| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.977924944812361, | |
| "eval_steps": 500, | |
| "global_step": 1130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008830022075055188, | |
| "grad_norm": 31.472214126236935, | |
| "learning_rate": 2.9411764705882356e-07, | |
| "loss": 2.1083, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017660044150110375, | |
| "grad_norm": 32.15833353719434, | |
| "learning_rate": 5.882352941176471e-07, | |
| "loss": 2.0871, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.026490066225165563, | |
| "grad_norm": 34.588808039057916, | |
| "learning_rate": 8.823529411764707e-07, | |
| "loss": 2.0878, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03532008830022075, | |
| "grad_norm": 30.636831811168822, | |
| "learning_rate": 1.1764705882352942e-06, | |
| "loss": 2.0747, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04415011037527594, | |
| "grad_norm": 30.05777723127091, | |
| "learning_rate": 1.4705882352941177e-06, | |
| "loss": 2.0485, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.052980132450331126, | |
| "grad_norm": 29.120771169243604, | |
| "learning_rate": 1.7647058823529414e-06, | |
| "loss": 1.9477, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06181015452538632, | |
| "grad_norm": 24.866737214984997, | |
| "learning_rate": 2.058823529411765e-06, | |
| "loss": 1.6497, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0706401766004415, | |
| "grad_norm": 23.230376850062278, | |
| "learning_rate": 2.3529411764705885e-06, | |
| "loss": 1.5604, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07947019867549669, | |
| "grad_norm": 19.647598453021406, | |
| "learning_rate": 2.647058823529412e-06, | |
| "loss": 1.1006, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08830022075055188, | |
| "grad_norm": 18.06390068820111, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 1.0531, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09713024282560706, | |
| "grad_norm": 16.569743051324362, | |
| "learning_rate": 3.2352941176470594e-06, | |
| "loss": 0.9419, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10596026490066225, | |
| "grad_norm": 12.784285647776292, | |
| "learning_rate": 3.529411764705883e-06, | |
| "loss": 0.5246, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.11479028697571744, | |
| "grad_norm": 11.605982508137956, | |
| "learning_rate": 3.8235294117647055e-06, | |
| "loss": 0.4798, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.12362030905077263, | |
| "grad_norm": 6.256037905480305, | |
| "learning_rate": 4.11764705882353e-06, | |
| "loss": 0.451, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.13245033112582782, | |
| "grad_norm": 7.103961274104865, | |
| "learning_rate": 4.411764705882353e-06, | |
| "loss": 0.4337, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.141280353200883, | |
| "grad_norm": 4.825119292404456, | |
| "learning_rate": 4.705882352941177e-06, | |
| "loss": 0.3948, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.15011037527593818, | |
| "grad_norm": 2.893093128603599, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3771, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.15894039735099338, | |
| "grad_norm": 8.328356735410495, | |
| "learning_rate": 5.294117647058824e-06, | |
| "loss": 0.3881, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.16777041942604856, | |
| "grad_norm": 4.006731967853461, | |
| "learning_rate": 5.588235294117647e-06, | |
| "loss": 0.3757, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17660044150110377, | |
| "grad_norm": 3.2628794858983867, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.3714, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18543046357615894, | |
| "grad_norm": 4.125441400823564, | |
| "learning_rate": 6.176470588235295e-06, | |
| "loss": 0.3708, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.19426048565121412, | |
| "grad_norm": 2.58341906282714, | |
| "learning_rate": 6.470588235294119e-06, | |
| "loss": 0.3569, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.20309050772626933, | |
| "grad_norm": 2.993594317696669, | |
| "learning_rate": 6.764705882352942e-06, | |
| "loss": 0.3523, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2119205298013245, | |
| "grad_norm": 4.5872224660898056, | |
| "learning_rate": 7.058823529411766e-06, | |
| "loss": 0.3691, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.22075055187637968, | |
| "grad_norm": 5.759010359441343, | |
| "learning_rate": 7.352941176470589e-06, | |
| "loss": 0.3516, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.22958057395143489, | |
| "grad_norm": 3.2537432872956993, | |
| "learning_rate": 7.647058823529411e-06, | |
| "loss": 0.3493, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23841059602649006, | |
| "grad_norm": 3.0852147543420654, | |
| "learning_rate": 7.941176470588236e-06, | |
| "loss": 0.3388, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.24724061810154527, | |
| "grad_norm": 4.186410339980927, | |
| "learning_rate": 8.23529411764706e-06, | |
| "loss": 0.3475, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2560706401766004, | |
| "grad_norm": 9.7375714331209, | |
| "learning_rate": 8.529411764705883e-06, | |
| "loss": 0.3482, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.26490066225165565, | |
| "grad_norm": 4.931105919661555, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.3677, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2737306843267108, | |
| "grad_norm": 6.412332226369269, | |
| "learning_rate": 9.11764705882353e-06, | |
| "loss": 0.3393, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.282560706401766, | |
| "grad_norm": 3.5350734062425526, | |
| "learning_rate": 9.411764705882354e-06, | |
| "loss": 0.3405, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2913907284768212, | |
| "grad_norm": 3.072099986244548, | |
| "learning_rate": 9.705882352941177e-06, | |
| "loss": 0.344, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.30022075055187636, | |
| "grad_norm": 3.0371243929196607, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3493, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3090507726269316, | |
| "grad_norm": 4.191051968737573, | |
| "learning_rate": 9.99997945915297e-06, | |
| "loss": 0.3467, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.31788079470198677, | |
| "grad_norm": 3.010164769069098, | |
| "learning_rate": 9.999917836780642e-06, | |
| "loss": 0.3449, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.32671081677704195, | |
| "grad_norm": 2.674113505340197, | |
| "learning_rate": 9.999815133389334e-06, | |
| "loss": 0.334, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3355408388520971, | |
| "grad_norm": 2.7453045941831102, | |
| "learning_rate": 9.999671349822887e-06, | |
| "loss": 0.3487, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3443708609271523, | |
| "grad_norm": 2.9404787627680915, | |
| "learning_rate": 9.999486487262677e-06, | |
| "loss": 0.3347, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.35320088300220753, | |
| "grad_norm": 3.7818713279477403, | |
| "learning_rate": 9.999260547227599e-06, | |
| "loss": 0.3381, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3620309050772627, | |
| "grad_norm": 4.458660179761209, | |
| "learning_rate": 9.998993531574048e-06, | |
| "loss": 0.3405, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3708609271523179, | |
| "grad_norm": 31.001580427161752, | |
| "learning_rate": 9.998685442495921e-06, | |
| "loss": 0.3367, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.37969094922737306, | |
| "grad_norm": 5.282504835702034, | |
| "learning_rate": 9.998336282524579e-06, | |
| "loss": 0.334, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.38852097130242824, | |
| "grad_norm": 11.298914380090567, | |
| "learning_rate": 9.997946054528837e-06, | |
| "loss": 0.3176, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3973509933774834, | |
| "grad_norm": 5.734658537901168, | |
| "learning_rate": 9.997514761714946e-06, | |
| "loss": 0.3348, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.40618101545253865, | |
| "grad_norm": 3.8830807632361863, | |
| "learning_rate": 9.99704240762655e-06, | |
| "loss": 0.308, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.41501103752759383, | |
| "grad_norm": 3.835669764673382, | |
| "learning_rate": 9.996528996144668e-06, | |
| "loss": 0.3189, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.423841059602649, | |
| "grad_norm": 3.5361428814007834, | |
| "learning_rate": 9.995974531487668e-06, | |
| "loss": 0.3291, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4326710816777042, | |
| "grad_norm": 3.7424429068660947, | |
| "learning_rate": 9.995379018211215e-06, | |
| "loss": 0.3275, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.44150110375275936, | |
| "grad_norm": 4.232373525339259, | |
| "learning_rate": 9.994742461208251e-06, | |
| "loss": 0.3354, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4503311258278146, | |
| "grad_norm": 2.93031806096149, | |
| "learning_rate": 9.994064865708944e-06, | |
| "loss": 0.3284, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.45916114790286977, | |
| "grad_norm": 2.8163989461585777, | |
| "learning_rate": 9.993346237280646e-06, | |
| "loss": 0.3238, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.46799116997792495, | |
| "grad_norm": 3.6855007504645325, | |
| "learning_rate": 9.992586581827853e-06, | |
| "loss": 0.3217, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.4768211920529801, | |
| "grad_norm": 6.785207552790701, | |
| "learning_rate": 9.991785905592149e-06, | |
| "loss": 0.3193, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4856512141280353, | |
| "grad_norm": 4.517525332788386, | |
| "learning_rate": 9.990944215152166e-06, | |
| "loss": 0.3221, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.49448123620309054, | |
| "grad_norm": 3.46545292217024, | |
| "learning_rate": 9.990061517423513e-06, | |
| "loss": 0.3157, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5033112582781457, | |
| "grad_norm": 4.553647135052966, | |
| "learning_rate": 9.989137819658738e-06, | |
| "loss": 0.328, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5121412803532008, | |
| "grad_norm": 3.401290053482735, | |
| "learning_rate": 9.988173129447251e-06, | |
| "loss": 0.3382, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5209713024282561, | |
| "grad_norm": 2.9344717025661655, | |
| "learning_rate": 9.987167454715277e-06, | |
| "loss": 0.3225, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5298013245033113, | |
| "grad_norm": 3.8235319736600974, | |
| "learning_rate": 9.986120803725776e-06, | |
| "loss": 0.315, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5386313465783664, | |
| "grad_norm": 3.425687806394113, | |
| "learning_rate": 9.985033185078392e-06, | |
| "loss": 0.311, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5474613686534217, | |
| "grad_norm": 3.692220938665743, | |
| "learning_rate": 9.983904607709365e-06, | |
| "loss": 0.3181, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5562913907284768, | |
| "grad_norm": 7.265669180018436, | |
| "learning_rate": 9.982735080891471e-06, | |
| "loss": 0.3206, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.565121412803532, | |
| "grad_norm": 4.825970130501887, | |
| "learning_rate": 9.981524614233938e-06, | |
| "loss": 0.3196, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5739514348785872, | |
| "grad_norm": 3.3039217714979685, | |
| "learning_rate": 9.98027321768237e-06, | |
| "loss": 0.308, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5827814569536424, | |
| "grad_norm": 4.1933921842003326, | |
| "learning_rate": 9.978980901518663e-06, | |
| "loss": 0.323, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5916114790286976, | |
| "grad_norm": 3.729103761925266, | |
| "learning_rate": 9.977647676360927e-06, | |
| "loss": 0.3224, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6004415011037527, | |
| "grad_norm": 5.5392456644015375, | |
| "learning_rate": 9.976273553163393e-06, | |
| "loss": 0.3102, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.609271523178808, | |
| "grad_norm": 4.552841845254194, | |
| "learning_rate": 9.974858543216319e-06, | |
| "loss": 0.3209, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6181015452538632, | |
| "grad_norm": 6.717370392353334, | |
| "learning_rate": 9.973402658145908e-06, | |
| "loss": 0.3085, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6269315673289183, | |
| "grad_norm": 4.8626221578786994, | |
| "learning_rate": 9.971905909914206e-06, | |
| "loss": 0.3204, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6357615894039735, | |
| "grad_norm": 4.151842172528407, | |
| "learning_rate": 9.970368310819e-06, | |
| "loss": 0.3071, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6445916114790287, | |
| "grad_norm": 3.8472536914005624, | |
| "learning_rate": 9.96878987349373e-06, | |
| "loss": 0.3092, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6534216335540839, | |
| "grad_norm": 3.2092988951248556, | |
| "learning_rate": 9.96717061090737e-06, | |
| "loss": 0.3091, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6622516556291391, | |
| "grad_norm": 3.4125331719718592, | |
| "learning_rate": 9.965510536364329e-06, | |
| "loss": 0.3098, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6710816777041942, | |
| "grad_norm": 4.940061254399842, | |
| "learning_rate": 9.96380966350434e-06, | |
| "loss": 0.3066, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6799116997792495, | |
| "grad_norm": 4.605143685171599, | |
| "learning_rate": 9.962068006302357e-06, | |
| "loss": 0.3067, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6887417218543046, | |
| "grad_norm": 6.978340444144683, | |
| "learning_rate": 9.960285579068419e-06, | |
| "loss": 0.31, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6975717439293598, | |
| "grad_norm": 3.891436802395962, | |
| "learning_rate": 9.958462396447556e-06, | |
| "loss": 0.3151, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7064017660044151, | |
| "grad_norm": 4.119146509983581, | |
| "learning_rate": 9.956598473419652e-06, | |
| "loss": 0.3075, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7152317880794702, | |
| "grad_norm": 4.705337914894693, | |
| "learning_rate": 9.954693825299333e-06, | |
| "loss": 0.3187, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7240618101545254, | |
| "grad_norm": 2.947984899144426, | |
| "learning_rate": 9.95274846773583e-06, | |
| "loss": 0.3068, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7328918322295805, | |
| "grad_norm": 5.138263095308943, | |
| "learning_rate": 9.950762416712862e-06, | |
| "loss": 0.314, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7417218543046358, | |
| "grad_norm": 3.188362465594982, | |
| "learning_rate": 9.948735688548496e-06, | |
| "loss": 0.311, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7505518763796909, | |
| "grad_norm": 3.042723770882341, | |
| "learning_rate": 9.946668299895017e-06, | |
| "loss": 0.3109, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7593818984547461, | |
| "grad_norm": 2.8522063112096583, | |
| "learning_rate": 9.944560267738792e-06, | |
| "loss": 0.3121, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7682119205298014, | |
| "grad_norm": 2.9238925275725434, | |
| "learning_rate": 9.942411609400127e-06, | |
| "loss": 0.3093, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7770419426048565, | |
| "grad_norm": 2.591722124536507, | |
| "learning_rate": 9.940222342533126e-06, | |
| "loss": 0.3068, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7858719646799117, | |
| "grad_norm": 2.5785034732705014, | |
| "learning_rate": 9.937992485125547e-06, | |
| "loss": 0.2942, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7947019867549668, | |
| "grad_norm": 4.138256637711546, | |
| "learning_rate": 9.935722055498655e-06, | |
| "loss": 0.3022, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8035320088300221, | |
| "grad_norm": 4.086329806341238, | |
| "learning_rate": 9.933411072307071e-06, | |
| "loss": 0.2904, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8123620309050773, | |
| "grad_norm": 4.03080901689932, | |
| "learning_rate": 9.931059554538613e-06, | |
| "loss": 0.3153, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8211920529801324, | |
| "grad_norm": 3.4137385241906086, | |
| "learning_rate": 9.928667521514149e-06, | |
| "loss": 0.3009, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8300220750551877, | |
| "grad_norm": 2.9594504347852912, | |
| "learning_rate": 9.92623499288743e-06, | |
| "loss": 0.3037, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8388520971302428, | |
| "grad_norm": 10.264141036298046, | |
| "learning_rate": 9.92376198864494e-06, | |
| "loss": 0.3032, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.847682119205298, | |
| "grad_norm": 3.1835922899288605, | |
| "learning_rate": 9.921248529105716e-06, | |
| "loss": 0.3022, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8565121412803532, | |
| "grad_norm": 14.325694406880958, | |
| "learning_rate": 9.918694634921195e-06, | |
| "loss": 0.3065, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8653421633554084, | |
| "grad_norm": 2.714403826120161, | |
| "learning_rate": 9.916100327075038e-06, | |
| "loss": 0.3022, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8741721854304636, | |
| "grad_norm": 2.380216318721788, | |
| "learning_rate": 9.913465626882954e-06, | |
| "loss": 0.3025, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8830022075055187, | |
| "grad_norm": 2.707321155905504, | |
| "learning_rate": 9.910790555992536e-06, | |
| "loss": 0.2948, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.891832229580574, | |
| "grad_norm": 2.7280642740110017, | |
| "learning_rate": 9.908075136383068e-06, | |
| "loss": 0.2972, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9006622516556292, | |
| "grad_norm": 4.172217839092669, | |
| "learning_rate": 9.905319390365364e-06, | |
| "loss": 0.3037, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9094922737306843, | |
| "grad_norm": 3.0609907097439693, | |
| "learning_rate": 9.902523340581562e-06, | |
| "loss": 0.3081, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9183222958057395, | |
| "grad_norm": 3.095543219759857, | |
| "learning_rate": 9.899687010004956e-06, | |
| "loss": 0.3046, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9271523178807947, | |
| "grad_norm": 2.779926187941415, | |
| "learning_rate": 9.896810421939797e-06, | |
| "loss": 0.304, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9359823399558499, | |
| "grad_norm": 4.043503414484149, | |
| "learning_rate": 9.893893600021112e-06, | |
| "loss": 0.3004, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9448123620309051, | |
| "grad_norm": 3.123314639461237, | |
| "learning_rate": 9.890936568214493e-06, | |
| "loss": 0.3098, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9536423841059603, | |
| "grad_norm": 2.981784037078161, | |
| "learning_rate": 9.88793935081592e-06, | |
| "loss": 0.3076, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9624724061810155, | |
| "grad_norm": 2.6275887763260317, | |
| "learning_rate": 9.884901972451542e-06, | |
| "loss": 0.306, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9713024282560706, | |
| "grad_norm": 3.0626145222937176, | |
| "learning_rate": 9.881824458077491e-06, | |
| "loss": 0.2907, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9801324503311258, | |
| "grad_norm": 7.625334642748094, | |
| "learning_rate": 9.878706832979668e-06, | |
| "loss": 0.3097, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9889624724061811, | |
| "grad_norm": 9.294252705300167, | |
| "learning_rate": 9.875549122773536e-06, | |
| "loss": 0.3067, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9977924944812362, | |
| "grad_norm": 2.65087564784751, | |
| "learning_rate": 9.872351353403912e-06, | |
| "loss": 0.3002, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0066225165562914, | |
| "grad_norm": 6.243134976658234, | |
| "learning_rate": 9.869113551144754e-06, | |
| "loss": 0.287, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0154525386313467, | |
| "grad_norm": 2.64422179535934, | |
| "learning_rate": 9.865835742598942e-06, | |
| "loss": 0.304, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0242825607064017, | |
| "grad_norm": 2.9183431870572, | |
| "learning_rate": 9.86251795469806e-06, | |
| "loss": 0.2866, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.033112582781457, | |
| "grad_norm": 3.034070209394834, | |
| "learning_rate": 9.85916021470218e-06, | |
| "loss": 0.2841, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0419426048565121, | |
| "grad_norm": 3.35101137770141, | |
| "learning_rate": 9.85576255019963e-06, | |
| "loss": 0.2954, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.0507726269315674, | |
| "grad_norm": 3.6017568334438543, | |
| "learning_rate": 9.852324989106772e-06, | |
| "loss": 0.2965, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0596026490066226, | |
| "grad_norm": 3.416933194221564, | |
| "learning_rate": 9.848847559667774e-06, | |
| "loss": 0.286, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0684326710816776, | |
| "grad_norm": 4.026068402110005, | |
| "learning_rate": 9.845330290454373e-06, | |
| "loss": 0.2784, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0772626931567328, | |
| "grad_norm": 3.949752899521531, | |
| "learning_rate": 9.841773210365646e-06, | |
| "loss": 0.2903, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.086092715231788, | |
| "grad_norm": 4.659419881904087, | |
| "learning_rate": 9.838176348627768e-06, | |
| "loss": 0.2786, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0949227373068433, | |
| "grad_norm": 3.9554546696632062, | |
| "learning_rate": 9.834539734793774e-06, | |
| "loss": 0.2765, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1037527593818985, | |
| "grad_norm": 3.3263059677806144, | |
| "learning_rate": 9.830863398743313e-06, | |
| "loss": 0.289, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1125827814569536, | |
| "grad_norm": 3.753486244769864, | |
| "learning_rate": 9.82714737068241e-06, | |
| "loss": 0.2734, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.1214128035320088, | |
| "grad_norm": 3.119442815671178, | |
| "learning_rate": 9.82339168114321e-06, | |
| "loss": 0.2837, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.130242825607064, | |
| "grad_norm": 3.0092230067566113, | |
| "learning_rate": 9.81959636098373e-06, | |
| "loss": 0.2873, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.1390728476821192, | |
| "grad_norm": 2.918428900141731, | |
| "learning_rate": 9.815761441387609e-06, | |
| "loss": 0.2899, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1479028697571745, | |
| "grad_norm": 3.531980341444317, | |
| "learning_rate": 9.811886953863841e-06, | |
| "loss": 0.2927, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1567328918322295, | |
| "grad_norm": 2.790497908261151, | |
| "learning_rate": 9.807972930246531e-06, | |
| "loss": 0.2788, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1655629139072847, | |
| "grad_norm": 3.249465838523664, | |
| "learning_rate": 9.804019402694627e-06, | |
| "loss": 0.2809, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.17439293598234, | |
| "grad_norm": 3.698315301608005, | |
| "learning_rate": 9.800026403691643e-06, | |
| "loss": 0.2774, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1832229580573952, | |
| "grad_norm": 5.910754956225486, | |
| "learning_rate": 9.795993966045418e-06, | |
| "loss": 0.2895, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1920529801324504, | |
| "grad_norm": 3.3223887611454845, | |
| "learning_rate": 9.791922122887823e-06, | |
| "loss": 0.2658, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2008830022075054, | |
| "grad_norm": 5.12533018347921, | |
| "learning_rate": 9.7878109076745e-06, | |
| "loss": 0.2984, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.2097130242825607, | |
| "grad_norm": 3.032435185891481, | |
| "learning_rate": 9.783660354184589e-06, | |
| "loss": 0.2938, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.218543046357616, | |
| "grad_norm": 5.112618572472365, | |
| "learning_rate": 9.779470496520442e-06, | |
| "loss": 0.2814, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.2273730684326711, | |
| "grad_norm": 4.598820891916754, | |
| "learning_rate": 9.77524136910735e-06, | |
| "loss": 0.2719, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.2362030905077264, | |
| "grad_norm": 3.407999383294316, | |
| "learning_rate": 9.770973006693256e-06, | |
| "loss": 0.2639, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2450331125827814, | |
| "grad_norm": 2.812300162623802, | |
| "learning_rate": 9.766665444348472e-06, | |
| "loss": 0.285, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2538631346578366, | |
| "grad_norm": 3.957009194516272, | |
| "learning_rate": 9.76231871746539e-06, | |
| "loss": 0.2777, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2626931567328918, | |
| "grad_norm": 3.0050891545381213, | |
| "learning_rate": 9.757932861758188e-06, | |
| "loss": 0.2785, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.271523178807947, | |
| "grad_norm": 4.8122954425792095, | |
| "learning_rate": 9.753507913262548e-06, | |
| "loss": 0.2675, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.280353200883002, | |
| "grad_norm": 3.2032683763953074, | |
| "learning_rate": 9.749043908335343e-06, | |
| "loss": 0.2697, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2891832229580573, | |
| "grad_norm": 4.011066621589908, | |
| "learning_rate": 9.744540883654348e-06, | |
| "loss": 0.278, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2980132450331126, | |
| "grad_norm": 3.9914909828317278, | |
| "learning_rate": 9.739998876217943e-06, | |
| "loss": 0.2898, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.3068432671081678, | |
| "grad_norm": 3.239939213802266, | |
| "learning_rate": 9.735417923344798e-06, | |
| "loss": 0.2782, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.315673289183223, | |
| "grad_norm": 3.1214484422893576, | |
| "learning_rate": 9.730798062673575e-06, | |
| "loss": 0.2982, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.3245033112582782, | |
| "grad_norm": 3.6051213726705553, | |
| "learning_rate": 9.726139332162613e-06, | |
| "loss": 0.2852, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 2.65930833924462, | |
| "learning_rate": 9.721441770089621e-06, | |
| "loss": 0.2762, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.3421633554083885, | |
| "grad_norm": 6.422275927400332, | |
| "learning_rate": 9.716705415051362e-06, | |
| "loss": 0.2773, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3509933774834437, | |
| "grad_norm": 2.9106522864082582, | |
| "learning_rate": 9.711930305963333e-06, | |
| "loss": 0.2784, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.359823399558499, | |
| "grad_norm": 2.942935349375758, | |
| "learning_rate": 9.707116482059447e-06, | |
| "loss": 0.276, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.368653421633554, | |
| "grad_norm": 3.4083975456037763, | |
| "learning_rate": 9.702263982891712e-06, | |
| "loss": 0.2655, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3774834437086092, | |
| "grad_norm": 2.8413100759896723, | |
| "learning_rate": 9.697372848329905e-06, | |
| "loss": 0.2743, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.3863134657836644, | |
| "grad_norm": 3.6858688107675044, | |
| "learning_rate": 9.692443118561248e-06, | |
| "loss": 0.2535, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3951434878587197, | |
| "grad_norm": 3.6023827654035205, | |
| "learning_rate": 9.68747483409007e-06, | |
| "loss": 0.2731, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.403973509933775, | |
| "grad_norm": 4.212735867483803, | |
| "learning_rate": 9.682468035737475e-06, | |
| "loss": 0.2803, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.4128035320088301, | |
| "grad_norm": 4.533995980777807, | |
| "learning_rate": 9.677422764641021e-06, | |
| "loss": 0.273, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.4216335540838851, | |
| "grad_norm": 7.698914599282975, | |
| "learning_rate": 9.672339062254359e-06, | |
| "loss": 0.2775, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.4304635761589404, | |
| "grad_norm": 3.5156241492335267, | |
| "learning_rate": 9.667216970346916e-06, | |
| "loss": 0.2757, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.4392935982339956, | |
| "grad_norm": 3.26262084050584, | |
| "learning_rate": 9.662056531003528e-06, | |
| "loss": 0.2744, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4481236203090508, | |
| "grad_norm": 4.415132138115277, | |
| "learning_rate": 9.656857786624119e-06, | |
| "loss": 0.2816, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4569536423841059, | |
| "grad_norm": 2.937391399969304, | |
| "learning_rate": 9.651620779923332e-06, | |
| "loss": 0.2715, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.465783664459161, | |
| "grad_norm": 3.9161994953200456, | |
| "learning_rate": 9.646345553930187e-06, | |
| "loss": 0.2684, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4746136865342163, | |
| "grad_norm": 3.245027966967772, | |
| "learning_rate": 9.64103215198773e-06, | |
| "loss": 0.2819, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.4834437086092715, | |
| "grad_norm": 3.1766817840431787, | |
| "learning_rate": 9.63568061775267e-06, | |
| "loss": 0.2799, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.4922737306843268, | |
| "grad_norm": 2.9692095940294236, | |
| "learning_rate": 9.630290995195028e-06, | |
| "loss": 0.278, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.501103752759382, | |
| "grad_norm": 5.859224280094942, | |
| "learning_rate": 9.624863328597767e-06, | |
| "loss": 0.2629, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5099337748344372, | |
| "grad_norm": 8.580606750008384, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.2871, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.5187637969094923, | |
| "grad_norm": 5.141915372978526, | |
| "learning_rate": 9.613894041978795e-06, | |
| "loss": 0.289, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.5275938189845475, | |
| "grad_norm": 2.9803065345701105, | |
| "learning_rate": 9.60835251208446e-06, | |
| "loss": 0.2695, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.5364238410596025, | |
| "grad_norm": 14.553698242525966, | |
| "learning_rate": 9.602773118404518e-06, | |
| "loss": 0.2709, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.5452538631346577, | |
| "grad_norm": 4.985779893976861, | |
| "learning_rate": 9.597155906781154e-06, | |
| "loss": 0.2802, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.554083885209713, | |
| "grad_norm": 3.8079488049866805, | |
| "learning_rate": 9.591500923367287e-06, | |
| "loss": 0.2754, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5629139072847682, | |
| "grad_norm": 3.4878082048235988, | |
| "learning_rate": 9.585808214626173e-06, | |
| "loss": 0.2665, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5717439293598234, | |
| "grad_norm": 3.1802324525612335, | |
| "learning_rate": 9.580077827331038e-06, | |
| "loss": 0.2919, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5805739514348787, | |
| "grad_norm": 2.618673109920962, | |
| "learning_rate": 9.574309808564682e-06, | |
| "loss": 0.281, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.589403973509934, | |
| "grad_norm": 2.7260712262034823, | |
| "learning_rate": 9.568504205719106e-06, | |
| "loss": 0.2628, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.598233995584989, | |
| "grad_norm": 3.388075861806477, | |
| "learning_rate": 9.562661066495108e-06, | |
| "loss": 0.2752, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.6070640176600441, | |
| "grad_norm": 3.6768479091539326, | |
| "learning_rate": 9.556780438901899e-06, | |
| "loss": 0.2668, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.6158940397350994, | |
| "grad_norm": 3.0390503984415154, | |
| "learning_rate": 9.550862371256705e-06, | |
| "loss": 0.2724, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.6247240618101544, | |
| "grad_norm": 5.468403403351693, | |
| "learning_rate": 9.544906912184383e-06, | |
| "loss": 0.2629, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.6335540838852096, | |
| "grad_norm": 2.9775518077313157, | |
| "learning_rate": 9.538914110616995e-06, | |
| "loss": 0.2616, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.6423841059602649, | |
| "grad_norm": 4.091861587160806, | |
| "learning_rate": 9.532884015793432e-06, | |
| "loss": 0.2767, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.65121412803532, | |
| "grad_norm": 5.523753816170061, | |
| "learning_rate": 9.526816677258995e-06, | |
| "loss": 0.2643, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.6600441501103753, | |
| "grad_norm": 3.009934476105004, | |
| "learning_rate": 9.520712144864997e-06, | |
| "loss": 0.2603, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6688741721854305, | |
| "grad_norm": 3.434710756320171, | |
| "learning_rate": 9.514570468768338e-06, | |
| "loss": 0.2705, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.6777041942604858, | |
| "grad_norm": 3.441152716205931, | |
| "learning_rate": 9.508391699431114e-06, | |
| "loss": 0.2608, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6865342163355408, | |
| "grad_norm": 5.074281008610461, | |
| "learning_rate": 9.502175887620188e-06, | |
| "loss": 0.256, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.695364238410596, | |
| "grad_norm": 3.50131405452359, | |
| "learning_rate": 9.495923084406773e-06, | |
| "loss": 0.2689, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.7041942604856513, | |
| "grad_norm": 3.8992604608113437, | |
| "learning_rate": 9.48963334116602e-06, | |
| "loss": 0.2671, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.7130242825607063, | |
| "grad_norm": 4.75791802439842, | |
| "learning_rate": 9.48330670957659e-06, | |
| "loss": 0.2715, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.7218543046357615, | |
| "grad_norm": 7.899252866668247, | |
| "learning_rate": 9.476943241620233e-06, | |
| "loss": 0.2713, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.7306843267108167, | |
| "grad_norm": 3.7222276908690937, | |
| "learning_rate": 9.470542989581357e-06, | |
| "loss": 0.2684, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.739514348785872, | |
| "grad_norm": 4.933844116299028, | |
| "learning_rate": 9.464106006046602e-06, | |
| "loss": 0.2571, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.7483443708609272, | |
| "grad_norm": 3.5987457528932962, | |
| "learning_rate": 9.457632343904404e-06, | |
| "loss": 0.2633, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7571743929359824, | |
| "grad_norm": 4.630895507273217, | |
| "learning_rate": 9.451122056344564e-06, | |
| "loss": 0.2506, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7660044150110377, | |
| "grad_norm": 4.853234260117955, | |
| "learning_rate": 9.444575196857814e-06, | |
| "loss": 0.2651, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7748344370860927, | |
| "grad_norm": 9.787702337965728, | |
| "learning_rate": 9.437991819235366e-06, | |
| "loss": 0.2739, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.783664459161148, | |
| "grad_norm": 7.374843445266783, | |
| "learning_rate": 9.431371977568483e-06, | |
| "loss": 0.2767, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.7924944812362031, | |
| "grad_norm": 5.318804005582171, | |
| "learning_rate": 9.424715726248027e-06, | |
| "loss": 0.2629, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.8013245033112582, | |
| "grad_norm": 4.8155223639383085, | |
| "learning_rate": 9.418023119964012e-06, | |
| "loss": 0.2664, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.8101545253863134, | |
| "grad_norm": 3.7834946464817425, | |
| "learning_rate": 9.411294213705162e-06, | |
| "loss": 0.2776, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.8189845474613686, | |
| "grad_norm": 3.8437162679852674, | |
| "learning_rate": 9.404529062758447e-06, | |
| "loss": 0.2518, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.8278145695364238, | |
| "grad_norm": 4.063545937455502, | |
| "learning_rate": 9.397727722708643e-06, | |
| "loss": 0.2667, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.836644591611479, | |
| "grad_norm": 3.5521457888425387, | |
| "learning_rate": 9.390890249437863e-06, | |
| "loss": 0.2613, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.8454746136865343, | |
| "grad_norm": 3.28448167697288, | |
| "learning_rate": 9.384016699125102e-06, | |
| "loss": 0.2493, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.8543046357615895, | |
| "grad_norm": 3.1820784576806984, | |
| "learning_rate": 9.377107128245782e-06, | |
| "loss": 0.2547, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8631346578366446, | |
| "grad_norm": 4.3420830156672166, | |
| "learning_rate": 9.370161593571274e-06, | |
| "loss": 0.2577, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.8719646799116998, | |
| "grad_norm": 9.654840118485705, | |
| "learning_rate": 9.363180152168448e-06, | |
| "loss": 0.2499, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.8807947019867548, | |
| "grad_norm": 10.791335205553247, | |
| "learning_rate": 9.356162861399188e-06, | |
| "loss": 0.2559, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.88962472406181, | |
| "grad_norm": 4.418269951620248, | |
| "learning_rate": 9.349109778919938e-06, | |
| "loss": 0.2731, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.8984547461368653, | |
| "grad_norm": 4.350311708598762, | |
| "learning_rate": 9.342020962681206e-06, | |
| "loss": 0.239, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.9072847682119205, | |
| "grad_norm": 3.7003028440684425, | |
| "learning_rate": 9.334896470927115e-06, | |
| "loss": 0.2624, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.9161147902869757, | |
| "grad_norm": 3.571424918897275, | |
| "learning_rate": 9.327736362194899e-06, | |
| "loss": 0.2466, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.924944812362031, | |
| "grad_norm": 4.9725016331177985, | |
| "learning_rate": 9.32054069531444e-06, | |
| "loss": 0.253, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.9337748344370862, | |
| "grad_norm": 3.9580856412457397, | |
| "learning_rate": 9.313309529407773e-06, | |
| "loss": 0.2543, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.9426048565121414, | |
| "grad_norm": 3.1846344968638176, | |
| "learning_rate": 9.306042923888607e-06, | |
| "loss": 0.2564, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9514348785871964, | |
| "grad_norm": 3.64080658809458, | |
| "learning_rate": 9.298740938461835e-06, | |
| "loss": 0.2637, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.9602649006622517, | |
| "grad_norm": 3.913114700217387, | |
| "learning_rate": 9.291403633123046e-06, | |
| "loss": 0.2485, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.9690949227373067, | |
| "grad_norm": 3.913668252474471, | |
| "learning_rate": 9.284031068158023e-06, | |
| "loss": 0.256, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.977924944812362, | |
| "grad_norm": 5.317494032090104, | |
| "learning_rate": 9.27662330414226e-06, | |
| "loss": 0.2614, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.9867549668874172, | |
| "grad_norm": 3.2228188535283104, | |
| "learning_rate": 9.269180401940455e-06, | |
| "loss": 0.256, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9955849889624724, | |
| "grad_norm": 3.0834227247287584, | |
| "learning_rate": 9.261702422706014e-06, | |
| "loss": 0.2532, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.0044150110375276, | |
| "grad_norm": 2.754471183541835, | |
| "learning_rate": 9.254189427880548e-06, | |
| "loss": 0.2305, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.013245033112583, | |
| "grad_norm": 3.240594675285172, | |
| "learning_rate": 9.24664147919337e-06, | |
| "loss": 0.2159, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.022075055187638, | |
| "grad_norm": 3.905148433958218, | |
| "learning_rate": 9.239058638660983e-06, | |
| "loss": 0.2164, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.0309050772626933, | |
| "grad_norm": 3.7158334815586143, | |
| "learning_rate": 9.231440968586572e-06, | |
| "loss": 0.2102, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0397350993377485, | |
| "grad_norm": 7.577503540942662, | |
| "learning_rate": 9.223788531559495e-06, | |
| "loss": 0.2116, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.0485651214128033, | |
| "grad_norm": 3.3948723966954977, | |
| "learning_rate": 9.216101390454771e-06, | |
| "loss": 0.2294, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.0573951434878586, | |
| "grad_norm": 5.038865023441443, | |
| "learning_rate": 9.208379608432552e-06, | |
| "loss": 0.2194, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.066225165562914, | |
| "grad_norm": 3.2008190096465565, | |
| "learning_rate": 9.200623248937619e-06, | |
| "loss": 0.2029, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.075055187637969, | |
| "grad_norm": 3.1245414330497194, | |
| "learning_rate": 9.192832375698845e-06, | |
| "loss": 0.2147, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.0838852097130243, | |
| "grad_norm": 3.9785034361704907, | |
| "learning_rate": 9.185007052728689e-06, | |
| "loss": 0.207, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.0927152317880795, | |
| "grad_norm": 3.3099121521259853, | |
| "learning_rate": 9.177147344322651e-06, | |
| "loss": 0.2144, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.1015452538631347, | |
| "grad_norm": 3.3179804707480303, | |
| "learning_rate": 9.169253315058764e-06, | |
| "loss": 0.2343, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.11037527593819, | |
| "grad_norm": 3.2343991624366732, | |
| "learning_rate": 9.161325029797044e-06, | |
| "loss": 0.2231, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.119205298013245, | |
| "grad_norm": 3.0923923112652045, | |
| "learning_rate": 9.153362553678967e-06, | |
| "loss": 0.1985, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1280353200883004, | |
| "grad_norm": 3.496135412895667, | |
| "learning_rate": 9.145365952126937e-06, | |
| "loss": 0.2149, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.136865342163355, | |
| "grad_norm": 3.2947689817010977, | |
| "learning_rate": 9.13733529084374e-06, | |
| "loss": 0.2172, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.1456953642384105, | |
| "grad_norm": 3.584791681462026, | |
| "learning_rate": 9.129270635812013e-06, | |
| "loss": 0.2189, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.1545253863134657, | |
| "grad_norm": 2.927582122461923, | |
| "learning_rate": 9.12117205329369e-06, | |
| "loss": 0.2202, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.163355408388521, | |
| "grad_norm": 4.393656310500425, | |
| "learning_rate": 9.113039609829472e-06, | |
| "loss": 0.2287, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.172185430463576, | |
| "grad_norm": 5.088586933067697, | |
| "learning_rate": 9.104873372238269e-06, | |
| "loss": 0.2264, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.1810154525386314, | |
| "grad_norm": 4.213132483597989, | |
| "learning_rate": 9.096673407616656e-06, | |
| "loss": 0.2214, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.1898454746136866, | |
| "grad_norm": 5.938160403918625, | |
| "learning_rate": 9.08843978333832e-06, | |
| "loss": 0.2216, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.198675496688742, | |
| "grad_norm": 3.6512930868488063, | |
| "learning_rate": 9.08017256705351e-06, | |
| "loss": 0.2253, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.207505518763797, | |
| "grad_norm": 3.57730563624783, | |
| "learning_rate": 9.071871826688472e-06, | |
| "loss": 0.2213, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.216335540838852, | |
| "grad_norm": 5.531462813491203, | |
| "learning_rate": 9.063537630444903e-06, | |
| "loss": 0.2077, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.225165562913907, | |
| "grad_norm": 3.4930151415891846, | |
| "learning_rate": 9.055170046799386e-06, | |
| "loss": 0.2224, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.2339955849889623, | |
| "grad_norm": 11.619309660675812, | |
| "learning_rate": 9.046769144502818e-06, | |
| "loss": 0.2286, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.2428256070640176, | |
| "grad_norm": 4.352362444845245, | |
| "learning_rate": 9.038334992579863e-06, | |
| "loss": 0.2228, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.251655629139073, | |
| "grad_norm": 5.074146906922235, | |
| "learning_rate": 9.029867660328369e-06, | |
| "loss": 0.222, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.260485651214128, | |
| "grad_norm": 3.5950314065428777, | |
| "learning_rate": 9.021367217318808e-06, | |
| "loss": 0.2176, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.2693156732891833, | |
| "grad_norm": 5.124257655952894, | |
| "learning_rate": 9.012833733393697e-06, | |
| "loss": 0.2277, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.2781456953642385, | |
| "grad_norm": 4.444867006726019, | |
| "learning_rate": 9.004267278667032e-06, | |
| "loss": 0.2257, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.2869757174392937, | |
| "grad_norm": 4.493447361344442, | |
| "learning_rate": 8.99566792352371e-06, | |
| "loss": 0.2352, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.295805739514349, | |
| "grad_norm": 3.323601813819435, | |
| "learning_rate": 8.987035738618943e-06, | |
| "loss": 0.2132, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.304635761589404, | |
| "grad_norm": 3.860553287166379, | |
| "learning_rate": 8.978370794877691e-06, | |
| "loss": 0.2195, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.313465783664459, | |
| "grad_norm": 5.0148973973292845, | |
| "learning_rate": 8.969673163494063e-06, | |
| "loss": 0.2141, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.322295805739514, | |
| "grad_norm": 7.678576605179586, | |
| "learning_rate": 8.960942915930749e-06, | |
| "loss": 0.2027, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.3311258278145695, | |
| "grad_norm": 4.5442853121719855, | |
| "learning_rate": 8.952180123918419e-06, | |
| "loss": 0.2187, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.3399558498896247, | |
| "grad_norm": 7.940242464239972, | |
| "learning_rate": 8.94338485945514e-06, | |
| "loss": 0.2179, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.34878587196468, | |
| "grad_norm": 4.31129181853736, | |
| "learning_rate": 8.934557194805787e-06, | |
| "loss": 0.2329, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.357615894039735, | |
| "grad_norm": 4.963957697311404, | |
| "learning_rate": 8.925697202501442e-06, | |
| "loss": 0.2316, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.3664459161147904, | |
| "grad_norm": 5.9919413399852965, | |
| "learning_rate": 8.916804955338807e-06, | |
| "loss": 0.2158, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.3752759381898456, | |
| "grad_norm": 3.228523793152186, | |
| "learning_rate": 8.907880526379594e-06, | |
| "loss": 0.2105, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.384105960264901, | |
| "grad_norm": 4.118589624269719, | |
| "learning_rate": 8.898923988949936e-06, | |
| "loss": 0.2166, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.3929359823399556, | |
| "grad_norm": 3.7803217386760353, | |
| "learning_rate": 8.88993541663978e-06, | |
| "loss": 0.2011, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.401766004415011, | |
| "grad_norm": 4.676450417731594, | |
| "learning_rate": 8.880914883302278e-06, | |
| "loss": 0.2167, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.410596026490066, | |
| "grad_norm": 3.445405479767293, | |
| "learning_rate": 8.871862463053193e-06, | |
| "loss": 0.207, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.4194260485651213, | |
| "grad_norm": 3.6863018648075068, | |
| "learning_rate": 8.862778230270276e-06, | |
| "loss": 0.2269, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.4282560706401766, | |
| "grad_norm": 5.190640473637881, | |
| "learning_rate": 8.85366225959266e-06, | |
| "loss": 0.219, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.437086092715232, | |
| "grad_norm": 4.806776073461402, | |
| "learning_rate": 8.844514625920246e-06, | |
| "loss": 0.2146, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.445916114790287, | |
| "grad_norm": 3.59639616032018, | |
| "learning_rate": 8.835335404413096e-06, | |
| "loss": 0.2204, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.4547461368653423, | |
| "grad_norm": 3.158498425687896, | |
| "learning_rate": 8.826124670490804e-06, | |
| "loss": 0.2125, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.4635761589403975, | |
| "grad_norm": 3.653768910564694, | |
| "learning_rate": 8.816882499831877e-06, | |
| "loss": 0.2223, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.4724061810154527, | |
| "grad_norm": 3.214436783450957, | |
| "learning_rate": 8.807608968373123e-06, | |
| "loss": 0.1975, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.481236203090508, | |
| "grad_norm": 3.5565280747702643, | |
| "learning_rate": 8.798304152309019e-06, | |
| "loss": 0.2103, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.4900662251655628, | |
| "grad_norm": 4.146688170703499, | |
| "learning_rate": 8.788968128091084e-06, | |
| "loss": 0.2264, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.498896247240618, | |
| "grad_norm": 4.823076720971229, | |
| "learning_rate": 8.779600972427257e-06, | |
| "loss": 0.222, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.507726269315673, | |
| "grad_norm": 6.306262609682544, | |
| "learning_rate": 8.770202762281267e-06, | |
| "loss": 0.204, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.5165562913907285, | |
| "grad_norm": 3.9797044182684913, | |
| "learning_rate": 8.760773574871985e-06, | |
| "loss": 0.2097, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.5253863134657837, | |
| "grad_norm": 12.415005857310733, | |
| "learning_rate": 8.751313487672815e-06, | |
| "loss": 0.2371, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.534216335540839, | |
| "grad_norm": 8.015171500397944, | |
| "learning_rate": 8.741822578411036e-06, | |
| "loss": 0.2244, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.543046357615894, | |
| "grad_norm": 6.084039839876109, | |
| "learning_rate": 8.732300925067177e-06, | |
| "loss": 0.2266, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.5518763796909494, | |
| "grad_norm": 6.20793148620874, | |
| "learning_rate": 8.722748605874365e-06, | |
| "loss": 0.2279, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.560706401766004, | |
| "grad_norm": 3.383405867426473, | |
| "learning_rate": 8.71316569931769e-06, | |
| "loss": 0.2252, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.5695364238410594, | |
| "grad_norm": 4.172850628157777, | |
| "learning_rate": 8.703552284133565e-06, | |
| "loss": 0.1988, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.5783664459161146, | |
| "grad_norm": 3.240053888269314, | |
| "learning_rate": 8.69390843930906e-06, | |
| "loss": 0.2083, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.58719646799117, | |
| "grad_norm": 3.980646183741661, | |
| "learning_rate": 8.684234244081274e-06, | |
| "loss": 0.212, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.596026490066225, | |
| "grad_norm": 4.40607180793288, | |
| "learning_rate": 8.674529777936674e-06, | |
| "loss": 0.197, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.6048565121412803, | |
| "grad_norm": 5.481061626988722, | |
| "learning_rate": 8.66479512061044e-06, | |
| "loss": 0.2226, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.6136865342163356, | |
| "grad_norm": 4.097961955582914, | |
| "learning_rate": 8.655030352085816e-06, | |
| "loss": 0.2234, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.622516556291391, | |
| "grad_norm": 4.188776522065794, | |
| "learning_rate": 8.645235552593447e-06, | |
| "loss": 0.201, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.631346578366446, | |
| "grad_norm": 3.3412050458100433, | |
| "learning_rate": 8.635410802610724e-06, | |
| "loss": 0.2187, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.6401766004415013, | |
| "grad_norm": 9.094374070598343, | |
| "learning_rate": 8.625556182861126e-06, | |
| "loss": 0.2052, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.6490066225165565, | |
| "grad_norm": 5.396561951272275, | |
| "learning_rate": 8.615671774313543e-06, | |
| "loss": 0.234, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.6578366445916117, | |
| "grad_norm": 4.234846903606196, | |
| "learning_rate": 8.605757658181626e-06, | |
| "loss": 0.2061, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 3.3872759211252883, | |
| "learning_rate": 8.595813915923113e-06, | |
| "loss": 0.2265, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.6754966887417218, | |
| "grad_norm": 3.220657792625397, | |
| "learning_rate": 8.585840629239158e-06, | |
| "loss": 0.1916, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.684326710816777, | |
| "grad_norm": 3.9998187335590263, | |
| "learning_rate": 8.575837880073663e-06, | |
| "loss": 0.2107, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.693156732891832, | |
| "grad_norm": 3.5556400871751115, | |
| "learning_rate": 8.565805750612607e-06, | |
| "loss": 0.2153, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.7019867549668874, | |
| "grad_norm": 4.04091912413314, | |
| "learning_rate": 8.555744323283364e-06, | |
| "loss": 0.2047, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.7108167770419427, | |
| "grad_norm": 5.627106480694921, | |
| "learning_rate": 8.545653680754029e-06, | |
| "loss": 0.2263, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.719646799116998, | |
| "grad_norm": 6.261924330638717, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.2014, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.7284768211920527, | |
| "grad_norm": 3.386185798358175, | |
| "learning_rate": 8.525385081966992e-06, | |
| "loss": 0.1875, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.737306843267108, | |
| "grad_norm": 5.775925981033664, | |
| "learning_rate": 8.515207292242969e-06, | |
| "loss": 0.2052, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.746136865342163, | |
| "grad_norm": 10.381711164468916, | |
| "learning_rate": 8.505000620384834e-06, | |
| "loss": 0.2087, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.7549668874172184, | |
| "grad_norm": 3.912905801933265, | |
| "learning_rate": 8.494765150254063e-06, | |
| "loss": 0.2112, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.7637969094922736, | |
| "grad_norm": 3.46039634713119, | |
| "learning_rate": 8.484500965948746e-06, | |
| "loss": 0.2018, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.772626931567329, | |
| "grad_norm": 7.829843169267625, | |
| "learning_rate": 8.474208151802898e-06, | |
| "loss": 0.2035, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.781456953642384, | |
| "grad_norm": 4.147934839778988, | |
| "learning_rate": 8.46388679238577e-06, | |
| "loss": 0.221, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.7902869757174393, | |
| "grad_norm": 8.850274608034498, | |
| "learning_rate": 8.453536972501146e-06, | |
| "loss": 0.2118, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.7991169977924946, | |
| "grad_norm": 3.1292281287164414, | |
| "learning_rate": 8.443158777186652e-06, | |
| "loss": 0.2153, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.80794701986755, | |
| "grad_norm": 3.591659182205305, | |
| "learning_rate": 8.432752291713058e-06, | |
| "loss": 0.2165, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.816777041942605, | |
| "grad_norm": 3.1861139586316183, | |
| "learning_rate": 8.422317601583576e-06, | |
| "loss": 0.2159, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.8256070640176603, | |
| "grad_norm": 3.8319532895168513, | |
| "learning_rate": 8.411854792533154e-06, | |
| "loss": 0.2269, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.8344370860927155, | |
| "grad_norm": 2.7192331323642143, | |
| "learning_rate": 8.401363950527777e-06, | |
| "loss": 0.2071, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.8432671081677703, | |
| "grad_norm": 3.081253585938768, | |
| "learning_rate": 8.390845161763756e-06, | |
| "loss": 0.2068, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.8520971302428255, | |
| "grad_norm": 2.6131199812099446, | |
| "learning_rate": 8.380298512667023e-06, | |
| "loss": 0.2059, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.8609271523178808, | |
| "grad_norm": 2.552577027630616, | |
| "learning_rate": 8.369724089892423e-06, | |
| "loss": 0.2188, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.869757174392936, | |
| "grad_norm": 3.1513145301940546, | |
| "learning_rate": 8.359121980322992e-06, | |
| "loss": 0.2037, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.878587196467991, | |
| "grad_norm": 3.589197229443645, | |
| "learning_rate": 8.34849227106926e-06, | |
| "loss": 0.2243, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.8874172185430464, | |
| "grad_norm": 3.606446179417492, | |
| "learning_rate": 8.337835049468517e-06, | |
| "loss": 0.2117, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.8962472406181017, | |
| "grad_norm": 2.6134224843944716, | |
| "learning_rate": 8.327150403084105e-06, | |
| "loss": 0.2, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.9050772626931565, | |
| "grad_norm": 3.4889239593356023, | |
| "learning_rate": 8.3164384197047e-06, | |
| "loss": 0.2096, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.9139072847682117, | |
| "grad_norm": 3.2867170089775946, | |
| "learning_rate": 8.305699187343586e-06, | |
| "loss": 0.2171, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.922737306843267, | |
| "grad_norm": 3.8295346691664336, | |
| "learning_rate": 8.294932794237936e-06, | |
| "loss": 0.2051, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.931567328918322, | |
| "grad_norm": 2.9022622399356717, | |
| "learning_rate": 8.284139328848083e-06, | |
| "loss": 0.2099, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.9403973509933774, | |
| "grad_norm": 2.9956324090522948, | |
| "learning_rate": 8.273318879856794e-06, | |
| "loss": 0.2069, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.9492273730684326, | |
| "grad_norm": 3.758631386904656, | |
| "learning_rate": 8.262471536168547e-06, | |
| "loss": 0.2228, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.958057395143488, | |
| "grad_norm": 3.2675257839695733, | |
| "learning_rate": 8.251597386908791e-06, | |
| "loss": 0.2098, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.966887417218543, | |
| "grad_norm": 15.222126927013242, | |
| "learning_rate": 8.240696521423221e-06, | |
| "loss": 0.2161, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.9757174392935983, | |
| "grad_norm": 3.898571990074825, | |
| "learning_rate": 8.229769029277044e-06, | |
| "loss": 0.2296, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.9845474613686536, | |
| "grad_norm": 5.123319713408517, | |
| "learning_rate": 8.218815000254233e-06, | |
| "loss": 0.2338, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.993377483443709, | |
| "grad_norm": 9.695546719920452, | |
| "learning_rate": 8.207834524356804e-06, | |
| "loss": 0.2241, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.0022075055187636, | |
| "grad_norm": 2.9756219009808795, | |
| "learning_rate": 8.196827691804066e-06, | |
| "loss": 0.2023, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.011037527593819, | |
| "grad_norm": 4.71342842505975, | |
| "learning_rate": 8.185794593031889e-06, | |
| "loss": 0.1765, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.019867549668874, | |
| "grad_norm": 3.78359648565944, | |
| "learning_rate": 8.174735318691946e-06, | |
| "loss": 0.1609, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.0286975717439293, | |
| "grad_norm": 6.163470908192064, | |
| "learning_rate": 8.163649959650983e-06, | |
| "loss": 0.1734, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.0375275938189845, | |
| "grad_norm": 5.128252606396241, | |
| "learning_rate": 8.152538606990065e-06, | |
| "loss": 0.1771, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.0463576158940397, | |
| "grad_norm": 4.009434226111314, | |
| "learning_rate": 8.141401352003834e-06, | |
| "loss": 0.1604, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.055187637969095, | |
| "grad_norm": 8.63516832320924, | |
| "learning_rate": 8.130238286199747e-06, | |
| "loss": 0.1685, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.06401766004415, | |
| "grad_norm": 4.45655573796746, | |
| "learning_rate": 8.119049501297336e-06, | |
| "loss": 0.1719, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.0728476821192054, | |
| "grad_norm": 3.896539237810484, | |
| "learning_rate": 8.107835089227446e-06, | |
| "loss": 0.1595, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.0816777041942607, | |
| "grad_norm": 4.586962092336582, | |
| "learning_rate": 8.096595142131491e-06, | |
| "loss": 0.1652, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.0905077262693155, | |
| "grad_norm": 3.1918663729543053, | |
| "learning_rate": 8.085329752360683e-06, | |
| "loss": 0.1744, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.0993377483443707, | |
| "grad_norm": 3.4882836490862994, | |
| "learning_rate": 8.074039012475277e-06, | |
| "loss": 0.1625, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.108167770419426, | |
| "grad_norm": 4.0363353366173405, | |
| "learning_rate": 8.062723015243821e-06, | |
| "loss": 0.1715, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.116997792494481, | |
| "grad_norm": 3.957658235500611, | |
| "learning_rate": 8.051381853642385e-06, | |
| "loss": 0.1629, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.1258278145695364, | |
| "grad_norm": 3.3377347317374513, | |
| "learning_rate": 8.04001562085379e-06, | |
| "loss": 0.1566, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.1346578366445916, | |
| "grad_norm": 5.0012135040521075, | |
| "learning_rate": 8.028624410266856e-06, | |
| "loss": 0.1723, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.143487858719647, | |
| "grad_norm": 8.030452222463463, | |
| "learning_rate": 8.017208315475633e-06, | |
| "loss": 0.1726, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.152317880794702, | |
| "grad_norm": 5.674580443068153, | |
| "learning_rate": 8.005767430278619e-06, | |
| "loss": 0.1668, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.1611479028697573, | |
| "grad_norm": 7.01666675859102, | |
| "learning_rate": 7.994301848678006e-06, | |
| "loss": 0.1766, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.1699779249448126, | |
| "grad_norm": 3.6539093837361403, | |
| "learning_rate": 7.982811664878897e-06, | |
| "loss": 0.176, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.1788079470198674, | |
| "grad_norm": 7.580302713085105, | |
| "learning_rate": 7.971296973288534e-06, | |
| "loss": 0.1695, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.1876379690949226, | |
| "grad_norm": 3.016656770337476, | |
| "learning_rate": 7.959757868515526e-06, | |
| "loss": 0.1706, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.196467991169978, | |
| "grad_norm": 3.8547585995862983, | |
| "learning_rate": 7.948194445369065e-06, | |
| "loss": 0.1668, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.205298013245033, | |
| "grad_norm": 4.615745175328835, | |
| "learning_rate": 7.936606798858154e-06, | |
| "loss": 0.1648, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.2141280353200883, | |
| "grad_norm": 4.511257505845132, | |
| "learning_rate": 7.924995024190825e-06, | |
| "loss": 0.157, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.2229580573951435, | |
| "grad_norm": 3.078311803533395, | |
| "learning_rate": 7.91335921677335e-06, | |
| "loss": 0.159, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.2317880794701987, | |
| "grad_norm": 3.315712636214642, | |
| "learning_rate": 7.901699472209467e-06, | |
| "loss": 0.1501, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.240618101545254, | |
| "grad_norm": 5.021727897238176, | |
| "learning_rate": 7.890015886299587e-06, | |
| "loss": 0.1586, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.249448123620309, | |
| "grad_norm": 5.812780194753577, | |
| "learning_rate": 7.878308555040012e-06, | |
| "loss": 0.1527, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.258278145695364, | |
| "grad_norm": 3.4678745486983438, | |
| "learning_rate": 7.86657757462214e-06, | |
| "loss": 0.1741, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.2671081677704192, | |
| "grad_norm": 3.2121277292742016, | |
| "learning_rate": 7.85482304143168e-06, | |
| "loss": 0.159, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.2759381898454745, | |
| "grad_norm": 3.967427420259324, | |
| "learning_rate": 7.843045052047863e-06, | |
| "loss": 0.1551, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.2847682119205297, | |
| "grad_norm": 3.6890161832059776, | |
| "learning_rate": 7.831243703242636e-06, | |
| "loss": 0.1753, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.293598233995585, | |
| "grad_norm": 4.113092600187377, | |
| "learning_rate": 7.819419091979884e-06, | |
| "loss": 0.1588, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.30242825607064, | |
| "grad_norm": 2.915758959364885, | |
| "learning_rate": 7.807571315414616e-06, | |
| "loss": 0.1697, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.3112582781456954, | |
| "grad_norm": 3.271707888485366, | |
| "learning_rate": 7.795700470892177e-06, | |
| "loss": 0.1692, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.3200883002207506, | |
| "grad_norm": 3.6620765401175026, | |
| "learning_rate": 7.783806655947454e-06, | |
| "loss": 0.1746, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.328918322295806, | |
| "grad_norm": 14.608756176216396, | |
| "learning_rate": 7.771889968304054e-06, | |
| "loss": 0.1788, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.337748344370861, | |
| "grad_norm": 4.626237283717204, | |
| "learning_rate": 7.759950505873523e-06, | |
| "loss": 0.1716, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.3465783664459163, | |
| "grad_norm": 3.839772250478355, | |
| "learning_rate": 7.747988366754529e-06, | |
| "loss": 0.1703, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.355408388520971, | |
| "grad_norm": 2.44458016386387, | |
| "learning_rate": 7.736003649232058e-06, | |
| "loss": 0.1741, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.3642384105960264, | |
| "grad_norm": 7.276860238846509, | |
| "learning_rate": 7.723996451776615e-06, | |
| "loss": 0.1629, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.3730684326710816, | |
| "grad_norm": 3.583073207932735, | |
| "learning_rate": 7.711966873043396e-06, | |
| "loss": 0.1707, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.381898454746137, | |
| "grad_norm": 3.8529715527196915, | |
| "learning_rate": 7.699915011871502e-06, | |
| "loss": 0.1888, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.390728476821192, | |
| "grad_norm": 4.514448439317309, | |
| "learning_rate": 7.687840967283102e-06, | |
| "loss": 0.183, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.3995584988962473, | |
| "grad_norm": 3.4831062760797016, | |
| "learning_rate": 7.675744838482641e-06, | |
| "loss": 0.1801, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.4083885209713025, | |
| "grad_norm": 4.724684045841327, | |
| "learning_rate": 7.66362672485601e-06, | |
| "loss": 0.1754, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.4172185430463577, | |
| "grad_norm": 3.8738785833187563, | |
| "learning_rate": 7.651486725969736e-06, | |
| "loss": 0.1714, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.426048565121413, | |
| "grad_norm": 3.2900131984816072, | |
| "learning_rate": 7.639324941570165e-06, | |
| "loss": 0.161, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.4348785871964678, | |
| "grad_norm": 3.09472102314459, | |
| "learning_rate": 7.627141471582635e-06, | |
| "loss": 0.1539, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.443708609271523, | |
| "grad_norm": 5.673624169661576, | |
| "learning_rate": 7.614936416110668e-06, | |
| "loss": 0.1609, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.4525386313465782, | |
| "grad_norm": 4.391300979982457, | |
| "learning_rate": 7.6027098754351306e-06, | |
| "loss": 0.1765, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.4613686534216335, | |
| "grad_norm": 3.284573564669799, | |
| "learning_rate": 7.590461950013424e-06, | |
| "loss": 0.1629, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.4701986754966887, | |
| "grad_norm": 3.1656711556433215, | |
| "learning_rate": 7.578192740478656e-06, | |
| "loss": 0.1645, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.479028697571744, | |
| "grad_norm": 2.855773024349452, | |
| "learning_rate": 7.565902347638806e-06, | |
| "loss": 0.1681, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.487858719646799, | |
| "grad_norm": 4.259407404192956, | |
| "learning_rate": 7.553590872475909e-06, | |
| "loss": 0.1676, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.4966887417218544, | |
| "grad_norm": 4.387110260375, | |
| "learning_rate": 7.541258416145212e-06, | |
| "loss": 0.1581, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.5055187637969096, | |
| "grad_norm": 3.5971408233782616, | |
| "learning_rate": 7.528905079974358e-06, | |
| "loss": 0.168, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.514348785871965, | |
| "grad_norm": 4.52421688500951, | |
| "learning_rate": 7.5165309654625405e-06, | |
| "loss": 0.1732, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.52317880794702, | |
| "grad_norm": 3.541280979708034, | |
| "learning_rate": 7.504136174279679e-06, | |
| "loss": 0.1567, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.5320088300220753, | |
| "grad_norm": 3.802914499292579, | |
| "learning_rate": 7.491720808265576e-06, | |
| "loss": 0.1665, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.54083885209713, | |
| "grad_norm": 5.25027041276173, | |
| "learning_rate": 7.4792849694290846e-06, | |
| "loss": 0.1554, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.5496688741721854, | |
| "grad_norm": 3.3143764951136987, | |
| "learning_rate": 7.466828759947271e-06, | |
| "loss": 0.1598, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.5584988962472406, | |
| "grad_norm": 3.1573477161500394, | |
| "learning_rate": 7.454352282164572e-06, | |
| "loss": 0.1651, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.567328918322296, | |
| "grad_norm": 3.548186224479581, | |
| "learning_rate": 7.441855638591958e-06, | |
| "loss": 0.1661, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.576158940397351, | |
| "grad_norm": 6.0742794254798, | |
| "learning_rate": 7.429338931906085e-06, | |
| "loss": 0.1668, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.5849889624724063, | |
| "grad_norm": 3.371102130237448, | |
| "learning_rate": 7.416802264948455e-06, | |
| "loss": 0.1555, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.5938189845474615, | |
| "grad_norm": 3.54371624942368, | |
| "learning_rate": 7.404245740724573e-06, | |
| "loss": 0.1848, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.6026490066225163, | |
| "grad_norm": 3.5954098106659513, | |
| "learning_rate": 7.391669462403096e-06, | |
| "loss": 0.1585, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.6114790286975715, | |
| "grad_norm": 2.789117239847806, | |
| "learning_rate": 7.379073533314988e-06, | |
| "loss": 0.1668, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.6203090507726268, | |
| "grad_norm": 4.019771265201474, | |
| "learning_rate": 7.366458056952668e-06, | |
| "loss": 0.1858, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.629139072847682, | |
| "grad_norm": 4.754065106387509, | |
| "learning_rate": 7.353823136969167e-06, | |
| "loss": 0.1814, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.6379690949227372, | |
| "grad_norm": 2.8823692657539555, | |
| "learning_rate": 7.341168877177267e-06, | |
| "loss": 0.1884, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.6467991169977925, | |
| "grad_norm": 3.2918096411460285, | |
| "learning_rate": 7.328495381548655e-06, | |
| "loss": 0.1602, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.6556291390728477, | |
| "grad_norm": 3.6016781065575816, | |
| "learning_rate": 7.315802754213062e-06, | |
| "loss": 0.1768, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.664459161147903, | |
| "grad_norm": 3.7417513231039656, | |
| "learning_rate": 7.303091099457418e-06, | |
| "loss": 0.1739, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.673289183222958, | |
| "grad_norm": 3.089014613349081, | |
| "learning_rate": 7.290360521724984e-06, | |
| "loss": 0.1673, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.6821192052980134, | |
| "grad_norm": 3.791257553784634, | |
| "learning_rate": 7.277611125614499e-06, | |
| "loss": 0.1745, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.6909492273730686, | |
| "grad_norm": 3.9016886544968687, | |
| "learning_rate": 7.264843015879321e-06, | |
| "loss": 0.1584, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.699779249448124, | |
| "grad_norm": 2.7880378105790555, | |
| "learning_rate": 7.25205629742657e-06, | |
| "loss": 0.1676, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.7086092715231787, | |
| "grad_norm": 6.822766487375965, | |
| "learning_rate": 7.2392510753162516e-06, | |
| "loss": 0.1636, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.717439293598234, | |
| "grad_norm": 3.6212564613485787, | |
| "learning_rate": 7.226427454760412e-06, | |
| "loss": 0.1794, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.726269315673289, | |
| "grad_norm": 4.224075079539843, | |
| "learning_rate": 7.213585541122261e-06, | |
| "loss": 0.1656, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.7350993377483444, | |
| "grad_norm": 5.793457492397086, | |
| "learning_rate": 7.200725439915314e-06, | |
| "loss": 0.1632, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.7439293598233996, | |
| "grad_norm": 2.8793838616576832, | |
| "learning_rate": 7.187847256802518e-06, | |
| "loss": 0.1547, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 3.752759381898455, | |
| "grad_norm": 3.764417483532706, | |
| "learning_rate": 7.174951097595389e-06, | |
| "loss": 0.1698, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.76158940397351, | |
| "grad_norm": 5.11129788606143, | |
| "learning_rate": 7.162037068253141e-06, | |
| "loss": 0.1623, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 3.7704194260485653, | |
| "grad_norm": 5.163050772812261, | |
| "learning_rate": 7.149105274881815e-06, | |
| "loss": 0.1671, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 3.77924944812362, | |
| "grad_norm": 3.679809209988342, | |
| "learning_rate": 7.136155823733405e-06, | |
| "loss": 0.1651, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.7880794701986753, | |
| "grad_norm": 3.8327447163429196, | |
| "learning_rate": 7.12318882120499e-06, | |
| "loss": 0.1675, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 3.7969094922737305, | |
| "grad_norm": 5.7977982980015925, | |
| "learning_rate": 7.110204373837857e-06, | |
| "loss": 0.1567, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.8057395143487858, | |
| "grad_norm": 43.95049168197346, | |
| "learning_rate": 7.097202588316625e-06, | |
| "loss": 0.1766, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 3.814569536423841, | |
| "grad_norm": 3.46599845688097, | |
| "learning_rate": 7.084183571468368e-06, | |
| "loss": 0.1621, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.8233995584988962, | |
| "grad_norm": 3.237998350417883, | |
| "learning_rate": 7.071147430261738e-06, | |
| "loss": 0.1704, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 3.8322295805739515, | |
| "grad_norm": 4.995380765836811, | |
| "learning_rate": 7.058094271806091e-06, | |
| "loss": 0.17, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 3.8410596026490067, | |
| "grad_norm": 5.624162765645761, | |
| "learning_rate": 7.045024203350598e-06, | |
| "loss": 0.1676, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.849889624724062, | |
| "grad_norm": 3.992830925617219, | |
| "learning_rate": 7.031937332283367e-06, | |
| "loss": 0.1757, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 3.858719646799117, | |
| "grad_norm": 3.004239599946409, | |
| "learning_rate": 7.018833766130571e-06, | |
| "loss": 0.1785, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 3.8675496688741724, | |
| "grad_norm": 2.7567357728888107, | |
| "learning_rate": 7.0057136125555456e-06, | |
| "loss": 0.1679, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.8763796909492276, | |
| "grad_norm": 2.549572826449668, | |
| "learning_rate": 6.9925769793579165e-06, | |
| "loss": 0.1417, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 3.8852097130242824, | |
| "grad_norm": 3.335875012526486, | |
| "learning_rate": 6.979423974472714e-06, | |
| "loss": 0.1612, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.8940397350993377, | |
| "grad_norm": 3.310311140159905, | |
| "learning_rate": 6.966254705969484e-06, | |
| "loss": 0.1604, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 3.902869757174393, | |
| "grad_norm": 3.080253298108051, | |
| "learning_rate": 6.953069282051397e-06, | |
| "loss": 0.1724, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 3.911699779249448, | |
| "grad_norm": 6.654784387915054, | |
| "learning_rate": 6.939867811054365e-06, | |
| "loss": 0.1462, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 3.9205298013245033, | |
| "grad_norm": 2.6187563263017517, | |
| "learning_rate": 6.9266504014461425e-06, | |
| "loss": 0.1541, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 3.9293598233995586, | |
| "grad_norm": 2.829142516260226, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.1729, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.938189845474614, | |
| "grad_norm": 2.7492277750432046, | |
| "learning_rate": 6.900168200921065e-06, | |
| "loss": 0.1586, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 3.9470198675496686, | |
| "grad_norm": 3.506062972307051, | |
| "learning_rate": 6.886903627590938e-06, | |
| "loss": 0.1615, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 3.955849889624724, | |
| "grad_norm": 14.402205490977623, | |
| "learning_rate": 6.8736235508213024e-06, | |
| "loss": 0.1495, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 3.964679911699779, | |
| "grad_norm": 4.327608579958209, | |
| "learning_rate": 6.860328079725764e-06, | |
| "loss": 0.1672, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 3.9735099337748343, | |
| "grad_norm": 3.1363405344268744, | |
| "learning_rate": 6.84701732354442e-06, | |
| "loss": 0.1624, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.9823399558498895, | |
| "grad_norm": 3.5348964182291747, | |
| "learning_rate": 6.8336913916429515e-06, | |
| "loss": 0.1712, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 3.9911699779249448, | |
| "grad_norm": 4.178109743705544, | |
| "learning_rate": 6.820350393511732e-06, | |
| "loss": 0.1528, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.034613557476163, | |
| "learning_rate": 6.806994438764922e-06, | |
| "loss": 0.1784, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.008830022075055, | |
| "grad_norm": 4.310488466467659, | |
| "learning_rate": 6.79362363713957e-06, | |
| "loss": 0.1289, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.0176600441501105, | |
| "grad_norm": 10.397212384375132, | |
| "learning_rate": 6.780238098494711e-06, | |
| "loss": 0.1199, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.026490066225166, | |
| "grad_norm": 3.441456634395186, | |
| "learning_rate": 6.766837932810468e-06, | |
| "loss": 0.1285, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.035320088300221, | |
| "grad_norm": 2.651022643605939, | |
| "learning_rate": 6.7534232501871425e-06, | |
| "loss": 0.1316, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 4.044150110375276, | |
| "grad_norm": 3.666205164429965, | |
| "learning_rate": 6.7399941608443096e-06, | |
| "loss": 0.126, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.052980132450331, | |
| "grad_norm": 3.3186796817591793, | |
| "learning_rate": 6.72655077511992e-06, | |
| "loss": 0.1221, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 4.061810154525387, | |
| "grad_norm": 3.084317801755122, | |
| "learning_rate": 6.713093203469384e-06, | |
| "loss": 0.1244, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.070640176600442, | |
| "grad_norm": 3.5749931933976535, | |
| "learning_rate": 6.6996215564646705e-06, | |
| "loss": 0.1191, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 4.079470198675497, | |
| "grad_norm": 4.843054954303335, | |
| "learning_rate": 6.686135944793395e-06, | |
| "loss": 0.1266, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.088300220750552, | |
| "grad_norm": 6.66417013195998, | |
| "learning_rate": 6.672636479257912e-06, | |
| "loss": 0.1299, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 4.097130242825607, | |
| "grad_norm": 5.637186076347759, | |
| "learning_rate": 6.659123270774406e-06, | |
| "loss": 0.1288, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.105960264900662, | |
| "grad_norm": 3.1279459891235346, | |
| "learning_rate": 6.645596430371976e-06, | |
| "loss": 0.1191, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 4.114790286975717, | |
| "grad_norm": 4.192007750019593, | |
| "learning_rate": 6.632056069191723e-06, | |
| "loss": 0.1195, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.123620309050772, | |
| "grad_norm": 3.311595746717904, | |
| "learning_rate": 6.618502298485844e-06, | |
| "loss": 0.1247, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.132450331125828, | |
| "grad_norm": 3.0585126783701293, | |
| "learning_rate": 6.604935229616711e-06, | |
| "loss": 0.1211, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.141280353200883, | |
| "grad_norm": 6.57086640007419, | |
| "learning_rate": 6.5913549740559606e-06, | |
| "loss": 0.1298, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.150110375275938, | |
| "grad_norm": 4.015952386626973, | |
| "learning_rate": 6.57776164338357e-06, | |
| "loss": 0.1265, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.158940397350993, | |
| "grad_norm": 5.960132106677637, | |
| "learning_rate": 6.564155349286952e-06, | |
| "loss": 0.1253, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.1677704194260485, | |
| "grad_norm": 4.177329071996041, | |
| "learning_rate": 6.550536203560028e-06, | |
| "loss": 0.1273, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.176600441501104, | |
| "grad_norm": 5.901015387557947, | |
| "learning_rate": 6.536904318102314e-06, | |
| "loss": 0.1205, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.185430463576159, | |
| "grad_norm": 3.902080005226539, | |
| "learning_rate": 6.523259804918001e-06, | |
| "loss": 0.1292, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.194260485651214, | |
| "grad_norm": 3.378227140949703, | |
| "learning_rate": 6.509602776115029e-06, | |
| "loss": 0.1151, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.2030905077262695, | |
| "grad_norm": 3.5917973988119978, | |
| "learning_rate": 6.4959333439041775e-06, | |
| "loss": 0.1178, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.211920529801325, | |
| "grad_norm": 2.7739845678814543, | |
| "learning_rate": 6.482251620598129e-06, | |
| "loss": 0.1243, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.22075055187638, | |
| "grad_norm": 3.3854437052146245, | |
| "learning_rate": 6.4685577186105595e-06, | |
| "loss": 0.1232, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.229580573951435, | |
| "grad_norm": 2.7771505830257452, | |
| "learning_rate": 6.45485175045521e-06, | |
| "loss": 0.1276, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.23841059602649, | |
| "grad_norm": 3.206550531809967, | |
| "learning_rate": 6.441133828744954e-06, | |
| "loss": 0.1146, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.247240618101546, | |
| "grad_norm": 3.046593840520393, | |
| "learning_rate": 6.427404066190889e-06, | |
| "loss": 0.127, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.256070640176601, | |
| "grad_norm": 5.486967683529114, | |
| "learning_rate": 6.413662575601391e-06, | |
| "loss": 0.1272, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.264900662251655, | |
| "grad_norm": 4.714191840435068, | |
| "learning_rate": 6.3999094698812055e-06, | |
| "loss": 0.1207, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 4.27373068432671, | |
| "grad_norm": 3.2459853374333, | |
| "learning_rate": 6.386144862030508e-06, | |
| "loss": 0.1105, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.282560706401766, | |
| "grad_norm": 4.281126761891633, | |
| "learning_rate": 6.3723688651439806e-06, | |
| "loss": 0.1289, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.291390728476821, | |
| "grad_norm": 3.831774668567211, | |
| "learning_rate": 6.358581592409881e-06, | |
| "loss": 0.1173, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.300220750551876, | |
| "grad_norm": 3.8338064659443223, | |
| "learning_rate": 6.344783157109114e-06, | |
| "loss": 0.1235, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 4.309050772626931, | |
| "grad_norm": 3.166014796852472, | |
| "learning_rate": 6.3309736726142965e-06, | |
| "loss": 0.1135, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.317880794701987, | |
| "grad_norm": 3.810884649923512, | |
| "learning_rate": 6.317153252388834e-06, | |
| "loss": 0.1335, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 4.326710816777042, | |
| "grad_norm": 3.5826711577092336, | |
| "learning_rate": 6.303322009985984e-06, | |
| "loss": 0.1174, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.335540838852097, | |
| "grad_norm": 4.868623015316583, | |
| "learning_rate": 6.289480059047915e-06, | |
| "loss": 0.1232, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 4.344370860927152, | |
| "grad_norm": 3.0030671476392663, | |
| "learning_rate": 6.27562751330479e-06, | |
| "loss": 0.1272, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.3532008830022075, | |
| "grad_norm": 6.0050086874338735, | |
| "learning_rate": 6.261764486573816e-06, | |
| "loss": 0.1262, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 4.362030905077263, | |
| "grad_norm": 3.1240446946258453, | |
| "learning_rate": 6.247891092758319e-06, | |
| "loss": 0.1229, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.370860927152318, | |
| "grad_norm": 3.7314574663506206, | |
| "learning_rate": 6.2340074458468014e-06, | |
| "loss": 0.1235, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.379690949227373, | |
| "grad_norm": 3.6212119224154615, | |
| "learning_rate": 6.220113659912012e-06, | |
| "loss": 0.1204, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.3885209713024285, | |
| "grad_norm": 3.8059191182004644, | |
| "learning_rate": 6.206209849110001e-06, | |
| "loss": 0.1294, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 4.397350993377484, | |
| "grad_norm": 3.5799883787764553, | |
| "learning_rate": 6.1922961276791925e-06, | |
| "loss": 0.1331, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.406181015452539, | |
| "grad_norm": 3.6288692302119627, | |
| "learning_rate": 6.1783726099394324e-06, | |
| "loss": 0.1211, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 4.415011037527594, | |
| "grad_norm": 3.591077898604039, | |
| "learning_rate": 6.164439410291061e-06, | |
| "loss": 0.1236, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.423841059602649, | |
| "grad_norm": 3.3496888954441495, | |
| "learning_rate": 6.150496643213969e-06, | |
| "loss": 0.122, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 4.432671081677704, | |
| "grad_norm": 3.442503005211259, | |
| "learning_rate": 6.136544423266651e-06, | |
| "loss": 0.1218, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 4.441501103752759, | |
| "grad_norm": 3.0393452192641446, | |
| "learning_rate": 6.122582865085278e-06, | |
| "loss": 0.1331, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 4.450331125827814, | |
| "grad_norm": 3.7451626002092833, | |
| "learning_rate": 6.108612083382739e-06, | |
| "loss": 0.1203, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 4.459161147902869, | |
| "grad_norm": 2.8171549167813676, | |
| "learning_rate": 6.094632192947711e-06, | |
| "loss": 0.1253, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.467991169977925, | |
| "grad_norm": 3.0005030501015066, | |
| "learning_rate": 6.08064330864371e-06, | |
| "loss": 0.1194, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 4.47682119205298, | |
| "grad_norm": 3.446034614704431, | |
| "learning_rate": 6.066645545408149e-06, | |
| "loss": 0.1242, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 4.485651214128035, | |
| "grad_norm": 4.1612917799048645, | |
| "learning_rate": 6.052639018251394e-06, | |
| "loss": 0.1305, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 4.49448123620309, | |
| "grad_norm": 3.2764056643756367, | |
| "learning_rate": 6.03862384225582e-06, | |
| "loss": 0.1158, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 4.503311258278146, | |
| "grad_norm": 2.9771987936665467, | |
| "learning_rate": 6.024600132574855e-06, | |
| "loss": 0.1248, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.512141280353201, | |
| "grad_norm": 3.4583533371552613, | |
| "learning_rate": 6.010568004432055e-06, | |
| "loss": 0.1154, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 4.520971302428256, | |
| "grad_norm": 2.7293827349676225, | |
| "learning_rate": 5.9965275731201364e-06, | |
| "loss": 0.1182, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 4.529801324503311, | |
| "grad_norm": 4.927258817051789, | |
| "learning_rate": 5.982478954000042e-06, | |
| "loss": 0.1279, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 4.5386313465783665, | |
| "grad_norm": 4.805592514114645, | |
| "learning_rate": 5.968422262499983e-06, | |
| "loss": 0.1258, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 4.547461368653422, | |
| "grad_norm": 4.265484292040181, | |
| "learning_rate": 5.9543576141145035e-06, | |
| "loss": 0.1175, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.556291390728477, | |
| "grad_norm": 3.0248837566549027, | |
| "learning_rate": 5.940285124403517e-06, | |
| "loss": 0.1311, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 4.565121412803532, | |
| "grad_norm": 3.641176693734621, | |
| "learning_rate": 5.926204908991366e-06, | |
| "loss": 0.1249, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 4.5739514348785875, | |
| "grad_norm": 5.1894209773075985, | |
| "learning_rate": 5.912117083565874e-06, | |
| "loss": 0.127, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 4.582781456953643, | |
| "grad_norm": 3.5669728878951465, | |
| "learning_rate": 5.898021763877388e-06, | |
| "loss": 0.1392, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 4.591611479028698, | |
| "grad_norm": 4.7615414125034885, | |
| "learning_rate": 5.883919065737827e-06, | |
| "loss": 0.1327, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.600441501103752, | |
| "grad_norm": 3.6770168680430966, | |
| "learning_rate": 5.869809105019738e-06, | |
| "loss": 0.1217, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 4.609271523178808, | |
| "grad_norm": 3.3713572186023755, | |
| "learning_rate": 5.85569199765534e-06, | |
| "loss": 0.1125, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 4.618101545253863, | |
| "grad_norm": 3.0913932636208026, | |
| "learning_rate": 5.841567859635572e-06, | |
| "loss": 0.1291, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 4.626931567328918, | |
| "grad_norm": 4.9934293415631705, | |
| "learning_rate": 5.827436807009133e-06, | |
| "loss": 0.1307, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 4.635761589403973, | |
| "grad_norm": 2.8705429059333145, | |
| "learning_rate": 5.813298955881542e-06, | |
| "loss": 0.1237, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.644591611479028, | |
| "grad_norm": 3.5167517180154344, | |
| "learning_rate": 5.799154422414174e-06, | |
| "loss": 0.1202, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 4.653421633554084, | |
| "grad_norm": 3.2585695646199357, | |
| "learning_rate": 5.785003322823307e-06, | |
| "loss": 0.1143, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 4.662251655629139, | |
| "grad_norm": 3.6582695725781194, | |
| "learning_rate": 5.7708457733791715e-06, | |
| "loss": 0.129, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 4.671081677704194, | |
| "grad_norm": 4.214177518301865, | |
| "learning_rate": 5.756681890404987e-06, | |
| "loss": 0.1208, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 4.679911699779249, | |
| "grad_norm": 4.816984959120091, | |
| "learning_rate": 5.7425117902760195e-06, | |
| "loss": 0.1281, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.688741721854305, | |
| "grad_norm": 3.6596193915847146, | |
| "learning_rate": 5.72833558941861e-06, | |
| "loss": 0.1249, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 4.69757174392936, | |
| "grad_norm": 3.613652221877597, | |
| "learning_rate": 5.714153404309228e-06, | |
| "loss": 0.1275, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 4.706401766004415, | |
| "grad_norm": 3.502281640812166, | |
| "learning_rate": 5.6999653514735124e-06, | |
| "loss": 0.1273, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 4.71523178807947, | |
| "grad_norm": 3.1938561964909113, | |
| "learning_rate": 5.685771547485312e-06, | |
| "loss": 0.1212, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 4.7240618101545255, | |
| "grad_norm": 2.650864800191451, | |
| "learning_rate": 5.671572108965729e-06, | |
| "loss": 0.1168, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.732891832229581, | |
| "grad_norm": 3.9444065764216742, | |
| "learning_rate": 5.65736715258216e-06, | |
| "loss": 0.1167, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 4.741721854304636, | |
| "grad_norm": 3.3249979553308533, | |
| "learning_rate": 5.643156795047343e-06, | |
| "loss": 0.1165, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 4.750551876379691, | |
| "grad_norm": 3.078467649444777, | |
| "learning_rate": 5.628941153118388e-06, | |
| "loss": 0.1124, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 4.7593818984547465, | |
| "grad_norm": 2.975540061747619, | |
| "learning_rate": 5.6147203435958246e-06, | |
| "loss": 0.1224, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 4.768211920529802, | |
| "grad_norm": 5.628155349580698, | |
| "learning_rate": 5.600494483322643e-06, | |
| "loss": 0.1121, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.777041942604857, | |
| "grad_norm": 3.307670376851456, | |
| "learning_rate": 5.586263689183332e-06, | |
| "loss": 0.1169, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 4.785871964679911, | |
| "grad_norm": 3.195864557983689, | |
| "learning_rate": 5.572028078102917e-06, | |
| "loss": 0.1239, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 4.7947019867549665, | |
| "grad_norm": 3.444073709671454, | |
| "learning_rate": 5.557787767046001e-06, | |
| "loss": 0.117, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 4.803532008830022, | |
| "grad_norm": 3.6483842184037507, | |
| "learning_rate": 5.543542873015806e-06, | |
| "loss": 0.1235, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 4.812362030905077, | |
| "grad_norm": 3.0846288659562515, | |
| "learning_rate": 5.529293513053207e-06, | |
| "loss": 0.1168, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.821192052980132, | |
| "grad_norm": 3.510732471218089, | |
| "learning_rate": 5.515039804235772e-06, | |
| "loss": 0.1133, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 4.830022075055187, | |
| "grad_norm": 3.9399472199228485, | |
| "learning_rate": 5.5007818636768055e-06, | |
| "loss": 0.1186, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 4.838852097130243, | |
| "grad_norm": 4.286368966597208, | |
| "learning_rate": 5.486519808524374e-06, | |
| "loss": 0.1236, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 4.847682119205298, | |
| "grad_norm": 3.250883345974673, | |
| "learning_rate": 5.472253755960358e-06, | |
| "loss": 0.1154, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 4.856512141280353, | |
| "grad_norm": 8.8175392593957, | |
| "learning_rate": 5.457983823199475e-06, | |
| "loss": 0.1318, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.865342163355408, | |
| "grad_norm": 3.6702454418250077, | |
| "learning_rate": 5.443710127488331e-06, | |
| "loss": 0.1248, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 4.874172185430464, | |
| "grad_norm": 6.984303526600391, | |
| "learning_rate": 5.429432786104446e-06, | |
| "loss": 0.1382, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 4.883002207505519, | |
| "grad_norm": 10.155902456843055, | |
| "learning_rate": 5.415151916355292e-06, | |
| "loss": 0.12, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 4.891832229580574, | |
| "grad_norm": 2.871691194061296, | |
| "learning_rate": 5.400867635577335e-06, | |
| "loss": 0.1179, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 4.900662251655629, | |
| "grad_norm": 3.1656329943567343, | |
| "learning_rate": 5.3865800611350634e-06, | |
| "loss": 0.1246, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 4.9094922737306845, | |
| "grad_norm": 3.952924417000145, | |
| "learning_rate": 5.372289310420032e-06, | |
| "loss": 0.12, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 4.91832229580574, | |
| "grad_norm": 3.1919392731929856, | |
| "learning_rate": 5.35799550084989e-06, | |
| "loss": 0.1233, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 4.927152317880795, | |
| "grad_norm": 7.4591084325571515, | |
| "learning_rate": 5.343698749867421e-06, | |
| "loss": 0.1074, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 4.93598233995585, | |
| "grad_norm": 2.6679369117650973, | |
| "learning_rate": 5.329399174939572e-06, | |
| "loss": 0.1193, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 4.9448123620309055, | |
| "grad_norm": 3.4742447014481592, | |
| "learning_rate": 5.315096893556497e-06, | |
| "loss": 0.1259, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.95364238410596, | |
| "grad_norm": 5.4465711698009995, | |
| "learning_rate": 5.300792023230587e-06, | |
| "loss": 0.1195, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 4.962472406181016, | |
| "grad_norm": 3.968307067870439, | |
| "learning_rate": 5.2864846814955e-06, | |
| "loss": 0.118, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 4.97130242825607, | |
| "grad_norm": 3.2506827692595475, | |
| "learning_rate": 5.272174985905207e-06, | |
| "loss": 0.1179, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 4.9801324503311255, | |
| "grad_norm": 3.8263525595866374, | |
| "learning_rate": 5.257863054033012e-06, | |
| "loss": 0.1292, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 4.988962472406181, | |
| "grad_norm": 4.674100426113167, | |
| "learning_rate": 5.243549003470599e-06, | |
| "loss": 0.1238, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 4.997792494481236, | |
| "grad_norm": 3.3561839120463244, | |
| "learning_rate": 5.229232951827054e-06, | |
| "loss": 0.118, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 5.006622516556291, | |
| "grad_norm": 3.146917355280202, | |
| "learning_rate": 5.2149150167279106e-06, | |
| "loss": 0.1024, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 5.015452538631346, | |
| "grad_norm": 3.22130119170267, | |
| "learning_rate": 5.200595315814174e-06, | |
| "loss": 0.0797, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 5.024282560706402, | |
| "grad_norm": 3.4535124173857583, | |
| "learning_rate": 5.18627396674136e-06, | |
| "loss": 0.0892, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 5.033112582781457, | |
| "grad_norm": 4.461267702272426, | |
| "learning_rate": 5.17195108717852e-06, | |
| "loss": 0.0826, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 5.041942604856512, | |
| "grad_norm": 5.477703530057339, | |
| "learning_rate": 5.15762679480729e-06, | |
| "loss": 0.0892, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 5.050772626931567, | |
| "grad_norm": 3.141786969910022, | |
| "learning_rate": 5.143301207320909e-06, | |
| "loss": 0.0758, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 5.059602649006623, | |
| "grad_norm": 3.839846617800994, | |
| "learning_rate": 5.128974442423254e-06, | |
| "loss": 0.0817, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 5.068432671081678, | |
| "grad_norm": 3.118483557718335, | |
| "learning_rate": 5.114646617827884e-06, | |
| "loss": 0.0833, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 5.077262693156733, | |
| "grad_norm": 4.089376916511554, | |
| "learning_rate": 5.100317851257057e-06, | |
| "loss": 0.0835, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 5.086092715231788, | |
| "grad_norm": 4.861334723664738, | |
| "learning_rate": 5.085988260440776e-06, | |
| "loss": 0.0823, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 5.0949227373068435, | |
| "grad_norm": 4.098126371178599, | |
| "learning_rate": 5.0716579631158124e-06, | |
| "loss": 0.0775, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 5.103752759381899, | |
| "grad_norm": 3.2774918605394205, | |
| "learning_rate": 5.057327077024745e-06, | |
| "loss": 0.0748, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 5.112582781456954, | |
| "grad_norm": 3.1759022392991496, | |
| "learning_rate": 5.0429957199149905e-06, | |
| "loss": 0.076, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 5.121412803532009, | |
| "grad_norm": 3.0604826124075113, | |
| "learning_rate": 5.028664009537835e-06, | |
| "loss": 0.0729, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 5.1302428256070645, | |
| "grad_norm": 3.578159593479843, | |
| "learning_rate": 5.014332063647462e-06, | |
| "loss": 0.075, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 5.139072847682119, | |
| "grad_norm": 4.62235484562011, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0796, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 5.147902869757174, | |
| "grad_norm": 5.19079093056643, | |
| "learning_rate": 4.985667936352538e-06, | |
| "loss": 0.0853, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 5.156732891832229, | |
| "grad_norm": 3.3663631673095127, | |
| "learning_rate": 4.971335990462168e-06, | |
| "loss": 0.0836, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 5.1655629139072845, | |
| "grad_norm": 5.810957450856354, | |
| "learning_rate": 4.95700428008501e-06, | |
| "loss": 0.0815, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 5.17439293598234, | |
| "grad_norm": 16.388157198181638, | |
| "learning_rate": 4.942672922975255e-06, | |
| "loss": 0.0796, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 5.183222958057395, | |
| "grad_norm": 4.60123438336755, | |
| "learning_rate": 4.928342036884189e-06, | |
| "loss": 0.0899, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 5.19205298013245, | |
| "grad_norm": 15.655999777602075, | |
| "learning_rate": 4.914011739559225e-06, | |
| "loss": 0.0756, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 5.200883002207505, | |
| "grad_norm": 3.9165053741788496, | |
| "learning_rate": 4.899682148742944e-06, | |
| "loss": 0.0753, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 5.209713024282561, | |
| "grad_norm": 2.9838330327043483, | |
| "learning_rate": 4.8853533821721175e-06, | |
| "loss": 0.0792, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.218543046357616, | |
| "grad_norm": 3.9762696232752983, | |
| "learning_rate": 4.871025557576747e-06, | |
| "loss": 0.084, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 5.227373068432671, | |
| "grad_norm": 3.613534212301652, | |
| "learning_rate": 4.8566987926790946e-06, | |
| "loss": 0.077, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 5.236203090507726, | |
| "grad_norm": 4.094916192512793, | |
| "learning_rate": 4.8423732051927115e-06, | |
| "loss": 0.0832, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 5.245033112582782, | |
| "grad_norm": 3.2032702626183305, | |
| "learning_rate": 4.82804891282148e-06, | |
| "loss": 0.0867, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 5.253863134657837, | |
| "grad_norm": 2.8963088368027043, | |
| "learning_rate": 4.813726033258643e-06, | |
| "loss": 0.0774, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 5.262693156732892, | |
| "grad_norm": 3.5228731375214544, | |
| "learning_rate": 4.799404684185828e-06, | |
| "loss": 0.0831, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 5.271523178807947, | |
| "grad_norm": 4.284381838964143, | |
| "learning_rate": 4.78508498327209e-06, | |
| "loss": 0.0777, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 5.2803532008830025, | |
| "grad_norm": 2.9560596560607406, | |
| "learning_rate": 4.770767048172948e-06, | |
| "loss": 0.0849, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 5.289183222958058, | |
| "grad_norm": 4.987385111477333, | |
| "learning_rate": 4.756450996529403e-06, | |
| "loss": 0.081, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 5.298013245033113, | |
| "grad_norm": 3.6167651983388187, | |
| "learning_rate": 4.742136945966991e-06, | |
| "loss": 0.0837, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.306843267108167, | |
| "grad_norm": 3.09609946262437, | |
| "learning_rate": 4.727825014094795e-06, | |
| "loss": 0.0793, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 5.315673289183223, | |
| "grad_norm": 10.109207349939986, | |
| "learning_rate": 4.713515318504501e-06, | |
| "loss": 0.0834, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 5.324503311258278, | |
| "grad_norm": 4.0032568996691085, | |
| "learning_rate": 4.699207976769416e-06, | |
| "loss": 0.0833, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 3.0226202484588844, | |
| "learning_rate": 4.684903106443504e-06, | |
| "loss": 0.0795, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 5.342163355408388, | |
| "grad_norm": 4.173970428799518, | |
| "learning_rate": 4.670600825060429e-06, | |
| "loss": 0.085, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 5.3509933774834435, | |
| "grad_norm": 2.9678997176733466, | |
| "learning_rate": 4.656301250132581e-06, | |
| "loss": 0.0799, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 5.359823399558499, | |
| "grad_norm": 3.9871235739201376, | |
| "learning_rate": 4.6420044991501104e-06, | |
| "loss": 0.0706, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 5.368653421633554, | |
| "grad_norm": 2.8970879482029943, | |
| "learning_rate": 4.627710689579968e-06, | |
| "loss": 0.0797, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 5.377483443708609, | |
| "grad_norm": 4.30580422384717, | |
| "learning_rate": 4.613419938864937e-06, | |
| "loss": 0.0914, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 5.386313465783664, | |
| "grad_norm": 3.7076631476401816, | |
| "learning_rate": 4.599132364422666e-06, | |
| "loss": 0.0761, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.39514348785872, | |
| "grad_norm": 3.661784528175195, | |
| "learning_rate": 4.58484808364471e-06, | |
| "loss": 0.0805, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 5.403973509933775, | |
| "grad_norm": 4.3336584620699785, | |
| "learning_rate": 4.570567213895555e-06, | |
| "loss": 0.0842, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 5.41280353200883, | |
| "grad_norm": 3.6059544174962697, | |
| "learning_rate": 4.556289872511669e-06, | |
| "loss": 0.0874, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 5.421633554083885, | |
| "grad_norm": 3.1768601421786746, | |
| "learning_rate": 4.542016176800527e-06, | |
| "loss": 0.0727, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 5.430463576158941, | |
| "grad_norm": 4.646440617522754, | |
| "learning_rate": 4.527746244039644e-06, | |
| "loss": 0.091, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 5.439293598233996, | |
| "grad_norm": 3.417104016568786, | |
| "learning_rate": 4.513480191475627e-06, | |
| "loss": 0.0862, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 5.448123620309051, | |
| "grad_norm": 3.748333035741698, | |
| "learning_rate": 4.499218136323197e-06, | |
| "loss": 0.0797, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 5.456953642384106, | |
| "grad_norm": 4.706982156041434, | |
| "learning_rate": 4.4849601957642295e-06, | |
| "loss": 0.0839, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 5.4657836644591615, | |
| "grad_norm": 13.040031956770356, | |
| "learning_rate": 4.470706486946797e-06, | |
| "loss": 0.0895, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 5.474613686534216, | |
| "grad_norm": 12.154055803835261, | |
| "learning_rate": 4.456457126984196e-06, | |
| "loss": 0.0934, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.483443708609272, | |
| "grad_norm": 3.2227180307772314, | |
| "learning_rate": 4.4422122329539996e-06, | |
| "loss": 0.0874, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 5.492273730684326, | |
| "grad_norm": 3.949315508621148, | |
| "learning_rate": 4.427971921897086e-06, | |
| "loss": 0.087, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 5.501103752759382, | |
| "grad_norm": 6.060786587045195, | |
| "learning_rate": 4.413736310816669e-06, | |
| "loss": 0.0969, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 5.509933774834437, | |
| "grad_norm": 3.513581979781065, | |
| "learning_rate": 4.399505516677358e-06, | |
| "loss": 0.0957, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 5.518763796909492, | |
| "grad_norm": 3.2801102852428503, | |
| "learning_rate": 4.385279656404178e-06, | |
| "loss": 0.0899, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 5.527593818984547, | |
| "grad_norm": 4.6833146539430945, | |
| "learning_rate": 4.371058846881614e-06, | |
| "loss": 0.0828, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 5.5364238410596025, | |
| "grad_norm": 3.111198232980169, | |
| "learning_rate": 4.356843204952657e-06, | |
| "loss": 0.0794, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 5.545253863134658, | |
| "grad_norm": 4.6270477782388, | |
| "learning_rate": 4.3426328474178405e-06, | |
| "loss": 0.0873, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 5.554083885209713, | |
| "grad_norm": 3.1907808379567144, | |
| "learning_rate": 4.328427891034273e-06, | |
| "loss": 0.0929, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 5.562913907284768, | |
| "grad_norm": 3.3337597317002907, | |
| "learning_rate": 4.3142284525146915e-06, | |
| "loss": 0.0884, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.571743929359823, | |
| "grad_norm": 4.059587213941394, | |
| "learning_rate": 4.300034648526489e-06, | |
| "loss": 0.0854, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 5.580573951434879, | |
| "grad_norm": 3.6449230362767815, | |
| "learning_rate": 4.2858465956907726e-06, | |
| "loss": 0.0778, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 5.589403973509934, | |
| "grad_norm": 2.8635560754823204, | |
| "learning_rate": 4.271664410581392e-06, | |
| "loss": 0.0834, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 5.598233995584989, | |
| "grad_norm": 3.8467733362478063, | |
| "learning_rate": 4.257488209723981e-06, | |
| "loss": 0.0773, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 5.607064017660044, | |
| "grad_norm": 4.272168759114605, | |
| "learning_rate": 4.243318109595014e-06, | |
| "loss": 0.0785, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 5.6158940397351, | |
| "grad_norm": 3.731464888209812, | |
| "learning_rate": 4.229154226620832e-06, | |
| "loss": 0.0835, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 5.624724061810155, | |
| "grad_norm": 4.504268895815233, | |
| "learning_rate": 4.2149966771766945e-06, | |
| "loss": 0.0856, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 5.63355408388521, | |
| "grad_norm": 2.7666800996784935, | |
| "learning_rate": 4.200845577585827e-06, | |
| "loss": 0.081, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 5.642384105960264, | |
| "grad_norm": 2.5726250105979114, | |
| "learning_rate": 4.186701044118459e-06, | |
| "loss": 0.0775, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 5.6512141280353205, | |
| "grad_norm": 3.138450144905294, | |
| "learning_rate": 4.1725631929908684e-06, | |
| "loss": 0.0766, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.660044150110375, | |
| "grad_norm": 3.1299019210977743, | |
| "learning_rate": 4.158432140364431e-06, | |
| "loss": 0.0882, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 5.66887417218543, | |
| "grad_norm": 2.6664807336372722, | |
| "learning_rate": 4.1443080023446605e-06, | |
| "loss": 0.0782, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 5.677704194260485, | |
| "grad_norm": 4.889136489228772, | |
| "learning_rate": 4.130190894980262e-06, | |
| "loss": 0.078, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 5.686534216335541, | |
| "grad_norm": 3.1885543878873195, | |
| "learning_rate": 4.116080934262175e-06, | |
| "loss": 0.0704, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 5.695364238410596, | |
| "grad_norm": 3.371577596141025, | |
| "learning_rate": 4.101978236122613e-06, | |
| "loss": 0.0814, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 5.704194260485651, | |
| "grad_norm": 4.978559831267471, | |
| "learning_rate": 4.087882916434126e-06, | |
| "loss": 0.0799, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 5.713024282560706, | |
| "grad_norm": 3.7887338629987357, | |
| "learning_rate": 4.0737950910086354e-06, | |
| "loss": 0.0895, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 5.7218543046357615, | |
| "grad_norm": 3.260423658931666, | |
| "learning_rate": 4.059714875596486e-06, | |
| "loss": 0.0811, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 5.730684326710817, | |
| "grad_norm": 3.052678502999751, | |
| "learning_rate": 4.045642385885497e-06, | |
| "loss": 0.0662, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 5.739514348785872, | |
| "grad_norm": 3.502216524172591, | |
| "learning_rate": 4.0315777375000185e-06, | |
| "loss": 0.0759, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.748344370860927, | |
| "grad_norm": 5.2989447051538034, | |
| "learning_rate": 4.017521045999961e-06, | |
| "loss": 0.072, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 5.757174392935982, | |
| "grad_norm": 4.15705549541637, | |
| "learning_rate": 4.003472426879866e-06, | |
| "loss": 0.0791, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 5.766004415011038, | |
| "grad_norm": 3.24016448418588, | |
| "learning_rate": 3.989431995567947e-06, | |
| "loss": 0.0844, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 5.774834437086093, | |
| "grad_norm": 6.233232207035722, | |
| "learning_rate": 3.975399867425146e-06, | |
| "loss": 0.0843, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 5.783664459161148, | |
| "grad_norm": 4.514298208802796, | |
| "learning_rate": 3.961376157744183e-06, | |
| "loss": 0.0918, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 5.792494481236203, | |
| "grad_norm": 5.8826321368622, | |
| "learning_rate": 3.947360981748607e-06, | |
| "loss": 0.0923, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 5.801324503311259, | |
| "grad_norm": 3.6129241718657417, | |
| "learning_rate": 3.933354454591851e-06, | |
| "loss": 0.0836, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 5.810154525386314, | |
| "grad_norm": 3.795101819789711, | |
| "learning_rate": 3.9193566913562915e-06, | |
| "loss": 0.0731, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 5.818984547461369, | |
| "grad_norm": 3.542191525075181, | |
| "learning_rate": 3.9053678070522904e-06, | |
| "loss": 0.0785, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 5.827814569536423, | |
| "grad_norm": 2.722735213413911, | |
| "learning_rate": 3.891387916617261e-06, | |
| "loss": 0.0782, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.8366445916114795, | |
| "grad_norm": 3.3247381022626166, | |
| "learning_rate": 3.877417134914724e-06, | |
| "loss": 0.0745, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 5.845474613686534, | |
| "grad_norm": 2.6483222669930866, | |
| "learning_rate": 3.863455576733349e-06, | |
| "loss": 0.0801, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 5.854304635761589, | |
| "grad_norm": 3.0329370213321165, | |
| "learning_rate": 3.849503356786034e-06, | |
| "loss": 0.0786, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 5.863134657836644, | |
| "grad_norm": 3.691105022986808, | |
| "learning_rate": 3.83556058970894e-06, | |
| "loss": 0.087, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 5.8719646799117, | |
| "grad_norm": 3.8162843564535036, | |
| "learning_rate": 3.821627390060568e-06, | |
| "loss": 0.0831, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 5.880794701986755, | |
| "grad_norm": 4.261183833481365, | |
| "learning_rate": 3.807703872320809e-06, | |
| "loss": 0.096, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 5.88962472406181, | |
| "grad_norm": 4.100395781225104, | |
| "learning_rate": 3.79379015089e-06, | |
| "loss": 0.0762, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 5.898454746136865, | |
| "grad_norm": 2.9729742915357225, | |
| "learning_rate": 3.7798863400879894e-06, | |
| "loss": 0.072, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 5.9072847682119205, | |
| "grad_norm": 3.2063765133460937, | |
| "learning_rate": 3.7659925541532006e-06, | |
| "loss": 0.0769, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 5.916114790286976, | |
| "grad_norm": 3.3133844049925525, | |
| "learning_rate": 3.752108907241682e-06, | |
| "loss": 0.0901, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.924944812362031, | |
| "grad_norm": 3.875509620548704, | |
| "learning_rate": 3.738235513426184e-06, | |
| "loss": 0.0808, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 5.933774834437086, | |
| "grad_norm": 4.453172114953024, | |
| "learning_rate": 3.7243724866952114e-06, | |
| "loss": 0.0802, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 5.942604856512141, | |
| "grad_norm": 3.883050398417393, | |
| "learning_rate": 3.710519940952085e-06, | |
| "loss": 0.0786, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 5.951434878587197, | |
| "grad_norm": 4.886476235877721, | |
| "learning_rate": 3.6966779900140193e-06, | |
| "loss": 0.095, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 5.960264900662252, | |
| "grad_norm": 4.136874645111521, | |
| "learning_rate": 3.6828467476111664e-06, | |
| "loss": 0.0857, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 5.969094922737307, | |
| "grad_norm": 3.097076951530958, | |
| "learning_rate": 3.6690263273857035e-06, | |
| "loss": 0.0897, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 5.977924944812362, | |
| "grad_norm": 3.7197135406624877, | |
| "learning_rate": 3.6552168428908886e-06, | |
| "loss": 0.0807, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 5.986754966887418, | |
| "grad_norm": 3.8598772604803866, | |
| "learning_rate": 3.6414184075901206e-06, | |
| "loss": 0.0776, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 5.995584988962472, | |
| "grad_norm": 3.763686510723587, | |
| "learning_rate": 3.62763113485602e-06, | |
| "loss": 0.0868, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 6.004415011037527, | |
| "grad_norm": 2.5935623419254257, | |
| "learning_rate": 3.6138551379694936e-06, | |
| "loss": 0.0637, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 6.013245033112582, | |
| "grad_norm": 2.119450996938125, | |
| "learning_rate": 3.6000905301187953e-06, | |
| "loss": 0.0455, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 6.022075055187638, | |
| "grad_norm": 7.1116345607405735, | |
| "learning_rate": 3.586337424398609e-06, | |
| "loss": 0.0482, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 6.030905077262693, | |
| "grad_norm": 2.427775209387265, | |
| "learning_rate": 3.5725959338091133e-06, | |
| "loss": 0.0444, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 6.039735099337748, | |
| "grad_norm": 2.2894656942221374, | |
| "learning_rate": 3.5588661712550464e-06, | |
| "loss": 0.0438, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 6.048565121412803, | |
| "grad_norm": 3.1225628225418616, | |
| "learning_rate": 3.545148249544793e-06, | |
| "loss": 0.0432, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 6.057395143487859, | |
| "grad_norm": 3.904197381263508, | |
| "learning_rate": 3.5314422813894413e-06, | |
| "loss": 0.044, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 6.066225165562914, | |
| "grad_norm": 2.980767396999886, | |
| "learning_rate": 3.517748379401872e-06, | |
| "loss": 0.0387, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 6.075055187637969, | |
| "grad_norm": 4.055772492425105, | |
| "learning_rate": 3.5040666560958246e-06, | |
| "loss": 0.0499, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 6.083885209713024, | |
| "grad_norm": 2.95448216758592, | |
| "learning_rate": 3.4903972238849727e-06, | |
| "loss": 0.0454, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 6.0927152317880795, | |
| "grad_norm": 3.745803810242431, | |
| "learning_rate": 3.4767401950820003e-06, | |
| "loss": 0.0472, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 6.101545253863135, | |
| "grad_norm": 4.096845749537238, | |
| "learning_rate": 3.4630956818976875e-06, | |
| "loss": 0.0488, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 6.11037527593819, | |
| "grad_norm": 4.849803606877378, | |
| "learning_rate": 3.4494637964399723e-06, | |
| "loss": 0.0558, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 6.119205298013245, | |
| "grad_norm": 4.154358398574059, | |
| "learning_rate": 3.4358446507130503e-06, | |
| "loss": 0.0472, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 6.1280353200883, | |
| "grad_norm": 3.3830561830802335, | |
| "learning_rate": 3.4222383566164314e-06, | |
| "loss": 0.0475, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 6.136865342163356, | |
| "grad_norm": 3.442379096610827, | |
| "learning_rate": 3.408645025944042e-06, | |
| "loss": 0.0483, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 6.145695364238411, | |
| "grad_norm": 4.742598864942118, | |
| "learning_rate": 3.3950647703832907e-06, | |
| "loss": 0.0431, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 6.154525386313466, | |
| "grad_norm": 2.697395245426002, | |
| "learning_rate": 3.3814977015141576e-06, | |
| "loss": 0.0408, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 6.163355408388521, | |
| "grad_norm": 3.1100176849979584, | |
| "learning_rate": 3.3679439308082777e-06, | |
| "loss": 0.0442, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 6.172185430463577, | |
| "grad_norm": 3.976759773598317, | |
| "learning_rate": 3.3544035696280264e-06, | |
| "loss": 0.0459, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 6.181015452538631, | |
| "grad_norm": 2.84177079378505, | |
| "learning_rate": 3.340876729225595e-06, | |
| "loss": 0.046, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.189845474613686, | |
| "grad_norm": 4.812222367835127, | |
| "learning_rate": 3.327363520742087e-06, | |
| "loss": 0.0461, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 6.198675496688741, | |
| "grad_norm": 2.7388890441297034, | |
| "learning_rate": 3.313864055206607e-06, | |
| "loss": 0.0469, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 6.207505518763797, | |
| "grad_norm": 4.039957059405307, | |
| "learning_rate": 3.3003784435353304e-06, | |
| "loss": 0.0525, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 6.216335540838852, | |
| "grad_norm": 5.104091111546962, | |
| "learning_rate": 3.2869067965306178e-06, | |
| "loss": 0.0426, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 6.225165562913907, | |
| "grad_norm": 3.574412350338948, | |
| "learning_rate": 3.273449224880081e-06, | |
| "loss": 0.0446, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 6.233995584988962, | |
| "grad_norm": 3.8987394939809965, | |
| "learning_rate": 3.260005839155691e-06, | |
| "loss": 0.0476, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 6.242825607064018, | |
| "grad_norm": 3.5137110339373736, | |
| "learning_rate": 3.2465767498128596e-06, | |
| "loss": 0.0437, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 6.251655629139073, | |
| "grad_norm": 4.725561434628457, | |
| "learning_rate": 3.233162067189533e-06, | |
| "loss": 0.0426, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 6.260485651214128, | |
| "grad_norm": 3.458802854096362, | |
| "learning_rate": 3.2197619015052893e-06, | |
| "loss": 0.0426, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 6.269315673289183, | |
| "grad_norm": 3.9639317370860603, | |
| "learning_rate": 3.206376362860432e-06, | |
| "loss": 0.0399, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 6.2781456953642385, | |
| "grad_norm": 4.263982376223075, | |
| "learning_rate": 3.1930055612350795e-06, | |
| "loss": 0.0458, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 6.286975717439294, | |
| "grad_norm": 4.373524348005642, | |
| "learning_rate": 3.1796496064882677e-06, | |
| "loss": 0.0473, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 6.295805739514349, | |
| "grad_norm": 2.9952083824369478, | |
| "learning_rate": 3.1663086083570493e-06, | |
| "loss": 0.0419, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 6.304635761589404, | |
| "grad_norm": 10.102286671281984, | |
| "learning_rate": 3.152982676455581e-06, | |
| "loss": 0.0455, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 6.313465783664459, | |
| "grad_norm": 2.991715308021915, | |
| "learning_rate": 3.1396719202742375e-06, | |
| "loss": 0.0401, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 6.322295805739515, | |
| "grad_norm": 2.897320064166137, | |
| "learning_rate": 3.1263764491786984e-06, | |
| "loss": 0.0443, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 6.33112582781457, | |
| "grad_norm": 3.158797117236713, | |
| "learning_rate": 3.1130963724090626e-06, | |
| "loss": 0.0437, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 6.339955849889625, | |
| "grad_norm": 3.6137999356757504, | |
| "learning_rate": 3.0998317990789378e-06, | |
| "loss": 0.0449, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 6.3487858719646795, | |
| "grad_norm": 2.807789197244593, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.047, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 6.357615894039735, | |
| "grad_norm": 5.127468598454896, | |
| "learning_rate": 3.0733495985538575e-06, | |
| "loss": 0.0556, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.36644591611479, | |
| "grad_norm": 4.1290994562688885, | |
| "learning_rate": 3.0601321889456378e-06, | |
| "loss": 0.0456, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 6.375275938189845, | |
| "grad_norm": 4.334626352337251, | |
| "learning_rate": 3.046930717948604e-06, | |
| "loss": 0.0417, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 6.3841059602649, | |
| "grad_norm": 2.3687486063543313, | |
| "learning_rate": 3.033745294030517e-06, | |
| "loss": 0.0435, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 6.392935982339956, | |
| "grad_norm": 4.692486058475576, | |
| "learning_rate": 3.0205760255272874e-06, | |
| "loss": 0.0482, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 6.401766004415011, | |
| "grad_norm": 3.9335564896963024, | |
| "learning_rate": 3.007423020642084e-06, | |
| "loss": 0.0489, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 6.410596026490066, | |
| "grad_norm": 2.524593452654292, | |
| "learning_rate": 2.9942863874444565e-06, | |
| "loss": 0.0385, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 6.419426048565121, | |
| "grad_norm": 3.0344790063597635, | |
| "learning_rate": 2.981166233869429e-06, | |
| "loss": 0.0393, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 6.428256070640177, | |
| "grad_norm": 2.9403456255407785, | |
| "learning_rate": 2.9680626677166324e-06, | |
| "loss": 0.0385, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 6.437086092715232, | |
| "grad_norm": 2.783740570779497, | |
| "learning_rate": 2.9549757966494053e-06, | |
| "loss": 0.045, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 6.445916114790287, | |
| "grad_norm": 3.0509378537516665, | |
| "learning_rate": 2.9419057281939106e-06, | |
| "loss": 0.0468, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.454746136865342, | |
| "grad_norm": 2.9824593633203103, | |
| "learning_rate": 2.9288525697382623e-06, | |
| "loss": 0.0447, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 6.4635761589403975, | |
| "grad_norm": 3.27801247712669, | |
| "learning_rate": 2.9158164285316356e-06, | |
| "loss": 0.0458, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.472406181015453, | |
| "grad_norm": 3.91821933231558, | |
| "learning_rate": 2.9027974116833756e-06, | |
| "loss": 0.0514, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 6.481236203090508, | |
| "grad_norm": 8.478846928606096, | |
| "learning_rate": 2.889795626162143e-06, | |
| "loss": 0.0499, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 6.490066225165563, | |
| "grad_norm": 3.715500921167974, | |
| "learning_rate": 2.8768111787950105e-06, | |
| "loss": 0.042, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 6.498896247240618, | |
| "grad_norm": 3.242093553710394, | |
| "learning_rate": 2.8638441762665957e-06, | |
| "loss": 0.0446, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 6.507726269315674, | |
| "grad_norm": 3.816859764003571, | |
| "learning_rate": 2.8508947251181885e-06, | |
| "loss": 0.0603, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 6.516556291390728, | |
| "grad_norm": 3.01932983549725, | |
| "learning_rate": 2.8379629317468604e-06, | |
| "loss": 0.0487, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 6.525386313465784, | |
| "grad_norm": 3.058789798977923, | |
| "learning_rate": 2.825048902404612e-06, | |
| "loss": 0.039, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 6.5342163355408385, | |
| "grad_norm": 3.357980266487943, | |
| "learning_rate": 2.8121527431974838e-06, | |
| "loss": 0.0468, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 6.543046357615894, | |
| "grad_norm": 2.490771021108029, | |
| "learning_rate": 2.799274560084688e-06, | |
| "loss": 0.0458, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 6.551876379690949, | |
| "grad_norm": 2.487429950078564, | |
| "learning_rate": 2.7864144588777403e-06, | |
| "loss": 0.0418, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 6.560706401766004, | |
| "grad_norm": 2.641391890727438, | |
| "learning_rate": 2.77357254523959e-06, | |
| "loss": 0.0386, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 6.569536423841059, | |
| "grad_norm": 3.302767843136541, | |
| "learning_rate": 2.7607489246837505e-06, | |
| "loss": 0.0506, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 6.578366445916115, | |
| "grad_norm": 3.258562996727361, | |
| "learning_rate": 2.74794370257343e-06, | |
| "loss": 0.0458, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 6.58719646799117, | |
| "grad_norm": 4.346618366259902, | |
| "learning_rate": 2.7351569841206792e-06, | |
| "loss": 0.0371, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 6.596026490066225, | |
| "grad_norm": 2.3730837128294104, | |
| "learning_rate": 2.722388874385503e-06, | |
| "loss": 0.0357, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 6.60485651214128, | |
| "grad_norm": 4.284591795453865, | |
| "learning_rate": 2.7096394782750186e-06, | |
| "loss": 0.0486, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 6.613686534216336, | |
| "grad_norm": 2.92935872222048, | |
| "learning_rate": 2.696908900542584e-06, | |
| "loss": 0.0442, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 6.622516556291391, | |
| "grad_norm": 9.296385299682987, | |
| "learning_rate": 2.684197245786938e-06, | |
| "loss": 0.0513, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.631346578366446, | |
| "grad_norm": 3.376367154477601, | |
| "learning_rate": 2.671504618451348e-06, | |
| "loss": 0.0444, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 6.640176600441501, | |
| "grad_norm": 3.917892048977837, | |
| "learning_rate": 2.658831122822735e-06, | |
| "loss": 0.0486, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 6.6490066225165565, | |
| "grad_norm": 17.231818383168626, | |
| "learning_rate": 2.6461768630308326e-06, | |
| "loss": 0.0434, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 6.657836644591612, | |
| "grad_norm": 3.863863279585976, | |
| "learning_rate": 2.633541943047334e-06, | |
| "loss": 0.0362, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 5.481114691878978, | |
| "learning_rate": 2.620926466685013e-06, | |
| "loss": 0.0434, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 6.675496688741722, | |
| "grad_norm": 3.8037293862780017, | |
| "learning_rate": 2.608330537596907e-06, | |
| "loss": 0.0469, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 6.684326710816777, | |
| "grad_norm": 3.45549685183023, | |
| "learning_rate": 2.595754259275428e-06, | |
| "loss": 0.0509, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 6.693156732891833, | |
| "grad_norm": 4.302516265446614, | |
| "learning_rate": 2.583197735051546e-06, | |
| "loss": 0.0489, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 6.701986754966887, | |
| "grad_norm": 4.114088123214382, | |
| "learning_rate": 2.5706610680939186e-06, | |
| "loss": 0.0458, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 6.710816777041942, | |
| "grad_norm": 3.483665626667719, | |
| "learning_rate": 2.5581443614080433e-06, | |
| "loss": 0.0445, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.7196467991169975, | |
| "grad_norm": 2.938065660292941, | |
| "learning_rate": 2.545647717835428e-06, | |
| "loss": 0.0444, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 6.728476821192053, | |
| "grad_norm": 2.6772235779624802, | |
| "learning_rate": 2.53317124005273e-06, | |
| "loss": 0.0394, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 6.737306843267108, | |
| "grad_norm": 2.58024034538905, | |
| "learning_rate": 2.5207150305709167e-06, | |
| "loss": 0.0416, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 6.746136865342163, | |
| "grad_norm": 4.97478354884514, | |
| "learning_rate": 2.5082791917344256e-06, | |
| "loss": 0.05, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 6.754966887417218, | |
| "grad_norm": 3.32999481725283, | |
| "learning_rate": 2.495863825720322e-06, | |
| "loss": 0.0482, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 6.763796909492274, | |
| "grad_norm": 3.968977263707575, | |
| "learning_rate": 2.4834690345374608e-06, | |
| "loss": 0.0417, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 6.772626931567329, | |
| "grad_norm": 5.59969112630162, | |
| "learning_rate": 2.471094920025644e-06, | |
| "loss": 0.0468, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 6.781456953642384, | |
| "grad_norm": 3.47081108347733, | |
| "learning_rate": 2.45874158385479e-06, | |
| "loss": 0.0453, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 6.790286975717439, | |
| "grad_norm": 3.3006938110439594, | |
| "learning_rate": 2.446409127524094e-06, | |
| "loss": 0.0454, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 6.799116997792495, | |
| "grad_norm": 3.582456542441678, | |
| "learning_rate": 2.4340976523611957e-06, | |
| "loss": 0.0408, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.80794701986755, | |
| "grad_norm": 2.5300104870568934, | |
| "learning_rate": 2.4218072595213467e-06, | |
| "loss": 0.0402, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 6.816777041942605, | |
| "grad_norm": 8.097072035556838, | |
| "learning_rate": 2.409538049986576e-06, | |
| "loss": 0.0487, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 6.82560706401766, | |
| "grad_norm": 3.5062271944130416, | |
| "learning_rate": 2.3972901245648724e-06, | |
| "loss": 0.0382, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 6.8344370860927155, | |
| "grad_norm": 3.1945184097837083, | |
| "learning_rate": 2.385063583889335e-06, | |
| "loss": 0.0437, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 6.843267108167771, | |
| "grad_norm": 5.82776778240583, | |
| "learning_rate": 2.3728585284173646e-06, | |
| "loss": 0.0459, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 6.852097130242826, | |
| "grad_norm": 3.4184169416602947, | |
| "learning_rate": 2.3606750584298375e-06, | |
| "loss": 0.0457, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 6.860927152317881, | |
| "grad_norm": 4.6993767838466995, | |
| "learning_rate": 2.348513274030264e-06, | |
| "loss": 0.0455, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 6.8697571743929355, | |
| "grad_norm": 6.543265758489795, | |
| "learning_rate": 2.3363732751439926e-06, | |
| "loss": 0.0401, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 6.878587196467992, | |
| "grad_norm": 2.6541651345708397, | |
| "learning_rate": 2.32425516151736e-06, | |
| "loss": 0.0434, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 6.887417218543046, | |
| "grad_norm": 12.077556689022126, | |
| "learning_rate": 2.3121590327168987e-06, | |
| "loss": 0.0477, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.896247240618101, | |
| "grad_norm": 2.6986067449140974, | |
| "learning_rate": 2.3000849881285016e-06, | |
| "loss": 0.0382, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 6.9050772626931565, | |
| "grad_norm": 3.3763887783500226, | |
| "learning_rate": 2.2880331269566043e-06, | |
| "loss": 0.0488, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 6.913907284768212, | |
| "grad_norm": 3.9259709955934725, | |
| "learning_rate": 2.2760035482233868e-06, | |
| "loss": 0.0446, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 6.922737306843267, | |
| "grad_norm": 3.865114268075013, | |
| "learning_rate": 2.263996350767942e-06, | |
| "loss": 0.0446, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 6.931567328918322, | |
| "grad_norm": 3.3426583787498054, | |
| "learning_rate": 2.2520116332454726e-06, | |
| "loss": 0.0476, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 6.940397350993377, | |
| "grad_norm": 2.4924270184506248, | |
| "learning_rate": 2.240049494126479e-06, | |
| "loss": 0.0416, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 6.949227373068433, | |
| "grad_norm": 11.135568386845106, | |
| "learning_rate": 2.2281100316959476e-06, | |
| "loss": 0.0414, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 6.958057395143488, | |
| "grad_norm": 4.282987321727844, | |
| "learning_rate": 2.2161933440525474e-06, | |
| "loss": 0.0392, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 6.966887417218543, | |
| "grad_norm": 8.49426009428992, | |
| "learning_rate": 2.2042995291078227e-06, | |
| "loss": 0.0419, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 6.975717439293598, | |
| "grad_norm": 14.206836677987118, | |
| "learning_rate": 2.192428684585386e-06, | |
| "loss": 0.0455, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.984547461368654, | |
| "grad_norm": 4.41512924968368, | |
| "learning_rate": 2.180580908020117e-06, | |
| "loss": 0.0442, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 6.993377483443709, | |
| "grad_norm": 2.145978217080188, | |
| "learning_rate": 2.1687562967573645e-06, | |
| "loss": 0.0407, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 7.002207505518764, | |
| "grad_norm": 3.373344674936557, | |
| "learning_rate": 2.156954947952139e-06, | |
| "loss": 0.039, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 7.011037527593819, | |
| "grad_norm": 2.8364513915581777, | |
| "learning_rate": 2.1451769585683196e-06, | |
| "loss": 0.0194, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 7.0198675496688745, | |
| "grad_norm": 2.378533215346128, | |
| "learning_rate": 2.1334224253778628e-06, | |
| "loss": 0.0182, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 7.02869757174393, | |
| "grad_norm": 2.1163864840029305, | |
| "learning_rate": 2.1216914449599905e-06, | |
| "loss": 0.0216, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 7.037527593818985, | |
| "grad_norm": 2.426034363946138, | |
| "learning_rate": 2.109984113700413e-06, | |
| "loss": 0.0175, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 7.046357615894039, | |
| "grad_norm": 2.1292392780706173, | |
| "learning_rate": 2.0983005277905348e-06, | |
| "loss": 0.0218, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 7.0551876379690945, | |
| "grad_norm": 5.143418342254754, | |
| "learning_rate": 2.0866407832266506e-06, | |
| "loss": 0.0191, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 7.06401766004415, | |
| "grad_norm": 2.76679698998833, | |
| "learning_rate": 2.0750049758091778e-06, | |
| "loss": 0.0214, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.072847682119205, | |
| "grad_norm": 2.592273503755067, | |
| "learning_rate": 2.063393201141846e-06, | |
| "loss": 0.0145, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 7.08167770419426, | |
| "grad_norm": 4.436668977194447, | |
| "learning_rate": 2.0518055546309362e-06, | |
| "loss": 0.0161, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 7.0905077262693155, | |
| "grad_norm": 3.3171318601697295, | |
| "learning_rate": 2.0402421314844774e-06, | |
| "loss": 0.0163, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 7.099337748344371, | |
| "grad_norm": 2.0261212217657527, | |
| "learning_rate": 2.0287030267114665e-06, | |
| "loss": 0.0205, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 7.108167770419426, | |
| "grad_norm": 2.3882323675051187, | |
| "learning_rate": 2.0171883351211038e-06, | |
| "loss": 0.0189, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 7.116997792494481, | |
| "grad_norm": 2.844393454412206, | |
| "learning_rate": 2.0056981513219944e-06, | |
| "loss": 0.0144, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 7.125827814569536, | |
| "grad_norm": 4.923841998471848, | |
| "learning_rate": 1.9942325697213817e-06, | |
| "loss": 0.017, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 7.134657836644592, | |
| "grad_norm": 2.56353141744233, | |
| "learning_rate": 1.9827916845243687e-06, | |
| "loss": 0.018, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 7.143487858719647, | |
| "grad_norm": 2.427889756719139, | |
| "learning_rate": 1.971375589733145e-06, | |
| "loss": 0.0178, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 7.152317880794702, | |
| "grad_norm": 1.8797780460221156, | |
| "learning_rate": 1.9599843791462123e-06, | |
| "loss": 0.0147, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 7.161147902869757, | |
| "grad_norm": 4.2013764200898, | |
| "learning_rate": 1.9486181463576176e-06, | |
| "loss": 0.0174, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 7.169977924944813, | |
| "grad_norm": 4.086210517276765, | |
| "learning_rate": 1.937276984756179e-06, | |
| "loss": 0.0159, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 7.178807947019868, | |
| "grad_norm": 3.2917005232629415, | |
| "learning_rate": 1.925960987524724e-06, | |
| "loss": 0.0214, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 7.187637969094923, | |
| "grad_norm": 2.7476381332120443, | |
| "learning_rate": 1.91467024763932e-06, | |
| "loss": 0.0184, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 7.196467991169978, | |
| "grad_norm": 2.1384411876184313, | |
| "learning_rate": 1.9034048578685099e-06, | |
| "loss": 0.0194, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 7.2052980132450335, | |
| "grad_norm": 3.56984543616482, | |
| "learning_rate": 1.8921649107725525e-06, | |
| "loss": 0.0157, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 7.214128035320089, | |
| "grad_norm": 2.499760035826477, | |
| "learning_rate": 1.880950498702666e-06, | |
| "loss": 0.0186, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 7.222958057395143, | |
| "grad_norm": 3.1744800720072686, | |
| "learning_rate": 1.8697617138002545e-06, | |
| "loss": 0.017, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 7.231788079470198, | |
| "grad_norm": 1.7838431844788865, | |
| "learning_rate": 1.8585986479961653e-06, | |
| "loss": 0.0155, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 7.2406181015452535, | |
| "grad_norm": 1.8862850317579425, | |
| "learning_rate": 1.8474613930099356e-06, | |
| "loss": 0.0125, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 7.249448123620309, | |
| "grad_norm": 3.5402867932239395, | |
| "learning_rate": 1.8363500403490175e-06, | |
| "loss": 0.0166, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 7.258278145695364, | |
| "grad_norm": 2.1688552483141272, | |
| "learning_rate": 1.8252646813080566e-06, | |
| "loss": 0.0153, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 7.267108167770419, | |
| "grad_norm": 1.79219323499248, | |
| "learning_rate": 1.814205406968112e-06, | |
| "loss": 0.0169, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 7.2759381898454745, | |
| "grad_norm": 2.831938125236168, | |
| "learning_rate": 1.8031723081959334e-06, | |
| "loss": 0.0178, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 7.28476821192053, | |
| "grad_norm": 1.722561665077389, | |
| "learning_rate": 1.792165475643199e-06, | |
| "loss": 0.0139, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 7.293598233995585, | |
| "grad_norm": 11.202227492452131, | |
| "learning_rate": 1.7811849997457681e-06, | |
| "loss": 0.0156, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 7.30242825607064, | |
| "grad_norm": 2.212982624248021, | |
| "learning_rate": 1.7702309707229576e-06, | |
| "loss": 0.0167, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 7.311258278145695, | |
| "grad_norm": 3.1007517698985785, | |
| "learning_rate": 1.7593034785767788e-06, | |
| "loss": 0.0226, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 7.320088300220751, | |
| "grad_norm": 2.757612857427265, | |
| "learning_rate": 1.7484026130912097e-06, | |
| "loss": 0.0227, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 7.328918322295806, | |
| "grad_norm": 2.2721148609205515, | |
| "learning_rate": 1.737528463831456e-06, | |
| "loss": 0.0155, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 7.337748344370861, | |
| "grad_norm": 3.99636025707727, | |
| "learning_rate": 1.726681120143207e-06, | |
| "loss": 0.017, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 7.346578366445916, | |
| "grad_norm": 2.2641157595476282, | |
| "learning_rate": 1.7158606711519193e-06, | |
| "loss": 0.0206, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 7.355408388520972, | |
| "grad_norm": 5.094505684513167, | |
| "learning_rate": 1.7050672057620666e-06, | |
| "loss": 0.0198, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 7.364238410596027, | |
| "grad_norm": 2.0556565937653657, | |
| "learning_rate": 1.6943008126564164e-06, | |
| "loss": 0.0143, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 7.373068432671082, | |
| "grad_norm": 3.1355693856781515, | |
| "learning_rate": 1.6835615802953026e-06, | |
| "loss": 0.0215, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 7.381898454746137, | |
| "grad_norm": 2.9513304853848834, | |
| "learning_rate": 1.6728495969158976e-06, | |
| "loss": 0.017, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 7.390728476821192, | |
| "grad_norm": 3.263540253351405, | |
| "learning_rate": 1.6621649505314853e-06, | |
| "loss": 0.018, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 7.399558498896247, | |
| "grad_norm": 5.194990676292726, | |
| "learning_rate": 1.6515077289307391e-06, | |
| "loss": 0.0158, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 7.408388520971302, | |
| "grad_norm": 2.107112405618394, | |
| "learning_rate": 1.640878019677008e-06, | |
| "loss": 0.0158, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 7.417218543046357, | |
| "grad_norm": 16.732017284842502, | |
| "learning_rate": 1.6302759101075788e-06, | |
| "loss": 0.0183, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.4260485651214125, | |
| "grad_norm": 3.0253668035429735, | |
| "learning_rate": 1.619701487332978e-06, | |
| "loss": 0.0184, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 7.434878587196468, | |
| "grad_norm": 2.923106683503437, | |
| "learning_rate": 1.609154838236246e-06, | |
| "loss": 0.0162, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 7.443708609271523, | |
| "grad_norm": 3.0339568808487996, | |
| "learning_rate": 1.5986360494722237e-06, | |
| "loss": 0.012, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 7.452538631346578, | |
| "grad_norm": 2.716505449959593, | |
| "learning_rate": 1.5881452074668474e-06, | |
| "loss": 0.0176, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 7.4613686534216335, | |
| "grad_norm": 2.5093021733900454, | |
| "learning_rate": 1.577682398416424e-06, | |
| "loss": 0.0171, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 7.470198675496689, | |
| "grad_norm": 2.423953596419337, | |
| "learning_rate": 1.567247708286942e-06, | |
| "loss": 0.017, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 7.479028697571744, | |
| "grad_norm": 3.607049773997457, | |
| "learning_rate": 1.5568412228133506e-06, | |
| "loss": 0.0219, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 7.487858719646799, | |
| "grad_norm": 2.739396145699267, | |
| "learning_rate": 1.5464630274988558e-06, | |
| "loss": 0.0166, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 7.496688741721854, | |
| "grad_norm": 3.1460175401960417, | |
| "learning_rate": 1.5361132076142316e-06, | |
| "loss": 0.0178, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 7.50551876379691, | |
| "grad_norm": 3.1265183781548824, | |
| "learning_rate": 1.5257918481971028e-06, | |
| "loss": 0.0195, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.514348785871965, | |
| "grad_norm": 2.1824819207181316, | |
| "learning_rate": 1.515499034051256e-06, | |
| "loss": 0.0139, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 7.52317880794702, | |
| "grad_norm": 3.1505051795329937, | |
| "learning_rate": 1.50523484974594e-06, | |
| "loss": 0.0154, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 7.532008830022075, | |
| "grad_norm": 4.254477530458638, | |
| "learning_rate": 1.4949993796151675e-06, | |
| "loss": 0.0133, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 7.540838852097131, | |
| "grad_norm": 2.3198932523987152, | |
| "learning_rate": 1.4847927077570324e-06, | |
| "loss": 0.0153, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.549668874172186, | |
| "grad_norm": 3.5603554442961496, | |
| "learning_rate": 1.4746149180330082e-06, | |
| "loss": 0.0164, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 7.55849889624724, | |
| "grad_norm": 2.3376069646760387, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.0159, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 7.567328918322296, | |
| "grad_norm": 2.368387160146722, | |
| "learning_rate": 1.4543463192459728e-06, | |
| "loss": 0.0137, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 7.576158940397351, | |
| "grad_norm": 2.5418824834778815, | |
| "learning_rate": 1.4442556767166371e-06, | |
| "loss": 0.0161, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 7.584988962472406, | |
| "grad_norm": 2.143348390657134, | |
| "learning_rate": 1.4341942493873934e-06, | |
| "loss": 0.0156, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 7.593818984547461, | |
| "grad_norm": 1.9257000569002491, | |
| "learning_rate": 1.4241621199263362e-06, | |
| "loss": 0.0132, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 7.602649006622516, | |
| "grad_norm": 2.466043045760216, | |
| "learning_rate": 1.4141593707608441e-06, | |
| "loss": 0.0185, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 7.6114790286975715, | |
| "grad_norm": 1.8695926443533988, | |
| "learning_rate": 1.40418608407689e-06, | |
| "loss": 0.014, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 7.620309050772627, | |
| "grad_norm": 3.9111267128869898, | |
| "learning_rate": 1.3942423418183764e-06, | |
| "loss": 0.0165, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 7.629139072847682, | |
| "grad_norm": 2.4922570381059, | |
| "learning_rate": 1.3843282256864599e-06, | |
| "loss": 0.0187, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 7.637969094922737, | |
| "grad_norm": 5.217690756083938, | |
| "learning_rate": 1.3744438171388752e-06, | |
| "loss": 0.0155, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 7.6467991169977925, | |
| "grad_norm": 1.8101355352608983, | |
| "learning_rate": 1.3645891973892772e-06, | |
| "loss": 0.0139, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 7.655629139072848, | |
| "grad_norm": 5.545891187556725, | |
| "learning_rate": 1.3547644474065557e-06, | |
| "loss": 0.0129, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 7.664459161147903, | |
| "grad_norm": 4.225172676007295, | |
| "learning_rate": 1.3449696479141855e-06, | |
| "loss": 0.0192, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 7.673289183222958, | |
| "grad_norm": 4.520628600478915, | |
| "learning_rate": 1.3352048793895623e-06, | |
| "loss": 0.0156, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 7.682119205298013, | |
| "grad_norm": 3.944492738552924, | |
| "learning_rate": 1.325470222063327e-06, | |
| "loss": 0.0217, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 7.690949227373069, | |
| "grad_norm": 3.0246291720472382, | |
| "learning_rate": 1.3157657559187264e-06, | |
| "loss": 0.0145, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 7.699779249448124, | |
| "grad_norm": 2.8899719439535994, | |
| "learning_rate": 1.3060915606909413e-06, | |
| "loss": 0.0161, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 7.708609271523179, | |
| "grad_norm": 3.027013167154667, | |
| "learning_rate": 1.2964477158664367e-06, | |
| "loss": 0.0159, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 7.717439293598234, | |
| "grad_norm": 3.1400303793707534, | |
| "learning_rate": 1.2868343006823113e-06, | |
| "loss": 0.0177, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 7.72626931567329, | |
| "grad_norm": 2.7296818780684853, | |
| "learning_rate": 1.2772513941256371e-06, | |
| "loss": 0.0144, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 7.735099337748345, | |
| "grad_norm": 2.443109174791123, | |
| "learning_rate": 1.2676990749328255e-06, | |
| "loss": 0.0146, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 7.743929359823399, | |
| "grad_norm": 3.8645559979710606, | |
| "learning_rate": 1.2581774215889653e-06, | |
| "loss": 0.0154, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 7.752759381898454, | |
| "grad_norm": 4.316002412063608, | |
| "learning_rate": 1.2486865123271868e-06, | |
| "loss": 0.0143, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 7.76158940397351, | |
| "grad_norm": 2.854224646717579, | |
| "learning_rate": 1.2392264251280167e-06, | |
| "loss": 0.0156, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 7.770419426048565, | |
| "grad_norm": 4.151908328285103, | |
| "learning_rate": 1.2297972377187361e-06, | |
| "loss": 0.0167, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 7.77924944812362, | |
| "grad_norm": 2.1779776586316193, | |
| "learning_rate": 1.2203990275727435e-06, | |
| "loss": 0.0171, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 7.788079470198675, | |
| "grad_norm": 2.530292045085328, | |
| "learning_rate": 1.211031871908916e-06, | |
| "loss": 0.0155, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 7.7969094922737305, | |
| "grad_norm": 2.554154765902391, | |
| "learning_rate": 1.201695847690983e-06, | |
| "loss": 0.0163, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 7.805739514348786, | |
| "grad_norm": 3.2873636940890756, | |
| "learning_rate": 1.1923910316268783e-06, | |
| "loss": 0.0167, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 7.814569536423841, | |
| "grad_norm": 4.718578485487324, | |
| "learning_rate": 1.183117500168125e-06, | |
| "loss": 0.0185, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 7.823399558498896, | |
| "grad_norm": 2.852464714262919, | |
| "learning_rate": 1.1738753295091986e-06, | |
| "loss": 0.016, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 7.8322295805739515, | |
| "grad_norm": 2.352581618839849, | |
| "learning_rate": 1.164664595586904e-06, | |
| "loss": 0.0138, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 7.841059602649007, | |
| "grad_norm": 2.7982772682644232, | |
| "learning_rate": 1.1554853740797556e-06, | |
| "loss": 0.0168, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 7.849889624724062, | |
| "grad_norm": 3.585955553449094, | |
| "learning_rate": 1.1463377404073433e-06, | |
| "loss": 0.0172, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 7.858719646799117, | |
| "grad_norm": 2.7082009380924514, | |
| "learning_rate": 1.137221769729725e-06, | |
| "loss": 0.0156, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.867549668874172, | |
| "grad_norm": 2.5240320142035344, | |
| "learning_rate": 1.1281375369468078e-06, | |
| "loss": 0.0127, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 7.876379690949228, | |
| "grad_norm": 3.235127121053359, | |
| "learning_rate": 1.1190851166977218e-06, | |
| "loss": 0.0159, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 7.885209713024283, | |
| "grad_norm": 3.0343922974864452, | |
| "learning_rate": 1.1100645833602231e-06, | |
| "loss": 0.0122, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 7.894039735099338, | |
| "grad_norm": 2.581977301531322, | |
| "learning_rate": 1.1010760110500652e-06, | |
| "loss": 0.0179, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 7.902869757174393, | |
| "grad_norm": 7.38800345715505, | |
| "learning_rate": 1.0921194736204066e-06, | |
| "loss": 0.0165, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 7.911699779249448, | |
| "grad_norm": 2.962956477533454, | |
| "learning_rate": 1.083195044661195e-06, | |
| "loss": 0.0139, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 7.920529801324504, | |
| "grad_norm": 2.3479876198259193, | |
| "learning_rate": 1.0743027974985576e-06, | |
| "loss": 0.015, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 7.929359823399558, | |
| "grad_norm": 2.678489593793483, | |
| "learning_rate": 1.065442805194214e-06, | |
| "loss": 0.0163, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 7.938189845474613, | |
| "grad_norm": 2.1090448936516926, | |
| "learning_rate": 1.056615140544861e-06, | |
| "loss": 0.0166, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 7.947019867549669, | |
| "grad_norm": 14.335019974415568, | |
| "learning_rate": 1.0478198760815833e-06, | |
| "loss": 0.0183, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.955849889624724, | |
| "grad_norm": 2.754108250727163, | |
| "learning_rate": 1.0390570840692527e-06, | |
| "loss": 0.0184, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 7.964679911699779, | |
| "grad_norm": 2.047834899241759, | |
| "learning_rate": 1.0303268365059383e-06, | |
| "loss": 0.0134, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 7.973509933774834, | |
| "grad_norm": 2.908581712678836, | |
| "learning_rate": 1.021629205122311e-06, | |
| "loss": 0.0178, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 7.9823399558498895, | |
| "grad_norm": 2.7092806374488507, | |
| "learning_rate": 1.0129642613810576e-06, | |
| "loss": 0.0135, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 7.991169977924945, | |
| "grad_norm": 8.427340467929495, | |
| "learning_rate": 1.0043320764762915e-06, | |
| "loss": 0.0141, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.8826244906591318, | |
| "learning_rate": 9.957327213329687e-07, | |
| "loss": 0.0149, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 8.008830022075054, | |
| "grad_norm": 0.9216642527760257, | |
| "learning_rate": 9.871662666063054e-07, | |
| "loss": 0.0049, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 8.01766004415011, | |
| "grad_norm": 1.586809078537705, | |
| "learning_rate": 9.786327826811942e-07, | |
| "loss": 0.0062, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 8.026490066225165, | |
| "grad_norm": 1.5105791113977531, | |
| "learning_rate": 9.701323396716312e-07, | |
| "loss": 0.0058, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 8.035320088300221, | |
| "grad_norm": 1.3122806248350998, | |
| "learning_rate": 9.616650074201383e-07, | |
| "loss": 0.0046, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 8.044150110375275, | |
| "grad_norm": 0.902563437247834, | |
| "learning_rate": 9.532308554971831e-07, | |
| "loss": 0.0044, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 8.052980132450331, | |
| "grad_norm": 2.4957384961901106, | |
| "learning_rate": 9.44829953200615e-07, | |
| "loss": 0.0047, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 8.061810154525386, | |
| "grad_norm": 2.125245251707967, | |
| "learning_rate": 9.364623695550979e-07, | |
| "loss": 0.006, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 8.070640176600442, | |
| "grad_norm": 2.4620509605661662, | |
| "learning_rate": 9.281281733115288e-07, | |
| "loss": 0.006, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 8.079470198675496, | |
| "grad_norm": 1.8720057057125778, | |
| "learning_rate": 9.198274329464929e-07, | |
| "loss": 0.0046, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 8.088300220750552, | |
| "grad_norm": 0.6970482985245559, | |
| "learning_rate": 9.115602166616805e-07, | |
| "loss": 0.0042, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 8.097130242825607, | |
| "grad_norm": 1.34177049630106, | |
| "learning_rate": 9.033265923833446e-07, | |
| "loss": 0.0047, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 8.105960264900663, | |
| "grad_norm": 1.2384572459282441, | |
| "learning_rate": 8.951266277617326e-07, | |
| "loss": 0.0052, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 8.114790286975717, | |
| "grad_norm": 1.1841347644308235, | |
| "learning_rate": 8.869603901705287e-07, | |
| "loss": 0.0036, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 8.123620309050773, | |
| "grad_norm": 1.0865504633097303, | |
| "learning_rate": 8.78827946706311e-07, | |
| "loss": 0.0046, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 8.132450331125828, | |
| "grad_norm": 1.3834817601097062, | |
| "learning_rate": 8.707293641879888e-07, | |
| "loss": 0.0053, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 8.141280353200884, | |
| "grad_norm": 1.6947959780214412, | |
| "learning_rate": 8.626647091562612e-07, | |
| "loss": 0.0048, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 8.150110375275938, | |
| "grad_norm": 1.4963551013193799, | |
| "learning_rate": 8.546340478730647e-07, | |
| "loss": 0.0037, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 8.158940397350994, | |
| "grad_norm": 5.288959785053701, | |
| "learning_rate": 8.466374463210348e-07, | |
| "loss": 0.0061, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 8.167770419426049, | |
| "grad_norm": 2.109180394557091, | |
| "learning_rate": 8.386749702029578e-07, | |
| "loss": 0.0055, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 8.176600441501105, | |
| "grad_norm": 1.2223125757703008, | |
| "learning_rate": 8.307466849412365e-07, | |
| "loss": 0.0046, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 8.185430463576159, | |
| "grad_norm": 1.3187516991254513, | |
| "learning_rate": 8.228526556773486e-07, | |
| "loss": 0.0048, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 8.194260485651213, | |
| "grad_norm": 1.2630767224054738, | |
| "learning_rate": 8.149929472713126e-07, | |
| "loss": 0.0036, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 8.20309050772627, | |
| "grad_norm": 2.0643197245119014, | |
| "learning_rate": 8.071676243011556e-07, | |
| "loss": 0.0067, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 8.211920529801324, | |
| "grad_norm": 1.1582456033233397, | |
| "learning_rate": 7.993767510623834e-07, | |
| "loss": 0.0038, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 8.22075055187638, | |
| "grad_norm": 1.114563153977826, | |
| "learning_rate": 7.91620391567448e-07, | |
| "loss": 0.0046, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 8.229580573951434, | |
| "grad_norm": 1.1948683848879686, | |
| "learning_rate": 7.838986095452311e-07, | |
| "loss": 0.0055, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 8.23841059602649, | |
| "grad_norm": 2.0219350248789567, | |
| "learning_rate": 7.762114684405064e-07, | |
| "loss": 0.0069, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 8.247240618101545, | |
| "grad_norm": 2.9139477661374227, | |
| "learning_rate": 7.685590314134294e-07, | |
| "loss": 0.0065, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 8.2560706401766, | |
| "grad_norm": 1.293021061880487, | |
| "learning_rate": 7.609413613390199e-07, | |
| "loss": 0.0045, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 8.264900662251655, | |
| "grad_norm": 2.953715103210279, | |
| "learning_rate": 7.533585208066302e-07, | |
| "loss": 0.0073, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 8.273730684326711, | |
| "grad_norm": 1.2460632394899884, | |
| "learning_rate": 7.458105721194525e-07, | |
| "loss": 0.0047, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 8.282560706401766, | |
| "grad_norm": 1.284980054716128, | |
| "learning_rate": 7.382975772939866e-07, | |
| "loss": 0.0051, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 8.291390728476822, | |
| "grad_norm": 2.3736622260844014, | |
| "learning_rate": 7.308195980595462e-07, | |
| "loss": 0.0047, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 8.300220750551876, | |
| "grad_norm": 1.2219382747072445, | |
| "learning_rate": 7.233766958577421e-07, | |
| "loss": 0.0036, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 8.309050772626932, | |
| "grad_norm": 1.1091857226178634, | |
| "learning_rate": 7.159689318419777e-07, | |
| "loss": 0.0031, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 8.317880794701987, | |
| "grad_norm": 1.3077010085263248, | |
| "learning_rate": 7.085963668769552e-07, | |
| "loss": 0.0052, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 8.326710816777043, | |
| "grad_norm": 0.8203794075909709, | |
| "learning_rate": 7.012590615381654e-07, | |
| "loss": 0.0025, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 8.335540838852097, | |
| "grad_norm": 1.4995044409627258, | |
| "learning_rate": 6.939570761113939e-07, | |
| "loss": 0.0035, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 8.344370860927153, | |
| "grad_norm": 1.562116658932747, | |
| "learning_rate": 6.866904705922284e-07, | |
| "loss": 0.0043, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 8.353200883002208, | |
| "grad_norm": 0.9167520304185413, | |
| "learning_rate": 6.794593046855613e-07, | |
| "loss": 0.0035, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 8.362030905077262, | |
| "grad_norm": 4.326122941504541, | |
| "learning_rate": 6.722636378051011e-07, | |
| "loss": 0.0053, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 8.370860927152318, | |
| "grad_norm": 1.5494926100315216, | |
| "learning_rate": 6.651035290728858e-07, | |
| "loss": 0.0053, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 8.379690949227372, | |
| "grad_norm": 1.6209750572966557, | |
| "learning_rate": 6.579790373187944e-07, | |
| "loss": 0.0053, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 8.388520971302428, | |
| "grad_norm": 1.2500030153858688, | |
| "learning_rate": 6.508902210800649e-07, | |
| "loss": 0.0047, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 8.397350993377483, | |
| "grad_norm": 1.2236201766489472, | |
| "learning_rate": 6.43837138600813e-07, | |
| "loss": 0.0033, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 8.406181015452539, | |
| "grad_norm": 1.3122107248792492, | |
| "learning_rate": 6.36819847831554e-07, | |
| "loss": 0.0034, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 8.415011037527593, | |
| "grad_norm": 1.8016103648711728, | |
| "learning_rate": 6.298384064287261e-07, | |
| "loss": 0.0053, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 8.42384105960265, | |
| "grad_norm": 0.7804923933562897, | |
| "learning_rate": 6.228928717542205e-07, | |
| "loss": 0.0027, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 8.432671081677704, | |
| "grad_norm": 1.077294274779956, | |
| "learning_rate": 6.159833008748988e-07, | |
| "loss": 0.0033, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 8.44150110375276, | |
| "grad_norm": 1.4893790375135336, | |
| "learning_rate": 6.091097505621374e-07, | |
| "loss": 0.0053, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 8.450331125827814, | |
| "grad_norm": 1.0682095206167823, | |
| "learning_rate": 6.022722772913581e-07, | |
| "loss": 0.0038, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 8.45916114790287, | |
| "grad_norm": 1.1276769369142745, | |
| "learning_rate": 5.954709372415524e-07, | |
| "loss": 0.0057, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 8.467991169977925, | |
| "grad_norm": 2.0609484200927066, | |
| "learning_rate": 5.887057862948403e-07, | |
| "loss": 0.006, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 8.47682119205298, | |
| "grad_norm": 1.7276986171727993, | |
| "learning_rate": 5.819768800359882e-07, | |
| "loss": 0.0048, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.485651214128035, | |
| "grad_norm": 2.955819366787621, | |
| "learning_rate": 5.752842737519743e-07, | |
| "loss": 0.0074, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 8.494481236203091, | |
| "grad_norm": 3.0522570867077246, | |
| "learning_rate": 5.686280224315189e-07, | |
| "loss": 0.0046, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 8.503311258278146, | |
| "grad_norm": 2.306994902901723, | |
| "learning_rate": 5.62008180764635e-07, | |
| "loss": 0.0066, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 8.512141280353202, | |
| "grad_norm": 1.076159080481545, | |
| "learning_rate": 5.554248031421872e-07, | |
| "loss": 0.0036, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 8.520971302428256, | |
| "grad_norm": 1.1516433631751446, | |
| "learning_rate": 5.488779436554359e-07, | |
| "loss": 0.0035, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 8.52980132450331, | |
| "grad_norm": 1.927000618878644, | |
| "learning_rate": 5.423676560955976e-07, | |
| "loss": 0.0034, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 8.538631346578367, | |
| "grad_norm": 1.8645495423929004, | |
| "learning_rate": 5.358939939534002e-07, | |
| "loss": 0.0073, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 8.54746136865342, | |
| "grad_norm": 2.0548525160077915, | |
| "learning_rate": 5.294570104186436e-07, | |
| "loss": 0.0043, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 8.556291390728477, | |
| "grad_norm": 1.856532869436402, | |
| "learning_rate": 5.230567583797674e-07, | |
| "loss": 0.0058, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 8.565121412803531, | |
| "grad_norm": 0.9776362465838028, | |
| "learning_rate": 5.166932904234101e-07, | |
| "loss": 0.0024, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 8.573951434878587, | |
| "grad_norm": 1.642426484620163, | |
| "learning_rate": 5.103666588339812e-07, | |
| "loss": 0.0042, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 8.582781456953642, | |
| "grad_norm": 1.9824756275567288, | |
| "learning_rate": 5.040769155932285e-07, | |
| "loss": 0.0052, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 8.591611479028698, | |
| "grad_norm": 3.4366301957917353, | |
| "learning_rate": 4.978241123798133e-07, | |
| "loss": 0.0034, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 8.600441501103752, | |
| "grad_norm": 1.9948600484771069, | |
| "learning_rate": 4.916083005688865e-07, | |
| "loss": 0.0042, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 8.609271523178808, | |
| "grad_norm": 2.15430359156571, | |
| "learning_rate": 4.85429531231662e-07, | |
| "loss": 0.0037, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 8.618101545253863, | |
| "grad_norm": 1.1766278739240257, | |
| "learning_rate": 4.792878551350055e-07, | |
| "loss": 0.003, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.626931567328919, | |
| "grad_norm": 3.7154516077034163, | |
| "learning_rate": 4.7318332274100595e-07, | |
| "loss": 0.0053, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 8.635761589403973, | |
| "grad_norm": 1.6448343441626732, | |
| "learning_rate": 4.671159842065698e-07, | |
| "loss": 0.0055, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 8.64459161147903, | |
| "grad_norm": 1.3601236016666638, | |
| "learning_rate": 4.6108588938300725e-07, | |
| "loss": 0.0066, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 8.653421633554084, | |
| "grad_norm": 2.2958751161079345, | |
| "learning_rate": 4.5509308781561846e-07, | |
| "loss": 0.0033, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.66225165562914, | |
| "grad_norm": 1.388373437779568, | |
| "learning_rate": 4.4913762874329527e-07, | |
| "loss": 0.0053, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 8.671081677704194, | |
| "grad_norm": 1.8958198094426935, | |
| "learning_rate": 4.4321956109810327e-07, | |
| "loss": 0.008, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 8.67991169977925, | |
| "grad_norm": 1.616244863701459, | |
| "learning_rate": 4.3733893350489386e-07, | |
| "loss": 0.0058, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 8.688741721854305, | |
| "grad_norm": 6.2773693867354625, | |
| "learning_rate": 4.314957942808956e-07, | |
| "loss": 0.0058, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 8.697571743929359, | |
| "grad_norm": 2.1813377197727943, | |
| "learning_rate": 4.2569019143531845e-07, | |
| "loss": 0.0065, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 8.706401766004415, | |
| "grad_norm": 2.16459760537303, | |
| "learning_rate": 4.199221726689634e-07, | |
| "loss": 0.0066, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 8.71523178807947, | |
| "grad_norm": 2.329687588049389, | |
| "learning_rate": 4.1419178537382756e-07, | |
| "loss": 0.0052, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 8.724061810154526, | |
| "grad_norm": 1.4744251427566957, | |
| "learning_rate": 4.0849907663271346e-07, | |
| "loss": 0.0043, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 8.73289183222958, | |
| "grad_norm": 2.059534878483957, | |
| "learning_rate": 4.028440932188465e-07, | |
| "loss": 0.0049, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 8.741721854304636, | |
| "grad_norm": 0.8689790885213784, | |
| "learning_rate": 3.972268815954833e-07, | |
| "loss": 0.003, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 8.75055187637969, | |
| "grad_norm": 1.4910172150909442, | |
| "learning_rate": 3.916474879155402e-07, | |
| "loss": 0.0045, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 8.759381898454746, | |
| "grad_norm": 1.2298349573738518, | |
| "learning_rate": 3.8610595802120564e-07, | |
| "loss": 0.0037, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 8.7682119205298, | |
| "grad_norm": 0.9265755140542287, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.0028, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 8.777041942604857, | |
| "grad_norm": 0.9373215469637189, | |
| "learning_rate": 3.751366714022342e-07, | |
| "loss": 0.0053, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 8.785871964679911, | |
| "grad_norm": 0.7661590237021436, | |
| "learning_rate": 3.6970900480497287e-07, | |
| "loss": 0.0031, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 8.794701986754967, | |
| "grad_norm": 1.5711093730790697, | |
| "learning_rate": 3.643193822473301e-07, | |
| "loss": 0.005, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 8.803532008830022, | |
| "grad_norm": 1.4371456444021427, | |
| "learning_rate": 3.5896784801227046e-07, | |
| "loss": 0.0043, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 8.812362030905078, | |
| "grad_norm": 4.519438937897612, | |
| "learning_rate": 3.5365444606981434e-07, | |
| "loss": 0.0056, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 8.821192052980132, | |
| "grad_norm": 1.4009009497014246, | |
| "learning_rate": 3.4837922007667e-07, | |
| "loss": 0.0043, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 8.830022075055188, | |
| "grad_norm": 1.3255555079300094, | |
| "learning_rate": 3.4314221337588217e-07, | |
| "loss": 0.0043, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.838852097130243, | |
| "grad_norm": 0.9979698124388939, | |
| "learning_rate": 3.379434689964728e-07, | |
| "loss": 0.0044, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 8.847682119205299, | |
| "grad_norm": 1.5779568428376927, | |
| "learning_rate": 3.3278302965308593e-07, | |
| "loss": 0.0039, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 8.856512141280353, | |
| "grad_norm": 1.004612384300377, | |
| "learning_rate": 3.276609377456419e-07, | |
| "loss": 0.004, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 8.865342163355407, | |
| "grad_norm": 0.9906129424262534, | |
| "learning_rate": 3.2257723535898177e-07, | |
| "loss": 0.004, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 8.874172185430464, | |
| "grad_norm": 2.200082020610832, | |
| "learning_rate": 3.1753196426252573e-07, | |
| "loss": 0.0041, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 8.883002207505518, | |
| "grad_norm": 2.2396795819915565, | |
| "learning_rate": 3.125251659099332e-07, | |
| "loss": 0.0056, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 8.891832229580574, | |
| "grad_norm": 2.5986716560055916, | |
| "learning_rate": 3.0755688143875253e-07, | |
| "loss": 0.004, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 8.900662251655628, | |
| "grad_norm": 1.702540665694322, | |
| "learning_rate": 3.026271516700946e-07, | |
| "loss": 0.0032, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 8.909492273730685, | |
| "grad_norm": 1.9964934466245583, | |
| "learning_rate": 2.9773601710828937e-07, | |
| "loss": 0.0051, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 8.918322295805739, | |
| "grad_norm": 1.4111955374392817, | |
| "learning_rate": 2.928835179405548e-07, | |
| "loss": 0.0038, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.927152317880795, | |
| "grad_norm": 1.3022027357530646, | |
| "learning_rate": 2.8806969403666897e-07, | |
| "loss": 0.0057, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 8.93598233995585, | |
| "grad_norm": 1.554274938533939, | |
| "learning_rate": 2.8329458494863846e-07, | |
| "loss": 0.0058, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 8.944812362030905, | |
| "grad_norm": 1.5772165971517251, | |
| "learning_rate": 2.7855822991037895e-07, | |
| "loss": 0.0062, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 8.95364238410596, | |
| "grad_norm": 1.551087200212398, | |
| "learning_rate": 2.738606678373873e-07, | |
| "loss": 0.0055, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 8.962472406181016, | |
| "grad_norm": 1.8406416998541573, | |
| "learning_rate": 2.6920193732642594e-07, | |
| "loss": 0.0057, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 8.97130242825607, | |
| "grad_norm": 6.7829324366727635, | |
| "learning_rate": 2.6458207665520266e-07, | |
| "loss": 0.0061, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 8.980132450331126, | |
| "grad_norm": 3.158192243428083, | |
| "learning_rate": 2.600011237820577e-07, | |
| "loss": 0.0044, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 8.98896247240618, | |
| "grad_norm": 1.7985070168148436, | |
| "learning_rate": 2.5545911634565266e-07, | |
| "loss": 0.0046, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 8.997792494481237, | |
| "grad_norm": 1.5683538645092074, | |
| "learning_rate": 2.5095609166465805e-07, | |
| "loss": 0.0049, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 9.006622516556291, | |
| "grad_norm": 2.5471409476536846, | |
| "learning_rate": 2.4649208673745317e-07, | |
| "loss": 0.0028, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 9.015452538631347, | |
| "grad_norm": 0.7700158467560422, | |
| "learning_rate": 2.420671382418122e-07, | |
| "loss": 0.0017, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 9.024282560706402, | |
| "grad_norm": 1.2067140795630868, | |
| "learning_rate": 2.3768128253461253e-07, | |
| "loss": 0.003, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 9.033112582781458, | |
| "grad_norm": 1.1242161264147874, | |
| "learning_rate": 2.333345556515304e-07, | |
| "loss": 0.0021, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 9.041942604856512, | |
| "grad_norm": 0.45696992448989654, | |
| "learning_rate": 2.290269933067457e-07, | |
| "loss": 0.0022, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 9.050772626931566, | |
| "grad_norm": 0.3783063541872817, | |
| "learning_rate": 2.2475863089265193e-07, | |
| "loss": 0.0013, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 9.059602649006623, | |
| "grad_norm": 1.9181847282581403, | |
| "learning_rate": 2.205295034795596e-07, | |
| "loss": 0.0019, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 9.068432671081677, | |
| "grad_norm": 1.1869922883999164, | |
| "learning_rate": 2.1633964581541212e-07, | |
| "loss": 0.0024, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 9.077262693156733, | |
| "grad_norm": 0.46941428356892356, | |
| "learning_rate": 2.1218909232550156e-07, | |
| "loss": 0.0015, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 9.086092715231787, | |
| "grad_norm": 0.806693065905521, | |
| "learning_rate": 2.0807787711217887e-07, | |
| "loss": 0.0017, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 9.094922737306844, | |
| "grad_norm": 1.055725069444773, | |
| "learning_rate": 2.0400603395458408e-07, | |
| "loss": 0.0015, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 9.103752759381898, | |
| "grad_norm": 0.4333000949149235, | |
| "learning_rate": 1.999735963083571e-07, | |
| "loss": 0.0018, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 9.112582781456954, | |
| "grad_norm": 0.38891892510898857, | |
| "learning_rate": 1.9598059730537465e-07, | |
| "loss": 0.0013, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 9.121412803532008, | |
| "grad_norm": 0.9673287003029005, | |
| "learning_rate": 1.9202706975346875e-07, | |
| "loss": 0.001, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 9.130242825607064, | |
| "grad_norm": 0.5073316613315646, | |
| "learning_rate": 1.881130461361591e-07, | |
| "loss": 0.0011, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 9.139072847682119, | |
| "grad_norm": 0.7395240265584485, | |
| "learning_rate": 1.8423855861239238e-07, | |
| "loss": 0.0016, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 9.147902869757175, | |
| "grad_norm": 0.8474114415798074, | |
| "learning_rate": 1.8040363901627001e-07, | |
| "loss": 0.0028, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 9.15673289183223, | |
| "grad_norm": 1.4593710303340477, | |
| "learning_rate": 1.7660831885679074e-07, | |
| "loss": 0.002, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 9.165562913907285, | |
| "grad_norm": 0.3376918976776975, | |
| "learning_rate": 1.7285262931759084e-07, | |
| "loss": 0.001, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 9.17439293598234, | |
| "grad_norm": 1.0290815165500646, | |
| "learning_rate": 1.6913660125668806e-07, | |
| "loss": 0.0026, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 9.183222958057396, | |
| "grad_norm": 0.581331466133525, | |
| "learning_rate": 1.654602652062276e-07, | |
| "loss": 0.0013, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 9.19205298013245, | |
| "grad_norm": 0.43451930064326927, | |
| "learning_rate": 1.6182365137223266e-07, | |
| "loss": 0.0014, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 9.200883002207506, | |
| "grad_norm": 0.2724268617015149, | |
| "learning_rate": 1.5822678963435479e-07, | |
| "loss": 0.0011, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 9.20971302428256, | |
| "grad_norm": 0.41211312128483507, | |
| "learning_rate": 1.5466970954562786e-07, | |
| "loss": 0.0021, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 9.218543046357617, | |
| "grad_norm": 1.1138391698665862, | |
| "learning_rate": 1.5115244033222732e-07, | |
| "loss": 0.007, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 9.227373068432671, | |
| "grad_norm": 3.911068240086975, | |
| "learning_rate": 1.47675010893229e-07, | |
| "loss": 0.001, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 9.236203090507725, | |
| "grad_norm": 0.7214880501282108, | |
| "learning_rate": 1.4423744980037068e-07, | |
| "loss": 0.0013, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 9.245033112582782, | |
| "grad_norm": 1.1740806551343546, | |
| "learning_rate": 1.408397852978205e-07, | |
| "loss": 0.0016, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 9.253863134657836, | |
| "grad_norm": 0.49395759998971256, | |
| "learning_rate": 1.3748204530193987e-07, | |
| "loss": 0.0012, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 9.262693156732892, | |
| "grad_norm": 0.7000985887232735, | |
| "learning_rate": 1.341642574010582e-07, | |
| "loss": 0.0017, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 9.271523178807946, | |
| "grad_norm": 0.6624183152454347, | |
| "learning_rate": 1.3088644885524637e-07, | |
| "loss": 0.0018, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 9.280353200883003, | |
| "grad_norm": 0.39977781777076876, | |
| "learning_rate": 1.27648646596088e-07, | |
| "loss": 0.0011, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 9.289183222958057, | |
| "grad_norm": 0.6945355488069699, | |
| "learning_rate": 1.2445087722646576e-07, | |
| "loss": 0.0019, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 9.298013245033113, | |
| "grad_norm": 0.4948417931941682, | |
| "learning_rate": 1.212931670203338e-07, | |
| "loss": 0.0016, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 9.306843267108167, | |
| "grad_norm": 0.49462025635574236, | |
| "learning_rate": 1.1817554192251002e-07, | |
| "loss": 0.0017, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 9.315673289183223, | |
| "grad_norm": 6.856576924850025, | |
| "learning_rate": 1.1509802754845978e-07, | |
| "loss": 0.0017, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 9.324503311258278, | |
| "grad_norm": 0.48773722744048326, | |
| "learning_rate": 1.1206064918408143e-07, | |
| "loss": 0.0013, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 0.5334956783318224, | |
| "learning_rate": 1.0906343178550715e-07, | |
| "loss": 0.0018, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 9.342163355408388, | |
| "grad_norm": 0.5336988621705372, | |
| "learning_rate": 1.0610639997888917e-07, | |
| "loss": 0.0016, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 9.350993377483444, | |
| "grad_norm": 0.4004075989274986, | |
| "learning_rate": 1.0318957806020269e-07, | |
| "loss": 0.0013, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 9.359823399558499, | |
| "grad_norm": 0.5673105181565646, | |
| "learning_rate": 1.0031298999504557e-07, | |
| "loss": 0.0013, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 9.368653421633555, | |
| "grad_norm": 0.6608172119810469, | |
| "learning_rate": 9.747665941843953e-08, | |
| "loss": 0.0021, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 9.37748344370861, | |
| "grad_norm": 0.41089386065043493, | |
| "learning_rate": 9.468060963463754e-08, | |
| "loss": 0.0017, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 9.386313465783665, | |
| "grad_norm": 0.9707171476786928, | |
| "learning_rate": 9.192486361693175e-08, | |
| "loss": 0.0019, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 9.39514348785872, | |
| "grad_norm": 0.5928364920966972, | |
| "learning_rate": 8.920944400746589e-08, | |
| "loss": 0.0012, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 9.403973509933774, | |
| "grad_norm": 0.5008273553812124, | |
| "learning_rate": 8.653437311704648e-08, | |
| "loss": 0.0019, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 9.41280353200883, | |
| "grad_norm": 0.24528745545983557, | |
| "learning_rate": 8.38996729249636e-08, | |
| "loss": 0.0013, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 9.421633554083884, | |
| "grad_norm": 0.38243824324579545, | |
| "learning_rate": 8.130536507880538e-08, | |
| "loss": 0.0017, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 9.43046357615894, | |
| "grad_norm": 0.438811270050038, | |
| "learning_rate": 7.875147089428436e-08, | |
| "loss": 0.0009, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 9.439293598233995, | |
| "grad_norm": 0.4250111638421303, | |
| "learning_rate": 7.623801135506148e-08, | |
| "loss": 0.001, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 9.448123620309051, | |
| "grad_norm": 0.3826980836890277, | |
| "learning_rate": 7.376500711257062e-08, | |
| "loss": 0.0014, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 9.456953642384105, | |
| "grad_norm": 0.8476036859255105, | |
| "learning_rate": 7.133247848585268e-08, | |
| "loss": 0.0014, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 9.465783664459162, | |
| "grad_norm": 0.6322766189316223, | |
| "learning_rate": 6.894044546138845e-08, | |
| "loss": 0.0016, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 9.474613686534216, | |
| "grad_norm": 1.6438462370271978, | |
| "learning_rate": 6.65889276929299e-08, | |
| "loss": 0.0018, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 9.483443708609272, | |
| "grad_norm": 0.5032114653583847, | |
| "learning_rate": 6.427794450134529e-08, | |
| "loss": 0.0012, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 9.492273730684326, | |
| "grad_norm": 0.45357447273965656, | |
| "learning_rate": 6.200751487445367e-08, | |
| "loss": 0.0016, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 9.501103752759382, | |
| "grad_norm": 0.35914176465983044, | |
| "learning_rate": 5.977765746687569e-08, | |
| "loss": 0.001, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 9.509933774834437, | |
| "grad_norm": 0.46757119926456336, | |
| "learning_rate": 5.758839059987531e-08, | |
| "loss": 0.0015, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 9.518763796909493, | |
| "grad_norm": 0.23766439614586482, | |
| "learning_rate": 5.5439732261209356e-08, | |
| "loss": 0.0008, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 9.527593818984547, | |
| "grad_norm": 0.46482063769914506, | |
| "learning_rate": 5.333170010498434e-08, | |
| "loss": 0.0011, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 9.536423841059603, | |
| "grad_norm": 0.4384873123644464, | |
| "learning_rate": 5.126431145150546e-08, | |
| "loss": 0.0012, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.545253863134658, | |
| "grad_norm": 0.27307457325375584, | |
| "learning_rate": 4.9237583287139454e-08, | |
| "loss": 0.0011, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 9.554083885209714, | |
| "grad_norm": 0.5135990338049987, | |
| "learning_rate": 4.72515322641709e-08, | |
| "loss": 0.0011, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 9.562913907284768, | |
| "grad_norm": 0.2655670207808881, | |
| "learning_rate": 4.530617470066834e-08, | |
| "loss": 0.0013, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 9.571743929359823, | |
| "grad_norm": 0.3279672339662895, | |
| "learning_rate": 4.340152658034835e-08, | |
| "loss": 0.0012, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 9.580573951434879, | |
| "grad_norm": 0.5852889740425053, | |
| "learning_rate": 4.153760355244507e-08, | |
| "loss": 0.0012, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 9.589403973509933, | |
| "grad_norm": 0.3692934546349742, | |
| "learning_rate": 3.971442093158195e-08, | |
| "loss": 0.001, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 9.59823399558499, | |
| "grad_norm": 0.7262635407703527, | |
| "learning_rate": 3.7931993697644664e-08, | |
| "loss": 0.0013, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 9.607064017660043, | |
| "grad_norm": 0.4241779422028337, | |
| "learning_rate": 3.6190336495659504e-08, | |
| "loss": 0.0011, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 9.6158940397351, | |
| "grad_norm": 0.3581702129006641, | |
| "learning_rate": 3.448946363567296e-08, | |
| "loss": 0.001, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 9.624724061810154, | |
| "grad_norm": 0.6719776466853286, | |
| "learning_rate": 3.282938909263122e-08, | |
| "loss": 0.0013, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 9.63355408388521, | |
| "grad_norm": 0.35948401676170405, | |
| "learning_rate": 3.121012650627031e-08, | |
| "loss": 0.0012, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 9.642384105960264, | |
| "grad_norm": 0.2951673376699998, | |
| "learning_rate": 2.9631689180999457e-08, | |
| "loss": 0.0009, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 9.65121412803532, | |
| "grad_norm": 0.5490569151381255, | |
| "learning_rate": 2.8094090085795112e-08, | |
| "loss": 0.0011, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 9.660044150110375, | |
| "grad_norm": 1.8123675650759785, | |
| "learning_rate": 2.6597341854092685e-08, | |
| "loss": 0.0032, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 9.668874172185431, | |
| "grad_norm": 0.7750946979722031, | |
| "learning_rate": 2.514145678368163e-08, | |
| "loss": 0.0014, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 9.677704194260485, | |
| "grad_norm": 0.9327158544524087, | |
| "learning_rate": 2.3726446836608298e-08, | |
| "loss": 0.0026, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 9.686534216335541, | |
| "grad_norm": 0.9856592262516735, | |
| "learning_rate": 2.235232363907269e-08, | |
| "loss": 0.002, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 9.695364238410596, | |
| "grad_norm": 0.43878420684933445, | |
| "learning_rate": 2.101909848133743e-08, | |
| "loss": 0.0013, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.704194260485652, | |
| "grad_norm": 0.7614724669148992, | |
| "learning_rate": 1.9726782317632255e-08, | |
| "loss": 0.0013, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 9.713024282560706, | |
| "grad_norm": 1.0478575318798933, | |
| "learning_rate": 1.8475385766063002e-08, | |
| "loss": 0.0012, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.721854304635762, | |
| "grad_norm": 1.8065062800725666, | |
| "learning_rate": 1.7264919108529455e-08, | |
| "loss": 0.0019, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 9.730684326710817, | |
| "grad_norm": 0.3995756996586655, | |
| "learning_rate": 1.6095392290635393e-08, | |
| "loss": 0.0009, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 9.739514348785871, | |
| "grad_norm": 0.30625964643708703, | |
| "learning_rate": 1.4966814921608674e-08, | |
| "loss": 0.0009, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 9.748344370860927, | |
| "grad_norm": 0.7467419056881797, | |
| "learning_rate": 1.3879196274224626e-08, | |
| "loss": 0.0011, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 9.757174392935982, | |
| "grad_norm": 0.36392106960645637, | |
| "learning_rate": 1.2832545284724995e-08, | |
| "loss": 0.0011, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 9.766004415011038, | |
| "grad_norm": 0.7988446480966244, | |
| "learning_rate": 1.1826870552749669e-08, | |
| "loss": 0.0019, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 9.774834437086092, | |
| "grad_norm": 1.769706504414356, | |
| "learning_rate": 1.0862180341263962e-08, | |
| "loss": 0.0018, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 9.783664459161148, | |
| "grad_norm": 0.4826885661800605, | |
| "learning_rate": 9.938482576487551e-09, | |
| "loss": 0.0015, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 9.792494481236202, | |
| "grad_norm": 0.37989176939802677, | |
| "learning_rate": 9.055784847836202e-09, | |
| "loss": 0.0011, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 9.801324503311259, | |
| "grad_norm": 0.39147684294871354, | |
| "learning_rate": 8.214094407851814e-09, | |
| "loss": 0.0012, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 9.810154525386313, | |
| "grad_norm": 0.8161157507883349, | |
| "learning_rate": 7.413418172149689e-09, | |
| "loss": 0.001, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 9.818984547461369, | |
| "grad_norm": 0.5745629308176634, | |
| "learning_rate": 6.6537627193558055e-09, | |
| "loss": 0.0009, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 9.827814569536423, | |
| "grad_norm": 0.35754287885276564, | |
| "learning_rate": 5.93513429105741e-09, | |
| "loss": 0.0012, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 9.83664459161148, | |
| "grad_norm": 0.2837516782334886, | |
| "learning_rate": 5.257538791749173e-09, | |
| "loss": 0.0008, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 9.845474613686534, | |
| "grad_norm": 1.0021887268331198, | |
| "learning_rate": 4.6209817887848955e-09, | |
| "loss": 0.0023, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 9.85430463576159, | |
| "grad_norm": 1.0563979382196924, | |
| "learning_rate": 4.025468512333098e-09, | |
| "loss": 0.0011, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 9.863134657836644, | |
| "grad_norm": 0.5120449598547939, | |
| "learning_rate": 3.471003855332611e-09, | |
| "loss": 0.0021, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 9.8719646799117, | |
| "grad_norm": 0.27917578033380663, | |
| "learning_rate": 2.9575923734520562e-09, | |
| "loss": 0.0009, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 9.880794701986755, | |
| "grad_norm": 0.4032369259810942, | |
| "learning_rate": 2.4852382850554245e-09, | |
| "loss": 0.0009, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 9.889624724061811, | |
| "grad_norm": 0.42591200374620364, | |
| "learning_rate": 2.0539454711626663e-09, | |
| "loss": 0.0011, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.898454746136865, | |
| "grad_norm": 0.37601095057682765, | |
| "learning_rate": 1.6637174754230435e-09, | |
| "loss": 0.0009, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 9.90728476821192, | |
| "grad_norm": 0.9674450779942547, | |
| "learning_rate": 1.3145575040801605e-09, | |
| "loss": 0.0035, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 9.916114790286976, | |
| "grad_norm": 0.5463054948503889, | |
| "learning_rate": 1.0064684259525337e-09, | |
| "loss": 0.0016, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 9.92494481236203, | |
| "grad_norm": 1.300516596073848, | |
| "learning_rate": 7.394527724030598e-10, | |
| "loss": 0.0016, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 9.933774834437086, | |
| "grad_norm": 0.31089171621966494, | |
| "learning_rate": 5.13512737324029e-10, | |
| "loss": 0.0012, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 9.94260485651214, | |
| "grad_norm": 0.3887599954018867, | |
| "learning_rate": 3.2865017711380955e-10, | |
| "loss": 0.001, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 9.951434878587197, | |
| "grad_norm": 1.6911658758865258, | |
| "learning_rate": 1.848666106674113e-10, | |
| "loss": 0.001, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 9.960264900662251, | |
| "grad_norm": 0.7156035931908286, | |
| "learning_rate": 8.216321935816673e-11, | |
| "loss": 0.0015, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 9.969094922737307, | |
| "grad_norm": 0.7043895600699439, | |
| "learning_rate": 2.0540847032179955e-11, | |
| "loss": 0.0014, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 9.977924944812361, | |
| "grad_norm": 0.7718803468823796, | |
| "learning_rate": 0.0, | |
| "loss": 0.0017, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 9.977924944812361, | |
| "step": 1130, | |
| "total_flos": 161800223023104.0, | |
| "train_loss": 0.13931474798921478, | |
| "train_runtime": 7878.4693, | |
| "train_samples_per_second": 9.21, | |
| "train_steps_per_second": 0.143 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 1130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 565, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 161800223023104.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |