| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9968, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 2.4969531033325163, | |
| "learning_rate": 6.25e-08, | |
| "loss": 0.1252, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 2.008890344969409, | |
| "learning_rate": 1.25e-07, | |
| "loss": 0.1336, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 2.1487872414082574, | |
| "learning_rate": 1.875e-07, | |
| "loss": 0.1239, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 2.404877068082983, | |
| "learning_rate": 2.5e-07, | |
| "loss": 0.1534, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.9977341287068444, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "loss": 0.127, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 2.10253617342413, | |
| "learning_rate": 3.75e-07, | |
| "loss": 0.1295, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 1.9813074191320224, | |
| "learning_rate": 4.375e-07, | |
| "loss": 0.1443, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 2.1426342290851705, | |
| "learning_rate": 5e-07, | |
| "loss": 0.1489, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 2.3977225328887846, | |
| "learning_rate": 5.625e-07, | |
| "loss": 0.1715, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 2.4910499854065375, | |
| "learning_rate": 6.249999999999999e-07, | |
| "loss": 0.1481, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 2.0036101364507264, | |
| "learning_rate": 6.875e-07, | |
| "loss": 0.133, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.31397843050416, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.1197, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 2.789398930843854, | |
| "learning_rate": 8.125e-07, | |
| "loss": 0.1026, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.3679931786396997, | |
| "learning_rate": 8.75e-07, | |
| "loss": 0.1473, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.146960838741869, | |
| "learning_rate": 9.374999999999999e-07, | |
| "loss": 0.1499, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 2.8095519343937547, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1312, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 3.1559803771267987, | |
| "learning_rate": 1.0625e-06, | |
| "loss": 0.1332, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 2.6394797568137442, | |
| "learning_rate": 1.125e-06, | |
| "loss": 0.1516, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 2.0903056059241214, | |
| "learning_rate": 1.1874999999999999e-06, | |
| "loss": 0.1401, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 2.9496671534502785, | |
| "learning_rate": 1.2499999999999999e-06, | |
| "loss": 0.1422, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 2.2510408864309746, | |
| "learning_rate": 1.3125e-06, | |
| "loss": 0.1339, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 2.988822164056312, | |
| "learning_rate": 1.375e-06, | |
| "loss": 0.1372, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 2.425664327575325, | |
| "learning_rate": 1.4375e-06, | |
| "loss": 0.1419, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.2878647901959135, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.1643, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.373440045449061, | |
| "learning_rate": 1.5624999999999999e-06, | |
| "loss": 0.1611, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 3.1720114517373186, | |
| "learning_rate": 1.625e-06, | |
| "loss": 0.1241, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 2.888254088184939, | |
| "learning_rate": 1.6875e-06, | |
| "loss": 0.1432, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.4223635511865513, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.1013, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 2.540051698989579, | |
| "learning_rate": 1.8125e-06, | |
| "loss": 0.1524, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 3.0841788486084516, | |
| "learning_rate": 1.8749999999999998e-06, | |
| "loss": 0.1583, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 2.961021534474499, | |
| "learning_rate": 1.9375e-06, | |
| "loss": 0.1331, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 2.0666296501147112, | |
| "learning_rate": 2e-06, | |
| "loss": 0.1235, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 2.8434105826490987, | |
| "learning_rate": 1.9999370567547003e-06, | |
| "loss": 0.1728, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 3.2571488442302927, | |
| "learning_rate": 1.9997482349425066e-06, | |
| "loss": 0.1423, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 2.737978793809459, | |
| "learning_rate": 1.9994335583335335e-06, | |
| "loss": 0.1468, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 3.1545505286005997, | |
| "learning_rate": 1.9989930665413145e-06, | |
| "loss": 0.1169, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 2.797274179036167, | |
| "learning_rate": 1.9984268150178167e-06, | |
| "loss": 0.1319, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 2.8094170725156715, | |
| "learning_rate": 1.997734875046456e-06, | |
| "loss": 0.1052, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 3.265503821648089, | |
| "learning_rate": 1.996917333733128e-06, | |
| "loss": 0.1425, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 3.150913553614633, | |
| "learning_rate": 1.995974293995239e-06, | |
| "loss": 0.1335, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 2.8775764914997604, | |
| "learning_rate": 1.994905874548752e-06, | |
| "loss": 0.1226, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 3.2180785068208575, | |
| "learning_rate": 1.9937122098932426e-06, | |
| "loss": 0.1329, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 5.857949780062274, | |
| "learning_rate": 1.9923934502949643e-06, | |
| "loss": 0.152, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 4.521459145995853, | |
| "learning_rate": 1.9909497617679347e-06, | |
| "loss": 0.1747, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 3.1404514498803136, | |
| "learning_rate": 1.9893813260530367e-06, | |
| "loss": 0.11, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 4.201009323986255, | |
| "learning_rate": 1.9876883405951377e-06, | |
| "loss": 0.1634, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 2.9577036484677004, | |
| "learning_rate": 1.9858710185182355e-06, | |
| "loss": 0.1561, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 3.535826754862616, | |
| "learning_rate": 1.9839295885986295e-06, | |
| "loss": 0.1619, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 4.120602596813965, | |
| "learning_rate": 1.9818642952361183e-06, | |
| "loss": 0.1495, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 4.259764877810664, | |
| "learning_rate": 1.9796753984232355e-06, | |
| "loss": 0.1307, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 4.242488888827966, | |
| "learning_rate": 1.977363173712519e-06, | |
| "loss": 0.1285, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 4.069178515100502, | |
| "learning_rate": 1.9749279121818236e-06, | |
| "loss": 0.1352, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 4.544140717519671, | |
| "learning_rate": 1.9723699203976766e-06, | |
| "loss": 0.1571, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 3.9107597581959426, | |
| "learning_rate": 1.9696895203766866e-06, | |
| "loss": 0.1686, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 3.583429566299793, | |
| "learning_rate": 1.966887049545006e-06, | |
| "loss": 0.1378, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 3.517952555975872, | |
| "learning_rate": 1.9639628606958534e-06, | |
| "loss": 0.1386, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 3.579939014596777, | |
| "learning_rate": 1.9609173219450997e-06, | |
| "loss": 0.1448, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 4.961408389397768, | |
| "learning_rate": 1.9577508166849303e-06, | |
| "loss": 0.1334, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 3.881854573486764, | |
| "learning_rate": 1.9544637435355806e-06, | |
| "loss": 0.1453, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 4.160469207615619, | |
| "learning_rate": 1.9510565162951534e-06, | |
| "loss": 0.1514, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 3.7376823245313995, | |
| "learning_rate": 1.947529563887529e-06, | |
| "loss": 0.1756, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 4.745526341279435, | |
| "learning_rate": 1.9438833303083674e-06, | |
| "loss": 0.1903, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 3.5872890144465877, | |
| "learning_rate": 1.9401182745692187e-06, | |
| "loss": 0.1327, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 4.895892914029687, | |
| "learning_rate": 1.936234870639737e-06, | |
| "loss": 0.1348, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 3.4298662566625913, | |
| "learning_rate": 1.9322336073880143e-06, | |
| "loss": 0.1262, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 6.841317138496493, | |
| "learning_rate": 1.928114988519039e-06, | |
| "loss": 0.172, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 4.250165113109143, | |
| "learning_rate": 1.9238795325112867e-06, | |
| "loss": 0.1541, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 4.4033621224434025, | |
| "learning_rate": 1.9195277725514506e-06, | |
| "loss": 0.1745, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 3.98075383182803, | |
| "learning_rate": 1.91506025646732e-06, | |
| "loss": 0.1474, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 3.627291803014999, | |
| "learning_rate": 1.9104775466588157e-06, | |
| "loss": 0.1807, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 4.766210491599441, | |
| "learning_rate": 1.905780220027194e-06, | |
| "loss": 0.1502, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 3.6759268834811625, | |
| "learning_rate": 1.9009688679024189e-06, | |
| "loss": 0.1393, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 4.101009698917842, | |
| "learning_rate": 1.8960440959687252e-06, | |
| "loss": 0.1369, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 4.59569886944552, | |
| "learning_rate": 1.8910065241883678e-06, | |
| "loss": 0.1619, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.7413886299218797, | |
| "learning_rate": 1.8858567867235798e-06, | |
| "loss": 0.1666, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 3.8867223584125457, | |
| "learning_rate": 1.8805955318567379e-06, | |
| "loss": 0.1704, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 4.493023053104012, | |
| "learning_rate": 1.8752234219087537e-06, | |
| "loss": 0.1661, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 4.759475180875903, | |
| "learning_rate": 1.8697411331556953e-06, | |
| "loss": 0.1488, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 4.539234264514853, | |
| "learning_rate": 1.8641493557436548e-06, | |
| "loss": 0.1524, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 3.493235818987802, | |
| "learning_rate": 1.858448793601866e-06, | |
| "loss": 0.162, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 6.5293887794567596, | |
| "learning_rate": 1.852640164354092e-06, | |
| "loss": 0.1533, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 6.663555301541236, | |
| "learning_rate": 1.8467241992282841e-06, | |
| "loss": 0.1789, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 4.9493441573099695, | |
| "learning_rate": 1.8407016429645302e-06, | |
| "loss": 0.1563, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 3.5356024848568492, | |
| "learning_rate": 1.8345732537213026e-06, | |
| "loss": 0.1331, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 3.331011840972039, | |
| "learning_rate": 1.8283398029800164e-06, | |
| "loss": 0.1725, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 3.1323607027132603, | |
| "learning_rate": 1.82200207544791e-06, | |
| "loss": 0.1399, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 4.543709410888673, | |
| "learning_rate": 1.8155608689592601e-06, | |
| "loss": 0.14, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 4.45673006932675, | |
| "learning_rate": 1.8090169943749474e-06, | |
| "loss": 0.1615, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 4.705373269348066, | |
| "learning_rate": 1.802371275480378e-06, | |
| "loss": 0.1564, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 4.329301032695692, | |
| "learning_rate": 1.795624548881781e-06, | |
| "loss": 0.1924, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 3.216831310097902, | |
| "learning_rate": 1.7887776639008912e-06, | |
| "loss": 0.1636, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 4.125027796241628, | |
| "learning_rate": 1.7818314824680298e-06, | |
| "loss": 0.177, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 4.962921250352651, | |
| "learning_rate": 1.774786879013601e-06, | |
| "loss": 0.1525, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 3.8404707219541683, | |
| "learning_rate": 1.767644740358011e-06, | |
| "loss": 0.1332, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 4.711974795123313, | |
| "learning_rate": 1.760405965600031e-06, | |
| "loss": 0.1422, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 4.199038511043031, | |
| "learning_rate": 1.753071466003611e-06, | |
| "loss": 0.13, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 3.8291089013264115, | |
| "learning_rate": 1.7456421648831654e-06, | |
| "loss": 0.1715, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 3.939969260526378, | |
| "learning_rate": 1.7381189974873407e-06, | |
| "loss": 0.1406, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 5.4527369894058175, | |
| "learning_rate": 1.7305029108812774e-06, | |
| "loss": 0.1551, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.494861371187384, | |
| "learning_rate": 1.7227948638273915e-06, | |
| "loss": 0.1701, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 4.232388610306294, | |
| "learning_rate": 1.7149958266646754e-06, | |
| "loss": 0.1796, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 3.7860648199901012, | |
| "learning_rate": 1.7071067811865474e-06, | |
| "loss": 0.1466, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 4.380310274735351, | |
| "learning_rate": 1.6991287205172574e-06, | |
| "loss": 0.1476, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 6.315476074209258, | |
| "learning_rate": 1.6910626489868648e-06, | |
| "loss": 0.1676, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 4.236074270422003, | |
| "learning_rate": 1.682909582004807e-06, | |
| "loss": 0.1732, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 3.887311488635536, | |
| "learning_rate": 1.6746705459320744e-06, | |
| "loss": 0.1819, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 4.324784278862244, | |
| "learning_rate": 1.6663465779520037e-06, | |
| "loss": 0.1514, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 3.5010867620695336, | |
| "learning_rate": 1.6579387259397126e-06, | |
| "loss": 0.1378, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 4.547638459770409, | |
| "learning_rate": 1.6494480483301835e-06, | |
| "loss": 0.1245, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 4.154442920017231, | |
| "learning_rate": 1.640875613985024e-06, | |
| "loss": 0.1507, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 4.2738669740461805, | |
| "learning_rate": 1.6322225020579096e-06, | |
| "loss": 0.1578, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 3.030419612133566, | |
| "learning_rate": 1.6234898018587336e-06, | |
| "loss": 0.1248, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 3.236869339406748, | |
| "learning_rate": 1.6146786127164771e-06, | |
| "loss": 0.1522, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 3.0781522550717986, | |
| "learning_rate": 1.6057900438408199e-06, | |
| "loss": 0.1186, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 4.235359931610902, | |
| "learning_rate": 1.5968252141825035e-06, | |
| "loss": 0.181, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 3.8393666889310145, | |
| "learning_rate": 1.587785252292473e-06, | |
| "loss": 0.1951, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 4.533557115616342, | |
| "learning_rate": 1.578671296179806e-06, | |
| "loss": 0.1162, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 4.155646517913304, | |
| "learning_rate": 1.569484493168452e-06, | |
| "loss": 0.1569, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 3.612304887656086, | |
| "learning_rate": 1.5602259997528027e-06, | |
| "loss": 0.1657, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 3.4718761590803457, | |
| "learning_rate": 1.5508969814521024e-06, | |
| "loss": 0.1006, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 2.749613882614241, | |
| "learning_rate": 1.5414986126637257e-06, | |
| "loss": 0.1225, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 3.9202126219465185, | |
| "learning_rate": 1.5320320765153365e-06, | |
| "loss": 0.1651, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 3.769859629804315, | |
| "learning_rate": 1.5224985647159488e-06, | |
| "loss": 0.1207, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 4.258299529843038, | |
| "learning_rate": 1.5128992774059062e-06, | |
| "loss": 0.1485, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.5138242342415755, | |
| "learning_rate": 1.5032354230058002e-06, | |
| "loss": 0.1716, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 3.1536366453296063, | |
| "learning_rate": 1.4935082180643467e-06, | |
| "loss": 0.1567, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 3.8255471768086466, | |
| "learning_rate": 1.4837188871052397e-06, | |
| "loss": 0.1289, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 3.998586103793301, | |
| "learning_rate": 1.4738686624729987e-06, | |
| "loss": 0.1708, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 6.108329633585067, | |
| "learning_rate": 1.463958784177834e-06, | |
| "loss": 0.1513, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 3.5116343920977755, | |
| "learning_rate": 1.4539904997395467e-06, | |
| "loss": 0.1426, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 3.8466991449364487, | |
| "learning_rate": 1.4439650640304821e-06, | |
| "loss": 0.1879, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 3.145269853440717, | |
| "learning_rate": 1.433883739117558e-06, | |
| "loss": 0.116, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 3.6531846559334458, | |
| "learning_rate": 1.4237477941033886e-06, | |
| "loss": 0.1667, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 3.870433185715427, | |
| "learning_rate": 1.4135585049665206e-06, | |
| "loss": 0.1159, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 2.9424803019594252, | |
| "learning_rate": 1.4033171544008051e-06, | |
| "loss": 0.1345, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 3.1917917000111022, | |
| "learning_rate": 1.3930250316539235e-06, | |
| "loss": 0.1809, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 3.779080435520245, | |
| "learning_rate": 1.3826834323650898e-06, | |
| "loss": 0.1667, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 4.015156839459903, | |
| "learning_rate": 1.3722936584019451e-06, | |
| "loss": 0.1435, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 3.3828511532441867, | |
| "learning_rate": 1.3618570176966722e-06, | |
| "loss": 0.1763, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 2.7378078741101004, | |
| "learning_rate": 1.3513748240813427e-06, | |
| "loss": 0.124, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 2.667510129275553, | |
| "learning_rate": 1.3408483971225249e-06, | |
| "loss": 0.1276, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 3.4600760953915493, | |
| "learning_rate": 1.3302790619551672e-06, | |
| "loss": 0.1604, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 2.949770798393727, | |
| "learning_rate": 1.3196681491157816e-06, | |
| "loss": 0.1327, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 3.0805919917128564, | |
| "learning_rate": 1.3090169943749473e-06, | |
| "loss": 0.1466, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 3.869134321743603, | |
| "learning_rate": 1.298326938569156e-06, | |
| "loss": 0.1477, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 3.273813944999775, | |
| "learning_rate": 1.2875993274320173e-06, | |
| "loss": 0.1535, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 2.607928868194842, | |
| "learning_rate": 1.2768355114248492e-06, | |
| "loss": 0.1175, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 3.5206710496602436, | |
| "learning_rate": 1.266036845566675e-06, | |
| "loss": 0.1447, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 2.9796516371379327, | |
| "learning_rate": 1.2552046892636426e-06, | |
| "loss": 0.1509, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.9436385585618674, | |
| "learning_rate": 1.244340406137894e-06, | |
| "loss": 0.1506, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 4.010655685945475, | |
| "learning_rate": 1.2334453638559054e-06, | |
| "loss": 0.1641, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 3.808704120734902, | |
| "learning_rate": 1.2225209339563143e-06, | |
| "loss": 0.1409, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 4.080366555503018, | |
| "learning_rate": 1.211568491677263e-06, | |
| "loss": 0.1542, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 3.32431009022062, | |
| "learning_rate": 1.2005894157832728e-06, | |
| "loss": 0.1554, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 4.109905324890129, | |
| "learning_rate": 1.1895850883916785e-06, | |
| "loss": 0.1356, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 3.1041868788537768, | |
| "learning_rate": 1.1785568947986366e-06, | |
| "loss": 0.1448, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0048, | |
| "grad_norm": 3.9549654789059807, | |
| "learning_rate": 1.1675062233047363e-06, | |
| "loss": 0.1323, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 5.265789444979294, | |
| "learning_rate": 1.156434465040231e-06, | |
| "loss": 0.1336, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 2.278886843906884, | |
| "learning_rate": 1.1453430137899128e-06, | |
| "loss": 0.1036, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 2.994965707347646, | |
| "learning_rate": 1.1342332658176555e-06, | |
| "loss": 0.1081, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0304, | |
| "grad_norm": 2.707133385634132, | |
| "learning_rate": 1.123106619690643e-06, | |
| "loss": 0.1407, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 2.151215553915982, | |
| "learning_rate": 1.1119644761033077e-06, | |
| "loss": 0.1119, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0432, | |
| "grad_norm": 2.0387940365716397, | |
| "learning_rate": 1.1008082377010045e-06, | |
| "loss": 0.1272, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 2.0188887298089755, | |
| "learning_rate": 1.0896393089034335e-06, | |
| "loss": 0.1244, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 3.5460776700681955, | |
| "learning_rate": 1.078459095727845e-06, | |
| "loss": 0.1104, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 2.360145390442641, | |
| "learning_rate": 1.0672690056120398e-06, | |
| "loss": 0.1368, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0688, | |
| "grad_norm": 2.7286076638804198, | |
| "learning_rate": 1.0560704472371917e-06, | |
| "loss": 0.098, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 3.9363201426551226, | |
| "learning_rate": 1.044864830350515e-06, | |
| "loss": 0.1683, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0816, | |
| "grad_norm": 3.2428028503112993, | |
| "learning_rate": 1.033653565587794e-06, | |
| "loss": 0.1142, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 3.113578605505914, | |
| "learning_rate": 1.022438064295805e-06, | |
| "loss": 0.1129, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 2.8114334244479684, | |
| "learning_rate": 1.0112197383546459e-06, | |
| "loss": 0.1279, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 3.1925580191789242, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1497, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1072, | |
| "grad_norm": 2.9392401643163497, | |
| "learning_rate": 9.88780261645354e-07, | |
| "loss": 0.1281, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 3.2004694564966583, | |
| "learning_rate": 9.77561935704195e-07, | |
| "loss": 0.113, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.7402823405153574, | |
| "learning_rate": 9.663464344122063e-07, | |
| "loss": 0.0722, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 3.870543794283185, | |
| "learning_rate": 9.551351696494853e-07, | |
| "loss": 0.145, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 3.12656262154436, | |
| "learning_rate": 9.43929552762808e-07, | |
| "loss": 0.1343, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 3.1524992206404527, | |
| "learning_rate": 9.327309943879603e-07, | |
| "loss": 0.129, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1456, | |
| "grad_norm": 3.693607421239943, | |
| "learning_rate": 9.215409042721551e-07, | |
| "loss": 0.1206, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 2.589250388541951, | |
| "learning_rate": 9.103606910965665e-07, | |
| "loss": 0.1046, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1584, | |
| "grad_norm": 3.9569120653267245, | |
| "learning_rate": 8.991917622989955e-07, | |
| "loss": 0.1437, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 3.0436417760418824, | |
| "learning_rate": 8.880355238966921e-07, | |
| "loss": 0.1131, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 2.7404896523288156, | |
| "learning_rate": 8.768933803093572e-07, | |
| "loss": 0.1182, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 2.978196221491147, | |
| "learning_rate": 8.657667341823448e-07, | |
| "loss": 0.1331, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 2.7953950740003948, | |
| "learning_rate": 8.546569862100875e-07, | |
| "loss": 0.1773, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 3.505293558413482, | |
| "learning_rate": 8.435655349597689e-07, | |
| "loss": 0.1289, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.1968, | |
| "grad_norm": 3.017122332004861, | |
| "learning_rate": 8.324937766952636e-07, | |
| "loss": 0.1264, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 3.4700813765221294, | |
| "learning_rate": 8.214431052013634e-07, | |
| "loss": 0.123, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 3.3325698466760905, | |
| "learning_rate": 8.104149116083216e-07, | |
| "loss": 0.1323, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 3.3688329432245645, | |
| "learning_rate": 7.994105842167272e-07, | |
| "loss": 0.1445, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2224, | |
| "grad_norm": 5.410904469002936, | |
| "learning_rate": 7.884315083227372e-07, | |
| "loss": 0.1348, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 2.8934452953623695, | |
| "learning_rate": 7.774790660436857e-07, | |
| "loss": 0.1042, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2352, | |
| "grad_norm": 4.515892407436011, | |
| "learning_rate": 7.665546361440949e-07, | |
| "loss": 0.1344, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 3.3336860990470583, | |
| "learning_rate": 7.556595938621058e-07, | |
| "loss": 0.1132, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 3.3781317545242686, | |
| "learning_rate": 7.447953107363574e-07, | |
| "loss": 0.1223, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 3.4052474284643175, | |
| "learning_rate": 7.33963154433325e-07, | |
| "loss": 0.1147, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.2608, | |
| "grad_norm": 3.260066938920247, | |
| "learning_rate": 7.231644885751507e-07, | |
| "loss": 0.1162, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 2.075129912437651, | |
| "learning_rate": 7.124006725679828e-07, | |
| "loss": 0.0891, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.2736, | |
| "grad_norm": 2.5796448583935994, | |
| "learning_rate": 7.016730614308439e-07, | |
| "loss": 0.1453, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 3.3625938910552935, | |
| "learning_rate": 6.909830056250526e-07, | |
| "loss": 0.1136, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 2.7459563636737707, | |
| "learning_rate": 6.803318508842186e-07, | |
| "loss": 0.135, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 3.6833307558438695, | |
| "learning_rate": 6.697209380448332e-07, | |
| "loss": 0.0941, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.2992, | |
| "grad_norm": 2.9375208875067464, | |
| "learning_rate": 6.59151602877475e-07, | |
| "loss": 0.1244, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 3.7877853756064708, | |
| "learning_rate": 6.486251759186572e-07, | |
| "loss": 0.0908, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.312, | |
| "grad_norm": 2.626489357460925, | |
| "learning_rate": 6.381429823033279e-07, | |
| "loss": 0.1147, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 2.7246829179484475, | |
| "learning_rate": 6.277063415980548e-07, | |
| "loss": 0.1111, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 2.793990841290057, | |
| "learning_rate": 6.173165676349102e-07, | |
| "loss": 0.1236, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 3.067033719132832, | |
| "learning_rate": 6.069749683460764e-07, | |
| "loss": 0.1269, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3376000000000001, | |
| "grad_norm": 2.62953934675259, | |
| "learning_rate": 5.96682845599195e-07, | |
| "loss": 0.1296, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 2.295428084618355, | |
| "learning_rate": 5.864414950334795e-07, | |
| "loss": 0.1067, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3504, | |
| "grad_norm": 2.494288637796771, | |
| "learning_rate": 5.762522058966113e-07, | |
| "loss": 0.1012, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 2.7289466545252856, | |
| "learning_rate": 5.661162608824419e-07, | |
| "loss": 0.1263, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 3.297503764337652, | |
| "learning_rate": 5.56034935969518e-07, | |
| "loss": 0.1175, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 2.733287987985122, | |
| "learning_rate": 5.460095002604532e-07, | |
| "loss": 0.1125, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 2.7994559067104574, | |
| "learning_rate": 5.36041215822166e-07, | |
| "loss": 0.1519, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 2.7375733905542643, | |
| "learning_rate": 5.261313375270013e-07, | |
| "loss": 0.1173, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.3888, | |
| "grad_norm": 3.492872700831482, | |
| "learning_rate": 5.162811128947602e-07, | |
| "loss": 0.1594, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 3.538990737016252, | |
| "learning_rate": 5.064917819356531e-07, | |
| "loss": 0.1121, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 3.456492027602956, | |
| "learning_rate": 4.967645769941999e-07, | |
| "loss": 0.1329, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 6.8845618461429225, | |
| "learning_rate": 4.871007225940939e-07, | |
| "loss": 0.1215, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4144, | |
| "grad_norm": 4.538034399979344, | |
| "learning_rate": 4.775014352840512e-07, | |
| "loss": 0.1148, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 3.381941347409583, | |
| "learning_rate": 4.6796792348466353e-07, | |
| "loss": 0.1102, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4272, | |
| "grad_norm": 2.3402894784761856, | |
| "learning_rate": 4.585013873362743e-07, | |
| "loss": 0.1208, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 2.92152467138924, | |
| "learning_rate": 4.4910301854789755e-07, | |
| "loss": 0.1377, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 2.5271124781868988, | |
| "learning_rate": 4.397740002471972e-07, | |
| "loss": 0.1004, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 3.2747270637549777, | |
| "learning_rate": 4.3051550683154804e-07, | |
| "loss": 0.1281, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.4527999999999999, | |
| "grad_norm": 2.6095079208452185, | |
| "learning_rate": 4.2132870382019427e-07, | |
| "loss": 0.1055, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 5.83458655577108, | |
| "learning_rate": 4.1221474770752696e-07, | |
| "loss": 0.1215, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4656, | |
| "grad_norm": 3.1153974927503763, | |
| "learning_rate": 4.031747858174964e-07, | |
| "loss": 0.0968, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 4.929073484618664, | |
| "learning_rate": 3.942099561591802e-07, | |
| "loss": 0.1048, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 2.5166547994261417, | |
| "learning_rate": 3.853213872835228e-07, | |
| "loss": 0.1058, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 2.4668836440326865, | |
| "learning_rate": 3.765101981412665e-07, | |
| "loss": 0.1308, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.4912, | |
| "grad_norm": 1.9951943331950452, | |
| "learning_rate": 3.677774979420903e-07, | |
| "loss": 0.1018, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 2.9908171754511437, | |
| "learning_rate": 3.5912438601497584e-07, | |
| "loss": 0.1161, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.504, | |
| "grad_norm": 3.3264145678441976, | |
| "learning_rate": 3.5055195166981646e-07, | |
| "loss": 0.128, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 3.0003805177406733, | |
| "learning_rate": 3.420612740602874e-07, | |
| "loss": 0.1273, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 4.210870037523149, | |
| "learning_rate": 3.3365342204799606e-07, | |
| "loss": 0.1269, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 3.1058783196727666, | |
| "learning_rate": 3.253294540679257e-07, | |
| "loss": 0.1253, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5295999999999998, | |
| "grad_norm": 2.7923304271341607, | |
| "learning_rate": 3.170904179951931e-07, | |
| "loss": 0.112, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 3.193465065928833, | |
| "learning_rate": 3.0893735101313535e-07, | |
| "loss": 0.1188, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5424, | |
| "grad_norm": 3.204335563563906, | |
| "learning_rate": 3.008712794827426e-07, | |
| "loss": 0.1384, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 2.5887872249352806, | |
| "learning_rate": 2.9289321881345254e-07, | |
| "loss": 0.124, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 2.6928108437347698, | |
| "learning_rate": 2.850041733353247e-07, | |
| "loss": 0.1071, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 2.6514260220506563, | |
| "learning_rate": 2.7720513617260855e-07, | |
| "loss": 0.1284, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 3.4173755675031807, | |
| "learning_rate": 2.6949708911872247e-07, | |
| "loss": 0.1106, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 3.964433823626616, | |
| "learning_rate": 2.6188100251265943e-07, | |
| "loss": 0.1087, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.5808, | |
| "grad_norm": 4.629149748841336, | |
| "learning_rate": 2.543578351168344e-07, | |
| "loss": 0.1148, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 3.4337305056813197, | |
| "learning_rate": 2.4692853399638913e-07, | |
| "loss": 0.1387, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 3.71914004045554, | |
| "learning_rate": 2.395940343999691e-07, | |
| "loss": 0.1216, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 6.720706070128215, | |
| "learning_rate": 2.3235525964198888e-07, | |
| "loss": 0.1501, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6064, | |
| "grad_norm": 3.30556531360598, | |
| "learning_rate": 2.252131209863991e-07, | |
| "loss": 0.1052, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 4.4642941608850055, | |
| "learning_rate": 2.181685175319702e-07, | |
| "loss": 0.139, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.6192, | |
| "grad_norm": 2.6766521332963777, | |
| "learning_rate": 2.11222336099109e-07, | |
| "loss": 0.0887, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 3.3097474207466444, | |
| "learning_rate": 2.043754511182191e-07, | |
| "loss": 0.1218, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 3.062260478274747, | |
| "learning_rate": 1.9762872451962208e-07, | |
| "loss": 0.1226, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 3.2630102453572882, | |
| "learning_rate": 1.9098300562505264e-07, | |
| "loss": 0.0844, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6448, | |
| "grad_norm": 2.7218126147886945, | |
| "learning_rate": 1.8443913104073982e-07, | |
| "loss": 0.1163, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 3.7971992271480186, | |
| "learning_rate": 1.7799792455209016e-07, | |
| "loss": 0.1181, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6576, | |
| "grad_norm": 3.738441130140733, | |
| "learning_rate": 1.716601970199836e-07, | |
| "loss": 0.1301, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 3.05018064730553, | |
| "learning_rate": 1.6542674627869734e-07, | |
| "loss": 0.0798, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 3.2523350421667363, | |
| "learning_rate": 1.592983570354699e-07, | |
| "loss": 0.1082, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 3.284724299450673, | |
| "learning_rate": 1.5327580077171588e-07, | |
| "loss": 0.1207, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.6832, | |
| "grad_norm": 3.92886529982871, | |
| "learning_rate": 1.473598356459078e-07, | |
| "loss": 0.0965, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 3.288393589630192, | |
| "learning_rate": 1.415512063981339e-07, | |
| "loss": 0.1356, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.696, | |
| "grad_norm": 2.4891167660291917, | |
| "learning_rate": 1.358506442563454e-07, | |
| "loss": 0.1205, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 2.749129320779883, | |
| "learning_rate": 1.3025886684430465e-07, | |
| "loss": 0.1114, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 2.9939547397600887, | |
| "learning_rate": 1.2477657809124632e-07, | |
| "loss": 0.1267, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 3.059481363897511, | |
| "learning_rate": 1.19404468143262e-07, | |
| "loss": 0.13, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7216, | |
| "grad_norm": 3.172719990097923, | |
| "learning_rate": 1.1414321327642019e-07, | |
| "loss": 0.1003, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 2.763074619615047, | |
| "learning_rate": 1.089934758116322e-07, | |
| "loss": 0.107, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7344, | |
| "grad_norm": 3.3513605159844477, | |
| "learning_rate": 1.0395590403127486e-07, | |
| "loss": 0.1, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 3.0130218075200284, | |
| "learning_rate": 9.903113209758096e-08, | |
| "loss": 0.1014, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 3.2984153616886442, | |
| "learning_rate": 9.421977997280594e-08, | |
| "loss": 0.1443, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 2.4116503417935977, | |
| "learning_rate": 8.952245334118413e-08, | |
| "loss": 0.129, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.303377440533526, | |
| "learning_rate": 8.493974353268019e-08, | |
| "loss": 0.1433, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 2.575473010066114, | |
| "learning_rate": 8.047222744854942e-08, | |
| "loss": 0.0928, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.7728000000000002, | |
| "grad_norm": 2.862390938762308, | |
| "learning_rate": 7.612046748871326e-08, | |
| "loss": 0.1003, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 2.512371847069459, | |
| "learning_rate": 7.188501148096116e-08, | |
| "loss": 0.1068, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 3.9557015566480285, | |
| "learning_rate": 6.77663926119858e-08, | |
| "loss": 0.111, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 3.237819756927862, | |
| "learning_rate": 6.376512936026279e-08, | |
| "loss": 0.1242, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7984, | |
| "grad_norm": 2.3340318791785646, | |
| "learning_rate": 5.988172543078096e-08, | |
| "loss": 0.1061, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 2.912775412360275, | |
| "learning_rate": 5.611666969163242e-08, | |
| "loss": 0.1288, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.8112, | |
| "grad_norm": 3.5993922569990877, | |
| "learning_rate": 5.2470436112471264e-08, | |
| "loss": 0.1502, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 3.1906595269687066, | |
| "learning_rate": 4.8943483704846465e-08, | |
| "loss": 0.1129, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 2.7886058393742177, | |
| "learning_rate": 4.553625646441928e-08, | |
| "loss": 0.1417, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 4.576338749251723, | |
| "learning_rate": 4.224918331506955e-08, | |
| "loss": 0.1079, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8368, | |
| "grad_norm": 2.3424965325338434, | |
| "learning_rate": 3.908267805490051e-08, | |
| "loss": 0.113, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 2.8896940146945, | |
| "learning_rate": 3.6037139304146756e-08, | |
| "loss": 0.132, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.8496000000000001, | |
| "grad_norm": 3.051287182293896, | |
| "learning_rate": 3.3112950454993625e-08, | |
| "loss": 0.1092, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 3.4677503211105645, | |
| "learning_rate": 3.0310479623313125e-08, | |
| "loss": 0.1255, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 2.629396760631384, | |
| "learning_rate": 2.7630079602323443e-08, | |
| "loss": 0.1082, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 3.060707277576519, | |
| "learning_rate": 2.507208781817638e-08, | |
| "loss": 0.1289, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.8752, | |
| "grad_norm": 4.498542316071858, | |
| "learning_rate": 2.263682628748087e-08, | |
| "loss": 0.1112, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 3.3242615389567653, | |
| "learning_rate": 2.032460157676452e-08, | |
| "loss": 0.1402, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.888, | |
| "grad_norm": 2.117823787343076, | |
| "learning_rate": 1.8135704763881598e-08, | |
| "loss": 0.1214, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 3.7999512049595867, | |
| "learning_rate": 1.607041140137033e-08, | |
| "loss": 0.106, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 2.9706181280580086, | |
| "learning_rate": 1.4128981481764113e-08, | |
| "loss": 0.1121, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 2.3832740990053702, | |
| "learning_rate": 1.231165940486234e-08, | |
| "loss": 0.1084, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9136, | |
| "grad_norm": 2.686907690774645, | |
| "learning_rate": 1.0618673946963364e-08, | |
| "loss": 0.1138, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.632510645620075, | |
| "learning_rate": 9.050238232065299e-09, | |
| "loss": 0.1069, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9264000000000001, | |
| "grad_norm": 3.3133984005584614, | |
| "learning_rate": 7.606549705035935e-09, | |
| "loss": 0.1158, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 3.033325678592628, | |
| "learning_rate": 6.2877901067573955e-09, | |
| "loss": 0.109, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 2.625891227751338, | |
| "learning_rate": 5.094125451247655e-09, | |
| "loss": 0.1198, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 4.0371687332878645, | |
| "learning_rate": 4.025706004760931e-09, | |
| "loss": 0.1089, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 2.681798202229694, | |
| "learning_rate": 3.082666266872036e-09, | |
| "loss": 0.1012, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 2.3816672358564093, | |
| "learning_rate": 2.2651249535439177e-09, | |
| "loss": 0.132, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.9647999999999999, | |
| "grad_norm": 2.731557615502218, | |
| "learning_rate": 1.5731849821833953e-09, | |
| "loss": 0.1205, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 2.3980291292550673, | |
| "learning_rate": 1.0069334586854105e-09, | |
| "loss": 0.0922, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 2.262332556662717, | |
| "learning_rate": 5.664416664666882e-10, | |
| "loss": 0.1116, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 3.7908540657365184, | |
| "learning_rate": 2.517650574934693e-10, | |
| "loss": 0.1022, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.9904, | |
| "grad_norm": 2.5746244022394906, | |
| "learning_rate": 6.29432452994294e-11, | |
| "loss": 0.1039, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 2.1856958417394132, | |
| "learning_rate": 0.0, | |
| "loss": 0.1103, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "step": 312, | |
| "total_flos": 211349000945664.0, | |
| "train_loss": 0.13338581045182088, | |
| "train_runtime": 10585.2413, | |
| "train_samples_per_second": 7.556, | |
| "train_steps_per_second": 0.029 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 211349000945664.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |