4886089265346527, "learning_rate": 9.916354870954671e-05, "loss": 0.0628, "step": 10540 }, { "grad_norm": 0.4304359257221222, "learning_rate": 9.91605342456734e-05, "loss": 0.0527, "step": 10550 }, { "grad_norm": 0.6569833755493164, "learning_rate": 9.915751440567795e-05, "loss": 0.0528, "step": 10560 }, { "grad_norm": 0.4880978465080261, "learning_rate": 9.915448918989066e-05, "loss": 0.0526, "step": 10570 }, { "grad_norm": 0.49765703082084656, "learning_rate": 9.915145859864232e-05, "loss": 0.0661, "step": 10580 }, { "grad_norm": 0.4521125853061676, "learning_rate": 9.914842263226437e-05, "loss": 0.0677, "step": 10590 }, { "grad_norm": 0.44766858220100403, "learning_rate": 9.914538129108882e-05, "loss": 0.0716, "step": 10600 }, { "grad_norm": 0.42137688398361206, "learning_rate": 9.914233457544825e-05, "loss": 0.0831, "step": 10610 }, { "grad_norm": 0.5621914863586426, "learning_rate": 9.913928248567586e-05, "loss": 0.0846, "step": 10620 }, { "grad_norm": 0.5556544661521912, "learning_rate": 9.913622502210542e-05, "loss": 0.065, "step": 10630 }, { "grad_norm": 0.5270330905914307, "learning_rate": 9.913316218507128e-05, "loss": 0.0761, "step": 10640 }, { "grad_norm": 0.6380853652954102, "learning_rate": 9.91300939749084e-05, "loss": 0.0713, "step": 10650 }, { "grad_norm": 0.4823300540447235, "learning_rate": 9.91270203919523e-05, "loss": 0.0643, "step": 10660 }, { "grad_norm": 0.6614582538604736, "learning_rate": 9.912394143653912e-05, "loss": 0.0624, "step": 10670 }, { "grad_norm": 0.7490034103393555, "learning_rate": 9.912085710900555e-05, "loss": 0.0541, "step": 10680 }, { "grad_norm": 0.42590823769569397, "learning_rate": 9.911776740968892e-05, "loss": 0.0672, "step": 10690 }, { "grad_norm": 0.3610494136810303, "learning_rate": 9.911467233892709e-05, "loss": 0.0858, "step": 10700 }, { "grad_norm": 0.38091012835502625, "learning_rate": 9.911157189705853e-05, "loss": 0.0736, "step": 10710 }, { "grad_norm": 0.6388312578201294, "learning_rate": 9.910846608442229e-05, "loss": 0.064, "step": 10720 }, { "grad_norm": 0.45791617035865784, "learning_rate": 9.910535490135805e-05, "loss": 0.0606, "step": 10730 }, { "grad_norm": 0.5158997774124146, "learning_rate": 9.910223834820603e-05, "loss": 0.0616, "step": 10740 }, { "grad_norm": 0.46154463291168213, "learning_rate": 9.909911642530703e-05, "loss": 0.0686, "step": 10750 }, { "grad_norm": 0.5349596738815308, "learning_rate": 9.909598913300249e-05, "loss": 0.0705, "step": 10760 }, { "grad_norm": 0.7954201698303223, "learning_rate": 9.909285647163438e-05, "loss": 0.0763, "step": 10770 }, { "grad_norm": 0.5556876063346863, "learning_rate": 9.908971844154531e-05, "loss": 0.0694, "step": 10780 }, { "grad_norm": 0.7259542942047119, "learning_rate": 9.908657504307843e-05, "loss": 0.0821, "step": 10790 }, { "grad_norm": 0.5605480670928955, "learning_rate": 9.908342627657751e-05, "loss": 0.0798, "step": 10800 }, { "grad_norm": 0.7077547907829285, "learning_rate": 9.908027214238689e-05, "loss": 0.0726, "step": 10810 }, { "grad_norm": 0.4733200967311859, "learning_rate": 9.90771126408515e-05, "loss": 0.0684, "step": 10820 }, { "grad_norm": 0.40319815278053284, "learning_rate": 9.907394777231685e-05, "loss": 0.0672, "step": 10830 }, { "grad_norm": 0.6216644644737244, "learning_rate": 9.907077753712905e-05, "loss": 0.0781, "step": 10840 }, { "grad_norm": 0.625967264175415, "learning_rate": 9.906760193563482e-05, "loss": 0.0759, "step": 10850 }, { "grad_norm": 0.4406677484512329, "learning_rate": 9.906442096818139e-05, "loss": 0.0527, "step": 10860 }, { "grad_norm": 0.5831046104431152, "learning_rate": 9.906123463511665e-05, "loss": 0.0621, "step": 10870 }, { "grad_norm": 0.5291848182678223, "learning_rate": 9.905804293678907e-05, "loss": 0.0674, "step": 10880 }, { "grad_norm": 0.411221444606781, "learning_rate": 9.905484587354766e-05, "loss": 0.0634, "step": 10890 }, { "grad_norm": 0.5117865800857544, "learning_rate": 9.905164344574205e-05, "loss": 0.0726, "step": 10900 }, { "grad_norm": 0.5142978429794312, "learning_rate": 9.904843565372248e-05, "loss": 0.0597, "step": 10910 }, { "grad_norm": 0.7242946028709412, "learning_rate": 9.904522249783972e-05, "loss": 0.0663, "step": 10920 }, { "grad_norm": 0.538405179977417, "learning_rate": 9.904200397844517e-05, "loss": 0.0831, "step": 10930 }, { "grad_norm": 0.6045331358909607, "learning_rate": 9.903878009589078e-05, "loss": 0.0728, "step": 10940 }, { "grad_norm": 0.570573091506958, "learning_rate": 9.903555085052915e-05, "loss": 0.068, "step": 10950 }, { "grad_norm": 0.3853062093257904, "learning_rate": 9.903231624271338e-05, "loss": 0.0537, "step": 10960 }, { "grad_norm": 0.5950727462768555, "learning_rate": 9.902907627279724e-05, "loss": 0.0708, "step": 10970 }, { "grad_norm": 0.6759238243103027, "learning_rate": 9.902583094113504e-05, "loss": 0.0726, "step": 10980 }, { "grad_norm": 0.8317592144012451, "learning_rate": 9.902258024808168e-05, "loss": 0.0674, "step": 10990 }, { "grad_norm": 0.5725681185722351, "learning_rate": 9.901932419399264e-05, "loss": 0.0731, "step": 11000 }, { "grad_norm": 0.6071702241897583, "learning_rate": 9.9016062779224e-05, "loss": 0.0645, "step": 11010 }, { "grad_norm": 0.6045878529548645, "learning_rate": 9.901279600413242e-05, "loss": 0.0619, "step": 11020 }, { "grad_norm": 0.5189881920814514, "learning_rate": 9.900952386907518e-05, "loss": 0.0654, "step": 11030 }, { "grad_norm": 0.552052915096283, "learning_rate": 9.90062463744101e-05, "loss": 0.0608, "step": 11040 }, { "grad_norm": 0.5090691447257996, "learning_rate": 9.900296352049558e-05, "loss": 0.0609, "step": 11050 }, { "grad_norm": 0.5223101377487183, "learning_rate": 9.899967530769065e-05, "loss": 0.0689, "step": 11060 }, { "grad_norm": 0.5479960441589355, "learning_rate": 9.899638173635489e-05, "loss": 0.068, "step": 11070 }, { "grad_norm": 0.669917643070221, "learning_rate": 9.899308280684849e-05, "loss": 0.0664, "step": 11080 }, { "grad_norm": 0.4911336600780487, "learning_rate": 9.898977851953222e-05, "loss": 0.0616, "step": 11090 }, { "grad_norm": 0.44860777258872986, "learning_rate": 9.898646887476741e-05, "loss": 0.0694, "step": 11100 }, { "grad_norm": 0.5243535041809082, "learning_rate": 9.898315387291603e-05, "loss": 0.0677, "step": 11110 }, { "grad_norm": 0.5619650483131409, "learning_rate": 9.89798335143406e-05, "loss": 0.0612, "step": 11120 }, { "grad_norm": 0.45142531394958496, "learning_rate": 9.897650779940419e-05, "loss": 0.069, "step": 11130 }, { "grad_norm": 0.43571949005126953, "learning_rate": 9.897317672847054e-05, "loss": 0.0611, "step": 11140 }, { "grad_norm": 0.4767857491970062, "learning_rate": 9.89698403019039e-05, "loss": 0.0501, "step": 11150 }, { "grad_norm": 0.6538906097412109, "learning_rate": 9.896649852006917e-05, "loss": 0.0613, "step": 11160 }, { "grad_norm": 0.45703259110450745, "learning_rate": 9.896315138333177e-05, "loss": 0.0579, "step": 11170 }, { "grad_norm": 0.4529726803302765, "learning_rate": 9.895979889205774e-05, "loss": 0.0585, "step": 11180 }, { "grad_norm": 0.4149405062198639, "learning_rate": 9.895644104661372e-05, "loss": 0.0588, "step": 11190 }, { "grad_norm": 0.5399737358093262, "learning_rate": 9.895307784736691e-05, "loss": 0.0632, "step": 11200 }, { "grad_norm": 0.616788387298584, "learning_rate": 9.894970929468512e-05, "loss": 0.0698, "step": 11210 }, { "grad_norm": 0.5725165605545044, "learning_rate": 9.89463353889367e-05, "loss": 0.0601, "step": 11220 }, { "grad_norm": 0.5857970714569092, "learning_rate": 9.894295613049065e-05, "loss": 0.0764, "step": 11230 }, { "grad_norm": 0.5324479937553406, "learning_rate": 9.893957151971649e-05, "loss": 0.066, "step": 11240 }, { "grad_norm": 0.5493262410163879, "learning_rate": 9.893618155698436e-05, "loss": 0.0512, "step": 11250 }, { "grad_norm": 0.4642956256866455, "learning_rate": 9.8932786242665e-05, "loss": 0.0487, "step": 11260 }, { "grad_norm": 0.4806196093559265, "learning_rate": 9.89293855771297e-05, "loss": 0.0572, "step": 11270 }, { "grad_norm": 0.6493856310844421, "learning_rate": 9.892597956075036e-05, "loss": 0.0813, "step": 11280 }, { "grad_norm": 0.549777626991272, "learning_rate": 9.892256819389947e-05, "loss": 0.071, "step": 11290 }, { "grad_norm": 0.5200958251953125, "learning_rate": 9.891915147695006e-05, "loss": 0.0639, "step": 11300 }, { "grad_norm": 0.4967650771141052, "learning_rate": 9.891572941027577e-05, "loss": 0.0634, "step": 11310 }, { "grad_norm": 0.45256248116493225, "learning_rate": 9.89123019942509e-05, "loss": 0.0557, "step": 11320 }, { "grad_norm": 0.5500040054321289, "learning_rate": 9.89088692292502e-05, "loss": 0.0567, "step": 11330 }, { "grad_norm": 0.6080765128135681, "learning_rate": 9.89054311156491e-05, "loss": 0.0579, "step": 11340 }, { "grad_norm": 0.6339036226272583, "learning_rate": 9.890198765382357e-05, "loss": 0.0841, "step": 11350 }, { "grad_norm": 0.43842223286628723, "learning_rate": 9.889853884415021e-05, "loss": 0.0623, "step": 11360 }, { "grad_norm": 0.5006853342056274, "learning_rate": 9.889508468700614e-05, "loss": 0.0506, "step": 11370 }, { "grad_norm": 0.3744708299636841, "learning_rate": 9.889162518276915e-05, "loss": 0.0551, "step": 11380 }, { "grad_norm": 0.5169289708137512, "learning_rate": 9.888816033181752e-05, "loss": 0.0667, "step": 11390 }, { "grad_norm": 0.5601144433021545, "learning_rate": 9.888469013453018e-05, "loss": 0.0748, "step": 11400 }, { "grad_norm": 0.4630681276321411, "learning_rate": 9.888121459128663e-05, "loss": 0.0565, "step": 11410 }, { "grad_norm": 0.4573352038860321, "learning_rate": 9.887773370246693e-05, "loss": 0.0757, "step": 11420 }, { "grad_norm": 0.4263344407081604, "learning_rate": 9.887424746845177e-05, "loss": 0.0623, "step": 11430 }, { "grad_norm": 0.4934084713459015, "learning_rate": 9.887075588962239e-05, "loss": 0.0586, "step": 11440 }, { "grad_norm": 0.5452814102172852, "learning_rate": 9.88672589663606e-05, "loss": 0.0533, "step": 11450 }, { "grad_norm": 0.5341313481330872, "learning_rate": 9.886375669904886e-05, "loss": 0.0724, "step": 11460 }, { "grad_norm": 0.48120537400245667, "learning_rate": 9.886024908807014e-05, "loss": 0.0555, "step": 11470 }, { "grad_norm": 0.5112374424934387, "learning_rate": 9.885673613380806e-05, "loss": 0.057, "step": 11480 }, { "grad_norm": 0.5288184285163879, "learning_rate": 9.885321783664676e-05, "loss": 0.0703, "step": 11490 }, { "grad_norm": 0.41658490896224976, "learning_rate": 9.884969419697101e-05, "loss": 0.0613, "step": 11500 }, { "grad_norm": 0.584334135055542, "learning_rate": 9.884616521516614e-05, "loss": 0.0645, "step": 11510 }, { "grad_norm": 0.4533812403678894, "learning_rate": 9.88426308916181e-05, "loss": 0.0644, "step": 11520 }, { "grad_norm": 0.5511714816093445, "learning_rate": 9.883909122671335e-05, "loss": 0.0523, "step": 11530 }, { "grad_norm": 0.3060801029205322, "learning_rate": 9.883554622083904e-05, "loss": 0.0556, "step": 11540 }, { "grad_norm": 0.6608258485794067, "learning_rate": 9.88319958743828e-05, "loss": 0.0662, "step": 11550 }, { "grad_norm": 0.5029768943786621, "learning_rate": 9.882844018773291e-05, "loss": 0.0539, "step": 11560 }, { "grad_norm": 0.4484965205192566, "learning_rate": 9.882487916127823e-05, "loss": 0.06, "step": 11570 }, { "grad_norm": 0.4178920388221741, "learning_rate": 9.882131279540815e-05, "loss": 0.0641, "step": 11580 }, { "grad_norm": 0.5116952657699585, "learning_rate": 9.881774109051271e-05, "loss": 0.0641, "step": 11590 }, { "grad_norm": 0.471833199262619, "learning_rate": 9.881416404698252e-05, "loss": 0.0544, "step": 11600 }, { "grad_norm": 0.4075809717178345, "learning_rate": 9.881058166520873e-05, "loss": 0.0784, "step": 11610 }, { "grad_norm": 0.5837434530258179, "learning_rate": 9.880699394558311e-05, "loss": 0.0786, "step": 11620 }, { "grad_norm": 0.6551673412322998, "learning_rate": 9.880340088849801e-05, "loss": 0.0875, "step": 11630 }, { "grad_norm": 0.49196723103523254, "learning_rate": 9.879980249434637e-05, "loss": 0.0564, "step": 11640 }, { "grad_norm": 0.4740079343318939, "learning_rate": 9.879619876352168e-05, "loss": 0.0604, "step": 11650 }, { "grad_norm": 0.3486393392086029, "learning_rate": 9.879258969641809e-05, "loss": 0.0615, "step": 11660 }, { "grad_norm": 0.33549582958221436, "learning_rate": 9.878897529343023e-05, "loss": 0.0531, "step": 11670 }, { "grad_norm": 0.4291069507598877, "learning_rate": 9.878535555495338e-05, "loss": 0.0649, "step": 11680 }, { "grad_norm": 0.5801488161087036, "learning_rate": 9.87817304813834e-05, "loss": 0.0659, "step": 11690 }, { "grad_norm": 0.40472108125686646, "learning_rate": 9.877810007311671e-05, "loss": 0.0656, "step": 11700 }, { "grad_norm": 0.5151432752609253, "learning_rate": 9.877446433055035e-05, "loss": 0.0618, "step": 11710 }, { "grad_norm": 0.6300022602081299, "learning_rate": 9.877082325408191e-05, "loss": 0.058, "step": 11720 }, { "grad_norm": 0.3507976233959198, "learning_rate": 9.876717684410954e-05, "loss": 0.0539, "step": 11730 }, { "grad_norm": 0.5036069750785828, "learning_rate": 9.876352510103204e-05, "loss": 0.0596, "step": 11740 }, { "grad_norm": 0.4204433560371399, "learning_rate": 9.875986802524875e-05, "loss": 0.0651, "step": 11750 }, { "grad_norm": 0.3201007544994354, "learning_rate": 9.87562056171596e-05, "loss": 0.0555, "step": 11760 }, { "grad_norm": 0.5011787414550781, "learning_rate": 9.875253787716511e-05, "loss": 0.0577, "step": 11770 }, { "grad_norm": 0.6114243865013123, "learning_rate": 9.874886480566637e-05, "loss": 0.0524, "step": 11780 }, { "grad_norm": 0.4563624858856201, "learning_rate": 9.874518640306507e-05, "loss": 0.0523, "step": 11790 }, { "grad_norm": 0.6137670278549194, "learning_rate": 9.874150266976347e-05, "loss": 0.0555, "step": 11800 }, { "grad_norm": 0.44055458903312683, "learning_rate": 9.873781360616443e-05, "loss": 0.0621, "step": 11810 }, { "grad_norm": 0.46008872985839844, "learning_rate": 9.873411921267137e-05, "loss": 0.0578, "step": 11820 }, { "grad_norm": 0.6290793418884277, "learning_rate": 9.873041948968829e-05, "loss": 0.073, "step": 11830 }, { "grad_norm": 0.7248063683509827, "learning_rate": 9.872671443761981e-05, "loss": 0.0504, "step": 11840 }, { "grad_norm": 0.3680097162723541, "learning_rate": 9.872300405687109e-05, "loss": 0.0682, "step": 11850 }, { "grad_norm": 0.5430225133895874, "learning_rate": 9.871928834784792e-05, "loss": 0.0626, "step": 11860 }, { "grad_norm": 0.4842098653316498, "learning_rate": 9.871556731095661e-05, "loss": 0.0561, "step": 11870 }, { "grad_norm": 0.45845872163772583, "learning_rate": 9.871184094660411e-05, "loss": 0.0698, "step": 11880 }, { "grad_norm": 0.4488903284072876, "learning_rate": 9.870810925519791e-05, "loss": 0.0589, "step": 11890 }, { "grad_norm": 0.5781821012496948, "learning_rate": 9.870437223714612e-05, "loss": 0.0652, "step": 11900 }, { "grad_norm": 0.42565661668777466, "learning_rate": 9.87006298928574e-05, "loss": 0.0683, "step": 11910 }, { "grad_norm": 0.5607888102531433, "learning_rate": 9.869688222274103e-05, "loss": 0.0573, "step": 11920 }, { "grad_norm": 0.5041768550872803, "learning_rate": 9.869312922720681e-05, "loss": 0.0585, "step": 11930 }, { "grad_norm": 0.34845367074012756, "learning_rate": 9.868937090666521e-05, "loss": 0.0557, "step": 11940 }, { "grad_norm": 0.33021628856658936, "learning_rate": 9.86856072615272e-05, "loss": 0.044, "step": 11950 }, { "grad_norm": 0.5024991035461426, "learning_rate": 9.868183829220438e-05, "loss": 0.0549, "step": 11960 }, { "grad_norm": 0.5620566606521606, "learning_rate": 9.867806399910893e-05, "loss": 0.066, "step": 11970 }, { "grad_norm": 0.5800288915634155, "learning_rate": 9.867428438265356e-05, "loss": 0.0631, "step": 11980 }, { "grad_norm": 0.3499724268913269, "learning_rate": 9.867049944325165e-05, "loss": 0.061, "step": 11990 }, { "grad_norm": 0.5144320130348206, "learning_rate": 9.86667091813171e-05, "loss": 0.0691, "step": 12000 }, { "grad_norm": 0.5431612730026245, "learning_rate": 9.866291359726438e-05, "loss": 0.0657, "step": 12010 }, { "grad_norm": 0.4357426166534424, "learning_rate": 9.865911269150861e-05, "loss": 0.0652, "step": 12020 }, { "grad_norm": 0.4060547351837158, "learning_rate": 9.865530646446544e-05, "loss": 0.0519, "step": 12030 }, { "grad_norm": 0.5133303999900818, "learning_rate": 9.86514949165511e-05, "loss": 0.0674, "step": 12040 }, { "grad_norm": 0.4797093868255615, "learning_rate": 9.864767804818243e-05, "loss": 0.0636, "step": 12050 }, { "grad_norm": 0.7075560092926025, "learning_rate": 9.86438558597768e-05, "loss": 0.0692, "step": 12060 }, { "grad_norm": 0.49195149540901184, "learning_rate": 9.864002835175225e-05, "loss": 0.0885, "step": 12070 }, { "grad_norm": 0.5660260915756226, "learning_rate": 9.863619552452734e-05, "loss": 0.0586, "step": 12080 }, { "grad_norm": 0.6015569567680359, "learning_rate": 9.863235737852119e-05, "loss": 0.0547, "step": 12090 }, { "grad_norm": 0.5186397433280945, "learning_rate": 9.862851391415356e-05, "loss": 0.0663, "step": 12100 }, { "grad_norm": 0.5375561714172363, "learning_rate": 9.862466513184477e-05, "loss": 0.0663, "step": 12110 }, { "grad_norm": 0.6170011758804321, "learning_rate": 9.86208110320157e-05, "loss": 0.0512, "step": 12120 }, { "grad_norm": 0.43264269828796387, "learning_rate": 9.861695161508784e-05, "loss": 0.058, "step": 12130 }, { "grad_norm": 0.7098499536514282, "learning_rate": 9.861308688148324e-05, "loss": 0.0601, "step": 12140 }, { "grad_norm": 0.4770154058933258, "learning_rate": 9.860921683162455e-05, "loss": 0.0559, "step": 12150 }, { "grad_norm": 0.46528828144073486, "learning_rate": 9.860534146593499e-05, "loss": 0.0485, "step": 12160 }, { "grad_norm": 0.4735589325428009, "learning_rate": 9.860146078483836e-05, "loss": 0.0541, "step": 12170 }, { "grad_norm": 0.6755074858665466, "learning_rate": 9.859757478875905e-05, "loss": 0.0692, "step": 12180 }, { "grad_norm": 0.4937996566295624, "learning_rate": 9.859368347812204e-05, "loss": 0.0513, "step": 12190 }, { "grad_norm": 0.3602108657360077, "learning_rate": 9.858978685335285e-05, "loss": 0.0617, "step": 12200 }, { "grad_norm": 0.3111427426338196, "learning_rate": 9.858588491487763e-05, "loss": 0.0806, "step": 12210 }, { "grad_norm": 0.43645429611206055, "learning_rate": 9.858197766312308e-05, "loss": 0.0572, "step": 12220 }, { "grad_norm": 0.5299776196479797, "learning_rate": 9.857806509851649e-05, "loss": 0.0663, "step": 12230 }, { "grad_norm": 0.33590927720069885, "learning_rate": 9.857414722148574e-05, "loss": 0.0588, "step": 12240 }, { "grad_norm": 0.4129096567630768, "learning_rate": 9.857022403245928e-05, "loss": 0.074, "step": 12250 }, { "grad_norm": 0.45253074169158936, "learning_rate": 9.856629553186615e-05, "loss": 0.08, "step": 12260 }, { "grad_norm": 0.4708222448825836, "learning_rate": 9.856236172013595e-05, "loss": 0.0663, "step": 12270 }, { "grad_norm": 0.419109970331192, "learning_rate": 9.85584225976989e-05, "loss": 0.0546, "step": 12280 }, { "grad_norm": 0.39352649450302124, "learning_rate": 9.855447816498575e-05, "loss": 0.0677, "step": 12290 }, { "grad_norm": 0.39799371361732483, "learning_rate": 9.855052842242787e-05, "loss": 0.0491, "step": 12300 }, { "grad_norm": 0.5362225770950317, "learning_rate": 9.85465733704572e-05, "loss": 0.0554, "step": 12310 }, { "grad_norm": 0.5652870535850525, "learning_rate": 9.854261300950624e-05, "loss": 0.0614, "step": 12320 }, { "grad_norm": 0.46994510293006897, "learning_rate": 9.853864734000813e-05, "loss": 0.058, "step": 12330 }, { "grad_norm": 0.5188853144645691, "learning_rate": 9.85346763623965e-05, "loss": 0.0701, "step": 12340 }, { "grad_norm": 0.4745475649833679, "learning_rate": 9.853070007710564e-05, "loss": 0.0619, "step": 12350 }, { "grad_norm": 0.5190359354019165, "learning_rate": 9.85267184845704e-05, "loss": 0.0687, "step": 12360 }, { "grad_norm": 0.3804573118686676, "learning_rate": 9.852273158522616e-05, "loss": 0.0532, "step": 12370 }, { "grad_norm": 0.33860138058662415, "learning_rate": 9.851873937950896e-05, "loss": 0.0524, "step": 12380 }, { "grad_norm": 0.47668275237083435, "learning_rate": 9.851474186785537e-05, "loss": 0.0567, "step": 12390 }, { "grad_norm": 0.521683394908905, "learning_rate": 9.851073905070254e-05, "loss": 0.0713, "step": 12400 }, { "grad_norm": 0.6076673865318298, "learning_rate": 9.850673092848824e-05, "loss": 0.0652, "step": 12410 }, { "grad_norm": 0.4082970917224884, "learning_rate": 9.850271750165077e-05, "loss": 0.0528, "step": 12420 }, { "grad_norm": 0.6762171983718872, "learning_rate": 9.849869877062902e-05, "loss": 0.0682, "step": 12430 }, { "grad_norm": 0.4697347581386566, "learning_rate": 9.849467473586252e-05, "loss": 0.081, "step": 12440 }, { "grad_norm": 0.5001893043518066, "learning_rate": 9.849064539779127e-05, "loss": 0.058, "step": 12450 }, { "grad_norm": 0.4866797626018524, "learning_rate": 9.848661075685594e-05, "loss": 0.0709, "step": 12460 }, { "grad_norm": 0.36787375807762146, "learning_rate": 9.848257081349778e-05, "loss": 0.0674, "step": 12470 }, { "grad_norm": 0.4295869469642639, "learning_rate": 9.847852556815856e-05, "loss": 0.0572, "step": 12480 }, { "grad_norm": 0.49570927023887634, "learning_rate": 9.847447502128067e-05, "loss": 0.0627, "step": 12490 }, { "grad_norm": 0.45744588971138, "learning_rate": 9.847041917330708e-05, "loss": 0.0629, "step": 12500 }, { "grad_norm": 0.7348610758781433, "learning_rate": 9.846635802468132e-05, "loss": 0.0763, "step": 12510 }, { "grad_norm": 0.3810877501964569, "learning_rate": 9.84622915758475e-05, "loss": 0.0593, "step": 12520 }, { "grad_norm": 0.41823244094848633, "learning_rate": 9.845821982725034e-05, "loss": 0.0575, "step": 12530 }, { "grad_norm": 0.7388051748275757, "learning_rate": 9.845414277933514e-05, "loss": 0.0822, "step": 12540 }, { "grad_norm": 0.6232948303222656, "learning_rate": 9.845006043254771e-05, "loss": 0.0653, "step": 12550 }, { "grad_norm": 0.5877264142036438, "learning_rate": 9.844597278733451e-05, "loss": 0.0606, "step": 12560 }, { "grad_norm": 0.5340073108673096, "learning_rate": 9.844187984414259e-05, "loss": 0.0713, "step": 12570 }, { "grad_norm": 0.34123754501342773, "learning_rate": 9.84377816034195e-05, "loss": 0.059, "step": 12580 }, { "grad_norm": 0.533689022064209, "learning_rate": 9.843367806561345e-05, "loss": 0.0757, "step": 12590 }, { "grad_norm": 0.5362411141395569, "learning_rate": 9.842956923117317e-05, "loss": 0.0711, "step": 12600 }, { "grad_norm": 0.38829928636550903, "learning_rate": 9.842545510054802e-05, "loss": 0.062, "step": 12610 }, { "grad_norm": 0.6973136067390442, "learning_rate": 9.842133567418792e-05, "loss": 0.0596, "step": 12620 }, { "grad_norm": 0.5627150535583496, "learning_rate": 9.841721095254333e-05, "loss": 0.085, "step": 12630 }, { "grad_norm": 0.6407228112220764, "learning_rate": 9.841308093606537e-05, "loss": 0.0793, "step": 12640 }, { "grad_norm": 0.6456556916236877, "learning_rate": 9.840894562520565e-05, "loss": 0.0631, "step": 12650 }, { "grad_norm": 1.0031306743621826, "learning_rate": 9.840480502041642e-05, "loss": 0.0818, "step": 12660 }, { "grad_norm": 0.5712264776229858, "learning_rate": 9.840065912215049e-05, "loss": 0.0599, "step": 12670 }, { "grad_norm": 0.49814388155937195, "learning_rate": 9.839650793086124e-05, "loss": 0.0763, "step": 12680 }, { "grad_norm": 0.40721794962882996, "learning_rate": 9.839235144700265e-05, "loss": 0.0669, "step": 12690 }, { "grad_norm": 0.5706475973129272, "learning_rate": 9.838818967102926e-05, "loss": 0.0649, "step": 12700 }, { "grad_norm": 0.5278543829917908, "learning_rate": 9.83840226033962e-05, "loss": 0.0613, "step": 12710 }, { "grad_norm": 0.4920456111431122, "learning_rate": 9.837985024455918e-05, "loss": 0.0567, "step": 12720 }, { "grad_norm": 0.4516405165195465, "learning_rate": 9.837567259497447e-05, "loss": 0.0555, "step": 12730 }, { "grad_norm": 0.5370510816574097, "learning_rate": 9.837148965509894e-05, "loss": 0.0647, "step": 12740 }, { "grad_norm": 0.47115716338157654, "learning_rate": 9.836730142539001e-05, "loss": 0.0487, "step": 12750 }, { "grad_norm": 0.6030642986297607, "learning_rate": 9.836310790630574e-05, "loss": 0.0662, "step": 12760 }, { "grad_norm": 0.5799939632415771, "learning_rate": 9.83589090983047e-05, "loss": 0.0703, "step": 12770 }, { "grad_norm": 0.5134831666946411, "learning_rate": 9.835470500184605e-05, "loss": 0.0723, "step": 12780 }, { "grad_norm": 0.39602190256118774, "learning_rate": 9.835049561738957e-05, "loss": 0.0669, "step": 12790 }, { "grad_norm": 0.39604952931404114, "learning_rate": 9.834628094539558e-05, "loss": 0.0568, "step": 12800 }, { "grad_norm": 0.4912037253379822, "learning_rate": 9.834206098632499e-05, "loss": 0.0564, "step": 12810 }, { "grad_norm": 0.8779373168945312, "learning_rate": 9.833783574063931e-05, "loss": 0.0682, "step": 12820 }, { "grad_norm": 0.5522615909576416, "learning_rate": 9.833360520880058e-05, "loss": 0.0728, "step": 12830 }, { "grad_norm": 0.4353196322917938, "learning_rate": 9.832936939127144e-05, "loss": 0.0605, "step": 12840 }, { "grad_norm": 0.389595627784729, "learning_rate": 9.832512828851515e-05, "loss": 0.0633, "step": 12850 }, { "grad_norm": 0.3861190974712372, "learning_rate": 9.832088190099546e-05, "loss": 0.0619, "step": 12860 }, { "grad_norm": 0.5336025357246399, "learning_rate": 9.831663022917679e-05, "loss": 0.0527, "step": 12870 }, { "grad_norm": 0.5530706644058228, "learning_rate": 9.831237327352407e-05, "loss": 0.0517, "step": 12880 }, { "grad_norm": 0.4653171896934509, "learning_rate": 9.830811103450286e-05, "loss": 0.0549, "step": 12890 }, { "grad_norm": 0.5723361372947693, "learning_rate": 9.830384351257924e-05, "loss": 0.0677, "step": 12900 }, { "grad_norm": 0.3731954097747803, "learning_rate": 9.829957070821993e-05, "loss": 0.0495, "step": 12910 }, { "grad_norm": 0.38092994689941406, "learning_rate": 9.829529262189218e-05, "loss": 0.0438, "step": 12920 }, { "grad_norm": 0.42551347613334656, "learning_rate": 9.829100925406385e-05, "loss": 0.0582, "step": 12930 }, { "grad_norm": 0.6220239996910095, "learning_rate": 9.828672060520333e-05, "loss": 0.0597, "step": 12940 }, { "grad_norm": 0.39352309703826904, "learning_rate": 9.828242667577966e-05, "loss": 0.0482, "step": 12950 }, { "grad_norm": 0.4418482780456543, "learning_rate": 9.82781274662624e-05, "loss": 0.0568, "step": 12960 }, { "grad_norm": 0.5016655325889587, "learning_rate": 9.82738229771217e-05, "loss": 0.0575, "step": 12970 }, { "grad_norm": 0.5003180503845215, "learning_rate": 9.826951320882829e-05, "loss": 0.0513, "step": 12980 }, { "grad_norm": 0.7355349063873291, "learning_rate": 9.826519816185351e-05, "loss": 0.0782, "step": 12990 }, { "grad_norm": 0.6269692182540894, "learning_rate": 9.826087783666921e-05, "loss": 0.0692, "step": 13000 }, { "grad_norm": 0.5963457226753235, "learning_rate": 9.825655223374787e-05, "loss": 0.0819, "step": 13010 }, { "grad_norm": 0.6987277865409851, "learning_rate": 9.825222135356253e-05, "loss": 0.0815, "step": 13020 }, { "grad_norm": 0.4221702218055725, "learning_rate": 9.82478851965868e-05, "loss": 0.0744, "step": 13030 }, { "grad_norm": 0.529055655002594, "learning_rate": 9.82435437632949e-05, "loss": 0.0634, "step": 13040 }, { "grad_norm": 0.4906289577484131, "learning_rate": 9.823919705416158e-05, "loss": 0.052, "step": 13050 }, { "grad_norm": 0.6105188727378845, "learning_rate": 9.82348450696622e-05, "loss": 0.058, "step": 13060 }, { "grad_norm": 0.5264946818351746, "learning_rate": 9.823048781027268e-05, "loss": 0.0713, "step": 13070 }, { "grad_norm": 0.586860179901123, "learning_rate": 9.822612527646953e-05, "loss": 0.0623, "step": 13080 }, { "grad_norm": 0.382286012172699, "learning_rate": 9.822175746872984e-05, "loss": 0.0618, "step": 13090 }, { "grad_norm": 0.886946976184845, "learning_rate": 9.821738438753123e-05, "loss": 0.0672, "step": 13100 }, { "grad_norm": 0.42588406801223755, "learning_rate": 9.821300603335196e-05, "loss": 0.0587, "step": 13110 }, { "grad_norm": 0.6678914427757263, "learning_rate": 9.820862240667085e-05, "loss": 0.059, "step": 13120 }, { "grad_norm": 0.5497949123382568, "learning_rate": 9.820423350796726e-05, "loss": 0.0716, "step": 13130 }, { "grad_norm": 0.4780993163585663, "learning_rate": 9.819983933772118e-05, "loss": 0.0662, "step": 13140 }, { "grad_norm": 0.7166250348091125, "learning_rate": 9.819543989641314e-05, "loss": 0.067, "step": 13150 }, { "grad_norm": 0.6006630659103394, "learning_rate": 9.819103518452423e-05, "loss": 0.0567, "step": 13160 }, { "grad_norm": 0.4347664415836334, "learning_rate": 9.818662520253618e-05, "loss": 0.0603, "step": 13170 }, { "grad_norm": 0.366131454706192, "learning_rate": 9.818220995093126e-05, "loss": 0.0606, "step": 13180 }, { "grad_norm": 0.6732580661773682, "learning_rate": 9.817778943019228e-05, "loss": 0.0482, "step": 13190 }, { "grad_norm": 0.45557039976119995, "learning_rate": 9.81733636408027e-05, "loss": 0.0677, "step": 13200 }, { "grad_norm": 0.3015366196632385, "learning_rate": 9.816893258324649e-05, "loss": 0.0499, "step": 13210 }, { "grad_norm": 0.4311295449733734, "learning_rate": 9.816449625800823e-05, "loss": 0.0517, "step": 13220 }, { "grad_norm": 0.3870006799697876, "learning_rate": 9.816005466557308e-05, "loss": 0.0613, "step": 13230 }, { "grad_norm": 0.6535587310791016, "learning_rate": 9.815560780642674e-05, "loss": 0.0669, "step": 13240 }, { "grad_norm": 0.4004146158695221, "learning_rate": 9.815115568105555e-05, "loss": 0.0545, "step": 13250 }, { "grad_norm": 0.6815268397331238, "learning_rate": 9.814669828994638e-05, "loss": 0.0571, "step": 13260 }, { "grad_norm": 0.4296134412288666, "learning_rate": 9.814223563358665e-05, "loss": 0.049, "step": 13270 }, { "grad_norm": 0.4592629373073578, "learning_rate": 9.813776771246443e-05, "loss": 0.0564, "step": 13280 }, { "grad_norm": 0.3628746569156647, "learning_rate": 9.813329452706829e-05, "loss": 0.0525, "step": 13290 }, { "grad_norm": 0.3539280593395233, "learning_rate": 9.812881607788744e-05, "loss": 0.0484, "step": 13300 }, { "grad_norm": 0.412788450717926, "learning_rate": 9.812433236541163e-05, "loss": 0.0624, "step": 13310 }, { "grad_norm": 0.4618065655231476, "learning_rate": 9.811984339013116e-05, "loss": 0.064, "step": 13320 }, { "grad_norm": 0.43213775753974915, "learning_rate": 9.811534915253698e-05, "loss": 0.0526, "step": 13330 }, { "grad_norm": 0.4880825877189636, "learning_rate": 9.811084965312056e-05, "loss": 0.0533, "step": 13340 }, { "grad_norm": 0.3912440836429596, "learning_rate": 9.810634489237396e-05, "loss": 0.046, "step": 13350 }, { "grad_norm": 0.7030535936355591, "learning_rate": 9.81018348707898e-05, "loss": 0.0628, "step": 13360 }, { "grad_norm": 0.31989455223083496, "learning_rate": 9.809731958886131e-05, "loss": 0.064, "step": 13370 }, { "grad_norm": 0.5379091501235962, "learning_rate": 9.809279904708224e-05, "loss": 0.0572, "step": 13380 }, { "grad_norm": 0.48966625332832336, "learning_rate": 9.808827324594699e-05, "loss": 0.0576, "step": 13390 }, { "grad_norm": 0.42169681191444397, "learning_rate": 9.808374218595046e-05, "loss": 0.0637, "step": 13400 }, { "grad_norm": 0.44419407844543457, "learning_rate": 9.80792058675882e-05, "loss": 0.0522, "step": 13410 }, { "grad_norm": 0.4426301121711731, "learning_rate": 9.807466429135627e-05, "loss": 0.0544, "step": 13420 }, { "grad_norm": 0.5991729497909546, "learning_rate": 9.807011745775132e-05, "loss": 0.058, "step": 13430 }, { "grad_norm": 0.3248061537742615, "learning_rate": 9.806556536727061e-05, "loss": 0.0616, "step": 13440 }, { "grad_norm": 0.5847064256668091, "learning_rate": 9.806100802041193e-05, "loss": 0.0731, "step": 13450 }, { "grad_norm": 0.43884676694869995, "learning_rate": 9.805644541767368e-05, "loss": 0.0665, "step": 13460 }, { "grad_norm": 0.6106776595115662, "learning_rate": 9.805187755955478e-05, "loss": 0.0724, "step": 13470 }, { "grad_norm": 0.36380255222320557, "learning_rate": 9.804730444655483e-05, "loss": 0.0708, "step": 13480 }, { "grad_norm": 0.5598996877670288, "learning_rate": 9.804272607917388e-05, "loss": 0.0623, "step": 13490 }, { "grad_norm": 0.5029104351997375, "learning_rate": 9.803814245791265e-05, "loss": 0.0574, "step": 13500 }, { "grad_norm": 0.6517794728279114, "learning_rate": 9.803355358327239e-05, "loss": 0.0709, "step": 13510 }, { "grad_norm": 0.6824939250946045, "learning_rate": 9.802895945575492e-05, "loss": 0.0753, "step": 13520 }, { "grad_norm": 0.8271418809890747, "learning_rate": 9.802436007586266e-05, "loss": 0.0735, "step": 13530 }, { "grad_norm": 0.4630602300167084, "learning_rate": 9.801975544409858e-05, "loss": 0.0583, "step": 13540 }, { "grad_norm": 0.5518075823783875, "learning_rate": 9.801514556096625e-05, "loss": 0.0511, "step": 13550 }, { "grad_norm": 0.559907078742981, "learning_rate": 9.801053042696977e-05, "loss": 0.0592, "step": 13560 }, { "grad_norm": 0.4458353519439697, "learning_rate": 9.800591004261388e-05, "loss": 0.0549, "step": 13570 }, { "grad_norm": 0.4682933986186981, "learning_rate": 9.800128440840385e-05, "loss": 0.0679, "step": 13580 }, { "grad_norm": 0.7460141181945801, "learning_rate": 9.799665352484552e-05, "loss": 0.0723, "step": 13590 }, { "grad_norm": 0.5984426736831665, "learning_rate": 9.799201739244532e-05, "loss": 0.084, "step": 13600 }, { "grad_norm": 0.41003796458244324, "learning_rate": 9.798737601171025e-05, "loss": 0.0563, "step": 13610 }, { "grad_norm": 0.568618893623352, "learning_rate": 9.79827293831479e-05, "loss": 0.0751, "step": 13620 }, { "grad_norm": 0.5256080031394958, "learning_rate": 9.797807750726638e-05, "loss": 0.0561, "step": 13630 }, { "grad_norm": 0.48954734206199646, "learning_rate": 9.797342038457446e-05, "loss": 0.0547, "step": 13640 }, { "grad_norm": 0.5141096711158752, "learning_rate": 9.796875801558141e-05, "loss": 0.0554, "step": 13650 }, { "grad_norm": 0.5409485101699829, "learning_rate": 9.79640904007971e-05, "loss": 0.0657, "step": 13660 }, { "grad_norm": 0.37448132038116455, "learning_rate": 9.795941754073199e-05, "loss": 0.06, "step": 13670 }, { "grad_norm": 0.551504909992218, "learning_rate": 9.795473943589705e-05, "loss": 0.0649, "step": 13680 }, { "grad_norm": 0.5422839522361755, "learning_rate": 9.795005608680394e-05, "loss": 0.055, "step": 13690 }, { "grad_norm": 0.4711022973060608, "learning_rate": 9.794536749396477e-05, "loss": 0.0589, "step": 13700 }, { "grad_norm": 0.5010373592376709, "learning_rate": 9.79406736578923e-05, "loss": 0.0622, "step": 13710 }, { "grad_norm": 0.5411672592163086, "learning_rate": 9.793597457909984e-05, "loss": 0.0518, "step": 13720 }, { "grad_norm": 0.4342072606086731, "learning_rate": 9.793127025810127e-05, "loss": 0.0526, "step": 13730 }, { "grad_norm": 0.36319780349731445, "learning_rate": 9.792656069541104e-05, "loss": 0.0549, "step": 13740 }, { "grad_norm": 0.5561398863792419, "learning_rate": 9.79218458915442e-05, "loss": 0.0582, "step": 13750 }, { "grad_norm": 0.5810701251029968, "learning_rate": 9.791712584701634e-05, "loss": 0.0769, "step": 13760 }, { "grad_norm": 0.5339515805244446, "learning_rate": 9.791240056234364e-05, "loss": 0.0565, "step": 13770 }, { "grad_norm": 0.46234557032585144, "learning_rate": 9.790767003804283e-05, "loss": 0.0508, "step": 13780 }, { "grad_norm": 0.49482375383377075, "learning_rate": 9.790293427463126e-05, "loss": 0.0524, "step": 13790 }, { "grad_norm": 0.4137812852859497, "learning_rate": 9.789819327262684e-05, "loss": 0.0548, "step": 13800 }, { "grad_norm": 0.4253414273262024, "learning_rate": 9.7893447032548e-05, "loss": 0.0626, "step": 13810 }, { "grad_norm": 0.5357803106307983, "learning_rate": 9.78886955549138e-05, "loss": 0.0615, "step": 13820 }, { "grad_norm": 0.5512624382972717, "learning_rate": 9.788393884024387e-05, "loss": 0.0622, "step": 13830 }, { "grad_norm": 0.427130788564682, "learning_rate": 9.787917688905836e-05, "loss": 0.0621, "step": 13840 }, { "grad_norm": 0.4544336199760437, "learning_rate": 9.787440970187807e-05, "loss": 0.0554, "step": 13850 }, { "grad_norm": 0.39730602502822876, "learning_rate": 9.786963727922429e-05, "loss": 0.0687, "step": 13860 }, { "grad_norm": 0.4066588878631592, "learning_rate": 9.786485962161897e-05, "loss": 0.0595, "step": 13870 }, { "grad_norm": 0.3810301423072815, "learning_rate": 9.786007672958455e-05, "loss": 0.046, "step": 13880 }, { "grad_norm": 0.6201302409172058, "learning_rate": 9.78552886036441e-05, "loss": 0.0707, "step": 13890 }, { "grad_norm": 0.5243633985519409, "learning_rate": 9.785049524432124e-05, "loss": 0.0498, "step": 13900 }, { "grad_norm": 0.38536059856414795, "learning_rate": 9.784569665214016e-05, "loss": 0.0522, "step": 13910 }, { "grad_norm": 0.5276191830635071, "learning_rate": 9.784089282762563e-05, "loss": 0.0682, "step": 13920 }, { "grad_norm": 0.48458346724510193, "learning_rate": 9.7836083771303e-05, "loss": 0.0546, "step": 13930 }, { "grad_norm": 0.5109743475914001, "learning_rate": 9.783126948369817e-05, "loss": 0.0606, "step": 13940 }, { "grad_norm": 0.9013217091560364, "learning_rate": 9.78264499653376e-05, "loss": 0.0605, "step": 13950 }, { "grad_norm": 0.38696661591529846, "learning_rate": 9.782162521674838e-05, "loss": 0.0534, "step": 13960 }, { "grad_norm": 0.42934054136276245, "learning_rate": 9.781679523845812e-05, "loss": 0.0912, "step": 13970 }, { "grad_norm": 0.5642206072807312, "learning_rate": 9.781196003099502e-05, "loss": 0.0681, "step": 13980 }, { "grad_norm": 0.47715628147125244, "learning_rate": 9.780711959488786e-05, "loss": 0.0709, "step": 13990 }, { "grad_norm": 0.5096250772476196, "learning_rate": 9.780227393066599e-05, "loss": 0.0604, "step": 14000 }, { "grad_norm": 0.36128202080726624, "learning_rate": 9.77974230388593e-05, "loss": 0.0536, "step": 14010 }, { "grad_norm": 0.4073948860168457, "learning_rate": 9.779256691999829e-05, "loss": 0.0688, "step": 14020 }, { "grad_norm": 0.4191538095474243, "learning_rate": 9.778770557461403e-05, "loss": 0.0656, "step": 14030 }, { "grad_norm": 0.45988067984580994, "learning_rate": 9.778283900323812e-05, "loss": 0.0566, "step": 14040 }, { "grad_norm": 0.5867590308189392, "learning_rate": 9.777796720640277e-05, "loss": 0.0614, "step": 14050 }, { "grad_norm": 0.520470142364502, "learning_rate": 9.777309018464078e-05, "loss": 0.0668, "step": 14060 }, { "grad_norm": 0.4773167073726654, "learning_rate": 9.776820793848547e-05, "loss": 0.0538, "step": 14070 }, { "grad_norm": 0.4592961370944977, "learning_rate": 9.776332046847075e-05, "loss": 0.0538, "step": 14080 }, { "grad_norm": 0.6215769648551941, "learning_rate": 9.775842777513111e-05, "loss": 0.0551, "step": 14090 }, { "grad_norm": 0.4129979908466339, "learning_rate": 9.775352985900163e-05, "loss": 0.0616, "step": 14100 }, { "grad_norm": 0.5435752272605896, "learning_rate": 9.774862672061791e-05, "loss": 0.0527, "step": 14110 }, { "grad_norm": 0.4685250222682953, "learning_rate": 9.774371836051616e-05, "loss": 0.0546, "step": 14120 }, { "grad_norm": 0.5856032371520996, "learning_rate": 9.773880477923315e-05, "loss": 0.0687, "step": 14130 }, { "grad_norm": 0.427096962928772, "learning_rate": 9.773388597730623e-05, "loss": 0.0679, "step": 14140 }, { "grad_norm": 0.47327786684036255, "learning_rate": 9.77289619552733e-05, "loss": 0.0541, "step": 14150 }, { "grad_norm": 0.5140852928161621, "learning_rate": 9.772403271367285e-05, "loss": 0.0813, "step": 14160 }, { "grad_norm": 0.5258190631866455, "learning_rate": 9.771909825304396e-05, "loss": 0.0618, "step": 14170 }, { "grad_norm": 0.6356775164604187, "learning_rate": 9.771415857392619e-05, "loss": 0.0662, "step": 14180 }, { "grad_norm": 0.6474205255508423, "learning_rate": 9.770921367685978e-05, "loss": 0.0757, "step": 14190 }, { "grad_norm": 0.47919273376464844, "learning_rate": 9.770426356238551e-05, "loss": 0.0627, "step": 14200 }, { "grad_norm": 0.4639676511287689, "learning_rate": 9.769930823104469e-05, "loss": 0.0548, "step": 14210 }, { "grad_norm": 0.5097461342811584, "learning_rate": 9.769434768337926e-05, "loss": 0.0678, "step": 14220 }, { "grad_norm": 0.4540160298347473, "learning_rate": 9.768938191993164e-05, "loss": 0.059, "step": 14230 }, { "grad_norm": 0.5864942073822021, "learning_rate": 9.768441094124494e-05, "loss": 0.0563, "step": 14240 }, { "grad_norm": 0.39951205253601074, "learning_rate": 9.767943474786275e-05, "loss": 0.0622, "step": 14250 }, { "grad_norm": 0.3744792640209198, "learning_rate": 9.767445334032923e-05, "loss": 0.0431, "step": 14260 }, { "grad_norm": 0.4375852644443512, "learning_rate": 9.766946671918919e-05, "loss": 0.055, "step": 14270 }, { "grad_norm": 0.45479774475097656, "learning_rate": 9.766447488498796e-05, "loss": 0.0461, "step": 14280 }, { "grad_norm": 0.5553463697433472, "learning_rate": 9.765947783827139e-05, "loss": 0.0523, "step": 14290 }, { "grad_norm": 0.6156951189041138, "learning_rate": 9.765447557958599e-05, "loss": 0.0679, "step": 14300 }, { "grad_norm": 0.48576197028160095, "learning_rate": 9.764946810947879e-05, "loss": 0.0517, "step": 14310 }, { "grad_norm": 0.40400856733322144, "learning_rate": 9.764445542849738e-05, "loss": 0.0613, "step": 14320 }, { "grad_norm": 0.460017591714859, "learning_rate": 9.763943753718998e-05, "loss": 0.0614, "step": 14330 }, { "grad_norm": 0.5033335089683533, "learning_rate": 9.76344144361053e-05, "loss": 0.0647, "step": 14340 }, { "grad_norm": 0.5202791690826416, "learning_rate": 9.762938612579269e-05, "loss": 0.0521, "step": 14350 }, { "grad_norm": 0.38523954153060913, "learning_rate": 9.762435260680202e-05, "loss": 0.045, "step": 14360 }, { "grad_norm": 0.5455633997917175, "learning_rate": 9.761931387968373e-05, "loss": 0.0545, "step": 14370 }, { "grad_norm": 0.7863775491714478, "learning_rate": 9.76142699449889e-05, "loss": 0.0521, "step": 14380 }, { "grad_norm": 0.3726328909397125, "learning_rate": 9.760922080326908e-05, "loss": 0.0631, "step": 14390 }, { "grad_norm": 0.4443886876106262, "learning_rate": 9.760416645507644e-05, "loss": 0.0523, "step": 14400 }, { "grad_norm": 0.7977640628814697, "learning_rate": 9.759910690096375e-05, "loss": 0.0578, "step": 14410 }, { "grad_norm": 0.726569414138794, "learning_rate": 9.759404214148429e-05, "loss": 0.0515, "step": 14420 }, { "grad_norm": 0.4605201482772827, "learning_rate": 9.758897217719191e-05, "loss": 0.0538, "step": 14430 }, { "grad_norm": 0.5882713198661804, "learning_rate": 9.758389700864113e-05, "loss": 0.0477, "step": 14440 }, { "grad_norm": 0.43685421347618103, "learning_rate": 9.757881663638688e-05, "loss": 0.0513, "step": 14450 }, { "grad_norm": 0.41898226737976074, "learning_rate": 9.757373106098478e-05, "loss": 0.0525, "step": 14460 }, { "grad_norm": 0.47877171635627747, "learning_rate": 9.756864028299097e-05, "loss": 0.0577, "step": 14470 }, { "grad_norm": 0.3748196065425873, "learning_rate": 9.75635443029622e-05, "loss": 0.045, "step": 14480 }, { "grad_norm": 0.4838070869445801, "learning_rate": 9.755844312145572e-05, "loss": 0.0601, "step": 14490 }, { "grad_norm": 0.3906629979610443, "learning_rate": 9.755333673902941e-05, "loss": 0.0694, "step": 14500 }, { "grad_norm": 0.4371432960033417, "learning_rate": 9.75482251562417e-05, "loss": 0.0599, "step": 14510 }, { "grad_norm": 0.4472096264362335, "learning_rate": 9.754310837365155e-05, "loss": 0.0582, "step": 14520 }, { "grad_norm": 0.45494434237480164, "learning_rate": 9.753798639181856e-05, "loss": 0.0803, "step": 14530 }, { "grad_norm": 0.45723581314086914, "learning_rate": 9.753285921130286e-05, "loss": 0.0561, "step": 14540 }, { "grad_norm": 0.4073367714881897, "learning_rate": 9.752772683266512e-05, "loss": 0.0576, "step": 14550 }, { "grad_norm": 0.5242099165916443, "learning_rate": 9.752258925646665e-05, "loss": 0.0567, "step": 14560 }, { "grad_norm": 0.3352718651294708, "learning_rate": 9.751744648326926e-05, "loss": 0.0458, "step": 14570 }, { "grad_norm": 0.4832824766635895, "learning_rate": 9.751229851363536e-05, "loss": 0.0653, "step": 14580 }, { "grad_norm": 0.3918011784553528, "learning_rate": 9.750714534812793e-05, "loss": 0.0592, "step": 14590 }, { "grad_norm": 0.48692893981933594, "learning_rate": 9.750198698731053e-05, "loss": 0.0538, "step": 14600 }, { "grad_norm": 0.3225610554218292, "learning_rate": 9.749682343174722e-05, "loss": 0.0477, "step": 14610 }, { "grad_norm": 0.4732306897640228, "learning_rate": 9.749165468200272e-05, "loss": 0.0563, "step": 14620 }, { "grad_norm": 0.5833715200424194, "learning_rate": 9.748648073864229e-05, "loss": 0.0515, "step": 14630 }, { "grad_norm": 0.5222834944725037, "learning_rate": 9.748130160223168e-05, "loss": 0.0496, "step": 14640 }, { "grad_norm": 0.5224114656448364, "learning_rate": 9.747611727333734e-05, "loss": 0.0537, "step": 14650 }, { "grad_norm": 0.47206398844718933, "learning_rate": 9.74709277525262e-05, "loss": 0.0523, "step": 14660 }, { "grad_norm": 0.5030981302261353, "learning_rate": 9.746573304036576e-05, "loss": 0.0607, "step": 14670 }, { "grad_norm": 0.5116593241691589, "learning_rate": 9.746053313742412e-05, "loss": 0.0495, "step": 14680 }, { "grad_norm": 0.4191858172416687, "learning_rate": 9.745532804426994e-05, "loss": 0.0562, "step": 14690 }, { "grad_norm": 0.5507869124412537, "learning_rate": 9.745011776147242e-05, "loss": 0.0604, "step": 14700 }, { "grad_norm": 0.5331596732139587, "learning_rate": 9.744490228960138e-05, "loss": 0.0648, "step": 14710 }, { "grad_norm": 0.45436808466911316, "learning_rate": 9.743968162922713e-05, "loss": 0.0598, "step": 14720 }, { "grad_norm": 0.38000285625457764, "learning_rate": 9.743445578092064e-05, "loss": 0.0577, "step": 14730 }, { "grad_norm": 0.29921379685401917, "learning_rate": 9.742922474525338e-05, "loss": 0.0748, "step": 14740 }, { "grad_norm": 0.540199875831604, "learning_rate": 9.742398852279741e-05, "loss": 0.0629, "step": 14750 }, { "grad_norm": 0.475923627614975, "learning_rate": 9.741874711412535e-05, "loss": 0.0564, "step": 14760 }, { "grad_norm": 0.5363209247589111, "learning_rate": 9.741350051981042e-05, "loss": 0.0608, "step": 14770 }, { "grad_norm": 0.45666760206222534, "learning_rate": 9.740824874042633e-05, "loss": 0.0645, "step": 14780 }, { "grad_norm": 0.39536088705062866, "learning_rate": 9.740299177654746e-05, "loss": 0.0436, "step": 14790 }, { "grad_norm": 0.517528235912323, "learning_rate": 9.739772962874867e-05, "loss": 0.0559, "step": 14800 }, { "grad_norm": 0.33347707986831665, "learning_rate": 9.739246229760541e-05, "loss": 0.046, "step": 14810 }, { "grad_norm": 0.40442582964897156, "learning_rate": 9.738718978369376e-05, "loss": 0.0558, "step": 14820 }, { "grad_norm": 0.5896177291870117, "learning_rate": 9.738191208759025e-05, "loss": 0.0664, "step": 14830 }, { "grad_norm": 0.4907852113246918, "learning_rate": 9.73766292098721e-05, "loss": 0.0754, "step": 14840 }, { "grad_norm": 0.41536834836006165, "learning_rate": 9.737134115111699e-05, "loss": 0.0751, "step": 14850 }, { "grad_norm": 0.41709059476852417, "learning_rate": 9.736604791190323e-05, "loss": 0.0538, "step": 14860 }, { "grad_norm": 0.3773876428604126, "learning_rate": 9.73607494928097e-05, "loss": 0.0499, "step": 14870 }, { "grad_norm": 0.4738220274448395, "learning_rate": 9.735544589441581e-05, "loss": 0.0568, "step": 14880 }, { "grad_norm": 0.5070257782936096, "learning_rate": 9.735013711730154e-05, "loss": 0.0558, "step": 14890 }, { "grad_norm": 0.691325843334198, "learning_rate": 9.734482316204747e-05, "loss": 0.0526, "step": 14900 }, { "grad_norm": 0.496967613697052, "learning_rate": 9.733950402923473e-05, "loss": 0.0559, "step": 14910 }, { "grad_norm": 0.40683692693710327, "learning_rate": 9.7334179719445e-05, "loss": 0.0554, "step": 14920 }, { "grad_norm": 0.5489476919174194, "learning_rate": 9.732885023326053e-05, "loss": 0.0499, "step": 14930 }, { "grad_norm": 0.5504977703094482, "learning_rate": 9.732351557126418e-05, "loss": 0.063, "step": 14940 }, { "grad_norm": 0.4344008266925812, "learning_rate": 9.731817573403929e-05, "loss": 0.0643, "step": 14950 }, { "grad_norm": 0.42040786147117615, "learning_rate": 9.731283072216985e-05, "loss": 0.0525, "step": 14960 }, { "grad_norm": 0.42134642601013184, "learning_rate": 9.730748053624039e-05, "loss": 0.054, "step": 14970 }, { "grad_norm": 0.4040990173816681, "learning_rate": 9.730212517683598e-05, "loss": 0.0471, "step": 14980 }, { "grad_norm": 0.3753746449947357, "learning_rate": 9.729676464454228e-05, "loss": 0.0551, "step": 14990 }, { "grad_norm": 0.37731850147247314, "learning_rate": 9.72913989399455e-05, "loss": 0.0561, "step": 15000 }, { "grad_norm": 0.3558387756347656, "learning_rate": 9.728602806363242e-05, "loss": 0.0475, "step": 15010 }, { "grad_norm": 0.5622401833534241, "learning_rate": 9.728065201619043e-05, "loss": 0.0502, "step": 15020 }, { "grad_norm": 0.5580646991729736, "learning_rate": 9.727527079820742e-05, "loss": 0.0549, "step": 15030 }, { "grad_norm": 0.3941076695919037, "learning_rate": 9.726988441027186e-05, "loss": 0.0662, "step": 15040 }, { "grad_norm": 0.3554348349571228, "learning_rate": 9.726449285297281e-05, "loss": 0.0721, "step": 15050 }, { "grad_norm": 0.54873126745224, "learning_rate": 9.72590961268999e-05, "loss": 0.0649, "step": 15060 }, { "grad_norm": 0.5601935386657715, "learning_rate": 9.725369423264328e-05, "loss": 0.0559, "step": 15070 }, { "grad_norm": 0.36713090538978577, "learning_rate": 9.72482871707937e-05, "loss": 0.0643, "step": 15080 }, { "grad_norm": 0.4157056212425232, "learning_rate": 9.724287494194247e-05, "loss": 0.0618, "step": 15090 }, { "grad_norm": 0.3649997413158417, "learning_rate": 9.723745754668147e-05, "loss": 0.0576, "step": 15100 }, { "grad_norm": 0.4560370445251465, "learning_rate": 9.723203498560313e-05, "loss": 0.0534, "step": 15110 }, { "grad_norm": 0.5241330862045288, "learning_rate": 9.722660725930046e-05, "loss": 0.0564, "step": 15120 }, { "grad_norm": 0.48011064529418945, "learning_rate": 9.722117436836702e-05, "loss": 0.0583, "step": 15130 }, { "grad_norm": 0.7919439673423767, "learning_rate": 9.721573631339696e-05, "loss": 0.058, "step": 15140 }, { "grad_norm": 0.528914749622345, "learning_rate": 9.721029309498494e-05, "loss": 0.061, "step": 15150 }, { "grad_norm": 0.48894786834716797, "learning_rate": 9.720484471372627e-05, "loss": 0.061, "step": 15160 }, { "grad_norm": 0.45950567722320557, "learning_rate": 9.719939117021673e-05, "loss": 0.0561, "step": 15170 }, { "grad_norm": 0.5070090293884277, "learning_rate": 9.719393246505275e-05, "loss": 0.055, "step": 15180 }, { "grad_norm": 0.5787073373794556, "learning_rate": 9.718846859883128e-05, "loss": 0.0535, "step": 15190 }, { "grad_norm": 0.4548698961734772, "learning_rate": 9.718299957214982e-05, "loss": 0.0583, "step": 15200 }, { "grad_norm": 0.5159264802932739, "learning_rate": 9.717752538560646e-05, "loss": 0.056, "step": 15210 }, { "grad_norm": 0.49721184372901917, "learning_rate": 9.717204603979986e-05, "loss": 0.0555, "step": 15220 }, { "grad_norm": 0.525342583656311, "learning_rate": 9.716656153532922e-05, "loss": 0.043, "step": 15230 }, { "grad_norm": 0.41119131445884705, "learning_rate": 9.716107187279434e-05, "loss": 0.0492, "step": 15240 }, { "grad_norm": 0.48846372961997986, "learning_rate": 9.715557705279555e-05, "loss": 0.0646, "step": 15250 }, { "grad_norm": 0.4647846519947052, "learning_rate": 9.715007707593372e-05, "loss": 0.0504, "step": 15260 }, { "grad_norm": 0.6427425146102905, "learning_rate": 9.714457194281036e-05, "loss": 0.0576, "step": 15270 }, { "grad_norm": 0.554074227809906, "learning_rate": 9.713906165402751e-05, "loss": 0.0501, "step": 15280 }, { "grad_norm": 0.4076119065284729, "learning_rate": 9.713354621018774e-05, "loss": 0.0709, "step": 15290 }, { "grad_norm": 0.741132378578186, "learning_rate": 9.712802561189422e-05, "loss": 0.0642, "step": 15300 }, { "grad_norm": 0.445149689912796, "learning_rate": 9.712249985975069e-05, "loss": 0.0691, "step": 15310 }, { "grad_norm": 0.5193228721618652, "learning_rate": 9.71169689543614e-05, "loss": 0.0467, "step": 15320 }, { "grad_norm": 0.5506672263145447, "learning_rate": 9.711143289633123e-05, "loss": 0.0456, "step": 15330 }, { "grad_norm": 0.3320715129375458, "learning_rate": 9.710589168626561e-05, "loss": 0.0599, "step": 15340 }, { "grad_norm": 0.5998912453651428, "learning_rate": 9.710034532477048e-05, "loss": 0.065, "step": 15350 }, { "grad_norm": 0.55107581615448, "learning_rate": 9.709479381245239e-05, "loss": 0.047, "step": 15360 }, { "grad_norm": 0.5396355986595154, "learning_rate": 9.708923714991847e-05, "loss": 0.0557, "step": 15370 }, { "grad_norm": 0.38576069474220276, "learning_rate": 9.708367533777638e-05, "loss": 0.0525, "step": 15380 }, { "grad_norm": 0.5609723925590515, "learning_rate": 9.707810837663431e-05, "loss": 0.0521, "step": 15390 }, { "grad_norm": 0.4783788323402405, "learning_rate": 9.707253626710113e-05, "loss": 0.0605, "step": 15400 }, { "grad_norm": 0.35754093527793884, "learning_rate": 9.706695900978613e-05, "loss": 0.0448, "step": 15410 }, { "grad_norm": 0.31287074089050293, "learning_rate": 9.706137660529926e-05, "loss": 0.0578, "step": 15420 }, { "grad_norm": 0.5883102416992188, "learning_rate": 9.705578905425101e-05, "loss": 0.0768, "step": 15430 }, { "grad_norm": 0.4590515196323395, "learning_rate": 9.705019635725241e-05, "loss": 0.0477, "step": 15440 }, { "grad_norm": 0.37248438596725464, "learning_rate": 9.704459851491508e-05, "loss": 0.0585, "step": 15450 }, { "grad_norm": 0.4880426526069641, "learning_rate": 9.703899552785118e-05, "loss": 0.0504, "step": 15460 }, { "grad_norm": 0.5556508898735046, "learning_rate": 9.703338739667346e-05, "loss": 0.0531, "step": 15470 }, { "grad_norm": 0.5828635096549988, "learning_rate": 9.70277741219952e-05, "loss": 0.0573, "step": 15480 }, { "grad_norm": 0.5271575450897217, "learning_rate": 9.702215570443027e-05, "loss": 0.057, "step": 15490 }, { "grad_norm": 0.35056066513061523, "learning_rate": 9.701653214459309e-05, "loss": 0.0506, "step": 15500 }, { "grad_norm": 0.39076927304267883, "learning_rate": 9.701090344309865e-05, "loss": 0.0588, "step": 15510 }, { "grad_norm": 0.7644540071487427, "learning_rate": 9.700526960056247e-05, "loss": 0.044, "step": 15520 }, { "grad_norm": 0.39921364188194275, "learning_rate": 9.699963061760068e-05, "loss": 0.0527, "step": 15530 }, { "grad_norm": 0.35688889026641846, "learning_rate": 9.699398649482997e-05, "loss": 0.0582, "step": 15540 }, { "grad_norm": 0.5273937582969666, "learning_rate": 9.698833723286753e-05, "loss": 0.0504, "step": 15550 }, { "grad_norm": 0.7533779144287109, "learning_rate": 9.698268283233118e-05, "loss": 0.0662, "step": 15560 }, { "grad_norm": 0.6841228604316711, "learning_rate": 9.697702329383929e-05, "loss": 0.0719, "step": 15570 }, { "grad_norm": 0.45422598719596863, "learning_rate": 9.697135861801074e-05, "loss": 0.0663, "step": 15580 }, { "grad_norm": 0.42446404695510864, "learning_rate": 9.696568880546505e-05, "loss": 0.0509, "step": 15590 }, { "grad_norm": 0.4651605486869812, "learning_rate": 9.696001385682223e-05, "loss": 0.0555, "step": 15600 }, { "grad_norm": 0.4166405498981476, "learning_rate": 9.695433377270291e-05, "loss": 0.0531, "step": 15610 }, { "grad_norm": 0.35095977783203125, "learning_rate": 9.694864855372824e-05, "loss": 0.0454, "step": 15620 }, { "grad_norm": 0.44331836700439453, "learning_rate": 9.694295820051995e-05, "loss": 0.0563, "step": 15630 }, { "grad_norm": 0.4896969497203827, "learning_rate": 9.693726271370032e-05, "loss": 0.051, "step": 15640 }, { "grad_norm": 0.4205925464630127, "learning_rate": 9.693156209389221e-05, "loss": 0.0538, "step": 15650 }, { "grad_norm": 0.5439686179161072, "learning_rate": 9.692585634171905e-05, "loss": 0.0584, "step": 15660 }, { "grad_norm": 0.5524373054504395, "learning_rate": 9.692014545780476e-05, "loss": 0.0524, "step": 15670 }, { "grad_norm": 0.3288308382034302, "learning_rate": 9.691442944277393e-05, "loss": 0.0381, "step": 15680 }, { "grad_norm": 0.38153594732284546, "learning_rate": 9.690870829725162e-05, "loss": 0.0495, "step": 15690 }, { "grad_norm": 0.4624873697757721, "learning_rate": 9.69029820218635e-05, "loss": 0.0537, "step": 15700 }, { "grad_norm": 0.363847941160202, "learning_rate": 9.689725061723579e-05, "loss": 0.0521, "step": 15710 }, { "grad_norm": 0.3813846707344055, "learning_rate": 9.689151408399527e-05, "loss": 0.0435, "step": 15720 }, { "grad_norm": 0.5071368217468262, "learning_rate": 9.688577242276924e-05, "loss": 0.0661, "step": 15730 }, { "grad_norm": 0.5097759962081909, "learning_rate": 9.688002563418566e-05, "loss": 0.0583, "step": 15740 }, { "grad_norm": 0.4743652641773224, "learning_rate": 9.687427371887293e-05, "loss": 0.061, "step": 15750 }, { "grad_norm": 0.6469278931617737, "learning_rate": 9.686851667746012e-05, "loss": 0.0565, "step": 15760 }, { "grad_norm": 0.5215329527854919, "learning_rate": 9.686275451057677e-05, "loss": 0.052, "step": 15770 }, { "grad_norm": 0.3916742205619812, "learning_rate": 9.685698721885308e-05, "loss": 0.045, "step": 15780 }, { "grad_norm": 0.5762841105461121, "learning_rate": 9.68512148029197e-05, "loss": 0.0628, "step": 15790 }, { "grad_norm": 0.47821229696273804, "learning_rate": 9.684543726340791e-05, "loss": 0.0592, "step": 15800 }, { "grad_norm": 0.500417172908783, "learning_rate": 9.683965460094952e-05, "loss": 0.0553, "step": 15810 }, { "grad_norm": 0.3473297655582428, "learning_rate": 9.683386681617694e-05, "loss": 0.0492, "step": 15820 }, { "grad_norm": 0.4670548439025879, "learning_rate": 9.68280739097231e-05, "loss": 0.0504, "step": 15830 }, { "grad_norm": 0.3981858789920807, "learning_rate": 9.682227588222148e-05, "loss": 0.0504, "step": 15840 }, { "grad_norm": 0.6072050333023071, "learning_rate": 9.681647273430618e-05, "loss": 0.0548, "step": 15850 }, { "grad_norm": 0.4850952923297882, "learning_rate": 9.681066446661182e-05, "loss": 0.051, "step": 15860 }, { "grad_norm": 0.342460960149765, "learning_rate": 9.680485107977357e-05, "loss": 0.0548, "step": 15870 }, { "grad_norm": 0.4541725218296051, "learning_rate": 9.679903257442716e-05, "loss": 0.0604, "step": 15880 }, { "grad_norm": 0.6991913914680481, "learning_rate": 9.679320895120891e-05, "loss": 0.0602, "step": 15890 }, { "grad_norm": 0.4346345067024231, "learning_rate": 9.67873802107557e-05, "loss": 0.0534, "step": 15900 }, { "grad_norm": 0.43627527356147766, "learning_rate": 9.67815463537049e-05, "loss": 0.0539, "step": 15910 }, { "grad_norm": 0.5177244544029236, "learning_rate": 9.677570738069457e-05, "loss": 0.0544, "step": 15920 }, { "grad_norm": 0.5829718112945557, "learning_rate": 9.676986329236318e-05, "loss": 0.0559, "step": 15930 }, { "grad_norm": 0.5397406220436096, "learning_rate": 9.676401408934987e-05, "loss": 0.0522, "step": 15940 }, { "grad_norm": 0.4576284885406494, "learning_rate": 9.675815977229428e-05, "loss": 0.0539, "step": 15950 }, { "grad_norm": 0.3464004695415497, "learning_rate": 9.675230034183664e-05, "loss": 0.0461, "step": 15960 }, { "grad_norm": 0.48302385210990906, "learning_rate": 9.674643579861773e-05, "loss": 0.0562, "step": 15970 }, { "grad_norm": 0.5508238673210144, "learning_rate": 9.674056614327886e-05, "loss": 0.0502, "step": 15980 }, { "grad_norm": 0.572272002696991, "learning_rate": 9.673469137646198e-05, "loss": 0.0538, "step": 15990 }, { "grad_norm": 0.5256220698356628, "learning_rate": 9.67288114988095e-05, "loss": 0.0537, "step": 16000 }, { "grad_norm": 0.5666032433509827, "learning_rate": 9.672292651096447e-05, "loss": 0.0593, "step": 16010 }, { "grad_norm": 0.4486767649650574, "learning_rate": 9.671703641357042e-05, "loss": 0.0588, "step": 16020 }, { "grad_norm": 0.3678446412086487, "learning_rate": 9.67111412072715e-05, "loss": 0.0534, "step": 16030 }, { "grad_norm": 0.3955906927585602, "learning_rate": 9.670524089271242e-05, "loss": 0.054, "step": 16040 }, { "grad_norm": 0.5948344469070435, "learning_rate": 9.669933547053842e-05, "loss": 0.0466, "step": 16050 }, { "grad_norm": 0.40659087896347046, "learning_rate": 9.669342494139531e-05, "loss": 0.0524, "step": 16060 }, { "grad_norm": 0.35977423191070557, "learning_rate": 9.668750930592943e-05, "loss": 0.0406, "step": 16070 }, { "grad_norm": 0.4676137864589691, "learning_rate": 9.668158856478775e-05, "loss": 0.0482, "step": 16080 }, { "grad_norm": 0.40026888251304626, "learning_rate": 9.66756627186177e-05, "loss": 0.0599, "step": 16090 }, { "grad_norm": 0.45040133595466614, "learning_rate": 9.666973176806737e-05, "loss": 0.0588, "step": 16100 }, { "grad_norm": 0.45479172468185425, "learning_rate": 9.666379571378534e-05, "loss": 0.059, "step": 16110 }, { "grad_norm": 0.3974035680294037, "learning_rate": 9.665785455642076e-05, "loss": 0.0662, "step": 16120 }, { "grad_norm": 0.6069784760475159, "learning_rate": 9.665190829662337e-05, "loss": 0.0613, "step": 16130 }, { "grad_norm": 0.5680529475212097, "learning_rate": 9.664595693504342e-05, "loss": 0.0591, "step": 16140 }, { "grad_norm": 0.391623854637146, "learning_rate": 9.664000047233175e-05, "loss": 0.056, "step": 16150 }, { "grad_norm": 0.3089551627635956, "learning_rate": 9.663403890913976e-05, "loss": 0.0485, "step": 16160 }, { "grad_norm": 0.4182678163051605, "learning_rate": 9.662807224611938e-05, "loss": 0.0425, "step": 16170 }, { "grad_norm": 0.36633822321891785, "learning_rate": 9.662210048392311e-05, "loss": 0.0411, "step": 16180 }, { "grad_norm": 0.4391915500164032, "learning_rate": 9.661612362320405e-05, "loss": 0.0459, "step": 16190 }, { "grad_norm": 0.3215637505054474, "learning_rate": 9.661014166461579e-05, "loss": 0.0472, "step": 16200 }, { "grad_norm": 0.3791619539260864, "learning_rate": 9.66041546088125e-05, "loss": 0.0563, "step": 16210 }, { "grad_norm": 0.4629876911640167, "learning_rate": 9.659816245644895e-05, "loss": 0.0445, "step": 16220 }, { "grad_norm": 0.38505885004997253, "learning_rate": 9.65921652081804e-05, "loss": 0.0584, "step": 16230 }, { "grad_norm": 0.5338959693908691, "learning_rate": 9.658616286466271e-05, "loss": 0.0514, "step": 16240 }, { "grad_norm": 0.46128448843955994, "learning_rate": 9.65801554265523e-05, "loss": 0.049, "step": 16250 }, { "grad_norm": 0.44327834248542786, "learning_rate": 9.657414289450612e-05, "loss": 0.0596, "step": 16260 }, { "grad_norm": 0.41430309414863586, "learning_rate": 9.656812526918171e-05, "loss": 0.0523, "step": 16270 }, { "grad_norm": 0.36549443006515503, "learning_rate": 9.656210255123712e-05, "loss": 0.0505, "step": 16280 }, { "grad_norm": 0.3432493209838867, "learning_rate": 9.6556074741331e-05, "loss": 0.0441, "step": 16290 }, { "grad_norm": 0.5726752877235413, "learning_rate": 9.655004184012256e-05, "loss": 0.04, "step": 16300 }, { "grad_norm": 0.6561883687973022, "learning_rate": 9.654400384827152e-05, "loss": 0.0581, "step": 16310 }, { "grad_norm": 0.4815542995929718, "learning_rate": 9.653796076643818e-05, "loss": 0.0498, "step": 16320 }, { "grad_norm": 0.42216938734054565, "learning_rate": 9.653191259528344e-05, "loss": 0.0578, "step": 16330 }, { "grad_norm": 0.46792536973953247, "learning_rate": 9.65258593354687e-05, "loss": 0.0449, "step": 16340 }, { "grad_norm": 0.33324822783470154, "learning_rate": 9.651980098765591e-05, "loss": 0.0497, "step": 16350 }, { "grad_norm": 0.406281441450119, "learning_rate": 9.651373755250765e-05, "loss": 0.0495, "step": 16360 }, { "grad_norm": 0.40836504101753235, "learning_rate": 9.650766903068697e-05, "loss": 0.055, "step": 16370 }, { "grad_norm": 0.515482485294342, "learning_rate": 9.650159542285753e-05, "loss": 0.0527, "step": 16380 }, { "grad_norm": 0.4627634584903717, "learning_rate": 9.649551672968353e-05, "loss": 0.0591, "step": 16390 }, { "grad_norm": 0.7404972910881042, "learning_rate": 9.648943295182973e-05, "loss": 0.0654, "step": 16400 }, { "grad_norm": 0.3879656195640564, "learning_rate": 9.648334408996144e-05, "loss": 0.0608, "step": 16410 }, { "grad_norm": 0.41433534026145935, "learning_rate": 9.647725014474452e-05, "loss": 0.0542, "step": 16420 }, { "grad_norm": 0.4816299080848694, "learning_rate": 9.64711511168454e-05, "loss": 0.0538, "step": 16430 }, { "grad_norm": 0.4443035423755646, "learning_rate": 9.646504700693108e-05, "loss": 0.0527, "step": 16440 }, { "grad_norm": 0.5642729997634888, "learning_rate": 9.645893781566907e-05, "loss": 0.0516, "step": 16450 }, { "grad_norm": 0.5134512782096863, "learning_rate": 9.645282354372744e-05, "loss": 0.0526, "step": 16460 }, { "grad_norm": 0.46109530329704285, "learning_rate": 9.644670419177491e-05, "loss": 0.0492, "step": 16470 }, { "grad_norm": 0.4495348632335663, "learning_rate": 9.644057976048062e-05, "loss": 0.0525, "step": 16480 }, { "grad_norm": 0.36197295784950256, "learning_rate": 9.643445025051435e-05, "loss": 0.0529, "step": 16490 }, { "grad_norm": 0.5240782499313354, "learning_rate": 9.642831566254641e-05, "loss": 0.0493, "step": 16500 }, { "grad_norm": 0.41829559206962585, "learning_rate": 9.642217599724769e-05, "loss": 0.052, "step": 16510 }, { "grad_norm": 0.4299930930137634, "learning_rate": 9.64160312552896e-05, "loss": 0.0541, "step": 16520 }, { "grad_norm": 0.5145165920257568, "learning_rate": 9.64098814373441e-05, "loss": 0.053, "step": 16530 }, { "grad_norm": 0.6633355617523193, "learning_rate": 9.640372654408374e-05, "loss": 0.0506, "step": 16540 }, { "grad_norm": 0.5536254644393921, "learning_rate": 9.639756657618162e-05, "loss": 0.0627, "step": 16550 }, { "grad_norm": 0.6220366358757019, "learning_rate": 9.639140153431138e-05, "loss": 0.0566, "step": 16560 }, { "grad_norm": 0.500329852104187, "learning_rate": 9.638523141914721e-05, "loss": 0.0532, "step": 16570 }, { "grad_norm": 0.4339722990989685, "learning_rate": 9.637905623136388e-05, "loss": 0.0626, "step": 16580 }, { "grad_norm": 0.7126352190971375, "learning_rate": 9.637287597163669e-05, "loss": 0.0434, "step": 16590 }, { "grad_norm": 0.5319318771362305, "learning_rate": 9.63666906406415e-05, "loss": 0.0391, "step": 16600 }, { "grad_norm": 0.4176725447177887, "learning_rate": 9.636050023905473e-05, "loss": 0.0529, "step": 16610 }, { "grad_norm": 0.48359251022338867, "learning_rate": 9.635430476755336e-05, "loss": 0.0478, "step": 16620 }, { "grad_norm": 0.3508477807044983, "learning_rate": 9.63481042268149e-05, "loss": 0.0416, "step": 16630 }, { "grad_norm": 0.5277535319328308, "learning_rate": 9.634189861751745e-05, "loss": 0.0645, "step": 16640 }, { "grad_norm": 0.4177386164665222, "learning_rate": 9.633568794033967e-05, "loss": 0.0601, "step": 16650 }, { "grad_norm": 0.5170790553092957, "learning_rate": 9.63294721959607e-05, "loss": 0.0518, "step": 16660 }, { "grad_norm": 0.43517595529556274, "learning_rate": 9.63232513850603e-05, "loss": 0.0691, "step": 16670 }, { "grad_norm": 0.4745773673057556, "learning_rate": 9.631702550831878e-05, "loss": 0.0557, "step": 16680 }, { "grad_norm": 0.5847225785255432, "learning_rate": 9.631079456641698e-05, "loss": 0.0573, "step": 16690 }, { "grad_norm": 0.44064074754714966, "learning_rate": 9.630455856003632e-05, "loss": 0.0503, "step": 16700 }, { "grad_norm": 0.6002745032310486, "learning_rate": 9.629831748985876e-05, "loss": 0.0546, "step": 16710 }, { "grad_norm": 0.47707757353782654, "learning_rate": 9.629207135656679e-05, "loss": 0.0655, "step": 16720 }, { "grad_norm": 0.46948665380477905, "learning_rate": 9.628582016084353e-05, "loss": 0.0546, "step": 16730 }, { "grad_norm": 0.6020501255989075, "learning_rate": 9.627956390337254e-05, "loss": 0.051, "step": 16740 }, { "grad_norm": 0.44284504652023315, "learning_rate": 9.627330258483802e-05, "loss": 0.0595, "step": 16750 }, { "grad_norm": 0.4370899498462677, "learning_rate": 9.62670362059247e-05, "loss": 0.0482, "step": 16760 }, { "grad_norm": 0.4550256133079529, "learning_rate": 9.626076476731786e-05, "loss": 0.0441, "step": 16770 }, { "grad_norm": 0.6075091361999512, "learning_rate": 9.625448826970336e-05, "loss": 0.0564, "step": 16780 }, { "grad_norm": 0.48089903593063354, "learning_rate": 9.624820671376755e-05, "loss": 0.0532, "step": 16790 }, { "grad_norm": 0.47023653984069824, "learning_rate": 9.62419201001974e-05, "loss": 0.0572, "step": 16800 }, { "grad_norm": 0.463351309299469, "learning_rate": 9.623562842968037e-05, "loss": 0.0548, "step": 16810 }, { "grad_norm": 0.38132333755493164, "learning_rate": 9.622933170290454e-05, "loss": 0.0444, "step": 16820 }, { "grad_norm": 0.6141583919525146, "learning_rate": 9.622302992055849e-05, "loss": 0.045, "step": 16830 }, { "grad_norm": 0.34501975774765015, "learning_rate": 9.62167230833314e-05, "loss": 0.0509, "step": 16840 }, { "grad_norm": 0.3450683057308197, "learning_rate": 9.621041119191295e-05, "loss": 0.051, "step": 16850 }, { "grad_norm": 0.4218332767486572, "learning_rate": 9.620409424699342e-05, "loss": 0.0564, "step": 16860 }, { "grad_norm": 0.609264612197876, "learning_rate": 9.619777224926359e-05, "loss": 0.0474, "step": 16870 }, { "grad_norm": 0.5513295531272888, "learning_rate": 9.619144519941485e-05, "loss": 0.0487, "step": 16880 }, { "grad_norm": 0.5222836136817932, "learning_rate": 9.618511309813912e-05, "loss": 0.0561, "step": 16890 }, { "grad_norm": 0.4066290259361267, "learning_rate": 9.617877594612886e-05, "loss": 0.0433, "step": 16900 }, { "grad_norm": 0.40690848231315613, "learning_rate": 9.617243374407707e-05, "loss": 0.0489, "step": 16910 }, { "grad_norm": 0.5269031524658203, "learning_rate": 9.616608649267736e-05, "loss": 0.0498, "step": 16920 }, { "grad_norm": 0.5372028350830078, "learning_rate": 9.615973419262385e-05, "loss": 0.0564, "step": 16930 }, { "grad_norm": 0.4899827837944031, "learning_rate": 9.615337684461119e-05, "loss": 0.0526, "step": 16940 }, { "grad_norm": 0.5256544947624207, "learning_rate": 9.614701444933465e-05, "loss": 0.0749, "step": 16950 }, { "grad_norm": 0.4690827429294586, "learning_rate": 9.614064700748997e-05, "loss": 0.063, "step": 16960 }, { "grad_norm": 0.43515297770500183, "learning_rate": 9.613427451977352e-05, "loss": 0.0604, "step": 16970 }, { "grad_norm": 0.45617860555648804, "learning_rate": 9.612789698688216e-05, "loss": 0.0533, "step": 16980 }, { "grad_norm": 0.5879470109939575, "learning_rate": 9.612151440951334e-05, "loss": 0.0588, "step": 16990 }, { "grad_norm": 0.6421018838882446, "learning_rate": 9.611512678836506e-05, "loss": 0.0521, "step": 17000 }, { "grad_norm": 0.5046072602272034, "learning_rate": 9.610873412413584e-05, "loss": 0.0603, "step": 17010 }, { "grad_norm": 0.46251124143600464, "learning_rate": 9.610233641752476e-05, "loss": 0.0608, "step": 17020 }, { "grad_norm": 0.40687525272369385, "learning_rate": 9.609593366923151e-05, "loss": 0.0488, "step": 17030 }, { "grad_norm": 0.47804686427116394, "learning_rate": 9.608952587995625e-05, "loss": 0.057, "step": 17040 }, { "grad_norm": 0.4894300103187561, "learning_rate": 9.608311305039972e-05, "loss": 0.0521, "step": 17050 }, { "grad_norm": 0.6901737451553345, "learning_rate": 9.607669518126326e-05, "loss": 0.0696, "step": 17060 }, { "grad_norm": 0.397847443819046, "learning_rate": 9.607027227324866e-05, "loss": 0.0451, "step": 17070 }, { "grad_norm": 0.8092025518417358, "learning_rate": 9.606384432705837e-05, "loss": 0.0492, "step": 17080 }, { "grad_norm": 0.43900829553604126, "learning_rate": 9.60574113433953e-05, "loss": 0.0494, "step": 17090 }, { "grad_norm": 0.469544917345047, "learning_rate": 9.6050973322963e-05, "loss": 0.0499, "step": 17100 }, { "grad_norm": 0.6043316125869751, "learning_rate": 9.604453026646547e-05, "loss": 0.0579, "step": 17110 }, { "grad_norm": 0.45314687490463257, "learning_rate": 9.603808217460735e-05, "loss": 0.0477, "step": 17120 }, { "grad_norm": 0.4597518742084503, "learning_rate": 9.603162904809377e-05, "loss": 0.048, "step": 17130 }, { "grad_norm": 0.5596561431884766, "learning_rate": 9.602517088763045e-05, "loss": 0.0484, "step": 17140 }, { "grad_norm": 0.4522288143634796, "learning_rate": 9.601870769392365e-05, "loss": 0.0518, "step": 17150 }, { "grad_norm": 0.5506749749183655, "learning_rate": 9.601223946768017e-05, "loss": 0.0536, "step": 17160 }, { "grad_norm": 0.4135389029979706, "learning_rate": 9.600576620960734e-05, "loss": 0.0548, "step": 17170 }, { "grad_norm": 0.5174310207366943, "learning_rate": 9.599928792041308e-05, "loss": 0.0525, "step": 17180 }, { "grad_norm": 0.5552428364753723, "learning_rate": 9.599280460080587e-05, "loss": 0.0505, "step": 17190 }, { "grad_norm": 0.4119274318218231, "learning_rate": 9.59863162514947e-05, "loss": 0.0493, "step": 17200 }, { "grad_norm": 0.47643208503723145, "learning_rate": 9.597982287318911e-05, "loss": 0.0511, "step": 17210 }, { "grad_norm": 0.4324658215045929, "learning_rate": 9.597332446659923e-05, "loss": 0.051, "step": 17220 }, { "grad_norm": 0.5703195929527283, "learning_rate": 9.59668210324357e-05, "loss": 0.0636, "step": 17230 }, { "grad_norm": 0.4747471511363983, "learning_rate": 9.596031257140974e-05, "loss": 0.0469, "step": 17240 }, { "grad_norm": 0.430350124835968, "learning_rate": 9.59537990842331e-05, "loss": 0.0635, "step": 17250 }, { "grad_norm": 0.48070263862609863, "learning_rate": 9.594728057161806e-05, "loss": 0.0549, "step": 17260 }, { "grad_norm": 0.3308330178260803, "learning_rate": 9.594075703427752e-05, "loss": 0.0488, "step": 17270 }, { "grad_norm": 0.4033336937427521, "learning_rate": 9.593422847292486e-05, "loss": 0.0484, "step": 17280 }, { "grad_norm": 0.3273792862892151, "learning_rate": 9.592769488827402e-05, "loss": 0.0551, "step": 17290 }, { "grad_norm": 0.4479403495788574, "learning_rate": 9.592115628103952e-05, "loss": 0.0549, "step": 17300 }, { "grad_norm": 0.5627068877220154, "learning_rate": 9.591461265193643e-05, "loss": 0.0568, "step": 17310 }, { "grad_norm": 0.534248411655426, "learning_rate": 9.590806400168032e-05, "loss": 0.0566, "step": 17320 }, { "grad_norm": 0.5240140557289124, "learning_rate": 9.590151033098735e-05, "loss": 0.0524, "step": 17330 }, { "grad_norm": 0.45024988055229187, "learning_rate": 9.589495164057423e-05, "loss": 0.0492, "step": 17340 }, { "grad_norm": 0.3444935083389282, "learning_rate": 9.58883879311582e-05, "loss": 0.0447, "step": 17350 }, { "grad_norm": 0.5383596420288086, "learning_rate": 9.588181920345705e-05, "loss": 0.0591, "step": 17360 }, { "grad_norm": 0.4003090262413025, "learning_rate": 9.587524545818913e-05, "loss": 0.056, "step": 17370 }, { "grad_norm": 0.39118504524230957, "learning_rate": 9.586866669607335e-05, "loss": 0.0446, "step": 17380 }, { "grad_norm": 0.3524579107761383, "learning_rate": 9.586208291782915e-05, "loss": 0.0451, "step": 17390 }, { "grad_norm": 1.0311899185180664, "learning_rate": 9.58554941241765e-05, "loss": 0.0537, "step": 17400 }, { "grad_norm": 0.4324147701263428, "learning_rate": 9.584890031583596e-05, "loss": 0.0461, "step": 17410 }, { "grad_norm": 0.5123440623283386, "learning_rate": 9.584230149352861e-05, "loss": 0.0505, "step": 17420 }, { "grad_norm": 0.7093375325202942, "learning_rate": 9.58356976579761e-05, "loss": 0.0527, "step": 17430 }, { "grad_norm": 0.5183936357498169, "learning_rate": 9.58290888099006e-05, "loss": 0.0575, "step": 17440 }, { "grad_norm": 0.3485107421875, "learning_rate": 9.582247495002486e-05, "loss": 0.049, "step": 17450 }, { "grad_norm": 0.4149213433265686, "learning_rate": 9.581585607907214e-05, "loss": 0.0548, "step": 17460 }, { "grad_norm": 0.47820040583610535, "learning_rate": 9.580923219776628e-05, "loss": 0.0467, "step": 17470 }, { "grad_norm": 0.4673596918582916, "learning_rate": 9.580260330683167e-05, "loss": 0.0593, "step": 17480 }, { "grad_norm": 0.3584863245487213, "learning_rate": 9.579596940699322e-05, "loss": 0.0459, "step": 17490 }, { "grad_norm": 0.5139623284339905, "learning_rate": 9.578933049897643e-05, "loss": 0.0595, "step": 17500 }, { "grad_norm": 0.6635710000991821, "learning_rate": 9.578268658350728e-05, "loss": 0.0647, "step": 17510 }, { "grad_norm": 0.5050578713417053, "learning_rate": 9.577603766131235e-05, "loss": 0.0667, "step": 17520 }, { "grad_norm": 0.5085577368736267, "learning_rate": 9.576938373311878e-05, "loss": 0.0589, "step": 17530 }, { "grad_norm": 0.46770820021629333, "learning_rate": 9.576272479965421e-05, "loss": 0.0592, "step": 17540 }, { "grad_norm": 0.4838998317718506, "learning_rate": 9.575606086164687e-05, "loss": 0.0519, "step": 17550 }, { "grad_norm": 0.5042723417282104, "learning_rate": 9.57493919198255e-05, "loss": 0.0691, "step": 17560 }, { "grad_norm": 0.48531782627105713, "learning_rate": 9.57427179749194e-05, "loss": 0.0561, "step": 17570 }, { "grad_norm": 0.43023934960365295, "learning_rate": 9.573603902765846e-05, "loss": 0.0619, "step": 17580 }, { "grad_norm": 0.4952029287815094, "learning_rate": 9.572935507877304e-05, "loss": 0.0567, "step": 17590 }, { "grad_norm": 0.41131773591041565, "learning_rate": 9.57226661289941e-05, "loss": 0.0663, "step": 17600 }, { "grad_norm": 0.5074775815010071, "learning_rate": 9.571597217905315e-05, "loss": 0.0627, "step": 17610 }, { "grad_norm": 0.4827404022216797, "learning_rate": 9.57092732296822e-05, "loss": 0.0506, "step": 17620 }, { "grad_norm": 0.4554319381713867, "learning_rate": 9.570256928161385e-05, "loss": 0.0473, "step": 17630 }, { "grad_norm": 0.5273383259773254, "learning_rate": 9.569586033558126e-05, "loss": 0.0526, "step": 17640 }, { "grad_norm": 0.41918236017227173, "learning_rate": 9.568914639231807e-05, "loss": 0.0533, "step": 17650 }, { "grad_norm": 0.7694292068481445, "learning_rate": 9.568242745255852e-05, "loss": 0.0653, "step": 17660 }, { "grad_norm": 0.5297472476959229, "learning_rate": 9.567570351703739e-05, "loss": 0.0603, "step": 17670 }, { "grad_norm": 0.5645979642868042, "learning_rate": 9.566897458649001e-05, "loss": 0.0509, "step": 17680 }, { "grad_norm": 0.30634239315986633, "learning_rate": 9.566224066165221e-05, "loss": 0.0558, "step": 17690 }, { "grad_norm": 0.4502127170562744, "learning_rate": 9.565550174326043e-05, "loss": 0.0703, "step": 17700 }, { "grad_norm": 0.5167814493179321, "learning_rate": 9.564875783205162e-05, "loss": 0.0643, "step": 17710 }, { "grad_norm": 0.4485641121864319, "learning_rate": 9.564200892876328e-05, "loss": 0.0679, "step": 17720 }, { "grad_norm": 0.6106651425361633, "learning_rate": 9.563525503413348e-05, "loss": 0.0534, "step": 17730 }, { "grad_norm": 0.415194034576416, "learning_rate": 9.562849614890079e-05, "loss": 0.0607, "step": 17740 }, { "grad_norm": 0.5326377153396606, "learning_rate": 9.562173227380436e-05, "loss": 0.0731, "step": 17750 }, { "grad_norm": 0.4479272663593292, "learning_rate": 9.561496340958389e-05, "loss": 0.0605, "step": 17760 }, { "grad_norm": 0.4446325898170471, "learning_rate": 9.560818955697959e-05, "loss": 0.0489, "step": 17770 }, { "grad_norm": 0.4151866137981415, "learning_rate": 9.560141071673228e-05, "loss": 0.0686, "step": 17780 }, { "grad_norm": 0.5125613212585449, "learning_rate": 9.559462688958323e-05, "loss": 0.0421, "step": 17790 }, { "grad_norm": 0.3670457899570465, "learning_rate": 9.558783807627434e-05, "loss": 0.0555, "step": 17800 }, { "grad_norm": 0.48645561933517456, "learning_rate": 9.558104427754801e-05, "loss": 0.0594, "step": 17810 }, { "grad_norm": 1.854928731918335, "learning_rate": 9.557424549414722e-05, "loss": 0.0596, "step": 17820 }, { "grad_norm": 0.40978607535362244, "learning_rate": 9.556744172681546e-05, "loss": 0.0511, "step": 17830 }, { "grad_norm": 0.5078470706939697, "learning_rate": 9.556063297629677e-05, "loss": 0.0529, "step": 17840 }, { "grad_norm": 0.5777053833007812, "learning_rate": 9.555381924333578e-05, "loss": 0.0562, "step": 17850 }, { "grad_norm": 0.6030309796333313, "learning_rate": 9.554700052867758e-05, "loss": 0.0477, "step": 17860 }, { "grad_norm": 0.5004217624664307, "learning_rate": 9.554017683306789e-05, "loss": 0.0537, "step": 17870 }, { "grad_norm": 0.3897063136100769, "learning_rate": 9.553334815725294e-05, "loss": 0.048, "step": 17880 }, { "grad_norm": 1.0040825605392456, "learning_rate": 9.552651450197949e-05, "loss": 0.0552, "step": 17890 }, { "grad_norm": 0.774681031703949, "learning_rate": 9.551967586799486e-05, "loss": 0.0504, "step": 17900 }, { "grad_norm": 0.5120039582252502, "learning_rate": 9.551283225604692e-05, "loss": 0.0593, "step": 17910 }, { "grad_norm": 0.4374562203884125, "learning_rate": 9.550598366688406e-05, "loss": 0.0507, "step": 17920 }, { "grad_norm": 0.46810421347618103, "learning_rate": 9.549913010125526e-05, "loss": 0.0513, "step": 17930 }, { "grad_norm": 0.4673312306404114, "learning_rate": 9.549227155990999e-05, "loss": 0.0455, "step": 17940 }, { "grad_norm": 0.4154157340526581, "learning_rate": 9.548540804359828e-05, "loss": 0.0625, "step": 17950 }, { "grad_norm": 0.49597206711769104, "learning_rate": 9.547853955307077e-05, "loss": 0.0519, "step": 17960 }, { "grad_norm": 0.43923527002334595, "learning_rate": 9.547166608907853e-05, "loss": 0.0471, "step": 17970 }, { "grad_norm": 0.5647384524345398, "learning_rate": 9.546478765237326e-05, "loss": 0.0662, "step": 17980 }, { "grad_norm": 0.5532541275024414, "learning_rate": 9.545790424370715e-05, "loss": 0.053, "step": 17990 }, { "grad_norm": 0.46947717666625977, "learning_rate": 9.5451015863833e-05, "loss": 0.052, "step": 18000 }, { "grad_norm": 0.44317033886909485, "learning_rate": 9.544412251350408e-05, "loss": 0.0601, "step": 18010 }, { "grad_norm": 0.407524973154068, "learning_rate": 9.543722419347422e-05, "loss": 0.0536, "step": 18020 }, { "grad_norm": 0.43635204434394836, "learning_rate": 9.543032090449788e-05, "loss": 0.0557, "step": 18030 }, { "grad_norm": 0.33009591698646545, "learning_rate": 9.542341264732992e-05, "loss": 0.0573, "step": 18040 }, { "grad_norm": 0.5563318133354187, "learning_rate": 9.541649942272585e-05, "loss": 0.0629, "step": 18050 }, { "grad_norm": 0.6491063833236694, "learning_rate": 9.54095812314417e-05, "loss": 0.0518, "step": 18060 }, { "grad_norm": 0.4117320477962494, "learning_rate": 9.540265807423401e-05, "loss": 0.0531, "step": 18070 }, { "grad_norm": 0.4510433077812195, "learning_rate": 9.53957299518599e-05, "loss": 0.0493, "step": 18080 }, { "grad_norm": 0.4839191734790802, "learning_rate": 9.5388796865077e-05, "loss": 0.0452, "step": 18090 }, { "grad_norm": 0.5007832050323486, "learning_rate": 9.538185881464353e-05, "loss": 0.0551, "step": 18100 }, { "grad_norm": 0.4739995300769806, "learning_rate": 9.537491580131821e-05, "loss": 0.0479, "step": 18110 }, { "grad_norm": 0.4510435163974762, "learning_rate": 9.53679678258603e-05, "loss": 0.0522, "step": 18120 }, { "grad_norm": 0.54203861951828, "learning_rate": 9.536101488902966e-05, "loss": 0.0527, "step": 18130 }, { "grad_norm": 0.4203409254550934, "learning_rate": 9.535405699158663e-05, "loss": 0.047, "step": 18140 }, { "grad_norm": 0.4391612112522125, "learning_rate": 9.53470941342921e-05, "loss": 0.0502, "step": 18150 }, { "grad_norm": 0.37155094742774963, "learning_rate": 9.534012631790756e-05, "loss": 0.0525, "step": 18160 }, { "grad_norm": 0.44113633036613464, "learning_rate": 9.533315354319494e-05, "loss": 0.0485, "step": 18170 }, { "grad_norm": 0.5177717804908752, "learning_rate": 9.532617581091682e-05, "loss": 0.0485, "step": 18180 }, { "grad_norm": 0.3917389512062073, "learning_rate": 9.531919312183629e-05, "loss": 0.057, "step": 18190 }, { "grad_norm": 0.5494665503501892, "learning_rate": 9.531220547671688e-05, "loss": 0.0461, "step": 18200 }, { "grad_norm": 0.3136879801750183, "learning_rate": 9.530521287632285e-05, "loss": 0.0534, "step": 18210 }, { "grad_norm": 0.4685707986354828, "learning_rate": 9.529821532141884e-05, "loss": 0.044, "step": 18220 }, { "grad_norm": 0.3900633454322815, "learning_rate": 9.52912128127701e-05, "loss": 0.05, "step": 18230 }, { "grad_norm": 0.522423505783081, "learning_rate": 9.528420535114244e-05, "loss": 0.0474, "step": 18240 }, { "grad_norm": 0.5609028339385986, "learning_rate": 9.527719293730215e-05, "loss": 0.0579, "step": 18250 }, { "grad_norm": 0.46520623564720154, "learning_rate": 9.527017557201611e-05, "loss": 0.0482, "step": 18260 }, { "grad_norm": 0.44182950258255005, "learning_rate": 9.526315325605176e-05, "loss": 0.0565, "step": 18270 }, { "grad_norm": 0.43284791707992554, "learning_rate": 9.525612599017699e-05, "loss": 0.0452, "step": 18280 }, { "grad_norm": 0.5525587201118469, "learning_rate": 9.524909377516033e-05, "loss": 0.0498, "step": 18290 }, { "grad_norm": 0.5297389030456543, "learning_rate": 9.524205661177081e-05, "loss": 0.0374, "step": 18300 }, { "grad_norm": 0.4921639561653137, "learning_rate": 9.523501450077801e-05, "loss": 0.0456, "step": 18310 }, { "grad_norm": 0.3893413543701172, "learning_rate": 9.522796744295202e-05, "loss": 0.0442, "step": 18320 }, { "grad_norm": 0.7221903800964355, "learning_rate": 9.522091543906352e-05, "loss": 0.0474, "step": 18330 }, { "grad_norm": 0.4696462154388428, "learning_rate": 9.521385848988369e-05, "loss": 0.0514, "step": 18340 }, { "grad_norm": 0.5626357793807983, "learning_rate": 9.520679659618428e-05, "loss": 0.0487, "step": 18350 }, { "grad_norm": 0.42241814732551575, "learning_rate": 9.519972975873754e-05, "loss": 0.0586, "step": 18360 }, { "grad_norm": 0.39175719022750854, "learning_rate": 9.519265797831633e-05, "loss": 0.05, "step": 18370 }, { "grad_norm": 0.665592610836029, "learning_rate": 9.518558125569399e-05, "loss": 0.06, "step": 18380 }, { "grad_norm": 0.4412371516227722, "learning_rate": 9.517849959164442e-05, "loss": 0.0434, "step": 18390 }, { "grad_norm": 0.47466254234313965, "learning_rate": 9.517141298694205e-05, "loss": 0.0479, "step": 18400 }, { "grad_norm": 0.3635716438293457, "learning_rate": 9.516432144236188e-05, "loss": 0.0529, "step": 18410 }, { "grad_norm": 0.5353749394416809, "learning_rate": 9.515722495867941e-05, "loss": 0.0514, "step": 18420 }, { "grad_norm": 0.3450605571269989, "learning_rate": 9.515012353667072e-05, "loss": 0.0389, "step": 18430 }, { "grad_norm": 0.5053106546401978, "learning_rate": 9.51430171771124e-05, "loss": 0.0473, "step": 18440 }, { "grad_norm": 0.4034360349178314, "learning_rate": 9.513590588078159e-05, "loss": 0.0509, "step": 18450 }, { "grad_norm": 0.4591408669948578, "learning_rate": 9.512878964845597e-05, "loss": 0.0553, "step": 18460 }, { "grad_norm": 0.5712065100669861, "learning_rate": 9.512166848091377e-05, "loss": 0.0559, "step": 18470 }, { "grad_norm": 0.4470430612564087, "learning_rate": 9.511454237893376e-05, "loss": 0.0511, "step": 18480 }, { "grad_norm": 0.3820447325706482, "learning_rate": 9.51074113432952e-05, "loss": 0.0435, "step": 18490 }, { "grad_norm": 0.4092724621295929, "learning_rate": 9.510027537477797e-05, "loss": 0.0576, "step": 18500 }, { "grad_norm": 0.540589451789856, "learning_rate": 9.509313447416242e-05, "loss": 0.0534, "step": 18510 }, { "grad_norm": 0.47474753856658936, "learning_rate": 9.508598864222949e-05, "loss": 0.0518, "step": 18520 }, { "grad_norm": 0.4479869604110718, "learning_rate": 9.507883787976062e-05, "loss": 0.049, "step": 18530 }, { "grad_norm": 0.4461944103240967, "learning_rate": 9.507168218753781e-05, "loss": 0.0452, "step": 18540 }, { "grad_norm": 0.34074312448501587, "learning_rate": 9.506452156634362e-05, "loss": 0.0466, "step": 18550 }, { "grad_norm": 0.38499242067337036, "learning_rate": 9.505735601696109e-05, "loss": 0.0645, "step": 18560 }, { "grad_norm": 0.5674631595611572, "learning_rate": 9.505018554017385e-05, "loss": 0.0528, "step": 18570 }, { "grad_norm": 0.5876055955886841, "learning_rate": 9.504301013676604e-05, "loss": 0.0493, "step": 18580 }, { "grad_norm": 0.5808190107345581, "learning_rate": 9.503582980752238e-05, "loss": 0.0478, "step": 18590 }, { "grad_norm": 0.6004931330680847, "learning_rate": 9.502864455322809e-05, "loss": 0.0694, "step": 18600 }, { "grad_norm": 0.3428039252758026, "learning_rate": 9.502145437466891e-05, "loss": 0.0491, "step": 18610 }, { "grad_norm": 0.36577799916267395, "learning_rate": 9.501425927263116e-05, "loss": 0.0551, "step": 18620 }, { "grad_norm": 0.6230515241622925, "learning_rate": 9.500705924790172e-05, "loss": 0.0773, "step": 18630 }, { "grad_norm": 0.41191330552101135, "learning_rate": 9.499985430126794e-05, "loss": 0.0479, "step": 18640 }, { "grad_norm": 0.5179558992385864, "learning_rate": 9.499264443351775e-05, "loss": 0.0517, "step": 18650 }, { "grad_norm": 0.3726932108402252, "learning_rate": 9.498542964543961e-05, "loss": 0.0507, "step": 18660 }, { "grad_norm": 0.3894786834716797, "learning_rate": 9.497820993782252e-05, "loss": 0.0543, "step": 18670 }, { "grad_norm": 0.4583604633808136, "learning_rate": 9.497098531145601e-05, "loss": 0.0509, "step": 18680 }, { "grad_norm": 0.4888085722923279, "learning_rate": 9.496375576713017e-05, "loss": 0.0525, "step": 18690 }, { "grad_norm": 0.37965384125709534, "learning_rate": 9.49565213056356e-05, "loss": 0.0494, "step": 18700 }, { "grad_norm": 0.42777496576309204, "learning_rate": 9.494928192776342e-05, "loss": 0.0477, "step": 18710 }, { "grad_norm": 0.6909065246582031, "learning_rate": 9.494203763430538e-05, "loss": 0.0404, "step": 18720 }, { "grad_norm": 0.5261523127555847, "learning_rate": 9.493478842605366e-05, "loss": 0.0788, "step": 18730 }, { "grad_norm": 0.31503787636756897, "learning_rate": 9.492753430380105e-05, "loss": 0.0442, "step": 18740 }, { "grad_norm": 0.37808525562286377, "learning_rate": 9.492027526834083e-05, "loss": 0.0458, "step": 18750 }, { "grad_norm": 0.41300368309020996, "learning_rate": 9.491301132046684e-05, "loss": 0.0439, "step": 18760 }, { "grad_norm": 0.5372617244720459, "learning_rate": 9.490574246097345e-05, "loss": 0.0739, "step": 18770 }, { "grad_norm": 0.5506138801574707, "learning_rate": 9.48984686906556e-05, "loss": 0.056, "step": 18780 }, { "grad_norm": 0.5373836159706116, "learning_rate": 9.489119001030871e-05, "loss": 0.0504, "step": 18790 }, { "grad_norm": 0.50945645570755, "learning_rate": 9.488390642072878e-05, "loss": 0.0562, "step": 18800 }, { "grad_norm": 0.31330710649490356, "learning_rate": 9.48766179227123e-05, "loss": 0.0509, "step": 18810 }, { "grad_norm": 0.48142436146736145, "learning_rate": 9.486932451705636e-05, "loss": 0.0539, "step": 18820 }, { "grad_norm": 0.433079332113266, "learning_rate": 9.486202620455857e-05, "loss": 0.0512, "step": 18830 }, { "grad_norm": 0.546686053276062, "learning_rate": 9.485472298601704e-05, "loss": 0.045, "step": 18840 }, { "grad_norm": 0.39655083417892456, "learning_rate": 9.484741486223043e-05, "loss": 0.0528, "step": 18850 }, { "grad_norm": 0.37906578183174133, "learning_rate": 9.484010183399797e-05, "loss": 0.0516, "step": 18860 }, { "grad_norm": 0.7344197034835815, "learning_rate": 9.483278390211938e-05, "loss": 0.0558, "step": 18870 }, { "grad_norm": 0.584588348865509, "learning_rate": 9.482546106739496e-05, "loss": 0.0464, "step": 18880 }, { "grad_norm": 0.5519776940345764, "learning_rate": 9.48181333306255e-05, "loss": 0.0527, "step": 18890 }, { "grad_norm": 0.44957271218299866, "learning_rate": 9.481080069261237e-05, "loss": 0.0636, "step": 18900 }, { "grad_norm": 0.40385904908180237, "learning_rate": 9.480346315415745e-05, "loss": 0.0492, "step": 18910 }, { "grad_norm": 1.3274081945419312, "learning_rate": 9.479612071606314e-05, "loss": 0.0537, "step": 18920 }, { "grad_norm": 0.5666117072105408, "learning_rate": 9.478877337913244e-05, "loss": 0.0675, "step": 18930 }, { "grad_norm": 0.6217034459114075, "learning_rate": 9.478142114416881e-05, "loss": 0.0582, "step": 18940 }, { "grad_norm": 0.39516904950141907, "learning_rate": 9.47740640119763e-05, "loss": 0.0471, "step": 18950 }, { "grad_norm": 0.3661578595638275, "learning_rate": 9.476670198335947e-05, "loss": 0.0524, "step": 18960 }, { "grad_norm": 0.42083677649497986, "learning_rate": 9.47593350591234e-05, "loss": 0.0526, "step": 18970 }, { "grad_norm": 0.4377959072589874, "learning_rate": 9.475196324007376e-05, "loss": 0.0492, "step": 18980 }, { "grad_norm": 0.5865117907524109, "learning_rate": 9.474458652701669e-05, "loss": 0.0452, "step": 18990 }, { "grad_norm": 0.4664327800273895, "learning_rate": 9.473720492075892e-05, "loss": 0.0544, "step": 19000 }, { "grad_norm": 0.5437051057815552, "learning_rate": 9.472981842210768e-05, "loss": 0.0579, "step": 19010 }, { "grad_norm": 0.4879036247730255, "learning_rate": 9.472242703187074e-05, "loss": 0.0611, "step": 19020 }, { "grad_norm": 0.45254749059677124, "learning_rate": 9.471503075085643e-05, "loss": 0.0383, "step": 19030 }, { "grad_norm": 0.4507561922073364, "learning_rate": 9.470762957987359e-05, "loss": 0.0585, "step": 19040 }, { "grad_norm": 0.4728252589702606, "learning_rate": 9.470022351973158e-05, "loss": 0.0519, "step": 19050 }, { "grad_norm": 0.43402010202407837, "learning_rate": 9.469281257124034e-05, "loss": 0.0507, "step": 19060 }, { "grad_norm": 0.475658655166626, "learning_rate": 9.46853967352103e-05, "loss": 0.0539, "step": 19070 }, { "grad_norm": 0.4480774998664856, "learning_rate": 9.467797601245246e-05, "loss": 0.0457, "step": 19080 }, { "grad_norm": 0.5420652627944946, "learning_rate": 9.467055040377834e-05, "loss": 0.0635, "step": 19090 }, { "grad_norm": 0.47771313786506653, "learning_rate": 9.466311990999999e-05, "loss": 0.06, "step": 19100 }, { "grad_norm": 0.3296869993209839, "learning_rate": 9.465568453193e-05, "loss": 0.0507, "step": 19110 }, { "grad_norm": 0.3667049705982208, "learning_rate": 9.464824427038148e-05, "loss": 0.0501, "step": 19120 }, { "grad_norm": 0.3796900808811188, "learning_rate": 9.46407991261681e-05, "loss": 0.0572, "step": 19130 }, { "grad_norm": 0.37323489785194397, "learning_rate": 9.463334910010404e-05, "loss": 0.0471, "step": 19140 }, { "grad_norm": 0.42197713255882263, "learning_rate": 9.462589419300403e-05, "loss": 0.0417, "step": 19150 }, { "grad_norm": 0.483585923910141, "learning_rate": 9.461843440568333e-05, "loss": 0.0495, "step": 19160 }, { "grad_norm": 0.5606008768081665, "learning_rate": 9.461096973895773e-05, "loss": 0.0445, "step": 19170 }, { "grad_norm": 0.4369919002056122, "learning_rate": 9.460350019364355e-05, "loss": 0.0578, "step": 19180 }, { "grad_norm": 0.4553016424179077, "learning_rate": 9.459602577055764e-05, "loss": 0.0618, "step": 19190 }, { "grad_norm": 0.4930660128593445, "learning_rate": 9.45885464705174e-05, "loss": 0.0493, "step": 19200 }, { "grad_norm": 0.3540933132171631, "learning_rate": 9.458106229434076e-05, "loss": 0.0456, "step": 19210 }, { "grad_norm": 0.546057939529419, "learning_rate": 9.457357324284617e-05, "loss": 0.0375, "step": 19220 }, { "grad_norm": 0.39140933752059937, "learning_rate": 9.456607931685262e-05, "loss": 0.0448, "step": 19230 }, { "grad_norm": 0.3957163989543915, "learning_rate": 9.455858051717965e-05, "loss": 0.0408, "step": 19240 }, { "grad_norm": 0.515529215335846, "learning_rate": 9.45510768446473e-05, "loss": 0.0475, "step": 19250 }, { "grad_norm": 0.39948874711990356, "learning_rate": 9.454356830007618e-05, "loss": 0.0423, "step": 19260 }, { "grad_norm": 0.5938382148742676, "learning_rate": 9.45360548842874e-05, "loss": 0.0372, "step": 19270 }, { "grad_norm": 0.40671825408935547, "learning_rate": 9.452853659810261e-05, "loss": 0.0434, "step": 19280 }, { "grad_norm": 0.4260565936565399, "learning_rate": 9.452101344234401e-05, "loss": 0.057, "step": 19290 }, { "grad_norm": 0.44567376375198364, "learning_rate": 9.451348541783431e-05, "loss": 0.0499, "step": 19300 }, { "grad_norm": 0.4881535768508911, "learning_rate": 9.450595252539678e-05, "loss": 0.0484, "step": 19310 }, { "grad_norm": 0.5192577242851257, "learning_rate": 9.449841476585518e-05, "loss": 0.0647, "step": 19320 }, { "grad_norm": 0.5051904320716858, "learning_rate": 9.449087214003384e-05, "loss": 0.0561, "step": 19330 }, { "grad_norm": 0.5250931978225708, "learning_rate": 9.448332464875765e-05, "loss": 0.0704, "step": 19340 }, { "grad_norm": 0.39174389839172363, "learning_rate": 9.447577229285192e-05, "loss": 0.0595, "step": 19350 }, { "grad_norm": 0.4097927510738373, "learning_rate": 9.446821507314261e-05, "loss": 0.0413, "step": 19360 }, { "grad_norm": 0.4862077832221985, "learning_rate": 9.446065299045617e-05, "loss": 0.0504, "step": 19370 }, { "grad_norm": 0.6286983489990234, "learning_rate": 9.445308604561955e-05, "loss": 0.0558, "step": 19380 }, { "grad_norm": 0.42296525835990906, "learning_rate": 9.444551423946028e-05, "loss": 0.0503, "step": 19390 }, { "grad_norm": 0.3878762125968933, "learning_rate": 9.443793757280638e-05, "loss": 0.0483, "step": 19400 }, { "grad_norm": 0.4748712480068207, "learning_rate": 9.443035604648646e-05, "loss": 0.0519, "step": 19410 }, { "grad_norm": 0.5142635703086853, "learning_rate": 9.44227696613296e-05, "loss": 0.0463, "step": 19420 }, { "grad_norm": 0.5391756892204285, "learning_rate": 9.441517841816542e-05, "loss": 0.0708, "step": 19430 }, { "grad_norm": 0.5727838277816772, "learning_rate": 9.440758231782413e-05, "loss": 0.0447, "step": 19440 }, { "grad_norm": 0.5115869641304016, "learning_rate": 9.439998136113639e-05, "loss": 0.0538, "step": 19450 }, { "grad_norm": 0.5334614515304565, "learning_rate": 9.439237554893344e-05, "loss": 0.0552, "step": 19460 }, { "grad_norm": 0.3495175838470459, "learning_rate": 9.438476488204705e-05, "loss": 0.0576, "step": 19470 }, { "grad_norm": 0.335957795381546, "learning_rate": 9.43771493613095e-05, "loss": 0.0473, "step": 19480 }, { "grad_norm": 0.5026257038116455, "learning_rate": 9.436952898755362e-05, "loss": 0.0588, "step": 19490 }, { "grad_norm": 0.4898719787597656, "learning_rate": 9.436190376161276e-05, "loss": 0.0481, "step": 19500 }, { "grad_norm": 0.5302513241767883, "learning_rate": 9.43542736843208e-05, "loss": 0.0631, "step": 19510 }, { "grad_norm": 0.4774457514286041, "learning_rate": 9.434663875651216e-05, "loss": 0.0437, "step": 19520 }, { "grad_norm": 0.4560105800628662, "learning_rate": 9.433899897902177e-05, "loss": 0.0493, "step": 19530 }, { "grad_norm": 0.5891119241714478, "learning_rate": 9.433135435268511e-05, "loss": 0.048, "step": 19540 }, { "grad_norm": 0.4225750267505646, "learning_rate": 9.432370487833819e-05, "loss": 0.0587, "step": 19550 }, { "grad_norm": 0.48492103815078735, "learning_rate": 9.431605055681756e-05, "loss": 0.0608, "step": 19560 }, { "grad_norm": 0.35438719391822815, "learning_rate": 9.430839138896026e-05, "loss": 0.0452, "step": 19570 }, { "grad_norm": 0.30930840969085693, "learning_rate": 9.43007273756039e-05, "loss": 0.0521, "step": 19580 }, { "grad_norm": 0.3507695496082306, "learning_rate": 9.429305851758658e-05, "loss": 0.0427, "step": 19590 }, { "grad_norm": 0.5875266790390015, "learning_rate": 9.428538481574699e-05, "loss": 0.0559, "step": 19600 }, { "grad_norm": 0.3679613173007965, "learning_rate": 9.42777062709243e-05, "loss": 0.0436, "step": 19610 }, { "grad_norm": 0.3449770510196686, "learning_rate": 9.427002288395821e-05, "loss": 0.0457, "step": 19620 }, { "grad_norm": 0.541044294834137, "learning_rate": 9.426233465568898e-05, "loss": 0.0499, "step": 19630 }, { "grad_norm": 0.29945874214172363, "learning_rate": 9.42546415869574e-05, "loss": 0.038, "step": 19640 }, { "grad_norm": 0.4064904451370239, "learning_rate": 9.424694367860473e-05, "loss": 0.0439, "step": 19650 }, { "grad_norm": 0.451157808303833, "learning_rate": 9.423924093147284e-05, "loss": 0.0436, "step": 19660 }, { "grad_norm": 0.40651118755340576, "learning_rate": 9.423153334640407e-05, "loss": 0.0453, "step": 19670 }, { "grad_norm": 0.47182783484458923, "learning_rate": 9.42238209242413e-05, "loss": 0.0445, "step": 19680 }, { "grad_norm": 0.2919749915599823, "learning_rate": 9.421610366582798e-05, "loss": 0.0517, "step": 19690 }, { "grad_norm": 0.3905106484889984, "learning_rate": 9.420838157200803e-05, "loss": 0.0485, "step": 19700 }, { "grad_norm": 0.3607258200645447, "learning_rate": 9.420065464362594e-05, "loss": 0.0595, "step": 19710 }, { "grad_norm": 0.39517587423324585, "learning_rate": 9.419292288152673e-05, "loss": 0.0485, "step": 19720 }, { "grad_norm": 0.4831292927265167, "learning_rate": 9.418518628655588e-05, "loss": 0.0483, "step": 19730 }, { "grad_norm": 0.7606999278068542, "learning_rate": 9.417744485955951e-05, "loss": 0.0515, "step": 19740 }, { "grad_norm": 0.35772597789764404, "learning_rate": 9.41696986013842e-05, "loss": 0.0434, "step": 19750 }, { "grad_norm": 0.3667316734790802, "learning_rate": 9.416194751287705e-05, "loss": 0.0579, "step": 19760 }, { "grad_norm": 0.4231189489364624, "learning_rate": 9.415419159488572e-05, "loss": 0.0491, "step": 19770 }, { "grad_norm": 0.45522022247314453, "learning_rate": 9.414643084825837e-05, "loss": 0.0576, "step": 19780 }, { "grad_norm": 0.5706002712249756, "learning_rate": 9.413866527384372e-05, "loss": 0.0543, "step": 19790 }, { "grad_norm": 0.4350029230117798, "learning_rate": 9.4130894872491e-05, "loss": 0.0499, "step": 19800 }, { "grad_norm": 0.42925581336021423, "learning_rate": 9.412311964504998e-05, "loss": 0.0513, "step": 19810 }, { "grad_norm": 0.530522882938385, "learning_rate": 9.411533959237091e-05, "loss": 0.0611, "step": 19820 }, { "grad_norm": 0.4644199311733246, "learning_rate": 9.410755471530464e-05, "loss": 0.0449, "step": 19830 }, { "grad_norm": 0.4749685823917389, "learning_rate": 9.40997650147025e-05, "loss": 0.0481, "step": 19840 }, { "grad_norm": 0.5252326130867004, "learning_rate": 9.409197049141637e-05, "loss": 0.0621, "step": 19850 }, { "grad_norm": 0.4967344403266907, "learning_rate": 9.408417114629863e-05, "loss": 0.0557, "step": 19860 }, { "grad_norm": 0.5068467259407043, "learning_rate": 9.40763669802022e-05, "loss": 0.0488, "step": 19870 }, { "grad_norm": 0.7673826217651367, "learning_rate": 9.406855799398056e-05, "loss": 0.0564, "step": 19880 }, { "grad_norm": 0.45441973209381104, "learning_rate": 9.406074418848767e-05, "loss": 0.0521, "step": 19890 }, { "grad_norm": 0.6308345198631287, "learning_rate": 9.405292556457805e-05, "loss": 0.0714, "step": 19900 }, { "grad_norm": 0.49532610177993774, "learning_rate": 9.404510212310671e-05, "loss": 0.0584, "step": 19910 }, { "grad_norm": 0.502753496170044, "learning_rate": 9.403727386492924e-05, "loss": 0.0544, "step": 19920 }, { "grad_norm": 0.34424078464508057, "learning_rate": 9.40294407909017e-05, "loss": 0.0543, "step": 19930 }, { "grad_norm": 0.45275866985321045, "learning_rate": 9.40216029018807e-05, "loss": 0.0509, "step": 19940 }, { "grad_norm": 0.8925970196723938, "learning_rate": 9.401376019872338e-05, "loss": 0.0523, "step": 19950 }, { "grad_norm": 0.5214274525642395, "learning_rate": 9.400591268228746e-05, "loss": 0.0616, "step": 19960 }, { "grad_norm": 0.4176652431488037, "learning_rate": 9.399806035343106e-05, "loss": 0.0497, "step": 19970 }, { "grad_norm": 0.5507798790931702, "learning_rate": 9.399020321301294e-05, "loss": 0.0488, "step": 19980 }, { "grad_norm": 0.6705978512763977, "learning_rate": 9.398234126189234e-05, "loss": 0.0534, "step": 19990 }, { "grad_norm": 0.5746465921401978, "learning_rate": 9.397447450092902e-05, "loss": 0.0472, "step": 20000 }, { "grad_norm": 0.44280296564102173, "learning_rate": 9.39666029309833e-05, "loss": 0.0438, "step": 20010 }, { "grad_norm": 0.6742819547653198, "learning_rate": 9.395872655291596e-05, "loss": 0.0467, "step": 20020 }, { "grad_norm": 0.4676345884799957, "learning_rate": 9.395084536758838e-05, "loss": 0.0411, "step": 20030 }, { "grad_norm": 0.4869939088821411, "learning_rate": 9.394295937586243e-05, "loss": 0.0589, "step": 20040 }, { "grad_norm": 0.5055854916572571, "learning_rate": 9.393506857860052e-05, "loss": 0.0487, "step": 20050 }, { "grad_norm": 0.5323932766914368, "learning_rate": 9.392717297666555e-05, "loss": 0.0367, "step": 20060 }, { "grad_norm": 0.4954873025417328, "learning_rate": 9.391927257092101e-05, "loss": 0.0537, "step": 20070 }, { "grad_norm": 0.4652024507522583, "learning_rate": 9.391136736223085e-05, "loss": 0.0475, "step": 20080 }, { "grad_norm": 0.47313857078552246, "learning_rate": 9.390345735145956e-05, "loss": 0.0562, "step": 20090 }, { "grad_norm": 0.47748711705207825, "learning_rate": 9.389554253947219e-05, "loss": 0.068, "step": 20100 }, { "grad_norm": 0.2936980426311493, "learning_rate": 9.388762292713428e-05, "loss": 0.0496, "step": 20110 }, { "grad_norm": 0.4932718873023987, "learning_rate": 9.38796985153119e-05, "loss": 0.044, "step": 20120 }, { "grad_norm": 0.3987932503223419, "learning_rate": 9.387176930487169e-05, "loss": 0.0502, "step": 20130 }, { "grad_norm": 0.6004276871681213, "learning_rate": 9.386383529668072e-05, "loss": 0.0463, "step": 20140 }, { "grad_norm": 0.5139516592025757, "learning_rate": 9.385589649160669e-05, "loss": 0.048, "step": 20150 }, { "grad_norm": 0.5847150087356567, "learning_rate": 9.384795289051775e-05, "loss": 0.0558, "step": 20160 }, { "grad_norm": 0.45258820056915283, "learning_rate": 9.384000449428261e-05, "loss": 0.0528, "step": 20170 }, { "grad_norm": 0.3850330710411072, "learning_rate": 9.383205130377048e-05, "loss": 0.0622, "step": 20180 }, { "grad_norm": 0.3635770082473755, "learning_rate": 9.382409331985114e-05, "loss": 0.0541, "step": 20190 }, { "grad_norm": 0.474658340215683, "learning_rate": 9.381613054339482e-05, "loss": 0.0471, "step": 20200 }, { "grad_norm": 0.38126108050346375, "learning_rate": 9.380816297527235e-05, "loss": 0.0397, "step": 20210 }, { "grad_norm": 0.6105748414993286, "learning_rate": 9.380019061635506e-05, "loss": 0.052, "step": 20220 }, { "grad_norm": 0.41751572489738464, "learning_rate": 9.379221346751474e-05, "loss": 0.0561, "step": 20230 }, { "grad_norm": 0.4644680619239807, "learning_rate": 9.378423152962382e-05, "loss": 0.0372, "step": 20240 }, { "grad_norm": 0.726047933101654, "learning_rate": 9.377624480355517e-05, "loss": 0.0437, "step": 20250 }, { "grad_norm": 0.5635186433792114, "learning_rate": 9.376825329018219e-05, "loss": 0.0405, "step": 20260 }, { "grad_norm": 0.4425557255744934, "learning_rate": 9.376025699037884e-05, "loss": 0.0451, "step": 20270 }, { "grad_norm": 0.46599945425987244, "learning_rate": 9.37522559050196e-05, "loss": 0.0433, "step": 20280 }, { "grad_norm": 0.42834973335266113, "learning_rate": 9.37442500349794e-05, "loss": 0.0558, "step": 20290 }, { "grad_norm": 0.4506751596927643, "learning_rate": 9.373623938113381e-05, "loss": 0.0445, "step": 20300 }, { "grad_norm": 0.3489236533641815, "learning_rate": 9.372822394435883e-05, "loss": 0.0553, "step": 20310 }, { "grad_norm": 0.6001182794570923, "learning_rate": 9.372020372553102e-05, "loss": 0.0476, "step": 20320 }, { "grad_norm": 0.4229689836502075, "learning_rate": 9.371217872552746e-05, "loss": 0.046, "step": 20330 }, { "grad_norm": 0.32951977849006653, "learning_rate": 9.370414894522576e-05, "loss": 0.0512, "step": 20340 }, { "grad_norm": 0.5057783126831055, "learning_rate": 9.369611438550406e-05, "loss": 0.0526, "step": 20350 }, { "grad_norm": 0.5058923959732056, "learning_rate": 9.368807504724095e-05, "loss": 0.0451, "step": 20360 }, { "grad_norm": 0.45289263129234314, "learning_rate": 9.368003093131565e-05, "loss": 0.054, "step": 20370 }, { "grad_norm": 0.4111449420452118, "learning_rate": 9.367198203860785e-05, "loss": 0.053, "step": 20380 }, { "grad_norm": 0.3750957250595093, "learning_rate": 9.366392836999774e-05, "loss": 0.042, "step": 20390 }, { "grad_norm": 0.260827898979187, "learning_rate": 9.365586992636607e-05, "loss": 0.0356, "step": 20400 }, { "grad_norm": 0.3707539141178131, "learning_rate": 9.364780670859412e-05, "loss": 0.0384, "step": 20410 }, { "grad_norm": 0.42773517966270447, "learning_rate": 9.363973871756364e-05, "loss": 0.0414, "step": 20420 }, { "grad_norm": 0.35280510783195496, "learning_rate": 9.363166595415696e-05, "loss": 0.0359, "step": 20430 }, { "grad_norm": 0.3992975950241089, "learning_rate": 9.362358841925686e-05, "loss": 0.0461, "step": 20440 }, { "grad_norm": 0.3926234543323517, "learning_rate": 9.361550611374674e-05, "loss": 0.0485, "step": 20450 }, { "grad_norm": 0.35423195362091064, "learning_rate": 9.360741903851043e-05, "loss": 0.0612, "step": 20460 }, { "grad_norm": 0.49429765343666077, "learning_rate": 9.359932719443236e-05, "loss": 0.0584, "step": 20470 }, { "grad_norm": 0.47858479619026184, "learning_rate": 9.35912305823974e-05, "loss": 0.0472, "step": 20480 }, { "grad_norm": 0.5044511556625366, "learning_rate": 9.358312920329101e-05, "loss": 0.0378, "step": 20490 }, { "grad_norm": 0.41789624094963074, "learning_rate": 9.357502305799914e-05, "loss": 0.045, "step": 20500 }, { "grad_norm": 0.4368109703063965, "learning_rate": 9.356691214740824e-05, "loss": 0.0472, "step": 20510 }, { "grad_norm": 0.3640827536582947, "learning_rate": 9.355879647240535e-05, "loss": 0.0449, "step": 20520 }, { "grad_norm": 0.43481308221817017, "learning_rate": 9.355067603387798e-05, "loss": 0.0558, "step": 20530 }, { "grad_norm": 0.47043758630752563, "learning_rate": 9.354255083271412e-05, "loss": 0.0455, "step": 20540 }, { "grad_norm": 0.5062864422798157, "learning_rate": 9.353442086980239e-05, "loss": 0.0521, "step": 20550 }, { "grad_norm": 0.45243269205093384, "learning_rate": 9.352628614603185e-05, "loss": 0.0416, "step": 20560 }, { "grad_norm": 0.43512189388275146, "learning_rate": 9.351814666229209e-05, "loss": 0.0455, "step": 20570 }, { "grad_norm": 0.3957905173301697, "learning_rate": 9.351000241947324e-05, "loss": 0.0457, "step": 20580 }, { "grad_norm": 0.4456646740436554, "learning_rate": 9.350185341846594e-05, "loss": 0.0486, "step": 20590 }, { "grad_norm": 0.5361337065696716, "learning_rate": 9.349369966016134e-05, "loss": 0.0447, "step": 20600 }, { "grad_norm": 0.4397842586040497, "learning_rate": 9.348554114545117e-05, "loss": 0.0616, "step": 20610 }, { "grad_norm": 0.39642345905303955, "learning_rate": 9.347737787522758e-05, "loss": 0.0461, "step": 20620 }, { "grad_norm": 0.40528205037117004, "learning_rate": 9.346920985038332e-05, "loss": 0.0515, "step": 20630 }, { "grad_norm": 0.47673484683036804, "learning_rate": 9.346103707181162e-05, "loss": 0.0503, "step": 20640 }, { "grad_norm": 0.5669925212860107, "learning_rate": 9.345285954040626e-05, "loss": 0.046, "step": 20650 }, { "grad_norm": 0.33866381645202637, "learning_rate": 9.34446772570615e-05, "loss": 0.0461, "step": 20660 }, { "grad_norm": 0.28761762380599976, "learning_rate": 9.343649022267214e-05, "loss": 0.0398, "step": 20670 }, { "grad_norm": 0.4410600960254669, "learning_rate": 9.342829843813353e-05, "loss": 0.0458, "step": 20680 }, { "grad_norm": 0.5437139272689819, "learning_rate": 9.342010190434149e-05, "loss": 0.048, "step": 20690 }, { "grad_norm": 0.48068997263908386, "learning_rate": 9.34119006221924e-05, "loss": 0.0459, "step": 20700 }, { "grad_norm": 0.4164782464504242, "learning_rate": 9.340369459258313e-05, "loss": 0.0423, "step": 20710 }, { "grad_norm": 0.44661495089530945, "learning_rate": 9.339548381641106e-05, "loss": 0.0529, "step": 20720 }, { "grad_norm": 0.6789365410804749, "learning_rate": 9.338726829457413e-05, "loss": 0.0503, "step": 20730 }, { "grad_norm": 0.4898708462715149, "learning_rate": 9.337904802797078e-05, "loss": 0.0376, "step": 20740 }, { "grad_norm": 0.5176379084587097, "learning_rate": 9.337082301749993e-05, "loss": 0.0522, "step": 20750 }, { "grad_norm": 0.3616733253002167, "learning_rate": 9.336259326406109e-05, "loss": 0.0492, "step": 20760 }, { "grad_norm": 0.39535078406333923, "learning_rate": 9.335435876855427e-05, "loss": 0.0543, "step": 20770 }, { "grad_norm": 0.4474828243255615, "learning_rate": 9.334611953187994e-05, "loss": 0.0492, "step": 20780 }, { "grad_norm": 0.43003931641578674, "learning_rate": 9.333787555493914e-05, "loss": 0.0407, "step": 20790 }, { "grad_norm": 0.42416608333587646, "learning_rate": 9.332962683863345e-05, "loss": 0.055, "step": 20800 }, { "grad_norm": 0.43892142176628113, "learning_rate": 9.332137338386489e-05, "loss": 0.044, "step": 20810 }, { "grad_norm": 0.33563101291656494, "learning_rate": 9.33131151915361e-05, "loss": 0.0576, "step": 20820 }, { "grad_norm": 0.3466498851776123, "learning_rate": 9.330485226255012e-05, "loss": 0.0433, "step": 20830 }, { "grad_norm": 0.4576544165611267, "learning_rate": 9.329658459781061e-05, "loss": 0.0512, "step": 20840 }, { "grad_norm": 0.3860737681388855, "learning_rate": 9.328831219822172e-05, "loss": 0.0476, "step": 20850 }, { "grad_norm": 0.71695476770401, "learning_rate": 9.328003506468808e-05, "loss": 0.0614, "step": 20860 }, { "grad_norm": 0.4413420557975769, "learning_rate": 9.327175319811488e-05, "loss": 0.046, "step": 20870 }, { "grad_norm": 0.3934735655784607, "learning_rate": 9.326346659940781e-05, "loss": 0.0467, "step": 20880 }, { "grad_norm": 0.38009193539619446, "learning_rate": 9.325517526947308e-05, "loss": 0.0459, "step": 20890 }, { "grad_norm": 0.5269112586975098, "learning_rate": 9.32468792092174e-05, "loss": 0.0472, "step": 20900 }, { "grad_norm": 0.433735191822052, "learning_rate": 9.323857841954803e-05, "loss": 0.0441, "step": 20910 }, { "grad_norm": 0.33255407214164734, "learning_rate": 9.323027290137276e-05, "loss": 0.0423, "step": 20920 }, { "grad_norm": 0.32273274660110474, "learning_rate": 9.322196265559981e-05, "loss": 0.0409, "step": 20930 }, { "grad_norm": 0.4127539396286011, "learning_rate": 9.321364768313803e-05, "loss": 0.0462, "step": 20940 }, { "grad_norm": 0.33253100514411926, "learning_rate": 9.32053279848967e-05, "loss": 0.0412, "step": 20950 }, { "grad_norm": 0.3965807855129242, "learning_rate": 9.319700356178567e-05, "loss": 0.0512, "step": 20960 }, { "grad_norm": 0.6071175932884216, "learning_rate": 9.318867441471527e-05, "loss": 0.051, "step": 20970 }, { "grad_norm": 0.48784786462783813, "learning_rate": 9.318034054459637e-05, "loss": 0.0379, "step": 20980 }, { "grad_norm": 0.5370292663574219, "learning_rate": 9.317200195234034e-05, "loss": 0.0442, "step": 20990 }, { "grad_norm": 0.45090770721435547, "learning_rate": 9.316365863885909e-05, "loss": 0.0474, "step": 21000 }, { "grad_norm": 0.42823630571365356, "learning_rate": 9.315531060506502e-05, "loss": 0.0595, "step": 21010 }, { "grad_norm": 0.344234436750412, "learning_rate": 9.314695785187108e-05, "loss": 0.0382, "step": 21020 }, { "grad_norm": 0.7027627229690552, "learning_rate": 9.313860038019069e-05, "loss": 0.0484, "step": 21030 }, { "grad_norm": 0.3222198188304901, "learning_rate": 9.313023819093782e-05, "loss": 0.0398, "step": 21040 }, { "grad_norm": 0.5403289198875427, "learning_rate": 9.312187128502695e-05, "loss": 0.0507, "step": 21050 }, { "grad_norm": 0.30722641944885254, "learning_rate": 9.311349966337307e-05, "loss": 0.0527, "step": 21060 }, { "grad_norm": 0.28567183017730713, "learning_rate": 9.310512332689169e-05, "loss": 0.0571, "step": 21070 }, { "grad_norm": 0.5375676155090332, "learning_rate": 9.309674227649883e-05, "loss": 0.052, "step": 21080 }, { "grad_norm": 0.48199838399887085, "learning_rate": 9.308835651311103e-05, "loss": 0.0533, "step": 21090 }, { "grad_norm": 0.4348083734512329, "learning_rate": 9.307996603764533e-05, "loss": 0.0553, "step": 21100 }, { "grad_norm": 0.548681914806366, "learning_rate": 9.307157085101932e-05, "loss": 0.0529, "step": 21110 }, { "grad_norm": 0.4986298382282257, "learning_rate": 9.306317095415109e-05, "loss": 0.041, "step": 21120 }, { "grad_norm": 0.4755251109600067, "learning_rate": 9.305476634795922e-05, "loss": 0.0467, "step": 21130 }, { "grad_norm": 0.32235440611839294, "learning_rate": 9.304635703336284e-05, "loss": 0.0419, "step": 21140 }, { "grad_norm": 0.40844225883483887, "learning_rate": 9.303794301128157e-05, "loss": 0.0406, "step": 21150 }, { "grad_norm": 0.4109501838684082, "learning_rate": 9.302952428263555e-05, "loss": 0.0494, "step": 21160 }, { "grad_norm": 0.3486318588256836, "learning_rate": 9.302110084834545e-05, "loss": 0.0473, "step": 21170 }, { "grad_norm": 0.3787235617637634, "learning_rate": 9.301267270933245e-05, "loss": 0.0401, "step": 21180 }, { "grad_norm": 0.4668937623500824, "learning_rate": 9.300423986651823e-05, "loss": 0.05, "step": 21190 }, { "grad_norm": 0.5459657311439514, "learning_rate": 9.299580232082501e-05, "loss": 0.048, "step": 21200 }, { "grad_norm": 0.5569335222244263, "learning_rate": 9.298736007317547e-05, "loss": 0.047, "step": 21210 }, { "grad_norm": 0.3558817505836487, "learning_rate": 9.297891312449288e-05, "loss": 0.0465, "step": 21220 }, { "grad_norm": 0.2491961419582367, "learning_rate": 9.297046147570094e-05, "loss": 0.0478, "step": 21230 }, { "grad_norm": 0.3467549979686737, "learning_rate": 9.296200512772396e-05, "loss": 0.0417, "step": 21240 }, { "grad_norm": 0.45671334862709045, "learning_rate": 9.295354408148668e-05, "loss": 0.0451, "step": 21250 }, { "grad_norm": 0.3677118420600891, "learning_rate": 9.294507833791441e-05, "loss": 0.048, "step": 21260 }, { "grad_norm": 0.40052470564842224, "learning_rate": 9.293660789793295e-05, "loss": 0.0525, "step": 21270 }, { "grad_norm": 0.4084414541721344, "learning_rate": 9.292813276246858e-05, "loss": 0.0583, "step": 21280 }, { "grad_norm": 0.42339009046554565, "learning_rate": 9.291965293244816e-05, "loss": 0.0434, "step": 21290 }, { "grad_norm": 0.37138673663139343, "learning_rate": 9.291116840879904e-05, "loss": 0.0478, "step": 21300 }, { "grad_norm": 0.36934760212898254, "learning_rate": 9.290267919244904e-05, "loss": 0.0501, "step": 21310 }, { "grad_norm": 0.26407361030578613, "learning_rate": 9.289418528432655e-05, "loss": 0.0466, "step": 21320 }, { "grad_norm": 0.39900538325309753, "learning_rate": 9.288568668536045e-05, "loss": 0.0639, "step": 21330 }, { "grad_norm": 0.41296717524528503, "learning_rate": 9.287718339648013e-05, "loss": 0.0462, "step": 21340 }, { "grad_norm": 0.3867717981338501, "learning_rate": 9.28686754186155e-05, "loss": 0.0511, "step": 21350 }, { "grad_norm": 0.532190203666687, "learning_rate": 9.286016275269698e-05, "loss": 0.0558, "step": 21360 }, { "grad_norm": 0.3607850670814514, "learning_rate": 9.285164539965551e-05, "loss": 0.0442, "step": 21370 }, { "grad_norm": 0.4481704533100128, "learning_rate": 9.284312336042251e-05, "loss": 0.0467, "step": 21380 }, { "grad_norm": 0.4642079174518585, "learning_rate": 9.283459663592996e-05, "loss": 0.0504, "step": 21390 }, { "grad_norm": 0.5889134407043457, "learning_rate": 9.282606522711033e-05, "loss": 0.0594, "step": 21400 }, { "grad_norm": 0.3653930425643921, "learning_rate": 9.281752913489657e-05, "loss": 0.0639, "step": 21410 }, { "grad_norm": 0.3816743791103363, "learning_rate": 9.280898836022222e-05, "loss": 0.0503, "step": 21420 }, { "grad_norm": 0.44995060563087463, "learning_rate": 9.280044290402126e-05, "loss": 0.0474, "step": 21430 }, { "grad_norm": 0.32598355412483215, "learning_rate": 9.279189276722821e-05, "loss": 0.0438, "step": 21440 }, { "grad_norm": 0.43017974495887756, "learning_rate": 9.278333795077812e-05, "loss": 0.0567, "step": 21450 }, { "grad_norm": 0.412961483001709, "learning_rate": 9.27747784556065e-05, "loss": 0.0408, "step": 21460 }, { "grad_norm": 0.5020239949226379, "learning_rate": 9.276621428264942e-05, "loss": 0.0617, "step": 21470 }, { "grad_norm": 0.38263455033302307, "learning_rate": 9.275764543284345e-05, "loss": 0.0416, "step": 21480 }, { "grad_norm": 0.5753520727157593, "learning_rate": 9.274907190712566e-05, "loss": 0.0437, "step": 21490 }, { "grad_norm": 0.49149078130722046, "learning_rate": 9.274049370643363e-05, "loss": 0.0478, "step": 21500 }, { "grad_norm": 0.3255866467952728, "learning_rate": 9.273191083170547e-05, "loss": 0.0505, "step": 21510 }, { "grad_norm": 0.7612363696098328, "learning_rate": 9.27233232838798e-05, "loss": 0.0467, "step": 21520 }, { "grad_norm": 0.32903364300727844, "learning_rate": 9.27147310638957e-05, "loss": 0.0444, "step": 21530 }, { "grad_norm": 0.40461763739585876, "learning_rate": 9.270613417269286e-05, "loss": 0.0493, "step": 21540 }, { "grad_norm": 0.4861052930355072, "learning_rate": 9.269753261121138e-05, "loss": 0.0489, "step": 21550 }, { "grad_norm": 0.490447461605072, "learning_rate": 9.268892638039194e-05, "loss": 0.0521, "step": 21560 }, { "grad_norm": 0.33161821961402893, "learning_rate": 9.268031548117569e-05, "loss": 0.0435, "step": 21570 }, { "grad_norm": 0.3296572268009186, "learning_rate": 9.26716999145043e-05, "loss": 0.0455, "step": 21580 }, { "grad_norm": 0.6505478620529175, "learning_rate": 9.266307968131998e-05, "loss": 0.0515, "step": 21590 }, { "grad_norm": 0.5537694692611694, "learning_rate": 9.26544547825654e-05, "loss": 0.0412, "step": 21600 }, { "grad_norm": 0.3429093658924103, "learning_rate": 9.264582521918376e-05, "loss": 0.0498, "step": 21610 }, { "grad_norm": 0.3374931514263153, "learning_rate": 9.263719099211881e-05, "loss": 0.0399, "step": 21620 }, { "grad_norm": 0.5018908381462097, "learning_rate": 9.262855210231476e-05, "loss": 0.0485, "step": 21630 }, { "grad_norm": 0.45422661304473877, "learning_rate": 9.261990855071633e-05, "loss": 0.0446, "step": 21640 }, { "grad_norm": 0.36638909578323364, "learning_rate": 9.261126033826878e-05, "loss": 0.046, "step": 21650 }, { "grad_norm": 0.39230069518089294, "learning_rate": 9.260260746591786e-05, "loss": 0.0452, "step": 21660 }, { "grad_norm": 0.37883615493774414, "learning_rate": 9.259394993460985e-05, "loss": 0.0406, "step": 21670 }, { "grad_norm": 0.32047808170318604, "learning_rate": 9.258528774529151e-05, "loss": 0.0365, "step": 21680 }, { "grad_norm": 0.4513372480869293, "learning_rate": 9.257662089891013e-05, "loss": 0.0389, "step": 21690 }, { "grad_norm": 0.4671926200389862, "learning_rate": 9.25679493964135e-05, "loss": 0.0446, "step": 21700 }, { "grad_norm": 0.4586036801338196, "learning_rate": 9.255927323874994e-05, "loss": 0.0383, "step": 21710 }, { "grad_norm": 0.5218340754508972, "learning_rate": 9.255059242686822e-05, "loss": 0.0498, "step": 21720 }, { "grad_norm": 0.4629426598548889, "learning_rate": 9.254190696171769e-05, "loss": 0.0355, "step": 21730 }, { "grad_norm": 0.4239157438278198, "learning_rate": 9.25332168442482e-05, "loss": 0.0432, "step": 21740 }, { "grad_norm": 0.515009343624115, "learning_rate": 9.252452207541004e-05, "loss": 0.0421, "step": 21750 }, { "grad_norm": 0.386435329914093, "learning_rate": 9.251582265615409e-05, "loss": 0.0537, "step": 21760 }, { "grad_norm": 0.5101167559623718, "learning_rate": 9.250711858743169e-05, "loss": 0.0499, "step": 21770 }, { "grad_norm": 0.35100823640823364, "learning_rate": 9.24984098701947e-05, "loss": 0.0549, "step": 21780 }, { "grad_norm": 0.3103024661540985, "learning_rate": 9.248969650539552e-05, "loss": 0.0431, "step": 21790 }, { "grad_norm": 0.5555477738380432, "learning_rate": 9.2480978493987e-05, "loss": 0.0433, "step": 21800 }, { "grad_norm": 0.3846818804740906, "learning_rate": 9.247225583692256e-05, "loss": 0.0378, "step": 21810 }, { "grad_norm": 0.3749888241291046, "learning_rate": 9.246352853515607e-05, "loss": 0.0463, "step": 21820 }, { "grad_norm": 0.45641374588012695, "learning_rate": 9.245479658964194e-05, "loss": 0.0446, "step": 21830 }, { "grad_norm": 0.44076964259147644, "learning_rate": 9.244606000133507e-05, "loss": 0.0486, "step": 21840 }, { "grad_norm": 0.7676374316215515, "learning_rate": 9.24373187711909e-05, "loss": 0.0539, "step": 21850 }, { "grad_norm": 0.4157990515232086, "learning_rate": 9.242857290016537e-05, "loss": 0.0543, "step": 21860 }, { "grad_norm": 0.6265017986297607, "learning_rate": 9.241982238921488e-05, "loss": 0.0495, "step": 21870 }, { "grad_norm": 0.3929927349090576, "learning_rate": 9.24110672392964e-05, "loss": 0.0394, "step": 21880 }, { "grad_norm": 0.3458910286426544, "learning_rate": 9.240230745136737e-05, "loss": 0.0411, "step": 21890 }, { "grad_norm": 0.497090607881546, "learning_rate": 9.239354302638575e-05, "loss": 0.0388, "step": 21900 }, { "grad_norm": 0.6300206184387207, "learning_rate": 9.238477396531e-05, "loss": 0.0499, "step": 21910 }, { "grad_norm": 0.40492889285087585, "learning_rate": 9.23760002690991e-05, "loss": 0.0632, "step": 21920 }, { "grad_norm": 0.4694874882698059, "learning_rate": 9.236722193871252e-05, "loss": 0.0435, "step": 21930 }, { "grad_norm": 0.4405157268047333, "learning_rate": 9.235843897511023e-05, "loss": 0.0452, "step": 21940 }, { "grad_norm": 0.42864376306533813, "learning_rate": 9.234965137925276e-05, "loss": 0.0408, "step": 21950 }, { "grad_norm": 0.4874095916748047, "learning_rate": 9.234085915210108e-05, "loss": 0.0437, "step": 21960 }, { "grad_norm": 0.4407656192779541, "learning_rate": 9.23320622946167e-05, "loss": 0.0348, "step": 21970 }, { "grad_norm": 0.38973551988601685, "learning_rate": 9.232326080776163e-05, "loss": 0.0433, "step": 21980 }, { "grad_norm": 0.37425893545150757, "learning_rate": 9.23144546924984e-05, "loss": 0.0427, "step": 21990 }, { "grad_norm": 0.3708851933479309, "learning_rate": 9.230564394979e-05, "loss": 0.0382, "step": 22000 }, { "grad_norm": 0.5081049203872681, "learning_rate": 9.22968285806e-05, "loss": 0.043, "step": 22010 }, { "grad_norm": 0.38288483023643494, "learning_rate": 9.228800858589242e-05, "loss": 0.0475, "step": 22020 }, { "grad_norm": 0.4378639757633209, "learning_rate": 9.227918396663179e-05, "loss": 0.0488, "step": 22030 }, { "grad_norm": 0.4489712119102478, "learning_rate": 9.227035472378319e-05, "loss": 0.0427, "step": 22040 }, { "grad_norm": 0.6546542048454285, "learning_rate": 9.226152085831213e-05, "loss": 0.039, "step": 22050 }, { "grad_norm": 0.39856836199760437, "learning_rate": 9.22526823711847e-05, "loss": 0.0392, "step": 22060 }, { "grad_norm": 0.5764376521110535, "learning_rate": 9.224383926336745e-05, "loss": 0.0398, "step": 22070 }, { "grad_norm": 0.3587905764579773, "learning_rate": 9.223499153582744e-05, "loss": 0.0423, "step": 22080 }, { "grad_norm": 0.5309158563613892, "learning_rate": 9.222613918953226e-05, "loss": 0.0584, "step": 22090 }, { "grad_norm": 0.4105774164199829, "learning_rate": 9.221728222544999e-05, "loss": 0.0477, "step": 22100 }, { "grad_norm": 0.4604174792766571, "learning_rate": 9.22084206445492e-05, "loss": 0.0507, "step": 22110 }, { "grad_norm": 0.38930293917655945, "learning_rate": 9.2199554447799e-05, "loss": 0.0542, "step": 22120 }, { "grad_norm": 0.3860056698322296, "learning_rate": 9.219068363616897e-05, "loss": 0.0426, "step": 22130 }, { "grad_norm": 0.3366868197917938, "learning_rate": 9.218180821062919e-05, "loss": 0.0444, "step": 22140 }, { "grad_norm": 0.4391324520111084, "learning_rate": 9.21729281721503e-05, "loss": 0.0504, "step": 22150 }, { "grad_norm": 0.4074971675872803, "learning_rate": 9.216404352170339e-05, "loss": 0.0394, "step": 22160 }, { "grad_norm": 0.29367610812187195, "learning_rate": 9.215515426026007e-05, "loss": 0.0404, "step": 22170 }, { "grad_norm": 0.4172815680503845, "learning_rate": 9.214626038879246e-05, "loss": 0.0383, "step": 22180 }, { "grad_norm": 0.319800466299057, "learning_rate": 9.21373619082732e-05, "loss": 0.0388, "step": 22190 }, { "grad_norm": 0.6536027789115906, "learning_rate": 9.212845881967535e-05, "loss": 0.0457, "step": 22200 }, { "grad_norm": 0.46295928955078125, "learning_rate": 9.211955112397262e-05, "loss": 0.0394, "step": 22210 }, { "grad_norm": 0.5294592976570129, "learning_rate": 9.211063882213909e-05, "loss": 0.0418, "step": 22220 }, { "grad_norm": 0.3522000014781952, "learning_rate": 9.210172191514942e-05, "loss": 0.0409, "step": 22230 }, { "grad_norm": 0.4959084987640381, "learning_rate": 9.209280040397874e-05, "loss": 0.0511, "step": 22240 }, { "grad_norm": 0.404263973236084, "learning_rate": 9.208387428960268e-05, "loss": 0.033, "step": 22250 }, { "grad_norm": 0.39226609468460083, "learning_rate": 9.20749435729974e-05, "loss": 0.0455, "step": 22260 }, { "grad_norm": 0.356109619140625, "learning_rate": 9.206600825513957e-05, "loss": 0.0364, "step": 22270 }, { "grad_norm": 0.39779436588287354, "learning_rate": 9.20570683370063e-05, "loss": 0.0435, "step": 22280 }, { "grad_norm": 0.3236343562602997, "learning_rate": 9.204812381957528e-05, "loss": 0.0375, "step": 22290 }, { "grad_norm": 0.5709640383720398, "learning_rate": 9.203917470382465e-05, "loss": 0.0433, "step": 22300 }, { "grad_norm": 0.35826948285102844, "learning_rate": 9.203022099073309e-05, "loss": 0.0426, "step": 22310 }, { "grad_norm": 0.45158159732818604, "learning_rate": 9.202126268127976e-05, "loss": 0.0474, "step": 22320 }, { "grad_norm": 0.32518479228019714, "learning_rate": 9.20122997764443e-05, "loss": 0.0447, "step": 22330 }, { "grad_norm": 0.52208012342453, "learning_rate": 9.200333227720692e-05, "loss": 0.0524, "step": 22340 }, { "grad_norm": 0.7104499936103821, "learning_rate": 9.199436018454826e-05, "loss": 0.0493, "step": 22350 }, { "grad_norm": 0.3424394726753235, "learning_rate": 9.198538349944952e-05, "loss": 0.0424, "step": 22360 }, { "grad_norm": 0.48281630873680115, "learning_rate": 9.197640222289234e-05, "loss": 0.0616, "step": 22370 }, { "grad_norm": 0.33948200941085815, "learning_rate": 9.196741635585895e-05, "loss": 0.044, "step": 22380 }, { "grad_norm": 0.3385647237300873, "learning_rate": 9.195842589933199e-05, "loss": 0.0376, "step": 22390 }, { "grad_norm": 0.5033378005027771, "learning_rate": 9.194943085429466e-05, "loss": 0.0545, "step": 22400 }, { "grad_norm": 0.4418729841709137, "learning_rate": 9.194043122173065e-05, "loss": 0.0557, "step": 22410 }, { "grad_norm": 0.4771473705768585, "learning_rate": 9.193142700262413e-05, "loss": 0.0462, "step": 22420 }, { "grad_norm": 0.360159695148468, "learning_rate": 9.192241819795979e-05, "loss": 0.041, "step": 22430 }, { "grad_norm": 0.5328552722930908, "learning_rate": 9.191340480872284e-05, "loss": 0.0518, "step": 22440 }, { "grad_norm": 0.40129393339157104, "learning_rate": 9.190438683589895e-05, "loss": 0.0458, "step": 22450 }, { "grad_norm": 0.33536913990974426, "learning_rate": 9.189536428047432e-05, "loss": 0.0529, "step": 22460 }, { "grad_norm": 0.42550358176231384, "learning_rate": 9.188633714343564e-05, "loss": 0.0535, "step": 22470 }, { "grad_norm": 0.506918728351593, "learning_rate": 9.18773054257701e-05, "loss": 0.0448, "step": 22480 }, { "grad_norm": 0.5088281035423279, "learning_rate": 9.18682691284654e-05, "loss": 0.0511, "step": 22490 }, { "grad_norm": 0.3649119734764099, "learning_rate": 9.185922825250974e-05, "loss": 0.0431, "step": 22500 }, { "grad_norm": 0.4179300367832184, "learning_rate": 9.185018279889181e-05, "loss": 0.0379, "step": 22510 }, { "grad_norm": 0.3395300507545471, "learning_rate": 9.184113276860082e-05, "loss": 0.0486, "step": 22520 }, { "grad_norm": 0.5445716977119446, "learning_rate": 9.183207816262645e-05, "loss": 0.0548, "step": 22530 }, { "grad_norm": 0.36440205574035645, "learning_rate": 9.182301898195891e-05, "loss": 0.0403, "step": 22540 }, { "grad_norm": 0.27407631278038025, "learning_rate": 9.181395522758889e-05, "loss": 0.0544, "step": 22550 }, { "grad_norm": 0.9012407660484314, "learning_rate": 9.180488690050759e-05, "loss": 0.0612, "step": 22560 }, { "grad_norm": 0.4409257471561432, "learning_rate": 9.179581400170671e-05, "loss": 0.0485, "step": 22570 }, { "grad_norm": 0.39449259638786316, "learning_rate": 9.178673653217845e-05, "loss": 0.0459, "step": 22580 }, { "grad_norm": 0.416026771068573, "learning_rate": 9.177765449291551e-05, "loss": 0.039, "step": 22590 }, { "grad_norm": 0.3230179250240326, "learning_rate": 9.176856788491109e-05, "loss": 0.0416, "step": 22600 }, { "grad_norm": 0.37470555305480957, "learning_rate": 9.175947670915887e-05, "loss": 0.0449, "step": 22610 }, { "grad_norm": 0.5254610180854797, "learning_rate": 9.175038096665309e-05, "loss": 0.0433, "step": 22620 }, { "grad_norm": 0.9171145558357239, "learning_rate": 9.17412806583884e-05, "loss": 0.0399, "step": 22630 }, { "grad_norm": 0.5549944043159485, "learning_rate": 9.173217578536002e-05, "loss": 0.0594, "step": 22640 }, { "grad_norm": 0.49616163969039917, "learning_rate": 9.172306634856362e-05, "loss": 0.0523, "step": 22650 }, { "grad_norm": 0.7610283493995667, "learning_rate": 9.171395234899545e-05, "loss": 0.0489, "step": 22660 }, { "grad_norm": 0.4047339856624603, "learning_rate": 9.170483378765214e-05, "loss": 0.0431, "step": 22670 }, { "grad_norm": 0.4247451424598694, "learning_rate": 9.169571066553091e-05, "loss": 0.0418, "step": 22680 }, { "grad_norm": 0.4025050103664398, "learning_rate": 9.168658298362946e-05, "loss": 0.0431, "step": 22690 }, { "grad_norm": 0.44224026799201965, "learning_rate": 9.167745074294598e-05, "loss": 0.0435, "step": 22700 }, { "grad_norm": 0.5696441531181335, "learning_rate": 9.166831394447913e-05, "loss": 0.053, "step": 22710 }, { "grad_norm": 0.4387897253036499, "learning_rate": 9.165917258922812e-05, "loss": 0.0398, "step": 22720 }, { "grad_norm": 0.4673297107219696, "learning_rate": 9.165002667819262e-05, "loss": 0.0433, "step": 22730 }, { "grad_norm": 0.3933289349079132, "learning_rate": 9.164087621237282e-05, "loss": 0.0456, "step": 22740 }, { "grad_norm": 0.407452791929245, "learning_rate": 9.163172119276942e-05, "loss": 0.0505, "step": 22750 }, { "grad_norm": 0.5164536833763123, "learning_rate": 9.162256162038358e-05, "loss": 0.053, "step": 22760 }, { "grad_norm": 0.38420289754867554, "learning_rate": 9.161339749621698e-05, "loss": 0.0465, "step": 22770 }, { "grad_norm": 0.45141223073005676, "learning_rate": 9.160422882127177e-05, "loss": 0.0419, "step": 22780 }, { "grad_norm": 0.43906331062316895, "learning_rate": 9.159505559655069e-05, "loss": 0.0359, "step": 22790 }, { "grad_norm": 0.9122899174690247, "learning_rate": 9.158587782305684e-05, "loss": 0.0511, "step": 22800 }, { "grad_norm": 0.4220622181892395, "learning_rate": 9.157669550179391e-05, "loss": 0.0424, "step": 22810 }, { "grad_norm": 0.4799509048461914, "learning_rate": 9.156750863376609e-05, "loss": 0.0548, "step": 22820 }, { "grad_norm": 0.43744441866874695, "learning_rate": 9.155831721997801e-05, "loss": 0.0445, "step": 22830 }, { "grad_norm": 0.43314722180366516, "learning_rate": 9.154912126143484e-05, "loss": 0.0475, "step": 22840 }, { "grad_norm": 0.6137633919715881, "learning_rate": 9.153992075914224e-05, "loss": 0.0536, "step": 22850 }, { "grad_norm": 0.3908355236053467, "learning_rate": 9.153071571410635e-05, "loss": 0.0346, "step": 22860 }, { "grad_norm": 0.5449622273445129, "learning_rate": 9.152150612733384e-05, "loss": 0.0382, "step": 22870 }, { "grad_norm": 0.37661275267601013, "learning_rate": 9.151229199983184e-05, "loss": 0.0505, "step": 22880 }, { "grad_norm": 0.4182891249656677, "learning_rate": 9.150307333260802e-05, "loss": 0.0434, "step": 22890 }, { "grad_norm": 0.4465920031070709, "learning_rate": 9.149385012667048e-05, "loss": 0.0601, "step": 22900 }, { "grad_norm": 0.4702576994895935, "learning_rate": 9.148462238302788e-05, "loss": 0.0546, "step": 22910 }, { "grad_norm": 0.5652656555175781, "learning_rate": 9.147539010268936e-05, "loss": 0.0735, "step": 22920 }, { "grad_norm": 0.48908284306526184, "learning_rate": 9.14661532866645e-05, "loss": 0.0477, "step": 22930 }, { "grad_norm": 0.7802523374557495, "learning_rate": 9.145691193596348e-05, "loss": 0.0493, "step": 22940 }, { "grad_norm": 0.435264527797699, "learning_rate": 9.144766605159691e-05, "loss": 0.0616, "step": 22950 }, { "grad_norm": 0.43476924300193787, "learning_rate": 9.14384156345759e-05, "loss": 0.0475, "step": 22960 }, { "grad_norm": 0.4762055575847626, "learning_rate": 9.142916068591204e-05, "loss": 0.0419, "step": 22970 }, { "grad_norm": 0.4532899856567383, "learning_rate": 9.141990120661746e-05, "loss": 0.0415, "step": 22980 }, { "grad_norm": 0.32821255922317505, "learning_rate": 9.141063719770475e-05, "loss": 0.0434, "step": 22990 }, { "grad_norm": 0.4002029299736023, "learning_rate": 9.140136866018704e-05, "loss": 0.0371, "step": 23000 }, { "grad_norm": 0.34202176332473755, "learning_rate": 9.139209559507788e-05, "loss": 0.0383, "step": 23010 }, { "grad_norm": 0.4542708396911621, "learning_rate": 9.13828180033914e-05, "loss": 0.0414, "step": 23020 }, { "grad_norm": 0.35972315073013306, "learning_rate": 9.137353588614212e-05, "loss": 0.0491, "step": 23030 }, { "grad_norm": 0.38797426223754883, "learning_rate": 9.136424924434519e-05, "loss": 0.0388, "step": 23040 }, { "grad_norm": 0.47944003343582153, "learning_rate": 9.135495807901615e-05, "loss": 0.0484, "step": 23050 }, { "grad_norm": 0.3644808530807495, "learning_rate": 9.134566239117108e-05, "loss": 0.037, "step": 23060 }, { "grad_norm": 0.4014102816581726, "learning_rate": 9.13363621818265e-05, "loss": 0.0412, "step": 23070 }, { "grad_norm": 0.4564666450023651, "learning_rate": 9.132705745199953e-05, "loss": 0.0395, "step": 23080 }, { "grad_norm": 0.5713633894920349, "learning_rate": 9.131774820270768e-05, "loss": 0.0377, "step": 23090 }, { "grad_norm": 0.3535519540309906, "learning_rate": 9.130843443496901e-05, "loss": 0.0414, "step": 23100 }, { "grad_norm": 0.309579998254776, "learning_rate": 9.129911614980206e-05, "loss": 0.0526, "step": 23110 }, { "grad_norm": 0.4039177894592285, "learning_rate": 9.128979334822584e-05, "loss": 0.0418, "step": 23120 }, { "grad_norm": 0.30321547389030457, "learning_rate": 9.128046603125992e-05, "loss": 0.0403, "step": 23130 }, { "grad_norm": 0.38695502281188965, "learning_rate": 9.12711341999243e-05, "loss": 0.0368, "step": 23140 }, { "grad_norm": 0.39760535955429077, "learning_rate": 9.12617978552395e-05, "loss": 0.0561, "step": 23150 }, { "grad_norm": 0.4454909563064575, "learning_rate": 9.12524569982265e-05, "loss": 0.0411, "step": 23160 }, { "grad_norm": 0.46020546555519104, "learning_rate": 9.124311162990684e-05, "loss": 0.0446, "step": 23170 }, { "grad_norm": 0.32811713218688965, "learning_rate": 9.12337617513025e-05, "loss": 0.039, "step": 23180 }, { "grad_norm": 0.417432576417923, "learning_rate": 9.122440736343596e-05, "loss": 0.0354, "step": 23190 }, { "grad_norm": 0.3750264644622803, "learning_rate": 9.12150484673302e-05, "loss": 0.0409, "step": 23200 }, { "grad_norm": 0.6080860495567322, "learning_rate": 9.120568506400873e-05, "loss": 0.0329, "step": 23210 }, { "grad_norm": 0.33114367723464966, "learning_rate": 9.119631715449548e-05, "loss": 0.0428, "step": 23220 }, { "grad_norm": 0.3842112421989441, "learning_rate": 9.118694473981493e-05, "loss": 0.042, "step": 23230 }, { "grad_norm": 0.35577312111854553, "learning_rate": 9.117756782099203e-05, "loss": 0.0342, "step": 23240 }, { "grad_norm": 0.37327316403388977, "learning_rate": 9.11681863990522e-05, "loss": 0.0494, "step": 23250 }, { "grad_norm": 0.47342953085899353, "learning_rate": 9.115880047502142e-05, "loss": 0.0461, "step": 23260 }, { "grad_norm": 0.4978363811969757, "learning_rate": 9.114941004992609e-05, "loss": 0.0545, "step": 23270 }, { "grad_norm": 0.49610814452171326, "learning_rate": 9.114001512479317e-05, "loss": 0.051, "step": 23280 }, { "grad_norm": 0.43594834208488464, "learning_rate": 9.113061570065003e-05, "loss": 0.0584, "step": 23290 }, { "grad_norm": 0.4538175165653229, "learning_rate": 9.112121177852459e-05, "loss": 0.05, "step": 23300 }, { "grad_norm": 0.43521392345428467, "learning_rate": 9.111180335944527e-05, "loss": 0.049, "step": 23310 }, { "grad_norm": 0.608026921749115, "learning_rate": 9.110239044444093e-05, "loss": 0.0444, "step": 23320 }, { "grad_norm": 0.3577221632003784, "learning_rate": 9.109297303454099e-05, "loss": 0.0467, "step": 23330 }, { "grad_norm": 0.506111741065979, "learning_rate": 9.108355113077526e-05, "loss": 0.0533, "step": 23340 }, { "grad_norm": 0.4432831406593323, "learning_rate": 9.107412473417419e-05, "loss": 0.0472, "step": 23350 }, { "grad_norm": 0.427778422832489, "learning_rate": 9.106469384576858e-05, "loss": 0.04, "step": 23360 }, { "grad_norm": 0.37331774830818176, "learning_rate": 9.105525846658978e-05, "loss": 0.0523, "step": 23370 }, { "grad_norm": 0.39985185861587524, "learning_rate": 9.104581859766965e-05, "loss": 0.051, "step": 23380 }, { "grad_norm": 0.43381935358047485, "learning_rate": 9.10363742400405e-05, "loss": 0.0538, "step": 23390 }, { "grad_norm": 0.3222205340862274, "learning_rate": 9.102692539473518e-05, "loss": 0.0487, "step": 23400 }, { "grad_norm": 0.6014277338981628, "learning_rate": 9.101747206278697e-05, "loss": 0.0541, "step": 23410 }, { "grad_norm": 0.5174510478973389, "learning_rate": 9.100801424522968e-05, "loss": 0.036, "step": 23420 }, { "grad_norm": 0.8628275990486145, "learning_rate": 9.099855194309762e-05, "loss": 0.0489, "step": 23430 }, { "grad_norm": 0.4181773066520691, "learning_rate": 9.098908515742554e-05, "loss": 0.0416, "step": 23440 }, { "grad_norm": 0.4602159261703491, "learning_rate": 9.097961388924873e-05, "loss": 0.045, "step": 23450 }, { "grad_norm": 0.4217230975627899, "learning_rate": 9.097013813960298e-05, "loss": 0.0421, "step": 23460 }, { "grad_norm": 1.2764745950698853, "learning_rate": 9.09606579095245e-05, "loss": 0.056, "step": 23470 }, { "grad_norm": 0.8208343386650085, "learning_rate": 9.095117320005008e-05, "loss": 0.0422, "step": 23480 }, { "grad_norm": 0.4606929421424866, "learning_rate": 9.094168401221691e-05, "loss": 0.0626, "step": 23490 }, { "grad_norm": 0.4912773668766022, "learning_rate": 9.093219034706273e-05, "loss": 0.0646, "step": 23500 }, { "grad_norm": 0.3590755760669708, "learning_rate": 9.092269220562577e-05, "loss": 0.0551, "step": 23510 }, { "grad_norm": 0.509632408618927, "learning_rate": 9.09131895889447e-05, "loss": 0.0705, "step": 23520 }, { "grad_norm": 0.49996140599250793, "learning_rate": 9.090368249805873e-05, "loss": 0.0552, "step": 23530 }, { "grad_norm": 0.3470512628555298, "learning_rate": 9.089417093400754e-05, "loss": 0.0477, "step": 23540 }, { "grad_norm": 0.5020407438278198, "learning_rate": 9.088465489783131e-05, "loss": 0.0478, "step": 23550 }, { "grad_norm": 0.3627971112728119, "learning_rate": 9.087513439057068e-05, "loss": 0.0439, "step": 23560 }, { "grad_norm": 0.3886120915412903, "learning_rate": 9.08656094132668e-05, "loss": 0.0525, "step": 23570 }, { "grad_norm": 0.5258092880249023, "learning_rate": 9.085607996696134e-05, "loss": 0.049, "step": 23580 }, { "grad_norm": 0.4387300908565521, "learning_rate": 9.084654605269639e-05, "loss": 0.0567, "step": 23590 }, { "grad_norm": 0.5107114911079407, "learning_rate": 9.083700767151457e-05, "loss": 0.0478, "step": 23600 }, { "grad_norm": 0.4374609589576721, "learning_rate": 9.082746482445898e-05, "loss": 0.0556, "step": 23610 }, { "grad_norm": 0.346963495016098, "learning_rate": 9.081791751257325e-05, "loss": 0.0425, "step": 23620 }, { "grad_norm": 0.4939481019973755, "learning_rate": 9.080836573690142e-05, "loss": 0.0406, "step": 23630 }, { "grad_norm": 0.45753851532936096, "learning_rate": 9.079880949848805e-05, "loss": 0.0537, "step": 23640 }, { "grad_norm": 2.7763898372650146, "learning_rate": 9.078924879837822e-05, "loss": 0.0486, "step": 23650 }, { "grad_norm": 0.4337669312953949, "learning_rate": 9.077968363761747e-05, "loss": 0.0527, "step": 23660 }, { "grad_norm": 0.430927038192749, "learning_rate": 9.077011401725182e-05, "loss": 0.0484, "step": 23670 }, { "grad_norm": 0.3934464454650879, "learning_rate": 9.07605399383278e-05, "loss": 0.0556, "step": 23680 }, { "grad_norm": 0.47084447741508484, "learning_rate": 9.075096140189243e-05, "loss": 0.0473, "step": 23690 }, { "grad_norm": 0.4850974678993225, "learning_rate": 9.074137840899318e-05, "loss": 0.0508, "step": 23700 }, { "grad_norm": 0.356893390417099, "learning_rate": 9.073179096067804e-05, "loss": 0.0459, "step": 23710 }, { "grad_norm": 0.2915450632572174, "learning_rate": 9.072219905799549e-05, "loss": 0.0387, "step": 23720 }, { "grad_norm": 0.5259132385253906, "learning_rate": 9.071260270199447e-05, "loss": 0.0554, "step": 23730 }, { "grad_norm": 0.6818866729736328, "learning_rate": 9.070300189372441e-05, "loss": 0.0379, "step": 23740 }, { "grad_norm": 0.3113372027873993, "learning_rate": 9.069339663423528e-05, "loss": 0.0381, "step": 23750 }, { "grad_norm": 0.5383308529853821, "learning_rate": 9.068378692457747e-05, "loss": 0.0572, "step": 23760 }, { "grad_norm": 0.7851988077163696, "learning_rate": 9.067417276580189e-05, "loss": 0.0477, "step": 23770 }, { "grad_norm": 0.5929152369499207, "learning_rate": 9.066455415895993e-05, "loss": 0.0493, "step": 23780 }, { "grad_norm": 0.5190828442573547, "learning_rate": 9.065493110510346e-05, "loss": 0.0427, "step": 23790 }, { "grad_norm": 0.5039796829223633, "learning_rate": 9.064530360528484e-05, "loss": 0.0444, "step": 23800 }, { "grad_norm": 0.30815064907073975, "learning_rate": 9.063567166055695e-05, "loss": 0.0426, "step": 23810 }, { "grad_norm": 0.4895677864551544, "learning_rate": 9.062603527197308e-05, "loss": 0.0428, "step": 23820 }, { "grad_norm": 0.44597581028938293, "learning_rate": 9.06163944405871e-05, "loss": 0.0439, "step": 23830 }, { "grad_norm": 0.3942187428474426, "learning_rate": 9.060674916745327e-05, "loss": 0.0523, "step": 23840 }, { "grad_norm": 0.5139894485473633, "learning_rate": 9.05970994536264e-05, "loss": 0.0402, "step": 23850 }, { "grad_norm": 0.4662807285785675, "learning_rate": 9.05874453001618e-05, "loss": 0.0388, "step": 23860 }, { "grad_norm": 0.5307244658470154, "learning_rate": 9.057778670811517e-05, "loss": 0.0468, "step": 23870 }, { "grad_norm": 0.37157055735588074, "learning_rate": 9.056812367854281e-05, "loss": 0.0375, "step": 23880 }, { "grad_norm": 0.3973899185657501, "learning_rate": 9.055845621250143e-05, "loss": 0.0483, "step": 23890 }, { "grad_norm": 0.40780001878738403, "learning_rate": 9.054878431104825e-05, "loss": 0.0385, "step": 23900 }, { "grad_norm": 0.3989911675453186, "learning_rate": 9.0539107975241e-05, "loss": 0.0449, "step": 23910 }, { "grad_norm": 0.368196576833725, "learning_rate": 9.052942720613784e-05, "loss": 0.0503, "step": 23920 }, { "grad_norm": 0.39519673585891724, "learning_rate": 9.051974200479745e-05, "loss": 0.037, "step": 23930 }, { "grad_norm": 0.5665845274925232, "learning_rate": 9.051005237227901e-05, "loss": 0.035, "step": 23940 }, { "grad_norm": 0.3747088313102722, "learning_rate": 9.050035830964215e-05, "loss": 0.0364, "step": 23950 }, { "grad_norm": 0.6933486461639404, "learning_rate": 9.049065981794698e-05, "loss": 0.0571, "step": 23960 }, { "grad_norm": 0.279793381690979, "learning_rate": 9.048095689825414e-05, "loss": 0.0399, "step": 23970 }, { "grad_norm": 0.4646567404270172, "learning_rate": 9.047124955162472e-05, "loss": 0.04, "step": 23980 }, { "grad_norm": 0.2719447910785675, "learning_rate": 9.046153777912028e-05, "loss": 0.0509, "step": 23990 }, { "grad_norm": 0.333146333694458, "learning_rate": 9.045182158180292e-05, "loss": 0.0429, "step": 24000 }, { "grad_norm": 0.3052952289581299, "learning_rate": 9.044210096073516e-05, "loss": 0.0316, "step": 24010 }, { "grad_norm": 0.4927136301994324, "learning_rate": 9.043237591698004e-05, "loss": 0.0364, "step": 24020 }, { "grad_norm": 0.47250962257385254, "learning_rate": 9.04226464516011e-05, "loss": 0.0397, "step": 24030 }, { "grad_norm": 0.41505151987075806, "learning_rate": 9.041291256566229e-05, "loss": 0.0455, "step": 24040 }, { "grad_norm": 0.3579067289829254, "learning_rate": 9.040317426022814e-05, "loss": 0.0456, "step": 24050 }, { "grad_norm": 0.31472665071487427, "learning_rate": 9.03934315363636e-05, "loss": 0.0427, "step": 24060 }, { "grad_norm": 0.2739294469356537, "learning_rate": 9.038368439513409e-05, "loss": 0.0351, "step": 24070 }, { "grad_norm": 0.5210661888122559, "learning_rate": 9.03739328376056e-05, "loss": 0.0465, "step": 24080 }, { "grad_norm": 0.327785849571228, "learning_rate": 9.036417686484451e-05, "loss": 0.0486, "step": 24090 }, { "grad_norm": 0.5060511231422424, "learning_rate": 9.035441647791773e-05, "loss": 0.0491, "step": 24100 }, { "grad_norm": 0.4247134327888489, "learning_rate": 9.034465167789263e-05, "loss": 0.0422, "step": 24110 }, { "grad_norm": 0.5483392477035522, "learning_rate": 9.033488246583706e-05, "loss": 0.0444, "step": 24120 }, { "grad_norm": 0.3524211347103119, "learning_rate": 9.032510884281941e-05, "loss": 0.0558, "step": 24130 }, { "grad_norm": 0.9861847162246704, "learning_rate": 9.031533080990848e-05, "loss": 0.0522, "step": 24140 }, { "grad_norm": 0.43651485443115234, "learning_rate": 9.030554836817358e-05, "loss": 0.0448, "step": 24150 }, { "grad_norm": 0.6884637475013733, "learning_rate": 9.029576151868451e-05, "loss": 0.0517, "step": 24160 }, { "grad_norm": 0.5053950548171997, "learning_rate": 9.028597026251155e-05, "loss": 0.0379, "step": 24170 }, { "grad_norm": 0.4788154363632202, "learning_rate": 9.027617460072547e-05, "loss": 0.0463, "step": 24180 }, { "grad_norm": 0.40567445755004883, "learning_rate": 9.026637453439745e-05, "loss": 0.0405, "step": 24190 }, { "grad_norm": 0.3672811686992645, "learning_rate": 9.025657006459927e-05, "loss": 0.0497, "step": 24200 }, { "grad_norm": 0.4091418385505676, "learning_rate": 9.024676119240311e-05, "loss": 0.0388, "step": 24210 }, { "grad_norm": 0.39011356234550476, "learning_rate": 9.023694791888166e-05, "loss": 0.0521, "step": 24220 }, { "grad_norm": 0.41964441537857056, "learning_rate": 9.022713024510808e-05, "loss": 0.0525, "step": 24230 }, { "grad_norm": 0.4123610556125641, "learning_rate": 9.021730817215601e-05, "loss": 0.048, "step": 24240 }, { "grad_norm": 0.316476970911026, "learning_rate": 9.02074817010996e-05, "loss": 0.0465, "step": 24250 }, { "grad_norm": 0.38605746626853943, "learning_rate": 9.019765083301342e-05, "loss": 0.0371, "step": 24260 }, { "grad_norm": 0.6130672693252563, "learning_rate": 9.01878155689726e-05, "loss": 0.0441, "step": 24270 }, { "grad_norm": 0.40201693773269653, "learning_rate": 9.017797591005268e-05, "loss": 0.0412, "step": 24280 }, { "grad_norm": 0.403691828250885, "learning_rate": 9.016813185732972e-05, "loss": 0.058, "step": 24290 }, { "grad_norm": 0.4459565579891205, "learning_rate": 9.015828341188027e-05, "loss": 0.0508, "step": 24300 }, { "grad_norm": 0.5472269058227539, "learning_rate": 9.01484305747813e-05, "loss": 0.0504, "step": 24310 }, { "grad_norm": 0.5983352065086365, "learning_rate": 9.013857334711033e-05, "loss": 0.0488, "step": 24320 }, { "grad_norm": 0.32008644938468933, "learning_rate": 9.012871172994534e-05, "loss": 0.0383, "step": 24330 }, { "grad_norm": 0.3902328908443451, "learning_rate": 9.011884572436476e-05, "loss": 0.0542, "step": 24340 }, { "grad_norm": 0.45982736349105835, "learning_rate": 9.010897533144754e-05, "loss": 0.0629, "step": 24350 }, { "grad_norm": 0.35087740421295166, "learning_rate": 9.009910055227306e-05, "loss": 0.0439, "step": 24360 }, { "grad_norm": 0.3578193187713623, "learning_rate": 9.008922138792124e-05, "loss": 0.0328, "step": 24370 }, { "grad_norm": 0.736243724822998, "learning_rate": 9.007933783947244e-05, "loss": 0.0406, "step": 24380 }, { "grad_norm": 0.36829984188079834, "learning_rate": 9.006944990800752e-05, "loss": 0.0366, "step": 24390 }, { "grad_norm": 0.42736226320266724, "learning_rate": 9.005955759460779e-05, "loss": 0.0372, "step": 24400 }, { "grad_norm": 0.4742386043071747, "learning_rate": 9.004966090035508e-05, "loss": 0.0382, "step": 24410 }, { "grad_norm": 0.2637013792991638, "learning_rate": 9.003975982633166e-05, "loss": 0.0443, "step": 24420 }, { "grad_norm": 0.3076235055923462, "learning_rate": 9.00298543736203e-05, "loss": 0.0464, "step": 24430 }, { "grad_norm": 0.2572341561317444, "learning_rate": 9.001994454330427e-05, "loss": 0.0357, "step": 24440 }, { "grad_norm": 0.5042076706886292, "learning_rate": 9.001003033646727e-05, "loss": 0.0462, "step": 24450 }, { "grad_norm": 0.5206143260002136, "learning_rate": 9.00001117541935e-05, "loss": 0.0458, "step": 24460 }, { "grad_norm": 0.4067910313606262, "learning_rate": 8.999018879756764e-05, "loss": 0.043, "step": 24470 }, { "grad_norm": 0.4408780336380005, "learning_rate": 8.998026146767487e-05, "loss": 0.0419, "step": 24480 }, { "grad_norm": 0.4787144958972931, "learning_rate": 8.99703297656008e-05, "loss": 0.0417, "step": 24490 }, { "grad_norm": 0.3943018913269043, "learning_rate": 8.996039369243156e-05, "loss": 0.0455, "step": 24500 }, { "grad_norm": 0.41167545318603516, "learning_rate": 8.995045324925378e-05, "loss": 0.0381, "step": 24510 }, { "grad_norm": 0.4360750913619995, "learning_rate": 8.994050843715448e-05, "loss": 0.0485, "step": 24520 }, { "grad_norm": 0.36751917004585266, "learning_rate": 8.993055925722121e-05, "loss": 0.0373, "step": 24530 }, { "grad_norm": 0.6513774991035461, "learning_rate": 8.992060571054202e-05, "loss": 0.0348, "step": 24540 }, { "grad_norm": 0.37279319763183594, "learning_rate": 8.991064779820542e-05, "loss": 0.0483, "step": 24550 }, { "grad_norm": 0.615170955657959, "learning_rate": 8.990068552130036e-05, "loss": 0.0369, "step": 24560 }, { "grad_norm": 0.4089623987674713, "learning_rate": 8.989071888091634e-05, "loss": 0.0455, "step": 24570 }, { "grad_norm": 0.4782070815563202, "learning_rate": 8.988074787814329e-05, "loss": 0.0423, "step": 24580 }, { "grad_norm": 0.3160964250564575, "learning_rate": 8.987077251407158e-05, "loss": 0.0387, "step": 24590 }, { "grad_norm": 0.38302212953567505, "learning_rate": 8.986079278979216e-05, "loss": 0.0375, "step": 24600 }, { "grad_norm": 0.3827918767929077, "learning_rate": 8.985080870639635e-05, "loss": 0.0438, "step": 24610 }, { "grad_norm": 0.34149008989334106, "learning_rate": 8.984082026497603e-05, "loss": 0.0339, "step": 24620 }, { "grad_norm": 0.39612486958503723, "learning_rate": 8.98308274666235e-05, "loss": 0.0523, "step": 24630 }, { "grad_norm": 0.402849018573761, "learning_rate": 8.982083031243155e-05, "loss": 0.0443, "step": 24640 }, { "grad_norm": 0.4077015519142151, "learning_rate": 8.98108288034935e-05, "loss": 0.0438, "step": 24650 }, { "grad_norm": 0.5412181615829468, "learning_rate": 8.980082294090305e-05, "loss": 0.0428, "step": 24660 }, { "grad_norm": 0.2878091633319855, "learning_rate": 8.979081272575443e-05, "loss": 0.0474, "step": 24670 }, { "grad_norm": 0.4812574088573456, "learning_rate": 8.978079815914236e-05, "loss": 0.0468, "step": 24680 }, { "grad_norm": 0.4352359473705292, "learning_rate": 8.977077924216202e-05, "loss": 0.0473, "step": 24690 }, { "grad_norm": 0.43821102380752563, "learning_rate": 8.976075597590905e-05, "loss": 0.0473, "step": 24700 }, { "grad_norm": 0.46641039848327637, "learning_rate": 8.975072836147958e-05, "loss": 0.0446, "step": 24710 }, { "grad_norm": 0.4825996458530426, "learning_rate": 8.974069639997025e-05, "loss": 0.0387, "step": 24720 }, { "grad_norm": 0.4250146448612213, "learning_rate": 8.973066009247808e-05, "loss": 0.0413, "step": 24730 }, { "grad_norm": 0.3536868095397949, "learning_rate": 8.972061944010066e-05, "loss": 0.0431, "step": 24740 }, { "grad_norm": 0.3427633047103882, "learning_rate": 8.971057444393603e-05, "loss": 0.0438, "step": 24750 }, { "grad_norm": 0.4046177268028259, "learning_rate": 8.970052510508268e-05, "loss": 0.0552, "step": 24760 }, { "grad_norm": 0.49528637528419495, "learning_rate": 8.969047142463959e-05, "loss": 0.048, "step": 24770 }, { "grad_norm": 0.4315946102142334, "learning_rate": 8.968041340370621e-05, "loss": 0.0429, "step": 24780 }, { "grad_norm": 0.4402289092540741, "learning_rate": 8.96703510433825e-05, "loss": 0.0416, "step": 24790 }, { "grad_norm": 0.3279387056827545, "learning_rate": 8.966028434476883e-05, "loss": 0.0359, "step": 24800 }, { "grad_norm": 0.7138353586196899, "learning_rate": 8.96502133089661e-05, "loss": 0.0366, "step": 24810 }, { "grad_norm": 0.3314097225666046, "learning_rate": 8.964013793707564e-05, "loss": 0.0365, "step": 24820 }, { "grad_norm": 0.3104144334793091, "learning_rate": 8.963005823019932e-05, "loss": 0.0409, "step": 24830 }, { "grad_norm": 0.42601680755615234, "learning_rate": 8.961997418943939e-05, "loss": 0.0411, "step": 24840 }, { "grad_norm": 0.3088757395744324, "learning_rate": 8.960988581589865e-05, "loss": 0.0353, "step": 24850 }, { "grad_norm": 0.378937304019928, "learning_rate": 8.959979311068037e-05, "loss": 0.0407, "step": 24860 }, { "grad_norm": 0.5012482404708862, "learning_rate": 8.958969607488823e-05, "loss": 0.059, "step": 24870 }, { "grad_norm": 0.3696201741695404, "learning_rate": 8.957959470962647e-05, "loss": 0.046, "step": 24880 }, { "grad_norm": 0.36228522658348083, "learning_rate": 8.956948901599971e-05, "loss": 0.0417, "step": 24890 }, { "grad_norm": 0.40637925267219543, "learning_rate": 8.955937899511315e-05, "loss": 0.0433, "step": 24900 }, { "grad_norm": 0.45398372411727905, "learning_rate": 8.954926464807238e-05, "loss": 0.0418, "step": 24910 }, { "grad_norm": 0.32533517479896545, "learning_rate": 8.953914597598347e-05, "loss": 0.0383, "step": 24920 }, { "grad_norm": 0.31427001953125, "learning_rate": 8.952902297995303e-05, "loss": 0.0488, "step": 24930 }, { "grad_norm": 0.5205745697021484, "learning_rate": 8.951889566108804e-05, "loss": 0.0427, "step": 24940 }, { "grad_norm": 0.34463170170783997, "learning_rate": 8.950876402049606e-05, "loss": 0.0411, "step": 24950 }, { "grad_norm": 0.4566899240016937, "learning_rate": 8.949862805928504e-05, "loss": 0.0487, "step": 24960 }, { "grad_norm": 0.5035616159439087, "learning_rate": 8.948848777856343e-05, "loss": 0.0391, "step": 24970 }, { "grad_norm": 0.4639624357223511, "learning_rate": 8.947834317944017e-05, "loss": 0.0417, "step": 24980 }, { "grad_norm": 0.42907771468162537, "learning_rate": 8.946819426302466e-05, "loss": 0.059, "step": 24990 }, { "grad_norm": 0.44582459330558777, "learning_rate": 8.945804103042676e-05, "loss": 0.0446, "step": 25000 }, { "grad_norm": 0.33903905749320984, "learning_rate": 8.944788348275681e-05, "loss": 0.0391, "step": 25010 }, { "grad_norm": 0.4487892985343933, "learning_rate": 8.943772162112565e-05, "loss": 0.0384, "step": 25020 }, { "grad_norm": 0.5916797518730164, "learning_rate": 8.942755544664454e-05, "loss": 0.0552, "step": 25030 }, { "grad_norm": 0.39399006962776184, "learning_rate": 8.941738496042525e-05, "loss": 0.0415, "step": 25040 }, { "grad_norm": 0.2892642915248871, "learning_rate": 8.940721016357999e-05, "loss": 0.0487, "step": 25050 }, { "grad_norm": 0.4205668568611145, "learning_rate": 8.939703105722148e-05, "loss": 0.0464, "step": 25060 }, { "grad_norm": 0.44630247354507446, "learning_rate": 8.93868476424629e-05, "loss": 0.0428, "step": 25070 }, { "grad_norm": 0.2696830630302429, "learning_rate": 8.937665992041786e-05, "loss": 0.0488, "step": 25080 }, { "grad_norm": 0.4631498157978058, "learning_rate": 8.93664678922005e-05, "loss": 0.0484, "step": 25090 }, { "grad_norm": 0.3922191560268402, "learning_rate": 8.93562715589254e-05, "loss": 0.0382, "step": 25100 }, { "grad_norm": 0.4329691231250763, "learning_rate": 8.934607092170762e-05, "loss": 0.0504, "step": 25110 }, { "grad_norm": 0.4544279873371124, "learning_rate": 8.933586598166266e-05, "loss": 0.0426, "step": 25120 }, { "grad_norm": 0.446625292301178, "learning_rate": 8.932565673990655e-05, "loss": 0.0547, "step": 25130 }, { "grad_norm": 0.3245214521884918, "learning_rate": 8.931544319755574e-05, "loss": 0.046, "step": 25140 }, { "grad_norm": 0.4154374897480011, "learning_rate": 8.930522535572718e-05, "loss": 0.0361, "step": 25150 }, { "grad_norm": 0.3931979537010193, "learning_rate": 8.929500321553826e-05, "loss": 0.0411, "step": 25160 }, { "grad_norm": 0.5473501086235046, "learning_rate": 8.928477677810686e-05, "loss": 0.0372, "step": 25170 }, { "grad_norm": 0.35957661271095276, "learning_rate": 8.927454604455137e-05, "loss": 0.0406, "step": 25180 }, { "grad_norm": 0.24614191055297852, "learning_rate": 8.926431101599053e-05, "loss": 0.0401, "step": 25190 }, { "grad_norm": 0.40698933601379395, "learning_rate": 8.925407169354369e-05, "loss": 0.0388, "step": 25200 }, { "grad_norm": 0.45928511023521423, "learning_rate": 8.92438280783306e-05, "loss": 0.0486, "step": 25210 }, { "grad_norm": 0.39805859327316284, "learning_rate": 8.923358017147146e-05, "loss": 0.0355, "step": 25220 }, { "grad_norm": 0.31197285652160645, "learning_rate": 8.922332797408697e-05, "loss": 0.0461, "step": 25230 }, { "grad_norm": 0.43880507349967957, "learning_rate": 8.921307148729831e-05, "loss": 0.038, "step": 25240 }, { "grad_norm": 0.43513917922973633, "learning_rate": 8.920281071222712e-05, "loss": 0.0464, "step": 25250 }, { "grad_norm": 0.42142102122306824, "learning_rate": 8.919254564999548e-05, "loss": 0.0441, "step": 25260 }, { "grad_norm": 0.6962120532989502, "learning_rate": 8.918227630172598e-05, "loss": 0.0456, "step": 25270 }, { "grad_norm": 0.5566393733024597, "learning_rate": 8.917200266854165e-05, "loss": 0.0397, "step": 25280 }, { "grad_norm": 0.5374945402145386, "learning_rate": 8.9161724751566e-05, "loss": 0.0463, "step": 25290 }, { "grad_norm": 0.34714075922966003, "learning_rate": 8.915144255192302e-05, "loss": 0.0398, "step": 25300 }, { "grad_norm": 0.4697699546813965, "learning_rate": 8.914115607073714e-05, "loss": 0.0396, "step": 25310 }, { "grad_norm": 0.570242702960968, "learning_rate": 8.913086530913327e-05, "loss": 0.04, "step": 25320 }, { "grad_norm": 0.3084549009799957, "learning_rate": 8.912057026823681e-05, "loss": 0.0482, "step": 25330 }, { "grad_norm": 0.6205167770385742, "learning_rate": 8.91102709491736e-05, "loss": 0.043, "step": 25340 }, { "grad_norm": 0.4992142617702484, "learning_rate": 8.909996735306996e-05, "loss": 0.037, "step": 25350 }, { "grad_norm": 0.5601912140846252, "learning_rate": 8.908965948105268e-05, "loss": 0.0531, "step": 25360 }, { "grad_norm": 0.5300737619400024, "learning_rate": 8.907934733424901e-05, "loss": 0.0465, "step": 25370 }, { "grad_norm": 0.3483542203903198, "learning_rate": 8.906903091378666e-05, "loss": 0.0431, "step": 25380 }, { "grad_norm": 0.3151978552341461, "learning_rate": 8.905871022079384e-05, "loss": 0.0385, "step": 25390 }, { "grad_norm": 0.40778499841690063, "learning_rate": 8.90483852563992e-05, "loss": 0.0358, "step": 25400 }, { "grad_norm": 0.34837743639945984, "learning_rate": 8.903805602173185e-05, "loss": 0.0417, "step": 25410 }, { "grad_norm": 0.42125892639160156, "learning_rate": 8.902772251792137e-05, "loss": 0.0312, "step": 25420 }, { "grad_norm": 0.4696665108203888, "learning_rate": 8.901738474609786e-05, "loss": 0.0492, "step": 25430 }, { "grad_norm": 0.4566618502140045, "learning_rate": 8.900704270739179e-05, "loss": 0.0316, "step": 25440 }, { "grad_norm": 0.4659746587276459, "learning_rate": 8.89966964029342e-05, "loss": 0.0526, "step": 25450 }, { "grad_norm": 0.47932156920433044, "learning_rate": 8.898634583385652e-05, "loss": 0.0408, "step": 25460 }, { "grad_norm": 0.3869185447692871, "learning_rate": 8.897599100129065e-05, "loss": 0.0455, "step": 25470 }, { "grad_norm": 0.5614513754844666, "learning_rate": 8.896563190636903e-05, "loss": 0.0388, "step": 25480 }, { "grad_norm": 0.38312581181526184, "learning_rate": 8.895526855022448e-05, "loss": 0.0432, "step": 25490 }, { "grad_norm": 0.48773568868637085, "learning_rate": 8.894490093399033e-05, "loss": 0.0461, "step": 25500 }, { "grad_norm": 0.38897454738616943, "learning_rate": 8.893452905880035e-05, "loss": 0.0349, "step": 25510 }, { "grad_norm": 0.64558345079422, "learning_rate": 8.892415292578883e-05, "loss": 0.0396, "step": 25520 }, { "grad_norm": 0.449135422706604, "learning_rate": 8.891377253609046e-05, "loss": 0.044, "step": 25530 }, { "grad_norm": 0.5013173222541809, "learning_rate": 8.890338789084043e-05, "loss": 0.0459, "step": 25540 }, { "grad_norm": 0.3840537369251251, "learning_rate": 8.88929989911744e-05, "loss": 0.0414, "step": 25550 }, { "grad_norm": 0.500403881072998, "learning_rate": 8.888260583822847e-05, "loss": 0.042, "step": 25560 }, { "grad_norm": 0.46830037236213684, "learning_rate": 8.887220843313921e-05, "loss": 0.0394, "step": 25570 }, { "grad_norm": 0.30470260977745056, "learning_rate": 8.88618067770437e-05, "loss": 0.0464, "step": 25580 }, { "grad_norm": 0.3682533800601959, "learning_rate": 8.885140087107942e-05, "loss": 0.0447, "step": 25590 }, { "grad_norm": 0.4132614731788635, "learning_rate": 8.884099071638436e-05, "loss": 0.0444, "step": 25600 }, { "grad_norm": 0.33086830377578735, "learning_rate": 8.883057631409695e-05, "loss": 0.0464, "step": 25610 }, { "grad_norm": 0.42462852597236633, "learning_rate": 8.882015766535608e-05, "loss": 0.0499, "step": 25620 }, { "grad_norm": 0.43838804960250854, "learning_rate": 8.880973477130115e-05, "loss": 0.0442, "step": 25630 }, { "grad_norm": 0.44339510798454285, "learning_rate": 8.879930763307197e-05, "loss": 0.0353, "step": 25640 }, { "grad_norm": 0.4439593255519867, "learning_rate": 8.878887625180884e-05, "loss": 0.046, "step": 25650 }, { "grad_norm": 0.3533964157104492, "learning_rate": 8.877844062865253e-05, "loss": 0.0325, "step": 25660 }, { "grad_norm": 0.4762118458747864, "learning_rate": 8.876800076474424e-05, "loss": 0.0494, "step": 25670 }, { "grad_norm": 0.4363310635089874, "learning_rate": 8.875755666122568e-05, "loss": 0.0386, "step": 25680 }, { "grad_norm": 0.3544519245624542, "learning_rate": 8.8747108319239e-05, "loss": 0.0469, "step": 25690 }, { "grad_norm": 0.4736771881580353, "learning_rate": 8.87366557399268e-05, "loss": 0.0393, "step": 25700 }, { "grad_norm": 0.40683984756469727, "learning_rate": 8.872619892443217e-05, "loss": 0.0403, "step": 25710 }, { "grad_norm": 0.4237653315067291, "learning_rate": 8.871573787389865e-05, "loss": 0.0382, "step": 25720 }, { "grad_norm": 0.4159677028656006, "learning_rate": 8.870527258947024e-05, "loss": 0.0455, "step": 25730 }, { "grad_norm": 0.5593675374984741, "learning_rate": 8.869480307229143e-05, "loss": 0.0435, "step": 25740 }, { "grad_norm": 0.49532803893089294, "learning_rate": 8.868432932350712e-05, "loss": 0.0425, "step": 25750 }, { "grad_norm": 0.4389176666736603, "learning_rate": 8.867385134426272e-05, "loss": 0.044, "step": 25760 }, { "grad_norm": 0.5431730151176453, "learning_rate": 8.866336913570407e-05, "loss": 0.0465, "step": 25770 }, { "grad_norm": 0.5705121159553528, "learning_rate": 8.865288269897751e-05, "loss": 0.036, "step": 25780 }, { "grad_norm": 0.5567966103553772, "learning_rate": 8.864239203522981e-05, "loss": 0.0444, "step": 25790 }, { "grad_norm": 1.0009793043136597, "learning_rate": 8.863189714560822e-05, "loss": 0.0425, "step": 25800 }, { "grad_norm": 0.4835280478000641, "learning_rate": 8.862139803126043e-05, "loss": 0.0454, "step": 25810 }, { "grad_norm": 0.6027415990829468, "learning_rate": 8.861089469333463e-05, "loss": 0.0416, "step": 25820 }, { "grad_norm": 0.42141494154930115, "learning_rate": 8.860038713297944e-05, "loss": 0.0478, "step": 25830 }, { "grad_norm": 0.3139091730117798, "learning_rate": 8.858987535134394e-05, "loss": 0.0434, "step": 25840 }, { "grad_norm": 0.349450945854187, "learning_rate": 8.857935934957769e-05, "loss": 0.0478, "step": 25850 }, { "grad_norm": 0.2553466856479645, "learning_rate": 8.856883912883071e-05, "loss": 0.0386, "step": 25860 }, { "grad_norm": 0.46584346890449524, "learning_rate": 8.855831469025346e-05, "loss": 0.0436, "step": 25870 }, { "grad_norm": 0.4516998529434204, "learning_rate": 8.854778603499689e-05, "loss": 0.0499, "step": 25880 }, { "grad_norm": 0.24761101603507996, "learning_rate": 8.85372531642124e-05, "loss": 0.0404, "step": 25890 }, { "grad_norm": 0.3381938636302948, "learning_rate": 8.852671607905185e-05, "loss": 0.0342, "step": 25900 }, { "grad_norm": 0.3823157548904419, "learning_rate": 8.851617478066754e-05, "loss": 0.051, "step": 25910 }, { "grad_norm": 0.45609748363494873, "learning_rate": 8.850562927021227e-05, "loss": 0.037, "step": 25920 }, { "grad_norm": 0.37985798716545105, "learning_rate": 8.849507954883928e-05, "loss": 0.0354, "step": 25930 }, { "grad_norm": 0.34904271364212036, "learning_rate": 8.848452561770226e-05, "loss": 0.0367, "step": 25940 }, { "grad_norm": 0.4564991593360901, "learning_rate": 8.847396747795538e-05, "loss": 0.0454, "step": 25950 }, { "grad_norm": 0.4630928039550781, "learning_rate": 8.846340513075327e-05, "loss": 0.0397, "step": 25960 }, { "grad_norm": 0.5612316131591797, "learning_rate": 8.845283857725099e-05, "loss": 0.036, "step": 25970 }, { "grad_norm": 0.57294762134552, "learning_rate": 8.844226781860409e-05, "loss": 0.0403, "step": 25980 }, { "grad_norm": 0.3549722135066986, "learning_rate": 8.84316928559686e-05, "loss": 0.0388, "step": 25990 }, { "grad_norm": 0.3939613401889801, "learning_rate": 8.842111369050094e-05, "loss": 0.0348, "step": 26000 }, { "grad_norm": 0.42519208788871765, "learning_rate": 8.841053032335808e-05, "loss": 0.0468, "step": 26010 }, { "grad_norm": 0.3744244873523712, "learning_rate": 8.839994275569735e-05, "loss": 0.043, "step": 26020 }, { "grad_norm": 0.5234925150871277, "learning_rate": 8.838935098867662e-05, "loss": 0.035, "step": 26030 }, { "grad_norm": 0.40380650758743286, "learning_rate": 8.837875502345418e-05, "loss": 0.0405, "step": 26040 }, { "grad_norm": 0.3486019968986511, "learning_rate": 8.83681548611888e-05, "loss": 0.0427, "step": 26050 }, { "grad_norm": 0.2722741663455963, "learning_rate": 8.835755050303969e-05, "loss": 0.0379, "step": 26060 }, { "grad_norm": 0.3180898427963257, "learning_rate": 8.834694195016653e-05, "loss": 0.0471, "step": 26070 }, { "grad_norm": 0.42848074436187744, "learning_rate": 8.833632920372942e-05, "loss": 0.0562, "step": 26080 }, { "grad_norm": 0.48612090945243835, "learning_rate": 8.832571226488903e-05, "loss": 0.0491, "step": 26090 }, { "grad_norm": 0.4658890962600708, "learning_rate": 8.831509113480634e-05, "loss": 0.0457, "step": 26100 }, { "grad_norm": 0.31873273849487305, "learning_rate": 8.83044658146429e-05, "loss": 0.0387, "step": 26110 }, { "grad_norm": 0.5339980721473694, "learning_rate": 8.829383630556067e-05, "loss": 0.0451, "step": 26120 }, { "grad_norm": 0.3627278506755829, "learning_rate": 8.828320260872207e-05, "loss": 0.0416, "step": 26130 }, { "grad_norm": 0.39463135600090027, "learning_rate": 8.827256472529e-05, "loss": 0.0375, "step": 26140 }, { "grad_norm": 0.47450485825538635, "learning_rate": 8.826192265642778e-05, "loss": 0.0455, "step": 26150 }, { "grad_norm": 0.47999975085258484, "learning_rate": 8.825127640329923e-05, "loss": 0.041, "step": 26160 }, { "grad_norm": 0.4834895133972168, "learning_rate": 8.824062596706861e-05, "loss": 0.0458, "step": 26170 }, { "grad_norm": 0.5798925757408142, "learning_rate": 8.822997134890062e-05, "loss": 0.0367, "step": 26180 }, { "grad_norm": 0.5309967994689941, "learning_rate": 8.821931254996044e-05, "loss": 0.0369, "step": 26190 }, { "grad_norm": 0.4167526662349701, "learning_rate": 8.82086495714137e-05, "loss": 0.0453, "step": 26200 }, { "grad_norm": 0.5299122333526611, "learning_rate": 8.81979824144265e-05, "loss": 0.0443, "step": 26210 }, { "grad_norm": 0.42486751079559326, "learning_rate": 8.818731108016536e-05, "loss": 0.0387, "step": 26220 }, { "grad_norm": 0.620071291923523, "learning_rate": 8.81766355697973e-05, "loss": 0.0474, "step": 26230 }, { "grad_norm": 0.38308364152908325, "learning_rate": 8.816595588448977e-05, "loss": 0.0356, "step": 26240 }, { "grad_norm": 0.4086436629295349, "learning_rate": 8.81552720254107e-05, "loss": 0.05, "step": 26250 }, { "grad_norm": 0.4302578568458557, "learning_rate": 8.814458399372842e-05, "loss": 0.0417, "step": 26260 }, { "grad_norm": 0.398613840341568, "learning_rate": 8.813389179061181e-05, "loss": 0.0333, "step": 26270 }, { "grad_norm": 0.37748071551322937, "learning_rate": 8.812319541723012e-05, "loss": 0.0497, "step": 26280 }, { "grad_norm": 0.3198792040348053, "learning_rate": 8.811249487475309e-05, "loss": 0.0515, "step": 26290 }, { "grad_norm": 0.48080524802207947, "learning_rate": 8.810179016435092e-05, "loss": 0.0531, "step": 26300 }, { "grad_norm": 0.5297011733055115, "learning_rate": 8.809108128719428e-05, "loss": 0.0509, "step": 26310 }, { "grad_norm": 0.5573291182518005, "learning_rate": 8.808036824445424e-05, "loss": 0.0472, "step": 26320 }, { "grad_norm": 0.5395667552947998, "learning_rate": 8.806965103730238e-05, "loss": 0.0451, "step": 26330 }, { "grad_norm": 0.44183799624443054, "learning_rate": 8.805892966691074e-05, "loss": 0.0409, "step": 26340 }, { "grad_norm": 0.29748034477233887, "learning_rate": 8.804820413445175e-05, "loss": 0.0386, "step": 26350 }, { "grad_norm": 0.48863258957862854, "learning_rate": 8.803747444109837e-05, "loss": 0.0424, "step": 26360 }, { "grad_norm": 1.2917084693908691, "learning_rate": 8.802674058802399e-05, "loss": 0.0405, "step": 26370 }, { "grad_norm": 0.5418098568916321, "learning_rate": 8.801600257640241e-05, "loss": 0.0395, "step": 26380 }, { "grad_norm": 0.49400633573532104, "learning_rate": 8.800526040740795e-05, "loss": 0.0451, "step": 26390 }, { "grad_norm": 0.6394933462142944, "learning_rate": 8.799451408221535e-05, "loss": 0.0456, "step": 26400 }, { "grad_norm": 0.3802223801612854, "learning_rate": 8.798376360199982e-05, "loss": 0.0417, "step": 26410 }, { "grad_norm": 0.42442771792411804, "learning_rate": 8.797300896793701e-05, "loss": 0.0363, "step": 26420 }, { "grad_norm": 0.32425153255462646, "learning_rate": 8.796225018120302e-05, "loss": 0.0445, "step": 26430 }, { "grad_norm": 0.2870631814002991, "learning_rate": 8.795148724297444e-05, "loss": 0.0422, "step": 26440 }, { "grad_norm": 0.40549758076667786, "learning_rate": 8.794072015442825e-05, "loss": 0.0454, "step": 26450 }, { "grad_norm": 0.3541848957538605, "learning_rate": 8.792994891674198e-05, "loss": 0.0442, "step": 26460 }, { "grad_norm": 0.4413270950317383, "learning_rate": 8.79191735310935e-05, "loss": 0.0322, "step": 26470 }, { "grad_norm": 0.363126665353775, "learning_rate": 8.790839399866122e-05, "loss": 0.0421, "step": 26480 }, { "grad_norm": 0.3517633378505707, "learning_rate": 8.789761032062397e-05, "loss": 0.0368, "step": 26490 }, { "grad_norm": 0.48863545060157776, "learning_rate": 8.788682249816103e-05, "loss": 0.0472, "step": 26500 }, { "grad_norm": 0.4045163094997406, "learning_rate": 8.787603053245215e-05, "loss": 0.0489, "step": 26510 }, { "grad_norm": 0.38597166538238525, "learning_rate": 8.78652344246775e-05, "loss": 0.046, "step": 26520 }, { "grad_norm": 0.420684278011322, "learning_rate": 8.785443417601776e-05, "loss": 0.0423, "step": 26530 }, { "grad_norm": 0.381433367729187, "learning_rate": 8.784362978765401e-05, "loss": 0.039, "step": 26540 }, { "grad_norm": 0.4687492847442627, "learning_rate": 8.783282126076779e-05, "loss": 0.0403, "step": 26550 }, { "grad_norm": 0.5205357074737549, "learning_rate": 8.782200859654112e-05, "loss": 0.0448, "step": 26560 }, { "grad_norm": 0.3058256208896637, "learning_rate": 8.781119179615646e-05, "loss": 0.0382, "step": 26570 }, { "grad_norm": 0.42532142996788025, "learning_rate": 8.780037086079674e-05, "loss": 0.0394, "step": 26580 }, { "grad_norm": 0.40846335887908936, "learning_rate": 8.778954579164527e-05, "loss": 0.0414, "step": 26590 }, { "grad_norm": 0.5286893844604492, "learning_rate": 8.777871658988588e-05, "loss": 0.0474, "step": 26600 }, { "grad_norm": 0.44914501905441284, "learning_rate": 8.776788325670285e-05, "loss": 0.0428, "step": 26610 }, { "grad_norm": 0.3342041075229645, "learning_rate": 8.775704579328089e-05, "loss": 0.0469, "step": 26620 }, { "grad_norm": 0.35114672780036926, "learning_rate": 8.774620420080517e-05, "loss": 0.0448, "step": 26630 }, { "grad_norm": 0.43151500821113586, "learning_rate": 8.773535848046131e-05, "loss": 0.0376, "step": 26640 }, { "grad_norm": 0.44216763973236084, "learning_rate": 8.772450863343538e-05, "loss": 0.0342, "step": 26650 }, { "grad_norm": 0.44450196623802185, "learning_rate": 8.77136546609139e-05, "loss": 0.0433, "step": 26660 }, { "grad_norm": 0.4126536548137665, "learning_rate": 8.770279656408385e-05, "loss": 0.0386, "step": 26670 }, { "grad_norm": 0.4208759665489197, "learning_rate": 8.769193434413265e-05, "loss": 0.0431, "step": 26680 }, { "grad_norm": 0.3812468349933624, "learning_rate": 8.76810680022482e-05, "loss": 0.0415, "step": 26690 }, { "grad_norm": 0.44994285702705383, "learning_rate": 8.767019753961878e-05, "loss": 0.0558, "step": 26700 }, { "grad_norm": 0.3646361231803894, "learning_rate": 8.765932295743321e-05, "loss": 0.0446, "step": 26710 }, { "grad_norm": 0.47740405797958374, "learning_rate": 8.764844425688068e-05, "loss": 0.0443, "step": 26720 }, { "grad_norm": 0.40756791830062866, "learning_rate": 8.763756143915092e-05, "loss": 0.0342, "step": 26730 }, { "grad_norm": 0.41087114810943604, "learning_rate": 8.7626674505434e-05, "loss": 0.0378, "step": 26740 }, { "grad_norm": 0.38844844698905945, "learning_rate": 8.761578345692053e-05, "loss": 0.0447, "step": 26750 }, { "grad_norm": 0.2863859236240387, "learning_rate": 8.760488829480156e-05, "loss": 0.0351, "step": 26760 }, { "grad_norm": 0.4132549464702606, "learning_rate": 8.759398902026854e-05, "loss": 0.0377, "step": 26770 }, { "grad_norm": 0.4778788387775421, "learning_rate": 8.758308563451339e-05, "loss": 0.0427, "step": 26780 }, { "grad_norm": 0.42227619886398315, "learning_rate": 8.75721781387285e-05, "loss": 0.0458, "step": 26790 }, { "grad_norm": 0.35708194971084595, "learning_rate": 8.75612665341067e-05, "loss": 0.0427, "step": 26800 }, { "grad_norm": 0.3545858860015869, "learning_rate": 8.755035082184126e-05, "loss": 0.0373, "step": 26810 }, { "grad_norm": 0.4030076563358307, "learning_rate": 8.753943100312592e-05, "loss": 0.0418, "step": 26820 }, { "grad_norm": 0.3853021264076233, "learning_rate": 8.752850707915484e-05, "loss": 0.0367, "step": 26830 }, { "grad_norm": 0.5687263011932373, "learning_rate": 8.751757905112264e-05, "loss": 0.0435, "step": 26840 }, { "grad_norm": 0.46637436747550964, "learning_rate": 8.75066469202244e-05, "loss": 0.0397, "step": 26850 }, { "grad_norm": 0.47972503304481506, "learning_rate": 8.749571068765567e-05, "loss": 0.0419, "step": 26860 }, { "grad_norm": 0.32699838280677795, "learning_rate": 8.748477035461238e-05, "loss": 0.0371, "step": 26870 }, { "grad_norm": 0.39590224623680115, "learning_rate": 8.747382592229095e-05, "loss": 0.0477, "step": 26880 }, { "grad_norm": 0.3582266867160797, "learning_rate": 8.746287739188828e-05, "loss": 0.0364, "step": 26890 }, { "grad_norm": 0.535392165184021, "learning_rate": 8.745192476460165e-05, "loss": 0.0409, "step": 26900 }, { "grad_norm": 0.3139107823371887, "learning_rate": 8.744096804162882e-05, "loss": 0.0428, "step": 26910 }, { "grad_norm": 0.6052684187889099, "learning_rate": 8.743000722416804e-05, "loss": 0.0463, "step": 26920 }, { "grad_norm": 0.44603651762008667, "learning_rate": 8.741904231341793e-05, "loss": 0.0415, "step": 26930 }, { "grad_norm": 0.5909829139709473, "learning_rate": 8.740807331057762e-05, "loss": 0.0514, "step": 26940 }, { "grad_norm": 0.43575412034988403, "learning_rate": 8.739710021684667e-05, "loss": 0.0447, "step": 26950 }, { "grad_norm": 0.41708311438560486, "learning_rate": 8.738612303342503e-05, "loss": 0.0391, "step": 26960 }, { "grad_norm": 0.33489546179771423, "learning_rate": 8.73751417615132e-05, "loss": 0.0444, "step": 26970 }, { "grad_norm": 0.393189936876297, "learning_rate": 8.736415640231208e-05, "loss": 0.0348, "step": 26980 }, { "grad_norm": 0.31897997856140137, "learning_rate": 8.735316695702297e-05, "loss": 0.037, "step": 26990 }, { "grad_norm": 0.38696226477622986, "learning_rate": 8.734217342684769e-05, "loss": 0.0374, "step": 27000 }, { "grad_norm": 0.4643322229385376, "learning_rate": 8.733117581298847e-05, "loss": 0.038, "step": 27010 }, { "grad_norm": 0.44392016530036926, "learning_rate": 8.732017411664796e-05, "loss": 0.0375, "step": 27020 }, { "grad_norm": 0.4320891201496124, "learning_rate": 8.730916833902936e-05, "loss": 0.0325, "step": 27030 }, { "grad_norm": 0.42647412419319153, "learning_rate": 8.729815848133618e-05, "loss": 0.0518, "step": 27040 }, { "grad_norm": 0.3832939863204956, "learning_rate": 8.728714454477247e-05, "loss": 0.0364, "step": 27050 }, { "grad_norm": 0.2934373915195465, "learning_rate": 8.727612653054269e-05, "loss": 0.0326, "step": 27060 }, { "grad_norm": 0.3049112558364868, "learning_rate": 8.726510443985176e-05, "loss": 0.0354, "step": 27070 }, { "grad_norm": 0.38006988167762756, "learning_rate": 8.725407827390503e-05, "loss": 0.039, "step": 27080 }, { "grad_norm": 0.3149518072605133, "learning_rate": 8.724304803390833e-05, "loss": 0.0382, "step": 27090 }, { "grad_norm": 0.32001930475234985, "learning_rate": 8.723201372106788e-05, "loss": 0.0411, "step": 27100 }, { "grad_norm": 0.4448491632938385, "learning_rate": 8.722097533659038e-05, "loss": 0.0469, "step": 27110 }, { "grad_norm": 0.726185142993927, "learning_rate": 8.720993288168299e-05, "loss": 0.0428, "step": 27120 }, { "grad_norm": 0.37880977988243103, "learning_rate": 8.719888635755327e-05, "loss": 0.0401, "step": 27130 }, { "grad_norm": 0.4211772084236145, "learning_rate": 8.718783576540928e-05, "loss": 0.0491, "step": 27140 }, { "grad_norm": 0.42436107993125916, "learning_rate": 8.717678110645948e-05, "loss": 0.04, "step": 27150 }, { "grad_norm": 0.3099361062049866, "learning_rate": 8.716572238191279e-05, "loss": 0.038, "step": 27160 }, { "grad_norm": 0.4044199287891388, "learning_rate": 8.715465959297857e-05, "loss": 0.0577, "step": 27170 }, { "grad_norm": 0.6095699071884155, "learning_rate": 8.714359274086665e-05, "loss": 0.0414, "step": 27180 }, { "grad_norm": 0.3066789507865906, "learning_rate": 8.713252182678726e-05, "loss": 0.0426, "step": 27190 }, { "grad_norm": 0.36684536933898926, "learning_rate": 8.712144685195112e-05, "loss": 0.0477, "step": 27200 }, { "grad_norm": 0.41802123188972473, "learning_rate": 8.711036781756936e-05, "loss": 0.043, "step": 27210 }, { "grad_norm": 0.6151764988899231, "learning_rate": 8.709928472485357e-05, "loss": 0.043, "step": 27220 }, { "grad_norm": 0.4136204123497009, "learning_rate": 8.708819757501579e-05, "loss": 0.036, "step": 27230 }, { "grad_norm": 0.35487625002861023, "learning_rate": 8.707710636926846e-05, "loss": 0.0407, "step": 27240 }, { "grad_norm": 0.46581000089645386, "learning_rate": 8.706601110882455e-05, "loss": 0.0477, "step": 27250 }, { "grad_norm": 0.6071615219116211, "learning_rate": 8.705491179489738e-05, "loss": 0.0388, "step": 27260 }, { "grad_norm": 0.524683952331543, "learning_rate": 8.704380842870077e-05, "loss": 0.0351, "step": 27270 }, { "grad_norm": 0.3290136456489563, "learning_rate": 8.703270101144895e-05, "loss": 0.0367, "step": 27280 }, { "grad_norm": 0.55184406042099, "learning_rate": 8.702158954435664e-05, "loss": 0.0353, "step": 27290 }, { "grad_norm": 0.2596089839935303, "learning_rate": 8.701047402863896e-05, "loss": 0.0325, "step": 27300 }, { "grad_norm": 0.41944923996925354, "learning_rate": 8.699935446551148e-05, "loss": 0.0418, "step": 27310 }, { "grad_norm": 0.40566569566726685, "learning_rate": 8.698823085619022e-05, "loss": 0.0315, "step": 27320 }, { "grad_norm": 0.29398584365844727, "learning_rate": 8.697710320189166e-05, "loss": 0.0311, "step": 27330 }, { "grad_norm": 0.45042338967323303, "learning_rate": 8.696597150383268e-05, "loss": 0.0399, "step": 27340 }, { "grad_norm": 0.4709487557411194, "learning_rate": 8.695483576323063e-05, "loss": 0.0415, "step": 27350 }, { "grad_norm": 0.40329697728157043, "learning_rate": 8.69436959813033e-05, "loss": 0.0455, "step": 27360 }, { "grad_norm": 0.28094130754470825, "learning_rate": 8.693255215926892e-05, "loss": 0.0351, "step": 27370 }, { "grad_norm": 0.4796828627586365, "learning_rate": 8.692140429834617e-05, "loss": 0.0376, "step": 27380 }, { "grad_norm": 0.3865768015384674, "learning_rate": 8.691025239975415e-05, "loss": 0.0384, "step": 27390 }, { "grad_norm": 0.6220869421958923, "learning_rate": 8.689909646471243e-05, "loss": 0.0372, "step": 27400 }, { "grad_norm": 0.3488219976425171, "learning_rate": 8.688793649444099e-05, "loss": 0.0446, "step": 27410 }, { "grad_norm": 0.48256731033325195, "learning_rate": 8.687677249016029e-05, "loss": 0.0397, "step": 27420 }, { "grad_norm": 0.5839152336120605, "learning_rate": 8.686560445309118e-05, "loss": 0.0422, "step": 27430 }, { "grad_norm": 0.628180205821991, "learning_rate": 8.685443238445499e-05, "loss": 0.0489, "step": 27440 }, { "grad_norm": 0.43919679522514343, "learning_rate": 8.68432562854735e-05, "loss": 0.0389, "step": 27450 }, { "grad_norm": 0.507077157497406, "learning_rate": 8.683207615736887e-05, "loss": 0.0432, "step": 27460 }, { "grad_norm": 0.5041781663894653, "learning_rate": 8.682089200136379e-05, "loss": 0.0522, "step": 27470 }, { "grad_norm": 0.5329259037971497, "learning_rate": 8.680970381868132e-05, "loss": 0.0457, "step": 27480 }, { "grad_norm": 0.44913598895072937, "learning_rate": 8.679851161054498e-05, "loss": 0.045, "step": 27490 }, { "grad_norm": 0.5478232502937317, "learning_rate": 8.678731537817873e-05, "loss": 0.0491, "step": 27500 }, { "grad_norm": 0.534076452255249, "learning_rate": 8.677611512280697e-05, "loss": 0.0423, "step": 27510 }, { "grad_norm": 0.34578582644462585, "learning_rate": 8.676491084565457e-05, "loss": 0.0514, "step": 27520 }, { "grad_norm": 0.4971577823162079, "learning_rate": 8.675370254794678e-05, "loss": 0.0573, "step": 27530 }, { "grad_norm": 0.43372973799705505, "learning_rate": 8.674249023090935e-05, "loss": 0.0415, "step": 27540 }, { "grad_norm": 0.34631770849227905, "learning_rate": 8.673127389576843e-05, "loss": 0.0442, "step": 27550 }, { "grad_norm": 0.2864795923233032, "learning_rate": 8.67200535437506e-05, "loss": 0.0351, "step": 27560 }, { "grad_norm": 0.44512319564819336, "learning_rate": 8.670882917608296e-05, "loss": 0.0389, "step": 27570 }, { "grad_norm": 0.40125271677970886, "learning_rate": 8.669760079399292e-05, "loss": 0.0357, "step": 27580 }, { "grad_norm": 0.4180833399295807, "learning_rate": 8.668636839870845e-05, "loss": 0.0404, "step": 27590 }, { "grad_norm": 0.4688946604728699, "learning_rate": 8.667513199145789e-05, "loss": 0.049, "step": 27600 }, { "grad_norm": 0.6365872621536255, "learning_rate": 8.666389157347002e-05, "loss": 0.0482, "step": 27610 }, { "grad_norm": 0.44132673740386963, "learning_rate": 8.66526471459741e-05, "loss": 0.0425, "step": 27620 }, { "grad_norm": 0.3260229825973511, "learning_rate": 8.66413987101998e-05, "loss": 0.0391, "step": 27630 }, { "grad_norm": 0.6920737624168396, "learning_rate": 8.663014626737723e-05, "loss": 0.048, "step": 27640 }, { "grad_norm": 0.4167085289955139, "learning_rate": 8.661888981873691e-05, "loss": 0.0408, "step": 27650 }, { "grad_norm": 0.4402405917644501, "learning_rate": 8.660762936550988e-05, "loss": 0.0563, "step": 27660 }, { "grad_norm": 0.41838905215263367, "learning_rate": 8.659636490892753e-05, "loss": 0.0518, "step": 27670 }, { "grad_norm": 0.5251144170761108, "learning_rate": 8.658509645022174e-05, "loss": 0.0466, "step": 27680 }, { "grad_norm": 0.40803733468055725, "learning_rate": 8.657382399062481e-05, "loss": 0.0384, "step": 27690 }, { "grad_norm": 0.4059421718120575, "learning_rate": 8.656254753136946e-05, "loss": 0.0397, "step": 27700 }, { "grad_norm": 0.32295429706573486, "learning_rate": 8.655126707368891e-05, "loss": 0.0366, "step": 27710 }, { "grad_norm": 0.39499735832214355, "learning_rate": 8.653998261881672e-05, "loss": 0.0367, "step": 27720 }, { "grad_norm": 0.41214242577552795, "learning_rate": 8.652869416798699e-05, "loss": 0.0342, "step": 27730 }, { "grad_norm": 0.7835559248924255, "learning_rate": 8.651740172243417e-05, "loss": 0.0395, "step": 27740 }, { "grad_norm": 0.4474407732486725, "learning_rate": 8.65061052833932e-05, "loss": 0.0442, "step": 27750 }, { "grad_norm": 0.3902808129787445, "learning_rate": 8.649480485209945e-05, "loss": 0.0391, "step": 27760 }, { "grad_norm": 0.3723195493221283, "learning_rate": 8.64835004297887e-05, "loss": 0.0454, "step": 27770 }, { "grad_norm": 0.32291293144226074, "learning_rate": 8.64721920176972e-05, "loss": 0.0358, "step": 27780 }, { "grad_norm": 0.5969051122665405, "learning_rate": 8.646087961706164e-05, "loss": 0.0473, "step": 27790 }, { "grad_norm": 0.38935476541519165, "learning_rate": 8.644956322911908e-05, "loss": 0.0334, "step": 27800 }, { "grad_norm": 0.5064986348152161, "learning_rate": 8.643824285510709e-05, "loss": 0.0395, "step": 27810 }, { "grad_norm": 0.24640044569969177, "learning_rate": 8.642691849626364e-05, "loss": 0.0325, "step": 27820 }, { "grad_norm": 0.38631170988082886, "learning_rate": 8.641559015382717e-05, "loss": 0.0374, "step": 27830 }, { "grad_norm": 0.3996540606021881, "learning_rate": 8.640425782903649e-05, "loss": 0.0451, "step": 27840 }, { "grad_norm": 0.381529301404953, "learning_rate": 8.639292152313091e-05, "loss": 0.0344, "step": 27850 }, { "grad_norm": 0.37720248103141785, "learning_rate": 8.638158123735015e-05, "loss": 0.0374, "step": 27860 }, { "grad_norm": 0.6730778217315674, "learning_rate": 8.637023697293436e-05, "loss": 0.0457, "step": 27870 }, { "grad_norm": 1.041640281677246, "learning_rate": 8.635888873112414e-05, "loss": 0.0565, "step": 27880 }, { "grad_norm": 0.5414908528327942, "learning_rate": 8.634753651316052e-05, "loss": 0.0492, "step": 27890 }, { "grad_norm": 0.4686753749847412, "learning_rate": 8.633618032028496e-05, "loss": 0.0417, "step": 27900 }, { "grad_norm": 0.695459246635437, "learning_rate": 8.632482015373934e-05, "loss": 0.0415, "step": 27910 }, { "grad_norm": 0.7204639315605164, "learning_rate": 8.6313456014766e-05, "loss": 0.0412, "step": 27920 }, { "grad_norm": 0.36319395899772644, "learning_rate": 8.630208790460771e-05, "loss": 0.0507, "step": 27930 }, { "grad_norm": 0.38519132137298584, "learning_rate": 8.629071582450768e-05, "loss": 0.042, "step": 27940 }, { "grad_norm": 0.5212339162826538, "learning_rate": 8.62793397757095e-05, "loss": 0.0523, "step": 27950 }, { "grad_norm": 0.4116954207420349, "learning_rate": 8.626795975945729e-05, "loss": 0.0565, "step": 27960 }, { "grad_norm": 0.6863014698028564, "learning_rate": 8.625657577699551e-05, "loss": 0.138, "step": 27970 }, { "grad_norm": 0.4659636914730072, "learning_rate": 8.624518782956914e-05, "loss": 0.0467, "step": 27980 }, { "grad_norm": 0.4196372628211975, "learning_rate": 8.62337959184235e-05, "loss": 0.04, "step": 27990 }, { "grad_norm": 0.41995522379875183, "learning_rate": 8.622240004480441e-05, "loss": 0.0456, "step": 28000 }, { "grad_norm": 0.4691149592399597, "learning_rate": 8.621100020995814e-05, "loss": 0.0391, "step": 28010 }, { "grad_norm": 0.5287653207778931, "learning_rate": 8.619959641513132e-05, "loss": 0.0424, "step": 28020 }, { "grad_norm": 0.5004647970199585, "learning_rate": 8.618818866157105e-05, "loss": 0.045, "step": 28030 }, { "grad_norm": 0.5394233465194702, "learning_rate": 8.617677695052487e-05, "loss": 0.0415, "step": 28040 }, { "grad_norm": 0.5857919454574585, "learning_rate": 8.616536128324078e-05, "loss": 0.0357, "step": 28050 }, { "grad_norm": 0.37409910559654236, "learning_rate": 8.615394166096712e-05, "loss": 0.0352, "step": 28060 }, { "grad_norm": 0.33871546387672424, "learning_rate": 8.614251808495279e-05, "loss": 0.0326, "step": 28070 }, { "grad_norm": 0.4901394844055176, "learning_rate": 8.6131090556447e-05, "loss": 0.0322, "step": 28080 }, { "grad_norm": 0.2848614752292633, "learning_rate": 8.611965907669947e-05, "loss": 0.041, "step": 28090 }, { "grad_norm": 0.42562782764434814, "learning_rate": 8.610822364696034e-05, "loss": 0.0431, "step": 28100 }, { "grad_norm": 0.3797306716442108, "learning_rate": 8.609678426848015e-05, "loss": 0.0454, "step": 28110 }, { "grad_norm": 0.4414806067943573, "learning_rate": 8.60853409425099e-05, "loss": 0.0448, "step": 28120 }, { "grad_norm": 0.5400820970535278, "learning_rate": 8.607389367030104e-05, "loss": 0.0388, "step": 28130 }, { "grad_norm": 0.41627565026283264, "learning_rate": 8.606244245310538e-05, "loss": 0.0324, "step": 28140 }, { "grad_norm": 0.4589884281158447, "learning_rate": 8.605098729217525e-05, "loss": 0.0436, "step": 28150 }, { "grad_norm": 0.36644667387008667, "learning_rate": 8.603952818876335e-05, "loss": 0.0394, "step": 28160 }, { "grad_norm": 0.39586353302001953, "learning_rate": 8.602806514412281e-05, "loss": 0.032, "step": 28170 }, { "grad_norm": 0.4671497642993927, "learning_rate": 8.601659815950726e-05, "loss": 0.0477, "step": 28180 }, { "grad_norm": 0.3927175998687744, "learning_rate": 8.600512723617067e-05, "loss": 0.0552, "step": 28190 }, { "grad_norm": 0.6207441687583923, "learning_rate": 8.59936523753675e-05, "loss": 0.0504, "step": 28200 }, { "grad_norm": 0.29844653606414795, "learning_rate": 8.598217357835264e-05, "loss": 0.0352, "step": 28210 }, { "grad_norm": 0.38874611258506775, "learning_rate": 8.597069084638135e-05, "loss": 0.0349, "step": 28220 }, { "grad_norm": 0.46981504559516907, "learning_rate": 8.595920418070939e-05, "loss": 0.0343, "step": 28230 }, { "grad_norm": 0.40519335865974426, "learning_rate": 8.594771358259295e-05, "loss": 0.0363, "step": 28240 }, { "grad_norm": 0.4900965988636017, "learning_rate": 8.593621905328858e-05, "loss": 0.0488, "step": 28250 }, { "grad_norm": 0.3998127579689026, "learning_rate": 8.592472059405333e-05, "loss": 0.0471, "step": 28260 }, { "grad_norm": 0.3547559082508087, "learning_rate": 8.591321820614464e-05, "loss": 0.0321, "step": 28270 }, { "grad_norm": 0.3565165400505066, "learning_rate": 8.590171189082041e-05, "loss": 0.0405, "step": 28280 }, { "grad_norm": 0.3849167227745056, "learning_rate": 8.589020164933894e-05, "loss": 0.0409, "step": 28290 }, { "grad_norm": 0.2866869866847992, "learning_rate": 8.587868748295898e-05, "loss": 0.0453, "step": 28300 }, { "grad_norm": 0.4063558578491211, "learning_rate": 8.586716939293971e-05, "loss": 0.0416, "step": 28310 }, { "grad_norm": 0.299403578042984, "learning_rate": 8.58556473805407e-05, "loss": 0.037, "step": 28320 }, { "grad_norm": 0.4834609031677246, "learning_rate": 8.584412144702202e-05, "loss": 0.0458, "step": 28330 }, { "grad_norm": 0.427270770072937, "learning_rate": 8.58325915936441e-05, "loss": 0.0391, "step": 28340 }, { "grad_norm": 0.4425901174545288, "learning_rate": 8.582105782166783e-05, "loss": 0.0391, "step": 28350 }, { "grad_norm": 0.30953627824783325, "learning_rate": 8.580952013235455e-05, "loss": 0.034, "step": 28360 }, { "grad_norm": 0.5235375761985779, "learning_rate": 8.579797852696596e-05, "loss": 0.0452, "step": 28370 }, { "grad_norm": 0.3881613314151764, "learning_rate": 8.578643300676428e-05, "loss": 0.0328, "step": 28380 }, { "grad_norm": 0.6984242796897888, "learning_rate": 8.577488357301209e-05, "loss": 0.0445, "step": 28390 }, { "grad_norm": 0.34654298424720764, "learning_rate": 8.576333022697242e-05, "loss": 0.0382, "step": 28400 }, { "grad_norm": 0.49306604266166687, "learning_rate": 8.575177296990873e-05, "loss": 0.042, "step": 28410 }, { "grad_norm": 0.5196893215179443, "learning_rate": 8.574021180308489e-05, "loss": 0.0458, "step": 28420 }, { "grad_norm": 0.4976309537887573, "learning_rate": 8.572864672776523e-05, "loss": 0.0363, "step": 28430 }, { "grad_norm": 0.4171842038631439, "learning_rate": 8.571707774521447e-05, "loss": 0.0366, "step": 28440 }, { "grad_norm": 0.4530760943889618, "learning_rate": 8.57055048566978e-05, "loss": 0.0327, "step": 28450 }, { "grad_norm": 0.48177626729011536, "learning_rate": 8.569392806348078e-05, "loss": 0.0396, "step": 28460 }, { "grad_norm": 0.28935763239860535, "learning_rate": 8.568234736682947e-05, "loss": 0.0371, "step": 28470 }, { "grad_norm": 0.5197708606719971, "learning_rate": 8.567076276801029e-05, "loss": 0.0443, "step": 28480 }, { "grad_norm": 0.38786032795906067, "learning_rate": 8.565917426829013e-05, "loss": 0.0362, "step": 28490 }, { "grad_norm": 0.3750564157962799, "learning_rate": 8.564758186893628e-05, "loss": 0.0437, "step": 28500 }, { "grad_norm": 0.417566180229187, "learning_rate": 8.563598557121649e-05, "loss": 0.0326, "step": 28510 }, { "grad_norm": 0.36102691292762756, "learning_rate": 8.562438537639888e-05, "loss": 0.0392, "step": 28520 }, { "grad_norm": 0.513648509979248, "learning_rate": 8.561278128575206e-05, "loss": 0.0428, "step": 28530 }, { "grad_norm": 0.37169232964515686, "learning_rate": 8.5601173300545e-05, "loss": 0.0378, "step": 28540 }, { "grad_norm": 0.5194171071052551, "learning_rate": 8.558956142204717e-05, "loss": 0.0401, "step": 28550 }, { "grad_norm": 0.627585768699646, "learning_rate": 8.55779456515284e-05, "loss": 0.0478, "step": 28560 }, { "grad_norm": 0.5288770794868469, "learning_rate": 8.556632599025898e-05, "loss": 0.0472, "step": 28570 }, { "grad_norm": 0.36778610944747925, "learning_rate": 8.555470243950964e-05, "loss": 0.0423, "step": 28580 }, { "grad_norm": 0.3816213309764862, "learning_rate": 8.554307500055148e-05, "loss": 0.0386, "step": 28590 }, { "grad_norm": 0.34892797470092773, "learning_rate": 8.553144367465609e-05, "loss": 0.0395, "step": 28600 }, { "grad_norm": 0.5360468029975891, "learning_rate": 8.551980846309544e-05, "loss": 0.0516, "step": 28610 }, { "grad_norm": 0.27354806661605835, "learning_rate": 8.550816936714193e-05, "loss": 0.0455, "step": 28620 }, { "grad_norm": 0.35134366154670715, "learning_rate": 8.549652638806841e-05, "loss": 0.0404, "step": 28630 }, { "grad_norm": 1.5835597515106201, "learning_rate": 8.548487952714812e-05, "loss": 0.0424, "step": 28640 }, { "grad_norm": 0.4709630012512207, "learning_rate": 8.547322878565478e-05, "loss": 0.0456, "step": 28650 }, { "grad_norm": 0.5429190397262573, "learning_rate": 8.546157416486245e-05, "loss": 0.0518, "step": 28660 }, { "grad_norm": 0.8013826608657837, "learning_rate": 8.54499156660457e-05, "loss": 0.0398, "step": 28670 }, { "grad_norm": 0.5329204201698303, "learning_rate": 8.543825329047947e-05, "loss": 0.0428, "step": 28680 }, { "grad_norm": 0.3355315625667572, "learning_rate": 8.542658703943913e-05, "loss": 0.0474, "step": 28690 }, { "grad_norm": 0.36551180481910706, "learning_rate": 8.541491691420051e-05, "loss": 0.0446, "step": 28700 }, { "grad_norm": 0.35214319825172424, "learning_rate": 8.54032429160398e-05, "loss": 0.0398, "step": 28710 }, { "grad_norm": 0.5073278546333313, "learning_rate": 8.539156504623369e-05, "loss": 0.0519, "step": 28720 }, { "grad_norm": 0.6553857922554016, "learning_rate": 8.537988330605923e-05, "loss": 0.0634, "step": 28730 }, { "grad_norm": 0.4936261475086212, "learning_rate": 8.536819769679393e-05, "loss": 0.0346, "step": 28740 }, { "grad_norm": 0.517982006072998, "learning_rate": 8.53565082197157e-05, "loss": 0.0491, "step": 28750 }, { "grad_norm": 0.347814679145813, "learning_rate": 8.534481487610289e-05, "loss": 0.0539, "step": 28760 }, { "grad_norm": 0.4261833727359772, "learning_rate": 8.533311766723428e-05, "loss": 0.038, "step": 28770 }, { "grad_norm": 0.39874017238616943, "learning_rate": 8.532141659438901e-05, "loss": 0.0397, "step": 28780 }, { "grad_norm": 0.2881341576576233, "learning_rate": 8.530971165884675e-05, "loss": 0.0396, "step": 28790 }, { "grad_norm": 0.47202637791633606, "learning_rate": 8.529800286188752e-05, "loss": 0.0398, "step": 28800 }, { "grad_norm": 0.39648497104644775, "learning_rate": 8.528629020479175e-05, "loss": 0.0459, "step": 28810 }, { "grad_norm": 0.3699898421764374, "learning_rate": 8.527457368884033e-05, "loss": 0.0496, "step": 28820 }, { "grad_norm": 0.34105828404426575, "learning_rate": 8.526285331531458e-05, "loss": 0.0387, "step": 28830 }, { "grad_norm": 0.5075435638427734, "learning_rate": 8.525112908549621e-05, "loss": 0.0421, "step": 28840 }, { "grad_norm": 0.5437759160995483, "learning_rate": 8.523940100066735e-05, "loss": 0.0416, "step": 28850 }, { "grad_norm": 0.5268001556396484, "learning_rate": 8.52276690621106e-05, "loss": 0.0435, "step": 28860 }, { "grad_norm": 0.40218889713287354, "learning_rate": 8.521593327110889e-05, "loss": 0.038, "step": 28870 }, { "grad_norm": 0.4013984203338623, "learning_rate": 8.520419362894569e-05, "loss": 0.0457, "step": 28880 }, { "grad_norm": 0.37208670377731323, "learning_rate": 8.51924501369048e-05, "loss": 0.0327, "step": 28890 }, { "grad_norm": 0.4135604202747345, "learning_rate": 8.518070279627047e-05, "loss": 0.0435, "step": 28900 }, { "grad_norm": 0.49188244342803955, "learning_rate": 8.516895160832737e-05, "loss": 0.0429, "step": 28910 }, { "grad_norm": 0.41243255138397217, "learning_rate": 8.515719657436061e-05, "loss": 0.0334, "step": 28920 }, { "grad_norm": 0.47622618079185486, "learning_rate": 8.514543769565568e-05, "loss": 0.038, "step": 28930 }, { "grad_norm": 0.46489042043685913, "learning_rate": 8.513367497349853e-05, "loss": 0.0496, "step": 28940 }, { "grad_norm": 0.33438563346862793, "learning_rate": 8.51219084091755e-05, "loss": 0.0423, "step": 28950 }, { "grad_norm": 0.31825047731399536, "learning_rate": 8.511013800397338e-05, "loss": 0.0371, "step": 28960 }, { "grad_norm": 0.42074650526046753, "learning_rate": 8.509836375917937e-05, "loss": 0.0373, "step": 28970 }, { "grad_norm": 0.4634209871292114, "learning_rate": 8.508658567608104e-05, "loss": 0.0356, "step": 28980 }, { "grad_norm": 0.3198583126068115, "learning_rate": 8.507480375596647e-05, "loss": 0.0332, "step": 28990 }, { "grad_norm": 0.49003866314888, "learning_rate": 8.506301800012408e-05, "loss": 0.0378, "step": 29000 }, { "grad_norm": 0.3229992091655731, "learning_rate": 8.505122840984278e-05, "loss": 0.0341, "step": 29010 }, { "grad_norm": 0.3656332194805145, "learning_rate": 8.503943498641182e-05, "loss": 0.0378, "step": 29020 }, { "grad_norm": 0.33678725361824036, "learning_rate": 8.502763773112095e-05, "loss": 0.0387, "step": 29030 }, { "grad_norm": 0.4165683090686798, "learning_rate": 8.501583664526026e-05, "loss": 0.0398, "step": 29040 }, { "grad_norm": 0.2980646789073944, "learning_rate": 8.500403173012032e-05, "loss": 0.0388, "step": 29050 }, { "grad_norm": 0.7100784778594971, "learning_rate": 8.499222298699211e-05, "loss": 0.0364, "step": 29060 }, { "grad_norm": 0.44144538044929504, "learning_rate": 8.498041041716701e-05, "loss": 0.0444, "step": 29070 }, { "grad_norm": 0.44582226872444153, "learning_rate": 8.496859402193681e-05, "loss": 0.0382, "step": 29080 }, { "grad_norm": 0.4320531189441681, "learning_rate": 8.495677380259374e-05, "loss": 0.0358, "step": 29090 }, { "grad_norm": 0.5420117378234863, "learning_rate": 8.494494976043045e-05, "loss": 0.0475, "step": 29100 }, { "grad_norm": 0.3223351836204529, "learning_rate": 8.493312189673998e-05, "loss": 0.0347, "step": 29110 }, { "grad_norm": 0.3352968096733093, "learning_rate": 8.492129021281584e-05, "loss": 0.0407, "step": 29120 }, { "grad_norm": 0.5237254500389099, "learning_rate": 8.490945470995188e-05, "loss": 0.0344, "step": 29130 }, { "grad_norm": 0.434608519077301, "learning_rate": 8.489761538944247e-05, "loss": 0.033, "step": 29140 }, { "grad_norm": 0.4536823630332947, "learning_rate": 8.48857722525823e-05, "loss": 0.0416, "step": 29150 }, { "grad_norm": 0.3600725829601288, "learning_rate": 8.487392530066652e-05, "loss": 0.0448, "step": 29160 }, { "grad_norm": 0.3675665557384491, "learning_rate": 8.486207453499069e-05, "loss": 0.0521, "step": 29170 }, { "grad_norm": 0.43177926540374756, "learning_rate": 8.485021995685082e-05, "loss": 0.0416, "step": 29180 }, { "grad_norm": 0.40225642919540405, "learning_rate": 8.483836156754328e-05, "loss": 0.0422, "step": 29190 }, { "grad_norm": 0.41974693536758423, "learning_rate": 8.482649936836491e-05, "loss": 0.0313, "step": 29200 }, { "grad_norm": 0.434594064950943, "learning_rate": 8.481463336061293e-05, "loss": 0.0376, "step": 29210 }, { "grad_norm": 0.3831857442855835, "learning_rate": 8.480276354558496e-05, "loss": 0.0493, "step": 29220 }, { "grad_norm": 0.42981162667274475, "learning_rate": 8.479088992457913e-05, "loss": 0.0408, "step": 29230 }, { "grad_norm": 0.42290693521499634, "learning_rate": 8.477901249889387e-05, "loss": 0.0439, "step": 29240 }, { "grad_norm": 0.46730101108551025, "learning_rate": 8.47671312698281e-05, "loss": 0.0384, "step": 29250 }, { "grad_norm": 0.4492683410644531, "learning_rate": 8.475524623868112e-05, "loss": 0.0415, "step": 29260 }, { "grad_norm": 0.6356605291366577, "learning_rate": 8.474335740675266e-05, "loss": 0.0377, "step": 29270 }, { "grad_norm": 0.4901793301105499, "learning_rate": 8.473146477534289e-05, "loss": 0.042, "step": 29280 }, { "grad_norm": 0.5110253691673279, "learning_rate": 8.471956834575232e-05, "loss": 0.0472, "step": 29290 }, { "grad_norm": 0.4197562634944916, "learning_rate": 8.470766811928197e-05, "loss": 0.045, "step": 29300 }, { "grad_norm": 0.3792935907840729, "learning_rate": 8.469576409723323e-05, "loss": 0.046, "step": 29310 }, { "grad_norm": 0.36531490087509155, "learning_rate": 8.468385628090788e-05, "loss": 0.0351, "step": 29320 }, { "grad_norm": 0.31039583683013916, "learning_rate": 8.467194467160815e-05, "loss": 0.0365, "step": 29330 }, { "grad_norm": 0.46833160519599915, "learning_rate": 8.466002927063667e-05, "loss": 0.033, "step": 29340 }, { "grad_norm": 0.4995599091053009, "learning_rate": 8.464811007929651e-05, "loss": 0.0372, "step": 29350 }, { "grad_norm": 0.42273059487342834, "learning_rate": 8.463618709889114e-05, "loss": 0.041, "step": 29360 }, { "grad_norm": 0.5984627604484558, "learning_rate": 8.462426033072442e-05, "loss": 0.0449, "step": 29370 }, { "grad_norm": 0.3910830318927765, "learning_rate": 8.461232977610061e-05, "loss": 0.0437, "step": 29380 }, { "grad_norm": 0.5653757452964783, "learning_rate": 8.46003954363245e-05, "loss": 0.042, "step": 29390 }, { "grad_norm": 0.4240376651287079, "learning_rate": 8.458845731270115e-05, "loss": 0.0462, "step": 29400 }, { "grad_norm": 0.46125173568725586, "learning_rate": 8.45765154065361e-05, "loss": 0.0366, "step": 29410 }, { "grad_norm": 0.24417994916439056, "learning_rate": 8.456456971913532e-05, "loss": 0.0311, "step": 29420 }, { "grad_norm": 0.2906039357185364, "learning_rate": 8.455262025180517e-05, "loss": 0.0357, "step": 29430 }, { "grad_norm": 0.36162614822387695, "learning_rate": 8.454066700585242e-05, "loss": 0.0368, "step": 29440 }, { "grad_norm": 0.3389662802219391, "learning_rate": 8.452870998258423e-05, "loss": 0.0336, "step": 29450 }, { "grad_norm": 0.4315180778503418, "learning_rate": 8.451674918330825e-05, "loss": 0.0298, "step": 29460 }, { "grad_norm": 0.435228168964386, "learning_rate": 8.450478460933246e-05, "loss": 0.0389, "step": 29470 }, { "grad_norm": 0.6768635511398315, "learning_rate": 8.449281626196532e-05, "loss": 0.0459, "step": 29480 }, { "grad_norm": 0.48891445994377136, "learning_rate": 8.448084414251564e-05, "loss": 0.046, "step": 29490 }, { "grad_norm": 0.3984028995037079, "learning_rate": 8.446886825229271e-05, "loss": 0.0473, "step": 29500 }, { "grad_norm": 0.41537296772003174, "learning_rate": 8.445688859260615e-05, "loss": 0.0404, "step": 29510 }, { "grad_norm": 0.4187251627445221, "learning_rate": 8.444490516476606e-05, "loss": 0.0354, "step": 29520 }, { "grad_norm": 0.39529865980148315, "learning_rate": 8.443291797008293e-05, "loss": 0.0451, "step": 29530 }, { "grad_norm": 0.38329029083251953, "learning_rate": 8.442092700986765e-05, "loss": 0.0515, "step": 29540 }, { "grad_norm": 0.38430002331733704, "learning_rate": 8.440893228543156e-05, "loss": 0.045, "step": 29550 }, { "grad_norm": 0.45228859782218933, "learning_rate": 8.439693379808638e-05, "loss": 0.0342, "step": 29560 }, { "grad_norm": 0.46751582622528076, "learning_rate": 8.43849315491442e-05, "loss": 0.0504, "step": 29570 }, { "grad_norm": 0.3935442268848419, "learning_rate": 8.437292553991763e-05, "loss": 0.0521, "step": 29580 }, { "grad_norm": 0.36638835072517395, "learning_rate": 8.436091577171959e-05, "loss": 0.0379, "step": 29590 }, { "grad_norm": 0.39115139842033386, "learning_rate": 8.434890224586347e-05, "loss": 0.0344, "step": 29600 }, { "grad_norm": 0.3890761435031891, "learning_rate": 8.433688496366303e-05, "loss": 0.0279, "step": 29610 }, { "grad_norm": 0.3932912051677704, "learning_rate": 8.432486392643248e-05, "loss": 0.0433, "step": 29620 }, { "grad_norm": 0.5113140344619751, "learning_rate": 8.431283913548643e-05, "loss": 0.0394, "step": 29630 }, { "grad_norm": 0.39244088530540466, "learning_rate": 8.430081059213985e-05, "loss": 0.0355, "step": 29640 }, { "grad_norm": 0.5075670480728149, "learning_rate": 8.428877829770823e-05, "loss": 0.0422, "step": 29650 }, { "grad_norm": 0.3981354236602783, "learning_rate": 8.427674225350735e-05, "loss": 0.0487, "step": 29660 }, { "grad_norm": 0.32736721634864807, "learning_rate": 8.426470246085347e-05, "loss": 0.044, "step": 29670 }, { "grad_norm": 0.5362238883972168, "learning_rate": 8.425265892106324e-05, "loss": 0.0478, "step": 29680 }, { "grad_norm": 0.37995675206184387, "learning_rate": 8.424061163545374e-05, "loss": 0.0423, "step": 29690 }, { "grad_norm": 0.3857271671295166, "learning_rate": 8.422856060534243e-05, "loss": 0.0399, "step": 29700 }, { "grad_norm": 0.6100767254829407, "learning_rate": 8.421650583204718e-05, "loss": 0.0457, "step": 29710 }, { "grad_norm": 0.3763330280780792, "learning_rate": 8.420444731688633e-05, "loss": 0.0355, "step": 29720 }, { "grad_norm": 0.44613537192344666, "learning_rate": 8.419238506117852e-05, "loss": 0.0534, "step": 29730 }, { "grad_norm": 0.5770670771598816, "learning_rate": 8.418031906624289e-05, "loss": 0.0375, "step": 29740 }, { "grad_norm": 0.35185572504997253, "learning_rate": 8.416824933339898e-05, "loss": 0.0403, "step": 29750 }, { "grad_norm": 0.4566335380077362, "learning_rate": 8.415617586396667e-05, "loss": 0.0377, "step": 29760 }, { "grad_norm": 0.4619210660457611, "learning_rate": 8.414409865926632e-05, "loss": 0.0411, "step": 29770 }, { "grad_norm": 0.551557719707489, "learning_rate": 8.413201772061867e-05, "loss": 0.0348, "step": 29780 }, { "grad_norm": 0.32131943106651306, "learning_rate": 8.411993304934488e-05, "loss": 0.031, "step": 29790 }, { "grad_norm": 0.2615411579608917, "learning_rate": 8.410784464676654e-05, "loss": 0.0346, "step": 29800 }, { "grad_norm": 0.34476473927497864, "learning_rate": 8.409575251420556e-05, "loss": 0.0344, "step": 29810 }, { "grad_norm": 0.33942246437072754, "learning_rate": 8.408365665298435e-05, "loss": 0.0387, "step": 29820 }, { "grad_norm": 0.6749054789543152, "learning_rate": 8.40715570644257e-05, "loss": 0.0522, "step": 29830 }, { "grad_norm": 0.47094807028770447, "learning_rate": 8.40594537498528e-05, "loss": 0.0472, "step": 29840 }, { "grad_norm": 0.5589328408241272, "learning_rate": 8.404734671058924e-05, "loss": 0.0477, "step": 29850 }, { "grad_norm": 0.5011661052703857, "learning_rate": 8.403523594795902e-05, "loss": 0.0386, "step": 29860 }, { "grad_norm": 0.3693699538707733, "learning_rate": 8.402312146328659e-05, "loss": 0.0367, "step": 29870 }, { "grad_norm": 0.5153509378433228, "learning_rate": 8.401100325789675e-05, "loss": 0.0449, "step": 29880 }, { "grad_norm": 0.4159702956676483, "learning_rate": 8.399888133311472e-05, "loss": 0.036, "step": 29890 }, { "grad_norm": 0.41286927461624146, "learning_rate": 8.398675569026613e-05, "loss": 0.0571, "step": 29900 }, { "grad_norm": 0.39004606008529663, "learning_rate": 8.397462633067705e-05, "loss": 0.044, "step": 29910 }, { "grad_norm": 1.59873366355896, "learning_rate": 8.396249325567392e-05, "loss": 0.0346, "step": 29920 }, { "grad_norm": 0.4715310335159302, "learning_rate": 8.395035646658357e-05, "loss": 0.0505, "step": 29930 }, { "grad_norm": 0.36959362030029297, "learning_rate": 8.39382159647333e-05, "loss": 0.0428, "step": 29940 }, { "grad_norm": 0.2659803032875061, "learning_rate": 8.392607175145075e-05, "loss": 0.0391, "step": 29950 }, { "grad_norm": 0.5300115942955017, "learning_rate": 8.3913923828064e-05, "loss": 0.0441, "step": 29960 }, { "grad_norm": 0.6668550968170166, "learning_rate": 8.390177219590152e-05, "loss": 0.0406, "step": 29970 }, { "grad_norm": 0.5122209191322327, "learning_rate": 8.388961685629222e-05, "loss": 0.0437, "step": 29980 }, { "grad_norm": 0.46400532126426697, "learning_rate": 8.387745781056536e-05, "loss": 0.0435, "step": 29990 }, { "grad_norm": 0.4295674264431, "learning_rate": 8.386529506005065e-05, "loss": 0.0405, "step": 30000 }, { "grad_norm": 0.48586416244506836, "learning_rate": 8.38531286060782e-05, "loss": 0.0413, "step": 30010 }, { "grad_norm": 0.43584319949150085, "learning_rate": 8.384095844997849e-05, "loss": 0.0328, "step": 30020 }, { "grad_norm": 0.5339021682739258, "learning_rate": 8.382878459308245e-05, "loss": 0.0286, "step": 30030 }, { "grad_norm": 0.414745569229126, "learning_rate": 8.381660703672138e-05, "loss": 0.0353, "step": 30040 }, { "grad_norm": 0.5063994526863098, "learning_rate": 8.380442578222702e-05, "loss": 0.0545, "step": 30050 }, { "grad_norm": 0.43110212683677673, "learning_rate": 8.379224083093146e-05, "loss": 0.0399, "step": 30060 }, { "grad_norm": 0.38241419196128845, "learning_rate": 8.378005218416727e-05, "loss": 0.0404, "step": 30070 }, { "grad_norm": 1.2049331665039062, "learning_rate": 8.376785984326735e-05, "loss": 0.0389, "step": 30080 }, { "grad_norm": 0.3867852985858917, "learning_rate": 8.375566380956506e-05, "loss": 0.0368, "step": 30090 }, { "grad_norm": 0.4919251501560211, "learning_rate": 8.374346408439411e-05, "loss": 0.0335, "step": 30100 }, { "grad_norm": 0.4182133972644806, "learning_rate": 8.373126066908868e-05, "loss": 0.0423, "step": 30110 }, { "grad_norm": 0.2965528964996338, "learning_rate": 8.371905356498326e-05, "loss": 0.0373, "step": 30120 }, { "grad_norm": 0.4499353766441345, "learning_rate": 8.370684277341288e-05, "loss": 0.0351, "step": 30130 }, { "grad_norm": 0.6490024924278259, "learning_rate": 8.369462829571282e-05, "loss": 0.0435, "step": 30140 }, { "grad_norm": 0.5169370770454407, "learning_rate": 8.36824101332189e-05, "loss": 0.0476, "step": 30150 }, { "grad_norm": 0.29047709703445435, "learning_rate": 8.367018828726721e-05, "loss": 0.0328, "step": 30160 }, { "grad_norm": 0.5080115795135498, "learning_rate": 8.365796275919438e-05, "loss": 0.0452, "step": 30170 }, { "grad_norm": 0.32969337701797485, "learning_rate": 8.364573355033734e-05, "loss": 0.0295, "step": 30180 }, { "grad_norm": 0.33744552731513977, "learning_rate": 8.363350066203346e-05, "loss": 0.0361, "step": 30190 }, { "grad_norm": 0.43103712797164917, "learning_rate": 8.362126409562053e-05, "loss": 0.0338, "step": 30200 }, { "grad_norm": 0.4236975908279419, "learning_rate": 8.360902385243667e-05, "loss": 0.0362, "step": 30210 }, { "grad_norm": 0.33900636434555054, "learning_rate": 8.359677993382052e-05, "loss": 0.0294, "step": 30220 }, { "grad_norm": 0.35659852623939514, "learning_rate": 8.358453234111103e-05, "loss": 0.037, "step": 30230 }, { "grad_norm": 0.29750341176986694, "learning_rate": 8.357228107564756e-05, "loss": 0.0484, "step": 30240 }, { "grad_norm": 0.48954784870147705, "learning_rate": 8.356002613876993e-05, "loss": 0.0322, "step": 30250 }, { "grad_norm": 0.34816473722457886, "learning_rate": 8.35477675318183e-05, "loss": 0.0293, "step": 30260 }, { "grad_norm": 0.40024054050445557, "learning_rate": 8.353550525613323e-05, "loss": 0.0429, "step": 30270 }, { "grad_norm": 0.42940667271614075, "learning_rate": 8.352323931305572e-05, "loss": 0.0524, "step": 30280 }, { "grad_norm": 0.5317011475563049, "learning_rate": 8.351096970392717e-05, "loss": 0.0378, "step": 30290 }, { "grad_norm": 0.3299408257007599, "learning_rate": 8.349869643008937e-05, "loss": 0.0343, "step": 30300 }, { "grad_norm": 0.3997906744480133, "learning_rate": 8.348641949288449e-05, "loss": 0.0466, "step": 30310 }, { "grad_norm": 0.4836638271808624, "learning_rate": 8.34741388936551e-05, "loss": 0.0401, "step": 30320 }, { "grad_norm": 0.4500376284122467, "learning_rate": 8.346185463374423e-05, "loss": 0.0398, "step": 30330 }, { "grad_norm": 0.3522997796535492, "learning_rate": 8.344956671449524e-05, "loss": 0.0373, "step": 30340 }, { "grad_norm": 0.49595925211906433, "learning_rate": 8.343727513725192e-05, "loss": 0.031, "step": 30350 }, { "grad_norm": 0.5163257718086243, "learning_rate": 8.342497990335847e-05, "loss": 0.0514, "step": 30360 }, { "grad_norm": 0.4660893976688385, "learning_rate": 8.341268101415946e-05, "loss": 0.0355, "step": 30370 }, { "grad_norm": 0.44998371601104736, "learning_rate": 8.34003784709999e-05, "loss": 0.0393, "step": 30380 }, { "grad_norm": 0.3556407392024994, "learning_rate": 8.338807227522516e-05, "loss": 0.0442, "step": 30390 }, { "grad_norm": 0.672804594039917, "learning_rate": 8.337576242818103e-05, "loss": 0.0474, "step": 30400 }, { "grad_norm": 0.305289089679718, "learning_rate": 8.336344893121372e-05, "loss": 0.0357, "step": 30410 }, { "grad_norm": 0.4262787997722626, "learning_rate": 8.335113178566977e-05, "loss": 0.0398, "step": 30420 }, { "grad_norm": 0.3062201142311096, "learning_rate": 8.33388109928962e-05, "loss": 0.0387, "step": 30430 }, { "grad_norm": 0.41626158356666565, "learning_rate": 8.33264865542404e-05, "loss": 0.0368, "step": 30440 }, { "grad_norm": 0.3104519546031952, "learning_rate": 8.331415847105013e-05, "loss": 0.0366, "step": 30450 }, { "grad_norm": 0.6351083517074585, "learning_rate": 8.330182674467357e-05, "loss": 0.0423, "step": 30460 }, { "grad_norm": 0.753291130065918, "learning_rate": 8.32894913764593e-05, "loss": 0.0314, "step": 30470 }, { "grad_norm": 0.4153631627559662, "learning_rate": 8.327715236775633e-05, "loss": 0.0416, "step": 30480 }, { "grad_norm": 0.4019031226634979, "learning_rate": 8.326480971991398e-05, "loss": 0.0365, "step": 30490 }, { "grad_norm": 0.30793991684913635, "learning_rate": 8.325246343428206e-05, "loss": 0.0387, "step": 30500 }, { "grad_norm": 0.4206324517726898, "learning_rate": 8.324011351221072e-05, "loss": 0.037, "step": 30510 }, { "grad_norm": 0.4947076439857483, "learning_rate": 8.322775995505057e-05, "loss": 0.0341, "step": 30520 }, { "grad_norm": 0.4230678379535675, "learning_rate": 8.321540276415254e-05, "loss": 0.0406, "step": 30530 }, { "grad_norm": 0.5553886890411377, "learning_rate": 8.320304194086798e-05, "loss": 0.0402, "step": 30540 }, { "grad_norm": 0.28436279296875, "learning_rate": 8.31906774865487e-05, "loss": 0.0309, "step": 30550 }, { "grad_norm": 0.2567470967769623, "learning_rate": 8.317830940254682e-05, "loss": 0.0307, "step": 30560 }, { "grad_norm": 0.4378148913383484, "learning_rate": 8.316593769021491e-05, "loss": 0.0424, "step": 30570 }, { "grad_norm": 0.431454062461853, "learning_rate": 8.315356235090592e-05, "loss": 0.0436, "step": 30580 }, { "grad_norm": 0.45683833956718445, "learning_rate": 8.314118338597319e-05, "loss": 0.0363, "step": 30590 }, { "grad_norm": 0.2534266710281372, "learning_rate": 8.312880079677048e-05, "loss": 0.041, "step": 30600 }, { "grad_norm": 0.3631252646446228, "learning_rate": 8.311641458465191e-05, "loss": 0.0394, "step": 30610 }, { "grad_norm": 0.43503183126449585, "learning_rate": 8.310402475097205e-05, "loss": 0.0469, "step": 30620 }, { "grad_norm": 0.35505160689353943, "learning_rate": 8.309163129708581e-05, "loss": 0.035, "step": 30630 }, { "grad_norm": 0.4422691762447357, "learning_rate": 8.307923422434852e-05, "loss": 0.0356, "step": 30640 }, { "grad_norm": 0.5181677937507629, "learning_rate": 8.30668335341159e-05, "loss": 0.0431, "step": 30650 }, { "grad_norm": 0.4716825485229492, "learning_rate": 8.305442922774408e-05, "loss": 0.036, "step": 30660 }, { "grad_norm": 0.31284964084625244, "learning_rate": 8.304202130658959e-05, "loss": 0.0358, "step": 30670 }, { "grad_norm": 0.4738682806491852, "learning_rate": 8.302960977200931e-05, "loss": 0.0369, "step": 30680 }, { "grad_norm": 0.5447167754173279, "learning_rate": 8.301719462536058e-05, "loss": 0.0359, "step": 30690 }, { "grad_norm": 0.4037087857723236, "learning_rate": 8.300477586800108e-05, "loss": 0.0474, "step": 30700 }, { "grad_norm": 0.350327730178833, "learning_rate": 8.299235350128892e-05, "loss": 0.0349, "step": 30710 }, { "grad_norm": 0.44049134850502014, "learning_rate": 8.297992752658258e-05, "loss": 0.0398, "step": 30720 }, { "grad_norm": 0.4919905364513397, "learning_rate": 8.296749794524094e-05, "loss": 0.0495, "step": 30730 }, { "grad_norm": 0.347993403673172, "learning_rate": 8.295506475862332e-05, "loss": 0.0375, "step": 30740 }, { "grad_norm": 0.30310070514678955, "learning_rate": 8.294262796808933e-05, "loss": 0.039, "step": 30750 }, { "grad_norm": 1.5065377950668335, "learning_rate": 8.293018757499909e-05, "loss": 0.0411, "step": 30760 }, { "grad_norm": 0.3832240700721741, "learning_rate": 8.291774358071305e-05, "loss": 0.0383, "step": 30770 }, { "grad_norm": 0.4227549731731415, "learning_rate": 8.290529598659205e-05, "loss": 0.0483, "step": 30780 }, { "grad_norm": 0.4254179000854492, "learning_rate": 8.289284479399738e-05, "loss": 0.0505, "step": 30790 }, { "grad_norm": 0.3956638276576996, "learning_rate": 8.288039000429064e-05, "loss": 0.0394, "step": 30800 }, { "grad_norm": 0.31913357973098755, "learning_rate": 8.286793161883388e-05, "loss": 0.0357, "step": 30810 }, { "grad_norm": 0.2995162606239319, "learning_rate": 8.285546963898954e-05, "loss": 0.0399, "step": 30820 }, { "grad_norm": 0.3947392702102661, "learning_rate": 8.284300406612044e-05, "loss": 0.0397, "step": 30830 }, { "grad_norm": 0.4795457124710083, "learning_rate": 8.283053490158978e-05, "loss": 0.0377, "step": 30840 }, { "grad_norm": 0.3489163815975189, "learning_rate": 8.28180621467612e-05, "loss": 0.0328, "step": 30850 }, { "grad_norm": 0.8080703020095825, "learning_rate": 8.280558580299867e-05, "loss": 0.041, "step": 30860 }, { "grad_norm": 0.38773804903030396, "learning_rate": 8.279310587166661e-05, "loss": 0.0362, "step": 30870 }, { "grad_norm": 0.37407612800598145, "learning_rate": 8.278062235412978e-05, "loss": 0.0353, "step": 30880 }, { "grad_norm": 0.39755040407180786, "learning_rate": 8.276813525175339e-05, "loss": 0.0423, "step": 30890 }, { "grad_norm": 0.40202564001083374, "learning_rate": 8.2755644565903e-05, "loss": 0.044, "step": 30900 }, { "grad_norm": 0.37797653675079346, "learning_rate": 8.274315029794454e-05, "loss": 0.0422, "step": 30910 }, { "grad_norm": 0.6371204257011414, "learning_rate": 8.273065244924443e-05, "loss": 0.047, "step": 30920 }, { "grad_norm": 0.42583394050598145, "learning_rate": 8.271815102116936e-05, "loss": 0.0415, "step": 30930 }, { "grad_norm": 0.4381627142429352, "learning_rate": 8.270564601508648e-05, "loss": 0.0358, "step": 30940 }, { "grad_norm": 0.44005534052848816, "learning_rate": 8.269313743236333e-05, "loss": 0.0448, "step": 30950 }, { "grad_norm": 0.33537042140960693, "learning_rate": 8.268062527436783e-05, "loss": 0.0351, "step": 30960 }, { "grad_norm": 0.429808109998703, "learning_rate": 8.266810954246828e-05, "loss": 0.0392, "step": 30970 }, { "grad_norm": 0.3684751093387604, "learning_rate": 8.265559023803338e-05, "loss": 0.0365, "step": 30980 }, { "grad_norm": 0.40109771490097046, "learning_rate": 8.264306736243223e-05, "loss": 0.0381, "step": 30990 }, { "grad_norm": 0.5139245986938477, "learning_rate": 8.263054091703432e-05, "loss": 0.0402, "step": 31000 }, { "grad_norm": 0.3126462697982788, "learning_rate": 8.26180109032095e-05, "loss": 0.0338, "step": 31010 }, { "grad_norm": 0.4432927966117859, "learning_rate": 8.260547732232807e-05, "loss": 0.0296, "step": 31020 }, { "grad_norm": 0.3795090913772583, "learning_rate": 8.259294017576063e-05, "loss": 0.0381, "step": 31030 }, { "grad_norm": 0.929949939250946, "learning_rate": 8.258039946487828e-05, "loss": 0.0449, "step": 31040 }, { "grad_norm": 0.34924814105033875, "learning_rate": 8.256785519105241e-05, "loss": 0.0353, "step": 31050 }, { "grad_norm": 1.3123183250427246, "learning_rate": 8.255530735565488e-05, "loss": 0.0327, "step": 31060 }, { "grad_norm": 0.38446733355522156, "learning_rate": 8.254275596005787e-05, "loss": 0.0323, "step": 31070 }, { "grad_norm": 0.43099913001060486, "learning_rate": 8.2530201005634e-05, "loss": 0.0375, "step": 31080 }, { "grad_norm": 0.36004751920700073, "learning_rate": 8.251764249375626e-05, "loss": 0.046, "step": 31090 }, { "grad_norm": 0.34088078141212463, "learning_rate": 8.250508042579803e-05, "loss": 0.0392, "step": 31100 }, { "grad_norm": 0.4767961800098419, "learning_rate": 8.249251480313307e-05, "loss": 0.0355, "step": 31110 }, { "grad_norm": 0.398281455039978, "learning_rate": 8.247994562713555e-05, "loss": 0.0356, "step": 31120 }, { "grad_norm": 0.32744383811950684, "learning_rate": 8.246737289917998e-05, "loss": 0.0355, "step": 31130 }, { "grad_norm": 0.6570373177528381, "learning_rate": 8.245479662064135e-05, "loss": 0.043, "step": 31140 }, { "grad_norm": 0.3851231634616852, "learning_rate": 8.244221679289496e-05, "loss": 0.0364, "step": 31150 }, { "grad_norm": 0.3151976466178894, "learning_rate": 8.242963341731652e-05, "loss": 0.0373, "step": 31160 }, { "grad_norm": 0.303340882062912, "learning_rate": 8.24170464952821e-05, "loss": 0.0378, "step": 31170 }, { "grad_norm": 0.43472644686698914, "learning_rate": 8.240445602816824e-05, "loss": 0.0469, "step": 31180 }, { "grad_norm": 0.7157021164894104, "learning_rate": 8.239186201735179e-05, "loss": 0.042, "step": 31190 }, { "grad_norm": 0.5808760523796082, "learning_rate": 8.237926446420998e-05, "loss": 0.033, "step": 31200 }, { "grad_norm": 0.4317631721496582, "learning_rate": 8.236666337012052e-05, "loss": 0.0383, "step": 31210 }, { "grad_norm": 0.3590041995048523, "learning_rate": 8.23540587364614e-05, "loss": 0.0327, "step": 31220 }, { "grad_norm": 0.388334721326828, "learning_rate": 8.234145056461107e-05, "loss": 0.038, "step": 31230 }, { "grad_norm": 0.36617517471313477, "learning_rate": 8.232883885594832e-05, "loss": 0.0384, "step": 31240 }, { "grad_norm": 0.3456323444843292, "learning_rate": 8.231622361185236e-05, "loss": 0.0321, "step": 31250 }, { "grad_norm": 0.4745836853981018, "learning_rate": 8.230360483370278e-05, "loss": 0.0592, "step": 31260 }, { "grad_norm": 0.3156106173992157, "learning_rate": 8.229098252287953e-05, "loss": 0.031, "step": 31270 }, { "grad_norm": 0.48111799359321594, "learning_rate": 8.2278356680763e-05, "loss": 0.0443, "step": 31280 }, { "grad_norm": 0.3995785415172577, "learning_rate": 8.22657273087339e-05, "loss": 0.045, "step": 31290 }, { "grad_norm": 0.6249603629112244, "learning_rate": 8.225309440817336e-05, "loss": 0.0548, "step": 31300 }, { "grad_norm": 0.2847881317138672, "learning_rate": 8.224045798046293e-05, "loss": 0.0606, "step": 31310 }, { "grad_norm": 0.3402876853942871, "learning_rate": 8.22278180269845e-05, "loss": 0.0367, "step": 31320 }, { "grad_norm": 0.5479305982589722, "learning_rate": 8.221517454912031e-05, "loss": 0.0387, "step": 31330 }, { "grad_norm": 0.4617674648761749, "learning_rate": 8.220252754825308e-05, "loss": 0.0436, "step": 31340 }, { "grad_norm": 0.7460483908653259, "learning_rate": 8.218987702576586e-05, "loss": 0.0403, "step": 31350 }, { "grad_norm": 0.3110179901123047, "learning_rate": 8.217722298304207e-05, "loss": 0.042, "step": 31360 }, { "grad_norm": 0.5663925409317017, "learning_rate": 8.216456542146557e-05, "loss": 0.0445, "step": 31370 }, { "grad_norm": 0.44776633381843567, "learning_rate": 8.215190434242055e-05, "loss": 0.0348, "step": 31380 }, { "grad_norm": 0.40410131216049194, "learning_rate": 8.213923974729161e-05, "loss": 0.0401, "step": 31390 }, { "grad_norm": 0.4659103751182556, "learning_rate": 8.212657163746373e-05, "loss": 0.0356, "step": 31400 }, { "grad_norm": 0.45570114254951477, "learning_rate": 8.211390001432227e-05, "loss": 0.033, "step": 31410 }, { "grad_norm": 0.5415859818458557, "learning_rate": 8.210122487925297e-05, "loss": 0.0319, "step": 31420 }, { "grad_norm": 0.4332423508167267, "learning_rate": 8.208854623364202e-05, "loss": 0.0381, "step": 31430 }, { "grad_norm": 0.35385796427726746, "learning_rate": 8.207586407887589e-05, "loss": 0.0329, "step": 31440 }, { "grad_norm": 0.29857945442199707, "learning_rate": 8.206317841634148e-05, "loss": 0.0334, "step": 31450 }, { "grad_norm": 0.891710102558136, "learning_rate": 8.205048924742609e-05, "loss": 0.0395, "step": 31460 }, { "grad_norm": 0.4374307692050934, "learning_rate": 8.203779657351738e-05, "loss": 0.0427, "step": 31470 }, { "grad_norm": 0.6235081553459167, "learning_rate": 8.20251003960034e-05, "loss": 0.0405, "step": 31480 }, { "grad_norm": 0.4278973639011383, "learning_rate": 8.201240071627258e-05, "loss": 0.0408, "step": 31490 }, { "grad_norm": 0.35485950112342834, "learning_rate": 8.199969753571377e-05, "loss": 0.0427, "step": 31500 }, { "grad_norm": 0.383038192987442, "learning_rate": 8.198699085571615e-05, "loss": 0.0367, "step": 31510 }, { "grad_norm": 0.48408442735671997, "learning_rate": 8.197428067766928e-05, "loss": 0.0444, "step": 31520 }, { "grad_norm": 0.447165846824646, "learning_rate": 8.196156700296316e-05, "loss": 0.0473, "step": 31530 }, { "grad_norm": 0.4929978549480438, "learning_rate": 8.194884983298814e-05, "loss": 0.0336, "step": 31540 }, { "grad_norm": 0.45425233244895935, "learning_rate": 8.193612916913491e-05, "loss": 0.038, "step": 31550 }, { "grad_norm": 0.34816640615463257, "learning_rate": 8.192340501279463e-05, "loss": 0.0404, "step": 31560 }, { "grad_norm": 0.46943849325180054, "learning_rate": 8.191067736535876e-05, "loss": 0.0357, "step": 31570 }, { "grad_norm": 0.40473541617393494, "learning_rate": 8.18979462282192e-05, "loss": 0.035, "step": 31580 }, { "grad_norm": 0.35536763072013855, "learning_rate": 8.188521160276819e-05, "loss": 0.0308, "step": 31590 }, { "grad_norm": 0.7826325297355652, "learning_rate": 8.187247349039837e-05, "loss": 0.0338, "step": 31600 }, { "grad_norm": 0.583611786365509, "learning_rate": 8.185973189250278e-05, "loss": 0.0454, "step": 31610 }, { "grad_norm": 0.40753212571144104, "learning_rate": 8.184698681047482e-05, "loss": 0.0386, "step": 31620 }, { "grad_norm": 0.31700292229652405, "learning_rate": 8.183423824570827e-05, "loss": 0.0363, "step": 31630 }, { "grad_norm": 0.3671077787876129, "learning_rate": 8.182148619959725e-05, "loss": 0.0404, "step": 31640 }, { "grad_norm": 0.32739976048469543, "learning_rate": 8.180873067353636e-05, "loss": 0.0398, "step": 31650 }, { "grad_norm": 0.670499324798584, "learning_rate": 8.179597166892052e-05, "loss": 0.0378, "step": 31660 }, { "grad_norm": 0.3951897621154785, "learning_rate": 8.178320918714501e-05, "loss": 0.0374, "step": 31670 }, { "grad_norm": 0.39014118909835815, "learning_rate": 8.177044322960554e-05, "loss": 0.0368, "step": 31680 }, { "grad_norm": 0.36867567896842957, "learning_rate": 8.175767379769816e-05, "loss": 0.0332, "step": 31690 }, { "grad_norm": 0.36666548252105713, "learning_rate": 8.174490089281932e-05, "loss": 0.0403, "step": 31700 }, { "grad_norm": 0.41385409235954285, "learning_rate": 8.173212451636584e-05, "loss": 0.0376, "step": 31710 }, { "grad_norm": 0.38045310974121094, "learning_rate": 8.171934466973493e-05, "loss": 0.0391, "step": 31720 }, { "grad_norm": 0.3778928816318512, "learning_rate": 8.170656135432418e-05, "loss": 0.0329, "step": 31730 }, { "grad_norm": 0.3828009366989136, "learning_rate": 8.169377457153155e-05, "loss": 0.039, "step": 31740 }, { "grad_norm": 0.27996042370796204, "learning_rate": 8.168098432275539e-05, "loss": 0.0329, "step": 31750 }, { "grad_norm": 0.46969833970069885, "learning_rate": 8.166819060939442e-05, "loss": 0.0486, "step": 31760 }, { "grad_norm": 0.3154881000518799, "learning_rate": 8.165539343284772e-05, "loss": 0.0335, "step": 31770 }, { "grad_norm": 0.3240743577480316, "learning_rate": 8.16425927945148e-05, "loss": 0.0431, "step": 31780 }, { "grad_norm": 0.25884050130844116, "learning_rate": 8.162978869579551e-05, "loss": 0.0473, "step": 31790 }, { "grad_norm": 0.41827720403671265, "learning_rate": 8.161698113809007e-05, "loss": 0.0397, "step": 31800 }, { "grad_norm": 0.3387698829174042, "learning_rate": 8.160417012279911e-05, "loss": 0.0449, "step": 31810 }, { "grad_norm": 0.43504273891448975, "learning_rate": 8.159135565132363e-05, "loss": 0.0352, "step": 31820 }, { "grad_norm": 0.28243085741996765, "learning_rate": 8.157853772506498e-05, "loss": 0.0483, "step": 31830 }, { "grad_norm": 0.41646817326545715, "learning_rate": 8.156571634542494e-05, "loss": 0.0464, "step": 31840 }, { "grad_norm": 0.39152729511260986, "learning_rate": 8.15528915138056e-05, "loss": 0.0392, "step": 31850 }, { "grad_norm": 0.3647712171077728, "learning_rate": 8.154006323160949e-05, "loss": 0.0293, "step": 31860 }, { "grad_norm": 0.3760889768600464, "learning_rate": 8.152723150023949e-05, "loss": 0.0387, "step": 31870 }, { "grad_norm": 0.3664554953575134, "learning_rate": 8.151439632109886e-05, "loss": 0.0359, "step": 31880 }, { "grad_norm": 0.4725692570209503, "learning_rate": 8.150155769559122e-05, "loss": 0.0527, "step": 31890 }, { "grad_norm": 0.4233403503894806, "learning_rate": 8.148871562512058e-05, "loss": 0.0339, "step": 31900 }, { "grad_norm": 0.3680943250656128, "learning_rate": 8.147587011109136e-05, "loss": 0.044, "step": 31910 }, { "grad_norm": 0.49502527713775635, "learning_rate": 8.14630211549083e-05, "loss": 0.0534, "step": 31920 }, { "grad_norm": 0.28928878903388977, "learning_rate": 8.145016875797655e-05, "loss": 0.0309, "step": 31930 }, { "grad_norm": 0.2873567044734955, "learning_rate": 8.143731292170164e-05, "loss": 0.0323, "step": 31940 }, { "grad_norm": 0.37036624550819397, "learning_rate": 8.142445364748944e-05, "loss": 0.044, "step": 31950 }, { "grad_norm": 0.3885038197040558, "learning_rate": 8.141159093674624e-05, "loss": 0.0418, "step": 31960 }, { "grad_norm": 0.5149439573287964, "learning_rate": 8.139872479087869e-05, "loss": 0.0342, "step": 31970 }, { "grad_norm": 0.5516887307167053, "learning_rate": 8.13858552112938e-05, "loss": 0.0358, "step": 31980 }, { "grad_norm": 0.716361939907074, "learning_rate": 8.137298219939895e-05, "loss": 0.0572, "step": 31990 }, { "grad_norm": 0.3642943203449249, "learning_rate": 8.136010575660196e-05, "loss": 0.0298, "step": 32000 }, { "grad_norm": 0.5495643019676208, "learning_rate": 8.134722588431095e-05, "loss": 0.0483, "step": 32010 }, { "grad_norm": 0.8902769684791565, "learning_rate": 8.133434258393444e-05, "loss": 0.0405, "step": 32020 }, { "grad_norm": 0.46102556586265564, "learning_rate": 8.132145585688134e-05, "loss": 0.0408, "step": 32030 }, { "grad_norm": 0.5217888951301575, "learning_rate": 8.130856570456093e-05, "loss": 0.045, "step": 32040 }, { "grad_norm": 0.3617262542247772, "learning_rate": 8.129567212838283e-05, "loss": 0.0376, "step": 32050 }, { "grad_norm": 0.4359557330608368, "learning_rate": 8.128277512975708e-05, "loss": 0.0434, "step": 32060 }, { "grad_norm": 0.4057367444038391, "learning_rate": 8.126987471009408e-05, "loss": 0.0401, "step": 32070 }, { "grad_norm": 0.5496452450752258, "learning_rate": 8.125697087080459e-05, "loss": 0.029, "step": 32080 }, { "grad_norm": 0.4008631706237793, "learning_rate": 8.124406361329976e-05, "loss": 0.0466, "step": 32090 }, { "grad_norm": 0.43189188838005066, "learning_rate": 8.123115293899111e-05, "loss": 0.0351, "step": 32100 }, { "grad_norm": 0.3775651454925537, "learning_rate": 8.121823884929055e-05, "loss": 0.0403, "step": 32110 }, { "grad_norm": 0.3598833978176117, "learning_rate": 8.12053213456103e-05, "loss": 0.0326, "step": 32120 }, { "grad_norm": 0.5818036794662476, "learning_rate": 8.119240042936303e-05, "loss": 0.0422, "step": 32130 }, { "grad_norm": 0.4180188775062561, "learning_rate": 8.117947610196175e-05, "loss": 0.0349, "step": 32140 }, { "grad_norm": 0.48866844177246094, "learning_rate": 8.116654836481982e-05, "loss": 0.0332, "step": 32150 }, { "grad_norm": 0.315908282995224, "learning_rate": 8.115361721935106e-05, "loss": 0.0306, "step": 32160 }, { "grad_norm": 0.3795575499534607, "learning_rate": 8.114068266696953e-05, "loss": 0.0376, "step": 32170 }, { "grad_norm": 0.37110477685928345, "learning_rate": 8.112774470908978e-05, "loss": 0.0301, "step": 32180 }, { "grad_norm": 0.35432082414627075, "learning_rate": 8.111480334712665e-05, "loss": 0.0303, "step": 32190 }, { "grad_norm": 0.32560214400291443, "learning_rate": 8.110185858249542e-05, "loss": 0.0371, "step": 32200 }, { "grad_norm": 0.42692047357559204, "learning_rate": 8.108891041661168e-05, "loss": 0.0405, "step": 32210 }, { "grad_norm": 0.5899509787559509, "learning_rate": 8.107595885089146e-05, "loss": 0.0398, "step": 32220 }, { "grad_norm": 0.3980068862438202, "learning_rate": 8.106300388675108e-05, "loss": 0.0326, "step": 32230 }, { "grad_norm": 0.44032126665115356, "learning_rate": 8.105004552560731e-05, "loss": 0.0431, "step": 32240 }, { "grad_norm": 0.560899019241333, "learning_rate": 8.103708376887724e-05, "loss": 0.0398, "step": 32250 }, { "grad_norm": 0.47836652398109436, "learning_rate": 8.102411861797836e-05, "loss": 0.0413, "step": 32260 }, { "grad_norm": 0.37181568145751953, "learning_rate": 8.101115007432851e-05, "loss": 0.0428, "step": 32270 }, { "grad_norm": 0.34115830063819885, "learning_rate": 8.09981781393459e-05, "loss": 0.0294, "step": 32280 }, { "grad_norm": 0.41351306438446045, "learning_rate": 8.098520281444915e-05, "loss": 0.032, "step": 32290 }, { "grad_norm": 0.4057536721229553, "learning_rate": 8.09722241010572e-05, "loss": 0.0344, "step": 32300 }, { "grad_norm": 0.301571786403656, "learning_rate": 8.095924200058939e-05, "loss": 0.0453, "step": 32310 }, { "grad_norm": 0.3529365360736847, "learning_rate": 8.094625651446541e-05, "loss": 0.0273, "step": 32320 }, { "grad_norm": 0.31449204683303833, "learning_rate": 8.093326764410536e-05, "loss": 0.0292, "step": 32330 }, { "grad_norm": 0.39378905296325684, "learning_rate": 8.092027539092966e-05, "loss": 0.0334, "step": 32340 }, { "grad_norm": 0.6658897995948792, "learning_rate": 8.090727975635913e-05, "loss": 0.0344, "step": 32350 }, { "grad_norm": 0.31832653284072876, "learning_rate": 8.089428074181497e-05, "loss": 0.0386, "step": 32360 }, { "grad_norm": 0.3679181933403015, "learning_rate": 8.088127834871871e-05, "loss": 0.035, "step": 32370 }, { "grad_norm": 0.30988913774490356, "learning_rate": 8.086827257849226e-05, "loss": 0.0358, "step": 32380 }, { "grad_norm": 0.3468816578388214, "learning_rate": 8.085526343255795e-05, "loss": 0.0367, "step": 32390 }, { "grad_norm": 0.4061000645160675, "learning_rate": 8.084225091233842e-05, "loss": 0.0347, "step": 32400 }, { "grad_norm": 0.5291264057159424, "learning_rate": 8.082923501925668e-05, "loss": 0.0408, "step": 32410 }, { "grad_norm": 0.5661615133285522, "learning_rate": 8.081621575473617e-05, "loss": 0.0377, "step": 32420 }, { "grad_norm": 0.5091941356658936, "learning_rate": 8.080319312020064e-05, "loss": 0.0448, "step": 32430 }, { "grad_norm": 0.36037206649780273, "learning_rate": 8.079016711707421e-05, "loss": 0.0324, "step": 32440 }, { "grad_norm": 0.27610012888908386, "learning_rate": 8.077713774678139e-05, "loss": 0.0388, "step": 32450 }, { "grad_norm": 0.34883761405944824, "learning_rate": 8.076410501074707e-05, "loss": 0.0353, "step": 32460 }, { "grad_norm": 0.40228503942489624, "learning_rate": 8.075106891039647e-05, "loss": 0.0417, "step": 32470 }, { "grad_norm": 0.29534342885017395, "learning_rate": 8.073802944715523e-05, "loss": 0.0423, "step": 32480 }, { "grad_norm": 0.44631752371788025, "learning_rate": 8.072498662244929e-05, "loss": 0.0471, "step": 32490 }, { "grad_norm": 0.4665133059024811, "learning_rate": 8.0711940437705e-05, "loss": 0.0378, "step": 32500 }, { "grad_norm": 0.3621627986431122, "learning_rate": 8.06988908943491e-05, "loss": 0.0276, "step": 32510 }, { "grad_norm": 0.4656003415584564, "learning_rate": 8.068583799380863e-05, "loss": 0.0356, "step": 32520 }, { "grad_norm": 0.4152032136917114, "learning_rate": 8.067278173751104e-05, "loss": 0.0375, "step": 32530 }, { "grad_norm": 0.541197657585144, "learning_rate": 8.065972212688417e-05, "loss": 0.0365, "step": 32540 }, { "grad_norm": 0.39300569891929626, "learning_rate": 8.064665916335618e-05, "loss": 0.0369, "step": 32550 }, { "grad_norm": 0.39119410514831543, "learning_rate": 8.063359284835564e-05, "loss": 0.0365, "step": 32560 }, { "grad_norm": 0.3017108738422394, "learning_rate": 8.062052318331142e-05, "loss": 0.0351, "step": 32570 }, { "grad_norm": 0.5323760509490967, "learning_rate": 8.060745016965283e-05, "loss": 0.0397, "step": 32580 }, { "grad_norm": 0.820115864276886, "learning_rate": 8.059437380880952e-05, "loss": 0.0318, "step": 32590 }, { "grad_norm": 0.43870946764945984, "learning_rate": 8.058129410221146e-05, "loss": 0.0326, "step": 32600 }, { "grad_norm": 0.28631821274757385, "learning_rate": 8.056821105128908e-05, "loss": 0.0377, "step": 32610 }, { "grad_norm": 0.3210770785808563, "learning_rate": 8.05551246574731e-05, "loss": 0.0332, "step": 32620 }, { "grad_norm": 0.4180472791194916, "learning_rate": 8.05420349221946e-05, "loss": 0.0337, "step": 32630 }, { "grad_norm": 0.35998672246932983, "learning_rate": 8.05289418468851e-05, "loss": 0.0331, "step": 32640 }, { "grad_norm": 0.3371773064136505, "learning_rate": 8.051584543297642e-05, "loss": 0.0271, "step": 32650 }, { "grad_norm": 0.3714428246021271, "learning_rate": 8.050274568190074e-05, "loss": 0.0339, "step": 32660 }, { "grad_norm": 0.2954934239387512, "learning_rate": 8.048964259509067e-05, "loss": 0.0496, "step": 32670 }, { "grad_norm": 0.37521421909332275, "learning_rate": 8.047653617397914e-05, "loss": 0.0392, "step": 32680 }, { "grad_norm": 0.46700039505958557, "learning_rate": 8.046342641999941e-05, "loss": 0.0405, "step": 32690 }, { "grad_norm": 0.5047705173492432, "learning_rate": 8.045031333458517e-05, "loss": 0.0403, "step": 32700 }, { "grad_norm": 0.4335077106952667, "learning_rate": 8.043719691917047e-05, "loss": 0.0448, "step": 32710 }, { "grad_norm": 0.28860607743263245, "learning_rate": 8.042407717518966e-05, "loss": 0.0305, "step": 32720 }, { "grad_norm": 0.27758780121803284, "learning_rate": 8.041095410407751e-05, "loss": 0.0331, "step": 32730 }, { "grad_norm": 0.4701482355594635, "learning_rate": 8.039782770726913e-05, "loss": 0.0415, "step": 32740 }, { "grad_norm": 0.35512056946754456, "learning_rate": 8.038469798620004e-05, "loss": 0.0359, "step": 32750 }, { "grad_norm": 0.5316081047058105, "learning_rate": 8.037156494230604e-05, "loss": 0.0383, "step": 32760 }, { "grad_norm": 0.29447540640830994, "learning_rate": 8.035842857702338e-05, "loss": 0.0394, "step": 32770 }, { "grad_norm": 0.48231393098831177, "learning_rate": 8.03452888917886e-05, "loss": 0.043, "step": 32780 }, { "grad_norm": 0.21783816814422607, "learning_rate": 8.033214588803866e-05, "loss": 0.0291, "step": 32790 }, { "grad_norm": 0.380781888961792, "learning_rate": 8.031899956721083e-05, "loss": 0.0325, "step": 32800 }, { "grad_norm": 0.36119526624679565, "learning_rate": 8.030584993074282e-05, "loss": 0.0298, "step": 32810 }, { "grad_norm": 0.4056541621685028, "learning_rate": 8.02926969800726e-05, "loss": 0.0361, "step": 32820 }, { "grad_norm": 0.40397387742996216, "learning_rate": 8.027954071663859e-05, "loss": 0.0441, "step": 32830 }, { "grad_norm": 0.3912581205368042, "learning_rate": 8.026638114187954e-05, "loss": 0.031, "step": 32840 }, { "grad_norm": 0.33285948634147644, "learning_rate": 8.025321825723456e-05, "loss": 0.0388, "step": 32850 }, { "grad_norm": 0.3159017562866211, "learning_rate": 8.02400520641431e-05, "loss": 0.0409, "step": 32860 }, { "grad_norm": 0.5336129665374756, "learning_rate": 8.022688256404501e-05, "loss": 0.0332, "step": 32870 }, { "grad_norm": 0.5110721588134766, "learning_rate": 8.02137097583805e-05, "loss": 0.0334, "step": 32880 }, { "grad_norm": 0.362362802028656, "learning_rate": 8.02005336485901e-05, "loss": 0.0261, "step": 32890 }, { "grad_norm": 0.49828025698661804, "learning_rate": 8.018735423611476e-05, "loss": 0.0438, "step": 32900 }, { "grad_norm": 0.41297873854637146, "learning_rate": 8.017417152239574e-05, "loss": 0.0344, "step": 32910 }, { "grad_norm": 0.3391721844673157, "learning_rate": 8.01609855088747e-05, "loss": 0.0382, "step": 32920 }, { "grad_norm": 0.2828434407711029, "learning_rate": 8.014779619699362e-05, "loss": 0.0276, "step": 32930 }, { "grad_norm": 0.29226699471473694, "learning_rate": 8.013460358819489e-05, "loss": 0.033, "step": 32940 }, { "grad_norm": 0.3296555280685425, "learning_rate": 8.01214076839212e-05, "loss": 0.0414, "step": 32950 }, { "grad_norm": 0.4943397045135498, "learning_rate": 8.010820848561565e-05, "loss": 0.0396, "step": 32960 }, { "grad_norm": 0.33294418454170227, "learning_rate": 8.009500599472171e-05, "loss": 0.0363, "step": 32970 }, { "grad_norm": 0.40679264068603516, "learning_rate": 8.008180021268314e-05, "loss": 0.0363, "step": 32980 }, { "grad_norm": 0.3569561541080475, "learning_rate": 8.006859114094414e-05, "loss": 0.0305, "step": 32990 }, { "grad_norm": 0.39974090456962585, "learning_rate": 8.005537878094921e-05, "loss": 0.0379, "step": 33000 }, { "grad_norm": 0.47211605310440063, "learning_rate": 8.004216313414323e-05, "loss": 0.0381, "step": 33010 }, { "grad_norm": 0.2623385190963745, "learning_rate": 8.002894420197149e-05, "loss": 0.0272, "step": 33020 }, { "grad_norm": 0.3252711892127991, "learning_rate": 8.001572198587954e-05, "loss": 0.0355, "step": 33030 }, { "grad_norm": 0.2960684299468994, "learning_rate": 8.000249648731338e-05, "loss": 0.0308, "step": 33040 }, { "grad_norm": 0.526465892791748, "learning_rate": 7.998926770771928e-05, "loss": 0.0511, "step": 33050 }, { "grad_norm": 0.4690508246421814, "learning_rate": 7.997603564854397e-05, "loss": 0.0421, "step": 33060 }, { "grad_norm": 0.4476502537727356, "learning_rate": 7.996280031123448e-05, "loss": 0.0443, "step": 33070 }, { "grad_norm": 0.3525829315185547, "learning_rate": 7.994956169723818e-05, "loss": 0.0536, "step": 33080 }, { "grad_norm": 0.41975101828575134, "learning_rate": 7.993631980800285e-05, "loss": 0.0368, "step": 33090 }, { "grad_norm": 0.3519164025783539, "learning_rate": 7.992307464497659e-05, "loss": 0.0451, "step": 33100 }, { "grad_norm": 0.3368517756462097, "learning_rate": 7.990982620960787e-05, "loss": 0.0329, "step": 33110 }, { "grad_norm": 0.29212090373039246, "learning_rate": 7.989657450334554e-05, "loss": 0.0347, "step": 33120 }, { "grad_norm": 0.5893731713294983, "learning_rate": 7.988331952763877e-05, "loss": 0.0421, "step": 33130 }, { "grad_norm": 0.8463014960289001, "learning_rate": 7.987006128393709e-05, "loss": 0.0572, "step": 33140 }, { "grad_norm": 0.3617856502532959, "learning_rate": 7.985679977369043e-05, "loss": 0.0427, "step": 33150 }, { "grad_norm": 0.4452109932899475, "learning_rate": 7.984353499834902e-05, "loss": 0.0387, "step": 33160 }, { "grad_norm": 0.5644485354423523, "learning_rate": 7.983026695936351e-05, "loss": 0.0394, "step": 33170 }, { "grad_norm": 0.48823457956314087, "learning_rate": 7.981699565818486e-05, "loss": 0.0359, "step": 33180 }, { "grad_norm": 0.4921768307685852, "learning_rate": 7.980372109626437e-05, "loss": 0.034, "step": 33190 }, { "grad_norm": 0.3636119067668915, "learning_rate": 7.979044327505375e-05, "loss": 0.0313, "step": 33200 }, { "grad_norm": 0.49461472034454346, "learning_rate": 7.977716219600506e-05, "loss": 0.0391, "step": 33210 }, { "grad_norm": 0.39226365089416504, "learning_rate": 7.97638778605707e-05, "loss": 0.0404, "step": 33220 }, { "grad_norm": 0.4250522255897522, "learning_rate": 7.975059027020338e-05, "loss": 0.0459, "step": 33230 }, { "grad_norm": 0.3539811670780182, "learning_rate": 7.973729942635623e-05, "loss": 0.0358, "step": 33240 }, { "grad_norm": 0.32326894998550415, "learning_rate": 7.972400533048273e-05, "loss": 0.044, "step": 33250 }, { "grad_norm": 0.35565537214279175, "learning_rate": 7.97107079840367e-05, "loss": 0.0399, "step": 33260 }, { "grad_norm": 0.5895673632621765, "learning_rate": 7.969740738847231e-05, "loss": 0.0406, "step": 33270 }, { "grad_norm": 0.34498533606529236, "learning_rate": 7.968410354524411e-05, "loss": 0.0339, "step": 33280 }, { "grad_norm": 0.4311908483505249, "learning_rate": 7.967079645580697e-05, "loss": 0.0389, "step": 33290 }, { "grad_norm": 0.40720507502555847, "learning_rate": 7.965748612161612e-05, "loss": 0.0332, "step": 33300 }, { "grad_norm": 0.4633391201496124, "learning_rate": 7.96441725441272e-05, "loss": 0.0341, "step": 33310 }, { "grad_norm": 0.38990217447280884, "learning_rate": 7.963085572479614e-05, "loss": 0.0338, "step": 33320 }, { "grad_norm": 0.3400762379169464, "learning_rate": 7.961753566507924e-05, "loss": 0.0327, "step": 33330 }, { "grad_norm": 0.2505190074443817, "learning_rate": 7.960421236643316e-05, "loss": 0.0413, "step": 33340 }, { "grad_norm": 0.2993389964103699, "learning_rate": 7.959088583031496e-05, "loss": 0.0462, "step": 33350 }, { "grad_norm": 0.4765803813934326, "learning_rate": 7.957755605818194e-05, "loss": 0.0334, "step": 33360 }, { "grad_norm": 0.3610600531101227, "learning_rate": 7.956422305149185e-05, "loss": 0.0491, "step": 33370 }, { "grad_norm": 0.35488560795783997, "learning_rate": 7.95508868117028e-05, "loss": 0.0414, "step": 33380 }, { "grad_norm": 0.4213213324546814, "learning_rate": 7.953754734027318e-05, "loss": 0.041, "step": 33390 }, { "grad_norm": 0.3091445565223694, "learning_rate": 7.952420463866182e-05, "loss": 0.0283, "step": 33400 }, { "grad_norm": 0.3475463092327118, "learning_rate": 7.951085870832782e-05, "loss": 0.0402, "step": 33410 }, { "grad_norm": 0.3745726943016052, "learning_rate": 7.949750955073067e-05, "loss": 0.0367, "step": 33420 }, { "grad_norm": 0.31653785705566406, "learning_rate": 7.948415716733022e-05, "loss": 0.0415, "step": 33430 }, { "grad_norm": 0.3539597690105438, "learning_rate": 7.947080155958669e-05, "loss": 0.0395, "step": 33440 }, { "grad_norm": 0.555534839630127, "learning_rate": 7.94574427289606e-05, "loss": 0.0337, "step": 33450 }, { "grad_norm": 0.34785032272338867, "learning_rate": 7.944408067691284e-05, "loss": 0.0352, "step": 33460 }, { "grad_norm": 0.23229052126407623, "learning_rate": 7.943071540490473e-05, "loss": 0.0308, "step": 33470 }, { "grad_norm": 0.38822442293167114, "learning_rate": 7.94173469143978e-05, "loss": 0.0302, "step": 33480 }, { "grad_norm": 0.28939902782440186, "learning_rate": 7.940397520685406e-05, "loss": 0.0262, "step": 33490 }, { "grad_norm": 0.44523343443870544, "learning_rate": 7.939060028373577e-05, "loss": 0.0332, "step": 33500 }, { "grad_norm": 0.2953184247016907, "learning_rate": 7.937722214650565e-05, "loss": 0.0403, "step": 33510 }, { "grad_norm": 0.35121941566467285, "learning_rate": 7.936384079662666e-05, "loss": 0.0313, "step": 33520 }, { "grad_norm": 0.3283657729625702, "learning_rate": 7.93504562355622e-05, "loss": 0.036, "step": 33530 }, { "grad_norm": 0.3845415711402893, "learning_rate": 7.933706846477599e-05, "loss": 0.0363, "step": 33540 }, { "grad_norm": 0.4392554759979248, "learning_rate": 7.932367748573206e-05, "loss": 0.0415, "step": 33550 }, { "grad_norm": 0.4150334298610687, "learning_rate": 7.931028329989485e-05, "loss": 0.036, "step": 33560 }, { "grad_norm": 0.4165506362915039, "learning_rate": 7.929688590872913e-05, "loss": 0.0395, "step": 33570 }, { "grad_norm": 0.41557472944259644, "learning_rate": 7.928348531370003e-05, "loss": 0.0275, "step": 33580 }, { "grad_norm": 0.37980708479881287, "learning_rate": 7.927008151627297e-05, "loss": 0.0313, "step": 33590 }, { "grad_norm": 0.2987217605113983, "learning_rate": 7.925667451791383e-05, "loss": 0.0332, "step": 33600 }, { "grad_norm": 0.46613654494285583, "learning_rate": 7.924326432008874e-05, "loss": 0.0369, "step": 33610 }, { "grad_norm": 0.37891003489494324, "learning_rate": 7.922985092426422e-05, "loss": 0.0357, "step": 33620 }, { "grad_norm": 0.4421810209751129, "learning_rate": 7.921643433190717e-05, "loss": 0.0365, "step": 33630 }, { "grad_norm": 0.5930172801017761, "learning_rate": 7.920301454448478e-05, "loss": 0.0419, "step": 33640 }, { "grad_norm": 0.37988948822021484, "learning_rate": 7.918959156346461e-05, "loss": 0.0318, "step": 33650 }, { "grad_norm": 0.43126168847084045, "learning_rate": 7.91761653903146e-05, "loss": 0.0344, "step": 33660 }, { "grad_norm": 0.4555242359638214, "learning_rate": 7.916273602650302e-05, "loss": 0.0399, "step": 33670 }, { "grad_norm": 0.35368043184280396, "learning_rate": 7.914930347349847e-05, "loss": 0.0293, "step": 33680 }, { "grad_norm": 0.2690638601779938, "learning_rate": 7.913586773276992e-05, "loss": 0.039, "step": 33690 }, { "grad_norm": 0.5014005899429321, "learning_rate": 7.912242880578667e-05, "loss": 0.0326, "step": 33700 }, { "grad_norm": 0.49703168869018555, "learning_rate": 7.910898669401839e-05, "loss": 0.0408, "step": 33710 }, { "grad_norm": 0.42327171564102173, "learning_rate": 7.909554139893511e-05, "loss": 0.0383, "step": 33720 }, { "grad_norm": 0.528640627861023, "learning_rate": 7.908209292200715e-05, "loss": 0.0398, "step": 33730 }, { "grad_norm": 0.3547811806201935, "learning_rate": 7.906864126470523e-05, "loss": 0.0349, "step": 33740 }, { "grad_norm": 0.530571699142456, "learning_rate": 7.905518642850041e-05, "loss": 0.0405, "step": 33750 }, { "grad_norm": 0.502318799495697, "learning_rate": 7.904172841486409e-05, "loss": 0.0506, "step": 33760 }, { "grad_norm": 0.5932299494743347, "learning_rate": 7.902826722526801e-05, "loss": 0.0454, "step": 33770 }, { "grad_norm": 0.28499534726142883, "learning_rate": 7.901480286118427e-05, "loss": 0.0314, "step": 33780 }, { "grad_norm": 0.42391708493232727, "learning_rate": 7.900133532408531e-05, "loss": 0.0415, "step": 33790 }, { "grad_norm": 0.410667359828949, "learning_rate": 7.898786461544395e-05, "loss": 0.0353, "step": 33800 }, { "grad_norm": 0.2707500755786896, "learning_rate": 7.897439073673325e-05, "loss": 0.0302, "step": 33810 }, { "grad_norm": 0.3815741539001465, "learning_rate": 7.896091368942677e-05, "loss": 0.0335, "step": 33820 }, { "grad_norm": 0.844696044921875, "learning_rate": 7.894743347499832e-05, "loss": 0.0372, "step": 33830 }, { "grad_norm": 0.411822646856308, "learning_rate": 7.893395009492203e-05, "loss": 0.0333, "step": 33840 }, { "grad_norm": 0.29494988918304443, "learning_rate": 7.892046355067248e-05, "loss": 0.036, "step": 33850 }, { "grad_norm": 0.7295587658882141, "learning_rate": 7.890697384372451e-05, "loss": 0.0364, "step": 33860 }, { "grad_norm": 0.5102230906486511, "learning_rate": 7.889348097555336e-05, "loss": 0.0309, "step": 33870 }, { "grad_norm": 0.43439456820487976, "learning_rate": 7.887998494763455e-05, "loss": 0.0336, "step": 33880 }, { "grad_norm": 0.4313559830188751, "learning_rate": 7.886648576144404e-05, "loss": 0.0305, "step": 33890 }, { "grad_norm": 0.41216400265693665, "learning_rate": 7.885298341845802e-05, "loss": 0.0354, "step": 33900 }, { "grad_norm": 0.4939497709274292, "learning_rate": 7.883947792015311e-05, "loss": 0.0445, "step": 33910 }, { "grad_norm": 0.8957787156105042, "learning_rate": 7.882596926800628e-05, "loss": 0.0618, "step": 33920 }, { "grad_norm": 0.580220639705658, "learning_rate": 7.881245746349477e-05, "loss": 0.0373, "step": 33930 }, { "grad_norm": 0.39498481154441833, "learning_rate": 7.879894250809623e-05, "loss": 0.0395, "step": 33940 }, { "grad_norm": 0.37725210189819336, "learning_rate": 7.878542440328865e-05, "loss": 0.0504, "step": 33950 }, { "grad_norm": 0.3193315267562866, "learning_rate": 7.877190315055031e-05, "loss": 0.0327, "step": 33960 }, { "grad_norm": 0.3457162082195282, "learning_rate": 7.875837875135991e-05, "loss": 0.0291, "step": 33970 }, { "grad_norm": 0.4329301416873932, "learning_rate": 7.874485120719646e-05, "loss": 0.0271, "step": 33980 }, { "grad_norm": 0.3136191666126251, "learning_rate": 7.873132051953928e-05, "loss": 0.0291, "step": 33990 }, { "grad_norm": 0.8346419930458069, "learning_rate": 7.87177866898681e-05, "loss": 0.0448, "step": 34000 }, { "grad_norm": 0.43315714597702026, "learning_rate": 7.870424971966294e-05, "loss": 0.0353, "step": 34010 }, { "grad_norm": 0.2967448830604553, "learning_rate": 7.869070961040419e-05, "loss": 0.0349, "step": 34020 }, { "grad_norm": 0.3601616322994232, "learning_rate": 7.867716636357257e-05, "loss": 0.0419, "step": 34030 }, { "grad_norm": 0.28746461868286133, "learning_rate": 7.866361998064915e-05, "loss": 0.0355, "step": 34040 }, { "grad_norm": 0.3190770745277405, "learning_rate": 7.865007046311534e-05, "loss": 0.0362, "step": 34050 }, { "grad_norm": 0.34197333455085754, "learning_rate": 7.86365178124529e-05, "loss": 0.0312, "step": 34060 }, { "grad_norm": 0.26920050382614136, "learning_rate": 7.862296203014394e-05, "loss": 0.0385, "step": 34070 }, { "grad_norm": 0.4484890103340149, "learning_rate": 7.860940311767088e-05, "loss": 0.033, "step": 34080 }, { "grad_norm": 0.5395636558532715, "learning_rate": 7.85958410765165e-05, "loss": 0.0393, "step": 34090 }, { "grad_norm": 0.5257156491279602, "learning_rate": 7.858227590816394e-05, "loss": 0.0324, "step": 34100 }, { "grad_norm": 0.4029703140258789, "learning_rate": 7.856870761409664e-05, "loss": 0.0388, "step": 34110 }, { "grad_norm": 0.5805256962776184, "learning_rate": 7.855513619579846e-05, "loss": 0.0424, "step": 34120 }, { "grad_norm": 0.31343385577201843, "learning_rate": 7.85415616547535e-05, "loss": 0.0303, "step": 34130 }, { "grad_norm": 0.27991560101509094, "learning_rate": 7.852798399244627e-05, "loss": 0.0275, "step": 34140 }, { "grad_norm": 0.5655466318130493, "learning_rate": 7.851440321036161e-05, "loss": 0.0348, "step": 34150 }, { "grad_norm": 0.40456581115722656, "learning_rate": 7.850081930998468e-05, "loss": 0.0373, "step": 34160 }, { "grad_norm": 0.40967172384262085, "learning_rate": 7.8487232292801e-05, "loss": 0.0416, "step": 34170 }, { "grad_norm": 0.3511572778224945, "learning_rate": 7.847364216029642e-05, "loss": 0.0366, "step": 34180 }, { "grad_norm": 0.4807809591293335, "learning_rate": 7.846004891395716e-05, "loss": 0.0364, "step": 34190 }, { "grad_norm": 0.42605361342430115, "learning_rate": 7.844645255526972e-05, "loss": 0.0303, "step": 34200 }, { "grad_norm": 0.46796292066574097, "learning_rate": 7.843285308572101e-05, "loss": 0.0363, "step": 34210 }, { "grad_norm": 0.32327157258987427, "learning_rate": 7.841925050679823e-05, "loss": 0.0298, "step": 34220 }, { "grad_norm": 0.3470956087112427, "learning_rate": 7.840564481998895e-05, "loss": 0.0322, "step": 34230 }, { "grad_norm": 0.2655753195285797, "learning_rate": 7.839203602678105e-05, "loss": 0.035, "step": 34240 }, { "grad_norm": 0.838781476020813, "learning_rate": 7.837842412866279e-05, "loss": 0.0349, "step": 34250 }, { "grad_norm": 0.3686838448047638, "learning_rate": 7.836480912712272e-05, "loss": 0.0394, "step": 34260 }, { "grad_norm": 0.5484349727630615, "learning_rate": 7.835119102364976e-05, "loss": 0.0453, "step": 34270 }, { "grad_norm": 0.5097535252571106, "learning_rate": 7.83375698197332e-05, "loss": 0.0321, "step": 34280 }, { "grad_norm": 0.331573486328125, "learning_rate": 7.83239455168626e-05, "loss": 0.0304, "step": 34290 }, { "grad_norm": 0.3020596206188202, "learning_rate": 7.83103181165279e-05, "loss": 0.0339, "step": 34300 }, { "grad_norm": 0.3543335199356079, "learning_rate": 7.829668762021937e-05, "loss": 0.0425, "step": 34310 }, { "grad_norm": 0.5985229015350342, "learning_rate": 7.828305402942764e-05, "loss": 0.0416, "step": 34320 }, { "grad_norm": 0.39754050970077515, "learning_rate": 7.826941734564363e-05, "loss": 0.0328, "step": 34330 }, { "grad_norm": 0.26387467980384827, "learning_rate": 7.825577757035865e-05, "loss": 0.0315, "step": 34340 }, { "grad_norm": 0.24390900135040283, "learning_rate": 7.824213470506431e-05, "loss": 0.0412, "step": 34350 }, { "grad_norm": 0.4873526990413666, "learning_rate": 7.822848875125257e-05, "loss": 0.0415, "step": 34360 }, { "grad_norm": 0.63516765832901, "learning_rate": 7.821483971041576e-05, "loss": 0.0443, "step": 34370 }, { "grad_norm": 0.499098002910614, "learning_rate": 7.820118758404649e-05, "loss": 0.0319, "step": 34380 }, { "grad_norm": 0.2254946082830429, "learning_rate": 7.818753237363776e-05, "loss": 0.0375, "step": 34390 }, { "grad_norm": 0.41870057582855225, "learning_rate": 7.817387408068286e-05, "loss": 0.0393, "step": 34400 }, { "grad_norm": 0.5739341378211975, "learning_rate": 7.816021270667544e-05, "loss": 0.0336, "step": 34410 }, { "grad_norm": 0.3939327299594879, "learning_rate": 7.81465482531095e-05, "loss": 0.0321, "step": 34420 }, { "grad_norm": 0.2911207377910614, "learning_rate": 7.813288072147938e-05, "loss": 0.0333, "step": 34430 }, { "grad_norm": 0.29450520873069763, "learning_rate": 7.811921011327972e-05, "loss": 0.0391, "step": 34440 }, { "grad_norm": 0.3759991526603699, "learning_rate": 7.810553643000549e-05, "loss": 0.0355, "step": 34450 }, { "grad_norm": 0.3455456495285034, "learning_rate": 7.809185967315206e-05, "loss": 0.0298, "step": 34460 }, { "grad_norm": 0.5247901678085327, "learning_rate": 7.80781798442151e-05, "loss": 0.0426, "step": 34470 }, { "grad_norm": 0.4613651633262634, "learning_rate": 7.806449694469061e-05, "loss": 0.0357, "step": 34480 }, { "grad_norm": 0.4268051087856293, "learning_rate": 7.805081097607492e-05, "loss": 0.0429, "step": 34490 }, { "grad_norm": 0.3382177948951721, "learning_rate": 7.803712193986474e-05, "loss": 0.0342, "step": 34500 }, { "grad_norm": 0.31603801250457764, "learning_rate": 7.802342983755702e-05, "loss": 0.037, "step": 34510 }, { "grad_norm": 0.44871586561203003, "learning_rate": 7.800973467064918e-05, "loss": 0.034, "step": 34520 }, { "grad_norm": 0.46887466311454773, "learning_rate": 7.799603644063884e-05, "loss": 0.0358, "step": 34530 }, { "grad_norm": 0.4358513057231903, "learning_rate": 7.798233514902405e-05, "loss": 0.028, "step": 34540 }, { "grad_norm": 0.3182416260242462, "learning_rate": 7.796863079730318e-05, "loss": 0.0332, "step": 34550 }, { "grad_norm": 0.4485641121864319, "learning_rate": 7.795492338697488e-05, "loss": 0.0232, "step": 34560 }, { "grad_norm": 0.3507183790206909, "learning_rate": 7.794121291953819e-05, "loss": 0.0411, "step": 34570 }, { "grad_norm": 0.36939653754234314, "learning_rate": 7.792749939649246e-05, "loss": 0.036, "step": 34580 }, { "grad_norm": 0.28270936012268066, "learning_rate": 7.79137828193374e-05, "loss": 0.0339, "step": 34590 }, { "grad_norm": 0.4263129234313965, "learning_rate": 7.790006318957301e-05, "loss": 0.0422, "step": 34600 }, { "grad_norm": 0.30968424677848816, "learning_rate": 7.788634050869965e-05, "loss": 0.0357, "step": 34610 }, { "grad_norm": 0.43102362751960754, "learning_rate": 7.787261477821803e-05, "loss": 0.0263, "step": 34620 }, { "grad_norm": 0.41463401913642883, "learning_rate": 7.785888599962916e-05, "loss": 0.0319, "step": 34630 }, { "grad_norm": 0.42821744084358215, "learning_rate": 7.784515417443439e-05, "loss": 0.0378, "step": 34640 }, { "grad_norm": 0.5024606585502625, "learning_rate": 7.783141930413545e-05, "loss": 0.0411, "step": 34650 }, { "grad_norm": 0.4273258447647095, "learning_rate": 7.78176813902343e-05, "loss": 0.0338, "step": 34660 }, { "grad_norm": 0.5324773788452148, "learning_rate": 7.780394043423336e-05, "loss": 0.0352, "step": 34670 }, { "grad_norm": 0.4260558485984802, "learning_rate": 7.77901964376353e-05, "loss": 0.0371, "step": 34680 }, { "grad_norm": 0.455831378698349, "learning_rate": 7.777644940194316e-05, "loss": 0.0327, "step": 34690 }, { "grad_norm": 0.40080004930496216, "learning_rate": 7.776269932866023e-05, "loss": 0.0335, "step": 34700 }, { "grad_norm": 0.38122648000717163, "learning_rate": 7.774894621929026e-05, "loss": 0.0398, "step": 34710 }, { "grad_norm": 0.3043213486671448, "learning_rate": 7.773519007533725e-05, "loss": 0.0354, "step": 34720 }, { "grad_norm": 0.31018123030662537, "learning_rate": 7.772143089830556e-05, "loss": 0.0383, "step": 34730 }, { "grad_norm": 0.35028085112571716, "learning_rate": 7.770766868969985e-05, "loss": 0.0369, "step": 34740 }, { "grad_norm": 0.4306255578994751, "learning_rate": 7.769390345102518e-05, "loss": 0.0437, "step": 34750 }, { "grad_norm": 0.3866455554962158, "learning_rate": 7.768013518378683e-05, "loss": 0.0428, "step": 34760 }, { "grad_norm": 0.5258846879005432, "learning_rate": 7.766636388949053e-05, "loss": 0.0322, "step": 34770 }, { "grad_norm": 0.2611868381500244, "learning_rate": 7.765258956964229e-05, "loss": 0.054, "step": 34780 }, { "grad_norm": 0.522405743598938, "learning_rate": 7.76388122257484e-05, "loss": 0.0324, "step": 34790 }, { "grad_norm": 0.344399631023407, "learning_rate": 7.762503185931558e-05, "loss": 0.0302, "step": 34800 }, { "grad_norm": 0.9022071361541748, "learning_rate": 7.76112484718508e-05, "loss": 0.0331, "step": 34810 }, { "grad_norm": 0.3013421595096588, "learning_rate": 7.75974620648614e-05, "loss": 0.0365, "step": 34820 }, { "grad_norm": 0.3586168587207794, "learning_rate": 7.758367263985503e-05, "loss": 0.0281, "step": 34830 }, { "grad_norm": 0.3440128564834595, "learning_rate": 7.75698801983397e-05, "loss": 0.0358, "step": 34840 }, { "grad_norm": 0.36901727318763733, "learning_rate": 7.755608474182372e-05, "loss": 0.0298, "step": 34850 }, { "grad_norm": 0.5963053107261658, "learning_rate": 7.754228627181574e-05, "loss": 0.0407, "step": 34860 }, { "grad_norm": 0.35870617628097534, "learning_rate": 7.752848478982476e-05, "loss": 0.0329, "step": 34870 }, { "grad_norm": 0.5050974488258362, "learning_rate": 7.751468029736006e-05, "loss": 0.0413, "step": 34880 }, { "grad_norm": 0.350374698638916, "learning_rate": 7.750087279593129e-05, "loss": 0.0454, "step": 34890 }, { "grad_norm": 0.5710947513580322, "learning_rate": 7.748706228704843e-05, "loss": 0.0389, "step": 34900 }, { "grad_norm": 0.330282986164093, "learning_rate": 7.747324877222176e-05, "loss": 0.0363, "step": 34910 }, { "grad_norm": 0.4614771604537964, "learning_rate": 7.745943225296188e-05, "loss": 0.0397, "step": 34920 }, { "grad_norm": 0.42445242404937744, "learning_rate": 7.744561273077981e-05, "loss": 0.0293, "step": 34930 }, { "grad_norm": 0.3552156686782837, "learning_rate": 7.743179020718678e-05, "loss": 0.0272, "step": 34940 }, { "grad_norm": 0.33350494503974915, "learning_rate": 7.741796468369443e-05, "loss": 0.0335, "step": 34950 }, { "grad_norm": 0.38774535059928894, "learning_rate": 7.740413616181466e-05, "loss": 0.0334, "step": 34960 }, { "grad_norm": 0.4594383239746094, "learning_rate": 7.739030464305978e-05, "loss": 0.0394, "step": 34970 }, { "grad_norm": 0.3579087257385254, "learning_rate": 7.737647012894235e-05, "loss": 0.038, "step": 34980 }, { "grad_norm": 0.31371384859085083, "learning_rate": 7.736263262097532e-05, "loss": 0.0392, "step": 34990 }, { "grad_norm": 0.3522431552410126, "learning_rate": 7.734879212067192e-05, "loss": 0.033, "step": 35000 }, { "grad_norm": 0.43970686197280884, "learning_rate": 7.733494862954573e-05, "loss": 0.0349, "step": 35010 }, { "grad_norm": 0.3555704653263092, "learning_rate": 7.732110214911066e-05, "loss": 0.0335, "step": 35020 }, { "grad_norm": 0.41723889112472534, "learning_rate": 7.730725268088094e-05, "loss": 0.0453, "step": 35030 }, { "grad_norm": 0.2462005615234375, "learning_rate": 7.729340022637112e-05, "loss": 0.0327, "step": 35040 }, { "grad_norm": 0.3951214551925659, "learning_rate": 7.727954478709607e-05, "loss": 0.0332, "step": 35050 }, { "grad_norm": 0.3440414071083069, "learning_rate": 7.726568636457103e-05, "loss": 0.0273, "step": 35060 }, { "grad_norm": 0.5091688632965088, "learning_rate": 7.725182496031153e-05, "loss": 0.0367, "step": 35070 }, { "grad_norm": 0.5667113661766052, "learning_rate": 7.723796057583341e-05, "loss": 0.0387, "step": 35080 }, { "grad_norm": 0.44276461005210876, "learning_rate": 7.722409321265287e-05, "loss": 0.0294, "step": 35090 }, { "grad_norm": 0.3011511266231537, "learning_rate": 7.721022287228645e-05, "loss": 0.0418, "step": 35100 }, { "grad_norm": 0.47669386863708496, "learning_rate": 7.719634955625094e-05, "loss": 0.0332, "step": 35110 }, { "grad_norm": 0.3746454119682312, "learning_rate": 7.718247326606355e-05, "loss": 0.0357, "step": 35120 }, { "grad_norm": 0.3717368245124817, "learning_rate": 7.716859400324175e-05, "loss": 0.0331, "step": 35130 }, { "grad_norm": 0.5621110200881958, "learning_rate": 7.715471176930335e-05, "loss": 0.0508, "step": 35140 }, { "grad_norm": 0.39820659160614014, "learning_rate": 7.714082656576651e-05, "loss": 0.0315, "step": 35150 }, { "grad_norm": 0.5831682682037354, "learning_rate": 7.712693839414968e-05, "loss": 0.0325, "step": 35160 }, { "grad_norm": 0.39373162388801575, "learning_rate": 7.711304725597164e-05, "loss": 0.041, "step": 35170 }, { "grad_norm": 0.7458943724632263, "learning_rate": 7.709915315275151e-05, "loss": 0.0343, "step": 35180 }, { "grad_norm": 0.3568446934223175, "learning_rate": 7.708525608600876e-05, "loss": 0.0464, "step": 35190 }, { "grad_norm": 0.37071338295936584, "learning_rate": 7.707135605726311e-05, "loss": 0.0342, "step": 35200 }, { "grad_norm": 0.3863879442214966, "learning_rate": 7.705745306803466e-05, "loss": 0.0534, "step": 35210 }, { "grad_norm": 0.3269311785697937, "learning_rate": 7.704354711984383e-05, "loss": 0.0293, "step": 35220 }, { "grad_norm": 0.26077163219451904, "learning_rate": 7.702963821421133e-05, "loss": 0.0312, "step": 35230 }, { "grad_norm": 0.45556333661079407, "learning_rate": 7.701572635265826e-05, "loss": 0.0369, "step": 35240 }, { "grad_norm": 0.4120641052722931, "learning_rate": 7.700181153670596e-05, "loss": 0.0302, "step": 35250 }, { "grad_norm": 0.6264234781265259, "learning_rate": 7.698789376787616e-05, "loss": 0.0307, "step": 35260 }, { "grad_norm": 0.46419641375541687, "learning_rate": 7.697397304769085e-05, "loss": 0.0341, "step": 35270 }, { "grad_norm": 0.37264490127563477, "learning_rate": 7.696004937767241e-05, "loss": 0.0494, "step": 35280 }, { "grad_norm": 0.32861217856407166, "learning_rate": 7.694612275934352e-05, "loss": 0.0346, "step": 35290 }, { "grad_norm": 0.3987441658973694, "learning_rate": 7.693219319422714e-05, "loss": 0.0305, "step": 35300 }, { "grad_norm": 0.41431912779808044, "learning_rate": 7.69182606838466e-05, "loss": 0.0425, "step": 35310 }, { "grad_norm": 0.45961740612983704, "learning_rate": 7.690432522972558e-05, "loss": 0.0376, "step": 35320 }, { "grad_norm": 0.4228561818599701, "learning_rate": 7.689038683338796e-05, "loss": 0.0311, "step": 35330 }, { "grad_norm": 0.3353624939918518, "learning_rate": 7.687644549635808e-05, "loss": 0.0279, "step": 35340 }, { "grad_norm": 0.3398837745189667, "learning_rate": 7.686250122016053e-05, "loss": 0.0301, "step": 35350 }, { "grad_norm": 0.35120680928230286, "learning_rate": 7.684855400632022e-05, "loss": 0.0349, "step": 35360 }, { "grad_norm": 0.3909093737602234, "learning_rate": 7.683460385636243e-05, "loss": 0.0286, "step": 35370 }, { "grad_norm": 0.3907926678657532, "learning_rate": 7.68206507718127e-05, "loss": 0.0382, "step": 35380 }, { "grad_norm": 0.5202562808990479, "learning_rate": 7.680669475419693e-05, "loss": 0.0411, "step": 35390 }, { "grad_norm": 0.7154792547225952, "learning_rate": 7.679273580504132e-05, "loss": 0.0442, "step": 35400 }, { "grad_norm": 0.43063145875930786, "learning_rate": 7.67787739258724e-05, "loss": 0.034, "step": 35410 }, { "grad_norm": 0.4277815520763397, "learning_rate": 7.676480911821705e-05, "loss": 0.0379, "step": 35420 }, { "grad_norm": 0.4226601719856262, "learning_rate": 7.675084138360238e-05, "loss": 0.0372, "step": 35430 }, { "grad_norm": 0.27886852622032166, "learning_rate": 7.673687072355592e-05, "loss": 0.0315, "step": 35440 }, { "grad_norm": 0.5028833150863647, "learning_rate": 7.67228971396055e-05, "loss": 0.028, "step": 35450 }, { "grad_norm": 0.2985464334487915, "learning_rate": 7.670892063327922e-05, "loss": 0.0328, "step": 35460 }, { "grad_norm": 0.2720070779323578, "learning_rate": 7.669494120610552e-05, "loss": 0.0367, "step": 35470 }, { "grad_norm": 0.37780091166496277, "learning_rate": 7.66809588596132e-05, "loss": 0.0288, "step": 35480 }, { "grad_norm": 0.5683345198631287, "learning_rate": 7.666697359533132e-05, "loss": 0.0315, "step": 35490 }, { "grad_norm": 0.4948287606239319, "learning_rate": 7.665298541478932e-05, "loss": 0.0436, "step": 35500 }, { "grad_norm": 0.39931732416152954, "learning_rate": 7.66389943195169e-05, "loss": 0.0313, "step": 35510 }, { "grad_norm": 0.3429691195487976, "learning_rate": 7.66250003110441e-05, "loss": 0.0398, "step": 35520 }, { "grad_norm": 0.41373831033706665, "learning_rate": 7.66110033909013e-05, "loss": 0.0314, "step": 35530 }, { "grad_norm": 0.33645060658454895, "learning_rate": 7.659700356061918e-05, "loss": 0.0283, "step": 35540 }, { "grad_norm": 0.3040839433670044, "learning_rate": 7.658300082172875e-05, "loss": 0.0278, "step": 35550 }, { "grad_norm": 0.3983301818370819, "learning_rate": 7.65689951757613e-05, "loss": 0.0299, "step": 35560 }, { "grad_norm": 0.3347741961479187, "learning_rate": 7.655498662424849e-05, "loss": 0.0296, "step": 35570 }, { "grad_norm": 0.2995540201663971, "learning_rate": 7.654097516872227e-05, "loss": 0.0315, "step": 35580 }, { "grad_norm": 0.5200126767158508, "learning_rate": 7.65269608107149e-05, "loss": 0.0336, "step": 35590 }, { "grad_norm": 0.43466734886169434, "learning_rate": 7.6512943551759e-05, "loss": 0.0455, "step": 35600 }, { "grad_norm": 0.3390241265296936, "learning_rate": 7.649892339338743e-05, "loss": 0.0382, "step": 35610 }, { "grad_norm": 0.26008614897727966, "learning_rate": 7.648490033713344e-05, "loss": 0.0299, "step": 35620 }, { "grad_norm": 0.2766122817993164, "learning_rate": 7.647087438453058e-05, "loss": 0.0273, "step": 35630 }, { "grad_norm": 0.3211420178413391, "learning_rate": 7.64568455371127e-05, "loss": 0.0348, "step": 35640 }, { "grad_norm": 0.3169976472854614, "learning_rate": 7.644281379641396e-05, "loss": 0.0312, "step": 35650 }, { "grad_norm": 0.3497352600097656, "learning_rate": 7.642877916396887e-05, "loss": 0.0288, "step": 35660 }, { "grad_norm": 0.3924674689769745, "learning_rate": 7.641474164131221e-05, "loss": 0.0352, "step": 35670 }, { "grad_norm": 0.31175118684768677, "learning_rate": 7.640070122997913e-05, "loss": 0.0287, "step": 35680 }, { "grad_norm": 0.28330644965171814, "learning_rate": 7.638665793150505e-05, "loss": 0.0378, "step": 35690 }, { "grad_norm": 0.47840070724487305, "learning_rate": 7.637261174742574e-05, "loss": 0.0326, "step": 35700 }, { "grad_norm": 0.43126946687698364, "learning_rate": 7.635856267927725e-05, "loss": 0.0321, "step": 35710 }, { "grad_norm": 0.35235512256622314, "learning_rate": 7.634451072859597e-05, "loss": 0.028, "step": 35720 }, { "grad_norm": 0.32888704538345337, "learning_rate": 7.633045589691863e-05, "loss": 0.0342, "step": 35730 }, { "grad_norm": 0.5203489661216736, "learning_rate": 7.63163981857822e-05, "loss": 0.0285, "step": 35740 }, { "grad_norm": 0.41726449131965637, "learning_rate": 7.630233759672403e-05, "loss": 0.0305, "step": 35750 }, { "grad_norm": 0.4399034082889557, "learning_rate": 7.628827413128177e-05, "loss": 0.0313, "step": 35760 }, { "grad_norm": 0.35112905502319336, "learning_rate": 7.627420779099338e-05, "loss": 0.0364, "step": 35770 }, { "grad_norm": 0.46955758333206177, "learning_rate": 7.626013857739711e-05, "loss": 0.0335, "step": 35780 }, { "grad_norm": 0.6773239970207214, "learning_rate": 7.624606649203158e-05, "loss": 0.0384, "step": 35790 }, { "grad_norm": 0.26991793513298035, "learning_rate": 7.623199153643569e-05, "loss": 0.0386, "step": 35800 }, { "grad_norm": 0.3110838532447815, "learning_rate": 7.621791371214863e-05, "loss": 0.0352, "step": 35810 }, { "grad_norm": 0.40950092673301697, "learning_rate": 7.620383302070995e-05, "loss": 0.0377, "step": 35820 }, { "grad_norm": 0.666553795337677, "learning_rate": 7.61897494636595e-05, "loss": 0.0331, "step": 35830 }, { "grad_norm": 0.32169556617736816, "learning_rate": 7.617566304253739e-05, "loss": 0.0448, "step": 35840 }, { "grad_norm": 0.47520872950553894, "learning_rate": 7.616157375888416e-05, "loss": 0.0298, "step": 35850 }, { "grad_norm": 0.2667853832244873, "learning_rate": 7.614748161424053e-05, "loss": 0.0283, "step": 35860 }, { "grad_norm": 0.2863723337650299, "learning_rate": 7.613338661014763e-05, "loss": 0.0273, "step": 35870 }, { "grad_norm": 0.3725821375846863, "learning_rate": 7.611928874814686e-05, "loss": 0.0332, "step": 35880 }, { "grad_norm": 0.33312907814979553, "learning_rate": 7.610518802977996e-05, "loss": 0.0293, "step": 35890 }, { "grad_norm": 0.5228464603424072, "learning_rate": 7.609108445658893e-05, "loss": 0.032, "step": 35900 }, { "grad_norm": 0.37117043137550354, "learning_rate": 7.607697803011612e-05, "loss": 0.052, "step": 35910 }, { "grad_norm": 0.3097042441368103, "learning_rate": 7.606286875190421e-05, "loss": 0.037, "step": 35920 }, { "grad_norm": 0.40920522809028625, "learning_rate": 7.604875662349617e-05, "loss": 0.0338, "step": 35930 }, { "grad_norm": 0.3507307171821594, "learning_rate": 7.603464164643525e-05, "loss": 0.0461, "step": 35940 }, { "grad_norm": 0.35025346279144287, "learning_rate": 7.602052382226507e-05, "loss": 0.0341, "step": 35950 }, { "grad_norm": 0.3842417001724243, "learning_rate": 7.600640315252954e-05, "loss": 0.0402, "step": 35960 }, { "grad_norm": 0.3006748855113983, "learning_rate": 7.599227963877284e-05, "loss": 0.0366, "step": 35970 }, { "grad_norm": 0.41754743456840515, "learning_rate": 7.597815328253954e-05, "loss": 0.0421, "step": 35980 }, { "grad_norm": 0.4081728458404541, "learning_rate": 7.596402408537443e-05, "loss": 0.0419, "step": 35990 }, { "grad_norm": 0.2798008322715759, "learning_rate": 7.59498920488227e-05, "loss": 0.0367, "step": 36000 }, { "grad_norm": 0.9492287039756775, "learning_rate": 7.593575717442979e-05, "loss": 0.0324, "step": 36010 }, { "grad_norm": 0.40499553084373474, "learning_rate": 7.592161946374147e-05, "loss": 0.0339, "step": 36020 }, { "grad_norm": 0.8046404123306274, "learning_rate": 7.590747891830381e-05, "loss": 0.0402, "step": 36030 }, { "grad_norm": 0.4131206274032593, "learning_rate": 7.58933355396632e-05, "loss": 0.0324, "step": 36040 }, { "grad_norm": 0.33744120597839355, "learning_rate": 7.587918932936636e-05, "loss": 0.0276, "step": 36050 }, { "grad_norm": 0.49018293619155884, "learning_rate": 7.586504028896028e-05, "loss": 0.0327, "step": 36060 }, { "grad_norm": 0.37769612669944763, "learning_rate": 7.585088841999228e-05, "loss": 0.0341, "step": 36070 }, { "grad_norm": 0.36429518461227417, "learning_rate": 7.583673372400999e-05, "loss": 0.0331, "step": 36080 }, { "grad_norm": 0.39852064847946167, "learning_rate": 7.582257620256134e-05, "loss": 0.0291, "step": 36090 }, { "grad_norm": 0.46416449546813965, "learning_rate": 7.580841585719458e-05, "loss": 0.0446, "step": 36100 }, { "grad_norm": 0.2601241171360016, "learning_rate": 7.579425268945825e-05, "loss": 0.0343, "step": 36110 }, { "grad_norm": 0.42641139030456543, "learning_rate": 7.578008670090127e-05, "loss": 0.0295, "step": 36120 }, { "grad_norm": 0.461740106344223, "learning_rate": 7.576591789307272e-05, "loss": 0.0372, "step": 36130 }, { "grad_norm": 0.3818996250629425, "learning_rate": 7.575174626752216e-05, "loss": 0.0346, "step": 36140 }, { "grad_norm": 0.4993993937969208, "learning_rate": 7.573757182579934e-05, "loss": 0.039, "step": 36150 }, { "grad_norm": 0.3405364453792572, "learning_rate": 7.572339456945435e-05, "loss": 0.0341, "step": 36160 }, { "grad_norm": 0.3179329037666321, "learning_rate": 7.570921450003762e-05, "loss": 0.0348, "step": 36170 }, { "grad_norm": 0.3541111648082733, "learning_rate": 7.569503161909983e-05, "loss": 0.0299, "step": 36180 }, { "grad_norm": 0.344411700963974, "learning_rate": 7.568084592819202e-05, "loss": 0.0321, "step": 36190 }, { "grad_norm": 0.39138782024383545, "learning_rate": 7.566665742886551e-05, "loss": 0.0352, "step": 36200 }, { "grad_norm": 0.5335773825645447, "learning_rate": 7.565246612267194e-05, "loss": 0.0333, "step": 36210 }, { "grad_norm": 0.3362475335597992, "learning_rate": 7.563827201116325e-05, "loss": 0.038, "step": 36220 }, { "grad_norm": 0.4744154214859009, "learning_rate": 7.562407509589164e-05, "loss": 0.0313, "step": 36230 }, { "grad_norm": 0.3308800160884857, "learning_rate": 7.560987537840973e-05, "loss": 0.0251, "step": 36240 }, { "grad_norm": 0.3051164448261261, "learning_rate": 7.559567286027036e-05, "loss": 0.0335, "step": 36250 }, { "grad_norm": 0.3647863566875458, "learning_rate": 7.558146754302668e-05, "loss": 0.0366, "step": 36260 }, { "grad_norm": 0.3009224534034729, "learning_rate": 7.556725942823217e-05, "loss": 0.033, "step": 36270 }, { "grad_norm": 0.3094457983970642, "learning_rate": 7.555304851744061e-05, "loss": 0.0352, "step": 36280 }, { "grad_norm": 0.3972824215888977, "learning_rate": 7.553883481220608e-05, "loss": 0.0417, "step": 36290 }, { "grad_norm": 0.7230464220046997, "learning_rate": 7.552461831408298e-05, "loss": 0.0293, "step": 36300 }, { "grad_norm": 0.34158816933631897, "learning_rate": 7.551039902462599e-05, "loss": 0.0342, "step": 36310 }, { "grad_norm": 0.3956478536128998, "learning_rate": 7.549617694539011e-05, "loss": 0.0311, "step": 36320 }, { "grad_norm": 0.42074429988861084, "learning_rate": 7.548195207793066e-05, "loss": 0.0428, "step": 36330 }, { "grad_norm": 0.3600137531757355, "learning_rate": 7.546772442380323e-05, "loss": 0.0291, "step": 36340 }, { "grad_norm": 0.34349435567855835, "learning_rate": 7.545349398456376e-05, "loss": 0.0316, "step": 36350 }, { "grad_norm": 0.3755649924278259, "learning_rate": 7.543926076176845e-05, "loss": 0.038, "step": 36360 }, { "grad_norm": 0.44063878059387207, "learning_rate": 7.542502475697385e-05, "loss": 0.0346, "step": 36370 }, { "grad_norm": 0.34665220975875854, "learning_rate": 7.541078597173675e-05, "loss": 0.0346, "step": 36380 }, { "grad_norm": 0.3656481206417084, "learning_rate": 7.539654440761431e-05, "loss": 0.0348, "step": 36390 }, { "grad_norm": 0.37812358140945435, "learning_rate": 7.538230006616395e-05, "loss": 0.0321, "step": 36400 }, { "grad_norm": 0.20877990126609802, "learning_rate": 7.536805294894342e-05, "loss": 0.0303, "step": 36410 }, { "grad_norm": 0.5118564367294312, "learning_rate": 7.535380305751076e-05, "loss": 0.0339, "step": 36420 }, { "grad_norm": 0.33090880513191223, "learning_rate": 7.533955039342431e-05, "loss": 0.0401, "step": 36430 }, { "grad_norm": 0.2805456519126892, "learning_rate": 7.532529495824274e-05, "loss": 0.0366, "step": 36440 }, { "grad_norm": 0.32989802956581116, "learning_rate": 7.531103675352497e-05, "loss": 0.0385, "step": 36450 }, { "grad_norm": 0.3253542184829712, "learning_rate": 7.529677578083029e-05, "loss": 0.0389, "step": 36460 }, { "grad_norm": 0.269228458404541, "learning_rate": 7.528251204171823e-05, "loss": 0.0229, "step": 36470 }, { "grad_norm": 0.4116588234901428, "learning_rate": 7.526824553774867e-05, "loss": 0.0368, "step": 36480 }, { "grad_norm": 0.4377199411392212, "learning_rate": 7.525397627048178e-05, "loss": 0.0309, "step": 36490 }, { "grad_norm": 0.3608086109161377, "learning_rate": 7.523970424147802e-05, "loss": 0.0369, "step": 36500 }, { "grad_norm": 0.2746692895889282, "learning_rate": 7.522542945229813e-05, "loss": 0.0319, "step": 36510 }, { "grad_norm": 0.2703666687011719, "learning_rate": 7.521115190450321e-05, "loss": 0.0346, "step": 36520 }, { "grad_norm": 0.324178546667099, "learning_rate": 7.519687159965462e-05, "loss": 0.0279, "step": 36530 }, { "grad_norm": 0.2951936423778534, "learning_rate": 7.518258853931403e-05, "loss": 0.0298, "step": 36540 }, { "grad_norm": 0.3794071674346924, "learning_rate": 7.516830272504342e-05, "loss": 0.0294, "step": 36550 }, { "grad_norm": 0.5887100696563721, "learning_rate": 7.515401415840509e-05, "loss": 0.0406, "step": 36560 }, { "grad_norm": 0.6985664963722229, "learning_rate": 7.513972284096155e-05, "loss": 0.0289, "step": 36570 }, { "grad_norm": 0.3541799485683441, "learning_rate": 7.512542877427574e-05, "loss": 0.0441, "step": 36580 }, { "grad_norm": 0.423201322555542, "learning_rate": 7.51111319599108e-05, "loss": 0.0435, "step": 36590 }, { "grad_norm": 0.4553878903388977, "learning_rate": 7.509683239943024e-05, "loss": 0.0338, "step": 36600 }, { "grad_norm": 0.5090067982673645, "learning_rate": 7.50825300943978e-05, "loss": 0.0324, "step": 36610 }, { "grad_norm": 0.4441375136375427, "learning_rate": 7.50682250463776e-05, "loss": 0.0347, "step": 36620 }, { "grad_norm": 0.3797697424888611, "learning_rate": 7.5053917256934e-05, "loss": 0.0416, "step": 36630 }, { "grad_norm": 0.23299619555473328, "learning_rate": 7.503960672763166e-05, "loss": 0.0327, "step": 36640 }, { "grad_norm": 0.2888309061527252, "learning_rate": 7.502529346003559e-05, "loss": 0.0317, "step": 36650 }, { "grad_norm": 0.4179849624633789, "learning_rate": 7.501097745571106e-05, "loss": 0.0365, "step": 36660 }, { "grad_norm": 0.42762327194213867, "learning_rate": 7.499665871622361e-05, "loss": 0.0417, "step": 36670 }, { "grad_norm": 0.35392332077026367, "learning_rate": 7.498233724313917e-05, "loss": 0.0404, "step": 36680 }, { "grad_norm": 0.3826090097427368, "learning_rate": 7.496801303802387e-05, "loss": 0.0412, "step": 36690 }, { "grad_norm": 0.2155860960483551, "learning_rate": 7.495368610244423e-05, "loss": 0.0319, "step": 36700 }, { "grad_norm": 0.43661731481552124, "learning_rate": 7.493935643796697e-05, "loss": 0.0316, "step": 36710 }, { "grad_norm": 0.3844558000564575, "learning_rate": 7.49250240461592e-05, "loss": 0.0365, "step": 36720 }, { "grad_norm": 0.5860258936882019, "learning_rate": 7.491068892858826e-05, "loss": 0.0348, "step": 36730 }, { "grad_norm": 0.30702096223831177, "learning_rate": 7.489635108682184e-05, "loss": 0.031, "step": 36740 }, { "grad_norm": 0.41511642932891846, "learning_rate": 7.48820105224279e-05, "loss": 0.0364, "step": 36750 }, { "grad_norm": 0.3865789473056793, "learning_rate": 7.48676672369747e-05, "loss": 0.0327, "step": 36760 }, { "grad_norm": 0.3951210081577301, "learning_rate": 7.485332123203079e-05, "loss": 0.032, "step": 36770 }, { "grad_norm": 0.2949102520942688, "learning_rate": 7.483897250916504e-05, "loss": 0.036, "step": 36780 }, { "grad_norm": 0.292131245136261, "learning_rate": 7.48246210699466e-05, "loss": 0.0353, "step": 36790 }, { "grad_norm": 0.3742341995239258, "learning_rate": 7.481026691594492e-05, "loss": 0.0349, "step": 36800 }, { "grad_norm": 0.2902800440788269, "learning_rate": 7.479591004872975e-05, "loss": 0.0341, "step": 36810 }, { "grad_norm": 0.4766395092010498, "learning_rate": 7.478155046987114e-05, "loss": 0.0397, "step": 36820 }, { "grad_norm": 0.2813335657119751, "learning_rate": 7.476718818093944e-05, "loss": 0.0326, "step": 36830 }, { "grad_norm": 0.3083386719226837, "learning_rate": 7.475282318350527e-05, "loss": 0.0335, "step": 36840 }, { "grad_norm": 0.3280448913574219, "learning_rate": 7.473845547913957e-05, "loss": 0.0328, "step": 36850 }, { "grad_norm": 0.4497562050819397, "learning_rate": 7.472408506941357e-05, "loss": 0.0409, "step": 36860 }, { "grad_norm": 0.35284799337387085, "learning_rate": 7.470971195589881e-05, "loss": 0.0317, "step": 36870 }, { "grad_norm": 0.41531240940093994, "learning_rate": 7.46953361401671e-05, "loss": 0.0435, "step": 36880 }, { "grad_norm": 0.3221638798713684, "learning_rate": 7.468095762379055e-05, "loss": 0.0302, "step": 36890 }, { "grad_norm": 0.32425621151924133, "learning_rate": 7.466657640834158e-05, "loss": 0.034, "step": 36900 }, { "grad_norm": 0.42436447739601135, "learning_rate": 7.46521924953929e-05, "loss": 0.0295, "step": 36910 }, { "grad_norm": 0.612923800945282, "learning_rate": 7.463780588651752e-05, "loss": 0.0331, "step": 36920 }, { "grad_norm": 0.4442134201526642, "learning_rate": 7.462341658328873e-05, "loss": 0.031, "step": 36930 }, { "grad_norm": 0.35312506556510925, "learning_rate": 7.460902458728012e-05, "loss": 0.0274, "step": 36940 }, { "grad_norm": 0.4413066506385803, "learning_rate": 7.459462990006558e-05, "loss": 0.0254, "step": 36950 }, { "grad_norm": 0.3437126874923706, "learning_rate": 7.458023252321929e-05, "loss": 0.0333, "step": 36960 }, { "grad_norm": 0.44201335310935974, "learning_rate": 7.456583245831573e-05, "loss": 0.027, "step": 36970 }, { "grad_norm": 0.27892598509788513, "learning_rate": 7.455142970692968e-05, "loss": 0.0467, "step": 36980 }, { "grad_norm": 0.46673256158828735, "learning_rate": 7.453702427063615e-05, "loss": 0.0345, "step": 36990 }, { "grad_norm": 0.521605908870697, "learning_rate": 7.452261615101057e-05, "loss": 0.0293, "step": 37000 }, { "grad_norm": 0.34245461225509644, "learning_rate": 7.450820534962856e-05, "loss": 0.0401, "step": 37010 }, { "grad_norm": 0.6796091198921204, "learning_rate": 7.449379186806604e-05, "loss": 0.0325, "step": 37020 }, { "grad_norm": 0.39027395844459534, "learning_rate": 7.44793757078993e-05, "loss": 0.0303, "step": 37030 }, { "grad_norm": 0.4255257248878479, "learning_rate": 7.44649568707048e-05, "loss": 0.036, "step": 37040 }, { "grad_norm": 0.3135553300380707, "learning_rate": 7.445053535805942e-05, "loss": 0.0366, "step": 37050 }, { "grad_norm": 0.2808677852153778, "learning_rate": 7.443611117154024e-05, "loss": 0.0326, "step": 37060 }, { "grad_norm": 0.3112480342388153, "learning_rate": 7.442168431272469e-05, "loss": 0.0379, "step": 37070 }, { "grad_norm": 0.492153137922287, "learning_rate": 7.440725478319044e-05, "loss": 0.0304, "step": 37080 }, { "grad_norm": 0.3069005608558655, "learning_rate": 7.439282258451553e-05, "loss": 0.0322, "step": 37090 }, { "grad_norm": 0.33083832263946533, "learning_rate": 7.43783877182782e-05, "loss": 0.0311, "step": 37100 }, { "grad_norm": 0.28368014097213745, "learning_rate": 7.436395018605707e-05, "loss": 0.0305, "step": 37110 }, { "grad_norm": 0.32182154059410095, "learning_rate": 7.434950998943094e-05, "loss": 0.0299, "step": 37120 }, { "grad_norm": 0.3670317530632019, "learning_rate": 7.433506712997902e-05, "loss": 0.0373, "step": 37130 }, { "grad_norm": 0.2479611486196518, "learning_rate": 7.432062160928077e-05, "loss": 0.0259, "step": 37140 }, { "grad_norm": 0.3443259000778198, "learning_rate": 7.430617342891588e-05, "loss": 0.024, "step": 37150 }, { "grad_norm": 0.3481704294681549, "learning_rate": 7.429172259046442e-05, "loss": 0.0345, "step": 37160 }, { "grad_norm": 0.9558371305465698, "learning_rate": 7.42772690955067e-05, "loss": 0.0258, "step": 37170 }, { "grad_norm": 0.4138137102127075, "learning_rate": 7.426281294562332e-05, "loss": 0.0445, "step": 37180 }, { "grad_norm": 0.41977810859680176, "learning_rate": 7.424835414239519e-05, "loss": 0.0317, "step": 37190 }, { "grad_norm": 0.3409653604030609, "learning_rate": 7.423389268740352e-05, "loss": 0.0309, "step": 37200 }, { "grad_norm": 0.40710580348968506, "learning_rate": 7.42194285822298e-05, "loss": 0.0301, "step": 37210 }, { "grad_norm": 0.2544859051704407, "learning_rate": 7.420496182845576e-05, "loss": 0.03, "step": 37220 }, { "grad_norm": 0.42253366112709045, "learning_rate": 7.419049242766348e-05, "loss": 0.0408, "step": 37230 }, { "grad_norm": 0.4477959871292114, "learning_rate": 7.417602038143533e-05, "loss": 0.0375, "step": 37240 }, { "grad_norm": 0.37540650367736816, "learning_rate": 7.416154569135393e-05, "loss": 0.033, "step": 37250 }, { "grad_norm": 0.3041248023509979, "learning_rate": 7.414706835900222e-05, "loss": 0.0346, "step": 37260 }, { "grad_norm": 0.36423951387405396, "learning_rate": 7.413258838596345e-05, "loss": 0.0392, "step": 37270 }, { "grad_norm": 0.3022841811180115, "learning_rate": 7.411810577382106e-05, "loss": 0.0318, "step": 37280 }, { "grad_norm": 0.4160495400428772, "learning_rate": 7.410362052415889e-05, "loss": 0.0324, "step": 37290 }, { "grad_norm": 0.293491929769516, "learning_rate": 7.408913263856102e-05, "loss": 0.0334, "step": 37300 }, { "grad_norm": 0.35731592774391174, "learning_rate": 7.407464211861183e-05, "loss": 0.0363, "step": 37310 }, { "grad_norm": 0.49417099356651306, "learning_rate": 7.406014896589597e-05, "loss": 0.0516, "step": 37320 }, { "grad_norm": 0.4194808304309845, "learning_rate": 7.404565318199838e-05, "loss": 0.0341, "step": 37330 }, { "grad_norm": 0.37769612669944763, "learning_rate": 7.403115476850432e-05, "loss": 0.0389, "step": 37340 }, { "grad_norm": 0.3823331296443939, "learning_rate": 7.401665372699932e-05, "loss": 0.0309, "step": 37350 }, { "grad_norm": 0.32830026745796204, "learning_rate": 7.400215005906916e-05, "loss": 0.0311, "step": 37360 }, { "grad_norm": 0.2521176338195801, "learning_rate": 7.398764376629996e-05, "loss": 0.0282, "step": 37370 }, { "grad_norm": 0.3097568452358246, "learning_rate": 7.39731348502781e-05, "loss": 0.0352, "step": 37380 }, { "grad_norm": 0.29747146368026733, "learning_rate": 7.395862331259028e-05, "loss": 0.0279, "step": 37390 }, { "grad_norm": 0.2748945653438568, "learning_rate": 7.394410915482341e-05, "loss": 0.0301, "step": 37400 }, { "grad_norm": 0.4199451506137848, "learning_rate": 7.392959237856476e-05, "loss": 0.0404, "step": 37410 }, { "grad_norm": 0.4153384864330292, "learning_rate": 7.391507298540187e-05, "loss": 0.0324, "step": 37420 }, { "grad_norm": 0.4707445502281189, "learning_rate": 7.390055097692257e-05, "loss": 0.0341, "step": 37430 }, { "grad_norm": 0.5450549721717834, "learning_rate": 7.388602635471494e-05, "loss": 0.051, "step": 37440 }, { "grad_norm": 0.7617719769477844, "learning_rate": 7.387149912036738e-05, "loss": 0.0518, "step": 37450 }, { "grad_norm": 0.5700081586837769, "learning_rate": 7.385696927546854e-05, "loss": 0.0366, "step": 37460 }, { "grad_norm": 0.4428249001502991, "learning_rate": 7.384243682160743e-05, "loss": 0.0473, "step": 37470 }, { "grad_norm": 0.2887082099914551, "learning_rate": 7.382790176037329e-05, "loss": 0.0381, "step": 37480 }, { "grad_norm": 0.38078126311302185, "learning_rate": 7.381336409335562e-05, "loss": 0.0359, "step": 37490 }, { "grad_norm": 0.396837443113327, "learning_rate": 7.379882382214426e-05, "loss": 0.0391, "step": 37500 }, { "grad_norm": 0.4836360514163971, "learning_rate": 7.378428094832931e-05, "loss": 0.0426, "step": 37510 }, { "grad_norm": 0.4422653913497925, "learning_rate": 7.376973547350114e-05, "loss": 0.0405, "step": 37520 }, { "grad_norm": 0.42717617750167847, "learning_rate": 7.375518739925046e-05, "loss": 0.0307, "step": 37530 }, { "grad_norm": 0.5133965015411377, "learning_rate": 7.374063672716818e-05, "loss": 0.0331, "step": 37540 }, { "grad_norm": 0.42903390526771545, "learning_rate": 7.372608345884558e-05, "loss": 0.034, "step": 37550 }, { "grad_norm": 0.5235249400138855, "learning_rate": 7.371152759587415e-05, "loss": 0.0387, "step": 37560 }, { "grad_norm": 0.3171375095844269, "learning_rate": 7.369696913984572e-05, "loss": 0.0336, "step": 37570 }, { "grad_norm": 0.2687748074531555, "learning_rate": 7.368240809235237e-05, "loss": 0.029, "step": 37580 }, { "grad_norm": 0.365714967250824, "learning_rate": 7.366784445498649e-05, "loss": 0.0321, "step": 37590 }, { "grad_norm": 0.2771085798740387, "learning_rate": 7.36532782293407e-05, "loss": 0.03, "step": 37600 }, { "grad_norm": 0.37523677945137024, "learning_rate": 7.363870941700797e-05, "loss": 0.0352, "step": 37610 }, { "grad_norm": 0.3404635190963745, "learning_rate": 7.362413801958152e-05, "loss": 0.0347, "step": 37620 }, { "grad_norm": 0.365079402923584, "learning_rate": 7.360956403865486e-05, "loss": 0.0371, "step": 37630 }, { "grad_norm": 0.4436698853969574, "learning_rate": 7.359498747582177e-05, "loss": 0.0331, "step": 37640 }, { "grad_norm": 0.21519352495670319, "learning_rate": 7.358040833267634e-05, "loss": 0.0336, "step": 37650 }, { "grad_norm": 0.40024229884147644, "learning_rate": 7.356582661081288e-05, "loss": 0.0338, "step": 37660 }, { "grad_norm": 0.3922363817691803, "learning_rate": 7.355124231182607e-05, "loss": 0.029, "step": 37670 }, { "grad_norm": 0.5239422917366028, "learning_rate": 7.353665543731083e-05, "loss": 0.0342, "step": 37680 }, { "grad_norm": 0.3817391097545624, "learning_rate": 7.352206598886229e-05, "loss": 0.0281, "step": 37690 }, { "grad_norm": 0.3492562174797058, "learning_rate": 7.350747396807601e-05, "loss": 0.0325, "step": 37700 }, { "grad_norm": 0.3437367379665375, "learning_rate": 7.349287937654772e-05, "loss": 0.0294, "step": 37710 }, { "grad_norm": 0.30042627453804016, "learning_rate": 7.347828221587345e-05, "loss": 0.0378, "step": 37720 }, { "grad_norm": 0.4912254214286804, "learning_rate": 7.346368248764955e-05, "loss": 0.0348, "step": 37730 }, { "grad_norm": 0.4318970739841461, "learning_rate": 7.34490801934726e-05, "loss": 0.03, "step": 37740 }, { "grad_norm": 0.37937602400779724, "learning_rate": 7.343447533493947e-05, "loss": 0.0355, "step": 37750 }, { "grad_norm": 0.49812841415405273, "learning_rate": 7.341986791364739e-05, "loss": 0.0499, "step": 37760 }, { "grad_norm": 0.32673323154449463, "learning_rate": 7.340525793119374e-05, "loss": 0.0433, "step": 37770 }, { "grad_norm": 0.46408629417419434, "learning_rate": 7.33906453891763e-05, "loss": 0.0303, "step": 37780 }, { "grad_norm": 0.37649425864219666, "learning_rate": 7.337603028919303e-05, "loss": 0.0303, "step": 37790 }, { "grad_norm": 0.33300793170928955, "learning_rate": 7.336141263284225e-05, "loss": 0.0292, "step": 37800 }, { "grad_norm": 0.4963834583759308, "learning_rate": 7.334679242172249e-05, "loss": 0.0399, "step": 37810 }, { "grad_norm": 0.3037944734096527, "learning_rate": 7.33321696574326e-05, "loss": 0.0321, "step": 37820 }, { "grad_norm": 0.4302045404911041, "learning_rate": 7.331754434157174e-05, "loss": 0.0317, "step": 37830 }, { "grad_norm": 0.454776406288147, "learning_rate": 7.330291647573928e-05, "loss": 0.0347, "step": 37840 }, { "grad_norm": 0.2976986765861511, "learning_rate": 7.32882860615349e-05, "loss": 0.0304, "step": 37850 }, { "grad_norm": 1.6750924587249756, "learning_rate": 7.327365310055858e-05, "loss": 0.0323, "step": 37860 }, { "grad_norm": 0.38401198387145996, "learning_rate": 7.325901759441055e-05, "loss": 0.0302, "step": 37870 }, { "grad_norm": 0.5251865983009338, "learning_rate": 7.324437954469132e-05, "loss": 0.0461, "step": 37880 }, { "grad_norm": 0.4233950078487396, "learning_rate": 7.32297389530017e-05, "loss": 0.0309, "step": 37890 }, { "grad_norm": 0.36709344387054443, "learning_rate": 7.321509582094274e-05, "loss": 0.0339, "step": 37900 }, { "grad_norm": 0.3007810115814209, "learning_rate": 7.320045015011582e-05, "loss": 0.0386, "step": 37910 }, { "grad_norm": 0.3675524592399597, "learning_rate": 7.318580194212253e-05, "loss": 0.0346, "step": 37920 }, { "grad_norm": 0.39938902854919434, "learning_rate": 7.317115119856483e-05, "loss": 0.0424, "step": 37930 }, { "grad_norm": 0.9531322121620178, "learning_rate": 7.315649792104485e-05, "loss": 0.0348, "step": 37940 }, { "grad_norm": 0.43710434436798096, "learning_rate": 7.314184211116507e-05, "loss": 0.0454, "step": 37950 }, { "grad_norm": 0.4442552626132965, "learning_rate": 7.312718377052823e-05, "loss": 0.0349, "step": 37960 }, { "grad_norm": 0.3282279670238495, "learning_rate": 7.311252290073737e-05, "loss": 0.0303, "step": 37970 }, { "grad_norm": 0.4017675518989563, "learning_rate": 7.309785950339572e-05, "loss": 0.0283, "step": 37980 }, { "grad_norm": 0.40571245551109314, "learning_rate": 7.308319358010692e-05, "loss": 0.0257, "step": 37990 }, { "grad_norm": 0.5306070446968079, "learning_rate": 7.306852513247474e-05, "loss": 0.0505, "step": 38000 }, { "grad_norm": 0.4027003347873688, "learning_rate": 7.305385416210336e-05, "loss": 0.0411, "step": 38010 }, { "grad_norm": 0.3338213562965393, "learning_rate": 7.303918067059711e-05, "loss": 0.0293, "step": 38020 }, { "grad_norm": 0.2880546450614929, "learning_rate": 7.302450465956075e-05, "loss": 0.0344, "step": 38030 }, { "grad_norm": 0.5603381395339966, "learning_rate": 7.300982613059914e-05, "loss": 0.0285, "step": 38040 }, { "grad_norm": 0.4020427465438843, "learning_rate": 7.299514508531757e-05, "loss": 0.0387, "step": 38050 }, { "grad_norm": 0.7964919209480286, "learning_rate": 7.298046152532148e-05, "loss": 0.0349, "step": 38060 }, { "grad_norm": 0.28855398297309875, "learning_rate": 7.296577545221668e-05, "loss": 0.0225, "step": 38070 }, { "grad_norm": 0.4242708683013916, "learning_rate": 7.29510868676092e-05, "loss": 0.0346, "step": 38080 }, { "grad_norm": 0.31504279375076294, "learning_rate": 7.293639577310538e-05, "loss": 0.0311, "step": 38090 }, { "grad_norm": 0.32214802503585815, "learning_rate": 7.292170217031179e-05, "loss": 0.0312, "step": 38100 }, { "grad_norm": 0.3401338458061218, "learning_rate": 7.290700606083532e-05, "loss": 0.0382, "step": 38110 }, { "grad_norm": 0.33853939175605774, "learning_rate": 7.28923074462831e-05, "loss": 0.0341, "step": 38120 }, { "grad_norm": 0.34683775901794434, "learning_rate": 7.287760632826255e-05, "loss": 0.039, "step": 38130 }, { "grad_norm": 0.4520412087440491, "learning_rate": 7.286290270838139e-05, "loss": 0.0458, "step": 38140 }, { "grad_norm": 0.32022544741630554, "learning_rate": 7.284819658824756e-05, "loss": 0.0269, "step": 38150 }, { "grad_norm": 0.6239699721336365, "learning_rate": 7.283348796946929e-05, "loss": 0.0322, "step": 38160 }, { "grad_norm": 0.8128901720046997, "learning_rate": 7.28187768536551e-05, "loss": 0.0334, "step": 38170 }, { "grad_norm": 0.37184008955955505, "learning_rate": 7.280406324241379e-05, "loss": 0.0423, "step": 38180 }, { "grad_norm": 0.3089345693588257, "learning_rate": 7.278934713735441e-05, "loss": 0.0394, "step": 38190 }, { "grad_norm": 0.42565473914146423, "learning_rate": 7.277462854008629e-05, "loss": 0.0335, "step": 38200 }, { "grad_norm": 0.48599299788475037, "learning_rate": 7.275990745221904e-05, "loss": 0.032, "step": 38210 }, { "grad_norm": 0.38187727332115173, "learning_rate": 7.274518387536252e-05, "loss": 0.0393, "step": 38220 }, { "grad_norm": 0.3447907865047455, "learning_rate": 7.273045781112689e-05, "loss": 0.0419, "step": 38230 }, { "grad_norm": 0.45305269956588745, "learning_rate": 7.271572926112257e-05, "loss": 0.0303, "step": 38240 }, { "grad_norm": 0.3505942225456238, "learning_rate": 7.270099822696024e-05, "loss": 0.0383, "step": 38250 }, { "grad_norm": 0.2992648482322693, "learning_rate": 7.26862647102509e-05, "loss": 0.0276, "step": 38260 }, { "grad_norm": 0.37451043725013733, "learning_rate": 7.267152871260573e-05, "loss": 0.0411, "step": 38270 }, { "grad_norm": 0.4265402853488922, "learning_rate": 7.265679023563629e-05, "loss": 0.0356, "step": 38280 }, { "grad_norm": 0.31100454926490784, "learning_rate": 7.264204928095433e-05, "loss": 0.03, "step": 38290 }, { "grad_norm": 0.48493677377700806, "learning_rate": 7.262730585017188e-05, "loss": 0.0358, "step": 38300 }, { "grad_norm": 0.3573066294193268, "learning_rate": 7.261255994490131e-05, "loss": 0.0343, "step": 38310 }, { "grad_norm": 0.36916905641555786, "learning_rate": 7.259781156675517e-05, "loss": 0.0351, "step": 38320 }, { "grad_norm": 0.3576589524745941, "learning_rate": 7.258306071734633e-05, "loss": 0.0292, "step": 38330 }, { "grad_norm": 0.34341928362846375, "learning_rate": 7.256830739828792e-05, "loss": 0.0274, "step": 38340 }, { "grad_norm": 0.27074477076530457, "learning_rate": 7.255355161119336e-05, "loss": 0.0227, "step": 38350 }, { "grad_norm": 0.28132200241088867, "learning_rate": 7.253879335767628e-05, "loss": 0.0252, "step": 38360 }, { "grad_norm": 0.3487521708011627, "learning_rate": 7.252403263935067e-05, "loss": 0.0292, "step": 38370 }, { "grad_norm": 0.4051937460899353, "learning_rate": 7.25092694578307e-05, "loss": 0.0393, "step": 38380 }, { "grad_norm": 0.40017232298851013, "learning_rate": 7.249450381473087e-05, "loss": 0.0277, "step": 38390 }, { "grad_norm": 0.3733394145965576, "learning_rate": 7.247973571166593e-05, "loss": 0.0344, "step": 38400 }, { "grad_norm": 0.4459158778190613, "learning_rate": 7.246496515025089e-05, "loss": 0.031, "step": 38410 }, { "grad_norm": 0.44212985038757324, "learning_rate": 7.245019213210101e-05, "loss": 0.0331, "step": 38420 }, { "grad_norm": 0.2861377000808716, "learning_rate": 7.24354166588319e-05, "loss": 0.0388, "step": 38430 }, { "grad_norm": 0.47197380661964417, "learning_rate": 7.242063873205935e-05, "loss": 0.0389, "step": 38440 }, { "grad_norm": 0.3580141067504883, "learning_rate": 7.240585835339946e-05, "loss": 0.0436, "step": 38450 }, { "grad_norm": 0.27979061007499695, "learning_rate": 7.239107552446857e-05, "loss": 0.0314, "step": 38460 }, { "grad_norm": 0.4700552523136139, "learning_rate": 7.237629024688334e-05, "loss": 0.0286, "step": 38470 }, { "grad_norm": 0.48670580983161926, "learning_rate": 7.236150252226064e-05, "loss": 0.0382, "step": 38480 }, { "grad_norm": 0.2502517104148865, "learning_rate": 7.234671235221765e-05, "loss": 0.042, "step": 38490 }, { "grad_norm": 0.3307342827320099, "learning_rate": 7.233191973837179e-05, "loss": 0.0332, "step": 38500 }, { "grad_norm": 0.3696059584617615, "learning_rate": 7.231712468234078e-05, "loss": 0.0325, "step": 38510 }, { "grad_norm": 0.4728197753429413, "learning_rate": 7.230232718574254e-05, "loss": 0.0301, "step": 38520 }, { "grad_norm": 0.6527722477912903, "learning_rate": 7.228752725019535e-05, "loss": 0.0377, "step": 38530 }, { "grad_norm": 0.3731437623500824, "learning_rate": 7.227272487731769e-05, "loss": 0.0322, "step": 38540 }, { "grad_norm": 0.4738205373287201, "learning_rate": 7.225792006872831e-05, "loss": 0.0309, "step": 38550 }, { "grad_norm": 0.24899278581142426, "learning_rate": 7.224311282604628e-05, "loss": 0.0295, "step": 38560 }, { "grad_norm": 0.349290668964386, "learning_rate": 7.222830315089085e-05, "loss": 0.0397, "step": 38570 }, { "grad_norm": 0.4628864824771881, "learning_rate": 7.22134910448816e-05, "loss": 0.0307, "step": 38580 }, { "grad_norm": 0.5145701766014099, "learning_rate": 7.219867650963839e-05, "loss": 0.0394, "step": 38590 }, { "grad_norm": 0.4278562664985657, "learning_rate": 7.21838595467813e-05, "loss": 0.0361, "step": 38600 }, { "grad_norm": 0.33018290996551514, "learning_rate": 7.216904015793067e-05, "loss": 0.0291, "step": 38610 }, { "grad_norm": 0.39696916937828064, "learning_rate": 7.215421834470713e-05, "loss": 0.0295, "step": 38620 }, { "grad_norm": 0.43575480580329895, "learning_rate": 7.213939410873158e-05, "loss": 0.0356, "step": 38630 }, { "grad_norm": 0.24102437496185303, "learning_rate": 7.212456745162519e-05, "loss": 0.0357, "step": 38640 }, { "grad_norm": 0.39658650755882263, "learning_rate": 7.210973837500937e-05, "loss": 0.034, "step": 38650 }, { "grad_norm": 0.3201777935028076, "learning_rate": 7.209490688050578e-05, "loss": 0.0286, "step": 38660 }, { "grad_norm": 0.29295849800109863, "learning_rate": 7.208007296973641e-05, "loss": 0.0296, "step": 38670 }, { "grad_norm": 0.42052462697029114, "learning_rate": 7.206523664432345e-05, "loss": 0.0344, "step": 38680 }, { "grad_norm": 0.3936426639556885, "learning_rate": 7.205039790588939e-05, "loss": 0.0277, "step": 38690 }, { "grad_norm": 0.49558496475219727, "learning_rate": 7.203555675605697e-05, "loss": 0.0372, "step": 38700 }, { "grad_norm": 0.43036943674087524, "learning_rate": 7.202071319644917e-05, "loss": 0.0439, "step": 38710 }, { "grad_norm": 0.4493442475795746, "learning_rate": 7.200586722868932e-05, "loss": 0.0355, "step": 38720 }, { "grad_norm": 0.3552796244621277, "learning_rate": 7.199101885440088e-05, "loss": 0.0294, "step": 38730 }, { "grad_norm": 0.38267508149147034, "learning_rate": 7.197616807520771e-05, "loss": 0.0297, "step": 38740 }, { "grad_norm": 0.4237494468688965, "learning_rate": 7.196131489273381e-05, "loss": 0.0257, "step": 38750 }, { "grad_norm": 0.26787859201431274, "learning_rate": 7.194645930860355e-05, "loss": 0.0307, "step": 38760 }, { "grad_norm": 0.3288700580596924, "learning_rate": 7.19316013244415e-05, "loss": 0.0316, "step": 38770 }, { "grad_norm": 0.2687138617038727, "learning_rate": 7.191674094187248e-05, "loss": 0.0317, "step": 38780 }, { "grad_norm": 0.42874887585639954, "learning_rate": 7.190187816252165e-05, "loss": 0.0247, "step": 38790 }, { "grad_norm": 0.26283764839172363, "learning_rate": 7.188701298801435e-05, "loss": 0.0223, "step": 38800 }, { "grad_norm": 0.4148457944393158, "learning_rate": 7.18721454199762e-05, "loss": 0.0352, "step": 38810 }, { "grad_norm": 0.3553873598575592, "learning_rate": 7.185727546003312e-05, "loss": 0.0329, "step": 38820 }, { "grad_norm": 0.5882664322853088, "learning_rate": 7.184240310981126e-05, "loss": 0.0401, "step": 38830 }, { "grad_norm": 0.33710771799087524, "learning_rate": 7.182752837093704e-05, "loss": 0.0345, "step": 38840 }, { "grad_norm": 0.3188742995262146, "learning_rate": 7.181265124503711e-05, "loss": 0.0432, "step": 38850 }, { "grad_norm": 0.2317187339067459, "learning_rate": 7.179777173373847e-05, "loss": 0.0283, "step": 38860 }, { "grad_norm": 0.2849605083465576, "learning_rate": 7.178288983866826e-05, "loss": 0.0287, "step": 38870 }, { "grad_norm": 0.4429578483104706, "learning_rate": 7.176800556145397e-05, "loss": 0.0398, "step": 38880 }, { "grad_norm": 0.37687918543815613, "learning_rate": 7.175311890372334e-05, "loss": 0.0326, "step": 38890 }, { "grad_norm": 0.4255314767360687, "learning_rate": 7.17382298671043e-05, "loss": 0.0299, "step": 38900 }, { "grad_norm": 0.5400063991546631, "learning_rate": 7.172333845322515e-05, "loss": 0.0354, "step": 38910 }, { "grad_norm": 0.34040871262550354, "learning_rate": 7.170844466371436e-05, "loss": 0.0382, "step": 38920 }, { "grad_norm": 0.46319350600242615, "learning_rate": 7.16935485002007e-05, "loss": 0.0322, "step": 38930 }, { "grad_norm": 0.25551408529281616, "learning_rate": 7.167864996431319e-05, "loss": 0.0382, "step": 38940 }, { "grad_norm": 0.6341482400894165, "learning_rate": 7.166374905768111e-05, "loss": 0.0307, "step": 38950 }, { "grad_norm": 0.4753851592540741, "learning_rate": 7.164884578193401e-05, "loss": 0.0377, "step": 38960 }, { "grad_norm": 0.36528316140174866, "learning_rate": 7.163394013870168e-05, "loss": 0.0311, "step": 38970 }, { "grad_norm": 0.5147162079811096, "learning_rate": 7.161903212961419e-05, "loss": 0.0456, "step": 38980 }, { "grad_norm": 0.288084477186203, "learning_rate": 7.160412175630183e-05, "loss": 0.031, "step": 38990 }, { "grad_norm": 0.39449506998062134, "learning_rate": 7.158920902039521e-05, "loss": 0.0293, "step": 39000 }, { "grad_norm": 0.3028942048549652, "learning_rate": 7.157429392352514e-05, "loss": 0.0264, "step": 39010 }, { "grad_norm": 0.34264710545539856, "learning_rate": 7.155937646732275e-05, "loss": 0.0308, "step": 39020 }, { "grad_norm": 0.40870627760887146, "learning_rate": 7.154445665341933e-05, "loss": 0.0289, "step": 39030 }, { "grad_norm": 0.4252842664718628, "learning_rate": 7.152953448344654e-05, "loss": 0.0293, "step": 39040 }, { "grad_norm": 0.3328307867050171, "learning_rate": 7.151460995903624e-05, "loss": 0.0315, "step": 39050 }, { "grad_norm": 0.5286023020744324, "learning_rate": 7.149968308182052e-05, "loss": 0.0296, "step": 39060 }, { "grad_norm": 0.22303056716918945, "learning_rate": 7.14847538534318e-05, "loss": 0.0292, "step": 39070 }, { "grad_norm": 0.3296741247177124, "learning_rate": 7.14698222755027e-05, "loss": 0.0295, "step": 39080 }, { "grad_norm": 0.3231498897075653, "learning_rate": 7.14548883496661e-05, "loss": 0.0334, "step": 39090 }, { "grad_norm": 0.31135210394859314, "learning_rate": 7.143995207755517e-05, "loss": 0.0309, "step": 39100 }, { "grad_norm": 0.3836762607097626, "learning_rate": 7.142501346080333e-05, "loss": 0.0365, "step": 39110 }, { "grad_norm": 0.43099263310432434, "learning_rate": 7.141007250104421e-05, "loss": 0.0289, "step": 39120 }, { "grad_norm": 0.3572421073913574, "learning_rate": 7.139512919991176e-05, "loss": 0.0399, "step": 39130 }, { "grad_norm": 0.32126501202583313, "learning_rate": 7.138018355904014e-05, "loss": 0.0304, "step": 39140 }, { "grad_norm": 0.6970922350883484, "learning_rate": 7.13652355800638e-05, "loss": 0.0423, "step": 39150 }, { "grad_norm": 0.2851942777633667, "learning_rate": 7.135028526461743e-05, "loss": 0.0326, "step": 39160 }, { "grad_norm": 0.35469961166381836, "learning_rate": 7.133533261433594e-05, "loss": 0.0257, "step": 39170 }, { "grad_norm": 0.5001198053359985, "learning_rate": 7.132037763085457e-05, "loss": 0.033, "step": 39180 }, { "grad_norm": 0.3922341465950012, "learning_rate": 7.130542031580875e-05, "loss": 0.0324, "step": 39190 }, { "grad_norm": 0.44596368074417114, "learning_rate": 7.12904606708342e-05, "loss": 0.0279, "step": 39200 }, { "grad_norm": 0.3866344392299652, "learning_rate": 7.127549869756687e-05, "loss": 0.0423, "step": 39210 }, { "grad_norm": 0.3383837640285492, "learning_rate": 7.126053439764299e-05, "loss": 0.0272, "step": 39220 }, { "grad_norm": 0.31984084844589233, "learning_rate": 7.124556777269904e-05, "loss": 0.0316, "step": 39230 }, { "grad_norm": 0.3232574462890625, "learning_rate": 7.123059882437174e-05, "loss": 0.0379, "step": 39240 }, { "grad_norm": 0.4241885840892792, "learning_rate": 7.121562755429807e-05, "loss": 0.0321, "step": 39250 }, { "grad_norm": 0.3233770728111267, "learning_rate": 7.120065396411528e-05, "loss": 0.0317, "step": 39260 }, { "grad_norm": 0.3650025725364685, "learning_rate": 7.118567805546084e-05, "loss": 0.031, "step": 39270 }, { "grad_norm": 0.470417857170105, "learning_rate": 7.117069982997248e-05, "loss": 0.0354, "step": 39280 }, { "grad_norm": 0.48548638820648193, "learning_rate": 7.115571928928825e-05, "loss": 0.05, "step": 39290 }, { "grad_norm": 0.38076916337013245, "learning_rate": 7.114073643504635e-05, "loss": 0.0357, "step": 39300 }, { "grad_norm": 0.3618517518043518, "learning_rate": 7.11257512688853e-05, "loss": 0.0313, "step": 39310 }, { "grad_norm": 0.5106916427612305, "learning_rate": 7.111076379244384e-05, "loss": 0.0315, "step": 39320 }, { "grad_norm": 0.26854121685028076, "learning_rate": 7.109577400736101e-05, "loss": 0.0275, "step": 39330 }, { "grad_norm": 0.3900196850299835, "learning_rate": 7.108078191527605e-05, "loss": 0.0313, "step": 39340 }, { "grad_norm": 0.3553740084171295, "learning_rate": 7.106578751782847e-05, "loss": 0.0372, "step": 39350 }, { "grad_norm": 0.35589370131492615, "learning_rate": 7.105079081665803e-05, "loss": 0.0301, "step": 39360 }, { "grad_norm": 0.4234510362148285, "learning_rate": 7.103579181340476e-05, "loss": 0.0326, "step": 39370 }, { "grad_norm": 0.2749917507171631, "learning_rate": 7.102079050970893e-05, "loss": 0.0304, "step": 39380 }, { "grad_norm": 0.26787471771240234, "learning_rate": 7.100578690721105e-05, "loss": 0.0348, "step": 39390 }, { "grad_norm": 0.6172516345977783, "learning_rate": 7.09907810075519e-05, "loss": 0.0389, "step": 39400 }, { "grad_norm": 0.505487322807312, "learning_rate": 7.097577281237249e-05, "loss": 0.0392, "step": 39410 }, { "grad_norm": 0.37655341625213623, "learning_rate": 7.09607623233141e-05, "loss": 0.0289, "step": 39420 }, { "grad_norm": 0.37269508838653564, "learning_rate": 7.094574954201827e-05, "loss": 0.0274, "step": 39430 }, { "grad_norm": 0.35292521119117737, "learning_rate": 7.093073447012675e-05, "loss": 0.0324, "step": 39440 }, { "grad_norm": 0.2623681426048279, "learning_rate": 7.09157171092816e-05, "loss": 0.0255, "step": 39450 }, { "grad_norm": 0.25931593775749207, "learning_rate": 7.090069746112504e-05, "loss": 0.0239, "step": 39460 }, { "grad_norm": 0.36411985754966736, "learning_rate": 7.088567552729965e-05, "loss": 0.0279, "step": 39470 }, { "grad_norm": 0.23150885105133057, "learning_rate": 7.087065130944818e-05, "loss": 0.0368, "step": 39480 }, { "grad_norm": 0.5088903307914734, "learning_rate": 7.085562480921366e-05, "loss": 0.0382, "step": 39490 }, { "grad_norm": 0.3155938982963562, "learning_rate": 7.084059602823937e-05, "loss": 0.0247, "step": 39500 }, { "grad_norm": 0.38682493567466736, "learning_rate": 7.082556496816882e-05, "loss": 0.0252, "step": 39510 }, { "grad_norm": 0.37488672137260437, "learning_rate": 7.081053163064582e-05, "loss": 0.034, "step": 39520 }, { "grad_norm": 0.41900286078453064, "learning_rate": 7.079549601731434e-05, "loss": 0.0364, "step": 39530 }, { "grad_norm": 0.3461170792579651, "learning_rate": 7.07804581298187e-05, "loss": 0.0326, "step": 39540 }, { "grad_norm": 0.45344051718711853, "learning_rate": 7.07654179698034e-05, "loss": 0.0488, "step": 39550 }, { "grad_norm": 0.4289153516292572, "learning_rate": 7.075037553891321e-05, "loss": 0.0374, "step": 39560 }, { "grad_norm": 0.37951353192329407, "learning_rate": 7.073533083879315e-05, "loss": 0.0319, "step": 39570 }, { "grad_norm": 0.4889448583126068, "learning_rate": 7.072028387108849e-05, "loss": 0.0339, "step": 39580 }, { "grad_norm": 0.28760814666748047, "learning_rate": 7.070523463744472e-05, "loss": 0.0361, "step": 39590 }, { "grad_norm": 0.3871575891971588, "learning_rate": 7.069018313950763e-05, "loss": 0.0293, "step": 39600 }, { "grad_norm": 0.3694900572299957, "learning_rate": 7.067512937892322e-05, "loss": 0.032, "step": 39610 }, { "grad_norm": 0.3124610483646393, "learning_rate": 7.066007335733775e-05, "loss": 0.0346, "step": 39620 }, { "grad_norm": 0.32336896657943726, "learning_rate": 7.064501507639772e-05, "loss": 0.0349, "step": 39630 }, { "grad_norm": 0.31307321786880493, "learning_rate": 7.062995453774987e-05, "loss": 0.0345, "step": 39640 }, { "grad_norm": 0.2783329486846924, "learning_rate": 7.061489174304121e-05, "loss": 0.0523, "step": 39650 }, { "grad_norm": 0.3615776598453522, "learning_rate": 7.0599826693919e-05, "loss": 0.0263, "step": 39660 }, { "grad_norm": 0.3656543493270874, "learning_rate": 7.05847593920307e-05, "loss": 0.0326, "step": 39670 }, { "grad_norm": 0.5109910368919373, "learning_rate": 7.056968983902406e-05, "loss": 0.0379, "step": 39680 }, { "grad_norm": 0.36892497539520264, "learning_rate": 7.055461803654706e-05, "loss": 0.0308, "step": 39690 }, { "grad_norm": 0.34794414043426514, "learning_rate": 7.053954398624794e-05, "loss": 0.0265, "step": 39700 }, { "grad_norm": 0.3086210787296295, "learning_rate": 7.052446768977518e-05, "loss": 0.0347, "step": 39710 }, { "grad_norm": 0.4413825571537018, "learning_rate": 7.050938914877748e-05, "loss": 0.0238, "step": 39720 }, { "grad_norm": 0.34418076276779175, "learning_rate": 7.049430836490382e-05, "loss": 0.0426, "step": 39730 }, { "grad_norm": 0.2353886514902115, "learning_rate": 7.047922533980342e-05, "loss": 0.0326, "step": 39740 }, { "grad_norm": 0.3317401111125946, "learning_rate": 7.046414007512571e-05, "loss": 0.0316, "step": 39750 }, { "grad_norm": 0.23946838080883026, "learning_rate": 7.044905257252042e-05, "loss": 0.0273, "step": 39760 }, { "grad_norm": 0.36108851432800293, "learning_rate": 7.04339628336375e-05, "loss": 0.0345, "step": 39770 }, { "grad_norm": 0.30655208230018616, "learning_rate": 7.041887086012711e-05, "loss": 0.0301, "step": 39780 }, { "grad_norm": 0.6849308609962463, "learning_rate": 7.040377665363969e-05, "loss": 0.0308, "step": 39790 }, { "grad_norm": 0.4334389865398407, "learning_rate": 7.038868021582594e-05, "loss": 0.0376, "step": 39800 }, { "grad_norm": 0.348750501871109, "learning_rate": 7.037358154833679e-05, "loss": 0.0329, "step": 39810 }, { "grad_norm": 0.4374648928642273, "learning_rate": 7.035848065282339e-05, "loss": 0.0377, "step": 39820 }, { "grad_norm": 0.46510323882102966, "learning_rate": 7.034337753093714e-05, "loss": 0.0268, "step": 39830 }, { "grad_norm": 0.27567189931869507, "learning_rate": 7.032827218432972e-05, "loss": 0.0285, "step": 39840 }, { "grad_norm": 0.3558785319328308, "learning_rate": 7.031316461465302e-05, "loss": 0.0261, "step": 39850 }, { "grad_norm": 0.4873988628387451, "learning_rate": 7.029805482355915e-05, "loss": 0.0261, "step": 39860 }, { "grad_norm": 0.36688560247421265, "learning_rate": 7.028294281270055e-05, "loss": 0.0295, "step": 39870 }, { "grad_norm": 0.3683277666568756, "learning_rate": 7.026782858372982e-05, "loss": 0.0381, "step": 39880 }, { "grad_norm": 0.33976301550865173, "learning_rate": 7.025271213829982e-05, "loss": 0.0327, "step": 39890 }, { "grad_norm": 0.3961979150772095, "learning_rate": 7.023759347806366e-05, "loss": 0.0252, "step": 39900 }, { "grad_norm": 0.33828186988830566, "learning_rate": 7.022247260467472e-05, "loss": 0.0322, "step": 39910 }, { "grad_norm": 0.3418874144554138, "learning_rate": 7.020734951978658e-05, "loss": 0.0328, "step": 39920 }, { "grad_norm": 0.33881422877311707, "learning_rate": 7.019222422505307e-05, "loss": 0.0333, "step": 39930 }, { "grad_norm": 0.32918694615364075, "learning_rate": 7.017709672212827e-05, "loss": 0.0291, "step": 39940 }, { "grad_norm": 0.29909801483154297, "learning_rate": 7.016196701266652e-05, "loss": 0.0345, "step": 39950 }, { "grad_norm": 0.6825445294380188, "learning_rate": 7.014683509832238e-05, "loss": 0.0389, "step": 39960 }, { "grad_norm": 0.34444761276245117, "learning_rate": 7.013170098075063e-05, "loss": 0.0436, "step": 39970 }, { "grad_norm": 0.4050128757953644, "learning_rate": 7.011656466160632e-05, "loss": 0.0315, "step": 39980 }, { "grad_norm": 0.25997158885002136, "learning_rate": 7.010142614254475e-05, "loss": 0.0261, "step": 39990 }, { "grad_norm": 0.31708040833473206, "learning_rate": 7.008628542522147e-05, "loss": 0.0311, "step": 40000 }, { "grad_norm": 0.6953687071800232, "learning_rate": 7.007114251129218e-05, "loss": 0.0358, "step": 40010 }, { "grad_norm": 0.4099067151546478, "learning_rate": 7.005599740241293e-05, "loss": 0.0277, "step": 40020 }, { "grad_norm": 0.4177510440349579, "learning_rate": 7.004085010023996e-05, "loss": 0.0391, "step": 40030 }, { "grad_norm": 0.3853084444999695, "learning_rate": 7.002570060642976e-05, "loss": 0.0335, "step": 40040 }, { "grad_norm": 0.3475624918937683, "learning_rate": 7.001054892263903e-05, "loss": 0.0331, "step": 40050 }, { "grad_norm": 0.3311547338962555, "learning_rate": 6.999539505052477e-05, "loss": 0.0328, "step": 40060 }, { "grad_norm": 0.4846521317958832, "learning_rate": 6.998023899174418e-05, "loss": 0.0329, "step": 40070 }, { "grad_norm": 0.30886730551719666, "learning_rate": 6.996508074795467e-05, "loss": 0.0257, "step": 40080 }, { "grad_norm": 0.5554499626159668, "learning_rate": 6.994992032081396e-05, "loss": 0.0419, "step": 40090 }, { "grad_norm": 0.21883508563041687, "learning_rate": 6.993475771197995e-05, "loss": 0.0302, "step": 40100 }, { "grad_norm": 0.25276440382003784, "learning_rate": 6.991959292311082e-05, "loss": 0.0307, "step": 40110 }, { "grad_norm": 0.3699369430541992, "learning_rate": 6.990442595586495e-05, "loss": 0.0308, "step": 40120 }, { "grad_norm": 0.394476056098938, "learning_rate": 6.988925681190098e-05, "loss": 0.0325, "step": 40130 }, { "grad_norm": 0.4366287291049957, "learning_rate": 6.987408549287778e-05, "loss": 0.0315, "step": 40140 }, { "grad_norm": 0.4543973207473755, "learning_rate": 6.985891200045449e-05, "loss": 0.0319, "step": 40150 }, { "grad_norm": 0.33081379532814026, "learning_rate": 6.984373633629045e-05, "loss": 0.0384, "step": 40160 }, { "grad_norm": 0.416840523481369, "learning_rate": 6.98285585020452e-05, "loss": 0.0274, "step": 40170 }, { "grad_norm": 0.2846469283103943, "learning_rate": 6.981337849937864e-05, "loss": 0.0295, "step": 40180 }, { "grad_norm": 0.30553773045539856, "learning_rate": 6.979819632995078e-05, "loss": 0.0297, "step": 40190 }, { "grad_norm": 0.32374894618988037, "learning_rate": 6.978301199542193e-05, "loss": 0.0303, "step": 40200 }, { "grad_norm": 0.33303573727607727, "learning_rate": 6.976782549745263e-05, "loss": 0.0298, "step": 40210 }, { "grad_norm": 0.4175799787044525, "learning_rate": 6.975263683770367e-05, "loss": 0.0322, "step": 40220 }, { "grad_norm": 0.3224183917045593, "learning_rate": 6.973744601783604e-05, "loss": 0.034, "step": 40230 }, { "grad_norm": 0.31026965379714966, "learning_rate": 6.972225303951097e-05, "loss": 0.0294, "step": 40240 }, { "grad_norm": 0.3726285398006439, "learning_rate": 6.970705790438998e-05, "loss": 0.0266, "step": 40250 }, { "grad_norm": 0.7188516855239868, "learning_rate": 6.969186061413477e-05, "loss": 0.0296, "step": 40260 }, { "grad_norm": 0.4159068167209625, "learning_rate": 6.967666117040727e-05, "loss": 0.0387, "step": 40270 }, { "grad_norm": 0.36855199933052063, "learning_rate": 6.966145957486972e-05, "loss": 0.0302, "step": 40280 }, { "grad_norm": 0.3768063187599182, "learning_rate": 6.964625582918449e-05, "loss": 0.0329, "step": 40290 }, { "grad_norm": 0.3904561698436737, "learning_rate": 6.963104993501425e-05, "loss": 0.0414, "step": 40300 }, { "grad_norm": 0.4484744966030121, "learning_rate": 6.961584189402192e-05, "loss": 0.0313, "step": 40310 }, { "grad_norm": 0.2403792440891266, "learning_rate": 6.96006317078706e-05, "loss": 0.0281, "step": 40320 }, { "grad_norm": 0.44802698493003845, "learning_rate": 6.958541937822367e-05, "loss": 0.033, "step": 40330 }, { "grad_norm": 0.4410144090652466, "learning_rate": 6.957020490674473e-05, "loss": 0.0299, "step": 40340 }, { "grad_norm": 0.37045684456825256, "learning_rate": 6.95549882950976e-05, "loss": 0.0328, "step": 40350 }, { "grad_norm": 0.623993456363678, "learning_rate": 6.953976954494635e-05, "loss": 0.0351, "step": 40360 }, { "grad_norm": 0.3466084599494934, "learning_rate": 6.952454865795528e-05, "loss": 0.0311, "step": 40370 }, { "grad_norm": 0.32343190908432007, "learning_rate": 6.950932563578892e-05, "loss": 0.0279, "step": 40380 }, { "grad_norm": 0.3441675901412964, "learning_rate": 6.949410048011206e-05, "loss": 0.0307, "step": 40390 }, { "grad_norm": 1.2046867609024048, "learning_rate": 6.947887319258966e-05, "loss": 0.039, "step": 40400 }, { "grad_norm": 0.27075329422950745, "learning_rate": 6.946364377488696e-05, "loss": 0.0327, "step": 40410 }, { "grad_norm": 0.301714152097702, "learning_rate": 6.944841222866947e-05, "loss": 0.0391, "step": 40420 }, { "grad_norm": 0.32992416620254517, "learning_rate": 6.943317855560284e-05, "loss": 0.0416, "step": 40430 }, { "grad_norm": 0.4220397472381592, "learning_rate": 6.941794275735302e-05, "loss": 0.0323, "step": 40440 }, { "grad_norm": 0.2847975790500641, "learning_rate": 6.94027048355862e-05, "loss": 0.027, "step": 40450 }, { "grad_norm": 0.26929348707199097, "learning_rate": 6.938746479196871e-05, "loss": 0.0325, "step": 40460 }, { "grad_norm": 0.36874401569366455, "learning_rate": 6.937222262816724e-05, "loss": 0.0256, "step": 40470 }, { "grad_norm": 0.389108806848526, "learning_rate": 6.935697834584865e-05, "loss": 0.037, "step": 40480 }, { "grad_norm": 0.3372257649898529, "learning_rate": 6.934173194667998e-05, "loss": 0.0295, "step": 40490 }, { "grad_norm": 0.31077268719673157, "learning_rate": 6.93264834323286e-05, "loss": 0.0302, "step": 40500 }, { "grad_norm": 0.3347148895263672, "learning_rate": 6.931123280446206e-05, "loss": 0.0313, "step": 40510 }, { "grad_norm": 0.32070326805114746, "learning_rate": 6.929598006474811e-05, "loss": 0.0262, "step": 40520 }, { "grad_norm": 0.36833348870277405, "learning_rate": 6.92807252148548e-05, "loss": 0.0363, "step": 40530 }, { "grad_norm": 0.3612001836299896, "learning_rate": 6.926546825645037e-05, "loss": 0.027, "step": 40540 }, { "grad_norm": 0.2283007651567459, "learning_rate": 6.92502091912033e-05, "loss": 0.0285, "step": 40550 }, { "grad_norm": 0.7317991852760315, "learning_rate": 6.923494802078229e-05, "loss": 0.0303, "step": 40560 }, { "grad_norm": 0.33626070618629456, "learning_rate": 6.921968474685628e-05, "loss": 0.029, "step": 40570 }, { "grad_norm": 0.5195502638816833, "learning_rate": 6.920441937109445e-05, "loss": 0.031, "step": 40580 }, { "grad_norm": 0.25574585795402527, "learning_rate": 6.91891518951662e-05, "loss": 0.0292, "step": 40590 }, { "grad_norm": 0.2770955562591553, "learning_rate": 6.917388232074114e-05, "loss": 0.0273, "step": 40600 }, { "grad_norm": 0.3490976393222809, "learning_rate": 6.915861064948914e-05, "loss": 0.0295, "step": 40610 }, { "grad_norm": 0.49834829568862915, "learning_rate": 6.914333688308025e-05, "loss": 0.0346, "step": 40620 }, { "grad_norm": 0.37942859530448914, "learning_rate": 6.912806102318487e-05, "loss": 0.0303, "step": 40630 }, { "grad_norm": 0.3190070390701294, "learning_rate": 6.911278307147347e-05, "loss": 0.0331, "step": 40640 }, { "grad_norm": 0.341877281665802, "learning_rate": 6.909750302961684e-05, "loss": 0.0364, "step": 40650 }, { "grad_norm": 0.3008157014846802, "learning_rate": 6.9082220899286e-05, "loss": 0.0451, "step": 40660 }, { "grad_norm": 0.4973756968975067, "learning_rate": 6.906693668215219e-05, "loss": 0.0384, "step": 40670 }, { "grad_norm": 0.315288245677948, "learning_rate": 6.905165037988683e-05, "loss": 0.0334, "step": 40680 }, { "grad_norm": 0.269363135099411, "learning_rate": 6.903636199416164e-05, "loss": 0.0281, "step": 40690 }, { "grad_norm": 0.41853970289230347, "learning_rate": 6.902107152664851e-05, "loss": 0.0273, "step": 40700 }, { "grad_norm": 0.3181133270263672, "learning_rate": 6.90057789790196e-05, "loss": 0.0263, "step": 40710 }, { "grad_norm": 0.32503196597099304, "learning_rate": 6.899048435294728e-05, "loss": 0.0314, "step": 40720 }, { "grad_norm": 0.30586886405944824, "learning_rate": 6.897518765010415e-05, "loss": 0.0284, "step": 40730 }, { "grad_norm": 0.29869359731674194, "learning_rate": 6.895988887216302e-05, "loss": 0.0283, "step": 40740 }, { "grad_norm": 0.3423612415790558, "learning_rate": 6.894458802079694e-05, "loss": 0.0355, "step": 40750 }, { "grad_norm": 0.36749008297920227, "learning_rate": 6.892928509767922e-05, "loss": 0.0291, "step": 40760 }, { "grad_norm": 0.2678740918636322, "learning_rate": 6.891398010448333e-05, "loss": 0.0265, "step": 40770 }, { "grad_norm": 0.38046354055404663, "learning_rate": 6.8898673042883e-05, "loss": 0.0351, "step": 40780 }, { "grad_norm": 0.44165366888046265, "learning_rate": 6.888336391455222e-05, "loss": 0.034, "step": 40790 }, { "grad_norm": 0.3875872492790222, "learning_rate": 6.886805272116513e-05, "loss": 0.0343, "step": 40800 }, { "grad_norm": 0.3995964825153351, "learning_rate": 6.885273946439617e-05, "loss": 0.0359, "step": 40810 }, { "grad_norm": 0.5084750056266785, "learning_rate": 6.883742414591998e-05, "loss": 0.0396, "step": 40820 }, { "grad_norm": 0.3133956789970398, "learning_rate": 6.88221067674114e-05, "loss": 0.0256, "step": 40830 }, { "grad_norm": 0.40791913866996765, "learning_rate": 6.88067873305455e-05, "loss": 0.0336, "step": 40840 }, { "grad_norm": 0.21550971269607544, "learning_rate": 6.879146583699765e-05, "loss": 0.029, "step": 40850 }, { "grad_norm": 0.24358247220516205, "learning_rate": 6.877614228844334e-05, "loss": 0.0278, "step": 40860 }, { "grad_norm": 0.41018611192703247, "learning_rate": 6.876081668655832e-05, "loss": 0.0344, "step": 40870 }, { "grad_norm": 0.5917119979858398, "learning_rate": 6.874548903301861e-05, "loss": 0.0411, "step": 40880 }, { "grad_norm": 0.3917681574821472, "learning_rate": 6.873015932950039e-05, "loss": 0.0356, "step": 40890 }, { "grad_norm": 0.47453707456588745, "learning_rate": 6.871482757768012e-05, "loss": 0.0341, "step": 40900 }, { "grad_norm": 0.3470626473426819, "learning_rate": 6.869949377923442e-05, "loss": 0.0345, "step": 40910 }, { "grad_norm": 0.5607727766036987, "learning_rate": 6.868415793584022e-05, "loss": 0.0319, "step": 40920 }, { "grad_norm": 0.38425007462501526, "learning_rate": 6.86688200491746e-05, "loss": 0.0315, "step": 40930 }, { "grad_norm": 0.3202134072780609, "learning_rate": 6.865348012091488e-05, "loss": 0.0282, "step": 40940 }, { "grad_norm": 0.2691190242767334, "learning_rate": 6.86381381527386e-05, "loss": 0.0311, "step": 40950 }, { "grad_norm": 0.34379205107688904, "learning_rate": 6.862279414632359e-05, "loss": 0.0284, "step": 40960 }, { "grad_norm": 0.35570159554481506, "learning_rate": 6.860744810334778e-05, "loss": 0.0343, "step": 40970 }, { "grad_norm": 0.28153863549232483, "learning_rate": 6.859210002548943e-05, "loss": 0.0262, "step": 40980 }, { "grad_norm": 0.2899901568889618, "learning_rate": 6.8576749914427e-05, "loss": 0.0238, "step": 40990 }, { "grad_norm": 0.4080057144165039, "learning_rate": 6.85613977718391e-05, "loss": 0.0331, "step": 41000 }, { "grad_norm": 0.413065642118454, "learning_rate": 6.854604359940464e-05, "loss": 0.0259, "step": 41010 }, { "grad_norm": 0.35918036103248596, "learning_rate": 6.853068739880276e-05, "loss": 0.024, "step": 41020 }, { "grad_norm": 0.39814692735671997, "learning_rate": 6.851532917171276e-05, "loss": 0.0381, "step": 41030 }, { "grad_norm": 0.4097086191177368, "learning_rate": 6.849996891981418e-05, "loss": 0.0385, "step": 41040 }, { "grad_norm": 0.32418903708457947, "learning_rate": 6.848460664478684e-05, "loss": 0.0338, "step": 41050 }, { "grad_norm": 0.26015329360961914, "learning_rate": 6.846924234831069e-05, "loss": 0.0298, "step": 41060 }, { "grad_norm": 0.4093790650367737, "learning_rate": 6.845387603206597e-05, "loss": 0.0349, "step": 41070 }, { "grad_norm": 0.29550090432167053, "learning_rate": 6.843850769773312e-05, "loss": 0.026, "step": 41080 }, { "grad_norm": 0.4072423279285431, "learning_rate": 6.842313734699278e-05, "loss": 0.0407, "step": 41090 }, { "grad_norm": 0.27095192670822144, "learning_rate": 6.840776498152584e-05, "loss": 0.0298, "step": 41100 }, { "grad_norm": 0.30660077929496765, "learning_rate": 6.83923906030134e-05, "loss": 0.0295, "step": 41110 }, { "grad_norm": 0.2910105288028717, "learning_rate": 6.837701421313677e-05, "loss": 0.0366, "step": 41120 }, { "grad_norm": 0.4539714753627777, "learning_rate": 6.83616358135775e-05, "loss": 0.0314, "step": 41130 }, { "grad_norm": 0.42898261547088623, "learning_rate": 6.834625540601734e-05, "loss": 0.0384, "step": 41140 }, { "grad_norm": 0.2851185202598572, "learning_rate": 6.833087299213829e-05, "loss": 0.0273, "step": 41150 }, { "grad_norm": 0.3374721109867096, "learning_rate": 6.83154885736225e-05, "loss": 0.038, "step": 41160 }, { "grad_norm": 0.5918672680854797, "learning_rate": 6.830010215215244e-05, "loss": 0.0312, "step": 41170 }, { "grad_norm": 0.3350222408771515, "learning_rate": 6.828471372941072e-05, "loss": 0.0309, "step": 41180 }, { "grad_norm": 0.4437936246395111, "learning_rate": 6.826932330708018e-05, "loss": 0.0347, "step": 41190 }, { "grad_norm": 0.3048757314682007, "learning_rate": 6.825393088684393e-05, "loss": 0.0253, "step": 41200 }, { "grad_norm": 0.3012840747833252, "learning_rate": 6.823853647038524e-05, "loss": 0.023, "step": 41210 }, { "grad_norm": 0.31798118352890015, "learning_rate": 6.822314005938762e-05, "loss": 0.0241, "step": 41220 }, { "grad_norm": 0.3751595616340637, "learning_rate": 6.820774165553479e-05, "loss": 0.0328, "step": 41230 }, { "grad_norm": 0.4023016691207886, "learning_rate": 6.819234126051072e-05, "loss": 0.0287, "step": 41240 }, { "grad_norm": 0.30113524198532104, "learning_rate": 6.817693887599956e-05, "loss": 0.0317, "step": 41250 }, { "grad_norm": 0.43332451581954956, "learning_rate": 6.816153450368568e-05, "loss": 0.0436, "step": 41260 }, { "grad_norm": 0.3720483183860779, "learning_rate": 6.814612814525372e-05, "loss": 0.041, "step": 41270 }, { "grad_norm": 0.41398748755455017, "learning_rate": 6.813071980238846e-05, "loss": 0.0286, "step": 41280 }, { "grad_norm": 0.558875322341919, "learning_rate": 6.811530947677492e-05, "loss": 0.0341, "step": 41290 }, { "grad_norm": 0.27390140295028687, "learning_rate": 6.809989717009839e-05, "loss": 0.0227, "step": 41300 }, { "grad_norm": 0.4166881740093231, "learning_rate": 6.808448288404431e-05, "loss": 0.0324, "step": 41310 }, { "grad_norm": 0.36808153986930847, "learning_rate": 6.806906662029838e-05, "loss": 0.0311, "step": 41320 }, { "grad_norm": 0.42317283153533936, "learning_rate": 6.805364838054647e-05, "loss": 0.0337, "step": 41330 }, { "grad_norm": 0.42139360308647156, "learning_rate": 6.803822816647474e-05, "loss": 0.0306, "step": 41340 }, { "grad_norm": 0.3206526041030884, "learning_rate": 6.802280597976949e-05, "loss": 0.026, "step": 41350 }, { "grad_norm": 0.4010836184024811, "learning_rate": 6.800738182211727e-05, "loss": 0.0327, "step": 41360 }, { "grad_norm": 0.3526090383529663, "learning_rate": 6.799195569520485e-05, "loss": 0.0345, "step": 41370 }, { "grad_norm": 0.22094786167144775, "learning_rate": 6.79765276007192e-05, "loss": 0.0237, "step": 41380 }, { "grad_norm": 0.4056732952594757, "learning_rate": 6.796109754034753e-05, "loss": 0.0247, "step": 41390 }, { "grad_norm": 0.45522892475128174, "learning_rate": 6.794566551577724e-05, "loss": 0.0247, "step": 41400 }, { "grad_norm": 0.30872881412506104, "learning_rate": 6.793023152869594e-05, "loss": 0.0273, "step": 41410 }, { "grad_norm": 0.2867003381252289, "learning_rate": 6.791479558079147e-05, "loss": 0.0258, "step": 41420 }, { "grad_norm": 0.3368889391422272, "learning_rate": 6.789935767375191e-05, "loss": 0.0315, "step": 41430 }, { "grad_norm": 0.40555694699287415, "learning_rate": 6.78839178092655e-05, "loss": 0.0234, "step": 41440 }, { "grad_norm": 0.5194580554962158, "learning_rate": 6.786847598902072e-05, "loss": 0.0293, "step": 41450 }, { "grad_norm": 0.34165313839912415, "learning_rate": 6.785303221470627e-05, "loss": 0.0346, "step": 41460 }, { "grad_norm": 0.4415517747402191, "learning_rate": 6.783758648801108e-05, "loss": 0.0319, "step": 41470 }, { "grad_norm": 0.4981130361557007, "learning_rate": 6.782213881062422e-05, "loss": 0.0408, "step": 41480 }, { "grad_norm": 0.42405954003334045, "learning_rate": 6.780668918423508e-05, "loss": 0.0308, "step": 41490 }, { "grad_norm": 0.23506200313568115, "learning_rate": 6.779123761053317e-05, "loss": 0.0277, "step": 41500 }, { "grad_norm": 0.46023333072662354, "learning_rate": 6.777578409120826e-05, "loss": 0.0345, "step": 41510 }, { "grad_norm": 0.4067228138446808, "learning_rate": 6.776032862795033e-05, "loss": 0.0388, "step": 41520 }, { "grad_norm": 0.30690687894821167, "learning_rate": 6.774487122244956e-05, "loss": 0.0288, "step": 41530 }, { "grad_norm": 0.2811371386051178, "learning_rate": 6.772941187639637e-05, "loss": 0.0272, "step": 41540 }, { "grad_norm": 0.2896729111671448, "learning_rate": 6.771395059148134e-05, "loss": 0.0256, "step": 41550 }, { "grad_norm": 0.2700364291667938, "learning_rate": 6.76984873693953e-05, "loss": 0.0294, "step": 41560 }, { "grad_norm": 0.49469631910324097, "learning_rate": 6.768302221182931e-05, "loss": 0.0338, "step": 41570 }, { "grad_norm": 0.49446627497673035, "learning_rate": 6.766755512047457e-05, "loss": 0.0392, "step": 41580 }, { "grad_norm": 0.2734972834587097, "learning_rate": 6.765208609702259e-05, "loss": 0.0304, "step": 41590 }, { "grad_norm": 0.30280229449272156, "learning_rate": 6.763661514316499e-05, "loss": 0.0277, "step": 41600 }, { "grad_norm": 0.4603520333766937, "learning_rate": 6.76211422605937e-05, "loss": 0.0356, "step": 41610 }, { "grad_norm": 0.3464196026325226, "learning_rate": 6.760566745100076e-05, "loss": 0.028, "step": 41620 }, { "grad_norm": 0.46090710163116455, "learning_rate": 6.75901907160785e-05, "loss": 0.0309, "step": 41630 }, { "grad_norm": 0.326217383146286, "learning_rate": 6.757471205751943e-05, "loss": 0.031, "step": 41640 }, { "grad_norm": 0.36746880412101746, "learning_rate": 6.75592314770163e-05, "loss": 0.047, "step": 41650 }, { "grad_norm": 0.39507734775543213, "learning_rate": 6.7543748976262e-05, "loss": 0.0368, "step": 41660 }, { "grad_norm": 0.34875765442848206, "learning_rate": 6.752826455694968e-05, "loss": 0.03, "step": 41670 }, { "grad_norm": 0.9413027763366699, "learning_rate": 6.751277822077271e-05, "loss": 0.0326, "step": 41680 }, { "grad_norm": 0.3099231719970703, "learning_rate": 6.749728996942465e-05, "loss": 0.0286, "step": 41690 }, { "grad_norm": 0.37047868967056274, "learning_rate": 6.748179980459924e-05, "loss": 0.0307, "step": 41700 }, { "grad_norm": 0.3474710285663605, "learning_rate": 6.746630772799052e-05, "loss": 0.0296, "step": 41710 }, { "grad_norm": 0.3741549849510193, "learning_rate": 6.745081374129262e-05, "loss": 0.0304, "step": 41720 }, { "grad_norm": 0.3949868679046631, "learning_rate": 6.743531784619998e-05, "loss": 0.0381, "step": 41730 }, { "grad_norm": 0.37384098768234253, "learning_rate": 6.741982004440719e-05, "loss": 0.0341, "step": 41740 }, { "grad_norm": 0.3432050347328186, "learning_rate": 6.740432033760907e-05, "loss": 0.034, "step": 41750 }, { "grad_norm": 0.4753934144973755, "learning_rate": 6.738881872750066e-05, "loss": 0.028, "step": 41760 }, { "grad_norm": 0.3772694170475006, "learning_rate": 6.737331521577715e-05, "loss": 0.0302, "step": 41770 }, { "grad_norm": 0.36508628726005554, "learning_rate": 6.735780980413403e-05, "loss": 0.0259, "step": 41780 }, { "grad_norm": 0.3086213171482086, "learning_rate": 6.734230249426692e-05, "loss": 0.0329, "step": 41790 }, { "grad_norm": 0.34870123863220215, "learning_rate": 6.732679328787168e-05, "loss": 0.0256, "step": 41800 }, { "grad_norm": 0.5750762820243835, "learning_rate": 6.731128218664438e-05, "loss": 0.0322, "step": 41810 }, { "grad_norm": 0.3780549168586731, "learning_rate": 6.729576919228129e-05, "loss": 0.0238, "step": 41820 }, { "grad_norm": 0.49636131525039673, "learning_rate": 6.728025430647888e-05, "loss": 0.0265, "step": 41830 }, { "grad_norm": 0.39426127076148987, "learning_rate": 6.726473753093383e-05, "loss": 0.0324, "step": 41840 }, { "grad_norm": 0.3239477872848511, "learning_rate": 6.724921886734305e-05, "loss": 0.025, "step": 41850 }, { "grad_norm": 0.4070124328136444, "learning_rate": 6.723369831740363e-05, "loss": 0.0476, "step": 41860 }, { "grad_norm": 0.389034241437912, "learning_rate": 6.721817588281288e-05, "loss": 0.0281, "step": 41870 }, { "grad_norm": 0.37702861428260803, "learning_rate": 6.720265156526828e-05, "loss": 0.0282, "step": 41880 }, { "grad_norm": 0.34233617782592773, "learning_rate": 6.718712536646758e-05, "loss": 0.0286, "step": 41890 }, { "grad_norm": 0.5386214852333069, "learning_rate": 6.71715972881087e-05, "loss": 0.0406, "step": 41900 }, { "grad_norm": 0.3733367919921875, "learning_rate": 6.715606733188974e-05, "loss": 0.0319, "step": 41910 }, { "grad_norm": 0.32211247086524963, "learning_rate": 6.714053549950909e-05, "loss": 0.037, "step": 41920 }, { "grad_norm": 0.4486488997936249, "learning_rate": 6.71250017926652e-05, "loss": 0.0296, "step": 41930 }, { "grad_norm": 0.34418368339538574, "learning_rate": 6.71094662130569e-05, "loss": 0.0251, "step": 41940 }, { "grad_norm": 0.3324718773365021, "learning_rate": 6.709392876238307e-05, "loss": 0.0312, "step": 41950 }, { "grad_norm": 0.6177857518196106, "learning_rate": 6.707838944234289e-05, "loss": 0.0337, "step": 41960 }, { "grad_norm": 0.3190675973892212, "learning_rate": 6.706284825463574e-05, "loss": 0.0276, "step": 41970 }, { "grad_norm": 0.5606040358543396, "learning_rate": 6.704730520096114e-05, "loss": 0.0268, "step": 41980 }, { "grad_norm": 0.41504839062690735, "learning_rate": 6.703176028301888e-05, "loss": 0.0291, "step": 41990 }, { "grad_norm": 0.6690750122070312, "learning_rate": 6.701621350250892e-05, "loss": 0.0293, "step": 42000 }, { "grad_norm": 0.4180886149406433, "learning_rate": 6.700066486113144e-05, "loss": 0.0303, "step": 42010 }, { "grad_norm": 0.2516728341579437, "learning_rate": 6.69851143605868e-05, "loss": 0.0348, "step": 42020 }, { "grad_norm": 0.5797470808029175, "learning_rate": 6.696956200257561e-05, "loss": 0.0353, "step": 42030 }, { "grad_norm": 0.5503591299057007, "learning_rate": 6.695400778879863e-05, "loss": 0.039, "step": 42040 }, { "grad_norm": 0.44965308904647827, "learning_rate": 6.693845172095683e-05, "loss": 0.041, "step": 42050 }, { "grad_norm": 0.367963969707489, "learning_rate": 6.692289380075142e-05, "loss": 0.0326, "step": 42060 }, { "grad_norm": 0.5016217231750488, "learning_rate": 6.690733402988379e-05, "loss": 0.0323, "step": 42070 }, { "grad_norm": 0.441642165184021, "learning_rate": 6.689177241005553e-05, "loss": 0.038, "step": 42080 }, { "grad_norm": 0.24824464321136475, "learning_rate": 6.687620894296844e-05, "loss": 0.0399, "step": 42090 }, { "grad_norm": 0.4156804382801056, "learning_rate": 6.686064363032451e-05, "loss": 0.0357, "step": 42100 }, { "grad_norm": 0.3979932963848114, "learning_rate": 6.684507647382596e-05, "loss": 0.0384, "step": 42110 }, { "grad_norm": 0.3877836763858795, "learning_rate": 6.682950747517513e-05, "loss": 0.0376, "step": 42120 }, { "grad_norm": 0.35858166217803955, "learning_rate": 6.68139366360747e-05, "loss": 0.0335, "step": 42130 }, { "grad_norm": 0.30283457040786743, "learning_rate": 6.679836395822744e-05, "loss": 0.0261, "step": 42140 }, { "grad_norm": 0.39214566349983215, "learning_rate": 6.678278944333633e-05, "loss": 0.0364, "step": 42150 }, { "grad_norm": 0.4044189751148224, "learning_rate": 6.676721309310462e-05, "loss": 0.0304, "step": 42160 }, { "grad_norm": 0.29335322976112366, "learning_rate": 6.67516349092357e-05, "loss": 0.0232, "step": 42170 }, { "grad_norm": 0.31580308079719543, "learning_rate": 6.673605489343315e-05, "loss": 0.0251, "step": 42180 }, { "grad_norm": 0.42940640449523926, "learning_rate": 6.672047304740081e-05, "loss": 0.0283, "step": 42190 }, { "grad_norm": 0.2277381867170334, "learning_rate": 6.670488937284268e-05, "loss": 0.0249, "step": 42200 }, { "grad_norm": 0.5472488403320312, "learning_rate": 6.668930387146294e-05, "loss": 0.0291, "step": 42210 }, { "grad_norm": 0.7136987447738647, "learning_rate": 6.667371654496605e-05, "loss": 0.0271, "step": 42220 }, { "grad_norm": 0.36007145047187805, "learning_rate": 6.665812739505659e-05, "loss": 0.0285, "step": 42230 }, { "grad_norm": 0.31582120060920715, "learning_rate": 6.664253642343935e-05, "loss": 0.0324, "step": 42240 }, { "grad_norm": 0.3028200566768646, "learning_rate": 6.662694363181935e-05, "loss": 0.0244, "step": 42250 }, { "grad_norm": 0.4522041380405426, "learning_rate": 6.66113490219018e-05, "loss": 0.0326, "step": 42260 }, { "grad_norm": 0.2793685793876648, "learning_rate": 6.659575259539207e-05, "loss": 0.0285, "step": 42270 }, { "grad_norm": 0.2872945964336395, "learning_rate": 6.65801543539958e-05, "loss": 0.028, "step": 42280 }, { "grad_norm": 0.2864031195640564, "learning_rate": 6.65645542994188e-05, "loss": 0.0278, "step": 42290 }, { "grad_norm": 0.27067261934280396, "learning_rate": 6.654895243336702e-05, "loss": 0.0241, "step": 42300 }, { "grad_norm": 0.3441922962665558, "learning_rate": 6.65333487575467e-05, "loss": 0.029, "step": 42310 }, { "grad_norm": 0.32191866636276245, "learning_rate": 6.651774327366422e-05, "loss": 0.0235, "step": 42320 }, { "grad_norm": 0.4347820281982422, "learning_rate": 6.650213598342613e-05, "loss": 0.0246, "step": 42330 }, { "grad_norm": 0.3230386972427368, "learning_rate": 6.648652688853931e-05, "loss": 0.0358, "step": 42340 }, { "grad_norm": 0.4476831555366516, "learning_rate": 6.647091599071066e-05, "loss": 0.0306, "step": 42350 }, { "grad_norm": 0.3247947096824646, "learning_rate": 6.645530329164743e-05, "loss": 0.0305, "step": 42360 }, { "grad_norm": 0.3044135272502899, "learning_rate": 6.643968879305694e-05, "loss": 0.0381, "step": 42370 }, { "grad_norm": 0.37045133113861084, "learning_rate": 6.642407249664683e-05, "loss": 0.0299, "step": 42380 }, { "grad_norm": 0.25688737630844116, "learning_rate": 6.640845440412483e-05, "loss": 0.036, "step": 42390 }, { "grad_norm": 0.3506454527378082, "learning_rate": 6.639283451719893e-05, "loss": 0.0328, "step": 42400 }, { "grad_norm": 0.2784423232078552, "learning_rate": 6.637721283757729e-05, "loss": 0.0285, "step": 42410 }, { "grad_norm": 0.321707546710968, "learning_rate": 6.636158936696828e-05, "loss": 0.0265, "step": 42420 }, { "grad_norm": 0.38605692982673645, "learning_rate": 6.634596410708047e-05, "loss": 0.0349, "step": 42430 }, { "grad_norm": 0.2771446108818054, "learning_rate": 6.633033705962257e-05, "loss": 0.0315, "step": 42440 }, { "grad_norm": 0.3356744647026062, "learning_rate": 6.631470822630359e-05, "loss": 0.0293, "step": 42450 }, { "grad_norm": 0.34526604413986206, "learning_rate": 6.629907760883263e-05, "loss": 0.0329, "step": 42460 }, { "grad_norm": 0.26889461278915405, "learning_rate": 6.628344520891905e-05, "loss": 0.0248, "step": 42470 }, { "grad_norm": 0.3466644287109375, "learning_rate": 6.62678110282724e-05, "loss": 0.0228, "step": 42480 }, { "grad_norm": 0.3878507614135742, "learning_rate": 6.625217506860237e-05, "loss": 0.0349, "step": 42490 }, { "grad_norm": 0.921842634677887, "learning_rate": 6.623653733161892e-05, "loss": 0.031, "step": 42500 }, { "grad_norm": 0.4537620544433594, "learning_rate": 6.622089781903217e-05, "loss": 0.0287, "step": 42510 }, { "grad_norm": 0.37914177775382996, "learning_rate": 6.620525653255241e-05, "loss": 0.0336, "step": 42520 }, { "grad_norm": 0.3650011122226715, "learning_rate": 6.618961347389015e-05, "loss": 0.0302, "step": 42530 }, { "grad_norm": 0.389435738325119, "learning_rate": 6.617396864475613e-05, "loss": 0.0261, "step": 42540 }, { "grad_norm": 0.4148279130458832, "learning_rate": 6.61583220468612e-05, "loss": 0.0288, "step": 42550 }, { "grad_norm": 0.26418906450271606, "learning_rate": 6.614267368191645e-05, "loss": 0.0257, "step": 42560 }, { "grad_norm": 0.32120466232299805, "learning_rate": 6.61270235516332e-05, "loss": 0.0232, "step": 42570 }, { "grad_norm": 0.4083940386772156, "learning_rate": 6.611137165772287e-05, "loss": 0.0251, "step": 42580 }, { "grad_norm": 0.42344799637794495, "learning_rate": 6.60957180018972e-05, "loss": 0.0328, "step": 42590 }, { "grad_norm": 0.34896212816238403, "learning_rate": 6.608006258586797e-05, "loss": 0.0235, "step": 42600 }, { "grad_norm": 0.24805252254009247, "learning_rate": 6.606440541134731e-05, "loss": 0.0244, "step": 42610 }, { "grad_norm": 0.24591132998466492, "learning_rate": 6.60487464800474e-05, "loss": 0.0356, "step": 42620 }, { "grad_norm": 0.4160960614681244, "learning_rate": 6.603308579368071e-05, "loss": 0.0241, "step": 42630 }, { "grad_norm": 0.26926904916763306, "learning_rate": 6.601742335395987e-05, "loss": 0.028, "step": 42640 }, { "grad_norm": 0.31342631578445435, "learning_rate": 6.600175916259769e-05, "loss": 0.0266, "step": 42650 }, { "grad_norm": 0.33056601881980896, "learning_rate": 6.598609322130718e-05, "loss": 0.0266, "step": 42660 }, { "grad_norm": 0.6756239533424377, "learning_rate": 6.597042553180155e-05, "loss": 0.027, "step": 42670 }, { "grad_norm": 0.32241731882095337, "learning_rate": 6.59547560957942e-05, "loss": 0.0286, "step": 42680 }, { "grad_norm": 0.313335657119751, "learning_rate": 6.59390849149987e-05, "loss": 0.0411, "step": 42690 }, { "grad_norm": 0.40582698583602905, "learning_rate": 6.592341199112886e-05, "loss": 0.0239, "step": 42700 }, { "grad_norm": 0.3319680094718933, "learning_rate": 6.590773732589861e-05, "loss": 0.0325, "step": 42710 }, { "grad_norm": 0.37540775537490845, "learning_rate": 6.589206092102213e-05, "loss": 0.0408, "step": 42720 }, { "grad_norm": 0.6960861682891846, "learning_rate": 6.587638277821375e-05, "loss": 0.0391, "step": 42730 }, { "grad_norm": 0.4423030614852905, "learning_rate": 6.586070289918803e-05, "loss": 0.0363, "step": 42740 }, { "grad_norm": 0.4493591785430908, "learning_rate": 6.584502128565968e-05, "loss": 0.0335, "step": 42750 }, { "grad_norm": 0.451185941696167, "learning_rate": 6.582933793934363e-05, "loss": 0.033, "step": 42760 }, { "grad_norm": 0.21899646520614624, "learning_rate": 6.581365286195498e-05, "loss": 0.0285, "step": 42770 }, { "grad_norm": 0.3146655559539795, "learning_rate": 6.579796605520903e-05, "loss": 0.0297, "step": 42780 }, { "grad_norm": 0.533613920211792, "learning_rate": 6.578227752082127e-05, "loss": 0.0392, "step": 42790 }, { "grad_norm": 0.326161652803421, "learning_rate": 6.576658726050735e-05, "loss": 0.0328, "step": 42800 }, { "grad_norm": 0.28606677055358887, "learning_rate": 6.575089527598316e-05, "loss": 0.0323, "step": 42810 }, { "grad_norm": 0.7921464443206787, "learning_rate": 6.573520156896476e-05, "loss": 0.0346, "step": 42820 }, { "grad_norm": 0.36993831396102905, "learning_rate": 6.571950614116835e-05, "loss": 0.033, "step": 42830 }, { "grad_norm": 0.27686187624931335, "learning_rate": 6.570380899431039e-05, "loss": 0.0509, "step": 42840 }, { "grad_norm": 0.4058866500854492, "learning_rate": 6.568811013010749e-05, "loss": 0.0278, "step": 42850 }, { "grad_norm": 0.47880351543426514, "learning_rate": 6.567240955027645e-05, "loss": 0.03, "step": 42860 }, { "grad_norm": 0.33384624123573303, "learning_rate": 6.565670725653427e-05, "loss": 0.0278, "step": 42870 }, { "grad_norm": 0.38378167152404785, "learning_rate": 6.564100325059812e-05, "loss": 0.0282, "step": 42880 }, { "grad_norm": 0.3656301200389862, "learning_rate": 6.562529753418538e-05, "loss": 0.0312, "step": 42890 }, { "grad_norm": 0.3229188323020935, "learning_rate": 6.56095901090136e-05, "loss": 0.0272, "step": 42900 }, { "grad_norm": 0.45601433515548706, "learning_rate": 6.55938809768005e-05, "loss": 0.0344, "step": 42910 }, { "grad_norm": 0.37798112630844116, "learning_rate": 6.557817013926402e-05, "loss": 0.0287, "step": 42920 }, { "grad_norm": 0.36758068203926086, "learning_rate": 6.556245759812227e-05, "loss": 0.0418, "step": 42930 }, { "grad_norm": 0.46354299783706665, "learning_rate": 6.554674335509357e-05, "loss": 0.0241, "step": 42940 }, { "grad_norm": 0.2778562605381012, "learning_rate": 6.553102741189638e-05, "loss": 0.0257, "step": 42950 }, { "grad_norm": 0.76362144947052, "learning_rate": 6.55153097702494e-05, "loss": 0.035, "step": 42960 }, { "grad_norm": 0.37097737193107605, "learning_rate": 6.549959043187146e-05, "loss": 0.0328, "step": 42970 }, { "grad_norm": 0.36528509855270386, "learning_rate": 6.54838693984816e-05, "loss": 0.0275, "step": 42980 }, { "grad_norm": 0.3882787525653839, "learning_rate": 6.546814667179909e-05, "loss": 0.0276, "step": 42990 }, { "grad_norm": 0.2992623746395111, "learning_rate": 6.545242225354328e-05, "loss": 0.0267, "step": 43000 }, { "grad_norm": 0.29900676012039185, "learning_rate": 6.543669614543383e-05, "loss": 0.0251, "step": 43010 }, { "grad_norm": 0.34647509455680847, "learning_rate": 6.542096834919049e-05, "loss": 0.0266, "step": 43020 }, { "grad_norm": 0.2856422960758209, "learning_rate": 6.540523886653324e-05, "loss": 0.036, "step": 43030 }, { "grad_norm": 0.27161896228790283, "learning_rate": 6.538950769918222e-05, "loss": 0.0276, "step": 43040 }, { "grad_norm": 0.516362190246582, "learning_rate": 6.537377484885779e-05, "loss": 0.0324, "step": 43050 }, { "grad_norm": 0.398701936006546, "learning_rate": 6.535804031728045e-05, "loss": 0.0327, "step": 43060 }, { "grad_norm": 0.409037321805954, "learning_rate": 6.534230410617092e-05, "loss": 0.0304, "step": 43070 }, { "grad_norm": 0.3516198396682739, "learning_rate": 6.532656621725007e-05, "loss": 0.0289, "step": 43080 }, { "grad_norm": 0.3972102999687195, "learning_rate": 6.531082665223898e-05, "loss": 0.029, "step": 43090 }, { "grad_norm": 0.44824299216270447, "learning_rate": 6.529508541285889e-05, "loss": 0.0381, "step": 43100 }, { "grad_norm": 0.4095861613750458, "learning_rate": 6.527934250083128e-05, "loss": 0.0339, "step": 43110 }, { "grad_norm": 0.3513810336589813, "learning_rate": 6.526359791787775e-05, "loss": 0.0227, "step": 43120 }, { "grad_norm": 0.3083207309246063, "learning_rate": 6.524785166572009e-05, "loss": 0.0241, "step": 43130 }, { "grad_norm": 0.3498956859111786, "learning_rate": 6.523210374608029e-05, "loss": 0.0261, "step": 43140 }, { "grad_norm": 0.38538914918899536, "learning_rate": 6.521635416068054e-05, "loss": 0.0249, "step": 43150 }, { "grad_norm": 0.30953821539878845, "learning_rate": 6.520060291124317e-05, "loss": 0.0446, "step": 43160 }, { "grad_norm": 0.35839733481407166, "learning_rate": 6.518484999949072e-05, "loss": 0.0313, "step": 43170 }, { "grad_norm": 0.5781462788581848, "learning_rate": 6.516909542714591e-05, "loss": 0.0373, "step": 43180 }, { "grad_norm": 0.43984273076057434, "learning_rate": 6.515333919593161e-05, "loss": 0.026, "step": 43190 }, { "grad_norm": 0.23810137808322906, "learning_rate": 6.513758130757094e-05, "loss": 0.0304, "step": 43200 }, { "grad_norm": 0.409981906414032, "learning_rate": 6.512182176378713e-05, "loss": 0.0347, "step": 43210 }, { "grad_norm": 0.44712722301483154, "learning_rate": 6.510606056630362e-05, "loss": 0.0278, "step": 43220 }, { "grad_norm": 0.3033158779144287, "learning_rate": 6.509029771684405e-05, "loss": 0.0307, "step": 43230 }, { "grad_norm": 0.3717784285545349, "learning_rate": 6.50745332171322e-05, "loss": 0.0308, "step": 43240 }, { "grad_norm": 0.37068191170692444, "learning_rate": 6.505876706889207e-05, "loss": 0.0328, "step": 43250 }, { "grad_norm": 0.27982527017593384, "learning_rate": 6.504299927384781e-05, "loss": 0.0311, "step": 43260 }, { "grad_norm": 0.37263935804367065, "learning_rate": 6.502722983372378e-05, "loss": 0.0259, "step": 43270 }, { "grad_norm": 0.3021630346775055, "learning_rate": 6.501145875024446e-05, "loss": 0.028, "step": 43280 }, { "grad_norm": 0.48460811376571655, "learning_rate": 6.499568602513459e-05, "loss": 0.0337, "step": 43290 }, { "grad_norm": 0.48452407121658325, "learning_rate": 6.497991166011903e-05, "loss": 0.0284, "step": 43300 }, { "grad_norm": 0.30905038118362427, "learning_rate": 6.49641356569229e-05, "loss": 0.0337, "step": 43310 }, { "grad_norm": 0.8532750606536865, "learning_rate": 6.494835801727135e-05, "loss": 0.0338, "step": 43320 }, { "grad_norm": 0.36108464002609253, "learning_rate": 6.493257874288986e-05, "loss": 0.0288, "step": 43330 }, { "grad_norm": 0.23886120319366455, "learning_rate": 6.491679783550399e-05, "loss": 0.0436, "step": 43340 }, { "grad_norm": 0.4892690181732178, "learning_rate": 6.490101529683957e-05, "loss": 0.0349, "step": 43350 }, { "grad_norm": 0.39465123414993286, "learning_rate": 6.488523112862248e-05, "loss": 0.0328, "step": 43360 }, { "grad_norm": 0.31653329730033875, "learning_rate": 6.486944533257891e-05, "loss": 0.0333, "step": 43370 }, { "grad_norm": 0.30015844106674194, "learning_rate": 6.485365791043517e-05, "loss": 0.0353, "step": 43380 }, { "grad_norm": 0.5117334723472595, "learning_rate": 6.483786886391772e-05, "loss": 0.0332, "step": 43390 }, { "grad_norm": 0.2642243206501007, "learning_rate": 6.482207819475323e-05, "loss": 0.0242, "step": 43400 }, { "grad_norm": 0.3202149569988251, "learning_rate": 6.480628590466857e-05, "loss": 0.0234, "step": 43410 }, { "grad_norm": 0.25820645689964294, "learning_rate": 6.479049199539074e-05, "loss": 0.025, "step": 43420 }, { "grad_norm": 0.34476763010025024, "learning_rate": 6.477469646864693e-05, "loss": 0.0298, "step": 43430 }, { "grad_norm": 0.3328821659088135, "learning_rate": 6.475889932616454e-05, "loss": 0.0268, "step": 43440 }, { "grad_norm": 0.35353049635887146, "learning_rate": 6.474310056967111e-05, "loss": 0.0303, "step": 43450 }, { "grad_norm": 0.2801663279533386, "learning_rate": 6.472730020089437e-05, "loss": 0.0263, "step": 43460 }, { "grad_norm": 0.2205038070678711, "learning_rate": 6.471149822156222e-05, "loss": 0.0289, "step": 43470 }, { "grad_norm": 0.2840270698070526, "learning_rate": 6.469569463340275e-05, "loss": 0.0254, "step": 43480 }, { "grad_norm": 0.45493635535240173, "learning_rate": 6.467988943814418e-05, "loss": 0.03, "step": 43490 }, { "grad_norm": 0.5457745790481567, "learning_rate": 6.4664082637515e-05, "loss": 0.0303, "step": 43500 }, { "grad_norm": 0.42084065079689026, "learning_rate": 6.464827423324377e-05, "loss": 0.027, "step": 43510 }, { "grad_norm": 0.29809972643852234, "learning_rate": 6.463246422705931e-05, "loss": 0.0258, "step": 43520 }, { "grad_norm": 0.3528716266155243, "learning_rate": 6.461665262069055e-05, "loss": 0.0269, "step": 43530 }, { "grad_norm": 0.418489933013916, "learning_rate": 6.460083941586665e-05, "loss": 0.029, "step": 43540 }, { "grad_norm": 0.5203394293785095, "learning_rate": 6.45850246143169e-05, "loss": 0.0314, "step": 43550 }, { "grad_norm": 0.36112135648727417, "learning_rate": 6.456920821777077e-05, "loss": 0.0355, "step": 43560 }, { "grad_norm": 0.37867283821105957, "learning_rate": 6.455339022795796e-05, "loss": 0.0375, "step": 43570 }, { "grad_norm": 0.3541131317615509, "learning_rate": 6.453757064660825e-05, "loss": 0.0341, "step": 43580 }, { "grad_norm": 0.2937225103378296, "learning_rate": 6.452174947545169e-05, "loss": 0.0215, "step": 43590 }, { "grad_norm": 0.4412861764431, "learning_rate": 6.450592671621842e-05, "loss": 0.0308, "step": 43600 }, { "grad_norm": 0.2695440351963043, "learning_rate": 6.449010237063882e-05, "loss": 0.0293, "step": 43610 }, { "grad_norm": 0.3658888041973114, "learning_rate": 6.447427644044343e-05, "loss": 0.0316, "step": 43620 }, { "grad_norm": 0.34756290912628174, "learning_rate": 6.44584489273629e-05, "loss": 0.0238, "step": 43630 }, { "grad_norm": 0.2924520671367645, "learning_rate": 6.444261983312817e-05, "loss": 0.0253, "step": 43640 }, { "grad_norm": 0.27161112427711487, "learning_rate": 6.442678915947023e-05, "loss": 0.0266, "step": 43650 }, { "grad_norm": 0.4073106646537781, "learning_rate": 6.441095690812032e-05, "loss": 0.0283, "step": 43660 }, { "grad_norm": 0.31460443139076233, "learning_rate": 6.439512308080982e-05, "loss": 0.0326, "step": 43670 }, { "grad_norm": 0.2727237343788147, "learning_rate": 6.43792876792703e-05, "loss": 0.0252, "step": 43680 }, { "grad_norm": 0.22619861364364624, "learning_rate": 6.43634507052335e-05, "loss": 0.0292, "step": 43690 }, { "grad_norm": 0.36589735746383667, "learning_rate": 6.434761216043135e-05, "loss": 0.0316, "step": 43700 }, { "grad_norm": 0.3482687473297119, "learning_rate": 6.433177204659589e-05, "loss": 0.0302, "step": 43710 }, { "grad_norm": 0.7450503706932068, "learning_rate": 6.431593036545938e-05, "loss": 0.0279, "step": 43720 }, { "grad_norm": 0.3473968803882599, "learning_rate": 6.430008711875426e-05, "loss": 0.0241, "step": 43730 }, { "grad_norm": 0.23326149582862854, "learning_rate": 6.42842423082131e-05, "loss": 0.0267, "step": 43740 }, { "grad_norm": 0.35657453536987305, "learning_rate": 6.426839593556868e-05, "loss": 0.0294, "step": 43750 }, { "grad_norm": 0.6402817964553833, "learning_rate": 6.425254800255392e-05, "loss": 0.0342, "step": 43760 }, { "grad_norm": 0.3663076162338257, "learning_rate": 6.423669851090196e-05, "loss": 0.033, "step": 43770 }, { "grad_norm": 0.6776834726333618, "learning_rate": 6.422084746234604e-05, "loss": 0.0335, "step": 43780 }, { "grad_norm": 0.3046480417251587, "learning_rate": 6.420499485861965e-05, "loss": 0.029, "step": 43790 }, { "grad_norm": 0.5361875891685486, "learning_rate": 6.418914070145636e-05, "loss": 0.0373, "step": 43800 }, { "grad_norm": 0.5886992812156677, "learning_rate": 6.417328499258996e-05, "loss": 0.0327, "step": 43810 }, { "grad_norm": 0.25573259592056274, "learning_rate": 6.415742773375446e-05, "loss": 0.0247, "step": 43820 }, { "grad_norm": 0.3409925401210785, "learning_rate": 6.414156892668393e-05, "loss": 0.0295, "step": 43830 }, { "grad_norm": 0.3211260437965393, "learning_rate": 6.412570857311267e-05, "loss": 0.0337, "step": 43840 }, { "grad_norm": 0.23374059796333313, "learning_rate": 6.410984667477518e-05, "loss": 0.0349, "step": 43850 }, { "grad_norm": 0.32048600912094116, "learning_rate": 6.409398323340607e-05, "loss": 0.0294, "step": 43860 }, { "grad_norm": 0.286347359418869, "learning_rate": 6.407811825074012e-05, "loss": 0.0373, "step": 43870 }, { "grad_norm": 0.34898173809051514, "learning_rate": 6.406225172851234e-05, "loss": 0.0343, "step": 43880 }, { "grad_norm": 0.46379539370536804, "learning_rate": 6.404638366845786e-05, "loss": 0.0304, "step": 43890 }, { "grad_norm": 0.45397117733955383, "learning_rate": 6.403051407231196e-05, "loss": 0.0304, "step": 43900 }, { "grad_norm": 0.3764728307723999, "learning_rate": 6.401464294181016e-05, "loss": 0.0289, "step": 43910 }, { "grad_norm": 0.3502747714519501, "learning_rate": 6.399877027868808e-05, "loss": 0.0306, "step": 43920 }, { "grad_norm": 0.48615264892578125, "learning_rate": 6.39828960846815e-05, "loss": 0.0303, "step": 43930 }, { "grad_norm": 0.2814576029777527, "learning_rate": 6.396702036152644e-05, "loss": 0.0349, "step": 43940 }, { "grad_norm": 0.316527396440506, "learning_rate": 6.395114311095905e-05, "loss": 0.0366, "step": 43950 }, { "grad_norm": 0.4115491211414337, "learning_rate": 6.393526433471559e-05, "loss": 0.0315, "step": 43960 }, { "grad_norm": 0.714805543422699, "learning_rate": 6.391938403453262e-05, "loss": 0.0336, "step": 43970 }, { "grad_norm": 0.42410528659820557, "learning_rate": 6.390350221214671e-05, "loss": 0.0345, "step": 43980 }, { "grad_norm": 0.4969673752784729, "learning_rate": 6.38876188692947e-05, "loss": 0.0419, "step": 43990 }, { "grad_norm": 0.36025339365005493, "learning_rate": 6.387173400771358e-05, "loss": 0.0334, "step": 44000 }, { "grad_norm": 0.28960633277893066, "learning_rate": 6.385584762914047e-05, "loss": 0.0256, "step": 44010 }, { "grad_norm": 0.3108094036579132, "learning_rate": 6.38399597353127e-05, "loss": 0.0327, "step": 44020 }, { "grad_norm": 0.33750393986701965, "learning_rate": 6.382407032796775e-05, "loss": 0.0301, "step": 44030 }, { "grad_norm": 0.4317125082015991, "learning_rate": 6.380817940884325e-05, "loss": 0.0305, "step": 44040 }, { "grad_norm": 0.3805946111679077, "learning_rate": 6.379228697967702e-05, "loss": 0.0308, "step": 44050 }, { "grad_norm": 0.40093737840652466, "learning_rate": 6.3776393042207e-05, "loss": 0.0229, "step": 44060 }, { "grad_norm": 0.3564160168170929, "learning_rate": 6.376049759817137e-05, "loss": 0.0216, "step": 44070 }, { "grad_norm": 0.37254711985588074, "learning_rate": 6.37446006493084e-05, "loss": 0.0318, "step": 44080 }, { "grad_norm": 1.072479009628296, "learning_rate": 6.372870219735656e-05, "loss": 0.032, "step": 44090 }, { "grad_norm": 0.360576868057251, "learning_rate": 6.371280224405452e-05, "loss": 0.0371, "step": 44100 }, { "grad_norm": 0.4481652081012726, "learning_rate": 6.369690079114101e-05, "loss": 0.0303, "step": 44110 }, { "grad_norm": 0.4426085948944092, "learning_rate": 6.368099784035504e-05, "loss": 0.0314, "step": 44120 }, { "grad_norm": 0.37655404210090637, "learning_rate": 6.366509339343572e-05, "loss": 0.0287, "step": 44130 }, { "grad_norm": 0.5068251490592957, "learning_rate": 6.364918745212232e-05, "loss": 0.0283, "step": 44140 }, { "grad_norm": 0.41697803139686584, "learning_rate": 6.363328001815431e-05, "loss": 0.0306, "step": 44150 }, { "grad_norm": 0.3489232361316681, "learning_rate": 6.361737109327128e-05, "loss": 0.0328, "step": 44160 }, { "grad_norm": 0.3583723306655884, "learning_rate": 6.360146067921301e-05, "loss": 0.0287, "step": 44170 }, { "grad_norm": 0.258199006319046, "learning_rate": 6.35855487777195e-05, "loss": 0.0292, "step": 44180 }, { "grad_norm": 0.2952999174594879, "learning_rate": 6.356963539053076e-05, "loss": 0.0382, "step": 44190 }, { "grad_norm": 0.29979345202445984, "learning_rate": 6.355372051938711e-05, "loss": 0.0352, "step": 44200 }, { "grad_norm": 0.3240439295768738, "learning_rate": 6.353780416602894e-05, "loss": 0.0352, "step": 44210 }, { "grad_norm": 0.7097965478897095, "learning_rate": 6.352188633219689e-05, "loss": 0.0262, "step": 44220 }, { "grad_norm": 0.5399842262268066, "learning_rate": 6.350596701963166e-05, "loss": 0.0321, "step": 44230 }, { "grad_norm": 0.37096932530403137, "learning_rate": 6.349004623007419e-05, "loss": 0.0312, "step": 44240 }, { "grad_norm": 0.44831398129463196, "learning_rate": 6.347412396526555e-05, "loss": 0.0339, "step": 44250 }, { "grad_norm": 0.4522140324115753, "learning_rate": 6.345820022694696e-05, "loss": 0.0388, "step": 44260 }, { "grad_norm": 0.2979099154472351, "learning_rate": 6.344227501685984e-05, "loss": 0.0275, "step": 44270 }, { "grad_norm": 0.4741308391094208, "learning_rate": 6.342634833674572e-05, "loss": 0.0341, "step": 44280 }, { "grad_norm": 0.5247727036476135, "learning_rate": 6.341042018834635e-05, "loss": 0.0328, "step": 44290 }, { "grad_norm": 0.32219502329826355, "learning_rate": 6.339449057340359e-05, "loss": 0.0269, "step": 44300 }, { "grad_norm": 0.2520892024040222, "learning_rate": 6.337855949365945e-05, "loss": 0.0262, "step": 44310 }, { "grad_norm": 0.26217129826545715, "learning_rate": 6.336262695085619e-05, "loss": 0.0327, "step": 44320 }, { "grad_norm": 0.5248078107833862, "learning_rate": 6.334669294673612e-05, "loss": 0.032, "step": 44330 }, { "grad_norm": 0.1841268390417099, "learning_rate": 6.333075748304179e-05, "loss": 0.0412, "step": 44340 }, { "grad_norm": 0.34008753299713135, "learning_rate": 6.331482056151585e-05, "loss": 0.0315, "step": 44350 }, { "grad_norm": 0.3450932502746582, "learning_rate": 6.329888218390117e-05, "loss": 0.0314, "step": 44360 }, { "grad_norm": 0.21030695736408234, "learning_rate": 6.328294235194072e-05, "loss": 0.0264, "step": 44370 }, { "grad_norm": 0.39144980907440186, "learning_rate": 6.326700106737765e-05, "loss": 0.0334, "step": 44380 }, { "grad_norm": 0.4494878947734833, "learning_rate": 6.325105833195531e-05, "loss": 0.028, "step": 44390 }, { "grad_norm": 0.3058971166610718, "learning_rate": 6.323511414741715e-05, "loss": 0.0303, "step": 44400 }, { "grad_norm": 0.43596330285072327, "learning_rate": 6.321916851550678e-05, "loss": 0.0411, "step": 44410 }, { "grad_norm": 0.314107745885849, "learning_rate": 6.320322143796806e-05, "loss": 0.0345, "step": 44420 }, { "grad_norm": 0.6516162157058716, "learning_rate": 6.318727291654488e-05, "loss": 0.025, "step": 44430 }, { "grad_norm": 0.28830304741859436, "learning_rate": 6.317132295298134e-05, "loss": 0.0266, "step": 44440 }, { "grad_norm": 0.3341674208641052, "learning_rate": 6.315537154902173e-05, "loss": 0.0228, "step": 44450 }, { "grad_norm": 0.5043071508407593, "learning_rate": 6.313941870641048e-05, "loss": 0.0304, "step": 44460 }, { "grad_norm": 0.3585417866706848, "learning_rate": 6.312346442689214e-05, "loss": 0.0308, "step": 44470 }, { "grad_norm": 0.3418324589729309, "learning_rate": 6.310750871221147e-05, "loss": 0.0365, "step": 44480 }, { "grad_norm": 0.36595892906188965, "learning_rate": 6.309155156411335e-05, "loss": 0.024, "step": 44490 }, { "grad_norm": 0.3723853528499603, "learning_rate": 6.307559298434284e-05, "loss": 0.0277, "step": 44500 }, { "grad_norm": 0.30113524198532104, "learning_rate": 6.305963297464512e-05, "loss": 0.0292, "step": 44510 }, { "grad_norm": 0.3990996479988098, "learning_rate": 6.304367153676561e-05, "loss": 0.0257, "step": 44520 }, { "grad_norm": 0.2820131480693817, "learning_rate": 6.302770867244978e-05, "loss": 0.0209, "step": 44530 }, { "grad_norm": 0.2787264883518219, "learning_rate": 6.301174438344328e-05, "loss": 0.0318, "step": 44540 }, { "grad_norm": 0.6101105213165283, "learning_rate": 6.299577867149202e-05, "loss": 0.0445, "step": 44550 }, { "grad_norm": 0.32891029119491577, "learning_rate": 6.297981153834192e-05, "loss": 0.0326, "step": 44560 }, { "grad_norm": 0.2533475160598755, "learning_rate": 6.296384298573916e-05, "loss": 0.0371, "step": 44570 }, { "grad_norm": 0.4063397943973541, "learning_rate": 6.294787301543001e-05, "loss": 0.0473, "step": 44580 }, { "grad_norm": 0.3270723521709442, "learning_rate": 6.293190162916095e-05, "loss": 0.0275, "step": 44590 }, { "grad_norm": 0.29469209909439087, "learning_rate": 6.291592882867855e-05, "loss": 0.0296, "step": 44600 }, { "grad_norm": 0.5798937678337097, "learning_rate": 6.28999546157296e-05, "loss": 0.0323, "step": 44610 }, { "grad_norm": 0.25740429759025574, "learning_rate": 6.288397899206102e-05, "loss": 0.0216, "step": 44620 }, { "grad_norm": 0.573202908039093, "learning_rate": 6.286800195941984e-05, "loss": 0.0413, "step": 44630 }, { "grad_norm": 0.40876078605651855, "learning_rate": 6.285202351955334e-05, "loss": 0.0348, "step": 44640 }, { "grad_norm": 0.3077259957790375, "learning_rate": 6.283604367420887e-05, "loss": 0.0364, "step": 44650 }, { "grad_norm": 0.26771509647369385, "learning_rate": 6.282006242513394e-05, "loss": 0.0274, "step": 44660 }, { "grad_norm": 0.3333996534347534, "learning_rate": 6.280407977407628e-05, "loss": 0.0352, "step": 44670 }, { "grad_norm": 0.3981972336769104, "learning_rate": 6.27880957227837e-05, "loss": 0.031, "step": 44680 }, { "grad_norm": 0.3553692102432251, "learning_rate": 6.27721102730042e-05, "loss": 0.0308, "step": 44690 }, { "grad_norm": 0.3244277834892273, "learning_rate": 6.275612342648592e-05, "loss": 0.0307, "step": 44700 }, { "grad_norm": 0.3899661600589752, "learning_rate": 6.274013518497716e-05, "loss": 0.0305, "step": 44710 }, { "grad_norm": 0.34752124547958374, "learning_rate": 6.272414555022636e-05, "loss": 0.0253, "step": 44720 }, { "grad_norm": 0.5892892479896545, "learning_rate": 6.270815452398215e-05, "loss": 0.0297, "step": 44730 }, { "grad_norm": 0.2618739604949951, "learning_rate": 6.269216210799326e-05, "loss": 0.031, "step": 44740 }, { "grad_norm": 0.5096508264541626, "learning_rate": 6.26761683040086e-05, "loss": 0.0403, "step": 44750 }, { "grad_norm": 0.3686636984348297, "learning_rate": 6.266017311377723e-05, "loss": 0.0368, "step": 44760 }, { "grad_norm": 0.3284100890159607, "learning_rate": 6.264417653904839e-05, "loss": 0.0393, "step": 44770 }, { "grad_norm": 0.3395973742008209, "learning_rate": 6.262817858157139e-05, "loss": 0.0314, "step": 44780 }, { "grad_norm": 0.6821990013122559, "learning_rate": 6.261217924309576e-05, "loss": 0.0491, "step": 44790 }, { "grad_norm": 0.43814313411712646, "learning_rate": 6.259617852537118e-05, "loss": 0.0258, "step": 44800 }, { "grad_norm": 0.4649648368358612, "learning_rate": 6.258017643014747e-05, "loss": 0.03, "step": 44810 }, { "grad_norm": 0.2746663987636566, "learning_rate": 6.256417295917456e-05, "loss": 0.0414, "step": 44820 }, { "grad_norm": 0.3335816562175751, "learning_rate": 6.254816811420258e-05, "loss": 0.0311, "step": 44830 }, { "grad_norm": 0.334651380777359, "learning_rate": 6.253216189698183e-05, "loss": 0.0301, "step": 44840 }, { "grad_norm": 0.26705077290534973, "learning_rate": 6.251615430926267e-05, "loss": 0.0287, "step": 44850 }, { "grad_norm": 0.46056413650512695, "learning_rate": 6.25001453527957e-05, "loss": 0.0259, "step": 44860 }, { "grad_norm": 0.32546699047088623, "learning_rate": 6.248413502933164e-05, "loss": 0.0317, "step": 44870 }, { "grad_norm": 0.32568782567977905, "learning_rate": 6.246812334062133e-05, "loss": 0.0295, "step": 44880 }, { "grad_norm": 0.47356995940208435, "learning_rate": 6.245211028841579e-05, "loss": 0.036, "step": 44890 }, { "grad_norm": 0.431306391954422, "learning_rate": 6.24360958744662e-05, "loss": 0.032, "step": 44900 }, { "grad_norm": 0.23199260234832764, "learning_rate": 6.242008010052387e-05, "loss": 0.025, "step": 44910 }, { "grad_norm": 0.31714174151420593, "learning_rate": 6.240406296834024e-05, "loss": 0.0342, "step": 44920 }, { "grad_norm": 0.3041900098323822, "learning_rate": 6.238804447966694e-05, "loss": 0.024, "step": 44930 }, { "grad_norm": 1.000173568725586, "learning_rate": 6.237202463625573e-05, "loss": 0.0316, "step": 44940 }, { "grad_norm": 0.43432340025901794, "learning_rate": 6.235600343985848e-05, "loss": 0.0373, "step": 44950 }, { "grad_norm": 0.1891235113143921, "learning_rate": 6.233998089222729e-05, "loss": 0.0254, "step": 44960 }, { "grad_norm": 0.29013386368751526, "learning_rate": 6.232395699511433e-05, "loss": 0.0357, "step": 44970 }, { "grad_norm": 0.30468782782554626, "learning_rate": 6.230793175027199e-05, "loss": 0.0381, "step": 44980 }, { "grad_norm": 0.34626343846321106, "learning_rate": 6.22919051594527e-05, "loss": 0.0271, "step": 44990 }, { "grad_norm": 0.48063457012176514, "learning_rate": 6.227587722440917e-05, "loss": 0.0366, "step": 45000 }, { "grad_norm": 0.4105531573295593, "learning_rate": 6.225984794689414e-05, "loss": 0.0269, "step": 45010 }, { "grad_norm": 0.5357885360717773, "learning_rate": 6.22438173286606e-05, "loss": 0.032, "step": 45020 }, { "grad_norm": 0.32795897126197815, "learning_rate": 6.222778537146159e-05, "loss": 0.0318, "step": 45030 }, { "grad_norm": 0.34829965233802795, "learning_rate": 6.221175207705037e-05, "loss": 0.0252, "step": 45040 }, { "grad_norm": 0.377785325050354, "learning_rate": 6.219571744718027e-05, "loss": 0.0237, "step": 45050 }, { "grad_norm": 0.25809356570243835, "learning_rate": 6.217968148360487e-05, "loss": 0.0327, "step": 45060 }, { "grad_norm": 0.3638227880001068, "learning_rate": 6.216364418807782e-05, "loss": 0.0256, "step": 45070 }, { "grad_norm": 0.34031742811203003, "learning_rate": 6.214760556235292e-05, "loss": 0.0286, "step": 45080 }, { "grad_norm": 0.5367836952209473, "learning_rate": 6.213156560818414e-05, "loss": 0.0316, "step": 45090 }, { "grad_norm": 0.36131203174591064, "learning_rate": 6.211552432732559e-05, "loss": 0.0315, "step": 45100 }, { "grad_norm": 0.26757144927978516, "learning_rate": 6.20994817215315e-05, "loss": 0.0291, "step": 45110 }, { "grad_norm": 0.2771430015563965, "learning_rate": 6.208343779255628e-05, "loss": 0.0262, "step": 45120 }, { "grad_norm": 0.42966529726982117, "learning_rate": 6.206739254215449e-05, "loss": 0.0285, "step": 45130 }, { "grad_norm": 0.2919232249259949, "learning_rate": 6.205134597208077e-05, "loss": 0.0251, "step": 45140 }, { "grad_norm": 0.3919123411178589, "learning_rate": 6.203529808409e-05, "loss": 0.0224, "step": 45150 }, { "grad_norm": 0.26835617423057556, "learning_rate": 6.201924887993712e-05, "loss": 0.0378, "step": 45160 }, { "grad_norm": 0.3775462806224823, "learning_rate": 6.200319836137725e-05, "loss": 0.0373, "step": 45170 }, { "grad_norm": 0.22202087938785553, "learning_rate": 6.198714653016565e-05, "loss": 0.032, "step": 45180 }, { "grad_norm": 0.2877192497253418, "learning_rate": 6.197109338805774e-05, "loss": 0.0268, "step": 45190 }, { "grad_norm": 0.35974228382110596, "learning_rate": 6.195503893680903e-05, "loss": 0.03, "step": 45200 }, { "grad_norm": 0.28286054730415344, "learning_rate": 6.193898317817524e-05, "loss": 0.0262, "step": 45210 }, { "grad_norm": 0.3689280152320862, "learning_rate": 6.192292611391221e-05, "loss": 0.0204, "step": 45220 }, { "grad_norm": 0.2912259101867676, "learning_rate": 6.190686774577591e-05, "loss": 0.0318, "step": 45230 }, { "grad_norm": 0.42984625697135925, "learning_rate": 6.189080807552245e-05, "loss": 0.033, "step": 45240 }, { "grad_norm": 0.2636837661266327, "learning_rate": 6.187474710490809e-05, "loss": 0.0287, "step": 45250 }, { "grad_norm": 0.30386316776275635, "learning_rate": 6.185868483568926e-05, "loss": 0.0276, "step": 45260 }, { "grad_norm": 0.2971532344818115, "learning_rate": 6.184262126962245e-05, "loss": 0.0265, "step": 45270 }, { "grad_norm": 0.42527198791503906, "learning_rate": 6.182655640846442e-05, "loss": 0.0273, "step": 45280 }, { "grad_norm": 0.3741663694381714, "learning_rate": 6.181049025397196e-05, "loss": 0.0269, "step": 45290 }, { "grad_norm": 0.288004606962204, "learning_rate": 6.179442280790202e-05, "loss": 0.0398, "step": 45300 }, { "grad_norm": 0.30798962712287903, "learning_rate": 6.177835407201174e-05, "loss": 0.0312, "step": 45310 }, { "grad_norm": 0.23711761832237244, "learning_rate": 6.176228404805839e-05, "loss": 0.0229, "step": 45320 }, { "grad_norm": 0.36329081654548645, "learning_rate": 6.174621273779932e-05, "loss": 0.0257, "step": 45330 }, { "grad_norm": 0.41320982575416565, "learning_rate": 6.17301401429921e-05, "loss": 0.0298, "step": 45340 }, { "grad_norm": 0.2925023138523102, "learning_rate": 6.171406626539436e-05, "loss": 0.0265, "step": 45350 }, { "grad_norm": 0.4263628125190735, "learning_rate": 6.169799110676398e-05, "loss": 0.0463, "step": 45360 }, { "grad_norm": 0.3996313512325287, "learning_rate": 6.168191466885885e-05, "loss": 0.0278, "step": 45370 }, { "grad_norm": 0.26900947093963623, "learning_rate": 6.166583695343711e-05, "loss": 0.0219, "step": 45380 }, { "grad_norm": 0.30367279052734375, "learning_rate": 6.164975796225698e-05, "loss": 0.0271, "step": 45390 }, { "grad_norm": 0.26515665650367737, "learning_rate": 6.163367769707683e-05, "loss": 0.0273, "step": 45400 }, { "grad_norm": 0.22379255294799805, "learning_rate": 6.161759615965519e-05, "loss": 0.0325, "step": 45410 }, { "grad_norm": 0.2745187282562256, "learning_rate": 6.16015133517507e-05, "loss": 0.0238, "step": 45420 }, { "grad_norm": 0.25483670830726624, "learning_rate": 6.158542927512214e-05, "loss": 0.0253, "step": 45430 }, { "grad_norm": 0.2630244791507721, "learning_rate": 6.156934393152846e-05, "loss": 0.0271, "step": 45440 }, { "grad_norm": 0.2880322337150574, "learning_rate": 6.15532573227287e-05, "loss": 0.0341, "step": 45450 }, { "grad_norm": 0.45821112394332886, "learning_rate": 6.153716945048212e-05, "loss": 0.0325, "step": 45460 }, { "grad_norm": 0.4740857481956482, "learning_rate": 6.152108031654802e-05, "loss": 0.0326, "step": 45470 }, { "grad_norm": 0.3695429265499115, "learning_rate": 6.15049899226859e-05, "loss": 0.0323, "step": 45480 }, { "grad_norm": 0.4864901304244995, "learning_rate": 6.148889827065537e-05, "loss": 0.0285, "step": 45490 }, { "grad_norm": 0.4058780372142792, "learning_rate": 6.147280536221622e-05, "loss": 0.024, "step": 45500 }, { "grad_norm": 0.36212703585624695, "learning_rate": 6.145671119912832e-05, "loss": 0.0264, "step": 45510 }, { "grad_norm": 0.3055894076824188, "learning_rate": 6.144061578315169e-05, "loss": 0.0235, "step": 45520 }, { "grad_norm": 0.2512526512145996, "learning_rate": 6.142451911604654e-05, "loss": 0.0204, "step": 45530 }, { "grad_norm": 0.3647357225418091, "learning_rate": 6.140842119957315e-05, "loss": 0.0235, "step": 45540 }, { "grad_norm": 0.37088823318481445, "learning_rate": 6.139232203549197e-05, "loss": 0.0258, "step": 45550 }, { "grad_norm": 0.23488663136959076, "learning_rate": 6.137622162556357e-05, "loss": 0.0243, "step": 45560 }, { "grad_norm": 0.3631010055541992, "learning_rate": 6.13601199715487e-05, "loss": 0.0283, "step": 45570 }, { "grad_norm": 0.3591521084308624, "learning_rate": 6.134401707520816e-05, "loss": 0.0216, "step": 45580 }, { "grad_norm": 0.26858338713645935, "learning_rate": 6.132791293830298e-05, "loss": 0.0227, "step": 45590 }, { "grad_norm": 0.3109428584575653, "learning_rate": 6.131180756259428e-05, "loss": 0.0274, "step": 45600 }, { "grad_norm": 0.2604982852935791, "learning_rate": 6.129570094984331e-05, "loss": 0.028, "step": 45610 }, { "grad_norm": 0.34633728861808777, "learning_rate": 6.127959310181145e-05, "loss": 0.0323, "step": 45620 }, { "grad_norm": 0.4443172514438629, "learning_rate": 6.126348402026026e-05, "loss": 0.0309, "step": 45630 }, { "grad_norm": 0.31257423758506775, "learning_rate": 6.12473737069514e-05, "loss": 0.0233, "step": 45640 }, { "grad_norm": 0.3913203477859497, "learning_rate": 6.123126216364665e-05, "loss": 0.031, "step": 45650 }, { "grad_norm": 0.7436754107475281, "learning_rate": 6.121514939210797e-05, "loss": 0.0276, "step": 45660 }, { "grad_norm": 0.38729098439216614, "learning_rate": 6.119903539409741e-05, "loss": 0.0259, "step": 45670 }, { "grad_norm": 1.079154372215271, "learning_rate": 6.118292017137716e-05, "loss": 0.0318, "step": 45680 }, { "grad_norm": 0.292633980512619, "learning_rate": 6.116680372570959e-05, "loss": 0.0279, "step": 45690 }, { "grad_norm": 0.32251524925231934, "learning_rate": 6.115068605885713e-05, "loss": 0.0267, "step": 45700 }, { "grad_norm": 0.32286694645881653, "learning_rate": 6.113456717258243e-05, "loss": 0.0356, "step": 45710 }, { "grad_norm": 0.32085907459259033, "learning_rate": 6.11184470686482e-05, "loss": 0.0272, "step": 45720 }, { "grad_norm": 0.37279021739959717, "learning_rate": 6.110232574881731e-05, "loss": 0.0229, "step": 45730 }, { "grad_norm": 0.28751182556152344, "learning_rate": 6.108620321485277e-05, "loss": 0.0236, "step": 45740 }, { "grad_norm": 0.263584166765213, "learning_rate": 6.107007946851773e-05, "loss": 0.025, "step": 45750 }, { "grad_norm": 0.3828915059566498, "learning_rate": 6.105395451157542e-05, "loss": 0.027, "step": 45760 }, { "grad_norm": 0.3449714183807373, "learning_rate": 6.103782834578928e-05, "loss": 0.0232, "step": 45770 }, { "grad_norm": 1.0396921634674072, "learning_rate": 6.102170097292281e-05, "loss": 0.0293, "step": 45780 }, { "grad_norm": 0.333847314119339, "learning_rate": 6.1005572394739716e-05, "loss": 0.0297, "step": 45790 }, { "grad_norm": 0.2797502279281616, "learning_rate": 6.0989442613003765e-05, "loss": 0.0327, "step": 45800 }, { "grad_norm": 0.36884063482284546, "learning_rate": 6.097331162947888e-05, "loss": 0.0325, "step": 45810 }, { "grad_norm": 0.36043938994407654, "learning_rate": 6.095717944592914e-05, "loss": 0.0229, "step": 45820 }, { "grad_norm": 0.8986244797706604, "learning_rate": 6.094104606411873e-05, "loss": 0.0401, "step": 45830 }, { "grad_norm": 0.35953062772750854, "learning_rate": 6.0924911485811966e-05, "loss": 0.0448, "step": 45840 }, { "grad_norm": 0.35403379797935486, "learning_rate": 6.090877571277331e-05, "loss": 0.035, "step": 45850 }, { "grad_norm": 0.38019078969955444, "learning_rate": 6.089263874676736e-05, "loss": 0.0291, "step": 45860 }, { "grad_norm": 0.34641194343566895, "learning_rate": 6.0876500589558796e-05, "loss": 0.0346, "step": 45870 }, { "grad_norm": 0.3249092996120453, "learning_rate": 6.086036124291248e-05, "loss": 0.0297, "step": 45880 }, { "grad_norm": 0.44874686002731323, "learning_rate": 6.084422070859339e-05, "loss": 0.0487, "step": 45890 }, { "grad_norm": 0.3605383634567261, "learning_rate": 6.082807898836663e-05, "loss": 0.0271, "step": 45900 }, { "grad_norm": 0.38536158204078674, "learning_rate": 6.081193608399742e-05, "loss": 0.0293, "step": 45910 }, { "grad_norm": 0.4191608130931854, "learning_rate": 6.0795791997251164e-05, "loss": 0.0328, "step": 45920 }, { "grad_norm": 0.35256722569465637, "learning_rate": 6.0779646729893294e-05, "loss": 0.023, "step": 45930 }, { "grad_norm": 0.4911859631538391, "learning_rate": 6.0763500283689476e-05, "loss": 0.0266, "step": 45940 }, { "grad_norm": 0.3264535367488861, "learning_rate": 6.0747352660405455e-05, "loss": 0.0349, "step": 45950 }, { "grad_norm": 0.45692145824432373, "learning_rate": 6.073120386180709e-05, "loss": 0.0294, "step": 45960 }, { "grad_norm": 0.3823382556438446, "learning_rate": 6.0715053889660425e-05, "loss": 0.0285, "step": 45970 }, { "grad_norm": 0.2846028804779053, "learning_rate": 6.069890274573157e-05, "loss": 0.0296, "step": 45980 }, { "grad_norm": 0.3400954008102417, "learning_rate": 6.068275043178679e-05, "loss": 0.038, "step": 45990 }, { "grad_norm": 0.3218591511249542, "learning_rate": 6.066659694959248e-05, "loss": 0.0277, "step": 46000 }, { "grad_norm": 0.41945409774780273, "learning_rate": 6.0650442300915176e-05, "loss": 0.0273, "step": 46010 }, { "grad_norm": 0.496161550283432, "learning_rate": 6.063428648752152e-05, "loss": 0.0245, "step": 46020 }, { "grad_norm": 0.5464702844619751, "learning_rate": 6.0618129511178266e-05, "loss": 0.0304, "step": 46030 }, { "grad_norm": 0.2990334630012512, "learning_rate": 6.060197137365234e-05, "loss": 0.0253, "step": 46040 }, { "grad_norm": 0.37697234749794006, "learning_rate": 6.058581207671077e-05, "loss": 0.0295, "step": 46050 }, { "grad_norm": 0.4541141986846924, "learning_rate": 6.056965162212072e-05, "loss": 0.0299, "step": 46060 }, { "grad_norm": 0.30866527557373047, "learning_rate": 6.055349001164943e-05, "loss": 0.0329, "step": 46070 }, { "grad_norm": 0.2871759831905365, "learning_rate": 6.0537327247064347e-05, "loss": 0.0309, "step": 46080 }, { "grad_norm": 0.566074013710022, "learning_rate": 6.0521163330133014e-05, "loss": 0.0359, "step": 46090 }, { "grad_norm": 0.3546768128871918, "learning_rate": 6.050499826262306e-05, "loss": 0.0264, "step": 46100 }, { "grad_norm": 0.3429046869277954, "learning_rate": 6.0488832046302294e-05, "loss": 0.0324, "step": 46110 }, { "grad_norm": 0.37265744805336, "learning_rate": 6.0472664682938626e-05, "loss": 0.0255, "step": 46120 }, { "grad_norm": 0.2585909962654114, "learning_rate": 6.045649617430009e-05, "loss": 0.0279, "step": 46130 }, { "grad_norm": 0.34130921959877014, "learning_rate": 6.0440326522154866e-05, "loss": 0.0287, "step": 46140 }, { "grad_norm": 0.431293249130249, "learning_rate": 6.0424155728271224e-05, "loss": 0.026, "step": 46150 }, { "grad_norm": 0.3013622760772705, "learning_rate": 6.040798379441758e-05, "loss": 0.0267, "step": 46160 }, { "grad_norm": 0.547649621963501, "learning_rate": 6.0391810722362485e-05, "loss": 0.0265, "step": 46170 }, { "grad_norm": 0.4239622950553894, "learning_rate": 6.037563651387458e-05, "loss": 0.0363, "step": 46180 }, { "grad_norm": 0.4334369897842407, "learning_rate": 6.0359461170722666e-05, "loss": 0.0307, "step": 46190 }, { "grad_norm": 0.27717792987823486, "learning_rate": 6.034328469467566e-05, "loss": 0.0335, "step": 46200 }, { "grad_norm": 0.5831218957901001, "learning_rate": 6.0327107087502596e-05, "loss": 0.0285, "step": 46210 }, { "grad_norm": 0.25677987933158875, "learning_rate": 6.031092835097262e-05, "loss": 0.0245, "step": 46220 }, { "grad_norm": 0.28708377480506897, "learning_rate": 6.0294748486855024e-05, "loss": 0.0283, "step": 46230 }, { "grad_norm": 0.26576900482177734, "learning_rate": 6.0278567496919216e-05, "loss": 0.0271, "step": 46240 }, { "grad_norm": 0.30557143688201904, "learning_rate": 6.026238538293472e-05, "loss": 0.0222, "step": 46250 }, { "grad_norm": 0.3165872097015381, "learning_rate": 6.024620214667118e-05, "loss": 0.03, "step": 46260 }, { "grad_norm": 0.37633323669433594, "learning_rate": 6.0230017789898384e-05, "loss": 0.0253, "step": 46270 }, { "grad_norm": 0.2839124798774719, "learning_rate": 6.021383231438622e-05, "loss": 0.0231, "step": 46280 }, { "grad_norm": 0.34886467456817627, "learning_rate": 6.0197645721904704e-05, "loss": 0.0258, "step": 46290 }, { "grad_norm": 0.3789544105529785, "learning_rate": 6.0181458014224e-05, "loss": 0.0301, "step": 46300 }, { "grad_norm": 0.30566465854644775, "learning_rate": 6.016526919311435e-05, "loss": 0.0266, "step": 46310 }, { "grad_norm": 0.41411030292510986, "learning_rate": 6.014907926034613e-05, "loss": 0.0281, "step": 46320 }, { "grad_norm": 0.35614266991615295, "learning_rate": 6.0132888217689875e-05, "loss": 0.0319, "step": 46330 }, { "grad_norm": 0.5763707756996155, "learning_rate": 6.01166960669162e-05, "loss": 0.04, "step": 46340 }, { "grad_norm": 0.2616027593612671, "learning_rate": 6.0100502809795845e-05, "loss": 0.0304, "step": 46350 }, { "grad_norm": 0.3118706941604614, "learning_rate": 6.00843084480997e-05, "loss": 0.0322, "step": 46360 }, { "grad_norm": 0.28734853863716125, "learning_rate": 6.0068112983598736e-05, "loss": 0.0255, "step": 46370 }, { "grad_norm": 0.2396399974822998, "learning_rate": 6.005191641806407e-05, "loss": 0.0284, "step": 46380 }, { "grad_norm": 0.3654792010784149, "learning_rate": 6.003571875326694e-05, "loss": 0.0251, "step": 46390 }, { "grad_norm": 0.19523468613624573, "learning_rate": 6.00195199909787e-05, "loss": 0.0228, "step": 46400 }, { "grad_norm": 0.33432018756866455, "learning_rate": 6.000332013297082e-05, "loss": 0.0279, "step": 46410 }, { "grad_norm": 0.45742547512054443, "learning_rate": 5.998711918101487e-05, "loss": 0.0312, "step": 46420 }, { "grad_norm": 0.33029425144195557, "learning_rate": 5.997091713688261e-05, "loss": 0.0266, "step": 46430 }, { "grad_norm": 0.3356790244579315, "learning_rate": 5.9954714002345836e-05, "loss": 0.0235, "step": 46440 }, { "grad_norm": 1.300704002380371, "learning_rate": 5.993850977917649e-05, "loss": 0.0295, "step": 46450 }, { "grad_norm": 0.3374480605125427, "learning_rate": 5.992230446914667e-05, "loss": 0.0284, "step": 46460 }, { "grad_norm": 0.3541247248649597, "learning_rate": 5.990609807402855e-05, "loss": 0.0308, "step": 46470 }, { "grad_norm": 0.26668572425842285, "learning_rate": 5.988989059559443e-05, "loss": 0.0341, "step": 46480 }, { "grad_norm": 0.2826586961746216, "learning_rate": 5.987368203561675e-05, "loss": 0.0263, "step": 46490 }, { "grad_norm": 0.5020946264266968, "learning_rate": 5.9857472395868055e-05, "loss": 0.0242, "step": 46500 }, { "grad_norm": 0.3043968677520752, "learning_rate": 5.9841261678120983e-05, "loss": 0.031, "step": 46510 }, { "grad_norm": 0.2619290351867676, "learning_rate": 5.982504988414834e-05, "loss": 0.0256, "step": 46520 }, { "grad_norm": 0.4940556585788727, "learning_rate": 5.9808837015723015e-05, "loss": 0.0409, "step": 46530 }, { "grad_norm": 0.4359317123889923, "learning_rate": 5.9792623074618016e-05, "loss": 0.0292, "step": 46540 }, { "grad_norm": 0.4487055838108063, "learning_rate": 5.977640806260648e-05, "loss": 0.0217, "step": 46550 }, { "grad_norm": 0.251916766166687, "learning_rate": 5.976019198146166e-05, "loss": 0.0196, "step": 46560 }, { "grad_norm": 0.30100569128990173, "learning_rate": 5.974397483295692e-05, "loss": 0.0257, "step": 46570 }, { "grad_norm": 0.8021897077560425, "learning_rate": 5.972775661886575e-05, "loss": 0.0386, "step": 46580 }, { "grad_norm": 0.3597385585308075, "learning_rate": 5.971153734096173e-05, "loss": 0.0251, "step": 46590 }, { "grad_norm": 0.2782129943370819, "learning_rate": 5.969531700101859e-05, "loss": 0.0229, "step": 46600 }, { "grad_norm": 0.26454591751098633, "learning_rate": 5.9679095600810155e-05, "loss": 0.0325, "step": 46610 }, { "grad_norm": 0.36536893248558044, "learning_rate": 5.9662873142110384e-05, "loss": 0.0263, "step": 46620 }, { "grad_norm": 0.2139371782541275, "learning_rate": 5.964664962669333e-05, "loss": 0.0259, "step": 46630 }, { "grad_norm": 0.6535186767578125, "learning_rate": 5.9630425056333186e-05, "loss": 0.0262, "step": 46640 }, { "grad_norm": 0.2895033359527588, "learning_rate": 5.961419943280422e-05, "loss": 0.0276, "step": 46650 }, { "grad_norm": 0.335397869348526, "learning_rate": 5.959797275788087e-05, "loss": 0.0239, "step": 46660 }, { "grad_norm": 0.30519792437553406, "learning_rate": 5.958174503333765e-05, "loss": 0.0331, "step": 46670 }, { "grad_norm": 0.28930386900901794, "learning_rate": 5.9565516260949195e-05, "loss": 0.0315, "step": 46680 }, { "grad_norm": 0.27848610281944275, "learning_rate": 5.9549286442490273e-05, "loss": 0.0261, "step": 46690 }, { "grad_norm": 0.5325188040733337, "learning_rate": 5.953305557973572e-05, "loss": 0.0321, "step": 46700 }, { "grad_norm": 0.5663842558860779, "learning_rate": 5.9516823674460564e-05, "loss": 0.0269, "step": 46710 }, { "grad_norm": 0.3721301257610321, "learning_rate": 5.9500590728439875e-05, "loss": 0.028, "step": 46720 }, { "grad_norm": 0.21805785596370697, "learning_rate": 5.948435674344886e-05, "loss": 0.026, "step": 46730 }, { "grad_norm": 0.3451869785785675, "learning_rate": 5.946812172126285e-05, "loss": 0.0236, "step": 46740 }, { "grad_norm": 0.6698355078697205, "learning_rate": 5.945188566365729e-05, "loss": 0.0299, "step": 46750 }, { "grad_norm": 0.2855869233608246, "learning_rate": 5.943564857240773e-05, "loss": 0.0321, "step": 46760 }, { "grad_norm": 0.2593078315258026, "learning_rate": 5.941941044928981e-05, "loss": 0.0285, "step": 46770 }, { "grad_norm": 0.39165735244750977, "learning_rate": 5.940317129607935e-05, "loss": 0.0342, "step": 46780 }, { "grad_norm": 0.24332912266254425, "learning_rate": 5.9386931114552204e-05, "loss": 0.0228, "step": 46790 }, { "grad_norm": 0.26724475622177124, "learning_rate": 5.937068990648438e-05, "loss": 0.0284, "step": 46800 }, { "grad_norm": 0.31470003724098206, "learning_rate": 5.935444767365199e-05, "loss": 0.0343, "step": 46810 }, { "grad_norm": 0.308749794960022, "learning_rate": 5.933820441783129e-05, "loss": 0.0301, "step": 46820 }, { "grad_norm": 0.3243982493877411, "learning_rate": 5.932196014079857e-05, "loss": 0.0264, "step": 46830 }, { "grad_norm": 0.2939325273036957, "learning_rate": 5.930571484433032e-05, "loss": 0.0253, "step": 46840 }, { "grad_norm": 0.2681284248828888, "learning_rate": 5.928946853020309e-05, "loss": 0.0211, "step": 46850 }, { "grad_norm": 0.28172752261161804, "learning_rate": 5.927322120019352e-05, "loss": 0.0234, "step": 46860 }, { "grad_norm": 0.22445736825466156, "learning_rate": 5.925697285607844e-05, "loss": 0.0307, "step": 46870 }, { "grad_norm": 0.34401267766952515, "learning_rate": 5.924072349963472e-05, "loss": 0.0283, "step": 46880 }, { "grad_norm": 0.3028211295604706, "learning_rate": 5.922447313263938e-05, "loss": 0.0311, "step": 46890 }, { "grad_norm": 0.26800182461738586, "learning_rate": 5.920822175686952e-05, "loss": 0.0248, "step": 46900 }, { "grad_norm": 0.33848413825035095, "learning_rate": 5.919196937410239e-05, "loss": 0.0341, "step": 46910 }, { "grad_norm": 0.5543503761291504, "learning_rate": 5.917571598611529e-05, "loss": 0.0317, "step": 46920 }, { "grad_norm": 0.30443698167800903, "learning_rate": 5.9159461594685686e-05, "loss": 0.0284, "step": 46930 }, { "grad_norm": 0.712874174118042, "learning_rate": 5.914320620159114e-05, "loss": 0.0362, "step": 46940 }, { "grad_norm": 0.38794946670532227, "learning_rate": 5.912694980860931e-05, "loss": 0.0259, "step": 46950 }, { "grad_norm": 0.3535034954547882, "learning_rate": 5.9110692417517964e-05, "loss": 0.0307, "step": 46960 }, { "grad_norm": 0.22545284032821655, "learning_rate": 5.9094434030095e-05, "loss": 0.027, "step": 46970 }, { "grad_norm": 0.26323941349983215, "learning_rate": 5.907817464811841e-05, "loss": 0.0247, "step": 46980 }, { "grad_norm": 0.3231678307056427, "learning_rate": 5.906191427336627e-05, "loss": 0.0273, "step": 46990 }, { "grad_norm": 0.2891395390033722, "learning_rate": 5.9045652907616835e-05, "loss": 0.0268, "step": 47000 }, { "grad_norm": 0.3166643977165222, "learning_rate": 5.902939055264838e-05, "loss": 0.0258, "step": 47010 }, { "grad_norm": 0.2778365910053253, "learning_rate": 5.901312721023935e-05, "loss": 0.0299, "step": 47020 }, { "grad_norm": 0.2669205069541931, "learning_rate": 5.8996862882168294e-05, "loss": 0.0287, "step": 47030 }, { "grad_norm": 0.4260578751564026, "learning_rate": 5.8980597570213826e-05, "loss": 0.041, "step": 47040 }, { "grad_norm": 0.31702959537506104, "learning_rate": 5.896433127615471e-05, "loss": 0.0264, "step": 47050 }, { "grad_norm": 0.2966172695159912, "learning_rate": 5.894806400176981e-05, "loss": 0.0279, "step": 47060 }, { "grad_norm": 0.25555771589279175, "learning_rate": 5.893179574883808e-05, "loss": 0.0286, "step": 47070 }, { "grad_norm": 0.5095371007919312, "learning_rate": 5.8915526519138585e-05, "loss": 0.0268, "step": 47080 }, { "grad_norm": 0.35308706760406494, "learning_rate": 5.889925631445053e-05, "loss": 0.0259, "step": 47090 }, { "grad_norm": 0.5740824937820435, "learning_rate": 5.888298513655318e-05, "loss": 0.0319, "step": 47100 }, { "grad_norm": 0.3473002016544342, "learning_rate": 5.886671298722592e-05, "loss": 0.0194, "step": 47110 }, { "grad_norm": 0.4567815661430359, "learning_rate": 5.885043986824828e-05, "loss": 0.0231, "step": 47120 }, { "grad_norm": 0.6311994194984436, "learning_rate": 5.883416578139982e-05, "loss": 0.0231, "step": 47130 }, { "grad_norm": 0.3695327937602997, "learning_rate": 5.881789072846029e-05, "loss": 0.0339, "step": 47140 }, { "grad_norm": 0.38129928708076477, "learning_rate": 5.8801614711209474e-05, "loss": 0.0204, "step": 47150 }, { "grad_norm": 0.3050748109817505, "learning_rate": 5.878533773142732e-05, "loss": 0.0359, "step": 47160 }, { "grad_norm": 0.39150819182395935, "learning_rate": 5.8769059790893834e-05, "loss": 0.0316, "step": 47170 }, { "grad_norm": 0.42980504035949707, "learning_rate": 5.8752780891389145e-05, "loss": 0.0279, "step": 47180 }, { "grad_norm": 0.2831290364265442, "learning_rate": 5.873650103469351e-05, "loss": 0.028, "step": 47190 }, { "grad_norm": 0.3687548339366913, "learning_rate": 5.872022022258725e-05, "loss": 0.0343, "step": 47200 }, { "grad_norm": 0.30757495760917664, "learning_rate": 5.8703938456850814e-05, "loss": 0.0227, "step": 47210 }, { "grad_norm": 0.3370412290096283, "learning_rate": 5.868765573926477e-05, "loss": 0.0247, "step": 47220 }, { "grad_norm": 0.38002511858940125, "learning_rate": 5.867137207160974e-05, "loss": 0.0279, "step": 47230 }, { "grad_norm": 0.47707799077033997, "learning_rate": 5.86550874556665e-05, "loss": 0.0242, "step": 47240 }, { "grad_norm": 0.34890204668045044, "learning_rate": 5.863880189321592e-05, "loss": 0.03, "step": 47250 }, { "grad_norm": 0.4548044204711914, "learning_rate": 5.862251538603894e-05, "loss": 0.0329, "step": 47260 }, { "grad_norm": 0.34652841091156006, "learning_rate": 5.8606227935916656e-05, "loss": 0.029, "step": 47270 }, { "grad_norm": 0.257477343082428, "learning_rate": 5.858993954463021e-05, "loss": 0.0211, "step": 47280 }, { "grad_norm": 0.30915379524230957, "learning_rate": 5.8573650213960896e-05, "loss": 0.0357, "step": 47290 }, { "grad_norm": 0.19911620020866394, "learning_rate": 5.8557359945690084e-05, "loss": 0.0217, "step": 47300 }, { "grad_norm": 0.3655030131340027, "learning_rate": 5.8541068741599246e-05, "loss": 0.0282, "step": 47310 }, { "grad_norm": 0.3185085952281952, "learning_rate": 5.8524776603469985e-05, "loss": 0.0226, "step": 47320 }, { "grad_norm": 0.2574828863143921, "learning_rate": 5.850848353308397e-05, "loss": 0.0208, "step": 47330 }, { "grad_norm": 0.3436118960380554, "learning_rate": 5.849218953222297e-05, "loss": 0.0292, "step": 47340 }, { "grad_norm": 0.24735990166664124, "learning_rate": 5.847589460266891e-05, "loss": 0.0205, "step": 47350 }, { "grad_norm": 0.32147276401519775, "learning_rate": 5.845959874620375e-05, "loss": 0.0269, "step": 47360 }, { "grad_norm": 0.31116193532943726, "learning_rate": 5.84433019646096e-05, "loss": 0.0327, "step": 47370 }, { "grad_norm": 0.33925071358680725, "learning_rate": 5.842700425966863e-05, "loss": 0.0304, "step": 47380 }, { "grad_norm": 0.44374361634254456, "learning_rate": 5.841070563316315e-05, "loss": 0.0324, "step": 47390 }, { "grad_norm": 0.9211646318435669, "learning_rate": 5.839440608687554e-05, "loss": 0.0314, "step": 47400 }, { "grad_norm": 0.3494893014431, "learning_rate": 5.837810562258831e-05, "loss": 0.0278, "step": 47410 }, { "grad_norm": 0.3532876670360565, "learning_rate": 5.836180424208405e-05, "loss": 0.0297, "step": 47420 }, { "grad_norm": 0.3060550093650818, "learning_rate": 5.834550194714543e-05, "loss": 0.0238, "step": 47430 }, { "grad_norm": 0.49789494276046753, "learning_rate": 5.832919873955528e-05, "loss": 0.0358, "step": 47440 }, { "grad_norm": 0.2880682349205017, "learning_rate": 5.831289462109648e-05, "loss": 0.0234, "step": 47450 }, { "grad_norm": 0.2271200716495514, "learning_rate": 5.829658959355202e-05, "loss": 0.0368, "step": 47460 }, { "grad_norm": 0.3438870906829834, "learning_rate": 5.8280283658704995e-05, "loss": 0.0265, "step": 47470 }, { "grad_norm": 0.2402055412530899, "learning_rate": 5.826397681833859e-05, "loss": 0.0258, "step": 47480 }, { "grad_norm": 0.3323028087615967, "learning_rate": 5.824766907423612e-05, "loss": 0.0244, "step": 47490 }, { "grad_norm": 0.23224377632141113, "learning_rate": 5.8231360428180945e-05, "loss": 0.0274, "step": 47500 }, { "grad_norm": 0.34656837582588196, "learning_rate": 5.821505088195658e-05, "loss": 0.0246, "step": 47510 }, { "grad_norm": 0.265708863735199, "learning_rate": 5.819874043734661e-05, "loss": 0.0277, "step": 47520 }, { "grad_norm": 0.5879507660865784, "learning_rate": 5.8182429096134695e-05, "loss": 0.0256, "step": 47530 }, { "grad_norm": 0.21261298656463623, "learning_rate": 5.816611686010465e-05, "loss": 0.0189, "step": 47540 }, { "grad_norm": 0.4921301007270813, "learning_rate": 5.814980373104033e-05, "loss": 0.0346, "step": 47550 }, { "grad_norm": 0.5023316144943237, "learning_rate": 5.813348971072572e-05, "loss": 0.0284, "step": 47560 }, { "grad_norm": 0.33681023120880127, "learning_rate": 5.811717480094492e-05, "loss": 0.03, "step": 47570 }, { "grad_norm": 0.43376675248146057, "learning_rate": 5.810085900348209e-05, "loss": 0.0292, "step": 47580 }, { "grad_norm": 0.49686872959136963, "learning_rate": 5.8084542320121483e-05, "loss": 0.0309, "step": 47590 }, { "grad_norm": 0.9412790536880493, "learning_rate": 5.8068224752647497e-05, "loss": 0.0369, "step": 47600 }, { "grad_norm": 0.3895578384399414, "learning_rate": 5.805190630284457e-05, "loss": 0.0249, "step": 47610 }, { "grad_norm": 0.4126943349838257, "learning_rate": 5.803558697249729e-05, "loss": 0.0266, "step": 47620 }, { "grad_norm": 0.3849808871746063, "learning_rate": 5.8019266763390276e-05, "loss": 0.0367, "step": 47630 }, { "grad_norm": 0.2825940251350403, "learning_rate": 5.800294567730833e-05, "loss": 0.0228, "step": 47640 }, { "grad_norm": 0.289060115814209, "learning_rate": 5.798662371603628e-05, "loss": 0.0299, "step": 47650 }, { "grad_norm": 0.587738037109375, "learning_rate": 5.797030088135904e-05, "loss": 0.0255, "step": 47660 }, { "grad_norm": 0.305765837430954, "learning_rate": 5.7953977175061705e-05, "loss": 0.0225, "step": 47670 }, { "grad_norm": 0.6376198530197144, "learning_rate": 5.793765259892938e-05, "loss": 0.029, "step": 47680 }, { "grad_norm": 0.36111828684806824, "learning_rate": 5.792132715474729e-05, "loss": 0.0224, "step": 47690 }, { "grad_norm": 0.38778313994407654, "learning_rate": 5.790500084430078e-05, "loss": 0.0218, "step": 47700 }, { "grad_norm": 0.27975428104400635, "learning_rate": 5.7888673669375274e-05, "loss": 0.0233, "step": 47710 }, { "grad_norm": 0.1862889528274536, "learning_rate": 5.787234563175625e-05, "loss": 0.0225, "step": 47720 }, { "grad_norm": 0.48965364694595337, "learning_rate": 5.7856016733229355e-05, "loss": 0.0284, "step": 47730 }, { "grad_norm": 0.6170341968536377, "learning_rate": 5.7839686975580297e-05, "loss": 0.0366, "step": 47740 }, { "grad_norm": 0.2540678083896637, "learning_rate": 5.782335636059484e-05, "loss": 0.0228, "step": 47750 }, { "grad_norm": 0.38962626457214355, "learning_rate": 5.780702489005889e-05, "loss": 0.0287, "step": 47760 }, { "grad_norm": 0.24540971219539642, "learning_rate": 5.779069256575845e-05, "loss": 0.0298, "step": 47770 }, { "grad_norm": 0.28799423575401306, "learning_rate": 5.7774359389479574e-05, "loss": 0.0414, "step": 47780 }, { "grad_norm": 0.3659072518348694, "learning_rate": 5.775802536300845e-05, "loss": 0.0263, "step": 47790 }, { "grad_norm": 0.19291765987873077, "learning_rate": 5.774169048813134e-05, "loss": 0.0297, "step": 47800 }, { "grad_norm": 0.2416277676820755, "learning_rate": 5.77253547666346e-05, "loss": 0.024, "step": 47810 }, { "grad_norm": 0.3974337875843048, "learning_rate": 5.770901820030465e-05, "loss": 0.0307, "step": 47820 }, { "grad_norm": 0.3118470311164856, "learning_rate": 5.769268079092809e-05, "loss": 0.0286, "step": 47830 }, { "grad_norm": 0.2996014952659607, "learning_rate": 5.767634254029151e-05, "loss": 0.0255, "step": 47840 }, { "grad_norm": 0.6453753113746643, "learning_rate": 5.7660003450181655e-05, "loss": 0.0498, "step": 47850 }, { "grad_norm": 0.29538479447364807, "learning_rate": 5.764366352238534e-05, "loss": 0.0198, "step": 47860 }, { "grad_norm": 0.3112446367740631, "learning_rate": 5.7627322758689474e-05, "loss": 0.0235, "step": 47870 }, { "grad_norm": 0.5762777924537659, "learning_rate": 5.761098116088105e-05, "loss": 0.0284, "step": 47880 }, { "grad_norm": 0.27512723207473755, "learning_rate": 5.759463873074717e-05, "loss": 0.0248, "step": 47890 }, { "grad_norm": 0.3065080940723419, "learning_rate": 5.757829547007504e-05, "loss": 0.0304, "step": 47900 }, { "grad_norm": 0.2890307903289795, "learning_rate": 5.756195138065189e-05, "loss": 0.0216, "step": 47910 }, { "grad_norm": 1.6450896263122559, "learning_rate": 5.754560646426511e-05, "loss": 0.0327, "step": 47920 }, { "grad_norm": 0.2960728406906128, "learning_rate": 5.752926072270216e-05, "loss": 0.0253, "step": 47930 }, { "grad_norm": 0.4884706735610962, "learning_rate": 5.7512914157750563e-05, "loss": 0.0246, "step": 47940 }, { "grad_norm": 0.5518668293952942, "learning_rate": 5.749656677119798e-05, "loss": 0.0251, "step": 47950 }, { "grad_norm": 0.45632240176200867, "learning_rate": 5.7480218564832125e-05, "loss": 0.0291, "step": 47960 }, { "grad_norm": 0.24945029616355896, "learning_rate": 5.746386954044082e-05, "loss": 0.0275, "step": 47970 }, { "grad_norm": 0.31300100684165955, "learning_rate": 5.744751969981195e-05, "loss": 0.0229, "step": 47980 }, { "grad_norm": 0.2604522407054901, "learning_rate": 5.7431169044733526e-05, "loss": 0.0324, "step": 47990 }, { "grad_norm": 0.24839723110198975, "learning_rate": 5.741481757699364e-05, "loss": 0.0268, "step": 48000 }, { "grad_norm": 0.3129292130470276, "learning_rate": 5.7398465298380434e-05, "loss": 0.0256, "step": 48010 }, { "grad_norm": 0.2560536563396454, "learning_rate": 5.7382112210682193e-05, "loss": 0.0336, "step": 48020 }, { "grad_norm": 0.3335108757019043, "learning_rate": 5.7365758315687266e-05, "loss": 0.0251, "step": 48030 }, { "grad_norm": 0.3457969129085541, "learning_rate": 5.734940361518407e-05, "loss": 0.0241, "step": 48040 }, { "grad_norm": 0.36020544171333313, "learning_rate": 5.733304811096116e-05, "loss": 0.0241, "step": 48050 }, { "grad_norm": 0.3003753423690796, "learning_rate": 5.731669180480713e-05, "loss": 0.0222, "step": 48060 }, { "grad_norm": 0.25086668133735657, "learning_rate": 5.730033469851067e-05, "loss": 0.0379, "step": 48070 }, { "grad_norm": 0.3553728759288788, "learning_rate": 5.7283976793860607e-05, "loss": 0.0296, "step": 48080 }, { "grad_norm": 0.273703932762146, "learning_rate": 5.726761809264577e-05, "loss": 0.0233, "step": 48090 }, { "grad_norm": 0.29746749997138977, "learning_rate": 5.7251258596655155e-05, "loss": 0.0265, "step": 48100 }, { "grad_norm": 0.4434432089328766, "learning_rate": 5.72348983076778e-05, "loss": 0.0281, "step": 48110 }, { "grad_norm": 0.4644680321216583, "learning_rate": 5.7218537227502854e-05, "loss": 0.0333, "step": 48120 }, { "grad_norm": 0.21978293359279633, "learning_rate": 5.720217535791951e-05, "loss": 0.0232, "step": 48130 }, { "grad_norm": 0.3157016932964325, "learning_rate": 5.718581270071711e-05, "loss": 0.0302, "step": 48140 }, { "grad_norm": 0.3761564791202545, "learning_rate": 5.716944925768505e-05, "loss": 0.0295, "step": 48150 }, { "grad_norm": 0.43223837018013, "learning_rate": 5.7153085030612786e-05, "loss": 0.0294, "step": 48160 }, { "grad_norm": 0.23675376176834106, "learning_rate": 5.713672002128989e-05, "loss": 0.0313, "step": 48170 }, { "grad_norm": 0.2617635726928711, "learning_rate": 5.712035423150602e-05, "loss": 0.0272, "step": 48180 }, { "grad_norm": 0.286639541387558, "learning_rate": 5.710398766305094e-05, "loss": 0.027, "step": 48190 }, { "grad_norm": 0.38898879289627075, "learning_rate": 5.708762031771442e-05, "loss": 0.0322, "step": 48200 }, { "grad_norm": 0.34512844681739807, "learning_rate": 5.707125219728642e-05, "loss": 0.0335, "step": 48210 }, { "grad_norm": 0.3141580820083618, "learning_rate": 5.7054883303556905e-05, "loss": 0.0249, "step": 48220 }, { "grad_norm": 0.3012251555919647, "learning_rate": 5.703851363831595e-05, "loss": 0.0292, "step": 48230 }, { "grad_norm": 0.4261223077774048, "learning_rate": 5.702214320335374e-05, "loss": 0.0432, "step": 48240 }, { "grad_norm": 0.26835930347442627, "learning_rate": 5.7005772000460514e-05, "loss": 0.0281, "step": 48250 }, { "grad_norm": 0.345981627702713, "learning_rate": 5.698940003142656e-05, "loss": 0.0227, "step": 48260 }, { "grad_norm": 0.3544243276119232, "learning_rate": 5.697302729804236e-05, "loss": 0.0241, "step": 48270 }, { "grad_norm": 0.5060890316963196, "learning_rate": 5.695665380209837e-05, "loss": 0.0226, "step": 48280 }, { "grad_norm": 0.27809473872184753, "learning_rate": 5.6940279545385165e-05, "loss": 0.0238, "step": 48290 }, { "grad_norm": 0.28619199991226196, "learning_rate": 5.692390452969344e-05, "loss": 0.0224, "step": 48300 }, { "grad_norm": 0.3421938121318817, "learning_rate": 5.690752875681392e-05, "loss": 0.0277, "step": 48310 }, { "grad_norm": 0.3912312090396881, "learning_rate": 5.6891152228537435e-05, "loss": 0.0304, "step": 48320 }, { "grad_norm": 0.34794825315475464, "learning_rate": 5.687477494665492e-05, "loss": 0.024, "step": 48330 }, { "grad_norm": 0.4016766846179962, "learning_rate": 5.685839691295733e-05, "loss": 0.0277, "step": 48340 }, { "grad_norm": 0.39136138558387756, "learning_rate": 5.6842018129235786e-05, "loss": 0.0282, "step": 48350 }, { "grad_norm": 0.26551392674446106, "learning_rate": 5.6825638597281404e-05, "loss": 0.0215, "step": 48360 }, { "grad_norm": 0.39975881576538086, "learning_rate": 5.680925831888546e-05, "loss": 0.0305, "step": 48370 }, { "grad_norm": 0.21896350383758545, "learning_rate": 5.6792877295839274e-05, "loss": 0.0292, "step": 48380 }, { "grad_norm": 0.4493141770362854, "learning_rate": 5.6776495529934224e-05, "loss": 0.031, "step": 48390 }, { "grad_norm": 0.3068813681602478, "learning_rate": 5.6760113022961824e-05, "loss": 0.0251, "step": 48400 }, { "grad_norm": 0.2617902457714081, "learning_rate": 5.6743729776713617e-05, "loss": 0.0194, "step": 48410 }, { "grad_norm": 0.24724715948104858, "learning_rate": 5.672734579298126e-05, "loss": 0.0264, "step": 48420 }, { "grad_norm": 0.33820870518684387, "learning_rate": 5.671096107355649e-05, "loss": 0.0265, "step": 48430 }, { "grad_norm": 0.33768901228904724, "learning_rate": 5.66945756202311e-05, "loss": 0.0245, "step": 48440 }, { "grad_norm": 0.3862093389034271, "learning_rate": 5.667818943479699e-05, "loss": 0.0287, "step": 48450 }, { "grad_norm": 0.31030088663101196, "learning_rate": 5.666180251904612e-05, "loss": 0.0374, "step": 48460 }, { "grad_norm": 0.44126859307289124, "learning_rate": 5.6645414874770555e-05, "loss": 0.0308, "step": 48470 }, { "grad_norm": 0.289997398853302, "learning_rate": 5.66290265037624e-05, "loss": 0.028, "step": 48480 }, { "grad_norm": 0.4546836316585541, "learning_rate": 5.661263740781386e-05, "loss": 0.0281, "step": 48490 }, { "grad_norm": 0.4004802405834198, "learning_rate": 5.6596247588717254e-05, "loss": 0.0245, "step": 48500 }, { "grad_norm": 0.22487089037895203, "learning_rate": 5.6579857048264926e-05, "loss": 0.0233, "step": 48510 }, { "grad_norm": 0.42757508158683777, "learning_rate": 5.6563465788249314e-05, "loss": 0.0228, "step": 48520 }, { "grad_norm": 0.4116780757904053, "learning_rate": 5.6547073810462956e-05, "loss": 0.0243, "step": 48530 }, { "grad_norm": 0.528059720993042, "learning_rate": 5.653068111669846e-05, "loss": 0.0277, "step": 48540 }, { "grad_norm": 0.4568864107131958, "learning_rate": 5.651428770874848e-05, "loss": 0.0316, "step": 48550 }, { "grad_norm": 0.31794390082359314, "learning_rate": 5.64978935884058e-05, "loss": 0.0261, "step": 48560 }, { "grad_norm": 0.4099048674106598, "learning_rate": 5.6481498757463244e-05, "loss": 0.0355, "step": 48570 }, { "grad_norm": 0.4476660490036011, "learning_rate": 5.646510321771373e-05, "loss": 0.0319, "step": 48580 }, { "grad_norm": 0.8257823586463928, "learning_rate": 5.644870697095024e-05, "loss": 0.0288, "step": 48590 }, { "grad_norm": 0.25464868545532227, "learning_rate": 5.643231001896586e-05, "loss": 0.0232, "step": 48600 }, { "grad_norm": 0.5099097490310669, "learning_rate": 5.6415912363553726e-05, "loss": 0.0584, "step": 48610 }, { "grad_norm": 0.481135755777359, "learning_rate": 5.639951400650706e-05, "loss": 0.0234, "step": 48620 }, { "grad_norm": 0.3733867108821869, "learning_rate": 5.6383114949619165e-05, "loss": 0.0268, "step": 48630 }, { "grad_norm": 0.2791765630245209, "learning_rate": 5.636671519468342e-05, "loss": 0.0261, "step": 48640 }, { "grad_norm": 0.42657455801963806, "learning_rate": 5.635031474349327e-05, "loss": 0.0266, "step": 48650 }, { "grad_norm": 0.28900185227394104, "learning_rate": 5.6333913597842246e-05, "loss": 0.0196, "step": 48660 }, { "grad_norm": 0.26351872086524963, "learning_rate": 5.6317511759523955e-05, "loss": 0.026, "step": 48670 }, { "grad_norm": 0.3797237277030945, "learning_rate": 5.630110923033207e-05, "loss": 0.024, "step": 48680 }, { "grad_norm": 0.33074289560317993, "learning_rate": 5.628470601206036e-05, "loss": 0.0306, "step": 48690 }, { "grad_norm": 0.22709253430366516, "learning_rate": 5.626830210650266e-05, "loss": 0.0278, "step": 48700 }, { "grad_norm": 0.2646973133087158, "learning_rate": 5.625189751545285e-05, "loss": 0.0301, "step": 48710 }, { "grad_norm": 0.5221923589706421, "learning_rate": 5.6235492240704936e-05, "loss": 0.029, "step": 48720 }, { "grad_norm": 0.40955421328544617, "learning_rate": 5.621908628405296e-05, "loss": 0.03, "step": 48730 }, { "grad_norm": 0.3194933533668518, "learning_rate": 5.620267964729106e-05, "loss": 0.0275, "step": 48740 }, { "grad_norm": 0.31305748224258423, "learning_rate": 5.618627233221344e-05, "loss": 0.0218, "step": 48750 }, { "grad_norm": 0.4531397819519043, "learning_rate": 5.616986434061438e-05, "loss": 0.03, "step": 48760 }, { "grad_norm": 0.22593702375888824, "learning_rate": 5.615345567428822e-05, "loss": 0.0255, "step": 48770 }, { "grad_norm": 0.27344000339508057, "learning_rate": 5.613704633502941e-05, "loss": 0.0238, "step": 48780 }, { "grad_norm": 0.2984413802623749, "learning_rate": 5.612063632463245e-05, "loss": 0.0182, "step": 48790 }, { "grad_norm": 0.3457522392272949, "learning_rate": 5.610422564489188e-05, "loss": 0.0216, "step": 48800 }, { "grad_norm": 0.3469234108924866, "learning_rate": 5.608781429760239e-05, "loss": 0.0276, "step": 48810 }, { "grad_norm": 0.3779674768447876, "learning_rate": 5.607140228455866e-05, "loss": 0.0285, "step": 48820 }, { "grad_norm": 0.33653944730758667, "learning_rate": 5.605498960755553e-05, "loss": 0.033, "step": 48830 }, { "grad_norm": 0.3570796847343445, "learning_rate": 5.603857626838782e-05, "loss": 0.0337, "step": 48840 }, { "grad_norm": 0.29118871688842773, "learning_rate": 5.60221622688505e-05, "loss": 0.0275, "step": 48850 }, { "grad_norm": 0.25983837246894836, "learning_rate": 5.6005747610738565e-05, "loss": 0.0271, "step": 48860 }, { "grad_norm": 0.31013572216033936, "learning_rate": 5.59893322958471e-05, "loss": 0.0255, "step": 48870 }, { "grad_norm": 0.34748294949531555, "learning_rate": 5.5972916325971256e-05, "loss": 0.0328, "step": 48880 }, { "grad_norm": 0.2917966842651367, "learning_rate": 5.595649970290628e-05, "loss": 0.0259, "step": 48890 }, { "grad_norm": 0.30707746744155884, "learning_rate": 5.5940082428447426e-05, "loss": 0.028, "step": 48900 }, { "grad_norm": 0.2890895903110504, "learning_rate": 5.592366450439012e-05, "loss": 0.0358, "step": 48910 }, { "grad_norm": 0.29803267121315, "learning_rate": 5.590724593252975e-05, "loss": 0.0307, "step": 48920 }, { "grad_norm": 0.25772324204444885, "learning_rate": 5.589082671466184e-05, "loss": 0.0209, "step": 48930 }, { "grad_norm": 0.3748464286327362, "learning_rate": 5.587440685258199e-05, "loss": 0.0224, "step": 48940 }, { "grad_norm": 0.321550190448761, "learning_rate": 5.585798634808583e-05, "loss": 0.0282, "step": 48950 }, { "grad_norm": 0.3595365583896637, "learning_rate": 5.584156520296909e-05, "loss": 0.0254, "step": 48960 }, { "grad_norm": 0.321625292301178, "learning_rate": 5.582514341902757e-05, "loss": 0.0243, "step": 48970 }, { "grad_norm": 0.42428073287010193, "learning_rate": 5.580872099805713e-05, "loss": 0.0276, "step": 48980 }, { "grad_norm": 0.319264680147171, "learning_rate": 5.5792297941853674e-05, "loss": 0.0273, "step": 48990 }, { "grad_norm": 0.360109806060791, "learning_rate": 5.5775874252213247e-05, "loss": 0.0268, "step": 49000 }, { "grad_norm": 0.4836645722389221, "learning_rate": 5.575944993093189e-05, "loss": 0.0323, "step": 49010 }, { "grad_norm": 0.3124876320362091, "learning_rate": 5.574302497980574e-05, "loss": 0.0254, "step": 49020 }, { "grad_norm": 0.35914939641952515, "learning_rate": 5.5726599400631005e-05, "loss": 0.0256, "step": 49030 }, { "grad_norm": 0.30790215730667114, "learning_rate": 5.571017319520399e-05, "loss": 0.0256, "step": 49040 }, { "grad_norm": 0.43245550990104675, "learning_rate": 5.5693746365320984e-05, "loss": 0.031, "step": 49050 }, { "grad_norm": 0.2900567352771759, "learning_rate": 5.567731891277848e-05, "loss": 0.0283, "step": 49060 }, { "grad_norm": 0.27457743883132935, "learning_rate": 5.566089083937288e-05, "loss": 0.0284, "step": 49070 }, { "grad_norm": 0.39724671840667725, "learning_rate": 5.564446214690079e-05, "loss": 0.0297, "step": 49080 }, { "grad_norm": 0.255079448223114, "learning_rate": 5.56280328371588e-05, "loss": 0.0252, "step": 49090 }, { "grad_norm": 0.29058837890625, "learning_rate": 5.5611602911943596e-05, "loss": 0.0275, "step": 49100 }, { "grad_norm": 0.3484536409378052, "learning_rate": 5.559517237305194e-05, "loss": 0.0287, "step": 49110 }, { "grad_norm": 0.3931635916233063, "learning_rate": 5.557874122228064e-05, "loss": 0.0319, "step": 49120 }, { "grad_norm": 0.3608801066875458, "learning_rate": 5.55623094614266e-05, "loss": 0.0253, "step": 49130 }, { "grad_norm": 0.2541634440422058, "learning_rate": 5.5545877092286743e-05, "loss": 0.024, "step": 49140 }, { "grad_norm": 0.2957706153392792, "learning_rate": 5.552944411665811e-05, "loss": 0.0237, "step": 49150 }, { "grad_norm": 0.4053652584552765, "learning_rate": 5.5513010536337795e-05, "loss": 0.0285, "step": 49160 }, { "grad_norm": 0.4064560830593109, "learning_rate": 5.5496576353122934e-05, "loss": 0.0296, "step": 49170 }, { "grad_norm": 0.29690423607826233, "learning_rate": 5.548014156881074e-05, "loss": 0.0249, "step": 49180 }, { "grad_norm": 0.23330023884773254, "learning_rate": 5.5463706185198494e-05, "loss": 0.0225, "step": 49190 }, { "grad_norm": 0.3750781714916229, "learning_rate": 5.544727020408358e-05, "loss": 0.0233, "step": 49200 }, { "grad_norm": 0.4784051775932312, "learning_rate": 5.5430833627263366e-05, "loss": 0.0263, "step": 49210 }, { "grad_norm": 0.4004388451576233, "learning_rate": 5.5414396456535354e-05, "loss": 0.0282, "step": 49220 }, { "grad_norm": 0.32076504826545715, "learning_rate": 5.539795869369709e-05, "loss": 0.0247, "step": 49230 }, { "grad_norm": 0.29138946533203125, "learning_rate": 5.538152034054618e-05, "loss": 0.0247, "step": 49240 }, { "grad_norm": 0.4638301730155945, "learning_rate": 5.536508139888028e-05, "loss": 0.0311, "step": 49250 }, { "grad_norm": 0.3220760226249695, "learning_rate": 5.534864187049716e-05, "loss": 0.0249, "step": 49260 }, { "grad_norm": 0.26494839787483215, "learning_rate": 5.5332201757194615e-05, "loss": 0.0267, "step": 49270 }, { "grad_norm": 0.33205464482307434, "learning_rate": 5.531576106077049e-05, "loss": 0.0257, "step": 49280 }, { "grad_norm": 0.3427906334400177, "learning_rate": 5.529931978302272e-05, "loss": 0.0293, "step": 49290 }, { "grad_norm": 0.26053929328918457, "learning_rate": 5.528287792574932e-05, "loss": 0.026, "step": 49300 }, { "grad_norm": 0.28428781032562256, "learning_rate": 5.5266435490748324e-05, "loss": 0.0235, "step": 49310 }, { "grad_norm": 0.4202880561351776, "learning_rate": 5.524999247981787e-05, "loss": 0.0273, "step": 49320 }, { "grad_norm": 0.2488909810781479, "learning_rate": 5.523354889475613e-05, "loss": 0.0214, "step": 49330 }, { "grad_norm": 0.30259981751441956, "learning_rate": 5.521710473736134e-05, "loss": 0.0298, "step": 49340 }, { "grad_norm": 0.3261134922504425, "learning_rate": 5.5200660009431826e-05, "loss": 0.0275, "step": 49350 }, { "grad_norm": 0.2562551498413086, "learning_rate": 5.518421471276596e-05, "loss": 0.0276, "step": 49360 }, { "grad_norm": 0.31533128023147583, "learning_rate": 5.5167768849162174e-05, "loss": 0.027, "step": 49370 }, { "grad_norm": 0.2859572470188141, "learning_rate": 5.515132242041893e-05, "loss": 0.0322, "step": 49380 }, { "grad_norm": 0.326915979385376, "learning_rate": 5.513487542833483e-05, "loss": 0.0297, "step": 49390 }, { "grad_norm": 0.324358195066452, "learning_rate": 5.5118427874708466e-05, "loss": 0.0242, "step": 49400 }, { "grad_norm": 0.24617139995098114, "learning_rate": 5.510197976133853e-05, "loss": 0.0251, "step": 49410 }, { "grad_norm": 0.4330430328845978, "learning_rate": 5.508553109002376e-05, "loss": 0.0279, "step": 49420 }, { "grad_norm": 0.3809587359428406, "learning_rate": 5.5069081862562957e-05, "loss": 0.0278, "step": 49430 }, { "grad_norm": 0.8295707702636719, "learning_rate": 5.5052632080754965e-05, "loss": 0.0289, "step": 49440 }, { "grad_norm": 0.2784287631511688, "learning_rate": 5.503618174639874e-05, "loss": 0.0255, "step": 49450 }, { "grad_norm": 0.326996773481369, "learning_rate": 5.501973086129325e-05, "loss": 0.0184, "step": 49460 }, { "grad_norm": 0.27212828397750854, "learning_rate": 5.500327942723753e-05, "loss": 0.0227, "step": 49470 }, { "grad_norm": 0.27110403776168823, "learning_rate": 5.498682744603071e-05, "loss": 0.0242, "step": 49480 }, { "grad_norm": 0.5245717763900757, "learning_rate": 5.497037491947194e-05, "loss": 0.0273, "step": 49490 }, { "grad_norm": 0.3588806986808777, "learning_rate": 5.4953921849360424e-05, "loss": 0.0344, "step": 49500 }, { "grad_norm": 0.32072609663009644, "learning_rate": 5.493746823749547e-05, "loss": 0.0366, "step": 49510 }, { "grad_norm": 0.4232206642627716, "learning_rate": 5.49210140856764e-05, "loss": 0.0315, "step": 49520 }, { "grad_norm": 0.27948737144470215, "learning_rate": 5.4904559395702647e-05, "loss": 0.0261, "step": 49530 }, { "grad_norm": 0.261772096157074, "learning_rate": 5.488810416937364e-05, "loss": 0.0226, "step": 49540 }, { "grad_norm": 0.3167647421360016, "learning_rate": 5.4871648408488926e-05, "loss": 0.0282, "step": 49550 }, { "grad_norm": 0.3903771638870239, "learning_rate": 5.485519211484807e-05, "loss": 0.0213, "step": 49560 }, { "grad_norm": 0.30818846821784973, "learning_rate": 5.4838735290250676e-05, "loss": 0.027, "step": 49570 }, { "grad_norm": 0.3538425862789154, "learning_rate": 5.482227793649649e-05, "loss": 0.0295, "step": 49580 }, { "grad_norm": 0.3024486005306244, "learning_rate": 5.480582005538524e-05, "loss": 0.0333, "step": 49590 }, { "grad_norm": 0.44989752769470215, "learning_rate": 5.478936164871671e-05, "loss": 0.0349, "step": 49600 }, { "grad_norm": 0.5263227820396423, "learning_rate": 5.477290271829081e-05, "loss": 0.0379, "step": 49610 }, { "grad_norm": 0.27189308404922485, "learning_rate": 5.475644326590744e-05, "loss": 0.024, "step": 49620 }, { "grad_norm": 0.7482415437698364, "learning_rate": 5.473998329336658e-05, "loss": 0.029, "step": 49630 }, { "grad_norm": 0.45875370502471924, "learning_rate": 5.4723522802468286e-05, "loss": 0.0261, "step": 49640 }, { "grad_norm": 0.2827671468257904, "learning_rate": 5.470706179501264e-05, "loss": 0.0293, "step": 49650 }, { "grad_norm": 0.27836981415748596, "learning_rate": 5.4690600272799776e-05, "loss": 0.0245, "step": 49660 }, { "grad_norm": 0.4496040940284729, "learning_rate": 5.467413823762993e-05, "loss": 0.0321, "step": 49670 }, { "grad_norm": 0.4038858115673065, "learning_rate": 5.465767569130335e-05, "loss": 0.0286, "step": 49680 }, { "grad_norm": 0.27325645089149475, "learning_rate": 5.464121263562036e-05, "loss": 0.0249, "step": 49690 }, { "grad_norm": 0.6159536242485046, "learning_rate": 5.4624749072381343e-05, "loss": 0.0238, "step": 49700 }, { "grad_norm": 0.31344592571258545, "learning_rate": 5.460828500338672e-05, "loss": 0.0198, "step": 49710 }, { "grad_norm": 0.365125834941864, "learning_rate": 5.459182043043698e-05, "loss": 0.0301, "step": 49720 }, { "grad_norm": 0.27996960282325745, "learning_rate": 5.457535535533265e-05, "loss": 0.0262, "step": 49730 }, { "grad_norm": 0.3497852683067322, "learning_rate": 5.4558889779874334e-05, "loss": 0.0214, "step": 49740 }, { "grad_norm": 0.26672571897506714, "learning_rate": 5.454242370586269e-05, "loss": 0.0274, "step": 49750 }, { "grad_norm": 0.2649017572402954, "learning_rate": 5.452595713509843e-05, "loss": 0.0259, "step": 49760 }, { "grad_norm": 0.3059203326702118, "learning_rate": 5.4509490069382274e-05, "loss": 0.0251, "step": 49770 }, { "grad_norm": 0.3257799446582794, "learning_rate": 5.449302251051509e-05, "loss": 0.0217, "step": 49780 }, { "grad_norm": 0.29618921875953674, "learning_rate": 5.44765544602977e-05, "loss": 0.0178, "step": 49790 }, { "grad_norm": 0.4170095920562744, "learning_rate": 5.446008592053107e-05, "loss": 0.0277, "step": 49800 }, { "grad_norm": 0.23701757192611694, "learning_rate": 5.4443616893016135e-05, "loss": 0.0285, "step": 49810 }, { "grad_norm": 0.3015690743923187, "learning_rate": 5.4427147379553934e-05, "loss": 0.0286, "step": 49820 }, { "grad_norm": 0.2789212167263031, "learning_rate": 5.441067738194556e-05, "loss": 0.0212, "step": 49830 }, { "grad_norm": 0.3845679759979248, "learning_rate": 5.439420690199214e-05, "loss": 0.0368, "step": 49840 }, { "grad_norm": 0.3952521085739136, "learning_rate": 5.437773594149484e-05, "loss": 0.028, "step": 49850 }, { "grad_norm": 0.24865971505641937, "learning_rate": 5.436126450225495e-05, "loss": 0.0245, "step": 49860 }, { "grad_norm": 0.3225994110107422, "learning_rate": 5.434479258607371e-05, "loss": 0.0262, "step": 49870 }, { "grad_norm": 0.32375219464302063, "learning_rate": 5.43283201947525e-05, "loss": 0.0284, "step": 49880 }, { "grad_norm": 0.2903909683227539, "learning_rate": 5.431184733009268e-05, "loss": 0.0336, "step": 49890 }, { "grad_norm": 0.2772846817970276, "learning_rate": 5.4295373993895736e-05, "loss": 0.0236, "step": 49900 }, { "grad_norm": 0.39492493867874146, "learning_rate": 5.4278900187963157e-05, "loss": 0.0274, "step": 49910 }, { "grad_norm": 0.2751333713531494, "learning_rate": 5.426242591409646e-05, "loss": 0.0299, "step": 49920 }, { "grad_norm": 0.49552881717681885, "learning_rate": 5.42459511740973e-05, "loss": 0.0248, "step": 49930 }, { "grad_norm": 0.33190152049064636, "learning_rate": 5.422947596976729e-05, "loss": 0.0236, "step": 49940 }, { "grad_norm": 0.44892433285713196, "learning_rate": 5.4213000302908134e-05, "loss": 0.0289, "step": 49950 }, { "grad_norm": 0.5508551597595215, "learning_rate": 5.419652417532162e-05, "loss": 0.0222, "step": 49960 }, { "grad_norm": 0.3464840054512024, "learning_rate": 5.4180047588809534e-05, "loss": 0.0248, "step": 49970 }, { "grad_norm": 0.41610807180404663, "learning_rate": 5.4163570545173704e-05, "loss": 0.028, "step": 49980 }, { "grad_norm": 0.3544944226741791, "learning_rate": 5.414709304621608e-05, "loss": 0.0247, "step": 49990 }, { "grad_norm": 0.3539729118347168, "learning_rate": 5.413061509373858e-05, "loss": 0.0309, "step": 50000 }, { "grad_norm": 0.5064467787742615, "learning_rate": 5.411413668954324e-05, "loss": 0.0284, "step": 50010 }, { "grad_norm": 0.3064897954463959, "learning_rate": 5.4097657835432083e-05, "loss": 0.0234, "step": 50020 }, { "grad_norm": 0.318765789270401, "learning_rate": 5.408117853320723e-05, "loss": 0.0194, "step": 50030 }, { "grad_norm": 0.291352778673172, "learning_rate": 5.406469878467084e-05, "loss": 0.0315, "step": 50040 }, { "grad_norm": 0.3400990664958954, "learning_rate": 5.404821859162509e-05, "loss": 0.0275, "step": 50050 }, { "grad_norm": 0.24631020426750183, "learning_rate": 5.403173795587225e-05, "loss": 0.0251, "step": 50060 }, { "grad_norm": 0.4708397686481476, "learning_rate": 5.4015256879214606e-05, "loss": 0.0254, "step": 50070 }, { "grad_norm": 0.3348610997200012, "learning_rate": 5.39987753634545e-05, "loss": 0.0252, "step": 50080 }, { "grad_norm": 0.21889059245586395, "learning_rate": 5.3982293410394335e-05, "loss": 0.0193, "step": 50090 }, { "grad_norm": 0.3132432699203491, "learning_rate": 5.396581102183655e-05, "loss": 0.0262, "step": 50100 }, { "grad_norm": 0.5913732647895813, "learning_rate": 5.3949328199583634e-05, "loss": 0.0218, "step": 50110 }, { "grad_norm": 0.39440885186195374, "learning_rate": 5.393284494543813e-05, "loss": 0.0313, "step": 50120 }, { "grad_norm": 0.21963901817798615, "learning_rate": 5.391636126120262e-05, "loss": 0.028, "step": 50130 }, { "grad_norm": 0.2912330627441406, "learning_rate": 5.389987714867971e-05, "loss": 0.0262, "step": 50140 }, { "grad_norm": 0.3225734531879425, "learning_rate": 5.3883392609672114e-05, "loss": 0.0207, "step": 50150 }, { "grad_norm": 0.36234042048454285, "learning_rate": 5.3866907645982545e-05, "loss": 0.034, "step": 50160 }, { "grad_norm": 0.3477833569049835, "learning_rate": 5.385042225941375e-05, "loss": 0.0283, "step": 50170 }, { "grad_norm": 0.35179808735847473, "learning_rate": 5.3833936451768585e-05, "loss": 0.0281, "step": 50180 }, { "grad_norm": 0.30882367491722107, "learning_rate": 5.3817450224849884e-05, "loss": 0.0266, "step": 50190 }, { "grad_norm": 0.36094069480895996, "learning_rate": 5.380096358046056e-05, "loss": 0.0406, "step": 50200 }, { "grad_norm": 0.3072097599506378, "learning_rate": 5.378447652040359e-05, "loss": 0.0256, "step": 50210 }, { "grad_norm": 0.24779003858566284, "learning_rate": 5.376798904648195e-05, "loss": 0.0228, "step": 50220 }, { "grad_norm": 0.35015740990638733, "learning_rate": 5.375150116049869e-05, "loss": 0.021, "step": 50230 }, { "grad_norm": 0.44033974409103394, "learning_rate": 5.373501286425691e-05, "loss": 0.0235, "step": 50240 }, { "grad_norm": 0.2491479516029358, "learning_rate": 5.3718524159559726e-05, "loss": 0.024, "step": 50250 }, { "grad_norm": 0.3713734745979309, "learning_rate": 5.370203504821034e-05, "loss": 0.0353, "step": 50260 }, { "grad_norm": 0.19890324771404266, "learning_rate": 5.368554553201196e-05, "loss": 0.0291, "step": 50270 }, { "grad_norm": 0.389494925737381, "learning_rate": 5.3669055612767874e-05, "loss": 0.0359, "step": 50280 }, { "grad_norm": 0.47988268733024597, "learning_rate": 5.3652565292281376e-05, "loss": 0.0255, "step": 50290 }, { "grad_norm": 0.803365170955658, "learning_rate": 5.3636074572355824e-05, "loss": 0.0268, "step": 50300 }, { "grad_norm": 0.46567946672439575, "learning_rate": 5.361958345479463e-05, "loss": 0.0302, "step": 50310 }, { "grad_norm": 0.35131245851516724, "learning_rate": 5.360309194140123e-05, "loss": 0.032, "step": 50320 }, { "grad_norm": 0.29298362135887146, "learning_rate": 5.3586600033979106e-05, "loss": 0.0276, "step": 50330 }, { "grad_norm": 0.3025685250759125, "learning_rate": 5.357010773433181e-05, "loss": 0.0186, "step": 50340 }, { "grad_norm": 0.3471362590789795, "learning_rate": 5.3553615044262895e-05, "loss": 0.0224, "step": 50350 }, { "grad_norm": 0.27721068263053894, "learning_rate": 5.3537121965575964e-05, "loss": 0.028, "step": 50360 }, { "grad_norm": 0.329842209815979, "learning_rate": 5.352062850007471e-05, "loss": 0.0272, "step": 50370 }, { "grad_norm": 0.334377259016037, "learning_rate": 5.350413464956282e-05, "loss": 0.0227, "step": 50380 }, { "grad_norm": 0.4240141808986664, "learning_rate": 5.348764041584403e-05, "loss": 0.0265, "step": 50390 }, { "grad_norm": 0.29607802629470825, "learning_rate": 5.347114580072212e-05, "loss": 0.0303, "step": 50400 }, { "grad_norm": 0.43969956040382385, "learning_rate": 5.345465080600093e-05, "loss": 0.0253, "step": 50410 }, { "grad_norm": 0.3778488039970398, "learning_rate": 5.3438155433484326e-05, "loss": 0.0198, "step": 50420 }, { "grad_norm": 0.3048415184020996, "learning_rate": 5.3421659684976197e-05, "loss": 0.0224, "step": 50430 }, { "grad_norm": 0.31425002217292786, "learning_rate": 5.340516356228052e-05, "loss": 0.0273, "step": 50440 }, { "grad_norm": 0.32025936245918274, "learning_rate": 5.338866706720128e-05, "loss": 0.0292, "step": 50450 }, { "grad_norm": 0.33806848526000977, "learning_rate": 5.337217020154249e-05, "loss": 0.0239, "step": 50460 }, { "grad_norm": 0.32815447449684143, "learning_rate": 5.335567296710825e-05, "loss": 0.0248, "step": 50470 }, { "grad_norm": 0.30787524580955505, "learning_rate": 5.333917536570265e-05, "loss": 0.0242, "step": 50480 }, { "grad_norm": 0.5901907086372375, "learning_rate": 5.332267739912986e-05, "loss": 0.021, "step": 50490 }, { "grad_norm": 0.23963597416877747, "learning_rate": 5.330617906919405e-05, "loss": 0.0291, "step": 50500 }, { "grad_norm": 0.22734971344470978, "learning_rate": 5.328968037769949e-05, "loss": 0.024, "step": 50510 }, { "grad_norm": 0.2862837016582489, "learning_rate": 5.327318132645042e-05, "loss": 0.0258, "step": 50520 }, { "grad_norm": 0.2805069386959076, "learning_rate": 5.325668191725116e-05, "loss": 0.022, "step": 50530 }, { "grad_norm": 0.2713545560836792, "learning_rate": 5.324018215190606e-05, "loss": 0.0226, "step": 50540 }, { "grad_norm": 0.46515974402427673, "learning_rate": 5.3223682032219515e-05, "loss": 0.0224, "step": 50550 }, { "grad_norm": 0.4667949974536896, "learning_rate": 5.320718155999595e-05, "loss": 0.0259, "step": 50560 }, { "grad_norm": 0.3201805353164673, "learning_rate": 5.3190680737039835e-05, "loss": 0.0293, "step": 50570 }, { "grad_norm": 0.29283350706100464, "learning_rate": 5.317417956515567e-05, "loss": 0.0217, "step": 50580 }, { "grad_norm": 0.4492204785346985, "learning_rate": 5.315767804614798e-05, "loss": 0.0238, "step": 50590 }, { "grad_norm": 0.4866330921649933, "learning_rate": 5.3141176181821395e-05, "loss": 0.032, "step": 50600 }, { "grad_norm": 0.6224801540374756, "learning_rate": 5.31246739739805e-05, "loss": 0.0194, "step": 50610 }, { "grad_norm": 0.26871129870414734, "learning_rate": 5.310817142442994e-05, "loss": 0.0304, "step": 50620 }, { "grad_norm": 0.4702262282371521, "learning_rate": 5.309166853497445e-05, "loss": 0.027, "step": 50630 }, { "grad_norm": 0.3537381887435913, "learning_rate": 5.307516530741873e-05, "loss": 0.0241, "step": 50640 }, { "grad_norm": 0.34290578961372375, "learning_rate": 5.305866174356754e-05, "loss": 0.0245, "step": 50650 }, { "grad_norm": 0.27099576592445374, "learning_rate": 5.304215784522571e-05, "loss": 0.024, "step": 50660 }, { "grad_norm": 0.24340113997459412, "learning_rate": 5.302565361419808e-05, "loss": 0.0322, "step": 50670 }, { "grad_norm": 0.25805267691612244, "learning_rate": 5.3009149052289507e-05, "loss": 0.0335, "step": 50680 }, { "grad_norm": 0.2569923400878906, "learning_rate": 5.299264416130493e-05, "loss": 0.0206, "step": 50690 }, { "grad_norm": 0.3437409996986389, "learning_rate": 5.297613894304928e-05, "loss": 0.0252, "step": 50700 }, { "grad_norm": 0.26115158200263977, "learning_rate": 5.2959633399327534e-05, "loss": 0.0221, "step": 50710 }, { "grad_norm": 0.2676180601119995, "learning_rate": 5.294312753194476e-05, "loss": 0.0262, "step": 50720 }, { "grad_norm": 0.308861643075943, "learning_rate": 5.292662134270596e-05, "loss": 0.0287, "step": 50730 }, { "grad_norm": 0.3003784120082855, "learning_rate": 5.291011483341626e-05, "loss": 0.0298, "step": 50740 }, { "grad_norm": 0.9565918445587158, "learning_rate": 5.2893608005880767e-05, "loss": 0.0326, "step": 50750 }, { "grad_norm": 0.48391997814178467, "learning_rate": 5.287710086190467e-05, "loss": 0.0324, "step": 50760 }, { "grad_norm": 0.5307325720787048, "learning_rate": 5.286059340329314e-05, "loss": 0.0223, "step": 50770 }, { "grad_norm": 0.5074331164360046, "learning_rate": 5.284408563185141e-05, "loss": 0.0298, "step": 50780 }, { "grad_norm": 0.9602169990539551, "learning_rate": 5.282757754938476e-05, "loss": 0.0258, "step": 50790 }, { "grad_norm": 0.32280534505844116, "learning_rate": 5.281106915769849e-05, "loss": 0.0297, "step": 50800 }, { "grad_norm": 0.2227359265089035, "learning_rate": 5.2794560458597897e-05, "loss": 0.0216, "step": 50810 }, { "grad_norm": 0.2775988280773163, "learning_rate": 5.27780514538884e-05, "loss": 0.0299, "step": 50820 }, { "grad_norm": 0.27864742279052734, "learning_rate": 5.2761542145375365e-05, "loss": 0.0251, "step": 50830 }, { "grad_norm": 0.3254223167896271, "learning_rate": 5.274503253486421e-05, "loss": 0.0258, "step": 50840 }, { "grad_norm": 0.505204975605011, "learning_rate": 5.272852262416046e-05, "loss": 0.0235, "step": 50850 }, { "grad_norm": 0.41259434819221497, "learning_rate": 5.2712012415069555e-05, "loss": 0.0311, "step": 50860 }, { "grad_norm": 0.5118359923362732, "learning_rate": 5.269550190939705e-05, "loss": 0.0218, "step": 50870 }, { "grad_norm": 0.4698472321033478, "learning_rate": 5.267899110894852e-05, "loss": 0.0214, "step": 50880 }, { "grad_norm": 0.22254633903503418, "learning_rate": 5.266248001552955e-05, "loss": 0.0205, "step": 50890 }, { "grad_norm": 0.3048785328865051, "learning_rate": 5.264596863094575e-05, "loss": 0.0256, "step": 50900 }, { "grad_norm": 0.5607914924621582, "learning_rate": 5.262945695700282e-05, "loss": 0.0248, "step": 50910 }, { "grad_norm": 0.30575665831565857, "learning_rate": 5.261294499550643e-05, "loss": 0.0282, "step": 50920 }, { "grad_norm": 0.3241969645023346, "learning_rate": 5.25964327482623e-05, "loss": 0.0234, "step": 50930 }, { "grad_norm": 0.33931824564933777, "learning_rate": 5.257992021707617e-05, "loss": 0.0261, "step": 50940 }, { "grad_norm": 0.45942145586013794, "learning_rate": 5.256340740375387e-05, "loss": 0.0299, "step": 50950 }, { "grad_norm": 0.2829381823539734, "learning_rate": 5.254689431010117e-05, "loss": 0.0294, "step": 50960 }, { "grad_norm": 0.2138226479291916, "learning_rate": 5.253038093792395e-05, "loss": 0.0182, "step": 50970 }, { "grad_norm": 0.3476429283618927, "learning_rate": 5.251386728902806e-05, "loss": 0.0224, "step": 50980 }, { "grad_norm": 0.37476611137390137, "learning_rate": 5.2497353365219446e-05, "loss": 0.0252, "step": 50990 }, { "grad_norm": 0.5039577484130859, "learning_rate": 5.2480839168304e-05, "loss": 0.0242, "step": 51000 }, { "grad_norm": 0.2997267246246338, "learning_rate": 5.246432470008773e-05, "loss": 0.0263, "step": 51010 }, { "grad_norm": 0.25927817821502686, "learning_rate": 5.2447809962376616e-05, "loss": 0.0213, "step": 51020 }, { "grad_norm": 0.39388617873191833, "learning_rate": 5.243129495697666e-05, "loss": 0.0274, "step": 51030 }, { "grad_norm": 0.28588417172431946, "learning_rate": 5.241477968569397e-05, "loss": 0.0291, "step": 51040 }, { "grad_norm": 0.4167303144931793, "learning_rate": 5.23982641503346e-05, "loss": 0.0239, "step": 51050 }, { "grad_norm": 0.3665262758731842, "learning_rate": 5.238174835270465e-05, "loss": 0.028, "step": 51060 }, { "grad_norm": 0.31835639476776123, "learning_rate": 5.23652322946103e-05, "loss": 0.0169, "step": 51070 }, { "grad_norm": 0.3819500803947449, "learning_rate": 5.234871597785771e-05, "loss": 0.0203, "step": 51080 }, { "grad_norm": 0.38617685437202454, "learning_rate": 5.233219940425306e-05, "loss": 0.0268, "step": 51090 }, { "grad_norm": 0.4534214735031128, "learning_rate": 5.2315682575602576e-05, "loss": 0.0407, "step": 51100 }, { "grad_norm": 0.3816022574901581, "learning_rate": 5.229916549371254e-05, "loss": 0.0219, "step": 51110 }, { "grad_norm": 0.2923247516155243, "learning_rate": 5.228264816038924e-05, "loss": 0.0253, "step": 51120 }, { "grad_norm": 0.33414173126220703, "learning_rate": 5.2266130577438935e-05, "loss": 0.0266, "step": 51130 }, { "grad_norm": 0.4312891960144043, "learning_rate": 5.224961274666802e-05, "loss": 0.0292, "step": 51140 }, { "grad_norm": 0.24804991483688354, "learning_rate": 5.223309466988283e-05, "loss": 0.0333, "step": 51150 }, { "grad_norm": 0.34536126255989075, "learning_rate": 5.221657634888976e-05, "loss": 0.0299, "step": 51160 }, { "grad_norm": 0.34597116708755493, "learning_rate": 5.2200057785495235e-05, "loss": 0.0324, "step": 51170 }, { "grad_norm": 0.30278849601745605, "learning_rate": 5.2183538981505706e-05, "loss": 0.0261, "step": 51180 }, { "grad_norm": 0.3500075340270996, "learning_rate": 5.216701993872762e-05, "loss": 0.0278, "step": 51190 }, { "grad_norm": 0.22025485336780548, "learning_rate": 5.215050065896752e-05, "loss": 0.034, "step": 51200 }, { "grad_norm": 0.49765926599502563, "learning_rate": 5.2133981144031875e-05, "loss": 0.0392, "step": 51210 }, { "grad_norm": 0.6812213659286499, "learning_rate": 5.211746139572727e-05, "loss": 0.0234, "step": 51220 }, { "grad_norm": 0.2649313509464264, "learning_rate": 5.210094141586026e-05, "loss": 0.0273, "step": 51230 }, { "grad_norm": 0.42350873351097107, "learning_rate": 5.2084421206237476e-05, "loss": 0.0349, "step": 51240 }, { "grad_norm": 0.28389009833335876, "learning_rate": 5.20679007686655e-05, "loss": 0.0321, "step": 51250 }, { "grad_norm": 0.5377870798110962, "learning_rate": 5.205138010495102e-05, "loss": 0.0291, "step": 51260 }, { "grad_norm": 0.43215498328208923, "learning_rate": 5.20348592169007e-05, "loss": 0.0376, "step": 51270 }, { "grad_norm": 0.5716677904129028, "learning_rate": 5.201833810632123e-05, "loss": 0.0298, "step": 51280 }, { "grad_norm": 0.3309156596660614, "learning_rate": 5.200181677501933e-05, "loss": 0.0251, "step": 51290 }, { "grad_norm": 0.22127273678779602, "learning_rate": 5.1985295224801777e-05, "loss": 0.0216, "step": 51300 }, { "grad_norm": 0.49426332116127014, "learning_rate": 5.196877345747531e-05, "loss": 0.0294, "step": 51310 }, { "grad_norm": 0.3257420063018799, "learning_rate": 5.195225147484674e-05, "loss": 0.0306, "step": 51320 }, { "grad_norm": 0.3992244005203247, "learning_rate": 5.193572927872288e-05, "loss": 0.0283, "step": 51330 }, { "grad_norm": 0.40848565101623535, "learning_rate": 5.191920687091059e-05, "loss": 0.028, "step": 51340 }, { "grad_norm": 0.3868333399295807, "learning_rate": 5.1902684253216695e-05, "loss": 0.029, "step": 51350 }, { "grad_norm": 0.2445162832736969, "learning_rate": 5.1886161427448134e-05, "loss": 0.0247, "step": 51360 }, { "grad_norm": 0.3459603190422058, "learning_rate": 5.186963839541179e-05, "loss": 0.0264, "step": 51370 }, { "grad_norm": 0.274763286113739, "learning_rate": 5.185311515891459e-05, "loss": 0.0214, "step": 51380 }, { "grad_norm": 0.32869023084640503, "learning_rate": 5.1836591719763516e-05, "loss": 0.0254, "step": 51390 }, { "grad_norm": 0.2797048091888428, "learning_rate": 5.182006807976554e-05, "loss": 0.0243, "step": 51400 }, { "grad_norm": 0.20721650123596191, "learning_rate": 5.180354424072763e-05, "loss": 0.0197, "step": 51410 }, { "grad_norm": 0.34058669209480286, "learning_rate": 5.178702020445686e-05, "loss": 0.028, "step": 51420 }, { "grad_norm": 0.2804257571697235, "learning_rate": 5.177049597276024e-05, "loss": 0.0224, "step": 51430 }, { "grad_norm": 0.2797638177871704, "learning_rate": 5.175397154744482e-05, "loss": 0.0305, "step": 51440 }, { "grad_norm": 0.2736355662345886, "learning_rate": 5.1737446930317736e-05, "loss": 0.018, "step": 51450 }, { "grad_norm": 0.290911465883255, "learning_rate": 5.172092212318604e-05, "loss": 0.0249, "step": 51460 }, { "grad_norm": 0.2280263602733612, "learning_rate": 5.170439712785692e-05, "loss": 0.027, "step": 51470 }, { "grad_norm": 0.6439094543457031, "learning_rate": 5.1687871946137467e-05, "loss": 0.0305, "step": 51480 }, { "grad_norm": 0.3240452706813812, "learning_rate": 5.16713465798349e-05, "loss": 0.0209, "step": 51490 }, { "grad_norm": 0.3647768497467041, "learning_rate": 5.165482103075637e-05, "loss": 0.0277, "step": 51500 }, { "grad_norm": 0.4379386007785797, "learning_rate": 5.163829530070909e-05, "loss": 0.0259, "step": 51510 }, { "grad_norm": 0.3340524137020111, "learning_rate": 5.162176939150032e-05, "loss": 0.0267, "step": 51520 }, { "grad_norm": 0.32181641459465027, "learning_rate": 5.160524330493728e-05, "loss": 0.0318, "step": 51530 }, { "grad_norm": 0.24099217355251312, "learning_rate": 5.158871704282724e-05, "loss": 0.0183, "step": 51540 }, { "grad_norm": 0.3267260789871216, "learning_rate": 5.15721906069775e-05, "loss": 0.0254, "step": 51550 }, { "grad_norm": 0.23159551620483398, "learning_rate": 5.155566399919536e-05, "loss": 0.0199, "step": 51560 }, { "grad_norm": 0.3462044596672058, "learning_rate": 5.153913722128813e-05, "loss": 0.0362, "step": 51570 }, { "grad_norm": 0.23095181584358215, "learning_rate": 5.1522610275063176e-05, "loss": 0.0233, "step": 51580 }, { "grad_norm": 0.3350890874862671, "learning_rate": 5.150608316232786e-05, "loss": 0.0237, "step": 51590 }, { "grad_norm": 0.3075977861881256, "learning_rate": 5.148955588488955e-05, "loss": 0.0272, "step": 51600 }, { "grad_norm": 0.3606542944908142, "learning_rate": 5.147302844455564e-05, "loss": 0.0224, "step": 51610 }, { "grad_norm": 0.3149911165237427, "learning_rate": 5.145650084313356e-05, "loss": 0.0245, "step": 51620 }, { "grad_norm": 0.3853667676448822, "learning_rate": 5.1439973082430734e-05, "loss": 0.0226, "step": 51630 }, { "grad_norm": 0.28645503520965576, "learning_rate": 5.142344516425461e-05, "loss": 0.0258, "step": 51640 }, { "grad_norm": 0.37014132738113403, "learning_rate": 5.1406917090412673e-05, "loss": 0.0349, "step": 51650 }, { "grad_norm": 0.5450400710105896, "learning_rate": 5.1390388862712404e-05, "loss": 0.0327, "step": 51660 }, { "grad_norm": 0.360297828912735, "learning_rate": 5.137386048296129e-05, "loss": 0.0249, "step": 51670 }, { "grad_norm": 0.49935224652290344, "learning_rate": 5.135733195296686e-05, "loss": 0.0322, "step": 51680 }, { "grad_norm": 0.3763391673564911, "learning_rate": 5.134080327453664e-05, "loss": 0.0245, "step": 51690 }, { "grad_norm": 0.30136051774024963, "learning_rate": 5.1324274449478204e-05, "loss": 0.0227, "step": 51700 }, { "grad_norm": 0.2700326144695282, "learning_rate": 5.130774547959909e-05, "loss": 0.025, "step": 51710 }, { "grad_norm": 0.4506356716156006, "learning_rate": 5.129121636670692e-05, "loss": 0.0315, "step": 51720 }, { "grad_norm": 0.2956100404262543, "learning_rate": 5.1274687112609246e-05, "loss": 0.023, "step": 51730 }, { "grad_norm": 0.4313170611858368, "learning_rate": 5.125815771911372e-05, "loss": 0.0292, "step": 51740 }, { "grad_norm": 0.4308822751045227, "learning_rate": 5.1241628188027956e-05, "loss": 0.0235, "step": 51750 }, { "grad_norm": 0.2969432771205902, "learning_rate": 5.1225098521159585e-05, "loss": 0.0212, "step": 51760 }, { "grad_norm": 0.33727800846099854, "learning_rate": 5.12085687203163e-05, "loss": 0.0228, "step": 51770 }, { "grad_norm": 0.30680471658706665, "learning_rate": 5.119203878730575e-05, "loss": 0.0242, "step": 51780 }, { "grad_norm": 0.36562657356262207, "learning_rate": 5.117550872393564e-05, "loss": 0.0295, "step": 51790 }, { "grad_norm": 0.3088717460632324, "learning_rate": 5.1158978532013645e-05, "loss": 0.0253, "step": 51800 }, { "grad_norm": 0.3396584987640381, "learning_rate": 5.114244821334751e-05, "loss": 0.0281, "step": 51810 }, { "grad_norm": 0.30952346324920654, "learning_rate": 5.112591776974497e-05, "loss": 0.0195, "step": 51820 }, { "grad_norm": 0.3118550777435303, "learning_rate": 5.110938720301374e-05, "loss": 0.025, "step": 51830 }, { "grad_norm": 0.2437748908996582, "learning_rate": 5.1092856514961595e-05, "loss": 0.0278, "step": 51840 }, { "grad_norm": 0.2857590615749359, "learning_rate": 5.107632570739631e-05, "loss": 0.0269, "step": 51850 }, { "grad_norm": 0.3603670597076416, "learning_rate": 5.1059794782125645e-05, "loss": 0.0219, "step": 51860 }, { "grad_norm": 0.4404747784137726, "learning_rate": 5.104326374095744e-05, "loss": 0.029, "step": 51870 }, { "grad_norm": 0.40155714750289917, "learning_rate": 5.102673258569947e-05, "loss": 0.0267, "step": 51880 }, { "grad_norm": 0.3565959334373474, "learning_rate": 5.101020131815956e-05, "loss": 0.0315, "step": 51890 }, { "grad_norm": 0.2876417338848114, "learning_rate": 5.099366994014556e-05, "loss": 0.0245, "step": 51900 }, { "grad_norm": 0.38855862617492676, "learning_rate": 5.097713845346532e-05, "loss": 0.0356, "step": 51910 }, { "grad_norm": 0.30874761939048767, "learning_rate": 5.096060685992668e-05, "loss": 0.0254, "step": 51920 }, { "grad_norm": 0.30423492193222046, "learning_rate": 5.0944075161337514e-05, "loss": 0.0267, "step": 51930 }, { "grad_norm": 0.3504886329174042, "learning_rate": 5.09275433595057e-05, "loss": 0.0285, "step": 51940 }, { "grad_norm": 0.30857574939727783, "learning_rate": 5.0911011456239157e-05, "loss": 0.0307, "step": 51950 }, { "grad_norm": 0.41507789492607117, "learning_rate": 5.089447945334574e-05, "loss": 0.0273, "step": 51960 }, { "grad_norm": 0.3095220923423767, "learning_rate": 5.087794735263343e-05, "loss": 0.0194, "step": 51970 }, { "grad_norm": 0.17470265924930573, "learning_rate": 5.086141515591011e-05, "loss": 0.0205, "step": 51980 }, { "grad_norm": 0.36336109042167664, "learning_rate": 5.084488286498371e-05, "loss": 0.0306, "step": 51990 }, { "grad_norm": 0.4513590335845947, "learning_rate": 5.08283504816622e-05, "loss": 0.0306, "step": 52000 }, { "grad_norm": 0.3735635578632355, "learning_rate": 5.081181800775353e-05, "loss": 0.0324, "step": 52010 }, { "grad_norm": 0.3252054452896118, "learning_rate": 5.079528544506565e-05, "loss": 0.0249, "step": 52020 }, { "grad_norm": 0.24784806370735168, "learning_rate": 5.077875279540657e-05, "loss": 0.0232, "step": 52030 }, { "grad_norm": 0.23753345012664795, "learning_rate": 5.076222006058424e-05, "loss": 0.0259, "step": 52040 }, { "grad_norm": 0.32624030113220215, "learning_rate": 5.074568724240667e-05, "loss": 0.0237, "step": 52050 }, { "grad_norm": 0.22988572716712952, "learning_rate": 5.072915434268187e-05, "loss": 0.0279, "step": 52060 }, { "grad_norm": 0.36771342158317566, "learning_rate": 5.071262136321785e-05, "loss": 0.0293, "step": 52070 }, { "grad_norm": 0.4027791917324066, "learning_rate": 5.069608830582262e-05, "loss": 0.0263, "step": 52080 }, { "grad_norm": 0.25662729144096375, "learning_rate": 5.067955517230423e-05, "loss": 0.0246, "step": 52090 }, { "grad_norm": 0.21485520899295807, "learning_rate": 5.06630219644707e-05, "loss": 0.0281, "step": 52100 }, { "grad_norm": 0.333342581987381, "learning_rate": 5.064648868413008e-05, "loss": 0.026, "step": 52110 }, { "grad_norm": 0.32975614070892334, "learning_rate": 5.062995533309044e-05, "loss": 0.0236, "step": 52120 }, { "grad_norm": 0.21670791506767273, "learning_rate": 5.061342191315983e-05, "loss": 0.0289, "step": 52130 }, { "grad_norm": 0.2830848693847656, "learning_rate": 5.0596888426146325e-05, "loss": 0.0212, "step": 52140 }, { "grad_norm": 0.30175963044166565, "learning_rate": 5.058035487385798e-05, "loss": 0.021, "step": 52150 }, { "grad_norm": 0.3878920078277588, "learning_rate": 5.0563821258102915e-05, "loss": 0.0195, "step": 52160 }, { "grad_norm": 0.2931576371192932, "learning_rate": 5.05472875806892e-05, "loss": 0.0184, "step": 52170 }, { "grad_norm": 0.5078902840614319, "learning_rate": 5.053075384342494e-05, "loss": 0.0322, "step": 52180 }, { "grad_norm": 0.9873950481414795, "learning_rate": 5.0514220048118234e-05, "loss": 0.0262, "step": 52190 }, { "grad_norm": 0.35370829701423645, "learning_rate": 5.049768619657721e-05, "loss": 0.0285, "step": 52200 }, { "grad_norm": 0.38510164618492126, "learning_rate": 5.048115229060996e-05, "loss": 0.0296, "step": 52210 }, { "grad_norm": 0.28007665276527405, "learning_rate": 5.046461833202462e-05, "loss": 0.0275, "step": 52220 }, { "grad_norm": 0.28462451696395874, "learning_rate": 5.044808432262932e-05, "loss": 0.0351, "step": 52230 }, { "grad_norm": 0.4234900176525116, "learning_rate": 5.043155026423219e-05, "loss": 0.0243, "step": 52240 }, { "grad_norm": 0.30540478229522705, "learning_rate": 5.041501615864137e-05, "loss": 0.0209, "step": 52250 }, { "grad_norm": 0.32651445269584656, "learning_rate": 5.039848200766502e-05, "loss": 0.0234, "step": 52260 }, { "grad_norm": 0.27373528480529785, "learning_rate": 5.038194781311125e-05, "loss": 0.0202, "step": 52270 }, { "grad_norm": 0.4195720851421356, "learning_rate": 5.0365413576788264e-05, "loss": 0.0206, "step": 52280 }, { "grad_norm": 0.294599711894989, "learning_rate": 5.0348879300504194e-05, "loss": 0.0236, "step": 52290 }, { "grad_norm": 0.2738566994667053, "learning_rate": 5.033234498606721e-05, "loss": 0.0193, "step": 52300 }, { "grad_norm": 0.34701022505760193, "learning_rate": 5.0315810635285476e-05, "loss": 0.0224, "step": 52310 }, { "grad_norm": 0.4265165627002716, "learning_rate": 5.0299276249967164e-05, "loss": 0.0248, "step": 52320 }, { "grad_norm": 0.2785396873950958, "learning_rate": 5.0282741831920454e-05, "loss": 0.0428, "step": 52330 }, { "grad_norm": 0.47883322834968567, "learning_rate": 5.026620738295351e-05, "loss": 0.0278, "step": 52340 }, { "grad_norm": 0.2584657073020935, "learning_rate": 5.024967290487455e-05, "loss": 0.0262, "step": 52350 }, { "grad_norm": 0.33698806166648865, "learning_rate": 5.023313839949172e-05, "loss": 0.0243, "step": 52360 }, { "grad_norm": 0.264157772064209, "learning_rate": 5.0216603868613224e-05, "loss": 0.0239, "step": 52370 }, { "grad_norm": 0.21853046119213104, "learning_rate": 5.020006931404726e-05, "loss": 0.0288, "step": 52380 }, { "grad_norm": 0.2733268141746521, "learning_rate": 5.0183534737602014e-05, "loss": 0.0214, "step": 52390 }, { "grad_norm": 0.3079827129840851, "learning_rate": 5.0167000141085686e-05, "loss": 0.0278, "step": 52400 }, { "grad_norm": 0.2829439640045166, "learning_rate": 5.015046552630647e-05, "loss": 0.0217, "step": 52410 }, { "grad_norm": 0.31652316451072693, "learning_rate": 5.013393089507258e-05, "loss": 0.0306, "step": 52420 }, { "grad_norm": 0.27164286375045776, "learning_rate": 5.011739624919219e-05, "loss": 0.0236, "step": 52430 }, { "grad_norm": 0.22081917524337769, "learning_rate": 5.0100861590473545e-05, "loss": 0.0216, "step": 52440 }, { "grad_norm": 0.29212507605552673, "learning_rate": 5.008432692072482e-05, "loss": 0.0204, "step": 52450 }, { "grad_norm": 0.27899324893951416, "learning_rate": 5.006779224175423e-05, "loss": 0.0273, "step": 52460 }, { "grad_norm": 0.3367869555950165, "learning_rate": 5.005125755536998e-05, "loss": 0.0272, "step": 52470 }, { "grad_norm": 0.22621263563632965, "learning_rate": 5.003472286338029e-05, "loss": 0.0269, "step": 52480 }, { "grad_norm": 0.2644486725330353, "learning_rate": 5.001818816759335e-05, "loss": 0.0296, "step": 52490 }, { "grad_norm": 0.2928752899169922, "learning_rate": 5.000165346981738e-05, "loss": 0.0237, "step": 52500 }, { "grad_norm": 0.33436909317970276, "learning_rate": 4.9985118771860595e-05, "loss": 0.0271, "step": 52510 }, { "grad_norm": 0.37021759152412415, "learning_rate": 4.996858407553119e-05, "loss": 0.0226, "step": 52520 }, { "grad_norm": 0.25110864639282227, "learning_rate": 4.99520493826374e-05, "loss": 0.0201, "step": 52530 }, { "grad_norm": 0.41873154044151306, "learning_rate": 4.9935514694987384e-05, "loss": 0.0283, "step": 52540 }, { "grad_norm": 0.3405953645706177, "learning_rate": 4.991898001438941e-05, "loss": 0.0226, "step": 52550 }, { "grad_norm": 0.27335911989212036, "learning_rate": 4.990244534265166e-05, "loss": 0.0299, "step": 52560 }, { "grad_norm": 0.44461119174957275, "learning_rate": 4.988591068158234e-05, "loss": 0.0268, "step": 52570 }, { "grad_norm": 0.2896435856819153, "learning_rate": 4.986937603298965e-05, "loss": 0.0251, "step": 52580 }, { "grad_norm": 0.30187100172042847, "learning_rate": 4.985284139868181e-05, "loss": 0.0178, "step": 52590 }, { "grad_norm": 0.32869577407836914, "learning_rate": 4.9836306780467e-05, "loss": 0.0237, "step": 52600 }, { "grad_norm": 0.2720457911491394, "learning_rate": 4.981977218015343e-05, "loss": 0.0227, "step": 52610 }, { "grad_norm": 0.28490543365478516, "learning_rate": 4.980323759954932e-05, "loss": 0.0218, "step": 52620 }, { "grad_norm": 0.27822914719581604, "learning_rate": 4.978670304046284e-05, "loss": 0.0214, "step": 52630 }, { "grad_norm": 0.3155827820301056, "learning_rate": 4.9770168504702195e-05, "loss": 0.0226, "step": 52640 }, { "grad_norm": 0.3299735188484192, "learning_rate": 4.975363399407559e-05, "loss": 0.0203, "step": 52650 }, { "grad_norm": 0.38631317019462585, "learning_rate": 4.9737099510391194e-05, "loss": 0.0267, "step": 52660 }, { "grad_norm": 0.2498200237751007, "learning_rate": 4.972056505545719e-05, "loss": 0.0285, "step": 52670 }, { "grad_norm": 0.3255429267883301, "learning_rate": 4.9704030631081775e-05, "loss": 0.0215, "step": 52680 }, { "grad_norm": 0.34613844752311707, "learning_rate": 4.968749623907313e-05, "loss": 0.0229, "step": 52690 }, { "grad_norm": 0.2630433142185211, "learning_rate": 4.967096188123943e-05, "loss": 0.0308, "step": 52700 }, { "grad_norm": 0.31784379482269287, "learning_rate": 4.965442755938884e-05, "loss": 0.0206, "step": 52710 }, { "grad_norm": 0.35422655940055847, "learning_rate": 4.963789327532954e-05, "loss": 0.0218, "step": 52720 }, { "grad_norm": 0.3334479331970215, "learning_rate": 4.962135903086968e-05, "loss": 0.0253, "step": 52730 }, { "grad_norm": 0.2652621865272522, "learning_rate": 4.960482482781741e-05, "loss": 0.0204, "step": 52740 }, { "grad_norm": 0.2783033549785614, "learning_rate": 4.958829066798092e-05, "loss": 0.0246, "step": 52750 }, { "grad_norm": 0.21885724365711212, "learning_rate": 4.957175655316835e-05, "loss": 0.0238, "step": 52760 }, { "grad_norm": 0.3895050585269928, "learning_rate": 4.9555222485187834e-05, "loss": 0.0276, "step": 52770 }, { "grad_norm": 0.3296109139919281, "learning_rate": 4.953868846584752e-05, "loss": 0.0227, "step": 52780 }, { "grad_norm": 0.35481464862823486, "learning_rate": 4.9522154496955544e-05, "loss": 0.0338, "step": 52790 }, { "grad_norm": 0.25916963815689087, "learning_rate": 4.950562058032001e-05, "loss": 0.0238, "step": 52800 }, { "grad_norm": 0.2324191927909851, "learning_rate": 4.948908671774909e-05, "loss": 0.0326, "step": 52810 }, { "grad_norm": 0.6224437355995178, "learning_rate": 4.947255291105087e-05, "loss": 0.026, "step": 52820 }, { "grad_norm": 0.2999218702316284, "learning_rate": 4.945601916203348e-05, "loss": 0.0324, "step": 52830 }, { "grad_norm": 0.3573894500732422, "learning_rate": 4.9439485472505016e-05, "loss": 0.0306, "step": 52840 }, { "grad_norm": 0.5682352781295776, "learning_rate": 4.942295184427356e-05, "loss": 0.0286, "step": 52850 }, { "grad_norm": 0.3286116421222687, "learning_rate": 4.9406418279147234e-05, "loss": 0.0277, "step": 52860 }, { "grad_norm": 0.7215491533279419, "learning_rate": 4.9389884778934116e-05, "loss": 0.0321, "step": 52870 }, { "grad_norm": 0.20405735075473785, "learning_rate": 4.937335134544228e-05, "loss": 0.0226, "step": 52880 }, { "grad_norm": 0.2245275378227234, "learning_rate": 4.935681798047979e-05, "loss": 0.0267, "step": 52890 }, { "grad_norm": 0.4528082311153412, "learning_rate": 4.934028468585473e-05, "loss": 0.0278, "step": 52900 }, { "grad_norm": 0.30033764243125916, "learning_rate": 4.932375146337514e-05, "loss": 0.0264, "step": 52910 }, { "grad_norm": 0.2612525522708893, "learning_rate": 4.930721831484908e-05, "loss": 0.0217, "step": 52920 }, { "grad_norm": 0.251029908657074, "learning_rate": 4.929068524208456e-05, "loss": 0.0252, "step": 52930 }, { "grad_norm": 0.33512336015701294, "learning_rate": 4.927415224688965e-05, "loss": 0.0365, "step": 52940 }, { "grad_norm": 0.2967614531517029, "learning_rate": 4.9257619331072375e-05, "loss": 0.0346, "step": 52950 }, { "grad_norm": 0.37906718254089355, "learning_rate": 4.924108649644072e-05, "loss": 0.0166, "step": 52960 }, { "grad_norm": 0.3172239065170288, "learning_rate": 4.922455374480271e-05, "loss": 0.0344, "step": 52970 }, { "grad_norm": 0.36289551854133606, "learning_rate": 4.920802107796634e-05, "loss": 0.0277, "step": 52980 }, { "grad_norm": 0.3535827696323395, "learning_rate": 4.9191488497739564e-05, "loss": 0.0216, "step": 52990 }, { "grad_norm": 0.21540047228336334, "learning_rate": 4.917495600593042e-05, "loss": 0.0257, "step": 53000 }, { "grad_norm": 0.35129037499427795, "learning_rate": 4.915842360434684e-05, "loss": 0.0312, "step": 53010 }, { "grad_norm": 0.5489148497581482, "learning_rate": 4.914189129479679e-05, "loss": 0.0396, "step": 53020 }, { "grad_norm": 0.35961419343948364, "learning_rate": 4.912535907908822e-05, "loss": 0.0237, "step": 53030 }, { "grad_norm": 0.378460168838501, "learning_rate": 4.910882695902907e-05, "loss": 0.0291, "step": 53040 }, { "grad_norm": 0.28767216205596924, "learning_rate": 4.9092294936427234e-05, "loss": 0.0252, "step": 53050 }, { "grad_norm": 0.3382442891597748, "learning_rate": 4.9075763013090685e-05, "loss": 0.0208, "step": 53060 }, { "grad_norm": 0.3203352093696594, "learning_rate": 4.9059231190827305e-05, "loss": 0.0334, "step": 53070 }, { "grad_norm": 0.2961473762989044, "learning_rate": 4.904269947144499e-05, "loss": 0.0474, "step": 53080 }, { "grad_norm": 0.3883046805858612, "learning_rate": 4.902616785675161e-05, "loss": 0.0263, "step": 53090 }, { "grad_norm": 0.39392876625061035, "learning_rate": 4.900963634855505e-05, "loss": 0.0301, "step": 53100 }, { "grad_norm": 0.30563271045684814, "learning_rate": 4.8993104948663173e-05, "loss": 0.0255, "step": 53110 }, { "grad_norm": 0.3549749553203583, "learning_rate": 4.8976573658883823e-05, "loss": 0.021, "step": 53120 }, { "grad_norm": 0.2089100331068039, "learning_rate": 4.896004248102484e-05, "loss": 0.0235, "step": 53130 }, { "grad_norm": 0.39067932963371277, "learning_rate": 4.8943511416894045e-05, "loss": 0.0255, "step": 53140 }, { "grad_norm": 0.240203857421875, "learning_rate": 4.892698046829926e-05, "loss": 0.0252, "step": 53150 }, { "grad_norm": 0.3184877932071686, "learning_rate": 4.8910449637048275e-05, "loss": 0.0315, "step": 53160 }, { "grad_norm": 0.3308296799659729, "learning_rate": 4.8893918924948884e-05, "loss": 0.0287, "step": 53170 }, { "grad_norm": 0.30961841344833374, "learning_rate": 4.8877388333808835e-05, "loss": 0.0261, "step": 53180 }, { "grad_norm": 0.3639664947986603, "learning_rate": 4.8860857865435925e-05, "loss": 0.0284, "step": 53190 }, { "grad_norm": 0.26523974537849426, "learning_rate": 4.884432752163789e-05, "loss": 0.0217, "step": 53200 }, { "grad_norm": 0.33016934990882874, "learning_rate": 4.8827797304222466e-05, "loss": 0.0232, "step": 53210 }, { "grad_norm": 0.342862606048584, "learning_rate": 4.8811267214997366e-05, "loss": 0.0198, "step": 53220 }, { "grad_norm": 0.3908994197845459, "learning_rate": 4.879473725577029e-05, "loss": 0.0251, "step": 53230 }, { "grad_norm": 0.3688780963420868, "learning_rate": 4.877820742834891e-05, "loss": 0.02, "step": 53240 }, { "grad_norm": 0.1954425573348999, "learning_rate": 4.8761677734540956e-05, "loss": 0.0208, "step": 53250 }, { "grad_norm": 0.2437206208705902, "learning_rate": 4.874514817615407e-05, "loss": 0.0272, "step": 53260 }, { "grad_norm": 0.22927971184253693, "learning_rate": 4.872861875499588e-05, "loss": 0.0254, "step": 53270 }, { "grad_norm": 0.35478198528289795, "learning_rate": 4.871208947287404e-05, "loss": 0.0323, "step": 53280 }, { "grad_norm": 0.4299429953098297, "learning_rate": 4.869556033159615e-05, "loss": 0.0216, "step": 53290 }, { "grad_norm": 0.19935566186904907, "learning_rate": 4.8679031332969835e-05, "loss": 0.0203, "step": 53300 }, { "grad_norm": 0.3279287815093994, "learning_rate": 4.8662502478802635e-05, "loss": 0.0299, "step": 53310 }, { "grad_norm": 0.36939364671707153, "learning_rate": 4.8645973770902176e-05, "loss": 0.0271, "step": 53320 }, { "grad_norm": 0.31847018003463745, "learning_rate": 4.862944521107599e-05, "loss": 0.0241, "step": 53330 }, { "grad_norm": 0.4491824209690094, "learning_rate": 4.861291680113162e-05, "loss": 0.0267, "step": 53340 }, { "grad_norm": 0.24621260166168213, "learning_rate": 4.8596388542876563e-05, "loss": 0.0176, "step": 53350 }, { "grad_norm": 0.2285156399011612, "learning_rate": 4.857986043811836e-05, "loss": 0.0245, "step": 53360 }, { "grad_norm": 0.4218977391719818, "learning_rate": 4.856333248866448e-05, "loss": 0.026, "step": 53370 }, { "grad_norm": 0.3381163477897644, "learning_rate": 4.85468046963224e-05, "loss": 0.0248, "step": 53380 }, { "grad_norm": 0.32044047117233276, "learning_rate": 4.853027706289957e-05, "loss": 0.0205, "step": 53390 }, { "grad_norm": 0.31385481357574463, "learning_rate": 4.851374959020344e-05, "loss": 0.0208, "step": 53400 }, { "grad_norm": 0.2611134648323059, "learning_rate": 4.849722228004141e-05, "loss": 0.0264, "step": 53410 }, { "grad_norm": 0.27294301986694336, "learning_rate": 4.848069513422091e-05, "loss": 0.0232, "step": 53420 }, { "grad_norm": 0.28658875823020935, "learning_rate": 4.8464168154549294e-05, "loss": 0.0212, "step": 53430 }, { "grad_norm": 0.4829787313938141, "learning_rate": 4.844764134283392e-05, "loss": 0.0281, "step": 53440 }, { "grad_norm": 0.2905789613723755, "learning_rate": 4.843111470088218e-05, "loss": 0.0265, "step": 53450 }, { "grad_norm": 0.22524389624595642, "learning_rate": 4.841458823050137e-05, "loss": 0.0246, "step": 53460 }, { "grad_norm": 0.34429270029067993, "learning_rate": 4.8398061933498816e-05, "loss": 0.029, "step": 53470 }, { "grad_norm": 0.3548995554447174, "learning_rate": 4.83815358116818e-05, "loss": 0.0257, "step": 53480 }, { "grad_norm": 0.2609153389930725, "learning_rate": 4.8365009866857584e-05, "loss": 0.0264, "step": 53490 }, { "grad_norm": 0.2555552124977112, "learning_rate": 4.834848410083342e-05, "loss": 0.0194, "step": 53500 }, { "grad_norm": 0.3331809937953949, "learning_rate": 4.833195851541657e-05, "loss": 0.0317, "step": 53510 }, { "grad_norm": 0.4954597055912018, "learning_rate": 4.831543311241423e-05, "loss": 0.0287, "step": 53520 }, { "grad_norm": 0.5193338394165039, "learning_rate": 4.829890789363359e-05, "loss": 0.0397, "step": 53530 }, { "grad_norm": 0.29948726296424866, "learning_rate": 4.828238286088182e-05, "loss": 0.0248, "step": 53540 }, { "grad_norm": 0.34200385212898254, "learning_rate": 4.826585801596609e-05, "loss": 0.0263, "step": 53550 }, { "grad_norm": 0.41258201003074646, "learning_rate": 4.8249333360693486e-05, "loss": 0.0232, "step": 53560 }, { "grad_norm": 0.31935715675354004, "learning_rate": 4.8232808896871185e-05, "loss": 0.0215, "step": 53570 }, { "grad_norm": 0.2657991647720337, "learning_rate": 4.821628462630624e-05, "loss": 0.0185, "step": 53580 }, { "grad_norm": 0.30094242095947266, "learning_rate": 4.8199760550805735e-05, "loss": 0.023, "step": 53590 }, { "grad_norm": 0.3382985293865204, "learning_rate": 4.818323667217669e-05, "loss": 0.0208, "step": 53600 }, { "grad_norm": 0.3452301323413849, "learning_rate": 4.8166712992226165e-05, "loss": 0.0217, "step": 53610 }, { "grad_norm": 0.4754352569580078, "learning_rate": 4.8150189512761154e-05, "loss": 0.0264, "step": 53620 }, { "grad_norm": 0.22937338054180145, "learning_rate": 4.813366623558862e-05, "loss": 0.0225, "step": 53630 }, { "grad_norm": 0.2832430899143219, "learning_rate": 4.811714316251554e-05, "loss": 0.0236, "step": 53640 }, { "grad_norm": 0.30890989303588867, "learning_rate": 4.810062029534886e-05, "loss": 0.0273, "step": 53650 }, { "grad_norm": 0.30198153853416443, "learning_rate": 4.8084097635895505e-05, "loss": 0.0334, "step": 53660 }, { "grad_norm": 0.30520114302635193, "learning_rate": 4.8067575185962335e-05, "loss": 0.0286, "step": 53670 }, { "grad_norm": 0.42172640562057495, "learning_rate": 4.805105294735625e-05, "loss": 0.0284, "step": 53680 }, { "grad_norm": 0.3944735825061798, "learning_rate": 4.803453092188406e-05, "loss": 0.0283, "step": 53690 }, { "grad_norm": 0.45397552847862244, "learning_rate": 4.801800911135263e-05, "loss": 0.027, "step": 53700 }, { "grad_norm": 0.3290870189666748, "learning_rate": 4.8001487517568744e-05, "loss": 0.0221, "step": 53710 }, { "grad_norm": 0.5084676742553711, "learning_rate": 4.798496614233918e-05, "loss": 0.0362, "step": 53720 }, { "grad_norm": 0.22541463375091553, "learning_rate": 4.796844498747069e-05, "loss": 0.0293, "step": 53730 }, { "grad_norm": 0.4318479299545288, "learning_rate": 4.795192405477e-05, "loss": 0.0268, "step": 53740 }, { "grad_norm": 0.34032493829727173, "learning_rate": 4.79354033460438e-05, "loss": 0.0293, "step": 53750 }, { "grad_norm": 0.9351160526275635, "learning_rate": 4.79188828630988e-05, "loss": 0.035, "step": 53760 }, { "grad_norm": 0.36336663365364075, "learning_rate": 4.790236260774165e-05, "loss": 0.022, "step": 53770 }, { "grad_norm": 0.3263956308364868, "learning_rate": 4.788584258177896e-05, "loss": 0.0268, "step": 53780 }, { "grad_norm": 0.3567354083061218, "learning_rate": 4.7869322787017356e-05, "loss": 0.0254, "step": 53790 }, { "grad_norm": 0.2498769462108612, "learning_rate": 4.78528032252634e-05, "loss": 0.0244, "step": 53800 }, { "grad_norm": 0.30184388160705566, "learning_rate": 4.783628389832366e-05, "loss": 0.0228, "step": 53810 }, { "grad_norm": 0.36060822010040283, "learning_rate": 4.7819764808004644e-05, "loss": 0.0235, "step": 53820 }, { "grad_norm": 0.42839691042900085, "learning_rate": 4.780324595611289e-05, "loss": 0.0272, "step": 53830 }, { "grad_norm": 0.3038712441921234, "learning_rate": 4.7786727344454854e-05, "loss": 0.0297, "step": 53840 }, { "grad_norm": 0.24048134684562683, "learning_rate": 4.777020897483697e-05, "loss": 0.0274, "step": 53850 }, { "grad_norm": 0.2416166365146637, "learning_rate": 4.775369084906569e-05, "loss": 0.0212, "step": 53860 }, { "grad_norm": 0.26767873764038086, "learning_rate": 4.773717296894742e-05, "loss": 0.027, "step": 53870 }, { "grad_norm": 0.2541724741458893, "learning_rate": 4.772065533628848e-05, "loss": 0.0252, "step": 53880 }, { "grad_norm": 0.3687029480934143, "learning_rate": 4.7704137952895265e-05, "loss": 0.0205, "step": 53890 }, { "grad_norm": 0.35440850257873535, "learning_rate": 4.768762082057407e-05, "loss": 0.028, "step": 53900 }, { "grad_norm": 0.409635066986084, "learning_rate": 4.7671103941131186e-05, "loss": 0.0244, "step": 53910 }, { "grad_norm": 0.34402355551719666, "learning_rate": 4.765458731637287e-05, "loss": 0.0245, "step": 53920 }, { "grad_norm": 0.33825573325157166, "learning_rate": 4.763807094810537e-05, "loss": 0.0296, "step": 53930 }, { "grad_norm": 0.2847020924091339, "learning_rate": 4.762155483813487e-05, "loss": 0.0287, "step": 53940 }, { "grad_norm": 0.5333057641983032, "learning_rate": 4.760503898826754e-05, "loss": 0.0303, "step": 53950 }, { "grad_norm": 0.3631042242050171, "learning_rate": 4.758852340030957e-05, "loss": 0.0321, "step": 53960 }, { "grad_norm": 0.35766202211380005, "learning_rate": 4.757200807606705e-05, "loss": 0.0277, "step": 53970 }, { "grad_norm": 0.3293415606021881, "learning_rate": 4.755549301734607e-05, "loss": 0.0251, "step": 53980 }, { "grad_norm": 0.26181674003601074, "learning_rate": 4.753897822595271e-05, "loss": 0.0364, "step": 53990 }, { "grad_norm": 0.3368326425552368, "learning_rate": 4.7522463703692994e-05, "loss": 0.0248, "step": 54000 }, { "grad_norm": 0.3816002905368805, "learning_rate": 4.75059494523729e-05, "loss": 0.0277, "step": 54010 }, { "grad_norm": 0.3643394410610199, "learning_rate": 4.748943547379843e-05, "loss": 0.0236, "step": 54020 }, { "grad_norm": 0.30030563473701477, "learning_rate": 4.7472921769775535e-05, "loss": 0.0285, "step": 54030 }, { "grad_norm": 0.4242136478424072, "learning_rate": 4.745640834211011e-05, "loss": 0.0307, "step": 54040 }, { "grad_norm": 0.31411945819854736, "learning_rate": 4.7439895192608054e-05, "loss": 0.0243, "step": 54050 }, { "grad_norm": 0.4001028537750244, "learning_rate": 4.7423382323075186e-05, "loss": 0.0252, "step": 54060 }, { "grad_norm": 0.27003657817840576, "learning_rate": 4.740686973531737e-05, "loss": 0.0175, "step": 54070 }, { "grad_norm": 0.43187659978866577, "learning_rate": 4.7390357431140375e-05, "loss": 0.0253, "step": 54080 }, { "grad_norm": 0.3417266309261322, "learning_rate": 4.737384541234998e-05, "loss": 0.0252, "step": 54090 }, { "grad_norm": 0.34904566407203674, "learning_rate": 4.735733368075189e-05, "loss": 0.0176, "step": 54100 }, { "grad_norm": 0.251997172832489, "learning_rate": 4.734082223815182e-05, "loss": 0.0252, "step": 54110 }, { "grad_norm": 0.3245823085308075, "learning_rate": 4.7324311086355436e-05, "loss": 0.018, "step": 54120 }, { "grad_norm": 0.46992239356040955, "learning_rate": 4.730780022716836e-05, "loss": 0.0235, "step": 54130 }, { "grad_norm": 0.40247642993927, "learning_rate": 4.729128966239619e-05, "loss": 0.0203, "step": 54140 }, { "grad_norm": 0.6919695138931274, "learning_rate": 4.7274779393844536e-05, "loss": 0.0252, "step": 54150 }, { "grad_norm": 0.46172192692756653, "learning_rate": 4.725826942331891e-05, "loss": 0.0398, "step": 54160 }, { "grad_norm": 0.23183563351631165, "learning_rate": 4.724175975262481e-05, "loss": 0.0212, "step": 54170 }, { "grad_norm": 0.37650126218795776, "learning_rate": 4.722525038356772e-05, "loss": 0.0229, "step": 54180 }, { "grad_norm": 0.34620437026023865, "learning_rate": 4.7208741317953075e-05, "loss": 0.0219, "step": 54190 }, { "grad_norm": 0.2907583713531494, "learning_rate": 4.7192232557586264e-05, "loss": 0.0338, "step": 54200 }, { "grad_norm": 0.49300259351730347, "learning_rate": 4.717572410427269e-05, "loss": 0.0266, "step": 54210 }, { "grad_norm": 0.3291338086128235, "learning_rate": 4.71592159598177e-05, "loss": 0.0314, "step": 54220 }, { "grad_norm": 0.4723157584667206, "learning_rate": 4.714270812602657e-05, "loss": 0.0294, "step": 54230 }, { "grad_norm": 0.3116743266582489, "learning_rate": 4.712620060470458e-05, "loss": 0.0295, "step": 54240 }, { "grad_norm": 0.23590071499347687, "learning_rate": 4.7109693397656986e-05, "loss": 0.0213, "step": 54250 }, { "grad_norm": 0.2309776097536087, "learning_rate": 4.709318650668894e-05, "loss": 0.0186, "step": 54260 }, { "grad_norm": 0.22152532637119293, "learning_rate": 4.7076679933605676e-05, "loss": 0.0206, "step": 54270 }, { "grad_norm": 0.28480064868927, "learning_rate": 4.70601736802123e-05, "loss": 0.0213, "step": 54280 }, { "grad_norm": 0.3218356966972351, "learning_rate": 4.704366774831391e-05, "loss": 0.0422, "step": 54290 }, { "grad_norm": 0.30291083455085754, "learning_rate": 4.702716213971557e-05, "loss": 0.0263, "step": 54300 }, { "grad_norm": 0.41449350118637085, "learning_rate": 4.70106568562223e-05, "loss": 0.0247, "step": 54310 }, { "grad_norm": 0.2666563391685486, "learning_rate": 4.6994151899639105e-05, "loss": 0.0252, "step": 54320 }, { "grad_norm": 0.2019711583852768, "learning_rate": 4.697764727177093e-05, "loss": 0.0276, "step": 54330 }, { "grad_norm": 0.419893354177475, "learning_rate": 4.696114297442272e-05, "loss": 0.0216, "step": 54340 }, { "grad_norm": 0.20062649250030518, "learning_rate": 4.694463900939933e-05, "loss": 0.0246, "step": 54350 }, { "grad_norm": 0.2600976824760437, "learning_rate": 4.692813537850564e-05, "loss": 0.0235, "step": 54360 }, { "grad_norm": 0.2913679778575897, "learning_rate": 4.6911632083546445e-05, "loss": 0.0165, "step": 54370 }, { "grad_norm": 0.3516009449958801, "learning_rate": 4.6895129126326525e-05, "loss": 0.03, "step": 54380 }, { "grad_norm": 0.3695443570613861, "learning_rate": 4.687862650865059e-05, "loss": 0.023, "step": 54390 }, { "grad_norm": 0.3052331507205963, "learning_rate": 4.686212423232339e-05, "loss": 0.0252, "step": 54400 }, { "grad_norm": 0.3021526634693146, "learning_rate": 4.6845622299149565e-05, "loss": 0.0186, "step": 54410 }, { "grad_norm": 0.3170534670352936, "learning_rate": 4.682912071093374e-05, "loss": 0.0266, "step": 54420 }, { "grad_norm": 0.22385042905807495, "learning_rate": 4.681261946948052e-05, "loss": 0.019, "step": 54430 }, { "grad_norm": 0.2537532448768616, "learning_rate": 4.679611857659443e-05, "loss": 0.0274, "step": 54440 }, { "grad_norm": 0.33790600299835205, "learning_rate": 4.6779618034079975e-05, "loss": 0.0242, "step": 54450 }, { "grad_norm": 0.21012842655181885, "learning_rate": 4.6763117843741675e-05, "loss": 0.0316, "step": 54460 }, { "grad_norm": 0.360489159822464, "learning_rate": 4.674661800738394e-05, "loss": 0.0203, "step": 54470 }, { "grad_norm": 0.43104395270347595, "learning_rate": 4.673011852681117e-05, "loss": 0.0244, "step": 54480 }, { "grad_norm": 0.26322537660598755, "learning_rate": 4.671361940382772e-05, "loss": 0.0254, "step": 54490 }, { "grad_norm": 0.26666948199272156, "learning_rate": 4.669712064023791e-05, "loss": 0.0289, "step": 54500 }, { "grad_norm": 0.3329467177391052, "learning_rate": 4.668062223784602e-05, "loss": 0.0206, "step": 54510 }, { "grad_norm": 0.37366411089897156, "learning_rate": 4.6664124198456274e-05, "loss": 0.0231, "step": 54520 }, { "grad_norm": 0.2383785992860794, "learning_rate": 4.6647626523872905e-05, "loss": 0.0282, "step": 54530 }, { "grad_norm": 0.27648717164993286, "learning_rate": 4.6631129215900066e-05, "loss": 0.029, "step": 54540 }, { "grad_norm": 0.24963800609111786, "learning_rate": 4.661463227634187e-05, "loss": 0.0384, "step": 54550 }, { "grad_norm": 0.25736886262893677, "learning_rate": 4.659813570700238e-05, "loss": 0.019, "step": 54560 }, { "grad_norm": 0.2674878239631653, "learning_rate": 4.658163950968568e-05, "loss": 0.0216, "step": 54570 }, { "grad_norm": 0.1899215430021286, "learning_rate": 4.6565143686195726e-05, "loss": 0.0238, "step": 54580 }, { "grad_norm": 0.4433162212371826, "learning_rate": 4.654864823833649e-05, "loss": 0.0252, "step": 54590 }, { "grad_norm": 0.2365773618221283, "learning_rate": 4.65321531679119e-05, "loss": 0.0246, "step": 54600 }, { "grad_norm": 0.3746635913848877, "learning_rate": 4.6515658476725834e-05, "loss": 0.0413, "step": 54610 }, { "grad_norm": 0.6373440027236938, "learning_rate": 4.649916416658212e-05, "loss": 0.0261, "step": 54620 }, { "grad_norm": 0.332673579454422, "learning_rate": 4.648267023928455e-05, "loss": 0.0227, "step": 54630 }, { "grad_norm": 0.3457983732223511, "learning_rate": 4.646617669663687e-05, "loss": 0.0236, "step": 54640 }, { "grad_norm": 0.4070361256599426, "learning_rate": 4.644968354044277e-05, "loss": 0.023, "step": 54650 }, { "grad_norm": 0.45081132650375366, "learning_rate": 4.643319077250596e-05, "loss": 0.0269, "step": 54660 }, { "grad_norm": 0.2300991714000702, "learning_rate": 4.6416698394630055e-05, "loss": 0.0204, "step": 54670 }, { "grad_norm": 0.43585461378097534, "learning_rate": 4.640020640861862e-05, "loss": 0.0239, "step": 54680 }, { "grad_norm": 0.7669989466667175, "learning_rate": 4.6383714816275195e-05, "loss": 0.0266, "step": 54690 }, { "grad_norm": 0.35457754135131836, "learning_rate": 4.636722361940329e-05, "loss": 0.0231, "step": 54700 }, { "grad_norm": 0.512930691242218, "learning_rate": 4.635073281980632e-05, "loss": 0.0261, "step": 54710 }, { "grad_norm": 0.4095747470855713, "learning_rate": 4.633424241928775e-05, "loss": 0.0281, "step": 54720 }, { "grad_norm": 0.33795109391212463, "learning_rate": 4.6317752419650914e-05, "loss": 0.0192, "step": 54730 }, { "grad_norm": 0.38075169920921326, "learning_rate": 4.630126282269914e-05, "loss": 0.0278, "step": 54740 }, { "grad_norm": 0.3479612171649933, "learning_rate": 4.628477363023569e-05, "loss": 0.0214, "step": 54750 }, { "grad_norm": 0.39252498745918274, "learning_rate": 4.626828484406383e-05, "loss": 0.0258, "step": 54760 }, { "grad_norm": 0.4671934247016907, "learning_rate": 4.625179646598669e-05, "loss": 0.0269, "step": 54770 }, { "grad_norm": 0.44791582226753235, "learning_rate": 4.6235308497807487e-05, "loss": 0.027, "step": 54780 }, { "grad_norm": 0.3607885539531708, "learning_rate": 4.6218820941329276e-05, "loss": 0.0274, "step": 54790 }, { "grad_norm": 0.2842702865600586, "learning_rate": 4.620233379835513e-05, "loss": 0.0205, "step": 54800 }, { "grad_norm": 0.42385783791542053, "learning_rate": 4.618584707068804e-05, "loss": 0.0233, "step": 54810 }, { "grad_norm": 0.8071926236152649, "learning_rate": 4.616936076013099e-05, "loss": 0.0271, "step": 54820 }, { "grad_norm": 0.37913429737091064, "learning_rate": 4.615287486848688e-05, "loss": 0.0281, "step": 54830 }, { "grad_norm": 0.3631656765937805, "learning_rate": 4.6136389397558584e-05, "loss": 0.0224, "step": 54840 }, { "grad_norm": 0.31025150418281555, "learning_rate": 4.6119904349148935e-05, "loss": 0.0193, "step": 54850 }, { "grad_norm": 0.7773411273956299, "learning_rate": 4.6103419725060726e-05, "loss": 0.0276, "step": 54860 }, { "grad_norm": 0.3335880935192108, "learning_rate": 4.608693552709667e-05, "loss": 0.0187, "step": 54870 }, { "grad_norm": 0.38346442580223083, "learning_rate": 4.607045175705947e-05, "loss": 0.0239, "step": 54880 }, { "grad_norm": 0.3996392488479614, "learning_rate": 4.605396841675176e-05, "loss": 0.0255, "step": 54890 }, { "grad_norm": 0.3658682107925415, "learning_rate": 4.6037485507976094e-05, "loss": 0.0262, "step": 54900 }, { "grad_norm": 0.35092392563819885, "learning_rate": 4.602100303253509e-05, "loss": 0.0219, "step": 54910 }, { "grad_norm": 0.36424392461776733, "learning_rate": 4.600452099223122e-05, "loss": 0.0325, "step": 54920 }, { "grad_norm": 0.3359244763851166, "learning_rate": 4.598803938886691e-05, "loss": 0.0243, "step": 54930 }, { "grad_norm": 0.3070162236690521, "learning_rate": 4.597155822424459e-05, "loss": 0.0216, "step": 54940 }, { "grad_norm": 0.3626038730144501, "learning_rate": 4.59550775001666e-05, "loss": 0.0267, "step": 54950 }, { "grad_norm": 0.3365055322647095, "learning_rate": 4.5938597218435234e-05, "loss": 0.0296, "step": 54960 }, { "grad_norm": 0.3394750952720642, "learning_rate": 4.592211738085279e-05, "loss": 0.0313, "step": 54970 }, { "grad_norm": 0.3106675446033478, "learning_rate": 4.590563798922146e-05, "loss": 0.0207, "step": 54980 }, { "grad_norm": 0.288106769323349, "learning_rate": 4.5889159045343404e-05, "loss": 0.0252, "step": 54990 }, { "grad_norm": 0.3198997378349304, "learning_rate": 4.5872680551020734e-05, "loss": 0.0221, "step": 55000 }, { "grad_norm": 0.6106293797492981, "learning_rate": 4.585620250805552e-05, "loss": 0.0237, "step": 55010 }, { "grad_norm": 0.31331175565719604, "learning_rate": 4.583972491824975e-05, "loss": 0.0209, "step": 55020 }, { "grad_norm": 0.23819532990455627, "learning_rate": 4.582324778340541e-05, "loss": 0.0243, "step": 55030 }, { "grad_norm": 0.33142533898353577, "learning_rate": 4.5806771105324424e-05, "loss": 0.0245, "step": 55040 }, { "grad_norm": 0.3024175763130188, "learning_rate": 4.579029488580865e-05, "loss": 0.0233, "step": 55050 }, { "grad_norm": 0.28962427377700806, "learning_rate": 4.5773819126659876e-05, "loss": 0.0196, "step": 55060 }, { "grad_norm": 0.2822283208370209, "learning_rate": 4.575734382967991e-05, "loss": 0.0224, "step": 55070 }, { "grad_norm": 0.31967225670814514, "learning_rate": 4.574086899667044e-05, "loss": 0.022, "step": 55080 }, { "grad_norm": 0.22823067009449005, "learning_rate": 4.5724394629433106e-05, "loss": 0.0207, "step": 55090 }, { "grad_norm": 0.29057255387306213, "learning_rate": 4.570792072976958e-05, "loss": 0.0303, "step": 55100 }, { "grad_norm": 0.3513332009315491, "learning_rate": 4.569144729948138e-05, "loss": 0.0258, "step": 55110 }, { "grad_norm": 0.3106125295162201, "learning_rate": 4.5674974340370033e-05, "loss": 0.0202, "step": 55120 }, { "grad_norm": 0.28034159541130066, "learning_rate": 4.565850185423699e-05, "loss": 0.0259, "step": 55130 }, { "grad_norm": 0.5894824266433716, "learning_rate": 4.564202984288365e-05, "loss": 0.0331, "step": 55140 }, { "grad_norm": 0.39445286989212036, "learning_rate": 4.5625558308111356e-05, "loss": 0.0276, "step": 55150 }, { "grad_norm": 0.23829925060272217, "learning_rate": 4.560908725172145e-05, "loss": 0.019, "step": 55160 }, { "grad_norm": 0.6376609206199646, "learning_rate": 4.559261667551516e-05, "loss": 0.0279, "step": 55170 }, { "grad_norm": 0.2592678666114807, "learning_rate": 4.5576146581293685e-05, "loss": 0.0294, "step": 55180 }, { "grad_norm": 0.18366408348083496, "learning_rate": 4.5559676970858164e-05, "loss": 0.02, "step": 55190 }, { "grad_norm": 0.4588509500026703, "learning_rate": 4.5543207846009705e-05, "loss": 0.0295, "step": 55200 }, { "grad_norm": 0.3498370051383972, "learning_rate": 4.552673920854933e-05, "loss": 0.0278, "step": 55210 }, { "grad_norm": 0.5445175170898438, "learning_rate": 4.5510271060278006e-05, "loss": 0.0266, "step": 55220 }, { "grad_norm": 0.4402787387371063, "learning_rate": 4.5493803402996704e-05, "loss": 0.0325, "step": 55230 }, { "grad_norm": 0.34456774592399597, "learning_rate": 4.5477336238506295e-05, "loss": 0.0227, "step": 55240 }, { "grad_norm": 0.3142775595188141, "learning_rate": 4.546086956860759e-05, "loss": 0.0227, "step": 55250 }, { "grad_norm": 0.30707788467407227, "learning_rate": 4.5444403395101374e-05, "loss": 0.021, "step": 55260 }, { "grad_norm": 0.34550178050994873, "learning_rate": 4.542793771978834e-05, "loss": 0.0306, "step": 55270 }, { "grad_norm": 0.4193084239959717, "learning_rate": 4.541147254446916e-05, "loss": 0.0312, "step": 55280 }, { "grad_norm": 0.24544520676136017, "learning_rate": 4.539500787094445e-05, "loss": 0.0289, "step": 55290 }, { "grad_norm": 0.30952250957489014, "learning_rate": 4.537854370101475e-05, "loss": 0.0235, "step": 55300 }, { "grad_norm": 0.2535974085330963, "learning_rate": 4.5362080036480566e-05, "loss": 0.0216, "step": 55310 }, { "grad_norm": 0.2094496488571167, "learning_rate": 4.534561687914235e-05, "loss": 0.0235, "step": 55320 }, { "grad_norm": 0.321267694234848, "learning_rate": 4.5329154230800466e-05, "loss": 0.0237, "step": 55330 }, { "grad_norm": 0.3263252377510071, "learning_rate": 4.531269209325525e-05, "loss": 0.0187, "step": 55340 }, { "grad_norm": 0.4102625846862793, "learning_rate": 4.529623046830696e-05, "loss": 0.0215, "step": 55350 }, { "grad_norm": 0.6234689950942993, "learning_rate": 4.5279769357755866e-05, "loss": 0.0286, "step": 55360 }, { "grad_norm": 0.3079240620136261, "learning_rate": 4.5263308763402084e-05, "loss": 0.0161, "step": 55370 }, { "grad_norm": 0.22463269531726837, "learning_rate": 4.524684868704574e-05, "loss": 0.0201, "step": 55380 }, { "grad_norm": 0.5420078635215759, "learning_rate": 4.523038913048688e-05, "loss": 0.0242, "step": 55390 }, { "grad_norm": 0.3474934697151184, "learning_rate": 4.521393009552549e-05, "loss": 0.0167, "step": 55400 }, { "grad_norm": 0.29195669293403625, "learning_rate": 4.5197471583961484e-05, "loss": 0.0292, "step": 55410 }, { "grad_norm": 0.21673132479190826, "learning_rate": 4.51810135975948e-05, "loss": 0.0282, "step": 55420 }, { "grad_norm": 0.2597915232181549, "learning_rate": 4.516455613822521e-05, "loss": 0.0247, "step": 55430 }, { "grad_norm": 0.2931508421897888, "learning_rate": 4.514809920765249e-05, "loss": 0.0237, "step": 55440 }, { "grad_norm": 0.448467493057251, "learning_rate": 4.5131642807676356e-05, "loss": 0.0267, "step": 55450 }, { "grad_norm": 0.2263343632221222, "learning_rate": 4.511518694009644e-05, "loss": 0.0228, "step": 55460 }, { "grad_norm": 0.28525587916374207, "learning_rate": 4.509873160671231e-05, "loss": 0.0225, "step": 55470 }, { "grad_norm": 0.34061092138290405, "learning_rate": 4.508227680932356e-05, "loss": 0.0242, "step": 55480 }, { "grad_norm": 0.44795534014701843, "learning_rate": 4.5065822549729605e-05, "loss": 0.031, "step": 55490 }, { "grad_norm": 0.49105456471443176, "learning_rate": 4.504936882972989e-05, "loss": 0.0208, "step": 55500 }, { "grad_norm": 0.43837836384773254, "learning_rate": 4.503291565112374e-05, "loss": 0.0216, "step": 55510 }, { "grad_norm": 0.33555036783218384, "learning_rate": 4.501646301571048e-05, "loss": 0.0254, "step": 55520 }, { "grad_norm": 0.2512364387512207, "learning_rate": 4.500001092528933e-05, "loss": 0.0262, "step": 55530 }, { "grad_norm": 0.3709087073802948, "learning_rate": 4.4983559381659455e-05, "loss": 0.0247, "step": 55540 }, { "grad_norm": 0.44659224152565, "learning_rate": 4.496710838661999e-05, "loss": 0.0203, "step": 55550 }, { "grad_norm": 0.2693752944469452, "learning_rate": 4.495065794196999e-05, "loss": 0.0272, "step": 55560 }, { "grad_norm": 0.1941225677728653, "learning_rate": 4.4934208049508455e-05, "loss": 0.0226, "step": 55570 }, { "grad_norm": 0.4370781481266022, "learning_rate": 4.4917758711034303e-05, "loss": 0.0294, "step": 55580 }, { "grad_norm": 0.27594196796417236, "learning_rate": 4.490130992834642e-05, "loss": 0.0223, "step": 55590 }, { "grad_norm": 0.36537623405456543, "learning_rate": 4.488486170324359e-05, "loss": 0.0337, "step": 55600 }, { "grad_norm": 0.3480689823627472, "learning_rate": 4.486841403752463e-05, "loss": 0.0257, "step": 55610 }, { "grad_norm": 0.3534529209136963, "learning_rate": 4.4851966932988174e-05, "loss": 0.022, "step": 55620 }, { "grad_norm": 0.33127090334892273, "learning_rate": 4.4835520391432894e-05, "loss": 0.0269, "step": 55630 }, { "grad_norm": 0.3460168242454529, "learning_rate": 4.481907441465733e-05, "loss": 0.0223, "step": 55640 }, { "grad_norm": 0.4479968547821045, "learning_rate": 4.480262900446e-05, "loss": 0.0329, "step": 55650 }, { "grad_norm": 0.3622623682022095, "learning_rate": 4.478618416263933e-05, "loss": 0.0232, "step": 55660 }, { "grad_norm": 0.39522838592529297, "learning_rate": 4.476973989099373e-05, "loss": 0.0232, "step": 55670 }, { "grad_norm": 0.27362245321273804, "learning_rate": 4.475329619132153e-05, "loss": 0.0287, "step": 55680 }, { "grad_norm": 0.2910440266132355, "learning_rate": 4.4736853065420956e-05, "loss": 0.0271, "step": 55690 }, { "grad_norm": 0.2487407922744751, "learning_rate": 4.472041051509023e-05, "loss": 0.0246, "step": 55700 }, { "grad_norm": 0.2979852259159088, "learning_rate": 4.470396854212746e-05, "loss": 0.0229, "step": 55710 }, { "grad_norm": 0.38206249475479126, "learning_rate": 4.468752714833072e-05, "loss": 0.0194, "step": 55720 }, { "grad_norm": 0.32475873827934265, "learning_rate": 4.467108633549803e-05, "loss": 0.0201, "step": 55730 }, { "grad_norm": 0.1977243572473526, "learning_rate": 4.4654646105427335e-05, "loss": 0.0277, "step": 55740 }, { "grad_norm": 0.3874416947364807, "learning_rate": 4.463820645991651e-05, "loss": 0.0217, "step": 55750 }, { "grad_norm": 0.2899089455604553, "learning_rate": 4.462176740076334e-05, "loss": 0.0182, "step": 55760 }, { "grad_norm": 0.28843241930007935, "learning_rate": 4.4605328929765624e-05, "loss": 0.0312, "step": 55770 }, { "grad_norm": 0.30925703048706055, "learning_rate": 4.458889104872102e-05, "loss": 0.0228, "step": 55780 }, { "grad_norm": 0.268537312746048, "learning_rate": 4.457245375942715e-05, "loss": 0.0329, "step": 55790 }, { "grad_norm": 0.4111349284648895, "learning_rate": 4.455601706368157e-05, "loss": 0.0279, "step": 55800 }, { "grad_norm": 0.3892782926559448, "learning_rate": 4.453958096328179e-05, "loss": 0.0217, "step": 55810 }, { "grad_norm": 0.3166223466396332, "learning_rate": 4.452314546002523e-05, "loss": 0.0237, "step": 55820 }, { "grad_norm": 0.3441419005393982, "learning_rate": 4.4506710555709246e-05, "loss": 0.0214, "step": 55830 }, { "grad_norm": 0.2810119688510895, "learning_rate": 4.449027625213114e-05, "loss": 0.026, "step": 55840 }, { "grad_norm": 0.2595841586589813, "learning_rate": 4.4473842551088135e-05, "loss": 0.0229, "step": 55850 }, { "grad_norm": 0.328118234872818, "learning_rate": 4.445740945437738e-05, "loss": 0.0228, "step": 55860 }, { "grad_norm": 0.271676629781723, "learning_rate": 4.444097696379602e-05, "loss": 0.0259, "step": 55870 }, { "grad_norm": 0.24952302873134613, "learning_rate": 4.442454508114106e-05, "loss": 0.023, "step": 55880 }, { "grad_norm": 0.40599265694618225, "learning_rate": 4.4408113808209465e-05, "loss": 0.0219, "step": 55890 }, { "grad_norm": 0.17422863841056824, "learning_rate": 4.439168314679813e-05, "loss": 0.0234, "step": 55900 }, { "grad_norm": 0.2166401445865631, "learning_rate": 4.43752530987039e-05, "loss": 0.0248, "step": 55910 }, { "grad_norm": 0.25174903869628906, "learning_rate": 4.4358823665723515e-05, "loss": 0.0174, "step": 55920 }, { "grad_norm": 0.3375873863697052, "learning_rate": 4.434239484965371e-05, "loss": 0.027, "step": 55930 }, { "grad_norm": 0.5674944519996643, "learning_rate": 4.4325966652291103e-05, "loss": 0.0209, "step": 55940 }, { "grad_norm": 0.26078566908836365, "learning_rate": 4.430953907543225e-05, "loss": 0.0213, "step": 55950 }, { "grad_norm": 0.35363340377807617, "learning_rate": 4.4293112120873645e-05, "loss": 0.0228, "step": 55960 }, { "grad_norm": 0.3064393997192383, "learning_rate": 4.427668579041172e-05, "loss": 0.0237, "step": 55970 }, { "grad_norm": 0.48150795698165894, "learning_rate": 4.4260260085842816e-05, "loss": 0.0246, "step": 55980 }, { "grad_norm": 0.2511182427406311, "learning_rate": 4.4243835008963256e-05, "loss": 0.0211, "step": 55990 }, { "grad_norm": 0.2524670362472534, "learning_rate": 4.422741056156925e-05, "loss": 0.019, "step": 56000 }, { "grad_norm": 0.27728071808815, "learning_rate": 4.421098674545693e-05, "loss": 0.0235, "step": 56010 }, { "grad_norm": 0.4001034200191498, "learning_rate": 4.419456356242241e-05, "loss": 0.022, "step": 56020 }, { "grad_norm": 0.2602594196796417, "learning_rate": 4.4178141014261685e-05, "loss": 0.0201, "step": 56030 }, { "grad_norm": 0.21511659026145935, "learning_rate": 4.416171910277071e-05, "loss": 0.0217, "step": 56040 }, { "grad_norm": 0.2789776027202606, "learning_rate": 4.414529782974533e-05, "loss": 0.0194, "step": 56050 }, { "grad_norm": 1.2760206460952759, "learning_rate": 4.41288771969814e-05, "loss": 0.0296, "step": 56060 }, { "grad_norm": 0.5070249438285828, "learning_rate": 4.411245720627464e-05, "loss": 0.0288, "step": 56070 }, { "grad_norm": 0.22569498419761658, "learning_rate": 4.40960378594207e-05, "loss": 0.0271, "step": 56080 }, { "grad_norm": 0.22113198041915894, "learning_rate": 4.407961915821519e-05, "loss": 0.0224, "step": 56090 }, { "grad_norm": 0.5586143732070923, "learning_rate": 4.406320110445362e-05, "loss": 0.0245, "step": 56100 }, { "grad_norm": 0.2322554886341095, "learning_rate": 4.404678369993144e-05, "loss": 0.0227, "step": 56110 }, { "grad_norm": 0.2589511573314667, "learning_rate": 4.403036694644406e-05, "loss": 0.021, "step": 56120 }, { "grad_norm": 0.43626195192337036, "learning_rate": 4.4013950845786764e-05, "loss": 0.0222, "step": 56130 }, { "grad_norm": 0.37863218784332275, "learning_rate": 4.399753539975482e-05, "loss": 0.0242, "step": 56140 }, { "grad_norm": 0.3365747034549713, "learning_rate": 4.398112061014337e-05, "loss": 0.0191, "step": 56150 }, { "grad_norm": 0.4377533793449402, "learning_rate": 4.396470647874753e-05, "loss": 0.0223, "step": 56160 }, { "grad_norm": 0.35801413655281067, "learning_rate": 4.394829300736229e-05, "loss": 0.0191, "step": 56170 }, { "grad_norm": 0.2975959777832031, "learning_rate": 4.393188019778265e-05, "loss": 0.0214, "step": 56180 }, { "grad_norm": 0.3549628257751465, "learning_rate": 4.391546805180347e-05, "loss": 0.0212, "step": 56190 }, { "grad_norm": 0.2640584707260132, "learning_rate": 4.389905657121955e-05, "loss": 0.0203, "step": 56200 }, { "grad_norm": 0.37843313813209534, "learning_rate": 4.3882645757825635e-05, "loss": 0.0261, "step": 56210 }, { "grad_norm": 0.35354137420654297, "learning_rate": 4.386623561341637e-05, "loss": 0.0236, "step": 56220 }, { "grad_norm": 0.3064345121383667, "learning_rate": 4.384982613978637e-05, "loss": 0.0258, "step": 56230 }, { "grad_norm": 0.38205602765083313, "learning_rate": 4.383341733873012e-05, "loss": 0.0235, "step": 56240 }, { "grad_norm": 0.24376805126667023, "learning_rate": 4.381700921204209e-05, "loss": 0.0199, "step": 56250 }, { "grad_norm": 1.0342925786972046, "learning_rate": 4.380060176151663e-05, "loss": 0.0231, "step": 56260 }, { "grad_norm": 0.4346877634525299, "learning_rate": 4.378419498894805e-05, "loss": 0.0243, "step": 56270 }, { "grad_norm": 0.4166601002216339, "learning_rate": 4.376778889613056e-05, "loss": 0.0246, "step": 56280 }, { "grad_norm": 0.263690710067749, "learning_rate": 4.375138348485831e-05, "loss": 0.0236, "step": 56290 }, { "grad_norm": 0.3803103566169739, "learning_rate": 4.3734978756925346e-05, "loss": 0.0191, "step": 56300 }, { "grad_norm": 0.17462772130966187, "learning_rate": 4.3718574714125706e-05, "loss": 0.0197, "step": 56310 }, { "grad_norm": 0.37298598885536194, "learning_rate": 4.370217135825329e-05, "loss": 0.0252, "step": 56320 }, { "grad_norm": 0.21719954907894135, "learning_rate": 4.368576869110194e-05, "loss": 0.017, "step": 56330 }, { "grad_norm": 0.32863739132881165, "learning_rate": 4.366936671446544e-05, "loss": 0.0244, "step": 56340 }, { "grad_norm": 0.4507938027381897, "learning_rate": 4.3652965430137474e-05, "loss": 0.0283, "step": 56350 }, { "grad_norm": 0.25335437059402466, "learning_rate": 4.3636564839911646e-05, "loss": 0.0227, "step": 56360 }, { "grad_norm": 0.21175453066825867, "learning_rate": 4.3620164945581545e-05, "loss": 0.0154, "step": 56370 }, { "grad_norm": 0.25836774706840515, "learning_rate": 4.360376574894061e-05, "loss": 0.0228, "step": 56380 }, { "grad_norm": 0.23713096976280212, "learning_rate": 4.358736725178224e-05, "loss": 0.0198, "step": 56390 }, { "grad_norm": 0.35742267966270447, "learning_rate": 4.357096945589974e-05, "loss": 0.0282, "step": 56400 }, { "grad_norm": 0.44814038276672363, "learning_rate": 4.355457236308636e-05, "loss": 0.0226, "step": 56410 }, { "grad_norm": 0.28799277544021606, "learning_rate": 4.353817597513526e-05, "loss": 0.0209, "step": 56420 }, { "grad_norm": 0.23697148263454437, "learning_rate": 4.352178029383948e-05, "loss": 0.0178, "step": 56430 }, { "grad_norm": 0.295671671628952, "learning_rate": 4.3505385320992105e-05, "loss": 0.0198, "step": 56440 }, { "grad_norm": 0.3394761085510254, "learning_rate": 4.348899105838602e-05, "loss": 0.0214, "step": 56450 }, { "grad_norm": 0.42880070209503174, "learning_rate": 4.3472597507814087e-05, "loss": 0.0237, "step": 56460 }, { "grad_norm": 0.45488399267196655, "learning_rate": 4.3456204671069066e-05, "loss": 0.0277, "step": 56470 }, { "grad_norm": 0.27181458473205566, "learning_rate": 4.343981254994367e-05, "loss": 0.0285, "step": 56480 }, { "grad_norm": 0.27818286418914795, "learning_rate": 4.34234211462305e-05, "loss": 0.0244, "step": 56490 }, { "grad_norm": 0.27408403158187866, "learning_rate": 4.340703046172213e-05, "loss": 0.0316, "step": 56500 }, { "grad_norm": 1.5183278322219849, "learning_rate": 4.339064049821097e-05, "loss": 0.0258, "step": 56510 }, { "grad_norm": 0.36620667576789856, "learning_rate": 4.3374251257489446e-05, "loss": 0.0268, "step": 56520 }, { "grad_norm": 0.24369370937347412, "learning_rate": 4.335786274134984e-05, "loss": 0.0237, "step": 56530 }, { "grad_norm": 0.5996931791305542, "learning_rate": 4.334147495158438e-05, "loss": 0.0316, "step": 56540 }, { "grad_norm": 0.42284640669822693, "learning_rate": 4.33250878899852e-05, "loss": 0.0228, "step": 56550 }, { "grad_norm": 0.5917117595672607, "learning_rate": 4.330870155834437e-05, "loss": 0.0204, "step": 56560 }, { "grad_norm": 0.3013404607772827, "learning_rate": 4.3292315958453884e-05, "loss": 0.022, "step": 56570 }, { "grad_norm": 0.2823442220687866, "learning_rate": 4.3275931092105646e-05, "loss": 0.0177, "step": 56580 }, { "grad_norm": 0.354874849319458, "learning_rate": 4.325954696109147e-05, "loss": 0.0212, "step": 56590 }, { "grad_norm": 0.309047669172287, "learning_rate": 4.32431635672031e-05, "loss": 0.0332, "step": 56600 }, { "grad_norm": 0.3283984661102295, "learning_rate": 4.3226780912232214e-05, "loss": 0.0362, "step": 56610 }, { "grad_norm": 0.4373037815093994, "learning_rate": 4.321039899797036e-05, "loss": 0.0214, "step": 56620 }, { "grad_norm": 0.2920253574848175, "learning_rate": 4.319401782620908e-05, "loss": 0.0176, "step": 56630 }, { "grad_norm": 0.3264593183994293, "learning_rate": 4.317763739873978e-05, "loss": 0.0253, "step": 56640 }, { "grad_norm": 0.476372629404068, "learning_rate": 4.31612577173538e-05, "loss": 0.0288, "step": 56650 }, { "grad_norm": 0.2949889898300171, "learning_rate": 4.314487878384239e-05, "loss": 0.032, "step": 56660 }, { "grad_norm": 0.5441391468048096, "learning_rate": 4.3128500599996736e-05, "loss": 0.0229, "step": 56670 }, { "grad_norm": 0.1844426542520523, "learning_rate": 4.31121231676079e-05, "loss": 0.0171, "step": 56680 }, { "grad_norm": 0.4782353937625885, "learning_rate": 4.309574648846694e-05, "loss": 0.029, "step": 56690 }, { "grad_norm": 0.27723246812820435, "learning_rate": 4.3079370564364755e-05, "loss": 0.0263, "step": 56700 }, { "grad_norm": 0.33299681544303894, "learning_rate": 4.306299539709221e-05, "loss": 0.0191, "step": 56710 }, { "grad_norm": 0.30043771862983704, "learning_rate": 4.3046620988440046e-05, "loss": 0.0215, "step": 56720 }, { "grad_norm": 0.3973105847835541, "learning_rate": 4.303024734019897e-05, "loss": 0.0207, "step": 56730 }, { "grad_norm": 0.24162080883979797, "learning_rate": 4.3013874454159565e-05, "loss": 0.0236, "step": 56740 }, { "grad_norm": 0.2923186719417572, "learning_rate": 4.299750233211233e-05, "loss": 0.0268, "step": 56750 }, { "grad_norm": 0.5084725022315979, "learning_rate": 4.2981130975847715e-05, "loss": 0.02, "step": 56760 }, { "grad_norm": 0.2796019911766052, "learning_rate": 4.2964760387156076e-05, "loss": 0.0215, "step": 56770 }, { "grad_norm": 0.2515234053134918, "learning_rate": 4.294839056782767e-05, "loss": 0.0199, "step": 56780 }, { "grad_norm": 0.2765173316001892, "learning_rate": 4.293202151965266e-05, "loss": 0.0224, "step": 56790 }, { "grad_norm": 0.2504526674747467, "learning_rate": 4.291565324442115e-05, "loss": 0.0201, "step": 56800 }, { "grad_norm": 0.3415834605693817, "learning_rate": 4.2899285743923135e-05, "loss": 0.021, "step": 56810 }, { "grad_norm": 0.32590752840042114, "learning_rate": 4.288291901994857e-05, "loss": 0.0187, "step": 56820 }, { "grad_norm": 0.3514823317527771, "learning_rate": 4.286655307428729e-05, "loss": 0.0254, "step": 56830 }, { "grad_norm": 0.3919496536254883, "learning_rate": 4.285018790872903e-05, "loss": 0.0272, "step": 56840 }, { "grad_norm": 0.4022397994995117, "learning_rate": 4.283382352506347e-05, "loss": 0.0227, "step": 56850 }, { "grad_norm": 0.34168919920921326, "learning_rate": 4.2817459925080205e-05, "loss": 0.0273, "step": 56860 }, { "grad_norm": 0.3380872905254364, "learning_rate": 4.28010971105687e-05, "loss": 0.0308, "step": 56870 }, { "grad_norm": 0.3337394893169403, "learning_rate": 4.278473508331841e-05, "loss": 0.0203, "step": 56880 }, { "grad_norm": 0.3908296525478363, "learning_rate": 4.276837384511864e-05, "loss": 0.0317, "step": 56890 }, { "grad_norm": 0.2880444824695587, "learning_rate": 4.275201339775864e-05, "loss": 0.0183, "step": 56900 }, { "grad_norm": 0.30323970317840576, "learning_rate": 4.273565374302756e-05, "loss": 0.0271, "step": 56910 }, { "grad_norm": 0.2723901569843292, "learning_rate": 4.2719294882714454e-05, "loss": 0.0177, "step": 56920 }, { "grad_norm": 0.4021115303039551, "learning_rate": 4.270293681860831e-05, "loss": 0.0232, "step": 56930 }, { "grad_norm": 0.19762659072875977, "learning_rate": 4.268657955249802e-05, "loss": 0.0172, "step": 56940 }, { "grad_norm": 0.30149149894714355, "learning_rate": 4.267022308617241e-05, "loss": 0.0196, "step": 56950 }, { "grad_norm": 0.3903123438358307, "learning_rate": 4.265386742142018e-05, "loss": 0.0189, "step": 56960 }, { "grad_norm": 0.3670007884502411, "learning_rate": 4.263751256002996e-05, "loss": 0.0266, "step": 56970 }, { "grad_norm": 0.2608427107334137, "learning_rate": 4.26211585037903e-05, "loss": 0.0238, "step": 56980 }, { "grad_norm": 0.36121681332588196, "learning_rate": 4.2604805254489656e-05, "loss": 0.0246, "step": 56990 }, { "grad_norm": 0.33757585287094116, "learning_rate": 4.258845281391638e-05, "loss": 0.0207, "step": 57000 }, { "grad_norm": 0.26725345849990845, "learning_rate": 4.257210118385876e-05, "loss": 0.0326, "step": 57010 }, { "grad_norm": 0.3276941478252411, "learning_rate": 4.2555750366104994e-05, "loss": 0.0212, "step": 57020 }, { "grad_norm": 0.1825840026140213, "learning_rate": 4.2539400362443184e-05, "loss": 0.0148, "step": 57030 }, { "grad_norm": 0.3632349967956543, "learning_rate": 4.252305117466133e-05, "loss": 0.022, "step": 57040 }, { "grad_norm": 0.26490211486816406, "learning_rate": 4.2506702804547356e-05, "loss": 0.0205, "step": 57050 }, { "grad_norm": 0.23085573315620422, "learning_rate": 4.249035525388907e-05, "loss": 0.0246, "step": 57060 }, { "grad_norm": 0.3628125488758087, "learning_rate": 4.247400852447427e-05, "loss": 0.0198, "step": 57070 }, { "grad_norm": 0.22037838399410248, "learning_rate": 4.245766261809059e-05, "loss": 0.0173, "step": 57080 }, { "grad_norm": 0.24154968559741974, "learning_rate": 4.244131753652558e-05, "loss": 0.0249, "step": 57090 }, { "grad_norm": 0.29302695393562317, "learning_rate": 4.242497328156672e-05, "loss": 0.0187, "step": 57100 }, { "grad_norm": 0.29357272386550903, "learning_rate": 4.240862985500139e-05, "loss": 0.0253, "step": 57110 }, { "grad_norm": 0.2743384838104248, "learning_rate": 4.2392287258616894e-05, "loss": 0.0152, "step": 57120 }, { "grad_norm": 1.1683015823364258, "learning_rate": 4.2375945494200396e-05, "loss": 0.0177, "step": 57130 }, { "grad_norm": 0.3519916236400604, "learning_rate": 4.2359604563539054e-05, "loss": 0.0277, "step": 57140 }, { "grad_norm": 0.36631011962890625, "learning_rate": 4.2343264468419885e-05, "loss": 0.0286, "step": 57150 }, { "grad_norm": 0.26721715927124023, "learning_rate": 4.232692521062979e-05, "loss": 0.0215, "step": 57160 }, { "grad_norm": 0.3773793578147888, "learning_rate": 4.231058679195562e-05, "loss": 0.0218, "step": 57170 }, { "grad_norm": 0.1891297996044159, "learning_rate": 4.229424921418411e-05, "loss": 0.0193, "step": 57180 }, { "grad_norm": 0.37423408031463623, "learning_rate": 4.2277912479101916e-05, "loss": 0.0292, "step": 57190 }, { "grad_norm": 0.2169685959815979, "learning_rate": 4.226157658849562e-05, "loss": 0.0189, "step": 57200 }, { "grad_norm": 0.36946308612823486, "learning_rate": 4.2245241544151664e-05, "loss": 0.0237, "step": 57210 }, { "grad_norm": 0.2342456877231598, "learning_rate": 4.2228907347856425e-05, "loss": 0.0194, "step": 57220 }, { "grad_norm": 0.3327316641807556, "learning_rate": 4.2212574001396214e-05, "loss": 0.0214, "step": 57230 }, { "grad_norm": 0.3700745403766632, "learning_rate": 4.21962415065572e-05, "loss": 0.0236, "step": 57240 }, { "grad_norm": 0.22972555458545685, "learning_rate": 4.217990986512548e-05, "loss": 0.0218, "step": 57250 }, { "grad_norm": 0.25558188557624817, "learning_rate": 4.216357907888704e-05, "loss": 0.0343, "step": 57260 }, { "grad_norm": 0.33951205015182495, "learning_rate": 4.2147249149627824e-05, "loss": 0.0224, "step": 57270 }, { "grad_norm": 0.31914782524108887, "learning_rate": 4.213092007913364e-05, "loss": 0.0245, "step": 57280 }, { "grad_norm": 0.2762589752674103, "learning_rate": 4.2114591869190205e-05, "loss": 0.0243, "step": 57290 }, { "grad_norm": 0.24110381305217743, "learning_rate": 4.209826452158315e-05, "loss": 0.0284, "step": 57300 }, { "grad_norm": 0.3781615197658539, "learning_rate": 4.2081938038098e-05, "loss": 0.023, "step": 57310 }, { "grad_norm": 0.30877238512039185, "learning_rate": 4.2065612420520185e-05, "loss": 0.0262, "step": 57320 }, { "grad_norm": 0.2770221531391144, "learning_rate": 4.204928767063509e-05, "loss": 0.0191, "step": 57330 }, { "grad_norm": 0.39696699380874634, "learning_rate": 4.2032963790227944e-05, "loss": 0.0217, "step": 57340 }, { "grad_norm": 0.41925331950187683, "learning_rate": 4.2016640781083887e-05, "loss": 0.0213, "step": 57350 }, { "grad_norm": 0.24881352484226227, "learning_rate": 4.2000318644988004e-05, "loss": 0.0303, "step": 57360 }, { "grad_norm": 0.2763761878013611, "learning_rate": 4.198399738372524e-05, "loss": 0.0221, "step": 57370 }, { "grad_norm": 0.18927431106567383, "learning_rate": 4.196767699908045e-05, "loss": 0.0202, "step": 57380 }, { "grad_norm": 0.31454968452453613, "learning_rate": 4.195135749283845e-05, "loss": 0.0196, "step": 57390 }, { "grad_norm": 0.46866971254348755, "learning_rate": 4.1935038866783894e-05, "loss": 0.0245, "step": 57400 }, { "grad_norm": 0.2324599176645279, "learning_rate": 4.1918721122701355e-05, "loss": 0.0209, "step": 57410 }, { "grad_norm": 0.3489978611469269, "learning_rate": 4.190240426237533e-05, "loss": 0.0264, "step": 57420 }, { "grad_norm": 0.2761477828025818, "learning_rate": 4.188608828759019e-05, "loss": 0.0242, "step": 57430 }, { "grad_norm": 0.3573967516422272, "learning_rate": 4.1869773200130246e-05, "loss": 0.0328, "step": 57440 }, { "grad_norm": 0.4028514325618744, "learning_rate": 4.185345900177967e-05, "loss": 0.0355, "step": 57450 }, { "grad_norm": 0.45586463809013367, "learning_rate": 4.183714569432258e-05, "loss": 0.025, "step": 57460 }, { "grad_norm": 0.31297847628593445, "learning_rate": 4.182083327954295e-05, "loss": 0.0207, "step": 57470 }, { "grad_norm": 0.23809833824634552, "learning_rate": 4.1804521759224714e-05, "loss": 0.0237, "step": 57480 }, { "grad_norm": 0.22752094268798828, "learning_rate": 4.178821113515165e-05, "loss": 0.0256, "step": 57490 }, { "grad_norm": 0.24654865264892578, "learning_rate": 4.1771901409107476e-05, "loss": 0.0294, "step": 57500 }, { "grad_norm": 0.40267592668533325, "learning_rate": 4.175559258287577e-05, "loss": 0.0244, "step": 57510 }, { "grad_norm": 0.3087446689605713, "learning_rate": 4.173928465824008e-05, "loss": 0.0266, "step": 57520 }, { "grad_norm": 0.21531008183956146, "learning_rate": 4.17229776369838e-05, "loss": 0.0208, "step": 57530 }, { "grad_norm": 0.27429717779159546, "learning_rate": 4.170667152089025e-05, "loss": 0.0284, "step": 57540 }, { "grad_norm": 0.2556593418121338, "learning_rate": 4.1690366311742626e-05, "loss": 0.016, "step": 57550 }, { "grad_norm": 0.2660399377346039, "learning_rate": 4.1674062011324055e-05, "loss": 0.0156, "step": 57560 }, { "grad_norm": 0.24324429035186768, "learning_rate": 4.165775862141752e-05, "loss": 0.0188, "step": 57570 }, { "grad_norm": 0.4053959548473358, "learning_rate": 4.164145614380598e-05, "loss": 0.0205, "step": 57580 }, { "grad_norm": 0.4320553243160248, "learning_rate": 4.162515458027223e-05, "loss": 0.0225, "step": 57590 }, { "grad_norm": 0.2247440367937088, "learning_rate": 4.160885393259899e-05, "loss": 0.0239, "step": 57600 }, { "grad_norm": 0.2802678644657135, "learning_rate": 4.159255420256886e-05, "loss": 0.0211, "step": 57610 }, { "grad_norm": 0.3695315718650818, "learning_rate": 4.157625539196436e-05, "loss": 0.0221, "step": 57620 }, { "grad_norm": 0.21253587305545807, "learning_rate": 4.155995750256791e-05, "loss": 0.0275, "step": 57630 }, { "grad_norm": 0.25477150082588196, "learning_rate": 4.15436605361618e-05, "loss": 0.0218, "step": 57640 }, { "grad_norm": 0.2599729299545288, "learning_rate": 4.152736449452827e-05, "loss": 0.0287, "step": 57650 }, { "grad_norm": 0.5817946195602417, "learning_rate": 4.151106937944943e-05, "loss": 0.0202, "step": 57660 }, { "grad_norm": 0.17699049413204193, "learning_rate": 4.149477519270727e-05, "loss": 0.0191, "step": 57670 }, { "grad_norm": 0.25353556871414185, "learning_rate": 4.147848193608371e-05, "loss": 0.0186, "step": 57680 }, { "grad_norm": 0.3185930848121643, "learning_rate": 4.146218961136056e-05, "loss": 0.024, "step": 57690 }, { "grad_norm": 0.4376847445964813, "learning_rate": 4.14458982203195e-05, "loss": 0.0225, "step": 57700 }, { "grad_norm": 0.2541455924510956, "learning_rate": 4.1429607764742165e-05, "loss": 0.0235, "step": 57710 }, { "grad_norm": 0.2795118987560272, "learning_rate": 4.141331824641003e-05, "loss": 0.0299, "step": 57720 }, { "grad_norm": 0.27728497982025146, "learning_rate": 4.139702966710452e-05, "loss": 0.0167, "step": 57730 }, { "grad_norm": 0.27436718344688416, "learning_rate": 4.13807420286069e-05, "loss": 0.025, "step": 57740 }, { "grad_norm": 0.3347356915473938, "learning_rate": 4.136445533269837e-05, "loss": 0.0235, "step": 57750 }, { "grad_norm": 0.15982571244239807, "learning_rate": 4.134816958116003e-05, "loss": 0.0155, "step": 57760 }, { "grad_norm": 0.7173174023628235, "learning_rate": 4.133188477577283e-05, "loss": 0.0251, "step": 57770 }, { "grad_norm": 0.24564334750175476, "learning_rate": 4.131560091831771e-05, "loss": 0.0216, "step": 57780 }, { "grad_norm": 0.34834176301956177, "learning_rate": 4.1299318010575405e-05, "loss": 0.0189, "step": 57790 }, { "grad_norm": 0.2604215145111084, "learning_rate": 4.128303605432661e-05, "loss": 0.0435, "step": 57800 }, { "grad_norm": 0.3958019018173218, "learning_rate": 4.126675505135188e-05, "loss": 0.0229, "step": 57810 }, { "grad_norm": 0.3149384558200836, "learning_rate": 4.125047500343169e-05, "loss": 0.0229, "step": 57820 }, { "grad_norm": 0.3874509334564209, "learning_rate": 4.1234195912346376e-05, "loss": 0.0235, "step": 57830 }, { "grad_norm": 0.4792398512363434, "learning_rate": 4.1217917779876235e-05, "loss": 0.0219, "step": 57840 }, { "grad_norm": 0.2793099582195282, "learning_rate": 4.120164060780141e-05, "loss": 0.0263, "step": 57850 }, { "grad_norm": 0.23381629586219788, "learning_rate": 4.1185364397901935e-05, "loss": 0.0257, "step": 57860 }, { "grad_norm": 0.24046792089939117, "learning_rate": 4.116908915195776e-05, "loss": 0.0151, "step": 57870 }, { "grad_norm": 0.4952741265296936, "learning_rate": 4.115281487174872e-05, "loss": 0.0287, "step": 57880 }, { "grad_norm": 0.29779520630836487, "learning_rate": 4.113654155905452e-05, "loss": 0.0214, "step": 57890 }, { "grad_norm": 0.20266777276992798, "learning_rate": 4.112026921565483e-05, "loss": 0.0223, "step": 57900 }, { "grad_norm": 0.25876516103744507, "learning_rate": 4.1103997843329147e-05, "loss": 0.0181, "step": 57910 }, { "grad_norm": 0.38663578033447266, "learning_rate": 4.108772744385689e-05, "loss": 0.035, "step": 57920 }, { "grad_norm": 0.28284987807273865, "learning_rate": 4.1071458019017346e-05, "loss": 0.0225, "step": 57930 }, { "grad_norm": 0.3077891170978546, "learning_rate": 4.1055189570589734e-05, "loss": 0.017, "step": 57940 }, { "grad_norm": 0.3564761281013489, "learning_rate": 4.103892210035315e-05, "loss": 0.0227, "step": 57950 }, { "grad_norm": 0.3330244719982147, "learning_rate": 4.102265561008656e-05, "loss": 0.0267, "step": 57960 }, { "grad_norm": 0.29820960760116577, "learning_rate": 4.100639010156885e-05, "loss": 0.0223, "step": 57970 }, { "grad_norm": 0.27511969208717346, "learning_rate": 4.099012557657882e-05, "loss": 0.0189, "step": 57980 }, { "grad_norm": 0.27094921469688416, "learning_rate": 4.09738620368951e-05, "loss": 0.019, "step": 57990 }, { "grad_norm": 0.4631775915622711, "learning_rate": 4.095759948429626e-05, "loss": 0.0228, "step": 58000 }, { "grad_norm": 0.22614917159080505, "learning_rate": 4.094133792056075e-05, "loss": 0.0219, "step": 58010 }, { "grad_norm": 0.4069003164768219, "learning_rate": 4.092507734746688e-05, "loss": 0.0262, "step": 58020 }, { "grad_norm": 0.30880311131477356, "learning_rate": 4.090881776679293e-05, "loss": 0.0185, "step": 58030 }, { "grad_norm": 0.3114272952079773, "learning_rate": 4.0892559180316994e-05, "loss": 0.0211, "step": 58040 }, { "grad_norm": 0.2031809538602829, "learning_rate": 4.08763015898171e-05, "loss": 0.0176, "step": 58050 }, { "grad_norm": 0.23724684119224548, "learning_rate": 4.0860044997071145e-05, "loss": 0.0208, "step": 58060 }, { "grad_norm": 0.40601256489753723, "learning_rate": 4.084378940385693e-05, "loss": 0.0198, "step": 58070 }, { "grad_norm": 0.3642597794532776, "learning_rate": 4.082753481195211e-05, "loss": 0.0181, "step": 58080 }, { "grad_norm": 0.2659195363521576, "learning_rate": 4.0811281223134324e-05, "loss": 0.0176, "step": 58090 }, { "grad_norm": 0.31107285618782043, "learning_rate": 4.0795028639181e-05, "loss": 0.0219, "step": 58100 }, { "grad_norm": 0.29108095169067383, "learning_rate": 4.077877706186951e-05, "loss": 0.0168, "step": 58110 }, { "grad_norm": 0.5509617924690247, "learning_rate": 4.076252649297709e-05, "loss": 0.0263, "step": 58120 }, { "grad_norm": 0.35140183568000793, "learning_rate": 4.07462769342809e-05, "loss": 0.0227, "step": 58130 }, { "grad_norm": 0.3900038003921509, "learning_rate": 4.073002838755793e-05, "loss": 0.0257, "step": 58140 }, { "grad_norm": 0.26840487122535706, "learning_rate": 4.0713780854585115e-05, "loss": 0.0295, "step": 58150 }, { "grad_norm": 0.30682116746902466, "learning_rate": 4.069753433713927e-05, "loss": 0.0208, "step": 58160 }, { "grad_norm": 0.34239494800567627, "learning_rate": 4.068128883699709e-05, "loss": 0.03, "step": 58170 }, { "grad_norm": 0.2993837893009186, "learning_rate": 4.0665044355935155e-05, "loss": 0.0213, "step": 58180 }, { "grad_norm": 0.2953571379184723, "learning_rate": 4.064880089572993e-05, "loss": 0.02, "step": 58190 }, { "grad_norm": 0.3598936200141907, "learning_rate": 4.063255845815779e-05, "loss": 0.0358, "step": 58200 }, { "grad_norm": 0.2688233554363251, "learning_rate": 4.0616317044994945e-05, "loss": 0.0197, "step": 58210 }, { "grad_norm": 0.22156476974487305, "learning_rate": 4.0600076658017585e-05, "loss": 0.0235, "step": 58220 }, { "grad_norm": 0.24512968957424164, "learning_rate": 4.0583837299001706e-05, "loss": 0.0185, "step": 58230 }, { "grad_norm": 0.2624925374984741, "learning_rate": 4.056759896972322e-05, "loss": 0.0212, "step": 58240 }, { "grad_norm": 0.19585339725017548, "learning_rate": 4.055136167195793e-05, "loss": 0.0205, "step": 58250 }, { "grad_norm": 0.3166971206665039, "learning_rate": 4.053512540748152e-05, "loss": 0.0244, "step": 58260 }, { "grad_norm": 0.2607763707637787, "learning_rate": 4.051889017806955e-05, "loss": 0.0242, "step": 58270 }, { "grad_norm": 0.36935311555862427, "learning_rate": 4.0502655985497516e-05, "loss": 0.0203, "step": 58280 }, { "grad_norm": 0.300529420375824, "learning_rate": 4.048642283154074e-05, "loss": 0.0249, "step": 58290 }, { "grad_norm": 0.400375634431839, "learning_rate": 4.0470190717974464e-05, "loss": 0.0226, "step": 58300 }, { "grad_norm": 0.2242276519536972, "learning_rate": 4.045395964657381e-05, "loss": 0.0243, "step": 58310 }, { "grad_norm": 0.24520325660705566, "learning_rate": 4.043772961911376e-05, "loss": 0.0177, "step": 58320 }, { "grad_norm": 0.32980313897132874, "learning_rate": 4.042150063736924e-05, "loss": 0.0245, "step": 58330 }, { "grad_norm": 0.3868369162082672, "learning_rate": 4.0405272703114976e-05, "loss": 0.0221, "step": 58340 }, { "grad_norm": 0.25494882464408875, "learning_rate": 4.038904581812569e-05, "loss": 0.0179, "step": 58350 }, { "grad_norm": 0.3411004841327667, "learning_rate": 4.03728199841759e-05, "loss": 0.0188, "step": 58360 }, { "grad_norm": 0.28238996863365173, "learning_rate": 4.0356595203040044e-05, "loss": 0.0256, "step": 58370 }, { "grad_norm": 0.28662753105163574, "learning_rate": 4.0340371476492426e-05, "loss": 0.02, "step": 58380 }, { "grad_norm": 0.37065067887306213, "learning_rate": 4.032414880630727e-05, "loss": 0.0226, "step": 58390 }, { "grad_norm": 0.2746477723121643, "learning_rate": 4.0307927194258646e-05, "loss": 0.0157, "step": 58400 }, { "grad_norm": 0.5823516845703125, "learning_rate": 4.029170664212054e-05, "loss": 0.0192, "step": 58410 }, { "grad_norm": 0.30515027046203613, "learning_rate": 4.0275487151666794e-05, "loss": 0.0188, "step": 58420 }, { "grad_norm": 0.30932164192199707, "learning_rate": 4.025926872467116e-05, "loss": 0.0243, "step": 58430 }, { "grad_norm": 0.22208231687545776, "learning_rate": 4.024305136290725e-05, "loss": 0.0205, "step": 58440 }, { "grad_norm": 0.28441494703292847, "learning_rate": 4.022683506814858e-05, "loss": 0.0244, "step": 58450 }, { "grad_norm": 0.22153833508491516, "learning_rate": 4.021061984216853e-05, "loss": 0.0174, "step": 58460 }, { "grad_norm": 0.2914285361766815, "learning_rate": 4.019440568674035e-05, "loss": 0.0203, "step": 58470 }, { "grad_norm": 0.36018821597099304, "learning_rate": 4.017819260363724e-05, "loss": 0.0231, "step": 58480 }, { "grad_norm": 0.2946470379829407, "learning_rate": 4.016198059463223e-05, "loss": 0.0347, "step": 58490 }, { "grad_norm": 0.23025447130203247, "learning_rate": 4.0145769661498233e-05, "loss": 0.0226, "step": 58500 }, { "grad_norm": 0.2692283093929291, "learning_rate": 4.012955980600804e-05, "loss": 0.0254, "step": 58510 }, { "grad_norm": 0.37025341391563416, "learning_rate": 4.0113351029934355e-05, "loss": 0.0222, "step": 58520 }, { "grad_norm": 0.7289502024650574, "learning_rate": 4.0097143335049714e-05, "loss": 0.0226, "step": 58530 }, { "grad_norm": 0.26269814372062683, "learning_rate": 4.0080936723126606e-05, "loss": 0.0171, "step": 58540 }, { "grad_norm": 0.2965354323387146, "learning_rate": 4.006473119593735e-05, "loss": 0.019, "step": 58550 }, { "grad_norm": 0.37596526741981506, "learning_rate": 4.004852675525415e-05, "loss": 0.0191, "step": 58560 }, { "grad_norm": 0.24037714302539825, "learning_rate": 4.00323234028491e-05, "loss": 0.0231, "step": 58570 }, { "grad_norm": 0.3834361732006073, "learning_rate": 4.0016121140494186e-05, "loss": 0.0271, "step": 58580 }, { "grad_norm": 0.3697766661643982, "learning_rate": 3.999991996996122e-05, "loss": 0.0213, "step": 58590 }, { "grad_norm": 0.2936896085739136, "learning_rate": 3.9983719893022e-05, "loss": 0.0185, "step": 58600 }, { "grad_norm": 0.41784995794296265, "learning_rate": 3.996752091144812e-05, "loss": 0.0197, "step": 58610 }, { "grad_norm": 0.2307387888431549, "learning_rate": 3.9951323027011055e-05, "loss": 0.0223, "step": 58620 }, { "grad_norm": 0.234537735581398, "learning_rate": 3.993512624148219e-05, "loss": 0.0232, "step": 58630 }, { "grad_norm": 0.23663143813610077, "learning_rate": 3.99189305566328e-05, "loss": 0.0231, "step": 58640 }, { "grad_norm": 0.32963526248931885, "learning_rate": 3.990273597423401e-05, "loss": 0.0215, "step": 58650 }, { "grad_norm": 0.2885645627975464, "learning_rate": 3.988654249605681e-05, "loss": 0.0207, "step": 58660 }, { "grad_norm": 0.2881295084953308, "learning_rate": 3.987035012387213e-05, "loss": 0.0201, "step": 58670 }, { "grad_norm": 0.6162644028663635, "learning_rate": 3.9854158859450735e-05, "loss": 0.0194, "step": 58680 }, { "grad_norm": 0.36921629309654236, "learning_rate": 3.983796870456328e-05, "loss": 0.0258, "step": 58690 }, { "grad_norm": 0.3936885893344879, "learning_rate": 3.982177966098028e-05, "loss": 0.0184, "step": 58700 }, { "grad_norm": 0.24063992500305176, "learning_rate": 3.980559173047216e-05, "loss": 0.0306, "step": 58710 }, { "grad_norm": 0.33229461312294006, "learning_rate": 3.9789404914809176e-05, "loss": 0.0233, "step": 58720 }, { "grad_norm": 0.1824266016483307, "learning_rate": 3.9773219215761544e-05, "loss": 0.0194, "step": 58730 }, { "grad_norm": 0.3985096216201782, "learning_rate": 3.975703463509928e-05, "loss": 0.0251, "step": 58740 }, { "grad_norm": 0.35837432742118835, "learning_rate": 3.974085117459231e-05, "loss": 0.0337, "step": 58750 }, { "grad_norm": 0.5168352127075195, "learning_rate": 3.972466883601043e-05, "loss": 0.0281, "step": 58760 }, { "grad_norm": 0.304362028837204, "learning_rate": 3.9708487621123316e-05, "loss": 0.0265, "step": 58770 }, { "grad_norm": 0.5126231908798218, "learning_rate": 3.969230753170051e-05, "loss": 0.0277, "step": 58780 }, { "grad_norm": 0.2230077087879181, "learning_rate": 3.967612856951146e-05, "loss": 0.0208, "step": 58790 }, { "grad_norm": 0.30782416462898254, "learning_rate": 3.9659950736325476e-05, "loss": 0.0195, "step": 58800 }, { "grad_norm": 0.40167373418807983, "learning_rate": 3.964377403391174e-05, "loss": 0.0416, "step": 58810 }, { "grad_norm": 0.2653111517429352, "learning_rate": 3.9627598464039303e-05, "loss": 0.0386, "step": 58820 }, { "grad_norm": 0.17625197768211365, "learning_rate": 3.9611424028477096e-05, "loss": 0.0203, "step": 58830 }, { "grad_norm": 0.28667905926704407, "learning_rate": 3.959525072899394e-05, "loss": 0.0311, "step": 58840 }, { "grad_norm": 0.2934286892414093, "learning_rate": 3.9579078567358506e-05, "loss": 0.0201, "step": 58850 }, { "grad_norm": 0.4789142310619354, "learning_rate": 3.956290754533939e-05, "loss": 0.0235, "step": 58860 }, { "grad_norm": 0.27672046422958374, "learning_rate": 3.954673766470501e-05, "loss": 0.0293, "step": 58870 }, { "grad_norm": 0.2925417125225067, "learning_rate": 3.9530568927223674e-05, "loss": 0.023, "step": 58880 }, { "grad_norm": 0.23865413665771484, "learning_rate": 3.95144013346636e-05, "loss": 0.0197, "step": 58890 }, { "grad_norm": 0.14996719360351562, "learning_rate": 3.949823488879282e-05, "loss": 0.0162, "step": 58900 }, { "grad_norm": 0.41234707832336426, "learning_rate": 3.9482069591379276e-05, "loss": 0.0291, "step": 58910 }, { "grad_norm": 0.37779733538627625, "learning_rate": 3.9465905444190784e-05, "loss": 0.0221, "step": 58920 }, { "grad_norm": 0.2699580490589142, "learning_rate": 3.944974244899505e-05, "loss": 0.0218, "step": 58930 }, { "grad_norm": 0.2898901700973511, "learning_rate": 3.9433580607559616e-05, "loss": 0.0244, "step": 58940 }, { "grad_norm": 0.36573711037635803, "learning_rate": 3.941741992165193e-05, "loss": 0.0332, "step": 58950 }, { "grad_norm": 0.28646957874298096, "learning_rate": 3.940126039303928e-05, "loss": 0.0231, "step": 58960 }, { "grad_norm": 0.2121543288230896, "learning_rate": 3.938510202348886e-05, "loss": 0.0192, "step": 58970 }, { "grad_norm": 0.24664945900440216, "learning_rate": 3.93689448147677e-05, "loss": 0.0186, "step": 58980 }, { "grad_norm": 0.3106021583080292, "learning_rate": 3.935278876864278e-05, "loss": 0.017, "step": 58990 }, { "grad_norm": 0.2742455005645752, "learning_rate": 3.933663388688087e-05, "loss": 0.0364, "step": 59000 }, { "grad_norm": 0.37870728969573975, "learning_rate": 3.9320480171248645e-05, "loss": 0.0188, "step": 59010 }, { "grad_norm": 0.6610321998596191, "learning_rate": 3.930432762351265e-05, "loss": 0.0275, "step": 59020 }, { "grad_norm": 0.36775246262550354, "learning_rate": 3.928817624543931e-05, "loss": 0.0235, "step": 59030 }, { "grad_norm": 0.27557656168937683, "learning_rate": 3.9272026038794885e-05, "loss": 0.017, "step": 59040 }, { "grad_norm": 0.3080541491508484, "learning_rate": 3.9255877005345586e-05, "loss": 0.0273, "step": 59050 }, { "grad_norm": 0.6915799379348755, "learning_rate": 3.923972914685743e-05, "loss": 0.0292, "step": 59060 }, { "grad_norm": 0.3526228666305542, "learning_rate": 3.922358246509631e-05, "loss": 0.0227, "step": 59070 }, { "grad_norm": 0.3814578056335449, "learning_rate": 3.920743696182801e-05, "loss": 0.026, "step": 59080 }, { "grad_norm": 0.40012654662132263, "learning_rate": 3.9191292638818164e-05, "loss": 0.0219, "step": 59090 }, { "grad_norm": 0.21180318295955658, "learning_rate": 3.9175149497832316e-05, "loss": 0.0206, "step": 59100 }, { "grad_norm": 0.16541481018066406, "learning_rate": 3.915900754063584e-05, "loss": 0.0197, "step": 59110 }, { "grad_norm": 0.25820690393447876, "learning_rate": 3.9142866768994014e-05, "loss": 0.0215, "step": 59120 }, { "grad_norm": 0.21576254069805145, "learning_rate": 3.912672718467193e-05, "loss": 0.0256, "step": 59130 }, { "grad_norm": 0.26746830344200134, "learning_rate": 3.911058878943463e-05, "loss": 0.0179, "step": 59140 }, { "grad_norm": 0.26859039068222046, "learning_rate": 3.9094451585046974e-05, "loss": 0.0223, "step": 59150 }, { "grad_norm": 0.24539814889431, "learning_rate": 3.90783155732737e-05, "loss": 0.0257, "step": 59160 }, { "grad_norm": 0.35256069898605347, "learning_rate": 3.9062180755879373e-05, "loss": 0.0236, "step": 59170 }, { "grad_norm": 0.28847089409828186, "learning_rate": 3.904604713462855e-05, "loss": 0.0197, "step": 59180 }, { "grad_norm": 0.23566798865795135, "learning_rate": 3.902991471128554e-05, "loss": 0.0203, "step": 59190 }, { "grad_norm": 0.3387194573879242, "learning_rate": 3.901378348761456e-05, "loss": 0.0197, "step": 59200 }, { "grad_norm": 0.35906368494033813, "learning_rate": 3.899765346537971e-05, "loss": 0.0237, "step": 59210 }, { "grad_norm": 0.29806971549987793, "learning_rate": 3.898152464634493e-05, "loss": 0.0156, "step": 59220 }, { "grad_norm": 0.8287525773048401, "learning_rate": 3.896539703227402e-05, "loss": 0.023, "step": 59230 }, { "grad_norm": 0.2910185158252716, "learning_rate": 3.894927062493073e-05, "loss": 0.0224, "step": 59240 }, { "grad_norm": 0.19844907522201538, "learning_rate": 3.893314542607859e-05, "loss": 0.0181, "step": 59250 }, { "grad_norm": 0.6503843665122986, "learning_rate": 3.891702143748103e-05, "loss": 0.024, "step": 59260 }, { "grad_norm": 0.2983045279979706, "learning_rate": 3.890089866090133e-05, "loss": 0.0181, "step": 59270 }, { "grad_norm": 0.386570543050766, "learning_rate": 3.888477709810267e-05, "loss": 0.0238, "step": 59280 }, { "grad_norm": 0.2410307675600052, "learning_rate": 3.886865675084806e-05, "loss": 0.0246, "step": 59290 }, { "grad_norm": 0.24056927859783173, "learning_rate": 3.885253762090043e-05, "loss": 0.0202, "step": 59300 }, { "grad_norm": 0.23036955296993256, "learning_rate": 3.8836419710022534e-05, "loss": 0.029, "step": 59310 }, { "grad_norm": 0.4007314145565033, "learning_rate": 3.882030301997698e-05, "loss": 0.0234, "step": 59320 }, { "grad_norm": 0.33357787132263184, "learning_rate": 3.880418755252629e-05, "loss": 0.0224, "step": 59330 }, { "grad_norm": 0.24117475748062134, "learning_rate": 3.87880733094328e-05, "loss": 0.0276, "step": 59340 }, { "grad_norm": 0.28159594535827637, "learning_rate": 3.877196029245877e-05, "loss": 0.0243, "step": 59350 }, { "grad_norm": 0.26270928978919983, "learning_rate": 3.875584850336627e-05, "loss": 0.0259, "step": 59360 }, { "grad_norm": 0.3929254412651062, "learning_rate": 3.873973794391728e-05, "loss": 0.0211, "step": 59370 }, { "grad_norm": 0.2831258475780487, "learning_rate": 3.872362861587361e-05, "loss": 0.0218, "step": 59380 }, { "grad_norm": 0.3956146538257599, "learning_rate": 3.870752052099698e-05, "loss": 0.022, "step": 59390 }, { "grad_norm": 0.20438846945762634, "learning_rate": 3.8691413661048916e-05, "loss": 0.0219, "step": 59400 }, { "grad_norm": 0.269596129655838, "learning_rate": 3.867530803779085e-05, "loss": 0.0193, "step": 59410 }, { "grad_norm": 0.2714550793170929, "learning_rate": 3.8659203652984054e-05, "loss": 0.0231, "step": 59420 }, { "grad_norm": 0.3028890788555145, "learning_rate": 3.864310050838972e-05, "loss": 0.0274, "step": 59430 }, { "grad_norm": 0.32270920276641846, "learning_rate": 3.8626998605768835e-05, "loss": 0.0159, "step": 59440 }, { "grad_norm": 0.2575109899044037, "learning_rate": 3.8610897946882294e-05, "loss": 0.0252, "step": 59450 }, { "grad_norm": 0.34655919671058655, "learning_rate": 3.859479853349082e-05, "loss": 0.024, "step": 59460 }, { "grad_norm": 0.4036065936088562, "learning_rate": 3.857870036735503e-05, "loss": 0.0232, "step": 59470 }, { "grad_norm": 0.35005640983581543, "learning_rate": 3.856260345023538e-05, "loss": 0.0273, "step": 59480 }, { "grad_norm": 0.24252672493457794, "learning_rate": 3.854650778389224e-05, "loss": 0.021, "step": 59490 }, { "grad_norm": 0.2382974922657013, "learning_rate": 3.8530413370085786e-05, "loss": 0.0251, "step": 59500 }, { "grad_norm": 0.20701822638511658, "learning_rate": 3.851432021057608e-05, "loss": 0.0189, "step": 59510 }, { "grad_norm": 0.22148776054382324, "learning_rate": 3.849822830712306e-05, "loss": 0.0275, "step": 59520 }, { "grad_norm": 0.49283283948898315, "learning_rate": 3.848213766148649e-05, "loss": 0.0221, "step": 59530 }, { "grad_norm": 0.24774231016635895, "learning_rate": 3.846604827542602e-05, "loss": 0.0188, "step": 59540 }, { "grad_norm": 0.3303679823875427, "learning_rate": 3.844996015070117e-05, "loss": 0.0266, "step": 59550 }, { "grad_norm": 0.29732954502105713, "learning_rate": 3.843387328907132e-05, "loss": 0.0188, "step": 59560 }, { "grad_norm": 0.38776612281799316, "learning_rate": 3.8417787692295696e-05, "loss": 0.0261, "step": 59570 }, { "grad_norm": 0.35344234108924866, "learning_rate": 3.84017033621334e-05, "loss": 0.02, "step": 59580 }, { "grad_norm": 0.2716183662414551, "learning_rate": 3.838562030034338e-05, "loss": 0.0221, "step": 59590 }, { "grad_norm": 0.48250266909599304, "learning_rate": 3.836953850868447e-05, "loss": 0.029, "step": 59600 }, { "grad_norm": 0.3358168303966522, "learning_rate": 3.835345798891533e-05, "loss": 0.0212, "step": 59610 }, { "grad_norm": 0.215969055891037, "learning_rate": 3.833737874279452e-05, "loss": 0.0314, "step": 59620 }, { "grad_norm": 0.37189632654190063, "learning_rate": 3.8321300772080433e-05, "loss": 0.023, "step": 59630 }, { "grad_norm": 0.32624709606170654, "learning_rate": 3.8305224078531345e-05, "loss": 0.0176, "step": 59640 }, { "grad_norm": 0.38740256428718567, "learning_rate": 3.828914866390537e-05, "loss": 0.0301, "step": 59650 }, { "grad_norm": 0.41638925671577454, "learning_rate": 3.827307452996048e-05, "loss": 0.0293, "step": 59660 }, { "grad_norm": 0.3687521815299988, "learning_rate": 3.825700167845454e-05, "loss": 0.0299, "step": 59670 }, { "grad_norm": 0.3363018333911896, "learning_rate": 3.8240930111145214e-05, "loss": 0.0249, "step": 59680 }, { "grad_norm": 0.13401031494140625, "learning_rate": 3.8224859829790116e-05, "loss": 0.0188, "step": 59690 }, { "grad_norm": 0.20133596658706665, "learning_rate": 3.820879083614664e-05, "loss": 0.0236, "step": 59700 }, { "grad_norm": 0.4043578505516052, "learning_rate": 3.819272313197208e-05, "loss": 0.0249, "step": 59710 }, { "grad_norm": 0.24359892308712006, "learning_rate": 3.817665671902355e-05, "loss": 0.0145, "step": 59720 }, { "grad_norm": 0.3337486982345581, "learning_rate": 3.816059159905808e-05, "loss": 0.0268, "step": 59730 }, { "grad_norm": 0.20777420699596405, "learning_rate": 3.8144527773832476e-05, "loss": 0.0242, "step": 59740 }, { "grad_norm": 0.2959195077419281, "learning_rate": 3.812846524510352e-05, "loss": 0.0235, "step": 59750 }, { "grad_norm": 0.30880939960479736, "learning_rate": 3.811240401462775e-05, "loss": 0.0216, "step": 59760 }, { "grad_norm": 0.4117458760738373, "learning_rate": 3.809634408416162e-05, "loss": 0.0288, "step": 59770 }, { "grad_norm": 0.31345608830451965, "learning_rate": 3.8080285455461394e-05, "loss": 0.0238, "step": 59780 }, { "grad_norm": 0.22659863531589508, "learning_rate": 3.806422813028323e-05, "loss": 0.0199, "step": 59790 }, { "grad_norm": 0.3316585123538971, "learning_rate": 3.8048172110383114e-05, "loss": 0.02, "step": 59800 }, { "grad_norm": 0.6981910467147827, "learning_rate": 3.8032117397516944e-05, "loss": 0.0212, "step": 59810 }, { "grad_norm": 0.2866569757461548, "learning_rate": 3.801606399344042e-05, "loss": 0.0187, "step": 59820 }, { "grad_norm": 0.3330939710140228, "learning_rate": 3.800001189990913e-05, "loss": 0.0182, "step": 59830 }, { "grad_norm": 0.2729133069515228, "learning_rate": 3.798396111867847e-05, "loss": 0.0219, "step": 59840 }, { "grad_norm": 0.2949451804161072, "learning_rate": 3.796791165150378e-05, "loss": 0.021, "step": 59850 }, { "grad_norm": 0.33393868803977966, "learning_rate": 3.795186350014018e-05, "loss": 0.0202, "step": 59860 }, { "grad_norm": 0.3706522285938263, "learning_rate": 3.793581666634266e-05, "loss": 0.021, "step": 59870 }, { "grad_norm": 0.302409291267395, "learning_rate": 3.791977115186609e-05, "loss": 0.0156, "step": 59880 }, { "grad_norm": 0.31618762016296387, "learning_rate": 3.79037269584652e-05, "loss": 0.0224, "step": 59890 }, { "grad_norm": 0.4493381977081299, "learning_rate": 3.7887684087894545e-05, "loss": 0.0312, "step": 59900 }, { "grad_norm": 0.23466701805591583, "learning_rate": 3.787164254190854e-05, "loss": 0.0215, "step": 59910 }, { "grad_norm": 0.29828009009361267, "learning_rate": 3.7855602322261473e-05, "loss": 0.0182, "step": 59920 }, { "grad_norm": 0.31118905544281006, "learning_rate": 3.783956343070746e-05, "loss": 0.0287, "step": 59930 }, { "grad_norm": 0.3926687240600586, "learning_rate": 3.7823525869000526e-05, "loss": 0.0209, "step": 59940 }, { "grad_norm": 0.3901537358760834, "learning_rate": 3.78074896388945e-05, "loss": 0.0269, "step": 59950 }, { "grad_norm": 0.4166125953197479, "learning_rate": 3.7791454742143075e-05, "loss": 0.0205, "step": 59960 }, { "grad_norm": 0.2697688937187195, "learning_rate": 3.7775421180499795e-05, "loss": 0.0182, "step": 59970 }, { "grad_norm": 0.303322434425354, "learning_rate": 3.7759388955718086e-05, "loss": 0.0211, "step": 59980 }, { "grad_norm": 0.24285005033016205, "learning_rate": 3.774335806955117e-05, "loss": 0.0206, "step": 59990 }, { "grad_norm": 0.3382817208766937, "learning_rate": 3.7727328523752215e-05, "loss": 0.0211, "step": 60000 }, { "grad_norm": 0.2752213776111603, "learning_rate": 3.771130032007415e-05, "loss": 0.0264, "step": 60010 }, { "grad_norm": 0.26449453830718994, "learning_rate": 3.769527346026982e-05, "loss": 0.0188, "step": 60020 }, { "grad_norm": 0.34186986088752747, "learning_rate": 3.767924794609188e-05, "loss": 0.0228, "step": 60030 }, { "grad_norm": 0.24820157885551453, "learning_rate": 3.7663223779292854e-05, "loss": 0.0193, "step": 60040 }, { "grad_norm": 0.1981578767299652, "learning_rate": 3.764720096162512e-05, "loss": 0.0218, "step": 60050 }, { "grad_norm": 0.5539417862892151, "learning_rate": 3.763117949484092e-05, "loss": 0.0198, "step": 60060 }, { "grad_norm": 0.3159014582633972, "learning_rate": 3.7615159380692344e-05, "loss": 0.0175, "step": 60070 }, { "grad_norm": 0.25552108883857727, "learning_rate": 3.759914062093132e-05, "loss": 0.0223, "step": 60080 }, { "grad_norm": 0.22171807289123535, "learning_rate": 3.7583123217309615e-05, "loss": 0.019, "step": 60090 }, { "grad_norm": 0.3252074718475342, "learning_rate": 3.7567107171578904e-05, "loss": 0.0266, "step": 60100 }, { "grad_norm": 0.41715872287750244, "learning_rate": 3.755109248549066e-05, "loss": 0.0265, "step": 60110 }, { "grad_norm": 0.2756373882293701, "learning_rate": 3.75350791607962e-05, "loss": 0.0193, "step": 60120 }, { "grad_norm": 0.29591992497444153, "learning_rate": 3.751906719924676e-05, "loss": 0.0168, "step": 60130 }, { "grad_norm": 0.25947439670562744, "learning_rate": 3.750305660259337e-05, "loss": 0.025, "step": 60140 }, { "grad_norm": 0.24121429026126862, "learning_rate": 3.748704737258693e-05, "loss": 0.0196, "step": 60150 }, { "grad_norm": 0.2699805498123169, "learning_rate": 3.747103951097816e-05, "loss": 0.0269, "step": 60160 }, { "grad_norm": 0.3541908860206604, "learning_rate": 3.745503301951768e-05, "loss": 0.0211, "step": 60170 }, { "grad_norm": 0.4369378983974457, "learning_rate": 3.743902789995592e-05, "loss": 0.0206, "step": 60180 }, { "grad_norm": 0.3432653248310089, "learning_rate": 3.7423024154043183e-05, "loss": 0.0233, "step": 60190 }, { "grad_norm": 0.30162283778190613, "learning_rate": 3.740702178352963e-05, "loss": 0.0213, "step": 60200 }, { "grad_norm": 0.2065618634223938, "learning_rate": 3.739102079016523e-05, "loss": 0.0193, "step": 60210 }, { "grad_norm": 0.29259082674980164, "learning_rate": 3.7375021175699846e-05, "loss": 0.0171, "step": 60220 }, { "grad_norm": 0.3366262912750244, "learning_rate": 3.735902294188317e-05, "loss": 0.0224, "step": 60230 }, { "grad_norm": 0.2342335283756256, "learning_rate": 3.734302609046473e-05, "loss": 0.0167, "step": 60240 }, { "grad_norm": 0.310835063457489, "learning_rate": 3.732703062319391e-05, "loss": 0.0187, "step": 60250 }, { "grad_norm": 0.4064253270626068, "learning_rate": 3.731103654181999e-05, "loss": 0.0232, "step": 60260 }, { "grad_norm": 0.34632962942123413, "learning_rate": 3.729504384809203e-05, "loss": 0.0236, "step": 60270 }, { "grad_norm": 0.29557982087135315, "learning_rate": 3.727905254375898e-05, "loss": 0.0245, "step": 60280 }, { "grad_norm": 0.25936460494995117, "learning_rate": 3.7263062630569614e-05, "loss": 0.0176, "step": 60290 }, { "grad_norm": 0.18045403063297272, "learning_rate": 3.724707411027256e-05, "loss": 0.0175, "step": 60300 }, { "grad_norm": 0.22611723840236664, "learning_rate": 3.723108698461631e-05, "loss": 0.0214, "step": 60310 }, { "grad_norm": 0.2900105118751526, "learning_rate": 3.72151012553492e-05, "loss": 0.0188, "step": 60320 }, { "grad_norm": 0.2996436655521393, "learning_rate": 3.719911692421939e-05, "loss": 0.0192, "step": 60330 }, { "grad_norm": 0.32442066073417664, "learning_rate": 3.71831339929749e-05, "loss": 0.0269, "step": 60340 }, { "grad_norm": 0.2676234245300293, "learning_rate": 3.716715246336362e-05, "loss": 0.0238, "step": 60350 }, { "grad_norm": 0.2193911373615265, "learning_rate": 3.715117233713324e-05, "loss": 0.0146, "step": 60360 }, { "grad_norm": 0.3707737326622009, "learning_rate": 3.713519361603135e-05, "loss": 0.0197, "step": 60370 }, { "grad_norm": 0.40432873368263245, "learning_rate": 3.711921630180532e-05, "loss": 0.0237, "step": 60380 }, { "grad_norm": 0.25934964418411255, "learning_rate": 3.710324039620245e-05, "loss": 0.0163, "step": 60390 }, { "grad_norm": 0.6793568134307861, "learning_rate": 3.708726590096982e-05, "loss": 0.0228, "step": 60400 }, { "grad_norm": 0.2838200032711029, "learning_rate": 3.707129281785437e-05, "loss": 0.0239, "step": 60410 }, { "grad_norm": 0.3876804709434509, "learning_rate": 3.705532114860291e-05, "loss": 0.0213, "step": 60420 }, { "grad_norm": 0.3783573806285858, "learning_rate": 3.703935089496207e-05, "loss": 0.0298, "step": 60430 }, { "grad_norm": 0.2728075683116913, "learning_rate": 3.702338205867829e-05, "loss": 0.0193, "step": 60440 }, { "grad_norm": 0.3498280942440033, "learning_rate": 3.7007414641497976e-05, "loss": 0.0334, "step": 60450 }, { "grad_norm": 0.4037318825721741, "learning_rate": 3.699144864516726e-05, "loss": 0.0361, "step": 60460 }, { "grad_norm": 0.2583432197570801, "learning_rate": 3.6975484071432153e-05, "loss": 0.0232, "step": 60470 }, { "grad_norm": 0.23453789949417114, "learning_rate": 3.695952092203854e-05, "loss": 0.0229, "step": 60480 }, { "grad_norm": 0.3116205036640167, "learning_rate": 3.69435591987321e-05, "loss": 0.0271, "step": 60490 }, { "grad_norm": 0.17704540491104126, "learning_rate": 3.6927598903258374e-05, "loss": 0.0189, "step": 60500 }, { "grad_norm": 0.34035855531692505, "learning_rate": 3.691164003736281e-05, "loss": 0.0211, "step": 60510 }, { "grad_norm": 0.19622129201889038, "learning_rate": 3.689568260279061e-05, "loss": 0.0245, "step": 60520 }, { "grad_norm": 0.2876036763191223, "learning_rate": 3.687972660128686e-05, "loss": 0.018, "step": 60530 }, { "grad_norm": 0.3777940571308136, "learning_rate": 3.686377203459648e-05, "loss": 0.0199, "step": 60540 }, { "grad_norm": 0.24241603910923004, "learning_rate": 3.6847818904464226e-05, "loss": 0.02, "step": 60550 }, { "grad_norm": 0.2500145435333252, "learning_rate": 3.683186721263474e-05, "loss": 0.018, "step": 60560 }, { "grad_norm": 0.3182380199432373, "learning_rate": 3.6815916960852435e-05, "loss": 0.0229, "step": 60570 }, { "grad_norm": 0.16560304164886475, "learning_rate": 3.679996815086165e-05, "loss": 0.0204, "step": 60580 }, { "grad_norm": 0.20906682312488556, "learning_rate": 3.678402078440649e-05, "loss": 0.0192, "step": 60590 }, { "grad_norm": 0.19891291856765747, "learning_rate": 3.676807486323096e-05, "loss": 0.0186, "step": 60600 }, { "grad_norm": 0.2856200635433197, "learning_rate": 3.6752130389078864e-05, "loss": 0.0252, "step": 60610 }, { "grad_norm": 0.25247442722320557, "learning_rate": 3.673618736369388e-05, "loss": 0.0206, "step": 60620 }, { "grad_norm": 0.3201679289340973, "learning_rate": 3.6720245788819474e-05, "loss": 0.0195, "step": 60630 }, { "grad_norm": 0.2646290361881256, "learning_rate": 3.6704305666199044e-05, "loss": 0.0215, "step": 60640 }, { "grad_norm": 0.15738189220428467, "learning_rate": 3.668836699757576e-05, "loss": 0.019, "step": 60650 }, { "grad_norm": 0.45376449823379517, "learning_rate": 3.667242978469265e-05, "loss": 0.0239, "step": 60660 }, { "grad_norm": 0.3927919864654541, "learning_rate": 3.6656494029292575e-05, "loss": 0.0268, "step": 60670 }, { "grad_norm": 0.2922157347202301, "learning_rate": 3.664055973311825e-05, "loss": 0.0223, "step": 60680 }, { "grad_norm": 0.2845486104488373, "learning_rate": 3.662462689791221e-05, "loss": 0.0182, "step": 60690 }, { "grad_norm": 0.2846667766571045, "learning_rate": 3.660869552541689e-05, "loss": 0.0182, "step": 60700 }, { "grad_norm": 0.2917756140232086, "learning_rate": 3.659276561737448e-05, "loss": 0.0209, "step": 60710 }, { "grad_norm": 0.5394718050956726, "learning_rate": 3.657683717552708e-05, "loss": 0.0263, "step": 60720 }, { "grad_norm": 0.3041737675666809, "learning_rate": 3.6560910201616574e-05, "loss": 0.0172, "step": 60730 }, { "grad_norm": 0.3034529983997345, "learning_rate": 3.654498469738472e-05, "loss": 0.0234, "step": 60740 }, { "grad_norm": 0.3579464554786682, "learning_rate": 3.6529060664573126e-05, "loss": 0.0254, "step": 60750 }, { "grad_norm": 0.28305867314338684, "learning_rate": 3.6513138104923176e-05, "loss": 0.0253, "step": 60760 }, { "grad_norm": 0.4330471456050873, "learning_rate": 3.6497217020176176e-05, "loss": 0.0219, "step": 60770 }, { "grad_norm": 0.29034098982810974, "learning_rate": 3.648129741207323e-05, "loss": 0.0269, "step": 60780 }, { "grad_norm": 0.18406090140342712, "learning_rate": 3.646537928235527e-05, "loss": 0.02, "step": 60790 }, { "grad_norm": 0.32419171929359436, "learning_rate": 3.6449462632763075e-05, "loss": 0.0202, "step": 60800 }, { "grad_norm": 0.4176368713378906, "learning_rate": 3.643354746503729e-05, "loss": 0.0257, "step": 60810 }, { "grad_norm": 0.24391183257102966, "learning_rate": 3.6417633780918345e-05, "loss": 0.0159, "step": 60820 }, { "grad_norm": 0.2795977294445038, "learning_rate": 3.640172158214655e-05, "loss": 0.0171, "step": 60830 }, { "grad_norm": 0.35086458921432495, "learning_rate": 3.6385810870462035e-05, "loss": 0.0207, "step": 60840 }, { "grad_norm": 0.27834469079971313, "learning_rate": 3.636990164760479e-05, "loss": 0.0232, "step": 60850 }, { "grad_norm": 0.17824383080005646, "learning_rate": 3.6353993915314596e-05, "loss": 0.0173, "step": 60860 }, { "grad_norm": 0.21553152799606323, "learning_rate": 3.6338087675331124e-05, "loss": 0.0203, "step": 60870 }, { "grad_norm": 0.24650701880455017, "learning_rate": 3.632218292939383e-05, "loss": 0.0199, "step": 60880 }, { "grad_norm": 0.30363228917121887, "learning_rate": 3.6306279679242033e-05, "loss": 0.0227, "step": 60890 }, { "grad_norm": 0.2453974485397339, "learning_rate": 3.629037792661492e-05, "loss": 0.016, "step": 60900 }, { "grad_norm": 0.29483741521835327, "learning_rate": 3.627447767325147e-05, "loss": 0.0199, "step": 60910 }, { "grad_norm": 0.24119795858860016, "learning_rate": 3.62585789208905e-05, "loss": 0.0184, "step": 60920 }, { "grad_norm": 0.2601480185985565, "learning_rate": 3.624268167127068e-05, "loss": 0.0226, "step": 60930 }, { "grad_norm": 0.2257896065711975, "learning_rate": 3.622678592613051e-05, "loss": 0.0192, "step": 60940 }, { "grad_norm": 0.2775936424732208, "learning_rate": 3.6210891687208304e-05, "loss": 0.0234, "step": 60950 }, { "grad_norm": 0.19348761439323425, "learning_rate": 3.619499895624228e-05, "loss": 0.0209, "step": 60960 }, { "grad_norm": 0.26586511731147766, "learning_rate": 3.61791077349704e-05, "loss": 0.0233, "step": 60970 }, { "grad_norm": 0.3683241903781891, "learning_rate": 3.616321802513053e-05, "loss": 0.0226, "step": 60980 }, { "grad_norm": 0.33101561665534973, "learning_rate": 3.6147329828460325e-05, "loss": 0.0173, "step": 60990 }, { "grad_norm": 0.34854671359062195, "learning_rate": 3.6131443146697306e-05, "loss": 0.0201, "step": 61000 }, { "grad_norm": 0.21610720455646515, "learning_rate": 3.611555798157879e-05, "loss": 0.0173, "step": 61010 }, { "grad_norm": 0.2744556963443756, "learning_rate": 3.6099674334842e-05, "loss": 0.0199, "step": 61020 }, { "grad_norm": 0.19993771612644196, "learning_rate": 3.6083792208223934e-05, "loss": 0.0238, "step": 61030 }, { "grad_norm": 0.3551735281944275, "learning_rate": 3.606791160346142e-05, "loss": 0.0236, "step": 61040 }, { "grad_norm": 0.20999789237976074, "learning_rate": 3.6052032522291134e-05, "loss": 0.0225, "step": 61050 }, { "grad_norm": 0.5362053513526917, "learning_rate": 3.603615496644962e-05, "loss": 0.0294, "step": 61060 }, { "grad_norm": 0.408295214176178, "learning_rate": 3.60202789376732e-05, "loss": 0.0233, "step": 61070 }, { "grad_norm": 0.25883185863494873, "learning_rate": 3.6004404437698036e-05, "loss": 0.0214, "step": 61080 }, { "grad_norm": 0.3350380063056946, "learning_rate": 3.5988531468260184e-05, "loss": 0.0183, "step": 61090 }, { "grad_norm": 0.24801023304462433, "learning_rate": 3.597266003109546e-05, "loss": 0.0215, "step": 61100 }, { "grad_norm": 0.25499075651168823, "learning_rate": 3.5956790127939557e-05, "loss": 0.0211, "step": 61110 }, { "grad_norm": 0.2386559545993805, "learning_rate": 3.594092176052796e-05, "loss": 0.0234, "step": 61120 }, { "grad_norm": 0.283442884683609, "learning_rate": 3.592505493059603e-05, "loss": 0.0227, "step": 61130 }, { "grad_norm": 0.3579449951648712, "learning_rate": 3.5909189639878905e-05, "loss": 0.0251, "step": 61140 }, { "grad_norm": 0.4021552801132202, "learning_rate": 3.5893325890111626e-05, "loss": 0.0203, "step": 61150 }, { "grad_norm": 0.23771460354328156, "learning_rate": 3.5877463683029034e-05, "loss": 0.0267, "step": 61160 }, { "grad_norm": 0.29984667897224426, "learning_rate": 3.586160302036578e-05, "loss": 0.0246, "step": 61170 }, { "grad_norm": 0.2385241985321045, "learning_rate": 3.584574390385636e-05, "loss": 0.0182, "step": 61180 }, { "grad_norm": 0.18750576674938202, "learning_rate": 3.582988633523511e-05, "loss": 0.0201, "step": 61190 }, { "grad_norm": 0.5461944341659546, "learning_rate": 3.581403031623616e-05, "loss": 0.0226, "step": 61200 }, { "grad_norm": 0.22446231544017792, "learning_rate": 3.579817584859354e-05, "loss": 0.0193, "step": 61210 }, { "grad_norm": 0.3976004123687744, "learning_rate": 3.578232293404107e-05, "loss": 0.0232, "step": 61220 }, { "grad_norm": 0.4787865877151489, "learning_rate": 3.5766471574312385e-05, "loss": 0.0247, "step": 61230 }, { "grad_norm": 0.35922500491142273, "learning_rate": 3.5750621771140964e-05, "loss": 0.0267, "step": 61240 }, { "grad_norm": 0.27351275086402893, "learning_rate": 3.5734773526260115e-05, "loss": 0.0263, "step": 61250 }, { "grad_norm": 0.3554512560367584, "learning_rate": 3.5718926841402996e-05, "loss": 0.0293, "step": 61260 }, { "grad_norm": 0.26077741384506226, "learning_rate": 3.5703081718302555e-05, "loss": 0.0274, "step": 61270 }, { "grad_norm": 0.3886374533176422, "learning_rate": 3.568723815869162e-05, "loss": 0.0304, "step": 61280 }, { "grad_norm": 0.28389281034469604, "learning_rate": 3.5671396164302777e-05, "loss": 0.0247, "step": 61290 }, { "grad_norm": 0.2381024956703186, "learning_rate": 3.565555573686853e-05, "loss": 0.0219, "step": 61300 }, { "grad_norm": 0.3193003833293915, "learning_rate": 3.563971687812113e-05, "loss": 0.0185, "step": 61310 }, { "grad_norm": 0.31456273794174194, "learning_rate": 3.5623879589792706e-05, "loss": 0.0215, "step": 61320 }, { "grad_norm": 0.3616739511489868, "learning_rate": 3.5608043873615185e-05, "loss": 0.0259, "step": 61330 }, { "grad_norm": 0.18588882684707642, "learning_rate": 3.559220973132035e-05, "loss": 0.0189, "step": 61340 }, { "grad_norm": 0.1807446926832199, "learning_rate": 3.557637716463981e-05, "loss": 0.0161, "step": 61350 }, { "grad_norm": 0.2644173800945282, "learning_rate": 3.556054617530497e-05, "loss": 0.0222, "step": 61360 }, { "grad_norm": 0.15422363579273224, "learning_rate": 3.5544716765047106e-05, "loss": 0.014, "step": 61370 }, { "grad_norm": 0.33876657485961914, "learning_rate": 3.5528888935597274e-05, "loss": 0.0275, "step": 61380 }, { "grad_norm": 0.2545927166938782, "learning_rate": 3.551306268868636e-05, "loss": 0.0183, "step": 61390 }, { "grad_norm": 0.3400023281574249, "learning_rate": 3.5497238026045166e-05, "loss": 0.0195, "step": 61400 }, { "grad_norm": 0.2688771188259125, "learning_rate": 3.548141494940421e-05, "loss": 0.0175, "step": 61410 }, { "grad_norm": 0.336883544921875, "learning_rate": 3.546559346049388e-05, "loss": 0.0184, "step": 61420 }, { "grad_norm": 0.32148781418800354, "learning_rate": 3.544977356104441e-05, "loss": 0.0171, "step": 61430 }, { "grad_norm": 0.2290063202381134, "learning_rate": 3.543395525278582e-05, "loss": 0.0257, "step": 61440 }, { "grad_norm": 0.3792070150375366, "learning_rate": 3.5418138537447985e-05, "loss": 0.0156, "step": 61450 }, { "grad_norm": 0.42194440960884094, "learning_rate": 3.5402323416760574e-05, "loss": 0.0208, "step": 61460 }, { "grad_norm": 0.31972309947013855, "learning_rate": 3.538650989245316e-05, "loss": 0.0208, "step": 61470 }, { "grad_norm": 0.29518991708755493, "learning_rate": 3.537069796625504e-05, "loss": 0.0185, "step": 61480 }, { "grad_norm": 0.2156151384115219, "learning_rate": 3.53548876398954e-05, "loss": 0.0162, "step": 61490 }, { "grad_norm": 0.3264101445674896, "learning_rate": 3.5339078915103214e-05, "loss": 0.0236, "step": 61500 }, { "grad_norm": 0.3160388469696045, "learning_rate": 3.532327179360733e-05, "loss": 0.0236, "step": 61510 }, { "grad_norm": 0.3170883059501648, "learning_rate": 3.530746627713636e-05, "loss": 0.0178, "step": 61520 }, { "grad_norm": 0.2276279181241989, "learning_rate": 3.5291662367418805e-05, "loss": 0.0149, "step": 61530 }, { "grad_norm": 0.2999931275844574, "learning_rate": 3.527586006618293e-05, "loss": 0.0214, "step": 61540 }, { "grad_norm": 0.267211377620697, "learning_rate": 3.526005937515687e-05, "loss": 0.0302, "step": 61550 }, { "grad_norm": 0.2791493833065033, "learning_rate": 3.524426029606856e-05, "loss": 0.0311, "step": 61560 }, { "grad_norm": 0.3083677887916565, "learning_rate": 3.5228462830645746e-05, "loss": 0.0245, "step": 61570 }, { "grad_norm": 0.296842098236084, "learning_rate": 3.5212666980616036e-05, "loss": 0.0196, "step": 61580 }, { "grad_norm": 0.21313516795635223, "learning_rate": 3.519687274770681e-05, "loss": 0.0233, "step": 61590 }, { "grad_norm": 0.21098898351192474, "learning_rate": 3.518108013364535e-05, "loss": 0.0292, "step": 61600 }, { "grad_norm": 0.2978573739528656, "learning_rate": 3.516528914015868e-05, "loss": 0.0269, "step": 61610 }, { "grad_norm": 0.3507331609725952, "learning_rate": 3.5149499768973694e-05, "loss": 0.0264, "step": 61620 }, { "grad_norm": 0.3383287191390991, "learning_rate": 3.513371202181708e-05, "loss": 0.0194, "step": 61630 }, { "grad_norm": 0.3460216224193573, "learning_rate": 3.511792590041537e-05, "loss": 0.0189, "step": 61640 }, { "grad_norm": 0.34687864780426025, "learning_rate": 3.510214140649488e-05, "loss": 0.0326, "step": 61650 }, { "grad_norm": 0.32104215025901794, "learning_rate": 3.508635854178184e-05, "loss": 0.0262, "step": 61660 }, { "grad_norm": 0.35065388679504395, "learning_rate": 3.50705773080022e-05, "loss": 0.0247, "step": 61670 }, { "grad_norm": 0.33265992999076843, "learning_rate": 3.505479770688178e-05, "loss": 0.019, "step": 61680 }, { "grad_norm": 0.4071672558784485, "learning_rate": 3.503901974014622e-05, "loss": 0.0222, "step": 61690 }, { "grad_norm": 0.2709243893623352, "learning_rate": 3.502324340952095e-05, "loss": 0.0221, "step": 61700 }, { "grad_norm": 0.7414878606796265, "learning_rate": 3.500746871673126e-05, "loss": 0.0212, "step": 61710 }, { "grad_norm": 0.29482683539390564, "learning_rate": 3.499169566350227e-05, "loss": 0.0256, "step": 61720 }, { "grad_norm": 0.31871891021728516, "learning_rate": 3.4975924251558876e-05, "loss": 0.0189, "step": 61730 }, { "grad_norm": 0.23114091157913208, "learning_rate": 3.496015448262582e-05, "loss": 0.02, "step": 61740 }, { "grad_norm": 0.6101691722869873, "learning_rate": 3.494438635842764e-05, "loss": 0.0223, "step": 61750 }, { "grad_norm": 0.40841901302337646, "learning_rate": 3.492861988068874e-05, "loss": 0.0208, "step": 61760 }, { "grad_norm": 0.9659653902053833, "learning_rate": 3.491285505113332e-05, "loss": 0.0266, "step": 61770 }, { "grad_norm": 0.20593637228012085, "learning_rate": 3.4897091871485376e-05, "loss": 0.0252, "step": 61780 }, { "grad_norm": 0.2051093876361847, "learning_rate": 3.488133034346875e-05, "loss": 0.0192, "step": 61790 }, { "grad_norm": 0.2352920025587082, "learning_rate": 3.486557046880713e-05, "loss": 0.0195, "step": 61800 }, { "grad_norm": 0.24604785442352295, "learning_rate": 3.484981224922395e-05, "loss": 0.0255, "step": 61810 }, { "grad_norm": 0.36910006403923035, "learning_rate": 3.4834055686442544e-05, "loss": 0.0208, "step": 61820 }, { "grad_norm": 0.45187896490097046, "learning_rate": 3.4818300782186e-05, "loss": 0.0226, "step": 61830 }, { "grad_norm": 0.2921312749385834, "learning_rate": 3.4802547538177236e-05, "loss": 0.0199, "step": 61840 }, { "grad_norm": 0.24930183589458466, "learning_rate": 3.478679595613905e-05, "loss": 0.0214, "step": 61850 }, { "grad_norm": 0.29824933409690857, "learning_rate": 3.4771046037793975e-05, "loss": 0.0259, "step": 61860 }, { "grad_norm": 0.20372989773750305, "learning_rate": 3.4755297784864416e-05, "loss": 0.0215, "step": 61870 }, { "grad_norm": 0.24748030304908752, "learning_rate": 3.473955119907257e-05, "loss": 0.0192, "step": 61880 }, { "grad_norm": 0.2815631031990051, "learning_rate": 3.472380628214046e-05, "loss": 0.0221, "step": 61890 }, { "grad_norm": 0.2662128806114197, "learning_rate": 3.4708063035789906e-05, "loss": 0.0188, "step": 61900 }, { "grad_norm": 0.31534168124198914, "learning_rate": 3.4692321461742614e-05, "loss": 0.0216, "step": 61910 }, { "grad_norm": 0.30456432700157166, "learning_rate": 3.4676581561720023e-05, "loss": 0.0208, "step": 61920 }, { "grad_norm": 0.3171454966068268, "learning_rate": 3.466084333744344e-05, "loss": 0.0251, "step": 61930 }, { "grad_norm": 0.2008330225944519, "learning_rate": 3.4645106790633964e-05, "loss": 0.0257, "step": 61940 }, { "grad_norm": 0.1530432552099228, "learning_rate": 3.4629371923012526e-05, "loss": 0.0227, "step": 61950 }, { "grad_norm": 0.3305770456790924, "learning_rate": 3.4613638736299856e-05, "loss": 0.023, "step": 61960 }, { "grad_norm": 0.17199039459228516, "learning_rate": 3.459790723221652e-05, "loss": 0.0134, "step": 61970 }, { "grad_norm": 0.29666680097579956, "learning_rate": 3.45821774124829e-05, "loss": 0.0223, "step": 61980 }, { "grad_norm": 0.24828466773033142, "learning_rate": 3.456644927881918e-05, "loss": 0.0184, "step": 61990 }, { "grad_norm": 0.26166534423828125, "learning_rate": 3.4550722832945354e-05, "loss": 0.0208, "step": 62000 }, { "grad_norm": 0.6749250888824463, "learning_rate": 3.453499807658126e-05, "loss": 0.0209, "step": 62010 }, { "grad_norm": 0.2985232174396515, "learning_rate": 3.451927501144653e-05, "loss": 0.0211, "step": 62020 }, { "grad_norm": 0.3957742750644684, "learning_rate": 3.450355363926059e-05, "loss": 0.0216, "step": 62030 }, { "grad_norm": 0.32668060064315796, "learning_rate": 3.448783396174273e-05, "loss": 0.0214, "step": 62040 }, { "grad_norm": 0.5792550444602966, "learning_rate": 3.4472115980612034e-05, "loss": 0.0227, "step": 62050 }, { "grad_norm": 0.17949526011943817, "learning_rate": 3.4456399697587396e-05, "loss": 0.0185, "step": 62060 }, { "grad_norm": 0.3556705415248871, "learning_rate": 3.4440685114387506e-05, "loss": 0.0205, "step": 62070 }, { "grad_norm": 0.2531745433807373, "learning_rate": 3.442497223273091e-05, "loss": 0.0199, "step": 62080 }, { "grad_norm": 0.3462311625480652, "learning_rate": 3.4409261054335926e-05, "loss": 0.0301, "step": 62090 }, { "grad_norm": 0.34154441952705383, "learning_rate": 3.439355158092069e-05, "loss": 0.0297, "step": 62100 }, { "grad_norm": 0.3007054030895233, "learning_rate": 3.437784381420322e-05, "loss": 0.0213, "step": 62110 }, { "grad_norm": 0.25603803992271423, "learning_rate": 3.436213775590126e-05, "loss": 0.0186, "step": 62120 }, { "grad_norm": 0.5280859470367432, "learning_rate": 3.4346433407732405e-05, "loss": 0.0247, "step": 62130 }, { "grad_norm": 0.29474034905433655, "learning_rate": 3.433073077141405e-05, "loss": 0.02, "step": 62140 }, { "grad_norm": 0.2528005540370941, "learning_rate": 3.431502984866343e-05, "loss": 0.0221, "step": 62150 }, { "grad_norm": 0.48705416917800903, "learning_rate": 3.429933064119753e-05, "loss": 0.0346, "step": 62160 }, { "grad_norm": 0.273112028837204, "learning_rate": 3.4283633150733255e-05, "loss": 0.0175, "step": 62170 }, { "grad_norm": 0.2788912355899811, "learning_rate": 3.4267937378987225e-05, "loss": 0.0253, "step": 62180 }, { "grad_norm": 0.25120213627815247, "learning_rate": 3.425224332767592e-05, "loss": 0.0185, "step": 62190 }, { "grad_norm": 0.35238707065582275, "learning_rate": 3.4236550998515595e-05, "loss": 0.0226, "step": 62200 }, { "grad_norm": 0.40856000781059265, "learning_rate": 3.422086039322235e-05, "loss": 0.0197, "step": 62210 }, { "grad_norm": 0.33025428652763367, "learning_rate": 3.420517151351209e-05, "loss": 0.0182, "step": 62220 }, { "grad_norm": 0.33077454566955566, "learning_rate": 3.418948436110053e-05, "loss": 0.0257, "step": 62230 }, { "grad_norm": 0.28254207968711853, "learning_rate": 3.417379893770319e-05, "loss": 0.022, "step": 62240 }, { "grad_norm": 0.3161621689796448, "learning_rate": 3.41581152450354e-05, "loss": 0.024, "step": 62250 }, { "grad_norm": 0.32179149985313416, "learning_rate": 3.414243328481231e-05, "loss": 0.023, "step": 62260 }, { "grad_norm": 0.24790754914283752, "learning_rate": 3.412675305874888e-05, "loss": 0.0148, "step": 62270 }, { "grad_norm": 0.28637224435806274, "learning_rate": 3.4111074568559876e-05, "loss": 0.0202, "step": 62280 }, { "grad_norm": 0.3085092306137085, "learning_rate": 3.4095397815959846e-05, "loss": 0.0274, "step": 62290 }, { "grad_norm": 0.4004705846309662, "learning_rate": 3.407972280266322e-05, "loss": 0.021, "step": 62300 }, { "grad_norm": 0.7017695903778076, "learning_rate": 3.4064049530384166e-05, "loss": 0.0252, "step": 62310 }, { "grad_norm": 0.18425285816192627, "learning_rate": 3.40483780008367e-05, "loss": 0.0238, "step": 62320 }, { "grad_norm": 0.43911370635032654, "learning_rate": 3.4032708215734635e-05, "loss": 0.0253, "step": 62330 }, { "grad_norm": 0.6056832671165466, "learning_rate": 3.4017040176791584e-05, "loss": 0.0269, "step": 62340 }, { "grad_norm": 0.9671433568000793, "learning_rate": 3.4001373885720976e-05, "loss": 0.022, "step": 62350 }, { "grad_norm": 0.3708006739616394, "learning_rate": 3.398570934423608e-05, "loss": 0.0223, "step": 62360 }, { "grad_norm": 0.23849788308143616, "learning_rate": 3.397004655404994e-05, "loss": 0.0205, "step": 62370 }, { "grad_norm": 0.3330485224723816, "learning_rate": 3.395438551687541e-05, "loss": 0.0259, "step": 62380 }, { "grad_norm": 0.303312748670578, "learning_rate": 3.393872623442515e-05, "loss": 0.0194, "step": 62390 }, { "grad_norm": 0.3738276958465576, "learning_rate": 3.392306870841164e-05, "loss": 0.0156, "step": 62400 }, { "grad_norm": 0.4144148528575897, "learning_rate": 3.390741294054715e-05, "loss": 0.0226, "step": 62410 }, { "grad_norm": 0.298971951007843, "learning_rate": 3.38917589325438e-05, "loss": 0.0184, "step": 62420 }, { "grad_norm": 0.252362996339798, "learning_rate": 3.387610668611349e-05, "loss": 0.0179, "step": 62430 }, { "grad_norm": 0.23160922527313232, "learning_rate": 3.38604562029679e-05, "loss": 0.0176, "step": 62440 }, { "grad_norm": 0.2057291865348816, "learning_rate": 3.384480748481855e-05, "loss": 0.0188, "step": 62450 }, { "grad_norm": 0.19686397910118103, "learning_rate": 3.382916053337676e-05, "loss": 0.0161, "step": 62460 }, { "grad_norm": 0.22431698441505432, "learning_rate": 3.381351535035366e-05, "loss": 0.0201, "step": 62470 }, { "grad_norm": 0.265236496925354, "learning_rate": 3.379787193746019e-05, "loss": 0.017, "step": 62480 }, { "grad_norm": 0.2326219379901886, "learning_rate": 3.378223029640709e-05, "loss": 0.0141, "step": 62490 }, { "grad_norm": 0.3102295696735382, "learning_rate": 3.376659042890489e-05, "loss": 0.0169, "step": 62500 }, { "grad_norm": 0.266726553440094, "learning_rate": 3.375095233666397e-05, "loss": 0.017, "step": 62510 }, { "grad_norm": 0.23169071972370148, "learning_rate": 3.373531602139447e-05, "loss": 0.0209, "step": 62520 }, { "grad_norm": 0.46828755736351013, "learning_rate": 3.371968148480635e-05, "loss": 0.0199, "step": 62530 }, { "grad_norm": 0.2100098431110382, "learning_rate": 3.370404872860937e-05, "loss": 0.0167, "step": 62540 }, { "grad_norm": 0.3391802906990051, "learning_rate": 3.3688417754513145e-05, "loss": 0.0189, "step": 62550 }, { "grad_norm": 0.22256314754486084, "learning_rate": 3.367278856422703e-05, "loss": 0.0226, "step": 62560 }, { "grad_norm": 0.3553479313850403, "learning_rate": 3.365716115946021e-05, "loss": 0.0221, "step": 62570 }, { "grad_norm": 0.2571852505207062, "learning_rate": 3.3641535541921666e-05, "loss": 0.03, "step": 62580 }, { "grad_norm": 0.2967525124549866, "learning_rate": 3.36259117133202e-05, "loss": 0.0212, "step": 62590 }, { "grad_norm": 0.32967081665992737, "learning_rate": 3.36102896753644e-05, "loss": 0.0199, "step": 62600 }, { "grad_norm": 0.21257126331329346, "learning_rate": 3.35946694297627e-05, "loss": 0.0192, "step": 62610 }, { "grad_norm": 0.20329874753952026, "learning_rate": 3.357905097822328e-05, "loss": 0.0196, "step": 62620 }, { "grad_norm": 0.43487268686294556, "learning_rate": 3.356343432245416e-05, "loss": 0.0189, "step": 62630 }, { "grad_norm": 0.2668378949165344, "learning_rate": 3.354781946416315e-05, "loss": 0.031, "step": 62640 }, { "grad_norm": 0.2610883116722107, "learning_rate": 3.3532206405057864e-05, "loss": 0.0201, "step": 62650 }, { "grad_norm": 0.6716023087501526, "learning_rate": 3.351659514684573e-05, "loss": 0.0187, "step": 62660 }, { "grad_norm": 0.24580247700214386, "learning_rate": 3.350098569123395e-05, "loss": 0.0193, "step": 62670 }, { "grad_norm": 0.2512175440788269, "learning_rate": 3.348537803992958e-05, "loss": 0.019, "step": 62680 }, { "grad_norm": 0.18162941932678223, "learning_rate": 3.346977219463945e-05, "loss": 0.0251, "step": 62690 }, { "grad_norm": 0.2947453558444977, "learning_rate": 3.345416815707017e-05, "loss": 0.0202, "step": 62700 }, { "grad_norm": 0.2941213548183441, "learning_rate": 3.3438565928928185e-05, "loss": 0.02, "step": 62710 }, { "grad_norm": 0.2116820514202118, "learning_rate": 3.3422965511919736e-05, "loss": 0.018, "step": 62720 }, { "grad_norm": 0.32561036944389343, "learning_rate": 3.3407366907750846e-05, "loss": 0.0178, "step": 62730 }, { "grad_norm": 0.22268171608448029, "learning_rate": 3.339177011812737e-05, "loss": 0.0202, "step": 62740 }, { "grad_norm": 0.40107354521751404, "learning_rate": 3.3376175144754935e-05, "loss": 0.0284, "step": 62750 }, { "grad_norm": 0.1751369684934616, "learning_rate": 3.3360581989339004e-05, "loss": 0.0201, "step": 62760 }, { "grad_norm": 0.29881513118743896, "learning_rate": 3.3344990653584804e-05, "loss": 0.0204, "step": 62770 }, { "grad_norm": 0.29030969738960266, "learning_rate": 3.332940113919739e-05, "loss": 0.0236, "step": 62780 }, { "grad_norm": 0.21641120314598083, "learning_rate": 3.331381344788161e-05, "loss": 0.0189, "step": 62790 }, { "grad_norm": 0.3255714178085327, "learning_rate": 3.3298227581342076e-05, "loss": 0.0222, "step": 62800 }, { "grad_norm": 0.22908525168895721, "learning_rate": 3.328264354128329e-05, "loss": 0.0315, "step": 62810 }, { "grad_norm": 0.2655884027481079, "learning_rate": 3.3267061329409465e-05, "loss": 0.021, "step": 62820 }, { "grad_norm": 0.28424209356307983, "learning_rate": 3.325148094742466e-05, "loss": 0.0206, "step": 62830 }, { "grad_norm": 0.22375446557998657, "learning_rate": 3.323590239703271e-05, "loss": 0.0205, "step": 62840 }, { "grad_norm": 0.3081974387168884, "learning_rate": 3.322032567993728e-05, "loss": 0.0207, "step": 62850 }, { "grad_norm": 0.27226969599723816, "learning_rate": 3.3204750797841774e-05, "loss": 0.0206, "step": 62860 }, { "grad_norm": 0.2778530716896057, "learning_rate": 3.31891777524495e-05, "loss": 0.0189, "step": 62870 }, { "grad_norm": 0.23995210230350494, "learning_rate": 3.317360654546347e-05, "loss": 0.0201, "step": 62880 }, { "grad_norm": 0.23937691748142242, "learning_rate": 3.3158037178586534e-05, "loss": 0.0203, "step": 62890 }, { "grad_norm": 0.3518136441707611, "learning_rate": 3.3142469653521316e-05, "loss": 0.0268, "step": 62900 }, { "grad_norm": 0.2670005261898041, "learning_rate": 3.3126903971970284e-05, "loss": 0.0168, "step": 62910 }, { "grad_norm": 0.40794941782951355, "learning_rate": 3.311134013563564e-05, "loss": 0.0186, "step": 62920 }, { "grad_norm": 0.24735353887081146, "learning_rate": 3.309577814621947e-05, "loss": 0.0187, "step": 62930 }, { "grad_norm": 0.2718370258808136, "learning_rate": 3.308021800542359e-05, "loss": 0.0178, "step": 62940 }, { "grad_norm": 0.2673853635787964, "learning_rate": 3.306465971494963e-05, "loss": 0.0265, "step": 62950 }, { "grad_norm": 0.3192259669303894, "learning_rate": 3.304910327649901e-05, "loss": 0.0216, "step": 62960 }, { "grad_norm": 0.42281222343444824, "learning_rate": 3.303354869177297e-05, "loss": 0.0163, "step": 62970 }, { "grad_norm": 0.24091094732284546, "learning_rate": 3.301799596247255e-05, "loss": 0.0181, "step": 62980 }, { "grad_norm": 0.2910759747028351, "learning_rate": 3.300244509029854e-05, "loss": 0.0189, "step": 62990 }, { "grad_norm": 0.307688444852829, "learning_rate": 3.2986896076951585e-05, "loss": 0.0166, "step": 63000 }, { "grad_norm": 0.4134184420108795, "learning_rate": 3.297134892413211e-05, "loss": 0.0211, "step": 63010 }, { "grad_norm": 0.27366095781326294, "learning_rate": 3.295580363354031e-05, "loss": 0.0172, "step": 63020 }, { "grad_norm": 0.3566730320453644, "learning_rate": 3.2940260206876194e-05, "loss": 0.0237, "step": 63030 }, { "grad_norm": 0.22279541194438934, "learning_rate": 3.292471864583958e-05, "loss": 0.0212, "step": 63040 }, { "grad_norm": 0.26393935084342957, "learning_rate": 3.290917895213004e-05, "loss": 0.0257, "step": 63050 }, { "grad_norm": 0.30380597710609436, "learning_rate": 3.289364112744701e-05, "loss": 0.0244, "step": 63060 }, { "grad_norm": 0.31071361899375916, "learning_rate": 3.287810517348967e-05, "loss": 0.0242, "step": 63070 }, { "grad_norm": 0.3982275128364563, "learning_rate": 3.2862571091957014e-05, "loss": 0.022, "step": 63080 }, { "grad_norm": 0.20723675191402435, "learning_rate": 3.284703888454781e-05, "loss": 0.0201, "step": 63090 }, { "grad_norm": 0.3715226948261261, "learning_rate": 3.2831508552960646e-05, "loss": 0.0244, "step": 63100 }, { "grad_norm": 0.2932375967502594, "learning_rate": 3.281598009889387e-05, "loss": 0.02, "step": 63110 }, { "grad_norm": 0.21146006882190704, "learning_rate": 3.28004535240457e-05, "loss": 0.0293, "step": 63120 }, { "grad_norm": 0.4080887734889984, "learning_rate": 3.278492883011407e-05, "loss": 0.0219, "step": 63130 }, { "grad_norm": 0.2934236228466034, "learning_rate": 3.276940601879675e-05, "loss": 0.0206, "step": 63140 }, { "grad_norm": 0.3094453811645508, "learning_rate": 3.275388509179127e-05, "loss": 0.0229, "step": 63150 }, { "grad_norm": 0.2611497640609741, "learning_rate": 3.2738366050794985e-05, "loss": 0.0262, "step": 63160 }, { "grad_norm": 0.39049869775772095, "learning_rate": 3.2722848897505034e-05, "loss": 0.0201, "step": 63170 }, { "grad_norm": 0.22477562725543976, "learning_rate": 3.2707333633618356e-05, "loss": 0.0141, "step": 63180 }, { "grad_norm": 0.19300758838653564, "learning_rate": 3.269182026083167e-05, "loss": 0.0183, "step": 63190 }, { "grad_norm": 0.5334663987159729, "learning_rate": 3.2676308780841516e-05, "loss": 0.0249, "step": 63200 }, { "grad_norm": 0.4865274727344513, "learning_rate": 3.266079919534416e-05, "loss": 0.0235, "step": 63210 }, { "grad_norm": 0.3231602609157562, "learning_rate": 3.264529150603575e-05, "loss": 0.0201, "step": 63220 }, { "grad_norm": 0.33563464879989624, "learning_rate": 3.262978571461217e-05, "loss": 0.02, "step": 63230 }, { "grad_norm": 0.23719315230846405, "learning_rate": 3.26142818227691e-05, "loss": 0.0183, "step": 63240 }, { "grad_norm": 0.2865428626537323, "learning_rate": 3.259877983220203e-05, "loss": 0.0223, "step": 63250 }, { "grad_norm": 0.24800832569599152, "learning_rate": 3.2583279744606246e-05, "loss": 0.0313, "step": 63260 }, { "grad_norm": 0.28551462292671204, "learning_rate": 3.25677815616768e-05, "loss": 0.0256, "step": 63270 }, { "grad_norm": 0.31672731041908264, "learning_rate": 3.255228528510856e-05, "loss": 0.0243, "step": 63280 }, { "grad_norm": 0.2545925974845886, "learning_rate": 3.253679091659617e-05, "loss": 0.0219, "step": 63290 }, { "grad_norm": 0.34302979707717896, "learning_rate": 3.252129845783404e-05, "loss": 0.0248, "step": 63300 }, { "grad_norm": 0.3585922420024872, "learning_rate": 3.2505807910516456e-05, "loss": 0.0196, "step": 63310 }, { "grad_norm": 0.3831574618816376, "learning_rate": 3.249031927633742e-05, "loss": 0.0233, "step": 63320 }, { "grad_norm": 0.3117402493953705, "learning_rate": 3.2474832556990737e-05, "loss": 0.0198, "step": 63330 }, { "grad_norm": 0.6685828566551208, "learning_rate": 3.245934775417002e-05, "loss": 0.0208, "step": 63340 }, { "grad_norm": 0.2894003689289093, "learning_rate": 3.244386486956866e-05, "loss": 0.0227, "step": 63350 }, { "grad_norm": 0.24043801426887512, "learning_rate": 3.2428383904879846e-05, "loss": 0.0183, "step": 63360 }, { "grad_norm": 0.2799140214920044, "learning_rate": 3.241290486179652e-05, "loss": 0.0343, "step": 63370 }, { "grad_norm": 0.3334488868713379, "learning_rate": 3.239742774201151e-05, "loss": 0.0204, "step": 63380 }, { "grad_norm": 0.2566649317741394, "learning_rate": 3.238195254721733e-05, "loss": 0.0203, "step": 63390 }, { "grad_norm": 0.2748544216156006, "learning_rate": 3.236647927910633e-05, "loss": 0.0183, "step": 63400 }, { "grad_norm": 0.4931245744228363, "learning_rate": 3.2351007939370646e-05, "loss": 0.0247, "step": 63410 }, { "grad_norm": 0.2922115623950958, "learning_rate": 3.2335538529702184e-05, "loss": 0.0176, "step": 63420 }, { "grad_norm": 0.22720125317573547, "learning_rate": 3.232007105179268e-05, "loss": 0.0188, "step": 63430 }, { "grad_norm": 0.37900251150131226, "learning_rate": 3.230460550733363e-05, "loss": 0.0176, "step": 63440 }, { "grad_norm": 0.24490635097026825, "learning_rate": 3.2289141898016315e-05, "loss": 0.0229, "step": 63450 }, { "grad_norm": 0.5562359094619751, "learning_rate": 3.227368022553181e-05, "loss": 0.0231, "step": 63460 }, { "grad_norm": 0.5400556921958923, "learning_rate": 3.225822049157099e-05, "loss": 0.0268, "step": 63470 }, { "grad_norm": 0.5272133946418762, "learning_rate": 3.2242762697824494e-05, "loss": 0.0173, "step": 63480 }, { "grad_norm": 0.3470805585384369, "learning_rate": 3.222730684598279e-05, "loss": 0.0322, "step": 63490 }, { "grad_norm": 0.3181706368923187, "learning_rate": 3.221185293773605e-05, "loss": 0.023, "step": 63500 }, { "grad_norm": 0.2974272668361664, "learning_rate": 3.2196400974774354e-05, "loss": 0.0201, "step": 63510 }, { "grad_norm": 0.20486564934253693, "learning_rate": 3.2180950958787485e-05, "loss": 0.0283, "step": 63520 }, { "grad_norm": 0.23184342682361603, "learning_rate": 3.2165502891465024e-05, "loss": 0.0198, "step": 63530 }, { "grad_norm": 0.3749663233757019, "learning_rate": 3.215005677449636e-05, "loss": 0.0205, "step": 63540 }, { "grad_norm": 0.2994743287563324, "learning_rate": 3.2134612609570644e-05, "loss": 0.0198, "step": 63550 }, { "grad_norm": 0.272943913936615, "learning_rate": 3.211917039837682e-05, "loss": 0.0205, "step": 63560 }, { "grad_norm": 0.38716378808021545, "learning_rate": 3.210373014260365e-05, "loss": 0.027, "step": 63570 }, { "grad_norm": 0.251169890165329, "learning_rate": 3.208829184393965e-05, "loss": 0.0202, "step": 63580 }, { "grad_norm": 0.2540537714958191, "learning_rate": 3.207285550407312e-05, "loss": 0.0303, "step": 63590 }, { "grad_norm": 0.2995606064796448, "learning_rate": 3.205742112469217e-05, "loss": 0.0174, "step": 63600 }, { "grad_norm": 0.44821053743362427, "learning_rate": 3.204198870748465e-05, "loss": 0.0299, "step": 63610 }, { "grad_norm": 0.2665182650089264, "learning_rate": 3.202655825413824e-05, "loss": 0.0236, "step": 63620 }, { "grad_norm": 1.2435811758041382, "learning_rate": 3.2011129766340416e-05, "loss": 0.0315, "step": 63630 }, { "grad_norm": 0.38590919971466064, "learning_rate": 3.1995703245778384e-05, "loss": 0.0206, "step": 63640 }, { "grad_norm": 0.32706770300865173, "learning_rate": 3.1980278694139186e-05, "loss": 0.0235, "step": 63650 }, { "grad_norm": 0.32617315649986267, "learning_rate": 3.196485611310962e-05, "loss": 0.0278, "step": 63660 }, { "grad_norm": 0.24612554907798767, "learning_rate": 3.1949435504376257e-05, "loss": 0.0208, "step": 63670 }, { "grad_norm": 0.3376757502555847, "learning_rate": 3.1934016869625505e-05, "loss": 0.0233, "step": 63680 }, { "grad_norm": 0.4394502341747284, "learning_rate": 3.191860021054349e-05, "loss": 0.0324, "step": 63690 }, { "grad_norm": 0.31130918860435486, "learning_rate": 3.190318552881618e-05, "loss": 0.0305, "step": 63700 }, { "grad_norm": 0.25127485394477844, "learning_rate": 3.1887772826129285e-05, "loss": 0.0191, "step": 63710 }, { "grad_norm": 0.22509290277957916, "learning_rate": 3.1872362104168335e-05, "loss": 0.0198, "step": 63720 }, { "grad_norm": 0.29988160729408264, "learning_rate": 3.185695336461861e-05, "loss": 0.0167, "step": 63730 }, { "grad_norm": 0.17037765681743622, "learning_rate": 3.184154660916519e-05, "loss": 0.0128, "step": 63740 }, { "grad_norm": 0.2918488681316376, "learning_rate": 3.1826141839492906e-05, "loss": 0.0185, "step": 63750 }, { "grad_norm": 0.2961828410625458, "learning_rate": 3.181073905728644e-05, "loss": 0.0193, "step": 63760 }, { "grad_norm": 0.32092854380607605, "learning_rate": 3.1795338264230214e-05, "loss": 0.0155, "step": 63770 }, { "grad_norm": 0.393294095993042, "learning_rate": 3.1779939462008426e-05, "loss": 0.0264, "step": 63780 }, { "grad_norm": 0.3412855565547943, "learning_rate": 3.1764542652305066e-05, "loss": 0.0208, "step": 63790 }, { "grad_norm": 0.23822808265686035, "learning_rate": 3.17491478368039e-05, "loss": 0.019, "step": 63800 }, { "grad_norm": 0.3610702455043793, "learning_rate": 3.173375501718847e-05, "loss": 0.0219, "step": 63810 }, { "grad_norm": 0.3152911961078644, "learning_rate": 3.171836419514215e-05, "loss": 0.0229, "step": 63820 }, { "grad_norm": 0.2673879563808441, "learning_rate": 3.170297537234803e-05, "loss": 0.0208, "step": 63830 }, { "grad_norm": 0.4450437128543854, "learning_rate": 3.168758855048901e-05, "loss": 0.0243, "step": 63840 }, { "grad_norm": 0.2535747289657593, "learning_rate": 3.1672203731247785e-05, "loss": 0.0222, "step": 63850 }, { "grad_norm": 0.2929549515247345, "learning_rate": 3.16568209163068e-05, "loss": 0.0176, "step": 63860 }, { "grad_norm": 0.31281018257141113, "learning_rate": 3.1641440107348295e-05, "loss": 0.0205, "step": 63870 }, { "grad_norm": 0.892566978931427, "learning_rate": 3.162606130605429e-05, "loss": 0.0243, "step": 63880 }, { "grad_norm": 0.2443799078464508, "learning_rate": 3.16106845141066e-05, "loss": 0.0181, "step": 63890 }, { "grad_norm": 0.3000389337539673, "learning_rate": 3.159530973318682e-05, "loss": 0.0189, "step": 63900 }, { "grad_norm": 0.30163681507110596, "learning_rate": 3.1579936964976286e-05, "loss": 0.0182, "step": 63910 }, { "grad_norm": 0.3130345642566681, "learning_rate": 3.156456621115615e-05, "loss": 0.0205, "step": 63920 }, { "grad_norm": 0.27155032753944397, "learning_rate": 3.154919747340734e-05, "loss": 0.0144, "step": 63930 }, { "grad_norm": 0.3902532160282135, "learning_rate": 3.153383075341054e-05, "loss": 0.0223, "step": 63940 }, { "grad_norm": 0.2827689051628113, "learning_rate": 3.1518466052846264e-05, "loss": 0.0205, "step": 63950 }, { "grad_norm": 0.47991400957107544, "learning_rate": 3.150310337339474e-05, "loss": 0.0215, "step": 63960 }, { "grad_norm": 0.21886014938354492, "learning_rate": 3.148774271673603e-05, "loss": 0.0135, "step": 63970 }, { "grad_norm": 0.2951955795288086, "learning_rate": 3.147238408454995e-05, "loss": 0.0255, "step": 63980 }, { "grad_norm": 0.23818644881248474, "learning_rate": 3.1457027478516086e-05, "loss": 0.0201, "step": 63990 }, { "grad_norm": 0.4697161912918091, "learning_rate": 3.144167290031382e-05, "loss": 0.0192, "step": 64000 }, { "grad_norm": 0.26473355293273926, "learning_rate": 3.142632035162227e-05, "loss": 0.0214, "step": 64010 }, { "grad_norm": 0.48734182119369507, "learning_rate": 3.141096983412044e-05, "loss": 0.0174, "step": 64020 }, { "grad_norm": 0.2194158136844635, "learning_rate": 3.139562134948699e-05, "loss": 0.024, "step": 64030 }, { "grad_norm": 0.23382727801799774, "learning_rate": 3.1380274899400424e-05, "loss": 0.0216, "step": 64040 }, { "grad_norm": 0.36785250902175903, "learning_rate": 3.1364930485539e-05, "loss": 0.0158, "step": 64050 }, { "grad_norm": 0.6830514669418335, "learning_rate": 3.134958810958076e-05, "loss": 0.0206, "step": 64060 }, { "grad_norm": 0.3143464922904968, "learning_rate": 3.13342477732035e-05, "loss": 0.0175, "step": 64070 }, { "grad_norm": 0.24133005738258362, "learning_rate": 3.131890947808487e-05, "loss": 0.018, "step": 64080 }, { "grad_norm": 0.28086620569229126, "learning_rate": 3.13035732259022e-05, "loss": 0.0183, "step": 64090 }, { "grad_norm": 0.24580293893814087, "learning_rate": 3.1288239018332656e-05, "loss": 0.0197, "step": 64100 }, { "grad_norm": 0.2730365991592407, "learning_rate": 3.1272906857053164e-05, "loss": 0.0209, "step": 64110 }, { "grad_norm": 0.4326326549053192, "learning_rate": 3.125757674374042e-05, "loss": 0.0221, "step": 64120 }, { "grad_norm": 0.2539856433868408, "learning_rate": 3.124224868007088e-05, "loss": 0.0219, "step": 64130 }, { "grad_norm": 0.3360884487628937, "learning_rate": 3.1226922667720854e-05, "loss": 0.0179, "step": 64140 }, { "grad_norm": 0.368061900138855, "learning_rate": 3.121159870836633e-05, "loss": 0.0246, "step": 64150 }, { "grad_norm": 0.18285803496837616, "learning_rate": 3.119627680368312e-05, "loss": 0.0165, "step": 64160 }, { "grad_norm": 0.24417367577552795, "learning_rate": 3.1180956955346816e-05, "loss": 0.0214, "step": 64170 }, { "grad_norm": 0.2829147279262543, "learning_rate": 3.116563916503276e-05, "loss": 0.0229, "step": 64180 }, { "grad_norm": 0.3899003267288208, "learning_rate": 3.1150323434416095e-05, "loss": 0.0232, "step": 64190 }, { "grad_norm": 0.2841544449329376, "learning_rate": 3.1135009765171696e-05, "loss": 0.0202, "step": 64200 }, { "grad_norm": 0.408877968788147, "learning_rate": 3.111969815897429e-05, "loss": 0.0223, "step": 64210 }, { "grad_norm": 0.3629082441329956, "learning_rate": 3.1104388617498295e-05, "loss": 0.0213, "step": 64220 }, { "grad_norm": 0.29996728897094727, "learning_rate": 3.1089081142417966e-05, "loss": 0.0241, "step": 64230 }, { "grad_norm": 0.2265547513961792, "learning_rate": 3.1073775735407285e-05, "loss": 0.0159, "step": 64240 }, { "grad_norm": 0.5748194456100464, "learning_rate": 3.1058472398140035e-05, "loss": 0.02, "step": 64250 }, { "grad_norm": 0.2533856928348541, "learning_rate": 3.1043171132289736e-05, "loss": 0.0211, "step": 64260 }, { "grad_norm": 0.26151809096336365, "learning_rate": 3.1027871939529774e-05, "loss": 0.0202, "step": 64270 }, { "grad_norm": 0.41758736968040466, "learning_rate": 3.1012574821533214e-05, "loss": 0.0336, "step": 64280 }, { "grad_norm": 0.30512046813964844, "learning_rate": 3.099727977997291e-05, "loss": 0.0221, "step": 64290 }, { "grad_norm": 0.2671513855457306, "learning_rate": 3.098198681652154e-05, "loss": 0.0206, "step": 64300 }, { "grad_norm": 0.3003798723220825, "learning_rate": 3.0966695932851475e-05, "loss": 0.0165, "step": 64310 }, { "grad_norm": 0.2749507427215576, "learning_rate": 3.0951407130634926e-05, "loss": 0.0218, "step": 64320 }, { "grad_norm": 0.313699871301651, "learning_rate": 3.0936120411543864e-05, "loss": 0.0211, "step": 64330 }, { "grad_norm": 0.27936288714408875, "learning_rate": 3.0920835777250015e-05, "loss": 0.0234, "step": 64340 }, { "grad_norm": 0.31538572907447815, "learning_rate": 3.090555322942488e-05, "loss": 0.0162, "step": 64350 }, { "grad_norm": 0.27228766679763794, "learning_rate": 3.089027276973973e-05, "loss": 0.0206, "step": 64360 }, { "grad_norm": 0.4107418656349182, "learning_rate": 3.0874994399865615e-05, "loss": 0.0215, "step": 64370 }, { "grad_norm": 1.179237723350525, "learning_rate": 3.085971812147337e-05, "loss": 0.0281, "step": 64380 }, { "grad_norm": 0.22650252282619476, "learning_rate": 3.0844443936233557e-05, "loss": 0.0211, "step": 64390 }, { "grad_norm": 0.2365972399711609, "learning_rate": 3.082917184581657e-05, "loss": 0.0185, "step": 64400 }, { "grad_norm": 0.482066810131073, "learning_rate": 3.0813901851892513e-05, "loss": 0.0146, "step": 64410 }, { "grad_norm": 0.25832298398017883, "learning_rate": 3.0798633956131316e-05, "loss": 0.0255, "step": 64420 }, { "grad_norm": 0.3377988040447235, "learning_rate": 3.078336816020264e-05, "loss": 0.0174, "step": 64430 }, { "grad_norm": 0.44335857033729553, "learning_rate": 3.0768104465775935e-05, "loss": 0.0219, "step": 64440 }, { "grad_norm": 0.19713230431079865, "learning_rate": 3.0752842874520396e-05, "loss": 0.0236, "step": 64450 }, { "grad_norm": 0.3495037257671356, "learning_rate": 3.073758338810503e-05, "loss": 0.0168, "step": 64460 }, { "grad_norm": 0.24875976145267487, "learning_rate": 3.072232600819859e-05, "loss": 0.0224, "step": 64470 }, { "grad_norm": 1.1985564231872559, "learning_rate": 3.070707073646959e-05, "loss": 0.0214, "step": 64480 }, { "grad_norm": 0.29915115237236023, "learning_rate": 3.069181757458633e-05, "loss": 0.0153, "step": 64490 }, { "grad_norm": 0.2441285103559494, "learning_rate": 3.067656652421687e-05, "loss": 0.0173, "step": 64500 }, { "grad_norm": 0.48097699880599976, "learning_rate": 3.0661317587029025e-05, "loss": 0.022, "step": 64510 }, { "grad_norm": 0.3189336061477661, "learning_rate": 3.064607076469043e-05, "loss": 0.0252, "step": 64520 }, { "grad_norm": 0.36899328231811523, "learning_rate": 3.063082605886845e-05, "loss": 0.0193, "step": 64530 }, { "grad_norm": 0.20401430130004883, "learning_rate": 3.061558347123019e-05, "loss": 0.027, "step": 64540 }, { "grad_norm": 0.48535478115081787, "learning_rate": 3.060034300344259e-05, "loss": 0.0215, "step": 64550 }, { "grad_norm": 0.5013400912284851, "learning_rate": 3.058510465717231e-05, "loss": 0.0273, "step": 64560 }, { "grad_norm": 0.21799765527248383, "learning_rate": 3.0569868434085804e-05, "loss": 0.0211, "step": 64570 }, { "grad_norm": 0.37507808208465576, "learning_rate": 3.055463433584924e-05, "loss": 0.0202, "step": 64580 }, { "grad_norm": 0.317518413066864, "learning_rate": 3.0539402364128656e-05, "loss": 0.0191, "step": 64590 }, { "grad_norm": 0.26891008019447327, "learning_rate": 3.052417252058977e-05, "loss": 0.0154, "step": 64600 }, { "grad_norm": 0.34345346689224243, "learning_rate": 3.0508944806898092e-05, "loss": 0.0261, "step": 64610 }, { "grad_norm": 0.28484296798706055, "learning_rate": 3.04937192247189e-05, "loss": 0.0272, "step": 64620 }, { "grad_norm": 0.20609523355960846, "learning_rate": 3.047849577571725e-05, "loss": 0.0206, "step": 64630 }, { "grad_norm": 0.3320721685886383, "learning_rate": 3.046327446155795e-05, "loss": 0.0247, "step": 64640 }, { "grad_norm": 0.2751416265964508, "learning_rate": 3.0448055283905587e-05, "loss": 0.0203, "step": 64650 }, { "grad_norm": 0.31559130549430847, "learning_rate": 3.0432838244424494e-05, "loss": 0.0246, "step": 64660 }, { "grad_norm": 0.21396881341934204, "learning_rate": 3.04176233447788e-05, "loss": 0.0149, "step": 64670 }, { "grad_norm": 0.2626374363899231, "learning_rate": 3.040241058663238e-05, "loss": 0.0174, "step": 64680 }, { "grad_norm": 0.23605872690677643, "learning_rate": 3.038719997164887e-05, "loss": 0.0236, "step": 64690 }, { "grad_norm": 0.30633705854415894, "learning_rate": 3.0371991501491683e-05, "loss": 0.0266, "step": 64700 }, { "grad_norm": 0.1879899799823761, "learning_rate": 3.0356785177823975e-05, "loss": 0.0227, "step": 64710 }, { "grad_norm": 0.2029908299446106, "learning_rate": 3.0341581002308726e-05, "loss": 0.0185, "step": 64720 }, { "grad_norm": 0.3191395103931427, "learning_rate": 3.0326378976608627e-05, "loss": 0.0155, "step": 64730 }, { "grad_norm": 0.41724538803100586, "learning_rate": 3.0311179102386146e-05, "loss": 0.0269, "step": 64740 }, { "grad_norm": 0.2166086733341217, "learning_rate": 3.0295981381303513e-05, "loss": 0.0143, "step": 64750 }, { "grad_norm": 0.27724993228912354, "learning_rate": 3.0280785815022727e-05, "loss": 0.0207, "step": 64760 }, { "grad_norm": 0.23700985312461853, "learning_rate": 3.0265592405205544e-05, "loss": 0.0174, "step": 64770 }, { "grad_norm": 0.36491885781288147, "learning_rate": 3.0250401153513517e-05, "loss": 0.0215, "step": 64780 }, { "grad_norm": 0.5345734357833862, "learning_rate": 3.0235212061607936e-05, "loss": 0.0303, "step": 64790 }, { "grad_norm": 0.3879144787788391, "learning_rate": 3.0220025131149842e-05, "loss": 0.0191, "step": 64800 }, { "grad_norm": 0.5320960879325867, "learning_rate": 3.020484036380007e-05, "loss": 0.0188, "step": 64810 }, { "grad_norm": 0.2976419925689697, "learning_rate": 3.0189657761219188e-05, "loss": 0.0212, "step": 64820 }, { "grad_norm": 0.2165924459695816, "learning_rate": 3.017447732506753e-05, "loss": 0.0209, "step": 64830 }, { "grad_norm": 0.2803557813167572, "learning_rate": 3.015929905700524e-05, "loss": 0.0179, "step": 64840 }, { "grad_norm": 0.21811780333518982, "learning_rate": 3.014412295869218e-05, "loss": 0.0154, "step": 64850 }, { "grad_norm": 0.33147725462913513, "learning_rate": 3.0128949031787977e-05, "loss": 0.0177, "step": 64860 }, { "grad_norm": 0.22061537206172943, "learning_rate": 3.0113777277952022e-05, "loss": 0.0241, "step": 64870 }, { "grad_norm": 0.2722041606903076, "learning_rate": 3.0098607698843496e-05, "loss": 0.0157, "step": 64880 }, { "grad_norm": 0.24175521731376648, "learning_rate": 3.0083440296121312e-05, "loss": 0.0143, "step": 64890 }, { "grad_norm": 0.8034968376159668, "learning_rate": 3.0068275071444142e-05, "loss": 0.0186, "step": 64900 }, { "grad_norm": 0.490418016910553, "learning_rate": 3.0053112026470442e-05, "loss": 0.0311, "step": 64910 }, { "grad_norm": 0.272786945104599, "learning_rate": 3.0037951162858434e-05, "loss": 0.0269, "step": 64920 }, { "grad_norm": 0.2564142048358917, "learning_rate": 3.002279248226607e-05, "loss": 0.0195, "step": 64930 }, { "grad_norm": 0.2663421332836151, "learning_rate": 3.000763598635109e-05, "loss": 0.0178, "step": 64940 }, { "grad_norm": 0.355263352394104, "learning_rate": 2.999248167677098e-05, "loss": 0.0174, "step": 64950 }, { "grad_norm": 0.2785555124282837, "learning_rate": 2.9977329555182976e-05, "loss": 0.0187, "step": 64960 }, { "grad_norm": 0.29746007919311523, "learning_rate": 2.9962179623244124e-05, "loss": 0.0186, "step": 64970 }, { "grad_norm": 0.2358151376247406, "learning_rate": 2.9947031882611175e-05, "loss": 0.0182, "step": 64980 }, { "grad_norm": 0.28001415729522705, "learning_rate": 2.9931886334940685e-05, "loss": 0.0172, "step": 64990 }, { "grad_norm": 0.38611099123954773, "learning_rate": 2.9916742981888923e-05, "loss": 0.0196, "step": 65000 }, { "grad_norm": 0.3766859471797943, "learning_rate": 2.990160182511196e-05, "loss": 0.0256, "step": 65010 }, { "grad_norm": 0.42256760597229004, "learning_rate": 2.9886462866265574e-05, "loss": 0.0159, "step": 65020 }, { "grad_norm": 0.2686053216457367, "learning_rate": 2.9871326107005392e-05, "loss": 0.0206, "step": 65030 }, { "grad_norm": 0.3282708525657654, "learning_rate": 2.9856191548986735e-05, "loss": 0.0204, "step": 65040 }, { "grad_norm": 0.27501630783081055, "learning_rate": 2.9841059193864672e-05, "loss": 0.0232, "step": 65050 }, { "grad_norm": 0.2560320198535919, "learning_rate": 2.982592904329407e-05, "loss": 0.0247, "step": 65060 }, { "grad_norm": 0.23617644608020782, "learning_rate": 2.981080109892954e-05, "loss": 0.0201, "step": 65070 }, { "grad_norm": 0.1775086671113968, "learning_rate": 2.9795675362425434e-05, "loss": 0.0159, "step": 65080 }, { "grad_norm": 0.22098298370838165, "learning_rate": 2.97805518354359e-05, "loss": 0.0171, "step": 65090 }, { "grad_norm": 0.21316760778427124, "learning_rate": 2.9765430519614825e-05, "loss": 0.0154, "step": 65100 }, { "grad_norm": 0.21596619486808777, "learning_rate": 2.975031141661584e-05, "loss": 0.0211, "step": 65110 }, { "grad_norm": 0.29663291573524475, "learning_rate": 2.9735194528092347e-05, "loss": 0.0189, "step": 65120 }, { "grad_norm": 0.26638463139533997, "learning_rate": 2.9720079855697525e-05, "loss": 0.026, "step": 65130 }, { "grad_norm": 0.28754422068595886, "learning_rate": 2.970496740108427e-05, "loss": 0.0248, "step": 65140 }, { "grad_norm": 0.12582121789455414, "learning_rate": 2.968985716590527e-05, "loss": 0.0137, "step": 65150 }, { "grad_norm": 0.3459341526031494, "learning_rate": 2.9674749151812953e-05, "loss": 0.0186, "step": 65160 }, { "grad_norm": 0.28054213523864746, "learning_rate": 2.965964336045951e-05, "loss": 0.0186, "step": 65170 }, { "grad_norm": 0.18860523402690887, "learning_rate": 2.9644539793496894e-05, "loss": 0.0223, "step": 65180 }, { "grad_norm": 0.2555910050868988, "learning_rate": 2.9629438452576807e-05, "loss": 0.0174, "step": 65190 }, { "grad_norm": 0.27399885654449463, "learning_rate": 2.9614339339350698e-05, "loss": 0.023, "step": 65200 }, { "grad_norm": 0.25229835510253906, "learning_rate": 2.959924245546979e-05, "loss": 0.0177, "step": 65210 }, { "grad_norm": 0.317234605550766, "learning_rate": 2.9584147802585038e-05, "loss": 0.0283, "step": 65220 }, { "grad_norm": 0.7249207496643066, "learning_rate": 2.956905538234721e-05, "loss": 0.0197, "step": 65230 }, { "grad_norm": 0.2517150938510895, "learning_rate": 2.9553965196406764e-05, "loss": 0.021, "step": 65240 }, { "grad_norm": 0.3113192915916443, "learning_rate": 2.9538877246413943e-05, "loss": 0.0177, "step": 65250 }, { "grad_norm": 0.29257622361183167, "learning_rate": 2.952379153401874e-05, "loss": 0.0228, "step": 65260 }, { "grad_norm": 0.36041060090065, "learning_rate": 2.9508708060870914e-05, "loss": 0.0207, "step": 65270 }, { "grad_norm": 0.23102647066116333, "learning_rate": 2.9493626828619948e-05, "loss": 0.0158, "step": 65280 }, { "grad_norm": 0.26530539989471436, "learning_rate": 2.9478547838915137e-05, "loss": 0.0244, "step": 65290 }, { "grad_norm": 0.2757171094417572, "learning_rate": 2.946347109340548e-05, "loss": 0.0165, "step": 65300 }, { "grad_norm": 0.2669827938079834, "learning_rate": 2.944839659373976e-05, "loss": 0.0215, "step": 65310 }, { "grad_norm": 0.3716092109680176, "learning_rate": 2.943332434156647e-05, "loss": 0.0201, "step": 65320 }, { "grad_norm": 0.39151522517204285, "learning_rate": 2.9418254338533912e-05, "loss": 0.0241, "step": 65330 }, { "grad_norm": 0.2620571255683899, "learning_rate": 2.940318658629011e-05, "loss": 0.0208, "step": 65340 }, { "grad_norm": 0.15916651487350464, "learning_rate": 2.938812108648287e-05, "loss": 0.0155, "step": 65350 }, { "grad_norm": 0.3069498538970947, "learning_rate": 2.937305784075972e-05, "loss": 0.02, "step": 65360 }, { "grad_norm": 0.2498091757297516, "learning_rate": 2.9357996850767937e-05, "loss": 0.0163, "step": 65370 }, { "grad_norm": 0.38179826736450195, "learning_rate": 2.9342938118154595e-05, "loss": 0.0162, "step": 65380 }, { "grad_norm": 0.23696278035640717, "learning_rate": 2.9327881644566485e-05, "loss": 0.0232, "step": 65390 }, { "grad_norm": 0.31971827149391174, "learning_rate": 2.9312827431650157e-05, "loss": 0.0212, "step": 65400 }, { "grad_norm": 0.20887064933776855, "learning_rate": 2.92977754810519e-05, "loss": 0.0193, "step": 65410 }, { "grad_norm": 0.3171866834163666, "learning_rate": 2.928272579441781e-05, "loss": 0.0152, "step": 65420 }, { "grad_norm": 0.24704140424728394, "learning_rate": 2.9267678373393676e-05, "loss": 0.0236, "step": 65430 }, { "grad_norm": 0.251842200756073, "learning_rate": 2.925263321962507e-05, "loss": 0.0241, "step": 65440 }, { "grad_norm": 0.2574668228626251, "learning_rate": 2.9237590334757292e-05, "loss": 0.0218, "step": 65450 }, { "grad_norm": 0.4343648850917816, "learning_rate": 2.922254972043543e-05, "loss": 0.0181, "step": 65460 }, { "grad_norm": 0.2674936056137085, "learning_rate": 2.9207511378304265e-05, "loss": 0.0176, "step": 65470 }, { "grad_norm": 0.2742835581302643, "learning_rate": 2.9192475310008428e-05, "loss": 0.0197, "step": 65480 }, { "grad_norm": 0.1599225550889969, "learning_rate": 2.917744151719219e-05, "loss": 0.0133, "step": 65490 }, { "grad_norm": 0.27443715929985046, "learning_rate": 2.916241000149964e-05, "loss": 0.0255, "step": 65500 }, { "grad_norm": 0.29650747776031494, "learning_rate": 2.9147380764574622e-05, "loss": 0.0146, "step": 65510 }, { "grad_norm": 0.3453841507434845, "learning_rate": 2.9132353808060674e-05, "loss": 0.0306, "step": 65520 }, { "grad_norm": 0.2669091820716858, "learning_rate": 2.9117329133601134e-05, "loss": 0.0289, "step": 65530 }, { "grad_norm": 0.20539498329162598, "learning_rate": 2.9102306742839093e-05, "loss": 0.016, "step": 65540 }, { "grad_norm": 0.2476949840784073, "learning_rate": 2.9087286637417375e-05, "loss": 0.0169, "step": 65550 }, { "grad_norm": 0.3032086491584778, "learning_rate": 2.9072268818978537e-05, "loss": 0.0303, "step": 65560 }, { "grad_norm": 0.3078829050064087, "learning_rate": 2.905725328916493e-05, "loss": 0.0223, "step": 65570 }, { "grad_norm": 0.22559118270874023, "learning_rate": 2.9042240049618608e-05, "loss": 0.0227, "step": 65580 }, { "grad_norm": 0.24548836052417755, "learning_rate": 2.9027229101981413e-05, "loss": 0.0223, "step": 65590 }, { "grad_norm": 0.23162442445755005, "learning_rate": 2.901222044789488e-05, "loss": 0.0208, "step": 65600 }, { "grad_norm": 0.3275775611400604, "learning_rate": 2.89972140890004e-05, "loss": 0.0191, "step": 65610 }, { "grad_norm": 0.39694878458976746, "learning_rate": 2.8982210026938988e-05, "loss": 0.0236, "step": 65620 }, { "grad_norm": 0.31212490797042847, "learning_rate": 2.896720826335151e-05, "loss": 0.0313, "step": 65630 }, { "grad_norm": 0.243242084980011, "learning_rate": 2.8952208799878493e-05, "loss": 0.019, "step": 65640 }, { "grad_norm": 0.2890620231628418, "learning_rate": 2.89372116381603e-05, "loss": 0.0172, "step": 65650 }, { "grad_norm": 0.24902142584323883, "learning_rate": 2.8922216779836932e-05, "loss": 0.016, "step": 65660 }, { "grad_norm": 0.286579966545105, "learning_rate": 2.8907224226548285e-05, "loss": 0.0203, "step": 65670 }, { "grad_norm": 0.430766224861145, "learning_rate": 2.8892233979933868e-05, "loss": 0.0211, "step": 65680 }, { "grad_norm": 0.300371915102005, "learning_rate": 2.8877246041633026e-05, "loss": 0.016, "step": 65690 }, { "grad_norm": 0.22616346180438995, "learning_rate": 2.886226041328478e-05, "loss": 0.0141, "step": 65700 }, { "grad_norm": 0.23020599782466888, "learning_rate": 2.8847277096527957e-05, "loss": 0.0186, "step": 65710 }, { "grad_norm": 0.2708148956298828, "learning_rate": 2.8832296093001105e-05, "loss": 0.0166, "step": 65720 }, { "grad_norm": 0.37674179673194885, "learning_rate": 2.8817317404342547e-05, "loss": 0.0251, "step": 65730 }, { "grad_norm": 0.6088005900382996, "learning_rate": 2.880234103219029e-05, "loss": 0.0181, "step": 65740 }, { "grad_norm": 0.30796003341674805, "learning_rate": 2.878736697818215e-05, "loss": 0.0178, "step": 65750 }, { "grad_norm": 0.23348791897296906, "learning_rate": 2.877239524395568e-05, "loss": 0.0183, "step": 65760 }, { "grad_norm": 0.2590622007846832, "learning_rate": 2.8757425831148134e-05, "loss": 0.0204, "step": 65770 }, { "grad_norm": 0.18338482081890106, "learning_rate": 2.874245874139658e-05, "loss": 0.02, "step": 65780 }, { "grad_norm": 0.17346379160881042, "learning_rate": 2.8727493976337737e-05, "loss": 0.0198, "step": 65790 }, { "grad_norm": 0.30561476945877075, "learning_rate": 2.8712531537608218e-05, "loss": 0.0153, "step": 65800 }, { "grad_norm": 0.24705009162425995, "learning_rate": 2.8697571426844217e-05, "loss": 0.0161, "step": 65810 }, { "grad_norm": 0.17480109632015228, "learning_rate": 2.86826136456818e-05, "loss": 0.0178, "step": 65820 }, { "grad_norm": 0.3081245422363281, "learning_rate": 2.8667658195756685e-05, "loss": 0.0184, "step": 65830 }, { "grad_norm": 0.25643017888069153, "learning_rate": 2.8652705078704412e-05, "loss": 0.024, "step": 65840 }, { "grad_norm": 0.3256997764110565, "learning_rate": 2.863775429616018e-05, "loss": 0.0189, "step": 65850 }, { "grad_norm": 0.2627318799495697, "learning_rate": 2.862280584975906e-05, "loss": 0.0143, "step": 65860 }, { "grad_norm": 0.38685107231140137, "learning_rate": 2.8607859741135735e-05, "loss": 0.018, "step": 65870 }, { "grad_norm": 0.3185310661792755, "learning_rate": 2.8592915971924723e-05, "loss": 0.0278, "step": 65880 }, { "grad_norm": 0.3043217658996582, "learning_rate": 2.8577974543760212e-05, "loss": 0.0202, "step": 65890 }, { "grad_norm": 0.19904443621635437, "learning_rate": 2.8563035458276222e-05, "loss": 0.0179, "step": 65900 }, { "grad_norm": 0.18386240303516388, "learning_rate": 2.8548098717106424e-05, "loss": 0.0236, "step": 65910 }, { "grad_norm": 0.16287164390087128, "learning_rate": 2.85331643218843e-05, "loss": 0.0157, "step": 65920 }, { "grad_norm": 0.6902846693992615, "learning_rate": 2.8518232274243046e-05, "loss": 0.0238, "step": 65930 }, { "grad_norm": 0.39399439096450806, "learning_rate": 2.8503302575815638e-05, "loss": 0.0246, "step": 65940 }, { "grad_norm": 0.38830116391181946, "learning_rate": 2.8488375228234716e-05, "loss": 0.0224, "step": 65950 }, { "grad_norm": 0.3971133828163147, "learning_rate": 2.8473450233132738e-05, "loss": 0.0234, "step": 65960 }, { "grad_norm": 0.24807465076446533, "learning_rate": 2.8458527592141898e-05, "loss": 0.0146, "step": 65970 }, { "grad_norm": 0.2785507142543793, "learning_rate": 2.8443607306894073e-05, "loss": 0.0271, "step": 65980 }, { "grad_norm": 0.32592251896858215, "learning_rate": 2.8428689379020946e-05, "loss": 0.0212, "step": 65990 }, { "grad_norm": 0.3693418800830841, "learning_rate": 2.841377381015391e-05, "loss": 0.0182, "step": 66000 }, { "grad_norm": 0.24205225706100464, "learning_rate": 2.8398860601924144e-05, "loss": 0.0205, "step": 66010 }, { "grad_norm": 0.2832408547401428, "learning_rate": 2.838394975596248e-05, "loss": 0.0134, "step": 66020 }, { "grad_norm": 0.246765598654747, "learning_rate": 2.83690412738996e-05, "loss": 0.021, "step": 66030 }, { "grad_norm": 0.2742258906364441, "learning_rate": 2.835413515736579e-05, "loss": 0.0213, "step": 66040 }, { "grad_norm": 0.3445776104927063, "learning_rate": 2.8339231407991267e-05, "loss": 0.0209, "step": 66050 }, { "grad_norm": 0.1938779652118683, "learning_rate": 2.832433002740581e-05, "loss": 0.0145, "step": 66060 }, { "grad_norm": 0.21135437488555908, "learning_rate": 2.8309431017239047e-05, "loss": 0.019, "step": 66070 }, { "grad_norm": 0.2808440327644348, "learning_rate": 2.829453437912028e-05, "loss": 0.024, "step": 66080 }, { "grad_norm": 0.2829906940460205, "learning_rate": 2.8279640114678617e-05, "loss": 0.018, "step": 66090 }, { "grad_norm": 0.2058321237564087, "learning_rate": 2.8264748225542836e-05, "loss": 0.0188, "step": 66100 }, { "grad_norm": 0.2679795026779175, "learning_rate": 2.8249858713341516e-05, "loss": 0.0175, "step": 66110 }, { "grad_norm": 0.3971303403377533, "learning_rate": 2.823497157970294e-05, "loss": 0.0176, "step": 66120 }, { "grad_norm": 0.41787442564964294, "learning_rate": 2.8220086826255164e-05, "loss": 0.0217, "step": 66130 }, { "grad_norm": 0.21527792513370514, "learning_rate": 2.8205204454625934e-05, "loss": 0.0182, "step": 66140 }, { "grad_norm": 0.27435848116874695, "learning_rate": 2.8190324466442798e-05, "loss": 0.0184, "step": 66150 }, { "grad_norm": 0.24037234485149384, "learning_rate": 2.8175446863332965e-05, "loss": 0.021, "step": 66160 }, { "grad_norm": 0.3236182630062103, "learning_rate": 2.816057164692345e-05, "loss": 0.0208, "step": 66170 }, { "grad_norm": 0.20476894080638885, "learning_rate": 2.8145698818840992e-05, "loss": 0.0209, "step": 66180 }, { "grad_norm": 0.31281399726867676, "learning_rate": 2.8130828380712065e-05, "loss": 0.0213, "step": 66190 }, { "grad_norm": 0.3127630054950714, "learning_rate": 2.811596033416285e-05, "loss": 0.0173, "step": 66200 }, { "grad_norm": 0.2840817868709564, "learning_rate": 2.8101094680819307e-05, "loss": 0.0198, "step": 66210 }, { "grad_norm": 0.30739182233810425, "learning_rate": 2.808623142230715e-05, "loss": 0.0234, "step": 66220 }, { "grad_norm": 0.362857848405838, "learning_rate": 2.8071370560251757e-05, "loss": 0.0217, "step": 66230 }, { "grad_norm": 0.384921133518219, "learning_rate": 2.805651209627831e-05, "loss": 0.0193, "step": 66240 }, { "grad_norm": 0.31128209829330444, "learning_rate": 2.804165603201171e-05, "loss": 0.0189, "step": 66250 }, { "grad_norm": 0.2317395955324173, "learning_rate": 2.802680236907661e-05, "loss": 0.0229, "step": 66260 }, { "grad_norm": 0.3368068337440491, "learning_rate": 2.8011951109097344e-05, "loss": 0.0179, "step": 66270 }, { "grad_norm": 0.30166032910346985, "learning_rate": 2.7997102253698072e-05, "loss": 0.022, "step": 66280 }, { "grad_norm": 0.2473183423280716, "learning_rate": 2.7982255804502588e-05, "loss": 0.0173, "step": 66290 }, { "grad_norm": 0.26668235659599304, "learning_rate": 2.7967411763134516e-05, "loss": 0.0191, "step": 66300 }, { "grad_norm": 0.4222175180912018, "learning_rate": 2.7952570131217166e-05, "loss": 0.0218, "step": 66310 }, { "grad_norm": 0.3712581992149353, "learning_rate": 2.7937730910373618e-05, "loss": 0.0227, "step": 66320 }, { "grad_norm": 0.4136229157447815, "learning_rate": 2.7922894102226628e-05, "loss": 0.0204, "step": 66330 }, { "grad_norm": 0.18806201219558716, "learning_rate": 2.790805970839876e-05, "loss": 0.0242, "step": 66340 }, { "grad_norm": 0.6623275279998779, "learning_rate": 2.789322773051225e-05, "loss": 0.0219, "step": 66350 }, { "grad_norm": 0.4612981975078583, "learning_rate": 2.7878398170189124e-05, "loss": 0.0243, "step": 66360 }, { "grad_norm": 0.19373264908790588, "learning_rate": 2.7863571029051112e-05, "loss": 0.019, "step": 66370 }, { "grad_norm": 0.3566507399082184, "learning_rate": 2.784874630871971e-05, "loss": 0.025, "step": 66380 }, { "grad_norm": 0.3623688519001007, "learning_rate": 2.7833924010816086e-05, "loss": 0.0133, "step": 66390 }, { "grad_norm": 0.20918600261211395, "learning_rate": 2.7819104136961226e-05, "loss": 0.0213, "step": 66400 }, { "grad_norm": 0.23353956639766693, "learning_rate": 2.780428668877577e-05, "loss": 0.0184, "step": 66410 }, { "grad_norm": 0.21738998591899872, "learning_rate": 2.7789471667880152e-05, "loss": 0.0188, "step": 66420 }, { "grad_norm": 0.1667342483997345, "learning_rate": 2.7774659075894523e-05, "loss": 0.0218, "step": 66430 }, { "grad_norm": 0.21236690878868103, "learning_rate": 2.7759848914438778e-05, "loss": 0.019, "step": 66440 }, { "grad_norm": 0.21788007020950317, "learning_rate": 2.7745041185132496e-05, "loss": 0.0125, "step": 66450 }, { "grad_norm": 0.31898170709609985, "learning_rate": 2.773023588959506e-05, "loss": 0.0228, "step": 66460 }, { "grad_norm": 0.316967636346817, "learning_rate": 2.7715433029445555e-05, "loss": 0.023, "step": 66470 }, { "grad_norm": 0.26384755969047546, "learning_rate": 2.7700632606302766e-05, "loss": 0.0181, "step": 66480 }, { "grad_norm": 0.625370442867279, "learning_rate": 2.768583462178528e-05, "loss": 0.0235, "step": 66490 }, { "grad_norm": 0.20917117595672607, "learning_rate": 2.767103907751136e-05, "loss": 0.0147, "step": 66500 }, { "grad_norm": 0.2646182179450989, "learning_rate": 2.7656245975099065e-05, "loss": 0.0172, "step": 66510 }, { "grad_norm": 0.4711340069770813, "learning_rate": 2.7641455316166093e-05, "loss": 0.0194, "step": 66520 }, { "grad_norm": 0.16348391771316528, "learning_rate": 2.7626667102329973e-05, "loss": 0.0197, "step": 66530 }, { "grad_norm": 0.31912970542907715, "learning_rate": 2.761188133520788e-05, "loss": 0.02, "step": 66540 }, { "grad_norm": 0.21402058005332947, "learning_rate": 2.7597098016416784e-05, "loss": 0.0153, "step": 66550 }, { "grad_norm": 0.24727757275104523, "learning_rate": 2.7582317147573368e-05, "loss": 0.0154, "step": 66560 }, { "grad_norm": 0.2314518690109253, "learning_rate": 2.7567538730294064e-05, "loss": 0.0256, "step": 66570 }, { "grad_norm": 0.21787585318088531, "learning_rate": 2.7552762766194973e-05, "loss": 0.0161, "step": 66580 }, { "grad_norm": 0.37699228525161743, "learning_rate": 2.753798925689202e-05, "loss": 0.0265, "step": 66590 }, { "grad_norm": 0.3082462251186371, "learning_rate": 2.7523218204000774e-05, "loss": 0.0192, "step": 66600 }, { "grad_norm": 0.3285446763038635, "learning_rate": 2.7508449609136612e-05, "loss": 0.0242, "step": 66610 }, { "grad_norm": 0.2869502305984497, "learning_rate": 2.749368347391454e-05, "loss": 0.0175, "step": 66620 }, { "grad_norm": 0.27527716755867004, "learning_rate": 2.7478919799949443e-05, "loss": 0.0167, "step": 66630 }, { "grad_norm": 0.2634907364845276, "learning_rate": 2.7464158588855803e-05, "loss": 0.0156, "step": 66640 }, { "grad_norm": 0.33669137954711914, "learning_rate": 2.7449399842247918e-05, "loss": 0.0198, "step": 66650 }, { "grad_norm": 0.2986108660697937, "learning_rate": 2.7434643561739736e-05, "loss": 0.0195, "step": 66660 }, { "grad_norm": 0.3893348276615143, "learning_rate": 2.7419889748945016e-05, "loss": 0.0203, "step": 66670 }, { "grad_norm": 0.4447568356990814, "learning_rate": 2.7405138405477203e-05, "loss": 0.0206, "step": 66680 }, { "grad_norm": 0.30826810002326965, "learning_rate": 2.73903895329495e-05, "loss": 0.0195, "step": 66690 }, { "grad_norm": 0.1937275230884552, "learning_rate": 2.737564313297478e-05, "loss": 0.0205, "step": 66700 }, { "grad_norm": 0.28530153632164, "learning_rate": 2.736089920716572e-05, "loss": 0.0176, "step": 66710 }, { "grad_norm": 0.4100198745727539, "learning_rate": 2.7346157757134706e-05, "loss": 0.0208, "step": 66720 }, { "grad_norm": 0.23122097551822662, "learning_rate": 2.733141878449379e-05, "loss": 0.0164, "step": 66730 }, { "grad_norm": 0.15524771809577942, "learning_rate": 2.7316682290854844e-05, "loss": 0.0187, "step": 66740 }, { "grad_norm": 0.5763899087905884, "learning_rate": 2.7301948277829414e-05, "loss": 0.0216, "step": 66750 }, { "grad_norm": 0.3707168698310852, "learning_rate": 2.7287216747028814e-05, "loss": 0.0181, "step": 66760 }, { "grad_norm": 0.18931950628757477, "learning_rate": 2.7272487700064025e-05, "loss": 0.0173, "step": 66770 }, { "grad_norm": 0.22070752084255219, "learning_rate": 2.725776113854583e-05, "loss": 0.0202, "step": 66780 }, { "grad_norm": 0.3527454435825348, "learning_rate": 2.7243037064084664e-05, "loss": 0.0188, "step": 66790 }, { "grad_norm": 0.5085651278495789, "learning_rate": 2.7228315478290766e-05, "loss": 0.0298, "step": 66800 }, { "grad_norm": 0.36422109603881836, "learning_rate": 2.721359638277402e-05, "loss": 0.0213, "step": 66810 }, { "grad_norm": 0.2302241027355194, "learning_rate": 2.7198879779144148e-05, "loss": 0.0183, "step": 66820 }, { "grad_norm": 0.4599485695362091, "learning_rate": 2.7184165669010485e-05, "loss": 0.0157, "step": 66830 }, { "grad_norm": 0.2781764268875122, "learning_rate": 2.716945405398218e-05, "loss": 0.0168, "step": 66840 }, { "grad_norm": 0.31918877363204956, "learning_rate": 2.715474493566803e-05, "loss": 0.0205, "step": 66850 }, { "grad_norm": 0.16965092718601227, "learning_rate": 2.7140038315676652e-05, "loss": 0.019, "step": 66860 }, { "grad_norm": 0.391759991645813, "learning_rate": 2.7125334195616275e-05, "loss": 0.0321, "step": 66870 }, { "grad_norm": 0.23777291178703308, "learning_rate": 2.7110632577094997e-05, "loss": 0.0185, "step": 66880 }, { "grad_norm": 0.3211301863193512, "learning_rate": 2.709593346172051e-05, "loss": 0.0216, "step": 66890 }, { "grad_norm": 0.29351767897605896, "learning_rate": 2.708123685110032e-05, "loss": 0.0194, "step": 66900 }, { "grad_norm": 0.5315475463867188, "learning_rate": 2.7066542746841588e-05, "loss": 0.0249, "step": 66910 }, { "grad_norm": 0.5643700957298279, "learning_rate": 2.7051851150551267e-05, "loss": 0.016, "step": 66920 }, { "grad_norm": 0.18966513872146606, "learning_rate": 2.7037162063835996e-05, "loss": 0.0145, "step": 66930 }, { "grad_norm": 0.24928557872772217, "learning_rate": 2.702247548830218e-05, "loss": 0.0125, "step": 66940 }, { "grad_norm": 0.18446961045265198, "learning_rate": 2.7007791425555883e-05, "loss": 0.017, "step": 66950 }, { "grad_norm": 0.31671932339668274, "learning_rate": 2.6993109877202945e-05, "loss": 0.0243, "step": 66960 }, { "grad_norm": 0.2605171501636505, "learning_rate": 2.6978430844848935e-05, "loss": 0.0193, "step": 66970 }, { "grad_norm": 0.3454415798187256, "learning_rate": 2.69637543300991e-05, "loss": 0.016, "step": 66980 }, { "grad_norm": 0.30440202355384827, "learning_rate": 2.694908033455848e-05, "loss": 0.016, "step": 66990 }, { "grad_norm": 0.27188733220100403, "learning_rate": 2.6934408859831733e-05, "loss": 0.0195, "step": 67000 }, { "grad_norm": 0.2582627236843109, "learning_rate": 2.69197399075234e-05, "loss": 0.0195, "step": 67010 }, { "grad_norm": 0.3194962739944458, "learning_rate": 2.6905073479237584e-05, "loss": 0.0192, "step": 67020 }, { "grad_norm": 0.26897403597831726, "learning_rate": 2.6890409576578235e-05, "loss": 0.0251, "step": 67030 }, { "grad_norm": 0.27959564328193665, "learning_rate": 2.687574820114892e-05, "loss": 0.0187, "step": 67040 }, { "grad_norm": 0.46283018589019775, "learning_rate": 2.686108935455305e-05, "loss": 0.0344, "step": 67050 }, { "grad_norm": 0.4219576120376587, "learning_rate": 2.6846433038393603e-05, "loss": 0.0203, "step": 67060 }, { "grad_norm": 0.30097466707229614, "learning_rate": 2.6831779254273482e-05, "loss": 0.0178, "step": 67070 }, { "grad_norm": 0.3789542317390442, "learning_rate": 2.6817128003795122e-05, "loss": 0.0159, "step": 67080 }, { "grad_norm": 0.17314326763153076, "learning_rate": 2.680247928856081e-05, "loss": 0.0126, "step": 67090 }, { "grad_norm": 0.3743618130683899, "learning_rate": 2.6787833110172466e-05, "loss": 0.015, "step": 67100 }, { "grad_norm": 0.1850036233663559, "learning_rate": 2.6773189470231807e-05, "loss": 0.0249, "step": 67110 }, { "grad_norm": 0.3078375458717346, "learning_rate": 2.675854837034022e-05, "loss": 0.018, "step": 67120 }, { "grad_norm": 0.15259893238544464, "learning_rate": 2.6743909812098833e-05, "loss": 0.0211, "step": 67130 }, { "grad_norm": 0.33093541860580444, "learning_rate": 2.67292737971085e-05, "loss": 0.024, "step": 67140 }, { "grad_norm": 0.281282514333725, "learning_rate": 2.671464032696982e-05, "loss": 0.0141, "step": 67150 }, { "grad_norm": 0.5696012377738953, "learning_rate": 2.6700009403283055e-05, "loss": 0.0204, "step": 67160 }, { "grad_norm": 0.2634124159812927, "learning_rate": 2.6685381027648214e-05, "loss": 0.018, "step": 67170 }, { "grad_norm": 0.36049050092697144, "learning_rate": 2.6670755201665086e-05, "loss": 0.0308, "step": 67180 }, { "grad_norm": 0.3088938593864441, "learning_rate": 2.6656131926933066e-05, "loss": 0.0191, "step": 67190 }, { "grad_norm": 0.29969322681427, "learning_rate": 2.664151120505136e-05, "loss": 0.0162, "step": 67200 }, { "grad_norm": 0.32893311977386475, "learning_rate": 2.6626893037618873e-05, "loss": 0.0202, "step": 67210 }, { "grad_norm": 0.5832938551902771, "learning_rate": 2.6612277426234243e-05, "loss": 0.0247, "step": 67220 }, { "grad_norm": 0.2924582064151764, "learning_rate": 2.659766437249577e-05, "loss": 0.0266, "step": 67230 }, { "grad_norm": 0.2655043601989746, "learning_rate": 2.6583053878001557e-05, "loss": 0.0143, "step": 67240 }, { "grad_norm": 0.40267419815063477, "learning_rate": 2.6568445944349323e-05, "loss": 0.0208, "step": 67250 }, { "grad_norm": 0.24795979261398315, "learning_rate": 2.6553840573136657e-05, "loss": 0.0161, "step": 67260 }, { "grad_norm": 0.6179730892181396, "learning_rate": 2.6539237765960706e-05, "loss": 0.0228, "step": 67270 }, { "grad_norm": 0.36550647020339966, "learning_rate": 2.6524637524418465e-05, "loss": 0.0203, "step": 67280 }, { "grad_norm": 0.21759845316410065, "learning_rate": 2.651003985010655e-05, "loss": 0.014, "step": 67290 }, { "grad_norm": 0.39666882157325745, "learning_rate": 2.649544474462138e-05, "loss": 0.0175, "step": 67300 }, { "grad_norm": 0.2672281861305237, "learning_rate": 2.6480852209559015e-05, "loss": 0.0158, "step": 67310 }, { "grad_norm": 0.2155960202217102, "learning_rate": 2.6466262246515296e-05, "loss": 0.0207, "step": 67320 }, { "grad_norm": 0.283743679523468, "learning_rate": 2.6451674857085746e-05, "loss": 0.0182, "step": 67330 }, { "grad_norm": 0.2532055377960205, "learning_rate": 2.6437090042865655e-05, "loss": 0.0221, "step": 67340 }, { "grad_norm": 0.24829965829849243, "learning_rate": 2.6422507805449947e-05, "loss": 0.0157, "step": 67350 }, { "grad_norm": 0.32365554571151733, "learning_rate": 2.6407928146433357e-05, "loss": 0.0227, "step": 67360 }, { "grad_norm": 0.28005489706993103, "learning_rate": 2.639335106741026e-05, "loss": 0.0163, "step": 67370 }, { "grad_norm": 0.30480918288230896, "learning_rate": 2.6378776569974796e-05, "loss": 0.0211, "step": 67380 }, { "grad_norm": 0.3340696394443512, "learning_rate": 2.6364204655720814e-05, "loss": 0.0151, "step": 67390 }, { "grad_norm": 0.37509429454803467, "learning_rate": 2.6349635326241894e-05, "loss": 0.0213, "step": 67400 }, { "grad_norm": 0.24653995037078857, "learning_rate": 2.633506858313127e-05, "loss": 0.0187, "step": 67410 }, { "grad_norm": 0.29062575101852417, "learning_rate": 2.6320504427981975e-05, "loss": 0.0274, "step": 67420 }, { "grad_norm": 0.33286717534065247, "learning_rate": 2.6305942862386734e-05, "loss": 0.015, "step": 67430 }, { "grad_norm": 0.34480270743370056, "learning_rate": 2.629138388793793e-05, "loss": 0.0159, "step": 67440 }, { "grad_norm": 0.8653836250305176, "learning_rate": 2.6276827506227743e-05, "loss": 0.0236, "step": 67450 }, { "grad_norm": 0.2769213914871216, "learning_rate": 2.6262273718848037e-05, "loss": 0.0156, "step": 67460 }, { "grad_norm": 0.2902722656726837, "learning_rate": 2.6247722527390407e-05, "loss": 0.0172, "step": 67470 }, { "grad_norm": 0.22332195937633514, "learning_rate": 2.6233173933446104e-05, "loss": 0.0207, "step": 67480 }, { "grad_norm": 0.21260884404182434, "learning_rate": 2.6218627938606188e-05, "loss": 0.0291, "step": 67490 }, { "grad_norm": 0.2686167359352112, "learning_rate": 2.620408454446135e-05, "loss": 0.0176, "step": 67500 }, { "grad_norm": 0.22395730018615723, "learning_rate": 2.6189543752602046e-05, "loss": 0.0226, "step": 67510 }, { "grad_norm": 0.3137400448322296, "learning_rate": 2.6175005564618444e-05, "loss": 0.0194, "step": 67520 }, { "grad_norm": 0.3504070043563843, "learning_rate": 2.6160469982100428e-05, "loss": 0.0226, "step": 67530 }, { "grad_norm": 0.41251707077026367, "learning_rate": 2.6145937006637555e-05, "loss": 0.0202, "step": 67540 }, { "grad_norm": 0.40183010697364807, "learning_rate": 2.6131406639819166e-05, "loss": 0.0226, "step": 67550 }, { "grad_norm": 0.6934511661529541, "learning_rate": 2.611687888323424e-05, "loss": 0.0227, "step": 67560 }, { "grad_norm": 0.2566652297973633, "learning_rate": 2.610235373847153e-05, "loss": 0.0232, "step": 67570 }, { "grad_norm": 0.34679076075553894, "learning_rate": 2.6087831207119494e-05, "loss": 0.0276, "step": 67580 }, { "grad_norm": 0.22266243398189545, "learning_rate": 2.60733112907663e-05, "loss": 0.0251, "step": 67590 }, { "grad_norm": 0.27606722712516785, "learning_rate": 2.6058793990999786e-05, "loss": 0.0294, "step": 67600 }, { "grad_norm": 0.35894691944122314, "learning_rate": 2.6044279309407593e-05, "loss": 0.0217, "step": 67610 }, { "grad_norm": 0.2997201681137085, "learning_rate": 2.6029767247576976e-05, "loss": 0.0248, "step": 67620 }, { "grad_norm": 0.2737341821193695, "learning_rate": 2.601525780709497e-05, "loss": 0.0285, "step": 67630 }, { "grad_norm": 0.19046592712402344, "learning_rate": 2.6000750989548318e-05, "loss": 0.0196, "step": 67640 }, { "grad_norm": 0.29237252473831177, "learning_rate": 2.5986246796523462e-05, "loss": 0.019, "step": 67650 }, { "grad_norm": 0.2194930613040924, "learning_rate": 2.5971745229606546e-05, "loss": 0.0145, "step": 67660 }, { "grad_norm": 0.44250887632369995, "learning_rate": 2.5957246290383442e-05, "loss": 0.0297, "step": 67670 }, { "grad_norm": 0.33091121912002563, "learning_rate": 2.594274998043975e-05, "loss": 0.0177, "step": 67680 }, { "grad_norm": 0.1732260286808014, "learning_rate": 2.5928256301360733e-05, "loss": 0.0181, "step": 67690 }, { "grad_norm": 0.2883419990539551, "learning_rate": 2.591376525473142e-05, "loss": 0.0215, "step": 67700 }, { "grad_norm": 0.3509216010570526, "learning_rate": 2.5899276842136523e-05, "loss": 0.0214, "step": 67710 }, { "grad_norm": 0.2976056933403015, "learning_rate": 2.5884791065160495e-05, "loss": 0.023, "step": 67720 }, { "grad_norm": 0.3302970230579376, "learning_rate": 2.5870307925387436e-05, "loss": 0.0183, "step": 67730 }, { "grad_norm": 0.23458819091320038, "learning_rate": 2.585582742440125e-05, "loss": 0.0141, "step": 67740 }, { "grad_norm": 0.29033970832824707, "learning_rate": 2.5841349563785455e-05, "loss": 0.0214, "step": 67750 }, { "grad_norm": 0.1475861370563507, "learning_rate": 2.5826874345123355e-05, "loss": 0.0175, "step": 67760 }, { "grad_norm": 0.27361026406288147, "learning_rate": 2.581240176999792e-05, "loss": 0.015, "step": 67770 }, { "grad_norm": 0.24869677424430847, "learning_rate": 2.579793183999189e-05, "loss": 0.019, "step": 67780 }, { "grad_norm": 0.2683373987674713, "learning_rate": 2.5783464556687627e-05, "loss": 0.0194, "step": 67790 }, { "grad_norm": 0.4463196098804474, "learning_rate": 2.576899992166729e-05, "loss": 0.0147, "step": 67800 }, { "grad_norm": 0.22162050008773804, "learning_rate": 2.5754537936512668e-05, "loss": 0.0221, "step": 67810 }, { "grad_norm": 0.4076273739337921, "learning_rate": 2.574007860280535e-05, "loss": 0.022, "step": 67820 }, { "grad_norm": 0.29831260442733765, "learning_rate": 2.5725621922126518e-05, "loss": 0.0199, "step": 67830 }, { "grad_norm": 0.35982581973075867, "learning_rate": 2.571116789605722e-05, "loss": 0.0165, "step": 67840 }, { "grad_norm": 0.41366034746170044, "learning_rate": 2.5696716526178054e-05, "loss": 0.0246, "step": 67850 }, { "grad_norm": 0.23462282121181488, "learning_rate": 2.568226781406946e-05, "loss": 0.0241, "step": 67860 }, { "grad_norm": 0.28566473722457886, "learning_rate": 2.5667821761311474e-05, "loss": 0.02, "step": 67870 }, { "grad_norm": 0.22932001948356628, "learning_rate": 2.5653378369483917e-05, "loss": 0.0196, "step": 67880 }, { "grad_norm": 0.28351444005966187, "learning_rate": 2.5638937640166292e-05, "loss": 0.0168, "step": 67890 }, { "grad_norm": 0.2922253906726837, "learning_rate": 2.5624499574937842e-05, "loss": 0.0181, "step": 67900 }, { "grad_norm": 0.358439564704895, "learning_rate": 2.5610064175377456e-05, "loss": 0.0244, "step": 67910 }, { "grad_norm": 0.19527120888233185, "learning_rate": 2.5595631443063782e-05, "loss": 0.0166, "step": 67920 }, { "grad_norm": 0.19401109218597412, "learning_rate": 2.5581201379575177e-05, "loss": 0.0235, "step": 67930 }, { "grad_norm": 0.16970287263393402, "learning_rate": 2.5566773986489655e-05, "loss": 0.0183, "step": 67940 }, { "grad_norm": 0.2908762991428375, "learning_rate": 2.5552349265385e-05, "loss": 0.0281, "step": 67950 }, { "grad_norm": 0.38537871837615967, "learning_rate": 2.5537927217838675e-05, "loss": 0.0205, "step": 67960 }, { "grad_norm": 0.3196740448474884, "learning_rate": 2.5523507845427864e-05, "loss": 0.0198, "step": 67970 }, { "grad_norm": 0.285606324672699, "learning_rate": 2.5509091149729413e-05, "loss": 0.0201, "step": 67980 }, { "grad_norm": 0.2997127175331116, "learning_rate": 2.5494677132319962e-05, "loss": 0.0217, "step": 67990 }, { "grad_norm": 0.3755938410758972, "learning_rate": 2.548026579477575e-05, "loss": 0.0262, "step": 68000 }, { "grad_norm": 0.29608821868896484, "learning_rate": 2.5465857138672834e-05, "loss": 0.0192, "step": 68010 }, { "grad_norm": 0.27302053570747375, "learning_rate": 2.5451451165586848e-05, "loss": 0.0236, "step": 68020 }, { "grad_norm": 0.19034668803215027, "learning_rate": 2.543704787709329e-05, "loss": 0.017, "step": 68030 }, { "grad_norm": 0.28691810369491577, "learning_rate": 2.542264727476723e-05, "loss": 0.0155, "step": 68040 }, { "grad_norm": 0.2020455300807953, "learning_rate": 2.5408249360183535e-05, "loss": 0.0256, "step": 68050 }, { "grad_norm": 0.2891235053539276, "learning_rate": 2.5393854134916694e-05, "loss": 0.0174, "step": 68060 }, { "grad_norm": 0.6119390726089478, "learning_rate": 2.5379461600540993e-05, "loss": 0.0173, "step": 68070 }, { "grad_norm": 0.3171771168708801, "learning_rate": 2.5365071758630317e-05, "loss": 0.0167, "step": 68080 }, { "grad_norm": 0.38369831442832947, "learning_rate": 2.535068461075839e-05, "loss": 0.0251, "step": 68090 }, { "grad_norm": 0.3505890369415283, "learning_rate": 2.5336300158498516e-05, "loss": 0.0239, "step": 68100 }, { "grad_norm": 0.2063063532114029, "learning_rate": 2.5321918403423793e-05, "loss": 0.0227, "step": 68110 }, { "grad_norm": 0.284515917301178, "learning_rate": 2.5307539347106957e-05, "loss": 0.0186, "step": 68120 }, { "grad_norm": 0.41925695538520813, "learning_rate": 2.529316299112049e-05, "loss": 0.0205, "step": 68130 }, { "grad_norm": 0.29586857557296753, "learning_rate": 2.5278789337036567e-05, "loss": 0.0152, "step": 68140 }, { "grad_norm": 0.3245324492454529, "learning_rate": 2.52644183864271e-05, "loss": 0.0181, "step": 68150 }, { "grad_norm": 0.24013292789459229, "learning_rate": 2.525005014086363e-05, "loss": 0.0214, "step": 68160 }, { "grad_norm": 0.2838113009929657, "learning_rate": 2.5235684601917465e-05, "loss": 0.0213, "step": 68170 }, { "grad_norm": 0.1571701169013977, "learning_rate": 2.5221321771159618e-05, "loss": 0.0151, "step": 68180 }, { "grad_norm": 0.2659137547016144, "learning_rate": 2.5206961650160743e-05, "loss": 0.0173, "step": 68190 }, { "grad_norm": 0.19293086230754852, "learning_rate": 2.519260424049129e-05, "loss": 0.0227, "step": 68200 }, { "grad_norm": 0.30795496702194214, "learning_rate": 2.5178249543721304e-05, "loss": 0.0188, "step": 68210 }, { "grad_norm": 0.17000767588615417, "learning_rate": 2.5163897561420668e-05, "loss": 0.0202, "step": 68220 }, { "grad_norm": 0.34255823493003845, "learning_rate": 2.514954829515883e-05, "loss": 0.0224, "step": 68230 }, { "grad_norm": 0.21893997490406036, "learning_rate": 2.513520174650505e-05, "loss": 0.0212, "step": 68240 }, { "grad_norm": 0.4396972060203552, "learning_rate": 2.5120857917028207e-05, "loss": 0.0175, "step": 68250 }, { "grad_norm": 0.2478908896446228, "learning_rate": 2.5106516808296953e-05, "loss": 0.015, "step": 68260 }, { "grad_norm": 0.3709245026111603, "learning_rate": 2.5092178421879552e-05, "loss": 0.0182, "step": 68270 }, { "grad_norm": 0.34336063265800476, "learning_rate": 2.5077842759344116e-05, "loss": 0.0297, "step": 68280 }, { "grad_norm": 0.17722058296203613, "learning_rate": 2.5063509822258304e-05, "loss": 0.0223, "step": 68290 }, { "grad_norm": 0.22764785587787628, "learning_rate": 2.5049179612189576e-05, "loss": 0.0196, "step": 68300 }, { "grad_norm": 0.478645920753479, "learning_rate": 2.5034852130705043e-05, "loss": 0.0228, "step": 68310 }, { "grad_norm": 0.3658584952354431, "learning_rate": 2.5020527379371572e-05, "loss": 0.027, "step": 68320 }, { "grad_norm": 0.2811371684074402, "learning_rate": 2.500620535975564e-05, "loss": 0.0178, "step": 68330 }, { "grad_norm": 0.2086026817560196, "learning_rate": 2.4991886073423515e-05, "loss": 0.0147, "step": 68340 }, { "grad_norm": 0.2026270627975464, "learning_rate": 2.4977569521941134e-05, "loss": 0.0168, "step": 68350 }, { "grad_norm": 0.206195667386055, "learning_rate": 2.4963255706874143e-05, "loss": 0.0206, "step": 68360 }, { "grad_norm": 0.29822978377342224, "learning_rate": 2.4948944629787853e-05, "loss": 0.0235, "step": 68370 }, { "grad_norm": 0.2939487397670746, "learning_rate": 2.493463629224731e-05, "loss": 0.0221, "step": 68380 }, { "grad_norm": 0.4176885485649109, "learning_rate": 2.492033069581728e-05, "loss": 0.0187, "step": 68390 }, { "grad_norm": 0.17361986637115479, "learning_rate": 2.4906027842062137e-05, "loss": 0.0221, "step": 68400 }, { "grad_norm": 0.21884585916996002, "learning_rate": 2.48917277325461e-05, "loss": 0.0198, "step": 68410 }, { "grad_norm": 0.27033793926239014, "learning_rate": 2.487743036883295e-05, "loss": 0.0199, "step": 68420 }, { "grad_norm": 0.3116534948348999, "learning_rate": 2.486313575248626e-05, "loss": 0.0193, "step": 68430 }, { "grad_norm": 0.30831578373908997, "learning_rate": 2.4848843885069235e-05, "loss": 0.0194, "step": 68440 }, { "grad_norm": 0.2171044647693634, "learning_rate": 2.483455476814484e-05, "loss": 0.0183, "step": 68450 }, { "grad_norm": 0.3877778947353363, "learning_rate": 2.4820268403275672e-05, "loss": 0.0203, "step": 68460 }, { "grad_norm": 0.3598494231700897, "learning_rate": 2.480598479202413e-05, "loss": 0.0233, "step": 68470 }, { "grad_norm": 0.28718602657318115, "learning_rate": 2.4791703935952193e-05, "loss": 0.0203, "step": 68480 }, { "grad_norm": 0.2327709048986435, "learning_rate": 2.477742583662163e-05, "loss": 0.0272, "step": 68490 }, { "grad_norm": 0.4167768657207489, "learning_rate": 2.476315049559384e-05, "loss": 0.0192, "step": 68500 }, { "grad_norm": 0.3017202317714691, "learning_rate": 2.4748877914429987e-05, "loss": 0.0186, "step": 68510 }, { "grad_norm": 0.2265785187482834, "learning_rate": 2.4734608094690857e-05, "loss": 0.0227, "step": 68520 }, { "grad_norm": 0.2625340223312378, "learning_rate": 2.4720341037937006e-05, "loss": 0.0249, "step": 68530 }, { "grad_norm": 0.4039916694164276, "learning_rate": 2.4706076745728645e-05, "loss": 0.0286, "step": 68540 }, { "grad_norm": 0.3217073380947113, "learning_rate": 2.469181521962573e-05, "loss": 0.0214, "step": 68550 }, { "grad_norm": 0.20240873098373413, "learning_rate": 2.467755646118783e-05, "loss": 0.0143, "step": 68560 }, { "grad_norm": 0.2566196024417877, "learning_rate": 2.46633004719743e-05, "loss": 0.0209, "step": 68570 }, { "grad_norm": 0.3992047905921936, "learning_rate": 2.4649047253544123e-05, "loss": 0.0163, "step": 68580 }, { "grad_norm": 0.48821401596069336, "learning_rate": 2.4634796807456016e-05, "loss": 0.0179, "step": 68590 }, { "grad_norm": 0.2968871593475342, "learning_rate": 2.46205491352684e-05, "loss": 0.0213, "step": 68600 }, { "grad_norm": 0.31447264552116394, "learning_rate": 2.4606304238539375e-05, "loss": 0.0195, "step": 68610 }, { "grad_norm": 0.320304274559021, "learning_rate": 2.4592062118826753e-05, "loss": 0.017, "step": 68620 }, { "grad_norm": 0.2960079312324524, "learning_rate": 2.4577822777688004e-05, "loss": 0.0199, "step": 68630 }, { "grad_norm": 0.43352752923965454, "learning_rate": 2.4563586216680347e-05, "loss": 0.0214, "step": 68640 }, { "grad_norm": 0.19597157835960388, "learning_rate": 2.4549352437360622e-05, "loss": 0.0201, "step": 68650 }, { "grad_norm": 0.22546105086803436, "learning_rate": 2.4535121441285493e-05, "loss": 0.0173, "step": 68660 }, { "grad_norm": 0.30863314867019653, "learning_rate": 2.4520893230011172e-05, "loss": 0.0175, "step": 68670 }, { "grad_norm": 0.177863210439682, "learning_rate": 2.450666780509368e-05, "loss": 0.0218, "step": 68680 }, { "grad_norm": 0.2563799321651459, "learning_rate": 2.4492445168088656e-05, "loss": 0.0188, "step": 68690 }, { "grad_norm": 0.24171939492225647, "learning_rate": 2.447822532055149e-05, "loss": 0.0127, "step": 68700 }, { "grad_norm": 0.3602461516857147, "learning_rate": 2.4464008264037212e-05, "loss": 0.0183, "step": 68710 }, { "grad_norm": 0.2534983456134796, "learning_rate": 2.4449794000100605e-05, "loss": 0.0198, "step": 68720 }, { "grad_norm": 0.8888373374938965, "learning_rate": 2.4435582530296108e-05, "loss": 0.0172, "step": 68730 }, { "grad_norm": 0.22220666706562042, "learning_rate": 2.4421373856177887e-05, "loss": 0.0178, "step": 68740 }, { "grad_norm": 0.31279680132865906, "learning_rate": 2.440716797929975e-05, "loss": 0.0187, "step": 68750 }, { "grad_norm": 0.2633315622806549, "learning_rate": 2.439296490121526e-05, "loss": 0.0155, "step": 68760 }, { "grad_norm": 0.15512803196907043, "learning_rate": 2.4378764623477608e-05, "loss": 0.0134, "step": 68770 }, { "grad_norm": 0.3312704265117645, "learning_rate": 2.436456714763974e-05, "loss": 0.0194, "step": 68780 }, { "grad_norm": 0.3483356833457947, "learning_rate": 2.4350372475254262e-05, "loss": 0.0188, "step": 68790 }, { "grad_norm": 0.22471314668655396, "learning_rate": 2.433618060787351e-05, "loss": 0.0187, "step": 68800 }, { "grad_norm": 0.20053988695144653, "learning_rate": 2.432199154704944e-05, "loss": 0.0257, "step": 68810 }, { "grad_norm": 0.25948667526245117, "learning_rate": 2.430780529433377e-05, "loss": 0.0174, "step": 68820 }, { "grad_norm": 0.25752848386764526, "learning_rate": 2.42936218512779e-05, "loss": 0.0267, "step": 68830 }, { "grad_norm": 0.2506134510040283, "learning_rate": 2.427944121943288e-05, "loss": 0.0171, "step": 68840 }, { "grad_norm": 0.35951900482177734, "learning_rate": 2.4265263400349497e-05, "loss": 0.0225, "step": 68850 }, { "grad_norm": 0.36475399136543274, "learning_rate": 2.4251088395578213e-05, "loss": 0.0211, "step": 68860 }, { "grad_norm": 0.29303640127182007, "learning_rate": 2.4236916206669214e-05, "loss": 0.0174, "step": 68870 }, { "grad_norm": 0.2357783317565918, "learning_rate": 2.4222746835172304e-05, "loss": 0.0141, "step": 68880 }, { "grad_norm": 0.3570748269557953, "learning_rate": 2.4208580282637065e-05, "loss": 0.0214, "step": 68890 }, { "grad_norm": 0.21206840872764587, "learning_rate": 2.4194416550612687e-05, "loss": 0.0155, "step": 68900 }, { "grad_norm": 0.36300525069236755, "learning_rate": 2.418025564064812e-05, "loss": 0.0172, "step": 68910 }, { "grad_norm": 0.17004525661468506, "learning_rate": 2.4166097554291978e-05, "loss": 0.0199, "step": 68920 }, { "grad_norm": 0.35705485939979553, "learning_rate": 2.4151942293092583e-05, "loss": 0.0215, "step": 68930 }, { "grad_norm": 0.2833733558654785, "learning_rate": 2.41377898585979e-05, "loss": 0.014, "step": 68940 }, { "grad_norm": 0.39085322618484497, "learning_rate": 2.4123640252355652e-05, "loss": 0.0211, "step": 68950 }, { "grad_norm": 0.2589625418186188, "learning_rate": 2.410949347591318e-05, "loss": 0.0163, "step": 68960 }, { "grad_norm": 0.4163057208061218, "learning_rate": 2.4095349530817578e-05, "loss": 0.0352, "step": 68970 }, { "grad_norm": 0.2449168562889099, "learning_rate": 2.4081208418615603e-05, "loss": 0.028, "step": 68980 }, { "grad_norm": 0.1658928245306015, "learning_rate": 2.4067070140853725e-05, "loss": 0.0175, "step": 68990 }, { "grad_norm": 0.2624630033969879, "learning_rate": 2.405293469907805e-05, "loss": 0.0166, "step": 69000 }, { "grad_norm": 0.2807163596153259, "learning_rate": 2.4038802094834446e-05, "loss": 0.0217, "step": 69010 }, { "grad_norm": 0.26162466406822205, "learning_rate": 2.4024672329668395e-05, "loss": 0.0157, "step": 69020 }, { "grad_norm": 0.26867538690567017, "learning_rate": 2.4010545405125125e-05, "loss": 0.0151, "step": 69030 }, { "grad_norm": 0.26705431938171387, "learning_rate": 2.3996421322749528e-05, "loss": 0.0175, "step": 69040 }, { "grad_norm": 0.33093854784965515, "learning_rate": 2.3982300084086224e-05, "loss": 0.0238, "step": 69050 }, { "grad_norm": 0.4192166328430176, "learning_rate": 2.396818169067944e-05, "loss": 0.016, "step": 69060 }, { "grad_norm": 0.20891419053077698, "learning_rate": 2.3954066144073178e-05, "loss": 0.0165, "step": 69070 }, { "grad_norm": 0.4101039171218872, "learning_rate": 2.39399534458111e-05, "loss": 0.0218, "step": 69080 }, { "grad_norm": 0.23517616093158722, "learning_rate": 2.392584359743651e-05, "loss": 0.0185, "step": 69090 }, { "grad_norm": 0.17775286734104156, "learning_rate": 2.3911736600492463e-05, "loss": 0.0135, "step": 69100 }, { "grad_norm": 0.4172225892543793, "learning_rate": 2.3897632456521672e-05, "loss": 0.0194, "step": 69110 }, { "grad_norm": 0.31065526604652405, "learning_rate": 2.388353116706658e-05, "loss": 0.0233, "step": 69120 }, { "grad_norm": 0.22540271282196045, "learning_rate": 2.3869432733669232e-05, "loss": 0.0228, "step": 69130 }, { "grad_norm": 0.28013479709625244, "learning_rate": 2.385533715787145e-05, "loss": 0.0181, "step": 69140 }, { "grad_norm": 0.22835251688957214, "learning_rate": 2.3841244441214673e-05, "loss": 0.0171, "step": 69150 }, { "grad_norm": 0.37214940786361694, "learning_rate": 2.3827154585240073e-05, "loss": 0.0165, "step": 69160 }, { "grad_norm": 0.46063095331192017, "learning_rate": 2.3813067591488498e-05, "loss": 0.0246, "step": 69170 }, { "grad_norm": 0.24230460822582245, "learning_rate": 2.3798983461500494e-05, "loss": 0.0215, "step": 69180 }, { "grad_norm": 0.14753834903240204, "learning_rate": 2.3784902196816256e-05, "loss": 0.0158, "step": 69190 }, { "grad_norm": 0.2429760843515396, "learning_rate": 2.3770823798975716e-05, "loss": 0.0215, "step": 69200 }, { "grad_norm": 0.18787701427936554, "learning_rate": 2.3756748269518424e-05, "loss": 0.0127, "step": 69210 }, { "grad_norm": 0.21525853872299194, "learning_rate": 2.374267560998371e-05, "loss": 0.0194, "step": 69220 }, { "grad_norm": 0.2971363365650177, "learning_rate": 2.3728605821910477e-05, "loss": 0.0256, "step": 69230 }, { "grad_norm": 0.4760010540485382, "learning_rate": 2.3714538906837453e-05, "loss": 0.0278, "step": 69240 }, { "grad_norm": 0.23281918466091156, "learning_rate": 2.3700474866302913e-05, "loss": 0.015, "step": 69250 }, { "grad_norm": 0.27182820439338684, "learning_rate": 2.3686413701844917e-05, "loss": 0.0156, "step": 69260 }, { "grad_norm": 0.2154955118894577, "learning_rate": 2.367235541500114e-05, "loss": 0.0201, "step": 69270 }, { "grad_norm": 0.3262418806552887, "learning_rate": 2.3658300007308993e-05, "loss": 0.017, "step": 69280 }, { "grad_norm": 0.15081888437271118, "learning_rate": 2.364424748030555e-05, "loss": 0.0127, "step": 69290 }, { "grad_norm": 0.20534561574459076, "learning_rate": 2.3630197835527595e-05, "loss": 0.0167, "step": 69300 }, { "grad_norm": 0.26609086990356445, "learning_rate": 2.3616151074511546e-05, "loss": 0.0216, "step": 69310 }, { "grad_norm": 0.3634340167045593, "learning_rate": 2.3602107198793545e-05, "loss": 0.0188, "step": 69320 }, { "grad_norm": 0.18782256543636322, "learning_rate": 2.3588066209909427e-05, "loss": 0.0218, "step": 69330 }, { "grad_norm": 0.3002335727214813, "learning_rate": 2.3574028109394662e-05, "loss": 0.0233, "step": 69340 }, { "grad_norm": 0.29609131813049316, "learning_rate": 2.3559992898784454e-05, "loss": 0.0243, "step": 69350 }, { "grad_norm": 0.13646438717842102, "learning_rate": 2.354596057961367e-05, "loss": 0.0129, "step": 69360 }, { "grad_norm": 0.26708757877349854, "learning_rate": 2.353193115341688e-05, "loss": 0.0317, "step": 69370 }, { "grad_norm": 0.2716245949268341, "learning_rate": 2.3517904621728294e-05, "loss": 0.0217, "step": 69380 }, { "grad_norm": 0.19813716411590576, "learning_rate": 2.3503880986081855e-05, "loss": 0.0251, "step": 69390 }, { "grad_norm": 0.21524304151535034, "learning_rate": 2.3489860248011136e-05, "loss": 0.0165, "step": 69400 }, { "grad_norm": 0.2839912474155426, "learning_rate": 2.3475842409049464e-05, "loss": 0.0214, "step": 69410 }, { "grad_norm": 0.2884373366832733, "learning_rate": 2.346182747072975e-05, "loss": 0.0252, "step": 69420 }, { "grad_norm": 0.3462439775466919, "learning_rate": 2.3447815434584718e-05, "loss": 0.0176, "step": 69430 }, { "grad_norm": 0.4220311939716339, "learning_rate": 2.3433806302146656e-05, "loss": 0.0178, "step": 69440 }, { "grad_norm": 0.4971824586391449, "learning_rate": 2.3419800074947612e-05, "loss": 0.0185, "step": 69450 }, { "grad_norm": 0.1948045939207077, "learning_rate": 2.3405796754519248e-05, "loss": 0.0215, "step": 69460 }, { "grad_norm": 0.26586589217185974, "learning_rate": 2.3391796342392992e-05, "loss": 0.0126, "step": 69470 }, { "grad_norm": 0.5078818202018738, "learning_rate": 2.337779884009984e-05, "loss": 0.0203, "step": 69480 }, { "grad_norm": 0.2276420146226883, "learning_rate": 2.3363804249170624e-05, "loss": 0.0225, "step": 69490 }, { "grad_norm": 0.3145105242729187, "learning_rate": 2.3349812571135703e-05, "loss": 0.0162, "step": 69500 }, { "grad_norm": 0.28399431705474854, "learning_rate": 2.333582380752523e-05, "loss": 0.0201, "step": 69510 }, { "grad_norm": 0.21832334995269775, "learning_rate": 2.3321837959868963e-05, "loss": 0.0156, "step": 69520 }, { "grad_norm": 0.1731681227684021, "learning_rate": 2.3307855029696383e-05, "loss": 0.015, "step": 69530 }, { "grad_norm": 0.25111812353134155, "learning_rate": 2.3293875018536648e-05, "loss": 0.0149, "step": 69540 }, { "grad_norm": 0.1940762847661972, "learning_rate": 2.327989792791861e-05, "loss": 0.0174, "step": 69550 }, { "grad_norm": 0.3023168444633484, "learning_rate": 2.326592375937074e-05, "loss": 0.0223, "step": 69560 }, { "grad_norm": 0.256346195936203, "learning_rate": 2.3251952514421248e-05, "loss": 0.0173, "step": 69570 }, { "grad_norm": 0.32697075605392456, "learning_rate": 2.323798419459804e-05, "loss": 0.0194, "step": 69580 }, { "grad_norm": 0.2308606058359146, "learning_rate": 2.3224018801428625e-05, "loss": 0.015, "step": 69590 }, { "grad_norm": 0.3999271094799042, "learning_rate": 2.3210056336440273e-05, "loss": 0.023, "step": 69600 }, { "grad_norm": 0.1775510311126709, "learning_rate": 2.3196096801159844e-05, "loss": 0.0166, "step": 69610 }, { "grad_norm": 0.33198457956314087, "learning_rate": 2.3182140197114012e-05, "loss": 0.0159, "step": 69620 }, { "grad_norm": 0.30576661229133606, "learning_rate": 2.3168186525828985e-05, "loss": 0.0206, "step": 69630 }, { "grad_norm": 0.4302988648414612, "learning_rate": 2.3154235788830757e-05, "loss": 0.0229, "step": 69640 }, { "grad_norm": 1.7661268711090088, "learning_rate": 2.3140287987644927e-05, "loss": 0.0233, "step": 69650 }, { "grad_norm": 0.30558574199676514, "learning_rate": 2.3126343123796833e-05, "loss": 0.0167, "step": 69660 }, { "grad_norm": 0.23081223666667938, "learning_rate": 2.3112401198811418e-05, "loss": 0.0234, "step": 69670 }, { "grad_norm": 0.41968613862991333, "learning_rate": 2.3098462214213418e-05, "loss": 0.0232, "step": 69680 }, { "grad_norm": 0.2170220911502838, "learning_rate": 2.308452617152712e-05, "loss": 0.0159, "step": 69690 }, { "grad_norm": 0.29947200417518616, "learning_rate": 2.3070593072276603e-05, "loss": 0.0194, "step": 69700 }, { "grad_norm": 0.2184055745601654, "learning_rate": 2.3056662917985518e-05, "loss": 0.0188, "step": 69710 }, { "grad_norm": 0.15886835753917694, "learning_rate": 2.3042735710177283e-05, "loss": 0.0313, "step": 69720 }, { "grad_norm": 0.2749366760253906, "learning_rate": 2.302881145037492e-05, "loss": 0.016, "step": 69730 }, { "grad_norm": 0.6107994318008423, "learning_rate": 2.3014890140101192e-05, "loss": 0.0192, "step": 69740 }, { "grad_norm": 0.3000996708869934, "learning_rate": 2.30009717808785e-05, "loss": 0.017, "step": 69750 }, { "grad_norm": 0.3215000331401825, "learning_rate": 2.2987056374228967e-05, "loss": 0.0168, "step": 69760 }, { "grad_norm": 0.4876407980918884, "learning_rate": 2.297314392167431e-05, "loss": 0.0212, "step": 69770 }, { "grad_norm": 0.3680923879146576, "learning_rate": 2.295923442473601e-05, "loss": 0.0177, "step": 69780 }, { "grad_norm": 0.18702161312103271, "learning_rate": 2.294532788493519e-05, "loss": 0.0165, "step": 69790 }, { "grad_norm": 0.4033478796482086, "learning_rate": 2.293142430379262e-05, "loss": 0.0196, "step": 69800 }, { "grad_norm": 0.27068766951560974, "learning_rate": 2.291752368282879e-05, "loss": 0.0141, "step": 69810 }, { "grad_norm": 0.2941875457763672, "learning_rate": 2.290362602356385e-05, "loss": 0.0169, "step": 69820 }, { "grad_norm": 0.27293041348457336, "learning_rate": 2.288973132751765e-05, "loss": 0.0142, "step": 69830 }, { "grad_norm": 0.29243800044059753, "learning_rate": 2.287583959620965e-05, "loss": 0.0271, "step": 69840 }, { "grad_norm": 0.2190294712781906, "learning_rate": 2.2861950831159074e-05, "loss": 0.0186, "step": 69850 }, { "grad_norm": 0.207861989736557, "learning_rate": 2.284806503388472e-05, "loss": 0.0212, "step": 69860 }, { "grad_norm": 0.2837592363357544, "learning_rate": 2.283418220590518e-05, "loss": 0.0246, "step": 69870 }, { "grad_norm": 0.451692134141922, "learning_rate": 2.282030234873862e-05, "loss": 0.0205, "step": 69880 }, { "grad_norm": 0.26124337315559387, "learning_rate": 2.2806425463902948e-05, "loss": 0.0155, "step": 69890 }, { "grad_norm": 0.3190125524997711, "learning_rate": 2.279255155291568e-05, "loss": 0.0221, "step": 69900 }, { "grad_norm": 0.3557928204536438, "learning_rate": 2.277868061729409e-05, "loss": 0.0177, "step": 69910 }, { "grad_norm": 0.25833749771118164, "learning_rate": 2.276481265855504e-05, "loss": 0.0202, "step": 69920 }, { "grad_norm": 0.5016780495643616, "learning_rate": 2.2750947678215128e-05, "loss": 0.0262, "step": 69930 }, { "grad_norm": 0.23103095591068268, "learning_rate": 2.2737085677790616e-05, "loss": 0.0162, "step": 69940 }, { "grad_norm": 0.7540201544761658, "learning_rate": 2.2723226658797443e-05, "loss": 0.0187, "step": 69950 }, { "grad_norm": 0.3145594000816345, "learning_rate": 2.270937062275117e-05, "loss": 0.0235, "step": 69960 }, { "grad_norm": 0.5347645878791809, "learning_rate": 2.2695517571167113e-05, "loss": 0.0236, "step": 69970 }, { "grad_norm": 0.7142834067344666, "learning_rate": 2.268166750556019e-05, "loss": 0.015, "step": 69980 }, { "grad_norm": 0.38106751441955566, "learning_rate": 2.266782042744503e-05, "loss": 0.0144, "step": 69990 }, { "grad_norm": 0.1664426624774933, "learning_rate": 2.2653976338335935e-05, "loss": 0.0132, "step": 70000 }, { "grad_norm": 0.6052603125572205, "learning_rate": 2.2640135239746895e-05, "loss": 0.021, "step": 70010 }, { "grad_norm": 0.2054184228181839, "learning_rate": 2.2626297133191515e-05, "loss": 0.0192, "step": 70020 }, { "grad_norm": 0.29315313696861267, "learning_rate": 2.261246202018312e-05, "loss": 0.0181, "step": 70030 }, { "grad_norm": 0.1843707114458084, "learning_rate": 2.2598629902234718e-05, "loss": 0.0248, "step": 70040 }, { "grad_norm": 0.27124959230422974, "learning_rate": 2.258480078085894e-05, "loss": 0.0164, "step": 70050 }, { "grad_norm": 0.23287010192871094, "learning_rate": 2.2570974657568126e-05, "loss": 0.0177, "step": 70060 }, { "grad_norm": 0.3364385664463043, "learning_rate": 2.2557151533874286e-05, "loss": 0.0171, "step": 70070 }, { "grad_norm": 0.2368917316198349, "learning_rate": 2.2543331411289116e-05, "loss": 0.0163, "step": 70080 }, { "grad_norm": 0.13841313123703003, "learning_rate": 2.252951429132392e-05, "loss": 0.0182, "step": 70090 }, { "grad_norm": 0.3370126485824585, "learning_rate": 2.2515700175489762e-05, "loss": 0.0198, "step": 70100 }, { "grad_norm": 0.627129316329956, "learning_rate": 2.2501889065297288e-05, "loss": 0.0169, "step": 70110 }, { "grad_norm": 0.44618821144104004, "learning_rate": 2.2488080962256887e-05, "loss": 0.0232, "step": 70120 }, { "grad_norm": 0.3587641417980194, "learning_rate": 2.2474275867878585e-05, "loss": 0.0224, "step": 70130 }, { "grad_norm": 1.7375726699829102, "learning_rate": 2.246047378367211e-05, "loss": 0.0276, "step": 70140 }, { "grad_norm": 0.6108624935150146, "learning_rate": 2.2446674711146798e-05, "loss": 0.0226, "step": 70150 }, { "grad_norm": 0.49408838152885437, "learning_rate": 2.2432878651811734e-05, "loss": 0.0197, "step": 70160 }, { "grad_norm": 0.17630313336849213, "learning_rate": 2.2419085607175594e-05, "loss": 0.0259, "step": 70170 }, { "grad_norm": 0.15526658296585083, "learning_rate": 2.240529557874678e-05, "loss": 0.0284, "step": 70180 }, { "grad_norm": 0.23975074291229248, "learning_rate": 2.2391508568033358e-05, "loss": 0.0158, "step": 70190 }, { "grad_norm": 0.25880762934684753, "learning_rate": 2.237772457654307e-05, "loss": 0.0145, "step": 70200 }, { "grad_norm": 0.3825399875640869, "learning_rate": 2.236394360578327e-05, "loss": 0.0226, "step": 70210 }, { "grad_norm": 0.2105836421251297, "learning_rate": 2.2350165657261073e-05, "loss": 0.0133, "step": 70220 }, { "grad_norm": 0.2732999920845032, "learning_rate": 2.2336390732483164e-05, "loss": 0.0195, "step": 70230 }, { "grad_norm": 0.19355852901935577, "learning_rate": 2.2322618832955976e-05, "loss": 0.0205, "step": 70240 }, { "grad_norm": 0.2942565381526947, "learning_rate": 2.2308849960185578e-05, "loss": 0.0144, "step": 70250 }, { "grad_norm": 0.26846975088119507, "learning_rate": 2.229508411567774e-05, "loss": 0.0218, "step": 70260 }, { "grad_norm": 0.25134924054145813, "learning_rate": 2.2281321300937823e-05, "loss": 0.0196, "step": 70270 }, { "grad_norm": 0.19712048768997192, "learning_rate": 2.2267561517470937e-05, "loss": 0.0138, "step": 70280 }, { "grad_norm": 0.4143908619880676, "learning_rate": 2.2253804766781845e-05, "loss": 0.0235, "step": 70290 }, { "grad_norm": 0.4116044342517853, "learning_rate": 2.2240051050374933e-05, "loss": 0.0175, "step": 70300 }, { "grad_norm": 0.3531509339809418, "learning_rate": 2.22263003697543e-05, "loss": 0.0183, "step": 70310 }, { "grad_norm": 0.2314496785402298, "learning_rate": 2.2212552726423702e-05, "loss": 0.0208, "step": 70320 }, { "grad_norm": 0.2558360695838928, "learning_rate": 2.2198808121886578e-05, "loss": 0.0163, "step": 70330 }, { "grad_norm": 0.1578306406736374, "learning_rate": 2.218506655764598e-05, "loss": 0.016, "step": 70340 }, { "grad_norm": 0.20847749710083008, "learning_rate": 2.2171328035204704e-05, "loss": 0.0157, "step": 70350 }, { "grad_norm": 0.28355950117111206, "learning_rate": 2.2157592556065133e-05, "loss": 0.0178, "step": 70360 }, { "grad_norm": 0.36396336555480957, "learning_rate": 2.2143860121729382e-05, "loss": 0.0143, "step": 70370 }, { "grad_norm": 0.2087821215391159, "learning_rate": 2.2130130733699206e-05, "loss": 0.0136, "step": 70380 }, { "grad_norm": 0.3051110506057739, "learning_rate": 2.2116404393476052e-05, "loss": 0.0157, "step": 70390 }, { "grad_norm": 0.2903865873813629, "learning_rate": 2.2102681102560967e-05, "loss": 0.0175, "step": 70400 }, { "grad_norm": 0.16755157709121704, "learning_rate": 2.208896086245476e-05, "loss": 0.0152, "step": 70410 }, { "grad_norm": 0.16539013385772705, "learning_rate": 2.2075243674657808e-05, "loss": 0.0118, "step": 70420 }, { "grad_norm": 0.3699314296245575, "learning_rate": 2.206152954067024e-05, "loss": 0.0182, "step": 70430 }, { "grad_norm": 0.15188917517662048, "learning_rate": 2.2047818461991758e-05, "loss": 0.0213, "step": 70440 }, { "grad_norm": 0.28515252470970154, "learning_rate": 2.2034110440121864e-05, "loss": 0.018, "step": 70450 }, { "grad_norm": 0.29173341393470764, "learning_rate": 2.2020405476559585e-05, "loss": 0.0233, "step": 70460 }, { "grad_norm": 0.3254372477531433, "learning_rate": 2.2006703572803722e-05, "loss": 0.0209, "step": 70470 }, { "grad_norm": 0.24097003042697906, "learning_rate": 2.1993004730352646e-05, "loss": 0.0166, "step": 70480 }, { "grad_norm": 0.16029848158359528, "learning_rate": 2.1979308950704464e-05, "loss": 0.0179, "step": 70490 }, { "grad_norm": 0.2376510500907898, "learning_rate": 2.1965616235356923e-05, "loss": 0.0177, "step": 70500 }, { "grad_norm": 0.25026455521583557, "learning_rate": 2.1951926585807465e-05, "loss": 0.0163, "step": 70510 }, { "grad_norm": 0.2229241132736206, "learning_rate": 2.193824000355313e-05, "loss": 0.0169, "step": 70520 }, { "grad_norm": 0.21612654626369476, "learning_rate": 2.192455649009067e-05, "loss": 0.0163, "step": 70530 }, { "grad_norm": 0.3542534410953522, "learning_rate": 2.1910876046916523e-05, "loss": 0.0166, "step": 70540 }, { "grad_norm": 0.2378130406141281, "learning_rate": 2.1897198675526725e-05, "loss": 0.0205, "step": 70550 }, { "grad_norm": 0.39445382356643677, "learning_rate": 2.1883524377417024e-05, "loss": 0.0206, "step": 70560 }, { "grad_norm": 0.3311583399772644, "learning_rate": 2.1869853154082825e-05, "loss": 0.0167, "step": 70570 }, { "grad_norm": 0.18918007612228394, "learning_rate": 2.1856185007019215e-05, "loss": 0.0145, "step": 70580 }, { "grad_norm": 0.20752829313278198, "learning_rate": 2.1842519937720874e-05, "loss": 0.0182, "step": 70590 }, { "grad_norm": 0.24854889512062073, "learning_rate": 2.182885794768224e-05, "loss": 0.0158, "step": 70600 }, { "grad_norm": 0.3109031617641449, "learning_rate": 2.1815199038397326e-05, "loss": 0.017, "step": 70610 }, { "grad_norm": 0.21110840141773224, "learning_rate": 2.1801543211359893e-05, "loss": 0.0204, "step": 70620 }, { "grad_norm": 0.2598766088485718, "learning_rate": 2.178789046806326e-05, "loss": 0.0144, "step": 70630 }, { "grad_norm": 0.39332079887390137, "learning_rate": 2.1774240810000547e-05, "loss": 0.0206, "step": 70640 }, { "grad_norm": 0.2583237290382385, "learning_rate": 2.17605942386644e-05, "loss": 0.0187, "step": 70650 }, { "grad_norm": 0.2597702741622925, "learning_rate": 2.174695075554723e-05, "loss": 0.0255, "step": 70660 }, { "grad_norm": 0.2563655376434326, "learning_rate": 2.173331036214103e-05, "loss": 0.0309, "step": 70670 }, { "grad_norm": 0.1750147044658661, "learning_rate": 2.171967305993753e-05, "loss": 0.0177, "step": 70680 }, { "grad_norm": 0.22627346217632294, "learning_rate": 2.1706038850428024e-05, "loss": 0.0129, "step": 70690 }, { "grad_norm": 0.27078038454055786, "learning_rate": 2.169240773510361e-05, "loss": 0.0132, "step": 70700 }, { "grad_norm": 0.3423754572868347, "learning_rate": 2.1678779715454906e-05, "loss": 0.0144, "step": 70710 }, { "grad_norm": 0.224361851811409, "learning_rate": 2.166515479297229e-05, "loss": 0.015, "step": 70720 }, { "grad_norm": 0.17129945755004883, "learning_rate": 2.1651532969145732e-05, "loss": 0.0112, "step": 70730 }, { "grad_norm": 0.4141070246696472, "learning_rate": 2.1637914245464903e-05, "loss": 0.0243, "step": 70740 }, { "grad_norm": 0.36437973380088806, "learning_rate": 2.1624298623419133e-05, "loss": 0.0153, "step": 70750 }, { "grad_norm": 0.3298783600330353, "learning_rate": 2.1610686104497413e-05, "loss": 0.0181, "step": 70760 }, { "grad_norm": 0.5422900915145874, "learning_rate": 2.1597076690188366e-05, "loss": 0.0268, "step": 70770 }, { "grad_norm": 0.27472367882728577, "learning_rate": 2.1583470381980315e-05, "loss": 0.0191, "step": 70780 }, { "grad_norm": 0.19192135334014893, "learning_rate": 2.156986718136123e-05, "loss": 0.0265, "step": 70790 }, { "grad_norm": 0.2659989893436432, "learning_rate": 2.155626708981871e-05, "loss": 0.0199, "step": 70800 }, { "grad_norm": 0.32651036977767944, "learning_rate": 2.1542670108840075e-05, "loss": 0.0188, "step": 70810 }, { "grad_norm": 0.42981502413749695, "learning_rate": 2.152907623991221e-05, "loss": 0.0229, "step": 70820 }, { "grad_norm": 0.2726651728153229, "learning_rate": 2.1515485484521802e-05, "loss": 0.0169, "step": 70830 }, { "grad_norm": 0.2614828050136566, "learning_rate": 2.150189784415506e-05, "loss": 0.0181, "step": 70840 }, { "grad_norm": 0.26823025941848755, "learning_rate": 2.148831332029794e-05, "loss": 0.0141, "step": 70850 }, { "grad_norm": 0.12279842793941498, "learning_rate": 2.147473191443599e-05, "loss": 0.0213, "step": 70860 }, { "grad_norm": 0.2637643814086914, "learning_rate": 2.1461153628054492e-05, "loss": 0.0211, "step": 70870 }, { "grad_norm": 0.36230260133743286, "learning_rate": 2.1447578462638284e-05, "loss": 0.0196, "step": 70880 }, { "grad_norm": 0.2050219029188156, "learning_rate": 2.143400641967201e-05, "loss": 0.016, "step": 70890 }, { "grad_norm": 0.28827032446861267, "learning_rate": 2.1420437500639823e-05, "loss": 0.0139, "step": 70900 }, { "grad_norm": 0.1843646615743637, "learning_rate": 2.1406871707025646e-05, "loss": 0.0204, "step": 70910 }, { "grad_norm": 0.5523907542228699, "learning_rate": 2.1393309040312963e-05, "loss": 0.0217, "step": 70920 }, { "grad_norm": 0.20463258028030396, "learning_rate": 2.137974950198502e-05, "loss": 0.0176, "step": 70930 }, { "grad_norm": 0.2775607109069824, "learning_rate": 2.1366193093524617e-05, "loss": 0.0168, "step": 70940 }, { "grad_norm": 0.4386969208717346, "learning_rate": 2.1352639816414288e-05, "loss": 0.0154, "step": 70950 }, { "grad_norm": 0.30035853385925293, "learning_rate": 2.1339089672136202e-05, "loss": 0.0201, "step": 70960 }, { "grad_norm": 0.2740505039691925, "learning_rate": 2.1325542662172193e-05, "loss": 0.0215, "step": 70970 }, { "grad_norm": 0.18345113098621368, "learning_rate": 2.1311998788003712e-05, "loss": 0.0157, "step": 70980 }, { "grad_norm": 0.252106636762619, "learning_rate": 2.1298458051111907e-05, "loss": 0.0182, "step": 70990 }, { "grad_norm": 0.27516892552375793, "learning_rate": 2.1284920452977602e-05, "loss": 0.0174, "step": 71000 }, { "grad_norm": 0.28476521372795105, "learning_rate": 2.12713859950812e-05, "loss": 0.0215, "step": 71010 }, { "grad_norm": 0.21976450085639954, "learning_rate": 2.1257854678902832e-05, "loss": 0.0134, "step": 71020 }, { "grad_norm": 0.3257816433906555, "learning_rate": 2.124432650592227e-05, "loss": 0.0228, "step": 71030 }, { "grad_norm": 0.26400426030158997, "learning_rate": 2.1230801477618937e-05, "loss": 0.0218, "step": 71040 }, { "grad_norm": 0.48131629824638367, "learning_rate": 2.121727959547189e-05, "loss": 0.0265, "step": 71050 }, { "grad_norm": 0.27938538789749146, "learning_rate": 2.1203760860959886e-05, "loss": 0.0226, "step": 71060 }, { "grad_norm": 0.3213844895362854, "learning_rate": 2.119024527556127e-05, "loss": 0.0238, "step": 71070 }, { "grad_norm": 0.8066507577896118, "learning_rate": 2.117673284075415e-05, "loss": 0.0137, "step": 71080 }, { "grad_norm": 0.1617484837770462, "learning_rate": 2.1163223558016165e-05, "loss": 0.0158, "step": 71090 }, { "grad_norm": 1.4296493530273438, "learning_rate": 2.1149717428824722e-05, "loss": 0.0194, "step": 71100 }, { "grad_norm": 0.40189266204833984, "learning_rate": 2.1136214454656784e-05, "loss": 0.0237, "step": 71110 }, { "grad_norm": 0.21819178760051727, "learning_rate": 2.1122714636989056e-05, "loss": 0.0212, "step": 71120 }, { "grad_norm": 0.24356867372989655, "learning_rate": 2.1109217977297823e-05, "loss": 0.014, "step": 71130 }, { "grad_norm": 0.2315232902765274, "learning_rate": 2.1095724477059077e-05, "loss": 0.0182, "step": 71140 }, { "grad_norm": 0.28521326184272766, "learning_rate": 2.1082234137748434e-05, "loss": 0.0211, "step": 71150 }, { "grad_norm": 0.20677857100963593, "learning_rate": 2.106874696084122e-05, "loss": 0.0193, "step": 71160 }, { "grad_norm": 0.17904484272003174, "learning_rate": 2.1055262947812315e-05, "loss": 0.0153, "step": 71170 }, { "grad_norm": 0.22538217902183533, "learning_rate": 2.104178210013636e-05, "loss": 0.0229, "step": 71180 }, { "grad_norm": 0.46591490507125854, "learning_rate": 2.1028304419287554e-05, "loss": 0.0248, "step": 71190 }, { "grad_norm": 0.31646376848220825, "learning_rate": 2.101482990673982e-05, "loss": 0.0215, "step": 71200 }, { "grad_norm": 0.31853410601615906, "learning_rate": 2.1001358563966707e-05, "loss": 0.0216, "step": 71210 }, { "grad_norm": 0.29089292883872986, "learning_rate": 2.098789039244144e-05, "loss": 0.0134, "step": 71220 }, { "grad_norm": 0.18508075177669525, "learning_rate": 2.097442539363685e-05, "loss": 0.0162, "step": 71230 }, { "grad_norm": 0.4024141728878021, "learning_rate": 2.0960963569025448e-05, "loss": 0.0157, "step": 71240 }, { "grad_norm": 0.15007071197032928, "learning_rate": 2.0947504920079435e-05, "loss": 0.0146, "step": 71250 }, { "grad_norm": 0.1999998390674591, "learning_rate": 2.093404944827059e-05, "loss": 0.0176, "step": 71260 }, { "grad_norm": 0.16131384670734406, "learning_rate": 2.0920597155070394e-05, "loss": 0.0157, "step": 71270 }, { "grad_norm": 0.3623846769332886, "learning_rate": 2.090714804194997e-05, "loss": 0.0247, "step": 71280 }, { "grad_norm": 0.25540876388549805, "learning_rate": 2.0893702110380114e-05, "loss": 0.0169, "step": 71290 }, { "grad_norm": 0.6964212656021118, "learning_rate": 2.0880259361831216e-05, "loss": 0.0252, "step": 71300 }, { "grad_norm": 0.40477609634399414, "learning_rate": 2.0866819797773397e-05, "loss": 0.019, "step": 71310 }, { "grad_norm": 0.46914955973625183, "learning_rate": 2.085338341967634e-05, "loss": 0.0187, "step": 71320 }, { "grad_norm": 0.22187547385692596, "learning_rate": 2.083995022900946e-05, "loss": 0.018, "step": 71330 }, { "grad_norm": 0.17830754816532135, "learning_rate": 2.0826520227241768e-05, "loss": 0.0162, "step": 71340 }, { "grad_norm": 0.29395315051078796, "learning_rate": 2.081309341584199e-05, "loss": 0.0213, "step": 71350 }, { "grad_norm": 0.21606037020683289, "learning_rate": 2.0799669796278413e-05, "loss": 0.0246, "step": 71360 }, { "grad_norm": 0.2193666249513626, "learning_rate": 2.078624937001906e-05, "loss": 0.0201, "step": 71370 }, { "grad_norm": 0.3776893615722656, "learning_rate": 2.0772832138531544e-05, "loss": 0.0353, "step": 71380 }, { "grad_norm": 0.30204153060913086, "learning_rate": 2.0759418103283158e-05, "loss": 0.0279, "step": 71390 }, { "grad_norm": 0.240901380777359, "learning_rate": 2.0746007265740846e-05, "loss": 0.0211, "step": 71400 }, { "grad_norm": 0.33169975876808167, "learning_rate": 2.0732599627371215e-05, "loss": 0.0181, "step": 71410 }, { "grad_norm": 0.27154454588890076, "learning_rate": 2.0719195189640468e-05, "loss": 0.0171, "step": 71420 }, { "grad_norm": 0.11971092969179153, "learning_rate": 2.0705793954014528e-05, "loss": 0.0128, "step": 71430 }, { "grad_norm": 0.25973665714263916, "learning_rate": 2.06923959219589e-05, "loss": 0.0166, "step": 71440 }, { "grad_norm": 0.4772321879863739, "learning_rate": 2.0679001094938793e-05, "loss": 0.0188, "step": 71450 }, { "grad_norm": 0.2232261747121811, "learning_rate": 2.066560947441904e-05, "loss": 0.0172, "step": 71460 }, { "grad_norm": 1.8980250358581543, "learning_rate": 2.0652221061864152e-05, "loss": 0.019, "step": 71470 }, { "grad_norm": 0.21278691291809082, "learning_rate": 2.0638835858738226e-05, "loss": 0.0182, "step": 71480 }, { "grad_norm": 0.4083247482776642, "learning_rate": 2.0625453866505068e-05, "loss": 0.0196, "step": 71490 }, { "grad_norm": 0.20623207092285156, "learning_rate": 2.0612075086628123e-05, "loss": 0.0222, "step": 71500 }, { "grad_norm": 0.16645070910453796, "learning_rate": 2.059869952057045e-05, "loss": 0.0158, "step": 71510 }, { "grad_norm": 0.3401128947734833, "learning_rate": 2.0585327169794793e-05, "loss": 0.0232, "step": 71520 }, { "grad_norm": 0.21363820135593414, "learning_rate": 2.0571958035763534e-05, "loss": 0.0131, "step": 71530 }, { "grad_norm": 0.22287486493587494, "learning_rate": 2.0558592119938715e-05, "loss": 0.0136, "step": 71540 }, { "grad_norm": 0.21311065554618835, "learning_rate": 2.054522942378198e-05, "loss": 0.0155, "step": 71550 }, { "grad_norm": 0.3374159038066864, "learning_rate": 2.053186994875469e-05, "loss": 0.0181, "step": 71560 }, { "grad_norm": 0.2863840162754059, "learning_rate": 2.0518513696317775e-05, "loss": 0.0194, "step": 71570 }, { "grad_norm": 0.21954913437366486, "learning_rate": 2.050516066793188e-05, "loss": 0.0244, "step": 71580 }, { "grad_norm": 0.22517001628875732, "learning_rate": 2.0491810865057276e-05, "loss": 0.016, "step": 71590 }, { "grad_norm": 0.321911096572876, "learning_rate": 2.0478464289153882e-05, "loss": 0.018, "step": 71600 }, { "grad_norm": 0.19591039419174194, "learning_rate": 2.0465120941681232e-05, "loss": 0.017, "step": 71610 }, { "grad_norm": 0.3003051280975342, "learning_rate": 2.0451780824098566e-05, "loss": 0.0165, "step": 71620 }, { "grad_norm": 0.2481805682182312, "learning_rate": 2.043844393786471e-05, "loss": 0.0155, "step": 71630 }, { "grad_norm": 0.4628787040710449, "learning_rate": 2.0425110284438193e-05, "loss": 0.0183, "step": 71640 }, { "grad_norm": 0.3085075318813324, "learning_rate": 2.041177986527712e-05, "loss": 0.0133, "step": 71650 }, { "grad_norm": 0.2363879233598709, "learning_rate": 2.0398452681839347e-05, "loss": 0.0167, "step": 71660 }, { "grad_norm": 0.30562251806259155, "learning_rate": 2.0385128735582272e-05, "loss": 0.0293, "step": 71670 }, { "grad_norm": 0.22438789904117584, "learning_rate": 2.0371808027963003e-05, "loss": 0.0142, "step": 71680 }, { "grad_norm": 0.7656699419021606, "learning_rate": 2.035849056043825e-05, "loss": 0.0222, "step": 71690 }, { "grad_norm": 0.3693995773792267, "learning_rate": 2.03451763344644e-05, "loss": 0.0202, "step": 71700 }, { "grad_norm": 0.17643755674362183, "learning_rate": 2.0331865351497482e-05, "loss": 0.014, "step": 71710 }, { "grad_norm": 0.5287994146347046, "learning_rate": 2.0318557612993185e-05, "loss": 0.0184, "step": 71720 }, { "grad_norm": 0.33413612842559814, "learning_rate": 2.0305253120406782e-05, "loss": 0.019, "step": 71730 }, { "grad_norm": 0.25831812620162964, "learning_rate": 2.029195187519325e-05, "loss": 0.0173, "step": 71740 }, { "grad_norm": 0.30629852414131165, "learning_rate": 2.0278653878807217e-05, "loss": 0.0189, "step": 71750 }, { "grad_norm": 0.20284004509449005, "learning_rate": 2.0265359132702892e-05, "loss": 0.0188, "step": 71760 }, { "grad_norm": 0.16541606187820435, "learning_rate": 2.0252067638334182e-05, "loss": 0.0179, "step": 71770 }, { "grad_norm": 0.19376175105571747, "learning_rate": 2.023877939715464e-05, "loss": 0.0169, "step": 71780 }, { "grad_norm": 0.23395371437072754, "learning_rate": 2.0225494410617448e-05, "loss": 0.0164, "step": 71790 }, { "grad_norm": 0.589833676815033, "learning_rate": 2.02122126801754e-05, "loss": 0.0153, "step": 71800 }, { "grad_norm": 0.21827825903892517, "learning_rate": 2.0198934207281018e-05, "loss": 0.0182, "step": 71810 }, { "grad_norm": 0.2669699490070343, "learning_rate": 2.0185658993386357e-05, "loss": 0.0174, "step": 71820 }, { "grad_norm": 0.25371482968330383, "learning_rate": 2.0172387039943226e-05, "loss": 0.0154, "step": 71830 }, { "grad_norm": 0.24745631217956543, "learning_rate": 2.0159118348402968e-05, "loss": 0.0226, "step": 71840 }, { "grad_norm": 0.186926931142807, "learning_rate": 2.0145852920216697e-05, "loss": 0.0176, "step": 71850 }, { "grad_norm": 0.5035653114318848, "learning_rate": 2.0132590756835045e-05, "loss": 0.023, "step": 71860 }, { "grad_norm": 0.2816905677318573, "learning_rate": 2.0119331859708386e-05, "loss": 0.0236, "step": 71870 }, { "grad_norm": 0.578622579574585, "learning_rate": 2.0106076230286654e-05, "loss": 0.0228, "step": 71880 }, { "grad_norm": 0.26257121562957764, "learning_rate": 2.0092823870019488e-05, "loss": 0.0174, "step": 71890 }, { "grad_norm": 0.27611324191093445, "learning_rate": 2.0079574780356115e-05, "loss": 0.0159, "step": 71900 }, { "grad_norm": 0.300435334444046, "learning_rate": 2.0066328962745486e-05, "loss": 0.0185, "step": 71910 }, { "grad_norm": 0.2109588086605072, "learning_rate": 2.00530864186361e-05, "loss": 0.018, "step": 71920 }, { "grad_norm": 0.18330056965351105, "learning_rate": 2.003984714947618e-05, "loss": 0.0196, "step": 71930 }, { "grad_norm": 0.3566165864467621, "learning_rate": 2.0026611156713515e-05, "loss": 0.0173, "step": 71940 }, { "grad_norm": 0.2384694367647171, "learning_rate": 2.0013378441795584e-05, "loss": 0.0114, "step": 71950 }, { "grad_norm": 0.24870014190673828, "learning_rate": 2.0000149006169506e-05, "loss": 0.019, "step": 71960 }, { "grad_norm": 0.3252832293510437, "learning_rate": 1.9986922851282048e-05, "loss": 0.0181, "step": 71970 }, { "grad_norm": 0.2723969519138336, "learning_rate": 1.997369997857956e-05, "loss": 0.0207, "step": 71980 }, { "grad_norm": 0.3160371482372284, "learning_rate": 1.9960480389508103e-05, "loss": 0.0166, "step": 71990 }, { "grad_norm": 0.3322054445743561, "learning_rate": 1.9947264085513363e-05, "loss": 0.0204, "step": 72000 }, { "grad_norm": 0.3781481385231018, "learning_rate": 1.993405106804062e-05, "loss": 0.0184, "step": 72010 }, { "grad_norm": 0.28877052664756775, "learning_rate": 1.9920841338534867e-05, "loss": 0.018, "step": 72020 }, { "grad_norm": 0.29289594292640686, "learning_rate": 1.9907634898440646e-05, "loss": 0.0142, "step": 72030 }, { "grad_norm": 0.2526799738407135, "learning_rate": 1.989443174920227e-05, "loss": 0.0135, "step": 72040 }, { "grad_norm": 0.3383001983165741, "learning_rate": 1.988123189226355e-05, "loss": 0.0272, "step": 72050 }, { "grad_norm": 0.2927813231945038, "learning_rate": 1.986803532906805e-05, "loss": 0.0272, "step": 72060 }, { "grad_norm": 0.18782640993595123, "learning_rate": 1.985484206105889e-05, "loss": 0.0177, "step": 72070 }, { "grad_norm": 0.2628365755081177, "learning_rate": 1.9841652089678887e-05, "loss": 0.0136, "step": 72080 }, { "grad_norm": 0.2373044341802597, "learning_rate": 1.9828465416370435e-05, "loss": 0.0227, "step": 72090 }, { "grad_norm": 0.253135085105896, "learning_rate": 1.9815282042575685e-05, "loss": 0.0234, "step": 72100 }, { "grad_norm": 0.18422827124595642, "learning_rate": 1.9802101969736285e-05, "loss": 0.0207, "step": 72110 }, { "grad_norm": 0.5936967730522156, "learning_rate": 1.9788925199293633e-05, "loss": 0.0235, "step": 72120 }, { "grad_norm": 0.30330345034599304, "learning_rate": 1.977575173268868e-05, "loss": 0.0195, "step": 72130 }, { "grad_norm": 0.2214639037847519, "learning_rate": 1.9762581571362094e-05, "loss": 0.0176, "step": 72140 }, { "grad_norm": 0.30347388982772827, "learning_rate": 1.9749414716754112e-05, "loss": 0.023, "step": 72150 }, { "grad_norm": 0.2855777144432068, "learning_rate": 1.973625117030465e-05, "loss": 0.0194, "step": 72160 }, { "grad_norm": 0.2035660743713379, "learning_rate": 1.972309093345326e-05, "loss": 0.0187, "step": 72170 }, { "grad_norm": 0.24374420940876007, "learning_rate": 1.970993400763915e-05, "loss": 0.0252, "step": 72180 }, { "grad_norm": 0.18790805339813232, "learning_rate": 1.969678039430109e-05, "loss": 0.0271, "step": 72190 }, { "grad_norm": 0.343091756105423, "learning_rate": 1.9683630094877564e-05, "loss": 0.0219, "step": 72200 }, { "grad_norm": 0.24808870255947113, "learning_rate": 1.9670483110806696e-05, "loss": 0.0225, "step": 72210 }, { "grad_norm": 0.25687745213508606, "learning_rate": 1.9657339443526172e-05, "loss": 0.0222, "step": 72220 }, { "grad_norm": 0.27036336064338684, "learning_rate": 1.964419909447339e-05, "loss": 0.0206, "step": 72230 }, { "grad_norm": 0.29758545756340027, "learning_rate": 1.963106206508536e-05, "loss": 0.0135, "step": 72240 }, { "grad_norm": 0.17909416556358337, "learning_rate": 1.9617928356798738e-05, "loss": 0.0133, "step": 72250 }, { "grad_norm": 2.9691336154937744, "learning_rate": 1.960479797104977e-05, "loss": 0.0235, "step": 72260 }, { "grad_norm": 0.19409798085689545, "learning_rate": 1.9591670909274418e-05, "loss": 0.0158, "step": 72270 }, { "grad_norm": 0.2343316227197647, "learning_rate": 1.9578547172908184e-05, "loss": 0.0203, "step": 72280 }, { "grad_norm": 0.26942211389541626, "learning_rate": 1.9565426763386325e-05, "loss": 0.0145, "step": 72290 }, { "grad_norm": 0.24771441519260406, "learning_rate": 1.9552309682143622e-05, "loss": 0.018, "step": 72300 }, { "grad_norm": 0.22465795278549194, "learning_rate": 1.9539195930614574e-05, "loss": 0.0182, "step": 72310 }, { "grad_norm": 0.20714202523231506, "learning_rate": 1.952608551023324e-05, "loss": 0.0155, "step": 72320 }, { "grad_norm": 0.3116222023963928, "learning_rate": 1.9512978422433397e-05, "loss": 0.0179, "step": 72330 }, { "grad_norm": 0.27090391516685486, "learning_rate": 1.9499874668648376e-05, "loss": 0.0153, "step": 72340 }, { "grad_norm": 0.18203648924827576, "learning_rate": 1.9486774250311213e-05, "loss": 0.0153, "step": 72350 }, { "grad_norm": 0.30164337158203125, "learning_rate": 1.9473677168854537e-05, "loss": 0.0198, "step": 72360 }, { "grad_norm": 0.930228590965271, "learning_rate": 1.946058342571065e-05, "loss": 0.0237, "step": 72370 }, { "grad_norm": 0.3031705319881439, "learning_rate": 1.9447493022311424e-05, "loss": 0.0173, "step": 72380 }, { "grad_norm": 0.3277326226234436, "learning_rate": 1.9434405960088442e-05, "loss": 0.0183, "step": 72390 }, { "grad_norm": 0.285334974527359, "learning_rate": 1.9421322240472857e-05, "loss": 0.0178, "step": 72400 }, { "grad_norm": 0.2255663424730301, "learning_rate": 1.9408241864895494e-05, "loss": 0.0199, "step": 72410 }, { "grad_norm": 0.35892435908317566, "learning_rate": 1.939516483478681e-05, "loss": 0.0188, "step": 72420 }, { "grad_norm": 0.38697466254234314, "learning_rate": 1.9382091151576902e-05, "loss": 0.0176, "step": 72430 }, { "grad_norm": 0.36630961298942566, "learning_rate": 1.9369020816695454e-05, "loss": 0.0208, "step": 72440 }, { "grad_norm": 0.266996294260025, "learning_rate": 1.935595383157184e-05, "loss": 0.0134, "step": 72450 }, { "grad_norm": 0.26895982027053833, "learning_rate": 1.9342890197635057e-05, "loss": 0.0133, "step": 72460 }, { "grad_norm": 0.3718297779560089, "learning_rate": 1.9329829916313685e-05, "loss": 0.0169, "step": 72470 }, { "grad_norm": 0.2786268889904022, "learning_rate": 1.9316772989036013e-05, "loss": 0.0206, "step": 72480 }, { "grad_norm": 0.2246483713388443, "learning_rate": 1.930371941722991e-05, "loss": 0.0126, "step": 72490 }, { "grad_norm": 0.18169823288917542, "learning_rate": 1.929066920232292e-05, "loss": 0.0157, "step": 72500 }, { "grad_norm": 0.3488375246524811, "learning_rate": 1.9277622345742152e-05, "loss": 0.0183, "step": 72510 }, { "grad_norm": 0.2869708240032196, "learning_rate": 1.926457884891444e-05, "loss": 0.0196, "step": 72520 }, { "grad_norm": 0.31652578711509705, "learning_rate": 1.9251538713266147e-05, "loss": 0.0157, "step": 72530 }, { "grad_norm": 0.37961480021476746, "learning_rate": 1.9238501940223358e-05, "loss": 0.0236, "step": 72540 }, { "grad_norm": 0.3960185647010803, "learning_rate": 1.922546853121175e-05, "loss": 0.0233, "step": 72550 }, { "grad_norm": 0.24953636527061462, "learning_rate": 1.9212438487656647e-05, "loss": 0.026, "step": 72560 }, { "grad_norm": 0.40233370661735535, "learning_rate": 1.9199411810982965e-05, "loss": 0.0209, "step": 72570 }, { "grad_norm": 0.2738818824291229, "learning_rate": 1.918638850261532e-05, "loss": 0.0154, "step": 72580 }, { "grad_norm": 0.18217939138412476, "learning_rate": 1.9173368563977885e-05, "loss": 0.0177, "step": 72590 }, { "grad_norm": 0.2521006166934967, "learning_rate": 1.9160351996494513e-05, "loss": 0.0148, "step": 72600 }, { "grad_norm": 0.23888447880744934, "learning_rate": 1.9147338801588677e-05, "loss": 0.0208, "step": 72610 }, { "grad_norm": 0.1465705931186676, "learning_rate": 1.913432898068351e-05, "loss": 0.0196, "step": 72620 }, { "grad_norm": 0.3024745285511017, "learning_rate": 1.912132253520169e-05, "loss": 0.0186, "step": 72630 }, { "grad_norm": 0.28341153264045715, "learning_rate": 1.910831946656564e-05, "loss": 0.025, "step": 72640 }, { "grad_norm": 0.26014024019241333, "learning_rate": 1.909531977619731e-05, "loss": 0.0137, "step": 72650 }, { "grad_norm": 0.42102766036987305, "learning_rate": 1.908232346551834e-05, "loss": 0.0124, "step": 72660 }, { "grad_norm": 0.321414589881897, "learning_rate": 1.9069330535949993e-05, "loss": 0.0172, "step": 72670 }, { "grad_norm": 0.2299356609582901, "learning_rate": 1.9056340988913167e-05, "loss": 0.0131, "step": 72680 }, { "grad_norm": 0.19137923419475555, "learning_rate": 1.9043354825828357e-05, "loss": 0.0171, "step": 72690 }, { "grad_norm": 0.25285565853118896, "learning_rate": 1.9030372048115724e-05, "loss": 0.0173, "step": 72700 }, { "grad_norm": 0.3191892206668854, "learning_rate": 1.9017392657195056e-05, "loss": 0.0262, "step": 72710 }, { "grad_norm": 0.3258081078529358, "learning_rate": 1.9004416654485724e-05, "loss": 0.0179, "step": 72720 }, { "grad_norm": 0.3130019009113312, "learning_rate": 1.8991444041406785e-05, "loss": 0.0203, "step": 72730 }, { "grad_norm": 0.12926122546195984, "learning_rate": 1.8978474819376912e-05, "loss": 0.0204, "step": 72740 }, { "grad_norm": 0.26390793919563293, "learning_rate": 1.8965508989814406e-05, "loss": 0.0205, "step": 72750 }, { "grad_norm": 0.28750669956207275, "learning_rate": 1.8952546554137158e-05, "loss": 0.0167, "step": 72760 }, { "grad_norm": 0.3183882534503937, "learning_rate": 1.8939587513762757e-05, "loss": 0.0133, "step": 72770 }, { "grad_norm": 0.32555800676345825, "learning_rate": 1.892663187010835e-05, "loss": 0.0156, "step": 72780 }, { "grad_norm": 0.2916109263896942, "learning_rate": 1.891367962459077e-05, "loss": 0.0178, "step": 72790 }, { "grad_norm": 1.2492241859436035, "learning_rate": 1.890073077862644e-05, "loss": 0.0138, "step": 72800 }, { "grad_norm": 0.27420973777770996, "learning_rate": 1.888778533363146e-05, "loss": 0.0154, "step": 72810 }, { "grad_norm": 0.14434842765331268, "learning_rate": 1.8874843291021478e-05, "loss": 0.0158, "step": 72820 }, { "grad_norm": 1.3160951137542725, "learning_rate": 1.8861904652211853e-05, "loss": 0.0168, "step": 72830 }, { "grad_norm": 0.3246558904647827, "learning_rate": 1.8848969418617507e-05, "loss": 0.0147, "step": 72840 }, { "grad_norm": 0.24141621589660645, "learning_rate": 1.8836037591653044e-05, "loss": 0.0169, "step": 72850 }, { "grad_norm": 0.23857955634593964, "learning_rate": 1.882310917273262e-05, "loss": 0.0181, "step": 72860 }, { "grad_norm": 0.32021042704582214, "learning_rate": 1.881018416327014e-05, "loss": 0.0171, "step": 72870 }, { "grad_norm": 0.23279596865177155, "learning_rate": 1.8797262564679014e-05, "loss": 0.0221, "step": 72880 }, { "grad_norm": 0.19303518533706665, "learning_rate": 1.8784344378372355e-05, "loss": 0.0147, "step": 72890 }, { "grad_norm": 0.2852869927883148, "learning_rate": 1.8771429605762843e-05, "loss": 0.0169, "step": 72900 }, { "grad_norm": 0.21223163604736328, "learning_rate": 1.8758518248262835e-05, "loss": 0.0176, "step": 72910 }, { "grad_norm": 0.2692083418369293, "learning_rate": 1.8745610307284305e-05, "loss": 0.0237, "step": 72920 }, { "grad_norm": 0.3100298345088959, "learning_rate": 1.8732705784238862e-05, "loss": 0.017, "step": 72930 }, { "grad_norm": 0.28054994344711304, "learning_rate": 1.8719804680537682e-05, "loss": 0.0209, "step": 72940 }, { "grad_norm": 0.18852657079696655, "learning_rate": 1.870690699759164e-05, "loss": 0.0145, "step": 72950 }, { "grad_norm": 0.2919384241104126, "learning_rate": 1.8694012736811213e-05, "loss": 0.0195, "step": 72960 }, { "grad_norm": 0.2849702835083008, "learning_rate": 1.8681121899606463e-05, "loss": 0.0159, "step": 72970 }, { "grad_norm": 0.133802130818367, "learning_rate": 1.8668234487387144e-05, "loss": 0.019, "step": 72980 }, { "grad_norm": 0.22391968965530396, "learning_rate": 1.8655350501562592e-05, "loss": 0.0132, "step": 72990 }, { "grad_norm": 0.22230595350265503, "learning_rate": 1.8642469943541796e-05, "loss": 0.0279, "step": 73000 }, { "grad_norm": 0.24513530731201172, "learning_rate": 1.8629592814733328e-05, "loss": 0.0149, "step": 73010 }, { "grad_norm": 0.21384163200855255, "learning_rate": 1.8616719116545438e-05, "loss": 0.0186, "step": 73020 }, { "grad_norm": 0.17776507139205933, "learning_rate": 1.8603848850385952e-05, "loss": 0.011, "step": 73030 }, { "grad_norm": 0.21619245409965515, "learning_rate": 1.859098201766236e-05, "loss": 0.0159, "step": 73040 }, { "grad_norm": 0.1604684293270111, "learning_rate": 1.8578118619781726e-05, "loss": 0.02, "step": 73050 }, { "grad_norm": 0.27823588252067566, "learning_rate": 1.856525865815083e-05, "loss": 0.0232, "step": 73060 }, { "grad_norm": 0.1785169392824173, "learning_rate": 1.8552402134175966e-05, "loss": 0.0136, "step": 73070 }, { "grad_norm": 0.2750975489616394, "learning_rate": 1.8539549049263143e-05, "loss": 0.0169, "step": 73080 }, { "grad_norm": 0.31710413098335266, "learning_rate": 1.8526699404817917e-05, "loss": 0.0199, "step": 73090 }, { "grad_norm": 0.32328248023986816, "learning_rate": 1.8513853202245546e-05, "loss": 0.0232, "step": 73100 }, { "grad_norm": 0.20658372342586517, "learning_rate": 1.850101044295081e-05, "loss": 0.0109, "step": 73110 }, { "grad_norm": 0.3639248311519623, "learning_rate": 1.848817112833825e-05, "loss": 0.0154, "step": 73120 }, { "grad_norm": 1.683560848236084, "learning_rate": 1.8475335259811905e-05, "loss": 0.0166, "step": 73130 }, { "grad_norm": 0.28785133361816406, "learning_rate": 1.8462502838775508e-05, "loss": 0.0183, "step": 73140 }, { "grad_norm": 0.27603548765182495, "learning_rate": 1.8449673866632378e-05, "loss": 0.0209, "step": 73150 }, { "grad_norm": 0.15410716831684113, "learning_rate": 1.8436848344785473e-05, "loss": 0.0186, "step": 73160 }, { "grad_norm": 0.2088192105293274, "learning_rate": 1.8424026274637378e-05, "loss": 0.0139, "step": 73170 }, { "grad_norm": 0.2547758221626282, "learning_rate": 1.8411207657590312e-05, "loss": 0.0157, "step": 73180 }, { "grad_norm": 0.19382448494434357, "learning_rate": 1.839839249504607e-05, "loss": 0.017, "step": 73190 }, { "grad_norm": 0.2611779570579529, "learning_rate": 1.8385580788406108e-05, "loss": 0.0157, "step": 73200 }, { "grad_norm": 0.19813480973243713, "learning_rate": 1.837277253907152e-05, "loss": 0.0187, "step": 73210 }, { "grad_norm": 0.23506726324558258, "learning_rate": 1.8359967748442957e-05, "loss": 0.0177, "step": 73220 }, { "grad_norm": 0.29130110144615173, "learning_rate": 1.8347166417920774e-05, "loss": 0.0127, "step": 73230 }, { "grad_norm": 0.14945818483829498, "learning_rate": 1.833436854890484e-05, "loss": 0.019, "step": 73240 }, { "grad_norm": 0.29909640550613403, "learning_rate": 1.8321574142794796e-05, "loss": 0.0157, "step": 73250 }, { "grad_norm": 0.20246009528636932, "learning_rate": 1.8308783200989748e-05, "loss": 0.0161, "step": 73260 }, { "grad_norm": 0.20988582074642181, "learning_rate": 1.829599572488855e-05, "loss": 0.0154, "step": 73270 }, { "grad_norm": 0.23369623720645905, "learning_rate": 1.828321171588957e-05, "loss": 0.0176, "step": 73280 }, { "grad_norm": 0.24264247715473175, "learning_rate": 1.827043117539089e-05, "loss": 0.0167, "step": 73290 }, { "grad_norm": 0.24765901267528534, "learning_rate": 1.8257654104790116e-05, "loss": 0.0176, "step": 73300 }, { "grad_norm": 0.25901442766189575, "learning_rate": 1.82448805054846e-05, "loss": 0.0189, "step": 73310 }, { "grad_norm": 0.18036583065986633, "learning_rate": 1.823211037887119e-05, "loss": 0.0195, "step": 73320 }, { "grad_norm": 0.33153656125068665, "learning_rate": 1.821934372634645e-05, "loss": 0.0149, "step": 73330 }, { "grad_norm": 0.20231781899929047, "learning_rate": 1.8206580549306472e-05, "loss": 0.0145, "step": 73340 }, { "grad_norm": 0.12901709973812103, "learning_rate": 1.819382084914707e-05, "loss": 0.0139, "step": 73350 }, { "grad_norm": 0.2951401472091675, "learning_rate": 1.8181064627263577e-05, "loss": 0.0133, "step": 73360 }, { "grad_norm": 0.27043864130973816, "learning_rate": 1.8168311885051022e-05, "loss": 0.0208, "step": 73370 }, { "grad_norm": 0.20350761711597443, "learning_rate": 1.815556262390402e-05, "loss": 0.0145, "step": 73380 }, { "grad_norm": 0.25771746039390564, "learning_rate": 1.8142816845216826e-05, "loss": 0.0147, "step": 73390 }, { "grad_norm": 0.17986486852169037, "learning_rate": 1.8130074550383276e-05, "loss": 0.015, "step": 73400 }, { "grad_norm": 0.33032870292663574, "learning_rate": 1.8117335740796848e-05, "loss": 0.0185, "step": 73410 }, { "grad_norm": 0.5143213272094727, "learning_rate": 1.810460041785067e-05, "loss": 0.0194, "step": 73420 }, { "grad_norm": 0.14625313878059387, "learning_rate": 1.809186858293742e-05, "loss": 0.0256, "step": 73430 }, { "grad_norm": 0.2505306899547577, "learning_rate": 1.8079140237449448e-05, "loss": 0.0216, "step": 73440 }, { "grad_norm": 0.6385701298713684, "learning_rate": 1.8066415382778707e-05, "loss": 0.0229, "step": 73450 }, { "grad_norm": 0.171608105301857, "learning_rate": 1.8053694020316786e-05, "loss": 0.015, "step": 73460 }, { "grad_norm": 0.28240782022476196, "learning_rate": 1.8040976151454837e-05, "loss": 0.0164, "step": 73470 }, { "grad_norm": 0.2232588529586792, "learning_rate": 1.8028261777583704e-05, "loss": 0.0254, "step": 73480 }, { "grad_norm": 0.2747848927974701, "learning_rate": 1.8015550900093758e-05, "loss": 0.0197, "step": 73490 }, { "grad_norm": 0.2631928622722626, "learning_rate": 1.800284352037512e-05, "loss": 0.0194, "step": 73500 }, { "grad_norm": 0.22426612675189972, "learning_rate": 1.799013963981738e-05, "loss": 0.0149, "step": 73510 }, { "grad_norm": 0.31224194169044495, "learning_rate": 1.797743925980987e-05, "loss": 0.0197, "step": 73520 }, { "grad_norm": 0.17711780965328217, "learning_rate": 1.7964742381741435e-05, "loss": 0.0137, "step": 73530 }, { "grad_norm": 0.46911150217056274, "learning_rate": 1.7952049007000627e-05, "loss": 0.0165, "step": 73540 }, { "grad_norm": 0.30899739265441895, "learning_rate": 1.7939359136975535e-05, "loss": 0.0182, "step": 73550 }, { "grad_norm": 0.1678076684474945, "learning_rate": 1.7926672773053932e-05, "loss": 0.0131, "step": 73560 }, { "grad_norm": 0.23641660809516907, "learning_rate": 1.791398991662317e-05, "loss": 0.0232, "step": 73570 }, { "grad_norm": 0.1637393683195114, "learning_rate": 1.7901310569070244e-05, "loss": 0.0199, "step": 73580 }, { "grad_norm": 0.30874407291412354, "learning_rate": 1.7888634731781724e-05, "loss": 0.0206, "step": 73590 }, { "grad_norm": 0.28518298268318176, "learning_rate": 1.7875962406143844e-05, "loss": 0.0117, "step": 73600 }, { "grad_norm": 0.2638711929321289, "learning_rate": 1.7863293593542403e-05, "loss": 0.0308, "step": 73610 }, { "grad_norm": 0.2103806734085083, "learning_rate": 1.7850628295362853e-05, "loss": 0.0154, "step": 73620 }, { "grad_norm": 0.3407757878303528, "learning_rate": 1.7837966512990263e-05, "loss": 0.019, "step": 73630 }, { "grad_norm": 0.38086435198783875, "learning_rate": 1.782530824780931e-05, "loss": 0.0197, "step": 73640 }, { "grad_norm": 0.20316867530345917, "learning_rate": 1.7812653501204263e-05, "loss": 0.0207, "step": 73650 }, { "grad_norm": 0.2537958025932312, "learning_rate": 1.780000227455903e-05, "loss": 0.0194, "step": 73660 }, { "grad_norm": 0.18211409449577332, "learning_rate": 1.778735456925716e-05, "loss": 0.021, "step": 73670 }, { "grad_norm": 0.31894224882125854, "learning_rate": 1.7774710386681737e-05, "loss": 0.0143, "step": 73680 }, { "grad_norm": 0.2400982826948166, "learning_rate": 1.7762069728215545e-05, "loss": 0.0202, "step": 73690 }, { "grad_norm": 0.256893515586853, "learning_rate": 1.7749432595240932e-05, "loss": 0.0137, "step": 73700 }, { "grad_norm": 0.2516809403896332, "learning_rate": 1.773679898913991e-05, "loss": 0.0187, "step": 73710 }, { "grad_norm": 0.4225715696811676, "learning_rate": 1.7724168911294015e-05, "loss": 0.0157, "step": 73720 }, { "grad_norm": 0.5162205696105957, "learning_rate": 1.7711542363084504e-05, "loss": 0.017, "step": 73730 }, { "grad_norm": 0.2043614238500595, "learning_rate": 1.7698919345892158e-05, "loss": 0.0146, "step": 73740 }, { "grad_norm": 0.2145114690065384, "learning_rate": 1.768629986109743e-05, "loss": 0.0159, "step": 73750 }, { "grad_norm": 0.43279755115509033, "learning_rate": 1.7673683910080357e-05, "loss": 0.0184, "step": 73760 }, { "grad_norm": 0.20847570896148682, "learning_rate": 1.7661071494220638e-05, "loss": 0.0189, "step": 73770 }, { "grad_norm": 0.2143632471561432, "learning_rate": 1.7648462614897498e-05, "loss": 0.0134, "step": 73780 }, { "grad_norm": 0.26248061656951904, "learning_rate": 1.7635857273489863e-05, "loss": 0.0226, "step": 73790 }, { "grad_norm": 0.356224924325943, "learning_rate": 1.7623255471376198e-05, "loss": 0.0161, "step": 73800 }, { "grad_norm": 0.28709033131599426, "learning_rate": 1.7610657209934646e-05, "loss": 0.0207, "step": 73810 }, { "grad_norm": 0.2625741958618164, "learning_rate": 1.7598062490542915e-05, "loss": 0.018, "step": 73820 }, { "grad_norm": 0.5612736940383911, "learning_rate": 1.7585471314578377e-05, "loss": 0.0195, "step": 73830 }, { "grad_norm": 0.39719533920288086, "learning_rate": 1.7572883683417945e-05, "loss": 0.0193, "step": 73840 }, { "grad_norm": 0.3866536319255829, "learning_rate": 1.7560299598438212e-05, "loss": 0.0209, "step": 73850 }, { "grad_norm": 0.22492615878582, "learning_rate": 1.7547719061015334e-05, "loss": 0.0165, "step": 73860 }, { "grad_norm": 0.17819222807884216, "learning_rate": 1.75351420725251e-05, "loss": 0.0153, "step": 73870 }, { "grad_norm": 0.3208363652229309, "learning_rate": 1.7522568634342923e-05, "loss": 0.0185, "step": 73880 }, { "grad_norm": 0.23473237454891205, "learning_rate": 1.7509998747843825e-05, "loss": 0.0128, "step": 73890 }, { "grad_norm": 0.19252096116542816, "learning_rate": 1.7497432414402403e-05, "loss": 0.0169, "step": 73900 }, { "grad_norm": 0.5913712978363037, "learning_rate": 1.7484869635392903e-05, "loss": 0.0192, "step": 73910 }, { "grad_norm": 0.38727259635925293, "learning_rate": 1.747231041218919e-05, "loss": 0.017, "step": 73920 }, { "grad_norm": 0.31932252645492554, "learning_rate": 1.7459754746164692e-05, "loss": 0.0155, "step": 73930 }, { "grad_norm": 0.24622364342212677, "learning_rate": 1.744720263869249e-05, "loss": 0.0257, "step": 73940 }, { "grad_norm": 0.21306860446929932, "learning_rate": 1.7434654091145263e-05, "loss": 0.0188, "step": 73950 }, { "grad_norm": 0.25946366786956787, "learning_rate": 1.742210910489532e-05, "loss": 0.0166, "step": 73960 }, { "grad_norm": 0.23291043937206268, "learning_rate": 1.7409567681314532e-05, "loss": 0.0151, "step": 73970 }, { "grad_norm": 1.4841609001159668, "learning_rate": 1.7397029821774434e-05, "loss": 0.0185, "step": 73980 }, { "grad_norm": 0.2957931458950043, "learning_rate": 1.7384495527646126e-05, "loss": 0.0168, "step": 73990 }, { "grad_norm": 0.16403406858444214, "learning_rate": 1.737196480030035e-05, "loss": 0.0205, "step": 74000 }, { "grad_norm": 0.28876903653144836, "learning_rate": 1.7359437641107447e-05, "loss": 0.0205, "step": 74010 }, { "grad_norm": 0.19640976190567017, "learning_rate": 1.7346914051437387e-05, "loss": 0.0203, "step": 74020 }, { "grad_norm": 1.2555302381515503, "learning_rate": 1.733439403265969e-05, "loss": 0.0157, "step": 74030 }, { "grad_norm": 0.2727750837802887, "learning_rate": 1.732187758614357e-05, "loss": 0.0274, "step": 74040 }, { "grad_norm": 0.2817787230014801, "learning_rate": 1.730936471325777e-05, "loss": 0.0164, "step": 74050 }, { "grad_norm": 0.2191927134990692, "learning_rate": 1.7296855415370718e-05, "loss": 0.0232, "step": 74060 }, { "grad_norm": 0.3709229826927185, "learning_rate": 1.728434969385035e-05, "loss": 0.015, "step": 74070 }, { "grad_norm": 0.17678190767765045, "learning_rate": 1.7271847550064346e-05, "loss": 0.0124, "step": 74080 }, { "grad_norm": 0.20645084977149963, "learning_rate": 1.725934898537987e-05, "loss": 0.0206, "step": 74090 }, { "grad_norm": 0.6509246826171875, "learning_rate": 1.7246854001163783e-05, "loss": 0.0147, "step": 74100 }, { "grad_norm": 0.31193435192108154, "learning_rate": 1.7234362598782476e-05, "loss": 0.0159, "step": 74110 }, { "grad_norm": 0.35401543974876404, "learning_rate": 1.722187477960202e-05, "loss": 0.0204, "step": 74120 }, { "grad_norm": 0.2083984911441803, "learning_rate": 1.7209390544988057e-05, "loss": 0.0186, "step": 74130 }, { "grad_norm": 0.31470730900764465, "learning_rate": 1.7196909896305856e-05, "loss": 0.0249, "step": 74140 }, { "grad_norm": 0.26858067512512207, "learning_rate": 1.718443283492026e-05, "loss": 0.0182, "step": 74150 }, { "grad_norm": 0.22551329433918, "learning_rate": 1.7171959362195746e-05, "loss": 0.0191, "step": 74160 }, { "grad_norm": 0.32157671451568604, "learning_rate": 1.715948947949642e-05, "loss": 0.0177, "step": 74170 }, { "grad_norm": 0.3485499620437622, "learning_rate": 1.714702318818593e-05, "loss": 0.0217, "step": 74180 }, { "grad_norm": 0.8047034740447998, "learning_rate": 1.7134560489627592e-05, "loss": 0.0186, "step": 74190 }, { "grad_norm": 0.27687376737594604, "learning_rate": 1.712210138518431e-05, "loss": 0.0174, "step": 74200 }, { "grad_norm": 0.3009498119354248, "learning_rate": 1.71096458762186e-05, "loss": 0.024, "step": 74210 }, { "grad_norm": 0.1806754618883133, "learning_rate": 1.7097193964092552e-05, "loss": 0.0177, "step": 74220 }, { "grad_norm": 0.25541558861732483, "learning_rate": 1.708474565016792e-05, "loss": 0.0172, "step": 74230 }, { "grad_norm": 0.32707223296165466, "learning_rate": 1.7072300935806e-05, "loss": 0.0176, "step": 74240 }, { "grad_norm": 0.317710280418396, "learning_rate": 1.705985982236776e-05, "loss": 0.0153, "step": 74250 }, { "grad_norm": 0.22279566526412964, "learning_rate": 1.7047422311213697e-05, "loss": 0.017, "step": 74260 }, { "grad_norm": 0.3512846827507019, "learning_rate": 1.703498840370401e-05, "loss": 0.0221, "step": 74270 }, { "grad_norm": 0.32185715436935425, "learning_rate": 1.702255810119842e-05, "loss": 0.0155, "step": 74280 }, { "grad_norm": 0.1671139895915985, "learning_rate": 1.7010131405056305e-05, "loss": 0.0139, "step": 74290 }, { "grad_norm": 0.3495543301105499, "learning_rate": 1.6997708316636613e-05, "loss": 0.0156, "step": 74300 }, { "grad_norm": 0.23519906401634216, "learning_rate": 1.6985288837297932e-05, "loss": 0.018, "step": 74310 }, { "grad_norm": 0.19875700771808624, "learning_rate": 1.6972872968398394e-05, "loss": 0.0159, "step": 74320 }, { "grad_norm": 0.24012857675552368, "learning_rate": 1.6960460711295845e-05, "loss": 0.0165, "step": 74330 }, { "grad_norm": 0.37176313996315, "learning_rate": 1.6948052067347623e-05, "loss": 0.0164, "step": 74340 }, { "grad_norm": 0.5625442862510681, "learning_rate": 1.693564703791075e-05, "loss": 0.0256, "step": 74350 }, { "grad_norm": 0.14995886385440826, "learning_rate": 1.6923245624341793e-05, "loss": 0.014, "step": 74360 }, { "grad_norm": 0.2563854455947876, "learning_rate": 1.6910847827996962e-05, "loss": 0.0202, "step": 74370 }, { "grad_norm": 0.2618156969547272, "learning_rate": 1.689845365023206e-05, "loss": 0.0198, "step": 74380 }, { "grad_norm": 0.3286064565181732, "learning_rate": 1.6886063092402505e-05, "loss": 0.0159, "step": 74390 }, { "grad_norm": 0.402049720287323, "learning_rate": 1.6873676155863317e-05, "loss": 0.0197, "step": 74400 }, { "grad_norm": 0.31726953387260437, "learning_rate": 1.6861292841969085e-05, "loss": 0.0142, "step": 74410 }, { "grad_norm": 0.2650703191757202, "learning_rate": 1.684891315207406e-05, "loss": 0.0229, "step": 74420 }, { "grad_norm": 0.200269415974617, "learning_rate": 1.6836537087532034e-05, "loss": 0.0131, "step": 74430 }, { "grad_norm": 0.24615687131881714, "learning_rate": 1.6824164649696465e-05, "loss": 0.0257, "step": 74440 }, { "grad_norm": 0.2594777047634125, "learning_rate": 1.681179583992034e-05, "loss": 0.0147, "step": 74450 }, { "grad_norm": 0.14997000992298126, "learning_rate": 1.679943065955636e-05, "loss": 0.0147, "step": 74460 }, { "grad_norm": 0.21866531670093536, "learning_rate": 1.678706910995671e-05, "loss": 0.0133, "step": 74470 }, { "grad_norm": 0.1860847771167755, "learning_rate": 1.6774711192473262e-05, "loss": 0.0187, "step": 74480 }, { "grad_norm": 0.33512160181999207, "learning_rate": 1.676235690845743e-05, "loss": 0.0156, "step": 74490 }, { "grad_norm": 0.4202205240726471, "learning_rate": 1.6750006259260285e-05, "loss": 0.0235, "step": 74500 }, { "grad_norm": 0.22291299700737, "learning_rate": 1.673765924623244e-05, "loss": 0.0184, "step": 74510 }, { "grad_norm": 0.37758660316467285, "learning_rate": 1.67253158707242e-05, "loss": 0.0148, "step": 74520 }, { "grad_norm": 0.2639842927455902, "learning_rate": 1.671297613408537e-05, "loss": 0.0154, "step": 74530 }, { "grad_norm": 0.30327409505844116, "learning_rate": 1.6700640037665444e-05, "loss": 0.0146, "step": 74540 }, { "grad_norm": 0.26011165976524353, "learning_rate": 1.668830758281344e-05, "loss": 0.017, "step": 74550 }, { "grad_norm": 0.20899683237075806, "learning_rate": 1.6675978770878052e-05, "loss": 0.0193, "step": 74560 }, { "grad_norm": 0.19020403921604156, "learning_rate": 1.6663653603207506e-05, "loss": 0.0194, "step": 74570 }, { "grad_norm": 0.2811110019683838, "learning_rate": 1.6651332081149685e-05, "loss": 0.0166, "step": 74580 }, { "grad_norm": 1.8134233951568604, "learning_rate": 1.6639014206052046e-05, "loss": 0.0204, "step": 74590 }, { "grad_norm": 0.21674145758152008, "learning_rate": 1.6626699979261657e-05, "loss": 0.0191, "step": 74600 }, { "grad_norm": 0.3265121281147003, "learning_rate": 1.6614389402125197e-05, "loss": 0.0155, "step": 74610 }, { "grad_norm": 0.46343284845352173, "learning_rate": 1.66020824759889e-05, "loss": 0.0211, "step": 74620 }, { "grad_norm": 0.2824503183364868, "learning_rate": 1.6589779202198675e-05, "loss": 0.0139, "step": 74630 }, { "grad_norm": 0.22269733250141144, "learning_rate": 1.6577479582099924e-05, "loss": 0.0216, "step": 74640 }, { "grad_norm": 0.43281590938568115, "learning_rate": 1.6565183617037788e-05, "loss": 0.03, "step": 74650 }, { "grad_norm": 0.36569228768348694, "learning_rate": 1.6552891308356894e-05, "loss": 0.0223, "step": 74660 }, { "grad_norm": 0.21618078649044037, "learning_rate": 1.6540602657401526e-05, "loss": 0.0268, "step": 74670 }, { "grad_norm": 0.28613218665122986, "learning_rate": 1.6528317665515537e-05, "loss": 0.0266, "step": 74680 }, { "grad_norm": 0.21332693099975586, "learning_rate": 1.6516036334042416e-05, "loss": 0.0186, "step": 74690 }, { "grad_norm": 0.19911381602287292, "learning_rate": 1.650375866432519e-05, "loss": 0.0186, "step": 74700 }, { "grad_norm": 0.18108567595481873, "learning_rate": 1.649148465770659e-05, "loss": 0.0124, "step": 74710 }, { "grad_norm": 0.3175783157348633, "learning_rate": 1.647921431552883e-05, "loss": 0.0133, "step": 74720 }, { "grad_norm": 0.27950039505958557, "learning_rate": 1.646694763913381e-05, "loss": 0.0204, "step": 74730 }, { "grad_norm": 0.31855887174606323, "learning_rate": 1.6454684629862966e-05, "loss": 0.0133, "step": 74740 }, { "grad_norm": 0.24482393264770508, "learning_rate": 1.6442425289057388e-05, "loss": 0.0158, "step": 74750 }, { "grad_norm": 0.33158063888549805, "learning_rate": 1.6430169618057718e-05, "loss": 0.0158, "step": 74760 }, { "grad_norm": 0.3340226411819458, "learning_rate": 1.641791761820422e-05, "loss": 0.0137, "step": 74770 }, { "grad_norm": 0.22261209785938263, "learning_rate": 1.640566929083676e-05, "loss": 0.0177, "step": 74780 }, { "grad_norm": 0.19573356211185455, "learning_rate": 1.6393424637294823e-05, "loss": 0.0257, "step": 74790 }, { "grad_norm": 0.19962270557880402, "learning_rate": 1.638118365891742e-05, "loss": 0.0185, "step": 74800 }, { "grad_norm": 0.23545660078525543, "learning_rate": 1.636894635704323e-05, "loss": 0.0199, "step": 74810 }, { "grad_norm": 0.2268936038017273, "learning_rate": 1.6356712733010517e-05, "loss": 0.0143, "step": 74820 }, { "grad_norm": 0.5452356934547424, "learning_rate": 1.6344482788157096e-05, "loss": 0.0212, "step": 74830 }, { "grad_norm": 0.5091189742088318, "learning_rate": 1.6332256523820444e-05, "loss": 0.0151, "step": 74840 }, { "grad_norm": 0.34144026041030884, "learning_rate": 1.6320033941337597e-05, "loss": 0.019, "step": 74850 }, { "grad_norm": 0.2703653872013092, "learning_rate": 1.6307815042045216e-05, "loss": 0.0154, "step": 74860 }, { "grad_norm": 0.28064146637916565, "learning_rate": 1.6295599827279513e-05, "loss": 0.0154, "step": 74870 }, { "grad_norm": 0.2759929299354553, "learning_rate": 1.6283388298376356e-05, "loss": 0.0167, "step": 74880 }, { "grad_norm": 0.34501391649246216, "learning_rate": 1.6271180456671125e-05, "loss": 0.0141, "step": 74890 }, { "grad_norm": 0.2345898300409317, "learning_rate": 1.625897630349893e-05, "loss": 0.0205, "step": 74900 }, { "grad_norm": 0.2466401904821396, "learning_rate": 1.6246775840194335e-05, "loss": 0.0188, "step": 74910 }, { "grad_norm": 0.28655627369880676, "learning_rate": 1.6234579068091604e-05, "loss": 0.015, "step": 74920 }, { "grad_norm": 0.349894255399704, "learning_rate": 1.6222385988524524e-05, "loss": 0.022, "step": 74930 }, { "grad_norm": 0.19576303660869598, "learning_rate": 1.6210196602826544e-05, "loss": 0.0197, "step": 74940 }, { "grad_norm": 0.251146525144577, "learning_rate": 1.6198010912330646e-05, "loss": 0.0207, "step": 74950 }, { "grad_norm": 0.27122804522514343, "learning_rate": 1.6185828918369456e-05, "loss": 0.0202, "step": 74960 }, { "grad_norm": 0.42562779784202576, "learning_rate": 1.617365062227517e-05, "loss": 0.0159, "step": 74970 }, { "grad_norm": 0.32931602001190186, "learning_rate": 1.616147602537962e-05, "loss": 0.0173, "step": 74980 }, { "grad_norm": 0.19261297583580017, "learning_rate": 1.614930512901415e-05, "loss": 0.0156, "step": 74990 }, { "grad_norm": 0.364484041929245, "learning_rate": 1.6137137934509804e-05, "loss": 0.0171, "step": 75000 }, { "grad_norm": 0.2413894087076187, "learning_rate": 1.6124974443197123e-05, "loss": 0.0157, "step": 75010 }, { "grad_norm": 0.18818119168281555, "learning_rate": 1.6112814656406306e-05, "loss": 0.0188, "step": 75020 }, { "grad_norm": 0.12482833117246628, "learning_rate": 1.6100658575467125e-05, "loss": 0.0139, "step": 75030 }, { "grad_norm": 0.20017316937446594, "learning_rate": 1.6088506201708974e-05, "loss": 0.018, "step": 75040 }, { "grad_norm": 0.23664496839046478, "learning_rate": 1.6076357536460785e-05, "loss": 0.0181, "step": 75050 }, { "grad_norm": 0.2500390112400055, "learning_rate": 1.606421258105113e-05, "loss": 0.0164, "step": 75060 }, { "grad_norm": 0.31871235370635986, "learning_rate": 1.6052071336808183e-05, "loss": 0.0202, "step": 75070 }, { "grad_norm": 0.36482933163642883, "learning_rate": 1.6039933805059658e-05, "loss": 0.0215, "step": 75080 }, { "grad_norm": 0.28378579020500183, "learning_rate": 1.6027799987132915e-05, "loss": 0.0146, "step": 75090 }, { "grad_norm": 0.3404388129711151, "learning_rate": 1.6015669884354888e-05, "loss": 0.0228, "step": 75100 }, { "grad_norm": 0.3447308838367462, "learning_rate": 1.6003543498052125e-05, "loss": 0.0181, "step": 75110 }, { "grad_norm": 0.3645228147506714, "learning_rate": 1.5991420829550715e-05, "loss": 0.0183, "step": 75120 }, { "grad_norm": 0.37132784724235535, "learning_rate": 1.5979301880176405e-05, "loss": 0.0208, "step": 75130 }, { "grad_norm": 0.21439184248447418, "learning_rate": 1.5967186651254485e-05, "loss": 0.0149, "step": 75140 }, { "grad_norm": 0.2324465662240982, "learning_rate": 1.5955075144109853e-05, "loss": 0.015, "step": 75150 }, { "grad_norm": 0.6335278749465942, "learning_rate": 1.594296736006703e-05, "loss": 0.0223, "step": 75160 }, { "grad_norm": 0.2880246639251709, "learning_rate": 1.59308633004501e-05, "loss": 0.0255, "step": 75170 }, { "grad_norm": 0.33757174015045166, "learning_rate": 1.5918762966582724e-05, "loss": 0.0145, "step": 75180 }, { "grad_norm": 0.19102899730205536, "learning_rate": 1.5906666359788203e-05, "loss": 0.0131, "step": 75190 }, { "grad_norm": 0.2104026973247528, "learning_rate": 1.589457348138938e-05, "loss": 0.0161, "step": 75200 }, { "grad_norm": 0.3813754618167877, "learning_rate": 1.588248433270872e-05, "loss": 0.0189, "step": 75210 }, { "grad_norm": 0.25385352969169617, "learning_rate": 1.587039891506828e-05, "loss": 0.0151, "step": 75220 }, { "grad_norm": 0.2582734525203705, "learning_rate": 1.5858317229789715e-05, "loss": 0.022, "step": 75230 }, { "grad_norm": 0.15866754949092865, "learning_rate": 1.584623927819423e-05, "loss": 0.0163, "step": 75240 }, { "grad_norm": 0.26139122247695923, "learning_rate": 1.583416506160269e-05, "loss": 0.0171, "step": 75250 }, { "grad_norm": 0.2299882173538208, "learning_rate": 1.5822094581335472e-05, "loss": 0.0233, "step": 75260 }, { "grad_norm": 0.17349384725093842, "learning_rate": 1.5810027838712604e-05, "loss": 0.019, "step": 75270 }, { "grad_norm": 0.2593039572238922, "learning_rate": 1.5797964835053687e-05, "loss": 0.0158, "step": 75280 }, { "grad_norm": 0.3255350589752197, "learning_rate": 1.5785905571677933e-05, "loss": 0.018, "step": 75290 }, { "grad_norm": 0.37735095620155334, "learning_rate": 1.5773850049904086e-05, "loss": 0.0172, "step": 75300 }, { "grad_norm": 0.24119096994400024, "learning_rate": 1.5761798271050543e-05, "loss": 0.0162, "step": 75310 }, { "grad_norm": 0.262135773897171, "learning_rate": 1.5749750236435277e-05, "loss": 0.0174, "step": 75320 }, { "grad_norm": 0.2801772356033325, "learning_rate": 1.5737705947375815e-05, "loss": 0.0161, "step": 75330 }, { "grad_norm": 0.20540490746498108, "learning_rate": 1.5725665405189322e-05, "loss": 0.0246, "step": 75340 }, { "grad_norm": 0.2249598652124405, "learning_rate": 1.571362861119253e-05, "loss": 0.0187, "step": 75350 }, { "grad_norm": 0.3096427023410797, "learning_rate": 1.570159556670177e-05, "loss": 0.021, "step": 75360 }, { "grad_norm": 0.17923060059547424, "learning_rate": 1.5689566273032946e-05, "loss": 0.0103, "step": 75370 }, { "grad_norm": 0.2855367064476013, "learning_rate": 1.5677540731501583e-05, "loss": 0.0162, "step": 75380 }, { "grad_norm": 0.32318806648254395, "learning_rate": 1.5665518943422746e-05, "loss": 0.0183, "step": 75390 }, { "grad_norm": 0.240348681807518, "learning_rate": 1.5653500910111134e-05, "loss": 0.0175, "step": 75400 }, { "grad_norm": 0.3490214943885803, "learning_rate": 1.5641486632881023e-05, "loss": 0.017, "step": 75410 }, { "grad_norm": 0.279553085565567, "learning_rate": 1.56294761130463e-05, "loss": 0.0167, "step": 75420 }, { "grad_norm": 0.2605268061161041, "learning_rate": 1.5617469351920373e-05, "loss": 0.0136, "step": 75430 }, { "grad_norm": 0.24110956490039825, "learning_rate": 1.5605466350816315e-05, "loss": 0.0186, "step": 75440 }, { "grad_norm": 0.2789836823940277, "learning_rate": 1.559346711104674e-05, "loss": 0.0168, "step": 75450 }, { "grad_norm": 0.32114264369010925, "learning_rate": 1.5581471633923883e-05, "loss": 0.0206, "step": 75460 }, { "grad_norm": 0.35909122228622437, "learning_rate": 1.5569479920759505e-05, "loss": 0.0199, "step": 75470 }, { "grad_norm": 0.21136268973350525, "learning_rate": 1.5557491972865072e-05, "loss": 0.0125, "step": 75480 }, { "grad_norm": 0.27226826548576355, "learning_rate": 1.554550779155152e-05, "loss": 0.0177, "step": 75490 }, { "grad_norm": 0.23193244636058807, "learning_rate": 1.5533527378129453e-05, "loss": 0.0133, "step": 75500 }, { "grad_norm": 0.293995201587677, "learning_rate": 1.552155073390899e-05, "loss": 0.0225, "step": 75510 }, { "grad_norm": 0.28969359397888184, "learning_rate": 1.5509577860199903e-05, "loss": 0.0204, "step": 75520 }, { "grad_norm": 0.23591627180576324, "learning_rate": 1.5497608758311525e-05, "loss": 0.0136, "step": 75530 }, { "grad_norm": 0.25025710463523865, "learning_rate": 1.5485643429552797e-05, "loss": 0.0165, "step": 75540 }, { "grad_norm": 0.17968407273292542, "learning_rate": 1.5473681875232197e-05, "loss": 0.0188, "step": 75550 }, { "grad_norm": 0.6145690679550171, "learning_rate": 1.5461724096657837e-05, "loss": 0.0103, "step": 75560 }, { "grad_norm": 0.42231446504592896, "learning_rate": 1.5449770095137416e-05, "loss": 0.0153, "step": 75570 }, { "grad_norm": 0.22506028413772583, "learning_rate": 1.5437819871978178e-05, "loss": 0.0223, "step": 75580 }, { "grad_norm": 0.4030134975910187, "learning_rate": 1.5425873428486992e-05, "loss": 0.0227, "step": 75590 }, { "grad_norm": 0.21379537880420685, "learning_rate": 1.5413930765970312e-05, "loss": 0.0209, "step": 75600 }, { "grad_norm": 0.2732982933521271, "learning_rate": 1.5401991885734175e-05, "loss": 0.0197, "step": 75610 }, { "grad_norm": 0.21848967671394348, "learning_rate": 1.5390056789084174e-05, "loss": 0.0179, "step": 75620 }, { "grad_norm": 0.21439582109451294, "learning_rate": 1.5378125477325545e-05, "loss": 0.0146, "step": 75630 }, { "grad_norm": 0.3203623294830322, "learning_rate": 1.5366197951763046e-05, "loss": 0.0162, "step": 75640 }, { "grad_norm": 0.3839911222457886, "learning_rate": 1.5354274213701075e-05, "loss": 0.0203, "step": 75650 }, { "grad_norm": 0.17990297079086304, "learning_rate": 1.5342354264443566e-05, "loss": 0.0158, "step": 75660 }, { "grad_norm": 0.20600725710391998, "learning_rate": 1.533043810529411e-05, "loss": 0.0192, "step": 75670 }, { "grad_norm": 0.3875741958618164, "learning_rate": 1.5318525737555807e-05, "loss": 0.0236, "step": 75680 }, { "grad_norm": 0.5148217082023621, "learning_rate": 1.53066171625314e-05, "loss": 0.0224, "step": 75690 }, { "grad_norm": 0.252686470746994, "learning_rate": 1.529471238152317e-05, "loss": 0.0217, "step": 75700 }, { "grad_norm": 0.18131355941295624, "learning_rate": 1.5282811395833023e-05, "loss": 0.0118, "step": 75710 }, { "grad_norm": 0.27479109168052673, "learning_rate": 1.5270914206762397e-05, "loss": 0.0145, "step": 75720 }, { "grad_norm": 0.2966312766075134, "learning_rate": 1.525902081561241e-05, "loss": 0.016, "step": 75730 }, { "grad_norm": 0.20976123213768005, "learning_rate": 1.5247131223683664e-05, "loss": 0.0139, "step": 75740 }, { "grad_norm": 0.20261608064174652, "learning_rate": 1.5235245432276408e-05, "loss": 0.0166, "step": 75750 }, { "grad_norm": 0.23287440836429596, "learning_rate": 1.5223363442690424e-05, "loss": 0.0119, "step": 75760 }, { "grad_norm": 0.27848735451698303, "learning_rate": 1.5211485256225133e-05, "loss": 0.0214, "step": 75770 }, { "grad_norm": 0.24208687245845795, "learning_rate": 1.519961087417951e-05, "loss": 0.0158, "step": 75780 }, { "grad_norm": 0.31492212414741516, "learning_rate": 1.5187740297852137e-05, "loss": 0.0153, "step": 75790 }, { "grad_norm": 0.28382790088653564, "learning_rate": 1.517587352854113e-05, "loss": 0.0116, "step": 75800 }, { "grad_norm": 0.2545543909072876, "learning_rate": 1.516401056754423e-05, "loss": 0.0167, "step": 75810 }, { "grad_norm": 0.3859982192516327, "learning_rate": 1.5152151416158783e-05, "loss": 0.0218, "step": 75820 }, { "grad_norm": 0.27147820591926575, "learning_rate": 1.5140296075681642e-05, "loss": 0.0183, "step": 75830 }, { "grad_norm": 0.15219151973724365, "learning_rate": 1.5128444547409331e-05, "loss": 0.0174, "step": 75840 }, { "grad_norm": 0.2932848036289215, "learning_rate": 1.5116596832637864e-05, "loss": 0.0159, "step": 75850 }, { "grad_norm": 0.4193494915962219, "learning_rate": 1.5104752932662952e-05, "loss": 0.0142, "step": 75860 }, { "grad_norm": 0.32853105664253235, "learning_rate": 1.5092912848779778e-05, "loss": 0.0148, "step": 75870 }, { "grad_norm": 0.317562460899353, "learning_rate": 1.5081076582283194e-05, "loss": 0.0136, "step": 75880 }, { "grad_norm": 0.2098056823015213, "learning_rate": 1.5069244134467553e-05, "loss": 0.0169, "step": 75890 }, { "grad_norm": 0.2686080038547516, "learning_rate": 1.5057415506626876e-05, "loss": 0.0184, "step": 75900 }, { "grad_norm": 0.36104515194892883, "learning_rate": 1.5045590700054668e-05, "loss": 0.019, "step": 75910 }, { "grad_norm": 0.1868588626384735, "learning_rate": 1.5033769716044144e-05, "loss": 0.0112, "step": 75920 }, { "grad_norm": 0.25902724266052246, "learning_rate": 1.5021952555887975e-05, "loss": 0.0155, "step": 75930 }, { "grad_norm": 0.2701323628425598, "learning_rate": 1.50101392208785e-05, "loss": 0.013, "step": 75940 }, { "grad_norm": 0.2555774450302124, "learning_rate": 1.4998329712307579e-05, "loss": 0.0157, "step": 75950 }, { "grad_norm": 0.35382646322250366, "learning_rate": 1.4986524031466708e-05, "loss": 0.0135, "step": 75960 }, { "grad_norm": 0.2912231981754303, "learning_rate": 1.4974722179646916e-05, "loss": 0.0146, "step": 75970 }, { "grad_norm": 0.3396250009536743, "learning_rate": 1.4962924158138847e-05, "loss": 0.0226, "step": 75980 }, { "grad_norm": 0.5272423624992371, "learning_rate": 1.4951129968232714e-05, "loss": 0.0114, "step": 75990 }, { "grad_norm": 0.18677549064159393, "learning_rate": 1.4939339611218329e-05, "loss": 0.015, "step": 76000 }, { "grad_norm": 0.3383999466896057, "learning_rate": 1.4927553088385032e-05, "loss": 0.0171, "step": 76010 }, { "grad_norm": 0.3225552439689636, "learning_rate": 1.4915770401021801e-05, "loss": 0.0131, "step": 76020 }, { "grad_norm": 0.20667140185832977, "learning_rate": 1.4903991550417185e-05, "loss": 0.0132, "step": 76030 }, { "grad_norm": 0.3619353771209717, "learning_rate": 1.4892216537859277e-05, "loss": 0.0186, "step": 76040 }, { "grad_norm": 0.335850328207016, "learning_rate": 1.4880445364635775e-05, "loss": 0.018, "step": 76050 }, { "grad_norm": 0.37992024421691895, "learning_rate": 1.4868678032033973e-05, "loss": 0.02, "step": 76060 }, { "grad_norm": 0.18823055922985077, "learning_rate": 1.4856914541340739e-05, "loss": 0.0197, "step": 76070 }, { "grad_norm": 0.19027893245220184, "learning_rate": 1.4845154893842472e-05, "loss": 0.0239, "step": 76080 }, { "grad_norm": 0.3844069838523865, "learning_rate": 1.4833399090825234e-05, "loss": 0.0186, "step": 76090 }, { "grad_norm": 0.3302120566368103, "learning_rate": 1.4821647133574562e-05, "loss": 0.0153, "step": 76100 }, { "grad_norm": 0.3602628707885742, "learning_rate": 1.4809899023375706e-05, "loss": 0.0164, "step": 76110 }, { "grad_norm": 0.33272379636764526, "learning_rate": 1.4798154761513366e-05, "loss": 0.0166, "step": 76120 }, { "grad_norm": 0.22212162613868713, "learning_rate": 1.4786414349271916e-05, "loss": 0.0187, "step": 76130 }, { "grad_norm": 0.2800440192222595, "learning_rate": 1.4774677787935226e-05, "loss": 0.0221, "step": 76140 }, { "grad_norm": 0.2649056911468506, "learning_rate": 1.4762945078786838e-05, "loss": 0.0165, "step": 76150 }, { "grad_norm": 0.2042037844657898, "learning_rate": 1.4751216223109777e-05, "loss": 0.0131, "step": 76160 }, { "grad_norm": 0.30726388096809387, "learning_rate": 1.4739491222186714e-05, "loss": 0.0149, "step": 76170 }, { "grad_norm": 0.20068888366222382, "learning_rate": 1.4727770077299879e-05, "loss": 0.0185, "step": 76180 }, { "grad_norm": 0.2190074920654297, "learning_rate": 1.4716052789731094e-05, "loss": 0.0146, "step": 76190 }, { "grad_norm": 0.12946385145187378, "learning_rate": 1.470433936076171e-05, "loss": 0.0129, "step": 76200 }, { "grad_norm": 0.16422535479068756, "learning_rate": 1.4692629791672719e-05, "loss": 0.0121, "step": 76210 }, { "grad_norm": 0.19761227071285248, "learning_rate": 1.468092408374464e-05, "loss": 0.0157, "step": 76220 }, { "grad_norm": 0.3493879437446594, "learning_rate": 1.4669222238257602e-05, "loss": 0.0185, "step": 76230 }, { "grad_norm": 0.20915427803993225, "learning_rate": 1.4657524256491301e-05, "loss": 0.0155, "step": 76240 }, { "grad_norm": 0.35848838090896606, "learning_rate": 1.464583013972502e-05, "loss": 0.0158, "step": 76250 }, { "grad_norm": 0.187003493309021, "learning_rate": 1.4634139889237591e-05, "loss": 0.0161, "step": 76260 }, { "grad_norm": 0.6164683699607849, "learning_rate": 1.4622453506307448e-05, "loss": 0.014, "step": 76270 }, { "grad_norm": 0.3263244926929474, "learning_rate": 1.461077099221262e-05, "loss": 0.0166, "step": 76280 }, { "grad_norm": 0.21502894163131714, "learning_rate": 1.4599092348230652e-05, "loss": 0.0166, "step": 76290 }, { "grad_norm": 0.23421929776668549, "learning_rate": 1.4587417575638724e-05, "loss": 0.0125, "step": 76300 }, { "grad_norm": 0.3407308757305145, "learning_rate": 1.4575746675713563e-05, "loss": 0.0218, "step": 76310 }, { "grad_norm": 0.3500726521015167, "learning_rate": 1.4564079649731504e-05, "loss": 0.0224, "step": 76320 }, { "grad_norm": 0.3922211527824402, "learning_rate": 1.4552416498968402e-05, "loss": 0.0141, "step": 76330 }, { "grad_norm": 0.21177037060260773, "learning_rate": 1.4540757224699747e-05, "loss": 0.015, "step": 76340 }, { "grad_norm": 0.36679768562316895, "learning_rate": 1.452910182820056e-05, "loss": 0.0154, "step": 76350 }, { "grad_norm": 0.3311832845211029, "learning_rate": 1.4517450310745468e-05, "loss": 0.0182, "step": 76360 }, { "grad_norm": 0.25263944268226624, "learning_rate": 1.450580267360866e-05, "loss": 0.015, "step": 76370 }, { "grad_norm": 0.0895075798034668, "learning_rate": 1.4494158918063922e-05, "loss": 0.0133, "step": 76380 }, { "grad_norm": 0.19463424384593964, "learning_rate": 1.4482519045384568e-05, "loss": 0.0175, "step": 76390 }, { "grad_norm": 0.21165409684181213, "learning_rate": 1.4470883056843548e-05, "loss": 0.0177, "step": 76400 }, { "grad_norm": 0.30153024196624756, "learning_rate": 1.4459250953713322e-05, "loss": 0.0195, "step": 76410 }, { "grad_norm": 0.22613529860973358, "learning_rate": 1.444762273726598e-05, "loss": 0.0174, "step": 76420 }, { "grad_norm": 0.3150424361228943, "learning_rate": 1.443599840877316e-05, "loss": 0.0174, "step": 76430 }, { "grad_norm": 0.2360956370830536, "learning_rate": 1.4424377969506103e-05, "loss": 0.0163, "step": 76440 }, { "grad_norm": 0.14432792365550995, "learning_rate": 1.4412761420735565e-05, "loss": 0.0151, "step": 76450 }, { "grad_norm": 0.2796921730041504, "learning_rate": 1.4401148763731953e-05, "loss": 0.0156, "step": 76460 }, { "grad_norm": 0.3473690152168274, "learning_rate": 1.4389539999765173e-05, "loss": 0.013, "step": 76470 }, { "grad_norm": 0.16098010540008545, "learning_rate": 1.4377935130104758e-05, "loss": 0.0172, "step": 76480 }, { "grad_norm": 0.6874793171882629, "learning_rate": 1.4366334156019801e-05, "loss": 0.0253, "step": 76490 }, { "grad_norm": 0.5281513929367065, "learning_rate": 1.435473707877898e-05, "loss": 0.0159, "step": 76500 }, { "grad_norm": 0.3623359501361847, "learning_rate": 1.4343143899650508e-05, "loss": 0.0148, "step": 76510 }, { "grad_norm": 0.2908633053302765, "learning_rate": 1.4331554619902204e-05, "loss": 0.0147, "step": 76520 }, { "grad_norm": 0.25358355045318604, "learning_rate": 1.4319969240801484e-05, "loss": 0.0182, "step": 76530 }, { "grad_norm": 0.3431106507778168, "learning_rate": 1.430838776361526e-05, "loss": 0.0157, "step": 76540 }, { "grad_norm": 0.3514471650123596, "learning_rate": 1.4296810189610093e-05, "loss": 0.0177, "step": 76550 }, { "grad_norm": 0.35361719131469727, "learning_rate": 1.4285236520052087e-05, "loss": 0.0131, "step": 76560 }, { "grad_norm": 0.19107583165168762, "learning_rate": 1.4273666756206933e-05, "loss": 0.0147, "step": 76570 }, { "grad_norm": 0.2064668834209442, "learning_rate": 1.4262100899339848e-05, "loss": 0.0167, "step": 76580 }, { "grad_norm": 0.2544557750225067, "learning_rate": 1.4250538950715696e-05, "loss": 0.0205, "step": 76590 }, { "grad_norm": 0.24461187422275543, "learning_rate": 1.4238980911598842e-05, "loss": 0.0148, "step": 76600 }, { "grad_norm": 0.34053775668144226, "learning_rate": 1.4227426783253267e-05, "loss": 0.0165, "step": 76610 }, { "grad_norm": 0.37965407967567444, "learning_rate": 1.4215876566942521e-05, "loss": 0.0212, "step": 76620 }, { "grad_norm": 0.2770576477050781, "learning_rate": 1.4204330263929721e-05, "loss": 0.0145, "step": 76630 }, { "grad_norm": 0.21013060212135315, "learning_rate": 1.4192787875477537e-05, "loss": 0.0169, "step": 76640 }, { "grad_norm": 0.1906418800354004, "learning_rate": 1.4181249402848246e-05, "loss": 0.0153, "step": 76650 }, { "grad_norm": 0.241229847073555, "learning_rate": 1.4169714847303656e-05, "loss": 0.0135, "step": 76660 }, { "grad_norm": 0.2258269041776657, "learning_rate": 1.4158184210105197e-05, "loss": 0.024, "step": 76670 }, { "grad_norm": 0.24196983873844147, "learning_rate": 1.4146657492513787e-05, "loss": 0.0189, "step": 76680 }, { "grad_norm": 0.19660218060016632, "learning_rate": 1.4135134695790047e-05, "loss": 0.0148, "step": 76690 }, { "grad_norm": 0.3096666932106018, "learning_rate": 1.4123615821194037e-05, "loss": 0.0177, "step": 76700 }, { "grad_norm": 0.14831854403018951, "learning_rate": 1.4112100869985473e-05, "loss": 0.0128, "step": 76710 }, { "grad_norm": 0.3170464038848877, "learning_rate": 1.4100589843423589e-05, "loss": 0.0191, "step": 76720 }, { "grad_norm": 0.48179706931114197, "learning_rate": 1.408908274276723e-05, "loss": 0.0224, "step": 76730 }, { "grad_norm": 0.3510138988494873, "learning_rate": 1.4077579569274786e-05, "loss": 0.0168, "step": 76740 }, { "grad_norm": 0.3562397062778473, "learning_rate": 1.4066080324204245e-05, "loss": 0.0209, "step": 76750 }, { "grad_norm": 0.2320597618818283, "learning_rate": 1.405458500881312e-05, "loss": 0.0148, "step": 76760 }, { "grad_norm": 0.19014154374599457, "learning_rate": 1.4043093624358538e-05, "loss": 0.0148, "step": 76770 }, { "grad_norm": 0.3678337335586548, "learning_rate": 1.4031606172097189e-05, "loss": 0.0209, "step": 76780 }, { "grad_norm": 0.1493891328573227, "learning_rate": 1.40201226532853e-05, "loss": 0.0154, "step": 76790 }, { "grad_norm": 0.19988538324832916, "learning_rate": 1.4008643069178701e-05, "loss": 0.0134, "step": 76800 }, { "grad_norm": 0.19971665740013123, "learning_rate": 1.399716742103278e-05, "loss": 0.0116, "step": 76810 }, { "grad_norm": 0.3083600699901581, "learning_rate": 1.398569571010252e-05, "loss": 0.015, "step": 76820 }, { "grad_norm": 0.21716873347759247, "learning_rate": 1.3974227937642414e-05, "loss": 0.0157, "step": 76830 }, { "grad_norm": 0.2000347375869751, "learning_rate": 1.3962764104906595e-05, "loss": 0.0139, "step": 76840 }, { "grad_norm": 0.24801242351531982, "learning_rate": 1.3951304213148696e-05, "loss": 0.018, "step": 76850 }, { "grad_norm": 0.2549522817134857, "learning_rate": 1.3939848263621979e-05, "loss": 0.0138, "step": 76860 }, { "grad_norm": 0.5127119421958923, "learning_rate": 1.3928396257579219e-05, "loss": 0.0166, "step": 76870 }, { "grad_norm": 0.4070410132408142, "learning_rate": 1.3916948196272833e-05, "loss": 0.017, "step": 76880 }, { "grad_norm": 0.25962236523628235, "learning_rate": 1.3905504080954734e-05, "loss": 0.0231, "step": 76890 }, { "grad_norm": 0.33330658078193665, "learning_rate": 1.3894063912876454e-05, "loss": 0.0162, "step": 76900 }, { "grad_norm": 0.22669583559036255, "learning_rate": 1.3882627693289046e-05, "loss": 0.0144, "step": 76910 }, { "grad_norm": 0.27145832777023315, "learning_rate": 1.3871195423443183e-05, "loss": 0.0188, "step": 76920 }, { "grad_norm": 0.43672820925712585, "learning_rate": 1.3859767104589044e-05, "loss": 0.0116, "step": 76930 }, { "grad_norm": 0.17736637592315674, "learning_rate": 1.3848342737976467e-05, "loss": 0.012, "step": 76940 }, { "grad_norm": 0.36322635412216187, "learning_rate": 1.3836922324854757e-05, "loss": 0.0173, "step": 76950 }, { "grad_norm": 0.18610608577728271, "learning_rate": 1.382550586647287e-05, "loss": 0.0127, "step": 76960 }, { "grad_norm": 0.17125099897384644, "learning_rate": 1.3814093364079257e-05, "loss": 0.012, "step": 76970 }, { "grad_norm": 0.3414466381072998, "learning_rate": 1.3802684818921991e-05, "loss": 0.0168, "step": 76980 }, { "grad_norm": 0.2547434866428375, "learning_rate": 1.3791280232248687e-05, "loss": 0.0181, "step": 76990 }, { "grad_norm": 0.2953726649284363, "learning_rate": 1.3779879605306555e-05, "loss": 0.0159, "step": 77000 }, { "grad_norm": 0.3806239366531372, "learning_rate": 1.376848293934232e-05, "loss": 0.0171, "step": 77010 }, { "grad_norm": 0.2523222267627716, "learning_rate": 1.3757090235602316e-05, "loss": 0.0188, "step": 77020 }, { "grad_norm": 0.44781845808029175, "learning_rate": 1.3745701495332447e-05, "loss": 0.0177, "step": 77030 }, { "grad_norm": 0.5644428133964539, "learning_rate": 1.3734316719778145e-05, "loss": 0.0161, "step": 77040 }, { "grad_norm": 0.30556637048721313, "learning_rate": 1.3722935910184454e-05, "loss": 0.0139, "step": 77050 }, { "grad_norm": 0.19291305541992188, "learning_rate": 1.3711559067795921e-05, "loss": 0.0153, "step": 77060 }, { "grad_norm": 0.19714519381523132, "learning_rate": 1.3700186193856756e-05, "loss": 0.0155, "step": 77070 }, { "grad_norm": 0.32548069953918457, "learning_rate": 1.368881728961064e-05, "loss": 0.0136, "step": 77080 }, { "grad_norm": 0.1584455668926239, "learning_rate": 1.3677452356300885e-05, "loss": 0.0202, "step": 77090 }, { "grad_norm": 0.25993612408638, "learning_rate": 1.3666091395170321e-05, "loss": 0.0184, "step": 77100 }, { "grad_norm": 0.2590329349040985, "learning_rate": 1.3654734407461384e-05, "loss": 0.0209, "step": 77110 }, { "grad_norm": 0.36543065309524536, "learning_rate": 1.3643381394416022e-05, "loss": 0.0191, "step": 77120 }, { "grad_norm": 0.22074227035045624, "learning_rate": 1.3632032357275838e-05, "loss": 0.0203, "step": 77130 }, { "grad_norm": 0.3148939311504364, "learning_rate": 1.3620687297281898e-05, "loss": 0.0172, "step": 77140 }, { "grad_norm": 0.38356855511665344, "learning_rate": 1.3609346215674917e-05, "loss": 0.0125, "step": 77150 }, { "grad_norm": 0.32110753655433655, "learning_rate": 1.3598009113695109e-05, "loss": 0.0152, "step": 77160 }, { "grad_norm": 0.23228155076503754, "learning_rate": 1.358667599258231e-05, "loss": 0.0182, "step": 77170 }, { "grad_norm": 0.15232405066490173, "learning_rate": 1.357534685357586e-05, "loss": 0.0139, "step": 77180 }, { "grad_norm": 0.2712881863117218, "learning_rate": 1.3564021697914715e-05, "loss": 0.016, "step": 77190 }, { "grad_norm": 0.15822620689868927, "learning_rate": 1.3552700526837381e-05, "loss": 0.0141, "step": 77200 }, { "grad_norm": 0.25437068939208984, "learning_rate": 1.3541383341581925e-05, "loss": 0.0145, "step": 77210 }, { "grad_norm": 0.3842693269252777, "learning_rate": 1.3530070143385965e-05, "loss": 0.0152, "step": 77220 }, { "grad_norm": 0.14254936575889587, "learning_rate": 1.3518760933486702e-05, "loss": 0.0125, "step": 77230 }, { "grad_norm": 0.2770236134529114, "learning_rate": 1.3507455713120908e-05, "loss": 0.0195, "step": 77240 }, { "grad_norm": 0.1251964569091797, "learning_rate": 1.3496154483524876e-05, "loss": 0.0185, "step": 77250 }, { "grad_norm": 0.20826275646686554, "learning_rate": 1.3484857245934506e-05, "loss": 0.0136, "step": 77260 }, { "grad_norm": 0.32949450612068176, "learning_rate": 1.3473564001585248e-05, "loss": 0.014, "step": 77270 }, { "grad_norm": 0.2275938093662262, "learning_rate": 1.3462274751712129e-05, "loss": 0.0157, "step": 77280 }, { "grad_norm": 0.20868849754333496, "learning_rate": 1.3450989497549698e-05, "loss": 0.0147, "step": 77290 }, { "grad_norm": 0.4720548987388611, "learning_rate": 1.3439708240332115e-05, "loss": 0.0215, "step": 77300 }, { "grad_norm": 0.31868040561676025, "learning_rate": 1.3428430981293043e-05, "loss": 0.0189, "step": 77310 }, { "grad_norm": 0.2420894205570221, "learning_rate": 1.3417157721665808e-05, "loss": 0.0196, "step": 77320 }, { "grad_norm": 0.21566130220890045, "learning_rate": 1.3405888462683186e-05, "loss": 0.013, "step": 77330 }, { "grad_norm": 0.19824425876140594, "learning_rate": 1.3394623205577595e-05, "loss": 0.0153, "step": 77340 }, { "grad_norm": 0.503115177154541, "learning_rate": 1.3383361951580963e-05, "loss": 0.0143, "step": 77350 }, { "grad_norm": 0.19719800353050232, "learning_rate": 1.337210470192483e-05, "loss": 0.014, "step": 77360 }, { "grad_norm": 0.44014400243759155, "learning_rate": 1.3360851457840245e-05, "loss": 0.0156, "step": 77370 }, { "grad_norm": 0.24875840544700623, "learning_rate": 1.3349602220557856e-05, "loss": 0.0131, "step": 77380 }, { "grad_norm": 0.3449510931968689, "learning_rate": 1.3338356991307865e-05, "loss": 0.0155, "step": 77390 }, { "grad_norm": 0.32913511991500854, "learning_rate": 1.3327115771320048e-05, "loss": 0.0209, "step": 77400 }, { "grad_norm": 0.20889489352703094, "learning_rate": 1.3315878561823697e-05, "loss": 0.0153, "step": 77410 }, { "grad_norm": 0.3089931607246399, "learning_rate": 1.330464536404773e-05, "loss": 0.0129, "step": 77420 }, { "grad_norm": 0.3410319983959198, "learning_rate": 1.3293416179220563e-05, "loss": 0.0186, "step": 77430 }, { "grad_norm": 0.36105477809906006, "learning_rate": 1.328219100857021e-05, "loss": 0.0138, "step": 77440 }, { "grad_norm": 0.4738839566707611, "learning_rate": 1.327096985332425e-05, "loss": 0.0199, "step": 77450 }, { "grad_norm": 0.2629057765007019, "learning_rate": 1.3259752714709816e-05, "loss": 0.014, "step": 77460 }, { "grad_norm": 0.39637237787246704, "learning_rate": 1.3248539593953569e-05, "loss": 0.015, "step": 77470 }, { "grad_norm": 0.17575299739837646, "learning_rate": 1.323733049228178e-05, "loss": 0.0176, "step": 77480 }, { "grad_norm": 0.3278827369213104, "learning_rate": 1.3226125410920276e-05, "loss": 0.0184, "step": 77490 }, { "grad_norm": 0.21216657757759094, "learning_rate": 1.3214924351094388e-05, "loss": 0.014, "step": 77500 }, { "grad_norm": 0.21463491022586823, "learning_rate": 1.3203727314029069e-05, "loss": 0.0135, "step": 77510 }, { "grad_norm": 0.1667236089706421, "learning_rate": 1.3192534300948811e-05, "loss": 0.0124, "step": 77520 }, { "grad_norm": 0.2472933530807495, "learning_rate": 1.3181345313077675e-05, "loss": 0.0137, "step": 77530 }, { "grad_norm": 0.28925034403800964, "learning_rate": 1.3170160351639243e-05, "loss": 0.0165, "step": 77540 }, { "grad_norm": 0.275640070438385, "learning_rate": 1.3158979417856721e-05, "loss": 0.0224, "step": 77550 }, { "grad_norm": 0.2114468663930893, "learning_rate": 1.31478025129528e-05, "loss": 0.0161, "step": 77560 }, { "grad_norm": 0.6599632501602173, "learning_rate": 1.3136629638149794e-05, "loss": 0.0137, "step": 77570 }, { "grad_norm": 0.2237047553062439, "learning_rate": 1.3125460794669547e-05, "loss": 0.0197, "step": 77580 }, { "grad_norm": 0.23122723400592804, "learning_rate": 1.3114295983733483e-05, "loss": 0.0162, "step": 77590 }, { "grad_norm": 0.3484392762184143, "learning_rate": 1.3103135206562534e-05, "loss": 0.014, "step": 77600 }, { "grad_norm": 0.2779693305492401, "learning_rate": 1.3091978464377264e-05, "loss": 0.0226, "step": 77610 }, { "grad_norm": 0.2476910799741745, "learning_rate": 1.3080825758397724e-05, "loss": 0.016, "step": 77620 }, { "grad_norm": 0.5412909984588623, "learning_rate": 1.3069677089843568e-05, "loss": 0.0144, "step": 77630 }, { "grad_norm": 0.28585284948349, "learning_rate": 1.305853245993401e-05, "loss": 0.0251, "step": 77640 }, { "grad_norm": 0.23382505774497986, "learning_rate": 1.3047391869887809e-05, "loss": 0.0122, "step": 77650 }, { "grad_norm": 0.2895863652229309, "learning_rate": 1.3036255320923263e-05, "loss": 0.0168, "step": 77660 }, { "grad_norm": 0.2889266908168793, "learning_rate": 1.3025122814258273e-05, "loss": 0.0143, "step": 77670 }, { "grad_norm": 0.5890244841575623, "learning_rate": 1.3013994351110249e-05, "loss": 0.0184, "step": 77680 }, { "grad_norm": 0.12389693409204483, "learning_rate": 1.3002869932696194e-05, "loss": 0.0147, "step": 77690 }, { "grad_norm": 0.2884441018104553, "learning_rate": 1.2991749560232664e-05, "loss": 0.0167, "step": 77700 }, { "grad_norm": 0.25573018193244934, "learning_rate": 1.2980633234935768e-05, "loss": 0.0191, "step": 77710 }, { "grad_norm": 0.2845834195613861, "learning_rate": 1.2969520958021153e-05, "loss": 0.0167, "step": 77720 }, { "grad_norm": 0.27950432896614075, "learning_rate": 1.295841273070405e-05, "loss": 0.0129, "step": 77730 }, { "grad_norm": 0.32045766711235046, "learning_rate": 1.2947308554199255e-05, "loss": 0.0136, "step": 77740 }, { "grad_norm": 0.38986197113990784, "learning_rate": 1.2936208429721075e-05, "loss": 0.0183, "step": 77750 }, { "grad_norm": 0.20878145098686218, "learning_rate": 1.2925112358483416e-05, "loss": 0.0214, "step": 77760 }, { "grad_norm": 0.17958101630210876, "learning_rate": 1.2914020341699728e-05, "loss": 0.0165, "step": 77770 }, { "grad_norm": 0.3902932405471802, "learning_rate": 1.2902932380583032e-05, "loss": 0.0163, "step": 77780 }, { "grad_norm": 0.16756191849708557, "learning_rate": 1.2891848476345863e-05, "loss": 0.0161, "step": 77790 }, { "grad_norm": 0.2811896502971649, "learning_rate": 1.2880768630200368e-05, "loss": 0.0167, "step": 77800 }, { "grad_norm": 0.287760466337204, "learning_rate": 1.286969284335819e-05, "loss": 0.0174, "step": 77810 }, { "grad_norm": 0.4835546314716339, "learning_rate": 1.2858621117030578e-05, "loss": 0.0184, "step": 77820 }, { "grad_norm": 0.27955955266952515, "learning_rate": 1.2847553452428324e-05, "loss": 0.0172, "step": 77830 }, { "grad_norm": 0.3999861776828766, "learning_rate": 1.283648985076178e-05, "loss": 0.0115, "step": 77840 }, { "grad_norm": 0.2441781908273697, "learning_rate": 1.2825430313240817e-05, "loss": 0.0142, "step": 77850 }, { "grad_norm": 0.47286006808280945, "learning_rate": 1.281437484107491e-05, "loss": 0.0199, "step": 77860 }, { "grad_norm": 0.19856898486614227, "learning_rate": 1.2803323435473058e-05, "loss": 0.0194, "step": 77870 }, { "grad_norm": 0.21873170137405396, "learning_rate": 1.2792276097643841e-05, "loss": 0.0165, "step": 77880 }, { "grad_norm": 0.3598770797252655, "learning_rate": 1.2781232828795342e-05, "loss": 0.0143, "step": 77890 }, { "grad_norm": 0.2727137804031372, "learning_rate": 1.2770193630135292e-05, "loss": 0.0232, "step": 77900 }, { "grad_norm": 0.2074972540140152, "learning_rate": 1.2759158502870877e-05, "loss": 0.0261, "step": 77910 }, { "grad_norm": 0.5150943994522095, "learning_rate": 1.2748127448208913e-05, "loss": 0.0233, "step": 77920 }, { "grad_norm": 0.3591483235359192, "learning_rate": 1.2737100467355706e-05, "loss": 0.0198, "step": 77930 }, { "grad_norm": 0.27155816555023193, "learning_rate": 1.2726077561517169e-05, "loss": 0.0301, "step": 77940 }, { "grad_norm": 0.21785520017147064, "learning_rate": 1.2715058731898755e-05, "loss": 0.0116, "step": 77950 }, { "grad_norm": 0.2865329384803772, "learning_rate": 1.2704043979705471e-05, "loss": 0.0131, "step": 77960 }, { "grad_norm": 0.37684333324432373, "learning_rate": 1.269303330614185e-05, "loss": 0.0172, "step": 77970 }, { "grad_norm": 0.2087186872959137, "learning_rate": 1.2682026712412015e-05, "loss": 0.017, "step": 77980 }, { "grad_norm": 0.22531640529632568, "learning_rate": 1.2671024199719644e-05, "loss": 0.0138, "step": 77990 }, { "grad_norm": 0.27009427547454834, "learning_rate": 1.2660025769267936e-05, "loss": 0.0175, "step": 78000 }, { "grad_norm": 0.5672930479049683, "learning_rate": 1.2649031422259667e-05, "loss": 0.0172, "step": 78010 }, { "grad_norm": 0.22648125886917114, "learning_rate": 1.2638041159897163e-05, "loss": 0.0152, "step": 78020 }, { "grad_norm": 0.3305927515029907, "learning_rate": 1.262705498338232e-05, "loss": 0.0142, "step": 78030 }, { "grad_norm": 0.19365178048610687, "learning_rate": 1.2616072893916542e-05, "loss": 0.0242, "step": 78040 }, { "grad_norm": 0.22971153259277344, "learning_rate": 1.2605094892700836e-05, "loss": 0.0154, "step": 78050 }, { "grad_norm": 0.30256229639053345, "learning_rate": 1.259412098093572e-05, "loss": 0.0142, "step": 78060 }, { "grad_norm": 0.41710638999938965, "learning_rate": 1.258315115982131e-05, "loss": 0.0147, "step": 78070 }, { "grad_norm": 0.21561473608016968, "learning_rate": 1.2572185430557203e-05, "loss": 0.0118, "step": 78080 }, { "grad_norm": 0.2617904245853424, "learning_rate": 1.2561223794342659e-05, "loss": 0.014, "step": 78090 }, { "grad_norm": 0.3239823579788208, "learning_rate": 1.255026625237637e-05, "loss": 0.0168, "step": 78100 }, { "grad_norm": 0.21278761327266693, "learning_rate": 1.2539312805856678e-05, "loss": 0.0193, "step": 78110 }, { "grad_norm": 0.16380412876605988, "learning_rate": 1.25283634559814e-05, "loss": 0.0131, "step": 78120 }, { "grad_norm": 0.20979559421539307, "learning_rate": 1.2517418203947972e-05, "loss": 0.0129, "step": 78130 }, { "grad_norm": 0.4669211506843567, "learning_rate": 1.2506477050953308e-05, "loss": 0.0137, "step": 78140 }, { "grad_norm": 0.31117814779281616, "learning_rate": 1.2495539998193972e-05, "loss": 0.0146, "step": 78150 }, { "grad_norm": 0.3816847503185272, "learning_rate": 1.2484607046865975e-05, "loss": 0.0181, "step": 78160 }, { "grad_norm": 0.2154785692691803, "learning_rate": 1.2473678198164968e-05, "loss": 0.0179, "step": 78170 }, { "grad_norm": 0.40694761276245117, "learning_rate": 1.2462753453286075e-05, "loss": 0.0188, "step": 78180 }, { "grad_norm": 0.23081150650978088, "learning_rate": 1.2451832813424031e-05, "loss": 0.0315, "step": 78190 }, { "grad_norm": 0.2731872498989105, "learning_rate": 1.2440916279773095e-05, "loss": 0.0183, "step": 78200 }, { "grad_norm": 0.18238738179206848, "learning_rate": 1.2430003853527101e-05, "loss": 0.0143, "step": 78210 }, { "grad_norm": 0.6872625350952148, "learning_rate": 1.2419095535879383e-05, "loss": 0.0243, "step": 78220 }, { "grad_norm": 0.3960539698600769, "learning_rate": 1.2408191328022878e-05, "loss": 0.0188, "step": 78230 }, { "grad_norm": 0.4513959586620331, "learning_rate": 1.2397291231150066e-05, "loss": 0.0155, "step": 78240 }, { "grad_norm": 0.33218541741371155, "learning_rate": 1.2386395246452937e-05, "loss": 0.0211, "step": 78250 }, { "grad_norm": 0.26611900329589844, "learning_rate": 1.2375503375123083e-05, "loss": 0.0154, "step": 78260 }, { "grad_norm": 0.26239827275276184, "learning_rate": 1.2364615618351583e-05, "loss": 0.0206, "step": 78270 }, { "grad_norm": 0.2684067189693451, "learning_rate": 1.2353731977329169e-05, "loss": 0.0128, "step": 78280 }, { "grad_norm": 0.7074528932571411, "learning_rate": 1.2342852453246007e-05, "loss": 0.0162, "step": 78290 }, { "grad_norm": 0.29703909158706665, "learning_rate": 1.2331977047291898e-05, "loss": 0.0192, "step": 78300 }, { "grad_norm": 0.32078641653060913, "learning_rate": 1.2321105760656137e-05, "loss": 0.0145, "step": 78310 }, { "grad_norm": 0.3128480017185211, "learning_rate": 1.2310238594527613e-05, "loss": 0.0174, "step": 78320 }, { "grad_norm": 0.2585955262184143, "learning_rate": 1.2299375550094693e-05, "loss": 0.0147, "step": 78330 }, { "grad_norm": 0.19587133824825287, "learning_rate": 1.2288516628545421e-05, "loss": 0.0119, "step": 78340 }, { "grad_norm": 0.17926059663295746, "learning_rate": 1.2277661831067255e-05, "loss": 0.0179, "step": 78350 }, { "grad_norm": 0.22974447906017303, "learning_rate": 1.2266811158847286e-05, "loss": 0.0159, "step": 78360 }, { "grad_norm": 0.3317130208015442, "learning_rate": 1.2255964613072107e-05, "loss": 0.0133, "step": 78370 }, { "grad_norm": 0.1695360690355301, "learning_rate": 1.2245122194927905e-05, "loss": 0.0196, "step": 78380 }, { "grad_norm": 0.2125944197177887, "learning_rate": 1.2234283905600364e-05, "loss": 0.0159, "step": 78390 }, { "grad_norm": 0.265095055103302, "learning_rate": 1.222344974627475e-05, "loss": 0.0176, "step": 78400 }, { "grad_norm": 0.2631811201572418, "learning_rate": 1.2212619718135875e-05, "loss": 0.0148, "step": 78410 }, { "grad_norm": 0.18048600852489471, "learning_rate": 1.2201793822368118e-05, "loss": 0.016, "step": 78420 }, { "grad_norm": 0.18303480744361877, "learning_rate": 1.2190972060155337e-05, "loss": 0.0112, "step": 78430 }, { "grad_norm": 0.34182241559028625, "learning_rate": 1.2180154432681013e-05, "loss": 0.016, "step": 78440 }, { "grad_norm": 0.3798076808452606, "learning_rate": 1.2169340941128148e-05, "loss": 0.0156, "step": 78450 }, { "grad_norm": 0.19400560855865479, "learning_rate": 1.2158531586679267e-05, "loss": 0.0134, "step": 78460 }, { "grad_norm": 0.18785087764263153, "learning_rate": 1.2147726370516477e-05, "loss": 0.0147, "step": 78470 }, { "grad_norm": 0.20193946361541748, "learning_rate": 1.2136925293821422e-05, "loss": 0.017, "step": 78480 }, { "grad_norm": 0.4071231186389923, "learning_rate": 1.2126128357775306e-05, "loss": 0.0158, "step": 78490 }, { "grad_norm": 0.36126595735549927, "learning_rate": 1.2115335563558838e-05, "loss": 0.0132, "step": 78500 }, { "grad_norm": 0.11024591326713562, "learning_rate": 1.2104546912352327e-05, "loss": 0.0145, "step": 78510 }, { "grad_norm": 0.2237308770418167, "learning_rate": 1.2093762405335557e-05, "loss": 0.0114, "step": 78520 }, { "grad_norm": 0.3781435191631317, "learning_rate": 1.2082982043687974e-05, "loss": 0.0145, "step": 78530 }, { "grad_norm": 0.21190449595451355, "learning_rate": 1.207220582858845e-05, "loss": 0.0118, "step": 78540 }, { "grad_norm": 0.25935986638069153, "learning_rate": 1.206143376121549e-05, "loss": 0.0143, "step": 78550 }, { "grad_norm": 0.1605454981327057, "learning_rate": 1.2050665842747078e-05, "loss": 0.0125, "step": 78560 }, { "grad_norm": 0.37467560172080994, "learning_rate": 1.2039902074360804e-05, "loss": 0.0113, "step": 78570 }, { "grad_norm": 0.15186499059200287, "learning_rate": 1.2029142457233755e-05, "loss": 0.0122, "step": 78580 }, { "grad_norm": 0.1767839938402176, "learning_rate": 1.2018386992542601e-05, "loss": 0.0138, "step": 78590 }, { "grad_norm": 0.1943179965019226, "learning_rate": 1.2007635681463542e-05, "loss": 0.014, "step": 78600 }, { "grad_norm": 0.16859716176986694, "learning_rate": 1.1996888525172334e-05, "loss": 0.0229, "step": 78610 }, { "grad_norm": 0.2734510004520416, "learning_rate": 1.198614552484425e-05, "loss": 0.0167, "step": 78620 }, { "grad_norm": 0.2432447373867035, "learning_rate": 1.197540668165415e-05, "loss": 0.0153, "step": 78630 }, { "grad_norm": 0.25180450081825256, "learning_rate": 1.1964671996776395e-05, "loss": 0.0218, "step": 78640 }, { "grad_norm": 0.20517946779727936, "learning_rate": 1.1953941471384922e-05, "loss": 0.0208, "step": 78650 }, { "grad_norm": 0.205011248588562, "learning_rate": 1.1943215106653205e-05, "loss": 0.0183, "step": 78660 }, { "grad_norm": 0.17514142394065857, "learning_rate": 1.193249290375429e-05, "loss": 0.0156, "step": 78670 }, { "grad_norm": 0.2956189513206482, "learning_rate": 1.1921774863860696e-05, "loss": 0.0147, "step": 78680 }, { "grad_norm": 0.24997290968894958, "learning_rate": 1.1911060988144556e-05, "loss": 0.0129, "step": 78690 }, { "grad_norm": 0.46728813648223877, "learning_rate": 1.1900351277777533e-05, "loss": 0.0184, "step": 78700 }, { "grad_norm": 0.49721798300743103, "learning_rate": 1.1889645733930798e-05, "loss": 0.0202, "step": 78710 }, { "grad_norm": 0.22219210863113403, "learning_rate": 1.1878944357775107e-05, "loss": 0.019, "step": 78720 }, { "grad_norm": 0.18221968412399292, "learning_rate": 1.1868247150480749e-05, "loss": 0.012, "step": 78730 }, { "grad_norm": 0.26719406247138977, "learning_rate": 1.1857554113217567e-05, "loss": 0.0155, "step": 78740 }, { "grad_norm": 0.2998570501804352, "learning_rate": 1.1846865247154903e-05, "loss": 0.0142, "step": 78750 }, { "grad_norm": 0.1658896952867508, "learning_rate": 1.1836180553461711e-05, "loss": 0.017, "step": 78760 }, { "grad_norm": 0.27969059348106384, "learning_rate": 1.1825500033306419e-05, "loss": 0.0284, "step": 78770 }, { "grad_norm": 0.20655688643455505, "learning_rate": 1.1814823687857052e-05, "loss": 0.0144, "step": 78780 }, { "grad_norm": 0.3204764723777771, "learning_rate": 1.1804151518281153e-05, "loss": 0.0171, "step": 78790 }, { "grad_norm": 0.4768572449684143, "learning_rate": 1.179348352574583e-05, "loss": 0.028, "step": 78800 }, { "grad_norm": 0.1014423742890358, "learning_rate": 1.1782819711417697e-05, "loss": 0.018, "step": 78810 }, { "grad_norm": 0.22621197998523712, "learning_rate": 1.1772160076462951e-05, "loss": 0.0145, "step": 78820 }, { "grad_norm": 0.2425076812505722, "learning_rate": 1.1761504622047293e-05, "loss": 0.0157, "step": 78830 }, { "grad_norm": 0.257375568151474, "learning_rate": 1.1750853349335999e-05, "loss": 0.0106, "step": 78840 }, { "grad_norm": 0.20991650223731995, "learning_rate": 1.1740206259493874e-05, "loss": 0.0201, "step": 78850 }, { "grad_norm": 0.3179958760738373, "learning_rate": 1.1729563353685286e-05, "loss": 0.015, "step": 78860 }, { "grad_norm": 0.18466590344905853, "learning_rate": 1.1718924633074091e-05, "loss": 0.0175, "step": 78870 }, { "grad_norm": 0.17547215521335602, "learning_rate": 1.1708290098823765e-05, "loss": 0.0192, "step": 78880 }, { "grad_norm": 0.22730354964733124, "learning_rate": 1.1697659752097245e-05, "loss": 0.017, "step": 78890 }, { "grad_norm": 0.3592078387737274, "learning_rate": 1.1687033594057068e-05, "loss": 0.0355, "step": 78900 }, { "grad_norm": 0.2304716557264328, "learning_rate": 1.1676411625865297e-05, "loss": 0.0113, "step": 78910 }, { "grad_norm": 0.24010980129241943, "learning_rate": 1.1665793848683542e-05, "loss": 0.0156, "step": 78920 }, { "grad_norm": 0.5264191031455994, "learning_rate": 1.1655180263672927e-05, "loss": 0.0235, "step": 78930 }, { "grad_norm": 0.15581925213336945, "learning_rate": 1.1644570871994142e-05, "loss": 0.0127, "step": 78940 }, { "grad_norm": 0.1846766173839569, "learning_rate": 1.163396567480744e-05, "loss": 0.0148, "step": 78950 }, { "grad_norm": 0.3460353910923004, "learning_rate": 1.1623364673272552e-05, "loss": 0.0162, "step": 78960 }, { "grad_norm": 0.24367713928222656, "learning_rate": 1.1612767868548807e-05, "loss": 0.0203, "step": 78970 }, { "grad_norm": 0.3180862367153168, "learning_rate": 1.1602175261795057e-05, "loss": 0.0216, "step": 78980 }, { "grad_norm": 0.29895639419555664, "learning_rate": 1.1591586854169696e-05, "loss": 0.0159, "step": 78990 }, { "grad_norm": 0.23783732950687408, "learning_rate": 1.158100264683064e-05, "loss": 0.0171, "step": 79000 }, { "grad_norm": 0.2774977385997772, "learning_rate": 1.1570422640935386e-05, "loss": 0.0204, "step": 79010 }, { "grad_norm": 0.1573941707611084, "learning_rate": 1.155984683764092e-05, "loss": 0.0237, "step": 79020 }, { "grad_norm": 0.37201815843582153, "learning_rate": 1.1549275238103812e-05, "loss": 0.0145, "step": 79030 }, { "grad_norm": 0.45508116483688354, "learning_rate": 1.1538707843480152e-05, "loss": 0.0173, "step": 79040 }, { "grad_norm": 0.2558605968952179, "learning_rate": 1.1528144654925593e-05, "loss": 0.0197, "step": 79050 }, { "grad_norm": 0.23508542776107788, "learning_rate": 1.1517585673595277e-05, "loss": 0.0179, "step": 79060 }, { "grad_norm": 0.20616993308067322, "learning_rate": 1.150703090064395e-05, "loss": 0.017, "step": 79070 }, { "grad_norm": 0.25924333930015564, "learning_rate": 1.1496480337225835e-05, "loss": 0.0178, "step": 79080 }, { "grad_norm": 0.3372266888618469, "learning_rate": 1.1485933984494751e-05, "loss": 0.0161, "step": 79090 }, { "grad_norm": 0.2535143792629242, "learning_rate": 1.1475391843604e-05, "loss": 0.0183, "step": 79100 }, { "grad_norm": 0.3749805986881256, "learning_rate": 1.1464853915706503e-05, "loss": 0.0207, "step": 79110 }, { "grad_norm": 0.2756129801273346, "learning_rate": 1.1454320201954626e-05, "loss": 0.0155, "step": 79120 }, { "grad_norm": 0.15669794380664825, "learning_rate": 1.1443790703500356e-05, "loss": 0.0162, "step": 79130 }, { "grad_norm": 0.2580418288707733, "learning_rate": 1.1433265421495154e-05, "loss": 0.0163, "step": 79140 }, { "grad_norm": 0.22093819081783295, "learning_rate": 1.1422744357090065e-05, "loss": 0.0259, "step": 79150 }, { "grad_norm": 0.2841292917728424, "learning_rate": 1.1412227511435647e-05, "loss": 0.0172, "step": 79160 }, { "grad_norm": 0.27922388911247253, "learning_rate": 1.1401714885682025e-05, "loss": 0.0153, "step": 79170 }, { "grad_norm": 0.2338474988937378, "learning_rate": 1.1391206480978823e-05, "loss": 0.0136, "step": 79180 }, { "grad_norm": 0.29367703199386597, "learning_rate": 1.1380702298475227e-05, "loss": 0.0165, "step": 79190 }, { "grad_norm": 0.2941833436489105, "learning_rate": 1.1370202339319985e-05, "loss": 0.0175, "step": 79200 }, { "grad_norm": 0.23569828271865845, "learning_rate": 1.1359706604661319e-05, "loss": 0.0163, "step": 79210 }, { "grad_norm": 0.2976457476615906, "learning_rate": 1.1349215095647042e-05, "loss": 0.0154, "step": 79220 }, { "grad_norm": 0.23280733823776245, "learning_rate": 1.1338727813424493e-05, "loss": 0.0137, "step": 79230 }, { "grad_norm": 0.20229753851890564, "learning_rate": 1.132824475914056e-05, "loss": 0.0157, "step": 79240 }, { "grad_norm": 0.37949442863464355, "learning_rate": 1.131776593394162e-05, "loss": 0.0157, "step": 79250 }, { "grad_norm": 0.38829705119132996, "learning_rate": 1.1307291338973657e-05, "loss": 0.0174, "step": 79260 }, { "grad_norm": 0.27350130677223206, "learning_rate": 1.1296820975382121e-05, "loss": 0.0144, "step": 79270 }, { "grad_norm": 0.43230384588241577, "learning_rate": 1.1286354844312074e-05, "loss": 0.0252, "step": 79280 }, { "grad_norm": 0.19554202258586884, "learning_rate": 1.1275892946908023e-05, "loss": 0.0232, "step": 79290 }, { "grad_norm": 0.3379306495189667, "learning_rate": 1.126543528431413e-05, "loss": 0.0236, "step": 79300 }, { "grad_norm": 0.21492810547351837, "learning_rate": 1.125498185767398e-05, "loss": 0.014, "step": 79310 }, { "grad_norm": 0.29721957445144653, "learning_rate": 1.1244532668130781e-05, "loss": 0.0171, "step": 79320 }, { "grad_norm": 0.26992881298065186, "learning_rate": 1.1234087716827203e-05, "loss": 0.02, "step": 79330 }, { "grad_norm": 0.2028600573539734, "learning_rate": 1.1223647004905524e-05, "loss": 0.0189, "step": 79340 }, { "grad_norm": 0.8212866187095642, "learning_rate": 1.1213210533507485e-05, "loss": 0.0171, "step": 79350 }, { "grad_norm": 0.26795804500579834, "learning_rate": 1.1202778303774447e-05, "loss": 0.0135, "step": 79360 }, { "grad_norm": 0.16474555432796478, "learning_rate": 1.1192350316847234e-05, "loss": 0.0169, "step": 79370 }, { "grad_norm": 0.2191684991121292, "learning_rate": 1.1181926573866258e-05, "loss": 0.0175, "step": 79380 }, { "grad_norm": 0.20200292766094208, "learning_rate": 1.1171507075971416e-05, "loss": 0.0189, "step": 79390 }, { "grad_norm": 0.18406778573989868, "learning_rate": 1.1161091824302177e-05, "loss": 0.0223, "step": 79400 }, { "grad_norm": 0.3340529203414917, "learning_rate": 1.115068081999755e-05, "loss": 0.0187, "step": 79410 }, { "grad_norm": 0.2941057085990906, "learning_rate": 1.1140274064196071e-05, "loss": 0.0202, "step": 79420 }, { "grad_norm": 0.26465904712677, "learning_rate": 1.112987155803578e-05, "loss": 0.0141, "step": 79430 }, { "grad_norm": 0.38617411255836487, "learning_rate": 1.1119473302654298e-05, "loss": 0.0192, "step": 79440 }, { "grad_norm": 0.193215474486351, "learning_rate": 1.1109079299188769e-05, "loss": 0.0155, "step": 79450 }, { "grad_norm": 0.233674556016922, "learning_rate": 1.1098689548775847e-05, "loss": 0.0176, "step": 79460 }, { "grad_norm": 0.26884758472442627, "learning_rate": 1.1088304052551762e-05, "loss": 0.0172, "step": 79470 }, { "grad_norm": 0.2863708436489105, "learning_rate": 1.1077922811652215e-05, "loss": 0.0165, "step": 79480 }, { "grad_norm": 0.20682388544082642, "learning_rate": 1.1067545827212534e-05, "loss": 0.0204, "step": 79490 }, { "grad_norm": 0.2435240000486374, "learning_rate": 1.1057173100367496e-05, "loss": 0.0131, "step": 79500 }, { "grad_norm": 0.19274009764194489, "learning_rate": 1.1046804632251474e-05, "loss": 0.0138, "step": 79510 }, { "grad_norm": 0.11578690260648727, "learning_rate": 1.1036440423998318e-05, "loss": 0.0132, "step": 79520 }, { "grad_norm": 0.29798758029937744, "learning_rate": 1.1026080476741469e-05, "loss": 0.0243, "step": 79530 }, { "grad_norm": 0.18668974936008453, "learning_rate": 1.1015724791613836e-05, "loss": 0.0156, "step": 79540 }, { "grad_norm": 0.20782773196697235, "learning_rate": 1.1005373369747951e-05, "loss": 0.0142, "step": 79550 }, { "grad_norm": 0.25365152955055237, "learning_rate": 1.09950262122758e-05, "loss": 0.0157, "step": 79560 }, { "grad_norm": 0.8214876651763916, "learning_rate": 1.0984683320328948e-05, "loss": 0.0203, "step": 79570 }, { "grad_norm": 0.47640612721443176, "learning_rate": 1.0974344695038458e-05, "loss": 0.0139, "step": 79580 }, { "grad_norm": 0.5200612545013428, "learning_rate": 1.0964010337534968e-05, "loss": 0.0164, "step": 79590 }, { "grad_norm": 0.2959233522415161, "learning_rate": 1.0953680248948611e-05, "loss": 0.0149, "step": 79600 }, { "grad_norm": 0.22448761761188507, "learning_rate": 1.0943354430409069e-05, "loss": 0.0125, "step": 79610 }, { "grad_norm": 0.1842467337846756, "learning_rate": 1.0933032883045574e-05, "loss": 0.0126, "step": 79620 }, { "grad_norm": 0.3866032660007477, "learning_rate": 1.0922715607986872e-05, "loss": 0.0189, "step": 79630 }, { "grad_norm": 0.2707255780696869, "learning_rate": 1.0912402606361234e-05, "loss": 0.019, "step": 79640 }, { "grad_norm": 0.42759114503860474, "learning_rate": 1.090209387929647e-05, "loss": 0.0145, "step": 79650 }, { "grad_norm": 0.22385939955711365, "learning_rate": 1.0891789427919957e-05, "loss": 0.0165, "step": 79660 }, { "grad_norm": 0.5297273993492126, "learning_rate": 1.088148925335854e-05, "loss": 0.0163, "step": 79670 }, { "grad_norm": 0.251358300447464, "learning_rate": 1.087119335673864e-05, "loss": 0.013, "step": 79680 }, { "grad_norm": 0.21751263737678528, "learning_rate": 1.0860901739186208e-05, "loss": 0.0134, "step": 79690 }, { "grad_norm": 0.20117314159870148, "learning_rate": 1.085061440182673e-05, "loss": 0.0156, "step": 79700 }, { "grad_norm": 0.16861459612846375, "learning_rate": 1.0840331345785193e-05, "loss": 0.0142, "step": 79710 }, { "grad_norm": 0.30024510622024536, "learning_rate": 1.083005257218615e-05, "loss": 0.0149, "step": 79720 }, { "grad_norm": 0.2752189040184021, "learning_rate": 1.0819778082153642e-05, "loss": 0.0155, "step": 79730 }, { "grad_norm": 0.16769134998321533, "learning_rate": 1.0809507876811325e-05, "loss": 0.0175, "step": 79740 }, { "grad_norm": 0.23908808827400208, "learning_rate": 1.0799241957282292e-05, "loss": 0.0143, "step": 79750 }, { "grad_norm": 0.37563788890838623, "learning_rate": 1.078898032468923e-05, "loss": 0.0179, "step": 79760 }, { "grad_norm": 0.21144211292266846, "learning_rate": 1.0778722980154315e-05, "loss": 0.0166, "step": 79770 }, { "grad_norm": 0.2395535260438919, "learning_rate": 1.0768469924799302e-05, "loss": 0.0125, "step": 79780 }, { "grad_norm": 0.2696355879306793, "learning_rate": 1.0758221159745419e-05, "loss": 0.0181, "step": 79790 }, { "grad_norm": 0.7202374935150146, "learning_rate": 1.074797668611347e-05, "loss": 0.0193, "step": 79800 }, { "grad_norm": 0.1419685333967209, "learning_rate": 1.0737736505023777e-05, "loss": 0.0169, "step": 79810 }, { "grad_norm": 0.18345746397972107, "learning_rate": 1.0727500617596203e-05, "loss": 0.017, "step": 79820 }, { "grad_norm": 0.2706604599952698, "learning_rate": 1.0717269024950104e-05, "loss": 0.0162, "step": 79830 }, { "grad_norm": 0.1714143007993698, "learning_rate": 1.070704172820442e-05, "loss": 0.0168, "step": 79840 }, { "grad_norm": 0.21049880981445312, "learning_rate": 1.0696818728477559e-05, "loss": 0.0163, "step": 79850 }, { "grad_norm": 0.2609007656574249, "learning_rate": 1.0686600026887511e-05, "loss": 0.0133, "step": 79860 }, { "grad_norm": 0.4004053473472595, "learning_rate": 1.0676385624551777e-05, "loss": 0.0173, "step": 79870 }, { "grad_norm": 0.2509816288948059, "learning_rate": 1.0666175522587402e-05, "loss": 0.019, "step": 79880 }, { "grad_norm": 0.41752538084983826, "learning_rate": 1.0655969722110925e-05, "loss": 0.0162, "step": 79890 }, { "grad_norm": 0.16271528601646423, "learning_rate": 1.0645768224238439e-05, "loss": 0.0148, "step": 79900 }, { "grad_norm": 0.35953494906425476, "learning_rate": 1.063557103008559e-05, "loss": 0.0155, "step": 79910 }, { "grad_norm": 0.32421764731407166, "learning_rate": 1.0625378140767494e-05, "loss": 0.0176, "step": 79920 }, { "grad_norm": 0.22581124305725098, "learning_rate": 1.0615189557398846e-05, "loss": 0.016, "step": 79930 }, { "grad_norm": 0.40104320645332336, "learning_rate": 1.060500528109386e-05, "loss": 0.0156, "step": 79940 }, { "grad_norm": 0.15532584488391876, "learning_rate": 1.0594825312966272e-05, "loss": 0.0136, "step": 79950 }, { "grad_norm": 0.2973516583442688, "learning_rate": 1.0584649654129336e-05, "loss": 0.0191, "step": 79960 }, { "grad_norm": 0.36765894293785095, "learning_rate": 1.0574478305695863e-05, "loss": 0.0143, "step": 79970 }, { "grad_norm": 0.3387444019317627, "learning_rate": 1.0564311268778154e-05, "loss": 0.0147, "step": 79980 }, { "grad_norm": 0.5231142044067383, "learning_rate": 1.0554148544488073e-05, "loss": 0.0215, "step": 79990 }, { "grad_norm": 0.23567886650562286, "learning_rate": 1.0543990133937004e-05, "loss": 0.0163, "step": 80000 } ], "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }