| { | |
| "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": 11.928018600966656, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 1.3988, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008097165991902834, | |
| "grad_norm": 11.971590288318163, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 1.4112, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012145748987854251, | |
| "grad_norm": 11.84560058135189, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 1.3838, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016194331983805668, | |
| "grad_norm": 11.55828682128216, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 1.39, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.020242914979757085, | |
| "grad_norm": 10.077145348112152, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 1.3448, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024291497975708502, | |
| "grad_norm": 9.587556653929228, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.3676, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02834008097165992, | |
| "grad_norm": 6.405381639175268, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 1.3091, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032388663967611336, | |
| "grad_norm": 5.951597880873701, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 1.2842, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03643724696356275, | |
| "grad_norm": 3.7532825696129346, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 1.2331, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04048582995951417, | |
| "grad_norm": 3.4153462885988337, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 1.2168, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044534412955465584, | |
| "grad_norm": 3.2317552335029154, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 1.2066, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048582995951417005, | |
| "grad_norm": 5.028721629612329, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 1.1553, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 5.373822019519315, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 1.1538, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05668016194331984, | |
| "grad_norm": 4.813541869164253, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 1.1472, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06072874493927125, | |
| "grad_norm": 4.277332637537192, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 1.13, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06477732793522267, | |
| "grad_norm": 2.83355646326921, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 1.0738, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06882591093117409, | |
| "grad_norm": 2.931363395139026, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 1.0565, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0728744939271255, | |
| "grad_norm": 2.500563437028688, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 1.0321, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 1.8783999341563729, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 1.0106, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08097165991902834, | |
| "grad_norm": 1.7228870281962572, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 1.0009, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08502024291497975, | |
| "grad_norm": 1.4563937172329025, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 0.982, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08906882591093117, | |
| "grad_norm": 1.3289870031897932, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 0.9763, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0931174089068826, | |
| "grad_norm": 1.685845469475488, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 0.9646, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09716599190283401, | |
| "grad_norm": 1.6059386687141395, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 0.9419, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10121457489878542, | |
| "grad_norm": 1.0950505375486572, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.9612, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 1.2231206585198313, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 0.9567, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.10931174089068826, | |
| "grad_norm": 1.3047970699796139, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 0.9485, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11336032388663968, | |
| "grad_norm": 1.2279455764191831, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 0.9435, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11740890688259109, | |
| "grad_norm": 1.105195237047331, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 0.9323, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.1214574898785425, | |
| "grad_norm": 1.1319754795525587, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.9243, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12550607287449392, | |
| "grad_norm": 1.0652355570445666, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9251, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12955465587044535, | |
| "grad_norm": 0.8968806380988843, | |
| "learning_rate": 1.0322580645161291e-05, | |
| "loss": 0.911, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13360323886639677, | |
| "grad_norm": 1.0805128613852644, | |
| "learning_rate": 1.0645161290322582e-05, | |
| "loss": 0.9272, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13765182186234817, | |
| "grad_norm": 0.932653221108751, | |
| "learning_rate": 1.096774193548387e-05, | |
| "loss": 0.8944, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1417004048582996, | |
| "grad_norm": 1.0668560140326413, | |
| "learning_rate": 1.1290322580645164e-05, | |
| "loss": 0.9021, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.145748987854251, | |
| "grad_norm": 1.0115247516436254, | |
| "learning_rate": 1.1612903225806453e-05, | |
| "loss": 0.8965, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14979757085020243, | |
| "grad_norm": 1.0515589899415758, | |
| "learning_rate": 1.1935483870967743e-05, | |
| "loss": 0.8885, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.8584965086245059, | |
| "learning_rate": 1.2258064516129034e-05, | |
| "loss": 0.9043, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 1.0253904806362593, | |
| "learning_rate": 1.2580645161290324e-05, | |
| "loss": 0.8823, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16194331983805668, | |
| "grad_norm": 0.8867676612128969, | |
| "learning_rate": 1.2903225806451613e-05, | |
| "loss": 0.8896, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1659919028340081, | |
| "grad_norm": 0.875793124297728, | |
| "learning_rate": 1.3225806451612903e-05, | |
| "loss": 0.8866, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1700404858299595, | |
| "grad_norm": 0.8911009306050302, | |
| "learning_rate": 1.3548387096774194e-05, | |
| "loss": 0.8699, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.17408906882591094, | |
| "grad_norm": 0.892567843179215, | |
| "learning_rate": 1.3870967741935486e-05, | |
| "loss": 0.8676, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17813765182186234, | |
| "grad_norm": 0.8452343699026483, | |
| "learning_rate": 1.4193548387096776e-05, | |
| "loss": 0.8877, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18218623481781376, | |
| "grad_norm": 0.8842077593358864, | |
| "learning_rate": 1.4516129032258066e-05, | |
| "loss": 0.8649, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1862348178137652, | |
| "grad_norm": 0.9778411244039633, | |
| "learning_rate": 1.4838709677419357e-05, | |
| "loss": 0.8842, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1902834008097166, | |
| "grad_norm": 0.811777900908325, | |
| "learning_rate": 1.5161290322580646e-05, | |
| "loss": 0.8574, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.19433198380566802, | |
| "grad_norm": 0.8700805459042197, | |
| "learning_rate": 1.5483870967741936e-05, | |
| "loss": 0.8697, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19838056680161945, | |
| "grad_norm": 0.9592203657807011, | |
| "learning_rate": 1.5806451612903226e-05, | |
| "loss": 0.8727, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.20242914979757085, | |
| "grad_norm": 0.9131161259100298, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.8522, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20647773279352227, | |
| "grad_norm": 0.8822263834191204, | |
| "learning_rate": 1.6451612903225807e-05, | |
| "loss": 0.8547, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.9803120437409316, | |
| "learning_rate": 1.6774193548387098e-05, | |
| "loss": 0.8527, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2145748987854251, | |
| "grad_norm": 0.8751123459515155, | |
| "learning_rate": 1.7096774193548388e-05, | |
| "loss": 0.855, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21862348178137653, | |
| "grad_norm": 1.197819165577258, | |
| "learning_rate": 1.741935483870968e-05, | |
| "loss": 0.8466, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22267206477732793, | |
| "grad_norm": 1.0432712532286252, | |
| "learning_rate": 1.774193548387097e-05, | |
| "loss": 0.8555, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22672064777327935, | |
| "grad_norm": 1.3956159825369678, | |
| "learning_rate": 1.806451612903226e-05, | |
| "loss": 0.8377, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 1.1206869461533875, | |
| "learning_rate": 1.838709677419355e-05, | |
| "loss": 0.836, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.23481781376518218, | |
| "grad_norm": 1.3388158863899136, | |
| "learning_rate": 1.870967741935484e-05, | |
| "loss": 0.8352, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2388663967611336, | |
| "grad_norm": 1.1343904199519586, | |
| "learning_rate": 1.903225806451613e-05, | |
| "loss": 0.8473, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.242914979757085, | |
| "grad_norm": 1.3066170567339175, | |
| "learning_rate": 1.935483870967742e-05, | |
| "loss": 0.8423, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24696356275303644, | |
| "grad_norm": 1.2120525298327263, | |
| "learning_rate": 1.967741935483871e-05, | |
| "loss": 0.8373, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.25101214574898784, | |
| "grad_norm": 1.1253037117819134, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8355, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2550607287449393, | |
| "grad_norm": 1.2049021617862234, | |
| "learning_rate": 2.0322580645161292e-05, | |
| "loss": 0.8175, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2591093117408907, | |
| "grad_norm": 1.0329466363024724, | |
| "learning_rate": 2.0645161290322582e-05, | |
| "loss": 0.84, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.3859581282831428, | |
| "learning_rate": 2.0967741935483873e-05, | |
| "loss": 0.8214, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26720647773279355, | |
| "grad_norm": 0.9607265263502548, | |
| "learning_rate": 2.1290322580645163e-05, | |
| "loss": 0.8478, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.27125506072874495, | |
| "grad_norm": 1.5287120146656288, | |
| "learning_rate": 2.161290322580645e-05, | |
| "loss": 0.8462, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.27530364372469635, | |
| "grad_norm": 1.0528634723279071, | |
| "learning_rate": 2.193548387096774e-05, | |
| "loss": 0.8097, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2793522267206478, | |
| "grad_norm": 1.872251324430632, | |
| "learning_rate": 2.225806451612903e-05, | |
| "loss": 0.8234, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2834008097165992, | |
| "grad_norm": 1.8327638643966244, | |
| "learning_rate": 2.2580645161290328e-05, | |
| "loss": 0.8339, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2874493927125506, | |
| "grad_norm": 0.9870033342877917, | |
| "learning_rate": 2.2903225806451615e-05, | |
| "loss": 0.8175, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.291497975708502, | |
| "grad_norm": 1.4236376583659134, | |
| "learning_rate": 2.3225806451612906e-05, | |
| "loss": 0.8287, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.29554655870445345, | |
| "grad_norm": 0.9946164867153161, | |
| "learning_rate": 2.3548387096774196e-05, | |
| "loss": 0.8246, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.29959514170040485, | |
| "grad_norm": 1.2594464165272394, | |
| "learning_rate": 2.3870967741935486e-05, | |
| "loss": 0.8395, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.30364372469635625, | |
| "grad_norm": 0.8676656689845658, | |
| "learning_rate": 2.4193548387096777e-05, | |
| "loss": 0.8296, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 1.2902636692240612, | |
| "learning_rate": 2.4516129032258067e-05, | |
| "loss": 0.8259, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3117408906882591, | |
| "grad_norm": 1.188592021900784, | |
| "learning_rate": 2.4838709677419358e-05, | |
| "loss": 0.8196, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 1.0004068234187906, | |
| "learning_rate": 2.5161290322580648e-05, | |
| "loss": 0.8155, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31983805668016196, | |
| "grad_norm": 1.0122408302265429, | |
| "learning_rate": 2.548387096774194e-05, | |
| "loss": 0.849, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32388663967611336, | |
| "grad_norm": 0.8965973124390435, | |
| "learning_rate": 2.5806451612903226e-05, | |
| "loss": 0.8197, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.32793522267206476, | |
| "grad_norm": 1.0432783079208694, | |
| "learning_rate": 2.6129032258064516e-05, | |
| "loss": 0.8192, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3319838056680162, | |
| "grad_norm": 1.2524032374136493, | |
| "learning_rate": 2.6451612903225806e-05, | |
| "loss": 0.8297, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3360323886639676, | |
| "grad_norm": 0.9057405541306237, | |
| "learning_rate": 2.6774193548387097e-05, | |
| "loss": 0.802, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.340080971659919, | |
| "grad_norm": 1.1360211295179483, | |
| "learning_rate": 2.7096774193548387e-05, | |
| "loss": 0.8206, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3441295546558704, | |
| "grad_norm": 0.9579736010591917, | |
| "learning_rate": 2.7419354838709678e-05, | |
| "loss": 0.8082, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3481781376518219, | |
| "grad_norm": 1.2010602787896734, | |
| "learning_rate": 2.774193548387097e-05, | |
| "loss": 0.7888, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3522267206477733, | |
| "grad_norm": 0.8984346172619635, | |
| "learning_rate": 2.8064516129032262e-05, | |
| "loss": 0.8071, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3562753036437247, | |
| "grad_norm": 1.2615693755492838, | |
| "learning_rate": 2.8387096774193552e-05, | |
| "loss": 0.8155, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3603238866396761, | |
| "grad_norm": 1.0526723648793794, | |
| "learning_rate": 2.8709677419354843e-05, | |
| "loss": 0.8155, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3643724696356275, | |
| "grad_norm": 1.1500428118135162, | |
| "learning_rate": 2.9032258064516133e-05, | |
| "loss": 0.8027, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 1.0804831787691072, | |
| "learning_rate": 2.9354838709677423e-05, | |
| "loss": 0.8029, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3724696356275304, | |
| "grad_norm": 1.0355693010971474, | |
| "learning_rate": 2.9677419354838714e-05, | |
| "loss": 0.8188, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3765182186234818, | |
| "grad_norm": 0.9972006778574262, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.8122, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3805668016194332, | |
| "grad_norm": 0.9973660527973609, | |
| "learning_rate": 3.032258064516129e-05, | |
| "loss": 0.8219, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 1.028715921382547, | |
| "learning_rate": 3.0645161290322585e-05, | |
| "loss": 0.7976, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.38866396761133604, | |
| "grad_norm": 0.9072490401613238, | |
| "learning_rate": 3.096774193548387e-05, | |
| "loss": 0.8118, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.39271255060728744, | |
| "grad_norm": 1.0279071156893875, | |
| "learning_rate": 3.1290322580645166e-05, | |
| "loss": 0.8166, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3967611336032389, | |
| "grad_norm": 0.9784233035190142, | |
| "learning_rate": 3.161290322580645e-05, | |
| "loss": 0.7918, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.4008097165991903, | |
| "grad_norm": 1.041212188069313, | |
| "learning_rate": 3.193548387096774e-05, | |
| "loss": 0.8035, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4048582995951417, | |
| "grad_norm": 1.5224674967899223, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.8154, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4089068825910931, | |
| "grad_norm": 1.287779535262822, | |
| "learning_rate": 3.258064516129033e-05, | |
| "loss": 0.7909, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.41295546558704455, | |
| "grad_norm": 0.9354490321952226, | |
| "learning_rate": 3.2903225806451614e-05, | |
| "loss": 0.7986, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.41700404858299595, | |
| "grad_norm": 1.063164163676534, | |
| "learning_rate": 3.322580645161291e-05, | |
| "loss": 0.8161, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 1.2225524854852559, | |
| "learning_rate": 3.3548387096774195e-05, | |
| "loss": 0.8143, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4251012145748988, | |
| "grad_norm": 0.8791269553349176, | |
| "learning_rate": 3.387096774193549e-05, | |
| "loss": 0.8148, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4291497975708502, | |
| "grad_norm": 0.9953902788540505, | |
| "learning_rate": 3.4193548387096776e-05, | |
| "loss": 0.8032, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4331983805668016, | |
| "grad_norm": 0.9470959921846567, | |
| "learning_rate": 3.451612903225807e-05, | |
| "loss": 0.7934, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.43724696356275305, | |
| "grad_norm": 0.9434810321214073, | |
| "learning_rate": 3.483870967741936e-05, | |
| "loss": 0.7957, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.44129554655870445, | |
| "grad_norm": 0.8507978606612988, | |
| "learning_rate": 3.516129032258065e-05, | |
| "loss": 0.7953, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44534412955465585, | |
| "grad_norm": 1.0644595541667905, | |
| "learning_rate": 3.548387096774194e-05, | |
| "loss": 0.7994, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4493927125506073, | |
| "grad_norm": 1.0719267607563627, | |
| "learning_rate": 3.5806451612903225e-05, | |
| "loss": 0.8166, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4534412955465587, | |
| "grad_norm": 0.9577993019256591, | |
| "learning_rate": 3.612903225806452e-05, | |
| "loss": 0.8033, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4574898785425101, | |
| "grad_norm": 0.9759400972185263, | |
| "learning_rate": 3.6451612903225805e-05, | |
| "loss": 0.8098, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.8888338162998822, | |
| "learning_rate": 3.67741935483871e-05, | |
| "loss": 0.7997, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46558704453441296, | |
| "grad_norm": 1.041987342632825, | |
| "learning_rate": 3.7096774193548386e-05, | |
| "loss": 0.7995, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.46963562753036436, | |
| "grad_norm": 1.0536170793825304, | |
| "learning_rate": 3.741935483870968e-05, | |
| "loss": 0.8088, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 1.1969018988301086, | |
| "learning_rate": 3.7741935483870974e-05, | |
| "loss": 0.7971, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.4777327935222672, | |
| "grad_norm": 0.7811410963303791, | |
| "learning_rate": 3.806451612903226e-05, | |
| "loss": 0.7668, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4817813765182186, | |
| "grad_norm": 1.1647468136180446, | |
| "learning_rate": 3.8387096774193555e-05, | |
| "loss": 0.8009, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48582995951417, | |
| "grad_norm": 1.0409380135017239, | |
| "learning_rate": 3.870967741935484e-05, | |
| "loss": 0.7991, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4898785425101215, | |
| "grad_norm": 1.2068129874166615, | |
| "learning_rate": 3.9032258064516135e-05, | |
| "loss": 0.7954, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4939271255060729, | |
| "grad_norm": 1.0163316443431556, | |
| "learning_rate": 3.935483870967742e-05, | |
| "loss": 0.7905, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4979757085020243, | |
| "grad_norm": 1.314453195641561, | |
| "learning_rate": 3.9677419354838716e-05, | |
| "loss": 0.7968, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5020242914979757, | |
| "grad_norm": 1.0913271126458255, | |
| "learning_rate": 4e-05, | |
| "loss": 0.8133, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5060728744939271, | |
| "grad_norm": 1.048009999014979, | |
| "learning_rate": 3.9999920040266474e-05, | |
| "loss": 0.796, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5101214574898786, | |
| "grad_norm": 1.2211837666030723, | |
| "learning_rate": 3.999968016170525e-05, | |
| "loss": 0.818, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5141700404858299, | |
| "grad_norm": 0.9223595742648767, | |
| "learning_rate": 3.9999280366234394e-05, | |
| "loss": 0.7791, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5182186234817814, | |
| "grad_norm": 0.9470447935355637, | |
| "learning_rate": 3.999872065705065e-05, | |
| "loss": 0.8137, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5222672064777328, | |
| "grad_norm": 1.257717396772088, | |
| "learning_rate": 3.999800103862945e-05, | |
| "loss": 0.805, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.0194251268181216, | |
| "learning_rate": 3.9997121516724834e-05, | |
| "loss": 0.8103, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5303643724696356, | |
| "grad_norm": 0.9113114180107196, | |
| "learning_rate": 3.999608209836944e-05, | |
| "loss": 0.7992, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5344129554655871, | |
| "grad_norm": 1.1032106004963393, | |
| "learning_rate": 3.999488279187443e-05, | |
| "loss": 0.7801, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.8334298883365372, | |
| "learning_rate": 3.9993523606829425e-05, | |
| "loss": 0.7693, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5425101214574899, | |
| "grad_norm": 1.0233748055645573, | |
| "learning_rate": 3.9992004554102435e-05, | |
| "loss": 0.804, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5465587044534413, | |
| "grad_norm": 1.0066280191198014, | |
| "learning_rate": 3.999032564583976e-05, | |
| "loss": 0.7869, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5506072874493927, | |
| "grad_norm": 0.8716561991934588, | |
| "learning_rate": 3.998848689546591e-05, | |
| "loss": 0.8011, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5546558704453441, | |
| "grad_norm": 1.0332418678453272, | |
| "learning_rate": 3.9986488317683496e-05, | |
| "loss": 0.7991, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5587044534412956, | |
| "grad_norm": 0.894797651837729, | |
| "learning_rate": 3.9984329928473067e-05, | |
| "loss": 0.7843, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.562753036437247, | |
| "grad_norm": 0.9089634929206187, | |
| "learning_rate": 3.998201174509306e-05, | |
| "loss": 0.7943, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5668016194331984, | |
| "grad_norm": 0.9900063101953183, | |
| "learning_rate": 3.997953378607961e-05, | |
| "loss": 0.7932, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5708502024291497, | |
| "grad_norm": 0.910456770796043, | |
| "learning_rate": 3.9976896071246403e-05, | |
| "loss": 0.7781, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5748987854251012, | |
| "grad_norm": 1.0586825613254516, | |
| "learning_rate": 3.997409862168454e-05, | |
| "loss": 0.789, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.9962112364369954, | |
| "learning_rate": 3.997114145976237e-05, | |
| "loss": 0.8079, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.582995951417004, | |
| "grad_norm": 0.7963003838671762, | |
| "learning_rate": 3.9968024609125245e-05, | |
| "loss": 0.786, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5870445344129555, | |
| "grad_norm": 1.2766971889787178, | |
| "learning_rate": 3.996474809469545e-05, | |
| "loss": 0.7829, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5910931174089069, | |
| "grad_norm": 0.807947309629623, | |
| "learning_rate": 3.996131194267189e-05, | |
| "loss": 0.7842, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5951417004048583, | |
| "grad_norm": 0.9918744506621566, | |
| "learning_rate": 3.995771618052995e-05, | |
| "loss": 0.7818, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5991902834008097, | |
| "grad_norm": 0.9574316796499617, | |
| "learning_rate": 3.9953960837021256e-05, | |
| "loss": 0.7866, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6032388663967612, | |
| "grad_norm": 0.8996263785365758, | |
| "learning_rate": 3.995004594217343e-05, | |
| "loss": 0.7969, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6072874493927125, | |
| "grad_norm": 0.8906469120006995, | |
| "learning_rate": 3.994597152728987e-05, | |
| "loss": 0.7876, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.611336032388664, | |
| "grad_norm": 0.7703961216577253, | |
| "learning_rate": 3.994173762494947e-05, | |
| "loss": 0.7803, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.9149453584530621, | |
| "learning_rate": 3.9937344269006424e-05, | |
| "loss": 0.7913, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6194331983805668, | |
| "grad_norm": 1.057647357746797, | |
| "learning_rate": 3.993279149458988e-05, | |
| "loss": 0.7809, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6234817813765182, | |
| "grad_norm": 1.1590284953988936, | |
| "learning_rate": 3.9928079338103695e-05, | |
| "loss": 0.7945, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6275303643724697, | |
| "grad_norm": 0.8503700201225144, | |
| "learning_rate": 3.992320783722616e-05, | |
| "loss": 0.7734, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 1.4524784413700682, | |
| "learning_rate": 3.991817703090965e-05, | |
| "loss": 0.7798, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6356275303643725, | |
| "grad_norm": 0.8709012309131625, | |
| "learning_rate": 3.991298695938038e-05, | |
| "loss": 0.7815, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6396761133603239, | |
| "grad_norm": 1.16396381782402, | |
| "learning_rate": 3.9907637664138e-05, | |
| "loss": 0.7737, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6437246963562753, | |
| "grad_norm": 0.865858172051471, | |
| "learning_rate": 3.990212918795534e-05, | |
| "loss": 0.7908, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6477732793522267, | |
| "grad_norm": 1.063922298581379, | |
| "learning_rate": 3.9896461574878036e-05, | |
| "loss": 0.7772, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6518218623481782, | |
| "grad_norm": 0.8596003387973127, | |
| "learning_rate": 3.989063487022416e-05, | |
| "loss": 0.7756, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6558704453441295, | |
| "grad_norm": 1.1807674175020897, | |
| "learning_rate": 3.9884649120583903e-05, | |
| "loss": 0.7614, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.659919028340081, | |
| "grad_norm": 0.9072779329078736, | |
| "learning_rate": 3.987850437381915e-05, | |
| "loss": 0.7831, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6639676113360324, | |
| "grad_norm": 1.1706617634329022, | |
| "learning_rate": 3.987220067906313e-05, | |
| "loss": 0.7826, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6680161943319838, | |
| "grad_norm": 0.8594104248024539, | |
| "learning_rate": 3.986573808672002e-05, | |
| "loss": 0.7593, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6720647773279352, | |
| "grad_norm": 1.0408199523714607, | |
| "learning_rate": 3.9859116648464534e-05, | |
| "loss": 0.7883, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6761133603238867, | |
| "grad_norm": 1.0764085705341881, | |
| "learning_rate": 3.985233641724152e-05, | |
| "loss": 0.7719, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.680161943319838, | |
| "grad_norm": 0.9065066892355266, | |
| "learning_rate": 3.984539744726554e-05, | |
| "loss": 0.781, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.9768150276628332, | |
| "learning_rate": 3.983829979402038e-05, | |
| "loss": 0.7737, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6882591093117408, | |
| "grad_norm": 1.0136738496785376, | |
| "learning_rate": 3.983104351425871e-05, | |
| "loss": 0.7786, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.786320577335759, | |
| "learning_rate": 3.982362866600154e-05, | |
| "loss": 0.7743, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6963562753036437, | |
| "grad_norm": 0.944433671538922, | |
| "learning_rate": 3.9816055308537806e-05, | |
| "loss": 0.7836, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7004048582995951, | |
| "grad_norm": 1.0076292588498705, | |
| "learning_rate": 3.980832350242388e-05, | |
| "loss": 0.7818, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7044534412955465, | |
| "grad_norm": 0.7319994949911552, | |
| "learning_rate": 3.980043330948306e-05, | |
| "loss": 0.758, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.708502024291498, | |
| "grad_norm": 0.8209878274379594, | |
| "learning_rate": 3.9792384792805124e-05, | |
| "loss": 0.7557, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7125506072874493, | |
| "grad_norm": 0.7787530388098032, | |
| "learning_rate": 3.978417801674579e-05, | |
| "loss": 0.7788, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7165991902834008, | |
| "grad_norm": 0.9210376102888569, | |
| "learning_rate": 3.9775813046926244e-05, | |
| "loss": 0.7592, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7206477732793523, | |
| "grad_norm": 0.8960899762833353, | |
| "learning_rate": 3.976728995023254e-05, | |
| "loss": 0.777, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7246963562753036, | |
| "grad_norm": 0.8302872857380716, | |
| "learning_rate": 3.975860879481514e-05, | |
| "loss": 0.773, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.728744939271255, | |
| "grad_norm": 0.8805608759905658, | |
| "learning_rate": 3.974976965008833e-05, | |
| "loss": 0.7675, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7327935222672065, | |
| "grad_norm": 0.8379754702623596, | |
| "learning_rate": 3.9740772586729674e-05, | |
| "loss": 0.7741, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.8050269526898328, | |
| "learning_rate": 3.973161767667946e-05, | |
| "loss": 0.7707, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7408906882591093, | |
| "grad_norm": 0.783221697502902, | |
| "learning_rate": 3.972230499314009e-05, | |
| "loss": 0.7601, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7449392712550608, | |
| "grad_norm": 0.8347727235165061, | |
| "learning_rate": 3.9712834610575545e-05, | |
| "loss": 0.7634, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7489878542510121, | |
| "grad_norm": 0.7879576553398304, | |
| "learning_rate": 3.9703206604710746e-05, | |
| "loss": 0.7575, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7530364372469636, | |
| "grad_norm": 0.7334333976000407, | |
| "learning_rate": 3.969342105253098e-05, | |
| "loss": 0.7645, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.757085020242915, | |
| "grad_norm": 0.9482124827613034, | |
| "learning_rate": 3.9683478032281246e-05, | |
| "loss": 0.754, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7611336032388664, | |
| "grad_norm": 0.7964998596649242, | |
| "learning_rate": 3.967337762346568e-05, | |
| "loss": 0.7917, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7651821862348178, | |
| "grad_norm": 0.8196346538025542, | |
| "learning_rate": 3.9663119906846885e-05, | |
| "loss": 0.7687, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.7506519893284438, | |
| "learning_rate": 3.965270496444528e-05, | |
| "loss": 0.7667, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7732793522267206, | |
| "grad_norm": 0.7763957802846687, | |
| "learning_rate": 3.9642132879538465e-05, | |
| "loss": 0.7731, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7773279352226721, | |
| "grad_norm": 0.7759219084418187, | |
| "learning_rate": 3.9631403736660566e-05, | |
| "loss": 0.7505, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7813765182186235, | |
| "grad_norm": 0.7855294899566686, | |
| "learning_rate": 3.962051762160151e-05, | |
| "loss": 0.7576, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7854251012145749, | |
| "grad_norm": 0.7932394222427307, | |
| "learning_rate": 3.960947462140638e-05, | |
| "loss": 0.765, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.6861142943401001, | |
| "learning_rate": 3.959827482437472e-05, | |
| "loss": 0.7411, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7935222672064778, | |
| "grad_norm": 0.8128703949867644, | |
| "learning_rate": 3.958691832005981e-05, | |
| "loss": 0.7645, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7975708502024291, | |
| "grad_norm": 0.765708378246437, | |
| "learning_rate": 3.9575405199267944e-05, | |
| "loss": 0.7583, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8016194331983806, | |
| "grad_norm": 0.6963956229967145, | |
| "learning_rate": 3.956373555405774e-05, | |
| "loss": 0.7587, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.805668016194332, | |
| "grad_norm": 0.75355603549981, | |
| "learning_rate": 3.955190947773937e-05, | |
| "loss": 0.7639, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8097165991902834, | |
| "grad_norm": 0.8525467777565977, | |
| "learning_rate": 3.9539927064873824e-05, | |
| "loss": 0.7712, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8137651821862348, | |
| "grad_norm": 0.7004232777425698, | |
| "learning_rate": 3.952778841127214e-05, | |
| "loss": 0.7733, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8178137651821862, | |
| "grad_norm": 0.7947820736350497, | |
| "learning_rate": 3.95154936139947e-05, | |
| "loss": 0.7661, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8218623481781376, | |
| "grad_norm": 0.7885054590300452, | |
| "learning_rate": 3.950304277135034e-05, | |
| "loss": 0.7477, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8259109311740891, | |
| "grad_norm": 0.6434610289343516, | |
| "learning_rate": 3.94904359828957e-05, | |
| "loss": 0.7577, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8299595141700404, | |
| "grad_norm": 0.8395380775621569, | |
| "learning_rate": 3.94776733494343e-05, | |
| "loss": 0.7516, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8340080971659919, | |
| "grad_norm": 0.7695791402632737, | |
| "learning_rate": 3.9464754973015826e-05, | |
| "loss": 0.7663, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8380566801619433, | |
| "grad_norm": 0.753971936092867, | |
| "learning_rate": 3.9451680956935275e-05, | |
| "loss": 0.7613, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.7189354083716368, | |
| "learning_rate": 3.943845140573213e-05, | |
| "loss": 0.766, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.8210880518443732, | |
| "learning_rate": 3.942506642518952e-05, | |
| "loss": 0.7553, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8502024291497976, | |
| "grad_norm": 0.7659965996411644, | |
| "learning_rate": 3.9411526122333396e-05, | |
| "loss": 0.7567, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.854251012145749, | |
| "grad_norm": 0.6998742615594465, | |
| "learning_rate": 3.939783060543167e-05, | |
| "loss": 0.7557, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8582995951417004, | |
| "grad_norm": 0.7867922703754036, | |
| "learning_rate": 3.938397998399333e-05, | |
| "loss": 0.7603, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.8623481781376519, | |
| "grad_norm": 0.7157600657792184, | |
| "learning_rate": 3.936997436876756e-05, | |
| "loss": 0.7727, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8663967611336032, | |
| "grad_norm": 0.7623079774175654, | |
| "learning_rate": 3.935581387174289e-05, | |
| "loss": 0.7538, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.8704453441295547, | |
| "grad_norm": 0.8325547651981385, | |
| "learning_rate": 3.9341498606146295e-05, | |
| "loss": 0.7645, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8744939271255061, | |
| "grad_norm": 0.8739576411154997, | |
| "learning_rate": 3.932702868644224e-05, | |
| "loss": 0.7686, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.8785425101214575, | |
| "grad_norm": 0.7754166854197276, | |
| "learning_rate": 3.931240422833181e-05, | |
| "loss": 0.7443, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8825910931174089, | |
| "grad_norm": 0.8255535748196109, | |
| "learning_rate": 3.9297625348751806e-05, | |
| "loss": 0.7429, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8866396761133604, | |
| "grad_norm": 0.852869389449532, | |
| "learning_rate": 3.9282692165873744e-05, | |
| "loss": 0.7421, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8906882591093117, | |
| "grad_norm": 0.8290870432215759, | |
| "learning_rate": 3.9267604799102945e-05, | |
| "loss": 0.7509, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.8035834678839668, | |
| "learning_rate": 3.9252363369077615e-05, | |
| "loss": 0.7537, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8987854251012146, | |
| "grad_norm": 0.7974385129652811, | |
| "learning_rate": 3.9236967997667796e-05, | |
| "loss": 0.7652, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.902834008097166, | |
| "grad_norm": 0.7842931006546915, | |
| "learning_rate": 3.9221418807974496e-05, | |
| "loss": 0.747, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9068825910931174, | |
| "grad_norm": 0.7482259973673658, | |
| "learning_rate": 3.92057159243286e-05, | |
| "loss": 0.7468, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9109311740890689, | |
| "grad_norm": 0.8361620173435462, | |
| "learning_rate": 3.9189859472289956e-05, | |
| "loss": 0.7517, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9149797570850202, | |
| "grad_norm": 0.7255325168095813, | |
| "learning_rate": 3.9173849578646324e-05, | |
| "loss": 0.7354, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9190283400809717, | |
| "grad_norm": 0.7529583547088349, | |
| "learning_rate": 3.9157686371412395e-05, | |
| "loss": 0.7488, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.720456149645914, | |
| "learning_rate": 3.914136997982874e-05, | |
| "loss": 0.7325, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9271255060728745, | |
| "grad_norm": 0.7910700220166, | |
| "learning_rate": 3.912490053436079e-05, | |
| "loss": 0.7479, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9311740890688259, | |
| "grad_norm": 0.7173589215219779, | |
| "learning_rate": 3.9108278166697805e-05, | |
| "loss": 0.755, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9352226720647774, | |
| "grad_norm": 0.7573784301242199, | |
| "learning_rate": 3.909150300975178e-05, | |
| "loss": 0.7472, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9392712550607287, | |
| "grad_norm": 0.7354136194969779, | |
| "learning_rate": 3.907457519765642e-05, | |
| "loss": 0.7513, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9433198380566802, | |
| "grad_norm": 0.7524944218584617, | |
| "learning_rate": 3.905749486576607e-05, | |
| "loss": 0.7591, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.7361416386096875, | |
| "learning_rate": 3.90402621506546e-05, | |
| "loss": 0.739, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.951417004048583, | |
| "grad_norm": 0.6737290751914601, | |
| "learning_rate": 3.902287719011435e-05, | |
| "loss": 0.7357, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9554655870445344, | |
| "grad_norm": 0.639730387931973, | |
| "learning_rate": 3.900534012315499e-05, | |
| "loss": 0.7469, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9595141700404858, | |
| "grad_norm": 0.683894976229601, | |
| "learning_rate": 3.8987651090002454e-05, | |
| "loss": 0.7448, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9635627530364372, | |
| "grad_norm": 0.6927706681783905, | |
| "learning_rate": 3.896981023209777e-05, | |
| "loss": 0.7341, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9676113360323887, | |
| "grad_norm": 0.7419206804032332, | |
| "learning_rate": 3.8951817692095966e-05, | |
| "loss": 0.7304, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.97165991902834, | |
| "grad_norm": 0.7008608943401118, | |
| "learning_rate": 3.893367361386491e-05, | |
| "loss": 0.7272, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9757085020242915, | |
| "grad_norm": 0.7728792013278537, | |
| "learning_rate": 3.891537814248417e-05, | |
| "loss": 0.7551, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.979757085020243, | |
| "grad_norm": 0.6749383570452424, | |
| "learning_rate": 3.889693142424384e-05, | |
| "loss": 0.7348, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9838056680161943, | |
| "grad_norm": 0.708778620326891, | |
| "learning_rate": 3.8878333606643405e-05, | |
| "loss": 0.7425, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9878542510121457, | |
| "grad_norm": 0.7483042922428739, | |
| "learning_rate": 3.885958483839049e-05, | |
| "loss": 0.7486, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9919028340080972, | |
| "grad_norm": 0.6755461028121474, | |
| "learning_rate": 3.8840685269399786e-05, | |
| "loss": 0.7464, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9959514170040485, | |
| "grad_norm": 0.7531516009766981, | |
| "learning_rate": 3.882163505079171e-05, | |
| "loss": 0.7386, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6943031343776864, | |
| "learning_rate": 3.880243433489132e-05, | |
| "loss": 0.7378, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0040485829959513, | |
| "grad_norm": 0.9026101430543176, | |
| "learning_rate": 3.878308327522702e-05, | |
| "loss": 0.5947, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.008097165991903, | |
| "grad_norm": 0.7733572068278654, | |
| "learning_rate": 3.876358202652936e-05, | |
| "loss": 0.5884, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0121457489878543, | |
| "grad_norm": 1.0398952007794906, | |
| "learning_rate": 3.8743930744729835e-05, | |
| "loss": 0.5732, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0161943319838056, | |
| "grad_norm": 0.8391396922904131, | |
| "learning_rate": 3.8724129586959535e-05, | |
| "loss": 0.5701, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0202429149797572, | |
| "grad_norm": 0.8341568612234538, | |
| "learning_rate": 3.870417871154801e-05, | |
| "loss": 0.5554, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0242914979757085, | |
| "grad_norm": 0.8648052982378912, | |
| "learning_rate": 3.868407827802192e-05, | |
| "loss": 0.5527, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0283400809716599, | |
| "grad_norm": 0.805091716165642, | |
| "learning_rate": 3.8663828447103804e-05, | |
| "loss": 0.5613, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.0323886639676114, | |
| "grad_norm": 0.9054910137596444, | |
| "learning_rate": 3.8643429380710765e-05, | |
| "loss": 0.5641, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0364372469635628, | |
| "grad_norm": 0.9445651051679907, | |
| "learning_rate": 3.862288124195319e-05, | |
| "loss": 0.5763, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.040485829959514, | |
| "grad_norm": 0.8501572004021589, | |
| "learning_rate": 3.8602184195133464e-05, | |
| "loss": 0.5861, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0445344129554657, | |
| "grad_norm": 0.8611590271248558, | |
| "learning_rate": 3.8581338405744604e-05, | |
| "loss": 0.5519, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.048582995951417, | |
| "grad_norm": 0.8087742679123083, | |
| "learning_rate": 3.8560344040468996e-05, | |
| "loss": 0.5757, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.7796486689039578, | |
| "learning_rate": 3.853920126717703e-05, | |
| "loss": 0.5731, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.05668016194332, | |
| "grad_norm": 0.7570386758304427, | |
| "learning_rate": 3.851791025492574e-05, | |
| "loss": 0.572, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0607287449392713, | |
| "grad_norm": 0.8424663381937973, | |
| "learning_rate": 3.8496471173957515e-05, | |
| "loss": 0.5628, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0647773279352226, | |
| "grad_norm": 0.7343657162634528, | |
| "learning_rate": 3.847488419569866e-05, | |
| "loss": 0.5604, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0688259109311742, | |
| "grad_norm": 0.7573343789644214, | |
| "learning_rate": 3.8453149492758077e-05, | |
| "loss": 0.5655, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0728744939271255, | |
| "grad_norm": 0.7741573138963245, | |
| "learning_rate": 3.8431267238925886e-05, | |
| "loss": 0.5643, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.6572765886679524, | |
| "learning_rate": 3.840923760917199e-05, | |
| "loss": 0.5605, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0809716599190284, | |
| "grad_norm": 0.7910186492476671, | |
| "learning_rate": 3.838706077964473e-05, | |
| "loss": 0.5542, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0850202429149798, | |
| "grad_norm": 0.6736039436420589, | |
| "learning_rate": 3.836473692766944e-05, | |
| "loss": 0.5471, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0890688259109311, | |
| "grad_norm": 0.8413493289801188, | |
| "learning_rate": 3.834226623174705e-05, | |
| "loss": 0.564, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0931174089068827, | |
| "grad_norm": 0.8579068832410611, | |
| "learning_rate": 3.831964887155264e-05, | |
| "loss": 0.5608, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.097165991902834, | |
| "grad_norm": 0.6797977400170663, | |
| "learning_rate": 3.829688502793403e-05, | |
| "loss": 0.5439, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.1012145748987854, | |
| "grad_norm": 0.8418093399634271, | |
| "learning_rate": 3.827397488291029e-05, | |
| "loss": 0.5692, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.715864269090179, | |
| "learning_rate": 3.8250918619670343e-05, | |
| "loss": 0.5482, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.1093117408906883, | |
| "grad_norm": 0.7537175464746101, | |
| "learning_rate": 3.822771642257145e-05, | |
| "loss": 0.5664, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1133603238866396, | |
| "grad_norm": 0.7873483858857175, | |
| "learning_rate": 3.820436847713776e-05, | |
| "loss": 0.5571, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.117408906882591, | |
| "grad_norm": 0.6863020867338843, | |
| "learning_rate": 3.8180874970058826e-05, | |
| "loss": 0.5608, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.1214574898785425, | |
| "grad_norm": 0.7519639451044865, | |
| "learning_rate": 3.8157236089188103e-05, | |
| "loss": 0.5576, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.125506072874494, | |
| "grad_norm": 0.7431282976639487, | |
| "learning_rate": 3.813345202354145e-05, | |
| "loss": 0.5683, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1295546558704452, | |
| "grad_norm": 0.7441021873150574, | |
| "learning_rate": 3.810952296329563e-05, | |
| "loss": 0.5678, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1336032388663968, | |
| "grad_norm": 0.6930231257646605, | |
| "learning_rate": 3.808544909978676e-05, | |
| "loss": 0.5467, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1376518218623481, | |
| "grad_norm": 0.7572857499438642, | |
| "learning_rate": 3.806123062550882e-05, | |
| "loss": 0.5654, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1417004048582995, | |
| "grad_norm": 0.685020463878595, | |
| "learning_rate": 3.803686773411208e-05, | |
| "loss": 0.5576, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.145748987854251, | |
| "grad_norm": 0.7186891025455111, | |
| "learning_rate": 3.801236062040158e-05, | |
| "loss": 0.5633, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1497975708502024, | |
| "grad_norm": 0.6909511146720823, | |
| "learning_rate": 3.798770948033553e-05, | |
| "loss": 0.5411, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.7036286134363424, | |
| "learning_rate": 3.796291451102381e-05, | |
| "loss": 0.5668, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.6634060208790806, | |
| "learning_rate": 3.793797591072631e-05, | |
| "loss": 0.5448, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1619433198380567, | |
| "grad_norm": 0.7450799831274377, | |
| "learning_rate": 3.791289387885144e-05, | |
| "loss": 0.5857, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.165991902834008, | |
| "grad_norm": 0.6393279773964702, | |
| "learning_rate": 3.788766861595443e-05, | |
| "loss": 0.5584, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1700404858299596, | |
| "grad_norm": 0.6751760334139698, | |
| "learning_rate": 3.7862300323735835e-05, | |
| "loss": 0.5715, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.174089068825911, | |
| "grad_norm": 0.6410637358369029, | |
| "learning_rate": 3.783678920503983e-05, | |
| "loss": 0.5637, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1781376518218623, | |
| "grad_norm": 0.6730347687268685, | |
| "learning_rate": 3.781113546385265e-05, | |
| "loss": 0.5561, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1821862348178138, | |
| "grad_norm": 0.765074962359393, | |
| "learning_rate": 3.7785339305300916e-05, | |
| "loss": 0.5803, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.1862348178137652, | |
| "grad_norm": 0.729968518635501, | |
| "learning_rate": 3.7759400935650025e-05, | |
| "loss": 0.5778, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1902834008097165, | |
| "grad_norm": 0.788204508396452, | |
| "learning_rate": 3.77333205623025e-05, | |
| "loss": 0.5652, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.194331983805668, | |
| "grad_norm": 0.679627186497987, | |
| "learning_rate": 3.7707098393796295e-05, | |
| "loss": 0.5666, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1983805668016194, | |
| "grad_norm": 0.7253528362246822, | |
| "learning_rate": 3.7680734639803185e-05, | |
| "loss": 0.5606, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2024291497975708, | |
| "grad_norm": 0.8819192724082157, | |
| "learning_rate": 3.765422951112704e-05, | |
| "loss": 0.5663, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2064777327935223, | |
| "grad_norm": 0.6953151328716948, | |
| "learning_rate": 3.762758321970216e-05, | |
| "loss": 0.5521, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.7431692394171407, | |
| "learning_rate": 3.7600795978591584e-05, | |
| "loss": 0.5632, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.214574898785425, | |
| "grad_norm": 0.6867143487987581, | |
| "learning_rate": 3.757386800198538e-05, | |
| "loss": 0.5829, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2186234817813766, | |
| "grad_norm": 0.6898398649877547, | |
| "learning_rate": 3.7546799505198925e-05, | |
| "loss": 0.5635, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.222672064777328, | |
| "grad_norm": 0.7085210940474705, | |
| "learning_rate": 3.7519590704671197e-05, | |
| "loss": 0.589, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.2267206477732793, | |
| "grad_norm": 0.7132473618425982, | |
| "learning_rate": 3.749224181796305e-05, | |
| "loss": 0.5623, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.6787475853216298, | |
| "learning_rate": 3.746475306375545e-05, | |
| "loss": 0.5933, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2348178137651822, | |
| "grad_norm": 0.7172578761061179, | |
| "learning_rate": 3.743712466184774e-05, | |
| "loss": 0.5576, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2388663967611335, | |
| "grad_norm": 0.6662731434920639, | |
| "learning_rate": 3.7409356833155885e-05, | |
| "loss": 0.5652, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.242914979757085, | |
| "grad_norm": 0.7473878887581146, | |
| "learning_rate": 3.73814497997107e-05, | |
| "loss": 0.5802, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2469635627530364, | |
| "grad_norm": 0.6927347641374217, | |
| "learning_rate": 3.7353403784656095e-05, | |
| "loss": 0.5599, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2510121457489878, | |
| "grad_norm": 0.7970091233487554, | |
| "learning_rate": 3.732521901224724e-05, | |
| "loss": 0.5579, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.2550607287449393, | |
| "grad_norm": 0.6462805534754538, | |
| "learning_rate": 3.7296895707848844e-05, | |
| "loss": 0.5677, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2591093117408907, | |
| "grad_norm": 0.8601521806897308, | |
| "learning_rate": 3.7268434097933275e-05, | |
| "loss": 0.5575, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.6753459772820677, | |
| "learning_rate": 3.723983441007882e-05, | |
| "loss": 0.5554, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2672064777327936, | |
| "grad_norm": 0.8512281779126234, | |
| "learning_rate": 3.721109687296781e-05, | |
| "loss": 0.5666, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.271255060728745, | |
| "grad_norm": 0.6594469903707109, | |
| "learning_rate": 3.718222171638484e-05, | |
| "loss": 0.5704, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2753036437246963, | |
| "grad_norm": 0.7432894397655959, | |
| "learning_rate": 3.715320917121488e-05, | |
| "loss": 0.5727, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2793522267206479, | |
| "grad_norm": 0.6331153639823924, | |
| "learning_rate": 3.712405946944149e-05, | |
| "loss": 0.5832, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2834008097165992, | |
| "grad_norm": 0.7463092459364726, | |
| "learning_rate": 3.709477284414488e-05, | |
| "loss": 0.5796, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.2874493927125505, | |
| "grad_norm": 0.6602846177683184, | |
| "learning_rate": 3.7065349529500136e-05, | |
| "loss": 0.5568, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.291497975708502, | |
| "grad_norm": 0.6921117429214206, | |
| "learning_rate": 3.7035789760775305e-05, | |
| "loss": 0.5771, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2955465587044535, | |
| "grad_norm": 0.6797853710985479, | |
| "learning_rate": 3.70060937743295e-05, | |
| "loss": 0.5718, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2995951417004048, | |
| "grad_norm": 0.7636191987484189, | |
| "learning_rate": 3.6976261807611045e-05, | |
| "loss": 0.5691, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.3036437246963564, | |
| "grad_norm": 0.7227138677728605, | |
| "learning_rate": 3.694629409915555e-05, | |
| "loss": 0.5694, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.7509985483602789, | |
| "learning_rate": 3.691619088858401e-05, | |
| "loss": 0.5802, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.311740890688259, | |
| "grad_norm": 0.7375536810735944, | |
| "learning_rate": 3.68859524166009e-05, | |
| "loss": 0.5655, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.6827982181324351, | |
| "learning_rate": 3.6855578924992226e-05, | |
| "loss": 0.5658, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.319838056680162, | |
| "grad_norm": 0.7669796557435152, | |
| "learning_rate": 3.6825070656623626e-05, | |
| "loss": 0.5741, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.3238866396761133, | |
| "grad_norm": 0.813834171717492, | |
| "learning_rate": 3.67944278554384e-05, | |
| "loss": 0.5661, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.3279352226720649, | |
| "grad_norm": 0.6678244184195242, | |
| "learning_rate": 3.676365076645557e-05, | |
| "loss": 0.5632, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.3319838056680162, | |
| "grad_norm": 0.8478858069107125, | |
| "learning_rate": 3.673273963576791e-05, | |
| "loss": 0.5694, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3360323886639676, | |
| "grad_norm": 0.6482957622073146, | |
| "learning_rate": 3.670169471054001e-05, | |
| "loss": 0.5665, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.3400809716599191, | |
| "grad_norm": 0.7897711347933256, | |
| "learning_rate": 3.6670516239006254e-05, | |
| "loss": 0.5687, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3441295546558705, | |
| "grad_norm": 0.8418073833152356, | |
| "learning_rate": 3.663920447046888e-05, | |
| "loss": 0.5596, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.3481781376518218, | |
| "grad_norm": 0.7213998889938051, | |
| "learning_rate": 3.6607759655295954e-05, | |
| "loss": 0.5581, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3522267206477734, | |
| "grad_norm": 0.744045933876794, | |
| "learning_rate": 3.657618204491937e-05, | |
| "loss": 0.557, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3562753036437247, | |
| "grad_norm": 0.738461435264195, | |
| "learning_rate": 3.6544471891832875e-05, | |
| "loss": 0.565, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.360323886639676, | |
| "grad_norm": 0.6805994454380724, | |
| "learning_rate": 3.6512629449589983e-05, | |
| "loss": 0.5792, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3643724696356276, | |
| "grad_norm": 0.7517991161524344, | |
| "learning_rate": 3.6480654972802045e-05, | |
| "loss": 0.5747, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.6549177110508912, | |
| "learning_rate": 3.644854871713611e-05, | |
| "loss": 0.5893, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3724696356275303, | |
| "grad_norm": 0.6909237364557033, | |
| "learning_rate": 3.6416310939312924e-05, | |
| "loss": 0.5494, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.376518218623482, | |
| "grad_norm": 0.6504014669305204, | |
| "learning_rate": 3.638394189710493e-05, | |
| "loss": 0.5674, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3805668016194332, | |
| "grad_norm": 0.780643719923325, | |
| "learning_rate": 3.635144184933412e-05, | |
| "loss": 0.5781, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.68109229688528, | |
| "learning_rate": 3.6318811055869994e-05, | |
| "loss": 0.5603, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3886639676113361, | |
| "grad_norm": 0.6917552206967442, | |
| "learning_rate": 3.628604977762752e-05, | |
| "loss": 0.5733, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.3927125506072875, | |
| "grad_norm": 0.6413587447319985, | |
| "learning_rate": 3.6253158276565004e-05, | |
| "loss": 0.5731, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3967611336032388, | |
| "grad_norm": 0.7332151848909559, | |
| "learning_rate": 3.6220136815682016e-05, | |
| "loss": 0.5745, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.4008097165991904, | |
| "grad_norm": 0.6307557287443971, | |
| "learning_rate": 3.618698565901727e-05, | |
| "loss": 0.5919, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.4048582995951417, | |
| "grad_norm": 0.7164835909421181, | |
| "learning_rate": 3.615370507164654e-05, | |
| "loss": 0.5788, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.408906882591093, | |
| "grad_norm": 0.644032162014878, | |
| "learning_rate": 3.61202953196805e-05, | |
| "loss": 0.5682, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.4129554655870447, | |
| "grad_norm": 0.7384179089707573, | |
| "learning_rate": 3.6086756670262654e-05, | |
| "loss": 0.5641, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.417004048582996, | |
| "grad_norm": 0.6981151298432864, | |
| "learning_rate": 3.6053089391567135e-05, | |
| "loss": 0.5794, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.7353705328526747, | |
| "learning_rate": 3.601929375279662e-05, | |
| "loss": 0.5777, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.425101214574899, | |
| "grad_norm": 0.6706224649380867, | |
| "learning_rate": 3.598537002418012e-05, | |
| "loss": 0.5713, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.4291497975708503, | |
| "grad_norm": 0.7696464128682874, | |
| "learning_rate": 3.5951318476970885e-05, | |
| "loss": 0.5645, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.4331983805668016, | |
| "grad_norm": 0.618837135836642, | |
| "learning_rate": 3.591713938344416e-05, | |
| "loss": 0.5535, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4372469635627532, | |
| "grad_norm": 0.8617885444749871, | |
| "learning_rate": 3.588283301689507e-05, | |
| "loss": 0.5704, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4412955465587045, | |
| "grad_norm": 0.7546707360775482, | |
| "learning_rate": 3.5848399651636424e-05, | |
| "loss": 0.5662, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4453441295546559, | |
| "grad_norm": 0.6538437150874089, | |
| "learning_rate": 3.5813839562996473e-05, | |
| "loss": 0.5609, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4493927125506074, | |
| "grad_norm": 0.7597445953393107, | |
| "learning_rate": 3.5779153027316775e-05, | |
| "loss": 0.5788, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4534412955465588, | |
| "grad_norm": 0.6486237881356858, | |
| "learning_rate": 3.574434032194994e-05, | |
| "loss": 0.5644, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.45748987854251, | |
| "grad_norm": 0.6628944881030545, | |
| "learning_rate": 3.5709401725257436e-05, | |
| "loss": 0.5741, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.6565629691318268, | |
| "learning_rate": 3.5674337516607346e-05, | |
| "loss": 0.5794, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.465587044534413, | |
| "grad_norm": 0.6788190008000546, | |
| "learning_rate": 3.563914797637216e-05, | |
| "loss": 0.5712, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.4696356275303644, | |
| "grad_norm": 0.6619529113035857, | |
| "learning_rate": 3.560383338592649e-05, | |
| "loss": 0.5703, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.6516194739711656, | |
| "learning_rate": 3.556839402764487e-05, | |
| "loss": 0.5722, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4777327935222673, | |
| "grad_norm": 0.7271673698057437, | |
| "learning_rate": 3.553283018489946e-05, | |
| "loss": 0.5627, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4817813765182186, | |
| "grad_norm": 0.6159028724293308, | |
| "learning_rate": 3.54971421420578e-05, | |
| "loss": 0.5658, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.48582995951417, | |
| "grad_norm": 0.6678081440440446, | |
| "learning_rate": 3.5461330184480524e-05, | |
| "loss": 0.5578, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4898785425101215, | |
| "grad_norm": 0.6692156075416775, | |
| "learning_rate": 3.54253945985191e-05, | |
| "loss": 0.5652, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4939271255060729, | |
| "grad_norm": 0.6862517071427146, | |
| "learning_rate": 3.5389335671513524e-05, | |
| "loss": 0.5728, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4979757085020242, | |
| "grad_norm": 0.6976054157854464, | |
| "learning_rate": 3.5353153691789986e-05, | |
| "loss": 0.5694, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.5020242914979756, | |
| "grad_norm": 0.6559517965424679, | |
| "learning_rate": 3.531684894865866e-05, | |
| "loss": 0.5536, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.5060728744939271, | |
| "grad_norm": 0.7510974425157274, | |
| "learning_rate": 3.528042173241128e-05, | |
| "loss": 0.5795, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.5101214574898787, | |
| "grad_norm": 0.6839396213201548, | |
| "learning_rate": 3.524387233431892e-05, | |
| "loss": 0.5669, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.5141700404858298, | |
| "grad_norm": 0.7420526817279469, | |
| "learning_rate": 3.520720104662958e-05, | |
| "loss": 0.5677, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5182186234817814, | |
| "grad_norm": 0.6545326506481609, | |
| "learning_rate": 3.5170408162565904e-05, | |
| "loss": 0.5729, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.522267206477733, | |
| "grad_norm": 0.6982172764965742, | |
| "learning_rate": 3.513349397632281e-05, | |
| "loss": 0.5571, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.5955365819265078, | |
| "learning_rate": 3.5096458783065145e-05, | |
| "loss": 0.571, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.5303643724696356, | |
| "grad_norm": 0.7264417494166947, | |
| "learning_rate": 3.505930287892533e-05, | |
| "loss": 0.5666, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.5344129554655872, | |
| "grad_norm": 0.6131648245693184, | |
| "learning_rate": 3.502202656100099e-05, | |
| "loss": 0.5697, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.6328817916283279, | |
| "learning_rate": 3.498463012735256e-05, | |
| "loss": 0.5739, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.54251012145749, | |
| "grad_norm": 0.6782069947447884, | |
| "learning_rate": 3.494711387700094e-05, | |
| "loss": 0.5507, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5465587044534415, | |
| "grad_norm": 0.6353464586387195, | |
| "learning_rate": 3.490947810992505e-05, | |
| "loss": 0.5678, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5506072874493926, | |
| "grad_norm": 0.6468559070599997, | |
| "learning_rate": 3.48717231270595e-05, | |
| "loss": 0.5737, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5546558704453441, | |
| "grad_norm": 0.7076028746429917, | |
| "learning_rate": 3.483384923029211e-05, | |
| "loss": 0.5885, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5587044534412957, | |
| "grad_norm": 0.6484575568511354, | |
| "learning_rate": 3.479585672246156e-05, | |
| "loss": 0.5875, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5627530364372468, | |
| "grad_norm": 0.6595027825235992, | |
| "learning_rate": 3.4757745907354924e-05, | |
| "loss": 0.5667, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5668016194331984, | |
| "grad_norm": 0.6098067830938682, | |
| "learning_rate": 3.471951708970527e-05, | |
| "loss": 0.5478, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.5708502024291497, | |
| "grad_norm": 0.6682694624569763, | |
| "learning_rate": 3.468117057518921e-05, | |
| "loss": 0.562, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.574898785425101, | |
| "grad_norm": 0.6697924628669244, | |
| "learning_rate": 3.4642706670424426e-05, | |
| "loss": 0.5664, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.6756559938048056, | |
| "learning_rate": 3.460412568296731e-05, | |
| "loss": 0.5794, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.582995951417004, | |
| "grad_norm": 0.7378694258226965, | |
| "learning_rate": 3.456542792131039e-05, | |
| "loss": 0.5683, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5870445344129553, | |
| "grad_norm": 0.6696701600667049, | |
| "learning_rate": 3.4526613694879936e-05, | |
| "loss": 0.5757, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.591093117408907, | |
| "grad_norm": 0.7262421972406361, | |
| "learning_rate": 3.448768331403348e-05, | |
| "loss": 0.5634, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5951417004048583, | |
| "grad_norm": 0.6403501478582839, | |
| "learning_rate": 3.44486370900573e-05, | |
| "loss": 0.5629, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5991902834008096, | |
| "grad_norm": 0.7475046946380149, | |
| "learning_rate": 3.440947533516396e-05, | |
| "loss": 0.5531, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.6032388663967612, | |
| "grad_norm": 0.7012401353195518, | |
| "learning_rate": 3.4370198362489816e-05, | |
| "loss": 0.5873, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.6072874493927125, | |
| "grad_norm": 0.6671329587202336, | |
| "learning_rate": 3.43308064860925e-05, | |
| "loss": 0.5788, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.6113360323886639, | |
| "grad_norm": 0.7478317145445765, | |
| "learning_rate": 3.4291300020948393e-05, | |
| "loss": 0.5731, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.7028356767970474, | |
| "learning_rate": 3.425167928295015e-05, | |
| "loss": 0.5648, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6194331983805668, | |
| "grad_norm": 0.646164312808452, | |
| "learning_rate": 3.421194458890411e-05, | |
| "loss": 0.5772, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.623481781376518, | |
| "grad_norm": 0.6962185524550453, | |
| "learning_rate": 3.4172096256527865e-05, | |
| "loss": 0.5769, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.6275303643724697, | |
| "grad_norm": 0.7332780298792553, | |
| "learning_rate": 3.4132134604447595e-05, | |
| "loss": 0.5743, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.6385840086895984, | |
| "learning_rate": 3.409205995219561e-05, | |
| "loss": 0.5666, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.6356275303643724, | |
| "grad_norm": 0.6586275709326784, | |
| "learning_rate": 3.4051872620207765e-05, | |
| "loss": 0.5759, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.639676113360324, | |
| "grad_norm": 0.6144531086719933, | |
| "learning_rate": 3.40115729298209e-05, | |
| "loss": 0.5694, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6437246963562753, | |
| "grad_norm": 0.6420413700259122, | |
| "learning_rate": 3.397116120327025e-05, | |
| "loss": 0.5762, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6477732793522266, | |
| "grad_norm": 0.6513536444925546, | |
| "learning_rate": 3.3930637763686916e-05, | |
| "loss": 0.5753, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6518218623481782, | |
| "grad_norm": 0.6341356477260882, | |
| "learning_rate": 3.389000293509524e-05, | |
| "loss": 0.5524, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6558704453441295, | |
| "grad_norm": 0.6452585520041425, | |
| "learning_rate": 3.3849257042410225e-05, | |
| "loss": 0.5722, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6599190283400809, | |
| "grad_norm": 0.6311831764422249, | |
| "learning_rate": 3.380840041143494e-05, | |
| "loss": 0.5559, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6639676113360324, | |
| "grad_norm": 0.6515612111151772, | |
| "learning_rate": 3.376743336885792e-05, | |
| "loss": 0.5844, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6680161943319838, | |
| "grad_norm": 0.6162609329784474, | |
| "learning_rate": 3.372635624225055e-05, | |
| "loss": 0.565, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6720647773279351, | |
| "grad_norm": 0.6436253598533966, | |
| "learning_rate": 3.3685169360064445e-05, | |
| "loss": 0.5705, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6761133603238867, | |
| "grad_norm": 0.6076329473959148, | |
| "learning_rate": 3.3643873051628804e-05, | |
| "loss": 0.5781, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.680161943319838, | |
| "grad_norm": 0.6760087865520652, | |
| "learning_rate": 3.3602467647147814e-05, | |
| "loss": 0.5833, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.6580370177409874, | |
| "learning_rate": 3.356095347769799e-05, | |
| "loss": 0.5638, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.688259109311741, | |
| "grad_norm": 0.6434363064073096, | |
| "learning_rate": 3.351933087522552e-05, | |
| "loss": 0.5718, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.6390770983845615, | |
| "learning_rate": 3.3477600172543626e-05, | |
| "loss": 0.5617, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6963562753036436, | |
| "grad_norm": 0.6424653840323984, | |
| "learning_rate": 3.3435761703329894e-05, | |
| "loss": 0.5847, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.7004048582995952, | |
| "grad_norm": 0.6260391706390812, | |
| "learning_rate": 3.3393815802123606e-05, | |
| "loss": 0.5549, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.7044534412955465, | |
| "grad_norm": 0.6424895032310081, | |
| "learning_rate": 3.335176280432308e-05, | |
| "loss": 0.5604, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.708502024291498, | |
| "grad_norm": 0.6621124860671304, | |
| "learning_rate": 3.330960304618295e-05, | |
| "loss": 0.5731, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.7125506072874495, | |
| "grad_norm": 0.6202568403195762, | |
| "learning_rate": 3.326733686481153e-05, | |
| "loss": 0.5663, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.7165991902834008, | |
| "grad_norm": 0.7073549734109129, | |
| "learning_rate": 3.3224964598168084e-05, | |
| "loss": 0.5643, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7206477732793521, | |
| "grad_norm": 0.6294346563103701, | |
| "learning_rate": 3.318248658506012e-05, | |
| "loss": 0.5728, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.7246963562753037, | |
| "grad_norm": 0.6822467165412236, | |
| "learning_rate": 3.313990316514069e-05, | |
| "loss": 0.5652, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.728744939271255, | |
| "grad_norm": 0.6772885130478778, | |
| "learning_rate": 3.309721467890571e-05, | |
| "loss": 0.5766, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.7327935222672064, | |
| "grad_norm": 0.6136822596370768, | |
| "learning_rate": 3.305442146769115e-05, | |
| "loss": 0.5767, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.6748953227507035, | |
| "learning_rate": 3.30115238736704e-05, | |
| "loss": 0.5674, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.7408906882591093, | |
| "grad_norm": 0.6196891929493058, | |
| "learning_rate": 3.296852223985148e-05, | |
| "loss": 0.5529, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7449392712550607, | |
| "grad_norm": 0.6624106954336721, | |
| "learning_rate": 3.2925416910074315e-05, | |
| "loss": 0.5781, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7489878542510122, | |
| "grad_norm": 0.599505731652241, | |
| "learning_rate": 3.288220822900796e-05, | |
| "loss": 0.5611, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7530364372469636, | |
| "grad_norm": 0.6775027635791951, | |
| "learning_rate": 3.283889654214788e-05, | |
| "loss": 0.552, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.757085020242915, | |
| "grad_norm": 0.6141835758539164, | |
| "learning_rate": 3.279548219581317e-05, | |
| "loss": 0.5584, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7611336032388665, | |
| "grad_norm": 0.6408910962989158, | |
| "learning_rate": 3.275196553714379e-05, | |
| "loss": 0.5619, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7651821862348178, | |
| "grad_norm": 0.6416408649572236, | |
| "learning_rate": 3.270834691409778e-05, | |
| "loss": 0.574, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.6276364278217696, | |
| "learning_rate": 3.266462667544849e-05, | |
| "loss": 0.5699, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7732793522267207, | |
| "grad_norm": 0.5999891347041133, | |
| "learning_rate": 3.262080517078178e-05, | |
| "loss": 0.5719, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.777327935222672, | |
| "grad_norm": 0.6265428249837802, | |
| "learning_rate": 3.257688275049323e-05, | |
| "loss": 0.5714, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7813765182186234, | |
| "grad_norm": 0.632234300414542, | |
| "learning_rate": 3.253285976578535e-05, | |
| "loss": 0.584, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.785425101214575, | |
| "grad_norm": 0.6120494243592122, | |
| "learning_rate": 3.2488736568664756e-05, | |
| "loss": 0.5538, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.6798718819918832, | |
| "learning_rate": 3.244451351193935e-05, | |
| "loss": 0.5747, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7935222672064777, | |
| "grad_norm": 0.6154683343430098, | |
| "learning_rate": 3.240019094921551e-05, | |
| "loss": 0.5627, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7975708502024292, | |
| "grad_norm": 0.660814847559166, | |
| "learning_rate": 3.235576923489527e-05, | |
| "loss": 0.5558, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.8016194331983806, | |
| "grad_norm": 0.5953419854829365, | |
| "learning_rate": 3.2311248724173476e-05, | |
| "loss": 0.561, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.805668016194332, | |
| "grad_norm": 0.710234348967718, | |
| "learning_rate": 3.226662977303494e-05, | |
| "loss": 0.5793, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.8097165991902835, | |
| "grad_norm": 0.6716553777368264, | |
| "learning_rate": 3.2221912738251626e-05, | |
| "loss": 0.5732, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.8137651821862348, | |
| "grad_norm": 0.6863062272064133, | |
| "learning_rate": 3.2177097977379736e-05, | |
| "loss": 0.5748, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.8178137651821862, | |
| "grad_norm": 0.6680683796001475, | |
| "learning_rate": 3.21321858487569e-05, | |
| "loss": 0.5655, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8218623481781377, | |
| "grad_norm": 0.7182472797453232, | |
| "learning_rate": 3.208717671149931e-05, | |
| "loss": 0.5507, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.825910931174089, | |
| "grad_norm": 0.6563951016175841, | |
| "learning_rate": 3.204207092549882e-05, | |
| "loss": 0.5712, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.8299595141700404, | |
| "grad_norm": 0.6300779236112205, | |
| "learning_rate": 3.19968688514201e-05, | |
| "loss": 0.5609, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.834008097165992, | |
| "grad_norm": 0.7141447700165718, | |
| "learning_rate": 3.195157085069774e-05, | |
| "loss": 0.57, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8380566801619433, | |
| "grad_norm": 0.6221614796011921, | |
| "learning_rate": 3.190617728553332e-05, | |
| "loss": 0.5769, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.7059037648981354, | |
| "learning_rate": 3.1860688518892606e-05, | |
| "loss": 0.5655, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.6600345789060437, | |
| "learning_rate": 3.181510491450254e-05, | |
| "loss": 0.5641, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8502024291497976, | |
| "grad_norm": 0.6511355258389494, | |
| "learning_rate": 3.176942683684842e-05, | |
| "loss": 0.5762, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.854251012145749, | |
| "grad_norm": 0.6431687694949807, | |
| "learning_rate": 3.1723654651170934e-05, | |
| "loss": 0.5652, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8582995951417005, | |
| "grad_norm": 0.6063993068464859, | |
| "learning_rate": 3.167778872346327e-05, | |
| "loss": 0.5634, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8623481781376519, | |
| "grad_norm": 0.648297902703422, | |
| "learning_rate": 3.163182942046815e-05, | |
| "loss": 0.5764, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8663967611336032, | |
| "grad_norm": 0.6160583079513099, | |
| "learning_rate": 3.158577710967494e-05, | |
| "loss": 0.5505, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8704453441295548, | |
| "grad_norm": 0.6343615400302368, | |
| "learning_rate": 3.153963215931669e-05, | |
| "loss": 0.5592, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.874493927125506, | |
| "grad_norm": 0.5899272809619991, | |
| "learning_rate": 3.14933949383672e-05, | |
| "loss": 0.571, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8785425101214575, | |
| "grad_norm": 0.6268853440117061, | |
| "learning_rate": 3.144706581653805e-05, | |
| "loss": 0.5592, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.882591093117409, | |
| "grad_norm": 0.5890126388900635, | |
| "learning_rate": 3.140064516427566e-05, | |
| "loss": 0.5681, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8866396761133604, | |
| "grad_norm": 0.6803983950740835, | |
| "learning_rate": 3.135413335275833e-05, | |
| "loss": 0.5768, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8906882591093117, | |
| "grad_norm": 0.6206219443830828, | |
| "learning_rate": 3.130753075389327e-05, | |
| "loss": 0.5588, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.6427344116487448, | |
| "learning_rate": 3.1260837740313625e-05, | |
| "loss": 0.5672, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8987854251012146, | |
| "grad_norm": 0.6593711904465325, | |
| "learning_rate": 3.121405468537547e-05, | |
| "loss": 0.5613, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.902834008097166, | |
| "grad_norm": 0.6282677744037833, | |
| "learning_rate": 3.116718196315488e-05, | |
| "loss": 0.5424, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.9068825910931175, | |
| "grad_norm": 0.6741937086428942, | |
| "learning_rate": 3.1120219948444894e-05, | |
| "loss": 0.5647, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.9109311740890689, | |
| "grad_norm": 0.6720879887538889, | |
| "learning_rate": 3.107316901675252e-05, | |
| "loss": 0.5674, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.9149797570850202, | |
| "grad_norm": 0.6706048309401765, | |
| "learning_rate": 3.1026029544295745e-05, | |
| "loss": 0.5577, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.9190283400809718, | |
| "grad_norm": 0.6729094150423004, | |
| "learning_rate": 3.097880190800057e-05, | |
| "loss": 0.5697, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.6072964905372156, | |
| "learning_rate": 3.093148648549788e-05, | |
| "loss": 0.561, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.9271255060728745, | |
| "grad_norm": 0.7520313397378537, | |
| "learning_rate": 3.088408365512055e-05, | |
| "loss": 0.5741, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.931174089068826, | |
| "grad_norm": 0.645583411349521, | |
| "learning_rate": 3.083659379590034e-05, | |
| "loss": 0.5814, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.9352226720647774, | |
| "grad_norm": 0.6736465178247456, | |
| "learning_rate": 3.0789017287564913e-05, | |
| "loss": 0.5822, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.9392712550607287, | |
| "grad_norm": 0.6421969392434398, | |
| "learning_rate": 3.074135451053475e-05, | |
| "loss": 0.5529, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.9433198380566803, | |
| "grad_norm": 0.5956593414153253, | |
| "learning_rate": 3.069360584592015e-05, | |
| "loss": 0.5655, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.6177814050367759, | |
| "learning_rate": 3.064577167551816e-05, | |
| "loss": 0.5597, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.951417004048583, | |
| "grad_norm": 0.6450907033559564, | |
| "learning_rate": 3.059785238180953e-05, | |
| "loss": 0.576, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9554655870445345, | |
| "grad_norm": 0.5789890096893527, | |
| "learning_rate": 3.054984834795566e-05, | |
| "loss": 0.5464, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9595141700404857, | |
| "grad_norm": 0.633566041177122, | |
| "learning_rate": 3.0501759957795537e-05, | |
| "loss": 0.5559, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.9635627530364372, | |
| "grad_norm": 0.5990017128238597, | |
| "learning_rate": 3.0453587595842627e-05, | |
| "loss": 0.5625, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9676113360323888, | |
| "grad_norm": 0.6407432257524849, | |
| "learning_rate": 3.0405331647281856e-05, | |
| "loss": 0.5706, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.97165991902834, | |
| "grad_norm": 0.566403065326569, | |
| "learning_rate": 3.0356992497966508e-05, | |
| "loss": 0.5502, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9757085020242915, | |
| "grad_norm": 0.6003225378309228, | |
| "learning_rate": 3.0308570534415137e-05, | |
| "loss": 0.5426, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.979757085020243, | |
| "grad_norm": 0.6017176887151184, | |
| "learning_rate": 3.0260066143808467e-05, | |
| "loss": 0.5724, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9838056680161942, | |
| "grad_norm": 0.6432474563119132, | |
| "learning_rate": 3.021147971398632e-05, | |
| "loss": 0.5517, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9878542510121457, | |
| "grad_norm": 0.5966361299127056, | |
| "learning_rate": 3.0162811633444478e-05, | |
| "loss": 0.5426, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9919028340080973, | |
| "grad_norm": 0.6335050476237478, | |
| "learning_rate": 3.0114062291331635e-05, | |
| "loss": 0.5709, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9959514170040484, | |
| "grad_norm": 0.6536566628602579, | |
| "learning_rate": 3.0065232077446217e-05, | |
| "loss": 0.5573, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.661822625205828, | |
| "learning_rate": 3.001632138223332e-05, | |
| "loss": 0.5482, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.0040485829959516, | |
| "grad_norm": 1.0391118391952434, | |
| "learning_rate": 2.9967330596781565e-05, | |
| "loss": 0.3492, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0080971659919027, | |
| "grad_norm": 0.6946279990333432, | |
| "learning_rate": 2.9918260112819952e-05, | |
| "loss": 0.3314, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0121457489878543, | |
| "grad_norm": 1.0973346897887462, | |
| "learning_rate": 2.9869110322714778e-05, | |
| "loss": 0.3233, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.016194331983806, | |
| "grad_norm": 0.9635205601462292, | |
| "learning_rate": 2.9819881619466447e-05, | |
| "loss": 0.329, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.020242914979757, | |
| "grad_norm": 0.7555874267840388, | |
| "learning_rate": 2.9770574396706364e-05, | |
| "loss": 0.3253, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0242914979757085, | |
| "grad_norm": 0.6355833025178687, | |
| "learning_rate": 2.972118904869376e-05, | |
| "loss": 0.3171, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02834008097166, | |
| "grad_norm": 0.7938754085734964, | |
| "learning_rate": 2.9671725970312563e-05, | |
| "loss": 0.3255, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.032388663967611, | |
| "grad_norm": 0.6486132862001646, | |
| "learning_rate": 2.9622185557068238e-05, | |
| "loss": 0.3242, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.0364372469635628, | |
| "grad_norm": 0.6539125124331022, | |
| "learning_rate": 2.95725682050846e-05, | |
| "loss": 0.3271, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.0404858299595143, | |
| "grad_norm": 0.6888847822416841, | |
| "learning_rate": 2.9522874311100684e-05, | |
| "loss": 0.3236, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.0445344129554655, | |
| "grad_norm": 0.6520165659087178, | |
| "learning_rate": 2.9473104272467526e-05, | |
| "loss": 0.3341, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.048582995951417, | |
| "grad_norm": 0.657904243679822, | |
| "learning_rate": 2.942325848714505e-05, | |
| "loss": 0.3231, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.6175545863177528, | |
| "learning_rate": 2.9373337353698804e-05, | |
| "loss": 0.3255, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.0566801619433197, | |
| "grad_norm": 0.651099111444449, | |
| "learning_rate": 2.932334127129686e-05, | |
| "loss": 0.3165, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.0607287449392713, | |
| "grad_norm": 0.610691362810703, | |
| "learning_rate": 2.9273270639706547e-05, | |
| "loss": 0.3282, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.064777327935223, | |
| "grad_norm": 0.593757138349224, | |
| "learning_rate": 2.922312585929131e-05, | |
| "loss": 0.3257, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.068825910931174, | |
| "grad_norm": 0.6020368825794592, | |
| "learning_rate": 2.9172907331007478e-05, | |
| "loss": 0.3156, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.0728744939271255, | |
| "grad_norm": 0.5967907279534934, | |
| "learning_rate": 2.9122615456401053e-05, | |
| "loss": 0.3198, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.5974824535948471, | |
| "learning_rate": 2.9072250637604535e-05, | |
| "loss": 0.3182, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.080971659919028, | |
| "grad_norm": 0.57416044139218, | |
| "learning_rate": 2.9021813277333667e-05, | |
| "loss": 0.3127, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.08502024291498, | |
| "grad_norm": 0.62683167408279, | |
| "learning_rate": 2.8971303778884244e-05, | |
| "loss": 0.3195, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.0890688259109313, | |
| "grad_norm": 0.5768988583361194, | |
| "learning_rate": 2.8920722546128863e-05, | |
| "loss": 0.3265, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0931174089068825, | |
| "grad_norm": 0.5963770427427139, | |
| "learning_rate": 2.887006998351372e-05, | |
| "loss": 0.3328, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.097165991902834, | |
| "grad_norm": 0.5804823681195685, | |
| "learning_rate": 2.881934649605535e-05, | |
| "loss": 0.3261, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.1012145748987856, | |
| "grad_norm": 0.6105946747777484, | |
| "learning_rate": 2.8768552489337408e-05, | |
| "loss": 0.3231, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.6242426585115816, | |
| "learning_rate": 2.8717688369507425e-05, | |
| "loss": 0.3227, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.1093117408906883, | |
| "grad_norm": 0.6017369521242725, | |
| "learning_rate": 2.8666754543273535e-05, | |
| "loss": 0.3198, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.11336032388664, | |
| "grad_norm": 0.6115858316714067, | |
| "learning_rate": 2.8615751417901273e-05, | |
| "loss": 0.3116, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.117408906882591, | |
| "grad_norm": 0.6393876659394231, | |
| "learning_rate": 2.856467940121025e-05, | |
| "loss": 0.3288, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.1214574898785425, | |
| "grad_norm": 0.6562738947656499, | |
| "learning_rate": 2.8513538901570964e-05, | |
| "loss": 0.3298, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.125506072874494, | |
| "grad_norm": 0.6357703748532301, | |
| "learning_rate": 2.846233032790149e-05, | |
| "loss": 0.3152, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.1295546558704452, | |
| "grad_norm": 0.6562518121983382, | |
| "learning_rate": 2.84110540896642e-05, | |
| "loss": 0.3324, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.133603238866397, | |
| "grad_norm": 0.6342178023701707, | |
| "learning_rate": 2.8359710596862555e-05, | |
| "loss": 0.3195, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.1376518218623484, | |
| "grad_norm": 0.6110482595724881, | |
| "learning_rate": 2.8308300260037734e-05, | |
| "loss": 0.3143, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.1417004048582995, | |
| "grad_norm": 0.6141293280741167, | |
| "learning_rate": 2.8256823490265424e-05, | |
| "loss": 0.3225, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.145748987854251, | |
| "grad_norm": 0.6464357831249691, | |
| "learning_rate": 2.8205280699152515e-05, | |
| "loss": 0.3345, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.1497975708502026, | |
| "grad_norm": 0.6118271115666903, | |
| "learning_rate": 2.8153672298833777e-05, | |
| "loss": 0.3204, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.649508330278151, | |
| "learning_rate": 2.8101998701968615e-05, | |
| "loss": 0.3215, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.5793578415140994, | |
| "learning_rate": 2.805026032173773e-05, | |
| "loss": 0.3238, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.161943319838057, | |
| "grad_norm": 0.5988297966294568, | |
| "learning_rate": 2.7998457571839832e-05, | |
| "loss": 0.3185, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.165991902834008, | |
| "grad_norm": 0.5990874189109103, | |
| "learning_rate": 2.7946590866488322e-05, | |
| "loss": 0.3319, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.1700404858299596, | |
| "grad_norm": 0.5445525670878916, | |
| "learning_rate": 2.7894660620408e-05, | |
| "loss": 0.3227, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.174089068825911, | |
| "grad_norm": 0.5640812642931451, | |
| "learning_rate": 2.784266724883173e-05, | |
| "loss": 0.3249, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.1781376518218623, | |
| "grad_norm": 0.5784280455373063, | |
| "learning_rate": 2.7790611167497128e-05, | |
| "loss": 0.3286, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.182186234817814, | |
| "grad_norm": 0.5687633577837329, | |
| "learning_rate": 2.7738492792643224e-05, | |
| "loss": 0.3317, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.1862348178137654, | |
| "grad_norm": 0.5740754746529344, | |
| "learning_rate": 2.768631254100716e-05, | |
| "loss": 0.3376, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.1902834008097165, | |
| "grad_norm": 0.581336817815702, | |
| "learning_rate": 2.7634070829820836e-05, | |
| "loss": 0.3294, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.194331983805668, | |
| "grad_norm": 0.5461329122274747, | |
| "learning_rate": 2.7581768076807592e-05, | |
| "loss": 0.3216, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.1983805668016196, | |
| "grad_norm": 0.5878317643491633, | |
| "learning_rate": 2.7529404700178832e-05, | |
| "loss": 0.3214, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.2024291497975708, | |
| "grad_norm": 0.5671672814797761, | |
| "learning_rate": 2.7476981118630735e-05, | |
| "loss": 0.327, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.2064777327935223, | |
| "grad_norm": 0.5828358103801321, | |
| "learning_rate": 2.742449775134085e-05, | |
| "loss": 0.3225, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.5871400746274831, | |
| "learning_rate": 2.7371955017964788e-05, | |
| "loss": 0.3321, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.214574898785425, | |
| "grad_norm": 0.6259211978356387, | |
| "learning_rate": 2.7319353338632845e-05, | |
| "loss": 0.3305, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.2186234817813766, | |
| "grad_norm": 0.5994021649264132, | |
| "learning_rate": 2.7266693133946645e-05, | |
| "loss": 0.3269, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.2226720647773277, | |
| "grad_norm": 0.5908556456560057, | |
| "learning_rate": 2.7213974824975785e-05, | |
| "loss": 0.3155, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2267206477732793, | |
| "grad_norm": 0.6142324218226479, | |
| "learning_rate": 2.7161198833254464e-05, | |
| "loss": 0.3276, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.5668554887385494, | |
| "learning_rate": 2.7108365580778095e-05, | |
| "loss": 0.3279, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.234817813765182, | |
| "grad_norm": 0.6725246092642019, | |
| "learning_rate": 2.705547548999996e-05, | |
| "loss": 0.3354, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.2388663967611335, | |
| "grad_norm": 0.5775626704987575, | |
| "learning_rate": 2.7002528983827817e-05, | |
| "loss": 0.3335, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.242914979757085, | |
| "grad_norm": 0.6178095432571559, | |
| "learning_rate": 2.6949526485620517e-05, | |
| "loss": 0.3296, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.246963562753036, | |
| "grad_norm": 0.5771796381446491, | |
| "learning_rate": 2.6896468419184623e-05, | |
| "loss": 0.3423, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.251012145748988, | |
| "grad_norm": 0.6003469473575598, | |
| "learning_rate": 2.6843355208771024e-05, | |
| "loss": 0.3321, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.2550607287449393, | |
| "grad_norm": 0.5615884951074165, | |
| "learning_rate": 2.6790187279071536e-05, | |
| "loss": 0.3192, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.2591093117408905, | |
| "grad_norm": 0.5753621733715386, | |
| "learning_rate": 2.67369650552155e-05, | |
| "loss": 0.3277, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.6163189507817494, | |
| "learning_rate": 2.6683688962766407e-05, | |
| "loss": 0.3266, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.2672064777327936, | |
| "grad_norm": 0.5577399543917514, | |
| "learning_rate": 2.6630359427718468e-05, | |
| "loss": 0.3375, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2712550607287447, | |
| "grad_norm": 0.5954794901084847, | |
| "learning_rate": 2.6576976876493225e-05, | |
| "loss": 0.3122, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.2753036437246963, | |
| "grad_norm": 0.5731523510102583, | |
| "learning_rate": 2.652354173593614e-05, | |
| "loss": 0.3346, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.279352226720648, | |
| "grad_norm": 0.6342305276336915, | |
| "learning_rate": 2.647005443331317e-05, | |
| "loss": 0.3329, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.283400809716599, | |
| "grad_norm": 0.6475517298051523, | |
| "learning_rate": 2.6416515396307354e-05, | |
| "loss": 0.3352, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.2874493927125505, | |
| "grad_norm": 0.6019415624015845, | |
| "learning_rate": 2.636292505301542e-05, | |
| "loss": 0.3482, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.291497975708502, | |
| "grad_norm": 0.6041190772691362, | |
| "learning_rate": 2.630928383194431e-05, | |
| "loss": 0.3249, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.2955465587044532, | |
| "grad_norm": 0.5922696961273499, | |
| "learning_rate": 2.625559216200781e-05, | |
| "loss": 0.3258, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.299595141700405, | |
| "grad_norm": 0.5749714682364586, | |
| "learning_rate": 2.620185047252308e-05, | |
| "loss": 0.3367, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.3036437246963564, | |
| "grad_norm": 0.5866052524973794, | |
| "learning_rate": 2.614805919320723e-05, | |
| "loss": 0.3356, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.5775777005159505, | |
| "learning_rate": 2.6094218754173895e-05, | |
| "loss": 0.3343, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.311740890688259, | |
| "grad_norm": 0.5732952532130144, | |
| "learning_rate": 2.60403295859298e-05, | |
| "loss": 0.3273, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.5963841600555134, | |
| "learning_rate": 2.5986392119371297e-05, | |
| "loss": 0.3329, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.3198380566801617, | |
| "grad_norm": 0.5933861056164098, | |
| "learning_rate": 2.5932406785780928e-05, | |
| "loss": 0.3422, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.3238866396761133, | |
| "grad_norm": 0.6202901181034397, | |
| "learning_rate": 2.5878374016823987e-05, | |
| "loss": 0.3347, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.327935222672065, | |
| "grad_norm": 0.5947382388218028, | |
| "learning_rate": 2.5824294244545045e-05, | |
| "loss": 0.3331, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.331983805668016, | |
| "grad_norm": 0.567320896337724, | |
| "learning_rate": 2.577016790136453e-05, | |
| "loss": 0.343, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.3360323886639676, | |
| "grad_norm": 0.607180298418831, | |
| "learning_rate": 2.5715995420075225e-05, | |
| "loss": 0.3258, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.340080971659919, | |
| "grad_norm": 0.5922974191202687, | |
| "learning_rate": 2.5661777233838864e-05, | |
| "loss": 0.3215, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.3441295546558703, | |
| "grad_norm": 0.5964725872620796, | |
| "learning_rate": 2.5607513776182616e-05, | |
| "loss": 0.3237, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.348178137651822, | |
| "grad_norm": 0.5930787351225197, | |
| "learning_rate": 2.5553205480995633e-05, | |
| "loss": 0.3254, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.3522267206477734, | |
| "grad_norm": 0.6113085565256515, | |
| "learning_rate": 2.5498852782525604e-05, | |
| "loss": 0.3395, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.3562753036437245, | |
| "grad_norm": 0.552560777478873, | |
| "learning_rate": 2.5444456115375252e-05, | |
| "loss": 0.3244, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.360323886639676, | |
| "grad_norm": 0.5979415587871841, | |
| "learning_rate": 2.539001591449889e-05, | |
| "loss": 0.3275, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.3643724696356276, | |
| "grad_norm": 0.5653824643941705, | |
| "learning_rate": 2.53355326151989e-05, | |
| "loss": 0.3215, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.5587368531141382, | |
| "learning_rate": 2.5281006653122288e-05, | |
| "loss": 0.3286, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.3724696356275303, | |
| "grad_norm": 0.5530969887214858, | |
| "learning_rate": 2.5226438464257217e-05, | |
| "loss": 0.3351, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.376518218623482, | |
| "grad_norm": 0.6012219589134333, | |
| "learning_rate": 2.5171828484929447e-05, | |
| "loss": 0.3442, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.380566801619433, | |
| "grad_norm": 0.5617049446733384, | |
| "learning_rate": 2.5117177151798934e-05, | |
| "loss": 0.3377, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.5588978216622987, | |
| "learning_rate": 2.506248490185627e-05, | |
| "loss": 0.336, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.388663967611336, | |
| "grad_norm": 0.6026586022024909, | |
| "learning_rate": 2.500775217241923e-05, | |
| "loss": 0.3392, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.3927125506072873, | |
| "grad_norm": 0.5877021733792994, | |
| "learning_rate": 2.4952979401129266e-05, | |
| "loss": 0.3303, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.396761133603239, | |
| "grad_norm": 0.5785501789528618, | |
| "learning_rate": 2.4898167025948e-05, | |
| "loss": 0.3352, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.4008097165991904, | |
| "grad_norm": 0.6006051007044411, | |
| "learning_rate": 2.4843315485153703e-05, | |
| "loss": 0.3501, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.4048582995951415, | |
| "grad_norm": 0.5973790682241997, | |
| "learning_rate": 2.4788425217337856e-05, | |
| "loss": 0.3331, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.408906882591093, | |
| "grad_norm": 0.6007771394823106, | |
| "learning_rate": 2.4733496661401572e-05, | |
| "loss": 0.329, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.4129554655870447, | |
| "grad_norm": 0.5896306378290673, | |
| "learning_rate": 2.4678530256552114e-05, | |
| "loss": 0.3444, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.417004048582996, | |
| "grad_norm": 0.5858748293502342, | |
| "learning_rate": 2.46235264422994e-05, | |
| "loss": 0.3336, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.5762339239073828, | |
| "learning_rate": 2.4568485658452452e-05, | |
| "loss": 0.3334, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.425101214574899, | |
| "grad_norm": 0.565927219478946, | |
| "learning_rate": 2.451340834511592e-05, | |
| "loss": 0.336, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.42914979757085, | |
| "grad_norm": 0.5890860193688824, | |
| "learning_rate": 2.4458294942686528e-05, | |
| "loss": 0.3277, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.4331983805668016, | |
| "grad_norm": 0.579715039754551, | |
| "learning_rate": 2.440314589184958e-05, | |
| "loss": 0.3337, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.437246963562753, | |
| "grad_norm": 0.5927815244989495, | |
| "learning_rate": 2.434796163357541e-05, | |
| "loss": 0.3373, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.4412955465587043, | |
| "grad_norm": 0.5891974976765698, | |
| "learning_rate": 2.429274260911588e-05, | |
| "loss": 0.3354, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.445344129554656, | |
| "grad_norm": 0.5910868931879784, | |
| "learning_rate": 2.423748926000084e-05, | |
| "loss": 0.3315, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.4493927125506074, | |
| "grad_norm": 0.5493527833895605, | |
| "learning_rate": 2.418220202803459e-05, | |
| "loss": 0.3362, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.4534412955465585, | |
| "grad_norm": 0.5685870900828257, | |
| "learning_rate": 2.412688135529237e-05, | |
| "loss": 0.3266, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.45748987854251, | |
| "grad_norm": 0.579923903117834, | |
| "learning_rate": 2.4071527684116807e-05, | |
| "loss": 0.3383, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.5602750915492556, | |
| "learning_rate": 2.4016141457114372e-05, | |
| "loss": 0.3335, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.465587044534413, | |
| "grad_norm": 0.5767072620078489, | |
| "learning_rate": 2.396072311715187e-05, | |
| "loss": 0.3376, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.4696356275303644, | |
| "grad_norm": 0.5598589616965857, | |
| "learning_rate": 2.3905273107352863e-05, | |
| "loss": 0.3274, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.5768807725866827, | |
| "learning_rate": 2.3849791871094157e-05, | |
| "loss": 0.3373, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.477732793522267, | |
| "grad_norm": 0.539113881882867, | |
| "learning_rate": 2.379427985200223e-05, | |
| "loss": 0.3419, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.4817813765182186, | |
| "grad_norm": 0.5690138704398798, | |
| "learning_rate": 2.373873749394972e-05, | |
| "loss": 0.334, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.48582995951417, | |
| "grad_norm": 0.5526396358523753, | |
| "learning_rate": 2.3683165241051836e-05, | |
| "loss": 0.3336, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.4898785425101213, | |
| "grad_norm": 0.5502181035952757, | |
| "learning_rate": 2.3627563537662823e-05, | |
| "loss": 0.3358, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.493927125506073, | |
| "grad_norm": 0.5668773764810557, | |
| "learning_rate": 2.3571932828372433e-05, | |
| "loss": 0.3336, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.4979757085020244, | |
| "grad_norm": 0.5893669319032543, | |
| "learning_rate": 2.351627355800233e-05, | |
| "loss": 0.3282, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.5020242914979756, | |
| "grad_norm": 0.5636112170422841, | |
| "learning_rate": 2.346058617160256e-05, | |
| "loss": 0.338, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.506072874493927, | |
| "grad_norm": 0.5707939456555897, | |
| "learning_rate": 2.3404871114447976e-05, | |
| "loss": 0.3374, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.5101214574898787, | |
| "grad_norm": 0.615350662180328, | |
| "learning_rate": 2.334912883203469e-05, | |
| "loss": 0.3369, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.51417004048583, | |
| "grad_norm": 0.5535007940394283, | |
| "learning_rate": 2.3293359770076505e-05, | |
| "loss": 0.3369, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.5182186234817814, | |
| "grad_norm": 0.5735104159070171, | |
| "learning_rate": 2.323756437450136e-05, | |
| "loss": 0.3294, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.522267206477733, | |
| "grad_norm": 0.5899203862807674, | |
| "learning_rate": 2.3181743091447754e-05, | |
| "loss": 0.3337, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.5657868458211172, | |
| "learning_rate": 2.3125896367261164e-05, | |
| "loss": 0.332, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5303643724696356, | |
| "grad_norm": 0.5849534135196368, | |
| "learning_rate": 2.307002464849052e-05, | |
| "loss": 0.3201, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.534412955465587, | |
| "grad_norm": 0.5627581484511709, | |
| "learning_rate": 2.30141283818846e-05, | |
| "loss": 0.3354, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.5508563117748102, | |
| "learning_rate": 2.2958208014388462e-05, | |
| "loss": 0.3279, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.54251012145749, | |
| "grad_norm": 0.5498610376945295, | |
| "learning_rate": 2.2902263993139857e-05, | |
| "loss": 0.3355, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.5465587044534415, | |
| "grad_norm": 0.572443345483025, | |
| "learning_rate": 2.2846296765465708e-05, | |
| "loss": 0.331, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.5506072874493926, | |
| "grad_norm": 0.5426340854921058, | |
| "learning_rate": 2.2790306778878465e-05, | |
| "loss": 0.3402, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.554655870445344, | |
| "grad_norm": 0.5576428547402056, | |
| "learning_rate": 2.273429448107257e-05, | |
| "loss": 0.3268, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.5587044534412957, | |
| "grad_norm": 0.5614950961103626, | |
| "learning_rate": 2.2678260319920868e-05, | |
| "loss": 0.3289, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.562753036437247, | |
| "grad_norm": 0.5274081859466022, | |
| "learning_rate": 2.262220474347102e-05, | |
| "loss": 0.3295, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.5668016194331984, | |
| "grad_norm": 0.5720962220081428, | |
| "learning_rate": 2.2566128199941918e-05, | |
| "loss": 0.332, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.57085020242915, | |
| "grad_norm": 0.5493271074330017, | |
| "learning_rate": 2.2510031137720104e-05, | |
| "loss": 0.3379, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.574898785425101, | |
| "grad_norm": 0.570557203625282, | |
| "learning_rate": 2.2453914005356206e-05, | |
| "loss": 0.337, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.5527666144924104, | |
| "learning_rate": 2.2397777251561302e-05, | |
| "loss": 0.3332, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.582995951417004, | |
| "grad_norm": 0.5950499287932448, | |
| "learning_rate": 2.2341621325203393e-05, | |
| "loss": 0.3388, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.5870445344129553, | |
| "grad_norm": 0.585606633943845, | |
| "learning_rate": 2.2285446675303764e-05, | |
| "loss": 0.3256, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.591093117408907, | |
| "grad_norm": 0.5755519957403183, | |
| "learning_rate": 2.2229253751033423e-05, | |
| "loss": 0.3251, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.5951417004048585, | |
| "grad_norm": 0.6128690888063102, | |
| "learning_rate": 2.2173043001709492e-05, | |
| "loss": 0.3401, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.5991902834008096, | |
| "grad_norm": 0.5614299963169289, | |
| "learning_rate": 2.2116814876791627e-05, | |
| "loss": 0.3362, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.603238866396761, | |
| "grad_norm": 0.5765713491798111, | |
| "learning_rate": 2.2060569825878413e-05, | |
| "loss": 0.3396, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.6072874493927127, | |
| "grad_norm": 0.5612104510913393, | |
| "learning_rate": 2.2004308298703777e-05, | |
| "loss": 0.3274, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.611336032388664, | |
| "grad_norm": 0.5825781427652414, | |
| "learning_rate": 2.1948030745133396e-05, | |
| "loss": 0.3408, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.5940528848995072, | |
| "learning_rate": 2.1891737615161083e-05, | |
| "loss": 0.3365, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.619433198380567, | |
| "grad_norm": 0.5622057267626622, | |
| "learning_rate": 2.1835429358905218e-05, | |
| "loss": 0.3321, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.623481781376518, | |
| "grad_norm": 0.580299770387165, | |
| "learning_rate": 2.17791064266051e-05, | |
| "loss": 0.3288, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.6275303643724697, | |
| "grad_norm": 0.5489381991316346, | |
| "learning_rate": 2.172276926861741e-05, | |
| "loss": 0.337, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.5625329132801877, | |
| "learning_rate": 2.166641833541255e-05, | |
| "loss": 0.3296, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.6356275303643724, | |
| "grad_norm": 0.5503943414273333, | |
| "learning_rate": 2.161005407757109e-05, | |
| "loss": 0.3385, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.639676113360324, | |
| "grad_norm": 0.560557050221961, | |
| "learning_rate": 2.155367694578013e-05, | |
| "loss": 0.3461, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.6437246963562755, | |
| "grad_norm": 0.5529207032883245, | |
| "learning_rate": 2.1497287390829712e-05, | |
| "loss": 0.3388, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.6477732793522266, | |
| "grad_norm": 0.554932464433224, | |
| "learning_rate": 2.1440885863609208e-05, | |
| "loss": 0.3373, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.651821862348178, | |
| "grad_norm": 0.5272569267782395, | |
| "learning_rate": 2.1384472815103737e-05, | |
| "loss": 0.3406, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.6558704453441297, | |
| "grad_norm": 0.5614359160221601, | |
| "learning_rate": 2.1328048696390534e-05, | |
| "loss": 0.3335, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.659919028340081, | |
| "grad_norm": 0.5515570028139887, | |
| "learning_rate": 2.1271613958635342e-05, | |
| "loss": 0.3362, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.6639676113360324, | |
| "grad_norm": 0.533079266671356, | |
| "learning_rate": 2.1215169053088825e-05, | |
| "loss": 0.3233, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.668016194331984, | |
| "grad_norm": 0.5718437758871524, | |
| "learning_rate": 2.1158714431082937e-05, | |
| "loss": 0.3325, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.672064777327935, | |
| "grad_norm": 0.5283049556591785, | |
| "learning_rate": 2.110225054402733e-05, | |
| "loss": 0.3343, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.6761133603238867, | |
| "grad_norm": 0.5620910158866274, | |
| "learning_rate": 2.1045777843405747e-05, | |
| "loss": 0.3342, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.6801619433198383, | |
| "grad_norm": 0.583959540200689, | |
| "learning_rate": 2.09892967807724e-05, | |
| "loss": 0.337, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.5521163575825638, | |
| "learning_rate": 2.093280780774835e-05, | |
| "loss": 0.3313, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.688259109311741, | |
| "grad_norm": 0.6198598601873652, | |
| "learning_rate": 2.087631137601793e-05, | |
| "loss": 0.3161, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.5521503222347048, | |
| "learning_rate": 2.0819807937325104e-05, | |
| "loss": 0.3285, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.6963562753036436, | |
| "grad_norm": 0.5836070294462439, | |
| "learning_rate": 2.0763297943469847e-05, | |
| "loss": 0.3288, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.700404858299595, | |
| "grad_norm": 0.5621519009260704, | |
| "learning_rate": 2.070678184630458e-05, | |
| "loss": 0.3323, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.7044534412955468, | |
| "grad_norm": 0.5669542073023939, | |
| "learning_rate": 2.06502600977305e-05, | |
| "loss": 0.3302, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.708502024291498, | |
| "grad_norm": 0.5693081360173637, | |
| "learning_rate": 2.0593733149694008e-05, | |
| "loss": 0.3344, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.7125506072874495, | |
| "grad_norm": 0.5586182605281562, | |
| "learning_rate": 2.0537201454183074e-05, | |
| "loss": 0.3323, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.716599190283401, | |
| "grad_norm": 0.5498776208120761, | |
| "learning_rate": 2.0480665463223626e-05, | |
| "loss": 0.3369, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.720647773279352, | |
| "grad_norm": 0.5318085060815756, | |
| "learning_rate": 2.042412562887594e-05, | |
| "loss": 0.3259, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.7246963562753037, | |
| "grad_norm": 0.5403404905694669, | |
| "learning_rate": 2.0367582403231033e-05, | |
| "loss": 0.3232, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.7287449392712553, | |
| "grad_norm": 0.5468790789773774, | |
| "learning_rate": 2.0311036238407023e-05, | |
| "loss": 0.3389, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7327935222672064, | |
| "grad_norm": 0.5285229732852461, | |
| "learning_rate": 2.0254487586545534e-05, | |
| "loss": 0.3369, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.5656977419111607, | |
| "learning_rate": 2.0197936899808088e-05, | |
| "loss": 0.3345, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.7408906882591095, | |
| "grad_norm": 0.5438781262329058, | |
| "learning_rate": 2.0141384630372466e-05, | |
| "loss": 0.3224, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7449392712550607, | |
| "grad_norm": 0.577771463017823, | |
| "learning_rate": 2.0084831230429106e-05, | |
| "loss": 0.3362, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.748987854251012, | |
| "grad_norm": 0.533077547595891, | |
| "learning_rate": 2.0028277152177492e-05, | |
| "loss": 0.3346, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.753036437246964, | |
| "grad_norm": 0.6196740484996521, | |
| "learning_rate": 1.9971722847822518e-05, | |
| "loss": 0.3403, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.757085020242915, | |
| "grad_norm": 0.5335604030943, | |
| "learning_rate": 1.99151687695709e-05, | |
| "loss": 0.3317, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7611336032388665, | |
| "grad_norm": 0.5824234082780467, | |
| "learning_rate": 1.9858615369627537e-05, | |
| "loss": 0.336, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.765182186234818, | |
| "grad_norm": 0.5815201851873978, | |
| "learning_rate": 1.9802063100191916e-05, | |
| "loss": 0.3366, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.5543283511547564, | |
| "learning_rate": 1.974551241345447e-05, | |
| "loss": 0.3365, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.7732793522267207, | |
| "grad_norm": 0.6061835169168608, | |
| "learning_rate": 1.9688963761592984e-05, | |
| "loss": 0.3345, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7773279352226723, | |
| "grad_norm": 0.5629488698409981, | |
| "learning_rate": 1.963241759676897e-05, | |
| "loss": 0.3327, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.7813765182186234, | |
| "grad_norm": 0.6338013185995467, | |
| "learning_rate": 1.9575874371124066e-05, | |
| "loss": 0.3304, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.785425101214575, | |
| "grad_norm": 0.5284247225820906, | |
| "learning_rate": 1.951933453677638e-05, | |
| "loss": 0.3128, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.5727641169673803, | |
| "learning_rate": 1.9462798545816932e-05, | |
| "loss": 0.3362, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.7935222672064777, | |
| "grad_norm": 0.5452967193898547, | |
| "learning_rate": 1.9406266850305996e-05, | |
| "loss": 0.3272, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7975708502024292, | |
| "grad_norm": 0.5431303980335477, | |
| "learning_rate": 1.9349739902269503e-05, | |
| "loss": 0.3224, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.801619433198381, | |
| "grad_norm": 0.557769272660263, | |
| "learning_rate": 1.9293218153695425e-05, | |
| "loss": 0.3309, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.805668016194332, | |
| "grad_norm": 0.5281264360788727, | |
| "learning_rate": 1.9236702056530153e-05, | |
| "loss": 0.3265, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.8097165991902835, | |
| "grad_norm": 0.5552567333827647, | |
| "learning_rate": 1.9180192062674907e-05, | |
| "loss": 0.3368, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.813765182186235, | |
| "grad_norm": 0.5534477867479235, | |
| "learning_rate": 1.9123688623982076e-05, | |
| "loss": 0.333, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.817813765182186, | |
| "grad_norm": 0.5673213983996134, | |
| "learning_rate": 1.906719219225165e-05, | |
| "loss": 0.3321, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.8218623481781377, | |
| "grad_norm": 0.5440143228006966, | |
| "learning_rate": 1.9010703219227603e-05, | |
| "loss": 0.3371, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.8259109311740893, | |
| "grad_norm": 0.5345607352572898, | |
| "learning_rate": 1.8954222156594256e-05, | |
| "loss": 0.3366, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.8299595141700404, | |
| "grad_norm": 0.5316103831243224, | |
| "learning_rate": 1.8897749455972673e-05, | |
| "loss": 0.3332, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.834008097165992, | |
| "grad_norm": 0.5809628866547711, | |
| "learning_rate": 1.884128556891707e-05, | |
| "loss": 0.3301, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8380566801619436, | |
| "grad_norm": 0.5362041688729495, | |
| "learning_rate": 1.878483094691119e-05, | |
| "loss": 0.3392, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.5569426892895448, | |
| "learning_rate": 1.8728386041364664e-05, | |
| "loss": 0.3315, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.5540379060903055, | |
| "learning_rate": 1.867195130360947e-05, | |
| "loss": 0.3273, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.850202429149798, | |
| "grad_norm": 0.5304159407233333, | |
| "learning_rate": 1.8615527184896263e-05, | |
| "loss": 0.3264, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.854251012145749, | |
| "grad_norm": 0.5530304490995382, | |
| "learning_rate": 1.8559114136390795e-05, | |
| "loss": 0.3407, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.8582995951417005, | |
| "grad_norm": 0.5490142199443762, | |
| "learning_rate": 1.8502712609170298e-05, | |
| "loss": 0.3156, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.862348178137652, | |
| "grad_norm": 0.529837998863273, | |
| "learning_rate": 1.8446323054219876e-05, | |
| "loss": 0.3213, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.866396761133603, | |
| "grad_norm": 0.5443933806815265, | |
| "learning_rate": 1.838994592242891e-05, | |
| "loss": 0.3355, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.8704453441295548, | |
| "grad_norm": 0.5440225994909329, | |
| "learning_rate": 1.8333581664587453e-05, | |
| "loss": 0.3244, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.8744939271255063, | |
| "grad_norm": 0.5564177990011538, | |
| "learning_rate": 1.8277230731382593e-05, | |
| "loss": 0.3329, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.8785425101214575, | |
| "grad_norm": 0.538683792598412, | |
| "learning_rate": 1.82208935733949e-05, | |
| "loss": 0.3314, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.882591093117409, | |
| "grad_norm": 0.5486767069204396, | |
| "learning_rate": 1.8164570641094793e-05, | |
| "loss": 0.3195, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.8866396761133606, | |
| "grad_norm": 0.5505859326348537, | |
| "learning_rate": 1.810826238483892e-05, | |
| "loss": 0.3194, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.8906882591093117, | |
| "grad_norm": 0.5470468076712892, | |
| "learning_rate": 1.8051969254866608e-05, | |
| "loss": 0.3368, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.5607222069063801, | |
| "learning_rate": 1.7995691701296226e-05, | |
| "loss": 0.3321, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.898785425101215, | |
| "grad_norm": 0.5782704314031526, | |
| "learning_rate": 1.7939430174121593e-05, | |
| "loss": 0.328, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.902834008097166, | |
| "grad_norm": 0.5588353116260328, | |
| "learning_rate": 1.788318512320838e-05, | |
| "loss": 0.3295, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.9068825910931175, | |
| "grad_norm": 0.5401329339271393, | |
| "learning_rate": 1.782695699829051e-05, | |
| "loss": 0.3211, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.910931174089069, | |
| "grad_norm": 0.5750791725789443, | |
| "learning_rate": 1.7770746248966587e-05, | |
| "loss": 0.3158, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.91497975708502, | |
| "grad_norm": 0.5703579080343582, | |
| "learning_rate": 1.7714553324696243e-05, | |
| "loss": 0.3237, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.919028340080972, | |
| "grad_norm": 0.5575433818733025, | |
| "learning_rate": 1.7658378674796614e-05, | |
| "loss": 0.3372, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.5440284227911192, | |
| "learning_rate": 1.76022227484387e-05, | |
| "loss": 0.3375, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.9271255060728745, | |
| "grad_norm": 0.5860886911425814, | |
| "learning_rate": 1.7546085994643807e-05, | |
| "loss": 0.3313, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.931174089068826, | |
| "grad_norm": 0.5509116529085467, | |
| "learning_rate": 1.7489968862279902e-05, | |
| "loss": 0.3291, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9352226720647776, | |
| "grad_norm": 0.5518667749150153, | |
| "learning_rate": 1.743387180005809e-05, | |
| "loss": 0.3296, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.9392712550607287, | |
| "grad_norm": 0.5815460037971854, | |
| "learning_rate": 1.737779525652899e-05, | |
| "loss": 0.3346, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.9433198380566803, | |
| "grad_norm": 0.5375259505412489, | |
| "learning_rate": 1.732173968007914e-05, | |
| "loss": 0.3338, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.5870142575839559, | |
| "learning_rate": 1.7265705518927437e-05, | |
| "loss": 0.3346, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.951417004048583, | |
| "grad_norm": 0.5462830527635123, | |
| "learning_rate": 1.7209693221121542e-05, | |
| "loss": 0.3165, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.9554655870445345, | |
| "grad_norm": 0.5192873795526657, | |
| "learning_rate": 1.7153703234534302e-05, | |
| "loss": 0.3199, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.9595141700404857, | |
| "grad_norm": 0.5520830086173941, | |
| "learning_rate": 1.709773600686015e-05, | |
| "loss": 0.3316, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.9635627530364372, | |
| "grad_norm": 0.5453140345748625, | |
| "learning_rate": 1.7041791985611548e-05, | |
| "loss": 0.3443, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.967611336032389, | |
| "grad_norm": 0.5275843595414752, | |
| "learning_rate": 1.69858716181154e-05, | |
| "loss": 0.3201, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.97165991902834, | |
| "grad_norm": 0.5381147717198677, | |
| "learning_rate": 1.692997535150948e-05, | |
| "loss": 0.3288, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.9757085020242915, | |
| "grad_norm": 0.5462749557280593, | |
| "learning_rate": 1.687410363273884e-05, | |
| "loss": 0.3312, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.979757085020243, | |
| "grad_norm": 0.5362343833156603, | |
| "learning_rate": 1.6818256908552257e-05, | |
| "loss": 0.3205, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.983805668016194, | |
| "grad_norm": 0.5247761351913595, | |
| "learning_rate": 1.6762435625498646e-05, | |
| "loss": 0.3316, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.9878542510121457, | |
| "grad_norm": 0.527716707045007, | |
| "learning_rate": 1.67066402299235e-05, | |
| "loss": 0.3252, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.9919028340080973, | |
| "grad_norm": 0.5421079021178479, | |
| "learning_rate": 1.6650871167965313e-05, | |
| "loss": 0.3274, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.9959514170040484, | |
| "grad_norm": 0.5270184820669256, | |
| "learning_rate": 1.6595128885552028e-05, | |
| "loss": 0.3272, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5453031386019743, | |
| "learning_rate": 1.653941382839745e-05, | |
| "loss": 0.3156, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 3.0040485829959516, | |
| "grad_norm": 0.6922841678897117, | |
| "learning_rate": 1.6483726441997673e-05, | |
| "loss": 0.1699, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 3.0080971659919027, | |
| "grad_norm": 0.6013621015606296, | |
| "learning_rate": 1.642806717162757e-05, | |
| "loss": 0.1582, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 3.0121457489878543, | |
| "grad_norm": 0.46881093675501667, | |
| "learning_rate": 1.637243646233718e-05, | |
| "loss": 0.1541, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.016194331983806, | |
| "grad_norm": 0.46864191376572917, | |
| "learning_rate": 1.6316834758948174e-05, | |
| "loss": 0.1544, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.020242914979757, | |
| "grad_norm": 0.6133251288901207, | |
| "learning_rate": 1.6261262506050282e-05, | |
| "loss": 0.1573, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 3.0242914979757085, | |
| "grad_norm": 0.6845212362081912, | |
| "learning_rate": 1.620572014799777e-05, | |
| "loss": 0.1554, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 3.02834008097166, | |
| "grad_norm": 0.6780689744275799, | |
| "learning_rate": 1.6150208128905857e-05, | |
| "loss": 0.1607, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 3.032388663967611, | |
| "grad_norm": 0.5539492376592169, | |
| "learning_rate": 1.6094726892647147e-05, | |
| "loss": 0.1516, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0364372469635628, | |
| "grad_norm": 0.5089383920074327, | |
| "learning_rate": 1.6039276882848135e-05, | |
| "loss": 0.1497, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0404858299595143, | |
| "grad_norm": 0.5168878161961428, | |
| "learning_rate": 1.5983858542885635e-05, | |
| "loss": 0.1561, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 3.0445344129554655, | |
| "grad_norm": 0.47541446246171104, | |
| "learning_rate": 1.5928472315883203e-05, | |
| "loss": 0.1461, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 3.048582995951417, | |
| "grad_norm": 0.4709989614653716, | |
| "learning_rate": 1.5873118644707633e-05, | |
| "loss": 0.1449, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.49534341799005727, | |
| "learning_rate": 1.5817797971965413e-05, | |
| "loss": 0.1451, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 3.0566801619433197, | |
| "grad_norm": 0.466292188613086, | |
| "learning_rate": 1.576251073999917e-05, | |
| "loss": 0.1437, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.0607287449392713, | |
| "grad_norm": 0.472551318090471, | |
| "learning_rate": 1.5707257390884126e-05, | |
| "loss": 0.1483, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 3.064777327935223, | |
| "grad_norm": 0.493251093064373, | |
| "learning_rate": 1.5652038366424595e-05, | |
| "loss": 0.1511, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 3.068825910931174, | |
| "grad_norm": 0.4901731411146901, | |
| "learning_rate": 1.5596854108150424e-05, | |
| "loss": 0.1541, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 3.0728744939271255, | |
| "grad_norm": 0.4915571211673923, | |
| "learning_rate": 1.5541705057313476e-05, | |
| "loss": 0.1505, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.4842526214072446, | |
| "learning_rate": 1.5486591654884086e-05, | |
| "loss": 0.1531, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.080971659919028, | |
| "grad_norm": 0.4786989086567994, | |
| "learning_rate": 1.543151434154755e-05, | |
| "loss": 0.1469, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 3.08502024291498, | |
| "grad_norm": 0.47001917149657263, | |
| "learning_rate": 1.5376473557700615e-05, | |
| "loss": 0.1467, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 3.0890688259109313, | |
| "grad_norm": 0.4802649852909784, | |
| "learning_rate": 1.5321469743447893e-05, | |
| "loss": 0.1464, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 3.0931174089068825, | |
| "grad_norm": 0.4670303075665406, | |
| "learning_rate": 1.5266503338598434e-05, | |
| "loss": 0.143, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 3.097165991902834, | |
| "grad_norm": 0.47567120643706456, | |
| "learning_rate": 1.5211574782662149e-05, | |
| "loss": 0.1441, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1012145748987856, | |
| "grad_norm": 0.45243404992180186, | |
| "learning_rate": 1.5156684514846307e-05, | |
| "loss": 0.1503, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.4654459701383649, | |
| "learning_rate": 1.5101832974052012e-05, | |
| "loss": 0.1453, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 3.1093117408906883, | |
| "grad_norm": 0.467481245779972, | |
| "learning_rate": 1.5047020598870737e-05, | |
| "loss": 0.1502, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 3.11336032388664, | |
| "grad_norm": 0.45660613021485325, | |
| "learning_rate": 1.4992247827580778e-05, | |
| "loss": 0.1457, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 3.117408906882591, | |
| "grad_norm": 0.45519828254153394, | |
| "learning_rate": 1.493751509814374e-05, | |
| "loss": 0.1483, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.1214574898785425, | |
| "grad_norm": 0.4684529811961153, | |
| "learning_rate": 1.4882822848201073e-05, | |
| "loss": 0.141, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 3.125506072874494, | |
| "grad_norm": 0.48535190328107114, | |
| "learning_rate": 1.4828171515070553e-05, | |
| "loss": 0.1549, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 3.1295546558704452, | |
| "grad_norm": 0.47976085398410584, | |
| "learning_rate": 1.4773561535742793e-05, | |
| "loss": 0.1499, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 3.133603238866397, | |
| "grad_norm": 0.4595584538363997, | |
| "learning_rate": 1.4718993346877715e-05, | |
| "loss": 0.1434, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.1376518218623484, | |
| "grad_norm": 0.4487108542691756, | |
| "learning_rate": 1.466446738480111e-05, | |
| "loss": 0.1477, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.1417004048582995, | |
| "grad_norm": 0.4731897566097362, | |
| "learning_rate": 1.460998408550112e-05, | |
| "loss": 0.1464, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 3.145748987854251, | |
| "grad_norm": 0.4547533149998213, | |
| "learning_rate": 1.4555543884624751e-05, | |
| "loss": 0.1443, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 3.1497975708502026, | |
| "grad_norm": 0.4725067069830284, | |
| "learning_rate": 1.4501147217474402e-05, | |
| "loss": 0.141, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 3.1538461538461537, | |
| "grad_norm": 0.4629105526970313, | |
| "learning_rate": 1.444679451900437e-05, | |
| "loss": 0.1466, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.46435963081468506, | |
| "learning_rate": 1.4392486223817397e-05, | |
| "loss": 0.1451, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.161943319838057, | |
| "grad_norm": 0.4948900449042696, | |
| "learning_rate": 1.4338222766161143e-05, | |
| "loss": 0.1573, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 3.165991902834008, | |
| "grad_norm": 0.46228648972295844, | |
| "learning_rate": 1.4284004579924777e-05, | |
| "loss": 0.1372, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 3.1700404858299596, | |
| "grad_norm": 0.46247550970743945, | |
| "learning_rate": 1.4229832098635479e-05, | |
| "loss": 0.1461, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 3.174089068825911, | |
| "grad_norm": 0.47061135391751924, | |
| "learning_rate": 1.4175705755454963e-05, | |
| "loss": 0.1464, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 3.1781376518218623, | |
| "grad_norm": 0.45881943190225316, | |
| "learning_rate": 1.412162598317602e-05, | |
| "loss": 0.1474, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.182186234817814, | |
| "grad_norm": 0.4745005361710274, | |
| "learning_rate": 1.406759321421907e-05, | |
| "loss": 0.1452, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 3.1862348178137654, | |
| "grad_norm": 0.46482665765346964, | |
| "learning_rate": 1.401360788062871e-05, | |
| "loss": 0.1489, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.1902834008097165, | |
| "grad_norm": 0.4717277912056015, | |
| "learning_rate": 1.3959670414070204e-05, | |
| "loss": 0.1402, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.194331983805668, | |
| "grad_norm": 0.46823320269366786, | |
| "learning_rate": 1.3905781245826108e-05, | |
| "loss": 0.1423, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.1983805668016196, | |
| "grad_norm": 0.46749595009650813, | |
| "learning_rate": 1.3851940806792778e-05, | |
| "loss": 0.1446, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.2024291497975708, | |
| "grad_norm": 0.4731515396777185, | |
| "learning_rate": 1.379814952747693e-05, | |
| "loss": 0.1464, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.2064777327935223, | |
| "grad_norm": 0.4456284588080441, | |
| "learning_rate": 1.3744407837992193e-05, | |
| "loss": 0.1409, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.47107569008979716, | |
| "learning_rate": 1.3690716168055692e-05, | |
| "loss": 0.1464, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.214574898785425, | |
| "grad_norm": 0.4476264836318455, | |
| "learning_rate": 1.363707494698459e-05, | |
| "loss": 0.1424, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.2186234817813766, | |
| "grad_norm": 0.456835308296189, | |
| "learning_rate": 1.358348460369265e-05, | |
| "loss": 0.1415, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.2226720647773277, | |
| "grad_norm": 0.4696604746632926, | |
| "learning_rate": 1.352994556668684e-05, | |
| "loss": 0.1472, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.2267206477732793, | |
| "grad_norm": 0.4647929081970695, | |
| "learning_rate": 1.3476458264063865e-05, | |
| "loss": 0.1475, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.4795864032790066, | |
| "learning_rate": 1.342302312350678e-05, | |
| "loss": 0.1475, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.234817813765182, | |
| "grad_norm": 0.45201608090833106, | |
| "learning_rate": 1.3369640572281537e-05, | |
| "loss": 0.1463, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.2388663967611335, | |
| "grad_norm": 0.47838994891064895, | |
| "learning_rate": 1.3316311037233596e-05, | |
| "loss": 0.1443, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.242914979757085, | |
| "grad_norm": 0.4691553579064072, | |
| "learning_rate": 1.326303494478451e-05, | |
| "loss": 0.1462, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.246963562753036, | |
| "grad_norm": 0.4591889073163608, | |
| "learning_rate": 1.3209812720928472e-05, | |
| "loss": 0.1524, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.251012145748988, | |
| "grad_norm": 0.45342508130251175, | |
| "learning_rate": 1.315664479122898e-05, | |
| "loss": 0.1417, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.2550607287449393, | |
| "grad_norm": 0.4718736476596431, | |
| "learning_rate": 1.3103531580815378e-05, | |
| "loss": 0.1459, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.2591093117408905, | |
| "grad_norm": 0.44525036872298274, | |
| "learning_rate": 1.3050473514379488e-05, | |
| "loss": 0.1415, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.47365409019945415, | |
| "learning_rate": 1.2997471016172188e-05, | |
| "loss": 0.1426, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.2672064777327936, | |
| "grad_norm": 0.47710098537738976, | |
| "learning_rate": 1.2944524510000042e-05, | |
| "loss": 0.1431, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.2712550607287447, | |
| "grad_norm": 0.4744850893186009, | |
| "learning_rate": 1.2891634419221908e-05, | |
| "loss": 0.1542, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.2753036437246963, | |
| "grad_norm": 0.47122082528227777, | |
| "learning_rate": 1.2838801166745545e-05, | |
| "loss": 0.1457, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.279352226720648, | |
| "grad_norm": 0.4610548129168195, | |
| "learning_rate": 1.2786025175024216e-05, | |
| "loss": 0.1491, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.283400809716599, | |
| "grad_norm": 0.46762406205356416, | |
| "learning_rate": 1.2733306866053357e-05, | |
| "loss": 0.1451, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.2874493927125505, | |
| "grad_norm": 0.4817581716348365, | |
| "learning_rate": 1.2680646661367163e-05, | |
| "loss": 0.1485, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.291497975708502, | |
| "grad_norm": 0.4849025055703871, | |
| "learning_rate": 1.2628044982035219e-05, | |
| "loss": 0.1472, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.2955465587044532, | |
| "grad_norm": 0.45302959901374407, | |
| "learning_rate": 1.2575502248659155e-05, | |
| "loss": 0.1435, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.299595141700405, | |
| "grad_norm": 0.4693840965021741, | |
| "learning_rate": 1.2523018881369269e-05, | |
| "loss": 0.1498, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.3036437246963564, | |
| "grad_norm": 0.4819045318483372, | |
| "learning_rate": 1.2470595299821176e-05, | |
| "loss": 0.1459, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.3076923076923075, | |
| "grad_norm": 0.4608756538812948, | |
| "learning_rate": 1.2418231923192416e-05, | |
| "loss": 0.1506, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.311740890688259, | |
| "grad_norm": 0.4690122384065877, | |
| "learning_rate": 1.2365929170179165e-05, | |
| "loss": 0.1467, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.45282604766331064, | |
| "learning_rate": 1.2313687458992848e-05, | |
| "loss": 0.1475, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.3198380566801617, | |
| "grad_norm": 0.4502359286963051, | |
| "learning_rate": 1.2261507207356785e-05, | |
| "loss": 0.1345, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.3238866396761133, | |
| "grad_norm": 0.4657353651008627, | |
| "learning_rate": 1.220938883250288e-05, | |
| "loss": 0.1485, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.327935222672065, | |
| "grad_norm": 0.4634736518172549, | |
| "learning_rate": 1.215733275116827e-05, | |
| "loss": 0.1392, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.331983805668016, | |
| "grad_norm": 0.44453421334233345, | |
| "learning_rate": 1.2105339379592005e-05, | |
| "loss": 0.1417, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.3360323886639676, | |
| "grad_norm": 0.4656530905385027, | |
| "learning_rate": 1.2053409133511681e-05, | |
| "loss": 0.1464, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.340080971659919, | |
| "grad_norm": 0.4579492160985602, | |
| "learning_rate": 1.2001542428160173e-05, | |
| "loss": 0.1447, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.3441295546558703, | |
| "grad_norm": 0.45125813544131266, | |
| "learning_rate": 1.1949739678262278e-05, | |
| "loss": 0.1459, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.348178137651822, | |
| "grad_norm": 0.45119308760918464, | |
| "learning_rate": 1.189800129803139e-05, | |
| "loss": 0.1416, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.3522267206477734, | |
| "grad_norm": 0.45495133367761725, | |
| "learning_rate": 1.1846327701166228e-05, | |
| "loss": 0.143, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.3562753036437245, | |
| "grad_norm": 0.4460966478918341, | |
| "learning_rate": 1.1794719300847493e-05, | |
| "loss": 0.1488, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.360323886639676, | |
| "grad_norm": 0.45529791208163223, | |
| "learning_rate": 1.1743176509734582e-05, | |
| "loss": 0.1471, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.3643724696356276, | |
| "grad_norm": 0.4523895652577792, | |
| "learning_rate": 1.1691699739962275e-05, | |
| "loss": 0.1395, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.4451310563436803, | |
| "learning_rate": 1.1640289403137456e-05, | |
| "loss": 0.1472, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.3724696356275303, | |
| "grad_norm": 0.4444512896449249, | |
| "learning_rate": 1.1588945910335802e-05, | |
| "loss": 0.1459, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.376518218623482, | |
| "grad_norm": 0.45395203223948843, | |
| "learning_rate": 1.1537669672098517e-05, | |
| "loss": 0.14, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.380566801619433, | |
| "grad_norm": 0.4501878330587206, | |
| "learning_rate": 1.1486461098429036e-05, | |
| "loss": 0.1423, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.44881749165192397, | |
| "learning_rate": 1.1435320598789747e-05, | |
| "loss": 0.1495, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.388663967611336, | |
| "grad_norm": 0.459772405089828, | |
| "learning_rate": 1.1384248582098736e-05, | |
| "loss": 0.1457, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.3927125506072873, | |
| "grad_norm": 0.4482454362358956, | |
| "learning_rate": 1.1333245456726467e-05, | |
| "loss": 0.1414, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.396761133603239, | |
| "grad_norm": 0.4611266162118473, | |
| "learning_rate": 1.128231163049258e-05, | |
| "loss": 0.1419, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.4008097165991904, | |
| "grad_norm": 0.44830903895097585, | |
| "learning_rate": 1.123144751066259e-05, | |
| "loss": 0.1449, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.4048582995951415, | |
| "grad_norm": 0.4447366820130688, | |
| "learning_rate": 1.1180653503944657e-05, | |
| "loss": 0.1442, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.408906882591093, | |
| "grad_norm": 0.46720194269781157, | |
| "learning_rate": 1.1129930016486287e-05, | |
| "loss": 0.1428, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.4129554655870447, | |
| "grad_norm": 0.45872529934464645, | |
| "learning_rate": 1.107927745387114e-05, | |
| "loss": 0.1396, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.417004048582996, | |
| "grad_norm": 0.4345830585681101, | |
| "learning_rate": 1.1028696221115766e-05, | |
| "loss": 0.1407, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.44630774079660857, | |
| "learning_rate": 1.0978186722666341e-05, | |
| "loss": 0.1411, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.425101214574899, | |
| "grad_norm": 0.4503473863325322, | |
| "learning_rate": 1.0927749362395475e-05, | |
| "loss": 0.1435, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.42914979757085, | |
| "grad_norm": 0.459783654604343, | |
| "learning_rate": 1.0877384543598954e-05, | |
| "loss": 0.1445, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.4331983805668016, | |
| "grad_norm": 0.45466784311660635, | |
| "learning_rate": 1.0827092668992529e-05, | |
| "loss": 0.1418, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.437246963562753, | |
| "grad_norm": 0.4594277290724805, | |
| "learning_rate": 1.077687414070869e-05, | |
| "loss": 0.1411, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.4412955465587043, | |
| "grad_norm": 0.4519338575047202, | |
| "learning_rate": 1.0726729360293451e-05, | |
| "loss": 0.1403, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.445344129554656, | |
| "grad_norm": 0.4520544555668104, | |
| "learning_rate": 1.0676658728703147e-05, | |
| "loss": 0.152, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.4493927125506074, | |
| "grad_norm": 0.44879465071237673, | |
| "learning_rate": 1.0626662646301198e-05, | |
| "loss": 0.139, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.4534412955465585, | |
| "grad_norm": 0.4483346848898802, | |
| "learning_rate": 1.0576741512854958e-05, | |
| "loss": 0.1452, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.45748987854251, | |
| "grad_norm": 0.4528263510339483, | |
| "learning_rate": 1.052689572753247e-05, | |
| "loss": 0.1455, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.4464721145778219, | |
| "learning_rate": 1.0477125688899328e-05, | |
| "loss": 0.1439, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.465587044534413, | |
| "grad_norm": 0.4455523378010689, | |
| "learning_rate": 1.0427431794915406e-05, | |
| "loss": 0.1455, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.4696356275303644, | |
| "grad_norm": 0.440008924274864, | |
| "learning_rate": 1.0377814442931769e-05, | |
| "loss": 0.1438, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.44302284673948067, | |
| "learning_rate": 1.0328274029687439e-05, | |
| "loss": 0.1472, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.477732793522267, | |
| "grad_norm": 0.454579546471386, | |
| "learning_rate": 1.0278810951306253e-05, | |
| "loss": 0.1465, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.4817813765182186, | |
| "grad_norm": 0.44761636259354526, | |
| "learning_rate": 1.0229425603293647e-05, | |
| "loss": 0.1504, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.48582995951417, | |
| "grad_norm": 0.45026347615404255, | |
| "learning_rate": 1.0180118380533558e-05, | |
| "loss": 0.1423, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.4898785425101213, | |
| "grad_norm": 0.4517673445419881, | |
| "learning_rate": 1.0130889677285225e-05, | |
| "loss": 0.1431, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.493927125506073, | |
| "grad_norm": 0.4515463354556758, | |
| "learning_rate": 1.008173988718005e-05, | |
| "loss": 0.1483, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.4979757085020244, | |
| "grad_norm": 0.44010161651195234, | |
| "learning_rate": 1.0032669403218439e-05, | |
| "loss": 0.1433, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.5020242914979756, | |
| "grad_norm": 0.45356028104940027, | |
| "learning_rate": 9.983678617766677e-06, | |
| "loss": 0.1432, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.506072874493927, | |
| "grad_norm": 0.45092377798627414, | |
| "learning_rate": 9.934767922553788e-06, | |
| "loss": 0.1417, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.5101214574898787, | |
| "grad_norm": 0.46259930243826575, | |
| "learning_rate": 9.885937708668371e-06, | |
| "loss": 0.1451, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.51417004048583, | |
| "grad_norm": 0.4428491554198212, | |
| "learning_rate": 9.837188366555524e-06, | |
| "loss": 0.1443, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.5182186234817814, | |
| "grad_norm": 0.46011933088424645, | |
| "learning_rate": 9.788520286013691e-06, | |
| "loss": 0.146, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.522267206477733, | |
| "grad_norm": 0.4437370064601154, | |
| "learning_rate": 9.739933856191537e-06, | |
| "loss": 0.1387, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.4516429391132442, | |
| "learning_rate": 9.691429465584866e-06, | |
| "loss": 0.1407, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.5303643724696356, | |
| "grad_norm": 0.4450442689779026, | |
| "learning_rate": 9.643007502033492e-06, | |
| "loss": 0.1454, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.534412955465587, | |
| "grad_norm": 0.4602451766297544, | |
| "learning_rate": 9.594668352718152e-06, | |
| "loss": 0.1465, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.4639745448872694, | |
| "learning_rate": 9.546412404157383e-06, | |
| "loss": 0.151, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.54251012145749, | |
| "grad_norm": 0.4465304412193125, | |
| "learning_rate": 9.498240042204471e-06, | |
| "loss": 0.1522, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.5465587044534415, | |
| "grad_norm": 0.4612360701313509, | |
| "learning_rate": 9.450151652044339e-06, | |
| "loss": 0.1456, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.5506072874493926, | |
| "grad_norm": 0.45637334933477414, | |
| "learning_rate": 9.402147618190472e-06, | |
| "loss": 0.1399, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.554655870445344, | |
| "grad_norm": 0.4527077992516098, | |
| "learning_rate": 9.354228324481841e-06, | |
| "loss": 0.1476, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.5587044534412957, | |
| "grad_norm": 0.46264278670961795, | |
| "learning_rate": 9.306394154079852e-06, | |
| "loss": 0.1427, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.562753036437247, | |
| "grad_norm": 0.4660769574073461, | |
| "learning_rate": 9.258645489465254e-06, | |
| "loss": 0.1357, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.5668016194331984, | |
| "grad_norm": 0.43949088758381083, | |
| "learning_rate": 9.21098271243509e-06, | |
| "loss": 0.1442, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.57085020242915, | |
| "grad_norm": 0.46147931287249466, | |
| "learning_rate": 9.163406204099659e-06, | |
| "loss": 0.1505, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.574898785425101, | |
| "grad_norm": 0.46371680600430715, | |
| "learning_rate": 9.115916344879452e-06, | |
| "loss": 0.1492, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.4545283284287348, | |
| "learning_rate": 9.068513514502126e-06, | |
| "loss": 0.1439, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.582995951417004, | |
| "grad_norm": 0.4533667392485297, | |
| "learning_rate": 9.021198091999439e-06, | |
| "loss": 0.1386, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.5870445344129553, | |
| "grad_norm": 0.4473801793180498, | |
| "learning_rate": 8.973970455704251e-06, | |
| "loss": 0.1404, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.591093117408907, | |
| "grad_norm": 0.4542388283700099, | |
| "learning_rate": 8.926830983247495e-06, | |
| "loss": 0.145, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.5951417004048585, | |
| "grad_norm": 0.44650971532943295, | |
| "learning_rate": 8.879780051555116e-06, | |
| "loss": 0.144, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.5991902834008096, | |
| "grad_norm": 0.43517172386479874, | |
| "learning_rate": 8.83281803684512e-06, | |
| "loss": 0.1375, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.603238866396761, | |
| "grad_norm": 0.4490492533428969, | |
| "learning_rate": 8.785945314624532e-06, | |
| "loss": 0.1378, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.6072874493927127, | |
| "grad_norm": 0.4371451577631009, | |
| "learning_rate": 8.739162259686385e-06, | |
| "loss": 0.1401, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.611336032388664, | |
| "grad_norm": 0.45459846504498913, | |
| "learning_rate": 8.692469246106734e-06, | |
| "loss": 0.149, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 0.442529253035985, | |
| "learning_rate": 8.645866647241677e-06, | |
| "loss": 0.1422, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.619433198380567, | |
| "grad_norm": 0.44358339477029607, | |
| "learning_rate": 8.599354835724348e-06, | |
| "loss": 0.1427, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.623481781376518, | |
| "grad_norm": 0.44085775001637845, | |
| "learning_rate": 8.552934183461958e-06, | |
| "loss": 0.1426, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.6275303643724697, | |
| "grad_norm": 0.4398272410340316, | |
| "learning_rate": 8.506605061632805e-06, | |
| "loss": 0.1462, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.4503421534747265, | |
| "learning_rate": 8.460367840683308e-06, | |
| "loss": 0.1489, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.6356275303643724, | |
| "grad_norm": 0.46508979968831715, | |
| "learning_rate": 8.414222890325065e-06, | |
| "loss": 0.1444, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.639676113360324, | |
| "grad_norm": 0.45597709190710234, | |
| "learning_rate": 8.368170579531855e-06, | |
| "loss": 0.1405, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.6437246963562755, | |
| "grad_norm": 0.4563960452061645, | |
| "learning_rate": 8.322211276536733e-06, | |
| "loss": 0.1427, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.6477732793522266, | |
| "grad_norm": 0.43317853947264684, | |
| "learning_rate": 8.27634534882907e-06, | |
| "loss": 0.1389, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.651821862348178, | |
| "grad_norm": 0.44912934665088833, | |
| "learning_rate": 8.230573163151585e-06, | |
| "loss": 0.138, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.6558704453441297, | |
| "grad_norm": 0.4574591417204284, | |
| "learning_rate": 8.184895085497466e-06, | |
| "loss": 0.1453, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.659919028340081, | |
| "grad_norm": 0.4522152863184393, | |
| "learning_rate": 8.139311481107401e-06, | |
| "loss": 0.1411, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.6639676113360324, | |
| "grad_norm": 0.461475835426998, | |
| "learning_rate": 8.093822714466686e-06, | |
| "loss": 0.1411, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.668016194331984, | |
| "grad_norm": 0.45527834960879715, | |
| "learning_rate": 8.048429149302271e-06, | |
| "loss": 0.1446, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.672064777327935, | |
| "grad_norm": 0.46281240323916106, | |
| "learning_rate": 8.003131148579902e-06, | |
| "loss": 0.1427, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.6761133603238867, | |
| "grad_norm": 0.4515424372637035, | |
| "learning_rate": 7.957929074501182e-06, | |
| "loss": 0.1416, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.6801619433198383, | |
| "grad_norm": 0.45099098942897, | |
| "learning_rate": 7.912823288500695e-06, | |
| "loss": 0.1455, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.44389797544779463, | |
| "learning_rate": 7.867814151243102e-06, | |
| "loss": 0.1413, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.688259109311741, | |
| "grad_norm": 0.4444229598255846, | |
| "learning_rate": 7.822902022620267e-06, | |
| "loss": 0.1435, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.4423378290771326, | |
| "learning_rate": 7.778087261748377e-06, | |
| "loss": 0.1433, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.6963562753036436, | |
| "grad_norm": 0.44519694207102095, | |
| "learning_rate": 7.733370226965056e-06, | |
| "loss": 0.1411, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.700404858299595, | |
| "grad_norm": 0.4378872717248358, | |
| "learning_rate": 7.688751275826527e-06, | |
| "loss": 0.1386, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.7044534412955468, | |
| "grad_norm": 0.4358615100161545, | |
| "learning_rate": 7.644230765104735e-06, | |
| "loss": 0.1452, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.708502024291498, | |
| "grad_norm": 0.4402558567458703, | |
| "learning_rate": 7.599809050784502e-06, | |
| "loss": 0.1372, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.7125506072874495, | |
| "grad_norm": 0.44669000434106704, | |
| "learning_rate": 7.555486488060661e-06, | |
| "loss": 0.138, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.716599190283401, | |
| "grad_norm": 0.4423369519789511, | |
| "learning_rate": 7.511263431335249e-06, | |
| "loss": 0.1412, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.720647773279352, | |
| "grad_norm": 0.45362473940757075, | |
| "learning_rate": 7.467140234214656e-06, | |
| "loss": 0.1463, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.7246963562753037, | |
| "grad_norm": 0.44040579182290995, | |
| "learning_rate": 7.423117249506777e-06, | |
| "loss": 0.1406, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.7287449392712553, | |
| "grad_norm": 0.4436725620283639, | |
| "learning_rate": 7.379194829218231e-06, | |
| "loss": 0.1391, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.7327935222672064, | |
| "grad_norm": 0.43278502537875696, | |
| "learning_rate": 7.335373324551518e-06, | |
| "loss": 0.1385, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.4397190727883781, | |
| "learning_rate": 7.291653085902224e-06, | |
| "loss": 0.1424, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.7408906882591095, | |
| "grad_norm": 0.4496558941626698, | |
| "learning_rate": 7.248034462856213e-06, | |
| "loss": 0.1474, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.7449392712550607, | |
| "grad_norm": 0.4410274102928882, | |
| "learning_rate": 7.20451780418683e-06, | |
| "loss": 0.1388, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.748987854251012, | |
| "grad_norm": 0.4406318538547849, | |
| "learning_rate": 7.161103457852128e-06, | |
| "loss": 0.1398, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.753036437246964, | |
| "grad_norm": 0.4369886444176131, | |
| "learning_rate": 7.117791770992046e-06, | |
| "loss": 0.1426, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.757085020242915, | |
| "grad_norm": 0.43898967284141305, | |
| "learning_rate": 7.074583089925688e-06, | |
| "loss": 0.1378, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.7611336032388665, | |
| "grad_norm": 0.4434899888314164, | |
| "learning_rate": 7.0314777601485155e-06, | |
| "loss": 0.1436, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.765182186234818, | |
| "grad_norm": 0.43909438089725916, | |
| "learning_rate": 6.988476126329604e-06, | |
| "loss": 0.1387, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.42819508967183767, | |
| "learning_rate": 6.945578532308857e-06, | |
| "loss": 0.1398, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.7732793522267207, | |
| "grad_norm": 0.4277133634691284, | |
| "learning_rate": 6.902785321094301e-06, | |
| "loss": 0.1419, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.7773279352226723, | |
| "grad_norm": 0.44318379230186467, | |
| "learning_rate": 6.8600968348593115e-06, | |
| "loss": 0.1438, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.7813765182186234, | |
| "grad_norm": 0.4416585432584095, | |
| "learning_rate": 6.817513414939892e-06, | |
| "loss": 0.1474, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.785425101214575, | |
| "grad_norm": 0.44300943229915374, | |
| "learning_rate": 6.775035401831924e-06, | |
| "loss": 0.1373, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.44409027456429795, | |
| "learning_rate": 6.732663135188473e-06, | |
| "loss": 0.1421, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.7935222672064777, | |
| "grad_norm": 0.43969018702885254, | |
| "learning_rate": 6.690396953817055e-06, | |
| "loss": 0.1453, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.7975708502024292, | |
| "grad_norm": 0.43363207297909717, | |
| "learning_rate": 6.648237195676927e-06, | |
| "loss": 0.1381, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.801619433198381, | |
| "grad_norm": 0.44099721930092983, | |
| "learning_rate": 6.606184197876395e-06, | |
| "loss": 0.1396, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.805668016194332, | |
| "grad_norm": 0.4615089801294735, | |
| "learning_rate": 6.564238296670109e-06, | |
| "loss": 0.1435, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.8097165991902835, | |
| "grad_norm": 0.4309995961318543, | |
| "learning_rate": 6.52239982745638e-06, | |
| "loss": 0.1434, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.813765182186235, | |
| "grad_norm": 0.4535262678948814, | |
| "learning_rate": 6.480669124774484e-06, | |
| "loss": 0.1472, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.817813765182186, | |
| "grad_norm": 0.4474811138109604, | |
| "learning_rate": 6.439046522302013e-06, | |
| "loss": 0.1422, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.8218623481781377, | |
| "grad_norm": 0.44456153274632365, | |
| "learning_rate": 6.397532352852191e-06, | |
| "loss": 0.1379, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.8259109311740893, | |
| "grad_norm": 0.43872938359285285, | |
| "learning_rate": 6.356126948371202e-06, | |
| "loss": 0.1419, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.8299595141700404, | |
| "grad_norm": 0.45010795244005897, | |
| "learning_rate": 6.314830639935559e-06, | |
| "loss": 0.1394, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.834008097165992, | |
| "grad_norm": 0.44206977494902094, | |
| "learning_rate": 6.273643757749448e-06, | |
| "loss": 0.141, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.8380566801619436, | |
| "grad_norm": 0.4355176954127249, | |
| "learning_rate": 6.2325666311420856e-06, | |
| "loss": 0.1425, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.43351427467457854, | |
| "learning_rate": 6.1915995885650675e-06, | |
| "loss": 0.136, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.4437481038688106, | |
| "learning_rate": 6.150742957589784e-06, | |
| "loss": 0.1435, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.850202429149798, | |
| "grad_norm": 0.4468343263482415, | |
| "learning_rate": 6.109997064904774e-06, | |
| "loss": 0.1394, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.854251012145749, | |
| "grad_norm": 0.4434575464398586, | |
| "learning_rate": 6.069362236313092e-06, | |
| "loss": 0.1395, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.8582995951417005, | |
| "grad_norm": 0.44765330289879895, | |
| "learning_rate": 6.0288387967297565e-06, | |
| "loss": 0.1402, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.862348178137652, | |
| "grad_norm": 0.441068229517719, | |
| "learning_rate": 5.988427070179108e-06, | |
| "loss": 0.1389, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.866396761133603, | |
| "grad_norm": 0.4421770693777091, | |
| "learning_rate": 5.948127379792234e-06, | |
| "loss": 0.1407, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.8704453441295548, | |
| "grad_norm": 0.44073028096292866, | |
| "learning_rate": 5.90794004780439e-06, | |
| "loss": 0.1376, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.8744939271255063, | |
| "grad_norm": 0.4457714006935401, | |
| "learning_rate": 5.867865395552405e-06, | |
| "loss": 0.1373, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.8785425101214575, | |
| "grad_norm": 0.4467440708969397, | |
| "learning_rate": 5.8279037434721364e-06, | |
| "loss": 0.1428, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.882591093117409, | |
| "grad_norm": 0.438603582550361, | |
| "learning_rate": 5.788055411095892e-06, | |
| "loss": 0.1419, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.8866396761133606, | |
| "grad_norm": 0.44396389877003933, | |
| "learning_rate": 5.7483207170498625e-06, | |
| "loss": 0.1422, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.8906882591093117, | |
| "grad_norm": 0.44520196351229385, | |
| "learning_rate": 5.708699979051613e-06, | |
| "loss": 0.1391, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.4296660140908561, | |
| "learning_rate": 5.66919351390751e-06, | |
| "loss": 0.1332, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.898785425101215, | |
| "grad_norm": 0.43453482089540413, | |
| "learning_rate": 5.629801637510188e-06, | |
| "loss": 0.1342, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.902834008097166, | |
| "grad_norm": 0.441319122990895, | |
| "learning_rate": 5.590524664836043e-06, | |
| "loss": 0.1369, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.9068825910931175, | |
| "grad_norm": 0.4289269998859761, | |
| "learning_rate": 5.551362909942708e-06, | |
| "loss": 0.1408, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.910931174089069, | |
| "grad_norm": 0.4409891890364828, | |
| "learning_rate": 5.51231668596653e-06, | |
| "loss": 0.1419, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.91497975708502, | |
| "grad_norm": 0.4243578081980491, | |
| "learning_rate": 5.47338630512007e-06, | |
| "loss": 0.1361, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.919028340080972, | |
| "grad_norm": 0.4297281117544812, | |
| "learning_rate": 5.434572078689617e-06, | |
| "loss": 0.1451, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.9230769230769234, | |
| "grad_norm": 0.4231791514570246, | |
| "learning_rate": 5.395874317032694e-06, | |
| "loss": 0.1314, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.9271255060728745, | |
| "grad_norm": 0.43877080151820286, | |
| "learning_rate": 5.357293329575573e-06, | |
| "loss": 0.1363, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.931174089068826, | |
| "grad_norm": 0.44732531762006117, | |
| "learning_rate": 5.318829424810797e-06, | |
| "loss": 0.1405, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.9352226720647776, | |
| "grad_norm": 0.4448550314498391, | |
| "learning_rate": 5.280482910294728e-06, | |
| "loss": 0.1484, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.9392712550607287, | |
| "grad_norm": 0.4294622236518271, | |
| "learning_rate": 5.242254092645078e-06, | |
| "loss": 0.1417, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.9433198380566803, | |
| "grad_norm": 0.4405113779241834, | |
| "learning_rate": 5.2041432775384445e-06, | |
| "loss": 0.1356, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.443436552463814, | |
| "learning_rate": 5.166150769707892e-06, | |
| "loss": 0.1393, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.951417004048583, | |
| "grad_norm": 0.4274487767398105, | |
| "learning_rate": 5.128276872940503e-06, | |
| "loss": 0.1394, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.9554655870445345, | |
| "grad_norm": 0.43402509423456803, | |
| "learning_rate": 5.090521890074953e-06, | |
| "loss": 0.1377, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.9595141700404857, | |
| "grad_norm": 0.42937241961045475, | |
| "learning_rate": 5.0528861229990675e-06, | |
| "loss": 0.1378, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.9635627530364372, | |
| "grad_norm": 0.43018401958362645, | |
| "learning_rate": 5.015369872647442e-06, | |
| "loss": 0.1332, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.967611336032389, | |
| "grad_norm": 0.4298287347710815, | |
| "learning_rate": 4.9779734389990195e-06, | |
| "loss": 0.1387, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97165991902834, | |
| "grad_norm": 0.42345338717441466, | |
| "learning_rate": 4.940697121074675e-06, | |
| "loss": 0.1369, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.9757085020242915, | |
| "grad_norm": 0.4434250699426475, | |
| "learning_rate": 4.903541216934864e-06, | |
| "loss": 0.1405, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.979757085020243, | |
| "grad_norm": 0.4307347290328425, | |
| "learning_rate": 4.8665060236772e-06, | |
| "loss": 0.1371, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.983805668016194, | |
| "grad_norm": 0.428205175753726, | |
| "learning_rate": 4.829591837434102e-06, | |
| "loss": 0.1373, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.9878542510121457, | |
| "grad_norm": 0.438468899462132, | |
| "learning_rate": 4.792798953370424e-06, | |
| "loss": 0.1375, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.9919028340080973, | |
| "grad_norm": 0.4376703322862341, | |
| "learning_rate": 4.756127665681083e-06, | |
| "loss": 0.1428, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.9959514170040484, | |
| "grad_norm": 0.43470080225810726, | |
| "learning_rate": 4.719578267588725e-06, | |
| "loss": 0.1423, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.4155659275587284, | |
| "learning_rate": 4.6831510513413525e-06, | |
| "loss": 0.1288, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 4.004048582995951, | |
| "grad_norm": 0.4194445299984874, | |
| "learning_rate": 4.6468463082100176e-06, | |
| "loss": 0.0701, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 4.008097165991903, | |
| "grad_norm": 0.3879315254176349, | |
| "learning_rate": 4.610664328486485e-06, | |
| "loss": 0.0685, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.012145748987854, | |
| "grad_norm": 0.3829313254146428, | |
| "learning_rate": 4.5746054014809e-06, | |
| "loss": 0.0708, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 4.016194331983805, | |
| "grad_norm": 0.3438998363731362, | |
| "learning_rate": 4.5386698155194766e-06, | |
| "loss": 0.0713, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.020242914979757, | |
| "grad_norm": 0.3221541172697996, | |
| "learning_rate": 4.502857857942204e-06, | |
| "loss": 0.0686, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 4.0242914979757085, | |
| "grad_norm": 0.3084234402790859, | |
| "learning_rate": 4.467169815100549e-06, | |
| "loss": 0.0648, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 4.02834008097166, | |
| "grad_norm": 0.3293354837655308, | |
| "learning_rate": 4.431605972355137e-06, | |
| "loss": 0.0726, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.032388663967612, | |
| "grad_norm": 0.32500558539477475, | |
| "learning_rate": 4.396166614073516e-06, | |
| "loss": 0.0677, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 4.036437246963563, | |
| "grad_norm": 0.34145299794834927, | |
| "learning_rate": 4.360852023627846e-06, | |
| "loss": 0.0667, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 4.040485829959514, | |
| "grad_norm": 0.34790705014156625, | |
| "learning_rate": 4.325662483392656e-06, | |
| "loss": 0.0638, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 4.044534412955466, | |
| "grad_norm": 0.35760507596691493, | |
| "learning_rate": 4.29059827474257e-06, | |
| "loss": 0.0631, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 4.048582995951417, | |
| "grad_norm": 0.37399129835443823, | |
| "learning_rate": 4.255659678050063e-06, | |
| "loss": 0.0658, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.3745746156592753, | |
| "learning_rate": 4.22084697268323e-06, | |
| "loss": 0.0671, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 4.05668016194332, | |
| "grad_norm": 0.3631544190316488, | |
| "learning_rate": 4.186160437003535e-06, | |
| "loss": 0.0633, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 4.060728744939271, | |
| "grad_norm": 0.35783761375321377, | |
| "learning_rate": 4.151600348363582e-06, | |
| "loss": 0.0662, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 4.064777327935222, | |
| "grad_norm": 0.3494088537201085, | |
| "learning_rate": 4.11716698310493e-06, | |
| "loss": 0.0648, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 4.068825910931174, | |
| "grad_norm": 0.32839464570432897, | |
| "learning_rate": 4.082860616555848e-06, | |
| "loss": 0.0624, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.0728744939271255, | |
| "grad_norm": 0.33443553083644145, | |
| "learning_rate": 4.048681523029125e-06, | |
| "loss": 0.0643, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 4.076923076923077, | |
| "grad_norm": 0.3217765575227432, | |
| "learning_rate": 4.014629975819882e-06, | |
| "loss": 0.0619, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 4.080971659919029, | |
| "grad_norm": 0.3135530985676136, | |
| "learning_rate": 3.9807062472033855e-06, | |
| "loss": 0.064, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 4.08502024291498, | |
| "grad_norm": 0.32769397666994604, | |
| "learning_rate": 3.94691060843287e-06, | |
| "loss": 0.0649, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 4.089068825910931, | |
| "grad_norm": 0.30921012114762886, | |
| "learning_rate": 3.913243329737353e-06, | |
| "loss": 0.0626, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.093117408906883, | |
| "grad_norm": 0.3095560590795697, | |
| "learning_rate": 3.879704680319503e-06, | |
| "loss": 0.0673, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 4.097165991902834, | |
| "grad_norm": 0.29963503643894746, | |
| "learning_rate": 3.846294928353471e-06, | |
| "loss": 0.0636, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 4.101214574898785, | |
| "grad_norm": 0.3034610416928676, | |
| "learning_rate": 3.8130143409827345e-06, | |
| "loss": 0.0639, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.32040400240830275, | |
| "learning_rate": 3.7798631843179887e-06, | |
| "loss": 0.0649, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 4.109311740890688, | |
| "grad_norm": 0.3040983721575812, | |
| "learning_rate": 3.7468417234349973e-06, | |
| "loss": 0.0621, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.113360323886639, | |
| "grad_norm": 0.3052080352048146, | |
| "learning_rate": 3.7139502223724823e-06, | |
| "loss": 0.0652, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 4.117408906882591, | |
| "grad_norm": 0.3104113852442432, | |
| "learning_rate": 3.6811889441300095e-06, | |
| "loss": 0.0656, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 4.1214574898785425, | |
| "grad_norm": 0.29982175429523356, | |
| "learning_rate": 3.648558150665886e-06, | |
| "loss": 0.0598, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 4.125506072874494, | |
| "grad_norm": 0.30875678948190116, | |
| "learning_rate": 3.616058102895075e-06, | |
| "loss": 0.0633, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 4.129554655870446, | |
| "grad_norm": 0.32162001139719015, | |
| "learning_rate": 3.583689060687079e-06, | |
| "loss": 0.0659, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.133603238866397, | |
| "grad_norm": 0.30829060649471884, | |
| "learning_rate": 3.5514512828639023e-06, | |
| "loss": 0.0632, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 4.137651821862348, | |
| "grad_norm": 0.30346330011821016, | |
| "learning_rate": 3.51934502719796e-06, | |
| "loss": 0.0617, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 4.1417004048583, | |
| "grad_norm": 0.31011321722848206, | |
| "learning_rate": 3.4873705504100185e-06, | |
| "loss": 0.0623, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 4.145748987854251, | |
| "grad_norm": 0.30928083648631294, | |
| "learning_rate": 3.4555281081671345e-06, | |
| "loss": 0.0607, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 4.149797570850202, | |
| "grad_norm": 0.31053812544910947, | |
| "learning_rate": 3.4238179550806326e-06, | |
| "loss": 0.0613, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 0.3149727601213779, | |
| "learning_rate": 3.3922403447040497e-06, | |
| "loss": 0.061, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.3197296087004964, | |
| "learning_rate": 3.3607955295311247e-06, | |
| "loss": 0.0631, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 4.161943319838056, | |
| "grad_norm": 0.3233974301950112, | |
| "learning_rate": 3.32948376099375e-06, | |
| "loss": 0.0651, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 4.165991902834008, | |
| "grad_norm": 0.3171322249600333, | |
| "learning_rate": 3.2983052894599976e-06, | |
| "loss": 0.0609, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 4.17004048582996, | |
| "grad_norm": 0.3239740710040186, | |
| "learning_rate": 3.2672603642320945e-06, | |
| "loss": 0.0648, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.174089068825911, | |
| "grad_norm": 0.31861964344047106, | |
| "learning_rate": 3.236349233544438e-06, | |
| "loss": 0.0662, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 4.178137651821863, | |
| "grad_norm": 0.3316440372659825, | |
| "learning_rate": 3.205572144561604e-06, | |
| "loss": 0.0642, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.182186234817814, | |
| "grad_norm": 0.30800554683390746, | |
| "learning_rate": 3.174929343376374e-06, | |
| "loss": 0.0622, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 4.186234817813765, | |
| "grad_norm": 0.3180684285768857, | |
| "learning_rate": 3.1444210750077777e-06, | |
| "loss": 0.0642, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 4.190283400809717, | |
| "grad_norm": 0.32556244942816004, | |
| "learning_rate": 3.1140475833991046e-06, | |
| "loss": 0.0636, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.194331983805668, | |
| "grad_norm": 0.31792932532411483, | |
| "learning_rate": 3.0838091114159895e-06, | |
| "loss": 0.0626, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 4.198380566801619, | |
| "grad_norm": 0.3088711216365181, | |
| "learning_rate": 3.0537059008444524e-06, | |
| "loss": 0.0591, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 4.202429149797571, | |
| "grad_norm": 0.3201336131459994, | |
| "learning_rate": 3.0237381923889563e-06, | |
| "loss": 0.0654, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 4.206477732793522, | |
| "grad_norm": 0.30560905898205404, | |
| "learning_rate": 2.9939062256704997e-06, | |
| "loss": 0.0628, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.3028512624546449, | |
| "learning_rate": 2.964210239224696e-06, | |
| "loss": 0.0613, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.2145748987854255, | |
| "grad_norm": 0.30985111654052705, | |
| "learning_rate": 2.934650470499869e-06, | |
| "loss": 0.0607, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 4.218623481781377, | |
| "grad_norm": 0.30971513823225344, | |
| "learning_rate": 2.905227155855128e-06, | |
| "loss": 0.0622, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 4.222672064777328, | |
| "grad_norm": 0.3143819386968968, | |
| "learning_rate": 2.875940530558519e-06, | |
| "loss": 0.0627, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 4.22672064777328, | |
| "grad_norm": 0.3229064011678917, | |
| "learning_rate": 2.8467908287851154e-06, | |
| "loss": 0.0653, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.313938687247288, | |
| "learning_rate": 2.8177782836151603e-06, | |
| "loss": 0.0623, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.234817813765182, | |
| "grad_norm": 0.3109687634250265, | |
| "learning_rate": 2.7889031270321874e-06, | |
| "loss": 0.0665, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 4.238866396761134, | |
| "grad_norm": 0.30506580820047297, | |
| "learning_rate": 2.760165589921184e-06, | |
| "loss": 0.0611, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 4.242914979757085, | |
| "grad_norm": 0.31578857825690015, | |
| "learning_rate": 2.7315659020667305e-06, | |
| "loss": 0.0612, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 4.246963562753036, | |
| "grad_norm": 0.3139910453613134, | |
| "learning_rate": 2.703104292151164e-06, | |
| "loss": 0.0636, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 4.251012145748988, | |
| "grad_norm": 0.3058748833904642, | |
| "learning_rate": 2.6747809877527607e-06, | |
| "loss": 0.0622, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.255060728744939, | |
| "grad_norm": 0.3100277650211013, | |
| "learning_rate": 2.6465962153439082e-06, | |
| "loss": 0.0634, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.2591093117408905, | |
| "grad_norm": 0.31486640942796223, | |
| "learning_rate": 2.618550200289303e-06, | |
| "loss": 0.0633, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.31973887313048277, | |
| "learning_rate": 2.59064316684412e-06, | |
| "loss": 0.0631, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.267206477732794, | |
| "grad_norm": 0.30809865633784933, | |
| "learning_rate": 2.562875338152264e-06, | |
| "loss": 0.0591, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.271255060728745, | |
| "grad_norm": 0.31451553133898347, | |
| "learning_rate": 2.5352469362445552e-06, | |
| "loss": 0.0633, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.275303643724697, | |
| "grad_norm": 0.3102949935970655, | |
| "learning_rate": 2.5077581820369523e-06, | |
| "loss": 0.064, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.279352226720648, | |
| "grad_norm": 0.30895431500739745, | |
| "learning_rate": 2.480409295328803e-06, | |
| "loss": 0.062, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.283400809716599, | |
| "grad_norm": 0.3037873724331983, | |
| "learning_rate": 2.45320049480108e-06, | |
| "loss": 0.0611, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.287449392712551, | |
| "grad_norm": 0.3251287971593738, | |
| "learning_rate": 2.4261319980146293e-06, | |
| "loss": 0.0676, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.291497975708502, | |
| "grad_norm": 0.316897798574019, | |
| "learning_rate": 2.399204021408421e-06, | |
| "loss": 0.0623, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.295546558704453, | |
| "grad_norm": 0.29516751484031023, | |
| "learning_rate": 2.3724167802978395e-06, | |
| "loss": 0.0591, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.299595141700405, | |
| "grad_norm": 0.31361180927979865, | |
| "learning_rate": 2.3457704888729625e-06, | |
| "loss": 0.0647, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.303643724696356, | |
| "grad_norm": 0.3080445317404377, | |
| "learning_rate": 2.3192653601968162e-06, | |
| "loss": 0.0601, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.3076923076923075, | |
| "grad_norm": 0.32171006663391993, | |
| "learning_rate": 2.2929016062037057e-06, | |
| "loss": 0.0622, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.3117408906882595, | |
| "grad_norm": 0.3166525792985942, | |
| "learning_rate": 2.2666794376975055e-06, | |
| "loss": 0.0615, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.31251954613721206, | |
| "learning_rate": 2.24059906434998e-06, | |
| "loss": 0.0622, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.319838056680162, | |
| "grad_norm": 0.31190489887128964, | |
| "learning_rate": 2.214660694699091e-06, | |
| "loss": 0.064, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.323886639676114, | |
| "grad_norm": 0.3149952391301792, | |
| "learning_rate": 2.188864536147355e-06, | |
| "loss": 0.0621, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.327935222672065, | |
| "grad_norm": 0.3060507787067717, | |
| "learning_rate": 2.1632107949601757e-06, | |
| "loss": 0.064, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.331983805668016, | |
| "grad_norm": 0.3110095058017396, | |
| "learning_rate": 2.137699676264171e-06, | |
| "loss": 0.0597, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.336032388663968, | |
| "grad_norm": 0.3227763529484236, | |
| "learning_rate": 2.1123313840455738e-06, | |
| "loss": 0.0651, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.340080971659919, | |
| "grad_norm": 0.3141965504353822, | |
| "learning_rate": 2.08710612114857e-06, | |
| "loss": 0.0637, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.34412955465587, | |
| "grad_norm": 0.30942699206019525, | |
| "learning_rate": 2.0620240892736953e-06, | |
| "loss": 0.0595, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.348178137651822, | |
| "grad_norm": 0.30169528167189885, | |
| "learning_rate": 2.0370854889762e-06, | |
| "loss": 0.0589, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.352226720647773, | |
| "grad_norm": 0.30316474143115607, | |
| "learning_rate": 2.0122905196644726e-06, | |
| "loss": 0.0569, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.3562753036437245, | |
| "grad_norm": 0.3015201542747948, | |
| "learning_rate": 1.987639379598427e-06, | |
| "loss": 0.0602, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.3603238866396765, | |
| "grad_norm": 0.31461625566768536, | |
| "learning_rate": 1.963132265887919e-06, | |
| "loss": 0.0636, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.364372469635628, | |
| "grad_norm": 0.3118293483387524, | |
| "learning_rate": 1.9387693744911784e-06, | |
| "loss": 0.0609, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.31948955473879326, | |
| "learning_rate": 1.9145509002132368e-06, | |
| "loss": 0.0641, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.372469635627531, | |
| "grad_norm": 0.306987549065969, | |
| "learning_rate": 1.8904770367043723e-06, | |
| "loss": 0.0595, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.376518218623482, | |
| "grad_norm": 0.32133791225167063, | |
| "learning_rate": 1.8665479764585504e-06, | |
| "loss": 0.0648, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.380566801619433, | |
| "grad_norm": 0.30940162376895736, | |
| "learning_rate": 1.8427639108118977e-06, | |
| "loss": 0.0641, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.384615384615385, | |
| "grad_norm": 0.30942537280886284, | |
| "learning_rate": 1.8191250299411755e-06, | |
| "loss": 0.0592, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.388663967611336, | |
| "grad_norm": 0.31045727400606515, | |
| "learning_rate": 1.7956315228622446e-06, | |
| "loss": 0.0621, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.392712550607287, | |
| "grad_norm": 0.32467698976372433, | |
| "learning_rate": 1.7722835774285553e-06, | |
| "loss": 0.065, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.396761133603239, | |
| "grad_norm": 0.3065212155517021, | |
| "learning_rate": 1.7490813803296624e-06, | |
| "loss": 0.061, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.40080971659919, | |
| "grad_norm": 0.3133006621678502, | |
| "learning_rate": 1.726025117089718e-06, | |
| "loss": 0.0617, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.4048582995951415, | |
| "grad_norm": 0.3086996742608691, | |
| "learning_rate": 1.70311497206598e-06, | |
| "loss": 0.0594, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.4089068825910935, | |
| "grad_norm": 0.30566665766163886, | |
| "learning_rate": 1.6803511284473638e-06, | |
| "loss": 0.0586, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.412955465587045, | |
| "grad_norm": 0.3107263513311672, | |
| "learning_rate": 1.6577337682529538e-06, | |
| "loss": 0.0625, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.417004048582996, | |
| "grad_norm": 0.3092036740207505, | |
| "learning_rate": 1.6352630723305617e-06, | |
| "loss": 0.0633, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.315577453029893, | |
| "learning_rate": 1.612939220355272e-06, | |
| "loss": 0.0628, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.425101214574899, | |
| "grad_norm": 0.31214373050026006, | |
| "learning_rate": 1.5907623908280112e-06, | |
| "loss": 0.0645, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.42914979757085, | |
| "grad_norm": 0.31323281236292333, | |
| "learning_rate": 1.568732761074121e-06, | |
| "loss": 0.0635, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.433198380566802, | |
| "grad_norm": 0.3314905469825467, | |
| "learning_rate": 1.5468505072419237e-06, | |
| "loss": 0.0665, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.437246963562753, | |
| "grad_norm": 0.3049043414886667, | |
| "learning_rate": 1.5251158043013448e-06, | |
| "loss": 0.058, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.441295546558704, | |
| "grad_norm": 0.31647355639539254, | |
| "learning_rate": 1.5035288260424885e-06, | |
| "loss": 0.0646, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.445344129554655, | |
| "grad_norm": 0.3086852223682441, | |
| "learning_rate": 1.4820897450742623e-06, | |
| "loss": 0.062, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.449392712550607, | |
| "grad_norm": 0.3038431638645785, | |
| "learning_rate": 1.4607987328229767e-06, | |
| "loss": 0.0589, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.4534412955465585, | |
| "grad_norm": 0.3206868120636026, | |
| "learning_rate": 1.4396559595310056e-06, | |
| "loss": 0.0641, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.4574898785425106, | |
| "grad_norm": 0.302570998930909, | |
| "learning_rate": 1.4186615942553973e-06, | |
| "loss": 0.0588, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.461538461538462, | |
| "grad_norm": 0.3105321530143476, | |
| "learning_rate": 1.3978158048665424e-06, | |
| "loss": 0.0606, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.465587044534413, | |
| "grad_norm": 0.3106403844024443, | |
| "learning_rate": 1.3771187580468115e-06, | |
| "loss": 0.0616, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.469635627530364, | |
| "grad_norm": 0.3098029511740475, | |
| "learning_rate": 1.3565706192892392e-06, | |
| "loss": 0.0627, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.30831379731854, | |
| "learning_rate": 1.3361715528961993e-06, | |
| "loss": 0.061, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.477732793522267, | |
| "grad_norm": 0.3171480748865212, | |
| "learning_rate": 1.3159217219780817e-06, | |
| "loss": 0.0637, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.481781376518219, | |
| "grad_norm": 0.3205497889371435, | |
| "learning_rate": 1.2958212884519949e-06, | |
| "loss": 0.0638, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.48582995951417, | |
| "grad_norm": 0.3110517288523301, | |
| "learning_rate": 1.2758704130404675e-06, | |
| "loss": 0.0608, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.489878542510121, | |
| "grad_norm": 0.32276654120251874, | |
| "learning_rate": 1.2560692552701738e-06, | |
| "loss": 0.063, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.493927125506072, | |
| "grad_norm": 0.31521757914686666, | |
| "learning_rate": 1.2364179734706382e-06, | |
| "loss": 0.0608, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.497975708502024, | |
| "grad_norm": 0.3091178327927371, | |
| "learning_rate": 1.2169167247729852e-06, | |
| "loss": 0.0625, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.502024291497976, | |
| "grad_norm": 0.30861461668289514, | |
| "learning_rate": 1.1975656651086864e-06, | |
| "loss": 0.0614, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.506072874493928, | |
| "grad_norm": 0.30471404194635815, | |
| "learning_rate": 1.1783649492082926e-06, | |
| "loss": 0.061, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.510121457489879, | |
| "grad_norm": 0.304560821905185, | |
| "learning_rate": 1.1593147306002183e-06, | |
| "loss": 0.0597, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.51417004048583, | |
| "grad_norm": 0.31382574008301617, | |
| "learning_rate": 1.140415161609505e-06, | |
| "loss": 0.0645, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.518218623481781, | |
| "grad_norm": 0.3124966908627871, | |
| "learning_rate": 1.121666393356604e-06, | |
| "loss": 0.0664, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.522267206477733, | |
| "grad_norm": 0.3061817136455212, | |
| "learning_rate": 1.1030685757561632e-06, | |
| "loss": 0.0607, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.3091083120275891, | |
| "learning_rate": 1.0846218575158373e-06, | |
| "loss": 0.0635, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.530364372469636, | |
| "grad_norm": 0.3076678066287034, | |
| "learning_rate": 1.066326386135097e-06, | |
| "loss": 0.063, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.534412955465587, | |
| "grad_norm": 0.3013185419746391, | |
| "learning_rate": 1.0481823079040398e-06, | |
| "loss": 0.0598, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.538461538461538, | |
| "grad_norm": 0.3119265732204636, | |
| "learning_rate": 1.0301897679022344e-06, | |
| "loss": 0.067, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.5425101214574894, | |
| "grad_norm": 0.30527348355556383, | |
| "learning_rate": 1.0123489099975491e-06, | |
| "loss": 0.0638, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.5465587044534415, | |
| "grad_norm": 0.3027475433196661, | |
| "learning_rate": 9.9465987684501e-07, | |
| "loss": 0.0594, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.550607287449393, | |
| "grad_norm": 0.30182872449989334, | |
| "learning_rate": 9.771228098856534e-07, | |
| "loss": 0.0579, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.554655870445345, | |
| "grad_norm": 0.30594650858613515, | |
| "learning_rate": 9.597378493454013e-07, | |
| "loss": 0.06, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.558704453441296, | |
| "grad_norm": 0.3155373268365066, | |
| "learning_rate": 9.425051342339353e-07, | |
| "loss": 0.0616, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.562753036437247, | |
| "grad_norm": 0.3054372241954927, | |
| "learning_rate": 9.254248023435864e-07, | |
| "loss": 0.0582, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.566801619433198, | |
| "grad_norm": 0.31866719608535476, | |
| "learning_rate": 9.084969902482288e-07, | |
| "loss": 0.0606, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.57085020242915, | |
| "grad_norm": 0.316639873599575, | |
| "learning_rate": 8.917218333021993e-07, | |
| "loss": 0.0653, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.574898785425101, | |
| "grad_norm": 0.30874481608154764, | |
| "learning_rate": 8.750994656392109e-07, | |
| "loss": 0.0617, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.31099503648311394, | |
| "learning_rate": 8.58630020171265e-07, | |
| "loss": 0.06, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.582995951417004, | |
| "grad_norm": 0.30797677457025163, | |
| "learning_rate": 8.423136285876099e-07, | |
| "loss": 0.0589, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.587044534412955, | |
| "grad_norm": 0.31490288661110916, | |
| "learning_rate": 8.26150421353682e-07, | |
| "loss": 0.0642, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.5910931174089065, | |
| "grad_norm": 0.30830254226126136, | |
| "learning_rate": 8.101405277100549e-07, | |
| "loss": 0.0608, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.5951417004048585, | |
| "grad_norm": 0.3135368461812759, | |
| "learning_rate": 7.942840756714077e-07, | |
| "loss": 0.0632, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.59919028340081, | |
| "grad_norm": 0.3202985530001462, | |
| "learning_rate": 7.785811920255093e-07, | |
| "loss": 0.0642, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.603238866396762, | |
| "grad_norm": 0.3134161587718647, | |
| "learning_rate": 7.630320023322024e-07, | |
| "loss": 0.0621, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.607287449392713, | |
| "grad_norm": 0.29883122380854393, | |
| "learning_rate": 7.476366309223903e-07, | |
| "loss": 0.0604, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.611336032388664, | |
| "grad_norm": 0.3060885226522882, | |
| "learning_rate": 7.323952008970537e-07, | |
| "loss": 0.0602, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.3010226767927793, | |
| "learning_rate": 7.173078341262618e-07, | |
| "loss": 0.0595, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.619433198380567, | |
| "grad_norm": 0.30987905116028747, | |
| "learning_rate": 7.023746512481989e-07, | |
| "loss": 0.0636, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.623481781376518, | |
| "grad_norm": 0.2953244821671931, | |
| "learning_rate": 6.875957716681902e-07, | |
| "loss": 0.0619, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.62753036437247, | |
| "grad_norm": 0.30456854954799356, | |
| "learning_rate": 6.729713135577687e-07, | |
| "loss": 0.0622, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.31471837177260653, | |
| "learning_rate": 6.585013938537143e-07, | |
| "loss": 0.0639, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.635627530364372, | |
| "grad_norm": 0.31016308202605813, | |
| "learning_rate": 6.441861282571115e-07, | |
| "loss": 0.063, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.6396761133603235, | |
| "grad_norm": 0.3093380662636829, | |
| "learning_rate": 6.300256312324471e-07, | |
| "loss": 0.0606, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.6437246963562755, | |
| "grad_norm": 0.30663696904258864, | |
| "learning_rate": 6.160200160066776e-07, | |
| "loss": 0.0598, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.647773279352227, | |
| "grad_norm": 0.3227469960910792, | |
| "learning_rate": 6.021693945683327e-07, | |
| "loss": 0.066, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.651821862348179, | |
| "grad_norm": 0.3162722820945616, | |
| "learning_rate": 5.884738776666044e-07, | |
| "loss": 0.062, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.65587044534413, | |
| "grad_norm": 0.30405442097239277, | |
| "learning_rate": 5.749335748104878e-07, | |
| "loss": 0.0629, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.659919028340081, | |
| "grad_norm": 0.3200512554123665, | |
| "learning_rate": 5.615485942678778e-07, | |
| "loss": 0.0639, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.663967611336032, | |
| "grad_norm": 0.3168319399395486, | |
| "learning_rate": 5.483190430647289e-07, | |
| "loss": 0.0598, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.668016194331984, | |
| "grad_norm": 0.3062038108685827, | |
| "learning_rate": 5.352450269841747e-07, | |
| "loss": 0.0586, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.672064777327935, | |
| "grad_norm": 0.30271376728734345, | |
| "learning_rate": 5.22326650565701e-07, | |
| "loss": 0.058, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.676113360323887, | |
| "grad_norm": 0.3212370455884682, | |
| "learning_rate": 5.095640171043048e-07, | |
| "loss": 0.0656, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.680161943319838, | |
| "grad_norm": 0.30083493779751, | |
| "learning_rate": 4.969572286496594e-07, | |
| "loss": 0.0592, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.32009070106020454, | |
| "learning_rate": 4.845063860053079e-07, | |
| "loss": 0.067, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.6882591093117405, | |
| "grad_norm": 0.3105516446717859, | |
| "learning_rate": 4.7221158872785997e-07, | |
| "loss": 0.0583, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.6923076923076925, | |
| "grad_norm": 0.316298655240376, | |
| "learning_rate": 4.600729351261857e-07, | |
| "loss": 0.06, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.696356275303644, | |
| "grad_norm": 0.31425556354432843, | |
| "learning_rate": 4.4809052226063355e-07, | |
| "loss": 0.0635, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.700404858299595, | |
| "grad_norm": 0.3060937495693123, | |
| "learning_rate": 4.3626444594226046e-07, | |
| "loss": 0.0598, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.704453441295547, | |
| "grad_norm": 0.2954955182320145, | |
| "learning_rate": 4.245948007320588e-07, | |
| "loss": 0.0567, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.708502024291498, | |
| "grad_norm": 0.30456056633382217, | |
| "learning_rate": 4.130816799401971e-07, | |
| "loss": 0.0586, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.712550607287449, | |
| "grad_norm": 0.30759459187617744, | |
| "learning_rate": 4.017251756252827e-07, | |
| "loss": 0.0579, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.716599190283401, | |
| "grad_norm": 0.3269776766986231, | |
| "learning_rate": 3.9052537859362473e-07, | |
| "loss": 0.0636, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.720647773279352, | |
| "grad_norm": 0.3072397672523831, | |
| "learning_rate": 3.7948237839849687e-07, | |
| "loss": 0.0609, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.724696356275303, | |
| "grad_norm": 0.3026146045811555, | |
| "learning_rate": 3.6859626333943797e-07, | |
| "loss": 0.0598, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.728744939271255, | |
| "grad_norm": 0.3145552822917229, | |
| "learning_rate": 3.578671204615347e-07, | |
| "loss": 0.0589, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.732793522267206, | |
| "grad_norm": 0.3129268209242089, | |
| "learning_rate": 3.472950355547267e-07, | |
| "loss": 0.0621, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.3069187550705592, | |
| "learning_rate": 3.3688009315312244e-07, | |
| "loss": 0.06, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.7408906882591095, | |
| "grad_norm": 0.3109751102528461, | |
| "learning_rate": 3.266223765343224e-07, | |
| "loss": 0.0615, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.744939271255061, | |
| "grad_norm": 0.3090905777503705, | |
| "learning_rate": 3.16521967718757e-07, | |
| "loss": 0.0613, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.748987854251012, | |
| "grad_norm": 0.31059046747884184, | |
| "learning_rate": 3.0657894746902726e-07, | |
| "loss": 0.0614, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.753036437246964, | |
| "grad_norm": 0.31330039872550147, | |
| "learning_rate": 2.9679339528925654e-07, | |
| "loss": 0.0621, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.757085020242915, | |
| "grad_norm": 0.3081944980337946, | |
| "learning_rate": 2.871653894244597e-07, | |
| "loss": 0.0612, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.761133603238866, | |
| "grad_norm": 0.3037265591715824, | |
| "learning_rate": 2.7769500685991266e-07, | |
| "loss": 0.0552, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.765182186234818, | |
| "grad_norm": 0.315180391756106, | |
| "learning_rate": 2.683823233205485e-07, | |
| "loss": 0.0631, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.769230769230769, | |
| "grad_norm": 0.29781482734212433, | |
| "learning_rate": 2.592274132703287e-07, | |
| "loss": 0.0585, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.77327935222672, | |
| "grad_norm": 0.30584406057707386, | |
| "learning_rate": 2.502303499116754e-07, | |
| "loss": 0.0628, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.777327935222672, | |
| "grad_norm": 0.31101572334964467, | |
| "learning_rate": 2.413912051848666e-07, | |
| "loss": 0.0628, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.781376518218623, | |
| "grad_norm": 0.2996638035223322, | |
| "learning_rate": 2.3271004976746613e-07, | |
| "loss": 0.06, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.7854251012145745, | |
| "grad_norm": 0.3134905137404126, | |
| "learning_rate": 2.2418695307376393e-07, | |
| "loss": 0.0608, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.3082753400934631, | |
| "learning_rate": 2.1582198325420968e-07, | |
| "loss": 0.0602, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.793522267206478, | |
| "grad_norm": 0.30075652578465595, | |
| "learning_rate": 2.0761520719488447e-07, | |
| "loss": 0.0601, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.797570850202429, | |
| "grad_norm": 0.31541221146463533, | |
| "learning_rate": 1.9956669051694798e-07, | |
| "loss": 0.063, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.801619433198381, | |
| "grad_norm": 0.2982518424333577, | |
| "learning_rate": 1.916764975761254e-07, | |
| "loss": 0.062, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.805668016194332, | |
| "grad_norm": 0.3099224446117604, | |
| "learning_rate": 1.839446914621923e-07, | |
| "loss": 0.065, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.809716599190283, | |
| "grad_norm": 0.3074292549415339, | |
| "learning_rate": 1.7637133399845962e-07, | |
| "loss": 0.0607, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.813765182186235, | |
| "grad_norm": 0.2994730632391032, | |
| "learning_rate": 1.6895648574129397e-07, | |
| "loss": 0.0612, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.817813765182186, | |
| "grad_norm": 0.30990599276397, | |
| "learning_rate": 1.6170020597962465e-07, | |
| "loss": 0.062, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.821862348178137, | |
| "grad_norm": 0.31638134598048, | |
| "learning_rate": 1.5460255273447078e-07, | |
| "loss": 0.0612, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.825910931174089, | |
| "grad_norm": 0.3142419880321012, | |
| "learning_rate": 1.4766358275847715e-07, | |
| "loss": 0.0634, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.82995951417004, | |
| "grad_norm": 0.30400853893246754, | |
| "learning_rate": 1.4088335153546573e-07, | |
| "loss": 0.0575, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.834008097165992, | |
| "grad_norm": 0.3106920516591474, | |
| "learning_rate": 1.3426191327998496e-07, | |
| "loss": 0.0645, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.838056680161944, | |
| "grad_norm": 0.3135677774316032, | |
| "learning_rate": 1.277993209368744e-07, | |
| "loss": 0.0658, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.301772733783362, | |
| "learning_rate": 1.214956261808542e-07, | |
| "loss": 0.0591, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.846153846153846, | |
| "grad_norm": 0.31575566627912083, | |
| "learning_rate": 1.1535087941609846e-07, | |
| "loss": 0.065, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.850202429149798, | |
| "grad_norm": 0.318695893171791, | |
| "learning_rate": 1.0936512977583802e-07, | |
| "loss": 0.0606, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.854251012145749, | |
| "grad_norm": 0.30038774652578465, | |
| "learning_rate": 1.0353842512196732e-07, | |
| "loss": 0.0591, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.8582995951417, | |
| "grad_norm": 0.31845495349886527, | |
| "learning_rate": 9.787081204466253e-08, | |
| "loss": 0.0642, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.862348178137652, | |
| "grad_norm": 0.3090378889240763, | |
| "learning_rate": 9.236233586200627e-08, | |
| "loss": 0.0603, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.866396761133603, | |
| "grad_norm": 0.31337149617271176, | |
| "learning_rate": 8.70130406196279e-08, | |
| "loss": 0.0643, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.870445344129554, | |
| "grad_norm": 0.3146627780792552, | |
| "learning_rate": 8.182296909035047e-08, | |
| "loss": 0.0644, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.874493927125506, | |
| "grad_norm": 0.31602599305682766, | |
| "learning_rate": 7.679216277384439e-08, | |
| "loss": 0.0635, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.8785425101214575, | |
| "grad_norm": 0.3160471620831649, | |
| "learning_rate": 7.19206618963053e-08, | |
| "loss": 0.0656, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.882591093117409, | |
| "grad_norm": 0.3190372810375494, | |
| "learning_rate": 6.720850541012347e-08, | |
| "loss": 0.0637, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.886639676113361, | |
| "grad_norm": 0.3062598237953661, | |
| "learning_rate": 6.265573099357714e-08, | |
| "loss": 0.0601, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.890688259109312, | |
| "grad_norm": 0.3057980436500378, | |
| "learning_rate": 5.826237505053067e-08, | |
| "loss": 0.0593, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.31299542761128224, | |
| "learning_rate": 5.4028472710137e-08, | |
| "loss": 0.0626, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.898785425101215, | |
| "grad_norm": 0.3099642937230172, | |
| "learning_rate": 4.9954057826571146e-08, | |
| "loss": 0.06, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.902834008097166, | |
| "grad_norm": 0.301989462855112, | |
| "learning_rate": 4.6039162978743776e-08, | |
| "loss": 0.0595, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.906882591093117, | |
| "grad_norm": 0.300529807136507, | |
| "learning_rate": 4.228381947004812e-08, | |
| "loss": 0.0605, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.910931174089069, | |
| "grad_norm": 0.3085882383542687, | |
| "learning_rate": 3.868805732811343e-08, | |
| "loss": 0.0637, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.91497975708502, | |
| "grad_norm": 0.3060684116744934, | |
| "learning_rate": 3.525190530455636e-08, | |
| "loss": 0.0621, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.919028340080971, | |
| "grad_norm": 0.3286245187006214, | |
| "learning_rate": 3.197539087475887e-08, | |
| "loss": 0.0663, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 0.30919055974770504, | |
| "learning_rate": 2.8858540237641786e-08, | |
| "loss": 0.0618, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.9271255060728745, | |
| "grad_norm": 0.3087464536539745, | |
| "learning_rate": 2.5901378315458247e-08, | |
| "loss": 0.0628, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.931174089068826, | |
| "grad_norm": 0.30194598111060456, | |
| "learning_rate": 2.3103928753598347e-08, | |
| "loss": 0.0609, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.935222672064778, | |
| "grad_norm": 0.3203152851299244, | |
| "learning_rate": 2.0466213920393718e-08, | |
| "loss": 0.0637, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.939271255060729, | |
| "grad_norm": 0.3096754980933418, | |
| "learning_rate": 1.7988254906942115e-08, | |
| "loss": 0.0657, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94331983805668, | |
| "grad_norm": 0.317210770671885, | |
| "learning_rate": 1.5670071526936447e-08, | |
| "loss": 0.0636, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.31131031710350787, | |
| "learning_rate": 1.3511682316509344e-08, | |
| "loss": 0.0621, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.951417004048583, | |
| "grad_norm": 0.3159374439951046, | |
| "learning_rate": 1.1513104534086605e-08, | |
| "loss": 0.0631, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.955465587044534, | |
| "grad_norm": 0.30676100996497896, | |
| "learning_rate": 9.67435416023843e-09, | |
| "loss": 0.0627, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.959514170040486, | |
| "grad_norm": 0.30829826111969133, | |
| "learning_rate": 7.995445897566178e-09, | |
| "loss": 0.0582, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.963562753036437, | |
| "grad_norm": 0.31935616227640934, | |
| "learning_rate": 6.476393170578021e-09, | |
| "loss": 0.0667, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.967611336032388, | |
| "grad_norm": 0.3114554220888009, | |
| "learning_rate": 5.11720812557348e-09, | |
| "loss": 0.0584, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.97165991902834, | |
| "grad_norm": 0.3137743332979701, | |
| "learning_rate": 3.917901630563492e-09, | |
| "loss": 0.0636, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.9757085020242915, | |
| "grad_norm": 0.29829013275449523, | |
| "learning_rate": 2.878483275170485e-09, | |
| "loss": 0.0588, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.979757085020243, | |
| "grad_norm": 0.3015792230945446, | |
| "learning_rate": 1.998961370555108e-09, | |
| "loss": 0.0599, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.983805668016195, | |
| "grad_norm": 0.3221209723965366, | |
| "learning_rate": 1.2793429493518361e-09, | |
| "loss": 0.0653, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.987854251012146, | |
| "grad_norm": 0.31844844201201317, | |
| "learning_rate": 7.196337656112384e-10, | |
| "loss": 0.066, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.991902834008097, | |
| "grad_norm": 0.3028370510621291, | |
| "learning_rate": 3.198382947511292e-10, | |
| "loss": 0.0634, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.995951417004049, | |
| "grad_norm": 0.3099780252026984, | |
| "learning_rate": 7.995973352770137e-11, | |
| "loss": 0.0625, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.29016466202663677, | |
| "learning_rate": 0.0, | |
| "loss": 0.0568, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1235, | |
| "total_flos": 720746397892608.0, | |
| "train_loss": 0.38790296928180373, | |
| "train_runtime": 6364.4209, | |
| "train_samples_per_second": 24.826, | |
| "train_steps_per_second": 0.194 | |
| } | |
| ], | |
| "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": 720746397892608.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |