| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.6008230452674898, | |
| "eval_steps": 110, | |
| "global_step": 438, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0013717421124828531, | |
| "grad_norm": 13.75906753540039, | |
| "learning_rate": 1.8281535648994516e-06, | |
| "loss": 1.2468, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0027434842249657062, | |
| "grad_norm": 14.05058765411377, | |
| "learning_rate": 3.6563071297989032e-06, | |
| "loss": 1.4692, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00411522633744856, | |
| "grad_norm": 13.991771697998047, | |
| "learning_rate": 5.484460694698355e-06, | |
| "loss": 1.2457, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0054869684499314125, | |
| "grad_norm": 13.429465293884277, | |
| "learning_rate": 7.3126142595978065e-06, | |
| "loss": 1.1859, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006858710562414266, | |
| "grad_norm": 12.403002738952637, | |
| "learning_rate": 9.140767824497258e-06, | |
| "loss": 1.2404, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00823045267489712, | |
| "grad_norm": 1.5524662733078003, | |
| "learning_rate": 1.096892138939671e-05, | |
| "loss": 0.042, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.009602194787379973, | |
| "grad_norm": 10.1494140625, | |
| "learning_rate": 1.2797074954296162e-05, | |
| "loss": 0.8856, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.010973936899862825, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.2797074954296162e-05, | |
| "loss": 1.5417, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.012345679012345678, | |
| "grad_norm": 0.4553964138031006, | |
| "learning_rate": 1.4625228519195613e-05, | |
| "loss": 0.023, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.013717421124828532, | |
| "grad_norm": 10.332369804382324, | |
| "learning_rate": 1.6453382084095062e-05, | |
| "loss": 0.8655, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015089163237311385, | |
| "grad_norm": 7.544310092926025, | |
| "learning_rate": 1.8281535648994517e-05, | |
| "loss": 0.5894, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01646090534979424, | |
| "grad_norm": 8.16427230834961, | |
| "learning_rate": 2.0109689213893968e-05, | |
| "loss": 0.7053, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01783264746227709, | |
| "grad_norm": 7.403252601623535, | |
| "learning_rate": 2.193784277879342e-05, | |
| "loss": 0.5857, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.019204389574759947, | |
| "grad_norm": 8.974674224853516, | |
| "learning_rate": 2.376599634369287e-05, | |
| "loss": 0.8375, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0205761316872428, | |
| "grad_norm": 13.417745590209961, | |
| "learning_rate": 2.5594149908592324e-05, | |
| "loss": 0.9043, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02194787379972565, | |
| "grad_norm": 12.881294250488281, | |
| "learning_rate": 2.742230347349177e-05, | |
| "loss": 0.8756, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.023319615912208505, | |
| "grad_norm": 7.439205169677734, | |
| "learning_rate": 2.9250457038391226e-05, | |
| "loss": 0.5076, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.024691358024691357, | |
| "grad_norm": 8.46964168548584, | |
| "learning_rate": 3.107861060329068e-05, | |
| "loss": 0.4757, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02606310013717421, | |
| "grad_norm": 17.02773666381836, | |
| "learning_rate": 3.2906764168190124e-05, | |
| "loss": 0.9993, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.027434842249657063, | |
| "grad_norm": 6.2668776512146, | |
| "learning_rate": 3.473491773308958e-05, | |
| "loss": 0.2622, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02880658436213992, | |
| "grad_norm": 8.273824691772461, | |
| "learning_rate": 3.656307129798903e-05, | |
| "loss": 0.3497, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03017832647462277, | |
| "grad_norm": 5.5460944175720215, | |
| "learning_rate": 3.839122486288849e-05, | |
| "loss": 0.2514, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.03155006858710562, | |
| "grad_norm": 4.283128261566162, | |
| "learning_rate": 4.0219378427787935e-05, | |
| "loss": 0.1673, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03292181069958848, | |
| "grad_norm": 4.708792209625244, | |
| "learning_rate": 4.204753199268738e-05, | |
| "loss": 0.203, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03429355281207133, | |
| "grad_norm": 14.491021156311035, | |
| "learning_rate": 4.387568555758684e-05, | |
| "loss": 0.698, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03566529492455418, | |
| "grad_norm": 7.903520584106445, | |
| "learning_rate": 4.570383912248629e-05, | |
| "loss": 0.3401, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.037037037037037035, | |
| "grad_norm": 7.333080291748047, | |
| "learning_rate": 4.753199268738574e-05, | |
| "loss": 0.2185, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.038408779149519894, | |
| "grad_norm": 8.625358581542969, | |
| "learning_rate": 4.936014625228519e-05, | |
| "loss": 0.4424, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.039780521262002745, | |
| "grad_norm": 1.5588488578796387, | |
| "learning_rate": 5.118829981718465e-05, | |
| "loss": 0.0381, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0411522633744856, | |
| "grad_norm": 12.401138305664062, | |
| "learning_rate": 5.3016453382084095e-05, | |
| "loss": 0.8215, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04252400548696845, | |
| "grad_norm": 5.405845642089844, | |
| "learning_rate": 5.484460694698354e-05, | |
| "loss": 0.1542, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0438957475994513, | |
| "grad_norm": 8.558808326721191, | |
| "learning_rate": 5.6672760511883e-05, | |
| "loss": 0.6893, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04526748971193416, | |
| "grad_norm": 7.206741809844971, | |
| "learning_rate": 5.850091407678245e-05, | |
| "loss": 0.3773, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.04663923182441701, | |
| "grad_norm": 8.300729751586914, | |
| "learning_rate": 6.0329067641681906e-05, | |
| "loss": 0.538, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04801097393689986, | |
| "grad_norm": 0.2500181496143341, | |
| "learning_rate": 6.215722120658135e-05, | |
| "loss": 0.0073, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04938271604938271, | |
| "grad_norm": 17.457223892211914, | |
| "learning_rate": 6.398537477148081e-05, | |
| "loss": 2.378, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05075445816186557, | |
| "grad_norm": 10.884990692138672, | |
| "learning_rate": 6.581352833638025e-05, | |
| "loss": 0.5949, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05212620027434842, | |
| "grad_norm": 10.013723373413086, | |
| "learning_rate": 6.764168190127972e-05, | |
| "loss": 0.7071, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.053497942386831275, | |
| "grad_norm": 4.653324604034424, | |
| "learning_rate": 6.946983546617916e-05, | |
| "loss": 0.1607, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.05486968449931413, | |
| "grad_norm": 9.527400970458984, | |
| "learning_rate": 7.129798903107861e-05, | |
| "loss": 0.7735, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.056241426611796985, | |
| "grad_norm": 12.477531433105469, | |
| "learning_rate": 7.312614259597807e-05, | |
| "loss": 0.7594, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.05761316872427984, | |
| "grad_norm": 5.369799613952637, | |
| "learning_rate": 7.495429616087751e-05, | |
| "loss": 0.3569, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.05898491083676269, | |
| "grad_norm": 5.1385908126831055, | |
| "learning_rate": 7.678244972577697e-05, | |
| "loss": 0.2454, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.06035665294924554, | |
| "grad_norm": 6.1807708740234375, | |
| "learning_rate": 7.861060329067642e-05, | |
| "loss": 0.2723, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 7.941879749298096, | |
| "learning_rate": 8.043875685557587e-05, | |
| "loss": 0.5338, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06310013717421124, | |
| "grad_norm": 5.015410423278809, | |
| "learning_rate": 8.226691042047532e-05, | |
| "loss": 0.1891, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0644718792866941, | |
| "grad_norm": 7.299699306488037, | |
| "learning_rate": 8.409506398537477e-05, | |
| "loss": 0.3647, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.06584362139917696, | |
| "grad_norm": 8.421393394470215, | |
| "learning_rate": 8.592321755027423e-05, | |
| "loss": 0.383, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.06721536351165981, | |
| "grad_norm": 5.5915937423706055, | |
| "learning_rate": 8.775137111517367e-05, | |
| "loss": 0.2353, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06858710562414266, | |
| "grad_norm": 8.187829971313477, | |
| "learning_rate": 8.957952468007313e-05, | |
| "loss": 0.5541, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06995884773662552, | |
| "grad_norm": 6.386786460876465, | |
| "learning_rate": 9.140767824497258e-05, | |
| "loss": 0.4908, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07133058984910837, | |
| "grad_norm": 8.64050006866455, | |
| "learning_rate": 9.323583180987204e-05, | |
| "loss": 0.586, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07270233196159122, | |
| "grad_norm": 5.879551410675049, | |
| "learning_rate": 9.506398537477148e-05, | |
| "loss": 0.2241, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 7.824138164520264, | |
| "learning_rate": 9.689213893967093e-05, | |
| "loss": 0.6046, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.07544581618655692, | |
| "grad_norm": 6.351109504699707, | |
| "learning_rate": 9.872029250457039e-05, | |
| "loss": 0.231, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07681755829903979, | |
| "grad_norm": 9.437410354614258, | |
| "learning_rate": 0.00010054844606946984, | |
| "loss": 0.7105, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07818930041152264, | |
| "grad_norm": 8.40911865234375, | |
| "learning_rate": 0.0001023765996343693, | |
| "loss": 0.5591, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.07956104252400549, | |
| "grad_norm": 7.631382942199707, | |
| "learning_rate": 0.00010420475319926874, | |
| "loss": 0.5194, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08093278463648834, | |
| "grad_norm": 5.773220062255859, | |
| "learning_rate": 0.00010603290676416819, | |
| "loss": 0.3297, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0823045267489712, | |
| "grad_norm": 1.3606321811676025, | |
| "learning_rate": 0.00010786106032906765, | |
| "loss": 0.0299, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08367626886145405, | |
| "grad_norm": 7.216275215148926, | |
| "learning_rate": 0.00010968921389396709, | |
| "loss": 0.3514, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0850480109739369, | |
| "grad_norm": 4.70477294921875, | |
| "learning_rate": 0.00011151736745886655, | |
| "loss": 0.1932, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.08641975308641975, | |
| "grad_norm": 6.754104137420654, | |
| "learning_rate": 0.000113345521023766, | |
| "loss": 0.4035, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.0877914951989026, | |
| "grad_norm": 0.19067375361919403, | |
| "learning_rate": 0.00011517367458866546, | |
| "loss": 0.0094, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.08916323731138547, | |
| "grad_norm": 1.1715893745422363, | |
| "learning_rate": 0.0001170018281535649, | |
| "loss": 0.0148, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09053497942386832, | |
| "grad_norm": 1.6287739276885986, | |
| "learning_rate": 0.00011882998171846434, | |
| "loss": 0.0231, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.09190672153635117, | |
| "grad_norm": 7.027708053588867, | |
| "learning_rate": 0.00012065813528336381, | |
| "loss": 0.3204, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.09327846364883402, | |
| "grad_norm": 7.248253345489502, | |
| "learning_rate": 0.00012248628884826325, | |
| "loss": 0.3011, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.09465020576131687, | |
| "grad_norm": 9.592718124389648, | |
| "learning_rate": 0.0001243144424131627, | |
| "loss": 0.3871, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.09602194787379972, | |
| "grad_norm": 5.128874778747559, | |
| "learning_rate": 0.00012614259597806216, | |
| "loss": 0.1823, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09739368998628258, | |
| "grad_norm": 6.496853351593018, | |
| "learning_rate": 0.00012797074954296162, | |
| "loss": 0.3572, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.09876543209876543, | |
| "grad_norm": 6.564659118652344, | |
| "learning_rate": 0.00012979890310786104, | |
| "loss": 0.5289, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.10013717421124829, | |
| "grad_norm": 6.480371952056885, | |
| "learning_rate": 0.0001316270566727605, | |
| "loss": 0.3223, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.10150891632373114, | |
| "grad_norm": 7.222306728363037, | |
| "learning_rate": 0.00013345521023765998, | |
| "loss": 0.3247, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.102880658436214, | |
| "grad_norm": 5.406076431274414, | |
| "learning_rate": 0.00013528336380255943, | |
| "loss": 0.2133, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.10425240054869685, | |
| "grad_norm": 11.029163360595703, | |
| "learning_rate": 0.00013711151736745886, | |
| "loss": 0.8249, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1056241426611797, | |
| "grad_norm": 7.284115314483643, | |
| "learning_rate": 0.00013893967093235832, | |
| "loss": 0.4341, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.10699588477366255, | |
| "grad_norm": 6.240738868713379, | |
| "learning_rate": 0.00014076782449725777, | |
| "loss": 0.2932, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1083676268861454, | |
| "grad_norm": 0.1745665967464447, | |
| "learning_rate": 0.00014259597806215722, | |
| "loss": 0.0099, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.10973936899862825, | |
| "grad_norm": 5.460353851318359, | |
| "learning_rate": 0.00014442413162705668, | |
| "loss": 0.3348, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 8.869246482849121, | |
| "learning_rate": 0.00014625228519195613, | |
| "loss": 0.6405, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.11248285322359397, | |
| "grad_norm": 4.475996971130371, | |
| "learning_rate": 0.0001480804387568556, | |
| "loss": 0.1536, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.11385459533607682, | |
| "grad_norm": 2.700299024581909, | |
| "learning_rate": 0.00014990859232175501, | |
| "loss": 0.1299, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.11522633744855967, | |
| "grad_norm": 7.5515618324279785, | |
| "learning_rate": 0.00015173674588665447, | |
| "loss": 0.5863, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.11659807956104253, | |
| "grad_norm": 9.869407653808594, | |
| "learning_rate": 0.00015356489945155395, | |
| "loss": 0.7205, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11796982167352538, | |
| "grad_norm": 8.208423614501953, | |
| "learning_rate": 0.00015539305301645338, | |
| "loss": 0.4052, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.11934156378600823, | |
| "grad_norm": 6.408420562744141, | |
| "learning_rate": 0.00015722120658135283, | |
| "loss": 0.3953, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.12071330589849108, | |
| "grad_norm": 7.050099849700928, | |
| "learning_rate": 0.00015904936014625229, | |
| "loss": 0.5598, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.12208504801097393, | |
| "grad_norm": 5.326991558074951, | |
| "learning_rate": 0.00016087751371115174, | |
| "loss": 0.2856, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 4.510193347930908, | |
| "learning_rate": 0.0001627056672760512, | |
| "loss": 0.2277, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12482853223593965, | |
| "grad_norm": 5.769596576690674, | |
| "learning_rate": 0.00016453382084095065, | |
| "loss": 0.3296, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1262002743484225, | |
| "grad_norm": 6.066390037536621, | |
| "learning_rate": 0.0001663619744058501, | |
| "loss": 0.3079, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.12757201646090535, | |
| "grad_norm": 6.80173921585083, | |
| "learning_rate": 0.00016819012797074953, | |
| "loss": 0.4867, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1289437585733882, | |
| "grad_norm": 6.219693183898926, | |
| "learning_rate": 0.00017001828153564899, | |
| "loss": 0.4319, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.13031550068587106, | |
| "grad_norm": 5.316290855407715, | |
| "learning_rate": 0.00017184643510054847, | |
| "loss": 0.2952, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.13168724279835392, | |
| "grad_norm": 6.86447811126709, | |
| "learning_rate": 0.00017367458866544792, | |
| "loss": 0.5531, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.13305898491083676, | |
| "grad_norm": 1.2648167610168457, | |
| "learning_rate": 0.00017550274223034735, | |
| "loss": 0.0296, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.13443072702331962, | |
| "grad_norm": 8.14661979675293, | |
| "learning_rate": 0.0001773308957952468, | |
| "loss": 0.8536, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.13580246913580246, | |
| "grad_norm": 8.927884101867676, | |
| "learning_rate": 0.00017915904936014626, | |
| "loss": 0.4879, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.13717421124828533, | |
| "grad_norm": 9.555243492126465, | |
| "learning_rate": 0.00018098720292504568, | |
| "loss": 0.67, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13854595336076816, | |
| "grad_norm": 7.783656120300293, | |
| "learning_rate": 0.00018281535648994517, | |
| "loss": 0.4813, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.13991769547325103, | |
| "grad_norm": 0.5169872641563416, | |
| "learning_rate": 0.00018464351005484462, | |
| "loss": 0.0488, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1412894375857339, | |
| "grad_norm": 6.967692852020264, | |
| "learning_rate": 0.00018647166361974407, | |
| "loss": 0.5388, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.14266117969821673, | |
| "grad_norm": 6.324373245239258, | |
| "learning_rate": 0.0001882998171846435, | |
| "loss": 0.376, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1440329218106996, | |
| "grad_norm": 0.7642683982849121, | |
| "learning_rate": 0.00019012797074954296, | |
| "loss": 0.017, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.14540466392318244, | |
| "grad_norm": 8.600672721862793, | |
| "learning_rate": 0.00019195612431444244, | |
| "loss": 0.7542, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1467764060356653, | |
| "grad_norm": 7.111880302429199, | |
| "learning_rate": 0.00019378427787934186, | |
| "loss": 0.4063, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 6.075577735900879, | |
| "learning_rate": 0.00019561243144424132, | |
| "loss": 0.3658, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.149519890260631, | |
| "grad_norm": 6.12313175201416, | |
| "learning_rate": 0.00019744058500914077, | |
| "loss": 0.4389, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "grad_norm": 5.813235759735107, | |
| "learning_rate": 0.00019926873857404023, | |
| "loss": 0.3803, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_Qnli-dev_cosine_accuracy": 0.705078125, | |
| "eval_Qnli-dev_cosine_accuracy_threshold": 0.6866907477378845, | |
| "eval_Qnli-dev_cosine_ap": 0.7567018413685389, | |
| "eval_Qnli-dev_cosine_f1": 0.6931818181818182, | |
| "eval_Qnli-dev_cosine_f1_threshold": 0.6343963146209717, | |
| "eval_Qnli-dev_cosine_precision": 0.6267123287671232, | |
| "eval_Qnli-dev_cosine_recall": 0.7754237288135594, | |
| "eval_allNLI-dev_cosine_accuracy": 0.76953125, | |
| "eval_allNLI-dev_cosine_accuracy_threshold": 0.7752166986465454, | |
| "eval_allNLI-dev_cosine_ap": 0.6627175481841632, | |
| "eval_allNLI-dev_cosine_f1": 0.6624737945492662, | |
| "eval_allNLI-dev_cosine_f1_threshold": 0.6564935445785522, | |
| "eval_allNLI-dev_cosine_precision": 0.5197368421052632, | |
| "eval_allNLI-dev_cosine_recall": 0.9132947976878613, | |
| "eval_sequential_score": 0.7567018413685389, | |
| "eval_sts-test_pearson_cosine": 0.9026620207137961, | |
| "eval_sts-test_spearman_cosine": 0.913678627606199, | |
| "eval_vitaminc-pairs_loss": 2.009296178817749, | |
| "eval_vitaminc-pairs_runtime": 14.3224, | |
| "eval_vitaminc-pairs_samples_per_second": 8.937, | |
| "eval_vitaminc-pairs_steps_per_second": 0.07, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_negation-triplets_loss": 1.59572434425354, | |
| "eval_negation-triplets_runtime": 1.1528, | |
| "eval_negation-triplets_samples_per_second": 111.029, | |
| "eval_negation-triplets_steps_per_second": 0.867, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_scitail-pairs-pos_loss": 0.061776161193847656, | |
| "eval_scitail-pairs-pos_runtime": 1.5728, | |
| "eval_scitail-pairs-pos_samples_per_second": 81.383, | |
| "eval_scitail-pairs-pos_steps_per_second": 0.636, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_scitail-pairs-qa_loss": 0.009187542833387852, | |
| "eval_scitail-pairs-qa_runtime": 1.2102, | |
| "eval_scitail-pairs-qa_samples_per_second": 105.771, | |
| "eval_scitail-pairs-qa_steps_per_second": 0.826, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_xsum-pairs_loss": 0.37210211157798767, | |
| "eval_xsum-pairs_runtime": 6.2854, | |
| "eval_xsum-pairs_samples_per_second": 20.365, | |
| "eval_xsum-pairs_steps_per_second": 0.159, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_sciq_pairs_loss": 0.04122849553823471, | |
| "eval_sciq_pairs_runtime": 8.8116, | |
| "eval_sciq_pairs_samples_per_second": 14.526, | |
| "eval_sciq_pairs_steps_per_second": 0.113, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_qasc_pairs_loss": 0.4748501479625702, | |
| "eval_qasc_pairs_runtime": 1.3827, | |
| "eval_qasc_pairs_samples_per_second": 92.573, | |
| "eval_qasc_pairs_steps_per_second": 0.723, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_openbookqa_pairs_loss": 1.1540580987930298, | |
| "eval_openbookqa_pairs_runtime": 1.1788, | |
| "eval_openbookqa_pairs_samples_per_second": 108.581, | |
| "eval_openbookqa_pairs_steps_per_second": 0.848, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_nq_pairs_loss": 0.2363465428352356, | |
| "eval_nq_pairs_runtime": 7.8515, | |
| "eval_nq_pairs_samples_per_second": 16.303, | |
| "eval_nq_pairs_steps_per_second": 0.127, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_trivia_pairs_loss": 0.6520176529884338, | |
| "eval_trivia_pairs_runtime": 8.9067, | |
| "eval_trivia_pairs_samples_per_second": 14.371, | |
| "eval_trivia_pairs_steps_per_second": 0.112, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_gooaq_pairs_loss": 0.22620199620723724, | |
| "eval_gooaq_pairs_runtime": 2.067, | |
| "eval_gooaq_pairs_samples_per_second": 61.924, | |
| "eval_gooaq_pairs_steps_per_second": 0.484, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_paws-pos_loss": 0.02822125516831875, | |
| "eval_paws-pos_runtime": 1.5117, | |
| "eval_paws-pos_samples_per_second": 84.672, | |
| "eval_paws-pos_steps_per_second": 0.662, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15089163237311384, | |
| "eval_global_dataset_loss": 0.30668479204177856, | |
| "eval_global_dataset_runtime": 33.2591, | |
| "eval_global_dataset_samples_per_second": 11.546, | |
| "eval_global_dataset_steps_per_second": 0.06, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1522633744855967, | |
| "grad_norm": 4.30504846572876, | |
| "learning_rate": 0.00020109689213893968, | |
| "loss": 0.2478, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.15363511659807957, | |
| "grad_norm": 6.559568881988525, | |
| "learning_rate": 0.00020292504570383914, | |
| "loss": 0.8402, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1550068587105624, | |
| "grad_norm": 5.812280654907227, | |
| "learning_rate": 0.0002047531992687386, | |
| "loss": 0.6608, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.15637860082304528, | |
| "grad_norm": 2.0805885791778564, | |
| "learning_rate": 0.00020658135283363802, | |
| "loss": 0.0934, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.15775034293552812, | |
| "grad_norm": 5.199294090270996, | |
| "learning_rate": 0.00020840950639853747, | |
| "loss": 0.3907, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.15912208504801098, | |
| "grad_norm": 6.3685078620910645, | |
| "learning_rate": 0.00021023765996343693, | |
| "loss": 0.449, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.16049382716049382, | |
| "grad_norm": 6.4199652671813965, | |
| "learning_rate": 0.00021206581352833638, | |
| "loss": 0.4041, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.16186556927297668, | |
| "grad_norm": 6.015898704528809, | |
| "learning_rate": 0.00021389396709323584, | |
| "loss": 0.6749, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.16323731138545952, | |
| "grad_norm": 7.721911430358887, | |
| "learning_rate": 0.0002157221206581353, | |
| "loss": 0.4847, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.1646090534979424, | |
| "grad_norm": 1.8774610757827759, | |
| "learning_rate": 0.00021755027422303474, | |
| "loss": 0.0526, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16598079561042525, | |
| "grad_norm": 8.094359397888184, | |
| "learning_rate": 0.00021937842778793417, | |
| "loss": 0.6795, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1673525377229081, | |
| "grad_norm": 0.33090323209762573, | |
| "learning_rate": 0.00022120658135283365, | |
| "loss": 0.0064, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.16872427983539096, | |
| "grad_norm": 7.3609418869018555, | |
| "learning_rate": 0.0002230347349177331, | |
| "loss": 0.5918, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1700960219478738, | |
| "grad_norm": 6.189216613769531, | |
| "learning_rate": 0.00022486288848263253, | |
| "loss": 0.3544, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.17146776406035666, | |
| "grad_norm": 5.588890075683594, | |
| "learning_rate": 0.000226691042047532, | |
| "loss": 0.3849, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1728395061728395, | |
| "grad_norm": 3.4582345485687256, | |
| "learning_rate": 0.00022851919561243144, | |
| "loss": 0.2051, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.17421124828532236, | |
| "grad_norm": 4.075862407684326, | |
| "learning_rate": 0.00023034734917733092, | |
| "loss": 0.2129, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1755829903978052, | |
| "grad_norm": 15.110091209411621, | |
| "learning_rate": 0.00023217550274223035, | |
| "loss": 2.7937, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.17695473251028807, | |
| "grad_norm": 0.35791516304016113, | |
| "learning_rate": 0.0002340036563071298, | |
| "loss": 0.0166, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.17832647462277093, | |
| "grad_norm": 7.5200090408325195, | |
| "learning_rate": 0.00023583180987202926, | |
| "loss": 0.7856, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.17969821673525377, | |
| "grad_norm": 6.566864490509033, | |
| "learning_rate": 0.0002376599634369287, | |
| "loss": 0.8368, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.18106995884773663, | |
| "grad_norm": 4.958701133728027, | |
| "learning_rate": 0.00023948811700182814, | |
| "loss": 0.3813, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.18244170096021947, | |
| "grad_norm": 5.745133876800537, | |
| "learning_rate": 0.00024131627056672762, | |
| "loss": 0.5695, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.18381344307270234, | |
| "grad_norm": 4.952736854553223, | |
| "learning_rate": 0.00024314442413162708, | |
| "loss": 0.351, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 5.733601093292236, | |
| "learning_rate": 0.0002449725776965265, | |
| "loss": 0.3821, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.18655692729766804, | |
| "grad_norm": 5.019097328186035, | |
| "learning_rate": 0.00024680073126142596, | |
| "loss": 0.3249, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.18792866941015088, | |
| "grad_norm": 5.300777912139893, | |
| "learning_rate": 0.0002486288848263254, | |
| "loss": 0.3404, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.18930041152263374, | |
| "grad_norm": 4.518141269683838, | |
| "learning_rate": 0.00025045703839122487, | |
| "loss": 0.4535, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1906721536351166, | |
| "grad_norm": 1.0158088207244873, | |
| "learning_rate": 0.0002522851919561243, | |
| "loss": 0.0577, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.19204389574759945, | |
| "grad_norm": 5.966796398162842, | |
| "learning_rate": 0.0002541133455210238, | |
| "loss": 0.7431, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1934156378600823, | |
| "grad_norm": 6.123642921447754, | |
| "learning_rate": 0.00025594149908592323, | |
| "loss": 0.6778, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.19478737997256515, | |
| "grad_norm": 5.842874050140381, | |
| "learning_rate": 0.0002577696526508227, | |
| "loss": 0.5436, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.19615912208504802, | |
| "grad_norm": 4.759068012237549, | |
| "learning_rate": 0.0002595978062157221, | |
| "loss": 0.3582, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 4.080338478088379, | |
| "learning_rate": 0.00026142595978062154, | |
| "loss": 0.316, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.19890260631001372, | |
| "grad_norm": 4.1391448974609375, | |
| "learning_rate": 0.000263254113345521, | |
| "loss": 0.4446, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.20027434842249658, | |
| "grad_norm": 5.856256008148193, | |
| "learning_rate": 0.0002650822669104205, | |
| "loss": 0.7792, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.20164609053497942, | |
| "grad_norm": 7.747331142425537, | |
| "learning_rate": 0.00026691042047531996, | |
| "loss": 1.1147, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2030178326474623, | |
| "grad_norm": 6.825289249420166, | |
| "learning_rate": 0.0002687385740402194, | |
| "loss": 0.8267, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.20438957475994513, | |
| "grad_norm": 7.336719512939453, | |
| "learning_rate": 0.00027056672760511887, | |
| "loss": 0.8149, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.205761316872428, | |
| "grad_norm": 6.731626510620117, | |
| "learning_rate": 0.00027239488117001827, | |
| "loss": 0.942, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.20713305898491083, | |
| "grad_norm": 10.727692604064941, | |
| "learning_rate": 0.0002742230347349177, | |
| "loss": 2.4865, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2085048010973937, | |
| "grad_norm": 8.583380699157715, | |
| "learning_rate": 0.0002760511882998172, | |
| "loss": 1.0715, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.20987654320987653, | |
| "grad_norm": 6.236877918243408, | |
| "learning_rate": 0.00027787934186471663, | |
| "loss": 0.6219, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2112482853223594, | |
| "grad_norm": 6.254538536071777, | |
| "learning_rate": 0.0002797074954296161, | |
| "loss": 0.8705, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.21262002743484226, | |
| "grad_norm": 3.0917959213256836, | |
| "learning_rate": 0.00028153564899451554, | |
| "loss": 0.2407, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2139917695473251, | |
| "grad_norm": 4.438024997711182, | |
| "learning_rate": 0.000283363802559415, | |
| "loss": 0.4925, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.21536351165980797, | |
| "grad_norm": 0.43344631791114807, | |
| "learning_rate": 0.00028519195612431445, | |
| "loss": 0.0316, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2167352537722908, | |
| "grad_norm": 5.73934268951416, | |
| "learning_rate": 0.0002870201096892139, | |
| "loss": 0.3935, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.21810699588477367, | |
| "grad_norm": 4.532804012298584, | |
| "learning_rate": 0.00028884826325411336, | |
| "loss": 0.2083, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2194787379972565, | |
| "grad_norm": 4.846848487854004, | |
| "learning_rate": 0.0002906764168190128, | |
| "loss": 0.2798, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.22085048010973937, | |
| "grad_norm": 7.060863018035889, | |
| "learning_rate": 0.00029250457038391227, | |
| "loss": 0.8777, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.012754157185554504, | |
| "learning_rate": 0.0002943327239488117, | |
| "loss": 0.0002, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.22359396433470508, | |
| "grad_norm": 4.094379901885986, | |
| "learning_rate": 0.0002961608775137112, | |
| "loss": 0.2736, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.22496570644718794, | |
| "grad_norm": 10.741785049438477, | |
| "learning_rate": 0.0002979890310786106, | |
| "loss": 2.4185, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.22633744855967078, | |
| "grad_norm": 4.820891380310059, | |
| "learning_rate": 0.00029981718464351003, | |
| "loss": 0.7767, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.22770919067215364, | |
| "grad_norm": 6.423076152801514, | |
| "learning_rate": 0.0003016453382084095, | |
| "loss": 0.7971, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.22908093278463648, | |
| "grad_norm": 4.492727756500244, | |
| "learning_rate": 0.00030347349177330894, | |
| "loss": 0.4535, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.23045267489711935, | |
| "grad_norm": 5.301379680633545, | |
| "learning_rate": 0.00030530164533820845, | |
| "loss": 0.6654, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.23182441700960219, | |
| "grad_norm": 5.155853748321533, | |
| "learning_rate": 0.0003071297989031079, | |
| "loss": 0.3985, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.23319615912208505, | |
| "grad_norm": 0.4378865361213684, | |
| "learning_rate": 0.00030895795246800735, | |
| "loss": 0.0338, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2345679012345679, | |
| "grad_norm": 4.022473335266113, | |
| "learning_rate": 0.00031078610603290675, | |
| "loss": 0.1834, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.23593964334705075, | |
| "grad_norm": 7.863429069519043, | |
| "learning_rate": 0.0003126142595978062, | |
| "loss": 0.603, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.23731138545953362, | |
| "grad_norm": 8.951998710632324, | |
| "learning_rate": 0.00031444241316270566, | |
| "loss": 0.7871, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.23868312757201646, | |
| "grad_norm": 6.265102386474609, | |
| "learning_rate": 0.0003162705667276051, | |
| "loss": 0.4304, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.24005486968449932, | |
| "grad_norm": 6.6486005783081055, | |
| "learning_rate": 0.00031809872029250457, | |
| "loss": 0.649, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.24142661179698216, | |
| "grad_norm": 0.47100114822387695, | |
| "learning_rate": 0.000319926873857404, | |
| "loss": 0.048, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.24279835390946503, | |
| "grad_norm": 4.884115695953369, | |
| "learning_rate": 0.0003217550274223035, | |
| "loss": 0.4079, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.24417009602194786, | |
| "grad_norm": 4.508667469024658, | |
| "learning_rate": 0.0003235831809872029, | |
| "loss": 0.4627, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.24554183813443073, | |
| "grad_norm": 3.22367262840271, | |
| "learning_rate": 0.0003254113345521024, | |
| "loss": 0.3703, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 7.695303916931152, | |
| "learning_rate": 0.00032723948811700184, | |
| "loss": 0.8343, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24828532235939643, | |
| "grad_norm": 7.249318599700928, | |
| "learning_rate": 0.0003290676416819013, | |
| "loss": 0.692, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.2496570644718793, | |
| "grad_norm": 11.686202049255371, | |
| "learning_rate": 0.00033089579524680075, | |
| "loss": 2.7071, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.25102880658436216, | |
| "grad_norm": 6.061092376708984, | |
| "learning_rate": 0.0003327239488117002, | |
| "loss": 0.8451, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.252400548696845, | |
| "grad_norm": 5.932607650756836, | |
| "learning_rate": 0.00033455210237659966, | |
| "loss": 0.635, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.25377229080932784, | |
| "grad_norm": 3.491114616394043, | |
| "learning_rate": 0.00033638025594149906, | |
| "loss": 0.312, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2551440329218107, | |
| "grad_norm": 6.4914164543151855, | |
| "learning_rate": 0.0003382084095063985, | |
| "loss": 0.6996, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.25651577503429357, | |
| "grad_norm": 6.15857458114624, | |
| "learning_rate": 0.00034003656307129797, | |
| "loss": 0.4432, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2578875171467764, | |
| "grad_norm": 4.767185211181641, | |
| "learning_rate": 0.0003418647166361974, | |
| "loss": 0.375, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.25925925925925924, | |
| "grad_norm": 7.944342613220215, | |
| "learning_rate": 0.00034369287020109693, | |
| "loss": 0.9366, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2606310013717421, | |
| "grad_norm": 6.573953628540039, | |
| "learning_rate": 0.0003455210237659964, | |
| "loss": 0.755, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.262002743484225, | |
| "grad_norm": 4.173367023468018, | |
| "learning_rate": 0.00034734917733089584, | |
| "loss": 0.6068, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.26337448559670784, | |
| "grad_norm": 5.26171875, | |
| "learning_rate": 0.00034917733089579524, | |
| "loss": 0.5336, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.26474622770919065, | |
| "grad_norm": 6.669304370880127, | |
| "learning_rate": 0.0003510054844606947, | |
| "loss": 0.8783, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2661179698216735, | |
| "grad_norm": 4.4192938804626465, | |
| "learning_rate": 0.00035283363802559415, | |
| "loss": 0.3576, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2674897119341564, | |
| "grad_norm": 10.117819786071777, | |
| "learning_rate": 0.0003546617915904936, | |
| "loss": 2.1854, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.26886145404663925, | |
| "grad_norm": 5.256247520446777, | |
| "learning_rate": 0.00035648994515539306, | |
| "loss": 0.7835, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.27023319615912206, | |
| "grad_norm": 5.784887313842773, | |
| "learning_rate": 0.0003583180987202925, | |
| "loss": 0.5668, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.2716049382716049, | |
| "grad_norm": 4.977567672729492, | |
| "learning_rate": 0.00036014625228519197, | |
| "loss": 0.7033, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2729766803840878, | |
| "grad_norm": 0.011424711905419827, | |
| "learning_rate": 0.00036197440585009137, | |
| "loss": 0.0002, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.27434842249657065, | |
| "grad_norm": 5.805008411407471, | |
| "learning_rate": 0.0003638025594149909, | |
| "loss": 0.5791, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2757201646090535, | |
| "grad_norm": 3.8826043605804443, | |
| "learning_rate": 0.00036563071297989033, | |
| "loss": 0.2697, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.27709190672153633, | |
| "grad_norm": 6.563521385192871, | |
| "learning_rate": 0.0003674588665447898, | |
| "loss": 0.6261, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2784636488340192, | |
| "grad_norm": 4.584529399871826, | |
| "learning_rate": 0.00036928702010968924, | |
| "loss": 0.3253, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.27983539094650206, | |
| "grad_norm": 6.636009216308594, | |
| "learning_rate": 0.0003711151736745887, | |
| "loss": 0.8323, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2812071330589849, | |
| "grad_norm": 5.0911359786987305, | |
| "learning_rate": 0.00037294332723948815, | |
| "loss": 0.4472, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2825788751714678, | |
| "grad_norm": 3.9219255447387695, | |
| "learning_rate": 0.00037477148080438755, | |
| "loss": 0.3342, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.2839506172839506, | |
| "grad_norm": 5.114777565002441, | |
| "learning_rate": 0.000376599634369287, | |
| "loss": 0.6313, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.28532235939643347, | |
| "grad_norm": 0.3298715353012085, | |
| "learning_rate": 0.00037842778793418646, | |
| "loss": 0.059, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.28669410150891633, | |
| "grad_norm": 1.5965046882629395, | |
| "learning_rate": 0.0003802559414990859, | |
| "loss": 0.1195, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2880658436213992, | |
| "grad_norm": 0.39121323823928833, | |
| "learning_rate": 0.00038208409506398537, | |
| "loss": 0.0296, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.289437585733882, | |
| "grad_norm": 4.317224025726318, | |
| "learning_rate": 0.0003839122486288849, | |
| "loss": 0.5316, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.2908093278463649, | |
| "grad_norm": 4.000308036804199, | |
| "learning_rate": 0.00038574040219378433, | |
| "loss": 0.5201, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.29218106995884774, | |
| "grad_norm": 6.2192301750183105, | |
| "learning_rate": 0.00038756855575868373, | |
| "loss": 0.6602, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.2935528120713306, | |
| "grad_norm": 6.702320098876953, | |
| "learning_rate": 0.0003893967093235832, | |
| "loss": 0.9578, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.29492455418381347, | |
| "grad_norm": 3.9136242866516113, | |
| "learning_rate": 0.00039122486288848264, | |
| "loss": 0.2089, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 6.901303768157959, | |
| "learning_rate": 0.0003930530164533821, | |
| "loss": 1.2112, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.29766803840877915, | |
| "grad_norm": 4.04884672164917, | |
| "learning_rate": 0.00039488117001828155, | |
| "loss": 0.3294, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.299039780521262, | |
| "grad_norm": 5.46201753616333, | |
| "learning_rate": 0.000396709323583181, | |
| "loss": 0.867, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3004115226337449, | |
| "grad_norm": 5.559458255767822, | |
| "learning_rate": 0.00039853747714808046, | |
| "loss": 1.1745, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "grad_norm": 4.930731296539307, | |
| "learning_rate": 0.00040036563071297986, | |
| "loss": 0.7287, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_Qnli-dev_cosine_accuracy": 0.736328125, | |
| "eval_Qnli-dev_cosine_accuracy_threshold": 0.8059661388397217, | |
| "eval_Qnli-dev_cosine_ap": 0.7693397822540732, | |
| "eval_Qnli-dev_cosine_f1": 0.720136518771331, | |
| "eval_Qnli-dev_cosine_f1_threshold": 0.7346209287643433, | |
| "eval_Qnli-dev_cosine_precision": 0.6028571428571429, | |
| "eval_Qnli-dev_cosine_recall": 0.8940677966101694, | |
| "eval_allNLI-dev_cosine_accuracy": 0.751953125, | |
| "eval_allNLI-dev_cosine_accuracy_threshold": 0.8584632873535156, | |
| "eval_allNLI-dev_cosine_ap": 0.6542482370347211, | |
| "eval_allNLI-dev_cosine_f1": 0.6681127982646421, | |
| "eval_allNLI-dev_cosine_f1_threshold": 0.7551147937774658, | |
| "eval_allNLI-dev_cosine_precision": 0.5347222222222222, | |
| "eval_allNLI-dev_cosine_recall": 0.8901734104046243, | |
| "eval_sequential_score": 0.7693397822540732, | |
| "eval_sts-test_pearson_cosine": 0.8635423204589071, | |
| "eval_sts-test_spearman_cosine": 0.8907274956890058, | |
| "eval_vitaminc-pairs_loss": 2.5131773948669434, | |
| "eval_vitaminc-pairs_runtime": 14.3452, | |
| "eval_vitaminc-pairs_samples_per_second": 8.923, | |
| "eval_vitaminc-pairs_steps_per_second": 0.07, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_negation-triplets_loss": 1.5992437601089478, | |
| "eval_negation-triplets_runtime": 1.1695, | |
| "eval_negation-triplets_samples_per_second": 109.451, | |
| "eval_negation-triplets_steps_per_second": 0.855, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_scitail-pairs-pos_loss": 0.12100159376859665, | |
| "eval_scitail-pairs-pos_runtime": 1.5843, | |
| "eval_scitail-pairs-pos_samples_per_second": 80.791, | |
| "eval_scitail-pairs-pos_steps_per_second": 0.631, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_scitail-pairs-qa_loss": 0.029047677293419838, | |
| "eval_scitail-pairs-qa_runtime": 1.2089, | |
| "eval_scitail-pairs-qa_samples_per_second": 105.881, | |
| "eval_scitail-pairs-qa_steps_per_second": 0.827, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_xsum-pairs_loss": 0.6063941717147827, | |
| "eval_xsum-pairs_runtime": 6.3291, | |
| "eval_xsum-pairs_samples_per_second": 20.224, | |
| "eval_xsum-pairs_steps_per_second": 0.158, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_sciq_pairs_loss": 0.05205194652080536, | |
| "eval_sciq_pairs_runtime": 8.8533, | |
| "eval_sciq_pairs_samples_per_second": 14.458, | |
| "eval_sciq_pairs_steps_per_second": 0.113, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_qasc_pairs_loss": 0.7244825959205627, | |
| "eval_qasc_pairs_runtime": 1.3845, | |
| "eval_qasc_pairs_samples_per_second": 92.449, | |
| "eval_qasc_pairs_steps_per_second": 0.722, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_openbookqa_pairs_loss": 1.4260488748550415, | |
| "eval_openbookqa_pairs_runtime": 1.1784, | |
| "eval_openbookqa_pairs_samples_per_second": 108.623, | |
| "eval_openbookqa_pairs_steps_per_second": 0.849, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_nq_pairs_loss": 0.7104523777961731, | |
| "eval_nq_pairs_runtime": 7.8911, | |
| "eval_nq_pairs_samples_per_second": 16.221, | |
| "eval_nq_pairs_steps_per_second": 0.127, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_trivia_pairs_loss": 0.8537120223045349, | |
| "eval_trivia_pairs_runtime": 8.9305, | |
| "eval_trivia_pairs_samples_per_second": 14.333, | |
| "eval_trivia_pairs_steps_per_second": 0.112, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_gooaq_pairs_loss": 0.5029886960983276, | |
| "eval_gooaq_pairs_runtime": 2.0675, | |
| "eval_gooaq_pairs_samples_per_second": 61.91, | |
| "eval_gooaq_pairs_steps_per_second": 0.484, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_paws-pos_loss": 0.03150199353694916, | |
| "eval_paws-pos_runtime": 1.5078, | |
| "eval_paws-pos_samples_per_second": 84.892, | |
| "eval_paws-pos_steps_per_second": 0.663, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3017832647462277, | |
| "eval_global_dataset_loss": 0.18923546373844147, | |
| "eval_global_dataset_runtime": 33.3421, | |
| "eval_global_dataset_samples_per_second": 11.517, | |
| "eval_global_dataset_steps_per_second": 0.06, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.30315500685871055, | |
| "grad_norm": 3.582907199859619, | |
| "learning_rate": 0.00040219378427787936, | |
| "loss": 0.5484, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.3045267489711934, | |
| "grad_norm": 4.960206031799316, | |
| "learning_rate": 0.0004040219378427788, | |
| "loss": 0.9396, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3058984910836763, | |
| "grad_norm": 4.219746112823486, | |
| "learning_rate": 0.0004058500914076783, | |
| "loss": 0.4335, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.30727023319615915, | |
| "grad_norm": 6.449894428253174, | |
| "learning_rate": 0.00040767824497257773, | |
| "loss": 0.9026, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.30864197530864196, | |
| "grad_norm": 6.239223003387451, | |
| "learning_rate": 0.0004095063985374772, | |
| "loss": 0.7214, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3100137174211248, | |
| "grad_norm": 4.418921947479248, | |
| "learning_rate": 0.00041133455210237664, | |
| "loss": 0.4794, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.3113854595336077, | |
| "grad_norm": 0.012527555227279663, | |
| "learning_rate": 0.00041316270566727604, | |
| "loss": 0.0003, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.31275720164609055, | |
| "grad_norm": 2.672603130340576, | |
| "learning_rate": 0.0004149908592321755, | |
| "loss": 0.3003, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.31412894375857336, | |
| "grad_norm": 4.433743476867676, | |
| "learning_rate": 0.00041681901279707495, | |
| "loss": 0.4667, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.31550068587105623, | |
| "grad_norm": 4.458980083465576, | |
| "learning_rate": 0.0004186471663619744, | |
| "loss": 0.5006, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3168724279835391, | |
| "grad_norm": 5.0898237228393555, | |
| "learning_rate": 0.00042047531992687385, | |
| "loss": 0.5555, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.31824417009602196, | |
| "grad_norm": 4.338139533996582, | |
| "learning_rate": 0.00042230347349177336, | |
| "loss": 0.4437, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3196159122085048, | |
| "grad_norm": 5.023694038391113, | |
| "learning_rate": 0.00042413162705667276, | |
| "loss": 0.8813, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.32098765432098764, | |
| "grad_norm": 6.410233497619629, | |
| "learning_rate": 0.0004259597806215722, | |
| "loss": 1.1836, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.3223593964334705, | |
| "grad_norm": 3.8459813594818115, | |
| "learning_rate": 0.00042778793418647167, | |
| "loss": 0.3176, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.32373113854595337, | |
| "grad_norm": 5.539570331573486, | |
| "learning_rate": 0.0004296160877513711, | |
| "loss": 0.6248, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.32510288065843623, | |
| "grad_norm": 3.5191774368286133, | |
| "learning_rate": 0.0004314442413162706, | |
| "loss": 0.3623, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.32647462277091904, | |
| "grad_norm": 3.2997043132781982, | |
| "learning_rate": 0.00043327239488117003, | |
| "loss": 0.3205, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.3278463648834019, | |
| "grad_norm": 3.9236536026000977, | |
| "learning_rate": 0.0004351005484460695, | |
| "loss": 0.439, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3292181069958848, | |
| "grad_norm": 3.9750499725341797, | |
| "learning_rate": 0.0004369287020109689, | |
| "loss": 0.653, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.33058984910836764, | |
| "grad_norm": 4.344120502471924, | |
| "learning_rate": 0.00043875685557586834, | |
| "loss": 0.5743, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.3319615912208505, | |
| "grad_norm": 3.2905893325805664, | |
| "learning_rate": 0.0004405850091407678, | |
| "loss": 0.5844, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 5.996461391448975, | |
| "learning_rate": 0.0004424131627056673, | |
| "loss": 0.9451, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3347050754458162, | |
| "grad_norm": 4.085718631744385, | |
| "learning_rate": 0.00044424131627056676, | |
| "loss": 0.7071, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.33607681755829905, | |
| "grad_norm": 0.26297450065612793, | |
| "learning_rate": 0.0004460694698354662, | |
| "loss": 0.0226, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3374485596707819, | |
| "grad_norm": 5.191401481628418, | |
| "learning_rate": 0.00044789762340036567, | |
| "loss": 1.0585, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3388203017832647, | |
| "grad_norm": 5.426116466522217, | |
| "learning_rate": 0.00044972577696526507, | |
| "loss": 1.0764, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3401920438957476, | |
| "grad_norm": 0.17406082153320312, | |
| "learning_rate": 0.0004515539305301645, | |
| "loss": 0.0289, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.34156378600823045, | |
| "grad_norm": 4.900349140167236, | |
| "learning_rate": 0.000453382084095064, | |
| "loss": 0.5588, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.3429355281207133, | |
| "grad_norm": 5.373581409454346, | |
| "learning_rate": 0.00045521023765996343, | |
| "loss": 0.7509, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3443072702331962, | |
| "grad_norm": 5.4629106521606445, | |
| "learning_rate": 0.0004570383912248629, | |
| "loss": 0.8388, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.345679012345679, | |
| "grad_norm": 4.140360355377197, | |
| "learning_rate": 0.00045886654478976234, | |
| "loss": 0.5444, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.34705075445816186, | |
| "grad_norm": 5.176646709442139, | |
| "learning_rate": 0.00046069469835466185, | |
| "loss": 1.2432, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.3484224965706447, | |
| "grad_norm": 5.172772407531738, | |
| "learning_rate": 0.00046252285191956125, | |
| "loss": 0.8329, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3497942386831276, | |
| "grad_norm": 4.968120574951172, | |
| "learning_rate": 0.0004643510054844607, | |
| "loss": 0.7158, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3511659807956104, | |
| "grad_norm": 5.91867208480835, | |
| "learning_rate": 0.00046617915904936016, | |
| "loss": 0.935, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.35253772290809327, | |
| "grad_norm": 6.223313808441162, | |
| "learning_rate": 0.0004680073126142596, | |
| "loss": 0.8777, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.35390946502057613, | |
| "grad_norm": 11.474114418029785, | |
| "learning_rate": 0.00046983546617915907, | |
| "loss": 3.1178, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.355281207133059, | |
| "grad_norm": 6.462806701660156, | |
| "learning_rate": 0.0004716636197440585, | |
| "loss": 0.9418, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.35665294924554186, | |
| "grad_norm": 5.5286173820495605, | |
| "learning_rate": 0.000473491773308958, | |
| "loss": 0.6252, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.35802469135802467, | |
| "grad_norm": 8.872392654418945, | |
| "learning_rate": 0.0004753199268738574, | |
| "loss": 2.9074, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.35939643347050754, | |
| "grad_norm": 2.4954137802124023, | |
| "learning_rate": 0.00047714808043875683, | |
| "loss": 0.303, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3607681755829904, | |
| "grad_norm": 3.22896671295166, | |
| "learning_rate": 0.0004789762340036563, | |
| "loss": 0.7146, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.36213991769547327, | |
| "grad_norm": 4.280979156494141, | |
| "learning_rate": 0.0004808043875685558, | |
| "loss": 0.6848, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.3635116598079561, | |
| "grad_norm": 3.9235236644744873, | |
| "learning_rate": 0.00048263254113345525, | |
| "loss": 0.47, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.36488340192043894, | |
| "grad_norm": 3.460500717163086, | |
| "learning_rate": 0.0004844606946983547, | |
| "loss": 0.5762, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.3662551440329218, | |
| "grad_norm": 11.026795387268066, | |
| "learning_rate": 0.00048628884826325416, | |
| "loss": 3.3872, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.3676268861454047, | |
| "grad_norm": 3.2708449363708496, | |
| "learning_rate": 0.00048811700182815356, | |
| "loss": 0.4195, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.36899862825788754, | |
| "grad_norm": 5.177340030670166, | |
| "learning_rate": 0.000489945155393053, | |
| "loss": 1.2292, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 5.413723945617676, | |
| "learning_rate": 0.0004917733089579525, | |
| "loss": 1.1249, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3717421124828532, | |
| "grad_norm": 4.922053337097168, | |
| "learning_rate": 0.0004936014625228519, | |
| "loss": 1.0863, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.3731138545953361, | |
| "grad_norm": 4.433996677398682, | |
| "learning_rate": 0.0004954296160877514, | |
| "loss": 0.9361, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.37448559670781895, | |
| "grad_norm": 5.205246448516846, | |
| "learning_rate": 0.0004972577696526508, | |
| "loss": 0.7965, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.37585733882030176, | |
| "grad_norm": 4.139344215393066, | |
| "learning_rate": 0.0004990859232175503, | |
| "loss": 0.7914, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3772290809327846, | |
| "grad_norm": 0.1397838592529297, | |
| "learning_rate": 0.0005009140767824497, | |
| "loss": 0.0027, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3786008230452675, | |
| "grad_norm": 3.3880808353424072, | |
| "learning_rate": 0.0005027422303473492, | |
| "loss": 0.6585, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.37997256515775035, | |
| "grad_norm": 4.524999141693115, | |
| "learning_rate": 0.0005045703839122486, | |
| "loss": 0.5388, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.3813443072702332, | |
| "grad_norm": 6.445588111877441, | |
| "learning_rate": 0.000506398537477148, | |
| "loss": 1.238, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.38271604938271603, | |
| "grad_norm": 2.8341169357299805, | |
| "learning_rate": 0.0005082266910420476, | |
| "loss": 0.3782, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.3840877914951989, | |
| "grad_norm": 5.0054240226745605, | |
| "learning_rate": 0.0005100548446069469, | |
| "loss": 1.1769, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.38545953360768176, | |
| "grad_norm": 3.6371870040893555, | |
| "learning_rate": 0.0005118829981718465, | |
| "loss": 0.4808, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3868312757201646, | |
| "grad_norm": 8.928922653198242, | |
| "learning_rate": 0.0005137111517367459, | |
| "loss": 3.2104, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.38820301783264743, | |
| "grad_norm": 2.7686235904693604, | |
| "learning_rate": 0.0005155393053016454, | |
| "loss": 0.3027, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.3895747599451303, | |
| "grad_norm": 1.498235821723938, | |
| "learning_rate": 0.0005173674588665448, | |
| "loss": 0.1422, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.39094650205761317, | |
| "grad_norm": 3.578543186187744, | |
| "learning_rate": 0.0005191956124314442, | |
| "loss": 0.6059, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.39231824417009603, | |
| "grad_norm": 2.900531053543091, | |
| "learning_rate": 0.0005210237659963437, | |
| "loss": 0.3491, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3936899862825789, | |
| "grad_norm": 5.693866729736328, | |
| "learning_rate": 0.0005228519195612431, | |
| "loss": 1.1603, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 3.7944750785827637, | |
| "learning_rate": 0.0005246800731261426, | |
| "loss": 0.5784, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.39643347050754457, | |
| "grad_norm": 4.433256149291992, | |
| "learning_rate": 0.000526508226691042, | |
| "loss": 0.6532, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.39780521262002744, | |
| "grad_norm": 3.986520290374756, | |
| "learning_rate": 0.0005283363802559416, | |
| "loss": 0.5613, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3991769547325103, | |
| "grad_norm": 4.399818420410156, | |
| "learning_rate": 0.000530164533820841, | |
| "loss": 0.8469, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.40054869684499317, | |
| "grad_norm": 3.0586366653442383, | |
| "learning_rate": 0.0005319926873857404, | |
| "loss": 0.4484, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.401920438957476, | |
| "grad_norm": 0.10376634448766708, | |
| "learning_rate": 0.0005338208409506399, | |
| "loss": 0.0034, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.40329218106995884, | |
| "grad_norm": 8.303990364074707, | |
| "learning_rate": 0.0005356489945155393, | |
| "loss": 3.0798, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.4046639231824417, | |
| "grad_norm": 4.507086277008057, | |
| "learning_rate": 0.0005374771480804388, | |
| "loss": 0.8632, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4060356652949246, | |
| "grad_norm": 5.233419895172119, | |
| "learning_rate": 0.0005393053016453382, | |
| "loss": 1.2788, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4074074074074074, | |
| "grad_norm": 4.281100273132324, | |
| "learning_rate": 0.0005411334552102377, | |
| "loss": 1.2439, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.40877914951989025, | |
| "grad_norm": 1.219199299812317, | |
| "learning_rate": 0.0005429616087751371, | |
| "loss": 0.1067, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4101508916323731, | |
| "grad_norm": 3.0503711700439453, | |
| "learning_rate": 0.0005447897623400365, | |
| "loss": 0.4197, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.411522633744856, | |
| "grad_norm": 3.5557351112365723, | |
| "learning_rate": 0.000546617915904936, | |
| "loss": 0.56, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.41289437585733885, | |
| "grad_norm": 4.1112470626831055, | |
| "learning_rate": 0.0005484460694698354, | |
| "loss": 0.7235, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.41426611796982166, | |
| "grad_norm": 2.915947675704956, | |
| "learning_rate": 0.000550274223034735, | |
| "loss": 0.3506, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4156378600823045, | |
| "grad_norm": 4.125770568847656, | |
| "learning_rate": 0.0005521023765996344, | |
| "loss": 0.6808, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.4170096021947874, | |
| "grad_norm": 5.084654808044434, | |
| "learning_rate": 0.0005539305301645339, | |
| "loss": 0.9406, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.41838134430727025, | |
| "grad_norm": 4.542891025543213, | |
| "learning_rate": 0.0005557586837294333, | |
| "loss": 0.6707, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.41975308641975306, | |
| "grad_norm": 4.285159587860107, | |
| "learning_rate": 0.0005575868372943327, | |
| "loss": 1.0718, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.42112482853223593, | |
| "grad_norm": 5.053350925445557, | |
| "learning_rate": 0.0005594149908592322, | |
| "loss": 0.9847, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4224965706447188, | |
| "grad_norm": 1.4923471212387085, | |
| "learning_rate": 0.0005612431444241316, | |
| "loss": 0.122, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.42386831275720166, | |
| "grad_norm": 4.332481384277344, | |
| "learning_rate": 0.0005630712979890311, | |
| "loss": 0.8221, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.4252400548696845, | |
| "grad_norm": 4.23899507522583, | |
| "learning_rate": 0.0005648994515539305, | |
| "loss": 0.9891, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.42661179698216734, | |
| "grad_norm": 4.370994567871094, | |
| "learning_rate": 0.00056672760511883, | |
| "loss": 0.926, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.4279835390946502, | |
| "grad_norm": 0.23886460065841675, | |
| "learning_rate": 0.0005685557586837294, | |
| "loss": 0.0151, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.42935528120713307, | |
| "grad_norm": 7.701839447021484, | |
| "learning_rate": 0.0005703839122486289, | |
| "loss": 2.8429, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.43072702331961593, | |
| "grad_norm": 4.738073825836182, | |
| "learning_rate": 0.0005722120658135283, | |
| "loss": 0.9917, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.43209876543209874, | |
| "grad_norm": 3.7907347679138184, | |
| "learning_rate": 0.0005740402193784278, | |
| "loss": 0.9199, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.4334705075445816, | |
| "grad_norm": 2.7892837524414062, | |
| "learning_rate": 0.0005758683729433273, | |
| "loss": 0.3931, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.4348422496570645, | |
| "grad_norm": 3.414641857147217, | |
| "learning_rate": 0.0005776965265082267, | |
| "loss": 0.7119, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.43621399176954734, | |
| "grad_norm": 0.023860761895775795, | |
| "learning_rate": 0.0005795246800731262, | |
| "loss": 0.0008, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.4375857338820302, | |
| "grad_norm": 3.0215470790863037, | |
| "learning_rate": 0.0005813528336380256, | |
| "loss": 0.4985, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.438957475994513, | |
| "grad_norm": 3.1884336471557617, | |
| "learning_rate": 0.000583180987202925, | |
| "loss": 0.7732, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4403292181069959, | |
| "grad_norm": 3.4848649501800537, | |
| "learning_rate": 0.0005850091407678245, | |
| "loss": 0.7515, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.44170096021947874, | |
| "grad_norm": 0.9857578873634338, | |
| "learning_rate": 0.0005868372943327239, | |
| "loss": 0.0885, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.4430727023319616, | |
| "grad_norm": 4.590999126434326, | |
| "learning_rate": 0.0005886654478976234, | |
| "loss": 1.1677, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 4.161310195922852, | |
| "learning_rate": 0.0005904936014625229, | |
| "loss": 0.7439, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.4458161865569273, | |
| "grad_norm": 4.842323303222656, | |
| "learning_rate": 0.0005923217550274223, | |
| "loss": 1.3473, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.44718792866941015, | |
| "grad_norm": 3.332562208175659, | |
| "learning_rate": 0.0005941499085923218, | |
| "loss": 0.7273, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.448559670781893, | |
| "grad_norm": 4.295160293579102, | |
| "learning_rate": 0.0005959780621572211, | |
| "loss": 0.843, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.4499314128943759, | |
| "grad_norm": 3.641636848449707, | |
| "learning_rate": 0.0005978062157221207, | |
| "loss": 0.7881, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.4513031550068587, | |
| "grad_norm": 2.767233371734619, | |
| "learning_rate": 0.0005996343692870201, | |
| "loss": 0.5319, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "grad_norm": 2.925886869430542, | |
| "learning_rate": 0.0006014625228519196, | |
| "loss": 0.4826, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_Qnli-dev_cosine_accuracy": 0.708984375, | |
| "eval_Qnli-dev_cosine_accuracy_threshold": 0.8169091939926147, | |
| "eval_Qnli-dev_cosine_ap": 0.7472481805376167, | |
| "eval_Qnli-dev_cosine_f1": 0.7189781021897811, | |
| "eval_Qnli-dev_cosine_f1_threshold": 0.7571755051612854, | |
| "eval_Qnli-dev_cosine_precision": 0.6314102564102564, | |
| "eval_Qnli-dev_cosine_recall": 0.8347457627118644, | |
| "eval_allNLI-dev_cosine_accuracy": 0.748046875, | |
| "eval_allNLI-dev_cosine_accuracy_threshold": 0.8621972799301147, | |
| "eval_allNLI-dev_cosine_ap": 0.6527104471447597, | |
| "eval_allNLI-dev_cosine_f1": 0.6606334841628959, | |
| "eval_allNLI-dev_cosine_f1_threshold": 0.7864561080932617, | |
| "eval_allNLI-dev_cosine_precision": 0.5427509293680297, | |
| "eval_allNLI-dev_cosine_recall": 0.8439306358381503, | |
| "eval_sequential_score": 0.7472481805376167, | |
| "eval_sts-test_pearson_cosine": 0.8465015878560311, | |
| "eval_sts-test_spearman_cosine": 0.8833058569973334, | |
| "eval_vitaminc-pairs_loss": 2.5387091636657715, | |
| "eval_vitaminc-pairs_runtime": 14.3065, | |
| "eval_vitaminc-pairs_samples_per_second": 8.947, | |
| "eval_vitaminc-pairs_steps_per_second": 0.07, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_negation-triplets_loss": 1.8608626127243042, | |
| "eval_negation-triplets_runtime": 1.1519, | |
| "eval_negation-triplets_samples_per_second": 111.119, | |
| "eval_negation-triplets_steps_per_second": 0.868, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_scitail-pairs-pos_loss": 0.07322188466787338, | |
| "eval_scitail-pairs-pos_runtime": 1.5539, | |
| "eval_scitail-pairs-pos_samples_per_second": 82.372, | |
| "eval_scitail-pairs-pos_steps_per_second": 0.644, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_scitail-pairs-qa_loss": 0.026681702584028244, | |
| "eval_scitail-pairs-qa_runtime": 1.2098, | |
| "eval_scitail-pairs-qa_samples_per_second": 105.799, | |
| "eval_scitail-pairs-qa_steps_per_second": 0.827, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_xsum-pairs_loss": 0.6542444825172424, | |
| "eval_xsum-pairs_runtime": 6.2906, | |
| "eval_xsum-pairs_samples_per_second": 20.348, | |
| "eval_xsum-pairs_steps_per_second": 0.159, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_sciq_pairs_loss": 0.06421030312776566, | |
| "eval_sciq_pairs_runtime": 8.8514, | |
| "eval_sciq_pairs_samples_per_second": 14.461, | |
| "eval_sciq_pairs_steps_per_second": 0.113, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_qasc_pairs_loss": 0.8813464641571045, | |
| "eval_qasc_pairs_runtime": 1.3875, | |
| "eval_qasc_pairs_samples_per_second": 92.249, | |
| "eval_qasc_pairs_steps_per_second": 0.721, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_openbookqa_pairs_loss": 1.4074363708496094, | |
| "eval_openbookqa_pairs_runtime": 1.1826, | |
| "eval_openbookqa_pairs_samples_per_second": 108.237, | |
| "eval_openbookqa_pairs_steps_per_second": 0.846, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_nq_pairs_loss": 0.62897789478302, | |
| "eval_nq_pairs_runtime": 7.8684, | |
| "eval_nq_pairs_samples_per_second": 16.268, | |
| "eval_nq_pairs_steps_per_second": 0.127, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_trivia_pairs_loss": 1.084182620048523, | |
| "eval_trivia_pairs_runtime": 8.9262, | |
| "eval_trivia_pairs_samples_per_second": 14.34, | |
| "eval_trivia_pairs_steps_per_second": 0.112, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_gooaq_pairs_loss": 0.6594768762588501, | |
| "eval_gooaq_pairs_runtime": 2.0651, | |
| "eval_gooaq_pairs_samples_per_second": 61.982, | |
| "eval_gooaq_pairs_steps_per_second": 0.484, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_paws-pos_loss": 0.03268549218773842, | |
| "eval_paws-pos_runtime": 1.5103, | |
| "eval_paws-pos_samples_per_second": 84.754, | |
| "eval_paws-pos_steps_per_second": 0.662, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "eval_global_dataset_loss": 0.21171291172504425, | |
| "eval_global_dataset_runtime": 33.2988, | |
| "eval_global_dataset_samples_per_second": 11.532, | |
| "eval_global_dataset_steps_per_second": 0.06, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4540466392318244, | |
| "grad_norm": 2.6243412494659424, | |
| "learning_rate": 0.000603290676416819, | |
| "loss": 0.6096, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.4554183813443073, | |
| "grad_norm": 2.776013135910034, | |
| "learning_rate": 0.0006051188299817185, | |
| "loss": 0.3687, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.4567901234567901, | |
| "grad_norm": 4.418542385101318, | |
| "learning_rate": 0.0006069469835466179, | |
| "loss": 0.9713, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.45816186556927296, | |
| "grad_norm": 5.112300872802734, | |
| "learning_rate": 0.0006087751371115173, | |
| "loss": 1.3203, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.45953360768175583, | |
| "grad_norm": 2.1659023761749268, | |
| "learning_rate": 0.0006106032906764169, | |
| "loss": 0.3443, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4609053497942387, | |
| "grad_norm": 4.783431529998779, | |
| "learning_rate": 0.0006124314442413162, | |
| "loss": 1.4592, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.46227709190672156, | |
| "grad_norm": 3.507357597351074, | |
| "learning_rate": 0.0006142595978062158, | |
| "loss": 0.8, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.46364883401920437, | |
| "grad_norm": 3.118370771408081, | |
| "learning_rate": 0.0006160877513711151, | |
| "loss": 0.5481, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.46502057613168724, | |
| "grad_norm": 4.21981143951416, | |
| "learning_rate": 0.0006179159049360147, | |
| "loss": 1.4286, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.4663923182441701, | |
| "grad_norm": 2.675670862197876, | |
| "learning_rate": 0.0006197440585009141, | |
| "loss": 0.4012, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.46776406035665297, | |
| "grad_norm": 1.6964771747589111, | |
| "learning_rate": 0.0006215722120658135, | |
| "loss": 0.1474, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.4691358024691358, | |
| "grad_norm": 4.83234167098999, | |
| "learning_rate": 0.000623400365630713, | |
| "loss": 1.2755, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.47050754458161864, | |
| "grad_norm": 2.7494046688079834, | |
| "learning_rate": 0.0006252285191956124, | |
| "loss": 0.4935, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.4718792866941015, | |
| "grad_norm": 4.708520412445068, | |
| "learning_rate": 0.0006270566727605119, | |
| "loss": 1.0101, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.4732510288065844, | |
| "grad_norm": 2.6878857612609863, | |
| "learning_rate": 0.0006288848263254113, | |
| "loss": 0.4529, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.47462277091906724, | |
| "grad_norm": 2.571988105773926, | |
| "learning_rate": 0.0006307129798903109, | |
| "loss": 0.3516, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.47599451303155005, | |
| "grad_norm": 3.210439443588257, | |
| "learning_rate": 0.0006325411334552102, | |
| "loss": 0.4045, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.4773662551440329, | |
| "grad_norm": 4.047224521636963, | |
| "learning_rate": 0.0006343692870201097, | |
| "loss": 1.2326, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.4787379972565158, | |
| "grad_norm": 4.121635437011719, | |
| "learning_rate": 0.0006361974405850091, | |
| "loss": 0.8951, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.48010973936899864, | |
| "grad_norm": 3.480602741241455, | |
| "learning_rate": 0.0006380255941499086, | |
| "loss": 0.6783, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.48148148148148145, | |
| "grad_norm": 4.286612033843994, | |
| "learning_rate": 0.000639853747714808, | |
| "loss": 1.1821, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.4828532235939643, | |
| "grad_norm": 3.077362537384033, | |
| "learning_rate": 0.0006416819012797075, | |
| "loss": 0.5101, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.4842249657064472, | |
| "grad_norm": 3.127272844314575, | |
| "learning_rate": 0.000643510054844607, | |
| "loss": 0.844, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.48559670781893005, | |
| "grad_norm": 2.6955454349517822, | |
| "learning_rate": 0.0006453382084095064, | |
| "loss": 0.5413, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.4869684499314129, | |
| "grad_norm": 3.285903215408325, | |
| "learning_rate": 0.0006471663619744058, | |
| "loss": 0.9982, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.4883401920438957, | |
| "grad_norm": 3.6762568950653076, | |
| "learning_rate": 0.0006489945155393053, | |
| "loss": 0.937, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.4897119341563786, | |
| "grad_norm": 3.365633964538574, | |
| "learning_rate": 0.0006508226691042048, | |
| "loss": 0.7977, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.49108367626886146, | |
| "grad_norm": 3.7572262287139893, | |
| "learning_rate": 0.0006526508226691042, | |
| "loss": 0.8697, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.4924554183813443, | |
| "grad_norm": 3.309539794921875, | |
| "learning_rate": 0.0006544789762340037, | |
| "loss": 1.1136, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 4.250339508056641, | |
| "learning_rate": 0.0006563071297989031, | |
| "loss": 1.3018, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.49519890260631, | |
| "grad_norm": 0.07554444670677185, | |
| "learning_rate": 0.0006581352833638026, | |
| "loss": 0.0023, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.49657064471879286, | |
| "grad_norm": 3.6443750858306885, | |
| "learning_rate": 0.0006599634369287019, | |
| "loss": 0.7019, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.49794238683127573, | |
| "grad_norm": 0.19442614912986755, | |
| "learning_rate": 0.0006617915904936015, | |
| "loss": 0.0481, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.4993141289437586, | |
| "grad_norm": 3.7148313522338867, | |
| "learning_rate": 0.0006636197440585009, | |
| "loss": 0.891, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5006858710562414, | |
| "grad_norm": 2.7239511013031006, | |
| "learning_rate": 0.0006654478976234004, | |
| "loss": 0.6353, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5020576131687243, | |
| "grad_norm": 2.5572762489318848, | |
| "learning_rate": 0.0006672760511882999, | |
| "loss": 0.5181, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5034293552812071, | |
| "grad_norm": 3.162834405899048, | |
| "learning_rate": 0.0006691042047531993, | |
| "loss": 0.8311, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.50480109739369, | |
| "grad_norm": 0.06661587208509445, | |
| "learning_rate": 0.0006709323583180988, | |
| "loss": 0.0026, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5061728395061729, | |
| "grad_norm": 2.4806196689605713, | |
| "learning_rate": 0.0006727605118829981, | |
| "loss": 0.5004, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5075445816186557, | |
| "grad_norm": 2.5774953365325928, | |
| "learning_rate": 0.0006745886654478977, | |
| "loss": 0.4511, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5089163237311386, | |
| "grad_norm": 2.2276206016540527, | |
| "learning_rate": 0.000676416819012797, | |
| "loss": 0.4558, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5102880658436214, | |
| "grad_norm": 2.652674674987793, | |
| "learning_rate": 0.0006782449725776966, | |
| "loss": 0.5073, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5116598079561042, | |
| "grad_norm": 2.2147669792175293, | |
| "learning_rate": 0.0006800731261425959, | |
| "loss": 0.3979, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.5130315500685871, | |
| "grad_norm": 1.5278068780899048, | |
| "learning_rate": 0.0006819012797074955, | |
| "loss": 0.1665, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.51440329218107, | |
| "grad_norm": 2.9153432846069336, | |
| "learning_rate": 0.0006837294332723948, | |
| "loss": 0.8231, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5157750342935528, | |
| "grad_norm": 4.252976894378662, | |
| "learning_rate": 0.0006855575868372943, | |
| "loss": 1.0406, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5171467764060357, | |
| "grad_norm": 3.768838405609131, | |
| "learning_rate": 0.0006873857404021939, | |
| "loss": 0.725, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 1.8634569644927979, | |
| "learning_rate": 0.0006892138939670932, | |
| "loss": 0.2603, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5198902606310014, | |
| "grad_norm": 5.095807075500488, | |
| "learning_rate": 0.0006910420475319928, | |
| "loss": 1.7357, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5212620027434842, | |
| "grad_norm": 2.8866710662841797, | |
| "learning_rate": 0.0006928702010968921, | |
| "loss": 0.9147, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.522633744855967, | |
| "grad_norm": 2.369819402694702, | |
| "learning_rate": 0.0006946983546617917, | |
| "loss": 0.4277, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.52400548696845, | |
| "grad_norm": 1.7865092754364014, | |
| "learning_rate": 0.000696526508226691, | |
| "loss": 0.4788, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5253772290809328, | |
| "grad_norm": 3.0729360580444336, | |
| "learning_rate": 0.0006983546617915905, | |
| "loss": 0.7666, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5267489711934157, | |
| "grad_norm": 3.023810386657715, | |
| "learning_rate": 0.0007001828153564899, | |
| "loss": 0.7728, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5281207133058985, | |
| "grad_norm": 2.5811986923217773, | |
| "learning_rate": 0.0007020109689213894, | |
| "loss": 0.4744, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5294924554183813, | |
| "grad_norm": 2.748720407485962, | |
| "learning_rate": 0.0007038391224862888, | |
| "loss": 1.1014, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5308641975308642, | |
| "grad_norm": 3.950869560241699, | |
| "learning_rate": 0.0007056672760511883, | |
| "loss": 1.5588, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.532235939643347, | |
| "grad_norm": 1.595503330230713, | |
| "learning_rate": 0.0007074954296160879, | |
| "loss": 0.3185, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.53360768175583, | |
| "grad_norm": 3.505636692047119, | |
| "learning_rate": 0.0007093235831809872, | |
| "loss": 1.3348, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5349794238683128, | |
| "grad_norm": 1.0456945896148682, | |
| "learning_rate": 0.0007111517367458867, | |
| "loss": 0.1656, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5363511659807956, | |
| "grad_norm": 2.747938871383667, | |
| "learning_rate": 0.0007129798903107861, | |
| "loss": 0.9375, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5377229080932785, | |
| "grad_norm": 4.741049289703369, | |
| "learning_rate": 0.0007148080438756856, | |
| "loss": 1.4665, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.5390946502057613, | |
| "grad_norm": 4.830301284790039, | |
| "learning_rate": 0.000716636197440585, | |
| "loss": 1.4635, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5404663923182441, | |
| "grad_norm": 2.938199758529663, | |
| "learning_rate": 0.0007184643510054845, | |
| "loss": 0.8677, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.541838134430727, | |
| "grad_norm": 6.291453838348389, | |
| "learning_rate": 0.0007202925045703839, | |
| "loss": 3.033, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5432098765432098, | |
| "grad_norm": 3.0533947944641113, | |
| "learning_rate": 0.0007221206581352834, | |
| "loss": 1.4375, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.5445816186556928, | |
| "grad_norm": 3.13800048828125, | |
| "learning_rate": 0.0007239488117001827, | |
| "loss": 0.9762, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5459533607681756, | |
| "grad_norm": 1.1184616088867188, | |
| "learning_rate": 0.0007257769652650823, | |
| "loss": 0.1333, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.5473251028806584, | |
| "grad_norm": 3.610217571258545, | |
| "learning_rate": 0.0007276051188299818, | |
| "loss": 1.1823, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.5486968449931413, | |
| "grad_norm": 3.6696395874023438, | |
| "learning_rate": 0.0007294332723948812, | |
| "loss": 1.2443, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5500685871056241, | |
| "grad_norm": 2.967648506164551, | |
| "learning_rate": 0.0007312614259597807, | |
| "loss": 0.662, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.551440329218107, | |
| "grad_norm": 0.8750410079956055, | |
| "learning_rate": 0.0007330895795246801, | |
| "loss": 0.0709, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.5528120713305898, | |
| "grad_norm": 3.0801315307617188, | |
| "learning_rate": 0.0007349177330895796, | |
| "loss": 0.5822, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.5541838134430727, | |
| "grad_norm": 3.701993227005005, | |
| "learning_rate": 0.0007367458866544789, | |
| "loss": 1.0826, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 3.439502716064453, | |
| "learning_rate": 0.0007385740402193785, | |
| "loss": 0.7953, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5569272976680384, | |
| "grad_norm": 4.8917341232299805, | |
| "learning_rate": 0.0007404021937842778, | |
| "loss": 1.6109, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.5582990397805213, | |
| "grad_norm": 3.793834924697876, | |
| "learning_rate": 0.0007422303473491774, | |
| "loss": 1.2505, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.5596707818930041, | |
| "grad_norm": 3.0604796409606934, | |
| "learning_rate": 0.0007440585009140767, | |
| "loss": 1.0019, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.5610425240054869, | |
| "grad_norm": 2.872400999069214, | |
| "learning_rate": 0.0007458866544789763, | |
| "loss": 0.8224, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.5624142661179699, | |
| "grad_norm": 2.631157398223877, | |
| "learning_rate": 0.0007477148080438758, | |
| "loss": 0.6592, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5637860082304527, | |
| "grad_norm": 2.843379020690918, | |
| "learning_rate": 0.0007495429616087751, | |
| "loss": 0.7099, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.5651577503429356, | |
| "grad_norm": 2.3366591930389404, | |
| "learning_rate": 0.0007513711151736747, | |
| "loss": 0.5484, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.5665294924554184, | |
| "grad_norm": 3.4202780723571777, | |
| "learning_rate": 0.000753199268738574, | |
| "loss": 1.3019, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.5679012345679012, | |
| "grad_norm": 0.9125491976737976, | |
| "learning_rate": 0.0007550274223034736, | |
| "loss": 0.1266, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.5692729766803841, | |
| "grad_norm": 2.8945682048797607, | |
| "learning_rate": 0.0007568555758683729, | |
| "loss": 0.8932, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5706447187928669, | |
| "grad_norm": 2.898399591445923, | |
| "learning_rate": 0.0007586837294332725, | |
| "loss": 0.7547, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.5720164609053497, | |
| "grad_norm": 2.9406332969665527, | |
| "learning_rate": 0.0007605118829981718, | |
| "loss": 0.7614, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.5733882030178327, | |
| "grad_norm": 2.1753182411193848, | |
| "learning_rate": 0.0007623400365630713, | |
| "loss": 0.7364, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.5747599451303155, | |
| "grad_norm": 2.5140652656555176, | |
| "learning_rate": 0.0007641681901279707, | |
| "loss": 0.6539, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.5761316872427984, | |
| "grad_norm": 2.0174620151519775, | |
| "learning_rate": 0.0007659963436928702, | |
| "loss": 0.3848, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5775034293552812, | |
| "grad_norm": 0.11105114966630936, | |
| "learning_rate": 0.0007678244972577697, | |
| "loss": 0.0394, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.578875171467764, | |
| "grad_norm": 1.8194284439086914, | |
| "learning_rate": 0.0007696526508226691, | |
| "loss": 0.4623, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.5802469135802469, | |
| "grad_norm": 0.7781994342803955, | |
| "learning_rate": 0.0007714808043875687, | |
| "loss": 0.0783, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.5816186556927297, | |
| "grad_norm": 3.7422642707824707, | |
| "learning_rate": 0.000773308957952468, | |
| "loss": 1.4366, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.5829903978052127, | |
| "grad_norm": 3.9761717319488525, | |
| "learning_rate": 0.0007751371115173675, | |
| "loss": 1.3568, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5843621399176955, | |
| "grad_norm": 2.948404550552368, | |
| "learning_rate": 0.0007769652650822669, | |
| "loss": 0.9065, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.5857338820301783, | |
| "grad_norm": 2.7700140476226807, | |
| "learning_rate": 0.0007787934186471664, | |
| "loss": 0.6723, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.5871056241426612, | |
| "grad_norm": 2.8618569374084473, | |
| "learning_rate": 0.0007806215722120658, | |
| "loss": 0.7596, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.588477366255144, | |
| "grad_norm": 0.956656813621521, | |
| "learning_rate": 0.0007824497257769653, | |
| "loss": 0.1426, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.5898491083676269, | |
| "grad_norm": 2.8430604934692383, | |
| "learning_rate": 0.0007842778793418648, | |
| "loss": 1.3264, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5912208504801097, | |
| "grad_norm": 0.11261007934808731, | |
| "learning_rate": 0.0007861060329067642, | |
| "loss": 0.0442, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 2.7201974391937256, | |
| "learning_rate": 0.0007879341864716636, | |
| "loss": 0.6046, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.5939643347050755, | |
| "grad_norm": 2.5923287868499756, | |
| "learning_rate": 0.0007897623400365631, | |
| "loss": 0.6173, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.5953360768175583, | |
| "grad_norm": 4.77182674407959, | |
| "learning_rate": 0.0007915904936014625, | |
| "loss": 2.8892, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.5967078189300411, | |
| "grad_norm": 3.1731245517730713, | |
| "learning_rate": 0.000793418647166362, | |
| "loss": 1.3149, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.598079561042524, | |
| "grad_norm": 2.849473237991333, | |
| "learning_rate": 0.0007952468007312615, | |
| "loss": 1.1886, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.5994513031550068, | |
| "grad_norm": 3.5986573696136475, | |
| "learning_rate": 0.0007970749542961609, | |
| "loss": 1.2799, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.6008230452674898, | |
| "grad_norm": 2.668875217437744, | |
| "learning_rate": 0.0007989031078610604, | |
| "loss": 0.7527, | |
| "step": 438 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 2187, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 219, | |
| "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": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |