| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5352960856473737, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "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 | |
| } | |
| ], | |
| "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": 2.225232452601774e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |