diff --git "a/checkpoint-610/trainer_state.json" "b/checkpoint-610/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-610/trainer_state.json" @@ -0,0 +1,4570 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.6003937007874016, + "eval_steps": 153, + "global_step": 610, + "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 + } + ], + "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 +}