| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0695884911341587, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0026764804282368685, | |
| "grad_norm": 416.7599792480469, | |
| "learning_rate": 0.0, | |
| "loss": 3.2698, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005352960856473737, | |
| "grad_norm": 316.08477783203125, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 2.9631, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008029441284710606, | |
| "grad_norm": 472.76885986328125, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 3.9863, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010705921712947474, | |
| "grad_norm": 251.41015625, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 3.2654, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013382402141184342, | |
| "grad_norm": 217.40838623046875, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 3.4505, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01605888256942121, | |
| "grad_norm": 136.2932586669922, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 3.2498, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01873536299765808, | |
| "grad_norm": 135.7803192138672, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 3.0983, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.021411843425894948, | |
| "grad_norm": 93.63102722167969, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 2.965, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.024088323854131816, | |
| "grad_norm": 82.7247085571289, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 2.6341, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.026764804282368684, | |
| "grad_norm": 94.11923217773438, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 3.2072, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.029441284710605555, | |
| "grad_norm": 65.76191711425781, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.6866, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03211776513884242, | |
| "grad_norm": 89.41738891601562, | |
| "learning_rate": 2.2e-06, | |
| "loss": 2.8421, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03479424556707929, | |
| "grad_norm": 82.1783676147461, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 2.5489, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03747072599531616, | |
| "grad_norm": 72.41712951660156, | |
| "learning_rate": 2.6e-06, | |
| "loss": 2.9365, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04014720642355303, | |
| "grad_norm": 77.541015625, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 2.0516, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042823686851789895, | |
| "grad_norm": 125.65862274169922, | |
| "learning_rate": 3e-06, | |
| "loss": 2.7676, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04550016728002677, | |
| "grad_norm": 94.85375213623047, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 2.5114, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04817664770826363, | |
| "grad_norm": 77.89604187011719, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 2.536, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0508531281365005, | |
| "grad_norm": 59.268680572509766, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 2.1877, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05352960856473737, | |
| "grad_norm": 57.32849884033203, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 2.2084, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05620608899297424, | |
| "grad_norm": 58.27803421020508, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.191, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05888256942121111, | |
| "grad_norm": 32.10304260253906, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 2.0009, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.061559049849447975, | |
| "grad_norm": 17.106046676635742, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.7496, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06423553027768485, | |
| "grad_norm": 16.371326446533203, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 2.0558, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06691201070592172, | |
| "grad_norm": 14.423291206359863, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 2.1511, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06958849113415858, | |
| "grad_norm": 16.198841094970703, | |
| "learning_rate": 5e-06, | |
| "loss": 1.9571, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07226497156239545, | |
| "grad_norm": 14.810019493103027, | |
| "learning_rate": 5.2e-06, | |
| "loss": 1.9937, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07494145199063232, | |
| "grad_norm": 13.464497566223145, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 1.7204, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07761793241886919, | |
| "grad_norm": 14.364913940429688, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.7, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08029441284710606, | |
| "grad_norm": 12.647878646850586, | |
| "learning_rate": 5.8e-06, | |
| "loss": 1.7034, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08297089327534292, | |
| "grad_norm": 12.305310249328613, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7873, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08564737370357979, | |
| "grad_norm": 9.316176414489746, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 1.7204, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08832385413181666, | |
| "grad_norm": 10.51429271697998, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.401, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09100033456005353, | |
| "grad_norm": 10.570704460144043, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 1.6091, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0936768149882904, | |
| "grad_norm": 12.805994987487793, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.5669, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09635329541652726, | |
| "grad_norm": 10.955977439880371, | |
| "learning_rate": 7e-06, | |
| "loss": 1.4644, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09902977584476413, | |
| "grad_norm": 15.71002197265625, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.5497, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.101706256273001, | |
| "grad_norm": 11.375880241394043, | |
| "learning_rate": 7.4e-06, | |
| "loss": 1.5646, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10438273670123788, | |
| "grad_norm": 12.987727165222168, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 1.4149, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10705921712947473, | |
| "grad_norm": 10.491798400878906, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 1.5017, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1097356975577116, | |
| "grad_norm": 12.43316650390625, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4759, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11241217798594848, | |
| "grad_norm": 12.077725410461426, | |
| "learning_rate": 8.2e-06, | |
| "loss": 1.4892, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11508865841418535, | |
| "grad_norm": 12.642361640930176, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.458, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11776513884242222, | |
| "grad_norm": 15.535161018371582, | |
| "learning_rate": 8.6e-06, | |
| "loss": 1.6861, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.12044161927065908, | |
| "grad_norm": 11.678996086120605, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.6968, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12311809969889595, | |
| "grad_norm": 11.906571388244629, | |
| "learning_rate": 9e-06, | |
| "loss": 1.44, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.12579458012713282, | |
| "grad_norm": 11.492127418518066, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 1.3714, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1284710605553697, | |
| "grad_norm": 15.786943435668945, | |
| "learning_rate": 9.4e-06, | |
| "loss": 1.4132, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13114754098360656, | |
| "grad_norm": 12.31556224822998, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.166, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13382402141184344, | |
| "grad_norm": 12.496630668640137, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 1.2859, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1365005018400803, | |
| "grad_norm": 11.003588676452637, | |
| "learning_rate": 1e-05, | |
| "loss": 1.2614, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.13917698226831715, | |
| "grad_norm": 12.946131706237793, | |
| "learning_rate": 9.99994935597953e-06, | |
| "loss": 1.4555, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.14185346269655402, | |
| "grad_norm": 13.676822662353516, | |
| "learning_rate": 9.999797424944041e-06, | |
| "loss": 1.2167, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.1445299431247909, | |
| "grad_norm": 14.827473640441895, | |
| "learning_rate": 9.999544209971299e-06, | |
| "loss": 1.3456, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14720642355302777, | |
| "grad_norm": 11.921989440917969, | |
| "learning_rate": 9.99918971619083e-06, | |
| "loss": 1.2535, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14988290398126464, | |
| "grad_norm": 12.726682662963867, | |
| "learning_rate": 9.99873395078383e-06, | |
| "loss": 1.2172, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1525593844095015, | |
| "grad_norm": 18.52472496032715, | |
| "learning_rate": 9.998176922983017e-06, | |
| "loss": 1.4418, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.15523586483773838, | |
| "grad_norm": 13.32449722290039, | |
| "learning_rate": 9.99751864407244e-06, | |
| "loss": 1.18, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.15791234526597525, | |
| "grad_norm": 10.778595924377441, | |
| "learning_rate": 9.996759127387259e-06, | |
| "loss": 1.1879, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.16058882569421212, | |
| "grad_norm": 10.155744552612305, | |
| "learning_rate": 9.99589838831346e-06, | |
| "loss": 1.1274, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 13.751497268676758, | |
| "learning_rate": 9.994936444287565e-06, | |
| "loss": 1.1499, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.16594178655068584, | |
| "grad_norm": 17.752635955810547, | |
| "learning_rate": 9.993873314796253e-06, | |
| "loss": 1.1425, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1686182669789227, | |
| "grad_norm": 13.581279754638672, | |
| "learning_rate": 9.992709021375987e-06, | |
| "loss": 1.0341, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.17129474740715958, | |
| "grad_norm": 11.867856979370117, | |
| "learning_rate": 9.991443587612568e-06, | |
| "loss": 1.024, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.17397122783539645, | |
| "grad_norm": 13.20876693725586, | |
| "learning_rate": 9.990077039140655e-06, | |
| "loss": 1.0375, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.17664770826363332, | |
| "grad_norm": 12.211270332336426, | |
| "learning_rate": 9.988609403643254e-06, | |
| "loss": 1.1334, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1793241886918702, | |
| "grad_norm": 11.89450454711914, | |
| "learning_rate": 9.987040710851148e-06, | |
| "loss": 1.0447, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.18200066912010707, | |
| "grad_norm": 10.132635116577148, | |
| "learning_rate": 9.9853709925423e-06, | |
| "loss": 0.7626, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.18467714954834394, | |
| "grad_norm": 9.661967277526855, | |
| "learning_rate": 9.983600282541213e-06, | |
| "loss": 0.968, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1873536299765808, | |
| "grad_norm": 11.19580078125, | |
| "learning_rate": 9.981728616718234e-06, | |
| "loss": 0.9161, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19003011040481765, | |
| "grad_norm": 9.165005683898926, | |
| "learning_rate": 9.979756032988837e-06, | |
| "loss": 0.8821, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.19270659083305453, | |
| "grad_norm": 9.17795181274414, | |
| "learning_rate": 9.977682571312847e-06, | |
| "loss": 1.0452, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.1953830712612914, | |
| "grad_norm": 7.483065128326416, | |
| "learning_rate": 9.975508273693643e-06, | |
| "loss": 1.0135, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.19805955168952827, | |
| "grad_norm": 7.2489728927612305, | |
| "learning_rate": 9.97323318417729e-06, | |
| "loss": 0.9675, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.20073603211776514, | |
| "grad_norm": 7.151744365692139, | |
| "learning_rate": 9.970857348851667e-06, | |
| "loss": 0.829, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.203412512546002, | |
| "grad_norm": 7.535038948059082, | |
| "learning_rate": 9.968380815845504e-06, | |
| "loss": 0.9383, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.20608899297423888, | |
| "grad_norm": 5.609306812286377, | |
| "learning_rate": 9.965803635327445e-06, | |
| "loss": 0.8914, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.20876547340247575, | |
| "grad_norm": 7.203204154968262, | |
| "learning_rate": 9.963125859505e-06, | |
| "loss": 1.0889, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.21144195383071263, | |
| "grad_norm": 4.907174587249756, | |
| "learning_rate": 9.960347542623506e-06, | |
| "loss": 0.7694, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.21411843425894947, | |
| "grad_norm": 5.539646148681641, | |
| "learning_rate": 9.95746874096501e-06, | |
| "loss": 0.9965, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.21679491468718634, | |
| "grad_norm": 5.0046868324279785, | |
| "learning_rate": 9.954489512847156e-06, | |
| "loss": 0.7795, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2194713951154232, | |
| "grad_norm": 5.138219833374023, | |
| "learning_rate": 9.951409918621977e-06, | |
| "loss": 0.8877, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.22214787554366008, | |
| "grad_norm": 4.388546466827393, | |
| "learning_rate": 9.948230020674685e-06, | |
| "loss": 0.9253, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.22482435597189696, | |
| "grad_norm": 4.456934928894043, | |
| "learning_rate": 9.944949883422409e-06, | |
| "loss": 0.7335, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.22750083640013383, | |
| "grad_norm": 4.645761013031006, | |
| "learning_rate": 9.941569573312882e-06, | |
| "loss": 0.8997, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2301773168283707, | |
| "grad_norm": 4.329288482666016, | |
| "learning_rate": 9.938089158823101e-06, | |
| "loss": 0.852, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.23285379725660757, | |
| "grad_norm": 5.044316291809082, | |
| "learning_rate": 9.934508710457944e-06, | |
| "loss": 0.8236, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.23553027768484444, | |
| "grad_norm": 5.356631755828857, | |
| "learning_rate": 9.930828300748726e-06, | |
| "loss": 0.9572, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.23820675811308128, | |
| "grad_norm": 3.933661699295044, | |
| "learning_rate": 9.927048004251748e-06, | |
| "loss": 0.8479, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.24088323854131816, | |
| "grad_norm": 5.228504180908203, | |
| "learning_rate": 9.923167897546773e-06, | |
| "loss": 0.741, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.24355971896955503, | |
| "grad_norm": 4.520639419555664, | |
| "learning_rate": 9.919188059235483e-06, | |
| "loss": 0.746, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2462361993977919, | |
| "grad_norm": 4.191668510437012, | |
| "learning_rate": 9.915108569939884e-06, | |
| "loss": 0.6904, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.24891267982602877, | |
| "grad_norm": 4.300325393676758, | |
| "learning_rate": 9.910929512300673e-06, | |
| "loss": 0.8753, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.25158916025426564, | |
| "grad_norm": 5.979667663574219, | |
| "learning_rate": 9.90665097097556e-06, | |
| "loss": 0.7355, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2542656406825025, | |
| "grad_norm": 5.374980449676514, | |
| "learning_rate": 9.902273032637558e-06, | |
| "loss": 0.7932, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2569421211107394, | |
| "grad_norm": 3.9075119495391846, | |
| "learning_rate": 9.897795785973227e-06, | |
| "loss": 0.9069, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.25961860153897626, | |
| "grad_norm": 4.439886569976807, | |
| "learning_rate": 9.89321932168088e-06, | |
| "loss": 0.8408, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.26229508196721313, | |
| "grad_norm": 4.403755187988281, | |
| "learning_rate": 9.888543732468732e-06, | |
| "loss": 0.709, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.26497156239545, | |
| "grad_norm": 5.023388862609863, | |
| "learning_rate": 9.883769113053039e-06, | |
| "loss": 0.8928, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.26764804282368687, | |
| "grad_norm": 5.586976528167725, | |
| "learning_rate": 9.878895560156172e-06, | |
| "loss": 0.9644, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.27032452325192374, | |
| "grad_norm": 15.905163764953613, | |
| "learning_rate": 9.873923172504653e-06, | |
| "loss": 0.8362, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2730010036801606, | |
| "grad_norm": 5.6663432121276855, | |
| "learning_rate": 9.868852050827167e-06, | |
| "loss": 0.7585, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.27567748410839743, | |
| "grad_norm": 4.5769124031066895, | |
| "learning_rate": 9.863682297852506e-06, | |
| "loss": 0.9748, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2783539645366343, | |
| "grad_norm": 4.305916786193848, | |
| "learning_rate": 9.858414018307503e-06, | |
| "loss": 0.805, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.2810304449648712, | |
| "grad_norm": 4.756903648376465, | |
| "learning_rate": 9.853047318914898e-06, | |
| "loss": 0.8221, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.28370692539310804, | |
| "grad_norm": 5.709683418273926, | |
| "learning_rate": 9.847582308391189e-06, | |
| "loss": 0.7416, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2863834058213449, | |
| "grad_norm": 3.933176279067993, | |
| "learning_rate": 9.842019097444414e-06, | |
| "loss": 0.7298, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2890598862495818, | |
| "grad_norm": 4.9454264640808105, | |
| "learning_rate": 9.836357798771922e-06, | |
| "loss": 0.7683, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.29173636667781866, | |
| "grad_norm": 4.274379730224609, | |
| "learning_rate": 9.830598527058083e-06, | |
| "loss": 0.7788, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.29441284710605553, | |
| "grad_norm": 3.6148972511291504, | |
| "learning_rate": 9.824741398971966e-06, | |
| "loss": 0.6941, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2970893275342924, | |
| "grad_norm": 3.8976197242736816, | |
| "learning_rate": 9.81878653316498e-06, | |
| "loss": 0.8783, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2997658079625293, | |
| "grad_norm": 4.712560176849365, | |
| "learning_rate": 9.81273405026846e-06, | |
| "loss": 0.9022, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.30244228839076615, | |
| "grad_norm": 5.358889579772949, | |
| "learning_rate": 9.806584072891234e-06, | |
| "loss": 0.8587, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.305118768819003, | |
| "grad_norm": 4.296327114105225, | |
| "learning_rate": 9.800336725617136e-06, | |
| "loss": 0.8813, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3077952492472399, | |
| "grad_norm": 4.895114421844482, | |
| "learning_rate": 9.793992135002476e-06, | |
| "loss": 0.8184, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.31047172967547676, | |
| "grad_norm": 4.782822132110596, | |
| "learning_rate": 9.787550429573487e-06, | |
| "loss": 0.6049, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.31314821010371363, | |
| "grad_norm": 5.41484260559082, | |
| "learning_rate": 9.781011739823715e-06, | |
| "loss": 0.7804, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3158246905319505, | |
| "grad_norm": 3.639780044555664, | |
| "learning_rate": 9.77437619821137e-06, | |
| "loss": 0.7241, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3185011709601874, | |
| "grad_norm": 4.5406880378723145, | |
| "learning_rate": 9.767643939156658e-06, | |
| "loss": 0.8398, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.32117765138842425, | |
| "grad_norm": 4.575490951538086, | |
| "learning_rate": 9.760815099039045e-06, | |
| "loss": 0.6951, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3238541318166611, | |
| "grad_norm": 3.868921995162964, | |
| "learning_rate": 9.753889816194498e-06, | |
| "loss": 0.706, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 4.064291477203369, | |
| "learning_rate": 9.746868230912683e-06, | |
| "loss": 0.7386, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3292070926731348, | |
| "grad_norm": 4.059055805206299, | |
| "learning_rate": 9.739750485434126e-06, | |
| "loss": 0.7171, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3318835731013717, | |
| "grad_norm": 4.522129535675049, | |
| "learning_rate": 9.73253672394732e-06, | |
| "loss": 0.6679, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.33456005352960855, | |
| "grad_norm": 3.669271945953369, | |
| "learning_rate": 9.725227092585824e-06, | |
| "loss": 0.7409, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3372365339578454, | |
| "grad_norm": 3.603746175765991, | |
| "learning_rate": 9.717821739425286e-06, | |
| "loss": 0.7286, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3399130143860823, | |
| "grad_norm": 4.118206977844238, | |
| "learning_rate": 9.710320814480448e-06, | |
| "loss": 0.7164, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.34258949481431916, | |
| "grad_norm": 3.6243247985839844, | |
| "learning_rate": 9.702724469702107e-06, | |
| "loss": 0.7772, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.34526597524255603, | |
| "grad_norm": 3.8515329360961914, | |
| "learning_rate": 9.695032858974042e-06, | |
| "loss": 0.7173, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3479424556707929, | |
| "grad_norm": 4.385611534118652, | |
| "learning_rate": 9.687246138109888e-06, | |
| "loss": 0.8672, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3506189360990298, | |
| "grad_norm": 3.6981892585754395, | |
| "learning_rate": 9.679364464849983e-06, | |
| "loss": 0.6571, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.35329541652726665, | |
| "grad_norm": 4.061744213104248, | |
| "learning_rate": 9.671387998858178e-06, | |
| "loss": 0.7114, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3559718969555035, | |
| "grad_norm": 4.793642997741699, | |
| "learning_rate": 9.663316901718599e-06, | |
| "loss": 0.7537, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3586483773837404, | |
| "grad_norm": 3.751180648803711, | |
| "learning_rate": 9.655151336932362e-06, | |
| "loss": 0.6826, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.36132485781197726, | |
| "grad_norm": 3.8843295574188232, | |
| "learning_rate": 9.646891469914285e-06, | |
| "loss": 0.6868, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.36400133824021413, | |
| "grad_norm": 4.41197395324707, | |
| "learning_rate": 9.638537467989517e-06, | |
| "loss": 0.7224, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.366677818668451, | |
| "grad_norm": 3.2272307872772217, | |
| "learning_rate": 9.630089500390154e-06, | |
| "loss": 0.5735, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3693542990966879, | |
| "grad_norm": 3.5232367515563965, | |
| "learning_rate": 9.621547738251816e-06, | |
| "loss": 0.7621, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.37203077952492475, | |
| "grad_norm": 3.4856607913970947, | |
| "learning_rate": 9.61291235461017e-06, | |
| "loss": 0.6517, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3747072599531616, | |
| "grad_norm": 3.958531379699707, | |
| "learning_rate": 9.604183524397439e-06, | |
| "loss": 0.8994, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.37738374038139844, | |
| "grad_norm": 3.3670690059661865, | |
| "learning_rate": 9.595361424438841e-06, | |
| "loss": 0.5975, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3800602208096353, | |
| "grad_norm": 4.388239860534668, | |
| "learning_rate": 9.586446233449024e-06, | |
| "loss": 0.7168, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3827367012378722, | |
| "grad_norm": 3.513667345046997, | |
| "learning_rate": 9.577438132028431e-06, | |
| "loss": 0.7645, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.38541318166610905, | |
| "grad_norm": 3.383009195327759, | |
| "learning_rate": 9.568337302659652e-06, | |
| "loss": 0.7159, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3880896620943459, | |
| "grad_norm": 3.7690727710723877, | |
| "learning_rate": 9.559143929703724e-06, | |
| "loss": 0.8527, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3907661425225828, | |
| "grad_norm": 3.1912662982940674, | |
| "learning_rate": 9.549858199396394e-06, | |
| "loss": 0.6288, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.39344262295081966, | |
| "grad_norm": 3.616091012954712, | |
| "learning_rate": 9.540480299844345e-06, | |
| "loss": 0.6928, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.39611910337905654, | |
| "grad_norm": 3.7126312255859375, | |
| "learning_rate": 9.531010421021396e-06, | |
| "loss": 0.7583, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3987955838072934, | |
| "grad_norm": 3.61950421333313, | |
| "learning_rate": 9.52144875476464e-06, | |
| "loss": 0.6796, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.4014720642355303, | |
| "grad_norm": 3.0952789783477783, | |
| "learning_rate": 9.511795494770563e-06, | |
| "loss": 0.6761, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.40414854466376715, | |
| "grad_norm": 3.4935519695281982, | |
| "learning_rate": 9.50205083659113e-06, | |
| "loss": 0.622, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.406825025092004, | |
| "grad_norm": 4.367190837860107, | |
| "learning_rate": 9.492214977629804e-06, | |
| "loss": 0.8146, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.4095015055202409, | |
| "grad_norm": 3.501554489135742, | |
| "learning_rate": 9.482288117137561e-06, | |
| "loss": 0.6101, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.41217798594847777, | |
| "grad_norm": 3.7483065128326416, | |
| "learning_rate": 9.472270456208856e-06, | |
| "loss": 0.6095, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.41485446637671464, | |
| "grad_norm": 2.7559807300567627, | |
| "learning_rate": 9.462162197777533e-06, | |
| "loss": 0.6167, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.4175309468049515, | |
| "grad_norm": 3.520737409591675, | |
| "learning_rate": 9.451963546612737e-06, | |
| "loss": 0.5388, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4202074272331884, | |
| "grad_norm": 4.042897701263428, | |
| "learning_rate": 9.441674709314743e-06, | |
| "loss": 0.7228, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.42288390766142525, | |
| "grad_norm": 4.090766429901123, | |
| "learning_rate": 9.431295894310786e-06, | |
| "loss": 0.5569, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.42556038808966207, | |
| "grad_norm": 4.141130447387695, | |
| "learning_rate": 9.420827311850836e-06, | |
| "loss": 0.6297, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.42823686851789894, | |
| "grad_norm": 3.472973346710205, | |
| "learning_rate": 9.410269174003333e-06, | |
| "loss": 0.6868, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4309133489461358, | |
| "grad_norm": 3.8000664710998535, | |
| "learning_rate": 9.399621694650898e-06, | |
| "loss": 0.6547, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.4335898293743727, | |
| "grad_norm": 4.582136154174805, | |
| "learning_rate": 9.388885089485995e-06, | |
| "loss": 0.7552, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.43626630980260955, | |
| "grad_norm": 3.4165501594543457, | |
| "learning_rate": 9.378059576006567e-06, | |
| "loss": 0.622, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4389427902308464, | |
| "grad_norm": 3.642669916152954, | |
| "learning_rate": 9.36714537351162e-06, | |
| "loss": 0.6302, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4416192706590833, | |
| "grad_norm": 3.53709077835083, | |
| "learning_rate": 9.356142703096793e-06, | |
| "loss": 0.5357, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.44429575108732017, | |
| "grad_norm": 3.6043941974639893, | |
| "learning_rate": 9.345051787649877e-06, | |
| "loss": 0.6978, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.44697223151555704, | |
| "grad_norm": 4.232686519622803, | |
| "learning_rate": 9.333872851846285e-06, | |
| "loss": 0.6698, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.4496487119437939, | |
| "grad_norm": 3.1254665851593018, | |
| "learning_rate": 9.322606122144524e-06, | |
| "loss": 0.5173, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4523251923720308, | |
| "grad_norm": 4.0060930252075195, | |
| "learning_rate": 9.311251826781587e-06, | |
| "loss": 0.6674, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.45500167280026765, | |
| "grad_norm": 4.720844268798828, | |
| "learning_rate": 9.299810195768341e-06, | |
| "loss": 0.8349, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4576781532285045, | |
| "grad_norm": 4.017560005187988, | |
| "learning_rate": 9.288281460884864e-06, | |
| "loss": 0.8489, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.4603546336567414, | |
| "grad_norm": 4.482756614685059, | |
| "learning_rate": 9.276665855675751e-06, | |
| "loss": 0.6877, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.46303111408497827, | |
| "grad_norm": 3.9189202785491943, | |
| "learning_rate": 9.264963615445378e-06, | |
| "loss": 0.58, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.46570759451321514, | |
| "grad_norm": 3.285618543624878, | |
| "learning_rate": 9.25317497725315e-06, | |
| "loss": 0.585, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.468384074941452, | |
| "grad_norm": 3.4484951496124268, | |
| "learning_rate": 9.241300179908672e-06, | |
| "loss": 0.5898, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4710605553696889, | |
| "grad_norm": 3.892336845397949, | |
| "learning_rate": 9.229339463966942e-06, | |
| "loss": 0.7691, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.47373703579792575, | |
| "grad_norm": 3.3191475868225098, | |
| "learning_rate": 9.217293071723455e-06, | |
| "loss": 0.6468, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.47641351622616257, | |
| "grad_norm": 3.2376253604888916, | |
| "learning_rate": 9.205161247209303e-06, | |
| "loss": 0.6233, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.47908999665439944, | |
| "grad_norm": 3.029269218444824, | |
| "learning_rate": 9.192944236186237e-06, | |
| "loss": 0.5845, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4817664770826363, | |
| "grad_norm": 3.5170111656188965, | |
| "learning_rate": 9.180642286141678e-06, | |
| "loss": 0.6815, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4844429575108732, | |
| "grad_norm": 4.0202155113220215, | |
| "learning_rate": 9.16825564628371e-06, | |
| "loss": 0.7131, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.48711943793911006, | |
| "grad_norm": 3.566044330596924, | |
| "learning_rate": 9.15578456753603e-06, | |
| "loss": 0.6628, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 3.9581563472747803, | |
| "learning_rate": 9.143229302532866e-06, | |
| "loss": 0.6497, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4924723987955838, | |
| "grad_norm": 3.2238948345184326, | |
| "learning_rate": 9.130590105613854e-06, | |
| "loss": 0.5773, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.49514887922382067, | |
| "grad_norm": 4.199437618255615, | |
| "learning_rate": 9.117867232818897e-06, | |
| "loss": 0.5552, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.49782535965205754, | |
| "grad_norm": 3.223834753036499, | |
| "learning_rate": 9.105060941882966e-06, | |
| "loss": 0.4985, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.5005018400802944, | |
| "grad_norm": 3.2973475456237793, | |
| "learning_rate": 9.092171492230883e-06, | |
| "loss": 0.6637, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5031783205085313, | |
| "grad_norm": 3.9001471996307373, | |
| "learning_rate": 9.079199144972072e-06, | |
| "loss": 0.7487, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5058548009367682, | |
| "grad_norm": 3.913593292236328, | |
| "learning_rate": 9.066144162895259e-06, | |
| "loss": 0.5707, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.508531281365005, | |
| "grad_norm": 2.6204168796539307, | |
| "learning_rate": 9.053006810463156e-06, | |
| "loss": 0.4685, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5112077617932419, | |
| "grad_norm": 3.507556200027466, | |
| "learning_rate": 9.039787353807101e-06, | |
| "loss": 0.5564, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5138842422214788, | |
| "grad_norm": 3.072106122970581, | |
| "learning_rate": 9.026486060721668e-06, | |
| "loss": 0.5427, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5165607226497156, | |
| "grad_norm": 2.578962802886963, | |
| "learning_rate": 9.01310320065924e-06, | |
| "loss": 0.3969, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5192372030779525, | |
| "grad_norm": 3.3628005981445312, | |
| "learning_rate": 8.999639044724555e-06, | |
| "loss": 0.6076, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5219136835061894, | |
| "grad_norm": 3.64412784576416, | |
| "learning_rate": 8.986093865669205e-06, | |
| "loss": 0.6826, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5245901639344263, | |
| "grad_norm": 3.7886106967926025, | |
| "learning_rate": 8.972467937886122e-06, | |
| "loss": 0.5279, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5272666443626631, | |
| "grad_norm": 3.543992757797241, | |
| "learning_rate": 8.958761537404012e-06, | |
| "loss": 0.5748, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5299431247909, | |
| "grad_norm": 4.291012763977051, | |
| "learning_rate": 8.944974941881766e-06, | |
| "loss": 0.6919, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5326196052191369, | |
| "grad_norm": 4.906763553619385, | |
| "learning_rate": 8.931108430602834e-06, | |
| "loss": 0.5633, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5352960856473737, | |
| "grad_norm": 3.177494764328003, | |
| "learning_rate": 8.917162284469569e-06, | |
| "loss": 0.5792, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5379725660756106, | |
| "grad_norm": 3.5298800468444824, | |
| "learning_rate": 8.903136785997533e-06, | |
| "loss": 0.5616, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5406490465038475, | |
| "grad_norm": 3.5532355308532715, | |
| "learning_rate": 8.889032219309781e-06, | |
| "loss": 0.6682, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.5433255269320844, | |
| "grad_norm": 4.2544026374816895, | |
| "learning_rate": 8.874848870131098e-06, | |
| "loss": 0.544, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.5460020073603212, | |
| "grad_norm": 3.0393524169921875, | |
| "learning_rate": 8.860587025782215e-06, | |
| "loss": 0.6333, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.5486784877885581, | |
| "grad_norm": 4.1217122077941895, | |
| "learning_rate": 8.846246975173985e-06, | |
| "loss": 0.597, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5513549682167949, | |
| "grad_norm": 3.7197227478027344, | |
| "learning_rate": 8.831829008801536e-06, | |
| "loss": 0.6568, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.5540314486450317, | |
| "grad_norm": 3.254206657409668, | |
| "learning_rate": 8.817333418738382e-06, | |
| "loss": 0.5321, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5567079290732686, | |
| "grad_norm": 3.084484577178955, | |
| "learning_rate": 8.802760498630507e-06, | |
| "loss": 0.4752, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5593844095015055, | |
| "grad_norm": 4.343369483947754, | |
| "learning_rate": 8.788110543690415e-06, | |
| "loss": 0.6444, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5620608899297423, | |
| "grad_norm": 3.27128529548645, | |
| "learning_rate": 8.773383850691155e-06, | |
| "loss": 0.5282, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5647373703579792, | |
| "grad_norm": 3.0274081230163574, | |
| "learning_rate": 8.758580717960303e-06, | |
| "loss": 0.6607, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5674138507862161, | |
| "grad_norm": 3.622535228729248, | |
| "learning_rate": 8.743701445373922e-06, | |
| "loss": 0.5787, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.570090331214453, | |
| "grad_norm": 3.6681504249572754, | |
| "learning_rate": 8.728746334350483e-06, | |
| "loss": 0.5879, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5727668116426898, | |
| "grad_norm": 3.163616895675659, | |
| "learning_rate": 8.713715687844772e-06, | |
| "loss": 0.5888, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5754432920709267, | |
| "grad_norm": 3.0883967876434326, | |
| "learning_rate": 8.698609810341733e-06, | |
| "loss": 0.5189, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5781197724991636, | |
| "grad_norm": 3.443776845932007, | |
| "learning_rate": 8.683429007850313e-06, | |
| "loss": 0.5303, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5807962529274004, | |
| "grad_norm": 3.3351690769195557, | |
| "learning_rate": 8.668173587897261e-06, | |
| "loss": 0.5296, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5834727333556373, | |
| "grad_norm": 3.8925840854644775, | |
| "learning_rate": 8.6528438595209e-06, | |
| "loss": 0.499, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5861492137838742, | |
| "grad_norm": 3.4368419647216797, | |
| "learning_rate": 8.637440133264858e-06, | |
| "loss": 0.4166, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5888256942121111, | |
| "grad_norm": 3.439969539642334, | |
| "learning_rate": 8.621962721171789e-06, | |
| "loss": 0.5417, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5915021746403479, | |
| "grad_norm": 3.2590739727020264, | |
| "learning_rate": 8.60641193677704e-06, | |
| "loss": 0.4925, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5941786550685848, | |
| "grad_norm": 3.918884754180908, | |
| "learning_rate": 8.59078809510231e-06, | |
| "loss": 0.75, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5968551354968217, | |
| "grad_norm": 4.21658182144165, | |
| "learning_rate": 8.57509151264926e-06, | |
| "loss": 0.5138, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5995316159250585, | |
| "grad_norm": 3.2700538635253906, | |
| "learning_rate": 8.55932250739311e-06, | |
| "loss": 0.4855, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.6022080963532954, | |
| "grad_norm": 3.5396180152893066, | |
| "learning_rate": 8.543481398776188e-06, | |
| "loss": 0.626, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6048845767815323, | |
| "grad_norm": 3.267137289047241, | |
| "learning_rate": 8.527568507701467e-06, | |
| "loss": 0.5043, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6075610572097692, | |
| "grad_norm": 4.908371448516846, | |
| "learning_rate": 8.511584156526059e-06, | |
| "loss": 0.6942, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.610237537638006, | |
| "grad_norm": 3.7445895671844482, | |
| "learning_rate": 8.495528669054688e-06, | |
| "loss": 0.5475, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6129140180662429, | |
| "grad_norm": 4.3304033279418945, | |
| "learning_rate": 8.479402370533127e-06, | |
| "loss": 0.6943, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.6155904984944798, | |
| "grad_norm": 3.6123831272125244, | |
| "learning_rate": 8.463205587641614e-06, | |
| "loss": 0.5365, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6182669789227166, | |
| "grad_norm": 4.3004045486450195, | |
| "learning_rate": 8.44693864848823e-06, | |
| "loss": 0.6377, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6209434593509535, | |
| "grad_norm": 3.1220314502716064, | |
| "learning_rate": 8.430601882602256e-06, | |
| "loss": 0.4646, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6236199397791904, | |
| "grad_norm": 4.166379451751709, | |
| "learning_rate": 8.414195620927491e-06, | |
| "loss": 0.5738, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.6262964202074273, | |
| "grad_norm": 4.061609745025635, | |
| "learning_rate": 8.397720195815561e-06, | |
| "loss": 0.4654, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.6289729006356641, | |
| "grad_norm": 3.618021249771118, | |
| "learning_rate": 8.381175941019171e-06, | |
| "loss": 0.5699, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.631649381063901, | |
| "grad_norm": 3.2099597454071045, | |
| "learning_rate": 8.364563191685348e-06, | |
| "loss": 0.5481, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6343258614921379, | |
| "grad_norm": 3.3331351280212402, | |
| "learning_rate": 8.347882284348665e-06, | |
| "loss": 0.5351, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6370023419203747, | |
| "grad_norm": 3.91636323928833, | |
| "learning_rate": 8.331133556924404e-06, | |
| "loss": 0.6588, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6396788223486116, | |
| "grad_norm": 3.613023519515991, | |
| "learning_rate": 8.314317348701724e-06, | |
| "loss": 0.6157, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.6423553027768485, | |
| "grad_norm": 3.674731731414795, | |
| "learning_rate": 8.297434000336781e-06, | |
| "loss": 0.5251, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6450317832050854, | |
| "grad_norm": 3.4246506690979004, | |
| "learning_rate": 8.280483853845831e-06, | |
| "loss": 0.5449, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.6477082636333222, | |
| "grad_norm": 3.308436155319214, | |
| "learning_rate": 8.263467252598303e-06, | |
| "loss": 0.6432, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.650384744061559, | |
| "grad_norm": 3.3527414798736572, | |
| "learning_rate": 8.246384541309835e-06, | |
| "loss": 0.6035, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 3.5172030925750732, | |
| "learning_rate": 8.2292360660353e-06, | |
| "loss": 0.522, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "grad_norm": 3.086130142211914, | |
| "learning_rate": 8.21202217416179e-06, | |
| "loss": 0.4867, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6584141853462696, | |
| "grad_norm": 4.54761266708374, | |
| "learning_rate": 8.194743214401587e-06, | |
| "loss": 0.5607, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.6610906657745065, | |
| "grad_norm": 3.190037250518799, | |
| "learning_rate": 8.17739953678508e-06, | |
| "loss": 0.4674, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.6637671462027434, | |
| "grad_norm": 3.303924083709717, | |
| "learning_rate": 8.1599914926537e-06, | |
| "loss": 0.5116, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6664436266309802, | |
| "grad_norm": 3.0905206203460693, | |
| "learning_rate": 8.142519434652782e-06, | |
| "loss": 0.5041, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6691201070592171, | |
| "grad_norm": 3.9465432167053223, | |
| "learning_rate": 8.124983716724434e-06, | |
| "loss": 0.5009, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.671796587487454, | |
| "grad_norm": 3.0109636783599854, | |
| "learning_rate": 8.107384694100355e-06, | |
| "loss": 0.5041, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6744730679156908, | |
| "grad_norm": 3.4623425006866455, | |
| "learning_rate": 8.089722723294654e-06, | |
| "loss": 0.5509, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6771495483439277, | |
| "grad_norm": 4.234053134918213, | |
| "learning_rate": 8.071998162096613e-06, | |
| "loss": 0.6164, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6798260287721646, | |
| "grad_norm": 3.2587759494781494, | |
| "learning_rate": 8.054211369563448e-06, | |
| "loss": 0.5531, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6825025092004015, | |
| "grad_norm": 3.6140904426574707, | |
| "learning_rate": 8.036362706013033e-06, | |
| "loss": 0.5566, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6851789896286383, | |
| "grad_norm": 2.9970974922180176, | |
| "learning_rate": 8.018452533016604e-06, | |
| "loss": 0.4614, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.6878554700568752, | |
| "grad_norm": 3.6728920936584473, | |
| "learning_rate": 8.000481213391422e-06, | |
| "loss": 0.5529, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6905319504851121, | |
| "grad_norm": 4.451485633850098, | |
| "learning_rate": 7.982449111193445e-06, | |
| "loss": 0.5997, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6932084309133489, | |
| "grad_norm": 3.2221109867095947, | |
| "learning_rate": 7.96435659170993e-06, | |
| "loss": 0.4782, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6958849113415858, | |
| "grad_norm": 3.3525805473327637, | |
| "learning_rate": 7.946204021452049e-06, | |
| "loss": 0.5988, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6985613917698227, | |
| "grad_norm": 3.346829652786255, | |
| "learning_rate": 7.92799176814746e-06, | |
| "loss": 0.461, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.7012378721980596, | |
| "grad_norm": 3.4841763973236084, | |
| "learning_rate": 7.90972020073285e-06, | |
| "loss": 0.5972, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.7039143526262964, | |
| "grad_norm": 3.1365015506744385, | |
| "learning_rate": 7.891389689346479e-06, | |
| "loss": 0.4601, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7065908330545333, | |
| "grad_norm": 2.762209177017212, | |
| "learning_rate": 7.873000605320658e-06, | |
| "loss": 0.4713, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7092673134827702, | |
| "grad_norm": 3.6923787593841553, | |
| "learning_rate": 7.85455332117425e-06, | |
| "loss": 0.5868, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.711943793911007, | |
| "grad_norm": 3.483915090560913, | |
| "learning_rate": 7.836048210605109e-06, | |
| "loss": 0.5488, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7146202743392439, | |
| "grad_norm": 3.7506582736968994, | |
| "learning_rate": 7.817485648482514e-06, | |
| "loss": 0.506, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.7172967547674808, | |
| "grad_norm": 3.148474931716919, | |
| "learning_rate": 7.798866010839577e-06, | |
| "loss": 0.4793, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.7199732351957177, | |
| "grad_norm": 3.194519281387329, | |
| "learning_rate": 7.780189674865617e-06, | |
| "loss": 0.5573, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.7226497156239545, | |
| "grad_norm": 3.321631669998169, | |
| "learning_rate": 7.761457018898536e-06, | |
| "loss": 0.4949, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7253261960521914, | |
| "grad_norm": 3.322242021560669, | |
| "learning_rate": 7.742668422417137e-06, | |
| "loss": 0.5329, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.7280026764804283, | |
| "grad_norm": 3.460432291030884, | |
| "learning_rate": 7.723824266033444e-06, | |
| "loss": 0.5798, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.7306791569086651, | |
| "grad_norm": 3.0027997493743896, | |
| "learning_rate": 7.704924931484997e-06, | |
| "loss": 0.5512, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.733355637336902, | |
| "grad_norm": 3.621713399887085, | |
| "learning_rate": 7.685970801627108e-06, | |
| "loss": 0.6221, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.7360321177651389, | |
| "grad_norm": 2.8740806579589844, | |
| "learning_rate": 7.666962260425113e-06, | |
| "loss": 0.532, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7387085981933758, | |
| "grad_norm": 2.9761555194854736, | |
| "learning_rate": 7.647899692946594e-06, | |
| "loss": 0.4211, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7413850786216126, | |
| "grad_norm": 2.8922770023345947, | |
| "learning_rate": 7.628783485353573e-06, | |
| "loss": 0.4465, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7440615590498495, | |
| "grad_norm": 2.775923252105713, | |
| "learning_rate": 7.609614024894694e-06, | |
| "loss": 0.46, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7467380394780864, | |
| "grad_norm": 3.38106632232666, | |
| "learning_rate": 7.5903916998973745e-06, | |
| "loss": 0.6024, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.7494145199063232, | |
| "grad_norm": 2.9495768547058105, | |
| "learning_rate": 7.571116899759945e-06, | |
| "loss": 0.4792, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.75209100033456, | |
| "grad_norm": 3.18288516998291, | |
| "learning_rate": 7.551790014943752e-06, | |
| "loss": 0.4439, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.7547674807627969, | |
| "grad_norm": 3.185044288635254, | |
| "learning_rate": 7.532411436965258e-06, | |
| "loss": 0.5251, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.7574439611910337, | |
| "grad_norm": 2.8244149684906006, | |
| "learning_rate": 7.512981558388101e-06, | |
| "loss": 0.4443, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.7601204416192706, | |
| "grad_norm": 3.172478437423706, | |
| "learning_rate": 7.49350077281515e-06, | |
| "loss": 0.5057, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.7627969220475075, | |
| "grad_norm": 3.2806384563446045, | |
| "learning_rate": 7.473969474880527e-06, | |
| "loss": 0.5944, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.7654734024757444, | |
| "grad_norm": 2.7359542846679688, | |
| "learning_rate": 7.45438806024161e-06, | |
| "loss": 0.4296, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.7681498829039812, | |
| "grad_norm": 2.7269349098205566, | |
| "learning_rate": 7.4347569255710254e-06, | |
| "loss": 0.4004, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.7708263633322181, | |
| "grad_norm": 3.779465913772583, | |
| "learning_rate": 7.41507646854861e-06, | |
| "loss": 0.5674, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.773502843760455, | |
| "grad_norm": 2.7215170860290527, | |
| "learning_rate": 7.395347087853349e-06, | |
| "loss": 0.3874, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.7761793241886918, | |
| "grad_norm": 4.240821838378906, | |
| "learning_rate": 7.375569183155306e-06, | |
| "loss": 0.5878, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7788558046169287, | |
| "grad_norm": 3.43664288520813, | |
| "learning_rate": 7.355743155107526e-06, | |
| "loss": 0.4482, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7815322850451656, | |
| "grad_norm": 3.580437660217285, | |
| "learning_rate": 7.335869405337919e-06, | |
| "loss": 0.5135, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7842087654734025, | |
| "grad_norm": 3.794210433959961, | |
| "learning_rate": 7.3159483364411175e-06, | |
| "loss": 0.4546, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.7868852459016393, | |
| "grad_norm": 2.8526501655578613, | |
| "learning_rate": 7.295980351970331e-06, | |
| "loss": 0.416, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7895617263298762, | |
| "grad_norm": 3.249462127685547, | |
| "learning_rate": 7.275965856429167e-06, | |
| "loss": 0.4575, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7922382067581131, | |
| "grad_norm": 4.065032482147217, | |
| "learning_rate": 7.255905255263434e-06, | |
| "loss": 0.4073, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7949146871863499, | |
| "grad_norm": 2.9181621074676514, | |
| "learning_rate": 7.235798954852929e-06, | |
| "loss": 0.467, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7975911676145868, | |
| "grad_norm": 3.012425661087036, | |
| "learning_rate": 7.2156473625032075e-06, | |
| "loss": 0.5308, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.8002676480428237, | |
| "grad_norm": 3.2600486278533936, | |
| "learning_rate": 7.195450886437334e-06, | |
| "loss": 0.5594, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.8029441284710606, | |
| "grad_norm": 3.5234806537628174, | |
| "learning_rate": 7.175209935787605e-06, | |
| "loss": 0.4377, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8056206088992974, | |
| "grad_norm": 3.3013510704040527, | |
| "learning_rate": 7.154924920587269e-06, | |
| "loss": 0.5039, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8082970893275343, | |
| "grad_norm": 3.6095550060272217, | |
| "learning_rate": 7.134596251762217e-06, | |
| "loss": 0.5333, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8109735697557712, | |
| "grad_norm": 3.533142328262329, | |
| "learning_rate": 7.114224341122655e-06, | |
| "loss": 0.4358, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.813650050184008, | |
| "grad_norm": 3.41983699798584, | |
| "learning_rate": 7.093809601354769e-06, | |
| "loss": 0.5369, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 2.5269010066986084, | |
| "learning_rate": 7.073352446012357e-06, | |
| "loss": 0.4014, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.8190030110404818, | |
| "grad_norm": 3.3515594005584717, | |
| "learning_rate": 7.052853289508458e-06, | |
| "loss": 0.4681, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.8216794914687187, | |
| "grad_norm": 3.920032501220703, | |
| "learning_rate": 7.03231254710695e-06, | |
| "loss": 0.5876, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.8243559718969555, | |
| "grad_norm": 2.734036684036255, | |
| "learning_rate": 7.0117306349141485e-06, | |
| "loss": 0.4294, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.8270324523251924, | |
| "grad_norm": 2.5630767345428467, | |
| "learning_rate": 6.991107969870363e-06, | |
| "loss": 0.3869, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.8297089327534293, | |
| "grad_norm": 2.897214889526367, | |
| "learning_rate": 6.970444969741462e-06, | |
| "loss": 0.4756, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8323854131816661, | |
| "grad_norm": 3.645385980606079, | |
| "learning_rate": 6.949742053110408e-06, | |
| "loss": 0.465, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.835061893609903, | |
| "grad_norm": 2.9441306591033936, | |
| "learning_rate": 6.928999639368773e-06, | |
| "loss": 0.4803, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.8377383740381399, | |
| "grad_norm": 3.1111350059509277, | |
| "learning_rate": 6.908218148708248e-06, | |
| "loss": 0.4119, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.8404148544663768, | |
| "grad_norm": 2.8598246574401855, | |
| "learning_rate": 6.887398002112129e-06, | |
| "loss": 0.5, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.8430913348946136, | |
| "grad_norm": 2.703573703765869, | |
| "learning_rate": 6.866539621346786e-06, | |
| "loss": 0.3387, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8457678153228505, | |
| "grad_norm": 2.924374580383301, | |
| "learning_rate": 6.845643428953127e-06, | |
| "loss": 0.4498, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8484442957510874, | |
| "grad_norm": 3.2768142223358154, | |
| "learning_rate": 6.824709848238028e-06, | |
| "loss": 0.4514, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.8511207761793241, | |
| "grad_norm": 3.422415256500244, | |
| "learning_rate": 6.8037393032657665e-06, | |
| "loss": 0.4806, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.853797256607561, | |
| "grad_norm": 3.6226656436920166, | |
| "learning_rate": 6.782732218849425e-06, | |
| "loss": 0.431, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.8564737370357979, | |
| "grad_norm": 3.154444932937622, | |
| "learning_rate": 6.761689020542288e-06, | |
| "loss": 0.4059, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8591502174640347, | |
| "grad_norm": 3.5488181114196777, | |
| "learning_rate": 6.740610134629224e-06, | |
| "loss": 0.5335, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.8618266978922716, | |
| "grad_norm": 3.483675718307495, | |
| "learning_rate": 6.719495988118043e-06, | |
| "loss": 0.5676, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.8645031783205085, | |
| "grad_norm": 3.83905291557312, | |
| "learning_rate": 6.698347008730854e-06, | |
| "loss": 0.5997, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.8671796587487454, | |
| "grad_norm": 3.451472759246826, | |
| "learning_rate": 6.677163624895393e-06, | |
| "loss": 0.4285, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.8698561391769822, | |
| "grad_norm": 3.9657328128814697, | |
| "learning_rate": 6.6559462657363525e-06, | |
| "loss": 0.5505, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8725326196052191, | |
| "grad_norm": 3.474752902984619, | |
| "learning_rate": 6.634695361066679e-06, | |
| "loss": 0.577, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.875209100033456, | |
| "grad_norm": 3.0896995067596436, | |
| "learning_rate": 6.613411341378872e-06, | |
| "loss": 0.4052, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.8778855804616928, | |
| "grad_norm": 2.541867256164551, | |
| "learning_rate": 6.592094637836266e-06, | |
| "loss": 0.3609, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.8805620608899297, | |
| "grad_norm": 3.222644329071045, | |
| "learning_rate": 6.570745682264288e-06, | |
| "loss": 0.4346, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.8832385413181666, | |
| "grad_norm": 2.9932363033294678, | |
| "learning_rate": 6.549364907141713e-06, | |
| "loss": 0.3862, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8859150217464035, | |
| "grad_norm": 2.610246181488037, | |
| "learning_rate": 6.527952745591911e-06, | |
| "loss": 0.3075, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8885915021746403, | |
| "grad_norm": 3.310206174850464, | |
| "learning_rate": 6.506509631374056e-06, | |
| "loss": 0.4288, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8912679826028772, | |
| "grad_norm": 3.0393242835998535, | |
| "learning_rate": 6.485035998874356e-06, | |
| "loss": 0.5224, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8939444630311141, | |
| "grad_norm": 3.029529571533203, | |
| "learning_rate": 6.4635322830972465e-06, | |
| "loss": 0.5228, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.896620943459351, | |
| "grad_norm": 2.9739904403686523, | |
| "learning_rate": 6.441998919656575e-06, | |
| "loss": 0.4464, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8992974238875878, | |
| "grad_norm": 3.4808287620544434, | |
| "learning_rate": 6.420436344766781e-06, | |
| "loss": 0.4787, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.9019739043158247, | |
| "grad_norm": 3.4027199745178223, | |
| "learning_rate": 6.398844995234057e-06, | |
| "loss": 0.4964, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.9046503847440616, | |
| "grad_norm": 3.5456416606903076, | |
| "learning_rate": 6.377225308447503e-06, | |
| "loss": 0.5533, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.9073268651722984, | |
| "grad_norm": 3.416761875152588, | |
| "learning_rate": 6.355577722370264e-06, | |
| "loss": 0.4679, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.9100033456005353, | |
| "grad_norm": 3.634335994720459, | |
| "learning_rate": 6.333902675530657e-06, | |
| "loss": 0.4683, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9126798260287722, | |
| "grad_norm": 2.5649242401123047, | |
| "learning_rate": 6.312200607013287e-06, | |
| "loss": 0.3649, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.915356306457009, | |
| "grad_norm": 3.457108736038208, | |
| "learning_rate": 6.2904719564501545e-06, | |
| "loss": 0.559, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.9180327868852459, | |
| "grad_norm": 3.2710628509521484, | |
| "learning_rate": 6.268717164011751e-06, | |
| "loss": 0.4526, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.9207092673134828, | |
| "grad_norm": 2.584167003631592, | |
| "learning_rate": 6.246936670398136e-06, | |
| "loss": 0.3223, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.9233857477417197, | |
| "grad_norm": 3.2143352031707764, | |
| "learning_rate": 6.225130916830017e-06, | |
| "loss": 0.3996, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.9260622281699565, | |
| "grad_norm": 2.900538206100464, | |
| "learning_rate": 6.203300345039804e-06, | |
| "loss": 0.4241, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.9287387085981934, | |
| "grad_norm": 2.566956043243408, | |
| "learning_rate": 6.181445397262671e-06, | |
| "loss": 0.3754, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.9314151890264303, | |
| "grad_norm": 3.051973819732666, | |
| "learning_rate": 6.159566516227582e-06, | |
| "loss": 0.464, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.9340916694546672, | |
| "grad_norm": 2.648303985595703, | |
| "learning_rate": 6.137664145148339e-06, | |
| "loss": 0.3669, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.936768149882904, | |
| "grad_norm": 3.3887152671813965, | |
| "learning_rate": 6.115738727714593e-06, | |
| "loss": 0.4347, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9394446303111409, | |
| "grad_norm": 2.8967878818511963, | |
| "learning_rate": 6.093790708082861e-06, | |
| "loss": 0.4126, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.9421211107393778, | |
| "grad_norm": 3.301456928253174, | |
| "learning_rate": 6.071820530867524e-06, | |
| "loss": 0.4932, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.9447975911676146, | |
| "grad_norm": 2.467073678970337, | |
| "learning_rate": 6.0498286411318255e-06, | |
| "loss": 0.3714, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.9474740715958515, | |
| "grad_norm": 2.737948417663574, | |
| "learning_rate": 6.027815484378848e-06, | |
| "loss": 0.3832, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.9501505520240883, | |
| "grad_norm": 3.305079221725464, | |
| "learning_rate": 6.005781506542498e-06, | |
| "loss": 0.3813, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9528270324523251, | |
| "grad_norm": 3.1508371829986572, | |
| "learning_rate": 5.983727153978467e-06, | |
| "loss": 0.5005, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.955503512880562, | |
| "grad_norm": 3.4938628673553467, | |
| "learning_rate": 5.961652873455186e-06, | |
| "loss": 0.4586, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9581799933087989, | |
| "grad_norm": 2.7031075954437256, | |
| "learning_rate": 5.939559112144781e-06, | |
| "loss": 0.3995, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9608564737370358, | |
| "grad_norm": 3.1249899864196777, | |
| "learning_rate": 5.917446317614012e-06, | |
| "loss": 0.4738, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.9635329541652726, | |
| "grad_norm": 3.1881392002105713, | |
| "learning_rate": 5.895314937815206e-06, | |
| "loss": 0.4234, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9662094345935095, | |
| "grad_norm": 3.020379066467285, | |
| "learning_rate": 5.873165421077186e-06, | |
| "loss": 0.347, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.9688859150217464, | |
| "grad_norm": 2.737119436264038, | |
| "learning_rate": 5.850998216096181e-06, | |
| "loss": 0.3805, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.9715623954499832, | |
| "grad_norm": 3.5535435676574707, | |
| "learning_rate": 5.828813771926746e-06, | |
| "loss": 0.5056, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.9742388758782201, | |
| "grad_norm": 3.3006703853607178, | |
| "learning_rate": 5.806612537972658e-06, | |
| "loss": 0.3342, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.976915356306457, | |
| "grad_norm": 3.176393508911133, | |
| "learning_rate": 5.784394963977815e-06, | |
| "loss": 0.4834, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 3.3865625858306885, | |
| "learning_rate": 5.762161500017128e-06, | |
| "loss": 0.4573, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.9822683171629307, | |
| "grad_norm": 2.942143678665161, | |
| "learning_rate": 5.739912596487396e-06, | |
| "loss": 0.4826, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.9849447975911676, | |
| "grad_norm": 2.834310531616211, | |
| "learning_rate": 5.717648704098191e-06, | |
| "loss": 0.444, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.9876212780194045, | |
| "grad_norm": 2.967898368835449, | |
| "learning_rate": 5.6953702738627215e-06, | |
| "loss": 0.3811, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.9902977584476413, | |
| "grad_norm": 2.998699426651001, | |
| "learning_rate": 5.6730777570887e-06, | |
| "loss": 0.4694, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9929742388758782, | |
| "grad_norm": 3.34025502204895, | |
| "learning_rate": 5.6507716053691916e-06, | |
| "loss": 0.5136, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.9956507193041151, | |
| "grad_norm": 2.5834507942199707, | |
| "learning_rate": 5.628452270573483e-06, | |
| "loss": 0.3702, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.998327199732352, | |
| "grad_norm": 2.72577166557312, | |
| "learning_rate": 5.6061202048379125e-06, | |
| "loss": 0.3757, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.7107036113739014, | |
| "learning_rate": 5.583775860556717e-06, | |
| "loss": 0.3578, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.0026764804282369, | |
| "grad_norm": 2.568798542022705, | |
| "learning_rate": 5.561419690372869e-06, | |
| "loss": 0.1397, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.0053529608564737, | |
| "grad_norm": 1.9347107410430908, | |
| "learning_rate": 5.539052147168903e-06, | |
| "loss": 0.2252, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.0080294412847106, | |
| "grad_norm": 2.0152370929718018, | |
| "learning_rate": 5.516673684057747e-06, | |
| "loss": 0.2206, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.0107059217129475, | |
| "grad_norm": 2.0257060527801514, | |
| "learning_rate": 5.494284754373538e-06, | |
| "loss": 0.1776, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.0133824021411844, | |
| "grad_norm": 2.107743263244629, | |
| "learning_rate": 5.471885811662442e-06, | |
| "loss": 0.1799, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.0160588825694212, | |
| "grad_norm": 2.221683979034424, | |
| "learning_rate": 5.449477309673462e-06, | |
| "loss": 0.1642, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.018735362997658, | |
| "grad_norm": 2.8024303913116455, | |
| "learning_rate": 5.427059702349255e-06, | |
| "loss": 0.1712, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.021411843425895, | |
| "grad_norm": 2.7197983264923096, | |
| "learning_rate": 5.4046334438169245e-06, | |
| "loss": 0.2057, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.0240883238541318, | |
| "grad_norm": 1.90606689453125, | |
| "learning_rate": 5.382198988378829e-06, | |
| "loss": 0.1793, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.0267648042823687, | |
| "grad_norm": 2.782726764678955, | |
| "learning_rate": 5.359756790503376e-06, | |
| "loss": 0.2036, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.0294412847106056, | |
| "grad_norm": 2.3498716354370117, | |
| "learning_rate": 5.337307304815817e-06, | |
| "loss": 0.2134, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.0321177651388425, | |
| "grad_norm": 3.6276447772979736, | |
| "learning_rate": 5.31485098608904e-06, | |
| "loss": 0.1908, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.0347942455670793, | |
| "grad_norm": 3.2593178749084473, | |
| "learning_rate": 5.292388289234349e-06, | |
| "loss": 0.2326, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.0374707259953162, | |
| "grad_norm": 3.1643850803375244, | |
| "learning_rate": 5.2699196692922546e-06, | |
| "loss": 0.2165, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.040147206423553, | |
| "grad_norm": 2.73394513130188, | |
| "learning_rate": 5.247445581423257e-06, | |
| "loss": 0.1906, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.04282368685179, | |
| "grad_norm": 3.0761959552764893, | |
| "learning_rate": 5.224966480898624e-06, | |
| "loss": 0.1903, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0455001672800268, | |
| "grad_norm": 3.4978442192077637, | |
| "learning_rate": 5.202482823091165e-06, | |
| "loss": 0.211, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.0481766477082637, | |
| "grad_norm": 2.406022548675537, | |
| "learning_rate": 5.179995063466011e-06, | |
| "loss": 0.2095, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.0508531281365006, | |
| "grad_norm": 3.4254419803619385, | |
| "learning_rate": 5.157503657571386e-06, | |
| "loss": 0.2339, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.0535296085647374, | |
| "grad_norm": 3.1183996200561523, | |
| "learning_rate": 5.1350090610293765e-06, | |
| "loss": 0.1881, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.0562060889929743, | |
| "grad_norm": 2.5185012817382812, | |
| "learning_rate": 5.112511729526708e-06, | |
| "loss": 0.1261, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.0588825694212112, | |
| "grad_norm": 2.3276853561401367, | |
| "learning_rate": 5.090012118805505e-06, | |
| "loss": 0.1423, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.061559049849448, | |
| "grad_norm": 2.8159985542297363, | |
| "learning_rate": 5.067510684654069e-06, | |
| "loss": 0.1786, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.064235530277685, | |
| "grad_norm": 2.6511971950531006, | |
| "learning_rate": 5.0450078828976326e-06, | |
| "loss": 0.2184, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.0669120107059218, | |
| "grad_norm": 2.8638663291931152, | |
| "learning_rate": 5.02250416938914e-06, | |
| "loss": 0.1745, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.0695884911341587, | |
| "grad_norm": 2.6509993076324463, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1678, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 748, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.392429255582024e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |