| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1235, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004048582995951417, | |
| "grad_norm": 6.448712376675753, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 1.041, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008097165991902834, | |
| "grad_norm": 6.571817448430504, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 1.0543, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012145748987854251, | |
| "grad_norm": 6.515895120322345, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 1.0617, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016194331983805668, | |
| "grad_norm": 6.19295838461633, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 1.0139, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020242914979757085, | |
| "grad_norm": 6.137447167839709, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 1.0375, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024291497975708502, | |
| "grad_norm": 5.846159308241376, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.0335, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02834008097165992, | |
| "grad_norm": 4.728729020577783, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 1.0135, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032388663967611336, | |
| "grad_norm": 4.236510770438479, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 0.9822, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03643724696356275, | |
| "grad_norm": 2.802384325971826, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 0.9892, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04048582995951417, | |
| "grad_norm": 2.633668574950178, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.9511, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044534412955465584, | |
| "grad_norm": 2.529537487624366, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 0.9484, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048582995951417005, | |
| "grad_norm": 3.743919273766966, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.9432, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 4.085177917901176, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 0.9394, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05668016194331984, | |
| "grad_norm": 3.9215870276440605, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.9457, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06072874493927125, | |
| "grad_norm": 3.5326113491038735, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 0.9554, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06477732793522267, | |
| "grad_norm": 2.597735909957419, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 0.8888, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06882591093117409, | |
| "grad_norm": 2.4329759093724346, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 0.8913, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0728744939271255, | |
| "grad_norm": 2.006831249329057, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.8712, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.3944756732793868, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 0.8386, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08097165991902834, | |
| "grad_norm": 1.0835710728587022, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.843, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08502024291497975, | |
| "grad_norm": 1.0201271328722668, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.8109, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08906882591093117, | |
| "grad_norm": 1.117305930907855, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.7953, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0931174089068826, | |
| "grad_norm": 1.0445605823630433, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.8207, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09716599190283401, | |
| "grad_norm": 0.9149129442769901, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.8037, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10121457489878542, | |
| "grad_norm": 0.9100306882028654, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.783, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 0.8294913234414353, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.7691, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10931174089068826, | |
| "grad_norm": 0.8121647636702842, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.783, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11336032388663968, | |
| "grad_norm": 0.7809449096995666, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.7723, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11740890688259109, | |
| "grad_norm": 0.8335357436015843, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.7537, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1214574898785425, | |
| "grad_norm": 0.7986755062616927, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.7475, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12550607287449392, | |
| "grad_norm": 0.7424475971042445, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7548, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12955465587044535, | |
| "grad_norm": 0.6711306609484193, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.7434, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13360323886639677, | |
| "grad_norm": 0.6198215890298625, | |
| "learning_rate": 1.0645161290322582e-05, | |
| "loss": 0.7284, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13765182186234817, | |
| "grad_norm": 0.6292873861689773, | |
| "learning_rate": 1.096774193548387e-05, | |
| "loss": 0.7191, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1417004048582996, | |
| "grad_norm": 0.573138434614111, | |
| "learning_rate": 1.1290322580645164e-05, | |
| "loss": 0.7445, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.145748987854251, | |
| "grad_norm": 0.6222655247198949, | |
| "learning_rate": 1.1612903225806453e-05, | |
| "loss": 0.7277, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14979757085020243, | |
| "grad_norm": 0.6685492385113305, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.7193, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.6271310615806173, | |
| "learning_rate": 1.2258064516129034e-05, | |
| "loss": 0.7129, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 0.5568137339781766, | |
| "learning_rate": 1.2580645161290324e-05, | |
| "loss": 0.717, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16194331983805668, | |
| "grad_norm": 0.6124574961197972, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.7113, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1659919028340081, | |
| "grad_norm": 0.6648196419705169, | |
| "learning_rate": 1.3225806451612903e-05, | |
| "loss": 0.7235, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1700404858299595, | |
| "grad_norm": 0.5451196287322313, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.7087, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17408906882591094, | |
| "grad_norm": 0.6239631869414103, | |
| "learning_rate": 1.3870967741935486e-05, | |
| "loss": 0.7164, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17813765182186234, | |
| "grad_norm": 0.48508430395561464, | |
| "learning_rate": 1.4193548387096776e-05, | |
| "loss": 0.6975, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18218623481781376, | |
| "grad_norm": 0.6205896903060076, | |
| "learning_rate": 1.4516129032258066e-05, | |
| "loss": 0.7137, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1862348178137652, | |
| "grad_norm": 0.46782624556430985, | |
| "learning_rate": 1.4838709677419357e-05, | |
| "loss": 0.7237, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1902834008097166, | |
| "grad_norm": 0.4743434226370183, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.6842, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19433198380566802, | |
| "grad_norm": 0.48559141222239616, | |
| "learning_rate": 1.5483870967741936e-05, | |
| "loss": 0.6909, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19838056680161945, | |
| "grad_norm": 0.4520917955107327, | |
| "learning_rate": 1.5806451612903226e-05, | |
| "loss": 0.6913, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20242914979757085, | |
| "grad_norm": 0.48747392965489245, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.7164, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20647773279352227, | |
| "grad_norm": 0.48625752142540163, | |
| "learning_rate": 1.6451612903225807e-05, | |
| "loss": 0.6772, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.5571652681916701, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.7109, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2145748987854251, | |
| "grad_norm": 0.5666389578604575, | |
| "learning_rate": 1.7096774193548388e-05, | |
| "loss": 0.7024, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21862348178137653, | |
| "grad_norm": 0.5346372642627782, | |
| "learning_rate": 1.741935483870968e-05, | |
| "loss": 0.6903, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22267206477732793, | |
| "grad_norm": 0.5594087220179628, | |
| "learning_rate": 1.774193548387097e-05, | |
| "loss": 0.6802, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22672064777327935, | |
| "grad_norm": 0.6454053476126799, | |
| "learning_rate": 1.806451612903226e-05, | |
| "loss": 0.6849, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.5474185647383353, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.6776, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23481781376518218, | |
| "grad_norm": 0.6920190231303881, | |
| "learning_rate": 1.870967741935484e-05, | |
| "loss": 0.6804, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2388663967611336, | |
| "grad_norm": 0.8798368043500211, | |
| "learning_rate": 1.903225806451613e-05, | |
| "loss": 0.6765, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.242914979757085, | |
| "grad_norm": 0.6071754760753234, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.6663, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24696356275303644, | |
| "grad_norm": 0.7305267293785266, | |
| "learning_rate": 1.967741935483871e-05, | |
| "loss": 0.6751, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.25101214574898784, | |
| "grad_norm": 0.6558255239692878, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6792, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2550607287449393, | |
| "grad_norm": 0.6933334304145264, | |
| "learning_rate": 2.0322580645161292e-05, | |
| "loss": 0.6588, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2591093117408907, | |
| "grad_norm": 0.5404597592795586, | |
| "learning_rate": 2.0645161290322582e-05, | |
| "loss": 0.675, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 0.6411338688209707, | |
| "learning_rate": 2.0967741935483873e-05, | |
| "loss": 0.6772, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26720647773279355, | |
| "grad_norm": 0.558805147337778, | |
| "learning_rate": 2.1290322580645163e-05, | |
| "loss": 0.6606, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27125506072874495, | |
| "grad_norm": 0.5914611844911203, | |
| "learning_rate": 2.161290322580645e-05, | |
| "loss": 0.6811, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27530364372469635, | |
| "grad_norm": 0.531794331335956, | |
| "learning_rate": 2.193548387096774e-05, | |
| "loss": 0.6621, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2793522267206478, | |
| "grad_norm": 0.48875146814954595, | |
| "learning_rate": 2.225806451612903e-05, | |
| "loss": 0.6604, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2834008097165992, | |
| "grad_norm": 0.6122059225719468, | |
| "learning_rate": 2.2580645161290328e-05, | |
| "loss": 0.6741, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2874493927125506, | |
| "grad_norm": 0.665083976670287, | |
| "learning_rate": 2.2903225806451615e-05, | |
| "loss": 0.6767, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.291497975708502, | |
| "grad_norm": 0.7917216757752762, | |
| "learning_rate": 2.3225806451612906e-05, | |
| "loss": 0.6607, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.29554655870445345, | |
| "grad_norm": 0.9756143499434393, | |
| "learning_rate": 2.3548387096774196e-05, | |
| "loss": 0.664, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29959514170040485, | |
| "grad_norm": 1.4264383720203815, | |
| "learning_rate": 2.3870967741935486e-05, | |
| "loss": 0.6805, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30364372469635625, | |
| "grad_norm": 0.6012651355664433, | |
| "learning_rate": 2.4193548387096777e-05, | |
| "loss": 0.659, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.0958619192095234, | |
| "learning_rate": 2.4516129032258067e-05, | |
| "loss": 0.666, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3117408906882591, | |
| "grad_norm": 1.4721290468740154, | |
| "learning_rate": 2.4838709677419358e-05, | |
| "loss": 0.6676, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 0.7484697578039731, | |
| "learning_rate": 2.5161290322580648e-05, | |
| "loss": 0.6475, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31983805668016196, | |
| "grad_norm": 1.6750579210157144, | |
| "learning_rate": 2.548387096774194e-05, | |
| "loss": 0.6677, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32388663967611336, | |
| "grad_norm": 0.8583135017377609, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.6684, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32793522267206476, | |
| "grad_norm": 1.477712115499635, | |
| "learning_rate": 2.6129032258064516e-05, | |
| "loss": 0.6443, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3319838056680162, | |
| "grad_norm": 1.0213086492686927, | |
| "learning_rate": 2.6451612903225806e-05, | |
| "loss": 0.6759, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3360323886639676, | |
| "grad_norm": 1.9484258582582834, | |
| "learning_rate": 2.6774193548387097e-05, | |
| "loss": 0.6505, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.340080971659919, | |
| "grad_norm": 1.7943892198238425, | |
| "learning_rate": 2.7096774193548387e-05, | |
| "loss": 0.6631, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3441295546558704, | |
| "grad_norm": 1.2040221069135388, | |
| "learning_rate": 2.7419354838709678e-05, | |
| "loss": 0.6613, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3481781376518219, | |
| "grad_norm": 1.1367017251427853, | |
| "learning_rate": 2.774193548387097e-05, | |
| "loss": 0.6694, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3522267206477733, | |
| "grad_norm": 1.200473266926972, | |
| "learning_rate": 2.8064516129032262e-05, | |
| "loss": 0.6524, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3562753036437247, | |
| "grad_norm": 1.0695421798033862, | |
| "learning_rate": 2.8387096774193552e-05, | |
| "loss": 0.6421, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3603238866396761, | |
| "grad_norm": 1.2773845821098537, | |
| "learning_rate": 2.8709677419354843e-05, | |
| "loss": 0.6542, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3643724696356275, | |
| "grad_norm": 1.158823905653716, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.6538, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 1.692004809009881, | |
| "learning_rate": 2.9354838709677423e-05, | |
| "loss": 0.6452, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3724696356275304, | |
| "grad_norm": 1.0716738885999277, | |
| "learning_rate": 2.9677419354838714e-05, | |
| "loss": 0.6622, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3765182186234818, | |
| "grad_norm": 1.8745762768496204, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.6496, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3805668016194332, | |
| "grad_norm": 1.5190937642178448, | |
| "learning_rate": 3.032258064516129e-05, | |
| "loss": 0.6388, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 1.4736426758765353, | |
| "learning_rate": 3.0645161290322585e-05, | |
| "loss": 0.6806, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 1.396553731694665, | |
| "learning_rate": 3.096774193548387e-05, | |
| "loss": 0.666, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39271255060728744, | |
| "grad_norm": 1.0361758030885917, | |
| "learning_rate": 3.1290322580645166e-05, | |
| "loss": 0.6566, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3967611336032389, | |
| "grad_norm": 1.1415274797141521, | |
| "learning_rate": 3.161290322580645e-05, | |
| "loss": 0.6664, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4008097165991903, | |
| "grad_norm": 1.3217313848383427, | |
| "learning_rate": 3.193548387096774e-05, | |
| "loss": 0.6549, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4048582995951417, | |
| "grad_norm": 1.1538225887305784, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.646, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4089068825910931, | |
| "grad_norm": 1.1171312580085944, | |
| "learning_rate": 3.258064516129033e-05, | |
| "loss": 0.6377, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.41295546558704455, | |
| "grad_norm": 0.8576589044181462, | |
| "learning_rate": 3.2903225806451614e-05, | |
| "loss": 0.6363, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.41700404858299595, | |
| "grad_norm": 1.5816405563528901, | |
| "learning_rate": 3.322580645161291e-05, | |
| "loss": 0.6294, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.9496614468176897, | |
| "learning_rate": 3.3548387096774195e-05, | |
| "loss": 0.6737, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4251012145748988, | |
| "grad_norm": 1.0854328755291738, | |
| "learning_rate": 3.387096774193549e-05, | |
| "loss": 0.6429, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4291497975708502, | |
| "grad_norm": 0.9852556840907373, | |
| "learning_rate": 3.4193548387096776e-05, | |
| "loss": 0.6625, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4331983805668016, | |
| "grad_norm": 1.3671876992278906, | |
| "learning_rate": 3.451612903225807e-05, | |
| "loss": 0.6649, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.43724696356275305, | |
| "grad_norm": 0.8267458859663657, | |
| "learning_rate": 3.483870967741936e-05, | |
| "loss": 0.6466, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.44129554655870445, | |
| "grad_norm": 1.3336520322371557, | |
| "learning_rate": 3.516129032258065e-05, | |
| "loss": 0.6744, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44534412955465585, | |
| "grad_norm": 0.8242234034126853, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.658, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4493927125506073, | |
| "grad_norm": 1.1635699633816798, | |
| "learning_rate": 3.5806451612903225e-05, | |
| "loss": 0.6443, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4534412955465587, | |
| "grad_norm": 0.7854569522818275, | |
| "learning_rate": 3.612903225806452e-05, | |
| "loss": 0.6601, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4574898785425101, | |
| "grad_norm": 1.2467922183639621, | |
| "learning_rate": 3.6451612903225805e-05, | |
| "loss": 0.6452, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.9903943379424068, | |
| "learning_rate": 3.67741935483871e-05, | |
| "loss": 0.6487, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46558704453441296, | |
| "grad_norm": 1.1011039444222088, | |
| "learning_rate": 3.7096774193548386e-05, | |
| "loss": 0.6626, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.46963562753036436, | |
| "grad_norm": 1.2470891209172554, | |
| "learning_rate": 3.741935483870968e-05, | |
| "loss": 0.6558, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 1.1738043085956735, | |
| "learning_rate": 3.7741935483870974e-05, | |
| "loss": 0.6547, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4777327935222672, | |
| "grad_norm": 1.0311169260421873, | |
| "learning_rate": 3.806451612903226e-05, | |
| "loss": 0.6667, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4817813765182186, | |
| "grad_norm": 0.9300031336625245, | |
| "learning_rate": 3.8387096774193555e-05, | |
| "loss": 0.6406, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48582995951417, | |
| "grad_norm": 0.9078227932508798, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.6429, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4898785425101215, | |
| "grad_norm": 1.1113849832364322, | |
| "learning_rate": 3.9032258064516135e-05, | |
| "loss": 0.6382, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4939271255060729, | |
| "grad_norm": 1.102247515725923, | |
| "learning_rate": 3.935483870967742e-05, | |
| "loss": 0.6526, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4979757085020243, | |
| "grad_norm": 1.340049090504809, | |
| "learning_rate": 3.9677419354838716e-05, | |
| "loss": 0.6481, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5020242914979757, | |
| "grad_norm": 1.2230207668330755, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6473, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5060728744939271, | |
| "grad_norm": 1.2486074747406835, | |
| "learning_rate": 3.9999920040266474e-05, | |
| "loss": 0.6304, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5101214574898786, | |
| "grad_norm": 1.1221126565176316, | |
| "learning_rate": 3.999968016170525e-05, | |
| "loss": 0.6459, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5141700404858299, | |
| "grad_norm": 1.427448393597703, | |
| "learning_rate": 3.9999280366234394e-05, | |
| "loss": 0.6332, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5182186234817814, | |
| "grad_norm": 0.9016423402291581, | |
| "learning_rate": 3.999872065705065e-05, | |
| "loss": 0.6262, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5222672064777328, | |
| "grad_norm": 1.2245668878182911, | |
| "learning_rate": 3.999800103862945e-05, | |
| "loss": 0.6424, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.237490413448666, | |
| "learning_rate": 3.9997121516724834e-05, | |
| "loss": 0.6532, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5303643724696356, | |
| "grad_norm": 1.045533702334215, | |
| "learning_rate": 3.999608209836944e-05, | |
| "loss": 0.6408, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5344129554655871, | |
| "grad_norm": 1.2434795961907634, | |
| "learning_rate": 3.999488279187443e-05, | |
| "loss": 0.6471, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.0848314344964733, | |
| "learning_rate": 3.9993523606829425e-05, | |
| "loss": 0.6324, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5425101214574899, | |
| "grad_norm": 1.092362143363375, | |
| "learning_rate": 3.9992004554102435e-05, | |
| "loss": 0.6407, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5465587044534413, | |
| "grad_norm": 0.826840739387905, | |
| "learning_rate": 3.999032564583976e-05, | |
| "loss": 0.6378, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5506072874493927, | |
| "grad_norm": 0.8814342908479853, | |
| "learning_rate": 3.998848689546591e-05, | |
| "loss": 0.6368, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5546558704453441, | |
| "grad_norm": 0.9006833137416743, | |
| "learning_rate": 3.9986488317683496e-05, | |
| "loss": 0.6462, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5587044534412956, | |
| "grad_norm": 0.9112412490678907, | |
| "learning_rate": 3.9984329928473067e-05, | |
| "loss": 0.6469, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.562753036437247, | |
| "grad_norm": 0.9370196379001619, | |
| "learning_rate": 3.998201174509306e-05, | |
| "loss": 0.6421, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5668016194331984, | |
| "grad_norm": 0.9572342058834716, | |
| "learning_rate": 3.997953378607961e-05, | |
| "loss": 0.6272, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5708502024291497, | |
| "grad_norm": 0.9944210295047771, | |
| "learning_rate": 3.9976896071246403e-05, | |
| "loss": 0.6414, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5748987854251012, | |
| "grad_norm": 0.7855656438674631, | |
| "learning_rate": 3.997409862168454e-05, | |
| "loss": 0.6418, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.7991357906130065, | |
| "learning_rate": 3.997114145976237e-05, | |
| "loss": 0.6537, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.582995951417004, | |
| "grad_norm": 0.9570306473017499, | |
| "learning_rate": 3.9968024609125245e-05, | |
| "loss": 0.6283, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5870445344129555, | |
| "grad_norm": 0.8791316236703429, | |
| "learning_rate": 3.996474809469545e-05, | |
| "loss": 0.6574, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5910931174089069, | |
| "grad_norm": 0.7689439720720929, | |
| "learning_rate": 3.996131194267189e-05, | |
| "loss": 0.6419, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5951417004048583, | |
| "grad_norm": 1.1250962384863523, | |
| "learning_rate": 3.995771618052995e-05, | |
| "loss": 0.6435, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5991902834008097, | |
| "grad_norm": 0.9370794818602229, | |
| "learning_rate": 3.9953960837021256e-05, | |
| "loss": 0.6365, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6032388663967612, | |
| "grad_norm": 0.6888314745652535, | |
| "learning_rate": 3.995004594217343e-05, | |
| "loss": 0.647, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6072874493927125, | |
| "grad_norm": 0.8312207551279359, | |
| "learning_rate": 3.994597152728987e-05, | |
| "loss": 0.6333, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.611336032388664, | |
| "grad_norm": 0.7354358454221117, | |
| "learning_rate": 3.994173762494947e-05, | |
| "loss": 0.6326, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.643480775243013, | |
| "learning_rate": 3.9937344269006424e-05, | |
| "loss": 0.6396, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6194331983805668, | |
| "grad_norm": 0.7535528830238153, | |
| "learning_rate": 3.993279149458988e-05, | |
| "loss": 0.6596, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6234817813765182, | |
| "grad_norm": 0.5437549785312074, | |
| "learning_rate": 3.9928079338103695e-05, | |
| "loss": 0.6438, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6275303643724697, | |
| "grad_norm": 0.7244530547518417, | |
| "learning_rate": 3.992320783722616e-05, | |
| "loss": 0.6373, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.9929118187008792, | |
| "learning_rate": 3.991817703090965e-05, | |
| "loss": 0.6346, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6356275303643725, | |
| "grad_norm": 0.9124228948570662, | |
| "learning_rate": 3.991298695938038e-05, | |
| "loss": 0.6376, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6396761133603239, | |
| "grad_norm": 0.7552832215945303, | |
| "learning_rate": 3.9907637664138e-05, | |
| "loss": 0.6294, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6437246963562753, | |
| "grad_norm": 0.6384745577725681, | |
| "learning_rate": 3.990212918795534e-05, | |
| "loss": 0.653, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6477732793522267, | |
| "grad_norm": 0.7615614073066913, | |
| "learning_rate": 3.9896461574878036e-05, | |
| "loss": 0.6386, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6518218623481782, | |
| "grad_norm": 1.331212525704376, | |
| "learning_rate": 3.989063487022416e-05, | |
| "loss": 0.6793, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6558704453441295, | |
| "grad_norm": 0.6361091051036739, | |
| "learning_rate": 3.9884649120583903e-05, | |
| "loss": 0.6283, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.659919028340081, | |
| "grad_norm": 0.720300022555699, | |
| "learning_rate": 3.987850437381915e-05, | |
| "loss": 0.61, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6639676113360324, | |
| "grad_norm": 0.80320600350198, | |
| "learning_rate": 3.987220067906313e-05, | |
| "loss": 0.6279, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6680161943319838, | |
| "grad_norm": 0.8500023111703493, | |
| "learning_rate": 3.986573808672002e-05, | |
| "loss": 0.6303, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6720647773279352, | |
| "grad_norm": 0.9401055608384464, | |
| "learning_rate": 3.9859116648464534e-05, | |
| "loss": 0.6256, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6761133603238867, | |
| "grad_norm": 0.9899532549912121, | |
| "learning_rate": 3.985233641724152e-05, | |
| "loss": 0.6372, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.680161943319838, | |
| "grad_norm": 0.7763544249966322, | |
| "learning_rate": 3.984539744726554e-05, | |
| "loss": 0.6399, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.6931847372157176, | |
| "learning_rate": 3.983829979402038e-05, | |
| "loss": 0.6354, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6882591093117408, | |
| "grad_norm": 0.8503929343244705, | |
| "learning_rate": 3.983104351425871e-05, | |
| "loss": 0.643, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.8600136753830666, | |
| "learning_rate": 3.982362866600154e-05, | |
| "loss": 0.6313, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6963562753036437, | |
| "grad_norm": 0.7095113929958955, | |
| "learning_rate": 3.9816055308537806e-05, | |
| "loss": 0.6192, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7004048582995951, | |
| "grad_norm": 1.1839408806799325, | |
| "learning_rate": 3.980832350242388e-05, | |
| "loss": 0.6543, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7044534412955465, | |
| "grad_norm": 0.6754932882188525, | |
| "learning_rate": 3.980043330948306e-05, | |
| "loss": 0.622, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708502024291498, | |
| "grad_norm": 0.974087034074319, | |
| "learning_rate": 3.9792384792805124e-05, | |
| "loss": 0.6378, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7125506072874493, | |
| "grad_norm": 0.9700983823890822, | |
| "learning_rate": 3.978417801674579e-05, | |
| "loss": 0.6401, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7165991902834008, | |
| "grad_norm": 0.6634322973185122, | |
| "learning_rate": 3.9775813046926244e-05, | |
| "loss": 0.6378, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7206477732793523, | |
| "grad_norm": 0.6318366546205781, | |
| "learning_rate": 3.976728995023254e-05, | |
| "loss": 0.6163, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7246963562753036, | |
| "grad_norm": 0.9778354975905699, | |
| "learning_rate": 3.975860879481514e-05, | |
| "loss": 0.6335, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.728744939271255, | |
| "grad_norm": 0.730278511988347, | |
| "learning_rate": 3.974976965008833e-05, | |
| "loss": 0.6297, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7327935222672065, | |
| "grad_norm": 0.5934711333692605, | |
| "learning_rate": 3.9740772586729674e-05, | |
| "loss": 0.6399, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.742366821273482, | |
| "learning_rate": 3.973161767667946e-05, | |
| "loss": 0.628, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7408906882591093, | |
| "grad_norm": 0.7707668499065367, | |
| "learning_rate": 3.972230499314009e-05, | |
| "loss": 0.6332, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7449392712550608, | |
| "grad_norm": 0.5914421575524721, | |
| "learning_rate": 3.9712834610575545e-05, | |
| "loss": 0.6344, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7489878542510121, | |
| "grad_norm": 0.823293370994319, | |
| "learning_rate": 3.9703206604710746e-05, | |
| "loss": 0.627, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7530364372469636, | |
| "grad_norm": 0.8520423873828534, | |
| "learning_rate": 3.969342105253098e-05, | |
| "loss": 0.6407, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.757085020242915, | |
| "grad_norm": 0.5895428519121637, | |
| "learning_rate": 3.9683478032281246e-05, | |
| "loss": 0.628, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7611336032388664, | |
| "grad_norm": 0.7168056197156862, | |
| "learning_rate": 3.967337762346568e-05, | |
| "loss": 0.6274, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7651821862348178, | |
| "grad_norm": 0.8436725093432914, | |
| "learning_rate": 3.9663119906846885e-05, | |
| "loss": 0.6297, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.5256821318942483, | |
| "learning_rate": 3.965270496444528e-05, | |
| "loss": 0.6133, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7732793522267206, | |
| "grad_norm": 0.5234662507183212, | |
| "learning_rate": 3.9642132879538465e-05, | |
| "loss": 0.6183, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.8337532725110053, | |
| "learning_rate": 3.9631403736660566e-05, | |
| "loss": 0.645, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7813765182186235, | |
| "grad_norm": 0.7668720841595759, | |
| "learning_rate": 3.962051762160151e-05, | |
| "loss": 0.6266, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7854251012145749, | |
| "grad_norm": 0.5140229134379459, | |
| "learning_rate": 3.960947462140638e-05, | |
| "loss": 0.6361, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.6914883006037478, | |
| "learning_rate": 3.959827482437472e-05, | |
| "loss": 0.6248, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7935222672064778, | |
| "grad_norm": 0.6932979550320546, | |
| "learning_rate": 3.958691832005981e-05, | |
| "loss": 0.6281, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7975708502024291, | |
| "grad_norm": 0.6184541495297794, | |
| "learning_rate": 3.9575405199267944e-05, | |
| "loss": 0.626, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8016194331983806, | |
| "grad_norm": 0.706892556005753, | |
| "learning_rate": 3.956373555405774e-05, | |
| "loss": 0.6352, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.805668016194332, | |
| "grad_norm": 0.6449175667235093, | |
| "learning_rate": 3.955190947773937e-05, | |
| "loss": 0.6193, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8097165991902834, | |
| "grad_norm": 0.5554864487862031, | |
| "learning_rate": 3.9539927064873824e-05, | |
| "loss": 0.6264, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8137651821862348, | |
| "grad_norm": 0.6747157481735445, | |
| "learning_rate": 3.952778841127214e-05, | |
| "loss": 0.632, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8178137651821862, | |
| "grad_norm": 0.5502629713846755, | |
| "learning_rate": 3.95154936139947e-05, | |
| "loss": 0.6087, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8218623481781376, | |
| "grad_norm": 0.5559193355553335, | |
| "learning_rate": 3.950304277135034e-05, | |
| "loss": 0.6159, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8259109311740891, | |
| "grad_norm": 0.6442470233046814, | |
| "learning_rate": 3.94904359828957e-05, | |
| "loss": 0.6164, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8299595141700404, | |
| "grad_norm": 0.5133438805237442, | |
| "learning_rate": 3.94776733494343e-05, | |
| "loss": 0.6283, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8340080971659919, | |
| "grad_norm": 0.586494474661796, | |
| "learning_rate": 3.9464754973015826e-05, | |
| "loss": 0.6097, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8380566801619433, | |
| "grad_norm": 0.4868877900682267, | |
| "learning_rate": 3.9451680956935275e-05, | |
| "loss": 0.6083, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.5347783261214675, | |
| "learning_rate": 3.943845140573213e-05, | |
| "loss": 0.6219, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.5616924351528346, | |
| "learning_rate": 3.942506642518952e-05, | |
| "loss": 0.622, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8502024291497976, | |
| "grad_norm": 0.528218773624546, | |
| "learning_rate": 3.9411526122333396e-05, | |
| "loss": 0.6329, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.854251012145749, | |
| "grad_norm": 0.5743524414100961, | |
| "learning_rate": 3.939783060543167e-05, | |
| "loss": 0.624, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8582995951417004, | |
| "grad_norm": 0.6034438935608605, | |
| "learning_rate": 3.938397998399333e-05, | |
| "loss": 0.6415, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8623481781376519, | |
| "grad_norm": 0.4712640724176737, | |
| "learning_rate": 3.936997436876756e-05, | |
| "loss": 0.6093, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8663967611336032, | |
| "grad_norm": 0.5249185388079481, | |
| "learning_rate": 3.935581387174289e-05, | |
| "loss": 0.6217, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8704453441295547, | |
| "grad_norm": 0.6056040223597903, | |
| "learning_rate": 3.9341498606146295e-05, | |
| "loss": 0.6261, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8744939271255061, | |
| "grad_norm": 0.5547110537089253, | |
| "learning_rate": 3.932702868644224e-05, | |
| "loss": 0.632, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8785425101214575, | |
| "grad_norm": 0.5339950769631198, | |
| "learning_rate": 3.931240422833181e-05, | |
| "loss": 0.6228, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8825910931174089, | |
| "grad_norm": 0.6213199864112682, | |
| "learning_rate": 3.9297625348751806e-05, | |
| "loss": 0.6166, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8866396761133604, | |
| "grad_norm": 0.5146668147139875, | |
| "learning_rate": 3.9282692165873744e-05, | |
| "loss": 0.6075, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8906882591093117, | |
| "grad_norm": 0.5032615648727539, | |
| "learning_rate": 3.9267604799102945e-05, | |
| "loss": 0.6179, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.4666790172818089, | |
| "learning_rate": 3.9252363369077615e-05, | |
| "loss": 0.6152, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8987854251012146, | |
| "grad_norm": 0.4620513460954615, | |
| "learning_rate": 3.9236967997667796e-05, | |
| "loss": 0.5941, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.902834008097166, | |
| "grad_norm": 0.5053219845262146, | |
| "learning_rate": 3.9221418807974496e-05, | |
| "loss": 0.6097, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9068825910931174, | |
| "grad_norm": 0.5831098187979585, | |
| "learning_rate": 3.92057159243286e-05, | |
| "loss": 0.6091, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9109311740890689, | |
| "grad_norm": 0.6149034850656219, | |
| "learning_rate": 3.9189859472289956e-05, | |
| "loss": 0.6256, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9149797570850202, | |
| "grad_norm": 0.6048522536592987, | |
| "learning_rate": 3.9173849578646324e-05, | |
| "loss": 0.6135, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9190283400809717, | |
| "grad_norm": 0.6948003448726063, | |
| "learning_rate": 3.9157686371412395e-05, | |
| "loss": 0.611, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.680314717156262, | |
| "learning_rate": 3.914136997982874e-05, | |
| "loss": 0.6265, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9271255060728745, | |
| "grad_norm": 0.5282090400660675, | |
| "learning_rate": 3.912490053436079e-05, | |
| "loss": 0.6174, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9311740890688259, | |
| "grad_norm": 0.6589769589162725, | |
| "learning_rate": 3.9108278166697805e-05, | |
| "loss": 0.6077, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9352226720647774, | |
| "grad_norm": 0.6279958438843974, | |
| "learning_rate": 3.909150300975178e-05, | |
| "loss": 0.6287, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9392712550607287, | |
| "grad_norm": 0.5396307483815266, | |
| "learning_rate": 3.907457519765642e-05, | |
| "loss": 0.6254, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9433198380566802, | |
| "grad_norm": 0.6184847456222738, | |
| "learning_rate": 3.905749486576607e-05, | |
| "loss": 0.6172, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.6310858104751806, | |
| "learning_rate": 3.90402621506546e-05, | |
| "loss": 0.6116, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.951417004048583, | |
| "grad_norm": 0.49812644188765937, | |
| "learning_rate": 3.902287719011435e-05, | |
| "loss": 0.6221, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9554655870445344, | |
| "grad_norm": 1.4161546031853591, | |
| "learning_rate": 3.900534012315499e-05, | |
| "loss": 0.6772, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9595141700404858, | |
| "grad_norm": 0.5196617540389556, | |
| "learning_rate": 3.8987651090002454e-05, | |
| "loss": 0.6427, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9635627530364372, | |
| "grad_norm": 0.7316234721249052, | |
| "learning_rate": 3.896981023209777e-05, | |
| "loss": 0.6254, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9676113360323887, | |
| "grad_norm": 0.4893688053547945, | |
| "learning_rate": 3.8951817692095966e-05, | |
| "loss": 0.6078, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.97165991902834, | |
| "grad_norm": 0.6803685597656999, | |
| "learning_rate": 3.893367361386491e-05, | |
| "loss": 0.6292, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9757085020242915, | |
| "grad_norm": 0.6387869544825641, | |
| "learning_rate": 3.891537814248417e-05, | |
| "loss": 0.6224, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.979757085020243, | |
| "grad_norm": 0.6169006563155737, | |
| "learning_rate": 3.889693142424384e-05, | |
| "loss": 0.6123, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9838056680161943, | |
| "grad_norm": 0.5140463181385796, | |
| "learning_rate": 3.8878333606643405e-05, | |
| "loss": 0.6155, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9878542510121457, | |
| "grad_norm": 0.7466039014798256, | |
| "learning_rate": 3.885958483839049e-05, | |
| "loss": 0.6094, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9919028340080972, | |
| "grad_norm": 0.6205041007785772, | |
| "learning_rate": 3.8840685269399786e-05, | |
| "loss": 0.6214, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9959514170040485, | |
| "grad_norm": 0.5570441860603544, | |
| "learning_rate": 3.882163505079171e-05, | |
| "loss": 0.6468, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7757806415291847, | |
| "learning_rate": 3.880243433489132e-05, | |
| "loss": 0.6216, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0040485829959513, | |
| "grad_norm": 0.4590593925345867, | |
| "learning_rate": 3.878308327522702e-05, | |
| "loss": 0.5648, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.008097165991903, | |
| "grad_norm": 0.6521724088544122, | |
| "learning_rate": 3.876358202652936e-05, | |
| "loss": 0.5532, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0121457489878543, | |
| "grad_norm": 0.9144413534659048, | |
| "learning_rate": 3.8743930744729835e-05, | |
| "loss": 0.5728, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0161943319838056, | |
| "grad_norm": 0.8160121685045223, | |
| "learning_rate": 3.8724129586959535e-05, | |
| "loss": 0.5667, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0202429149797572, | |
| "grad_norm": 0.610809156172711, | |
| "learning_rate": 3.870417871154801e-05, | |
| "loss": 0.5657, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0242914979757085, | |
| "grad_norm": 0.45428207545847904, | |
| "learning_rate": 3.868407827802192e-05, | |
| "loss": 0.5492, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0283400809716599, | |
| "grad_norm": 0.5675590809452851, | |
| "learning_rate": 3.8663828447103804e-05, | |
| "loss": 0.5591, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0323886639676114, | |
| "grad_norm": 0.579086725857133, | |
| "learning_rate": 3.8643429380710765e-05, | |
| "loss": 0.5655, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0364372469635628, | |
| "grad_norm": 0.5612159808712044, | |
| "learning_rate": 3.862288124195319e-05, | |
| "loss": 0.5643, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.040485829959514, | |
| "grad_norm": 0.7988220247401013, | |
| "learning_rate": 3.8602184195133464e-05, | |
| "loss": 0.5733, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0445344129554657, | |
| "grad_norm": 0.5753264591785938, | |
| "learning_rate": 3.8581338405744604e-05, | |
| "loss": 0.5676, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.048582995951417, | |
| "grad_norm": 0.6860478309661131, | |
| "learning_rate": 3.8560344040468996e-05, | |
| "loss": 0.5655, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.8021601344559687, | |
| "learning_rate": 3.853920126717703e-05, | |
| "loss": 0.5697, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.05668016194332, | |
| "grad_norm": 0.4550205115857696, | |
| "learning_rate": 3.851791025492574e-05, | |
| "loss": 0.5634, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0607287449392713, | |
| "grad_norm": 0.6487779553499391, | |
| "learning_rate": 3.8496471173957515e-05, | |
| "loss": 0.5503, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0647773279352226, | |
| "grad_norm": 0.5140457228901137, | |
| "learning_rate": 3.847488419569866e-05, | |
| "loss": 0.5632, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0688259109311742, | |
| "grad_norm": 0.5104057162095357, | |
| "learning_rate": 3.8453149492758077e-05, | |
| "loss": 0.5658, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0728744939271255, | |
| "grad_norm": 0.5393191891118115, | |
| "learning_rate": 3.8431267238925886e-05, | |
| "loss": 0.5676, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.505098439937945, | |
| "learning_rate": 3.840923760917199e-05, | |
| "loss": 0.577, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0809716599190284, | |
| "grad_norm": 0.6834513243834368, | |
| "learning_rate": 3.838706077964473e-05, | |
| "loss": 0.5596, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0850202429149798, | |
| "grad_norm": 0.8958487946850145, | |
| "learning_rate": 3.836473692766944e-05, | |
| "loss": 0.5597, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0890688259109311, | |
| "grad_norm": 0.8125529612647241, | |
| "learning_rate": 3.834226623174705e-05, | |
| "loss": 0.561, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0931174089068827, | |
| "grad_norm": 0.4861046757963102, | |
| "learning_rate": 3.831964887155264e-05, | |
| "loss": 0.5699, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.097165991902834, | |
| "grad_norm": 0.8244959082768244, | |
| "learning_rate": 3.829688502793403e-05, | |
| "loss": 0.5706, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.1012145748987854, | |
| "grad_norm": 0.7374363017872935, | |
| "learning_rate": 3.827397488291029e-05, | |
| "loss": 0.5591, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.49262883409089425, | |
| "learning_rate": 3.8250918619670343e-05, | |
| "loss": 0.5464, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.1093117408906883, | |
| "grad_norm": 0.5418701500592275, | |
| "learning_rate": 3.822771642257145e-05, | |
| "loss": 0.5594, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1133603238866396, | |
| "grad_norm": 0.4636698730142695, | |
| "learning_rate": 3.820436847713776e-05, | |
| "loss": 0.5719, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.117408906882591, | |
| "grad_norm": 0.40463164218456366, | |
| "learning_rate": 3.8180874970058826e-05, | |
| "loss": 0.5636, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.1214574898785425, | |
| "grad_norm": 0.4266578930025979, | |
| "learning_rate": 3.8157236089188103e-05, | |
| "loss": 0.554, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.125506072874494, | |
| "grad_norm": 0.5186066337395122, | |
| "learning_rate": 3.813345202354145e-05, | |
| "loss": 0.555, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1295546558704452, | |
| "grad_norm": 0.42217711856822937, | |
| "learning_rate": 3.810952296329563e-05, | |
| "loss": 0.5529, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1336032388663968, | |
| "grad_norm": 0.44027903341196645, | |
| "learning_rate": 3.808544909978676e-05, | |
| "loss": 0.5557, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1376518218623481, | |
| "grad_norm": 0.47956664588249703, | |
| "learning_rate": 3.806123062550882e-05, | |
| "loss": 0.572, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1417004048582995, | |
| "grad_norm": 0.45141145209185696, | |
| "learning_rate": 3.803686773411208e-05, | |
| "loss": 0.5496, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.145748987854251, | |
| "grad_norm": 0.4977476478240117, | |
| "learning_rate": 3.801236062040158e-05, | |
| "loss": 0.5479, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1497975708502024, | |
| "grad_norm": 0.45216953828146983, | |
| "learning_rate": 3.798770948033553e-05, | |
| "loss": 0.5566, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.4775853287712311, | |
| "learning_rate": 3.796291451102381e-05, | |
| "loss": 0.5562, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.41588373860935735, | |
| "learning_rate": 3.793797591072631e-05, | |
| "loss": 0.5643, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1619433198380567, | |
| "grad_norm": 0.430566751879627, | |
| "learning_rate": 3.791289387885144e-05, | |
| "loss": 0.5488, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.165991902834008, | |
| "grad_norm": 0.41777307946677167, | |
| "learning_rate": 3.788766861595443e-05, | |
| "loss": 0.562, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1700404858299596, | |
| "grad_norm": 0.4054354279169243, | |
| "learning_rate": 3.7862300323735835e-05, | |
| "loss": 0.5756, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.174089068825911, | |
| "grad_norm": 0.4192795232710301, | |
| "learning_rate": 3.783678920503983e-05, | |
| "loss": 0.5511, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1781376518218623, | |
| "grad_norm": 0.4689691639917982, | |
| "learning_rate": 3.781113546385265e-05, | |
| "loss": 0.5419, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1821862348178138, | |
| "grad_norm": 0.4617589069862704, | |
| "learning_rate": 3.7785339305300916e-05, | |
| "loss": 0.5483, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1862348178137652, | |
| "grad_norm": 0.36422814553399935, | |
| "learning_rate": 3.7759400935650025e-05, | |
| "loss": 0.5648, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1902834008097165, | |
| "grad_norm": 0.4012542088890448, | |
| "learning_rate": 3.77333205623025e-05, | |
| "loss": 0.5567, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.194331983805668, | |
| "grad_norm": 0.36457896515796245, | |
| "learning_rate": 3.7707098393796295e-05, | |
| "loss": 0.5614, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1983805668016194, | |
| "grad_norm": 0.4008193418528227, | |
| "learning_rate": 3.7680734639803185e-05, | |
| "loss": 0.5612, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2024291497975708, | |
| "grad_norm": 0.43285452096724175, | |
| "learning_rate": 3.765422951112704e-05, | |
| "loss": 0.5671, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2064777327935223, | |
| "grad_norm": 0.42722126582059106, | |
| "learning_rate": 3.762758321970216e-05, | |
| "loss": 0.5512, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.4567321401824608, | |
| "learning_rate": 3.7600795978591584e-05, | |
| "loss": 0.5596, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.214574898785425, | |
| "grad_norm": 0.43565497046029317, | |
| "learning_rate": 3.757386800198538e-05, | |
| "loss": 0.5614, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2186234817813766, | |
| "grad_norm": 0.538296021317049, | |
| "learning_rate": 3.7546799505198925e-05, | |
| "loss": 0.5515, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.222672064777328, | |
| "grad_norm": 0.43344156742620243, | |
| "learning_rate": 3.7519590704671197e-05, | |
| "loss": 0.549, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2267206477732793, | |
| "grad_norm": 0.4885136830588201, | |
| "learning_rate": 3.749224181796305e-05, | |
| "loss": 0.5704, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.5424036596881996, | |
| "learning_rate": 3.746475306375545e-05, | |
| "loss": 0.5837, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2348178137651822, | |
| "grad_norm": 0.5156904751314169, | |
| "learning_rate": 3.743712466184774e-05, | |
| "loss": 0.5623, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2388663967611335, | |
| "grad_norm": 0.5496324809111625, | |
| "learning_rate": 3.7409356833155885e-05, | |
| "loss": 0.5709, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.242914979757085, | |
| "grad_norm": 0.3965718602615464, | |
| "learning_rate": 3.73814497997107e-05, | |
| "loss": 0.5517, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2469635627530364, | |
| "grad_norm": 0.49481713209156336, | |
| "learning_rate": 3.7353403784656095e-05, | |
| "loss": 0.5763, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2510121457489878, | |
| "grad_norm": 0.40384249807116984, | |
| "learning_rate": 3.732521901224724e-05, | |
| "loss": 0.5756, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2550607287449393, | |
| "grad_norm": 0.4045142089884226, | |
| "learning_rate": 3.7296895707848844e-05, | |
| "loss": 0.5661, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2591093117408907, | |
| "grad_norm": 0.41823128836596074, | |
| "learning_rate": 3.7268434097933275e-05, | |
| "loss": 0.5671, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.39476877070478517, | |
| "learning_rate": 3.723983441007882e-05, | |
| "loss": 0.5754, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2672064777327936, | |
| "grad_norm": 0.43329000626228914, | |
| "learning_rate": 3.721109687296781e-05, | |
| "loss": 0.5457, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.271255060728745, | |
| "grad_norm": 0.47239406908184906, | |
| "learning_rate": 3.718222171638484e-05, | |
| "loss": 0.566, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2753036437246963, | |
| "grad_norm": 0.9375233920154142, | |
| "learning_rate": 3.715320917121488e-05, | |
| "loss": 0.6034, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2793522267206479, | |
| "grad_norm": 0.4874935411549607, | |
| "learning_rate": 3.712405946944149e-05, | |
| "loss": 0.587, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2834008097165992, | |
| "grad_norm": 0.5445068394105309, | |
| "learning_rate": 3.709477284414488e-05, | |
| "loss": 0.5563, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2874493927125505, | |
| "grad_norm": 0.7375327056756558, | |
| "learning_rate": 3.7065349529500136e-05, | |
| "loss": 0.5708, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.291497975708502, | |
| "grad_norm": 0.46771940325091826, | |
| "learning_rate": 3.7035789760775305e-05, | |
| "loss": 0.5646, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2955465587044535, | |
| "grad_norm": 0.48405249600394146, | |
| "learning_rate": 3.70060937743295e-05, | |
| "loss": 0.5471, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2995951417004048, | |
| "grad_norm": 0.4633601002604183, | |
| "learning_rate": 3.6976261807611045e-05, | |
| "loss": 0.5493, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.3036437246963564, | |
| "grad_norm": 0.3874970935683487, | |
| "learning_rate": 3.694629409915555e-05, | |
| "loss": 0.5491, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.4045662040531869, | |
| "learning_rate": 3.691619088858401e-05, | |
| "loss": 0.5566, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.311740890688259, | |
| "grad_norm": 0.41379371913521734, | |
| "learning_rate": 3.68859524166009e-05, | |
| "loss": 0.562, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.4135805256259969, | |
| "learning_rate": 3.6855578924992226e-05, | |
| "loss": 0.554, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.319838056680162, | |
| "grad_norm": 0.41098296249833005, | |
| "learning_rate": 3.6825070656623626e-05, | |
| "loss": 0.5708, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.3238866396761133, | |
| "grad_norm": 0.41827761225240395, | |
| "learning_rate": 3.67944278554384e-05, | |
| "loss": 0.5556, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3279352226720649, | |
| "grad_norm": 0.47255479916758725, | |
| "learning_rate": 3.676365076645557e-05, | |
| "loss": 0.5555, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.43804114345597245, | |
| "learning_rate": 3.673273963576791e-05, | |
| "loss": 0.5591, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3360323886639676, | |
| "grad_norm": 0.5268033003019378, | |
| "learning_rate": 3.670169471054001e-05, | |
| "loss": 0.5587, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3400809716599191, | |
| "grad_norm": 0.49413983579125403, | |
| "learning_rate": 3.6670516239006254e-05, | |
| "loss": 0.5498, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3441295546558705, | |
| "grad_norm": 0.4485263545899135, | |
| "learning_rate": 3.663920447046888e-05, | |
| "loss": 0.5691, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3481781376518218, | |
| "grad_norm": 0.4590750063857331, | |
| "learning_rate": 3.6607759655295954e-05, | |
| "loss": 0.5492, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3522267206477734, | |
| "grad_norm": 0.4595441542998001, | |
| "learning_rate": 3.657618204491937e-05, | |
| "loss": 0.5739, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3562753036437247, | |
| "grad_norm": 0.38864180031147366, | |
| "learning_rate": 3.6544471891832875e-05, | |
| "loss": 0.552, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.360323886639676, | |
| "grad_norm": 0.5083084945771971, | |
| "learning_rate": 3.6512629449589983e-05, | |
| "loss": 0.5437, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3643724696356276, | |
| "grad_norm": 0.4637092127569107, | |
| "learning_rate": 3.6480654972802045e-05, | |
| "loss": 0.5559, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.48441567991577, | |
| "learning_rate": 3.644854871713611e-05, | |
| "loss": 0.5573, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3724696356275303, | |
| "grad_norm": 0.660782325393081, | |
| "learning_rate": 3.6416310939312924e-05, | |
| "loss": 0.5494, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.376518218623482, | |
| "grad_norm": 0.7496729697288195, | |
| "learning_rate": 3.638394189710493e-05, | |
| "loss": 0.5659, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3805668016194332, | |
| "grad_norm": 0.406576805674725, | |
| "learning_rate": 3.635144184933412e-05, | |
| "loss": 0.5506, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.7106489062577356, | |
| "learning_rate": 3.6318811055869994e-05, | |
| "loss": 0.5547, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3886639676113361, | |
| "grad_norm": 0.4786423339713834, | |
| "learning_rate": 3.628604977762752e-05, | |
| "loss": 0.5688, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3927125506072875, | |
| "grad_norm": 0.4870543773289513, | |
| "learning_rate": 3.6253158276565004e-05, | |
| "loss": 0.5547, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3967611336032388, | |
| "grad_norm": 0.43610330546753867, | |
| "learning_rate": 3.6220136815682016e-05, | |
| "loss": 0.5687, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4008097165991904, | |
| "grad_norm": 0.5466408606051341, | |
| "learning_rate": 3.618698565901727e-05, | |
| "loss": 0.5616, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.4048582995951417, | |
| "grad_norm": 0.5096876411752935, | |
| "learning_rate": 3.615370507164654e-05, | |
| "loss": 0.5686, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.408906882591093, | |
| "grad_norm": 0.5004986850532005, | |
| "learning_rate": 3.61202953196805e-05, | |
| "loss": 0.5479, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.4129554655870447, | |
| "grad_norm": 0.5247607885026978, | |
| "learning_rate": 3.6086756670262654e-05, | |
| "loss": 0.5511, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.417004048582996, | |
| "grad_norm": 0.42394782233919676, | |
| "learning_rate": 3.6053089391567135e-05, | |
| "loss": 0.5701, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.44492931048506806, | |
| "learning_rate": 3.601929375279662e-05, | |
| "loss": 0.5559, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.425101214574899, | |
| "grad_norm": 0.5187171869795293, | |
| "learning_rate": 3.598537002418012e-05, | |
| "loss": 0.548, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4291497975708503, | |
| "grad_norm": 0.3606798663159909, | |
| "learning_rate": 3.5951318476970885e-05, | |
| "loss": 0.565, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4331983805668016, | |
| "grad_norm": 0.5032155338445629, | |
| "learning_rate": 3.591713938344416e-05, | |
| "loss": 0.5517, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4372469635627532, | |
| "grad_norm": 0.5530244631432956, | |
| "learning_rate": 3.588283301689507e-05, | |
| "loss": 0.5556, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4412955465587045, | |
| "grad_norm": 0.41528198861174337, | |
| "learning_rate": 3.5848399651636424e-05, | |
| "loss": 0.5592, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4453441295546559, | |
| "grad_norm": 0.5494245164391978, | |
| "learning_rate": 3.5813839562996473e-05, | |
| "loss": 0.5574, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4493927125506074, | |
| "grad_norm": 0.5474543309868355, | |
| "learning_rate": 3.5779153027316775e-05, | |
| "loss": 0.5594, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4534412955465588, | |
| "grad_norm": 0.47731029389929464, | |
| "learning_rate": 3.574434032194994e-05, | |
| "loss": 0.5513, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.45748987854251, | |
| "grad_norm": 0.47874948379808757, | |
| "learning_rate": 3.5709401725257436e-05, | |
| "loss": 0.5624, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.5049991620797376, | |
| "learning_rate": 3.5674337516607346e-05, | |
| "loss": 0.5582, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.465587044534413, | |
| "grad_norm": 0.3846060830993011, | |
| "learning_rate": 3.563914797637216e-05, | |
| "loss": 0.5418, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4696356275303644, | |
| "grad_norm": 0.48197157415493064, | |
| "learning_rate": 3.560383338592649e-05, | |
| "loss": 0.5613, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.6189643425045648, | |
| "learning_rate": 3.556839402764487e-05, | |
| "loss": 0.5605, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4777327935222673, | |
| "grad_norm": 0.412744364270222, | |
| "learning_rate": 3.553283018489946e-05, | |
| "loss": 0.552, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4817813765182186, | |
| "grad_norm": 0.47450645922471446, | |
| "learning_rate": 3.54971421420578e-05, | |
| "loss": 0.5657, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.48582995951417, | |
| "grad_norm": 0.4625921566853585, | |
| "learning_rate": 3.5461330184480524e-05, | |
| "loss": 0.549, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4898785425101215, | |
| "grad_norm": 0.4764343518126904, | |
| "learning_rate": 3.54253945985191e-05, | |
| "loss": 0.5356, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4939271255060729, | |
| "grad_norm": 0.35699012282513404, | |
| "learning_rate": 3.5389335671513524e-05, | |
| "loss": 0.5437, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4979757085020242, | |
| "grad_norm": 0.44528918479968377, | |
| "learning_rate": 3.5353153691789986e-05, | |
| "loss": 0.5566, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5020242914979756, | |
| "grad_norm": 0.48426675888728804, | |
| "learning_rate": 3.531684894865866e-05, | |
| "loss": 0.5438, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.5060728744939271, | |
| "grad_norm": 0.4044180676470221, | |
| "learning_rate": 3.528042173241128e-05, | |
| "loss": 0.5696, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5101214574898787, | |
| "grad_norm": 0.3887645644860446, | |
| "learning_rate": 3.524387233431892e-05, | |
| "loss": 0.5706, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.5141700404858298, | |
| "grad_norm": 0.39605071122297575, | |
| "learning_rate": 3.520720104662958e-05, | |
| "loss": 0.5557, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5182186234817814, | |
| "grad_norm": 0.4220426854526296, | |
| "learning_rate": 3.5170408162565904e-05, | |
| "loss": 0.5717, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.522267206477733, | |
| "grad_norm": 0.41363712583714735, | |
| "learning_rate": 3.513349397632281e-05, | |
| "loss": 0.5554, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.4385190789796589, | |
| "learning_rate": 3.5096458783065145e-05, | |
| "loss": 0.5602, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5303643724696356, | |
| "grad_norm": 0.39761418303143353, | |
| "learning_rate": 3.505930287892533e-05, | |
| "loss": 0.551, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.5344129554655872, | |
| "grad_norm": 0.43483860767715904, | |
| "learning_rate": 3.502202656100099e-05, | |
| "loss": 0.5474, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.4551055502892514, | |
| "learning_rate": 3.498463012735256e-05, | |
| "loss": 0.5532, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.54251012145749, | |
| "grad_norm": 0.4699854518983839, | |
| "learning_rate": 3.494711387700094e-05, | |
| "loss": 0.5715, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5465587044534415, | |
| "grad_norm": 0.4314721383962728, | |
| "learning_rate": 3.490947810992505e-05, | |
| "loss": 0.5433, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5506072874493926, | |
| "grad_norm": 0.4658317653928339, | |
| "learning_rate": 3.48717231270595e-05, | |
| "loss": 0.5424, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5546558704453441, | |
| "grad_norm": 0.37723315991208517, | |
| "learning_rate": 3.483384923029211e-05, | |
| "loss": 0.5598, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5587044534412957, | |
| "grad_norm": 0.44689216908108387, | |
| "learning_rate": 3.479585672246156e-05, | |
| "loss": 0.5424, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5627530364372468, | |
| "grad_norm": 0.4782899383070214, | |
| "learning_rate": 3.4757745907354924e-05, | |
| "loss": 0.5646, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5668016194331984, | |
| "grad_norm": 0.3731802795020949, | |
| "learning_rate": 3.471951708970527e-05, | |
| "loss": 0.5603, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5708502024291497, | |
| "grad_norm": 0.46298190582291365, | |
| "learning_rate": 3.468117057518921e-05, | |
| "loss": 0.56, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.574898785425101, | |
| "grad_norm": 0.40947044354603845, | |
| "learning_rate": 3.4642706670424426e-05, | |
| "loss": 0.5592, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.38698691991361167, | |
| "learning_rate": 3.460412568296731e-05, | |
| "loss": 0.5539, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.582995951417004, | |
| "grad_norm": 0.39981280592798785, | |
| "learning_rate": 3.456542792131039e-05, | |
| "loss": 0.5524, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5870445344129553, | |
| "grad_norm": 0.43488615140419656, | |
| "learning_rate": 3.4526613694879936e-05, | |
| "loss": 0.5595, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.591093117408907, | |
| "grad_norm": 0.39983299584443904, | |
| "learning_rate": 3.448768331403348e-05, | |
| "loss": 0.5436, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5951417004048583, | |
| "grad_norm": 0.4295789508215197, | |
| "learning_rate": 3.44486370900573e-05, | |
| "loss": 0.5445, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5991902834008096, | |
| "grad_norm": 0.5112853417242073, | |
| "learning_rate": 3.440947533516396e-05, | |
| "loss": 0.5721, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.6032388663967612, | |
| "grad_norm": 0.37870143888646524, | |
| "learning_rate": 3.4370198362489816e-05, | |
| "loss": 0.5751, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.6072874493927125, | |
| "grad_norm": 0.39867028209400646, | |
| "learning_rate": 3.43308064860925e-05, | |
| "loss": 0.5624, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.6113360323886639, | |
| "grad_norm": 0.4112518415052006, | |
| "learning_rate": 3.4291300020948393e-05, | |
| "loss": 0.5501, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.4591747267386976, | |
| "learning_rate": 3.425167928295015e-05, | |
| "loss": 0.5439, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6194331983805668, | |
| "grad_norm": 0.43770930976375694, | |
| "learning_rate": 3.421194458890411e-05, | |
| "loss": 0.5614, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.623481781376518, | |
| "grad_norm": 0.4254028181280734, | |
| "learning_rate": 3.4172096256527865e-05, | |
| "loss": 0.5626, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.6275303643724697, | |
| "grad_norm": 0.4658329958831512, | |
| "learning_rate": 3.4132134604447595e-05, | |
| "loss": 0.554, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.4789543460611861, | |
| "learning_rate": 3.409205995219561e-05, | |
| "loss": 0.5506, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6356275303643724, | |
| "grad_norm": 0.3900568738845315, | |
| "learning_rate": 3.4051872620207765e-05, | |
| "loss": 0.559, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.639676113360324, | |
| "grad_norm": 0.49671641495178637, | |
| "learning_rate": 3.40115729298209e-05, | |
| "loss": 0.5696, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6437246963562753, | |
| "grad_norm": 0.4712081454388575, | |
| "learning_rate": 3.397116120327025e-05, | |
| "loss": 0.5348, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6477732793522266, | |
| "grad_norm": 0.46362561125909035, | |
| "learning_rate": 3.3930637763686916e-05, | |
| "loss": 0.5369, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6518218623481782, | |
| "grad_norm": 0.489792863559607, | |
| "learning_rate": 3.389000293509524e-05, | |
| "loss": 0.5607, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6558704453441295, | |
| "grad_norm": 0.5707221335617982, | |
| "learning_rate": 3.3849257042410225e-05, | |
| "loss": 0.5465, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6599190283400809, | |
| "grad_norm": 0.392844018568974, | |
| "learning_rate": 3.380840041143494e-05, | |
| "loss": 0.5498, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6639676113360324, | |
| "grad_norm": 0.37725595371159326, | |
| "learning_rate": 3.376743336885792e-05, | |
| "loss": 0.5491, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6680161943319838, | |
| "grad_norm": 0.5585091637991101, | |
| "learning_rate": 3.372635624225055e-05, | |
| "loss": 0.5423, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6720647773279351, | |
| "grad_norm": 0.3555832739423202, | |
| "learning_rate": 3.3685169360064445e-05, | |
| "loss": 0.5572, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6761133603238867, | |
| "grad_norm": 0.5040767341966229, | |
| "learning_rate": 3.3643873051628804e-05, | |
| "loss": 0.5452, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.680161943319838, | |
| "grad_norm": 0.43577436227926514, | |
| "learning_rate": 3.3602467647147814e-05, | |
| "loss": 0.5611, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.45138319293743656, | |
| "learning_rate": 3.356095347769799e-05, | |
| "loss": 0.555, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.688259109311741, | |
| "grad_norm": 0.3818669720074758, | |
| "learning_rate": 3.351933087522552e-05, | |
| "loss": 0.5652, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.5740637895169447, | |
| "learning_rate": 3.3477600172543626e-05, | |
| "loss": 0.5658, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6963562753036436, | |
| "grad_norm": 0.43560773701795547, | |
| "learning_rate": 3.3435761703329894e-05, | |
| "loss": 0.5644, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.7004048582995952, | |
| "grad_norm": 0.38510893785841777, | |
| "learning_rate": 3.3393815802123606e-05, | |
| "loss": 0.5595, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7044534412955465, | |
| "grad_norm": 0.47993508677948815, | |
| "learning_rate": 3.335176280432308e-05, | |
| "loss": 0.5473, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.708502024291498, | |
| "grad_norm": 0.5036468161659058, | |
| "learning_rate": 3.330960304618295e-05, | |
| "loss": 0.5582, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.7125506072874495, | |
| "grad_norm": 0.3280607438796661, | |
| "learning_rate": 3.326733686481153e-05, | |
| "loss": 0.5445, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.7165991902834008, | |
| "grad_norm": 0.3886434900457605, | |
| "learning_rate": 3.3224964598168084e-05, | |
| "loss": 0.5656, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.5043617157739533, | |
| "learning_rate": 3.318248658506012e-05, | |
| "loss": 0.5601, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7246963562753037, | |
| "grad_norm": 0.39933349449406413, | |
| "learning_rate": 3.313990316514069e-05, | |
| "loss": 0.5646, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.728744939271255, | |
| "grad_norm": 0.3862012705735096, | |
| "learning_rate": 3.309721467890571e-05, | |
| "loss": 0.5681, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.7327935222672064, | |
| "grad_norm": 0.47093903543671484, | |
| "learning_rate": 3.305442146769115e-05, | |
| "loss": 0.5496, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.40152346566014113, | |
| "learning_rate": 3.30115238736704e-05, | |
| "loss": 0.5489, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7408906882591093, | |
| "grad_norm": 0.35448288432190583, | |
| "learning_rate": 3.296852223985148e-05, | |
| "loss": 0.5652, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7449392712550607, | |
| "grad_norm": 0.4719323734000785, | |
| "learning_rate": 3.2925416910074315e-05, | |
| "loss": 0.5494, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7489878542510122, | |
| "grad_norm": 0.4558549862959935, | |
| "learning_rate": 3.288220822900796e-05, | |
| "loss": 0.5429, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7530364372469636, | |
| "grad_norm": 0.39997893721177136, | |
| "learning_rate": 3.283889654214788e-05, | |
| "loss": 0.5505, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.757085020242915, | |
| "grad_norm": 0.5120071236436007, | |
| "learning_rate": 3.279548219581317e-05, | |
| "loss": 0.5487, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7611336032388665, | |
| "grad_norm": 0.4240026731180767, | |
| "learning_rate": 3.275196553714379e-05, | |
| "loss": 0.5587, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7651821862348178, | |
| "grad_norm": 0.4608960303503504, | |
| "learning_rate": 3.270834691409778e-05, | |
| "loss": 0.5474, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.5535681319593274, | |
| "learning_rate": 3.266462667544849e-05, | |
| "loss": 0.5519, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7732793522267207, | |
| "grad_norm": 0.3940613513982405, | |
| "learning_rate": 3.262080517078178e-05, | |
| "loss": 0.553, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.777327935222672, | |
| "grad_norm": 0.5034400215351699, | |
| "learning_rate": 3.257688275049323e-05, | |
| "loss": 0.5533, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7813765182186234, | |
| "grad_norm": 0.5075651463550881, | |
| "learning_rate": 3.253285976578535e-05, | |
| "loss": 0.5667, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.785425101214575, | |
| "grad_norm": 0.38063733403247074, | |
| "learning_rate": 3.2488736568664756e-05, | |
| "loss": 0.553, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.3712316491083618, | |
| "learning_rate": 3.244451351193935e-05, | |
| "loss": 0.5553, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7935222672064777, | |
| "grad_norm": 0.3690052550417436, | |
| "learning_rate": 3.240019094921551e-05, | |
| "loss": 0.5566, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7975708502024292, | |
| "grad_norm": 0.3607595034110213, | |
| "learning_rate": 3.235576923489527e-05, | |
| "loss": 0.5628, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.8016194331983806, | |
| "grad_norm": 0.458842370097784, | |
| "learning_rate": 3.2311248724173476e-05, | |
| "loss": 0.5542, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.805668016194332, | |
| "grad_norm": 0.4031042916189146, | |
| "learning_rate": 3.226662977303494e-05, | |
| "loss": 0.5407, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.8097165991902835, | |
| "grad_norm": 0.4063486121649146, | |
| "learning_rate": 3.2221912738251626e-05, | |
| "loss": 0.564, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8137651821862348, | |
| "grad_norm": 0.4241259589047649, | |
| "learning_rate": 3.2177097977379736e-05, | |
| "loss": 0.5586, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.8178137651821862, | |
| "grad_norm": 0.4604181322001097, | |
| "learning_rate": 3.21321858487569e-05, | |
| "loss": 0.5663, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8218623481781377, | |
| "grad_norm": 0.41492021101433607, | |
| "learning_rate": 3.208717671149931e-05, | |
| "loss": 0.5626, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.825910931174089, | |
| "grad_norm": 0.41002819692340065, | |
| "learning_rate": 3.204207092549882e-05, | |
| "loss": 0.5422, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.8299595141700404, | |
| "grad_norm": 0.3346375538594348, | |
| "learning_rate": 3.19968688514201e-05, | |
| "loss": 0.5421, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.834008097165992, | |
| "grad_norm": 0.38115691566070303, | |
| "learning_rate": 3.195157085069774e-05, | |
| "loss": 0.5476, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8380566801619433, | |
| "grad_norm": 0.3113873114726482, | |
| "learning_rate": 3.190617728553332e-05, | |
| "loss": 0.566, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.3553381602243142, | |
| "learning_rate": 3.1860688518892606e-05, | |
| "loss": 0.5399, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.37945786108688634, | |
| "learning_rate": 3.181510491450254e-05, | |
| "loss": 0.5642, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8502024291497976, | |
| "grad_norm": 0.4450304045024324, | |
| "learning_rate": 3.176942683684842e-05, | |
| "loss": 0.5442, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.854251012145749, | |
| "grad_norm": 0.41762682586974836, | |
| "learning_rate": 3.1723654651170934e-05, | |
| "loss": 0.5555, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8582995951417005, | |
| "grad_norm": 0.413606355522318, | |
| "learning_rate": 3.167778872346327e-05, | |
| "loss": 0.5604, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8623481781376519, | |
| "grad_norm": 0.49182937872572274, | |
| "learning_rate": 3.163182942046815e-05, | |
| "loss": 0.5608, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8663967611336032, | |
| "grad_norm": 0.44415056964337457, | |
| "learning_rate": 3.158577710967494e-05, | |
| "loss": 0.548, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8704453441295548, | |
| "grad_norm": 0.3474263431386141, | |
| "learning_rate": 3.153963215931669e-05, | |
| "loss": 0.5418, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.874493927125506, | |
| "grad_norm": 0.4541346437588395, | |
| "learning_rate": 3.14933949383672e-05, | |
| "loss": 0.5489, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8785425101214575, | |
| "grad_norm": 0.36591063617054775, | |
| "learning_rate": 3.144706581653805e-05, | |
| "loss": 0.5502, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.882591093117409, | |
| "grad_norm": 0.5099803815491926, | |
| "learning_rate": 3.140064516427566e-05, | |
| "loss": 0.5659, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8866396761133604, | |
| "grad_norm": 0.3848890990018038, | |
| "learning_rate": 3.135413335275833e-05, | |
| "loss": 0.5665, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8906882591093117, | |
| "grad_norm": 0.4901830841053478, | |
| "learning_rate": 3.130753075389327e-05, | |
| "loss": 0.5494, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.4368793910102728, | |
| "learning_rate": 3.1260837740313625e-05, | |
| "loss": 0.5444, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8987854251012146, | |
| "grad_norm": 0.4167674575210837, | |
| "learning_rate": 3.121405468537547e-05, | |
| "loss": 0.5512, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.902834008097166, | |
| "grad_norm": 0.46844744657657816, | |
| "learning_rate": 3.116718196315488e-05, | |
| "loss": 0.5278, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9068825910931175, | |
| "grad_norm": 0.3505272721671307, | |
| "learning_rate": 3.1120219948444894e-05, | |
| "loss": 0.5459, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.9109311740890689, | |
| "grad_norm": 0.42807381385223586, | |
| "learning_rate": 3.107316901675252e-05, | |
| "loss": 0.5569, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.9149797570850202, | |
| "grad_norm": 0.3848385152695767, | |
| "learning_rate": 3.1026029544295745e-05, | |
| "loss": 0.553, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.9190283400809718, | |
| "grad_norm": 0.3111796579280068, | |
| "learning_rate": 3.097880190800057e-05, | |
| "loss": 0.5604, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.3309080115646451, | |
| "learning_rate": 3.093148648549788e-05, | |
| "loss": 0.5655, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9271255060728745, | |
| "grad_norm": 0.3608434216078505, | |
| "learning_rate": 3.088408365512055e-05, | |
| "loss": 0.5633, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.931174089068826, | |
| "grad_norm": 0.3634246174744796, | |
| "learning_rate": 3.083659379590034e-05, | |
| "loss": 0.5585, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.9352226720647774, | |
| "grad_norm": 0.46452152480279546, | |
| "learning_rate": 3.0789017287564913e-05, | |
| "loss": 0.5477, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9392712550607287, | |
| "grad_norm": 0.3687492816085792, | |
| "learning_rate": 3.074135451053475e-05, | |
| "loss": 0.5571, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9433198380566803, | |
| "grad_norm": 0.4362083431217131, | |
| "learning_rate": 3.069360584592015e-05, | |
| "loss": 0.5541, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.4616636834929283, | |
| "learning_rate": 3.064577167551816e-05, | |
| "loss": 0.5462, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.951417004048583, | |
| "grad_norm": 0.3335580101448725, | |
| "learning_rate": 3.059785238180953e-05, | |
| "loss": 0.5561, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9554655870445345, | |
| "grad_norm": 0.4952204349630398, | |
| "learning_rate": 3.054984834795566e-05, | |
| "loss": 0.5633, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9595141700404857, | |
| "grad_norm": 0.341045894330387, | |
| "learning_rate": 3.0501759957795537e-05, | |
| "loss": 0.5561, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9635627530364372, | |
| "grad_norm": 0.5009707469275333, | |
| "learning_rate": 3.0453587595842627e-05, | |
| "loss": 0.5458, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9676113360323888, | |
| "grad_norm": 0.5054167375795247, | |
| "learning_rate": 3.0405331647281856e-05, | |
| "loss": 0.5583, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.97165991902834, | |
| "grad_norm": 0.3960563457202511, | |
| "learning_rate": 3.0356992497966508e-05, | |
| "loss": 0.5269, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9757085020242915, | |
| "grad_norm": 0.5626907615779736, | |
| "learning_rate": 3.0308570534415137e-05, | |
| "loss": 0.5576, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.979757085020243, | |
| "grad_norm": 0.4669919973604278, | |
| "learning_rate": 3.0260066143808467e-05, | |
| "loss": 0.549, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9838056680161942, | |
| "grad_norm": 0.3392295801182429, | |
| "learning_rate": 3.021147971398632e-05, | |
| "loss": 0.5508, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9878542510121457, | |
| "grad_norm": 0.42855995874179437, | |
| "learning_rate": 3.0162811633444478e-05, | |
| "loss": 0.5657, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9919028340080973, | |
| "grad_norm": 0.4155220372546249, | |
| "learning_rate": 3.0114062291331635e-05, | |
| "loss": 0.5534, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9959514170040484, | |
| "grad_norm": 0.3570198541411215, | |
| "learning_rate": 3.0065232077446217e-05, | |
| "loss": 0.5479, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.4264136567706597, | |
| "learning_rate": 3.001632138223332e-05, | |
| "loss": 0.5445, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.0040485829959516, | |
| "grad_norm": 0.5255964693879461, | |
| "learning_rate": 2.9967330596781565e-05, | |
| "loss": 0.4761, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0080971659919027, | |
| "grad_norm": 0.6965829761669385, | |
| "learning_rate": 2.9918260112819952e-05, | |
| "loss": 0.4798, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0121457489878543, | |
| "grad_norm": 0.5592827222562774, | |
| "learning_rate": 2.9869110322714778e-05, | |
| "loss": 0.4808, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.016194331983806, | |
| "grad_norm": 0.4920174105089216, | |
| "learning_rate": 2.9819881619466447e-05, | |
| "loss": 0.4758, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.020242914979757, | |
| "grad_norm": 0.4999913429699286, | |
| "learning_rate": 2.9770574396706364e-05, | |
| "loss": 0.4911, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0242914979757085, | |
| "grad_norm": 0.5715872348484083, | |
| "learning_rate": 2.972118904869376e-05, | |
| "loss": 0.4733, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02834008097166, | |
| "grad_norm": 0.47196921634390304, | |
| "learning_rate": 2.9671725970312563e-05, | |
| "loss": 0.4775, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.032388663967611, | |
| "grad_norm": 0.545988311567732, | |
| "learning_rate": 2.9622185557068238e-05, | |
| "loss": 0.4876, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.0364372469635628, | |
| "grad_norm": 0.4099366295819739, | |
| "learning_rate": 2.95725682050846e-05, | |
| "loss": 0.4862, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0404858299595143, | |
| "grad_norm": 0.45982356593228474, | |
| "learning_rate": 2.9522874311100684e-05, | |
| "loss": 0.4855, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0445344129554655, | |
| "grad_norm": 0.44256741247401815, | |
| "learning_rate": 2.9473104272467526e-05, | |
| "loss": 0.4728, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.048582995951417, | |
| "grad_norm": 0.4278733678886794, | |
| "learning_rate": 2.942325848714505e-05, | |
| "loss": 0.4857, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.4657703840069526, | |
| "learning_rate": 2.9373337353698804e-05, | |
| "loss": 0.4869, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.0566801619433197, | |
| "grad_norm": 0.3828074947587904, | |
| "learning_rate": 2.932334127129686e-05, | |
| "loss": 0.4834, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.0607287449392713, | |
| "grad_norm": 0.4244219651001206, | |
| "learning_rate": 2.9273270639706547e-05, | |
| "loss": 0.4828, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.064777327935223, | |
| "grad_norm": 0.42159709947760177, | |
| "learning_rate": 2.922312585929131e-05, | |
| "loss": 0.492, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.068825910931174, | |
| "grad_norm": 0.31557142848801795, | |
| "learning_rate": 2.9172907331007478e-05, | |
| "loss": 0.4924, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.0728744939271255, | |
| "grad_norm": 0.42133146594842114, | |
| "learning_rate": 2.9122615456401053e-05, | |
| "loss": 0.4747, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.3480957420911763, | |
| "learning_rate": 2.9072250637604535e-05, | |
| "loss": 0.4732, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.080971659919028, | |
| "grad_norm": 0.3743857533798433, | |
| "learning_rate": 2.9021813277333667e-05, | |
| "loss": 0.4794, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.08502024291498, | |
| "grad_norm": 0.39916724054863567, | |
| "learning_rate": 2.8971303778884244e-05, | |
| "loss": 0.4959, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.0890688259109313, | |
| "grad_norm": 0.3204847821809214, | |
| "learning_rate": 2.8920722546128863e-05, | |
| "loss": 0.48, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0931174089068825, | |
| "grad_norm": 0.3515394043400751, | |
| "learning_rate": 2.887006998351372e-05, | |
| "loss": 0.4658, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.097165991902834, | |
| "grad_norm": 0.3434410732445909, | |
| "learning_rate": 2.881934649605535e-05, | |
| "loss": 0.4707, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.1012145748987856, | |
| "grad_norm": 0.3597533987014403, | |
| "learning_rate": 2.8768552489337408e-05, | |
| "loss": 0.4743, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.3211913397227537, | |
| "learning_rate": 2.8717688369507425e-05, | |
| "loss": 0.4773, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1093117408906883, | |
| "grad_norm": 0.40454291649481644, | |
| "learning_rate": 2.8666754543273535e-05, | |
| "loss": 0.4798, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.11336032388664, | |
| "grad_norm": 0.33733785833994895, | |
| "learning_rate": 2.8615751417901273e-05, | |
| "loss": 0.4811, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.117408906882591, | |
| "grad_norm": 0.38237218076159757, | |
| "learning_rate": 2.856467940121025e-05, | |
| "loss": 0.5113, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.1214574898785425, | |
| "grad_norm": 0.38568119587480415, | |
| "learning_rate": 2.8513538901570964e-05, | |
| "loss": 0.4798, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.125506072874494, | |
| "grad_norm": 0.34546556852635707, | |
| "learning_rate": 2.846233032790149e-05, | |
| "loss": 0.4811, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1295546558704452, | |
| "grad_norm": 0.4171349023924531, | |
| "learning_rate": 2.84110540896642e-05, | |
| "loss": 0.4649, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.133603238866397, | |
| "grad_norm": 0.3478707990350369, | |
| "learning_rate": 2.8359710596862555e-05, | |
| "loss": 0.481, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.1376518218623484, | |
| "grad_norm": 0.45158014649975997, | |
| "learning_rate": 2.8308300260037734e-05, | |
| "loss": 0.4787, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.1417004048582995, | |
| "grad_norm": 0.33399871129009595, | |
| "learning_rate": 2.8256823490265424e-05, | |
| "loss": 0.4734, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.145748987854251, | |
| "grad_norm": 0.4153482709955574, | |
| "learning_rate": 2.8205280699152515e-05, | |
| "loss": 0.4741, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1497975708502026, | |
| "grad_norm": 0.4267155948795778, | |
| "learning_rate": 2.8153672298833777e-05, | |
| "loss": 0.4836, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.3768508667954057, | |
| "learning_rate": 2.8101998701968615e-05, | |
| "loss": 0.487, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.3672893654806205, | |
| "learning_rate": 2.805026032173773e-05, | |
| "loss": 0.4833, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.161943319838057, | |
| "grad_norm": 0.3268063453507024, | |
| "learning_rate": 2.7998457571839832e-05, | |
| "loss": 0.486, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.165991902834008, | |
| "grad_norm": 0.38061006082391696, | |
| "learning_rate": 2.7946590866488322e-05, | |
| "loss": 0.4815, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1700404858299596, | |
| "grad_norm": 0.31861383419032996, | |
| "learning_rate": 2.7894660620408e-05, | |
| "loss": 0.4867, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.174089068825911, | |
| "grad_norm": 0.37031292494134627, | |
| "learning_rate": 2.784266724883173e-05, | |
| "loss": 0.4682, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.1781376518218623, | |
| "grad_norm": 0.33582509367856256, | |
| "learning_rate": 2.7790611167497128e-05, | |
| "loss": 0.4914, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.182186234817814, | |
| "grad_norm": 0.3436402249342084, | |
| "learning_rate": 2.7738492792643224e-05, | |
| "loss": 0.4841, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.1862348178137654, | |
| "grad_norm": 0.3764580709883491, | |
| "learning_rate": 2.768631254100716e-05, | |
| "loss": 0.4864, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1902834008097165, | |
| "grad_norm": 0.3472163483951373, | |
| "learning_rate": 2.7634070829820836e-05, | |
| "loss": 0.487, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.194331983805668, | |
| "grad_norm": 0.4261083669564123, | |
| "learning_rate": 2.7581768076807592e-05, | |
| "loss": 0.4583, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.1983805668016196, | |
| "grad_norm": 0.4001357030736576, | |
| "learning_rate": 2.7529404700178832e-05, | |
| "loss": 0.4779, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.2024291497975708, | |
| "grad_norm": 0.33803835274619953, | |
| "learning_rate": 2.7476981118630735e-05, | |
| "loss": 0.4636, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.2064777327935223, | |
| "grad_norm": 0.3551004535296609, | |
| "learning_rate": 2.742449775134085e-05, | |
| "loss": 0.4748, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.3351110063164218, | |
| "learning_rate": 2.7371955017964788e-05, | |
| "loss": 0.4621, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.214574898785425, | |
| "grad_norm": 0.4436396189365972, | |
| "learning_rate": 2.7319353338632845e-05, | |
| "loss": 0.4697, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.2186234817813766, | |
| "grad_norm": 0.3324035372645233, | |
| "learning_rate": 2.7266693133946645e-05, | |
| "loss": 0.4668, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.2226720647773277, | |
| "grad_norm": 0.3721943997158495, | |
| "learning_rate": 2.7213974824975785e-05, | |
| "loss": 0.4652, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2267206477732793, | |
| "grad_norm": 0.3613790375250415, | |
| "learning_rate": 2.7161198833254464e-05, | |
| "loss": 0.4714, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.3945178935371662, | |
| "learning_rate": 2.7108365580778095e-05, | |
| "loss": 0.4808, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.234817813765182, | |
| "grad_norm": 0.3322510870092138, | |
| "learning_rate": 2.705547548999996e-05, | |
| "loss": 0.4754, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.2388663967611335, | |
| "grad_norm": 0.33246644212367266, | |
| "learning_rate": 2.7002528983827817e-05, | |
| "loss": 0.4888, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.242914979757085, | |
| "grad_norm": 0.34545219569241686, | |
| "learning_rate": 2.6949526485620517e-05, | |
| "loss": 0.479, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.246963562753036, | |
| "grad_norm": 0.33036372064219877, | |
| "learning_rate": 2.6896468419184623e-05, | |
| "loss": 0.4982, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.251012145748988, | |
| "grad_norm": 0.3017124629426259, | |
| "learning_rate": 2.6843355208771024e-05, | |
| "loss": 0.4779, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.2550607287449393, | |
| "grad_norm": 0.37890099755780277, | |
| "learning_rate": 2.6790187279071536e-05, | |
| "loss": 0.4796, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.2591093117408905, | |
| "grad_norm": 0.32424420335355325, | |
| "learning_rate": 2.67369650552155e-05, | |
| "loss": 0.4835, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.37504873826628843, | |
| "learning_rate": 2.6683688962766407e-05, | |
| "loss": 0.4792, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.2672064777327936, | |
| "grad_norm": 0.3482786869640209, | |
| "learning_rate": 2.6630359427718468e-05, | |
| "loss": 0.4811, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2712550607287447, | |
| "grad_norm": 0.3828251902018629, | |
| "learning_rate": 2.6576976876493225e-05, | |
| "loss": 0.4721, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.3784212495364127, | |
| "learning_rate": 2.652354173593614e-05, | |
| "loss": 0.4806, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.279352226720648, | |
| "grad_norm": 0.3820167295077175, | |
| "learning_rate": 2.647005443331317e-05, | |
| "loss": 0.4856, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.283400809716599, | |
| "grad_norm": 0.33919050846179793, | |
| "learning_rate": 2.6416515396307354e-05, | |
| "loss": 0.5064, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.2874493927125505, | |
| "grad_norm": 0.3981920470377488, | |
| "learning_rate": 2.636292505301542e-05, | |
| "loss": 0.5045, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.291497975708502, | |
| "grad_norm": 0.3282928187093333, | |
| "learning_rate": 2.630928383194431e-05, | |
| "loss": 0.4728, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.2955465587044532, | |
| "grad_norm": 0.3729801723563401, | |
| "learning_rate": 2.625559216200781e-05, | |
| "loss": 0.4908, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.299595141700405, | |
| "grad_norm": 0.36768861610459475, | |
| "learning_rate": 2.620185047252308e-05, | |
| "loss": 0.519, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.3036437246963564, | |
| "grad_norm": 0.3278921501357504, | |
| "learning_rate": 2.614805919320723e-05, | |
| "loss": 0.4807, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.3420959617877313, | |
| "learning_rate": 2.6094218754173895e-05, | |
| "loss": 0.4727, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.311740890688259, | |
| "grad_norm": 0.32187711864973034, | |
| "learning_rate": 2.60403295859298e-05, | |
| "loss": 0.4623, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.34473098585928796, | |
| "learning_rate": 2.5986392119371297e-05, | |
| "loss": 0.4868, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.3198380566801617, | |
| "grad_norm": 0.3784351604369565, | |
| "learning_rate": 2.5932406785780928e-05, | |
| "loss": 0.4892, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.3238866396761133, | |
| "grad_norm": 0.3253665066610285, | |
| "learning_rate": 2.5878374016823987e-05, | |
| "loss": 0.4934, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.327935222672065, | |
| "grad_norm": 0.40232950990596195, | |
| "learning_rate": 2.5824294244545045e-05, | |
| "loss": 0.4803, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.331983805668016, | |
| "grad_norm": 0.3543332133333822, | |
| "learning_rate": 2.577016790136453e-05, | |
| "loss": 0.4827, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.3360323886639676, | |
| "grad_norm": 0.3258066603140043, | |
| "learning_rate": 2.5715995420075225e-05, | |
| "loss": 0.4684, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.340080971659919, | |
| "grad_norm": 0.32765625923015856, | |
| "learning_rate": 2.5661777233838864e-05, | |
| "loss": 0.4883, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.3441295546558703, | |
| "grad_norm": 0.33092591562232326, | |
| "learning_rate": 2.5607513776182616e-05, | |
| "loss": 0.4826, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.348178137651822, | |
| "grad_norm": 0.5212338177448698, | |
| "learning_rate": 2.5553205480995633e-05, | |
| "loss": 0.4773, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3522267206477734, | |
| "grad_norm": 0.3477730770135436, | |
| "learning_rate": 2.5498852782525604e-05, | |
| "loss": 0.4838, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.3562753036437245, | |
| "grad_norm": 0.3140536131919354, | |
| "learning_rate": 2.5444456115375252e-05, | |
| "loss": 0.4737, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.360323886639676, | |
| "grad_norm": 0.3361313645881251, | |
| "learning_rate": 2.539001591449889e-05, | |
| "loss": 0.4859, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.3643724696356276, | |
| "grad_norm": 0.3490435683312682, | |
| "learning_rate": 2.53355326151989e-05, | |
| "loss": 0.4776, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.2832986346497351, | |
| "learning_rate": 2.5281006653122288e-05, | |
| "loss": 0.4816, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.3724696356275303, | |
| "grad_norm": 0.377078791346549, | |
| "learning_rate": 2.5226438464257217e-05, | |
| "loss": 0.4978, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.376518218623482, | |
| "grad_norm": 0.35350365039852927, | |
| "learning_rate": 2.5171828484929447e-05, | |
| "loss": 0.4766, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.380566801619433, | |
| "grad_norm": 0.3142476615985253, | |
| "learning_rate": 2.5117177151798934e-05, | |
| "loss": 0.4777, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.4206283452628263, | |
| "learning_rate": 2.506248490185627e-05, | |
| "loss": 0.5056, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.388663967611336, | |
| "grad_norm": 3.5194600775134477, | |
| "learning_rate": 2.500775217241923e-05, | |
| "loss": 0.4813, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.3927125506072873, | |
| "grad_norm": 0.4913319118727765, | |
| "learning_rate": 2.4952979401129266e-05, | |
| "loss": 0.4727, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.396761133603239, | |
| "grad_norm": 0.42095510762807253, | |
| "learning_rate": 2.4898167025948e-05, | |
| "loss": 0.47, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.4008097165991904, | |
| "grad_norm": 0.3971684701998389, | |
| "learning_rate": 2.4843315485153703e-05, | |
| "loss": 0.4836, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.4048582995951415, | |
| "grad_norm": 0.44739727070635266, | |
| "learning_rate": 2.4788425217337856e-05, | |
| "loss": 0.4938, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.408906882591093, | |
| "grad_norm": 0.3225402197511607, | |
| "learning_rate": 2.4733496661401572e-05, | |
| "loss": 0.4794, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.4129554655870447, | |
| "grad_norm": 0.4356066506415827, | |
| "learning_rate": 2.4678530256552114e-05, | |
| "loss": 0.4887, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.417004048582996, | |
| "grad_norm": 0.31189667633794854, | |
| "learning_rate": 2.46235264422994e-05, | |
| "loss": 0.4834, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.40195977720162973, | |
| "learning_rate": 2.4568485658452452e-05, | |
| "loss": 0.498, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.425101214574899, | |
| "grad_norm": 0.3076855808055376, | |
| "learning_rate": 2.451340834511592e-05, | |
| "loss": 0.4892, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.42914979757085, | |
| "grad_norm": 0.3165425295280976, | |
| "learning_rate": 2.4458294942686528e-05, | |
| "loss": 0.4833, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4331983805668016, | |
| "grad_norm": 0.3427742338307665, | |
| "learning_rate": 2.440314589184958e-05, | |
| "loss": 0.48, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.437246963562753, | |
| "grad_norm": 0.3318565717616365, | |
| "learning_rate": 2.434796163357541e-05, | |
| "loss": 0.4865, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.4412955465587043, | |
| "grad_norm": 0.34540774912712935, | |
| "learning_rate": 2.429274260911588e-05, | |
| "loss": 0.4755, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.445344129554656, | |
| "grad_norm": 0.3308820933389972, | |
| "learning_rate": 2.423748926000084e-05, | |
| "loss": 0.4798, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.4493927125506074, | |
| "grad_norm": 0.34110261534065395, | |
| "learning_rate": 2.418220202803459e-05, | |
| "loss": 0.4818, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.4534412955465585, | |
| "grad_norm": 0.3721372923670884, | |
| "learning_rate": 2.412688135529237e-05, | |
| "loss": 0.4731, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.45748987854251, | |
| "grad_norm": 0.30401810187273326, | |
| "learning_rate": 2.4071527684116807e-05, | |
| "loss": 0.4898, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.36366493579391046, | |
| "learning_rate": 2.4016141457114372e-05, | |
| "loss": 0.4868, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.465587044534413, | |
| "grad_norm": 0.3180935893724043, | |
| "learning_rate": 2.396072311715187e-05, | |
| "loss": 0.4826, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.4696356275303644, | |
| "grad_norm": 0.40558437306330414, | |
| "learning_rate": 2.3905273107352863e-05, | |
| "loss": 0.4829, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.3085626073369214, | |
| "learning_rate": 2.3849791871094157e-05, | |
| "loss": 0.4825, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.477732793522267, | |
| "grad_norm": 0.4031287815601151, | |
| "learning_rate": 2.379427985200223e-05, | |
| "loss": 0.4777, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.4817813765182186, | |
| "grad_norm": 0.33825826460956754, | |
| "learning_rate": 2.373873749394972e-05, | |
| "loss": 0.4976, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.48582995951417, | |
| "grad_norm": 0.3596226754360303, | |
| "learning_rate": 2.3683165241051836e-05, | |
| "loss": 0.4784, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.4898785425101213, | |
| "grad_norm": 0.34193805773850716, | |
| "learning_rate": 2.3627563537662823e-05, | |
| "loss": 0.4856, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.493927125506073, | |
| "grad_norm": 0.3666892381392251, | |
| "learning_rate": 2.3571932828372433e-05, | |
| "loss": 0.4716, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.4979757085020244, | |
| "grad_norm": 0.3014132233315583, | |
| "learning_rate": 2.351627355800233e-05, | |
| "loss": 0.492, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.5020242914979756, | |
| "grad_norm": 0.3035545499301961, | |
| "learning_rate": 2.346058617160256e-05, | |
| "loss": 0.4882, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.506072874493927, | |
| "grad_norm": 0.32729056933417294, | |
| "learning_rate": 2.3404871114447976e-05, | |
| "loss": 0.4942, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.5101214574898787, | |
| "grad_norm": 0.29413096800304256, | |
| "learning_rate": 2.334912883203469e-05, | |
| "loss": 0.4965, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.51417004048583, | |
| "grad_norm": 0.35936167859832635, | |
| "learning_rate": 2.3293359770076505e-05, | |
| "loss": 0.4978, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.5182186234817814, | |
| "grad_norm": 0.29249210014862026, | |
| "learning_rate": 2.323756437450136e-05, | |
| "loss": 0.4903, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.522267206477733, | |
| "grad_norm": 0.34125361675315413, | |
| "learning_rate": 2.3181743091447754e-05, | |
| "loss": 0.4927, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.2923142512976395, | |
| "learning_rate": 2.3125896367261164e-05, | |
| "loss": 0.4776, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5303643724696356, | |
| "grad_norm": 0.3246172517203694, | |
| "learning_rate": 2.307002464849052e-05, | |
| "loss": 0.4723, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.534412955465587, | |
| "grad_norm": 0.31355119264426834, | |
| "learning_rate": 2.30141283818846e-05, | |
| "loss": 0.4983, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.3065554843923178, | |
| "learning_rate": 2.2958208014388462e-05, | |
| "loss": 0.4905, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.54251012145749, | |
| "grad_norm": 0.3566545417937911, | |
| "learning_rate": 2.2902263993139857e-05, | |
| "loss": 0.4866, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.5465587044534415, | |
| "grad_norm": 0.3305459612422374, | |
| "learning_rate": 2.2846296765465708e-05, | |
| "loss": 0.4761, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.5506072874493926, | |
| "grad_norm": 0.35605056102418065, | |
| "learning_rate": 2.2790306778878465e-05, | |
| "loss": 0.4862, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.554655870445344, | |
| "grad_norm": 0.2896448261323579, | |
| "learning_rate": 2.273429448107257e-05, | |
| "loss": 0.4794, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.5587044534412957, | |
| "grad_norm": 0.3880527497892241, | |
| "learning_rate": 2.2678260319920868e-05, | |
| "loss": 0.4819, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.562753036437247, | |
| "grad_norm": 0.30653457860557154, | |
| "learning_rate": 2.262220474347102e-05, | |
| "loss": 0.4779, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.5668016194331984, | |
| "grad_norm": 0.3606880582691859, | |
| "learning_rate": 2.2566128199941918e-05, | |
| "loss": 0.4843, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.57085020242915, | |
| "grad_norm": 0.3468669542278885, | |
| "learning_rate": 2.2510031137720104e-05, | |
| "loss": 0.4779, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.574898785425101, | |
| "grad_norm": 0.35749417555704177, | |
| "learning_rate": 2.2453914005356206e-05, | |
| "loss": 0.4785, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.3495014773312508, | |
| "learning_rate": 2.2397777251561302e-05, | |
| "loss": 0.4897, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.582995951417004, | |
| "grad_norm": 0.44009991826508665, | |
| "learning_rate": 2.2341621325203393e-05, | |
| "loss": 0.4725, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.5870445344129553, | |
| "grad_norm": 0.2926440116675323, | |
| "learning_rate": 2.2285446675303764e-05, | |
| "loss": 0.4827, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.591093117408907, | |
| "grad_norm": 0.41059504292476057, | |
| "learning_rate": 2.2229253751033423e-05, | |
| "loss": 0.5001, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.5951417004048585, | |
| "grad_norm": 0.3270496141776584, | |
| "learning_rate": 2.2173043001709492e-05, | |
| "loss": 0.4828, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.5991902834008096, | |
| "grad_norm": 0.3918957434418543, | |
| "learning_rate": 2.2116814876791627e-05, | |
| "loss": 0.4713, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.603238866396761, | |
| "grad_norm": 0.3074699261175081, | |
| "learning_rate": 2.2060569825878413e-05, | |
| "loss": 0.4718, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.6072874493927127, | |
| "grad_norm": 0.35435391915803793, | |
| "learning_rate": 2.2004308298703777e-05, | |
| "loss": 0.4758, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.611336032388664, | |
| "grad_norm": 0.326751939504499, | |
| "learning_rate": 2.1948030745133396e-05, | |
| "loss": 0.4814, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.3301911618219912, | |
| "learning_rate": 2.1891737615161083e-05, | |
| "loss": 0.4958, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.619433198380567, | |
| "grad_norm": 0.3482556774420088, | |
| "learning_rate": 2.1835429358905218e-05, | |
| "loss": 0.4821, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.623481781376518, | |
| "grad_norm": 0.3457968717322539, | |
| "learning_rate": 2.17791064266051e-05, | |
| "loss": 0.4918, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.6275303643724697, | |
| "grad_norm": 0.34922479699068504, | |
| "learning_rate": 2.172276926861741e-05, | |
| "loss": 0.469, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.36349147460539827, | |
| "learning_rate": 2.166641833541255e-05, | |
| "loss": 0.4831, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6356275303643724, | |
| "grad_norm": 0.39314877996718806, | |
| "learning_rate": 2.161005407757109e-05, | |
| "loss": 0.4852, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.639676113360324, | |
| "grad_norm": 0.36836755823201545, | |
| "learning_rate": 2.155367694578013e-05, | |
| "loss": 0.4857, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.6437246963562755, | |
| "grad_norm": 0.39652281339582435, | |
| "learning_rate": 2.1497287390829712e-05, | |
| "loss": 0.4871, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.6477732793522266, | |
| "grad_norm": 0.36324946553901877, | |
| "learning_rate": 2.1440885863609208e-05, | |
| "loss": 0.4977, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.651821862348178, | |
| "grad_norm": 0.325504654561306, | |
| "learning_rate": 2.1384472815103737e-05, | |
| "loss": 0.4815, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6558704453441297, | |
| "grad_norm": 0.34509719445788856, | |
| "learning_rate": 2.1328048696390534e-05, | |
| "loss": 0.4919, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.659919028340081, | |
| "grad_norm": 0.33289389281123755, | |
| "learning_rate": 2.1271613958635342e-05, | |
| "loss": 0.4959, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.6639676113360324, | |
| "grad_norm": 0.30097686946979735, | |
| "learning_rate": 2.1215169053088825e-05, | |
| "loss": 0.4653, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.668016194331984, | |
| "grad_norm": 0.384684035236388, | |
| "learning_rate": 2.1158714431082937e-05, | |
| "loss": 0.4875, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.672064777327935, | |
| "grad_norm": 0.29471802008836634, | |
| "learning_rate": 2.110225054402733e-05, | |
| "loss": 0.4822, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.6761133603238867, | |
| "grad_norm": 0.38488869267017617, | |
| "learning_rate": 2.1045777843405747e-05, | |
| "loss": 0.4853, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.6801619433198383, | |
| "grad_norm": 0.3093098523087038, | |
| "learning_rate": 2.09892967807724e-05, | |
| "loss": 0.4738, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.3887444238560476, | |
| "learning_rate": 2.093280780774835e-05, | |
| "loss": 0.4767, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.688259109311741, | |
| "grad_norm": 0.30820236313394606, | |
| "learning_rate": 2.087631137601793e-05, | |
| "loss": 0.478, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.3520143593536292, | |
| "learning_rate": 2.0819807937325104e-05, | |
| "loss": 0.4924, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.6963562753036436, | |
| "grad_norm": 0.2816885470930205, | |
| "learning_rate": 2.0763297943469847e-05, | |
| "loss": 0.4956, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.700404858299595, | |
| "grad_norm": 0.3047438320520578, | |
| "learning_rate": 2.070678184630458e-05, | |
| "loss": 0.49, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.7044534412955468, | |
| "grad_norm": 0.27931313193269336, | |
| "learning_rate": 2.06502600977305e-05, | |
| "loss": 0.4786, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.708502024291498, | |
| "grad_norm": 0.3041397009153398, | |
| "learning_rate": 2.0593733149694008e-05, | |
| "loss": 0.4923, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.7125506072874495, | |
| "grad_norm": 0.2814771232928221, | |
| "learning_rate": 2.0537201454183074e-05, | |
| "loss": 0.4884, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.716599190283401, | |
| "grad_norm": 0.3227904299763256, | |
| "learning_rate": 2.0480665463223626e-05, | |
| "loss": 0.4829, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.720647773279352, | |
| "grad_norm": 0.3178300340936118, | |
| "learning_rate": 2.042412562887594e-05, | |
| "loss": 0.4776, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7246963562753037, | |
| "grad_norm": 0.2658879989153444, | |
| "learning_rate": 2.0367582403231033e-05, | |
| "loss": 0.4838, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.7287449392712553, | |
| "grad_norm": 0.31462779417459064, | |
| "learning_rate": 2.0311036238407023e-05, | |
| "loss": 0.4841, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7327935222672064, | |
| "grad_norm": 0.277866936311208, | |
| "learning_rate": 2.0254487586545534e-05, | |
| "loss": 0.476, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.3215715448737897, | |
| "learning_rate": 2.0197936899808088e-05, | |
| "loss": 0.4766, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.7408906882591095, | |
| "grad_norm": 0.2875857571251426, | |
| "learning_rate": 2.0141384630372466e-05, | |
| "loss": 0.4706, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7449392712550607, | |
| "grad_norm": 0.3241431734750217, | |
| "learning_rate": 2.0084831230429106e-05, | |
| "loss": 0.4917, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.748987854251012, | |
| "grad_norm": 0.30746467569066493, | |
| "learning_rate": 2.0028277152177492e-05, | |
| "loss": 0.4806, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.753036437246964, | |
| "grad_norm": 0.29822397249441795, | |
| "learning_rate": 1.9971722847822518e-05, | |
| "loss": 0.4636, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.757085020242915, | |
| "grad_norm": 0.35038580766020216, | |
| "learning_rate": 1.99151687695709e-05, | |
| "loss": 0.4784, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7611336032388665, | |
| "grad_norm": 0.3358697687752019, | |
| "learning_rate": 1.9858615369627537e-05, | |
| "loss": 0.4813, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.765182186234818, | |
| "grad_norm": 0.31693406878310376, | |
| "learning_rate": 1.9802063100191916e-05, | |
| "loss": 0.4854, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.3491022349607036, | |
| "learning_rate": 1.974551241345447e-05, | |
| "loss": 0.4874, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.7732793522267207, | |
| "grad_norm": 0.3509181799726959, | |
| "learning_rate": 1.9688963761592984e-05, | |
| "loss": 0.4795, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7773279352226723, | |
| "grad_norm": 0.35195624613050674, | |
| "learning_rate": 1.963241759676897e-05, | |
| "loss": 0.4807, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.7813765182186234, | |
| "grad_norm": 0.32682551551821004, | |
| "learning_rate": 1.9575874371124066e-05, | |
| "loss": 0.4893, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.785425101214575, | |
| "grad_norm": 0.2976488176315341, | |
| "learning_rate": 1.951933453677638e-05, | |
| "loss": 0.4858, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.30952878579153437, | |
| "learning_rate": 1.9462798545816932e-05, | |
| "loss": 0.4785, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.7935222672064777, | |
| "grad_norm": 0.3008675141800297, | |
| "learning_rate": 1.9406266850305996e-05, | |
| "loss": 0.4669, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7975708502024292, | |
| "grad_norm": 0.2824045349392727, | |
| "learning_rate": 1.9349739902269503e-05, | |
| "loss": 0.4796, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.801619433198381, | |
| "grad_norm": 0.33882740302370146, | |
| "learning_rate": 1.9293218153695425e-05, | |
| "loss": 0.474, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.805668016194332, | |
| "grad_norm": 0.32219942232578547, | |
| "learning_rate": 1.9236702056530153e-05, | |
| "loss": 0.4686, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.8097165991902835, | |
| "grad_norm": 0.31478150835648566, | |
| "learning_rate": 1.9180192062674907e-05, | |
| "loss": 0.5089, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.813765182186235, | |
| "grad_norm": 0.33138388818461023, | |
| "learning_rate": 1.9123688623982076e-05, | |
| "loss": 0.4687, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.817813765182186, | |
| "grad_norm": 0.29457296088289214, | |
| "learning_rate": 1.906719219225165e-05, | |
| "loss": 0.4785, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.8218623481781377, | |
| "grad_norm": 0.3076939529566145, | |
| "learning_rate": 1.9010703219227603e-05, | |
| "loss": 0.4781, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.8259109311740893, | |
| "grad_norm": 0.32763432467876624, | |
| "learning_rate": 1.8954222156594256e-05, | |
| "loss": 0.4888, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.8299595141700404, | |
| "grad_norm": 0.2832982868710207, | |
| "learning_rate": 1.8897749455972673e-05, | |
| "loss": 0.4719, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.834008097165992, | |
| "grad_norm": 0.31427106790343573, | |
| "learning_rate": 1.884128556891707e-05, | |
| "loss": 0.5005, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8380566801619436, | |
| "grad_norm": 0.30943504001399885, | |
| "learning_rate": 1.878483094691119e-05, | |
| "loss": 0.4673, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.31566890901294437, | |
| "learning_rate": 1.8728386041364664e-05, | |
| "loss": 0.4876, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.3245529732755377, | |
| "learning_rate": 1.867195130360947e-05, | |
| "loss": 0.4858, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.850202429149798, | |
| "grad_norm": 0.31272542941122244, | |
| "learning_rate": 1.8615527184896263e-05, | |
| "loss": 0.4812, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.854251012145749, | |
| "grad_norm": 0.2856025788065554, | |
| "learning_rate": 1.8559114136390795e-05, | |
| "loss": 0.4971, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8582995951417005, | |
| "grad_norm": 0.3118571792835259, | |
| "learning_rate": 1.8502712609170298e-05, | |
| "loss": 0.4948, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.862348178137652, | |
| "grad_norm": 0.27930407729334966, | |
| "learning_rate": 1.8446323054219876e-05, | |
| "loss": 0.4789, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.866396761133603, | |
| "grad_norm": 0.32304122368049953, | |
| "learning_rate": 1.838994592242891e-05, | |
| "loss": 0.4898, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.8704453441295548, | |
| "grad_norm": 0.29688863079751826, | |
| "learning_rate": 1.8333581664587453e-05, | |
| "loss": 0.4659, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.8744939271255063, | |
| "grad_norm": 0.2853577584069504, | |
| "learning_rate": 1.8277230731382593e-05, | |
| "loss": 0.4904, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8785425101214575, | |
| "grad_norm": 0.2817863860313454, | |
| "learning_rate": 1.82208935733949e-05, | |
| "loss": 0.4852, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.882591093117409, | |
| "grad_norm": 0.31637283656100545, | |
| "learning_rate": 1.8164570641094793e-05, | |
| "loss": 0.4901, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.8866396761133606, | |
| "grad_norm": 0.2918159719023521, | |
| "learning_rate": 1.810826238483892e-05, | |
| "loss": 0.4778, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.8906882591093117, | |
| "grad_norm": 0.31238293332178846, | |
| "learning_rate": 1.8051969254866608e-05, | |
| "loss": 0.4788, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.3455577519445618, | |
| "learning_rate": 1.7995691701296226e-05, | |
| "loss": 0.4968, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.898785425101215, | |
| "grad_norm": 0.2797140397010844, | |
| "learning_rate": 1.7939430174121593e-05, | |
| "loss": 0.4786, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.902834008097166, | |
| "grad_norm": 0.2941047817426251, | |
| "learning_rate": 1.788318512320838e-05, | |
| "loss": 0.4632, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.9068825910931175, | |
| "grad_norm": 0.307233667789525, | |
| "learning_rate": 1.782695699829051e-05, | |
| "loss": 0.4742, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.910931174089069, | |
| "grad_norm": 0.28605448800775735, | |
| "learning_rate": 1.7770746248966587e-05, | |
| "loss": 0.4712, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.91497975708502, | |
| "grad_norm": 0.35804084598700425, | |
| "learning_rate": 1.7714553324696243e-05, | |
| "loss": 0.4809, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.919028340080972, | |
| "grad_norm": 0.2974691723504908, | |
| "learning_rate": 1.7658378674796614e-05, | |
| "loss": 0.4854, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.3489026450904155, | |
| "learning_rate": 1.76022227484387e-05, | |
| "loss": 0.4774, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.9271255060728745, | |
| "grad_norm": 0.2875068404234572, | |
| "learning_rate": 1.7546085994643807e-05, | |
| "loss": 0.4974, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.931174089068826, | |
| "grad_norm": 0.3167527435066739, | |
| "learning_rate": 1.7489968862279902e-05, | |
| "loss": 0.4797, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9352226720647776, | |
| "grad_norm": 0.2856467578190936, | |
| "learning_rate": 1.743387180005809e-05, | |
| "loss": 0.4763, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.9392712550607287, | |
| "grad_norm": 0.3075405232762281, | |
| "learning_rate": 1.737779525652899e-05, | |
| "loss": 0.4714, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.9433198380566803, | |
| "grad_norm": 0.3099184726395637, | |
| "learning_rate": 1.732173968007914e-05, | |
| "loss": 0.4665, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.27860054457997296, | |
| "learning_rate": 1.7265705518927437e-05, | |
| "loss": 0.4869, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.951417004048583, | |
| "grad_norm": 0.3131038467966113, | |
| "learning_rate": 1.7209693221121542e-05, | |
| "loss": 0.4637, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.9554655870445345, | |
| "grad_norm": 0.33361265884388336, | |
| "learning_rate": 1.7153703234534302e-05, | |
| "loss": 0.4782, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9595141700404857, | |
| "grad_norm": 0.3023894838232881, | |
| "learning_rate": 1.709773600686015e-05, | |
| "loss": 0.4857, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.9635627530364372, | |
| "grad_norm": 0.35496209149939506, | |
| "learning_rate": 1.7041791985611548e-05, | |
| "loss": 0.4859, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.967611336032389, | |
| "grad_norm": 0.31074535966715583, | |
| "learning_rate": 1.69858716181154e-05, | |
| "loss": 0.4826, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.97165991902834, | |
| "grad_norm": 0.30442416366159136, | |
| "learning_rate": 1.692997535150948e-05, | |
| "loss": 0.4679, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.9757085020242915, | |
| "grad_norm": 0.34774886416726697, | |
| "learning_rate": 1.687410363273884e-05, | |
| "loss": 0.4709, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.979757085020243, | |
| "grad_norm": 0.2587572917174982, | |
| "learning_rate": 1.6818256908552257e-05, | |
| "loss": 0.4923, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.983805668016194, | |
| "grad_norm": 0.42782506325763536, | |
| "learning_rate": 1.6762435625498646e-05, | |
| "loss": 0.4805, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.9878542510121457, | |
| "grad_norm": 0.26812330581026905, | |
| "learning_rate": 1.67066402299235e-05, | |
| "loss": 0.4785, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.9919028340080973, | |
| "grad_norm": 0.34407914938322287, | |
| "learning_rate": 1.6650871167965313e-05, | |
| "loss": 0.475, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.9959514170040484, | |
| "grad_norm": 0.3190147071211478, | |
| "learning_rate": 1.6595128885552028e-05, | |
| "loss": 0.4962, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.29058334246519, | |
| "learning_rate": 1.653941382839745e-05, | |
| "loss": 0.4784, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.0040485829959516, | |
| "grad_norm": 0.4852002617452993, | |
| "learning_rate": 1.6483726441997673e-05, | |
| "loss": 0.4246, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.0080971659919027, | |
| "grad_norm": 0.32986291099468185, | |
| "learning_rate": 1.642806717162757e-05, | |
| "loss": 0.4194, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.0121457489878543, | |
| "grad_norm": 0.7709645778836366, | |
| "learning_rate": 1.637243646233718e-05, | |
| "loss": 0.4211, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.016194331983806, | |
| "grad_norm": 0.49960282932566347, | |
| "learning_rate": 1.6316834758948174e-05, | |
| "loss": 0.41, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.020242914979757, | |
| "grad_norm": 0.4867414189230154, | |
| "learning_rate": 1.6261262506050282e-05, | |
| "loss": 0.4288, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.0242914979757085, | |
| "grad_norm": 0.4151330358938428, | |
| "learning_rate": 1.620572014799777e-05, | |
| "loss": 0.4098, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.02834008097166, | |
| "grad_norm": 0.4409775665736684, | |
| "learning_rate": 1.6150208128905857e-05, | |
| "loss": 0.4213, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.032388663967611, | |
| "grad_norm": 0.47548961241427784, | |
| "learning_rate": 1.6094726892647147e-05, | |
| "loss": 0.4131, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0364372469635628, | |
| "grad_norm": 0.4248162571750495, | |
| "learning_rate": 1.6039276882848135e-05, | |
| "loss": 0.4109, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0404858299595143, | |
| "grad_norm": 0.4199106985148589, | |
| "learning_rate": 1.5983858542885635e-05, | |
| "loss": 0.4115, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.0445344129554655, | |
| "grad_norm": 0.4599508007689629, | |
| "learning_rate": 1.5928472315883203e-05, | |
| "loss": 0.4282, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.048582995951417, | |
| "grad_norm": 0.38995770713070443, | |
| "learning_rate": 1.5873118644707633e-05, | |
| "loss": 0.4116, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.40916634978917427, | |
| "learning_rate": 1.5817797971965413e-05, | |
| "loss": 0.4194, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.0566801619433197, | |
| "grad_norm": 0.4090342361205192, | |
| "learning_rate": 1.576251073999917e-05, | |
| "loss": 0.4205, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.0607287449392713, | |
| "grad_norm": 0.4197476586147074, | |
| "learning_rate": 1.5707257390884126e-05, | |
| "loss": 0.4137, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.064777327935223, | |
| "grad_norm": 0.35434560417383315, | |
| "learning_rate": 1.5652038366424595e-05, | |
| "loss": 0.4182, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.068825910931174, | |
| "grad_norm": 0.45298464517272247, | |
| "learning_rate": 1.5596854108150424e-05, | |
| "loss": 0.4312, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.0728744939271255, | |
| "grad_norm": 0.3433455920438952, | |
| "learning_rate": 1.5541705057313476e-05, | |
| "loss": 0.4252, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.35284959727889725, | |
| "learning_rate": 1.5486591654884086e-05, | |
| "loss": 0.4235, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.080971659919028, | |
| "grad_norm": 0.3566226583609998, | |
| "learning_rate": 1.543151434154755e-05, | |
| "loss": 0.4082, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.08502024291498, | |
| "grad_norm": 0.27884889328181167, | |
| "learning_rate": 1.5376473557700615e-05, | |
| "loss": 0.4213, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.0890688259109313, | |
| "grad_norm": 0.36009405440478903, | |
| "learning_rate": 1.5321469743447893e-05, | |
| "loss": 0.3962, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.0931174089068825, | |
| "grad_norm": 0.3200622221808826, | |
| "learning_rate": 1.5266503338598434e-05, | |
| "loss": 0.4092, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.097165991902834, | |
| "grad_norm": 0.2819336970800046, | |
| "learning_rate": 1.5211574782662149e-05, | |
| "loss": 0.416, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1012145748987856, | |
| "grad_norm": 0.32391062575861806, | |
| "learning_rate": 1.5156684514846307e-05, | |
| "loss": 0.4099, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.32963765318786953, | |
| "learning_rate": 1.5101832974052012e-05, | |
| "loss": 0.4195, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.1093117408906883, | |
| "grad_norm": 0.28421919202418283, | |
| "learning_rate": 1.5047020598870737e-05, | |
| "loss": 0.4248, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.11336032388664, | |
| "grad_norm": 0.31204592586794994, | |
| "learning_rate": 1.4992247827580778e-05, | |
| "loss": 0.4202, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.117408906882591, | |
| "grad_norm": 0.35810090531722716, | |
| "learning_rate": 1.493751509814374e-05, | |
| "loss": 0.4031, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1214574898785425, | |
| "grad_norm": 0.3020395628823993, | |
| "learning_rate": 1.4882822848201073e-05, | |
| "loss": 0.4087, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.125506072874494, | |
| "grad_norm": 0.2873351973856373, | |
| "learning_rate": 1.4828171515070553e-05, | |
| "loss": 0.408, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.1295546558704452, | |
| "grad_norm": 0.30770332517887217, | |
| "learning_rate": 1.4773561535742793e-05, | |
| "loss": 0.416, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.133603238866397, | |
| "grad_norm": 0.28200614543492286, | |
| "learning_rate": 1.4718993346877715e-05, | |
| "loss": 0.4045, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.1376518218623484, | |
| "grad_norm": 0.3033979771741979, | |
| "learning_rate": 1.466446738480111e-05, | |
| "loss": 0.4233, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1417004048582995, | |
| "grad_norm": 0.278098661126838, | |
| "learning_rate": 1.460998408550112e-05, | |
| "loss": 0.4025, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.145748987854251, | |
| "grad_norm": 0.272030749385453, | |
| "learning_rate": 1.4555543884624751e-05, | |
| "loss": 0.4346, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.1497975708502026, | |
| "grad_norm": 0.30515184955642866, | |
| "learning_rate": 1.4501147217474402e-05, | |
| "loss": 0.4283, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.25788854114875615, | |
| "learning_rate": 1.444679451900437e-05, | |
| "loss": 0.4318, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.29552113597931295, | |
| "learning_rate": 1.4392486223817397e-05, | |
| "loss": 0.4057, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.161943319838057, | |
| "grad_norm": 0.2893985012186451, | |
| "learning_rate": 1.4338222766161143e-05, | |
| "loss": 0.4229, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.165991902834008, | |
| "grad_norm": 0.3101548883991377, | |
| "learning_rate": 1.4284004579924777e-05, | |
| "loss": 0.4152, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.1700404858299596, | |
| "grad_norm": 0.2696285179380415, | |
| "learning_rate": 1.4229832098635479e-05, | |
| "loss": 0.4208, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.174089068825911, | |
| "grad_norm": 0.28899100820394713, | |
| "learning_rate": 1.4175705755454963e-05, | |
| "loss": 0.4379, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.1781376518218623, | |
| "grad_norm": 0.2776972534510234, | |
| "learning_rate": 1.412162598317602e-05, | |
| "loss": 0.4281, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.182186234817814, | |
| "grad_norm": 0.265892327080463, | |
| "learning_rate": 1.406759321421907e-05, | |
| "loss": 0.4214, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.1862348178137654, | |
| "grad_norm": 0.28396343854204864, | |
| "learning_rate": 1.401360788062871e-05, | |
| "loss": 0.4138, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.1902834008097165, | |
| "grad_norm": 0.26913791742043913, | |
| "learning_rate": 1.3959670414070204e-05, | |
| "loss": 0.4128, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.194331983805668, | |
| "grad_norm": 0.27930409573602105, | |
| "learning_rate": 1.3905781245826108e-05, | |
| "loss": 0.4196, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.1983805668016196, | |
| "grad_norm": 0.29570996630201335, | |
| "learning_rate": 1.3851940806792778e-05, | |
| "loss": 0.4245, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2024291497975708, | |
| "grad_norm": 0.26032491773196975, | |
| "learning_rate": 1.379814952747693e-05, | |
| "loss": 0.4125, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.2064777327935223, | |
| "grad_norm": 0.29909410746368253, | |
| "learning_rate": 1.3744407837992193e-05, | |
| "loss": 0.4026, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.28585377552255986, | |
| "learning_rate": 1.3690716168055692e-05, | |
| "loss": 0.4219, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.214574898785425, | |
| "grad_norm": 0.2833139332651386, | |
| "learning_rate": 1.363707494698459e-05, | |
| "loss": 0.4201, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.2186234817813766, | |
| "grad_norm": 0.2685427168017872, | |
| "learning_rate": 1.358348460369265e-05, | |
| "loss": 0.4198, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.2226720647773277, | |
| "grad_norm": 0.3094613316856183, | |
| "learning_rate": 1.352994556668684e-05, | |
| "loss": 0.427, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.2267206477732793, | |
| "grad_norm": 0.27683702697892953, | |
| "learning_rate": 1.3476458264063865e-05, | |
| "loss": 0.4201, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.28388644729351614, | |
| "learning_rate": 1.342302312350678e-05, | |
| "loss": 0.4196, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.234817813765182, | |
| "grad_norm": 0.28545540610408343, | |
| "learning_rate": 1.3369640572281537e-05, | |
| "loss": 0.4009, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.2388663967611335, | |
| "grad_norm": 0.2758858763335519, | |
| "learning_rate": 1.3316311037233596e-05, | |
| "loss": 0.4205, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.242914979757085, | |
| "grad_norm": 0.30329265243919973, | |
| "learning_rate": 1.326303494478451e-05, | |
| "loss": 0.411, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.246963562753036, | |
| "grad_norm": 0.26529250078234684, | |
| "learning_rate": 1.3209812720928472e-05, | |
| "loss": 0.4138, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.251012145748988, | |
| "grad_norm": 0.29978768377815374, | |
| "learning_rate": 1.315664479122898e-05, | |
| "loss": 0.412, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.2550607287449393, | |
| "grad_norm": 0.2815392465899716, | |
| "learning_rate": 1.3103531580815378e-05, | |
| "loss": 0.407, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.2591093117408905, | |
| "grad_norm": 0.29752752241299174, | |
| "learning_rate": 1.3050473514379488e-05, | |
| "loss": 0.4119, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.2636980965226637, | |
| "learning_rate": 1.2997471016172188e-05, | |
| "loss": 0.411, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.2672064777327936, | |
| "grad_norm": 0.2846169815095903, | |
| "learning_rate": 1.2944524510000042e-05, | |
| "loss": 0.3974, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.2712550607287447, | |
| "grad_norm": 0.2847241751279045, | |
| "learning_rate": 1.2891634419221908e-05, | |
| "loss": 0.4352, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.2753036437246963, | |
| "grad_norm": 0.29905238270525997, | |
| "learning_rate": 1.2838801166745545e-05, | |
| "loss": 0.4052, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.279352226720648, | |
| "grad_norm": 0.25296346050922386, | |
| "learning_rate": 1.2786025175024216e-05, | |
| "loss": 0.4233, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.283400809716599, | |
| "grad_norm": 0.25733623477676304, | |
| "learning_rate": 1.2733306866053357e-05, | |
| "loss": 0.4115, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.2874493927125505, | |
| "grad_norm": 0.29908323760045097, | |
| "learning_rate": 1.2680646661367163e-05, | |
| "loss": 0.4442, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.291497975708502, | |
| "grad_norm": 0.2941653827104627, | |
| "learning_rate": 1.2628044982035219e-05, | |
| "loss": 0.4065, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.2955465587044532, | |
| "grad_norm": 0.27259698502896407, | |
| "learning_rate": 1.2575502248659155e-05, | |
| "loss": 0.3976, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.299595141700405, | |
| "grad_norm": 0.28750677459358737, | |
| "learning_rate": 1.2523018881369269e-05, | |
| "loss": 0.4258, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.3036437246963564, | |
| "grad_norm": 0.2626149200847516, | |
| "learning_rate": 1.2470595299821176e-05, | |
| "loss": 0.4205, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.28633017738894423, | |
| "learning_rate": 1.2418231923192416e-05, | |
| "loss": 0.4254, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.311740890688259, | |
| "grad_norm": 0.2775345500862741, | |
| "learning_rate": 1.2365929170179165e-05, | |
| "loss": 0.4148, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.2789073625148514, | |
| "learning_rate": 1.2313687458992848e-05, | |
| "loss": 0.4282, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.3198380566801617, | |
| "grad_norm": 0.27281947887199787, | |
| "learning_rate": 1.2261507207356785e-05, | |
| "loss": 0.4258, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.3238866396761133, | |
| "grad_norm": 0.26719471969226294, | |
| "learning_rate": 1.220938883250288e-05, | |
| "loss": 0.4127, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.327935222672065, | |
| "grad_norm": 0.2626825401689675, | |
| "learning_rate": 1.215733275116827e-05, | |
| "loss": 0.4172, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.331983805668016, | |
| "grad_norm": 0.28834081400741257, | |
| "learning_rate": 1.2105339379592005e-05, | |
| "loss": 0.4068, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.3360323886639676, | |
| "grad_norm": 0.2593948402788786, | |
| "learning_rate": 1.2053409133511681e-05, | |
| "loss": 0.4258, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.340080971659919, | |
| "grad_norm": 0.26281284626700396, | |
| "learning_rate": 1.2001542428160173e-05, | |
| "loss": 0.4144, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.3441295546558703, | |
| "grad_norm": 0.27520659221033217, | |
| "learning_rate": 1.1949739678262278e-05, | |
| "loss": 0.4124, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.348178137651822, | |
| "grad_norm": 0.26410367907777504, | |
| "learning_rate": 1.189800129803139e-05, | |
| "loss": 0.4131, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.3522267206477734, | |
| "grad_norm": 0.2744687562515528, | |
| "learning_rate": 1.1846327701166228e-05, | |
| "loss": 0.4229, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.3562753036437245, | |
| "grad_norm": 0.2813872474601481, | |
| "learning_rate": 1.1794719300847493e-05, | |
| "loss": 0.4143, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.360323886639676, | |
| "grad_norm": 0.2649635843302524, | |
| "learning_rate": 1.1743176509734582e-05, | |
| "loss": 0.3924, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.3643724696356276, | |
| "grad_norm": 0.3117929884248612, | |
| "learning_rate": 1.1691699739962275e-05, | |
| "loss": 0.4127, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.2618016010123577, | |
| "learning_rate": 1.1640289403137456e-05, | |
| "loss": 0.4205, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.3724696356275303, | |
| "grad_norm": 0.2571925896709088, | |
| "learning_rate": 1.1588945910335802e-05, | |
| "loss": 0.4196, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.376518218623482, | |
| "grad_norm": 0.26902215493240156, | |
| "learning_rate": 1.1537669672098517e-05, | |
| "loss": 0.4048, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.380566801619433, | |
| "grad_norm": 0.23072054505841688, | |
| "learning_rate": 1.1486461098429036e-05, | |
| "loss": 0.4287, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.2620743288218488, | |
| "learning_rate": 1.1435320598789747e-05, | |
| "loss": 0.4225, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.388663967611336, | |
| "grad_norm": 0.2691914181859769, | |
| "learning_rate": 1.1384248582098736e-05, | |
| "loss": 0.4204, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.3927125506072873, | |
| "grad_norm": 0.26838104502636273, | |
| "learning_rate": 1.1333245456726467e-05, | |
| "loss": 0.4203, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.396761133603239, | |
| "grad_norm": 0.28114049216369325, | |
| "learning_rate": 1.128231163049258e-05, | |
| "loss": 0.4318, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.4008097165991904, | |
| "grad_norm": 0.27904158074015917, | |
| "learning_rate": 1.123144751066259e-05, | |
| "loss": 0.4095, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4048582995951415, | |
| "grad_norm": 0.24727612194916695, | |
| "learning_rate": 1.1180653503944657e-05, | |
| "loss": 0.4156, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.408906882591093, | |
| "grad_norm": 0.2524137837562779, | |
| "learning_rate": 1.1129930016486287e-05, | |
| "loss": 0.4129, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.4129554655870447, | |
| "grad_norm": 0.26472482093984806, | |
| "learning_rate": 1.107927745387114e-05, | |
| "loss": 0.4064, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.417004048582996, | |
| "grad_norm": 0.24558320515190007, | |
| "learning_rate": 1.1028696221115766e-05, | |
| "loss": 0.3969, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.2511066474679512, | |
| "learning_rate": 1.0978186722666341e-05, | |
| "loss": 0.4048, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.425101214574899, | |
| "grad_norm": 0.26809938699089275, | |
| "learning_rate": 1.0927749362395475e-05, | |
| "loss": 0.4123, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.42914979757085, | |
| "grad_norm": 0.24377114537371256, | |
| "learning_rate": 1.0877384543598954e-05, | |
| "loss": 0.4163, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.4331983805668016, | |
| "grad_norm": 0.25931215095266547, | |
| "learning_rate": 1.0827092668992529e-05, | |
| "loss": 0.4204, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.437246963562753, | |
| "grad_norm": 0.2500698429617878, | |
| "learning_rate": 1.077687414070869e-05, | |
| "loss": 0.4142, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.4412955465587043, | |
| "grad_norm": 0.2527860865152629, | |
| "learning_rate": 1.0726729360293451e-05, | |
| "loss": 0.4192, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.445344129554656, | |
| "grad_norm": 0.2658504045425895, | |
| "learning_rate": 1.0676658728703147e-05, | |
| "loss": 0.4195, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.4493927125506074, | |
| "grad_norm": 0.25143047185657597, | |
| "learning_rate": 1.0626662646301198e-05, | |
| "loss": 0.4163, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.4534412955465585, | |
| "grad_norm": 0.258642018549662, | |
| "learning_rate": 1.0576741512854958e-05, | |
| "loss": 0.3983, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.45748987854251, | |
| "grad_norm": 0.2635973401255197, | |
| "learning_rate": 1.052689572753247e-05, | |
| "loss": 0.422, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.24009771041774305, | |
| "learning_rate": 1.0477125688899328e-05, | |
| "loss": 0.4248, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.465587044534413, | |
| "grad_norm": 0.2435659741404332, | |
| "learning_rate": 1.0427431794915406e-05, | |
| "loss": 0.4246, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.4696356275303644, | |
| "grad_norm": 0.2760508644285671, | |
| "learning_rate": 1.0377814442931769e-05, | |
| "loss": 0.4074, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.2550434604225144, | |
| "learning_rate": 1.0328274029687439e-05, | |
| "loss": 0.4102, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.477732793522267, | |
| "grad_norm": 0.26319331596504447, | |
| "learning_rate": 1.0278810951306253e-05, | |
| "loss": 0.4228, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.4817813765182186, | |
| "grad_norm": 0.266654337755455, | |
| "learning_rate": 1.0229425603293647e-05, | |
| "loss": 0.4194, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.48582995951417, | |
| "grad_norm": 0.27173103089079087, | |
| "learning_rate": 1.0180118380533558e-05, | |
| "loss": 0.4174, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.4898785425101213, | |
| "grad_norm": 0.24116290919800826, | |
| "learning_rate": 1.0130889677285225e-05, | |
| "loss": 0.4186, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.493927125506073, | |
| "grad_norm": 0.2576292024117734, | |
| "learning_rate": 1.008173988718005e-05, | |
| "loss": 0.4197, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.4979757085020244, | |
| "grad_norm": 0.23631171598030654, | |
| "learning_rate": 1.0032669403218439e-05, | |
| "loss": 0.4282, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.5020242914979756, | |
| "grad_norm": 0.2544417946763229, | |
| "learning_rate": 9.983678617766677e-06, | |
| "loss": 0.4162, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.506072874493927, | |
| "grad_norm": 0.2715382827462546, | |
| "learning_rate": 9.934767922553788e-06, | |
| "loss": 0.4184, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.5101214574898787, | |
| "grad_norm": 0.24536985851927426, | |
| "learning_rate": 9.885937708668371e-06, | |
| "loss": 0.4062, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.51417004048583, | |
| "grad_norm": 0.2843628784876327, | |
| "learning_rate": 9.837188366555524e-06, | |
| "loss": 0.4169, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.5182186234817814, | |
| "grad_norm": 0.25521444237209767, | |
| "learning_rate": 9.788520286013691e-06, | |
| "loss": 0.4037, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.522267206477733, | |
| "grad_norm": 0.24430924493526104, | |
| "learning_rate": 9.739933856191537e-06, | |
| "loss": 0.4127, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.2587616995440698, | |
| "learning_rate": 9.691429465584866e-06, | |
| "loss": 0.4026, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.5303643724696356, | |
| "grad_norm": 0.24799534782114344, | |
| "learning_rate": 9.643007502033492e-06, | |
| "loss": 0.4189, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.534412955465587, | |
| "grad_norm": 0.258518582439613, | |
| "learning_rate": 9.594668352718152e-06, | |
| "loss": 0.4182, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.24670437741090856, | |
| "learning_rate": 9.546412404157383e-06, | |
| "loss": 0.4151, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.54251012145749, | |
| "grad_norm": 0.2433093013883759, | |
| "learning_rate": 9.498240042204471e-06, | |
| "loss": 0.44, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.5465587044534415, | |
| "grad_norm": 0.25085801950369013, | |
| "learning_rate": 9.450151652044339e-06, | |
| "loss": 0.4168, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.5506072874493926, | |
| "grad_norm": 0.2315058340254269, | |
| "learning_rate": 9.402147618190472e-06, | |
| "loss": 0.4192, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.554655870445344, | |
| "grad_norm": 0.26093559518320314, | |
| "learning_rate": 9.354228324481841e-06, | |
| "loss": 0.4146, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.5587044534412957, | |
| "grad_norm": 0.2907308399654384, | |
| "learning_rate": 9.306394154079852e-06, | |
| "loss": 0.4073, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.562753036437247, | |
| "grad_norm": 0.2567607063375344, | |
| "learning_rate": 9.258645489465254e-06, | |
| "loss": 0.4078, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.5668016194331984, | |
| "grad_norm": 0.2389586558331976, | |
| "learning_rate": 9.21098271243509e-06, | |
| "loss": 0.4125, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.57085020242915, | |
| "grad_norm": 0.2407033628086288, | |
| "learning_rate": 9.163406204099659e-06, | |
| "loss": 0.4163, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.574898785425101, | |
| "grad_norm": 0.24618145872605418, | |
| "learning_rate": 9.115916344879452e-06, | |
| "loss": 0.4075, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.2507495776960942, | |
| "learning_rate": 9.068513514502126e-06, | |
| "loss": 0.4232, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.582995951417004, | |
| "grad_norm": 0.2520849429126169, | |
| "learning_rate": 9.021198091999439e-06, | |
| "loss": 0.3967, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.5870445344129553, | |
| "grad_norm": 0.2505734242773023, | |
| "learning_rate": 8.973970455704251e-06, | |
| "loss": 0.4004, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.591093117408907, | |
| "grad_norm": 0.27156432935276337, | |
| "learning_rate": 8.926830983247495e-06, | |
| "loss": 0.3989, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.5951417004048585, | |
| "grad_norm": 0.2608624752298989, | |
| "learning_rate": 8.879780051555116e-06, | |
| "loss": 0.4083, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.5991902834008096, | |
| "grad_norm": 0.2488400335299692, | |
| "learning_rate": 8.83281803684512e-06, | |
| "loss": 0.4222, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.603238866396761, | |
| "grad_norm": 0.2558327681188481, | |
| "learning_rate": 8.785945314624532e-06, | |
| "loss": 0.4053, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.6072874493927127, | |
| "grad_norm": 0.2546484946441488, | |
| "learning_rate": 8.739162259686385e-06, | |
| "loss": 0.4055, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.611336032388664, | |
| "grad_norm": 0.2601725875652405, | |
| "learning_rate": 8.692469246106734e-06, | |
| "loss": 0.3986, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.2867445533902131, | |
| "learning_rate": 8.645866647241677e-06, | |
| "loss": 0.4368, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.619433198380567, | |
| "grad_norm": 0.278439981584336, | |
| "learning_rate": 8.599354835724348e-06, | |
| "loss": 0.4302, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.623481781376518, | |
| "grad_norm": 0.2618555708278032, | |
| "learning_rate": 8.552934183461958e-06, | |
| "loss": 0.4306, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6275303643724697, | |
| "grad_norm": 0.27947905017286206, | |
| "learning_rate": 8.506605061632805e-06, | |
| "loss": 0.4111, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.24697719459241682, | |
| "learning_rate": 8.460367840683308e-06, | |
| "loss": 0.4388, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.6356275303643724, | |
| "grad_norm": 0.27174539719239194, | |
| "learning_rate": 8.414222890325065e-06, | |
| "loss": 0.4128, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.639676113360324, | |
| "grad_norm": 0.2410919456723375, | |
| "learning_rate": 8.368170579531855e-06, | |
| "loss": 0.4027, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.6437246963562755, | |
| "grad_norm": 0.25021451076812184, | |
| "learning_rate": 8.322211276536733e-06, | |
| "loss": 0.4238, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6477732793522266, | |
| "grad_norm": 0.2657700377890334, | |
| "learning_rate": 8.27634534882907e-06, | |
| "loss": 0.4286, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.651821862348178, | |
| "grad_norm": 0.24156492946448097, | |
| "learning_rate": 8.230573163151585e-06, | |
| "loss": 0.4155, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.6558704453441297, | |
| "grad_norm": 0.2594742649682914, | |
| "learning_rate": 8.184895085497466e-06, | |
| "loss": 0.4302, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.659919028340081, | |
| "grad_norm": 0.24047295000985097, | |
| "learning_rate": 8.139311481107401e-06, | |
| "loss": 0.4114, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.6639676113360324, | |
| "grad_norm": 0.24172468260504001, | |
| "learning_rate": 8.093822714466686e-06, | |
| "loss": 0.4097, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.668016194331984, | |
| "grad_norm": 0.25349780970761243, | |
| "learning_rate": 8.048429149302271e-06, | |
| "loss": 0.4192, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.672064777327935, | |
| "grad_norm": 0.2551343227908141, | |
| "learning_rate": 8.003131148579902e-06, | |
| "loss": 0.4055, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.6761133603238867, | |
| "grad_norm": 0.24437647044286265, | |
| "learning_rate": 7.957929074501182e-06, | |
| "loss": 0.4381, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.6801619433198383, | |
| "grad_norm": 0.2315457447496859, | |
| "learning_rate": 7.912823288500695e-06, | |
| "loss": 0.4199, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.26202846125142193, | |
| "learning_rate": 7.867814151243102e-06, | |
| "loss": 0.4203, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.688259109311741, | |
| "grad_norm": 0.25507245802961104, | |
| "learning_rate": 7.822902022620267e-06, | |
| "loss": 0.4112, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.2590152818772919, | |
| "learning_rate": 7.778087261748377e-06, | |
| "loss": 0.4119, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.6963562753036436, | |
| "grad_norm": 0.23669488449937295, | |
| "learning_rate": 7.733370226965056e-06, | |
| "loss": 0.4246, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.700404858299595, | |
| "grad_norm": 0.25475691901014386, | |
| "learning_rate": 7.688751275826527e-06, | |
| "loss": 0.405, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.7044534412955468, | |
| "grad_norm": 0.26568490981450993, | |
| "learning_rate": 7.644230765104735e-06, | |
| "loss": 0.4172, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.708502024291498, | |
| "grad_norm": 0.2455252582063566, | |
| "learning_rate": 7.599809050784502e-06, | |
| "loss": 0.4079, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.7125506072874495, | |
| "grad_norm": 0.2450954300928559, | |
| "learning_rate": 7.555486488060661e-06, | |
| "loss": 0.4238, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.716599190283401, | |
| "grad_norm": 0.24239424400350965, | |
| "learning_rate": 7.511263431335249e-06, | |
| "loss": 0.4149, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.720647773279352, | |
| "grad_norm": 0.25131912663109746, | |
| "learning_rate": 7.467140234214656e-06, | |
| "loss": 0.4223, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.7246963562753037, | |
| "grad_norm": 0.24187260612511852, | |
| "learning_rate": 7.423117249506777e-06, | |
| "loss": 0.416, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7287449392712553, | |
| "grad_norm": 0.24795668336111396, | |
| "learning_rate": 7.379194829218231e-06, | |
| "loss": 0.3995, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.7327935222672064, | |
| "grad_norm": 0.23120091027789744, | |
| "learning_rate": 7.335373324551518e-06, | |
| "loss": 0.4205, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.2330515571434964, | |
| "learning_rate": 7.291653085902224e-06, | |
| "loss": 0.4165, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.7408906882591095, | |
| "grad_norm": 0.23738225366751256, | |
| "learning_rate": 7.248034462856213e-06, | |
| "loss": 0.4297, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.7449392712550607, | |
| "grad_norm": 0.26170438021567677, | |
| "learning_rate": 7.20451780418683e-06, | |
| "loss": 0.3998, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.748987854251012, | |
| "grad_norm": 0.23816494266347113, | |
| "learning_rate": 7.161103457852128e-06, | |
| "loss": 0.4104, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.753036437246964, | |
| "grad_norm": 0.22056394601427598, | |
| "learning_rate": 7.117791770992046e-06, | |
| "loss": 0.4137, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.757085020242915, | |
| "grad_norm": 0.24311473441776643, | |
| "learning_rate": 7.074583089925688e-06, | |
| "loss": 0.4248, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.7611336032388665, | |
| "grad_norm": 0.26123404941468287, | |
| "learning_rate": 7.0314777601485155e-06, | |
| "loss": 0.4094, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.765182186234818, | |
| "grad_norm": 0.2465462540183578, | |
| "learning_rate": 6.988476126329604e-06, | |
| "loss": 0.4094, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.24400844404900224, | |
| "learning_rate": 6.945578532308857e-06, | |
| "loss": 0.4214, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.7732793522267207, | |
| "grad_norm": 0.23607742420954472, | |
| "learning_rate": 6.902785321094301e-06, | |
| "loss": 0.4243, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.7773279352226723, | |
| "grad_norm": 0.24210362958688797, | |
| "learning_rate": 6.8600968348593115e-06, | |
| "loss": 0.4214, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.7813765182186234, | |
| "grad_norm": 0.23636034010212947, | |
| "learning_rate": 6.817513414939892e-06, | |
| "loss": 0.4103, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.785425101214575, | |
| "grad_norm": 0.29013260375072036, | |
| "learning_rate": 6.775035401831924e-06, | |
| "loss": 0.4269, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.27406760323470314, | |
| "learning_rate": 6.732663135188473e-06, | |
| "loss": 0.4126, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.7935222672064777, | |
| "grad_norm": 0.24045754206203387, | |
| "learning_rate": 6.690396953817055e-06, | |
| "loss": 0.4081, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.7975708502024292, | |
| "grad_norm": 0.2574833069954563, | |
| "learning_rate": 6.648237195676927e-06, | |
| "loss": 0.4103, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.801619433198381, | |
| "grad_norm": 0.2949162459932832, | |
| "learning_rate": 6.606184197876395e-06, | |
| "loss": 0.4082, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.805668016194332, | |
| "grad_norm": 0.2427403819469884, | |
| "learning_rate": 6.564238296670109e-06, | |
| "loss": 0.4123, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.8097165991902835, | |
| "grad_norm": 0.24241469773524746, | |
| "learning_rate": 6.52239982745638e-06, | |
| "loss": 0.4037, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.813765182186235, | |
| "grad_norm": 0.2580135730120143, | |
| "learning_rate": 6.480669124774484e-06, | |
| "loss": 0.4135, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.817813765182186, | |
| "grad_norm": 0.26516638781281304, | |
| "learning_rate": 6.439046522302013e-06, | |
| "loss": 0.4023, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.8218623481781377, | |
| "grad_norm": 0.24230294139712136, | |
| "learning_rate": 6.397532352852191e-06, | |
| "loss": 0.4113, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.8259109311740893, | |
| "grad_norm": 0.242006563816342, | |
| "learning_rate": 6.356126948371202e-06, | |
| "loss": 0.4291, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.8299595141700404, | |
| "grad_norm": 0.271139964513911, | |
| "learning_rate": 6.314830639935559e-06, | |
| "loss": 0.4112, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.834008097165992, | |
| "grad_norm": 0.2519922776611533, | |
| "learning_rate": 6.273643757749448e-06, | |
| "loss": 0.4097, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.8380566801619436, | |
| "grad_norm": 0.22871659566170832, | |
| "learning_rate": 6.2325666311420856e-06, | |
| "loss": 0.3974, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.26318767394698533, | |
| "learning_rate": 6.1915995885650675e-06, | |
| "loss": 0.4085, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.27255093648547685, | |
| "learning_rate": 6.150742957589784e-06, | |
| "loss": 0.4312, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.850202429149798, | |
| "grad_norm": 0.22602417221993584, | |
| "learning_rate": 6.109997064904774e-06, | |
| "loss": 0.4283, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.854251012145749, | |
| "grad_norm": 0.2576948040362773, | |
| "learning_rate": 6.069362236313092e-06, | |
| "loss": 0.4161, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.8582995951417005, | |
| "grad_norm": 0.24593167285805276, | |
| "learning_rate": 6.0288387967297565e-06, | |
| "loss": 0.4228, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.862348178137652, | |
| "grad_norm": 0.23948763515347704, | |
| "learning_rate": 5.988427070179108e-06, | |
| "loss": 0.4113, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.866396761133603, | |
| "grad_norm": 0.24907347238091787, | |
| "learning_rate": 5.948127379792234e-06, | |
| "loss": 0.422, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.8704453441295548, | |
| "grad_norm": 0.23839824414459468, | |
| "learning_rate": 5.90794004780439e-06, | |
| "loss": 0.4074, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.8744939271255063, | |
| "grad_norm": 0.2501393012743726, | |
| "learning_rate": 5.867865395552405e-06, | |
| "loss": 0.4226, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.8785425101214575, | |
| "grad_norm": 0.24072965474910124, | |
| "learning_rate": 5.8279037434721364e-06, | |
| "loss": 0.4124, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.882591093117409, | |
| "grad_norm": 0.23554115360490796, | |
| "learning_rate": 5.788055411095892e-06, | |
| "loss": 0.4198, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.8866396761133606, | |
| "grad_norm": 0.23620112792022624, | |
| "learning_rate": 5.7483207170498625e-06, | |
| "loss": 0.4067, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8906882591093117, | |
| "grad_norm": 0.24108869758134996, | |
| "learning_rate": 5.708699979051613e-06, | |
| "loss": 0.4082, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.22999827211171686, | |
| "learning_rate": 5.66919351390751e-06, | |
| "loss": 0.4143, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.898785425101215, | |
| "grad_norm": 0.219839299102655, | |
| "learning_rate": 5.629801637510188e-06, | |
| "loss": 0.418, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.902834008097166, | |
| "grad_norm": 0.2382661645254843, | |
| "learning_rate": 5.590524664836043e-06, | |
| "loss": 0.4102, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.9068825910931175, | |
| "grad_norm": 0.23494821764701931, | |
| "learning_rate": 5.551362909942708e-06, | |
| "loss": 0.4208, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.910931174089069, | |
| "grad_norm": 0.23407189915708193, | |
| "learning_rate": 5.51231668596653e-06, | |
| "loss": 0.4127, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.91497975708502, | |
| "grad_norm": 0.21308654791158174, | |
| "learning_rate": 5.47338630512007e-06, | |
| "loss": 0.3998, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.919028340080972, | |
| "grad_norm": 0.233502477272743, | |
| "learning_rate": 5.434572078689617e-06, | |
| "loss": 0.4348, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.25595385876488286, | |
| "learning_rate": 5.395874317032694e-06, | |
| "loss": 0.4081, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.9271255060728745, | |
| "grad_norm": 0.240729260157137, | |
| "learning_rate": 5.357293329575573e-06, | |
| "loss": 0.4211, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.931174089068826, | |
| "grad_norm": 0.22852772481671998, | |
| "learning_rate": 5.318829424810797e-06, | |
| "loss": 0.4088, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.9352226720647776, | |
| "grad_norm": 0.2462832113382296, | |
| "learning_rate": 5.280482910294728e-06, | |
| "loss": 0.4075, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.9392712550607287, | |
| "grad_norm": 0.24381076417971487, | |
| "learning_rate": 5.242254092645078e-06, | |
| "loss": 0.4196, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.9433198380566803, | |
| "grad_norm": 0.23933963930563265, | |
| "learning_rate": 5.2041432775384445e-06, | |
| "loss": 0.4021, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.22834762523795654, | |
| "learning_rate": 5.166150769707892e-06, | |
| "loss": 0.4069, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.951417004048583, | |
| "grad_norm": 0.2377334022884601, | |
| "learning_rate": 5.128276872940503e-06, | |
| "loss": 0.4185, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.9554655870445345, | |
| "grad_norm": 0.2347255180363036, | |
| "learning_rate": 5.090521890074953e-06, | |
| "loss": 0.4082, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.9595141700404857, | |
| "grad_norm": 0.2390997586490745, | |
| "learning_rate": 5.0528861229990675e-06, | |
| "loss": 0.4233, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.9635627530364372, | |
| "grad_norm": 0.2298889704694621, | |
| "learning_rate": 5.015369872647442e-06, | |
| "loss": 0.4098, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.967611336032389, | |
| "grad_norm": 0.22238496678410072, | |
| "learning_rate": 4.9779734389990195e-06, | |
| "loss": 0.4171, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97165991902834, | |
| "grad_norm": 0.2479767042346718, | |
| "learning_rate": 4.940697121074675e-06, | |
| "loss": 0.3986, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.9757085020242915, | |
| "grad_norm": 0.22385780710178893, | |
| "learning_rate": 4.903541216934864e-06, | |
| "loss": 0.4178, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.979757085020243, | |
| "grad_norm": 0.2292304958428387, | |
| "learning_rate": 4.8665060236772e-06, | |
| "loss": 0.4194, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.983805668016194, | |
| "grad_norm": 0.22664576273054665, | |
| "learning_rate": 4.829591837434102e-06, | |
| "loss": 0.4155, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.9878542510121457, | |
| "grad_norm": 0.2300953487018585, | |
| "learning_rate": 4.792798953370424e-06, | |
| "loss": 0.4216, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.9919028340080973, | |
| "grad_norm": 0.2370184013891261, | |
| "learning_rate": 4.756127665681083e-06, | |
| "loss": 0.3988, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.9959514170040484, | |
| "grad_norm": 0.22448569988259964, | |
| "learning_rate": 4.719578267588725e-06, | |
| "loss": 0.4167, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.22125162049902486, | |
| "learning_rate": 4.6831510513413525e-06, | |
| "loss": 0.4165, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.004048582995951, | |
| "grad_norm": 0.39252950179589735, | |
| "learning_rate": 4.6468463082100176e-06, | |
| "loss": 0.3891, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.008097165991903, | |
| "grad_norm": 0.3182662465780301, | |
| "learning_rate": 4.610664328486485e-06, | |
| "loss": 0.3724, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.012145748987854, | |
| "grad_norm": 0.25131691610959145, | |
| "learning_rate": 4.5746054014809e-06, | |
| "loss": 0.3793, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.016194331983805, | |
| "grad_norm": 0.4417281290195976, | |
| "learning_rate": 4.5386698155194766e-06, | |
| "loss": 0.3804, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.020242914979757, | |
| "grad_norm": 0.36996354439193124, | |
| "learning_rate": 4.502857857942204e-06, | |
| "loss": 0.353, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.0242914979757085, | |
| "grad_norm": 0.2836384430442476, | |
| "learning_rate": 4.467169815100549e-06, | |
| "loss": 0.3677, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.02834008097166, | |
| "grad_norm": 0.3078431295232452, | |
| "learning_rate": 4.431605972355137e-06, | |
| "loss": 0.3663, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.032388663967612, | |
| "grad_norm": 0.35300679681636177, | |
| "learning_rate": 4.396166614073516e-06, | |
| "loss": 0.3801, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.036437246963563, | |
| "grad_norm": 0.2802759686090966, | |
| "learning_rate": 4.360852023627846e-06, | |
| "loss": 0.3741, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.040485829959514, | |
| "grad_norm": 0.2409136161971281, | |
| "learning_rate": 4.325662483392656e-06, | |
| "loss": 0.3627, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.044534412955466, | |
| "grad_norm": 0.27905983212925745, | |
| "learning_rate": 4.29059827474257e-06, | |
| "loss": 0.3795, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.048582995951417, | |
| "grad_norm": 0.30227369994161446, | |
| "learning_rate": 4.255659678050063e-06, | |
| "loss": 0.3585, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.28628146753272443, | |
| "learning_rate": 4.22084697268323e-06, | |
| "loss": 0.3817, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.05668016194332, | |
| "grad_norm": 0.2570601916532132, | |
| "learning_rate": 4.186160437003535e-06, | |
| "loss": 0.3844, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.060728744939271, | |
| "grad_norm": 0.2561756214186277, | |
| "learning_rate": 4.151600348363582e-06, | |
| "loss": 0.3589, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.064777327935222, | |
| "grad_norm": 0.2736495478866887, | |
| "learning_rate": 4.11716698310493e-06, | |
| "loss": 0.3757, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.068825910931174, | |
| "grad_norm": 0.2659276360178028, | |
| "learning_rate": 4.082860616555848e-06, | |
| "loss": 0.3595, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.0728744939271255, | |
| "grad_norm": 0.2856278105594851, | |
| "learning_rate": 4.048681523029125e-06, | |
| "loss": 0.3597, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.2434318051947398, | |
| "learning_rate": 4.014629975819882e-06, | |
| "loss": 0.3559, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.080971659919029, | |
| "grad_norm": 0.2589894448799751, | |
| "learning_rate": 3.9807062472033855e-06, | |
| "loss": 0.3625, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.08502024291498, | |
| "grad_norm": 0.26685895974880386, | |
| "learning_rate": 3.94691060843287e-06, | |
| "loss": 0.3789, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.089068825910931, | |
| "grad_norm": 0.2468777206926549, | |
| "learning_rate": 3.913243329737353e-06, | |
| "loss": 0.369, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.093117408906883, | |
| "grad_norm": 0.2438292235073446, | |
| "learning_rate": 3.879704680319503e-06, | |
| "loss": 0.3694, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.097165991902834, | |
| "grad_norm": 0.22884680700339524, | |
| "learning_rate": 3.846294928353471e-06, | |
| "loss": 0.3886, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.101214574898785, | |
| "grad_norm": 0.235718920738506, | |
| "learning_rate": 3.8130143409827345e-06, | |
| "loss": 0.3721, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.2520010610484568, | |
| "learning_rate": 3.7798631843179887e-06, | |
| "loss": 0.3504, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.109311740890688, | |
| "grad_norm": 0.22937642188826762, | |
| "learning_rate": 3.7468417234349973e-06, | |
| "loss": 0.3529, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.113360323886639, | |
| "grad_norm": 0.24679696525547187, | |
| "learning_rate": 3.7139502223724823e-06, | |
| "loss": 0.3972, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.117408906882591, | |
| "grad_norm": 0.23974182433427418, | |
| "learning_rate": 3.6811889441300095e-06, | |
| "loss": 0.3738, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.1214574898785425, | |
| "grad_norm": 0.24700991379250686, | |
| "learning_rate": 3.648558150665886e-06, | |
| "loss": 0.3816, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.125506072874494, | |
| "grad_norm": 0.22821850591271442, | |
| "learning_rate": 3.616058102895075e-06, | |
| "loss": 0.3906, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.129554655870446, | |
| "grad_norm": 0.23074519813149919, | |
| "learning_rate": 3.583689060687079e-06, | |
| "loss": 0.3699, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.133603238866397, | |
| "grad_norm": 0.2143724252871462, | |
| "learning_rate": 3.5514512828639023e-06, | |
| "loss": 0.3801, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.137651821862348, | |
| "grad_norm": 0.22740796868594398, | |
| "learning_rate": 3.51934502719796e-06, | |
| "loss": 0.3589, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.1417004048583, | |
| "grad_norm": 0.22297795461056738, | |
| "learning_rate": 3.4873705504100185e-06, | |
| "loss": 0.358, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.145748987854251, | |
| "grad_norm": 0.22547206686339066, | |
| "learning_rate": 3.4555281081671345e-06, | |
| "loss": 0.3834, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.149797570850202, | |
| "grad_norm": 0.21900422330682578, | |
| "learning_rate": 3.4238179550806326e-06, | |
| "loss": 0.377, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.2275693696676007, | |
| "learning_rate": 3.3922403447040497e-06, | |
| "loss": 0.3586, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.21774693379039112, | |
| "learning_rate": 3.3607955295311247e-06, | |
| "loss": 0.3599, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.161943319838056, | |
| "grad_norm": 0.2169866935996568, | |
| "learning_rate": 3.32948376099375e-06, | |
| "loss": 0.3671, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.165991902834008, | |
| "grad_norm": 0.2174647084080635, | |
| "learning_rate": 3.2983052894599976e-06, | |
| "loss": 0.3649, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.17004048582996, | |
| "grad_norm": 0.19988101384709261, | |
| "learning_rate": 3.2672603642320945e-06, | |
| "loss": 0.3685, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.174089068825911, | |
| "grad_norm": 0.23809563676754059, | |
| "learning_rate": 3.236349233544438e-06, | |
| "loss": 0.3619, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.178137651821863, | |
| "grad_norm": 0.2153068250894809, | |
| "learning_rate": 3.205572144561604e-06, | |
| "loss": 0.3729, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.182186234817814, | |
| "grad_norm": 0.21802681798832452, | |
| "learning_rate": 3.174929343376374e-06, | |
| "loss": 0.3634, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.186234817813765, | |
| "grad_norm": 0.21605038746273808, | |
| "learning_rate": 3.1444210750077777e-06, | |
| "loss": 0.3989, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.190283400809717, | |
| "grad_norm": 0.21660572437456413, | |
| "learning_rate": 3.1140475833991046e-06, | |
| "loss": 0.375, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.194331983805668, | |
| "grad_norm": 0.22735452089637245, | |
| "learning_rate": 3.0838091114159895e-06, | |
| "loss": 0.3396, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.198380566801619, | |
| "grad_norm": 0.21060728899782163, | |
| "learning_rate": 3.0537059008444524e-06, | |
| "loss": 0.3748, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.202429149797571, | |
| "grad_norm": 0.20773038524439189, | |
| "learning_rate": 3.0237381923889563e-06, | |
| "loss": 0.3969, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.206477732793522, | |
| "grad_norm": 0.2070162724023983, | |
| "learning_rate": 2.9939062256704997e-06, | |
| "loss": 0.3677, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.21038464893671294, | |
| "learning_rate": 2.964210239224696e-06, | |
| "loss": 0.3948, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.2145748987854255, | |
| "grad_norm": 0.2226270769609997, | |
| "learning_rate": 2.934650470499869e-06, | |
| "loss": 0.371, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.218623481781377, | |
| "grad_norm": 0.21544560821031292, | |
| "learning_rate": 2.905227155855128e-06, | |
| "loss": 0.3707, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.222672064777328, | |
| "grad_norm": 0.21833524877104019, | |
| "learning_rate": 2.875940530558519e-06, | |
| "loss": 0.3754, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.22672064777328, | |
| "grad_norm": 0.20897341739758415, | |
| "learning_rate": 2.8467908287851154e-06, | |
| "loss": 0.3747, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.21231650693252707, | |
| "learning_rate": 2.8177782836151603e-06, | |
| "loss": 0.3665, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.234817813765182, | |
| "grad_norm": 0.2194703613821168, | |
| "learning_rate": 2.7889031270321874e-06, | |
| "loss": 0.3704, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.238866396761134, | |
| "grad_norm": 0.2163595988971289, | |
| "learning_rate": 2.760165589921184e-06, | |
| "loss": 0.3787, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.242914979757085, | |
| "grad_norm": 0.21676342202724155, | |
| "learning_rate": 2.7315659020667305e-06, | |
| "loss": 0.363, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.246963562753036, | |
| "grad_norm": 0.21397392223492234, | |
| "learning_rate": 2.703104292151164e-06, | |
| "loss": 0.3633, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.251012145748988, | |
| "grad_norm": 0.21017945120449524, | |
| "learning_rate": 2.6747809877527607e-06, | |
| "loss": 0.3789, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.255060728744939, | |
| "grad_norm": 0.20095008320046884, | |
| "learning_rate": 2.6465962153439082e-06, | |
| "loss": 0.3633, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.2591093117408905, | |
| "grad_norm": 0.20686147700924407, | |
| "learning_rate": 2.618550200289303e-06, | |
| "loss": 0.3936, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.2080259207916012, | |
| "learning_rate": 2.59064316684412e-06, | |
| "loss": 0.3706, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.267206477732794, | |
| "grad_norm": 0.21194225859584725, | |
| "learning_rate": 2.562875338152264e-06, | |
| "loss": 0.3666, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.271255060728745, | |
| "grad_norm": 0.21467405736407916, | |
| "learning_rate": 2.5352469362445552e-06, | |
| "loss": 0.3662, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.275303643724697, | |
| "grad_norm": 0.21318168617577496, | |
| "learning_rate": 2.5077581820369523e-06, | |
| "loss": 0.3646, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.279352226720648, | |
| "grad_norm": 0.21621518778020482, | |
| "learning_rate": 2.480409295328803e-06, | |
| "loss": 0.3773, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.283400809716599, | |
| "grad_norm": 0.20255614723264437, | |
| "learning_rate": 2.45320049480108e-06, | |
| "loss": 0.3663, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.287449392712551, | |
| "grad_norm": 0.22635289395704802, | |
| "learning_rate": 2.4261319980146293e-06, | |
| "loss": 0.3579, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.291497975708502, | |
| "grad_norm": 0.21225259028067509, | |
| "learning_rate": 2.399204021408421e-06, | |
| "loss": 0.3829, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.295546558704453, | |
| "grad_norm": 0.21231181096967974, | |
| "learning_rate": 2.3724167802978395e-06, | |
| "loss": 0.371, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.299595141700405, | |
| "grad_norm": 0.21455103909045126, | |
| "learning_rate": 2.3457704888729625e-06, | |
| "loss": 0.3899, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.303643724696356, | |
| "grad_norm": 0.21737225682098485, | |
| "learning_rate": 2.3192653601968162e-06, | |
| "loss": 0.3496, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.2167080969031002, | |
| "learning_rate": 2.2929016062037057e-06, | |
| "loss": 0.3686, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.3117408906882595, | |
| "grad_norm": 0.217898325487751, | |
| "learning_rate": 2.2666794376975055e-06, | |
| "loss": 0.3775, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.2123745389032504, | |
| "learning_rate": 2.24059906434998e-06, | |
| "loss": 0.3597, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.319838056680162, | |
| "grad_norm": 0.21820128268905606, | |
| "learning_rate": 2.214660694699091e-06, | |
| "loss": 0.3551, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.323886639676114, | |
| "grad_norm": 0.19913805922345426, | |
| "learning_rate": 2.188864536147355e-06, | |
| "loss": 0.4004, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.327935222672065, | |
| "grad_norm": 0.20022906625943135, | |
| "learning_rate": 2.1632107949601757e-06, | |
| "loss": 0.3829, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.331983805668016, | |
| "grad_norm": 0.2077382513167395, | |
| "learning_rate": 2.137699676264171e-06, | |
| "loss": 0.3692, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.336032388663968, | |
| "grad_norm": 0.2219348912058595, | |
| "learning_rate": 2.1123313840455738e-06, | |
| "loss": 0.3509, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.340080971659919, | |
| "grad_norm": 0.20441464563943523, | |
| "learning_rate": 2.08710612114857e-06, | |
| "loss": 0.3693, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.34412955465587, | |
| "grad_norm": 0.22028971525221952, | |
| "learning_rate": 2.0620240892736953e-06, | |
| "loss": 0.3717, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.348178137651822, | |
| "grad_norm": 0.21083289486387968, | |
| "learning_rate": 2.0370854889762e-06, | |
| "loss": 0.3903, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.352226720647773, | |
| "grad_norm": 0.20647125861179438, | |
| "learning_rate": 2.0122905196644726e-06, | |
| "loss": 0.3778, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3562753036437245, | |
| "grad_norm": 0.21368317567517972, | |
| "learning_rate": 1.987639379598427e-06, | |
| "loss": 0.3605, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.3603238866396765, | |
| "grad_norm": 0.21111150478944304, | |
| "learning_rate": 1.963132265887919e-06, | |
| "loss": 0.374, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.364372469635628, | |
| "grad_norm": 0.2303524438818984, | |
| "learning_rate": 1.9387693744911784e-06, | |
| "loss": 0.3826, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.21307623386091945, | |
| "learning_rate": 1.9145509002132368e-06, | |
| "loss": 0.36, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.372469635627531, | |
| "grad_norm": 0.19449705747164803, | |
| "learning_rate": 1.8904770367043723e-06, | |
| "loss": 0.3799, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.376518218623482, | |
| "grad_norm": 0.20027662571042937, | |
| "learning_rate": 1.8665479764585504e-06, | |
| "loss": 0.3843, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.380566801619433, | |
| "grad_norm": 0.2194877201430139, | |
| "learning_rate": 1.8427639108118977e-06, | |
| "loss": 0.3587, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.2087189784420618, | |
| "learning_rate": 1.8191250299411755e-06, | |
| "loss": 0.3594, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.388663967611336, | |
| "grad_norm": 0.22148421963355813, | |
| "learning_rate": 1.7956315228622446e-06, | |
| "loss": 0.3863, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.392712550607287, | |
| "grad_norm": 0.21502420777305256, | |
| "learning_rate": 1.7722835774285553e-06, | |
| "loss": 0.3746, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.396761133603239, | |
| "grad_norm": 0.2059999554987837, | |
| "learning_rate": 1.7490813803296624e-06, | |
| "loss": 0.3602, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.40080971659919, | |
| "grad_norm": 0.20713320037140015, | |
| "learning_rate": 1.726025117089718e-06, | |
| "loss": 0.3811, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.4048582995951415, | |
| "grad_norm": 0.20226306939732056, | |
| "learning_rate": 1.70311497206598e-06, | |
| "loss": 0.3822, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.4089068825910935, | |
| "grad_norm": 0.20289833726191872, | |
| "learning_rate": 1.6803511284473638e-06, | |
| "loss": 0.3795, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.412955465587045, | |
| "grad_norm": 0.20674564812922736, | |
| "learning_rate": 1.6577337682529538e-06, | |
| "loss": 0.3776, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.417004048582996, | |
| "grad_norm": 0.20837457334968215, | |
| "learning_rate": 1.6352630723305617e-06, | |
| "loss": 0.3732, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.2039667926762729, | |
| "learning_rate": 1.612939220355272e-06, | |
| "loss": 0.3802, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.425101214574899, | |
| "grad_norm": 0.20802751877243994, | |
| "learning_rate": 1.5907623908280112e-06, | |
| "loss": 0.3717, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.42914979757085, | |
| "grad_norm": 0.20808212362702605, | |
| "learning_rate": 1.568732761074121e-06, | |
| "loss": 0.3587, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.433198380566802, | |
| "grad_norm": 0.21325074476643338, | |
| "learning_rate": 1.5468505072419237e-06, | |
| "loss": 0.3518, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.437246963562753, | |
| "grad_norm": 0.20364093777155476, | |
| "learning_rate": 1.5251158043013448e-06, | |
| "loss": 0.3876, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.441295546558704, | |
| "grad_norm": 0.20688374945857943, | |
| "learning_rate": 1.5035288260424885e-06, | |
| "loss": 0.3677, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.445344129554655, | |
| "grad_norm": 0.20565906597165348, | |
| "learning_rate": 1.4820897450742623e-06, | |
| "loss": 0.3834, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.449392712550607, | |
| "grad_norm": 0.20151896186987597, | |
| "learning_rate": 1.4607987328229767e-06, | |
| "loss": 0.3798, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.4534412955465585, | |
| "grad_norm": 0.21633505265905997, | |
| "learning_rate": 1.4396559595310056e-06, | |
| "loss": 0.3721, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.4574898785425106, | |
| "grad_norm": 0.20434431276946882, | |
| "learning_rate": 1.4186615942553973e-06, | |
| "loss": 0.3598, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.20086293893792875, | |
| "learning_rate": 1.3978158048665424e-06, | |
| "loss": 0.3832, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.465587044534413, | |
| "grad_norm": 0.20098982307091853, | |
| "learning_rate": 1.3771187580468115e-06, | |
| "loss": 0.379, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.469635627530364, | |
| "grad_norm": 0.20660864055684847, | |
| "learning_rate": 1.3565706192892392e-06, | |
| "loss": 0.367, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.2040797618146867, | |
| "learning_rate": 1.3361715528961993e-06, | |
| "loss": 0.3745, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.477732793522267, | |
| "grad_norm": 0.218360077625037, | |
| "learning_rate": 1.3159217219780817e-06, | |
| "loss": 0.3955, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.481781376518219, | |
| "grad_norm": 0.21373117692718963, | |
| "learning_rate": 1.2958212884519949e-06, | |
| "loss": 0.3868, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.48582995951417, | |
| "grad_norm": 0.20539077038990125, | |
| "learning_rate": 1.2758704130404675e-06, | |
| "loss": 0.3476, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.489878542510121, | |
| "grad_norm": 0.19639744208741114, | |
| "learning_rate": 1.2560692552701738e-06, | |
| "loss": 0.3801, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.493927125506072, | |
| "grad_norm": 0.22283825211059533, | |
| "learning_rate": 1.2364179734706382e-06, | |
| "loss": 0.3433, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.497975708502024, | |
| "grad_norm": 0.21058631537857014, | |
| "learning_rate": 1.2169167247729852e-06, | |
| "loss": 0.3747, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.502024291497976, | |
| "grad_norm": 0.19786779162369167, | |
| "learning_rate": 1.1975656651086864e-06, | |
| "loss": 0.3813, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.506072874493928, | |
| "grad_norm": 0.21013017503156464, | |
| "learning_rate": 1.1783649492082926e-06, | |
| "loss": 0.3573, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.510121457489879, | |
| "grad_norm": 0.19812123825206696, | |
| "learning_rate": 1.1593147306002183e-06, | |
| "loss": 0.3881, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.51417004048583, | |
| "grad_norm": 0.21134764763481323, | |
| "learning_rate": 1.140415161609505e-06, | |
| "loss": 0.3841, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.518218623481781, | |
| "grad_norm": 0.20920213902616955, | |
| "learning_rate": 1.121666393356604e-06, | |
| "loss": 0.3735, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.522267206477733, | |
| "grad_norm": 0.2021970360156431, | |
| "learning_rate": 1.1030685757561632e-06, | |
| "loss": 0.3956, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.19776763154707772, | |
| "learning_rate": 1.0846218575158373e-06, | |
| "loss": 0.3738, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.530364372469636, | |
| "grad_norm": 0.2030002907346198, | |
| "learning_rate": 1.066326386135097e-06, | |
| "loss": 0.3659, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.534412955465587, | |
| "grad_norm": 0.20661901154355267, | |
| "learning_rate": 1.0481823079040398e-06, | |
| "loss": 0.3735, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.19995442183084533, | |
| "learning_rate": 1.0301897679022344e-06, | |
| "loss": 0.3819, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.5425101214574894, | |
| "grad_norm": 0.20075835616248647, | |
| "learning_rate": 1.0123489099975491e-06, | |
| "loss": 0.3834, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.5465587044534415, | |
| "grad_norm": 0.20577071386109538, | |
| "learning_rate": 9.9465987684501e-07, | |
| "loss": 0.3688, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.550607287449393, | |
| "grad_norm": 0.21011970841276625, | |
| "learning_rate": 9.771228098856534e-07, | |
| "loss": 0.3825, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.554655870445345, | |
| "grad_norm": 0.2145048882699763, | |
| "learning_rate": 9.597378493454013e-07, | |
| "loss": 0.3581, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.558704453441296, | |
| "grad_norm": 0.19782943327082878, | |
| "learning_rate": 9.425051342339353e-07, | |
| "loss": 0.3813, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.562753036437247, | |
| "grad_norm": 0.1973160105536076, | |
| "learning_rate": 9.254248023435864e-07, | |
| "loss": 0.3746, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.566801619433198, | |
| "grad_norm": 0.2057327108690158, | |
| "learning_rate": 9.084969902482288e-07, | |
| "loss": 0.3719, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.57085020242915, | |
| "grad_norm": 0.20236696023615902, | |
| "learning_rate": 8.917218333021993e-07, | |
| "loss": 0.3743, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.574898785425101, | |
| "grad_norm": 0.21154610277970848, | |
| "learning_rate": 8.750994656392109e-07, | |
| "loss": 0.3556, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.21005333393901887, | |
| "learning_rate": 8.58630020171265e-07, | |
| "loss": 0.3647, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.582995951417004, | |
| "grad_norm": 0.2256078382507988, | |
| "learning_rate": 8.423136285876099e-07, | |
| "loss": 0.3631, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.587044534412955, | |
| "grad_norm": 0.2001394205506842, | |
| "learning_rate": 8.26150421353682e-07, | |
| "loss": 0.3771, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.5910931174089065, | |
| "grad_norm": 0.21474940905856316, | |
| "learning_rate": 8.101405277100549e-07, | |
| "loss": 0.356, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.5951417004048585, | |
| "grad_norm": 0.20878722063269964, | |
| "learning_rate": 7.942840756714077e-07, | |
| "loss": 0.3666, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.59919028340081, | |
| "grad_norm": 0.19900919831352587, | |
| "learning_rate": 7.785811920255093e-07, | |
| "loss": 0.3757, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.603238866396762, | |
| "grad_norm": 0.23234480562616305, | |
| "learning_rate": 7.630320023322024e-07, | |
| "loss": 0.3493, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.607287449392713, | |
| "grad_norm": 0.20083209213069644, | |
| "learning_rate": 7.476366309223903e-07, | |
| "loss": 0.3683, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.611336032388664, | |
| "grad_norm": 0.20937724574356198, | |
| "learning_rate": 7.323952008970537e-07, | |
| "loss": 0.388, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.20708182933620914, | |
| "learning_rate": 7.173078341262618e-07, | |
| "loss": 0.3641, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.619433198380567, | |
| "grad_norm": 0.20859289304299053, | |
| "learning_rate": 7.023746512481989e-07, | |
| "loss": 0.3661, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.623481781376518, | |
| "grad_norm": 0.20275442149147027, | |
| "learning_rate": 6.875957716681902e-07, | |
| "loss": 0.3848, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.62753036437247, | |
| "grad_norm": 0.2044594664567228, | |
| "learning_rate": 6.729713135577687e-07, | |
| "loss": 0.3597, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.20343416835715308, | |
| "learning_rate": 6.585013938537143e-07, | |
| "loss": 0.3677, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.635627530364372, | |
| "grad_norm": 0.2106602455949804, | |
| "learning_rate": 6.441861282571115e-07, | |
| "loss": 0.3735, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.6396761133603235, | |
| "grad_norm": 0.2017005937236046, | |
| "learning_rate": 6.300256312324471e-07, | |
| "loss": 0.3769, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.6437246963562755, | |
| "grad_norm": 0.20891313529758837, | |
| "learning_rate": 6.160200160066776e-07, | |
| "loss": 0.3641, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.647773279352227, | |
| "grad_norm": 0.1968712671172605, | |
| "learning_rate": 6.021693945683327e-07, | |
| "loss": 0.3812, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.651821862348179, | |
| "grad_norm": 0.20650393117723095, | |
| "learning_rate": 5.884738776666044e-07, | |
| "loss": 0.3603, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.65587044534413, | |
| "grad_norm": 0.19590842178305257, | |
| "learning_rate": 5.749335748104878e-07, | |
| "loss": 0.3744, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.659919028340081, | |
| "grad_norm": 0.20573775682828943, | |
| "learning_rate": 5.615485942678778e-07, | |
| "loss": 0.3904, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.663967611336032, | |
| "grad_norm": 0.19593710382546925, | |
| "learning_rate": 5.483190430647289e-07, | |
| "loss": 0.382, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.668016194331984, | |
| "grad_norm": 0.19728642708702357, | |
| "learning_rate": 5.352450269841747e-07, | |
| "loss": 0.3601, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.672064777327935, | |
| "grad_norm": 0.2021876337439171, | |
| "learning_rate": 5.22326650565701e-07, | |
| "loss": 0.383, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.676113360323887, | |
| "grad_norm": 0.2051097348323144, | |
| "learning_rate": 5.095640171043048e-07, | |
| "loss": 0.3807, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.680161943319838, | |
| "grad_norm": 0.5286804911955623, | |
| "learning_rate": 4.969572286496594e-07, | |
| "loss": 0.3566, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.19889920805010683, | |
| "learning_rate": 4.845063860053079e-07, | |
| "loss": 0.3811, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.6882591093117405, | |
| "grad_norm": 0.19993468332913772, | |
| "learning_rate": 4.7221158872785997e-07, | |
| "loss": 0.3681, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 0.20725550673961504, | |
| "learning_rate": 4.600729351261857e-07, | |
| "loss": 0.3902, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.696356275303644, | |
| "grad_norm": 0.2127451164461561, | |
| "learning_rate": 4.4809052226063355e-07, | |
| "loss": 0.3521, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.700404858299595, | |
| "grad_norm": 0.2093813946322679, | |
| "learning_rate": 4.3626444594226046e-07, | |
| "loss": 0.3621, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.704453441295547, | |
| "grad_norm": 0.1969991353906767, | |
| "learning_rate": 4.245948007320588e-07, | |
| "loss": 0.3744, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.708502024291498, | |
| "grad_norm": 0.20778268465283944, | |
| "learning_rate": 4.130816799401971e-07, | |
| "loss": 0.3789, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.712550607287449, | |
| "grad_norm": 0.20367601828313875, | |
| "learning_rate": 4.017251756252827e-07, | |
| "loss": 0.3654, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.716599190283401, | |
| "grad_norm": 0.19769349667780697, | |
| "learning_rate": 3.9052537859362473e-07, | |
| "loss": 0.3747, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.720647773279352, | |
| "grad_norm": 0.20045752844140344, | |
| "learning_rate": 3.7948237839849687e-07, | |
| "loss": 0.3603, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.724696356275303, | |
| "grad_norm": 0.20080765476071388, | |
| "learning_rate": 3.6859626333943797e-07, | |
| "loss": 0.3588, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.728744939271255, | |
| "grad_norm": 0.20319166959330368, | |
| "learning_rate": 3.578671204615347e-07, | |
| "loss": 0.3656, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.732793522267206, | |
| "grad_norm": 0.1971213632726664, | |
| "learning_rate": 3.472950355547267e-07, | |
| "loss": 0.3603, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.20505899135774802, | |
| "learning_rate": 3.3688009315312244e-07, | |
| "loss": 0.3634, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.7408906882591095, | |
| "grad_norm": 0.2047858384602914, | |
| "learning_rate": 3.266223765343224e-07, | |
| "loss": 0.3915, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.744939271255061, | |
| "grad_norm": 0.19913567850945607, | |
| "learning_rate": 3.16521967718757e-07, | |
| "loss": 0.3667, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.748987854251012, | |
| "grad_norm": 0.2487175025967134, | |
| "learning_rate": 3.0657894746902726e-07, | |
| "loss": 0.3654, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.753036437246964, | |
| "grad_norm": 0.20211512711148824, | |
| "learning_rate": 2.9679339528925654e-07, | |
| "loss": 0.3798, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.757085020242915, | |
| "grad_norm": 0.19908250359966267, | |
| "learning_rate": 2.871653894244597e-07, | |
| "loss": 0.3969, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.761133603238866, | |
| "grad_norm": 0.2070298175677068, | |
| "learning_rate": 2.7769500685991266e-07, | |
| "loss": 0.3752, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.765182186234818, | |
| "grad_norm": 0.2016274659463227, | |
| "learning_rate": 2.683823233205485e-07, | |
| "loss": 0.3707, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.2209065473950627, | |
| "learning_rate": 2.592274132703287e-07, | |
| "loss": 0.371, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.77327935222672, | |
| "grad_norm": 0.19551035876560155, | |
| "learning_rate": 2.502303499116754e-07, | |
| "loss": 0.3626, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.777327935222672, | |
| "grad_norm": 0.2038038460302735, | |
| "learning_rate": 2.413912051848666e-07, | |
| "loss": 0.3627, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.781376518218623, | |
| "grad_norm": 0.2054717642221217, | |
| "learning_rate": 2.3271004976746613e-07, | |
| "loss": 0.3695, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.7854251012145745, | |
| "grad_norm": 0.1993028585164368, | |
| "learning_rate": 2.2418695307376393e-07, | |
| "loss": 0.3692, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.20805557911908407, | |
| "learning_rate": 2.1582198325420968e-07, | |
| "loss": 0.3529, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.793522267206478, | |
| "grad_norm": 0.20445341667414793, | |
| "learning_rate": 2.0761520719488447e-07, | |
| "loss": 0.3636, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.797570850202429, | |
| "grad_norm": 0.20507411749029636, | |
| "learning_rate": 1.9956669051694798e-07, | |
| "loss": 0.363, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.801619433198381, | |
| "grad_norm": 0.20824999116103637, | |
| "learning_rate": 1.916764975761254e-07, | |
| "loss": 0.3572, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.805668016194332, | |
| "grad_norm": 0.212633196853488, | |
| "learning_rate": 1.839446914621923e-07, | |
| "loss": 0.3747, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.809716599190283, | |
| "grad_norm": 0.21522170228789297, | |
| "learning_rate": 1.7637133399845962e-07, | |
| "loss": 0.3779, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.813765182186235, | |
| "grad_norm": 0.19801899204590845, | |
| "learning_rate": 1.6895648574129397e-07, | |
| "loss": 0.3771, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.817813765182186, | |
| "grad_norm": 0.2103192886029066, | |
| "learning_rate": 1.6170020597962465e-07, | |
| "loss": 0.3565, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.821862348178137, | |
| "grad_norm": 0.22466353624872093, | |
| "learning_rate": 1.5460255273447078e-07, | |
| "loss": 0.3699, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.825910931174089, | |
| "grad_norm": 0.20140696605698727, | |
| "learning_rate": 1.4766358275847715e-07, | |
| "loss": 0.3741, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.82995951417004, | |
| "grad_norm": 0.20818742509109409, | |
| "learning_rate": 1.4088335153546573e-07, | |
| "loss": 0.381, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.834008097165992, | |
| "grad_norm": 0.19876831775779943, | |
| "learning_rate": 1.3426191327998496e-07, | |
| "loss": 0.3663, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.838056680161944, | |
| "grad_norm": 0.19378679393334136, | |
| "learning_rate": 1.277993209368744e-07, | |
| "loss": 0.3791, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.2021511198794935, | |
| "learning_rate": 1.214956261808542e-07, | |
| "loss": 0.3859, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.20683217210184562, | |
| "learning_rate": 1.1535087941609846e-07, | |
| "loss": 0.3667, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.850202429149798, | |
| "grad_norm": 0.2050870619558301, | |
| "learning_rate": 1.0936512977583802e-07, | |
| "loss": 0.381, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.854251012145749, | |
| "grad_norm": 0.200237544366357, | |
| "learning_rate": 1.0353842512196732e-07, | |
| "loss": 0.3629, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.8582995951417, | |
| "grad_norm": 0.2042585915401584, | |
| "learning_rate": 9.787081204466253e-08, | |
| "loss": 0.3608, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.862348178137652, | |
| "grad_norm": 0.20859600135213022, | |
| "learning_rate": 9.236233586200627e-08, | |
| "loss": 0.3653, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.866396761133603, | |
| "grad_norm": 0.19975405201134153, | |
| "learning_rate": 8.70130406196279e-08, | |
| "loss": 0.3793, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.870445344129554, | |
| "grad_norm": 0.2128484577367486, | |
| "learning_rate": 8.182296909035047e-08, | |
| "loss": 0.3631, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.874493927125506, | |
| "grad_norm": 0.20621750314176976, | |
| "learning_rate": 7.679216277384439e-08, | |
| "loss": 0.3525, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.8785425101214575, | |
| "grad_norm": 0.1994427651885493, | |
| "learning_rate": 7.19206618963053e-08, | |
| "loss": 0.3854, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.882591093117409, | |
| "grad_norm": 0.1920741140454697, | |
| "learning_rate": 6.720850541012347e-08, | |
| "loss": 0.383, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.886639676113361, | |
| "grad_norm": 0.20284512876700123, | |
| "learning_rate": 6.265573099357714e-08, | |
| "loss": 0.3683, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.890688259109312, | |
| "grad_norm": 0.1996066990716179, | |
| "learning_rate": 5.826237505053067e-08, | |
| "loss": 0.3737, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.21127161007819645, | |
| "learning_rate": 5.4028472710137e-08, | |
| "loss": 0.3565, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.898785425101215, | |
| "grad_norm": 0.2025120132285612, | |
| "learning_rate": 4.9954057826571146e-08, | |
| "loss": 0.3589, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.902834008097166, | |
| "grad_norm": 0.2133128381884382, | |
| "learning_rate": 4.6039162978743776e-08, | |
| "loss": 0.3518, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.906882591093117, | |
| "grad_norm": 0.20623996631677874, | |
| "learning_rate": 4.228381947004812e-08, | |
| "loss": 0.3448, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.910931174089069, | |
| "grad_norm": 0.2183073792264033, | |
| "learning_rate": 3.868805732811343e-08, | |
| "loss": 0.3549, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.91497975708502, | |
| "grad_norm": 0.20045401168965837, | |
| "learning_rate": 3.525190530455636e-08, | |
| "loss": 0.3744, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.919028340080971, | |
| "grad_norm": 0.22012259688905417, | |
| "learning_rate": 3.197539087475887e-08, | |
| "loss": 0.3823, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.2047560415405844, | |
| "learning_rate": 2.8858540237641786e-08, | |
| "loss": 0.3359, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.9271255060728745, | |
| "grad_norm": 0.20288854480286928, | |
| "learning_rate": 2.5901378315458247e-08, | |
| "loss": 0.3645, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.931174089068826, | |
| "grad_norm": 0.20172279198707205, | |
| "learning_rate": 2.3103928753598347e-08, | |
| "loss": 0.3694, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.935222672064778, | |
| "grad_norm": 0.20002629604205055, | |
| "learning_rate": 2.0466213920393718e-08, | |
| "loss": 0.3782, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.939271255060729, | |
| "grad_norm": 0.20688179864407355, | |
| "learning_rate": 1.7988254906942115e-08, | |
| "loss": 0.3574, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94331983805668, | |
| "grad_norm": 0.208324806719985, | |
| "learning_rate": 1.5670071526936447e-08, | |
| "loss": 0.369, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.19857330959681366, | |
| "learning_rate": 1.3511682316509344e-08, | |
| "loss": 0.381, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.951417004048583, | |
| "grad_norm": 0.21701545124565066, | |
| "learning_rate": 1.1513104534086605e-08, | |
| "loss": 0.345, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.955465587044534, | |
| "grad_norm": 0.21489155559924641, | |
| "learning_rate": 9.67435416023843e-09, | |
| "loss": 0.3613, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.959514170040486, | |
| "grad_norm": 0.19909751150480057, | |
| "learning_rate": 7.995445897566178e-09, | |
| "loss": 0.3802, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.963562753036437, | |
| "grad_norm": 0.2001673297323576, | |
| "learning_rate": 6.476393170578021e-09, | |
| "loss": 0.355, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.967611336032388, | |
| "grad_norm": 0.20916743338695562, | |
| "learning_rate": 5.11720812557348e-09, | |
| "loss": 0.3659, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.97165991902834, | |
| "grad_norm": 0.2238312563660954, | |
| "learning_rate": 3.917901630563492e-09, | |
| "loss": 0.3713, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.9757085020242915, | |
| "grad_norm": 0.2117126414684857, | |
| "learning_rate": 2.878483275170485e-09, | |
| "loss": 0.3613, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.979757085020243, | |
| "grad_norm": 0.21495166538914537, | |
| "learning_rate": 1.998961370555108e-09, | |
| "loss": 0.3577, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.983805668016195, | |
| "grad_norm": 0.21154998514815876, | |
| "learning_rate": 1.2793429493518361e-09, | |
| "loss": 0.3688, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.987854251012146, | |
| "grad_norm": 0.20637853131213268, | |
| "learning_rate": 7.196337656112384e-10, | |
| "loss": 0.3576, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.991902834008097, | |
| "grad_norm": 0.20055957311330078, | |
| "learning_rate": 3.198382947511292e-10, | |
| "loss": 0.3569, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.995951417004049, | |
| "grad_norm": 0.20851407912611256, | |
| "learning_rate": 7.995973352770137e-11, | |
| "loss": 0.3569, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.19508840445303455, | |
| "learning_rate": 0.0, | |
| "loss": 0.3712, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1235, | |
| "total_flos": 4.729547622979731e+18, | |
| "train_loss": 0.5010926865855692, | |
| "train_runtime": 29200.8941, | |
| "train_samples_per_second": 5.411, | |
| "train_steps_per_second": 0.042 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1235, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.729547622979731e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |