| { | |
| "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": 12.182062107416865, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 1.4095, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008097165991902834, | |
| "grad_norm": 11.717496718152374, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 1.3877, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012145748987854251, | |
| "grad_norm": 12.262839551374327, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 1.4096, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016194331983805668, | |
| "grad_norm": 11.32692490411073, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 1.3998, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020242914979757085, | |
| "grad_norm": 10.138778823196647, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 1.3665, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024291497975708502, | |
| "grad_norm": 9.461281723956844, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.3468, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02834008097165992, | |
| "grad_norm": 6.451372280298322, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 1.2908, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032388663967611336, | |
| "grad_norm": 5.956618325676995, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 1.3001, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03643724696356275, | |
| "grad_norm": 3.751692133815938, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 1.239, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04048582995951417, | |
| "grad_norm": 3.348396685883471, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 1.2078, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044534412955465584, | |
| "grad_norm": 3.08122654044921, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 1.1914, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048582995951417005, | |
| "grad_norm": 5.17900869509297, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 1.1697, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 5.500586093283521, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 1.1499, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05668016194331984, | |
| "grad_norm": 4.937204162414813, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 1.1434, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06072874493927125, | |
| "grad_norm": 4.48293123780571, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 1.1377, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06477732793522267, | |
| "grad_norm": 3.061050700642893, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 1.0766, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06882591093117409, | |
| "grad_norm": 2.8389716794560673, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 1.0344, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0728744939271255, | |
| "grad_norm": 2.4110596840110725, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.9976, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.9719853435514052, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 1.0341, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08097165991902834, | |
| "grad_norm": 1.6900634888374673, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 1.0039, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08502024291497975, | |
| "grad_norm": 1.473089415402246, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.9934, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08906882591093117, | |
| "grad_norm": 1.421719090360208, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.9604, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0931174089068826, | |
| "grad_norm": 1.620382674890822, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.9754, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09716599190283401, | |
| "grad_norm": 1.502238387591459, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.9604, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10121457489878542, | |
| "grad_norm": 1.133997640361694, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.9419, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 1.3552075421183924, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.9332, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10931174089068826, | |
| "grad_norm": 1.2696167054442145, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.9507, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11336032388663968, | |
| "grad_norm": 1.1724259884898933, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.9324, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11740890688259109, | |
| "grad_norm": 1.1204067747043172, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.9329, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1214574898785425, | |
| "grad_norm": 1.0870954568107625, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.9022, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12550607287449392, | |
| "grad_norm": 1.052831854000786, | |
| "learning_rate": 1e-05, | |
| "loss": 0.924, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12955465587044535, | |
| "grad_norm": 0.9048207402859777, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.9011, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13360323886639677, | |
| "grad_norm": 1.0040365017353026, | |
| "learning_rate": 1.0645161290322582e-05, | |
| "loss": 0.9038, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13765182186234817, | |
| "grad_norm": 0.988291698143235, | |
| "learning_rate": 1.096774193548387e-05, | |
| "loss": 0.8906, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1417004048582996, | |
| "grad_norm": 0.9293185106775377, | |
| "learning_rate": 1.1290322580645164e-05, | |
| "loss": 0.8922, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.145748987854251, | |
| "grad_norm": 0.9002638017226141, | |
| "learning_rate": 1.1612903225806453e-05, | |
| "loss": 0.8781, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14979757085020243, | |
| "grad_norm": 0.9730787199298694, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.8928, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.7991697305132978, | |
| "learning_rate": 1.2258064516129034e-05, | |
| "loss": 0.8856, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 0.9136446060305378, | |
| "learning_rate": 1.2580645161290324e-05, | |
| "loss": 0.8641, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16194331983805668, | |
| "grad_norm": 0.8915844975391511, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.874, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1659919028340081, | |
| "grad_norm": 0.8126280199620886, | |
| "learning_rate": 1.3225806451612903e-05, | |
| "loss": 0.8626, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1700404858299595, | |
| "grad_norm": 0.8377986941412183, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.8635, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17408906882591094, | |
| "grad_norm": 0.850210159261465, | |
| "learning_rate": 1.3870967741935486e-05, | |
| "loss": 0.8717, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17813765182186234, | |
| "grad_norm": 0.796106544618932, | |
| "learning_rate": 1.4193548387096776e-05, | |
| "loss": 0.8658, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18218623481781376, | |
| "grad_norm": 0.8284643464429826, | |
| "learning_rate": 1.4516129032258066e-05, | |
| "loss": 0.8553, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1862348178137652, | |
| "grad_norm": 0.8856296048007727, | |
| "learning_rate": 1.4838709677419357e-05, | |
| "loss": 0.8576, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1902834008097166, | |
| "grad_norm": 0.8383815583916348, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.8563, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19433198380566802, | |
| "grad_norm": 0.8424643316363467, | |
| "learning_rate": 1.5483870967741936e-05, | |
| "loss": 0.8502, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19838056680161945, | |
| "grad_norm": 0.913264332169559, | |
| "learning_rate": 1.5806451612903226e-05, | |
| "loss": 0.8546, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20242914979757085, | |
| "grad_norm": 0.8347958187505958, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.8518, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20647773279352227, | |
| "grad_norm": 0.7999061254649359, | |
| "learning_rate": 1.6451612903225807e-05, | |
| "loss": 0.8308, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.9537132090457249, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.8446, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2145748987854251, | |
| "grad_norm": 0.8555888765326704, | |
| "learning_rate": 1.7096774193548388e-05, | |
| "loss": 0.8341, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21862348178137653, | |
| "grad_norm": 0.9530606061555009, | |
| "learning_rate": 1.741935483870968e-05, | |
| "loss": 0.8226, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22267206477732793, | |
| "grad_norm": 0.8597185829619078, | |
| "learning_rate": 1.774193548387097e-05, | |
| "loss": 0.8286, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22672064777327935, | |
| "grad_norm": 0.8944381295696394, | |
| "learning_rate": 1.806451612903226e-05, | |
| "loss": 0.8275, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.9957166379570176, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.8418, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23481781376518218, | |
| "grad_norm": 1.053543433802956, | |
| "learning_rate": 1.870967741935484e-05, | |
| "loss": 0.8026, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2388663967611336, | |
| "grad_norm": 0.8879667449577124, | |
| "learning_rate": 1.903225806451613e-05, | |
| "loss": 0.8098, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.242914979757085, | |
| "grad_norm": 0.9953034141696906, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.8429, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24696356275303644, | |
| "grad_norm": 0.8932078029736371, | |
| "learning_rate": 1.967741935483871e-05, | |
| "loss": 0.8077, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.25101214574898784, | |
| "grad_norm": 0.8185527975807287, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8181, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2550607287449393, | |
| "grad_norm": 0.9317746824118319, | |
| "learning_rate": 2.0322580645161292e-05, | |
| "loss": 0.8156, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2591093117408907, | |
| "grad_norm": 0.9253147144179968, | |
| "learning_rate": 2.0645161290322582e-05, | |
| "loss": 0.8027, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 0.9761890646894145, | |
| "learning_rate": 2.0967741935483873e-05, | |
| "loss": 0.8286, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26720647773279355, | |
| "grad_norm": 0.9692099761463319, | |
| "learning_rate": 2.1290322580645163e-05, | |
| "loss": 0.8295, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27125506072874495, | |
| "grad_norm": 0.937204398317383, | |
| "learning_rate": 2.161290322580645e-05, | |
| "loss": 0.8164, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27530364372469635, | |
| "grad_norm": 0.9425289429021542, | |
| "learning_rate": 2.193548387096774e-05, | |
| "loss": 0.8121, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2793522267206478, | |
| "grad_norm": 1.0462044572566471, | |
| "learning_rate": 2.225806451612903e-05, | |
| "loss": 0.817, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2834008097165992, | |
| "grad_norm": 1.2015734495140493, | |
| "learning_rate": 2.2580645161290328e-05, | |
| "loss": 0.8156, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2874493927125506, | |
| "grad_norm": 1.000226147508421, | |
| "learning_rate": 2.2903225806451615e-05, | |
| "loss": 0.8148, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.291497975708502, | |
| "grad_norm": 1.02182891756832, | |
| "learning_rate": 2.3225806451612906e-05, | |
| "loss": 0.8122, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.29554655870445345, | |
| "grad_norm": 1.2107692252482407, | |
| "learning_rate": 2.3548387096774196e-05, | |
| "loss": 0.8009, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29959514170040485, | |
| "grad_norm": 1.070806312452508, | |
| "learning_rate": 2.3870967741935486e-05, | |
| "loss": 0.7956, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30364372469635625, | |
| "grad_norm": 1.3311354666288953, | |
| "learning_rate": 2.4193548387096777e-05, | |
| "loss": 0.7873, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.9596229700824749, | |
| "learning_rate": 2.4516129032258067e-05, | |
| "loss": 0.8011, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3117408906882591, | |
| "grad_norm": 1.4682495839191074, | |
| "learning_rate": 2.4838709677419358e-05, | |
| "loss": 0.8061, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 0.9569580629239876, | |
| "learning_rate": 2.5161290322580648e-05, | |
| "loss": 0.8063, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31983805668016196, | |
| "grad_norm": 1.0573305468765168, | |
| "learning_rate": 2.548387096774194e-05, | |
| "loss": 0.7817, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32388663967611336, | |
| "grad_norm": 1.0247879312909849, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.783, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32793522267206476, | |
| "grad_norm": 1.0221924900448256, | |
| "learning_rate": 2.6129032258064516e-05, | |
| "loss": 0.808, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3319838056680162, | |
| "grad_norm": 0.9834555598730277, | |
| "learning_rate": 2.6451612903225806e-05, | |
| "loss": 0.7835, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3360323886639676, | |
| "grad_norm": 0.9000600719842645, | |
| "learning_rate": 2.6774193548387097e-05, | |
| "loss": 0.7784, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.340080971659919, | |
| "grad_norm": 0.8700428081870981, | |
| "learning_rate": 2.7096774193548387e-05, | |
| "loss": 0.7838, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3441295546558704, | |
| "grad_norm": 0.9013803485706032, | |
| "learning_rate": 2.7419354838709678e-05, | |
| "loss": 0.788, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3481781376518219, | |
| "grad_norm": 0.9700783164559821, | |
| "learning_rate": 2.774193548387097e-05, | |
| "loss": 0.8057, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3522267206477733, | |
| "grad_norm": 1.2767816408785253, | |
| "learning_rate": 2.8064516129032262e-05, | |
| "loss": 0.7729, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3562753036437247, | |
| "grad_norm": 0.8885548059426953, | |
| "learning_rate": 2.8387096774193552e-05, | |
| "loss": 0.7735, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3603238866396761, | |
| "grad_norm": 1.2225591134993845, | |
| "learning_rate": 2.8709677419354843e-05, | |
| "loss": 0.7858, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3643724696356275, | |
| "grad_norm": 0.861380043624081, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.7767, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 1.0345985735077377, | |
| "learning_rate": 2.9354838709677423e-05, | |
| "loss": 0.7859, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3724696356275304, | |
| "grad_norm": 1.078594185581657, | |
| "learning_rate": 2.9677419354838714e-05, | |
| "loss": 0.7895, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3765182186234818, | |
| "grad_norm": 1.2255575497768838, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.7828, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3805668016194332, | |
| "grad_norm": 0.9237286537206904, | |
| "learning_rate": 3.032258064516129e-05, | |
| "loss": 0.7803, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 1.3028050835507063, | |
| "learning_rate": 3.0645161290322585e-05, | |
| "loss": 0.7954, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 1.2637204222864442, | |
| "learning_rate": 3.096774193548387e-05, | |
| "loss": 0.767, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39271255060728744, | |
| "grad_norm": 1.0684663623638133, | |
| "learning_rate": 3.1290322580645166e-05, | |
| "loss": 0.7723, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3967611336032389, | |
| "grad_norm": 1.2883115185387857, | |
| "learning_rate": 3.161290322580645e-05, | |
| "loss": 0.7691, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4008097165991903, | |
| "grad_norm": 0.8873463734137309, | |
| "learning_rate": 3.193548387096774e-05, | |
| "loss": 0.7835, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4048582995951417, | |
| "grad_norm": 1.0447689716147934, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.7699, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4089068825910931, | |
| "grad_norm": 1.0788287525253393, | |
| "learning_rate": 3.258064516129033e-05, | |
| "loss": 0.7668, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.41295546558704455, | |
| "grad_norm": 0.9533779227961438, | |
| "learning_rate": 3.2903225806451614e-05, | |
| "loss": 0.7462, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.41700404858299595, | |
| "grad_norm": 1.189123021717784, | |
| "learning_rate": 3.322580645161291e-05, | |
| "loss": 0.7718, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.8759430085224712, | |
| "learning_rate": 3.3548387096774195e-05, | |
| "loss": 0.7607, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4251012145748988, | |
| "grad_norm": 1.0959168711301916, | |
| "learning_rate": 3.387096774193549e-05, | |
| "loss": 0.76, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4291497975708502, | |
| "grad_norm": 0.858481187949417, | |
| "learning_rate": 3.4193548387096776e-05, | |
| "loss": 0.7616, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4331983805668016, | |
| "grad_norm": 0.9308535380357283, | |
| "learning_rate": 3.451612903225807e-05, | |
| "loss": 0.7491, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.43724696356275305, | |
| "grad_norm": 0.8075932533226653, | |
| "learning_rate": 3.483870967741936e-05, | |
| "loss": 0.7461, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.44129554655870445, | |
| "grad_norm": 1.0963428138549698, | |
| "learning_rate": 3.516129032258065e-05, | |
| "loss": 0.7654, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44534412955465585, | |
| "grad_norm": 1.3181391812586403, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.7652, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4493927125506073, | |
| "grad_norm": 0.9705590539771749, | |
| "learning_rate": 3.5806451612903225e-05, | |
| "loss": 0.7653, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4534412955465587, | |
| "grad_norm": 1.0291883694445132, | |
| "learning_rate": 3.612903225806452e-05, | |
| "loss": 0.7512, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4574898785425101, | |
| "grad_norm": 1.210905200665075, | |
| "learning_rate": 3.6451612903225805e-05, | |
| "loss": 0.7507, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.6024927549337018, | |
| "learning_rate": 3.67741935483871e-05, | |
| "loss": 0.7595, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46558704453441296, | |
| "grad_norm": 1.1618001770215418, | |
| "learning_rate": 3.7096774193548386e-05, | |
| "loss": 0.7578, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.46963562753036436, | |
| "grad_norm": 1.3699344481158298, | |
| "learning_rate": 3.741935483870968e-05, | |
| "loss": 0.7391, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 1.10140626004365, | |
| "learning_rate": 3.7741935483870974e-05, | |
| "loss": 0.7699, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4777327935222672, | |
| "grad_norm": 1.418834219976673, | |
| "learning_rate": 3.806451612903226e-05, | |
| "loss": 0.7424, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4817813765182186, | |
| "grad_norm": 1.0973828487187547, | |
| "learning_rate": 3.8387096774193555e-05, | |
| "loss": 0.7289, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48582995951417, | |
| "grad_norm": 1.3099236147761113, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.7658, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4898785425101215, | |
| "grad_norm": 1.2654968317266189, | |
| "learning_rate": 3.9032258064516135e-05, | |
| "loss": 0.7585, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4939271255060729, | |
| "grad_norm": 0.9944465765282847, | |
| "learning_rate": 3.935483870967742e-05, | |
| "loss": 0.7469, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4979757085020243, | |
| "grad_norm": 1.2512516734337622, | |
| "learning_rate": 3.9677419354838716e-05, | |
| "loss": 0.7584, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5020242914979757, | |
| "grad_norm": 1.0918433536021444, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7754, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5060728744939271, | |
| "grad_norm": 1.0381333268866684, | |
| "learning_rate": 3.9999920040266474e-05, | |
| "loss": 0.7807, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5101214574898786, | |
| "grad_norm": 1.0261214512989851, | |
| "learning_rate": 3.999968016170525e-05, | |
| "loss": 0.7647, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5141700404858299, | |
| "grad_norm": 0.894412515363923, | |
| "learning_rate": 3.9999280366234394e-05, | |
| "loss": 0.7291, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5182186234817814, | |
| "grad_norm": 1.1062692682253494, | |
| "learning_rate": 3.999872065705065e-05, | |
| "loss": 0.7393, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5222672064777328, | |
| "grad_norm": 0.9538070839354019, | |
| "learning_rate": 3.999800103862945e-05, | |
| "loss": 0.7201, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.3393917888721534, | |
| "learning_rate": 3.9997121516724834e-05, | |
| "loss": 0.724, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5303643724696356, | |
| "grad_norm": 1.3018554223515284, | |
| "learning_rate": 3.999608209836944e-05, | |
| "loss": 0.7459, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5344129554655871, | |
| "grad_norm": 1.0006079803065342, | |
| "learning_rate": 3.999488279187443e-05, | |
| "loss": 0.7274, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.3890571794515558, | |
| "learning_rate": 3.9993523606829425e-05, | |
| "loss": 0.7359, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5425101214574899, | |
| "grad_norm": 1.0074862604893813, | |
| "learning_rate": 3.9992004554102435e-05, | |
| "loss": 0.738, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5465587044534413, | |
| "grad_norm": 1.0096255652537907, | |
| "learning_rate": 3.999032564583976e-05, | |
| "loss": 0.7427, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5506072874493927, | |
| "grad_norm": 1.2863325802080772, | |
| "learning_rate": 3.998848689546591e-05, | |
| "loss": 0.7414, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5546558704453441, | |
| "grad_norm": 0.9522865641039402, | |
| "learning_rate": 3.9986488317683496e-05, | |
| "loss": 0.7447, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5587044534412956, | |
| "grad_norm": 1.1369938144547977, | |
| "learning_rate": 3.9984329928473067e-05, | |
| "loss": 0.7067, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.562753036437247, | |
| "grad_norm": 0.9746993733840074, | |
| "learning_rate": 3.998201174509306e-05, | |
| "loss": 0.7245, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5668016194331984, | |
| "grad_norm": 1.0886760047605528, | |
| "learning_rate": 3.997953378607961e-05, | |
| "loss": 0.7385, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5708502024291497, | |
| "grad_norm": 0.8073745800484787, | |
| "learning_rate": 3.9976896071246403e-05, | |
| "loss": 0.7199, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5748987854251012, | |
| "grad_norm": 1.064873901418392, | |
| "learning_rate": 3.997409862168454e-05, | |
| "loss": 0.7135, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.9188433546912571, | |
| "learning_rate": 3.997114145976237e-05, | |
| "loss": 0.7564, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.582995951417004, | |
| "grad_norm": 0.9951789609055286, | |
| "learning_rate": 3.9968024609125245e-05, | |
| "loss": 0.7282, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5870445344129555, | |
| "grad_norm": 0.9586251943136666, | |
| "learning_rate": 3.996474809469545e-05, | |
| "loss": 0.7385, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5910931174089069, | |
| "grad_norm": 1.1419719005106206, | |
| "learning_rate": 3.996131194267189e-05, | |
| "loss": 0.7279, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5951417004048583, | |
| "grad_norm": 1.2120822278340764, | |
| "learning_rate": 3.995771618052995e-05, | |
| "loss": 0.7207, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5991902834008097, | |
| "grad_norm": 0.9557274891379403, | |
| "learning_rate": 3.9953960837021256e-05, | |
| "loss": 0.741, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6032388663967612, | |
| "grad_norm": 1.2212710750280011, | |
| "learning_rate": 3.995004594217343e-05, | |
| "loss": 0.6971, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6072874493927125, | |
| "grad_norm": 1.3001767936417377, | |
| "learning_rate": 3.994597152728987e-05, | |
| "loss": 0.7266, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.611336032388664, | |
| "grad_norm": 1.0016546538302296, | |
| "learning_rate": 3.994173762494947e-05, | |
| "loss": 0.7053, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.142272939392001, | |
| "learning_rate": 3.9937344269006424e-05, | |
| "loss": 0.7156, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6194331983805668, | |
| "grad_norm": 1.2121576288193212, | |
| "learning_rate": 3.993279149458988e-05, | |
| "loss": 0.7309, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6234817813765182, | |
| "grad_norm": 0.9574760602516821, | |
| "learning_rate": 3.9928079338103695e-05, | |
| "loss": 0.7255, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6275303643724697, | |
| "grad_norm": 1.4518541232276934, | |
| "learning_rate": 3.992320783722616e-05, | |
| "loss": 0.722, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 1.0934710343135225, | |
| "learning_rate": 3.991817703090965e-05, | |
| "loss": 0.7074, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6356275303643725, | |
| "grad_norm": 1.4092060532029405, | |
| "learning_rate": 3.991298695938038e-05, | |
| "loss": 0.7048, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6396761133603239, | |
| "grad_norm": 1.30113439873749, | |
| "learning_rate": 3.9907637664138e-05, | |
| "loss": 0.7128, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6437246963562753, | |
| "grad_norm": 1.2310871227683433, | |
| "learning_rate": 3.990212918795534e-05, | |
| "loss": 0.7024, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6477732793522267, | |
| "grad_norm": 1.3953900474100263, | |
| "learning_rate": 3.9896461574878036e-05, | |
| "loss": 0.7228, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6518218623481782, | |
| "grad_norm": 1.2117170742885421, | |
| "learning_rate": 3.989063487022416e-05, | |
| "loss": 0.723, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6558704453441295, | |
| "grad_norm": 1.3712568744794509, | |
| "learning_rate": 3.9884649120583903e-05, | |
| "loss": 0.6994, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.659919028340081, | |
| "grad_norm": 1.101182278064963, | |
| "learning_rate": 3.987850437381915e-05, | |
| "loss": 0.6943, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6639676113360324, | |
| "grad_norm": 1.3660301915408697, | |
| "learning_rate": 3.987220067906313e-05, | |
| "loss": 0.7099, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6680161943319838, | |
| "grad_norm": 0.91156020521562, | |
| "learning_rate": 3.986573808672002e-05, | |
| "loss": 0.7017, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6720647773279352, | |
| "grad_norm": 1.3734506938470623, | |
| "learning_rate": 3.9859116648464534e-05, | |
| "loss": 0.6867, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6761133603238867, | |
| "grad_norm": 0.9262119072703058, | |
| "learning_rate": 3.985233641724152e-05, | |
| "loss": 0.6784, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.680161943319838, | |
| "grad_norm": 1.277361300881098, | |
| "learning_rate": 3.984539744726554e-05, | |
| "loss": 0.7165, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 1.085366218739523, | |
| "learning_rate": 3.983829979402038e-05, | |
| "loss": 0.7074, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6882591093117408, | |
| "grad_norm": 1.1112867814909375, | |
| "learning_rate": 3.983104351425871e-05, | |
| "loss": 0.6834, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 1.209068016870128, | |
| "learning_rate": 3.982362866600154e-05, | |
| "loss": 0.698, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6963562753036437, | |
| "grad_norm": 1.067787236775526, | |
| "learning_rate": 3.9816055308537806e-05, | |
| "loss": 0.7267, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7004048582995951, | |
| "grad_norm": 1.2912825992225403, | |
| "learning_rate": 3.980832350242388e-05, | |
| "loss": 0.7066, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7044534412955465, | |
| "grad_norm": 1.0409620351474573, | |
| "learning_rate": 3.980043330948306e-05, | |
| "loss": 0.6956, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708502024291498, | |
| "grad_norm": 1.1202390032988538, | |
| "learning_rate": 3.9792384792805124e-05, | |
| "loss": 0.689, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7125506072874493, | |
| "grad_norm": 1.1013614762371824, | |
| "learning_rate": 3.978417801674579e-05, | |
| "loss": 0.679, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7165991902834008, | |
| "grad_norm": 0.9330342577629378, | |
| "learning_rate": 3.9775813046926244e-05, | |
| "loss": 0.6906, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7206477732793523, | |
| "grad_norm": 1.2685416605554474, | |
| "learning_rate": 3.976728995023254e-05, | |
| "loss": 0.6835, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7246963562753036, | |
| "grad_norm": 0.9163629852035429, | |
| "learning_rate": 3.975860879481514e-05, | |
| "loss": 0.6723, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.728744939271255, | |
| "grad_norm": 1.2537361193214511, | |
| "learning_rate": 3.974976965008833e-05, | |
| "loss": 0.7042, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7327935222672065, | |
| "grad_norm": 0.9134656975654133, | |
| "learning_rate": 3.9740772586729674e-05, | |
| "loss": 0.7029, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 1.0017485332864073, | |
| "learning_rate": 3.973161767667946e-05, | |
| "loss": 0.6847, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7408906882591093, | |
| "grad_norm": 0.8493384495927144, | |
| "learning_rate": 3.972230499314009e-05, | |
| "loss": 0.6872, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7449392712550608, | |
| "grad_norm": 0.9013224953347614, | |
| "learning_rate": 3.9712834610575545e-05, | |
| "loss": 0.6826, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7489878542510121, | |
| "grad_norm": 0.8385032032667946, | |
| "learning_rate": 3.9703206604710746e-05, | |
| "loss": 0.672, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7530364372469636, | |
| "grad_norm": 0.9876240232739915, | |
| "learning_rate": 3.969342105253098e-05, | |
| "loss": 0.6613, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.757085020242915, | |
| "grad_norm": 0.7955167495647031, | |
| "learning_rate": 3.9683478032281246e-05, | |
| "loss": 0.6484, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7611336032388664, | |
| "grad_norm": 1.0122196584174195, | |
| "learning_rate": 3.967337762346568e-05, | |
| "loss": 0.6977, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7651821862348178, | |
| "grad_norm": 0.9190766178006726, | |
| "learning_rate": 3.9663119906846885e-05, | |
| "loss": 0.6676, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.8426876120471108, | |
| "learning_rate": 3.965270496444528e-05, | |
| "loss": 0.6847, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7732793522267206, | |
| "grad_norm": 0.9155453705086761, | |
| "learning_rate": 3.9642132879538465e-05, | |
| "loss": 0.6618, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.7975653828020102, | |
| "learning_rate": 3.9631403736660566e-05, | |
| "loss": 0.678, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7813765182186235, | |
| "grad_norm": 0.7573054407326352, | |
| "learning_rate": 3.962051762160151e-05, | |
| "loss": 0.6539, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7854251012145749, | |
| "grad_norm": 0.90286167611589, | |
| "learning_rate": 3.960947462140638e-05, | |
| "loss": 0.6881, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.8202045188597511, | |
| "learning_rate": 3.959827482437472e-05, | |
| "loss": 0.6803, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7935222672064778, | |
| "grad_norm": 0.8585516779832999, | |
| "learning_rate": 3.958691832005981e-05, | |
| "loss": 0.6594, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7975708502024291, | |
| "grad_norm": 0.9355654548866901, | |
| "learning_rate": 3.9575405199267944e-05, | |
| "loss": 0.6605, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8016194331983806, | |
| "grad_norm": 0.8323170601585129, | |
| "learning_rate": 3.956373555405774e-05, | |
| "loss": 0.6644, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.805668016194332, | |
| "grad_norm": 0.9039229638338878, | |
| "learning_rate": 3.955190947773937e-05, | |
| "loss": 0.6694, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8097165991902834, | |
| "grad_norm": 0.8013428381189868, | |
| "learning_rate": 3.9539927064873824e-05, | |
| "loss": 0.6778, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8137651821862348, | |
| "grad_norm": 0.7996628500390034, | |
| "learning_rate": 3.952778841127214e-05, | |
| "loss": 0.6711, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8178137651821862, | |
| "grad_norm": 0.8687270539581922, | |
| "learning_rate": 3.95154936139947e-05, | |
| "loss": 0.6695, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8218623481781376, | |
| "grad_norm": 1.1124106247215944, | |
| "learning_rate": 3.950304277135034e-05, | |
| "loss": 0.6331, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8259109311740891, | |
| "grad_norm": 0.8231000162685119, | |
| "learning_rate": 3.94904359828957e-05, | |
| "loss": 0.6565, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8299595141700404, | |
| "grad_norm": 0.8946455083930984, | |
| "learning_rate": 3.94776733494343e-05, | |
| "loss": 0.6715, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8340080971659919, | |
| "grad_norm": 1.005716962200229, | |
| "learning_rate": 3.9464754973015826e-05, | |
| "loss": 0.6673, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8380566801619433, | |
| "grad_norm": 0.7160090082354215, | |
| "learning_rate": 3.9451680956935275e-05, | |
| "loss": 0.6934, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.8433255978659079, | |
| "learning_rate": 3.943845140573213e-05, | |
| "loss": 0.6488, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.7671586609945361, | |
| "learning_rate": 3.942506642518952e-05, | |
| "loss": 0.6581, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8502024291497976, | |
| "grad_norm": 0.8582767346372916, | |
| "learning_rate": 3.9411526122333396e-05, | |
| "loss": 0.6791, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.854251012145749, | |
| "grad_norm": 0.880995522727754, | |
| "learning_rate": 3.939783060543167e-05, | |
| "loss": 0.6787, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8582995951417004, | |
| "grad_norm": 0.7988474757664359, | |
| "learning_rate": 3.938397998399333e-05, | |
| "loss": 0.6804, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8623481781376519, | |
| "grad_norm": 0.8387514515298007, | |
| "learning_rate": 3.936997436876756e-05, | |
| "loss": 0.6484, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8663967611336032, | |
| "grad_norm": 0.7568333212503778, | |
| "learning_rate": 3.935581387174289e-05, | |
| "loss": 0.6607, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8704453441295547, | |
| "grad_norm": 0.7628074933166019, | |
| "learning_rate": 3.9341498606146295e-05, | |
| "loss": 0.6476, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8744939271255061, | |
| "grad_norm": 0.8013359922534841, | |
| "learning_rate": 3.932702868644224e-05, | |
| "loss": 0.6478, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8785425101214575, | |
| "grad_norm": 0.8005468135404453, | |
| "learning_rate": 3.931240422833181e-05, | |
| "loss": 0.6489, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8825910931174089, | |
| "grad_norm": 0.8226699363408041, | |
| "learning_rate": 3.9297625348751806e-05, | |
| "loss": 0.6692, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8866396761133604, | |
| "grad_norm": 0.811896899725631, | |
| "learning_rate": 3.9282692165873744e-05, | |
| "loss": 0.6457, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8906882591093117, | |
| "grad_norm": 0.7809200213586357, | |
| "learning_rate": 3.9267604799102945e-05, | |
| "loss": 0.6784, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.8776919494575084, | |
| "learning_rate": 3.9252363369077615e-05, | |
| "loss": 0.6287, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8987854251012146, | |
| "grad_norm": 0.9711695146270455, | |
| "learning_rate": 3.9236967997667796e-05, | |
| "loss": 0.6265, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.902834008097166, | |
| "grad_norm": 0.7666498489167548, | |
| "learning_rate": 3.9221418807974496e-05, | |
| "loss": 0.6466, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9068825910931174, | |
| "grad_norm": 0.9298992164795116, | |
| "learning_rate": 3.92057159243286e-05, | |
| "loss": 0.6455, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9109311740890689, | |
| "grad_norm": 0.813005462905074, | |
| "learning_rate": 3.9189859472289956e-05, | |
| "loss": 0.6703, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9149797570850202, | |
| "grad_norm": 0.7303344049862439, | |
| "learning_rate": 3.9173849578646324e-05, | |
| "loss": 0.6335, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9190283400809717, | |
| "grad_norm": 0.786607387780103, | |
| "learning_rate": 3.9157686371412395e-05, | |
| "loss": 0.652, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.7511467381336169, | |
| "learning_rate": 3.914136997982874e-05, | |
| "loss": 0.6728, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9271255060728745, | |
| "grad_norm": 0.8761917668444121, | |
| "learning_rate": 3.912490053436079e-05, | |
| "loss": 0.6195, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9311740890688259, | |
| "grad_norm": 0.7224100448274569, | |
| "learning_rate": 3.9108278166697805e-05, | |
| "loss": 0.627, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9352226720647774, | |
| "grad_norm": 0.7511152211317839, | |
| "learning_rate": 3.909150300975178e-05, | |
| "loss": 0.6099, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9392712550607287, | |
| "grad_norm": 0.812237682308071, | |
| "learning_rate": 3.907457519765642e-05, | |
| "loss": 0.642, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9433198380566802, | |
| "grad_norm": 0.7799960604477554, | |
| "learning_rate": 3.905749486576607e-05, | |
| "loss": 0.6377, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.9768533386339057, | |
| "learning_rate": 3.90402621506546e-05, | |
| "loss": 0.6608, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.951417004048583, | |
| "grad_norm": 0.8144557373984253, | |
| "learning_rate": 3.902287719011435e-05, | |
| "loss": 0.6568, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9554655870445344, | |
| "grad_norm": 0.8260567265802632, | |
| "learning_rate": 3.900534012315499e-05, | |
| "loss": 0.6234, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9595141700404858, | |
| "grad_norm": 0.7554052279807324, | |
| "learning_rate": 3.8987651090002454e-05, | |
| "loss": 0.6293, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9635627530364372, | |
| "grad_norm": 0.8482841626487434, | |
| "learning_rate": 3.896981023209777e-05, | |
| "loss": 0.6443, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9676113360323887, | |
| "grad_norm": 0.740868513709928, | |
| "learning_rate": 3.8951817692095966e-05, | |
| "loss": 0.6182, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.97165991902834, | |
| "grad_norm": 0.7459937262741059, | |
| "learning_rate": 3.893367361386491e-05, | |
| "loss": 0.6327, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9757085020242915, | |
| "grad_norm": 0.7692196746844617, | |
| "learning_rate": 3.891537814248417e-05, | |
| "loss": 0.6128, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.979757085020243, | |
| "grad_norm": 1.0195404451444292, | |
| "learning_rate": 3.889693142424384e-05, | |
| "loss": 0.6397, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9838056680161943, | |
| "grad_norm": 0.7646556762786737, | |
| "learning_rate": 3.8878333606643405e-05, | |
| "loss": 0.6135, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9878542510121457, | |
| "grad_norm": 0.9993448547170458, | |
| "learning_rate": 3.885958483839049e-05, | |
| "loss": 0.6325, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9919028340080972, | |
| "grad_norm": 0.7951511842737364, | |
| "learning_rate": 3.8840685269399786e-05, | |
| "loss": 0.6321, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9959514170040485, | |
| "grad_norm": 0.7888293808671422, | |
| "learning_rate": 3.882163505079171e-05, | |
| "loss": 0.5738, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7225367903004931, | |
| "learning_rate": 3.880243433489132e-05, | |
| "loss": 0.6175, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0040485829959513, | |
| "grad_norm": 0.9748201914680213, | |
| "learning_rate": 3.878308327522702e-05, | |
| "loss": 0.4582, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.008097165991903, | |
| "grad_norm": 0.993623025232321, | |
| "learning_rate": 3.876358202652936e-05, | |
| "loss": 0.4426, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0121457489878543, | |
| "grad_norm": 0.9243762785899933, | |
| "learning_rate": 3.8743930744729835e-05, | |
| "loss": 0.4384, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0161943319838056, | |
| "grad_norm": 0.956310546650985, | |
| "learning_rate": 3.8724129586959535e-05, | |
| "loss": 0.4381, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0202429149797572, | |
| "grad_norm": 0.9664449071464359, | |
| "learning_rate": 3.870417871154801e-05, | |
| "loss": 0.4404, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0242914979757085, | |
| "grad_norm": 0.7539226918500501, | |
| "learning_rate": 3.868407827802192e-05, | |
| "loss": 0.4166, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0283400809716599, | |
| "grad_norm": 0.9165531287081911, | |
| "learning_rate": 3.8663828447103804e-05, | |
| "loss": 0.4633, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0323886639676114, | |
| "grad_norm": 0.7292258145855004, | |
| "learning_rate": 3.8643429380710765e-05, | |
| "loss": 0.4384, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0364372469635628, | |
| "grad_norm": 0.900601972489518, | |
| "learning_rate": 3.862288124195319e-05, | |
| "loss": 0.441, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.040485829959514, | |
| "grad_norm": 0.795110205548103, | |
| "learning_rate": 3.8602184195133464e-05, | |
| "loss": 0.442, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0445344129554657, | |
| "grad_norm": 0.6949364070905548, | |
| "learning_rate": 3.8581338405744604e-05, | |
| "loss": 0.4261, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.048582995951417, | |
| "grad_norm": 0.8607135258099478, | |
| "learning_rate": 3.8560344040468996e-05, | |
| "loss": 0.4705, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.6993516507119473, | |
| "learning_rate": 3.853920126717703e-05, | |
| "loss": 0.4365, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.05668016194332, | |
| "grad_norm": 0.8347362189070607, | |
| "learning_rate": 3.851791025492574e-05, | |
| "loss": 0.4477, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0607287449392713, | |
| "grad_norm": 0.7772431565664104, | |
| "learning_rate": 3.8496471173957515e-05, | |
| "loss": 0.409, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0647773279352226, | |
| "grad_norm": 1.01365561855499, | |
| "learning_rate": 3.847488419569866e-05, | |
| "loss": 0.4448, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0688259109311742, | |
| "grad_norm": 0.6947352030335455, | |
| "learning_rate": 3.8453149492758077e-05, | |
| "loss": 0.4261, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0728744939271255, | |
| "grad_norm": 0.9941569192388304, | |
| "learning_rate": 3.8431267238925886e-05, | |
| "loss": 0.4199, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.6793098785506314, | |
| "learning_rate": 3.840923760917199e-05, | |
| "loss": 0.4201, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0809716599190284, | |
| "grad_norm": 0.8454753444694649, | |
| "learning_rate": 3.838706077964473e-05, | |
| "loss": 0.4158, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0850202429149798, | |
| "grad_norm": 0.8171542888671784, | |
| "learning_rate": 3.836473692766944e-05, | |
| "loss": 0.4283, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0890688259109311, | |
| "grad_norm": 0.6723779874936695, | |
| "learning_rate": 3.834226623174705e-05, | |
| "loss": 0.4211, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0931174089068827, | |
| "grad_norm": 0.7747075218244401, | |
| "learning_rate": 3.831964887155264e-05, | |
| "loss": 0.4419, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.097165991902834, | |
| "grad_norm": 0.8042420011213908, | |
| "learning_rate": 3.829688502793403e-05, | |
| "loss": 0.3853, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.1012145748987854, | |
| "grad_norm": 0.8913746193261449, | |
| "learning_rate": 3.827397488291029e-05, | |
| "loss": 0.451, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.7478960001306808, | |
| "learning_rate": 3.8250918619670343e-05, | |
| "loss": 0.4151, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.1093117408906883, | |
| "grad_norm": 0.9351723241689474, | |
| "learning_rate": 3.822771642257145e-05, | |
| "loss": 0.4259, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1133603238866396, | |
| "grad_norm": 0.6500692255480931, | |
| "learning_rate": 3.820436847713776e-05, | |
| "loss": 0.4139, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.117408906882591, | |
| "grad_norm": 0.7458464714453845, | |
| "learning_rate": 3.8180874970058826e-05, | |
| "loss": 0.418, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.1214574898785425, | |
| "grad_norm": 0.7293491285405405, | |
| "learning_rate": 3.8157236089188103e-05, | |
| "loss": 0.4122, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.125506072874494, | |
| "grad_norm": 0.7573780271514868, | |
| "learning_rate": 3.813345202354145e-05, | |
| "loss": 0.4388, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1295546558704452, | |
| "grad_norm": 0.6492534223221046, | |
| "learning_rate": 3.810952296329563e-05, | |
| "loss": 0.4116, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1336032388663968, | |
| "grad_norm": 0.6677324220483424, | |
| "learning_rate": 3.808544909978676e-05, | |
| "loss": 0.4287, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1376518218623481, | |
| "grad_norm": 0.7466108559386191, | |
| "learning_rate": 3.806123062550882e-05, | |
| "loss": 0.4396, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1417004048582995, | |
| "grad_norm": 0.7104330333627169, | |
| "learning_rate": 3.803686773411208e-05, | |
| "loss": 0.4392, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.145748987854251, | |
| "grad_norm": 0.7688990910427622, | |
| "learning_rate": 3.801236062040158e-05, | |
| "loss": 0.4101, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1497975708502024, | |
| "grad_norm": 0.7476415941299864, | |
| "learning_rate": 3.798770948033553e-05, | |
| "loss": 0.4181, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.7175008885124321, | |
| "learning_rate": 3.796291451102381e-05, | |
| "loss": 0.4195, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.8383689882133241, | |
| "learning_rate": 3.793797591072631e-05, | |
| "loss": 0.4294, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1619433198380567, | |
| "grad_norm": 0.6815637361602364, | |
| "learning_rate": 3.791289387885144e-05, | |
| "loss": 0.4256, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.165991902834008, | |
| "grad_norm": 0.8002000435472675, | |
| "learning_rate": 3.788766861595443e-05, | |
| "loss": 0.443, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1700404858299596, | |
| "grad_norm": 0.7200346326317898, | |
| "learning_rate": 3.7862300323735835e-05, | |
| "loss": 0.4036, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.174089068825911, | |
| "grad_norm": 0.7004993992251936, | |
| "learning_rate": 3.783678920503983e-05, | |
| "loss": 0.4036, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1781376518218623, | |
| "grad_norm": 0.6968100562242444, | |
| "learning_rate": 3.781113546385265e-05, | |
| "loss": 0.3979, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1821862348178138, | |
| "grad_norm": 0.8055202626177287, | |
| "learning_rate": 3.7785339305300916e-05, | |
| "loss": 0.435, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1862348178137652, | |
| "grad_norm": 0.9544460211148613, | |
| "learning_rate": 3.7759400935650025e-05, | |
| "loss": 0.4035, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1902834008097165, | |
| "grad_norm": 0.6932996115415344, | |
| "learning_rate": 3.77333205623025e-05, | |
| "loss": 0.4097, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.194331983805668, | |
| "grad_norm": 0.837540356393131, | |
| "learning_rate": 3.7707098393796295e-05, | |
| "loss": 0.406, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1983805668016194, | |
| "grad_norm": 0.8141251041078297, | |
| "learning_rate": 3.7680734639803185e-05, | |
| "loss": 0.4144, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2024291497975708, | |
| "grad_norm": 0.6841688738768549, | |
| "learning_rate": 3.765422951112704e-05, | |
| "loss": 0.4162, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2064777327935223, | |
| "grad_norm": 0.7463036472998463, | |
| "learning_rate": 3.762758321970216e-05, | |
| "loss": 0.3996, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.766011538460647, | |
| "learning_rate": 3.7600795978591584e-05, | |
| "loss": 0.4068, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.214574898785425, | |
| "grad_norm": 0.6358735971004372, | |
| "learning_rate": 3.757386800198538e-05, | |
| "loss": 0.4436, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2186234817813766, | |
| "grad_norm": 0.7325470892734588, | |
| "learning_rate": 3.7546799505198925e-05, | |
| "loss": 0.4208, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.222672064777328, | |
| "grad_norm": 0.6822956186532906, | |
| "learning_rate": 3.7519590704671197e-05, | |
| "loss": 0.4213, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2267206477732793, | |
| "grad_norm": 0.7262050379087895, | |
| "learning_rate": 3.749224181796305e-05, | |
| "loss": 0.4215, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.7451494326633437, | |
| "learning_rate": 3.746475306375545e-05, | |
| "loss": 0.4385, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2348178137651822, | |
| "grad_norm": 0.6704807912380671, | |
| "learning_rate": 3.743712466184774e-05, | |
| "loss": 0.4379, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2388663967611335, | |
| "grad_norm": 0.8212062218625626, | |
| "learning_rate": 3.7409356833155885e-05, | |
| "loss": 0.4352, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.242914979757085, | |
| "grad_norm": 0.727420854305335, | |
| "learning_rate": 3.73814497997107e-05, | |
| "loss": 0.4249, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2469635627530364, | |
| "grad_norm": 0.8118315575244557, | |
| "learning_rate": 3.7353403784656095e-05, | |
| "loss": 0.3795, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2510121457489878, | |
| "grad_norm": 0.774687179429063, | |
| "learning_rate": 3.732521901224724e-05, | |
| "loss": 0.4327, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2550607287449393, | |
| "grad_norm": 0.6782778352703934, | |
| "learning_rate": 3.7296895707848844e-05, | |
| "loss": 0.4032, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2591093117408907, | |
| "grad_norm": 0.6503266304686203, | |
| "learning_rate": 3.7268434097933275e-05, | |
| "loss": 0.4138, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.6510301783553996, | |
| "learning_rate": 3.723983441007882e-05, | |
| "loss": 0.4053, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2672064777327936, | |
| "grad_norm": 0.6888076249907793, | |
| "learning_rate": 3.721109687296781e-05, | |
| "loss": 0.4417, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.271255060728745, | |
| "grad_norm": 0.7213399363233239, | |
| "learning_rate": 3.718222171638484e-05, | |
| "loss": 0.4102, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2753036437246963, | |
| "grad_norm": 0.7303914771326256, | |
| "learning_rate": 3.715320917121488e-05, | |
| "loss": 0.4028, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2793522267206479, | |
| "grad_norm": 0.7211430633964273, | |
| "learning_rate": 3.712405946944149e-05, | |
| "loss": 0.4433, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2834008097165992, | |
| "grad_norm": 0.7017640922173776, | |
| "learning_rate": 3.709477284414488e-05, | |
| "loss": 0.4202, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2874493927125505, | |
| "grad_norm": 0.6344771850984858, | |
| "learning_rate": 3.7065349529500136e-05, | |
| "loss": 0.3872, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.291497975708502, | |
| "grad_norm": 0.7105536894885824, | |
| "learning_rate": 3.7035789760775305e-05, | |
| "loss": 0.4295, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2955465587044535, | |
| "grad_norm": 0.6810045355475922, | |
| "learning_rate": 3.70060937743295e-05, | |
| "loss": 0.377, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2995951417004048, | |
| "grad_norm": 0.6255556113326413, | |
| "learning_rate": 3.6976261807611045e-05, | |
| "loss": 0.4318, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.3036437246963564, | |
| "grad_norm": 0.6707681478707466, | |
| "learning_rate": 3.694629409915555e-05, | |
| "loss": 0.4225, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.6998395632719493, | |
| "learning_rate": 3.691619088858401e-05, | |
| "loss": 0.4235, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.311740890688259, | |
| "grad_norm": 0.7266954457424479, | |
| "learning_rate": 3.68859524166009e-05, | |
| "loss": 0.407, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.6432261847968216, | |
| "learning_rate": 3.6855578924992226e-05, | |
| "loss": 0.3937, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.319838056680162, | |
| "grad_norm": 0.6729651650070371, | |
| "learning_rate": 3.6825070656623626e-05, | |
| "loss": 0.4067, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.3238866396761133, | |
| "grad_norm": 0.6532796747970667, | |
| "learning_rate": 3.67944278554384e-05, | |
| "loss": 0.4055, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3279352226720649, | |
| "grad_norm": 0.7159352057668014, | |
| "learning_rate": 3.676365076645557e-05, | |
| "loss": 0.4187, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.6788183638014359, | |
| "learning_rate": 3.673273963576791e-05, | |
| "loss": 0.4105, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3360323886639676, | |
| "grad_norm": 0.6640123580490781, | |
| "learning_rate": 3.670169471054001e-05, | |
| "loss": 0.412, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3400809716599191, | |
| "grad_norm": 0.6828719166912848, | |
| "learning_rate": 3.6670516239006254e-05, | |
| "loss": 0.4087, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3441295546558705, | |
| "grad_norm": 0.735288533230758, | |
| "learning_rate": 3.663920447046888e-05, | |
| "loss": 0.426, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3481781376518218, | |
| "grad_norm": 0.8071501105788881, | |
| "learning_rate": 3.6607759655295954e-05, | |
| "loss": 0.3902, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3522267206477734, | |
| "grad_norm": 0.6600462188196065, | |
| "learning_rate": 3.657618204491937e-05, | |
| "loss": 0.3762, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3562753036437247, | |
| "grad_norm": 1.003717972045279, | |
| "learning_rate": 3.6544471891832875e-05, | |
| "loss": 0.3932, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.360323886639676, | |
| "grad_norm": 0.6512629989942073, | |
| "learning_rate": 3.6512629449589983e-05, | |
| "loss": 0.3929, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3643724696356276, | |
| "grad_norm": 0.8166351280381657, | |
| "learning_rate": 3.6480654972802045e-05, | |
| "loss": 0.4088, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.6532606953723291, | |
| "learning_rate": 3.644854871713611e-05, | |
| "loss": 0.4007, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3724696356275303, | |
| "grad_norm": 0.6735427699163394, | |
| "learning_rate": 3.6416310939312924e-05, | |
| "loss": 0.3801, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.376518218623482, | |
| "grad_norm": 0.7296991763372398, | |
| "learning_rate": 3.638394189710493e-05, | |
| "loss": 0.3778, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3805668016194332, | |
| "grad_norm": 0.6659796373351617, | |
| "learning_rate": 3.635144184933412e-05, | |
| "loss": 0.4041, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.7049353518658542, | |
| "learning_rate": 3.6318811055869994e-05, | |
| "loss": 0.4315, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3886639676113361, | |
| "grad_norm": 0.6188813332140634, | |
| "learning_rate": 3.628604977762752e-05, | |
| "loss": 0.3884, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3927125506072875, | |
| "grad_norm": 0.8372683219430923, | |
| "learning_rate": 3.6253158276565004e-05, | |
| "loss": 0.4611, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3967611336032388, | |
| "grad_norm": 0.6340473322181044, | |
| "learning_rate": 3.6220136815682016e-05, | |
| "loss": 0.3815, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4008097165991904, | |
| "grad_norm": 0.6463718463390473, | |
| "learning_rate": 3.618698565901727e-05, | |
| "loss": 0.413, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.4048582995951417, | |
| "grad_norm": 0.6459246833064364, | |
| "learning_rate": 3.615370507164654e-05, | |
| "loss": 0.3901, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.408906882591093, | |
| "grad_norm": 0.6972805956978931, | |
| "learning_rate": 3.61202953196805e-05, | |
| "loss": 0.3935, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.4129554655870447, | |
| "grad_norm": 0.6240512178401184, | |
| "learning_rate": 3.6086756670262654e-05, | |
| "loss": 0.4031, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.417004048582996, | |
| "grad_norm": 0.7121236051463878, | |
| "learning_rate": 3.6053089391567135e-05, | |
| "loss": 0.4031, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.6871659955120933, | |
| "learning_rate": 3.601929375279662e-05, | |
| "loss": 0.3907, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.425101214574899, | |
| "grad_norm": 0.6361719258687897, | |
| "learning_rate": 3.598537002418012e-05, | |
| "loss": 0.3735, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4291497975708503, | |
| "grad_norm": 0.6705644036548064, | |
| "learning_rate": 3.5951318476970885e-05, | |
| "loss": 0.4105, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4331983805668016, | |
| "grad_norm": 0.7385435540839745, | |
| "learning_rate": 3.591713938344416e-05, | |
| "loss": 0.3674, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4372469635627532, | |
| "grad_norm": 0.6463574334834491, | |
| "learning_rate": 3.588283301689507e-05, | |
| "loss": 0.4021, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4412955465587045, | |
| "grad_norm": 0.7380150463292473, | |
| "learning_rate": 3.5848399651636424e-05, | |
| "loss": 0.4158, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4453441295546559, | |
| "grad_norm": 0.6342068416633311, | |
| "learning_rate": 3.5813839562996473e-05, | |
| "loss": 0.3912, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4493927125506074, | |
| "grad_norm": 0.7205568825789411, | |
| "learning_rate": 3.5779153027316775e-05, | |
| "loss": 0.3696, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4534412955465588, | |
| "grad_norm": 0.6633455473916287, | |
| "learning_rate": 3.574434032194994e-05, | |
| "loss": 0.4103, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.45748987854251, | |
| "grad_norm": 0.6183963035127608, | |
| "learning_rate": 3.5709401725257436e-05, | |
| "loss": 0.3913, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.6613099340867201, | |
| "learning_rate": 3.5674337516607346e-05, | |
| "loss": 0.3696, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.465587044534413, | |
| "grad_norm": 0.648431905508475, | |
| "learning_rate": 3.563914797637216e-05, | |
| "loss": 0.3961, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4696356275303644, | |
| "grad_norm": 0.7375028348756888, | |
| "learning_rate": 3.560383338592649e-05, | |
| "loss": 0.395, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.8092324820116858, | |
| "learning_rate": 3.556839402764487e-05, | |
| "loss": 0.3684, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4777327935222673, | |
| "grad_norm": 0.6447645872796207, | |
| "learning_rate": 3.553283018489946e-05, | |
| "loss": 0.3665, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4817813765182186, | |
| "grad_norm": 0.7526371831374633, | |
| "learning_rate": 3.54971421420578e-05, | |
| "loss": 0.3904, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.48582995951417, | |
| "grad_norm": 0.6727599938544688, | |
| "learning_rate": 3.5461330184480524e-05, | |
| "loss": 0.3998, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4898785425101215, | |
| "grad_norm": 0.6668641358680505, | |
| "learning_rate": 3.54253945985191e-05, | |
| "loss": 0.4031, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4939271255060729, | |
| "grad_norm": 0.7933199323995338, | |
| "learning_rate": 3.5389335671513524e-05, | |
| "loss": 0.4, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4979757085020242, | |
| "grad_norm": 0.6704336374945502, | |
| "learning_rate": 3.5353153691789986e-05, | |
| "loss": 0.3655, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5020242914979756, | |
| "grad_norm": 0.7494734739789745, | |
| "learning_rate": 3.531684894865866e-05, | |
| "loss": 0.3641, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.5060728744939271, | |
| "grad_norm": 0.6694362510321302, | |
| "learning_rate": 3.528042173241128e-05, | |
| "loss": 0.3775, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5101214574898787, | |
| "grad_norm": 0.6650963158007247, | |
| "learning_rate": 3.524387233431892e-05, | |
| "loss": 0.4345, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.5141700404858298, | |
| "grad_norm": 0.783301480757075, | |
| "learning_rate": 3.520720104662958e-05, | |
| "loss": 0.376, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5182186234817814, | |
| "grad_norm": 0.6236107158664598, | |
| "learning_rate": 3.5170408162565904e-05, | |
| "loss": 0.3688, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.522267206477733, | |
| "grad_norm": 0.9527916715788473, | |
| "learning_rate": 3.513349397632281e-05, | |
| "loss": 0.3921, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.7565850209534924, | |
| "learning_rate": 3.5096458783065145e-05, | |
| "loss": 0.37, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5303643724696356, | |
| "grad_norm": 0.7878266275379665, | |
| "learning_rate": 3.505930287892533e-05, | |
| "loss": 0.3957, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.5344129554655872, | |
| "grad_norm": 0.6311139979298284, | |
| "learning_rate": 3.502202656100099e-05, | |
| "loss": 0.3755, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.9658564962032372, | |
| "learning_rate": 3.498463012735256e-05, | |
| "loss": 0.4001, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.54251012145749, | |
| "grad_norm": 0.6642616494778204, | |
| "learning_rate": 3.494711387700094e-05, | |
| "loss": 0.3907, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5465587044534415, | |
| "grad_norm": 0.7808584841169934, | |
| "learning_rate": 3.490947810992505e-05, | |
| "loss": 0.3735, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5506072874493926, | |
| "grad_norm": 0.6634093692537301, | |
| "learning_rate": 3.48717231270595e-05, | |
| "loss": 0.4026, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5546558704453441, | |
| "grad_norm": 0.6352043683603928, | |
| "learning_rate": 3.483384923029211e-05, | |
| "loss": 0.3659, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5587044534412957, | |
| "grad_norm": 0.7912926022127001, | |
| "learning_rate": 3.479585672246156e-05, | |
| "loss": 0.4052, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5627530364372468, | |
| "grad_norm": 0.6418028574272334, | |
| "learning_rate": 3.4757745907354924e-05, | |
| "loss": 0.3769, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5668016194331984, | |
| "grad_norm": 0.9139216619528555, | |
| "learning_rate": 3.471951708970527e-05, | |
| "loss": 0.3338, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5708502024291497, | |
| "grad_norm": 0.6025475611451261, | |
| "learning_rate": 3.468117057518921e-05, | |
| "loss": 0.3491, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.574898785425101, | |
| "grad_norm": 1.046321538873015, | |
| "learning_rate": 3.4642706670424426e-05, | |
| "loss": 0.4098, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.6052427039963192, | |
| "learning_rate": 3.460412568296731e-05, | |
| "loss": 0.3467, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.582995951417004, | |
| "grad_norm": 0.723798865874273, | |
| "learning_rate": 3.456542792131039e-05, | |
| "loss": 0.3783, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5870445344129553, | |
| "grad_norm": 0.676561299835144, | |
| "learning_rate": 3.4526613694879936e-05, | |
| "loss": 0.3907, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.591093117408907, | |
| "grad_norm": 0.6982525864070904, | |
| "learning_rate": 3.448768331403348e-05, | |
| "loss": 0.3817, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5951417004048583, | |
| "grad_norm": 0.6826294435041715, | |
| "learning_rate": 3.44486370900573e-05, | |
| "loss": 0.3603, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5991902834008096, | |
| "grad_norm": 0.625090227352634, | |
| "learning_rate": 3.440947533516396e-05, | |
| "loss": 0.3573, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.6032388663967612, | |
| "grad_norm": 0.6676344709043406, | |
| "learning_rate": 3.4370198362489816e-05, | |
| "loss": 0.3754, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.6072874493927125, | |
| "grad_norm": 0.6566449927118926, | |
| "learning_rate": 3.43308064860925e-05, | |
| "loss": 0.4001, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.6113360323886639, | |
| "grad_norm": 0.6556009991904479, | |
| "learning_rate": 3.4291300020948393e-05, | |
| "loss": 0.3544, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.7832010636196122, | |
| "learning_rate": 3.425167928295015e-05, | |
| "loss": 0.3687, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6194331983805668, | |
| "grad_norm": 0.656711542812388, | |
| "learning_rate": 3.421194458890411e-05, | |
| "loss": 0.3892, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.623481781376518, | |
| "grad_norm": 0.6893468978062148, | |
| "learning_rate": 3.4172096256527865e-05, | |
| "loss": 0.3607, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.6275303643724697, | |
| "grad_norm": 0.7191118659547315, | |
| "learning_rate": 3.4132134604447595e-05, | |
| "loss": 0.3577, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.9331330141221553, | |
| "learning_rate": 3.409205995219561e-05, | |
| "loss": 0.3767, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6356275303643724, | |
| "grad_norm": 0.6809770101601113, | |
| "learning_rate": 3.4051872620207765e-05, | |
| "loss": 0.3497, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.639676113360324, | |
| "grad_norm": 0.811073169356484, | |
| "learning_rate": 3.40115729298209e-05, | |
| "loss": 0.3766, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6437246963562753, | |
| "grad_norm": 0.6768342507485269, | |
| "learning_rate": 3.397116120327025e-05, | |
| "loss": 0.3564, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6477732793522266, | |
| "grad_norm": 0.6627576482561457, | |
| "learning_rate": 3.3930637763686916e-05, | |
| "loss": 0.3545, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6518218623481782, | |
| "grad_norm": 0.6201799575140006, | |
| "learning_rate": 3.389000293509524e-05, | |
| "loss": 0.352, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6558704453441295, | |
| "grad_norm": 0.7272368626109466, | |
| "learning_rate": 3.3849257042410225e-05, | |
| "loss": 0.3705, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6599190283400809, | |
| "grad_norm": 0.6758870179301117, | |
| "learning_rate": 3.380840041143494e-05, | |
| "loss": 0.3737, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6639676113360324, | |
| "grad_norm": 0.6872984566917261, | |
| "learning_rate": 3.376743336885792e-05, | |
| "loss": 0.3669, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6680161943319838, | |
| "grad_norm": 0.6355999245090816, | |
| "learning_rate": 3.372635624225055e-05, | |
| "loss": 0.3504, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6720647773279351, | |
| "grad_norm": 0.6311492548917269, | |
| "learning_rate": 3.3685169360064445e-05, | |
| "loss": 0.3737, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6761133603238867, | |
| "grad_norm": 0.5858591229080635, | |
| "learning_rate": 3.3643873051628804e-05, | |
| "loss": 0.3562, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.680161943319838, | |
| "grad_norm": 0.6574560950139171, | |
| "learning_rate": 3.3602467647147814e-05, | |
| "loss": 0.3704, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.5840193931642242, | |
| "learning_rate": 3.356095347769799e-05, | |
| "loss": 0.3597, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.688259109311741, | |
| "grad_norm": 0.6012991435243238, | |
| "learning_rate": 3.351933087522552e-05, | |
| "loss": 0.3529, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.7159301286882909, | |
| "learning_rate": 3.3477600172543626e-05, | |
| "loss": 0.3594, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6963562753036436, | |
| "grad_norm": 0.6808723062743018, | |
| "learning_rate": 3.3435761703329894e-05, | |
| "loss": 0.3614, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.7004048582995952, | |
| "grad_norm": 0.6389200961290191, | |
| "learning_rate": 3.3393815802123606e-05, | |
| "loss": 0.3307, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7044534412955465, | |
| "grad_norm": 0.6078305174568407, | |
| "learning_rate": 3.335176280432308e-05, | |
| "loss": 0.3716, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.708502024291498, | |
| "grad_norm": 0.7773436171440138, | |
| "learning_rate": 3.330960304618295e-05, | |
| "loss": 0.3357, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.7125506072874495, | |
| "grad_norm": 0.6107914957155542, | |
| "learning_rate": 3.326733686481153e-05, | |
| "loss": 0.3437, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.7165991902834008, | |
| "grad_norm": 0.7834581246028476, | |
| "learning_rate": 3.3224964598168084e-05, | |
| "loss": 0.3639, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.6081499369088955, | |
| "learning_rate": 3.318248658506012e-05, | |
| "loss": 0.3701, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7246963562753037, | |
| "grad_norm": 0.6505073329190354, | |
| "learning_rate": 3.313990316514069e-05, | |
| "loss": 0.351, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.728744939271255, | |
| "grad_norm": 0.6175520846030192, | |
| "learning_rate": 3.309721467890571e-05, | |
| "loss": 0.3587, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.7327935222672064, | |
| "grad_norm": 0.6189117415322185, | |
| "learning_rate": 3.305442146769115e-05, | |
| "loss": 0.3615, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.6166873984705731, | |
| "learning_rate": 3.30115238736704e-05, | |
| "loss": 0.3439, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7408906882591093, | |
| "grad_norm": 0.6489028304811535, | |
| "learning_rate": 3.296852223985148e-05, | |
| "loss": 0.3266, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7449392712550607, | |
| "grad_norm": 0.7548984921620322, | |
| "learning_rate": 3.2925416910074315e-05, | |
| "loss": 0.3662, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7489878542510122, | |
| "grad_norm": 0.6496804115799346, | |
| "learning_rate": 3.288220822900796e-05, | |
| "loss": 0.3701, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7530364372469636, | |
| "grad_norm": 0.9126263615483103, | |
| "learning_rate": 3.283889654214788e-05, | |
| "loss": 0.3272, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.757085020242915, | |
| "grad_norm": 0.6209751206566996, | |
| "learning_rate": 3.279548219581317e-05, | |
| "loss": 0.3339, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7611336032388665, | |
| "grad_norm": 0.9198176527241435, | |
| "learning_rate": 3.275196553714379e-05, | |
| "loss": 0.3556, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7651821862348178, | |
| "grad_norm": 0.6746128180711853, | |
| "learning_rate": 3.270834691409778e-05, | |
| "loss": 0.3505, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.6799762568433642, | |
| "learning_rate": 3.266462667544849e-05, | |
| "loss": 0.3554, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7732793522267207, | |
| "grad_norm": 0.6735358876189036, | |
| "learning_rate": 3.262080517078178e-05, | |
| "loss": 0.4067, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.777327935222672, | |
| "grad_norm": 0.6206620181883336, | |
| "learning_rate": 3.257688275049323e-05, | |
| "loss": 0.3532, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7813765182186234, | |
| "grad_norm": 0.6295746287095364, | |
| "learning_rate": 3.253285976578535e-05, | |
| "loss": 0.3355, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.785425101214575, | |
| "grad_norm": 0.6905452232093593, | |
| "learning_rate": 3.2488736568664756e-05, | |
| "loss": 0.3314, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.5878584077986101, | |
| "learning_rate": 3.244451351193935e-05, | |
| "loss": 0.3322, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7935222672064777, | |
| "grad_norm": 0.6954617249662703, | |
| "learning_rate": 3.240019094921551e-05, | |
| "loss": 0.3297, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7975708502024292, | |
| "grad_norm": 0.6782089954439077, | |
| "learning_rate": 3.235576923489527e-05, | |
| "loss": 0.3314, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.8016194331983806, | |
| "grad_norm": 0.7205514723548773, | |
| "learning_rate": 3.2311248724173476e-05, | |
| "loss": 0.3265, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.805668016194332, | |
| "grad_norm": 0.7977910802029401, | |
| "learning_rate": 3.226662977303494e-05, | |
| "loss": 0.3582, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.8097165991902835, | |
| "grad_norm": 0.7518140053231699, | |
| "learning_rate": 3.2221912738251626e-05, | |
| "loss": 0.3334, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8137651821862348, | |
| "grad_norm": 0.6544680163346921, | |
| "learning_rate": 3.2177097977379736e-05, | |
| "loss": 0.3505, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.8178137651821862, | |
| "grad_norm": 0.5997187729959513, | |
| "learning_rate": 3.21321858487569e-05, | |
| "loss": 0.3262, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8218623481781377, | |
| "grad_norm": 0.6102387352800385, | |
| "learning_rate": 3.208717671149931e-05, | |
| "loss": 0.3175, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.825910931174089, | |
| "grad_norm": 0.5904403529369553, | |
| "learning_rate": 3.204207092549882e-05, | |
| "loss": 0.3259, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.8299595141700404, | |
| "grad_norm": 0.6138533309729416, | |
| "learning_rate": 3.19968688514201e-05, | |
| "loss": 0.3616, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.834008097165992, | |
| "grad_norm": 0.5738743173857487, | |
| "learning_rate": 3.195157085069774e-05, | |
| "loss": 0.3455, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8380566801619433, | |
| "grad_norm": 0.6297017107209488, | |
| "learning_rate": 3.190617728553332e-05, | |
| "loss": 0.3654, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.6226534652435004, | |
| "learning_rate": 3.1860688518892606e-05, | |
| "loss": 0.353, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.6487486032864627, | |
| "learning_rate": 3.181510491450254e-05, | |
| "loss": 0.3461, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8502024291497976, | |
| "grad_norm": 0.6177569974726697, | |
| "learning_rate": 3.176942683684842e-05, | |
| "loss": 0.3319, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.854251012145749, | |
| "grad_norm": 0.6237352831192434, | |
| "learning_rate": 3.1723654651170934e-05, | |
| "loss": 0.3471, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8582995951417005, | |
| "grad_norm": 0.6471889766590868, | |
| "learning_rate": 3.167778872346327e-05, | |
| "loss": 0.3152, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8623481781376519, | |
| "grad_norm": 0.6366545850709868, | |
| "learning_rate": 3.163182942046815e-05, | |
| "loss": 0.3227, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8663967611336032, | |
| "grad_norm": 0.7069022957357695, | |
| "learning_rate": 3.158577710967494e-05, | |
| "loss": 0.3257, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8704453441295548, | |
| "grad_norm": 0.5488910592834189, | |
| "learning_rate": 3.153963215931669e-05, | |
| "loss": 0.323, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.874493927125506, | |
| "grad_norm": 0.6212834404041688, | |
| "learning_rate": 3.14933949383672e-05, | |
| "loss": 0.3164, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8785425101214575, | |
| "grad_norm": 0.5738873203261604, | |
| "learning_rate": 3.144706581653805e-05, | |
| "loss": 0.3333, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.882591093117409, | |
| "grad_norm": 0.5911230573756899, | |
| "learning_rate": 3.140064516427566e-05, | |
| "loss": 0.3205, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8866396761133604, | |
| "grad_norm": 0.5535783620972551, | |
| "learning_rate": 3.135413335275833e-05, | |
| "loss": 0.3181, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8906882591093117, | |
| "grad_norm": 0.5992721522987052, | |
| "learning_rate": 3.130753075389327e-05, | |
| "loss": 0.3123, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.6255037103099289, | |
| "learning_rate": 3.1260837740313625e-05, | |
| "loss": 0.3381, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8987854251012146, | |
| "grad_norm": 0.6031869794941884, | |
| "learning_rate": 3.121405468537547e-05, | |
| "loss": 0.3497, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.902834008097166, | |
| "grad_norm": 0.6346486842804068, | |
| "learning_rate": 3.116718196315488e-05, | |
| "loss": 0.2908, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9068825910931175, | |
| "grad_norm": 0.625833280506256, | |
| "learning_rate": 3.1120219948444894e-05, | |
| "loss": 0.3375, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.9109311740890689, | |
| "grad_norm": 0.5942972057023159, | |
| "learning_rate": 3.107316901675252e-05, | |
| "loss": 0.3457, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.9149797570850202, | |
| "grad_norm": 0.607552009242185, | |
| "learning_rate": 3.1026029544295745e-05, | |
| "loss": 0.3312, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.9190283400809718, | |
| "grad_norm": 0.6449431896288145, | |
| "learning_rate": 3.097880190800057e-05, | |
| "loss": 0.3231, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.6210608996457025, | |
| "learning_rate": 3.093148648549788e-05, | |
| "loss": 0.3301, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9271255060728745, | |
| "grad_norm": 0.8011105611863388, | |
| "learning_rate": 3.088408365512055e-05, | |
| "loss": 0.341, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.931174089068826, | |
| "grad_norm": 0.6878746343247292, | |
| "learning_rate": 3.083659379590034e-05, | |
| "loss": 0.3303, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.9352226720647774, | |
| "grad_norm": 0.7126761343699256, | |
| "learning_rate": 3.0789017287564913e-05, | |
| "loss": 0.3148, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9392712550607287, | |
| "grad_norm": 0.6141059483611218, | |
| "learning_rate": 3.074135451053475e-05, | |
| "loss": 0.3032, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9433198380566803, | |
| "grad_norm": 0.7708913837857332, | |
| "learning_rate": 3.069360584592015e-05, | |
| "loss": 0.3199, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.5808456752072106, | |
| "learning_rate": 3.064577167551816e-05, | |
| "loss": 0.3084, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.951417004048583, | |
| "grad_norm": 0.7623494556498668, | |
| "learning_rate": 3.059785238180953e-05, | |
| "loss": 0.3346, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9554655870445345, | |
| "grad_norm": 0.5765583295203529, | |
| "learning_rate": 3.054984834795566e-05, | |
| "loss": 0.3183, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9595141700404857, | |
| "grad_norm": 0.7257996002970939, | |
| "learning_rate": 3.0501759957795537e-05, | |
| "loss": 0.3034, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9635627530364372, | |
| "grad_norm": 0.5985262213716662, | |
| "learning_rate": 3.0453587595842627e-05, | |
| "loss": 0.322, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9676113360323888, | |
| "grad_norm": 0.6204231481085439, | |
| "learning_rate": 3.0405331647281856e-05, | |
| "loss": 0.3235, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.97165991902834, | |
| "grad_norm": 0.6018425206120479, | |
| "learning_rate": 3.0356992497966508e-05, | |
| "loss": 0.3195, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9757085020242915, | |
| "grad_norm": 0.5789881146671522, | |
| "learning_rate": 3.0308570534415137e-05, | |
| "loss": 0.2977, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.979757085020243, | |
| "grad_norm": 0.5834097923011218, | |
| "learning_rate": 3.0260066143808467e-05, | |
| "loss": 0.3184, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9838056680161942, | |
| "grad_norm": 0.5765375750439212, | |
| "learning_rate": 3.021147971398632e-05, | |
| "loss": 0.29, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9878542510121457, | |
| "grad_norm": 0.5931341261525634, | |
| "learning_rate": 3.0162811633444478e-05, | |
| "loss": 0.2878, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9919028340080973, | |
| "grad_norm": 0.7404139135900863, | |
| "learning_rate": 3.0114062291331635e-05, | |
| "loss": 0.3375, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9959514170040484, | |
| "grad_norm": 0.6505218455747304, | |
| "learning_rate": 3.0065232077446217e-05, | |
| "loss": 0.3178, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6944398149616482, | |
| "learning_rate": 3.001632138223332e-05, | |
| "loss": 0.3156, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.0040485829959516, | |
| "grad_norm": 0.5446878848326366, | |
| "learning_rate": 2.9967330596781565e-05, | |
| "loss": 0.1518, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0080971659919027, | |
| "grad_norm": 0.49678500055816754, | |
| "learning_rate": 2.9918260112819952e-05, | |
| "loss": 0.1534, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0121457489878543, | |
| "grad_norm": 0.5957171634257802, | |
| "learning_rate": 2.9869110322714778e-05, | |
| "loss": 0.1598, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.016194331983806, | |
| "grad_norm": 0.5328140338079657, | |
| "learning_rate": 2.9819881619466447e-05, | |
| "loss": 0.1482, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.020242914979757, | |
| "grad_norm": 0.5102762347224472, | |
| "learning_rate": 2.9770574396706364e-05, | |
| "loss": 0.1526, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0242914979757085, | |
| "grad_norm": 0.5002578368559704, | |
| "learning_rate": 2.972118904869376e-05, | |
| "loss": 0.1316, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02834008097166, | |
| "grad_norm": 0.5226464067377673, | |
| "learning_rate": 2.9671725970312563e-05, | |
| "loss": 0.1425, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.032388663967611, | |
| "grad_norm": 0.49637629952599416, | |
| "learning_rate": 2.9622185557068238e-05, | |
| "loss": 0.1592, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.0364372469635628, | |
| "grad_norm": 0.5049287777897218, | |
| "learning_rate": 2.95725682050846e-05, | |
| "loss": 0.1401, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0404858299595143, | |
| "grad_norm": 0.5271575296911113, | |
| "learning_rate": 2.9522874311100684e-05, | |
| "loss": 0.149, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0445344129554655, | |
| "grad_norm": 0.5225359399437434, | |
| "learning_rate": 2.9473104272467526e-05, | |
| "loss": 0.1571, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.048582995951417, | |
| "grad_norm": 0.4865925351233659, | |
| "learning_rate": 2.942325848714505e-05, | |
| "loss": 0.1483, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.49711946945830887, | |
| "learning_rate": 2.9373337353698804e-05, | |
| "loss": 0.1587, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.0566801619433197, | |
| "grad_norm": 0.5051504699400007, | |
| "learning_rate": 2.932334127129686e-05, | |
| "loss": 0.1375, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.0607287449392713, | |
| "grad_norm": 0.464607303759286, | |
| "learning_rate": 2.9273270639706547e-05, | |
| "loss": 0.14, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.064777327935223, | |
| "grad_norm": 0.49301774575349305, | |
| "learning_rate": 2.922312585929131e-05, | |
| "loss": 0.1522, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.068825910931174, | |
| "grad_norm": 0.5103365710466667, | |
| "learning_rate": 2.9172907331007478e-05, | |
| "loss": 0.1509, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.0728744939271255, | |
| "grad_norm": 0.49412370013437795, | |
| "learning_rate": 2.9122615456401053e-05, | |
| "loss": 0.1594, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.4602633295716823, | |
| "learning_rate": 2.9072250637604535e-05, | |
| "loss": 0.1268, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.080971659919028, | |
| "grad_norm": 0.4857912858562402, | |
| "learning_rate": 2.9021813277333667e-05, | |
| "loss": 0.1405, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.08502024291498, | |
| "grad_norm": 0.4719975487418815, | |
| "learning_rate": 2.8971303778884244e-05, | |
| "loss": 0.1539, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.0890688259109313, | |
| "grad_norm": 0.48793505773151696, | |
| "learning_rate": 2.8920722546128863e-05, | |
| "loss": 0.1378, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0931174089068825, | |
| "grad_norm": 0.4772420524489314, | |
| "learning_rate": 2.887006998351372e-05, | |
| "loss": 0.1526, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.097165991902834, | |
| "grad_norm": 0.5130614075264299, | |
| "learning_rate": 2.881934649605535e-05, | |
| "loss": 0.1551, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.1012145748987856, | |
| "grad_norm": 0.4777059751001841, | |
| "learning_rate": 2.8768552489337408e-05, | |
| "loss": 0.1382, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.5106552509847153, | |
| "learning_rate": 2.8717688369507425e-05, | |
| "loss": 0.1297, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1093117408906883, | |
| "grad_norm": 0.5061461344776289, | |
| "learning_rate": 2.8666754543273535e-05, | |
| "loss": 0.1632, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.11336032388664, | |
| "grad_norm": 0.47438937839206946, | |
| "learning_rate": 2.8615751417901273e-05, | |
| "loss": 0.1305, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.117408906882591, | |
| "grad_norm": 0.5182441570091887, | |
| "learning_rate": 2.856467940121025e-05, | |
| "loss": 0.1562, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.1214574898785425, | |
| "grad_norm": 0.4731133554490589, | |
| "learning_rate": 2.8513538901570964e-05, | |
| "loss": 0.1293, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.125506072874494, | |
| "grad_norm": 0.49548934223563634, | |
| "learning_rate": 2.846233032790149e-05, | |
| "loss": 0.1625, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1295546558704452, | |
| "grad_norm": 0.4793008475061938, | |
| "learning_rate": 2.84110540896642e-05, | |
| "loss": 0.1343, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.133603238866397, | |
| "grad_norm": 0.484338428271236, | |
| "learning_rate": 2.8359710596862555e-05, | |
| "loss": 0.1431, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.1376518218623484, | |
| "grad_norm": 0.4650298844627208, | |
| "learning_rate": 2.8308300260037734e-05, | |
| "loss": 0.1326, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.1417004048582995, | |
| "grad_norm": 0.48309002016285585, | |
| "learning_rate": 2.8256823490265424e-05, | |
| "loss": 0.1404, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.145748987854251, | |
| "grad_norm": 0.4698520505466549, | |
| "learning_rate": 2.8205280699152515e-05, | |
| "loss": 0.1279, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1497975708502026, | |
| "grad_norm": 0.4695052840261044, | |
| "learning_rate": 2.8153672298833777e-05, | |
| "loss": 0.1511, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.4968278279926343, | |
| "learning_rate": 2.8101998701968615e-05, | |
| "loss": 0.1504, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.49643729448595736, | |
| "learning_rate": 2.805026032173773e-05, | |
| "loss": 0.1487, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.161943319838057, | |
| "grad_norm": 0.47588255772410004, | |
| "learning_rate": 2.7998457571839832e-05, | |
| "loss": 0.1393, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.165991902834008, | |
| "grad_norm": 0.47189941712881145, | |
| "learning_rate": 2.7946590866488322e-05, | |
| "loss": 0.1462, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1700404858299596, | |
| "grad_norm": 0.482511811080346, | |
| "learning_rate": 2.7894660620408e-05, | |
| "loss": 0.1353, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.174089068825911, | |
| "grad_norm": 0.4902086218894055, | |
| "learning_rate": 2.784266724883173e-05, | |
| "loss": 0.1456, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.1781376518218623, | |
| "grad_norm": 0.4765084222226641, | |
| "learning_rate": 2.7790611167497128e-05, | |
| "loss": 0.138, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.182186234817814, | |
| "grad_norm": 0.516338413049325, | |
| "learning_rate": 2.7738492792643224e-05, | |
| "loss": 0.1406, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.1862348178137654, | |
| "grad_norm": 0.48625279251646136, | |
| "learning_rate": 2.768631254100716e-05, | |
| "loss": 0.142, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1902834008097165, | |
| "grad_norm": 0.4812948891047256, | |
| "learning_rate": 2.7634070829820836e-05, | |
| "loss": 0.1401, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.194331983805668, | |
| "grad_norm": 0.4727496645539855, | |
| "learning_rate": 2.7581768076807592e-05, | |
| "loss": 0.139, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.1983805668016196, | |
| "grad_norm": 0.4754238476963871, | |
| "learning_rate": 2.7529404700178832e-05, | |
| "loss": 0.1309, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.2024291497975708, | |
| "grad_norm": 0.4634866984833147, | |
| "learning_rate": 2.7476981118630735e-05, | |
| "loss": 0.1274, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.2064777327935223, | |
| "grad_norm": 0.4846875013480393, | |
| "learning_rate": 2.742449775134085e-05, | |
| "loss": 0.146, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.47007758572096986, | |
| "learning_rate": 2.7371955017964788e-05, | |
| "loss": 0.1289, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.214574898785425, | |
| "grad_norm": 0.46699261945553516, | |
| "learning_rate": 2.7319353338632845e-05, | |
| "loss": 0.1293, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.2186234817813766, | |
| "grad_norm": 0.47989785491183556, | |
| "learning_rate": 2.7266693133946645e-05, | |
| "loss": 0.1394, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.2226720647773277, | |
| "grad_norm": 0.48746429365513333, | |
| "learning_rate": 2.7213974824975785e-05, | |
| "loss": 0.1447, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2267206477732793, | |
| "grad_norm": 0.4683501288898338, | |
| "learning_rate": 2.7161198833254464e-05, | |
| "loss": 0.1401, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.4716139560301076, | |
| "learning_rate": 2.7108365580778095e-05, | |
| "loss": 0.1425, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.234817813765182, | |
| "grad_norm": 0.46422097575353055, | |
| "learning_rate": 2.705547548999996e-05, | |
| "loss": 0.1445, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.2388663967611335, | |
| "grad_norm": 0.48055387517043546, | |
| "learning_rate": 2.7002528983827817e-05, | |
| "loss": 0.1469, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.242914979757085, | |
| "grad_norm": 0.47015026352164946, | |
| "learning_rate": 2.6949526485620517e-05, | |
| "loss": 0.1358, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.246963562753036, | |
| "grad_norm": 0.47391950526617194, | |
| "learning_rate": 2.6896468419184623e-05, | |
| "loss": 0.1281, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.251012145748988, | |
| "grad_norm": 0.46386727615257883, | |
| "learning_rate": 2.6843355208771024e-05, | |
| "loss": 0.133, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.2550607287449393, | |
| "grad_norm": 0.4429983684163096, | |
| "learning_rate": 2.6790187279071536e-05, | |
| "loss": 0.1242, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.2591093117408905, | |
| "grad_norm": 0.4735928341938856, | |
| "learning_rate": 2.67369650552155e-05, | |
| "loss": 0.1389, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.47063801607288996, | |
| "learning_rate": 2.6683688962766407e-05, | |
| "loss": 0.1312, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.2672064777327936, | |
| "grad_norm": 0.4639817853953108, | |
| "learning_rate": 2.6630359427718468e-05, | |
| "loss": 0.1483, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2712550607287447, | |
| "grad_norm": 0.46715218408487025, | |
| "learning_rate": 2.6576976876493225e-05, | |
| "loss": 0.1298, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.4688840092225887, | |
| "learning_rate": 2.652354173593614e-05, | |
| "loss": 0.1289, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.279352226720648, | |
| "grad_norm": 0.515346982629467, | |
| "learning_rate": 2.647005443331317e-05, | |
| "loss": 0.14, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.283400809716599, | |
| "grad_norm": 0.45629550495072935, | |
| "learning_rate": 2.6416515396307354e-05, | |
| "loss": 0.1533, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.2874493927125505, | |
| "grad_norm": 0.4933247350735794, | |
| "learning_rate": 2.636292505301542e-05, | |
| "loss": 0.1396, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.291497975708502, | |
| "grad_norm": 0.4788387357990163, | |
| "learning_rate": 2.630928383194431e-05, | |
| "loss": 0.141, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.2955465587044532, | |
| "grad_norm": 0.47273154437931936, | |
| "learning_rate": 2.625559216200781e-05, | |
| "loss": 0.1322, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.299595141700405, | |
| "grad_norm": 0.49473101506951983, | |
| "learning_rate": 2.620185047252308e-05, | |
| "loss": 0.1363, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.3036437246963564, | |
| "grad_norm": 0.4684259225255795, | |
| "learning_rate": 2.614805919320723e-05, | |
| "loss": 0.1328, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.48600201712501423, | |
| "learning_rate": 2.6094218754173895e-05, | |
| "loss": 0.1405, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.311740890688259, | |
| "grad_norm": 0.4900885526260744, | |
| "learning_rate": 2.60403295859298e-05, | |
| "loss": 0.1411, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.5125612984691664, | |
| "learning_rate": 2.5986392119371297e-05, | |
| "loss": 0.1592, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.3198380566801617, | |
| "grad_norm": 0.4694429386481874, | |
| "learning_rate": 2.5932406785780928e-05, | |
| "loss": 0.1393, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.3238866396761133, | |
| "grad_norm": 0.48275064152087505, | |
| "learning_rate": 2.5878374016823987e-05, | |
| "loss": 0.1332, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.327935222672065, | |
| "grad_norm": 0.4557559235287692, | |
| "learning_rate": 2.5824294244545045e-05, | |
| "loss": 0.1322, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.331983805668016, | |
| "grad_norm": 0.4801554633528288, | |
| "learning_rate": 2.577016790136453e-05, | |
| "loss": 0.1385, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.3360323886639676, | |
| "grad_norm": 0.47377384210190104, | |
| "learning_rate": 2.5715995420075225e-05, | |
| "loss": 0.1418, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.340080971659919, | |
| "grad_norm": 0.46469636117289354, | |
| "learning_rate": 2.5661777233838864e-05, | |
| "loss": 0.1493, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.3441295546558703, | |
| "grad_norm": 0.480404708476999, | |
| "learning_rate": 2.5607513776182616e-05, | |
| "loss": 0.1311, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.348178137651822, | |
| "grad_norm": 0.46574392413957033, | |
| "learning_rate": 2.5553205480995633e-05, | |
| "loss": 0.1434, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3522267206477734, | |
| "grad_norm": 0.49370897864591695, | |
| "learning_rate": 2.5498852782525604e-05, | |
| "loss": 0.1554, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.3562753036437245, | |
| "grad_norm": 0.4783371163781702, | |
| "learning_rate": 2.5444456115375252e-05, | |
| "loss": 0.1572, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.360323886639676, | |
| "grad_norm": 0.4800591928734701, | |
| "learning_rate": 2.539001591449889e-05, | |
| "loss": 0.1364, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.3643724696356276, | |
| "grad_norm": 0.45366139512101317, | |
| "learning_rate": 2.53355326151989e-05, | |
| "loss": 0.1294, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.47412244218892813, | |
| "learning_rate": 2.5281006653122288e-05, | |
| "loss": 0.1399, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.3724696356275303, | |
| "grad_norm": 0.48467297574880175, | |
| "learning_rate": 2.5226438464257217e-05, | |
| "loss": 0.1563, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.376518218623482, | |
| "grad_norm": 0.4795359920158383, | |
| "learning_rate": 2.5171828484929447e-05, | |
| "loss": 0.1453, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.380566801619433, | |
| "grad_norm": 0.45551886331511415, | |
| "learning_rate": 2.5117177151798934e-05, | |
| "loss": 0.1394, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.492488737040852, | |
| "learning_rate": 2.506248490185627e-05, | |
| "loss": 0.1449, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.388663967611336, | |
| "grad_norm": 0.45059602372394825, | |
| "learning_rate": 2.500775217241923e-05, | |
| "loss": 0.1332, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.3927125506072873, | |
| "grad_norm": 0.4786442389815131, | |
| "learning_rate": 2.4952979401129266e-05, | |
| "loss": 0.136, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.396761133603239, | |
| "grad_norm": 0.4934459683178504, | |
| "learning_rate": 2.4898167025948e-05, | |
| "loss": 0.1343, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.4008097165991904, | |
| "grad_norm": 0.4517918900162648, | |
| "learning_rate": 2.4843315485153703e-05, | |
| "loss": 0.1338, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.4048582995951415, | |
| "grad_norm": 0.4732649452092136, | |
| "learning_rate": 2.4788425217337856e-05, | |
| "loss": 0.1394, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.408906882591093, | |
| "grad_norm": 0.46432409136730385, | |
| "learning_rate": 2.4733496661401572e-05, | |
| "loss": 0.1306, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.4129554655870447, | |
| "grad_norm": 0.4360993064415112, | |
| "learning_rate": 2.4678530256552114e-05, | |
| "loss": 0.1174, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.417004048582996, | |
| "grad_norm": 0.4589561595607134, | |
| "learning_rate": 2.46235264422994e-05, | |
| "loss": 0.1321, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.45229475682455583, | |
| "learning_rate": 2.4568485658452452e-05, | |
| "loss": 0.1346, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.425101214574899, | |
| "grad_norm": 0.513375584446629, | |
| "learning_rate": 2.451340834511592e-05, | |
| "loss": 0.1394, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.42914979757085, | |
| "grad_norm": 0.45276521597333275, | |
| "learning_rate": 2.4458294942686528e-05, | |
| "loss": 0.129, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4331983805668016, | |
| "grad_norm": 0.49258253634828, | |
| "learning_rate": 2.440314589184958e-05, | |
| "loss": 0.1298, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.437246963562753, | |
| "grad_norm": 0.4889897169108998, | |
| "learning_rate": 2.434796163357541e-05, | |
| "loss": 0.1525, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.4412955465587043, | |
| "grad_norm": 0.4809882574855909, | |
| "learning_rate": 2.429274260911588e-05, | |
| "loss": 0.1402, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.445344129554656, | |
| "grad_norm": 0.5010736976953899, | |
| "learning_rate": 2.423748926000084e-05, | |
| "loss": 0.1452, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.4493927125506074, | |
| "grad_norm": 0.4449353664028447, | |
| "learning_rate": 2.418220202803459e-05, | |
| "loss": 0.129, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.4534412955465585, | |
| "grad_norm": 0.48159672177069224, | |
| "learning_rate": 2.412688135529237e-05, | |
| "loss": 0.1198, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.45748987854251, | |
| "grad_norm": 0.4600884992030372, | |
| "learning_rate": 2.4071527684116807e-05, | |
| "loss": 0.1404, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.47031483164878135, | |
| "learning_rate": 2.4016141457114372e-05, | |
| "loss": 0.127, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.465587044534413, | |
| "grad_norm": 0.4290481473398923, | |
| "learning_rate": 2.396072311715187e-05, | |
| "loss": 0.1093, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.4696356275303644, | |
| "grad_norm": 0.4356571913796565, | |
| "learning_rate": 2.3905273107352863e-05, | |
| "loss": 0.1134, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.5251431226928566, | |
| "learning_rate": 2.3849791871094157e-05, | |
| "loss": 0.1341, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.477732793522267, | |
| "grad_norm": 0.47313852013260566, | |
| "learning_rate": 2.379427985200223e-05, | |
| "loss": 0.1186, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.4817813765182186, | |
| "grad_norm": 0.44163569989500145, | |
| "learning_rate": 2.373873749394972e-05, | |
| "loss": 0.1244, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.48582995951417, | |
| "grad_norm": 0.48267001626127015, | |
| "learning_rate": 2.3683165241051836e-05, | |
| "loss": 0.1474, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.4898785425101213, | |
| "grad_norm": 0.4655166750478048, | |
| "learning_rate": 2.3627563537662823e-05, | |
| "loss": 0.1465, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.493927125506073, | |
| "grad_norm": 0.45480087015393206, | |
| "learning_rate": 2.3571932828372433e-05, | |
| "loss": 0.134, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.4979757085020244, | |
| "grad_norm": 0.46495931721561357, | |
| "learning_rate": 2.351627355800233e-05, | |
| "loss": 0.1279, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.5020242914979756, | |
| "grad_norm": 0.461838497963151, | |
| "learning_rate": 2.346058617160256e-05, | |
| "loss": 0.1394, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.506072874493927, | |
| "grad_norm": 0.4534387750610306, | |
| "learning_rate": 2.3404871114447976e-05, | |
| "loss": 0.1336, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.5101214574898787, | |
| "grad_norm": 0.4396830200828266, | |
| "learning_rate": 2.334912883203469e-05, | |
| "loss": 0.1216, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.51417004048583, | |
| "grad_norm": 0.4434418538038561, | |
| "learning_rate": 2.3293359770076505e-05, | |
| "loss": 0.1344, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.5182186234817814, | |
| "grad_norm": 0.43488554067915985, | |
| "learning_rate": 2.323756437450136e-05, | |
| "loss": 0.1339, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.522267206477733, | |
| "grad_norm": 0.43787251881544037, | |
| "learning_rate": 2.3181743091447754e-05, | |
| "loss": 0.1218, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.42044643031383616, | |
| "learning_rate": 2.3125896367261164e-05, | |
| "loss": 0.122, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5303643724696356, | |
| "grad_norm": 0.43376174854028743, | |
| "learning_rate": 2.307002464849052e-05, | |
| "loss": 0.1252, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.534412955465587, | |
| "grad_norm": 0.4640969247602629, | |
| "learning_rate": 2.30141283818846e-05, | |
| "loss": 0.1409, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.41660559423971444, | |
| "learning_rate": 2.2958208014388462e-05, | |
| "loss": 0.1076, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.54251012145749, | |
| "grad_norm": 0.43237665247102713, | |
| "learning_rate": 2.2902263993139857e-05, | |
| "loss": 0.1314, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.5465587044534415, | |
| "grad_norm": 0.4510568197539964, | |
| "learning_rate": 2.2846296765465708e-05, | |
| "loss": 0.1244, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.5506072874493926, | |
| "grad_norm": 0.4208374339463754, | |
| "learning_rate": 2.2790306778878465e-05, | |
| "loss": 0.1121, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.554655870445344, | |
| "grad_norm": 0.43044678820204635, | |
| "learning_rate": 2.273429448107257e-05, | |
| "loss": 0.1339, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.5587044534412957, | |
| "grad_norm": 0.45606331306187875, | |
| "learning_rate": 2.2678260319920868e-05, | |
| "loss": 0.1356, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.562753036437247, | |
| "grad_norm": 0.43000162044318296, | |
| "learning_rate": 2.262220474347102e-05, | |
| "loss": 0.1208, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.5668016194331984, | |
| "grad_norm": 0.43933695375366677, | |
| "learning_rate": 2.2566128199941918e-05, | |
| "loss": 0.1251, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.57085020242915, | |
| "grad_norm": 0.45045049933668435, | |
| "learning_rate": 2.2510031137720104e-05, | |
| "loss": 0.1291, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.574898785425101, | |
| "grad_norm": 0.4213846963035745, | |
| "learning_rate": 2.2453914005356206e-05, | |
| "loss": 0.1178, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.45382115958104263, | |
| "learning_rate": 2.2397777251561302e-05, | |
| "loss": 0.1158, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.582995951417004, | |
| "grad_norm": 0.41997817371613183, | |
| "learning_rate": 2.2341621325203393e-05, | |
| "loss": 0.1223, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.5870445344129553, | |
| "grad_norm": 0.43030517421842024, | |
| "learning_rate": 2.2285446675303764e-05, | |
| "loss": 0.12, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.591093117408907, | |
| "grad_norm": 0.4586356091704196, | |
| "learning_rate": 2.2229253751033423e-05, | |
| "loss": 0.1132, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.5951417004048585, | |
| "grad_norm": 0.48534941292914374, | |
| "learning_rate": 2.2173043001709492e-05, | |
| "loss": 0.1529, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.5991902834008096, | |
| "grad_norm": 0.4567340895136368, | |
| "learning_rate": 2.2116814876791627e-05, | |
| "loss": 0.1194, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.603238866396761, | |
| "grad_norm": 0.44849133947265674, | |
| "learning_rate": 2.2060569825878413e-05, | |
| "loss": 0.129, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.6072874493927127, | |
| "grad_norm": 0.4477160761282708, | |
| "learning_rate": 2.2004308298703777e-05, | |
| "loss": 0.1292, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.611336032388664, | |
| "grad_norm": 0.44371256140539106, | |
| "learning_rate": 2.1948030745133396e-05, | |
| "loss": 0.1122, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.44271973937661735, | |
| "learning_rate": 2.1891737615161083e-05, | |
| "loss": 0.1183, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.619433198380567, | |
| "grad_norm": 0.4782914777308083, | |
| "learning_rate": 2.1835429358905218e-05, | |
| "loss": 0.1286, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.623481781376518, | |
| "grad_norm": 0.44952455523274026, | |
| "learning_rate": 2.17791064266051e-05, | |
| "loss": 0.1279, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.6275303643724697, | |
| "grad_norm": 0.44560057038781553, | |
| "learning_rate": 2.172276926861741e-05, | |
| "loss": 0.1301, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.4506997614916057, | |
| "learning_rate": 2.166641833541255e-05, | |
| "loss": 0.1269, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6356275303643724, | |
| "grad_norm": 0.4745989996925604, | |
| "learning_rate": 2.161005407757109e-05, | |
| "loss": 0.1475, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.639676113360324, | |
| "grad_norm": 0.4384242034181404, | |
| "learning_rate": 2.155367694578013e-05, | |
| "loss": 0.1354, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.6437246963562755, | |
| "grad_norm": 0.44218847615091045, | |
| "learning_rate": 2.1497287390829712e-05, | |
| "loss": 0.1284, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.6477732793522266, | |
| "grad_norm": 0.43887992072452825, | |
| "learning_rate": 2.1440885863609208e-05, | |
| "loss": 0.1223, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.651821862348178, | |
| "grad_norm": 0.4457391573594413, | |
| "learning_rate": 2.1384472815103737e-05, | |
| "loss": 0.1381, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6558704453441297, | |
| "grad_norm": 0.4321400765031413, | |
| "learning_rate": 2.1328048696390534e-05, | |
| "loss": 0.1149, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.659919028340081, | |
| "grad_norm": 0.4552777461439712, | |
| "learning_rate": 2.1271613958635342e-05, | |
| "loss": 0.1285, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.6639676113360324, | |
| "grad_norm": 0.42783013019862265, | |
| "learning_rate": 2.1215169053088825e-05, | |
| "loss": 0.1164, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.668016194331984, | |
| "grad_norm": 0.4232093110550775, | |
| "learning_rate": 2.1158714431082937e-05, | |
| "loss": 0.1123, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.672064777327935, | |
| "grad_norm": 0.422512008026287, | |
| "learning_rate": 2.110225054402733e-05, | |
| "loss": 0.1062, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.6761133603238867, | |
| "grad_norm": 0.4493158534850418, | |
| "learning_rate": 2.1045777843405747e-05, | |
| "loss": 0.1239, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.6801619433198383, | |
| "grad_norm": 0.43393440382523696, | |
| "learning_rate": 2.09892967807724e-05, | |
| "loss": 0.1087, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.44502888644219674, | |
| "learning_rate": 2.093280780774835e-05, | |
| "loss": 0.1216, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.688259109311741, | |
| "grad_norm": 0.4183840268792954, | |
| "learning_rate": 2.087631137601793e-05, | |
| "loss": 0.1051, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.43884305451477745, | |
| "learning_rate": 2.0819807937325104e-05, | |
| "loss": 0.1071, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.6963562753036436, | |
| "grad_norm": 0.4190610187704408, | |
| "learning_rate": 2.0763297943469847e-05, | |
| "loss": 0.119, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.700404858299595, | |
| "grad_norm": 0.41996816323539576, | |
| "learning_rate": 2.070678184630458e-05, | |
| "loss": 0.1159, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.7044534412955468, | |
| "grad_norm": 0.44368601928623086, | |
| "learning_rate": 2.06502600977305e-05, | |
| "loss": 0.1152, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.708502024291498, | |
| "grad_norm": 0.4437271451712315, | |
| "learning_rate": 2.0593733149694008e-05, | |
| "loss": 0.1379, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.7125506072874495, | |
| "grad_norm": 0.46969326542691087, | |
| "learning_rate": 2.0537201454183074e-05, | |
| "loss": 0.1345, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.716599190283401, | |
| "grad_norm": 0.4360001288579485, | |
| "learning_rate": 2.0480665463223626e-05, | |
| "loss": 0.1292, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.720647773279352, | |
| "grad_norm": 0.4150794062397875, | |
| "learning_rate": 2.042412562887594e-05, | |
| "loss": 0.1112, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7246963562753037, | |
| "grad_norm": 0.4393877251621434, | |
| "learning_rate": 2.0367582403231033e-05, | |
| "loss": 0.1209, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.7287449392712553, | |
| "grad_norm": 0.44428301463129727, | |
| "learning_rate": 2.0311036238407023e-05, | |
| "loss": 0.1206, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7327935222672064, | |
| "grad_norm": 0.42281932052082116, | |
| "learning_rate": 2.0254487586545534e-05, | |
| "loss": 0.1269, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.435461352345069, | |
| "learning_rate": 2.0197936899808088e-05, | |
| "loss": 0.1235, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.7408906882591095, | |
| "grad_norm": 0.42855412998451603, | |
| "learning_rate": 2.0141384630372466e-05, | |
| "loss": 0.1309, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7449392712550607, | |
| "grad_norm": 0.42959335520047, | |
| "learning_rate": 2.0084831230429106e-05, | |
| "loss": 0.1246, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.748987854251012, | |
| "grad_norm": 0.42024647752838606, | |
| "learning_rate": 2.0028277152177492e-05, | |
| "loss": 0.1232, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.753036437246964, | |
| "grad_norm": 0.4299769593711143, | |
| "learning_rate": 1.9971722847822518e-05, | |
| "loss": 0.1214, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.757085020242915, | |
| "grad_norm": 0.4124007681779938, | |
| "learning_rate": 1.99151687695709e-05, | |
| "loss": 0.1104, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7611336032388665, | |
| "grad_norm": 0.42476553859793, | |
| "learning_rate": 1.9858615369627537e-05, | |
| "loss": 0.1135, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.765182186234818, | |
| "grad_norm": 0.4086675107722901, | |
| "learning_rate": 1.9802063100191916e-05, | |
| "loss": 0.1113, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.43975896387905583, | |
| "learning_rate": 1.974551241345447e-05, | |
| "loss": 0.1236, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.7732793522267207, | |
| "grad_norm": 0.4066581400104201, | |
| "learning_rate": 1.9688963761592984e-05, | |
| "loss": 0.1006, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7773279352226723, | |
| "grad_norm": 0.43764496660627816, | |
| "learning_rate": 1.963241759676897e-05, | |
| "loss": 0.1225, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.7813765182186234, | |
| "grad_norm": 0.4089191520828651, | |
| "learning_rate": 1.9575874371124066e-05, | |
| "loss": 0.107, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.785425101214575, | |
| "grad_norm": 0.4225050689823692, | |
| "learning_rate": 1.951933453677638e-05, | |
| "loss": 0.1208, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.4201015862434235, | |
| "learning_rate": 1.9462798545816932e-05, | |
| "loss": 0.1154, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.7935222672064777, | |
| "grad_norm": 0.4260208974082497, | |
| "learning_rate": 1.9406266850305996e-05, | |
| "loss": 0.1173, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7975708502024292, | |
| "grad_norm": 0.41092827876733784, | |
| "learning_rate": 1.9349739902269503e-05, | |
| "loss": 0.1157, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.801619433198381, | |
| "grad_norm": 0.40525990468737283, | |
| "learning_rate": 1.9293218153695425e-05, | |
| "loss": 0.1078, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.805668016194332, | |
| "grad_norm": 0.4239914793818898, | |
| "learning_rate": 1.9236702056530153e-05, | |
| "loss": 0.1101, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.8097165991902835, | |
| "grad_norm": 0.40061281460206144, | |
| "learning_rate": 1.9180192062674907e-05, | |
| "loss": 0.1059, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.813765182186235, | |
| "grad_norm": 0.4582148684528146, | |
| "learning_rate": 1.9123688623982076e-05, | |
| "loss": 0.1194, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.817813765182186, | |
| "grad_norm": 0.4209263641703365, | |
| "learning_rate": 1.906719219225165e-05, | |
| "loss": 0.105, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.8218623481781377, | |
| "grad_norm": 0.4612551067605902, | |
| "learning_rate": 1.9010703219227603e-05, | |
| "loss": 0.1395, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.8259109311740893, | |
| "grad_norm": 0.44312566708887247, | |
| "learning_rate": 1.8954222156594256e-05, | |
| "loss": 0.1367, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.8299595141700404, | |
| "grad_norm": 0.4386469931078378, | |
| "learning_rate": 1.8897749455972673e-05, | |
| "loss": 0.1218, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.834008097165992, | |
| "grad_norm": 0.4636550056260926, | |
| "learning_rate": 1.884128556891707e-05, | |
| "loss": 0.1122, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8380566801619436, | |
| "grad_norm": 0.4310526047353378, | |
| "learning_rate": 1.878483094691119e-05, | |
| "loss": 0.1128, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.43144958877116646, | |
| "learning_rate": 1.8728386041364664e-05, | |
| "loss": 0.1176, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.44816949429363584, | |
| "learning_rate": 1.867195130360947e-05, | |
| "loss": 0.1079, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.850202429149798, | |
| "grad_norm": 0.4547982506296234, | |
| "learning_rate": 1.8615527184896263e-05, | |
| "loss": 0.1254, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.854251012145749, | |
| "grad_norm": 0.4123823889199154, | |
| "learning_rate": 1.8559114136390795e-05, | |
| "loss": 0.1038, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8582995951417005, | |
| "grad_norm": 0.41648769993185625, | |
| "learning_rate": 1.8502712609170298e-05, | |
| "loss": 0.097, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.862348178137652, | |
| "grad_norm": 0.4227333863013822, | |
| "learning_rate": 1.8446323054219876e-05, | |
| "loss": 0.1248, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.866396761133603, | |
| "grad_norm": 0.4419026168670899, | |
| "learning_rate": 1.838994592242891e-05, | |
| "loss": 0.1208, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.8704453441295548, | |
| "grad_norm": 0.39762404224466014, | |
| "learning_rate": 1.8333581664587453e-05, | |
| "loss": 0.0913, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.8744939271255063, | |
| "grad_norm": 0.4192372156120713, | |
| "learning_rate": 1.8277230731382593e-05, | |
| "loss": 0.1035, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8785425101214575, | |
| "grad_norm": 0.46679014906918204, | |
| "learning_rate": 1.82208935733949e-05, | |
| "loss": 0.124, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.882591093117409, | |
| "grad_norm": 0.3838348752460877, | |
| "learning_rate": 1.8164570641094793e-05, | |
| "loss": 0.0916, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.8866396761133606, | |
| "grad_norm": 0.42260025567609966, | |
| "learning_rate": 1.810826238483892e-05, | |
| "loss": 0.0987, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.8906882591093117, | |
| "grad_norm": 0.41803279484805667, | |
| "learning_rate": 1.8051969254866608e-05, | |
| "loss": 0.1071, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.4254963829524555, | |
| "learning_rate": 1.7995691701296226e-05, | |
| "loss": 0.1063, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.898785425101215, | |
| "grad_norm": 0.43266963496692395, | |
| "learning_rate": 1.7939430174121593e-05, | |
| "loss": 0.1231, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.902834008097166, | |
| "grad_norm": 0.3989473986522358, | |
| "learning_rate": 1.788318512320838e-05, | |
| "loss": 0.1037, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.9068825910931175, | |
| "grad_norm": 0.4171070318338755, | |
| "learning_rate": 1.782695699829051e-05, | |
| "loss": 0.096, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.910931174089069, | |
| "grad_norm": 0.3875161241443959, | |
| "learning_rate": 1.7770746248966587e-05, | |
| "loss": 0.0959, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.91497975708502, | |
| "grad_norm": 0.4040058204677392, | |
| "learning_rate": 1.7714553324696243e-05, | |
| "loss": 0.0927, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.919028340080972, | |
| "grad_norm": 0.45844894354646293, | |
| "learning_rate": 1.7658378674796614e-05, | |
| "loss": 0.1134, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.40683537021661975, | |
| "learning_rate": 1.76022227484387e-05, | |
| "loss": 0.1094, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.9271255060728745, | |
| "grad_norm": 0.41845260796927186, | |
| "learning_rate": 1.7546085994643807e-05, | |
| "loss": 0.1134, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.931174089068826, | |
| "grad_norm": 0.4092219396245873, | |
| "learning_rate": 1.7489968862279902e-05, | |
| "loss": 0.101, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9352226720647776, | |
| "grad_norm": 0.3798723791046161, | |
| "learning_rate": 1.743387180005809e-05, | |
| "loss": 0.0897, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.9392712550607287, | |
| "grad_norm": 0.4122779975037354, | |
| "learning_rate": 1.737779525652899e-05, | |
| "loss": 0.1075, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.9433198380566803, | |
| "grad_norm": 0.4493893711809922, | |
| "learning_rate": 1.732173968007914e-05, | |
| "loss": 0.1329, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.4136541847323124, | |
| "learning_rate": 1.7265705518927437e-05, | |
| "loss": 0.121, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.951417004048583, | |
| "grad_norm": 0.4479927107496375, | |
| "learning_rate": 1.7209693221121542e-05, | |
| "loss": 0.1228, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.9554655870445345, | |
| "grad_norm": 0.3870659340186111, | |
| "learning_rate": 1.7153703234534302e-05, | |
| "loss": 0.0797, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9595141700404857, | |
| "grad_norm": 0.3901033090504797, | |
| "learning_rate": 1.709773600686015e-05, | |
| "loss": 0.0929, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.9635627530364372, | |
| "grad_norm": 0.4263177483907374, | |
| "learning_rate": 1.7041791985611548e-05, | |
| "loss": 0.1057, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.967611336032389, | |
| "grad_norm": 0.3863077603814813, | |
| "learning_rate": 1.69858716181154e-05, | |
| "loss": 0.0905, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.97165991902834, | |
| "grad_norm": 0.39186017708503185, | |
| "learning_rate": 1.692997535150948e-05, | |
| "loss": 0.0968, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.9757085020242915, | |
| "grad_norm": 0.44475919346679244, | |
| "learning_rate": 1.687410363273884e-05, | |
| "loss": 0.1136, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.979757085020243, | |
| "grad_norm": 0.41261512597302713, | |
| "learning_rate": 1.6818256908552257e-05, | |
| "loss": 0.1045, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.983805668016194, | |
| "grad_norm": 0.38889138407040286, | |
| "learning_rate": 1.6762435625498646e-05, | |
| "loss": 0.0993, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.9878542510121457, | |
| "grad_norm": 0.39765244139867495, | |
| "learning_rate": 1.67066402299235e-05, | |
| "loss": 0.1072, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.9919028340080973, | |
| "grad_norm": 0.38856655564439907, | |
| "learning_rate": 1.6650871167965313e-05, | |
| "loss": 0.0952, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.9959514170040484, | |
| "grad_norm": 0.3842611743895166, | |
| "learning_rate": 1.6595128885552028e-05, | |
| "loss": 0.097, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.41452912250319873, | |
| "learning_rate": 1.653941382839745e-05, | |
| "loss": 0.1087, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.0040485829959516, | |
| "grad_norm": 0.3352436752967678, | |
| "learning_rate": 1.6483726441997673e-05, | |
| "loss": 0.0391, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.0080971659919027, | |
| "grad_norm": 0.2924171113421083, | |
| "learning_rate": 1.642806717162757e-05, | |
| "loss": 0.0384, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.0121457489878543, | |
| "grad_norm": 0.3173474838793183, | |
| "learning_rate": 1.637243646233718e-05, | |
| "loss": 0.0436, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.016194331983806, | |
| "grad_norm": 0.3270684773333412, | |
| "learning_rate": 1.6316834758948174e-05, | |
| "loss": 0.0392, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.020242914979757, | |
| "grad_norm": 0.3396066830448493, | |
| "learning_rate": 1.6261262506050282e-05, | |
| "loss": 0.0395, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.0242914979757085, | |
| "grad_norm": 0.34027769969986643, | |
| "learning_rate": 1.620572014799777e-05, | |
| "loss": 0.0416, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.02834008097166, | |
| "grad_norm": 0.33005525829907345, | |
| "learning_rate": 1.6150208128905857e-05, | |
| "loss": 0.0461, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.032388663967611, | |
| "grad_norm": 0.2979876487713673, | |
| "learning_rate": 1.6094726892647147e-05, | |
| "loss": 0.0362, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0364372469635628, | |
| "grad_norm": 0.30355320424682875, | |
| "learning_rate": 1.6039276882848135e-05, | |
| "loss": 0.0384, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0404858299595143, | |
| "grad_norm": 0.3285185393354326, | |
| "learning_rate": 1.5983858542885635e-05, | |
| "loss": 0.0445, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.0445344129554655, | |
| "grad_norm": 0.3094507306827858, | |
| "learning_rate": 1.5928472315883203e-05, | |
| "loss": 0.0431, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.048582995951417, | |
| "grad_norm": 0.2866421683727671, | |
| "learning_rate": 1.5873118644707633e-05, | |
| "loss": 0.0361, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.3171148279567297, | |
| "learning_rate": 1.5817797971965413e-05, | |
| "loss": 0.0436, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.0566801619433197, | |
| "grad_norm": 0.2951058220963285, | |
| "learning_rate": 1.576251073999917e-05, | |
| "loss": 0.0355, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.0607287449392713, | |
| "grad_norm": 0.3129832135815809, | |
| "learning_rate": 1.5707257390884126e-05, | |
| "loss": 0.0385, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.064777327935223, | |
| "grad_norm": 0.2929144027299848, | |
| "learning_rate": 1.5652038366424595e-05, | |
| "loss": 0.0326, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.068825910931174, | |
| "grad_norm": 0.3036965722500406, | |
| "learning_rate": 1.5596854108150424e-05, | |
| "loss": 0.039, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.0728744939271255, | |
| "grad_norm": 0.30740784047310193, | |
| "learning_rate": 1.5541705057313476e-05, | |
| "loss": 0.036, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.30921843665048937, | |
| "learning_rate": 1.5486591654884086e-05, | |
| "loss": 0.0425, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.080971659919028, | |
| "grad_norm": 0.2862420773656848, | |
| "learning_rate": 1.543151434154755e-05, | |
| "loss": 0.0396, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.08502024291498, | |
| "grad_norm": 0.2785297645226252, | |
| "learning_rate": 1.5376473557700615e-05, | |
| "loss": 0.0355, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.0890688259109313, | |
| "grad_norm": 0.29615868694247494, | |
| "learning_rate": 1.5321469743447893e-05, | |
| "loss": 0.043, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.0931174089068825, | |
| "grad_norm": 0.2826499143346129, | |
| "learning_rate": 1.5266503338598434e-05, | |
| "loss": 0.0384, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.097165991902834, | |
| "grad_norm": 0.3036533559454366, | |
| "learning_rate": 1.5211574782662149e-05, | |
| "loss": 0.0425, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1012145748987856, | |
| "grad_norm": 0.2971419416102655, | |
| "learning_rate": 1.5156684514846307e-05, | |
| "loss": 0.0407, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.30703520449965294, | |
| "learning_rate": 1.5101832974052012e-05, | |
| "loss": 0.037, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.1093117408906883, | |
| "grad_norm": 0.3402511325206599, | |
| "learning_rate": 1.5047020598870737e-05, | |
| "loss": 0.0433, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.11336032388664, | |
| "grad_norm": 0.30528342204546466, | |
| "learning_rate": 1.4992247827580778e-05, | |
| "loss": 0.0394, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.117408906882591, | |
| "grad_norm": 0.30424834385047095, | |
| "learning_rate": 1.493751509814374e-05, | |
| "loss": 0.0374, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1214574898785425, | |
| "grad_norm": 0.31243512892179903, | |
| "learning_rate": 1.4882822848201073e-05, | |
| "loss": 0.0434, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.125506072874494, | |
| "grad_norm": 0.3036154979530872, | |
| "learning_rate": 1.4828171515070553e-05, | |
| "loss": 0.0486, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.1295546558704452, | |
| "grad_norm": 0.4307356104241326, | |
| "learning_rate": 1.4773561535742793e-05, | |
| "loss": 0.0427, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.133603238866397, | |
| "grad_norm": 0.3683582977572595, | |
| "learning_rate": 1.4718993346877715e-05, | |
| "loss": 0.0406, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.1376518218623484, | |
| "grad_norm": 0.39220571629785594, | |
| "learning_rate": 1.466446738480111e-05, | |
| "loss": 0.0398, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1417004048582995, | |
| "grad_norm": 0.4441334939978718, | |
| "learning_rate": 1.460998408550112e-05, | |
| "loss": 0.0386, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.145748987854251, | |
| "grad_norm": 0.2813458847940639, | |
| "learning_rate": 1.4555543884624751e-05, | |
| "loss": 0.0348, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.1497975708502026, | |
| "grad_norm": 0.2777132025536206, | |
| "learning_rate": 1.4501147217474402e-05, | |
| "loss": 0.0338, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.2756716002598391, | |
| "learning_rate": 1.444679451900437e-05, | |
| "loss": 0.0353, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.30484485631109504, | |
| "learning_rate": 1.4392486223817397e-05, | |
| "loss": 0.0418, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.161943319838057, | |
| "grad_norm": 0.31082708732475933, | |
| "learning_rate": 1.4338222766161143e-05, | |
| "loss": 0.0415, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.165991902834008, | |
| "grad_norm": 0.30957826577195297, | |
| "learning_rate": 1.4284004579924777e-05, | |
| "loss": 0.0388, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.1700404858299596, | |
| "grad_norm": 0.2766949063373334, | |
| "learning_rate": 1.4229832098635479e-05, | |
| "loss": 0.0344, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.174089068825911, | |
| "grad_norm": 0.28309257781037445, | |
| "learning_rate": 1.4175705755454963e-05, | |
| "loss": 0.037, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.1781376518218623, | |
| "grad_norm": 0.2970950156205057, | |
| "learning_rate": 1.412162598317602e-05, | |
| "loss": 0.037, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.182186234817814, | |
| "grad_norm": 0.28249296167776256, | |
| "learning_rate": 1.406759321421907e-05, | |
| "loss": 0.0388, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.1862348178137654, | |
| "grad_norm": 0.2943951310285605, | |
| "learning_rate": 1.401360788062871e-05, | |
| "loss": 0.0377, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.1902834008097165, | |
| "grad_norm": 0.2839833088536315, | |
| "learning_rate": 1.3959670414070204e-05, | |
| "loss": 0.0373, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.194331983805668, | |
| "grad_norm": 0.26928584210597145, | |
| "learning_rate": 1.3905781245826108e-05, | |
| "loss": 0.0266, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.1983805668016196, | |
| "grad_norm": 0.27747677399021986, | |
| "learning_rate": 1.3851940806792778e-05, | |
| "loss": 0.0333, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2024291497975708, | |
| "grad_norm": 0.2659551296897198, | |
| "learning_rate": 1.379814952747693e-05, | |
| "loss": 0.032, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.2064777327935223, | |
| "grad_norm": 0.29776118442996014, | |
| "learning_rate": 1.3744407837992193e-05, | |
| "loss": 0.0368, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.2988812641488589, | |
| "learning_rate": 1.3690716168055692e-05, | |
| "loss": 0.0392, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.214574898785425, | |
| "grad_norm": 0.27756961676428116, | |
| "learning_rate": 1.363707494698459e-05, | |
| "loss": 0.035, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.2186234817813766, | |
| "grad_norm": 0.2705523873972445, | |
| "learning_rate": 1.358348460369265e-05, | |
| "loss": 0.0327, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.2226720647773277, | |
| "grad_norm": 0.2985751708536574, | |
| "learning_rate": 1.352994556668684e-05, | |
| "loss": 0.0366, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.2267206477732793, | |
| "grad_norm": 0.2938718720180687, | |
| "learning_rate": 1.3476458264063865e-05, | |
| "loss": 0.0417, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.28309330504530844, | |
| "learning_rate": 1.342302312350678e-05, | |
| "loss": 0.0351, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.234817813765182, | |
| "grad_norm": 0.27293234980278347, | |
| "learning_rate": 1.3369640572281537e-05, | |
| "loss": 0.0353, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.2388663967611335, | |
| "grad_norm": 0.2964908998848756, | |
| "learning_rate": 1.3316311037233596e-05, | |
| "loss": 0.0402, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.242914979757085, | |
| "grad_norm": 0.28577654211853715, | |
| "learning_rate": 1.326303494478451e-05, | |
| "loss": 0.0383, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.246963562753036, | |
| "grad_norm": 0.2697213433148471, | |
| "learning_rate": 1.3209812720928472e-05, | |
| "loss": 0.033, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.251012145748988, | |
| "grad_norm": 0.2906114710970122, | |
| "learning_rate": 1.315664479122898e-05, | |
| "loss": 0.0373, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.2550607287449393, | |
| "grad_norm": 0.3071996709734444, | |
| "learning_rate": 1.3103531580815378e-05, | |
| "loss": 0.041, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.2591093117408905, | |
| "grad_norm": 0.27983498472992846, | |
| "learning_rate": 1.3050473514379488e-05, | |
| "loss": 0.0378, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.31118549487889446, | |
| "learning_rate": 1.2997471016172188e-05, | |
| "loss": 0.0386, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.2672064777327936, | |
| "grad_norm": 0.3064859566182475, | |
| "learning_rate": 1.2944524510000042e-05, | |
| "loss": 0.0401, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.2712550607287447, | |
| "grad_norm": 0.277279045959964, | |
| "learning_rate": 1.2891634419221908e-05, | |
| "loss": 0.0332, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.2753036437246963, | |
| "grad_norm": 0.2680605840687216, | |
| "learning_rate": 1.2838801166745545e-05, | |
| "loss": 0.03, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.279352226720648, | |
| "grad_norm": 0.2790133213341529, | |
| "learning_rate": 1.2786025175024216e-05, | |
| "loss": 0.0356, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.283400809716599, | |
| "grad_norm": 0.2741958256864115, | |
| "learning_rate": 1.2733306866053357e-05, | |
| "loss": 0.0345, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.2874493927125505, | |
| "grad_norm": 0.2855460847136566, | |
| "learning_rate": 1.2680646661367163e-05, | |
| "loss": 0.038, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.291497975708502, | |
| "grad_norm": 0.2879832272949949, | |
| "learning_rate": 1.2628044982035219e-05, | |
| "loss": 0.0398, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.2955465587044532, | |
| "grad_norm": 0.28155960579992556, | |
| "learning_rate": 1.2575502248659155e-05, | |
| "loss": 0.0369, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.299595141700405, | |
| "grad_norm": 0.27831492820020703, | |
| "learning_rate": 1.2523018881369269e-05, | |
| "loss": 0.0325, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.3036437246963564, | |
| "grad_norm": 0.30914633104811046, | |
| "learning_rate": 1.2470595299821176e-05, | |
| "loss": 0.0436, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.2895110868422758, | |
| "learning_rate": 1.2418231923192416e-05, | |
| "loss": 0.0353, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.311740890688259, | |
| "grad_norm": 0.28348353275351557, | |
| "learning_rate": 1.2365929170179165e-05, | |
| "loss": 0.0399, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.309110622942695, | |
| "learning_rate": 1.2313687458992848e-05, | |
| "loss": 0.0394, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.3198380566801617, | |
| "grad_norm": 0.2661068915042164, | |
| "learning_rate": 1.2261507207356785e-05, | |
| "loss": 0.0293, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.3238866396761133, | |
| "grad_norm": 0.24923184610009227, | |
| "learning_rate": 1.220938883250288e-05, | |
| "loss": 0.0255, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.327935222672065, | |
| "grad_norm": 0.29528844163306484, | |
| "learning_rate": 1.215733275116827e-05, | |
| "loss": 0.0367, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.331983805668016, | |
| "grad_norm": 0.2786940884113045, | |
| "learning_rate": 1.2105339379592005e-05, | |
| "loss": 0.0368, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.3360323886639676, | |
| "grad_norm": 0.2628716258832749, | |
| "learning_rate": 1.2053409133511681e-05, | |
| "loss": 0.0323, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.340080971659919, | |
| "grad_norm": 0.2708932415693049, | |
| "learning_rate": 1.2001542428160173e-05, | |
| "loss": 0.0323, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.3441295546558703, | |
| "grad_norm": 0.26220298610752085, | |
| "learning_rate": 1.1949739678262278e-05, | |
| "loss": 0.0295, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.348178137651822, | |
| "grad_norm": 0.28992214665061533, | |
| "learning_rate": 1.189800129803139e-05, | |
| "loss": 0.0362, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.3522267206477734, | |
| "grad_norm": 0.2759139601992978, | |
| "learning_rate": 1.1846327701166228e-05, | |
| "loss": 0.0344, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.3562753036437245, | |
| "grad_norm": 0.2852235723249053, | |
| "learning_rate": 1.1794719300847493e-05, | |
| "loss": 0.0369, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.360323886639676, | |
| "grad_norm": 0.295316209446337, | |
| "learning_rate": 1.1743176509734582e-05, | |
| "loss": 0.0394, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.3643724696356276, | |
| "grad_norm": 0.2474916124446917, | |
| "learning_rate": 1.1691699739962275e-05, | |
| "loss": 0.0298, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.26214051167204094, | |
| "learning_rate": 1.1640289403137456e-05, | |
| "loss": 0.0339, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.3724696356275303, | |
| "grad_norm": 0.28381441042005373, | |
| "learning_rate": 1.1588945910335802e-05, | |
| "loss": 0.0357, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.376518218623482, | |
| "grad_norm": 0.27296206336432705, | |
| "learning_rate": 1.1537669672098517e-05, | |
| "loss": 0.0339, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.380566801619433, | |
| "grad_norm": 0.29170130078331824, | |
| "learning_rate": 1.1486461098429036e-05, | |
| "loss": 0.0319, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.26090470017252293, | |
| "learning_rate": 1.1435320598789747e-05, | |
| "loss": 0.0273, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.388663967611336, | |
| "grad_norm": 0.2987896646593945, | |
| "learning_rate": 1.1384248582098736e-05, | |
| "loss": 0.0408, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.3927125506072873, | |
| "grad_norm": 0.28324439111358096, | |
| "learning_rate": 1.1333245456726467e-05, | |
| "loss": 0.0352, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.396761133603239, | |
| "grad_norm": 0.26698161368373996, | |
| "learning_rate": 1.128231163049258e-05, | |
| "loss": 0.0303, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.4008097165991904, | |
| "grad_norm": 0.310068044752786, | |
| "learning_rate": 1.123144751066259e-05, | |
| "loss": 0.046, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4048582995951415, | |
| "grad_norm": 0.2859434479401864, | |
| "learning_rate": 1.1180653503944657e-05, | |
| "loss": 0.0375, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.408906882591093, | |
| "grad_norm": 0.26954592826713747, | |
| "learning_rate": 1.1129930016486287e-05, | |
| "loss": 0.0337, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.4129554655870447, | |
| "grad_norm": 0.2865956617167514, | |
| "learning_rate": 1.107927745387114e-05, | |
| "loss": 0.0372, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.417004048582996, | |
| "grad_norm": 0.2855046445797711, | |
| "learning_rate": 1.1028696221115766e-05, | |
| "loss": 0.0364, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.3120304927789088, | |
| "learning_rate": 1.0978186722666341e-05, | |
| "loss": 0.0411, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.425101214574899, | |
| "grad_norm": 0.2951674709662551, | |
| "learning_rate": 1.0927749362395475e-05, | |
| "loss": 0.0366, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.42914979757085, | |
| "grad_norm": 0.2648736698609751, | |
| "learning_rate": 1.0877384543598954e-05, | |
| "loss": 0.0284, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.4331983805668016, | |
| "grad_norm": 0.27706824911982236, | |
| "learning_rate": 1.0827092668992529e-05, | |
| "loss": 0.0358, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.437246963562753, | |
| "grad_norm": 0.265103938476179, | |
| "learning_rate": 1.077687414070869e-05, | |
| "loss": 0.0302, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.4412955465587043, | |
| "grad_norm": 0.2796360067167226, | |
| "learning_rate": 1.0726729360293451e-05, | |
| "loss": 0.036, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.445344129554656, | |
| "grad_norm": 0.259415695581298, | |
| "learning_rate": 1.0676658728703147e-05, | |
| "loss": 0.0291, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.4493927125506074, | |
| "grad_norm": 0.2508011584994525, | |
| "learning_rate": 1.0626662646301198e-05, | |
| "loss": 0.0266, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.4534412955465585, | |
| "grad_norm": 0.2404678288690985, | |
| "learning_rate": 1.0576741512854958e-05, | |
| "loss": 0.0264, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.45748987854251, | |
| "grad_norm": 0.2746243221195842, | |
| "learning_rate": 1.052689572753247e-05, | |
| "loss": 0.0353, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.2632323752492928, | |
| "learning_rate": 1.0477125688899328e-05, | |
| "loss": 0.0295, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.465587044534413, | |
| "grad_norm": 0.25845627727855636, | |
| "learning_rate": 1.0427431794915406e-05, | |
| "loss": 0.0309, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.4696356275303644, | |
| "grad_norm": 0.2495753866096952, | |
| "learning_rate": 1.0377814442931769e-05, | |
| "loss": 0.0308, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.263979512050116, | |
| "learning_rate": 1.0328274029687439e-05, | |
| "loss": 0.0313, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.477732793522267, | |
| "grad_norm": 0.2824398709601557, | |
| "learning_rate": 1.0278810951306253e-05, | |
| "loss": 0.0336, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.4817813765182186, | |
| "grad_norm": 0.309207836768734, | |
| "learning_rate": 1.0229425603293647e-05, | |
| "loss": 0.0434, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.48582995951417, | |
| "grad_norm": 0.28215624426109226, | |
| "learning_rate": 1.0180118380533558e-05, | |
| "loss": 0.0358, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.4898785425101213, | |
| "grad_norm": 0.2687048018867749, | |
| "learning_rate": 1.0130889677285225e-05, | |
| "loss": 0.0329, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.493927125506073, | |
| "grad_norm": 0.29285141492709027, | |
| "learning_rate": 1.008173988718005e-05, | |
| "loss": 0.0422, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.4979757085020244, | |
| "grad_norm": 0.2719186172673911, | |
| "learning_rate": 1.0032669403218439e-05, | |
| "loss": 0.035, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.5020242914979756, | |
| "grad_norm": 0.2568674454968868, | |
| "learning_rate": 9.983678617766677e-06, | |
| "loss": 0.0307, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.506072874493927, | |
| "grad_norm": 0.24376806509790808, | |
| "learning_rate": 9.934767922553788e-06, | |
| "loss": 0.0267, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.5101214574898787, | |
| "grad_norm": 0.25187693048153337, | |
| "learning_rate": 9.885937708668371e-06, | |
| "loss": 0.0287, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.51417004048583, | |
| "grad_norm": 0.26876477151049766, | |
| "learning_rate": 9.837188366555524e-06, | |
| "loss": 0.0295, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.5182186234817814, | |
| "grad_norm": 0.2627052958965245, | |
| "learning_rate": 9.788520286013691e-06, | |
| "loss": 0.0287, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.522267206477733, | |
| "grad_norm": 0.25645441699954574, | |
| "learning_rate": 9.739933856191537e-06, | |
| "loss": 0.0312, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.28254412288560915, | |
| "learning_rate": 9.691429465584866e-06, | |
| "loss": 0.0338, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.5303643724696356, | |
| "grad_norm": 0.28922334946468964, | |
| "learning_rate": 9.643007502033492e-06, | |
| "loss": 0.0355, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.534412955465587, | |
| "grad_norm": 0.29998697446263956, | |
| "learning_rate": 9.594668352718152e-06, | |
| "loss": 0.0388, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.2823584394625869, | |
| "learning_rate": 9.546412404157383e-06, | |
| "loss": 0.0346, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.54251012145749, | |
| "grad_norm": 0.26897418616628943, | |
| "learning_rate": 9.498240042204471e-06, | |
| "loss": 0.0325, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.5465587044534415, | |
| "grad_norm": 0.25487124338438966, | |
| "learning_rate": 9.450151652044339e-06, | |
| "loss": 0.0294, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.5506072874493926, | |
| "grad_norm": 0.26010424648894853, | |
| "learning_rate": 9.402147618190472e-06, | |
| "loss": 0.0302, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.554655870445344, | |
| "grad_norm": 0.26493813163505553, | |
| "learning_rate": 9.354228324481841e-06, | |
| "loss": 0.032, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.5587044534412957, | |
| "grad_norm": 0.2683156412068524, | |
| "learning_rate": 9.306394154079852e-06, | |
| "loss": 0.0335, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.562753036437247, | |
| "grad_norm": 0.2790216687434528, | |
| "learning_rate": 9.258645489465254e-06, | |
| "loss": 0.0362, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.5668016194331984, | |
| "grad_norm": 0.2726667114989407, | |
| "learning_rate": 9.21098271243509e-06, | |
| "loss": 0.0332, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.57085020242915, | |
| "grad_norm": 0.26949052709344545, | |
| "learning_rate": 9.163406204099659e-06, | |
| "loss": 0.0311, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.574898785425101, | |
| "grad_norm": 0.27103952688109323, | |
| "learning_rate": 9.115916344879452e-06, | |
| "loss": 0.0329, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.2513782527746535, | |
| "learning_rate": 9.068513514502126e-06, | |
| "loss": 0.0278, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.582995951417004, | |
| "grad_norm": 0.2410769152449758, | |
| "learning_rate": 9.021198091999439e-06, | |
| "loss": 0.0269, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.5870445344129553, | |
| "grad_norm": 0.25484344290650845, | |
| "learning_rate": 8.973970455704251e-06, | |
| "loss": 0.0323, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.591093117408907, | |
| "grad_norm": 0.2637867891164724, | |
| "learning_rate": 8.926830983247495e-06, | |
| "loss": 0.0333, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.5951417004048585, | |
| "grad_norm": 0.22943341570701786, | |
| "learning_rate": 8.879780051555116e-06, | |
| "loss": 0.0233, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.5991902834008096, | |
| "grad_norm": 0.24194733152955303, | |
| "learning_rate": 8.83281803684512e-06, | |
| "loss": 0.0266, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.603238866396761, | |
| "grad_norm": 0.26400951466760725, | |
| "learning_rate": 8.785945314624532e-06, | |
| "loss": 0.0292, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.6072874493927127, | |
| "grad_norm": 0.23848020584590868, | |
| "learning_rate": 8.739162259686385e-06, | |
| "loss": 0.0253, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.611336032388664, | |
| "grad_norm": 0.2604228632779668, | |
| "learning_rate": 8.692469246106734e-06, | |
| "loss": 0.0306, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.2842888154228456, | |
| "learning_rate": 8.645866647241677e-06, | |
| "loss": 0.036, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.619433198380567, | |
| "grad_norm": 0.28882481060786214, | |
| "learning_rate": 8.599354835724348e-06, | |
| "loss": 0.0368, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.623481781376518, | |
| "grad_norm": 0.2676899079902395, | |
| "learning_rate": 8.552934183461958e-06, | |
| "loss": 0.0333, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6275303643724697, | |
| "grad_norm": 0.26103566551486795, | |
| "learning_rate": 8.506605061632805e-06, | |
| "loss": 0.0275, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.2658775844192551, | |
| "learning_rate": 8.460367840683308e-06, | |
| "loss": 0.0332, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.6356275303643724, | |
| "grad_norm": 0.2423079773520941, | |
| "learning_rate": 8.414222890325065e-06, | |
| "loss": 0.0286, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.639676113360324, | |
| "grad_norm": 0.24962250540199904, | |
| "learning_rate": 8.368170579531855e-06, | |
| "loss": 0.0288, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.6437246963562755, | |
| "grad_norm": 0.2656358895283444, | |
| "learning_rate": 8.322211276536733e-06, | |
| "loss": 0.0308, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6477732793522266, | |
| "grad_norm": 0.24955662812079604, | |
| "learning_rate": 8.27634534882907e-06, | |
| "loss": 0.0287, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.651821862348178, | |
| "grad_norm": 0.2581561566746039, | |
| "learning_rate": 8.230573163151585e-06, | |
| "loss": 0.0304, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.6558704453441297, | |
| "grad_norm": 0.24465875017675737, | |
| "learning_rate": 8.184895085497466e-06, | |
| "loss": 0.0273, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.659919028340081, | |
| "grad_norm": 0.2581339595638995, | |
| "learning_rate": 8.139311481107401e-06, | |
| "loss": 0.0318, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.6639676113360324, | |
| "grad_norm": 0.25999068072318293, | |
| "learning_rate": 8.093822714466686e-06, | |
| "loss": 0.031, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.668016194331984, | |
| "grad_norm": 0.2525135170855626, | |
| "learning_rate": 8.048429149302271e-06, | |
| "loss": 0.0316, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.672064777327935, | |
| "grad_norm": 0.24842135738688753, | |
| "learning_rate": 8.003131148579902e-06, | |
| "loss": 0.0281, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.6761133603238867, | |
| "grad_norm": 0.25923593430139935, | |
| "learning_rate": 7.957929074501182e-06, | |
| "loss": 0.0317, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.6801619433198383, | |
| "grad_norm": 0.2539662681781545, | |
| "learning_rate": 7.912823288500695e-06, | |
| "loss": 0.0279, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.2499141298645139, | |
| "learning_rate": 7.867814151243102e-06, | |
| "loss": 0.0273, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.688259109311741, | |
| "grad_norm": 0.24513027256164574, | |
| "learning_rate": 7.822902022620267e-06, | |
| "loss": 0.0287, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.3067332746696152, | |
| "learning_rate": 7.778087261748377e-06, | |
| "loss": 0.0368, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.6963562753036436, | |
| "grad_norm": 0.2648285462040098, | |
| "learning_rate": 7.733370226965056e-06, | |
| "loss": 0.0323, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.700404858299595, | |
| "grad_norm": 0.24686015905692285, | |
| "learning_rate": 7.688751275826527e-06, | |
| "loss": 0.0282, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.7044534412955468, | |
| "grad_norm": 0.23448644381840145, | |
| "learning_rate": 7.644230765104735e-06, | |
| "loss": 0.0277, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.708502024291498, | |
| "grad_norm": 0.2676316902558523, | |
| "learning_rate": 7.599809050784502e-06, | |
| "loss": 0.0331, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.7125506072874495, | |
| "grad_norm": 0.2669227831155205, | |
| "learning_rate": 7.555486488060661e-06, | |
| "loss": 0.0299, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.716599190283401, | |
| "grad_norm": 0.27482778845289924, | |
| "learning_rate": 7.511263431335249e-06, | |
| "loss": 0.0383, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.720647773279352, | |
| "grad_norm": 0.26360975939582865, | |
| "learning_rate": 7.467140234214656e-06, | |
| "loss": 0.0356, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.7246963562753037, | |
| "grad_norm": 0.24763525989054547, | |
| "learning_rate": 7.423117249506777e-06, | |
| "loss": 0.0283, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7287449392712553, | |
| "grad_norm": 0.24250193690179273, | |
| "learning_rate": 7.379194829218231e-06, | |
| "loss": 0.0309, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.7327935222672064, | |
| "grad_norm": 0.24451306574598808, | |
| "learning_rate": 7.335373324551518e-06, | |
| "loss": 0.0244, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.2471000247503652, | |
| "learning_rate": 7.291653085902224e-06, | |
| "loss": 0.0279, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.7408906882591095, | |
| "grad_norm": 0.26134890871193284, | |
| "learning_rate": 7.248034462856213e-06, | |
| "loss": 0.0324, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.7449392712550607, | |
| "grad_norm": 0.25131061622990053, | |
| "learning_rate": 7.20451780418683e-06, | |
| "loss": 0.0276, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.748987854251012, | |
| "grad_norm": 0.2667463488868631, | |
| "learning_rate": 7.161103457852128e-06, | |
| "loss": 0.0314, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.753036437246964, | |
| "grad_norm": 0.2542724325663606, | |
| "learning_rate": 7.117791770992046e-06, | |
| "loss": 0.029, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.757085020242915, | |
| "grad_norm": 0.24868216019686132, | |
| "learning_rate": 7.074583089925688e-06, | |
| "loss": 0.0301, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.7611336032388665, | |
| "grad_norm": 0.24891197318799205, | |
| "learning_rate": 7.0314777601485155e-06, | |
| "loss": 0.028, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.765182186234818, | |
| "grad_norm": 0.26773421091109956, | |
| "learning_rate": 6.988476126329604e-06, | |
| "loss": 0.0329, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.25755880201854314, | |
| "learning_rate": 6.945578532308857e-06, | |
| "loss": 0.03, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.7732793522267207, | |
| "grad_norm": 0.24566001500770698, | |
| "learning_rate": 6.902785321094301e-06, | |
| "loss": 0.0293, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.7773279352226723, | |
| "grad_norm": 0.24970252732362008, | |
| "learning_rate": 6.8600968348593115e-06, | |
| "loss": 0.0268, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.7813765182186234, | |
| "grad_norm": 0.24234711571547435, | |
| "learning_rate": 6.817513414939892e-06, | |
| "loss": 0.0278, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.785425101214575, | |
| "grad_norm": 0.2674435109427063, | |
| "learning_rate": 6.775035401831924e-06, | |
| "loss": 0.0335, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.24501062599552217, | |
| "learning_rate": 6.732663135188473e-06, | |
| "loss": 0.0306, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.7935222672064777, | |
| "grad_norm": 0.22853788388896276, | |
| "learning_rate": 6.690396953817055e-06, | |
| "loss": 0.0233, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.7975708502024292, | |
| "grad_norm": 0.25692417663592426, | |
| "learning_rate": 6.648237195676927e-06, | |
| "loss": 0.0299, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.801619433198381, | |
| "grad_norm": 0.24936474114684865, | |
| "learning_rate": 6.606184197876395e-06, | |
| "loss": 0.027, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.805668016194332, | |
| "grad_norm": 0.2451086201232552, | |
| "learning_rate": 6.564238296670109e-06, | |
| "loss": 0.0282, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.8097165991902835, | |
| "grad_norm": 0.23803801868589414, | |
| "learning_rate": 6.52239982745638e-06, | |
| "loss": 0.0272, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.813765182186235, | |
| "grad_norm": 0.26340085925253814, | |
| "learning_rate": 6.480669124774484e-06, | |
| "loss": 0.0333, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.817813765182186, | |
| "grad_norm": 0.24643300331636903, | |
| "learning_rate": 6.439046522302013e-06, | |
| "loss": 0.0317, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.8218623481781377, | |
| "grad_norm": 0.25611556470985514, | |
| "learning_rate": 6.397532352852191e-06, | |
| "loss": 0.0333, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.8259109311740893, | |
| "grad_norm": 0.2828645233451167, | |
| "learning_rate": 6.356126948371202e-06, | |
| "loss": 0.037, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.8299595141700404, | |
| "grad_norm": 0.25638918666719723, | |
| "learning_rate": 6.314830639935559e-06, | |
| "loss": 0.0338, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.834008097165992, | |
| "grad_norm": 0.237725370591467, | |
| "learning_rate": 6.273643757749448e-06, | |
| "loss": 0.0281, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.8380566801619436, | |
| "grad_norm": 0.2516329275150387, | |
| "learning_rate": 6.2325666311420856e-06, | |
| "loss": 0.0265, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.24191377698483732, | |
| "learning_rate": 6.1915995885650675e-06, | |
| "loss": 0.0245, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.2312243373667707, | |
| "learning_rate": 6.150742957589784e-06, | |
| "loss": 0.024, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.850202429149798, | |
| "grad_norm": 0.25531629789070037, | |
| "learning_rate": 6.109997064904774e-06, | |
| "loss": 0.0327, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.854251012145749, | |
| "grad_norm": 0.24947438793627705, | |
| "learning_rate": 6.069362236313092e-06, | |
| "loss": 0.0282, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.8582995951417005, | |
| "grad_norm": 0.24251380408198905, | |
| "learning_rate": 6.0288387967297565e-06, | |
| "loss": 0.0308, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.862348178137652, | |
| "grad_norm": 0.23583576417975477, | |
| "learning_rate": 5.988427070179108e-06, | |
| "loss": 0.0308, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.866396761133603, | |
| "grad_norm": 0.23974245412368303, | |
| "learning_rate": 5.948127379792234e-06, | |
| "loss": 0.0277, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.8704453441295548, | |
| "grad_norm": 0.23502261457108042, | |
| "learning_rate": 5.90794004780439e-06, | |
| "loss": 0.0288, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.8744939271255063, | |
| "grad_norm": 0.24177399514887044, | |
| "learning_rate": 5.867865395552405e-06, | |
| "loss": 0.0321, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.8785425101214575, | |
| "grad_norm": 0.2300000018208566, | |
| "learning_rate": 5.8279037434721364e-06, | |
| "loss": 0.027, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.882591093117409, | |
| "grad_norm": 0.24475523966463855, | |
| "learning_rate": 5.788055411095892e-06, | |
| "loss": 0.0325, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.8866396761133606, | |
| "grad_norm": 0.2387884536618001, | |
| "learning_rate": 5.7483207170498625e-06, | |
| "loss": 0.0291, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8906882591093117, | |
| "grad_norm": 0.26117436024035695, | |
| "learning_rate": 5.708699979051613e-06, | |
| "loss": 0.0326, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.24513478302858585, | |
| "learning_rate": 5.66919351390751e-06, | |
| "loss": 0.0276, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.898785425101215, | |
| "grad_norm": 0.26969765678696955, | |
| "learning_rate": 5.629801637510188e-06, | |
| "loss": 0.0324, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.902834008097166, | |
| "grad_norm": 0.22440951166208645, | |
| "learning_rate": 5.590524664836043e-06, | |
| "loss": 0.0221, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.9068825910931175, | |
| "grad_norm": 0.24884734366945277, | |
| "learning_rate": 5.551362909942708e-06, | |
| "loss": 0.0269, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.910931174089069, | |
| "grad_norm": 0.2506535272190872, | |
| "learning_rate": 5.51231668596653e-06, | |
| "loss": 0.0297, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.91497975708502, | |
| "grad_norm": 0.23210879065742235, | |
| "learning_rate": 5.47338630512007e-06, | |
| "loss": 0.0272, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.919028340080972, | |
| "grad_norm": 0.2687472323593363, | |
| "learning_rate": 5.434572078689617e-06, | |
| "loss": 0.0344, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.2424467259735395, | |
| "learning_rate": 5.395874317032694e-06, | |
| "loss": 0.0283, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.9271255060728745, | |
| "grad_norm": 0.24469952667658154, | |
| "learning_rate": 5.357293329575573e-06, | |
| "loss": 0.0289, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.931174089068826, | |
| "grad_norm": 0.24229793435183208, | |
| "learning_rate": 5.318829424810797e-06, | |
| "loss": 0.0298, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.9352226720647776, | |
| "grad_norm": 0.244968227642017, | |
| "learning_rate": 5.280482910294728e-06, | |
| "loss": 0.0312, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.9392712550607287, | |
| "grad_norm": 0.221294897889363, | |
| "learning_rate": 5.242254092645078e-06, | |
| "loss": 0.027, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.9433198380566803, | |
| "grad_norm": 0.24360547615981312, | |
| "learning_rate": 5.2041432775384445e-06, | |
| "loss": 0.0302, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.23378843455003775, | |
| "learning_rate": 5.166150769707892e-06, | |
| "loss": 0.0276, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.951417004048583, | |
| "grad_norm": 0.22572185770712128, | |
| "learning_rate": 5.128276872940503e-06, | |
| "loss": 0.0252, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.9554655870445345, | |
| "grad_norm": 0.23442553312485512, | |
| "learning_rate": 5.090521890074953e-06, | |
| "loss": 0.0294, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.9595141700404857, | |
| "grad_norm": 0.22966744331705288, | |
| "learning_rate": 5.0528861229990675e-06, | |
| "loss": 0.0253, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.9635627530364372, | |
| "grad_norm": 0.22218775212944292, | |
| "learning_rate": 5.015369872647442e-06, | |
| "loss": 0.0266, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.967611336032389, | |
| "grad_norm": 0.2771183571261403, | |
| "learning_rate": 4.9779734389990195e-06, | |
| "loss": 0.0325, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97165991902834, | |
| "grad_norm": 0.2311687371510812, | |
| "learning_rate": 4.940697121074675e-06, | |
| "loss": 0.0271, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.9757085020242915, | |
| "grad_norm": 0.22488940294698198, | |
| "learning_rate": 4.903541216934864e-06, | |
| "loss": 0.026, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.979757085020243, | |
| "grad_norm": 0.24869380860881143, | |
| "learning_rate": 4.8665060236772e-06, | |
| "loss": 0.0309, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.983805668016194, | |
| "grad_norm": 0.2541467793714208, | |
| "learning_rate": 4.829591837434102e-06, | |
| "loss": 0.0322, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.9878542510121457, | |
| "grad_norm": 0.2226945738738167, | |
| "learning_rate": 4.792798953370424e-06, | |
| "loss": 0.027, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.9919028340080973, | |
| "grad_norm": 0.2485554133026516, | |
| "learning_rate": 4.756127665681083e-06, | |
| "loss": 0.0323, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.9959514170040484, | |
| "grad_norm": 0.22408744007680326, | |
| "learning_rate": 4.719578267588725e-06, | |
| "loss": 0.0262, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.22054406646105573, | |
| "learning_rate": 4.6831510513413525e-06, | |
| "loss": 0.0247, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.004048582995951, | |
| "grad_norm": 0.20371955994245364, | |
| "learning_rate": 4.6468463082100176e-06, | |
| "loss": 0.012, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.008097165991903, | |
| "grad_norm": 0.19866768074328828, | |
| "learning_rate": 4.610664328486485e-06, | |
| "loss": 0.0125, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.012145748987854, | |
| "grad_norm": 0.17077918345732307, | |
| "learning_rate": 4.5746054014809e-06, | |
| "loss": 0.0107, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.016194331983805, | |
| "grad_norm": 0.16635575228406493, | |
| "learning_rate": 4.5386698155194766e-06, | |
| "loss": 0.0121, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.020242914979757, | |
| "grad_norm": 0.1401455153714774, | |
| "learning_rate": 4.502857857942204e-06, | |
| "loss": 0.0107, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.0242914979757085, | |
| "grad_norm": 0.1473790650581133, | |
| "learning_rate": 4.467169815100549e-06, | |
| "loss": 0.0113, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.02834008097166, | |
| "grad_norm": 0.14213892993886568, | |
| "learning_rate": 4.431605972355137e-06, | |
| "loss": 0.0119, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.032388663967612, | |
| "grad_norm": 0.136391634726847, | |
| "learning_rate": 4.396166614073516e-06, | |
| "loss": 0.01, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.036437246963563, | |
| "grad_norm": 0.16368435440304963, | |
| "learning_rate": 4.360852023627846e-06, | |
| "loss": 0.0119, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.040485829959514, | |
| "grad_norm": 0.170634910537191, | |
| "learning_rate": 4.325662483392656e-06, | |
| "loss": 0.0101, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.044534412955466, | |
| "grad_norm": 0.17517521076041476, | |
| "learning_rate": 4.29059827474257e-06, | |
| "loss": 0.0098, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.048582995951417, | |
| "grad_norm": 0.1922387576514228, | |
| "learning_rate": 4.255659678050063e-06, | |
| "loss": 0.0115, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.17498450484958397, | |
| "learning_rate": 4.22084697268323e-06, | |
| "loss": 0.0108, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.05668016194332, | |
| "grad_norm": 0.15744174424854623, | |
| "learning_rate": 4.186160437003535e-06, | |
| "loss": 0.0092, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.060728744939271, | |
| "grad_norm": 0.1756393827258928, | |
| "learning_rate": 4.151600348363582e-06, | |
| "loss": 0.0123, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.064777327935222, | |
| "grad_norm": 0.1715035425440024, | |
| "learning_rate": 4.11716698310493e-06, | |
| "loss": 0.0113, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.068825910931174, | |
| "grad_norm": 0.15240000700750814, | |
| "learning_rate": 4.082860616555848e-06, | |
| "loss": 0.0104, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.0728744939271255, | |
| "grad_norm": 0.14908317277371289, | |
| "learning_rate": 4.048681523029125e-06, | |
| "loss": 0.0098, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.1461066508260859, | |
| "learning_rate": 4.014629975819882e-06, | |
| "loss": 0.0112, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.080971659919029, | |
| "grad_norm": 0.1669549915733771, | |
| "learning_rate": 3.9807062472033855e-06, | |
| "loss": 0.012, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.08502024291498, | |
| "grad_norm": 0.14195472592993524, | |
| "learning_rate": 3.94691060843287e-06, | |
| "loss": 0.01, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.089068825910931, | |
| "grad_norm": 0.15920426639856652, | |
| "learning_rate": 3.913243329737353e-06, | |
| "loss": 0.0137, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.093117408906883, | |
| "grad_norm": 0.1297479668459364, | |
| "learning_rate": 3.879704680319503e-06, | |
| "loss": 0.0099, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.097165991902834, | |
| "grad_norm": 0.13626885230231667, | |
| "learning_rate": 3.846294928353471e-06, | |
| "loss": 0.0123, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.101214574898785, | |
| "grad_norm": 0.13517209890938914, | |
| "learning_rate": 3.8130143409827345e-06, | |
| "loss": 0.0093, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.15560096969347134, | |
| "learning_rate": 3.7798631843179887e-06, | |
| "loss": 0.0113, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.109311740890688, | |
| "grad_norm": 0.14720209758336034, | |
| "learning_rate": 3.7468417234349973e-06, | |
| "loss": 0.0113, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.113360323886639, | |
| "grad_norm": 0.11610646383070639, | |
| "learning_rate": 3.7139502223724823e-06, | |
| "loss": 0.0083, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.117408906882591, | |
| "grad_norm": 0.13530260658081264, | |
| "learning_rate": 3.6811889441300095e-06, | |
| "loss": 0.01, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.1214574898785425, | |
| "grad_norm": 0.1546368113755922, | |
| "learning_rate": 3.648558150665886e-06, | |
| "loss": 0.0123, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.125506072874494, | |
| "grad_norm": 0.1428938215729518, | |
| "learning_rate": 3.616058102895075e-06, | |
| "loss": 0.0105, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.129554655870446, | |
| "grad_norm": 0.1485757169128447, | |
| "learning_rate": 3.583689060687079e-06, | |
| "loss": 0.0099, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.133603238866397, | |
| "grad_norm": 0.14590619696859067, | |
| "learning_rate": 3.5514512828639023e-06, | |
| "loss": 0.0116, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.137651821862348, | |
| "grad_norm": 0.15198662044846317, | |
| "learning_rate": 3.51934502719796e-06, | |
| "loss": 0.0119, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.1417004048583, | |
| "grad_norm": 0.1324197026076446, | |
| "learning_rate": 3.4873705504100185e-06, | |
| "loss": 0.0092, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.145748987854251, | |
| "grad_norm": 0.14585588276581982, | |
| "learning_rate": 3.4555281081671345e-06, | |
| "loss": 0.0077, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.149797570850202, | |
| "grad_norm": 0.14274927206206617, | |
| "learning_rate": 3.4238179550806326e-06, | |
| "loss": 0.0101, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.16662436568582625, | |
| "learning_rate": 3.3922403447040497e-06, | |
| "loss": 0.0118, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.14826032444897203, | |
| "learning_rate": 3.3607955295311247e-06, | |
| "loss": 0.0092, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.161943319838056, | |
| "grad_norm": 0.13985667298864743, | |
| "learning_rate": 3.32948376099375e-06, | |
| "loss": 0.0097, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.165991902834008, | |
| "grad_norm": 0.13334277357692417, | |
| "learning_rate": 3.2983052894599976e-06, | |
| "loss": 0.0089, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.17004048582996, | |
| "grad_norm": 0.14186870713968672, | |
| "learning_rate": 3.2672603642320945e-06, | |
| "loss": 0.0098, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.174089068825911, | |
| "grad_norm": 0.15536441164910983, | |
| "learning_rate": 3.236349233544438e-06, | |
| "loss": 0.0113, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.178137651821863, | |
| "grad_norm": 0.13906436662064445, | |
| "learning_rate": 3.205572144561604e-06, | |
| "loss": 0.0102, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.182186234817814, | |
| "grad_norm": 0.15869061153264416, | |
| "learning_rate": 3.174929343376374e-06, | |
| "loss": 0.0119, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.186234817813765, | |
| "grad_norm": 0.14471016427374767, | |
| "learning_rate": 3.1444210750077777e-06, | |
| "loss": 0.0113, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.190283400809717, | |
| "grad_norm": 0.14528911979682707, | |
| "learning_rate": 3.1140475833991046e-06, | |
| "loss": 0.0107, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.194331983805668, | |
| "grad_norm": 0.14050503604458717, | |
| "learning_rate": 3.0838091114159895e-06, | |
| "loss": 0.0105, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.198380566801619, | |
| "grad_norm": 0.14571761830538685, | |
| "learning_rate": 3.0537059008444524e-06, | |
| "loss": 0.0107, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.202429149797571, | |
| "grad_norm": 0.13216267639107204, | |
| "learning_rate": 3.0237381923889563e-06, | |
| "loss": 0.0095, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.206477732793522, | |
| "grad_norm": 0.12332747692787496, | |
| "learning_rate": 2.9939062256704997e-06, | |
| "loss": 0.0088, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.1377289250514664, | |
| "learning_rate": 2.964210239224696e-06, | |
| "loss": 0.0103, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.2145748987854255, | |
| "grad_norm": 0.12832515162059027, | |
| "learning_rate": 2.934650470499869e-06, | |
| "loss": 0.0099, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.218623481781377, | |
| "grad_norm": 0.15841954790684326, | |
| "learning_rate": 2.905227155855128e-06, | |
| "loss": 0.0128, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.222672064777328, | |
| "grad_norm": 0.14695945418968231, | |
| "learning_rate": 2.875940530558519e-06, | |
| "loss": 0.0111, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.22672064777328, | |
| "grad_norm": 0.13396735055372241, | |
| "learning_rate": 2.8467908287851154e-06, | |
| "loss": 0.009, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.14709525587698397, | |
| "learning_rate": 2.8177782836151603e-06, | |
| "loss": 0.0097, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.234817813765182, | |
| "grad_norm": 0.14395474098505487, | |
| "learning_rate": 2.7889031270321874e-06, | |
| "loss": 0.01, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.238866396761134, | |
| "grad_norm": 0.14685961755572974, | |
| "learning_rate": 2.760165589921184e-06, | |
| "loss": 0.0118, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.242914979757085, | |
| "grad_norm": 0.13367611817260355, | |
| "learning_rate": 2.7315659020667305e-06, | |
| "loss": 0.0102, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.246963562753036, | |
| "grad_norm": 0.12200784143787627, | |
| "learning_rate": 2.703104292151164e-06, | |
| "loss": 0.0078, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.251012145748988, | |
| "grad_norm": 0.14521693890348067, | |
| "learning_rate": 2.6747809877527607e-06, | |
| "loss": 0.0103, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.255060728744939, | |
| "grad_norm": 0.14265232441432707, | |
| "learning_rate": 2.6465962153439082e-06, | |
| "loss": 0.0095, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.2591093117408905, | |
| "grad_norm": 0.14882793402463995, | |
| "learning_rate": 2.618550200289303e-06, | |
| "loss": 0.0113, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.1462744399929931, | |
| "learning_rate": 2.59064316684412e-06, | |
| "loss": 0.0111, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.267206477732794, | |
| "grad_norm": 0.133228896043334, | |
| "learning_rate": 2.562875338152264e-06, | |
| "loss": 0.0099, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.271255060728745, | |
| "grad_norm": 0.12866017446219807, | |
| "learning_rate": 2.5352469362445552e-06, | |
| "loss": 0.0089, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.275303643724697, | |
| "grad_norm": 0.12198271647583872, | |
| "learning_rate": 2.5077581820369523e-06, | |
| "loss": 0.0087, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.279352226720648, | |
| "grad_norm": 0.14125874686021123, | |
| "learning_rate": 2.480409295328803e-06, | |
| "loss": 0.0084, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.283400809716599, | |
| "grad_norm": 0.13147426774586868, | |
| "learning_rate": 2.45320049480108e-06, | |
| "loss": 0.0091, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.287449392712551, | |
| "grad_norm": 0.12646612934700904, | |
| "learning_rate": 2.4261319980146293e-06, | |
| "loss": 0.0088, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.291497975708502, | |
| "grad_norm": 0.13047424696617388, | |
| "learning_rate": 2.399204021408421e-06, | |
| "loss": 0.0097, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.295546558704453, | |
| "grad_norm": 0.14152001605813855, | |
| "learning_rate": 2.3724167802978395e-06, | |
| "loss": 0.011, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.299595141700405, | |
| "grad_norm": 0.13967427013663206, | |
| "learning_rate": 2.3457704888729625e-06, | |
| "loss": 0.0085, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.303643724696356, | |
| "grad_norm": 0.1378949615364226, | |
| "learning_rate": 2.3192653601968162e-06, | |
| "loss": 0.0098, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.14134913277090017, | |
| "learning_rate": 2.2929016062037057e-06, | |
| "loss": 0.009, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.3117408906882595, | |
| "grad_norm": 0.13787435710291962, | |
| "learning_rate": 2.2666794376975055e-06, | |
| "loss": 0.0098, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.14200432291562226, | |
| "learning_rate": 2.24059906434998e-06, | |
| "loss": 0.0102, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.319838056680162, | |
| "grad_norm": 0.1434581012679992, | |
| "learning_rate": 2.214660694699091e-06, | |
| "loss": 0.0105, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.323886639676114, | |
| "grad_norm": 0.13165738407314154, | |
| "learning_rate": 2.188864536147355e-06, | |
| "loss": 0.0093, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.327935222672065, | |
| "grad_norm": 0.12811722851791585, | |
| "learning_rate": 2.1632107949601757e-06, | |
| "loss": 0.0092, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.331983805668016, | |
| "grad_norm": 0.1346169607145899, | |
| "learning_rate": 2.137699676264171e-06, | |
| "loss": 0.0099, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.336032388663968, | |
| "grad_norm": 0.1419453599939057, | |
| "learning_rate": 2.1123313840455738e-06, | |
| "loss": 0.0102, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.340080971659919, | |
| "grad_norm": 0.130679930767489, | |
| "learning_rate": 2.08710612114857e-06, | |
| "loss": 0.0086, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.34412955465587, | |
| "grad_norm": 0.1469804020098759, | |
| "learning_rate": 2.0620240892736953e-06, | |
| "loss": 0.0107, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.348178137651822, | |
| "grad_norm": 0.12470233826101945, | |
| "learning_rate": 2.0370854889762e-06, | |
| "loss": 0.0096, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.352226720647773, | |
| "grad_norm": 0.15070581411387604, | |
| "learning_rate": 2.0122905196644726e-06, | |
| "loss": 0.011, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3562753036437245, | |
| "grad_norm": 0.15738966215572384, | |
| "learning_rate": 1.987639379598427e-06, | |
| "loss": 0.0113, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.3603238866396765, | |
| "grad_norm": 0.15470061680839808, | |
| "learning_rate": 1.963132265887919e-06, | |
| "loss": 0.0094, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.364372469635628, | |
| "grad_norm": 0.3062020908176644, | |
| "learning_rate": 1.9387693744911784e-06, | |
| "loss": 0.0086, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.15007565427589425, | |
| "learning_rate": 1.9145509002132368e-06, | |
| "loss": 0.0108, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.372469635627531, | |
| "grad_norm": 0.13453780586907893, | |
| "learning_rate": 1.8904770367043723e-06, | |
| "loss": 0.009, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.376518218623482, | |
| "grad_norm": 0.11065108711990922, | |
| "learning_rate": 1.8665479764585504e-06, | |
| "loss": 0.0075, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.380566801619433, | |
| "grad_norm": 0.14670116930341254, | |
| "learning_rate": 1.8427639108118977e-06, | |
| "loss": 0.0107, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.13485953452666688, | |
| "learning_rate": 1.8191250299411755e-06, | |
| "loss": 0.009, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.388663967611336, | |
| "grad_norm": 0.12931706925305644, | |
| "learning_rate": 1.7956315228622446e-06, | |
| "loss": 0.0094, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.392712550607287, | |
| "grad_norm": 0.1723708305175637, | |
| "learning_rate": 1.7722835774285553e-06, | |
| "loss": 0.0131, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.396761133603239, | |
| "grad_norm": 0.14604117836904942, | |
| "learning_rate": 1.7490813803296624e-06, | |
| "loss": 0.0111, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.40080971659919, | |
| "grad_norm": 0.13769348598618616, | |
| "learning_rate": 1.726025117089718e-06, | |
| "loss": 0.0091, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.4048582995951415, | |
| "grad_norm": 0.13533747740178362, | |
| "learning_rate": 1.70311497206598e-06, | |
| "loss": 0.0092, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.4089068825910935, | |
| "grad_norm": 0.15182869054624035, | |
| "learning_rate": 1.6803511284473638e-06, | |
| "loss": 0.0115, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.412955465587045, | |
| "grad_norm": 0.1361983037372098, | |
| "learning_rate": 1.6577337682529538e-06, | |
| "loss": 0.011, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.417004048582996, | |
| "grad_norm": 0.13809441894275623, | |
| "learning_rate": 1.6352630723305617e-06, | |
| "loss": 0.0091, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.1214681270390892, | |
| "learning_rate": 1.612939220355272e-06, | |
| "loss": 0.0081, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.425101214574899, | |
| "grad_norm": 0.13522246193803203, | |
| "learning_rate": 1.5907623908280112e-06, | |
| "loss": 0.0096, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.42914979757085, | |
| "grad_norm": 0.14504387621688156, | |
| "learning_rate": 1.568732761074121e-06, | |
| "loss": 0.0107, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.433198380566802, | |
| "grad_norm": 0.11315488204154023, | |
| "learning_rate": 1.5468505072419237e-06, | |
| "loss": 0.0077, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.437246963562753, | |
| "grad_norm": 0.1199839206706225, | |
| "learning_rate": 1.5251158043013448e-06, | |
| "loss": 0.008, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.441295546558704, | |
| "grad_norm": 0.14560696129142447, | |
| "learning_rate": 1.5035288260424885e-06, | |
| "loss": 0.0106, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.445344129554655, | |
| "grad_norm": 0.12640235250130455, | |
| "learning_rate": 1.4820897450742623e-06, | |
| "loss": 0.0088, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.449392712550607, | |
| "grad_norm": 0.1395104075450439, | |
| "learning_rate": 1.4607987328229767e-06, | |
| "loss": 0.011, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.4534412955465585, | |
| "grad_norm": 0.12096705021256599, | |
| "learning_rate": 1.4396559595310056e-06, | |
| "loss": 0.0084, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.4574898785425106, | |
| "grad_norm": 0.1294873265890372, | |
| "learning_rate": 1.4186615942553973e-06, | |
| "loss": 0.0098, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.140389945698987, | |
| "learning_rate": 1.3978158048665424e-06, | |
| "loss": 0.0106, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.465587044534413, | |
| "grad_norm": 0.14335415089203743, | |
| "learning_rate": 1.3771187580468115e-06, | |
| "loss": 0.0097, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.469635627530364, | |
| "grad_norm": 0.1331555406740756, | |
| "learning_rate": 1.3565706192892392e-06, | |
| "loss": 0.0091, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.13981385413300754, | |
| "learning_rate": 1.3361715528961993e-06, | |
| "loss": 0.0092, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.477732793522267, | |
| "grad_norm": 0.15543453567650006, | |
| "learning_rate": 1.3159217219780817e-06, | |
| "loss": 0.0123, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.481781376518219, | |
| "grad_norm": 0.15363558287750484, | |
| "learning_rate": 1.2958212884519949e-06, | |
| "loss": 0.0119, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.48582995951417, | |
| "grad_norm": 0.15501510376842173, | |
| "learning_rate": 1.2758704130404675e-06, | |
| "loss": 0.0114, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.489878542510121, | |
| "grad_norm": 0.1350243308209304, | |
| "learning_rate": 1.2560692552701738e-06, | |
| "loss": 0.0097, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.493927125506072, | |
| "grad_norm": 0.21199619003722486, | |
| "learning_rate": 1.2364179734706382e-06, | |
| "loss": 0.0097, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.497975708502024, | |
| "grad_norm": 0.13737685553845683, | |
| "learning_rate": 1.2169167247729852e-06, | |
| "loss": 0.0103, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.502024291497976, | |
| "grad_norm": 0.13290722405772276, | |
| "learning_rate": 1.1975656651086864e-06, | |
| "loss": 0.0096, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.506072874493928, | |
| "grad_norm": 0.13461267865017346, | |
| "learning_rate": 1.1783649492082926e-06, | |
| "loss": 0.0097, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.510121457489879, | |
| "grad_norm": 0.1315490105457004, | |
| "learning_rate": 1.1593147306002183e-06, | |
| "loss": 0.0085, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.51417004048583, | |
| "grad_norm": 0.13452830816679334, | |
| "learning_rate": 1.140415161609505e-06, | |
| "loss": 0.0096, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.518218623481781, | |
| "grad_norm": 0.13851801822832382, | |
| "learning_rate": 1.121666393356604e-06, | |
| "loss": 0.0105, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.522267206477733, | |
| "grad_norm": 0.13401278163460878, | |
| "learning_rate": 1.1030685757561632e-06, | |
| "loss": 0.0096, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.13892218072822068, | |
| "learning_rate": 1.0846218575158373e-06, | |
| "loss": 0.0095, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.530364372469636, | |
| "grad_norm": 0.13514646220194168, | |
| "learning_rate": 1.066326386135097e-06, | |
| "loss": 0.0099, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.534412955465587, | |
| "grad_norm": 0.12082536218895079, | |
| "learning_rate": 1.0481823079040398e-06, | |
| "loss": 0.0084, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.1264931661097679, | |
| "learning_rate": 1.0301897679022344e-06, | |
| "loss": 0.0088, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.5425101214574894, | |
| "grad_norm": 0.13333859804773432, | |
| "learning_rate": 1.0123489099975491e-06, | |
| "loss": 0.0103, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.5465587044534415, | |
| "grad_norm": 0.14411520669979547, | |
| "learning_rate": 9.9465987684501e-07, | |
| "loss": 0.0109, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.550607287449393, | |
| "grad_norm": 0.14287787657264028, | |
| "learning_rate": 9.771228098856534e-07, | |
| "loss": 0.0107, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.554655870445345, | |
| "grad_norm": 0.12053578233197319, | |
| "learning_rate": 9.597378493454013e-07, | |
| "loss": 0.0086, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.558704453441296, | |
| "grad_norm": 0.13363164684282242, | |
| "learning_rate": 9.425051342339353e-07, | |
| "loss": 0.0101, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.562753036437247, | |
| "grad_norm": 0.1258747410581103, | |
| "learning_rate": 9.254248023435864e-07, | |
| "loss": 0.0079, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.566801619433198, | |
| "grad_norm": 0.12420983302059593, | |
| "learning_rate": 9.084969902482288e-07, | |
| "loss": 0.009, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.57085020242915, | |
| "grad_norm": 0.11545740227758816, | |
| "learning_rate": 8.917218333021993e-07, | |
| "loss": 0.0086, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.574898785425101, | |
| "grad_norm": 0.12964832122457454, | |
| "learning_rate": 8.750994656392109e-07, | |
| "loss": 0.0093, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.12930245702243826, | |
| "learning_rate": 8.58630020171265e-07, | |
| "loss": 0.009, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.582995951417004, | |
| "grad_norm": 0.14633633741509744, | |
| "learning_rate": 8.423136285876099e-07, | |
| "loss": 0.0105, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.587044534412955, | |
| "grad_norm": 0.12792668404475804, | |
| "learning_rate": 8.26150421353682e-07, | |
| "loss": 0.0085, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.5910931174089065, | |
| "grad_norm": 0.13416685116208066, | |
| "learning_rate": 8.101405277100549e-07, | |
| "loss": 0.0104, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.5951417004048585, | |
| "grad_norm": 0.13077952921647173, | |
| "learning_rate": 7.942840756714077e-07, | |
| "loss": 0.0096, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.59919028340081, | |
| "grad_norm": 0.11909649960621903, | |
| "learning_rate": 7.785811920255093e-07, | |
| "loss": 0.008, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.603238866396762, | |
| "grad_norm": 0.1304140833572685, | |
| "learning_rate": 7.630320023322024e-07, | |
| "loss": 0.0087, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.607287449392713, | |
| "grad_norm": 0.1347439369366911, | |
| "learning_rate": 7.476366309223903e-07, | |
| "loss": 0.0101, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.611336032388664, | |
| "grad_norm": 0.11913624770007529, | |
| "learning_rate": 7.323952008970537e-07, | |
| "loss": 0.0073, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.15374418260371822, | |
| "learning_rate": 7.173078341262618e-07, | |
| "loss": 0.011, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.619433198380567, | |
| "grad_norm": 0.1420093989809516, | |
| "learning_rate": 7.023746512481989e-07, | |
| "loss": 0.0097, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.623481781376518, | |
| "grad_norm": 0.10565448318635932, | |
| "learning_rate": 6.875957716681902e-07, | |
| "loss": 0.0072, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.62753036437247, | |
| "grad_norm": 0.13382127125694934, | |
| "learning_rate": 6.729713135577687e-07, | |
| "loss": 0.0102, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.14497244064690074, | |
| "learning_rate": 6.585013938537143e-07, | |
| "loss": 0.0106, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.635627530364372, | |
| "grad_norm": 0.13777845735341343, | |
| "learning_rate": 6.441861282571115e-07, | |
| "loss": 0.0103, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.6396761133603235, | |
| "grad_norm": 0.14086379701590535, | |
| "learning_rate": 6.300256312324471e-07, | |
| "loss": 0.0107, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.6437246963562755, | |
| "grad_norm": 0.14249939647897553, | |
| "learning_rate": 6.160200160066776e-07, | |
| "loss": 0.0107, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.647773279352227, | |
| "grad_norm": 0.1217130966018751, | |
| "learning_rate": 6.021693945683327e-07, | |
| "loss": 0.0081, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.651821862348179, | |
| "grad_norm": 0.15189793566578852, | |
| "learning_rate": 5.884738776666044e-07, | |
| "loss": 0.0114, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.65587044534413, | |
| "grad_norm": 0.13551106073496236, | |
| "learning_rate": 5.749335748104878e-07, | |
| "loss": 0.0115, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.659919028340081, | |
| "grad_norm": 0.1286845373409889, | |
| "learning_rate": 5.615485942678778e-07, | |
| "loss": 0.0094, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.663967611336032, | |
| "grad_norm": 0.1396878853476864, | |
| "learning_rate": 5.483190430647289e-07, | |
| "loss": 0.0102, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.668016194331984, | |
| "grad_norm": 0.1250185118838258, | |
| "learning_rate": 5.352450269841747e-07, | |
| "loss": 0.0072, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.672064777327935, | |
| "grad_norm": 0.1210332661867685, | |
| "learning_rate": 5.22326650565701e-07, | |
| "loss": 0.0081, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.676113360323887, | |
| "grad_norm": 0.1427519660745095, | |
| "learning_rate": 5.095640171043048e-07, | |
| "loss": 0.0087, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.680161943319838, | |
| "grad_norm": 0.13480261701245977, | |
| "learning_rate": 4.969572286496594e-07, | |
| "loss": 0.0091, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.13176895966075935, | |
| "learning_rate": 4.845063860053079e-07, | |
| "loss": 0.0087, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.6882591093117405, | |
| "grad_norm": 0.13023850734922265, | |
| "learning_rate": 4.7221158872785997e-07, | |
| "loss": 0.0087, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 0.14650603560910638, | |
| "learning_rate": 4.600729351261857e-07, | |
| "loss": 0.0102, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.696356275303644, | |
| "grad_norm": 0.12930013974657575, | |
| "learning_rate": 4.4809052226063355e-07, | |
| "loss": 0.0089, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.700404858299595, | |
| "grad_norm": 0.15287275020227598, | |
| "learning_rate": 4.3626444594226046e-07, | |
| "loss": 0.0107, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.704453441295547, | |
| "grad_norm": 0.14308889585709947, | |
| "learning_rate": 4.245948007320588e-07, | |
| "loss": 0.0101, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.708502024291498, | |
| "grad_norm": 0.14015218396608906, | |
| "learning_rate": 4.130816799401971e-07, | |
| "loss": 0.0107, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.712550607287449, | |
| "grad_norm": 0.13845585539113459, | |
| "learning_rate": 4.017251756252827e-07, | |
| "loss": 0.0103, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.716599190283401, | |
| "grad_norm": 0.1399052865574843, | |
| "learning_rate": 3.9052537859362473e-07, | |
| "loss": 0.0102, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.720647773279352, | |
| "grad_norm": 0.13783952256713264, | |
| "learning_rate": 3.7948237839849687e-07, | |
| "loss": 0.0088, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.724696356275303, | |
| "grad_norm": 0.1592176738878414, | |
| "learning_rate": 3.6859626333943797e-07, | |
| "loss": 0.0131, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.728744939271255, | |
| "grad_norm": 0.12092852096770772, | |
| "learning_rate": 3.578671204615347e-07, | |
| "loss": 0.0087, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.732793522267206, | |
| "grad_norm": 0.13203076072526815, | |
| "learning_rate": 3.472950355547267e-07, | |
| "loss": 0.0088, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.13136170926192742, | |
| "learning_rate": 3.3688009315312244e-07, | |
| "loss": 0.0097, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.7408906882591095, | |
| "grad_norm": 0.13712212655924413, | |
| "learning_rate": 3.266223765343224e-07, | |
| "loss": 0.0098, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.744939271255061, | |
| "grad_norm": 0.13487812298425933, | |
| "learning_rate": 3.16521967718757e-07, | |
| "loss": 0.0096, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.748987854251012, | |
| "grad_norm": 0.12850820750748695, | |
| "learning_rate": 3.0657894746902726e-07, | |
| "loss": 0.0086, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.753036437246964, | |
| "grad_norm": 0.11077600809932152, | |
| "learning_rate": 2.9679339528925654e-07, | |
| "loss": 0.0071, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.757085020242915, | |
| "grad_norm": 0.12809407570954262, | |
| "learning_rate": 2.871653894244597e-07, | |
| "loss": 0.01, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.761133603238866, | |
| "grad_norm": 0.13028962405179959, | |
| "learning_rate": 2.7769500685991266e-07, | |
| "loss": 0.008, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.765182186234818, | |
| "grad_norm": 0.12935462613990528, | |
| "learning_rate": 2.683823233205485e-07, | |
| "loss": 0.0088, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.1434972756685281, | |
| "learning_rate": 2.592274132703287e-07, | |
| "loss": 0.0098, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.77327935222672, | |
| "grad_norm": 0.12084060944494333, | |
| "learning_rate": 2.502303499116754e-07, | |
| "loss": 0.0079, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.777327935222672, | |
| "grad_norm": 0.13677075481427253, | |
| "learning_rate": 2.413912051848666e-07, | |
| "loss": 0.0094, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.781376518218623, | |
| "grad_norm": 0.1404359629572804, | |
| "learning_rate": 2.3271004976746613e-07, | |
| "loss": 0.0093, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.7854251012145745, | |
| "grad_norm": 0.1358564068558122, | |
| "learning_rate": 2.2418695307376393e-07, | |
| "loss": 0.0095, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.14111100344173463, | |
| "learning_rate": 2.1582198325420968e-07, | |
| "loss": 0.009, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.793522267206478, | |
| "grad_norm": 0.11580354527592233, | |
| "learning_rate": 2.0761520719488447e-07, | |
| "loss": 0.0089, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.797570850202429, | |
| "grad_norm": 0.1328931562904537, | |
| "learning_rate": 1.9956669051694798e-07, | |
| "loss": 0.0084, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.801619433198381, | |
| "grad_norm": 0.13601127261236606, | |
| "learning_rate": 1.916764975761254e-07, | |
| "loss": 0.0095, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.805668016194332, | |
| "grad_norm": 0.13654479197112798, | |
| "learning_rate": 1.839446914621923e-07, | |
| "loss": 0.0113, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.809716599190283, | |
| "grad_norm": 0.13143931136986953, | |
| "learning_rate": 1.7637133399845962e-07, | |
| "loss": 0.0099, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.813765182186235, | |
| "grad_norm": 0.12627069899995413, | |
| "learning_rate": 1.6895648574129397e-07, | |
| "loss": 0.0088, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.817813765182186, | |
| "grad_norm": 0.12204405685407808, | |
| "learning_rate": 1.6170020597962465e-07, | |
| "loss": 0.0084, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.821862348178137, | |
| "grad_norm": 0.11484788573462601, | |
| "learning_rate": 1.5460255273447078e-07, | |
| "loss": 0.0083, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.825910931174089, | |
| "grad_norm": 0.12276691316818931, | |
| "learning_rate": 1.4766358275847715e-07, | |
| "loss": 0.0088, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.82995951417004, | |
| "grad_norm": 0.1247730825421336, | |
| "learning_rate": 1.4088335153546573e-07, | |
| "loss": 0.009, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.834008097165992, | |
| "grad_norm": 0.1516413929943811, | |
| "learning_rate": 1.3426191327998496e-07, | |
| "loss": 0.0112, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.838056680161944, | |
| "grad_norm": 0.1542143058469344, | |
| "learning_rate": 1.277993209368744e-07, | |
| "loss": 0.0122, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.13299699546733856, | |
| "learning_rate": 1.214956261808542e-07, | |
| "loss": 0.0101, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.14182165891563583, | |
| "learning_rate": 1.1535087941609846e-07, | |
| "loss": 0.0109, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.850202429149798, | |
| "grad_norm": 0.11385582753092119, | |
| "learning_rate": 1.0936512977583802e-07, | |
| "loss": 0.0076, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.854251012145749, | |
| "grad_norm": 0.14741716580696576, | |
| "learning_rate": 1.0353842512196732e-07, | |
| "loss": 0.0102, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.8582995951417, | |
| "grad_norm": 0.13673906454615714, | |
| "learning_rate": 9.787081204466253e-08, | |
| "loss": 0.0103, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.862348178137652, | |
| "grad_norm": 0.12524520677414064, | |
| "learning_rate": 9.236233586200627e-08, | |
| "loss": 0.0085, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.866396761133603, | |
| "grad_norm": 0.12635233810226854, | |
| "learning_rate": 8.70130406196279e-08, | |
| "loss": 0.0086, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.870445344129554, | |
| "grad_norm": 0.12489351474996944, | |
| "learning_rate": 8.182296909035047e-08, | |
| "loss": 0.0088, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.874493927125506, | |
| "grad_norm": 0.1331521195077243, | |
| "learning_rate": 7.679216277384439e-08, | |
| "loss": 0.0099, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.8785425101214575, | |
| "grad_norm": 0.12238877445353931, | |
| "learning_rate": 7.19206618963053e-08, | |
| "loss": 0.0083, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.882591093117409, | |
| "grad_norm": 0.14037798094312645, | |
| "learning_rate": 6.720850541012347e-08, | |
| "loss": 0.0101, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.886639676113361, | |
| "grad_norm": 0.13813593417707254, | |
| "learning_rate": 6.265573099357714e-08, | |
| "loss": 0.0102, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.890688259109312, | |
| "grad_norm": 0.12473370909331241, | |
| "learning_rate": 5.826237505053067e-08, | |
| "loss": 0.0089, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.12918603831190886, | |
| "learning_rate": 5.4028472710137e-08, | |
| "loss": 0.0092, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.898785425101215, | |
| "grad_norm": 0.1201342952232311, | |
| "learning_rate": 4.9954057826571146e-08, | |
| "loss": 0.008, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.902834008097166, | |
| "grad_norm": 0.13011875332663725, | |
| "learning_rate": 4.6039162978743776e-08, | |
| "loss": 0.0096, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.906882591093117, | |
| "grad_norm": 0.11591349184351346, | |
| "learning_rate": 4.228381947004812e-08, | |
| "loss": 0.0087, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.910931174089069, | |
| "grad_norm": 0.14171672796545617, | |
| "learning_rate": 3.868805732811343e-08, | |
| "loss": 0.0109, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.91497975708502, | |
| "grad_norm": 0.10786412727398516, | |
| "learning_rate": 3.525190530455636e-08, | |
| "loss": 0.007, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.919028340080971, | |
| "grad_norm": 0.11246380844002148, | |
| "learning_rate": 3.197539087475887e-08, | |
| "loss": 0.0083, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.14053015454121565, | |
| "learning_rate": 2.8858540237641786e-08, | |
| "loss": 0.0109, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.9271255060728745, | |
| "grad_norm": 0.11751417912812889, | |
| "learning_rate": 2.5901378315458247e-08, | |
| "loss": 0.0074, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.931174089068826, | |
| "grad_norm": 0.11485240601814557, | |
| "learning_rate": 2.3103928753598347e-08, | |
| "loss": 0.0083, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.935222672064778, | |
| "grad_norm": 0.12356139510155939, | |
| "learning_rate": 2.0466213920393718e-08, | |
| "loss": 0.0085, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.939271255060729, | |
| "grad_norm": 0.12090586138558747, | |
| "learning_rate": 1.7988254906942115e-08, | |
| "loss": 0.0094, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94331983805668, | |
| "grad_norm": 0.14620674553456045, | |
| "learning_rate": 1.5670071526936447e-08, | |
| "loss": 0.0094, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.12268295659168563, | |
| "learning_rate": 1.3511682316509344e-08, | |
| "loss": 0.0089, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.951417004048583, | |
| "grad_norm": 0.13594351330337037, | |
| "learning_rate": 1.1513104534086605e-08, | |
| "loss": 0.0098, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.955465587044534, | |
| "grad_norm": 0.1152586615995607, | |
| "learning_rate": 9.67435416023843e-09, | |
| "loss": 0.0081, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.959514170040486, | |
| "grad_norm": 0.13546201458958967, | |
| "learning_rate": 7.995445897566178e-09, | |
| "loss": 0.0097, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.963562753036437, | |
| "grad_norm": 0.1273627290833216, | |
| "learning_rate": 6.476393170578021e-09, | |
| "loss": 0.0093, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.967611336032388, | |
| "grad_norm": 0.12677587231211798, | |
| "learning_rate": 5.11720812557348e-09, | |
| "loss": 0.0084, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.97165991902834, | |
| "grad_norm": 0.13147173354348343, | |
| "learning_rate": 3.917901630563492e-09, | |
| "loss": 0.0098, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.9757085020242915, | |
| "grad_norm": 0.1306920100465347, | |
| "learning_rate": 2.878483275170485e-09, | |
| "loss": 0.0098, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.979757085020243, | |
| "grad_norm": 0.15672861251104997, | |
| "learning_rate": 1.998961370555108e-09, | |
| "loss": 0.0123, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.983805668016195, | |
| "grad_norm": 0.11180231382658604, | |
| "learning_rate": 1.2793429493518361e-09, | |
| "loss": 0.0078, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.987854251012146, | |
| "grad_norm": 0.11330540343611202, | |
| "learning_rate": 7.196337656112384e-10, | |
| "loss": 0.0082, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.991902834008097, | |
| "grad_norm": 0.1322457912532674, | |
| "learning_rate": 3.198382947511292e-10, | |
| "loss": 0.0097, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.995951417004049, | |
| "grad_norm": 0.13876739765359994, | |
| "learning_rate": 7.995973352770137e-11, | |
| "loss": 0.0108, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.12226882667684377, | |
| "learning_rate": 0.0, | |
| "loss": 0.0079, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1235, | |
| "total_flos": 721077208055808.0, | |
| "train_loss": 0.26725307092115824, | |
| "train_runtime": 6318.7629, | |
| "train_samples_per_second": 25.005, | |
| "train_steps_per_second": 0.195 | |
| } | |
| ], | |
| "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": 721077208055808.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |