| { | |
| "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": 7.195065521935694, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 1.1466, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008097165991902834, | |
| "grad_norm": 7.000139506321542, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 1.1235, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012145748987854251, | |
| "grad_norm": 7.316577756989413, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 1.1452, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016194331983805668, | |
| "grad_norm": 7.023038649688676, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 1.1187, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020242914979757085, | |
| "grad_norm": 6.79506329764893, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 1.1219, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024291497975708502, | |
| "grad_norm": 6.769874803813742, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.1186, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02834008097165992, | |
| "grad_norm": 5.423736808726153, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 1.0746, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032388663967611336, | |
| "grad_norm": 5.217006432594199, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 1.0905, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03643724696356275, | |
| "grad_norm": 2.9674390013088288, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 1.0415, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04048582995951417, | |
| "grad_norm": 2.354289100730901, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.9588, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044534412955465584, | |
| "grad_norm": 2.41463036088528, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 0.9886, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048582995951417005, | |
| "grad_norm": 4.455895065386493, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.9912, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 4.635320275371106, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 0.9703, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05668016194331984, | |
| "grad_norm": 4.72982769014657, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 1.007, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06072874493927125, | |
| "grad_norm": 4.213946023691154, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 0.9295, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06477732793522267, | |
| "grad_norm": 3.1506513177406683, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 0.9224, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06882591093117409, | |
| "grad_norm": 2.9167377770548164, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 0.9326, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0728744939271255, | |
| "grad_norm": 2.9421886795750805, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.9052, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 2.3286760645188194, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 0.8864, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08097165991902834, | |
| "grad_norm": 1.9807292272115053, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.8993, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08502024291497975, | |
| "grad_norm": 1.563602927128434, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.8838, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08906882591093117, | |
| "grad_norm": 1.6806350225213729, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.8405, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0931174089068826, | |
| "grad_norm": 1.8258014646781615, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.7959, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09716599190283401, | |
| "grad_norm": 1.8425927214498854, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.8637, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10121457489878542, | |
| "grad_norm": 1.5395758491809977, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.8211, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 1.2870575090015717, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.7886, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10931174089068826, | |
| "grad_norm": 1.3647796083774733, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.782, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11336032388663968, | |
| "grad_norm": 1.0626974519408134, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.8121, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11740890688259109, | |
| "grad_norm": 1.0691793549397983, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.8046, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1214574898785425, | |
| "grad_norm": 1.0354212554396836, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.8084, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12550607287449392, | |
| "grad_norm": 0.891762617058733, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8277, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12955465587044535, | |
| "grad_norm": 0.7825391114986583, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.786, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13360323886639677, | |
| "grad_norm": 0.7630976123552771, | |
| "learning_rate": 1.0645161290322582e-05, | |
| "loss": 0.7952, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13765182186234817, | |
| "grad_norm": 0.9711849481037858, | |
| "learning_rate": 1.096774193548387e-05, | |
| "loss": 0.7817, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1417004048582996, | |
| "grad_norm": 0.8008824659105409, | |
| "learning_rate": 1.1290322580645164e-05, | |
| "loss": 0.8045, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.145748987854251, | |
| "grad_norm": 0.7331448800035693, | |
| "learning_rate": 1.1612903225806453e-05, | |
| "loss": 0.7591, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14979757085020243, | |
| "grad_norm": 0.7967220939129485, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.7939, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.7859184659071131, | |
| "learning_rate": 1.2258064516129034e-05, | |
| "loss": 0.7878, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 0.7171671080465287, | |
| "learning_rate": 1.2580645161290324e-05, | |
| "loss": 0.7506, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16194331983805668, | |
| "grad_norm": 0.8799669945101638, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.7671, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1659919028340081, | |
| "grad_norm": 0.7015516284202606, | |
| "learning_rate": 1.3225806451612903e-05, | |
| "loss": 0.756, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1700404858299595, | |
| "grad_norm": 0.7725509505583343, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.7788, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17408906882591094, | |
| "grad_norm": 0.8240876205104878, | |
| "learning_rate": 1.3870967741935486e-05, | |
| "loss": 0.7537, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17813765182186234, | |
| "grad_norm": 0.7557641384711952, | |
| "learning_rate": 1.4193548387096776e-05, | |
| "loss": 0.7823, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18218623481781376, | |
| "grad_norm": 0.8143074028658217, | |
| "learning_rate": 1.4516129032258066e-05, | |
| "loss": 0.7678, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1862348178137652, | |
| "grad_norm": 0.951344318319013, | |
| "learning_rate": 1.4838709677419357e-05, | |
| "loss": 0.7662, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1902834008097166, | |
| "grad_norm": 0.7117906281399105, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.7428, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19433198380566802, | |
| "grad_norm": 0.7917437654303368, | |
| "learning_rate": 1.5483870967741936e-05, | |
| "loss": 0.7508, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19838056680161945, | |
| "grad_norm": 0.7977013605064187, | |
| "learning_rate": 1.5806451612903226e-05, | |
| "loss": 0.7416, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20242914979757085, | |
| "grad_norm": 0.6669483166713627, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.7514, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20647773279352227, | |
| "grad_norm": 0.8551633961030181, | |
| "learning_rate": 1.6451612903225807e-05, | |
| "loss": 0.7397, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.6717460744890548, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.7174, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2145748987854251, | |
| "grad_norm": 0.8123765902158401, | |
| "learning_rate": 1.7096774193548388e-05, | |
| "loss": 0.7373, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21862348178137653, | |
| "grad_norm": 0.8260155357488893, | |
| "learning_rate": 1.741935483870968e-05, | |
| "loss": 0.7593, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22267206477732793, | |
| "grad_norm": 0.6779148401344507, | |
| "learning_rate": 1.774193548387097e-05, | |
| "loss": 0.7583, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22672064777327935, | |
| "grad_norm": 0.8388871179507195, | |
| "learning_rate": 1.806451612903226e-05, | |
| "loss": 0.7299, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.8435952396871678, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.748, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23481781376518218, | |
| "grad_norm": 0.8180628432423388, | |
| "learning_rate": 1.870967741935484e-05, | |
| "loss": 0.7659, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2388663967611336, | |
| "grad_norm": 0.7849059313017999, | |
| "learning_rate": 1.903225806451613e-05, | |
| "loss": 0.7239, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.242914979757085, | |
| "grad_norm": 0.7668097078239308, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.742, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24696356275303644, | |
| "grad_norm": 0.7219481191372216, | |
| "learning_rate": 1.967741935483871e-05, | |
| "loss": 0.7169, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.25101214574898784, | |
| "grad_norm": 0.6268845020965849, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7207, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2550607287449393, | |
| "grad_norm": 0.6691078822116006, | |
| "learning_rate": 2.0322580645161292e-05, | |
| "loss": 0.7151, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2591093117408907, | |
| "grad_norm": 0.7737443773536367, | |
| "learning_rate": 2.0645161290322582e-05, | |
| "loss": 0.7031, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 0.7585727919273387, | |
| "learning_rate": 2.0967741935483873e-05, | |
| "loss": 0.7503, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26720647773279355, | |
| "grad_norm": 0.8817269967195435, | |
| "learning_rate": 2.1290322580645163e-05, | |
| "loss": 0.762, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27125506072874495, | |
| "grad_norm": 0.8280115656412732, | |
| "learning_rate": 2.161290322580645e-05, | |
| "loss": 0.7561, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27530364372469635, | |
| "grad_norm": 0.9187458105863074, | |
| "learning_rate": 2.193548387096774e-05, | |
| "loss": 0.7461, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2793522267206478, | |
| "grad_norm": 0.6804343756457157, | |
| "learning_rate": 2.225806451612903e-05, | |
| "loss": 0.7086, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2834008097165992, | |
| "grad_norm": 0.8534591666063066, | |
| "learning_rate": 2.2580645161290328e-05, | |
| "loss": 0.7339, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2874493927125506, | |
| "grad_norm": 0.7935543289980963, | |
| "learning_rate": 2.2903225806451615e-05, | |
| "loss": 0.7107, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.291497975708502, | |
| "grad_norm": 0.8486696052988059, | |
| "learning_rate": 2.3225806451612906e-05, | |
| "loss": 0.7247, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.29554655870445345, | |
| "grad_norm": 0.8996265425810784, | |
| "learning_rate": 2.3548387096774196e-05, | |
| "loss": 0.7517, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29959514170040485, | |
| "grad_norm": 0.7694863609778891, | |
| "learning_rate": 2.3870967741935486e-05, | |
| "loss": 0.6973, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30364372469635625, | |
| "grad_norm": 1.1990838964160588, | |
| "learning_rate": 2.4193548387096777e-05, | |
| "loss": 0.7838, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.7916638368888684, | |
| "learning_rate": 2.4516129032258067e-05, | |
| "loss": 0.7325, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3117408906882591, | |
| "grad_norm": 1.1555213770531279, | |
| "learning_rate": 2.4838709677419358e-05, | |
| "loss": 0.7389, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 0.833288732038675, | |
| "learning_rate": 2.5161290322580648e-05, | |
| "loss": 0.7335, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31983805668016196, | |
| "grad_norm": 0.7804564280257076, | |
| "learning_rate": 2.548387096774194e-05, | |
| "loss": 0.7134, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32388663967611336, | |
| "grad_norm": 0.8911172840276674, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.6932, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32793522267206476, | |
| "grad_norm": 0.9478631432885956, | |
| "learning_rate": 2.6129032258064516e-05, | |
| "loss": 0.7434, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3319838056680162, | |
| "grad_norm": 0.8623932018451369, | |
| "learning_rate": 2.6451612903225806e-05, | |
| "loss": 0.7326, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3360323886639676, | |
| "grad_norm": 0.9200130594244436, | |
| "learning_rate": 2.6774193548387097e-05, | |
| "loss": 0.7258, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.340080971659919, | |
| "grad_norm": 1.0212745135739252, | |
| "learning_rate": 2.7096774193548387e-05, | |
| "loss": 0.72, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3441295546558704, | |
| "grad_norm": 1.0142399457987423, | |
| "learning_rate": 2.7419354838709678e-05, | |
| "loss": 0.7087, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3481781376518219, | |
| "grad_norm": 1.0462941740948752, | |
| "learning_rate": 2.774193548387097e-05, | |
| "loss": 0.7162, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3522267206477733, | |
| "grad_norm": 0.9048788427885954, | |
| "learning_rate": 2.8064516129032262e-05, | |
| "loss": 0.7467, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3562753036437247, | |
| "grad_norm": 0.9599524333629735, | |
| "learning_rate": 2.8387096774193552e-05, | |
| "loss": 0.7105, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3603238866396761, | |
| "grad_norm": 0.7428275755177126, | |
| "learning_rate": 2.8709677419354843e-05, | |
| "loss": 0.7312, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3643724696356275, | |
| "grad_norm": 0.9773495492337466, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.7154, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 0.7996466709869239, | |
| "learning_rate": 2.9354838709677423e-05, | |
| "loss": 0.7263, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3724696356275304, | |
| "grad_norm": 0.9670409860343432, | |
| "learning_rate": 2.9677419354838714e-05, | |
| "loss": 0.7592, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3765182186234818, | |
| "grad_norm": 0.8291022857313779, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.6681, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3805668016194332, | |
| "grad_norm": 0.9954476774483646, | |
| "learning_rate": 3.032258064516129e-05, | |
| "loss": 0.7108, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.9367683594842714, | |
| "learning_rate": 3.0645161290322585e-05, | |
| "loss": 0.7199, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 0.9180172891491688, | |
| "learning_rate": 3.096774193548387e-05, | |
| "loss": 0.7472, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39271255060728744, | |
| "grad_norm": 1.0485420089455582, | |
| "learning_rate": 3.1290322580645166e-05, | |
| "loss": 0.7111, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3967611336032389, | |
| "grad_norm": 1.0194267381814601, | |
| "learning_rate": 3.161290322580645e-05, | |
| "loss": 0.7244, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4008097165991903, | |
| "grad_norm": 0.7856847985592265, | |
| "learning_rate": 3.193548387096774e-05, | |
| "loss": 0.6739, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4048582995951417, | |
| "grad_norm": 0.9455950738112902, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.7105, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4089068825910931, | |
| "grad_norm": 0.8892692746245751, | |
| "learning_rate": 3.258064516129033e-05, | |
| "loss": 0.7387, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.41295546558704455, | |
| "grad_norm": 1.0636145075207581, | |
| "learning_rate": 3.2903225806451614e-05, | |
| "loss": 0.7412, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.41700404858299595, | |
| "grad_norm": 0.9129148201099412, | |
| "learning_rate": 3.322580645161291e-05, | |
| "loss": 0.724, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.7369522152186451, | |
| "learning_rate": 3.3548387096774195e-05, | |
| "loss": 0.717, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4251012145748988, | |
| "grad_norm": 0.8036140807339396, | |
| "learning_rate": 3.387096774193549e-05, | |
| "loss": 0.6858, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4291497975708502, | |
| "grad_norm": 0.8483468089426119, | |
| "learning_rate": 3.4193548387096776e-05, | |
| "loss": 0.7126, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4331983805668016, | |
| "grad_norm": 1.0136528311850648, | |
| "learning_rate": 3.451612903225807e-05, | |
| "loss": 0.681, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.43724696356275305, | |
| "grad_norm": 0.9347996871975378, | |
| "learning_rate": 3.483870967741936e-05, | |
| "loss": 0.7362, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.44129554655870445, | |
| "grad_norm": 1.1116449962553616, | |
| "learning_rate": 3.516129032258065e-05, | |
| "loss": 0.7243, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44534412955465585, | |
| "grad_norm": 0.9792783789136374, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.7125, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4493927125506073, | |
| "grad_norm": 0.8593162940762274, | |
| "learning_rate": 3.5806451612903225e-05, | |
| "loss": 0.7336, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4534412955465587, | |
| "grad_norm": 0.8525151499889524, | |
| "learning_rate": 3.612903225806452e-05, | |
| "loss": 0.6893, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4574898785425101, | |
| "grad_norm": 0.9147758411233519, | |
| "learning_rate": 3.6451612903225805e-05, | |
| "loss": 0.6886, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.0337262411625825, | |
| "learning_rate": 3.67741935483871e-05, | |
| "loss": 0.7387, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46558704453441296, | |
| "grad_norm": 0.7825683004821407, | |
| "learning_rate": 3.7096774193548386e-05, | |
| "loss": 0.7074, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.46963562753036436, | |
| "grad_norm": 0.8720416033645897, | |
| "learning_rate": 3.741935483870968e-05, | |
| "loss": 0.7022, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 0.7403843052813942, | |
| "learning_rate": 3.7741935483870974e-05, | |
| "loss": 0.6571, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4777327935222672, | |
| "grad_norm": 1.0022752659634524, | |
| "learning_rate": 3.806451612903226e-05, | |
| "loss": 0.7258, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4817813765182186, | |
| "grad_norm": 0.816159884795137, | |
| "learning_rate": 3.8387096774193555e-05, | |
| "loss": 0.7124, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48582995951417, | |
| "grad_norm": 0.9582701043387787, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.7383, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4898785425101215, | |
| "grad_norm": 0.9114139035895283, | |
| "learning_rate": 3.9032258064516135e-05, | |
| "loss": 0.7472, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4939271255060729, | |
| "grad_norm": 0.811443672020182, | |
| "learning_rate": 3.935483870967742e-05, | |
| "loss": 0.7159, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4979757085020243, | |
| "grad_norm": 0.8443179798916379, | |
| "learning_rate": 3.9677419354838716e-05, | |
| "loss": 0.7125, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5020242914979757, | |
| "grad_norm": 0.8088367307693484, | |
| "learning_rate": 4e-05, | |
| "loss": 0.702, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5060728744939271, | |
| "grad_norm": 0.9733325006696112, | |
| "learning_rate": 3.9999920040266474e-05, | |
| "loss": 0.7133, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5101214574898786, | |
| "grad_norm": 0.7871623098944788, | |
| "learning_rate": 3.999968016170525e-05, | |
| "loss": 0.7548, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5141700404858299, | |
| "grad_norm": 0.7199048315629579, | |
| "learning_rate": 3.9999280366234394e-05, | |
| "loss": 0.7287, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5182186234817814, | |
| "grad_norm": 0.7712587472492017, | |
| "learning_rate": 3.999872065705065e-05, | |
| "loss": 0.6783, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5222672064777328, | |
| "grad_norm": 0.9200775771971991, | |
| "learning_rate": 3.999800103862945e-05, | |
| "loss": 0.7221, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.8243570000250011, | |
| "learning_rate": 3.9997121516724834e-05, | |
| "loss": 0.7257, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5303643724696356, | |
| "grad_norm": 0.7663302968604523, | |
| "learning_rate": 3.999608209836944e-05, | |
| "loss": 0.6943, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5344129554655871, | |
| "grad_norm": 0.8514394594121165, | |
| "learning_rate": 3.999488279187443e-05, | |
| "loss": 0.6804, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.7222425248639243, | |
| "learning_rate": 3.9993523606829425e-05, | |
| "loss": 0.6867, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5425101214574899, | |
| "grad_norm": 0.7849533597758447, | |
| "learning_rate": 3.9992004554102435e-05, | |
| "loss": 0.7136, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5465587044534413, | |
| "grad_norm": 0.7745689483179966, | |
| "learning_rate": 3.999032564583976e-05, | |
| "loss": 0.7337, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5506072874493927, | |
| "grad_norm": 0.7387880790312601, | |
| "learning_rate": 3.998848689546591e-05, | |
| "loss": 0.7436, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5546558704453441, | |
| "grad_norm": 0.8543827785892651, | |
| "learning_rate": 3.9986488317683496e-05, | |
| "loss": 0.7121, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5587044534412956, | |
| "grad_norm": 0.7407824614854854, | |
| "learning_rate": 3.9984329928473067e-05, | |
| "loss": 0.7267, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.562753036437247, | |
| "grad_norm": 0.8867865004893184, | |
| "learning_rate": 3.998201174509306e-05, | |
| "loss": 0.762, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5668016194331984, | |
| "grad_norm": 0.8899722984556165, | |
| "learning_rate": 3.997953378607961e-05, | |
| "loss": 0.6978, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5708502024291497, | |
| "grad_norm": 0.77611421284881, | |
| "learning_rate": 3.9976896071246403e-05, | |
| "loss": 0.6897, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5748987854251012, | |
| "grad_norm": 0.8149419354790826, | |
| "learning_rate": 3.997409862168454e-05, | |
| "loss": 0.7367, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.8879367738879624, | |
| "learning_rate": 3.997114145976237e-05, | |
| "loss": 0.7404, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.582995951417004, | |
| "grad_norm": 0.7048502359807165, | |
| "learning_rate": 3.9968024609125245e-05, | |
| "loss": 0.7281, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5870445344129555, | |
| "grad_norm": 0.6653555301985642, | |
| "learning_rate": 3.996474809469545e-05, | |
| "loss": 0.7394, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5910931174089069, | |
| "grad_norm": 0.8676737433971777, | |
| "learning_rate": 3.996131194267189e-05, | |
| "loss": 0.7344, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5951417004048583, | |
| "grad_norm": 0.8084963237990159, | |
| "learning_rate": 3.995771618052995e-05, | |
| "loss": 0.6966, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5991902834008097, | |
| "grad_norm": 0.6466641246804055, | |
| "learning_rate": 3.9953960837021256e-05, | |
| "loss": 0.6926, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6032388663967612, | |
| "grad_norm": 0.8040568695968681, | |
| "learning_rate": 3.995004594217343e-05, | |
| "loss": 0.6952, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6072874493927125, | |
| "grad_norm": 0.8298594213701527, | |
| "learning_rate": 3.994597152728987e-05, | |
| "loss": 0.7226, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.611336032388664, | |
| "grad_norm": 0.7642588574972398, | |
| "learning_rate": 3.994173762494947e-05, | |
| "loss": 0.6774, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.6222817574278533, | |
| "learning_rate": 3.9937344269006424e-05, | |
| "loss": 0.7165, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6194331983805668, | |
| "grad_norm": 0.6666283540526751, | |
| "learning_rate": 3.993279149458988e-05, | |
| "loss": 0.6805, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6234817813765182, | |
| "grad_norm": 0.6890072763187951, | |
| "learning_rate": 3.9928079338103695e-05, | |
| "loss": 0.6654, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6275303643724697, | |
| "grad_norm": 0.6311796211739726, | |
| "learning_rate": 3.992320783722616e-05, | |
| "loss": 0.7198, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.8171303352549184, | |
| "learning_rate": 3.991817703090965e-05, | |
| "loss": 0.7393, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6356275303643725, | |
| "grad_norm": 0.7585359461873679, | |
| "learning_rate": 3.991298695938038e-05, | |
| "loss": 0.6721, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6396761133603239, | |
| "grad_norm": 0.8066839924638346, | |
| "learning_rate": 3.9907637664138e-05, | |
| "loss": 0.6958, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6437246963562753, | |
| "grad_norm": 0.9329070244527191, | |
| "learning_rate": 3.990212918795534e-05, | |
| "loss": 0.6951, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6477732793522267, | |
| "grad_norm": 0.616920243275045, | |
| "learning_rate": 3.9896461574878036e-05, | |
| "loss": 0.6734, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6518218623481782, | |
| "grad_norm": 0.7660764423913196, | |
| "learning_rate": 3.989063487022416e-05, | |
| "loss": 0.679, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6558704453441295, | |
| "grad_norm": 0.6136598083725247, | |
| "learning_rate": 3.9884649120583903e-05, | |
| "loss": 0.6412, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.659919028340081, | |
| "grad_norm": 0.7153404152851133, | |
| "learning_rate": 3.987850437381915e-05, | |
| "loss": 0.6875, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6639676113360324, | |
| "grad_norm": 0.7673995962175238, | |
| "learning_rate": 3.987220067906313e-05, | |
| "loss": 0.7166, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6680161943319838, | |
| "grad_norm": 0.7041665069547666, | |
| "learning_rate": 3.986573808672002e-05, | |
| "loss": 0.6835, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6720647773279352, | |
| "grad_norm": 0.7622420338781171, | |
| "learning_rate": 3.9859116648464534e-05, | |
| "loss": 0.6975, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6761133603238867, | |
| "grad_norm": 0.661092756731041, | |
| "learning_rate": 3.985233641724152e-05, | |
| "loss": 0.7304, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.680161943319838, | |
| "grad_norm": 0.6736604526116492, | |
| "learning_rate": 3.984539744726554e-05, | |
| "loss": 0.6873, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.676976162432414, | |
| "learning_rate": 3.983829979402038e-05, | |
| "loss": 0.7105, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6882591093117408, | |
| "grad_norm": 0.6173672439344426, | |
| "learning_rate": 3.983104351425871e-05, | |
| "loss": 0.7266, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.677890180480186, | |
| "learning_rate": 3.982362866600154e-05, | |
| "loss": 0.6812, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6963562753036437, | |
| "grad_norm": 0.7050557226993639, | |
| "learning_rate": 3.9816055308537806e-05, | |
| "loss": 0.7141, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7004048582995951, | |
| "grad_norm": 0.6412702621559053, | |
| "learning_rate": 3.980832350242388e-05, | |
| "loss": 0.7106, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7044534412955465, | |
| "grad_norm": 0.7693747522782185, | |
| "learning_rate": 3.980043330948306e-05, | |
| "loss": 0.6888, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708502024291498, | |
| "grad_norm": 0.6630035594004723, | |
| "learning_rate": 3.9792384792805124e-05, | |
| "loss": 0.6936, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7125506072874493, | |
| "grad_norm": 0.6303327310359051, | |
| "learning_rate": 3.978417801674579e-05, | |
| "loss": 0.6822, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7165991902834008, | |
| "grad_norm": 0.6664395038560121, | |
| "learning_rate": 3.9775813046926244e-05, | |
| "loss": 0.6866, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7206477732793523, | |
| "grad_norm": 0.7662308584720872, | |
| "learning_rate": 3.976728995023254e-05, | |
| "loss": 0.7111, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7246963562753036, | |
| "grad_norm": 0.6185836642812562, | |
| "learning_rate": 3.975860879481514e-05, | |
| "loss": 0.6728, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.728744939271255, | |
| "grad_norm": 0.6917036829422804, | |
| "learning_rate": 3.974976965008833e-05, | |
| "loss": 0.727, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7327935222672065, | |
| "grad_norm": 0.7015253552566695, | |
| "learning_rate": 3.9740772586729674e-05, | |
| "loss": 0.6932, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.6533191935550603, | |
| "learning_rate": 3.973161767667946e-05, | |
| "loss": 0.6583, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7408906882591093, | |
| "grad_norm": 0.6569557675121401, | |
| "learning_rate": 3.972230499314009e-05, | |
| "loss": 0.6908, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7449392712550608, | |
| "grad_norm": 0.6267789070211041, | |
| "learning_rate": 3.9712834610575545e-05, | |
| "loss": 0.6938, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7489878542510121, | |
| "grad_norm": 0.7048766223296957, | |
| "learning_rate": 3.9703206604710746e-05, | |
| "loss": 0.6535, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7530364372469636, | |
| "grad_norm": 0.7295168910315042, | |
| "learning_rate": 3.969342105253098e-05, | |
| "loss": 0.7049, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.757085020242915, | |
| "grad_norm": 0.6546358171364175, | |
| "learning_rate": 3.9683478032281246e-05, | |
| "loss": 0.6748, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7611336032388664, | |
| "grad_norm": 0.5679257208097711, | |
| "learning_rate": 3.967337762346568e-05, | |
| "loss": 0.6978, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7651821862348178, | |
| "grad_norm": 0.5941350241998313, | |
| "learning_rate": 3.9663119906846885e-05, | |
| "loss": 0.6669, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.6207157805164277, | |
| "learning_rate": 3.965270496444528e-05, | |
| "loss": 0.6632, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7732793522267206, | |
| "grad_norm": 0.6487150085030405, | |
| "learning_rate": 3.9642132879538465e-05, | |
| "loss": 0.725, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.6417633054064433, | |
| "learning_rate": 3.9631403736660566e-05, | |
| "loss": 0.6929, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7813765182186235, | |
| "grad_norm": 0.7646631100277187, | |
| "learning_rate": 3.962051762160151e-05, | |
| "loss": 0.6831, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7854251012145749, | |
| "grad_norm": 0.7869180942402786, | |
| "learning_rate": 3.960947462140638e-05, | |
| "loss": 0.641, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.6239219444802597, | |
| "learning_rate": 3.959827482437472e-05, | |
| "loss": 0.7355, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7935222672064778, | |
| "grad_norm": 0.8006684858549897, | |
| "learning_rate": 3.958691832005981e-05, | |
| "loss": 0.6934, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7975708502024291, | |
| "grad_norm": 0.6316537995510485, | |
| "learning_rate": 3.9575405199267944e-05, | |
| "loss": 0.6931, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8016194331983806, | |
| "grad_norm": 0.6129641877941455, | |
| "learning_rate": 3.956373555405774e-05, | |
| "loss": 0.6713, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.805668016194332, | |
| "grad_norm": 0.5936996641828995, | |
| "learning_rate": 3.955190947773937e-05, | |
| "loss": 0.6831, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8097165991902834, | |
| "grad_norm": 0.5436887243347416, | |
| "learning_rate": 3.9539927064873824e-05, | |
| "loss": 0.6874, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8137651821862348, | |
| "grad_norm": 0.6705750635757758, | |
| "learning_rate": 3.952778841127214e-05, | |
| "loss": 0.716, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8178137651821862, | |
| "grad_norm": 0.6581842937860904, | |
| "learning_rate": 3.95154936139947e-05, | |
| "loss": 0.7272, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8218623481781376, | |
| "grad_norm": 0.7199462804232438, | |
| "learning_rate": 3.950304277135034e-05, | |
| "loss": 0.6984, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8259109311740891, | |
| "grad_norm": 0.6569912282014467, | |
| "learning_rate": 3.94904359828957e-05, | |
| "loss": 0.6997, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8299595141700404, | |
| "grad_norm": 0.6246182993927791, | |
| "learning_rate": 3.94776733494343e-05, | |
| "loss": 0.6597, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8340080971659919, | |
| "grad_norm": 0.5687665046546535, | |
| "learning_rate": 3.9464754973015826e-05, | |
| "loss": 0.658, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8380566801619433, | |
| "grad_norm": 0.6055414468913521, | |
| "learning_rate": 3.9451680956935275e-05, | |
| "loss": 0.6531, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.6010946238705605, | |
| "learning_rate": 3.943845140573213e-05, | |
| "loss": 0.6928, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.721046045698023, | |
| "learning_rate": 3.942506642518952e-05, | |
| "loss": 0.6963, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8502024291497976, | |
| "grad_norm": 0.6219629336136466, | |
| "learning_rate": 3.9411526122333396e-05, | |
| "loss": 0.6955, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.854251012145749, | |
| "grad_norm": 0.6781072652911881, | |
| "learning_rate": 3.939783060543167e-05, | |
| "loss": 0.6959, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8582995951417004, | |
| "grad_norm": 0.5858999817203563, | |
| "learning_rate": 3.938397998399333e-05, | |
| "loss": 0.6696, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8623481781376519, | |
| "grad_norm": 0.6097186952085879, | |
| "learning_rate": 3.936997436876756e-05, | |
| "loss": 0.6861, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8663967611336032, | |
| "grad_norm": 0.555836389540804, | |
| "learning_rate": 3.935581387174289e-05, | |
| "loss": 0.7286, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8704453441295547, | |
| "grad_norm": 0.5560571547743615, | |
| "learning_rate": 3.9341498606146295e-05, | |
| "loss": 0.6711, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8744939271255061, | |
| "grad_norm": 0.6076946765491926, | |
| "learning_rate": 3.932702868644224e-05, | |
| "loss": 0.668, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8785425101214575, | |
| "grad_norm": 0.5967598812275031, | |
| "learning_rate": 3.931240422833181e-05, | |
| "loss": 0.6803, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8825910931174089, | |
| "grad_norm": 0.6041391705776588, | |
| "learning_rate": 3.9297625348751806e-05, | |
| "loss": 0.6763, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8866396761133604, | |
| "grad_norm": 0.6312823834043627, | |
| "learning_rate": 3.9282692165873744e-05, | |
| "loss": 0.6899, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8906882591093117, | |
| "grad_norm": 0.6270556911781027, | |
| "learning_rate": 3.9267604799102945e-05, | |
| "loss": 0.6691, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.6213930360761881, | |
| "learning_rate": 3.9252363369077615e-05, | |
| "loss": 0.6619, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8987854251012146, | |
| "grad_norm": 0.6933879726053501, | |
| "learning_rate": 3.9236967997667796e-05, | |
| "loss": 0.672, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.902834008097166, | |
| "grad_norm": 0.5539651204658643, | |
| "learning_rate": 3.9221418807974496e-05, | |
| "loss": 0.7111, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9068825910931174, | |
| "grad_norm": 0.8560103441274269, | |
| "learning_rate": 3.92057159243286e-05, | |
| "loss": 0.6655, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9109311740890689, | |
| "grad_norm": 0.6152163465850327, | |
| "learning_rate": 3.9189859472289956e-05, | |
| "loss": 0.6724, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9149797570850202, | |
| "grad_norm": 0.6444505831444888, | |
| "learning_rate": 3.9173849578646324e-05, | |
| "loss": 0.6957, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9190283400809717, | |
| "grad_norm": 0.6872210708263651, | |
| "learning_rate": 3.9157686371412395e-05, | |
| "loss": 0.6689, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.6918696818560587, | |
| "learning_rate": 3.914136997982874e-05, | |
| "loss": 0.7162, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9271255060728745, | |
| "grad_norm": 0.5875629093038148, | |
| "learning_rate": 3.912490053436079e-05, | |
| "loss": 0.6515, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9311740890688259, | |
| "grad_norm": 0.6327521761647066, | |
| "learning_rate": 3.9108278166697805e-05, | |
| "loss": 0.6792, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9352226720647774, | |
| "grad_norm": 0.627556070502507, | |
| "learning_rate": 3.909150300975178e-05, | |
| "loss": 0.7144, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9392712550607287, | |
| "grad_norm": 0.8905910299403569, | |
| "learning_rate": 3.907457519765642e-05, | |
| "loss": 0.6842, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9433198380566802, | |
| "grad_norm": 0.6906472138576186, | |
| "learning_rate": 3.905749486576607e-05, | |
| "loss": 0.7163, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.6080150717570272, | |
| "learning_rate": 3.90402621506546e-05, | |
| "loss": 0.6402, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.951417004048583, | |
| "grad_norm": 0.5732035994972682, | |
| "learning_rate": 3.902287719011435e-05, | |
| "loss": 0.6556, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9554655870445344, | |
| "grad_norm": 0.7199262860966557, | |
| "learning_rate": 3.900534012315499e-05, | |
| "loss": 0.7067, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9595141700404858, | |
| "grad_norm": 0.6486563488260335, | |
| "learning_rate": 3.8987651090002454e-05, | |
| "loss": 0.7028, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9635627530364372, | |
| "grad_norm": 0.6914851280931159, | |
| "learning_rate": 3.896981023209777e-05, | |
| "loss": 0.6166, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9676113360323887, | |
| "grad_norm": 0.6622291478260064, | |
| "learning_rate": 3.8951817692095966e-05, | |
| "loss": 0.6879, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.97165991902834, | |
| "grad_norm": 0.612845641456122, | |
| "learning_rate": 3.893367361386491e-05, | |
| "loss": 0.7032, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9757085020242915, | |
| "grad_norm": 0.6832630996426141, | |
| "learning_rate": 3.891537814248417e-05, | |
| "loss": 0.6529, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.979757085020243, | |
| "grad_norm": 0.7094625570440616, | |
| "learning_rate": 3.889693142424384e-05, | |
| "loss": 0.683, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9838056680161943, | |
| "grad_norm": 0.5821649841832206, | |
| "learning_rate": 3.8878333606643405e-05, | |
| "loss": 0.6703, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9878542510121457, | |
| "grad_norm": 0.6214562927452656, | |
| "learning_rate": 3.885958483839049e-05, | |
| "loss": 0.676, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9919028340080972, | |
| "grad_norm": 0.6340673350243471, | |
| "learning_rate": 3.8840685269399786e-05, | |
| "loss": 0.6672, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9959514170040485, | |
| "grad_norm": 0.5568705865113611, | |
| "learning_rate": 3.882163505079171e-05, | |
| "loss": 0.6639, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7431532582658508, | |
| "learning_rate": 3.880243433489132e-05, | |
| "loss": 0.6705, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0040485829959513, | |
| "grad_norm": 0.6810625192467499, | |
| "learning_rate": 3.878308327522702e-05, | |
| "loss": 0.5476, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.008097165991903, | |
| "grad_norm": 0.6293162815924903, | |
| "learning_rate": 3.876358202652936e-05, | |
| "loss": 0.5458, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0121457489878543, | |
| "grad_norm": 0.7371515558330818, | |
| "learning_rate": 3.8743930744729835e-05, | |
| "loss": 0.5514, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0161943319838056, | |
| "grad_norm": 0.6922073803929327, | |
| "learning_rate": 3.8724129586959535e-05, | |
| "loss": 0.5712, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0202429149797572, | |
| "grad_norm": 0.6561578326980045, | |
| "learning_rate": 3.870417871154801e-05, | |
| "loss": 0.5307, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0242914979757085, | |
| "grad_norm": 0.7118390141144403, | |
| "learning_rate": 3.868407827802192e-05, | |
| "loss": 0.5146, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0283400809716599, | |
| "grad_norm": 0.5827211817431118, | |
| "learning_rate": 3.8663828447103804e-05, | |
| "loss": 0.5425, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0323886639676114, | |
| "grad_norm": 0.6365397359045377, | |
| "learning_rate": 3.8643429380710765e-05, | |
| "loss": 0.5403, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0364372469635628, | |
| "grad_norm": 0.678163636819451, | |
| "learning_rate": 3.862288124195319e-05, | |
| "loss": 0.5189, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.040485829959514, | |
| "grad_norm": 0.6385452025463348, | |
| "learning_rate": 3.8602184195133464e-05, | |
| "loss": 0.5752, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0445344129554657, | |
| "grad_norm": 0.6392227717430902, | |
| "learning_rate": 3.8581338405744604e-05, | |
| "loss": 0.5408, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.048582995951417, | |
| "grad_norm": 0.6646522636073441, | |
| "learning_rate": 3.8560344040468996e-05, | |
| "loss": 0.5578, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.5865573313066954, | |
| "learning_rate": 3.853920126717703e-05, | |
| "loss": 0.5301, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.05668016194332, | |
| "grad_norm": 0.6794361902917602, | |
| "learning_rate": 3.851791025492574e-05, | |
| "loss": 0.5487, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0607287449392713, | |
| "grad_norm": 0.5848241545278178, | |
| "learning_rate": 3.8496471173957515e-05, | |
| "loss": 0.5369, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0647773279352226, | |
| "grad_norm": 0.6229827743682884, | |
| "learning_rate": 3.847488419569866e-05, | |
| "loss": 0.5376, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0688259109311742, | |
| "grad_norm": 0.7000010730513091, | |
| "learning_rate": 3.8453149492758077e-05, | |
| "loss": 0.5498, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0728744939271255, | |
| "grad_norm": 0.6364670929241698, | |
| "learning_rate": 3.8431267238925886e-05, | |
| "loss": 0.519, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.5742053043719915, | |
| "learning_rate": 3.840923760917199e-05, | |
| "loss": 0.5078, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0809716599190284, | |
| "grad_norm": 0.7199885079491375, | |
| "learning_rate": 3.838706077964473e-05, | |
| "loss": 0.5352, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0850202429149798, | |
| "grad_norm": 0.5924339127134981, | |
| "learning_rate": 3.836473692766944e-05, | |
| "loss": 0.5135, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0890688259109311, | |
| "grad_norm": 0.6748407018585711, | |
| "learning_rate": 3.834226623174705e-05, | |
| "loss": 0.5497, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0931174089068827, | |
| "grad_norm": 0.5677622919594815, | |
| "learning_rate": 3.831964887155264e-05, | |
| "loss": 0.5159, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.097165991902834, | |
| "grad_norm": 0.5896839174698049, | |
| "learning_rate": 3.829688502793403e-05, | |
| "loss": 0.5193, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.1012145748987854, | |
| "grad_norm": 0.6689337428997033, | |
| "learning_rate": 3.827397488291029e-05, | |
| "loss": 0.5507, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.5886825399697059, | |
| "learning_rate": 3.8250918619670343e-05, | |
| "loss": 0.5153, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.1093117408906883, | |
| "grad_norm": 0.6219194528064476, | |
| "learning_rate": 3.822771642257145e-05, | |
| "loss": 0.5408, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1133603238866396, | |
| "grad_norm": 0.6262557303090941, | |
| "learning_rate": 3.820436847713776e-05, | |
| "loss": 0.554, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.117408906882591, | |
| "grad_norm": 0.6376826166358697, | |
| "learning_rate": 3.8180874970058826e-05, | |
| "loss": 0.5284, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.1214574898785425, | |
| "grad_norm": 0.6941750074358048, | |
| "learning_rate": 3.8157236089188103e-05, | |
| "loss": 0.5268, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.125506072874494, | |
| "grad_norm": 0.596763822453056, | |
| "learning_rate": 3.813345202354145e-05, | |
| "loss": 0.5252, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1295546558704452, | |
| "grad_norm": 0.7524897128698806, | |
| "learning_rate": 3.810952296329563e-05, | |
| "loss": 0.5309, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1336032388663968, | |
| "grad_norm": 0.6263399153876753, | |
| "learning_rate": 3.808544909978676e-05, | |
| "loss": 0.5357, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1376518218623481, | |
| "grad_norm": 0.7833828552542226, | |
| "learning_rate": 3.806123062550882e-05, | |
| "loss": 0.5412, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1417004048582995, | |
| "grad_norm": 0.6092492555583845, | |
| "learning_rate": 3.803686773411208e-05, | |
| "loss": 0.5301, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.145748987854251, | |
| "grad_norm": 0.7189918833299878, | |
| "learning_rate": 3.801236062040158e-05, | |
| "loss": 0.5619, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1497975708502024, | |
| "grad_norm": 0.6021006520940442, | |
| "learning_rate": 3.798770948033553e-05, | |
| "loss": 0.5422, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.6009512421361342, | |
| "learning_rate": 3.796291451102381e-05, | |
| "loss": 0.5432, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.6010217116932527, | |
| "learning_rate": 3.793797591072631e-05, | |
| "loss": 0.5355, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1619433198380567, | |
| "grad_norm": 0.6648676346478087, | |
| "learning_rate": 3.791289387885144e-05, | |
| "loss": 0.5826, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.165991902834008, | |
| "grad_norm": 0.6300161266871862, | |
| "learning_rate": 3.788766861595443e-05, | |
| "loss": 0.5394, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1700404858299596, | |
| "grad_norm": 0.6007345701650114, | |
| "learning_rate": 3.7862300323735835e-05, | |
| "loss": 0.5555, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.174089068825911, | |
| "grad_norm": 0.6003604373574547, | |
| "learning_rate": 3.783678920503983e-05, | |
| "loss": 0.5517, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1781376518218623, | |
| "grad_norm": 0.5831279271761669, | |
| "learning_rate": 3.781113546385265e-05, | |
| "loss": 0.5519, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1821862348178138, | |
| "grad_norm": 0.6118872166040743, | |
| "learning_rate": 3.7785339305300916e-05, | |
| "loss": 0.5588, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1862348178137652, | |
| "grad_norm": 0.5354970908096294, | |
| "learning_rate": 3.7759400935650025e-05, | |
| "loss": 0.5354, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1902834008097165, | |
| "grad_norm": 0.7042049027131789, | |
| "learning_rate": 3.77333205623025e-05, | |
| "loss": 0.5639, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.194331983805668, | |
| "grad_norm": 0.5834965502619284, | |
| "learning_rate": 3.7707098393796295e-05, | |
| "loss": 0.5276, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1983805668016194, | |
| "grad_norm": 0.7150777002410401, | |
| "learning_rate": 3.7680734639803185e-05, | |
| "loss": 0.5679, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2024291497975708, | |
| "grad_norm": 0.6278212235106412, | |
| "learning_rate": 3.765422951112704e-05, | |
| "loss": 0.5371, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2064777327935223, | |
| "grad_norm": 0.6953895979489594, | |
| "learning_rate": 3.762758321970216e-05, | |
| "loss": 0.5582, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.5504372790417611, | |
| "learning_rate": 3.7600795978591584e-05, | |
| "loss": 0.5409, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.214574898785425, | |
| "grad_norm": 0.5669220073634287, | |
| "learning_rate": 3.757386800198538e-05, | |
| "loss": 0.5445, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2186234817813766, | |
| "grad_norm": 0.5572864533628222, | |
| "learning_rate": 3.7546799505198925e-05, | |
| "loss": 0.5277, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.222672064777328, | |
| "grad_norm": 0.6124166605488833, | |
| "learning_rate": 3.7519590704671197e-05, | |
| "loss": 0.5234, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2267206477732793, | |
| "grad_norm": 0.5462690864182183, | |
| "learning_rate": 3.749224181796305e-05, | |
| "loss": 0.541, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.5775938845822195, | |
| "learning_rate": 3.746475306375545e-05, | |
| "loss": 0.537, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2348178137651822, | |
| "grad_norm": 0.6457843445891075, | |
| "learning_rate": 3.743712466184774e-05, | |
| "loss": 0.5701, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2388663967611335, | |
| "grad_norm": 0.5681704152125098, | |
| "learning_rate": 3.7409356833155885e-05, | |
| "loss": 0.5157, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.242914979757085, | |
| "grad_norm": 0.5897999659382217, | |
| "learning_rate": 3.73814497997107e-05, | |
| "loss": 0.5499, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2469635627530364, | |
| "grad_norm": 0.6259687674717236, | |
| "learning_rate": 3.7353403784656095e-05, | |
| "loss": 0.5448, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2510121457489878, | |
| "grad_norm": 0.6417441069615476, | |
| "learning_rate": 3.732521901224724e-05, | |
| "loss": 0.5528, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2550607287449393, | |
| "grad_norm": 0.5754915989392557, | |
| "learning_rate": 3.7296895707848844e-05, | |
| "loss": 0.544, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2591093117408907, | |
| "grad_norm": 0.5106040191109841, | |
| "learning_rate": 3.7268434097933275e-05, | |
| "loss": 0.5578, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.545244136482479, | |
| "learning_rate": 3.723983441007882e-05, | |
| "loss": 0.5606, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2672064777327936, | |
| "grad_norm": 0.5872273864896124, | |
| "learning_rate": 3.721109687296781e-05, | |
| "loss": 0.5468, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.271255060728745, | |
| "grad_norm": 0.5754993879748829, | |
| "learning_rate": 3.718222171638484e-05, | |
| "loss": 0.5351, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2753036437246963, | |
| "grad_norm": 0.5882399380283807, | |
| "learning_rate": 3.715320917121488e-05, | |
| "loss": 0.5382, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2793522267206479, | |
| "grad_norm": 0.6131648109022995, | |
| "learning_rate": 3.712405946944149e-05, | |
| "loss": 0.5456, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2834008097165992, | |
| "grad_norm": 0.5740254099924404, | |
| "learning_rate": 3.709477284414488e-05, | |
| "loss": 0.5608, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2874493927125505, | |
| "grad_norm": 0.5546168473482324, | |
| "learning_rate": 3.7065349529500136e-05, | |
| "loss": 0.5355, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.291497975708502, | |
| "grad_norm": 0.5797902372179542, | |
| "learning_rate": 3.7035789760775305e-05, | |
| "loss": 0.5414, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2955465587044535, | |
| "grad_norm": 0.5862294906217036, | |
| "learning_rate": 3.70060937743295e-05, | |
| "loss": 0.5637, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2995951417004048, | |
| "grad_norm": 0.569593447723245, | |
| "learning_rate": 3.6976261807611045e-05, | |
| "loss": 0.5428, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.3036437246963564, | |
| "grad_norm": 0.6002991469764011, | |
| "learning_rate": 3.694629409915555e-05, | |
| "loss": 0.5221, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.567061228385956, | |
| "learning_rate": 3.691619088858401e-05, | |
| "loss": 0.5476, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.311740890688259, | |
| "grad_norm": 0.6632965857303368, | |
| "learning_rate": 3.68859524166009e-05, | |
| "loss": 0.5263, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.5343289153766135, | |
| "learning_rate": 3.6855578924992226e-05, | |
| "loss": 0.5456, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.319838056680162, | |
| "grad_norm": 0.5867644322617519, | |
| "learning_rate": 3.6825070656623626e-05, | |
| "loss": 0.57, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.3238866396761133, | |
| "grad_norm": 0.5772992795615821, | |
| "learning_rate": 3.67944278554384e-05, | |
| "loss": 0.5593, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3279352226720649, | |
| "grad_norm": 0.6110762225963609, | |
| "learning_rate": 3.676365076645557e-05, | |
| "loss": 0.5392, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.5695374438272516, | |
| "learning_rate": 3.673273963576791e-05, | |
| "loss": 0.5494, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3360323886639676, | |
| "grad_norm": 0.5629262111401677, | |
| "learning_rate": 3.670169471054001e-05, | |
| "loss": 0.5521, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3400809716599191, | |
| "grad_norm": 0.6095153050904223, | |
| "learning_rate": 3.6670516239006254e-05, | |
| "loss": 0.5493, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3441295546558705, | |
| "grad_norm": 0.5326737643943468, | |
| "learning_rate": 3.663920447046888e-05, | |
| "loss": 0.5401, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3481781376518218, | |
| "grad_norm": 0.6557233373155459, | |
| "learning_rate": 3.6607759655295954e-05, | |
| "loss": 0.5536, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3522267206477734, | |
| "grad_norm": 0.6739890446587424, | |
| "learning_rate": 3.657618204491937e-05, | |
| "loss": 0.5466, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3562753036437247, | |
| "grad_norm": 0.6721067694332786, | |
| "learning_rate": 3.6544471891832875e-05, | |
| "loss": 0.5318, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.360323886639676, | |
| "grad_norm": 0.6661207144274902, | |
| "learning_rate": 3.6512629449589983e-05, | |
| "loss": 0.5388, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3643724696356276, | |
| "grad_norm": 0.5566881010894584, | |
| "learning_rate": 3.6480654972802045e-05, | |
| "loss": 0.5373, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.6950586299262039, | |
| "learning_rate": 3.644854871713611e-05, | |
| "loss": 0.5377, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3724696356275303, | |
| "grad_norm": 0.5219071720830948, | |
| "learning_rate": 3.6416310939312924e-05, | |
| "loss": 0.5298, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.376518218623482, | |
| "grad_norm": 0.5557450469846117, | |
| "learning_rate": 3.638394189710493e-05, | |
| "loss": 0.5315, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3805668016194332, | |
| "grad_norm": 0.5955808271916493, | |
| "learning_rate": 3.635144184933412e-05, | |
| "loss": 0.5304, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.559552947015133, | |
| "learning_rate": 3.6318811055869994e-05, | |
| "loss": 0.5299, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3886639676113361, | |
| "grad_norm": 0.6041286553522512, | |
| "learning_rate": 3.628604977762752e-05, | |
| "loss": 0.5479, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3927125506072875, | |
| "grad_norm": 0.5765495940261955, | |
| "learning_rate": 3.6253158276565004e-05, | |
| "loss": 0.5297, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3967611336032388, | |
| "grad_norm": 0.6649788630549403, | |
| "learning_rate": 3.6220136815682016e-05, | |
| "loss": 0.5537, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4008097165991904, | |
| "grad_norm": 0.5589894269640473, | |
| "learning_rate": 3.618698565901727e-05, | |
| "loss": 0.5589, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.4048582995951417, | |
| "grad_norm": 0.6683224137523348, | |
| "learning_rate": 3.615370507164654e-05, | |
| "loss": 0.5521, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.408906882591093, | |
| "grad_norm": 0.547556007325787, | |
| "learning_rate": 3.61202953196805e-05, | |
| "loss": 0.5347, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.4129554655870447, | |
| "grad_norm": 0.6399196619907583, | |
| "learning_rate": 3.6086756670262654e-05, | |
| "loss": 0.551, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.417004048582996, | |
| "grad_norm": 0.6611682151625241, | |
| "learning_rate": 3.6053089391567135e-05, | |
| "loss": 0.5763, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.5579018463871099, | |
| "learning_rate": 3.601929375279662e-05, | |
| "loss": 0.5411, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.425101214574899, | |
| "grad_norm": 0.7563094231618028, | |
| "learning_rate": 3.598537002418012e-05, | |
| "loss": 0.5229, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4291497975708503, | |
| "grad_norm": 0.5887367919853754, | |
| "learning_rate": 3.5951318476970885e-05, | |
| "loss": 0.5524, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4331983805668016, | |
| "grad_norm": 0.6543502616702487, | |
| "learning_rate": 3.591713938344416e-05, | |
| "loss": 0.5146, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4372469635627532, | |
| "grad_norm": 0.6057250327871209, | |
| "learning_rate": 3.588283301689507e-05, | |
| "loss": 0.5353, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4412955465587045, | |
| "grad_norm": 0.5820032058713939, | |
| "learning_rate": 3.5848399651636424e-05, | |
| "loss": 0.5562, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4453441295546559, | |
| "grad_norm": 0.6385631617320232, | |
| "learning_rate": 3.5813839562996473e-05, | |
| "loss": 0.5412, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4493927125506074, | |
| "grad_norm": 0.5865934693824049, | |
| "learning_rate": 3.5779153027316775e-05, | |
| "loss": 0.5425, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4534412955465588, | |
| "grad_norm": 0.5550142272368002, | |
| "learning_rate": 3.574434032194994e-05, | |
| "loss": 0.5289, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.45748987854251, | |
| "grad_norm": 0.5934744183274968, | |
| "learning_rate": 3.5709401725257436e-05, | |
| "loss": 0.5451, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.5629221450756171, | |
| "learning_rate": 3.5674337516607346e-05, | |
| "loss": 0.5789, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.465587044534413, | |
| "grad_norm": 0.6082145521760911, | |
| "learning_rate": 3.563914797637216e-05, | |
| "loss": 0.5456, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4696356275303644, | |
| "grad_norm": 0.5641789786767606, | |
| "learning_rate": 3.560383338592649e-05, | |
| "loss": 0.5548, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.5795094290481424, | |
| "learning_rate": 3.556839402764487e-05, | |
| "loss": 0.54, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4777327935222673, | |
| "grad_norm": 0.5956859698207708, | |
| "learning_rate": 3.553283018489946e-05, | |
| "loss": 0.5286, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4817813765182186, | |
| "grad_norm": 0.5792549404250356, | |
| "learning_rate": 3.54971421420578e-05, | |
| "loss": 0.551, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.48582995951417, | |
| "grad_norm": 0.6270285715283056, | |
| "learning_rate": 3.5461330184480524e-05, | |
| "loss": 0.5569, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4898785425101215, | |
| "grad_norm": 0.5815526370985188, | |
| "learning_rate": 3.54253945985191e-05, | |
| "loss": 0.5351, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4939271255060729, | |
| "grad_norm": 0.5968846340100084, | |
| "learning_rate": 3.5389335671513524e-05, | |
| "loss": 0.5162, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4979757085020242, | |
| "grad_norm": 0.5546346664663212, | |
| "learning_rate": 3.5353153691789986e-05, | |
| "loss": 0.5273, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5020242914979756, | |
| "grad_norm": 0.5608288966154504, | |
| "learning_rate": 3.531684894865866e-05, | |
| "loss": 0.5108, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.5060728744939271, | |
| "grad_norm": 0.5644599650292882, | |
| "learning_rate": 3.528042173241128e-05, | |
| "loss": 0.5774, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5101214574898787, | |
| "grad_norm": 0.6082633491707109, | |
| "learning_rate": 3.524387233431892e-05, | |
| "loss": 0.5459, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.5141700404858298, | |
| "grad_norm": 0.5419590961228211, | |
| "learning_rate": 3.520720104662958e-05, | |
| "loss": 0.5067, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5182186234817814, | |
| "grad_norm": 0.5662785352434235, | |
| "learning_rate": 3.5170408162565904e-05, | |
| "loss": 0.5164, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.522267206477733, | |
| "grad_norm": 0.536198525658535, | |
| "learning_rate": 3.513349397632281e-05, | |
| "loss": 0.5593, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.5844478050606557, | |
| "learning_rate": 3.5096458783065145e-05, | |
| "loss": 0.5668, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5303643724696356, | |
| "grad_norm": 0.6122167346599451, | |
| "learning_rate": 3.505930287892533e-05, | |
| "loss": 0.5655, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.5344129554655872, | |
| "grad_norm": 0.5369959092619967, | |
| "learning_rate": 3.502202656100099e-05, | |
| "loss": 0.5325, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.6338099215625523, | |
| "learning_rate": 3.498463012735256e-05, | |
| "loss": 0.5894, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.54251012145749, | |
| "grad_norm": 0.5744651359692033, | |
| "learning_rate": 3.494711387700094e-05, | |
| "loss": 0.5672, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5465587044534415, | |
| "grad_norm": 0.529511992578701, | |
| "learning_rate": 3.490947810992505e-05, | |
| "loss": 0.55, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5506072874493926, | |
| "grad_norm": 0.5279033292032922, | |
| "learning_rate": 3.48717231270595e-05, | |
| "loss": 0.553, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5546558704453441, | |
| "grad_norm": 0.5186115238733202, | |
| "learning_rate": 3.483384923029211e-05, | |
| "loss": 0.539, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5587044534412957, | |
| "grad_norm": 0.5882203534596091, | |
| "learning_rate": 3.479585672246156e-05, | |
| "loss": 0.5455, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5627530364372468, | |
| "grad_norm": 0.514069310206094, | |
| "learning_rate": 3.4757745907354924e-05, | |
| "loss": 0.5462, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5668016194331984, | |
| "grad_norm": 0.5502091616925835, | |
| "learning_rate": 3.471951708970527e-05, | |
| "loss": 0.5307, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5708502024291497, | |
| "grad_norm": 0.5771966363124096, | |
| "learning_rate": 3.468117057518921e-05, | |
| "loss": 0.5383, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.574898785425101, | |
| "grad_norm": 0.5340839813302571, | |
| "learning_rate": 3.4642706670424426e-05, | |
| "loss": 0.5332, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.6656264978893606, | |
| "learning_rate": 3.460412568296731e-05, | |
| "loss": 0.556, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.582995951417004, | |
| "grad_norm": 0.5519183995519735, | |
| "learning_rate": 3.456542792131039e-05, | |
| "loss": 0.544, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5870445344129553, | |
| "grad_norm": 0.5873050823156413, | |
| "learning_rate": 3.4526613694879936e-05, | |
| "loss": 0.5524, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.591093117408907, | |
| "grad_norm": 0.6168539598242996, | |
| "learning_rate": 3.448768331403348e-05, | |
| "loss": 0.5752, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5951417004048583, | |
| "grad_norm": 0.6044858312983241, | |
| "learning_rate": 3.44486370900573e-05, | |
| "loss": 0.5524, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5991902834008096, | |
| "grad_norm": 0.5633922374434579, | |
| "learning_rate": 3.440947533516396e-05, | |
| "loss": 0.5389, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.6032388663967612, | |
| "grad_norm": 0.5330770463866848, | |
| "learning_rate": 3.4370198362489816e-05, | |
| "loss": 0.5599, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.6072874493927125, | |
| "grad_norm": 0.5447616192273551, | |
| "learning_rate": 3.43308064860925e-05, | |
| "loss": 0.5259, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.6113360323886639, | |
| "grad_norm": 0.5112292537853833, | |
| "learning_rate": 3.4291300020948393e-05, | |
| "loss": 0.5238, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.4912327772029144, | |
| "learning_rate": 3.425167928295015e-05, | |
| "loss": 0.5019, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6194331983805668, | |
| "grad_norm": 0.6075709234869082, | |
| "learning_rate": 3.421194458890411e-05, | |
| "loss": 0.536, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.623481781376518, | |
| "grad_norm": 0.5218679372887354, | |
| "learning_rate": 3.4172096256527865e-05, | |
| "loss": 0.5414, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.6275303643724697, | |
| "grad_norm": 0.5839809665990388, | |
| "learning_rate": 3.4132134604447595e-05, | |
| "loss": 0.5667, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.6129771127558679, | |
| "learning_rate": 3.409205995219561e-05, | |
| "loss": 0.5715, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6356275303643724, | |
| "grad_norm": 0.566283992858257, | |
| "learning_rate": 3.4051872620207765e-05, | |
| "loss": 0.5548, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.639676113360324, | |
| "grad_norm": 0.6790463210859209, | |
| "learning_rate": 3.40115729298209e-05, | |
| "loss": 0.5487, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6437246963562753, | |
| "grad_norm": 0.6438336007530644, | |
| "learning_rate": 3.397116120327025e-05, | |
| "loss": 0.5427, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6477732793522266, | |
| "grad_norm": 0.5945159545850142, | |
| "learning_rate": 3.3930637763686916e-05, | |
| "loss": 0.5488, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6518218623481782, | |
| "grad_norm": 0.6806373441817244, | |
| "learning_rate": 3.389000293509524e-05, | |
| "loss": 0.55, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6558704453441295, | |
| "grad_norm": 0.5336334902961299, | |
| "learning_rate": 3.3849257042410225e-05, | |
| "loss": 0.529, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6599190283400809, | |
| "grad_norm": 0.5002109782008192, | |
| "learning_rate": 3.380840041143494e-05, | |
| "loss": 0.5322, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6639676113360324, | |
| "grad_norm": 0.5768329713699226, | |
| "learning_rate": 3.376743336885792e-05, | |
| "loss": 0.5551, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6680161943319838, | |
| "grad_norm": 0.5001956399956409, | |
| "learning_rate": 3.372635624225055e-05, | |
| "loss": 0.5419, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6720647773279351, | |
| "grad_norm": 0.5145189509761202, | |
| "learning_rate": 3.3685169360064445e-05, | |
| "loss": 0.5353, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6761133603238867, | |
| "grad_norm": 0.49320154564804985, | |
| "learning_rate": 3.3643873051628804e-05, | |
| "loss": 0.55, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.680161943319838, | |
| "grad_norm": 0.5238348597462116, | |
| "learning_rate": 3.3602467647147814e-05, | |
| "loss": 0.5771, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.5643572128725495, | |
| "learning_rate": 3.356095347769799e-05, | |
| "loss": 0.5841, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.688259109311741, | |
| "grad_norm": 0.558343384731941, | |
| "learning_rate": 3.351933087522552e-05, | |
| "loss": 0.5287, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.4529906739260511, | |
| "learning_rate": 3.3477600172543626e-05, | |
| "loss": 0.5343, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6963562753036436, | |
| "grad_norm": 0.5947032085934537, | |
| "learning_rate": 3.3435761703329894e-05, | |
| "loss": 0.5336, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.7004048582995952, | |
| "grad_norm": 0.5640322426540256, | |
| "learning_rate": 3.3393815802123606e-05, | |
| "loss": 0.5308, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7044534412955465, | |
| "grad_norm": 0.5724692380655311, | |
| "learning_rate": 3.335176280432308e-05, | |
| "loss": 0.562, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.708502024291498, | |
| "grad_norm": 0.5944370672293678, | |
| "learning_rate": 3.330960304618295e-05, | |
| "loss": 0.546, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.7125506072874495, | |
| "grad_norm": 0.5576422057916952, | |
| "learning_rate": 3.326733686481153e-05, | |
| "loss": 0.5701, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.7165991902834008, | |
| "grad_norm": 0.5616689591267434, | |
| "learning_rate": 3.3224964598168084e-05, | |
| "loss": 0.5511, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.5571043410963156, | |
| "learning_rate": 3.318248658506012e-05, | |
| "loss": 0.5624, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7246963562753037, | |
| "grad_norm": 0.539162929806453, | |
| "learning_rate": 3.313990316514069e-05, | |
| "loss": 0.5399, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.728744939271255, | |
| "grad_norm": 0.5413554734432351, | |
| "learning_rate": 3.309721467890571e-05, | |
| "loss": 0.5482, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.7327935222672064, | |
| "grad_norm": 0.5521913520272806, | |
| "learning_rate": 3.305442146769115e-05, | |
| "loss": 0.5375, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.5444437237173824, | |
| "learning_rate": 3.30115238736704e-05, | |
| "loss": 0.5473, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7408906882591093, | |
| "grad_norm": 0.6055519244496669, | |
| "learning_rate": 3.296852223985148e-05, | |
| "loss": 0.5261, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7449392712550607, | |
| "grad_norm": 0.4653868044530118, | |
| "learning_rate": 3.2925416910074315e-05, | |
| "loss": 0.5312, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7489878542510122, | |
| "grad_norm": 0.6011992637620894, | |
| "learning_rate": 3.288220822900796e-05, | |
| "loss": 0.5556, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7530364372469636, | |
| "grad_norm": 0.49568927000828217, | |
| "learning_rate": 3.283889654214788e-05, | |
| "loss": 0.5582, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.757085020242915, | |
| "grad_norm": 0.5831165347767162, | |
| "learning_rate": 3.279548219581317e-05, | |
| "loss": 0.5529, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7611336032388665, | |
| "grad_norm": 0.530048335838685, | |
| "learning_rate": 3.275196553714379e-05, | |
| "loss": 0.5448, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7651821862348178, | |
| "grad_norm": 0.5727268029651259, | |
| "learning_rate": 3.270834691409778e-05, | |
| "loss": 0.526, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.5021908546509489, | |
| "learning_rate": 3.266462667544849e-05, | |
| "loss": 0.5455, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7732793522267207, | |
| "grad_norm": 0.5205377756259043, | |
| "learning_rate": 3.262080517078178e-05, | |
| "loss": 0.5742, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.777327935222672, | |
| "grad_norm": 0.5823634286645456, | |
| "learning_rate": 3.257688275049323e-05, | |
| "loss": 0.5546, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7813765182186234, | |
| "grad_norm": 0.5444535885126472, | |
| "learning_rate": 3.253285976578535e-05, | |
| "loss": 0.5307, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.785425101214575, | |
| "grad_norm": 0.6300360383737398, | |
| "learning_rate": 3.2488736568664756e-05, | |
| "loss": 0.563, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.5970166335647262, | |
| "learning_rate": 3.244451351193935e-05, | |
| "loss": 0.5617, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7935222672064777, | |
| "grad_norm": 0.6338193924827893, | |
| "learning_rate": 3.240019094921551e-05, | |
| "loss": 0.5564, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7975708502024292, | |
| "grad_norm": 0.5767859802877245, | |
| "learning_rate": 3.235576923489527e-05, | |
| "loss": 0.5306, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.8016194331983806, | |
| "grad_norm": 0.5737943792727452, | |
| "learning_rate": 3.2311248724173476e-05, | |
| "loss": 0.558, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.805668016194332, | |
| "grad_norm": 0.5689720674508204, | |
| "learning_rate": 3.226662977303494e-05, | |
| "loss": 0.5494, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.8097165991902835, | |
| "grad_norm": 0.5364716270591355, | |
| "learning_rate": 3.2221912738251626e-05, | |
| "loss": 0.5359, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8137651821862348, | |
| "grad_norm": 0.569966398821834, | |
| "learning_rate": 3.2177097977379736e-05, | |
| "loss": 0.5442, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.8178137651821862, | |
| "grad_norm": 0.5656885254515, | |
| "learning_rate": 3.21321858487569e-05, | |
| "loss": 0.5421, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8218623481781377, | |
| "grad_norm": 0.5774244787011716, | |
| "learning_rate": 3.208717671149931e-05, | |
| "loss": 0.5293, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.825910931174089, | |
| "grad_norm": 0.5374855598707884, | |
| "learning_rate": 3.204207092549882e-05, | |
| "loss": 0.5449, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.8299595141700404, | |
| "grad_norm": 0.6067614817978467, | |
| "learning_rate": 3.19968688514201e-05, | |
| "loss": 0.5478, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.834008097165992, | |
| "grad_norm": 0.5543718086941022, | |
| "learning_rate": 3.195157085069774e-05, | |
| "loss": 0.5247, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8380566801619433, | |
| "grad_norm": 0.61010393795452, | |
| "learning_rate": 3.190617728553332e-05, | |
| "loss": 0.5517, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.5572788154819422, | |
| "learning_rate": 3.1860688518892606e-05, | |
| "loss": 0.5613, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.6664612906206048, | |
| "learning_rate": 3.181510491450254e-05, | |
| "loss": 0.5552, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8502024291497976, | |
| "grad_norm": 0.5931313478411449, | |
| "learning_rate": 3.176942683684842e-05, | |
| "loss": 0.5458, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.854251012145749, | |
| "grad_norm": 0.5275488897673468, | |
| "learning_rate": 3.1723654651170934e-05, | |
| "loss": 0.5407, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8582995951417005, | |
| "grad_norm": 0.5958881562496423, | |
| "learning_rate": 3.167778872346327e-05, | |
| "loss": 0.5365, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8623481781376519, | |
| "grad_norm": 0.6452587052874706, | |
| "learning_rate": 3.163182942046815e-05, | |
| "loss": 0.5391, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8663967611336032, | |
| "grad_norm": 0.5348781634011814, | |
| "learning_rate": 3.158577710967494e-05, | |
| "loss": 0.5521, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8704453441295548, | |
| "grad_norm": 0.6130594700515292, | |
| "learning_rate": 3.153963215931669e-05, | |
| "loss": 0.5304, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.874493927125506, | |
| "grad_norm": 0.5544837270894496, | |
| "learning_rate": 3.14933949383672e-05, | |
| "loss": 0.5349, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8785425101214575, | |
| "grad_norm": 0.5081747795656953, | |
| "learning_rate": 3.144706581653805e-05, | |
| "loss": 0.546, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.882591093117409, | |
| "grad_norm": 0.6281753831828862, | |
| "learning_rate": 3.140064516427566e-05, | |
| "loss": 0.5565, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8866396761133604, | |
| "grad_norm": 0.49985151638828496, | |
| "learning_rate": 3.135413335275833e-05, | |
| "loss": 0.5383, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8906882591093117, | |
| "grad_norm": 0.5270765766045398, | |
| "learning_rate": 3.130753075389327e-05, | |
