diff --git "a/checkpoint-565/trainer_state.json" "b/checkpoint-565/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-565/trainer_state.json" @@ -0,0 +1,3988 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.988962472406181, + "eval_steps": 500, + "global_step": 565, + "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 + } + ], + "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": false + }, + "attributes": {} + } + }, + "total_flos": 80916831789056.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}