| { | |
| "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": 25.06905818289659, | |
| "learning_rate": 2.9411764705882356e-07, | |
| "loss": 1.9387, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017660044150110375, | |
| "grad_norm": 25.68568630022093, | |
| "learning_rate": 5.882352941176471e-07, | |
| "loss": 1.928, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.026490066225165563, | |
| "grad_norm": 26.502734598525898, | |
| "learning_rate": 8.823529411764707e-07, | |
| "loss": 1.9155, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03532008830022075, | |
| "grad_norm": 25.310906985150506, | |
| "learning_rate": 1.1764705882352942e-06, | |
| "loss": 1.9235, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04415011037527594, | |
| "grad_norm": 24.72651604573988, | |
| "learning_rate": 1.4705882352941177e-06, | |
| "loss": 1.8924, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.052980132450331126, | |
| "grad_norm": 24.205412937319906, | |
| "learning_rate": 1.7647058823529414e-06, | |
| "loss": 1.8026, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06181015452538632, | |
| "grad_norm": 22.15412876974167, | |
| "learning_rate": 2.058823529411765e-06, | |
| "loss": 1.5564, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0706401766004415, | |
| "grad_norm": 19.764774939212682, | |
| "learning_rate": 2.3529411764705885e-06, | |
| "loss": 1.4926, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07947019867549669, | |
| "grad_norm": 18.085419039016024, | |
| "learning_rate": 2.647058823529412e-06, | |
| "loss": 1.036, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08830022075055188, | |
| "grad_norm": 15.607291712181967, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 0.9977, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09713024282560706, | |
| "grad_norm": 14.945489252455502, | |
| "learning_rate": 3.2352941176470594e-06, | |
| "loss": 0.9075, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10596026490066225, | |
| "grad_norm": 27.120860454783713, | |
| "learning_rate": 3.529411764705883e-06, | |
| "loss": 0.503, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.11479028697571744, | |
| "grad_norm": 13.65940891061477, | |
| "learning_rate": 3.8235294117647055e-06, | |
| "loss": 0.4615, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.12362030905077263, | |
| "grad_norm": 13.175755458095395, | |
| "learning_rate": 4.11764705882353e-06, | |
| "loss": 0.4129, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.13245033112582782, | |
| "grad_norm": 11.148898315396657, | |
| "learning_rate": 4.411764705882353e-06, | |
| "loss": 0.3516, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.141280353200883, | |
| "grad_norm": 7.428424652374598, | |
| "learning_rate": 4.705882352941177e-06, | |
| "loss": 0.3313, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.15011037527593818, | |
| "grad_norm": 11.107291096031085, | |
| "learning_rate": 5e-06, | |
| "loss": 0.322, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.15894039735099338, | |
| "grad_norm": 8.331582263755678, | |
| "learning_rate": 5.294117647058824e-06, | |
| "loss": 0.2905, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.16777041942604856, | |
| "grad_norm": 7.123373742560772, | |
| "learning_rate": 5.588235294117647e-06, | |
| "loss": 0.2977, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17660044150110377, | |
| "grad_norm": 5.2783172446030395, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.2857, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18543046357615894, | |
| "grad_norm": 4.180778519803163, | |
| "learning_rate": 6.176470588235295e-06, | |
| "loss": 0.285, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.19426048565121412, | |
| "grad_norm": 4.943481049285879, | |
| "learning_rate": 6.470588235294119e-06, | |
| "loss": 0.2707, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.20309050772626933, | |
| "grad_norm": 6.261139315191096, | |
| "learning_rate": 6.764705882352942e-06, | |
| "loss": 0.2713, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2119205298013245, | |
| "grad_norm": 5.205543023717728, | |
| "learning_rate": 7.058823529411766e-06, | |
| "loss": 0.3015, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.22075055187637968, | |
| "grad_norm": 7.1297538914315925, | |
| "learning_rate": 7.352941176470589e-06, | |
| "loss": 0.2687, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.22958057395143489, | |
| "grad_norm": 9.380313333856593, | |
| "learning_rate": 7.647058823529411e-06, | |
| "loss": 0.2752, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23841059602649006, | |
| "grad_norm": 5.461368987186966, | |
| "learning_rate": 7.941176470588236e-06, | |
| "loss": 0.268, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.24724061810154527, | |
| "grad_norm": 6.130043348495507, | |
| "learning_rate": 8.23529411764706e-06, | |
| "loss": 0.2691, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2560706401766004, | |
| "grad_norm": 4.667685505341346, | |
| "learning_rate": 8.529411764705883e-06, | |
| "loss": 0.2696, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.26490066225165565, | |
| "grad_norm": 4.33954165829827, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.2782, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2737306843267108, | |
| "grad_norm": 4.142057835286045, | |
| "learning_rate": 9.11764705882353e-06, | |
| "loss": 0.2661, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.282560706401766, | |
| "grad_norm": 5.125587426107246, | |
| "learning_rate": 9.411764705882354e-06, | |
| "loss": 0.2658, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2913907284768212, | |
| "grad_norm": 6.160516391015107, | |
| "learning_rate": 9.705882352941177e-06, | |
| "loss": 0.2724, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.30022075055187636, | |
| "grad_norm": 7.3062949667119215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2795, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3090507726269316, | |
| "grad_norm": 4.75477126908679, | |
| "learning_rate": 9.99997945915297e-06, | |
| "loss": 0.2771, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.31788079470198677, | |
| "grad_norm": 6.469930861735197, | |
| "learning_rate": 9.999917836780642e-06, | |
| "loss": 0.2746, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.32671081677704195, | |
| "grad_norm": 6.774915528699888, | |
| "learning_rate": 9.999815133389334e-06, | |
| "loss": 0.2625, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3355408388520971, | |
| "grad_norm": 3.532349122431978, | |
| "learning_rate": 9.999671349822887e-06, | |
| "loss": 0.2574, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3443708609271523, | |
| "grad_norm": 4.296282052033279, | |
| "learning_rate": 9.999486487262677e-06, | |
| "loss": 0.2878, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.35320088300220753, | |
| "grad_norm": 5.6368589176942665, | |
| "learning_rate": 9.999260547227599e-06, | |
| "loss": 0.2695, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3620309050772627, | |
| "grad_norm": 5.904653496281388, | |
| "learning_rate": 9.998993531574048e-06, | |
| "loss": 0.26, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3708609271523179, | |
| "grad_norm": 6.201716230094972, | |
| "learning_rate": 9.998685442495921e-06, | |
| "loss": 0.277, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.37969094922737306, | |
| "grad_norm": 3.9609425619162084, | |
| "learning_rate": 9.998336282524579e-06, | |
| "loss": 0.2445, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.38852097130242824, | |
| "grad_norm": 4.548164086756688, | |
| "learning_rate": 9.997946054528837e-06, | |
| "loss": 0.249, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3973509933774834, | |
| "grad_norm": 5.446596078550373, | |
| "learning_rate": 9.997514761714946e-06, | |
| "loss": 0.2644, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.40618101545253865, | |
| "grad_norm": 6.853964332896531, | |
| "learning_rate": 9.99704240762655e-06, | |
| "loss": 0.2407, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.41501103752759383, | |
| "grad_norm": 5.248615557746856, | |
| "learning_rate": 9.996528996144668e-06, | |
| "loss": 0.2343, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.423841059602649, | |
| "grad_norm": 3.851621358245188, | |
| "learning_rate": 9.995974531487668e-06, | |
| "loss": 0.2456, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4326710816777042, | |
| "grad_norm": 7.360655763951351, | |
| "learning_rate": 9.995379018211215e-06, | |
| "loss": 0.2591, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.44150110375275936, | |
| "grad_norm": 3.8007967239727547, | |
| "learning_rate": 9.994742461208251e-06, | |
| "loss": 0.2464, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4503311258278146, | |
| "grad_norm": 6.54251380380296, | |
| "learning_rate": 9.994064865708944e-06, | |
| "loss": 0.2474, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.45916114790286977, | |
| "grad_norm": 4.345236663462358, | |
| "learning_rate": 9.993346237280646e-06, | |
| "loss": 0.241, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.46799116997792495, | |
| "grad_norm": 4.340752142993626, | |
| "learning_rate": 9.992586581827853e-06, | |
| "loss": 0.2445, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.4768211920529801, | |
| "grad_norm": 8.680644031503022, | |
| "learning_rate": 9.991785905592149e-06, | |
| "loss": 0.2382, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4856512141280353, | |
| "grad_norm": 3.9551260449633228, | |
| "learning_rate": 9.990944215152166e-06, | |
| "loss": 0.2415, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.49448123620309054, | |
| "grad_norm": 4.393232827634922, | |
| "learning_rate": 9.990061517423513e-06, | |
| "loss": 0.2348, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5033112582781457, | |
| "grad_norm": 4.711307408468823, | |
| "learning_rate": 9.989137819658738e-06, | |
| "loss": 0.2453, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5121412803532008, | |
| "grad_norm": 3.562557321457758, | |
| "learning_rate": 9.988173129447251e-06, | |
| "loss": 0.2544, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5209713024282561, | |
| "grad_norm": 3.1473783838712714, | |
| "learning_rate": 9.987167454715277e-06, | |
| "loss": 0.2496, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5298013245033113, | |
| "grad_norm": 4.258342746195014, | |
| "learning_rate": 9.986120803725776e-06, | |
| "loss": 0.2417, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5386313465783664, | |
| "grad_norm": 3.0661639870207944, | |
| "learning_rate": 9.985033185078392e-06, | |
| "loss": 0.2416, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5474613686534217, | |
| "grad_norm": 6.350094923609783, | |
| "learning_rate": 9.983904607709365e-06, | |
| "loss": 0.2364, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5562913907284768, | |
| "grad_norm": 3.218801211881891, | |
| "learning_rate": 9.982735080891471e-06, | |
| "loss": 0.2427, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.565121412803532, | |
| "grad_norm": 4.189964625089664, | |
| "learning_rate": 9.981524614233938e-06, | |
| "loss": 0.2414, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5739514348785872, | |
| "grad_norm": 4.295839759826754, | |
| "learning_rate": 9.98027321768237e-06, | |
| "loss": 0.2334, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5827814569536424, | |
| "grad_norm": 5.620588539030716, | |
| "learning_rate": 9.978980901518663e-06, | |
| "loss": 0.2594, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5916114790286976, | |
| "grad_norm": 3.0744693491498944, | |
| "learning_rate": 9.977647676360927e-06, | |
| "loss": 0.2394, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6004415011037527, | |
| "grad_norm": 6.359908476906296, | |
| "learning_rate": 9.976273553163393e-06, | |
| "loss": 0.231, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.609271523178808, | |
| "grad_norm": 6.316442880887049, | |
| "learning_rate": 9.974858543216319e-06, | |
| "loss": 0.2274, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6181015452538632, | |
| "grad_norm": 5.324156655672741, | |
| "learning_rate": 9.973402658145908e-06, | |
| "loss": 0.2288, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6269315673289183, | |
| "grad_norm": 3.9538127249812374, | |
| "learning_rate": 9.971905909914206e-06, | |
| "loss": 0.2395, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6357615894039735, | |
| "grad_norm": 4.015855350861166, | |
| "learning_rate": 9.970368310819e-06, | |
| "loss": 0.2078, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6445916114790287, | |
| "grad_norm": 4.719041520529964, | |
| "learning_rate": 9.96878987349373e-06, | |
| "loss": 0.2311, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6534216335540839, | |
| "grad_norm": 4.1070724317700975, | |
| "learning_rate": 9.96717061090737e-06, | |
| "loss": 0.2262, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6622516556291391, | |
| "grad_norm": 6.469366294468439, | |
| "learning_rate": 9.965510536364329e-06, | |
| "loss": 0.2151, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6710816777041942, | |
| "grad_norm": 4.930762891482967, | |
| "learning_rate": 9.96380966350434e-06, | |
| "loss": 0.2335, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6799116997792495, | |
| "grad_norm": 3.6682691137381753, | |
| "learning_rate": 9.962068006302357e-06, | |
| "loss": 0.2228, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6887417218543046, | |
| "grad_norm": 3.8312756603355096, | |
| "learning_rate": 9.960285579068419e-06, | |
| "loss": 0.2376, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6975717439293598, | |
| "grad_norm": 5.690370014674548, | |
| "learning_rate": 9.958462396447556e-06, | |
| "loss": 0.2254, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7064017660044151, | |
| "grad_norm": 7.10575289465107, | |
| "learning_rate": 9.956598473419652e-06, | |
| "loss": 0.2367, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7152317880794702, | |
| "grad_norm": 3.786799193333354, | |
| "learning_rate": 9.954693825299333e-06, | |
| "loss": 0.2507, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7240618101545254, | |
| "grad_norm": 4.164532105917261, | |
| "learning_rate": 9.95274846773583e-06, | |
| "loss": 0.2304, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7328918322295805, | |
| "grad_norm": 2.957167457981859, | |
| "learning_rate": 9.950762416712862e-06, | |
| "loss": 0.2309, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7417218543046358, | |
| "grad_norm": 3.950243607284656, | |
| "learning_rate": 9.948735688548496e-06, | |
| "loss": 0.2394, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7505518763796909, | |
| "grad_norm": 4.199393340080547, | |
| "learning_rate": 9.946668299895017e-06, | |
| "loss": 0.2232, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7593818984547461, | |
| "grad_norm": 5.353432006990455, | |
| "learning_rate": 9.944560267738792e-06, | |
| "loss": 0.2295, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7682119205298014, | |
| "grad_norm": 5.187956574955682, | |
| "learning_rate": 9.942411609400127e-06, | |
| "loss": 0.2366, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7770419426048565, | |
| "grad_norm": 3.857122403820624, | |
| "learning_rate": 9.940222342533126e-06, | |
| "loss": 0.2302, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7858719646799117, | |
| "grad_norm": 3.8235009485898135, | |
| "learning_rate": 9.937992485125547e-06, | |
| "loss": 0.2221, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7947019867549668, | |
| "grad_norm": 4.994710419417956, | |
| "learning_rate": 9.935722055498655e-06, | |
| "loss": 0.2441, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8035320088300221, | |
| "grad_norm": 3.6274639946148177, | |
| "learning_rate": 9.933411072307071e-06, | |
| "loss": 0.2126, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8123620309050773, | |
| "grad_norm": 3.494744001694175, | |
| "learning_rate": 9.931059554538613e-06, | |
| "loss": 0.2329, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8211920529801324, | |
| "grad_norm": 3.78348592809509, | |
| "learning_rate": 9.928667521514149e-06, | |
| "loss": 0.2127, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8300220750551877, | |
| "grad_norm": 5.340365599979518, | |
| "learning_rate": 9.92623499288743e-06, | |
| "loss": 0.2222, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8388520971302428, | |
| "grad_norm": 6.5579314449709445, | |
| "learning_rate": 9.92376198864494e-06, | |
| "loss": 0.2183, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.847682119205298, | |
| "grad_norm": 4.959651982342057, | |
| "learning_rate": 9.921248529105716e-06, | |
| "loss": 0.223, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8565121412803532, | |
| "grad_norm": 3.4308915999621528, | |
| "learning_rate": 9.918694634921195e-06, | |
| "loss": 0.2263, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8653421633554084, | |
| "grad_norm": 4.279324835441828, | |
| "learning_rate": 9.916100327075038e-06, | |
| "loss": 0.2106, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8741721854304636, | |
| "grad_norm": 5.118504467469523, | |
| "learning_rate": 9.913465626882954e-06, | |
| "loss": 0.2313, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8830022075055187, | |
| "grad_norm": 5.611920388270936, | |
| "learning_rate": 9.910790555992536e-06, | |
| "loss": 0.2241, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.891832229580574, | |
| "grad_norm": 3.201069614707369, | |
| "learning_rate": 9.908075136383068e-06, | |
| "loss": 0.2045, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9006622516556292, | |
| "grad_norm": 6.023860874306649, | |
| "learning_rate": 9.905319390365364e-06, | |
| "loss": 0.2361, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9094922737306843, | |
| "grad_norm": 4.541003622505249, | |
| "learning_rate": 9.902523340581562e-06, | |
| "loss": 0.2231, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9183222958057395, | |
| "grad_norm": 3.892822194098379, | |
| "learning_rate": 9.899687010004956e-06, | |
| "loss": 0.2346, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9271523178807947, | |
| "grad_norm": 4.550335472806411, | |
| "learning_rate": 9.896810421939797e-06, | |
| "loss": 0.2332, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9359823399558499, | |
| "grad_norm": 3.85828415445978, | |
| "learning_rate": 9.893893600021112e-06, | |
| "loss": 0.2159, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9448123620309051, | |
| "grad_norm": 5.446836037485189, | |
| "learning_rate": 9.890936568214493e-06, | |
| "loss": 0.2336, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9536423841059603, | |
| "grad_norm": 5.648084319092979, | |
| "learning_rate": 9.88793935081592e-06, | |
| "loss": 0.219, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9624724061810155, | |
| "grad_norm": 10.700799326300302, | |
| "learning_rate": 9.884901972451542e-06, | |
| "loss": 0.2257, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9713024282560706, | |
| "grad_norm": 6.281034440727802, | |
| "learning_rate": 9.881824458077491e-06, | |
| "loss": 0.2295, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9801324503311258, | |
| "grad_norm": 4.701514002631264, | |
| "learning_rate": 9.878706832979668e-06, | |
| "loss": 0.2129, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9889624724061811, | |
| "grad_norm": 4.383624556135786, | |
| "learning_rate": 9.875549122773536e-06, | |
| "loss": 0.2312, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9977924944812362, | |
| "grad_norm": 3.576068647188603, | |
| "learning_rate": 9.872351353403912e-06, | |
| "loss": 0.2141, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0066225165562914, | |
| "grad_norm": 6.0830817761682985, | |
| "learning_rate": 9.869113551144754e-06, | |
| "loss": 0.2014, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0154525386313467, | |
| "grad_norm": 4.7531813854125655, | |
| "learning_rate": 9.865835742598942e-06, | |
| "loss": 0.2022, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0242825607064017, | |
| "grad_norm": 3.2824105375326478, | |
| "learning_rate": 9.86251795469806e-06, | |
| "loss": 0.1814, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.033112582781457, | |
| "grad_norm": 2.786895822375094, | |
| "learning_rate": 9.85916021470218e-06, | |
| "loss": 0.1898, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0419426048565121, | |
| "grad_norm": 4.098036626468437, | |
| "learning_rate": 9.85576255019963e-06, | |
| "loss": 0.1923, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.0507726269315674, | |
| "grad_norm": 3.916652138771565, | |
| "learning_rate": 9.852324989106772e-06, | |
| "loss": 0.1877, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0596026490066226, | |
| "grad_norm": 5.455813349247477, | |
| "learning_rate": 9.848847559667774e-06, | |
| "loss": 0.1846, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0684326710816776, | |
| "grad_norm": 10.542661969730387, | |
| "learning_rate": 9.845330290454373e-06, | |
| "loss": 0.1718, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0772626931567328, | |
| "grad_norm": 4.02767225020254, | |
| "learning_rate": 9.841773210365646e-06, | |
| "loss": 0.19, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.086092715231788, | |
| "grad_norm": 6.839436084957277, | |
| "learning_rate": 9.838176348627768e-06, | |
| "loss": 0.1708, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0949227373068433, | |
| "grad_norm": 4.798409291630464, | |
| "learning_rate": 9.834539734793774e-06, | |
| "loss": 0.1846, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1037527593818985, | |
| "grad_norm": 4.137394904622889, | |
| "learning_rate": 9.830863398743313e-06, | |
| "loss": 0.1789, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1125827814569536, | |
| "grad_norm": 3.88794974694025, | |
| "learning_rate": 9.82714737068241e-06, | |
| "loss": 0.1737, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.1214128035320088, | |
| "grad_norm": 3.4537997696122362, | |
| "learning_rate": 9.82339168114321e-06, | |
| "loss": 0.1814, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.130242825607064, | |
| "grad_norm": 3.743862822832632, | |
| "learning_rate": 9.81959636098373e-06, | |
| "loss": 0.1776, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.1390728476821192, | |
| "grad_norm": 3.64783731609154, | |
| "learning_rate": 9.815761441387609e-06, | |
| "loss": 0.1863, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1479028697571745, | |
| "grad_norm": 4.945406809993386, | |
| "learning_rate": 9.811886953863841e-06, | |
| "loss": 0.1948, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1567328918322295, | |
| "grad_norm": 3.995572235306309, | |
| "learning_rate": 9.807972930246531e-06, | |
| "loss": 0.189, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1655629139072847, | |
| "grad_norm": 5.315570871040949, | |
| "learning_rate": 9.804019402694627e-06, | |
| "loss": 0.1878, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.17439293598234, | |
| "grad_norm": 7.942697593904259, | |
| "learning_rate": 9.800026403691643e-06, | |
| "loss": 0.1822, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1832229580573952, | |
| "grad_norm": 3.7100381437064605, | |
| "learning_rate": 9.795993966045418e-06, | |
| "loss": 0.184, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1920529801324504, | |
| "grad_norm": 3.515806231432155, | |
| "learning_rate": 9.791922122887823e-06, | |
| "loss": 0.1557, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2008830022075054, | |
| "grad_norm": 11.905142017862433, | |
| "learning_rate": 9.7878109076745e-06, | |
| "loss": 0.1886, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.2097130242825607, | |
| "grad_norm": 5.549791017896183, | |
| "learning_rate": 9.783660354184589e-06, | |
| "loss": 0.1911, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.218543046357616, | |
| "grad_norm": 3.1831986903773575, | |
| "learning_rate": 9.779470496520442e-06, | |
| "loss": 0.1845, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.2273730684326711, | |
| "grad_norm": 7.241858960668914, | |
| "learning_rate": 9.77524136910735e-06, | |
| "loss": 0.1746, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.2362030905077264, | |
| "grad_norm": 3.2346972578304034, | |
| "learning_rate": 9.770973006693256e-06, | |
| "loss": 0.1519, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2450331125827814, | |
| "grad_norm": 6.403826372889126, | |
| "learning_rate": 9.766665444348472e-06, | |
| "loss": 0.1758, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2538631346578366, | |
| "grad_norm": 4.4524344269491625, | |
| "learning_rate": 9.76231871746539e-06, | |
| "loss": 0.1979, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2626931567328918, | |
| "grad_norm": 3.862760163174027, | |
| "learning_rate": 9.757932861758188e-06, | |
| "loss": 0.2054, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.271523178807947, | |
| "grad_norm": 3.2208460551997553, | |
| "learning_rate": 9.753507913262548e-06, | |
| "loss": 0.1711, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.280353200883002, | |
| "grad_norm": 3.9685569501042948, | |
| "learning_rate": 9.749043908335343e-06, | |
| "loss": 0.154, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2891832229580573, | |
| "grad_norm": 4.700535965091022, | |
| "learning_rate": 9.744540883654348e-06, | |
| "loss": 0.1803, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2980132450331126, | |
| "grad_norm": 3.290798769098629, | |
| "learning_rate": 9.739998876217943e-06, | |
| "loss": 0.1904, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.3068432671081678, | |
| "grad_norm": 4.379650402165434, | |
| "learning_rate": 9.735417923344798e-06, | |
| "loss": 0.1828, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.315673289183223, | |
| "grad_norm": 4.4512235264983335, | |
| "learning_rate": 9.730798062673575e-06, | |
| "loss": 0.1963, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.3245033112582782, | |
| "grad_norm": 4.928447363077229, | |
| "learning_rate": 9.726139332162613e-06, | |
| "loss": 0.1986, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 3.167430281144184, | |
| "learning_rate": 9.721441770089621e-06, | |
| "loss": 0.1783, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.3421633554083885, | |
| "grad_norm": 5.16299387787309, | |
| "learning_rate": 9.716705415051362e-06, | |
| "loss": 0.1817, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3509933774834437, | |
| "grad_norm": 5.22565296604379, | |
| "learning_rate": 9.711930305963333e-06, | |
| "loss": 0.1606, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.359823399558499, | |
| "grad_norm": 6.989669414948377, | |
| "learning_rate": 9.707116482059447e-06, | |
| "loss": 0.1924, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.368653421633554, | |
| "grad_norm": 4.288151185944291, | |
| "learning_rate": 9.702263982891712e-06, | |
| "loss": 0.1651, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3774834437086092, | |
| "grad_norm": 4.567897047089374, | |
| "learning_rate": 9.697372848329905e-06, | |
| "loss": 0.1842, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.3863134657836644, | |
| "grad_norm": 4.29424871998317, | |
| "learning_rate": 9.692443118561248e-06, | |
| "loss": 0.173, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3951434878587197, | |
| "grad_norm": 7.122508056007231, | |
| "learning_rate": 9.68747483409007e-06, | |
| "loss": 0.1813, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.403973509933775, | |
| "grad_norm": 20.80552868339202, | |
| "learning_rate": 9.682468035737475e-06, | |
| "loss": 0.1948, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.4128035320088301, | |
| "grad_norm": 7.039552426013288, | |
| "learning_rate": 9.677422764641021e-06, | |
| "loss": 0.1782, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.4216335540838851, | |
| "grad_norm": 5.217781974573682, | |
| "learning_rate": 9.672339062254359e-06, | |
| "loss": 0.1813, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.4304635761589404, | |
| "grad_norm": 3.4648301225351474, | |
| "learning_rate": 9.667216970346916e-06, | |
| "loss": 0.1789, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.4392935982339956, | |
| "grad_norm": 23.4115100661157, | |
| "learning_rate": 9.662056531003528e-06, | |
| "loss": 0.1783, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4481236203090508, | |
| "grad_norm": 4.149843420090105, | |
| "learning_rate": 9.656857786624119e-06, | |
| "loss": 0.1895, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4569536423841059, | |
| "grad_norm": 6.867974570202788, | |
| "learning_rate": 9.651620779923332e-06, | |
| "loss": 0.1766, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.465783664459161, | |
| "grad_norm": 5.828193293904613, | |
| "learning_rate": 9.646345553930187e-06, | |
| "loss": 0.175, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4746136865342163, | |
| "grad_norm": 6.2210187622807265, | |
| "learning_rate": 9.64103215198773e-06, | |
| "loss": 0.1891, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.4834437086092715, | |
| "grad_norm": 9.016484689070838, | |
| "learning_rate": 9.63568061775267e-06, | |
| "loss": 0.189, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.4922737306843268, | |
| "grad_norm": 3.531196386481693, | |
| "learning_rate": 9.630290995195028e-06, | |
| "loss": 0.1937, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.501103752759382, | |
| "grad_norm": 4.641638553308808, | |
| "learning_rate": 9.624863328597767e-06, | |
| "loss": 0.1763, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5099337748344372, | |
| "grad_norm": 4.68576860426708, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.1859, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.5187637969094923, | |
| "grad_norm": 5.383190718687304, | |
| "learning_rate": 9.613894041978795e-06, | |
| "loss": 0.1931, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.5275938189845475, | |
| "grad_norm": 4.33787132489928, | |
| "learning_rate": 9.60835251208446e-06, | |
| "loss": 0.181, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.5364238410596025, | |
| "grad_norm": 5.0471966768108585, | |
| "learning_rate": 9.602773118404518e-06, | |
| "loss": 0.1828, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.5452538631346577, | |
| "grad_norm": 4.52992338383335, | |
| "learning_rate": 9.597155906781154e-06, | |
| "loss": 0.1805, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.554083885209713, | |
| "grad_norm": 4.631621416482625, | |
| "learning_rate": 9.591500923367287e-06, | |
| "loss": 0.1764, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5629139072847682, | |
| "grad_norm": 4.130494606770307, | |
| "learning_rate": 9.585808214626173e-06, | |
| "loss": 0.1811, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5717439293598234, | |
| "grad_norm": 9.27347846169718, | |
| "learning_rate": 9.580077827331038e-06, | |
| "loss": 0.1794, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5805739514348787, | |
| "grad_norm": 4.156796060089488, | |
| "learning_rate": 9.574309808564682e-06, | |
| "loss": 0.1791, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.589403973509934, | |
| "grad_norm": 5.7164314447515885, | |
| "learning_rate": 9.568504205719106e-06, | |
| "loss": 0.1583, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.598233995584989, | |
| "grad_norm": 4.1934597232319035, | |
| "learning_rate": 9.562661066495108e-06, | |
| "loss": 0.1632, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.6070640176600441, | |
| "grad_norm": 3.7973289921753484, | |
| "learning_rate": 9.556780438901899e-06, | |
| "loss": 0.1832, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.6158940397350994, | |
| "grad_norm": 6.698619995777142, | |
| "learning_rate": 9.550862371256705e-06, | |
| "loss": 0.1686, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.6247240618101544, | |
| "grad_norm": 5.1689172402316075, | |
| "learning_rate": 9.544906912184383e-06, | |
| "loss": 0.1565, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.6335540838852096, | |
| "grad_norm": 3.654331376556239, | |
| "learning_rate": 9.538914110616995e-06, | |
| "loss": 0.1708, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.6423841059602649, | |
| "grad_norm": 6.304736886487628, | |
| "learning_rate": 9.532884015793432e-06, | |
| "loss": 0.1825, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.65121412803532, | |
| "grad_norm": 3.859565209636668, | |
| "learning_rate": 9.526816677258995e-06, | |
| "loss": 0.1735, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.6600441501103753, | |
| "grad_norm": 3.5711915158076932, | |
| "learning_rate": 9.520712144864997e-06, | |
| "loss": 0.1681, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6688741721854305, | |
| "grad_norm": 4.435977147295718, | |
| "learning_rate": 9.514570468768338e-06, | |
| "loss": 0.1789, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.6777041942604858, | |
| "grad_norm": 4.0501977712489765, | |
| "learning_rate": 9.508391699431114e-06, | |
| "loss": 0.1528, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6865342163355408, | |
| "grad_norm": 4.071633625203301, | |
| "learning_rate": 9.502175887620188e-06, | |
| "loss": 0.1729, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.695364238410596, | |
| "grad_norm": 3.6268611641097297, | |
| "learning_rate": 9.495923084406773e-06, | |
| "loss": 0.1661, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.7041942604856513, | |
| "grad_norm": 16.20474671557156, | |
| "learning_rate": 9.48963334116602e-06, | |
| "loss": 0.1748, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.7130242825607063, | |
| "grad_norm": 4.423640904498312, | |
| "learning_rate": 9.48330670957659e-06, | |
| "loss": 0.1836, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.7218543046357615, | |
| "grad_norm": 6.3298241198543135, | |
| "learning_rate": 9.476943241620233e-06, | |
| "loss": 0.1748, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.7306843267108167, | |
| "grad_norm": 4.032022414185224, | |
| "learning_rate": 9.470542989581357e-06, | |
| "loss": 0.1801, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.739514348785872, | |
| "grad_norm": 5.580442910747046, | |
| "learning_rate": 9.464106006046602e-06, | |
| "loss": 0.18, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.7483443708609272, | |
| "grad_norm": 4.9625941262808935, | |
| "learning_rate": 9.457632343904404e-06, | |
| "loss": 0.1641, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7571743929359824, | |
| "grad_norm": 4.036510413716825, | |
| "learning_rate": 9.451122056344564e-06, | |
| "loss": 0.1623, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7660044150110377, | |
| "grad_norm": 4.351769673457381, | |
| "learning_rate": 9.444575196857814e-06, | |
| "loss": 0.1667, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7748344370860927, | |
| "grad_norm": 3.503055068180116, | |
| "learning_rate": 9.437991819235366e-06, | |
| "loss": 0.1761, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.783664459161148, | |
| "grad_norm": 4.040453399695261, | |
| "learning_rate": 9.431371977568483e-06, | |
| "loss": 0.1845, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.7924944812362031, | |
| "grad_norm": 4.543045949052899, | |
| "learning_rate": 9.424715726248027e-06, | |
| "loss": 0.1763, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.8013245033112582, | |
| "grad_norm": 4.994252724100901, | |
| "learning_rate": 9.418023119964012e-06, | |
| "loss": 0.19, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.8101545253863134, | |
| "grad_norm": 5.806173853040929, | |
| "learning_rate": 9.411294213705162e-06, | |
| "loss": 0.1947, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.8189845474613686, | |
| "grad_norm": 5.481669048806609, | |
| "learning_rate": 9.404529062758447e-06, | |
| "loss": 0.16, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.8278145695364238, | |
| "grad_norm": 3.2196597657326893, | |
| "learning_rate": 9.397727722708643e-06, | |
| "loss": 0.1881, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.836644591611479, | |
| "grad_norm": 4.75501651474976, | |
| "learning_rate": 9.390890249437863e-06, | |
| "loss": 0.1652, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.8454746136865343, | |
| "grad_norm": 6.469771054964556, | |
| "learning_rate": 9.384016699125102e-06, | |
| "loss": 0.1548, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.8543046357615895, | |
| "grad_norm": 3.7820316395600924, | |
| "learning_rate": 9.377107128245782e-06, | |
| "loss": 0.172, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8631346578366446, | |
| "grad_norm": 4.667257391932444, | |
| "learning_rate": 9.370161593571274e-06, | |
| "loss": 0.1751, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.8719646799116998, | |
| "grad_norm": 5.767881384215297, | |
| "learning_rate": 9.363180152168448e-06, | |
| "loss": 0.1753, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.8807947019867548, | |
| "grad_norm": 3.62378559765123, | |
| "learning_rate": 9.356162861399188e-06, | |
| "loss": 0.1595, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.88962472406181, | |
| "grad_norm": 5.406688558269064, | |
| "learning_rate": 9.349109778919938e-06, | |
| "loss": 0.1757, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.8984547461368653, | |
| "grad_norm": 4.5151455175909, | |
| "learning_rate": 9.342020962681206e-06, | |
| "loss": 0.1591, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.9072847682119205, | |
| "grad_norm": 7.874317090048882, | |
| "learning_rate": 9.334896470927115e-06, | |
| "loss": 0.177, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.9161147902869757, | |
| "grad_norm": 4.916816313808601, | |
| "learning_rate": 9.327736362194899e-06, | |
| "loss": 0.1708, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.924944812362031, | |
| "grad_norm": 4.778734339774051, | |
| "learning_rate": 9.32054069531444e-06, | |
| "loss": 0.1749, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.9337748344370862, | |
| "grad_norm": 6.049032923004645, | |
| "learning_rate": 9.313309529407773e-06, | |
| "loss": 0.1697, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.9426048565121414, | |
| "grad_norm": 4.146956823403823, | |
| "learning_rate": 9.306042923888607e-06, | |
| "loss": 0.177, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9514348785871964, | |
| "grad_norm": 6.77533831621212, | |
| "learning_rate": 9.298740938461835e-06, | |
| "loss": 0.1758, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.9602649006622517, | |
| "grad_norm": 3.206891132737421, | |
| "learning_rate": 9.291403633123046e-06, | |
| "loss": 0.152, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.9690949227373067, | |
| "grad_norm": 4.306726275574873, | |
| "learning_rate": 9.284031068158023e-06, | |
| "loss": 0.1691, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.977924944812362, | |
| "grad_norm": 4.0544766400174765, | |
| "learning_rate": 9.27662330414226e-06, | |
| "loss": 0.1724, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.9867549668874172, | |
| "grad_norm": 7.120274337729496, | |
| "learning_rate": 9.269180401940455e-06, | |
| "loss": 0.1743, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9955849889624724, | |
| "grad_norm": 3.4193567491490855, | |
| "learning_rate": 9.261702422706014e-06, | |
| "loss": 0.1597, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.0044150110375276, | |
| "grad_norm": 4.106288088301316, | |
| "learning_rate": 9.254189427880548e-06, | |
| "loss": 0.141, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.013245033112583, | |
| "grad_norm": 2.882770025798958, | |
| "learning_rate": 9.24664147919337e-06, | |
| "loss": 0.1305, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.022075055187638, | |
| "grad_norm": 3.511207421364191, | |
| "learning_rate": 9.239058638660983e-06, | |
| "loss": 0.123, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.0309050772626933, | |
| "grad_norm": 5.765668062012182, | |
| "learning_rate": 9.231440968586572e-06, | |
| "loss": 0.1284, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0397350993377485, | |
| "grad_norm": 6.6705151722336895, | |
| "learning_rate": 9.223788531559495e-06, | |
| "loss": 0.1166, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.0485651214128033, | |
| "grad_norm": 5.678256758749395, | |
| "learning_rate": 9.216101390454771e-06, | |
| "loss": 0.1467, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.0573951434878586, | |
| "grad_norm": 4.597940897671489, | |
| "learning_rate": 9.208379608432552e-06, | |
| "loss": 0.1406, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.066225165562914, | |
| "grad_norm": 5.145951489387972, | |
| "learning_rate": 9.200623248937619e-06, | |
| "loss": 0.1201, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.075055187637969, | |
| "grad_norm": 6.588636998795935, | |
| "learning_rate": 9.192832375698845e-06, | |
| "loss": 0.1285, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.0838852097130243, | |
| "grad_norm": 5.135646392206335, | |
| "learning_rate": 9.185007052728689e-06, | |
| "loss": 0.116, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.0927152317880795, | |
| "grad_norm": 4.220282717321963, | |
| "learning_rate": 9.177147344322651e-06, | |
| "loss": 0.1326, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.1015452538631347, | |
| "grad_norm": 5.711014478089566, | |
| "learning_rate": 9.169253315058764e-06, | |
| "loss": 0.1352, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.11037527593819, | |
| "grad_norm": 4.054495215920748, | |
| "learning_rate": 9.161325029797044e-06, | |
| "loss": 0.1145, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.119205298013245, | |
| "grad_norm": 4.521366476282431, | |
| "learning_rate": 9.153362553678967e-06, | |
| "loss": 0.1226, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1280353200883004, | |
| "grad_norm": 4.722112139773477, | |
| "learning_rate": 9.145365952126937e-06, | |
| "loss": 0.1185, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.136865342163355, | |
| "grad_norm": 7.469499081939174, | |
| "learning_rate": 9.13733529084374e-06, | |
| "loss": 0.1279, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.1456953642384105, | |
| "grad_norm": 5.488709858524371, | |
| "learning_rate": 9.129270635812013e-06, | |
| "loss": 0.1371, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.1545253863134657, | |
| "grad_norm": 4.729490430275487, | |
| "learning_rate": 9.12117205329369e-06, | |
| "loss": 0.1232, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.163355408388521, | |
| "grad_norm": 3.697980948897901, | |
| "learning_rate": 9.113039609829472e-06, | |
| "loss": 0.1262, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.172185430463576, | |
| "grad_norm": 5.729648089011882, | |
| "learning_rate": 9.104873372238269e-06, | |
| "loss": 0.1253, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.1810154525386314, | |
| "grad_norm": 4.416650629008491, | |
| "learning_rate": 9.096673407616656e-06, | |
| "loss": 0.1255, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.1898454746136866, | |
| "grad_norm": 3.6240003911621166, | |
| "learning_rate": 9.08843978333832e-06, | |
| "loss": 0.1326, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.198675496688742, | |
| "grad_norm": 5.710954987113158, | |
| "learning_rate": 9.08017256705351e-06, | |
| "loss": 0.1209, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.207505518763797, | |
| "grad_norm": 3.4673951330264674, | |
| "learning_rate": 9.071871826688472e-06, | |
| "loss": 0.1335, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.216335540838852, | |
| "grad_norm": 6.636636329580233, | |
| "learning_rate": 9.063537630444903e-06, | |
| "loss": 0.1039, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.225165562913907, | |
| "grad_norm": 4.560755557071873, | |
| "learning_rate": 9.055170046799386e-06, | |
| "loss": 0.1227, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.2339955849889623, | |
| "grad_norm": 4.623925302817527, | |
| "learning_rate": 9.046769144502818e-06, | |
| "loss": 0.1261, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.2428256070640176, | |
| "grad_norm": 4.233240951948466, | |
| "learning_rate": 9.038334992579863e-06, | |
| "loss": 0.1211, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.251655629139073, | |
| "grad_norm": 4.332831951430898, | |
| "learning_rate": 9.029867660328369e-06, | |
| "loss": 0.1281, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.260485651214128, | |
| "grad_norm": 5.35275306470717, | |
| "learning_rate": 9.021367217318808e-06, | |
| "loss": 0.1266, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.2693156732891833, | |
| "grad_norm": 5.683503750461502, | |
| "learning_rate": 9.012833733393697e-06, | |
| "loss": 0.1308, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.2781456953642385, | |
| "grad_norm": 16.76988877462198, | |
| "learning_rate": 9.004267278667032e-06, | |
| "loss": 0.1281, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.2869757174392937, | |
| "grad_norm": 8.92628471689616, | |
| "learning_rate": 8.99566792352371e-06, | |
| "loss": 0.1386, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.295805739514349, | |
| "grad_norm": 3.6085083460551646, | |
| "learning_rate": 8.987035738618943e-06, | |
| "loss": 0.122, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.304635761589404, | |
| "grad_norm": 5.51707997275494, | |
| "learning_rate": 8.978370794877691e-06, | |
| "loss": 0.1302, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.313465783664459, | |
| "grad_norm": 4.397721834795367, | |
| "learning_rate": 8.969673163494063e-06, | |
| "loss": 0.1199, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.322295805739514, | |
| "grad_norm": 3.3276046272638395, | |
| "learning_rate": 8.960942915930749e-06, | |
| "loss": 0.1374, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.3311258278145695, | |
| "grad_norm": 4.481239495813595, | |
| "learning_rate": 8.952180123918419e-06, | |
| "loss": 0.1155, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.3399558498896247, | |
| "grad_norm": 3.3371182566992528, | |
| "learning_rate": 8.94338485945514e-06, | |
| "loss": 0.1272, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.34878587196468, | |
| "grad_norm": 9.526212734949459, | |
| "learning_rate": 8.934557194805787e-06, | |
| "loss": 0.1375, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.357615894039735, | |
| "grad_norm": 3.2449386868810666, | |
| "learning_rate": 8.925697202501442e-06, | |
| "loss": 0.1288, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.3664459161147904, | |
| "grad_norm": 10.65763509588136, | |
| "learning_rate": 8.916804955338807e-06, | |
| "loss": 0.1257, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.3752759381898456, | |
| "grad_norm": 4.285910268021899, | |
| "learning_rate": 8.907880526379594e-06, | |
| "loss": 0.1197, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.384105960264901, | |
| "grad_norm": 4.138365874339128, | |
| "learning_rate": 8.898923988949936e-06, | |
| "loss": 0.1189, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.3929359823399556, | |
| "grad_norm": 4.922900685358142, | |
| "learning_rate": 8.88993541663978e-06, | |
| "loss": 0.1211, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.401766004415011, | |
| "grad_norm": 7.231740869831, | |
| "learning_rate": 8.880914883302278e-06, | |
| "loss": 0.1215, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.410596026490066, | |
| "grad_norm": 5.019824722823812, | |
| "learning_rate": 8.871862463053193e-06, | |
| "loss": 0.1186, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.4194260485651213, | |
| "grad_norm": 13.177984589352702, | |
| "learning_rate": 8.862778230270276e-06, | |
| "loss": 0.1295, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.4282560706401766, | |
| "grad_norm": 5.092803874683864, | |
| "learning_rate": 8.85366225959266e-06, | |
| "loss": 0.1193, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.437086092715232, | |
| "grad_norm": 4.645436040352056, | |
| "learning_rate": 8.844514625920246e-06, | |
| "loss": 0.1195, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.445916114790287, | |
| "grad_norm": 4.4792099516559345, | |
| "learning_rate": 8.835335404413096e-06, | |
| "loss": 0.1315, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.4547461368653423, | |
| "grad_norm": 4.376327551414001, | |
| "learning_rate": 8.826124670490804e-06, | |
| "loss": 0.1265, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.4635761589403975, | |
| "grad_norm": 5.143967207001637, | |
| "learning_rate": 8.816882499831877e-06, | |
| "loss": 0.1385, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.4724061810154527, | |
| "grad_norm": 4.727297780865657, | |
| "learning_rate": 8.807608968373123e-06, | |
| "loss": 0.1213, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.481236203090508, | |
| "grad_norm": 4.348943546633772, | |
| "learning_rate": 8.798304152309019e-06, | |
| "loss": 0.1194, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.4900662251655628, | |
| "grad_norm": 3.5448573901775275, | |
| "learning_rate": 8.788968128091084e-06, | |
| "loss": 0.1455, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.498896247240618, | |
| "grad_norm": 4.0224898315004065, | |
| "learning_rate": 8.779600972427257e-06, | |
| "loss": 0.1375, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.507726269315673, | |
| "grad_norm": 6.783149660122743, | |
| "learning_rate": 8.770202762281267e-06, | |
| "loss": 0.1232, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.5165562913907285, | |
| "grad_norm": 3.097212078028231, | |
| "learning_rate": 8.760773574871985e-06, | |
| "loss": 0.115, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.5253863134657837, | |
| "grad_norm": 7.372711877322491, | |
| "learning_rate": 8.751313487672815e-06, | |
| "loss": 0.134, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.534216335540839, | |
| "grad_norm": 3.8347010004386894, | |
| "learning_rate": 8.741822578411036e-06, | |
| "loss": 0.1376, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.543046357615894, | |
| "grad_norm": 3.2301374236057434, | |
| "learning_rate": 8.732300925067177e-06, | |
| "loss": 0.1315, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.5518763796909494, | |
| "grad_norm": 6.855372976047249, | |
| "learning_rate": 8.722748605874365e-06, | |
| "loss": 0.1355, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.560706401766004, | |
| "grad_norm": 6.1942078029864405, | |
| "learning_rate": 8.71316569931769e-06, | |
| "loss": 0.132, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.5695364238410594, | |
| "grad_norm": 3.7145168130345305, | |
| "learning_rate": 8.703552284133565e-06, | |
| "loss": 0.1336, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.5783664459161146, | |
| "grad_norm": 5.010946900888269, | |
| "learning_rate": 8.69390843930906e-06, | |
| "loss": 0.1308, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.58719646799117, | |
| "grad_norm": 4.886631546645197, | |
| "learning_rate": 8.684234244081274e-06, | |
| "loss": 0.1259, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.596026490066225, | |
| "grad_norm": 2.940069775298397, | |
| "learning_rate": 8.674529777936674e-06, | |
| "loss": 0.1138, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.6048565121412803, | |
| "grad_norm": 5.251924590305838, | |
| "learning_rate": 8.66479512061044e-06, | |
| "loss": 0.1417, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.6136865342163356, | |
| "grad_norm": 3.7733208885863867, | |
| "learning_rate": 8.655030352085816e-06, | |
| "loss": 0.1374, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.622516556291391, | |
| "grad_norm": 3.9170101048822246, | |
| "learning_rate": 8.645235552593447e-06, | |
| "loss": 0.1253, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.631346578366446, | |
| "grad_norm": 4.150269659617099, | |
| "learning_rate": 8.635410802610724e-06, | |
| "loss": 0.1276, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.6401766004415013, | |
| "grad_norm": 3.5267586217995732, | |
| "learning_rate": 8.625556182861126e-06, | |
| "loss": 0.132, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.6490066225165565, | |
| "grad_norm": 3.9458309081426877, | |
| "learning_rate": 8.615671774313543e-06, | |
| "loss": 0.1444, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.6578366445916117, | |
| "grad_norm": 4.0021555334427195, | |
| "learning_rate": 8.605757658181626e-06, | |
| "loss": 0.1249, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 3.4783433195038156, | |
| "learning_rate": 8.595813915923113e-06, | |
| "loss": 0.1368, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.6754966887417218, | |
| "grad_norm": 3.9613428910630617, | |
| "learning_rate": 8.585840629239158e-06, | |
| "loss": 0.1149, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.684326710816777, | |
| "grad_norm": 3.8749406279119936, | |
| "learning_rate": 8.575837880073663e-06, | |
| "loss": 0.1251, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.693156732891832, | |
| "grad_norm": 7.232470603311793, | |
| "learning_rate": 8.565805750612607e-06, | |
| "loss": 0.1424, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.7019867549668874, | |
| "grad_norm": 3.6550319288550157, | |
| "learning_rate": 8.555744323283364e-06, | |
| "loss": 0.1277, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.7108167770419427, | |
| "grad_norm": 6.097422404372455, | |
| "learning_rate": 8.545653680754029e-06, | |
| "loss": 0.1315, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.719646799116998, | |
| "grad_norm": 3.473890295262665, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.1207, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.7284768211920527, | |
| "grad_norm": 3.1376859437887052, | |
| "learning_rate": 8.525385081966992e-06, | |
| "loss": 0.1188, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.737306843267108, | |
| "grad_norm": 3.729731832819541, | |
| "learning_rate": 8.515207292242969e-06, | |
| "loss": 0.1281, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.746136865342163, | |
| "grad_norm": 7.591563114540939, | |
| "learning_rate": 8.505000620384834e-06, | |
| "loss": 0.14, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.7549668874172184, | |
| "grad_norm": 6.158725034789448, | |
| "learning_rate": 8.494765150254063e-06, | |
| "loss": 0.1284, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.7637969094922736, | |
| "grad_norm": 3.3869145111066437, | |
| "learning_rate": 8.484500965948746e-06, | |
| "loss": 0.1252, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.772626931567329, | |
| "grad_norm": 9.32989798703237, | |
| "learning_rate": 8.474208151802898e-06, | |
| "loss": 0.136, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.781456953642384, | |
| "grad_norm": 3.964154637019035, | |
| "learning_rate": 8.46388679238577e-06, | |
| "loss": 0.1269, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.7902869757174393, | |
| "grad_norm": 3.8668966734142876, | |
| "learning_rate": 8.453536972501146e-06, | |
| "loss": 0.1319, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.7991169977924946, | |
| "grad_norm": 4.800270912440046, | |
| "learning_rate": 8.443158777186652e-06, | |
| "loss": 0.125, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.80794701986755, | |
| "grad_norm": 8.17005200004958, | |
| "learning_rate": 8.432752291713058e-06, | |
| "loss": 0.1407, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.816777041942605, | |
| "grad_norm": 4.195643135069415, | |
| "learning_rate": 8.422317601583576e-06, | |
| "loss": 0.1259, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.8256070640176603, | |
| "grad_norm": 5.061888356796782, | |
| "learning_rate": 8.411854792533154e-06, | |
| "loss": 0.151, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.8344370860927155, | |
| "grad_norm": 4.503620333584586, | |
| "learning_rate": 8.401363950527777e-06, | |
| "loss": 0.1152, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.8432671081677703, | |
| "grad_norm": 5.13276069257213, | |
| "learning_rate": 8.390845161763756e-06, | |
| "loss": 0.1334, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.8520971302428255, | |
| "grad_norm": 4.374037184205516, | |
| "learning_rate": 8.380298512667023e-06, | |
| "loss": 0.1323, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.8609271523178808, | |
| "grad_norm": 3.0639753366300453, | |
| "learning_rate": 8.369724089892423e-06, | |
| "loss": 0.1252, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.869757174392936, | |
| "grad_norm": 4.19388003319346, | |
| "learning_rate": 8.359121980322992e-06, | |
| "loss": 0.1193, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.878587196467991, | |
| "grad_norm": 15.214151684387948, | |
| "learning_rate": 8.34849227106926e-06, | |
| "loss": 0.1336, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.8874172185430464, | |
| "grad_norm": 3.7275946590748967, | |
| "learning_rate": 8.337835049468517e-06, | |
| "loss": 0.1251, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.8962472406181017, | |
| "grad_norm": 4.290520924678558, | |
| "learning_rate": 8.327150403084105e-06, | |
| "loss": 0.1133, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.9050772626931565, | |
| "grad_norm": 8.977560548421655, | |
| "learning_rate": 8.3164384197047e-06, | |
| "loss": 0.1228, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.9139072847682117, | |
| "grad_norm": 3.655985008304991, | |
| "learning_rate": 8.305699187343586e-06, | |
| "loss": 0.1252, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.922737306843267, | |
| "grad_norm": 3.1746423345857178, | |
| "learning_rate": 8.294932794237936e-06, | |
| "loss": 0.1269, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.931567328918322, | |
| "grad_norm": 5.169537592761684, | |
| "learning_rate": 8.284139328848083e-06, | |
| "loss": 0.1287, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.9403973509933774, | |
| "grad_norm": 6.020170962551394, | |
| "learning_rate": 8.273318879856794e-06, | |
| "loss": 0.125, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.9492273730684326, | |
| "grad_norm": 3.9184909634093397, | |
| "learning_rate": 8.262471536168547e-06, | |
| "loss": 0.1386, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.958057395143488, | |
| "grad_norm": 3.6631131424597103, | |
| "learning_rate": 8.251597386908791e-06, | |
| "loss": 0.127, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.966887417218543, | |
| "grad_norm": 3.70785572766733, | |
| "learning_rate": 8.240696521423221e-06, | |
| "loss": 0.1508, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.9757174392935983, | |
| "grad_norm": 3.1503977758349584, | |
| "learning_rate": 8.229769029277044e-06, | |
| "loss": 0.1286, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.9845474613686536, | |
| "grad_norm": 6.870504017240963, | |
| "learning_rate": 8.218815000254233e-06, | |
| "loss": 0.1427, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.993377483443709, | |
| "grad_norm": 2.768532609030032, | |
| "learning_rate": 8.207834524356804e-06, | |
| "loss": 0.1138, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.0022075055187636, | |
| "grad_norm": 2.8207046730318344, | |
| "learning_rate": 8.196827691804066e-06, | |
| "loss": 0.1205, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.011037527593819, | |
| "grad_norm": 4.402177665136975, | |
| "learning_rate": 8.185794593031889e-06, | |
| "loss": 0.0963, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.019867549668874, | |
| "grad_norm": 2.874087994226284, | |
| "learning_rate": 8.174735318691946e-06, | |
| "loss": 0.0985, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.0286975717439293, | |
| "grad_norm": 2.773091790681681, | |
| "learning_rate": 8.163649959650983e-06, | |
| "loss": 0.095, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.0375275938189845, | |
| "grad_norm": 3.1852082070922703, | |
| "learning_rate": 8.152538606990065e-06, | |
| "loss": 0.0919, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.0463576158940397, | |
| "grad_norm": 4.97982244086657, | |
| "learning_rate": 8.141401352003834e-06, | |
| "loss": 0.0997, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.055187637969095, | |
| "grad_norm": 8.541612200165412, | |
| "learning_rate": 8.130238286199747e-06, | |
| "loss": 0.103, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.06401766004415, | |
| "grad_norm": 4.486284012587778, | |
| "learning_rate": 8.119049501297336e-06, | |
| "loss": 0.1161, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.0728476821192054, | |
| "grad_norm": 3.3537490721423926, | |
| "learning_rate": 8.107835089227446e-06, | |
| "loss": 0.0997, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.0816777041942607, | |
| "grad_norm": 5.17366168346374, | |
| "learning_rate": 8.096595142131491e-06, | |
| "loss": 0.0941, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.0905077262693155, | |
| "grad_norm": 4.052922245836476, | |
| "learning_rate": 8.085329752360683e-06, | |
| "loss": 0.1003, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.0993377483443707, | |
| "grad_norm": 10.466691824028937, | |
| "learning_rate": 8.074039012475277e-06, | |
| "loss": 0.0972, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.108167770419426, | |
| "grad_norm": 6.899695990615844, | |
| "learning_rate": 8.062723015243821e-06, | |
| "loss": 0.1018, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.116997792494481, | |
| "grad_norm": 6.935623054251069, | |
| "learning_rate": 8.051381853642385e-06, | |
| "loss": 0.0959, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.1258278145695364, | |
| "grad_norm": 5.380240033624375, | |
| "learning_rate": 8.04001562085379e-06, | |
| "loss": 0.1009, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.1346578366445916, | |
| "grad_norm": 5.204167174531336, | |
| "learning_rate": 8.028624410266856e-06, | |
| "loss": 0.1044, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.143487858719647, | |
| "grad_norm": 5.267521866348581, | |
| "learning_rate": 8.017208315475633e-06, | |
| "loss": 0.0925, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.152317880794702, | |
| "grad_norm": 6.366799712282832, | |
| "learning_rate": 8.005767430278619e-06, | |
| "loss": 0.1012, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.1611479028697573, | |
| "grad_norm": 3.9450113952746904, | |
| "learning_rate": 7.994301848678006e-06, | |
| "loss": 0.0997, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.1699779249448126, | |
| "grad_norm": 3.1063823479949377, | |
| "learning_rate": 7.982811664878897e-06, | |
| "loss": 0.1008, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.1788079470198674, | |
| "grad_norm": 4.66629360897943, | |
| "learning_rate": 7.971296973288534e-06, | |
| "loss": 0.0905, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.1876379690949226, | |
| "grad_norm": 4.158029256356261, | |
| "learning_rate": 7.959757868515526e-06, | |
| "loss": 0.1065, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.196467991169978, | |
| "grad_norm": 4.761932913825092, | |
| "learning_rate": 7.948194445369065e-06, | |
| "loss": 0.0925, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.205298013245033, | |
| "grad_norm": 3.8759763859212755, | |
| "learning_rate": 7.936606798858154e-06, | |
| "loss": 0.0925, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.2141280353200883, | |
| "grad_norm": 4.532153769879546, | |
| "learning_rate": 7.924995024190825e-06, | |
| "loss": 0.0927, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.2229580573951435, | |
| "grad_norm": 2.8255168649376614, | |
| "learning_rate": 7.91335921677335e-06, | |
| "loss": 0.0979, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.2317880794701987, | |
| "grad_norm": 6.973702691843859, | |
| "learning_rate": 7.901699472209467e-06, | |
| "loss": 0.0883, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.240618101545254, | |
| "grad_norm": 4.809908522267683, | |
| "learning_rate": 7.890015886299587e-06, | |
| "loss": 0.1006, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.249448123620309, | |
| "grad_norm": 4.7958905830128895, | |
| "learning_rate": 7.878308555040012e-06, | |
| "loss": 0.0928, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.258278145695364, | |
| "grad_norm": 2.86896545751829, | |
| "learning_rate": 7.86657757462214e-06, | |
| "loss": 0.0988, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.2671081677704192, | |
| "grad_norm": 3.6073305272465515, | |
| "learning_rate": 7.85482304143168e-06, | |
| "loss": 0.0879, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.2759381898454745, | |
| "grad_norm": 6.14779480895514, | |
| "learning_rate": 7.843045052047863e-06, | |
| "loss": 0.0968, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.2847682119205297, | |
| "grad_norm": 3.576211782488104, | |
| "learning_rate": 7.831243703242636e-06, | |
| "loss": 0.1012, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.293598233995585, | |
| "grad_norm": 4.598562775247985, | |
| "learning_rate": 7.819419091979884e-06, | |
| "loss": 0.0894, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.30242825607064, | |
| "grad_norm": 10.407962558531288, | |
| "learning_rate": 7.807571315414616e-06, | |
| "loss": 0.0998, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.3112582781456954, | |
| "grad_norm": 4.0838619374407354, | |
| "learning_rate": 7.795700470892177e-06, | |
| "loss": 0.1108, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.3200883002207506, | |
| "grad_norm": 3.390233963930005, | |
| "learning_rate": 7.783806655947454e-06, | |
| "loss": 0.0996, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.328918322295806, | |
| "grad_norm": 3.4343376189192925, | |
| "learning_rate": 7.771889968304054e-06, | |
| "loss": 0.1022, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.337748344370861, | |
| "grad_norm": 3.9967675391286464, | |
| "learning_rate": 7.759950505873523e-06, | |
| "loss": 0.0966, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.3465783664459163, | |
| "grad_norm": 3.931184294041522, | |
| "learning_rate": 7.747988366754529e-06, | |
| "loss": 0.1019, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.355408388520971, | |
| "grad_norm": 3.6434067114246695, | |
| "learning_rate": 7.736003649232058e-06, | |
| "loss": 0.101, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.3642384105960264, | |
| "grad_norm": 3.0415486494416197, | |
| "learning_rate": 7.723996451776615e-06, | |
| "loss": 0.095, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.3730684326710816, | |
| "grad_norm": 3.0405884839128516, | |
| "learning_rate": 7.711966873043396e-06, | |
| "loss": 0.0989, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.381898454746137, | |
| "grad_norm": 3.8412175131705473, | |
| "learning_rate": 7.699915011871502e-06, | |
| "loss": 0.1193, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.390728476821192, | |
| "grad_norm": 2.993701017330564, | |
| "learning_rate": 7.687840967283102e-06, | |
| "loss": 0.0947, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.3995584988962473, | |
| "grad_norm": 5.643942446934809, | |
| "learning_rate": 7.675744838482641e-06, | |
| "loss": 0.1089, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.4083885209713025, | |
| "grad_norm": 3.798921749626721, | |
| "learning_rate": 7.66362672485601e-06, | |
| "loss": 0.1111, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.4172185430463577, | |
| "grad_norm": 4.07519615976915, | |
| "learning_rate": 7.651486725969736e-06, | |
| "loss": 0.0995, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.426048565121413, | |
| "grad_norm": 3.109263108741034, | |
| "learning_rate": 7.639324941570165e-06, | |
| "loss": 0.0957, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.4348785871964678, | |
| "grad_norm": 3.966120155328651, | |
| "learning_rate": 7.627141471582635e-06, | |
| "loss": 0.0898, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.443708609271523, | |
| "grad_norm": 5.637058048676252, | |
| "learning_rate": 7.614936416110668e-06, | |
| "loss": 0.093, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.4525386313465782, | |
| "grad_norm": 2.741187454133543, | |
| "learning_rate": 7.6027098754351306e-06, | |
| "loss": 0.1018, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.4613686534216335, | |
| "grad_norm": 3.8393555875989978, | |
| "learning_rate": 7.590461950013424e-06, | |
| "loss": 0.0919, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.4701986754966887, | |
| "grad_norm": 2.919287435009137, | |
| "learning_rate": 7.578192740478656e-06, | |
| "loss": 0.1025, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.479028697571744, | |
| "grad_norm": 3.3130879815381604, | |
| "learning_rate": 7.565902347638806e-06, | |
| "loss": 0.0869, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.487858719646799, | |
| "grad_norm": 3.0285191115691843, | |
| "learning_rate": 7.553590872475909e-06, | |
| "loss": 0.1024, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.4966887417218544, | |
| "grad_norm": 2.6155203207930295, | |
| "learning_rate": 7.541258416145212e-06, | |
| "loss": 0.0897, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.5055187637969096, | |
| "grad_norm": 5.716915177287957, | |
| "learning_rate": 7.528905079974358e-06, | |
| "loss": 0.0958, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.514348785871965, | |
| "grad_norm": 3.3529943708033945, | |
| "learning_rate": 7.5165309654625405e-06, | |
| "loss": 0.0917, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.52317880794702, | |
| "grad_norm": 4.407394855256224, | |
| "learning_rate": 7.504136174279679e-06, | |
| "loss": 0.104, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.5320088300220753, | |
| "grad_norm": 3.141064942786671, | |
| "learning_rate": 7.491720808265576e-06, | |
| "loss": 0.0963, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.54083885209713, | |
| "grad_norm": 3.2198136199963856, | |
| "learning_rate": 7.4792849694290846e-06, | |
| "loss": 0.1089, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.5496688741721854, | |
| "grad_norm": 3.9253303059749833, | |
| "learning_rate": 7.466828759947271e-06, | |
| "loss": 0.0929, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.5584988962472406, | |
| "grad_norm": 4.280334098281072, | |
| "learning_rate": 7.454352282164572e-06, | |
| "loss": 0.0973, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.567328918322296, | |
| "grad_norm": 3.969696029648232, | |
| "learning_rate": 7.441855638591958e-06, | |
| "loss": 0.1006, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.576158940397351, | |
| "grad_norm": 19.883828843473825, | |
| "learning_rate": 7.429338931906085e-06, | |
| "loss": 0.1017, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.5849889624724063, | |
| "grad_norm": 4.198510642360416, | |
| "learning_rate": 7.416802264948455e-06, | |
| "loss": 0.1006, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.5938189845474615, | |
| "grad_norm": 2.822853105650156, | |
| "learning_rate": 7.404245740724573e-06, | |
| "loss": 0.1079, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.6026490066225163, | |
| "grad_norm": 3.353228799417008, | |
| "learning_rate": 7.391669462403096e-06, | |
| "loss": 0.0903, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.6114790286975715, | |
| "grad_norm": 5.576752865540119, | |
| "learning_rate": 7.379073533314988e-06, | |
| "loss": 0.1009, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.6203090507726268, | |
| "grad_norm": 4.017600918158823, | |
| "learning_rate": 7.366458056952668e-06, | |
| "loss": 0.1022, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.629139072847682, | |
| "grad_norm": 3.0289707352387065, | |
| "learning_rate": 7.353823136969167e-06, | |
| "loss": 0.1118, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.6379690949227372, | |
| "grad_norm": 3.2030363014750387, | |
| "learning_rate": 7.341168877177267e-06, | |
| "loss": 0.111, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.6467991169977925, | |
| "grad_norm": 2.8373469235474413, | |
| "learning_rate": 7.328495381548655e-06, | |
| "loss": 0.0984, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.6556291390728477, | |
| "grad_norm": 4.237589768108583, | |
| "learning_rate": 7.315802754213062e-06, | |
| "loss": 0.0928, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.664459161147903, | |
| "grad_norm": 3.293617760716342, | |
| "learning_rate": 7.303091099457418e-06, | |
| "loss": 0.1035, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.673289183222958, | |
| "grad_norm": 3.695166960242193, | |
| "learning_rate": 7.290360521724984e-06, | |
| "loss": 0.1051, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.6821192052980134, | |
| "grad_norm": 3.2546124145541944, | |
| "learning_rate": 7.277611125614499e-06, | |
| "loss": 0.0999, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.6909492273730686, | |
| "grad_norm": 4.98428989250065, | |
| "learning_rate": 7.264843015879321e-06, | |
| "loss": 0.1002, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.699779249448124, | |
| "grad_norm": 6.408155086873857, | |
| "learning_rate": 7.25205629742657e-06, | |
| "loss": 0.1004, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.7086092715231787, | |
| "grad_norm": 11.152466179896583, | |
| "learning_rate": 7.2392510753162516e-06, | |
| "loss": 0.0996, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.717439293598234, | |
| "grad_norm": 3.558825070857552, | |
| "learning_rate": 7.226427454760412e-06, | |
| "loss": 0.1001, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.726269315673289, | |
| "grad_norm": 6.4326311484879, | |
| "learning_rate": 7.213585541122261e-06, | |
| "loss": 0.0899, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.7350993377483444, | |
| "grad_norm": 3.107269880751236, | |
| "learning_rate": 7.200725439915314e-06, | |
| "loss": 0.1054, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.7439293598233996, | |
| "grad_norm": 3.496564737590626, | |
| "learning_rate": 7.187847256802518e-06, | |
| "loss": 0.0861, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 3.752759381898455, | |
| "grad_norm": 3.888379002270393, | |
| "learning_rate": 7.174951097595389e-06, | |
| "loss": 0.1022, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.76158940397351, | |
| "grad_norm": 5.99226737328626, | |
| "learning_rate": 7.162037068253141e-06, | |
| "loss": 0.1046, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 3.7704194260485653, | |
| "grad_norm": 3.7128788164666875, | |
| "learning_rate": 7.149105274881815e-06, | |
| "loss": 0.103, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 3.77924944812362, | |
| "grad_norm": 5.407802305998047, | |
| "learning_rate": 7.136155823733405e-06, | |
| "loss": 0.0967, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.7880794701986753, | |
| "grad_norm": 3.817443194172372, | |
| "learning_rate": 7.12318882120499e-06, | |
| "loss": 0.108, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 3.7969094922737305, | |
| "grad_norm": 3.5868651536922886, | |
| "learning_rate": 7.110204373837857e-06, | |
| "loss": 0.1057, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.8057395143487858, | |
| "grad_norm": 3.2056096939347043, | |
| "learning_rate": 7.097202588316625e-06, | |
| "loss": 0.1156, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 3.814569536423841, | |
| "grad_norm": 2.7349141183729357, | |
| "learning_rate": 7.084183571468368e-06, | |
| "loss": 0.0983, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.8233995584988962, | |
| "grad_norm": 4.246208306345596, | |
| "learning_rate": 7.071147430261738e-06, | |
| "loss": 0.0978, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 3.8322295805739515, | |
| "grad_norm": 5.8984888612881585, | |
| "learning_rate": 7.058094271806091e-06, | |
| "loss": 0.1004, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 3.8410596026490067, | |
| "grad_norm": 3.3105818478273696, | |
| "learning_rate": 7.045024203350598e-06, | |
| "loss": 0.0973, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.849889624724062, | |
| "grad_norm": 4.809955878560633, | |
| "learning_rate": 7.031937332283367e-06, | |
| "loss": 0.0982, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 3.858719646799117, | |
| "grad_norm": 3.4599118483793365, | |
| "learning_rate": 7.018833766130571e-06, | |
| "loss": 0.1121, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 3.8675496688741724, | |
| "grad_norm": 3.962277461297394, | |
| "learning_rate": 7.0057136125555456e-06, | |
| "loss": 0.0954, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.8763796909492276, | |
| "grad_norm": 2.521073351235719, | |
| "learning_rate": 6.9925769793579165e-06, | |
| "loss": 0.082, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 3.8852097130242824, | |
| "grad_norm": 3.669379634426741, | |
| "learning_rate": 6.979423974472714e-06, | |
| "loss": 0.0916, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.8940397350993377, | |
| "grad_norm": 2.810339070177215, | |
| "learning_rate": 6.966254705969484e-06, | |
| "loss": 0.0971, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 3.902869757174393, | |
| "grad_norm": 3.3113163980254345, | |
| "learning_rate": 6.953069282051397e-06, | |
| "loss": 0.0984, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 3.911699779249448, | |
| "grad_norm": 2.873394765233237, | |
| "learning_rate": 6.939867811054365e-06, | |
| "loss": 0.0835, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 3.9205298013245033, | |
| "grad_norm": 6.471682111167088, | |
| "learning_rate": 6.9266504014461425e-06, | |
| "loss": 0.0947, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 3.9293598233995586, | |
| "grad_norm": 4.5132940160523125, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.1179, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 3.938189845474614, | |
| "grad_norm": 4.273369420328601, | |
| "learning_rate": 6.900168200921065e-06, | |
| "loss": 0.0941, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 3.9470198675496686, | |
| "grad_norm": 3.136852814569383, | |
| "learning_rate": 6.886903627590938e-06, | |
| "loss": 0.0903, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 3.955849889624724, | |
| "grad_norm": 24.428120897543273, | |
| "learning_rate": 6.8736235508213024e-06, | |
| "loss": 0.0996, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 3.964679911699779, | |
| "grad_norm": 2.5560553413107474, | |
| "learning_rate": 6.860328079725764e-06, | |
| "loss": 0.0988, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 3.9735099337748343, | |
| "grad_norm": 5.943985947216547, | |
| "learning_rate": 6.84701732354442e-06, | |
| "loss": 0.0984, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.9823399558498895, | |
| "grad_norm": 3.765915925509269, | |
| "learning_rate": 6.8336913916429515e-06, | |
| "loss": 0.0944, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 3.9911699779249448, | |
| "grad_norm": 3.0838860947525606, | |
| "learning_rate": 6.820350393511732e-06, | |
| "loss": 0.0919, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.258262477534644, | |
| "learning_rate": 6.806994438764922e-06, | |
| "loss": 0.1015, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.008830022075055, | |
| "grad_norm": 2.487600464960827, | |
| "learning_rate": 6.79362363713957e-06, | |
| "loss": 0.0811, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.0176600441501105, | |
| "grad_norm": 2.704929925231827, | |
| "learning_rate": 6.780238098494711e-06, | |
| "loss": 0.0747, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.026490066225166, | |
| "grad_norm": 4.590794377813223, | |
| "learning_rate": 6.766837932810468e-06, | |
| "loss": 0.0773, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.035320088300221, | |
| "grad_norm": 3.694952643706577, | |
| "learning_rate": 6.7534232501871425e-06, | |
| "loss": 0.0839, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 4.044150110375276, | |
| "grad_norm": 6.619965935098318, | |
| "learning_rate": 6.7399941608443096e-06, | |
| "loss": 0.0849, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.052980132450331, | |
| "grad_norm": 3.2744763933991017, | |
| "learning_rate": 6.72655077511992e-06, | |
| "loss": 0.0814, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 4.061810154525387, | |
| "grad_norm": 2.2788263144306162, | |
| "learning_rate": 6.713093203469384e-06, | |
| "loss": 0.073, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.070640176600442, | |
| "grad_norm": 2.7293401871470646, | |
| "learning_rate": 6.6996215564646705e-06, | |
| "loss": 0.0833, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 4.079470198675497, | |
| "grad_norm": 2.7878106813417296, | |
| "learning_rate": 6.686135944793395e-06, | |
| "loss": 0.0764, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.088300220750552, | |
| "grad_norm": 3.4508666937690347, | |
| "learning_rate": 6.672636479257912e-06, | |
| "loss": 0.0685, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 4.097130242825607, | |
| "grad_norm": 3.2939345109347338, | |
| "learning_rate": 6.659123270774406e-06, | |
| "loss": 0.0849, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.105960264900662, | |
| "grad_norm": 3.760887373221355, | |
| "learning_rate": 6.645596430371976e-06, | |
| "loss": 0.0808, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 4.114790286975717, | |
| "grad_norm": 3.2597140513855005, | |
| "learning_rate": 6.632056069191723e-06, | |
| "loss": 0.0797, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.123620309050772, | |
| "grad_norm": 6.450410996460324, | |
| "learning_rate": 6.618502298485844e-06, | |
| "loss": 0.0861, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.132450331125828, | |
| "grad_norm": 3.3573492020873656, | |
| "learning_rate": 6.604935229616711e-06, | |
| "loss": 0.0782, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.141280353200883, | |
| "grad_norm": 6.126384157468954, | |
| "learning_rate": 6.5913549740559606e-06, | |
| "loss": 0.0821, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.150110375275938, | |
| "grad_norm": 3.438150104347235, | |
| "learning_rate": 6.57776164338357e-06, | |
| "loss": 0.082, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.158940397350993, | |
| "grad_norm": 2.9135640923053896, | |
| "learning_rate": 6.564155349286952e-06, | |
| "loss": 0.0776, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.1677704194260485, | |
| "grad_norm": 2.9999523172662523, | |
| "learning_rate": 6.550536203560028e-06, | |
| "loss": 0.0699, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.176600441501104, | |
| "grad_norm": 3.746386815870089, | |
| "learning_rate": 6.536904318102314e-06, | |
| "loss": 0.0769, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.185430463576159, | |
| "grad_norm": 2.9718529127994286, | |
| "learning_rate": 6.523259804918001e-06, | |
| "loss": 0.0837, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.194260485651214, | |
| "grad_norm": 3.474007331109214, | |
| "learning_rate": 6.509602776115029e-06, | |
| "loss": 0.0755, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.2030905077262695, | |
| "grad_norm": 3.8851000736196526, | |
| "learning_rate": 6.4959333439041775e-06, | |
| "loss": 0.0774, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.211920529801325, | |
| "grad_norm": 3.3949435150028924, | |
| "learning_rate": 6.482251620598129e-06, | |
| "loss": 0.0694, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.22075055187638, | |
| "grad_norm": 4.26273289010399, | |
| "learning_rate": 6.4685577186105595e-06, | |
| "loss": 0.0778, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.229580573951435, | |
| "grad_norm": 3.1626694349683224, | |
| "learning_rate": 6.45485175045521e-06, | |
| "loss": 0.0799, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.23841059602649, | |
| "grad_norm": 3.2387645523490214, | |
| "learning_rate": 6.441133828744954e-06, | |
| "loss": 0.0817, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.247240618101546, | |
| "grad_norm": 3.6360260747998914, | |
| "learning_rate": 6.427404066190889e-06, | |
| "loss": 0.0841, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.256070640176601, | |
| "grad_norm": 2.5866840173292402, | |
| "learning_rate": 6.413662575601391e-06, | |
| "loss": 0.079, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.264900662251655, | |
| "grad_norm": 2.7677550181684603, | |
| "learning_rate": 6.3999094698812055e-06, | |
| "loss": 0.0815, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 4.27373068432671, | |
| "grad_norm": 3.16309495676388, | |
| "learning_rate": 6.386144862030508e-06, | |
| "loss": 0.0647, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.282560706401766, | |
| "grad_norm": 2.4645179267970736, | |
| "learning_rate": 6.3723688651439806e-06, | |
| "loss": 0.0846, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.291390728476821, | |
| "grad_norm": 6.38903880108792, | |
| "learning_rate": 6.358581592409881e-06, | |
| "loss": 0.0731, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.300220750551876, | |
| "grad_norm": 3.0558384340208686, | |
| "learning_rate": 6.344783157109114e-06, | |
| "loss": 0.0747, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 4.309050772626931, | |
| "grad_norm": 2.9983291623446764, | |
| "learning_rate": 6.3309736726142965e-06, | |
| "loss": 0.075, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.317880794701987, | |
| "grad_norm": 3.2690342947361493, | |
| "learning_rate": 6.317153252388834e-06, | |
| "loss": 0.0867, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 4.326710816777042, | |
| "grad_norm": 3.610248860150095, | |
| "learning_rate": 6.303322009985984e-06, | |
| "loss": 0.0724, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.335540838852097, | |
| "grad_norm": 3.8688332246602752, | |
| "learning_rate": 6.289480059047915e-06, | |
| "loss": 0.0753, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 4.344370860927152, | |
| "grad_norm": 7.441887079342132, | |
| "learning_rate": 6.27562751330479e-06, | |
| "loss": 0.0842, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.3532008830022075, | |
| "grad_norm": 2.9839561092138718, | |
| "learning_rate": 6.261764486573816e-06, | |
| "loss": 0.079, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 4.362030905077263, | |
| "grad_norm": 4.200459499442376, | |
| "learning_rate": 6.247891092758319e-06, | |
| "loss": 0.0736, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.370860927152318, | |
| "grad_norm": 3.4070365441270467, | |
| "learning_rate": 6.2340074458468014e-06, | |
| "loss": 0.0806, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.379690949227373, | |
| "grad_norm": 3.441955046739519, | |
| "learning_rate": 6.220113659912012e-06, | |
| "loss": 0.0731, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.3885209713024285, | |
| "grad_norm": 3.1561631964440076, | |
| "learning_rate": 6.206209849110001e-06, | |
| "loss": 0.0758, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 4.397350993377484, | |
| "grad_norm": 3.624346507958278, | |
| "learning_rate": 6.1922961276791925e-06, | |
| "loss": 0.0794, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.406181015452539, | |
| "grad_norm": 6.080763290404623, | |
| "learning_rate": 6.1783726099394324e-06, | |
| "loss": 0.0804, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 4.415011037527594, | |
| "grad_norm": 3.1460557464112293, | |
| "learning_rate": 6.164439410291061e-06, | |
| "loss": 0.0786, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.423841059602649, | |
| "grad_norm": 4.32533416371131, | |
| "learning_rate": 6.150496643213969e-06, | |
| "loss": 0.0813, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 4.432671081677704, | |
| "grad_norm": 4.384447144192398, | |
| "learning_rate": 6.136544423266651e-06, | |
| "loss": 0.078, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 4.441501103752759, | |
| "grad_norm": 3.527942957160671, | |
| "learning_rate": 6.122582865085278e-06, | |
| "loss": 0.0843, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 4.450331125827814, | |
| "grad_norm": 3.1467040214710527, | |
| "learning_rate": 6.108612083382739e-06, | |
| "loss": 0.0689, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 4.459161147902869, | |
| "grad_norm": 5.934612259061866, | |
| "learning_rate": 6.094632192947711e-06, | |
| "loss": 0.0806, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.467991169977925, | |
| "grad_norm": 3.195416678057104, | |
| "learning_rate": 6.08064330864371e-06, | |
| "loss": 0.0841, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 4.47682119205298, | |
| "grad_norm": 5.946520342587703, | |
| "learning_rate": 6.066645545408149e-06, | |
| "loss": 0.0832, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 4.485651214128035, | |
| "grad_norm": 2.3820570403659693, | |
| "learning_rate": 6.052639018251394e-06, | |
| "loss": 0.0829, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 4.49448123620309, | |
| "grad_norm": 3.115965495879289, | |
| "learning_rate": 6.03862384225582e-06, | |
| "loss": 0.0741, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 4.503311258278146, | |
| "grad_norm": 4.214892939498798, | |
| "learning_rate": 6.024600132574855e-06, | |
| "loss": 0.0742, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.512141280353201, | |
| "grad_norm": 4.138254792265556, | |
| "learning_rate": 6.010568004432055e-06, | |
| "loss": 0.0825, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 4.520971302428256, | |
| "grad_norm": 2.729416010034935, | |
| "learning_rate": 5.9965275731201364e-06, | |
| "loss": 0.0808, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 4.529801324503311, | |
| "grad_norm": 13.118641333902108, | |
| "learning_rate": 5.982478954000042e-06, | |
| "loss": 0.08, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 4.5386313465783665, | |
| "grad_norm": 2.335880462238545, | |
| "learning_rate": 5.968422262499983e-06, | |
| "loss": 0.0739, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 4.547461368653422, | |
| "grad_norm": 3.6375779968448496, | |
| "learning_rate": 5.9543576141145035e-06, | |
| "loss": 0.0622, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.556291390728477, | |
| "grad_norm": 3.6160696452303984, | |
| "learning_rate": 5.940285124403517e-06, | |
| "loss": 0.0784, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 4.565121412803532, | |
| "grad_norm": 3.3711175360832777, | |
| "learning_rate": 5.926204908991366e-06, | |
| "loss": 0.0735, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 4.5739514348785875, | |
| "grad_norm": 3.2269124481467104, | |
| "learning_rate": 5.912117083565874e-06, | |
| "loss": 0.0791, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 4.582781456953643, | |
| "grad_norm": 5.900071692486368, | |
| "learning_rate": 5.898021763877388e-06, | |
| "loss": 0.0914, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 4.591611479028698, | |
| "grad_norm": 4.418513990528456, | |
| "learning_rate": 5.883919065737827e-06, | |
| "loss": 0.081, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.600441501103752, | |
| "grad_norm": 3.3722000807665866, | |
| "learning_rate": 5.869809105019738e-06, | |
| "loss": 0.0744, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 4.609271523178808, | |
| "grad_norm": 7.469587100913114, | |
| "learning_rate": 5.85569199765534e-06, | |
| "loss": 0.0723, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 4.618101545253863, | |
| "grad_norm": 2.9175696518893486, | |
| "learning_rate": 5.841567859635572e-06, | |
| "loss": 0.0728, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 4.626931567328918, | |
| "grad_norm": 4.026258728317068, | |
| "learning_rate": 5.827436807009133e-06, | |
| "loss": 0.0765, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 4.635761589403973, | |
| "grad_norm": 4.023701049292052, | |
| "learning_rate": 5.813298955881542e-06, | |
| "loss": 0.0795, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.644591611479028, | |
| "grad_norm": 7.492669294902294, | |
| "learning_rate": 5.799154422414174e-06, | |
| "loss": 0.0832, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 4.653421633554084, | |
| "grad_norm": 27.135072765679574, | |
| "learning_rate": 5.785003322823307e-06, | |
| "loss": 0.0641, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 4.662251655629139, | |
| "grad_norm": 2.1533625604567357, | |
| "learning_rate": 5.7708457733791715e-06, | |
| "loss": 0.0773, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 4.671081677704194, | |
| "grad_norm": 2.9620640076509615, | |
| "learning_rate": 5.756681890404987e-06, | |
| "loss": 0.0853, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 4.679911699779249, | |
| "grad_norm": 4.678163808609413, | |
| "learning_rate": 5.7425117902760195e-06, | |
| "loss": 0.0713, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.688741721854305, | |
| "grad_norm": 4.296864154061136, | |
| "learning_rate": 5.72833558941861e-06, | |
| "loss": 0.0859, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 4.69757174392936, | |
| "grad_norm": 2.922605079813754, | |
| "learning_rate": 5.714153404309228e-06, | |
| "loss": 0.078, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 4.706401766004415, | |
| "grad_norm": 3.9382223946197015, | |
| "learning_rate": 5.6999653514735124e-06, | |
| "loss": 0.071, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 4.71523178807947, | |
| "grad_norm": 3.147586504010664, | |
| "learning_rate": 5.685771547485312e-06, | |
| "loss": 0.0813, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 4.7240618101545255, | |
| "grad_norm": 2.966590218091343, | |
| "learning_rate": 5.671572108965729e-06, | |
| "loss": 0.0751, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.732891832229581, | |
| "grad_norm": 3.2622534882340446, | |
| "learning_rate": 5.65736715258216e-06, | |
| "loss": 0.0753, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 4.741721854304636, | |
| "grad_norm": 2.3211945267996716, | |
| "learning_rate": 5.643156795047343e-06, | |
| "loss": 0.0769, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 4.750551876379691, | |
| "grad_norm": 8.119722637541681, | |
| "learning_rate": 5.628941153118388e-06, | |
| "loss": 0.0822, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 4.7593818984547465, | |
| "grad_norm": 8.441365529225875, | |
| "learning_rate": 5.6147203435958246e-06, | |
| "loss": 0.0829, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 4.768211920529802, | |
| "grad_norm": 2.451783795869697, | |
| "learning_rate": 5.600494483322643e-06, | |
| "loss": 0.0763, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.777041942604857, | |
| "grad_norm": 2.991753091402444, | |
| "learning_rate": 5.586263689183332e-06, | |
| "loss": 0.0741, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 4.785871964679911, | |
| "grad_norm": 3.539488911553519, | |
| "learning_rate": 5.572028078102917e-06, | |
| "loss": 0.0682, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 4.7947019867549665, | |
| "grad_norm": 5.018439865551765, | |
| "learning_rate": 5.557787767046001e-06, | |
| "loss": 0.084, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 4.803532008830022, | |
| "grad_norm": 2.8442572450341257, | |
| "learning_rate": 5.543542873015806e-06, | |
| "loss": 0.0769, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 4.812362030905077, | |
| "grad_norm": 2.7357924462195466, | |
| "learning_rate": 5.529293513053207e-06, | |
| "loss": 0.0661, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.821192052980132, | |
| "grad_norm": 4.7717542869998475, | |
| "learning_rate": 5.515039804235772e-06, | |
| "loss": 0.0797, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 4.830022075055187, | |
| "grad_norm": 2.361311252969642, | |
| "learning_rate": 5.5007818636768055e-06, | |
| "loss": 0.0756, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 4.838852097130243, | |
| "grad_norm": 3.6882765130368793, | |
| "learning_rate": 5.486519808524374e-06, | |
| "loss": 0.0798, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 4.847682119205298, | |
| "grad_norm": 3.9327149256112035, | |
| "learning_rate": 5.472253755960358e-06, | |
| "loss": 0.0808, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 4.856512141280353, | |
| "grad_norm": 3.5572765179325034, | |
| "learning_rate": 5.457983823199475e-06, | |
| "loss": 0.0772, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.865342163355408, | |
| "grad_norm": 3.0090033893509402, | |
| "learning_rate": 5.443710127488331e-06, | |
| "loss": 0.0887, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 4.874172185430464, | |
| "grad_norm": 4.076041046123783, | |
| "learning_rate": 5.429432786104446e-06, | |
| "loss": 0.0901, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 4.883002207505519, | |
| "grad_norm": 5.244647456159231, | |
| "learning_rate": 5.415151916355292e-06, | |
| "loss": 0.0786, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 4.891832229580574, | |
| "grad_norm": 2.8707489082851136, | |
| "learning_rate": 5.400867635577335e-06, | |
| "loss": 0.0639, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 4.900662251655629, | |
| "grad_norm": 6.188907034334066, | |
| "learning_rate": 5.3865800611350634e-06, | |
| "loss": 0.0803, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 4.9094922737306845, | |
| "grad_norm": 3.791444437800692, | |
| "learning_rate": 5.372289310420032e-06, | |
| "loss": 0.0823, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 4.91832229580574, | |
| "grad_norm": 3.383269845290552, | |
| "learning_rate": 5.35799550084989e-06, | |
| "loss": 0.0894, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 4.927152317880795, | |
| "grad_norm": 3.564403614663905, | |
| "learning_rate": 5.343698749867421e-06, | |
| "loss": 0.0702, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 4.93598233995585, | |
| "grad_norm": 3.2097656260996086, | |
| "learning_rate": 5.329399174939572e-06, | |
| "loss": 0.0725, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 4.9448123620309055, | |
| "grad_norm": 2.796838877718886, | |
| "learning_rate": 5.315096893556497e-06, | |
| "loss": 0.0829, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.95364238410596, | |
| "grad_norm": 2.6046376261906548, | |
| "learning_rate": 5.300792023230587e-06, | |
| "loss": 0.0813, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 4.962472406181016, | |
| "grad_norm": 2.5662770693138475, | |
| "learning_rate": 5.2864846814955e-06, | |
| "loss": 0.0676, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 4.97130242825607, | |
| "grad_norm": 2.5457460369187963, | |
| "learning_rate": 5.272174985905207e-06, | |
| "loss": 0.0815, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 4.9801324503311255, | |
| "grad_norm": 3.023794300410492, | |
| "learning_rate": 5.257863054033012e-06, | |
| "loss": 0.078, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 4.988962472406181, | |
| "grad_norm": 3.422355333754827, | |
| "learning_rate": 5.243549003470599e-06, | |
| "loss": 0.0823, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 4.997792494481236, | |
| "grad_norm": 2.7100217336415446, | |
| "learning_rate": 5.229232951827054e-06, | |
| "loss": 0.0793, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 5.006622516556291, | |
| "grad_norm": 2.9333502511024467, | |
| "learning_rate": 5.2149150167279106e-06, | |
| "loss": 0.0636, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 5.015452538631346, | |
| "grad_norm": 2.0589587062649177, | |
| "learning_rate": 5.200595315814174e-06, | |
| "loss": 0.0548, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 5.024282560706402, | |
| "grad_norm": 2.525518229510344, | |
| "learning_rate": 5.18627396674136e-06, | |
| "loss": 0.0591, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 5.033112582781457, | |
| "grad_norm": 2.5607980347948427, | |
| "learning_rate": 5.17195108717852e-06, | |
| "loss": 0.0591, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 5.041942604856512, | |
| "grad_norm": 3.5828058180185227, | |
| "learning_rate": 5.15762679480729e-06, | |
| "loss": 0.06, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 5.050772626931567, | |
| "grad_norm": 4.618347231171923, | |
| "learning_rate": 5.143301207320909e-06, | |
| "loss": 0.0684, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 5.059602649006623, | |
| "grad_norm": 3.274180106367918, | |
| "learning_rate": 5.128974442423254e-06, | |
| "loss": 0.0647, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 5.068432671081678, | |
| "grad_norm": 2.6041223294383395, | |
| "learning_rate": 5.114646617827884e-06, | |
| "loss": 0.0541, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 5.077262693156733, | |
| "grad_norm": 2.4320109870435465, | |
| "learning_rate": 5.100317851257057e-06, | |
| "loss": 0.0565, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 5.086092715231788, | |
| "grad_norm": 3.4063716575895966, | |
| "learning_rate": 5.085988260440776e-06, | |
| "loss": 0.0527, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 5.0949227373068435, | |
| "grad_norm": 2.668641717499427, | |
| "learning_rate": 5.0716579631158124e-06, | |
| "loss": 0.0535, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 5.103752759381899, | |
| "grad_norm": 4.031985013996196, | |
| "learning_rate": 5.057327077024745e-06, | |
| "loss": 0.0578, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 5.112582781456954, | |
| "grad_norm": 5.842387157699151, | |
| "learning_rate": 5.0429957199149905e-06, | |
| "loss": 0.0665, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 5.121412803532009, | |
| "grad_norm": 3.169377757087287, | |
| "learning_rate": 5.028664009537835e-06, | |
| "loss": 0.062, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 5.1302428256070645, | |
| "grad_norm": 4.058506270727854, | |
| "learning_rate": 5.014332063647462e-06, | |
| "loss": 0.0671, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 5.139072847682119, | |
| "grad_norm": 5.591261528615111, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0623, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 5.147902869757174, | |
| "grad_norm": 4.048689762276835, | |
| "learning_rate": 4.985667936352538e-06, | |
| "loss": 0.0613, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 5.156732891832229, | |
| "grad_norm": 3.2808596186674586, | |
| "learning_rate": 4.971335990462168e-06, | |
| "loss": 0.0654, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 5.1655629139072845, | |
| "grad_norm": 14.727042643318551, | |
| "learning_rate": 4.95700428008501e-06, | |
| "loss": 0.0654, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 5.17439293598234, | |
| "grad_norm": 2.5426941014392885, | |
| "learning_rate": 4.942672922975255e-06, | |
| "loss": 0.0686, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 5.183222958057395, | |
| "grad_norm": 3.227106789070159, | |
| "learning_rate": 4.928342036884189e-06, | |
| "loss": 0.0665, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 5.19205298013245, | |
| "grad_norm": 2.3563330632848296, | |
| "learning_rate": 4.914011739559225e-06, | |
| "loss": 0.0564, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 5.200883002207505, | |
| "grad_norm": 2.907564408294844, | |
| "learning_rate": 4.899682148742944e-06, | |
| "loss": 0.0563, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 5.209713024282561, | |
| "grad_norm": 5.070972886177633, | |
| "learning_rate": 4.8853533821721175e-06, | |
| "loss": 0.0647, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.218543046357616, | |
| "grad_norm": 2.797286246322686, | |
| "learning_rate": 4.871025557576747e-06, | |
| "loss": 0.0573, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 5.227373068432671, | |
| "grad_norm": 5.185940824607066, | |
| "learning_rate": 4.8566987926790946e-06, | |
| "loss": 0.0508, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 5.236203090507726, | |
| "grad_norm": 2.1389405609805503, | |
| "learning_rate": 4.8423732051927115e-06, | |
| "loss": 0.0638, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 5.245033112582782, | |
| "grad_norm": 5.257867292451715, | |
| "learning_rate": 4.82804891282148e-06, | |
| "loss": 0.0564, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 5.253863134657837, | |
| "grad_norm": 2.266351603099357, | |
| "learning_rate": 4.813726033258643e-06, | |
| "loss": 0.0571, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 5.262693156732892, | |
| "grad_norm": 2.1944338348449737, | |
| "learning_rate": 4.799404684185828e-06, | |
| "loss": 0.0587, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 5.271523178807947, | |
| "grad_norm": 2.6901167546355236, | |
| "learning_rate": 4.78508498327209e-06, | |
| "loss": 0.0607, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 5.2803532008830025, | |
| "grad_norm": 2.2184857872398416, | |
| "learning_rate": 4.770767048172948e-06, | |
| "loss": 0.0543, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 5.289183222958058, | |
| "grad_norm": 2.0863728960818406, | |
| "learning_rate": 4.756450996529403e-06, | |
| "loss": 0.0548, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 5.298013245033113, | |
| "grad_norm": 2.703484755869529, | |
| "learning_rate": 4.742136945966991e-06, | |
| "loss": 0.0574, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.306843267108167, | |
| "grad_norm": 5.97260271492099, | |
| "learning_rate": 4.727825014094795e-06, | |
| "loss": 0.0672, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 5.315673289183223, | |
| "grad_norm": 3.851667211093075, | |
| "learning_rate": 4.713515318504501e-06, | |
| "loss": 0.0606, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 5.324503311258278, | |
| "grad_norm": 2.9899659713595788, | |
| "learning_rate": 4.699207976769416e-06, | |
| "loss": 0.0587, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 4.97742887290687, | |
| "learning_rate": 4.684903106443504e-06, | |
| "loss": 0.06, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 5.342163355408388, | |
| "grad_norm": 3.707184682144467, | |
| "learning_rate": 4.670600825060429e-06, | |
| "loss": 0.061, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 5.3509933774834435, | |
| "grad_norm": 3.6916199658567215, | |
| "learning_rate": 4.656301250132581e-06, | |
| "loss": 0.0616, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 5.359823399558499, | |
| "grad_norm": 4.400818586786533, | |
| "learning_rate": 4.6420044991501104e-06, | |
| "loss": 0.0595, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 5.368653421633554, | |
| "grad_norm": 3.371746830193681, | |
| "learning_rate": 4.627710689579968e-06, | |
| "loss": 0.0564, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 5.377483443708609, | |
| "grad_norm": 4.801597687377842, | |
| "learning_rate": 4.613419938864937e-06, | |
| "loss": 0.0637, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 5.386313465783664, | |
| "grad_norm": 5.4024251876154485, | |
| "learning_rate": 4.599132364422666e-06, | |
| "loss": 0.0593, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.39514348785872, | |
| "grad_norm": 4.504605782454792, | |
| "learning_rate": 4.58484808364471e-06, | |
| "loss": 0.0595, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 5.403973509933775, | |
| "grad_norm": 2.961163622235218, | |
| "learning_rate": 4.570567213895555e-06, | |
| "loss": 0.0653, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 5.41280353200883, | |
| "grad_norm": 3.249541622380529, | |
| "learning_rate": 4.556289872511669e-06, | |
| "loss": 0.067, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 5.421633554083885, | |
| "grad_norm": 3.2924656864763606, | |
| "learning_rate": 4.542016176800527e-06, | |
| "loss": 0.0632, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 5.430463576158941, | |
| "grad_norm": 4.428423801068884, | |
| "learning_rate": 4.527746244039644e-06, | |
| "loss": 0.0649, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 5.439293598233996, | |
| "grad_norm": 2.812339137274197, | |
| "learning_rate": 4.513480191475627e-06, | |
| "loss": 0.0608, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 5.448123620309051, | |
| "grad_norm": 2.8118735950822247, | |
| "learning_rate": 4.499218136323197e-06, | |
| "loss": 0.0575, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 5.456953642384106, | |
| "grad_norm": 3.06477145300682, | |
| "learning_rate": 4.4849601957642295e-06, | |
| "loss": 0.0644, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 5.4657836644591615, | |
| "grad_norm": 2.3243117309542356, | |
| "learning_rate": 4.470706486946797e-06, | |
| "loss": 0.0574, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 5.474613686534216, | |
| "grad_norm": 3.2891319383531683, | |
| "learning_rate": 4.456457126984196e-06, | |
| "loss": 0.0574, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.483443708609272, | |
| "grad_norm": 4.309593005252989, | |
| "learning_rate": 4.4422122329539996e-06, | |
| "loss": 0.0587, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 5.492273730684326, | |
| "grad_norm": 2.85722651778961, | |
| "learning_rate": 4.427971921897086e-06, | |
| "loss": 0.0548, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 5.501103752759382, | |
| "grad_norm": 2.734709311715621, | |
| "learning_rate": 4.413736310816669e-06, | |
| "loss": 0.0617, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 5.509933774834437, | |
| "grad_norm": 3.117299230497021, | |
| "learning_rate": 4.399505516677358e-06, | |
| "loss": 0.059, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 5.518763796909492, | |
| "grad_norm": 2.3050646384088243, | |
| "learning_rate": 4.385279656404178e-06, | |
| "loss": 0.0503, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 5.527593818984547, | |
| "grad_norm": 2.3459181641244973, | |
| "learning_rate": 4.371058846881614e-06, | |
| "loss": 0.0577, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 5.5364238410596025, | |
| "grad_norm": 2.496472650047635, | |
| "learning_rate": 4.356843204952657e-06, | |
| "loss": 0.0547, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 5.545253863134658, | |
| "grad_norm": 2.071427451636752, | |
| "learning_rate": 4.3426328474178405e-06, | |
| "loss": 0.0549, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 5.554083885209713, | |
| "grad_norm": 2.1923973923593527, | |
| "learning_rate": 4.328427891034273e-06, | |
| "loss": 0.0525, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 5.562913907284768, | |
| "grad_norm": 3.111741941666927, | |
| "learning_rate": 4.3142284525146915e-06, | |
| "loss": 0.0556, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.571743929359823, | |
| "grad_norm": 2.7566511531837614, | |
| "learning_rate": 4.300034648526489e-06, | |
| "loss": 0.0541, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 5.580573951434879, | |
| "grad_norm": 2.986680596223554, | |
| "learning_rate": 4.2858465956907726e-06, | |
| "loss": 0.0599, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 5.589403973509934, | |
| "grad_norm": 3.586357046609581, | |
| "learning_rate": 4.271664410581392e-06, | |
| "loss": 0.0597, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 5.598233995584989, | |
| "grad_norm": 2.6007768970135663, | |
| "learning_rate": 4.257488209723981e-06, | |
| "loss": 0.056, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 5.607064017660044, | |
| "grad_norm": 3.409352912881948, | |
| "learning_rate": 4.243318109595014e-06, | |
| "loss": 0.0615, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 5.6158940397351, | |
| "grad_norm": 2.586344873890741, | |
| "learning_rate": 4.229154226620832e-06, | |
| "loss": 0.0571, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 5.624724061810155, | |
| "grad_norm": 3.921062187367027, | |
| "learning_rate": 4.2149966771766945e-06, | |
| "loss": 0.0597, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 5.63355408388521, | |
| "grad_norm": 2.5356425300437127, | |
| "learning_rate": 4.200845577585827e-06, | |
| "loss": 0.0583, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 5.642384105960264, | |
| "grad_norm": 2.3279147591454956, | |
| "learning_rate": 4.186701044118459e-06, | |
| "loss": 0.0548, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 5.6512141280353205, | |
| "grad_norm": 2.5923385633306646, | |
| "learning_rate": 4.1725631929908684e-06, | |
| "loss": 0.0548, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.660044150110375, | |
| "grad_norm": 2.9259179557765584, | |
| "learning_rate": 4.158432140364431e-06, | |
| "loss": 0.0569, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 5.66887417218543, | |
| "grad_norm": 2.4472802188382663, | |
| "learning_rate": 4.1443080023446605e-06, | |
| "loss": 0.0599, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 5.677704194260485, | |
| "grad_norm": 3.2062058758996783, | |
| "learning_rate": 4.130190894980262e-06, | |
| "loss": 0.0634, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 5.686534216335541, | |
| "grad_norm": 5.42270104288098, | |
| "learning_rate": 4.116080934262175e-06, | |
| "loss": 0.0571, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 5.695364238410596, | |
| "grad_norm": 2.5896418417547977, | |
| "learning_rate": 4.101978236122613e-06, | |
| "loss": 0.0626, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 5.704194260485651, | |
| "grad_norm": 2.7665877644799353, | |
| "learning_rate": 4.087882916434126e-06, | |
| "loss": 0.0659, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 5.713024282560706, | |
| "grad_norm": 4.599495908781595, | |
| "learning_rate": 4.0737950910086354e-06, | |
| "loss": 0.0672, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 5.7218543046357615, | |
| "grad_norm": 2.94462656426052, | |
| "learning_rate": 4.059714875596486e-06, | |
| "loss": 0.0575, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 5.730684326710817, | |
| "grad_norm": 4.065245430121945, | |
| "learning_rate": 4.045642385885497e-06, | |
| "loss": 0.0557, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 5.739514348785872, | |
| "grad_norm": 3.4049333346217736, | |
| "learning_rate": 4.0315777375000185e-06, | |
| "loss": 0.0597, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.748344370860927, | |
| "grad_norm": 2.9741565355682145, | |
| "learning_rate": 4.017521045999961e-06, | |
| "loss": 0.0547, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 5.757174392935982, | |
| "grad_norm": 2.8839101833352956, | |
| "learning_rate": 4.003472426879866e-06, | |
| "loss": 0.055, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 5.766004415011038, | |
| "grad_norm": 2.278214477282001, | |
| "learning_rate": 3.989431995567947e-06, | |
| "loss": 0.0511, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 5.774834437086093, | |
| "grad_norm": 3.756698727631742, | |
| "learning_rate": 3.975399867425146e-06, | |
| "loss": 0.0579, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 5.783664459161148, | |
| "grad_norm": 3.649024366025531, | |
| "learning_rate": 3.961376157744183e-06, | |
| "loss": 0.0639, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 5.792494481236203, | |
| "grad_norm": 3.3760620252813456, | |
| "learning_rate": 3.947360981748607e-06, | |
| "loss": 0.0656, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 5.801324503311259, | |
| "grad_norm": 2.4829829222682376, | |
| "learning_rate": 3.933354454591851e-06, | |
| "loss": 0.0572, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 5.810154525386314, | |
| "grad_norm": 5.466130917091564, | |
| "learning_rate": 3.9193566913562915e-06, | |
| "loss": 0.0577, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 5.818984547461369, | |
| "grad_norm": 2.996268788176131, | |
| "learning_rate": 3.9053678070522904e-06, | |
| "loss": 0.0589, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 5.827814569536423, | |
| "grad_norm": 2.7141967033954066, | |
| "learning_rate": 3.891387916617261e-06, | |
| "loss": 0.0596, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.8366445916114795, | |
| "grad_norm": 2.6052802271605686, | |
| "learning_rate": 3.877417134914724e-06, | |
| "loss": 0.0569, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 5.845474613686534, | |
| "grad_norm": 3.133859570438111, | |
| "learning_rate": 3.863455576733349e-06, | |
| "loss": 0.056, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 5.854304635761589, | |
| "grad_norm": 10.873722442866486, | |
| "learning_rate": 3.849503356786034e-06, | |
| "loss": 0.0584, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 5.863134657836644, | |
| "grad_norm": 3.652069793001449, | |
| "learning_rate": 3.83556058970894e-06, | |
| "loss": 0.0681, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 5.8719646799117, | |
| "grad_norm": 3.0068993319937567, | |
| "learning_rate": 3.821627390060568e-06, | |
| "loss": 0.0575, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 5.880794701986755, | |
| "grad_norm": 3.015486680749582, | |
| "learning_rate": 3.807703872320809e-06, | |
| "loss": 0.0661, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 5.88962472406181, | |
| "grad_norm": 3.1884013214321345, | |
| "learning_rate": 3.79379015089e-06, | |
| "loss": 0.0656, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 5.898454746136865, | |
| "grad_norm": 3.8980532887343657, | |
| "learning_rate": 3.7798863400879894e-06, | |
| "loss": 0.0627, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 5.9072847682119205, | |
| "grad_norm": 2.9067546061467313, | |
| "learning_rate": 3.7659925541532006e-06, | |
| "loss": 0.062, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 5.916114790286976, | |
| "grad_norm": 2.082373981665909, | |
| "learning_rate": 3.752108907241682e-06, | |
| "loss": 0.0484, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.924944812362031, | |
| "grad_norm": 2.8414745300376985, | |
| "learning_rate": 3.738235513426184e-06, | |
| "loss": 0.0538, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 5.933774834437086, | |
| "grad_norm": 2.7350597659868887, | |
| "learning_rate": 3.7243724866952114e-06, | |
| "loss": 0.0591, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 5.942604856512141, | |
| "grad_norm": 2.950452086994668, | |
| "learning_rate": 3.710519940952085e-06, | |
| "loss": 0.0619, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 5.951434878587197, | |
| "grad_norm": 3.3971453739654947, | |
| "learning_rate": 3.6966779900140193e-06, | |
| "loss": 0.0629, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 5.960264900662252, | |
| "grad_norm": 3.4801957062345545, | |
| "learning_rate": 3.6828467476111664e-06, | |
| "loss": 0.0568, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 5.969094922737307, | |
| "grad_norm": 3.890846507295426, | |
| "learning_rate": 3.6690263273857035e-06, | |
| "loss": 0.0582, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 5.977924944812362, | |
| "grad_norm": 5.583327342274176, | |
| "learning_rate": 3.6552168428908886e-06, | |
| "loss": 0.0527, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 5.986754966887418, | |
| "grad_norm": 3.0507993466730583, | |
| "learning_rate": 3.6414184075901206e-06, | |
| "loss": 0.0502, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 5.995584988962472, | |
| "grad_norm": 2.907342135871787, | |
| "learning_rate": 3.62763113485602e-06, | |
| "loss": 0.0585, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 6.004415011037527, | |
| "grad_norm": 2.5372341157116653, | |
| "learning_rate": 3.6138551379694936e-06, | |
| "loss": 0.049, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 6.013245033112582, | |
| "grad_norm": 2.901139380705564, | |
| "learning_rate": 3.6000905301187953e-06, | |
| "loss": 0.0392, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 6.022075055187638, | |
| "grad_norm": 2.132587134784585, | |
| "learning_rate": 3.586337424398609e-06, | |
| "loss": 0.0367, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 6.030905077262693, | |
| "grad_norm": 3.0119134167439716, | |
| "learning_rate": 3.5725959338091133e-06, | |
| "loss": 0.0391, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 6.039735099337748, | |
| "grad_norm": 2.497663295122042, | |
| "learning_rate": 3.5588661712550464e-06, | |
| "loss": 0.0409, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 6.048565121412803, | |
| "grad_norm": 5.219603199736772, | |
| "learning_rate": 3.545148249544793e-06, | |
| "loss": 0.042, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 6.057395143487859, | |
| "grad_norm": 7.696899888197639, | |
| "learning_rate": 3.5314422813894413e-06, | |
| "loss": 0.0357, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 6.066225165562914, | |
| "grad_norm": 3.0317914717988073, | |
| "learning_rate": 3.517748379401872e-06, | |
| "loss": 0.0401, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 6.075055187637969, | |
| "grad_norm": 5.738303523610005, | |
| "learning_rate": 3.5040666560958246e-06, | |
| "loss": 0.0406, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 6.083885209713024, | |
| "grad_norm": 2.566024467496126, | |
| "learning_rate": 3.4903972238849727e-06, | |
| "loss": 0.0433, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 6.0927152317880795, | |
| "grad_norm": 14.693362185579907, | |
| "learning_rate": 3.4767401950820003e-06, | |
| "loss": 0.0408, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 6.101545253863135, | |
| "grad_norm": 3.051023280314456, | |
| "learning_rate": 3.4630956818976875e-06, | |
| "loss": 0.0401, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 6.11037527593819, | |
| "grad_norm": 3.1996583158698377, | |
| "learning_rate": 3.4494637964399723e-06, | |
| "loss": 0.0418, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 6.119205298013245, | |
| "grad_norm": 2.4568209242710832, | |
| "learning_rate": 3.4358446507130503e-06, | |
| "loss": 0.038, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 6.1280353200883, | |
| "grad_norm": 3.784789713649103, | |
| "learning_rate": 3.4222383566164314e-06, | |
| "loss": 0.0349, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 6.136865342163356, | |
| "grad_norm": 2.506468109318104, | |
| "learning_rate": 3.408645025944042e-06, | |
| "loss": 0.0481, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 6.145695364238411, | |
| "grad_norm": 10.277226238218848, | |
| "learning_rate": 3.3950647703832907e-06, | |
| "loss": 0.0399, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 6.154525386313466, | |
| "grad_norm": 3.446697091550699, | |
| "learning_rate": 3.3814977015141576e-06, | |
| "loss": 0.0402, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 6.163355408388521, | |
| "grad_norm": 3.1440652404590326, | |
| "learning_rate": 3.3679439308082777e-06, | |
| "loss": 0.0427, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 6.172185430463577, | |
| "grad_norm": 1.8022214396900063, | |
| "learning_rate": 3.3544035696280264e-06, | |
| "loss": 0.0311, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 6.181015452538631, | |
| "grad_norm": 7.203643491781946, | |
| "learning_rate": 3.340876729225595e-06, | |
| "loss": 0.0452, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.189845474613686, | |
| "grad_norm": 5.443104722512851, | |
| "learning_rate": 3.327363520742087e-06, | |
| "loss": 0.0462, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 6.198675496688741, | |
| "grad_norm": 2.892866724062855, | |
| "learning_rate": 3.313864055206607e-06, | |
| "loss": 0.0451, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 6.207505518763797, | |
| "grad_norm": 2.5825036634770955, | |
| "learning_rate": 3.3003784435353304e-06, | |
| "loss": 0.0447, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 6.216335540838852, | |
| "grad_norm": 2.371081962330087, | |
| "learning_rate": 3.2869067965306178e-06, | |
| "loss": 0.0395, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 6.225165562913907, | |
| "grad_norm": 5.2888513009273215, | |
| "learning_rate": 3.273449224880081e-06, | |
| "loss": 0.0459, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 6.233995584988962, | |
| "grad_norm": 3.246689980218029, | |
| "learning_rate": 3.260005839155691e-06, | |
| "loss": 0.0439, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 6.242825607064018, | |
| "grad_norm": 2.664991670569899, | |
| "learning_rate": 3.2465767498128596e-06, | |
| "loss": 0.037, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 6.251655629139073, | |
| "grad_norm": 2.8051287977254886, | |
| "learning_rate": 3.233162067189533e-06, | |
| "loss": 0.043, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 6.260485651214128, | |
| "grad_norm": 2.133632773820073, | |
| "learning_rate": 3.2197619015052893e-06, | |
| "loss": 0.0385, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 6.269315673289183, | |
| "grad_norm": 2.505052974779182, | |
| "learning_rate": 3.206376362860432e-06, | |
| "loss": 0.0349, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 6.2781456953642385, | |
| "grad_norm": 1.6900753857878728, | |
| "learning_rate": 3.1930055612350795e-06, | |
| "loss": 0.036, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 6.286975717439294, | |
| "grad_norm": 2.540432626556844, | |
| "learning_rate": 3.1796496064882677e-06, | |
| "loss": 0.0384, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 6.295805739514349, | |
| "grad_norm": 3.868753172143831, | |
| "learning_rate": 3.1663086083570493e-06, | |
| "loss": 0.0348, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 6.304635761589404, | |
| "grad_norm": 2.151059410846429, | |
| "learning_rate": 3.152982676455581e-06, | |
| "loss": 0.037, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 6.313465783664459, | |
| "grad_norm": 2.1492905792405326, | |
| "learning_rate": 3.1396719202742375e-06, | |
| "loss": 0.0384, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 6.322295805739515, | |
| "grad_norm": 2.7591878809667767, | |
| "learning_rate": 3.1263764491786984e-06, | |
| "loss": 0.039, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 6.33112582781457, | |
| "grad_norm": 3.7396816809123026, | |
| "learning_rate": 3.1130963724090626e-06, | |
| "loss": 0.0395, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 6.339955849889625, | |
| "grad_norm": 2.7892607629887944, | |
| "learning_rate": 3.0998317990789378e-06, | |
| "loss": 0.0382, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 6.3487858719646795, | |
| "grad_norm": 2.446511315008329, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.0369, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 6.357615894039735, | |
| "grad_norm": 4.465803535799041, | |
| "learning_rate": 3.0733495985538575e-06, | |
| "loss": 0.0382, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.36644591611479, | |
| "grad_norm": 3.7975803644599395, | |
| "learning_rate": 3.0601321889456378e-06, | |
| "loss": 0.0479, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 6.375275938189845, | |
| "grad_norm": 5.910694681761274, | |
| "learning_rate": 3.046930717948604e-06, | |
| "loss": 0.0446, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 6.3841059602649, | |
| "grad_norm": 3.5801704875383056, | |
| "learning_rate": 3.033745294030517e-06, | |
| "loss": 0.0481, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 6.392935982339956, | |
| "grad_norm": 2.7983101793342344, | |
| "learning_rate": 3.0205760255272874e-06, | |
| "loss": 0.0422, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 6.401766004415011, | |
| "grad_norm": 2.923394588563541, | |
| "learning_rate": 3.007423020642084e-06, | |
| "loss": 0.04, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 6.410596026490066, | |
| "grad_norm": 2.670813931896076, | |
| "learning_rate": 2.9942863874444565e-06, | |
| "loss": 0.0364, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 6.419426048565121, | |
| "grad_norm": 4.334736430377144, | |
| "learning_rate": 2.981166233869429e-06, | |
| "loss": 0.0341, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 6.428256070640177, | |
| "grad_norm": 2.101335167632437, | |
| "learning_rate": 2.9680626677166324e-06, | |
| "loss": 0.0341, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 6.437086092715232, | |
| "grad_norm": 2.8152196075893126, | |
| "learning_rate": 2.9549757966494053e-06, | |
| "loss": 0.0431, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 6.445916114790287, | |
| "grad_norm": 2.521300943288849, | |
| "learning_rate": 2.9419057281939106e-06, | |
| "loss": 0.0406, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.454746136865342, | |
| "grad_norm": 3.2931844374433976, | |
| "learning_rate": 2.9288525697382623e-06, | |
| "loss": 0.0362, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 6.4635761589403975, | |
| "grad_norm": 2.614155855220592, | |
| "learning_rate": 2.9158164285316356e-06, | |
| "loss": 0.0481, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.472406181015453, | |
| "grad_norm": 2.370532842877096, | |
| "learning_rate": 2.9027974116833756e-06, | |
| "loss": 0.0361, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 6.481236203090508, | |
| "grad_norm": 3.792527309831017, | |
| "learning_rate": 2.889795626162143e-06, | |
| "loss": 0.0398, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 6.490066225165563, | |
| "grad_norm": 2.2254029916470053, | |
| "learning_rate": 2.8768111787950105e-06, | |
| "loss": 0.0357, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 6.498896247240618, | |
| "grad_norm": 2.9758775001184237, | |
| "learning_rate": 2.8638441762665957e-06, | |
| "loss": 0.0424, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 6.507726269315674, | |
| "grad_norm": 2.7928679400305167, | |
| "learning_rate": 2.8508947251181885e-06, | |
| "loss": 0.0409, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 6.516556291390728, | |
| "grad_norm": 2.7552105391367467, | |
| "learning_rate": 2.8379629317468604e-06, | |
| "loss": 0.0376, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 6.525386313465784, | |
| "grad_norm": 4.608769183414743, | |
| "learning_rate": 2.825048902404612e-06, | |
| "loss": 0.0422, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 6.5342163355408385, | |
| "grad_norm": 2.7068232049413656, | |
| "learning_rate": 2.8121527431974838e-06, | |
| "loss": 0.0398, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 6.543046357615894, | |
| "grad_norm": 2.3005458239974317, | |
| "learning_rate": 2.799274560084688e-06, | |
| "loss": 0.0361, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 6.551876379690949, | |
| "grad_norm": 2.061368895244273, | |
| "learning_rate": 2.7864144588777403e-06, | |
| "loss": 0.0375, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 6.560706401766004, | |
| "grad_norm": 3.488560417322964, | |
| "learning_rate": 2.77357254523959e-06, | |
| "loss": 0.0404, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 6.569536423841059, | |
| "grad_norm": 4.516566926696822, | |
| "learning_rate": 2.7607489246837505e-06, | |
| "loss": 0.0373, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 6.578366445916115, | |
| "grad_norm": 2.162092538735688, | |
| "learning_rate": 2.74794370257343e-06, | |
| "loss": 0.0367, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 6.58719646799117, | |
| "grad_norm": 2.209712238867527, | |
| "learning_rate": 2.7351569841206792e-06, | |
| "loss": 0.0338, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 6.596026490066225, | |
| "grad_norm": 4.902409366826644, | |
| "learning_rate": 2.722388874385503e-06, | |
| "loss": 0.0296, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 6.60485651214128, | |
| "grad_norm": 2.991720229224731, | |
| "learning_rate": 2.7096394782750186e-06, | |
| "loss": 0.0403, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 6.613686534216336, | |
| "grad_norm": 4.394681767828003, | |
| "learning_rate": 2.696908900542584e-06, | |
| "loss": 0.0427, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 6.622516556291391, | |
| "grad_norm": 2.241917975111583, | |
| "learning_rate": 2.684197245786938e-06, | |
| "loss": 0.0389, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.631346578366446, | |
| "grad_norm": 2.5929391814373317, | |
| "learning_rate": 2.671504618451348e-06, | |
| "loss": 0.0428, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 6.640176600441501, | |
| "grad_norm": 3.986667280450995, | |
| "learning_rate": 2.658831122822735e-06, | |
| "loss": 0.0385, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 6.6490066225165565, | |
| "grad_norm": 2.4146196048021515, | |
| "learning_rate": 2.6461768630308326e-06, | |
| "loss": 0.0345, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 6.657836644591612, | |
| "grad_norm": 2.215591523500783, | |
| "learning_rate": 2.633541943047334e-06, | |
| "loss": 0.0413, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 3.2975620777429784, | |
| "learning_rate": 2.620926466685013e-06, | |
| "loss": 0.0398, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 6.675496688741722, | |
| "grad_norm": 4.223972999480639, | |
| "learning_rate": 2.608330537596907e-06, | |
| "loss": 0.0389, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 6.684326710816777, | |
| "grad_norm": 5.9344165036781895, | |
| "learning_rate": 2.595754259275428e-06, | |
| "loss": 0.0376, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 6.693156732891833, | |
| "grad_norm": 3.292430486809714, | |
| "learning_rate": 2.583197735051546e-06, | |
| "loss": 0.0421, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 6.701986754966887, | |
| "grad_norm": 2.849761239414322, | |
| "learning_rate": 2.5706610680939186e-06, | |
| "loss": 0.0416, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 6.710816777041942, | |
| "grad_norm": 2.2524832012223888, | |
| "learning_rate": 2.5581443614080433e-06, | |
| "loss": 0.0375, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.7196467991169975, | |
| "grad_norm": 2.732715070931953, | |
| "learning_rate": 2.545647717835428e-06, | |
| "loss": 0.0421, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 6.728476821192053, | |
| "grad_norm": 2.1389488417703504, | |
| "learning_rate": 2.53317124005273e-06, | |
| "loss": 0.0425, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 6.737306843267108, | |
| "grad_norm": 3.1275909013272707, | |
| "learning_rate": 2.5207150305709167e-06, | |
| "loss": 0.0353, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 6.746136865342163, | |
| "grad_norm": 3.4875687811241685, | |
| "learning_rate": 2.5082791917344256e-06, | |
| "loss": 0.0379, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 6.754966887417218, | |
| "grad_norm": 8.559694789319202, | |
| "learning_rate": 2.495863825720322e-06, | |
| "loss": 0.0336, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 6.763796909492274, | |
| "grad_norm": 2.685216387232715, | |
| "learning_rate": 2.4834690345374608e-06, | |
| "loss": 0.0425, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 6.772626931567329, | |
| "grad_norm": 2.5515950966722523, | |
| "learning_rate": 2.471094920025644e-06, | |
| "loss": 0.0378, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 6.781456953642384, | |
| "grad_norm": 2.6487457621063086, | |
| "learning_rate": 2.45874158385479e-06, | |
| "loss": 0.035, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 6.790286975717439, | |
| "grad_norm": 2.8955292191315163, | |
| "learning_rate": 2.446409127524094e-06, | |
| "loss": 0.0374, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 6.799116997792495, | |
| "grad_norm": 3.225668023606972, | |
| "learning_rate": 2.4340976523611957e-06, | |
| "loss": 0.0368, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.80794701986755, | |
| "grad_norm": 3.2143169626555577, | |
| "learning_rate": 2.4218072595213467e-06, | |
| "loss": 0.036, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 6.816777041942605, | |
| "grad_norm": 2.316528139483897, | |
| "learning_rate": 2.409538049986576e-06, | |
| "loss": 0.0389, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 6.82560706401766, | |
| "grad_norm": 2.1265023326583568, | |
| "learning_rate": 2.3972901245648724e-06, | |
| "loss": 0.0368, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 6.8344370860927155, | |
| "grad_norm": 2.748400166671758, | |
| "learning_rate": 2.385063583889335e-06, | |
| "loss": 0.0351, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 6.843267108167771, | |
| "grad_norm": 3.1784434215970077, | |
| "learning_rate": 2.3728585284173646e-06, | |
| "loss": 0.0389, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 6.852097130242826, | |
| "grad_norm": 3.093985098754174, | |
| "learning_rate": 2.3606750584298375e-06, | |
| "loss": 0.0423, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 6.860927152317881, | |
| "grad_norm": 3.1085189340885697, | |
| "learning_rate": 2.348513274030264e-06, | |
| "loss": 0.0383, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 6.8697571743929355, | |
| "grad_norm": 2.332572892938318, | |
| "learning_rate": 2.3363732751439926e-06, | |
| "loss": 0.0407, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 6.878587196467992, | |
| "grad_norm": 3.3589262508003097, | |
| "learning_rate": 2.32425516151736e-06, | |
| "loss": 0.0373, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 6.887417218543046, | |
| "grad_norm": 1.9386944155345733, | |
| "learning_rate": 2.3121590327168987e-06, | |
| "loss": 0.0324, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.896247240618101, | |
| "grad_norm": 1.9199888954422202, | |
| "learning_rate": 2.3000849881285016e-06, | |
| "loss": 0.0346, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 6.9050772626931565, | |
| "grad_norm": 2.012927623083998, | |
| "learning_rate": 2.2880331269566043e-06, | |
| "loss": 0.0349, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 6.913907284768212, | |
| "grad_norm": 2.5454387144102624, | |
| "learning_rate": 2.2760035482233868e-06, | |
| "loss": 0.0381, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 6.922737306843267, | |
| "grad_norm": 1.836927639977041, | |
| "learning_rate": 2.263996350767942e-06, | |
| "loss": 0.0297, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 6.931567328918322, | |
| "grad_norm": 5.426472131691331, | |
| "learning_rate": 2.2520116332454726e-06, | |
| "loss": 0.0355, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 6.940397350993377, | |
| "grad_norm": 2.4857248120488515, | |
| "learning_rate": 2.240049494126479e-06, | |
| "loss": 0.0417, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 6.949227373068433, | |
| "grad_norm": 2.1624227633805053, | |
| "learning_rate": 2.2281100316959476e-06, | |
| "loss": 0.0346, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 6.958057395143488, | |
| "grad_norm": 4.623829014905759, | |
| "learning_rate": 2.2161933440525474e-06, | |
| "loss": 0.0426, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 6.966887417218543, | |
| "grad_norm": 2.946162852577001, | |
| "learning_rate": 2.2042995291078227e-06, | |
| "loss": 0.0354, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 6.975717439293598, | |
| "grad_norm": 2.5327062019367, | |
| "learning_rate": 2.192428684585386e-06, | |
| "loss": 0.038, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.984547461368654, | |
| "grad_norm": 2.6894541258986044, | |
| "learning_rate": 2.180580908020117e-06, | |
| "loss": 0.0421, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 6.993377483443709, | |
| "grad_norm": 3.3879149653062215, | |
| "learning_rate": 2.1687562967573645e-06, | |
| "loss": 0.031, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 7.002207505518764, | |
| "grad_norm": 1.9395684157763193, | |
| "learning_rate": 2.156954947952139e-06, | |
| "loss": 0.0376, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 7.011037527593819, | |
| "grad_norm": 1.3351262364052678, | |
| "learning_rate": 2.1451769585683196e-06, | |
| "loss": 0.0242, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 7.0198675496688745, | |
| "grad_norm": 1.838802636867857, | |
| "learning_rate": 2.1334224253778628e-06, | |
| "loss": 0.0283, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 7.02869757174393, | |
| "grad_norm": 3.943880828831373, | |
| "learning_rate": 2.1216914449599905e-06, | |
| "loss": 0.0224, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 7.037527593818985, | |
| "grad_norm": 6.979679476282558, | |
| "learning_rate": 2.109984113700413e-06, | |
| "loss": 0.0213, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 7.046357615894039, | |
| "grad_norm": 1.7676454435450142, | |
| "learning_rate": 2.0983005277905348e-06, | |
| "loss": 0.028, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 7.0551876379690945, | |
| "grad_norm": 2.2950302206098256, | |
| "learning_rate": 2.0866407832266506e-06, | |
| "loss": 0.0173, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 7.06401766004415, | |
| "grad_norm": 3.5619001044641116, | |
| "learning_rate": 2.0750049758091778e-06, | |
| "loss": 0.0268, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.072847682119205, | |
| "grad_norm": 2.184888168844001, | |
| "learning_rate": 2.063393201141846e-06, | |
| "loss": 0.0216, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 7.08167770419426, | |
| "grad_norm": 1.7980721349330822, | |
| "learning_rate": 2.0518055546309362e-06, | |
| "loss": 0.0206, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 7.0905077262693155, | |
| "grad_norm": 2.16361649558258, | |
| "learning_rate": 2.0402421314844774e-06, | |
| "loss": 0.0179, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 7.099337748344371, | |
| "grad_norm": 2.5322936668005074, | |
| "learning_rate": 2.0287030267114665e-06, | |
| "loss": 0.0182, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 7.108167770419426, | |
| "grad_norm": 2.8585159988830156, | |
| "learning_rate": 2.0171883351211038e-06, | |
| "loss": 0.0182, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 7.116997792494481, | |
| "grad_norm": 2.508620475053591, | |
| "learning_rate": 2.0056981513219944e-06, | |
| "loss": 0.0185, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 7.125827814569536, | |
| "grad_norm": 2.321001585490897, | |
| "learning_rate": 1.9942325697213817e-06, | |
| "loss": 0.0218, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 7.134657836644592, | |
| "grad_norm": 2.8456076364746155, | |
| "learning_rate": 1.9827916845243687e-06, | |
| "loss": 0.0284, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 7.143487858719647, | |
| "grad_norm": 5.21332677622318, | |
| "learning_rate": 1.971375589733145e-06, | |
| "loss": 0.019, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 7.152317880794702, | |
| "grad_norm": 2.600603391190043, | |
| "learning_rate": 1.9599843791462123e-06, | |
| "loss": 0.0226, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 7.161147902869757, | |
| "grad_norm": 2.4291651186702357, | |
| "learning_rate": 1.9486181463576176e-06, | |
| "loss": 0.0168, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 7.169977924944813, | |
| "grad_norm": 2.06300021341791, | |
| "learning_rate": 1.937276984756179e-06, | |
| "loss": 0.0202, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 7.178807947019868, | |
| "grad_norm": 2.430343025724867, | |
| "learning_rate": 1.925960987524724e-06, | |
| "loss": 0.0183, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 7.187637969094923, | |
| "grad_norm": 2.4638001879801994, | |
| "learning_rate": 1.91467024763932e-06, | |
| "loss": 0.0178, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 7.196467991169978, | |
| "grad_norm": 2.966513980989748, | |
| "learning_rate": 1.9034048578685099e-06, | |
| "loss": 0.0259, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 7.2052980132450335, | |
| "grad_norm": 4.97492019147766, | |
| "learning_rate": 1.8921649107725525e-06, | |
| "loss": 0.0206, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 7.214128035320089, | |
| "grad_norm": 3.7362996609089203, | |
| "learning_rate": 1.880950498702666e-06, | |
| "loss": 0.0221, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 7.222958057395143, | |
| "grad_norm": 5.330062168647716, | |
| "learning_rate": 1.8697617138002545e-06, | |
| "loss": 0.0203, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 7.231788079470198, | |
| "grad_norm": 2.4682768719456907, | |
| "learning_rate": 1.8585986479961653e-06, | |
| "loss": 0.0236, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 7.2406181015452535, | |
| "grad_norm": 2.3107033724617723, | |
| "learning_rate": 1.8474613930099356e-06, | |
| "loss": 0.0198, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 7.249448123620309, | |
| "grad_norm": 2.074339676493327, | |
| "learning_rate": 1.8363500403490175e-06, | |
| "loss": 0.0206, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 7.258278145695364, | |
| "grad_norm": 2.6070772200880747, | |
| "learning_rate": 1.8252646813080566e-06, | |
| "loss": 0.0248, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 7.267108167770419, | |
| "grad_norm": 3.005859370147622, | |
| "learning_rate": 1.814205406968112e-06, | |
| "loss": 0.0231, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 7.2759381898454745, | |
| "grad_norm": 2.4914405243393523, | |
| "learning_rate": 1.8031723081959334e-06, | |
| "loss": 0.0268, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 7.28476821192053, | |
| "grad_norm": 2.0961570467071406, | |
| "learning_rate": 1.792165475643199e-06, | |
| "loss": 0.0166, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 7.293598233995585, | |
| "grad_norm": 2.044423079381866, | |
| "learning_rate": 1.7811849997457681e-06, | |
| "loss": 0.0202, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 7.30242825607064, | |
| "grad_norm": 2.241605271360007, | |
| "learning_rate": 1.7702309707229576e-06, | |
| "loss": 0.0254, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 7.311258278145695, | |
| "grad_norm": 2.544290234878124, | |
| "learning_rate": 1.7593034785767788e-06, | |
| "loss": 0.0305, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 7.320088300220751, | |
| "grad_norm": 2.2185422029084787, | |
| "learning_rate": 1.7484026130912097e-06, | |
| "loss": 0.0199, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 7.328918322295806, | |
| "grad_norm": 3.385549191337894, | |
| "learning_rate": 1.737528463831456e-06, | |
| "loss": 0.0232, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 7.337748344370861, | |
| "grad_norm": 2.0846787775644144, | |
| "learning_rate": 1.726681120143207e-06, | |
| "loss": 0.0219, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 7.346578366445916, | |
| "grad_norm": 2.483671973493014, | |
| "learning_rate": 1.7158606711519193e-06, | |
| "loss": 0.0215, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 7.355408388520972, | |
| "grad_norm": 1.9008728371682941, | |
| "learning_rate": 1.7050672057620666e-06, | |
| "loss": 0.019, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 7.364238410596027, | |
| "grad_norm": 2.025237707875373, | |
| "learning_rate": 1.6943008126564164e-06, | |
| "loss": 0.0221, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 7.373068432671082, | |
| "grad_norm": 2.538304228617941, | |
| "learning_rate": 1.6835615802953026e-06, | |
| "loss": 0.0238, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 7.381898454746137, | |
| "grad_norm": 2.1945407829106642, | |
| "learning_rate": 1.6728495969158976e-06, | |
| "loss": 0.0233, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 7.390728476821192, | |
| "grad_norm": 2.1102890644808014, | |
| "learning_rate": 1.6621649505314853e-06, | |
| "loss": 0.0215, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 7.399558498896247, | |
| "grad_norm": 2.6134856388418926, | |
| "learning_rate": 1.6515077289307391e-06, | |
| "loss": 0.0182, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 7.408388520971302, | |
| "grad_norm": 6.901364649837025, | |
| "learning_rate": 1.640878019677008e-06, | |
| "loss": 0.0241, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 7.417218543046357, | |
| "grad_norm": 2.70801903001521, | |
| "learning_rate": 1.6302759101075788e-06, | |
| "loss": 0.0205, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.4260485651214125, | |
| "grad_norm": 6.919268025262333, | |
| "learning_rate": 1.619701487332978e-06, | |
| "loss": 0.0174, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 7.434878587196468, | |
| "grad_norm": 2.4424662065293896, | |
| "learning_rate": 1.609154838236246e-06, | |
| "loss": 0.0174, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 7.443708609271523, | |
| "grad_norm": 3.80128705163988, | |
| "learning_rate": 1.5986360494722237e-06, | |
| "loss": 0.0183, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 7.452538631346578, | |
| "grad_norm": 3.6044220961477333, | |
| "learning_rate": 1.5881452074668474e-06, | |
| "loss": 0.0215, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 7.4613686534216335, | |
| "grad_norm": 2.6228062867466617, | |
| "learning_rate": 1.577682398416424e-06, | |
| "loss": 0.0171, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 7.470198675496689, | |
| "grad_norm": 2.0977523566035017, | |
| "learning_rate": 1.567247708286942e-06, | |
| "loss": 0.0217, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 7.479028697571744, | |
| "grad_norm": 2.793220603051959, | |
| "learning_rate": 1.5568412228133506e-06, | |
| "loss": 0.0216, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 7.487858719646799, | |
| "grad_norm": 2.02924183324416, | |
| "learning_rate": 1.5464630274988558e-06, | |
| "loss": 0.0193, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 7.496688741721854, | |
| "grad_norm": 2.9310793919597002, | |
| "learning_rate": 1.5361132076142316e-06, | |
| "loss": 0.022, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 7.50551876379691, | |
| "grad_norm": 2.1719118985345136, | |
| "learning_rate": 1.5257918481971028e-06, | |
| "loss": 0.0249, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.514348785871965, | |
| "grad_norm": 3.0190872228806676, | |
| "learning_rate": 1.515499034051256e-06, | |
| "loss": 0.0254, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 7.52317880794702, | |
| "grad_norm": 2.1279187042699226, | |
| "learning_rate": 1.50523484974594e-06, | |
| "loss": 0.0212, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 7.532008830022075, | |
| "grad_norm": 2.507090582734221, | |
| "learning_rate": 1.4949993796151675e-06, | |
| "loss": 0.0247, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 7.540838852097131, | |
| "grad_norm": 2.458174732413524, | |
| "learning_rate": 1.4847927077570324e-06, | |
| "loss": 0.0172, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.549668874172186, | |
| "grad_norm": 2.473671700510561, | |
| "learning_rate": 1.4746149180330082e-06, | |
| "loss": 0.0179, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 7.55849889624724, | |
| "grad_norm": 2.2559671767662985, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.019, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 7.567328918322296, | |
| "grad_norm": 5.587791991196285, | |
| "learning_rate": 1.4543463192459728e-06, | |
| "loss": 0.0232, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 7.576158940397351, | |
| "grad_norm": 4.176103681184144, | |
| "learning_rate": 1.4442556767166371e-06, | |
| "loss": 0.0199, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 7.584988962472406, | |
| "grad_norm": 2.952365079659642, | |
| "learning_rate": 1.4341942493873934e-06, | |
| "loss": 0.0208, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 7.593818984547461, | |
| "grad_norm": 2.6443262718204403, | |
| "learning_rate": 1.4241621199263362e-06, | |
| "loss": 0.0209, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 7.602649006622516, | |
| "grad_norm": 2.789233213497681, | |
| "learning_rate": 1.4141593707608441e-06, | |
| "loss": 0.0266, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 7.6114790286975715, | |
| "grad_norm": 2.0102562040904384, | |
| "learning_rate": 1.40418608407689e-06, | |
| "loss": 0.0181, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 7.620309050772627, | |
| "grad_norm": 3.2658211403555164, | |
| "learning_rate": 1.3942423418183764e-06, | |
| "loss": 0.0245, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 7.629139072847682, | |
| "grad_norm": 2.6224642656954975, | |
| "learning_rate": 1.3843282256864599e-06, | |
| "loss": 0.022, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 7.637969094922737, | |
| "grad_norm": 2.5685133896281283, | |
| "learning_rate": 1.3744438171388752e-06, | |
| "loss": 0.0212, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 7.6467991169977925, | |
| "grad_norm": 2.319203592892883, | |
| "learning_rate": 1.3645891973892772e-06, | |
| "loss": 0.0204, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 7.655629139072848, | |
| "grad_norm": 1.656658157416864, | |
| "learning_rate": 1.3547644474065557e-06, | |
| "loss": 0.0171, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 7.664459161147903, | |
| "grad_norm": 2.3832665868336287, | |
| "learning_rate": 1.3449696479141855e-06, | |
| "loss": 0.0186, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 7.673289183222958, | |
| "grad_norm": 3.2514803641340686, | |
| "learning_rate": 1.3352048793895623e-06, | |
| "loss": 0.0148, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 7.682119205298013, | |
| "grad_norm": 2.271952293563063, | |
| "learning_rate": 1.325470222063327e-06, | |
| "loss": 0.0225, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 7.690949227373069, | |
| "grad_norm": 2.598045236141581, | |
| "learning_rate": 1.3157657559187264e-06, | |
| "loss": 0.0158, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 7.699779249448124, | |
| "grad_norm": 2.2037954681254224, | |
| "learning_rate": 1.3060915606909413e-06, | |
| "loss": 0.0241, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 7.708609271523179, | |
| "grad_norm": 1.7284528351938868, | |
| "learning_rate": 1.2964477158664367e-06, | |
| "loss": 0.0191, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 7.717439293598234, | |
| "grad_norm": 2.365122065428329, | |
| "learning_rate": 1.2868343006823113e-06, | |
| "loss": 0.0224, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 7.72626931567329, | |
| "grad_norm": 2.273506107793251, | |
| "learning_rate": 1.2772513941256371e-06, | |
| "loss": 0.0236, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 7.735099337748345, | |
| "grad_norm": 2.008267127031853, | |
| "learning_rate": 1.2676990749328255e-06, | |
| "loss": 0.0153, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 7.743929359823399, | |
| "grad_norm": 2.48839562244605, | |
| "learning_rate": 1.2581774215889653e-06, | |
| "loss": 0.0213, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 7.752759381898454, | |
| "grad_norm": 1.7178855291698532, | |
| "learning_rate": 1.2486865123271868e-06, | |
| "loss": 0.0189, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 7.76158940397351, | |
| "grad_norm": 1.8489588055111108, | |
| "learning_rate": 1.2392264251280167e-06, | |
| "loss": 0.0161, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 7.770419426048565, | |
| "grad_norm": 1.8146961936190737, | |
| "learning_rate": 1.2297972377187361e-06, | |
| "loss": 0.0206, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 7.77924944812362, | |
| "grad_norm": 1.8749300582474253, | |
| "learning_rate": 1.2203990275727435e-06, | |
| "loss": 0.019, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 7.788079470198675, | |
| "grad_norm": 2.9053006214179473, | |
| "learning_rate": 1.211031871908916e-06, | |
| "loss": 0.0164, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 7.7969094922737305, | |
| "grad_norm": 1.6936916234587374, | |
| "learning_rate": 1.201695847690983e-06, | |
| "loss": 0.0153, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 7.805739514348786, | |
| "grad_norm": 1.776853857970458, | |
| "learning_rate": 1.1923910316268783e-06, | |
| "loss": 0.0178, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 7.814569536423841, | |
| "grad_norm": 2.2451712890117683, | |
| "learning_rate": 1.183117500168125e-06, | |
| "loss": 0.0151, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 7.823399558498896, | |
| "grad_norm": 2.562938392111076, | |
| "learning_rate": 1.1738753295091986e-06, | |
| "loss": 0.0257, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 7.8322295805739515, | |
| "grad_norm": 2.2393583493432065, | |
| "learning_rate": 1.164664595586904e-06, | |
| "loss": 0.0192, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 7.841059602649007, | |
| "grad_norm": 2.11442244493206, | |
| "learning_rate": 1.1554853740797556e-06, | |
| "loss": 0.0208, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 7.849889624724062, | |
| "grad_norm": 2.823907846227811, | |
| "learning_rate": 1.1463377404073433e-06, | |
| "loss": 0.0193, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 7.858719646799117, | |
| "grad_norm": 2.1444744833664817, | |
| "learning_rate": 1.137221769729725e-06, | |
| "loss": 0.0214, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.867549668874172, | |
| "grad_norm": 2.4488895115951177, | |
| "learning_rate": 1.1281375369468078e-06, | |
| "loss": 0.022, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 7.876379690949228, | |
| "grad_norm": 1.905459279890981, | |
| "learning_rate": 1.1190851166977218e-06, | |
| "loss": 0.0185, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 7.885209713024283, | |
| "grad_norm": 2.7364856840584575, | |
| "learning_rate": 1.1100645833602231e-06, | |
| "loss": 0.0198, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 7.894039735099338, | |
| "grad_norm": 2.295609038935843, | |
| "learning_rate": 1.1010760110500652e-06, | |
| "loss": 0.0195, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 7.902869757174393, | |
| "grad_norm": 1.9420471727188124, | |
| "learning_rate": 1.0921194736204066e-06, | |
| "loss": 0.0169, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 7.911699779249448, | |
| "grad_norm": 2.3765948390021148, | |
| "learning_rate": 1.083195044661195e-06, | |
| "loss": 0.019, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 7.920529801324504, | |
| "grad_norm": 2.9487851656499213, | |
| "learning_rate": 1.0743027974985576e-06, | |
| "loss": 0.0164, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 7.929359823399558, | |
| "grad_norm": 2.0278573964902713, | |
| "learning_rate": 1.065442805194214e-06, | |
| "loss": 0.0213, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 7.938189845474613, | |
| "grad_norm": 2.4507086572112287, | |
| "learning_rate": 1.056615140544861e-06, | |
| "loss": 0.0209, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 7.947019867549669, | |
| "grad_norm": 2.504870886604773, | |
| "learning_rate": 1.0478198760815833e-06, | |
| "loss": 0.0203, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.955849889624724, | |
| "grad_norm": 2.6906909041136586, | |
| "learning_rate": 1.0390570840692527e-06, | |
| "loss": 0.0188, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 7.964679911699779, | |
| "grad_norm": 2.5738303951773567, | |
| "learning_rate": 1.0303268365059383e-06, | |
| "loss": 0.0223, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 7.973509933774834, | |
| "grad_norm": 2.254353943389869, | |
| "learning_rate": 1.021629205122311e-06, | |
| "loss": 0.0208, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 7.9823399558498895, | |
| "grad_norm": 2.1558770703060017, | |
| "learning_rate": 1.0129642613810576e-06, | |
| "loss": 0.0185, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 7.991169977924945, | |
| "grad_norm": 2.6229552158930534, | |
| "learning_rate": 1.0043320764762915e-06, | |
| "loss": 0.0172, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.054375538718464, | |
| "learning_rate": 9.957327213329687e-07, | |
| "loss": 0.0219, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 8.008830022075054, | |
| "grad_norm": 1.9782066500836768, | |
| "learning_rate": 9.871662666063054e-07, | |
| "loss": 0.0112, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 8.01766004415011, | |
| "grad_norm": 1.4589642331986807, | |
| "learning_rate": 9.786327826811942e-07, | |
| "loss": 0.0103, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 8.026490066225165, | |
| "grad_norm": 1.1710309196675532, | |
| "learning_rate": 9.701323396716312e-07, | |
| "loss": 0.0093, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 8.035320088300221, | |
| "grad_norm": 1.4787353632980826, | |
| "learning_rate": 9.616650074201383e-07, | |
| "loss": 0.0088, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 8.044150110375275, | |
| "grad_norm": 1.2424731443976285, | |
| "learning_rate": 9.532308554971831e-07, | |
| "loss": 0.0095, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 8.052980132450331, | |
| "grad_norm": 1.3858499824497619, | |
| "learning_rate": 9.44829953200615e-07, | |
| "loss": 0.0092, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 8.061810154525386, | |
| "grad_norm": 1.0441016719090177, | |
| "learning_rate": 9.364623695550979e-07, | |
| "loss": 0.0085, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 8.070640176600442, | |
| "grad_norm": 1.527007095102573, | |
| "learning_rate": 9.281281733115288e-07, | |
| "loss": 0.0062, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 8.079470198675496, | |
| "grad_norm": 1.322023181950295, | |
| "learning_rate": 9.198274329464929e-07, | |
| "loss": 0.0078, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 8.088300220750552, | |
| "grad_norm": 1.808110774556195, | |
| "learning_rate": 9.115602166616805e-07, | |
| "loss": 0.009, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 8.097130242825607, | |
| "grad_norm": 1.884820688944253, | |
| "learning_rate": 9.033265923833446e-07, | |
| "loss": 0.0111, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 8.105960264900663, | |
| "grad_norm": 5.222067863820351, | |
| "learning_rate": 8.951266277617326e-07, | |
| "loss": 0.0098, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 8.114790286975717, | |
| "grad_norm": 1.4923255141846794, | |
| "learning_rate": 8.869603901705287e-07, | |
| "loss": 0.0083, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 8.123620309050773, | |
| "grad_norm": 1.561848050316407, | |
| "learning_rate": 8.78827946706311e-07, | |
| "loss": 0.0068, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 8.132450331125828, | |
| "grad_norm": 2.162317799091395, | |
| "learning_rate": 8.707293641879888e-07, | |
| "loss": 0.0074, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 8.141280353200884, | |
| "grad_norm": 1.9874712733917315, | |
| "learning_rate": 8.626647091562612e-07, | |
| "loss": 0.0114, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 8.150110375275938, | |
| "grad_norm": 2.0943718459590337, | |
| "learning_rate": 8.546340478730647e-07, | |
| "loss": 0.0078, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 8.158940397350994, | |
| "grad_norm": 1.3673660557038076, | |
| "learning_rate": 8.466374463210348e-07, | |
| "loss": 0.0087, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 8.167770419426049, | |
| "grad_norm": 1.6761272635756412, | |
| "learning_rate": 8.386749702029578e-07, | |
| "loss": 0.0089, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 8.176600441501105, | |
| "grad_norm": 1.6665924733012645, | |
| "learning_rate": 8.307466849412365e-07, | |
| "loss": 0.0058, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 8.185430463576159, | |
| "grad_norm": 1.5188450405366778, | |
| "learning_rate": 8.228526556773486e-07, | |
| "loss": 0.0084, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 8.194260485651213, | |
| "grad_norm": 7.9268953076354185, | |
| "learning_rate": 8.149929472713126e-07, | |
| "loss": 0.0125, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 8.20309050772627, | |
| "grad_norm": 2.0104696255501375, | |
| "learning_rate": 8.071676243011556e-07, | |
| "loss": 0.0104, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 8.211920529801324, | |
| "grad_norm": 1.2932107196941836, | |
| "learning_rate": 7.993767510623834e-07, | |
| "loss": 0.0067, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 8.22075055187638, | |
| "grad_norm": 3.9671954240585143, | |
| "learning_rate": 7.91620391567448e-07, | |
| "loss": 0.0097, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 8.229580573951434, | |
| "grad_norm": 1.4295056007256273, | |
| "learning_rate": 7.838986095452311e-07, | |
| "loss": 0.0082, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 8.23841059602649, | |
| "grad_norm": 11.51906295848909, | |
| "learning_rate": 7.762114684405064e-07, | |
| "loss": 0.008, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 8.247240618101545, | |
| "grad_norm": 1.8867833389233322, | |
| "learning_rate": 7.685590314134294e-07, | |
| "loss": 0.0086, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 8.2560706401766, | |
| "grad_norm": 1.6644258333313648, | |
| "learning_rate": 7.609413613390199e-07, | |
| "loss": 0.0099, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 8.264900662251655, | |
| "grad_norm": 1.6556750947882068, | |
| "learning_rate": 7.533585208066302e-07, | |
| "loss": 0.0083, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 8.273730684326711, | |
| "grad_norm": 1.800392752517919, | |
| "learning_rate": 7.458105721194525e-07, | |
| "loss": 0.0077, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 8.282560706401766, | |
| "grad_norm": 2.8177788729531326, | |
| "learning_rate": 7.382975772939866e-07, | |
| "loss": 0.0134, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 8.291390728476822, | |
| "grad_norm": 1.2225627298623025, | |
| "learning_rate": 7.308195980595462e-07, | |
| "loss": 0.0067, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 8.300220750551876, | |
| "grad_norm": 1.5252393790858616, | |
| "learning_rate": 7.233766958577421e-07, | |
| "loss": 0.0059, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 8.309050772626932, | |
| "grad_norm": 5.528368862350887, | |
| "learning_rate": 7.159689318419777e-07, | |
| "loss": 0.0116, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 8.317880794701987, | |
| "grad_norm": 1.23171581946235, | |
| "learning_rate": 7.085963668769552e-07, | |
| "loss": 0.0058, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 8.326710816777043, | |
| "grad_norm": 4.182313247524454, | |
| "learning_rate": 7.012590615381654e-07, | |
| "loss": 0.0065, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 8.335540838852097, | |
| "grad_norm": 1.7272665433036212, | |
| "learning_rate": 6.939570761113939e-07, | |
| "loss": 0.0096, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 8.344370860927153, | |
| "grad_norm": 2.837511603762501, | |
| "learning_rate": 6.866904705922284e-07, | |
| "loss": 0.0099, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 8.353200883002208, | |
| "grad_norm": 1.9424168876082293, | |
| "learning_rate": 6.794593046855613e-07, | |
| "loss": 0.0082, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 8.362030905077262, | |
| "grad_norm": 1.789363790465278, | |
| "learning_rate": 6.722636378051011e-07, | |
| "loss": 0.0076, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 8.370860927152318, | |
| "grad_norm": 1.7961195151994216, | |
| "learning_rate": 6.651035290728858e-07, | |
| "loss": 0.0097, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 8.379690949227372, | |
| "grad_norm": 3.5193218300229767, | |
| "learning_rate": 6.579790373187944e-07, | |
| "loss": 0.0062, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 8.388520971302428, | |
| "grad_norm": 2.6219461290045234, | |
| "learning_rate": 6.508902210800649e-07, | |
| "loss": 0.0086, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 8.397350993377483, | |
| "grad_norm": 1.5054755510817834, | |
| "learning_rate": 6.43837138600813e-07, | |
| "loss": 0.0068, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 8.406181015452539, | |
| "grad_norm": 1.2785143149456388, | |
| "learning_rate": 6.36819847831554e-07, | |
| "loss": 0.0067, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 8.415011037527593, | |
| "grad_norm": 1.286133310602691, | |
| "learning_rate": 6.298384064287261e-07, | |
| "loss": 0.0063, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 8.42384105960265, | |
| "grad_norm": 1.9538071544134283, | |
| "learning_rate": 6.228928717542205e-07, | |
| "loss": 0.0094, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 8.432671081677704, | |
| "grad_norm": 1.4973697003418378, | |
| "learning_rate": 6.159833008748988e-07, | |
| "loss": 0.0061, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 8.44150110375276, | |
| "grad_norm": 2.011947552820601, | |
| "learning_rate": 6.091097505621374e-07, | |
| "loss": 0.0091, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 8.450331125827814, | |
| "grad_norm": 1.5223779669820798, | |
| "learning_rate": 6.022722772913581e-07, | |
| "loss": 0.0068, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 8.45916114790287, | |
| "grad_norm": 1.6701309181354644, | |
| "learning_rate": 5.954709372415524e-07, | |
| "loss": 0.0072, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 8.467991169977925, | |
| "grad_norm": 1.2600493296646673, | |
| "learning_rate": 5.887057862948403e-07, | |
| "loss": 0.005, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 8.47682119205298, | |
| "grad_norm": 1.479733222516066, | |
| "learning_rate": 5.819768800359882e-07, | |
| "loss": 0.006, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.485651214128035, | |
| "grad_norm": 2.3916841797239203, | |
| "learning_rate": 5.752842737519743e-07, | |
| "loss": 0.0102, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 8.494481236203091, | |
| "grad_norm": 1.3596706350545904, | |
| "learning_rate": 5.686280224315189e-07, | |
| "loss": 0.0079, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 8.503311258278146, | |
| "grad_norm": 2.9243317528540502, | |
| "learning_rate": 5.62008180764635e-07, | |
| "loss": 0.0092, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 8.512141280353202, | |
| "grad_norm": 1.9050107889088481, | |
| "learning_rate": 5.554248031421872e-07, | |
| "loss": 0.008, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 8.520971302428256, | |
| "grad_norm": 1.5685044235666967, | |
| "learning_rate": 5.488779436554359e-07, | |
| "loss": 0.0057, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 8.52980132450331, | |
| "grad_norm": 2.1213280526372356, | |
| "learning_rate": 5.423676560955976e-07, | |
| "loss": 0.0067, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 8.538631346578367, | |
| "grad_norm": 1.5391693052825097, | |
| "learning_rate": 5.358939939534002e-07, | |
| "loss": 0.0066, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 8.54746136865342, | |
| "grad_norm": 1.999296080204376, | |
| "learning_rate": 5.294570104186436e-07, | |
| "loss": 0.0088, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 8.556291390728477, | |
| "grad_norm": 1.7744438943867837, | |
| "learning_rate": 5.230567583797674e-07, | |
| "loss": 0.0084, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 8.565121412803531, | |
| "grad_norm": 1.787302435136351, | |
| "learning_rate": 5.166932904234101e-07, | |
| "loss": 0.0072, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 8.573951434878587, | |
| "grad_norm": 1.806438247395929, | |
| "learning_rate": 5.103666588339812e-07, | |
| "loss": 0.0084, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 8.582781456953642, | |
| "grad_norm": 2.5509700305938665, | |
| "learning_rate": 5.040769155932285e-07, | |
| "loss": 0.0116, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 8.591611479028698, | |
| "grad_norm": 2.8299817396227644, | |
| "learning_rate": 4.978241123798133e-07, | |
| "loss": 0.0095, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 8.600441501103752, | |
| "grad_norm": 2.1420639341772216, | |
| "learning_rate": 4.916083005688865e-07, | |
| "loss": 0.0109, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 8.609271523178808, | |
| "grad_norm": 4.136222751554592, | |
| "learning_rate": 4.85429531231662e-07, | |
| "loss": 0.0091, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 8.618101545253863, | |
| "grad_norm": 1.3874549800784268, | |
| "learning_rate": 4.792878551350055e-07, | |
| "loss": 0.0075, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.626931567328919, | |
| "grad_norm": 1.8402714605625108, | |
| "learning_rate": 4.7318332274100595e-07, | |
| "loss": 0.0061, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 8.635761589403973, | |
| "grad_norm": 1.568407987242629, | |
| "learning_rate": 4.671159842065698e-07, | |
| "loss": 0.0097, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 8.64459161147903, | |
| "grad_norm": 1.593952578835664, | |
| "learning_rate": 4.6108588938300725e-07, | |
| "loss": 0.0071, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 8.653421633554084, | |
| "grad_norm": 5.240056073675244, | |
| "learning_rate": 4.5509308781561846e-07, | |
| "loss": 0.0081, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.66225165562914, | |
| "grad_norm": 3.278304982478669, | |
| "learning_rate": 4.4913762874329527e-07, | |
| "loss": 0.0105, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 8.671081677704194, | |
| "grad_norm": 1.9001265439536104, | |
| "learning_rate": 4.4321956109810327e-07, | |
| "loss": 0.0122, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 8.67991169977925, | |
| "grad_norm": 4.087562508063904, | |
| "learning_rate": 4.3733893350489386e-07, | |
| "loss": 0.0097, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 8.688741721854305, | |
| "grad_norm": 1.3913865427325924, | |
| "learning_rate": 4.314957942808956e-07, | |
| "loss": 0.0081, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 8.697571743929359, | |
| "grad_norm": 1.801118499567221, | |
| "learning_rate": 4.2569019143531845e-07, | |
| "loss": 0.0087, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 8.706401766004415, | |
| "grad_norm": 3.1681646901425595, | |
| "learning_rate": 4.199221726689634e-07, | |
| "loss": 0.0065, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 8.71523178807947, | |
| "grad_norm": 1.8559490757415693, | |
| "learning_rate": 4.1419178537382756e-07, | |
| "loss": 0.0075, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 8.724061810154526, | |
| "grad_norm": 1.4108941305897273, | |
| "learning_rate": 4.0849907663271346e-07, | |
| "loss": 0.0081, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 8.73289183222958, | |
| "grad_norm": 2.0050699380763697, | |
| "learning_rate": 4.028440932188465e-07, | |
| "loss": 0.0074, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 8.741721854304636, | |
| "grad_norm": 2.516758153283756, | |
| "learning_rate": 3.972268815954833e-07, | |
| "loss": 0.0192, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 8.75055187637969, | |
| "grad_norm": 4.258771207036382, | |
| "learning_rate": 3.916474879155402e-07, | |
| "loss": 0.0081, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 8.759381898454746, | |
| "grad_norm": 1.883683920095607, | |
| "learning_rate": 3.8610595802120564e-07, | |
| "loss": 0.0096, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 8.7682119205298, | |
| "grad_norm": 1.6034464568127333, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.0074, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 8.777041942604857, | |
| "grad_norm": 2.7127973138685224, | |
| "learning_rate": 3.751366714022342e-07, | |
| "loss": 0.0075, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 8.785871964679911, | |
| "grad_norm": 1.3352905399398, | |
| "learning_rate": 3.6970900480497287e-07, | |
| "loss": 0.006, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 8.794701986754967, | |
| "grad_norm": 1.2277884898477545, | |
| "learning_rate": 3.643193822473301e-07, | |
| "loss": 0.0052, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 8.803532008830022, | |
| "grad_norm": 1.7466179933217658, | |
| "learning_rate": 3.5896784801227046e-07, | |
| "loss": 0.0064, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 8.812362030905078, | |
| "grad_norm": 1.3488521850304223, | |
| "learning_rate": 3.5365444606981434e-07, | |
| "loss": 0.0046, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 8.821192052980132, | |
| "grad_norm": 1.1864613104433188, | |
| "learning_rate": 3.4837922007667e-07, | |
| "loss": 0.0052, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 8.830022075055188, | |
| "grad_norm": 1.6044723168492465, | |
| "learning_rate": 3.4314221337588217e-07, | |
| "loss": 0.0087, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.838852097130243, | |
| "grad_norm": 2.0336967189490327, | |
| "learning_rate": 3.379434689964728e-07, | |
| "loss": 0.0085, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 8.847682119205299, | |
| "grad_norm": 2.444042213569914, | |
| "learning_rate": 3.3278302965308593e-07, | |
| "loss": 0.0077, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 8.856512141280353, | |
| "grad_norm": 1.5583088114691361, | |
| "learning_rate": 3.276609377456419e-07, | |
| "loss": 0.008, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 8.865342163355407, | |
| "grad_norm": 1.3298586801450094, | |
| "learning_rate": 3.2257723535898177e-07, | |
| "loss": 0.008, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 8.874172185430464, | |
| "grad_norm": 1.860676668053892, | |
| "learning_rate": 3.1753196426252573e-07, | |
| "loss": 0.0096, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 8.883002207505518, | |
| "grad_norm": 1.9067443672345188, | |
| "learning_rate": 3.125251659099332e-07, | |
| "loss": 0.0093, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 8.891832229580574, | |
| "grad_norm": 2.086498283141313, | |
| "learning_rate": 3.0755688143875253e-07, | |
| "loss": 0.0092, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 8.900662251655628, | |
| "grad_norm": 1.3656695975208246, | |
| "learning_rate": 3.026271516700946e-07, | |
| "loss": 0.0051, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 8.909492273730685, | |
| "grad_norm": 1.3317150264636313, | |
| "learning_rate": 2.9773601710828937e-07, | |
| "loss": 0.0048, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 8.918322295805739, | |
| "grad_norm": 1.1977480165931524, | |
| "learning_rate": 2.928835179405548e-07, | |
| "loss": 0.006, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.927152317880795, | |
| "grad_norm": 1.1261365470696467, | |
| "learning_rate": 2.8806969403666897e-07, | |
| "loss": 0.0064, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 8.93598233995585, | |
| "grad_norm": 2.051586500624973, | |
| "learning_rate": 2.8329458494863846e-07, | |
| "loss": 0.0053, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 8.944812362030905, | |
| "grad_norm": 2.911804488736531, | |
| "learning_rate": 2.7855822991037895e-07, | |
| "loss": 0.008, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 8.95364238410596, | |
| "grad_norm": 1.3207535673566064, | |
| "learning_rate": 2.738606678373873e-07, | |
| "loss": 0.0098, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 8.962472406181016, | |
| "grad_norm": 0.77785943385093, | |
| "learning_rate": 2.6920193732642594e-07, | |
| "loss": 0.0048, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 8.97130242825607, | |
| "grad_norm": 1.4521384438631695, | |
| "learning_rate": 2.6458207665520266e-07, | |
| "loss": 0.0057, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 8.980132450331126, | |
| "grad_norm": 1.4526428738205903, | |
| "learning_rate": 2.600011237820577e-07, | |
| "loss": 0.0078, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 8.98896247240618, | |
| "grad_norm": 1.3819720802342337, | |
| "learning_rate": 2.5545911634565266e-07, | |
| "loss": 0.0087, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 8.997792494481237, | |
| "grad_norm": 1.297569933767477, | |
| "learning_rate": 2.5095609166465805e-07, | |
| "loss": 0.0067, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 9.006622516556291, | |
| "grad_norm": 0.792907712282856, | |
| "learning_rate": 2.4649208673745317e-07, | |
| "loss": 0.0035, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 9.015452538631347, | |
| "grad_norm": 0.8378040152065405, | |
| "learning_rate": 2.420671382418122e-07, | |
| "loss": 0.0033, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 9.024282560706402, | |
| "grad_norm": 0.4814999947409412, | |
| "learning_rate": 2.3768128253461253e-07, | |
| "loss": 0.0022, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 9.033112582781458, | |
| "grad_norm": 0.8269140916055675, | |
| "learning_rate": 2.333345556515304e-07, | |
| "loss": 0.0032, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 9.041942604856512, | |
| "grad_norm": 0.5950843442771953, | |
| "learning_rate": 2.290269933067457e-07, | |
| "loss": 0.0033, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 9.050772626931566, | |
| "grad_norm": 0.853773234550327, | |
| "learning_rate": 2.2475863089265193e-07, | |
| "loss": 0.0029, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 9.059602649006623, | |
| "grad_norm": 1.1638310856407534, | |
| "learning_rate": 2.205295034795596e-07, | |
| "loss": 0.0031, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 9.068432671081677, | |
| "grad_norm": 1.2586900984371077, | |
| "learning_rate": 2.1633964581541212e-07, | |
| "loss": 0.0023, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 9.077262693156733, | |
| "grad_norm": 3.938052567593757, | |
| "learning_rate": 2.1218909232550156e-07, | |
| "loss": 0.0041, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 9.086092715231787, | |
| "grad_norm": 0.6754018372402402, | |
| "learning_rate": 2.0807787711217887e-07, | |
| "loss": 0.003, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 9.094922737306844, | |
| "grad_norm": 0.7536603680896521, | |
| "learning_rate": 2.0400603395458408e-07, | |
| "loss": 0.0026, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 9.103752759381898, | |
| "grad_norm": 1.1727061945693198, | |
| "learning_rate": 1.999735963083571e-07, | |
| "loss": 0.0036, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 9.112582781456954, | |
| "grad_norm": 0.7895373436299211, | |
| "learning_rate": 1.9598059730537465e-07, | |
| "loss": 0.0029, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 9.121412803532008, | |
| "grad_norm": 1.3191861405378926, | |
| "learning_rate": 1.9202706975346875e-07, | |
| "loss": 0.003, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 9.130242825607064, | |
| "grad_norm": 2.3985956634341927, | |
| "learning_rate": 1.881130461361591e-07, | |
| "loss": 0.0026, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 9.139072847682119, | |
| "grad_norm": 0.5808790596533611, | |
| "learning_rate": 1.8423855861239238e-07, | |
| "loss": 0.0036, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 9.147902869757175, | |
| "grad_norm": 0.9644447189163927, | |
| "learning_rate": 1.8040363901627001e-07, | |
| "loss": 0.0038, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 9.15673289183223, | |
| "grad_norm": 0.7845234749084711, | |
| "learning_rate": 1.7660831885679074e-07, | |
| "loss": 0.0034, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 9.165562913907285, | |
| "grad_norm": 0.9380752046416668, | |
| "learning_rate": 1.7285262931759084e-07, | |
| "loss": 0.0035, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 9.17439293598234, | |
| "grad_norm": 0.5554430646485682, | |
| "learning_rate": 1.6913660125668806e-07, | |
| "loss": 0.0025, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 9.183222958057396, | |
| "grad_norm": 1.24527330749529, | |
| "learning_rate": 1.654602652062276e-07, | |
| "loss": 0.004, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 9.19205298013245, | |
| "grad_norm": 1.5936290649516498, | |
| "learning_rate": 1.6182365137223266e-07, | |
| "loss": 0.0039, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 9.200883002207506, | |
| "grad_norm": 0.8490016124768498, | |
| "learning_rate": 1.5822678963435479e-07, | |
| "loss": 0.0028, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 9.20971302428256, | |
| "grad_norm": 0.979565463427429, | |
| "learning_rate": 1.5466970954562786e-07, | |
| "loss": 0.0027, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 9.218543046357617, | |
| "grad_norm": 0.659865609696697, | |
| "learning_rate": 1.5115244033222732e-07, | |
| "loss": 0.003, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 9.227373068432671, | |
| "grad_norm": 2.130077445744013, | |
| "learning_rate": 1.47675010893229e-07, | |
| "loss": 0.0043, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 9.236203090507725, | |
| "grad_norm": 2.542325080904649, | |
| "learning_rate": 1.4423744980037068e-07, | |
| "loss": 0.0026, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 9.245033112582782, | |
| "grad_norm": 1.2325673483443094, | |
| "learning_rate": 1.408397852978205e-07, | |
| "loss": 0.0039, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 9.253863134657836, | |
| "grad_norm": 0.40114081038550564, | |
| "learning_rate": 1.3748204530193987e-07, | |
| "loss": 0.0021, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 9.262693156732892, | |
| "grad_norm": 0.784869098937524, | |
| "learning_rate": 1.341642574010582e-07, | |
| "loss": 0.0024, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 9.271523178807946, | |
| "grad_norm": 1.9661787215807498, | |
| "learning_rate": 1.3088644885524637e-07, | |
| "loss": 0.0035, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 9.280353200883003, | |
| "grad_norm": 0.5392536865076786, | |
| "learning_rate": 1.27648646596088e-07, | |
| "loss": 0.0019, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 9.289183222958057, | |
| "grad_norm": 0.8040736649311998, | |
| "learning_rate": 1.2445087722646576e-07, | |
| "loss": 0.003, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 9.298013245033113, | |
| "grad_norm": 0.6160784160573874, | |
| "learning_rate": 1.212931670203338e-07, | |
| "loss": 0.0023, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 9.306843267108167, | |
| "grad_norm": 6.994993198287031, | |
| "learning_rate": 1.1817554192251002e-07, | |
| "loss": 0.0046, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 9.315673289183223, | |
| "grad_norm": 0.6796010220847084, | |
| "learning_rate": 1.1509802754845978e-07, | |
| "loss": 0.0034, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 9.324503311258278, | |
| "grad_norm": 1.2132628310167854, | |
| "learning_rate": 1.1206064918408143e-07, | |
| "loss": 0.0022, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 1.0126429350214408, | |
| "learning_rate": 1.0906343178550715e-07, | |
| "loss": 0.0036, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 9.342163355408388, | |
| "grad_norm": 0.5782259402303945, | |
| "learning_rate": 1.0610639997888917e-07, | |
| "loss": 0.0021, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 9.350993377483444, | |
| "grad_norm": 1.1728476138124861, | |
| "learning_rate": 1.0318957806020269e-07, | |
| "loss": 0.003, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 9.359823399558499, | |
| "grad_norm": 0.8926747140898281, | |
| "learning_rate": 1.0031298999504557e-07, | |
| "loss": 0.0024, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 9.368653421633555, | |
| "grad_norm": 1.0889544500641504, | |
| "learning_rate": 9.747665941843953e-08, | |
| "loss": 0.004, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 9.37748344370861, | |
| "grad_norm": 2.396863460721648, | |
| "learning_rate": 9.468060963463754e-08, | |
| "loss": 0.0029, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 9.386313465783665, | |
| "grad_norm": 0.7609162671501313, | |
| "learning_rate": 9.192486361693175e-08, | |
| "loss": 0.0028, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 9.39514348785872, | |
| "grad_norm": 0.9587801304321354, | |
| "learning_rate": 8.920944400746589e-08, | |
| "loss": 0.0026, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 9.403973509933774, | |
| "grad_norm": 1.239601885332687, | |
| "learning_rate": 8.653437311704648e-08, | |
| "loss": 0.0032, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 9.41280353200883, | |
| "grad_norm": 0.4353285405500985, | |
| "learning_rate": 8.38996729249636e-08, | |
| "loss": 0.0018, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 9.421633554083884, | |
| "grad_norm": 0.558435622083429, | |
| "learning_rate": 8.130536507880538e-08, | |
| "loss": 0.0028, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 9.43046357615894, | |
| "grad_norm": 0.6937357506246238, | |
| "learning_rate": 7.875147089428436e-08, | |
| "loss": 0.0028, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 9.439293598233995, | |
| "grad_norm": 0.7655478062452088, | |
| "learning_rate": 7.623801135506148e-08, | |
| "loss": 0.0023, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 9.448123620309051, | |
| "grad_norm": 0.9614689357256824, | |
| "learning_rate": 7.376500711257062e-08, | |
| "loss": 0.0026, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 9.456953642384105, | |
| "grad_norm": 0.7046079919509814, | |
| "learning_rate": 7.133247848585268e-08, | |
| "loss": 0.0032, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 9.465783664459162, | |
| "grad_norm": 0.9297323795035541, | |
| "learning_rate": 6.894044546138845e-08, | |
| "loss": 0.0029, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 9.474613686534216, | |
| "grad_norm": 0.8867212494844405, | |
| "learning_rate": 6.65889276929299e-08, | |
| "loss": 0.003, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 9.483443708609272, | |
| "grad_norm": 2.096643223334579, | |
| "learning_rate": 6.427794450134529e-08, | |
| "loss": 0.0025, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 9.492273730684326, | |
| "grad_norm": 1.0306091176110177, | |
| "learning_rate": 6.200751487445367e-08, | |
| "loss": 0.0046, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 9.501103752759382, | |
| "grad_norm": 1.014331915070882, | |
| "learning_rate": 5.977765746687569e-08, | |
| "loss": 0.0023, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 9.509933774834437, | |
| "grad_norm": 3.673154065443917, | |
| "learning_rate": 5.758839059987531e-08, | |
| "loss": 0.0043, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 9.518763796909493, | |
| "grad_norm": 0.5807594028862743, | |
| "learning_rate": 5.5439732261209356e-08, | |
| "loss": 0.0029, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 9.527593818984547, | |
| "grad_norm": 1.242363318839941, | |
| "learning_rate": 5.333170010498434e-08, | |
| "loss": 0.0029, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 9.536423841059603, | |
| "grad_norm": 0.6247380011491397, | |
| "learning_rate": 5.126431145150546e-08, | |
| "loss": 0.0024, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.545253863134658, | |
| "grad_norm": 1.0975436092612265, | |
| "learning_rate": 4.9237583287139454e-08, | |
| "loss": 0.0042, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 9.554083885209714, | |
| "grad_norm": 1.2079739433316141, | |
| "learning_rate": 4.72515322641709e-08, | |
| "loss": 0.0031, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 9.562913907284768, | |
| "grad_norm": 0.7613882964492794, | |
| "learning_rate": 4.530617470066834e-08, | |
| "loss": 0.0031, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 9.571743929359823, | |
| "grad_norm": 1.2387457682643679, | |
| "learning_rate": 4.340152658034835e-08, | |
| "loss": 0.0036, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 9.580573951434879, | |
| "grad_norm": 1.0608242023116785, | |
| "learning_rate": 4.153760355244507e-08, | |
| "loss": 0.0028, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 9.589403973509933, | |
| "grad_norm": 0.6370232957084921, | |
| "learning_rate": 3.971442093158195e-08, | |
| "loss": 0.0015, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 9.59823399558499, | |
| "grad_norm": 0.8193368479818421, | |
| "learning_rate": 3.7931993697644664e-08, | |
| "loss": 0.0028, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 9.607064017660043, | |
| "grad_norm": 1.2941407201695256, | |
| "learning_rate": 3.6190336495659504e-08, | |
| "loss": 0.0021, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 9.6158940397351, | |
| "grad_norm": 1.0375034894346489, | |
| "learning_rate": 3.448946363567296e-08, | |
| "loss": 0.0039, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 9.624724061810154, | |
| "grad_norm": 1.043687527074888, | |
| "learning_rate": 3.282938909263122e-08, | |
| "loss": 0.0041, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 9.63355408388521, | |
| "grad_norm": 1.4855826707518094, | |
| "learning_rate": 3.121012650627031e-08, | |
| "loss": 0.006, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 9.642384105960264, | |
| "grad_norm": 1.7283966454401098, | |
| "learning_rate": 2.9631689180999457e-08, | |
| "loss": 0.0029, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 9.65121412803532, | |
| "grad_norm": 1.271455283397445, | |
| "learning_rate": 2.8094090085795112e-08, | |
| "loss": 0.0036, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 9.660044150110375, | |
| "grad_norm": 0.7580687095180034, | |
| "learning_rate": 2.6597341854092685e-08, | |
| "loss": 0.0031, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 9.668874172185431, | |
| "grad_norm": 0.6883431962052938, | |
| "learning_rate": 2.514145678368163e-08, | |
| "loss": 0.0026, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 9.677704194260485, | |
| "grad_norm": 1.0652514963488309, | |
| "learning_rate": 2.3726446836608298e-08, | |
| "loss": 0.0024, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 9.686534216335541, | |
| "grad_norm": 0.8473764440047209, | |
| "learning_rate": 2.235232363907269e-08, | |
| "loss": 0.0027, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 9.695364238410596, | |
| "grad_norm": 1.2523167599432252, | |
| "learning_rate": 2.101909848133743e-08, | |
| "loss": 0.0021, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.704194260485652, | |
| "grad_norm": 0.911129829918068, | |
| "learning_rate": 1.9726782317632255e-08, | |
| "loss": 0.0018, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 9.713024282560706, | |
| "grad_norm": 0.9192458510730664, | |
| "learning_rate": 1.8475385766063002e-08, | |
| "loss": 0.0022, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.721854304635762, | |
| "grad_norm": 0.6610469616129485, | |
| "learning_rate": 1.7264919108529455e-08, | |
| "loss": 0.0025, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 9.730684326710817, | |
| "grad_norm": 0.7080936572847712, | |
| "learning_rate": 1.6095392290635393e-08, | |
| "loss": 0.002, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 9.739514348785871, | |
| "grad_norm": 1.0787509205951389, | |
| "learning_rate": 1.4966814921608674e-08, | |
| "loss": 0.0022, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 9.748344370860927, | |
| "grad_norm": 0.5862652270690087, | |
| "learning_rate": 1.3879196274224626e-08, | |
| "loss": 0.0018, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 9.757174392935982, | |
| "grad_norm": 0.8603115386937105, | |
| "learning_rate": 1.2832545284724995e-08, | |
| "loss": 0.0035, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 9.766004415011038, | |
| "grad_norm": 12.538876923821956, | |
| "learning_rate": 1.1826870552749669e-08, | |
| "loss": 0.0081, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 9.774834437086092, | |
| "grad_norm": 0.9243735095294218, | |
| "learning_rate": 1.0862180341263962e-08, | |
| "loss": 0.0031, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 9.783664459161148, | |
| "grad_norm": 1.1855461359872093, | |
| "learning_rate": 9.938482576487551e-09, | |
| "loss": 0.0036, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 9.792494481236202, | |
| "grad_norm": 1.345018294494326, | |
| "learning_rate": 9.055784847836202e-09, | |
| "loss": 0.0045, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 9.801324503311259, | |
| "grad_norm": 0.8163415746521859, | |
| "learning_rate": 8.214094407851814e-09, | |
| "loss": 0.0028, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 9.810154525386313, | |
| "grad_norm": 1.2667912702993898, | |
| "learning_rate": 7.413418172149689e-09, | |
| "loss": 0.0042, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 9.818984547461369, | |
| "grad_norm": 0.5885221490598476, | |
| "learning_rate": 6.6537627193558055e-09, | |
| "loss": 0.0022, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 9.827814569536423, | |
| "grad_norm": 1.0832213789329375, | |
| "learning_rate": 5.93513429105741e-09, | |
| "loss": 0.0028, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 9.83664459161148, | |
| "grad_norm": 0.8397940384714349, | |
| "learning_rate": 5.257538791749173e-09, | |
| "loss": 0.0023, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 9.845474613686534, | |
| "grad_norm": 1.028187655365053, | |
| "learning_rate": 4.6209817887848955e-09, | |
| "loss": 0.0027, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 9.85430463576159, | |
| "grad_norm": 1.572908153202505, | |
| "learning_rate": 4.025468512333098e-09, | |
| "loss": 0.0039, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 9.863134657836644, | |
| "grad_norm": 0.6206454486532419, | |
| "learning_rate": 3.471003855332611e-09, | |
| "loss": 0.0027, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 9.8719646799117, | |
| "grad_norm": 0.480850997778566, | |
| "learning_rate": 2.9575923734520562e-09, | |
| "loss": 0.002, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 9.880794701986755, | |
| "grad_norm": 1.097220607267479, | |
| "learning_rate": 2.4852382850554245e-09, | |
| "loss": 0.0028, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 9.889624724061811, | |
| "grad_norm": 0.6847361038908694, | |
| "learning_rate": 2.0539454711626663e-09, | |
| "loss": 0.003, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.898454746136865, | |
| "grad_norm": 0.43652180289021436, | |
| "learning_rate": 1.6637174754230435e-09, | |
| "loss": 0.0018, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 9.90728476821192, | |
| "grad_norm": 1.0842639854903184, | |
| "learning_rate": 1.3145575040801605e-09, | |
| "loss": 0.004, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 9.916114790286976, | |
| "grad_norm": 0.5558639714191853, | |
| "learning_rate": 1.0064684259525337e-09, | |
| "loss": 0.0022, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 9.92494481236203, | |
| "grad_norm": 1.2720016586301384, | |
| "learning_rate": 7.394527724030598e-10, | |
| "loss": 0.0041, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 9.933774834437086, | |
| "grad_norm": 1.6350616090766645, | |
| "learning_rate": 5.13512737324029e-10, | |
| "loss": 0.0029, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 9.94260485651214, | |
| "grad_norm": 0.6608522240240123, | |
| "learning_rate": 3.2865017711380955e-10, | |
| "loss": 0.0022, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 9.951434878587197, | |
| "grad_norm": 0.6803776235396081, | |
| "learning_rate": 1.848666106674113e-10, | |
| "loss": 0.003, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 9.960264900662251, | |
| "grad_norm": 0.6608776431366015, | |
| "learning_rate": 8.216321935816673e-11, | |
| "loss": 0.0022, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 9.969094922737307, | |
| "grad_norm": 0.6062467708083688, | |
| "learning_rate": 2.0540847032179955e-11, | |
| "loss": 0.0017, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 9.977924944812361, | |
| "grad_norm": 0.7336777043773473, | |
| "learning_rate": 0.0, | |
| "loss": 0.0021, | |
| "step": 1130 | |
| } | |
| ], | |
| "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 | |
| } | |