| "loss": 0.5592, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.5254788893661382, | |
| "learning_rate": 3.1260837740313625e-05, | |
| "loss": 0.5102, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8987854251012146, | |
| "grad_norm": 0.5158942980783344, | |
| "learning_rate": 3.121405468537547e-05, | |
| "loss": 0.5286, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.902834008097166, | |
| "grad_norm": 0.6116143270668203, | |
| "learning_rate": 3.116718196315488e-05, | |
| "loss": 0.5584, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9068825910931175, | |
| "grad_norm": 0.536586506698827, | |
| "learning_rate": 3.1120219948444894e-05, | |
| "loss": 0.5364, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.9109311740890689, | |
| "grad_norm": 0.631067530857748, | |
| "learning_rate": 3.107316901675252e-05, | |
| "loss": 0.5588, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.9149797570850202, | |
| "grad_norm": 0.6190403558272248, | |
| "learning_rate": 3.1026029544295745e-05, | |
| "loss": 0.5473, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.9190283400809718, | |
| "grad_norm": 0.526030730587988, | |
| "learning_rate": 3.097880190800057e-05, | |
| "loss": 0.5283, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.6904296733831176, | |
| "learning_rate": 3.093148648549788e-05, | |
| "loss": 0.5474, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9271255060728745, | |
| "grad_norm": 0.5025286521482776, | |
| "learning_rate": 3.088408365512055e-05, | |
| "loss": 0.5393, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.931174089068826, | |
| "grad_norm": 0.6331148379906872, | |
| "learning_rate": 3.083659379590034e-05, | |
| "loss": 0.5428, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.9352226720647774, | |
| "grad_norm": 0.584055353480832, | |
| "learning_rate": 3.0789017287564913e-05, | |
| "loss": 0.5295, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9392712550607287, | |
| "grad_norm": 0.5164731051140695, | |
| "learning_rate": 3.074135451053475e-05, | |
| "loss": 0.5157, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9433198380566803, | |
| "grad_norm": 0.6197466020349894, | |
| "learning_rate": 3.069360584592015e-05, | |
| "loss": 0.5695, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.5719739711328328, | |
| "learning_rate": 3.064577167551816e-05, | |
| "loss": 0.5604, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.951417004048583, | |
| "grad_norm": 0.49644960536962585, | |
| "learning_rate": 3.059785238180953e-05, | |
| "loss": 0.5458, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9554655870445345, | |
| "grad_norm": 0.561350589777537, | |
| "learning_rate": 3.054984834795566e-05, | |
| "loss": 0.5345, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9595141700404857, | |
| "grad_norm": 0.5271431795586062, | |
| "learning_rate": 3.0501759957795537e-05, | |
| "loss": 0.5553, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9635627530364372, | |
| "grad_norm": 0.615054608837459, | |
| "learning_rate": 3.0453587595842627e-05, | |
| "loss": 0.5563, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9676113360323888, | |
| "grad_norm": 0.5814784397552187, | |
| "learning_rate": 3.0405331647281856e-05, | |
| "loss": 0.5334, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.97165991902834, | |
| "grad_norm": 0.6223839353202709, | |
| "learning_rate": 3.0356992497966508e-05, | |
| "loss": 0.5769, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9757085020242915, | |
| "grad_norm": 0.5052312623545493, | |
| "learning_rate": 3.0308570534415137e-05, | |
| "loss": 0.5377, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.979757085020243, | |
| "grad_norm": 0.6195507857689725, | |
| "learning_rate": 3.0260066143808467e-05, | |
| "loss": 0.5501, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9838056680161942, | |
| "grad_norm": 0.4913975732511621, | |
| "learning_rate": 3.021147971398632e-05, | |
| "loss": 0.521, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9878542510121457, | |
| "grad_norm": 0.5944431021570141, | |
| "learning_rate": 3.0162811633444478e-05, | |
| "loss": 0.5517, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9919028340080973, | |
| "grad_norm": 0.583142815087593, | |
| "learning_rate": 3.0114062291331635e-05, | |
| "loss": 0.5681, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9959514170040484, | |
| "grad_norm": 0.535225913484151, | |
| "learning_rate": 3.0065232077446217e-05, | |
| "loss": 0.5344, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5162834709847981, | |
| "learning_rate": 3.001632138223332e-05, | |
| "loss": 0.5217, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.0040485829959516, | |
| "grad_norm": 0.8233404625130921, | |
| "learning_rate": 2.9967330596781565e-05, | |
| "loss": 0.3824, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0080971659919027, | |
| "grad_norm": 0.4807300670595259, | |
| "learning_rate": 2.9918260112819952e-05, | |
| "loss": 0.3432, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0121457489878543, | |
| "grad_norm": 1.5832761680955654, | |
| "learning_rate": 2.9869110322714778e-05, | |
| "loss": 0.3656, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.016194331983806, | |
| "grad_norm": 0.8365859077289611, | |
| "learning_rate": 2.9819881619466447e-05, | |
| "loss": 0.4002, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.020242914979757, | |
| "grad_norm": 0.6180493599864167, | |
| "learning_rate": 2.9770574396706364e-05, | |
| "loss": 0.3635, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0242914979757085, | |
| "grad_norm": 0.6370516348195722, | |
| "learning_rate": 2.972118904869376e-05, | |
| "loss": 0.351, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02834008097166, | |
| "grad_norm": 0.6055755925156562, | |
| "learning_rate": 2.9671725970312563e-05, | |
| "loss": 0.3708, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.032388663967611, | |
| "grad_norm": 0.6609206505517798, | |
| "learning_rate": 2.9622185557068238e-05, | |
| "loss": 0.3645, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.0364372469635628, | |
| "grad_norm": 0.6176511751254831, | |
| "learning_rate": 2.95725682050846e-05, | |
| "loss": 0.3873, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0404858299595143, | |
| "grad_norm": 0.6349895381103957, | |
| "learning_rate": 2.9522874311100684e-05, | |
| "loss": 0.3877, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0445344129554655, | |
| "grad_norm": 0.617644333979201, | |
| "learning_rate": 2.9473104272467526e-05, | |
| "loss": 0.3724, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.048582995951417, | |
| "grad_norm": 0.5964703309716946, | |
| "learning_rate": 2.942325848714505e-05, | |
| "loss": 0.3888, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.6198621582504845, | |
| "learning_rate": 2.9373337353698804e-05, | |
| "loss": 0.3774, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.0566801619433197, | |
| "grad_norm": 0.5519777006090475, | |
| "learning_rate": 2.932334127129686e-05, | |
| "loss": 0.349, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.0607287449392713, | |
| "grad_norm": 0.5668495912768947, | |
| "learning_rate": 2.9273270639706547e-05, | |
| "loss": 0.3701, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.064777327935223, | |
| "grad_norm": 0.6159025219544635, | |
| "learning_rate": 2.922312585929131e-05, | |
| "loss": 0.3775, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.068825910931174, | |
| "grad_norm": 0.5824619327176122, | |
| "learning_rate": 2.9172907331007478e-05, | |
| "loss": 0.3632, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.0728744939271255, | |
| "grad_norm": 0.6185861432850418, | |
| "learning_rate": 2.9122615456401053e-05, | |
| "loss": 0.3589, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.5127900445159829, | |
| "learning_rate": 2.9072250637604535e-05, | |
| "loss": 0.3624, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.080971659919028, | |
| "grad_norm": 0.5443410616158665, | |
| "learning_rate": 2.9021813277333667e-05, | |
| "loss": 0.3601, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.08502024291498, | |
| "grad_norm": 0.5397146215405616, | |
| "learning_rate": 2.8971303778884244e-05, | |
| "loss": 0.3797, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.0890688259109313, | |
| "grad_norm": 0.529247135872998, | |
| "learning_rate": 2.8920722546128863e-05, | |
| "loss": 0.3842, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0931174089068825, | |
| "grad_norm": 0.6127073701502763, | |
| "learning_rate": 2.887006998351372e-05, | |
| "loss": 0.367, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.097165991902834, | |
| "grad_norm": 0.45085050608245836, | |
| "learning_rate": 2.881934649605535e-05, | |
| "loss": 0.3541, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.1012145748987856, | |
| "grad_norm": 0.5981308745768894, | |
| "learning_rate": 2.8768552489337408e-05, | |
| "loss": 0.3456, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.5086960901426772, | |
| "learning_rate": 2.8717688369507425e-05, | |
| "loss": 0.3528, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1093117408906883, | |
| "grad_norm": 0.5539005799129516, | |
| "learning_rate": 2.8666754543273535e-05, | |
| "loss": 0.363, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.11336032388664, | |
| "grad_norm": 0.5028629444615814, | |
| "learning_rate": 2.8615751417901273e-05, | |
| "loss": 0.3341, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.117408906882591, | |
| "grad_norm": 0.6111563818040204, | |
| "learning_rate": 2.856467940121025e-05, | |
| "loss": 0.3833, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.1214574898785425, | |
| "grad_norm": 0.5877364804655288, | |
| "learning_rate": 2.8513538901570964e-05, | |
| "loss": 0.3348, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.125506072874494, | |
| "grad_norm": 0.5869264260928779, | |
| "learning_rate": 2.846233032790149e-05, | |
| "loss": 0.3363, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1295546558704452, | |
| "grad_norm": 0.5384586229822572, | |
| "learning_rate": 2.84110540896642e-05, | |
| "loss": 0.3685, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.133603238866397, | |
| "grad_norm": 0.6213359450676945, | |
| "learning_rate": 2.8359710596862555e-05, | |
| "loss": 0.365, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.1376518218623484, | |
| "grad_norm": 0.48379554799834434, | |
| "learning_rate": 2.8308300260037734e-05, | |
| "loss": 0.362, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.1417004048582995, | |
| "grad_norm": 0.5858134307026769, | |
| "learning_rate": 2.8256823490265424e-05, | |
| "loss": 0.3402, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.145748987854251, | |
| "grad_norm": 0.5169601895290833, | |
| "learning_rate": 2.8205280699152515e-05, | |
| "loss": 0.3779, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1497975708502026, | |
| "grad_norm": 0.5768606195663104, | |
| "learning_rate": 2.8153672298833777e-05, | |
| "loss": 0.3694, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.5858363029110957, | |
| "learning_rate": 2.8101998701968615e-05, | |
| "loss": 0.3396, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.5608182135983992, | |
| "learning_rate": 2.805026032173773e-05, | |
| "loss": 0.3708, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.161943319838057, | |
| "grad_norm": 0.5756578171652131, | |
| "learning_rate": 2.7998457571839832e-05, | |
| "loss": 0.3691, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.165991902834008, | |
| "grad_norm": 0.5642796780765593, | |
| "learning_rate": 2.7946590866488322e-05, | |
| "loss": 0.3471, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1700404858299596, | |
| "grad_norm": 0.5412212238589666, | |
| "learning_rate": 2.7894660620408e-05, | |
| "loss": 0.3579, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.174089068825911, | |
| "grad_norm": 0.4999358059706049, | |
| "learning_rate": 2.784266724883173e-05, | |
| "loss": 0.3498, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.1781376518218623, | |
| "grad_norm": 0.6002158972090321, | |
| "learning_rate": 2.7790611167497128e-05, | |
| "loss": 0.3907, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.182186234817814, | |
| "grad_norm": 0.48735000305637977, | |
| "learning_rate": 2.7738492792643224e-05, | |
| "loss": 0.3512, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.1862348178137654, | |
| "grad_norm": 0.5310245710941207, | |
| "learning_rate": 2.768631254100716e-05, | |
| "loss": 0.3333, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1902834008097165, | |
| "grad_norm": 0.49262412083547447, | |
| "learning_rate": 2.7634070829820836e-05, | |
| "loss": 0.3489, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.194331983805668, | |
| "grad_norm": 0.4895186738526013, | |
| "learning_rate": 2.7581768076807592e-05, | |
| "loss": 0.3534, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.1983805668016196, | |
| "grad_norm": 0.498658532067327, | |
| "learning_rate": 2.7529404700178832e-05, | |
| "loss": 0.3492, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.2024291497975708, | |
| "grad_norm": 0.5056448117284382, | |
| "learning_rate": 2.7476981118630735e-05, | |
| "loss": 0.3706, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.2064777327935223, | |
| "grad_norm": 0.495741234903099, | |
| "learning_rate": 2.742449775134085e-05, | |
| "loss": 0.3587, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.5038931078395993, | |
| "learning_rate": 2.7371955017964788e-05, | |
| "loss": 0.3605, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.214574898785425, | |
| "grad_norm": 0.44839504438729466, | |
| "learning_rate": 2.7319353338632845e-05, | |
| "loss": 0.3782, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.2186234817813766, | |
| "grad_norm": 0.5550509523243179, | |
| "learning_rate": 2.7266693133946645e-05, | |
| "loss": 0.3609, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.2226720647773277, | |
| "grad_norm": 0.49299101297006614, | |
| "learning_rate": 2.7213974824975785e-05, | |
| "loss": 0.3425, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2267206477732793, | |
| "grad_norm": 0.47256315425281326, | |
| "learning_rate": 2.7161198833254464e-05, | |
| "loss": 0.3758, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.4941581048655325, | |
| "learning_rate": 2.7108365580778095e-05, | |
| "loss": 0.3848, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.234817813765182, | |
| "grad_norm": 0.4814562001206006, | |
| "learning_rate": 2.705547548999996e-05, | |
| "loss": 0.3721, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.2388663967611335, | |
| "grad_norm": 0.5077770185284652, | |
| "learning_rate": 2.7002528983827817e-05, | |
| "loss": 0.3723, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.242914979757085, | |
| "grad_norm": 0.5028572112429468, | |
| "learning_rate": 2.6949526485620517e-05, | |
| "loss": 0.3626, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.246963562753036, | |
| "grad_norm": 0.5268097358376536, | |
| "learning_rate": 2.6896468419184623e-05, | |
| "loss": 0.3851, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.251012145748988, | |
| "grad_norm": 0.5274137817083523, | |
| "learning_rate": 2.6843355208771024e-05, | |
| "loss": 0.3625, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.2550607287449393, | |
| "grad_norm": 0.5209596784380258, | |
| "learning_rate": 2.6790187279071536e-05, | |
| "loss": 0.3656, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.2591093117408905, | |
| "grad_norm": 0.4994520440631946, | |
| "learning_rate": 2.67369650552155e-05, | |
| "loss": 0.3714, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.503030881585558, | |
| "learning_rate": 2.6683688962766407e-05, | |
| "loss": 0.3476, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.2672064777327936, | |
| "grad_norm": 0.4789032682799846, | |
| "learning_rate": 2.6630359427718468e-05, | |
| "loss": 0.359, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2712550607287447, | |
| "grad_norm": 0.5300663342160954, | |
| "learning_rate": 2.6576976876493225e-05, | |
| "loss": 0.3603, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.5153370754263088, | |
| "learning_rate": 2.652354173593614e-05, | |
| "loss": 0.3697, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.279352226720648, | |
| "grad_norm": 0.4871085255554419, | |
| "learning_rate": 2.647005443331317e-05, | |
| "loss": 0.3782, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.283400809716599, | |
| "grad_norm": 0.5279970778709258, | |
| "learning_rate": 2.6416515396307354e-05, | |
| "loss": 0.3657, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.2874493927125505, | |
| "grad_norm": 0.5241347802299953, | |
| "learning_rate": 2.636292505301542e-05, | |
| "loss": 0.3568, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.291497975708502, | |
| "grad_norm": 0.5113046751548369, | |
| "learning_rate": 2.630928383194431e-05, | |
| "loss": 0.3691, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.2955465587044532, | |
| "grad_norm": 0.5556081829364512, | |
| "learning_rate": 2.625559216200781e-05, | |
| "loss": 0.3713, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.299595141700405, | |
| "grad_norm": 0.49513720441179676, | |
| "learning_rate": 2.620185047252308e-05, | |
| "loss": 0.3759, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.3036437246963564, | |
| "grad_norm": 0.5204102999521909, | |
| "learning_rate": 2.614805919320723e-05, | |
| "loss": 0.3653, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.48789458233862315, | |
| "learning_rate": 2.6094218754173895e-05, | |
| "loss": 0.3593, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.311740890688259, | |
| "grad_norm": 0.5256392467649599, | |
| "learning_rate": 2.60403295859298e-05, | |
| "loss": 0.3548, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.44959938081483203, | |
| "learning_rate": 2.5986392119371297e-05, | |
| "loss": 0.361, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.3198380566801617, | |
| "grad_norm": 0.5142476941789333, | |
| "learning_rate": 2.5932406785780928e-05, | |
| "loss": 0.3656, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.3238866396761133, | |
| "grad_norm": 0.47370024059031873, | |
| "learning_rate": 2.5878374016823987e-05, | |
| "loss": 0.3531, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.327935222672065, | |
| "grad_norm": 0.5429046529407853, | |
| "learning_rate": 2.5824294244545045e-05, | |
| "loss": 0.3785, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.331983805668016, | |
| "grad_norm": 0.5250033999333329, | |
| "learning_rate": 2.577016790136453e-05, | |
| "loss": 0.3615, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.3360323886639676, | |
| "grad_norm": 0.5059276052765423, | |
| "learning_rate": 2.5715995420075225e-05, | |
| "loss": 0.3775, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.340080971659919, | |
| "grad_norm": 0.5202793319030032, | |
| "learning_rate": 2.5661777233838864e-05, | |
| "loss": 0.3629, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.3441295546558703, | |
| "grad_norm": 0.4498902242892306, | |
| "learning_rate": 2.5607513776182616e-05, | |
| "loss": 0.34, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.348178137651822, | |
| "grad_norm": 0.5034493741934669, | |
| "learning_rate": 2.5553205480995633e-05, | |
| "loss": 0.3655, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3522267206477734, | |
| "grad_norm": 0.4869121939466513, | |
| "learning_rate": 2.5498852782525604e-05, | |
| "loss": 0.3645, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.3562753036437245, | |
| "grad_norm": 0.4958200914926886, | |
| "learning_rate": 2.5444456115375252e-05, | |
| "loss": 0.3552, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.360323886639676, | |
| "grad_norm": 0.4976374594766584, | |
| "learning_rate": 2.539001591449889e-05, | |
| "loss": 0.3783, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.3643724696356276, | |
| "grad_norm": 0.5580228046835354, | |
| "learning_rate": 2.53355326151989e-05, | |
| "loss": 0.3625, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.5538656902919011, | |
| "learning_rate": 2.5281006653122288e-05, | |
| "loss": 0.3827, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.3724696356275303, | |
| "grad_norm": 0.5557923393745527, | |
| "learning_rate": 2.5226438464257217e-05, | |
| "loss": 0.3824, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.376518218623482, | |
| "grad_norm": 0.5056834111660493, | |
| "learning_rate": 2.5171828484929447e-05, | |
| "loss": 0.3621, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.380566801619433, | |
| "grad_norm": 0.5383626583232584, | |
| "learning_rate": 2.5117177151798934e-05, | |
| "loss": 0.3666, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.47613470940743435, | |
| "learning_rate": 2.506248490185627e-05, | |
| "loss": 0.3751, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.388663967611336, | |
| "grad_norm": 0.4848384968995998, | |
| "learning_rate": 2.500775217241923e-05, | |
| "loss": 0.3758, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.3927125506072873, | |
| "grad_norm": 0.5546230134852332, | |
| "learning_rate": 2.4952979401129266e-05, | |
| "loss": 0.3653, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.396761133603239, | |
| "grad_norm": 0.4910344236359216, | |
| "learning_rate": 2.4898167025948e-05, | |
| "loss": 0.384, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.4008097165991904, | |
| "grad_norm": 0.5123251722277935, | |
| "learning_rate": 2.4843315485153703e-05, | |
| "loss": 0.3682, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.4048582995951415, | |
| "grad_norm": 0.5137792641420066, | |
| "learning_rate": 2.4788425217337856e-05, | |
| "loss": 0.3739, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.408906882591093, | |
| "grad_norm": 0.45802945724389804, | |
| "learning_rate": 2.4733496661401572e-05, | |
| "loss": 0.3487, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.4129554655870447, | |
| "grad_norm": 0.5190945859081991, | |
| "learning_rate": 2.4678530256552114e-05, | |
| "loss": 0.3808, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.417004048582996, | |
| "grad_norm": 0.5098564234705092, | |
| "learning_rate": 2.46235264422994e-05, | |
| "loss": 0.3751, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.478833431701644, | |
| "learning_rate": 2.4568485658452452e-05, | |
| "loss": 0.384, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.425101214574899, | |
| "grad_norm": 0.45984426782121, | |
| "learning_rate": 2.451340834511592e-05, | |
| "loss": 0.3607, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.42914979757085, | |
| "grad_norm": 0.445411008159572, | |
| "learning_rate": 2.4458294942686528e-05, | |
| "loss": 0.3296, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4331983805668016, | |
| "grad_norm": 0.4700225810887517, | |
| "learning_rate": 2.440314589184958e-05, | |
| "loss": 0.3675, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.437246963562753, | |
| "grad_norm": 0.49475720418209085, | |
| "learning_rate": 2.434796163357541e-05, | |
| "loss": 0.3658, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.4412955465587043, | |
| "grad_norm": 0.4935815752484955, | |
| "learning_rate": 2.429274260911588e-05, | |
| "loss": 0.3701, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.445344129554656, | |
| "grad_norm": 0.476617707410897, | |
| "learning_rate": 2.423748926000084e-05, | |
| "loss": 0.3804, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.4493927125506074, | |
| "grad_norm": 0.5313510412817136, | |
| "learning_rate": 2.418220202803459e-05, | |
| "loss": 0.3797, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.4534412955465585, | |
| "grad_norm": 0.4935871892468845, | |
| "learning_rate": 2.412688135529237e-05, | |
| "loss": 0.3684, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.45748987854251, | |
| "grad_norm": 0.4896732327920852, | |
| "learning_rate": 2.4071527684116807e-05, | |
| "loss": 0.3602, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.5137832697222736, | |
| "learning_rate": 2.4016141457114372e-05, | |
| "loss": 0.3843, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.465587044534413, | |
| "grad_norm": 0.5059205753777065, | |
| "learning_rate": 2.396072311715187e-05, | |
| "loss": 0.3712, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.4696356275303644, | |
| "grad_norm": 0.46364485630406865, | |
| "learning_rate": 2.3905273107352863e-05, | |
| "loss": 0.3537, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.507775018880395, | |
| "learning_rate": 2.3849791871094157e-05, | |
| "loss": 0.3743, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.477732793522267, | |
| "grad_norm": 0.5091170979740561, | |
| "learning_rate": 2.379427985200223e-05, | |
| "loss": 0.3629, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.4817813765182186, | |
| "grad_norm": 0.4973624119510038, | |
| "learning_rate": 2.373873749394972e-05, | |
| "loss": 0.374, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.48582995951417, | |
| "grad_norm": 0.5127316465904181, | |
| "learning_rate": 2.3683165241051836e-05, | |
| "loss": 0.3598, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.4898785425101213, | |
| "grad_norm": 0.45937266688558426, | |
| "learning_rate": 2.3627563537662823e-05, | |
| "loss": 0.3631, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.493927125506073, | |
| "grad_norm": 0.5853928580271655, | |
| "learning_rate": 2.3571932828372433e-05, | |
| "loss": 0.376, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.4979757085020244, | |
| "grad_norm": 0.4675659283027485, | |
| "learning_rate": 2.351627355800233e-05, | |
| "loss": 0.3535, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.5020242914979756, | |
| "grad_norm": 0.5403792064431929, | |
| "learning_rate": 2.346058617160256e-05, | |
| "loss": 0.3791, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.506072874493927, | |
| "grad_norm": 0.5437503394329618, | |
| "learning_rate": 2.3404871114447976e-05, | |
| "loss": 0.3967, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.5101214574898787, | |
| "grad_norm": 0.5299862945816842, | |
| "learning_rate": 2.334912883203469e-05, | |
| "loss": 0.3665, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.51417004048583, | |
| "grad_norm": 0.46326743141040005, | |
| "learning_rate": 2.3293359770076505e-05, | |
| "loss": 0.3533, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.5182186234817814, | |
| "grad_norm": 0.5355681521244215, | |
| "learning_rate": 2.323756437450136e-05, | |
| "loss": 0.3558, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.522267206477733, | |
| "grad_norm": 0.4883829575378139, | |
| "learning_rate": 2.3181743091447754e-05, | |
| "loss": 0.3651, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.47193089610327993, | |
| "learning_rate": 2.3125896367261164e-05, | |
| "loss": 0.3769, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5303643724696356, | |
| "grad_norm": 0.47434930697899425, | |
| "learning_rate": 2.307002464849052e-05, | |
| "loss": 0.3519, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.534412955465587, | |
| "grad_norm": 0.4554578145171081, | |
| "learning_rate": 2.30141283818846e-05, | |
| "loss": 0.3698, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.5254669934336076, | |
| "learning_rate": 2.2958208014388462e-05, | |
| "loss": 0.3842, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.54251012145749, | |
| "grad_norm": 0.48773062933099026, | |
| "learning_rate": 2.2902263993139857e-05, | |
| "loss": 0.3707, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.5465587044534415, | |
| "grad_norm": 0.4696856107926319, | |
| "learning_rate": 2.2846296765465708e-05, | |
| "loss": 0.3544, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.5506072874493926, | |
| "grad_norm": 0.4369181486890408, | |
| "learning_rate": 2.2790306778878465e-05, | |
| "loss": 0.3527, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.554655870445344, | |
| "grad_norm": 0.44243121466450236, | |
| "learning_rate": 2.273429448107257e-05, | |
| "loss": 0.352, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.5587044534412957, | |
| "grad_norm": 0.4763765673857041, | |
| "learning_rate": 2.2678260319920868e-05, | |
| "loss": 0.3526, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.562753036437247, | |
| "grad_norm": 0.4784626692095596, | |
| "learning_rate": 2.262220474347102e-05, | |
| "loss": 0.3639, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.5668016194331984, | |
| "grad_norm": 0.4947967115725402, | |
| "learning_rate": 2.2566128199941918e-05, | |
| "loss": 0.36, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.57085020242915, | |
| "grad_norm": 0.46287039632103183, | |
| "learning_rate": 2.2510031137720104e-05, | |
| "loss": 0.3778, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.574898785425101, | |
| "grad_norm": 0.5147668193914432, | |
| "learning_rate": 2.2453914005356206e-05, | |
| "loss": 0.3669, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.4984398736163127, | |
| "learning_rate": 2.2397777251561302e-05, | |
| "loss": 0.3844, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.582995951417004, | |
| "grad_norm": 0.4497335441559265, | |
| "learning_rate": 2.2341621325203393e-05, | |
| "loss": 0.3578, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.5870445344129553, | |
| "grad_norm": 0.4721594853721678, | |
| "learning_rate": 2.2285446675303764e-05, | |
| "loss": 0.3761, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.591093117408907, | |
| "grad_norm": 0.49539240632177983, | |
| "learning_rate": 2.2229253751033423e-05, | |
| "loss": 0.38, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.5951417004048585, | |
| "grad_norm": 0.4779037348563006, | |
| "learning_rate": 2.2173043001709492e-05, | |
| "loss": 0.3886, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.5991902834008096, | |
| "grad_norm": 0.47345570219826594, | |
| "learning_rate": 2.2116814876791627e-05, | |
| "loss": 0.382, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.603238866396761, | |
| "grad_norm": 0.4784774053948373, | |
| "learning_rate": 2.2060569825878413e-05, | |
| "loss": 0.3693, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.6072874493927127, | |
| "grad_norm": 0.45127955664189395, | |
| "learning_rate": 2.2004308298703777e-05, | |
| "loss": 0.3487, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.611336032388664, | |
| "grad_norm": 0.525540901549924, | |
| "learning_rate": 2.1948030745133396e-05, | |
| "loss": 0.3878, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.4864447120884368, | |
| "learning_rate": 2.1891737615161083e-05, | |
| "loss": 0.3804, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.619433198380567, | |
| "grad_norm": 0.49990679076530375, | |
| "learning_rate": 2.1835429358905218e-05, | |
| "loss": 0.3565, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.623481781376518, | |
| "grad_norm": 0.47704643941950026, | |
| "learning_rate": 2.17791064266051e-05, | |
| "loss": 0.3699, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.6275303643724697, | |
| "grad_norm": 0.5201751130093851, | |
| "learning_rate": 2.172276926861741e-05, | |
| "loss": 0.3467, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.44353614328366797, | |
| "learning_rate": 2.166641833541255e-05, | |
| "loss": 0.3402, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6356275303643724, | |
| "grad_norm": 0.5097553277331798, | |
| "learning_rate": 2.161005407757109e-05, | |
| "loss": 0.3744, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.639676113360324, | |
| "grad_norm": 0.5418726453048757, | |
| "learning_rate": 2.155367694578013e-05, | |
| "loss": 0.3771, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.6437246963562755, | |
| "grad_norm": 0.5075195127587114, | |
| "learning_rate": 2.1497287390829712e-05, | |
| "loss": 0.3797, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.6477732793522266, | |
| "grad_norm": 0.506357205881397, | |
| "learning_rate": 2.1440885863609208e-05, | |
| "loss": 0.3559, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.651821862348178, | |
| "grad_norm": 0.4797313270337654, | |
| "learning_rate": 2.1384472815103737e-05, | |
| "loss": 0.3558, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6558704453441297, | |
| "grad_norm": 0.47552454630286345, | |
| "learning_rate": 2.1328048696390534e-05, | |
| "loss": 0.3646, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.659919028340081, | |
| "grad_norm": 0.46303529049973463, | |
| "learning_rate": 2.1271613958635342e-05, | |
| "loss": 0.3617, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.6639676113360324, | |
| "grad_norm": 0.46875948908119347, | |
| "learning_rate": 2.1215169053088825e-05, | |
| "loss": 0.3478, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.668016194331984, | |
| "grad_norm": 0.47698936398504876, | |
| "learning_rate": 2.1158714431082937e-05, | |
| "loss": 0.3695, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.672064777327935, | |
| "grad_norm": 0.46417313689736034, | |
| "learning_rate": 2.110225054402733e-05, | |
| "loss": 0.3815, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.6761133603238867, | |
| "grad_norm": 0.48317477956061416, | |
| "learning_rate": 2.1045777843405747e-05, | |
| "loss": 0.3743, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.6801619433198383, | |
| "grad_norm": 0.4687566391750218, | |
| "learning_rate": 2.09892967807724e-05, | |
| "loss": 0.3724, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.4630499433513492, | |
| "learning_rate": 2.093280780774835e-05, | |
| "loss": 0.3589, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.688259109311741, | |
| "grad_norm": 0.4759409727984925, | |
| "learning_rate": 2.087631137601793e-05, | |
| "loss": 0.3904, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.4620366427791594, | |
| "learning_rate": 2.0819807937325104e-05, | |
| "loss": 0.3636, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.6963562753036436, | |
| "grad_norm": 0.4438215631519308, | |
| "learning_rate": 2.0763297943469847e-05, | |
| "loss": 0.3647, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.700404858299595, | |
| "grad_norm": 0.4842978195404141, | |
| "learning_rate": 2.070678184630458e-05, | |
| "loss": 0.3532, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.7044534412955468, | |
| "grad_norm": 0.4862805815601221, | |
| "learning_rate": 2.06502600977305e-05, | |
| "loss": 0.3799, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.708502024291498, | |
| "grad_norm": 0.46092937753023233, | |
| "learning_rate": 2.0593733149694008e-05, | |
| "loss": 0.3321, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.7125506072874495, | |
| "grad_norm": 0.4810766400328739, | |
| "learning_rate": 2.0537201454183074e-05, | |
| "loss": 0.3496, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.716599190283401, | |
| "grad_norm": 0.4589365203280722, | |
| "learning_rate": 2.0480665463223626e-05, | |
| "loss": 0.3784, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.720647773279352, | |
| "grad_norm": 0.4851737703568432, | |
| "learning_rate": 2.042412562887594e-05, | |
| "loss": 0.3269, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7246963562753037, | |
| "grad_norm": 0.4569976796244981, | |
| "learning_rate": 2.0367582403231033e-05, | |
| "loss": 0.3699, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.7287449392712553, | |
| "grad_norm": 0.4463151943613483, | |
| "learning_rate": 2.0311036238407023e-05, | |
| "loss": 0.3639, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7327935222672064, | |
| "grad_norm": 0.5037642597350784, | |
| "learning_rate": 2.0254487586545534e-05, | |
| "loss": 0.3713, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.4347045110070311, | |
| "learning_rate": 2.0197936899808088e-05, | |
| "loss": 0.341, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.7408906882591095, | |
| "grad_norm": 0.45968476803779285, | |
| "learning_rate": 2.0141384630372466e-05, | |
| "loss": 0.3643, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7449392712550607, | |
| "grad_norm": 0.49865079157736075, | |
| "learning_rate": 2.0084831230429106e-05, | |
| "loss": 0.3743, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.748987854251012, | |
| "grad_norm": 0.47399222173629535, | |
| "learning_rate": 2.0028277152177492e-05, | |
| "loss": 0.3828, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.753036437246964, | |
| "grad_norm": 0.4606581929107441, | |
| "learning_rate": 1.9971722847822518e-05, | |
| "loss": 0.3734, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.757085020242915, | |
| "grad_norm": 0.47862895502103103, | |
| "learning_rate": 1.99151687695709e-05, | |
| "loss": 0.3386, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7611336032388665, | |
| "grad_norm": 0.4912356998943867, | |
| "learning_rate": 1.9858615369627537e-05, | |
| "loss": 0.3783, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.765182186234818, | |
| "grad_norm": 0.4816789756678036, | |
| "learning_rate": 1.9802063100191916e-05, | |
| "loss": 0.3529, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.4255647613182946, | |
| "learning_rate": 1.974551241345447e-05, | |
| "loss": 0.3661, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.7732793522267207, | |
| "grad_norm": 0.498514879441169, | |
| "learning_rate": 1.9688963761592984e-05, | |
| "loss": 0.3584, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7773279352226723, | |
| "grad_norm": 0.45555135717723255, | |
| "learning_rate": 1.963241759676897e-05, | |
| "loss": 0.3613, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.7813765182186234, | |
| "grad_norm": 0.47723069951142516, | |
| "learning_rate": 1.9575874371124066e-05, | |
| "loss": 0.3478, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.785425101214575, | |
| "grad_norm": 0.4720371996366464, | |
| "learning_rate": 1.951933453677638e-05, | |
| "loss": 0.3856, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.4566915477305463, | |
| "learning_rate": 1.9462798545816932e-05, | |
| "loss": 0.3648, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.7935222672064777, | |
| "grad_norm": 0.45903458608715303, | |
| "learning_rate": 1.9406266850305996e-05, | |
| "loss": 0.3533, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7975708502024292, | |
| "grad_norm": 0.45378578290395266, | |
| "learning_rate": 1.9349739902269503e-05, | |
| "loss": 0.3713, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.801619433198381, | |
| "grad_norm": 0.48290989035550985, | |
| "learning_rate": 1.9293218153695425e-05, | |
| "loss": 0.379, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.805668016194332, | |
| "grad_norm": 0.45363621797098574, | |
| "learning_rate": 1.9236702056530153e-05, | |
| "loss": 0.3519, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.8097165991902835, | |
| "grad_norm": 0.4670351930315145, | |
| "learning_rate": 1.9180192062674907e-05, | |
| "loss": 0.3774, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.813765182186235, | |
| "grad_norm": 0.518862060717548, | |
| "learning_rate": 1.9123688623982076e-05, | |
| "loss": 0.3669, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.817813765182186, | |
| "grad_norm": 0.4473220289066495, | |
| "learning_rate": 1.906719219225165e-05, | |
| "loss": 0.36, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.8218623481781377, | |
| "grad_norm": 0.5561993267876191, | |
| "learning_rate": 1.9010703219227603e-05, | |
| "loss": 0.4036, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.8259109311740893, | |
| "grad_norm": 0.46674539209818405, | |
| "learning_rate": 1.8954222156594256e-05, | |
| "loss": 0.3441, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.8299595141700404, | |
| "grad_norm": 0.5168979210196667, | |
| "learning_rate": 1.8897749455972673e-05, | |
| "loss": 0.3642, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.834008097165992, | |
| "grad_norm": 0.4681670411000131, | |
| "learning_rate": 1.884128556891707e-05, | |
| "loss": 0.3788, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8380566801619436, | |
| "grad_norm": 0.5341284231599533, | |
| "learning_rate": 1.878483094691119e-05, | |
| "loss": 0.3802, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.4703669748785538, | |
| "learning_rate": 1.8728386041364664e-05, | |
| "loss": 0.3481, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.459617483034287, | |
| "learning_rate": 1.867195130360947e-05, | |
| "loss": 0.369, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.850202429149798, | |
| "grad_norm": 0.5074356590193903, | |
| "learning_rate": 1.8615527184896263e-05, | |
| "loss": 0.3733, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.854251012145749, | |
| "grad_norm": 0.4449380609262646, | |
| "learning_rate": 1.8559114136390795e-05, | |
| "loss": 0.3736, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8582995951417005, | |
| "grad_norm": 0.467611807280845, | |
| "learning_rate": 1.8502712609170298e-05, | |
| "loss": 0.3482, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.862348178137652, | |
| "grad_norm": 0.48819532942448096, | |
| "learning_rate": 1.8446323054219876e-05, | |
| "loss": 0.3568, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.866396761133603, | |
| "grad_norm": 0.46287839889868143, | |
| "learning_rate": 1.838994592242891e-05, | |
| "loss": 0.3615, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.8704453441295548, | |
| "grad_norm": 0.45940615314448713, | |
| "learning_rate": 1.8333581664587453e-05, | |
| "loss": 0.3718, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.8744939271255063, | |
| "grad_norm": 0.4462048692650306, | |
| "learning_rate": 1.8277230731382593e-05, | |
| "loss": 0.3698, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8785425101214575, | |
| "grad_norm": 0.46584083304668, | |
| "learning_rate": 1.82208935733949e-05, | |
| "loss": 0.3804, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.882591093117409, | |
| "grad_norm": 0.43869390779432016, | |
| "learning_rate": 1.8164570641094793e-05, | |
| "loss": 0.3889, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.8866396761133606, | |
| "grad_norm": 0.44141596449943316, | |
| "learning_rate": 1.810826238483892e-05, | |
| "loss": 0.3549, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.8906882591093117, | |
| "grad_norm": 0.4826977813463676, | |
| "learning_rate": 1.8051969254866608e-05, | |
| "loss": 0.379, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.4883768016797113, | |
| "learning_rate": 1.7995691701296226e-05, | |
| "loss": 0.3465, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.898785425101215, | |
| "grad_norm": 0.4866016059377446, | |
| "learning_rate": 1.7939430174121593e-05, | |
| "loss": 0.3724, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.902834008097166, | |
| "grad_norm": 0.48415773828353037, | |
| "learning_rate": 1.788318512320838e-05, | |
| "loss": 0.3621, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.9068825910931175, | |
| "grad_norm": 0.44183216054011965, | |
| "learning_rate": 1.782695699829051e-05, | |
| "loss": 0.3584, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.910931174089069, | |
| "grad_norm": 0.4348189875167545, | |
| "learning_rate": 1.7770746248966587e-05, | |
| "loss": 0.3591, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.91497975708502, | |
| "grad_norm": 0.5166371634244631, | |
| "learning_rate": 1.7714553324696243e-05, | |
| "loss": 0.3668, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.919028340080972, | |
| "grad_norm": 0.4113627183208117, | |
| "learning_rate": 1.7658378674796614e-05, | |
| "loss": 0.3599, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.45569840539577927, | |
| "learning_rate": 1.76022227484387e-05, | |
| "loss": 0.3633, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.9271255060728745, | |
| "grad_norm": 0.49500603537396376, | |
| "learning_rate": 1.7546085994643807e-05, | |
| "loss": 0.3534, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.931174089068826, | |
| "grad_norm": 0.46974049155702424, | |
| "learning_rate": 1.7489968862279902e-05, | |
| "loss": 0.3608, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9352226720647776, | |
| "grad_norm": 0.4312885080463309, | |
| "learning_rate": 1.743387180005809e-05, | |
| "loss": 0.3538, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.9392712550607287, | |
| "grad_norm": 0.4577879787586369, | |
| "learning_rate": 1.737779525652899e-05, | |
| "loss": 0.3576, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.9433198380566803, | |
| "grad_norm": 0.48332579943184484, | |
| "learning_rate": 1.732173968007914e-05, | |
| "loss": 0.3757, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.41833935648923554, | |
| "learning_rate": 1.7265705518927437e-05, | |
| "loss": 0.3535, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.951417004048583, | |
| "grad_norm": 0.46165561120659115, | |
| "learning_rate": 1.7209693221121542e-05, | |
| "loss": 0.3596, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.9554655870445345, | |
| "grad_norm": 0.4697727029608329, | |
| "learning_rate": 1.7153703234534302e-05, | |
| "loss": 0.352, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9595141700404857, | |
| "grad_norm": 0.4438532263698274, | |
| "learning_rate": 1.709773600686015e-05, | |
| "loss": 0.3775, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.9635627530364372, | |
| "grad_norm": 0.4759182542285938, | |
| "learning_rate": 1.7041791985611548e-05, | |
| "loss": 0.3725, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.967611336032389, | |
| "grad_norm": 0.43630884492431266, | |
| "learning_rate": 1.69858716181154e-05, | |
| "loss": 0.3767, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.97165991902834, | |
| "grad_norm": 0.44330418200855953, | |
| "learning_rate": 1.692997535150948e-05, | |
| "loss": 0.3427, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.9757085020242915, | |
| "grad_norm": 0.43192238897582746, | |
| "learning_rate": 1.687410363273884e-05, | |
| "loss": 0.3702, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.979757085020243, | |
| "grad_norm": 0.4260342647462933, | |
| "learning_rate": 1.6818256908552257e-05, | |
| "loss": 0.3501, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.983805668016194, | |
| "grad_norm": 0.4487315797229072, | |
| "learning_rate": 1.6762435625498646e-05, | |
| "loss": 0.3663, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.9878542510121457, | |
| "grad_norm": 0.43491389940842035, | |
| "learning_rate": 1.67066402299235e-05, | |
| "loss": 0.3629, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.9919028340080973, | |
| "grad_norm": 0.442345557795626, | |
| "learning_rate": 1.6650871167965313e-05, | |
| "loss": 0.3752, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.9959514170040484, | |
| "grad_norm": 0.461419483682596, | |
| "learning_rate": 1.6595128885552028e-05, | |
| "loss": 0.3524, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.45805831060291474, | |
| "learning_rate": 1.653941382839745e-05, | |
| "loss": 0.33, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.0040485829959516, | |
| "grad_norm": 0.6078204170535164, | |
| "learning_rate": 1.6483726441997673e-05, | |
| "loss": 0.2184, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.0080971659919027, | |
| "grad_norm": 0.46934289668618584, | |
| "learning_rate": 1.642806717162757e-05, | |
| "loss": 0.2246, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.0121457489878543, | |
| "grad_norm": 0.4693782842939853, | |
| "learning_rate": 1.637243646233718e-05, | |
| "loss": 0.2189, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.016194331983806, | |
| "grad_norm": 0.6428739483190475, | |
| "learning_rate": 1.6316834758948174e-05, | |
| "loss": 0.2264, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.020242914979757, | |
| "grad_norm": 0.6369857214113686, | |
| "learning_rate": 1.6261262506050282e-05, | |
| "loss": 0.2163, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.0242914979757085, | |
| "grad_norm": 0.49500912380027257, | |
| "learning_rate": 1.620572014799777e-05, | |
| "loss": 0.2292, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.02834008097166, | |
| "grad_norm": 0.5963215238951878, | |
| "learning_rate": 1.6150208128905857e-05, | |
| "loss": 0.2139, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.032388663967611, | |
| "grad_norm": 0.5297504858502328, | |
| "learning_rate": 1.6094726892647147e-05, | |
| "loss": 0.2286, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0364372469635628, | |
| "grad_norm": 0.46281273983013693, | |
| "learning_rate": 1.6039276882848135e-05, | |
| "loss": 0.21, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0404858299595143, | |
| "grad_norm": 0.4375938001194552, | |
| "learning_rate": 1.5983858542885635e-05, | |
| "loss": 0.2146, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.0445344129554655, | |
| "grad_norm": 0.46919326801037764, | |
| "learning_rate": 1.5928472315883203e-05, | |
| "loss": 0.2286, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.048582995951417, | |
| "grad_norm": 0.47648558204107677, | |
| "learning_rate": 1.5873118644707633e-05, | |
| "loss": 0.2181, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.4681386221087982, | |
| "learning_rate": 1.5817797971965413e-05, | |
| "loss": 0.204, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.0566801619433197, | |
| "grad_norm": 0.4343793732678376, | |
| "learning_rate": 1.576251073999917e-05, | |
| "loss": 0.215, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.0607287449392713, | |
| "grad_norm": 0.46107041797132425, | |
| "learning_rate": 1.5707257390884126e-05, | |
| "loss": 0.2195, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.064777327935223, | |
| "grad_norm": 0.42724952561749285, | |
| "learning_rate": 1.5652038366424595e-05, | |
| "loss": 0.2268, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.068825910931174, | |
| "grad_norm": 0.43454889460125307, | |
| "learning_rate": 1.5596854108150424e-05, | |
| "loss": 0.2158, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.0728744939271255, | |
| "grad_norm": 0.4546736122709604, | |
| "learning_rate": 1.5541705057313476e-05, | |
| "loss": 0.2077, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.4090435330600373, | |
| "learning_rate": 1.5486591654884086e-05, | |
| "loss": 0.2311, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.080971659919028, | |
| "grad_norm": 0.41571973884867947, | |
| "learning_rate": 1.543151434154755e-05, | |
| "loss": 0.2126, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.08502024291498, | |
| "grad_norm": 0.4773412003480402, | |
| "learning_rate": 1.5376473557700615e-05, | |
| "loss": 0.2219, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.0890688259109313, | |
| "grad_norm": 0.44538045969894535, | |
| "learning_rate": 1.5321469743447893e-05, | |
| "loss": 0.201, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.0931174089068825, | |
| "grad_norm": 0.4358257869787928, | |
| "learning_rate": 1.5266503338598434e-05, | |
| "loss": 0.2166, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.097165991902834, | |
| "grad_norm": 0.4191469763154963, | |
| "learning_rate": 1.5211574782662149e-05, | |
| "loss": 0.221, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1012145748987856, | |
| "grad_norm": 0.4222529450477182, | |
| "learning_rate": 1.5156684514846307e-05, | |
| "loss": 0.2088, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.44173976913198487, | |
| "learning_rate": 1.5101832974052012e-05, | |
| "loss": 0.193, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.1093117408906883, | |
| "grad_norm": 0.43301455607251826, | |
| "learning_rate": 1.5047020598870737e-05, | |
| "loss": 0.2132, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.11336032388664, | |
| "grad_norm": 0.42933437170665845, | |
| "learning_rate": 1.4992247827580778e-05, | |
| "loss": 0.2173, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.117408906882591, | |
| "grad_norm": 0.42621495416091487, | |
| "learning_rate": 1.493751509814374e-05, | |
| "loss": 0.2127, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1214574898785425, | |
| "grad_norm": 0.4248879025549845, | |
| "learning_rate": 1.4882822848201073e-05, | |
| "loss": 0.2095, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.125506072874494, | |
| "grad_norm": 0.43068684620097636, | |
| "learning_rate": 1.4828171515070553e-05, | |
| "loss": 0.2044, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.1295546558704452, | |
| "grad_norm": 0.4191957931765077, | |
| "learning_rate": 1.4773561535742793e-05, | |
| "loss": 0.2153, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.133603238866397, | |
| "grad_norm": 0.42860429876077144, | |
| "learning_rate": 1.4718993346877715e-05, | |
| "loss": 0.2022, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.1376518218623484, | |
| "grad_norm": 0.4191859208574585, | |
| "learning_rate": 1.466446738480111e-05, | |
| "loss": 0.2479, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1417004048582995, | |
| "grad_norm": 0.4124678681714068, | |
| "learning_rate": 1.460998408550112e-05, | |
| "loss": 0.2168, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.145748987854251, | |
| "grad_norm": 0.4282239736027295, | |
| "learning_rate": 1.4555543884624751e-05, | |
| "loss": 0.2067, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.1497975708502026, | |
| "grad_norm": 0.41496411990220367, | |
| "learning_rate": 1.4501147217474402e-05, | |
| "loss": 0.235, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.40380861278833335, | |
| "learning_rate": 1.444679451900437e-05, | |
| "loss": 0.2011, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.42027945378656656, | |
| "learning_rate": 1.4392486223817397e-05, | |
| "loss": 0.2233, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.161943319838057, | |
| "grad_norm": 0.41322428766664276, | |
| "learning_rate": 1.4338222766161143e-05, | |
| "loss": 0.2136, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.165991902834008, | |
| "grad_norm": 0.4173547195487682, | |
| "learning_rate": 1.4284004579924777e-05, | |
| "loss": 0.2137, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.1700404858299596, | |
| "grad_norm": 0.42242733336000476, | |
| "learning_rate": 1.4229832098635479e-05, | |
| "loss": 0.2042, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.174089068825911, | |
| "grad_norm": 0.4362369220473824, | |
| "learning_rate": 1.4175705755454963e-05, | |
| "loss": 0.2195, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.1781376518218623, | |
| "grad_norm": 0.42619076351290813, | |
| "learning_rate": 1.412162598317602e-05, | |
| "loss": 0.2065, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.182186234817814, | |
| "grad_norm": 0.4255688971604276, | |
| "learning_rate": 1.406759321421907e-05, | |
| "loss": 0.2018, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.1862348178137654, | |
| "grad_norm": 0.4228051628813469, | |
| "learning_rate": 1.401360788062871e-05, | |
| "loss": 0.2289, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.1902834008097165, | |
| "grad_norm": 0.4018272215987163, | |
| "learning_rate": 1.3959670414070204e-05, | |
| "loss": 0.2118, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.194331983805668, | |
| "grad_norm": 0.4040885331182461, | |
| "learning_rate": 1.3905781245826108e-05, | |
| "loss": 0.221, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.1983805668016196, | |
| "grad_norm": 0.4252662391204368, | |
| "learning_rate": 1.3851940806792778e-05, | |
| "loss": 0.2184, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2024291497975708, | |
| "grad_norm": 0.41431024010481304, | |
| "learning_rate": 1.379814952747693e-05, | |
| "loss": 0.2095, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.2064777327935223, | |
| "grad_norm": 0.42918838387744757, | |
| "learning_rate": 1.3744407837992193e-05, | |
| "loss": 0.2012, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.41260937267873893, | |
| "learning_rate": 1.3690716168055692e-05, | |
| "loss": 0.2154, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.214574898785425, | |
| "grad_norm": 0.41226117626044634, | |
| "learning_rate": 1.363707494698459e-05, | |
| "loss": 0.1853, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.2186234817813766, | |
| "grad_norm": 0.4245373396360888, | |
| "learning_rate": 1.358348460369265e-05, | |
| "loss": 0.1839, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.2226720647773277, | |
| "grad_norm": 0.39187452353499397, | |
| "learning_rate": 1.352994556668684e-05, | |
| "loss": 0.2182, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.2267206477732793, | |
| "grad_norm": 0.38806667886101104, | |
| "learning_rate": 1.3476458264063865e-05, | |
| "loss": 0.2182, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.40157988166104563, | |
| "learning_rate": 1.342302312350678e-05, | |
| "loss": 0.2564, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.234817813765182, | |
| "grad_norm": 0.42986730159999553, | |
| "learning_rate": 1.3369640572281537e-05, | |
| "loss": 0.2158, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.2388663967611335, | |
| "grad_norm": 0.3913842654813449, | |
| "learning_rate": 1.3316311037233596e-05, | |
| "loss": 0.2067, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.242914979757085, | |
| "grad_norm": 0.4363957911244246, | |
| "learning_rate": 1.326303494478451e-05, | |
| "loss": 0.2239, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.246963562753036, | |
| "grad_norm": 0.44180189846881707, | |
| "learning_rate": 1.3209812720928472e-05, | |
| "loss": 0.2217, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.251012145748988, | |
| "grad_norm": 0.3895070031054075, | |
| "learning_rate": 1.315664479122898e-05, | |
| "loss": 0.2195, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.2550607287449393, | |
| "grad_norm": 0.40814235160112056, | |
| "learning_rate": 1.3103531580815378e-05, | |
| "loss": 0.2075, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.2591093117408905, | |
| "grad_norm": 0.42705958063637794, | |
| "learning_rate": 1.3050473514379488e-05, | |
| "loss": 0.2315, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.413863415912222, | |
| "learning_rate": 1.2997471016172188e-05, | |
| "loss": 0.2108, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.2672064777327936, | |
| "grad_norm": 0.4279213758331502, | |
| "learning_rate": 1.2944524510000042e-05, | |
| "loss": 0.1986, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.2712550607287447, | |
| "grad_norm": 0.4271380014279438, | |
| "learning_rate": 1.2891634419221908e-05, | |
| "loss": 0.2172, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.2753036437246963, | |
| "grad_norm": 0.4268110656682208, | |
| "learning_rate": 1.2838801166745545e-05, | |
| "loss": 0.2106, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.279352226720648, | |
| "grad_norm": 0.41223054333058035, | |
| "learning_rate": 1.2786025175024216e-05, | |
| "loss": 0.2289, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.283400809716599, | |
| "grad_norm": 0.4343726457568752, | |
| "learning_rate": 1.2733306866053357e-05, | |
| "loss": 0.2142, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.2874493927125505, | |
| "grad_norm": 0.43275673008462634, | |
| "learning_rate": 1.2680646661367163e-05, | |
| "loss": 0.2036, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.291497975708502, | |
| "grad_norm": 0.4298906643206618, | |
| "learning_rate": 1.2628044982035219e-05, | |
| "loss": 0.2172, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.2955465587044532, | |
| "grad_norm": 0.42751147268494744, | |
| "learning_rate": 1.2575502248659155e-05, | |
| "loss": 0.2168, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.299595141700405, | |
| "grad_norm": 0.4066541789012753, | |
| "learning_rate": 1.2523018881369269e-05, | |
| "loss": 0.2077, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.3036437246963564, | |
| "grad_norm": 0.45112134940345217, | |
| "learning_rate": 1.2470595299821176e-05, | |
| "loss": 0.2053, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.3989067385523775, | |
| "learning_rate": 1.2418231923192416e-05, | |
| "loss": 0.2076, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.311740890688259, | |
| "grad_norm": 0.42474116333024264, | |
| "learning_rate": 1.2365929170179165e-05, | |
| "loss": 0.2014, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.428409701492181, | |
| "learning_rate": 1.2313687458992848e-05, | |
| "loss": 0.2047, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.3198380566801617, | |
| "grad_norm": 0.3990637023075555, | |
| "learning_rate": 1.2261507207356785e-05, | |
| "loss": 0.2032, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.3238866396761133, | |
| "grad_norm": 0.4298702123022994, | |
| "learning_rate": 1.220938883250288e-05, | |
| "loss": 0.2042, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.327935222672065, | |
| "grad_norm": 0.4107894954807658, | |
| "learning_rate": 1.215733275116827e-05, | |
| "loss": 0.2094, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.331983805668016, | |
| "grad_norm": 0.41640881058426177, | |
| "learning_rate": 1.2105339379592005e-05, | |
| "loss": 0.205, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.3360323886639676, | |
| "grad_norm": 0.3933958307446595, | |
| "learning_rate": 1.2053409133511681e-05, | |
| "loss": 0.2182, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.340080971659919, | |
| "grad_norm": 0.43400741430815254, | |
| "learning_rate": 1.2001542428160173e-05, | |
| "loss": 0.2138, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.3441295546558703, | |
| "grad_norm": 0.44873712842577845, | |
| "learning_rate": 1.1949739678262278e-05, | |
| "loss": 0.2221, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.348178137651822, | |
| "grad_norm": 0.4072837132292443, | |
| "learning_rate": 1.189800129803139e-05, | |
| "loss": 0.2057, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.3522267206477734, | |
| "grad_norm": 0.4131878060714796, | |
| "learning_rate": 1.1846327701166228e-05, | |
| "loss": 0.2243, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.3562753036437245, | |
| "grad_norm": 0.40274996332502194, | |
| "learning_rate": 1.1794719300847493e-05, | |
| "loss": 0.2061, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.360323886639676, | |
| "grad_norm": 0.4275890184412437, | |
| "learning_rate": 1.1743176509734582e-05, | |
| "loss": 0.2395, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.3643724696356276, | |
| "grad_norm": 0.41425801635849097, | |
| "learning_rate": 1.1691699739962275e-05, | |
| "loss": 0.2366, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.41503119188734217, | |
| "learning_rate": 1.1640289403137456e-05, | |
| "loss": 0.2282, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.3724696356275303, | |
| "grad_norm": 0.42001880907345224, | |
| "learning_rate": 1.1588945910335802e-05, | |
| "loss": 0.2126, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.376518218623482, | |
| "grad_norm": 0.42516448924154376, | |
| "learning_rate": 1.1537669672098517e-05, | |
| "loss": 0.2227, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.380566801619433, | |
| "grad_norm": 0.45865406719620316, | |
| "learning_rate": 1.1486461098429036e-05, | |
| "loss": 0.2046, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.42210512304535824, | |
| "learning_rate": 1.1435320598789747e-05, | |
| "loss": 0.2182, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.388663967611336, | |
| "grad_norm": 0.4032405475465058, | |
| "learning_rate": 1.1384248582098736e-05, | |
| "loss": 0.2173, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.3927125506072873, | |
| "grad_norm": 0.41851861356552966, | |
| "learning_rate": 1.1333245456726467e-05, | |
| "loss": 0.1957, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.396761133603239, | |
| "grad_norm": 0.38620080497016623, | |
| "learning_rate": 1.128231163049258e-05, | |
| "loss": 0.2467, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.4008097165991904, | |
| "grad_norm": 0.37364777135875715, | |
| "learning_rate": 1.123144751066259e-05, | |
| "loss": 0.2122, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4048582995951415, | |
| "grad_norm": 0.3897965649688975, | |
| "learning_rate": 1.1180653503944657e-05, | |
| "loss": 0.22, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.408906882591093, | |
| "grad_norm": 0.4061296118435523, | |
| "learning_rate": 1.1129930016486287e-05, | |
| "loss": 0.2209, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.4129554655870447, | |
| "grad_norm": 0.4062153963308523, | |
| "learning_rate": 1.107927745387114e-05, | |
| "loss": 0.2054, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.417004048582996, | |
| "grad_norm": 0.4294399469150004, | |
| "learning_rate": 1.1028696221115766e-05, | |
| "loss": 0.2194, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.40152612907352764, | |
| "learning_rate": 1.0978186722666341e-05, | |
| "loss": 0.2108, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.425101214574899, | |
| "grad_norm": 0.43868970582381905, | |
| "learning_rate": 1.0927749362395475e-05, | |
| "loss": 0.2009, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.42914979757085, | |
| "grad_norm": 0.40819844572269987, | |
| "learning_rate": 1.0877384543598954e-05, | |
| "loss": 0.2161, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.4331983805668016, | |
| "grad_norm": 0.41659851956538735, | |
| "learning_rate": 1.0827092668992529e-05, | |
| "loss": 0.2221, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.437246963562753, | |
| "grad_norm": 0.42297326024008075, | |
| "learning_rate": 1.077687414070869e-05, | |
| "loss": 0.2019, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.4412955465587043, | |
| "grad_norm": 0.4046246351547333, | |
| "learning_rate": 1.0726729360293451e-05, | |
| "loss": 0.2313, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.445344129554656, | |
| "grad_norm": 0.4035839231958286, | |
| "learning_rate": 1.0676658728703147e-05, | |
| "loss": 0.2265, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.4493927125506074, | |
| "grad_norm": 0.38398868335504915, | |
| "learning_rate": 1.0626662646301198e-05, | |
| "loss": 0.2152, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.4534412955465585, | |
| "grad_norm": 0.3847381266208618, | |
| "learning_rate": 1.0576741512854958e-05, | |
| "loss": 0.2061, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.45748987854251, | |
| "grad_norm": 0.40950706158812056, | |
| "learning_rate": 1.052689572753247e-05, | |
| "loss": 0.2123, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.376549334660143, | |
| "learning_rate": 1.0477125688899328e-05, | |
| "loss": 0.2058, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.465587044534413, | |
| "grad_norm": 0.39775857046746566, | |
| "learning_rate": 1.0427431794915406e-05, | |
| "loss": 0.1929, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.4696356275303644, | |
| "grad_norm": 0.4259035516736771, | |
| "learning_rate": 1.0377814442931769e-05, | |
| "loss": 0.1947, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.391111993682781, | |
| "learning_rate": 1.0328274029687439e-05, | |
| "loss": 0.1839, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.477732793522267, | |
| "grad_norm": 0.41380541400193177, | |
| "learning_rate": 1.0278810951306253e-05, | |
| "loss": 0.237, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.4817813765182186, | |
| "grad_norm": 0.40383122633037205, | |
| "learning_rate": 1.0229425603293647e-05, | |
| "loss": 0.219, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.48582995951417, | |
| "grad_norm": 0.3983822993588476, | |
| "learning_rate": 1.0180118380533558e-05, | |
| "loss": 0.2092, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.4898785425101213, | |
| "grad_norm": 0.42471858960830167, | |
| "learning_rate": 1.0130889677285225e-05, | |
| "loss": 0.2199, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.493927125506073, | |
| "grad_norm": 0.40011470084349127, | |
| "learning_rate": 1.008173988718005e-05, | |
| "loss": 0.2175, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.4979757085020244, | |
| "grad_norm": 0.4102858476587693, | |
| "learning_rate": 1.0032669403218439e-05, | |
| "loss": 0.217, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.5020242914979756, | |
| "grad_norm": 0.4311998560895837, | |
| "learning_rate": 9.983678617766677e-06, | |
| "loss": 0.2331, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.506072874493927, | |
| "grad_norm": 0.4182270470660038, | |
| "learning_rate": 9.934767922553788e-06, | |
| "loss": 0.2357, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.5101214574898787, | |
| "grad_norm": 0.4298759197478185, | |
| "learning_rate": 9.885937708668371e-06, | |
| "loss": 0.2049, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.51417004048583, | |
| "grad_norm": 0.3910627716077731, | |
| "learning_rate": 9.837188366555524e-06, | |
| "loss": 0.2085, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.5182186234817814, | |
| "grad_norm": 0.4159838864288812, | |
| "learning_rate": 9.788520286013691e-06, | |
| "loss": 0.2056, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.522267206477733, | |
| "grad_norm": 0.4099832846653073, | |
| "learning_rate": 9.739933856191537e-06, | |
| "loss": 0.225, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.3821537770060851, | |
| "learning_rate": 9.691429465584866e-06, | |
| "loss": 0.1953, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.5303643724696356, | |
| "grad_norm": 0.42154426523986294, | |
| "learning_rate": 9.643007502033492e-06, | |
| "loss": 0.2249, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.534412955465587, | |
| "grad_norm": 0.42051478292218697, | |
| "learning_rate": 9.594668352718152e-06, | |
| "loss": 0.2341, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.4151292414159816, | |
| "learning_rate": 9.546412404157383e-06, | |
| "loss": 0.204, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.54251012145749, | |
| "grad_norm": 0.3923854761729001, | |
| "learning_rate": 9.498240042204471e-06, | |
| "loss": 0.2118, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.5465587044534415, | |
| "grad_norm": 0.4134281403963182, | |
| "learning_rate": 9.450151652044339e-06, | |
| "loss": 0.2061, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.5506072874493926, | |
| "grad_norm": 0.4102576778426014, | |
| "learning_rate": 9.402147618190472e-06, | |
| "loss": 0.2041, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.554655870445344, | |
| "grad_norm": 0.4224771559754108, | |
| "learning_rate": 9.354228324481841e-06, | |
| "loss": 0.2131, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.5587044534412957, | |
| "grad_norm": 0.42110989929456527, | |
| "learning_rate": 9.306394154079852e-06, | |
| "loss": 0.2378, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.562753036437247, | |
| "grad_norm": 0.4441996652147314, | |
| "learning_rate": 9.258645489465254e-06, | |
| "loss": 0.2129, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.5668016194331984, | |
| "grad_norm": 0.4168804894543135, | |
| "learning_rate": 9.21098271243509e-06, | |
| "loss": 0.2087, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.57085020242915, | |
| "grad_norm": 0.40126742398861165, | |
| "learning_rate": 9.163406204099659e-06, | |
| "loss": 0.1989, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.574898785425101, | |
| "grad_norm": 0.40915977443700285, | |
| "learning_rate": 9.115916344879452e-06, | |
| "loss": 0.2397, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.39882793821230694, | |
| "learning_rate": 9.068513514502126e-06, | |
| "loss": 0.2149, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.582995951417004, | |
| "grad_norm": 0.39907987317925236, | |
| "learning_rate": 9.021198091999439e-06, | |
| "loss": 0.2291, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.5870445344129553, | |
| "grad_norm": 0.42105905571618074, | |
| "learning_rate": 8.973970455704251e-06, | |
| "loss": 0.1942, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.591093117408907, | |
| "grad_norm": 0.3918422971655026, | |
| "learning_rate": 8.926830983247495e-06, | |
| "loss": 0.204, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.5951417004048585, | |
| "grad_norm": 0.43072437953364706, | |
| "learning_rate": 8.879780051555116e-06, | |
| "loss": 0.2057, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.5991902834008096, | |
| "grad_norm": 0.4025009402824398, | |
| "learning_rate": 8.83281803684512e-06, | |
| "loss": 0.1971, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.603238866396761, | |
| "grad_norm": 0.4148328579728651, | |
| "learning_rate": 8.785945314624532e-06, | |
| "loss": 0.2145, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.6072874493927127, | |
| "grad_norm": 0.4307999917414275, | |
| "learning_rate": 8.739162259686385e-06, | |
| "loss": 0.2248, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.611336032388664, | |
| "grad_norm": 0.4131716239709606, | |
| "learning_rate": 8.692469246106734e-06, | |
| "loss": 0.2143, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.44040644631918363, | |
| "learning_rate": 8.645866647241677e-06, | |
| "loss": 0.2123, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.619433198380567, | |
| "grad_norm": 0.42440571433772256, | |
| "learning_rate": 8.599354835724348e-06, | |
| "loss": 0.206, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.623481781376518, | |
| "grad_norm": 0.41492708736579753, | |
| "learning_rate": 8.552934183461958e-06, | |
| "loss": 0.2034, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6275303643724697, | |
| "grad_norm": 0.4225697640930608, | |
| "learning_rate": 8.506605061632805e-06, | |
| "loss": 0.1948, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.42653112014037087, | |
| "learning_rate": 8.460367840683308e-06, | |
| "loss": 0.2053, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.6356275303643724, | |
| "grad_norm": 0.45218310297227643, | |
| "learning_rate": 8.414222890325065e-06, | |
| "loss": 0.2047, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.639676113360324, | |
| "grad_norm": 0.4464914995597135, | |
| "learning_rate": 8.368170579531855e-06, | |
| "loss": 0.1891, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.6437246963562755, | |
| "grad_norm": 0.4376841798173919, | |
| "learning_rate": 8.322211276536733e-06, | |
| "loss": 0.2251, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6477732793522266, | |
| "grad_norm": 0.39720902676744285, | |
| "learning_rate": 8.27634534882907e-06, | |
| "loss": 0.2288, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.651821862348178, | |
| "grad_norm": 0.4233778895683264, | |
| "learning_rate": 8.230573163151585e-06, | |
| "loss": 0.2159, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.6558704453441297, | |
| "grad_norm": 0.44409655442029505, | |
| "learning_rate": 8.184895085497466e-06, | |
| "loss": 0.2147, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.659919028340081, | |
| "grad_norm": 0.3761121034968118, | |
| "learning_rate": 8.139311481107401e-06, | |
| "loss": 0.2328, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.6639676113360324, | |
| "grad_norm": 0.4439733988973326, | |
| "learning_rate": 8.093822714466686e-06, | |
| "loss": 0.2122, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.668016194331984, | |
| "grad_norm": 0.43530234241105165, | |
| "learning_rate": 8.048429149302271e-06, | |
| "loss": 0.205, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.672064777327935, | |
| "grad_norm": 0.4069185815707309, | |
| "learning_rate": 8.003131148579902e-06, | |
| "loss": 0.218, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.6761133603238867, | |
| "grad_norm": 0.4400088261059275, | |
| "learning_rate": 7.957929074501182e-06, | |
| "loss": 0.2135, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.6801619433198383, | |
| "grad_norm": 0.3886458941576382, | |
| "learning_rate": 7.912823288500695e-06, | |
| "loss": 0.2045, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.4037379272669396, | |
| "learning_rate": 7.867814151243102e-06, | |
| "loss": 0.2134, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.688259109311741, | |
| "grad_norm": 0.4103056210503837, | |
| "learning_rate": 7.822902022620267e-06, | |
| "loss": 0.2018, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.4231463219265299, | |
| "learning_rate": 7.778087261748377e-06, | |
| "loss": 0.2074, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.6963562753036436, | |
| "grad_norm": 0.4181783112585586, | |
| "learning_rate": 7.733370226965056e-06, | |
| "loss": 0.2073, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.700404858299595, | |
| "grad_norm": 0.39974979963701723, | |
| "learning_rate": 7.688751275826527e-06, | |
| "loss": 0.224, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.7044534412955468, | |
| "grad_norm": 0.3978220892629539, | |
| "learning_rate": 7.644230765104735e-06, | |
| "loss": 0.2144, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.708502024291498, | |
| "grad_norm": 0.4170680753965977, | |
| "learning_rate": 7.599809050784502e-06, | |
| "loss": 0.1953, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.7125506072874495, | |
| "grad_norm": 0.41596671141145297, | |
| "learning_rate": 7.555486488060661e-06, | |
| "loss": 0.2212, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.716599190283401, | |
| "grad_norm": 0.42168603848949027, | |
| "learning_rate": 7.511263431335249e-06, | |
| "loss": 0.2169, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.720647773279352, | |
| "grad_norm": 0.4255468513486364, | |
| "learning_rate": 7.467140234214656e-06, | |
| "loss": 0.2153, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.7246963562753037, | |
| "grad_norm": 0.3926648395739187, | |
| "learning_rate": 7.423117249506777e-06, | |
| "loss": 0.2245, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7287449392712553, | |
| "grad_norm": 0.4218996494280218, | |
| "learning_rate": 7.379194829218231e-06, | |
| "loss": 0.2308, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.7327935222672064, | |
| "grad_norm": 0.398246587057973, | |
| "learning_rate": 7.335373324551518e-06, | |
| "loss": 0.2021, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.3988212397779606, | |
| "learning_rate": 7.291653085902224e-06, | |
| "loss": 0.2082, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.7408906882591095, | |
| "grad_norm": 0.41895675702350843, | |
| "learning_rate": 7.248034462856213e-06, | |
| "loss": 0.2151, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.7449392712550607, | |
| "grad_norm": 0.43653742152131353, | |
| "learning_rate": 7.20451780418683e-06, | |
| "loss": 0.1948, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.748987854251012, | |
| "grad_norm": 0.40618860774697296, | |
| "learning_rate": 7.161103457852128e-06, | |
| "loss": 0.2047, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.753036437246964, | |
| "grad_norm": 0.4196556590060039, | |
| "learning_rate": 7.117791770992046e-06, | |
| "loss": 0.2163, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.757085020242915, | |
| "grad_norm": 0.4034698022363297, | |
| "learning_rate": 7.074583089925688e-06, | |
| "loss": 0.2081, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.7611336032388665, | |
| "grad_norm": 0.4062414576298064, | |
| "learning_rate": 7.0314777601485155e-06, | |
| "loss": 0.2003, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.765182186234818, | |
| "grad_norm": 0.40874385159890086, | |
| "learning_rate": 6.988476126329604e-06, | |
| "loss": 0.2122, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.4078529886833036, | |
| "learning_rate": 6.945578532308857e-06, | |
| "loss": 0.2044, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.7732793522267207, | |
| "grad_norm": 0.3991290276047874, | |
| "learning_rate": 6.902785321094301e-06, | |
| "loss": 0.2207, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.7773279352226723, | |
| "grad_norm": 0.42533588464603084, | |
| "learning_rate": 6.8600968348593115e-06, | |
| "loss": 0.2077, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.7813765182186234, | |
| "grad_norm": 0.4130659075112586, | |
| "learning_rate": 6.817513414939892e-06, | |
| "loss": 0.218, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.785425101214575, | |
| "grad_norm": 0.38126254628022355, | |
| "learning_rate": 6.775035401831924e-06, | |
| "loss": 0.2023, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.3998923367997668, | |
| "learning_rate": 6.732663135188473e-06, | |
| "loss": 0.2186, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.7935222672064777, | |
| "grad_norm": 0.4174690947334511, | |
| "learning_rate": 6.690396953817055e-06, | |
| "loss": 0.2033, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.7975708502024292, | |
| "grad_norm": 0.4107050866794266, | |
| "learning_rate": 6.648237195676927e-06, | |
| "loss": 0.1837, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.801619433198381, | |
| "grad_norm": 0.3866703446485205, | |
| "learning_rate": 6.606184197876395e-06, | |
| "loss": 0.2133, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.805668016194332, | |
| "grad_norm": 0.39854732981502716, | |
| "learning_rate": 6.564238296670109e-06, | |
| "loss": 0.2081, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.8097165991902835, | |
| "grad_norm": 0.4101567953262562, | |
| "learning_rate": 6.52239982745638e-06, | |
| "loss": 0.1947, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.813765182186235, | |
| "grad_norm": 0.3978294617450139, | |
| "learning_rate": 6.480669124774484e-06, | |
| "loss": 0.2456, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.817813765182186, | |
| "grad_norm": 0.4032920701642121, | |
| "learning_rate": 6.439046522302013e-06, | |
| "loss": 0.1943, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.8218623481781377, | |
| "grad_norm": 0.40134738987562846, | |
| "learning_rate": 6.397532352852191e-06, | |
| "loss": 0.1942, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.8259109311740893, | |
| "grad_norm": 0.44659782326448644, | |
| "learning_rate": 6.356126948371202e-06, | |
| "loss": 0.1854, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.8299595141700404, | |
| "grad_norm": 0.4053578598326146, | |
| "learning_rate": 6.314830639935559e-06, | |
| "loss": 0.2093, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.834008097165992, | |
| "grad_norm": 0.41167583083621045, | |
| "learning_rate": 6.273643757749448e-06, | |
| "loss": 0.2051, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.8380566801619436, | |
| "grad_norm": 0.6218223016507936, | |
| "learning_rate": 6.2325666311420856e-06, | |
| "loss": 0.2374, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.4089927314461644, | |
| "learning_rate": 6.1915995885650675e-06, | |
| "loss": 0.1793, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.4021585958859696, | |
| "learning_rate": 6.150742957589784e-06, | |
| "loss": 0.2178, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.850202429149798, | |
| "grad_norm": 0.4058113025076413, | |
| "learning_rate": 6.109997064904774e-06, | |
| "loss": 0.2427, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.854251012145749, | |
| "grad_norm": 0.39271737671576734, | |
| "learning_rate": 6.069362236313092e-06, | |
| "loss": 0.2106, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.8582995951417005, | |
| "grad_norm": 0.43162049043876, | |
| "learning_rate": 6.0288387967297565e-06, | |
| "loss": 0.2109, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.862348178137652, | |
| "grad_norm": 0.41249990056501346, | |
| "learning_rate": 5.988427070179108e-06, | |
| "loss": 0.203, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.866396761133603, | |
| "grad_norm": 0.38931605420075743, | |
| "learning_rate": 5.948127379792234e-06, | |
| "loss": 0.2166, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.8704453441295548, | |
| "grad_norm": 0.40670332625059796, | |
| "learning_rate": 5.90794004780439e-06, | |
| "loss": 0.2028, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.8744939271255063, | |
| "grad_norm": 0.3871828524984619, | |
| "learning_rate": 5.867865395552405e-06, | |
| "loss": 0.2101, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.8785425101214575, | |
| "grad_norm": 0.39683226415858774, | |
| "learning_rate": 5.8279037434721364e-06, | |
| "loss": 0.2003, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.882591093117409, | |
| "grad_norm": 0.4085894923075928, | |
| "learning_rate": 5.788055411095892e-06, | |
| "loss": 0.2388, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.8866396761133606, | |
| "grad_norm": 0.3822815776821334, | |
| "learning_rate": 5.7483207170498625e-06, | |
| "loss": 0.2116, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8906882591093117, | |
| "grad_norm": 0.41405155839597707, | |
| "learning_rate": 5.708699979051613e-06, | |
| "loss": 0.2177, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.39074293157528767, | |
| "learning_rate": 5.66919351390751e-06, | |
| "loss": 0.2118, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.898785425101215, | |
| "grad_norm": 0.38345516564800136, | |
| "learning_rate": 5.629801637510188e-06, | |
| "loss": 0.2103, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.902834008097166, | |
| "grad_norm": 0.4114944146790465, | |
| "learning_rate": 5.590524664836043e-06, | |
| "loss": 0.184, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.9068825910931175, | |
| "grad_norm": 0.4043319689711557, | |
| "learning_rate": 5.551362909942708e-06, | |
| "loss": 0.2216, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.910931174089069, | |
| "grad_norm": 0.3874631564300556, | |
| "learning_rate": 5.51231668596653e-06, | |
| "loss": 0.2082, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.91497975708502, | |
| "grad_norm": 0.3879391140875748, | |
| "learning_rate": 5.47338630512007e-06, | |
| "loss": 0.218, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.919028340080972, | |
| "grad_norm": 0.39654119082769795, | |
| "learning_rate": 5.434572078689617e-06, | |
| "loss": 0.2194, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.40411410727515473, | |
| "learning_rate": 5.395874317032694e-06, | |
| "loss": 0.2043, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.9271255060728745, | |
| "grad_norm": 0.4068868717909876, | |
| "learning_rate": 5.357293329575573e-06, | |
| "loss": 0.2171, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.931174089068826, | |
| "grad_norm": 0.39659600846004167, | |
| "learning_rate": 5.318829424810797e-06, | |
| "loss": 0.2278, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.9352226720647776, | |
| "grad_norm": 0.3947554375353607, | |
| "learning_rate": 5.280482910294728e-06, | |
| "loss": 0.1986, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.9392712550607287, | |
| "grad_norm": 0.3941293403993755, | |
| "learning_rate": 5.242254092645078e-06, | |
| "loss": 0.229, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.9433198380566803, | |
| "grad_norm": 0.41715636182665916, | |
| "learning_rate": 5.2041432775384445e-06, | |
| "loss": 0.2195, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.40339006129647886, | |
| "learning_rate": 5.166150769707892e-06, | |
| "loss": 0.2094, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.951417004048583, | |
| "grad_norm": 0.3968436593312863, | |
| "learning_rate": 5.128276872940503e-06, | |
| "loss": 0.2052, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.9554655870445345, | |
| "grad_norm": 0.38666371316414033, | |
| "learning_rate": 5.090521890074953e-06, | |
| "loss": 0.2184, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.9595141700404857, | |
| "grad_norm": 0.39406605831259145, | |
| "learning_rate": 5.0528861229990675e-06, | |
| "loss": 0.2144, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.9635627530364372, | |
| "grad_norm": 0.41687723873386234, | |
| "learning_rate": 5.015369872647442e-06, | |
| "loss": 0.203, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.967611336032389, | |
| "grad_norm": 0.4031852310594219, | |
| "learning_rate": 4.9779734389990195e-06, | |
| "loss": 0.2041, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97165991902834, | |
| "grad_norm": 0.3916684289933084, | |
| "learning_rate": 4.940697121074675e-06, | |
| "loss": 0.2221, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.9757085020242915, | |
| "grad_norm": 0.3776644976575847, | |
| "learning_rate": 4.903541216934864e-06, | |
| "loss": 0.2297, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.979757085020243, | |
| "grad_norm": 0.40911791501330746, | |
| "learning_rate": 4.8665060236772e-06, | |
| "loss": 0.187, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.983805668016194, | |
| "grad_norm": 0.35909481216993605, | |
| "learning_rate": 4.829591837434102e-06, | |
| "loss": 0.2254, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.9878542510121457, | |
| "grad_norm": 0.3980498419255356, | |
| "learning_rate": 4.792798953370424e-06, | |
| "loss": 0.211, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.9919028340080973, | |
| "grad_norm": 0.38577283820044633, | |
| "learning_rate": 4.756127665681083e-06, | |
| "loss": 0.2186, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.9959514170040484, | |
| "grad_norm": 0.388249963849127, | |
| "learning_rate": 4.719578267588725e-06, | |
| "loss": 0.1917, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3994036880357296, | |
| "learning_rate": 4.6831510513413525e-06, | |
| "loss": 0.2105, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.004048582995951, | |
| "grad_norm": 0.568827263979545, | |
| "learning_rate": 4.6468463082100176e-06, | |
| "loss": 0.1341, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.008097165991903, | |
| "grad_norm": 0.5085493316873685, | |
| "learning_rate": 4.610664328486485e-06, | |
| "loss": 0.1321, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.012145748987854, | |
| "grad_norm": 0.3798541970848187, | |
| "learning_rate": 4.5746054014809e-06, | |
| "loss": 0.1243, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.016194331983805, | |
| "grad_norm": 0.3321662718172259, | |
| "learning_rate": 4.5386698155194766e-06, | |
| "loss": 0.1407, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.020242914979757, | |
| "grad_norm": 0.3455506419472206, | |
| "learning_rate": 4.502857857942204e-06, | |
| "loss": 0.1447, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.0242914979757085, | |
| "grad_norm": 0.41382736238353185, | |
| "learning_rate": 4.467169815100549e-06, | |
| "loss": 0.1177, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.02834008097166, | |
| "grad_norm": 0.4847635850664381, | |
| "learning_rate": 4.431605972355137e-06, | |
| "loss": 0.1348, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.032388663967612, | |
| "grad_norm": 0.5127700534358939, | |
| "learning_rate": 4.396166614073516e-06, | |
| "loss": 0.1327, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.036437246963563, | |
| "grad_norm": 0.4687775324606054, | |
| "learning_rate": 4.360852023627846e-06, | |
| "loss": 0.1395, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.040485829959514, | |
| "grad_norm": 0.4295998868895013, | |
| "learning_rate": 4.325662483392656e-06, | |
| "loss": 0.1374, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.044534412955466, | |
| "grad_norm": 0.3870897654521579, | |
| "learning_rate": 4.29059827474257e-06, | |
| "loss": 0.134, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.048582995951417, | |
| "grad_norm": 0.3590574457512256, | |
| "learning_rate": 4.255659678050063e-06, | |
| "loss": 0.1302, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.3885453299693591, | |
| "learning_rate": 4.22084697268323e-06, | |
| "loss": 0.1485, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.05668016194332, | |
| "grad_norm": 0.411891439694812, | |
| "learning_rate": 4.186160437003535e-06, | |
| "loss": 0.141, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.060728744939271, | |
| "grad_norm": 0.42514717506723276, | |
| "learning_rate": 4.151600348363582e-06, | |
| "loss": 0.1303, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.064777327935222, | |
| "grad_norm": 0.3586026520449302, | |
| "learning_rate": 4.11716698310493e-06, | |
| "loss": 0.1331, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.068825910931174, | |
| "grad_norm": 0.32283432509107457, | |
| "learning_rate": 4.082860616555848e-06, | |
| "loss": 0.1179, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.0728744939271255, | |
| "grad_norm": 0.3322854956632521, | |
| "learning_rate": 4.048681523029125e-06, | |
| "loss": 0.1222, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.3518488243047731, | |
| "learning_rate": 4.014629975819882e-06, | |
| "loss": 0.1331, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.080971659919029, | |
| "grad_norm": 0.3673647117666814, | |
| "learning_rate": 3.9807062472033855e-06, | |
| "loss": 0.1313, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.08502024291498, | |
| "grad_norm": 0.3763760576187429, | |
| "learning_rate": 3.94691060843287e-06, | |
| "loss": 0.1223, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.089068825910931, | |
| "grad_norm": 0.3676630033938589, | |
| "learning_rate": 3.913243329737353e-06, | |
| "loss": 0.1157, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.093117408906883, | |
| "grad_norm": 0.3584072670422102, | |
| "learning_rate": 3.879704680319503e-06, | |
| "loss": 0.126, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.097165991902834, | |
| "grad_norm": 0.34646826686569276, | |
| "learning_rate": 3.846294928353471e-06, | |
| "loss": 0.1221, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.101214574898785, | |
| "grad_norm": 0.34284596018562136, | |
| "learning_rate": 3.8130143409827345e-06, | |
| "loss": 0.1363, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.34344271322853154, | |
| "learning_rate": 3.7798631843179887e-06, | |
| "loss": 0.1445, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.109311740890688, | |
| "grad_norm": 0.33755673328052926, | |
| "learning_rate": 3.7468417234349973e-06, | |
| "loss": 0.1231, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.113360323886639, | |
| "grad_norm": 0.3524417727223887, | |
| "learning_rate": 3.7139502223724823e-06, | |
| "loss": 0.1422, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.117408906882591, | |
| "grad_norm": 0.34884868454565715, | |
| "learning_rate": 3.6811889441300095e-06, | |
| "loss": 0.1469, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.1214574898785425, | |
| "grad_norm": 0.34702129490515515, | |
| "learning_rate": 3.648558150665886e-06, | |
| "loss": 0.132, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.125506072874494, | |
| "grad_norm": 0.3409211496086895, | |
| "learning_rate": 3.616058102895075e-06, | |
| "loss": 0.1154, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.129554655870446, | |
| "grad_norm": 0.3312768312766913, | |
| "learning_rate": 3.583689060687079e-06, | |
| "loss": 0.1287, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.133603238866397, | |
| "grad_norm": 0.31463584314862286, | |
| "learning_rate": 3.5514512828639023e-06, | |
| "loss": 0.1572, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.137651821862348, | |
| "grad_norm": 0.3322553641088711, | |
| "learning_rate": 3.51934502719796e-06, | |
| "loss": 0.1272, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.1417004048583, | |
| "grad_norm": 0.33590692610470635, | |
| "learning_rate": 3.4873705504100185e-06, | |
| "loss": 0.1222, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.145748987854251, | |
| "grad_norm": 0.34550208476620664, | |
| "learning_rate": 3.4555281081671345e-06, | |
| "loss": 0.1251, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.149797570850202, | |
| "grad_norm": 0.34120351690062806, | |
| "learning_rate": 3.4238179550806326e-06, | |
| "loss": 0.1244, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.3353633923507429, | |
| "learning_rate": 3.3922403447040497e-06, | |
| "loss": 0.1221, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.3217607791157494, | |
| "learning_rate": 3.3607955295311247e-06, | |
| "loss": 0.1689, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.161943319838056, | |
| "grad_norm": 0.3301967327686324, | |
| "learning_rate": 3.32948376099375e-06, | |
| "loss": 0.1146, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.165991902834008, | |
| "grad_norm": 0.34267398008128813, | |
| "learning_rate": 3.2983052894599976e-06, | |
| "loss": 0.1276, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.17004048582996, | |
| "grad_norm": 0.32719378586011205, | |
| "learning_rate": 3.2672603642320945e-06, | |
| "loss": 0.1309, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.174089068825911, | |
| "grad_norm": 0.3362016950790297, | |
| "learning_rate": 3.236349233544438e-06, | |
| "loss": 0.1398, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.178137651821863, | |
| "grad_norm": 0.32348956802375783, | |
| "learning_rate": 3.205572144561604e-06, | |
| "loss": 0.1349, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.182186234817814, | |
| "grad_norm": 0.3294292881464577, | |
| "learning_rate": 3.174929343376374e-06, | |
| "loss": 0.1415, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.186234817813765, | |
| "grad_norm": 0.32909084273892647, | |
| "learning_rate": 3.1444210750077777e-06, | |
| "loss": 0.131, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.190283400809717, | |
| "grad_norm": 0.31135044300332654, | |
| "learning_rate": 3.1140475833991046e-06, | |
| "loss": 0.1362, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.194331983805668, | |
| "grad_norm": 0.34809164102676676, | |
| "learning_rate": 3.0838091114159895e-06, | |
| "loss": 0.1325, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.198380566801619, | |
| "grad_norm": 0.32693502376263217, | |
| "learning_rate": 3.0537059008444524e-06, | |
| "loss": 0.1163, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.202429149797571, | |
| "grad_norm": 0.36004391753004433, | |
| "learning_rate": 3.0237381923889563e-06, | |
| "loss": 0.1388, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.206477732793522, | |
| "grad_norm": 0.3273544883847716, | |
| "learning_rate": 2.9939062256704997e-06, | |
| "loss": 0.111, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.34486850685612486, | |
| "learning_rate": 2.964210239224696e-06, | |
| "loss": 0.105, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.2145748987854255, | |
| "grad_norm": 0.3372731693814375, | |
| "learning_rate": 2.934650470499869e-06, | |
| "loss": 0.129, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.218623481781377, | |
| "grad_norm": 0.3271987330239122, | |
| "learning_rate": 2.905227155855128e-06, | |
| "loss": 0.1406, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.222672064777328, | |
| "grad_norm": 0.33266703988580504, | |
| "learning_rate": 2.875940530558519e-06, | |
| "loss": 0.1365, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.22672064777328, | |
| "grad_norm": 0.33387592065381994, | |
| "learning_rate": 2.8467908287851154e-06, | |
| "loss": 0.1075, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.3362864781391286, | |
| "learning_rate": 2.8177782836151603e-06, | |
| "loss": 0.1096, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.234817813765182, | |
| "grad_norm": 0.32160069420740245, | |
| "learning_rate": 2.7889031270321874e-06, | |
| "loss": 0.1391, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.238866396761134, | |
| "grad_norm": 0.3411603668562679, | |
| "learning_rate": 2.760165589921184e-06, | |
| "loss": 0.1065, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.242914979757085, | |
| "grad_norm": 0.33457734888881674, | |
| "learning_rate": 2.7315659020667305e-06, | |
| "loss": 0.1379, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.246963562753036, | |
| "grad_norm": 0.30925121688403345, | |
| "learning_rate": 2.703104292151164e-06, | |
| "loss": 0.1384, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.251012145748988, | |
| "grad_norm": 0.3428185329358589, | |
| "learning_rate": 2.6747809877527607e-06, | |
| "loss": 0.1125, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.255060728744939, | |
| "grad_norm": 0.33179884042910435, | |
| "learning_rate": 2.6465962153439082e-06, | |
| "loss": 0.1377, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.2591093117408905, | |
| "grad_norm": 0.3487031070470737, | |
| "learning_rate": 2.618550200289303e-06, | |
| "loss": 0.1241, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.3382652746513643, | |
| "learning_rate": 2.59064316684412e-06, | |
| "loss": 0.127, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.267206477732794, | |
| "grad_norm": 0.33967703511527153, | |
| "learning_rate": 2.562875338152264e-06, | |
| "loss": 0.1293, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.271255060728745, | |
| "grad_norm": 0.35242121361971157, | |
| "learning_rate": 2.5352469362445552e-06, | |
| "loss": 0.1302, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.275303643724697, | |
| "grad_norm": 0.31950430082874887, | |
| "learning_rate": 2.5077581820369523e-06, | |
| "loss": 0.1071, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.279352226720648, | |
| "grad_norm": 0.3465602956209444, | |
| "learning_rate": 2.480409295328803e-06, | |
| "loss": 0.1399, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.283400809716599, | |
| "grad_norm": 0.3164952398497548, | |
| "learning_rate": 2.45320049480108e-06, | |
| "loss": 0.1401, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.287449392712551, | |
| "grad_norm": 0.3319296783790085, | |
| "learning_rate": 2.4261319980146293e-06, | |
| "loss": 0.1319, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.291497975708502, | |
| "grad_norm": 0.3370963359929347, | |
| "learning_rate": 2.399204021408421e-06, | |
| "loss": 0.1229, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.295546558704453, | |
| "grad_norm": 0.3450055138854635, | |
| "learning_rate": 2.3724167802978395e-06, | |
| "loss": 0.1207, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.299595141700405, | |
| "grad_norm": 0.34620004666839366, | |
| "learning_rate": 2.3457704888729625e-06, | |
| "loss": 0.1253, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.303643724696356, | |
| "grad_norm": 0.3446915216270338, | |
| "learning_rate": 2.3192653601968162e-06, | |
| "loss": 0.1207, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.3307420013894255, | |
| "learning_rate": 2.2929016062037057e-06, | |
| "loss": 0.1257, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.3117408906882595, | |
| "grad_norm": 0.3368803696054483, | |
| "learning_rate": 2.2666794376975055e-06, | |
| "loss": 0.1219, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.3206487444188798, | |
| "learning_rate": 2.24059906434998e-06, | |
| "loss": 0.1328, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.319838056680162, | |
| "grad_norm": 0.32555278582388664, | |
| "learning_rate": 2.214660694699091e-06, | |
| "loss": 0.1262, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.323886639676114, | |
| "grad_norm": 0.3374216802978804, | |
| "learning_rate": 2.188864536147355e-06, | |
| "loss": 0.1519, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.327935222672065, | |
| "grad_norm": 0.32128267526186005, | |
| "learning_rate": 2.1632107949601757e-06, | |
| "loss": 0.1206, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.331983805668016, | |
| "grad_norm": 0.33053447078668513, | |
| "learning_rate": 2.137699676264171e-06, | |
| "loss": 0.1146, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.336032388663968, | |
| "grad_norm": 0.3290549975341042, | |
| "learning_rate": 2.1123313840455738e-06, | |
| "loss": 0.1349, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.340080971659919, | |
| "grad_norm": 0.33923456384085254, | |
| "learning_rate": 2.08710612114857e-06, | |
| "loss": 0.1332, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.34412955465587, | |
| "grad_norm": 0.31833595195586295, | |
| "learning_rate": 2.0620240892736953e-06, | |
| "loss": 0.113, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.348178137651822, | |
| "grad_norm": 0.3220617073001673, | |
| "learning_rate": 2.0370854889762e-06, | |
| "loss": 0.1311, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.352226720647773, | |
| "grad_norm": 0.3254333634997375, | |
| "learning_rate": 2.0122905196644726e-06, | |
| "loss": 0.1147, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3562753036437245, | |
| "grad_norm": 0.33286224571921824, | |
| "learning_rate": 1.987639379598427e-06, | |
| "loss": 0.1258, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.3603238866396765, | |
| "grad_norm": 0.3288840297605256, | |
| "learning_rate": 1.963132265887919e-06, | |
| "loss": 0.1045, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.364372469635628, | |
| "grad_norm": 0.34423273528600573, | |
| "learning_rate": 1.9387693744911784e-06, | |
| "loss": 0.1236, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.33546286358934774, | |
| "learning_rate": 1.9145509002132368e-06, | |
| "loss": 0.1666, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.372469635627531, | |
| "grad_norm": 0.3362465716558694, | |
| "learning_rate": 1.8904770367043723e-06, | |
| "loss": 0.1402, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.376518218623482, | |
| "grad_norm": 0.33310746009798875, | |
| "learning_rate": 1.8665479764585504e-06, | |
| "loss": 0.1607, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.380566801619433, | |
| "grad_norm": 0.32752770718186663, | |
| "learning_rate": 1.8427639108118977e-06, | |
| "loss": 0.124, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.32848843654232757, | |
| "learning_rate": 1.8191250299411755e-06, | |
| "loss": 0.1305, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.388663967611336, | |
| "grad_norm": 0.3483053609106138, | |
| "learning_rate": 1.7956315228622446e-06, | |
| "loss": 0.1242, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.392712550607287, | |
| "grad_norm": 0.31730128269500146, | |
| "learning_rate": 1.7722835774285553e-06, | |
| "loss": 0.1295, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.396761133603239, | |
| "grad_norm": 0.3265902285828709, | |
| "learning_rate": 1.7490813803296624e-06, | |
| "loss": 0.1213, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.40080971659919, | |
| "grad_norm": 0.3382408892319782, | |
| "learning_rate": 1.726025117089718e-06, | |
| "loss": 0.1299, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.4048582995951415, | |
| "grad_norm": 0.33217972967292997, | |
| "learning_rate": 1.70311497206598e-06, | |
| "loss": 0.1265, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.4089068825910935, | |
| "grad_norm": 0.334859008057648, | |
| "learning_rate": 1.6803511284473638e-06, | |
| "loss": 0.1304, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.412955465587045, | |
| "grad_norm": 0.35186974841161806, | |
| "learning_rate": 1.6577337682529538e-06, | |
| "loss": 0.1119, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.417004048582996, | |
| "grad_norm": 0.31658088254286987, | |
| "learning_rate": 1.6352630723305617e-06, | |
| "loss": 0.1121, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.36826736474339966, | |
| "learning_rate": 1.612939220355272e-06, | |
| "loss": 0.1267, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.425101214574899, | |
| "grad_norm": 0.34382550310759313, | |
| "learning_rate": 1.5907623908280112e-06, | |
| "loss": 0.1395, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.42914979757085, | |
| "grad_norm": 0.3172574642438205, | |
| "learning_rate": 1.568732761074121e-06, | |
| "loss": 0.1178, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.433198380566802, | |
| "grad_norm": 0.34884222775479856, | |
| "learning_rate": 1.5468505072419237e-06, | |
| "loss": 0.1436, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.437246963562753, | |
| "grad_norm": 0.3365424322397863, | |
| "learning_rate": 1.5251158043013448e-06, | |
| "loss": 0.1375, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.441295546558704, | |
| "grad_norm": 0.34422774090308084, | |
| "learning_rate": 1.5035288260424885e-06, | |
| "loss": 0.1119, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.445344129554655, | |
| "grad_norm": 0.32069491387848187, | |
| "learning_rate": 1.4820897450742623e-06, | |
| "loss": 0.1197, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.449392712550607, | |
| "grad_norm": 0.337357819774447, | |
| "learning_rate": 1.4607987328229767e-06, | |
| "loss": 0.1403, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.4534412955465585, | |
| "grad_norm": 0.36058924926875635, | |
| "learning_rate": 1.4396559595310056e-06, | |
| "loss": 0.1552, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.4574898785425106, | |
| "grad_norm": 0.337278177517377, | |
| "learning_rate": 1.4186615942553973e-06, | |
| "loss": 0.1166, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.3319222266583466, | |
| "learning_rate": 1.3978158048665424e-06, | |
| "loss": 0.1181, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.465587044534413, | |
| "grad_norm": 0.33583058498147245, | |
| "learning_rate": 1.3771187580468115e-06, | |
| "loss": 0.1354, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.469635627530364, | |
| "grad_norm": 0.32867951723063044, | |
| "learning_rate": 1.3565706192892392e-06, | |
| "loss": 0.1293, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.3312797295404151, | |
| "learning_rate": 1.3361715528961993e-06, | |
| "loss": 0.1405, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.477732793522267, | |
| "grad_norm": 0.34011977487921335, | |
| "learning_rate": 1.3159217219780817e-06, | |
| "loss": 0.1171, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.481781376518219, | |
| "grad_norm": 0.3442332356698423, | |
| "learning_rate": 1.2958212884519949e-06, | |
| "loss": 0.1256, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.48582995951417, | |
| "grad_norm": 0.32857457510562743, | |
| "learning_rate": 1.2758704130404675e-06, | |
| "loss": 0.1344, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.489878542510121, | |
| "grad_norm": 0.3410053747594066, | |
| "learning_rate": 1.2560692552701738e-06, | |
| "loss": 0.1193, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.493927125506072, | |
| "grad_norm": 0.35522933664168593, | |
| "learning_rate": 1.2364179734706382e-06, | |
| "loss": 0.1103, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.497975708502024, | |
| "grad_norm": 0.3263821267914822, | |
| "learning_rate": 1.2169167247729852e-06, | |
| "loss": 0.1141, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.502024291497976, | |
| "grad_norm": 0.31943000771826063, | |
| "learning_rate": 1.1975656651086864e-06, | |
| "loss": 0.1269, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.506072874493928, | |
| "grad_norm": 0.32765854414829465, | |
| "learning_rate": 1.1783649492082926e-06, | |
| "loss": 0.1129, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.510121457489879, | |
| "grad_norm": 0.32338537316206356, | |
| "learning_rate": 1.1593147306002183e-06, | |
| "loss": 0.1284, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.51417004048583, | |
| "grad_norm": 0.34263145303560005, | |
| "learning_rate": 1.140415161609505e-06, | |
| "loss": 0.1252, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.518218623481781, | |
| "grad_norm": 0.32832982511253034, | |
| "learning_rate": 1.121666393356604e-06, | |
| "loss": 0.1242, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.522267206477733, | |
| "grad_norm": 0.3341896668090987, | |
| "learning_rate": 1.1030685757561632e-06, | |
| "loss": 0.142, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.32539525998292546, | |
| "learning_rate": 1.0846218575158373e-06, | |
| "loss": 0.1483, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.530364372469636, | |
| "grad_norm": 0.31639740023778895, | |
| "learning_rate": 1.066326386135097e-06, | |
| "loss": 0.1265, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.534412955465587, | |
| "grad_norm": 0.34609778295468857, | |
| "learning_rate": 1.0481823079040398e-06, | |
| "loss": 0.1209, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.3455943009145169, | |
| "learning_rate": 1.0301897679022344e-06, | |
| "loss": 0.1267, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.5425101214574894, | |
| "grad_norm": 0.3191637100281669, | |
| "learning_rate": 1.0123489099975491e-06, | |
| "loss": 0.1235, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.5465587044534415, | |
| "grad_norm": 0.3366709940780492, | |
| "learning_rate": 9.9465987684501e-07, | |
| "loss": 0.1282, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.550607287449393, | |
| "grad_norm": 0.32008991871812686, | |
| "learning_rate": 9.771228098856534e-07, | |
| "loss": 0.1296, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.554655870445345, | |
| "grad_norm": 0.3228852091893408, | |
| "learning_rate": 9.597378493454013e-07, | |
| "loss": 0.1318, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.558704453441296, | |
| "grad_norm": 0.33547186816042845, | |
| "learning_rate": 9.425051342339353e-07, | |
| "loss": 0.127, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.562753036437247, | |
| "grad_norm": 0.34653072951234576, | |
| "learning_rate": 9.254248023435864e-07, | |
| "loss": 0.1281, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.566801619433198, | |
| "grad_norm": 0.31238133644262384, | |
| "learning_rate": 9.084969902482288e-07, | |
| "loss": 0.1321, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.57085020242915, | |
| "grad_norm": 0.32492848911902816, | |
| "learning_rate": 8.917218333021993e-07, | |
| "loss": 0.117, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.574898785425101, | |
| "grad_norm": 0.31879824684545705, | |
| "learning_rate": 8.750994656392109e-07, | |
| "loss": 0.1347, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.32118515681576126, | |
| "learning_rate": 8.58630020171265e-07, | |
| "loss": 0.1127, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.582995951417004, | |
| "grad_norm": 0.34663878371891466, | |
| "learning_rate": 8.423136285876099e-07, | |
| "loss": 0.1265, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.587044534412955, | |
| "grad_norm": 0.324825662048009, | |
| "learning_rate": 8.26150421353682e-07, | |
| "loss": 0.139, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.5910931174089065, | |
| "grad_norm": 0.32914357557023544, | |
| "learning_rate": 8.101405277100549e-07, | |
| "loss": 0.1255, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.5951417004048585, | |
| "grad_norm": 0.33374370530029485, | |
| "learning_rate": 7.942840756714077e-07, | |
| "loss": 0.106, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.59919028340081, | |
| "grad_norm": 0.3263159161354798, | |
| "learning_rate": 7.785811920255093e-07, | |
| "loss": 0.1444, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.603238866396762, | |
| "grad_norm": 0.32467413994993777, | |
| "learning_rate": 7.630320023322024e-07, | |
| "loss": 0.1277, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.607287449392713, | |
| "grad_norm": 0.31963485690415616, | |
| "learning_rate": 7.476366309223903e-07, | |
| "loss": 0.1343, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.611336032388664, | |
| "grad_norm": 0.3528229763206559, | |
| "learning_rate": 7.323952008970537e-07, | |
| "loss": 0.1251, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.3287144874114821, | |
| "learning_rate": 7.173078341262618e-07, | |
| "loss": 0.1251, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.619433198380567, | |
| "grad_norm": 0.32368173946899176, | |
| "learning_rate": 7.023746512481989e-07, | |
| "loss": 0.1414, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.623481781376518, | |
| "grad_norm": 0.33147087831282, | |
| "learning_rate": 6.875957716681902e-07, | |
| "loss": 0.1437, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.62753036437247, | |
| "grad_norm": 0.33919141521925805, | |
| "learning_rate": 6.729713135577687e-07, | |
| "loss": 0.1384, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.3389888789658539, | |
| "learning_rate": 6.585013938537143e-07, | |
| "loss": 0.1374, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.635627530364372, | |
| "grad_norm": 0.33245852835402817, | |
| "learning_rate": 6.441861282571115e-07, | |
| "loss": 0.1171, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.6396761133603235, | |
| "grad_norm": 0.3374953300409512, | |
| "learning_rate": 6.300256312324471e-07, | |
| "loss": 0.1239, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.6437246963562755, | |
| "grad_norm": 0.3182326092813078, | |
| "learning_rate": 6.160200160066776e-07, | |
| "loss": 0.1108, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.647773279352227, | |
| "grad_norm": 0.3316480090669831, | |
| "learning_rate": 6.021693945683327e-07, | |
| "loss": 0.1433, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.651821862348179, | |
| "grad_norm": 0.33405896679038477, | |
| "learning_rate": 5.884738776666044e-07, | |
| "loss": 0.1436, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.65587044534413, | |
| "grad_norm": 0.3501796464527722, | |
| "learning_rate": 5.749335748104878e-07, | |
| "loss": 0.1351, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.659919028340081, | |
| "grad_norm": 0.31973443692331655, | |
| "learning_rate": 5.615485942678778e-07, | |
| "loss": 0.1305, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.663967611336032, | |
| "grad_norm": 0.33436006124193657, | |
| "learning_rate": 5.483190430647289e-07, | |
| "loss": 0.125, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.668016194331984, | |
| "grad_norm": 0.3423824920345662, | |
| "learning_rate": 5.352450269841747e-07, | |
| "loss": 0.1212, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.672064777327935, | |
| "grad_norm": 0.3334672105592075, | |
| "learning_rate": 5.22326650565701e-07, | |
| "loss": 0.1076, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.676113360323887, | |
| "grad_norm": 0.3386173367925981, | |
| "learning_rate": 5.095640171043048e-07, | |
| "loss": 0.1166, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.680161943319838, | |
| "grad_norm": 0.32366386482309495, | |
| "learning_rate": 4.969572286496594e-07, | |
| "loss": 0.1116, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.31979111401484567, | |
| "learning_rate": 4.845063860053079e-07, | |
| "loss": 0.1163, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.6882591093117405, | |
| "grad_norm": 0.31922998476342845, | |
| "learning_rate": 4.7221158872785997e-07, | |
| "loss": 0.1177, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 0.332837142954603, | |
| "learning_rate": 4.600729351261857e-07, | |
| "loss": 0.1447, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.696356275303644, | |
| "grad_norm": 0.32828377023730304, | |
| "learning_rate": 4.4809052226063355e-07, | |
| "loss": 0.1375, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.700404858299595, | |
| "grad_norm": 0.3274992039641488, | |
| "learning_rate": 4.3626444594226046e-07, | |
| "loss": 0.1166, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.704453441295547, | |
| "grad_norm": 0.3260288359149715, | |
| "learning_rate": 4.245948007320588e-07, | |
| "loss": 0.1058, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.708502024291498, | |
| "grad_norm": 0.34286616490246513, | |
| "learning_rate": 4.130816799401971e-07, | |
| "loss": 0.1361, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.712550607287449, | |
| "grad_norm": 0.34038913166573087, | |
| "learning_rate": 4.017251756252827e-07, | |
| "loss": 0.1169, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.716599190283401, | |
| "grad_norm": 0.3319871928944993, | |
| "learning_rate": 3.9052537859362473e-07, | |
| "loss": 0.1301, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.720647773279352, | |
| "grad_norm": 0.3159888994598253, | |
| "learning_rate": 3.7948237839849687e-07, | |
| "loss": 0.1314, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.724696356275303, | |
| "grad_norm": 0.33079466431547594, | |
| "learning_rate": 3.6859626333943797e-07, | |
| "loss": 0.1215, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.728744939271255, | |
| "grad_norm": 0.32855172985634923, | |
| "learning_rate": 3.578671204615347e-07, | |
| "loss": 0.1209, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.732793522267206, | |
| "grad_norm": 0.3292846593734423, | |
| "learning_rate": 3.472950355547267e-07, | |
| "loss": 0.1361, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.3363380066487142, | |
| "learning_rate": 3.3688009315312244e-07, | |
| "loss": 0.1361, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.7408906882591095, | |
| "grad_norm": 0.31980816027340103, | |
| "learning_rate": 3.266223765343224e-07, | |
| "loss": 0.1126, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.744939271255061, | |
| "grad_norm": 0.3605095833424521, | |
| "learning_rate": 3.16521967718757e-07, | |
| "loss": 0.1466, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.748987854251012, | |
| "grad_norm": 0.33047878383808815, | |
| "learning_rate": 3.0657894746902726e-07, | |
| "loss": 0.1205, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.753036437246964, | |
| "grad_norm": 0.3365212132960834, | |
| "learning_rate": 2.9679339528925654e-07, | |
| "loss": 0.1374, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.757085020242915, | |
| "grad_norm": 0.32848887456536086, | |
| "learning_rate": 2.871653894244597e-07, | |
| "loss": 0.1151, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.761133603238866, | |
| "grad_norm": 0.35696630315830347, | |
| "learning_rate": 2.7769500685991266e-07, | |
| "loss": 0.1251, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.765182186234818, | |
| "grad_norm": 0.3240889456835772, | |
| "learning_rate": 2.683823233205485e-07, | |
| "loss": 0.1362, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.3339885541983915, | |
| "learning_rate": 2.592274132703287e-07, | |
| "loss": 0.1484, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.77327935222672, | |
| "grad_norm": 0.33253043693556594, | |
| "learning_rate": 2.502303499116754e-07, | |
| "loss": 0.1423, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.777327935222672, | |
| "grad_norm": 0.32356207980841395, | |
| "learning_rate": 2.413912051848666e-07, | |
| "loss": 0.1191, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.781376518218623, | |
| "grad_norm": 0.32303470979605653, | |
| "learning_rate": 2.3271004976746613e-07, | |
| "loss": 0.1413, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.7854251012145745, | |
| "grad_norm": 0.34066354652880243, | |
| "learning_rate": 2.2418695307376393e-07, | |
| "loss": 0.126, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.3196825472013292, | |
| "learning_rate": 2.1582198325420968e-07, | |
| "loss": 0.118, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.793522267206478, | |
| "grad_norm": 0.3209302819834496, | |
| "learning_rate": 2.0761520719488447e-07, | |
| "loss": 0.1108, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.797570850202429, | |
| "grad_norm": 0.33737292878488634, | |
| "learning_rate": 1.9956669051694798e-07, | |
| "loss": 0.1232, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.801619433198381, | |
| "grad_norm": 0.33618028113158677, | |
| "learning_rate": 1.916764975761254e-07, | |
| "loss": 0.122, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.805668016194332, | |
| "grad_norm": 0.32786806554497855, | |
| "learning_rate": 1.839446914621923e-07, | |
| "loss": 0.1108, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.809716599190283, | |
| "grad_norm": 0.3211739235990876, | |
| "learning_rate": 1.7637133399845962e-07, | |
| "loss": 0.1397, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.813765182186235, | |
| "grad_norm": 0.32251617473542776, | |
| "learning_rate": 1.6895648574129397e-07, | |
| "loss": 0.1216, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.817813765182186, | |
| "grad_norm": 0.30384505509885257, | |
| "learning_rate": 1.6170020597962465e-07, | |
| "loss": 0.1243, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.821862348178137, | |
| "grad_norm": 0.3239686760708486, | |
| "learning_rate": 1.5460255273447078e-07, | |
| "loss": 0.1247, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.825910931174089, | |
| "grad_norm": 0.3181590943369618, | |
| "learning_rate": 1.4766358275847715e-07, | |
| "loss": 0.1282, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.82995951417004, | |
| "grad_norm": 0.34581203737117877, | |
| "learning_rate": 1.4088335153546573e-07, | |
| "loss": 0.1247, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.834008097165992, | |
| "grad_norm": 0.33563854287791484, | |
| "learning_rate": 1.3426191327998496e-07, | |
| "loss": 0.1302, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.838056680161944, | |
| "grad_norm": 0.33172959888528847, | |
| "learning_rate": 1.277993209368744e-07, | |
| "loss": 0.1171, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.34583738085264437, | |
| "learning_rate": 1.214956261808542e-07, | |
| "loss": 0.1381, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.32793280407451403, | |
| "learning_rate": 1.1535087941609846e-07, | |
| "loss": 0.1567, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.850202429149798, | |
| "grad_norm": 0.32561400686372705, | |
| "learning_rate": 1.0936512977583802e-07, | |
| "loss": 0.1352, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.854251012145749, | |
| "grad_norm": 0.3342588050470388, | |
| "learning_rate": 1.0353842512196732e-07, | |
| "loss": 0.1286, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.8582995951417, | |
| "grad_norm": 0.34272068705739084, | |
| "learning_rate": 9.787081204466253e-08, | |
| "loss": 0.1193, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.862348178137652, | |
| "grad_norm": 0.31540991665735324, | |
| "learning_rate": 9.236233586200627e-08, | |
| "loss": 0.1545, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.866396761133603, | |
| "grad_norm": 0.34605510745219814, | |
| "learning_rate": 8.70130406196279e-08, | |
| "loss": 0.1203, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.870445344129554, | |
| "grad_norm": 0.3192240725076089, | |
| "learning_rate": 8.182296909035047e-08, | |
| "loss": 0.1336, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.874493927125506, | |
| "grad_norm": 0.3245515309612596, | |
| "learning_rate": 7.679216277384439e-08, | |
| "loss": 0.1223, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.8785425101214575, | |
| "grad_norm": 0.317014818634635, | |
| "learning_rate": 7.19206618963053e-08, | |
| "loss": 0.1284, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.882591093117409, | |
| "grad_norm": 0.33791965303814275, | |
| "learning_rate": 6.720850541012347e-08, | |
| "loss": 0.1506, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.886639676113361, | |
| "grad_norm": 0.3186095548086024, | |
| "learning_rate": 6.265573099357714e-08, | |
| "loss": 0.1225, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.890688259109312, | |
| "grad_norm": 0.32745265404841545, | |
| "learning_rate": 5.826237505053067e-08, | |
| "loss": 0.1106, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.33361702357917433, | |
| "learning_rate": 5.4028472710137e-08, | |
| "loss": 0.1324, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.898785425101215, | |
| "grad_norm": 0.33054494412119406, | |
| "learning_rate": 4.9954057826571146e-08, | |
| "loss": 0.1447, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.902834008097166, | |
| "grad_norm": 0.3258762722144293, | |
| "learning_rate": 4.6039162978743776e-08, | |
| "loss": 0.1391, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.906882591093117, | |
| "grad_norm": 0.3204132184562727, | |
| "learning_rate": 4.228381947004812e-08, | |
| "loss": 0.1259, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.910931174089069, | |
| "grad_norm": 0.3287726412928029, | |
| "learning_rate": 3.868805732811343e-08, | |
| "loss": 0.124, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.91497975708502, | |
| "grad_norm": 0.32226632408839956, | |
| "learning_rate": 3.525190530455636e-08, | |
| "loss": 0.1374, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.919028340080971, | |
| "grad_norm": 0.3413421070617115, | |
| "learning_rate": 3.197539087475887e-08, | |
| "loss": 0.1263, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.34145063213917337, | |
| "learning_rate": 2.8858540237641786e-08, | |
| "loss": 0.132, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.9271255060728745, | |
| "grad_norm": 0.34381757167919064, | |
| "learning_rate": 2.5901378315458247e-08, | |
| "loss": 0.1334, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.931174089068826, | |
| "grad_norm": 0.3359366312178854, | |
| "learning_rate": 2.3103928753598347e-08, | |
| "loss": 0.1371, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.935222672064778, | |
| "grad_norm": 0.31311766252237705, | |
| "learning_rate": 2.0466213920393718e-08, | |
| "loss": 0.1241, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.939271255060729, | |
| "grad_norm": 0.32988709325971555, | |
| "learning_rate": 1.7988254906942115e-08, | |
| "loss": 0.1177, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94331983805668, | |
| "grad_norm": 0.32266280919667006, | |
| "learning_rate": 1.5670071526936447e-08, | |
| "loss": 0.1416, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.3476029513760411, | |
| "learning_rate": 1.3511682316509344e-08, | |
| "loss": 0.1301, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.951417004048583, | |
| "grad_norm": 0.3306357036385511, | |
| "learning_rate": 1.1513104534086605e-08, | |
| "loss": 0.1476, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.955465587044534, | |
| "grad_norm": 0.3328926278311347, | |
| "learning_rate": 9.67435416023843e-09, | |
| "loss": 0.1274, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.959514170040486, | |
| "grad_norm": 0.31695716444501576, | |
| "learning_rate": 7.995445897566178e-09, | |
| "loss": 0.1409, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.963562753036437, | |
| "grad_norm": 0.3270844142706258, | |
| "learning_rate": 6.476393170578021e-09, | |
| "loss": 0.1413, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.967611336032388, | |
| "grad_norm": 0.3380241531232909, | |
| "learning_rate": 5.11720812557348e-09, | |
| "loss": 0.1308, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.97165991902834, | |
| "grad_norm": 0.3263464268465559, | |
| "learning_rate": 3.917901630563492e-09, | |
| "loss": 0.1191, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.9757085020242915, | |
| "grad_norm": 0.33872317110316447, | |
| "learning_rate": 2.878483275170485e-09, | |
| "loss": 0.1283, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.979757085020243, | |
| "grad_norm": 0.33384688641566684, | |
| "learning_rate": 1.998961370555108e-09, | |
| "loss": 0.1143, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.983805668016195, | |
| "grad_norm": 0.31872993868379224, | |
| "learning_rate": 1.2793429493518361e-09, | |
| "loss": 0.1393, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.987854251012146, | |
| "grad_norm": 0.3026318352337657, | |
| "learning_rate": 7.196337656112384e-10, | |
| "loss": 0.1452, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.991902834008097, | |
| "grad_norm": 0.32897463630818113, | |
| "learning_rate": 3.198382947511292e-10, | |
| "loss": 0.1296, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.995951417004049, | |
| "grad_norm": 0.3283323665148413, | |
| "learning_rate": 7.995973352770137e-11, | |
| "loss": 0.1248, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.33409118073484556, | |
| "learning_rate": 0.0, | |
| "loss": 0.1149, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1235, | |
| "total_flos": 1.1404623141058642e+18, | |
| "train_loss": 0.39826868520936504, | |
| "train_runtime": 17206.7809, | |
| "train_samples_per_second": 9.182, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "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": 1.1404623141058642e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |