| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1235, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004048582995951417, | |
| "grad_norm": 6.372999775536321, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 1.0257, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008097165991902834, | |
| "grad_norm": 6.56674922274851, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 1.0483, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012145748987854251, | |
| "grad_norm": 6.373509146795322, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 1.0435, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016194331983805668, | |
| "grad_norm": 6.352216977730505, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 1.0494, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020242914979757085, | |
| "grad_norm": 5.946492139217835, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 1.0271, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024291497975708502, | |
| "grad_norm": 5.889056555557102, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.044, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02834008097165992, | |
| "grad_norm": 4.473750482599924, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 0.9843, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032388663967611336, | |
| "grad_norm": 4.205269992856126, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 0.9898, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03643724696356275, | |
| "grad_norm": 2.729667793257588, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 0.9669, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04048582995951417, | |
| "grad_norm": 2.552488293224467, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.9532, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044534412955465584, | |
| "grad_norm": 2.4527640448677386, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 0.9507, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048582995951417005, | |
| "grad_norm": 3.8432695613920314, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 0.935, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 4.230015747560021, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 0.9444, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05668016194331984, | |
| "grad_norm": 4.019960070363596, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 0.9312, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06072874493927125, | |
| "grad_norm": 3.7137002527213374, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 0.9309, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06477732793522267, | |
| "grad_norm": 2.7219898680125243, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 0.8797, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06882591093117409, | |
| "grad_norm": 2.4987694725322265, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 0.8734, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0728744939271255, | |
| "grad_norm": 2.0745068006878173, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 0.8639, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.600576863973362, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 0.8614, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08097165991902834, | |
| "grad_norm": 1.1046803449845817, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.8438, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08502024291497975, | |
| "grad_norm": 1.0556630020519169, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.8282, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08906882591093117, | |
| "grad_norm": 1.1977535584532837, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.801, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0931174089068826, | |
| "grad_norm": 1.129551884690838, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.7892, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09716599190283401, | |
| "grad_norm": 0.9347463309836743, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.7927, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10121457489878542, | |
| "grad_norm": 0.9064126111606521, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.7879, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 0.8655276208216961, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.781, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10931174089068826, | |
| "grad_norm": 0.8825316131732711, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.7595, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11336032388663968, | |
| "grad_norm": 0.7809096276467044, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.7807, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11740890688259109, | |
| "grad_norm": 0.8389974420946148, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.7655, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1214574898785425, | |
| "grad_norm": 0.7967468705470595, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.7687, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12550607287449392, | |
| "grad_norm": 0.8016834013067525, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7429, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12955465587044535, | |
| "grad_norm": 0.6821473685377043, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.7437, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13360323886639677, | |
| "grad_norm": 0.7789145055117247, | |
| "learning_rate": 1.0645161290322582e-05, | |
| "loss": 0.7388, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13765182186234817, | |
| "grad_norm": 0.8275627674919379, | |
| "learning_rate": 1.096774193548387e-05, | |
| "loss": 0.7476, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1417004048582996, | |
| "grad_norm": 0.6858028466109612, | |
| "learning_rate": 1.1290322580645164e-05, | |
| "loss": 0.7267, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.145748987854251, | |
| "grad_norm": 0.7303965554581344, | |
| "learning_rate": 1.1612903225806453e-05, | |
| "loss": 0.7205, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14979757085020243, | |
| "grad_norm": 0.7249343327437402, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.7086, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.6041262985297398, | |
| "learning_rate": 1.2258064516129034e-05, | |
| "loss": 0.7193, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 0.7151039998851911, | |
| "learning_rate": 1.2580645161290324e-05, | |
| "loss": 0.7367, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16194331983805668, | |
| "grad_norm": 0.5469598969443545, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.7229, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1659919028340081, | |
| "grad_norm": 0.7540757185047104, | |
| "learning_rate": 1.3225806451612903e-05, | |
| "loss": 0.7073, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1700404858299595, | |
| "grad_norm": 0.5443906878401364, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.7085, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17408906882591094, | |
| "grad_norm": 0.5969600487119051, | |
| "learning_rate": 1.3870967741935486e-05, | |
| "loss": 0.7088, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17813765182186234, | |
| "grad_norm": 0.5834262627186286, | |
| "learning_rate": 1.4193548387096776e-05, | |
| "loss": 0.7071, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18218623481781376, | |
| "grad_norm": 0.7017624848787908, | |
| "learning_rate": 1.4516129032258066e-05, | |
| "loss": 0.7145, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1862348178137652, | |
| "grad_norm": 0.45834064769370203, | |
| "learning_rate": 1.4838709677419357e-05, | |
| "loss": 0.7035, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1902834008097166, | |
| "grad_norm": 0.5830371295385899, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.7131, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19433198380566802, | |
| "grad_norm": 0.5795895237696818, | |
| "learning_rate": 1.5483870967741936e-05, | |
| "loss": 0.7111, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19838056680161945, | |
| "grad_norm": 0.5758578625842882, | |
| "learning_rate": 1.5806451612903226e-05, | |
| "loss": 0.6865, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20242914979757085, | |
| "grad_norm": 0.5220718363805363, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.7049, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20647773279352227, | |
| "grad_norm": 0.5727985333126907, | |
| "learning_rate": 1.6451612903225807e-05, | |
| "loss": 0.6903, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.4613786501784483, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.704, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2145748987854251, | |
| "grad_norm": 0.6873739500032742, | |
| "learning_rate": 1.7096774193548388e-05, | |
| "loss": 0.6877, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21862348178137653, | |
| "grad_norm": 0.4955929425888984, | |
| "learning_rate": 1.741935483870968e-05, | |
| "loss": 0.6933, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22267206477732793, | |
| "grad_norm": 0.6446911354962467, | |
| "learning_rate": 1.774193548387097e-05, | |
| "loss": 0.7118, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22672064777327935, | |
| "grad_norm": 0.6956374929229022, | |
| "learning_rate": 1.806451612903226e-05, | |
| "loss": 0.6922, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.4343219370573566, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.6959, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23481781376518218, | |
| "grad_norm": 0.6445633806406904, | |
| "learning_rate": 1.870967741935484e-05, | |
| "loss": 0.6778, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2388663967611336, | |
| "grad_norm": 0.522123681755494, | |
| "learning_rate": 1.903225806451613e-05, | |
| "loss": 0.6767, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.242914979757085, | |
| "grad_norm": 0.5146040065598788, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.6622, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24696356275303644, | |
| "grad_norm": 0.6737001760950656, | |
| "learning_rate": 1.967741935483871e-05, | |
| "loss": 0.6877, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.25101214574898784, | |
| "grad_norm": 0.6161273383400716, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6728, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2550607287449393, | |
| "grad_norm": 0.5539014554549692, | |
| "learning_rate": 2.0322580645161292e-05, | |
| "loss": 0.6832, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2591093117408907, | |
| "grad_norm": 0.5732908474902535, | |
| "learning_rate": 2.0645161290322582e-05, | |
| "loss": 0.6735, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 0.5590313255867089, | |
| "learning_rate": 2.0967741935483873e-05, | |
| "loss": 0.6699, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26720647773279355, | |
| "grad_norm": 0.5699955045833435, | |
| "learning_rate": 2.1290322580645163e-05, | |
| "loss": 0.6835, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27125506072874495, | |
| "grad_norm": 0.5179363248613448, | |
| "learning_rate": 2.161290322580645e-05, | |
| "loss": 0.6832, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27530364372469635, | |
| "grad_norm": 0.8765277993361901, | |
| "learning_rate": 2.193548387096774e-05, | |
| "loss": 0.6589, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2793522267206478, | |
| "grad_norm": 1.17670768275659, | |
| "learning_rate": 2.225806451612903e-05, | |
| "loss": 0.6728, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2834008097165992, | |
| "grad_norm": 0.7823319161157729, | |
| "learning_rate": 2.2580645161290328e-05, | |
| "loss": 0.6751, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2874493927125506, | |
| "grad_norm": 0.772288869504426, | |
| "learning_rate": 2.2903225806451615e-05, | |
| "loss": 0.6666, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.291497975708502, | |
| "grad_norm": 0.6170525595194073, | |
| "learning_rate": 2.3225806451612906e-05, | |
| "loss": 0.6756, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.29554655870445345, | |
| "grad_norm": 0.9167628311073598, | |
| "learning_rate": 2.3548387096774196e-05, | |
| "loss": 0.6535, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29959514170040485, | |
| "grad_norm": 0.9893253909206395, | |
| "learning_rate": 2.3870967741935486e-05, | |
| "loss": 0.6555, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30364372469635625, | |
| "grad_norm": 0.9834499018994526, | |
| "learning_rate": 2.4193548387096777e-05, | |
| "loss": 0.6705, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.144734493054604, | |
| "learning_rate": 2.4516129032258067e-05, | |
| "loss": 0.6601, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3117408906882591, | |
| "grad_norm": 0.8749639307143532, | |
| "learning_rate": 2.4838709677419358e-05, | |
| "loss": 0.6567, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 1.0790363068518367, | |
| "learning_rate": 2.5161290322580648e-05, | |
| "loss": 0.6622, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31983805668016196, | |
| "grad_norm": 0.8068351467131251, | |
| "learning_rate": 2.548387096774194e-05, | |
| "loss": 0.6494, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32388663967611336, | |
| "grad_norm": 0.7358217857581465, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.6691, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32793522267206476, | |
| "grad_norm": 0.7910404484026751, | |
| "learning_rate": 2.6129032258064516e-05, | |
| "loss": 0.6581, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3319838056680162, | |
| "grad_norm": 1.0189281163152317, | |
| "learning_rate": 2.6451612903225806e-05, | |
| "loss": 0.6604, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3360323886639676, | |
| "grad_norm": 1.250707232094781, | |
| "learning_rate": 2.6774193548387097e-05, | |
| "loss": 0.6584, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.340080971659919, | |
| "grad_norm": 1.0221010559076413, | |
| "learning_rate": 2.7096774193548387e-05, | |
| "loss": 0.6654, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3441295546558704, | |
| "grad_norm": 0.9747487894817757, | |
| "learning_rate": 2.7419354838709678e-05, | |
| "loss": 0.668, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3481781376518219, | |
| "grad_norm": 1.1488440035276524, | |
| "learning_rate": 2.774193548387097e-05, | |
| "loss": 0.6547, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3522267206477733, | |
| "grad_norm": 1.0433437567794352, | |
| "learning_rate": 2.8064516129032262e-05, | |
| "loss": 0.6559, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3562753036437247, | |
| "grad_norm": 1.1271192882939096, | |
| "learning_rate": 2.8387096774193552e-05, | |
| "loss": 0.6437, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3603238866396761, | |
| "grad_norm": 0.8110248464355594, | |
| "learning_rate": 2.8709677419354843e-05, | |
| "loss": 0.6601, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3643724696356275, | |
| "grad_norm": 0.8448782614483569, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.646, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 1.0582363522720288, | |
| "learning_rate": 2.9354838709677423e-05, | |
| "loss": 0.6567, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3724696356275304, | |
| "grad_norm": 0.9252738028082241, | |
| "learning_rate": 2.9677419354838714e-05, | |
| "loss": 0.649, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3765182186234818, | |
| "grad_norm": 1.2026768414286737, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.6532, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3805668016194332, | |
| "grad_norm": 0.7804485442120414, | |
| "learning_rate": 3.032258064516129e-05, | |
| "loss": 0.6636, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 1.1269008353956613, | |
| "learning_rate": 3.0645161290322585e-05, | |
| "loss": 0.6483, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 1.2422834508677367, | |
| "learning_rate": 3.096774193548387e-05, | |
| "loss": 0.6557, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39271255060728744, | |
| "grad_norm": 1.080902738918357, | |
| "learning_rate": 3.1290322580645166e-05, | |
| "loss": 0.6565, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3967611336032389, | |
| "grad_norm": 1.207127966624223, | |
| "learning_rate": 3.161290322580645e-05, | |
| "loss": 0.645, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4008097165991903, | |
| "grad_norm": 0.8345242535508659, | |
| "learning_rate": 3.193548387096774e-05, | |
| "loss": 0.658, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4048582995951417, | |
| "grad_norm": 1.2834060470398054, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.6465, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4089068825910931, | |
| "grad_norm": 0.8462393488328828, | |
| "learning_rate": 3.258064516129033e-05, | |
| "loss": 0.6439, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.41295546558704455, | |
| "grad_norm": 1.0851316072291375, | |
| "learning_rate": 3.2903225806451614e-05, | |
| "loss": 0.6533, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.41700404858299595, | |
| "grad_norm": 0.8515368569240155, | |
| "learning_rate": 3.322580645161291e-05, | |
| "loss": 0.6663, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 1.120902785628779, | |
| "learning_rate": 3.3548387096774195e-05, | |
| "loss": 0.6484, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4251012145748988, | |
| "grad_norm": 0.8273884194044672, | |
| "learning_rate": 3.387096774193549e-05, | |
| "loss": 0.6501, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4291497975708502, | |
| "grad_norm": 1.043975383203534, | |
| "learning_rate": 3.4193548387096776e-05, | |
| "loss": 0.6559, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4331983805668016, | |
| "grad_norm": 0.9558606090576687, | |
| "learning_rate": 3.451612903225807e-05, | |
| "loss": 0.6622, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.43724696356275305, | |
| "grad_norm": 1.409602348684208, | |
| "learning_rate": 3.483870967741936e-05, | |
| "loss": 0.638, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.44129554655870445, | |
| "grad_norm": 0.9077425913929449, | |
| "learning_rate": 3.516129032258065e-05, | |
| "loss": 0.6403, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44534412955465585, | |
| "grad_norm": 1.4456907110675636, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.6386, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4493927125506073, | |
| "grad_norm": 0.8850507470193456, | |
| "learning_rate": 3.5806451612903225e-05, | |
| "loss": 0.6436, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4534412955465587, | |
| "grad_norm": 1.3598941384583307, | |
| "learning_rate": 3.612903225806452e-05, | |
| "loss": 0.6489, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4574898785425101, | |
| "grad_norm": 1.0554552128183237, | |
| "learning_rate": 3.6451612903225805e-05, | |
| "loss": 0.6429, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.2978921250974293, | |
| "learning_rate": 3.67741935483871e-05, | |
| "loss": 0.6544, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46558704453441296, | |
| "grad_norm": 1.1999953008092246, | |
| "learning_rate": 3.7096774193548386e-05, | |
| "loss": 0.6502, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.46963562753036436, | |
| "grad_norm": 0.9028813371129263, | |
| "learning_rate": 3.741935483870968e-05, | |
| "loss": 0.6449, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 1.498172577600328, | |
| "learning_rate": 3.7741935483870974e-05, | |
| "loss": 0.635, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4777327935222672, | |
| "grad_norm": 0.8628684902592916, | |
| "learning_rate": 3.806451612903226e-05, | |
| "loss": 0.651, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4817813765182186, | |
| "grad_norm": 1.325356878123798, | |
| "learning_rate": 3.8387096774193555e-05, | |
| "loss": 0.6509, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48582995951417, | |
| "grad_norm": 1.3024384322952347, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.6372, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4898785425101215, | |
| "grad_norm": 0.8931893048327549, | |
| "learning_rate": 3.9032258064516135e-05, | |
| "loss": 0.6295, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4939271255060729, | |
| "grad_norm": 1.6660214744528012, | |
| "learning_rate": 3.935483870967742e-05, | |
| "loss": 0.6418, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4979757085020243, | |
| "grad_norm": 0.7006044284126614, | |
| "learning_rate": 3.9677419354838716e-05, | |
| "loss": 0.6335, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5020242914979757, | |
| "grad_norm": 1.416424327753051, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6278, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5060728744939271, | |
| "grad_norm": 1.4968005996647389, | |
| "learning_rate": 3.9999920040266474e-05, | |
| "loss": 0.6416, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5101214574898786, | |
| "grad_norm": 0.7385106364982954, | |
| "learning_rate": 3.999968016170525e-05, | |
| "loss": 0.6437, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5141700404858299, | |
| "grad_norm": 1.2015286325568517, | |
| "learning_rate": 3.9999280366234394e-05, | |
| "loss": 0.6325, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5182186234817814, | |
| "grad_norm": 1.1101803815664573, | |
| "learning_rate": 3.999872065705065e-05, | |
| "loss": 0.634, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5222672064777328, | |
| "grad_norm": 1.2091813223734664, | |
| "learning_rate": 3.999800103862945e-05, | |
| "loss": 0.6409, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.1233944820169528, | |
| "learning_rate": 3.9997121516724834e-05, | |
| "loss": 0.6493, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5303643724696356, | |
| "grad_norm": 1.4084358933603047, | |
| "learning_rate": 3.999608209836944e-05, | |
| "loss": 0.637, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5344129554655871, | |
| "grad_norm": 0.7787358372642558, | |
| "learning_rate": 3.999488279187443e-05, | |
| "loss": 0.6429, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.2067694799153441, | |
| "learning_rate": 3.9993523606829425e-05, | |
| "loss": 0.6368, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5425101214574899, | |
| "grad_norm": 1.2569091416829608, | |
| "learning_rate": 3.9992004554102435e-05, | |
| "loss": 0.6279, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5465587044534413, | |
| "grad_norm": 0.7464647034016333, | |
| "learning_rate": 3.999032564583976e-05, | |
| "loss": 0.6624, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5506072874493927, | |
| "grad_norm": 0.82421085218986, | |
| "learning_rate": 3.998848689546591e-05, | |
| "loss": 0.6331, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5546558704453441, | |
| "grad_norm": 0.868166836133922, | |
| "learning_rate": 3.9986488317683496e-05, | |
| "loss": 0.6291, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5587044534412956, | |
| "grad_norm": 0.8446917005226834, | |
| "learning_rate": 3.9984329928473067e-05, | |
| "loss": 0.6325, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.562753036437247, | |
| "grad_norm": 1.2408895740679893, | |
| "learning_rate": 3.998201174509306e-05, | |
| "loss": 0.6189, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5668016194331984, | |
| "grad_norm": 0.7474843610290718, | |
| "learning_rate": 3.997953378607961e-05, | |
| "loss": 0.6293, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5708502024291497, | |
| "grad_norm": 1.1144423905084961, | |
| "learning_rate": 3.9976896071246403e-05, | |
| "loss": 0.6298, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5748987854251012, | |
| "grad_norm": 0.8987567089344991, | |
| "learning_rate": 3.997409862168454e-05, | |
| "loss": 0.6428, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 1.0169941491430243, | |
| "learning_rate": 3.997114145976237e-05, | |
| "loss": 0.6349, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.582995951417004, | |
| "grad_norm": 1.0467719923221595, | |
| "learning_rate": 3.9968024609125245e-05, | |
| "loss": 0.6398, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5870445344129555, | |
| "grad_norm": 0.7530470550280801, | |
| "learning_rate": 3.996474809469545e-05, | |
| "loss": 0.6447, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5910931174089069, | |
| "grad_norm": 1.0470077294769857, | |
| "learning_rate": 3.996131194267189e-05, | |
| "loss": 0.6179, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5951417004048583, | |
| "grad_norm": 0.9720160267515444, | |
| "learning_rate": 3.995771618052995e-05, | |
| "loss": 0.6244, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5991902834008097, | |
| "grad_norm": 1.0110779048189436, | |
| "learning_rate": 3.9953960837021256e-05, | |
| "loss": 0.6206, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6032388663967612, | |
| "grad_norm": 0.9585091363346829, | |
| "learning_rate": 3.995004594217343e-05, | |
| "loss": 0.6242, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6072874493927125, | |
| "grad_norm": 0.9481297272203146, | |
| "learning_rate": 3.994597152728987e-05, | |
| "loss": 0.6576, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.611336032388664, | |
| "grad_norm": 1.0088553011847603, | |
| "learning_rate": 3.994173762494947e-05, | |
| "loss": 0.6193, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.8574833850121605, | |
| "learning_rate": 3.9937344269006424e-05, | |
| "loss": 0.6425, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6194331983805668, | |
| "grad_norm": 1.2446194013165088, | |
| "learning_rate": 3.993279149458988e-05, | |
| "loss": 0.6188, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6234817813765182, | |
| "grad_norm": 0.6743588896540974, | |
| "learning_rate": 3.9928079338103695e-05, | |
| "loss": 0.6122, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6275303643724697, | |
| "grad_norm": 1.0341104798753165, | |
| "learning_rate": 3.992320783722616e-05, | |
| "loss": 0.6359, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.6577839962052963, | |
| "learning_rate": 3.991817703090965e-05, | |
| "loss": 0.6268, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6356275303643725, | |
| "grad_norm": 0.9837835263033738, | |
| "learning_rate": 3.991298695938038e-05, | |
| "loss": 0.6236, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6396761133603239, | |
| "grad_norm": 0.685178410575326, | |
| "learning_rate": 3.9907637664138e-05, | |
| "loss": 0.6308, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6437246963562753, | |
| "grad_norm": 0.7340026392656439, | |
| "learning_rate": 3.990212918795534e-05, | |
| "loss": 0.6285, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6477732793522267, | |
| "grad_norm": 0.7716577181238026, | |
| "learning_rate": 3.9896461574878036e-05, | |
| "loss": 0.6273, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6518218623481782, | |
| "grad_norm": 0.7276666862905775, | |
| "learning_rate": 3.989063487022416e-05, | |
| "loss": 0.6201, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6558704453441295, | |
| "grad_norm": 0.946540379526742, | |
| "learning_rate": 3.9884649120583903e-05, | |
| "loss": 0.6365, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.659919028340081, | |
| "grad_norm": 0.9280603658861549, | |
| "learning_rate": 3.987850437381915e-05, | |
| "loss": 0.6243, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6639676113360324, | |
| "grad_norm": 0.6894104855347177, | |
| "learning_rate": 3.987220067906313e-05, | |
| "loss": 0.6284, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6680161943319838, | |
| "grad_norm": 0.9533414407012497, | |
| "learning_rate": 3.986573808672002e-05, | |
| "loss": 0.6269, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6720647773279352, | |
| "grad_norm": 0.7544257846015593, | |
| "learning_rate": 3.9859116648464534e-05, | |
| "loss": 0.6107, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6761133603238867, | |
| "grad_norm": 1.1808258878813995, | |
| "learning_rate": 3.985233641724152e-05, | |
| "loss": 0.6222, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.680161943319838, | |
| "grad_norm": 0.7708280003559577, | |
| "learning_rate": 3.984539744726554e-05, | |
| "loss": 0.6179, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.7741473601330331, | |
| "learning_rate": 3.983829979402038e-05, | |
| "loss": 0.6213, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6882591093117408, | |
| "grad_norm": 0.6370403112590626, | |
| "learning_rate": 3.983104351425871e-05, | |
| "loss": 0.6086, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.636000083846134, | |
| "learning_rate": 3.982362866600154e-05, | |
| "loss": 0.5996, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6963562753036437, | |
| "grad_norm": 0.7310227901795401, | |
| "learning_rate": 3.9816055308537806e-05, | |
| "loss": 0.628, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7004048582995951, | |
| "grad_norm": 0.7044382914846494, | |
| "learning_rate": 3.980832350242388e-05, | |
| "loss": 0.6177, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7044534412955465, | |
| "grad_norm": 0.6778178868533852, | |
| "learning_rate": 3.980043330948306e-05, | |
| "loss": 0.6134, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708502024291498, | |
| "grad_norm": 0.5795979592297114, | |
| "learning_rate": 3.9792384792805124e-05, | |
| "loss": 0.62, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7125506072874493, | |
| "grad_norm": 0.5555545527092377, | |
| "learning_rate": 3.978417801674579e-05, | |
| "loss": 0.6178, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7165991902834008, | |
| "grad_norm": 0.6759407074307792, | |
| "learning_rate": 3.9775813046926244e-05, | |
| "loss": 0.6458, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7206477732793523, | |
| "grad_norm": 0.5813643047432081, | |
| "learning_rate": 3.976728995023254e-05, | |
| "loss": 0.631, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7246963562753036, | |
| "grad_norm": 0.7633671610926105, | |
| "learning_rate": 3.975860879481514e-05, | |
| "loss": 0.6243, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.728744939271255, | |
| "grad_norm": 0.7736048319921064, | |
| "learning_rate": 3.974976965008833e-05, | |
| "loss": 0.6228, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7327935222672065, | |
| "grad_norm": 0.6481800611628884, | |
| "learning_rate": 3.9740772586729674e-05, | |
| "loss": 0.6143, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.4547289801645608, | |
| "learning_rate": 3.973161767667946e-05, | |
| "loss": 0.6225, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7408906882591093, | |
| "grad_norm": 0.607964214506139, | |
| "learning_rate": 3.972230499314009e-05, | |
| "loss": 0.6081, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7449392712550608, | |
| "grad_norm": 0.49947613300257027, | |
| "learning_rate": 3.9712834610575545e-05, | |
| "loss": 0.6225, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7489878542510121, | |
| "grad_norm": 0.535937854108031, | |
| "learning_rate": 3.9703206604710746e-05, | |
| "loss": 0.6262, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7530364372469636, | |
| "grad_norm": 0.5196880567676438, | |
| "learning_rate": 3.969342105253098e-05, | |
| "loss": 0.6297, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.757085020242915, | |
| "grad_norm": 0.47905972531379215, | |
| "learning_rate": 3.9683478032281246e-05, | |
| "loss": 0.6361, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7611336032388664, | |
| "grad_norm": 0.4877598414862695, | |
| "learning_rate": 3.967337762346568e-05, | |
| "loss": 0.6122, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7651821862348178, | |
| "grad_norm": 0.5615696905260811, | |
| "learning_rate": 3.9663119906846885e-05, | |
| "loss": 0.6287, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.5385044796318006, | |
| "learning_rate": 3.965270496444528e-05, | |
| "loss": 0.6023, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7732793522267206, | |
| "grad_norm": 0.48334164450292466, | |
| "learning_rate": 3.9642132879538465e-05, | |
| "loss": 0.6147, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.4829449534676736, | |
| "learning_rate": 3.9631403736660566e-05, | |
| "loss": 0.6194, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7813765182186235, | |
| "grad_norm": 0.5804687241024078, | |
| "learning_rate": 3.962051762160151e-05, | |
| "loss": 0.6151, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7854251012145749, | |
| "grad_norm": 0.6215285796973908, | |
| "learning_rate": 3.960947462140638e-05, | |
| "loss": 0.6136, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.5446729615297259, | |
| "learning_rate": 3.959827482437472e-05, | |
| "loss": 0.6124, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7935222672064778, | |
| "grad_norm": 0.4685381769964206, | |
| "learning_rate": 3.958691832005981e-05, | |
| "loss": 0.6094, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7975708502024291, | |
| "grad_norm": 0.4687285452293428, | |
| "learning_rate": 3.9575405199267944e-05, | |
| "loss": 0.6263, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8016194331983806, | |
| "grad_norm": 0.5770739882048039, | |
| "learning_rate": 3.956373555405774e-05, | |
| "loss": 0.6107, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.805668016194332, | |
| "grad_norm": 0.6242387822147193, | |
| "learning_rate": 3.955190947773937e-05, | |
| "loss": 0.6091, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8097165991902834, | |
| "grad_norm": 0.6776755981256427, | |
| "learning_rate": 3.9539927064873824e-05, | |
| "loss": 0.6244, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8137651821862348, | |
| "grad_norm": 0.7509513557885094, | |
| "learning_rate": 3.952778841127214e-05, | |
| "loss": 0.6178, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8178137651821862, | |
| "grad_norm": 0.7677532325769164, | |
| "learning_rate": 3.95154936139947e-05, | |
| "loss": 0.6133, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8218623481781376, | |
| "grad_norm": 0.6881023145091432, | |
| "learning_rate": 3.950304277135034e-05, | |
| "loss": 0.6156, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8259109311740891, | |
| "grad_norm": 0.526640981632382, | |
| "learning_rate": 3.94904359828957e-05, | |
| "loss": 0.6086, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8299595141700404, | |
| "grad_norm": 0.5999188019422967, | |
| "learning_rate": 3.94776733494343e-05, | |
| "loss": 0.6313, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8340080971659919, | |
| "grad_norm": 0.6572149072647956, | |
| "learning_rate": 3.9464754973015826e-05, | |
| "loss": 0.6004, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8380566801619433, | |
| "grad_norm": 0.668330381035694, | |
| "learning_rate": 3.9451680956935275e-05, | |
| "loss": 0.627, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.6055193171436035, | |
| "learning_rate": 3.943845140573213e-05, | |
| "loss": 0.6178, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.4578668887366359, | |
| "learning_rate": 3.942506642518952e-05, | |
| "loss": 0.617, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8502024291497976, | |
| "grad_norm": 0.5131830783920847, | |
| "learning_rate": 3.9411526122333396e-05, | |
| "loss": 0.6119, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.854251012145749, | |
| "grad_norm": 0.6839867498939525, | |
| "learning_rate": 3.939783060543167e-05, | |
| "loss": 0.6043, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8582995951417004, | |
| "grad_norm": 0.6996978827777254, | |
| "learning_rate": 3.938397998399333e-05, | |
| "loss": 0.6086, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8623481781376519, | |
| "grad_norm": 0.5230350853870152, | |
| "learning_rate": 3.936997436876756e-05, | |
| "loss": 0.6103, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8663967611336032, | |
| "grad_norm": 0.5603870779390425, | |
| "learning_rate": 3.935581387174289e-05, | |
| "loss": 0.6107, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8704453441295547, | |
| "grad_norm": 0.4621144512503493, | |
| "learning_rate": 3.9341498606146295e-05, | |
| "loss": 0.5941, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8744939271255061, | |
| "grad_norm": 0.4514272131595286, | |
| "learning_rate": 3.932702868644224e-05, | |
| "loss": 0.6033, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8785425101214575, | |
| "grad_norm": 0.6515572427986207, | |
| "learning_rate": 3.931240422833181e-05, | |
| "loss": 0.6217, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8825910931174089, | |
| "grad_norm": 0.5763340177659431, | |
| "learning_rate": 3.9297625348751806e-05, | |
| "loss": 0.6151, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8866396761133604, | |
| "grad_norm": 0.5400088061169059, | |
| "learning_rate": 3.9282692165873744e-05, | |
| "loss": 0.6042, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8906882591093117, | |
| "grad_norm": 0.46195275151611204, | |
| "learning_rate": 3.9267604799102945e-05, | |
| "loss": 0.6162, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.5302110370121319, | |
| "learning_rate": 3.9252363369077615e-05, | |
| "loss": 0.6135, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8987854251012146, | |
| "grad_norm": 0.6090590872783986, | |
| "learning_rate": 3.9236967997667796e-05, | |
| "loss": 0.6087, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.902834008097166, | |
| "grad_norm": 0.771174381970499, | |
| "learning_rate": 3.9221418807974496e-05, | |
| "loss": 0.6199, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9068825910931174, | |
| "grad_norm": 0.782126552446252, | |
| "learning_rate": 3.92057159243286e-05, | |
| "loss": 0.605, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9109311740890689, | |
| "grad_norm": 0.5833257265858036, | |
| "learning_rate": 3.9189859472289956e-05, | |
| "loss": 0.5919, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9149797570850202, | |
| "grad_norm": 0.43138133801564266, | |
| "learning_rate": 3.9173849578646324e-05, | |
| "loss": 0.6189, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9190283400809717, | |
| "grad_norm": 0.48608134582956036, | |
| "learning_rate": 3.9157686371412395e-05, | |
| "loss": 0.6093, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.7605909730717876, | |
| "learning_rate": 3.914136997982874e-05, | |
| "loss": 0.6096, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9271255060728745, | |
| "grad_norm": 0.5593764732750317, | |
| "learning_rate": 3.912490053436079e-05, | |
| "loss": 0.6174, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9311740890688259, | |
| "grad_norm": 0.46609834419770724, | |
| "learning_rate": 3.9108278166697805e-05, | |
| "loss": 0.6095, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9352226720647774, | |
| "grad_norm": 0.5218099369303045, | |
| "learning_rate": 3.909150300975178e-05, | |
| "loss": 0.5957, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9392712550607287, | |
| "grad_norm": 0.6406655070077478, | |
| "learning_rate": 3.907457519765642e-05, | |
| "loss": 0.6066, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9433198380566802, | |
| "grad_norm": 0.6401348012429077, | |
| "learning_rate": 3.905749486576607e-05, | |
| "loss": 0.6092, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.4178925181919035, | |
| "learning_rate": 3.90402621506546e-05, | |
| "loss": 0.6097, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.951417004048583, | |
| "grad_norm": 0.4729374858780956, | |
| "learning_rate": 3.902287719011435e-05, | |
| "loss": 0.6045, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9554655870445344, | |
| "grad_norm": 0.6329417226251981, | |
| "learning_rate": 3.900534012315499e-05, | |
| "loss": 0.6114, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9595141700404858, | |
| "grad_norm": 0.6907538623600455, | |
| "learning_rate": 3.8987651090002454e-05, | |
| "loss": 0.6142, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9635627530364372, | |
| "grad_norm": 0.6630663405888653, | |
| "learning_rate": 3.896981023209777e-05, | |
| "loss": 0.5965, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9676113360323887, | |
| "grad_norm": 0.5878430233810128, | |
| "learning_rate": 3.8951817692095966e-05, | |
| "loss": 0.6105, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.97165991902834, | |
| "grad_norm": 0.61734889113668, | |
| "learning_rate": 3.893367361386491e-05, | |
| "loss": 0.607, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9757085020242915, | |
| "grad_norm": 0.5671079822680488, | |
| "learning_rate": 3.891537814248417e-05, | |
| "loss": 0.5973, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.979757085020243, | |
| "grad_norm": 0.6306752626233983, | |
| "learning_rate": 3.889693142424384e-05, | |
| "loss": 0.6028, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9838056680161943, | |
| "grad_norm": 0.6289871748136849, | |
| "learning_rate": 3.8878333606643405e-05, | |
| "loss": 0.612, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9878542510121457, | |
| "grad_norm": 0.4156639973200423, | |
| "learning_rate": 3.885958483839049e-05, | |
| "loss": 0.6213, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9919028340080972, | |
| "grad_norm": 0.4511848085290255, | |
| "learning_rate": 3.8840685269399786e-05, | |
| "loss": 0.6003, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9959514170040485, | |
| "grad_norm": 0.45837833678431184, | |
| "learning_rate": 3.882163505079171e-05, | |
| "loss": 0.6132, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5333597277197615, | |
| "learning_rate": 3.880243433489132e-05, | |
| "loss": 0.6031, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0040485829959513, | |
| "grad_norm": 0.5781739956075361, | |
| "learning_rate": 3.878308327522702e-05, | |
| "loss": 0.5704, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.008097165991903, | |
| "grad_norm": 0.3930256481349083, | |
| "learning_rate": 3.876358202652936e-05, | |
| "loss": 0.5502, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0121457489878543, | |
| "grad_norm": 0.6708988376376661, | |
| "learning_rate": 3.8743930744729835e-05, | |
| "loss": 0.5551, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0161943319838056, | |
| "grad_norm": 0.690325824408422, | |
| "learning_rate": 3.8724129586959535e-05, | |
| "loss": 0.5607, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0202429149797572, | |
| "grad_norm": 0.7451322644645656, | |
| "learning_rate": 3.870417871154801e-05, | |
| "loss": 0.5475, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0242914979757085, | |
| "grad_norm": 0.5663372466758823, | |
| "learning_rate": 3.868407827802192e-05, | |
| "loss": 0.5512, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0283400809716599, | |
| "grad_norm": 0.6168057716947042, | |
| "learning_rate": 3.8663828447103804e-05, | |
| "loss": 0.5594, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0323886639676114, | |
| "grad_norm": 0.7216016763043562, | |
| "learning_rate": 3.8643429380710765e-05, | |
| "loss": 0.5584, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0364372469635628, | |
| "grad_norm": 0.6384038249513801, | |
| "learning_rate": 3.862288124195319e-05, | |
| "loss": 0.5607, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.040485829959514, | |
| "grad_norm": 0.5592596100400233, | |
| "learning_rate": 3.8602184195133464e-05, | |
| "loss": 0.5506, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0445344129554657, | |
| "grad_norm": 0.5082145284346814, | |
| "learning_rate": 3.8581338405744604e-05, | |
| "loss": 0.5485, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.048582995951417, | |
| "grad_norm": 0.7882542732966339, | |
| "learning_rate": 3.8560344040468996e-05, | |
| "loss": 0.575, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.5933373627628633, | |
| "learning_rate": 3.853920126717703e-05, | |
| "loss": 0.556, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.05668016194332, | |
| "grad_norm": 0.661917357205393, | |
| "learning_rate": 3.851791025492574e-05, | |
| "loss": 0.5592, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0607287449392713, | |
| "grad_norm": 0.6324908030840547, | |
| "learning_rate": 3.8496471173957515e-05, | |
| "loss": 0.5454, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0647773279352226, | |
| "grad_norm": 0.5058717063271357, | |
| "learning_rate": 3.847488419569866e-05, | |
| "loss": 0.5499, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0688259109311742, | |
| "grad_norm": 0.5860370885481088, | |
| "learning_rate": 3.8453149492758077e-05, | |
| "loss": 0.544, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0728744939271255, | |
| "grad_norm": 0.4991682841491159, | |
| "learning_rate": 3.8431267238925886e-05, | |
| "loss": 0.5597, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.496038439641373, | |
| "learning_rate": 3.840923760917199e-05, | |
| "loss": 0.5682, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0809716599190284, | |
| "grad_norm": 0.47911779307841285, | |
| "learning_rate": 3.838706077964473e-05, | |
| "loss": 0.5565, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0850202429149798, | |
| "grad_norm": 0.5463875983297761, | |
| "learning_rate": 3.836473692766944e-05, | |
| "loss": 0.5544, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0890688259109311, | |
| "grad_norm": 0.5286162744220438, | |
| "learning_rate": 3.834226623174705e-05, | |
| "loss": 0.5541, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0931174089068827, | |
| "grad_norm": 0.46621542988101566, | |
| "learning_rate": 3.831964887155264e-05, | |
| "loss": 0.5683, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.097165991902834, | |
| "grad_norm": 0.4905331814846084, | |
| "learning_rate": 3.829688502793403e-05, | |
| "loss": 0.5628, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.1012145748987854, | |
| "grad_norm": 0.5977422905308619, | |
| "learning_rate": 3.827397488291029e-05, | |
| "loss": 0.5538, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.5439026686814479, | |
| "learning_rate": 3.8250918619670343e-05, | |
| "loss": 0.5592, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.1093117408906883, | |
| "grad_norm": 0.4716887012332386, | |
| "learning_rate": 3.822771642257145e-05, | |
| "loss": 0.5626, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1133603238866396, | |
| "grad_norm": 0.67115965411025, | |
| "learning_rate": 3.820436847713776e-05, | |
| "loss": 0.5412, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.117408906882591, | |
| "grad_norm": 0.49966319549920724, | |
| "learning_rate": 3.8180874970058826e-05, | |
| "loss": 0.5523, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.1214574898785425, | |
| "grad_norm": 0.4800759748504716, | |
| "learning_rate": 3.8157236089188103e-05, | |
| "loss": 0.5724, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.125506072874494, | |
| "grad_norm": 0.6188269299338136, | |
| "learning_rate": 3.813345202354145e-05, | |
| "loss": 0.5565, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1295546558704452, | |
| "grad_norm": 0.5860103418720398, | |
| "learning_rate": 3.810952296329563e-05, | |
| "loss": 0.5535, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1336032388663968, | |
| "grad_norm": 0.41226336980137956, | |
| "learning_rate": 3.808544909978676e-05, | |
| "loss": 0.5441, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1376518218623481, | |
| "grad_norm": 0.586057945792376, | |
| "learning_rate": 3.806123062550882e-05, | |
| "loss": 0.5793, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1417004048582995, | |
| "grad_norm": 0.6103720123359673, | |
| "learning_rate": 3.803686773411208e-05, | |
| "loss": 0.5491, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.145748987854251, | |
| "grad_norm": 0.6033213462051763, | |
| "learning_rate": 3.801236062040158e-05, | |
| "loss": 0.547, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1497975708502024, | |
| "grad_norm": 0.6609040859045737, | |
| "learning_rate": 3.798770948033553e-05, | |
| "loss": 0.5635, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.5898313441837735, | |
| "learning_rate": 3.796291451102381e-05, | |
| "loss": 0.5511, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.43968814443446946, | |
| "learning_rate": 3.793797591072631e-05, | |
| "loss": 0.5572, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1619433198380567, | |
| "grad_norm": 0.651515945747009, | |
| "learning_rate": 3.791289387885144e-05, | |
| "loss": 0.5519, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.165991902834008, | |
| "grad_norm": 0.759509386813495, | |
| "learning_rate": 3.788766861595443e-05, | |
| "loss": 0.5694, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1700404858299596, | |
| "grad_norm": 0.47437255850859544, | |
| "learning_rate": 3.7862300323735835e-05, | |
| "loss": 0.5437, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.174089068825911, | |
| "grad_norm": 0.5565111883015389, | |
| "learning_rate": 3.783678920503983e-05, | |
| "loss": 0.5412, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1781376518218623, | |
| "grad_norm": 0.7096470131746911, | |
| "learning_rate": 3.781113546385265e-05, | |
| "loss": 0.5417, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1821862348178138, | |
| "grad_norm": 0.6060273193677553, | |
| "learning_rate": 3.7785339305300916e-05, | |
| "loss": 0.5578, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1862348178137652, | |
| "grad_norm": 0.40546348381442066, | |
| "learning_rate": 3.7759400935650025e-05, | |
| "loss": 0.5512, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1902834008097165, | |
| "grad_norm": 0.42935173306430263, | |
| "learning_rate": 3.77333205623025e-05, | |
| "loss": 0.5579, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.194331983805668, | |
| "grad_norm": 0.4359593397136448, | |
| "learning_rate": 3.7707098393796295e-05, | |
| "loss": 0.5471, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1983805668016194, | |
| "grad_norm": 0.33165910639749374, | |
| "learning_rate": 3.7680734639803185e-05, | |
| "loss": 0.5428, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2024291497975708, | |
| "grad_norm": 0.368858039581001, | |
| "learning_rate": 3.765422951112704e-05, | |
| "loss": 0.5591, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2064777327935223, | |
| "grad_norm": 0.4124294413921431, | |
| "learning_rate": 3.762758321970216e-05, | |
| "loss": 0.5563, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.3851222178039451, | |
| "learning_rate": 3.7600795978591584e-05, | |
| "loss": 0.5621, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.214574898785425, | |
| "grad_norm": 0.4339825800141171, | |
| "learning_rate": 3.757386800198538e-05, | |
| "loss": 0.5594, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2186234817813766, | |
| "grad_norm": 0.4653828659957958, | |
| "learning_rate": 3.7546799505198925e-05, | |
| "loss": 0.5483, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.222672064777328, | |
| "grad_norm": 0.4383154979819392, | |
| "learning_rate": 3.7519590704671197e-05, | |
| "loss": 0.5612, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2267206477732793, | |
| "grad_norm": 0.36437347775616985, | |
| "learning_rate": 3.749224181796305e-05, | |
| "loss": 0.5575, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.39449444129314043, | |
| "learning_rate": 3.746475306375545e-05, | |
| "loss": 0.5691, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2348178137651822, | |
| "grad_norm": 0.45923071932906545, | |
| "learning_rate": 3.743712466184774e-05, | |
| "loss": 0.561, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2388663967611335, | |
| "grad_norm": 0.37719653270581155, | |
| "learning_rate": 3.7409356833155885e-05, | |
| "loss": 0.5502, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.242914979757085, | |
| "grad_norm": 0.4271173375554501, | |
| "learning_rate": 3.73814497997107e-05, | |
| "loss": 0.5724, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2469635627530364, | |
| "grad_norm": 0.469321378671175, | |
| "learning_rate": 3.7353403784656095e-05, | |
| "loss": 0.5579, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2510121457489878, | |
| "grad_norm": 0.4168055648991831, | |
| "learning_rate": 3.732521901224724e-05, | |
| "loss": 0.5449, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2550607287449393, | |
| "grad_norm": 0.4145296077450622, | |
| "learning_rate": 3.7296895707848844e-05, | |
| "loss": 0.5554, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2591093117408907, | |
| "grad_norm": 0.47819112658983753, | |
| "learning_rate": 3.7268434097933275e-05, | |
| "loss": 0.5546, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.4895915335424623, | |
| "learning_rate": 3.723983441007882e-05, | |
| "loss": 0.5511, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2672064777327936, | |
| "grad_norm": 0.41171228421641026, | |
| "learning_rate": 3.721109687296781e-05, | |
| "loss": 0.5489, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.271255060728745, | |
| "grad_norm": 0.5428891684752148, | |
| "learning_rate": 3.718222171638484e-05, | |
| "loss": 0.5602, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2753036437246963, | |
| "grad_norm": 0.4539521356485337, | |
| "learning_rate": 3.715320917121488e-05, | |
| "loss": 0.5564, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2793522267206479, | |
| "grad_norm": 0.4568470500646606, | |
| "learning_rate": 3.712405946944149e-05, | |
| "loss": 0.5387, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2834008097165992, | |
| "grad_norm": 0.47325123007566056, | |
| "learning_rate": 3.709477284414488e-05, | |
| "loss": 0.5434, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2874493927125505, | |
| "grad_norm": 0.45124561015248765, | |
| "learning_rate": 3.7065349529500136e-05, | |
| "loss": 0.5289, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.291497975708502, | |
| "grad_norm": 0.4472716197778247, | |
| "learning_rate": 3.7035789760775305e-05, | |
| "loss": 0.5416, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2955465587044535, | |
| "grad_norm": 0.5278691411537134, | |
| "learning_rate": 3.70060937743295e-05, | |
| "loss": 0.5379, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2995951417004048, | |
| "grad_norm": 0.45347255351526744, | |
| "learning_rate": 3.6976261807611045e-05, | |
| "loss": 0.5429, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.3036437246963564, | |
| "grad_norm": 0.5661659462928471, | |
| "learning_rate": 3.694629409915555e-05, | |
| "loss": 0.5635, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.5539938041706618, | |
| "learning_rate": 3.691619088858401e-05, | |
| "loss": 0.5412, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.311740890688259, | |
| "grad_norm": 0.5868350529060926, | |
| "learning_rate": 3.68859524166009e-05, | |
| "loss": 0.5583, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.7213140061003013, | |
| "learning_rate": 3.6855578924992226e-05, | |
| "loss": 0.5531, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.319838056680162, | |
| "grad_norm": 0.6087057700769508, | |
| "learning_rate": 3.6825070656623626e-05, | |
| "loss": 0.5528, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.3238866396761133, | |
| "grad_norm": 0.4024216055174575, | |
| "learning_rate": 3.67944278554384e-05, | |
| "loss": 0.5443, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3279352226720649, | |
| "grad_norm": 0.7127908889948412, | |
| "learning_rate": 3.676365076645557e-05, | |
| "loss": 0.5455, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.538914678649597, | |
| "learning_rate": 3.673273963576791e-05, | |
| "loss": 0.5406, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3360323886639676, | |
| "grad_norm": 0.45525432998758136, | |
| "learning_rate": 3.670169471054001e-05, | |
| "loss": 0.558, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3400809716599191, | |
| "grad_norm": 0.5481792886558613, | |
| "learning_rate": 3.6670516239006254e-05, | |
| "loss": 0.5877, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3441295546558705, | |
| "grad_norm": 0.519507158458117, | |
| "learning_rate": 3.663920447046888e-05, | |
| "loss": 0.5456, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3481781376518218, | |
| "grad_norm": 0.4450407609137937, | |
| "learning_rate": 3.6607759655295954e-05, | |
| "loss": 0.5517, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3522267206477734, | |
| "grad_norm": 0.4223367870804306, | |
| "learning_rate": 3.657618204491937e-05, | |
| "loss": 0.5514, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3562753036437247, | |
| "grad_norm": 0.3707518423504386, | |
| "learning_rate": 3.6544471891832875e-05, | |
| "loss": 0.5579, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.360323886639676, | |
| "grad_norm": 0.38767203809826734, | |
| "learning_rate": 3.6512629449589983e-05, | |
| "loss": 0.536, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3643724696356276, | |
| "grad_norm": 0.42742065515290606, | |
| "learning_rate": 3.6480654972802045e-05, | |
| "loss": 0.5481, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.3759391270184466, | |
| "learning_rate": 3.644854871713611e-05, | |
| "loss": 0.5469, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3724696356275303, | |
| "grad_norm": 0.3806074072986547, | |
| "learning_rate": 3.6416310939312924e-05, | |
| "loss": 0.5452, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.376518218623482, | |
| "grad_norm": 0.4357322502971459, | |
| "learning_rate": 3.638394189710493e-05, | |
| "loss": 0.5537, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3805668016194332, | |
| "grad_norm": 0.38797949865374004, | |
| "learning_rate": 3.635144184933412e-05, | |
| "loss": 0.5522, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.389758201820098, | |
| "learning_rate": 3.6318811055869994e-05, | |
| "loss": 0.5443, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3886639676113361, | |
| "grad_norm": 0.372912360074584, | |
| "learning_rate": 3.628604977762752e-05, | |
| "loss": 0.5359, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3927125506072875, | |
| "grad_norm": 0.48150453156717343, | |
| "learning_rate": 3.6253158276565004e-05, | |
| "loss": 0.5501, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3967611336032388, | |
| "grad_norm": 0.504648727353078, | |
| "learning_rate": 3.6220136815682016e-05, | |
| "loss": 0.569, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4008097165991904, | |
| "grad_norm": 0.46747441483691565, | |
| "learning_rate": 3.618698565901727e-05, | |
| "loss": 0.5562, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.4048582995951417, | |
| "grad_norm": 0.6038749898143284, | |
| "learning_rate": 3.615370507164654e-05, | |
| "loss": 0.5653, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.408906882591093, | |
| "grad_norm": 0.32142701068180357, | |
| "learning_rate": 3.61202953196805e-05, | |
| "loss": 0.5448, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.4129554655870447, | |
| "grad_norm": 0.41574032351461837, | |
| "learning_rate": 3.6086756670262654e-05, | |
| "loss": 0.5554, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.417004048582996, | |
| "grad_norm": 0.41473452160474056, | |
| "learning_rate": 3.6053089391567135e-05, | |
| "loss": 0.5448, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.4259468180063578, | |
| "learning_rate": 3.601929375279662e-05, | |
| "loss": 0.5462, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.425101214574899, | |
| "grad_norm": 0.3598093593431287, | |
| "learning_rate": 3.598537002418012e-05, | |
| "loss": 0.5651, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4291497975708503, | |
| "grad_norm": 0.42051315384452137, | |
| "learning_rate": 3.5951318476970885e-05, | |
| "loss": 0.5569, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4331983805668016, | |
| "grad_norm": 0.316290096195694, | |
| "learning_rate": 3.591713938344416e-05, | |
| "loss": 0.5574, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4372469635627532, | |
| "grad_norm": 0.35104402090698394, | |
| "learning_rate": 3.588283301689507e-05, | |
| "loss": 0.5582, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4412955465587045, | |
| "grad_norm": 0.3877013898187867, | |
| "learning_rate": 3.5848399651636424e-05, | |
| "loss": 0.5449, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4453441295546559, | |
| "grad_norm": 0.396355479366867, | |
| "learning_rate": 3.5813839562996473e-05, | |
| "loss": 0.5445, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4493927125506074, | |
| "grad_norm": 0.4192323554164813, | |
| "learning_rate": 3.5779153027316775e-05, | |
| "loss": 0.554, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4534412955465588, | |
| "grad_norm": 0.38288191966858337, | |
| "learning_rate": 3.574434032194994e-05, | |
| "loss": 0.5591, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.45748987854251, | |
| "grad_norm": 0.3889896332826583, | |
| "learning_rate": 3.5709401725257436e-05, | |
| "loss": 0.5594, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.5246359388281405, | |
| "learning_rate": 3.5674337516607346e-05, | |
| "loss": 0.5483, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.465587044534413, | |
| "grad_norm": 0.39840905177966307, | |
| "learning_rate": 3.563914797637216e-05, | |
| "loss": 0.5659, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4696356275303644, | |
| "grad_norm": 0.5059796021499885, | |
| "learning_rate": 3.560383338592649e-05, | |
| "loss": 0.5515, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.40517808322365867, | |
| "learning_rate": 3.556839402764487e-05, | |
| "loss": 0.5531, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4777327935222673, | |
| "grad_norm": 0.498686172009117, | |
| "learning_rate": 3.553283018489946e-05, | |
| "loss": 0.5401, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4817813765182186, | |
| "grad_norm": 0.4840030703668601, | |
| "learning_rate": 3.54971421420578e-05, | |
| "loss": 0.5343, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.48582995951417, | |
| "grad_norm": 0.39340792089173077, | |
| "learning_rate": 3.5461330184480524e-05, | |
| "loss": 0.5533, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4898785425101215, | |
| "grad_norm": 0.4473409455453468, | |
| "learning_rate": 3.54253945985191e-05, | |
| "loss": 0.5443, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4939271255060729, | |
| "grad_norm": 0.4413666662827651, | |
| "learning_rate": 3.5389335671513524e-05, | |
| "loss": 0.5594, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4979757085020242, | |
| "grad_norm": 0.4294652059760945, | |
| "learning_rate": 3.5353153691789986e-05, | |
| "loss": 0.5503, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5020242914979756, | |
| "grad_norm": 0.443302614306386, | |
| "learning_rate": 3.531684894865866e-05, | |
| "loss": 0.5583, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.5060728744939271, | |
| "grad_norm": 0.46835383587532076, | |
| "learning_rate": 3.528042173241128e-05, | |
| "loss": 0.5304, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5101214574898787, | |
| "grad_norm": 0.38851656884225866, | |
| "learning_rate": 3.524387233431892e-05, | |
| "loss": 0.5398, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.5141700404858298, | |
| "grad_norm": 0.519253433695992, | |
| "learning_rate": 3.520720104662958e-05, | |
| "loss": 0.5497, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5182186234817814, | |
| "grad_norm": 0.3865535893621698, | |
| "learning_rate": 3.5170408162565904e-05, | |
| "loss": 0.5435, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.522267206477733, | |
| "grad_norm": 0.47252875158237856, | |
| "learning_rate": 3.513349397632281e-05, | |
| "loss": 0.552, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.3669654375443796, | |
| "learning_rate": 3.5096458783065145e-05, | |
| "loss": 0.5432, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5303643724696356, | |
| "grad_norm": 0.42079906280293133, | |
| "learning_rate": 3.505930287892533e-05, | |
| "loss": 0.5651, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.5344129554655872, | |
| "grad_norm": 0.4857096795121452, | |
| "learning_rate": 3.502202656100099e-05, | |
| "loss": 0.5508, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.46641685295608515, | |
| "learning_rate": 3.498463012735256e-05, | |
| "loss": 0.5604, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.54251012145749, | |
| "grad_norm": 0.45020359923461156, | |
| "learning_rate": 3.494711387700094e-05, | |
| "loss": 0.5475, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5465587044534415, | |
| "grad_norm": 0.3259877440866667, | |
| "learning_rate": 3.490947810992505e-05, | |
| "loss": 0.5466, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5506072874493926, | |
| "grad_norm": 0.4318006800292944, | |
| "learning_rate": 3.48717231270595e-05, | |
| "loss": 0.5368, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5546558704453441, | |
| "grad_norm": 0.4287572417128566, | |
| "learning_rate": 3.483384923029211e-05, | |
| "loss": 0.5561, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5587044534412957, | |
| "grad_norm": 0.3489164466847689, | |
| "learning_rate": 3.479585672246156e-05, | |
| "loss": 0.556, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5627530364372468, | |
| "grad_norm": 0.5157097587485331, | |
| "learning_rate": 3.4757745907354924e-05, | |
| "loss": 0.5403, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5668016194331984, | |
| "grad_norm": 0.5798994139136568, | |
| "learning_rate": 3.471951708970527e-05, | |
| "loss": 0.5472, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5708502024291497, | |
| "grad_norm": 0.40434565792777016, | |
| "learning_rate": 3.468117057518921e-05, | |
| "loss": 0.5507, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.574898785425101, | |
| "grad_norm": 0.42134805725274344, | |
| "learning_rate": 3.4642706670424426e-05, | |
| "loss": 0.5562, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.48219185881048915, | |
| "learning_rate": 3.460412568296731e-05, | |
| "loss": 0.5693, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.582995951417004, | |
| "grad_norm": 0.46268729279786036, | |
| "learning_rate": 3.456542792131039e-05, | |
| "loss": 0.5604, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5870445344129553, | |
| "grad_norm": 0.41573048783494604, | |
| "learning_rate": 3.4526613694879936e-05, | |
| "loss": 0.56, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.591093117408907, | |
| "grad_norm": 0.37563110401112215, | |
| "learning_rate": 3.448768331403348e-05, | |
| "loss": 0.5428, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5951417004048583, | |
| "grad_norm": 0.43004916237792873, | |
| "learning_rate": 3.44486370900573e-05, | |
| "loss": 0.529, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5991902834008096, | |
| "grad_norm": 0.37174694293642635, | |
| "learning_rate": 3.440947533516396e-05, | |
| "loss": 0.5463, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.6032388663967612, | |
| "grad_norm": 0.4794084735104106, | |
| "learning_rate": 3.4370198362489816e-05, | |
| "loss": 0.5574, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.6072874493927125, | |
| "grad_norm": 0.5077632607380789, | |
| "learning_rate": 3.43308064860925e-05, | |
| "loss": 0.5539, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.6113360323886639, | |
| "grad_norm": 0.40765846657546956, | |
| "learning_rate": 3.4291300020948393e-05, | |
| "loss": 0.5549, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.4760193774430813, | |
| "learning_rate": 3.425167928295015e-05, | |
| "loss": 0.5451, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6194331983805668, | |
| "grad_norm": 0.35859129973506165, | |
| "learning_rate": 3.421194458890411e-05, | |
| "loss": 0.5521, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.623481781376518, | |
| "grad_norm": 0.4403985710056867, | |
| "learning_rate": 3.4172096256527865e-05, | |
| "loss": 0.5478, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.6275303643724697, | |
| "grad_norm": 0.31618364175644664, | |
| "learning_rate": 3.4132134604447595e-05, | |
| "loss": 0.5592, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.3936012438501045, | |
| "learning_rate": 3.409205995219561e-05, | |
| "loss": 0.5499, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6356275303643724, | |
| "grad_norm": 0.3916542568652986, | |
| "learning_rate": 3.4051872620207765e-05, | |
| "loss": 0.5424, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.639676113360324, | |
| "grad_norm": 0.36595274519358856, | |
| "learning_rate": 3.40115729298209e-05, | |
| "loss": 0.5488, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6437246963562753, | |
| "grad_norm": 0.36114421653096784, | |
| "learning_rate": 3.397116120327025e-05, | |
| "loss": 0.5614, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6477732793522266, | |
| "grad_norm": 0.4318367505355158, | |
| "learning_rate": 3.3930637763686916e-05, | |
| "loss": 0.5473, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6518218623481782, | |
| "grad_norm": 0.4001610197979546, | |
| "learning_rate": 3.389000293509524e-05, | |
| "loss": 0.5504, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6558704453441295, | |
| "grad_norm": 0.43104676335651726, | |
| "learning_rate": 3.3849257042410225e-05, | |
| "loss": 0.5414, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6599190283400809, | |
| "grad_norm": 0.46380196988403577, | |
| "learning_rate": 3.380840041143494e-05, | |
| "loss": 0.5511, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6639676113360324, | |
| "grad_norm": 0.3754909610408835, | |
| "learning_rate": 3.376743336885792e-05, | |
| "loss": 0.5677, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6680161943319838, | |
| "grad_norm": 0.41339815710776184, | |
| "learning_rate": 3.372635624225055e-05, | |
| "loss": 0.5377, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6720647773279351, | |
| "grad_norm": 0.44408298963445464, | |
| "learning_rate": 3.3685169360064445e-05, | |
| "loss": 0.5542, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6761133603238867, | |
| "grad_norm": 0.43518516788535155, | |
| "learning_rate": 3.3643873051628804e-05, | |
| "loss": 0.554, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.680161943319838, | |
| "grad_norm": 0.37274910675920114, | |
| "learning_rate": 3.3602467647147814e-05, | |
| "loss": 0.545, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.45952230742755107, | |
| "learning_rate": 3.356095347769799e-05, | |
| "loss": 0.5501, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.688259109311741, | |
| "grad_norm": 0.41741042647914256, | |
| "learning_rate": 3.351933087522552e-05, | |
| "loss": 0.5561, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.4142258861401265, | |
| "learning_rate": 3.3477600172543626e-05, | |
| "loss": 0.5316, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6963562753036436, | |
| "grad_norm": 0.4600162283877399, | |
| "learning_rate": 3.3435761703329894e-05, | |
| "loss": 0.5589, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.7004048582995952, | |
| "grad_norm": 0.42058816972802, | |
| "learning_rate": 3.3393815802123606e-05, | |
| "loss": 0.5426, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7044534412955465, | |
| "grad_norm": 0.43688096437818236, | |
| "learning_rate": 3.335176280432308e-05, | |
| "loss": 0.5505, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.708502024291498, | |
| "grad_norm": 0.4642035053734633, | |
| "learning_rate": 3.330960304618295e-05, | |
| "loss": 0.5422, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.7125506072874495, | |
| "grad_norm": 0.4008341040883464, | |
| "learning_rate": 3.326733686481153e-05, | |
| "loss": 0.5573, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.7165991902834008, | |
| "grad_norm": 0.36683119619811955, | |
| "learning_rate": 3.3224964598168084e-05, | |
| "loss": 0.5495, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.41683602212986914, | |
| "learning_rate": 3.318248658506012e-05, | |
| "loss": 0.5488, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7246963562753037, | |
| "grad_norm": 0.3555599947072005, | |
| "learning_rate": 3.313990316514069e-05, | |
| "loss": 0.56, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.728744939271255, | |
| "grad_norm": 0.34140235029736526, | |
| "learning_rate": 3.309721467890571e-05, | |
| "loss": 0.5485, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.7327935222672064, | |
| "grad_norm": 0.431585362244648, | |
| "learning_rate": 3.305442146769115e-05, | |
| "loss": 0.5551, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.31814977963608365, | |
| "learning_rate": 3.30115238736704e-05, | |
| "loss": 0.5493, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7408906882591093, | |
| "grad_norm": 0.39184048570884483, | |
| "learning_rate": 3.296852223985148e-05, | |
| "loss": 0.5515, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7449392712550607, | |
| "grad_norm": 0.4438062741327208, | |
| "learning_rate": 3.2925416910074315e-05, | |
| "loss": 0.5503, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7489878542510122, | |
| "grad_norm": 0.3575873050941277, | |
| "learning_rate": 3.288220822900796e-05, | |
| "loss": 0.5535, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7530364372469636, | |
| "grad_norm": 0.33921903705557227, | |
| "learning_rate": 3.283889654214788e-05, | |
| "loss": 0.5485, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.757085020242915, | |
| "grad_norm": 0.3861344855911588, | |
| "learning_rate": 3.279548219581317e-05, | |
| "loss": 0.5412, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7611336032388665, | |
| "grad_norm": 0.3175072443084405, | |
| "learning_rate": 3.275196553714379e-05, | |
| "loss": 0.5449, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7651821862348178, | |
| "grad_norm": 0.3833838761838536, | |
| "learning_rate": 3.270834691409778e-05, | |
| "loss": 0.5544, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.3962827795156874, | |
| "learning_rate": 3.266462667544849e-05, | |
| "loss": 0.5546, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7732793522267207, | |
| "grad_norm": 0.4090104631013175, | |
| "learning_rate": 3.262080517078178e-05, | |
| "loss": 0.5471, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.777327935222672, | |
| "grad_norm": 0.3578179936933714, | |
| "learning_rate": 3.257688275049323e-05, | |
| "loss": 0.5462, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7813765182186234, | |
| "grad_norm": 0.3400987090972382, | |
| "learning_rate": 3.253285976578535e-05, | |
| "loss": 0.5502, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.785425101214575, | |
| "grad_norm": 0.37457444831083186, | |
| "learning_rate": 3.2488736568664756e-05, | |
| "loss": 0.5431, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.3855013246705002, | |
| "learning_rate": 3.244451351193935e-05, | |
| "loss": 0.5496, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7935222672064777, | |
| "grad_norm": 0.41646490212304954, | |
| "learning_rate": 3.240019094921551e-05, | |
| "loss": 0.5544, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7975708502024292, | |
| "grad_norm": 0.43639554705656025, | |
| "learning_rate": 3.235576923489527e-05, | |
| "loss": 0.5594, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.8016194331983806, | |
| "grad_norm": 0.3691907174299561, | |
| "learning_rate": 3.2311248724173476e-05, | |
| "loss": 0.5378, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.805668016194332, | |
| "grad_norm": 0.42150143819473795, | |
| "learning_rate": 3.226662977303494e-05, | |
| "loss": 0.5522, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.8097165991902835, | |
| "grad_norm": 0.433987603287083, | |
| "learning_rate": 3.2221912738251626e-05, | |
| "loss": 0.5627, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8137651821862348, | |
| "grad_norm": 0.425950199681544, | |
| "learning_rate": 3.2177097977379736e-05, | |
| "loss": 0.5291, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.8178137651821862, | |
| "grad_norm": 0.39864953742781867, | |
| "learning_rate": 3.21321858487569e-05, | |
| "loss": 0.5467, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8218623481781377, | |
| "grad_norm": 0.4426667707639431, | |
| "learning_rate": 3.208717671149931e-05, | |
| "loss": 0.5427, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.825910931174089, | |
| "grad_norm": 0.348225414700469, | |
| "learning_rate": 3.204207092549882e-05, | |
| "loss": 0.5512, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.8299595141700404, | |
| "grad_norm": 0.4212247643556705, | |
| "learning_rate": 3.19968688514201e-05, | |
| "loss": 0.5564, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.834008097165992, | |
| "grad_norm": 0.3339601182188654, | |
| "learning_rate": 3.195157085069774e-05, | |
| "loss": 0.5406, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8380566801619433, | |
| "grad_norm": 0.4318635495473487, | |
| "learning_rate": 3.190617728553332e-05, | |
| "loss": 0.543, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.3596069222271307, | |
| "learning_rate": 3.1860688518892606e-05, | |
| "loss": 0.5374, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.35067420341705613, | |
| "learning_rate": 3.181510491450254e-05, | |
| "loss": 0.5546, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8502024291497976, | |
| "grad_norm": 0.40509869569145945, | |
| "learning_rate": 3.176942683684842e-05, | |
| "loss": 0.5582, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.854251012145749, | |
| "grad_norm": 0.34771114688028903, | |
| "learning_rate": 3.1723654651170934e-05, | |
| "loss": 0.5522, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8582995951417005, | |
| "grad_norm": 0.3141090828207029, | |
| "learning_rate": 3.167778872346327e-05, | |
| "loss": 0.5419, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8623481781376519, | |
| "grad_norm": 0.33846003021559, | |
| "learning_rate": 3.163182942046815e-05, | |
| "loss": 0.5384, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8663967611336032, | |
| "grad_norm": 0.32474323546821876, | |
| "learning_rate": 3.158577710967494e-05, | |
| "loss": 0.5325, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8704453441295548, | |
| "grad_norm": 0.33177364775268287, | |
| "learning_rate": 3.153963215931669e-05, | |
| "loss": 0.5433, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.874493927125506, | |
| "grad_norm": 0.33686547143741097, | |
| "learning_rate": 3.14933949383672e-05, | |
| "loss": 0.5426, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8785425101214575, | |
| "grad_norm": 0.42253843863660007, | |
| "learning_rate": 3.144706581653805e-05, | |
| "loss": 0.5398, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.882591093117409, | |
| "grad_norm": 0.3199143784589805, | |
| "learning_rate": 3.140064516427566e-05, | |
| "loss": 0.5387, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8866396761133604, | |
| "grad_norm": 0.3243516036012094, | |
| "learning_rate": 3.135413335275833e-05, | |
| "loss": 0.5438, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8906882591093117, | |
| "grad_norm": 0.3871220903767453, | |
| "learning_rate": 3.130753075389327e-05, | |
| "loss": 0.5295, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.32219974384524463, | |
| "learning_rate": 3.1260837740313625e-05, | |
| "loss": 0.5527, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8987854251012146, | |
| "grad_norm": 0.38935312949756323, | |
| "learning_rate": 3.121405468537547e-05, | |
| "loss": 0.5457, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.902834008097166, | |
| "grad_norm": 0.3540734384166846, | |
| "learning_rate": 3.116718196315488e-05, | |
| "loss": 0.5464, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9068825910931175, | |
| "grad_norm": 0.40423215043963845, | |
| "learning_rate": 3.1120219948444894e-05, | |
| "loss": 0.5513, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.9109311740890689, | |
| "grad_norm": 0.29139047583874617, | |
| "learning_rate": 3.107316901675252e-05, | |
| "loss": 0.5384, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.9149797570850202, | |
| "grad_norm": 0.37089094103145387, | |
| "learning_rate": 3.1026029544295745e-05, | |
| "loss": 0.5489, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.9190283400809718, | |
| "grad_norm": 0.3300944169582965, | |
| "learning_rate": 3.097880190800057e-05, | |
| "loss": 0.5406, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.3675622016859678, | |
| "learning_rate": 3.093148648549788e-05, | |
| "loss": 0.5398, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9271255060728745, | |
| "grad_norm": 0.3737713641110076, | |
| "learning_rate": 3.088408365512055e-05, | |
| "loss": 0.5408, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.931174089068826, | |
| "grad_norm": 0.530742038696287, | |
| "learning_rate": 3.083659379590034e-05, | |
| "loss": 0.565, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.9352226720647774, | |
| "grad_norm": 0.40072532113784226, | |
| "learning_rate": 3.0789017287564913e-05, | |
| "loss": 0.5494, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9392712550607287, | |
| "grad_norm": 0.333526386161166, | |
| "learning_rate": 3.074135451053475e-05, | |
| "loss": 0.5456, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9433198380566803, | |
| "grad_norm": 0.46027923478417426, | |
| "learning_rate": 3.069360584592015e-05, | |
| "loss": 0.5424, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.3893010725077418, | |
| "learning_rate": 3.064577167551816e-05, | |
| "loss": 0.5315, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.951417004048583, | |
| "grad_norm": 0.33418354438780346, | |
| "learning_rate": 3.059785238180953e-05, | |
| "loss": 0.534, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9554655870445345, | |
| "grad_norm": 0.40475104271130113, | |
| "learning_rate": 3.054984834795566e-05, | |
| "loss": 0.5362, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9595141700404857, | |
| "grad_norm": 0.41302310570742995, | |
| "learning_rate": 3.0501759957795537e-05, | |
| "loss": 0.5578, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9635627530364372, | |
| "grad_norm": 0.3655137236734688, | |
| "learning_rate": 3.0453587595842627e-05, | |
| "loss": 0.5422, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9676113360323888, | |
| "grad_norm": 0.4361898480186977, | |
| "learning_rate": 3.0405331647281856e-05, | |
| "loss": 0.5405, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.97165991902834, | |
| "grad_norm": 0.3925781039870444, | |
| "learning_rate": 3.0356992497966508e-05, | |
| "loss": 0.5449, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9757085020242915, | |
| "grad_norm": 0.424778845297431, | |
| "learning_rate": 3.0308570534415137e-05, | |
| "loss": 0.5368, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.979757085020243, | |
| "grad_norm": 0.44143807459667644, | |
| "learning_rate": 3.0260066143808467e-05, | |
| "loss": 0.5436, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9838056680161942, | |
| "grad_norm": 0.40380055943777804, | |
| "learning_rate": 3.021147971398632e-05, | |
| "loss": 0.5331, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9878542510121457, | |
| "grad_norm": 0.35665457720361365, | |
| "learning_rate": 3.0162811633444478e-05, | |
| "loss": 0.5441, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9919028340080973, | |
| "grad_norm": 0.4581405134726967, | |
| "learning_rate": 3.0114062291331635e-05, | |
| "loss": 0.5333, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9959514170040484, | |
| "grad_norm": 0.4208255199516352, | |
| "learning_rate": 3.0065232077446217e-05, | |
| "loss": 0.556, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.40540702274497814, | |
| "learning_rate": 3.001632138223332e-05, | |
| "loss": 0.5357, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.0040485829959516, | |
| "grad_norm": 0.46205009793668234, | |
| "learning_rate": 2.9967330596781565e-05, | |
| "loss": 0.4923, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0080971659919027, | |
| "grad_norm": 0.6407762229190318, | |
| "learning_rate": 2.9918260112819952e-05, | |
| "loss": 0.4871, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0121457489878543, | |
| "grad_norm": 0.596946607315424, | |
| "learning_rate": 2.9869110322714778e-05, | |
| "loss": 0.4734, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.016194331983806, | |
| "grad_norm": 0.4730381046916302, | |
| "learning_rate": 2.9819881619466447e-05, | |
| "loss": 0.4775, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.020242914979757, | |
| "grad_norm": 0.4422543865908775, | |
| "learning_rate": 2.9770574396706364e-05, | |
| "loss": 0.4713, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0242914979757085, | |
| "grad_norm": 0.5298713651700834, | |
| "learning_rate": 2.972118904869376e-05, | |
| "loss": 0.4629, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02834008097166, | |
| "grad_norm": 0.43858633617484516, | |
| "learning_rate": 2.9671725970312563e-05, | |
| "loss": 0.4719, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.032388663967611, | |
| "grad_norm": 0.5082726206478139, | |
| "learning_rate": 2.9622185557068238e-05, | |
| "loss": 0.4775, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.0364372469635628, | |
| "grad_norm": 0.5374411165879134, | |
| "learning_rate": 2.95725682050846e-05, | |
| "loss": 0.4697, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0404858299595143, | |
| "grad_norm": 0.4081644811809975, | |
| "learning_rate": 2.9522874311100684e-05, | |
| "loss": 0.4697, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0445344129554655, | |
| "grad_norm": 0.42827900326335056, | |
| "learning_rate": 2.9473104272467526e-05, | |
| "loss": 0.4763, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.048582995951417, | |
| "grad_norm": 0.4801291930730278, | |
| "learning_rate": 2.942325848714505e-05, | |
| "loss": 0.4852, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.37176392700166694, | |
| "learning_rate": 2.9373337353698804e-05, | |
| "loss": 0.4844, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.0566801619433197, | |
| "grad_norm": 0.45190277404195706, | |
| "learning_rate": 2.932334127129686e-05, | |
| "loss": 0.4736, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.0607287449392713, | |
| "grad_norm": 0.38314603222757176, | |
| "learning_rate": 2.9273270639706547e-05, | |
| "loss": 0.4749, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.064777327935223, | |
| "grad_norm": 0.44211419562223314, | |
| "learning_rate": 2.922312585929131e-05, | |
| "loss": 0.4651, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.068825910931174, | |
| "grad_norm": 0.3791536593371242, | |
| "learning_rate": 2.9172907331007478e-05, | |
| "loss": 0.4879, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.0728744939271255, | |
| "grad_norm": 0.37741922434310393, | |
| "learning_rate": 2.9122615456401053e-05, | |
| "loss": 0.4856, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.3933483698614429, | |
| "learning_rate": 2.9072250637604535e-05, | |
| "loss": 0.4833, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.080971659919028, | |
| "grad_norm": 0.33114034729245695, | |
| "learning_rate": 2.9021813277333667e-05, | |
| "loss": 0.4787, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.08502024291498, | |
| "grad_norm": 0.36892257837943215, | |
| "learning_rate": 2.8971303778884244e-05, | |
| "loss": 0.4789, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.0890688259109313, | |
| "grad_norm": 0.31733077619553235, | |
| "learning_rate": 2.8920722546128863e-05, | |
| "loss": 0.4853, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0931174089068825, | |
| "grad_norm": 0.33007433701974626, | |
| "learning_rate": 2.887006998351372e-05, | |
| "loss": 0.4867, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.097165991902834, | |
| "grad_norm": 0.34188180523315165, | |
| "learning_rate": 2.881934649605535e-05, | |
| "loss": 0.4802, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.1012145748987856, | |
| "grad_norm": 0.4025092027982093, | |
| "learning_rate": 2.8768552489337408e-05, | |
| "loss": 0.4838, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.37140759024324044, | |
| "learning_rate": 2.8717688369507425e-05, | |
| "loss": 0.4732, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1093117408906883, | |
| "grad_norm": 0.39007322512295817, | |
| "learning_rate": 2.8666754543273535e-05, | |
| "loss": 0.4732, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.11336032388664, | |
| "grad_norm": 0.4070632343315749, | |
| "learning_rate": 2.8615751417901273e-05, | |
| "loss": 0.4505, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.117408906882591, | |
| "grad_norm": 0.39061476872502454, | |
| "learning_rate": 2.856467940121025e-05, | |
| "loss": 0.4845, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.1214574898785425, | |
| "grad_norm": 0.5017099172686267, | |
| "learning_rate": 2.8513538901570964e-05, | |
| "loss": 0.4722, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.125506072874494, | |
| "grad_norm": 0.3641457415724582, | |
| "learning_rate": 2.846233032790149e-05, | |
| "loss": 0.473, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1295546558704452, | |
| "grad_norm": 0.3626874493491777, | |
| "learning_rate": 2.84110540896642e-05, | |
| "loss": 0.4811, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.133603238866397, | |
| "grad_norm": 0.3211827806002674, | |
| "learning_rate": 2.8359710596862555e-05, | |
| "loss": 0.4714, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.1376518218623484, | |
| "grad_norm": 0.31845513062778447, | |
| "learning_rate": 2.8308300260037734e-05, | |
| "loss": 0.4685, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.1417004048582995, | |
| "grad_norm": 0.3719875828360997, | |
| "learning_rate": 2.8256823490265424e-05, | |
| "loss": 0.4819, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.145748987854251, | |
| "grad_norm": 0.30063061583278633, | |
| "learning_rate": 2.8205280699152515e-05, | |
| "loss": 0.4745, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1497975708502026, | |
| "grad_norm": 0.36365278852964866, | |
| "learning_rate": 2.8153672298833777e-05, | |
| "loss": 0.473, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.35858853852333133, | |
| "learning_rate": 2.8101998701968615e-05, | |
| "loss": 0.4924, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.31174247631293683, | |
| "learning_rate": 2.805026032173773e-05, | |
| "loss": 0.4783, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.161943319838057, | |
| "grad_norm": 0.32477886343247875, | |
| "learning_rate": 2.7998457571839832e-05, | |
| "loss": 0.4676, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.165991902834008, | |
| "grad_norm": 0.3154120133804794, | |
| "learning_rate": 2.7946590866488322e-05, | |
| "loss": 0.4677, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1700404858299596, | |
| "grad_norm": 0.3592402340729052, | |
| "learning_rate": 2.7894660620408e-05, | |
| "loss": 0.4718, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.174089068825911, | |
| "grad_norm": 0.3177448563887081, | |
| "learning_rate": 2.784266724883173e-05, | |
| "loss": 0.4801, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.1781376518218623, | |
| "grad_norm": 0.3415220859179897, | |
| "learning_rate": 2.7790611167497128e-05, | |
| "loss": 0.4868, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.182186234817814, | |
| "grad_norm": 0.34959305796216855, | |
| "learning_rate": 2.7738492792643224e-05, | |
| "loss": 0.4795, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.1862348178137654, | |
| "grad_norm": 0.32348164835646037, | |
| "learning_rate": 2.768631254100716e-05, | |
| "loss": 0.4885, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1902834008097165, | |
| "grad_norm": 0.376387335945229, | |
| "learning_rate": 2.7634070829820836e-05, | |
| "loss": 0.4737, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.194331983805668, | |
| "grad_norm": 0.36326794792472933, | |
| "learning_rate": 2.7581768076807592e-05, | |
| "loss": 0.4731, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.1983805668016196, | |
| "grad_norm": 0.3669977458232998, | |
| "learning_rate": 2.7529404700178832e-05, | |
| "loss": 0.4763, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.2024291497975708, | |
| "grad_norm": 0.39000099007999306, | |
| "learning_rate": 2.7476981118630735e-05, | |
| "loss": 0.48, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.2064777327935223, | |
| "grad_norm": 0.3954887819525922, | |
| "learning_rate": 2.742449775134085e-05, | |
| "loss": 0.4842, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.3632775415594135, | |
| "learning_rate": 2.7371955017964788e-05, | |
| "loss": 0.4869, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.214574898785425, | |
| "grad_norm": 0.40866829842723157, | |
| "learning_rate": 2.7319353338632845e-05, | |
| "loss": 0.4666, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.2186234817813766, | |
| "grad_norm": 0.3814871715738774, | |
| "learning_rate": 2.7266693133946645e-05, | |
| "loss": 0.4731, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.2226720647773277, | |
| "grad_norm": 0.41759713743644306, | |
| "learning_rate": 2.7213974824975785e-05, | |
| "loss": 0.4659, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2267206477732793, | |
| "grad_norm": 0.3501215177249891, | |
| "learning_rate": 2.7161198833254464e-05, | |
| "loss": 0.4868, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.37429499398326027, | |
| "learning_rate": 2.7108365580778095e-05, | |
| "loss": 0.48, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.234817813765182, | |
| "grad_norm": 0.3884470037827856, | |
| "learning_rate": 2.705547548999996e-05, | |
| "loss": 0.4801, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.2388663967611335, | |
| "grad_norm": 0.3721402701079292, | |
| "learning_rate": 2.7002528983827817e-05, | |
| "loss": 0.4696, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.242914979757085, | |
| "grad_norm": 0.3347550958816996, | |
| "learning_rate": 2.6949526485620517e-05, | |
| "loss": 0.4891, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.246963562753036, | |
| "grad_norm": 0.3963406155139975, | |
| "learning_rate": 2.6896468419184623e-05, | |
| "loss": 0.486, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.251012145748988, | |
| "grad_norm": 0.33197518816210114, | |
| "learning_rate": 2.6843355208771024e-05, | |
| "loss": 0.487, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.2550607287449393, | |
| "grad_norm": 0.3507097403760946, | |
| "learning_rate": 2.6790187279071536e-05, | |
| "loss": 0.4604, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.2591093117408905, | |
| "grad_norm": 0.33178444134838186, | |
| "learning_rate": 2.67369650552155e-05, | |
| "loss": 0.4855, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.3725557053470955, | |
| "learning_rate": 2.6683688962766407e-05, | |
| "loss": 0.4671, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.2672064777327936, | |
| "grad_norm": 0.31899792200517846, | |
| "learning_rate": 2.6630359427718468e-05, | |
| "loss": 0.4748, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2712550607287447, | |
| "grad_norm": 0.34794906418998367, | |
| "learning_rate": 2.6576976876493225e-05, | |
| "loss": 0.483, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.30791045125885913, | |
| "learning_rate": 2.652354173593614e-05, | |
| "loss": 0.4689, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.279352226720648, | |
| "grad_norm": 0.3274236953448683, | |
| "learning_rate": 2.647005443331317e-05, | |
| "loss": 0.4657, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.283400809716599, | |
| "grad_norm": 0.2959433848732403, | |
| "learning_rate": 2.6416515396307354e-05, | |
| "loss": 0.4682, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.2874493927125505, | |
| "grad_norm": 0.32802799945156164, | |
| "learning_rate": 2.636292505301542e-05, | |
| "loss": 0.4818, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.291497975708502, | |
| "grad_norm": 0.3049720434502394, | |
| "learning_rate": 2.630928383194431e-05, | |
| "loss": 0.4839, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.2955465587044532, | |
| "grad_norm": 0.27627645093208963, | |
| "learning_rate": 2.625559216200781e-05, | |
| "loss": 0.4754, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.299595141700405, | |
| "grad_norm": 0.3434300734979695, | |
| "learning_rate": 2.620185047252308e-05, | |
| "loss": 0.4737, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.3036437246963564, | |
| "grad_norm": 0.27293351498277724, | |
| "learning_rate": 2.614805919320723e-05, | |
| "loss": 0.4853, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.3394482740350358, | |
| "learning_rate": 2.6094218754173895e-05, | |
| "loss": 0.4822, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.311740890688259, | |
| "grad_norm": 0.32327716480323243, | |
| "learning_rate": 2.60403295859298e-05, | |
| "loss": 0.4783, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.3655281273082933, | |
| "learning_rate": 2.5986392119371297e-05, | |
| "loss": 0.4883, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.3198380566801617, | |
| "grad_norm": 0.3319989955223436, | |
| "learning_rate": 2.5932406785780928e-05, | |
| "loss": 0.4759, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.3238866396761133, | |
| "grad_norm": 0.3526111156569753, | |
| "learning_rate": 2.5878374016823987e-05, | |
| "loss": 0.4712, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.327935222672065, | |
| "grad_norm": 0.3514487343022425, | |
| "learning_rate": 2.5824294244545045e-05, | |
| "loss": 0.4686, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.331983805668016, | |
| "grad_norm": 0.33176560314173476, | |
| "learning_rate": 2.577016790136453e-05, | |
| "loss": 0.4899, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.3360323886639676, | |
| "grad_norm": 0.31270825371252314, | |
| "learning_rate": 2.5715995420075225e-05, | |
| "loss": 0.4872, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.340080971659919, | |
| "grad_norm": 0.33325012194238746, | |
| "learning_rate": 2.5661777233838864e-05, | |
| "loss": 0.4761, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.3441295546558703, | |
| "grad_norm": 0.3381192941847261, | |
| "learning_rate": 2.5607513776182616e-05, | |
| "loss": 0.4767, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.348178137651822, | |
| "grad_norm": 0.34227698333354795, | |
| "learning_rate": 2.5553205480995633e-05, | |
| "loss": 0.494, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3522267206477734, | |
| "grad_norm": 0.3785247730038111, | |
| "learning_rate": 2.5498852782525604e-05, | |
| "loss": 0.4824, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.3562753036437245, | |
| "grad_norm": 0.3053023628784075, | |
| "learning_rate": 2.5444456115375252e-05, | |
| "loss": 0.4768, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.360323886639676, | |
| "grad_norm": 0.3544189248828337, | |
| "learning_rate": 2.539001591449889e-05, | |
| "loss": 0.4674, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.3643724696356276, | |
| "grad_norm": 0.32654887071253547, | |
| "learning_rate": 2.53355326151989e-05, | |
| "loss": 0.4786, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.3277418434358984, | |
| "learning_rate": 2.5281006653122288e-05, | |
| "loss": 0.4857, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.3724696356275303, | |
| "grad_norm": 0.3728537909696274, | |
| "learning_rate": 2.5226438464257217e-05, | |
| "loss": 0.4979, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.376518218623482, | |
| "grad_norm": 0.31652638741530303, | |
| "learning_rate": 2.5171828484929447e-05, | |
| "loss": 0.4838, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.380566801619433, | |
| "grad_norm": 0.32548754913133593, | |
| "learning_rate": 2.5117177151798934e-05, | |
| "loss": 0.469, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.34754068146711664, | |
| "learning_rate": 2.506248490185627e-05, | |
| "loss": 0.4755, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.388663967611336, | |
| "grad_norm": 0.33123794311584287, | |
| "learning_rate": 2.500775217241923e-05, | |
| "loss": 0.4859, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.3927125506072873, | |
| "grad_norm": 0.28083766933559867, | |
| "learning_rate": 2.4952979401129266e-05, | |
| "loss": 0.4891, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.396761133603239, | |
| "grad_norm": 0.34018493278015, | |
| "learning_rate": 2.4898167025948e-05, | |
| "loss": 0.4632, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.4008097165991904, | |
| "grad_norm": 0.324478825876648, | |
| "learning_rate": 2.4843315485153703e-05, | |
| "loss": 0.4851, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.4048582995951415, | |
| "grad_norm": 0.30868613790586275, | |
| "learning_rate": 2.4788425217337856e-05, | |
| "loss": 0.476, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.408906882591093, | |
| "grad_norm": 0.3169357568726334, | |
| "learning_rate": 2.4733496661401572e-05, | |
| "loss": 0.4899, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.4129554655870447, | |
| "grad_norm": 0.35459090880635424, | |
| "learning_rate": 2.4678530256552114e-05, | |
| "loss": 0.4962, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.417004048582996, | |
| "grad_norm": 0.28187715416610293, | |
| "learning_rate": 2.46235264422994e-05, | |
| "loss": 0.4888, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.32652116004819437, | |
| "learning_rate": 2.4568485658452452e-05, | |
| "loss": 0.471, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.425101214574899, | |
| "grad_norm": 0.3301366350676471, | |
| "learning_rate": 2.451340834511592e-05, | |
| "loss": 0.4749, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.42914979757085, | |
| "grad_norm": 0.29349135455474873, | |
| "learning_rate": 2.4458294942686528e-05, | |
| "loss": 0.4749, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4331983805668016, | |
| "grad_norm": 0.2766921647896269, | |
| "learning_rate": 2.440314589184958e-05, | |
| "loss": 0.4912, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.437246963562753, | |
| "grad_norm": 0.33108476546954146, | |
| "learning_rate": 2.434796163357541e-05, | |
| "loss": 0.4831, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.4412955465587043, | |
| "grad_norm": 0.31065914374059356, | |
| "learning_rate": 2.429274260911588e-05, | |
| "loss": 0.4852, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.445344129554656, | |
| "grad_norm": 0.3409358457625251, | |
| "learning_rate": 2.423748926000084e-05, | |
| "loss": 0.477, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.4493927125506074, | |
| "grad_norm": 0.29020353811498123, | |
| "learning_rate": 2.418220202803459e-05, | |
| "loss": 0.4855, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.4534412955465585, | |
| "grad_norm": 0.33972003737665984, | |
| "learning_rate": 2.412688135529237e-05, | |
| "loss": 0.4933, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.45748987854251, | |
| "grad_norm": 0.2840535724086914, | |
| "learning_rate": 2.4071527684116807e-05, | |
| "loss": 0.4867, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.3398999053488498, | |
| "learning_rate": 2.4016141457114372e-05, | |
| "loss": 0.4844, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.465587044534413, | |
| "grad_norm": 0.2609251555034588, | |
| "learning_rate": 2.396072311715187e-05, | |
| "loss": 0.4881, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.4696356275303644, | |
| "grad_norm": 0.320533966996985, | |
| "learning_rate": 2.3905273107352863e-05, | |
| "loss": 0.4778, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.3121198899037089, | |
| "learning_rate": 2.3849791871094157e-05, | |
| "loss": 0.4714, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.477732793522267, | |
| "grad_norm": 0.3071669928736951, | |
| "learning_rate": 2.379427985200223e-05, | |
| "loss": 0.4727, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.4817813765182186, | |
| "grad_norm": 0.2895009349547413, | |
| "learning_rate": 2.373873749394972e-05, | |
| "loss": 0.4803, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.48582995951417, | |
| "grad_norm": 0.28898409043323126, | |
| "learning_rate": 2.3683165241051836e-05, | |
| "loss": 0.481, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.4898785425101213, | |
| "grad_norm": 0.29855898063570263, | |
| "learning_rate": 2.3627563537662823e-05, | |
| "loss": 0.4685, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.493927125506073, | |
| "grad_norm": 0.31017323577661027, | |
| "learning_rate": 2.3571932828372433e-05, | |
| "loss": 0.4758, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.4979757085020244, | |
| "grad_norm": 0.29839855822725897, | |
| "learning_rate": 2.351627355800233e-05, | |
| "loss": 0.4851, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.5020242914979756, | |
| "grad_norm": 0.34053774686690175, | |
| "learning_rate": 2.346058617160256e-05, | |
| "loss": 0.4686, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.506072874493927, | |
| "grad_norm": 0.300854947019146, | |
| "learning_rate": 2.3404871114447976e-05, | |
| "loss": 0.4813, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.5101214574898787, | |
| "grad_norm": 0.43991278690741376, | |
| "learning_rate": 2.334912883203469e-05, | |
| "loss": 0.4787, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.51417004048583, | |
| "grad_norm": 0.3120882168495416, | |
| "learning_rate": 2.3293359770076505e-05, | |
| "loss": 0.4862, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.5182186234817814, | |
| "grad_norm": 0.35114526752579933, | |
| "learning_rate": 2.323756437450136e-05, | |
| "loss": 0.4756, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.522267206477733, | |
| "grad_norm": 0.29977115618972505, | |
| "learning_rate": 2.3181743091447754e-05, | |
| "loss": 0.4829, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.3189511315100978, | |
| "learning_rate": 2.3125896367261164e-05, | |
| "loss": 0.486, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5303643724696356, | |
| "grad_norm": 0.3050771981148409, | |
| "learning_rate": 2.307002464849052e-05, | |
| "loss": 0.4738, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.534412955465587, | |
| "grad_norm": 0.3358024590529171, | |
| "learning_rate": 2.30141283818846e-05, | |
| "loss": 0.4766, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.313292693404395, | |
| "learning_rate": 2.2958208014388462e-05, | |
| "loss": 0.4723, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.54251012145749, | |
| "grad_norm": 0.32033291468096464, | |
| "learning_rate": 2.2902263993139857e-05, | |
| "loss": 0.4752, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.5465587044534415, | |
| "grad_norm": 0.3345372221883433, | |
| "learning_rate": 2.2846296765465708e-05, | |
| "loss": 0.483, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.5506072874493926, | |
| "grad_norm": 0.319946213797381, | |
| "learning_rate": 2.2790306778878465e-05, | |
| "loss": 0.48, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.554655870445344, | |
| "grad_norm": 0.3310946237927437, | |
| "learning_rate": 2.273429448107257e-05, | |
| "loss": 0.4839, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.5587044534412957, | |
| "grad_norm": 0.2912676438397428, | |
| "learning_rate": 2.2678260319920868e-05, | |
| "loss": 0.4831, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.562753036437247, | |
| "grad_norm": 0.3047760555209865, | |
| "learning_rate": 2.262220474347102e-05, | |
| "loss": 0.4881, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.5668016194331984, | |
| "grad_norm": 0.3236255067373046, | |
| "learning_rate": 2.2566128199941918e-05, | |
| "loss": 0.4787, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.57085020242915, | |
| "grad_norm": 0.2834795238825436, | |
| "learning_rate": 2.2510031137720104e-05, | |
| "loss": 0.4797, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.574898785425101, | |
| "grad_norm": 0.29610205735162304, | |
| "learning_rate": 2.2453914005356206e-05, | |
| "loss": 0.4746, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.29038570081754583, | |
| "learning_rate": 2.2397777251561302e-05, | |
| "loss": 0.4873, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.582995951417004, | |
| "grad_norm": 0.27242401910885033, | |
| "learning_rate": 2.2341621325203393e-05, | |
| "loss": 0.4863, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.5870445344129553, | |
| "grad_norm": 0.3254402326151251, | |
| "learning_rate": 2.2285446675303764e-05, | |
| "loss": 0.4849, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.591093117408907, | |
| "grad_norm": 0.27558182199015857, | |
| "learning_rate": 2.2229253751033423e-05, | |
| "loss": 0.4752, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.5951417004048585, | |
| "grad_norm": 0.29935369416620133, | |
| "learning_rate": 2.2173043001709492e-05, | |
| "loss": 0.4709, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.5991902834008096, | |
| "grad_norm": 0.29244124648363623, | |
| "learning_rate": 2.2116814876791627e-05, | |
| "loss": 0.4946, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.603238866396761, | |
| "grad_norm": 0.32078290611911753, | |
| "learning_rate": 2.2060569825878413e-05, | |
| "loss": 0.4738, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.6072874493927127, | |
| "grad_norm": 0.27711431928911806, | |
| "learning_rate": 2.2004308298703777e-05, | |
| "loss": 0.4668, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.611336032388664, | |
| "grad_norm": 0.340906026568954, | |
| "learning_rate": 2.1948030745133396e-05, | |
| "loss": 0.4848, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.28622790421843736, | |
| "learning_rate": 2.1891737615161083e-05, | |
| "loss": 0.4773, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.619433198380567, | |
| "grad_norm": 0.3269498292618273, | |
| "learning_rate": 2.1835429358905218e-05, | |
| "loss": 0.4829, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.623481781376518, | |
| "grad_norm": 0.2548321257462986, | |
| "learning_rate": 2.17791064266051e-05, | |
| "loss": 0.4879, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.6275303643724697, | |
| "grad_norm": 0.30664312047295156, | |
| "learning_rate": 2.172276926861741e-05, | |
| "loss": 0.4874, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.27837157110921545, | |
| "learning_rate": 2.166641833541255e-05, | |
| "loss": 0.4744, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6356275303643724, | |
| "grad_norm": 0.3025822085486682, | |
| "learning_rate": 2.161005407757109e-05, | |
| "loss": 0.4825, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.639676113360324, | |
| "grad_norm": 0.3926404461980294, | |
| "learning_rate": 2.155367694578013e-05, | |
| "loss": 0.4861, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.6437246963562755, | |
| "grad_norm": 0.28966513080011363, | |
| "learning_rate": 2.1497287390829712e-05, | |
| "loss": 0.4631, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.6477732793522266, | |
| "grad_norm": 0.38538383111288377, | |
| "learning_rate": 2.1440885863609208e-05, | |
| "loss": 0.4881, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.651821862348178, | |
| "grad_norm": 0.2786465108113584, | |
| "learning_rate": 2.1384472815103737e-05, | |
| "loss": 0.4848, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6558704453441297, | |
| "grad_norm": 0.305503208785887, | |
| "learning_rate": 2.1328048696390534e-05, | |
| "loss": 0.4845, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.659919028340081, | |
| "grad_norm": 0.31949504539270973, | |
| "learning_rate": 2.1271613958635342e-05, | |
| "loss": 0.4809, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.6639676113360324, | |
| "grad_norm": 0.3118625355101463, | |
| "learning_rate": 2.1215169053088825e-05, | |
| "loss": 0.4783, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.668016194331984, | |
| "grad_norm": 0.32433462136500363, | |
| "learning_rate": 2.1158714431082937e-05, | |
| "loss": 0.4713, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.672064777327935, | |
| "grad_norm": 0.290362245980769, | |
| "learning_rate": 2.110225054402733e-05, | |
| "loss": 0.4825, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.6761133603238867, | |
| "grad_norm": 0.34433360791430484, | |
| "learning_rate": 2.1045777843405747e-05, | |
| "loss": 0.4789, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.6801619433198383, | |
| "grad_norm": 0.3129809114985225, | |
| "learning_rate": 2.09892967807724e-05, | |
| "loss": 0.4848, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.30365990914125723, | |
| "learning_rate": 2.093280780774835e-05, | |
| "loss": 0.4719, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.688259109311741, | |
| "grad_norm": 0.32319042444293833, | |
| "learning_rate": 2.087631137601793e-05, | |
| "loss": 0.4767, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.31589692714971196, | |
| "learning_rate": 2.0819807937325104e-05, | |
| "loss": 0.4778, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.6963562753036436, | |
| "grad_norm": 0.3260105200839393, | |
| "learning_rate": 2.0763297943469847e-05, | |
| "loss": 0.4858, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.700404858299595, | |
| "grad_norm": 0.3711121641607142, | |
| "learning_rate": 2.070678184630458e-05, | |
| "loss": 0.5035, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.7044534412955468, | |
| "grad_norm": 0.31182187148499463, | |
| "learning_rate": 2.06502600977305e-05, | |
| "loss": 0.4757, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.708502024291498, | |
| "grad_norm": 0.4451753045770485, | |
| "learning_rate": 2.0593733149694008e-05, | |
| "loss": 0.4786, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.7125506072874495, | |
| "grad_norm": 0.31564005080930574, | |
| "learning_rate": 2.0537201454183074e-05, | |
| "loss": 0.4785, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.716599190283401, | |
| "grad_norm": 0.3430885011349374, | |
| "learning_rate": 2.0480665463223626e-05, | |
| "loss": 0.4886, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.720647773279352, | |
| "grad_norm": 0.2759924466301294, | |
| "learning_rate": 2.042412562887594e-05, | |
| "loss": 0.4922, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7246963562753037, | |
| "grad_norm": 0.29514656572204007, | |
| "learning_rate": 2.0367582403231033e-05, | |
| "loss": 0.4651, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.7287449392712553, | |
| "grad_norm": 0.2812407772263676, | |
| "learning_rate": 2.0311036238407023e-05, | |
| "loss": 0.4866, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7327935222672064, | |
| "grad_norm": 0.32991322893826264, | |
| "learning_rate": 2.0254487586545534e-05, | |
| "loss": 0.4673, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.3040963193557163, | |
| "learning_rate": 2.0197936899808088e-05, | |
| "loss": 0.4646, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.7408906882591095, | |
| "grad_norm": 0.35700830992756233, | |
| "learning_rate": 2.0141384630372466e-05, | |
| "loss": 0.4717, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7449392712550607, | |
| "grad_norm": 0.26950411589623946, | |
| "learning_rate": 2.0084831230429106e-05, | |
| "loss": 0.4725, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.748987854251012, | |
| "grad_norm": 0.312614077570178, | |
| "learning_rate": 2.0028277152177492e-05, | |
| "loss": 0.4942, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.753036437246964, | |
| "grad_norm": 0.29551089791997165, | |
| "learning_rate": 1.9971722847822518e-05, | |
| "loss": 0.478, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.757085020242915, | |
| "grad_norm": 0.35455269748228313, | |
| "learning_rate": 1.99151687695709e-05, | |
| "loss": 0.4717, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7611336032388665, | |
| "grad_norm": 0.32196564701510866, | |
| "learning_rate": 1.9858615369627537e-05, | |
| "loss": 0.4845, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.765182186234818, | |
| "grad_norm": 0.32971301502610834, | |
| "learning_rate": 1.9802063100191916e-05, | |
| "loss": 0.4797, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.311922413148334, | |
| "learning_rate": 1.974551241345447e-05, | |
| "loss": 0.4843, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.7732793522267207, | |
| "grad_norm": 0.28246475852775865, | |
| "learning_rate": 1.9688963761592984e-05, | |
| "loss": 0.4802, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7773279352226723, | |
| "grad_norm": 0.28983240800149046, | |
| "learning_rate": 1.963241759676897e-05, | |
| "loss": 0.4664, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.7813765182186234, | |
| "grad_norm": 0.2979821745183721, | |
| "learning_rate": 1.9575874371124066e-05, | |
| "loss": 0.4813, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.785425101214575, | |
| "grad_norm": 0.3045347954756811, | |
| "learning_rate": 1.951933453677638e-05, | |
| "loss": 0.4878, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.28156086436739036, | |
| "learning_rate": 1.9462798545816932e-05, | |
| "loss": 0.4881, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.7935222672064777, | |
| "grad_norm": 0.3244721658906315, | |
| "learning_rate": 1.9406266850305996e-05, | |
| "loss": 0.4805, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7975708502024292, | |
| "grad_norm": 0.2700498321032519, | |
| "learning_rate": 1.9349739902269503e-05, | |
| "loss": 0.4664, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.801619433198381, | |
| "grad_norm": 0.3504078829568257, | |
| "learning_rate": 1.9293218153695425e-05, | |
| "loss": 0.4898, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.805668016194332, | |
| "grad_norm": 0.2920728117409114, | |
| "learning_rate": 1.9236702056530153e-05, | |
| "loss": 0.4764, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.8097165991902835, | |
| "grad_norm": 0.3370050518868706, | |
| "learning_rate": 1.9180192062674907e-05, | |
| "loss": 0.4867, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.813765182186235, | |
| "grad_norm": 0.292862415136467, | |
| "learning_rate": 1.9123688623982076e-05, | |
| "loss": 0.4834, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.817813765182186, | |
| "grad_norm": 0.3176176267748534, | |
| "learning_rate": 1.906719219225165e-05, | |
| "loss": 0.4855, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.8218623481781377, | |
| "grad_norm": 0.2869394755700951, | |
| "learning_rate": 1.9010703219227603e-05, | |
| "loss": 0.4859, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.8259109311740893, | |
| "grad_norm": 0.3193011364913069, | |
| "learning_rate": 1.8954222156594256e-05, | |
| "loss": 0.4841, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.8299595141700404, | |
| "grad_norm": 0.2808785450626554, | |
| "learning_rate": 1.8897749455972673e-05, | |
| "loss": 0.4772, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.834008097165992, | |
| "grad_norm": 0.3647139938041626, | |
| "learning_rate": 1.884128556891707e-05, | |
| "loss": 0.4733, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8380566801619436, | |
| "grad_norm": 0.279319013127972, | |
| "learning_rate": 1.878483094691119e-05, | |
| "loss": 0.4897, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.325481220961451, | |
| "learning_rate": 1.8728386041364664e-05, | |
| "loss": 0.4827, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.30637044529836505, | |
| "learning_rate": 1.867195130360947e-05, | |
| "loss": 0.4686, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.850202429149798, | |
| "grad_norm": 0.31505077357652966, | |
| "learning_rate": 1.8615527184896263e-05, | |
| "loss": 0.4857, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.854251012145749, | |
| "grad_norm": 0.3334918977863746, | |
| "learning_rate": 1.8559114136390795e-05, | |
| "loss": 0.4745, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8582995951417005, | |
| "grad_norm": 0.32762919985149747, | |
| "learning_rate": 1.8502712609170298e-05, | |
| "loss": 0.4832, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.862348178137652, | |
| "grad_norm": 0.34118007955738416, | |
| "learning_rate": 1.8446323054219876e-05, | |
| "loss": 0.4742, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.866396761133603, | |
| "grad_norm": 0.30570950777028116, | |
| "learning_rate": 1.838994592242891e-05, | |
| "loss": 0.4743, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.8704453441295548, | |
| "grad_norm": 0.31243584510066746, | |
| "learning_rate": 1.8333581664587453e-05, | |
| "loss": 0.486, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.8744939271255063, | |
| "grad_norm": 0.2948532209255868, | |
| "learning_rate": 1.8277230731382593e-05, | |
| "loss": 0.4575, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8785425101214575, | |
| "grad_norm": 0.31456708953249585, | |
| "learning_rate": 1.82208935733949e-05, | |
| "loss": 0.4832, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.882591093117409, | |
| "grad_norm": 0.2943223138446065, | |
| "learning_rate": 1.8164570641094793e-05, | |
| "loss": 0.4879, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.8866396761133606, | |
| "grad_norm": 0.31850784869559984, | |
| "learning_rate": 1.810826238483892e-05, | |
| "loss": 0.4742, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.8906882591093117, | |
| "grad_norm": 0.2939851303271168, | |
| "learning_rate": 1.8051969254866608e-05, | |
| "loss": 0.4842, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.3366571932863712, | |
| "learning_rate": 1.7995691701296226e-05, | |
| "loss": 0.4603, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.898785425101215, | |
| "grad_norm": 0.2929268822691067, | |
| "learning_rate": 1.7939430174121593e-05, | |
| "loss": 0.4908, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.902834008097166, | |
| "grad_norm": 0.3305934306754584, | |
| "learning_rate": 1.788318512320838e-05, | |
| "loss": 0.479, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.9068825910931175, | |
| "grad_norm": 0.37152949519982165, | |
| "learning_rate": 1.782695699829051e-05, | |
| "loss": 0.4841, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.910931174089069, | |
| "grad_norm": 0.27230948030323554, | |
| "learning_rate": 1.7770746248966587e-05, | |
| "loss": 0.4726, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.91497975708502, | |
| "grad_norm": 0.3592021120861283, | |
| "learning_rate": 1.7714553324696243e-05, | |
| "loss": 0.4754, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.919028340080972, | |
| "grad_norm": 0.2633214443009097, | |
| "learning_rate": 1.7658378674796614e-05, | |
| "loss": 0.4795, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.4038302466305554, | |
| "learning_rate": 1.76022227484387e-05, | |
| "loss": 0.4785, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.9271255060728745, | |
| "grad_norm": 0.2603617286650775, | |
| "learning_rate": 1.7546085994643807e-05, | |
| "loss": 0.4823, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.931174089068826, | |
| "grad_norm": 0.3408534161761949, | |
| "learning_rate": 1.7489968862279902e-05, | |
| "loss": 0.4789, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9352226720647776, | |
| "grad_norm": 0.28845695625042433, | |
| "learning_rate": 1.743387180005809e-05, | |
| "loss": 0.4857, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.9392712550607287, | |
| "grad_norm": 0.3038412527732128, | |
| "learning_rate": 1.737779525652899e-05, | |
| "loss": 0.4754, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.9433198380566803, | |
| "grad_norm": 0.3107204151501639, | |
| "learning_rate": 1.732173968007914e-05, | |
| "loss": 0.4901, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.30441024705502817, | |
| "learning_rate": 1.7265705518927437e-05, | |
| "loss": 0.4779, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.951417004048583, | |
| "grad_norm": 0.29516464031741324, | |
| "learning_rate": 1.7209693221121542e-05, | |
| "loss": 0.4814, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.9554655870445345, | |
| "grad_norm": 0.2910047619095074, | |
| "learning_rate": 1.7153703234534302e-05, | |
| "loss": 0.4794, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9595141700404857, | |
| "grad_norm": 0.2756119247676066, | |
| "learning_rate": 1.709773600686015e-05, | |
| "loss": 0.4749, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.9635627530364372, | |
| "grad_norm": 0.2800311187129712, | |
| "learning_rate": 1.7041791985611548e-05, | |
| "loss": 0.4895, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.967611336032389, | |
| "grad_norm": 0.2928914305477768, | |
| "learning_rate": 1.69858716181154e-05, | |
| "loss": 0.4888, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.97165991902834, | |
| "grad_norm": 0.27469695969976615, | |
| "learning_rate": 1.692997535150948e-05, | |
| "loss": 0.476, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.9757085020242915, | |
| "grad_norm": 0.30964819075001576, | |
| "learning_rate": 1.687410363273884e-05, | |
| "loss": 0.4702, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.979757085020243, | |
| "grad_norm": 0.2613933759496942, | |
| "learning_rate": 1.6818256908552257e-05, | |
| "loss": 0.484, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.983805668016194, | |
| "grad_norm": 0.3393436273418995, | |
| "learning_rate": 1.6762435625498646e-05, | |
| "loss": 0.4856, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.9878542510121457, | |
| "grad_norm": 0.2940758827446396, | |
| "learning_rate": 1.67066402299235e-05, | |
| "loss": 0.4805, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.9919028340080973, | |
| "grad_norm": 0.30260649299742604, | |
| "learning_rate": 1.6650871167965313e-05, | |
| "loss": 0.4677, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.9959514170040484, | |
| "grad_norm": 0.2952801147421949, | |
| "learning_rate": 1.6595128885552028e-05, | |
| "loss": 0.4842, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.2841978548583681, | |
| "learning_rate": 1.653941382839745e-05, | |
| "loss": 0.4811, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.0040485829959516, | |
| "grad_norm": 0.48575838070179433, | |
| "learning_rate": 1.6483726441997673e-05, | |
| "loss": 0.4167, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.0080971659919027, | |
| "grad_norm": 0.42406214387932323, | |
| "learning_rate": 1.642806717162757e-05, | |
| "loss": 0.4136, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.0121457489878543, | |
| "grad_norm": 0.5541797976805712, | |
| "learning_rate": 1.637243646233718e-05, | |
| "loss": 0.428, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.016194331983806, | |
| "grad_norm": 0.5364986836156773, | |
| "learning_rate": 1.6316834758948174e-05, | |
| "loss": 0.397, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.020242914979757, | |
| "grad_norm": 0.40149577456745766, | |
| "learning_rate": 1.6261262506050282e-05, | |
| "loss": 0.4126, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.0242914979757085, | |
| "grad_norm": 0.39661867437448556, | |
| "learning_rate": 1.620572014799777e-05, | |
| "loss": 0.4144, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.02834008097166, | |
| "grad_norm": 0.4179295675323336, | |
| "learning_rate": 1.6150208128905857e-05, | |
| "loss": 0.4247, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.032388663967611, | |
| "grad_norm": 0.4474180676616743, | |
| "learning_rate": 1.6094726892647147e-05, | |
| "loss": 0.4075, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0364372469635628, | |
| "grad_norm": 0.3052479258776868, | |
| "learning_rate": 1.6039276882848135e-05, | |
| "loss": 0.4111, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0404858299595143, | |
| "grad_norm": 0.3750421600645596, | |
| "learning_rate": 1.5983858542885635e-05, | |
| "loss": 0.4222, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.0445344129554655, | |
| "grad_norm": 0.33095041147100973, | |
| "learning_rate": 1.5928472315883203e-05, | |
| "loss": 0.4181, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.048582995951417, | |
| "grad_norm": 0.33480733250853356, | |
| "learning_rate": 1.5873118644707633e-05, | |
| "loss": 0.4212, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.3187571673644541, | |
| "learning_rate": 1.5817797971965413e-05, | |
| "loss": 0.4102, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.0566801619433197, | |
| "grad_norm": 0.32902827178738664, | |
| "learning_rate": 1.576251073999917e-05, | |
| "loss": 0.4084, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.0607287449392713, | |
| "grad_norm": 0.3360612202769251, | |
| "learning_rate": 1.5707257390884126e-05, | |
| "loss": 0.4176, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.064777327935223, | |
| "grad_norm": 0.31172078703082906, | |
| "learning_rate": 1.5652038366424595e-05, | |
| "loss": 0.4017, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.068825910931174, | |
| "grad_norm": 0.3149093057804946, | |
| "learning_rate": 1.5596854108150424e-05, | |
| "loss": 0.4105, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.0728744939271255, | |
| "grad_norm": 0.32430906341124177, | |
| "learning_rate": 1.5541705057313476e-05, | |
| "loss": 0.4153, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.2878438587230576, | |
| "learning_rate": 1.5486591654884086e-05, | |
| "loss": 0.4153, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.080971659919028, | |
| "grad_norm": 0.30705407885850405, | |
| "learning_rate": 1.543151434154755e-05, | |
| "loss": 0.4047, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.08502024291498, | |
| "grad_norm": 0.32497240738953365, | |
| "learning_rate": 1.5376473557700615e-05, | |
| "loss": 0.4087, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.0890688259109313, | |
| "grad_norm": 0.29719163008295457, | |
| "learning_rate": 1.5321469743447893e-05, | |
| "loss": 0.4053, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.0931174089068825, | |
| "grad_norm": 0.29980133514479346, | |
| "learning_rate": 1.5266503338598434e-05, | |
| "loss": 0.4252, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.097165991902834, | |
| "grad_norm": 0.27574489721634615, | |
| "learning_rate": 1.5211574782662149e-05, | |
| "loss": 0.3995, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1012145748987856, | |
| "grad_norm": 0.2696360483535082, | |
| "learning_rate": 1.5156684514846307e-05, | |
| "loss": 0.4146, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.2685568551092582, | |
| "learning_rate": 1.5101832974052012e-05, | |
| "loss": 0.4072, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.1093117408906883, | |
| "grad_norm": 0.26447150555173576, | |
| "learning_rate": 1.5047020598870737e-05, | |
| "loss": 0.418, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.11336032388664, | |
| "grad_norm": 0.28045652733915877, | |
| "learning_rate": 1.4992247827580778e-05, | |
| "loss": 0.4089, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.117408906882591, | |
| "grad_norm": 0.26069958870928883, | |
| "learning_rate": 1.493751509814374e-05, | |
| "loss": 0.4234, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1214574898785425, | |
| "grad_norm": 0.264656765075795, | |
| "learning_rate": 1.4882822848201073e-05, | |
| "loss": 0.4114, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.125506072874494, | |
| "grad_norm": 0.27075822258377286, | |
| "learning_rate": 1.4828171515070553e-05, | |
| "loss": 0.4191, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.1295546558704452, | |
| "grad_norm": 0.27061373657202925, | |
| "learning_rate": 1.4773561535742793e-05, | |
| "loss": 0.4053, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.133603238866397, | |
| "grad_norm": 0.25500096272262707, | |
| "learning_rate": 1.4718993346877715e-05, | |
| "loss": 0.4305, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.1376518218623484, | |
| "grad_norm": 0.2903305470644584, | |
| "learning_rate": 1.466446738480111e-05, | |
| "loss": 0.4222, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1417004048582995, | |
| "grad_norm": 0.2632849434789534, | |
| "learning_rate": 1.460998408550112e-05, | |
| "loss": 0.4133, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.145748987854251, | |
| "grad_norm": 0.29733002817957516, | |
| "learning_rate": 1.4555543884624751e-05, | |
| "loss": 0.3958, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.1497975708502026, | |
| "grad_norm": 0.2805663263618633, | |
| "learning_rate": 1.4501147217474402e-05, | |
| "loss": 0.4191, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.29450460577076765, | |
| "learning_rate": 1.444679451900437e-05, | |
| "loss": 0.4004, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.28977821455582337, | |
| "learning_rate": 1.4392486223817397e-05, | |
| "loss": 0.3971, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.161943319838057, | |
| "grad_norm": 0.26744717755157116, | |
| "learning_rate": 1.4338222766161143e-05, | |
| "loss": 0.396, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.165991902834008, | |
| "grad_norm": 0.281369086174738, | |
| "learning_rate": 1.4284004579924777e-05, | |
| "loss": 0.3978, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.1700404858299596, | |
| "grad_norm": 0.267445994576062, | |
| "learning_rate": 1.4229832098635479e-05, | |
| "loss": 0.4081, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.174089068825911, | |
| "grad_norm": 0.2647875270562826, | |
| "learning_rate": 1.4175705755454963e-05, | |
| "loss": 0.4128, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.1781376518218623, | |
| "grad_norm": 0.25388157064783534, | |
| "learning_rate": 1.412162598317602e-05, | |
| "loss": 0.4208, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.182186234817814, | |
| "grad_norm": 0.25544746702809074, | |
| "learning_rate": 1.406759321421907e-05, | |
| "loss": 0.4212, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.1862348178137654, | |
| "grad_norm": 0.2940816972878137, | |
| "learning_rate": 1.401360788062871e-05, | |
| "loss": 0.4053, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.1902834008097165, | |
| "grad_norm": 0.2668067871554974, | |
| "learning_rate": 1.3959670414070204e-05, | |
| "loss": 0.4352, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.194331983805668, | |
| "grad_norm": 0.2648213659309414, | |
| "learning_rate": 1.3905781245826108e-05, | |
| "loss": 0.4154, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.1983805668016196, | |
| "grad_norm": 0.24922911016366578, | |
| "learning_rate": 1.3851940806792778e-05, | |
| "loss": 0.4232, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2024291497975708, | |
| "grad_norm": 0.27610905671343416, | |
| "learning_rate": 1.379814952747693e-05, | |
| "loss": 0.4092, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.2064777327935223, | |
| "grad_norm": 0.2568990215163064, | |
| "learning_rate": 1.3744407837992193e-05, | |
| "loss": 0.4227, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.28843624766529236, | |
| "learning_rate": 1.3690716168055692e-05, | |
| "loss": 0.416, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.214574898785425, | |
| "grad_norm": 0.26831424040104374, | |
| "learning_rate": 1.363707494698459e-05, | |
| "loss": 0.4121, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.2186234817813766, | |
| "grad_norm": 0.29487016929880117, | |
| "learning_rate": 1.358348460369265e-05, | |
| "loss": 0.4195, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.2226720647773277, | |
| "grad_norm": 0.2642309882179861, | |
| "learning_rate": 1.352994556668684e-05, | |
| "loss": 0.4112, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.2267206477732793, | |
| "grad_norm": 0.28899482545821237, | |
| "learning_rate": 1.3476458264063865e-05, | |
| "loss": 0.4013, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.30987831431128154, | |
| "learning_rate": 1.342302312350678e-05, | |
| "loss": 0.4312, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.234817813765182, | |
| "grad_norm": 0.2842887164034997, | |
| "learning_rate": 1.3369640572281537e-05, | |
| "loss": 0.4101, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.2388663967611335, | |
| "grad_norm": 0.26181088343767206, | |
| "learning_rate": 1.3316311037233596e-05, | |
| "loss": 0.4051, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.242914979757085, | |
| "grad_norm": 0.34642799611892705, | |
| "learning_rate": 1.326303494478451e-05, | |
| "loss": 0.4028, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.246963562753036, | |
| "grad_norm": 0.28310275220217634, | |
| "learning_rate": 1.3209812720928472e-05, | |
| "loss": 0.4249, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.251012145748988, | |
| "grad_norm": 0.2865157137718032, | |
| "learning_rate": 1.315664479122898e-05, | |
| "loss": 0.4094, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.2550607287449393, | |
| "grad_norm": 0.344867689695805, | |
| "learning_rate": 1.3103531580815378e-05, | |
| "loss": 0.429, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.2591093117408905, | |
| "grad_norm": 0.25760108226887757, | |
| "learning_rate": 1.3050473514379488e-05, | |
| "loss": 0.4033, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.3503551904220396, | |
| "learning_rate": 1.2997471016172188e-05, | |
| "loss": 0.4191, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.2672064777327936, | |
| "grad_norm": 0.25802217039036696, | |
| "learning_rate": 1.2944524510000042e-05, | |
| "loss": 0.4142, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.2712550607287447, | |
| "grad_norm": 0.31158482647278163, | |
| "learning_rate": 1.2891634419221908e-05, | |
| "loss": 0.4026, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.2753036437246963, | |
| "grad_norm": 0.27081720820174265, | |
| "learning_rate": 1.2838801166745545e-05, | |
| "loss": 0.4209, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.279352226720648, | |
| "grad_norm": 0.2959039614844649, | |
| "learning_rate": 1.2786025175024216e-05, | |
| "loss": 0.4175, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.283400809716599, | |
| "grad_norm": 0.33308105598459326, | |
| "learning_rate": 1.2733306866053357e-05, | |
| "loss": 0.4233, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.2874493927125505, | |
| "grad_norm": 0.2553297997765034, | |
| "learning_rate": 1.2680646661367163e-05, | |
| "loss": 0.4275, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.291497975708502, | |
| "grad_norm": 0.3293111252907255, | |
| "learning_rate": 1.2628044982035219e-05, | |
| "loss": 0.407, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.2955465587044532, | |
| "grad_norm": 0.25881650134222833, | |
| "learning_rate": 1.2575502248659155e-05, | |
| "loss": 0.412, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.299595141700405, | |
| "grad_norm": 0.2713407268541946, | |
| "learning_rate": 1.2523018881369269e-05, | |
| "loss": 0.4072, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.3036437246963564, | |
| "grad_norm": 0.25893410722848864, | |
| "learning_rate": 1.2470595299821176e-05, | |
| "loss": 0.4271, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.2639107679731469, | |
| "learning_rate": 1.2418231923192416e-05, | |
| "loss": 0.4159, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.311740890688259, | |
| "grad_norm": 0.27827928273340824, | |
| "learning_rate": 1.2365929170179165e-05, | |
| "loss": 0.4036, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.26360637672622933, | |
| "learning_rate": 1.2313687458992848e-05, | |
| "loss": 0.4179, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.3198380566801617, | |
| "grad_norm": 0.26201312831151846, | |
| "learning_rate": 1.2261507207356785e-05, | |
| "loss": 0.4252, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.3238866396761133, | |
| "grad_norm": 0.2521691998217544, | |
| "learning_rate": 1.220938883250288e-05, | |
| "loss": 0.4121, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.327935222672065, | |
| "grad_norm": 0.24006074055354865, | |
| "learning_rate": 1.215733275116827e-05, | |
| "loss": 0.4241, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.331983805668016, | |
| "grad_norm": 0.24820487655555196, | |
| "learning_rate": 1.2105339379592005e-05, | |
| "loss": 0.4296, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.3360323886639676, | |
| "grad_norm": 0.23691468224463333, | |
| "learning_rate": 1.2053409133511681e-05, | |
| "loss": 0.4136, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.340080971659919, | |
| "grad_norm": 0.2538809750852651, | |
| "learning_rate": 1.2001542428160173e-05, | |
| "loss": 0.4095, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.3441295546558703, | |
| "grad_norm": 0.24663240021686594, | |
| "learning_rate": 1.1949739678262278e-05, | |
| "loss": 0.4244, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.348178137651822, | |
| "grad_norm": 0.24015455155454962, | |
| "learning_rate": 1.189800129803139e-05, | |
| "loss": 0.4104, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.3522267206477734, | |
| "grad_norm": 0.24366115941948588, | |
| "learning_rate": 1.1846327701166228e-05, | |
| "loss": 0.4232, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.3562753036437245, | |
| "grad_norm": 0.2264750501136474, | |
| "learning_rate": 1.1794719300847493e-05, | |
| "loss": 0.4087, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.360323886639676, | |
| "grad_norm": 0.2557012361045378, | |
| "learning_rate": 1.1743176509734582e-05, | |
| "loss": 0.4185, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.3643724696356276, | |
| "grad_norm": 0.24748786933261685, | |
| "learning_rate": 1.1691699739962275e-05, | |
| "loss": 0.4043, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.2487020187957994, | |
| "learning_rate": 1.1640289403137456e-05, | |
| "loss": 0.4175, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.3724696356275303, | |
| "grad_norm": 0.250100360626892, | |
| "learning_rate": 1.1588945910335802e-05, | |
| "loss": 0.41, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.376518218623482, | |
| "grad_norm": 0.23970538649514372, | |
| "learning_rate": 1.1537669672098517e-05, | |
| "loss": 0.4128, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.380566801619433, | |
| "grad_norm": 0.2472068497572778, | |
| "learning_rate": 1.1486461098429036e-05, | |
| "loss": 0.427, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.23966279291130876, | |
| "learning_rate": 1.1435320598789747e-05, | |
| "loss": 0.4082, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.388663967611336, | |
| "grad_norm": 0.23895097775919735, | |
| "learning_rate": 1.1384248582098736e-05, | |
| "loss": 0.413, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.3927125506072873, | |
| "grad_norm": 0.24211605944152123, | |
| "learning_rate": 1.1333245456726467e-05, | |
| "loss": 0.3934, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.396761133603239, | |
| "grad_norm": 0.22966737927866532, | |
| "learning_rate": 1.128231163049258e-05, | |
| "loss": 0.4121, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.4008097165991904, | |
| "grad_norm": 0.22295593147484424, | |
| "learning_rate": 1.123144751066259e-05, | |
| "loss": 0.4203, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4048582995951415, | |
| "grad_norm": 0.23313928924213537, | |
| "learning_rate": 1.1180653503944657e-05, | |
| "loss": 0.4127, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.408906882591093, | |
| "grad_norm": 0.23350414615986287, | |
| "learning_rate": 1.1129930016486287e-05, | |
| "loss": 0.415, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.4129554655870447, | |
| "grad_norm": 0.23143715451971905, | |
| "learning_rate": 1.107927745387114e-05, | |
| "loss": 0.4045, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.417004048582996, | |
| "grad_norm": 0.23908285484410194, | |
| "learning_rate": 1.1028696221115766e-05, | |
| "loss": 0.418, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.24684163658320055, | |
| "learning_rate": 1.0978186722666341e-05, | |
| "loss": 0.4263, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.425101214574899, | |
| "grad_norm": 0.2589960352604857, | |
| "learning_rate": 1.0927749362395475e-05, | |
| "loss": 0.4201, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.42914979757085, | |
| "grad_norm": 0.2167972873818071, | |
| "learning_rate": 1.0877384543598954e-05, | |
| "loss": 0.4104, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.4331983805668016, | |
| "grad_norm": 0.21975797722112025, | |
| "learning_rate": 1.0827092668992529e-05, | |
| "loss": 0.4066, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.437246963562753, | |
| "grad_norm": 0.23662623954008496, | |
| "learning_rate": 1.077687414070869e-05, | |
| "loss": 0.4231, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.4412955465587043, | |
| "grad_norm": 0.23345867517565463, | |
| "learning_rate": 1.0726729360293451e-05, | |
| "loss": 0.4028, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.445344129554656, | |
| "grad_norm": 0.22800237580705676, | |
| "learning_rate": 1.0676658728703147e-05, | |
| "loss": 0.428, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.4493927125506074, | |
| "grad_norm": 0.24486492378256522, | |
| "learning_rate": 1.0626662646301198e-05, | |
| "loss": 0.4039, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.4534412955465585, | |
| "grad_norm": 0.23905252083715273, | |
| "learning_rate": 1.0576741512854958e-05, | |
| "loss": 0.409, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.45748987854251, | |
| "grad_norm": 0.22856195862689221, | |
| "learning_rate": 1.052689572753247e-05, | |
| "loss": 0.3979, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.25606221108105953, | |
| "learning_rate": 1.0477125688899328e-05, | |
| "loss": 0.4155, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.465587044534413, | |
| "grad_norm": 0.23977156311678005, | |
| "learning_rate": 1.0427431794915406e-05, | |
| "loss": 0.4253, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.4696356275303644, | |
| "grad_norm": 0.22479510681217435, | |
| "learning_rate": 1.0377814442931769e-05, | |
| "loss": 0.4121, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.23221095154029686, | |
| "learning_rate": 1.0328274029687439e-05, | |
| "loss": 0.4043, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.477732793522267, | |
| "grad_norm": 0.235249448702866, | |
| "learning_rate": 1.0278810951306253e-05, | |
| "loss": 0.4171, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.4817813765182186, | |
| "grad_norm": 0.23754207118494297, | |
| "learning_rate": 1.0229425603293647e-05, | |
| "loss": 0.4077, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.48582995951417, | |
| "grad_norm": 0.25767839091767347, | |
| "learning_rate": 1.0180118380533558e-05, | |
| "loss": 0.4206, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.4898785425101213, | |
| "grad_norm": 0.23860877160656285, | |
| "learning_rate": 1.0130889677285225e-05, | |
| "loss": 0.4177, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.493927125506073, | |
| "grad_norm": 0.25080563984697624, | |
| "learning_rate": 1.008173988718005e-05, | |
| "loss": 0.4075, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.4979757085020244, | |
| "grad_norm": 0.23788210929675413, | |
| "learning_rate": 1.0032669403218439e-05, | |
| "loss": 0.4207, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.5020242914979756, | |
| "grad_norm": 0.23659118360726303, | |
| "learning_rate": 9.983678617766677e-06, | |
| "loss": 0.4144, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.506072874493927, | |
| "grad_norm": 0.2593351008905294, | |
| "learning_rate": 9.934767922553788e-06, | |
| "loss": 0.4103, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.5101214574898787, | |
| "grad_norm": 0.22050467980354266, | |
| "learning_rate": 9.885937708668371e-06, | |
| "loss": 0.4091, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.51417004048583, | |
| "grad_norm": 0.25067710115802244, | |
| "learning_rate": 9.837188366555524e-06, | |
| "loss": 0.4111, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.5182186234817814, | |
| "grad_norm": 0.24671391308068305, | |
| "learning_rate": 9.788520286013691e-06, | |
| "loss": 0.4253, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.522267206477733, | |
| "grad_norm": 0.23388582068585273, | |
| "learning_rate": 9.739933856191537e-06, | |
| "loss": 0.4156, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.263703792587332, | |
| "learning_rate": 9.691429465584866e-06, | |
| "loss": 0.4082, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.5303643724696356, | |
| "grad_norm": 0.2463177560328071, | |
| "learning_rate": 9.643007502033492e-06, | |
| "loss": 0.406, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.534412955465587, | |
| "grad_norm": 0.25508834043614, | |
| "learning_rate": 9.594668352718152e-06, | |
| "loss": 0.3971, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.24238423461309214, | |
| "learning_rate": 9.546412404157383e-06, | |
| "loss": 0.41, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.54251012145749, | |
| "grad_norm": 0.2429718786708057, | |
| "learning_rate": 9.498240042204471e-06, | |
| "loss": 0.403, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.5465587044534415, | |
| "grad_norm": 0.2531070903512779, | |
| "learning_rate": 9.450151652044339e-06, | |
| "loss": 0.4059, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.5506072874493926, | |
| "grad_norm": 0.23822745593533745, | |
| "learning_rate": 9.402147618190472e-06, | |
| "loss": 0.4227, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.554655870445344, | |
| "grad_norm": 0.24876962614353584, | |
| "learning_rate": 9.354228324481841e-06, | |
| "loss": 0.4116, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.5587044534412957, | |
| "grad_norm": 0.27785092827182134, | |
| "learning_rate": 9.306394154079852e-06, | |
| "loss": 0.4132, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.562753036437247, | |
| "grad_norm": 0.24132230422532822, | |
| "learning_rate": 9.258645489465254e-06, | |
| "loss": 0.4269, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.5668016194331984, | |
| "grad_norm": 0.29327070030388, | |
| "learning_rate": 9.21098271243509e-06, | |
| "loss": 0.4096, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.57085020242915, | |
| "grad_norm": 0.23547427228070564, | |
| "learning_rate": 9.163406204099659e-06, | |
| "loss": 0.4033, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.574898785425101, | |
| "grad_norm": 0.2323742515794626, | |
| "learning_rate": 9.115916344879452e-06, | |
| "loss": 0.4211, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.2662105655339641, | |
| "learning_rate": 9.068513514502126e-06, | |
| "loss": 0.4291, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.582995951417004, | |
| "grad_norm": 0.24492582632403403, | |
| "learning_rate": 9.021198091999439e-06, | |
| "loss": 0.4134, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.5870445344129553, | |
| "grad_norm": 0.2514695494110977, | |
| "learning_rate": 8.973970455704251e-06, | |
| "loss": 0.4054, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.591093117408907, | |
| "grad_norm": 0.26913217510072135, | |
| "learning_rate": 8.926830983247495e-06, | |
| "loss": 0.4056, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.5951417004048585, | |
| "grad_norm": 0.23229262075814605, | |
| "learning_rate": 8.879780051555116e-06, | |
| "loss": 0.4215, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.5991902834008096, | |
| "grad_norm": 0.22868736813382462, | |
| "learning_rate": 8.83281803684512e-06, | |
| "loss": 0.4045, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.603238866396761, | |
| "grad_norm": 0.25059700083943015, | |
| "learning_rate": 8.785945314624532e-06, | |
| "loss": 0.3942, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.6072874493927127, | |
| "grad_norm": 0.23305808242452242, | |
| "learning_rate": 8.739162259686385e-06, | |
| "loss": 0.4091, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.611336032388664, | |
| "grad_norm": 0.22516435988952027, | |
| "learning_rate": 8.692469246106734e-06, | |
| "loss": 0.4208, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.22875054047288096, | |
| "learning_rate": 8.645866647241677e-06, | |
| "loss": 0.4038, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.619433198380567, | |
| "grad_norm": 0.221698577632437, | |
| "learning_rate": 8.599354835724348e-06, | |
| "loss": 0.4287, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.623481781376518, | |
| "grad_norm": 0.2251775461374014, | |
| "learning_rate": 8.552934183461958e-06, | |
| "loss": 0.4197, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6275303643724697, | |
| "grad_norm": 0.22383864100095577, | |
| "learning_rate": 8.506605061632805e-06, | |
| "loss": 0.4051, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.22488071557676764, | |
| "learning_rate": 8.460367840683308e-06, | |
| "loss": 0.4123, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.6356275303643724, | |
| "grad_norm": 0.21471393944657227, | |
| "learning_rate": 8.414222890325065e-06, | |
| "loss": 0.401, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.639676113360324, | |
| "grad_norm": 0.22679304741343073, | |
| "learning_rate": 8.368170579531855e-06, | |
| "loss": 0.417, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.6437246963562755, | |
| "grad_norm": 0.22406302045943277, | |
| "learning_rate": 8.322211276536733e-06, | |
| "loss": 0.4214, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6477732793522266, | |
| "grad_norm": 0.22280069397781485, | |
| "learning_rate": 8.27634534882907e-06, | |
| "loss": 0.4238, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.651821862348178, | |
| "grad_norm": 0.20893738457246483, | |
| "learning_rate": 8.230573163151585e-06, | |
| "loss": 0.433, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.6558704453441297, | |
| "grad_norm": 0.22820681136250076, | |
| "learning_rate": 8.184895085497466e-06, | |
| "loss": 0.4096, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.659919028340081, | |
| "grad_norm": 0.2163696954402914, | |
| "learning_rate": 8.139311481107401e-06, | |
| "loss": 0.4136, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.6639676113360324, | |
| "grad_norm": 0.2255463325491583, | |
| "learning_rate": 8.093822714466686e-06, | |
| "loss": 0.4024, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.668016194331984, | |
| "grad_norm": 0.23088731458513842, | |
| "learning_rate": 8.048429149302271e-06, | |
| "loss": 0.4243, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.672064777327935, | |
| "grad_norm": 0.23602967298233632, | |
| "learning_rate": 8.003131148579902e-06, | |
| "loss": 0.4019, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.6761133603238867, | |
| "grad_norm": 0.23080182348792755, | |
| "learning_rate": 7.957929074501182e-06, | |
| "loss": 0.4123, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.6801619433198383, | |
| "grad_norm": 0.2238287048074134, | |
| "learning_rate": 7.912823288500695e-06, | |
| "loss": 0.4237, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.22735751146869326, | |
| "learning_rate": 7.867814151243102e-06, | |
| "loss": 0.4216, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.688259109311741, | |
| "grad_norm": 0.23817234939538678, | |
| "learning_rate": 7.822902022620267e-06, | |
| "loss": 0.4111, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.2223413257518064, | |
| "learning_rate": 7.778087261748377e-06, | |
| "loss": 0.4146, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.6963562753036436, | |
| "grad_norm": 0.2232572560853659, | |
| "learning_rate": 7.733370226965056e-06, | |
| "loss": 0.4196, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.700404858299595, | |
| "grad_norm": 0.2326495266257994, | |
| "learning_rate": 7.688751275826527e-06, | |
| "loss": 0.4302, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.7044534412955468, | |
| "grad_norm": 0.2285678690041097, | |
| "learning_rate": 7.644230765104735e-06, | |
| "loss": 0.4033, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.708502024291498, | |
| "grad_norm": 0.21533182524721395, | |
| "learning_rate": 7.599809050784502e-06, | |
| "loss": 0.4025, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.7125506072874495, | |
| "grad_norm": 0.2221603887032431, | |
| "learning_rate": 7.555486488060661e-06, | |
| "loss": 0.419, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.716599190283401, | |
| "grad_norm": 0.25144793812506355, | |
| "learning_rate": 7.511263431335249e-06, | |
| "loss": 0.4189, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.720647773279352, | |
| "grad_norm": 0.2179050852865918, | |
| "learning_rate": 7.467140234214656e-06, | |
| "loss": 0.419, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.7246963562753037, | |
| "grad_norm": 0.23661090497796883, | |
| "learning_rate": 7.423117249506777e-06, | |
| "loss": 0.4186, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7287449392712553, | |
| "grad_norm": 0.2400443842514219, | |
| "learning_rate": 7.379194829218231e-06, | |
| "loss": 0.4207, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.7327935222672064, | |
| "grad_norm": 0.23195031345783584, | |
| "learning_rate": 7.335373324551518e-06, | |
| "loss": 0.419, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.2339913524596655, | |
| "learning_rate": 7.291653085902224e-06, | |
| "loss": 0.4158, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.7408906882591095, | |
| "grad_norm": 0.22656385611961596, | |
| "learning_rate": 7.248034462856213e-06, | |
| "loss": 0.4147, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.7449392712550607, | |
| "grad_norm": 0.2176000074216103, | |
| "learning_rate": 7.20451780418683e-06, | |
| "loss": 0.4049, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.748987854251012, | |
| "grad_norm": 0.23243900801302583, | |
| "learning_rate": 7.161103457852128e-06, | |
| "loss": 0.421, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.753036437246964, | |
| "grad_norm": 0.22835635707465324, | |
| "learning_rate": 7.117791770992046e-06, | |
| "loss": 0.4208, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.757085020242915, | |
| "grad_norm": 0.24457916712739644, | |
| "learning_rate": 7.074583089925688e-06, | |
| "loss": 0.4217, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.7611336032388665, | |
| "grad_norm": 0.2232399696078186, | |
| "learning_rate": 7.0314777601485155e-06, | |
| "loss": 0.4271, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.765182186234818, | |
| "grad_norm": 0.2272152351865259, | |
| "learning_rate": 6.988476126329604e-06, | |
| "loss": 0.4155, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.22872501381479668, | |
| "learning_rate": 6.945578532308857e-06, | |
| "loss": 0.407, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.7732793522267207, | |
| "grad_norm": 0.21679147242476574, | |
| "learning_rate": 6.902785321094301e-06, | |
| "loss": 0.4218, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.7773279352226723, | |
| "grad_norm": 0.23944741902795927, | |
| "learning_rate": 6.8600968348593115e-06, | |
| "loss": 0.4016, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.7813765182186234, | |
| "grad_norm": 0.23521064691087173, | |
| "learning_rate": 6.817513414939892e-06, | |
| "loss": 0.4041, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.785425101214575, | |
| "grad_norm": 0.232005445338959, | |
| "learning_rate": 6.775035401831924e-06, | |
| "loss": 0.4022, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.24484962577874878, | |
| "learning_rate": 6.732663135188473e-06, | |
| "loss": 0.4034, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.7935222672064777, | |
| "grad_norm": 0.22927974900294817, | |
| "learning_rate": 6.690396953817055e-06, | |
| "loss": 0.4223, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.7975708502024292, | |
| "grad_norm": 0.24890343028608244, | |
| "learning_rate": 6.648237195676927e-06, | |
| "loss": 0.4004, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.801619433198381, | |
| "grad_norm": 0.22107025648143833, | |
| "learning_rate": 6.606184197876395e-06, | |
| "loss": 0.4338, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.805668016194332, | |
| "grad_norm": 0.21334529432493124, | |
| "learning_rate": 6.564238296670109e-06, | |
| "loss": 0.4315, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.8097165991902835, | |
| "grad_norm": 0.22653474483760117, | |
| "learning_rate": 6.52239982745638e-06, | |
| "loss": 0.4049, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.813765182186235, | |
| "grad_norm": 0.22115581471905885, | |
| "learning_rate": 6.480669124774484e-06, | |
| "loss": 0.4169, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.817813765182186, | |
| "grad_norm": 0.22239042271762677, | |
| "learning_rate": 6.439046522302013e-06, | |
| "loss": 0.4049, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.8218623481781377, | |
| "grad_norm": 0.2240046213754753, | |
| "learning_rate": 6.397532352852191e-06, | |
| "loss": 0.4135, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.8259109311740893, | |
| "grad_norm": 0.2177395565359218, | |
| "learning_rate": 6.356126948371202e-06, | |
| "loss": 0.4116, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.8299595141700404, | |
| "grad_norm": 0.2223251577791594, | |
| "learning_rate": 6.314830639935559e-06, | |
| "loss": 0.3995, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.834008097165992, | |
| "grad_norm": 0.22817470931353886, | |
| "learning_rate": 6.273643757749448e-06, | |
| "loss": 0.4108, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.8380566801619436, | |
| "grad_norm": 0.22413940625863973, | |
| "learning_rate": 6.2325666311420856e-06, | |
| "loss": 0.4176, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.22265578193483632, | |
| "learning_rate": 6.1915995885650675e-06, | |
| "loss": 0.4096, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.22854114442421095, | |
| "learning_rate": 6.150742957589784e-06, | |
| "loss": 0.404, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.850202429149798, | |
| "grad_norm": 0.23835589065551038, | |
| "learning_rate": 6.109997064904774e-06, | |
| "loss": 0.4065, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.854251012145749, | |
| "grad_norm": 0.2307426293186405, | |
| "learning_rate": 6.069362236313092e-06, | |
| "loss": 0.4119, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.8582995951417005, | |
| "grad_norm": 0.2196032044066672, | |
| "learning_rate": 6.0288387967297565e-06, | |
| "loss": 0.418, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.862348178137652, | |
| "grad_norm": 0.24104926018198716, | |
| "learning_rate": 5.988427070179108e-06, | |
| "loss": 0.4171, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.866396761133603, | |
| "grad_norm": 0.22356127575692764, | |
| "learning_rate": 5.948127379792234e-06, | |
| "loss": 0.4173, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.8704453441295548, | |
| "grad_norm": 0.21817576145811945, | |
| "learning_rate": 5.90794004780439e-06, | |
| "loss": 0.416, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.8744939271255063, | |
| "grad_norm": 0.22308250686030393, | |
| "learning_rate": 5.867865395552405e-06, | |
| "loss": 0.4065, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.8785425101214575, | |
| "grad_norm": 0.2156489046187672, | |
| "learning_rate": 5.8279037434721364e-06, | |
| "loss": 0.4205, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.882591093117409, | |
| "grad_norm": 0.22792113857742813, | |
| "learning_rate": 5.788055411095892e-06, | |
| "loss": 0.4078, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.8866396761133606, | |
| "grad_norm": 0.2268558645405396, | |
| "learning_rate": 5.7483207170498625e-06, | |
| "loss": 0.4058, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8906882591093117, | |
| "grad_norm": 0.2302601783993815, | |
| "learning_rate": 5.708699979051613e-06, | |
| "loss": 0.4173, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.2185822982321844, | |
| "learning_rate": 5.66919351390751e-06, | |
| "loss": 0.417, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.898785425101215, | |
| "grad_norm": 0.22448332540503804, | |
| "learning_rate": 5.629801637510188e-06, | |
| "loss": 0.4235, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.902834008097166, | |
| "grad_norm": 0.21547821610858342, | |
| "learning_rate": 5.590524664836043e-06, | |
| "loss": 0.4029, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.9068825910931175, | |
| "grad_norm": 0.21352587061839362, | |
| "learning_rate": 5.551362909942708e-06, | |
| "loss": 0.433, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.910931174089069, | |
| "grad_norm": 0.20862492438051605, | |
| "learning_rate": 5.51231668596653e-06, | |
| "loss": 0.415, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.91497975708502, | |
| "grad_norm": 0.2183393969471234, | |
| "learning_rate": 5.47338630512007e-06, | |
| "loss": 0.4135, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.919028340080972, | |
| "grad_norm": 0.21646507577466567, | |
| "learning_rate": 5.434572078689617e-06, | |
| "loss": 0.4197, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.21553801461452862, | |
| "learning_rate": 5.395874317032694e-06, | |
| "loss": 0.4155, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.9271255060728745, | |
| "grad_norm": 0.22477490722234772, | |
| "learning_rate": 5.357293329575573e-06, | |
| "loss": 0.4018, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.931174089068826, | |
| "grad_norm": 0.21708280491201826, | |
| "learning_rate": 5.318829424810797e-06, | |
| "loss": 0.4097, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.9352226720647776, | |
| "grad_norm": 0.22425604673605462, | |
| "learning_rate": 5.280482910294728e-06, | |
| "loss": 0.4148, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.9392712550607287, | |
| "grad_norm": 0.21590540930735513, | |
| "learning_rate": 5.242254092645078e-06, | |
| "loss": 0.4143, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.9433198380566803, | |
| "grad_norm": 0.2070426903948516, | |
| "learning_rate": 5.2041432775384445e-06, | |
| "loss": 0.4101, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.21811766415448117, | |
| "learning_rate": 5.166150769707892e-06, | |
| "loss": 0.4189, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.951417004048583, | |
| "grad_norm": 0.21114653116233437, | |
| "learning_rate": 5.128276872940503e-06, | |
| "loss": 0.4182, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.9554655870445345, | |
| "grad_norm": 0.2216259829197288, | |
| "learning_rate": 5.090521890074953e-06, | |
| "loss": 0.4113, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.9595141700404857, | |
| "grad_norm": 0.21062230454673436, | |
| "learning_rate": 5.0528861229990675e-06, | |
| "loss": 0.4091, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.9635627530364372, | |
| "grad_norm": 0.21089363414699636, | |
| "learning_rate": 5.015369872647442e-06, | |
| "loss": 0.4197, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.967611336032389, | |
| "grad_norm": 0.21525757800030698, | |
| "learning_rate": 4.9779734389990195e-06, | |
| "loss": 0.41, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97165991902834, | |
| "grad_norm": 0.2185848430568364, | |
| "learning_rate": 4.940697121074675e-06, | |
| "loss": 0.399, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.9757085020242915, | |
| "grad_norm": 0.20336275059267048, | |
| "learning_rate": 4.903541216934864e-06, | |
| "loss": 0.4031, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.979757085020243, | |
| "grad_norm": 0.2129259463857742, | |
| "learning_rate": 4.8665060236772e-06, | |
| "loss": 0.3875, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.983805668016194, | |
| "grad_norm": 0.21118809685379542, | |
| "learning_rate": 4.829591837434102e-06, | |
| "loss": 0.4219, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.9878542510121457, | |
| "grad_norm": 0.21109562018564912, | |
| "learning_rate": 4.792798953370424e-06, | |
| "loss": 0.4133, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.9919028340080973, | |
| "grad_norm": 0.21279586749470228, | |
| "learning_rate": 4.756127665681083e-06, | |
| "loss": 0.4189, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.9959514170040484, | |
| "grad_norm": 0.21870873722256806, | |
| "learning_rate": 4.719578267588725e-06, | |
| "loss": 0.4025, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.2105295667178359, | |
| "learning_rate": 4.6831510513413525e-06, | |
| "loss": 0.4225, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.004048582995951, | |
| "grad_norm": 0.43988420327726196, | |
| "learning_rate": 4.6468463082100176e-06, | |
| "loss": 0.3585, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.008097165991903, | |
| "grad_norm": 0.29709652812321524, | |
| "learning_rate": 4.610664328486485e-06, | |
| "loss": 0.36, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.012145748987854, | |
| "grad_norm": 0.3056825276032711, | |
| "learning_rate": 4.5746054014809e-06, | |
| "loss": 0.3789, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.016194331983805, | |
| "grad_norm": 0.5445630700265527, | |
| "learning_rate": 4.5386698155194766e-06, | |
| "loss": 0.3835, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.020242914979757, | |
| "grad_norm": 0.32301035601923217, | |
| "learning_rate": 4.502857857942204e-06, | |
| "loss": 0.3616, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.0242914979757085, | |
| "grad_norm": 0.34361345487648703, | |
| "learning_rate": 4.467169815100549e-06, | |
| "loss": 0.3736, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.02834008097166, | |
| "grad_norm": 0.4587317690779146, | |
| "learning_rate": 4.431605972355137e-06, | |
| "loss": 0.3592, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.032388663967612, | |
| "grad_norm": 0.30958909055021183, | |
| "learning_rate": 4.396166614073516e-06, | |
| "loss": 0.3721, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.036437246963563, | |
| "grad_norm": 0.24585148455876782, | |
| "learning_rate": 4.360852023627846e-06, | |
| "loss": 0.3713, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.040485829959514, | |
| "grad_norm": 0.3305272921049497, | |
| "learning_rate": 4.325662483392656e-06, | |
| "loss": 0.3652, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.044534412955466, | |
| "grad_norm": 0.34965841246856955, | |
| "learning_rate": 4.29059827474257e-06, | |
| "loss": 0.3645, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.048582995951417, | |
| "grad_norm": 0.28015814095022495, | |
| "learning_rate": 4.255659678050063e-06, | |
| "loss": 0.3529, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.24304337940601628, | |
| "learning_rate": 4.22084697268323e-06, | |
| "loss": 0.3772, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.05668016194332, | |
| "grad_norm": 0.28479117867694753, | |
| "learning_rate": 4.186160437003535e-06, | |
| "loss": 0.3638, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.060728744939271, | |
| "grad_norm": 0.3039619037162753, | |
| "learning_rate": 4.151600348363582e-06, | |
| "loss": 0.3574, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.064777327935222, | |
| "grad_norm": 0.24967658798501077, | |
| "learning_rate": 4.11716698310493e-06, | |
| "loss": 0.3673, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.068825910931174, | |
| "grad_norm": 0.23513246202895532, | |
| "learning_rate": 4.082860616555848e-06, | |
| "loss": 0.3567, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.0728744939271255, | |
| "grad_norm": 0.24555550008898952, | |
| "learning_rate": 4.048681523029125e-06, | |
| "loss": 0.3689, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.27888040989421237, | |
| "learning_rate": 4.014629975819882e-06, | |
| "loss": 0.365, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.080971659919029, | |
| "grad_norm": 0.2471919520450784, | |
| "learning_rate": 3.9807062472033855e-06, | |
| "loss": 0.3634, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.08502024291498, | |
| "grad_norm": 0.2386714044585816, | |
| "learning_rate": 3.94691060843287e-06, | |
| "loss": 0.374, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.089068825910931, | |
| "grad_norm": 0.2232488441599435, | |
| "learning_rate": 3.913243329737353e-06, | |
| "loss": 0.3717, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.093117408906883, | |
| "grad_norm": 0.2530858224641521, | |
| "learning_rate": 3.879704680319503e-06, | |
| "loss": 0.3637, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.097165991902834, | |
| "grad_norm": 0.2593117466270855, | |
| "learning_rate": 3.846294928353471e-06, | |
| "loss": 0.3738, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.101214574898785, | |
| "grad_norm": 0.22512843759895304, | |
| "learning_rate": 3.8130143409827345e-06, | |
| "loss": 0.3717, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.2177828673995546, | |
| "learning_rate": 3.7798631843179887e-06, | |
| "loss": 0.3576, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.109311740890688, | |
| "grad_norm": 0.24748886126883773, | |
| "learning_rate": 3.7468417234349973e-06, | |
| "loss": 0.3648, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.113360323886639, | |
| "grad_norm": 0.257709341453407, | |
| "learning_rate": 3.7139502223724823e-06, | |
| "loss": 0.3645, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.117408906882591, | |
| "grad_norm": 0.23880413976433265, | |
| "learning_rate": 3.6811889441300095e-06, | |
| "loss": 0.3895, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.1214574898785425, | |
| "grad_norm": 0.22249462198980202, | |
| "learning_rate": 3.648558150665886e-06, | |
| "loss": 0.3595, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.125506072874494, | |
| "grad_norm": 0.2346908402703715, | |
| "learning_rate": 3.616058102895075e-06, | |
| "loss": 0.3711, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.129554655870446, | |
| "grad_norm": 0.2421664695281808, | |
| "learning_rate": 3.583689060687079e-06, | |
| "loss": 0.3824, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.133603238866397, | |
| "grad_norm": 0.2192153899289673, | |
| "learning_rate": 3.5514512828639023e-06, | |
| "loss": 0.3747, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.137651821862348, | |
| "grad_norm": 0.22641429838504762, | |
| "learning_rate": 3.51934502719796e-06, | |
| "loss": 0.3673, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.1417004048583, | |
| "grad_norm": 0.24609701604085082, | |
| "learning_rate": 3.4873705504100185e-06, | |
| "loss": 0.3516, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.145748987854251, | |
| "grad_norm": 0.22583183824294592, | |
| "learning_rate": 3.4555281081671345e-06, | |
| "loss": 0.3601, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.149797570850202, | |
| "grad_norm": 0.22592778503318248, | |
| "learning_rate": 3.4238179550806326e-06, | |
| "loss": 0.3683, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.2131674286987314, | |
| "learning_rate": 3.3922403447040497e-06, | |
| "loss": 0.3611, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.2348280011650784, | |
| "learning_rate": 3.3607955295311247e-06, | |
| "loss": 0.3655, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.161943319838056, | |
| "grad_norm": 0.2121225962179995, | |
| "learning_rate": 3.32948376099375e-06, | |
| "loss": 0.3685, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.165991902834008, | |
| "grad_norm": 0.2203192364839038, | |
| "learning_rate": 3.2983052894599976e-06, | |
| "loss": 0.3557, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.17004048582996, | |
| "grad_norm": 0.2101284248919076, | |
| "learning_rate": 3.2672603642320945e-06, | |
| "loss": 0.3646, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.174089068825911, | |
| "grad_norm": 0.22438785142043235, | |
| "learning_rate": 3.236349233544438e-06, | |
| "loss": 0.3671, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.178137651821863, | |
| "grad_norm": 0.22419016367775726, | |
| "learning_rate": 3.205572144561604e-06, | |
| "loss": 0.3655, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.182186234817814, | |
| "grad_norm": 0.21793371380252557, | |
| "learning_rate": 3.174929343376374e-06, | |
| "loss": 0.3595, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.186234817813765, | |
| "grad_norm": 0.21741282672878431, | |
| "learning_rate": 3.1444210750077777e-06, | |
| "loss": 0.3745, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.190283400809717, | |
| "grad_norm": 0.22515646266291253, | |
| "learning_rate": 3.1140475833991046e-06, | |
| "loss": 0.3727, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.194331983805668, | |
| "grad_norm": 0.2228792408348791, | |
| "learning_rate": 3.0838091114159895e-06, | |
| "loss": 0.3749, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.198380566801619, | |
| "grad_norm": 0.22111547801342404, | |
| "learning_rate": 3.0537059008444524e-06, | |
| "loss": 0.3565, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.202429149797571, | |
| "grad_norm": 0.22814838092484999, | |
| "learning_rate": 3.0237381923889563e-06, | |
| "loss": 0.37, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.206477732793522, | |
| "grad_norm": 0.2108329447462992, | |
| "learning_rate": 2.9939062256704997e-06, | |
| "loss": 0.3731, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.22512911002311467, | |
| "learning_rate": 2.964210239224696e-06, | |
| "loss": 0.364, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.2145748987854255, | |
| "grad_norm": 0.2068274783735355, | |
| "learning_rate": 2.934650470499869e-06, | |
| "loss": 0.3608, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.218623481781377, | |
| "grad_norm": 0.23106458558576565, | |
| "learning_rate": 2.905227155855128e-06, | |
| "loss": 0.3721, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.222672064777328, | |
| "grad_norm": 0.2175464654546572, | |
| "learning_rate": 2.875940530558519e-06, | |
| "loss": 0.3741, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.22672064777328, | |
| "grad_norm": 0.21990131427375362, | |
| "learning_rate": 2.8467908287851154e-06, | |
| "loss": 0.3547, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.21724095498302204, | |
| "learning_rate": 2.8177782836151603e-06, | |
| "loss": 0.3813, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.234817813765182, | |
| "grad_norm": 0.23612057239759127, | |
| "learning_rate": 2.7889031270321874e-06, | |
| "loss": 0.3677, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.238866396761134, | |
| "grad_norm": 0.2102128696241325, | |
| "learning_rate": 2.760165589921184e-06, | |
| "loss": 0.3724, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.242914979757085, | |
| "grad_norm": 0.20942800932685685, | |
| "learning_rate": 2.7315659020667305e-06, | |
| "loss": 0.3781, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.246963562753036, | |
| "grad_norm": 0.21500927689948812, | |
| "learning_rate": 2.703104292151164e-06, | |
| "loss": 0.3699, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.251012145748988, | |
| "grad_norm": 0.20954052405926582, | |
| "learning_rate": 2.6747809877527607e-06, | |
| "loss": 0.3616, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.255060728744939, | |
| "grad_norm": 0.20530364560158554, | |
| "learning_rate": 2.6465962153439082e-06, | |
| "loss": 0.3704, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.2591093117408905, | |
| "grad_norm": 0.20905347906793081, | |
| "learning_rate": 2.618550200289303e-06, | |
| "loss": 0.3788, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.36771143769165665, | |
| "learning_rate": 2.59064316684412e-06, | |
| "loss": 0.3649, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.267206477732794, | |
| "grad_norm": 0.2211940074973393, | |
| "learning_rate": 2.562875338152264e-06, | |
| "loss": 0.3577, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.271255060728745, | |
| "grad_norm": 0.21366482669932402, | |
| "learning_rate": 2.5352469362445552e-06, | |
| "loss": 0.3736, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.275303643724697, | |
| "grad_norm": 0.21504319197363841, | |
| "learning_rate": 2.5077581820369523e-06, | |
| "loss": 0.3554, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.279352226720648, | |
| "grad_norm": 0.21351235137163188, | |
| "learning_rate": 2.480409295328803e-06, | |
| "loss": 0.3546, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.283400809716599, | |
| "grad_norm": 0.21765519826165877, | |
| "learning_rate": 2.45320049480108e-06, | |
| "loss": 0.3684, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.287449392712551, | |
| "grad_norm": 0.2103330104607481, | |
| "learning_rate": 2.4261319980146293e-06, | |
| "loss": 0.369, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.291497975708502, | |
| "grad_norm": 0.21254406378380516, | |
| "learning_rate": 2.399204021408421e-06, | |
| "loss": 0.3772, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.295546558704453, | |
| "grad_norm": 0.21306594276144258, | |
| "learning_rate": 2.3724167802978395e-06, | |
| "loss": 0.3559, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.299595141700405, | |
| "grad_norm": 0.21207755479355514, | |
| "learning_rate": 2.3457704888729625e-06, | |
| "loss": 0.3739, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.303643724696356, | |
| "grad_norm": 0.2038793993963804, | |
| "learning_rate": 2.3192653601968162e-06, | |
| "loss": 0.3692, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.2227526198757001, | |
| "learning_rate": 2.2929016062037057e-06, | |
| "loss": 0.3447, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.3117408906882595, | |
| "grad_norm": 0.21559353621087868, | |
| "learning_rate": 2.2666794376975055e-06, | |
| "loss": 0.3615, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.2168730905186857, | |
| "learning_rate": 2.24059906434998e-06, | |
| "loss": 0.3639, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.319838056680162, | |
| "grad_norm": 0.2120694884150577, | |
| "learning_rate": 2.214660694699091e-06, | |
| "loss": 0.3568, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.323886639676114, | |
| "grad_norm": 0.20652709346887033, | |
| "learning_rate": 2.188864536147355e-06, | |
| "loss": 0.3709, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.327935222672065, | |
| "grad_norm": 0.2024511471750982, | |
| "learning_rate": 2.1632107949601757e-06, | |
| "loss": 0.3758, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.331983805668016, | |
| "grad_norm": 0.20200429596876862, | |
| "learning_rate": 2.137699676264171e-06, | |
| "loss": 0.3588, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.336032388663968, | |
| "grad_norm": 0.2034990815178409, | |
| "learning_rate": 2.1123313840455738e-06, | |
| "loss": 0.3781, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.340080971659919, | |
| "grad_norm": 0.20349496286313154, | |
| "learning_rate": 2.08710612114857e-06, | |
| "loss": 0.374, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.34412955465587, | |
| "grad_norm": 0.20263687980135467, | |
| "learning_rate": 2.0620240892736953e-06, | |
| "loss": 0.3789, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.348178137651822, | |
| "grad_norm": 0.2039935768492433, | |
| "learning_rate": 2.0370854889762e-06, | |
| "loss": 0.3696, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.352226720647773, | |
| "grad_norm": 0.2133228457128409, | |
| "learning_rate": 2.0122905196644726e-06, | |
| "loss": 0.3535, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3562753036437245, | |
| "grad_norm": 0.20060880274881626, | |
| "learning_rate": 1.987639379598427e-06, | |
| "loss": 0.3644, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.3603238866396765, | |
| "grad_norm": 0.1990953778078439, | |
| "learning_rate": 1.963132265887919e-06, | |
| "loss": 0.3686, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.364372469635628, | |
| "grad_norm": 0.2088174610139731, | |
| "learning_rate": 1.9387693744911784e-06, | |
| "loss": 0.3692, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.21865699299132155, | |
| "learning_rate": 1.9145509002132368e-06, | |
| "loss": 0.375, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.372469635627531, | |
| "grad_norm": 0.20525293863363717, | |
| "learning_rate": 1.8904770367043723e-06, | |
| "loss": 0.3581, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.376518218623482, | |
| "grad_norm": 0.2077243892759442, | |
| "learning_rate": 1.8665479764585504e-06, | |
| "loss": 0.3749, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.380566801619433, | |
| "grad_norm": 0.1982057069700626, | |
| "learning_rate": 1.8427639108118977e-06, | |
| "loss": 0.3808, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.20818395051264252, | |
| "learning_rate": 1.8191250299411755e-06, | |
| "loss": 0.3541, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.388663967611336, | |
| "grad_norm": 0.22147235250418407, | |
| "learning_rate": 1.7956315228622446e-06, | |
| "loss": 0.3691, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.392712550607287, | |
| "grad_norm": 0.1998016524372679, | |
| "learning_rate": 1.7722835774285553e-06, | |
| "loss": 0.3679, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.396761133603239, | |
| "grad_norm": 0.2100472045311932, | |
| "learning_rate": 1.7490813803296624e-06, | |
| "loss": 0.3698, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.40080971659919, | |
| "grad_norm": 0.20530288552617545, | |
| "learning_rate": 1.726025117089718e-06, | |
| "loss": 0.3773, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.4048582995951415, | |
| "grad_norm": 0.20687755625054843, | |
| "learning_rate": 1.70311497206598e-06, | |
| "loss": 0.3707, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.4089068825910935, | |
| "grad_norm": 0.21193969741667862, | |
| "learning_rate": 1.6803511284473638e-06, | |
| "loss": 0.3546, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.412955465587045, | |
| "grad_norm": 0.20236539523932698, | |
| "learning_rate": 1.6577337682529538e-06, | |
| "loss": 0.3665, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.417004048582996, | |
| "grad_norm": 0.21062571298863786, | |
| "learning_rate": 1.6352630723305617e-06, | |
| "loss": 0.3601, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.20079503030045542, | |
| "learning_rate": 1.612939220355272e-06, | |
| "loss": 0.3651, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.425101214574899, | |
| "grad_norm": 0.21665014664929078, | |
| "learning_rate": 1.5907623908280112e-06, | |
| "loss": 0.3584, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.42914979757085, | |
| "grad_norm": 0.1957170908699381, | |
| "learning_rate": 1.568732761074121e-06, | |
| "loss": 0.3772, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.433198380566802, | |
| "grad_norm": 0.19707789744430154, | |
| "learning_rate": 1.5468505072419237e-06, | |
| "loss": 0.3675, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.437246963562753, | |
| "grad_norm": 0.21554516530653295, | |
| "learning_rate": 1.5251158043013448e-06, | |
| "loss": 0.3642, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.441295546558704, | |
| "grad_norm": 0.2117549442431771, | |
| "learning_rate": 1.5035288260424885e-06, | |
| "loss": 0.3645, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.445344129554655, | |
| "grad_norm": 0.21063325973657257, | |
| "learning_rate": 1.4820897450742623e-06, | |
| "loss": 0.3677, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.449392712550607, | |
| "grad_norm": 0.20728735541520224, | |
| "learning_rate": 1.4607987328229767e-06, | |
| "loss": 0.3686, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.4534412955465585, | |
| "grad_norm": 0.20377512677098236, | |
| "learning_rate": 1.4396559595310056e-06, | |
| "loss": 0.3846, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.4574898785425106, | |
| "grad_norm": 0.21264504287417904, | |
| "learning_rate": 1.4186615942553973e-06, | |
| "loss": 0.3683, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.20513395694441056, | |
| "learning_rate": 1.3978158048665424e-06, | |
| "loss": 0.3649, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.465587044534413, | |
| "grad_norm": 0.21944606230722857, | |
| "learning_rate": 1.3771187580468115e-06, | |
| "loss": 0.3699, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.469635627530364, | |
| "grad_norm": 0.20530597403589154, | |
| "learning_rate": 1.3565706192892392e-06, | |
| "loss": 0.3661, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.20141881922790378, | |
| "learning_rate": 1.3361715528961993e-06, | |
| "loss": 0.3654, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.477732793522267, | |
| "grad_norm": 0.21752707078932787, | |
| "learning_rate": 1.3159217219780817e-06, | |
| "loss": 0.3663, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.481781376518219, | |
| "grad_norm": 0.2138277902316521, | |
| "learning_rate": 1.2958212884519949e-06, | |
| "loss": 0.361, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.48582995951417, | |
| "grad_norm": 0.19776624040359297, | |
| "learning_rate": 1.2758704130404675e-06, | |
| "loss": 0.3777, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.489878542510121, | |
| "grad_norm": 0.20218747548020016, | |
| "learning_rate": 1.2560692552701738e-06, | |
| "loss": 0.3661, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.493927125506072, | |
| "grad_norm": 0.2163389166544215, | |
| "learning_rate": 1.2364179734706382e-06, | |
| "loss": 0.3588, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.497975708502024, | |
| "grad_norm": 0.20962844545043868, | |
| "learning_rate": 1.2169167247729852e-06, | |
| "loss": 0.3845, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.502024291497976, | |
| "grad_norm": 0.20196553867931172, | |
| "learning_rate": 1.1975656651086864e-06, | |
| "loss": 0.3726, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.506072874493928, | |
| "grad_norm": 0.20366181499693203, | |
| "learning_rate": 1.1783649492082926e-06, | |
| "loss": 0.3694, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.510121457489879, | |
| "grad_norm": 0.20695895280769958, | |
| "learning_rate": 1.1593147306002183e-06, | |
| "loss": 0.3569, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.51417004048583, | |
| "grad_norm": 0.21410848472739444, | |
| "learning_rate": 1.140415161609505e-06, | |
| "loss": 0.3727, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.518218623481781, | |
| "grad_norm": 0.22334137738564325, | |
| "learning_rate": 1.121666393356604e-06, | |
| "loss": 0.382, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.522267206477733, | |
| "grad_norm": 0.2071799904833287, | |
| "learning_rate": 1.1030685757561632e-06, | |
| "loss": 0.3541, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.20233565272289328, | |
| "learning_rate": 1.0846218575158373e-06, | |
| "loss": 0.38, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.530364372469636, | |
| "grad_norm": 0.20017132478094582, | |
| "learning_rate": 1.066326386135097e-06, | |
| "loss": 0.368, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.534412955465587, | |
| "grad_norm": 0.20074108564733012, | |
| "learning_rate": 1.0481823079040398e-06, | |
| "loss": 0.3662, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.2055336980557157, | |
| "learning_rate": 1.0301897679022344e-06, | |
| "loss": 0.3645, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.5425101214574894, | |
| "grad_norm": 0.20333624122619015, | |
| "learning_rate": 1.0123489099975491e-06, | |
| "loss": 0.3643, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.5465587044534415, | |
| "grad_norm": 0.20296601360075228, | |
| "learning_rate": 9.9465987684501e-07, | |
| "loss": 0.3646, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.550607287449393, | |
| "grad_norm": 0.21470391397723296, | |
| "learning_rate": 9.771228098856534e-07, | |
| "loss": 0.3556, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.554655870445345, | |
| "grad_norm": 0.1995525501199136, | |
| "learning_rate": 9.597378493454013e-07, | |
| "loss": 0.3651, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.558704453441296, | |
| "grad_norm": 0.21283380580417433, | |
| "learning_rate": 9.425051342339353e-07, | |
| "loss": 0.3773, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.562753036437247, | |
| "grad_norm": 0.2019868408656294, | |
| "learning_rate": 9.254248023435864e-07, | |
| "loss": 0.3729, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.566801619433198, | |
| "grad_norm": 0.20497028986490334, | |
| "learning_rate": 9.084969902482288e-07, | |
| "loss": 0.3703, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.57085020242915, | |
| "grad_norm": 0.20245025234077096, | |
| "learning_rate": 8.917218333021993e-07, | |
| "loss": 0.3756, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.574898785425101, | |
| "grad_norm": 0.19810639506621755, | |
| "learning_rate": 8.750994656392109e-07, | |
| "loss": 0.3763, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.2047817713076563, | |
| "learning_rate": 8.58630020171265e-07, | |
| "loss": 0.3643, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.582995951417004, | |
| "grad_norm": 0.20583437007765112, | |
| "learning_rate": 8.423136285876099e-07, | |
| "loss": 0.3575, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.587044534412955, | |
| "grad_norm": 0.2070756401137745, | |
| "learning_rate": 8.26150421353682e-07, | |
| "loss": 0.3608, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.5910931174089065, | |
| "grad_norm": 0.2036219035944316, | |
| "learning_rate": 8.101405277100549e-07, | |
| "loss": 0.3637, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.5951417004048585, | |
| "grad_norm": 0.2013190699807321, | |
| "learning_rate": 7.942840756714077e-07, | |
| "loss": 0.3641, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.59919028340081, | |
| "grad_norm": 0.2024573710290679, | |
| "learning_rate": 7.785811920255093e-07, | |
| "loss": 0.3604, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.603238866396762, | |
| "grad_norm": 0.189424814325877, | |
| "learning_rate": 7.630320023322024e-07, | |
| "loss": 0.37, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.607287449392713, | |
| "grad_norm": 0.19852639062390065, | |
| "learning_rate": 7.476366309223903e-07, | |
| "loss": 0.3718, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.611336032388664, | |
| "grad_norm": 0.19823785410781508, | |
| "learning_rate": 7.323952008970537e-07, | |
| "loss": 0.3765, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.21473574656131536, | |
| "learning_rate": 7.173078341262618e-07, | |
| "loss": 0.3345, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.619433198380567, | |
| "grad_norm": 0.2040683038607736, | |
| "learning_rate": 7.023746512481989e-07, | |
| "loss": 0.371, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.623481781376518, | |
| "grad_norm": 0.202055681702731, | |
| "learning_rate": 6.875957716681902e-07, | |
| "loss": 0.3701, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.62753036437247, | |
| "grad_norm": 0.21589729652577389, | |
| "learning_rate": 6.729713135577687e-07, | |
| "loss": 0.3532, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.20426191803874266, | |
| "learning_rate": 6.585013938537143e-07, | |
| "loss": 0.3759, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.635627530364372, | |
| "grad_norm": 0.1935729496011885, | |
| "learning_rate": 6.441861282571115e-07, | |
| "loss": 0.3868, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.6396761133603235, | |
| "grad_norm": 0.21587873976228916, | |
| "learning_rate": 6.300256312324471e-07, | |
| "loss": 0.3602, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.6437246963562755, | |
| "grad_norm": 0.19863625575461055, | |
| "learning_rate": 6.160200160066776e-07, | |
| "loss": 0.358, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.647773279352227, | |
| "grad_norm": 0.21535842550658527, | |
| "learning_rate": 6.021693945683327e-07, | |
| "loss": 0.3755, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.651821862348179, | |
| "grad_norm": 0.20224677981661188, | |
| "learning_rate": 5.884738776666044e-07, | |
| "loss": 0.3705, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.65587044534413, | |
| "grad_norm": 0.21339542311614876, | |
| "learning_rate": 5.749335748104878e-07, | |
| "loss": 0.3614, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.659919028340081, | |
| "grad_norm": 0.20316253719815278, | |
| "learning_rate": 5.615485942678778e-07, | |
| "loss": 0.3673, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.663967611336032, | |
| "grad_norm": 0.19934242794468543, | |
| "learning_rate": 5.483190430647289e-07, | |
| "loss": 0.3498, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.668016194331984, | |
| "grad_norm": 0.20056618370012835, | |
| "learning_rate": 5.352450269841747e-07, | |
| "loss": 0.3604, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.672064777327935, | |
| "grad_norm": 0.2034864894848634, | |
| "learning_rate": 5.22326650565701e-07, | |
| "loss": 0.3605, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.676113360323887, | |
| "grad_norm": 0.20370138308033525, | |
| "learning_rate": 5.095640171043048e-07, | |
| "loss": 0.3628, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.680161943319838, | |
| "grad_norm": 0.1984153968111599, | |
| "learning_rate": 4.969572286496594e-07, | |
| "loss": 0.3498, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.2016956533866842, | |
| "learning_rate": 4.845063860053079e-07, | |
| "loss": 0.3595, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.6882591093117405, | |
| "grad_norm": 0.1955091887836674, | |
| "learning_rate": 4.7221158872785997e-07, | |
| "loss": 0.3613, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 0.19727411603979447, | |
| "learning_rate": 4.600729351261857e-07, | |
| "loss": 0.3837, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.696356275303644, | |
| "grad_norm": 0.21018713712394557, | |
| "learning_rate": 4.4809052226063355e-07, | |
| "loss": 0.3431, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.700404858299595, | |
| "grad_norm": 0.2001253345699822, | |
| "learning_rate": 4.3626444594226046e-07, | |
| "loss": 0.3445, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.704453441295547, | |
| "grad_norm": 0.21204449002803663, | |
| "learning_rate": 4.245948007320588e-07, | |
| "loss": 0.3548, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.708502024291498, | |
| "grad_norm": 0.20346651804506632, | |
| "learning_rate": 4.130816799401971e-07, | |
| "loss": 0.3681, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.712550607287449, | |
| "grad_norm": 0.20091250733950347, | |
| "learning_rate": 4.017251756252827e-07, | |
| "loss": 0.3783, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.716599190283401, | |
| "grad_norm": 0.19806438826209666, | |
| "learning_rate": 3.9052537859362473e-07, | |
| "loss": 0.3737, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.720647773279352, | |
| "grad_norm": 0.20264487357130123, | |
| "learning_rate": 3.7948237839849687e-07, | |
| "loss": 0.3651, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.724696356275303, | |
| "grad_norm": 0.20701506018725, | |
| "learning_rate": 3.6859626333943797e-07, | |
| "loss": 0.3602, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.728744939271255, | |
| "grad_norm": 0.1978865635253937, | |
| "learning_rate": 3.578671204615347e-07, | |
| "loss": 0.3606, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.732793522267206, | |
| "grad_norm": 0.20552404331066385, | |
| "learning_rate": 3.472950355547267e-07, | |
| "loss": 0.3595, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.19481063854165304, | |
| "learning_rate": 3.3688009315312244e-07, | |
| "loss": 0.3846, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.7408906882591095, | |
| "grad_norm": 0.1931526481549935, | |
| "learning_rate": 3.266223765343224e-07, | |
| "loss": 0.3681, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.744939271255061, | |
| "grad_norm": 0.20303810932838623, | |
| "learning_rate": 3.16521967718757e-07, | |
| "loss": 0.3484, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.748987854251012, | |
| "grad_norm": 0.20017770242603988, | |
| "learning_rate": 3.0657894746902726e-07, | |
| "loss": 0.3807, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.753036437246964, | |
| "grad_norm": 0.2020715390910552, | |
| "learning_rate": 2.9679339528925654e-07, | |
| "loss": 0.3651, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.757085020242915, | |
| "grad_norm": 0.19684569396093196, | |
| "learning_rate": 2.871653894244597e-07, | |
| "loss": 0.3699, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.761133603238866, | |
| "grad_norm": 0.19462441400414604, | |
| "learning_rate": 2.7769500685991266e-07, | |
| "loss": 0.3789, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.765182186234818, | |
| "grad_norm": 0.1988006046023525, | |
| "learning_rate": 2.683823233205485e-07, | |
| "loss": 0.3696, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.1859559161803913, | |
| "learning_rate": 2.592274132703287e-07, | |
| "loss": 0.3789, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.77327935222672, | |
| "grad_norm": 0.20336851900490685, | |
| "learning_rate": 2.502303499116754e-07, | |
| "loss": 0.3673, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.777327935222672, | |
| "grad_norm": 0.2017553782912709, | |
| "learning_rate": 2.413912051848666e-07, | |
| "loss": 0.3843, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.781376518218623, | |
| "grad_norm": 0.2056327140662509, | |
| "learning_rate": 2.3271004976746613e-07, | |
| "loss": 0.3728, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.7854251012145745, | |
| "grad_norm": 0.20029435933412107, | |
| "learning_rate": 2.2418695307376393e-07, | |
| "loss": 0.375, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.19548339683544133, | |
| "learning_rate": 2.1582198325420968e-07, | |
| "loss": 0.3817, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.793522267206478, | |
| "grad_norm": 0.19480387739998042, | |
| "learning_rate": 2.0761520719488447e-07, | |
| "loss": 0.3683, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.797570850202429, | |
| "grad_norm": 0.1961688261571079, | |
| "learning_rate": 1.9956669051694798e-07, | |
| "loss": 0.3721, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.801619433198381, | |
| "grad_norm": 0.20007253058724614, | |
| "learning_rate": 1.916764975761254e-07, | |
| "loss": 0.3501, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.805668016194332, | |
| "grad_norm": 0.2099756285692963, | |
| "learning_rate": 1.839446914621923e-07, | |
| "loss": 0.3598, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.809716599190283, | |
| "grad_norm": 0.19744468212707478, | |
| "learning_rate": 1.7637133399845962e-07, | |
| "loss": 0.3708, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.813765182186235, | |
| "grad_norm": 0.19611934434506437, | |
| "learning_rate": 1.6895648574129397e-07, | |
| "loss": 0.3702, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.817813765182186, | |
| "grad_norm": 0.19199107945606356, | |
| "learning_rate": 1.6170020597962465e-07, | |
| "loss": 0.3706, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.821862348178137, | |
| "grad_norm": 0.20429881146216095, | |
| "learning_rate": 1.5460255273447078e-07, | |
| "loss": 0.3648, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.825910931174089, | |
| "grad_norm": 0.19954257074415876, | |
| "learning_rate": 1.4766358275847715e-07, | |
| "loss": 0.3593, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.82995951417004, | |
| "grad_norm": 0.2002836637199141, | |
| "learning_rate": 1.4088335153546573e-07, | |
| "loss": 0.3669, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.834008097165992, | |
| "grad_norm": 0.20692083812472387, | |
| "learning_rate": 1.3426191327998496e-07, | |
| "loss": 0.3784, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.838056680161944, | |
| "grad_norm": 0.19981185268912346, | |
| "learning_rate": 1.277993209368744e-07, | |
| "loss": 0.3554, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.19479008423135766, | |
| "learning_rate": 1.214956261808542e-07, | |
| "loss": 0.3762, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.2063932918341991, | |
| "learning_rate": 1.1535087941609846e-07, | |
| "loss": 0.3698, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.850202429149798, | |
| "grad_norm": 0.1999644097864471, | |
| "learning_rate": 1.0936512977583802e-07, | |
| "loss": 0.3656, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.854251012145749, | |
| "grad_norm": 0.20033373294737533, | |
| "learning_rate": 1.0353842512196732e-07, | |
| "loss": 0.3693, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.8582995951417, | |
| "grad_norm": 0.1980884069545729, | |
| "learning_rate": 9.787081204466253e-08, | |
| "loss": 0.3774, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.862348178137652, | |
| "grad_norm": 0.19526567494798341, | |
| "learning_rate": 9.236233586200627e-08, | |
| "loss": 0.3692, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.866396761133603, | |
| "grad_norm": 0.20145958748033096, | |
| "learning_rate": 8.70130406196279e-08, | |
| "loss": 0.3548, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.870445344129554, | |
| "grad_norm": 0.19573891910498648, | |
| "learning_rate": 8.182296909035047e-08, | |
| "loss": 0.3788, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.874493927125506, | |
| "grad_norm": 0.2119332323710691, | |
| "learning_rate": 7.679216277384439e-08, | |
| "loss": 0.3696, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.8785425101214575, | |
| "grad_norm": 0.20641419686189585, | |
| "learning_rate": 7.19206618963053e-08, | |
| "loss": 0.374, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.882591093117409, | |
| "grad_norm": 0.19314670124093294, | |
| "learning_rate": 6.720850541012347e-08, | |
| "loss": 0.3787, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.886639676113361, | |
| "grad_norm": 0.2048378712119293, | |
| "learning_rate": 6.265573099357714e-08, | |
| "loss": 0.3474, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.890688259109312, | |
| "grad_norm": 0.19789567656411364, | |
| "learning_rate": 5.826237505053067e-08, | |
| "loss": 0.3696, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.19928824490018412, | |
| "learning_rate": 5.4028472710137e-08, | |
| "loss": 0.3721, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.898785425101215, | |
| "grad_norm": 0.2097778832090056, | |
| "learning_rate": 4.9954057826571146e-08, | |
| "loss": 0.3639, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.902834008097166, | |
| "grad_norm": 0.19818566831778567, | |
| "learning_rate": 4.6039162978743776e-08, | |
| "loss": 0.3627, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.906882591093117, | |
| "grad_norm": 0.20497056332589894, | |
| "learning_rate": 4.228381947004812e-08, | |
| "loss": 0.3617, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.910931174089069, | |
| "grad_norm": 0.19798551858995633, | |
| "learning_rate": 3.868805732811343e-08, | |
| "loss": 0.3757, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.91497975708502, | |
| "grad_norm": 0.204663246595883, | |
| "learning_rate": 3.525190530455636e-08, | |
| "loss": 0.3608, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.919028340080971, | |
| "grad_norm": 0.20055949274239243, | |
| "learning_rate": 3.197539087475887e-08, | |
| "loss": 0.3637, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.20194826392753137, | |
| "learning_rate": 2.8858540237641786e-08, | |
| "loss": 0.3696, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.9271255060728745, | |
| "grad_norm": 0.1999860866482001, | |
| "learning_rate": 2.5901378315458247e-08, | |
| "loss": 0.3595, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.931174089068826, | |
| "grad_norm": 0.1886764180609573, | |
| "learning_rate": 2.3103928753598347e-08, | |
| "loss": 0.3892, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.935222672064778, | |
| "grad_norm": 0.20181989524418878, | |
| "learning_rate": 2.0466213920393718e-08, | |
| "loss": 0.3437, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.939271255060729, | |
| "grad_norm": 0.1996475557471868, | |
| "learning_rate": 1.7988254906942115e-08, | |
| "loss": 0.3596, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94331983805668, | |
| "grad_norm": 0.2083233023250047, | |
| "learning_rate": 1.5670071526936447e-08, | |
| "loss": 0.3466, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.1983409520050114, | |
| "learning_rate": 1.3511682316509344e-08, | |
| "loss": 0.3622, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.951417004048583, | |
| "grad_norm": 0.20506363420655346, | |
| "learning_rate": 1.1513104534086605e-08, | |
| "loss": 0.368, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.955465587044534, | |
| "grad_norm": 0.19672882117606416, | |
| "learning_rate": 9.67435416023843e-09, | |
| "loss": 0.3707, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.959514170040486, | |
| "grad_norm": 0.19626491777892482, | |
| "learning_rate": 7.995445897566178e-09, | |
| "loss": 0.363, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.963562753036437, | |
| "grad_norm": 0.19520965436667842, | |
| "learning_rate": 6.476393170578021e-09, | |
| "loss": 0.3792, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.967611336032388, | |
| "grad_norm": 0.19786206163795564, | |
| "learning_rate": 5.11720812557348e-09, | |
| "loss": 0.3558, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.97165991902834, | |
| "grad_norm": 0.18842386520977678, | |
| "learning_rate": 3.917901630563492e-09, | |
| "loss": 0.376, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.9757085020242915, | |
| "grad_norm": 0.19696524312597324, | |
| "learning_rate": 2.878483275170485e-09, | |
| "loss": 0.3681, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.979757085020243, | |
| "grad_norm": 0.19579986268192207, | |
| "learning_rate": 1.998961370555108e-09, | |
| "loss": 0.362, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.983805668016195, | |
| "grad_norm": 0.19216686326176435, | |
| "learning_rate": 1.2793429493518361e-09, | |
| "loss": 0.3798, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.987854251012146, | |
| "grad_norm": 0.1980128217204017, | |
| "learning_rate": 7.196337656112384e-10, | |
| "loss": 0.3584, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.991902834008097, | |
| "grad_norm": 0.20214734404899665, | |
| "learning_rate": 3.198382947511292e-10, | |
| "loss": 0.3633, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.995951417004049, | |
| "grad_norm": 0.20323061903510864, | |
| "learning_rate": 7.995973352770137e-11, | |
| "loss": 0.3718, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.1935039532546334, | |
| "learning_rate": 0.0, | |
| "loss": 0.364, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1235, | |
| "total_flos": 5.028117034995548e+18, | |
| "train_loss": 0.49688222311768937, | |
| "train_runtime": 29417.885, | |
| "train_samples_per_second": 5.371, | |
| "train_steps_per_second": 0.042 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1235, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.028117034995548e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |