|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 1.2007874015748032, |
|
"eval_steps": 153, |
|
"global_step": 1220, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.000984251968503937, |
|
"grad_norm": NaN, |
|
"learning_rate": 0.0, |
|
"loss": 18.7427, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.001968503937007874, |
|
"grad_norm": NaN, |
|
"learning_rate": 0.0, |
|
"loss": 11.6434, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 0.002952755905511811, |
|
"grad_norm": Infinity, |
|
"learning_rate": 0.0, |
|
"loss": 7.4859, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 0.003937007874015748, |
|
"grad_norm": 31.9741268157959, |
|
"learning_rate": 9.940357852882705e-10, |
|
"loss": 7.3779, |
|
"step": 4 |
|
}, |
|
{ |
|
"epoch": 0.004921259842519685, |
|
"grad_norm": Infinity, |
|
"learning_rate": 9.940357852882705e-10, |
|
"loss": 17.5878, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 0.005905511811023622, |
|
"grad_norm": 40.571693420410156, |
|
"learning_rate": 1.988071570576541e-09, |
|
"loss": 8.4984, |
|
"step": 6 |
|
}, |
|
{ |
|
"epoch": 0.006889763779527559, |
|
"grad_norm": 44.114723205566406, |
|
"learning_rate": 2.9821073558648116e-09, |
|
"loss": 8.375, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 0.007874015748031496, |
|
"grad_norm": 39.95664978027344, |
|
"learning_rate": 3.976143141153082e-09, |
|
"loss": 7.3241, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 0.008858267716535433, |
|
"grad_norm": 54.75717544555664, |
|
"learning_rate": 4.970178926441353e-09, |
|
"loss": 10.3081, |
|
"step": 9 |
|
}, |
|
{ |
|
"epoch": 0.00984251968503937, |
|
"grad_norm": 48.129844665527344, |
|
"learning_rate": 5.964214711729623e-09, |
|
"loss": 8.5363, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.010826771653543307, |
|
"grad_norm": Infinity, |
|
"learning_rate": 5.964214711729623e-09, |
|
"loss": 17.2241, |
|
"step": 11 |
|
}, |
|
{ |
|
"epoch": 0.011811023622047244, |
|
"grad_norm": 37.79861068725586, |
|
"learning_rate": 6.9582504970178946e-09, |
|
"loss": 7.575, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 0.012795275590551181, |
|
"grad_norm": 37.935638427734375, |
|
"learning_rate": 7.952286282306164e-09, |
|
"loss": 9.1905, |
|
"step": 13 |
|
}, |
|
{ |
|
"epoch": 0.013779527559055118, |
|
"grad_norm": 59.75738525390625, |
|
"learning_rate": 8.946322067594435e-09, |
|
"loss": 11.7727, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 0.014763779527559055, |
|
"grad_norm": 46.43831253051758, |
|
"learning_rate": 9.940357852882705e-09, |
|
"loss": 9.5827, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 0.015748031496062992, |
|
"grad_norm": 36.75105667114258, |
|
"learning_rate": 1.0934393638170978e-08, |
|
"loss": 7.4432, |
|
"step": 16 |
|
}, |
|
{ |
|
"epoch": 0.01673228346456693, |
|
"grad_norm": 30.298437118530273, |
|
"learning_rate": 1.1928429423459246e-08, |
|
"loss": 7.1573, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 0.017716535433070866, |
|
"grad_norm": 96.71277618408203, |
|
"learning_rate": 1.2922465208747517e-08, |
|
"loss": 19.8016, |
|
"step": 18 |
|
}, |
|
{ |
|
"epoch": 0.018700787401574805, |
|
"grad_norm": 84.41168212890625, |
|
"learning_rate": 1.3916500994035789e-08, |
|
"loss": 19.5118, |
|
"step": 19 |
|
}, |
|
{ |
|
"epoch": 0.01968503937007874, |
|
"grad_norm": 35.94807815551758, |
|
"learning_rate": 1.4910536779324056e-08, |
|
"loss": 7.9062, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.02066929133858268, |
|
"grad_norm": 41.35914611816406, |
|
"learning_rate": 1.590457256461233e-08, |
|
"loss": 8.6791, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 0.021653543307086614, |
|
"grad_norm": 37.65285873413086, |
|
"learning_rate": 1.68986083499006e-08, |
|
"loss": 7.7318, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 0.022637795275590553, |
|
"grad_norm": 40.24616622924805, |
|
"learning_rate": 1.789264413518887e-08, |
|
"loss": 7.9319, |
|
"step": 23 |
|
}, |
|
{ |
|
"epoch": 0.023622047244094488, |
|
"grad_norm": 32.05257034301758, |
|
"learning_rate": 1.888667992047714e-08, |
|
"loss": 7.192, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 0.024606299212598427, |
|
"grad_norm": 87.60692596435547, |
|
"learning_rate": 1.988071570576541e-08, |
|
"loss": 15.5799, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.025590551181102362, |
|
"grad_norm": 46.73486328125, |
|
"learning_rate": 2.087475149105368e-08, |
|
"loss": 9.7859, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 0.0265748031496063, |
|
"grad_norm": 50.18130111694336, |
|
"learning_rate": 2.1868787276341955e-08, |
|
"loss": 9.9259, |
|
"step": 27 |
|
}, |
|
{ |
|
"epoch": 0.027559055118110236, |
|
"grad_norm": 29.22075653076172, |
|
"learning_rate": 2.2862823061630224e-08, |
|
"loss": 6.3076, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 0.028543307086614175, |
|
"grad_norm": 37.356693267822266, |
|
"learning_rate": 2.3856858846918493e-08, |
|
"loss": 7.4471, |
|
"step": 29 |
|
}, |
|
{ |
|
"epoch": 0.02952755905511811, |
|
"grad_norm": 35.80938720703125, |
|
"learning_rate": 2.4850894632206765e-08, |
|
"loss": 7.1246, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.03051181102362205, |
|
"grad_norm": 25.738622665405273, |
|
"learning_rate": 2.5844930417495034e-08, |
|
"loss": 6.5505, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 0.031496062992125984, |
|
"grad_norm": 83.83589935302734, |
|
"learning_rate": 2.6838966202783303e-08, |
|
"loss": 18.5194, |
|
"step": 32 |
|
}, |
|
{ |
|
"epoch": 0.03248031496062992, |
|
"grad_norm": 33.8433837890625, |
|
"learning_rate": 2.7833001988071578e-08, |
|
"loss": 7.0747, |
|
"step": 33 |
|
}, |
|
{ |
|
"epoch": 0.03346456692913386, |
|
"grad_norm": 74.36174011230469, |
|
"learning_rate": 2.8827037773359847e-08, |
|
"loss": 14.9456, |
|
"step": 34 |
|
}, |
|
{ |
|
"epoch": 0.0344488188976378, |
|
"grad_norm": 26.985536575317383, |
|
"learning_rate": 2.982107355864811e-08, |
|
"loss": 6.608, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 0.03543307086614173, |
|
"grad_norm": 40.41023635864258, |
|
"learning_rate": 3.081510934393639e-08, |
|
"loss": 8.4672, |
|
"step": 36 |
|
}, |
|
{ |
|
"epoch": 0.03641732283464567, |
|
"grad_norm": 33.05155944824219, |
|
"learning_rate": 3.180914512922466e-08, |
|
"loss": 6.8853, |
|
"step": 37 |
|
}, |
|
{ |
|
"epoch": 0.03740157480314961, |
|
"grad_norm": 74.3410873413086, |
|
"learning_rate": 3.280318091451293e-08, |
|
"loss": 13.6063, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 0.038385826771653545, |
|
"grad_norm": 35.997493743896484, |
|
"learning_rate": 3.37972166998012e-08, |
|
"loss": 7.2625, |
|
"step": 39 |
|
}, |
|
{ |
|
"epoch": 0.03937007874015748, |
|
"grad_norm": 31.67265510559082, |
|
"learning_rate": 3.479125248508947e-08, |
|
"loss": 6.2234, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.040354330708661415, |
|
"grad_norm": 78.0589599609375, |
|
"learning_rate": 3.578528827037774e-08, |
|
"loss": 14.9675, |
|
"step": 41 |
|
}, |
|
{ |
|
"epoch": 0.04133858267716536, |
|
"grad_norm": 27.73711395263672, |
|
"learning_rate": 3.6779324055666005e-08, |
|
"loss": 6.6038, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 0.04232283464566929, |
|
"grad_norm": 61.068092346191406, |
|
"learning_rate": 3.777335984095428e-08, |
|
"loss": 13.1173, |
|
"step": 43 |
|
}, |
|
{ |
|
"epoch": 0.04330708661417323, |
|
"grad_norm": 84.73564910888672, |
|
"learning_rate": 3.8767395626242556e-08, |
|
"loss": 16.6992, |
|
"step": 44 |
|
}, |
|
{ |
|
"epoch": 0.04429133858267716, |
|
"grad_norm": 27.289846420288086, |
|
"learning_rate": 3.976143141153082e-08, |
|
"loss": 6.4828, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 0.045275590551181105, |
|
"grad_norm": 25.921550750732422, |
|
"learning_rate": 4.0755467196819094e-08, |
|
"loss": 5.9815, |
|
"step": 46 |
|
}, |
|
{ |
|
"epoch": 0.04625984251968504, |
|
"grad_norm": 28.005834579467773, |
|
"learning_rate": 4.174950298210736e-08, |
|
"loss": 6.1738, |
|
"step": 47 |
|
}, |
|
{ |
|
"epoch": 0.047244094488188976, |
|
"grad_norm": 33.678253173828125, |
|
"learning_rate": 4.274353876739563e-08, |
|
"loss": 7.134, |
|
"step": 48 |
|
}, |
|
{ |
|
"epoch": 0.04822834645669291, |
|
"grad_norm": 50.81912612915039, |
|
"learning_rate": 4.373757455268391e-08, |
|
"loss": 9.3933, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 0.04921259842519685, |
|
"grad_norm": 53.16002655029297, |
|
"learning_rate": 4.4731610337972176e-08, |
|
"loss": 10.8085, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.05019685039370079, |
|
"grad_norm": 53.343692779541016, |
|
"learning_rate": 4.572564612326045e-08, |
|
"loss": 11.4172, |
|
"step": 51 |
|
}, |
|
{ |
|
"epoch": 0.051181102362204724, |
|
"grad_norm": 29.51181983947754, |
|
"learning_rate": 4.6719681908548713e-08, |
|
"loss": 7.3397, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 0.05216535433070866, |
|
"grad_norm": 28.54021644592285, |
|
"learning_rate": 4.7713717693836986e-08, |
|
"loss": 5.8851, |
|
"step": 53 |
|
}, |
|
{ |
|
"epoch": 0.0531496062992126, |
|
"grad_norm": 29.310325622558594, |
|
"learning_rate": 4.870775347912525e-08, |
|
"loss": 6.8105, |
|
"step": 54 |
|
}, |
|
{ |
|
"epoch": 0.054133858267716536, |
|
"grad_norm": 19.80927276611328, |
|
"learning_rate": 4.970178926441353e-08, |
|
"loss": 5.3637, |
|
"step": 55 |
|
}, |
|
{ |
|
"epoch": 0.05511811023622047, |
|
"grad_norm": 30.965606689453125, |
|
"learning_rate": 5.06958250497018e-08, |
|
"loss": 6.2628, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 0.05610236220472441, |
|
"grad_norm": 21.39442253112793, |
|
"learning_rate": 5.168986083499007e-08, |
|
"loss": 6.0039, |
|
"step": 57 |
|
}, |
|
{ |
|
"epoch": 0.05708661417322835, |
|
"grad_norm": 38.37812805175781, |
|
"learning_rate": 5.268389662027834e-08, |
|
"loss": 7.5859, |
|
"step": 58 |
|
}, |
|
{ |
|
"epoch": 0.058070866141732284, |
|
"grad_norm": 22.37571144104004, |
|
"learning_rate": 5.3677932405566605e-08, |
|
"loss": 6.0802, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 0.05905511811023622, |
|
"grad_norm": 27.690608978271484, |
|
"learning_rate": 5.467196819085488e-08, |
|
"loss": 5.5822, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.060039370078740155, |
|
"grad_norm": 23.312894821166992, |
|
"learning_rate": 5.5666003976143156e-08, |
|
"loss": 5.8773, |
|
"step": 61 |
|
}, |
|
{ |
|
"epoch": 0.0610236220472441, |
|
"grad_norm": 19.575923919677734, |
|
"learning_rate": 5.666003976143142e-08, |
|
"loss": 6.0814, |
|
"step": 62 |
|
}, |
|
{ |
|
"epoch": 0.06200787401574803, |
|
"grad_norm": 20.126571655273438, |
|
"learning_rate": 5.7654075546719694e-08, |
|
"loss": 5.4483, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 0.06299212598425197, |
|
"grad_norm": 57.36194610595703, |
|
"learning_rate": 5.864811133200796e-08, |
|
"loss": 10.2506, |
|
"step": 64 |
|
}, |
|
{ |
|
"epoch": 0.0639763779527559, |
|
"grad_norm": 48.520416259765625, |
|
"learning_rate": 5.964214711729623e-08, |
|
"loss": 10.5976, |
|
"step": 65 |
|
}, |
|
{ |
|
"epoch": 0.06496062992125984, |
|
"grad_norm": 24.259727478027344, |
|
"learning_rate": 6.06361829025845e-08, |
|
"loss": 6.9942, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 0.06594488188976377, |
|
"grad_norm": 18.457971572875977, |
|
"learning_rate": 6.163021868787278e-08, |
|
"loss": 5.4813, |
|
"step": 67 |
|
}, |
|
{ |
|
"epoch": 0.06692913385826772, |
|
"grad_norm": 27.634117126464844, |
|
"learning_rate": 6.262425447316104e-08, |
|
"loss": 7.045, |
|
"step": 68 |
|
}, |
|
{ |
|
"epoch": 0.06791338582677166, |
|
"grad_norm": 18.592876434326172, |
|
"learning_rate": 6.361829025844931e-08, |
|
"loss": 5.8549, |
|
"step": 69 |
|
}, |
|
{ |
|
"epoch": 0.0688976377952756, |
|
"grad_norm": 37.92665481567383, |
|
"learning_rate": 6.461232604373759e-08, |
|
"loss": 8.8514, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.06988188976377953, |
|
"grad_norm": 19.356266021728516, |
|
"learning_rate": 6.560636182902586e-08, |
|
"loss": 5.2557, |
|
"step": 71 |
|
}, |
|
{ |
|
"epoch": 0.07086614173228346, |
|
"grad_norm": 13.698501586914062, |
|
"learning_rate": 6.660039761431412e-08, |
|
"loss": 5.1181, |
|
"step": 72 |
|
}, |
|
{ |
|
"epoch": 0.0718503937007874, |
|
"grad_norm": 19.626392364501953, |
|
"learning_rate": 6.75944333996024e-08, |
|
"loss": 5.5331, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 0.07283464566929133, |
|
"grad_norm": 15.392860412597656, |
|
"learning_rate": 6.858846918489067e-08, |
|
"loss": 5.5944, |
|
"step": 74 |
|
}, |
|
{ |
|
"epoch": 0.07381889763779527, |
|
"grad_norm": 16.435239791870117, |
|
"learning_rate": 6.958250497017893e-08, |
|
"loss": 4.6332, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.07480314960629922, |
|
"grad_norm": 16.286462783813477, |
|
"learning_rate": 7.057654075546721e-08, |
|
"loss": 4.9532, |
|
"step": 76 |
|
}, |
|
{ |
|
"epoch": 0.07578740157480315, |
|
"grad_norm": 17.462007522583008, |
|
"learning_rate": 7.157057654075548e-08, |
|
"loss": 5.055, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 0.07677165354330709, |
|
"grad_norm": 15.096108436584473, |
|
"learning_rate": 7.256461232604374e-08, |
|
"loss": 4.5005, |
|
"step": 78 |
|
}, |
|
{ |
|
"epoch": 0.07775590551181102, |
|
"grad_norm": 14.233656883239746, |
|
"learning_rate": 7.355864811133201e-08, |
|
"loss": 5.1997, |
|
"step": 79 |
|
}, |
|
{ |
|
"epoch": 0.07874015748031496, |
|
"grad_norm": 19.416706085205078, |
|
"learning_rate": 7.455268389662029e-08, |
|
"loss": 5.1479, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.0797244094488189, |
|
"grad_norm": 18.06542205810547, |
|
"learning_rate": 7.554671968190855e-08, |
|
"loss": 5.1777, |
|
"step": 81 |
|
}, |
|
{ |
|
"epoch": 0.08070866141732283, |
|
"grad_norm": 16.027755737304688, |
|
"learning_rate": 7.654075546719683e-08, |
|
"loss": 5.5565, |
|
"step": 82 |
|
}, |
|
{ |
|
"epoch": 0.08169291338582677, |
|
"grad_norm": 16.291139602661133, |
|
"learning_rate": 7.753479125248511e-08, |
|
"loss": 4.6999, |
|
"step": 83 |
|
}, |
|
{ |
|
"epoch": 0.08267716535433071, |
|
"grad_norm": 16.005495071411133, |
|
"learning_rate": 7.852882703777338e-08, |
|
"loss": 5.0681, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 0.08366141732283465, |
|
"grad_norm": 19.505733489990234, |
|
"learning_rate": 7.952286282306164e-08, |
|
"loss": 5.2208, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 0.08464566929133858, |
|
"grad_norm": 18.256221771240234, |
|
"learning_rate": 8.051689860834992e-08, |
|
"loss": 4.56, |
|
"step": 86 |
|
}, |
|
{ |
|
"epoch": 0.08562992125984252, |
|
"grad_norm": 14.99673843383789, |
|
"learning_rate": 8.151093439363819e-08, |
|
"loss": 4.6793, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 0.08661417322834646, |
|
"grad_norm": 13.485681533813477, |
|
"learning_rate": 8.250497017892645e-08, |
|
"loss": 4.4611, |
|
"step": 88 |
|
}, |
|
{ |
|
"epoch": 0.08759842519685039, |
|
"grad_norm": 52.90532302856445, |
|
"learning_rate": 8.349900596421472e-08, |
|
"loss": 9.623, |
|
"step": 89 |
|
}, |
|
{ |
|
"epoch": 0.08858267716535433, |
|
"grad_norm": 18.402536392211914, |
|
"learning_rate": 8.4493041749503e-08, |
|
"loss": 5.0316, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.08956692913385826, |
|
"grad_norm": 12.668858528137207, |
|
"learning_rate": 8.548707753479126e-08, |
|
"loss": 4.1771, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 0.09055118110236221, |
|
"grad_norm": 17.02667236328125, |
|
"learning_rate": 8.648111332007953e-08, |
|
"loss": 4.9652, |
|
"step": 92 |
|
}, |
|
{ |
|
"epoch": 0.09153543307086615, |
|
"grad_norm": 51.03838348388672, |
|
"learning_rate": 8.747514910536782e-08, |
|
"loss": 8.7432, |
|
"step": 93 |
|
}, |
|
{ |
|
"epoch": 0.09251968503937008, |
|
"grad_norm": 15.051873207092285, |
|
"learning_rate": 8.846918489065609e-08, |
|
"loss": 4.6234, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 0.09350393700787402, |
|
"grad_norm": 15.075822830200195, |
|
"learning_rate": 8.946322067594435e-08, |
|
"loss": 4.4016, |
|
"step": 95 |
|
}, |
|
{ |
|
"epoch": 0.09448818897637795, |
|
"grad_norm": 14.73617172241211, |
|
"learning_rate": 9.045725646123262e-08, |
|
"loss": 4.9903, |
|
"step": 96 |
|
}, |
|
{ |
|
"epoch": 0.09547244094488189, |
|
"grad_norm": 12.564230918884277, |
|
"learning_rate": 9.14512922465209e-08, |
|
"loss": 4.5606, |
|
"step": 97 |
|
}, |
|
{ |
|
"epoch": 0.09645669291338582, |
|
"grad_norm": 12.062474250793457, |
|
"learning_rate": 9.244532803180916e-08, |
|
"loss": 4.9534, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 0.09744094488188976, |
|
"grad_norm": 42.591461181640625, |
|
"learning_rate": 9.343936381709743e-08, |
|
"loss": 8.1838, |
|
"step": 99 |
|
}, |
|
{ |
|
"epoch": 0.0984251968503937, |
|
"grad_norm": 12.18688678741455, |
|
"learning_rate": 9.44333996023857e-08, |
|
"loss": 4.9736, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.09940944881889764, |
|
"grad_norm": 14.301939964294434, |
|
"learning_rate": 9.542743538767397e-08, |
|
"loss": 4.4733, |
|
"step": 101 |
|
}, |
|
{ |
|
"epoch": 0.10039370078740158, |
|
"grad_norm": 15.062311172485352, |
|
"learning_rate": 9.642147117296224e-08, |
|
"loss": 4.9725, |
|
"step": 102 |
|
}, |
|
{ |
|
"epoch": 0.10137795275590551, |
|
"grad_norm": 11.707622528076172, |
|
"learning_rate": 9.74155069582505e-08, |
|
"loss": 4.5861, |
|
"step": 103 |
|
}, |
|
{ |
|
"epoch": 0.10236220472440945, |
|
"grad_norm": 42.08045196533203, |
|
"learning_rate": 9.840954274353878e-08, |
|
"loss": 7.7634, |
|
"step": 104 |
|
}, |
|
{ |
|
"epoch": 0.10334645669291338, |
|
"grad_norm": 13.918876647949219, |
|
"learning_rate": 9.940357852882706e-08, |
|
"loss": 4.9915, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 0.10433070866141732, |
|
"grad_norm": 12.0372953414917, |
|
"learning_rate": 1.0039761431411533e-07, |
|
"loss": 5.1391, |
|
"step": 106 |
|
}, |
|
{ |
|
"epoch": 0.10531496062992125, |
|
"grad_norm": 16.739591598510742, |
|
"learning_rate": 1.013916500994036e-07, |
|
"loss": 5.0157, |
|
"step": 107 |
|
}, |
|
{ |
|
"epoch": 0.1062992125984252, |
|
"grad_norm": 16.80640983581543, |
|
"learning_rate": 1.0238568588469187e-07, |
|
"loss": 4.0982, |
|
"step": 108 |
|
}, |
|
{ |
|
"epoch": 0.10728346456692914, |
|
"grad_norm": 10.70914077758789, |
|
"learning_rate": 1.0337972166998014e-07, |
|
"loss": 4.2178, |
|
"step": 109 |
|
}, |
|
{ |
|
"epoch": 0.10826771653543307, |
|
"grad_norm": 11.523154258728027, |
|
"learning_rate": 1.043737574552684e-07, |
|
"loss": 4.6193, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.10925196850393701, |
|
"grad_norm": 14.624194145202637, |
|
"learning_rate": 1.0536779324055668e-07, |
|
"loss": 4.7638, |
|
"step": 111 |
|
}, |
|
{ |
|
"epoch": 0.11023622047244094, |
|
"grad_norm": 11.593772888183594, |
|
"learning_rate": 1.0636182902584495e-07, |
|
"loss": 4.1207, |
|
"step": 112 |
|
}, |
|
{ |
|
"epoch": 0.11122047244094488, |
|
"grad_norm": 18.1988468170166, |
|
"learning_rate": 1.0735586481113321e-07, |
|
"loss": 5.2034, |
|
"step": 113 |
|
}, |
|
{ |
|
"epoch": 0.11220472440944881, |
|
"grad_norm": 14.054649353027344, |
|
"learning_rate": 1.0834990059642149e-07, |
|
"loss": 5.0693, |
|
"step": 114 |
|
}, |
|
{ |
|
"epoch": 0.11318897637795275, |
|
"grad_norm": 12.262086868286133, |
|
"learning_rate": 1.0934393638170976e-07, |
|
"loss": 4.7895, |
|
"step": 115 |
|
}, |
|
{ |
|
"epoch": 0.1141732283464567, |
|
"grad_norm": 12.955352783203125, |
|
"learning_rate": 1.1033797216699802e-07, |
|
"loss": 4.9486, |
|
"step": 116 |
|
}, |
|
{ |
|
"epoch": 0.11515748031496063, |
|
"grad_norm": 14.151504516601562, |
|
"learning_rate": 1.1133200795228631e-07, |
|
"loss": 4.6552, |
|
"step": 117 |
|
}, |
|
{ |
|
"epoch": 0.11614173228346457, |
|
"grad_norm": 11.60208511352539, |
|
"learning_rate": 1.1232604373757458e-07, |
|
"loss": 4.4555, |
|
"step": 118 |
|
}, |
|
{ |
|
"epoch": 0.1171259842519685, |
|
"grad_norm": 11.05080509185791, |
|
"learning_rate": 1.1332007952286284e-07, |
|
"loss": 4.8977, |
|
"step": 119 |
|
}, |
|
{ |
|
"epoch": 0.11811023622047244, |
|
"grad_norm": 43.464908599853516, |
|
"learning_rate": 1.1431411530815111e-07, |
|
"loss": 7.6836, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.11909448818897637, |
|
"grad_norm": 11.060967445373535, |
|
"learning_rate": 1.1530815109343939e-07, |
|
"loss": 4.8106, |
|
"step": 121 |
|
}, |
|
{ |
|
"epoch": 0.12007874015748031, |
|
"grad_norm": 11.406394004821777, |
|
"learning_rate": 1.1630218687872765e-07, |
|
"loss": 4.9958, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 0.12106299212598425, |
|
"grad_norm": 12.630412101745605, |
|
"learning_rate": 1.1729622266401592e-07, |
|
"loss": 4.4585, |
|
"step": 123 |
|
}, |
|
{ |
|
"epoch": 0.1220472440944882, |
|
"grad_norm": 42.50083541870117, |
|
"learning_rate": 1.182902584493042e-07, |
|
"loss": 7.5559, |
|
"step": 124 |
|
}, |
|
{ |
|
"epoch": 0.12303149606299213, |
|
"grad_norm": 12.229009628295898, |
|
"learning_rate": 1.1928429423459245e-07, |
|
"loss": 4.2636, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.12401574803149606, |
|
"grad_norm": 9.794486999511719, |
|
"learning_rate": 1.2027833001988073e-07, |
|
"loss": 4.0436, |
|
"step": 126 |
|
}, |
|
{ |
|
"epoch": 0.125, |
|
"grad_norm": 11.505509376525879, |
|
"learning_rate": 1.21272365805169e-07, |
|
"loss": 4.7416, |
|
"step": 127 |
|
}, |
|
{ |
|
"epoch": 0.12598425196850394, |
|
"grad_norm": 12.046993255615234, |
|
"learning_rate": 1.222664015904573e-07, |
|
"loss": 4.2215, |
|
"step": 128 |
|
}, |
|
{ |
|
"epoch": 0.12696850393700787, |
|
"grad_norm": 33.795265197753906, |
|
"learning_rate": 1.2326043737574557e-07, |
|
"loss": 6.3561, |
|
"step": 129 |
|
}, |
|
{ |
|
"epoch": 0.1279527559055118, |
|
"grad_norm": 33.15379333496094, |
|
"learning_rate": 1.2425447316103382e-07, |
|
"loss": 6.2299, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.12893700787401574, |
|
"grad_norm": 11.174860000610352, |
|
"learning_rate": 1.2524850894632207e-07, |
|
"loss": 4.3492, |
|
"step": 131 |
|
}, |
|
{ |
|
"epoch": 0.12992125984251968, |
|
"grad_norm": 10.696345329284668, |
|
"learning_rate": 1.2624254473161035e-07, |
|
"loss": 4.0216, |
|
"step": 132 |
|
}, |
|
{ |
|
"epoch": 0.1309055118110236, |
|
"grad_norm": 35.72112274169922, |
|
"learning_rate": 1.2723658051689863e-07, |
|
"loss": 6.963, |
|
"step": 133 |
|
}, |
|
{ |
|
"epoch": 0.13188976377952755, |
|
"grad_norm": 9.074990272521973, |
|
"learning_rate": 1.282306163021869e-07, |
|
"loss": 3.9474, |
|
"step": 134 |
|
}, |
|
{ |
|
"epoch": 0.1328740157480315, |
|
"grad_norm": 11.409770965576172, |
|
"learning_rate": 1.2922465208747519e-07, |
|
"loss": 4.3437, |
|
"step": 135 |
|
}, |
|
{ |
|
"epoch": 0.13385826771653545, |
|
"grad_norm": 10.008737564086914, |
|
"learning_rate": 1.3021868787276344e-07, |
|
"loss": 3.6267, |
|
"step": 136 |
|
}, |
|
{ |
|
"epoch": 0.13484251968503938, |
|
"grad_norm": 11.481470108032227, |
|
"learning_rate": 1.3121272365805172e-07, |
|
"loss": 3.9896, |
|
"step": 137 |
|
}, |
|
{ |
|
"epoch": 0.13582677165354332, |
|
"grad_norm": 12.856668472290039, |
|
"learning_rate": 1.3220675944333997e-07, |
|
"loss": 4.8156, |
|
"step": 138 |
|
}, |
|
{ |
|
"epoch": 0.13681102362204725, |
|
"grad_norm": 12.462326049804688, |
|
"learning_rate": 1.3320079522862825e-07, |
|
"loss": 4.9751, |
|
"step": 139 |
|
}, |
|
{ |
|
"epoch": 0.1377952755905512, |
|
"grad_norm": 13.430492401123047, |
|
"learning_rate": 1.3419483101391653e-07, |
|
"loss": 4.4144, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.13877952755905512, |
|
"grad_norm": 24.038984298706055, |
|
"learning_rate": 1.351888667992048e-07, |
|
"loss": 4.7213, |
|
"step": 141 |
|
}, |
|
{ |
|
"epoch": 0.13976377952755906, |
|
"grad_norm": 39.7520637512207, |
|
"learning_rate": 1.3618290258449306e-07, |
|
"loss": 6.6081, |
|
"step": 142 |
|
}, |
|
{ |
|
"epoch": 0.140748031496063, |
|
"grad_norm": 10.726151466369629, |
|
"learning_rate": 1.3717693836978134e-07, |
|
"loss": 4.2929, |
|
"step": 143 |
|
}, |
|
{ |
|
"epoch": 0.14173228346456693, |
|
"grad_norm": 11.658100128173828, |
|
"learning_rate": 1.381709741550696e-07, |
|
"loss": 4.2537, |
|
"step": 144 |
|
}, |
|
{ |
|
"epoch": 0.14271653543307086, |
|
"grad_norm": 9.662211418151855, |
|
"learning_rate": 1.3916500994035787e-07, |
|
"loss": 4.0647, |
|
"step": 145 |
|
}, |
|
{ |
|
"epoch": 0.1437007874015748, |
|
"grad_norm": 10.004919052124023, |
|
"learning_rate": 1.4015904572564615e-07, |
|
"loss": 3.937, |
|
"step": 146 |
|
}, |
|
{ |
|
"epoch": 0.14468503937007873, |
|
"grad_norm": 35.38102340698242, |
|
"learning_rate": 1.4115308151093443e-07, |
|
"loss": 5.6582, |
|
"step": 147 |
|
}, |
|
{ |
|
"epoch": 0.14566929133858267, |
|
"grad_norm": 11.280171394348145, |
|
"learning_rate": 1.421471172962227e-07, |
|
"loss": 4.2648, |
|
"step": 148 |
|
}, |
|
{ |
|
"epoch": 0.1466535433070866, |
|
"grad_norm": 10.569673538208008, |
|
"learning_rate": 1.4314115308151096e-07, |
|
"loss": 4.4429, |
|
"step": 149 |
|
}, |
|
{ |
|
"epoch": 0.14763779527559054, |
|
"grad_norm": 14.092275619506836, |
|
"learning_rate": 1.4413518886679924e-07, |
|
"loss": 3.6197, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.1486220472440945, |
|
"grad_norm": 8.960445404052734, |
|
"learning_rate": 1.451292246520875e-07, |
|
"loss": 3.7953, |
|
"step": 151 |
|
}, |
|
{ |
|
"epoch": 0.14960629921259844, |
|
"grad_norm": 11.712779998779297, |
|
"learning_rate": 1.4612326043737577e-07, |
|
"loss": 3.8175, |
|
"step": 152 |
|
}, |
|
{ |
|
"epoch": 0.15059055118110237, |
|
"grad_norm": 11.68183708190918, |
|
"learning_rate": 1.4711729622266402e-07, |
|
"loss": 4.5137, |
|
"step": 153 |
|
}, |
|
{ |
|
"epoch": 0.15059055118110237, |
|
"eval_Qnli-dev_cosine_accuracy": 0.58203125, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.9368094801902771, |
|
"eval_Qnli-dev_cosine_ap": 0.5484497034083067, |
|
"eval_Qnli-dev_cosine_f1": 0.6300268096514745, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.802739143371582, |
|
"eval_Qnli-dev_cosine_precision": 0.46078431372549017, |
|
"eval_Qnli-dev_cosine_recall": 0.9957627118644068, |
|
"eval_Qnli-dev_dot_accuracy": 0.58203125, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 719.7518310546875, |
|
"eval_Qnli-dev_dot_ap": 0.548461685358088, |
|
"eval_Qnli-dev_dot_f1": 0.6300268096514745, |
|
"eval_Qnli-dev_dot_f1_threshold": 616.7227783203125, |
|
"eval_Qnli-dev_dot_precision": 0.46078431372549017, |
|
"eval_Qnli-dev_dot_recall": 0.9957627118644068, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.58203125, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 9.853867530822754, |
|
"eval_Qnli-dev_euclidean_ap": 0.5484497034083067, |
|
"eval_Qnli-dev_euclidean_f1": 0.6300268096514745, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 17.40953254699707, |
|
"eval_Qnli-dev_euclidean_precision": 0.46078431372549017, |
|
"eval_Qnli-dev_euclidean_recall": 0.9957627118644068, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.607421875, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 182.1275177001953, |
|
"eval_Qnli-dev_manhattan_ap": 0.5750034744442096, |
|
"eval_Qnli-dev_manhattan_f1": 0.6303724928366763, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 230.0565185546875, |
|
"eval_Qnli-dev_manhattan_precision": 0.47619047619047616, |
|
"eval_Qnli-dev_manhattan_recall": 0.9322033898305084, |
|
"eval_Qnli-dev_max_accuracy": 0.607421875, |
|
"eval_Qnli-dev_max_accuracy_threshold": 719.7518310546875, |
|
"eval_Qnli-dev_max_ap": 0.5750034744442096, |
|
"eval_Qnli-dev_max_f1": 0.6303724928366763, |
|
"eval_Qnli-dev_max_f1_threshold": 616.7227783203125, |
|
"eval_Qnli-dev_max_precision": 0.47619047619047616, |
|
"eval_Qnli-dev_max_recall": 0.9957627118644068, |
|
"eval_allNLI-dev_cosine_accuracy": 0.66796875, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9721524119377136, |
|
"eval_allNLI-dev_cosine_ap": 0.3857994503224615, |
|
"eval_allNLI-dev_cosine_f1": 0.5029239766081871, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.821484386920929, |
|
"eval_allNLI-dev_cosine_precision": 0.33659491193737767, |
|
"eval_allNLI-dev_cosine_recall": 0.9942196531791907, |
|
"eval_allNLI-dev_dot_accuracy": 0.66796875, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 746.914794921875, |
|
"eval_allNLI-dev_dot_ap": 0.38572844452312516, |
|
"eval_allNLI-dev_dot_f1": 0.5029239766081871, |
|
"eval_allNLI-dev_dot_f1_threshold": 631.138916015625, |
|
"eval_allNLI-dev_dot_precision": 0.33659491193737767, |
|
"eval_allNLI-dev_dot_recall": 0.9942196531791907, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.66796875, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 6.541449546813965, |
|
"eval_allNLI-dev_euclidean_ap": 0.3858031188548441, |
|
"eval_allNLI-dev_euclidean_f1": 0.5029239766081871, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 16.558998107910156, |
|
"eval_allNLI-dev_euclidean_precision": 0.33659491193737767, |
|
"eval_allNLI-dev_euclidean_recall": 0.9942196531791907, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.666015625, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 95.24527740478516, |
|
"eval_allNLI-dev_manhattan_ap": 0.39193409293721965, |
|
"eval_allNLI-dev_manhattan_f1": 0.5045317220543807, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 254.973388671875, |
|
"eval_allNLI-dev_manhattan_precision": 0.34151329243353784, |
|
"eval_allNLI-dev_manhattan_recall": 0.9653179190751445, |
|
"eval_allNLI-dev_max_accuracy": 0.66796875, |
|
"eval_allNLI-dev_max_accuracy_threshold": 746.914794921875, |
|
"eval_allNLI-dev_max_ap": 0.39193409293721965, |
|
"eval_allNLI-dev_max_f1": 0.5045317220543807, |
|
"eval_allNLI-dev_max_f1_threshold": 631.138916015625, |
|
"eval_allNLI-dev_max_precision": 0.34151329243353784, |
|
"eval_allNLI-dev_max_recall": 0.9942196531791907, |
|
"eval_loss": 3.3210110664367676, |
|
"eval_runtime": 53.0927, |
|
"eval_samples_per_second": 31.341, |
|
"eval_sequential_score": 0.5750034744442096, |
|
"eval_steps_per_second": 0.132, |
|
"eval_sts-test_pearson_cosine": 0.12009124140478655, |
|
"eval_sts-test_pearson_dot": 0.11997652374043644, |
|
"eval_sts-test_pearson_euclidean": 0.15529980522625675, |
|
"eval_sts-test_pearson_manhattan": 0.18492770691981375, |
|
"eval_sts-test_pearson_max": 0.18492770691981375, |
|
"eval_sts-test_spearman_cosine": 0.180573622028628, |
|
"eval_sts-test_spearman_dot": 0.18041242798509616, |
|
"eval_sts-test_spearman_euclidean": 0.18058248277838349, |
|
"eval_sts-test_spearman_manhattan": 0.21139381574888486, |
|
"eval_sts-test_spearman_max": 0.21139381574888486, |
|
"step": 153 |
|
}, |
|
{ |
|
"epoch": 0.1515748031496063, |
|
"grad_norm": 12.480842590332031, |
|
"learning_rate": 1.4811133200795232e-07, |
|
"loss": 4.3528, |
|
"step": 154 |
|
}, |
|
{ |
|
"epoch": 0.15255905511811024, |
|
"grad_norm": 10.134323120117188, |
|
"learning_rate": 1.4910536779324058e-07, |
|
"loss": 3.6573, |
|
"step": 155 |
|
}, |
|
{ |
|
"epoch": 0.15354330708661418, |
|
"grad_norm": 11.572721481323242, |
|
"learning_rate": 1.5009940357852886e-07, |
|
"loss": 3.5248, |
|
"step": 156 |
|
}, |
|
{ |
|
"epoch": 0.1545275590551181, |
|
"grad_norm": 12.186054229736328, |
|
"learning_rate": 1.510934393638171e-07, |
|
"loss": 3.9275, |
|
"step": 157 |
|
}, |
|
{ |
|
"epoch": 0.15551181102362205, |
|
"grad_norm": 42.491851806640625, |
|
"learning_rate": 1.5208747514910539e-07, |
|
"loss": 7.1868, |
|
"step": 158 |
|
}, |
|
{ |
|
"epoch": 0.15649606299212598, |
|
"grad_norm": 10.867408752441406, |
|
"learning_rate": 1.5308151093439367e-07, |
|
"loss": 3.6294, |
|
"step": 159 |
|
}, |
|
{ |
|
"epoch": 0.15748031496062992, |
|
"grad_norm": 11.463488578796387, |
|
"learning_rate": 1.5407554671968192e-07, |
|
"loss": 3.6886, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.15846456692913385, |
|
"grad_norm": 10.748167991638184, |
|
"learning_rate": 1.5506958250497022e-07, |
|
"loss": 3.1873, |
|
"step": 161 |
|
}, |
|
{ |
|
"epoch": 0.1594488188976378, |
|
"grad_norm": 38.69701385498047, |
|
"learning_rate": 1.5606361829025848e-07, |
|
"loss": 6.1951, |
|
"step": 162 |
|
}, |
|
{ |
|
"epoch": 0.16043307086614172, |
|
"grad_norm": 11.077349662780762, |
|
"learning_rate": 1.5705765407554675e-07, |
|
"loss": 3.9747, |
|
"step": 163 |
|
}, |
|
{ |
|
"epoch": 0.16141732283464566, |
|
"grad_norm": 47.21928405761719, |
|
"learning_rate": 1.58051689860835e-07, |
|
"loss": 7.004, |
|
"step": 164 |
|
}, |
|
{ |
|
"epoch": 0.1624015748031496, |
|
"grad_norm": 15.23488712310791, |
|
"learning_rate": 1.5904572564612329e-07, |
|
"loss": 4.3221, |
|
"step": 165 |
|
}, |
|
{ |
|
"epoch": 0.16338582677165353, |
|
"grad_norm": 10.329791069030762, |
|
"learning_rate": 1.6003976143141154e-07, |
|
"loss": 3.5963, |
|
"step": 166 |
|
}, |
|
{ |
|
"epoch": 0.1643700787401575, |
|
"grad_norm": 9.624051094055176, |
|
"learning_rate": 1.6103379721669984e-07, |
|
"loss": 3.1988, |
|
"step": 167 |
|
}, |
|
{ |
|
"epoch": 0.16535433070866143, |
|
"grad_norm": 11.511842727661133, |
|
"learning_rate": 1.620278330019881e-07, |
|
"loss": 3.8236, |
|
"step": 168 |
|
}, |
|
{ |
|
"epoch": 0.16633858267716536, |
|
"grad_norm": 10.053628921508789, |
|
"learning_rate": 1.6302186878727637e-07, |
|
"loss": 3.5063, |
|
"step": 169 |
|
}, |
|
{ |
|
"epoch": 0.1673228346456693, |
|
"grad_norm": 30.96078109741211, |
|
"learning_rate": 1.6401590457256465e-07, |
|
"loss": 5.9843, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.16830708661417323, |
|
"grad_norm": 27.598793029785156, |
|
"learning_rate": 1.650099403578529e-07, |
|
"loss": 5.884, |
|
"step": 171 |
|
}, |
|
{ |
|
"epoch": 0.16929133858267717, |
|
"grad_norm": 10.295365333557129, |
|
"learning_rate": 1.6600397614314118e-07, |
|
"loss": 4.1317, |
|
"step": 172 |
|
}, |
|
{ |
|
"epoch": 0.1702755905511811, |
|
"grad_norm": 10.183586120605469, |
|
"learning_rate": 1.6699801192842944e-07, |
|
"loss": 3.9255, |
|
"step": 173 |
|
}, |
|
{ |
|
"epoch": 0.17125984251968504, |
|
"grad_norm": 11.306191444396973, |
|
"learning_rate": 1.6799204771371774e-07, |
|
"loss": 4.1121, |
|
"step": 174 |
|
}, |
|
{ |
|
"epoch": 0.17224409448818898, |
|
"grad_norm": 9.866985321044922, |
|
"learning_rate": 1.68986083499006e-07, |
|
"loss": 3.7748, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.1732283464566929, |
|
"grad_norm": 25.779844284057617, |
|
"learning_rate": 1.6998011928429427e-07, |
|
"loss": 5.1602, |
|
"step": 176 |
|
}, |
|
{ |
|
"epoch": 0.17421259842519685, |
|
"grad_norm": 22.994916915893555, |
|
"learning_rate": 1.7097415506958253e-07, |
|
"loss": 4.8807, |
|
"step": 177 |
|
}, |
|
{ |
|
"epoch": 0.17519685039370078, |
|
"grad_norm": 9.323945045471191, |
|
"learning_rate": 1.719681908548708e-07, |
|
"loss": 3.4643, |
|
"step": 178 |
|
}, |
|
{ |
|
"epoch": 0.17618110236220472, |
|
"grad_norm": 9.101490020751953, |
|
"learning_rate": 1.7296222664015906e-07, |
|
"loss": 3.4937, |
|
"step": 179 |
|
}, |
|
{ |
|
"epoch": 0.17716535433070865, |
|
"grad_norm": 18.585084915161133, |
|
"learning_rate": 1.7395626242544734e-07, |
|
"loss": 5.2731, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.1781496062992126, |
|
"grad_norm": 17.947126388549805, |
|
"learning_rate": 1.7495029821073564e-07, |
|
"loss": 4.6416, |
|
"step": 181 |
|
}, |
|
{ |
|
"epoch": 0.17913385826771652, |
|
"grad_norm": 9.495512008666992, |
|
"learning_rate": 1.759443339960239e-07, |
|
"loss": 3.5226, |
|
"step": 182 |
|
}, |
|
{ |
|
"epoch": 0.18011811023622049, |
|
"grad_norm": 21.095630645751953, |
|
"learning_rate": 1.7693836978131217e-07, |
|
"loss": 4.7794, |
|
"step": 183 |
|
}, |
|
{ |
|
"epoch": 0.18110236220472442, |
|
"grad_norm": 9.708288192749023, |
|
"learning_rate": 1.7793240556660042e-07, |
|
"loss": 3.8504, |
|
"step": 184 |
|
}, |
|
{ |
|
"epoch": 0.18208661417322836, |
|
"grad_norm": 10.588545799255371, |
|
"learning_rate": 1.789264413518887e-07, |
|
"loss": 3.5391, |
|
"step": 185 |
|
}, |
|
{ |
|
"epoch": 0.1830708661417323, |
|
"grad_norm": 10.238434791564941, |
|
"learning_rate": 1.7992047713717695e-07, |
|
"loss": 4.0291, |
|
"step": 186 |
|
}, |
|
{ |
|
"epoch": 0.18405511811023623, |
|
"grad_norm": 12.710492134094238, |
|
"learning_rate": 1.8091451292246523e-07, |
|
"loss": 3.5606, |
|
"step": 187 |
|
}, |
|
{ |
|
"epoch": 0.18503937007874016, |
|
"grad_norm": 13.008979797363281, |
|
"learning_rate": 1.819085487077535e-07, |
|
"loss": 3.8957, |
|
"step": 188 |
|
}, |
|
{ |
|
"epoch": 0.1860236220472441, |
|
"grad_norm": 16.594804763793945, |
|
"learning_rate": 1.829025844930418e-07, |
|
"loss": 4.3657, |
|
"step": 189 |
|
}, |
|
{ |
|
"epoch": 0.18700787401574803, |
|
"grad_norm": 25.25087547302246, |
|
"learning_rate": 1.8389662027833004e-07, |
|
"loss": 5.0173, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.18799212598425197, |
|
"grad_norm": 18.24939727783203, |
|
"learning_rate": 1.8489065606361832e-07, |
|
"loss": 4.3915, |
|
"step": 191 |
|
}, |
|
{ |
|
"epoch": 0.1889763779527559, |
|
"grad_norm": 11.081143379211426, |
|
"learning_rate": 1.8588469184890657e-07, |
|
"loss": 3.4613, |
|
"step": 192 |
|
}, |
|
{ |
|
"epoch": 0.18996062992125984, |
|
"grad_norm": 10.244654655456543, |
|
"learning_rate": 1.8687872763419485e-07, |
|
"loss": 3.2005, |
|
"step": 193 |
|
}, |
|
{ |
|
"epoch": 0.19094488188976377, |
|
"grad_norm": 10.305020332336426, |
|
"learning_rate": 1.8787276341948313e-07, |
|
"loss": 3.3986, |
|
"step": 194 |
|
}, |
|
{ |
|
"epoch": 0.1919291338582677, |
|
"grad_norm": 12.405226707458496, |
|
"learning_rate": 1.888667992047714e-07, |
|
"loss": 3.7937, |
|
"step": 195 |
|
}, |
|
{ |
|
"epoch": 0.19291338582677164, |
|
"grad_norm": 10.247261047363281, |
|
"learning_rate": 1.898608349900597e-07, |
|
"loss": 3.8981, |
|
"step": 196 |
|
}, |
|
{ |
|
"epoch": 0.19389763779527558, |
|
"grad_norm": 10.535922050476074, |
|
"learning_rate": 1.9085487077534794e-07, |
|
"loss": 3.7051, |
|
"step": 197 |
|
}, |
|
{ |
|
"epoch": 0.19488188976377951, |
|
"grad_norm": 9.828015327453613, |
|
"learning_rate": 1.9184890656063622e-07, |
|
"loss": 3.8028, |
|
"step": 198 |
|
}, |
|
{ |
|
"epoch": 0.19586614173228348, |
|
"grad_norm": 9.550689697265625, |
|
"learning_rate": 1.9284294234592447e-07, |
|
"loss": 3.3294, |
|
"step": 199 |
|
}, |
|
{ |
|
"epoch": 0.1968503937007874, |
|
"grad_norm": 16.766008377075195, |
|
"learning_rate": 1.9383697813121275e-07, |
|
"loss": 4.1252, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.19783464566929135, |
|
"grad_norm": 16.56619644165039, |
|
"learning_rate": 1.94831013916501e-07, |
|
"loss": 4.2564, |
|
"step": 201 |
|
}, |
|
{ |
|
"epoch": 0.19881889763779528, |
|
"grad_norm": 17.413206100463867, |
|
"learning_rate": 1.958250497017893e-07, |
|
"loss": 3.8258, |
|
"step": 202 |
|
}, |
|
{ |
|
"epoch": 0.19980314960629922, |
|
"grad_norm": 10.175580978393555, |
|
"learning_rate": 1.9681908548707756e-07, |
|
"loss": 3.1025, |
|
"step": 203 |
|
}, |
|
{ |
|
"epoch": 0.20078740157480315, |
|
"grad_norm": 11.58014965057373, |
|
"learning_rate": 1.9781312127236584e-07, |
|
"loss": 3.5038, |
|
"step": 204 |
|
}, |
|
{ |
|
"epoch": 0.2017716535433071, |
|
"grad_norm": 9.23029613494873, |
|
"learning_rate": 1.9880715705765412e-07, |
|
"loss": 3.6021, |
|
"step": 205 |
|
}, |
|
{ |
|
"epoch": 0.20275590551181102, |
|
"grad_norm": 11.87929630279541, |
|
"learning_rate": 1.9980119284294237e-07, |
|
"loss": 3.7637, |
|
"step": 206 |
|
}, |
|
{ |
|
"epoch": 0.20374015748031496, |
|
"grad_norm": 10.55318546295166, |
|
"learning_rate": 2.0079522862823065e-07, |
|
"loss": 3.2563, |
|
"step": 207 |
|
}, |
|
{ |
|
"epoch": 0.2047244094488189, |
|
"grad_norm": 15.530010223388672, |
|
"learning_rate": 2.017892644135189e-07, |
|
"loss": 3.9323, |
|
"step": 208 |
|
}, |
|
{ |
|
"epoch": 0.20570866141732283, |
|
"grad_norm": 12.805971145629883, |
|
"learning_rate": 2.027833001988072e-07, |
|
"loss": 3.489, |
|
"step": 209 |
|
}, |
|
{ |
|
"epoch": 0.20669291338582677, |
|
"grad_norm": 11.042547225952148, |
|
"learning_rate": 2.0377733598409546e-07, |
|
"loss": 3.6549, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.2076771653543307, |
|
"grad_norm": 13.565040588378906, |
|
"learning_rate": 2.0477137176938374e-07, |
|
"loss": 3.1609, |
|
"step": 211 |
|
}, |
|
{ |
|
"epoch": 0.20866141732283464, |
|
"grad_norm": 10.574054718017578, |
|
"learning_rate": 2.05765407554672e-07, |
|
"loss": 3.2467, |
|
"step": 212 |
|
}, |
|
{ |
|
"epoch": 0.20964566929133857, |
|
"grad_norm": 9.994040489196777, |
|
"learning_rate": 2.0675944333996027e-07, |
|
"loss": 3.4514, |
|
"step": 213 |
|
}, |
|
{ |
|
"epoch": 0.2106299212598425, |
|
"grad_norm": 11.422471046447754, |
|
"learning_rate": 2.0775347912524852e-07, |
|
"loss": 3.4945, |
|
"step": 214 |
|
}, |
|
{ |
|
"epoch": 0.21161417322834647, |
|
"grad_norm": 13.696518898010254, |
|
"learning_rate": 2.087475149105368e-07, |
|
"loss": 3.5932, |
|
"step": 215 |
|
}, |
|
{ |
|
"epoch": 0.2125984251968504, |
|
"grad_norm": 9.801803588867188, |
|
"learning_rate": 2.097415506958251e-07, |
|
"loss": 3.2289, |
|
"step": 216 |
|
}, |
|
{ |
|
"epoch": 0.21358267716535434, |
|
"grad_norm": 10.710221290588379, |
|
"learning_rate": 2.1073558648111336e-07, |
|
"loss": 3.3279, |
|
"step": 217 |
|
}, |
|
{ |
|
"epoch": 0.21456692913385828, |
|
"grad_norm": 12.974212646484375, |
|
"learning_rate": 2.1172962226640164e-07, |
|
"loss": 3.8141, |
|
"step": 218 |
|
}, |
|
{ |
|
"epoch": 0.2155511811023622, |
|
"grad_norm": 10.196144104003906, |
|
"learning_rate": 2.127236580516899e-07, |
|
"loss": 3.1171, |
|
"step": 219 |
|
}, |
|
{ |
|
"epoch": 0.21653543307086615, |
|
"grad_norm": 12.518829345703125, |
|
"learning_rate": 2.1371769383697817e-07, |
|
"loss": 3.6287, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.21751968503937008, |
|
"grad_norm": 18.030723571777344, |
|
"learning_rate": 2.1471172962226642e-07, |
|
"loss": 3.8517, |
|
"step": 221 |
|
}, |
|
{ |
|
"epoch": 0.21850393700787402, |
|
"grad_norm": 11.197093963623047, |
|
"learning_rate": 2.1570576540755473e-07, |
|
"loss": 3.3836, |
|
"step": 222 |
|
}, |
|
{ |
|
"epoch": 0.21948818897637795, |
|
"grad_norm": 10.613561630249023, |
|
"learning_rate": 2.1669980119284298e-07, |
|
"loss": 3.425, |
|
"step": 223 |
|
}, |
|
{ |
|
"epoch": 0.2204724409448819, |
|
"grad_norm": 11.70770263671875, |
|
"learning_rate": 2.1769383697813126e-07, |
|
"loss": 3.6246, |
|
"step": 224 |
|
}, |
|
{ |
|
"epoch": 0.22145669291338582, |
|
"grad_norm": 10.802409172058105, |
|
"learning_rate": 2.186878727634195e-07, |
|
"loss": 3.5682, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.22244094488188976, |
|
"grad_norm": 10.422605514526367, |
|
"learning_rate": 2.196819085487078e-07, |
|
"loss": 3.3034, |
|
"step": 226 |
|
}, |
|
{ |
|
"epoch": 0.2234251968503937, |
|
"grad_norm": 17.00844383239746, |
|
"learning_rate": 2.2067594433399604e-07, |
|
"loss": 3.9251, |
|
"step": 227 |
|
}, |
|
{ |
|
"epoch": 0.22440944881889763, |
|
"grad_norm": 11.708518028259277, |
|
"learning_rate": 2.2166998011928432e-07, |
|
"loss": 3.146, |
|
"step": 228 |
|
}, |
|
{ |
|
"epoch": 0.22539370078740156, |
|
"grad_norm": 11.237220764160156, |
|
"learning_rate": 2.2266401590457263e-07, |
|
"loss": 3.8859, |
|
"step": 229 |
|
}, |
|
{ |
|
"epoch": 0.2263779527559055, |
|
"grad_norm": 10.216686248779297, |
|
"learning_rate": 2.2365805168986088e-07, |
|
"loss": 3.2977, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.22736220472440946, |
|
"grad_norm": 11.36294937133789, |
|
"learning_rate": 2.2465208747514916e-07, |
|
"loss": 3.2664, |
|
"step": 231 |
|
}, |
|
{ |
|
"epoch": 0.2283464566929134, |
|
"grad_norm": 9.619734764099121, |
|
"learning_rate": 2.256461232604374e-07, |
|
"loss": 3.1275, |
|
"step": 232 |
|
}, |
|
{ |
|
"epoch": 0.22933070866141733, |
|
"grad_norm": 10.229348182678223, |
|
"learning_rate": 2.266401590457257e-07, |
|
"loss": 3.2408, |
|
"step": 233 |
|
}, |
|
{ |
|
"epoch": 0.23031496062992127, |
|
"grad_norm": 11.304025650024414, |
|
"learning_rate": 2.2763419483101394e-07, |
|
"loss": 2.907, |
|
"step": 234 |
|
}, |
|
{ |
|
"epoch": 0.2312992125984252, |
|
"grad_norm": 11.743476867675781, |
|
"learning_rate": 2.2862823061630222e-07, |
|
"loss": 2.9178, |
|
"step": 235 |
|
}, |
|
{ |
|
"epoch": 0.23228346456692914, |
|
"grad_norm": 12.260607719421387, |
|
"learning_rate": 2.296222664015905e-07, |
|
"loss": 3.324, |
|
"step": 236 |
|
}, |
|
{ |
|
"epoch": 0.23326771653543307, |
|
"grad_norm": 12.115522384643555, |
|
"learning_rate": 2.3061630218687878e-07, |
|
"loss": 2.9172, |
|
"step": 237 |
|
}, |
|
{ |
|
"epoch": 0.234251968503937, |
|
"grad_norm": 13.81004810333252, |
|
"learning_rate": 2.3161033797216703e-07, |
|
"loss": 3.4324, |
|
"step": 238 |
|
}, |
|
{ |
|
"epoch": 0.23523622047244094, |
|
"grad_norm": 28.633853912353516, |
|
"learning_rate": 2.326043737574553e-07, |
|
"loss": 4.0563, |
|
"step": 239 |
|
}, |
|
{ |
|
"epoch": 0.23622047244094488, |
|
"grad_norm": 10.444801330566406, |
|
"learning_rate": 2.335984095427436e-07, |
|
"loss": 2.8736, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.2372047244094488, |
|
"grad_norm": 30.760478973388672, |
|
"learning_rate": 2.3459244532803184e-07, |
|
"loss": 4.7174, |
|
"step": 241 |
|
}, |
|
{ |
|
"epoch": 0.23818897637795275, |
|
"grad_norm": 11.556520462036133, |
|
"learning_rate": 2.3558648111332012e-07, |
|
"loss": 3.2025, |
|
"step": 242 |
|
}, |
|
{ |
|
"epoch": 0.23917322834645668, |
|
"grad_norm": 10.850605010986328, |
|
"learning_rate": 2.365805168986084e-07, |
|
"loss": 2.7835, |
|
"step": 243 |
|
}, |
|
{ |
|
"epoch": 0.24015748031496062, |
|
"grad_norm": 26.828187942504883, |
|
"learning_rate": 2.3757455268389668e-07, |
|
"loss": 4.3158, |
|
"step": 244 |
|
}, |
|
{ |
|
"epoch": 0.24114173228346455, |
|
"grad_norm": 10.565204620361328, |
|
"learning_rate": 2.385685884691849e-07, |
|
"loss": 2.8619, |
|
"step": 245 |
|
}, |
|
{ |
|
"epoch": 0.2421259842519685, |
|
"grad_norm": 11.693890571594238, |
|
"learning_rate": 2.395626242544732e-07, |
|
"loss": 2.5156, |
|
"step": 246 |
|
}, |
|
{ |
|
"epoch": 0.24311023622047245, |
|
"grad_norm": 14.487508773803711, |
|
"learning_rate": 2.4055666003976146e-07, |
|
"loss": 3.2144, |
|
"step": 247 |
|
}, |
|
{ |
|
"epoch": 0.2440944881889764, |
|
"grad_norm": 22.58393669128418, |
|
"learning_rate": 2.4155069582504976e-07, |
|
"loss": 3.5927, |
|
"step": 248 |
|
}, |
|
{ |
|
"epoch": 0.24507874015748032, |
|
"grad_norm": 10.880494117736816, |
|
"learning_rate": 2.42544731610338e-07, |
|
"loss": 2.6059, |
|
"step": 249 |
|
}, |
|
{ |
|
"epoch": 0.24606299212598426, |
|
"grad_norm": 10.982386589050293, |
|
"learning_rate": 2.4353876739562627e-07, |
|
"loss": 2.9758, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.2470472440944882, |
|
"grad_norm": 20.552265167236328, |
|
"learning_rate": 2.445328031809146e-07, |
|
"loss": 3.9214, |
|
"step": 251 |
|
}, |
|
{ |
|
"epoch": 0.24803149606299213, |
|
"grad_norm": 11.680928230285645, |
|
"learning_rate": 2.4552683896620283e-07, |
|
"loss": 3.2892, |
|
"step": 252 |
|
}, |
|
{ |
|
"epoch": 0.24901574803149606, |
|
"grad_norm": 10.980914115905762, |
|
"learning_rate": 2.4652087475149113e-07, |
|
"loss": 2.9503, |
|
"step": 253 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 11.629937171936035, |
|
"learning_rate": 2.475149105367794e-07, |
|
"loss": 2.5969, |
|
"step": 254 |
|
}, |
|
{ |
|
"epoch": 0.25098425196850394, |
|
"grad_norm": 11.606547355651855, |
|
"learning_rate": 2.4850894632206764e-07, |
|
"loss": 2.9908, |
|
"step": 255 |
|
}, |
|
{ |
|
"epoch": 0.25196850393700787, |
|
"grad_norm": 10.955612182617188, |
|
"learning_rate": 2.495029821073559e-07, |
|
"loss": 2.8995, |
|
"step": 256 |
|
}, |
|
{ |
|
"epoch": 0.2529527559055118, |
|
"grad_norm": 11.561877250671387, |
|
"learning_rate": 2.5049701789264414e-07, |
|
"loss": 3.124, |
|
"step": 257 |
|
}, |
|
{ |
|
"epoch": 0.25393700787401574, |
|
"grad_norm": 12.040234565734863, |
|
"learning_rate": 2.5149105367793245e-07, |
|
"loss": 3.1197, |
|
"step": 258 |
|
}, |
|
{ |
|
"epoch": 0.2549212598425197, |
|
"grad_norm": 12.25498104095459, |
|
"learning_rate": 2.524850894632207e-07, |
|
"loss": 2.3073, |
|
"step": 259 |
|
}, |
|
{ |
|
"epoch": 0.2559055118110236, |
|
"grad_norm": 9.674198150634766, |
|
"learning_rate": 2.53479125248509e-07, |
|
"loss": 2.8441, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.25688976377952755, |
|
"grad_norm": 11.55859661102295, |
|
"learning_rate": 2.5447316103379726e-07, |
|
"loss": 1.9788, |
|
"step": 261 |
|
}, |
|
{ |
|
"epoch": 0.2578740157480315, |
|
"grad_norm": 12.343245506286621, |
|
"learning_rate": 2.554671968190855e-07, |
|
"loss": 2.1442, |
|
"step": 262 |
|
}, |
|
{ |
|
"epoch": 0.2588582677165354, |
|
"grad_norm": 35.242835998535156, |
|
"learning_rate": 2.564612326043738e-07, |
|
"loss": 4.9015, |
|
"step": 263 |
|
}, |
|
{ |
|
"epoch": 0.25984251968503935, |
|
"grad_norm": 13.894003868103027, |
|
"learning_rate": 2.5745526838966207e-07, |
|
"loss": 2.7866, |
|
"step": 264 |
|
}, |
|
{ |
|
"epoch": 0.2608267716535433, |
|
"grad_norm": 12.785796165466309, |
|
"learning_rate": 2.5844930417495037e-07, |
|
"loss": 2.4588, |
|
"step": 265 |
|
}, |
|
{ |
|
"epoch": 0.2618110236220472, |
|
"grad_norm": 13.115778923034668, |
|
"learning_rate": 2.5944333996023857e-07, |
|
"loss": 2.3909, |
|
"step": 266 |
|
}, |
|
{ |
|
"epoch": 0.26279527559055116, |
|
"grad_norm": 38.795326232910156, |
|
"learning_rate": 2.604373757455269e-07, |
|
"loss": 4.7394, |
|
"step": 267 |
|
}, |
|
{ |
|
"epoch": 0.2637795275590551, |
|
"grad_norm": 14.078845977783203, |
|
"learning_rate": 2.614314115308152e-07, |
|
"loss": 3.1581, |
|
"step": 268 |
|
}, |
|
{ |
|
"epoch": 0.26476377952755903, |
|
"grad_norm": 36.33890151977539, |
|
"learning_rate": 2.6242544731610343e-07, |
|
"loss": 3.973, |
|
"step": 269 |
|
}, |
|
{ |
|
"epoch": 0.265748031496063, |
|
"grad_norm": 32.12986755371094, |
|
"learning_rate": 2.634194831013917e-07, |
|
"loss": 4.1565, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.26673228346456695, |
|
"grad_norm": 13.353660583496094, |
|
"learning_rate": 2.6441351888667994e-07, |
|
"loss": 2.5183, |
|
"step": 271 |
|
}, |
|
{ |
|
"epoch": 0.2677165354330709, |
|
"grad_norm": 24.31439208984375, |
|
"learning_rate": 2.6540755467196824e-07, |
|
"loss": 3.614, |
|
"step": 272 |
|
}, |
|
{ |
|
"epoch": 0.2687007874015748, |
|
"grad_norm": 11.95603084564209, |
|
"learning_rate": 2.664015904572565e-07, |
|
"loss": 2.6858, |
|
"step": 273 |
|
}, |
|
{ |
|
"epoch": 0.26968503937007876, |
|
"grad_norm": 14.08092975616455, |
|
"learning_rate": 2.6739562624254475e-07, |
|
"loss": 3.1182, |
|
"step": 274 |
|
}, |
|
{ |
|
"epoch": 0.2706692913385827, |
|
"grad_norm": 11.759414672851562, |
|
"learning_rate": 2.6838966202783305e-07, |
|
"loss": 2.9628, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.27165354330708663, |
|
"grad_norm": 13.239761352539062, |
|
"learning_rate": 2.693836978131213e-07, |
|
"loss": 2.8376, |
|
"step": 276 |
|
}, |
|
{ |
|
"epoch": 0.27263779527559057, |
|
"grad_norm": 13.551993370056152, |
|
"learning_rate": 2.703777335984096e-07, |
|
"loss": 2.7858, |
|
"step": 277 |
|
}, |
|
{ |
|
"epoch": 0.2736220472440945, |
|
"grad_norm": 12.110513687133789, |
|
"learning_rate": 2.7137176938369786e-07, |
|
"loss": 2.1037, |
|
"step": 278 |
|
}, |
|
{ |
|
"epoch": 0.27460629921259844, |
|
"grad_norm": 14.467287063598633, |
|
"learning_rate": 2.723658051689861e-07, |
|
"loss": 3.0436, |
|
"step": 279 |
|
}, |
|
{ |
|
"epoch": 0.2755905511811024, |
|
"grad_norm": 17.148061752319336, |
|
"learning_rate": 2.7335984095427437e-07, |
|
"loss": 3.4125, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.2765748031496063, |
|
"grad_norm": 12.052011489868164, |
|
"learning_rate": 2.743538767395627e-07, |
|
"loss": 2.5027, |
|
"step": 281 |
|
}, |
|
{ |
|
"epoch": 0.27755905511811024, |
|
"grad_norm": 11.502729415893555, |
|
"learning_rate": 2.75347912524851e-07, |
|
"loss": 2.7922, |
|
"step": 282 |
|
}, |
|
{ |
|
"epoch": 0.2785433070866142, |
|
"grad_norm": 15.440553665161133, |
|
"learning_rate": 2.763419483101392e-07, |
|
"loss": 2.9762, |
|
"step": 283 |
|
}, |
|
{ |
|
"epoch": 0.2795275590551181, |
|
"grad_norm": 12.604522705078125, |
|
"learning_rate": 2.773359840954275e-07, |
|
"loss": 2.6458, |
|
"step": 284 |
|
}, |
|
{ |
|
"epoch": 0.28051181102362205, |
|
"grad_norm": 12.404953002929688, |
|
"learning_rate": 2.7833001988071574e-07, |
|
"loss": 2.962, |
|
"step": 285 |
|
}, |
|
{ |
|
"epoch": 0.281496062992126, |
|
"grad_norm": 12.527341842651367, |
|
"learning_rate": 2.7932405566600404e-07, |
|
"loss": 2.5439, |
|
"step": 286 |
|
}, |
|
{ |
|
"epoch": 0.2824803149606299, |
|
"grad_norm": 12.331761360168457, |
|
"learning_rate": 2.803180914512923e-07, |
|
"loss": 2.8437, |
|
"step": 287 |
|
}, |
|
{ |
|
"epoch": 0.28346456692913385, |
|
"grad_norm": 17.64342498779297, |
|
"learning_rate": 2.8131212723658055e-07, |
|
"loss": 3.2134, |
|
"step": 288 |
|
}, |
|
{ |
|
"epoch": 0.2844488188976378, |
|
"grad_norm": 12.74220085144043, |
|
"learning_rate": 2.8230616302186885e-07, |
|
"loss": 2.5655, |
|
"step": 289 |
|
}, |
|
{ |
|
"epoch": 0.2854330708661417, |
|
"grad_norm": 13.071759223937988, |
|
"learning_rate": 2.833001988071571e-07, |
|
"loss": 2.9465, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.28641732283464566, |
|
"grad_norm": 12.093460083007812, |
|
"learning_rate": 2.842942345924454e-07, |
|
"loss": 2.4653, |
|
"step": 291 |
|
}, |
|
{ |
|
"epoch": 0.2874015748031496, |
|
"grad_norm": 18.303346633911133, |
|
"learning_rate": 2.852882703777336e-07, |
|
"loss": 3.1467, |
|
"step": 292 |
|
}, |
|
{ |
|
"epoch": 0.28838582677165353, |
|
"grad_norm": 11.613899230957031, |
|
"learning_rate": 2.862823061630219e-07, |
|
"loss": 2.6551, |
|
"step": 293 |
|
}, |
|
{ |
|
"epoch": 0.28937007874015747, |
|
"grad_norm": 12.590253829956055, |
|
"learning_rate": 2.8727634194831017e-07, |
|
"loss": 2.5098, |
|
"step": 294 |
|
}, |
|
{ |
|
"epoch": 0.2903543307086614, |
|
"grad_norm": 13.879076957702637, |
|
"learning_rate": 2.8827037773359847e-07, |
|
"loss": 2.5988, |
|
"step": 295 |
|
}, |
|
{ |
|
"epoch": 0.29133858267716534, |
|
"grad_norm": 23.504112243652344, |
|
"learning_rate": 2.892644135188867e-07, |
|
"loss": 3.778, |
|
"step": 296 |
|
}, |
|
{ |
|
"epoch": 0.29232283464566927, |
|
"grad_norm": 11.794881820678711, |
|
"learning_rate": 2.90258449304175e-07, |
|
"loss": 2.6257, |
|
"step": 297 |
|
}, |
|
{ |
|
"epoch": 0.2933070866141732, |
|
"grad_norm": 11.491183280944824, |
|
"learning_rate": 2.912524850894633e-07, |
|
"loss": 2.5142, |
|
"step": 298 |
|
}, |
|
{ |
|
"epoch": 0.29429133858267714, |
|
"grad_norm": 11.801483154296875, |
|
"learning_rate": 2.9224652087475153e-07, |
|
"loss": 2.3182, |
|
"step": 299 |
|
}, |
|
{ |
|
"epoch": 0.2952755905511811, |
|
"grad_norm": 21.28721046447754, |
|
"learning_rate": 2.9324055666003984e-07, |
|
"loss": 3.3505, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.296259842519685, |
|
"grad_norm": 14.353005409240723, |
|
"learning_rate": 2.9423459244532804e-07, |
|
"loss": 2.9615, |
|
"step": 301 |
|
}, |
|
{ |
|
"epoch": 0.297244094488189, |
|
"grad_norm": 17.378894805908203, |
|
"learning_rate": 2.9522862823061634e-07, |
|
"loss": 2.9136, |
|
"step": 302 |
|
}, |
|
{ |
|
"epoch": 0.29822834645669294, |
|
"grad_norm": 12.216327667236328, |
|
"learning_rate": 2.9622266401590465e-07, |
|
"loss": 2.6192, |
|
"step": 303 |
|
}, |
|
{ |
|
"epoch": 0.2992125984251969, |
|
"grad_norm": 12.561356544494629, |
|
"learning_rate": 2.972166998011929e-07, |
|
"loss": 2.3255, |
|
"step": 304 |
|
}, |
|
{ |
|
"epoch": 0.3001968503937008, |
|
"grad_norm": 15.897427558898926, |
|
"learning_rate": 2.9821073558648115e-07, |
|
"loss": 2.7168, |
|
"step": 305 |
|
}, |
|
{ |
|
"epoch": 0.30118110236220474, |
|
"grad_norm": 20.39179039001465, |
|
"learning_rate": 2.992047713717694e-07, |
|
"loss": 2.9137, |
|
"step": 306 |
|
}, |
|
{ |
|
"epoch": 0.30118110236220474, |
|
"eval_Qnli-dev_cosine_accuracy": 0.59765625, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.9195176362991333, |
|
"eval_Qnli-dev_cosine_ap": 0.5806713513254954, |
|
"eval_Qnli-dev_cosine_f1": 0.6419400855920115, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.8419712781906128, |
|
"eval_Qnli-dev_cosine_precision": 0.4838709677419355, |
|
"eval_Qnli-dev_cosine_recall": 0.9533898305084746, |
|
"eval_Qnli-dev_dot_accuracy": 0.59765625, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 706.0679931640625, |
|
"eval_Qnli-dev_dot_ap": 0.5806120989705038, |
|
"eval_Qnli-dev_dot_f1": 0.641025641025641, |
|
"eval_Qnli-dev_dot_f1_threshold": 646.50732421875, |
|
"eval_Qnli-dev_dot_precision": 0.48283261802575106, |
|
"eval_Qnli-dev_dot_recall": 0.9533898305084746, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.59765625, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 11.117534637451172, |
|
"eval_Qnli-dev_euclidean_ap": 0.5806941879083722, |
|
"eval_Qnli-dev_euclidean_f1": 0.6419400855920115, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 15.578910827636719, |
|
"eval_Qnli-dev_euclidean_precision": 0.4838709677419355, |
|
"eval_Qnli-dev_euclidean_recall": 0.9533898305084746, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.607421875, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 233.11520385742188, |
|
"eval_Qnli-dev_manhattan_ap": 0.5947537164269667, |
|
"eval_Qnli-dev_manhattan_f1": 0.6514459665144596, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 283.4456481933594, |
|
"eval_Qnli-dev_manhattan_precision": 0.5083135391923991, |
|
"eval_Qnli-dev_manhattan_recall": 0.9067796610169492, |
|
"eval_Qnli-dev_max_accuracy": 0.607421875, |
|
"eval_Qnli-dev_max_accuracy_threshold": 706.0679931640625, |
|
"eval_Qnli-dev_max_ap": 0.5947537164269667, |
|
"eval_Qnli-dev_max_f1": 0.6514459665144596, |
|
"eval_Qnli-dev_max_f1_threshold": 646.50732421875, |
|
"eval_Qnli-dev_max_precision": 0.5083135391923991, |
|
"eval_Qnli-dev_max_recall": 0.9533898305084746, |
|
"eval_allNLI-dev_cosine_accuracy": 0.671875, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9753081798553467, |
|
"eval_allNLI-dev_cosine_ap": 0.41030395214559057, |
|
"eval_allNLI-dev_cosine_f1": 0.5126475548060708, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.8886709213256836, |
|
"eval_allNLI-dev_cosine_precision": 0.3619047619047619, |
|
"eval_allNLI-dev_cosine_recall": 0.8786127167630058, |
|
"eval_allNLI-dev_dot_accuracy": 0.669921875, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 751.201904296875, |
|
"eval_allNLI-dev_dot_ap": 0.40946066450522606, |
|
"eval_allNLI-dev_dot_f1": 0.5126475548060708, |
|
"eval_allNLI-dev_dot_f1_threshold": 682.495849609375, |
|
"eval_allNLI-dev_dot_precision": 0.3619047619047619, |
|
"eval_allNLI-dev_dot_recall": 0.8786127167630058, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.671875, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 6.158434867858887, |
|
"eval_allNLI-dev_euclidean_ap": 0.4103105551716732, |
|
"eval_allNLI-dev_euclidean_f1": 0.5126475548060708, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 13.076848983764648, |
|
"eval_allNLI-dev_euclidean_precision": 0.3619047619047619, |
|
"eval_allNLI-dev_euclidean_recall": 0.8786127167630058, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.66796875, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 125.19190979003906, |
|
"eval_allNLI-dev_manhattan_ap": 0.405662874748899, |
|
"eval_allNLI-dev_manhattan_f1": 0.5164644714038128, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 245.46649169921875, |
|
"eval_allNLI-dev_manhattan_precision": 0.3688118811881188, |
|
"eval_allNLI-dev_manhattan_recall": 0.861271676300578, |
|
"eval_allNLI-dev_max_accuracy": 0.671875, |
|
"eval_allNLI-dev_max_accuracy_threshold": 751.201904296875, |
|
"eval_allNLI-dev_max_ap": 0.4103105551716732, |
|
"eval_allNLI-dev_max_f1": 0.5164644714038128, |
|
"eval_allNLI-dev_max_f1_threshold": 682.495849609375, |
|
"eval_allNLI-dev_max_precision": 0.3688118811881188, |
|
"eval_allNLI-dev_max_recall": 0.8786127167630058, |
|
"eval_loss": 2.4280221462249756, |
|
"eval_runtime": 52.6639, |
|
"eval_samples_per_second": 31.597, |
|
"eval_sequential_score": 0.5947537164269667, |
|
"eval_steps_per_second": 0.133, |
|
"eval_sts-test_pearson_cosine": 0.20581042639062724, |
|
"eval_sts-test_pearson_dot": 0.2051481762928456, |
|
"eval_sts-test_pearson_euclidean": 0.22966459127158345, |
|
"eval_sts-test_pearson_manhattan": 0.23576329100758983, |
|
"eval_sts-test_pearson_max": 0.23576329100758983, |
|
"eval_sts-test_spearman_cosine": 0.25074163083484124, |
|
"eval_sts-test_spearman_dot": 0.250431638847183, |
|
"eval_sts-test_spearman_euclidean": 0.2507327700850857, |
|
"eval_sts-test_spearman_manhattan": 0.256412465927098, |
|
"eval_sts-test_spearman_max": 0.256412465927098, |
|
"step": 306 |
|
}, |
|
{ |
|
"epoch": 0.3021653543307087, |
|
"grad_norm": 16.60762596130371, |
|
"learning_rate": 3.001988071570577e-07, |
|
"loss": 2.6681, |
|
"step": 307 |
|
}, |
|
{ |
|
"epoch": 0.3031496062992126, |
|
"grad_norm": 15.935681343078613, |
|
"learning_rate": 3.01192842942346e-07, |
|
"loss": 2.7219, |
|
"step": 308 |
|
}, |
|
{ |
|
"epoch": 0.30413385826771655, |
|
"grad_norm": 12.386747360229492, |
|
"learning_rate": 3.021868787276342e-07, |
|
"loss": 2.4057, |
|
"step": 309 |
|
}, |
|
{ |
|
"epoch": 0.3051181102362205, |
|
"grad_norm": 11.695796966552734, |
|
"learning_rate": 3.031809145129225e-07, |
|
"loss": 2.7402, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.3061023622047244, |
|
"grad_norm": 14.046793937683105, |
|
"learning_rate": 3.0417495029821077e-07, |
|
"loss": 2.5512, |
|
"step": 311 |
|
}, |
|
{ |
|
"epoch": 0.30708661417322836, |
|
"grad_norm": 11.52070426940918, |
|
"learning_rate": 3.051689860834991e-07, |
|
"loss": 2.8553, |
|
"step": 312 |
|
}, |
|
{ |
|
"epoch": 0.3080708661417323, |
|
"grad_norm": 12.387203216552734, |
|
"learning_rate": 3.0616302186878733e-07, |
|
"loss": 2.598, |
|
"step": 313 |
|
}, |
|
{ |
|
"epoch": 0.3090551181102362, |
|
"grad_norm": 14.028305053710938, |
|
"learning_rate": 3.071570576540756e-07, |
|
"loss": 2.6186, |
|
"step": 314 |
|
}, |
|
{ |
|
"epoch": 0.31003937007874016, |
|
"grad_norm": 11.698431968688965, |
|
"learning_rate": 3.0815109343936384e-07, |
|
"loss": 2.3678, |
|
"step": 315 |
|
}, |
|
{ |
|
"epoch": 0.3110236220472441, |
|
"grad_norm": 12.938248634338379, |
|
"learning_rate": 3.0914512922465214e-07, |
|
"loss": 2.886, |
|
"step": 316 |
|
}, |
|
{ |
|
"epoch": 0.31200787401574803, |
|
"grad_norm": 12.183907508850098, |
|
"learning_rate": 3.1013916500994045e-07, |
|
"loss": 2.1738, |
|
"step": 317 |
|
}, |
|
{ |
|
"epoch": 0.31299212598425197, |
|
"grad_norm": 15.912786483764648, |
|
"learning_rate": 3.1113320079522865e-07, |
|
"loss": 2.6619, |
|
"step": 318 |
|
}, |
|
{ |
|
"epoch": 0.3139763779527559, |
|
"grad_norm": 10.824213981628418, |
|
"learning_rate": 3.1212723658051695e-07, |
|
"loss": 2.1818, |
|
"step": 319 |
|
}, |
|
{ |
|
"epoch": 0.31496062992125984, |
|
"grad_norm": 13.062846183776855, |
|
"learning_rate": 3.131212723658052e-07, |
|
"loss": 3.0407, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.3159448818897638, |
|
"grad_norm": 11.457890510559082, |
|
"learning_rate": 3.141153081510935e-07, |
|
"loss": 2.464, |
|
"step": 321 |
|
}, |
|
{ |
|
"epoch": 0.3169291338582677, |
|
"grad_norm": 12.722960472106934, |
|
"learning_rate": 3.151093439363818e-07, |
|
"loss": 2.7415, |
|
"step": 322 |
|
}, |
|
{ |
|
"epoch": 0.31791338582677164, |
|
"grad_norm": 13.098994255065918, |
|
"learning_rate": 3.1610337972167e-07, |
|
"loss": 2.7455, |
|
"step": 323 |
|
}, |
|
{ |
|
"epoch": 0.3188976377952756, |
|
"grad_norm": 12.980748176574707, |
|
"learning_rate": 3.170974155069583e-07, |
|
"loss": 2.4061, |
|
"step": 324 |
|
}, |
|
{ |
|
"epoch": 0.3198818897637795, |
|
"grad_norm": 11.015922546386719, |
|
"learning_rate": 3.1809145129224657e-07, |
|
"loss": 2.0491, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.32086614173228345, |
|
"grad_norm": 23.543014526367188, |
|
"learning_rate": 3.190854870775349e-07, |
|
"loss": 3.3097, |
|
"step": 326 |
|
}, |
|
{ |
|
"epoch": 0.3218503937007874, |
|
"grad_norm": 13.360864639282227, |
|
"learning_rate": 3.200795228628231e-07, |
|
"loss": 2.3587, |
|
"step": 327 |
|
}, |
|
{ |
|
"epoch": 0.3228346456692913, |
|
"grad_norm": 11.620161056518555, |
|
"learning_rate": 3.210735586481114e-07, |
|
"loss": 1.9493, |
|
"step": 328 |
|
}, |
|
{ |
|
"epoch": 0.32381889763779526, |
|
"grad_norm": 16.858875274658203, |
|
"learning_rate": 3.220675944333997e-07, |
|
"loss": 2.5399, |
|
"step": 329 |
|
}, |
|
{ |
|
"epoch": 0.3248031496062992, |
|
"grad_norm": 16.233020782470703, |
|
"learning_rate": 3.2306163021868794e-07, |
|
"loss": 2.3569, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.3257874015748031, |
|
"grad_norm": 10.670516014099121, |
|
"learning_rate": 3.240556660039762e-07, |
|
"loss": 1.9024, |
|
"step": 331 |
|
}, |
|
{ |
|
"epoch": 0.32677165354330706, |
|
"grad_norm": 13.305569648742676, |
|
"learning_rate": 3.2504970178926444e-07, |
|
"loss": 2.3513, |
|
"step": 332 |
|
}, |
|
{ |
|
"epoch": 0.327755905511811, |
|
"grad_norm": 16.92148208618164, |
|
"learning_rate": 3.2604373757455275e-07, |
|
"loss": 2.2488, |
|
"step": 333 |
|
}, |
|
{ |
|
"epoch": 0.328740157480315, |
|
"grad_norm": 13.141341209411621, |
|
"learning_rate": 3.27037773359841e-07, |
|
"loss": 1.9141, |
|
"step": 334 |
|
}, |
|
{ |
|
"epoch": 0.3297244094488189, |
|
"grad_norm": 12.373048782348633, |
|
"learning_rate": 3.280318091451293e-07, |
|
"loss": 2.7065, |
|
"step": 335 |
|
}, |
|
{ |
|
"epoch": 0.33070866141732286, |
|
"grad_norm": 11.926918029785156, |
|
"learning_rate": 3.2902584493041756e-07, |
|
"loss": 2.139, |
|
"step": 336 |
|
}, |
|
{ |
|
"epoch": 0.3316929133858268, |
|
"grad_norm": 14.438637733459473, |
|
"learning_rate": 3.300198807157058e-07, |
|
"loss": 2.2345, |
|
"step": 337 |
|
}, |
|
{ |
|
"epoch": 0.33267716535433073, |
|
"grad_norm": 11.567721366882324, |
|
"learning_rate": 3.310139165009941e-07, |
|
"loss": 2.3612, |
|
"step": 338 |
|
}, |
|
{ |
|
"epoch": 0.33366141732283466, |
|
"grad_norm": 11.467390060424805, |
|
"learning_rate": 3.3200795228628237e-07, |
|
"loss": 2.1413, |
|
"step": 339 |
|
}, |
|
{ |
|
"epoch": 0.3346456692913386, |
|
"grad_norm": 13.05380630493164, |
|
"learning_rate": 3.330019880715706e-07, |
|
"loss": 2.2214, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.33562992125984253, |
|
"grad_norm": 13.208922386169434, |
|
"learning_rate": 3.3399602385685887e-07, |
|
"loss": 2.9006, |
|
"step": 341 |
|
}, |
|
{ |
|
"epoch": 0.33661417322834647, |
|
"grad_norm": 12.578964233398438, |
|
"learning_rate": 3.349900596421472e-07, |
|
"loss": 2.417, |
|
"step": 342 |
|
}, |
|
{ |
|
"epoch": 0.3375984251968504, |
|
"grad_norm": 11.79666519165039, |
|
"learning_rate": 3.359840954274355e-07, |
|
"loss": 2.2348, |
|
"step": 343 |
|
}, |
|
{ |
|
"epoch": 0.33858267716535434, |
|
"grad_norm": 12.465178489685059, |
|
"learning_rate": 3.369781312127237e-07, |
|
"loss": 2.4369, |
|
"step": 344 |
|
}, |
|
{ |
|
"epoch": 0.3395669291338583, |
|
"grad_norm": 16.569061279296875, |
|
"learning_rate": 3.37972166998012e-07, |
|
"loss": 2.7623, |
|
"step": 345 |
|
}, |
|
{ |
|
"epoch": 0.3405511811023622, |
|
"grad_norm": 18.68236541748047, |
|
"learning_rate": 3.3896620278330024e-07, |
|
"loss": 2.6741, |
|
"step": 346 |
|
}, |
|
{ |
|
"epoch": 0.34153543307086615, |
|
"grad_norm": 22.597843170166016, |
|
"learning_rate": 3.3996023856858855e-07, |
|
"loss": 3.0515, |
|
"step": 347 |
|
}, |
|
{ |
|
"epoch": 0.3425196850393701, |
|
"grad_norm": 13.142956733703613, |
|
"learning_rate": 3.409542743538768e-07, |
|
"loss": 2.4952, |
|
"step": 348 |
|
}, |
|
{ |
|
"epoch": 0.343503937007874, |
|
"grad_norm": 10.252715110778809, |
|
"learning_rate": 3.4194831013916505e-07, |
|
"loss": 2.1265, |
|
"step": 349 |
|
}, |
|
{ |
|
"epoch": 0.34448818897637795, |
|
"grad_norm": 10.546085357666016, |
|
"learning_rate": 3.4294234592445336e-07, |
|
"loss": 2.0359, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.3454724409448819, |
|
"grad_norm": 15.560839653015137, |
|
"learning_rate": 3.439363817097416e-07, |
|
"loss": 3.107, |
|
"step": 351 |
|
}, |
|
{ |
|
"epoch": 0.3464566929133858, |
|
"grad_norm": 10.87291145324707, |
|
"learning_rate": 3.449304174950299e-07, |
|
"loss": 2.116, |
|
"step": 352 |
|
}, |
|
{ |
|
"epoch": 0.34744094488188976, |
|
"grad_norm": 11.75866413116455, |
|
"learning_rate": 3.459244532803181e-07, |
|
"loss": 2.1996, |
|
"step": 353 |
|
}, |
|
{ |
|
"epoch": 0.3484251968503937, |
|
"grad_norm": 14.872318267822266, |
|
"learning_rate": 3.469184890656064e-07, |
|
"loss": 2.9312, |
|
"step": 354 |
|
}, |
|
{ |
|
"epoch": 0.3494094488188976, |
|
"grad_norm": 14.699666023254395, |
|
"learning_rate": 3.4791252485089467e-07, |
|
"loss": 2.2885, |
|
"step": 355 |
|
}, |
|
{ |
|
"epoch": 0.35039370078740156, |
|
"grad_norm": 17.621458053588867, |
|
"learning_rate": 3.48906560636183e-07, |
|
"loss": 3.0302, |
|
"step": 356 |
|
}, |
|
{ |
|
"epoch": 0.3513779527559055, |
|
"grad_norm": 12.756147384643555, |
|
"learning_rate": 3.499005964214713e-07, |
|
"loss": 2.2163, |
|
"step": 357 |
|
}, |
|
{ |
|
"epoch": 0.35236220472440943, |
|
"grad_norm": 14.842194557189941, |
|
"learning_rate": 3.508946322067595e-07, |
|
"loss": 2.8304, |
|
"step": 358 |
|
}, |
|
{ |
|
"epoch": 0.35334645669291337, |
|
"grad_norm": 11.481346130371094, |
|
"learning_rate": 3.518886679920478e-07, |
|
"loss": 2.2715, |
|
"step": 359 |
|
}, |
|
{ |
|
"epoch": 0.3543307086614173, |
|
"grad_norm": 11.839008331298828, |
|
"learning_rate": 3.5288270377733604e-07, |
|
"loss": 2.3388, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.35531496062992124, |
|
"grad_norm": 12.406773567199707, |
|
"learning_rate": 3.5387673956262434e-07, |
|
"loss": 2.2098, |
|
"step": 361 |
|
}, |
|
{ |
|
"epoch": 0.3562992125984252, |
|
"grad_norm": 11.005632400512695, |
|
"learning_rate": 3.5487077534791254e-07, |
|
"loss": 2.0911, |
|
"step": 362 |
|
}, |
|
{ |
|
"epoch": 0.3572834645669291, |
|
"grad_norm": 11.084342002868652, |
|
"learning_rate": 3.5586481113320085e-07, |
|
"loss": 2.3582, |
|
"step": 363 |
|
}, |
|
{ |
|
"epoch": 0.35826771653543305, |
|
"grad_norm": 10.120821952819824, |
|
"learning_rate": 3.5685884691848915e-07, |
|
"loss": 1.8605, |
|
"step": 364 |
|
}, |
|
{ |
|
"epoch": 0.359251968503937, |
|
"grad_norm": 13.282901763916016, |
|
"learning_rate": 3.578528827037774e-07, |
|
"loss": 2.2252, |
|
"step": 365 |
|
}, |
|
{ |
|
"epoch": 0.36023622047244097, |
|
"grad_norm": 13.67626667022705, |
|
"learning_rate": 3.5884691848906566e-07, |
|
"loss": 2.2018, |
|
"step": 366 |
|
}, |
|
{ |
|
"epoch": 0.3612204724409449, |
|
"grad_norm": 13.320021629333496, |
|
"learning_rate": 3.598409542743539e-07, |
|
"loss": 2.1099, |
|
"step": 367 |
|
}, |
|
{ |
|
"epoch": 0.36220472440944884, |
|
"grad_norm": 11.790545463562012, |
|
"learning_rate": 3.608349900596422e-07, |
|
"loss": 2.1323, |
|
"step": 368 |
|
}, |
|
{ |
|
"epoch": 0.3631889763779528, |
|
"grad_norm": 13.921866416931152, |
|
"learning_rate": 3.6182902584493047e-07, |
|
"loss": 2.4203, |
|
"step": 369 |
|
}, |
|
{ |
|
"epoch": 0.3641732283464567, |
|
"grad_norm": 23.001510620117188, |
|
"learning_rate": 3.6282306163021877e-07, |
|
"loss": 2.7768, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.36515748031496065, |
|
"grad_norm": 14.317642211914062, |
|
"learning_rate": 3.63817097415507e-07, |
|
"loss": 2.3359, |
|
"step": 371 |
|
}, |
|
{ |
|
"epoch": 0.3661417322834646, |
|
"grad_norm": 16.946765899658203, |
|
"learning_rate": 3.648111332007953e-07, |
|
"loss": 2.3773, |
|
"step": 372 |
|
}, |
|
{ |
|
"epoch": 0.3671259842519685, |
|
"grad_norm": 12.836204528808594, |
|
"learning_rate": 3.658051689860836e-07, |
|
"loss": 2.4424, |
|
"step": 373 |
|
}, |
|
{ |
|
"epoch": 0.36811023622047245, |
|
"grad_norm": 13.058905601501465, |
|
"learning_rate": 3.6679920477137184e-07, |
|
"loss": 1.9478, |
|
"step": 374 |
|
}, |
|
{ |
|
"epoch": 0.3690944881889764, |
|
"grad_norm": 15.509428977966309, |
|
"learning_rate": 3.677932405566601e-07, |
|
"loss": 1.6047, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.3700787401574803, |
|
"grad_norm": 11.629678726196289, |
|
"learning_rate": 3.6878727634194834e-07, |
|
"loss": 1.7384, |
|
"step": 376 |
|
}, |
|
{ |
|
"epoch": 0.37106299212598426, |
|
"grad_norm": 13.820512771606445, |
|
"learning_rate": 3.6978131212723665e-07, |
|
"loss": 2.1147, |
|
"step": 377 |
|
}, |
|
{ |
|
"epoch": 0.3720472440944882, |
|
"grad_norm": 11.474569320678711, |
|
"learning_rate": 3.7077534791252495e-07, |
|
"loss": 1.8449, |
|
"step": 378 |
|
}, |
|
{ |
|
"epoch": 0.37303149606299213, |
|
"grad_norm": 15.554925918579102, |
|
"learning_rate": 3.7176938369781315e-07, |
|
"loss": 2.6009, |
|
"step": 379 |
|
}, |
|
{ |
|
"epoch": 0.37401574803149606, |
|
"grad_norm": 13.037839889526367, |
|
"learning_rate": 3.7276341948310145e-07, |
|
"loss": 2.4051, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.375, |
|
"grad_norm": 12.058863639831543, |
|
"learning_rate": 3.737574552683897e-07, |
|
"loss": 2.3035, |
|
"step": 381 |
|
}, |
|
{ |
|
"epoch": 0.37598425196850394, |
|
"grad_norm": 11.572028160095215, |
|
"learning_rate": 3.74751491053678e-07, |
|
"loss": 1.8955, |
|
"step": 382 |
|
}, |
|
{ |
|
"epoch": 0.37696850393700787, |
|
"grad_norm": 16.923925399780273, |
|
"learning_rate": 3.7574552683896626e-07, |
|
"loss": 2.287, |
|
"step": 383 |
|
}, |
|
{ |
|
"epoch": 0.3779527559055118, |
|
"grad_norm": 10.901126861572266, |
|
"learning_rate": 3.767395626242545e-07, |
|
"loss": 1.9123, |
|
"step": 384 |
|
}, |
|
{ |
|
"epoch": 0.37893700787401574, |
|
"grad_norm": 10.82137680053711, |
|
"learning_rate": 3.777335984095428e-07, |
|
"loss": 1.9369, |
|
"step": 385 |
|
}, |
|
{ |
|
"epoch": 0.3799212598425197, |
|
"grad_norm": 11.884252548217773, |
|
"learning_rate": 3.787276341948311e-07, |
|
"loss": 2.1367, |
|
"step": 386 |
|
}, |
|
{ |
|
"epoch": 0.3809055118110236, |
|
"grad_norm": 12.42955207824707, |
|
"learning_rate": 3.797216699801194e-07, |
|
"loss": 1.9437, |
|
"step": 387 |
|
}, |
|
{ |
|
"epoch": 0.38188976377952755, |
|
"grad_norm": 13.655105590820312, |
|
"learning_rate": 3.807157057654076e-07, |
|
"loss": 2.3873, |
|
"step": 388 |
|
}, |
|
{ |
|
"epoch": 0.3828740157480315, |
|
"grad_norm": 11.53514575958252, |
|
"learning_rate": 3.817097415506959e-07, |
|
"loss": 1.7463, |
|
"step": 389 |
|
}, |
|
{ |
|
"epoch": 0.3838582677165354, |
|
"grad_norm": 22.04655647277832, |
|
"learning_rate": 3.8270377733598414e-07, |
|
"loss": 2.8438, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.38484251968503935, |
|
"grad_norm": 17.399572372436523, |
|
"learning_rate": 3.8369781312127244e-07, |
|
"loss": 2.4875, |
|
"step": 391 |
|
}, |
|
{ |
|
"epoch": 0.3858267716535433, |
|
"grad_norm": 13.103324890136719, |
|
"learning_rate": 3.8469184890656075e-07, |
|
"loss": 2.0798, |
|
"step": 392 |
|
}, |
|
{ |
|
"epoch": 0.3868110236220472, |
|
"grad_norm": 13.370649337768555, |
|
"learning_rate": 3.8568588469184895e-07, |
|
"loss": 2.2242, |
|
"step": 393 |
|
}, |
|
{ |
|
"epoch": 0.38779527559055116, |
|
"grad_norm": 12.390323638916016, |
|
"learning_rate": 3.8667992047713725e-07, |
|
"loss": 1.8714, |
|
"step": 394 |
|
}, |
|
{ |
|
"epoch": 0.3887795275590551, |
|
"grad_norm": 13.5247802734375, |
|
"learning_rate": 3.876739562624255e-07, |
|
"loss": 1.5893, |
|
"step": 395 |
|
}, |
|
{ |
|
"epoch": 0.38976377952755903, |
|
"grad_norm": 10.50505542755127, |
|
"learning_rate": 3.886679920477138e-07, |
|
"loss": 1.5633, |
|
"step": 396 |
|
}, |
|
{ |
|
"epoch": 0.390748031496063, |
|
"grad_norm": 11.619284629821777, |
|
"learning_rate": 3.89662027833002e-07, |
|
"loss": 1.8645, |
|
"step": 397 |
|
}, |
|
{ |
|
"epoch": 0.39173228346456695, |
|
"grad_norm": 11.723767280578613, |
|
"learning_rate": 3.906560636182903e-07, |
|
"loss": 1.8928, |
|
"step": 398 |
|
}, |
|
{ |
|
"epoch": 0.3927165354330709, |
|
"grad_norm": 9.949848175048828, |
|
"learning_rate": 3.916500994035786e-07, |
|
"loss": 1.3352, |
|
"step": 399 |
|
}, |
|
{ |
|
"epoch": 0.3937007874015748, |
|
"grad_norm": 30.210622787475586, |
|
"learning_rate": 3.9264413518886687e-07, |
|
"loss": 3.3052, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.39468503937007876, |
|
"grad_norm": 12.481863975524902, |
|
"learning_rate": 3.936381709741551e-07, |
|
"loss": 1.6256, |
|
"step": 401 |
|
}, |
|
{ |
|
"epoch": 0.3956692913385827, |
|
"grad_norm": 12.854496955871582, |
|
"learning_rate": 3.946322067594434e-07, |
|
"loss": 1.8856, |
|
"step": 402 |
|
}, |
|
{ |
|
"epoch": 0.39665354330708663, |
|
"grad_norm": 11.837615013122559, |
|
"learning_rate": 3.956262425447317e-07, |
|
"loss": 1.8355, |
|
"step": 403 |
|
}, |
|
{ |
|
"epoch": 0.39763779527559057, |
|
"grad_norm": 14.314859390258789, |
|
"learning_rate": 3.9662027833001993e-07, |
|
"loss": 1.8944, |
|
"step": 404 |
|
}, |
|
{ |
|
"epoch": 0.3986220472440945, |
|
"grad_norm": 11.3548002243042, |
|
"learning_rate": 3.9761431411530824e-07, |
|
"loss": 1.7636, |
|
"step": 405 |
|
}, |
|
{ |
|
"epoch": 0.39960629921259844, |
|
"grad_norm": 25.222524642944336, |
|
"learning_rate": 3.986083499005965e-07, |
|
"loss": 2.8097, |
|
"step": 406 |
|
}, |
|
{ |
|
"epoch": 0.4005905511811024, |
|
"grad_norm": 13.39458179473877, |
|
"learning_rate": 3.9960238568588474e-07, |
|
"loss": 1.9121, |
|
"step": 407 |
|
}, |
|
{ |
|
"epoch": 0.4015748031496063, |
|
"grad_norm": 12.579276084899902, |
|
"learning_rate": 4.0059642147117305e-07, |
|
"loss": 1.9233, |
|
"step": 408 |
|
}, |
|
{ |
|
"epoch": 0.40255905511811024, |
|
"grad_norm": 12.75617504119873, |
|
"learning_rate": 4.015904572564613e-07, |
|
"loss": 1.543, |
|
"step": 409 |
|
}, |
|
{ |
|
"epoch": 0.4035433070866142, |
|
"grad_norm": 13.689290046691895, |
|
"learning_rate": 4.0258449304174955e-07, |
|
"loss": 1.7207, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.4045275590551181, |
|
"grad_norm": 11.717874526977539, |
|
"learning_rate": 4.035785288270378e-07, |
|
"loss": 1.6344, |
|
"step": 411 |
|
}, |
|
{ |
|
"epoch": 0.40551181102362205, |
|
"grad_norm": 22.84028434753418, |
|
"learning_rate": 4.045725646123261e-07, |
|
"loss": 2.4177, |
|
"step": 412 |
|
}, |
|
{ |
|
"epoch": 0.406496062992126, |
|
"grad_norm": 15.062310218811035, |
|
"learning_rate": 4.055666003976144e-07, |
|
"loss": 2.2995, |
|
"step": 413 |
|
}, |
|
{ |
|
"epoch": 0.4074803149606299, |
|
"grad_norm": 13.126559257507324, |
|
"learning_rate": 4.065606361829026e-07, |
|
"loss": 1.7681, |
|
"step": 414 |
|
}, |
|
{ |
|
"epoch": 0.40846456692913385, |
|
"grad_norm": 11.371601104736328, |
|
"learning_rate": 4.075546719681909e-07, |
|
"loss": 1.6562, |
|
"step": 415 |
|
}, |
|
{ |
|
"epoch": 0.4094488188976378, |
|
"grad_norm": 13.057047843933105, |
|
"learning_rate": 4.085487077534792e-07, |
|
"loss": 1.8896, |
|
"step": 416 |
|
}, |
|
{ |
|
"epoch": 0.4104330708661417, |
|
"grad_norm": 14.517732620239258, |
|
"learning_rate": 4.095427435387675e-07, |
|
"loss": 2.0671, |
|
"step": 417 |
|
}, |
|
{ |
|
"epoch": 0.41141732283464566, |
|
"grad_norm": 15.08285903930664, |
|
"learning_rate": 4.1053677932405573e-07, |
|
"loss": 1.6097, |
|
"step": 418 |
|
}, |
|
{ |
|
"epoch": 0.4124015748031496, |
|
"grad_norm": 23.114534378051758, |
|
"learning_rate": 4.11530815109344e-07, |
|
"loss": 2.8126, |
|
"step": 419 |
|
}, |
|
{ |
|
"epoch": 0.41338582677165353, |
|
"grad_norm": 12.036443710327148, |
|
"learning_rate": 4.125248508946323e-07, |
|
"loss": 1.7028, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.41437007874015747, |
|
"grad_norm": 11.2406644821167, |
|
"learning_rate": 4.1351888667992054e-07, |
|
"loss": 1.526, |
|
"step": 421 |
|
}, |
|
{ |
|
"epoch": 0.4153543307086614, |
|
"grad_norm": 23.773788452148438, |
|
"learning_rate": 4.1451292246520885e-07, |
|
"loss": 2.5029, |
|
"step": 422 |
|
}, |
|
{ |
|
"epoch": 0.41633858267716534, |
|
"grad_norm": 13.637556076049805, |
|
"learning_rate": 4.1550695825049705e-07, |
|
"loss": 1.7668, |
|
"step": 423 |
|
}, |
|
{ |
|
"epoch": 0.41732283464566927, |
|
"grad_norm": 17.130098342895508, |
|
"learning_rate": 4.1650099403578535e-07, |
|
"loss": 1.9065, |
|
"step": 424 |
|
}, |
|
{ |
|
"epoch": 0.4183070866141732, |
|
"grad_norm": 13.371904373168945, |
|
"learning_rate": 4.174950298210736e-07, |
|
"loss": 1.6645, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.41929133858267714, |
|
"grad_norm": 12.900718688964844, |
|
"learning_rate": 4.184890656063619e-07, |
|
"loss": 1.8075, |
|
"step": 426 |
|
}, |
|
{ |
|
"epoch": 0.4202755905511811, |
|
"grad_norm": 12.189275741577148, |
|
"learning_rate": 4.194831013916502e-07, |
|
"loss": 1.872, |
|
"step": 427 |
|
}, |
|
{ |
|
"epoch": 0.421259842519685, |
|
"grad_norm": 15.679601669311523, |
|
"learning_rate": 4.204771371769384e-07, |
|
"loss": 2.0487, |
|
"step": 428 |
|
}, |
|
{ |
|
"epoch": 0.422244094488189, |
|
"grad_norm": 11.605814933776855, |
|
"learning_rate": 4.214711729622267e-07, |
|
"loss": 1.535, |
|
"step": 429 |
|
}, |
|
{ |
|
"epoch": 0.42322834645669294, |
|
"grad_norm": 11.776004791259766, |
|
"learning_rate": 4.2246520874751497e-07, |
|
"loss": 1.8046, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.4242125984251969, |
|
"grad_norm": 14.196041107177734, |
|
"learning_rate": 4.234592445328033e-07, |
|
"loss": 2.2561, |
|
"step": 431 |
|
}, |
|
{ |
|
"epoch": 0.4251968503937008, |
|
"grad_norm": 13.536731719970703, |
|
"learning_rate": 4.244532803180915e-07, |
|
"loss": 2.0306, |
|
"step": 432 |
|
}, |
|
{ |
|
"epoch": 0.42618110236220474, |
|
"grad_norm": 10.69434642791748, |
|
"learning_rate": 4.254473161033798e-07, |
|
"loss": 2.1311, |
|
"step": 433 |
|
}, |
|
{ |
|
"epoch": 0.4271653543307087, |
|
"grad_norm": 14.568777084350586, |
|
"learning_rate": 4.264413518886681e-07, |
|
"loss": 2.3013, |
|
"step": 434 |
|
}, |
|
{ |
|
"epoch": 0.4281496062992126, |
|
"grad_norm": 11.407812118530273, |
|
"learning_rate": 4.2743538767395634e-07, |
|
"loss": 1.6402, |
|
"step": 435 |
|
}, |
|
{ |
|
"epoch": 0.42913385826771655, |
|
"grad_norm": 13.241829872131348, |
|
"learning_rate": 4.284294234592446e-07, |
|
"loss": 1.9572, |
|
"step": 436 |
|
}, |
|
{ |
|
"epoch": 0.4301181102362205, |
|
"grad_norm": 11.766160011291504, |
|
"learning_rate": 4.2942345924453284e-07, |
|
"loss": 1.6364, |
|
"step": 437 |
|
}, |
|
{ |
|
"epoch": 0.4311023622047244, |
|
"grad_norm": 12.111750602722168, |
|
"learning_rate": 4.3041749502982115e-07, |
|
"loss": 1.446, |
|
"step": 438 |
|
}, |
|
{ |
|
"epoch": 0.43208661417322836, |
|
"grad_norm": 12.901723861694336, |
|
"learning_rate": 4.3141153081510945e-07, |
|
"loss": 1.6009, |
|
"step": 439 |
|
}, |
|
{ |
|
"epoch": 0.4330708661417323, |
|
"grad_norm": 13.12215518951416, |
|
"learning_rate": 4.324055666003977e-07, |
|
"loss": 1.9469, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.4340551181102362, |
|
"grad_norm": 14.054530143737793, |
|
"learning_rate": 4.3339960238568596e-07, |
|
"loss": 2.1951, |
|
"step": 441 |
|
}, |
|
{ |
|
"epoch": 0.43503937007874016, |
|
"grad_norm": 13.3843412399292, |
|
"learning_rate": 4.343936381709742e-07, |
|
"loss": 1.675, |
|
"step": 442 |
|
}, |
|
{ |
|
"epoch": 0.4360236220472441, |
|
"grad_norm": 13.083832740783691, |
|
"learning_rate": 4.353876739562625e-07, |
|
"loss": 1.4182, |
|
"step": 443 |
|
}, |
|
{ |
|
"epoch": 0.43700787401574803, |
|
"grad_norm": 13.504453659057617, |
|
"learning_rate": 4.3638170974155077e-07, |
|
"loss": 2.2317, |
|
"step": 444 |
|
}, |
|
{ |
|
"epoch": 0.43799212598425197, |
|
"grad_norm": 13.614442825317383, |
|
"learning_rate": 4.37375745526839e-07, |
|
"loss": 2.1076, |
|
"step": 445 |
|
}, |
|
{ |
|
"epoch": 0.4389763779527559, |
|
"grad_norm": 12.694330215454102, |
|
"learning_rate": 4.383697813121273e-07, |
|
"loss": 1.6691, |
|
"step": 446 |
|
}, |
|
{ |
|
"epoch": 0.43996062992125984, |
|
"grad_norm": 12.593806266784668, |
|
"learning_rate": 4.393638170974156e-07, |
|
"loss": 1.6909, |
|
"step": 447 |
|
}, |
|
{ |
|
"epoch": 0.4409448818897638, |
|
"grad_norm": 37.6284065246582, |
|
"learning_rate": 4.403578528827039e-07, |
|
"loss": 3.1056, |
|
"step": 448 |
|
}, |
|
{ |
|
"epoch": 0.4419291338582677, |
|
"grad_norm": 12.051369667053223, |
|
"learning_rate": 4.413518886679921e-07, |
|
"loss": 1.4069, |
|
"step": 449 |
|
}, |
|
{ |
|
"epoch": 0.44291338582677164, |
|
"grad_norm": 15.523516654968262, |
|
"learning_rate": 4.423459244532804e-07, |
|
"loss": 2.1639, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.4438976377952756, |
|
"grad_norm": 11.322219848632812, |
|
"learning_rate": 4.4333996023856864e-07, |
|
"loss": 1.5531, |
|
"step": 451 |
|
}, |
|
{ |
|
"epoch": 0.4448818897637795, |
|
"grad_norm": 12.900222778320312, |
|
"learning_rate": 4.4433399602385695e-07, |
|
"loss": 2.1895, |
|
"step": 452 |
|
}, |
|
{ |
|
"epoch": 0.44586614173228345, |
|
"grad_norm": 14.88873291015625, |
|
"learning_rate": 4.4532803180914525e-07, |
|
"loss": 1.9384, |
|
"step": 453 |
|
}, |
|
{ |
|
"epoch": 0.4468503937007874, |
|
"grad_norm": 13.40294075012207, |
|
"learning_rate": 4.4632206759443345e-07, |
|
"loss": 1.7761, |
|
"step": 454 |
|
}, |
|
{ |
|
"epoch": 0.4478346456692913, |
|
"grad_norm": 27.584877014160156, |
|
"learning_rate": 4.4731610337972176e-07, |
|
"loss": 2.8286, |
|
"step": 455 |
|
}, |
|
{ |
|
"epoch": 0.44881889763779526, |
|
"grad_norm": 25.157712936401367, |
|
"learning_rate": 4.4831013916501e-07, |
|
"loss": 2.4877, |
|
"step": 456 |
|
}, |
|
{ |
|
"epoch": 0.4498031496062992, |
|
"grad_norm": 22.14501953125, |
|
"learning_rate": 4.493041749502983e-07, |
|
"loss": 1.7636, |
|
"step": 457 |
|
}, |
|
{ |
|
"epoch": 0.4507874015748031, |
|
"grad_norm": 10.37475872039795, |
|
"learning_rate": 4.502982107355865e-07, |
|
"loss": 1.1849, |
|
"step": 458 |
|
}, |
|
{ |
|
"epoch": 0.45177165354330706, |
|
"grad_norm": 14.996706008911133, |
|
"learning_rate": 4.512922465208748e-07, |
|
"loss": 1.8331, |
|
"step": 459 |
|
}, |
|
{ |
|
"epoch": 0.45177165354330706, |
|
"eval_Qnli-dev_cosine_accuracy": 0.634765625, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.9121971130371094, |
|
"eval_Qnli-dev_cosine_ap": 0.6377161139177543, |
|
"eval_Qnli-dev_cosine_f1": 0.6430868167202571, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.8449763059616089, |
|
"eval_Qnli-dev_cosine_precision": 0.5181347150259067, |
|
"eval_Qnli-dev_cosine_recall": 0.847457627118644, |
|
"eval_Qnli-dev_dot_accuracy": 0.63671875, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 699.1280517578125, |
|
"eval_Qnli-dev_dot_ap": 0.6388138195772171, |
|
"eval_Qnli-dev_dot_f1": 0.6430868167202571, |
|
"eval_Qnli-dev_dot_f1_threshold": 647.91845703125, |
|
"eval_Qnli-dev_dot_precision": 0.5181347150259067, |
|
"eval_Qnli-dev_dot_recall": 0.847457627118644, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.634765625, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 11.602351188659668, |
|
"eval_Qnli-dev_euclidean_ap": 0.6377918821678507, |
|
"eval_Qnli-dev_euclidean_f1": 0.6430868167202571, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 15.418830871582031, |
|
"eval_Qnli-dev_euclidean_precision": 0.5181347150259067, |
|
"eval_Qnli-dev_euclidean_recall": 0.847457627118644, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.642578125, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 233.09597778320312, |
|
"eval_Qnli-dev_manhattan_ap": 0.6510660300493925, |
|
"eval_Qnli-dev_manhattan_f1": 0.6605783866057838, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 315.7362976074219, |
|
"eval_Qnli-dev_manhattan_precision": 0.5154394299287411, |
|
"eval_Qnli-dev_manhattan_recall": 0.9194915254237288, |
|
"eval_Qnli-dev_max_accuracy": 0.642578125, |
|
"eval_Qnli-dev_max_accuracy_threshold": 699.1280517578125, |
|
"eval_Qnli-dev_max_ap": 0.6510660300493925, |
|
"eval_Qnli-dev_max_f1": 0.6605783866057838, |
|
"eval_Qnli-dev_max_f1_threshold": 647.91845703125, |
|
"eval_Qnli-dev_max_precision": 0.5181347150259067, |
|
"eval_Qnli-dev_max_recall": 0.9194915254237288, |
|
"eval_allNLI-dev_cosine_accuracy": 0.673828125, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9788761138916016, |
|
"eval_allNLI-dev_cosine_ap": 0.4327025118722887, |
|
"eval_allNLI-dev_cosine_f1": 0.5157894736842105, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.8889895081520081, |
|
"eval_allNLI-dev_cosine_precision": 0.3702770780856423, |
|
"eval_allNLI-dev_cosine_recall": 0.8497109826589595, |
|
"eval_allNLI-dev_dot_accuracy": 0.673828125, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 751.3026733398438, |
|
"eval_allNLI-dev_dot_ap": 0.43253813511417205, |
|
"eval_allNLI-dev_dot_f1": 0.5166959578207382, |
|
"eval_allNLI-dev_dot_f1_threshold": 682.115234375, |
|
"eval_allNLI-dev_dot_precision": 0.3712121212121212, |
|
"eval_allNLI-dev_dot_recall": 0.8497109826589595, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.673828125, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 5.694375038146973, |
|
"eval_allNLI-dev_euclidean_ap": 0.4325438108928368, |
|
"eval_allNLI-dev_euclidean_f1": 0.5157894736842105, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 13.050301551818848, |
|
"eval_allNLI-dev_euclidean_precision": 0.3702770780856423, |
|
"eval_allNLI-dev_euclidean_recall": 0.8497109826589595, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.671875, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 105.85862731933594, |
|
"eval_allNLI-dev_manhattan_ap": 0.4278948508381489, |
|
"eval_allNLI-dev_manhattan_f1": 0.5217391304347826, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 241.55101013183594, |
|
"eval_allNLI-dev_manhattan_precision": 0.38764044943820225, |
|
"eval_allNLI-dev_manhattan_recall": 0.7976878612716763, |
|
"eval_allNLI-dev_max_accuracy": 0.673828125, |
|
"eval_allNLI-dev_max_accuracy_threshold": 751.3026733398438, |
|
"eval_allNLI-dev_max_ap": 0.4327025118722887, |
|
"eval_allNLI-dev_max_f1": 0.5217391304347826, |
|
"eval_allNLI-dev_max_f1_threshold": 682.115234375, |
|
"eval_allNLI-dev_max_precision": 0.38764044943820225, |
|
"eval_allNLI-dev_max_recall": 0.8497109826589595, |
|
"eval_loss": 1.9853787422180176, |
|
"eval_runtime": 52.6119, |
|
"eval_samples_per_second": 31.628, |
|
"eval_sequential_score": 0.6510660300493925, |
|
"eval_steps_per_second": 0.133, |
|
"eval_sts-test_pearson_cosine": 0.27561429926791464, |
|
"eval_sts-test_pearson_dot": 0.27336219058729005, |
|
"eval_sts-test_pearson_euclidean": 0.29994864031298485, |
|
"eval_sts-test_pearson_manhattan": 0.3112396414398868, |
|
"eval_sts-test_pearson_max": 0.3112396414398868, |
|
"eval_sts-test_spearman_cosine": 0.32606859471811517, |
|
"eval_sts-test_spearman_dot": 0.3235796341494495, |
|
"eval_sts-test_spearman_euclidean": 0.3260361203897462, |
|
"eval_sts-test_spearman_manhattan": 0.3379918226318111, |
|
"eval_sts-test_spearman_max": 0.3379918226318111, |
|
"step": 459 |
|
}, |
|
{ |
|
"epoch": 0.452755905511811, |
|
"grad_norm": 15.168042182922363, |
|
"learning_rate": 4.522862823061631e-07, |
|
"loss": 2.0416, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.453740157480315, |
|
"grad_norm": 12.258943557739258, |
|
"learning_rate": 4.532803180914514e-07, |
|
"loss": 2.1907, |
|
"step": 461 |
|
}, |
|
{ |
|
"epoch": 0.4547244094488189, |
|
"grad_norm": 16.1510009765625, |
|
"learning_rate": 4.542743538767397e-07, |
|
"loss": 1.7478, |
|
"step": 462 |
|
}, |
|
{ |
|
"epoch": 0.45570866141732286, |
|
"grad_norm": 13.124279022216797, |
|
"learning_rate": 4.552683896620279e-07, |
|
"loss": 1.9, |
|
"step": 463 |
|
}, |
|
{ |
|
"epoch": 0.4566929133858268, |
|
"grad_norm": 11.794757843017578, |
|
"learning_rate": 4.562624254473162e-07, |
|
"loss": 1.6876, |
|
"step": 464 |
|
}, |
|
{ |
|
"epoch": 0.45767716535433073, |
|
"grad_norm": 13.155580520629883, |
|
"learning_rate": 4.5725646123260444e-07, |
|
"loss": 2.0035, |
|
"step": 465 |
|
}, |
|
{ |
|
"epoch": 0.45866141732283466, |
|
"grad_norm": 11.5040922164917, |
|
"learning_rate": 4.5825049701789274e-07, |
|
"loss": 1.4127, |
|
"step": 466 |
|
}, |
|
{ |
|
"epoch": 0.4596456692913386, |
|
"grad_norm": 11.074214935302734, |
|
"learning_rate": 4.59244532803181e-07, |
|
"loss": 1.5593, |
|
"step": 467 |
|
}, |
|
{ |
|
"epoch": 0.46062992125984253, |
|
"grad_norm": 13.357093811035156, |
|
"learning_rate": 4.6023856858846925e-07, |
|
"loss": 1.7, |
|
"step": 468 |
|
}, |
|
{ |
|
"epoch": 0.46161417322834647, |
|
"grad_norm": 14.933716773986816, |
|
"learning_rate": 4.6123260437375755e-07, |
|
"loss": 1.5157, |
|
"step": 469 |
|
}, |
|
{ |
|
"epoch": 0.4625984251968504, |
|
"grad_norm": 12.29159164428711, |
|
"learning_rate": 4.622266401590458e-07, |
|
"loss": 1.6554, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.46358267716535434, |
|
"grad_norm": 12.008153915405273, |
|
"learning_rate": 4.6322067594433406e-07, |
|
"loss": 1.7404, |
|
"step": 471 |
|
}, |
|
{ |
|
"epoch": 0.4645669291338583, |
|
"grad_norm": 13.209589958190918, |
|
"learning_rate": 4.642147117296223e-07, |
|
"loss": 2.1432, |
|
"step": 472 |
|
}, |
|
{ |
|
"epoch": 0.4655511811023622, |
|
"grad_norm": 13.247814178466797, |
|
"learning_rate": 4.652087475149106e-07, |
|
"loss": 1.7322, |
|
"step": 473 |
|
}, |
|
{ |
|
"epoch": 0.46653543307086615, |
|
"grad_norm": 12.133121490478516, |
|
"learning_rate": 4.662027833001989e-07, |
|
"loss": 1.7281, |
|
"step": 474 |
|
}, |
|
{ |
|
"epoch": 0.4675196850393701, |
|
"grad_norm": 12.511950492858887, |
|
"learning_rate": 4.671968190854872e-07, |
|
"loss": 1.5107, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.468503937007874, |
|
"grad_norm": 13.9373779296875, |
|
"learning_rate": 4.681908548707754e-07, |
|
"loss": 1.779, |
|
"step": 476 |
|
}, |
|
{ |
|
"epoch": 0.46948818897637795, |
|
"grad_norm": 10.878262519836426, |
|
"learning_rate": 4.691848906560637e-07, |
|
"loss": 1.325, |
|
"step": 477 |
|
}, |
|
{ |
|
"epoch": 0.4704724409448819, |
|
"grad_norm": 11.128897666931152, |
|
"learning_rate": 4.70178926441352e-07, |
|
"loss": 1.073, |
|
"step": 478 |
|
}, |
|
{ |
|
"epoch": 0.4714566929133858, |
|
"grad_norm": 14.144214630126953, |
|
"learning_rate": 4.7117296222664024e-07, |
|
"loss": 1.864, |
|
"step": 479 |
|
}, |
|
{ |
|
"epoch": 0.47244094488188976, |
|
"grad_norm": 15.306927680969238, |
|
"learning_rate": 4.721669980119285e-07, |
|
"loss": 2.3645, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.4734251968503937, |
|
"grad_norm": 10.306663513183594, |
|
"learning_rate": 4.731610337972168e-07, |
|
"loss": 1.181, |
|
"step": 481 |
|
}, |
|
{ |
|
"epoch": 0.4744094488188976, |
|
"grad_norm": 12.236723899841309, |
|
"learning_rate": 4.7415506958250505e-07, |
|
"loss": 1.4562, |
|
"step": 482 |
|
}, |
|
{ |
|
"epoch": 0.47539370078740156, |
|
"grad_norm": 12.010734558105469, |
|
"learning_rate": 4.7514910536779335e-07, |
|
"loss": 1.3105, |
|
"step": 483 |
|
}, |
|
{ |
|
"epoch": 0.4763779527559055, |
|
"grad_norm": 24.507387161254883, |
|
"learning_rate": 4.7614314115308155e-07, |
|
"loss": 2.8012, |
|
"step": 484 |
|
}, |
|
{ |
|
"epoch": 0.47736220472440943, |
|
"grad_norm": 15.473782539367676, |
|
"learning_rate": 4.771371769383698e-07, |
|
"loss": 2.0114, |
|
"step": 485 |
|
}, |
|
{ |
|
"epoch": 0.47834645669291337, |
|
"grad_norm": 13.31457805633545, |
|
"learning_rate": 4.781312127236581e-07, |
|
"loss": 1.6307, |
|
"step": 486 |
|
}, |
|
{ |
|
"epoch": 0.4793307086614173, |
|
"grad_norm": 26.823562622070312, |
|
"learning_rate": 4.791252485089464e-07, |
|
"loss": 2.7733, |
|
"step": 487 |
|
}, |
|
{ |
|
"epoch": 0.48031496062992124, |
|
"grad_norm": 13.975008010864258, |
|
"learning_rate": 4.801192842942347e-07, |
|
"loss": 1.8211, |
|
"step": 488 |
|
}, |
|
{ |
|
"epoch": 0.4812992125984252, |
|
"grad_norm": 12.852688789367676, |
|
"learning_rate": 4.811133200795229e-07, |
|
"loss": 1.574, |
|
"step": 489 |
|
}, |
|
{ |
|
"epoch": 0.4822834645669291, |
|
"grad_norm": 18.221059799194336, |
|
"learning_rate": 4.821073558648112e-07, |
|
"loss": 1.9713, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.48326771653543305, |
|
"grad_norm": 12.657098770141602, |
|
"learning_rate": 4.831013916500995e-07, |
|
"loss": 1.2774, |
|
"step": 491 |
|
}, |
|
{ |
|
"epoch": 0.484251968503937, |
|
"grad_norm": 16.00031280517578, |
|
"learning_rate": 4.840954274353877e-07, |
|
"loss": 2.58, |
|
"step": 492 |
|
}, |
|
{ |
|
"epoch": 0.48523622047244097, |
|
"grad_norm": 12.299332618713379, |
|
"learning_rate": 4.85089463220676e-07, |
|
"loss": 2.0594, |
|
"step": 493 |
|
}, |
|
{ |
|
"epoch": 0.4862204724409449, |
|
"grad_norm": 16.219257354736328, |
|
"learning_rate": 4.860834990059642e-07, |
|
"loss": 1.5857, |
|
"step": 494 |
|
}, |
|
{ |
|
"epoch": 0.48720472440944884, |
|
"grad_norm": 12.892383575439453, |
|
"learning_rate": 4.870775347912525e-07, |
|
"loss": 2.0028, |
|
"step": 495 |
|
}, |
|
{ |
|
"epoch": 0.4881889763779528, |
|
"grad_norm": 16.72523307800293, |
|
"learning_rate": 4.880715705765408e-07, |
|
"loss": 1.8863, |
|
"step": 496 |
|
}, |
|
{ |
|
"epoch": 0.4891732283464567, |
|
"grad_norm": 14.14151382446289, |
|
"learning_rate": 4.890656063618291e-07, |
|
"loss": 1.5171, |
|
"step": 497 |
|
}, |
|
{ |
|
"epoch": 0.49015748031496065, |
|
"grad_norm": 16.287216186523438, |
|
"learning_rate": 4.900596421471173e-07, |
|
"loss": 1.9355, |
|
"step": 498 |
|
}, |
|
{ |
|
"epoch": 0.4911417322834646, |
|
"grad_norm": 16.423860549926758, |
|
"learning_rate": 4.910536779324057e-07, |
|
"loss": 2.0675, |
|
"step": 499 |
|
}, |
|
{ |
|
"epoch": 0.4921259842519685, |
|
"grad_norm": 12.384491920471191, |
|
"learning_rate": 4.92047713717694e-07, |
|
"loss": 1.6017, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.49311023622047245, |
|
"grad_norm": 10.480212211608887, |
|
"learning_rate": 4.930417495029823e-07, |
|
"loss": 1.4089, |
|
"step": 501 |
|
}, |
|
{ |
|
"epoch": 0.4940944881889764, |
|
"grad_norm": 9.902453422546387, |
|
"learning_rate": 4.940357852882705e-07, |
|
"loss": 1.3836, |
|
"step": 502 |
|
}, |
|
{ |
|
"epoch": 0.4950787401574803, |
|
"grad_norm": 12.404056549072266, |
|
"learning_rate": 4.950298210735588e-07, |
|
"loss": 1.6033, |
|
"step": 503 |
|
}, |
|
{ |
|
"epoch": 0.49606299212598426, |
|
"grad_norm": 9.605865478515625, |
|
"learning_rate": 4.96023856858847e-07, |
|
"loss": 1.0891, |
|
"step": 504 |
|
}, |
|
{ |
|
"epoch": 0.4970472440944882, |
|
"grad_norm": 13.224746704101562, |
|
"learning_rate": 4.970178926441353e-07, |
|
"loss": 1.7119, |
|
"step": 505 |
|
}, |
|
{ |
|
"epoch": 0.49803149606299213, |
|
"grad_norm": 13.01671314239502, |
|
"learning_rate": 4.980119284294235e-07, |
|
"loss": 1.3685, |
|
"step": 506 |
|
}, |
|
{ |
|
"epoch": 0.49901574803149606, |
|
"grad_norm": 11.965983390808105, |
|
"learning_rate": 4.990059642147118e-07, |
|
"loss": 1.4252, |
|
"step": 507 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 13.703736305236816, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 1.5538, |
|
"step": 508 |
|
}, |
|
{ |
|
"epoch": 0.5009842519685039, |
|
"grad_norm": 16.96326446533203, |
|
"learning_rate": 5.009940357852883e-07, |
|
"loss": 1.7513, |
|
"step": 509 |
|
}, |
|
{ |
|
"epoch": 0.5019685039370079, |
|
"grad_norm": 9.837270736694336, |
|
"learning_rate": 5.019880715705767e-07, |
|
"loss": 1.1831, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.5029527559055118, |
|
"grad_norm": 13.675102233886719, |
|
"learning_rate": 5.029821073558649e-07, |
|
"loss": 1.7767, |
|
"step": 511 |
|
}, |
|
{ |
|
"epoch": 0.5039370078740157, |
|
"grad_norm": 12.791515350341797, |
|
"learning_rate": 5.039761431411532e-07, |
|
"loss": 1.4324, |
|
"step": 512 |
|
}, |
|
{ |
|
"epoch": 0.5049212598425197, |
|
"grad_norm": 18.81296157836914, |
|
"learning_rate": 5.049701789264414e-07, |
|
"loss": 2.1672, |
|
"step": 513 |
|
}, |
|
{ |
|
"epoch": 0.5059055118110236, |
|
"grad_norm": 13.022212028503418, |
|
"learning_rate": 5.059642147117297e-07, |
|
"loss": 1.6348, |
|
"step": 514 |
|
}, |
|
{ |
|
"epoch": 0.5068897637795275, |
|
"grad_norm": 16.152873992919922, |
|
"learning_rate": 5.06958250497018e-07, |
|
"loss": 1.7285, |
|
"step": 515 |
|
}, |
|
{ |
|
"epoch": 0.5078740157480315, |
|
"grad_norm": 19.10329818725586, |
|
"learning_rate": 5.079522862823062e-07, |
|
"loss": 2.0186, |
|
"step": 516 |
|
}, |
|
{ |
|
"epoch": 0.5088582677165354, |
|
"grad_norm": 12.92306137084961, |
|
"learning_rate": 5.089463220675945e-07, |
|
"loss": 1.382, |
|
"step": 517 |
|
}, |
|
{ |
|
"epoch": 0.5098425196850394, |
|
"grad_norm": 12.875118255615234, |
|
"learning_rate": 5.099403578528828e-07, |
|
"loss": 1.4509, |
|
"step": 518 |
|
}, |
|
{ |
|
"epoch": 0.5108267716535433, |
|
"grad_norm": 10.416156768798828, |
|
"learning_rate": 5.10934393638171e-07, |
|
"loss": 1.1043, |
|
"step": 519 |
|
}, |
|
{ |
|
"epoch": 0.5118110236220472, |
|
"grad_norm": 12.095355987548828, |
|
"learning_rate": 5.119284294234593e-07, |
|
"loss": 1.3322, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.5127952755905512, |
|
"grad_norm": 10.407857894897461, |
|
"learning_rate": 5.129224652087476e-07, |
|
"loss": 1.3267, |
|
"step": 521 |
|
}, |
|
{ |
|
"epoch": 0.5137795275590551, |
|
"grad_norm": 12.354635238647461, |
|
"learning_rate": 5.139165009940358e-07, |
|
"loss": 1.3639, |
|
"step": 522 |
|
}, |
|
{ |
|
"epoch": 0.514763779527559, |
|
"grad_norm": 11.841660499572754, |
|
"learning_rate": 5.149105367793241e-07, |
|
"loss": 1.203, |
|
"step": 523 |
|
}, |
|
{ |
|
"epoch": 0.515748031496063, |
|
"grad_norm": 15.488195419311523, |
|
"learning_rate": 5.159045725646124e-07, |
|
"loss": 1.8583, |
|
"step": 524 |
|
}, |
|
{ |
|
"epoch": 0.5167322834645669, |
|
"grad_norm": 18.47950553894043, |
|
"learning_rate": 5.168986083499007e-07, |
|
"loss": 2.267, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.5177165354330708, |
|
"grad_norm": 11.875829696655273, |
|
"learning_rate": 5.178926441351889e-07, |
|
"loss": 1.2935, |
|
"step": 526 |
|
}, |
|
{ |
|
"epoch": 0.5187007874015748, |
|
"grad_norm": 13.655136108398438, |
|
"learning_rate": 5.188866799204771e-07, |
|
"loss": 1.7431, |
|
"step": 527 |
|
}, |
|
{ |
|
"epoch": 0.5196850393700787, |
|
"grad_norm": 16.041044235229492, |
|
"learning_rate": 5.198807157057656e-07, |
|
"loss": 1.8484, |
|
"step": 528 |
|
}, |
|
{ |
|
"epoch": 0.5206692913385826, |
|
"grad_norm": 15.453495979309082, |
|
"learning_rate": 5.208747514910538e-07, |
|
"loss": 1.5626, |
|
"step": 529 |
|
}, |
|
{ |
|
"epoch": 0.5216535433070866, |
|
"grad_norm": 18.945079803466797, |
|
"learning_rate": 5.21868787276342e-07, |
|
"loss": 2.2645, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.5226377952755905, |
|
"grad_norm": 12.458165168762207, |
|
"learning_rate": 5.228628230616304e-07, |
|
"loss": 1.4313, |
|
"step": 531 |
|
}, |
|
{ |
|
"epoch": 0.5236220472440944, |
|
"grad_norm": 13.550048828125, |
|
"learning_rate": 5.238568588469186e-07, |
|
"loss": 1.8204, |
|
"step": 532 |
|
}, |
|
{ |
|
"epoch": 0.5246062992125984, |
|
"grad_norm": 14.208359718322754, |
|
"learning_rate": 5.248508946322069e-07, |
|
"loss": 1.5659, |
|
"step": 533 |
|
}, |
|
{ |
|
"epoch": 0.5255905511811023, |
|
"grad_norm": 11.332735061645508, |
|
"learning_rate": 5.258449304174951e-07, |
|
"loss": 1.2689, |
|
"step": 534 |
|
}, |
|
{ |
|
"epoch": 0.5265748031496063, |
|
"grad_norm": 12.082714080810547, |
|
"learning_rate": 5.268389662027834e-07, |
|
"loss": 1.8193, |
|
"step": 535 |
|
}, |
|
{ |
|
"epoch": 0.5275590551181102, |
|
"grad_norm": 15.39877700805664, |
|
"learning_rate": 5.278330019880717e-07, |
|
"loss": 2.2902, |
|
"step": 536 |
|
}, |
|
{ |
|
"epoch": 0.5285433070866141, |
|
"grad_norm": 11.852607727050781, |
|
"learning_rate": 5.288270377733599e-07, |
|
"loss": 1.6936, |
|
"step": 537 |
|
}, |
|
{ |
|
"epoch": 0.5295275590551181, |
|
"grad_norm": 12.435094833374023, |
|
"learning_rate": 5.298210735586483e-07, |
|
"loss": 1.7305, |
|
"step": 538 |
|
}, |
|
{ |
|
"epoch": 0.530511811023622, |
|
"grad_norm": 12.685184478759766, |
|
"learning_rate": 5.308151093439365e-07, |
|
"loss": 1.4449, |
|
"step": 539 |
|
}, |
|
{ |
|
"epoch": 0.531496062992126, |
|
"grad_norm": 12.26480484008789, |
|
"learning_rate": 5.318091451292247e-07, |
|
"loss": 1.5594, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.53248031496063, |
|
"grad_norm": 14.554289817810059, |
|
"learning_rate": 5.32803180914513e-07, |
|
"loss": 1.9678, |
|
"step": 541 |
|
}, |
|
{ |
|
"epoch": 0.5334645669291339, |
|
"grad_norm": 13.191105842590332, |
|
"learning_rate": 5.337972166998013e-07, |
|
"loss": 2.0327, |
|
"step": 542 |
|
}, |
|
{ |
|
"epoch": 0.5344488188976378, |
|
"grad_norm": 14.845519065856934, |
|
"learning_rate": 5.347912524850895e-07, |
|
"loss": 2.0456, |
|
"step": 543 |
|
}, |
|
{ |
|
"epoch": 0.5354330708661418, |
|
"grad_norm": 14.311629295349121, |
|
"learning_rate": 5.357852882703778e-07, |
|
"loss": 2.0452, |
|
"step": 544 |
|
}, |
|
{ |
|
"epoch": 0.5364173228346457, |
|
"grad_norm": 14.074420928955078, |
|
"learning_rate": 5.367793240556661e-07, |
|
"loss": 1.9435, |
|
"step": 545 |
|
}, |
|
{ |
|
"epoch": 0.5374015748031497, |
|
"grad_norm": 16.501874923706055, |
|
"learning_rate": 5.377733598409544e-07, |
|
"loss": 1.8963, |
|
"step": 546 |
|
}, |
|
{ |
|
"epoch": 0.5383858267716536, |
|
"grad_norm": 16.289310455322266, |
|
"learning_rate": 5.387673956262426e-07, |
|
"loss": 1.9536, |
|
"step": 547 |
|
}, |
|
{ |
|
"epoch": 0.5393700787401575, |
|
"grad_norm": 11.657831192016602, |
|
"learning_rate": 5.397614314115308e-07, |
|
"loss": 1.0665, |
|
"step": 548 |
|
}, |
|
{ |
|
"epoch": 0.5403543307086615, |
|
"grad_norm": 17.12558364868164, |
|
"learning_rate": 5.407554671968192e-07, |
|
"loss": 1.8067, |
|
"step": 549 |
|
}, |
|
{ |
|
"epoch": 0.5413385826771654, |
|
"grad_norm": 12.209671974182129, |
|
"learning_rate": 5.417495029821074e-07, |
|
"loss": 1.6227, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.5423228346456693, |
|
"grad_norm": 13.168261528015137, |
|
"learning_rate": 5.427435387673957e-07, |
|
"loss": 1.687, |
|
"step": 551 |
|
}, |
|
{ |
|
"epoch": 0.5433070866141733, |
|
"grad_norm": 11.820549011230469, |
|
"learning_rate": 5.43737574552684e-07, |
|
"loss": 1.5937, |
|
"step": 552 |
|
}, |
|
{ |
|
"epoch": 0.5442913385826772, |
|
"grad_norm": 12.037482261657715, |
|
"learning_rate": 5.447316103379722e-07, |
|
"loss": 0.9216, |
|
"step": 553 |
|
}, |
|
{ |
|
"epoch": 0.5452755905511811, |
|
"grad_norm": 15.135348320007324, |
|
"learning_rate": 5.457256461232605e-07, |
|
"loss": 1.3895, |
|
"step": 554 |
|
}, |
|
{ |
|
"epoch": 0.5462598425196851, |
|
"grad_norm": 16.537757873535156, |
|
"learning_rate": 5.467196819085487e-07, |
|
"loss": 1.7863, |
|
"step": 555 |
|
}, |
|
{ |
|
"epoch": 0.547244094488189, |
|
"grad_norm": 10.809349060058594, |
|
"learning_rate": 5.477137176938371e-07, |
|
"loss": 1.2574, |
|
"step": 556 |
|
}, |
|
{ |
|
"epoch": 0.5482283464566929, |
|
"grad_norm": 17.314552307128906, |
|
"learning_rate": 5.487077534791253e-07, |
|
"loss": 2.108, |
|
"step": 557 |
|
}, |
|
{ |
|
"epoch": 0.5492125984251969, |
|
"grad_norm": 11.581759452819824, |
|
"learning_rate": 5.497017892644135e-07, |
|
"loss": 1.2782, |
|
"step": 558 |
|
}, |
|
{ |
|
"epoch": 0.5501968503937008, |
|
"grad_norm": 11.438880920410156, |
|
"learning_rate": 5.50695825049702e-07, |
|
"loss": 1.4959, |
|
"step": 559 |
|
}, |
|
{ |
|
"epoch": 0.5511811023622047, |
|
"grad_norm": 12.535311698913574, |
|
"learning_rate": 5.516898608349902e-07, |
|
"loss": 1.9191, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.5521653543307087, |
|
"grad_norm": 17.22417449951172, |
|
"learning_rate": 5.526838966202784e-07, |
|
"loss": 2.0049, |
|
"step": 561 |
|
}, |
|
{ |
|
"epoch": 0.5531496062992126, |
|
"grad_norm": 11.439227104187012, |
|
"learning_rate": 5.536779324055667e-07, |
|
"loss": 1.2511, |
|
"step": 562 |
|
}, |
|
{ |
|
"epoch": 0.5541338582677166, |
|
"grad_norm": 13.34407901763916, |
|
"learning_rate": 5.54671968190855e-07, |
|
"loss": 1.3912, |
|
"step": 563 |
|
}, |
|
{ |
|
"epoch": 0.5551181102362205, |
|
"grad_norm": 13.90283203125, |
|
"learning_rate": 5.556660039761433e-07, |
|
"loss": 1.371, |
|
"step": 564 |
|
}, |
|
{ |
|
"epoch": 0.5561023622047244, |
|
"grad_norm": 18.280315399169922, |
|
"learning_rate": 5.566600397614315e-07, |
|
"loss": 1.6155, |
|
"step": 565 |
|
}, |
|
{ |
|
"epoch": 0.5570866141732284, |
|
"grad_norm": 10.993673324584961, |
|
"learning_rate": 5.576540755467198e-07, |
|
"loss": 1.4625, |
|
"step": 566 |
|
}, |
|
{ |
|
"epoch": 0.5580708661417323, |
|
"grad_norm": 9.647687911987305, |
|
"learning_rate": 5.586481113320081e-07, |
|
"loss": 0.86, |
|
"step": 567 |
|
}, |
|
{ |
|
"epoch": 0.5590551181102362, |
|
"grad_norm": 14.242208480834961, |
|
"learning_rate": 5.596421471172963e-07, |
|
"loss": 1.5753, |
|
"step": 568 |
|
}, |
|
{ |
|
"epoch": 0.5600393700787402, |
|
"grad_norm": 11.613588333129883, |
|
"learning_rate": 5.606361829025846e-07, |
|
"loss": 1.6126, |
|
"step": 569 |
|
}, |
|
{ |
|
"epoch": 0.5610236220472441, |
|
"grad_norm": 11.57154369354248, |
|
"learning_rate": 5.616302186878729e-07, |
|
"loss": 1.3171, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.562007874015748, |
|
"grad_norm": 15.930388450622559, |
|
"learning_rate": 5.626242544731611e-07, |
|
"loss": 1.9378, |
|
"step": 571 |
|
}, |
|
{ |
|
"epoch": 0.562992125984252, |
|
"grad_norm": 12.384969711303711, |
|
"learning_rate": 5.636182902584494e-07, |
|
"loss": 1.2736, |
|
"step": 572 |
|
}, |
|
{ |
|
"epoch": 0.5639763779527559, |
|
"grad_norm": 10.599091529846191, |
|
"learning_rate": 5.646123260437377e-07, |
|
"loss": 1.2368, |
|
"step": 573 |
|
}, |
|
{ |
|
"epoch": 0.5649606299212598, |
|
"grad_norm": 10.539170265197754, |
|
"learning_rate": 5.656063618290259e-07, |
|
"loss": 1.1005, |
|
"step": 574 |
|
}, |
|
{ |
|
"epoch": 0.5659448818897638, |
|
"grad_norm": 11.336695671081543, |
|
"learning_rate": 5.666003976143142e-07, |
|
"loss": 1.1765, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.5669291338582677, |
|
"grad_norm": 12.062975883483887, |
|
"learning_rate": 5.675944333996024e-07, |
|
"loss": 1.3557, |
|
"step": 576 |
|
}, |
|
{ |
|
"epoch": 0.5679133858267716, |
|
"grad_norm": 12.290358543395996, |
|
"learning_rate": 5.685884691848908e-07, |
|
"loss": 1.3224, |
|
"step": 577 |
|
}, |
|
{ |
|
"epoch": 0.5688976377952756, |
|
"grad_norm": 17.76947593688965, |
|
"learning_rate": 5.69582504970179e-07, |
|
"loss": 1.7914, |
|
"step": 578 |
|
}, |
|
{ |
|
"epoch": 0.5698818897637795, |
|
"grad_norm": 10.218934059143066, |
|
"learning_rate": 5.705765407554672e-07, |
|
"loss": 1.0633, |
|
"step": 579 |
|
}, |
|
{ |
|
"epoch": 0.5708661417322834, |
|
"grad_norm": 12.503771781921387, |
|
"learning_rate": 5.715705765407556e-07, |
|
"loss": 1.3624, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.5718503937007874, |
|
"grad_norm": 11.047056198120117, |
|
"learning_rate": 5.725646123260438e-07, |
|
"loss": 0.9804, |
|
"step": 581 |
|
}, |
|
{ |
|
"epoch": 0.5728346456692913, |
|
"grad_norm": 17.329021453857422, |
|
"learning_rate": 5.735586481113321e-07, |
|
"loss": 1.8246, |
|
"step": 582 |
|
}, |
|
{ |
|
"epoch": 0.5738188976377953, |
|
"grad_norm": 13.590007781982422, |
|
"learning_rate": 5.745526838966203e-07, |
|
"loss": 1.1806, |
|
"step": 583 |
|
}, |
|
{ |
|
"epoch": 0.5748031496062992, |
|
"grad_norm": 12.608687400817871, |
|
"learning_rate": 5.755467196819086e-07, |
|
"loss": 1.6243, |
|
"step": 584 |
|
}, |
|
{ |
|
"epoch": 0.5757874015748031, |
|
"grad_norm": 17.16189956665039, |
|
"learning_rate": 5.765407554671969e-07, |
|
"loss": 1.739, |
|
"step": 585 |
|
}, |
|
{ |
|
"epoch": 0.5767716535433071, |
|
"grad_norm": 11.89455509185791, |
|
"learning_rate": 5.775347912524851e-07, |
|
"loss": 1.2502, |
|
"step": 586 |
|
}, |
|
{ |
|
"epoch": 0.577755905511811, |
|
"grad_norm": 17.690635681152344, |
|
"learning_rate": 5.785288270377734e-07, |
|
"loss": 1.6328, |
|
"step": 587 |
|
}, |
|
{ |
|
"epoch": 0.5787401574803149, |
|
"grad_norm": 12.663801193237305, |
|
"learning_rate": 5.795228628230618e-07, |
|
"loss": 1.3618, |
|
"step": 588 |
|
}, |
|
{ |
|
"epoch": 0.5797244094488189, |
|
"grad_norm": 14.101471900939941, |
|
"learning_rate": 5.8051689860835e-07, |
|
"loss": 1.1535, |
|
"step": 589 |
|
}, |
|
{ |
|
"epoch": 0.5807086614173228, |
|
"grad_norm": 13.188057899475098, |
|
"learning_rate": 5.815109343936383e-07, |
|
"loss": 1.2214, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.5816929133858267, |
|
"grad_norm": 14.179088592529297, |
|
"learning_rate": 5.825049701789266e-07, |
|
"loss": 1.4884, |
|
"step": 591 |
|
}, |
|
{ |
|
"epoch": 0.5826771653543307, |
|
"grad_norm": 10.61578369140625, |
|
"learning_rate": 5.834990059642148e-07, |
|
"loss": 1.4029, |
|
"step": 592 |
|
}, |
|
{ |
|
"epoch": 0.5836614173228346, |
|
"grad_norm": 10.220196723937988, |
|
"learning_rate": 5.844930417495031e-07, |
|
"loss": 1.0542, |
|
"step": 593 |
|
}, |
|
{ |
|
"epoch": 0.5846456692913385, |
|
"grad_norm": 12.367060661315918, |
|
"learning_rate": 5.854870775347914e-07, |
|
"loss": 1.5848, |
|
"step": 594 |
|
}, |
|
{ |
|
"epoch": 0.5856299212598425, |
|
"grad_norm": 11.365569114685059, |
|
"learning_rate": 5.864811133200797e-07, |
|
"loss": 1.405, |
|
"step": 595 |
|
}, |
|
{ |
|
"epoch": 0.5866141732283464, |
|
"grad_norm": 12.305469512939453, |
|
"learning_rate": 5.874751491053679e-07, |
|
"loss": 1.6281, |
|
"step": 596 |
|
}, |
|
{ |
|
"epoch": 0.5875984251968503, |
|
"grad_norm": 12.690118789672852, |
|
"learning_rate": 5.884691848906561e-07, |
|
"loss": 1.5228, |
|
"step": 597 |
|
}, |
|
{ |
|
"epoch": 0.5885826771653543, |
|
"grad_norm": 14.444221496582031, |
|
"learning_rate": 5.894632206759445e-07, |
|
"loss": 1.8192, |
|
"step": 598 |
|
}, |
|
{ |
|
"epoch": 0.5895669291338582, |
|
"grad_norm": 10.509269714355469, |
|
"learning_rate": 5.904572564612327e-07, |
|
"loss": 1.2403, |
|
"step": 599 |
|
}, |
|
{ |
|
"epoch": 0.5905511811023622, |
|
"grad_norm": 16.332834243774414, |
|
"learning_rate": 5.914512922465209e-07, |
|
"loss": 1.9368, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.5915354330708661, |
|
"grad_norm": 14.727821350097656, |
|
"learning_rate": 5.924453280318093e-07, |
|
"loss": 1.6623, |
|
"step": 601 |
|
}, |
|
{ |
|
"epoch": 0.59251968503937, |
|
"grad_norm": 14.168722152709961, |
|
"learning_rate": 5.934393638170975e-07, |
|
"loss": 1.495, |
|
"step": 602 |
|
}, |
|
{ |
|
"epoch": 0.593503937007874, |
|
"grad_norm": 12.486218452453613, |
|
"learning_rate": 5.944333996023858e-07, |
|
"loss": 1.7079, |
|
"step": 603 |
|
}, |
|
{ |
|
"epoch": 0.594488188976378, |
|
"grad_norm": 11.449934959411621, |
|
"learning_rate": 5.95427435387674e-07, |
|
"loss": 1.0651, |
|
"step": 604 |
|
}, |
|
{ |
|
"epoch": 0.5954724409448819, |
|
"grad_norm": 11.699078559875488, |
|
"learning_rate": 5.964214711729623e-07, |
|
"loss": 1.2121, |
|
"step": 605 |
|
}, |
|
{ |
|
"epoch": 0.5964566929133859, |
|
"grad_norm": 12.642477989196777, |
|
"learning_rate": 5.974155069582506e-07, |
|
"loss": 1.5385, |
|
"step": 606 |
|
}, |
|
{ |
|
"epoch": 0.5974409448818898, |
|
"grad_norm": 13.445345878601074, |
|
"learning_rate": 5.984095427435388e-07, |
|
"loss": 1.1015, |
|
"step": 607 |
|
}, |
|
{ |
|
"epoch": 0.5984251968503937, |
|
"grad_norm": 15.42099380493164, |
|
"learning_rate": 5.994035785288272e-07, |
|
"loss": 1.7596, |
|
"step": 608 |
|
}, |
|
{ |
|
"epoch": 0.5994094488188977, |
|
"grad_norm": 13.161843299865723, |
|
"learning_rate": 6.003976143141154e-07, |
|
"loss": 1.5597, |
|
"step": 609 |
|
}, |
|
{ |
|
"epoch": 0.6003937007874016, |
|
"grad_norm": 12.604528427124023, |
|
"learning_rate": 6.013916500994036e-07, |
|
"loss": 1.3254, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.6013779527559056, |
|
"grad_norm": 12.85647964477539, |
|
"learning_rate": 6.02385685884692e-07, |
|
"loss": 1.3595, |
|
"step": 611 |
|
}, |
|
{ |
|
"epoch": 0.6023622047244095, |
|
"grad_norm": 11.856555938720703, |
|
"learning_rate": 6.033797216699802e-07, |
|
"loss": 1.0326, |
|
"step": 612 |
|
}, |
|
{ |
|
"epoch": 0.6023622047244095, |
|
"eval_Qnli-dev_cosine_accuracy": 0.646484375, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.8883535861968994, |
|
"eval_Qnli-dev_cosine_ap": 0.6529248124763861, |
|
"eval_Qnli-dev_cosine_f1": 0.656195462478185, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.8338419795036316, |
|
"eval_Qnli-dev_cosine_precision": 0.5578635014836796, |
|
"eval_Qnli-dev_cosine_recall": 0.7966101694915254, |
|
"eval_Qnli-dev_dot_accuracy": 0.64453125, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 681.0067138671875, |
|
"eval_Qnli-dev_dot_ap": 0.6519832432307548, |
|
"eval_Qnli-dev_dot_f1": 0.657293497363796, |
|
"eval_Qnli-dev_dot_f1_threshold": 639.3973999023438, |
|
"eval_Qnli-dev_dot_precision": 0.5615615615615616, |
|
"eval_Qnli-dev_dot_recall": 0.7923728813559322, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.646484375, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 13.06938362121582, |
|
"eval_Qnli-dev_euclidean_ap": 0.6529216999296377, |
|
"eval_Qnli-dev_euclidean_f1": 0.656195462478185, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 15.943912506103516, |
|
"eval_Qnli-dev_euclidean_precision": 0.5578635014836796, |
|
"eval_Qnli-dev_euclidean_recall": 0.7966101694915254, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.65234375, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 260.0885314941406, |
|
"eval_Qnli-dev_manhattan_ap": 0.6597671356125537, |
|
"eval_Qnli-dev_manhattan_f1": 0.6655948553054662, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 340.8906555175781, |
|
"eval_Qnli-dev_manhattan_precision": 0.5362694300518135, |
|
"eval_Qnli-dev_manhattan_recall": 0.8771186440677966, |
|
"eval_Qnli-dev_max_accuracy": 0.65234375, |
|
"eval_Qnli-dev_max_accuracy_threshold": 681.0067138671875, |
|
"eval_Qnli-dev_max_ap": 0.6597671356125537, |
|
"eval_Qnli-dev_max_f1": 0.6655948553054662, |
|
"eval_Qnli-dev_max_f1_threshold": 639.3973999023438, |
|
"eval_Qnli-dev_max_precision": 0.5615615615615616, |
|
"eval_Qnli-dev_max_recall": 0.8771186440677966, |
|
"eval_allNLI-dev_cosine_accuracy": 0.671875, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9779053926467896, |
|
"eval_allNLI-dev_cosine_ap": 0.447770779436667, |
|
"eval_allNLI-dev_cosine_f1": 0.5192629815745394, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.8607640862464905, |
|
"eval_allNLI-dev_cosine_precision": 0.36556603773584906, |
|
"eval_allNLI-dev_cosine_recall": 0.8959537572254336, |
|
"eval_allNLI-dev_dot_accuracy": 0.671875, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 749.451904296875, |
|
"eval_allNLI-dev_dot_ap": 0.4471854335007342, |
|
"eval_allNLI-dev_dot_f1": 0.5192629815745394, |
|
"eval_allNLI-dev_dot_f1_threshold": 659.1720581054688, |
|
"eval_allNLI-dev_dot_precision": 0.36556603773584906, |
|
"eval_allNLI-dev_dot_recall": 0.8959537572254336, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.671875, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 5.819397449493408, |
|
"eval_allNLI-dev_euclidean_ap": 0.44774506125516483, |
|
"eval_allNLI-dev_euclidean_f1": 0.5192629815745394, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 14.603199005126953, |
|
"eval_allNLI-dev_euclidean_precision": 0.36556603773584906, |
|
"eval_allNLI-dev_euclidean_recall": 0.8959537572254336, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.671875, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 119.47159576416016, |
|
"eval_allNLI-dev_manhattan_ap": 0.4448387653657968, |
|
"eval_allNLI-dev_manhattan_f1": 0.5315904139433552, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 245.42477416992188, |
|
"eval_allNLI-dev_manhattan_precision": 0.42657342657342656, |
|
"eval_allNLI-dev_manhattan_recall": 0.7052023121387283, |
|
"eval_allNLI-dev_max_accuracy": 0.671875, |
|
"eval_allNLI-dev_max_accuracy_threshold": 749.451904296875, |
|
"eval_allNLI-dev_max_ap": 0.447770779436667, |
|
"eval_allNLI-dev_max_f1": 0.5315904139433552, |
|
"eval_allNLI-dev_max_f1_threshold": 659.1720581054688, |
|
"eval_allNLI-dev_max_precision": 0.42657342657342656, |
|
"eval_allNLI-dev_max_recall": 0.8959537572254336, |
|
"eval_loss": 1.6315439939498901, |
|
"eval_runtime": 52.9207, |
|
"eval_samples_per_second": 31.443, |
|
"eval_sequential_score": 0.6597671356125537, |
|
"eval_steps_per_second": 0.132, |
|
"eval_sts-test_pearson_cosine": 0.31607787444026963, |
|
"eval_sts-test_pearson_dot": 0.31281708435791855, |
|
"eval_sts-test_pearson_euclidean": 0.33857077884567166, |
|
"eval_sts-test_pearson_manhattan": 0.3555732019703781, |
|
"eval_sts-test_pearson_max": 0.3555732019703781, |
|
"eval_sts-test_spearman_cosine": 0.35971175784560405, |
|
"eval_sts-test_spearman_dot": 0.3566100031673048, |
|
"eval_sts-test_spearman_euclidean": 0.35988199164393697, |
|
"eval_sts-test_spearman_manhattan": 0.376050443379604, |
|
"eval_sts-test_spearman_max": 0.376050443379604, |
|
"step": 612 |
|
}, |
|
{ |
|
"epoch": 0.6033464566929134, |
|
"grad_norm": 15.954797744750977, |
|
"learning_rate": 6.043737574552684e-07, |
|
"loss": 1.4822, |
|
"step": 613 |
|
}, |
|
{ |
|
"epoch": 0.6043307086614174, |
|
"grad_norm": 11.097407341003418, |
|
"learning_rate": 6.053677932405567e-07, |
|
"loss": 0.9997, |
|
"step": 614 |
|
}, |
|
{ |
|
"epoch": 0.6053149606299213, |
|
"grad_norm": 14.609465599060059, |
|
"learning_rate": 6.06361829025845e-07, |
|
"loss": 1.4946, |
|
"step": 615 |
|
}, |
|
{ |
|
"epoch": 0.6062992125984252, |
|
"grad_norm": 11.441244125366211, |
|
"learning_rate": 6.073558648111333e-07, |
|
"loss": 1.3491, |
|
"step": 616 |
|
}, |
|
{ |
|
"epoch": 0.6072834645669292, |
|
"grad_norm": 12.638875961303711, |
|
"learning_rate": 6.083499005964215e-07, |
|
"loss": 1.2118, |
|
"step": 617 |
|
}, |
|
{ |
|
"epoch": 0.6082677165354331, |
|
"grad_norm": 15.746068000793457, |
|
"learning_rate": 6.093439363817097e-07, |
|
"loss": 1.6948, |
|
"step": 618 |
|
}, |
|
{ |
|
"epoch": 0.609251968503937, |
|
"grad_norm": 20.515592575073242, |
|
"learning_rate": 6.103379721669982e-07, |
|
"loss": 2.0457, |
|
"step": 619 |
|
}, |
|
{ |
|
"epoch": 0.610236220472441, |
|
"grad_norm": 13.20988655090332, |
|
"learning_rate": 6.113320079522864e-07, |
|
"loss": 1.2724, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.6112204724409449, |
|
"grad_norm": 11.521592140197754, |
|
"learning_rate": 6.123260437375747e-07, |
|
"loss": 1.2677, |
|
"step": 621 |
|
}, |
|
{ |
|
"epoch": 0.6122047244094488, |
|
"grad_norm": 14.984801292419434, |
|
"learning_rate": 6.13320079522863e-07, |
|
"loss": 1.6639, |
|
"step": 622 |
|
}, |
|
{ |
|
"epoch": 0.6131889763779528, |
|
"grad_norm": 13.088191032409668, |
|
"learning_rate": 6.143141153081512e-07, |
|
"loss": 1.139, |
|
"step": 623 |
|
}, |
|
{ |
|
"epoch": 0.6141732283464567, |
|
"grad_norm": 14.182477951049805, |
|
"learning_rate": 6.153081510934395e-07, |
|
"loss": 1.2424, |
|
"step": 624 |
|
}, |
|
{ |
|
"epoch": 0.6151574803149606, |
|
"grad_norm": 10.056435585021973, |
|
"learning_rate": 6.163021868787277e-07, |
|
"loss": 0.8214, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.6161417322834646, |
|
"grad_norm": 11.403252601623535, |
|
"learning_rate": 6.172962226640161e-07, |
|
"loss": 1.4398, |
|
"step": 626 |
|
}, |
|
{ |
|
"epoch": 0.6171259842519685, |
|
"grad_norm": 12.417586326599121, |
|
"learning_rate": 6.182902584493043e-07, |
|
"loss": 1.6008, |
|
"step": 627 |
|
}, |
|
{ |
|
"epoch": 0.6181102362204725, |
|
"grad_norm": 14.504150390625, |
|
"learning_rate": 6.192842942345925e-07, |
|
"loss": 1.3225, |
|
"step": 628 |
|
}, |
|
{ |
|
"epoch": 0.6190944881889764, |
|
"grad_norm": 15.959015846252441, |
|
"learning_rate": 6.202783300198809e-07, |
|
"loss": 1.4452, |
|
"step": 629 |
|
}, |
|
{ |
|
"epoch": 0.6200787401574803, |
|
"grad_norm": 17.7681827545166, |
|
"learning_rate": 6.212723658051691e-07, |
|
"loss": 1.7723, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.6210629921259843, |
|
"grad_norm": 10.61798095703125, |
|
"learning_rate": 6.222664015904573e-07, |
|
"loss": 1.035, |
|
"step": 631 |
|
}, |
|
{ |
|
"epoch": 0.6220472440944882, |
|
"grad_norm": 13.359830856323242, |
|
"learning_rate": 6.232604373757457e-07, |
|
"loss": 1.6132, |
|
"step": 632 |
|
}, |
|
{ |
|
"epoch": 0.6230314960629921, |
|
"grad_norm": 11.638774871826172, |
|
"learning_rate": 6.242544731610339e-07, |
|
"loss": 1.4167, |
|
"step": 633 |
|
}, |
|
{ |
|
"epoch": 0.6240157480314961, |
|
"grad_norm": 11.718339920043945, |
|
"learning_rate": 6.252485089463222e-07, |
|
"loss": 1.1764, |
|
"step": 634 |
|
}, |
|
{ |
|
"epoch": 0.625, |
|
"grad_norm": 9.881278991699219, |
|
"learning_rate": 6.262425447316104e-07, |
|
"loss": 1.0702, |
|
"step": 635 |
|
}, |
|
{ |
|
"epoch": 0.6259842519685039, |
|
"grad_norm": 15.579713821411133, |
|
"learning_rate": 6.272365805168987e-07, |
|
"loss": 1.3096, |
|
"step": 636 |
|
}, |
|
{ |
|
"epoch": 0.6269685039370079, |
|
"grad_norm": 12.148575782775879, |
|
"learning_rate": 6.28230616302187e-07, |
|
"loss": 1.4174, |
|
"step": 637 |
|
}, |
|
{ |
|
"epoch": 0.6279527559055118, |
|
"grad_norm": 13.778498649597168, |
|
"learning_rate": 6.292246520874752e-07, |
|
"loss": 1.8742, |
|
"step": 638 |
|
}, |
|
{ |
|
"epoch": 0.6289370078740157, |
|
"grad_norm": 11.97413158416748, |
|
"learning_rate": 6.302186878727636e-07, |
|
"loss": 1.3372, |
|
"step": 639 |
|
}, |
|
{ |
|
"epoch": 0.6299212598425197, |
|
"grad_norm": 11.662862777709961, |
|
"learning_rate": 6.312127236580518e-07, |
|
"loss": 1.4453, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.6309055118110236, |
|
"grad_norm": 10.738568305969238, |
|
"learning_rate": 6.3220675944334e-07, |
|
"loss": 0.6792, |
|
"step": 641 |
|
}, |
|
{ |
|
"epoch": 0.6318897637795275, |
|
"grad_norm": 11.657475471496582, |
|
"learning_rate": 6.332007952286283e-07, |
|
"loss": 1.2363, |
|
"step": 642 |
|
}, |
|
{ |
|
"epoch": 0.6328740157480315, |
|
"grad_norm": 12.533220291137695, |
|
"learning_rate": 6.341948310139166e-07, |
|
"loss": 1.2037, |
|
"step": 643 |
|
}, |
|
{ |
|
"epoch": 0.6338582677165354, |
|
"grad_norm": 10.946784019470215, |
|
"learning_rate": 6.351888667992048e-07, |
|
"loss": 1.105, |
|
"step": 644 |
|
}, |
|
{ |
|
"epoch": 0.6348425196850394, |
|
"grad_norm": 17.624608993530273, |
|
"learning_rate": 6.361829025844931e-07, |
|
"loss": 1.8319, |
|
"step": 645 |
|
}, |
|
{ |
|
"epoch": 0.6358267716535433, |
|
"grad_norm": 13.263932228088379, |
|
"learning_rate": 6.371769383697813e-07, |
|
"loss": 1.4261, |
|
"step": 646 |
|
}, |
|
{ |
|
"epoch": 0.6368110236220472, |
|
"grad_norm": 11.63705062866211, |
|
"learning_rate": 6.381709741550698e-07, |
|
"loss": 1.0431, |
|
"step": 647 |
|
}, |
|
{ |
|
"epoch": 0.6377952755905512, |
|
"grad_norm": 13.618841171264648, |
|
"learning_rate": 6.39165009940358e-07, |
|
"loss": 1.2132, |
|
"step": 648 |
|
}, |
|
{ |
|
"epoch": 0.6387795275590551, |
|
"grad_norm": 12.975966453552246, |
|
"learning_rate": 6.401590457256462e-07, |
|
"loss": 1.1464, |
|
"step": 649 |
|
}, |
|
{ |
|
"epoch": 0.639763779527559, |
|
"grad_norm": 12.637139320373535, |
|
"learning_rate": 6.411530815109346e-07, |
|
"loss": 1.3076, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.640748031496063, |
|
"grad_norm": 12.68572998046875, |
|
"learning_rate": 6.421471172962228e-07, |
|
"loss": 0.9162, |
|
"step": 651 |
|
}, |
|
{ |
|
"epoch": 0.6417322834645669, |
|
"grad_norm": 15.307023048400879, |
|
"learning_rate": 6.431411530815111e-07, |
|
"loss": 1.6739, |
|
"step": 652 |
|
}, |
|
{ |
|
"epoch": 0.6427165354330708, |
|
"grad_norm": 16.65964698791504, |
|
"learning_rate": 6.441351888667994e-07, |
|
"loss": 1.3627, |
|
"step": 653 |
|
}, |
|
{ |
|
"epoch": 0.6437007874015748, |
|
"grad_norm": 10.542638778686523, |
|
"learning_rate": 6.451292246520876e-07, |
|
"loss": 1.084, |
|
"step": 654 |
|
}, |
|
{ |
|
"epoch": 0.6446850393700787, |
|
"grad_norm": 11.227409362792969, |
|
"learning_rate": 6.461232604373759e-07, |
|
"loss": 0.8349, |
|
"step": 655 |
|
}, |
|
{ |
|
"epoch": 0.6456692913385826, |
|
"grad_norm": 14.71562671661377, |
|
"learning_rate": 6.471172962226641e-07, |
|
"loss": 1.5575, |
|
"step": 656 |
|
}, |
|
{ |
|
"epoch": 0.6466535433070866, |
|
"grad_norm": 10.660284042358398, |
|
"learning_rate": 6.481113320079524e-07, |
|
"loss": 0.7503, |
|
"step": 657 |
|
}, |
|
{ |
|
"epoch": 0.6476377952755905, |
|
"grad_norm": 12.722469329833984, |
|
"learning_rate": 6.491053677932407e-07, |
|
"loss": 1.5593, |
|
"step": 658 |
|
}, |
|
{ |
|
"epoch": 0.6486220472440944, |
|
"grad_norm": 13.289987564086914, |
|
"learning_rate": 6.500994035785289e-07, |
|
"loss": 1.1788, |
|
"step": 659 |
|
}, |
|
{ |
|
"epoch": 0.6496062992125984, |
|
"grad_norm": 13.905250549316406, |
|
"learning_rate": 6.510934393638173e-07, |
|
"loss": 1.5119, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.6505905511811023, |
|
"grad_norm": 14.397658348083496, |
|
"learning_rate": 6.520874751491055e-07, |
|
"loss": 1.1793, |
|
"step": 661 |
|
}, |
|
{ |
|
"epoch": 0.6515748031496063, |
|
"grad_norm": 13.769511222839355, |
|
"learning_rate": 6.530815109343937e-07, |
|
"loss": 1.1577, |
|
"step": 662 |
|
}, |
|
{ |
|
"epoch": 0.6525590551181102, |
|
"grad_norm": 17.989856719970703, |
|
"learning_rate": 6.54075546719682e-07, |
|
"loss": 1.4391, |
|
"step": 663 |
|
}, |
|
{ |
|
"epoch": 0.6535433070866141, |
|
"grad_norm": 13.776409149169922, |
|
"learning_rate": 6.550695825049703e-07, |
|
"loss": 1.2291, |
|
"step": 664 |
|
}, |
|
{ |
|
"epoch": 0.6545275590551181, |
|
"grad_norm": 11.71041202545166, |
|
"learning_rate": 6.560636182902586e-07, |
|
"loss": 1.3118, |
|
"step": 665 |
|
}, |
|
{ |
|
"epoch": 0.655511811023622, |
|
"grad_norm": 12.15799331665039, |
|
"learning_rate": 6.570576540755468e-07, |
|
"loss": 1.1358, |
|
"step": 666 |
|
}, |
|
{ |
|
"epoch": 0.656496062992126, |
|
"grad_norm": 15.679516792297363, |
|
"learning_rate": 6.580516898608351e-07, |
|
"loss": 1.814, |
|
"step": 667 |
|
}, |
|
{ |
|
"epoch": 0.65748031496063, |
|
"grad_norm": 13.018841743469238, |
|
"learning_rate": 6.590457256461234e-07, |
|
"loss": 1.038, |
|
"step": 668 |
|
}, |
|
{ |
|
"epoch": 0.6584645669291339, |
|
"grad_norm": 12.75416088104248, |
|
"learning_rate": 6.600397614314116e-07, |
|
"loss": 1.2344, |
|
"step": 669 |
|
}, |
|
{ |
|
"epoch": 0.6594488188976378, |
|
"grad_norm": 12.071971893310547, |
|
"learning_rate": 6.610337972166998e-07, |
|
"loss": 1.1064, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.6604330708661418, |
|
"grad_norm": 13.168946266174316, |
|
"learning_rate": 6.620278330019882e-07, |
|
"loss": 1.8745, |
|
"step": 671 |
|
}, |
|
{ |
|
"epoch": 0.6614173228346457, |
|
"grad_norm": 14.136277198791504, |
|
"learning_rate": 6.630218687872764e-07, |
|
"loss": 1.8775, |
|
"step": 672 |
|
}, |
|
{ |
|
"epoch": 0.6624015748031497, |
|
"grad_norm": 11.940804481506348, |
|
"learning_rate": 6.640159045725647e-07, |
|
"loss": 1.2582, |
|
"step": 673 |
|
}, |
|
{ |
|
"epoch": 0.6633858267716536, |
|
"grad_norm": 11.855066299438477, |
|
"learning_rate": 6.65009940357853e-07, |
|
"loss": 1.3454, |
|
"step": 674 |
|
}, |
|
{ |
|
"epoch": 0.6643700787401575, |
|
"grad_norm": 15.009392738342285, |
|
"learning_rate": 6.660039761431412e-07, |
|
"loss": 1.2374, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.6653543307086615, |
|
"grad_norm": 9.056105613708496, |
|
"learning_rate": 6.669980119284295e-07, |
|
"loss": 0.7227, |
|
"step": 676 |
|
}, |
|
{ |
|
"epoch": 0.6663385826771654, |
|
"grad_norm": 11.413203239440918, |
|
"learning_rate": 6.679920477137177e-07, |
|
"loss": 1.2036, |
|
"step": 677 |
|
}, |
|
{ |
|
"epoch": 0.6673228346456693, |
|
"grad_norm": 12.716734886169434, |
|
"learning_rate": 6.689860834990062e-07, |
|
"loss": 1.5248, |
|
"step": 678 |
|
}, |
|
{ |
|
"epoch": 0.6683070866141733, |
|
"grad_norm": 11.400150299072266, |
|
"learning_rate": 6.699801192842944e-07, |
|
"loss": 1.0584, |
|
"step": 679 |
|
}, |
|
{ |
|
"epoch": 0.6692913385826772, |
|
"grad_norm": 10.649139404296875, |
|
"learning_rate": 6.709741550695826e-07, |
|
"loss": 0.7455, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.6702755905511811, |
|
"grad_norm": 11.595582962036133, |
|
"learning_rate": 6.71968190854871e-07, |
|
"loss": 1.0616, |
|
"step": 681 |
|
}, |
|
{ |
|
"epoch": 0.6712598425196851, |
|
"grad_norm": 13.45544147491455, |
|
"learning_rate": 6.729622266401592e-07, |
|
"loss": 1.1377, |
|
"step": 682 |
|
}, |
|
{ |
|
"epoch": 0.672244094488189, |
|
"grad_norm": 13.17196273803711, |
|
"learning_rate": 6.739562624254474e-07, |
|
"loss": 1.7022, |
|
"step": 683 |
|
}, |
|
{ |
|
"epoch": 0.6732283464566929, |
|
"grad_norm": 14.776825904846191, |
|
"learning_rate": 6.749502982107357e-07, |
|
"loss": 1.4008, |
|
"step": 684 |
|
}, |
|
{ |
|
"epoch": 0.6742125984251969, |
|
"grad_norm": 15.56824016571045, |
|
"learning_rate": 6.75944333996024e-07, |
|
"loss": 1.4894, |
|
"step": 685 |
|
}, |
|
{ |
|
"epoch": 0.6751968503937008, |
|
"grad_norm": 14.228174209594727, |
|
"learning_rate": 6.769383697813123e-07, |
|
"loss": 1.5464, |
|
"step": 686 |
|
}, |
|
{ |
|
"epoch": 0.6761811023622047, |
|
"grad_norm": 15.94443130493164, |
|
"learning_rate": 6.779324055666005e-07, |
|
"loss": 1.5018, |
|
"step": 687 |
|
}, |
|
{ |
|
"epoch": 0.6771653543307087, |
|
"grad_norm": 20.589385986328125, |
|
"learning_rate": 6.789264413518888e-07, |
|
"loss": 1.5803, |
|
"step": 688 |
|
}, |
|
{ |
|
"epoch": 0.6781496062992126, |
|
"grad_norm": 18.15032386779785, |
|
"learning_rate": 6.799204771371771e-07, |
|
"loss": 1.5654, |
|
"step": 689 |
|
}, |
|
{ |
|
"epoch": 0.6791338582677166, |
|
"grad_norm": 13.775176048278809, |
|
"learning_rate": 6.809145129224653e-07, |
|
"loss": 1.2537, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.6801181102362205, |
|
"grad_norm": 15.423773765563965, |
|
"learning_rate": 6.819085487077536e-07, |
|
"loss": 1.4901, |
|
"step": 691 |
|
}, |
|
{ |
|
"epoch": 0.6811023622047244, |
|
"grad_norm": 11.028498649597168, |
|
"learning_rate": 6.829025844930419e-07, |
|
"loss": 0.9597, |
|
"step": 692 |
|
}, |
|
{ |
|
"epoch": 0.6820866141732284, |
|
"grad_norm": 14.687771797180176, |
|
"learning_rate": 6.838966202783301e-07, |
|
"loss": 1.5515, |
|
"step": 693 |
|
}, |
|
{ |
|
"epoch": 0.6830708661417323, |
|
"grad_norm": 11.73508071899414, |
|
"learning_rate": 6.848906560636184e-07, |
|
"loss": 1.1314, |
|
"step": 694 |
|
}, |
|
{ |
|
"epoch": 0.6840551181102362, |
|
"grad_norm": 11.076573371887207, |
|
"learning_rate": 6.858846918489067e-07, |
|
"loss": 1.2442, |
|
"step": 695 |
|
}, |
|
{ |
|
"epoch": 0.6850393700787402, |
|
"grad_norm": 14.134485244750977, |
|
"learning_rate": 6.86878727634195e-07, |
|
"loss": 1.5542, |
|
"step": 696 |
|
}, |
|
{ |
|
"epoch": 0.6860236220472441, |
|
"grad_norm": 9.774703979492188, |
|
"learning_rate": 6.878727634194832e-07, |
|
"loss": 0.7617, |
|
"step": 697 |
|
}, |
|
{ |
|
"epoch": 0.687007874015748, |
|
"grad_norm": 14.868870735168457, |
|
"learning_rate": 6.888667992047714e-07, |
|
"loss": 1.4626, |
|
"step": 698 |
|
}, |
|
{ |
|
"epoch": 0.687992125984252, |
|
"grad_norm": 11.818559646606445, |
|
"learning_rate": 6.898608349900598e-07, |
|
"loss": 0.6546, |
|
"step": 699 |
|
}, |
|
{ |
|
"epoch": 0.6889763779527559, |
|
"grad_norm": 13.716598510742188, |
|
"learning_rate": 6.90854870775348e-07, |
|
"loss": 1.1726, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.6899606299212598, |
|
"grad_norm": 10.127545356750488, |
|
"learning_rate": 6.918489065606362e-07, |
|
"loss": 0.7522, |
|
"step": 701 |
|
}, |
|
{ |
|
"epoch": 0.6909448818897638, |
|
"grad_norm": 14.03452205657959, |
|
"learning_rate": 6.928429423459246e-07, |
|
"loss": 1.6203, |
|
"step": 702 |
|
}, |
|
{ |
|
"epoch": 0.6919291338582677, |
|
"grad_norm": 11.538923263549805, |
|
"learning_rate": 6.938369781312128e-07, |
|
"loss": 1.0822, |
|
"step": 703 |
|
}, |
|
{ |
|
"epoch": 0.6929133858267716, |
|
"grad_norm": 11.976446151733398, |
|
"learning_rate": 6.948310139165011e-07, |
|
"loss": 0.9652, |
|
"step": 704 |
|
}, |
|
{ |
|
"epoch": 0.6938976377952756, |
|
"grad_norm": 11.482376098632812, |
|
"learning_rate": 6.958250497017893e-07, |
|
"loss": 0.9884, |
|
"step": 705 |
|
}, |
|
{ |
|
"epoch": 0.6948818897637795, |
|
"grad_norm": 12.661773681640625, |
|
"learning_rate": 6.968190854870776e-07, |
|
"loss": 1.3876, |
|
"step": 706 |
|
}, |
|
{ |
|
"epoch": 0.6958661417322834, |
|
"grad_norm": 11.961230278015137, |
|
"learning_rate": 6.97813121272366e-07, |
|
"loss": 1.1726, |
|
"step": 707 |
|
}, |
|
{ |
|
"epoch": 0.6968503937007874, |
|
"grad_norm": 11.375978469848633, |
|
"learning_rate": 6.988071570576542e-07, |
|
"loss": 1.1426, |
|
"step": 708 |
|
}, |
|
{ |
|
"epoch": 0.6978346456692913, |
|
"grad_norm": 13.53009033203125, |
|
"learning_rate": 6.998011928429426e-07, |
|
"loss": 1.2303, |
|
"step": 709 |
|
}, |
|
{ |
|
"epoch": 0.6988188976377953, |
|
"grad_norm": 12.14291763305664, |
|
"learning_rate": 7.007952286282308e-07, |
|
"loss": 1.3438, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.6998031496062992, |
|
"grad_norm": 13.431678771972656, |
|
"learning_rate": 7.01789264413519e-07, |
|
"loss": 1.3823, |
|
"step": 711 |
|
}, |
|
{ |
|
"epoch": 0.7007874015748031, |
|
"grad_norm": 14.398938179016113, |
|
"learning_rate": 7.027833001988073e-07, |
|
"loss": 1.5166, |
|
"step": 712 |
|
}, |
|
{ |
|
"epoch": 0.7017716535433071, |
|
"grad_norm": 12.663697242736816, |
|
"learning_rate": 7.037773359840956e-07, |
|
"loss": 1.4148, |
|
"step": 713 |
|
}, |
|
{ |
|
"epoch": 0.702755905511811, |
|
"grad_norm": 11.511486053466797, |
|
"learning_rate": 7.047713717693838e-07, |
|
"loss": 0.9564, |
|
"step": 714 |
|
}, |
|
{ |
|
"epoch": 0.7037401574803149, |
|
"grad_norm": 14.274373054504395, |
|
"learning_rate": 7.057654075546721e-07, |
|
"loss": 1.1031, |
|
"step": 715 |
|
}, |
|
{ |
|
"epoch": 0.7047244094488189, |
|
"grad_norm": 22.680253982543945, |
|
"learning_rate": 7.067594433399604e-07, |
|
"loss": 1.6628, |
|
"step": 716 |
|
}, |
|
{ |
|
"epoch": 0.7057086614173228, |
|
"grad_norm": 14.918832778930664, |
|
"learning_rate": 7.077534791252487e-07, |
|
"loss": 1.3693, |
|
"step": 717 |
|
}, |
|
{ |
|
"epoch": 0.7066929133858267, |
|
"grad_norm": 9.6978120803833, |
|
"learning_rate": 7.087475149105369e-07, |
|
"loss": 1.0247, |
|
"step": 718 |
|
}, |
|
{ |
|
"epoch": 0.7076771653543307, |
|
"grad_norm": 10.600605010986328, |
|
"learning_rate": 7.097415506958251e-07, |
|
"loss": 0.7778, |
|
"step": 719 |
|
}, |
|
{ |
|
"epoch": 0.7086614173228346, |
|
"grad_norm": 13.40506649017334, |
|
"learning_rate": 7.107355864811135e-07, |
|
"loss": 1.3158, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.7096456692913385, |
|
"grad_norm": 14.183016777038574, |
|
"learning_rate": 7.117296222664017e-07, |
|
"loss": 1.5329, |
|
"step": 721 |
|
}, |
|
{ |
|
"epoch": 0.7106299212598425, |
|
"grad_norm": 12.718838691711426, |
|
"learning_rate": 7.1272365805169e-07, |
|
"loss": 1.2845, |
|
"step": 722 |
|
}, |
|
{ |
|
"epoch": 0.7116141732283464, |
|
"grad_norm": 10.078266143798828, |
|
"learning_rate": 7.137176938369783e-07, |
|
"loss": 0.7895, |
|
"step": 723 |
|
}, |
|
{ |
|
"epoch": 0.7125984251968503, |
|
"grad_norm": 9.505541801452637, |
|
"learning_rate": 7.147117296222665e-07, |
|
"loss": 0.9098, |
|
"step": 724 |
|
}, |
|
{ |
|
"epoch": 0.7135826771653543, |
|
"grad_norm": 9.941719055175781, |
|
"learning_rate": 7.157057654075548e-07, |
|
"loss": 0.7648, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.7145669291338582, |
|
"grad_norm": 13.63866138458252, |
|
"learning_rate": 7.16699801192843e-07, |
|
"loss": 1.0951, |
|
"step": 726 |
|
}, |
|
{ |
|
"epoch": 0.7155511811023622, |
|
"grad_norm": 11.125609397888184, |
|
"learning_rate": 7.176938369781313e-07, |
|
"loss": 1.1351, |
|
"step": 727 |
|
}, |
|
{ |
|
"epoch": 0.7165354330708661, |
|
"grad_norm": 13.634588241577148, |
|
"learning_rate": 7.186878727634196e-07, |
|
"loss": 1.0514, |
|
"step": 728 |
|
}, |
|
{ |
|
"epoch": 0.71751968503937, |
|
"grad_norm": 12.125421524047852, |
|
"learning_rate": 7.196819085487078e-07, |
|
"loss": 1.2328, |
|
"step": 729 |
|
}, |
|
{ |
|
"epoch": 0.718503937007874, |
|
"grad_norm": 10.89931583404541, |
|
"learning_rate": 7.206759443339962e-07, |
|
"loss": 1.0664, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.719488188976378, |
|
"grad_norm": 12.2229585647583, |
|
"learning_rate": 7.216699801192844e-07, |
|
"loss": 1.1058, |
|
"step": 731 |
|
}, |
|
{ |
|
"epoch": 0.7204724409448819, |
|
"grad_norm": 7.885047912597656, |
|
"learning_rate": 7.226640159045726e-07, |
|
"loss": 0.6698, |
|
"step": 732 |
|
}, |
|
{ |
|
"epoch": 0.7214566929133859, |
|
"grad_norm": 13.67959976196289, |
|
"learning_rate": 7.236580516898609e-07, |
|
"loss": 1.1438, |
|
"step": 733 |
|
}, |
|
{ |
|
"epoch": 0.7224409448818898, |
|
"grad_norm": 13.96220874786377, |
|
"learning_rate": 7.246520874751492e-07, |
|
"loss": 1.0473, |
|
"step": 734 |
|
}, |
|
{ |
|
"epoch": 0.7234251968503937, |
|
"grad_norm": 17.09735679626465, |
|
"learning_rate": 7.256461232604375e-07, |
|
"loss": 1.5307, |
|
"step": 735 |
|
}, |
|
{ |
|
"epoch": 0.7244094488188977, |
|
"grad_norm": 13.294286727905273, |
|
"learning_rate": 7.266401590457257e-07, |
|
"loss": 1.2537, |
|
"step": 736 |
|
}, |
|
{ |
|
"epoch": 0.7253937007874016, |
|
"grad_norm": 9.892070770263672, |
|
"learning_rate": 7.27634194831014e-07, |
|
"loss": 0.7006, |
|
"step": 737 |
|
}, |
|
{ |
|
"epoch": 0.7263779527559056, |
|
"grad_norm": 11.488076210021973, |
|
"learning_rate": 7.286282306163024e-07, |
|
"loss": 0.8487, |
|
"step": 738 |
|
}, |
|
{ |
|
"epoch": 0.7273622047244095, |
|
"grad_norm": 12.460606575012207, |
|
"learning_rate": 7.296222664015906e-07, |
|
"loss": 1.193, |
|
"step": 739 |
|
}, |
|
{ |
|
"epoch": 0.7283464566929134, |
|
"grad_norm": 12.237475395202637, |
|
"learning_rate": 7.306163021868788e-07, |
|
"loss": 0.9711, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.7293307086614174, |
|
"grad_norm": 8.606354713439941, |
|
"learning_rate": 7.316103379721672e-07, |
|
"loss": 0.5139, |
|
"step": 741 |
|
}, |
|
{ |
|
"epoch": 0.7303149606299213, |
|
"grad_norm": 14.769782066345215, |
|
"learning_rate": 7.326043737574554e-07, |
|
"loss": 1.1185, |
|
"step": 742 |
|
}, |
|
{ |
|
"epoch": 0.7312992125984252, |
|
"grad_norm": 11.785113334655762, |
|
"learning_rate": 7.335984095427437e-07, |
|
"loss": 0.9982, |
|
"step": 743 |
|
}, |
|
{ |
|
"epoch": 0.7322834645669292, |
|
"grad_norm": 10.399678230285645, |
|
"learning_rate": 7.34592445328032e-07, |
|
"loss": 0.7987, |
|
"step": 744 |
|
}, |
|
{ |
|
"epoch": 0.7332677165354331, |
|
"grad_norm": 14.253144264221191, |
|
"learning_rate": 7.355864811133202e-07, |
|
"loss": 1.161, |
|
"step": 745 |
|
}, |
|
{ |
|
"epoch": 0.734251968503937, |
|
"grad_norm": 12.234946250915527, |
|
"learning_rate": 7.365805168986085e-07, |
|
"loss": 0.9832, |
|
"step": 746 |
|
}, |
|
{ |
|
"epoch": 0.735236220472441, |
|
"grad_norm": 17.171052932739258, |
|
"learning_rate": 7.375745526838967e-07, |
|
"loss": 1.6041, |
|
"step": 747 |
|
}, |
|
{ |
|
"epoch": 0.7362204724409449, |
|
"grad_norm": 12.20317268371582, |
|
"learning_rate": 7.385685884691851e-07, |
|
"loss": 1.1084, |
|
"step": 748 |
|
}, |
|
{ |
|
"epoch": 0.7372047244094488, |
|
"grad_norm": 13.474609375, |
|
"learning_rate": 7.395626242544733e-07, |
|
"loss": 1.1264, |
|
"step": 749 |
|
}, |
|
{ |
|
"epoch": 0.7381889763779528, |
|
"grad_norm": 10.15322494506836, |
|
"learning_rate": 7.405566600397615e-07, |
|
"loss": 0.6759, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.7391732283464567, |
|
"grad_norm": 13.045278549194336, |
|
"learning_rate": 7.415506958250499e-07, |
|
"loss": 1.5241, |
|
"step": 751 |
|
}, |
|
{ |
|
"epoch": 0.7401574803149606, |
|
"grad_norm": 14.404647827148438, |
|
"learning_rate": 7.425447316103381e-07, |
|
"loss": 1.6181, |
|
"step": 752 |
|
}, |
|
{ |
|
"epoch": 0.7411417322834646, |
|
"grad_norm": 10.859138488769531, |
|
"learning_rate": 7.435387673956263e-07, |
|
"loss": 1.1862, |
|
"step": 753 |
|
}, |
|
{ |
|
"epoch": 0.7421259842519685, |
|
"grad_norm": 12.503321647644043, |
|
"learning_rate": 7.445328031809146e-07, |
|
"loss": 1.0147, |
|
"step": 754 |
|
}, |
|
{ |
|
"epoch": 0.7431102362204725, |
|
"grad_norm": 12.434426307678223, |
|
"learning_rate": 7.455268389662029e-07, |
|
"loss": 1.0435, |
|
"step": 755 |
|
}, |
|
{ |
|
"epoch": 0.7440944881889764, |
|
"grad_norm": 12.646941184997559, |
|
"learning_rate": 7.465208747514912e-07, |
|
"loss": 0.9175, |
|
"step": 756 |
|
}, |
|
{ |
|
"epoch": 0.7450787401574803, |
|
"grad_norm": 14.548798561096191, |
|
"learning_rate": 7.475149105367794e-07, |
|
"loss": 1.5608, |
|
"step": 757 |
|
}, |
|
{ |
|
"epoch": 0.7460629921259843, |
|
"grad_norm": 15.520113945007324, |
|
"learning_rate": 7.485089463220677e-07, |
|
"loss": 1.7027, |
|
"step": 758 |
|
}, |
|
{ |
|
"epoch": 0.7470472440944882, |
|
"grad_norm": 11.138642311096191, |
|
"learning_rate": 7.49502982107356e-07, |
|
"loss": 0.9282, |
|
"step": 759 |
|
}, |
|
{ |
|
"epoch": 0.7480314960629921, |
|
"grad_norm": 10.51353931427002, |
|
"learning_rate": 7.504970178926442e-07, |
|
"loss": 0.8072, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.7490157480314961, |
|
"grad_norm": 15.635726928710938, |
|
"learning_rate": 7.514910536779325e-07, |
|
"loss": 1.13, |
|
"step": 761 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 14.228267669677734, |
|
"learning_rate": 7.524850894632208e-07, |
|
"loss": 1.153, |
|
"step": 762 |
|
}, |
|
{ |
|
"epoch": 0.7509842519685039, |
|
"grad_norm": 18.568565368652344, |
|
"learning_rate": 7.53479125248509e-07, |
|
"loss": 1.7723, |
|
"step": 763 |
|
}, |
|
{ |
|
"epoch": 0.7519685039370079, |
|
"grad_norm": 11.066574096679688, |
|
"learning_rate": 7.544731610337973e-07, |
|
"loss": 0.9833, |
|
"step": 764 |
|
}, |
|
{ |
|
"epoch": 0.7529527559055118, |
|
"grad_norm": 15.546330451965332, |
|
"learning_rate": 7.554671968190856e-07, |
|
"loss": 1.1766, |
|
"step": 765 |
|
}, |
|
{ |
|
"epoch": 0.7529527559055118, |
|
"eval_Qnli-dev_cosine_accuracy": 0.650390625, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.8711802363395691, |
|
"eval_Qnli-dev_cosine_ap": 0.6636184479177414, |
|
"eval_Qnli-dev_cosine_f1": 0.6573208722741433, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.779100775718689, |
|
"eval_Qnli-dev_cosine_precision": 0.5197044334975369, |
|
"eval_Qnli-dev_cosine_recall": 0.8940677966101694, |
|
"eval_Qnli-dev_dot_accuracy": 0.650390625, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 664.5684814453125, |
|
"eval_Qnli-dev_dot_ap": 0.6615631968512774, |
|
"eval_Qnli-dev_dot_f1": 0.6568047337278107, |
|
"eval_Qnli-dev_dot_f1_threshold": 582.806396484375, |
|
"eval_Qnli-dev_dot_precision": 0.5045454545454545, |
|
"eval_Qnli-dev_dot_recall": 0.940677966101695, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.650390625, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 14.019062995910645, |
|
"eval_Qnli-dev_euclidean_ap": 0.663466792452131, |
|
"eval_Qnli-dev_euclidean_f1": 0.6573208722741433, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 18.369121551513672, |
|
"eval_Qnli-dev_euclidean_precision": 0.5197044334975369, |
|
"eval_Qnli-dev_euclidean_recall": 0.8940677966101694, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.65625, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 303.57794189453125, |
|
"eval_Qnli-dev_manhattan_ap": 0.6740758222672537, |
|
"eval_Qnli-dev_manhattan_f1": 0.6689655172413793, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 349.1427001953125, |
|
"eval_Qnli-dev_manhattan_precision": 0.563953488372093, |
|
"eval_Qnli-dev_manhattan_recall": 0.8220338983050848, |
|
"eval_Qnli-dev_max_accuracy": 0.65625, |
|
"eval_Qnli-dev_max_accuracy_threshold": 664.5684814453125, |
|
"eval_Qnli-dev_max_ap": 0.6740758222672537, |
|
"eval_Qnli-dev_max_f1": 0.6689655172413793, |
|
"eval_Qnli-dev_max_f1_threshold": 582.806396484375, |
|
"eval_Qnli-dev_max_precision": 0.563953488372093, |
|
"eval_Qnli-dev_max_recall": 0.940677966101695, |
|
"eval_allNLI-dev_cosine_accuracy": 0.677734375, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9697319269180298, |
|
"eval_allNLI-dev_cosine_ap": 0.47097839371304523, |
|
"eval_allNLI-dev_cosine_f1": 0.5189003436426117, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.8454164862632751, |
|
"eval_allNLI-dev_cosine_precision": 0.3691931540342298, |
|
"eval_allNLI-dev_cosine_recall": 0.8728323699421965, |
|
"eval_allNLI-dev_dot_accuracy": 0.677734375, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 737.8129272460938, |
|
"eval_allNLI-dev_dot_ap": 0.4710099600965258, |
|
"eval_allNLI-dev_dot_f1": 0.5189003436426117, |
|
"eval_allNLI-dev_dot_f1_threshold": 647.1653442382812, |
|
"eval_allNLI-dev_dot_precision": 0.3691931540342298, |
|
"eval_allNLI-dev_dot_recall": 0.8728323699421965, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.677734375, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 6.808244228363037, |
|
"eval_allNLI-dev_euclidean_ap": 0.47125375440506867, |
|
"eval_allNLI-dev_euclidean_f1": 0.5189003436426117, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 15.383974075317383, |
|
"eval_allNLI-dev_euclidean_precision": 0.3691931540342298, |
|
"eval_allNLI-dev_euclidean_recall": 0.8728323699421965, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.677734375, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 182.8551788330078, |
|
"eval_allNLI-dev_manhattan_ap": 0.4685839203620036, |
|
"eval_allNLI-dev_manhattan_f1": 0.5330296127562643, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 261.1466979980469, |
|
"eval_allNLI-dev_manhattan_precision": 0.4398496240601504, |
|
"eval_allNLI-dev_manhattan_recall": 0.6763005780346821, |
|
"eval_allNLI-dev_max_accuracy": 0.677734375, |
|
"eval_allNLI-dev_max_accuracy_threshold": 737.8129272460938, |
|
"eval_allNLI-dev_max_ap": 0.47125375440506867, |
|
"eval_allNLI-dev_max_f1": 0.5330296127562643, |
|
"eval_allNLI-dev_max_f1_threshold": 647.1653442382812, |
|
"eval_allNLI-dev_max_precision": 0.4398496240601504, |
|
"eval_allNLI-dev_max_recall": 0.8728323699421965, |
|
"eval_loss": 1.3821483850479126, |
|
"eval_runtime": 52.6629, |
|
"eval_samples_per_second": 31.597, |
|
"eval_sequential_score": 0.6740758222672537, |
|
"eval_steps_per_second": 0.133, |
|
"eval_sts-test_pearson_cosine": 0.41074334497722165, |
|
"eval_sts-test_pearson_dot": 0.407552380755139, |
|
"eval_sts-test_pearson_euclidean": 0.4333967868726292, |
|
"eval_sts-test_pearson_manhattan": 0.44332181906543383, |
|
"eval_sts-test_pearson_max": 0.44332181906543383, |
|
"eval_sts-test_spearman_cosine": 0.4583378890486231, |
|
"eval_sts-test_spearman_dot": 0.4553192375302001, |
|
"eval_sts-test_spearman_euclidean": 0.458448737920612, |
|
"eval_sts-test_spearman_manhattan": 0.4653892556459323, |
|
"eval_sts-test_spearman_max": 0.4653892556459323, |
|
"step": 765 |
|
}, |
|
{ |
|
"epoch": 0.7539370078740157, |
|
"grad_norm": 16.530473709106445, |
|
"learning_rate": 7.56461232604374e-07, |
|
"loss": 1.2351, |
|
"step": 766 |
|
}, |
|
{ |
|
"epoch": 0.7549212598425197, |
|
"grad_norm": 14.440428733825684, |
|
"learning_rate": 7.574552683896621e-07, |
|
"loss": 0.9521, |
|
"step": 767 |
|
}, |
|
{ |
|
"epoch": 0.7559055118110236, |
|
"grad_norm": 12.20114517211914, |
|
"learning_rate": 7.584493041749503e-07, |
|
"loss": 1.1981, |
|
"step": 768 |
|
}, |
|
{ |
|
"epoch": 0.7568897637795275, |
|
"grad_norm": 11.787071228027344, |
|
"learning_rate": 7.594433399602388e-07, |
|
"loss": 0.8031, |
|
"step": 769 |
|
}, |
|
{ |
|
"epoch": 0.7578740157480315, |
|
"grad_norm": 12.98501205444336, |
|
"learning_rate": 7.60437375745527e-07, |
|
"loss": 1.2502, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.7588582677165354, |
|
"grad_norm": 10.086615562438965, |
|
"learning_rate": 7.614314115308152e-07, |
|
"loss": 0.9313, |
|
"step": 771 |
|
}, |
|
{ |
|
"epoch": 0.7598425196850394, |
|
"grad_norm": 9.566267013549805, |
|
"learning_rate": 7.624254473161036e-07, |
|
"loss": 0.8738, |
|
"step": 772 |
|
}, |
|
{ |
|
"epoch": 0.7608267716535433, |
|
"grad_norm": 11.557473182678223, |
|
"learning_rate": 7.634194831013918e-07, |
|
"loss": 1.0316, |
|
"step": 773 |
|
}, |
|
{ |
|
"epoch": 0.7618110236220472, |
|
"grad_norm": 10.591485023498535, |
|
"learning_rate": 7.644135188866801e-07, |
|
"loss": 1.0679, |
|
"step": 774 |
|
}, |
|
{ |
|
"epoch": 0.7627952755905512, |
|
"grad_norm": 12.123056411743164, |
|
"learning_rate": 7.654075546719683e-07, |
|
"loss": 1.0308, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 0.7637795275590551, |
|
"grad_norm": 11.250408172607422, |
|
"learning_rate": 7.664015904572566e-07, |
|
"loss": 1.0663, |
|
"step": 776 |
|
}, |
|
{ |
|
"epoch": 0.764763779527559, |
|
"grad_norm": 13.228967666625977, |
|
"learning_rate": 7.673956262425449e-07, |
|
"loss": 1.3565, |
|
"step": 777 |
|
}, |
|
{ |
|
"epoch": 0.765748031496063, |
|
"grad_norm": 9.849665641784668, |
|
"learning_rate": 7.683896620278331e-07, |
|
"loss": 0.9103, |
|
"step": 778 |
|
}, |
|
{ |
|
"epoch": 0.7667322834645669, |
|
"grad_norm": 14.536798477172852, |
|
"learning_rate": 7.693836978131215e-07, |
|
"loss": 1.477, |
|
"step": 779 |
|
}, |
|
{ |
|
"epoch": 0.7677165354330708, |
|
"grad_norm": 10.974257469177246, |
|
"learning_rate": 7.703777335984097e-07, |
|
"loss": 1.2526, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 0.7687007874015748, |
|
"grad_norm": 9.089542388916016, |
|
"learning_rate": 7.713717693836979e-07, |
|
"loss": 0.8647, |
|
"step": 781 |
|
}, |
|
{ |
|
"epoch": 0.7696850393700787, |
|
"grad_norm": 11.030006408691406, |
|
"learning_rate": 7.723658051689862e-07, |
|
"loss": 0.8498, |
|
"step": 782 |
|
}, |
|
{ |
|
"epoch": 0.7706692913385826, |
|
"grad_norm": 12.539122581481934, |
|
"learning_rate": 7.733598409542745e-07, |
|
"loss": 1.5257, |
|
"step": 783 |
|
}, |
|
{ |
|
"epoch": 0.7716535433070866, |
|
"grad_norm": 8.903121948242188, |
|
"learning_rate": 7.743538767395627e-07, |
|
"loss": 0.5493, |
|
"step": 784 |
|
}, |
|
{ |
|
"epoch": 0.7726377952755905, |
|
"grad_norm": 12.023555755615234, |
|
"learning_rate": 7.75347912524851e-07, |
|
"loss": 1.2321, |
|
"step": 785 |
|
}, |
|
{ |
|
"epoch": 0.7736220472440944, |
|
"grad_norm": 14.207155227661133, |
|
"learning_rate": 7.763419483101393e-07, |
|
"loss": 1.4306, |
|
"step": 786 |
|
}, |
|
{ |
|
"epoch": 0.7746062992125984, |
|
"grad_norm": 12.97458553314209, |
|
"learning_rate": 7.773359840954276e-07, |
|
"loss": 1.0428, |
|
"step": 787 |
|
}, |
|
{ |
|
"epoch": 0.7755905511811023, |
|
"grad_norm": 14.624722480773926, |
|
"learning_rate": 7.783300198807158e-07, |
|
"loss": 1.3553, |
|
"step": 788 |
|
}, |
|
{ |
|
"epoch": 0.7765748031496063, |
|
"grad_norm": 11.624364852905273, |
|
"learning_rate": 7.79324055666004e-07, |
|
"loss": 1.1102, |
|
"step": 789 |
|
}, |
|
{ |
|
"epoch": 0.7775590551181102, |
|
"grad_norm": 11.759018898010254, |
|
"learning_rate": 7.803180914512924e-07, |
|
"loss": 0.9225, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 0.7785433070866141, |
|
"grad_norm": 11.817508697509766, |
|
"learning_rate": 7.813121272365806e-07, |
|
"loss": 1.2585, |
|
"step": 791 |
|
}, |
|
{ |
|
"epoch": 0.7795275590551181, |
|
"grad_norm": 15.326850891113281, |
|
"learning_rate": 7.823061630218689e-07, |
|
"loss": 1.1335, |
|
"step": 792 |
|
}, |
|
{ |
|
"epoch": 0.780511811023622, |
|
"grad_norm": 14.029367446899414, |
|
"learning_rate": 7.833001988071572e-07, |
|
"loss": 1.2781, |
|
"step": 793 |
|
}, |
|
{ |
|
"epoch": 0.781496062992126, |
|
"grad_norm": 15.042057037353516, |
|
"learning_rate": 7.842942345924454e-07, |
|
"loss": 1.0265, |
|
"step": 794 |
|
}, |
|
{ |
|
"epoch": 0.78248031496063, |
|
"grad_norm": 13.87203311920166, |
|
"learning_rate": 7.852882703777337e-07, |
|
"loss": 1.1846, |
|
"step": 795 |
|
}, |
|
{ |
|
"epoch": 0.7834645669291339, |
|
"grad_norm": 16.487451553344727, |
|
"learning_rate": 7.862823061630219e-07, |
|
"loss": 1.2123, |
|
"step": 796 |
|
}, |
|
{ |
|
"epoch": 0.7844488188976378, |
|
"grad_norm": 11.526494026184082, |
|
"learning_rate": 7.872763419483102e-07, |
|
"loss": 1.0861, |
|
"step": 797 |
|
}, |
|
{ |
|
"epoch": 0.7854330708661418, |
|
"grad_norm": 12.820661544799805, |
|
"learning_rate": 7.882703777335986e-07, |
|
"loss": 1.1929, |
|
"step": 798 |
|
}, |
|
{ |
|
"epoch": 0.7864173228346457, |
|
"grad_norm": 8.670584678649902, |
|
"learning_rate": 7.892644135188868e-07, |
|
"loss": 0.8202, |
|
"step": 799 |
|
}, |
|
{ |
|
"epoch": 0.7874015748031497, |
|
"grad_norm": 14.31804084777832, |
|
"learning_rate": 7.902584493041752e-07, |
|
"loss": 1.0745, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.7883858267716536, |
|
"grad_norm": 15.200295448303223, |
|
"learning_rate": 7.912524850894634e-07, |
|
"loss": 0.9162, |
|
"step": 801 |
|
}, |
|
{ |
|
"epoch": 0.7893700787401575, |
|
"grad_norm": 14.89583969116211, |
|
"learning_rate": 7.922465208747516e-07, |
|
"loss": 1.2077, |
|
"step": 802 |
|
}, |
|
{ |
|
"epoch": 0.7903543307086615, |
|
"grad_norm": 17.461294174194336, |
|
"learning_rate": 7.932405566600399e-07, |
|
"loss": 1.432, |
|
"step": 803 |
|
}, |
|
{ |
|
"epoch": 0.7913385826771654, |
|
"grad_norm": 13.266355514526367, |
|
"learning_rate": 7.942345924453282e-07, |
|
"loss": 0.9384, |
|
"step": 804 |
|
}, |
|
{ |
|
"epoch": 0.7923228346456693, |
|
"grad_norm": 13.99533748626709, |
|
"learning_rate": 7.952286282306165e-07, |
|
"loss": 1.4686, |
|
"step": 805 |
|
}, |
|
{ |
|
"epoch": 0.7933070866141733, |
|
"grad_norm": 11.035348892211914, |
|
"learning_rate": 7.962226640159047e-07, |
|
"loss": 0.8565, |
|
"step": 806 |
|
}, |
|
{ |
|
"epoch": 0.7942913385826772, |
|
"grad_norm": 14.682604789733887, |
|
"learning_rate": 7.97216699801193e-07, |
|
"loss": 1.1208, |
|
"step": 807 |
|
}, |
|
{ |
|
"epoch": 0.7952755905511811, |
|
"grad_norm": 11.655988693237305, |
|
"learning_rate": 7.982107355864813e-07, |
|
"loss": 1.006, |
|
"step": 808 |
|
}, |
|
{ |
|
"epoch": 0.7962598425196851, |
|
"grad_norm": 10.584676742553711, |
|
"learning_rate": 7.992047713717695e-07, |
|
"loss": 0.71, |
|
"step": 809 |
|
}, |
|
{ |
|
"epoch": 0.797244094488189, |
|
"grad_norm": 11.906569480895996, |
|
"learning_rate": 8.001988071570577e-07, |
|
"loss": 1.0219, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 0.7982283464566929, |
|
"grad_norm": 12.940180778503418, |
|
"learning_rate": 8.011928429423461e-07, |
|
"loss": 1.0836, |
|
"step": 811 |
|
}, |
|
{ |
|
"epoch": 0.7992125984251969, |
|
"grad_norm": 14.408658027648926, |
|
"learning_rate": 8.021868787276343e-07, |
|
"loss": 1.1318, |
|
"step": 812 |
|
}, |
|
{ |
|
"epoch": 0.8001968503937008, |
|
"grad_norm": 15.621986389160156, |
|
"learning_rate": 8.031809145129226e-07, |
|
"loss": 1.5411, |
|
"step": 813 |
|
}, |
|
{ |
|
"epoch": 0.8011811023622047, |
|
"grad_norm": 11.737788200378418, |
|
"learning_rate": 8.041749502982109e-07, |
|
"loss": 1.0063, |
|
"step": 814 |
|
}, |
|
{ |
|
"epoch": 0.8021653543307087, |
|
"grad_norm": 13.536646842956543, |
|
"learning_rate": 8.051689860834991e-07, |
|
"loss": 1.277, |
|
"step": 815 |
|
}, |
|
{ |
|
"epoch": 0.8031496062992126, |
|
"grad_norm": 12.623762130737305, |
|
"learning_rate": 8.061630218687874e-07, |
|
"loss": 1.0547, |
|
"step": 816 |
|
}, |
|
{ |
|
"epoch": 0.8041338582677166, |
|
"grad_norm": 10.43212890625, |
|
"learning_rate": 8.071570576540756e-07, |
|
"loss": 0.7084, |
|
"step": 817 |
|
}, |
|
{ |
|
"epoch": 0.8051181102362205, |
|
"grad_norm": 12.9672269821167, |
|
"learning_rate": 8.08151093439364e-07, |
|
"loss": 1.2581, |
|
"step": 818 |
|
}, |
|
{ |
|
"epoch": 0.8061023622047244, |
|
"grad_norm": 11.470852851867676, |
|
"learning_rate": 8.091451292246522e-07, |
|
"loss": 0.7857, |
|
"step": 819 |
|
}, |
|
{ |
|
"epoch": 0.8070866141732284, |
|
"grad_norm": 10.311899185180664, |
|
"learning_rate": 8.101391650099404e-07, |
|
"loss": 0.7219, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 0.8080708661417323, |
|
"grad_norm": 14.127314567565918, |
|
"learning_rate": 8.111332007952288e-07, |
|
"loss": 1.2658, |
|
"step": 821 |
|
}, |
|
{ |
|
"epoch": 0.8090551181102362, |
|
"grad_norm": 17.850603103637695, |
|
"learning_rate": 8.12127236580517e-07, |
|
"loss": 1.4563, |
|
"step": 822 |
|
}, |
|
{ |
|
"epoch": 0.8100393700787402, |
|
"grad_norm": 11.794126510620117, |
|
"learning_rate": 8.131212723658052e-07, |
|
"loss": 0.9098, |
|
"step": 823 |
|
}, |
|
{ |
|
"epoch": 0.8110236220472441, |
|
"grad_norm": 11.173970222473145, |
|
"learning_rate": 8.141153081510935e-07, |
|
"loss": 0.8606, |
|
"step": 824 |
|
}, |
|
{ |
|
"epoch": 0.812007874015748, |
|
"grad_norm": 13.826883316040039, |
|
"learning_rate": 8.151093439363818e-07, |
|
"loss": 1.1116, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 0.812992125984252, |
|
"grad_norm": 10.755367279052734, |
|
"learning_rate": 8.161033797216701e-07, |
|
"loss": 0.952, |
|
"step": 826 |
|
}, |
|
{ |
|
"epoch": 0.8139763779527559, |
|
"grad_norm": 11.222610473632812, |
|
"learning_rate": 8.170974155069583e-07, |
|
"loss": 0.6831, |
|
"step": 827 |
|
}, |
|
{ |
|
"epoch": 0.8149606299212598, |
|
"grad_norm": 10.882960319519043, |
|
"learning_rate": 8.180914512922467e-07, |
|
"loss": 0.8267, |
|
"step": 828 |
|
}, |
|
{ |
|
"epoch": 0.8159448818897638, |
|
"grad_norm": 12.082534790039062, |
|
"learning_rate": 8.19085487077535e-07, |
|
"loss": 0.934, |
|
"step": 829 |
|
}, |
|
{ |
|
"epoch": 0.8169291338582677, |
|
"grad_norm": 16.402433395385742, |
|
"learning_rate": 8.200795228628232e-07, |
|
"loss": 1.7555, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 0.8179133858267716, |
|
"grad_norm": 12.048799514770508, |
|
"learning_rate": 8.210735586481115e-07, |
|
"loss": 1.0109, |
|
"step": 831 |
|
}, |
|
{ |
|
"epoch": 0.8188976377952756, |
|
"grad_norm": 15.013277053833008, |
|
"learning_rate": 8.220675944333998e-07, |
|
"loss": 1.4509, |
|
"step": 832 |
|
}, |
|
{ |
|
"epoch": 0.8198818897637795, |
|
"grad_norm": 10.118528366088867, |
|
"learning_rate": 8.23061630218688e-07, |
|
"loss": 0.6548, |
|
"step": 833 |
|
}, |
|
{ |
|
"epoch": 0.8208661417322834, |
|
"grad_norm": 10.921168327331543, |
|
"learning_rate": 8.240556660039763e-07, |
|
"loss": 0.7777, |
|
"step": 834 |
|
}, |
|
{ |
|
"epoch": 0.8218503937007874, |
|
"grad_norm": 12.746010780334473, |
|
"learning_rate": 8.250497017892646e-07, |
|
"loss": 0.8929, |
|
"step": 835 |
|
}, |
|
{ |
|
"epoch": 0.8228346456692913, |
|
"grad_norm": 15.578157424926758, |
|
"learning_rate": 8.260437375745529e-07, |
|
"loss": 1.282, |
|
"step": 836 |
|
}, |
|
{ |
|
"epoch": 0.8238188976377953, |
|
"grad_norm": 11.15747356414795, |
|
"learning_rate": 8.270377733598411e-07, |
|
"loss": 0.6726, |
|
"step": 837 |
|
}, |
|
{ |
|
"epoch": 0.8248031496062992, |
|
"grad_norm": 14.893630027770996, |
|
"learning_rate": 8.280318091451293e-07, |
|
"loss": 1.0643, |
|
"step": 838 |
|
}, |
|
{ |
|
"epoch": 0.8257874015748031, |
|
"grad_norm": 11.360194206237793, |
|
"learning_rate": 8.290258449304177e-07, |
|
"loss": 1.01, |
|
"step": 839 |
|
}, |
|
{ |
|
"epoch": 0.8267716535433071, |
|
"grad_norm": 12.738455772399902, |
|
"learning_rate": 8.300198807157059e-07, |
|
"loss": 1.0631, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 0.827755905511811, |
|
"grad_norm": 13.322991371154785, |
|
"learning_rate": 8.310139165009941e-07, |
|
"loss": 1.092, |
|
"step": 841 |
|
}, |
|
{ |
|
"epoch": 0.8287401574803149, |
|
"grad_norm": 11.534586906433105, |
|
"learning_rate": 8.320079522862825e-07, |
|
"loss": 1.0219, |
|
"step": 842 |
|
}, |
|
{ |
|
"epoch": 0.8297244094488189, |
|
"grad_norm": 10.090067863464355, |
|
"learning_rate": 8.330019880715707e-07, |
|
"loss": 1.0437, |
|
"step": 843 |
|
}, |
|
{ |
|
"epoch": 0.8307086614173228, |
|
"grad_norm": 11.919770240783691, |
|
"learning_rate": 8.33996023856859e-07, |
|
"loss": 0.8789, |
|
"step": 844 |
|
}, |
|
{ |
|
"epoch": 0.8316929133858267, |
|
"grad_norm": 13.919114112854004, |
|
"learning_rate": 8.349900596421472e-07, |
|
"loss": 1.5409, |
|
"step": 845 |
|
}, |
|
{ |
|
"epoch": 0.8326771653543307, |
|
"grad_norm": 12.6904296875, |
|
"learning_rate": 8.359840954274355e-07, |
|
"loss": 1.365, |
|
"step": 846 |
|
}, |
|
{ |
|
"epoch": 0.8336614173228346, |
|
"grad_norm": 13.091736793518066, |
|
"learning_rate": 8.369781312127238e-07, |
|
"loss": 1.0567, |
|
"step": 847 |
|
}, |
|
{ |
|
"epoch": 0.8346456692913385, |
|
"grad_norm": 9.207596778869629, |
|
"learning_rate": 8.37972166998012e-07, |
|
"loss": 0.59, |
|
"step": 848 |
|
}, |
|
{ |
|
"epoch": 0.8356299212598425, |
|
"grad_norm": 9.879922866821289, |
|
"learning_rate": 8.389662027833004e-07, |
|
"loss": 0.839, |
|
"step": 849 |
|
}, |
|
{ |
|
"epoch": 0.8366141732283464, |
|
"grad_norm": 9.073246955871582, |
|
"learning_rate": 8.399602385685886e-07, |
|
"loss": 0.727, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.8375984251968503, |
|
"grad_norm": 10.825157165527344, |
|
"learning_rate": 8.409542743538768e-07, |
|
"loss": 0.8681, |
|
"step": 851 |
|
}, |
|
{ |
|
"epoch": 0.8385826771653543, |
|
"grad_norm": 14.727147102355957, |
|
"learning_rate": 8.419483101391651e-07, |
|
"loss": 1.5186, |
|
"step": 852 |
|
}, |
|
{ |
|
"epoch": 0.8395669291338582, |
|
"grad_norm": 11.10617446899414, |
|
"learning_rate": 8.429423459244534e-07, |
|
"loss": 0.8923, |
|
"step": 853 |
|
}, |
|
{ |
|
"epoch": 0.8405511811023622, |
|
"grad_norm": 10.533689498901367, |
|
"learning_rate": 8.439363817097416e-07, |
|
"loss": 0.8248, |
|
"step": 854 |
|
}, |
|
{ |
|
"epoch": 0.8415354330708661, |
|
"grad_norm": 13.720536231994629, |
|
"learning_rate": 8.449304174950299e-07, |
|
"loss": 0.9884, |
|
"step": 855 |
|
}, |
|
{ |
|
"epoch": 0.84251968503937, |
|
"grad_norm": 12.623149871826172, |
|
"learning_rate": 8.459244532803182e-07, |
|
"loss": 1.257, |
|
"step": 856 |
|
}, |
|
{ |
|
"epoch": 0.843503937007874, |
|
"grad_norm": 11.030608177185059, |
|
"learning_rate": 8.469184890656066e-07, |
|
"loss": 0.8841, |
|
"step": 857 |
|
}, |
|
{ |
|
"epoch": 0.844488188976378, |
|
"grad_norm": 14.188507080078125, |
|
"learning_rate": 8.479125248508948e-07, |
|
"loss": 1.232, |
|
"step": 858 |
|
}, |
|
{ |
|
"epoch": 0.8454724409448819, |
|
"grad_norm": 15.049591064453125, |
|
"learning_rate": 8.48906560636183e-07, |
|
"loss": 1.2214, |
|
"step": 859 |
|
}, |
|
{ |
|
"epoch": 0.8464566929133859, |
|
"grad_norm": 12.906851768493652, |
|
"learning_rate": 8.499005964214714e-07, |
|
"loss": 1.3332, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 0.8474409448818898, |
|
"grad_norm": 12.399016380310059, |
|
"learning_rate": 8.508946322067596e-07, |
|
"loss": 0.8272, |
|
"step": 861 |
|
}, |
|
{ |
|
"epoch": 0.8484251968503937, |
|
"grad_norm": 12.69371223449707, |
|
"learning_rate": 8.518886679920479e-07, |
|
"loss": 1.0901, |
|
"step": 862 |
|
}, |
|
{ |
|
"epoch": 0.8494094488188977, |
|
"grad_norm": 14.04008960723877, |
|
"learning_rate": 8.528827037773362e-07, |
|
"loss": 1.2892, |
|
"step": 863 |
|
}, |
|
{ |
|
"epoch": 0.8503937007874016, |
|
"grad_norm": 11.152538299560547, |
|
"learning_rate": 8.538767395626244e-07, |
|
"loss": 0.8817, |
|
"step": 864 |
|
}, |
|
{ |
|
"epoch": 0.8513779527559056, |
|
"grad_norm": 13.230735778808594, |
|
"learning_rate": 8.548707753479127e-07, |
|
"loss": 0.944, |
|
"step": 865 |
|
}, |
|
{ |
|
"epoch": 0.8523622047244095, |
|
"grad_norm": 10.027914047241211, |
|
"learning_rate": 8.558648111332009e-07, |
|
"loss": 0.6212, |
|
"step": 866 |
|
}, |
|
{ |
|
"epoch": 0.8533464566929134, |
|
"grad_norm": 14.65113639831543, |
|
"learning_rate": 8.568588469184892e-07, |
|
"loss": 1.4346, |
|
"step": 867 |
|
}, |
|
{ |
|
"epoch": 0.8543307086614174, |
|
"grad_norm": 9.519758224487305, |
|
"learning_rate": 8.578528827037775e-07, |
|
"loss": 0.7204, |
|
"step": 868 |
|
}, |
|
{ |
|
"epoch": 0.8553149606299213, |
|
"grad_norm": 17.111120223999023, |
|
"learning_rate": 8.588469184890657e-07, |
|
"loss": 1.4009, |
|
"step": 869 |
|
}, |
|
{ |
|
"epoch": 0.8562992125984252, |
|
"grad_norm": 13.902044296264648, |
|
"learning_rate": 8.598409542743541e-07, |
|
"loss": 1.1522, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 0.8572834645669292, |
|
"grad_norm": 11.630669593811035, |
|
"learning_rate": 8.608349900596423e-07, |
|
"loss": 0.9558, |
|
"step": 871 |
|
}, |
|
{ |
|
"epoch": 0.8582677165354331, |
|
"grad_norm": 10.162226676940918, |
|
"learning_rate": 8.618290258449305e-07, |
|
"loss": 0.8871, |
|
"step": 872 |
|
}, |
|
{ |
|
"epoch": 0.859251968503937, |
|
"grad_norm": 12.421677589416504, |
|
"learning_rate": 8.628230616302189e-07, |
|
"loss": 1.0025, |
|
"step": 873 |
|
}, |
|
{ |
|
"epoch": 0.860236220472441, |
|
"grad_norm": 11.895072937011719, |
|
"learning_rate": 8.638170974155071e-07, |
|
"loss": 0.8617, |
|
"step": 874 |
|
}, |
|
{ |
|
"epoch": 0.8612204724409449, |
|
"grad_norm": 12.787313461303711, |
|
"learning_rate": 8.648111332007954e-07, |
|
"loss": 1.1982, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 0.8622047244094488, |
|
"grad_norm": 18.052860260009766, |
|
"learning_rate": 8.658051689860836e-07, |
|
"loss": 1.2484, |
|
"step": 876 |
|
}, |
|
{ |
|
"epoch": 0.8631889763779528, |
|
"grad_norm": 15.959498405456543, |
|
"learning_rate": 8.667992047713719e-07, |
|
"loss": 1.0181, |
|
"step": 877 |
|
}, |
|
{ |
|
"epoch": 0.8641732283464567, |
|
"grad_norm": 15.16512680053711, |
|
"learning_rate": 8.677932405566602e-07, |
|
"loss": 0.8535, |
|
"step": 878 |
|
}, |
|
{ |
|
"epoch": 0.8651574803149606, |
|
"grad_norm": 15.01723575592041, |
|
"learning_rate": 8.687872763419484e-07, |
|
"loss": 1.144, |
|
"step": 879 |
|
}, |
|
{ |
|
"epoch": 0.8661417322834646, |
|
"grad_norm": 13.688557624816895, |
|
"learning_rate": 8.697813121272366e-07, |
|
"loss": 1.1927, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 0.8671259842519685, |
|
"grad_norm": 10.62550163269043, |
|
"learning_rate": 8.70775347912525e-07, |
|
"loss": 0.7245, |
|
"step": 881 |
|
}, |
|
{ |
|
"epoch": 0.8681102362204725, |
|
"grad_norm": 11.776147842407227, |
|
"learning_rate": 8.717693836978132e-07, |
|
"loss": 1.0118, |
|
"step": 882 |
|
}, |
|
{ |
|
"epoch": 0.8690944881889764, |
|
"grad_norm": 15.631773948669434, |
|
"learning_rate": 8.727634194831015e-07, |
|
"loss": 1.4399, |
|
"step": 883 |
|
}, |
|
{ |
|
"epoch": 0.8700787401574803, |
|
"grad_norm": 13.422005653381348, |
|
"learning_rate": 8.737574552683898e-07, |
|
"loss": 1.0666, |
|
"step": 884 |
|
}, |
|
{ |
|
"epoch": 0.8710629921259843, |
|
"grad_norm": 17.7437686920166, |
|
"learning_rate": 8.74751491053678e-07, |
|
"loss": 1.1947, |
|
"step": 885 |
|
}, |
|
{ |
|
"epoch": 0.8720472440944882, |
|
"grad_norm": 11.687212944030762, |
|
"learning_rate": 8.757455268389663e-07, |
|
"loss": 0.8908, |
|
"step": 886 |
|
}, |
|
{ |
|
"epoch": 0.8730314960629921, |
|
"grad_norm": 15.046998977661133, |
|
"learning_rate": 8.767395626242545e-07, |
|
"loss": 1.394, |
|
"step": 887 |
|
}, |
|
{ |
|
"epoch": 0.8740157480314961, |
|
"grad_norm": 11.96061897277832, |
|
"learning_rate": 8.77733598409543e-07, |
|
"loss": 0.802, |
|
"step": 888 |
|
}, |
|
{ |
|
"epoch": 0.875, |
|
"grad_norm": 12.169413566589355, |
|
"learning_rate": 8.787276341948312e-07, |
|
"loss": 0.9879, |
|
"step": 889 |
|
}, |
|
{ |
|
"epoch": 0.8759842519685039, |
|
"grad_norm": 11.521828651428223, |
|
"learning_rate": 8.797216699801194e-07, |
|
"loss": 0.9895, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 0.8769685039370079, |
|
"grad_norm": 12.837910652160645, |
|
"learning_rate": 8.807157057654078e-07, |
|
"loss": 1.0927, |
|
"step": 891 |
|
}, |
|
{ |
|
"epoch": 0.8779527559055118, |
|
"grad_norm": 14.709928512573242, |
|
"learning_rate": 8.81709741550696e-07, |
|
"loss": 1.4376, |
|
"step": 892 |
|
}, |
|
{ |
|
"epoch": 0.8789370078740157, |
|
"grad_norm": 13.70672607421875, |
|
"learning_rate": 8.827037773359842e-07, |
|
"loss": 1.1509, |
|
"step": 893 |
|
}, |
|
{ |
|
"epoch": 0.8799212598425197, |
|
"grad_norm": 12.354025840759277, |
|
"learning_rate": 8.836978131212726e-07, |
|
"loss": 0.7944, |
|
"step": 894 |
|
}, |
|
{ |
|
"epoch": 0.8809055118110236, |
|
"grad_norm": 14.6171236038208, |
|
"learning_rate": 8.846918489065608e-07, |
|
"loss": 1.4574, |
|
"step": 895 |
|
}, |
|
{ |
|
"epoch": 0.8818897637795275, |
|
"grad_norm": 12.5920991897583, |
|
"learning_rate": 8.856858846918491e-07, |
|
"loss": 1.2278, |
|
"step": 896 |
|
}, |
|
{ |
|
"epoch": 0.8828740157480315, |
|
"grad_norm": 13.8173189163208, |
|
"learning_rate": 8.866799204771373e-07, |
|
"loss": 1.1974, |
|
"step": 897 |
|
}, |
|
{ |
|
"epoch": 0.8838582677165354, |
|
"grad_norm": 12.641080856323242, |
|
"learning_rate": 8.876739562624256e-07, |
|
"loss": 1.0711, |
|
"step": 898 |
|
}, |
|
{ |
|
"epoch": 0.8848425196850394, |
|
"grad_norm": 14.381937026977539, |
|
"learning_rate": 8.886679920477139e-07, |
|
"loss": 1.3942, |
|
"step": 899 |
|
}, |
|
{ |
|
"epoch": 0.8858267716535433, |
|
"grad_norm": 13.973237037658691, |
|
"learning_rate": 8.896620278330021e-07, |
|
"loss": 1.1718, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.8868110236220472, |
|
"grad_norm": 12.490571022033691, |
|
"learning_rate": 8.906560636182905e-07, |
|
"loss": 1.1274, |
|
"step": 901 |
|
}, |
|
{ |
|
"epoch": 0.8877952755905512, |
|
"grad_norm": 10.961933135986328, |
|
"learning_rate": 8.916500994035787e-07, |
|
"loss": 0.9954, |
|
"step": 902 |
|
}, |
|
{ |
|
"epoch": 0.8887795275590551, |
|
"grad_norm": 10.65230655670166, |
|
"learning_rate": 8.926441351888669e-07, |
|
"loss": 1.0155, |
|
"step": 903 |
|
}, |
|
{ |
|
"epoch": 0.889763779527559, |
|
"grad_norm": 14.470539093017578, |
|
"learning_rate": 8.936381709741552e-07, |
|
"loss": 1.4032, |
|
"step": 904 |
|
}, |
|
{ |
|
"epoch": 0.890748031496063, |
|
"grad_norm": 15.179549217224121, |
|
"learning_rate": 8.946322067594435e-07, |
|
"loss": 1.4135, |
|
"step": 905 |
|
}, |
|
{ |
|
"epoch": 0.8917322834645669, |
|
"grad_norm": 11.116039276123047, |
|
"learning_rate": 8.956262425447318e-07, |
|
"loss": 1.1893, |
|
"step": 906 |
|
}, |
|
{ |
|
"epoch": 0.8927165354330708, |
|
"grad_norm": 10.903943061828613, |
|
"learning_rate": 8.9662027833002e-07, |
|
"loss": 0.7243, |
|
"step": 907 |
|
}, |
|
{ |
|
"epoch": 0.8937007874015748, |
|
"grad_norm": 12.204026222229004, |
|
"learning_rate": 8.976143141153082e-07, |
|
"loss": 1.1674, |
|
"step": 908 |
|
}, |
|
{ |
|
"epoch": 0.8946850393700787, |
|
"grad_norm": 12.262521743774414, |
|
"learning_rate": 8.986083499005966e-07, |
|
"loss": 1.3777, |
|
"step": 909 |
|
}, |
|
{ |
|
"epoch": 0.8956692913385826, |
|
"grad_norm": 12.773459434509277, |
|
"learning_rate": 8.996023856858848e-07, |
|
"loss": 1.0281, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 0.8966535433070866, |
|
"grad_norm": 16.213043212890625, |
|
"learning_rate": 9.00596421471173e-07, |
|
"loss": 1.2869, |
|
"step": 911 |
|
}, |
|
{ |
|
"epoch": 0.8976377952755905, |
|
"grad_norm": 10.814340591430664, |
|
"learning_rate": 9.015904572564614e-07, |
|
"loss": 1.0125, |
|
"step": 912 |
|
}, |
|
{ |
|
"epoch": 0.8986220472440944, |
|
"grad_norm": 11.734081268310547, |
|
"learning_rate": 9.025844930417496e-07, |
|
"loss": 1.0047, |
|
"step": 913 |
|
}, |
|
{ |
|
"epoch": 0.8996062992125984, |
|
"grad_norm": 10.440587997436523, |
|
"learning_rate": 9.035785288270379e-07, |
|
"loss": 0.8839, |
|
"step": 914 |
|
}, |
|
{ |
|
"epoch": 0.9005905511811023, |
|
"grad_norm": 13.46262264251709, |
|
"learning_rate": 9.045725646123262e-07, |
|
"loss": 1.4279, |
|
"step": 915 |
|
}, |
|
{ |
|
"epoch": 0.9015748031496063, |
|
"grad_norm": 13.135570526123047, |
|
"learning_rate": 9.055666003976144e-07, |
|
"loss": 1.053, |
|
"step": 916 |
|
}, |
|
{ |
|
"epoch": 0.9025590551181102, |
|
"grad_norm": 10.498846054077148, |
|
"learning_rate": 9.065606361829028e-07, |
|
"loss": 0.7841, |
|
"step": 917 |
|
}, |
|
{ |
|
"epoch": 0.9035433070866141, |
|
"grad_norm": 8.559603691101074, |
|
"learning_rate": 9.07554671968191e-07, |
|
"loss": 0.7738, |
|
"step": 918 |
|
}, |
|
{ |
|
"epoch": 0.9035433070866141, |
|
"eval_Qnli-dev_cosine_accuracy": 0.6484375, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.8556903004646301, |
|
"eval_Qnli-dev_cosine_ap": 0.6712050759826647, |
|
"eval_Qnli-dev_cosine_f1": 0.6836734693877551, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.817956805229187, |
|
"eval_Qnli-dev_cosine_precision": 0.5710227272727273, |
|
"eval_Qnli-dev_cosine_recall": 0.8516949152542372, |
|
"eval_Qnli-dev_dot_accuracy": 0.646484375, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 663.1381225585938, |
|
"eval_Qnli-dev_dot_ap": 0.6701392631858906, |
|
"eval_Qnli-dev_dot_f1": 0.6835871404399323, |
|
"eval_Qnli-dev_dot_f1_threshold": 623.9208984375, |
|
"eval_Qnli-dev_dot_precision": 0.5690140845070423, |
|
"eval_Qnli-dev_dot_recall": 0.8559322033898306, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.650390625, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 14.574091911315918, |
|
"eval_Qnli-dev_euclidean_ap": 0.6711308988804017, |
|
"eval_Qnli-dev_euclidean_f1": 0.6844741235392321, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 16.858455657958984, |
|
"eval_Qnli-dev_euclidean_precision": 0.5647382920110193, |
|
"eval_Qnli-dev_euclidean_recall": 0.8686440677966102, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.65625, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 308.8890380859375, |
|
"eval_Qnli-dev_manhattan_ap": 0.6758066941744509, |
|
"eval_Qnli-dev_manhattan_f1": 0.6814580031695722, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 359.8348083496094, |
|
"eval_Qnli-dev_manhattan_precision": 0.5443037974683544, |
|
"eval_Qnli-dev_manhattan_recall": 0.9110169491525424, |
|
"eval_Qnli-dev_max_accuracy": 0.65625, |
|
"eval_Qnli-dev_max_accuracy_threshold": 663.1381225585938, |
|
"eval_Qnli-dev_max_ap": 0.6758066941744509, |
|
"eval_Qnli-dev_max_f1": 0.6844741235392321, |
|
"eval_Qnli-dev_max_f1_threshold": 623.9208984375, |
|
"eval_Qnli-dev_max_precision": 0.5710227272727273, |
|
"eval_Qnli-dev_max_recall": 0.9110169491525424, |
|
"eval_allNLI-dev_cosine_accuracy": 0.681640625, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9492451548576355, |
|
"eval_allNLI-dev_cosine_ap": 0.4860204244756696, |
|
"eval_allNLI-dev_cosine_f1": 0.5408805031446541, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.8793689012527466, |
|
"eval_allNLI-dev_cosine_precision": 0.4243421052631579, |
|
"eval_allNLI-dev_cosine_recall": 0.7456647398843931, |
|
"eval_allNLI-dev_dot_accuracy": 0.681640625, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 719.29296875, |
|
"eval_allNLI-dev_dot_ap": 0.4850643671480781, |
|
"eval_allNLI-dev_dot_f1": 0.5408805031446541, |
|
"eval_allNLI-dev_dot_f1_threshold": 671.677001953125, |
|
"eval_allNLI-dev_dot_precision": 0.4243421052631579, |
|
"eval_allNLI-dev_dot_recall": 0.7456647398843931, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.681640625, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 8.806386947631836, |
|
"eval_allNLI-dev_euclidean_ap": 0.4857904504087363, |
|
"eval_allNLI-dev_euclidean_f1": 0.5408805031446541, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 13.574195861816406, |
|
"eval_allNLI-dev_euclidean_precision": 0.4243421052631579, |
|
"eval_allNLI-dev_euclidean_recall": 0.7456647398843931, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.6796875, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 179.25799560546875, |
|
"eval_allNLI-dev_manhattan_ap": 0.48263260956262277, |
|
"eval_allNLI-dev_manhattan_f1": 0.5512820512820513, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 281.2195739746094, |
|
"eval_allNLI-dev_manhattan_precision": 0.43728813559322033, |
|
"eval_allNLI-dev_manhattan_recall": 0.7456647398843931, |
|
"eval_allNLI-dev_max_accuracy": 0.681640625, |
|
"eval_allNLI-dev_max_accuracy_threshold": 719.29296875, |
|
"eval_allNLI-dev_max_ap": 0.4860204244756696, |
|
"eval_allNLI-dev_max_f1": 0.5512820512820513, |
|
"eval_allNLI-dev_max_f1_threshold": 671.677001953125, |
|
"eval_allNLI-dev_max_precision": 0.43728813559322033, |
|
"eval_allNLI-dev_max_recall": 0.7456647398843931, |
|
"eval_loss": 1.2464040517807007, |
|
"eval_runtime": 52.8966, |
|
"eval_samples_per_second": 31.458, |
|
"eval_sequential_score": 0.6758066941744509, |
|
"eval_steps_per_second": 0.132, |
|
"eval_sts-test_pearson_cosine": 0.49084594999467995, |
|
"eval_sts-test_pearson_dot": 0.48706933209590814, |
|
"eval_sts-test_pearson_euclidean": 0.5093765998387638, |
|
"eval_sts-test_pearson_manhattan": 0.5202808585227008, |
|
"eval_sts-test_pearson_max": 0.5202808585227008, |
|
"eval_sts-test_spearman_cosine": 0.5248642742038014, |
|
"eval_sts-test_spearman_dot": 0.5211632670331621, |
|
"eval_sts-test_spearman_euclidean": 0.5250007655510451, |
|
"eval_sts-test_spearman_manhattan": 0.5343063428447727, |
|
"eval_sts-test_spearman_max": 0.5343063428447727, |
|
"step": 918 |
|
}, |
|
{ |
|
"epoch": 0.9045275590551181, |
|
"grad_norm": 10.508577346801758, |
|
"learning_rate": 9.085487077534794e-07, |
|
"loss": 0.8998, |
|
"step": 919 |
|
}, |
|
{ |
|
"epoch": 0.905511811023622, |
|
"grad_norm": 15.358297348022461, |
|
"learning_rate": 9.095427435387676e-07, |
|
"loss": 1.7733, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 0.906496062992126, |
|
"grad_norm": 11.635780334472656, |
|
"learning_rate": 9.105367793240558e-07, |
|
"loss": 0.8032, |
|
"step": 921 |
|
}, |
|
{ |
|
"epoch": 0.90748031496063, |
|
"grad_norm": 9.231951713562012, |
|
"learning_rate": 9.115308151093442e-07, |
|
"loss": 0.6311, |
|
"step": 922 |
|
}, |
|
{ |
|
"epoch": 0.9084645669291339, |
|
"grad_norm": 10.531167030334473, |
|
"learning_rate": 9.125248508946324e-07, |
|
"loss": 0.9534, |
|
"step": 923 |
|
}, |
|
{ |
|
"epoch": 0.9094488188976378, |
|
"grad_norm": 12.956085205078125, |
|
"learning_rate": 9.135188866799206e-07, |
|
"loss": 1.2538, |
|
"step": 924 |
|
}, |
|
{ |
|
"epoch": 0.9104330708661418, |
|
"grad_norm": 13.205366134643555, |
|
"learning_rate": 9.145129224652089e-07, |
|
"loss": 0.5411, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 0.9114173228346457, |
|
"grad_norm": 11.612566947937012, |
|
"learning_rate": 9.155069582504972e-07, |
|
"loss": 0.797, |
|
"step": 926 |
|
}, |
|
{ |
|
"epoch": 0.9124015748031497, |
|
"grad_norm": 15.050639152526855, |
|
"learning_rate": 9.165009940357855e-07, |
|
"loss": 1.2115, |
|
"step": 927 |
|
}, |
|
{ |
|
"epoch": 0.9133858267716536, |
|
"grad_norm": 11.717726707458496, |
|
"learning_rate": 9.174950298210737e-07, |
|
"loss": 0.8412, |
|
"step": 928 |
|
}, |
|
{ |
|
"epoch": 0.9143700787401575, |
|
"grad_norm": 11.777231216430664, |
|
"learning_rate": 9.18489065606362e-07, |
|
"loss": 0.7325, |
|
"step": 929 |
|
}, |
|
{ |
|
"epoch": 0.9153543307086615, |
|
"grad_norm": 22.675230026245117, |
|
"learning_rate": 9.194831013916503e-07, |
|
"loss": 1.6631, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 0.9163385826771654, |
|
"grad_norm": 12.434297561645508, |
|
"learning_rate": 9.204771371769385e-07, |
|
"loss": 0.8573, |
|
"step": 931 |
|
}, |
|
{ |
|
"epoch": 0.9173228346456693, |
|
"grad_norm": 14.405913352966309, |
|
"learning_rate": 9.214711729622268e-07, |
|
"loss": 0.8194, |
|
"step": 932 |
|
}, |
|
{ |
|
"epoch": 0.9183070866141733, |
|
"grad_norm": 14.679107666015625, |
|
"learning_rate": 9.224652087475151e-07, |
|
"loss": 1.1627, |
|
"step": 933 |
|
}, |
|
{ |
|
"epoch": 0.9192913385826772, |
|
"grad_norm": 18.158557891845703, |
|
"learning_rate": 9.234592445328033e-07, |
|
"loss": 1.3729, |
|
"step": 934 |
|
}, |
|
{ |
|
"epoch": 0.9202755905511811, |
|
"grad_norm": 16.845603942871094, |
|
"learning_rate": 9.244532803180916e-07, |
|
"loss": 1.0679, |
|
"step": 935 |
|
}, |
|
{ |
|
"epoch": 0.9212598425196851, |
|
"grad_norm": 14.517767906188965, |
|
"learning_rate": 9.254473161033799e-07, |
|
"loss": 1.2123, |
|
"step": 936 |
|
}, |
|
{ |
|
"epoch": 0.922244094488189, |
|
"grad_norm": 13.135607719421387, |
|
"learning_rate": 9.264413518886681e-07, |
|
"loss": 0.9217, |
|
"step": 937 |
|
}, |
|
{ |
|
"epoch": 0.9232283464566929, |
|
"grad_norm": 11.955747604370117, |
|
"learning_rate": 9.274353876739564e-07, |
|
"loss": 0.6959, |
|
"step": 938 |
|
}, |
|
{ |
|
"epoch": 0.9242125984251969, |
|
"grad_norm": 12.874065399169922, |
|
"learning_rate": 9.284294234592446e-07, |
|
"loss": 0.9306, |
|
"step": 939 |
|
}, |
|
{ |
|
"epoch": 0.9251968503937008, |
|
"grad_norm": 15.19837760925293, |
|
"learning_rate": 9.29423459244533e-07, |
|
"loss": 1.2815, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 0.9261811023622047, |
|
"grad_norm": 11.53364372253418, |
|
"learning_rate": 9.304174950298212e-07, |
|
"loss": 1.1188, |
|
"step": 941 |
|
}, |
|
{ |
|
"epoch": 0.9271653543307087, |
|
"grad_norm": 10.0062837600708, |
|
"learning_rate": 9.314115308151094e-07, |
|
"loss": 0.8454, |
|
"step": 942 |
|
}, |
|
{ |
|
"epoch": 0.9281496062992126, |
|
"grad_norm": 10.954655647277832, |
|
"learning_rate": 9.324055666003978e-07, |
|
"loss": 0.7369, |
|
"step": 943 |
|
}, |
|
{ |
|
"epoch": 0.9291338582677166, |
|
"grad_norm": 12.967487335205078, |
|
"learning_rate": 9.33399602385686e-07, |
|
"loss": 1.2244, |
|
"step": 944 |
|
}, |
|
{ |
|
"epoch": 0.9301181102362205, |
|
"grad_norm": 12.993568420410156, |
|
"learning_rate": 9.343936381709743e-07, |
|
"loss": 1.1402, |
|
"step": 945 |
|
}, |
|
{ |
|
"epoch": 0.9311023622047244, |
|
"grad_norm": 7.9335832595825195, |
|
"learning_rate": 9.353876739562625e-07, |
|
"loss": 0.4213, |
|
"step": 946 |
|
}, |
|
{ |
|
"epoch": 0.9320866141732284, |
|
"grad_norm": 14.420933723449707, |
|
"learning_rate": 9.363817097415509e-07, |
|
"loss": 1.0047, |
|
"step": 947 |
|
}, |
|
{ |
|
"epoch": 0.9330708661417323, |
|
"grad_norm": 13.709823608398438, |
|
"learning_rate": 9.373757455268392e-07, |
|
"loss": 1.1353, |
|
"step": 948 |
|
}, |
|
{ |
|
"epoch": 0.9340551181102362, |
|
"grad_norm": 16.349287033081055, |
|
"learning_rate": 9.383697813121274e-07, |
|
"loss": 1.2862, |
|
"step": 949 |
|
}, |
|
{ |
|
"epoch": 0.9350393700787402, |
|
"grad_norm": 11.642127990722656, |
|
"learning_rate": 9.393638170974157e-07, |
|
"loss": 0.6358, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.9360236220472441, |
|
"grad_norm": 11.004044532775879, |
|
"learning_rate": 9.40357852882704e-07, |
|
"loss": 0.676, |
|
"step": 951 |
|
}, |
|
{ |
|
"epoch": 0.937007874015748, |
|
"grad_norm": 13.824701309204102, |
|
"learning_rate": 9.413518886679922e-07, |
|
"loss": 1.1628, |
|
"step": 952 |
|
}, |
|
{ |
|
"epoch": 0.937992125984252, |
|
"grad_norm": 12.448197364807129, |
|
"learning_rate": 9.423459244532805e-07, |
|
"loss": 0.861, |
|
"step": 953 |
|
}, |
|
{ |
|
"epoch": 0.9389763779527559, |
|
"grad_norm": 11.722825050354004, |
|
"learning_rate": 9.433399602385688e-07, |
|
"loss": 0.7121, |
|
"step": 954 |
|
}, |
|
{ |
|
"epoch": 0.9399606299212598, |
|
"grad_norm": 11.611307144165039, |
|
"learning_rate": 9.44333996023857e-07, |
|
"loss": 0.6625, |
|
"step": 955 |
|
}, |
|
{ |
|
"epoch": 0.9409448818897638, |
|
"grad_norm": 11.159773826599121, |
|
"learning_rate": 9.453280318091453e-07, |
|
"loss": 0.9343, |
|
"step": 956 |
|
}, |
|
{ |
|
"epoch": 0.9419291338582677, |
|
"grad_norm": 10.660136222839355, |
|
"learning_rate": 9.463220675944336e-07, |
|
"loss": 0.8103, |
|
"step": 957 |
|
}, |
|
{ |
|
"epoch": 0.9429133858267716, |
|
"grad_norm": 11.581384658813477, |
|
"learning_rate": 9.473161033797219e-07, |
|
"loss": 1.047, |
|
"step": 958 |
|
}, |
|
{ |
|
"epoch": 0.9438976377952756, |
|
"grad_norm": 10.32679557800293, |
|
"learning_rate": 9.483101391650101e-07, |
|
"loss": 0.6455, |
|
"step": 959 |
|
}, |
|
{ |
|
"epoch": 0.9448818897637795, |
|
"grad_norm": 12.044934272766113, |
|
"learning_rate": 9.493041749502983e-07, |
|
"loss": 1.0978, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 0.9458661417322834, |
|
"grad_norm": 11.067245483398438, |
|
"learning_rate": 9.502982107355867e-07, |
|
"loss": 0.9522, |
|
"step": 961 |
|
}, |
|
{ |
|
"epoch": 0.9468503937007874, |
|
"grad_norm": 11.00229549407959, |
|
"learning_rate": 9.512922465208749e-07, |
|
"loss": 0.9151, |
|
"step": 962 |
|
}, |
|
{ |
|
"epoch": 0.9478346456692913, |
|
"grad_norm": 13.090099334716797, |
|
"learning_rate": 9.522862823061631e-07, |
|
"loss": 1.2515, |
|
"step": 963 |
|
}, |
|
{ |
|
"epoch": 0.9488188976377953, |
|
"grad_norm": 13.134993553161621, |
|
"learning_rate": 9.532803180914515e-07, |
|
"loss": 1.0809, |
|
"step": 964 |
|
}, |
|
{ |
|
"epoch": 0.9498031496062992, |
|
"grad_norm": 10.051626205444336, |
|
"learning_rate": 9.542743538767396e-07, |
|
"loss": 0.6971, |
|
"step": 965 |
|
}, |
|
{ |
|
"epoch": 0.9507874015748031, |
|
"grad_norm": 15.127157211303711, |
|
"learning_rate": 9.55268389662028e-07, |
|
"loss": 1.4086, |
|
"step": 966 |
|
}, |
|
{ |
|
"epoch": 0.9517716535433071, |
|
"grad_norm": 13.528825759887695, |
|
"learning_rate": 9.562624254473162e-07, |
|
"loss": 1.3102, |
|
"step": 967 |
|
}, |
|
{ |
|
"epoch": 0.952755905511811, |
|
"grad_norm": 15.755900382995605, |
|
"learning_rate": 9.572564612326044e-07, |
|
"loss": 1.5008, |
|
"step": 968 |
|
}, |
|
{ |
|
"epoch": 0.9537401574803149, |
|
"grad_norm": 11.590021133422852, |
|
"learning_rate": 9.582504970178928e-07, |
|
"loss": 0.9449, |
|
"step": 969 |
|
}, |
|
{ |
|
"epoch": 0.9547244094488189, |
|
"grad_norm": 12.689349174499512, |
|
"learning_rate": 9.59244532803181e-07, |
|
"loss": 1.0705, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 0.9557086614173228, |
|
"grad_norm": 12.062538146972656, |
|
"learning_rate": 9.602385685884694e-07, |
|
"loss": 0.7788, |
|
"step": 971 |
|
}, |
|
{ |
|
"epoch": 0.9566929133858267, |
|
"grad_norm": 11.848228454589844, |
|
"learning_rate": 9.612326043737576e-07, |
|
"loss": 1.0007, |
|
"step": 972 |
|
}, |
|
{ |
|
"epoch": 0.9576771653543307, |
|
"grad_norm": 12.216765403747559, |
|
"learning_rate": 9.622266401590458e-07, |
|
"loss": 0.9345, |
|
"step": 973 |
|
}, |
|
{ |
|
"epoch": 0.9586614173228346, |
|
"grad_norm": 13.60435962677002, |
|
"learning_rate": 9.632206759443342e-07, |
|
"loss": 1.4525, |
|
"step": 974 |
|
}, |
|
{ |
|
"epoch": 0.9596456692913385, |
|
"grad_norm": 11.18852710723877, |
|
"learning_rate": 9.642147117296224e-07, |
|
"loss": 0.5892, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 0.9606299212598425, |
|
"grad_norm": 12.86732006072998, |
|
"learning_rate": 9.652087475149109e-07, |
|
"loss": 0.927, |
|
"step": 976 |
|
}, |
|
{ |
|
"epoch": 0.9616141732283464, |
|
"grad_norm": 11.5518798828125, |
|
"learning_rate": 9.66202783300199e-07, |
|
"loss": 0.7473, |
|
"step": 977 |
|
}, |
|
{ |
|
"epoch": 0.9625984251968503, |
|
"grad_norm": 10.896742820739746, |
|
"learning_rate": 9.671968190854873e-07, |
|
"loss": 0.8173, |
|
"step": 978 |
|
}, |
|
{ |
|
"epoch": 0.9635826771653543, |
|
"grad_norm": 9.958429336547852, |
|
"learning_rate": 9.681908548707755e-07, |
|
"loss": 0.6888, |
|
"step": 979 |
|
}, |
|
{ |
|
"epoch": 0.9645669291338582, |
|
"grad_norm": 11.125531196594238, |
|
"learning_rate": 9.691848906560639e-07, |
|
"loss": 0.8059, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 0.9655511811023622, |
|
"grad_norm": 10.909327507019043, |
|
"learning_rate": 9.70178926441352e-07, |
|
"loss": 0.9087, |
|
"step": 981 |
|
}, |
|
{ |
|
"epoch": 0.9665354330708661, |
|
"grad_norm": 16.282583236694336, |
|
"learning_rate": 9.711729622266403e-07, |
|
"loss": 1.5197, |
|
"step": 982 |
|
}, |
|
{ |
|
"epoch": 0.96751968503937, |
|
"grad_norm": 11.268091201782227, |
|
"learning_rate": 9.721669980119285e-07, |
|
"loss": 0.6625, |
|
"step": 983 |
|
}, |
|
{ |
|
"epoch": 0.968503937007874, |
|
"grad_norm": 12.072832107543945, |
|
"learning_rate": 9.731610337972169e-07, |
|
"loss": 1.0372, |
|
"step": 984 |
|
}, |
|
{ |
|
"epoch": 0.969488188976378, |
|
"grad_norm": 13.086760520935059, |
|
"learning_rate": 9.74155069582505e-07, |
|
"loss": 0.91, |
|
"step": 985 |
|
}, |
|
{ |
|
"epoch": 0.9704724409448819, |
|
"grad_norm": 11.189858436584473, |
|
"learning_rate": 9.751491053677933e-07, |
|
"loss": 0.7797, |
|
"step": 986 |
|
}, |
|
{ |
|
"epoch": 0.9714566929133859, |
|
"grad_norm": 13.82830810546875, |
|
"learning_rate": 9.761431411530817e-07, |
|
"loss": 0.838, |
|
"step": 987 |
|
}, |
|
{ |
|
"epoch": 0.9724409448818898, |
|
"grad_norm": 14.06183910369873, |
|
"learning_rate": 9.771371769383699e-07, |
|
"loss": 0.9568, |
|
"step": 988 |
|
}, |
|
{ |
|
"epoch": 0.9734251968503937, |
|
"grad_norm": 18.976341247558594, |
|
"learning_rate": 9.781312127236583e-07, |
|
"loss": 1.4704, |
|
"step": 989 |
|
}, |
|
{ |
|
"epoch": 0.9744094488188977, |
|
"grad_norm": 12.61073112487793, |
|
"learning_rate": 9.791252485089465e-07, |
|
"loss": 0.8209, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 0.9753937007874016, |
|
"grad_norm": 14.609846115112305, |
|
"learning_rate": 9.801192842942347e-07, |
|
"loss": 0.86, |
|
"step": 991 |
|
}, |
|
{ |
|
"epoch": 0.9763779527559056, |
|
"grad_norm": 11.316509246826172, |
|
"learning_rate": 9.811133200795231e-07, |
|
"loss": 0.7749, |
|
"step": 992 |
|
}, |
|
{ |
|
"epoch": 0.9773622047244095, |
|
"grad_norm": 11.958404541015625, |
|
"learning_rate": 9.821073558648113e-07, |
|
"loss": 0.7225, |
|
"step": 993 |
|
}, |
|
{ |
|
"epoch": 0.9783464566929134, |
|
"grad_norm": 14.443065643310547, |
|
"learning_rate": 9.831013916500995e-07, |
|
"loss": 0.8488, |
|
"step": 994 |
|
}, |
|
{ |
|
"epoch": 0.9793307086614174, |
|
"grad_norm": 19.713075637817383, |
|
"learning_rate": 9.84095427435388e-07, |
|
"loss": 1.4838, |
|
"step": 995 |
|
}, |
|
{ |
|
"epoch": 0.9803149606299213, |
|
"grad_norm": 15.081480979919434, |
|
"learning_rate": 9.850894632206761e-07, |
|
"loss": 1.3282, |
|
"step": 996 |
|
}, |
|
{ |
|
"epoch": 0.9812992125984252, |
|
"grad_norm": 14.910409927368164, |
|
"learning_rate": 9.860834990059645e-07, |
|
"loss": 0.9905, |
|
"step": 997 |
|
}, |
|
{ |
|
"epoch": 0.9822834645669292, |
|
"grad_norm": 21.365158081054688, |
|
"learning_rate": 9.870775347912527e-07, |
|
"loss": 1.7902, |
|
"step": 998 |
|
}, |
|
{ |
|
"epoch": 0.9832677165354331, |
|
"grad_norm": 14.85123062133789, |
|
"learning_rate": 9.88071570576541e-07, |
|
"loss": 0.7411, |
|
"step": 999 |
|
}, |
|
{ |
|
"epoch": 0.984251968503937, |
|
"grad_norm": 12.506738662719727, |
|
"learning_rate": 9.890656063618291e-07, |
|
"loss": 1.2985, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.985236220472441, |
|
"grad_norm": 12.294382095336914, |
|
"learning_rate": 9.900596421471175e-07, |
|
"loss": 0.9858, |
|
"step": 1001 |
|
}, |
|
{ |
|
"epoch": 0.9862204724409449, |
|
"grad_norm": 10.234291076660156, |
|
"learning_rate": 9.910536779324057e-07, |
|
"loss": 0.8951, |
|
"step": 1002 |
|
}, |
|
{ |
|
"epoch": 0.9872047244094488, |
|
"grad_norm": 8.601326942443848, |
|
"learning_rate": 9.92047713717694e-07, |
|
"loss": 0.5526, |
|
"step": 1003 |
|
}, |
|
{ |
|
"epoch": 0.9881889763779528, |
|
"grad_norm": 15.147130966186523, |
|
"learning_rate": 9.930417495029821e-07, |
|
"loss": 1.466, |
|
"step": 1004 |
|
}, |
|
{ |
|
"epoch": 0.9891732283464567, |
|
"grad_norm": 8.791632652282715, |
|
"learning_rate": 9.940357852882705e-07, |
|
"loss": 0.5767, |
|
"step": 1005 |
|
}, |
|
{ |
|
"epoch": 0.9901574803149606, |
|
"grad_norm": 13.056974411010742, |
|
"learning_rate": 9.950298210735587e-07, |
|
"loss": 0.8946, |
|
"step": 1006 |
|
}, |
|
{ |
|
"epoch": 0.9911417322834646, |
|
"grad_norm": 14.539746284484863, |
|
"learning_rate": 9.96023856858847e-07, |
|
"loss": 1.0825, |
|
"step": 1007 |
|
}, |
|
{ |
|
"epoch": 0.9921259842519685, |
|
"grad_norm": 12.327513694763184, |
|
"learning_rate": 9.970178926441354e-07, |
|
"loss": 1.0152, |
|
"step": 1008 |
|
}, |
|
{ |
|
"epoch": 0.9931102362204725, |
|
"grad_norm": 10.933964729309082, |
|
"learning_rate": 9.980119284294236e-07, |
|
"loss": 0.7586, |
|
"step": 1009 |
|
}, |
|
{ |
|
"epoch": 0.9940944881889764, |
|
"grad_norm": 12.646042823791504, |
|
"learning_rate": 9.99005964214712e-07, |
|
"loss": 1.3353, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 0.9950787401574803, |
|
"grad_norm": 11.157730102539062, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 0.9555, |
|
"step": 1011 |
|
}, |
|
{ |
|
"epoch": 0.9960629921259843, |
|
"grad_norm": 13.540596961975098, |
|
"learning_rate": 9.999994102361802e-07, |
|
"loss": 1.0585, |
|
"step": 1012 |
|
}, |
|
{ |
|
"epoch": 0.9970472440944882, |
|
"grad_norm": 9.059813499450684, |
|
"learning_rate": 9.999976409461164e-07, |
|
"loss": 0.6774, |
|
"step": 1013 |
|
}, |
|
{ |
|
"epoch": 0.9980314960629921, |
|
"grad_norm": 14.586808204650879, |
|
"learning_rate": 9.999946921339963e-07, |
|
"loss": 1.5326, |
|
"step": 1014 |
|
}, |
|
{ |
|
"epoch": 0.9990157480314961, |
|
"grad_norm": 13.352030754089355, |
|
"learning_rate": 9.999905638068e-07, |
|
"loss": 1.0034, |
|
"step": 1015 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 17.30084228515625, |
|
"learning_rate": 9.999852559742985e-07, |
|
"loss": 0.8925, |
|
"step": 1016 |
|
}, |
|
{ |
|
"epoch": 1.000984251968504, |
|
"grad_norm": 13.423869132995605, |
|
"learning_rate": 9.999787686490555e-07, |
|
"loss": 1.1286, |
|
"step": 1017 |
|
}, |
|
{ |
|
"epoch": 1.0019685039370079, |
|
"grad_norm": 13.345012664794922, |
|
"learning_rate": 9.99971101846426e-07, |
|
"loss": 0.9836, |
|
"step": 1018 |
|
}, |
|
{ |
|
"epoch": 1.0029527559055118, |
|
"grad_norm": 10.137012481689453, |
|
"learning_rate": 9.999622555845573e-07, |
|
"loss": 0.6727, |
|
"step": 1019 |
|
}, |
|
{ |
|
"epoch": 1.0039370078740157, |
|
"grad_norm": 12.473076820373535, |
|
"learning_rate": 9.999522298843875e-07, |
|
"loss": 0.8101, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 1.0049212598425197, |
|
"grad_norm": 12.631579399108887, |
|
"learning_rate": 9.99941024769647e-07, |
|
"loss": 0.7747, |
|
"step": 1021 |
|
}, |
|
{ |
|
"epoch": 1.0059055118110236, |
|
"grad_norm": 12.579475402832031, |
|
"learning_rate": 9.999286402668578e-07, |
|
"loss": 0.6931, |
|
"step": 1022 |
|
}, |
|
{ |
|
"epoch": 1.0068897637795275, |
|
"grad_norm": 17.56043815612793, |
|
"learning_rate": 9.999150764053334e-07, |
|
"loss": 1.2261, |
|
"step": 1023 |
|
}, |
|
{ |
|
"epoch": 1.0078740157480315, |
|
"grad_norm": 13.309534072875977, |
|
"learning_rate": 9.999003332171784e-07, |
|
"loss": 0.7315, |
|
"step": 1024 |
|
}, |
|
{ |
|
"epoch": 1.0088582677165354, |
|
"grad_norm": 12.705358505249023, |
|
"learning_rate": 9.998844107372894e-07, |
|
"loss": 0.61, |
|
"step": 1025 |
|
}, |
|
{ |
|
"epoch": 1.0098425196850394, |
|
"grad_norm": 10.458514213562012, |
|
"learning_rate": 9.99867309003354e-07, |
|
"loss": 0.5402, |
|
"step": 1026 |
|
}, |
|
{ |
|
"epoch": 1.0108267716535433, |
|
"grad_norm": 17.223058700561523, |
|
"learning_rate": 9.998490280558511e-07, |
|
"loss": 1.3353, |
|
"step": 1027 |
|
}, |
|
{ |
|
"epoch": 1.0118110236220472, |
|
"grad_norm": 15.71506118774414, |
|
"learning_rate": 9.998295679380506e-07, |
|
"loss": 0.9483, |
|
"step": 1028 |
|
}, |
|
{ |
|
"epoch": 1.0127952755905512, |
|
"grad_norm": 16.565704345703125, |
|
"learning_rate": 9.998089286960135e-07, |
|
"loss": 0.9226, |
|
"step": 1029 |
|
}, |
|
{ |
|
"epoch": 1.013779527559055, |
|
"grad_norm": 8.595297813415527, |
|
"learning_rate": 9.99787110378592e-07, |
|
"loss": 0.6536, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 1.014763779527559, |
|
"grad_norm": 13.655412673950195, |
|
"learning_rate": 9.997641130374284e-07, |
|
"loss": 0.7343, |
|
"step": 1031 |
|
}, |
|
{ |
|
"epoch": 1.015748031496063, |
|
"grad_norm": 13.065082550048828, |
|
"learning_rate": 9.997399367269565e-07, |
|
"loss": 0.7647, |
|
"step": 1032 |
|
}, |
|
{ |
|
"epoch": 1.016732283464567, |
|
"grad_norm": 11.816627502441406, |
|
"learning_rate": 9.997145815044003e-07, |
|
"loss": 0.7541, |
|
"step": 1033 |
|
}, |
|
{ |
|
"epoch": 1.0177165354330708, |
|
"grad_norm": 15.631732940673828, |
|
"learning_rate": 9.99688047429774e-07, |
|
"loss": 1.1746, |
|
"step": 1034 |
|
}, |
|
{ |
|
"epoch": 1.0187007874015748, |
|
"grad_norm": 17.95444679260254, |
|
"learning_rate": 9.996603345658824e-07, |
|
"loss": 1.3935, |
|
"step": 1035 |
|
}, |
|
{ |
|
"epoch": 1.0196850393700787, |
|
"grad_norm": 15.705811500549316, |
|
"learning_rate": 9.996314429783204e-07, |
|
"loss": 1.5224, |
|
"step": 1036 |
|
}, |
|
{ |
|
"epoch": 1.0206692913385826, |
|
"grad_norm": 13.482087135314941, |
|
"learning_rate": 9.99601372735473e-07, |
|
"loss": 0.9067, |
|
"step": 1037 |
|
}, |
|
{ |
|
"epoch": 1.0216535433070866, |
|
"grad_norm": 10.162127494812012, |
|
"learning_rate": 9.995701239085143e-07, |
|
"loss": 0.7184, |
|
"step": 1038 |
|
}, |
|
{ |
|
"epoch": 1.0226377952755905, |
|
"grad_norm": 10.461275100708008, |
|
"learning_rate": 9.99537696571409e-07, |
|
"loss": 0.5991, |
|
"step": 1039 |
|
}, |
|
{ |
|
"epoch": 1.0236220472440944, |
|
"grad_norm": 11.486288070678711, |
|
"learning_rate": 9.995040908009107e-07, |
|
"loss": 0.8841, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 1.0246062992125984, |
|
"grad_norm": 10.825572967529297, |
|
"learning_rate": 9.994693066765622e-07, |
|
"loss": 0.7474, |
|
"step": 1041 |
|
}, |
|
{ |
|
"epoch": 1.0255905511811023, |
|
"grad_norm": 12.2081298828125, |
|
"learning_rate": 9.994333442806959e-07, |
|
"loss": 0.7444, |
|
"step": 1042 |
|
}, |
|
{ |
|
"epoch": 1.0265748031496063, |
|
"grad_norm": 15.263197898864746, |
|
"learning_rate": 9.993962036984325e-07, |
|
"loss": 1.4098, |
|
"step": 1043 |
|
}, |
|
{ |
|
"epoch": 1.0275590551181102, |
|
"grad_norm": 9.019082069396973, |
|
"learning_rate": 9.99357885017682e-07, |
|
"loss": 0.6033, |
|
"step": 1044 |
|
}, |
|
{ |
|
"epoch": 1.0285433070866141, |
|
"grad_norm": 14.380988121032715, |
|
"learning_rate": 9.993183883291425e-07, |
|
"loss": 0.93, |
|
"step": 1045 |
|
}, |
|
{ |
|
"epoch": 1.029527559055118, |
|
"grad_norm": 12.470325469970703, |
|
"learning_rate": 9.992777137263007e-07, |
|
"loss": 0.7308, |
|
"step": 1046 |
|
}, |
|
{ |
|
"epoch": 1.030511811023622, |
|
"grad_norm": 11.483294486999512, |
|
"learning_rate": 9.99235861305431e-07, |
|
"loss": 0.7339, |
|
"step": 1047 |
|
}, |
|
{ |
|
"epoch": 1.031496062992126, |
|
"grad_norm": 17.470186233520508, |
|
"learning_rate": 9.991928311655956e-07, |
|
"loss": 1.3459, |
|
"step": 1048 |
|
}, |
|
{ |
|
"epoch": 1.0324803149606299, |
|
"grad_norm": 12.41200065612793, |
|
"learning_rate": 9.991486234086447e-07, |
|
"loss": 0.6866, |
|
"step": 1049 |
|
}, |
|
{ |
|
"epoch": 1.0334645669291338, |
|
"grad_norm": 11.915864944458008, |
|
"learning_rate": 9.991032381392158e-07, |
|
"loss": 0.8323, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 1.0344488188976377, |
|
"grad_norm": 10.167206764221191, |
|
"learning_rate": 9.990566754647328e-07, |
|
"loss": 0.5301, |
|
"step": 1051 |
|
}, |
|
{ |
|
"epoch": 1.0354330708661417, |
|
"grad_norm": 15.526559829711914, |
|
"learning_rate": 9.990089354954073e-07, |
|
"loss": 1.1091, |
|
"step": 1052 |
|
}, |
|
{ |
|
"epoch": 1.0364173228346456, |
|
"grad_norm": 11.07171630859375, |
|
"learning_rate": 9.989600183442373e-07, |
|
"loss": 0.7301, |
|
"step": 1053 |
|
}, |
|
{ |
|
"epoch": 1.0374015748031495, |
|
"grad_norm": 13.301454544067383, |
|
"learning_rate": 9.98909924127007e-07, |
|
"loss": 1.1041, |
|
"step": 1054 |
|
}, |
|
{ |
|
"epoch": 1.0383858267716535, |
|
"grad_norm": 7.886115074157715, |
|
"learning_rate": 9.988586529622864e-07, |
|
"loss": 0.433, |
|
"step": 1055 |
|
}, |
|
{ |
|
"epoch": 1.0393700787401574, |
|
"grad_norm": 13.891656875610352, |
|
"learning_rate": 9.98806204971432e-07, |
|
"loss": 0.8808, |
|
"step": 1056 |
|
}, |
|
{ |
|
"epoch": 1.0403543307086613, |
|
"grad_norm": 10.688889503479004, |
|
"learning_rate": 9.987525802785847e-07, |
|
"loss": 0.5836, |
|
"step": 1057 |
|
}, |
|
{ |
|
"epoch": 1.0413385826771653, |
|
"grad_norm": 12.516069412231445, |
|
"learning_rate": 9.986977790106715e-07, |
|
"loss": 0.6459, |
|
"step": 1058 |
|
}, |
|
{ |
|
"epoch": 1.0423228346456692, |
|
"grad_norm": 11.766775131225586, |
|
"learning_rate": 9.986418012974042e-07, |
|
"loss": 0.4647, |
|
"step": 1059 |
|
}, |
|
{ |
|
"epoch": 1.0433070866141732, |
|
"grad_norm": 18.46944236755371, |
|
"learning_rate": 9.985846472712786e-07, |
|
"loss": 1.4988, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 1.044291338582677, |
|
"grad_norm": 14.120805740356445, |
|
"learning_rate": 9.985263170675756e-07, |
|
"loss": 1.2506, |
|
"step": 1061 |
|
}, |
|
{ |
|
"epoch": 1.045275590551181, |
|
"grad_norm": 11.723356246948242, |
|
"learning_rate": 9.98466810824359e-07, |
|
"loss": 0.5674, |
|
"step": 1062 |
|
}, |
|
{ |
|
"epoch": 1.046259842519685, |
|
"grad_norm": 11.574830055236816, |
|
"learning_rate": 9.984061286824774e-07, |
|
"loss": 0.8192, |
|
"step": 1063 |
|
}, |
|
{ |
|
"epoch": 1.047244094488189, |
|
"grad_norm": 19.33531379699707, |
|
"learning_rate": 9.983442707855616e-07, |
|
"loss": 1.443, |
|
"step": 1064 |
|
}, |
|
{ |
|
"epoch": 1.0482283464566928, |
|
"grad_norm": 9.310123443603516, |
|
"learning_rate": 9.982812372800262e-07, |
|
"loss": 0.4771, |
|
"step": 1065 |
|
}, |
|
{ |
|
"epoch": 1.0492125984251968, |
|
"grad_norm": 11.541061401367188, |
|
"learning_rate": 9.982170283150678e-07, |
|
"loss": 0.5886, |
|
"step": 1066 |
|
}, |
|
{ |
|
"epoch": 1.0501968503937007, |
|
"grad_norm": 11.022961616516113, |
|
"learning_rate": 9.981516440426659e-07, |
|
"loss": 0.9119, |
|
"step": 1067 |
|
}, |
|
{ |
|
"epoch": 1.0511811023622046, |
|
"grad_norm": 12.574899673461914, |
|
"learning_rate": 9.98085084617581e-07, |
|
"loss": 0.7875, |
|
"step": 1068 |
|
}, |
|
{ |
|
"epoch": 1.0521653543307086, |
|
"grad_norm": 10.970486640930176, |
|
"learning_rate": 9.980173501973558e-07, |
|
"loss": 0.7041, |
|
"step": 1069 |
|
}, |
|
{ |
|
"epoch": 1.0531496062992125, |
|
"grad_norm": 10.811155319213867, |
|
"learning_rate": 9.97948440942314e-07, |
|
"loss": 0.5232, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 1.0541338582677164, |
|
"grad_norm": 10.792261123657227, |
|
"learning_rate": 9.978783570155597e-07, |
|
"loss": 0.57, |
|
"step": 1071 |
|
}, |
|
{ |
|
"epoch": 1.0541338582677164, |
|
"eval_Qnli-dev_cosine_accuracy": 0.6640625, |
|
"eval_Qnli-dev_cosine_accuracy_threshold": 0.8653759956359863, |
|
"eval_Qnli-dev_cosine_ap": 0.6846632153971953, |
|
"eval_Qnli-dev_cosine_f1": 0.6816608996539792, |
|
"eval_Qnli-dev_cosine_f1_threshold": 0.7897687554359436, |
|
"eval_Qnli-dev_cosine_precision": 0.5760233918128655, |
|
"eval_Qnli-dev_cosine_recall": 0.8347457627118644, |
|
"eval_Qnli-dev_dot_accuracy": 0.66015625, |
|
"eval_Qnli-dev_dot_accuracy_threshold": 657.5789794921875, |
|
"eval_Qnli-dev_dot_ap": 0.6832739733257174, |
|
"eval_Qnli-dev_dot_f1": 0.6814310051107325, |
|
"eval_Qnli-dev_dot_f1_threshold": 598.9862670898438, |
|
"eval_Qnli-dev_dot_precision": 0.5698005698005698, |
|
"eval_Qnli-dev_dot_recall": 0.847457627118644, |
|
"eval_Qnli-dev_euclidean_accuracy": 0.6640625, |
|
"eval_Qnli-dev_euclidean_accuracy_threshold": 14.331533432006836, |
|
"eval_Qnli-dev_euclidean_ap": 0.6849916947033118, |
|
"eval_Qnli-dev_euclidean_f1": 0.6804123711340205, |
|
"eval_Qnli-dev_euclidean_f1_threshold": 17.989498138427734, |
|
"eval_Qnli-dev_euclidean_precision": 0.5722543352601156, |
|
"eval_Qnli-dev_euclidean_recall": 0.8389830508474576, |
|
"eval_Qnli-dev_manhattan_accuracy": 0.6796875, |
|
"eval_Qnli-dev_manhattan_accuracy_threshold": 294.7474670410156, |
|
"eval_Qnli-dev_manhattan_ap": 0.6961691800504137, |
|
"eval_Qnli-dev_manhattan_f1": 0.6860068259385665, |
|
"eval_Qnli-dev_manhattan_f1_threshold": 376.3381652832031, |
|
"eval_Qnli-dev_manhattan_precision": 0.5742857142857143, |
|
"eval_Qnli-dev_manhattan_recall": 0.8516949152542372, |
|
"eval_Qnli-dev_max_accuracy": 0.6796875, |
|
"eval_Qnli-dev_max_accuracy_threshold": 657.5789794921875, |
|
"eval_Qnli-dev_max_ap": 0.6961691800504137, |
|
"eval_Qnli-dev_max_f1": 0.6860068259385665, |
|
"eval_Qnli-dev_max_f1_threshold": 598.9862670898438, |
|
"eval_Qnli-dev_max_precision": 0.5760233918128655, |
|
"eval_Qnli-dev_max_recall": 0.8516949152542372, |
|
"eval_allNLI-dev_cosine_accuracy": 0.69140625, |
|
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9270304441452026, |
|
"eval_allNLI-dev_cosine_ap": 0.49228397783239236, |
|
"eval_allNLI-dev_cosine_f1": 0.5363984674329503, |
|
"eval_allNLI-dev_cosine_f1_threshold": 0.8337951898574829, |
|
"eval_allNLI-dev_cosine_precision": 0.40114613180515757, |
|
"eval_allNLI-dev_cosine_recall": 0.8092485549132948, |
|
"eval_allNLI-dev_dot_accuracy": 0.69140625, |
|
"eval_allNLI-dev_dot_accuracy_threshold": 709.9082641601562, |
|
"eval_allNLI-dev_dot_ap": 0.4925267288107693, |
|
"eval_allNLI-dev_dot_f1": 0.5384615384615384, |
|
"eval_allNLI-dev_dot_f1_threshold": 638.5916748046875, |
|
"eval_allNLI-dev_dot_precision": 0.4034582132564842, |
|
"eval_allNLI-dev_dot_recall": 0.8092485549132948, |
|
"eval_allNLI-dev_euclidean_accuracy": 0.69140625, |
|
"eval_allNLI-dev_euclidean_accuracy_threshold": 10.571220397949219, |
|
"eval_allNLI-dev_euclidean_ap": 0.4923798851447917, |
|
"eval_allNLI-dev_euclidean_f1": 0.5363984674329503, |
|
"eval_allNLI-dev_euclidean_f1_threshold": 15.952140808105469, |
|
"eval_allNLI-dev_euclidean_precision": 0.40114613180515757, |
|
"eval_allNLI-dev_euclidean_recall": 0.8092485549132948, |
|
"eval_allNLI-dev_manhattan_accuracy": 0.689453125, |
|
"eval_allNLI-dev_manhattan_accuracy_threshold": 214.40182495117188, |
|
"eval_allNLI-dev_manhattan_ap": 0.49086550398875856, |
|
"eval_allNLI-dev_manhattan_f1": 0.5391621129326047, |
|
"eval_allNLI-dev_manhattan_f1_threshold": 346.060302734375, |
|
"eval_allNLI-dev_manhattan_precision": 0.39361702127659576, |
|
"eval_allNLI-dev_manhattan_recall": 0.8554913294797688, |
|
"eval_allNLI-dev_max_accuracy": 0.69140625, |
|
"eval_allNLI-dev_max_accuracy_threshold": 709.9082641601562, |
|
"eval_allNLI-dev_max_ap": 0.4925267288107693, |
|
"eval_allNLI-dev_max_f1": 0.5391621129326047, |
|
"eval_allNLI-dev_max_f1_threshold": 638.5916748046875, |
|
"eval_allNLI-dev_max_precision": 0.4034582132564842, |
|
"eval_allNLI-dev_max_recall": 0.8554913294797688, |
|
"eval_loss": 1.1570727825164795, |
|
"eval_runtime": 52.6333, |
|
"eval_samples_per_second": 31.615, |
|
"eval_sequential_score": 0.6961691800504137, |
|
"eval_steps_per_second": 0.133, |
|
"eval_sts-test_pearson_cosine": 0.509950648911252, |
|
"eval_sts-test_pearson_dot": 0.5045563506807493, |
|
"eval_sts-test_pearson_euclidean": 0.5284216887479147, |
|
"eval_sts-test_pearson_manhattan": 0.5423369169040173, |
|
"eval_sts-test_pearson_max": 0.5423369169040173, |
|
"eval_sts-test_spearman_cosine": 0.5372664597807683, |
|
"eval_sts-test_spearman_dot": 0.5314762283231191, |
|
"eval_sts-test_spearman_euclidean": 0.5375220789858359, |
|
"eval_sts-test_spearman_manhattan": 0.5495945820201323, |
|
"eval_sts-test_spearman_max": 0.5495945820201323, |
|
"step": 1071 |
|
}, |
|
{ |
|
"epoch": 1.0551181102362204, |
|
"grad_norm": 11.420138359069824, |
|
"learning_rate": 9.978070985829782e-07, |
|
"loss": 0.8948, |
|
"step": 1072 |
|
}, |
|
{ |
|
"epoch": 1.0561023622047243, |
|
"grad_norm": 14.187203407287598, |
|
"learning_rate": 9.97734665813234e-07, |
|
"loss": 1.0059, |
|
"step": 1073 |
|
}, |
|
{ |
|
"epoch": 1.0570866141732282, |
|
"grad_norm": 11.898282051086426, |
|
"learning_rate": 9.976610588777717e-07, |
|
"loss": 0.9796, |
|
"step": 1074 |
|
}, |
|
{ |
|
"epoch": 1.0580708661417322, |
|
"grad_norm": 12.518900871276855, |
|
"learning_rate": 9.975862779508149e-07, |
|
"loss": 0.9632, |
|
"step": 1075 |
|
}, |
|
{ |
|
"epoch": 1.0590551181102361, |
|
"grad_norm": 15.27397632598877, |
|
"learning_rate": 9.975103232093656e-07, |
|
"loss": 1.1254, |
|
"step": 1076 |
|
}, |
|
{ |
|
"epoch": 1.06003937007874, |
|
"grad_norm": 12.165975570678711, |
|
"learning_rate": 9.974331948332048e-07, |
|
"loss": 0.8915, |
|
"step": 1077 |
|
}, |
|
{ |
|
"epoch": 1.061023622047244, |
|
"grad_norm": 11.72553825378418, |
|
"learning_rate": 9.97354893004891e-07, |
|
"loss": 0.5213, |
|
"step": 1078 |
|
}, |
|
{ |
|
"epoch": 1.0620078740157481, |
|
"grad_norm": 10.605430603027344, |
|
"learning_rate": 9.972754179097602e-07, |
|
"loss": 0.6755, |
|
"step": 1079 |
|
}, |
|
{ |
|
"epoch": 1.0629921259842519, |
|
"grad_norm": 11.675688743591309, |
|
"learning_rate": 9.971947697359257e-07, |
|
"loss": 0.6665, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 1.063976377952756, |
|
"grad_norm": 11.114697456359863, |
|
"learning_rate": 9.971129486742777e-07, |
|
"loss": 0.729, |
|
"step": 1081 |
|
}, |
|
{ |
|
"epoch": 1.0649606299212597, |
|
"grad_norm": 13.48147201538086, |
|
"learning_rate": 9.970299549184814e-07, |
|
"loss": 0.9047, |
|
"step": 1082 |
|
}, |
|
{ |
|
"epoch": 1.0659448818897639, |
|
"grad_norm": 13.161925315856934, |
|
"learning_rate": 9.969457886649791e-07, |
|
"loss": 1.0514, |
|
"step": 1083 |
|
}, |
|
{ |
|
"epoch": 1.0669291338582678, |
|
"grad_norm": 18.14025115966797, |
|
"learning_rate": 9.968604501129877e-07, |
|
"loss": 1.6729, |
|
"step": 1084 |
|
}, |
|
{ |
|
"epoch": 1.0679133858267718, |
|
"grad_norm": 15.771471977233887, |
|
"learning_rate": 9.967739394644982e-07, |
|
"loss": 0.8698, |
|
"step": 1085 |
|
}, |
|
{ |
|
"epoch": 1.0688976377952757, |
|
"grad_norm": 13.432958602905273, |
|
"learning_rate": 9.966862569242773e-07, |
|
"loss": 1.1209, |
|
"step": 1086 |
|
}, |
|
{ |
|
"epoch": 1.0698818897637796, |
|
"grad_norm": 13.478239059448242, |
|
"learning_rate": 9.965974026998642e-07, |
|
"loss": 0.9425, |
|
"step": 1087 |
|
}, |
|
{ |
|
"epoch": 1.0708661417322836, |
|
"grad_norm": 12.611392974853516, |
|
"learning_rate": 9.965073770015722e-07, |
|
"loss": 0.7151, |
|
"step": 1088 |
|
}, |
|
{ |
|
"epoch": 1.0718503937007875, |
|
"grad_norm": 12.270779609680176, |
|
"learning_rate": 9.964161800424873e-07, |
|
"loss": 0.7336, |
|
"step": 1089 |
|
}, |
|
{ |
|
"epoch": 1.0728346456692914, |
|
"grad_norm": 18.48605728149414, |
|
"learning_rate": 9.963238120384677e-07, |
|
"loss": 1.0971, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 1.0738188976377954, |
|
"grad_norm": 11.107337951660156, |
|
"learning_rate": 9.962302732081432e-07, |
|
"loss": 0.5334, |
|
"step": 1091 |
|
}, |
|
{ |
|
"epoch": 1.0748031496062993, |
|
"grad_norm": 13.366519927978516, |
|
"learning_rate": 9.961355637729153e-07, |
|
"loss": 1.0171, |
|
"step": 1092 |
|
}, |
|
{ |
|
"epoch": 1.0757874015748032, |
|
"grad_norm": 12.98841381072998, |
|
"learning_rate": 9.960396839569557e-07, |
|
"loss": 1.0153, |
|
"step": 1093 |
|
}, |
|
{ |
|
"epoch": 1.0767716535433072, |
|
"grad_norm": 12.056806564331055, |
|
"learning_rate": 9.95942633987207e-07, |
|
"loss": 0.6662, |
|
"step": 1094 |
|
}, |
|
{ |
|
"epoch": 1.077755905511811, |
|
"grad_norm": 11.31603717803955, |
|
"learning_rate": 9.95844414093381e-07, |
|
"loss": 0.5451, |
|
"step": 1095 |
|
}, |
|
{ |
|
"epoch": 1.078740157480315, |
|
"grad_norm": 11.599329948425293, |
|
"learning_rate": 9.957450245079588e-07, |
|
"loss": 0.7743, |
|
"step": 1096 |
|
}, |
|
{ |
|
"epoch": 1.079724409448819, |
|
"grad_norm": 17.17649269104004, |
|
"learning_rate": 9.9564446546619e-07, |
|
"loss": 0.9187, |
|
"step": 1097 |
|
}, |
|
{ |
|
"epoch": 1.080708661417323, |
|
"grad_norm": 14.856340408325195, |
|
"learning_rate": 9.955427372060928e-07, |
|
"loss": 1.1919, |
|
"step": 1098 |
|
}, |
|
{ |
|
"epoch": 1.0816929133858268, |
|
"grad_norm": 12.51814079284668, |
|
"learning_rate": 9.954398399684515e-07, |
|
"loss": 0.9405, |
|
"step": 1099 |
|
}, |
|
{ |
|
"epoch": 1.0826771653543308, |
|
"grad_norm": 11.004323959350586, |
|
"learning_rate": 9.953357739968192e-07, |
|
"loss": 0.5361, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.0836614173228347, |
|
"grad_norm": 12.910338401794434, |
|
"learning_rate": 9.952305395375137e-07, |
|
"loss": 1.233, |
|
"step": 1101 |
|
}, |
|
{ |
|
"epoch": 1.0846456692913387, |
|
"grad_norm": 13.418724060058594, |
|
"learning_rate": 9.951241368396194e-07, |
|
"loss": 0.8222, |
|
"step": 1102 |
|
}, |
|
{ |
|
"epoch": 1.0856299212598426, |
|
"grad_norm": 11.835745811462402, |
|
"learning_rate": 9.950165661549856e-07, |
|
"loss": 0.906, |
|
"step": 1103 |
|
}, |
|
{ |
|
"epoch": 1.0866141732283465, |
|
"grad_norm": 10.636372566223145, |
|
"learning_rate": 9.949078277382262e-07, |
|
"loss": 0.62, |
|
"step": 1104 |
|
}, |
|
{ |
|
"epoch": 1.0875984251968505, |
|
"grad_norm": 13.1072416305542, |
|
"learning_rate": 9.947979218467193e-07, |
|
"loss": 0.887, |
|
"step": 1105 |
|
}, |
|
{ |
|
"epoch": 1.0885826771653544, |
|
"grad_norm": 9.505353927612305, |
|
"learning_rate": 9.946868487406057e-07, |
|
"loss": 0.6135, |
|
"step": 1106 |
|
}, |
|
{ |
|
"epoch": 1.0895669291338583, |
|
"grad_norm": 8.03498649597168, |
|
"learning_rate": 9.945746086827896e-07, |
|
"loss": 0.2279, |
|
"step": 1107 |
|
}, |
|
{ |
|
"epoch": 1.0905511811023623, |
|
"grad_norm": 12.267210006713867, |
|
"learning_rate": 9.94461201938937e-07, |
|
"loss": 1.0091, |
|
"step": 1108 |
|
}, |
|
{ |
|
"epoch": 1.0915354330708662, |
|
"grad_norm": 13.677517890930176, |
|
"learning_rate": 9.943466287774756e-07, |
|
"loss": 1.2457, |
|
"step": 1109 |
|
}, |
|
{ |
|
"epoch": 1.0925196850393701, |
|
"grad_norm": 10.236763954162598, |
|
"learning_rate": 9.942308894695932e-07, |
|
"loss": 0.6233, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 1.093503937007874, |
|
"grad_norm": 8.855351448059082, |
|
"learning_rate": 9.941139842892391e-07, |
|
"loss": 0.6004, |
|
"step": 1111 |
|
}, |
|
{ |
|
"epoch": 1.094488188976378, |
|
"grad_norm": 13.564655303955078, |
|
"learning_rate": 9.939959135131214e-07, |
|
"loss": 1.0716, |
|
"step": 1112 |
|
}, |
|
{ |
|
"epoch": 1.095472440944882, |
|
"grad_norm": 10.523872375488281, |
|
"learning_rate": 9.938766774207064e-07, |
|
"loss": 0.5854, |
|
"step": 1113 |
|
}, |
|
{ |
|
"epoch": 1.0964566929133859, |
|
"grad_norm": 12.336915016174316, |
|
"learning_rate": 9.937562762942202e-07, |
|
"loss": 0.5708, |
|
"step": 1114 |
|
}, |
|
{ |
|
"epoch": 1.0974409448818898, |
|
"grad_norm": 15.987162590026855, |
|
"learning_rate": 9.936347104186452e-07, |
|
"loss": 1.0223, |
|
"step": 1115 |
|
}, |
|
{ |
|
"epoch": 1.0984251968503937, |
|
"grad_norm": 11.978333473205566, |
|
"learning_rate": 9.935119800817215e-07, |
|
"loss": 0.6434, |
|
"step": 1116 |
|
}, |
|
{ |
|
"epoch": 1.0994094488188977, |
|
"grad_norm": 8.372782707214355, |
|
"learning_rate": 9.933880855739446e-07, |
|
"loss": 0.363, |
|
"step": 1117 |
|
}, |
|
{ |
|
"epoch": 1.1003937007874016, |
|
"grad_norm": 17.084253311157227, |
|
"learning_rate": 9.932630271885662e-07, |
|
"loss": 1.2779, |
|
"step": 1118 |
|
}, |
|
{ |
|
"epoch": 1.1013779527559056, |
|
"grad_norm": 13.697333335876465, |
|
"learning_rate": 9.93136805221593e-07, |
|
"loss": 0.8207, |
|
"step": 1119 |
|
}, |
|
{ |
|
"epoch": 1.1023622047244095, |
|
"grad_norm": 12.337996482849121, |
|
"learning_rate": 9.930094199717848e-07, |
|
"loss": 0.6738, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 1.1033464566929134, |
|
"grad_norm": 15.996990203857422, |
|
"learning_rate": 9.92880871740656e-07, |
|
"loss": 1.0475, |
|
"step": 1121 |
|
}, |
|
{ |
|
"epoch": 1.1043307086614174, |
|
"grad_norm": 12.84057331085205, |
|
"learning_rate": 9.927511608324732e-07, |
|
"loss": 0.9189, |
|
"step": 1122 |
|
}, |
|
{ |
|
"epoch": 1.1053149606299213, |
|
"grad_norm": 18.542404174804688, |
|
"learning_rate": 9.926202875542546e-07, |
|
"loss": 1.0991, |
|
"step": 1123 |
|
}, |
|
{ |
|
"epoch": 1.1062992125984252, |
|
"grad_norm": 11.175984382629395, |
|
"learning_rate": 9.924882522157706e-07, |
|
"loss": 0.7348, |
|
"step": 1124 |
|
}, |
|
{ |
|
"epoch": 1.1072834645669292, |
|
"grad_norm": 11.271822929382324, |
|
"learning_rate": 9.923550551295411e-07, |
|
"loss": 0.7178, |
|
"step": 1125 |
|
}, |
|
{ |
|
"epoch": 1.108267716535433, |
|
"grad_norm": 15.300597190856934, |
|
"learning_rate": 9.922206966108367e-07, |
|
"loss": 0.893, |
|
"step": 1126 |
|
}, |
|
{ |
|
"epoch": 1.109251968503937, |
|
"grad_norm": 12.767745018005371, |
|
"learning_rate": 9.920851769776761e-07, |
|
"loss": 0.7549, |
|
"step": 1127 |
|
}, |
|
{ |
|
"epoch": 1.110236220472441, |
|
"grad_norm": 15.920074462890625, |
|
"learning_rate": 9.919484965508276e-07, |
|
"loss": 1.4074, |
|
"step": 1128 |
|
}, |
|
{ |
|
"epoch": 1.111220472440945, |
|
"grad_norm": 15.358880996704102, |
|
"learning_rate": 9.918106556538057e-07, |
|
"loss": 0.8321, |
|
"step": 1129 |
|
}, |
|
{ |
|
"epoch": 1.1122047244094488, |
|
"grad_norm": 15.989846229553223, |
|
"learning_rate": 9.91671654612872e-07, |
|
"loss": 1.0289, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 1.1131889763779528, |
|
"grad_norm": 15.226240158081055, |
|
"learning_rate": 9.91531493757035e-07, |
|
"loss": 0.8296, |
|
"step": 1131 |
|
}, |
|
{ |
|
"epoch": 1.1141732283464567, |
|
"grad_norm": 12.149371147155762, |
|
"learning_rate": 9.913901734180476e-07, |
|
"loss": 0.7587, |
|
"step": 1132 |
|
}, |
|
{ |
|
"epoch": 1.1151574803149606, |
|
"grad_norm": 9.797683715820312, |
|
"learning_rate": 9.912476939304069e-07, |
|
"loss": 0.6198, |
|
"step": 1133 |
|
}, |
|
{ |
|
"epoch": 1.1161417322834646, |
|
"grad_norm": 12.730241775512695, |
|
"learning_rate": 9.911040556313543e-07, |
|
"loss": 0.7474, |
|
"step": 1134 |
|
}, |
|
{ |
|
"epoch": 1.1171259842519685, |
|
"grad_norm": 11.825194358825684, |
|
"learning_rate": 9.909592588608738e-07, |
|
"loss": 0.719, |
|
"step": 1135 |
|
}, |
|
{ |
|
"epoch": 1.1181102362204725, |
|
"grad_norm": 10.120687484741211, |
|
"learning_rate": 9.908133039616911e-07, |
|
"loss": 0.517, |
|
"step": 1136 |
|
}, |
|
{ |
|
"epoch": 1.1190944881889764, |
|
"grad_norm": 9.076939582824707, |
|
"learning_rate": 9.90666191279274e-07, |
|
"loss": 0.4452, |
|
"step": 1137 |
|
}, |
|
{ |
|
"epoch": 1.1200787401574803, |
|
"grad_norm": 17.70829200744629, |
|
"learning_rate": 9.905179211618296e-07, |
|
"loss": 1.1683, |
|
"step": 1138 |
|
}, |
|
{ |
|
"epoch": 1.1210629921259843, |
|
"grad_norm": 11.209772109985352, |
|
"learning_rate": 9.903684939603054e-07, |
|
"loss": 0.7705, |
|
"step": 1139 |
|
}, |
|
{ |
|
"epoch": 1.1220472440944882, |
|
"grad_norm": 12.0265531539917, |
|
"learning_rate": 9.902179100283875e-07, |
|
"loss": 0.7873, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 1.1230314960629921, |
|
"grad_norm": 12.67635440826416, |
|
"learning_rate": 9.900661697224995e-07, |
|
"loss": 0.7939, |
|
"step": 1141 |
|
}, |
|
{ |
|
"epoch": 1.124015748031496, |
|
"grad_norm": 15.988715171813965, |
|
"learning_rate": 9.899132734018024e-07, |
|
"loss": 0.6513, |
|
"step": 1142 |
|
}, |
|
{ |
|
"epoch": 1.125, |
|
"grad_norm": 16.87392234802246, |
|
"learning_rate": 9.897592214281936e-07, |
|
"loss": 1.4787, |
|
"step": 1143 |
|
}, |
|
{ |
|
"epoch": 1.125984251968504, |
|
"grad_norm": 10.136147499084473, |
|
"learning_rate": 9.896040141663054e-07, |
|
"loss": 0.7314, |
|
"step": 1144 |
|
}, |
|
{ |
|
"epoch": 1.1269685039370079, |
|
"grad_norm": 12.934366226196289, |
|
"learning_rate": 9.894476519835052e-07, |
|
"loss": 0.4927, |
|
"step": 1145 |
|
}, |
|
{ |
|
"epoch": 1.1279527559055118, |
|
"grad_norm": 12.322945594787598, |
|
"learning_rate": 9.892901352498932e-07, |
|
"loss": 0.9718, |
|
"step": 1146 |
|
}, |
|
{ |
|
"epoch": 1.1289370078740157, |
|
"grad_norm": 12.120319366455078, |
|
"learning_rate": 9.891314643383037e-07, |
|
"loss": 0.6738, |
|
"step": 1147 |
|
}, |
|
{ |
|
"epoch": 1.1299212598425197, |
|
"grad_norm": 13.021424293518066, |
|
"learning_rate": 9.88971639624301e-07, |
|
"loss": 1.2356, |
|
"step": 1148 |
|
}, |
|
{ |
|
"epoch": 1.1309055118110236, |
|
"grad_norm": 12.181797981262207, |
|
"learning_rate": 9.888106614861823e-07, |
|
"loss": 0.902, |
|
"step": 1149 |
|
}, |
|
{ |
|
"epoch": 1.1318897637795275, |
|
"grad_norm": 11.161940574645996, |
|
"learning_rate": 9.886485303049734e-07, |
|
"loss": 0.5957, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 1.1328740157480315, |
|
"grad_norm": 12.929612159729004, |
|
"learning_rate": 9.884852464644305e-07, |
|
"loss": 1.1595, |
|
"step": 1151 |
|
}, |
|
{ |
|
"epoch": 1.1338582677165354, |
|
"grad_norm": 12.247835159301758, |
|
"learning_rate": 9.883208103510367e-07, |
|
"loss": 0.9146, |
|
"step": 1152 |
|
}, |
|
{ |
|
"epoch": 1.1348425196850394, |
|
"grad_norm": 10.77420425415039, |
|
"learning_rate": 9.88155222354004e-07, |
|
"loss": 0.5402, |
|
"step": 1153 |
|
}, |
|
{ |
|
"epoch": 1.1358267716535433, |
|
"grad_norm": 10.38335132598877, |
|
"learning_rate": 9.879884828652698e-07, |
|
"loss": 0.6177, |
|
"step": 1154 |
|
}, |
|
{ |
|
"epoch": 1.1368110236220472, |
|
"grad_norm": 11.626049995422363, |
|
"learning_rate": 9.878205922794972e-07, |
|
"loss": 0.6705, |
|
"step": 1155 |
|
}, |
|
{ |
|
"epoch": 1.1377952755905512, |
|
"grad_norm": 10.968165397644043, |
|
"learning_rate": 9.876515509940742e-07, |
|
"loss": 0.603, |
|
"step": 1156 |
|
}, |
|
{ |
|
"epoch": 1.138779527559055, |
|
"grad_norm": 13.380332946777344, |
|
"learning_rate": 9.874813594091122e-07, |
|
"loss": 0.733, |
|
"step": 1157 |
|
}, |
|
{ |
|
"epoch": 1.139763779527559, |
|
"grad_norm": 15.77918815612793, |
|
"learning_rate": 9.873100179274453e-07, |
|
"loss": 1.1947, |
|
"step": 1158 |
|
}, |
|
{ |
|
"epoch": 1.140748031496063, |
|
"grad_norm": 13.412628173828125, |
|
"learning_rate": 9.871375269546292e-07, |
|
"loss": 0.6454, |
|
"step": 1159 |
|
}, |
|
{ |
|
"epoch": 1.141732283464567, |
|
"grad_norm": 18.430992126464844, |
|
"learning_rate": 9.869638868989407e-07, |
|
"loss": 1.0318, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 1.1427165354330708, |
|
"grad_norm": 15.241898536682129, |
|
"learning_rate": 9.867890981713764e-07, |
|
"loss": 1.0791, |
|
"step": 1161 |
|
}, |
|
{ |
|
"epoch": 1.1437007874015748, |
|
"grad_norm": 13.973746299743652, |
|
"learning_rate": 9.866131611856514e-07, |
|
"loss": 0.928, |
|
"step": 1162 |
|
}, |
|
{ |
|
"epoch": 1.1446850393700787, |
|
"grad_norm": 8.853302955627441, |
|
"learning_rate": 9.864360763581994e-07, |
|
"loss": 0.5488, |
|
"step": 1163 |
|
}, |
|
{ |
|
"epoch": 1.1456692913385826, |
|
"grad_norm": 11.555764198303223, |
|
"learning_rate": 9.862578441081699e-07, |
|
"loss": 0.7937, |
|
"step": 1164 |
|
}, |
|
{ |
|
"epoch": 1.1466535433070866, |
|
"grad_norm": 14.18860149383545, |
|
"learning_rate": 9.860784648574288e-07, |
|
"loss": 0.6262, |
|
"step": 1165 |
|
}, |
|
{ |
|
"epoch": 1.1476377952755905, |
|
"grad_norm": 11.897910118103027, |
|
"learning_rate": 9.858979390305573e-07, |
|
"loss": 0.6925, |
|
"step": 1166 |
|
}, |
|
{ |
|
"epoch": 1.1486220472440944, |
|
"grad_norm": 10.954864501953125, |
|
"learning_rate": 9.857162670548502e-07, |
|
"loss": 0.6949, |
|
"step": 1167 |
|
}, |
|
{ |
|
"epoch": 1.1496062992125984, |
|
"grad_norm": 8.025981903076172, |
|
"learning_rate": 9.855334493603147e-07, |
|
"loss": 0.2552, |
|
"step": 1168 |
|
}, |
|
{ |
|
"epoch": 1.1505905511811023, |
|
"grad_norm": 13.133017539978027, |
|
"learning_rate": 9.853494863796705e-07, |
|
"loss": 0.7877, |
|
"step": 1169 |
|
}, |
|
{ |
|
"epoch": 1.1515748031496063, |
|
"grad_norm": 13.997974395751953, |
|
"learning_rate": 9.851643785483478e-07, |
|
"loss": 0.7581, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 1.1525590551181102, |
|
"grad_norm": 10.509552001953125, |
|
"learning_rate": 9.849781263044867e-07, |
|
"loss": 0.4293, |
|
"step": 1171 |
|
}, |
|
{ |
|
"epoch": 1.1535433070866141, |
|
"grad_norm": 8.699508666992188, |
|
"learning_rate": 9.847907300889358e-07, |
|
"loss": 0.3826, |
|
"step": 1172 |
|
}, |
|
{ |
|
"epoch": 1.154527559055118, |
|
"grad_norm": 14.012443542480469, |
|
"learning_rate": 9.846021903452519e-07, |
|
"loss": 0.9302, |
|
"step": 1173 |
|
}, |
|
{ |
|
"epoch": 1.155511811023622, |
|
"grad_norm": 13.481894493103027, |
|
"learning_rate": 9.844125075196983e-07, |
|
"loss": 0.9708, |
|
"step": 1174 |
|
}, |
|
{ |
|
"epoch": 1.156496062992126, |
|
"grad_norm": 13.585330963134766, |
|
"learning_rate": 9.842216820612438e-07, |
|
"loss": 0.4477, |
|
"step": 1175 |
|
}, |
|
{ |
|
"epoch": 1.1574803149606299, |
|
"grad_norm": 13.615195274353027, |
|
"learning_rate": 9.840297144215615e-07, |
|
"loss": 0.7183, |
|
"step": 1176 |
|
}, |
|
{ |
|
"epoch": 1.1584645669291338, |
|
"grad_norm": 11.44372272491455, |
|
"learning_rate": 9.838366050550287e-07, |
|
"loss": 0.6087, |
|
"step": 1177 |
|
}, |
|
{ |
|
"epoch": 1.1594488188976377, |
|
"grad_norm": 15.111401557922363, |
|
"learning_rate": 9.836423544187245e-07, |
|
"loss": 0.8324, |
|
"step": 1178 |
|
}, |
|
{ |
|
"epoch": 1.1604330708661417, |
|
"grad_norm": 10.387323379516602, |
|
"learning_rate": 9.834469629724293e-07, |
|
"loss": 0.3906, |
|
"step": 1179 |
|
}, |
|
{ |
|
"epoch": 1.1614173228346456, |
|
"grad_norm": 19.592727661132812, |
|
"learning_rate": 9.832504311786238e-07, |
|
"loss": 1.5957, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 1.1624015748031495, |
|
"grad_norm": 16.09340476989746, |
|
"learning_rate": 9.830527595024885e-07, |
|
"loss": 1.2033, |
|
"step": 1181 |
|
}, |
|
{ |
|
"epoch": 1.1633858267716535, |
|
"grad_norm": 7.257636070251465, |
|
"learning_rate": 9.82853948411901e-07, |
|
"loss": 0.473, |
|
"step": 1182 |
|
}, |
|
{ |
|
"epoch": 1.1643700787401574, |
|
"grad_norm": 9.731563568115234, |
|
"learning_rate": 9.826539983774365e-07, |
|
"loss": 0.6048, |
|
"step": 1183 |
|
}, |
|
{ |
|
"epoch": 1.1653543307086613, |
|
"grad_norm": 13.61319637298584, |
|
"learning_rate": 9.824529098723657e-07, |
|
"loss": 0.861, |
|
"step": 1184 |
|
}, |
|
{ |
|
"epoch": 1.1663385826771653, |
|
"grad_norm": 10.9859619140625, |
|
"learning_rate": 9.822506833726537e-07, |
|
"loss": 0.9065, |
|
"step": 1185 |
|
}, |
|
{ |
|
"epoch": 1.1673228346456692, |
|
"grad_norm": 13.370589256286621, |
|
"learning_rate": 9.8204731935696e-07, |
|
"loss": 0.6253, |
|
"step": 1186 |
|
}, |
|
{ |
|
"epoch": 1.1683070866141732, |
|
"grad_norm": 16.47172737121582, |
|
"learning_rate": 9.818428183066356e-07, |
|
"loss": 1.227, |
|
"step": 1187 |
|
}, |
|
{ |
|
"epoch": 1.169291338582677, |
|
"grad_norm": 12.20095443725586, |
|
"learning_rate": 9.816371807057237e-07, |
|
"loss": 0.7943, |
|
"step": 1188 |
|
}, |
|
{ |
|
"epoch": 1.170275590551181, |
|
"grad_norm": 15.804914474487305, |
|
"learning_rate": 9.81430407040957e-07, |
|
"loss": 0.6892, |
|
"step": 1189 |
|
}, |
|
{ |
|
"epoch": 1.171259842519685, |
|
"grad_norm": 15.404899597167969, |
|
"learning_rate": 9.812224978017573e-07, |
|
"loss": 1.0856, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 1.172244094488189, |
|
"grad_norm": 14.905608177185059, |
|
"learning_rate": 9.810134534802346e-07, |
|
"loss": 0.7547, |
|
"step": 1191 |
|
}, |
|
{ |
|
"epoch": 1.1732283464566928, |
|
"grad_norm": 8.873392105102539, |
|
"learning_rate": 9.80803274571185e-07, |
|
"loss": 0.4079, |
|
"step": 1192 |
|
}, |
|
{ |
|
"epoch": 1.1742125984251968, |
|
"grad_norm": 15.257758140563965, |
|
"learning_rate": 9.80591961572091e-07, |
|
"loss": 0.6927, |
|
"step": 1193 |
|
}, |
|
{ |
|
"epoch": 1.1751968503937007, |
|
"grad_norm": 16.975690841674805, |
|
"learning_rate": 9.803795149831183e-07, |
|
"loss": 1.0321, |
|
"step": 1194 |
|
}, |
|
{ |
|
"epoch": 1.1761811023622046, |
|
"grad_norm": 18.011857986450195, |
|
"learning_rate": 9.801659353071165e-07, |
|
"loss": 1.1502, |
|
"step": 1195 |
|
}, |
|
{ |
|
"epoch": 1.1771653543307086, |
|
"grad_norm": 15.919508934020996, |
|
"learning_rate": 9.799512230496172e-07, |
|
"loss": 0.9344, |
|
"step": 1196 |
|
}, |
|
{ |
|
"epoch": 1.1781496062992125, |
|
"grad_norm": 16.511510848999023, |
|
"learning_rate": 9.79735378718832e-07, |
|
"loss": 1.0466, |
|
"step": 1197 |
|
}, |
|
{ |
|
"epoch": 1.1791338582677164, |
|
"grad_norm": 13.429428100585938, |
|
"learning_rate": 9.795184028256534e-07, |
|
"loss": 0.6004, |
|
"step": 1198 |
|
}, |
|
{ |
|
"epoch": 1.1801181102362204, |
|
"grad_norm": 15.42149543762207, |
|
"learning_rate": 9.793002958836506e-07, |
|
"loss": 0.7064, |
|
"step": 1199 |
|
}, |
|
{ |
|
"epoch": 1.1811023622047245, |
|
"grad_norm": 16.78218650817871, |
|
"learning_rate": 9.790810584090712e-07, |
|
"loss": 0.659, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.1820866141732282, |
|
"grad_norm": 12.032279968261719, |
|
"learning_rate": 9.78860690920838e-07, |
|
"loss": 0.6445, |
|
"step": 1201 |
|
}, |
|
{ |
|
"epoch": 1.1830708661417324, |
|
"grad_norm": 17.824844360351562, |
|
"learning_rate": 9.78639193940549e-07, |
|
"loss": 1.1254, |
|
"step": 1202 |
|
}, |
|
{ |
|
"epoch": 1.1840551181102361, |
|
"grad_norm": 11.511914253234863, |
|
"learning_rate": 9.784165679924752e-07, |
|
"loss": 0.7727, |
|
"step": 1203 |
|
}, |
|
{ |
|
"epoch": 1.1850393700787403, |
|
"grad_norm": 14.361802101135254, |
|
"learning_rate": 9.781928136035604e-07, |
|
"loss": 0.8912, |
|
"step": 1204 |
|
}, |
|
{ |
|
"epoch": 1.186023622047244, |
|
"grad_norm": 15.696497917175293, |
|
"learning_rate": 9.779679313034182e-07, |
|
"loss": 1.3996, |
|
"step": 1205 |
|
}, |
|
{ |
|
"epoch": 1.1870078740157481, |
|
"grad_norm": 12.164986610412598, |
|
"learning_rate": 9.777419216243333e-07, |
|
"loss": 0.6232, |
|
"step": 1206 |
|
}, |
|
{ |
|
"epoch": 1.1879921259842519, |
|
"grad_norm": 18.16672134399414, |
|
"learning_rate": 9.77514785101258e-07, |
|
"loss": 1.2074, |
|
"step": 1207 |
|
}, |
|
{ |
|
"epoch": 1.188976377952756, |
|
"grad_norm": 10.229595184326172, |
|
"learning_rate": 9.77286522271812e-07, |
|
"loss": 0.7199, |
|
"step": 1208 |
|
}, |
|
{ |
|
"epoch": 1.1899606299212597, |
|
"grad_norm": 11.71091079711914, |
|
"learning_rate": 9.77057133676281e-07, |
|
"loss": 0.5826, |
|
"step": 1209 |
|
}, |
|
{ |
|
"epoch": 1.1909448818897639, |
|
"grad_norm": 11.842780113220215, |
|
"learning_rate": 9.768266198576147e-07, |
|
"loss": 0.7618, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 1.1919291338582676, |
|
"grad_norm": 18.636600494384766, |
|
"learning_rate": 9.765949813614273e-07, |
|
"loss": 1.1995, |
|
"step": 1211 |
|
}, |
|
{ |
|
"epoch": 1.1929133858267718, |
|
"grad_norm": 11.783096313476562, |
|
"learning_rate": 9.763622187359942e-07, |
|
"loss": 0.6375, |
|
"step": 1212 |
|
}, |
|
{ |
|
"epoch": 1.1938976377952755, |
|
"grad_norm": 13.159656524658203, |
|
"learning_rate": 9.761283325322515e-07, |
|
"loss": 0.5623, |
|
"step": 1213 |
|
}, |
|
{ |
|
"epoch": 1.1948818897637796, |
|
"grad_norm": 13.676416397094727, |
|
"learning_rate": 9.758933233037952e-07, |
|
"loss": 0.7756, |
|
"step": 1214 |
|
}, |
|
{ |
|
"epoch": 1.1958661417322836, |
|
"grad_norm": 12.268596649169922, |
|
"learning_rate": 9.756571916068793e-07, |
|
"loss": 0.8011, |
|
"step": 1215 |
|
}, |
|
{ |
|
"epoch": 1.1968503937007875, |
|
"grad_norm": 9.940037727355957, |
|
"learning_rate": 9.754199380004145e-07, |
|
"loss": 0.4404, |
|
"step": 1216 |
|
}, |
|
{ |
|
"epoch": 1.1978346456692914, |
|
"grad_norm": 11.429154396057129, |
|
"learning_rate": 9.75181563045967e-07, |
|
"loss": 0.8401, |
|
"step": 1217 |
|
}, |
|
{ |
|
"epoch": 1.1988188976377954, |
|
"grad_norm": 11.402060508728027, |
|
"learning_rate": 9.749420673077574e-07, |
|
"loss": 0.8971, |
|
"step": 1218 |
|
}, |
|
{ |
|
"epoch": 1.1998031496062993, |
|
"grad_norm": 12.799181938171387, |
|
"learning_rate": 9.747014513526588e-07, |
|
"loss": 1.0111, |
|
"step": 1219 |
|
}, |
|
{ |
|
"epoch": 1.2007874015748032, |
|
"grad_norm": 10.88442611694336, |
|
"learning_rate": 9.744597157501957e-07, |
|
"loss": 0.8536, |
|
"step": 1220 |
|
} |
|
], |
|
"logging_steps": 1, |
|
"max_steps": 3048, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 3, |
|
"save_steps": 305, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 0.0, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|