diff --git "a/checkpoint-2135/trainer_state.json" "b/checkpoint-2135/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-2135/trainer_state.json" @@ -0,0 +1,16135 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.1013779527559056, + "eval_steps": 153, + "global_step": 2135, + "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": 5.4393, + "step": 1 + }, + { + "epoch": 0.001968503937007874, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 28.0074, + "step": 2 + }, + { + "epoch": 0.002952755905511811, + "grad_norm": 23.266517639160156, + "learning_rate": 9.940357852882705e-10, + "loss": 4.1448, + "step": 3 + }, + { + "epoch": 0.003937007874015748, + "grad_norm": Infinity, + "learning_rate": 9.940357852882705e-10, + "loss": 5.9042, + "step": 4 + }, + { + "epoch": 0.004921259842519685, + "grad_norm": 33.751468658447266, + "learning_rate": 1.988071570576541e-09, + "loss": 5.1153, + "step": 5 + }, + { + "epoch": 0.005905511811023622, + "grad_norm": 40.47331619262695, + "learning_rate": 2.9821073558648116e-09, + "loss": 6.2251, + "step": 6 + }, + { + "epoch": 0.006889763779527559, + "grad_norm": 40.0908203125, + "learning_rate": 3.976143141153082e-09, + "loss": 6.6748, + "step": 7 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": 37.691471099853516, + "learning_rate": 4.970178926441353e-09, + "loss": 6.8788, + "step": 8 + }, + { + "epoch": 0.008858267716535433, + "grad_norm": 25.829736709594727, + "learning_rate": 5.964214711729623e-09, + "loss": 4.4231, + "step": 9 + }, + { + "epoch": 0.00984251968503937, + "grad_norm": 28.65277862548828, + "learning_rate": 6.9582504970178946e-09, + "loss": 5.0295, + "step": 10 + }, + { + "epoch": 0.010826771653543307, + "grad_norm": Infinity, + "learning_rate": 6.9582504970178946e-09, + "loss": 28.0661, + "step": 11 + }, + { + "epoch": 0.011811023622047244, + "grad_norm": Infinity, + "learning_rate": 6.9582504970178946e-09, + "loss": 27.2917, + "step": 12 + }, + { + "epoch": 0.012795275590551181, + "grad_norm": 27.269742965698242, + "learning_rate": 7.952286282306164e-09, + "loss": 5.4481, + "step": 13 + }, + { + "epoch": 0.013779527559055118, + "grad_norm": 76.96902465820312, + "learning_rate": 8.946322067594435e-09, + "loss": 27.5355, + "step": 14 + }, + { + "epoch": 0.014763779527559055, + "grad_norm": 43.93783950805664, + "learning_rate": 9.940357852882705e-09, + "loss": 6.5198, + "step": 15 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 32.38773727416992, + "learning_rate": 1.0934393638170978e-08, + "loss": 5.3392, + "step": 16 + }, + { + "epoch": 0.01673228346456693, + "grad_norm": 41.67168045043945, + "learning_rate": 1.1928429423459246e-08, + "loss": 5.9224, + "step": 17 + }, + { + "epoch": 0.017716535433070866, + "grad_norm": 37.90800094604492, + "learning_rate": 1.2922465208747517e-08, + "loss": 6.0366, + "step": 18 + }, + { + "epoch": 0.018700787401574805, + "grad_norm": 37.6961784362793, + "learning_rate": 1.3916500994035789e-08, + "loss": 6.0526, + "step": 19 + }, + { + "epoch": 0.01968503937007874, + "grad_norm": 74.10464477539062, + "learning_rate": 1.4910536779324056e-08, + "loss": 27.6148, + "step": 20 + }, + { + "epoch": 0.02066929133858268, + "grad_norm": 33.596378326416016, + "learning_rate": 1.590457256461233e-08, + "loss": 5.3683, + "step": 21 + }, + { + "epoch": 0.021653543307086614, + "grad_norm": 49.686702728271484, + "learning_rate": 1.68986083499006e-08, + "loss": 7.389, + "step": 22 + }, + { + "epoch": 0.022637795275590553, + "grad_norm": 28.7089786529541, + "learning_rate": 1.789264413518887e-08, + "loss": 5.3282, + "step": 23 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 25.526540756225586, + "learning_rate": 1.888667992047714e-08, + "loss": 4.8274, + "step": 24 + }, + { + "epoch": 0.024606299212598427, + "grad_norm": 69.71428680419922, + "learning_rate": 1.988071570576541e-08, + "loss": 25.8878, + "step": 25 + }, + { + "epoch": 0.025590551181102362, + "grad_norm": 44.590267181396484, + "learning_rate": 2.087475149105368e-08, + "loss": 6.7162, + "step": 26 + }, + { + "epoch": 0.0265748031496063, + "grad_norm": 28.835615158081055, + "learning_rate": 2.1868787276341955e-08, + "loss": 5.2153, + "step": 27 + }, + { + "epoch": 0.027559055118110236, + "grad_norm": 24.665605545043945, + "learning_rate": 2.2862823061630224e-08, + "loss": 4.3749, + "step": 28 + }, + { + "epoch": 0.028543307086614175, + "grad_norm": 30.081247329711914, + "learning_rate": 2.3856858846918493e-08, + "loss": 5.6032, + "step": 29 + }, + { + "epoch": 0.02952755905511811, + "grad_norm": 17.228534698486328, + "learning_rate": 2.4850894632206765e-08, + "loss": 5.1131, + "step": 30 + }, + { + "epoch": 0.03051181102362205, + "grad_norm": 65.52853393554688, + "learning_rate": 2.5844930417495034e-08, + "loss": 24.2319, + "step": 31 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 33.676334381103516, + "learning_rate": 2.6838966202783303e-08, + "loss": 6.1904, + "step": 32 + }, + { + "epoch": 0.03248031496062992, + "grad_norm": 24.33664894104004, + "learning_rate": 2.7833001988071578e-08, + "loss": 4.8403, + "step": 33 + }, + { + "epoch": 0.03346456692913386, + "grad_norm": 65.39360046386719, + "learning_rate": 2.8827037773359847e-08, + "loss": 23.9613, + "step": 34 + }, + { + "epoch": 0.0344488188976378, + "grad_norm": 64.55318450927734, + "learning_rate": 2.982107355864811e-08, + "loss": 24.0617, + "step": 35 + }, + { + "epoch": 0.03543307086614173, + "grad_norm": 27.231170654296875, + "learning_rate": 3.081510934393639e-08, + "loss": 5.3386, + "step": 36 + }, + { + "epoch": 0.03641732283464567, + "grad_norm": 15.538535118103027, + "learning_rate": 3.180914512922466e-08, + "loss": 3.9959, + "step": 37 + }, + { + "epoch": 0.03740157480314961, + "grad_norm": 19.403221130371094, + "learning_rate": 3.280318091451293e-08, + "loss": 4.7363, + "step": 38 + }, + { + "epoch": 0.038385826771653545, + "grad_norm": 22.01759147644043, + "learning_rate": 3.37972166998012e-08, + "loss": 4.5966, + "step": 39 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 20.658782958984375, + "learning_rate": 3.479125248508947e-08, + "loss": 4.6072, + "step": 40 + }, + { + "epoch": 0.040354330708661415, + "grad_norm": 22.004575729370117, + "learning_rate": 3.578528827037774e-08, + "loss": 4.3863, + "step": 41 + }, + { + "epoch": 0.04133858267716536, + "grad_norm": 15.574060440063477, + "learning_rate": 3.6779324055666005e-08, + "loss": 3.5401, + "step": 42 + }, + { + "epoch": 0.04232283464566929, + "grad_norm": 53.6202507019043, + "learning_rate": 3.777335984095428e-08, + "loss": 21.6366, + "step": 43 + }, + { + "epoch": 0.04330708661417323, + "grad_norm": 15.613253593444824, + "learning_rate": 3.8767395626242556e-08, + "loss": 4.1459, + "step": 44 + }, + { + "epoch": 0.04429133858267716, + "grad_norm": 23.75933837890625, + "learning_rate": 3.976143141153082e-08, + "loss": 5.2601, + "step": 45 + }, + { + "epoch": 0.045275590551181105, + "grad_norm": 24.233673095703125, + "learning_rate": 4.0755467196819094e-08, + "loss": 5.1343, + "step": 46 + }, + { + "epoch": 0.04625984251968504, + "grad_norm": 52.074249267578125, + "learning_rate": 4.174950298210736e-08, + "loss": 20.0884, + "step": 47 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 39.622276306152344, + "learning_rate": 4.274353876739563e-08, + "loss": 15.8916, + "step": 48 + }, + { + "epoch": 0.04822834645669291, + "grad_norm": 20.208133697509766, + "learning_rate": 4.373757455268391e-08, + "loss": 4.5513, + "step": 49 + }, + { + "epoch": 0.04921259842519685, + "grad_norm": 17.529369354248047, + "learning_rate": 4.4731610337972176e-08, + "loss": 4.5784, + "step": 50 + }, + { + "epoch": 0.05019685039370079, + "grad_norm": 18.710325241088867, + "learning_rate": 4.572564612326045e-08, + "loss": 4.4681, + "step": 51 + }, + { + "epoch": 0.051181102362204724, + "grad_norm": 46.71225357055664, + "learning_rate": 4.6719681908548713e-08, + "loss": 19.1849, + "step": 52 + }, + { + "epoch": 0.05216535433070866, + "grad_norm": 46.563560485839844, + "learning_rate": 4.7713717693836986e-08, + "loss": 18.5769, + "step": 53 + }, + { + "epoch": 0.0531496062992126, + "grad_norm": 17.8726749420166, + "learning_rate": 4.870775347912525e-08, + "loss": 4.8428, + "step": 54 + }, + { + "epoch": 0.054133858267716536, + "grad_norm": 12.50363826751709, + "learning_rate": 4.970178926441353e-08, + "loss": 3.8768, + "step": 55 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 13.662007331848145, + "learning_rate": 5.06958250497018e-08, + "loss": 4.1905, + "step": 56 + }, + { + "epoch": 0.05610236220472441, + "grad_norm": 42.61918640136719, + "learning_rate": 5.168986083499007e-08, + "loss": 17.839, + "step": 57 + }, + { + "epoch": 0.05708661417322835, + "grad_norm": 11.281116485595703, + "learning_rate": 5.268389662027834e-08, + "loss": 4.2829, + "step": 58 + }, + { + "epoch": 0.058070866141732284, + "grad_norm": 13.894123077392578, + "learning_rate": 5.3677932405566605e-08, + "loss": 4.8802, + "step": 59 + }, + { + "epoch": 0.05905511811023622, + "grad_norm": 14.007708549499512, + "learning_rate": 5.467196819085488e-08, + "loss": 4.6199, + "step": 60 + }, + { + "epoch": 0.060039370078740155, + "grad_norm": 12.504571914672852, + "learning_rate": 5.5666003976143156e-08, + "loss": 4.3791, + "step": 61 + }, + { + "epoch": 0.0610236220472441, + "grad_norm": 40.982643127441406, + "learning_rate": 5.666003976143142e-08, + "loss": 16.2548, + "step": 62 + }, + { + "epoch": 0.06200787401574803, + "grad_norm": 13.279716491699219, + "learning_rate": 5.7654075546719694e-08, + "loss": 3.8545, + "step": 63 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 19.03230094909668, + "learning_rate": 5.864811133200796e-08, + "loss": 4.9071, + "step": 64 + }, + { + "epoch": 0.0639763779527559, + "grad_norm": 17.066438674926758, + "learning_rate": 5.964214711729623e-08, + "loss": 4.3138, + "step": 65 + }, + { + "epoch": 0.06496062992125984, + "grad_norm": 12.428396224975586, + "learning_rate": 6.06361829025845e-08, + "loss": 4.2379, + "step": 66 + }, + { + "epoch": 0.06594488188976377, + "grad_norm": 12.542096138000488, + "learning_rate": 6.163021868787278e-08, + "loss": 4.3083, + "step": 67 + }, + { + "epoch": 0.06692913385826772, + "grad_norm": 38.80957794189453, + "learning_rate": 6.262425447316104e-08, + "loss": 14.9322, + "step": 68 + }, + { + "epoch": 0.06791338582677166, + "grad_norm": 11.219073295593262, + "learning_rate": 6.361829025844931e-08, + "loss": 3.9923, + "step": 69 + }, + { + "epoch": 0.0688976377952756, + "grad_norm": 11.427291870117188, + "learning_rate": 6.461232604373759e-08, + "loss": 4.0024, + "step": 70 + }, + { + "epoch": 0.06988188976377953, + "grad_norm": 15.939499855041504, + "learning_rate": 6.560636182902586e-08, + "loss": 4.1997, + "step": 71 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 14.391752243041992, + "learning_rate": 6.660039761431412e-08, + "loss": 3.9745, + "step": 72 + }, + { + "epoch": 0.0718503937007874, + "grad_norm": 13.009647369384766, + "learning_rate": 6.75944333996024e-08, + "loss": 4.0912, + "step": 73 + }, + { + "epoch": 0.07283464566929133, + "grad_norm": 36.10318374633789, + "learning_rate": 6.858846918489067e-08, + "loss": 13.9359, + "step": 74 + }, + { + "epoch": 0.07381889763779527, + "grad_norm": 10.009491920471191, + "learning_rate": 6.958250497017893e-08, + "loss": 3.5901, + "step": 75 + }, + { + "epoch": 0.07480314960629922, + "grad_norm": 11.90031909942627, + "learning_rate": 7.057654075546721e-08, + "loss": 3.3922, + "step": 76 + }, + { + "epoch": 0.07578740157480315, + "grad_norm": 18.244056701660156, + "learning_rate": 7.157057654075548e-08, + "loss": 4.27, + "step": 77 + }, + { + "epoch": 0.07677165354330709, + "grad_norm": 11.25367546081543, + "learning_rate": 7.256461232604374e-08, + "loss": 3.7923, + "step": 78 + }, + { + "epoch": 0.07775590551181102, + "grad_norm": 11.265695571899414, + "learning_rate": 7.355864811133201e-08, + "loss": 4.1654, + "step": 79 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 13.79488754272461, + "learning_rate": 7.455268389662029e-08, + "loss": 4.6474, + "step": 80 + }, + { + "epoch": 0.0797244094488189, + "grad_norm": 10.221566200256348, + "learning_rate": 7.554671968190855e-08, + "loss": 4.1161, + "step": 81 + }, + { + "epoch": 0.08070866141732283, + "grad_norm": 10.750500679016113, + "learning_rate": 7.654075546719683e-08, + "loss": 4.5888, + "step": 82 + }, + { + "epoch": 0.08169291338582677, + "grad_norm": 12.23923110961914, + "learning_rate": 7.753479125248511e-08, + "loss": 3.6771, + "step": 83 + }, + { + "epoch": 0.08267716535433071, + "grad_norm": 10.162713050842285, + "learning_rate": 7.852882703777338e-08, + "loss": 3.7787, + "step": 84 + }, + { + "epoch": 0.08366141732283465, + "grad_norm": 14.139646530151367, + "learning_rate": 7.952286282306164e-08, + "loss": 4.5414, + "step": 85 + }, + { + "epoch": 0.08464566929133858, + "grad_norm": 10.491697311401367, + "learning_rate": 8.051689860834992e-08, + "loss": 3.6806, + "step": 86 + }, + { + "epoch": 0.08562992125984252, + "grad_norm": 35.544918060302734, + "learning_rate": 8.151093439363819e-08, + "loss": 12.2725, + "step": 87 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 10.688739776611328, + "learning_rate": 8.250497017892645e-08, + "loss": 3.7953, + "step": 88 + }, + { + "epoch": 0.08759842519685039, + "grad_norm": 11.720666885375977, + "learning_rate": 8.349900596421472e-08, + "loss": 4.2137, + "step": 89 + }, + { + "epoch": 0.08858267716535433, + "grad_norm": 9.828582763671875, + "learning_rate": 8.4493041749503e-08, + "loss": 3.7188, + "step": 90 + }, + { + "epoch": 0.08956692913385826, + "grad_norm": 9.70528507232666, + "learning_rate": 8.548707753479126e-08, + "loss": 3.048, + "step": 91 + }, + { + "epoch": 0.09055118110236221, + "grad_norm": 10.048575401306152, + "learning_rate": 8.648111332007953e-08, + "loss": 3.9118, + "step": 92 + }, + { + "epoch": 0.09153543307086615, + "grad_norm": 36.533756256103516, + "learning_rate": 8.747514910536782e-08, + "loss": 12.4685, + "step": 93 + }, + { + "epoch": 0.09251968503937008, + "grad_norm": 37.62527847290039, + "learning_rate": 8.846918489065609e-08, + "loss": 12.153, + "step": 94 + }, + { + "epoch": 0.09350393700787402, + "grad_norm": 11.075881958007812, + "learning_rate": 8.946322067594435e-08, + "loss": 3.5308, + "step": 95 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 11.461938858032227, + "learning_rate": 9.045725646123262e-08, + "loss": 4.0387, + "step": 96 + }, + { + "epoch": 0.09547244094488189, + "grad_norm": 39.349822998046875, + "learning_rate": 9.14512922465209e-08, + "loss": 11.3819, + "step": 97 + }, + { + "epoch": 0.09645669291338582, + "grad_norm": 12.11327838897705, + "learning_rate": 9.244532803180916e-08, + "loss": 4.0332, + "step": 98 + }, + { + "epoch": 0.09744094488188976, + "grad_norm": 10.561070442199707, + "learning_rate": 9.343936381709743e-08, + "loss": 4.4249, + "step": 99 + }, + { + "epoch": 0.0984251968503937, + "grad_norm": 11.298664093017578, + "learning_rate": 9.44333996023857e-08, + "loss": 4.3217, + "step": 100 + }, + { + "epoch": 0.09940944881889764, + "grad_norm": 40.283451080322266, + "learning_rate": 9.542743538767397e-08, + "loss": 10.6294, + "step": 101 + }, + { + "epoch": 0.10039370078740158, + "grad_norm": 29.307941436767578, + "learning_rate": 9.642147117296224e-08, + "loss": 8.6388, + "step": 102 + }, + { + "epoch": 0.10137795275590551, + "grad_norm": 36.60942459106445, + "learning_rate": 9.74155069582505e-08, + "loss": 10.9995, + "step": 103 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 30.35404396057129, + "learning_rate": 9.840954274353878e-08, + "loss": 9.6468, + "step": 104 + }, + { + "epoch": 0.10334645669291338, + "grad_norm": 12.698685646057129, + "learning_rate": 9.940357852882706e-08, + "loss": 4.2915, + "step": 105 + }, + { + "epoch": 0.10433070866141732, + "grad_norm": 15.266210556030273, + "learning_rate": 1.0039761431411533e-07, + "loss": 4.567, + "step": 106 + }, + { + "epoch": 0.10531496062992125, + "grad_norm": 14.179697036743164, + "learning_rate": 1.013916500994036e-07, + "loss": 4.449, + "step": 107 + }, + { + "epoch": 0.1062992125984252, + "grad_norm": 10.754941940307617, + "learning_rate": 1.0238568588469187e-07, + "loss": 3.7339, + "step": 108 + }, + { + "epoch": 0.10728346456692914, + "grad_norm": 35.972007751464844, + "learning_rate": 1.0337972166998014e-07, + "loss": 9.5283, + "step": 109 + }, + { + "epoch": 0.10826771653543307, + "grad_norm": 22.920154571533203, + "learning_rate": 1.043737574552684e-07, + "loss": 3.991, + "step": 110 + }, + { + "epoch": 0.10925196850393701, + "grad_norm": 13.55944538116455, + "learning_rate": 1.0536779324055668e-07, + "loss": 4.0756, + "step": 111 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 13.829747200012207, + "learning_rate": 1.0636182902584495e-07, + "loss": 3.6422, + "step": 112 + }, + { + "epoch": 0.11122047244094488, + "grad_norm": 16.949851989746094, + "learning_rate": 1.0735586481113321e-07, + "loss": 5.0637, + "step": 113 + }, + { + "epoch": 0.11220472440944881, + "grad_norm": 10.135372161865234, + "learning_rate": 1.0834990059642149e-07, + "loss": 4.0646, + "step": 114 + }, + { + "epoch": 0.11318897637795275, + "grad_norm": 13.799703598022461, + "learning_rate": 1.0934393638170976e-07, + "loss": 4.1526, + "step": 115 + }, + { + "epoch": 0.1141732283464567, + "grad_norm": 10.923892974853516, + "learning_rate": 1.1033797216699802e-07, + "loss": 3.8386, + "step": 116 + }, + { + "epoch": 0.11515748031496063, + "grad_norm": 9.54076099395752, + "learning_rate": 1.1133200795228631e-07, + "loss": 3.627, + "step": 117 + }, + { + "epoch": 0.11614173228346457, + "grad_norm": 11.477869033813477, + "learning_rate": 1.1232604373757458e-07, + "loss": 4.1605, + "step": 118 + }, + { + "epoch": 0.1171259842519685, + "grad_norm": 10.326825141906738, + "learning_rate": 1.1332007952286284e-07, + "loss": 4.0105, + "step": 119 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 11.820672988891602, + "learning_rate": 1.1431411530815111e-07, + "loss": 3.8861, + "step": 120 + }, + { + "epoch": 0.11909448818897637, + "grad_norm": 10.27392864227295, + "learning_rate": 1.1530815109343939e-07, + "loss": 3.6702, + "step": 121 + }, + { + "epoch": 0.12007874015748031, + "grad_norm": 35.950748443603516, + "learning_rate": 1.1630218687872765e-07, + "loss": 8.3001, + "step": 122 + }, + { + "epoch": 0.12106299212598425, + "grad_norm": 35.172950744628906, + "learning_rate": 1.1729622266401592e-07, + "loss": 9.0462, + "step": 123 + }, + { + "epoch": 0.1220472440944882, + "grad_norm": 10.133049964904785, + "learning_rate": 1.182902584493042e-07, + "loss": 3.5176, + "step": 124 + }, + { + "epoch": 0.12303149606299213, + "grad_norm": 9.260289192199707, + "learning_rate": 1.1928429423459245e-07, + "loss": 3.9299, + "step": 125 + }, + { + "epoch": 0.12401574803149606, + "grad_norm": 28.9150447845459, + "learning_rate": 1.2027833001988073e-07, + "loss": 8.2621, + "step": 126 + }, + { + "epoch": 0.125, + "grad_norm": 10.249862670898438, + "learning_rate": 1.21272365805169e-07, + "loss": 3.6949, + "step": 127 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 11.146471977233887, + "learning_rate": 1.222664015904573e-07, + "loss": 3.2706, + "step": 128 + }, + { + "epoch": 0.12696850393700787, + "grad_norm": 11.053058624267578, + "learning_rate": 1.2326043737574557e-07, + "loss": 3.6803, + "step": 129 + }, + { + "epoch": 0.1279527559055118, + "grad_norm": 11.563634872436523, + "learning_rate": 1.2425447316103382e-07, + "loss": 3.2926, + "step": 130 + }, + { + "epoch": 0.12893700787401574, + "grad_norm": 13.855484008789062, + "learning_rate": 1.2524850894632207e-07, + "loss": 3.7752, + "step": 131 + }, + { + "epoch": 0.12992125984251968, + "grad_norm": 14.643479347229004, + "learning_rate": 1.2624254473161035e-07, + "loss": 3.8276, + "step": 132 + }, + { + "epoch": 0.1309055118110236, + "grad_norm": 9.72199535369873, + "learning_rate": 1.2723658051689863e-07, + "loss": 3.7446, + "step": 133 + }, + { + "epoch": 0.13188976377952755, + "grad_norm": 10.56336498260498, + "learning_rate": 1.282306163021869e-07, + "loss": 3.8056, + "step": 134 + }, + { + "epoch": 0.1328740157480315, + "grad_norm": 11.192402839660645, + "learning_rate": 1.2922465208747519e-07, + "loss": 3.7077, + "step": 135 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 9.400568962097168, + "learning_rate": 1.3021868787276344e-07, + "loss": 3.2734, + "step": 136 + }, + { + "epoch": 0.13484251968503938, + "grad_norm": 13.436176300048828, + "learning_rate": 1.3121272365805172e-07, + "loss": 3.6188, + "step": 137 + }, + { + "epoch": 0.13582677165354332, + "grad_norm": 39.868553161621094, + "learning_rate": 1.3220675944333997e-07, + "loss": 8.0773, + "step": 138 + }, + { + "epoch": 0.13681102362204725, + "grad_norm": 13.665103912353516, + "learning_rate": 1.3320079522862825e-07, + "loss": 4.4859, + "step": 139 + }, + { + "epoch": 0.1377952755905512, + "grad_norm": 10.541531562805176, + "learning_rate": 1.3419483101391653e-07, + "loss": 3.7818, + "step": 140 + }, + { + "epoch": 0.13877952755905512, + "grad_norm": 10.38351821899414, + "learning_rate": 1.351888667992048e-07, + "loss": 3.7187, + "step": 141 + }, + { + "epoch": 0.13976377952755906, + "grad_norm": 10.672074317932129, + "learning_rate": 1.3618290258449306e-07, + "loss": 3.9732, + "step": 142 + }, + { + "epoch": 0.140748031496063, + "grad_norm": 11.76051139831543, + "learning_rate": 1.3717693836978134e-07, + "loss": 3.7551, + "step": 143 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 11.446274757385254, + "learning_rate": 1.381709741550696e-07, + "loss": 3.7604, + "step": 144 + }, + { + "epoch": 0.14271653543307086, + "grad_norm": 10.468277931213379, + "learning_rate": 1.3916500994035787e-07, + "loss": 3.2703, + "step": 145 + }, + { + "epoch": 0.1437007874015748, + "grad_norm": 12.42776107788086, + "learning_rate": 1.4015904572564615e-07, + "loss": 3.7582, + "step": 146 + }, + { + "epoch": 0.14468503937007873, + "grad_norm": 10.938421249389648, + "learning_rate": 1.4115308151093443e-07, + "loss": 3.7076, + "step": 147 + }, + { + "epoch": 0.14566929133858267, + "grad_norm": 10.445693969726562, + "learning_rate": 1.421471172962227e-07, + "loss": 3.2915, + "step": 148 + }, + { + "epoch": 0.1466535433070866, + "grad_norm": 10.35840892791748, + "learning_rate": 1.4314115308151096e-07, + "loss": 3.9108, + "step": 149 + }, + { + "epoch": 0.14763779527559054, + "grad_norm": 11.640554428100586, + "learning_rate": 1.4413518886679924e-07, + "loss": 3.2007, + "step": 150 + }, + { + "epoch": 0.1486220472440945, + "grad_norm": 10.536201477050781, + "learning_rate": 1.451292246520875e-07, + "loss": 4.063, + "step": 151 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 10.62363338470459, + "learning_rate": 1.4612326043737577e-07, + "loss": 3.7597, + "step": 152 + }, + { + "epoch": 0.15059055118110237, + "grad_norm": 11.645200729370117, + "learning_rate": 1.4711729622266402e-07, + "loss": 3.5123, + "step": 153 + }, + { + "epoch": 0.15059055118110237, + "eval_Qnli-dev_cosine_accuracy": 0.5859375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.9580137729644775, + "eval_Qnli-dev_cosine_ap": 0.559540310893035, + "eval_Qnli-dev_cosine_f1": 0.6291834002677376, + "eval_Qnli-dev_cosine_f1_threshold": 0.7614445686340332, + "eval_Qnli-dev_cosine_precision": 0.4598825831702544, + "eval_Qnli-dev_cosine_recall": 0.9957627118644068, + "eval_Qnli-dev_dot_accuracy": 0.5859375, + "eval_Qnli-dev_dot_accuracy_threshold": 735.8870849609375, + "eval_Qnli-dev_dot_ap": 0.5597189508701941, + "eval_Qnli-dev_dot_f1": 0.6291834002677376, + "eval_Qnli-dev_dot_f1_threshold": 584.8859252929688, + "eval_Qnli-dev_dot_precision": 0.4598825831702544, + "eval_Qnli-dev_dot_recall": 0.9957627118644068, + "eval_Qnli-dev_euclidean_accuracy": 0.5859375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 8.031293869018555, + "eval_Qnli-dev_euclidean_ap": 0.5595454049536515, + "eval_Qnli-dev_euclidean_f1": 0.6291834002677376, + "eval_Qnli-dev_euclidean_f1_threshold": 19.135658264160156, + "eval_Qnli-dev_euclidean_precision": 0.4598825831702544, + "eval_Qnli-dev_euclidean_recall": 0.9957627118644068, + "eval_Qnli-dev_manhattan_accuracy": 0.6015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 157.00613403320312, + "eval_Qnli-dev_manhattan_ap": 0.5792037854776722, + "eval_Qnli-dev_manhattan_f1": 0.6300268096514745, + "eval_Qnli-dev_manhattan_f1_threshold": 302.62664794921875, + "eval_Qnli-dev_manhattan_precision": 0.46078431372549017, + "eval_Qnli-dev_manhattan_recall": 0.9957627118644068, + "eval_Qnli-dev_max_accuracy": 0.6015625, + "eval_Qnli-dev_max_accuracy_threshold": 735.8870849609375, + "eval_Qnli-dev_max_ap": 0.5792037854776722, + "eval_Qnli-dev_max_f1": 0.6300268096514745, + "eval_Qnli-dev_max_f1_threshold": 584.8859252929688, + "eval_Qnli-dev_max_precision": 0.46078431372549017, + "eval_Qnli-dev_max_recall": 0.9957627118644068, + "eval_allNLI-dev_cosine_accuracy": 0.671875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9673999547958374, + "eval_allNLI-dev_cosine_ap": 0.3945317873705333, + "eval_allNLI-dev_cosine_f1": 0.5060975609756098, + "eval_allNLI-dev_cosine_f1_threshold": 0.8668580055236816, + "eval_allNLI-dev_cosine_precision": 0.34368530020703936, + "eval_allNLI-dev_cosine_recall": 0.9595375722543352, + "eval_allNLI-dev_dot_accuracy": 0.671875, + "eval_allNLI-dev_dot_accuracy_threshold": 743.0966796875, + "eval_allNLI-dev_dot_ap": 0.3944429395112889, + "eval_allNLI-dev_dot_f1": 0.5060975609756098, + "eval_allNLI-dev_dot_f1_threshold": 665.8623046875, + "eval_allNLI-dev_dot_precision": 0.34368530020703936, + "eval_allNLI-dev_dot_recall": 0.9595375722543352, + "eval_allNLI-dev_euclidean_accuracy": 0.671875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 7.076907157897949, + "eval_allNLI-dev_euclidean_ap": 0.3945317873705333, + "eval_allNLI-dev_euclidean_f1": 0.5060975609756098, + "eval_allNLI-dev_euclidean_f1_threshold": 14.301802635192871, + "eval_allNLI-dev_euclidean_precision": 0.34368530020703936, + "eval_allNLI-dev_euclidean_recall": 0.9595375722543352, + "eval_allNLI-dev_manhattan_accuracy": 0.669921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 126.07140350341797, + "eval_allNLI-dev_manhattan_ap": 0.39814770003394995, + "eval_allNLI-dev_manhattan_f1": 0.5043988269794721, + "eval_allNLI-dev_manhattan_f1_threshold": 295.8509521484375, + "eval_allNLI-dev_manhattan_precision": 0.3379174852652259, + "eval_allNLI-dev_manhattan_recall": 0.9942196531791907, + "eval_allNLI-dev_max_accuracy": 0.671875, + "eval_allNLI-dev_max_accuracy_threshold": 743.0966796875, + "eval_allNLI-dev_max_ap": 0.39814770003394995, + "eval_allNLI-dev_max_f1": 0.5060975609756098, + "eval_allNLI-dev_max_f1_threshold": 665.8623046875, + "eval_allNLI-dev_max_precision": 0.34368530020703936, + "eval_allNLI-dev_max_recall": 0.9942196531791907, + "eval_loss": 2.7143681049346924, + "eval_runtime": 51.8031, + "eval_samples_per_second": 32.122, + "eval_sequential_score": 0.5792037854776722, + "eval_steps_per_second": 0.135, + "eval_sts-test_pearson_cosine": 0.3034049047698802, + "eval_sts-test_pearson_dot": 0.3032511646889009, + "eval_sts-test_pearson_euclidean": 0.3314665868797849, + "eval_sts-test_pearson_manhattan": 0.34733318724391027, + "eval_sts-test_pearson_max": 0.34733318724391027, + "eval_sts-test_spearman_cosine": 0.352760544413678, + "eval_sts-test_spearman_dot": 0.3525915636506641, + "eval_sts-test_spearman_euclidean": 0.3527674808591936, + "eval_sts-test_spearman_manhattan": 0.36528297663000353, + "eval_sts-test_spearman_max": 0.36528297663000353, + "step": 153 + }, + { + "epoch": 0.1515748031496063, + "grad_norm": 12.879021644592285, + "learning_rate": 1.4811133200795232e-07, + "loss": 4.1012, + "step": 154 + }, + { + "epoch": 0.15255905511811024, + "grad_norm": 9.246201515197754, + "learning_rate": 1.4910536779324058e-07, + "loss": 3.212, + "step": 155 + }, + { + "epoch": 0.15354330708661418, + "grad_norm": 11.628643035888672, + "learning_rate": 1.5009940357852886e-07, + "loss": 3.2028, + "step": 156 + }, + { + "epoch": 0.1545275590551181, + "grad_norm": 12.011067390441895, + "learning_rate": 1.510934393638171e-07, + "loss": 3.9507, + "step": 157 + }, + { + "epoch": 0.15551181102362205, + "grad_norm": 54.29878616333008, + "learning_rate": 1.5208747514910539e-07, + "loss": 8.985, + "step": 158 + }, + { + "epoch": 0.15649606299212598, + "grad_norm": 9.819111824035645, + "learning_rate": 1.5308151093439367e-07, + "loss": 3.0407, + "step": 159 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 9.699429512023926, + "learning_rate": 1.5407554671968192e-07, + "loss": 3.0494, + "step": 160 + }, + { + "epoch": 0.15846456692913385, + "grad_norm": 12.42037582397461, + "learning_rate": 1.5506958250497022e-07, + "loss": 3.357, + "step": 161 + }, + { + "epoch": 0.1594488188976378, + "grad_norm": 12.423589706420898, + "learning_rate": 1.5606361829025848e-07, + "loss": 3.5956, + "step": 162 + }, + { + "epoch": 0.16043307086614172, + "grad_norm": 11.176725387573242, + "learning_rate": 1.5705765407554675e-07, + "loss": 3.5224, + "step": 163 + }, + { + "epoch": 0.16141732283464566, + "grad_norm": 9.707764625549316, + "learning_rate": 1.58051689860835e-07, + "loss": 4.1106, + "step": 164 + }, + { + "epoch": 0.1624015748031496, + "grad_norm": 53.60691452026367, + "learning_rate": 1.5904572564612329e-07, + "loss": 8.9825, + "step": 165 + }, + { + "epoch": 0.16338582677165353, + "grad_norm": 56.2020149230957, + "learning_rate": 1.6003976143141154e-07, + "loss": 8.8678, + "step": 166 + }, + { + "epoch": 0.1643700787401575, + "grad_norm": 8.789534568786621, + "learning_rate": 1.6103379721669984e-07, + "loss": 2.9505, + "step": 167 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 9.67978286743164, + "learning_rate": 1.620278330019881e-07, + "loss": 3.0805, + "step": 168 + }, + { + "epoch": 0.16633858267716536, + "grad_norm": 9.513910293579102, + "learning_rate": 1.6302186878727637e-07, + "loss": 3.5239, + "step": 169 + }, + { + "epoch": 0.1673228346456693, + "grad_norm": 9.413968086242676, + "learning_rate": 1.6401590457256465e-07, + "loss": 3.6057, + "step": 170 + }, + { + "epoch": 0.16830708661417323, + "grad_norm": 9.712397575378418, + "learning_rate": 1.650099403578529e-07, + "loss": 3.4248, + "step": 171 + }, + { + "epoch": 0.16929133858267717, + "grad_norm": 44.07297897338867, + "learning_rate": 1.6600397614314118e-07, + "loss": 7.8892, + "step": 172 + }, + { + "epoch": 0.1702755905511811, + "grad_norm": 54.28559494018555, + "learning_rate": 1.6699801192842944e-07, + "loss": 8.4579, + "step": 173 + }, + { + "epoch": 0.17125984251968504, + "grad_norm": 10.271102905273438, + "learning_rate": 1.6799204771371774e-07, + "loss": 3.0962, + "step": 174 + }, + { + "epoch": 0.17224409448818898, + "grad_norm": 9.830999374389648, + "learning_rate": 1.68986083499006e-07, + "loss": 3.3825, + "step": 175 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 11.632530212402344, + "learning_rate": 1.6998011928429427e-07, + "loss": 3.2841, + "step": 176 + }, + { + "epoch": 0.17421259842519685, + "grad_norm": 10.029848098754883, + "learning_rate": 1.7097415506958253e-07, + "loss": 2.9979, + "step": 177 + }, + { + "epoch": 0.17519685039370078, + "grad_norm": 10.078184127807617, + "learning_rate": 1.719681908548708e-07, + "loss": 3.0206, + "step": 178 + }, + { + "epoch": 0.17618110236220472, + "grad_norm": 8.961700439453125, + "learning_rate": 1.7296222664015906e-07, + "loss": 3.1492, + "step": 179 + }, + { + "epoch": 0.17716535433070865, + "grad_norm": 10.939624786376953, + "learning_rate": 1.7395626242544734e-07, + "loss": 3.4719, + "step": 180 + }, + { + "epoch": 0.1781496062992126, + "grad_norm": 12.779672622680664, + "learning_rate": 1.7495029821073564e-07, + "loss": 3.219, + "step": 181 + }, + { + "epoch": 0.17913385826771652, + "grad_norm": 10.443387031555176, + "learning_rate": 1.759443339960239e-07, + "loss": 3.2491, + "step": 182 + }, + { + "epoch": 0.18011811023622049, + "grad_norm": 11.280111312866211, + "learning_rate": 1.7693836978131217e-07, + "loss": 3.2704, + "step": 183 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 10.742427825927734, + "learning_rate": 1.7793240556660042e-07, + "loss": 2.7536, + "step": 184 + }, + { + "epoch": 0.18208661417322836, + "grad_norm": 10.527275085449219, + "learning_rate": 1.789264413518887e-07, + "loss": 3.1839, + "step": 185 + }, + { + "epoch": 0.1830708661417323, + "grad_norm": 11.380389213562012, + "learning_rate": 1.7992047713717695e-07, + "loss": 3.303, + "step": 186 + }, + { + "epoch": 0.18405511811023623, + "grad_norm": 12.696966171264648, + "learning_rate": 1.8091451292246523e-07, + "loss": 3.0331, + "step": 187 + }, + { + "epoch": 0.18503937007874016, + "grad_norm": 133.25595092773438, + "learning_rate": 1.819085487077535e-07, + "loss": 11.7292, + "step": 188 + }, + { + "epoch": 0.1860236220472441, + "grad_norm": 12.542126655578613, + "learning_rate": 1.829025844930418e-07, + "loss": 3.3332, + "step": 189 + }, + { + "epoch": 0.18700787401574803, + "grad_norm": 12.556474685668945, + "learning_rate": 1.8389662027833004e-07, + "loss": 3.2252, + "step": 190 + }, + { + "epoch": 0.18799212598425197, + "grad_norm": 11.937398910522461, + "learning_rate": 1.8489065606361832e-07, + "loss": 3.1748, + "step": 191 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 17.055953979492188, + "learning_rate": 1.8588469184890657e-07, + "loss": 3.2157, + "step": 192 + }, + { + "epoch": 0.18996062992125984, + "grad_norm": 110.42061614990234, + "learning_rate": 1.8687872763419485e-07, + "loss": 11.8472, + "step": 193 + }, + { + "epoch": 0.19094488188976377, + "grad_norm": 13.12432861328125, + "learning_rate": 1.8787276341948313e-07, + "loss": 3.009, + "step": 194 + }, + { + "epoch": 0.1919291338582677, + "grad_norm": 11.234591484069824, + "learning_rate": 1.888667992047714e-07, + "loss": 2.7096, + "step": 195 + }, + { + "epoch": 0.19291338582677164, + "grad_norm": 12.8180570602417, + "learning_rate": 1.898608349900597e-07, + "loss": 3.7063, + "step": 196 + }, + { + "epoch": 0.19389763779527558, + "grad_norm": 11.697596549987793, + "learning_rate": 1.9085487077534794e-07, + "loss": 2.7595, + "step": 197 + }, + { + "epoch": 0.19488188976377951, + "grad_norm": 13.939170837402344, + "learning_rate": 1.9184890656063622e-07, + "loss": 2.7522, + "step": 198 + }, + { + "epoch": 0.19586614173228348, + "grad_norm": 12.252857208251953, + "learning_rate": 1.9284294234592447e-07, + "loss": 2.6238, + "step": 199 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 12.42138671875, + "learning_rate": 1.9383697813121275e-07, + "loss": 3.0377, + "step": 200 + }, + { + "epoch": 0.19783464566929135, + "grad_norm": 11.15164566040039, + "learning_rate": 1.94831013916501e-07, + "loss": 2.9167, + "step": 201 + }, + { + "epoch": 0.19881889763779528, + "grad_norm": 14.99342155456543, + "learning_rate": 1.958250497017893e-07, + "loss": 3.4026, + "step": 202 + }, + { + "epoch": 0.19980314960629922, + "grad_norm": 11.762504577636719, + "learning_rate": 1.9681908548707756e-07, + "loss": 2.8645, + "step": 203 + }, + { + "epoch": 0.20078740157480315, + "grad_norm": 11.21902847290039, + "learning_rate": 1.9781312127236584e-07, + "loss": 2.8705, + "step": 204 + }, + { + "epoch": 0.2017716535433071, + "grad_norm": 11.662793159484863, + "learning_rate": 1.9880715705765412e-07, + "loss": 3.0029, + "step": 205 + }, + { + "epoch": 0.20275590551181102, + "grad_norm": 9.543460845947266, + "learning_rate": 1.9980119284294237e-07, + "loss": 2.2861, + "step": 206 + }, + { + "epoch": 0.20374015748031496, + "grad_norm": 108.6338882446289, + "learning_rate": 2.0079522862823065e-07, + "loss": 14.354, + "step": 207 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 11.800822257995605, + "learning_rate": 2.017892644135189e-07, + "loss": 2.7733, + "step": 208 + }, + { + "epoch": 0.20570866141732283, + "grad_norm": 11.760587692260742, + "learning_rate": 2.027833001988072e-07, + "loss": 2.7618, + "step": 209 + }, + { + "epoch": 0.20669291338582677, + "grad_norm": 12.588250160217285, + "learning_rate": 2.0377733598409546e-07, + "loss": 2.9012, + "step": 210 + }, + { + "epoch": 0.2076771653543307, + "grad_norm": 10.962112426757812, + "learning_rate": 2.0477137176938374e-07, + "loss": 2.4187, + "step": 211 + }, + { + "epoch": 0.20866141732283464, + "grad_norm": 11.708662986755371, + "learning_rate": 2.05765407554672e-07, + "loss": 3.0123, + "step": 212 + }, + { + "epoch": 0.20964566929133857, + "grad_norm": 107.84342193603516, + "learning_rate": 2.0675944333996027e-07, + "loss": 14.8579, + "step": 213 + }, + { + "epoch": 0.2106299212598425, + "grad_norm": 11.205227851867676, + "learning_rate": 2.0775347912524852e-07, + "loss": 2.3067, + "step": 214 + }, + { + "epoch": 0.21161417322834647, + "grad_norm": 11.632312774658203, + "learning_rate": 2.087475149105368e-07, + "loss": 2.4989, + "step": 215 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 10.67951488494873, + "learning_rate": 2.097415506958251e-07, + "loss": 2.6475, + "step": 216 + }, + { + "epoch": 0.21358267716535434, + "grad_norm": 105.90058898925781, + "learning_rate": 2.1073558648111336e-07, + "loss": 14.8199, + "step": 217 + }, + { + "epoch": 0.21456692913385828, + "grad_norm": 13.48009204864502, + "learning_rate": 2.1172962226640164e-07, + "loss": 2.7968, + "step": 218 + }, + { + "epoch": 0.2155511811023622, + "grad_norm": 13.132164001464844, + "learning_rate": 2.127236580516899e-07, + "loss": 3.0639, + "step": 219 + }, + { + "epoch": 0.21653543307086615, + "grad_norm": 11.862865447998047, + "learning_rate": 2.1371769383697817e-07, + "loss": 2.3427, + "step": 220 + }, + { + "epoch": 0.21751968503937008, + "grad_norm": 98.5626220703125, + "learning_rate": 2.1471172962226642e-07, + "loss": 13.288, + "step": 221 + }, + { + "epoch": 0.21850393700787402, + "grad_norm": 88.0949935913086, + "learning_rate": 2.1570576540755473e-07, + "loss": 12.4998, + "step": 222 + }, + { + "epoch": 0.21948818897637795, + "grad_norm": 78.91239929199219, + "learning_rate": 2.1669980119284298e-07, + "loss": 11.582, + "step": 223 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 12.989612579345703, + "learning_rate": 2.1769383697813126e-07, + "loss": 2.9272, + "step": 224 + }, + { + "epoch": 0.22145669291338582, + "grad_norm": 10.528528213500977, + "learning_rate": 2.186878727634195e-07, + "loss": 2.997, + "step": 225 + }, + { + "epoch": 0.22244094488188976, + "grad_norm": 12.239700317382812, + "learning_rate": 2.196819085487078e-07, + "loss": 3.0848, + "step": 226 + }, + { + "epoch": 0.2234251968503937, + "grad_norm": 9.70710563659668, + "learning_rate": 2.2067594433399604e-07, + "loss": 3.1213, + "step": 227 + }, + { + "epoch": 0.22440944881889763, + "grad_norm": 10.96924877166748, + "learning_rate": 2.2166998011928432e-07, + "loss": 2.801, + "step": 228 + }, + { + "epoch": 0.22539370078740156, + "grad_norm": 10.285512924194336, + "learning_rate": 2.2266401590457263e-07, + "loss": 3.3834, + "step": 229 + }, + { + "epoch": 0.2263779527559055, + "grad_norm": 11.169885635375977, + "learning_rate": 2.2365805168986088e-07, + "loss": 2.6838, + "step": 230 + }, + { + "epoch": 0.22736220472440946, + "grad_norm": 11.633996963500977, + "learning_rate": 2.2465208747514916e-07, + "loss": 2.8403, + "step": 231 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 11.630722999572754, + "learning_rate": 2.256461232604374e-07, + "loss": 2.8931, + "step": 232 + }, + { + "epoch": 0.22933070866141733, + "grad_norm": 11.04566478729248, + "learning_rate": 2.266401590457257e-07, + "loss": 2.6142, + "step": 233 + }, + { + "epoch": 0.23031496062992127, + "grad_norm": 11.077239036560059, + "learning_rate": 2.2763419483101394e-07, + "loss": 2.8142, + "step": 234 + }, + { + "epoch": 0.2312992125984252, + "grad_norm": 77.64453125, + "learning_rate": 2.2862823061630222e-07, + "loss": 10.9256, + "step": 235 + }, + { + "epoch": 0.23228346456692914, + "grad_norm": 12.177558898925781, + "learning_rate": 2.296222664015905e-07, + "loss": 2.7954, + "step": 236 + }, + { + "epoch": 0.23326771653543307, + "grad_norm": 11.819588661193848, + "learning_rate": 2.3061630218687878e-07, + "loss": 1.9992, + "step": 237 + }, + { + "epoch": 0.234251968503937, + "grad_norm": 13.254929542541504, + "learning_rate": 2.3161033797216703e-07, + "loss": 2.809, + "step": 238 + }, + { + "epoch": 0.23523622047244094, + "grad_norm": 14.092291831970215, + "learning_rate": 2.326043737574553e-07, + "loss": 2.9697, + "step": 239 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 90.33231353759766, + "learning_rate": 2.335984095427436e-07, + "loss": 12.0765, + "step": 240 + }, + { + "epoch": 0.2372047244094488, + "grad_norm": 89.21347045898438, + "learning_rate": 2.3459244532803184e-07, + "loss": 10.8043, + "step": 241 + }, + { + "epoch": 0.23818897637795275, + "grad_norm": 11.470010757446289, + "learning_rate": 2.3558648111332012e-07, + "loss": 2.4787, + "step": 242 + }, + { + "epoch": 0.23917322834645668, + "grad_norm": 93.6417465209961, + "learning_rate": 2.365805168986084e-07, + "loss": 11.889, + "step": 243 + }, + { + "epoch": 0.24015748031496062, + "grad_norm": 106.90039825439453, + "learning_rate": 2.3757455268389668e-07, + "loss": 12.1268, + "step": 244 + }, + { + "epoch": 0.24114173228346455, + "grad_norm": 14.848711013793945, + "learning_rate": 2.385685884691849e-07, + "loss": 2.6714, + "step": 245 + }, + { + "epoch": 0.2421259842519685, + "grad_norm": 68.94020080566406, + "learning_rate": 2.395626242544732e-07, + "loss": 9.5693, + "step": 246 + }, + { + "epoch": 0.24311023622047245, + "grad_norm": 11.783477783203125, + "learning_rate": 2.4055666003976146e-07, + "loss": 2.6957, + "step": 247 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 43.78407287597656, + "learning_rate": 2.4155069582504976e-07, + "loss": 7.6639, + "step": 248 + }, + { + "epoch": 0.24507874015748032, + "grad_norm": 11.159483909606934, + "learning_rate": 2.42544731610338e-07, + "loss": 2.2924, + "step": 249 + }, + { + "epoch": 0.24606299212598426, + "grad_norm": 42.60402297973633, + "learning_rate": 2.4353876739562627e-07, + "loss": 7.6361, + "step": 250 + }, + { + "epoch": 0.2470472440944882, + "grad_norm": 11.96988296508789, + "learning_rate": 2.445328031809146e-07, + "loss": 2.5172, + "step": 251 + }, + { + "epoch": 0.24803149606299213, + "grad_norm": 32.87655258178711, + "learning_rate": 2.4552683896620283e-07, + "loss": 6.0579, + "step": 252 + }, + { + "epoch": 0.24901574803149606, + "grad_norm": 9.832136154174805, + "learning_rate": 2.4652087475149113e-07, + "loss": 2.6152, + "step": 253 + }, + { + "epoch": 0.25, + "grad_norm": 9.719839096069336, + "learning_rate": 2.475149105367794e-07, + "loss": 2.5602, + "step": 254 + }, + { + "epoch": 0.25098425196850394, + "grad_norm": 11.7006196975708, + "learning_rate": 2.4850894632206764e-07, + "loss": 2.6714, + "step": 255 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 12.606962203979492, + "learning_rate": 2.495029821073559e-07, + "loss": 2.6705, + "step": 256 + }, + { + "epoch": 0.2529527559055118, + "grad_norm": 27.753028869628906, + "learning_rate": 2.5049701789264414e-07, + "loss": 5.7727, + "step": 257 + }, + { + "epoch": 0.25393700787401574, + "grad_norm": 31.346681594848633, + "learning_rate": 2.5149105367793245e-07, + "loss": 5.5458, + "step": 258 + }, + { + "epoch": 0.2549212598425197, + "grad_norm": 12.829514503479004, + "learning_rate": 2.524850894632207e-07, + "loss": 2.3155, + "step": 259 + }, + { + "epoch": 0.2559055118110236, + "grad_norm": 10.834115982055664, + "learning_rate": 2.53479125248509e-07, + "loss": 2.7764, + "step": 260 + }, + { + "epoch": 0.25688976377952755, + "grad_norm": 9.759571075439453, + "learning_rate": 2.5447316103379726e-07, + "loss": 1.972, + "step": 261 + }, + { + "epoch": 0.2578740157480315, + "grad_norm": 12.796414375305176, + "learning_rate": 2.554671968190855e-07, + "loss": 2.2353, + "step": 262 + }, + { + "epoch": 0.2588582677165354, + "grad_norm": 11.815021514892578, + "learning_rate": 2.564612326043738e-07, + "loss": 2.6839, + "step": 263 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 12.061202049255371, + "learning_rate": 2.5745526838966207e-07, + "loss": 2.7648, + "step": 264 + }, + { + "epoch": 0.2608267716535433, + "grad_norm": 15.068981170654297, + "learning_rate": 2.5844930417495037e-07, + "loss": 2.852, + "step": 265 + }, + { + "epoch": 0.2618110236220472, + "grad_norm": 13.547517776489258, + "learning_rate": 2.5944333996023857e-07, + "loss": 2.1087, + "step": 266 + }, + { + "epoch": 0.26279527559055116, + "grad_norm": 14.179494857788086, + "learning_rate": 2.604373757455269e-07, + "loss": 2.5804, + "step": 267 + }, + { + "epoch": 0.2637795275590551, + "grad_norm": 12.557324409484863, + "learning_rate": 2.614314115308152e-07, + "loss": 2.3923, + "step": 268 + }, + { + "epoch": 0.26476377952755903, + "grad_norm": 12.106611251831055, + "learning_rate": 2.6242544731610343e-07, + "loss": 2.2182, + "step": 269 + }, + { + "epoch": 0.265748031496063, + "grad_norm": 12.182107925415039, + "learning_rate": 2.634194831013917e-07, + "loss": 2.3111, + "step": 270 + }, + { + "epoch": 0.26673228346456695, + "grad_norm": 12.86404800415039, + "learning_rate": 2.6441351888667994e-07, + "loss": 2.5097, + "step": 271 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 54.05057144165039, + "learning_rate": 2.6540755467196824e-07, + "loss": 6.3327, + "step": 272 + }, + { + "epoch": 0.2687007874015748, + "grad_norm": 12.239849090576172, + "learning_rate": 2.664015904572565e-07, + "loss": 2.4999, + "step": 273 + }, + { + "epoch": 0.26968503937007876, + "grad_norm": 12.460600852966309, + "learning_rate": 2.6739562624254475e-07, + "loss": 3.1619, + "step": 274 + }, + { + "epoch": 0.2706692913385827, + "grad_norm": 14.582513809204102, + "learning_rate": 2.6838966202783305e-07, + "loss": 2.7525, + "step": 275 + }, + { + "epoch": 0.27165354330708663, + "grad_norm": 15.142477989196777, + "learning_rate": 2.693836978131213e-07, + "loss": 2.4016, + "step": 276 + }, + { + "epoch": 0.27263779527559057, + "grad_norm": 12.963899612426758, + "learning_rate": 2.703777335984096e-07, + "loss": 2.3184, + "step": 277 + }, + { + "epoch": 0.2736220472440945, + "grad_norm": 12.028178215026855, + "learning_rate": 2.7137176938369786e-07, + "loss": 1.6971, + "step": 278 + }, + { + "epoch": 0.27460629921259844, + "grad_norm": 13.630943298339844, + "learning_rate": 2.723658051689861e-07, + "loss": 2.6344, + "step": 279 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 14.304428100585938, + "learning_rate": 2.7335984095427437e-07, + "loss": 2.4414, + "step": 280 + }, + { + "epoch": 0.2765748031496063, + "grad_norm": 11.340383529663086, + "learning_rate": 2.743538767395627e-07, + "loss": 2.0357, + "step": 281 + }, + { + "epoch": 0.27755905511811024, + "grad_norm": 113.19358825683594, + "learning_rate": 2.75347912524851e-07, + "loss": 11.7209, + "step": 282 + }, + { + "epoch": 0.2785433070866142, + "grad_norm": 35.21394729614258, + "learning_rate": 2.763419483101392e-07, + "loss": 3.3792, + "step": 283 + }, + { + "epoch": 0.2795275590551181, + "grad_norm": 13.167503356933594, + "learning_rate": 2.773359840954275e-07, + "loss": 2.2311, + "step": 284 + }, + { + "epoch": 0.28051181102362205, + "grad_norm": 101.97720336914062, + "learning_rate": 2.7833001988071574e-07, + "loss": 11.0255, + "step": 285 + }, + { + "epoch": 0.281496062992126, + "grad_norm": 50.5713005065918, + "learning_rate": 2.7932405566600404e-07, + "loss": 7.4335, + "step": 286 + }, + { + "epoch": 0.2824803149606299, + "grad_norm": 15.233115196228027, + "learning_rate": 2.803180914512923e-07, + "loss": 2.1134, + "step": 287 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 13.013148307800293, + "learning_rate": 2.8131212723658055e-07, + "loss": 2.4226, + "step": 288 + }, + { + "epoch": 0.2844488188976378, + "grad_norm": 12.31765365600586, + "learning_rate": 2.8230616302186885e-07, + "loss": 2.4353, + "step": 289 + }, + { + "epoch": 0.2854330708661417, + "grad_norm": 11.408845901489258, + "learning_rate": 2.833001988071571e-07, + "loss": 2.1077, + "step": 290 + }, + { + "epoch": 0.28641732283464566, + "grad_norm": 10.369791984558105, + "learning_rate": 2.842942345924454e-07, + "loss": 1.8507, + "step": 291 + }, + { + "epoch": 0.2874015748031496, + "grad_norm": 11.332071304321289, + "learning_rate": 2.852882703777336e-07, + "loss": 1.881, + "step": 292 + }, + { + "epoch": 0.28838582677165353, + "grad_norm": 11.65276050567627, + "learning_rate": 2.862823061630219e-07, + "loss": 2.0215, + "step": 293 + }, + { + "epoch": 0.28937007874015747, + "grad_norm": 51.045013427734375, + "learning_rate": 2.8727634194831017e-07, + "loss": 7.9533, + "step": 294 + }, + { + "epoch": 0.2903543307086614, + "grad_norm": 10.009112358093262, + "learning_rate": 2.8827037773359847e-07, + "loss": 1.6268, + "step": 295 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 12.618308067321777, + "learning_rate": 2.892644135188867e-07, + "loss": 2.4842, + "step": 296 + }, + { + "epoch": 0.29232283464566927, + "grad_norm": 12.222711563110352, + "learning_rate": 2.90258449304175e-07, + "loss": 1.8853, + "step": 297 + }, + { + "epoch": 0.2933070866141732, + "grad_norm": 74.74982452392578, + "learning_rate": 2.912524850894633e-07, + "loss": 8.5807, + "step": 298 + }, + { + "epoch": 0.29429133858267714, + "grad_norm": 11.625731468200684, + "learning_rate": 2.9224652087475153e-07, + "loss": 1.7294, + "step": 299 + }, + { + "epoch": 0.2952755905511811, + "grad_norm": 12.836846351623535, + "learning_rate": 2.9324055666003984e-07, + "loss": 2.6489, + "step": 300 + }, + { + "epoch": 0.296259842519685, + "grad_norm": 76.93685913085938, + "learning_rate": 2.9423459244532804e-07, + "loss": 8.0587, + "step": 301 + }, + { + "epoch": 0.297244094488189, + "grad_norm": 12.567708015441895, + "learning_rate": 2.9522862823061634e-07, + "loss": 2.496, + "step": 302 + }, + { + "epoch": 0.29822834645669294, + "grad_norm": 13.918742179870605, + "learning_rate": 2.9622266401590465e-07, + "loss": 2.1661, + "step": 303 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 10.268260955810547, + "learning_rate": 2.972166998011929e-07, + "loss": 1.8613, + "step": 304 + }, + { + "epoch": 0.3001968503937008, + "grad_norm": 12.233942985534668, + "learning_rate": 2.9821073558648115e-07, + "loss": 2.4711, + "step": 305 + }, + { + "epoch": 0.30118110236220474, + "grad_norm": 49.10639572143555, + "learning_rate": 2.992047713717694e-07, + "loss": 7.3749, + "step": 306 + }, + { + "epoch": 0.30118110236220474, + "eval_Qnli-dev_cosine_accuracy": 0.61328125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.916123628616333, + "eval_Qnli-dev_cosine_ap": 0.5984453025586394, + "eval_Qnli-dev_cosine_f1": 0.6381766381766382, + "eval_Qnli-dev_cosine_f1_threshold": 0.82862389087677, + "eval_Qnli-dev_cosine_precision": 0.48068669527896996, + "eval_Qnli-dev_cosine_recall": 0.9491525423728814, + "eval_Qnli-dev_dot_accuracy": 0.61328125, + "eval_Qnli-dev_dot_accuracy_threshold": 703.0178833007812, + "eval_Qnli-dev_dot_ap": 0.5978318102098164, + "eval_Qnli-dev_dot_f1": 0.6381766381766382, + "eval_Qnli-dev_dot_f1_threshold": 635.887451171875, + "eval_Qnli-dev_dot_precision": 0.48068669527896996, + "eval_Qnli-dev_dot_recall": 0.9491525423728814, + "eval_Qnli-dev_euclidean_accuracy": 0.61328125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 11.345888137817383, + "eval_Qnli-dev_euclidean_ap": 0.5983041877213555, + "eval_Qnli-dev_euclidean_f1": 0.6390870185449358, + "eval_Qnli-dev_euclidean_f1_threshold": 16.20899200439453, + "eval_Qnli-dev_euclidean_precision": 0.4817204301075269, + "eval_Qnli-dev_euclidean_recall": 0.9491525423728814, + "eval_Qnli-dev_manhattan_accuracy": 0.625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 224.3476104736328, + "eval_Qnli-dev_manhattan_ap": 0.608776755528686, + "eval_Qnli-dev_manhattan_f1": 0.6521106259097526, + "eval_Qnli-dev_manhattan_f1_threshold": 307.41888427734375, + "eval_Qnli-dev_manhattan_precision": 0.49667405764966743, + "eval_Qnli-dev_manhattan_recall": 0.9491525423728814, + "eval_Qnli-dev_max_accuracy": 0.625, + "eval_Qnli-dev_max_accuracy_threshold": 703.0178833007812, + "eval_Qnli-dev_max_ap": 0.608776755528686, + "eval_Qnli-dev_max_f1": 0.6521106259097526, + "eval_Qnli-dev_max_f1_threshold": 635.887451171875, + "eval_Qnli-dev_max_precision": 0.49667405764966743, + "eval_Qnli-dev_max_recall": 0.9491525423728814, + "eval_allNLI-dev_cosine_accuracy": 0.673828125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9651205539703369, + "eval_allNLI-dev_cosine_ap": 0.4220306766615718, + "eval_allNLI-dev_cosine_f1": 0.5118829981718465, + "eval_allNLI-dev_cosine_f1_threshold": 0.901992678642273, + "eval_allNLI-dev_cosine_precision": 0.37433155080213903, + "eval_allNLI-dev_cosine_recall": 0.8092485549132948, + "eval_allNLI-dev_dot_accuracy": 0.673828125, + "eval_allNLI-dev_dot_accuracy_threshold": 740.8585205078125, + "eval_allNLI-dev_dot_ap": 0.42232904959997797, + "eval_allNLI-dev_dot_f1": 0.5117117117117117, + "eval_allNLI-dev_dot_f1_threshold": 691.3936157226562, + "eval_allNLI-dev_dot_precision": 0.3717277486910995, + "eval_allNLI-dev_dot_recall": 0.8208092485549133, + "eval_allNLI-dev_euclidean_accuracy": 0.673828125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 7.317618370056152, + "eval_allNLI-dev_euclidean_ap": 0.42204167495286793, + "eval_allNLI-dev_euclidean_f1": 0.5118829981718465, + "eval_allNLI-dev_euclidean_f1_threshold": 12.265130043029785, + "eval_allNLI-dev_euclidean_precision": 0.37433155080213903, + "eval_allNLI-dev_euclidean_recall": 0.8092485549132948, + "eval_allNLI-dev_manhattan_accuracy": 0.669921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 108.60018920898438, + "eval_allNLI-dev_manhattan_ap": 0.41673033336995857, + "eval_allNLI-dev_manhattan_f1": 0.5128205128205128, + "eval_allNLI-dev_manhattan_f1_threshold": 255.00796508789062, + "eval_allNLI-dev_manhattan_precision": 0.3640776699029126, + "eval_allNLI-dev_manhattan_recall": 0.8670520231213873, + "eval_allNLI-dev_max_accuracy": 0.673828125, + "eval_allNLI-dev_max_accuracy_threshold": 740.8585205078125, + "eval_allNLI-dev_max_ap": 0.42232904959997797, + "eval_allNLI-dev_max_f1": 0.5128205128205128, + "eval_allNLI-dev_max_f1_threshold": 691.3936157226562, + "eval_allNLI-dev_max_precision": 0.37433155080213903, + "eval_allNLI-dev_max_recall": 0.8670520231213873, + "eval_loss": 1.8866561651229858, + "eval_runtime": 51.4126, + "eval_samples_per_second": 32.366, + "eval_sequential_score": 0.608776755528686, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.2632449966092212, + "eval_sts-test_pearson_dot": 0.2622330345292296, + "eval_sts-test_pearson_euclidean": 0.28968666678748967, + "eval_sts-test_pearson_manhattan": 0.289998923343493, + "eval_sts-test_pearson_max": 0.289998923343493, + "eval_sts-test_spearman_cosine": 0.315323563437873, + "eval_sts-test_spearman_dot": 0.3138196525475536, + "eval_sts-test_spearman_euclidean": 0.3153666589025929, + "eval_sts-test_spearman_manhattan": 0.3171849026529254, + "eval_sts-test_spearman_max": 0.3171849026529254, + "step": 306 + }, + { + "epoch": 0.3021653543307087, + "grad_norm": 11.498650550842285, + "learning_rate": 3.001988071570577e-07, + "loss": 2.1621, + "step": 307 + }, + { + "epoch": 0.3031496062992126, + "grad_norm": 11.405529975891113, + "learning_rate": 3.01192842942346e-07, + "loss": 1.7807, + "step": 308 + }, + { + "epoch": 0.30413385826771655, + "grad_norm": 10.972796440124512, + "learning_rate": 3.021868787276342e-07, + "loss": 2.2104, + "step": 309 + }, + { + "epoch": 0.3051181102362205, + "grad_norm": 11.644588470458984, + "learning_rate": 3.031809145129225e-07, + "loss": 2.1281, + "step": 310 + }, + { + "epoch": 0.3061023622047244, + "grad_norm": 43.770240783691406, + "learning_rate": 3.0417495029821077e-07, + "loss": 6.9695, + "step": 311 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 13.784707069396973, + "learning_rate": 3.051689860834991e-07, + "loss": 2.3068, + "step": 312 + }, + { + "epoch": 0.3080708661417323, + "grad_norm": 49.54023361206055, + "learning_rate": 3.0616302186878733e-07, + "loss": 7.162, + "step": 313 + }, + { + "epoch": 0.3090551181102362, + "grad_norm": 10.574169158935547, + "learning_rate": 3.071570576540756e-07, + "loss": 2.3378, + "step": 314 + }, + { + "epoch": 0.31003937007874016, + "grad_norm": 10.255630493164062, + "learning_rate": 3.0815109343936384e-07, + "loss": 1.911, + "step": 315 + }, + { + "epoch": 0.3110236220472441, + "grad_norm": 11.315991401672363, + "learning_rate": 3.0914512922465214e-07, + "loss": 1.8118, + "step": 316 + }, + { + "epoch": 0.31200787401574803, + "grad_norm": 11.545084953308105, + "learning_rate": 3.1013916500994045e-07, + "loss": 2.3977, + "step": 317 + }, + { + "epoch": 0.31299212598425197, + "grad_norm": 12.347085952758789, + "learning_rate": 3.1113320079522865e-07, + "loss": 2.4472, + "step": 318 + }, + { + "epoch": 0.3139763779527559, + "grad_norm": 11.311092376708984, + "learning_rate": 3.1212723658051695e-07, + "loss": 1.9549, + "step": 319 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 11.856200218200684, + "learning_rate": 3.131212723658052e-07, + "loss": 2.1023, + "step": 320 + }, + { + "epoch": 0.3159448818897638, + "grad_norm": 14.64504337310791, + "learning_rate": 3.141153081510935e-07, + "loss": 2.3819, + "step": 321 + }, + { + "epoch": 0.3169291338582677, + "grad_norm": 14.122454643249512, + "learning_rate": 3.151093439363818e-07, + "loss": 2.5242, + "step": 322 + }, + { + "epoch": 0.31791338582677164, + "grad_norm": 12.046473503112793, + "learning_rate": 3.1610337972167e-07, + "loss": 2.1382, + "step": 323 + }, + { + "epoch": 0.3188976377952756, + "grad_norm": 11.19228744506836, + "learning_rate": 3.170974155069583e-07, + "loss": 1.6567, + "step": 324 + }, + { + "epoch": 0.3198818897637795, + "grad_norm": 10.96595287322998, + "learning_rate": 3.1809145129224657e-07, + "loss": 1.6929, + "step": 325 + }, + { + "epoch": 0.32086614173228345, + "grad_norm": 60.629520416259766, + "learning_rate": 3.190854870775349e-07, + "loss": 8.3233, + "step": 326 + }, + { + "epoch": 0.3218503937007874, + "grad_norm": 12.280549049377441, + "learning_rate": 3.200795228628231e-07, + "loss": 1.9625, + "step": 327 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 65.54190063476562, + "learning_rate": 3.210735586481114e-07, + "loss": 8.4215, + "step": 328 + }, + { + "epoch": 0.32381889763779526, + "grad_norm": 12.738523483276367, + "learning_rate": 3.220675944333997e-07, + "loss": 1.7966, + "step": 329 + }, + { + "epoch": 0.3248031496062992, + "grad_norm": 10.889781951904297, + "learning_rate": 3.2306163021868794e-07, + "loss": 1.5442, + "step": 330 + }, + { + "epoch": 0.3257874015748031, + "grad_norm": 11.906895637512207, + "learning_rate": 3.240556660039762e-07, + "loss": 1.5842, + "step": 331 + }, + { + "epoch": 0.32677165354330706, + "grad_norm": 14.563230514526367, + "learning_rate": 3.2504970178926444e-07, + "loss": 2.0283, + "step": 332 + }, + { + "epoch": 0.327755905511811, + "grad_norm": 53.128257751464844, + "learning_rate": 3.2604373757455275e-07, + "loss": 7.7629, + "step": 333 + }, + { + "epoch": 0.328740157480315, + "grad_norm": 10.362350463867188, + "learning_rate": 3.27037773359841e-07, + "loss": 1.5504, + "step": 334 + }, + { + "epoch": 0.3297244094488189, + "grad_norm": 14.863081932067871, + "learning_rate": 3.280318091451293e-07, + "loss": 1.9673, + "step": 335 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 68.6702651977539, + "learning_rate": 3.2902584493041756e-07, + "loss": 8.3486, + "step": 336 + }, + { + "epoch": 0.3316929133858268, + "grad_norm": 11.154709815979004, + "learning_rate": 3.300198807157058e-07, + "loss": 1.8449, + "step": 337 + }, + { + "epoch": 0.33267716535433073, + "grad_norm": 51.934730529785156, + "learning_rate": 3.310139165009941e-07, + "loss": 7.6838, + "step": 338 + }, + { + "epoch": 0.33366141732283466, + "grad_norm": 11.276252746582031, + "learning_rate": 3.3200795228628237e-07, + "loss": 2.0031, + "step": 339 + }, + { + "epoch": 0.3346456692913386, + "grad_norm": 10.661493301391602, + "learning_rate": 3.330019880715706e-07, + "loss": 1.6159, + "step": 340 + }, + { + "epoch": 0.33562992125984253, + "grad_norm": 11.6641845703125, + "learning_rate": 3.3399602385685887e-07, + "loss": 1.9979, + "step": 341 + }, + { + "epoch": 0.33661417322834647, + "grad_norm": 42.07444763183594, + "learning_rate": 3.349900596421472e-07, + "loss": 6.4111, + "step": 342 + }, + { + "epoch": 0.3375984251968504, + "grad_norm": 42.648006439208984, + "learning_rate": 3.359840954274355e-07, + "loss": 5.9771, + "step": 343 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 10.396349906921387, + "learning_rate": 3.369781312127237e-07, + "loss": 1.5946, + "step": 344 + }, + { + "epoch": 0.3395669291338583, + "grad_norm": 12.217790603637695, + "learning_rate": 3.37972166998012e-07, + "loss": 2.0369, + "step": 345 + }, + { + "epoch": 0.3405511811023622, + "grad_norm": 37.96440124511719, + "learning_rate": 3.3896620278330024e-07, + "loss": 5.3554, + "step": 346 + }, + { + "epoch": 0.34153543307086615, + "grad_norm": 11.637295722961426, + "learning_rate": 3.3996023856858855e-07, + "loss": 2.215, + "step": 347 + }, + { + "epoch": 0.3425196850393701, + "grad_norm": 12.787163734436035, + "learning_rate": 3.409542743538768e-07, + "loss": 2.5668, + "step": 348 + }, + { + "epoch": 0.343503937007874, + "grad_norm": 10.833976745605469, + "learning_rate": 3.4194831013916505e-07, + "loss": 1.688, + "step": 349 + }, + { + "epoch": 0.34448818897637795, + "grad_norm": 11.339617729187012, + "learning_rate": 3.4294234592445336e-07, + "loss": 1.9555, + "step": 350 + }, + { + "epoch": 0.3454724409448819, + "grad_norm": 12.769586563110352, + "learning_rate": 3.439363817097416e-07, + "loss": 1.7196, + "step": 351 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 10.027280807495117, + "learning_rate": 3.449304174950299e-07, + "loss": 1.7139, + "step": 352 + }, + { + "epoch": 0.34744094488188976, + "grad_norm": 11.467499732971191, + "learning_rate": 3.459244532803181e-07, + "loss": 1.6959, + "step": 353 + }, + { + "epoch": 0.3484251968503937, + "grad_norm": 12.446152687072754, + "learning_rate": 3.469184890656064e-07, + "loss": 2.4359, + "step": 354 + }, + { + "epoch": 0.3494094488188976, + "grad_norm": 13.018854141235352, + "learning_rate": 3.4791252485089467e-07, + "loss": 2.2009, + "step": 355 + }, + { + "epoch": 0.35039370078740156, + "grad_norm": 12.124582290649414, + "learning_rate": 3.48906560636183e-07, + "loss": 2.2464, + "step": 356 + }, + { + "epoch": 0.3513779527559055, + "grad_norm": 12.922340393066406, + "learning_rate": 3.499005964214713e-07, + "loss": 1.6181, + "step": 357 + }, + { + "epoch": 0.35236220472440943, + "grad_norm": 13.51753044128418, + "learning_rate": 3.508946322067595e-07, + "loss": 1.9289, + "step": 358 + }, + { + "epoch": 0.35334645669291337, + "grad_norm": 12.572702407836914, + "learning_rate": 3.518886679920478e-07, + "loss": 1.6492, + "step": 359 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 12.720595359802246, + "learning_rate": 3.5288270377733604e-07, + "loss": 2.065, + "step": 360 + }, + { + "epoch": 0.35531496062992124, + "grad_norm": 81.92599487304688, + "learning_rate": 3.5387673956262434e-07, + "loss": 7.899, + "step": 361 + }, + { + "epoch": 0.3562992125984252, + "grad_norm": 12.986297607421875, + "learning_rate": 3.5487077534791254e-07, + "loss": 1.5591, + "step": 362 + }, + { + "epoch": 0.3572834645669291, + "grad_norm": 13.615756034851074, + "learning_rate": 3.5586481113320085e-07, + "loss": 2.0124, + "step": 363 + }, + { + "epoch": 0.35826771653543305, + "grad_norm": 69.3979263305664, + "learning_rate": 3.5685884691848915e-07, + "loss": 7.3982, + "step": 364 + }, + { + "epoch": 0.359251968503937, + "grad_norm": 55.40796661376953, + "learning_rate": 3.578528827037774e-07, + "loss": 6.4459, + "step": 365 + }, + { + "epoch": 0.36023622047244097, + "grad_norm": 35.5084342956543, + "learning_rate": 3.5884691848906566e-07, + "loss": 3.3719, + "step": 366 + }, + { + "epoch": 0.3612204724409449, + "grad_norm": 12.582864761352539, + "learning_rate": 3.598409542743539e-07, + "loss": 1.9539, + "step": 367 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 11.056336402893066, + "learning_rate": 3.608349900596422e-07, + "loss": 1.5193, + "step": 368 + }, + { + "epoch": 0.3631889763779528, + "grad_norm": 13.587398529052734, + "learning_rate": 3.6182902584493047e-07, + "loss": 2.086, + "step": 369 + }, + { + "epoch": 0.3641732283464567, + "grad_norm": 55.30432891845703, + "learning_rate": 3.6282306163021877e-07, + "loss": 5.8079, + "step": 370 + }, + { + "epoch": 0.36515748031496065, + "grad_norm": 12.745539665222168, + "learning_rate": 3.63817097415507e-07, + "loss": 1.8882, + "step": 371 + }, + { + "epoch": 0.3661417322834646, + "grad_norm": 11.412108421325684, + "learning_rate": 3.648111332007953e-07, + "loss": 1.7741, + "step": 372 + }, + { + "epoch": 0.3671259842519685, + "grad_norm": 42.92417907714844, + "learning_rate": 3.658051689860836e-07, + "loss": 5.5435, + "step": 373 + }, + { + "epoch": 0.36811023622047245, + "grad_norm": 11.240915298461914, + "learning_rate": 3.6679920477137184e-07, + "loss": 1.6847, + "step": 374 + }, + { + "epoch": 0.3690944881889764, + "grad_norm": 24.1689510345459, + "learning_rate": 3.677932405566601e-07, + "loss": 2.1274, + "step": 375 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 11.673293113708496, + "learning_rate": 3.6878727634194834e-07, + "loss": 1.4802, + "step": 376 + }, + { + "epoch": 0.37106299212598426, + "grad_norm": 39.63169860839844, + "learning_rate": 3.6978131212723665e-07, + "loss": 5.5033, + "step": 377 + }, + { + "epoch": 0.3720472440944882, + "grad_norm": 10.161783218383789, + "learning_rate": 3.7077534791252495e-07, + "loss": 1.4748, + "step": 378 + }, + { + "epoch": 0.37303149606299213, + "grad_norm": 11.951128005981445, + "learning_rate": 3.7176938369781315e-07, + "loss": 2.5636, + "step": 379 + }, + { + "epoch": 0.37401574803149606, + "grad_norm": 12.307464599609375, + "learning_rate": 3.7276341948310145e-07, + "loss": 2.1026, + "step": 380 + }, + { + "epoch": 0.375, + "grad_norm": 12.582341194152832, + "learning_rate": 3.737574552683897e-07, + "loss": 1.9383, + "step": 381 + }, + { + "epoch": 0.37598425196850394, + "grad_norm": 10.883564949035645, + "learning_rate": 3.74751491053678e-07, + "loss": 1.8689, + "step": 382 + }, + { + "epoch": 0.37696850393700787, + "grad_norm": 10.29930305480957, + "learning_rate": 3.7574552683896626e-07, + "loss": 1.722, + "step": 383 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 40.94071578979492, + "learning_rate": 3.767395626242545e-07, + "loss": 5.37, + "step": 384 + }, + { + "epoch": 0.37893700787401574, + "grad_norm": 32.14348602294922, + "learning_rate": 3.777335984095428e-07, + "loss": 4.5745, + "step": 385 + }, + { + "epoch": 0.3799212598425197, + "grad_norm": 12.06587028503418, + "learning_rate": 3.787276341948311e-07, + "loss": 1.9023, + "step": 386 + }, + { + "epoch": 0.3809055118110236, + "grad_norm": 12.805564880371094, + "learning_rate": 3.797216699801194e-07, + "loss": 1.8358, + "step": 387 + }, + { + "epoch": 0.38188976377952755, + "grad_norm": 13.309209823608398, + "learning_rate": 3.807157057654076e-07, + "loss": 1.8491, + "step": 388 + }, + { + "epoch": 0.3828740157480315, + "grad_norm": 11.234875679016113, + "learning_rate": 3.817097415506959e-07, + "loss": 1.6519, + "step": 389 + }, + { + "epoch": 0.3838582677165354, + "grad_norm": 14.558342933654785, + "learning_rate": 3.8270377733598414e-07, + "loss": 2.4811, + "step": 390 + }, + { + "epoch": 0.38484251968503935, + "grad_norm": 14.53983211517334, + "learning_rate": 3.8369781312127244e-07, + "loss": 2.4058, + "step": 391 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 32.78592300415039, + "learning_rate": 3.8469184890656075e-07, + "loss": 4.7023, + "step": 392 + }, + { + "epoch": 0.3868110236220472, + "grad_norm": 27.27498435974121, + "learning_rate": 3.8568588469184895e-07, + "loss": 3.175, + "step": 393 + }, + { + "epoch": 0.38779527559055116, + "grad_norm": 11.310855865478516, + "learning_rate": 3.8667992047713725e-07, + "loss": 1.9329, + "step": 394 + }, + { + "epoch": 0.3887795275590551, + "grad_norm": 10.476078987121582, + "learning_rate": 3.876739562624255e-07, + "loss": 1.3376, + "step": 395 + }, + { + "epoch": 0.38976377952755903, + "grad_norm": 9.912825584411621, + "learning_rate": 3.886679920477138e-07, + "loss": 1.5034, + "step": 396 + }, + { + "epoch": 0.390748031496063, + "grad_norm": 10.36616039276123, + "learning_rate": 3.89662027833002e-07, + "loss": 1.6294, + "step": 397 + }, + { + "epoch": 0.39173228346456695, + "grad_norm": 10.961806297302246, + "learning_rate": 3.906560636182903e-07, + "loss": 1.6094, + "step": 398 + }, + { + "epoch": 0.3927165354330709, + "grad_norm": 10.372532844543457, + "learning_rate": 3.916500994035786e-07, + "loss": 1.5651, + "step": 399 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 13.756068229675293, + "learning_rate": 3.9264413518886687e-07, + "loss": 2.4392, + "step": 400 + }, + { + "epoch": 0.39468503937007876, + "grad_norm": 34.37030029296875, + "learning_rate": 3.936381709741551e-07, + "loss": 4.4179, + "step": 401 + }, + { + "epoch": 0.3956692913385827, + "grad_norm": 30.322948455810547, + "learning_rate": 3.946322067594434e-07, + "loss": 4.5763, + "step": 402 + }, + { + "epoch": 0.39665354330708663, + "grad_norm": 11.070450782775879, + "learning_rate": 3.956262425447317e-07, + "loss": 1.5081, + "step": 403 + }, + { + "epoch": 0.39763779527559057, + "grad_norm": 34.38623046875, + "learning_rate": 3.9662027833001993e-07, + "loss": 4.2634, + "step": 404 + }, + { + "epoch": 0.3986220472440945, + "grad_norm": 11.74915885925293, + "learning_rate": 3.9761431411530824e-07, + "loss": 1.6725, + "step": 405 + }, + { + "epoch": 0.39960629921259844, + "grad_norm": 36.9741096496582, + "learning_rate": 3.986083499005965e-07, + "loss": 4.2432, + "step": 406 + }, + { + "epoch": 0.4005905511811024, + "grad_norm": 36.66675567626953, + "learning_rate": 3.9960238568588474e-07, + "loss": 4.2339, + "step": 407 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 12.600472450256348, + "learning_rate": 4.0059642147117305e-07, + "loss": 2.1615, + "step": 408 + }, + { + "epoch": 0.40255905511811024, + "grad_norm": 10.42699146270752, + "learning_rate": 4.015904572564613e-07, + "loss": 1.6564, + "step": 409 + }, + { + "epoch": 0.4035433070866142, + "grad_norm": 12.76755428314209, + "learning_rate": 4.0258449304174955e-07, + "loss": 2.3105, + "step": 410 + }, + { + "epoch": 0.4045275590551181, + "grad_norm": 11.840320587158203, + "learning_rate": 4.035785288270378e-07, + "loss": 1.6068, + "step": 411 + }, + { + "epoch": 0.40551181102362205, + "grad_norm": 12.758131980895996, + "learning_rate": 4.045725646123261e-07, + "loss": 1.915, + "step": 412 + }, + { + "epoch": 0.406496062992126, + "grad_norm": 13.703764915466309, + "learning_rate": 4.055666003976144e-07, + "loss": 1.834, + "step": 413 + }, + { + "epoch": 0.4074803149606299, + "grad_norm": 12.14503288269043, + "learning_rate": 4.065606361829026e-07, + "loss": 1.6041, + "step": 414 + }, + { + "epoch": 0.40846456692913385, + "grad_norm": 11.629945755004883, + "learning_rate": 4.075546719681909e-07, + "loss": 1.821, + "step": 415 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 11.828859329223633, + "learning_rate": 4.085487077534792e-07, + "loss": 1.772, + "step": 416 + }, + { + "epoch": 0.4104330708661417, + "grad_norm": 13.074658393859863, + "learning_rate": 4.095427435387675e-07, + "loss": 2.0671, + "step": 417 + }, + { + "epoch": 0.41141732283464566, + "grad_norm": 11.73059368133545, + "learning_rate": 4.1053677932405573e-07, + "loss": 1.3104, + "step": 418 + }, + { + "epoch": 0.4124015748031496, + "grad_norm": 12.339275360107422, + "learning_rate": 4.11530815109344e-07, + "loss": 1.4044, + "step": 419 + }, + { + "epoch": 0.41338582677165353, + "grad_norm": 13.040380477905273, + "learning_rate": 4.125248508946323e-07, + "loss": 1.6926, + "step": 420 + }, + { + "epoch": 0.41437007874015747, + "grad_norm": 11.760157585144043, + "learning_rate": 4.1351888667992054e-07, + "loss": 1.5651, + "step": 421 + }, + { + "epoch": 0.4153543307086614, + "grad_norm": 13.127520561218262, + "learning_rate": 4.1451292246520885e-07, + "loss": 1.7494, + "step": 422 + }, + { + "epoch": 0.41633858267716534, + "grad_norm": 12.034996032714844, + "learning_rate": 4.1550695825049705e-07, + "loss": 1.4918, + "step": 423 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 42.95773696899414, + "learning_rate": 4.1650099403578535e-07, + "loss": 4.4006, + "step": 424 + }, + { + "epoch": 0.4183070866141732, + "grad_norm": 13.784770965576172, + "learning_rate": 4.174950298210736e-07, + "loss": 1.5258, + "step": 425 + }, + { + "epoch": 0.41929133858267714, + "grad_norm": 12.560863494873047, + "learning_rate": 4.184890656063619e-07, + "loss": 1.5475, + "step": 426 + }, + { + "epoch": 0.4202755905511811, + "grad_norm": 12.546041488647461, + "learning_rate": 4.194831013916502e-07, + "loss": 1.5044, + "step": 427 + }, + { + "epoch": 0.421259842519685, + "grad_norm": 13.269911766052246, + "learning_rate": 4.204771371769384e-07, + "loss": 1.5454, + "step": 428 + }, + { + "epoch": 0.422244094488189, + "grad_norm": 12.899785041809082, + "learning_rate": 4.214711729622267e-07, + "loss": 1.6404, + "step": 429 + }, + { + "epoch": 0.42322834645669294, + "grad_norm": 13.493300437927246, + "learning_rate": 4.2246520874751497e-07, + "loss": 1.4207, + "step": 430 + }, + { + "epoch": 0.4242125984251969, + "grad_norm": 14.371034622192383, + "learning_rate": 4.234592445328033e-07, + "loss": 1.7885, + "step": 431 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 13.716957092285156, + "learning_rate": 4.244532803180915e-07, + "loss": 1.6594, + "step": 432 + }, + { + "epoch": 0.42618110236220474, + "grad_norm": 43.595272064208984, + "learning_rate": 4.254473161033798e-07, + "loss": 5.8005, + "step": 433 + }, + { + "epoch": 0.4271653543307087, + "grad_norm": 13.151529312133789, + "learning_rate": 4.264413518886681e-07, + "loss": 1.5369, + "step": 434 + }, + { + "epoch": 0.4281496062992126, + "grad_norm": 68.17890930175781, + "learning_rate": 4.2743538767395634e-07, + "loss": 7.8479, + "step": 435 + }, + { + "epoch": 0.42913385826771655, + "grad_norm": 14.194901466369629, + "learning_rate": 4.284294234592446e-07, + "loss": 1.5969, + "step": 436 + }, + { + "epoch": 0.4301181102362205, + "grad_norm": 12.343461990356445, + "learning_rate": 4.2942345924453284e-07, + "loss": 1.3442, + "step": 437 + }, + { + "epoch": 0.4311023622047244, + "grad_norm": 12.324607849121094, + "learning_rate": 4.3041749502982115e-07, + "loss": 1.2976, + "step": 438 + }, + { + "epoch": 0.43208661417322836, + "grad_norm": 14.157360076904297, + "learning_rate": 4.3141153081510945e-07, + "loss": 1.7511, + "step": 439 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 12.092175483703613, + "learning_rate": 4.324055666003977e-07, + "loss": 1.5225, + "step": 440 + }, + { + "epoch": 0.4340551181102362, + "grad_norm": 38.61181640625, + "learning_rate": 4.3339960238568596e-07, + "loss": 5.16, + "step": 441 + }, + { + "epoch": 0.43503937007874016, + "grad_norm": 10.22101879119873, + "learning_rate": 4.343936381709742e-07, + "loss": 1.0734, + "step": 442 + }, + { + "epoch": 0.4360236220472441, + "grad_norm": 9.978960037231445, + "learning_rate": 4.353876739562625e-07, + "loss": 1.0368, + "step": 443 + }, + { + "epoch": 0.43700787401574803, + "grad_norm": 12.753045082092285, + "learning_rate": 4.3638170974155077e-07, + "loss": 1.7517, + "step": 444 + }, + { + "epoch": 0.43799212598425197, + "grad_norm": 13.640998840332031, + "learning_rate": 4.37375745526839e-07, + "loss": 1.9461, + "step": 445 + }, + { + "epoch": 0.4389763779527559, + "grad_norm": 11.576430320739746, + "learning_rate": 4.383697813121273e-07, + "loss": 1.4079, + "step": 446 + }, + { + "epoch": 0.43996062992125984, + "grad_norm": 38.03751754760742, + "learning_rate": 4.393638170974156e-07, + "loss": 4.2293, + "step": 447 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 12.281790733337402, + "learning_rate": 4.403578528827039e-07, + "loss": 1.4854, + "step": 448 + }, + { + "epoch": 0.4419291338582677, + "grad_norm": 11.98123836517334, + "learning_rate": 4.413518886679921e-07, + "loss": 1.4409, + "step": 449 + }, + { + "epoch": 0.44291338582677164, + "grad_norm": 14.723586082458496, + "learning_rate": 4.423459244532804e-07, + "loss": 1.7013, + "step": 450 + }, + { + "epoch": 0.4438976377952756, + "grad_norm": 12.312840461730957, + "learning_rate": 4.4333996023856864e-07, + "loss": 1.439, + "step": 451 + }, + { + "epoch": 0.4448818897637795, + "grad_norm": 10.803173065185547, + "learning_rate": 4.4433399602385695e-07, + "loss": 1.3878, + "step": 452 + }, + { + "epoch": 0.44586614173228345, + "grad_norm": 10.145612716674805, + "learning_rate": 4.4532803180914525e-07, + "loss": 1.106, + "step": 453 + }, + { + "epoch": 0.4468503937007874, + "grad_norm": 11.358169555664062, + "learning_rate": 4.4632206759443345e-07, + "loss": 1.4017, + "step": 454 + }, + { + "epoch": 0.4478346456692913, + "grad_norm": 14.369976997375488, + "learning_rate": 4.4731610337972176e-07, + "loss": 1.9968, + "step": 455 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 11.71997356414795, + "learning_rate": 4.4831013916501e-07, + "loss": 1.468, + "step": 456 + }, + { + "epoch": 0.4498031496062992, + "grad_norm": 10.118363380432129, + "learning_rate": 4.493041749502983e-07, + "loss": 1.1154, + "step": 457 + }, + { + "epoch": 0.4507874015748031, + "grad_norm": 11.167040824890137, + "learning_rate": 4.502982107355865e-07, + "loss": 1.0393, + "step": 458 + }, + { + "epoch": 0.45177165354330706, + "grad_norm": 48.0859489440918, + "learning_rate": 4.512922465208748e-07, + "loss": 5.0393, + "step": 459 + }, + { + "epoch": 0.45177165354330706, + "eval_Qnli-dev_cosine_accuracy": 0.642578125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8873454928398132, + "eval_Qnli-dev_cosine_ap": 0.6389617582073207, + "eval_Qnli-dev_cosine_f1": 0.6468085106382979, + "eval_Qnli-dev_cosine_f1_threshold": 0.7709820866584778, + "eval_Qnli-dev_cosine_precision": 0.4861407249466951, + "eval_Qnli-dev_cosine_recall": 0.9661016949152542, + "eval_Qnli-dev_dot_accuracy": 0.642578125, + "eval_Qnli-dev_dot_accuracy_threshold": 680.0654296875, + "eval_Qnli-dev_dot_ap": 0.6378866352495927, + "eval_Qnli-dev_dot_f1": 0.6468085106382979, + "eval_Qnli-dev_dot_f1_threshold": 590.9285888671875, + "eval_Qnli-dev_dot_precision": 0.4861407249466951, + "eval_Qnli-dev_dot_recall": 0.9661016949152542, + "eval_Qnli-dev_euclidean_accuracy": 0.642578125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.140678405761719, + "eval_Qnli-dev_euclidean_ap": 0.6388920776922351, + "eval_Qnli-dev_euclidean_f1": 0.6468926553672316, + "eval_Qnli-dev_euclidean_f1_threshold": 18.792827606201172, + "eval_Qnli-dev_euclidean_precision": 0.4851694915254237, + "eval_Qnli-dev_euclidean_recall": 0.9703389830508474, + "eval_Qnli-dev_manhattan_accuracy": 0.6484375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 267.9780578613281, + "eval_Qnli-dev_manhattan_ap": 0.6456498934750234, + "eval_Qnli-dev_manhattan_f1": 0.6539074960127591, + "eval_Qnli-dev_manhattan_f1_threshold": 340.74969482421875, + "eval_Qnli-dev_manhattan_precision": 0.5242966751918159, + "eval_Qnli-dev_manhattan_recall": 0.8686440677966102, + "eval_Qnli-dev_max_accuracy": 0.6484375, + "eval_Qnli-dev_max_accuracy_threshold": 680.0654296875, + "eval_Qnli-dev_max_ap": 0.6456498934750234, + "eval_Qnli-dev_max_f1": 0.6539074960127591, + "eval_Qnli-dev_max_f1_threshold": 590.9285888671875, + "eval_Qnli-dev_max_precision": 0.5242966751918159, + "eval_Qnli-dev_max_recall": 0.9703389830508474, + "eval_allNLI-dev_cosine_accuracy": 0.671875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9784883260726929, + "eval_allNLI-dev_cosine_ap": 0.4327293323599516, + "eval_allNLI-dev_cosine_f1": 0.513911620294599, + "eval_allNLI-dev_cosine_f1_threshold": 0.8457035422325134, + "eval_allNLI-dev_cosine_precision": 0.3584474885844749, + "eval_allNLI-dev_cosine_recall": 0.9075144508670521, + "eval_allNLI-dev_dot_accuracy": 0.671875, + "eval_allNLI-dev_dot_accuracy_threshold": 751.0498046875, + "eval_allNLI-dev_dot_ap": 0.4334997927998542, + "eval_allNLI-dev_dot_f1": 0.513911620294599, + "eval_allNLI-dev_dot_f1_threshold": 648.7015380859375, + "eval_allNLI-dev_dot_precision": 0.3584474885844749, + "eval_allNLI-dev_dot_recall": 0.9075144508670521, + "eval_allNLI-dev_euclidean_accuracy": 0.671875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 5.746514797210693, + "eval_allNLI-dev_euclidean_ap": 0.43273474057499156, + "eval_allNLI-dev_euclidean_f1": 0.513911620294599, + "eval_allNLI-dev_euclidean_f1_threshold": 15.385263442993164, + "eval_allNLI-dev_euclidean_precision": 0.3584474885844749, + "eval_allNLI-dev_euclidean_recall": 0.9075144508670521, + "eval_allNLI-dev_manhattan_accuracy": 0.669921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 109.5778579711914, + "eval_allNLI-dev_manhattan_ap": 0.4334808884960006, + "eval_allNLI-dev_manhattan_f1": 0.5152057245080501, + "eval_allNLI-dev_manhattan_f1_threshold": 274.7369384765625, + "eval_allNLI-dev_manhattan_precision": 0.37305699481865284, + "eval_allNLI-dev_manhattan_recall": 0.8323699421965318, + "eval_allNLI-dev_max_accuracy": 0.671875, + "eval_allNLI-dev_max_accuracy_threshold": 751.0498046875, + "eval_allNLI-dev_max_ap": 0.4334997927998542, + "eval_allNLI-dev_max_f1": 0.5152057245080501, + "eval_allNLI-dev_max_f1_threshold": 648.7015380859375, + "eval_allNLI-dev_max_precision": 0.37305699481865284, + "eval_allNLI-dev_max_recall": 0.9075144508670521, + "eval_loss": 1.401798963546753, + "eval_runtime": 51.2939, + "eval_samples_per_second": 32.441, + "eval_sequential_score": 0.6456498934750234, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.3066273602034609, + "eval_sts-test_pearson_dot": 0.3041194856492825, + "eval_sts-test_pearson_euclidean": 0.3345205919827309, + "eval_sts-test_pearson_manhattan": 0.3398796566403535, + "eval_sts-test_pearson_max": 0.3398796566403535, + "eval_sts-test_spearman_cosine": 0.3621057713249984, + "eval_sts-test_spearman_dot": 0.3582156784260282, + "eval_sts-test_spearman_euclidean": 0.3622225721172299, + "eval_sts-test_spearman_manhattan": 0.36877890041863826, + "eval_sts-test_spearman_max": 0.36877890041863826, + "step": 459 + }, + { + "epoch": 0.452755905511811, + "grad_norm": 82.39715576171875, + "learning_rate": 4.522862823061631e-07, + "loss": 8.3174, + "step": 460 + }, + { + "epoch": 0.453740157480315, + "grad_norm": 52.898155212402344, + "learning_rate": 4.532803180914514e-07, + "loss": 4.1278, + "step": 461 + }, + { + "epoch": 0.4547244094488189, + "grad_norm": 13.927783966064453, + "learning_rate": 4.542743538767397e-07, + "loss": 1.5574, + "step": 462 + }, + { + "epoch": 0.45570866141732286, + "grad_norm": 59.73894119262695, + "learning_rate": 4.552683896620279e-07, + "loss": 6.7434, + "step": 463 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 10.882119178771973, + "learning_rate": 4.562624254473162e-07, + "loss": 1.166, + "step": 464 + }, + { + "epoch": 0.45767716535433073, + "grad_norm": 10.511579513549805, + "learning_rate": 4.5725646123260444e-07, + "loss": 1.1528, + "step": 465 + }, + { + "epoch": 0.45866141732283466, + "grad_norm": 12.323670387268066, + "learning_rate": 4.5825049701789274e-07, + "loss": 1.2498, + "step": 466 + }, + { + "epoch": 0.4596456692913386, + "grad_norm": 10.636286735534668, + "learning_rate": 4.59244532803181e-07, + "loss": 1.0892, + "step": 467 + }, + { + "epoch": 0.46062992125984253, + "grad_norm": 10.837008476257324, + "learning_rate": 4.6023856858846925e-07, + "loss": 1.151, + "step": 468 + }, + { + "epoch": 0.46161417322834647, + "grad_norm": 9.60957145690918, + "learning_rate": 4.6123260437375755e-07, + "loss": 0.9736, + "step": 469 + }, + { + "epoch": 0.4625984251968504, + "grad_norm": 12.348689079284668, + "learning_rate": 4.622266401590458e-07, + "loss": 1.422, + "step": 470 + }, + { + "epoch": 0.46358267716535434, + "grad_norm": 11.842246055603027, + "learning_rate": 4.6322067594433406e-07, + "loss": 1.2426, + "step": 471 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 41.27305221557617, + "learning_rate": 4.642147117296223e-07, + "loss": 5.3049, + "step": 472 + }, + { + "epoch": 0.4655511811023622, + "grad_norm": 11.455384254455566, + "learning_rate": 4.652087475149106e-07, + "loss": 1.2425, + "step": 473 + }, + { + "epoch": 0.46653543307086615, + "grad_norm": 12.320660591125488, + "learning_rate": 4.662027833001989e-07, + "loss": 1.5747, + "step": 474 + }, + { + "epoch": 0.4675196850393701, + "grad_norm": 11.062726974487305, + "learning_rate": 4.671968190854872e-07, + "loss": 1.2091, + "step": 475 + }, + { + "epoch": 0.468503937007874, + "grad_norm": 12.084156036376953, + "learning_rate": 4.681908548707754e-07, + "loss": 1.6617, + "step": 476 + }, + { + "epoch": 0.46948818897637795, + "grad_norm": 51.530765533447266, + "learning_rate": 4.691848906560637e-07, + "loss": 5.6202, + "step": 477 + }, + { + "epoch": 0.4704724409448819, + "grad_norm": 43.16566848754883, + "learning_rate": 4.70178926441352e-07, + "loss": 4.6429, + "step": 478 + }, + { + "epoch": 0.4714566929133858, + "grad_norm": 13.047057151794434, + "learning_rate": 4.7117296222664024e-07, + "loss": 1.8271, + "step": 479 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 10.565884590148926, + "learning_rate": 4.721669980119285e-07, + "loss": 1.155, + "step": 480 + }, + { + "epoch": 0.4734251968503937, + "grad_norm": 42.160797119140625, + "learning_rate": 4.731610337972168e-07, + "loss": 4.2303, + "step": 481 + }, + { + "epoch": 0.4744094488188976, + "grad_norm": 10.417061805725098, + "learning_rate": 4.7415506958250505e-07, + "loss": 1.3812, + "step": 482 + }, + { + "epoch": 0.47539370078740156, + "grad_norm": 30.367698669433594, + "learning_rate": 4.7514910536779335e-07, + "loss": 3.9862, + "step": 483 + }, + { + "epoch": 0.4763779527559055, + "grad_norm": 30.068883895874023, + "learning_rate": 4.7614314115308155e-07, + "loss": 3.8449, + "step": 484 + }, + { + "epoch": 0.47736220472440943, + "grad_norm": 28.166189193725586, + "learning_rate": 4.771371769383698e-07, + "loss": 3.0548, + "step": 485 + }, + { + "epoch": 0.47834645669291337, + "grad_norm": 25.816070556640625, + "learning_rate": 4.781312127236581e-07, + "loss": 3.3619, + "step": 486 + }, + { + "epoch": 0.4793307086614173, + "grad_norm": 10.16846752166748, + "learning_rate": 4.791252485089464e-07, + "loss": 1.4412, + "step": 487 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 21.649477005004883, + "learning_rate": 4.801192842942347e-07, + "loss": 2.6257, + "step": 488 + }, + { + "epoch": 0.4812992125984252, + "grad_norm": 11.959589004516602, + "learning_rate": 4.811133200795229e-07, + "loss": 1.6072, + "step": 489 + }, + { + "epoch": 0.4822834645669291, + "grad_norm": 10.868080139160156, + "learning_rate": 4.821073558648112e-07, + "loss": 1.6318, + "step": 490 + }, + { + "epoch": 0.48326771653543305, + "grad_norm": 15.460439682006836, + "learning_rate": 4.831013916500995e-07, + "loss": 1.952, + "step": 491 + }, + { + "epoch": 0.484251968503937, + "grad_norm": 10.419231414794922, + "learning_rate": 4.840954274353877e-07, + "loss": 1.7529, + "step": 492 + }, + { + "epoch": 0.48523622047244097, + "grad_norm": 10.788681983947754, + "learning_rate": 4.85089463220676e-07, + "loss": 1.7482, + "step": 493 + }, + { + "epoch": 0.4862204724409449, + "grad_norm": 30.656389236450195, + "learning_rate": 4.860834990059642e-07, + "loss": 2.8361, + "step": 494 + }, + { + "epoch": 0.48720472440944884, + "grad_norm": 27.033105850219727, + "learning_rate": 4.870775347912525e-07, + "loss": 3.4978, + "step": 495 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 11.075968742370605, + "learning_rate": 4.880715705765408e-07, + "loss": 1.5524, + "step": 496 + }, + { + "epoch": 0.4891732283464567, + "grad_norm": 10.439245223999023, + "learning_rate": 4.890656063618291e-07, + "loss": 1.2908, + "step": 497 + }, + { + "epoch": 0.49015748031496065, + "grad_norm": 15.17002010345459, + "learning_rate": 4.900596421471173e-07, + "loss": 2.3283, + "step": 498 + }, + { + "epoch": 0.4911417322834646, + "grad_norm": 12.133261680603027, + "learning_rate": 4.910536779324057e-07, + "loss": 1.7446, + "step": 499 + }, + { + "epoch": 0.4921259842519685, + "grad_norm": 11.527321815490723, + "learning_rate": 4.92047713717694e-07, + "loss": 1.5288, + "step": 500 + }, + { + "epoch": 0.49311023622047245, + "grad_norm": 9.531623840332031, + "learning_rate": 4.930417495029823e-07, + "loss": 1.1935, + "step": 501 + }, + { + "epoch": 0.4940944881889764, + "grad_norm": 10.518357276916504, + "learning_rate": 4.940357852882705e-07, + "loss": 1.3287, + "step": 502 + }, + { + "epoch": 0.4950787401574803, + "grad_norm": 11.910670280456543, + "learning_rate": 4.950298210735588e-07, + "loss": 1.5486, + "step": 503 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 9.1786527633667, + "learning_rate": 4.96023856858847e-07, + "loss": 0.8555, + "step": 504 + }, + { + "epoch": 0.4970472440944882, + "grad_norm": 30.960664749145508, + "learning_rate": 4.970178926441353e-07, + "loss": 3.1332, + "step": 505 + }, + { + "epoch": 0.49803149606299213, + "grad_norm": 43.69401550292969, + "learning_rate": 4.980119284294235e-07, + "loss": 3.8618, + "step": 506 + }, + { + "epoch": 0.49901574803149606, + "grad_norm": 14.343463897705078, + "learning_rate": 4.990059642147118e-07, + "loss": 1.6118, + "step": 507 + }, + { + "epoch": 0.5, + "grad_norm": 24.55521011352539, + "learning_rate": 5.000000000000001e-07, + "loss": 2.1475, + "step": 508 + }, + { + "epoch": 0.5009842519685039, + "grad_norm": 10.86742877960205, + "learning_rate": 5.009940357852883e-07, + "loss": 1.057, + "step": 509 + }, + { + "epoch": 0.5019685039370079, + "grad_norm": 11.385946273803711, + "learning_rate": 5.019880715705767e-07, + "loss": 1.2357, + "step": 510 + }, + { + "epoch": 0.5029527559055118, + "grad_norm": 14.718344688415527, + "learning_rate": 5.029821073558649e-07, + "loss": 2.0833, + "step": 511 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 10.352972030639648, + "learning_rate": 5.039761431411532e-07, + "loss": 1.2028, + "step": 512 + }, + { + "epoch": 0.5049212598425197, + "grad_norm": 12.419795989990234, + "learning_rate": 5.049701789264414e-07, + "loss": 1.5992, + "step": 513 + }, + { + "epoch": 0.5059055118110236, + "grad_norm": 10.721029281616211, + "learning_rate": 5.059642147117297e-07, + "loss": 1.0734, + "step": 514 + }, + { + "epoch": 0.5068897637795275, + "grad_norm": 36.97587203979492, + "learning_rate": 5.06958250497018e-07, + "loss": 3.199, + "step": 515 + }, + { + "epoch": 0.5078740157480315, + "grad_norm": 11.912192344665527, + "learning_rate": 5.079522862823062e-07, + "loss": 1.5272, + "step": 516 + }, + { + "epoch": 0.5088582677165354, + "grad_norm": 30.46552848815918, + "learning_rate": 5.089463220675945e-07, + "loss": 2.4856, + "step": 517 + }, + { + "epoch": 0.5098425196850394, + "grad_norm": 11.717628479003906, + "learning_rate": 5.099403578528828e-07, + "loss": 1.6101, + "step": 518 + }, + { + "epoch": 0.5108267716535433, + "grad_norm": 10.724647521972656, + "learning_rate": 5.10934393638171e-07, + "loss": 1.254, + "step": 519 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 10.890359878540039, + "learning_rate": 5.119284294234593e-07, + "loss": 1.3034, + "step": 520 + }, + { + "epoch": 0.5127952755905512, + "grad_norm": 24.375886917114258, + "learning_rate": 5.129224652087476e-07, + "loss": 1.8605, + "step": 521 + }, + { + "epoch": 0.5137795275590551, + "grad_norm": 10.705530166625977, + "learning_rate": 5.139165009940358e-07, + "loss": 1.2163, + "step": 522 + }, + { + "epoch": 0.514763779527559, + "grad_norm": 11.374772071838379, + "learning_rate": 5.149105367793241e-07, + "loss": 1.1579, + "step": 523 + }, + { + "epoch": 0.515748031496063, + "grad_norm": 10.751642227172852, + "learning_rate": 5.159045725646124e-07, + "loss": 1.2257, + "step": 524 + }, + { + "epoch": 0.5167322834645669, + "grad_norm": 11.367935180664062, + "learning_rate": 5.168986083499007e-07, + "loss": 1.5041, + "step": 525 + }, + { + "epoch": 0.5177165354330708, + "grad_norm": 21.515865325927734, + "learning_rate": 5.178926441351889e-07, + "loss": 2.2019, + "step": 526 + }, + { + "epoch": 0.5187007874015748, + "grad_norm": 12.607844352722168, + "learning_rate": 5.188866799204771e-07, + "loss": 1.517, + "step": 527 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 11.534890174865723, + "learning_rate": 5.198807157057656e-07, + "loss": 1.2556, + "step": 528 + }, + { + "epoch": 0.5206692913385826, + "grad_norm": 29.27938461303711, + "learning_rate": 5.208747514910538e-07, + "loss": 3.0446, + "step": 529 + }, + { + "epoch": 0.5216535433070866, + "grad_norm": 13.452887535095215, + "learning_rate": 5.21868787276342e-07, + "loss": 1.3628, + "step": 530 + }, + { + "epoch": 0.5226377952755905, + "grad_norm": 13.540075302124023, + "learning_rate": 5.228628230616304e-07, + "loss": 1.615, + "step": 531 + }, + { + "epoch": 0.5236220472440944, + "grad_norm": 34.353275299072266, + "learning_rate": 5.238568588469186e-07, + "loss": 2.8944, + "step": 532 + }, + { + "epoch": 0.5246062992125984, + "grad_norm": 13.308038711547852, + "learning_rate": 5.248508946322069e-07, + "loss": 1.7452, + "step": 533 + }, + { + "epoch": 0.5255905511811023, + "grad_norm": 10.696693420410156, + "learning_rate": 5.258449304174951e-07, + "loss": 0.9379, + "step": 534 + }, + { + "epoch": 0.5265748031496063, + "grad_norm": 15.371025085449219, + "learning_rate": 5.268389662027834e-07, + "loss": 1.8645, + "step": 535 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 13.851642608642578, + "learning_rate": 5.278330019880717e-07, + "loss": 1.562, + "step": 536 + }, + { + "epoch": 0.5285433070866141, + "grad_norm": 14.475969314575195, + "learning_rate": 5.288270377733599e-07, + "loss": 1.5743, + "step": 537 + }, + { + "epoch": 0.5295275590551181, + "grad_norm": 12.969488143920898, + "learning_rate": 5.298210735586483e-07, + "loss": 1.3699, + "step": 538 + }, + { + "epoch": 0.530511811023622, + "grad_norm": 24.1462345123291, + "learning_rate": 5.308151093439365e-07, + "loss": 2.2463, + "step": 539 + }, + { + "epoch": 0.531496062992126, + "grad_norm": 12.548248291015625, + "learning_rate": 5.318091451292247e-07, + "loss": 1.2225, + "step": 540 + }, + { + "epoch": 0.53248031496063, + "grad_norm": 9.875529289245605, + "learning_rate": 5.32803180914513e-07, + "loss": 0.9797, + "step": 541 + }, + { + "epoch": 0.5334645669291339, + "grad_norm": 14.409546852111816, + "learning_rate": 5.337972166998013e-07, + "loss": 1.3656, + "step": 542 + }, + { + "epoch": 0.5344488188976378, + "grad_norm": 14.300630569458008, + "learning_rate": 5.347912524850895e-07, + "loss": 1.4248, + "step": 543 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 13.680397987365723, + "learning_rate": 5.357852882703778e-07, + "loss": 1.5803, + "step": 544 + }, + { + "epoch": 0.5364173228346457, + "grad_norm": 14.38206672668457, + "learning_rate": 5.367793240556661e-07, + "loss": 1.8014, + "step": 545 + }, + { + "epoch": 0.5374015748031497, + "grad_norm": 46.56623458862305, + "learning_rate": 5.377733598409544e-07, + "loss": 4.1583, + "step": 546 + }, + { + "epoch": 0.5383858267716536, + "grad_norm": 46.82794952392578, + "learning_rate": 5.387673956262426e-07, + "loss": 4.1559, + "step": 547 + }, + { + "epoch": 0.5393700787401575, + "grad_norm": 10.397107124328613, + "learning_rate": 5.397614314115308e-07, + "loss": 1.017, + "step": 548 + }, + { + "epoch": 0.5403543307086615, + "grad_norm": 10.988142967224121, + "learning_rate": 5.407554671968192e-07, + "loss": 1.004, + "step": 549 + }, + { + "epoch": 0.5413385826771654, + "grad_norm": 38.28085708618164, + "learning_rate": 5.417495029821074e-07, + "loss": 3.3581, + "step": 550 + }, + { + "epoch": 0.5423228346456693, + "grad_norm": 12.568646430969238, + "learning_rate": 5.427435387673957e-07, + "loss": 1.1344, + "step": 551 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 11.366560935974121, + "learning_rate": 5.43737574552684e-07, + "loss": 1.3354, + "step": 552 + }, + { + "epoch": 0.5442913385826772, + "grad_norm": 14.407459259033203, + "learning_rate": 5.447316103379722e-07, + "loss": 1.694, + "step": 553 + }, + { + "epoch": 0.5452755905511811, + "grad_norm": 29.43073081970215, + "learning_rate": 5.457256461232605e-07, + "loss": 3.2579, + "step": 554 + }, + { + "epoch": 0.5462598425196851, + "grad_norm": 11.890396118164062, + "learning_rate": 5.467196819085487e-07, + "loss": 1.2556, + "step": 555 + }, + { + "epoch": 0.547244094488189, + "grad_norm": 11.380683898925781, + "learning_rate": 5.477137176938371e-07, + "loss": 1.3283, + "step": 556 + }, + { + "epoch": 0.5482283464566929, + "grad_norm": 10.407194137573242, + "learning_rate": 5.487077534791253e-07, + "loss": 1.0437, + "step": 557 + }, + { + "epoch": 0.5492125984251969, + "grad_norm": 13.227530479431152, + "learning_rate": 5.497017892644135e-07, + "loss": 1.2431, + "step": 558 + }, + { + "epoch": 0.5501968503937008, + "grad_norm": 10.60537052154541, + "learning_rate": 5.50695825049702e-07, + "loss": 1.1519, + "step": 559 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 13.306818008422852, + "learning_rate": 5.516898608349902e-07, + "loss": 1.5753, + "step": 560 + }, + { + "epoch": 0.5521653543307087, + "grad_norm": 10.690176963806152, + "learning_rate": 5.526838966202784e-07, + "loss": 1.3356, + "step": 561 + }, + { + "epoch": 0.5531496062992126, + "grad_norm": 25.577571868896484, + "learning_rate": 5.536779324055667e-07, + "loss": 2.4032, + "step": 562 + }, + { + "epoch": 0.5541338582677166, + "grad_norm": 12.043152809143066, + "learning_rate": 5.54671968190855e-07, + "loss": 1.2915, + "step": 563 + }, + { + "epoch": 0.5551181102362205, + "grad_norm": 30.257396697998047, + "learning_rate": 5.556660039761433e-07, + "loss": 2.6656, + "step": 564 + }, + { + "epoch": 0.5561023622047244, + "grad_norm": 13.785453796386719, + "learning_rate": 5.566600397614315e-07, + "loss": 1.9347, + "step": 565 + }, + { + "epoch": 0.5570866141732284, + "grad_norm": 16.543371200561523, + "learning_rate": 5.576540755467198e-07, + "loss": 1.7852, + "step": 566 + }, + { + "epoch": 0.5580708661417323, + "grad_norm": 15.194260597229004, + "learning_rate": 5.586481113320081e-07, + "loss": 1.5072, + "step": 567 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 13.929033279418945, + "learning_rate": 5.596421471172963e-07, + "loss": 1.507, + "step": 568 + }, + { + "epoch": 0.5600393700787402, + "grad_norm": 11.420060157775879, + "learning_rate": 5.606361829025846e-07, + "loss": 1.0272, + "step": 569 + }, + { + "epoch": 0.5610236220472441, + "grad_norm": 13.144247055053711, + "learning_rate": 5.616302186878729e-07, + "loss": 1.4452, + "step": 570 + }, + { + "epoch": 0.562007874015748, + "grad_norm": 27.946935653686523, + "learning_rate": 5.626242544731611e-07, + "loss": 2.9127, + "step": 571 + }, + { + "epoch": 0.562992125984252, + "grad_norm": 11.884686470031738, + "learning_rate": 5.636182902584494e-07, + "loss": 0.9656, + "step": 572 + }, + { + "epoch": 0.5639763779527559, + "grad_norm": 11.995596885681152, + "learning_rate": 5.646123260437377e-07, + "loss": 1.0789, + "step": 573 + }, + { + "epoch": 0.5649606299212598, + "grad_norm": 11.087538719177246, + "learning_rate": 5.656063618290259e-07, + "loss": 1.0709, + "step": 574 + }, + { + "epoch": 0.5659448818897638, + "grad_norm": 27.41124725341797, + "learning_rate": 5.666003976143142e-07, + "loss": 2.4211, + "step": 575 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 11.643184661865234, + "learning_rate": 5.675944333996024e-07, + "loss": 0.9606, + "step": 576 + }, + { + "epoch": 0.5679133858267716, + "grad_norm": 12.335195541381836, + "learning_rate": 5.685884691848908e-07, + "loss": 1.2229, + "step": 577 + }, + { + "epoch": 0.5688976377952756, + "grad_norm": 13.128782272338867, + "learning_rate": 5.69582504970179e-07, + "loss": 1.1131, + "step": 578 + }, + { + "epoch": 0.5698818897637795, + "grad_norm": 14.412220001220703, + "learning_rate": 5.705765407554672e-07, + "loss": 1.4448, + "step": 579 + }, + { + "epoch": 0.5708661417322834, + "grad_norm": 15.162165641784668, + "learning_rate": 5.715705765407556e-07, + "loss": 1.4289, + "step": 580 + }, + { + "epoch": 0.5718503937007874, + "grad_norm": 27.540359497070312, + "learning_rate": 5.725646123260438e-07, + "loss": 2.4415, + "step": 581 + }, + { + "epoch": 0.5728346456692913, + "grad_norm": 12.708159446716309, + "learning_rate": 5.735586481113321e-07, + "loss": 0.9805, + "step": 582 + }, + { + "epoch": 0.5738188976377953, + "grad_norm": 10.591647148132324, + "learning_rate": 5.745526838966203e-07, + "loss": 0.8534, + "step": 583 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 32.671504974365234, + "learning_rate": 5.755467196819086e-07, + "loss": 2.9736, + "step": 584 + }, + { + "epoch": 0.5757874015748031, + "grad_norm": 30.118398666381836, + "learning_rate": 5.765407554671969e-07, + "loss": 2.5249, + "step": 585 + }, + { + "epoch": 0.5767716535433071, + "grad_norm": 10.852633476257324, + "learning_rate": 5.775347912524851e-07, + "loss": 0.8731, + "step": 586 + }, + { + "epoch": 0.577755905511811, + "grad_norm": 11.627351760864258, + "learning_rate": 5.785288270377734e-07, + "loss": 1.1817, + "step": 587 + }, + { + "epoch": 0.5787401574803149, + "grad_norm": 11.62166690826416, + "learning_rate": 5.795228628230618e-07, + "loss": 1.0803, + "step": 588 + }, + { + "epoch": 0.5797244094488189, + "grad_norm": 11.084623336791992, + "learning_rate": 5.8051689860835e-07, + "loss": 0.9559, + "step": 589 + }, + { + "epoch": 0.5807086614173228, + "grad_norm": 28.948238372802734, + "learning_rate": 5.815109343936383e-07, + "loss": 3.0132, + "step": 590 + }, + { + "epoch": 0.5816929133858267, + "grad_norm": 11.215713500976562, + "learning_rate": 5.825049701789266e-07, + "loss": 1.3511, + "step": 591 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 22.789833068847656, + "learning_rate": 5.834990059642148e-07, + "loss": 2.4456, + "step": 592 + }, + { + "epoch": 0.5836614173228346, + "grad_norm": 9.721744537353516, + "learning_rate": 5.844930417495031e-07, + "loss": 0.8408, + "step": 593 + }, + { + "epoch": 0.5846456692913385, + "grad_norm": 11.74154281616211, + "learning_rate": 5.854870775347914e-07, + "loss": 1.2301, + "step": 594 + }, + { + "epoch": 0.5856299212598425, + "grad_norm": 11.753499031066895, + "learning_rate": 5.864811133200797e-07, + "loss": 1.1377, + "step": 595 + }, + { + "epoch": 0.5866141732283464, + "grad_norm": 22.84195899963379, + "learning_rate": 5.874751491053679e-07, + "loss": 1.8652, + "step": 596 + }, + { + "epoch": 0.5875984251968503, + "grad_norm": 12.816361427307129, + "learning_rate": 5.884691848906561e-07, + "loss": 1.4308, + "step": 597 + }, + { + "epoch": 0.5885826771653543, + "grad_norm": 9.53377628326416, + "learning_rate": 5.894632206759445e-07, + "loss": 1.06, + "step": 598 + }, + { + "epoch": 0.5895669291338582, + "grad_norm": 10.471661567687988, + "learning_rate": 5.904572564612327e-07, + "loss": 1.1925, + "step": 599 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 24.913503646850586, + "learning_rate": 5.914512922465209e-07, + "loss": 1.4629, + "step": 600 + }, + { + "epoch": 0.5915354330708661, + "grad_norm": 22.783151626586914, + "learning_rate": 5.924453280318093e-07, + "loss": 2.6618, + "step": 601 + }, + { + "epoch": 0.59251968503937, + "grad_norm": 10.51607894897461, + "learning_rate": 5.934393638170975e-07, + "loss": 1.1027, + "step": 602 + }, + { + "epoch": 0.593503937007874, + "grad_norm": 10.741540908813477, + "learning_rate": 5.944333996023858e-07, + "loss": 1.3149, + "step": 603 + }, + { + "epoch": 0.594488188976378, + "grad_norm": 10.431660652160645, + "learning_rate": 5.95427435387674e-07, + "loss": 0.8886, + "step": 604 + }, + { + "epoch": 0.5954724409448819, + "grad_norm": 12.338500022888184, + "learning_rate": 5.964214711729623e-07, + "loss": 1.3366, + "step": 605 + }, + { + "epoch": 0.5964566929133859, + "grad_norm": 13.88163948059082, + "learning_rate": 5.974155069582506e-07, + "loss": 1.356, + "step": 606 + }, + { + "epoch": 0.5974409448818898, + "grad_norm": 14.141697883605957, + "learning_rate": 5.984095427435388e-07, + "loss": 1.28, + "step": 607 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 29.235586166381836, + "learning_rate": 5.994035785288272e-07, + "loss": 2.7501, + "step": 608 + }, + { + "epoch": 0.5994094488188977, + "grad_norm": 10.489766120910645, + "learning_rate": 6.003976143141154e-07, + "loss": 0.9256, + "step": 609 + }, + { + "epoch": 0.6003937007874016, + "grad_norm": 11.40336799621582, + "learning_rate": 6.013916500994036e-07, + "loss": 0.8477, + "step": 610 + }, + { + "epoch": 0.6013779527559056, + "grad_norm": 12.08163070678711, + "learning_rate": 6.02385685884692e-07, + "loss": 1.1828, + "step": 611 + }, + { + "epoch": 0.6023622047244095, + "grad_norm": 12.5928373336792, + "learning_rate": 6.033797216699802e-07, + "loss": 1.2434, + "step": 612 + }, + { + "epoch": 0.6023622047244095, + "eval_Qnli-dev_cosine_accuracy": 0.654296875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8779141902923584, + "eval_Qnli-dev_cosine_ap": 0.656311858267534, + "eval_Qnli-dev_cosine_f1": 0.6475849731663685, + "eval_Qnli-dev_cosine_f1_threshold": 0.8221863508224487, + "eval_Qnli-dev_cosine_precision": 0.5603715170278638, + "eval_Qnli-dev_cosine_recall": 0.7669491525423728, + "eval_Qnli-dev_dot_accuracy": 0.65625, + "eval_Qnli-dev_dot_accuracy_threshold": 671.3241577148438, + "eval_Qnli-dev_dot_ap": 0.6532475470373437, + "eval_Qnli-dev_dot_f1": 0.6480836236933798, + "eval_Qnli-dev_dot_f1_threshold": 625.4417724609375, + "eval_Qnli-dev_dot_precision": 0.5502958579881657, + "eval_Qnli-dev_dot_recall": 0.788135593220339, + "eval_Qnli-dev_euclidean_accuracy": 0.65234375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.653518676757812, + "eval_Qnli-dev_euclidean_ap": 0.6564392688132011, + "eval_Qnli-dev_euclidean_f1": 0.6475849731663685, + "eval_Qnli-dev_euclidean_f1_threshold": 16.493606567382812, + "eval_Qnli-dev_euclidean_precision": 0.5603715170278638, + "eval_Qnli-dev_euclidean_recall": 0.7669491525423728, + "eval_Qnli-dev_manhattan_accuracy": 0.65234375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 277.54986572265625, + "eval_Qnli-dev_manhattan_ap": 0.6627734722297267, + "eval_Qnli-dev_manhattan_f1": 0.6550522648083624, + "eval_Qnli-dev_manhattan_f1_threshold": 342.1537780761719, + "eval_Qnli-dev_manhattan_precision": 0.5562130177514792, + "eval_Qnli-dev_manhattan_recall": 0.7966101694915254, + "eval_Qnli-dev_max_accuracy": 0.65625, + "eval_Qnli-dev_max_accuracy_threshold": 671.3241577148438, + "eval_Qnli-dev_max_ap": 0.6627734722297267, + "eval_Qnli-dev_max_f1": 0.6550522648083624, + "eval_Qnli-dev_max_f1_threshold": 625.4417724609375, + "eval_Qnli-dev_max_precision": 0.5603715170278638, + "eval_Qnli-dev_max_recall": 0.7966101694915254, + "eval_allNLI-dev_cosine_accuracy": 0.673828125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9601989388465881, + "eval_allNLI-dev_cosine_ap": 0.4528401952578379, + "eval_allNLI-dev_cosine_f1": 0.5136436597110755, + "eval_allNLI-dev_cosine_f1_threshold": 0.8315722942352295, + "eval_allNLI-dev_cosine_precision": 0.35555555555555557, + "eval_allNLI-dev_cosine_recall": 0.9248554913294798, + "eval_allNLI-dev_dot_accuracy": 0.67578125, + "eval_allNLI-dev_dot_accuracy_threshold": 736.3153686523438, + "eval_allNLI-dev_dot_ap": 0.45325627248151745, + "eval_allNLI-dev_dot_f1": 0.5136436597110755, + "eval_allNLI-dev_dot_f1_threshold": 636.7808837890625, + "eval_allNLI-dev_dot_precision": 0.35555555555555557, + "eval_allNLI-dev_dot_recall": 0.9248554913294798, + "eval_allNLI-dev_euclidean_accuracy": 0.673828125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 7.8111114501953125, + "eval_allNLI-dev_euclidean_ap": 0.4529571995907292, + "eval_allNLI-dev_euclidean_f1": 0.5136436597110755, + "eval_allNLI-dev_euclidean_f1_threshold": 16.056068420410156, + "eval_allNLI-dev_euclidean_precision": 0.35555555555555557, + "eval_allNLI-dev_euclidean_recall": 0.9248554913294798, + "eval_allNLI-dev_manhattan_accuracy": 0.673828125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 177.4397735595703, + "eval_allNLI-dev_manhattan_ap": 0.45384369486637116, + "eval_allNLI-dev_manhattan_f1": 0.5263157894736843, + "eval_allNLI-dev_manhattan_f1_threshold": 267.63739013671875, + "eval_allNLI-dev_manhattan_precision": 0.39705882352941174, + "eval_allNLI-dev_manhattan_recall": 0.7803468208092486, + "eval_allNLI-dev_max_accuracy": 0.67578125, + "eval_allNLI-dev_max_accuracy_threshold": 736.3153686523438, + "eval_allNLI-dev_max_ap": 0.45384369486637116, + "eval_allNLI-dev_max_f1": 0.5263157894736843, + "eval_allNLI-dev_max_f1_threshold": 636.7808837890625, + "eval_allNLI-dev_max_precision": 0.39705882352941174, + "eval_allNLI-dev_max_recall": 0.9248554913294798, + "eval_loss": 1.2393934726715088, + "eval_runtime": 51.413, + "eval_samples_per_second": 32.365, + "eval_sequential_score": 0.6627734722297267, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.3425124347586146, + "eval_sts-test_pearson_dot": 0.3379845466930417, + "eval_sts-test_pearson_euclidean": 0.3693839678681028, + "eval_sts-test_pearson_manhattan": 0.37980012705427046, + "eval_sts-test_pearson_max": 0.37980012705427046, + "eval_sts-test_spearman_cosine": 0.39493399006765173, + "eval_sts-test_spearman_dot": 0.38992618977610344, + "eval_sts-test_spearman_euclidean": 0.3953308889960606, + "eval_sts-test_spearman_manhattan": 0.40510223719994237, + "eval_sts-test_spearman_max": 0.40510223719994237, + "step": 612 + }, + { + "epoch": 0.6033464566929134, + "grad_norm": 9.596250534057617, + "learning_rate": 6.043737574552684e-07, + "loss": 0.9043, + "step": 613 + }, + { + "epoch": 0.6043307086614174, + "grad_norm": 10.514294624328613, + "learning_rate": 6.053677932405567e-07, + "loss": 0.8613, + "step": 614 + }, + { + "epoch": 0.6053149606299213, + "grad_norm": 12.741471290588379, + "learning_rate": 6.06361829025845e-07, + "loss": 1.2901, + "step": 615 + }, + { + "epoch": 0.6062992125984252, + "grad_norm": 13.291648864746094, + "learning_rate": 6.073558648111333e-07, + "loss": 1.2323, + "step": 616 + }, + { + "epoch": 0.6072834645669292, + "grad_norm": 13.98913860321045, + "learning_rate": 6.083499005964215e-07, + "loss": 1.3886, + "step": 617 + }, + { + "epoch": 0.6082677165354331, + "grad_norm": 14.959402084350586, + "learning_rate": 6.093439363817097e-07, + "loss": 1.7351, + "step": 618 + }, + { + "epoch": 0.609251968503937, + "grad_norm": 49.22218322753906, + "learning_rate": 6.103379721669982e-07, + "loss": 4.0782, + "step": 619 + }, + { + "epoch": 0.610236220472441, + "grad_norm": 12.091609001159668, + "learning_rate": 6.113320079522864e-07, + "loss": 0.9729, + "step": 620 + }, + { + "epoch": 0.6112204724409449, + "grad_norm": 11.08301830291748, + "learning_rate": 6.123260437375747e-07, + "loss": 1.0137, + "step": 621 + }, + { + "epoch": 0.6122047244094488, + "grad_norm": 13.69589900970459, + "learning_rate": 6.13320079522863e-07, + "loss": 1.0514, + "step": 622 + }, + { + "epoch": 0.6131889763779528, + "grad_norm": 30.808237075805664, + "learning_rate": 6.143141153081512e-07, + "loss": 3.1936, + "step": 623 + }, + { + "epoch": 0.6141732283464567, + "grad_norm": 13.483572006225586, + "learning_rate": 6.153081510934395e-07, + "loss": 1.2081, + "step": 624 + }, + { + "epoch": 0.6151574803149606, + "grad_norm": 26.108518600463867, + "learning_rate": 6.163021868787277e-07, + "loss": 1.6647, + "step": 625 + }, + { + "epoch": 0.6161417322834646, + "grad_norm": 13.253960609436035, + "learning_rate": 6.172962226640161e-07, + "loss": 1.1733, + "step": 626 + }, + { + "epoch": 0.6171259842519685, + "grad_norm": 12.534513473510742, + "learning_rate": 6.182902584493043e-07, + "loss": 1.8176, + "step": 627 + }, + { + "epoch": 0.6181102362204725, + "grad_norm": 11.345523834228516, + "learning_rate": 6.192842942345925e-07, + "loss": 1.1713, + "step": 628 + }, + { + "epoch": 0.6190944881889764, + "grad_norm": 29.95690155029297, + "learning_rate": 6.202783300198809e-07, + "loss": 2.7953, + "step": 629 + }, + { + "epoch": 0.6200787401574803, + "grad_norm": 12.109441757202148, + "learning_rate": 6.212723658051691e-07, + "loss": 1.0633, + "step": 630 + }, + { + "epoch": 0.6210629921259843, + "grad_norm": 11.02467155456543, + "learning_rate": 6.222664015904573e-07, + "loss": 1.0282, + "step": 631 + }, + { + "epoch": 0.6220472440944882, + "grad_norm": 14.736310005187988, + "learning_rate": 6.232604373757457e-07, + "loss": 2.0119, + "step": 632 + }, + { + "epoch": 0.6230314960629921, + "grad_norm": 11.843450546264648, + "learning_rate": 6.242544731610339e-07, + "loss": 1.2355, + "step": 633 + }, + { + "epoch": 0.6240157480314961, + "grad_norm": 10.010743141174316, + "learning_rate": 6.252485089463222e-07, + "loss": 0.8702, + "step": 634 + }, + { + "epoch": 0.625, + "grad_norm": 23.628681182861328, + "learning_rate": 6.262425447316104e-07, + "loss": 1.3692, + "step": 635 + }, + { + "epoch": 0.6259842519685039, + "grad_norm": 9.567180633544922, + "learning_rate": 6.272365805168987e-07, + "loss": 1.034, + "step": 636 + }, + { + "epoch": 0.6269685039370079, + "grad_norm": 11.429587364196777, + "learning_rate": 6.28230616302187e-07, + "loss": 1.0292, + "step": 637 + }, + { + "epoch": 0.6279527559055118, + "grad_norm": 12.177762985229492, + "learning_rate": 6.292246520874752e-07, + "loss": 1.1059, + "step": 638 + }, + { + "epoch": 0.6289370078740157, + "grad_norm": 12.012341499328613, + "learning_rate": 6.302186878727636e-07, + "loss": 1.2945, + "step": 639 + }, + { + "epoch": 0.6299212598425197, + "grad_norm": 25.510358810424805, + "learning_rate": 6.312127236580518e-07, + "loss": 1.9812, + "step": 640 + }, + { + "epoch": 0.6309055118110236, + "grad_norm": 11.094367980957031, + "learning_rate": 6.3220675944334e-07, + "loss": 0.8846, + "step": 641 + }, + { + "epoch": 0.6318897637795275, + "grad_norm": 13.495521545410156, + "learning_rate": 6.332007952286283e-07, + "loss": 1.3874, + "step": 642 + }, + { + "epoch": 0.6328740157480315, + "grad_norm": 10.81843090057373, + "learning_rate": 6.341948310139166e-07, + "loss": 1.0868, + "step": 643 + }, + { + "epoch": 0.6338582677165354, + "grad_norm": 11.740157127380371, + "learning_rate": 6.351888667992048e-07, + "loss": 1.0356, + "step": 644 + }, + { + "epoch": 0.6348425196850394, + "grad_norm": 11.699848175048828, + "learning_rate": 6.361829025844931e-07, + "loss": 1.229, + "step": 645 + }, + { + "epoch": 0.6358267716535433, + "grad_norm": 59.38969802856445, + "learning_rate": 6.371769383697813e-07, + "loss": 3.729, + "step": 646 + }, + { + "epoch": 0.6368110236220472, + "grad_norm": 13.450605392456055, + "learning_rate": 6.381709741550698e-07, + "loss": 1.1382, + "step": 647 + }, + { + "epoch": 0.6377952755905512, + "grad_norm": 11.111218452453613, + "learning_rate": 6.39165009940358e-07, + "loss": 1.17, + "step": 648 + }, + { + "epoch": 0.6387795275590551, + "grad_norm": 9.77136516571045, + "learning_rate": 6.401590457256462e-07, + "loss": 0.6846, + "step": 649 + }, + { + "epoch": 0.639763779527559, + "grad_norm": 42.0877571105957, + "learning_rate": 6.411530815109346e-07, + "loss": 3.0024, + "step": 650 + }, + { + "epoch": 0.640748031496063, + "grad_norm": 11.444602012634277, + "learning_rate": 6.421471172962228e-07, + "loss": 0.8964, + "step": 651 + }, + { + "epoch": 0.6417322834645669, + "grad_norm": 12.272961616516113, + "learning_rate": 6.431411530815111e-07, + "loss": 1.4124, + "step": 652 + }, + { + "epoch": 0.6427165354330708, + "grad_norm": 11.764492988586426, + "learning_rate": 6.441351888667994e-07, + "loss": 1.1492, + "step": 653 + }, + { + "epoch": 0.6437007874015748, + "grad_norm": 15.19382381439209, + "learning_rate": 6.451292246520876e-07, + "loss": 1.2041, + "step": 654 + }, + { + "epoch": 0.6446850393700787, + "grad_norm": 13.139802932739258, + "learning_rate": 6.461232604373759e-07, + "loss": 1.3567, + "step": 655 + }, + { + "epoch": 0.6456692913385826, + "grad_norm": 13.58987045288086, + "learning_rate": 6.471172962226641e-07, + "loss": 1.2777, + "step": 656 + }, + { + "epoch": 0.6466535433070866, + "grad_norm": 12.390425682067871, + "learning_rate": 6.481113320079524e-07, + "loss": 1.1406, + "step": 657 + }, + { + "epoch": 0.6476377952755905, + "grad_norm": 11.243461608886719, + "learning_rate": 6.491053677932407e-07, + "loss": 0.9606, + "step": 658 + }, + { + "epoch": 0.6486220472440944, + "grad_norm": 10.577733039855957, + "learning_rate": 6.500994035785289e-07, + "loss": 0.8331, + "step": 659 + }, + { + "epoch": 0.6496062992125984, + "grad_norm": 33.58454513549805, + "learning_rate": 6.510934393638173e-07, + "loss": 3.4216, + "step": 660 + }, + { + "epoch": 0.6505905511811023, + "grad_norm": 11.1649808883667, + "learning_rate": 6.520874751491055e-07, + "loss": 0.8534, + "step": 661 + }, + { + "epoch": 0.6515748031496063, + "grad_norm": 11.563724517822266, + "learning_rate": 6.530815109343937e-07, + "loss": 1.0986, + "step": 662 + }, + { + "epoch": 0.6525590551181102, + "grad_norm": 9.949313163757324, + "learning_rate": 6.54075546719682e-07, + "loss": 0.7135, + "step": 663 + }, + { + "epoch": 0.6535433070866141, + "grad_norm": 11.389878273010254, + "learning_rate": 6.550695825049703e-07, + "loss": 1.0527, + "step": 664 + }, + { + "epoch": 0.6545275590551181, + "grad_norm": 10.103633880615234, + "learning_rate": 6.560636182902586e-07, + "loss": 0.7865, + "step": 665 + }, + { + "epoch": 0.655511811023622, + "grad_norm": 33.22819900512695, + "learning_rate": 6.570576540755468e-07, + "loss": 3.4573, + "step": 666 + }, + { + "epoch": 0.656496062992126, + "grad_norm": 14.148033142089844, + "learning_rate": 6.580516898608351e-07, + "loss": 1.1366, + "step": 667 + }, + { + "epoch": 0.65748031496063, + "grad_norm": 29.169261932373047, + "learning_rate": 6.590457256461234e-07, + "loss": 2.2528, + "step": 668 + }, + { + "epoch": 0.6584645669291339, + "grad_norm": 11.599723815917969, + "learning_rate": 6.600397614314116e-07, + "loss": 0.9883, + "step": 669 + }, + { + "epoch": 0.6594488188976378, + "grad_norm": 8.977407455444336, + "learning_rate": 6.610337972166998e-07, + "loss": 0.6787, + "step": 670 + }, + { + "epoch": 0.6604330708661418, + "grad_norm": 30.045438766479492, + "learning_rate": 6.620278330019882e-07, + "loss": 3.66, + "step": 671 + }, + { + "epoch": 0.6614173228346457, + "grad_norm": 12.010369300842285, + "learning_rate": 6.630218687872764e-07, + "loss": 1.092, + "step": 672 + }, + { + "epoch": 0.6624015748031497, + "grad_norm": 14.737366676330566, + "learning_rate": 6.640159045725647e-07, + "loss": 1.7895, + "step": 673 + }, + { + "epoch": 0.6633858267716536, + "grad_norm": 12.320960998535156, + "learning_rate": 6.65009940357853e-07, + "loss": 0.967, + "step": 674 + }, + { + "epoch": 0.6643700787401575, + "grad_norm": 13.666062355041504, + "learning_rate": 6.660039761431412e-07, + "loss": 0.7553, + "step": 675 + }, + { + "epoch": 0.6653543307086615, + "grad_norm": 20.1140193939209, + "learning_rate": 6.669980119284295e-07, + "loss": 1.3217, + "step": 676 + }, + { + "epoch": 0.6663385826771654, + "grad_norm": 13.900456428527832, + "learning_rate": 6.679920477137177e-07, + "loss": 1.2246, + "step": 677 + }, + { + "epoch": 0.6673228346456693, + "grad_norm": 12.594614028930664, + "learning_rate": 6.689860834990062e-07, + "loss": 1.242, + "step": 678 + }, + { + "epoch": 0.6683070866141733, + "grad_norm": 23.472980499267578, + "learning_rate": 6.699801192842944e-07, + "loss": 2.1658, + "step": 679 + }, + { + "epoch": 0.6692913385826772, + "grad_norm": 18.802461624145508, + "learning_rate": 6.709741550695826e-07, + "loss": 1.4381, + "step": 680 + }, + { + "epoch": 0.6702755905511811, + "grad_norm": 12.633712768554688, + "learning_rate": 6.71968190854871e-07, + "loss": 0.9602, + "step": 681 + }, + { + "epoch": 0.6712598425196851, + "grad_norm": 11.526813507080078, + "learning_rate": 6.729622266401592e-07, + "loss": 1.0741, + "step": 682 + }, + { + "epoch": 0.672244094488189, + "grad_norm": 11.790997505187988, + "learning_rate": 6.739562624254474e-07, + "loss": 1.3184, + "step": 683 + }, + { + "epoch": 0.6732283464566929, + "grad_norm": 11.810404777526855, + "learning_rate": 6.749502982107357e-07, + "loss": 0.9802, + "step": 684 + }, + { + "epoch": 0.6742125984251969, + "grad_norm": 14.490528106689453, + "learning_rate": 6.75944333996024e-07, + "loss": 1.144, + "step": 685 + }, + { + "epoch": 0.6751968503937008, + "grad_norm": 11.788384437561035, + "learning_rate": 6.769383697813123e-07, + "loss": 1.1724, + "step": 686 + }, + { + "epoch": 0.6761811023622047, + "grad_norm": 11.866044044494629, + "learning_rate": 6.779324055666005e-07, + "loss": 1.0144, + "step": 687 + }, + { + "epoch": 0.6771653543307087, + "grad_norm": 24.21769142150879, + "learning_rate": 6.789264413518888e-07, + "loss": 2.2625, + "step": 688 + }, + { + "epoch": 0.6781496062992126, + "grad_norm": 12.545021057128906, + "learning_rate": 6.799204771371771e-07, + "loss": 1.3074, + "step": 689 + }, + { + "epoch": 0.6791338582677166, + "grad_norm": 21.05637550354004, + "learning_rate": 6.809145129224653e-07, + "loss": 1.5017, + "step": 690 + }, + { + "epoch": 0.6801181102362205, + "grad_norm": 12.763387680053711, + "learning_rate": 6.819085487077536e-07, + "loss": 0.9315, + "step": 691 + }, + { + "epoch": 0.6811023622047244, + "grad_norm": 9.738778114318848, + "learning_rate": 6.829025844930419e-07, + "loss": 0.6666, + "step": 692 + }, + { + "epoch": 0.6820866141732284, + "grad_norm": 12.113200187683105, + "learning_rate": 6.838966202783301e-07, + "loss": 1.2388, + "step": 693 + }, + { + "epoch": 0.6830708661417323, + "grad_norm": 9.824642181396484, + "learning_rate": 6.848906560636184e-07, + "loss": 0.7314, + "step": 694 + }, + { + "epoch": 0.6840551181102362, + "grad_norm": 35.968177795410156, + "learning_rate": 6.858846918489067e-07, + "loss": 2.5126, + "step": 695 + }, + { + "epoch": 0.6850393700787402, + "grad_norm": 11.211616516113281, + "learning_rate": 6.86878727634195e-07, + "loss": 1.3168, + "step": 696 + }, + { + "epoch": 0.6860236220472441, + "grad_norm": 10.734878540039062, + "learning_rate": 6.878727634194832e-07, + "loss": 0.8281, + "step": 697 + }, + { + "epoch": 0.687007874015748, + "grad_norm": 21.839208602905273, + "learning_rate": 6.888667992047714e-07, + "loss": 2.5375, + "step": 698 + }, + { + "epoch": 0.687992125984252, + "grad_norm": 13.58681869506836, + "learning_rate": 6.898608349900598e-07, + "loss": 1.3509, + "step": 699 + }, + { + "epoch": 0.6889763779527559, + "grad_norm": 9.648022651672363, + "learning_rate": 6.90854870775348e-07, + "loss": 0.802, + "step": 700 + }, + { + "epoch": 0.6899606299212598, + "grad_norm": 11.510672569274902, + "learning_rate": 6.918489065606362e-07, + "loss": 1.0974, + "step": 701 + }, + { + "epoch": 0.6909448818897638, + "grad_norm": 22.108243942260742, + "learning_rate": 6.928429423459246e-07, + "loss": 2.3393, + "step": 702 + }, + { + "epoch": 0.6919291338582677, + "grad_norm": 11.486313819885254, + "learning_rate": 6.938369781312128e-07, + "loss": 1.6287, + "step": 703 + }, + { + "epoch": 0.6929133858267716, + "grad_norm": 10.629705429077148, + "learning_rate": 6.948310139165011e-07, + "loss": 1.1297, + "step": 704 + }, + { + "epoch": 0.6938976377952756, + "grad_norm": 10.482368469238281, + "learning_rate": 6.958250497017893e-07, + "loss": 0.9367, + "step": 705 + }, + { + "epoch": 0.6948818897637795, + "grad_norm": 11.394485473632812, + "learning_rate": 6.968190854870776e-07, + "loss": 0.8724, + "step": 706 + }, + { + "epoch": 0.6958661417322834, + "grad_norm": 10.056958198547363, + "learning_rate": 6.97813121272366e-07, + "loss": 0.7766, + "step": 707 + }, + { + "epoch": 0.6968503937007874, + "grad_norm": 17.666093826293945, + "learning_rate": 6.988071570576542e-07, + "loss": 1.5104, + "step": 708 + }, + { + "epoch": 0.6978346456692913, + "grad_norm": 14.270581245422363, + "learning_rate": 6.998011928429426e-07, + "loss": 1.2963, + "step": 709 + }, + { + "epoch": 0.6988188976377953, + "grad_norm": 15.917206764221191, + "learning_rate": 7.007952286282308e-07, + "loss": 1.3217, + "step": 710 + }, + { + "epoch": 0.6998031496062992, + "grad_norm": 12.223875045776367, + "learning_rate": 7.01789264413519e-07, + "loss": 1.511, + "step": 711 + }, + { + "epoch": 0.7007874015748031, + "grad_norm": 14.500188827514648, + "learning_rate": 7.027833001988073e-07, + "loss": 1.2841, + "step": 712 + }, + { + "epoch": 0.7017716535433071, + "grad_norm": 30.41779136657715, + "learning_rate": 7.037773359840956e-07, + "loss": 2.2693, + "step": 713 + }, + { + "epoch": 0.702755905511811, + "grad_norm": 12.932038307189941, + "learning_rate": 7.047713717693838e-07, + "loss": 1.1542, + "step": 714 + }, + { + "epoch": 0.7037401574803149, + "grad_norm": 16.0056095123291, + "learning_rate": 7.057654075546721e-07, + "loss": 1.3174, + "step": 715 + }, + { + "epoch": 0.7047244094488189, + "grad_norm": 25.923389434814453, + "learning_rate": 7.067594433399604e-07, + "loss": 2.4071, + "step": 716 + }, + { + "epoch": 0.7057086614173228, + "grad_norm": 14.635533332824707, + "learning_rate": 7.077534791252487e-07, + "loss": 1.9443, + "step": 717 + }, + { + "epoch": 0.7066929133858267, + "grad_norm": 12.386008262634277, + "learning_rate": 7.087475149105369e-07, + "loss": 1.2466, + "step": 718 + }, + { + "epoch": 0.7076771653543307, + "grad_norm": 21.118125915527344, + "learning_rate": 7.097415506958251e-07, + "loss": 1.702, + "step": 719 + }, + { + "epoch": 0.7086614173228346, + "grad_norm": 16.153949737548828, + "learning_rate": 7.107355864811135e-07, + "loss": 1.9423, + "step": 720 + }, + { + "epoch": 0.7096456692913385, + "grad_norm": 9.497204780578613, + "learning_rate": 7.117296222664017e-07, + "loss": 0.7641, + "step": 721 + }, + { + "epoch": 0.7106299212598425, + "grad_norm": 10.133769989013672, + "learning_rate": 7.1272365805169e-07, + "loss": 0.8676, + "step": 722 + }, + { + "epoch": 0.7116141732283464, + "grad_norm": 10.550159454345703, + "learning_rate": 7.137176938369783e-07, + "loss": 0.8088, + "step": 723 + }, + { + "epoch": 0.7125984251968503, + "grad_norm": 10.690868377685547, + "learning_rate": 7.147117296222665e-07, + "loss": 1.1074, + "step": 724 + }, + { + "epoch": 0.7135826771653543, + "grad_norm": 10.60130786895752, + "learning_rate": 7.157057654075548e-07, + "loss": 0.8919, + "step": 725 + }, + { + "epoch": 0.7145669291338582, + "grad_norm": 21.287555694580078, + "learning_rate": 7.16699801192843e-07, + "loss": 1.7894, + "step": 726 + }, + { + "epoch": 0.7155511811023622, + "grad_norm": 15.752532005310059, + "learning_rate": 7.176938369781313e-07, + "loss": 1.1245, + "step": 727 + }, + { + "epoch": 0.7165354330708661, + "grad_norm": 10.15629768371582, + "learning_rate": 7.186878727634196e-07, + "loss": 0.7873, + "step": 728 + }, + { + "epoch": 0.71751968503937, + "grad_norm": 13.381704330444336, + "learning_rate": 7.196819085487078e-07, + "loss": 1.8204, + "step": 729 + }, + { + "epoch": 0.718503937007874, + "grad_norm": 11.077107429504395, + "learning_rate": 7.206759443339962e-07, + "loss": 1.1784, + "step": 730 + }, + { + "epoch": 0.719488188976378, + "grad_norm": 10.849299430847168, + "learning_rate": 7.216699801192844e-07, + "loss": 0.8938, + "step": 731 + }, + { + "epoch": 0.7204724409448819, + "grad_norm": 8.247758865356445, + "learning_rate": 7.226640159045726e-07, + "loss": 0.6575, + "step": 732 + }, + { + "epoch": 0.7214566929133859, + "grad_norm": 22.916067123413086, + "learning_rate": 7.236580516898609e-07, + "loss": 1.4965, + "step": 733 + }, + { + "epoch": 0.7224409448818898, + "grad_norm": 23.232479095458984, + "learning_rate": 7.246520874751492e-07, + "loss": 1.8752, + "step": 734 + }, + { + "epoch": 0.7234251968503937, + "grad_norm": 13.055563926696777, + "learning_rate": 7.256461232604375e-07, + "loss": 1.2691, + "step": 735 + }, + { + "epoch": 0.7244094488188977, + "grad_norm": 11.828513145446777, + "learning_rate": 7.266401590457257e-07, + "loss": 0.8873, + "step": 736 + }, + { + "epoch": 0.7253937007874016, + "grad_norm": 11.598047256469727, + "learning_rate": 7.27634194831014e-07, + "loss": 0.9377, + "step": 737 + }, + { + "epoch": 0.7263779527559056, + "grad_norm": 8.214035034179688, + "learning_rate": 7.286282306163024e-07, + "loss": 0.5962, + "step": 738 + }, + { + "epoch": 0.7273622047244095, + "grad_norm": 11.756397247314453, + "learning_rate": 7.296222664015906e-07, + "loss": 0.9877, + "step": 739 + }, + { + "epoch": 0.7283464566929134, + "grad_norm": 9.81069278717041, + "learning_rate": 7.306163021868788e-07, + "loss": 0.9054, + "step": 740 + }, + { + "epoch": 0.7293307086614174, + "grad_norm": 10.654214859008789, + "learning_rate": 7.316103379721672e-07, + "loss": 0.8589, + "step": 741 + }, + { + "epoch": 0.7303149606299213, + "grad_norm": 12.728946685791016, + "learning_rate": 7.326043737574554e-07, + "loss": 1.2914, + "step": 742 + }, + { + "epoch": 0.7312992125984252, + "grad_norm": 12.277595520019531, + "learning_rate": 7.335984095427437e-07, + "loss": 0.9707, + "step": 743 + }, + { + "epoch": 0.7322834645669292, + "grad_norm": 12.706618309020996, + "learning_rate": 7.34592445328032e-07, + "loss": 1.0986, + "step": 744 + }, + { + "epoch": 0.7332677165354331, + "grad_norm": 12.07954216003418, + "learning_rate": 7.355864811133202e-07, + "loss": 1.1549, + "step": 745 + }, + { + "epoch": 0.734251968503937, + "grad_norm": 13.788519859313965, + "learning_rate": 7.365805168986085e-07, + "loss": 1.194, + "step": 746 + }, + { + "epoch": 0.735236220472441, + "grad_norm": 12.38054084777832, + "learning_rate": 7.375745526838967e-07, + "loss": 0.9507, + "step": 747 + }, + { + "epoch": 0.7362204724409449, + "grad_norm": 11.45421028137207, + "learning_rate": 7.385685884691851e-07, + "loss": 1.1306, + "step": 748 + }, + { + "epoch": 0.7372047244094488, + "grad_norm": 11.022470474243164, + "learning_rate": 7.395626242544733e-07, + "loss": 0.7992, + "step": 749 + }, + { + "epoch": 0.7381889763779528, + "grad_norm": 29.555896759033203, + "learning_rate": 7.405566600397615e-07, + "loss": 2.5297, + "step": 750 + }, + { + "epoch": 0.7391732283464567, + "grad_norm": 14.44088363647461, + "learning_rate": 7.415506958250499e-07, + "loss": 1.5363, + "step": 751 + }, + { + "epoch": 0.7401574803149606, + "grad_norm": 12.722126007080078, + "learning_rate": 7.425447316103381e-07, + "loss": 1.191, + "step": 752 + }, + { + "epoch": 0.7411417322834646, + "grad_norm": 12.448638916015625, + "learning_rate": 7.435387673956263e-07, + "loss": 0.9455, + "step": 753 + }, + { + "epoch": 0.7421259842519685, + "grad_norm": 11.549408912658691, + "learning_rate": 7.445328031809146e-07, + "loss": 0.8884, + "step": 754 + }, + { + "epoch": 0.7431102362204725, + "grad_norm": 14.364675521850586, + "learning_rate": 7.455268389662029e-07, + "loss": 1.1491, + "step": 755 + }, + { + "epoch": 0.7440944881889764, + "grad_norm": 12.005996704101562, + "learning_rate": 7.465208747514912e-07, + "loss": 0.7112, + "step": 756 + }, + { + "epoch": 0.7450787401574803, + "grad_norm": 33.205078125, + "learning_rate": 7.475149105367794e-07, + "loss": 3.0752, + "step": 757 + }, + { + "epoch": 0.7460629921259843, + "grad_norm": 13.928772926330566, + "learning_rate": 7.485089463220677e-07, + "loss": 1.1433, + "step": 758 + }, + { + "epoch": 0.7470472440944882, + "grad_norm": 26.089548110961914, + "learning_rate": 7.49502982107356e-07, + "loss": 2.142, + "step": 759 + }, + { + "epoch": 0.7480314960629921, + "grad_norm": 9.849699974060059, + "learning_rate": 7.504970178926442e-07, + "loss": 0.7859, + "step": 760 + }, + { + "epoch": 0.7490157480314961, + "grad_norm": 23.575334548950195, + "learning_rate": 7.514910536779325e-07, + "loss": 2.1668, + "step": 761 + }, + { + "epoch": 0.75, + "grad_norm": 13.810334205627441, + "learning_rate": 7.524850894632208e-07, + "loss": 1.2018, + "step": 762 + }, + { + "epoch": 0.7509842519685039, + "grad_norm": 36.599456787109375, + "learning_rate": 7.53479125248509e-07, + "loss": 2.9182, + "step": 763 + }, + { + "epoch": 0.7519685039370079, + "grad_norm": 13.759730339050293, + "learning_rate": 7.544731610337973e-07, + "loss": 1.1678, + "step": 764 + }, + { + "epoch": 0.7529527559055118, + "grad_norm": 28.865312576293945, + "learning_rate": 7.554671968190856e-07, + "loss": 2.5948, + "step": 765 + }, + { + "epoch": 0.7529527559055118, + "eval_Qnli-dev_cosine_accuracy": 0.654296875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8598544597625732, + "eval_Qnli-dev_cosine_ap": 0.6608111867356681, + "eval_Qnli-dev_cosine_f1": 0.6552901023890786, + "eval_Qnli-dev_cosine_f1_threshold": 0.8130647540092468, + "eval_Qnli-dev_cosine_precision": 0.5485714285714286, + "eval_Qnli-dev_cosine_recall": 0.8135593220338984, + "eval_Qnli-dev_dot_accuracy": 0.654296875, + "eval_Qnli-dev_dot_accuracy_threshold": 655.1024169921875, + "eval_Qnli-dev_dot_ap": 0.6591387960817412, + "eval_Qnli-dev_dot_f1": 0.6551724137931034, + "eval_Qnli-dev_dot_f1_threshold": 623.02685546875, + "eval_Qnli-dev_dot_precision": 0.5523255813953488, + "eval_Qnli-dev_dot_recall": 0.8050847457627118, + "eval_Qnli-dev_euclidean_accuracy": 0.654296875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.652008056640625, + "eval_Qnli-dev_euclidean_ap": 0.6608853977454573, + "eval_Qnli-dev_euclidean_f1": 0.6575342465753424, + "eval_Qnli-dev_euclidean_f1_threshold": 16.89258575439453, + "eval_Qnli-dev_euclidean_precision": 0.5517241379310345, + "eval_Qnli-dev_euclidean_recall": 0.8135593220338984, + "eval_Qnli-dev_manhattan_accuracy": 0.66015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 302.859375, + "eval_Qnli-dev_manhattan_ap": 0.6757856010816513, + "eval_Qnli-dev_manhattan_f1": 0.6619718309859155, + "eval_Qnli-dev_manhattan_f1_threshold": 341.82745361328125, + "eval_Qnli-dev_manhattan_precision": 0.5662650602409639, + "eval_Qnli-dev_manhattan_recall": 0.7966101694915254, + "eval_Qnli-dev_max_accuracy": 0.66015625, + "eval_Qnli-dev_max_accuracy_threshold": 655.1024169921875, + "eval_Qnli-dev_max_ap": 0.6757856010816513, + "eval_Qnli-dev_max_f1": 0.6619718309859155, + "eval_Qnli-dev_max_f1_threshold": 623.02685546875, + "eval_Qnli-dev_max_precision": 0.5662650602409639, + "eval_Qnli-dev_max_recall": 0.8135593220338984, + "eval_allNLI-dev_cosine_accuracy": 0.6875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9571083784103394, + "eval_allNLI-dev_cosine_ap": 0.46976477997672783, + "eval_allNLI-dev_cosine_f1": 0.5184, + "eval_allNLI-dev_cosine_f1_threshold": 0.8223001956939697, + "eval_allNLI-dev_cosine_precision": 0.3584070796460177, + "eval_allNLI-dev_cosine_recall": 0.9364161849710982, + "eval_allNLI-dev_dot_accuracy": 0.6875, + "eval_allNLI-dev_dot_accuracy_threshold": 733.13232421875, + "eval_allNLI-dev_dot_ap": 0.47038364604933863, + "eval_allNLI-dev_dot_f1": 0.5184, + "eval_allNLI-dev_dot_f1_threshold": 628.8274536132812, + "eval_allNLI-dev_dot_precision": 0.3584070796460177, + "eval_allNLI-dev_dot_recall": 0.9364161849710982, + "eval_allNLI-dev_euclidean_accuracy": 0.6875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 8.105023384094238, + "eval_allNLI-dev_euclidean_ap": 0.4697079763541016, + "eval_allNLI-dev_euclidean_f1": 0.5192307692307693, + "eval_allNLI-dev_euclidean_f1_threshold": 16.44765853881836, + "eval_allNLI-dev_euclidean_precision": 0.35920177383592017, + "eval_allNLI-dev_euclidean_recall": 0.9364161849710982, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 171.57086181640625, + "eval_allNLI-dev_manhattan_ap": 0.47562690305652516, + "eval_allNLI-dev_manhattan_f1": 0.5260960334029228, + "eval_allNLI-dev_manhattan_f1_threshold": 268.2262268066406, + "eval_allNLI-dev_manhattan_precision": 0.4117647058823529, + "eval_allNLI-dev_manhattan_recall": 0.7283236994219653, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 733.13232421875, + "eval_allNLI-dev_max_ap": 0.47562690305652516, + "eval_allNLI-dev_max_f1": 0.5260960334029228, + "eval_allNLI-dev_max_f1_threshold": 628.8274536132812, + "eval_allNLI-dev_max_precision": 0.4117647058823529, + "eval_allNLI-dev_max_recall": 0.9364161849710982, + "eval_loss": 1.1276968717575073, + "eval_runtime": 51.3915, + "eval_samples_per_second": 32.379, + "eval_sequential_score": 0.6757856010816513, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.3821953814433542, + "eval_sts-test_pearson_dot": 0.37751369661462264, + "eval_sts-test_pearson_euclidean": 0.4098652374465639, + "eval_sts-test_pearson_manhattan": 0.4210804679693259, + "eval_sts-test_pearson_max": 0.4210804679693259, + "eval_sts-test_spearman_cosine": 0.43466276880920734, + "eval_sts-test_spearman_dot": 0.42964168994121504, + "eval_sts-test_spearman_euclidean": 0.43483082245585475, + "eval_sts-test_spearman_manhattan": 0.4486004275758973, + "eval_sts-test_spearman_max": 0.4486004275758973, + "step": 765 + }, + { + "epoch": 0.7539370078740157, + "grad_norm": 9.897500038146973, + "learning_rate": 7.56461232604374e-07, + "loss": 1.0679, + "step": 766 + }, + { + "epoch": 0.7549212598425197, + "grad_norm": 9.520186424255371, + "learning_rate": 7.574552683896621e-07, + "loss": 0.7494, + "step": 767 + }, + { + "epoch": 0.7559055118110236, + "grad_norm": 10.916582107543945, + "learning_rate": 7.584493041749503e-07, + "loss": 0.8815, + "step": 768 + }, + { + "epoch": 0.7568897637795275, + "grad_norm": 10.711734771728516, + "learning_rate": 7.594433399602388e-07, + "loss": 0.8266, + "step": 769 + }, + { + "epoch": 0.7578740157480315, + "grad_norm": 17.2027530670166, + "learning_rate": 7.60437375745527e-07, + "loss": 1.5885, + "step": 770 + }, + { + "epoch": 0.7588582677165354, + "grad_norm": 10.653676986694336, + "learning_rate": 7.614314115308152e-07, + "loss": 1.0503, + "step": 771 + }, + { + "epoch": 0.7598425196850394, + "grad_norm": 15.49246883392334, + "learning_rate": 7.624254473161036e-07, + "loss": 1.404, + "step": 772 + }, + { + "epoch": 0.7608267716535433, + "grad_norm": 11.080362319946289, + "learning_rate": 7.634194831013918e-07, + "loss": 0.821, + "step": 773 + }, + { + "epoch": 0.7618110236220472, + "grad_norm": 22.17679214477539, + "learning_rate": 7.644135188866801e-07, + "loss": 1.9858, + "step": 774 + }, + { + "epoch": 0.7627952755905512, + "grad_norm": 10.551581382751465, + "learning_rate": 7.654075546719683e-07, + "loss": 0.9155, + "step": 775 + }, + { + "epoch": 0.7637795275590551, + "grad_norm": 14.220746994018555, + "learning_rate": 7.664015904572566e-07, + "loss": 1.6313, + "step": 776 + }, + { + "epoch": 0.764763779527559, + "grad_norm": 13.507702827453613, + "learning_rate": 7.673956262425449e-07, + "loss": 1.161, + "step": 777 + }, + { + "epoch": 0.765748031496063, + "grad_norm": 11.868473052978516, + "learning_rate": 7.683896620278331e-07, + "loss": 1.0967, + "step": 778 + }, + { + "epoch": 0.7667322834645669, + "grad_norm": 24.887378692626953, + "learning_rate": 7.693836978131215e-07, + "loss": 1.9419, + "step": 779 + }, + { + "epoch": 0.7677165354330708, + "grad_norm": 11.0693998336792, + "learning_rate": 7.703777335984097e-07, + "loss": 1.0386, + "step": 780 + }, + { + "epoch": 0.7687007874015748, + "grad_norm": 14.315882682800293, + "learning_rate": 7.713717693836979e-07, + "loss": 1.1734, + "step": 781 + }, + { + "epoch": 0.7696850393700787, + "grad_norm": 13.440196990966797, + "learning_rate": 7.723658051689862e-07, + "loss": 1.1683, + "step": 782 + }, + { + "epoch": 0.7706692913385826, + "grad_norm": 13.43931770324707, + "learning_rate": 7.733598409542745e-07, + "loss": 1.3971, + "step": 783 + }, + { + "epoch": 0.7716535433070866, + "grad_norm": 23.48846435546875, + "learning_rate": 7.743538767395627e-07, + "loss": 1.843, + "step": 784 + }, + { + "epoch": 0.7726377952755905, + "grad_norm": 12.930007934570312, + "learning_rate": 7.75347912524851e-07, + "loss": 1.3123, + "step": 785 + }, + { + "epoch": 0.7736220472440944, + "grad_norm": 14.505959510803223, + "learning_rate": 7.763419483101393e-07, + "loss": 1.3975, + "step": 786 + }, + { + "epoch": 0.7746062992125984, + "grad_norm": 12.660870552062988, + "learning_rate": 7.773359840954276e-07, + "loss": 0.9761, + "step": 787 + }, + { + "epoch": 0.7755905511811023, + "grad_norm": 13.688265800476074, + "learning_rate": 7.783300198807158e-07, + "loss": 1.2307, + "step": 788 + }, + { + "epoch": 0.7765748031496063, + "grad_norm": 12.699599266052246, + "learning_rate": 7.79324055666004e-07, + "loss": 1.1414, + "step": 789 + }, + { + "epoch": 0.7775590551181102, + "grad_norm": 12.13899040222168, + "learning_rate": 7.803180914512924e-07, + "loss": 1.0158, + "step": 790 + }, + { + "epoch": 0.7785433070866141, + "grad_norm": 13.28480052947998, + "learning_rate": 7.813121272365806e-07, + "loss": 1.2875, + "step": 791 + }, + { + "epoch": 0.7795275590551181, + "grad_norm": 10.770246505737305, + "learning_rate": 7.823061630218689e-07, + "loss": 0.7074, + "step": 792 + }, + { + "epoch": 0.780511811023622, + "grad_norm": 12.663084983825684, + "learning_rate": 7.833001988071572e-07, + "loss": 1.0812, + "step": 793 + }, + { + "epoch": 0.781496062992126, + "grad_norm": 11.055609703063965, + "learning_rate": 7.842942345924454e-07, + "loss": 0.8385, + "step": 794 + }, + { + "epoch": 0.78248031496063, + "grad_norm": 14.255220413208008, + "learning_rate": 7.852882703777337e-07, + "loss": 1.1102, + "step": 795 + }, + { + "epoch": 0.7834645669291339, + "grad_norm": 12.136215209960938, + "learning_rate": 7.862823061630219e-07, + "loss": 1.0973, + "step": 796 + }, + { + "epoch": 0.7844488188976378, + "grad_norm": 12.627120018005371, + "learning_rate": 7.872763419483102e-07, + "loss": 0.927, + "step": 797 + }, + { + "epoch": 0.7854330708661418, + "grad_norm": 10.780425071716309, + "learning_rate": 7.882703777335986e-07, + "loss": 0.7737, + "step": 798 + }, + { + "epoch": 0.7864173228346457, + "grad_norm": 7.835007667541504, + "learning_rate": 7.892644135188868e-07, + "loss": 0.4748, + "step": 799 + }, + { + "epoch": 0.7874015748031497, + "grad_norm": 13.980217933654785, + "learning_rate": 7.902584493041752e-07, + "loss": 1.4069, + "step": 800 + }, + { + "epoch": 0.7883858267716536, + "grad_norm": 10.69910717010498, + "learning_rate": 7.912524850894634e-07, + "loss": 0.9795, + "step": 801 + }, + { + "epoch": 0.7893700787401575, + "grad_norm": 12.988688468933105, + "learning_rate": 7.922465208747516e-07, + "loss": 0.9736, + "step": 802 + }, + { + "epoch": 0.7903543307086615, + "grad_norm": 12.868518829345703, + "learning_rate": 7.932405566600399e-07, + "loss": 0.9576, + "step": 803 + }, + { + "epoch": 0.7913385826771654, + "grad_norm": 8.793651580810547, + "learning_rate": 7.942345924453282e-07, + "loss": 0.4677, + "step": 804 + }, + { + "epoch": 0.7923228346456693, + "grad_norm": 9.964987754821777, + "learning_rate": 7.952286282306165e-07, + "loss": 0.637, + "step": 805 + }, + { + "epoch": 0.7933070866141733, + "grad_norm": 10.472431182861328, + "learning_rate": 7.962226640159047e-07, + "loss": 0.6411, + "step": 806 + }, + { + "epoch": 0.7942913385826772, + "grad_norm": 14.797800064086914, + "learning_rate": 7.97216699801193e-07, + "loss": 0.9679, + "step": 807 + }, + { + "epoch": 0.7952755905511811, + "grad_norm": 10.579925537109375, + "learning_rate": 7.982107355864813e-07, + "loss": 0.8034, + "step": 808 + }, + { + "epoch": 0.7962598425196851, + "grad_norm": 34.83205795288086, + "learning_rate": 7.992047713717695e-07, + "loss": 3.0881, + "step": 809 + }, + { + "epoch": 0.797244094488189, + "grad_norm": 9.906510353088379, + "learning_rate": 8.001988071570577e-07, + "loss": 0.5545, + "step": 810 + }, + { + "epoch": 0.7982283464566929, + "grad_norm": 27.75262451171875, + "learning_rate": 8.011928429423461e-07, + "loss": 1.837, + "step": 811 + }, + { + "epoch": 0.7992125984251969, + "grad_norm": 25.863445281982422, + "learning_rate": 8.021868787276343e-07, + "loss": 2.3376, + "step": 812 + }, + { + "epoch": 0.8001968503937008, + "grad_norm": 15.276238441467285, + "learning_rate": 8.031809145129226e-07, + "loss": 1.0845, + "step": 813 + }, + { + "epoch": 0.8011811023622047, + "grad_norm": 14.695995330810547, + "learning_rate": 8.041749502982109e-07, + "loss": 1.1629, + "step": 814 + }, + { + "epoch": 0.8021653543307087, + "grad_norm": 14.996781349182129, + "learning_rate": 8.051689860834991e-07, + "loss": 1.2243, + "step": 815 + }, + { + "epoch": 0.8031496062992126, + "grad_norm": 12.822697639465332, + "learning_rate": 8.061630218687874e-07, + "loss": 0.8564, + "step": 816 + }, + { + "epoch": 0.8041338582677166, + "grad_norm": 10.565481185913086, + "learning_rate": 8.071570576540756e-07, + "loss": 0.8796, + "step": 817 + }, + { + "epoch": 0.8051181102362205, + "grad_norm": 25.22612190246582, + "learning_rate": 8.08151093439364e-07, + "loss": 1.5817, + "step": 818 + }, + { + "epoch": 0.8061023622047244, + "grad_norm": 11.941892623901367, + "learning_rate": 8.091451292246522e-07, + "loss": 1.1185, + "step": 819 + }, + { + "epoch": 0.8070866141732284, + "grad_norm": 7.7961578369140625, + "learning_rate": 8.101391650099404e-07, + "loss": 0.394, + "step": 820 + }, + { + "epoch": 0.8080708661417323, + "grad_norm": 12.509605407714844, + "learning_rate": 8.111332007952288e-07, + "loss": 0.9968, + "step": 821 + }, + { + "epoch": 0.8090551181102362, + "grad_norm": 11.517319679260254, + "learning_rate": 8.12127236580517e-07, + "loss": 0.973, + "step": 822 + }, + { + "epoch": 0.8100393700787402, + "grad_norm": 8.818398475646973, + "learning_rate": 8.131212723658052e-07, + "loss": 0.6203, + "step": 823 + }, + { + "epoch": 0.8110236220472441, + "grad_norm": 22.124818801879883, + "learning_rate": 8.141153081510935e-07, + "loss": 1.6007, + "step": 824 + }, + { + "epoch": 0.812007874015748, + "grad_norm": 20.68657112121582, + "learning_rate": 8.151093439363818e-07, + "loss": 1.7027, + "step": 825 + }, + { + "epoch": 0.812992125984252, + "grad_norm": 21.102813720703125, + "learning_rate": 8.161033797216701e-07, + "loss": 1.8865, + "step": 826 + }, + { + "epoch": 0.8139763779527559, + "grad_norm": 12.814473152160645, + "learning_rate": 8.170974155069583e-07, + "loss": 1.2149, + "step": 827 + }, + { + "epoch": 0.8149606299212598, + "grad_norm": 23.65349769592285, + "learning_rate": 8.180914512922467e-07, + "loss": 2.0459, + "step": 828 + }, + { + "epoch": 0.8159448818897638, + "grad_norm": 23.712879180908203, + "learning_rate": 8.19085487077535e-07, + "loss": 2.2676, + "step": 829 + }, + { + "epoch": 0.8169291338582677, + "grad_norm": 11.891949653625488, + "learning_rate": 8.200795228628232e-07, + "loss": 1.3287, + "step": 830 + }, + { + "epoch": 0.8179133858267716, + "grad_norm": 11.260537147521973, + "learning_rate": 8.210735586481115e-07, + "loss": 0.9346, + "step": 831 + }, + { + "epoch": 0.8188976377952756, + "grad_norm": 10.765104293823242, + "learning_rate": 8.220675944333998e-07, + "loss": 0.835, + "step": 832 + }, + { + "epoch": 0.8198818897637795, + "grad_norm": 8.057222366333008, + "learning_rate": 8.23061630218688e-07, + "loss": 0.5872, + "step": 833 + }, + { + "epoch": 0.8208661417322834, + "grad_norm": 10.354174613952637, + "learning_rate": 8.240556660039763e-07, + "loss": 0.7393, + "step": 834 + }, + { + "epoch": 0.8218503937007874, + "grad_norm": 12.181092262268066, + "learning_rate": 8.250497017892646e-07, + "loss": 0.971, + "step": 835 + }, + { + "epoch": 0.8228346456692913, + "grad_norm": 11.355849266052246, + "learning_rate": 8.260437375745529e-07, + "loss": 1.1222, + "step": 836 + }, + { + "epoch": 0.8238188976377953, + "grad_norm": 11.0961275100708, + "learning_rate": 8.270377733598411e-07, + "loss": 1.1121, + "step": 837 + }, + { + "epoch": 0.8248031496062992, + "grad_norm": 9.565896987915039, + "learning_rate": 8.280318091451293e-07, + "loss": 0.7018, + "step": 838 + }, + { + "epoch": 0.8257874015748031, + "grad_norm": 11.490205764770508, + "learning_rate": 8.290258449304177e-07, + "loss": 1.0924, + "step": 839 + }, + { + "epoch": 0.8267716535433071, + "grad_norm": 17.25556755065918, + "learning_rate": 8.300198807157059e-07, + "loss": 1.1113, + "step": 840 + }, + { + "epoch": 0.827755905511811, + "grad_norm": 12.140558242797852, + "learning_rate": 8.310139165009941e-07, + "loss": 1.198, + "step": 841 + }, + { + "epoch": 0.8287401574803149, + "grad_norm": 9.053621292114258, + "learning_rate": 8.320079522862825e-07, + "loss": 0.7756, + "step": 842 + }, + { + "epoch": 0.8297244094488189, + "grad_norm": 19.02862548828125, + "learning_rate": 8.330019880715707e-07, + "loss": 1.5646, + "step": 843 + }, + { + "epoch": 0.8307086614173228, + "grad_norm": 11.701822280883789, + "learning_rate": 8.33996023856859e-07, + "loss": 0.7395, + "step": 844 + }, + { + "epoch": 0.8316929133858267, + "grad_norm": 11.258445739746094, + "learning_rate": 8.349900596421472e-07, + "loss": 0.8208, + "step": 845 + }, + { + "epoch": 0.8326771653543307, + "grad_norm": 19.854507446289062, + "learning_rate": 8.359840954274355e-07, + "loss": 1.9033, + "step": 846 + }, + { + "epoch": 0.8336614173228346, + "grad_norm": 15.992486953735352, + "learning_rate": 8.369781312127238e-07, + "loss": 1.3321, + "step": 847 + }, + { + "epoch": 0.8346456692913385, + "grad_norm": 12.462971687316895, + "learning_rate": 8.37972166998012e-07, + "loss": 0.7459, + "step": 848 + }, + { + "epoch": 0.8356299212598425, + "grad_norm": 10.48875617980957, + "learning_rate": 8.389662027833004e-07, + "loss": 0.6561, + "step": 849 + }, + { + "epoch": 0.8366141732283464, + "grad_norm": 9.815774917602539, + "learning_rate": 8.399602385685886e-07, + "loss": 0.6797, + "step": 850 + }, + { + "epoch": 0.8375984251968503, + "grad_norm": 30.559804916381836, + "learning_rate": 8.409542743538768e-07, + "loss": 2.2847, + "step": 851 + }, + { + "epoch": 0.8385826771653543, + "grad_norm": 13.01028823852539, + "learning_rate": 8.419483101391651e-07, + "loss": 1.0886, + "step": 852 + }, + { + "epoch": 0.8395669291338582, + "grad_norm": 10.173580169677734, + "learning_rate": 8.429423459244534e-07, + "loss": 0.598, + "step": 853 + }, + { + "epoch": 0.8405511811023622, + "grad_norm": 21.44631576538086, + "learning_rate": 8.439363817097416e-07, + "loss": 1.3609, + "step": 854 + }, + { + "epoch": 0.8415354330708661, + "grad_norm": 12.637033462524414, + "learning_rate": 8.449304174950299e-07, + "loss": 1.055, + "step": 855 + }, + { + "epoch": 0.84251968503937, + "grad_norm": 29.952192306518555, + "learning_rate": 8.459244532803182e-07, + "loss": 2.6634, + "step": 856 + }, + { + "epoch": 0.843503937007874, + "grad_norm": 19.379634857177734, + "learning_rate": 8.469184890656066e-07, + "loss": 1.7001, + "step": 857 + }, + { + "epoch": 0.844488188976378, + "grad_norm": 15.857636451721191, + "learning_rate": 8.479125248508948e-07, + "loss": 1.2224, + "step": 858 + }, + { + "epoch": 0.8454724409448819, + "grad_norm": 14.187600135803223, + "learning_rate": 8.48906560636183e-07, + "loss": 1.2708, + "step": 859 + }, + { + "epoch": 0.8464566929133859, + "grad_norm": 11.313172340393066, + "learning_rate": 8.499005964214714e-07, + "loss": 0.9535, + "step": 860 + }, + { + "epoch": 0.8474409448818898, + "grad_norm": 15.50436782836914, + "learning_rate": 8.508946322067596e-07, + "loss": 1.1954, + "step": 861 + }, + { + "epoch": 0.8484251968503937, + "grad_norm": 12.802027702331543, + "learning_rate": 8.518886679920479e-07, + "loss": 1.1258, + "step": 862 + }, + { + "epoch": 0.8494094488188977, + "grad_norm": 20.335342407226562, + "learning_rate": 8.528827037773362e-07, + "loss": 1.503, + "step": 863 + }, + { + "epoch": 0.8503937007874016, + "grad_norm": 12.412203788757324, + "learning_rate": 8.538767395626244e-07, + "loss": 1.2438, + "step": 864 + }, + { + "epoch": 0.8513779527559056, + "grad_norm": 16.102977752685547, + "learning_rate": 8.548707753479127e-07, + "loss": 1.9224, + "step": 865 + }, + { + "epoch": 0.8523622047244095, + "grad_norm": 11.874804496765137, + "learning_rate": 8.558648111332009e-07, + "loss": 0.9334, + "step": 866 + }, + { + "epoch": 0.8533464566929134, + "grad_norm": 19.06435775756836, + "learning_rate": 8.568588469184892e-07, + "loss": 1.5525, + "step": 867 + }, + { + "epoch": 0.8543307086614174, + "grad_norm": 9.16429615020752, + "learning_rate": 8.578528827037775e-07, + "loss": 0.502, + "step": 868 + }, + { + "epoch": 0.8553149606299213, + "grad_norm": 11.745601654052734, + "learning_rate": 8.588469184890657e-07, + "loss": 1.4609, + "step": 869 + }, + { + "epoch": 0.8562992125984252, + "grad_norm": 21.498977661132812, + "learning_rate": 8.598409542743541e-07, + "loss": 1.8758, + "step": 870 + }, + { + "epoch": 0.8572834645669292, + "grad_norm": 11.363224029541016, + "learning_rate": 8.608349900596423e-07, + "loss": 0.9021, + "step": 871 + }, + { + "epoch": 0.8582677165354331, + "grad_norm": 12.484760284423828, + "learning_rate": 8.618290258449305e-07, + "loss": 0.735, + "step": 872 + }, + { + "epoch": 0.859251968503937, + "grad_norm": 12.676814079284668, + "learning_rate": 8.628230616302189e-07, + "loss": 1.2601, + "step": 873 + }, + { + "epoch": 0.860236220472441, + "grad_norm": 13.334346771240234, + "learning_rate": 8.638170974155071e-07, + "loss": 1.4235, + "step": 874 + }, + { + "epoch": 0.8612204724409449, + "grad_norm": 11.478998184204102, + "learning_rate": 8.648111332007954e-07, + "loss": 1.0701, + "step": 875 + }, + { + "epoch": 0.8622047244094488, + "grad_norm": 10.054930686950684, + "learning_rate": 8.658051689860836e-07, + "loss": 0.7327, + "step": 876 + }, + { + "epoch": 0.8631889763779528, + "grad_norm": 10.54784870147705, + "learning_rate": 8.667992047713719e-07, + "loss": 1.0354, + "step": 877 + }, + { + "epoch": 0.8641732283464567, + "grad_norm": 9.850967407226562, + "learning_rate": 8.677932405566602e-07, + "loss": 0.6344, + "step": 878 + }, + { + "epoch": 0.8651574803149606, + "grad_norm": 12.492563247680664, + "learning_rate": 8.687872763419484e-07, + "loss": 0.8981, + "step": 879 + }, + { + "epoch": 0.8661417322834646, + "grad_norm": 16.78278923034668, + "learning_rate": 8.697813121272366e-07, + "loss": 1.0991, + "step": 880 + }, + { + "epoch": 0.8671259842519685, + "grad_norm": 11.904287338256836, + "learning_rate": 8.70775347912525e-07, + "loss": 0.7773, + "step": 881 + }, + { + "epoch": 0.8681102362204725, + "grad_norm": 16.986021041870117, + "learning_rate": 8.717693836978132e-07, + "loss": 1.5066, + "step": 882 + }, + { + "epoch": 0.8690944881889764, + "grad_norm": 12.554054260253906, + "learning_rate": 8.727634194831015e-07, + "loss": 0.7757, + "step": 883 + }, + { + "epoch": 0.8700787401574803, + "grad_norm": 16.78717041015625, + "learning_rate": 8.737574552683898e-07, + "loss": 0.9993, + "step": 884 + }, + { + "epoch": 0.8710629921259843, + "grad_norm": 14.430180549621582, + "learning_rate": 8.74751491053678e-07, + "loss": 1.1451, + "step": 885 + }, + { + "epoch": 0.8720472440944882, + "grad_norm": 12.971931457519531, + "learning_rate": 8.757455268389663e-07, + "loss": 0.9861, + "step": 886 + }, + { + "epoch": 0.8730314960629921, + "grad_norm": 12.929235458374023, + "learning_rate": 8.767395626242545e-07, + "loss": 0.9548, + "step": 887 + }, + { + "epoch": 0.8740157480314961, + "grad_norm": 9.33077335357666, + "learning_rate": 8.77733598409543e-07, + "loss": 0.4494, + "step": 888 + }, + { + "epoch": 0.875, + "grad_norm": 12.301645278930664, + "learning_rate": 8.787276341948312e-07, + "loss": 0.8269, + "step": 889 + }, + { + "epoch": 0.8759842519685039, + "grad_norm": 25.47489356994629, + "learning_rate": 8.797216699801194e-07, + "loss": 1.8873, + "step": 890 + }, + { + "epoch": 0.8769685039370079, + "grad_norm": 11.886186599731445, + "learning_rate": 8.807157057654078e-07, + "loss": 0.8112, + "step": 891 + }, + { + "epoch": 0.8779527559055118, + "grad_norm": 13.846529960632324, + "learning_rate": 8.81709741550696e-07, + "loss": 1.152, + "step": 892 + }, + { + "epoch": 0.8789370078740157, + "grad_norm": 16.300519943237305, + "learning_rate": 8.827037773359842e-07, + "loss": 1.4563, + "step": 893 + }, + { + "epoch": 0.8799212598425197, + "grad_norm": 24.21488380432129, + "learning_rate": 8.836978131212726e-07, + "loss": 1.7084, + "step": 894 + }, + { + "epoch": 0.8809055118110236, + "grad_norm": 12.841357231140137, + "learning_rate": 8.846918489065608e-07, + "loss": 1.0018, + "step": 895 + }, + { + "epoch": 0.8818897637795275, + "grad_norm": 8.953272819519043, + "learning_rate": 8.856858846918491e-07, + "loss": 0.6114, + "step": 896 + }, + { + "epoch": 0.8828740157480315, + "grad_norm": 11.888354301452637, + "learning_rate": 8.866799204771373e-07, + "loss": 0.8984, + "step": 897 + }, + { + "epoch": 0.8838582677165354, + "grad_norm": 16.15188217163086, + "learning_rate": 8.876739562624256e-07, + "loss": 1.4973, + "step": 898 + }, + { + "epoch": 0.8848425196850394, + "grad_norm": 11.510607719421387, + "learning_rate": 8.886679920477139e-07, + "loss": 0.8233, + "step": 899 + }, + { + "epoch": 0.8858267716535433, + "grad_norm": 12.673114776611328, + "learning_rate": 8.896620278330021e-07, + "loss": 1.0287, + "step": 900 + }, + { + "epoch": 0.8868110236220472, + "grad_norm": 10.993170738220215, + "learning_rate": 8.906560636182905e-07, + "loss": 0.9614, + "step": 901 + }, + { + "epoch": 0.8877952755905512, + "grad_norm": 15.443177223205566, + "learning_rate": 8.916500994035787e-07, + "loss": 1.0764, + "step": 902 + }, + { + "epoch": 0.8887795275590551, + "grad_norm": 14.781033515930176, + "learning_rate": 8.926441351888669e-07, + "loss": 0.9253, + "step": 903 + }, + { + "epoch": 0.889763779527559, + "grad_norm": 14.57919979095459, + "learning_rate": 8.936381709741552e-07, + "loss": 1.2771, + "step": 904 + }, + { + "epoch": 0.890748031496063, + "grad_norm": 11.156841278076172, + "learning_rate": 8.946322067594435e-07, + "loss": 0.8748, + "step": 905 + }, + { + "epoch": 0.8917322834645669, + "grad_norm": 11.762528419494629, + "learning_rate": 8.956262425447318e-07, + "loss": 0.7201, + "step": 906 + }, + { + "epoch": 0.8927165354330708, + "grad_norm": 19.710874557495117, + "learning_rate": 8.9662027833002e-07, + "loss": 1.4378, + "step": 907 + }, + { + "epoch": 0.8937007874015748, + "grad_norm": 13.278345108032227, + "learning_rate": 8.976143141153082e-07, + "loss": 1.222, + "step": 908 + }, + { + "epoch": 0.8946850393700787, + "grad_norm": 9.57332706451416, + "learning_rate": 8.986083499005966e-07, + "loss": 0.5576, + "step": 909 + }, + { + "epoch": 0.8956692913385826, + "grad_norm": 13.30394172668457, + "learning_rate": 8.996023856858848e-07, + "loss": 0.9889, + "step": 910 + }, + { + "epoch": 0.8966535433070866, + "grad_norm": 10.710693359375, + "learning_rate": 9.00596421471173e-07, + "loss": 0.752, + "step": 911 + }, + { + "epoch": 0.8976377952755905, + "grad_norm": 12.118535041809082, + "learning_rate": 9.015904572564614e-07, + "loss": 0.9586, + "step": 912 + }, + { + "epoch": 0.8986220472440944, + "grad_norm": 11.010882377624512, + "learning_rate": 9.025844930417496e-07, + "loss": 0.7206, + "step": 913 + }, + { + "epoch": 0.8996062992125984, + "grad_norm": 9.926494598388672, + "learning_rate": 9.035785288270379e-07, + "loss": 0.5809, + "step": 914 + }, + { + "epoch": 0.9005905511811023, + "grad_norm": 12.404060363769531, + "learning_rate": 9.045725646123262e-07, + "loss": 0.6836, + "step": 915 + }, + { + "epoch": 0.9015748031496063, + "grad_norm": 11.887187004089355, + "learning_rate": 9.055666003976144e-07, + "loss": 0.6633, + "step": 916 + }, + { + "epoch": 0.9025590551181102, + "grad_norm": 9.540690422058105, + "learning_rate": 9.065606361829028e-07, + "loss": 0.625, + "step": 917 + }, + { + "epoch": 0.9035433070866141, + "grad_norm": 9.185359001159668, + "learning_rate": 9.07554671968191e-07, + "loss": 0.6292, + "step": 918 + }, + { + "epoch": 0.9035433070866141, + "eval_Qnli-dev_cosine_accuracy": 0.67578125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8416056632995605, + "eval_Qnli-dev_cosine_ap": 0.6868588056341085, + "eval_Qnli-dev_cosine_f1": 0.6655290102389079, + "eval_Qnli-dev_cosine_f1_threshold": 0.7768508791923523, + "eval_Qnli-dev_cosine_precision": 0.5571428571428572, + "eval_Qnli-dev_cosine_recall": 0.826271186440678, + "eval_Qnli-dev_dot_accuracy": 0.673828125, + "eval_Qnli-dev_dot_accuracy_threshold": 643.5964965820312, + "eval_Qnli-dev_dot_ap": 0.6848126073198346, + "eval_Qnli-dev_dot_f1": 0.6666666666666666, + "eval_Qnli-dev_dot_f1_threshold": 592.97900390625, + "eval_Qnli-dev_dot_precision": 0.5568181818181818, + "eval_Qnli-dev_dot_recall": 0.8305084745762712, + "eval_Qnli-dev_euclidean_accuracy": 0.67578125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.5496826171875, + "eval_Qnli-dev_euclidean_ap": 0.687901776797238, + "eval_Qnli-dev_euclidean_f1": 0.6655290102389079, + "eval_Qnli-dev_euclidean_f1_threshold": 18.46298599243164, + "eval_Qnli-dev_euclidean_precision": 0.5571428571428572, + "eval_Qnli-dev_euclidean_recall": 0.826271186440678, + "eval_Qnli-dev_manhattan_accuracy": 0.673828125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 327.61651611328125, + "eval_Qnli-dev_manhattan_ap": 0.6955127635209091, + "eval_Qnli-dev_manhattan_f1": 0.6746987951807228, + "eval_Qnli-dev_manhattan_f1_threshold": 380.16864013671875, + "eval_Qnli-dev_manhattan_precision": 0.5681159420289855, + "eval_Qnli-dev_manhattan_recall": 0.8305084745762712, + "eval_Qnli-dev_max_accuracy": 0.67578125, + "eval_Qnli-dev_max_accuracy_threshold": 643.5964965820312, + "eval_Qnli-dev_max_ap": 0.6955127635209091, + "eval_Qnli-dev_max_f1": 0.6746987951807228, + "eval_Qnli-dev_max_f1_threshold": 592.97900390625, + "eval_Qnli-dev_max_precision": 0.5681159420289855, + "eval_Qnli-dev_max_recall": 0.8305084745762712, + "eval_allNLI-dev_cosine_accuracy": 0.681640625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9431064128875732, + "eval_allNLI-dev_cosine_ap": 0.4868215343781438, + "eval_allNLI-dev_cosine_f1": 0.5358649789029536, + "eval_allNLI-dev_cosine_f1_threshold": 0.8642745614051819, + "eval_allNLI-dev_cosine_precision": 0.4219269102990033, + "eval_allNLI-dev_cosine_recall": 0.7341040462427746, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 721.832763671875, + "eval_allNLI-dev_dot_ap": 0.48691473603156943, + "eval_allNLI-dev_dot_f1": 0.5361702127659574, + "eval_allNLI-dev_dot_f1_threshold": 663.1417236328125, + "eval_allNLI-dev_dot_precision": 0.42424242424242425, + "eval_allNLI-dev_dot_recall": 0.7283236994219653, + "eval_allNLI-dev_euclidean_accuracy": 0.681640625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.334846496582031, + "eval_allNLI-dev_euclidean_ap": 0.4866152335549325, + "eval_allNLI-dev_euclidean_f1": 0.5347368421052632, + "eval_allNLI-dev_euclidean_f1_threshold": 14.412881851196289, + "eval_allNLI-dev_euclidean_precision": 0.4205298013245033, + "eval_allNLI-dev_euclidean_recall": 0.7341040462427746, + "eval_allNLI-dev_manhattan_accuracy": 0.6796875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 195.29490661621094, + "eval_allNLI-dev_manhattan_ap": 0.488089181664835, + "eval_allNLI-dev_manhattan_f1": 0.5410628019323671, + "eval_allNLI-dev_manhattan_f1_threshold": 279.084228515625, + "eval_allNLI-dev_manhattan_precision": 0.46473029045643155, + "eval_allNLI-dev_manhattan_recall": 0.6473988439306358, + "eval_allNLI-dev_max_accuracy": 0.68359375, + "eval_allNLI-dev_max_accuracy_threshold": 721.832763671875, + "eval_allNLI-dev_max_ap": 0.488089181664835, + "eval_allNLI-dev_max_f1": 0.5410628019323671, + "eval_allNLI-dev_max_f1_threshold": 663.1417236328125, + "eval_allNLI-dev_max_precision": 0.46473029045643155, + "eval_allNLI-dev_max_recall": 0.7341040462427746, + "eval_loss": 0.9408406615257263, + "eval_runtime": 51.4094, + "eval_samples_per_second": 32.368, + "eval_sequential_score": 0.6955127635209091, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.47489010897580175, + "eval_sts-test_pearson_dot": 0.47173813637961304, + "eval_sts-test_pearson_euclidean": 0.4958211269942067, + "eval_sts-test_pearson_manhattan": 0.5071164653454794, + "eval_sts-test_pearson_max": 0.5071164653454794, + "eval_sts-test_spearman_cosine": 0.5108403029140102, + "eval_sts-test_spearman_dot": 0.5080144845134502, + "eval_sts-test_spearman_euclidean": 0.5111272032507904, + "eval_sts-test_spearman_manhattan": 0.5219595593294084, + "eval_sts-test_spearman_max": 0.5219595593294084, + "step": 918 + }, + { + "epoch": 0.9045275590551181, + "grad_norm": 13.31297779083252, + "learning_rate": 9.085487077534794e-07, + "loss": 1.1144, + "step": 919 + }, + { + "epoch": 0.905511811023622, + "grad_norm": 30.346559524536133, + "learning_rate": 9.095427435387676e-07, + "loss": 2.0274, + "step": 920 + }, + { + "epoch": 0.906496062992126, + "grad_norm": 14.480632781982422, + "learning_rate": 9.105367793240558e-07, + "loss": 1.1657, + "step": 921 + }, + { + "epoch": 0.90748031496063, + "grad_norm": 11.103870391845703, + "learning_rate": 9.115308151093442e-07, + "loss": 0.7091, + "step": 922 + }, + { + "epoch": 0.9084645669291339, + "grad_norm": 14.190383911132812, + "learning_rate": 9.125248508946324e-07, + "loss": 1.303, + "step": 923 + }, + { + "epoch": 0.9094488188976378, + "grad_norm": 36.14653015136719, + "learning_rate": 9.135188866799206e-07, + "loss": 2.4449, + "step": 924 + }, + { + "epoch": 0.9104330708661418, + "grad_norm": 11.780421257019043, + "learning_rate": 9.145129224652089e-07, + "loss": 0.7181, + "step": 925 + }, + { + "epoch": 0.9114173228346457, + "grad_norm": 12.563950538635254, + "learning_rate": 9.155069582504972e-07, + "loss": 1.2308, + "step": 926 + }, + { + "epoch": 0.9124015748031497, + "grad_norm": 12.789769172668457, + "learning_rate": 9.165009940357855e-07, + "loss": 1.0513, + "step": 927 + }, + { + "epoch": 0.9133858267716536, + "grad_norm": 11.681246757507324, + "learning_rate": 9.174950298210737e-07, + "loss": 0.8714, + "step": 928 + }, + { + "epoch": 0.9143700787401575, + "grad_norm": 10.838265419006348, + "learning_rate": 9.18489065606362e-07, + "loss": 0.9573, + "step": 929 + }, + { + "epoch": 0.9153543307086615, + "grad_norm": 19.145748138427734, + "learning_rate": 9.194831013916503e-07, + "loss": 2.0303, + "step": 930 + }, + { + "epoch": 0.9163385826771654, + "grad_norm": 9.689408302307129, + "learning_rate": 9.204771371769385e-07, + "loss": 0.7414, + "step": 931 + }, + { + "epoch": 0.9173228346456693, + "grad_norm": 17.340660095214844, + "learning_rate": 9.214711729622268e-07, + "loss": 1.4278, + "step": 932 + }, + { + "epoch": 0.9183070866141733, + "grad_norm": 13.9085111618042, + "learning_rate": 9.224652087475151e-07, + "loss": 1.2565, + "step": 933 + }, + { + "epoch": 0.9192913385826772, + "grad_norm": 11.080978393554688, + "learning_rate": 9.234592445328033e-07, + "loss": 0.6508, + "step": 934 + }, + { + "epoch": 0.9202755905511811, + "grad_norm": 9.10645580291748, + "learning_rate": 9.244532803180916e-07, + "loss": 0.5146, + "step": 935 + }, + { + "epoch": 0.9212598425196851, + "grad_norm": 8.178444862365723, + "learning_rate": 9.254473161033799e-07, + "loss": 0.5738, + "step": 936 + }, + { + "epoch": 0.922244094488189, + "grad_norm": 16.414060592651367, + "learning_rate": 9.264413518886681e-07, + "loss": 1.1354, + "step": 937 + }, + { + "epoch": 0.9232283464566929, + "grad_norm": 13.718335151672363, + "learning_rate": 9.274353876739564e-07, + "loss": 0.9445, + "step": 938 + }, + { + "epoch": 0.9242125984251969, + "grad_norm": 10.481086730957031, + "learning_rate": 9.284294234592446e-07, + "loss": 0.6232, + "step": 939 + }, + { + "epoch": 0.9251968503937008, + "grad_norm": 15.332398414611816, + "learning_rate": 9.29423459244533e-07, + "loss": 1.6567, + "step": 940 + }, + { + "epoch": 0.9261811023622047, + "grad_norm": 19.93722915649414, + "learning_rate": 9.304174950298212e-07, + "loss": 1.8855, + "step": 941 + }, + { + "epoch": 0.9271653543307087, + "grad_norm": 12.797704696655273, + "learning_rate": 9.314115308151094e-07, + "loss": 0.8225, + "step": 942 + }, + { + "epoch": 0.9281496062992126, + "grad_norm": 10.449248313903809, + "learning_rate": 9.324055666003978e-07, + "loss": 0.4741, + "step": 943 + }, + { + "epoch": 0.9291338582677166, + "grad_norm": 29.1734561920166, + "learning_rate": 9.33399602385686e-07, + "loss": 2.3002, + "step": 944 + }, + { + "epoch": 0.9301181102362205, + "grad_norm": 20.42205047607422, + "learning_rate": 9.343936381709743e-07, + "loss": 1.4051, + "step": 945 + }, + { + "epoch": 0.9311023622047244, + "grad_norm": 19.387802124023438, + "learning_rate": 9.353876739562625e-07, + "loss": 1.0528, + "step": 946 + }, + { + "epoch": 0.9320866141732284, + "grad_norm": 11.130048751831055, + "learning_rate": 9.363817097415509e-07, + "loss": 0.81, + "step": 947 + }, + { + "epoch": 0.9330708661417323, + "grad_norm": 9.055191993713379, + "learning_rate": 9.373757455268392e-07, + "loss": 0.6629, + "step": 948 + }, + { + "epoch": 0.9340551181102362, + "grad_norm": 13.656773567199707, + "learning_rate": 9.383697813121274e-07, + "loss": 0.9533, + "step": 949 + }, + { + "epoch": 0.9350393700787402, + "grad_norm": 8.283589363098145, + "learning_rate": 9.393638170974157e-07, + "loss": 0.4455, + "step": 950 + }, + { + "epoch": 0.9360236220472441, + "grad_norm": 9.452292442321777, + "learning_rate": 9.40357852882704e-07, + "loss": 0.7622, + "step": 951 + }, + { + "epoch": 0.937007874015748, + "grad_norm": 11.2641019821167, + "learning_rate": 9.413518886679922e-07, + "loss": 1.0084, + "step": 952 + }, + { + "epoch": 0.937992125984252, + "grad_norm": 8.543481826782227, + "learning_rate": 9.423459244532805e-07, + "loss": 0.553, + "step": 953 + }, + { + "epoch": 0.9389763779527559, + "grad_norm": 10.038814544677734, + "learning_rate": 9.433399602385688e-07, + "loss": 0.6533, + "step": 954 + }, + { + "epoch": 0.9399606299212598, + "grad_norm": 8.911643028259277, + "learning_rate": 9.44333996023857e-07, + "loss": 0.5864, + "step": 955 + }, + { + "epoch": 0.9409448818897638, + "grad_norm": 16.401081085205078, + "learning_rate": 9.453280318091453e-07, + "loss": 1.1066, + "step": 956 + }, + { + "epoch": 0.9419291338582677, + "grad_norm": 8.937928199768066, + "learning_rate": 9.463220675944336e-07, + "loss": 0.5942, + "step": 957 + }, + { + "epoch": 0.9429133858267716, + "grad_norm": 9.269134521484375, + "learning_rate": 9.473161033797219e-07, + "loss": 0.5911, + "step": 958 + }, + { + "epoch": 0.9438976377952756, + "grad_norm": 11.144103050231934, + "learning_rate": 9.483101391650101e-07, + "loss": 0.8383, + "step": 959 + }, + { + "epoch": 0.9448818897637795, + "grad_norm": 16.97820472717285, + "learning_rate": 9.493041749502983e-07, + "loss": 1.0647, + "step": 960 + }, + { + "epoch": 0.9458661417322834, + "grad_norm": 10.413034439086914, + "learning_rate": 9.502982107355867e-07, + "loss": 0.781, + "step": 961 + }, + { + "epoch": 0.9468503937007874, + "grad_norm": 12.050930976867676, + "learning_rate": 9.512922465208749e-07, + "loss": 0.6583, + "step": 962 + }, + { + "epoch": 0.9478346456692913, + "grad_norm": 11.038022994995117, + "learning_rate": 9.522862823061631e-07, + "loss": 0.8397, + "step": 963 + }, + { + "epoch": 0.9488188976377953, + "grad_norm": 15.14192008972168, + "learning_rate": 9.532803180914515e-07, + "loss": 1.1975, + "step": 964 + }, + { + "epoch": 0.9498031496062992, + "grad_norm": 8.80117130279541, + "learning_rate": 9.542743538767396e-07, + "loss": 0.3651, + "step": 965 + }, + { + "epoch": 0.9507874015748031, + "grad_norm": 10.17732048034668, + "learning_rate": 9.55268389662028e-07, + "loss": 0.6498, + "step": 966 + }, + { + "epoch": 0.9517716535433071, + "grad_norm": 26.41754722595215, + "learning_rate": 9.562624254473162e-07, + "loss": 2.2642, + "step": 967 + }, + { + "epoch": 0.952755905511811, + "grad_norm": 9.346402168273926, + "learning_rate": 9.572564612326044e-07, + "loss": 0.4515, + "step": 968 + }, + { + "epoch": 0.9537401574803149, + "grad_norm": 12.634305000305176, + "learning_rate": 9.582504970178928e-07, + "loss": 0.9761, + "step": 969 + }, + { + "epoch": 0.9547244094488189, + "grad_norm": 27.351110458374023, + "learning_rate": 9.59244532803181e-07, + "loss": 1.8422, + "step": 970 + }, + { + "epoch": 0.9557086614173228, + "grad_norm": 19.948509216308594, + "learning_rate": 9.602385685884694e-07, + "loss": 1.5052, + "step": 971 + }, + { + "epoch": 0.9566929133858267, + "grad_norm": 12.549859046936035, + "learning_rate": 9.612326043737576e-07, + "loss": 0.9086, + "step": 972 + }, + { + "epoch": 0.9576771653543307, + "grad_norm": 13.942103385925293, + "learning_rate": 9.622266401590458e-07, + "loss": 1.0606, + "step": 973 + }, + { + "epoch": 0.9586614173228346, + "grad_norm": 14.972650527954102, + "learning_rate": 9.632206759443342e-07, + "loss": 1.2389, + "step": 974 + }, + { + "epoch": 0.9596456692913385, + "grad_norm": 11.518607139587402, + "learning_rate": 9.642147117296224e-07, + "loss": 0.4794, + "step": 975 + }, + { + "epoch": 0.9606299212598425, + "grad_norm": 19.139293670654297, + "learning_rate": 9.652087475149109e-07, + "loss": 1.311, + "step": 976 + }, + { + "epoch": 0.9616141732283464, + "grad_norm": 16.821123123168945, + "learning_rate": 9.66202783300199e-07, + "loss": 1.2144, + "step": 977 + }, + { + "epoch": 0.9625984251968503, + "grad_norm": 9.898756980895996, + "learning_rate": 9.671968190854873e-07, + "loss": 0.5207, + "step": 978 + }, + { + "epoch": 0.9635826771653543, + "grad_norm": 14.798030853271484, + "learning_rate": 9.681908548707755e-07, + "loss": 0.8678, + "step": 979 + }, + { + "epoch": 0.9645669291338582, + "grad_norm": 16.760276794433594, + "learning_rate": 9.691848906560639e-07, + "loss": 1.0381, + "step": 980 + }, + { + "epoch": 0.9655511811023622, + "grad_norm": 14.848541259765625, + "learning_rate": 9.70178926441352e-07, + "loss": 0.8134, + "step": 981 + }, + { + "epoch": 0.9665354330708661, + "grad_norm": 14.691442489624023, + "learning_rate": 9.711729622266403e-07, + "loss": 1.1715, + "step": 982 + }, + { + "epoch": 0.96751968503937, + "grad_norm": 13.612778663635254, + "learning_rate": 9.721669980119285e-07, + "loss": 1.352, + "step": 983 + }, + { + "epoch": 0.968503937007874, + "grad_norm": 15.93306827545166, + "learning_rate": 9.731610337972169e-07, + "loss": 1.0196, + "step": 984 + }, + { + "epoch": 0.969488188976378, + "grad_norm": 11.488577842712402, + "learning_rate": 9.74155069582505e-07, + "loss": 0.8284, + "step": 985 + }, + { + "epoch": 0.9704724409448819, + "grad_norm": 10.168137550354004, + "learning_rate": 9.751491053677933e-07, + "loss": 0.7436, + "step": 986 + }, + { + "epoch": 0.9714566929133859, + "grad_norm": 10.076919555664062, + "learning_rate": 9.761431411530817e-07, + "loss": 0.5441, + "step": 987 + }, + { + "epoch": 0.9724409448818898, + "grad_norm": 22.378480911254883, + "learning_rate": 9.771371769383699e-07, + "loss": 1.2086, + "step": 988 + }, + { + "epoch": 0.9734251968503937, + "grad_norm": 19.3181209564209, + "learning_rate": 9.781312127236583e-07, + "loss": 1.1172, + "step": 989 + }, + { + "epoch": 0.9744094488188977, + "grad_norm": 10.783356666564941, + "learning_rate": 9.791252485089465e-07, + "loss": 0.6686, + "step": 990 + }, + { + "epoch": 0.9753937007874016, + "grad_norm": 9.608838081359863, + "learning_rate": 9.801192842942347e-07, + "loss": 0.696, + "step": 991 + }, + { + "epoch": 0.9763779527559056, + "grad_norm": 9.590912818908691, + "learning_rate": 9.811133200795231e-07, + "loss": 0.6819, + "step": 992 + }, + { + "epoch": 0.9773622047244095, + "grad_norm": 25.78241729736328, + "learning_rate": 9.821073558648113e-07, + "loss": 1.1375, + "step": 993 + }, + { + "epoch": 0.9783464566929134, + "grad_norm": 12.179429054260254, + "learning_rate": 9.831013916500995e-07, + "loss": 0.9031, + "step": 994 + }, + { + "epoch": 0.9793307086614174, + "grad_norm": 11.945301055908203, + "learning_rate": 9.84095427435388e-07, + "loss": 0.7194, + "step": 995 + }, + { + "epoch": 0.9803149606299213, + "grad_norm": 10.541696548461914, + "learning_rate": 9.850894632206761e-07, + "loss": 0.7496, + "step": 996 + }, + { + "epoch": 0.9812992125984252, + "grad_norm": 10.725748062133789, + "learning_rate": 9.860834990059645e-07, + "loss": 0.5843, + "step": 997 + }, + { + "epoch": 0.9822834645669292, + "grad_norm": 12.232301712036133, + "learning_rate": 9.870775347912527e-07, + "loss": 0.8935, + "step": 998 + }, + { + "epoch": 0.9832677165354331, + "grad_norm": 12.933833122253418, + "learning_rate": 9.88071570576541e-07, + "loss": 0.7518, + "step": 999 + }, + { + "epoch": 0.984251968503937, + "grad_norm": 16.762495040893555, + "learning_rate": 9.890656063618291e-07, + "loss": 1.3913, + "step": 1000 + }, + { + "epoch": 0.985236220472441, + "grad_norm": 11.918944358825684, + "learning_rate": 9.900596421471175e-07, + "loss": 0.6707, + "step": 1001 + }, + { + "epoch": 0.9862204724409449, + "grad_norm": 35.70449447631836, + "learning_rate": 9.910536779324057e-07, + "loss": 2.5054, + "step": 1002 + }, + { + "epoch": 0.9872047244094488, + "grad_norm": 29.08394432067871, + "learning_rate": 9.92047713717694e-07, + "loss": 2.2562, + "step": 1003 + }, + { + "epoch": 0.9881889763779528, + "grad_norm": 13.002362251281738, + "learning_rate": 9.930417495029821e-07, + "loss": 1.1758, + "step": 1004 + }, + { + "epoch": 0.9891732283464567, + "grad_norm": 10.518692016601562, + "learning_rate": 9.940357852882705e-07, + "loss": 0.7679, + "step": 1005 + }, + { + "epoch": 0.9901574803149606, + "grad_norm": 20.907424926757812, + "learning_rate": 9.950298210735587e-07, + "loss": 1.2762, + "step": 1006 + }, + { + "epoch": 0.9911417322834646, + "grad_norm": 10.026046752929688, + "learning_rate": 9.96023856858847e-07, + "loss": 0.7565, + "step": 1007 + }, + { + "epoch": 0.9921259842519685, + "grad_norm": 12.306164741516113, + "learning_rate": 9.970178926441354e-07, + "loss": 0.9347, + "step": 1008 + }, + { + "epoch": 0.9931102362204725, + "grad_norm": 7.531988620758057, + "learning_rate": 9.980119284294236e-07, + "loss": 0.4468, + "step": 1009 + }, + { + "epoch": 0.9940944881889764, + "grad_norm": 13.031584739685059, + "learning_rate": 9.99005964214712e-07, + "loss": 1.0553, + "step": 1010 + }, + { + "epoch": 0.9950787401574803, + "grad_norm": 12.94326114654541, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.9899, + "step": 1011 + }, + { + "epoch": 0.9960629921259843, + "grad_norm": 10.048609733581543, + "learning_rate": 9.999994102361802e-07, + "loss": 0.562, + "step": 1012 + }, + { + "epoch": 0.9970472440944882, + "grad_norm": 7.913425922393799, + "learning_rate": 9.999976409461164e-07, + "loss": 0.4784, + "step": 1013 + }, + { + "epoch": 0.9980314960629921, + "grad_norm": 18.53791046142578, + "learning_rate": 9.999946921339963e-07, + "loss": 1.4621, + "step": 1014 + }, + { + "epoch": 0.9990157480314961, + "grad_norm": 13.517660140991211, + "learning_rate": 9.999905638068e-07, + "loss": 0.9868, + "step": 1015 + }, + { + "epoch": 1.0, + "grad_norm": 11.711966514587402, + "learning_rate": 9.999852559742985e-07, + "loss": 0.6348, + "step": 1016 + }, + { + "epoch": 1.000984251968504, + "grad_norm": 12.872618675231934, + "learning_rate": 9.999787686490555e-07, + "loss": 1.0253, + "step": 1017 + }, + { + "epoch": 1.0019685039370079, + "grad_norm": 18.18792152404785, + "learning_rate": 9.99971101846426e-07, + "loss": 2.0283, + "step": 1018 + }, + { + "epoch": 1.0029527559055118, + "grad_norm": 8.895248413085938, + "learning_rate": 9.999622555845573e-07, + "loss": 0.5054, + "step": 1019 + }, + { + "epoch": 1.0039370078740157, + "grad_norm": 22.991552352905273, + "learning_rate": 9.999522298843875e-07, + "loss": 1.0792, + "step": 1020 + }, + { + "epoch": 1.0049212598425197, + "grad_norm": 24.473651885986328, + "learning_rate": 9.99941024769647e-07, + "loss": 1.2756, + "step": 1021 + }, + { + "epoch": 1.0059055118110236, + "grad_norm": 11.652467727661133, + "learning_rate": 9.999286402668578e-07, + "loss": 0.8319, + "step": 1022 + }, + { + "epoch": 1.0068897637795275, + "grad_norm": 19.23595428466797, + "learning_rate": 9.999150764053334e-07, + "loss": 1.191, + "step": 1023 + }, + { + "epoch": 1.0078740157480315, + "grad_norm": 19.412294387817383, + "learning_rate": 9.999003332171784e-07, + "loss": 1.3041, + "step": 1024 + }, + { + "epoch": 1.0088582677165354, + "grad_norm": 10.123785972595215, + "learning_rate": 9.998844107372894e-07, + "loss": 0.7303, + "step": 1025 + }, + { + "epoch": 1.0098425196850394, + "grad_norm": 12.623733520507812, + "learning_rate": 9.99867309003354e-07, + "loss": 0.8216, + "step": 1026 + }, + { + "epoch": 1.0108267716535433, + "grad_norm": 11.097447395324707, + "learning_rate": 9.998490280558511e-07, + "loss": 0.7877, + "step": 1027 + }, + { + "epoch": 1.0118110236220472, + "grad_norm": 11.763712882995605, + "learning_rate": 9.998295679380506e-07, + "loss": 0.7625, + "step": 1028 + }, + { + "epoch": 1.0127952755905512, + "grad_norm": 22.43312644958496, + "learning_rate": 9.998089286960135e-07, + "loss": 1.072, + "step": 1029 + }, + { + "epoch": 1.013779527559055, + "grad_norm": 10.953347206115723, + "learning_rate": 9.99787110378592e-07, + "loss": 0.9892, + "step": 1030 + }, + { + "epoch": 1.014763779527559, + "grad_norm": 13.673620223999023, + "learning_rate": 9.997641130374284e-07, + "loss": 1.3609, + "step": 1031 + }, + { + "epoch": 1.015748031496063, + "grad_norm": 18.985450744628906, + "learning_rate": 9.997399367269565e-07, + "loss": 1.1012, + "step": 1032 + }, + { + "epoch": 1.016732283464567, + "grad_norm": 13.805867195129395, + "learning_rate": 9.997145815044003e-07, + "loss": 1.1633, + "step": 1033 + }, + { + "epoch": 1.0177165354330708, + "grad_norm": 12.622978210449219, + "learning_rate": 9.99688047429774e-07, + "loss": 0.7493, + "step": 1034 + }, + { + "epoch": 1.0187007874015748, + "grad_norm": 11.227761268615723, + "learning_rate": 9.996603345658824e-07, + "loss": 0.8047, + "step": 1035 + }, + { + "epoch": 1.0196850393700787, + "grad_norm": 9.230905532836914, + "learning_rate": 9.996314429783204e-07, + "loss": 0.6312, + "step": 1036 + }, + { + "epoch": 1.0206692913385826, + "grad_norm": 12.647571563720703, + "learning_rate": 9.99601372735473e-07, + "loss": 0.9001, + "step": 1037 + }, + { + "epoch": 1.0216535433070866, + "grad_norm": 12.069604873657227, + "learning_rate": 9.995701239085143e-07, + "loss": 1.0504, + "step": 1038 + }, + { + "epoch": 1.0226377952755905, + "grad_norm": 18.552156448364258, + "learning_rate": 9.99537696571409e-07, + "loss": 1.1305, + "step": 1039 + }, + { + "epoch": 1.0236220472440944, + "grad_norm": 11.786487579345703, + "learning_rate": 9.995040908009107e-07, + "loss": 0.6795, + "step": 1040 + }, + { + "epoch": 1.0246062992125984, + "grad_norm": 10.507966995239258, + "learning_rate": 9.994693066765622e-07, + "loss": 0.6223, + "step": 1041 + }, + { + "epoch": 1.0255905511811023, + "grad_norm": 19.367963790893555, + "learning_rate": 9.994333442806959e-07, + "loss": 1.4211, + "step": 1042 + }, + { + "epoch": 1.0265748031496063, + "grad_norm": 12.668354988098145, + "learning_rate": 9.993962036984325e-07, + "loss": 0.7079, + "step": 1043 + }, + { + "epoch": 1.0275590551181102, + "grad_norm": 12.891959190368652, + "learning_rate": 9.99357885017682e-07, + "loss": 0.84, + "step": 1044 + }, + { + "epoch": 1.0285433070866141, + "grad_norm": 12.73304271697998, + "learning_rate": 9.993183883291425e-07, + "loss": 0.9718, + "step": 1045 + }, + { + "epoch": 1.029527559055118, + "grad_norm": 19.29059410095215, + "learning_rate": 9.992777137263007e-07, + "loss": 1.4166, + "step": 1046 + }, + { + "epoch": 1.030511811023622, + "grad_norm": 18.758543014526367, + "learning_rate": 9.99235861305431e-07, + "loss": 1.0734, + "step": 1047 + }, + { + "epoch": 1.031496062992126, + "grad_norm": 13.670878410339355, + "learning_rate": 9.991928311655956e-07, + "loss": 1.0404, + "step": 1048 + }, + { + "epoch": 1.0324803149606299, + "grad_norm": 8.387916564941406, + "learning_rate": 9.991486234086447e-07, + "loss": 0.2556, + "step": 1049 + }, + { + "epoch": 1.0334645669291338, + "grad_norm": 9.226862907409668, + "learning_rate": 9.991032381392158e-07, + "loss": 0.4799, + "step": 1050 + }, + { + "epoch": 1.0344488188976377, + "grad_norm": 10.872235298156738, + "learning_rate": 9.990566754647328e-07, + "loss": 0.6954, + "step": 1051 + }, + { + "epoch": 1.0354330708661417, + "grad_norm": 18.022737503051758, + "learning_rate": 9.990089354954073e-07, + "loss": 1.1927, + "step": 1052 + }, + { + "epoch": 1.0364173228346456, + "grad_norm": 11.728109359741211, + "learning_rate": 9.989600183442373e-07, + "loss": 0.825, + "step": 1053 + }, + { + "epoch": 1.0374015748031495, + "grad_norm": 15.314092636108398, + "learning_rate": 9.98909924127007e-07, + "loss": 1.303, + "step": 1054 + }, + { + "epoch": 1.0383858267716535, + "grad_norm": 9.027541160583496, + "learning_rate": 9.988586529622864e-07, + "loss": 0.6624, + "step": 1055 + }, + { + "epoch": 1.0393700787401574, + "grad_norm": 17.31993865966797, + "learning_rate": 9.98806204971432e-07, + "loss": 1.2299, + "step": 1056 + }, + { + "epoch": 1.0403543307086613, + "grad_norm": 10.081514358520508, + "learning_rate": 9.987525802785847e-07, + "loss": 0.5722, + "step": 1057 + }, + { + "epoch": 1.0413385826771653, + "grad_norm": 8.890061378479004, + "learning_rate": 9.986977790106715e-07, + "loss": 0.4275, + "step": 1058 + }, + { + "epoch": 1.0423228346456692, + "grad_norm": 18.210792541503906, + "learning_rate": 9.986418012974042e-07, + "loss": 1.2038, + "step": 1059 + }, + { + "epoch": 1.0433070866141732, + "grad_norm": 12.176244735717773, + "learning_rate": 9.985846472712786e-07, + "loss": 1.1635, + "step": 1060 + }, + { + "epoch": 1.044291338582677, + "grad_norm": 9.834606170654297, + "learning_rate": 9.985263170675756e-07, + "loss": 0.5874, + "step": 1061 + }, + { + "epoch": 1.045275590551181, + "grad_norm": 9.242854118347168, + "learning_rate": 9.98466810824359e-07, + "loss": 0.5254, + "step": 1062 + }, + { + "epoch": 1.046259842519685, + "grad_norm": 10.408865928649902, + "learning_rate": 9.984061286824774e-07, + "loss": 0.6062, + "step": 1063 + }, + { + "epoch": 1.047244094488189, + "grad_norm": 11.639140129089355, + "learning_rate": 9.983442707855616e-07, + "loss": 0.8536, + "step": 1064 + }, + { + "epoch": 1.0482283464566928, + "grad_norm": 17.28397560119629, + "learning_rate": 9.982812372800262e-07, + "loss": 1.0845, + "step": 1065 + }, + { + "epoch": 1.0492125984251968, + "grad_norm": 16.99485969543457, + "learning_rate": 9.982170283150678e-07, + "loss": 1.2005, + "step": 1066 + }, + { + "epoch": 1.0501968503937007, + "grad_norm": 14.910852432250977, + "learning_rate": 9.981516440426659e-07, + "loss": 1.1397, + "step": 1067 + }, + { + "epoch": 1.0511811023622046, + "grad_norm": 13.27585506439209, + "learning_rate": 9.98085084617581e-07, + "loss": 0.6992, + "step": 1068 + }, + { + "epoch": 1.0521653543307086, + "grad_norm": 13.180895805358887, + "learning_rate": 9.980173501973558e-07, + "loss": 1.1104, + "step": 1069 + }, + { + "epoch": 1.0531496062992125, + "grad_norm": 10.505107879638672, + "learning_rate": 9.97948440942314e-07, + "loss": 0.4401, + "step": 1070 + }, + { + "epoch": 1.0541338582677164, + "grad_norm": 11.06208324432373, + "learning_rate": 9.978783570155597e-07, + "loss": 0.6145, + "step": 1071 + }, + { + "epoch": 1.0541338582677164, + "eval_Qnli-dev_cosine_accuracy": 0.6796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8589683175086975, + "eval_Qnli-dev_cosine_ap": 0.6811526925251525, + "eval_Qnli-dev_cosine_f1": 0.6757679180887372, + "eval_Qnli-dev_cosine_f1_threshold": 0.7798424363136292, + "eval_Qnli-dev_cosine_precision": 0.5657142857142857, + "eval_Qnli-dev_cosine_recall": 0.8389830508474576, + "eval_Qnli-dev_dot_accuracy": 0.673828125, + "eval_Qnli-dev_dot_accuracy_threshold": 657.5110473632812, + "eval_Qnli-dev_dot_ap": 0.6806278763147593, + "eval_Qnli-dev_dot_f1": 0.6756756756756757, + "eval_Qnli-dev_dot_f1_threshold": 594.4506225585938, + "eval_Qnli-dev_dot_precision": 0.5617977528089888, + "eval_Qnli-dev_dot_recall": 0.847457627118644, + "eval_Qnli-dev_euclidean_accuracy": 0.6796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.64499282836914, + "eval_Qnli-dev_euclidean_ap": 0.6811071531134263, + "eval_Qnli-dev_euclidean_f1": 0.6768707482993197, + "eval_Qnli-dev_euclidean_f1_threshold": 18.386024475097656, + "eval_Qnli-dev_euclidean_precision": 0.5653409090909091, + "eval_Qnli-dev_euclidean_recall": 0.8432203389830508, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 313.95770263671875, + "eval_Qnli-dev_manhattan_ap": 0.687811012744714, + "eval_Qnli-dev_manhattan_f1": 0.6812816188870151, + "eval_Qnli-dev_manhattan_f1_threshold": 386.62255859375, + "eval_Qnli-dev_manhattan_precision": 0.5658263305322129, + "eval_Qnli-dev_manhattan_recall": 0.8559322033898306, + "eval_Qnli-dev_max_accuracy": 0.6796875, + "eval_Qnli-dev_max_accuracy_threshold": 657.5110473632812, + "eval_Qnli-dev_max_ap": 0.687811012744714, + "eval_Qnli-dev_max_f1": 0.6812816188870151, + "eval_Qnli-dev_max_f1_threshold": 594.4506225585938, + "eval_Qnli-dev_max_precision": 0.5658263305322129, + "eval_Qnli-dev_max_recall": 0.8559322033898306, + "eval_allNLI-dev_cosine_accuracy": 0.685546875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9436202049255371, + "eval_allNLI-dev_cosine_ap": 0.49491949487120757, + "eval_allNLI-dev_cosine_f1": 0.5363984674329503, + "eval_allNLI-dev_cosine_f1_threshold": 0.8473018407821655, + "eval_allNLI-dev_cosine_precision": 0.40114613180515757, + "eval_allNLI-dev_cosine_recall": 0.8092485549132948, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 720.7135009765625, + "eval_allNLI-dev_dot_ap": 0.4932256899651205, + "eval_allNLI-dev_dot_f1": 0.5400372439478585, + "eval_allNLI-dev_dot_f1_threshold": 645.960205078125, + "eval_allNLI-dev_dot_precision": 0.3983516483516483, + "eval_allNLI-dev_dot_recall": 0.838150289017341, + "eval_allNLI-dev_euclidean_accuracy": 0.685546875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.2955322265625, + "eval_allNLI-dev_euclidean_ap": 0.49518764112432034, + "eval_allNLI-dev_euclidean_f1": 0.5374280230326295, + "eval_allNLI-dev_euclidean_f1_threshold": 15.273109436035156, + "eval_allNLI-dev_euclidean_precision": 0.40229885057471265, + "eval_allNLI-dev_euclidean_recall": 0.8092485549132948, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 194.04481506347656, + "eval_allNLI-dev_manhattan_ap": 0.4970543460207635, + "eval_allNLI-dev_manhattan_f1": 0.5360824742268041, + "eval_allNLI-dev_manhattan_f1_threshold": 302.1809387207031, + "eval_allNLI-dev_manhattan_precision": 0.4166666666666667, + "eval_allNLI-dev_manhattan_recall": 0.7514450867052023, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 720.7135009765625, + "eval_allNLI-dev_max_ap": 0.4970543460207635, + "eval_allNLI-dev_max_f1": 0.5400372439478585, + "eval_allNLI-dev_max_f1_threshold": 645.960205078125, + "eval_allNLI-dev_max_precision": 0.4166666666666667, + "eval_allNLI-dev_max_recall": 0.838150289017341, + "eval_loss": 0.8866996169090271, + "eval_runtime": 51.3443, + "eval_samples_per_second": 32.409, + "eval_sequential_score": 0.687811012744714, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.46419451684421176, + "eval_sts-test_pearson_dot": 0.46134985517735105, + "eval_sts-test_pearson_euclidean": 0.48589958368920916, + "eval_sts-test_pearson_manhattan": 0.5015187277321582, + "eval_sts-test_pearson_max": 0.5015187277321582, + "eval_sts-test_spearman_cosine": 0.5020793037201544, + "eval_sts-test_spearman_dot": 0.49904221461456577, + "eval_sts-test_spearman_euclidean": 0.5024715932775115, + "eval_sts-test_spearman_manhattan": 0.5173440484832835, + "eval_sts-test_spearman_max": 0.5173440484832835, + "step": 1071 + }, + { + "epoch": 1.0551181102362204, + "grad_norm": 13.029767036437988, + "learning_rate": 9.978070985829782e-07, + "loss": 0.8673, + "step": 1072 + }, + { + "epoch": 1.0561023622047243, + "grad_norm": 11.501550674438477, + "learning_rate": 9.97734665813234e-07, + "loss": 0.4861, + "step": 1073 + }, + { + "epoch": 1.0570866141732282, + "grad_norm": 11.720154762268066, + "learning_rate": 9.976610588777717e-07, + "loss": 0.6278, + "step": 1074 + }, + { + "epoch": 1.0580708661417322, + "grad_norm": 14.068936347961426, + "learning_rate": 9.975862779508149e-07, + "loss": 1.1955, + "step": 1075 + }, + { + "epoch": 1.0590551181102361, + "grad_norm": 9.46675968170166, + "learning_rate": 9.975103232093656e-07, + "loss": 0.5441, + "step": 1076 + }, + { + "epoch": 1.06003937007874, + "grad_norm": 13.920763969421387, + "learning_rate": 9.974331948332048e-07, + "loss": 1.2355, + "step": 1077 + }, + { + "epoch": 1.061023622047244, + "grad_norm": 15.3096342086792, + "learning_rate": 9.97354893004891e-07, + "loss": 0.7504, + "step": 1078 + }, + { + "epoch": 1.0620078740157481, + "grad_norm": 14.229239463806152, + "learning_rate": 9.972754179097602e-07, + "loss": 0.5964, + "step": 1079 + }, + { + "epoch": 1.0629921259842519, + "grad_norm": 14.73090648651123, + "learning_rate": 9.971947697359257e-07, + "loss": 0.7314, + "step": 1080 + }, + { + "epoch": 1.063976377952756, + "grad_norm": 12.88805103302002, + "learning_rate": 9.971129486742777e-07, + "loss": 0.5723, + "step": 1081 + }, + { + "epoch": 1.0649606299212597, + "grad_norm": 11.795894622802734, + "learning_rate": 9.970299549184814e-07, + "loss": 0.5942, + "step": 1082 + }, + { + "epoch": 1.0659448818897639, + "grad_norm": 28.093719482421875, + "learning_rate": 9.969457886649791e-07, + "loss": 1.9039, + "step": 1083 + }, + { + "epoch": 1.0669291338582678, + "grad_norm": 13.558938026428223, + "learning_rate": 9.968604501129877e-07, + "loss": 1.2122, + "step": 1084 + }, + { + "epoch": 1.0679133858267718, + "grad_norm": 8.226652145385742, + "learning_rate": 9.967739394644982e-07, + "loss": 0.4748, + "step": 1085 + }, + { + "epoch": 1.0688976377952757, + "grad_norm": 10.694149017333984, + "learning_rate": 9.966862569242773e-07, + "loss": 0.6057, + "step": 1086 + }, + { + "epoch": 1.0698818897637796, + "grad_norm": 11.170210838317871, + "learning_rate": 9.965974026998642e-07, + "loss": 0.5627, + "step": 1087 + }, + { + "epoch": 1.0708661417322836, + "grad_norm": 7.397644996643066, + "learning_rate": 9.965073770015722e-07, + "loss": 0.2975, + "step": 1088 + }, + { + "epoch": 1.0718503937007875, + "grad_norm": 17.239303588867188, + "learning_rate": 9.964161800424873e-07, + "loss": 0.6175, + "step": 1089 + }, + { + "epoch": 1.0728346456692914, + "grad_norm": 9.651748657226562, + "learning_rate": 9.963238120384677e-07, + "loss": 0.5306, + "step": 1090 + }, + { + "epoch": 1.0738188976377954, + "grad_norm": 9.865948677062988, + "learning_rate": 9.962302732081432e-07, + "loss": 0.5697, + "step": 1091 + }, + { + "epoch": 1.0748031496062993, + "grad_norm": 9.394389152526855, + "learning_rate": 9.961355637729153e-07, + "loss": 0.466, + "step": 1092 + }, + { + "epoch": 1.0757874015748032, + "grad_norm": 11.819548606872559, + "learning_rate": 9.960396839569557e-07, + "loss": 0.9722, + "step": 1093 + }, + { + "epoch": 1.0767716535433072, + "grad_norm": 12.320271492004395, + "learning_rate": 9.95942633987207e-07, + "loss": 1.0696, + "step": 1094 + }, + { + "epoch": 1.077755905511811, + "grad_norm": 9.119085311889648, + "learning_rate": 9.95844414093381e-07, + "loss": 0.5533, + "step": 1095 + }, + { + "epoch": 1.078740157480315, + "grad_norm": 11.696290016174316, + "learning_rate": 9.957450245079588e-07, + "loss": 0.6409, + "step": 1096 + }, + { + "epoch": 1.079724409448819, + "grad_norm": 12.307863235473633, + "learning_rate": 9.9564446546619e-07, + "loss": 0.6046, + "step": 1097 + }, + { + "epoch": 1.080708661417323, + "grad_norm": 12.855661392211914, + "learning_rate": 9.955427372060928e-07, + "loss": 0.8067, + "step": 1098 + }, + { + "epoch": 1.0816929133858268, + "grad_norm": 11.116859436035156, + "learning_rate": 9.954398399684515e-07, + "loss": 0.5695, + "step": 1099 + }, + { + "epoch": 1.0826771653543308, + "grad_norm": 14.217082023620605, + "learning_rate": 9.953357739968192e-07, + "loss": 0.6817, + "step": 1100 + }, + { + "epoch": 1.0836614173228347, + "grad_norm": 12.979687690734863, + "learning_rate": 9.952305395375137e-07, + "loss": 0.9963, + "step": 1101 + }, + { + "epoch": 1.0846456692913387, + "grad_norm": 12.859124183654785, + "learning_rate": 9.951241368396194e-07, + "loss": 0.9799, + "step": 1102 + }, + { + "epoch": 1.0856299212598426, + "grad_norm": 15.441899299621582, + "learning_rate": 9.950165661549856e-07, + "loss": 1.3903, + "step": 1103 + }, + { + "epoch": 1.0866141732283465, + "grad_norm": 17.483436584472656, + "learning_rate": 9.949078277382262e-07, + "loss": 1.019, + "step": 1104 + }, + { + "epoch": 1.0875984251968505, + "grad_norm": 9.914593696594238, + "learning_rate": 9.947979218467193e-07, + "loss": 0.5109, + "step": 1105 + }, + { + "epoch": 1.0885826771653544, + "grad_norm": 20.513473510742188, + "learning_rate": 9.946868487406057e-07, + "loss": 1.0597, + "step": 1106 + }, + { + "epoch": 1.0895669291338583, + "grad_norm": 5.7887115478515625, + "learning_rate": 9.945746086827896e-07, + "loss": 0.1463, + "step": 1107 + }, + { + "epoch": 1.0905511811023623, + "grad_norm": 10.948200225830078, + "learning_rate": 9.94461201938937e-07, + "loss": 0.5552, + "step": 1108 + }, + { + "epoch": 1.0915354330708662, + "grad_norm": 11.003507614135742, + "learning_rate": 9.943466287774756e-07, + "loss": 0.5808, + "step": 1109 + }, + { + "epoch": 1.0925196850393701, + "grad_norm": 13.039937973022461, + "learning_rate": 9.942308894695932e-07, + "loss": 0.82, + "step": 1110 + }, + { + "epoch": 1.093503937007874, + "grad_norm": 25.417362213134766, + "learning_rate": 9.941139842892391e-07, + "loss": 1.4818, + "step": 1111 + }, + { + "epoch": 1.094488188976378, + "grad_norm": 27.56861686706543, + "learning_rate": 9.939959135131214e-07, + "loss": 1.7281, + "step": 1112 + }, + { + "epoch": 1.095472440944882, + "grad_norm": 6.9113240242004395, + "learning_rate": 9.938766774207064e-07, + "loss": 0.342, + "step": 1113 + }, + { + "epoch": 1.0964566929133859, + "grad_norm": 10.753077507019043, + "learning_rate": 9.937562762942202e-07, + "loss": 0.5543, + "step": 1114 + }, + { + "epoch": 1.0974409448818898, + "grad_norm": 15.725546836853027, + "learning_rate": 9.936347104186452e-07, + "loss": 1.0291, + "step": 1115 + }, + { + "epoch": 1.0984251968503937, + "grad_norm": 23.7658634185791, + "learning_rate": 9.935119800817215e-07, + "loss": 1.4693, + "step": 1116 + }, + { + "epoch": 1.0994094488188977, + "grad_norm": 9.444933891296387, + "learning_rate": 9.933880855739446e-07, + "loss": 0.6533, + "step": 1117 + }, + { + "epoch": 1.1003937007874016, + "grad_norm": 13.30677604675293, + "learning_rate": 9.932630271885662e-07, + "loss": 0.8667, + "step": 1118 + }, + { + "epoch": 1.1013779527559056, + "grad_norm": 17.388879776000977, + "learning_rate": 9.93136805221593e-07, + "loss": 1.3982, + "step": 1119 + }, + { + "epoch": 1.1023622047244095, + "grad_norm": 9.378705978393555, + "learning_rate": 9.930094199717848e-07, + "loss": 0.4305, + "step": 1120 + }, + { + "epoch": 1.1033464566929134, + "grad_norm": 20.85811996459961, + "learning_rate": 9.92880871740656e-07, + "loss": 1.248, + "step": 1121 + }, + { + "epoch": 1.1043307086614174, + "grad_norm": 18.253276824951172, + "learning_rate": 9.927511608324732e-07, + "loss": 1.1434, + "step": 1122 + }, + { + "epoch": 1.1053149606299213, + "grad_norm": 17.432559967041016, + "learning_rate": 9.926202875542546e-07, + "loss": 1.0944, + "step": 1123 + }, + { + "epoch": 1.1062992125984252, + "grad_norm": 19.3885555267334, + "learning_rate": 9.924882522157706e-07, + "loss": 1.2556, + "step": 1124 + }, + { + "epoch": 1.1072834645669292, + "grad_norm": 14.321420669555664, + "learning_rate": 9.923550551295411e-07, + "loss": 0.893, + "step": 1125 + }, + { + "epoch": 1.108267716535433, + "grad_norm": 10.655247688293457, + "learning_rate": 9.922206966108367e-07, + "loss": 0.6589, + "step": 1126 + }, + { + "epoch": 1.109251968503937, + "grad_norm": 13.070340156555176, + "learning_rate": 9.920851769776761e-07, + "loss": 0.6495, + "step": 1127 + }, + { + "epoch": 1.110236220472441, + "grad_norm": 11.607705116271973, + "learning_rate": 9.919484965508276e-07, + "loss": 0.4612, + "step": 1128 + }, + { + "epoch": 1.111220472440945, + "grad_norm": 12.572264671325684, + "learning_rate": 9.918106556538057e-07, + "loss": 0.8137, + "step": 1129 + }, + { + "epoch": 1.1122047244094488, + "grad_norm": 13.347660064697266, + "learning_rate": 9.91671654612872e-07, + "loss": 0.6577, + "step": 1130 + }, + { + "epoch": 1.1131889763779528, + "grad_norm": 16.497905731201172, + "learning_rate": 9.91531493757035e-07, + "loss": 1.077, + "step": 1131 + }, + { + "epoch": 1.1141732283464567, + "grad_norm": 16.14683723449707, + "learning_rate": 9.913901734180476e-07, + "loss": 0.9539, + "step": 1132 + }, + { + "epoch": 1.1151574803149606, + "grad_norm": 11.908422470092773, + "learning_rate": 9.912476939304069e-07, + "loss": 0.6905, + "step": 1133 + }, + { + "epoch": 1.1161417322834646, + "grad_norm": 16.775802612304688, + "learning_rate": 9.911040556313543e-07, + "loss": 0.8103, + "step": 1134 + }, + { + "epoch": 1.1171259842519685, + "grad_norm": 10.2266845703125, + "learning_rate": 9.909592588608738e-07, + "loss": 0.5007, + "step": 1135 + }, + { + "epoch": 1.1181102362204725, + "grad_norm": 18.091506958007812, + "learning_rate": 9.908133039616911e-07, + "loss": 1.208, + "step": 1136 + }, + { + "epoch": 1.1190944881889764, + "grad_norm": 9.337846755981445, + "learning_rate": 9.90666191279274e-07, + "loss": 0.5792, + "step": 1137 + }, + { + "epoch": 1.1200787401574803, + "grad_norm": 11.969504356384277, + "learning_rate": 9.905179211618296e-07, + "loss": 0.7462, + "step": 1138 + }, + { + "epoch": 1.1210629921259843, + "grad_norm": 9.322251319885254, + "learning_rate": 9.903684939603054e-07, + "loss": 0.3858, + "step": 1139 + }, + { + "epoch": 1.1220472440944882, + "grad_norm": 12.901102066040039, + "learning_rate": 9.902179100283875e-07, + "loss": 0.9128, + "step": 1140 + }, + { + "epoch": 1.1230314960629921, + "grad_norm": 13.293919563293457, + "learning_rate": 9.900661697224995e-07, + "loss": 0.9166, + "step": 1141 + }, + { + "epoch": 1.124015748031496, + "grad_norm": 13.659887313842773, + "learning_rate": 9.899132734018024e-07, + "loss": 0.8742, + "step": 1142 + }, + { + "epoch": 1.125, + "grad_norm": 9.943990707397461, + "learning_rate": 9.897592214281936e-07, + "loss": 0.8224, + "step": 1143 + }, + { + "epoch": 1.125984251968504, + "grad_norm": 8.73592758178711, + "learning_rate": 9.896040141663054e-07, + "loss": 0.566, + "step": 1144 + }, + { + "epoch": 1.1269685039370079, + "grad_norm": 9.734535217285156, + "learning_rate": 9.894476519835052e-07, + "loss": 0.6497, + "step": 1145 + }, + { + "epoch": 1.1279527559055118, + "grad_norm": 23.456430435180664, + "learning_rate": 9.892901352498932e-07, + "loss": 2.2927, + "step": 1146 + }, + { + "epoch": 1.1289370078740157, + "grad_norm": 9.371479988098145, + "learning_rate": 9.891314643383037e-07, + "loss": 0.5446, + "step": 1147 + }, + { + "epoch": 1.1299212598425197, + "grad_norm": 11.9376859664917, + "learning_rate": 9.88971639624301e-07, + "loss": 0.9981, + "step": 1148 + }, + { + "epoch": 1.1309055118110236, + "grad_norm": 7.961652755737305, + "learning_rate": 9.888106614861823e-07, + "loss": 0.3982, + "step": 1149 + }, + { + "epoch": 1.1318897637795275, + "grad_norm": 8.365242958068848, + "learning_rate": 9.886485303049734e-07, + "loss": 0.3324, + "step": 1150 + }, + { + "epoch": 1.1328740157480315, + "grad_norm": 10.67441177368164, + "learning_rate": 9.884852464644305e-07, + "loss": 0.851, + "step": 1151 + }, + { + "epoch": 1.1338582677165354, + "grad_norm": 19.849498748779297, + "learning_rate": 9.883208103510367e-07, + "loss": 1.6992, + "step": 1152 + }, + { + "epoch": 1.1348425196850394, + "grad_norm": 9.721522331237793, + "learning_rate": 9.88155222354004e-07, + "loss": 0.6466, + "step": 1153 + }, + { + "epoch": 1.1358267716535433, + "grad_norm": 11.709403038024902, + "learning_rate": 9.879884828652698e-07, + "loss": 0.8126, + "step": 1154 + }, + { + "epoch": 1.1368110236220472, + "grad_norm": 11.224492073059082, + "learning_rate": 9.878205922794972e-07, + "loss": 0.6029, + "step": 1155 + }, + { + "epoch": 1.1377952755905512, + "grad_norm": 11.349569320678711, + "learning_rate": 9.876515509940742e-07, + "loss": 0.7146, + "step": 1156 + }, + { + "epoch": 1.138779527559055, + "grad_norm": 6.832194805145264, + "learning_rate": 9.874813594091122e-07, + "loss": 0.3069, + "step": 1157 + }, + { + "epoch": 1.139763779527559, + "grad_norm": 15.690938949584961, + "learning_rate": 9.873100179274453e-07, + "loss": 1.2919, + "step": 1158 + }, + { + "epoch": 1.140748031496063, + "grad_norm": 11.24889850616455, + "learning_rate": 9.871375269546292e-07, + "loss": 0.6137, + "step": 1159 + }, + { + "epoch": 1.141732283464567, + "grad_norm": 22.022859573364258, + "learning_rate": 9.869638868989407e-07, + "loss": 1.5394, + "step": 1160 + }, + { + "epoch": 1.1427165354330708, + "grad_norm": 11.28795051574707, + "learning_rate": 9.867890981713764e-07, + "loss": 0.5581, + "step": 1161 + }, + { + "epoch": 1.1437007874015748, + "grad_norm": 12.344659805297852, + "learning_rate": 9.866131611856514e-07, + "loss": 1.1485, + "step": 1162 + }, + { + "epoch": 1.1446850393700787, + "grad_norm": 15.431796073913574, + "learning_rate": 9.864360763581994e-07, + "loss": 1.175, + "step": 1163 + }, + { + "epoch": 1.1456692913385826, + "grad_norm": 7.843120098114014, + "learning_rate": 9.862578441081699e-07, + "loss": 0.3461, + "step": 1164 + }, + { + "epoch": 1.1466535433070866, + "grad_norm": 24.42163848876953, + "learning_rate": 9.860784648574288e-07, + "loss": 1.3923, + "step": 1165 + }, + { + "epoch": 1.1476377952755905, + "grad_norm": 9.632267951965332, + "learning_rate": 9.858979390305573e-07, + "loss": 0.5471, + "step": 1166 + }, + { + "epoch": 1.1486220472440944, + "grad_norm": 19.2584285736084, + "learning_rate": 9.857162670548502e-07, + "loss": 1.0431, + "step": 1167 + }, + { + "epoch": 1.1496062992125984, + "grad_norm": 9.059511184692383, + "learning_rate": 9.855334493603147e-07, + "loss": 0.3017, + "step": 1168 + }, + { + "epoch": 1.1505905511811023, + "grad_norm": 14.917661666870117, + "learning_rate": 9.853494863796705e-07, + "loss": 1.1278, + "step": 1169 + }, + { + "epoch": 1.1515748031496063, + "grad_norm": 19.67047119140625, + "learning_rate": 9.851643785483478e-07, + "loss": 1.1442, + "step": 1170 + }, + { + "epoch": 1.1525590551181102, + "grad_norm": 15.289731979370117, + "learning_rate": 9.849781263044867e-07, + "loss": 0.6557, + "step": 1171 + }, + { + "epoch": 1.1535433070866141, + "grad_norm": 11.469537734985352, + "learning_rate": 9.847907300889358e-07, + "loss": 0.6248, + "step": 1172 + }, + { + "epoch": 1.154527559055118, + "grad_norm": 20.226198196411133, + "learning_rate": 9.846021903452519e-07, + "loss": 1.478, + "step": 1173 + }, + { + "epoch": 1.155511811023622, + "grad_norm": 23.517227172851562, + "learning_rate": 9.844125075196983e-07, + "loss": 1.4457, + "step": 1174 + }, + { + "epoch": 1.156496062992126, + "grad_norm": 12.229169845581055, + "learning_rate": 9.842216820612438e-07, + "loss": 0.8652, + "step": 1175 + }, + { + "epoch": 1.1574803149606299, + "grad_norm": 9.687929153442383, + "learning_rate": 9.840297144215615e-07, + "loss": 0.7371, + "step": 1176 + }, + { + "epoch": 1.1584645669291338, + "grad_norm": 14.011287689208984, + "learning_rate": 9.838366050550287e-07, + "loss": 1.0206, + "step": 1177 + }, + { + "epoch": 1.1594488188976377, + "grad_norm": 13.678019523620605, + "learning_rate": 9.836423544187245e-07, + "loss": 0.8996, + "step": 1178 + }, + { + "epoch": 1.1604330708661417, + "grad_norm": 11.131888389587402, + "learning_rate": 9.834469629724293e-07, + "loss": 0.7363, + "step": 1179 + }, + { + "epoch": 1.1614173228346456, + "grad_norm": 14.733101844787598, + "learning_rate": 9.832504311786238e-07, + "loss": 1.3426, + "step": 1180 + }, + { + "epoch": 1.1624015748031495, + "grad_norm": 13.642965316772461, + "learning_rate": 9.830527595024885e-07, + "loss": 0.9533, + "step": 1181 + }, + { + "epoch": 1.1633858267716535, + "grad_norm": 6.219560146331787, + "learning_rate": 9.82853948411901e-07, + "loss": 0.3364, + "step": 1182 + }, + { + "epoch": 1.1643700787401574, + "grad_norm": 6.48429536819458, + "learning_rate": 9.826539983774365e-07, + "loss": 0.2821, + "step": 1183 + }, + { + "epoch": 1.1653543307086613, + "grad_norm": 12.023889541625977, + "learning_rate": 9.824529098723657e-07, + "loss": 0.7553, + "step": 1184 + }, + { + "epoch": 1.1663385826771653, + "grad_norm": 19.615596771240234, + "learning_rate": 9.822506833726537e-07, + "loss": 1.4646, + "step": 1185 + }, + { + "epoch": 1.1673228346456692, + "grad_norm": 13.54229736328125, + "learning_rate": 9.8204731935696e-07, + "loss": 1.2136, + "step": 1186 + }, + { + "epoch": 1.1683070866141732, + "grad_norm": 14.583290100097656, + "learning_rate": 9.818428183066356e-07, + "loss": 1.3401, + "step": 1187 + }, + { + "epoch": 1.169291338582677, + "grad_norm": 13.6580228805542, + "learning_rate": 9.816371807057237e-07, + "loss": 1.1959, + "step": 1188 + }, + { + "epoch": 1.170275590551181, + "grad_norm": 11.358354568481445, + "learning_rate": 9.81430407040957e-07, + "loss": 0.667, + "step": 1189 + }, + { + "epoch": 1.171259842519685, + "grad_norm": 11.186590194702148, + "learning_rate": 9.812224978017573e-07, + "loss": 0.932, + "step": 1190 + }, + { + "epoch": 1.172244094488189, + "grad_norm": 18.10564613342285, + "learning_rate": 9.810134534802346e-07, + "loss": 0.8288, + "step": 1191 + }, + { + "epoch": 1.1732283464566928, + "grad_norm": 8.587007522583008, + "learning_rate": 9.80803274571185e-07, + "loss": 0.6271, + "step": 1192 + }, + { + "epoch": 1.1742125984251968, + "grad_norm": 7.676675319671631, + "learning_rate": 9.80591961572091e-07, + "loss": 0.4196, + "step": 1193 + }, + { + "epoch": 1.1751968503937007, + "grad_norm": 12.111679077148438, + "learning_rate": 9.803795149831183e-07, + "loss": 0.9061, + "step": 1194 + }, + { + "epoch": 1.1761811023622046, + "grad_norm": 11.08476448059082, + "learning_rate": 9.801659353071165e-07, + "loss": 0.8713, + "step": 1195 + }, + { + "epoch": 1.1771653543307086, + "grad_norm": 20.66461944580078, + "learning_rate": 9.799512230496172e-07, + "loss": 1.1829, + "step": 1196 + }, + { + "epoch": 1.1781496062992125, + "grad_norm": 8.07231616973877, + "learning_rate": 9.79735378718832e-07, + "loss": 0.3551, + "step": 1197 + }, + { + "epoch": 1.1791338582677164, + "grad_norm": 18.250873565673828, + "learning_rate": 9.795184028256534e-07, + "loss": 1.3069, + "step": 1198 + }, + { + "epoch": 1.1801181102362204, + "grad_norm": 11.728555679321289, + "learning_rate": 9.793002958836506e-07, + "loss": 0.8112, + "step": 1199 + }, + { + "epoch": 1.1811023622047245, + "grad_norm": 8.41722583770752, + "learning_rate": 9.790810584090712e-07, + "loss": 0.2843, + "step": 1200 + }, + { + "epoch": 1.1820866141732282, + "grad_norm": 8.391276359558105, + "learning_rate": 9.78860690920838e-07, + "loss": 0.4981, + "step": 1201 + }, + { + "epoch": 1.1830708661417324, + "grad_norm": 12.04411792755127, + "learning_rate": 9.78639193940549e-07, + "loss": 0.7478, + "step": 1202 + }, + { + "epoch": 1.1840551181102361, + "grad_norm": 11.410354614257812, + "learning_rate": 9.784165679924752e-07, + "loss": 0.9183, + "step": 1203 + }, + { + "epoch": 1.1850393700787403, + "grad_norm": 8.285551071166992, + "learning_rate": 9.781928136035604e-07, + "loss": 0.5247, + "step": 1204 + }, + { + "epoch": 1.186023622047244, + "grad_norm": 11.824106216430664, + "learning_rate": 9.779679313034182e-07, + "loss": 0.63, + "step": 1205 + }, + { + "epoch": 1.1870078740157481, + "grad_norm": 11.811593055725098, + "learning_rate": 9.777419216243333e-07, + "loss": 0.6917, + "step": 1206 + }, + { + "epoch": 1.1879921259842519, + "grad_norm": 23.336164474487305, + "learning_rate": 9.77514785101258e-07, + "loss": 1.427, + "step": 1207 + }, + { + "epoch": 1.188976377952756, + "grad_norm": 6.843215465545654, + "learning_rate": 9.77286522271812e-07, + "loss": 0.3583, + "step": 1208 + }, + { + "epoch": 1.1899606299212597, + "grad_norm": 9.87233829498291, + "learning_rate": 9.77057133676281e-07, + "loss": 0.383, + "step": 1209 + }, + { + "epoch": 1.1909448818897639, + "grad_norm": 15.967246055603027, + "learning_rate": 9.768266198576147e-07, + "loss": 0.9227, + "step": 1210 + }, + { + "epoch": 1.1919291338582676, + "grad_norm": 16.89198875427246, + "learning_rate": 9.765949813614273e-07, + "loss": 1.1553, + "step": 1211 + }, + { + "epoch": 1.1929133858267718, + "grad_norm": 12.008272171020508, + "learning_rate": 9.763622187359942e-07, + "loss": 0.9643, + "step": 1212 + }, + { + "epoch": 1.1938976377952755, + "grad_norm": 12.100934028625488, + "learning_rate": 9.761283325322515e-07, + "loss": 0.6882, + "step": 1213 + }, + { + "epoch": 1.1948818897637796, + "grad_norm": 9.986300468444824, + "learning_rate": 9.758933233037952e-07, + "loss": 0.7215, + "step": 1214 + }, + { + "epoch": 1.1958661417322836, + "grad_norm": 16.89799690246582, + "learning_rate": 9.756571916068793e-07, + "loss": 1.4159, + "step": 1215 + }, + { + "epoch": 1.1968503937007875, + "grad_norm": 11.459410667419434, + "learning_rate": 9.754199380004145e-07, + "loss": 0.7105, + "step": 1216 + }, + { + "epoch": 1.1978346456692914, + "grad_norm": 16.53590965270996, + "learning_rate": 9.75181563045967e-07, + "loss": 1.3628, + "step": 1217 + }, + { + "epoch": 1.1988188976377954, + "grad_norm": 12.679309844970703, + "learning_rate": 9.749420673077574e-07, + "loss": 1.0276, + "step": 1218 + }, + { + "epoch": 1.1998031496062993, + "grad_norm": 15.640876770019531, + "learning_rate": 9.747014513526588e-07, + "loss": 1.5007, + "step": 1219 + }, + { + "epoch": 1.2007874015748032, + "grad_norm": 14.455171585083008, + "learning_rate": 9.744597157501957e-07, + "loss": 0.8174, + "step": 1220 + }, + { + "epoch": 1.2017716535433072, + "grad_norm": 14.104186058044434, + "learning_rate": 9.742168610725437e-07, + "loss": 0.869, + "step": 1221 + }, + { + "epoch": 1.202755905511811, + "grad_norm": 10.396157264709473, + "learning_rate": 9.73972887894526e-07, + "loss": 0.6098, + "step": 1222 + }, + { + "epoch": 1.203740157480315, + "grad_norm": 9.877795219421387, + "learning_rate": 9.737277967936137e-07, + "loss": 0.6399, + "step": 1223 + }, + { + "epoch": 1.204724409448819, + "grad_norm": 9.88862133026123, + "learning_rate": 9.734815883499241e-07, + "loss": 0.4801, + "step": 1224 + }, + { + "epoch": 1.204724409448819, + "eval_Qnli-dev_cosine_accuracy": 0.650390625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.864111065864563, + "eval_Qnli-dev_cosine_ap": 0.6686175496002672, + "eval_Qnli-dev_cosine_f1": 0.6745762711864407, + "eval_Qnli-dev_cosine_f1_threshold": 0.7906274199485779, + "eval_Qnli-dev_cosine_precision": 0.5621468926553672, + "eval_Qnli-dev_cosine_recall": 0.8432203389830508, + "eval_Qnli-dev_dot_accuracy": 0.64453125, + "eval_Qnli-dev_dot_accuracy_threshold": 658.134765625, + "eval_Qnli-dev_dot_ap": 0.6671691099445898, + "eval_Qnli-dev_dot_f1": 0.674496644295302, + "eval_Qnli-dev_dot_f1_threshold": 598.698974609375, + "eval_Qnli-dev_dot_precision": 0.5583333333333333, + "eval_Qnli-dev_dot_recall": 0.8516949152542372, + "eval_Qnli-dev_euclidean_accuracy": 0.650390625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.402128219604492, + "eval_Qnli-dev_euclidean_ap": 0.668933387322661, + "eval_Qnli-dev_euclidean_f1": 0.6745762711864407, + "eval_Qnli-dev_euclidean_f1_threshold": 17.879480361938477, + "eval_Qnli-dev_euclidean_precision": 0.5621468926553672, + "eval_Qnli-dev_euclidean_recall": 0.8432203389830508, + "eval_Qnli-dev_manhattan_accuracy": 0.654296875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 296.47686767578125, + "eval_Qnli-dev_manhattan_ap": 0.6813781610580046, + "eval_Qnli-dev_manhattan_f1": 0.6813559322033899, + "eval_Qnli-dev_manhattan_f1_threshold": 374.47235107421875, + "eval_Qnli-dev_manhattan_precision": 0.5677966101694916, + "eval_Qnli-dev_manhattan_recall": 0.8516949152542372, + "eval_Qnli-dev_max_accuracy": 0.654296875, + "eval_Qnli-dev_max_accuracy_threshold": 658.134765625, + "eval_Qnli-dev_max_ap": 0.6813781610580046, + "eval_Qnli-dev_max_f1": 0.6813559322033899, + "eval_Qnli-dev_max_f1_threshold": 598.698974609375, + "eval_Qnli-dev_max_precision": 0.5677966101694916, + "eval_Qnli-dev_max_recall": 0.8516949152542372, + "eval_allNLI-dev_cosine_accuracy": 0.697265625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9384304881095886, + "eval_allNLI-dev_cosine_ap": 0.5009162286088922, + "eval_allNLI-dev_cosine_f1": 0.5342163355408388, + "eval_allNLI-dev_cosine_f1_threshold": 0.8694221377372742, + "eval_allNLI-dev_cosine_precision": 0.43214285714285716, + "eval_allNLI-dev_cosine_recall": 0.6994219653179191, + "eval_allNLI-dev_dot_accuracy": 0.701171875, + "eval_allNLI-dev_dot_accuracy_threshold": 716.254150390625, + "eval_allNLI-dev_dot_ap": 0.502121488648025, + "eval_allNLI-dev_dot_f1": 0.5357873210633947, + "eval_allNLI-dev_dot_f1_threshold": 653.184814453125, + "eval_allNLI-dev_dot_precision": 0.41455696202531644, + "eval_allNLI-dev_dot_recall": 0.7572254335260116, + "eval_allNLI-dev_euclidean_accuracy": 0.697265625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.702470779418945, + "eval_allNLI-dev_euclidean_ap": 0.5014088741949603, + "eval_allNLI-dev_euclidean_f1": 0.5336048879837068, + "eval_allNLI-dev_euclidean_f1_threshold": 14.84305191040039, + "eval_allNLI-dev_euclidean_precision": 0.4119496855345912, + "eval_allNLI-dev_euclidean_recall": 0.7572254335260116, + "eval_allNLI-dev_manhattan_accuracy": 0.693359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 204.68756103515625, + "eval_allNLI-dev_manhattan_ap": 0.5061285262379455, + "eval_allNLI-dev_manhattan_f1": 0.5371900826446281, + "eval_allNLI-dev_manhattan_f1_threshold": 306.1254577636719, + "eval_allNLI-dev_manhattan_precision": 0.4180064308681672, + "eval_allNLI-dev_manhattan_recall": 0.7514450867052023, + "eval_allNLI-dev_max_accuracy": 0.701171875, + "eval_allNLI-dev_max_accuracy_threshold": 716.254150390625, + "eval_allNLI-dev_max_ap": 0.5061285262379455, + "eval_allNLI-dev_max_f1": 0.5371900826446281, + "eval_allNLI-dev_max_f1_threshold": 653.184814453125, + "eval_allNLI-dev_max_precision": 0.43214285714285716, + "eval_allNLI-dev_max_recall": 0.7572254335260116, + "eval_loss": 0.8518065214157104, + "eval_runtime": 51.2973, + "eval_samples_per_second": 32.438, + "eval_sequential_score": 0.6813781610580046, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.5610266724163121, + "eval_sts-test_pearson_dot": 0.558918050371575, + "eval_sts-test_pearson_euclidean": 0.5834177940303891, + "eval_sts-test_pearson_manhattan": 0.593389043577112, + "eval_sts-test_pearson_max": 0.593389043577112, + "eval_sts-test_spearman_cosine": 0.5905739294186731, + "eval_sts-test_spearman_dot": 0.5888154377775213, + "eval_sts-test_spearman_euclidean": 0.5908653588080771, + "eval_sts-test_spearman_manhattan": 0.6004561358906522, + "eval_sts-test_spearman_max": 0.6004561358906522, + "step": 1224 + }, + { + "epoch": 1.205708661417323, + "grad_norm": 10.591872215270996, + "learning_rate": 9.73234263146219e-07, + "loss": 0.7651, + "step": 1225 + }, + { + "epoch": 1.2066929133858268, + "grad_norm": 10.7745361328125, + "learning_rate": 9.72985821767904e-07, + "loss": 0.4628, + "step": 1226 + }, + { + "epoch": 1.2076771653543308, + "grad_norm": 18.50623321533203, + "learning_rate": 9.727362648030254e-07, + "loss": 1.4387, + "step": 1227 + }, + { + "epoch": 1.2086614173228347, + "grad_norm": 12.07636547088623, + "learning_rate": 9.724855928422714e-07, + "loss": 0.7183, + "step": 1228 + }, + { + "epoch": 1.2096456692913387, + "grad_norm": 10.997964859008789, + "learning_rate": 9.722338064789688e-07, + "loss": 0.7154, + "step": 1229 + }, + { + "epoch": 1.2106299212598426, + "grad_norm": 11.254063606262207, + "learning_rate": 9.719809063090818e-07, + "loss": 0.5026, + "step": 1230 + }, + { + "epoch": 1.2116141732283465, + "grad_norm": 8.949058532714844, + "learning_rate": 9.717268929312112e-07, + "loss": 0.3789, + "step": 1231 + }, + { + "epoch": 1.2125984251968505, + "grad_norm": 10.556520462036133, + "learning_rate": 9.714717669465932e-07, + "loss": 0.5861, + "step": 1232 + }, + { + "epoch": 1.2135826771653544, + "grad_norm": 13.280324935913086, + "learning_rate": 9.712155289590964e-07, + "loss": 0.7763, + "step": 1233 + }, + { + "epoch": 1.2145669291338583, + "grad_norm": 9.581073760986328, + "learning_rate": 9.709581795752225e-07, + "loss": 0.4284, + "step": 1234 + }, + { + "epoch": 1.2155511811023623, + "grad_norm": 9.537132263183594, + "learning_rate": 9.70699719404103e-07, + "loss": 0.37, + "step": 1235 + }, + { + "epoch": 1.2165354330708662, + "grad_norm": 11.305273056030273, + "learning_rate": 9.704401490574993e-07, + "loss": 0.5274, + "step": 1236 + }, + { + "epoch": 1.2175196850393701, + "grad_norm": 13.992218017578125, + "learning_rate": 9.701794691497997e-07, + "loss": 0.8894, + "step": 1237 + }, + { + "epoch": 1.218503937007874, + "grad_norm": 12.191359519958496, + "learning_rate": 9.699176802980199e-07, + "loss": 0.5741, + "step": 1238 + }, + { + "epoch": 1.219488188976378, + "grad_norm": 19.28204345703125, + "learning_rate": 9.69654783121799e-07, + "loss": 1.1052, + "step": 1239 + }, + { + "epoch": 1.220472440944882, + "grad_norm": 13.538243293762207, + "learning_rate": 9.693907782434008e-07, + "loss": 1.0815, + "step": 1240 + }, + { + "epoch": 1.2214566929133859, + "grad_norm": 8.773369789123535, + "learning_rate": 9.691256662877099e-07, + "loss": 0.505, + "step": 1241 + }, + { + "epoch": 1.2224409448818898, + "grad_norm": 18.02227783203125, + "learning_rate": 9.688594478822317e-07, + "loss": 0.9539, + "step": 1242 + }, + { + "epoch": 1.2234251968503937, + "grad_norm": 13.074368476867676, + "learning_rate": 9.685921236570914e-07, + "loss": 0.8867, + "step": 1243 + }, + { + "epoch": 1.2244094488188977, + "grad_norm": 9.300528526306152, + "learning_rate": 9.683236942450294e-07, + "loss": 0.4681, + "step": 1244 + }, + { + "epoch": 1.2253937007874016, + "grad_norm": 14.433270454406738, + "learning_rate": 9.680541602814047e-07, + "loss": 1.1859, + "step": 1245 + }, + { + "epoch": 1.2263779527559056, + "grad_norm": 10.780652046203613, + "learning_rate": 9.677835224041885e-07, + "loss": 0.4907, + "step": 1246 + }, + { + "epoch": 1.2273622047244095, + "grad_norm": 11.958067893981934, + "learning_rate": 9.67511781253966e-07, + "loss": 0.6341, + "step": 1247 + }, + { + "epoch": 1.2283464566929134, + "grad_norm": 11.88088321685791, + "learning_rate": 9.67238937473934e-07, + "loss": 0.8029, + "step": 1248 + }, + { + "epoch": 1.2293307086614174, + "grad_norm": 10.48951244354248, + "learning_rate": 9.669649917098983e-07, + "loss": 0.5358, + "step": 1249 + }, + { + "epoch": 1.2303149606299213, + "grad_norm": 10.449585914611816, + "learning_rate": 9.666899446102736e-07, + "loss": 0.5571, + "step": 1250 + }, + { + "epoch": 1.2312992125984252, + "grad_norm": 12.562650680541992, + "learning_rate": 9.664137968260814e-07, + "loss": 0.5891, + "step": 1251 + }, + { + "epoch": 1.2322834645669292, + "grad_norm": 14.111005783081055, + "learning_rate": 9.661365490109482e-07, + "loss": 0.7619, + "step": 1252 + }, + { + "epoch": 1.233267716535433, + "grad_norm": 9.546327590942383, + "learning_rate": 9.658582018211045e-07, + "loss": 0.2854, + "step": 1253 + }, + { + "epoch": 1.234251968503937, + "grad_norm": 14.747997283935547, + "learning_rate": 9.655787559153824e-07, + "loss": 0.6662, + "step": 1254 + }, + { + "epoch": 1.235236220472441, + "grad_norm": 17.832862854003906, + "learning_rate": 9.652982119552154e-07, + "loss": 1.4318, + "step": 1255 + }, + { + "epoch": 1.236220472440945, + "grad_norm": 25.594974517822266, + "learning_rate": 9.650165706046359e-07, + "loss": 1.4687, + "step": 1256 + }, + { + "epoch": 1.2372047244094488, + "grad_norm": 12.508796691894531, + "learning_rate": 9.64733832530273e-07, + "loss": 0.6946, + "step": 1257 + }, + { + "epoch": 1.2381889763779528, + "grad_norm": 12.623579025268555, + "learning_rate": 9.644499984013525e-07, + "loss": 0.633, + "step": 1258 + }, + { + "epoch": 1.2391732283464567, + "grad_norm": 20.903850555419922, + "learning_rate": 9.64165068889694e-07, + "loss": 1.5568, + "step": 1259 + }, + { + "epoch": 1.2401574803149606, + "grad_norm": 10.497096061706543, + "learning_rate": 9.6387904466971e-07, + "loss": 0.6798, + "step": 1260 + }, + { + "epoch": 1.2411417322834646, + "grad_norm": 26.627531051635742, + "learning_rate": 9.635919264184043e-07, + "loss": 1.267, + "step": 1261 + }, + { + "epoch": 1.2421259842519685, + "grad_norm": 11.09723949432373, + "learning_rate": 9.6330371481537e-07, + "loss": 0.6641, + "step": 1262 + }, + { + "epoch": 1.2431102362204725, + "grad_norm": 12.233721733093262, + "learning_rate": 9.630144105427883e-07, + "loss": 0.8602, + "step": 1263 + }, + { + "epoch": 1.2440944881889764, + "grad_norm": 9.321574211120605, + "learning_rate": 9.627240142854262e-07, + "loss": 0.4017, + "step": 1264 + }, + { + "epoch": 1.2450787401574803, + "grad_norm": 9.705317497253418, + "learning_rate": 9.624325267306357e-07, + "loss": 0.4382, + "step": 1265 + }, + { + "epoch": 1.2460629921259843, + "grad_norm": 18.52342414855957, + "learning_rate": 9.621399485683523e-07, + "loss": 1.2464, + "step": 1266 + }, + { + "epoch": 1.2470472440944882, + "grad_norm": 10.046468734741211, + "learning_rate": 9.618462804910917e-07, + "loss": 0.8608, + "step": 1267 + }, + { + "epoch": 1.2480314960629921, + "grad_norm": 15.30111312866211, + "learning_rate": 9.61551523193951e-07, + "loss": 1.1932, + "step": 1268 + }, + { + "epoch": 1.249015748031496, + "grad_norm": 11.027518272399902, + "learning_rate": 9.612556773746038e-07, + "loss": 0.8773, + "step": 1269 + }, + { + "epoch": 1.25, + "grad_norm": 11.729086875915527, + "learning_rate": 9.609587437333011e-07, + "loss": 0.5899, + "step": 1270 + }, + { + "epoch": 1.250984251968504, + "grad_norm": 8.820747375488281, + "learning_rate": 9.606607229728688e-07, + "loss": 0.4145, + "step": 1271 + }, + { + "epoch": 1.2519685039370079, + "grad_norm": 13.87430477142334, + "learning_rate": 9.603616157987054e-07, + "loss": 0.9711, + "step": 1272 + }, + { + "epoch": 1.2529527559055118, + "grad_norm": 18.675867080688477, + "learning_rate": 9.600614229187812e-07, + "loss": 1.0946, + "step": 1273 + }, + { + "epoch": 1.2539370078740157, + "grad_norm": 15.559755325317383, + "learning_rate": 9.597601450436364e-07, + "loss": 1.2652, + "step": 1274 + }, + { + "epoch": 1.2549212598425197, + "grad_norm": 9.610017776489258, + "learning_rate": 9.594577828863794e-07, + "loss": 0.4312, + "step": 1275 + }, + { + "epoch": 1.2559055118110236, + "grad_norm": 11.332744598388672, + "learning_rate": 9.591543371626842e-07, + "loss": 0.9017, + "step": 1276 + }, + { + "epoch": 1.2568897637795275, + "grad_norm": 8.62760066986084, + "learning_rate": 9.588498085907905e-07, + "loss": 0.5428, + "step": 1277 + }, + { + "epoch": 1.2578740157480315, + "grad_norm": 11.455253601074219, + "learning_rate": 9.585441978915008e-07, + "loss": 0.6475, + "step": 1278 + }, + { + "epoch": 1.2588582677165354, + "grad_norm": 11.097201347351074, + "learning_rate": 9.582375057881785e-07, + "loss": 0.9587, + "step": 1279 + }, + { + "epoch": 1.2598425196850394, + "grad_norm": 15.259642601013184, + "learning_rate": 9.579297330067476e-07, + "loss": 0.9115, + "step": 1280 + }, + { + "epoch": 1.2608267716535433, + "grad_norm": 13.242481231689453, + "learning_rate": 9.576208802756888e-07, + "loss": 0.9348, + "step": 1281 + }, + { + "epoch": 1.2618110236220472, + "grad_norm": 12.079668998718262, + "learning_rate": 9.5731094832604e-07, + "loss": 0.8717, + "step": 1282 + }, + { + "epoch": 1.2627952755905512, + "grad_norm": 8.83189582824707, + "learning_rate": 9.569999378913925e-07, + "loss": 0.4587, + "step": 1283 + }, + { + "epoch": 1.263779527559055, + "grad_norm": 10.650629997253418, + "learning_rate": 9.566878497078914e-07, + "loss": 0.5246, + "step": 1284 + }, + { + "epoch": 1.264763779527559, + "grad_norm": 10.929993629455566, + "learning_rate": 9.563746845142326e-07, + "loss": 0.6744, + "step": 1285 + }, + { + "epoch": 1.265748031496063, + "grad_norm": 18.163965225219727, + "learning_rate": 9.5606044305166e-07, + "loss": 1.3601, + "step": 1286 + }, + { + "epoch": 1.266732283464567, + "grad_norm": 10.698195457458496, + "learning_rate": 9.557451260639665e-07, + "loss": 0.6256, + "step": 1287 + }, + { + "epoch": 1.2677165354330708, + "grad_norm": 15.834595680236816, + "learning_rate": 9.5542873429749e-07, + "loss": 0.7788, + "step": 1288 + }, + { + "epoch": 1.2687007874015748, + "grad_norm": 9.984282493591309, + "learning_rate": 9.551112685011127e-07, + "loss": 0.6486, + "step": 1289 + }, + { + "epoch": 1.2696850393700787, + "grad_norm": 16.016998291015625, + "learning_rate": 9.547927294262583e-07, + "loss": 1.1881, + "step": 1290 + }, + { + "epoch": 1.2706692913385826, + "grad_norm": 10.568839073181152, + "learning_rate": 9.544731178268913e-07, + "loss": 0.6957, + "step": 1291 + }, + { + "epoch": 1.2716535433070866, + "grad_norm": 13.368728637695312, + "learning_rate": 9.54152434459515e-07, + "loss": 1.0403, + "step": 1292 + }, + { + "epoch": 1.2726377952755905, + "grad_norm": 10.826345443725586, + "learning_rate": 9.538306800831695e-07, + "loss": 0.5265, + "step": 1293 + }, + { + "epoch": 1.2736220472440944, + "grad_norm": 9.606311798095703, + "learning_rate": 9.535078554594293e-07, + "loss": 0.4922, + "step": 1294 + }, + { + "epoch": 1.2746062992125984, + "grad_norm": 13.779792785644531, + "learning_rate": 9.531839613524029e-07, + "loss": 0.9464, + "step": 1295 + }, + { + "epoch": 1.2755905511811023, + "grad_norm": 13.541178703308105, + "learning_rate": 9.528589985287297e-07, + "loss": 0.7804, + "step": 1296 + }, + { + "epoch": 1.2765748031496063, + "grad_norm": 10.518589973449707, + "learning_rate": 9.525329677575787e-07, + "loss": 0.6297, + "step": 1297 + }, + { + "epoch": 1.2775590551181102, + "grad_norm": 10.690191268920898, + "learning_rate": 9.522058698106475e-07, + "loss": 0.5289, + "step": 1298 + }, + { + "epoch": 1.2785433070866141, + "grad_norm": 10.678605079650879, + "learning_rate": 9.518777054621582e-07, + "loss": 0.6719, + "step": 1299 + }, + { + "epoch": 1.279527559055118, + "grad_norm": 13.158123970031738, + "learning_rate": 9.515484754888582e-07, + "loss": 0.7858, + "step": 1300 + }, + { + "epoch": 1.280511811023622, + "grad_norm": 6.52417516708374, + "learning_rate": 9.512181806700167e-07, + "loss": 0.3309, + "step": 1301 + }, + { + "epoch": 1.281496062992126, + "grad_norm": 10.241775512695312, + "learning_rate": 9.508868217874233e-07, + "loss": 0.4839, + "step": 1302 + }, + { + "epoch": 1.2824803149606299, + "grad_norm": 9.741416931152344, + "learning_rate": 9.505543996253865e-07, + "loss": 0.4494, + "step": 1303 + }, + { + "epoch": 1.2834645669291338, + "grad_norm": 12.683394432067871, + "learning_rate": 9.502209149707312e-07, + "loss": 0.5807, + "step": 1304 + }, + { + "epoch": 1.2844488188976377, + "grad_norm": 12.149230003356934, + "learning_rate": 9.498863686127971e-07, + "loss": 0.6408, + "step": 1305 + }, + { + "epoch": 1.2854330708661417, + "grad_norm": 11.451055526733398, + "learning_rate": 9.495507613434374e-07, + "loss": 0.6289, + "step": 1306 + }, + { + "epoch": 1.2864173228346456, + "grad_norm": 19.40523338317871, + "learning_rate": 9.492140939570159e-07, + "loss": 0.7991, + "step": 1307 + }, + { + "epoch": 1.2874015748031495, + "grad_norm": 10.470606803894043, + "learning_rate": 9.488763672504058e-07, + "loss": 0.651, + "step": 1308 + }, + { + "epoch": 1.2883858267716535, + "grad_norm": 11.133830070495605, + "learning_rate": 9.485375820229876e-07, + "loss": 0.6239, + "step": 1309 + }, + { + "epoch": 1.2893700787401574, + "grad_norm": 18.81745147705078, + "learning_rate": 9.481977390766475e-07, + "loss": 1.3505, + "step": 1310 + }, + { + "epoch": 1.2903543307086613, + "grad_norm": 5.142672061920166, + "learning_rate": 9.47856839215775e-07, + "loss": 0.1423, + "step": 1311 + }, + { + "epoch": 1.2913385826771653, + "grad_norm": 14.626995086669922, + "learning_rate": 9.475148832472613e-07, + "loss": 1.109, + "step": 1312 + }, + { + "epoch": 1.2923228346456692, + "grad_norm": 20.07413673400879, + "learning_rate": 9.471718719804975e-07, + "loss": 1.3417, + "step": 1313 + }, + { + "epoch": 1.2933070866141732, + "grad_norm": 7.412285327911377, + "learning_rate": 9.468278062273726e-07, + "loss": 0.4346, + "step": 1314 + }, + { + "epoch": 1.294291338582677, + "grad_norm": 22.638029098510742, + "learning_rate": 9.464826868022709e-07, + "loss": 1.4748, + "step": 1315 + }, + { + "epoch": 1.295275590551181, + "grad_norm": 16.264259338378906, + "learning_rate": 9.461365145220715e-07, + "loss": 1.2036, + "step": 1316 + }, + { + "epoch": 1.296259842519685, + "grad_norm": 10.352934837341309, + "learning_rate": 9.457892902061452e-07, + "loss": 0.6295, + "step": 1317 + }, + { + "epoch": 1.297244094488189, + "grad_norm": 12.874444007873535, + "learning_rate": 9.454410146763525e-07, + "loss": 0.7632, + "step": 1318 + }, + { + "epoch": 1.298228346456693, + "grad_norm": 8.567391395568848, + "learning_rate": 9.450916887570429e-07, + "loss": 0.4074, + "step": 1319 + }, + { + "epoch": 1.2992125984251968, + "grad_norm": 5.803460597991943, + "learning_rate": 9.447413132750515e-07, + "loss": 0.2692, + "step": 1320 + }, + { + "epoch": 1.300196850393701, + "grad_norm": 11.235288619995117, + "learning_rate": 9.443898890596977e-07, + "loss": 0.7128, + "step": 1321 + }, + { + "epoch": 1.3011811023622046, + "grad_norm": 8.22232723236084, + "learning_rate": 9.440374169427834e-07, + "loss": 0.2916, + "step": 1322 + }, + { + "epoch": 1.3021653543307088, + "grad_norm": 9.740578651428223, + "learning_rate": 9.43683897758591e-07, + "loss": 0.5082, + "step": 1323 + }, + { + "epoch": 1.3031496062992125, + "grad_norm": 8.524904251098633, + "learning_rate": 9.433293323438807e-07, + "loss": 0.3651, + "step": 1324 + }, + { + "epoch": 1.3041338582677167, + "grad_norm": 11.907980918884277, + "learning_rate": 9.429737215378894e-07, + "loss": 0.7582, + "step": 1325 + }, + { + "epoch": 1.3051181102362204, + "grad_norm": 10.966760635375977, + "learning_rate": 9.426170661823285e-07, + "loss": 0.6235, + "step": 1326 + }, + { + "epoch": 1.3061023622047245, + "grad_norm": 11.651907920837402, + "learning_rate": 9.422593671213816e-07, + "loss": 0.7423, + "step": 1327 + }, + { + "epoch": 1.3070866141732282, + "grad_norm": 15.069643020629883, + "learning_rate": 9.419006252017027e-07, + "loss": 1.1407, + "step": 1328 + }, + { + "epoch": 1.3080708661417324, + "grad_norm": 8.573887825012207, + "learning_rate": 9.415408412724141e-07, + "loss": 0.4683, + "step": 1329 + }, + { + "epoch": 1.3090551181102361, + "grad_norm": 14.449766159057617, + "learning_rate": 9.411800161851048e-07, + "loss": 0.9656, + "step": 1330 + }, + { + "epoch": 1.3100393700787403, + "grad_norm": 12.743842124938965, + "learning_rate": 9.408181507938279e-07, + "loss": 0.5829, + "step": 1331 + }, + { + "epoch": 1.311023622047244, + "grad_norm": 9.47890853881836, + "learning_rate": 9.404552459550988e-07, + "loss": 0.4886, + "step": 1332 + }, + { + "epoch": 1.3120078740157481, + "grad_norm": 12.849534034729004, + "learning_rate": 9.400913025278937e-07, + "loss": 0.6078, + "step": 1333 + }, + { + "epoch": 1.3129921259842519, + "grad_norm": 11.463114738464355, + "learning_rate": 9.397263213736464e-07, + "loss": 0.5075, + "step": 1334 + }, + { + "epoch": 1.313976377952756, + "grad_norm": 10.758073806762695, + "learning_rate": 9.393603033562471e-07, + "loss": 0.4273, + "step": 1335 + }, + { + "epoch": 1.3149606299212597, + "grad_norm": 15.226777076721191, + "learning_rate": 9.389932493420407e-07, + "loss": 1.0254, + "step": 1336 + }, + { + "epoch": 1.3159448818897639, + "grad_norm": 14.52092456817627, + "learning_rate": 9.386251601998239e-07, + "loss": 0.895, + "step": 1337 + }, + { + "epoch": 1.3169291338582676, + "grad_norm": 12.966414451599121, + "learning_rate": 9.38256036800843e-07, + "loss": 0.6699, + "step": 1338 + }, + { + "epoch": 1.3179133858267718, + "grad_norm": 11.159317016601562, + "learning_rate": 9.378858800187935e-07, + "loss": 0.5603, + "step": 1339 + }, + { + "epoch": 1.3188976377952755, + "grad_norm": 11.317286491394043, + "learning_rate": 9.375146907298157e-07, + "loss": 0.6651, + "step": 1340 + }, + { + "epoch": 1.3198818897637796, + "grad_norm": 8.095876693725586, + "learning_rate": 9.371424698124945e-07, + "loss": 0.2131, + "step": 1341 + }, + { + "epoch": 1.3208661417322833, + "grad_norm": 11.953645706176758, + "learning_rate": 9.367692181478564e-07, + "loss": 0.4912, + "step": 1342 + }, + { + "epoch": 1.3218503937007875, + "grad_norm": 14.653837203979492, + "learning_rate": 9.363949366193673e-07, + "loss": 0.6365, + "step": 1343 + }, + { + "epoch": 1.3228346456692912, + "grad_norm": 8.630864143371582, + "learning_rate": 9.360196261129313e-07, + "loss": 0.4183, + "step": 1344 + }, + { + "epoch": 1.3238188976377954, + "grad_norm": 12.628782272338867, + "learning_rate": 9.356432875168876e-07, + "loss": 0.555, + "step": 1345 + }, + { + "epoch": 1.324803149606299, + "grad_norm": 10.981666564941406, + "learning_rate": 9.352659217220091e-07, + "loss": 0.7718, + "step": 1346 + }, + { + "epoch": 1.3257874015748032, + "grad_norm": 24.697113037109375, + "learning_rate": 9.348875296215001e-07, + "loss": 1.2605, + "step": 1347 + }, + { + "epoch": 1.326771653543307, + "grad_norm": 11.701473236083984, + "learning_rate": 9.345081121109935e-07, + "loss": 0.4897, + "step": 1348 + }, + { + "epoch": 1.327755905511811, + "grad_norm": 10.443194389343262, + "learning_rate": 9.341276700885501e-07, + "loss": 0.5364, + "step": 1349 + }, + { + "epoch": 1.328740157480315, + "grad_norm": 7.822117328643799, + "learning_rate": 9.337462044546548e-07, + "loss": 0.2149, + "step": 1350 + }, + { + "epoch": 1.329724409448819, + "grad_norm": 19.030441284179688, + "learning_rate": 9.333637161122166e-07, + "loss": 0.8194, + "step": 1351 + }, + { + "epoch": 1.330708661417323, + "grad_norm": 11.934989929199219, + "learning_rate": 9.329802059665637e-07, + "loss": 0.4944, + "step": 1352 + }, + { + "epoch": 1.3316929133858268, + "grad_norm": 14.144577980041504, + "learning_rate": 9.325956749254437e-07, + "loss": 0.9056, + "step": 1353 + }, + { + "epoch": 1.3326771653543308, + "grad_norm": 15.389364242553711, + "learning_rate": 9.322101238990206e-07, + "loss": 1.0409, + "step": 1354 + }, + { + "epoch": 1.3336614173228347, + "grad_norm": 21.147554397583008, + "learning_rate": 9.318235537998726e-07, + "loss": 1.379, + "step": 1355 + }, + { + "epoch": 1.3346456692913387, + "grad_norm": 8.598258972167969, + "learning_rate": 9.314359655429897e-07, + "loss": 0.3896, + "step": 1356 + }, + { + "epoch": 1.3356299212598426, + "grad_norm": 11.006596565246582, + "learning_rate": 9.310473600457721e-07, + "loss": 0.6206, + "step": 1357 + }, + { + "epoch": 1.3366141732283465, + "grad_norm": 8.743659973144531, + "learning_rate": 9.306577382280277e-07, + "loss": 0.5468, + "step": 1358 + }, + { + "epoch": 1.3375984251968505, + "grad_norm": 11.284257888793945, + "learning_rate": 9.3026710101197e-07, + "loss": 0.4268, + "step": 1359 + }, + { + "epoch": 1.3385826771653544, + "grad_norm": 8.214118003845215, + "learning_rate": 9.298754493222155e-07, + "loss": 0.4647, + "step": 1360 + }, + { + "epoch": 1.3395669291338583, + "grad_norm": 9.324970245361328, + "learning_rate": 9.294827840857827e-07, + "loss": 0.5162, + "step": 1361 + }, + { + "epoch": 1.3405511811023623, + "grad_norm": 10.994433403015137, + "learning_rate": 9.290891062320884e-07, + "loss": 0.6775, + "step": 1362 + }, + { + "epoch": 1.3415354330708662, + "grad_norm": 11.114113807678223, + "learning_rate": 9.286944166929466e-07, + "loss": 0.5889, + "step": 1363 + }, + { + "epoch": 1.3425196850393701, + "grad_norm": 13.406707763671875, + "learning_rate": 9.282987164025658e-07, + "loss": 0.8913, + "step": 1364 + }, + { + "epoch": 1.343503937007874, + "grad_norm": 7.464841842651367, + "learning_rate": 9.279020062975465e-07, + "loss": 0.4086, + "step": 1365 + }, + { + "epoch": 1.344488188976378, + "grad_norm": 9.809942245483398, + "learning_rate": 9.275042873168802e-07, + "loss": 0.5859, + "step": 1366 + }, + { + "epoch": 1.345472440944882, + "grad_norm": 11.583236694335938, + "learning_rate": 9.271055604019458e-07, + "loss": 0.531, + "step": 1367 + }, + { + "epoch": 1.3464566929133859, + "grad_norm": 9.490612030029297, + "learning_rate": 9.267058264965076e-07, + "loss": 0.4143, + "step": 1368 + }, + { + "epoch": 1.3474409448818898, + "grad_norm": 10.371625900268555, + "learning_rate": 9.263050865467143e-07, + "loss": 0.493, + "step": 1369 + }, + { + "epoch": 1.3484251968503937, + "grad_norm": 10.511662483215332, + "learning_rate": 9.259033415010952e-07, + "loss": 0.5863, + "step": 1370 + }, + { + "epoch": 1.3494094488188977, + "grad_norm": 11.5587739944458, + "learning_rate": 9.255005923105587e-07, + "loss": 0.7581, + "step": 1371 + }, + { + "epoch": 1.3503937007874016, + "grad_norm": 24.195449829101562, + "learning_rate": 9.250968399283902e-07, + "loss": 0.9708, + "step": 1372 + }, + { + "epoch": 1.3513779527559056, + "grad_norm": 7.567129135131836, + "learning_rate": 9.24692085310249e-07, + "loss": 0.2478, + "step": 1373 + }, + { + "epoch": 1.3523622047244095, + "grad_norm": 21.72224998474121, + "learning_rate": 9.242863294141675e-07, + "loss": 1.4765, + "step": 1374 + }, + { + "epoch": 1.3533464566929134, + "grad_norm": 12.258607864379883, + "learning_rate": 9.238795732005475e-07, + "loss": 0.6349, + "step": 1375 + }, + { + "epoch": 1.3543307086614174, + "grad_norm": 12.518092155456543, + "learning_rate": 9.234718176321586e-07, + "loss": 0.8818, + "step": 1376 + }, + { + "epoch": 1.3553149606299213, + "grad_norm": 11.761449813842773, + "learning_rate": 9.230630636741358e-07, + "loss": 0.3951, + "step": 1377 + }, + { + "epoch": 1.3553149606299213, + "eval_Qnli-dev_cosine_accuracy": 0.658203125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8371257781982422, + "eval_Qnli-dev_cosine_ap": 0.6811765025539783, + "eval_Qnli-dev_cosine_f1": 0.679245283018868, + "eval_Qnli-dev_cosine_f1_threshold": 0.760552167892456, + "eval_Qnli-dev_cosine_precision": 0.5706051873198847, + "eval_Qnli-dev_cosine_recall": 0.8389830508474576, + "eval_Qnli-dev_dot_accuracy": 0.650390625, + "eval_Qnli-dev_dot_accuracy_threshold": 649.9522705078125, + "eval_Qnli-dev_dot_ap": 0.678890264150892, + "eval_Qnli-dev_dot_f1": 0.6802721088435374, + "eval_Qnli-dev_dot_f1_threshold": 579.0703125, + "eval_Qnli-dev_dot_precision": 0.5681818181818182, + "eval_Qnli-dev_dot_recall": 0.847457627118644, + "eval_Qnli-dev_euclidean_accuracy": 0.658203125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.773345947265625, + "eval_Qnli-dev_euclidean_ap": 0.6814925241755564, + "eval_Qnli-dev_euclidean_f1": 0.6782608695652175, + "eval_Qnli-dev_euclidean_f1_threshold": 18.98737335205078, + "eval_Qnli-dev_euclidean_precision": 0.5752212389380531, + "eval_Qnli-dev_euclidean_recall": 0.826271186440678, + "eval_Qnli-dev_manhattan_accuracy": 0.666015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 324.795166015625, + "eval_Qnli-dev_manhattan_ap": 0.6892945079582472, + "eval_Qnli-dev_manhattan_f1": 0.6770833333333334, + "eval_Qnli-dev_manhattan_f1_threshold": 396.5395202636719, + "eval_Qnli-dev_manhattan_precision": 0.5735294117647058, + "eval_Qnli-dev_manhattan_recall": 0.826271186440678, + "eval_Qnli-dev_max_accuracy": 0.666015625, + "eval_Qnli-dev_max_accuracy_threshold": 649.9522705078125, + "eval_Qnli-dev_max_ap": 0.6892945079582472, + "eval_Qnli-dev_max_f1": 0.6802721088435374, + "eval_Qnli-dev_max_f1_threshold": 579.0703125, + "eval_Qnli-dev_max_precision": 0.5752212389380531, + "eval_Qnli-dev_max_recall": 0.847457627118644, + "eval_allNLI-dev_cosine_accuracy": 0.697265625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9207851886749268, + "eval_allNLI-dev_cosine_ap": 0.5175016908121097, + "eval_allNLI-dev_cosine_f1": 0.5428571428571429, + "eval_allNLI-dev_cosine_f1_threshold": 0.8632231950759888, + "eval_allNLI-dev_cosine_precision": 0.46153846153846156, + "eval_allNLI-dev_cosine_recall": 0.6589595375722543, + "eval_allNLI-dev_dot_accuracy": 0.697265625, + "eval_allNLI-dev_dot_accuracy_threshold": 708.7349853515625, + "eval_allNLI-dev_dot_ap": 0.5183962490789402, + "eval_allNLI-dev_dot_f1": 0.5395348837209303, + "eval_allNLI-dev_dot_f1_threshold": 658.3466796875, + "eval_allNLI-dev_dot_precision": 0.45136186770428016, + "eval_allNLI-dev_dot_recall": 0.6705202312138728, + "eval_allNLI-dev_euclidean_accuracy": 0.697265625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.946168899536133, + "eval_allNLI-dev_euclidean_ap": 0.5172653377212496, + "eval_allNLI-dev_euclidean_f1": 0.5415676959619952, + "eval_allNLI-dev_euclidean_f1_threshold": 14.458630561828613, + "eval_allNLI-dev_euclidean_precision": 0.4596774193548387, + "eval_allNLI-dev_euclidean_recall": 0.6589595375722543, + "eval_allNLI-dev_manhattan_accuracy": 0.69921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 227.57864379882812, + "eval_allNLI-dev_manhattan_ap": 0.5217577261199255, + "eval_allNLI-dev_manhattan_f1": 0.5491990846681922, + "eval_allNLI-dev_manhattan_f1_threshold": 308.6780090332031, + "eval_allNLI-dev_manhattan_precision": 0.45454545454545453, + "eval_allNLI-dev_manhattan_recall": 0.6936416184971098, + "eval_allNLI-dev_max_accuracy": 0.69921875, + "eval_allNLI-dev_max_accuracy_threshold": 708.7349853515625, + "eval_allNLI-dev_max_ap": 0.5217577261199255, + "eval_allNLI-dev_max_f1": 0.5491990846681922, + "eval_allNLI-dev_max_f1_threshold": 658.3466796875, + "eval_allNLI-dev_max_precision": 0.46153846153846156, + "eval_allNLI-dev_max_recall": 0.6936416184971098, + "eval_loss": 0.7521839141845703, + "eval_runtime": 51.2618, + "eval_samples_per_second": 32.461, + "eval_sequential_score": 0.6892945079582472, + "eval_steps_per_second": 0.137, + "eval_sts-test_pearson_cosine": 0.6002417990705935, + "eval_sts-test_pearson_dot": 0.5987837342557453, + "eval_sts-test_pearson_euclidean": 0.6155642117709486, + "eval_sts-test_pearson_manhattan": 0.6212566252771287, + "eval_sts-test_pearson_max": 0.6212566252771287, + "eval_sts-test_spearman_cosine": 0.6147386799687629, + "eval_sts-test_spearman_dot": 0.6141685488986365, + "eval_sts-test_spearman_euclidean": 0.6149686119497918, + "eval_sts-test_spearman_manhattan": 0.6215729583546733, + "eval_sts-test_spearman_max": 0.6215729583546733, + "step": 1377 + }, + { + "epoch": 1.3562992125984252, + "grad_norm": 11.374031066894531, + "learning_rate": 9.226533122939772e-07, + "loss": 0.5389, + "step": 1378 + }, + { + "epoch": 1.3572834645669292, + "grad_norm": 18.062849044799805, + "learning_rate": 9.222425644615421e-07, + "loss": 0.8855, + "step": 1379 + }, + { + "epoch": 1.358267716535433, + "grad_norm": 25.814117431640625, + "learning_rate": 9.218308211490479e-07, + "loss": 1.4853, + "step": 1380 + }, + { + "epoch": 1.359251968503937, + "grad_norm": 10.830196380615234, + "learning_rate": 9.214180833310684e-07, + "loss": 0.4334, + "step": 1381 + }, + { + "epoch": 1.360236220472441, + "grad_norm": 16.53986167907715, + "learning_rate": 9.210043519845313e-07, + "loss": 1.185, + "step": 1382 + }, + { + "epoch": 1.361220472440945, + "grad_norm": 14.258798599243164, + "learning_rate": 9.205896280887163e-07, + "loss": 0.9326, + "step": 1383 + }, + { + "epoch": 1.3622047244094488, + "grad_norm": 10.722103118896484, + "learning_rate": 9.201739126252518e-07, + "loss": 0.599, + "step": 1384 + }, + { + "epoch": 1.3631889763779528, + "grad_norm": 21.030916213989258, + "learning_rate": 9.197572065781133e-07, + "loss": 1.0678, + "step": 1385 + }, + { + "epoch": 1.3641732283464567, + "grad_norm": 13.463126182556152, + "learning_rate": 9.193395109336216e-07, + "loss": 0.8085, + "step": 1386 + }, + { + "epoch": 1.3651574803149606, + "grad_norm": 16.344621658325195, + "learning_rate": 9.189208266804389e-07, + "loss": 0.8788, + "step": 1387 + }, + { + "epoch": 1.3661417322834646, + "grad_norm": 11.256791114807129, + "learning_rate": 9.185011548095682e-07, + "loss": 0.67, + "step": 1388 + }, + { + "epoch": 1.3671259842519685, + "grad_norm": 13.486610412597656, + "learning_rate": 9.180804963143496e-07, + "loss": 0.8087, + "step": 1389 + }, + { + "epoch": 1.3681102362204725, + "grad_norm": 8.284912109375, + "learning_rate": 9.176588521904588e-07, + "loss": 0.3291, + "step": 1390 + }, + { + "epoch": 1.3690944881889764, + "grad_norm": 10.581992149353027, + "learning_rate": 9.172362234359041e-07, + "loss": 0.66, + "step": 1391 + }, + { + "epoch": 1.3700787401574803, + "grad_norm": 8.81778621673584, + "learning_rate": 9.168126110510246e-07, + "loss": 0.4315, + "step": 1392 + }, + { + "epoch": 1.3710629921259843, + "grad_norm": 8.121106147766113, + "learning_rate": 9.163880160384877e-07, + "loss": 0.5049, + "step": 1393 + }, + { + "epoch": 1.3720472440944882, + "grad_norm": 7.680241107940674, + "learning_rate": 9.159624394032864e-07, + "loss": 0.3882, + "step": 1394 + }, + { + "epoch": 1.3730314960629921, + "grad_norm": 13.15041732788086, + "learning_rate": 9.155358821527371e-07, + "loss": 1.0489, + "step": 1395 + }, + { + "epoch": 1.374015748031496, + "grad_norm": 8.546382904052734, + "learning_rate": 9.151083452964775e-07, + "loss": 0.48, + "step": 1396 + }, + { + "epoch": 1.375, + "grad_norm": 17.865381240844727, + "learning_rate": 9.146798298464642e-07, + "loss": 1.2194, + "step": 1397 + }, + { + "epoch": 1.375984251968504, + "grad_norm": 9.983588218688965, + "learning_rate": 9.14250336816969e-07, + "loss": 0.7008, + "step": 1398 + }, + { + "epoch": 1.3769685039370079, + "grad_norm": 12.46605110168457, + "learning_rate": 9.13819867224579e-07, + "loss": 0.894, + "step": 1399 + }, + { + "epoch": 1.3779527559055118, + "grad_norm": 16.217758178710938, + "learning_rate": 9.133884220881917e-07, + "loss": 1.24, + "step": 1400 + }, + { + "epoch": 1.3789370078740157, + "grad_norm": 11.264232635498047, + "learning_rate": 9.129560024290145e-07, + "loss": 0.7041, + "step": 1401 + }, + { + "epoch": 1.3799212598425197, + "grad_norm": 8.890203475952148, + "learning_rate": 9.125226092705603e-07, + "loss": 0.3816, + "step": 1402 + }, + { + "epoch": 1.3809055118110236, + "grad_norm": 10.996663093566895, + "learning_rate": 9.120882436386474e-07, + "loss": 0.4578, + "step": 1403 + }, + { + "epoch": 1.3818897637795275, + "grad_norm": 10.450602531433105, + "learning_rate": 9.116529065613954e-07, + "loss": 0.4671, + "step": 1404 + }, + { + "epoch": 1.3828740157480315, + "grad_norm": 12.85247802734375, + "learning_rate": 9.112165990692231e-07, + "loss": 1.05, + "step": 1405 + }, + { + "epoch": 1.3838582677165354, + "grad_norm": 30.172073364257812, + "learning_rate": 9.107793221948464e-07, + "loss": 1.8844, + "step": 1406 + }, + { + "epoch": 1.3848425196850394, + "grad_norm": 12.470993995666504, + "learning_rate": 9.103410769732755e-07, + "loss": 0.8993, + "step": 1407 + }, + { + "epoch": 1.3858267716535433, + "grad_norm": 19.326557159423828, + "learning_rate": 9.099018644418133e-07, + "loss": 0.8769, + "step": 1408 + }, + { + "epoch": 1.3868110236220472, + "grad_norm": 8.913108825683594, + "learning_rate": 9.094616856400512e-07, + "loss": 0.3715, + "step": 1409 + }, + { + "epoch": 1.3877952755905512, + "grad_norm": 11.698339462280273, + "learning_rate": 9.090205416098685e-07, + "loss": 0.862, + "step": 1410 + }, + { + "epoch": 1.388779527559055, + "grad_norm": 6.275789260864258, + "learning_rate": 9.08578433395429e-07, + "loss": 0.2813, + "step": 1411 + }, + { + "epoch": 1.389763779527559, + "grad_norm": 7.035228729248047, + "learning_rate": 9.081353620431783e-07, + "loss": 0.2469, + "step": 1412 + }, + { + "epoch": 1.390748031496063, + "grad_norm": 11.262689590454102, + "learning_rate": 9.07691328601842e-07, + "loss": 0.6913, + "step": 1413 + }, + { + "epoch": 1.391732283464567, + "grad_norm": 9.390658378601074, + "learning_rate": 9.072463341224232e-07, + "loss": 0.4501, + "step": 1414 + }, + { + "epoch": 1.3927165354330708, + "grad_norm": 11.353736877441406, + "learning_rate": 9.068003796581991e-07, + "loss": 0.4488, + "step": 1415 + }, + { + "epoch": 1.3937007874015748, + "grad_norm": 14.299334526062012, + "learning_rate": 9.063534662647195e-07, + "loss": 0.9175, + "step": 1416 + }, + { + "epoch": 1.3946850393700787, + "grad_norm": 9.409429550170898, + "learning_rate": 9.059055949998039e-07, + "loss": 0.5717, + "step": 1417 + }, + { + "epoch": 1.3956692913385826, + "grad_norm": 11.336569786071777, + "learning_rate": 9.05456766923539e-07, + "loss": 0.6312, + "step": 1418 + }, + { + "epoch": 1.3966535433070866, + "grad_norm": 8.701967239379883, + "learning_rate": 9.05006983098276e-07, + "loss": 0.4169, + "step": 1419 + }, + { + "epoch": 1.3976377952755905, + "grad_norm": 13.990185737609863, + "learning_rate": 9.045562445886289e-07, + "loss": 1.0509, + "step": 1420 + }, + { + "epoch": 1.3986220472440944, + "grad_norm": 14.266518592834473, + "learning_rate": 9.041045524614707e-07, + "loss": 0.7645, + "step": 1421 + }, + { + "epoch": 1.3996062992125984, + "grad_norm": 12.690227508544922, + "learning_rate": 9.03651907785932e-07, + "loss": 0.8793, + "step": 1422 + }, + { + "epoch": 1.4005905511811023, + "grad_norm": 9.04496955871582, + "learning_rate": 9.03198311633398e-07, + "loss": 0.4623, + "step": 1423 + }, + { + "epoch": 1.4015748031496063, + "grad_norm": 11.421829223632812, + "learning_rate": 9.027437650775055e-07, + "loss": 1.0466, + "step": 1424 + }, + { + "epoch": 1.4025590551181102, + "grad_norm": 10.869682312011719, + "learning_rate": 9.022882691941417e-07, + "loss": 0.9142, + "step": 1425 + }, + { + "epoch": 1.4035433070866141, + "grad_norm": 20.46613883972168, + "learning_rate": 9.018318250614402e-07, + "loss": 0.7341, + "step": 1426 + }, + { + "epoch": 1.404527559055118, + "grad_norm": 6.882892608642578, + "learning_rate": 9.01374433759779e-07, + "loss": 0.343, + "step": 1427 + }, + { + "epoch": 1.405511811023622, + "grad_norm": 20.45710563659668, + "learning_rate": 9.009160963717784e-07, + "loss": 0.9336, + "step": 1428 + }, + { + "epoch": 1.406496062992126, + "grad_norm": 6.482004165649414, + "learning_rate": 9.004568139822978e-07, + "loss": 0.2852, + "step": 1429 + }, + { + "epoch": 1.4074803149606299, + "grad_norm": 6.5446085929870605, + "learning_rate": 8.999965876784333e-07, + "loss": 0.258, + "step": 1430 + }, + { + "epoch": 1.4084645669291338, + "grad_norm": 10.215383529663086, + "learning_rate": 8.995354185495155e-07, + "loss": 0.6507, + "step": 1431 + }, + { + "epoch": 1.4094488188976377, + "grad_norm": 12.591222763061523, + "learning_rate": 8.990733076871063e-07, + "loss": 0.6733, + "step": 1432 + }, + { + "epoch": 1.4104330708661417, + "grad_norm": 15.16763973236084, + "learning_rate": 8.986102561849967e-07, + "loss": 0.9762, + "step": 1433 + }, + { + "epoch": 1.4114173228346456, + "grad_norm": 5.284968376159668, + "learning_rate": 8.981462651392041e-07, + "loss": 0.1865, + "step": 1434 + }, + { + "epoch": 1.4124015748031495, + "grad_norm": 11.886442184448242, + "learning_rate": 8.9768133564797e-07, + "loss": 0.5412, + "step": 1435 + }, + { + "epoch": 1.4133858267716535, + "grad_norm": 11.830586433410645, + "learning_rate": 8.972154688117568e-07, + "loss": 0.5507, + "step": 1436 + }, + { + "epoch": 1.4143700787401574, + "grad_norm": 14.710182189941406, + "learning_rate": 8.967486657332462e-07, + "loss": 0.6734, + "step": 1437 + }, + { + "epoch": 1.4153543307086613, + "grad_norm": 11.092493057250977, + "learning_rate": 8.962809275173347e-07, + "loss": 0.6331, + "step": 1438 + }, + { + "epoch": 1.4163385826771653, + "grad_norm": 11.088115692138672, + "learning_rate": 8.958122552711336e-07, + "loss": 0.5477, + "step": 1439 + }, + { + "epoch": 1.4173228346456692, + "grad_norm": 10.937018394470215, + "learning_rate": 8.953426501039641e-07, + "loss": 0.493, + "step": 1440 + }, + { + "epoch": 1.4183070866141732, + "grad_norm": 12.68562126159668, + "learning_rate": 8.94872113127356e-07, + "loss": 0.6708, + "step": 1441 + }, + { + "epoch": 1.419291338582677, + "grad_norm": 9.494450569152832, + "learning_rate": 8.944006454550443e-07, + "loss": 0.4402, + "step": 1442 + }, + { + "epoch": 1.420275590551181, + "grad_norm": 27.259986877441406, + "learning_rate": 8.939282482029671e-07, + "loss": 1.5426, + "step": 1443 + }, + { + "epoch": 1.421259842519685, + "grad_norm": 11.444234848022461, + "learning_rate": 8.934549224892629e-07, + "loss": 0.483, + "step": 1444 + }, + { + "epoch": 1.422244094488189, + "grad_norm": 10.555806159973145, + "learning_rate": 8.929806694342677e-07, + "loss": 0.6425, + "step": 1445 + }, + { + "epoch": 1.423228346456693, + "grad_norm": 12.313192367553711, + "learning_rate": 8.925054901605124e-07, + "loss": 0.8039, + "step": 1446 + }, + { + "epoch": 1.4242125984251968, + "grad_norm": 13.14362907409668, + "learning_rate": 8.9202938579272e-07, + "loss": 0.8185, + "step": 1447 + }, + { + "epoch": 1.425196850393701, + "grad_norm": 17.453140258789062, + "learning_rate": 8.915523574578038e-07, + "loss": 1.1124, + "step": 1448 + }, + { + "epoch": 1.4261811023622046, + "grad_norm": 11.470340728759766, + "learning_rate": 8.910744062848633e-07, + "loss": 0.9253, + "step": 1449 + }, + { + "epoch": 1.4271653543307088, + "grad_norm": 19.180891036987305, + "learning_rate": 8.90595533405183e-07, + "loss": 1.3097, + "step": 1450 + }, + { + "epoch": 1.4281496062992125, + "grad_norm": 8.698410987854004, + "learning_rate": 8.901157399522286e-07, + "loss": 0.4352, + "step": 1451 + }, + { + "epoch": 1.4291338582677167, + "grad_norm": 12.320472717285156, + "learning_rate": 8.896350270616449e-07, + "loss": 0.7809, + "step": 1452 + }, + { + "epoch": 1.4301181102362204, + "grad_norm": 8.019765853881836, + "learning_rate": 8.891533958712529e-07, + "loss": 0.5639, + "step": 1453 + }, + { + "epoch": 1.4311023622047245, + "grad_norm": 12.787984848022461, + "learning_rate": 8.88670847521047e-07, + "loss": 0.6546, + "step": 1454 + }, + { + "epoch": 1.4320866141732282, + "grad_norm": 12.937506675720215, + "learning_rate": 8.881873831531928e-07, + "loss": 0.644, + "step": 1455 + }, + { + "epoch": 1.4330708661417324, + "grad_norm": 12.214847564697266, + "learning_rate": 8.877030039120242e-07, + "loss": 0.891, + "step": 1456 + }, + { + "epoch": 1.4340551181102361, + "grad_norm": 13.094480514526367, + "learning_rate": 8.872177109440398e-07, + "loss": 0.7778, + "step": 1457 + }, + { + "epoch": 1.4350393700787403, + "grad_norm": 8.966718673706055, + "learning_rate": 8.867315053979016e-07, + "loss": 0.3828, + "step": 1458 + }, + { + "epoch": 1.436023622047244, + "grad_norm": 5.761513710021973, + "learning_rate": 8.862443884244314e-07, + "loss": 0.2598, + "step": 1459 + }, + { + "epoch": 1.4370078740157481, + "grad_norm": 13.386043548583984, + "learning_rate": 8.857563611766085e-07, + "loss": 0.9535, + "step": 1460 + }, + { + "epoch": 1.4379921259842519, + "grad_norm": 9.792831420898438, + "learning_rate": 8.852674248095663e-07, + "loss": 0.4564, + "step": 1461 + }, + { + "epoch": 1.438976377952756, + "grad_norm": 8.115612030029297, + "learning_rate": 8.847775804805907e-07, + "loss": 0.2435, + "step": 1462 + }, + { + "epoch": 1.4399606299212597, + "grad_norm": 6.284011363983154, + "learning_rate": 8.84286829349116e-07, + "loss": 0.3064, + "step": 1463 + }, + { + "epoch": 1.4409448818897639, + "grad_norm": 10.450715065002441, + "learning_rate": 8.837951725767234e-07, + "loss": 0.6287, + "step": 1464 + }, + { + "epoch": 1.4419291338582676, + "grad_norm": 8.95020580291748, + "learning_rate": 8.833026113271374e-07, + "loss": 0.3884, + "step": 1465 + }, + { + "epoch": 1.4429133858267718, + "grad_norm": 13.18276309967041, + "learning_rate": 8.828091467662234e-07, + "loss": 0.8437, + "step": 1466 + }, + { + "epoch": 1.4438976377952755, + "grad_norm": 8.962101936340332, + "learning_rate": 8.82314780061985e-07, + "loss": 0.4804, + "step": 1467 + }, + { + "epoch": 1.4448818897637796, + "grad_norm": 8.015385627746582, + "learning_rate": 8.81819512384561e-07, + "loss": 0.3724, + "step": 1468 + }, + { + "epoch": 1.4458661417322833, + "grad_norm": 9.255770683288574, + "learning_rate": 8.813233449062229e-07, + "loss": 0.5828, + "step": 1469 + }, + { + "epoch": 1.4468503937007875, + "grad_norm": 15.871583938598633, + "learning_rate": 8.808262788013718e-07, + "loss": 0.7696, + "step": 1470 + }, + { + "epoch": 1.4478346456692912, + "grad_norm": 13.32904052734375, + "learning_rate": 8.803283152465357e-07, + "loss": 0.8469, + "step": 1471 + }, + { + "epoch": 1.4488188976377954, + "grad_norm": 9.550171852111816, + "learning_rate": 8.798294554203674e-07, + "loss": 0.4554, + "step": 1472 + }, + { + "epoch": 1.449803149606299, + "grad_norm": 5.600789546966553, + "learning_rate": 8.793297005036402e-07, + "loss": 0.2414, + "step": 1473 + }, + { + "epoch": 1.4507874015748032, + "grad_norm": 11.869569778442383, + "learning_rate": 8.788290516792471e-07, + "loss": 0.5172, + "step": 1474 + }, + { + "epoch": 1.451771653543307, + "grad_norm": 10.772007942199707, + "learning_rate": 8.783275101321962e-07, + "loss": 0.5951, + "step": 1475 + }, + { + "epoch": 1.452755905511811, + "grad_norm": 20.940526962280273, + "learning_rate": 8.778250770496084e-07, + "loss": 1.1616, + "step": 1476 + }, + { + "epoch": 1.453740157480315, + "grad_norm": 14.303383827209473, + "learning_rate": 8.773217536207158e-07, + "loss": 0.8693, + "step": 1477 + }, + { + "epoch": 1.454724409448819, + "grad_norm": 15.649582862854004, + "learning_rate": 8.76817541036857e-07, + "loss": 0.524, + "step": 1478 + }, + { + "epoch": 1.455708661417323, + "grad_norm": 9.849818229675293, + "learning_rate": 8.763124404914754e-07, + "loss": 0.4673, + "step": 1479 + }, + { + "epoch": 1.4566929133858268, + "grad_norm": 8.481264114379883, + "learning_rate": 8.758064531801166e-07, + "loss": 0.3573, + "step": 1480 + }, + { + "epoch": 1.4576771653543308, + "grad_norm": 8.395500183105469, + "learning_rate": 8.752995803004247e-07, + "loss": 0.3688, + "step": 1481 + }, + { + "epoch": 1.4586614173228347, + "grad_norm": 11.030028343200684, + "learning_rate": 8.747918230521398e-07, + "loss": 0.6607, + "step": 1482 + }, + { + "epoch": 1.4596456692913387, + "grad_norm": 5.175130367279053, + "learning_rate": 8.742831826370956e-07, + "loss": 0.1564, + "step": 1483 + }, + { + "epoch": 1.4606299212598426, + "grad_norm": 17.556764602661133, + "learning_rate": 8.737736602592158e-07, + "loss": 1.0894, + "step": 1484 + }, + { + "epoch": 1.4616141732283465, + "grad_norm": 7.026211261749268, + "learning_rate": 8.732632571245121e-07, + "loss": 0.3176, + "step": 1485 + }, + { + "epoch": 1.4625984251968505, + "grad_norm": 9.502352714538574, + "learning_rate": 8.727519744410806e-07, + "loss": 0.511, + "step": 1486 + }, + { + "epoch": 1.4635826771653544, + "grad_norm": 12.687467575073242, + "learning_rate": 8.722398134190995e-07, + "loss": 0.7712, + "step": 1487 + }, + { + "epoch": 1.4645669291338583, + "grad_norm": 13.778657913208008, + "learning_rate": 8.717267752708257e-07, + "loss": 0.6294, + "step": 1488 + }, + { + "epoch": 1.4655511811023623, + "grad_norm": 9.563069343566895, + "learning_rate": 8.712128612105924e-07, + "loss": 0.518, + "step": 1489 + }, + { + "epoch": 1.4665354330708662, + "grad_norm": 12.011832237243652, + "learning_rate": 8.706980724548058e-07, + "loss": 0.6499, + "step": 1490 + }, + { + "epoch": 1.4675196850393701, + "grad_norm": 13.67669677734375, + "learning_rate": 8.701824102219428e-07, + "loss": 0.6263, + "step": 1491 + }, + { + "epoch": 1.468503937007874, + "grad_norm": 10.781290054321289, + "learning_rate": 8.696658757325474e-07, + "loss": 0.5123, + "step": 1492 + }, + { + "epoch": 1.469488188976378, + "grad_norm": 10.782546997070312, + "learning_rate": 8.691484702092286e-07, + "loss": 0.465, + "step": 1493 + }, + { + "epoch": 1.470472440944882, + "grad_norm": 8.449795722961426, + "learning_rate": 8.686301948766568e-07, + "loss": 0.5224, + "step": 1494 + }, + { + "epoch": 1.4714566929133859, + "grad_norm": 14.044183731079102, + "learning_rate": 8.681110509615612e-07, + "loss": 0.9398, + "step": 1495 + }, + { + "epoch": 1.4724409448818898, + "grad_norm": 8.62885856628418, + "learning_rate": 8.675910396927269e-07, + "loss": 0.4608, + "step": 1496 + }, + { + "epoch": 1.4734251968503937, + "grad_norm": 9.031576156616211, + "learning_rate": 8.670701623009921e-07, + "loss": 0.3787, + "step": 1497 + }, + { + "epoch": 1.4744094488188977, + "grad_norm": 9.516302108764648, + "learning_rate": 8.66548420019245e-07, + "loss": 0.4767, + "step": 1498 + }, + { + "epoch": 1.4753937007874016, + "grad_norm": 10.302477836608887, + "learning_rate": 8.66025814082421e-07, + "loss": 0.3481, + "step": 1499 + }, + { + "epoch": 1.4763779527559056, + "grad_norm": 10.279803276062012, + "learning_rate": 8.655023457274994e-07, + "loss": 0.5924, + "step": 1500 + }, + { + "epoch": 1.4773622047244095, + "grad_norm": 9.42895221710205, + "learning_rate": 8.649780161935012e-07, + "loss": 0.5202, + "step": 1501 + }, + { + "epoch": 1.4783464566929134, + "grad_norm": 20.95256233215332, + "learning_rate": 8.644528267214859e-07, + "loss": 1.1383, + "step": 1502 + }, + { + "epoch": 1.4793307086614174, + "grad_norm": 8.058876037597656, + "learning_rate": 8.639267785545478e-07, + "loss": 0.473, + "step": 1503 + }, + { + "epoch": 1.4803149606299213, + "grad_norm": 8.1046781539917, + "learning_rate": 8.63399872937814e-07, + "loss": 0.2584, + "step": 1504 + }, + { + "epoch": 1.4812992125984252, + "grad_norm": 17.544803619384766, + "learning_rate": 8.628721111184414e-07, + "loss": 1.0639, + "step": 1505 + }, + { + "epoch": 1.4822834645669292, + "grad_norm": 10.679337501525879, + "learning_rate": 8.623434943456132e-07, + "loss": 0.564, + "step": 1506 + }, + { + "epoch": 1.483267716535433, + "grad_norm": 13.62854290008545, + "learning_rate": 8.61814023870536e-07, + "loss": 0.7819, + "step": 1507 + }, + { + "epoch": 1.484251968503937, + "grad_norm": 10.745209693908691, + "learning_rate": 8.612837009464377e-07, + "loss": 0.7226, + "step": 1508 + }, + { + "epoch": 1.485236220472441, + "grad_norm": 9.823802947998047, + "learning_rate": 8.607525268285634e-07, + "loss": 0.5971, + "step": 1509 + }, + { + "epoch": 1.486220472440945, + "grad_norm": 11.389198303222656, + "learning_rate": 8.602205027741731e-07, + "loss": 0.6205, + "step": 1510 + }, + { + "epoch": 1.4872047244094488, + "grad_norm": 12.77564525604248, + "learning_rate": 8.596876300425384e-07, + "loss": 0.8551, + "step": 1511 + }, + { + "epoch": 1.4881889763779528, + "grad_norm": 17.55690574645996, + "learning_rate": 8.591539098949401e-07, + "loss": 1.4143, + "step": 1512 + }, + { + "epoch": 1.4891732283464567, + "grad_norm": 7.800512790679932, + "learning_rate": 8.586193435946641e-07, + "loss": 0.3514, + "step": 1513 + }, + { + "epoch": 1.4901574803149606, + "grad_norm": 34.540836334228516, + "learning_rate": 8.580839324069998e-07, + "loss": 2.243, + "step": 1514 + }, + { + "epoch": 1.4911417322834646, + "grad_norm": 8.57536506652832, + "learning_rate": 8.57547677599236e-07, + "loss": 0.3916, + "step": 1515 + }, + { + "epoch": 1.4921259842519685, + "grad_norm": 14.394399642944336, + "learning_rate": 8.570105804406583e-07, + "loss": 0.8289, + "step": 1516 + }, + { + "epoch": 1.4931102362204725, + "grad_norm": 8.080702781677246, + "learning_rate": 8.564726422025462e-07, + "loss": 0.5227, + "step": 1517 + }, + { + "epoch": 1.4940944881889764, + "grad_norm": 8.852377891540527, + "learning_rate": 8.5593386415817e-07, + "loss": 0.3593, + "step": 1518 + }, + { + "epoch": 1.4950787401574803, + "grad_norm": 8.874595642089844, + "learning_rate": 8.553942475827879e-07, + "loss": 0.3856, + "step": 1519 + }, + { + "epoch": 1.4960629921259843, + "grad_norm": 8.662959098815918, + "learning_rate": 8.548537937536425e-07, + "loss": 0.2703, + "step": 1520 + }, + { + "epoch": 1.4970472440944882, + "grad_norm": 19.03708839416504, + "learning_rate": 8.543125039499582e-07, + "loss": 1.0162, + "step": 1521 + }, + { + "epoch": 1.4980314960629921, + "grad_norm": 10.240668296813965, + "learning_rate": 8.537703794529387e-07, + "loss": 0.5613, + "step": 1522 + }, + { + "epoch": 1.499015748031496, + "grad_norm": 12.856542587280273, + "learning_rate": 8.532274215457625e-07, + "loss": 0.8964, + "step": 1523 + }, + { + "epoch": 1.5, + "grad_norm": 9.26098346710205, + "learning_rate": 8.526836315135815e-07, + "loss": 0.546, + "step": 1524 + }, + { + "epoch": 1.500984251968504, + "grad_norm": 7.665274620056152, + "learning_rate": 8.521390106435168e-07, + "loss": 0.3827, + "step": 1525 + }, + { + "epoch": 1.5019685039370079, + "grad_norm": 9.265157699584961, + "learning_rate": 8.515935602246558e-07, + "loss": 0.4451, + "step": 1526 + }, + { + "epoch": 1.5029527559055118, + "grad_norm": 12.346192359924316, + "learning_rate": 8.510472815480498e-07, + "loss": 0.7627, + "step": 1527 + }, + { + "epoch": 1.5039370078740157, + "grad_norm": 9.11153507232666, + "learning_rate": 8.505001759067106e-07, + "loss": 0.4021, + "step": 1528 + }, + { + "epoch": 1.5049212598425197, + "grad_norm": 10.442115783691406, + "learning_rate": 8.499522445956071e-07, + "loss": 0.9789, + "step": 1529 + }, + { + "epoch": 1.5059055118110236, + "grad_norm": 12.224811553955078, + "learning_rate": 8.494034889116626e-07, + "loss": 0.588, + "step": 1530 + }, + { + "epoch": 1.5059055118110236, + "eval_Qnli-dev_cosine_accuracy": 0.666015625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8726706504821777, + "eval_Qnli-dev_cosine_ap": 0.686138338031671, + "eval_Qnli-dev_cosine_f1": 0.6888111888111887, + "eval_Qnli-dev_cosine_f1_threshold": 0.7970194816589355, + "eval_Qnli-dev_cosine_precision": 0.5863095238095238, + "eval_Qnli-dev_cosine_recall": 0.8347457627118644, + "eval_Qnli-dev_dot_accuracy": 0.658203125, + "eval_Qnli-dev_dot_accuracy_threshold": 662.91357421875, + "eval_Qnli-dev_dot_ap": 0.6838401632111198, + "eval_Qnli-dev_dot_f1": 0.6871961102106969, + "eval_Qnli-dev_dot_f1_threshold": 589.2022705078125, + "eval_Qnli-dev_dot_precision": 0.5564304461942258, + "eval_Qnli-dev_dot_recall": 0.8983050847457628, + "eval_Qnli-dev_euclidean_accuracy": 0.666015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.948681831359863, + "eval_Qnli-dev_euclidean_ap": 0.6864197659765374, + "eval_Qnli-dev_euclidean_f1": 0.6877192982456141, + "eval_Qnli-dev_euclidean_f1_threshold": 17.551612854003906, + "eval_Qnli-dev_euclidean_precision": 0.5868263473053892, + "eval_Qnli-dev_euclidean_recall": 0.8305084745762712, + "eval_Qnli-dev_manhattan_accuracy": 0.6640625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 350.95172119140625, + "eval_Qnli-dev_manhattan_ap": 0.6968749918700499, + "eval_Qnli-dev_manhattan_f1": 0.6843853820598007, + "eval_Qnli-dev_manhattan_f1_threshold": 384.99029541015625, + "eval_Qnli-dev_manhattan_precision": 0.5628415300546448, + "eval_Qnli-dev_manhattan_recall": 0.8728813559322034, + "eval_Qnli-dev_max_accuracy": 0.666015625, + "eval_Qnli-dev_max_accuracy_threshold": 662.91357421875, + "eval_Qnli-dev_max_ap": 0.6968749918700499, + "eval_Qnli-dev_max_f1": 0.6888111888111887, + "eval_Qnli-dev_max_f1_threshold": 589.2022705078125, + "eval_Qnli-dev_max_precision": 0.5868263473053892, + "eval_Qnli-dev_max_recall": 0.8983050847457628, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9320794343948364, + "eval_allNLI-dev_cosine_ap": 0.5060248919066748, + "eval_allNLI-dev_cosine_f1": 0.5466666666666666, + "eval_allNLI-dev_cosine_f1_threshold": 0.8609572052955627, + "eval_allNLI-dev_cosine_precision": 0.44404332129963897, + "eval_allNLI-dev_cosine_recall": 0.7109826589595376, + "eval_allNLI-dev_dot_accuracy": 0.6875, + "eval_allNLI-dev_dot_accuracy_threshold": 710.0655517578125, + "eval_allNLI-dev_dot_ap": 0.5066897132253594, + "eval_allNLI-dev_dot_f1": 0.5426008968609866, + "eval_allNLI-dev_dot_f1_threshold": 657.003173828125, + "eval_allNLI-dev_dot_precision": 0.4432234432234432, + "eval_allNLI-dev_dot_recall": 0.6994219653179191, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.182336807250977, + "eval_allNLI-dev_euclidean_ap": 0.5062293920155956, + "eval_allNLI-dev_euclidean_f1": 0.5454545454545455, + "eval_allNLI-dev_euclidean_f1_threshold": 14.568036079406738, + "eval_allNLI-dev_euclidean_precision": 0.44244604316546765, + "eval_allNLI-dev_euclidean_recall": 0.7109826589595376, + "eval_allNLI-dev_manhattan_accuracy": 0.685546875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 200.71170043945312, + "eval_allNLI-dev_manhattan_ap": 0.5100764805616853, + "eval_allNLI-dev_manhattan_f1": 0.5505050505050505, + "eval_allNLI-dev_manhattan_f1_threshold": 286.9205627441406, + "eval_allNLI-dev_manhattan_precision": 0.48878923766816146, + "eval_allNLI-dev_manhattan_recall": 0.630057803468208, + "eval_allNLI-dev_max_accuracy": 0.689453125, + "eval_allNLI-dev_max_accuracy_threshold": 710.0655517578125, + "eval_allNLI-dev_max_ap": 0.5100764805616853, + "eval_allNLI-dev_max_f1": 0.5505050505050505, + "eval_allNLI-dev_max_f1_threshold": 657.003173828125, + "eval_allNLI-dev_max_precision": 0.48878923766816146, + "eval_allNLI-dev_max_recall": 0.7109826589595376, + "eval_loss": 0.7326235175132751, + "eval_runtime": 51.8019, + "eval_samples_per_second": 32.122, + "eval_sequential_score": 0.6968749918700499, + "eval_steps_per_second": 0.135, + "eval_sts-test_pearson_cosine": 0.6255681039687839, + "eval_sts-test_pearson_dot": 0.6250716332304032, + "eval_sts-test_pearson_euclidean": 0.646878963631455, + "eval_sts-test_pearson_manhattan": 0.6523584389407872, + "eval_sts-test_pearson_max": 0.6523584389407872, + "eval_sts-test_spearman_cosine": 0.6466049344240437, + "eval_sts-test_spearman_dot": 0.6462539950322124, + "eval_sts-test_spearman_euclidean": 0.647093752452222, + "eval_sts-test_spearman_manhattan": 0.6508617639101164, + "eval_sts-test_spearman_max": 0.6508617639101164, + "step": 1530 + }, + { + "epoch": 1.5068897637795275, + "grad_norm": 7.402773857116699, + "learning_rate": 8.488539101537518e-07, + "loss": 0.2848, + "step": 1531 + }, + { + "epoch": 1.5078740157480315, + "grad_norm": 10.254727363586426, + "learning_rate": 8.483035096226975e-07, + "loss": 0.4957, + "step": 1532 + }, + { + "epoch": 1.5088582677165354, + "grad_norm": 5.65178108215332, + "learning_rate": 8.477522886212673e-07, + "loss": 0.1843, + "step": 1533 + }, + { + "epoch": 1.5098425196850394, + "grad_norm": 18.793121337890625, + "learning_rate": 8.472002484541711e-07, + "loss": 1.4668, + "step": 1534 + }, + { + "epoch": 1.5108267716535433, + "grad_norm": 8.860997200012207, + "learning_rate": 8.466473904280579e-07, + "loss": 0.4216, + "step": 1535 + }, + { + "epoch": 1.5118110236220472, + "grad_norm": 10.477073669433594, + "learning_rate": 8.46093715851512e-07, + "loss": 0.6761, + "step": 1536 + }, + { + "epoch": 1.5127952755905512, + "grad_norm": 4.932042121887207, + "learning_rate": 8.455392260350507e-07, + "loss": 0.1538, + "step": 1537 + }, + { + "epoch": 1.513779527559055, + "grad_norm": 12.891571044921875, + "learning_rate": 8.449839222911214e-07, + "loss": 0.6587, + "step": 1538 + }, + { + "epoch": 1.514763779527559, + "grad_norm": 8.308895111083984, + "learning_rate": 8.444278059340971e-07, + "loss": 0.2875, + "step": 1539 + }, + { + "epoch": 1.515748031496063, + "grad_norm": 4.691012382507324, + "learning_rate": 8.438708782802748e-07, + "loss": 0.2036, + "step": 1540 + }, + { + "epoch": 1.516732283464567, + "grad_norm": 11.703125, + "learning_rate": 8.43313140647872e-07, + "loss": 0.715, + "step": 1541 + }, + { + "epoch": 1.5177165354330708, + "grad_norm": 10.412879943847656, + "learning_rate": 8.427545943570226e-07, + "loss": 0.5181, + "step": 1542 + }, + { + "epoch": 1.5187007874015748, + "grad_norm": 10.32512092590332, + "learning_rate": 8.421952407297752e-07, + "loss": 0.6573, + "step": 1543 + }, + { + "epoch": 1.5196850393700787, + "grad_norm": 14.16104793548584, + "learning_rate": 8.416350810900892e-07, + "loss": 0.7494, + "step": 1544 + }, + { + "epoch": 1.5206692913385826, + "grad_norm": 9.666444778442383, + "learning_rate": 8.410741167638317e-07, + "loss": 0.5275, + "step": 1545 + }, + { + "epoch": 1.5216535433070866, + "grad_norm": 11.393864631652832, + "learning_rate": 8.405123490787741e-07, + "loss": 0.4506, + "step": 1546 + }, + { + "epoch": 1.5226377952755905, + "grad_norm": 11.62814712524414, + "learning_rate": 8.3994977936459e-07, + "loss": 0.726, + "step": 1547 + }, + { + "epoch": 1.5236220472440944, + "grad_norm": 8.988724708557129, + "learning_rate": 8.393864089528514e-07, + "loss": 0.3336, + "step": 1548 + }, + { + "epoch": 1.5246062992125984, + "grad_norm": 14.445733070373535, + "learning_rate": 8.388222391770245e-07, + "loss": 1.3186, + "step": 1549 + }, + { + "epoch": 1.5255905511811023, + "grad_norm": 17.472166061401367, + "learning_rate": 8.382572713724683e-07, + "loss": 0.8029, + "step": 1550 + }, + { + "epoch": 1.5265748031496063, + "grad_norm": 14.064423561096191, + "learning_rate": 8.376915068764308e-07, + "loss": 0.6675, + "step": 1551 + }, + { + "epoch": 1.5275590551181102, + "grad_norm": 9.772034645080566, + "learning_rate": 8.371249470280453e-07, + "loss": 0.4343, + "step": 1552 + }, + { + "epoch": 1.5285433070866141, + "grad_norm": 9.490726470947266, + "learning_rate": 8.36557593168328e-07, + "loss": 0.4902, + "step": 1553 + }, + { + "epoch": 1.529527559055118, + "grad_norm": 9.363591194152832, + "learning_rate": 8.359894466401743e-07, + "loss": 0.4472, + "step": 1554 + }, + { + "epoch": 1.530511811023622, + "grad_norm": 11.04429817199707, + "learning_rate": 8.354205087883556e-07, + "loss": 0.6886, + "step": 1555 + }, + { + "epoch": 1.531496062992126, + "grad_norm": 22.000707626342773, + "learning_rate": 8.348507809595169e-07, + "loss": 1.133, + "step": 1556 + }, + { + "epoch": 1.53248031496063, + "grad_norm": 9.29854965209961, + "learning_rate": 8.342802645021727e-07, + "loss": 0.4037, + "step": 1557 + }, + { + "epoch": 1.5334645669291338, + "grad_norm": 7.053875923156738, + "learning_rate": 8.337089607667039e-07, + "loss": 0.2597, + "step": 1558 + }, + { + "epoch": 1.534448818897638, + "grad_norm": 15.32122802734375, + "learning_rate": 8.331368711053552e-07, + "loss": 0.7336, + "step": 1559 + }, + { + "epoch": 1.5354330708661417, + "grad_norm": 14.502772331237793, + "learning_rate": 8.325639968722313e-07, + "loss": 1.0777, + "step": 1560 + }, + { + "epoch": 1.5364173228346458, + "grad_norm": 24.384653091430664, + "learning_rate": 8.319903394232942e-07, + "loss": 1.3634, + "step": 1561 + }, + { + "epoch": 1.5374015748031495, + "grad_norm": 13.809354782104492, + "learning_rate": 8.314159001163596e-07, + "loss": 0.7729, + "step": 1562 + }, + { + "epoch": 1.5383858267716537, + "grad_norm": 29.76972770690918, + "learning_rate": 8.308406803110936e-07, + "loss": 1.8108, + "step": 1563 + }, + { + "epoch": 1.5393700787401574, + "grad_norm": 13.510603904724121, + "learning_rate": 8.302646813690101e-07, + "loss": 0.5194, + "step": 1564 + }, + { + "epoch": 1.5403543307086616, + "grad_norm": 5.753684043884277, + "learning_rate": 8.296879046534668e-07, + "loss": 0.2106, + "step": 1565 + }, + { + "epoch": 1.5413385826771653, + "grad_norm": 10.1058349609375, + "learning_rate": 8.291103515296628e-07, + "loss": 0.5281, + "step": 1566 + }, + { + "epoch": 1.5423228346456694, + "grad_norm": 9.436454772949219, + "learning_rate": 8.285320233646345e-07, + "loss": 0.3398, + "step": 1567 + }, + { + "epoch": 1.5433070866141732, + "grad_norm": 12.91254997253418, + "learning_rate": 8.279529215272526e-07, + "loss": 0.9779, + "step": 1568 + }, + { + "epoch": 1.5442913385826773, + "grad_norm": 9.625316619873047, + "learning_rate": 8.273730473882198e-07, + "loss": 0.6771, + "step": 1569 + }, + { + "epoch": 1.545275590551181, + "grad_norm": 13.374066352844238, + "learning_rate": 8.267924023200662e-07, + "loss": 0.8631, + "step": 1570 + }, + { + "epoch": 1.5462598425196852, + "grad_norm": 9.743898391723633, + "learning_rate": 8.262109876971466e-07, + "loss": 0.5849, + "step": 1571 + }, + { + "epoch": 1.547244094488189, + "grad_norm": 9.956043243408203, + "learning_rate": 8.256288048956379e-07, + "loss": 0.5975, + "step": 1572 + }, + { + "epoch": 1.548228346456693, + "grad_norm": 9.438084602355957, + "learning_rate": 8.250458552935343e-07, + "loss": 0.6461, + "step": 1573 + }, + { + "epoch": 1.5492125984251968, + "grad_norm": 7.46604585647583, + "learning_rate": 8.244621402706462e-07, + "loss": 0.3175, + "step": 1574 + }, + { + "epoch": 1.550196850393701, + "grad_norm": 11.996360778808594, + "learning_rate": 8.238776612085943e-07, + "loss": 0.7706, + "step": 1575 + }, + { + "epoch": 1.5511811023622046, + "grad_norm": 17.977930068969727, + "learning_rate": 8.232924194908088e-07, + "loss": 1.2396, + "step": 1576 + }, + { + "epoch": 1.5521653543307088, + "grad_norm": 11.14438247680664, + "learning_rate": 8.227064165025247e-07, + "loss": 0.5457, + "step": 1577 + }, + { + "epoch": 1.5531496062992125, + "grad_norm": 9.292198181152344, + "learning_rate": 8.221196536307788e-07, + "loss": 0.5688, + "step": 1578 + }, + { + "epoch": 1.5541338582677167, + "grad_norm": 8.607418060302734, + "learning_rate": 8.215321322644067e-07, + "loss": 0.6901, + "step": 1579 + }, + { + "epoch": 1.5551181102362204, + "grad_norm": 11.122705459594727, + "learning_rate": 8.209438537940393e-07, + "loss": 0.5043, + "step": 1580 + }, + { + "epoch": 1.5561023622047245, + "grad_norm": 19.077972412109375, + "learning_rate": 8.203548196120994e-07, + "loss": 1.3952, + "step": 1581 + }, + { + "epoch": 1.5570866141732282, + "grad_norm": 18.639930725097656, + "learning_rate": 8.197650311127981e-07, + "loss": 1.0306, + "step": 1582 + }, + { + "epoch": 1.5580708661417324, + "grad_norm": 10.341582298278809, + "learning_rate": 8.191744896921331e-07, + "loss": 0.6551, + "step": 1583 + }, + { + "epoch": 1.5590551181102361, + "grad_norm": 10.178255081176758, + "learning_rate": 8.185831967478831e-07, + "loss": 0.4972, + "step": 1584 + }, + { + "epoch": 1.5600393700787403, + "grad_norm": 8.843459129333496, + "learning_rate": 8.179911536796062e-07, + "loss": 0.479, + "step": 1585 + }, + { + "epoch": 1.561023622047244, + "grad_norm": 14.268756866455078, + "learning_rate": 8.173983618886358e-07, + "loss": 0.8781, + "step": 1586 + }, + { + "epoch": 1.5620078740157481, + "grad_norm": 14.749018669128418, + "learning_rate": 8.168048227780775e-07, + "loss": 1.0491, + "step": 1587 + }, + { + "epoch": 1.5629921259842519, + "grad_norm": 10.37008285522461, + "learning_rate": 8.162105377528057e-07, + "loss": 0.5971, + "step": 1588 + }, + { + "epoch": 1.563976377952756, + "grad_norm": 8.307663917541504, + "learning_rate": 8.156155082194607e-07, + "loss": 0.3664, + "step": 1589 + }, + { + "epoch": 1.5649606299212597, + "grad_norm": 9.04546070098877, + "learning_rate": 8.150197355864446e-07, + "loss": 0.4961, + "step": 1590 + }, + { + "epoch": 1.5659448818897639, + "grad_norm": 3.671900987625122, + "learning_rate": 8.144232212639184e-07, + "loss": 0.1262, + "step": 1591 + }, + { + "epoch": 1.5669291338582676, + "grad_norm": 8.696824073791504, + "learning_rate": 8.138259666637987e-07, + "loss": 0.3932, + "step": 1592 + }, + { + "epoch": 1.5679133858267718, + "grad_norm": 9.176651000976562, + "learning_rate": 8.132279731997544e-07, + "loss": 0.5438, + "step": 1593 + }, + { + "epoch": 1.5688976377952755, + "grad_norm": 9.702910423278809, + "learning_rate": 8.126292422872034e-07, + "loss": 0.366, + "step": 1594 + }, + { + "epoch": 1.5698818897637796, + "grad_norm": 12.509620666503906, + "learning_rate": 8.120297753433087e-07, + "loss": 0.8283, + "step": 1595 + }, + { + "epoch": 1.5708661417322833, + "grad_norm": 14.940932273864746, + "learning_rate": 8.114295737869759e-07, + "loss": 0.8132, + "step": 1596 + }, + { + "epoch": 1.5718503937007875, + "grad_norm": 11.618242263793945, + "learning_rate": 8.108286390388487e-07, + "loss": 0.5428, + "step": 1597 + }, + { + "epoch": 1.5728346456692912, + "grad_norm": 9.441998481750488, + "learning_rate": 8.102269725213071e-07, + "loss": 0.4288, + "step": 1598 + }, + { + "epoch": 1.5738188976377954, + "grad_norm": 17.75216293334961, + "learning_rate": 8.096245756584625e-07, + "loss": 0.7339, + "step": 1599 + }, + { + "epoch": 1.574803149606299, + "grad_norm": 10.789690017700195, + "learning_rate": 8.090214498761551e-07, + "loss": 0.6724, + "step": 1600 + }, + { + "epoch": 1.5757874015748032, + "grad_norm": 11.713319778442383, + "learning_rate": 8.084175966019507e-07, + "loss": 0.5732, + "step": 1601 + }, + { + "epoch": 1.576771653543307, + "grad_norm": 5.157113552093506, + "learning_rate": 8.078130172651368e-07, + "loss": 0.165, + "step": 1602 + }, + { + "epoch": 1.577755905511811, + "grad_norm": 12.99669075012207, + "learning_rate": 8.072077132967193e-07, + "loss": 0.7889, + "step": 1603 + }, + { + "epoch": 1.5787401574803148, + "grad_norm": 16.681533813476562, + "learning_rate": 8.066016861294198e-07, + "loss": 0.7041, + "step": 1604 + }, + { + "epoch": 1.579724409448819, + "grad_norm": 12.58165454864502, + "learning_rate": 8.059949371976711e-07, + "loss": 0.5951, + "step": 1605 + }, + { + "epoch": 1.5807086614173227, + "grad_norm": 8.467135429382324, + "learning_rate": 8.053874679376147e-07, + "loss": 0.4587, + "step": 1606 + }, + { + "epoch": 1.5816929133858268, + "grad_norm": 11.108519554138184, + "learning_rate": 8.047792797870968e-07, + "loss": 0.6667, + "step": 1607 + }, + { + "epoch": 1.5826771653543306, + "grad_norm": 8.499879837036133, + "learning_rate": 8.041703741856657e-07, + "loss": 0.4452, + "step": 1608 + }, + { + "epoch": 1.5836614173228347, + "grad_norm": 10.20980167388916, + "learning_rate": 8.035607525745672e-07, + "loss": 0.2984, + "step": 1609 + }, + { + "epoch": 1.5846456692913384, + "grad_norm": 7.794604301452637, + "learning_rate": 8.029504163967423e-07, + "loss": 0.2643, + "step": 1610 + }, + { + "epoch": 1.5856299212598426, + "grad_norm": 17.791414260864258, + "learning_rate": 8.023393670968235e-07, + "loss": 0.8992, + "step": 1611 + }, + { + "epoch": 1.5866141732283463, + "grad_norm": 11.386825561523438, + "learning_rate": 8.017276061211303e-07, + "loss": 0.608, + "step": 1612 + }, + { + "epoch": 1.5875984251968505, + "grad_norm": 15.693930625915527, + "learning_rate": 8.011151349176681e-07, + "loss": 0.8575, + "step": 1613 + }, + { + "epoch": 1.5885826771653542, + "grad_norm": 5.8269734382629395, + "learning_rate": 8.005019549361222e-07, + "loss": 0.2714, + "step": 1614 + }, + { + "epoch": 1.5895669291338583, + "grad_norm": 9.453310012817383, + "learning_rate": 7.998880676278558e-07, + "loss": 0.4766, + "step": 1615 + }, + { + "epoch": 1.590551181102362, + "grad_norm": 9.279977798461914, + "learning_rate": 7.992734744459068e-07, + "loss": 0.4286, + "step": 1616 + }, + { + "epoch": 1.5915354330708662, + "grad_norm": 9.635364532470703, + "learning_rate": 7.986581768449833e-07, + "loss": 0.384, + "step": 1617 + }, + { + "epoch": 1.59251968503937, + "grad_norm": 6.326999187469482, + "learning_rate": 7.98042176281461e-07, + "loss": 0.3322, + "step": 1618 + }, + { + "epoch": 1.593503937007874, + "grad_norm": 12.379164695739746, + "learning_rate": 7.974254742133795e-07, + "loss": 0.4917, + "step": 1619 + }, + { + "epoch": 1.594488188976378, + "grad_norm": 8.266642570495605, + "learning_rate": 7.968080721004386e-07, + "loss": 0.2052, + "step": 1620 + }, + { + "epoch": 1.595472440944882, + "grad_norm": 11.63606071472168, + "learning_rate": 7.961899714039954e-07, + "loss": 0.4616, + "step": 1621 + }, + { + "epoch": 1.5964566929133859, + "grad_norm": 14.06924819946289, + "learning_rate": 7.955711735870601e-07, + "loss": 0.6446, + "step": 1622 + }, + { + "epoch": 1.5974409448818898, + "grad_norm": 13.163519859313965, + "learning_rate": 7.949516801142932e-07, + "loss": 0.7428, + "step": 1623 + }, + { + "epoch": 1.5984251968503937, + "grad_norm": 11.67570972442627, + "learning_rate": 7.943314924520017e-07, + "loss": 0.7618, + "step": 1624 + }, + { + "epoch": 1.5994094488188977, + "grad_norm": 8.469634056091309, + "learning_rate": 7.93710612068136e-07, + "loss": 0.3268, + "step": 1625 + }, + { + "epoch": 1.6003937007874016, + "grad_norm": 6.006660461425781, + "learning_rate": 7.930890404322857e-07, + "loss": 0.2086, + "step": 1626 + }, + { + "epoch": 1.6013779527559056, + "grad_norm": 15.344389915466309, + "learning_rate": 7.924667790156766e-07, + "loss": 0.4383, + "step": 1627 + }, + { + "epoch": 1.6023622047244095, + "grad_norm": 10.652959823608398, + "learning_rate": 7.918438292911674e-07, + "loss": 0.5524, + "step": 1628 + }, + { + "epoch": 1.6033464566929134, + "grad_norm": 9.096324920654297, + "learning_rate": 7.912201927332461e-07, + "loss": 0.3591, + "step": 1629 + }, + { + "epoch": 1.6043307086614174, + "grad_norm": 5.367679119110107, + "learning_rate": 7.905958708180258e-07, + "loss": 0.1176, + "step": 1630 + }, + { + "epoch": 1.6053149606299213, + "grad_norm": 22.331052780151367, + "learning_rate": 7.899708650232426e-07, + "loss": 1.1557, + "step": 1631 + }, + { + "epoch": 1.6062992125984252, + "grad_norm": 18.126850128173828, + "learning_rate": 7.893451768282508e-07, + "loss": 0.6375, + "step": 1632 + }, + { + "epoch": 1.6072834645669292, + "grad_norm": 12.394740104675293, + "learning_rate": 7.8871880771402e-07, + "loss": 0.5563, + "step": 1633 + }, + { + "epoch": 1.608267716535433, + "grad_norm": 16.225257873535156, + "learning_rate": 7.880917591631312e-07, + "loss": 1.0859, + "step": 1634 + }, + { + "epoch": 1.609251968503937, + "grad_norm": 13.359918594360352, + "learning_rate": 7.874640326597741e-07, + "loss": 0.7384, + "step": 1635 + }, + { + "epoch": 1.610236220472441, + "grad_norm": 8.84333610534668, + "learning_rate": 7.868356296897431e-07, + "loss": 0.3059, + "step": 1636 + }, + { + "epoch": 1.611220472440945, + "grad_norm": 12.398821830749512, + "learning_rate": 7.862065517404334e-07, + "loss": 0.7034, + "step": 1637 + }, + { + "epoch": 1.6122047244094488, + "grad_norm": 13.815208435058594, + "learning_rate": 7.855768003008376e-07, + "loss": 0.6495, + "step": 1638 + }, + { + "epoch": 1.6131889763779528, + "grad_norm": 10.999493598937988, + "learning_rate": 7.849463768615432e-07, + "loss": 0.5121, + "step": 1639 + }, + { + "epoch": 1.6141732283464567, + "grad_norm": 8.639820098876953, + "learning_rate": 7.843152829147278e-07, + "loss": 0.3473, + "step": 1640 + }, + { + "epoch": 1.6151574803149606, + "grad_norm": 8.779716491699219, + "learning_rate": 7.836835199541562e-07, + "loss": 0.3447, + "step": 1641 + }, + { + "epoch": 1.6161417322834646, + "grad_norm": 12.756985664367676, + "learning_rate": 7.830510894751764e-07, + "loss": 0.5997, + "step": 1642 + }, + { + "epoch": 1.6171259842519685, + "grad_norm": 12.467016220092773, + "learning_rate": 7.824179929747168e-07, + "loss": 0.8563, + "step": 1643 + }, + { + "epoch": 1.6181102362204725, + "grad_norm": 9.96351146697998, + "learning_rate": 7.817842319512818e-07, + "loss": 0.3372, + "step": 1644 + }, + { + "epoch": 1.6190944881889764, + "grad_norm": 14.600238800048828, + "learning_rate": 7.811498079049492e-07, + "loss": 1.0262, + "step": 1645 + }, + { + "epoch": 1.6200787401574803, + "grad_norm": 9.461737632751465, + "learning_rate": 7.805147223373658e-07, + "loss": 0.393, + "step": 1646 + }, + { + "epoch": 1.6210629921259843, + "grad_norm": 9.289676666259766, + "learning_rate": 7.798789767517444e-07, + "loss": 0.3937, + "step": 1647 + }, + { + "epoch": 1.6220472440944882, + "grad_norm": 13.504961013793945, + "learning_rate": 7.792425726528596e-07, + "loss": 0.975, + "step": 1648 + }, + { + "epoch": 1.6230314960629921, + "grad_norm": 9.970810890197754, + "learning_rate": 7.786055115470454e-07, + "loss": 0.6906, + "step": 1649 + }, + { + "epoch": 1.624015748031496, + "grad_norm": 11.597799301147461, + "learning_rate": 7.779677949421902e-07, + "loss": 0.5405, + "step": 1650 + }, + { + "epoch": 1.625, + "grad_norm": 9.213786125183105, + "learning_rate": 7.77329424347734e-07, + "loss": 0.4693, + "step": 1651 + }, + { + "epoch": 1.625984251968504, + "grad_norm": 10.477055549621582, + "learning_rate": 7.766904012746653e-07, + "loss": 0.5488, + "step": 1652 + }, + { + "epoch": 1.6269685039370079, + "grad_norm": 9.528751373291016, + "learning_rate": 7.760507272355165e-07, + "loss": 0.4237, + "step": 1653 + }, + { + "epoch": 1.6279527559055118, + "grad_norm": 18.211416244506836, + "learning_rate": 7.754104037443609e-07, + "loss": 0.658, + "step": 1654 + }, + { + "epoch": 1.6289370078740157, + "grad_norm": 13.66952133178711, + "learning_rate": 7.747694323168091e-07, + "loss": 0.9406, + "step": 1655 + }, + { + "epoch": 1.6299212598425197, + "grad_norm": 11.695971488952637, + "learning_rate": 7.741278144700052e-07, + "loss": 0.5151, + "step": 1656 + }, + { + "epoch": 1.6309055118110236, + "grad_norm": 6.70000696182251, + "learning_rate": 7.734855517226234e-07, + "loss": 0.2421, + "step": 1657 + }, + { + "epoch": 1.6318897637795275, + "grad_norm": 16.58856773376465, + "learning_rate": 7.728426455948644e-07, + "loss": 0.8784, + "step": 1658 + }, + { + "epoch": 1.6328740157480315, + "grad_norm": 8.08621883392334, + "learning_rate": 7.721990976084517e-07, + "loss": 0.4185, + "step": 1659 + }, + { + "epoch": 1.6338582677165354, + "grad_norm": 10.188220024108887, + "learning_rate": 7.715549092866282e-07, + "loss": 0.4951, + "step": 1660 + }, + { + "epoch": 1.6348425196850394, + "grad_norm": 18.876163482666016, + "learning_rate": 7.70910082154152e-07, + "loss": 1.3294, + "step": 1661 + }, + { + "epoch": 1.6358267716535433, + "grad_norm": 13.313483238220215, + "learning_rate": 7.702646177372938e-07, + "loss": 0.6635, + "step": 1662 + }, + { + "epoch": 1.6368110236220472, + "grad_norm": 10.43400764465332, + "learning_rate": 7.696185175638325e-07, + "loss": 0.4767, + "step": 1663 + }, + { + "epoch": 1.6377952755905512, + "grad_norm": 17.89423179626465, + "learning_rate": 7.689717831630513e-07, + "loss": 0.8853, + "step": 1664 + }, + { + "epoch": 1.638779527559055, + "grad_norm": 7.940224647521973, + "learning_rate": 7.683244160657353e-07, + "loss": 0.3292, + "step": 1665 + }, + { + "epoch": 1.639763779527559, + "grad_norm": 8.252863883972168, + "learning_rate": 7.676764178041668e-07, + "loss": 0.4556, + "step": 1666 + }, + { + "epoch": 1.640748031496063, + "grad_norm": 8.355647087097168, + "learning_rate": 7.670277899121223e-07, + "loss": 0.3158, + "step": 1667 + }, + { + "epoch": 1.641732283464567, + "grad_norm": 13.903353691101074, + "learning_rate": 7.663785339248683e-07, + "loss": 0.7745, + "step": 1668 + }, + { + "epoch": 1.6427165354330708, + "grad_norm": 10.879101753234863, + "learning_rate": 7.657286513791578e-07, + "loss": 0.5706, + "step": 1669 + }, + { + "epoch": 1.6437007874015748, + "grad_norm": 10.882110595703125, + "learning_rate": 7.650781438132276e-07, + "loss": 0.4708, + "step": 1670 + }, + { + "epoch": 1.6446850393700787, + "grad_norm": 18.709218978881836, + "learning_rate": 7.644270127667932e-07, + "loss": 1.0001, + "step": 1671 + }, + { + "epoch": 1.6456692913385826, + "grad_norm": 10.047717094421387, + "learning_rate": 7.637752597810456e-07, + "loss": 0.3657, + "step": 1672 + }, + { + "epoch": 1.6466535433070866, + "grad_norm": 11.821809768676758, + "learning_rate": 7.631228863986491e-07, + "loss": 0.5173, + "step": 1673 + }, + { + "epoch": 1.6476377952755905, + "grad_norm": 13.471353530883789, + "learning_rate": 7.624698941637352e-07, + "loss": 0.584, + "step": 1674 + }, + { + "epoch": 1.6486220472440944, + "grad_norm": 10.301562309265137, + "learning_rate": 7.618162846219006e-07, + "loss": 0.4782, + "step": 1675 + }, + { + "epoch": 1.6496062992125984, + "grad_norm": 14.927230834960938, + "learning_rate": 7.611620593202037e-07, + "loss": 1.1817, + "step": 1676 + }, + { + "epoch": 1.6505905511811023, + "grad_norm": 10.375627517700195, + "learning_rate": 7.605072198071592e-07, + "loss": 0.5823, + "step": 1677 + }, + { + "epoch": 1.6515748031496063, + "grad_norm": 11.692891120910645, + "learning_rate": 7.598517676327368e-07, + "loss": 0.6579, + "step": 1678 + }, + { + "epoch": 1.6525590551181102, + "grad_norm": 9.509578704833984, + "learning_rate": 7.591957043483555e-07, + "loss": 0.6332, + "step": 1679 + }, + { + "epoch": 1.6535433070866141, + "grad_norm": 14.366110801696777, + "learning_rate": 7.585390315068812e-07, + "loss": 0.8395, + "step": 1680 + }, + { + "epoch": 1.654527559055118, + "grad_norm": 9.950905799865723, + "learning_rate": 7.578817506626224e-07, + "loss": 0.486, + "step": 1681 + }, + { + "epoch": 1.655511811023622, + "grad_norm": 15.582627296447754, + "learning_rate": 7.572238633713268e-07, + "loss": 1.1842, + "step": 1682 + }, + { + "epoch": 1.656496062992126, + "grad_norm": 12.522905349731445, + "learning_rate": 7.565653711901778e-07, + "loss": 0.5918, + "step": 1683 + }, + { + "epoch": 1.656496062992126, + "eval_Qnli-dev_cosine_accuracy": 0.65625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8509283065795898, + "eval_Qnli-dev_cosine_ap": 0.681054790098263, + "eval_Qnli-dev_cosine_f1": 0.6841243862520459, + "eval_Qnli-dev_cosine_f1_threshold": 0.7551895976066589, + "eval_Qnli-dev_cosine_precision": 0.5573333333333333, + "eval_Qnli-dev_cosine_recall": 0.885593220338983, + "eval_Qnli-dev_dot_accuracy": 0.66015625, + "eval_Qnli-dev_dot_accuracy_threshold": 645.9811401367188, + "eval_Qnli-dev_dot_ap": 0.679454111760124, + "eval_Qnli-dev_dot_f1": 0.6819672131147542, + "eval_Qnli-dev_dot_f1_threshold": 575.6328125, + "eval_Qnli-dev_dot_precision": 0.5561497326203209, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.65625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.084193229675293, + "eval_Qnli-dev_euclidean_ap": 0.6815843097482049, + "eval_Qnli-dev_euclidean_f1": 0.6862745098039215, + "eval_Qnli-dev_euclidean_f1_threshold": 19.3087158203125, + "eval_Qnli-dev_euclidean_precision": 0.5585106382978723, + "eval_Qnli-dev_euclidean_recall": 0.8898305084745762, + "eval_Qnli-dev_manhattan_accuracy": 0.666015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 373.87567138671875, + "eval_Qnli-dev_manhattan_ap": 0.6904196124211742, + "eval_Qnli-dev_manhattan_f1": 0.6850828729281769, + "eval_Qnli-dev_manhattan_f1_threshold": 374.1050109863281, + "eval_Qnli-dev_manhattan_precision": 0.6058631921824105, + "eval_Qnli-dev_manhattan_recall": 0.788135593220339, + "eval_Qnli-dev_max_accuracy": 0.666015625, + "eval_Qnli-dev_max_accuracy_threshold": 645.9811401367188, + "eval_Qnli-dev_max_ap": 0.6904196124211742, + "eval_Qnli-dev_max_f1": 0.6862745098039215, + "eval_Qnli-dev_max_f1_threshold": 575.6328125, + "eval_Qnli-dev_max_precision": 0.6058631921824105, + "eval_Qnli-dev_max_recall": 0.8898305084745762, + "eval_allNLI-dev_cosine_accuracy": 0.693359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9375313520431519, + "eval_allNLI-dev_cosine_ap": 0.5106322806856907, + "eval_allNLI-dev_cosine_f1": 0.5408163265306122, + "eval_allNLI-dev_cosine_f1_threshold": 0.8662025928497314, + "eval_allNLI-dev_cosine_precision": 0.4840182648401826, + "eval_allNLI-dev_cosine_recall": 0.6127167630057804, + "eval_allNLI-dev_dot_accuracy": 0.693359375, + "eval_allNLI-dev_dot_accuracy_threshold": 714.4664916992188, + "eval_allNLI-dev_dot_ap": 0.5118205711755635, + "eval_allNLI-dev_dot_f1": 0.5433255269320844, + "eval_allNLI-dev_dot_f1_threshold": 653.50634765625, + "eval_allNLI-dev_dot_precision": 0.4566929133858268, + "eval_allNLI-dev_dot_recall": 0.6705202312138728, + "eval_allNLI-dev_euclidean_accuracy": 0.693359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.759525299072266, + "eval_allNLI-dev_euclidean_ap": 0.5108767134241021, + "eval_allNLI-dev_euclidean_f1": 0.5409429280397022, + "eval_allNLI-dev_euclidean_f1_threshold": 14.420662879943848, + "eval_allNLI-dev_euclidean_precision": 0.47391304347826085, + "eval_allNLI-dev_euclidean_recall": 0.630057803468208, + "eval_allNLI-dev_manhattan_accuracy": 0.693359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 237.72364807128906, + "eval_allNLI-dev_manhattan_ap": 0.513985978728926, + "eval_allNLI-dev_manhattan_f1": 0.5611510791366907, + "eval_allNLI-dev_manhattan_f1_threshold": 307.5692138671875, + "eval_allNLI-dev_manhattan_precision": 0.47950819672131145, + "eval_allNLI-dev_manhattan_recall": 0.6763005780346821, + "eval_allNLI-dev_max_accuracy": 0.693359375, + "eval_allNLI-dev_max_accuracy_threshold": 714.4664916992188, + "eval_allNLI-dev_max_ap": 0.513985978728926, + "eval_allNLI-dev_max_f1": 0.5611510791366907, + "eval_allNLI-dev_max_f1_threshold": 653.50634765625, + "eval_allNLI-dev_max_precision": 0.4840182648401826, + "eval_allNLI-dev_max_recall": 0.6763005780346821, + "eval_loss": 0.7026398777961731, + "eval_runtime": 51.2642, + "eval_samples_per_second": 32.459, + "eval_sequential_score": 0.6904196124211742, + "eval_steps_per_second": 0.137, + "eval_sts-test_pearson_cosine": 0.6378753872048569, + "eval_sts-test_pearson_dot": 0.6379468233147109, + "eval_sts-test_pearson_euclidean": 0.6579907313824014, + "eval_sts-test_pearson_manhattan": 0.6641812458202475, + "eval_sts-test_pearson_max": 0.6641812458202475, + "eval_sts-test_spearman_cosine": 0.6598976701026994, + "eval_sts-test_spearman_dot": 0.6593824040790388, + "eval_sts-test_spearman_euclidean": 0.6600905480392965, + "eval_sts-test_spearman_manhattan": 0.6637495454244414, + "eval_sts-test_spearman_max": 0.6637495454244414, + "step": 1683 + }, + { + "epoch": 1.65748031496063, + "grad_norm": 11.333374977111816, + "learning_rate": 7.559062756777898e-07, + "loss": 0.4818, + "step": 1684 + }, + { + "epoch": 1.6584645669291338, + "grad_norm": 6.981270790100098, + "learning_rate": 7.552465783942058e-07, + "loss": 0.2778, + "step": 1685 + }, + { + "epoch": 1.659448818897638, + "grad_norm": 5.901653289794922, + "learning_rate": 7.545862809008931e-07, + "loss": 0.2085, + "step": 1686 + }, + { + "epoch": 1.6604330708661417, + "grad_norm": 11.936545372009277, + "learning_rate": 7.539253847607397e-07, + "loss": 0.825, + "step": 1687 + }, + { + "epoch": 1.6614173228346458, + "grad_norm": 8.547285079956055, + "learning_rate": 7.532638915380503e-07, + "loss": 0.3304, + "step": 1688 + }, + { + "epoch": 1.6624015748031495, + "grad_norm": 12.993122100830078, + "learning_rate": 7.526018027985434e-07, + "loss": 1.0776, + "step": 1689 + }, + { + "epoch": 1.6633858267716537, + "grad_norm": 9.556748390197754, + "learning_rate": 7.519391201093462e-07, + "loss": 0.4255, + "step": 1690 + }, + { + "epoch": 1.6643700787401574, + "grad_norm": 5.061547756195068, + "learning_rate": 7.512758450389925e-07, + "loss": 0.1358, + "step": 1691 + }, + { + "epoch": 1.6653543307086616, + "grad_norm": 10.19187068939209, + "learning_rate": 7.50611979157418e-07, + "loss": 0.3624, + "step": 1692 + }, + { + "epoch": 1.6663385826771653, + "grad_norm": 10.900358200073242, + "learning_rate": 7.499475240359567e-07, + "loss": 0.7943, + "step": 1693 + }, + { + "epoch": 1.6673228346456694, + "grad_norm": 13.84185791015625, + "learning_rate": 7.492824812473376e-07, + "loss": 0.7367, + "step": 1694 + }, + { + "epoch": 1.6683070866141732, + "grad_norm": 20.673704147338867, + "learning_rate": 7.4861685236568e-07, + "loss": 0.9489, + "step": 1695 + }, + { + "epoch": 1.6692913385826773, + "grad_norm": 9.632235527038574, + "learning_rate": 7.479506389664914e-07, + "loss": 0.4323, + "step": 1696 + }, + { + "epoch": 1.670275590551181, + "grad_norm": 10.106782913208008, + "learning_rate": 7.47283842626662e-07, + "loss": 0.4374, + "step": 1697 + }, + { + "epoch": 1.6712598425196852, + "grad_norm": 7.663733005523682, + "learning_rate": 7.466164649244625e-07, + "loss": 0.2878, + "step": 1698 + }, + { + "epoch": 1.672244094488189, + "grad_norm": 12.01664924621582, + "learning_rate": 7.45948507439539e-07, + "loss": 0.8243, + "step": 1699 + }, + { + "epoch": 1.673228346456693, + "grad_norm": 12.359426498413086, + "learning_rate": 7.452799717529102e-07, + "loss": 0.5028, + "step": 1700 + }, + { + "epoch": 1.6742125984251968, + "grad_norm": 11.55900764465332, + "learning_rate": 7.446108594469634e-07, + "loss": 0.4705, + "step": 1701 + }, + { + "epoch": 1.675196850393701, + "grad_norm": 11.337023735046387, + "learning_rate": 7.439411721054508e-07, + "loss": 0.4235, + "step": 1702 + }, + { + "epoch": 1.6761811023622046, + "grad_norm": 8.713342666625977, + "learning_rate": 7.432709113134856e-07, + "loss": 0.355, + "step": 1703 + }, + { + "epoch": 1.6771653543307088, + "grad_norm": 12.83935546875, + "learning_rate": 7.42600078657538e-07, + "loss": 0.8231, + "step": 1704 + }, + { + "epoch": 1.6781496062992125, + "grad_norm": 10.76926326751709, + "learning_rate": 7.419286757254324e-07, + "loss": 0.5603, + "step": 1705 + }, + { + "epoch": 1.6791338582677167, + "grad_norm": 3.749600887298584, + "learning_rate": 7.412567041063424e-07, + "loss": 0.0942, + "step": 1706 + }, + { + "epoch": 1.6801181102362204, + "grad_norm": 11.84522819519043, + "learning_rate": 7.40584165390788e-07, + "loss": 0.463, + "step": 1707 + }, + { + "epoch": 1.6811023622047245, + "grad_norm": 6.267951965332031, + "learning_rate": 7.399110611706316e-07, + "loss": 0.153, + "step": 1708 + }, + { + "epoch": 1.6820866141732282, + "grad_norm": 12.05251407623291, + "learning_rate": 7.392373930390738e-07, + "loss": 0.8333, + "step": 1709 + }, + { + "epoch": 1.6830708661417324, + "grad_norm": 8.769957542419434, + "learning_rate": 7.3856316259065e-07, + "loss": 0.3173, + "step": 1710 + }, + { + "epoch": 1.6840551181102361, + "grad_norm": 11.866000175476074, + "learning_rate": 7.378883714212268e-07, + "loss": 0.6905, + "step": 1711 + }, + { + "epoch": 1.6850393700787403, + "grad_norm": 12.132401466369629, + "learning_rate": 7.372130211279977e-07, + "loss": 0.5572, + "step": 1712 + }, + { + "epoch": 1.686023622047244, + "grad_norm": 7.610491752624512, + "learning_rate": 7.365371133094799e-07, + "loss": 0.2381, + "step": 1713 + }, + { + "epoch": 1.6870078740157481, + "grad_norm": 13.988032341003418, + "learning_rate": 7.3586064956551e-07, + "loss": 0.9946, + "step": 1714 + }, + { + "epoch": 1.6879921259842519, + "grad_norm": 10.767354011535645, + "learning_rate": 7.351836314972408e-07, + "loss": 0.5725, + "step": 1715 + }, + { + "epoch": 1.688976377952756, + "grad_norm": 7.5194172859191895, + "learning_rate": 7.345060607071365e-07, + "loss": 0.2278, + "step": 1716 + }, + { + "epoch": 1.6899606299212597, + "grad_norm": 10.975831031799316, + "learning_rate": 7.338279387989703e-07, + "loss": 0.5617, + "step": 1717 + }, + { + "epoch": 1.6909448818897639, + "grad_norm": 20.18640899658203, + "learning_rate": 7.331492673778193e-07, + "loss": 1.071, + "step": 1718 + }, + { + "epoch": 1.6919291338582676, + "grad_norm": 26.24131202697754, + "learning_rate": 7.324700480500615e-07, + "loss": 1.9854, + "step": 1719 + }, + { + "epoch": 1.6929133858267718, + "grad_norm": 11.812013626098633, + "learning_rate": 7.317902824233719e-07, + "loss": 0.6289, + "step": 1720 + }, + { + "epoch": 1.6938976377952755, + "grad_norm": 10.017646789550781, + "learning_rate": 7.311099721067183e-07, + "loss": 0.4823, + "step": 1721 + }, + { + "epoch": 1.6948818897637796, + "grad_norm": 5.729677677154541, + "learning_rate": 7.304291187103578e-07, + "loss": 0.2182, + "step": 1722 + }, + { + "epoch": 1.6958661417322833, + "grad_norm": 7.125960826873779, + "learning_rate": 7.297477238458333e-07, + "loss": 0.2221, + "step": 1723 + }, + { + "epoch": 1.6968503937007875, + "grad_norm": 19.171634674072266, + "learning_rate": 7.290657891259688e-07, + "loss": 0.8107, + "step": 1724 + }, + { + "epoch": 1.6978346456692912, + "grad_norm": 16.699453353881836, + "learning_rate": 7.283833161648664e-07, + "loss": 0.8205, + "step": 1725 + }, + { + "epoch": 1.6988188976377954, + "grad_norm": 8.69482421875, + "learning_rate": 7.277003065779022e-07, + "loss": 0.3824, + "step": 1726 + }, + { + "epoch": 1.699803149606299, + "grad_norm": 9.883689880371094, + "learning_rate": 7.270167619817222e-07, + "loss": 0.6142, + "step": 1727 + }, + { + "epoch": 1.7007874015748032, + "grad_norm": 15.141371726989746, + "learning_rate": 7.26332683994239e-07, + "loss": 0.7923, + "step": 1728 + }, + { + "epoch": 1.701771653543307, + "grad_norm": 9.613105773925781, + "learning_rate": 7.256480742346278e-07, + "loss": 0.6371, + "step": 1729 + }, + { + "epoch": 1.702755905511811, + "grad_norm": 18.075740814208984, + "learning_rate": 7.249629343233221e-07, + "loss": 0.9596, + "step": 1730 + }, + { + "epoch": 1.7037401574803148, + "grad_norm": 8.848851203918457, + "learning_rate": 7.242772658820105e-07, + "loss": 0.3392, + "step": 1731 + }, + { + "epoch": 1.704724409448819, + "grad_norm": 14.672804832458496, + "learning_rate": 7.235910705336327e-07, + "loss": 0.856, + "step": 1732 + }, + { + "epoch": 1.7057086614173227, + "grad_norm": 14.823372840881348, + "learning_rate": 7.229043499023753e-07, + "loss": 1.0407, + "step": 1733 + }, + { + "epoch": 1.7066929133858268, + "grad_norm": 12.815380096435547, + "learning_rate": 7.222171056136683e-07, + "loss": 0.6108, + "step": 1734 + }, + { + "epoch": 1.7076771653543306, + "grad_norm": 9.097896575927734, + "learning_rate": 7.215293392941814e-07, + "loss": 0.4197, + "step": 1735 + }, + { + "epoch": 1.7086614173228347, + "grad_norm": 14.844433784484863, + "learning_rate": 7.208410525718195e-07, + "loss": 0.945, + "step": 1736 + }, + { + "epoch": 1.7096456692913384, + "grad_norm": 7.867664813995361, + "learning_rate": 7.201522470757196e-07, + "loss": 0.2771, + "step": 1737 + }, + { + "epoch": 1.7106299212598426, + "grad_norm": 9.49281120300293, + "learning_rate": 7.194629244362462e-07, + "loss": 0.3732, + "step": 1738 + }, + { + "epoch": 1.7116141732283463, + "grad_norm": 6.025905132293701, + "learning_rate": 7.187730862849886e-07, + "loss": 0.249, + "step": 1739 + }, + { + "epoch": 1.7125984251968505, + "grad_norm": 9.20777416229248, + "learning_rate": 7.180827342547556e-07, + "loss": 0.331, + "step": 1740 + }, + { + "epoch": 1.7135826771653542, + "grad_norm": 12.778132438659668, + "learning_rate": 7.173918699795726e-07, + "loss": 0.6882, + "step": 1741 + }, + { + "epoch": 1.7145669291338583, + "grad_norm": 8.081833839416504, + "learning_rate": 7.167004950946774e-07, + "loss": 0.3609, + "step": 1742 + }, + { + "epoch": 1.715551181102362, + "grad_norm": 9.465290069580078, + "learning_rate": 7.160086112365163e-07, + "loss": 0.3878, + "step": 1743 + }, + { + "epoch": 1.7165354330708662, + "grad_norm": 8.498932838439941, + "learning_rate": 7.153162200427404e-07, + "loss": 0.3484, + "step": 1744 + }, + { + "epoch": 1.71751968503937, + "grad_norm": 16.855077743530273, + "learning_rate": 7.146233231522018e-07, + "loss": 1.1531, + "step": 1745 + }, + { + "epoch": 1.718503937007874, + "grad_norm": 8.619532585144043, + "learning_rate": 7.139299222049493e-07, + "loss": 0.486, + "step": 1746 + }, + { + "epoch": 1.719488188976378, + "grad_norm": 8.059650421142578, + "learning_rate": 7.132360188422249e-07, + "loss": 0.2869, + "step": 1747 + }, + { + "epoch": 1.720472440944882, + "grad_norm": 4.437140464782715, + "learning_rate": 7.125416147064597e-07, + "loss": 0.1983, + "step": 1748 + }, + { + "epoch": 1.7214566929133859, + "grad_norm": 9.482026100158691, + "learning_rate": 7.118467114412702e-07, + "loss": 0.377, + "step": 1749 + }, + { + "epoch": 1.7224409448818898, + "grad_norm": 6.485823631286621, + "learning_rate": 7.111513106914544e-07, + "loss": 0.373, + "step": 1750 + }, + { + "epoch": 1.7234251968503937, + "grad_norm": 13.507378578186035, + "learning_rate": 7.104554141029874e-07, + "loss": 0.7411, + "step": 1751 + }, + { + "epoch": 1.7244094488188977, + "grad_norm": 8.923224449157715, + "learning_rate": 7.097590233230186e-07, + "loss": 0.3041, + "step": 1752 + }, + { + "epoch": 1.7253937007874016, + "grad_norm": 7.605806350708008, + "learning_rate": 7.090621399998665e-07, + "loss": 0.3193, + "step": 1753 + }, + { + "epoch": 1.7263779527559056, + "grad_norm": 4.4992218017578125, + "learning_rate": 7.083647657830155e-07, + "loss": 0.1406, + "step": 1754 + }, + { + "epoch": 1.7273622047244095, + "grad_norm": 7.35247278213501, + "learning_rate": 7.076669023231126e-07, + "loss": 0.3105, + "step": 1755 + }, + { + "epoch": 1.7283464566929134, + "grad_norm": 7.219346523284912, + "learning_rate": 7.069685512719619e-07, + "loss": 0.365, + "step": 1756 + }, + { + "epoch": 1.7293307086614174, + "grad_norm": 15.006402015686035, + "learning_rate": 7.06269714282522e-07, + "loss": 0.478, + "step": 1757 + }, + { + "epoch": 1.7303149606299213, + "grad_norm": 21.270875930786133, + "learning_rate": 7.055703930089018e-07, + "loss": 1.0871, + "step": 1758 + }, + { + "epoch": 1.7312992125984252, + "grad_norm": 9.952723503112793, + "learning_rate": 7.048705891063562e-07, + "loss": 0.4599, + "step": 1759 + }, + { + "epoch": 1.7322834645669292, + "grad_norm": 6.753321170806885, + "learning_rate": 7.04170304231283e-07, + "loss": 0.2112, + "step": 1760 + }, + { + "epoch": 1.733267716535433, + "grad_norm": 12.478215217590332, + "learning_rate": 7.034695400412177e-07, + "loss": 0.6315, + "step": 1761 + }, + { + "epoch": 1.734251968503937, + "grad_norm": 13.235335350036621, + "learning_rate": 7.027682981948306e-07, + "loss": 0.7519, + "step": 1762 + }, + { + "epoch": 1.735236220472441, + "grad_norm": 22.36837387084961, + "learning_rate": 7.020665803519229e-07, + "loss": 1.1257, + "step": 1763 + }, + { + "epoch": 1.736220472440945, + "grad_norm": 9.113337516784668, + "learning_rate": 7.013643881734222e-07, + "loss": 0.4301, + "step": 1764 + }, + { + "epoch": 1.7372047244094488, + "grad_norm": 10.880926132202148, + "learning_rate": 7.006617233213787e-07, + "loss": 0.5459, + "step": 1765 + }, + { + "epoch": 1.7381889763779528, + "grad_norm": 12.829588890075684, + "learning_rate": 6.999585874589616e-07, + "loss": 0.688, + "step": 1766 + }, + { + "epoch": 1.7391732283464567, + "grad_norm": 12.976068496704102, + "learning_rate": 6.992549822504548e-07, + "loss": 0.7211, + "step": 1767 + }, + { + "epoch": 1.7401574803149606, + "grad_norm": 13.932178497314453, + "learning_rate": 6.985509093612534e-07, + "loss": 0.6589, + "step": 1768 + }, + { + "epoch": 1.7411417322834646, + "grad_norm": 9.614189147949219, + "learning_rate": 6.978463704578589e-07, + "loss": 0.3702, + "step": 1769 + }, + { + "epoch": 1.7421259842519685, + "grad_norm": 6.6707258224487305, + "learning_rate": 6.971413672078766e-07, + "loss": 0.2318, + "step": 1770 + }, + { + "epoch": 1.7431102362204725, + "grad_norm": 11.708477020263672, + "learning_rate": 6.9643590128001e-07, + "loss": 0.3045, + "step": 1771 + }, + { + "epoch": 1.7440944881889764, + "grad_norm": 9.63950252532959, + "learning_rate": 6.957299743440589e-07, + "loss": 0.381, + "step": 1772 + }, + { + "epoch": 1.7450787401574803, + "grad_norm": 18.139057159423828, + "learning_rate": 6.950235880709131e-07, + "loss": 0.984, + "step": 1773 + }, + { + "epoch": 1.7460629921259843, + "grad_norm": 11.944750785827637, + "learning_rate": 6.943167441325504e-07, + "loss": 0.4269, + "step": 1774 + }, + { + "epoch": 1.7470472440944882, + "grad_norm": 13.353084564208984, + "learning_rate": 6.936094442020314e-07, + "loss": 0.4809, + "step": 1775 + }, + { + "epoch": 1.7480314960629921, + "grad_norm": 10.593157768249512, + "learning_rate": 6.929016899534963e-07, + "loss": 0.5388, + "step": 1776 + }, + { + "epoch": 1.749015748031496, + "grad_norm": 14.205294609069824, + "learning_rate": 6.921934830621603e-07, + "loss": 1.0426, + "step": 1777 + }, + { + "epoch": 1.75, + "grad_norm": 12.602083206176758, + "learning_rate": 6.914848252043106e-07, + "loss": 0.4852, + "step": 1778 + }, + { + "epoch": 1.750984251968504, + "grad_norm": 17.39359474182129, + "learning_rate": 6.907757180573013e-07, + "loss": 0.6473, + "step": 1779 + }, + { + "epoch": 1.7519685039370079, + "grad_norm": 10.737658500671387, + "learning_rate": 6.9006616329955e-07, + "loss": 0.4637, + "step": 1780 + }, + { + "epoch": 1.7529527559055118, + "grad_norm": 16.11721420288086, + "learning_rate": 6.893561626105338e-07, + "loss": 0.7861, + "step": 1781 + }, + { + "epoch": 1.7539370078740157, + "grad_norm": 17.54054832458496, + "learning_rate": 6.886457176707855e-07, + "loss": 0.9019, + "step": 1782 + }, + { + "epoch": 1.7549212598425197, + "grad_norm": 10.458827018737793, + "learning_rate": 6.879348301618894e-07, + "loss": 0.5938, + "step": 1783 + }, + { + "epoch": 1.7559055118110236, + "grad_norm": 10.137784004211426, + "learning_rate": 6.872235017664769e-07, + "loss": 0.4754, + "step": 1784 + }, + { + "epoch": 1.7568897637795275, + "grad_norm": 15.559357643127441, + "learning_rate": 6.865117341682234e-07, + "loss": 1.0763, + "step": 1785 + }, + { + "epoch": 1.7578740157480315, + "grad_norm": 8.916090965270996, + "learning_rate": 6.857995290518439e-07, + "loss": 0.2995, + "step": 1786 + }, + { + "epoch": 1.7588582677165354, + "grad_norm": 8.638309478759766, + "learning_rate": 6.850868881030883e-07, + "loss": 0.535, + "step": 1787 + }, + { + "epoch": 1.7598425196850394, + "grad_norm": 8.75821304321289, + "learning_rate": 6.843738130087393e-07, + "loss": 0.3883, + "step": 1788 + }, + { + "epoch": 1.7608267716535433, + "grad_norm": 9.704635620117188, + "learning_rate": 6.836603054566059e-07, + "loss": 0.3746, + "step": 1789 + }, + { + "epoch": 1.7618110236220472, + "grad_norm": 12.514015197753906, + "learning_rate": 6.829463671355216e-07, + "loss": 0.7878, + "step": 1790 + }, + { + "epoch": 1.7627952755905512, + "grad_norm": 7.831786632537842, + "learning_rate": 6.822319997353391e-07, + "loss": 0.334, + "step": 1791 + }, + { + "epoch": 1.763779527559055, + "grad_norm": 13.068495750427246, + "learning_rate": 6.815172049469268e-07, + "loss": 0.9061, + "step": 1792 + }, + { + "epoch": 1.764763779527559, + "grad_norm": 12.966577529907227, + "learning_rate": 6.808019844621649e-07, + "loss": 0.5657, + "step": 1793 + }, + { + "epoch": 1.765748031496063, + "grad_norm": 15.134439468383789, + "learning_rate": 6.800863399739409e-07, + "loss": 0.8639, + "step": 1794 + }, + { + "epoch": 1.766732283464567, + "grad_norm": 10.75047492980957, + "learning_rate": 6.793702731761459e-07, + "loss": 0.648, + "step": 1795 + }, + { + "epoch": 1.7677165354330708, + "grad_norm": 10.715109825134277, + "learning_rate": 6.786537857636707e-07, + "loss": 0.5737, + "step": 1796 + }, + { + "epoch": 1.7687007874015748, + "grad_norm": 14.109840393066406, + "learning_rate": 6.77936879432402e-07, + "loss": 0.6318, + "step": 1797 + }, + { + "epoch": 1.7696850393700787, + "grad_norm": 16.420080184936523, + "learning_rate": 6.772195558792175e-07, + "loss": 1.003, + "step": 1798 + }, + { + "epoch": 1.7706692913385826, + "grad_norm": 12.585822105407715, + "learning_rate": 6.765018168019825e-07, + "loss": 0.7127, + "step": 1799 + }, + { + "epoch": 1.7716535433070866, + "grad_norm": 8.718441009521484, + "learning_rate": 6.757836638995461e-07, + "loss": 0.33, + "step": 1800 + }, + { + "epoch": 1.7726377952755905, + "grad_norm": 12.439270973205566, + "learning_rate": 6.75065098871737e-07, + "loss": 0.6338, + "step": 1801 + }, + { + "epoch": 1.7736220472440944, + "grad_norm": 16.2679386138916, + "learning_rate": 6.74346123419359e-07, + "loss": 0.8115, + "step": 1802 + }, + { + "epoch": 1.7746062992125984, + "grad_norm": 11.987645149230957, + "learning_rate": 6.736267392441874e-07, + "loss": 0.5626, + "step": 1803 + }, + { + "epoch": 1.7755905511811023, + "grad_norm": 12.127118110656738, + "learning_rate": 6.729069480489652e-07, + "loss": 0.8565, + "step": 1804 + }, + { + "epoch": 1.7765748031496063, + "grad_norm": 11.7110595703125, + "learning_rate": 6.721867515373987e-07, + "loss": 0.6183, + "step": 1805 + }, + { + "epoch": 1.7775590551181102, + "grad_norm": 9.546368598937988, + "learning_rate": 6.714661514141535e-07, + "loss": 0.4151, + "step": 1806 + }, + { + "epoch": 1.7785433070866141, + "grad_norm": 13.251776695251465, + "learning_rate": 6.707451493848502e-07, + "loss": 0.8111, + "step": 1807 + }, + { + "epoch": 1.779527559055118, + "grad_norm": 6.505370140075684, + "learning_rate": 6.700237471560614e-07, + "loss": 0.2495, + "step": 1808 + }, + { + "epoch": 1.780511811023622, + "grad_norm": 10.462722778320312, + "learning_rate": 6.693019464353065e-07, + "loss": 0.5267, + "step": 1809 + }, + { + "epoch": 1.781496062992126, + "grad_norm": 9.609010696411133, + "learning_rate": 6.685797489310481e-07, + "loss": 0.4687, + "step": 1810 + }, + { + "epoch": 1.78248031496063, + "grad_norm": 12.09762191772461, + "learning_rate": 6.67857156352688e-07, + "loss": 0.6576, + "step": 1811 + }, + { + "epoch": 1.7834645669291338, + "grad_norm": 9.396162033081055, + "learning_rate": 6.671341704105632e-07, + "loss": 0.4489, + "step": 1812 + }, + { + "epoch": 1.784448818897638, + "grad_norm": 6.937375545501709, + "learning_rate": 6.664107928159419e-07, + "loss": 0.2523, + "step": 1813 + }, + { + "epoch": 1.7854330708661417, + "grad_norm": 17.318628311157227, + "learning_rate": 6.656870252810187e-07, + "loss": 1.0483, + "step": 1814 + }, + { + "epoch": 1.7864173228346458, + "grad_norm": 6.78587007522583, + "learning_rate": 6.649628695189123e-07, + "loss": 0.3089, + "step": 1815 + }, + { + "epoch": 1.7874015748031495, + "grad_norm": 19.762271881103516, + "learning_rate": 6.642383272436589e-07, + "loss": 1.2254, + "step": 1816 + }, + { + "epoch": 1.7883858267716537, + "grad_norm": 11.951211929321289, + "learning_rate": 6.635134001702107e-07, + "loss": 0.8332, + "step": 1817 + }, + { + "epoch": 1.7893700787401574, + "grad_norm": 8.019990921020508, + "learning_rate": 6.627880900144303e-07, + "loss": 0.4108, + "step": 1818 + }, + { + "epoch": 1.7903543307086616, + "grad_norm": 8.551742553710938, + "learning_rate": 6.620623984930869e-07, + "loss": 0.3647, + "step": 1819 + }, + { + "epoch": 1.7913385826771653, + "grad_norm": 5.962251663208008, + "learning_rate": 6.613363273238527e-07, + "loss": 0.2217, + "step": 1820 + }, + { + "epoch": 1.7923228346456694, + "grad_norm": 14.370640754699707, + "learning_rate": 6.606098782252978e-07, + "loss": 0.5796, + "step": 1821 + }, + { + "epoch": 1.7933070866141732, + "grad_norm": 9.894375801086426, + "learning_rate": 6.598830529168877e-07, + "loss": 0.5114, + "step": 1822 + }, + { + "epoch": 1.7942913385826773, + "grad_norm": 11.380830764770508, + "learning_rate": 6.591558531189781e-07, + "loss": 0.2156, + "step": 1823 + }, + { + "epoch": 1.795275590551181, + "grad_norm": 9.474596977233887, + "learning_rate": 6.584282805528106e-07, + "loss": 0.4107, + "step": 1824 + }, + { + "epoch": 1.7962598425196852, + "grad_norm": 20.004667282104492, + "learning_rate": 6.577003369405099e-07, + "loss": 0.9586, + "step": 1825 + }, + { + "epoch": 1.797244094488189, + "grad_norm": 5.846341133117676, + "learning_rate": 6.569720240050784e-07, + "loss": 0.1446, + "step": 1826 + }, + { + "epoch": 1.798228346456693, + "grad_norm": 3.7557458877563477, + "learning_rate": 6.562433434703928e-07, + "loss": 0.0991, + "step": 1827 + }, + { + "epoch": 1.7992125984251968, + "grad_norm": 10.2771577835083, + "learning_rate": 6.555142970612001e-07, + "loss": 0.472, + "step": 1828 + }, + { + "epoch": 1.800196850393701, + "grad_norm": 12.053825378417969, + "learning_rate": 6.54784886503113e-07, + "loss": 0.6334, + "step": 1829 + }, + { + "epoch": 1.8011811023622046, + "grad_norm": 15.814884185791016, + "learning_rate": 6.540551135226061e-07, + "loss": 0.8573, + "step": 1830 + }, + { + "epoch": 1.8021653543307088, + "grad_norm": 10.095357894897461, + "learning_rate": 6.533249798470123e-07, + "loss": 0.4909, + "step": 1831 + }, + { + "epoch": 1.8031496062992125, + "grad_norm": 7.293158531188965, + "learning_rate": 6.525944872045175e-07, + "loss": 0.2107, + "step": 1832 + }, + { + "epoch": 1.8041338582677167, + "grad_norm": 8.286677360534668, + "learning_rate": 6.51863637324158e-07, + "loss": 0.3745, + "step": 1833 + }, + { + "epoch": 1.8051181102362204, + "grad_norm": 7.183446884155273, + "learning_rate": 6.511324319358154e-07, + "loss": 0.2491, + "step": 1834 + }, + { + "epoch": 1.8061023622047245, + "grad_norm": 19.922443389892578, + "learning_rate": 6.504008727702123e-07, + "loss": 1.4215, + "step": 1835 + }, + { + "epoch": 1.8070866141732282, + "grad_norm": 14.210881233215332, + "learning_rate": 6.496689615589096e-07, + "loss": 0.5273, + "step": 1836 + }, + { + "epoch": 1.8070866141732282, + "eval_Qnli-dev_cosine_accuracy": 0.65625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7958178520202637, + "eval_Qnli-dev_cosine_ap": 0.6889326859706882, + "eval_Qnli-dev_cosine_f1": 0.6853582554517135, + "eval_Qnli-dev_cosine_f1_threshold": 0.716392457485199, + "eval_Qnli-dev_cosine_precision": 0.541871921182266, + "eval_Qnli-dev_cosine_recall": 0.9322033898305084, + "eval_Qnli-dev_dot_accuracy": 0.654296875, + "eval_Qnli-dev_dot_accuracy_threshold": 613.7510375976562, + "eval_Qnli-dev_dot_ap": 0.6872149178601934, + "eval_Qnli-dev_dot_f1": 0.6842923794712287, + "eval_Qnli-dev_dot_f1_threshold": 545.7647094726562, + "eval_Qnli-dev_dot_precision": 0.5405405405405406, + "eval_Qnli-dev_dot_recall": 0.9322033898305084, + "eval_Qnli-dev_euclidean_accuracy": 0.65625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.662763595581055, + "eval_Qnli-dev_euclidean_ap": 0.6889310064166702, + "eval_Qnli-dev_euclidean_f1": 0.6853582554517135, + "eval_Qnli-dev_euclidean_f1_threshold": 20.800479888916016, + "eval_Qnli-dev_euclidean_precision": 0.541871921182266, + "eval_Qnli-dev_euclidean_recall": 0.9322033898305084, + "eval_Qnli-dev_manhattan_accuracy": 0.66015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 368.34844970703125, + "eval_Qnli-dev_manhattan_ap": 0.697194420363898, + "eval_Qnli-dev_manhattan_f1": 0.6830065359477124, + "eval_Qnli-dev_manhattan_f1_threshold": 423.89007568359375, + "eval_Qnli-dev_manhattan_precision": 0.5558510638297872, + "eval_Qnli-dev_manhattan_recall": 0.885593220338983, + "eval_Qnli-dev_max_accuracy": 0.66015625, + "eval_Qnli-dev_max_accuracy_threshold": 613.7510375976562, + "eval_Qnli-dev_max_ap": 0.697194420363898, + "eval_Qnli-dev_max_f1": 0.6853582554517135, + "eval_Qnli-dev_max_f1_threshold": 545.7647094726562, + "eval_Qnli-dev_max_precision": 0.5558510638297872, + "eval_Qnli-dev_max_recall": 0.9322033898305084, + "eval_allNLI-dev_cosine_accuracy": 0.693359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9289067983627319, + "eval_allNLI-dev_cosine_ap": 0.5208760455583072, + "eval_allNLI-dev_cosine_f1": 0.5569007263922519, + "eval_allNLI-dev_cosine_f1_threshold": 0.8536206483840942, + "eval_allNLI-dev_cosine_precision": 0.4791666666666667, + "eval_allNLI-dev_cosine_recall": 0.6647398843930635, + "eval_allNLI-dev_dot_accuracy": 0.6953125, + "eval_allNLI-dev_dot_accuracy_threshold": 710.2685546875, + "eval_allNLI-dev_dot_ap": 0.519996450033682, + "eval_allNLI-dev_dot_f1": 0.5528846153846153, + "eval_allNLI-dev_dot_f1_threshold": 651.2723388671875, + "eval_allNLI-dev_dot_precision": 0.4732510288065844, + "eval_allNLI-dev_dot_recall": 0.6647398843930635, + "eval_allNLI-dev_euclidean_accuracy": 0.693359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.431549072265625, + "eval_allNLI-dev_euclidean_ap": 0.5214983555266551, + "eval_allNLI-dev_euclidean_f1": 0.5555555555555556, + "eval_allNLI-dev_euclidean_f1_threshold": 14.99692153930664, + "eval_allNLI-dev_euclidean_precision": 0.47717842323651455, + "eval_allNLI-dev_euclidean_recall": 0.6647398843930635, + "eval_allNLI-dev_manhattan_accuracy": 0.693359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 223.4080047607422, + "eval_allNLI-dev_manhattan_ap": 0.5226497094892681, + "eval_allNLI-dev_manhattan_f1": 0.5613207547169812, + "eval_allNLI-dev_manhattan_f1_threshold": 321.18707275390625, + "eval_allNLI-dev_manhattan_precision": 0.47410358565737054, + "eval_allNLI-dev_manhattan_recall": 0.6878612716763006, + "eval_allNLI-dev_max_accuracy": 0.6953125, + "eval_allNLI-dev_max_accuracy_threshold": 710.2685546875, + "eval_allNLI-dev_max_ap": 0.5226497094892681, + "eval_allNLI-dev_max_f1": 0.5613207547169812, + "eval_allNLI-dev_max_f1_threshold": 651.2723388671875, + "eval_allNLI-dev_max_precision": 0.4791666666666667, + "eval_allNLI-dev_max_recall": 0.6878612716763006, + "eval_loss": 0.6583033204078674, + "eval_runtime": 51.5088, + "eval_samples_per_second": 32.305, + "eval_sequential_score": 0.697194420363898, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.6482615589072106, + "eval_sts-test_pearson_dot": 0.6469131583986065, + "eval_sts-test_pearson_euclidean": 0.6690645125650454, + "eval_sts-test_pearson_manhattan": 0.6718954949364839, + "eval_sts-test_pearson_max": 0.6718954949364839, + "eval_sts-test_spearman_cosine": 0.668762134212892, + "eval_sts-test_spearman_dot": 0.6674914668969447, + "eval_sts-test_spearman_euclidean": 0.6691267674907092, + "eval_sts-test_spearman_manhattan": 0.6709370902677755, + "eval_sts-test_spearman_max": 0.6709370902677755, + "step": 1836 + }, + { + "epoch": 1.8080708661417324, + "grad_norm": 10.738568305969238, + "learning_rate": 6.489367000343005e-07, + "loss": 0.4595, + "step": 1837 + }, + { + "epoch": 1.8090551181102361, + "grad_norm": 15.320980072021484, + "learning_rate": 6.482040899296081e-07, + "loss": 0.7292, + "step": 1838 + }, + { + "epoch": 1.8100393700787403, + "grad_norm": 7.67435884475708, + "learning_rate": 6.474711329788802e-07, + "loss": 0.2475, + "step": 1839 + }, + { + "epoch": 1.811023622047244, + "grad_norm": 5.039150714874268, + "learning_rate": 6.467378309169858e-07, + "loss": 0.1717, + "step": 1840 + }, + { + "epoch": 1.8120078740157481, + "grad_norm": 11.503175735473633, + "learning_rate": 6.460041854796105e-07, + "loss": 0.484, + "step": 1841 + }, + { + "epoch": 1.8129921259842519, + "grad_norm": 11.132969856262207, + "learning_rate": 6.452701984032526e-07, + "loss": 0.5428, + "step": 1842 + }, + { + "epoch": 1.813976377952756, + "grad_norm": 11.278051376342773, + "learning_rate": 6.445358714252195e-07, + "loss": 0.5466, + "step": 1843 + }, + { + "epoch": 1.8149606299212597, + "grad_norm": 9.864574432373047, + "learning_rate": 6.438012062836226e-07, + "loss": 0.4286, + "step": 1844 + }, + { + "epoch": 1.8159448818897639, + "grad_norm": 14.751577377319336, + "learning_rate": 6.430662047173742e-07, + "loss": 0.7834, + "step": 1845 + }, + { + "epoch": 1.8169291338582676, + "grad_norm": 14.805031776428223, + "learning_rate": 6.423308684661825e-07, + "loss": 0.9017, + "step": 1846 + }, + { + "epoch": 1.8179133858267718, + "grad_norm": 13.299103736877441, + "learning_rate": 6.415951992705482e-07, + "loss": 0.6578, + "step": 1847 + }, + { + "epoch": 1.8188976377952755, + "grad_norm": 10.048017501831055, + "learning_rate": 6.408591988717594e-07, + "loss": 0.3558, + "step": 1848 + }, + { + "epoch": 1.8198818897637796, + "grad_norm": 10.021560668945312, + "learning_rate": 6.401228690118893e-07, + "loss": 0.4255, + "step": 1849 + }, + { + "epoch": 1.8208661417322833, + "grad_norm": 5.557498931884766, + "learning_rate": 6.3938621143379e-07, + "loss": 0.1765, + "step": 1850 + }, + { + "epoch": 1.8218503937007875, + "grad_norm": 10.919778823852539, + "learning_rate": 6.386492278810894e-07, + "loss": 0.4543, + "step": 1851 + }, + { + "epoch": 1.8228346456692912, + "grad_norm": 13.752058029174805, + "learning_rate": 6.37911920098187e-07, + "loss": 0.4977, + "step": 1852 + }, + { + "epoch": 1.8238188976377954, + "grad_norm": 14.460210800170898, + "learning_rate": 6.371742898302503e-07, + "loss": 0.9087, + "step": 1853 + }, + { + "epoch": 1.824803149606299, + "grad_norm": 5.7258758544921875, + "learning_rate": 6.364363388232091e-07, + "loss": 0.1289, + "step": 1854 + }, + { + "epoch": 1.8257874015748032, + "grad_norm": 9.88226318359375, + "learning_rate": 6.356980688237535e-07, + "loss": 0.4927, + "step": 1855 + }, + { + "epoch": 1.826771653543307, + "grad_norm": 10.143924713134766, + "learning_rate": 6.349594815793275e-07, + "loss": 0.2055, + "step": 1856 + }, + { + "epoch": 1.827755905511811, + "grad_norm": 11.382137298583984, + "learning_rate": 6.34220578838127e-07, + "loss": 0.5921, + "step": 1857 + }, + { + "epoch": 1.8287401574803148, + "grad_norm": 10.316205024719238, + "learning_rate": 6.334813623490939e-07, + "loss": 0.5854, + "step": 1858 + }, + { + "epoch": 1.829724409448819, + "grad_norm": 13.60283374786377, + "learning_rate": 6.327418338619132e-07, + "loss": 0.65, + "step": 1859 + }, + { + "epoch": 1.8307086614173227, + "grad_norm": 9.580816268920898, + "learning_rate": 6.320019951270082e-07, + "loss": 0.3905, + "step": 1860 + }, + { + "epoch": 1.8316929133858268, + "grad_norm": 9.066387176513672, + "learning_rate": 6.312618478955368e-07, + "loss": 0.3033, + "step": 1861 + }, + { + "epoch": 1.8326771653543306, + "grad_norm": 16.17304229736328, + "learning_rate": 6.305213939193864e-07, + "loss": 0.9375, + "step": 1862 + }, + { + "epoch": 1.8336614173228347, + "grad_norm": 13.129256248474121, + "learning_rate": 6.297806349511711e-07, + "loss": 0.7983, + "step": 1863 + }, + { + "epoch": 1.8346456692913384, + "grad_norm": 11.40530776977539, + "learning_rate": 6.290395727442269e-07, + "loss": 0.2748, + "step": 1864 + }, + { + "epoch": 1.8356299212598426, + "grad_norm": 7.396024703979492, + "learning_rate": 6.282982090526071e-07, + "loss": 0.2819, + "step": 1865 + }, + { + "epoch": 1.8366141732283463, + "grad_norm": 13.383564949035645, + "learning_rate": 6.27556545631079e-07, + "loss": 0.5683, + "step": 1866 + }, + { + "epoch": 1.8375984251968505, + "grad_norm": 11.786568641662598, + "learning_rate": 6.268145842351191e-07, + "loss": 0.4445, + "step": 1867 + }, + { + "epoch": 1.8385826771653542, + "grad_norm": 12.825894355773926, + "learning_rate": 6.260723266209094e-07, + "loss": 0.6943, + "step": 1868 + }, + { + "epoch": 1.8395669291338583, + "grad_norm": 9.532464027404785, + "learning_rate": 6.253297745453329e-07, + "loss": 0.345, + "step": 1869 + }, + { + "epoch": 1.840551181102362, + "grad_norm": 7.8513007164001465, + "learning_rate": 6.245869297659694e-07, + "loss": 0.3122, + "step": 1870 + }, + { + "epoch": 1.8415354330708662, + "grad_norm": 12.317331314086914, + "learning_rate": 6.238437940410921e-07, + "loss": 0.3699, + "step": 1871 + }, + { + "epoch": 1.84251968503937, + "grad_norm": 10.58767318725586, + "learning_rate": 6.231003691296622e-07, + "loss": 0.7276, + "step": 1872 + }, + { + "epoch": 1.843503937007874, + "grad_norm": 16.209243774414062, + "learning_rate": 6.223566567913256e-07, + "loss": 0.7885, + "step": 1873 + }, + { + "epoch": 1.844488188976378, + "grad_norm": 8.840424537658691, + "learning_rate": 6.216126587864084e-07, + "loss": 0.3389, + "step": 1874 + }, + { + "epoch": 1.845472440944882, + "grad_norm": 14.408219337463379, + "learning_rate": 6.208683768759136e-07, + "loss": 0.7737, + "step": 1875 + }, + { + "epoch": 1.8464566929133859, + "grad_norm": 14.283469200134277, + "learning_rate": 6.201238128215152e-07, + "loss": 0.6218, + "step": 1876 + }, + { + "epoch": 1.8474409448818898, + "grad_norm": 10.20729923248291, + "learning_rate": 6.193789683855555e-07, + "loss": 0.5406, + "step": 1877 + }, + { + "epoch": 1.8484251968503937, + "grad_norm": 6.3633270263671875, + "learning_rate": 6.186338453310404e-07, + "loss": 0.2269, + "step": 1878 + }, + { + "epoch": 1.8494094488188977, + "grad_norm": 11.058782577514648, + "learning_rate": 6.178884454216355e-07, + "loss": 0.5565, + "step": 1879 + }, + { + "epoch": 1.8503937007874016, + "grad_norm": 9.93575668334961, + "learning_rate": 6.171427704216609e-07, + "loss": 0.5396, + "step": 1880 + }, + { + "epoch": 1.8513779527559056, + "grad_norm": 13.250054359436035, + "learning_rate": 6.163968220960891e-07, + "loss": 0.7901, + "step": 1881 + }, + { + "epoch": 1.8523622047244095, + "grad_norm": 11.081515312194824, + "learning_rate": 6.156506022105382e-07, + "loss": 0.4403, + "step": 1882 + }, + { + "epoch": 1.8533464566929134, + "grad_norm": 10.32402229309082, + "learning_rate": 6.1490411253127e-07, + "loss": 0.628, + "step": 1883 + }, + { + "epoch": 1.8543307086614174, + "grad_norm": 8.243292808532715, + "learning_rate": 6.141573548251847e-07, + "loss": 0.2079, + "step": 1884 + }, + { + "epoch": 1.8553149606299213, + "grad_norm": 13.154631614685059, + "learning_rate": 6.134103308598166e-07, + "loss": 0.8746, + "step": 1885 + }, + { + "epoch": 1.8562992125984252, + "grad_norm": 12.704285621643066, + "learning_rate": 6.126630424033303e-07, + "loss": 0.6964, + "step": 1886 + }, + { + "epoch": 1.8572834645669292, + "grad_norm": 10.261919021606445, + "learning_rate": 6.119154912245167e-07, + "loss": 0.324, + "step": 1887 + }, + { + "epoch": 1.858267716535433, + "grad_norm": 7.801597595214844, + "learning_rate": 6.111676790927884e-07, + "loss": 0.2366, + "step": 1888 + }, + { + "epoch": 1.859251968503937, + "grad_norm": 11.690486907958984, + "learning_rate": 6.104196077781757e-07, + "loss": 0.5403, + "step": 1889 + }, + { + "epoch": 1.860236220472441, + "grad_norm": 13.91586685180664, + "learning_rate": 6.096712790513222e-07, + "loss": 0.7417, + "step": 1890 + }, + { + "epoch": 1.861220472440945, + "grad_norm": 11.251380920410156, + "learning_rate": 6.08922694683481e-07, + "loss": 0.576, + "step": 1891 + }, + { + "epoch": 1.8622047244094488, + "grad_norm": 8.049712181091309, + "learning_rate": 6.081738564465103e-07, + "loss": 0.3191, + "step": 1892 + }, + { + "epoch": 1.8631889763779528, + "grad_norm": 11.583464622497559, + "learning_rate": 6.07424766112869e-07, + "loss": 0.7684, + "step": 1893 + }, + { + "epoch": 1.8641732283464567, + "grad_norm": 15.939492225646973, + "learning_rate": 6.066754254556127e-07, + "loss": 0.5849, + "step": 1894 + }, + { + "epoch": 1.8651574803149606, + "grad_norm": 13.468368530273438, + "learning_rate": 6.059258362483897e-07, + "loss": 0.6496, + "step": 1895 + }, + { + "epoch": 1.8661417322834646, + "grad_norm": 9.483004570007324, + "learning_rate": 6.051760002654364e-07, + "loss": 0.3471, + "step": 1896 + }, + { + "epoch": 1.8671259842519685, + "grad_norm": 11.548891067504883, + "learning_rate": 6.044259192815735e-07, + "loss": 0.3582, + "step": 1897 + }, + { + "epoch": 1.8681102362204725, + "grad_norm": 9.642685890197754, + "learning_rate": 6.036755950722014e-07, + "loss": 0.4418, + "step": 1898 + }, + { + "epoch": 1.8690944881889764, + "grad_norm": 14.921225547790527, + "learning_rate": 6.029250294132965e-07, + "loss": 0.7053, + "step": 1899 + }, + { + "epoch": 1.8700787401574803, + "grad_norm": 9.473539352416992, + "learning_rate": 6.021742240814067e-07, + "loss": 0.436, + "step": 1900 + }, + { + "epoch": 1.8710629921259843, + "grad_norm": 16.96040153503418, + "learning_rate": 6.014231808536463e-07, + "loss": 0.8283, + "step": 1901 + }, + { + "epoch": 1.8720472440944882, + "grad_norm": 13.724132537841797, + "learning_rate": 6.006719015076938e-07, + "loss": 0.5635, + "step": 1902 + }, + { + "epoch": 1.8730314960629921, + "grad_norm": 10.564257621765137, + "learning_rate": 5.999203878217861e-07, + "loss": 0.4921, + "step": 1903 + }, + { + "epoch": 1.874015748031496, + "grad_norm": 9.157358169555664, + "learning_rate": 5.99168641574715e-07, + "loss": 0.2843, + "step": 1904 + }, + { + "epoch": 1.875, + "grad_norm": 13.874457359313965, + "learning_rate": 5.984166645458224e-07, + "loss": 0.3578, + "step": 1905 + }, + { + "epoch": 1.875984251968504, + "grad_norm": 12.66700267791748, + "learning_rate": 5.976644585149967e-07, + "loss": 0.5495, + "step": 1906 + }, + { + "epoch": 1.8769685039370079, + "grad_norm": 10.231301307678223, + "learning_rate": 5.969120252626684e-07, + "loss": 0.567, + "step": 1907 + }, + { + "epoch": 1.8779527559055118, + "grad_norm": 12.036577224731445, + "learning_rate": 5.961593665698055e-07, + "loss": 0.6658, + "step": 1908 + }, + { + "epoch": 1.8789370078740157, + "grad_norm": 14.570939064025879, + "learning_rate": 5.954064842179099e-07, + "loss": 0.741, + "step": 1909 + }, + { + "epoch": 1.8799212598425197, + "grad_norm": 20.911849975585938, + "learning_rate": 5.946533799890127e-07, + "loss": 0.8485, + "step": 1910 + }, + { + "epoch": 1.8809055118110236, + "grad_norm": 7.768706321716309, + "learning_rate": 5.939000556656704e-07, + "loss": 0.3136, + "step": 1911 + }, + { + "epoch": 1.8818897637795275, + "grad_norm": 10.94675064086914, + "learning_rate": 5.931465130309604e-07, + "loss": 0.3753, + "step": 1912 + }, + { + "epoch": 1.8828740157480315, + "grad_norm": 8.576940536499023, + "learning_rate": 5.923927538684764e-07, + "loss": 0.3434, + "step": 1913 + }, + { + "epoch": 1.8838582677165354, + "grad_norm": 8.819656372070312, + "learning_rate": 5.916387799623253e-07, + "loss": 0.4035, + "step": 1914 + }, + { + "epoch": 1.8848425196850394, + "grad_norm": 9.173199653625488, + "learning_rate": 5.908845930971217e-07, + "loss": 0.6502, + "step": 1915 + }, + { + "epoch": 1.8858267716535433, + "grad_norm": 9.636296272277832, + "learning_rate": 5.901301950579845e-07, + "loss": 0.335, + "step": 1916 + }, + { + "epoch": 1.8868110236220472, + "grad_norm": 11.033160209655762, + "learning_rate": 5.893755876305328e-07, + "loss": 0.6374, + "step": 1917 + }, + { + "epoch": 1.8877952755905512, + "grad_norm": 12.411829948425293, + "learning_rate": 5.886207726008805e-07, + "loss": 0.8368, + "step": 1918 + }, + { + "epoch": 1.888779527559055, + "grad_norm": 5.822746753692627, + "learning_rate": 5.878657517556336e-07, + "loss": 0.2835, + "step": 1919 + }, + { + "epoch": 1.889763779527559, + "grad_norm": 12.679840087890625, + "learning_rate": 5.871105268818848e-07, + "loss": 0.8733, + "step": 1920 + }, + { + "epoch": 1.890748031496063, + "grad_norm": 9.119536399841309, + "learning_rate": 5.863550997672101e-07, + "loss": 0.2776, + "step": 1921 + }, + { + "epoch": 1.891732283464567, + "grad_norm": 9.011099815368652, + "learning_rate": 5.855994721996638e-07, + "loss": 0.4738, + "step": 1922 + }, + { + "epoch": 1.8927165354330708, + "grad_norm": 16.789323806762695, + "learning_rate": 5.848436459677752e-07, + "loss": 0.7217, + "step": 1923 + }, + { + "epoch": 1.8937007874015748, + "grad_norm": 10.968154907226562, + "learning_rate": 5.840876228605432e-07, + "loss": 0.8127, + "step": 1924 + }, + { + "epoch": 1.8946850393700787, + "grad_norm": 12.029850006103516, + "learning_rate": 5.833314046674329e-07, + "loss": 0.4527, + "step": 1925 + }, + { + "epoch": 1.8956692913385826, + "grad_norm": 14.161413192749023, + "learning_rate": 5.825749931783715e-07, + "loss": 0.7141, + "step": 1926 + }, + { + "epoch": 1.8966535433070866, + "grad_norm": 7.364201068878174, + "learning_rate": 5.818183901837435e-07, + "loss": 0.257, + "step": 1927 + }, + { + "epoch": 1.8976377952755905, + "grad_norm": 10.396787643432617, + "learning_rate": 5.810615974743864e-07, + "loss": 0.6537, + "step": 1928 + }, + { + "epoch": 1.8986220472440944, + "grad_norm": 14.781997680664062, + "learning_rate": 5.803046168415872e-07, + "loss": 0.6333, + "step": 1929 + }, + { + "epoch": 1.8996062992125984, + "grad_norm": 6.7298359870910645, + "learning_rate": 5.795474500770772e-07, + "loss": 0.236, + "step": 1930 + }, + { + "epoch": 1.9005905511811023, + "grad_norm": 6.864409923553467, + "learning_rate": 5.787900989730291e-07, + "loss": 0.2141, + "step": 1931 + }, + { + "epoch": 1.9015748031496063, + "grad_norm": 6.676103591918945, + "learning_rate": 5.780325653220509e-07, + "loss": 0.2644, + "step": 1932 + }, + { + "epoch": 1.9025590551181102, + "grad_norm": 8.446288108825684, + "learning_rate": 5.772748509171833e-07, + "loss": 0.2762, + "step": 1933 + }, + { + "epoch": 1.9035433070866141, + "grad_norm": 8.367350578308105, + "learning_rate": 5.765169575518948e-07, + "loss": 0.402, + "step": 1934 + }, + { + "epoch": 1.904527559055118, + "grad_norm": 14.271595001220703, + "learning_rate": 5.757588870200772e-07, + "loss": 0.727, + "step": 1935 + }, + { + "epoch": 1.905511811023622, + "grad_norm": 17.437908172607422, + "learning_rate": 5.750006411160421e-07, + "loss": 1.1796, + "step": 1936 + }, + { + "epoch": 1.906496062992126, + "grad_norm": 11.969321250915527, + "learning_rate": 5.742422216345157e-07, + "loss": 0.5995, + "step": 1937 + }, + { + "epoch": 1.90748031496063, + "grad_norm": 8.234703063964844, + "learning_rate": 5.734836303706352e-07, + "loss": 0.2578, + "step": 1938 + }, + { + "epoch": 1.9084645669291338, + "grad_norm": 11.856316566467285, + "learning_rate": 5.727248691199446e-07, + "loss": 0.6926, + "step": 1939 + }, + { + "epoch": 1.909448818897638, + "grad_norm": 12.148863792419434, + "learning_rate": 5.719659396783902e-07, + "loss": 0.6428, + "step": 1940 + }, + { + "epoch": 1.9104330708661417, + "grad_norm": 6.12877893447876, + "learning_rate": 5.712068438423161e-07, + "loss": 0.158, + "step": 1941 + }, + { + "epoch": 1.9114173228346458, + "grad_norm": 14.655328750610352, + "learning_rate": 5.704475834084608e-07, + "loss": 0.7288, + "step": 1942 + }, + { + "epoch": 1.9124015748031495, + "grad_norm": 11.98600959777832, + "learning_rate": 5.696881601739517e-07, + "loss": 0.6368, + "step": 1943 + }, + { + "epoch": 1.9133858267716537, + "grad_norm": 11.770471572875977, + "learning_rate": 5.689285759363022e-07, + "loss": 0.4142, + "step": 1944 + }, + { + "epoch": 1.9143700787401574, + "grad_norm": 14.175639152526855, + "learning_rate": 5.681688324934064e-07, + "loss": 0.6299, + "step": 1945 + }, + { + "epoch": 1.9153543307086616, + "grad_norm": 9.831664085388184, + "learning_rate": 5.674089316435353e-07, + "loss": 0.4991, + "step": 1946 + }, + { + "epoch": 1.9163385826771653, + "grad_norm": 11.123538970947266, + "learning_rate": 5.666488751853321e-07, + "loss": 0.5262, + "step": 1947 + }, + { + "epoch": 1.9173228346456694, + "grad_norm": 7.65354585647583, + "learning_rate": 5.658886649178096e-07, + "loss": 0.35, + "step": 1948 + }, + { + "epoch": 1.9183070866141732, + "grad_norm": 19.06454849243164, + "learning_rate": 5.651283026403427e-07, + "loss": 1.0507, + "step": 1949 + }, + { + "epoch": 1.9192913385826773, + "grad_norm": 24.62841033935547, + "learning_rate": 5.64367790152668e-07, + "loss": 1.1313, + "step": 1950 + }, + { + "epoch": 1.920275590551181, + "grad_norm": 9.834978103637695, + "learning_rate": 5.636071292548766e-07, + "loss": 0.3222, + "step": 1951 + }, + { + "epoch": 1.9212598425196852, + "grad_norm": 13.258596420288086, + "learning_rate": 5.628463217474108e-07, + "loss": 0.4616, + "step": 1952 + }, + { + "epoch": 1.922244094488189, + "grad_norm": 10.210066795349121, + "learning_rate": 5.620853694310604e-07, + "loss": 0.5276, + "step": 1953 + }, + { + "epoch": 1.923228346456693, + "grad_norm": 13.035691261291504, + "learning_rate": 5.613242741069578e-07, + "loss": 0.6014, + "step": 1954 + }, + { + "epoch": 1.9242125984251968, + "grad_norm": 9.69564437866211, + "learning_rate": 5.605630375765739e-07, + "loss": 0.3682, + "step": 1955 + }, + { + "epoch": 1.925196850393701, + "grad_norm": 11.325922012329102, + "learning_rate": 5.598016616417136e-07, + "loss": 0.6999, + "step": 1956 + }, + { + "epoch": 1.9261811023622046, + "grad_norm": 7.43300724029541, + "learning_rate": 5.590401481045118e-07, + "loss": 0.3223, + "step": 1957 + }, + { + "epoch": 1.9271653543307088, + "grad_norm": 11.16835880279541, + "learning_rate": 5.582784987674296e-07, + "loss": 0.5748, + "step": 1958 + }, + { + "epoch": 1.9281496062992125, + "grad_norm": 8.531428337097168, + "learning_rate": 5.575167154332489e-07, + "loss": 0.2562, + "step": 1959 + }, + { + "epoch": 1.9291338582677167, + "grad_norm": 15.335569381713867, + "learning_rate": 5.567547999050692e-07, + "loss": 0.7111, + "step": 1960 + }, + { + "epoch": 1.9301181102362204, + "grad_norm": 11.951147079467773, + "learning_rate": 5.559927539863024e-07, + "loss": 0.6036, + "step": 1961 + }, + { + "epoch": 1.9311023622047245, + "grad_norm": 8.56628704071045, + "learning_rate": 5.552305794806699e-07, + "loss": 0.2651, + "step": 1962 + }, + { + "epoch": 1.9320866141732282, + "grad_norm": 11.414342880249023, + "learning_rate": 5.544682781921964e-07, + "loss": 0.4267, + "step": 1963 + }, + { + "epoch": 1.9330708661417324, + "grad_norm": 7.9041266441345215, + "learning_rate": 5.537058519252076e-07, + "loss": 0.3572, + "step": 1964 + }, + { + "epoch": 1.9340551181102361, + "grad_norm": 7.771170616149902, + "learning_rate": 5.529433024843244e-07, + "loss": 0.1698, + "step": 1965 + }, + { + "epoch": 1.9350393700787403, + "grad_norm": 5.711643695831299, + "learning_rate": 5.521806316744594e-07, + "loss": 0.1659, + "step": 1966 + }, + { + "epoch": 1.936023622047244, + "grad_norm": 8.713125228881836, + "learning_rate": 5.514178413008125e-07, + "loss": 0.4387, + "step": 1967 + }, + { + "epoch": 1.9370078740157481, + "grad_norm": 12.458745002746582, + "learning_rate": 5.506549331688669e-07, + "loss": 0.8473, + "step": 1968 + }, + { + "epoch": 1.9379921259842519, + "grad_norm": 7.764037132263184, + "learning_rate": 5.49891909084384e-07, + "loss": 0.3323, + "step": 1969 + }, + { + "epoch": 1.938976377952756, + "grad_norm": 16.753019332885742, + "learning_rate": 5.491287708534e-07, + "loss": 0.6246, + "step": 1970 + }, + { + "epoch": 1.9399606299212597, + "grad_norm": 7.089112758636475, + "learning_rate": 5.483655202822214e-07, + "loss": 0.2985, + "step": 1971 + }, + { + "epoch": 1.9409448818897639, + "grad_norm": 7.975224018096924, + "learning_rate": 5.4760215917742e-07, + "loss": 0.2729, + "step": 1972 + }, + { + "epoch": 1.9419291338582676, + "grad_norm": 7.56587028503418, + "learning_rate": 5.4683868934583e-07, + "loss": 0.3262, + "step": 1973 + }, + { + "epoch": 1.9429133858267718, + "grad_norm": 7.976964950561523, + "learning_rate": 5.460751125945424e-07, + "loss": 0.2995, + "step": 1974 + }, + { + "epoch": 1.9438976377952755, + "grad_norm": 9.415502548217773, + "learning_rate": 5.453114307309015e-07, + "loss": 0.4751, + "step": 1975 + }, + { + "epoch": 1.9448818897637796, + "grad_norm": 8.0772705078125, + "learning_rate": 5.445476455625002e-07, + "loss": 0.3349, + "step": 1976 + }, + { + "epoch": 1.9458661417322833, + "grad_norm": 8.458420753479004, + "learning_rate": 5.437837588971761e-07, + "loss": 0.343, + "step": 1977 + }, + { + "epoch": 1.9468503937007875, + "grad_norm": 9.498231887817383, + "learning_rate": 5.430197725430071e-07, + "loss": 0.2648, + "step": 1978 + }, + { + "epoch": 1.9478346456692912, + "grad_norm": 9.385558128356934, + "learning_rate": 5.422556883083069e-07, + "loss": 0.389, + "step": 1979 + }, + { + "epoch": 1.9488188976377954, + "grad_norm": 20.827730178833008, + "learning_rate": 5.414915080016209e-07, + "loss": 1.2056, + "step": 1980 + }, + { + "epoch": 1.949803149606299, + "grad_norm": 11.813233375549316, + "learning_rate": 5.407272334317221e-07, + "loss": 0.4393, + "step": 1981 + }, + { + "epoch": 1.9507874015748032, + "grad_norm": 8.28148078918457, + "learning_rate": 5.399628664076062e-07, + "loss": 0.292, + "step": 1982 + }, + { + "epoch": 1.951771653543307, + "grad_norm": 16.118715286254883, + "learning_rate": 5.391984087384882e-07, + "loss": 0.7443, + "step": 1983 + }, + { + "epoch": 1.952755905511811, + "grad_norm": 7.399683952331543, + "learning_rate": 5.384338622337973e-07, + "loss": 0.1946, + "step": 1984 + }, + { + "epoch": 1.9537401574803148, + "grad_norm": 21.07978630065918, + "learning_rate": 5.37669228703173e-07, + "loss": 1.0972, + "step": 1985 + }, + { + "epoch": 1.954724409448819, + "grad_norm": 11.241035461425781, + "learning_rate": 5.369045099564613e-07, + "loss": 0.6114, + "step": 1986 + }, + { + "epoch": 1.9557086614173227, + "grad_norm": 9.055529594421387, + "learning_rate": 5.361397078037089e-07, + "loss": 0.5224, + "step": 1987 + }, + { + "epoch": 1.9566929133858268, + "grad_norm": 8.740368843078613, + "learning_rate": 5.353748240551612e-07, + "loss": 0.4773, + "step": 1988 + }, + { + "epoch": 1.9576771653543306, + "grad_norm": 11.645110130310059, + "learning_rate": 5.346098605212553e-07, + "loss": 0.8126, + "step": 1989 + }, + { + "epoch": 1.9576771653543306, + "eval_Qnli-dev_cosine_accuracy": 0.6640625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8005393743515015, + "eval_Qnli-dev_cosine_ap": 0.6963810565222734, + "eval_Qnli-dev_cosine_f1": 0.6898839137645109, + "eval_Qnli-dev_cosine_f1_threshold": 0.7475045323371887, + "eval_Qnli-dev_cosine_precision": 0.5667574931880109, + "eval_Qnli-dev_cosine_recall": 0.8813559322033898, + "eval_Qnli-dev_dot_accuracy": 0.66796875, + "eval_Qnli-dev_dot_accuracy_threshold": 605.2431640625, + "eval_Qnli-dev_dot_ap": 0.6933667039839441, + "eval_Qnli-dev_dot_f1": 0.6864686468646866, + "eval_Qnli-dev_dot_f1_threshold": 567.939208984375, + "eval_Qnli-dev_dot_precision": 0.5621621621621622, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.666015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.45079803466797, + "eval_Qnli-dev_euclidean_ap": 0.6964584994456302, + "eval_Qnli-dev_euclidean_f1": 0.6898839137645109, + "eval_Qnli-dev_euclidean_f1_threshold": 19.631027221679688, + "eval_Qnli-dev_euclidean_precision": 0.5667574931880109, + "eval_Qnli-dev_euclidean_recall": 0.8813559322033898, + "eval_Qnli-dev_manhattan_accuracy": 0.671875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 354.006103515625, + "eval_Qnli-dev_manhattan_ap": 0.7030057331700904, + "eval_Qnli-dev_manhattan_f1": 0.6949152542372881, + "eval_Qnli-dev_manhattan_f1_threshold": 410.61859130859375, + "eval_Qnli-dev_manhattan_precision": 0.5790960451977402, + "eval_Qnli-dev_manhattan_recall": 0.8686440677966102, + "eval_Qnli-dev_max_accuracy": 0.671875, + "eval_Qnli-dev_max_accuracy_threshold": 605.2431640625, + "eval_Qnli-dev_max_ap": 0.7030057331700904, + "eval_Qnli-dev_max_f1": 0.6949152542372881, + "eval_Qnli-dev_max_f1_threshold": 567.939208984375, + "eval_Qnli-dev_max_precision": 0.5790960451977402, + "eval_Qnli-dev_max_recall": 0.8813559322033898, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9275094270706177, + "eval_allNLI-dev_cosine_ap": 0.5163761733021546, + "eval_allNLI-dev_cosine_f1": 0.5508684863523573, + "eval_allNLI-dev_cosine_f1_threshold": 0.862393319606781, + "eval_allNLI-dev_cosine_precision": 0.4826086956521739, + "eval_allNLI-dev_cosine_recall": 0.6416184971098265, + "eval_allNLI-dev_dot_accuracy": 0.689453125, + "eval_allNLI-dev_dot_accuracy_threshold": 701.0382080078125, + "eval_allNLI-dev_dot_ap": 0.5170597922539218, + "eval_allNLI-dev_dot_f1": 0.5481481481481482, + "eval_allNLI-dev_dot_f1_threshold": 657.1117553710938, + "eval_allNLI-dev_dot_precision": 0.47844827586206895, + "eval_allNLI-dev_dot_recall": 0.6416184971098265, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.526958465576172, + "eval_allNLI-dev_euclidean_ap": 0.5163758899855585, + "eval_allNLI-dev_euclidean_f1": 0.5508684863523573, + "eval_allNLI-dev_euclidean_f1_threshold": 14.494762420654297, + "eval_allNLI-dev_euclidean_precision": 0.4826086956521739, + "eval_allNLI-dev_euclidean_recall": 0.6416184971098265, + "eval_allNLI-dev_manhattan_accuracy": 0.689453125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 217.35939025878906, + "eval_allNLI-dev_manhattan_ap": 0.5200494439911988, + "eval_allNLI-dev_manhattan_f1": 0.5686746987951807, + "eval_allNLI-dev_manhattan_f1_threshold": 311.76495361328125, + "eval_allNLI-dev_manhattan_precision": 0.48760330578512395, + "eval_allNLI-dev_manhattan_recall": 0.6820809248554913, + "eval_allNLI-dev_max_accuracy": 0.689453125, + "eval_allNLI-dev_max_accuracy_threshold": 701.0382080078125, + "eval_allNLI-dev_max_ap": 0.5200494439911988, + "eval_allNLI-dev_max_f1": 0.5686746987951807, + "eval_allNLI-dev_max_f1_threshold": 657.1117553710938, + "eval_allNLI-dev_max_precision": 0.48760330578512395, + "eval_allNLI-dev_max_recall": 0.6820809248554913, + "eval_loss": 0.6397632956504822, + "eval_runtime": 51.2966, + "eval_samples_per_second": 32.439, + "eval_sequential_score": 0.7030057331700904, + "eval_steps_per_second": 0.136, + "eval_sts-test_pearson_cosine": 0.6515835819558387, + "eval_sts-test_pearson_dot": 0.6496631696023635, + "eval_sts-test_pearson_euclidean": 0.6664666041698615, + "eval_sts-test_pearson_manhattan": 0.6689532703109837, + "eval_sts-test_pearson_max": 0.6689532703109837, + "eval_sts-test_spearman_cosine": 0.6624880521170792, + "eval_sts-test_spearman_dot": 0.660623311806161, + "eval_sts-test_spearman_euclidean": 0.6626712190299542, + "eval_sts-test_spearman_manhattan": 0.6642616788598046, + "eval_sts-test_spearman_max": 0.6642616788598046, + "step": 1989 + }, + { + "epoch": 1.9586614173228347, + "grad_norm": 14.053847312927246, + "learning_rate": 5.338448190126184e-07, + "loss": 0.9602, + "step": 1990 + }, + { + "epoch": 1.9596456692913384, + "grad_norm": 5.862058162689209, + "learning_rate": 5.330797013400614e-07, + "loss": 0.1188, + "step": 1991 + }, + { + "epoch": 1.9606299212598426, + "grad_norm": 13.629286766052246, + "learning_rate": 5.323145093145763e-07, + "loss": 0.6278, + "step": 1992 + }, + { + "epoch": 1.9616141732283463, + "grad_norm": 10.82761287689209, + "learning_rate": 5.3154924474733e-07, + "loss": 0.4647, + "step": 1993 + }, + { + "epoch": 1.9625984251968505, + "grad_norm": 11.246370315551758, + "learning_rate": 5.307839094496621e-07, + "loss": 0.4902, + "step": 1994 + }, + { + "epoch": 1.9635826771653542, + "grad_norm": 8.25075626373291, + "learning_rate": 5.300185052330793e-07, + "loss": 0.3656, + "step": 1995 + }, + { + "epoch": 1.9645669291338583, + "grad_norm": 15.318300247192383, + "learning_rate": 5.292530339092513e-07, + "loss": 0.5947, + "step": 1996 + }, + { + "epoch": 1.965551181102362, + "grad_norm": 5.805018901824951, + "learning_rate": 5.284874972900064e-07, + "loss": 0.2003, + "step": 1997 + }, + { + "epoch": 1.9665354330708662, + "grad_norm": 11.280017852783203, + "learning_rate": 5.277218971873281e-07, + "loss": 0.6722, + "step": 1998 + }, + { + "epoch": 1.96751968503937, + "grad_norm": 14.226088523864746, + "learning_rate": 5.269562354133496e-07, + "loss": 0.8103, + "step": 1999 + }, + { + "epoch": 1.968503937007874, + "grad_norm": 12.065138816833496, + "learning_rate": 5.261905137803504e-07, + "loss": 0.6414, + "step": 2000 + }, + { + "epoch": 1.969488188976378, + "grad_norm": 9.506181716918945, + "learning_rate": 5.254247341007514e-07, + "loss": 0.4922, + "step": 2001 + }, + { + "epoch": 1.970472440944882, + "grad_norm": 11.069506645202637, + "learning_rate": 5.246588981871113e-07, + "loss": 0.5083, + "step": 2002 + }, + { + "epoch": 1.9714566929133859, + "grad_norm": 7.399655818939209, + "learning_rate": 5.238930078521215e-07, + "loss": 0.3076, + "step": 2003 + }, + { + "epoch": 1.9724409448818898, + "grad_norm": 13.209633827209473, + "learning_rate": 5.231270649086023e-07, + "loss": 0.4796, + "step": 2004 + }, + { + "epoch": 1.9734251968503937, + "grad_norm": 7.998384952545166, + "learning_rate": 5.22361071169499e-07, + "loss": 0.2737, + "step": 2005 + }, + { + "epoch": 1.9744094488188977, + "grad_norm": 10.659584045410156, + "learning_rate": 5.215950284478763e-07, + "loss": 0.3725, + "step": 2006 + }, + { + "epoch": 1.9753937007874016, + "grad_norm": 8.86026382446289, + "learning_rate": 5.208289385569154e-07, + "loss": 0.3862, + "step": 2007 + }, + { + "epoch": 1.9763779527559056, + "grad_norm": 8.115157127380371, + "learning_rate": 5.20062803309909e-07, + "loss": 0.2906, + "step": 2008 + }, + { + "epoch": 1.9773622047244095, + "grad_norm": 4.823766231536865, + "learning_rate": 5.192966245202571e-07, + "loss": 0.1386, + "step": 2009 + }, + { + "epoch": 1.9783464566929134, + "grad_norm": 15.458700180053711, + "learning_rate": 5.18530404001463e-07, + "loss": 0.7797, + "step": 2010 + }, + { + "epoch": 1.9793307086614174, + "grad_norm": 7.782804489135742, + "learning_rate": 5.177641435671284e-07, + "loss": 0.183, + "step": 2011 + }, + { + "epoch": 1.9803149606299213, + "grad_norm": 6.583715915679932, + "learning_rate": 5.169978450309498e-07, + "loss": 0.2328, + "step": 2012 + }, + { + "epoch": 1.9812992125984252, + "grad_norm": 7.962342262268066, + "learning_rate": 5.162315102067139e-07, + "loss": 0.2509, + "step": 2013 + }, + { + "epoch": 1.9822834645669292, + "grad_norm": 16.081270217895508, + "learning_rate": 5.154651409082927e-07, + "loss": 1.0533, + "step": 2014 + }, + { + "epoch": 1.983267716535433, + "grad_norm": 9.272991180419922, + "learning_rate": 5.146987389496408e-07, + "loss": 0.5187, + "step": 2015 + }, + { + "epoch": 1.984251968503937, + "grad_norm": 12.794522285461426, + "learning_rate": 5.139323061447892e-07, + "loss": 0.8734, + "step": 2016 + }, + { + "epoch": 1.985236220472441, + "grad_norm": 11.748847961425781, + "learning_rate": 5.131658443078426e-07, + "loss": 0.592, + "step": 2017 + }, + { + "epoch": 1.986220472440945, + "grad_norm": 11.479252815246582, + "learning_rate": 5.123993552529736e-07, + "loss": 0.5189, + "step": 2018 + }, + { + "epoch": 1.9872047244094488, + "grad_norm": 11.677574157714844, + "learning_rate": 5.116328407944201e-07, + "loss": 0.7473, + "step": 2019 + }, + { + "epoch": 1.9881889763779528, + "grad_norm": 15.634410858154297, + "learning_rate": 5.108663027464795e-07, + "loss": 0.9871, + "step": 2020 + }, + { + "epoch": 1.9891732283464567, + "grad_norm": 10.726166725158691, + "learning_rate": 5.100997429235055e-07, + "loss": 0.4513, + "step": 2021 + }, + { + "epoch": 1.9901574803149606, + "grad_norm": 8.022065162658691, + "learning_rate": 5.093331631399029e-07, + "loss": 0.3456, + "step": 2022 + }, + { + "epoch": 1.9911417322834646, + "grad_norm": 6.33901309967041, + "learning_rate": 5.085665652101244e-07, + "loss": 0.2714, + "step": 2023 + }, + { + "epoch": 1.9921259842519685, + "grad_norm": 10.190364837646484, + "learning_rate": 5.077999509486645e-07, + "loss": 0.5281, + "step": 2024 + }, + { + "epoch": 1.9931102362204725, + "grad_norm": 7.32463264465332, + "learning_rate": 5.070333221700575e-07, + "loss": 0.2752, + "step": 2025 + }, + { + "epoch": 1.9940944881889764, + "grad_norm": 15.004027366638184, + "learning_rate": 5.062666806888715e-07, + "loss": 0.8317, + "step": 2026 + }, + { + "epoch": 1.9950787401574803, + "grad_norm": 10.293353080749512, + "learning_rate": 5.055000283197051e-07, + "loss": 0.4502, + "step": 2027 + }, + { + "epoch": 1.9960629921259843, + "grad_norm": 11.628711700439453, + "learning_rate": 5.047333668771821e-07, + "loss": 0.3672, + "step": 2028 + }, + { + "epoch": 1.9970472440944882, + "grad_norm": 3.886667490005493, + "learning_rate": 5.03966698175948e-07, + "loss": 0.2122, + "step": 2029 + }, + { + "epoch": 1.9980314960629921, + "grad_norm": 13.344625473022461, + "learning_rate": 5.032000240306655e-07, + "loss": 0.9165, + "step": 2030 + }, + { + "epoch": 1.999015748031496, + "grad_norm": 15.02889347076416, + "learning_rate": 5.024333462560104e-07, + "loss": 0.5364, + "step": 2031 + }, + { + "epoch": 2.0, + "grad_norm": 5.44955587387085, + "learning_rate": 5.016666666666668e-07, + "loss": 0.1625, + "step": 2032 + }, + { + "epoch": 2.000984251968504, + "grad_norm": 7.618643760681152, + "learning_rate": 5.008999870773232e-07, + "loss": 0.3005, + "step": 2033 + }, + { + "epoch": 2.001968503937008, + "grad_norm": 15.2062349319458, + "learning_rate": 5.001333093026681e-07, + "loss": 1.1874, + "step": 2034 + }, + { + "epoch": 2.002952755905512, + "grad_norm": 5.787307262420654, + "learning_rate": 4.993666351573856e-07, + "loss": 0.1789, + "step": 2035 + }, + { + "epoch": 2.0039370078740157, + "grad_norm": 8.523685455322266, + "learning_rate": 4.985999664561514e-07, + "loss": 0.5909, + "step": 2036 + }, + { + "epoch": 2.00492125984252, + "grad_norm": 13.194377899169922, + "learning_rate": 4.978333050136285e-07, + "loss": 0.392, + "step": 2037 + }, + { + "epoch": 2.0059055118110236, + "grad_norm": 7.510269641876221, + "learning_rate": 4.970666526444619e-07, + "loss": 0.366, + "step": 2038 + }, + { + "epoch": 2.0068897637795278, + "grad_norm": 7.141751289367676, + "learning_rate": 4.96300011163276e-07, + "loss": 0.2462, + "step": 2039 + }, + { + "epoch": 2.0078740157480315, + "grad_norm": 9.78703784942627, + "learning_rate": 4.955333823846691e-07, + "loss": 0.4829, + "step": 2040 + }, + { + "epoch": 2.0088582677165356, + "grad_norm": 8.474166870117188, + "learning_rate": 4.947667681232094e-07, + "loss": 0.3755, + "step": 2041 + }, + { + "epoch": 2.0098425196850394, + "grad_norm": 10.374906539916992, + "learning_rate": 4.940001701934305e-07, + "loss": 0.4269, + "step": 2042 + }, + { + "epoch": 2.0108267716535435, + "grad_norm": 11.179118156433105, + "learning_rate": 4.932335904098282e-07, + "loss": 0.6941, + "step": 2043 + }, + { + "epoch": 2.0118110236220472, + "grad_norm": 7.24183464050293, + "learning_rate": 4.924670305868541e-07, + "loss": 0.2839, + "step": 2044 + }, + { + "epoch": 2.0127952755905514, + "grad_norm": 7.524889945983887, + "learning_rate": 4.917004925389137e-07, + "loss": 0.1781, + "step": 2045 + }, + { + "epoch": 2.013779527559055, + "grad_norm": 8.83763599395752, + "learning_rate": 4.909339780803601e-07, + "loss": 0.4752, + "step": 2046 + }, + { + "epoch": 2.0147637795275593, + "grad_norm": 19.595184326171875, + "learning_rate": 4.901674890254911e-07, + "loss": 1.324, + "step": 2047 + }, + { + "epoch": 2.015748031496063, + "grad_norm": 7.941568851470947, + "learning_rate": 4.894010271885444e-07, + "loss": 0.2226, + "step": 2048 + }, + { + "epoch": 2.016732283464567, + "grad_norm": 12.508964538574219, + "learning_rate": 4.886345943836928e-07, + "loss": 0.7916, + "step": 2049 + }, + { + "epoch": 2.017716535433071, + "grad_norm": 9.916942596435547, + "learning_rate": 4.878681924250408e-07, + "loss": 0.5052, + "step": 2050 + }, + { + "epoch": 2.018700787401575, + "grad_norm": 13.577058792114258, + "learning_rate": 4.871018231266197e-07, + "loss": 0.7909, + "step": 2051 + }, + { + "epoch": 2.0196850393700787, + "grad_norm": 8.705831527709961, + "learning_rate": 4.863354883023838e-07, + "loss": 0.261, + "step": 2052 + }, + { + "epoch": 2.020669291338583, + "grad_norm": 16.489540100097656, + "learning_rate": 4.85569189766205e-07, + "loss": 0.4512, + "step": 2053 + }, + { + "epoch": 2.0216535433070866, + "grad_norm": 10.788933753967285, + "learning_rate": 4.848029293318705e-07, + "loss": 0.5187, + "step": 2054 + }, + { + "epoch": 2.0226377952755907, + "grad_norm": 6.642419338226318, + "learning_rate": 4.840367088130764e-07, + "loss": 0.1973, + "step": 2055 + }, + { + "epoch": 2.0236220472440944, + "grad_norm": 9.779131889343262, + "learning_rate": 4.832705300234245e-07, + "loss": 0.3391, + "step": 2056 + }, + { + "epoch": 2.0246062992125986, + "grad_norm": 6.594153881072998, + "learning_rate": 4.825043947764181e-07, + "loss": 0.3302, + "step": 2057 + }, + { + "epoch": 2.0255905511811023, + "grad_norm": 11.959356307983398, + "learning_rate": 4.817383048854572e-07, + "loss": 0.6158, + "step": 2058 + }, + { + "epoch": 2.0265748031496065, + "grad_norm": 9.14958667755127, + "learning_rate": 4.809722621638345e-07, + "loss": 0.3849, + "step": 2059 + }, + { + "epoch": 2.02755905511811, + "grad_norm": 9.941996574401855, + "learning_rate": 4.802062684247311e-07, + "loss": 0.3843, + "step": 2060 + }, + { + "epoch": 2.0285433070866143, + "grad_norm": 17.83205223083496, + "learning_rate": 4.79440325481212e-07, + "loss": 0.9799, + "step": 2061 + }, + { + "epoch": 2.029527559055118, + "grad_norm": 12.4343843460083, + "learning_rate": 4.786744351462222e-07, + "loss": 0.515, + "step": 2062 + }, + { + "epoch": 2.030511811023622, + "grad_norm": 6.367133140563965, + "learning_rate": 4.779085992325821e-07, + "loss": 0.1543, + "step": 2063 + }, + { + "epoch": 2.031496062992126, + "grad_norm": 10.106760025024414, + "learning_rate": 4.771428195529833e-07, + "loss": 0.5123, + "step": 2064 + }, + { + "epoch": 2.03248031496063, + "grad_norm": 4.3997883796691895, + "learning_rate": 4.763770979199841e-07, + "loss": 0.1001, + "step": 2065 + }, + { + "epoch": 2.033464566929134, + "grad_norm": 8.13400650024414, + "learning_rate": 4.756114361460057e-07, + "loss": 0.2056, + "step": 2066 + }, + { + "epoch": 2.034448818897638, + "grad_norm": 6.556519985198975, + "learning_rate": 4.748458360433273e-07, + "loss": 0.2022, + "step": 2067 + }, + { + "epoch": 2.0354330708661417, + "grad_norm": 9.841743469238281, + "learning_rate": 4.7408029942408244e-07, + "loss": 0.2532, + "step": 2068 + }, + { + "epoch": 2.036417322834646, + "grad_norm": 12.035931587219238, + "learning_rate": 4.733148281002543e-07, + "loss": 0.528, + "step": 2069 + }, + { + "epoch": 2.0374015748031495, + "grad_norm": 19.423419952392578, + "learning_rate": 4.7254942388367147e-07, + "loss": 1.1732, + "step": 2070 + }, + { + "epoch": 2.0383858267716537, + "grad_norm": 14.042691230773926, + "learning_rate": 4.717840885860036e-07, + "loss": 0.49, + "step": 2071 + }, + { + "epoch": 2.0393700787401574, + "grad_norm": 8.415985107421875, + "learning_rate": 4.7101882401875745e-07, + "loss": 0.4327, + "step": 2072 + }, + { + "epoch": 2.0403543307086616, + "grad_norm": 13.537103652954102, + "learning_rate": 4.7025363199327213e-07, + "loss": 0.7335, + "step": 2073 + }, + { + "epoch": 2.0413385826771653, + "grad_norm": 7.711862087249756, + "learning_rate": 4.694885143207153e-07, + "loss": 0.2256, + "step": 2074 + }, + { + "epoch": 2.0423228346456694, + "grad_norm": 8.378612518310547, + "learning_rate": 4.687234728120783e-07, + "loss": 0.3081, + "step": 2075 + }, + { + "epoch": 2.043307086614173, + "grad_norm": 10.325879096984863, + "learning_rate": 4.6795850927817256e-07, + "loss": 0.5843, + "step": 2076 + }, + { + "epoch": 2.0442913385826773, + "grad_norm": 16.89910125732422, + "learning_rate": 4.6719362552962457e-07, + "loss": 0.7457, + "step": 2077 + }, + { + "epoch": 2.045275590551181, + "grad_norm": 11.279486656188965, + "learning_rate": 4.664288233768724e-07, + "loss": 0.5651, + "step": 2078 + }, + { + "epoch": 2.046259842519685, + "grad_norm": 9.48285961151123, + "learning_rate": 4.6566410463016044e-07, + "loss": 0.2633, + "step": 2079 + }, + { + "epoch": 2.047244094488189, + "grad_norm": 9.818591117858887, + "learning_rate": 4.648994710995363e-07, + "loss": 0.3835, + "step": 2080 + }, + { + "epoch": 2.048228346456693, + "grad_norm": 10.789673805236816, + "learning_rate": 4.641349245948454e-07, + "loss": 0.4973, + "step": 2081 + }, + { + "epoch": 2.0492125984251968, + "grad_norm": 11.748615264892578, + "learning_rate": 4.633704669257273e-07, + "loss": 0.4388, + "step": 2082 + }, + { + "epoch": 2.050196850393701, + "grad_norm": 9.578168869018555, + "learning_rate": 4.626060999016115e-07, + "loss": 0.3956, + "step": 2083 + }, + { + "epoch": 2.0511811023622046, + "grad_norm": 15.06390380859375, + "learning_rate": 4.6184182533171257e-07, + "loss": 0.7272, + "step": 2084 + }, + { + "epoch": 2.052165354330709, + "grad_norm": 13.640119552612305, + "learning_rate": 4.6107764502502684e-07, + "loss": 0.596, + "step": 2085 + }, + { + "epoch": 2.0531496062992125, + "grad_norm": 8.436870574951172, + "learning_rate": 4.603135607903266e-07, + "loss": 0.3083, + "step": 2086 + }, + { + "epoch": 2.0541338582677167, + "grad_norm": 7.328392028808594, + "learning_rate": 4.5954957443615765e-07, + "loss": 0.2522, + "step": 2087 + }, + { + "epoch": 2.0551181102362204, + "grad_norm": 10.441664695739746, + "learning_rate": 4.5878568777083356e-07, + "loss": 0.3875, + "step": 2088 + }, + { + "epoch": 2.0561023622047245, + "grad_norm": 10.235856056213379, + "learning_rate": 4.5802190260243226e-07, + "loss": 0.3108, + "step": 2089 + }, + { + "epoch": 2.0570866141732282, + "grad_norm": 6.590058326721191, + "learning_rate": 4.572582207387912e-07, + "loss": 0.1853, + "step": 2090 + }, + { + "epoch": 2.0580708661417324, + "grad_norm": 19.546649932861328, + "learning_rate": 4.5649464398750356e-07, + "loss": 0.9936, + "step": 2091 + }, + { + "epoch": 2.059055118110236, + "grad_norm": 9.767743110656738, + "learning_rate": 4.557311741559135e-07, + "loss": 0.3712, + "step": 2092 + }, + { + "epoch": 2.0600393700787403, + "grad_norm": 10.898755073547363, + "learning_rate": 4.5496781305111215e-07, + "loss": 0.6181, + "step": 2093 + }, + { + "epoch": 2.061023622047244, + "grad_norm": 5.6319684982299805, + "learning_rate": 4.5420456247993354e-07, + "loss": 0.1535, + "step": 2094 + }, + { + "epoch": 2.062007874015748, + "grad_norm": 8.243206977844238, + "learning_rate": 4.5344142424894954e-07, + "loss": 0.2344, + "step": 2095 + }, + { + "epoch": 2.062992125984252, + "grad_norm": 10.080398559570312, + "learning_rate": 4.526784001644667e-07, + "loss": 0.408, + "step": 2096 + }, + { + "epoch": 2.063976377952756, + "grad_norm": 8.111991882324219, + "learning_rate": 4.5191549203252103e-07, + "loss": 0.2513, + "step": 2097 + }, + { + "epoch": 2.0649606299212597, + "grad_norm": 8.100790023803711, + "learning_rate": 4.511527016588742e-07, + "loss": 0.1991, + "step": 2098 + }, + { + "epoch": 2.065944881889764, + "grad_norm": 11.35762882232666, + "learning_rate": 4.503900308490092e-07, + "loss": 0.4484, + "step": 2099 + }, + { + "epoch": 2.0669291338582676, + "grad_norm": 12.64667797088623, + "learning_rate": 4.496274814081259e-07, + "loss": 0.7249, + "step": 2100 + }, + { + "epoch": 2.0679133858267718, + "grad_norm": 8.352909088134766, + "learning_rate": 4.4886505514113695e-07, + "loss": 0.3423, + "step": 2101 + }, + { + "epoch": 2.0688976377952755, + "grad_norm": 7.676507472991943, + "learning_rate": 4.481027538526636e-07, + "loss": 0.3492, + "step": 2102 + }, + { + "epoch": 2.0698818897637796, + "grad_norm": 8.027936935424805, + "learning_rate": 4.4734057934703095e-07, + "loss": 0.2461, + "step": 2103 + }, + { + "epoch": 2.0708661417322833, + "grad_norm": 5.40668249130249, + "learning_rate": 4.4657853342826435e-07, + "loss": 0.1769, + "step": 2104 + }, + { + "epoch": 2.0718503937007875, + "grad_norm": 7.17467737197876, + "learning_rate": 4.458166179000848e-07, + "loss": 0.1894, + "step": 2105 + }, + { + "epoch": 2.072834645669291, + "grad_norm": 5.848901271820068, + "learning_rate": 4.4505483456590413e-07, + "loss": 0.2176, + "step": 2106 + }, + { + "epoch": 2.0738188976377954, + "grad_norm": 9.032877922058105, + "learning_rate": 4.442931852288219e-07, + "loss": 0.361, + "step": 2107 + }, + { + "epoch": 2.074803149606299, + "grad_norm": 8.919492721557617, + "learning_rate": 4.435316716916202e-07, + "loss": 0.2532, + "step": 2108 + }, + { + "epoch": 2.0757874015748032, + "grad_norm": 9.545595169067383, + "learning_rate": 4.4277029575675983e-07, + "loss": 0.6373, + "step": 2109 + }, + { + "epoch": 2.076771653543307, + "grad_norm": 12.292963981628418, + "learning_rate": 4.4200905922637575e-07, + "loss": 0.8166, + "step": 2110 + }, + { + "epoch": 2.077755905511811, + "grad_norm": 8.164896965026855, + "learning_rate": 4.4124796390227313e-07, + "loss": 0.2628, + "step": 2111 + }, + { + "epoch": 2.078740157480315, + "grad_norm": 8.746564865112305, + "learning_rate": 4.4048701158592277e-07, + "loss": 0.4562, + "step": 2112 + }, + { + "epoch": 2.079724409448819, + "grad_norm": 9.169198989868164, + "learning_rate": 4.3972620407845714e-07, + "loss": 0.3518, + "step": 2113 + }, + { + "epoch": 2.0807086614173227, + "grad_norm": 10.58525562286377, + "learning_rate": 4.3896554318066556e-07, + "loss": 0.377, + "step": 2114 + }, + { + "epoch": 2.081692913385827, + "grad_norm": 16.02490997314453, + "learning_rate": 4.3820503069299077e-07, + "loss": 0.6541, + "step": 2115 + }, + { + "epoch": 2.0826771653543306, + "grad_norm": 7.947490692138672, + "learning_rate": 4.3744466841552416e-07, + "loss": 0.2178, + "step": 2116 + }, + { + "epoch": 2.0836614173228347, + "grad_norm": 21.175268173217773, + "learning_rate": 4.3668445814800137e-07, + "loss": 0.8861, + "step": 2117 + }, + { + "epoch": 2.0846456692913384, + "grad_norm": 9.052186012268066, + "learning_rate": 4.359244016897984e-07, + "loss": 0.5029, + "step": 2118 + }, + { + "epoch": 2.0856299212598426, + "grad_norm": 10.664506912231445, + "learning_rate": 4.3516450083992714e-07, + "loss": 0.4783, + "step": 2119 + }, + { + "epoch": 2.0866141732283463, + "grad_norm": 5.958987712860107, + "learning_rate": 4.344047573970313e-07, + "loss": 0.1502, + "step": 2120 + }, + { + "epoch": 2.0875984251968505, + "grad_norm": 7.741562843322754, + "learning_rate": 4.336451731593817e-07, + "loss": 0.2403, + "step": 2121 + }, + { + "epoch": 2.088582677165354, + "grad_norm": 6.969387054443359, + "learning_rate": 4.328857499248727e-07, + "loss": 0.2598, + "step": 2122 + }, + { + "epoch": 2.0895669291338583, + "grad_norm": 2.9592931270599365, + "learning_rate": 4.321264894910173e-07, + "loss": 0.0499, + "step": 2123 + }, + { + "epoch": 2.090551181102362, + "grad_norm": 8.456254959106445, + "learning_rate": 4.3136739365494334e-07, + "loss": 0.2388, + "step": 2124 + }, + { + "epoch": 2.091535433070866, + "grad_norm": 6.181542873382568, + "learning_rate": 4.306084642133889e-07, + "loss": 0.196, + "step": 2125 + }, + { + "epoch": 2.09251968503937, + "grad_norm": 9.54503059387207, + "learning_rate": 4.2984970296269856e-07, + "loss": 0.4343, + "step": 2126 + }, + { + "epoch": 2.093503937007874, + "grad_norm": 15.962824821472168, + "learning_rate": 4.290911116988181e-07, + "loss": 0.6956, + "step": 2127 + }, + { + "epoch": 2.094488188976378, + "grad_norm": 15.452770233154297, + "learning_rate": 4.2833269221729153e-07, + "loss": 0.6575, + "step": 2128 + }, + { + "epoch": 2.095472440944882, + "grad_norm": 6.801177978515625, + "learning_rate": 4.275744463132564e-07, + "loss": 0.2129, + "step": 2129 + }, + { + "epoch": 2.0964566929133857, + "grad_norm": 11.077638626098633, + "learning_rate": 4.268163757814388e-07, + "loss": 0.3383, + "step": 2130 + }, + { + "epoch": 2.09744094488189, + "grad_norm": 8.228487968444824, + "learning_rate": 4.2605848241615035e-07, + "loss": 0.3318, + "step": 2131 + }, + { + "epoch": 2.0984251968503935, + "grad_norm": 13.941864967346191, + "learning_rate": 4.2530076801128275e-07, + "loss": 0.7162, + "step": 2132 + }, + { + "epoch": 2.0994094488188977, + "grad_norm": 10.613702774047852, + "learning_rate": 4.2454323436030457e-07, + "loss": 0.4118, + "step": 2133 + }, + { + "epoch": 2.1003937007874014, + "grad_norm": 7.851850986480713, + "learning_rate": 4.2378588325625627e-07, + "loss": 0.2831, + "step": 2134 + }, + { + "epoch": 2.1013779527559056, + "grad_norm": 16.458694458007812, + "learning_rate": 4.2302871649174646e-07, + "loss": 1.014, + "step": 2135 + } + ], + "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 +}